EPA/600/R-12/043F I September 2013 I www.epa.gov
United States
Environmental Protection
Agency
Comprehensive Environmental Assessment
Applied to:
Multiwalled Carbon Nanotube Flame-Retardant
Coatings in Upholstery Textiles—
A Case Study Presenting Priority
Research Gaps for Future Risk Assessments
United States Environmental Protection Agency; Office of Research and Development
National Center for Environmental Assessment
-------
&EPA
United States
Environmental Protection
Agency
EPA/600/R-12/043F
Comprehensive Environmental Assessment
Applied to:
Multiwalled Carbon Nanotube Flame-Retardant
Coatings in Upholstery Textiles—
A Case Study Presenting Priority Research Gaps
for Future Risk Assessments
September 2013
U.S. Environmental Protection Agency
Washington, DC
-------
Disclaimer
This document has been reviewed in accordance with U.S. Environmental Protection Agency
policy and approved for publication. Mention of trade names or commercial products does not
constitute endorsement or recommendation for use.
-------
Table of Contents
Table of Contents
List of Tables
List of Figures
List of Text Boxes
List of Additional Information Highlight Boxes
Key Terms
Abbreviations and Acronyms
Authors, Contributors, and Reviewers
iii
viii
xii
xiii
xiv
XV
xvi
xxi
Preface xxiii
Executive Summary xxvi
Chapter 1. Introduction to this Document 1-1
1.1. Background 1-1
1.1.1. Introduction to Comprehensive Environmental Assessment 1-2
1.1.2. Purpose of this Document 1-6
1.1.3. How the CEA Framework and Process Were Applied 1-7
1.2. Introduction to Flame Retardants in Textiles 1-16
1.2.1. Standards for Textiles 1-17
1.2.2. Flame-Retardant Materials as Solutions to Flammability 1-19
1.3. DecaBDE and MWCNTs in Flame-Retardant Textiles 1-24
1.3.1. Introduction to DecaBDE 1-24
1.3.2. Introduction to MWCNTs 1-28
1.3.3. MWCNTs as Alternative Flame-Retardant Materials in Upholstery Textiles 1-32
Chapter 2. Product Life Cycle 2-1
2.1. Feedstocks 2-6
2.2. Manufacturing 2-6
2.2.1. Research and Development 2-6
2.2.2. Material Synthesis 2-8
2.2.3. Material Processing 2-13
2.2.4. Product Manufacturing 2-17
-------
2.3. Storage and Distribution 2-23
2.4. Use 2-23
2.4.1. Life-Cycle Processes 2-24
2.4.2. Potential Releases during the Use Stage 2-25
2.5. Reuse, Recycling, and End of Life 2-27
2.5.1. Reuse and Recycling 2-29
2.5.2. Incineration 2-30
2.5.3. Land-Filling 2-32
2.5.4. Wastewater Treatment Plants 2-33
Chapters. Transport, Transformation, and Fate 3-1
3.1. Physicochemical Factors Influencing Transport, Transformation, and Fate 3-2
3.2. Transport, Transformation, and Fate in Air 3-6
3.3. Transport, Transformation, and Fate in Water and Sediment 3-8
3.3.1. Surface Water and Sediment (Inland and Coastal) 3-8
3.3.2. Ground Water 3-11
3.3.3. Wastewater 3-12
3.4. Transport, Transformation, and Fate in Soil 3-13
3.5. Multimedia Models to Predict Environmental Fate and Transport 3-14
Chapter 4. Exposure-Dose 4-1
4.1. Detection, Measurement, and Characterization 4-3
4.1.1. Dose and Exposure Metrics 4-6
4.1.2. Concentrations in Environmental Media and Indoor Environments 4-9
4.2. Human Exposure and Kinetics Leading to Dose 4-16
4.2.1. Occupational Exposure Pathway Scenarios 4-17
4.2.2. Consumer Exposure Pathway Scenarios 4-23
4.2.3. General Public Exposure Pathway Scenarios through Environmental Media 4-28
4.2.4. Highly Exposed Populations 4-28
4.2.5. Exposure Reference Values and Recommendations 4-29
4.2.6. Toxicokinetics, Dose, and Body Burden 4-32
4.3. Ecological Exposure and Kinetics Leading to Dose 4-34
4.3.1. Factors Impacting Ecological Exposure 4-34
4.3.2. ADME in Ecological Receptors 4-34
4.3.3. Exposure Pathways in Aquatic Systems 4-35
4.3.4. Exposure Pathways in Terrestrial Systems 4-36
4.4. Aggregate Exposures 4-37
4.5. Cumulative Exposures 4-38
Chapters. Potential Human Health, Ecological, and Other Impacts 5-1
5.1. Human Health Effects 5-2
-------
5.1.1. Quantitative Toxicity Assessment 5-7
5.1.2. Systemic Toxicity 5-8
5.1.3. Pulmonary Toxicity 5-10
5.1.4. Eye Irritation 5-12
5.1.5. Skin Irritation 5-12
5.1.6. Reproductive Effects 5-12
5.1.7. Developmental Effects 5-13
5.1.8. Immune System Effects 5-13
5.1.9. In Vitro Data 5-14
5.1.10. Genotoxicity/Mutagenicity 5-16
5.1.11. Carcinogenicity 5-17
5.1.12. Susceptible Populations 5-18
5.2. Ecological Effects 5-19
5.2.1. Aquatic Receptors 5-20
5.2.2. Terrestrial Receptors 5-26
5.3. Other Impacts 5-26
5.3.1. Environmental Justice 5-28
5.3.2. Energy Demand and Natural Resource Depletion 5-28
5.3.3. Climate Change 5-33
5.3.4. Economics 5-33
Chapter 6. Identifying and Prioritizing Research Needs to Support Risk Assessment and
Risk Management 6-1
6.1. Context for Identifying and Prioritizing Research 6-1
6.2. Identification and Prioritization of Research Needs 6-3
6.3. Research Priorities Identified through Collective Judgment in this Application of CEA 6-7
6.3.1. Product Life Cycle 6-8
6.3.2. Environmental Transport, Transformation, & Fate 6-19
6.3.3. Exposure Route and Dose (Kinetics) 6-27
6.3.4. Impacts 6-35
6.4. Moving From "Assessment" to "Management" in the CEA Process 6-46
References R-1
Appendix A. Case Study Candidate Evaluation and Selection Process A-1
A.1. Background A-1
A.2. Candidate Identification Process A-2
A.3. Nanomaterial Application Candidates A-5
A.4. Summary of Factors Affecting Suitability A-5
A.5. U.S. EPA Program Involvement in Final Selection A-7
Appendix A References A-7
-------
Appendix B. Analytical Methods for Detecting, Measuring, and Characterizing BDE-209 and
Multiwalled Carbon Nanotubes B-1
B.1. Measuring and Characterizing PBDEs and MWCNTs B-1
B.2. Summary Tables B-2
Appendix B References B-7
Appendix C. Efficacy of Various Methods of Carbon Nanotube Purification C-1
Appendix C References C-4
Appendix D. Study Summaries on the Transport, Transformation, and Fate of BDE-209 and
MWCNTs in Environmental Systems D-1
Appendix D References D-9
Appendix E. Environmental Contaminant Concentrations E-1
E.1. Concentrations in Environmental Media E-1
E.2. Concentrations in Biota E-17
Appendix E References E-21
Appendix F. Toxicological and Ecological Effects F-1
F.1. Toxicological Effects F-2
F.2. Ecological Effects F-31
Appendix F References F-53
Appendix G. Unprioritized Areas of the CEA Framework for MWCNTs G-1
G.1. Introduction to this Appendix G-1
G.2. Product Life Cycle G-1
G.3. Transport, Transformation, and Fate G-6
G.4. Exposure-Dose G-7
G.5. Potential Human Health, Ecological, and Other Impacts G-12
Appendix G References G-17
Appendix H. Compilation of CEA Framework Data for DecaBDE H-1
H.1. Introduction to DecaBDE H-1
H.2. Product Life Cycle H-4
H.3. Transport, Transformation, and Fate H-18
H.4. Exposure-Dose H-28
H.5. Potential Human Health, Ecological, and Other Impacts H-60
-------
Appendix H References H-75
Appendix I. External Review Draft Comments and Agency Responses 1-1
1.1. Background 1-1
1.2. Responses to Comments I-8
1.3. Full Comment Excerpts 1-31
Appendix I References 1-71
Appendix J. Knowledge Map Pilot: Environmental Transport, Transformation, and Fate J-1
J.1. Knowledge Maps J-1
J.2. Transport, Transformation, and Fate Knowledge Maps: Overview J-1
J.3. Physicochemical Properties Maps J-2
J.4. Transport Map J-5
J.5. Transformation Map J-8
J.6. Future Applications of Knowledge Maps in CEA J-11
Appendix J References J-13
Appendix K. Peer Review Draft Comments and Agency Responses K-1
K.1. Background K-1
K.2. Responses to Comments K-6
K.3. Full Comment Excerpts K-31
Appendix K References K-66
VII
-------
List of Tables
Table 1-1. Existing state regulatory initiatives for decaBDE 1-9
Table 1-2. Common textile fibers, and degrees of flammability 1-17
Table 1-3. U.S. and international fire regulations for upholstery textiles 1-18
Table 1-4. Durability classifications of flame-retardant finishes 1-20
Table 1-5. Flame retardants summarized by chemical class, method of application, and mechanism of
flame-retardant action 1-23
Table 1-6. Major PBDE congeners 1-25
Table 1-7. Commercial formulations of PBDEs used as flame retardants 1-26
Table 1-8. Physical properties and chemical identity of decaBDE 1-27
Table 1-9. Physical properties and chemical identity of MWCNTs 1-30
Table 1-10. Overview of decaBDE and MWCNTs for flame-retardant textile application 1-33
Table 1-11. General qualitative comparisons of performance criteria for decaBDE and MWCNTs 1-34
Table 1-12. Flame test performance of decaBDE and MWCNTs 1-36
Table 1-13. Physicochemical properties of MWCNTs related to flame-retardant performance 1-37
Table 2-1. Summary of common CNT synthesis methods 2-9
Table 2-2. Current scale and projected growth in the CNT industry 2-10
Table 2-3. Potential release scenarios during material synthesis 2-13
Table 2-4. Examples of functionalization of MWCNTs 2-16
Table 2-5. Potential release scenarios during material processing of MWCNTs 2-17
Table 2-6. Potential release scenarios during product manufacturing 2-22
Table 2-7. Potential release scenarios during product use 2-27
Table 2-8. Potential release scenarios during reuse and recycling 2-30
Table 2-9. Potential release scenarios during incineration 2-31
Table 2-10. Potential release scenarios during land-filling 2-32
Table 2-11. Potential release scenarios during wastewater treatment 2-34
Table 3 -1. Summary of physicochemical properties that affect partitioning and fate of nanomaterials
such as multiwalled carbon nanotubes (MWCNTs) 3-3
Table 3-2. Predicted environmental concentrations of CNTs using fate and transport modeling 3-16
Table 4-1. Established inhalation reference values and recommendations applicable to MWCNTs 4-31
Table 5-1. Summary of effects observed after dermal, oral, and inhalation exposure to
decaBDE and MWCNTs 5-4
Table 5-2. Effects of decaBDE and MWCNTs on aquatic receptors: Algae, plants, and invertebrates 5-21
Table 5-3. Effects of decaBDE and MWCNTs on aquatic receptors: aquatic vertebrates 5-25
Table 5-4. Estimated minimum energy requirements and process rates for synthesis of CNTs 5-30
VIII
-------
Table 5-5. Environmental assessment of production of 1 kilogram (kg) of carbon nanofibers 5-31
Table 5-6. Environmental assessment of production of one SWCNT polymer mesh 5-32
Table 6-1. Examples of risk management decisions 6-5
Table 6-2. Example questions for problem formulation 6-6
Table 6-3. "Influential Factors" options for all areas 6-8
Table 6-4. Research identified by RTI workshop participants: MWCNT material processing 6-13
Table 6-5. Research identified by RTI workshop participants: MWCNT product manufacturing 6-15
Table 6-6. Research identified by RTI workshop participants: MWCNT product use 6-17
Table 6-7. Research identified by RTI workshop participants: MWCNT disposal / recycling 6-19
Table 6-8. Research identified by RTI workshop participants: MWCNT mobility & persistence in air 6-22
Table 6-9. Research identified by RTI workshop participants: MWCNT mobility & persistence
inwastewater 6-24
Table 6-10. Research identified by RTI workshop participants: MWCNT persistence in sediment 6-26
Table 6-11. Research identified by RTI workshop participants: MWCNT occupational exposure
via inhalation 6-30
Table 6-12. Research identified by RTI workshop participants: Human absorption, metabolism,
and excretion of MWCNTs 6-35
Table 6-13. Research identified by RTI workshop participants: Noncancer human health impacts
of MWCNTs 6-38
Table 6-14. Research identified by RTI workshop participants: Societal impacts of MWCNTs 6-45
Table A-l. Nanomaterial application candidates and non-nanoenabled products for comparison in a
case study A-5
Table A-2. Factors for consideration in selecting a candidate for case study A-6
Table B-l. Analytical techniques for detecting, measuring, and characterizing PBDEs B-3
Table B-2. Analytical techniques for detecting, measuring, and characterizing MWCNTs B-4
Table C-l. Purification methods for carbon nanotubes C-2
Table D-l. Relevant studies of transformation (debromination) of BDE-209 D-l
Table D-2. Relevant studies of MWCNTs in aqueous media D-4
Table D-3. Relevant studies of BDE-209 in soils and plants D-7
Table D-4. Relevant studies of carbon nanotubes (CNTs) in soils D-8
Table E-l. BDE-209 concentrations in building dust E-l
Table E-2. BDE-209 air concentrations in outdoor and indoor air E-3
Table E-3. BDE-209 concentrations in aquatic systems E-6
Table E-4. BDE-209 concentrations in sewage effluent and sludge E-9
Table E-5. BDE-209 concentration data in soil E-ll
Table E-6. Proxy data for estimating MWCNT concentrations in occupational air E-12
Table E-7. Measured concentrations of PBDEs in biota E-17
Table E-8. Mean concentration of PBDEs in media/biota in an aquatic ecosystem E-21
Table F-l. Select toxicokinetic studies for decaBDE F-2
-------
Table F-2. Select toxicokinetic studies for MWCNTs F-7
Table F-3. Select dermal and ocular studies for decaBDE F-9
Table F-4. Select dermal and ocular studies for MWCNTs F-ll
Table F-5. Select pulmonary exposure studies for decaBDE F-13
Table F-6. Select pulmonary exposure studies for MWCNTs F-14
Table F-7. Select oral and intragastric studies for decaBDE F-18
Table F-8. Select intubation and injection studies for MWCNTs F-21
Table F-9. Select genotoxicity, mutagenicity, and other in vitro studies for MWCNTs F-23
Table F-10. Select carcinogenicity studies for decaBDE F-28
Table F-ll. Select carcinogenicity studies for MWCNTs F-29
Table F-12. Effects of exposure to PBDEs in aquatic invertebrates F-31
Table F-13. Effects of exposure to PBDEs in fish and frogs F-35
Table F-14. Effects of exposure to MWCNTs in algae, macrophytes, and aquatic macroinvertebrates F-39
Table F-15. Effects of exposure to MWCNTs via water on Ceriodaphnia dubia F-41
Table F-16. Effects of exposure to MWCNTs on zebrafish and medaka embryos F-42
Table F-17. Immune responses in rainbow trout (Oncorhynchus mykiss) head kidney cells following
MWCNT exposure F-43
Table F-18. Effects of exposure to decaBDE in soil microbes, terrestrial invertebrates, and plants F-44
Table F-19. Effects of exposure to MWCNTs in bacteria F-47
Table F-20. Effects of exposure to MWCNTs on plants F-51
Table G-l. Percent yields for agglomerated growth of SWCNTs and MWCNTs using various
synthesis methods, processing temperatures, and catalysts G-2
Table G-2. Potential release scenarios during storage and distribution G-5
Table G-3. Effects of decaBDE and MWCNTs on soil microbes and invertebrates G-13
Table G-4. Effects of decaBDE and MWCNTs on plants G-15
Table H-l. MajorPBDE congeners H-2
Table H-2. Commercial formulations of PBDEs used as flame retardants H-3
Table H-3. Physical properties and chemical identity of decaBDE H-4
Table H-4. Summary of physicochemical properties that affect partitioning and fate of BDE-209 H-21
Table H-5. Median tissue concentration ranges (in ng/gram liquid weight) for three polybrominated
diphenyl ether congeners in humans H-52
Table H-6. Biomagnification factors of select PBDE congeners in an aquatic ecosystem H-57
Table 1-1. Expert affiliations and area of expertise 1-3
Table 1-2. Public commenters' affiliations and area of expertise 1-4
Table 1-3. Agency response categories 1-5
Table 1-4. Relevant to the general case study or multiple sections of the case study 1-8
Table 1-5. Relevant to Chapter 1 (including Preface and Executive Summary) 1-14
Table 1-6. Relevant to Chapter 2 1-17
-------
Table 1-7. Relevant to Chapter 3 1-19
Table 1-8. Relevant to Chapter 4 1-21
Table 1-9. Relevant to Chapter 5 1-24
Table 1-10. Relevant to Chapter 6 and Appendices 1-27
Table 1-11. Comments that required no action 1-28
Table 1-12. Comments received in response to the expert charge questions 1-31
Table 1-13. Free-form comments received from experts 1-52
Table 1-14. Comments received in response to the public comment period 1-62
Table K-l. Expert affiliations and area of expertise K-3
Table K-2. Agency response categories K-4
Table K-3. Revisions relevant to the general case study or multiple sections of the case study K-6
Table K-4. Revisions relevant to Chapter 1 (including Preface and Executive Summary) K-9
Table K-5. Revisions relevant to Chapter 2 K-12
Table K-6. Revisions relevant to Chapters K-13
Table K-7. Revisions relevant to Chapter 4 K-14
Table K-8. Revisions relevant to Chapters K-15
Table K-9. Revisions relevant to Chapter 6 K-18
Table K-l0. Revisions relevant to Appendices K-19
Table K-ll. Comments that required no action K-21
Table K-12. Comments for which no action was requested K-29
Table K-13. Comments received in response to the expert charge questions K-31
-------
List of Figures
Figure 1-1. Comprehensive environmental assessment framework 1-3
Figure 1-2. Comprehensive environmental assessment process 1-5
Figure 1-3. Detailed CEA framework used for the collective judgment prioritization process 1-14
Figure 1-4. Durability of additive flame retardants 1-21
Figure 1-5. The combustion process 1-22
Figure 2-1. Life-cycle stages, potential release scenarios, and forms of release for decaBDE and MWCNTs
in flame-retardant coatings applied to upholstery textiles 2-2
Figure 2-2. Variations in MWCNT formulations and functionalization along the product life-cycle 2-5
Figure 3-1. Variability in MWCNT chemistry and implications in terms of life cycle, exposure, and risk.0 3-5
Figure 6-1. Iterative communication flow in the CEA framework 6-1
Figure A-l. Candidate identification process A-4
Figure J-l. Physicochemical properties map for decabromodiphenyl ether J-4
Figure J-2. Physicochemical properties map for multiwalled carbon nanotubes J-5
Figure J-3. Transport map for decabromodiphenyl ether J-7
Figure J-4. Transport map for multiwalled carbon nanotubes J-8
Figure J-5. Transformation map for decabromodiphenyl ether J-10
Figure J-6. Transformation map for multiwalled carbon nanotubes J-ll
XII
-------
List of Text Boxes
Text Box 1-1. Physicochemical Properties of Multiwalled Carbon Nanotubes (MWCNTs) Affect Their
Release, Behavior in the Environment, and Interaction with Biota 1-29
Text Box 4-1. Detecting, Measuring, and Characterizing MWCNTs 4-4
Text Box 4-2. Specific Physicochemical Properties of MWCNTs Shown to Influence Exposure, Kinetics,
and Dose 4-5
Text Box 5-1. Specific Physicochemical Properties of Multiwalled Carbon Nanotubes Shown to Influence
Toxicity 5-3
Text Box H.3-1. BDE-209 Undergoes Biotic and Abiotic Debromination H-20
Text Box H.4-1. Detecting, Measuring, and Characterizing PBDEs H-31
XIII
-------
List of Additional Information
Highlight Boxes
1. Factors influencing selection of flame retardants 1-19
2. MWCNTs are not widely used as flame-retardants in textiles 1-31
3. MWCNTs are likely used in combination with other chemicals in flame-retardant applications 1-35
4. MWCNT release from various product matrices informs predictions of release from textiles 2-21
5. Impact of MWCNT release into wastewater treatment plants 2-33
6. Transformation throughout the product life cycle 3-2
7. Properties of the MWCNT formulation impact environmental release and transformation 3-4
8. Multimedia modeling of MWCNT environmental transport 3-17
9. Challenges related to MWCNT toxicokinetics 4-6
10. Weaknesses of current analytical techniques 4-10
11. MWCNT dermal absorption 4-18
12. Predicted dominant exposure routes for MWCNTs 4-24
13. MWCNT fibers resemble asbestos fibers 5-6
14. Applying traditional in vivo inhalation study design models to MWCNTs 5-11
15. Developing NexGen-style models for MWCNT toxicity testing 5-15
16. Mechanisms of toxicity for cancer and noncancer impacts of MWCNTs 5-18
17. Toxicity to benthic invertebrates 5-23
Gl. Uptake and Absorption in Aquatic Foodwebs G-10
G2. Toxicity to Terrestrial Invertebrates G-14
XIV
-------
Key Terms
Term
Upholstery textiles
Flame retardant
DecaBDE
BDE-209
Multiwalled carbon
nanotube
Meaning
Fabric or cloth material that is fixed to furniture (e.g., chairs), mattresses or
transportation industry components (e.g., seating, roof linings)
A chemical or other manufactured material that has the ability to inhibit the
combustion process and increase the resistance of textile products to degradation
associated with fire and flame.
The only polybrominated diphenyl ether (PBDE) that is fully brominated (i.e., all
available hydrogen atoms in the diphenyl rings have been substituted with bromine
atoms) and therefore exists as a single isomer (see BDE-209); commercial
formulations of decaBDE may contain trace amounts of lower brominated
congeners (e.g., nona- or octaBDEs) as impurities
The single isomer of deca-substituted BDE (see decaBDE) named as the final of
the 209 possible congeners of PBDE (many lower brominated PBDEs [e.g.,
heptaBDE] are composed of many different congeners based on the exact position
of the bromine atoms in the chemical conformation)
Hollow nanoscale (i.e., with one or more dimensions in the range of approximately
1-100 nm) tubes composed of multiple concentrically nested graphene sheets
XV
-------
Abbreviations and Acronyms
Term
A549
ACC
ACGIH
ADME
Ag
Al
APA
APPI
ASTM
ATSDR
B6C3F-,
BAF(s)
BAF/BCF(s)
BALB/c
BCF(s)
BDE
BDE-209
BDL
BfFV
BFR(s)
BHI
BMF
Br
BSA
BSEF
Meaning
human lung epithelial cells
American Chemistry Council
American Conference of
Government Industrial
Hygienists
absorption, distribution,
metabolism, and excretion
silver
aluminum
alkaline phosphatase (enzyme)
atmospheric pressure
photoionization
American Society for Testing
and Materials
Agency for Toxic Substances
and Disease Registry
mouse strain
bioaccumulation factor(s)
bioaccumulation or
bioconcentration factor(s)
mouse strain
bioconcentration factor(s)
brominated diphenyl ether
single isomer of decaBDE
(congener)
below detection limit
Bundesministerium fur
Frauenangelegenheiten und
Verbraucherschutz (Germany's
Federal ministry for women's
interests and consumer
protection)
brominated flame retardant(s)
brain heart infusion broth
biomagnification factor
Bromine
bovine serum albumin
Bromine Science and
Environmental Forum
Term
BSI
bw
C
°C
14C
C57BL/6
CATS
Ca
Cal/EPA
CalRecyle
Ce
CEA
CFR
CINAHL
Cl
CLF
cm3
CNF(s)
CNQ
CNT(s)
Co
CO2
COOH-MWCNT
CPC
CPTC
CVD
decaBDE
Dl
diBDE
Meaning
British Standards Institution
body weight
Carbon
degrees in Celsius
radiolabeled carbon
mouse strain
California (Bureau of Home
Furnishings and Thermal
Insulation) Technical Bulletin
Calcium
California EPA
California Department of
Resources Recycling and
Recovery
Cesium
Comprehensive Environmental
Assessment
Code of Federal Regulations
Cumulative Index to Nursing and
Allied Health Literature
Chlorine
Conservation Law Foundation
cubic centimeters
carbon nanofiber(s)
could not quantify
carbon nanotube(s)
Cobalt
carbon dioxide
carboxylated MWCNT
condensation particle counters
Consumer Product Testing
Company
chemical vapor deposition
decabromodiphenyl ether
deionized (water)
dibromodiphenyl ether
XVI
-------
Term Meaning
DIN Deutsches Institut fur Normung
(Germany)
DLS dynamic light scattering
DMA deoxyribonucleic acid
DOD U.S. Department of Defense
doi digital object identifier
DOM dissolved organic matter
DWCNT(s) double-walled carbon
nanotube(s)
E Element in CEA Framework
ECso median effective concentration
ECB European Chemicals Bureau
ECHA European Chemicals Agency
ECNI electron capture negative
ionization
ECNI-MS electron capture negative
ionization-mass spectrometry
EEA electron capture negative
ionization
EEB European Environmental Bureau
EEC European Economic Community
EI99 Eco Indicator 1999 (method)
ENM(s) engineered nanoscale
material(s)
EPA U.S. Environmental Protection
Agency
EROD ethoxyresorunfin-O-deethylase
enzyme
E-RRF Element / Risk-Relevance-
Factor (Pair in CEA Framework)
EU European Union
F344 rat strain
FBCVD fluidized bed chemical vapor
deposition
Fe iron
FLE forelimb emergence
FLM fluorescence microscopy
FMVSS Federal Motor Vehicle Safety
Standards
FTIR Fourier transform infrared
spectroscopy
g, mg, ug, ng, gram, milligram, microgram,
pg; kg nanogram, picogram; kilogram
Term
GC
GC/HR TOF MS
GD
GHG
GLP
GLRI
GPIIb/llla
gpt
GSI
GSRI
GWERD
heptaBDE
hexaBDE
HHPC-6
HiPCO®
HMVEC
hpf
HR
HRMS
HSDB
IARC
ICF
ICL
ICP-MS
ICR
ID
IF
INEL
IO
Meaning
gas chromatograph(y)
gas chromatography/high
resolution time-of-flight mass
spectroscopy
gestation day
green house gas(es)
Good Laboratory Practices
Great Lakes Research Institute
glycoprotein integrin receptor
guanine phosphoribosyl-
transferase
gonadosomatic index
Gulf South Research Institute
Ground Water and Ecosystems
Restoration Division of NRMRL
heptabromodiphenyl ether
hexabromodiphenyl ether
Hand-held airborne particle
counter
a high pressure carbon
monoxide synthesis process
human microvascular
endothelial cells
hours post fertilization
high resolution
High resolution mass
spectroscopy
Hazardous Substances Data
Bank
International Agency for
Research on Cancer
ICF International, Inc. (formerly
Inner City Fund; ICF-Kaiser; ICF
Consulting)
Israel Chemical Ltd.
inductively coupled plasma
mass spectrometry
mouse strain
inner diameter
Influential Factor
indicative (human) no-effect
level
Immediate Office
XVII
-------
Term
IPCS
IPEN
IRDC
IRIS
ISO
kg
KM
L
L, ml_
LC
LC/MS-MS
LC5o
LCA
LD5o
LDH
LOAEL
LOEC
LOEL
LOI
LRT
LSRI
u
M, mm, urn
m2
m3
Mg
MN
Mo
Meaning
International Programme on
Chemical Safety
International POPs Elimination
Network's Nanotechnology
Working Group
International Research and
Development Corporation
(U.S. EPA) Integrated Risk
Information System
International Organization for
Standardization
kilogram
Knowledge Map
Soil organic carbon/water
partition coefficient
octanol/water partition
coefficient
length
Liter, milliliter
liquid chromatography
liquid chromatography tandem-
coupled mass spectroscopy
median lethal concentration
Life-cycle Assessment
median lethal dose
layered double hydroxide
lowest-observed-adverse-effect
level
lowest observed effect
concentration
lowest observed effect level
limiting oxygen index
long-range atmospheric
transport
Life Science Research Israel
mu symbol, denoting 'micro' or
10'6
meter, millimeter, micrometer
square meters
cubic meters
magnesium
micronucleus
molybdenum
Term
MRL
mRNA
MS
MWCNT(s)
MWCNT-OH
MWCNT-NH2
MWCNT-NH3+
MWCNT-COOH
MWNT
n
NA
Na
NaCI
Nanomaterials
nCeo
NCC
NCCT
NCEA
NCSL
ND
NERL
NF
NFPA
NH2-MWCNT
NH3+-MWCNT
NHDF
Ni
NICNAS
NIEHS
Meaning
minimal risk level
messenger RNA
mass spectrometer
multiwalled carbon nanotube(s)
hydroxylated MWCNT
amine-functionalized MWCNT
ammonium-functionalized
MWCNT(s)
carboxylated MWCNT
multiwalled nanotube(s)
sample number
not applicable
sodium
sodium chloride (salt)
nanoscale materials
nanofullerene
nanocrystalline cellulose
National Center for
Computational Toxicology (U.S.
EPA /ORD)
National Center for
Environmental Assessment
(U.S. EPA/ORD)
National Conference of State
Legislators
Not determined; No data
identified
National Environmental
Research Laboratory (U.S. EPA
/ORD)
not functionalized
National Fire Protection
Association
amine-functionalized MWCNT
ammonium-functionalized
MWCNT
normal human dermal fibroblast
cells
nickel
(Australia) National Industrial
Chemicals Notification and
Assessment
National Institute of
Environmental Science (NIH)
XVIII
-------
Term
NIH
NIOSH
NLM
NMAM
NMRI
NOAEL
NOE
NOEC
NOEL
NOM
nonaBDE
NR
NRC
NRMRL
NRMRL/GWERD
NSTC
NTP
O
OAF
OCSPP
octaBDE
OD
OECD
OEL(s)
OH-MWCNT
OPC
OPP
ORD
ORISE
Meaning
National Institute of Health
National Institute for
Occupational Safety and Health
National Library of Medicine
(NIH)
NIOSH Manual of Analytical
Methods
mouse strain
no-observed-adverse-effect
level
no observed effect
no observed effect concentration
no-observed-effect level
natural organic matter
nonabromodiphenyl ether
not reported
National Research Council
National Risk Management
Research Laboratory (U.S. EPA,
ORD)
Ground Water and Ecosystems
Restoration Division of NRMRL
National Science and
Technology Council
National Toxicology Program
(NIEHS/NIH)
oxygen
overall assessment factor
Office of Chemical Safety and
Pollution Prevention (U.S. EPA)
octabromodiphenyl ether
outer diameter; optical density
Organisation for Economic Co-
operation and Development
occupational exposure limit(s)
hydroxylated MWCNT
optical particle counters
Office of Pesticide Programs
(U.S. EPA)
Office of Research and
Development (U.S. EPA)
Oak Ridge Institute for Science
and Education
Term Meaning
OSCP Office of Science Coordination
and Policy (in OCSPP; U.S.
EPA)
OST Office of Science and
Technology (in Office of Water;
U.S. EPA)
OVA ovalbumin
OW Office of Water (U.S. EPA)
P purity
p p-value, estimated probability,
level of statistical significance
p53+/" mouse strain with impaired gene
stability
PAH(s) polycyclic aromatic
hydrocarbon(s)
PBDD polybrominated dibenzo-p-dioxin
PBDE polybrominated diphenyl ether
PBDF polybrominated dibenzofuran
PBS phosphate buffered saline
(solution)
PBZ personal breathing zone
PCB(s) polychlorinated biphenyl(s)
PEC(s) predicted environmental
concentration(s)
PEI polyethyleneimine
PEL permissible exposure limit
pentaBDE pentabromodiphenyl ether
pH scale of acidity and alkalinity
PINFA Phosphorus, Inorganic and
Nitrogen Flame Retardants
Association
PMMA polymethyl methacrylate
PMN premanufacturing notice
PND postnatal day
POPs persistent organic compounds
ppb parts per billion
ppm parts per million
PPM Physicochemical Properties Map
ppt parts per trillion
R&D Research and Development
RAW 264.7 murine macrophage cell line
XIX
-------
Term Meaning
REACH Registration, Evaluation,
Authorisation and Restrictions of
Chemicals (EU)
REL recommended exposure limit
RfC reference concentration
RfD reference dose
RNA ribonucleic acid
RRF Risk Relevance Factor in CEA
Framework
RTI Research Triangle Institute
International
$ cost in U.S. dollars
S sulfur
SA surface area
SAFENANO Europe's Center of Excellence
on Nanotechnology Hazard and
Risk, based at the Institute of
Occupational Medicine
SD rat strain; standard deviation
SEM scanning electron microscopy;
standard error of mean
Si silicone
SiO2 silicone dioxide
SNUR (TSCA) Significant New Use
Rule
SOC(s) synthetic organic compound(s)
SWCNT(s) single-walled carbon
nanotube(s)
T3 free tri-iodothyronine
T4 free thyroxine
TB technical bulletin
TB total body
TEM transmission electron
microscopy
tetraBDE tetrabromodiphenyl ether
TfM Transformation Map
TGA thermogravimetric analysis
Ti titanium
TiO2 titanium dioxide
TLV(s) threshold limit value(s)
TOC total organic carbon
TOF time of flight
Term Meaning
TpM Transport Map
triBDE tribrominated diphenyl ether
TSCA Toxic Substances Control Act
TWA time weighted average
U.K. United Kingdom
U.S EPA U.S. Environmental Protection
Agency
U.S. United States of America
UV ultraviolet
V volume
V79 cells lung fibroblast cell line from
Chinese hamster lung tissue
w/w weight-for-weight measurement
WHO World Health Organization
wt weight
XPS x-ray photoelectron
spectroscopy
Zn zinc
xx
-------
Authors, Contributors, and Reviewers
EPA Project Leaders/Authors
Christina Powers, ORD/NCEA
Geniece Lehmann, ORD/NCEA
J. Michael Davis, ORD/NCEA
EPA Contributors
Jeff Gift, ORD/NCEA
Meredith Lassiter, ORD/NCEA
Patricia Gillespie, ORD/NCEA
Connie Meacham, ORD/NCEA
Ray Antonelli, ORISE
Emma McConnell, ORISE
Kyle Painter, ORISE
Internal Reviewers
Ambika Bathija, OW/OST
Dermont Bouchard, ORD/NERL
Lyle Burgoon, ORD/NCEA
Diana Eignor, OW/OST
Karen Hamernik, OCSPP,OSCP
Ross Highsmith, ORD/NERL
Keith Houck, ORD/NCCT
Connie Meacham, ORD/NCEA
Dave Meyer, ORD/NRMRL
Venkata Krishna K. Upadhyayula, ORD/NRMRL
Reeder Sams, III, ORD/NCEA
Philip Sayre, ORD/IO
John Vandenberg, ORD/NCEA
Debra Walsh, ORD/NCEA
XXI
-------
ICF International
This draft was prepared by ICF International under EPA Contract No. EP-C-09-009, with technical
direction by the National Center for Environmental Assessment.
Project Manager: David Burch
Primary Authors: Kristen Fedak, Adeline Harris, Pamela Hartman, Jocelyn Hospital, Bryan
Luukinen, Katherine Sullivan, Amalia Turner
Contributors: Michelle Cawley, Whitney Kihlstrom, Codi Sharp, Courtney Skuce, Audrey
Turley, Nicole Vetter, Ron White
Technical Editor: Penelope Kellar
Consultant Reviewer: Jo Anne Shatkin (CLF Ventures)
Interagency Reviewers
T. Matthew Cho, Navy and Marine Corps Public Health Center
Candis M. Hunter, Agency for Toxic Substances and Disease Registry
Treye Thomas, Consumer Product Safety Commission
Public Commenters
Samantha Dozier, People for the Ethical Treatment of Animals
Maureen Egan, Littleton/Englewood Wastewater Treatment Plant, Englewood, CO
Marcia L. Hardy, Albemarle Corporation
Fred Klaessig, Pennsylvania Bio Nano Systems, LLC
Xuyang Liu, National Research Council Associate, US EPA/NRMRL/GWERD
Letter Peer Reviewers
Karin Asheberger, European Commission - Joint Research Centre
Jaclyn Canas-Carrell, Texas Tech University
P. Lee Ferguson, Duke University
Fadri Gottschalk, Environmental, technical and scientific services (ETSS)
Stig I. Olsen, Technical University of Denmark
XXII
-------
Preface
This document is part of continuing efforts by the U.S. Environmental Protection Agency (EPA)
to understand the potential health and environmental impacts of nanotechnology, including associated
scientific issues and information. Accordingly it is consistent with recommendations in the U.S. EPA
Nanotechnology White Paper (2007) and U.S. EPA Nanomaterial Research Strategy (2009). Although no
national or international consensus definition for nanomaterials exists, a current working definition is a
material having at least one dimension on the order of 1 to 100 nm (NSTC. 20 lib).
Previous EPA documents similar to this one focused on (1) nanoscale titanium dioxide used in
drinking water treatment and in topical sunscreen (U.S. EPA. 2010d) and (2) nanoscale silver used in
disinfectant spray (U.S. EPA. 2010e). The nanomaterials considered in this case study are multiwalled
carbon nanotubes (MWCNTs), as incorporated into flame-retardant coatings for upholstery textiles. This
document neither represents a risk assessment, nor intends to serve as a basis for near-term risk
management decisions on possible uses of MWCNTs. Rather, it is a case study presenting research
priorities that, if pursued, could inform future assessments and subsequent risk management decisions for
MWCNTs in this application. The research priorities presented in this case study were identified through
a structured decision process that engaged experts in identifying and prioritizing MWCNT research gaps
after reviewing the External Review Draft (U.S. EPA, 2012b). In revising the External Review Draft
(U.S. EPA. 2012b) to create this final case study, EPA streamlined the document to clearly reflect the
research priorities that experts identified, as well as respond to public comments and recommendations
from expert reviewers.
Similar to previous case studies, this MWCNT case study is based on the comprehensive
environmental assessment (CEA) approach, which consists of both a framework and a process.
The organization of this document reflects the CEA framework, the principal elements of which are
described in Chapter 1 of this document and largely represented in Chapter 2 through Chapters.
This document also contains information about a traditional (i.e., "non-nanoenabled") product,
decabromodiphenyl ether flame-retardant upholstery coatings, to compare against available data for
MWCNT flame-retardant upholstery coatings (i.e., the "nanoenabled" product). The primary purpose of
including a comparative element in the case study (U.S. EPA. 2012b) was to provide a more robust
database as a foundation from which to identify data gaps related to the nanoenabled product. Because it
has served its primary purpose (i.e., to help identify data gaps for MWCNTs), most of this comparative
XXIII
-------
information has been moved to an appendix; how the information about the traditional product might
inform research planning for MWCNTs is included in succinct textboxes in appropriate areas of the
document.
Following a general introduction to the materials and selected application in this case study in
Chapter 1; Chapter 2 discusses stages of the product life cycle for the nanoenabled product and highlights
which stages in the product life cycle present opportunities for releases to the environment. Chapter 3
then provides information on the transport, transformation, and fate processes affecting the behavior of
MWCNTs, by-products, and transformation products in environmental compartments. Chapter 4
characterizes exposure, uptake, and dose for MWCNTs, by-products, and transformation products for
different human populations and ecological receptors, after which Chapter 5 describes the human health,
ecological, and other (i.e., social, economic, and environmental resource) impacts related to those
exposures.
Collectively, these chapters represent the assembly of information across the vertical spectrum of
the CEA framework (Figure 1-1); as outlined in Chapter 1, however, this step is merely the first in the
CEA process (Figure 1-2). Next, a group of expert stakeholders representing a variety of technical
backgrounds and sector perspectives used the External Review Draft (U.S. EPA. 2012b) in a collective
judgment process to rate areas of the CEA framework in terms of (1) importance for future risk
assessments of MWCNTs, and (2) confidence in the data to support risk management decisions.
Concurrently, the case study was posted for public comment. EPA then revised the case study to address
public comments and expert recommendations, including revisions to highlight (1) the outcomes of the
collective judgment step, and (2) key sections of the case study based on expert comments. EPA expects
these revisions to facilitate research that supports the compilation of new information in the CEA
framework for future iterations of the approach. Chapter 2 through Chapter 5 now reflect the areas of the
framework expert stakeholders identified as important research for future risk assessment and
management decisions. Areas identified as lower priorities are discussed in appendices of the document.
In addition, new information identified through public or reviewer comments on the External Review
Draft (U.S. EPA. 2012b) is highlighted throughout this document in text boxes.
As described in more detail in Chapter 6. the identification of priority areas by a diverse group of
expert stakeholders is a key part of connecting research, risk assessment, and risk management for
MWCNTs (areas that were not identified as priorities are discussed in Appendix G). The next critical step
in this process is to engage the broader scientific community in implementing research in areas identified
as important to consider in future MWCNT risk assessments, but which lack sufficient data to support
risk management decisions for MWCNTs. Doing so will support the subsequent steps of the CEA
XXIV
-------
process, which involve a continued, iterative communication flow across the continuum of research, risk
assessment, and risk management.
xxv
-------
Executive Summary
Chapter 1: Introduction to this Document
Background
As part of an ongoing effort to identify research needs and data gaps in assessing the broad
environmental implications of nanomaterials, this case study focuses on a specific nanomaterial in a
particular application: multiwalled carbon nanotubes (MWCNTs) in flame-retardant coatings applied to
upholstery textiles. The selection of this specific nanomaterial and particular application was made with
input from representatives across the U.S. Environmental Protection Agency (EPA) and was based in part
on its relevance to EPA programmatic interests and the similarity in the potential for release and exposure
over the product life cycle compared to conventional flame-retardant materials that are being phased out
of use.
Like previous case studies of nanoscale titanium dioxide and nanoscale silver, this case study is
built on the comprehensive environmental assessment (CEA) approach, which is both a framework and a
process. The CEA framework (Figure 1-1) starts with the inception of a material and encompasses
environmental fate, exposure-dose, and impacts associated with that material. The framework also
considers differences in environmental media and the physical, chemical, biological, and social conditions
in which the material occurs. Here, the framework is used to organize information about MWCNTs in the
case study systematically. This information does not represent a completed or even preliminary risk
assessment; rather, it is intended to inform research planning. The External Review Draft of the document
provided a basis for identifying and prioritizing data gaps and research needs for MWCNTs and other
nanomaterial assessments as part of the CEA process (Figure 1-2). Specifically, a group of expert
stakeholders representing diverse technical (e.g., human health effects, ecological effects, material
characterization) and sector (e.g., industry, academia, government) perspectives engaged in a structured,
collective judgment workshop process such that each individual had equal input in identifying research
priorities. To facilitate the identification of key research gaps related to assessing MWCNTs in this
application, the External Review Draft case study provided a comparative perspective by also presenting
information on a traditional flame retardant, decabromodiphenyl ether (decaBDE). The prioritized
research gaps that emerged are intended to inform decision-makers in the EPA and the broader scientific
community in developing research agendas that support future risk assessment and risk management
goals for MWCNTs. Strategic research planning focused on supporting future risk assessment and
XXVI
-------
management is responsive to guidance from the National Research Council and others to place more
emphasis on problem formulation and the identification of broader potential impacts of materials under
evaluation in the risk assessment [see Chapter 6; QJ.S.GAO. 2013; NRC. 2011. 2009)1. Moreover,
identifying research priorities for future risk assessment and management can help to ensure that the
effects of resources allocated to research are maximized (i.e., focused on the most pressing data gaps).
The Priority Research Areas that were identified for MWCNTs are the primary focus of this
revised document, with information on decaBDE that supported identifying the priorities in the previous
draft (U.S. EPA. 2012b) presented primarily in Appendix H. Background information on decaBDE,
however, is provided in Chapter 1 to give the necessary context for reviewing the research priorities
identified for MWCNTs. In addition, text boxes with the title "DecaBDE Can Inform MWCNT
Assessment" are provided throughout the document to succinctly note how information on the
conventional material might inform research planning for MWCNTs. Information on MWCNTs that
pertains to areas that were not prioritized for research is now located in Appendix G. Input on the
External Review Draft case study from public and expert stakeholders also is highlighted throughout the
document and is recorded in Appendix I.
Given the purpose of the document, this case study does not purport to be a comprehensive
literature review; rather, available sources were incorporated specifically to support prioritizing and
subsequently planning research, as described above. As this case study involves an emerging technology,
some information, particularly regarding background or general concepts, was occasionally obtained from
non-peer-reviewed sources to supplement the published literature available. The most recent literature
search for this case study was conducted in May 2012 using specific criteria relevant to MWCNTs in
flame retardants. Additional targeted literature searches were conducted on November 13, 2012, using
search terms specific to topic areas identified in public and expert comments. Specific references
suggested by experts through an independent Letter Peer Review were then incorporated into the
document in May 2013.
Introduction to decaBDE and MWCNT flame-retardant textiles
Production and importation of decaBDE are currently being phased out in the United States as a
result of voluntary commitments within the industry and EPA actions in response to concerns regarding
potential human health and ecological impacts. As a result, a range of alternative flame-retardant
technologies, including nanotechnologies, is being evaluated as potential replacements for this
extensively used material. This document presents information on a potential alternative flame-retardant
technology, MWCNTs, in the context of the research priorities that could support future assessments of
this product. The primary purpose of this document is to inform research planning efforts for MWCNTs
XXVII
-------
across the scientific community. In doing so, the document supports a key objective of the CEA approach;
to link research, risk assessment, and risk management efforts iteratively.
In developing research plans for MWCNTs, understanding the considerations involved in their
potential use in flame-retardant textiles is informative. Many manufacturers incorporated flame-retardant
materials into textiles to comply with state, federal, and industry fire-safety standards (i.e., certain flame
test performance criteria that must be met). Once applied, flame retardants act to inhibit the combustion
process through a variety of physical or chemical means (e.g., producing inert gases that dilute the oxygen
supply available to the flame, producing protective char barriers) (Section 1.2).
Both decaBDE and MWCNTs can be mixed with binding agents and applied as coatings to
increase the flame resistance of upholstery textiles. In this application, the two materials are both referred
to as barrier technologies because they exhibit similar mechanisms of flame-retardant action: decaBDE
forms a protective char barrier and MWCNTs form a network floccules layer (i.e., network of loosely
bound MWCNT bundles). The similarity in potential applications for decaBDE and MWCNTs was a
primary reason for including the comparison of the two materials as flame-retardant coatings in
upholstery textiles in the External Review Draft of the case study, as the comparison informed the
identification of data gaps related to assessing possible risks and benefits associated with MWCNTs.
Moreover, the comparison of these materials highlighted MWCNT- and nano-specific factors that might
influence future research directions for nanomaterials and nanoenabled products. For example, unlike
with decaBDE, the physicochemical properties of MWCNTs are often intentionally altered during
synthesis; thus MWCNTs are not a single material with a defined set of characteristics, but rather a
variety of materials—often present as mixtures—with vastly different physicochemical characteristics.
Such variation in the physicochemical characteristics of MWCNTs presents challenges in describing the
releases, behavior, and effects of exposure to MWCNTs as a class of materials (Section 1.3). Importantly,
MWCNTs likely will be used in combination with other flame-retardant materials to provide sufficient
efficacy for the standards noted above (Section 1.2). In addition to introducing greater variability in
MWCNT behavior, exposure, and effects, the use of MWCNTs in combination with other materials raises
important implications for the potential use of MWCNTs in this application (Additional Information
Highlight Box 3V
Chapter 2: Product Life Cycle
Little information is available on the commercial production and use of MWCNT flame-retardant
coatings, as few commercial-scale products currently exist. The manufacturing stages of MWCNT flame-
retardant textile coatings (Section 2.2). along with the use (Section 2.4) and reuse/recycling/end-of life
XXVIII
-------
stages (Section 2.5). were identified as Priority Research Areas for upholstery textiles treated with
MWCNT flame retardants.
Based on the available data, releases of MWCNTs to the environment are expected to occur
throughout the life cycle of MWCNT flame-retardant upholstery textiles. The projected increase in
MWCNT production likely will result in increased environmental releases of MWCNTs from flame-
retardant textiles or other MWCNT products. Most MWCNTs released in the manufacturing stages are
anticipated to be in the free or bundled form (Footnote J_3 in Chapter 2 explains this terminology), while
most releases later in the life cycle are anticipated to be in the polymer or textile matrix-bound form.
Upholstery textile products are expected to have a long lifespan and likely will be disposed of in
municipal landfills or incineration facilities.
Air and water releases of MWCNTs during manufacturing are expected to occur based on the
activities performed in manufacturing stages of the product life cycle. Although release is particularly
likely during mixing, handling, and equipment cleaning, releases are expected to be fairly well controlled
when proper ventilation and environmental controls are in place. Air releases of MWCNTs have been
measured during material synthesis but no data are available regarding release to water during
manufacturing. Additionally, MWCNTs typically require purification and functionalization, which also
could result in releases due to chemical and physical processing methods (Section 2.2). Activities like
textile and furniture processing might take place outside of closed systems and could result in
environmental releases of MWCNTs. Abrasion, washing, unintended use, and accidental exposure to high
heat or fire during the use stage could result in releases of MWCNTs (Sections 2.4 and 2.5).
No data are currently available on the volume or potential release of MWCNTs in the use stage of
the flame-retardant upholstery textile product life cycle. Based on decaBDE data, however, the potential
for release during this stage of the product life cycle could be relatively high. Similarly, no data currently
exist on the volume or potential release of MWCNTs in upholstery textiles at end of life. Nevertheless,
the physical and chemical processes (e.g., shredding, milling, chemical treatment) used to recycle textiles
also could lead to releases of MWCNTs. Air releases from land-filling of MWCNT flame-retardant
upholstery also could occur due to mixing and compacting. In addition, release in leachate from landfills
is possible if the product or polymer matrix degrades. Although incineration at end of life presents the
potential for airborne release of MWCNTs and by-products, preliminary experimental data suggest that
MWCNTs will not be released to the environment when exposed to the sufficiently high temperatures of
municipal incinerators (Sections 2.4 and 2.5). Incomplete incineration during other stages of the product
life cycle, however, is one of the most likely airborne release scenarios for CNT textile coatings.
XXIX
-------
Chapter 3: Transport, Transformation, and Fate
Although MWCNTs are incorporated into polymer matrices after the flame-retardant production
stage, little information exists that describes the environmental behavior of these polymer matrices. As a
result, Chapter 3 focuses on the transport, transformation, and fate of MWCNTs and not the polymer
matrices in which they are incorporated. Environmental transport, transformation, and fate of MWCNTs
in air, wastewater, and sediment were identified as Priority Research Areas. The environmental behavior
of MWCNTs is dictated by their physical and chemical properties—surface area, surface chemistry,
morphology (shape), solubility (Footnote Jj5 in Chapter 2 explains this terminology), presence or absence
of functionalization and surface coatings (e.g., engineered coatings or natural organic matter), and
hydrophobicity. The nanostructured morphology, small size, and high surface area-to-volume ratio of
MWCNTs can enhance chemical reactivity and propensity of MWCNTs to form bundles; however, single
MWCNTs, as compared to bundles, will differ in their behavior in the environment (Section 3.1).
Recent literature regarding the behavior of airborne MWCNTs is extremely limited, and
dominant fate, transport, and transformation processes for MWCNTs in indoor and outdoor air are
unknown. In aqueous media, such as wastewater, the hydrophobicity, and van der Waals interactions of
pure MWCNTs suggest they will bundle together or sorb to particles and be removed during the sewage
treatment process, or settle out into sediment in receiving water bodies. Physicochemical characteristics
of the MWCNTs and environmental conditions, however, can alter this behavior. For example, the
presence of dissolved organic matter has been shown to debundle MWCNTs causing to them to remain in
solution. Similarly, surface coatings can affect the sorption behavior of MWCNTs in these systems and
influence their mobility, dispersion, and bioavailability in environmental media (Sections 3.2.
13, and 14).
Scientists have demonstrated the use of simple, deterministic models and more complex
probabilistic models to simulate movement of carbon nanotubes through, and predict environmental
concentrations in, environmental compartments. Differences in modeling approaches, model scale, and
model input data make comparisons across models for predicting environmental concentrations of CNTs
difficult. Nevertheless, a recent life-cycle-based analysis predicted the impacts of CNT synthesis in
aquatic systems by using output data from a single model of environmental concentrations (Section 3.5).
Chapter 4: Exposure-Dose
Several analytical challenges for nanomaterials combined with the lack of historical use of
MWCNTs in consumer products have so far prevented MWCNTs from being detected in ambient media,
which could inform decisions related to potential exposures in human and ecological populations (Section
4.1). Human exposures to MWCNTs released throughout the flame-retardant textile coating life cycle are
xxx
-------
expected to differ for workers, consumers, and the general public. Based on available information,
occupational and consumer exposures were identified as Priority Research Areas in the CEA collective
judgment workshop process for MWCNTs. Workers can be exposed to various forms of MWCNTs (e.g.,
adsorbed to dust, as part of the polymer or textile matrix) via inhalation and ingestion of, and dermal
contact with, these substances during manufacturing, storage and distribution, and end-of-life activities. In
the workplace, the inhalation route is expected to represent the greatest potential for exposures, and
MWCNTs are expected to be in the particulate phase when inhaled. Little is reported about consumer
exposures to MWCNTs, especially those incorporated into flame-retardant textiles. Yet, based on
activities expected to occur during use, repurposing, or reuse of upholstered products, consumers might
be exposed to MWCNTs during each of these points in the product life cycle. The MWCNTs released
from finished products also are expected to be in particulate form, generally adsorbed to dust or
constituents of the polymer or textile matrix. The primary route of exposure (i.e., inhalation, ingestion, or
dermal) for consumers is unknown.
Developing exposure standards, guidelines, or recommendations for MWCNTs is complicated by
the heterogeneity in MWCNT configurations and challenges measuring MWCNTs in occupational or
environmental settings. The National Institute for Occupational Safety and Health (NIOSH) established a
recommended exposure limit for elemental carbon, and several other occupational exposure limits have
been proposed by industry and international agencies (Section 4.2.5). In general, MWCNTs appear to be
biopersistent and might remain in the lung for several months after inhalation. Limited studies show that,
after oral exposure, most ingested MWCNTs are eliminated with no detectable metabolism or transport
into the blood. Distribution to the liver, lungs, and spleen, however, has been reported following
intravenous exposure (Section 4.2). Notably, the bioavailability, and thus dose, of MWCNTs likely will
be based on whether they are bound in a textile matrix, bundled, or free (Footnote H in Chapter 2
explains this terminology).
No evidence is currently available to determine whether portions of the population might
experience higher exposure levels to MWCNTs compared to the general population; however, the activity
of children and workers might increase total exposure levels of MWCNTs relative to the general
population (Section 4.2).
Exposure and dose in ecological populations were not deemed Priority Research Areas for
MWCNTs in the CEA collective judgment workshop process, and thus information on these areas is now
located in Appendix G and Appendix H for MWCNTs and decaBDE, respectively. The anticipated
increase in MWCNT production (Section 2.2.2) along with increases in potential applications of the
material could lead to an increase in the number and type of exposures experienced by workers,
XXXI
-------
consumers, and ecological populations. These changes are expected to increase aggregate and cumulative
exposures to different formulations of MWCNTs, transformation products, and by-products.
Chapters: Potential Human Health, Ecological, and Other Impacts
Expert stakeholders participating in the CEA collective judgment workshop process identified
human health impacts as a Priority Research Area for MWCNTs. Toxicology studies conducted on
animals are the only identified data on human health impacts of MWCNTs because no human data on
effects of MWCNT exposure exist. All routes of exposure were examined in this case study because each
route (dermal, inhalation, and oral) offers potential for human exposures (Section 5.1). Toxicological
effects from MWCNT exposure in animal models have been evaluated predominantly after dermal and
inhalation exposures, rather than after oral exposure. Effects were generally localized and included
irritation (skin and ocular), sensitization (respiratory), and inflammation (respiratory). In addition,
MWCNTs altered immunological function after exposure via inhalation for 14 days or via a single
intranasal injection. The carcinogenicity of MWCNTs following inhalation exposure has not been
investigated; however, several studies using methods such as instillation indicate that some types of
MWCNTs behave like asbestos, potentially inducing mesotheliomas, and might be more toxic than
asbestos (Section 5.1).
Expert stakeholders identified impacts in aquatic, but not terrestrial, biota as a Priority Research
Area. Considerations for the ecological impact of MWCNTs include the toxicity toward different species,
types of effects, and potential for bioaccumulation and biomagnification. More than 20 studies have
investigated the effects of MWCNTs on aquatic species or aquatic systems; those studies indicate low
acute toxicity potential, with the effect level varying based on size and functionalization properties of the
MWCNTs. Chronic studies show that MWCNTs can elicit immune responses and produce developmental
impacts (Section 5.2).
Other impacts, including economic or societal effects and alterations in environmental resources,
were identified as a Priority Research Area by expert stakeholders. No empirical data exist relating
MWCNTs to other impacts, but the background literature on processes involved in manufacturing similar
materials (e.g., carbon nanofibers, single-walled carbon nanotubes) provides some basis for concern
regarding potential impacts of MWCNTs on energy demand, resource depletion, climate change, and
economics. These related studies provide a plausible foundation for suggesting that MWCNT
manufacturing can be an energy-intensive process potentially causing the depletion of nonrenewable
natural resources like fossil fuels, and that the synthesis of MWCNTs can result in emissions of other
compounds causing adverse environmental effects (e.g., volatile organic compounds; Section 5.3).
XXXII
-------
Chapter 6: Identifying and Prioritizing Research Needs to Support Risk Assessment
and Risk Management
The External Review Draft of this document served as the foundation from which expert
stakeholders participating in the CEA process could identify key data gaps and determine research
priorities. The information presented in this revised document focuses on those priorities to inform
ongoing research planning for nanotechnology in the general scientific community and at EPA. Results of
these research efforts could subsequently support future assessments and risk management efforts for
MWCNTs or other nanomaterials. Future evaluations of nanoenabled products, such as MWCNT in
flame-retardant textile coatings, could involve the consideration of risk-related trade-offs, for example,
thyroid health effects versus pulmonary health effects and environmental justice considerations versus
energy costs. This document therefore strives to inform research planning efforts that would support
conducting risk assessments that can inform risk management decisions about such trade-offs.
The research priorities discussed in the case study were identified by a group of diverse expert
stakeholders independently rating areas of the CEA framework based on two factors:
• Importance: how important an area is to consider in risk assessments of MWCNTs;
• Confidence: the availability and utility of current data to support risk management decisions
for MWCNTs.
For those areas they identified as "Important" to consider in future risk assessments of MWCNTs,
stakeholders were asked to rate the relative importance and confidence in data related to the relationship
of the area with risk factors that might be considered in risk assessment or risk management efforts for the
area. Areas that experts most commonly identified as being of high importance to risk assessment, and
were not confident in the data to support risk management decisions, are considered high priorities for
research. In contrast, areas rated as of high importance and for which experts had confidence in the data
might be of interest to decision-makers for evaluating risk management options for MWCNTs.
Most of the prioritized CEA framework areas were considered research priorities, including
release rates across the product life cycle; persistence and bioavailability in air, wastewater, or sediment,
and inhalation exposure in workers and consumers. Other areas identified as high Priority Research Areas
include absorption, metabolism, and excretion in humans, as well as impacts on human health, aquatic
biota, and other considerations (i.e., economic, societal, environmental resources). For a subset of these
areas, experts identified potential risk management decisions in the context of an example risk scenario
for that area and noted the type of assessment(s) that could inform those decisions. Specific research
questions to support such assessments also were identified, along with estimates of the financial and time
resources to carry out the research. Risk management decisions generally centered on choosing
XXXIII
-------
appropriate control technologies or personal protective equipment, modifications to MWCNTs (e.g.,
reducing residence time in air by increasing aggregation potential), or limits on production and use of the
materials. Assessments to inform these and other types of risk management efforts included human health
risk assessments, cost benefit analyses, and life cycle assessments. Research areas to support such
assessments can be grouped into five general themes: (1) the influence of MWCNT characteristics on
release from the product matrix; (2) the influence of MWCNT characteristics and the product matrix both
on environmental transport and transformation, and on absorption across biological barriers (e.g.,
gastrointestinal tract); (3) development of analytical methods or tools to detect MWCNTs in complex
matrices and measure exposures; (4) human health impacts of MWCNTs and co-factors (e.g., solvents)
after acute and chronic exposures; and (5) improving public engagement in and understanding of potential
benefits and risks of nanotechnology.
The connection of specific questions within Priority Research Areas to the assessments and risk
management decisions they would subsequently support demonstrates the focus within the CEA approach
on linking communication across the continuum of research, risk assessment, and risk management.
Moreover, the specific questions are intended to provide more concrete support for strategic research
planning that informs future decision-making about MWCNTs.
xxxiv
-------
This page intentionally left blank.
-------
Chapter 1. Introduction to this Document
1.1. Background
Nanoscale materials (nanomaterials) have been described as having at least one dimension
approximately 1-100 nm (NSTC. 20 lib). Although this definition is not universally accepted and
continues to evolve, 100 nm is typically used as an upper bound, and this working definition is used as the
size standard in this case study. Engineered nanomaterials are intentionally synthesized at the nanoscale,
rather than being produced as incidental by-products of combustion or a natural process such as erosion,
to exploit the unique or novel properties that can arise from their small size. Like all emerging
technologies, engineered nanomaterials offer the potential for both benefits and risks, the assessments of
which depend on the availability of relevant data and other information.
This document is part of an endeavor to identify what is known and, more importantly, what is
not known that could be of value in assessing the broad environmental implications of nanomaterials. As
a case study, this document presents information about a specific nanomaterial in a particular application.
It does not represent completed or even preliminary assessments; rather, the External Review Draft
provided a starting point in a process to identify and prioritize possible research directions to support
future risk assessments of nanomaterials. The prioritized research gaps that emerged are the focus of this
revised case study document. As with previous case studies, these research priorities are intended to
inform decision-makers in the U.S. Environmental Protection Agency (EPA) as well as the broader
scientific community in developing research agendas that support future risk assessment and risk
management goals. Such information is expected to be considered in the context of the particular focus,
budgetary constraints, ongoing research, and other considerations of any organization; however, as
discussed below, by using a holistic framework paired with input from a diverse group of expert
stakeholders, the priorities identified through the comprehensive environmental assessment (CEA)
approach employed in this case study can provide a unique perspective on research directions to support
future risk management goals.
The focus of this document is a specific application of a selected nanomaterial: the use of
engineered multiwalled carbon nanotubes (MWCNTs) as an agent in flame-retardant coatings on
1-1
-------
upholstery textiles.1 EPA previously completed case studies of nanoscale titanium dioxide used for
drinking water treatment and for topical sunscreen (U.S. EPA. 2010d) and nanoscale silver used as an
agent in disinfectant spray products (U.S. EPA. 2010e). Unlike previous case studies, this case study
incorporates information about a traditional (i.e., "non-nanoenabled") product, decabromodiphenyl ether
(decaBDE [a traditional flame-retardant]), against which the "nanoenabled" product (MWCNT flame-
retardant coating) is compared (see Section 1.1.3.1).
Part of the rationale for compiling a series of nanomaterial case studies is that the properties
associated with different nanomaterials are often complex and vary considerably within, between, or
among specific types of nanomaterial groups, nanomaterials in general, and different applications of
nanomaterials. As a result, applying generalities could result in overlooking key characteristics or
information. Focusing on a single example of an application of MWCNTs is not intended to represent all
ways in which this nanomaterial could be used or all issues that other applications might raise. However,
by considering this single application of MWCNTs, research directions can be identified that would
support future assessments of this material. Such information might be used more broadly as an analog
for other applications of MWCNTs or types of nanomaterials. For instance, research investigating the
influence of MWCNT surface treatment on potential release from flame-retardant textile coatings and
subsequent behavior in environmental media can also inform efforts to understand the influence of
surface treatment on the environmental behavior of MWCNTs in other applications.
1.1.1. Introduction to Comprehensive Environmental Assessment
This case study of MWCNTs, like the previous case studies of nanoscale titanium dioxide (U.S.
EPA. 2010d) and nanoscale silver (U.S. EPA. 2010e). is built on the CEA approach, which consists of
both a framework and a process, the principal elements of which are illustrated in Figure 1-1 and Figure
1-2. respectively. The uppermost box of Figure 1-1 lists typical stages of a product life cycle: research
and development (R&D), feedstock processing, manufacturing, storage and distribution, use, and disposal
(which would include reuse or recycling, if applicable).
Although not considered a life-cycle stage in typical life cycle analyses, R&D is included in
business models of product value chains. Because of the relatively large portion of resources and
information associated with this stage for emerging materials, such as nanomaterials, R&D is considered
Although flame retardants are commonly used in both upholstery textiles and furniture foam, this case study
focuses only on information relevant to the use of flame retardants as coatings on upholstery textiles. The extent to
which the information presented might be relevant to the use of flame retardants in furniture foam is not addressed.
1-2
-------
in the CEA framework. The actual volume of the material used in R&D is likely small but could represent
a significant proportion of the total market, particularly during product development, given the limited
number of full-scale commercial manufacturing efforts early in the life cycle for emerging materials. For
these materials, processes in R&D lend insight to full-scale commercial processes and might constitute an
important source of material release into the environment, as well as occupational exposures. Other CEA
applications focusing on traditional or more mature materials or technologies might provide minimal or
no information on the R&D portion of the product lifecycle, given that R&D would be less active.
Regardless of the material of focus, releases to the environment associated with any stage of the
product life cycle lead to what is depicted in the second box in Figure 1-1. which refers to transport,
transformation, and fate processes. These processes can result in the spatial distribution of both primary
and secondary contaminants in the environment. The chains of events represented in the CEA framework
occur within multiple environmental media (air, water, sediment, soil) and under various conditions
(physical, chemical, biological,
social). Also of note are the single
arrows connecting one facet of the
CEA framework to the next, which
represent a variety of linkages,
transfers, and feedback loops. For
example, the transfer of material from
one organism to another through the
food chain would represent a
bidirectional exchange between
transport, transformation, and fate and
exposure, uptake, and dose.
The third box in Figure 1-1.
exposure-dose, goes beyond
characterizing the occurrence of
contaminants in the environment, as
exposure refers to actual contact
/physical
Transport/Transformation/Fate
Primaryand Secondary Substances
Exposure-Dose
Humans Other Biota Abiotic Resources'?
Health
Impacts
Ecological
Aesthelk, Climate, Energy, Ethical Legal Social, Re:
Source: (U.S. EPA. 2011a)
Figure 1-1. Comprehensive environmental assessment
framework.
Note: The CEA framework is used to organize complex information systematically
in evaluations of the environmental implications of selected chemicals, products,
or technologies (i.e., materials). The framework starts with the inception of a
material and encompasses the environmental fate, exposure-dose, and impacts.
Notably, the sequence of events is not always linear when, for example, transfers
occur between media or via the food web. In addition, a variety of factors
. influence each event, including differences in environmental media and the
between a contaminant and a receptor, physical, chemical, biological, and social conditions in which the material event
occurs. Details on these influential factors are thus included throughout the
whether living or nonliving. Living framework when possible.
1-3
-------
organisms include humans and other biota.2 Examples of nonliving, or abiotic, receptors include features
of the natural landscape, structures such as buildings and statues, and painted surfaces of vehicles and
other objects. Exposure can involve aggregate exposure across routes (e.g., inhalation, ingestion, dermal),
cumulative exposure to multiple contaminants (both primary and secondary), and various spatiotemporal
dimensions (e.g., activity patterns, diurnal and seasonal changes). Dose is the amount of a substance that
enters an organism by crossing a biological barrier or which deposits on an inanimate object.
As part of a chain of cause-effect events, dose links exposure with potential impacts of various
types, as indicated in the last box of Figure 1-1. Human health effects might result when a certain
delivered dose reaches a target cell or organ. In an ecological context, effects might occur when a stressor
reaches a level sufficient to cause an adverse outcome in biotic or abiotic receptors. Impacts encompass
both qualitative hazards and quantitative exposure-response relationships and can extend to aesthetic
(e.g., alterations in visibility, taste, and odor), climate change, energy consumption, resource depletion,
socioeconomic, and other effects. Such effects are considered in the CEA framework, but their ultimate
inclusion would depend on whether the compiled information indicates that such effects could reasonably
be expected to occur. As discussed below, the inclusion of such information in the CEA framework
should influence the selection of the technical experts for the next step of the CEA process.
Not reflected in Figure 1-1 is the role of analytical methods that make detecting, measuring, and
characterizing nanomaterials in the environment and in organisms possible. Characterizing a substance of
interest (e.g., determining its chemical identity, reactivity, purity, and other properties) is fundamental to
the assessment of any material. Thus, if adequate analytical techniques have not yet been developed or
need refinement, methods development must be included in research efforts to inform future assessments.
For simplicity, such information is not included in this high-level view of the CEA framework. For the
purpose of this document, analytical methods for the materials in this case study are presented in detail in
Appendix B.
As previously mentioned, the CEA approach consists of both a framework and a process.
Compiling the information described above into the CEA framework is the first step of the CEA process
(Figure 1-2). Starting with the holistic perspective of the CEA framework facilitates identifying
information pertinent to consider for the material of focus, which in turn supports problem formulation
and scoping for assessment purposes. Next, a collective judgment process is used to evaluate and
prioritize this information. Collective judgment, as applied in the CEA process to date, refers to a formal,
structured procedure enabling a range of participants to be heard individually and to be represented in a
2The term biota is used throughout this document to refer to all living organisms other than humans.
1-4
-------
transparent record of the collectively
reached outcomes. Collective judgment
supports an essential feature of CEA:
the inclusion of diverse technical and
stakeholder perspectives to ensure that
a holistic evaluation is achieved (U.S.
EPA. 201 Oh).
Prioritization is a key objective
in this holistic evaluation within the
CEA process. Depending on one's
objectives and the state of the science
surrounding an issue, CEA can be used
to prioritize (1) information gaps
leading to development of a research
plan that will support future assessment
efforts and (2) risk trade-offs leading to
development of an adaptive risk
management plan. As depicted in
Figure 1-2. these uses of CEA cross
over from conducting assessments into
management efforts after the initial
identification and prioritization of
information. Specifically, this transition
c
Compile Information
in CEA Frameworb
Develop Risk
Management Plan
Monitor, Evaluate
Outcomes
Source: (U.S. EPA. 2011a)
Figure 1-2. Comprehensive environmental assessment
process.
Note: The CEA process involves a series of steps that result in judgments
about the implications of information contained in the CEA framework.
Compiling information in the CEA framework is fundamental for a given
material, but is only a first step in the CEA process. Next, the information in
the framework is evaluated using a collective judgment technique (i.e., a
structured process that allows the participants representing a variety of
technical and stakeholder viewpoints to learn from one another, yet form their
own independent judgments). The result of the collective judgment step is a
prioritized list of risk trade-offs or information gaps that then can be used in
planning research and developing adaptive risk management plans.
The knowledge gained from these research and risk management activities
feeds back in an iterative process of periodic CEA updates.
encompasses the use of prioritized information by research planners and risk managers in their
evaluations, which subsequently inform research and risk management decisions. In either instance, CEA
is meant to be iterative; thus, the results of research, assessments that are carried out with new research
results, and risk management efforts would be used to update the CEA framework after some period of
time determined by those conducting the CEA process. At present, the CEA framework and process are
being applied to help refine research planning for nanomaterials, with particular focus on a specific
nanomaterial application. As the knowledge base grows for nanomaterials, the availability of more
complete information will make the identification and prioritization of risk-risk and risk-benefit trade-offs
feasible, and the path leading to risk management (as shown in Figure 1-2) will be pursued. Such
prioritized risk-related trade-offs would be only one source of information that risk managers could
consider when making judgments about risk management options in the context of relevant legal,
1-5
-------
political, and other considerations. Yet, the use of the holistic CEA framework together with diverse
stakeholder input in the development of such priorities will make them a unique resource that is
responsive to current recommendations to further refine risk assessment and management approaches (see
Chapter 6).
1.1.2. Purpose of this Document
This document represents a revision of the External Review Draft that was used in the collective
judgment step of the CEA process applied to MWCNTs (Figure 1-2). As a revised case study, it provides
a basis for considering the outcomes of the collective judgment process to aid research planning that
supports long-term assessment efforts; it does not, however, purport to present an exhaustive review of
the literature. References included in the document were selected with the specific intention of supporting
research planning for MWCNTs in flame-retardant coatings applied to upholstery textiles. Literature
searches were conducted in November 2011 and January 2012, with a more targeted search conducted in
May 2012, using specific criteria to search the PubMed database, Academic Search Complete,
Environment Complete, and CINAHL (Cumulative Index to Nursing and Allied Health Literature).
Search terms included carbon nanotube*, carbon nanofiber*, CNT*, CNF*, MWNT*, MWCNT*, and
SWCNT*. Additional targeted literature searches were conducted in November 2012, using search terms
specific to topic areas identified by public commenters and expert reviewers. Finally, specific references
suggested by reviewers through an independent Letter Peer Review were incorporated into the document
in May 2013. Given the emerging state of the science surrounding MWCNTs, government reports or
other sources outside of the peer-reviewed literature are used in the case study when little or no peer-
reviewed information was found. As a document to support research planning, this case study is not an
actual risk assessment and does not provide conclusions on potential ecological or human health impacts
related to MWCNTs. As discussed in Chapter 6. the Toxic Substances Control Act (TSCA) is one of the
statutes under which EPA currently considers nanoscale substances. Considerations of nanoscale
substances under TSCA begin with a determination of whether the substance is already included on the
TSCA Chemical Substance Inventory based on whether the substance has the same molecular identity as
a substance listed on the Inventory (U.S. EPA. 2008c). Determinations of whether nanoscale substances
are new or existing substances are currently made on a case-by-case basis (U.S. EPA, 2008c).
A variety of efforts are underway to increase the scientific body of knowledge such that
regulatory decisions through TSCA or other statutes could move beyond a case-by-case approach. These
include several research frameworks applicable to, or exclusively for, nanomaterials and intended to
support future assessments and subsequent risk management of these materials [e.g., (U.S. EPA. 2009).
1-6
-------
(NSTC. 2011 a). (NRC. 2012). and (OECD. 2012)1. These frameworks and the CEA approach share
several common elements (e.g., focus on product life cycle, identifying environmental fate mechanisms
and exposure sources, importance of stakeholder engagement), yet as described above, the CEA approach
incorporates decision-support tools to engage stakeholders beyond those seen in other frameworks to date
(Figure 1-2).
1.1.3. How the CEA Framework and Process Were Applied
This document presents information in the CEA framework for MWCNTs, as potentially used in
flame-retardant coatings applied to upholstery textiles, in the context of the outcomes that emerged from
engaging expert stakeholders in the CEA process for this material. This case study begins with a general
overview of textiles and flame-retardant systems and where MWCNTs fit into that context, as well as
detailed introductory information on decaBDE in the context of textiles and flame-retardant systems
(Chapter 1). Throughout the main body of this document (Chapter 2 through Chapter 5). the focus is on
highlighting what is known and not known about each part of the CEA framework that was identified as a
research priority in the CEA collective judgment step for MWCNTs in flame-retardant coatings. In these
chapters, readers are referred to Appendix H for the detailed information regarding decaBDE at each
stage of the CEA framework. Highlight-level information on decaBDE, however, is presented in select
tables, figures, and text boxes (see Section 1.1.3.1 and Appendix I) to provide a succinct comparison of
MWCNTs to decaBDE in this particular application so that such comparisons might inform MWCNT
research planning. Information on MWCNTs relevant to areas identified as lower priorities for research is
located in Appendix G. Appendices A-F contain additional supporting information (e.g., study summary
tables) for Chapters 1-5. Information was placed in these supporting appendices, rather than the main text
of the document, if it was deemed informative for those interested in details re levant to a particular topic,
but not critical for understanding the key points of the topic. Appendix J summarizes a pilot study to
convey information visually in CEA case studies, while Appendix I and Appendix K summarize
comments and revisions in response to comments from external reviews of the case study document.
Throughout the document, text boxes are used to highlight key concepts or to summarize new information
added to the document based on reviewer feedback; in these boxes, citations are provided when specific
data are discussed, while readers are referred to the main text for more detail on general concepts.
1.1.3.1. Comparison of DecaBDE and MWCNT in the CEA Framework
An important aspect of the CEA approach is the ability to examine the relative risks and benefits
of, for example, different products or different formulation options, to aid in risk management decisions.
1-7
-------
The particular comparison to focus on in an application of CEA would be guided by risk management
objectives. For example, MWCNT flame-retardant coatings applied to upholstery textiles might be
compared to conventional flame-retardant products, a different nanoenabled flame-retardant formulation,
a flame retardant not applied as a coating, or some other variable. Although several different options
could be of interest to risk managers, considering every potential option in the present case study is not
feasible. Therefore, this document focuses solely on a comparison of MWCNTs and a traditional flame
retardant, decaBDE, as they might be used in flame-retardant coatings for upholstery textiles, including
those used in homes and nonresidential areas such as public buildings and automobiles.
As described in detail in Appendix A. several candidate carbon-based nanomaterials and
applications were identified as options for this case study using a systematic approach, and professional
judgment then was applied to narrow down the selection to a single nanomaterial and application. First,
candidate carbon-based nanomaterials were identified through initial strategic literature and Internet
searches, news reports, and basic literature search statistics (e.g., number of total hits, number of hits in
scientific databases). This approach provided an initial indication of overall data availability and research
interest within the nanotechnology and scientific communities for several different broad groups of
carbon-based nanomaterials (e.g., carbon nanotubes, carbon nanofibers, nanocrystalline cellulose). This
group was further narrowed using a more judgment-based approach to evaluating suitability, including
consideration of the available data for multiple applications of each nanomaterial. Finally, five feasible
candidates of unique nanomaterial and application pairs—carbon nanofibers in cement, MWCNTs in
flame-retardant coatings, single-walled carbon nanotubes (SWCNTs) in textiles, nanocrystalline cellulose
in biodegradable packaging, and MWCNTs in rubber tires—were selected based on additional
professional judgment of suitability.
The process for selecting the material-application pair of MWCNT flame-retardant coatings for
upholstery textiles as the subject of a CEA case study involved individuals representing EPA program
offices, regional offices, and Office of Research and Development laboratories and centers. Individuals
were appointed by their organization within EPA to be involved with development of nanomaterial case
study documents. They were encouraged to share information on the five selected candidate carbon-based
nanomaterials and applications with colleagues in their organization and to represent the views of their
organization in voting for their preferences. The two candidates receiving the most votes were MWCNTs
in flame-retardant coatings and composites and SWCNTs in textiles. Rationale for selecting MWCNTs
and SWCNTs in each respective application included: relevance of both materials to Agency programs,
similarity in potential release and exposure over the product life cycle of textiles compared to existing
flame-retardant materials being phased out of use, greater availability of data compared to other candidate
applications, and potential for market expansion of CNTs (see Appendix A). Based on input that
1-8
-------
MWCNTs were of greater interest (i.e., more widely produced than SWCNTs and might contain more
contaminants) and that an application involving textiles would be preferable, a hybrid option was selected
as the topic of this case study: MWCNTs in flame-retardant coatings applied to upholstery textiles. This
selection does not imply that MWCNTs in flame-retardant coatings applied to textiles represents the
carbon-based nanomaterial and application with the largest current market share (see Section 1.3.2). but
rather was based on the selection factors noted above.
DecaBDE was chosen as the traditional flame-retardant product to compare to MWCNTs due to
its extensive use since the 1970s and the robust scientific database available for it and for the family of
brominated flame retardants (BFR) in general. DecaBDE has been used widely in the textile industry to
meet fire safety standards (see Section 1.2.1). In 2001, decaBDE use accounted for 83% of total
polybrominated diphenyl ethers (PBDE) production worldwide (U.S. EPA. 2010b): an estimated 10-20%
of decaBDE use is in the textile industry (Pure Strategies Inc.. 2005). At the end of 2004, both octa- and
pentaBDE were voluntarily withdrawn from the U.S. marketplace due to evidence of environmental
persistence and toxicity, which left decaBDE as the sole PBDE available for use in commercial products
in the United States (U.S. EPA. 201 Ob).
Table 1-1. Existing state regulatory initiatives for decaBDE.
Regulatory Initiative State
Implemented studies to assess environmental and human health impacts of Illinois, Minnesota, Rhode Island
decaBDE to inform regulatory action
Restricted the use or sale of products containing decaBDE Oregon
Prohibited the manufacture, use, or sale of certain products containing Vermont, Maryland, Maine,
decaBDE Washington State
Source: National Conference of State Legislators (2011).
Yet, concern has grown regarding the potential impacts of decaBDE on ecological and human
health. For example, despite previous assumptions that decaBDE is relatively stable and inert in the
environment, recent studies have suggested that it can debrominate, or break down into lower weight
congeners, which have been much more widely studied and are known to be highly toxic [(Environment
Canada. 2010: U.S. EPA. 2010b: Siddiqi etal.. 2003: Rahman et al.. 2001): see Section 3J_ and Text Box
H.3-1]. In response to these concerns regarding potential adverse impacts on human health and the
environment (see Chapter 5). limitations or bans on the use of decaBDE have been imposed recently both
in the United States and abroad (see Section 1.3.1).
1-9
-------
As discussed further in Section 1.3 the comparison of the larger body of information on decaBDE
with the relatively small database for MWCNTs was intended to help pinpoint data gaps relating to this
specific MWCNT product. Although at least one commercial MWCNT flame-retardant product is
available, the use of MWCNTs as flame retardants is relatively new and is not abundant in the market
(see Section 1.3.2). Given the projected decline in decaBDE use, as described above, investigating these
nanoenabled products as a potential emerging alternative is relevant. The use of flame retardants in
textiles is of interest to EPA and also aligns with the needs of other organizations (e.g., Consumer Product
Safety Commission, National Institute of Standards and Technology).
In the External Review Draft of this case study, the comparison between decaBDE and MWCNTs
provided: (1) a more robust database (i.e., that of a traditional product that has been relatively well
characterized) as a reference for identifying data gaps relating to a nanoenabled product; and (2) a context
for identifying key factors and data gaps related to assessing the risk-risk and risk-benefit trade-offs
between a nanoenabled product and a non-nanoenabled product. Although the specific characteristics,
exposure patterns, and effects associated with the use of MWCNTs and decaBDE are expected to differ
substantially, the data needed to inform risk assessment and risk management decision-making are
comparable; thus, the comparative framework was used to help determine whether relevant information
(e.g., dominant exposure pathways, sensitive populations) is available and sufficient to inform future risk
decision-making, and by extension, to identify key MWCNT data gaps that could be pursued. Since key
MWNCT data gaps (i.e., research priorities) were identified by expert stakeholders after reviewing the
External Review Draft, the primary purpose of including decaBDE information in the case study has been
fulfilled. In turn, as outlined in Section 1.1.3, how decaBDE could inform MWCNT research planning is
highlighted in a series of text boxes in this final case study document, with detailed decaBDE information
available in Appendix H.
1.1.3.2. Application of the CEA Process: Identifying MWCNT Research Priorities
The External Review Draft of this document represented the "Compile Information in CEA
Framework" step of the CEA process (Figure 1-2). and thus supported the next step of the process:
evaluating the data in the framework using a collective judgment technique to identify and prioritize
information gaps about MWCNTs. The collective judgment prioritization technique used for this case
study was funded by EPA and conducted independently by an EPA contractor, RTI International. Details
related to the collective judgment method and its outcomes are described in a separate report prepared by
RTI International (RTI International. 2012). A summary of that process is described here with the
outcomes discussed in greater detail in Section 6.3.
1-10
-------
In the collective judgment step of CEA applied to MWCNTs, RTI International selected experts
such that the group as a whole represented a cross-section of sectors (e.g., industry, academia,
government) and technical backgrounds (e.g., material characterization, exposure and dose, human health
effects, ecological effects, risk assessment) [see Table 1-1 and RTI International (2012) for more detail on
participant sector and expertise affiliations]. Experts were first asked to read the External Review Draft of
the case study. Next, they were asked to consider what elements of the CEA framework were most
important to understanding, and therefore managing, the most significant risks associated with MWCNTs.
The experts identified important areas by independently rating areas of a more detailed view of
the CEA framework (Figure 1-3). This detailed CEA framework illustrates discrete elements (blue boxes
in top left of (Figure 1-3) or discrete pathways within the broad levels of the CEA framework (e.g.,
Product Life Cycle, Exposure in (Figure 1-1). Each element is associated with "risk relevance factors"
(green boxes in top left of (Figure 1-3). which might be considered in risk assessment or management
efforts of a material, such as MWCNTs.
Experts were asked to rate the importance of each element of the detailed CEA framework as
important, possibly important, or least important. If they rated the element important, they were then
asked to rate (1) the importance of each element-risk relevance factor (E-RRF) pair using the same scale,
and (2) their confidence in the availability and utility of current data for the E-RRF to support risk
management decisions (as confident, somewhat confident, or not confident).
These ratings were collected in each of the three rounds of collective judgment prioritization used
for this application of CEA:
Round 1: Thirty-one selected participants entered their individual opinions on the E-RRF pairs in
a spreadsheet and submitted the spreadsheet to a secure online platform (website).
Round 2: Twenty-eight of the original thirty-one participants3 viewed the compiled opinions of
the wider group through a series of bar charts and tables available via the website and were given
the opportunity to re-enter their opinions.
Round 3: A subset of participants (13) attended a structured workshop where they:
a. discussed their opinions in a structured collective judgment technique,
b. finalized research priorities through a third round of individually rating all E-RRFs and
compiling these ratings, and
c. developed detailed research questions for a subset of those priorities.
3Three participants from the first round of prioritization were unable to participate in the second round.
1-11
-------
The finalized priority areas determined in Round 3, part b of the technique, hereafter referred to as
"Priority Research Areas," are summarized in Figure 1-3 and discussed in greater detail in Section 6.3,
along with participants' rationales for selecting each priority area. As discussed below, these outcomes
were used to focus the information in this case study document.
1.1.3.3. How the Case Study Was Streamlined to Emphasize Research Priorities
Compared to the External Review Draft (U.S. EPA. 2012b). this draft of the case study document
has been streamlined to clearly reflect the outcomes of the collective judgment step of the CEA process.
New text boxes have been embedded in the document immediately following section headings4 that
correspond to elements of the detailed CEA framework (see Figure 1-3) to highlight the outcomes of the
RTI workshop (RTI International. 2012) related to the E-RRFs discussed in that section of the case study.
• Boxes outlined in red with the title "Priority Research Area Highlight" (e.g., Section 2.2.2)
indicate that the E-RRFs discussed in that section were deemed by participants in the RTI
workshop to be priorities for continuing research, based on (1) high importance of that area to
risk assessment and risk management, and (2) low confidence in the utility and availability of
the data on the topic.
• Boxes outlinedjn gray with the title "Unprioritized Research Area Highlight" (e.g.,
Section 2.1) indicate that the E-RRFs discussed in that section were not identified by
workshop participants as Priority Research Areas (i.e., the most commonly selected rating
was "possibly important" or "least important" rather than "important"; therefore, most
participants did not rate the Importance and Confidence for those E-RRFs). For these
sections, all text relevant to decaBDE and MWCNTs was moved to Appendix H and
Appendix G. respectively, to focus the main body of the document on the priority research
areas.
• Boxes outlined in black with title "Neutral Research Areas" denote case study sections that
present necessary supporting information for E-RRF pairs, but do not directly discuss a
specific E-RRF pair. For these cases, the text that originally appeared in the section remains,
as it supports understanding of other E-RRFs that are priorities for research.
In each "Priority Research Area" and "Unprioritized Research Area" highlight box, a graphic
appears that summarizes information on how the 13 workshop participants individually rated the
Importance of each element; and, for the subset of participants who stated the element was of highest
importance, their Importance and Confidence ratings for each E-RRF. The collective Importance and
4Each text box can be located using the Table of Contents because they correspond with the main section headings
in the document. These text boxes are not numbered because information in each text box is not intended to stand
alone, but rather to be read in the context of the text in that section.
1-12
-------
Confidence for each E-RRF is expressed using an Importance/Confidence Matrix image, where the three
Importance categories are shown on the Y-axis and the three Confidence categories are shown on the X-
axis, creating nine bins representing unique importance-confidence pairings (see lower left of Figure 1-3).
E-RRFs were assigned to a particular bin of the Importance/Confidence Matrix based on which ratings
were most commonly selected by expert stakeholders for Importance and for Confidence.5'6 The
prioritization of the framework areas (i.e., E-RRFs) is therefore based on the most frequently selected
rating for each factor (Importance or Confidence), rather than on the most commonly selected
combination of Importance and Confidence for each E-RRF. E-RRFs in Unprioritized Research Areas
were not assigned to a particular bin because only a small subset of participants rated the Importance and
Confidence of the E-RRF.
In most instances, the most commonly agreed-upon Importance and Confidence ratings align with
the portion of the matrix having the most stakeholders; however, in three instances,7 this is not the case.
This lack of concordance reflects a difference in how individuals combined Importance/Confidence
ratings compared to the overall rating combination of all stakeholders. In all cases, the most commonly
selected rating for Importance and the most commonly selected rating for Confidence determines the
placement of the E-RRF in the Importance/Confidence Matrix.
5In instances of a tie (i.e., 6 of 13 [or 46%] stakeholders rated an E-RRF "Important" and the same number rated the
E-RRF "Possibly Important") the more conservative rating was used as the most commonly selected rating (i.e., the
E-RRF was rated as "Important"). The same rule applies for Confidence ratings.
6For example, if 6 of 13 stakeholders rated an E-RRF "Important" and 3 of 13 stakeholders rated the E-RRF
"Possibly Important," the E-RRF was collectively rated "Important." Similarly, if 4 of 13 stakeholders, or 31%,
rated their confidence in an E-RRF as Not Confident and 3 of 13, or 23%, rated their confidence in the E-RRF as
Somewhat Confident, the E-RRF would be rated as "Not Confident." Based on both ratings, the E-RRF would be
placed in the "Important"/"Not Confident" bin of the matrix).
7(1) Environmental Transport, Transformation & Fate: Wastewater - Bioavailability; (2) Environmental Transport,
Transformation & Fate: Sediment - Mobility; and (3) Dose (Kinetics): Human - Distribution.
1-13
-------
Importance to Risk Assessment
Least possibly
Important important Important
•
Confident Somewhat No( Confjdent
Confident
Product Life Cycle
Environmental Transport, Transformation & Fate
Biota
Exposure Route
Confidence that Current Data Can Support
Risk Management Decisions
Human:
General Population
Aquatic
Biota
Terrestrial
Biota
Abiotic
[ Ingestion ]
[ inhalation ]
[ Dermal ]
( Injestion |
[ IngesTion |
( Direct Contact j
[ Direct Contact
[ Direct Contact ]
Dose (Kinetics)
Aquatic Biota
Absorption )
Distribution
( Metabolism ]
I Excretion
| Terrestrial Biota |
[ AbsJiplmr j
[ Distribution]
[ Metabolism
T
Abiotic
[ Absorption |
Impacts
Terrestrial Biota
[ Survival
Developmental ]
[ Reproductive
nomtc
Other Sublethsl
Endpoints
Figure 1-3. Detailed CEA framework used for the collective judgment prioritization process.
Note: The detailed CEA framework contains "elements" at each CEA level (product life cycle; environmental transport, transformation, and fate; exposure route; dose (kinetics); and
impacts). Each element is associated with several "risk relevance factors." In the original detailed CEA framework presented to participants in the collective judgment prioritization
process, each element was represented by a blue bar and each risk relevance factor was represented by a green bar (shown in the upper left of this figure). During the prioritization
process, participants assigned each element-risk relevance factor (E-RRF) pair a rating of "importance" and "confidence," placing each E-RRF into a bin of the Importance/Confidence
Matrix (shown in the bottom left of this figure). Ratings among all participants were tallied to determine the collective assignment for each E-RRF, which is shown in the detailed CEA
framework on the right side of this figure. Areas in white denote those deemed of lesser importance for future MWCNT risk assessments, while those in colors represent areas
collectively identified as of being high importance for future assessments. Areas in red are those of highest priority for research because participants most commonly rated the area as
"Important" to MWCNT risk assessments and were "Not Confident" that data could currently support risk management decisions. More information on the collective judgment results
for each E-RRF is presented in the "Priority Research Area Highlight" and "Unprioritized Research Area Highlight" boxes throughout the rest of this case study document.
1-1
-------
1.1.3.4. How the Case Study Was Revised to Respond to Public and Peer
Comments
Additional changes were made to the case study document, and new elements were added, in
response to written feedback from 23 experts involved in the prioritization process and in response to
comments from several members of the public (see Appendix I for more detail). Consistent with the
discussion above, these changes were implemented to emphasize research priorities identified through the
RTI workshop process and to improve the scientific accuracy and rigor of the compiled information.
First, as mentioned in Section 1.1.3 and discussed above, detailed information on decaBDE was
moved to Appendix H and replaced with text boxes that include highlights comparing information known
about decaBDE to what is known about MWCNTs in Priority Research Areas. These highlights are meant
to illustrate how understanding the data on decaBDE in flame-retardant upholstery textiles might help
guide research planning to elucidate potential risks of MWCNTs. These text boxes are consistently titled
"DecaBDE Can Inform MWCNT Assessment" and are outlined in green (e.g., Section 2.2.2).
Second, a series of "Additional Information Highlight Text Boxes," new figures, and new tables
were embedded in the case study to draw attention to scientific concepts related to the priority areas that
commenters felt were under-represented in the External Review Draft of the case study. These elements
supplement information presented in the main text or text boxes that existed in the External Review Draft.
They were added to emphasize scientific topics that were included in the External Review Draft but were
unclear or not clearly described, or to discuss a topic that was not included previously but is relevant to
the topic and discussion. Additional Information Highlight Text Boxes, new tables, and new figures are
(e.g., Table 2-2).
Finally, Chapter 6 was expanded to include a final section (Section 6.3) that discusses the priority
research areas in more detail. Section 6.3 builds on the "Priority Area Highlight" text boxes outlined in
red described above, which are intended to briefly outline how participant ratings resulted in the area
collectively identified as a priority. Examples of the rationale for prioritizing these areas are presented in
Section 6.3. along with factors that might be important to include in planning research for each area. In
addition, for some priority areas, public commenters, workshop participants, and targeted literature
searches identified relevant literature that had not been included in the External Review Draft of the case
study. This literature is discussed in Section 6.3 in the context of how it might influence research
planning. Finally, specific research questions identified by expert participants (or based on the available
literature) are listed for each priority area.
During a Letter Peer Review, five independently selected experts reviewed the revised External
Review Draft of the case study document, which is known as the Peer Review Draft (see Appendix K).
1-15
-------
These experts identified additional relevant literature or revisions to improve the document. The
suggested references were incorporated, as appropriate, in Section 6.3 or in Additional Information Text
Boxes, and appropriate revisions were implemented throughout the document, as detailed in Appendix K.
1.2. Introduction to Flame Retardants in Textiles
Textiles and fabrics, which are networks of fibers composing flexible woven or nonwoven
materials, are flammable to varying degrees due to their ignitability and their potential to propagate flame
and produce burning droplets (PINFA. 2010). The behavior of various untreated textiles when exposed to
flames depends on the chemical composition of the raw materials. Table 1-2 lists several common
categories of textile fibers along with their flammability characteristics. The flammability of these fibers,
when incorporated in different textile products, has led to the development of numerous fire safety
standards (PINFA. 2010). as discussed in Section 1.2.1.
1-16
-------
Table 1-2. Common textile fibers, and degrees of flammability.
Fiber
Flammability Characteristics of Untreated Fibers
Increasing Fire
Hazard
Cotton
Flax
Ignite easily, burn heavily; do not melt away from flame3
Viscose
Burns rapidly, similar to cotton
Acetates
Burn heavily; can melt away from flame; form burning droplets
Acrylics
Burn rapidly; form burning droplets; produce dense black smoke
Polyesters
Polyolefins
Polyamide
• Burn slowly and hotc, can melt away from flame; form burning droplets
Other synthetics
Wool
Difficult to ignite; burns slowly; might self-extinguish
Modified acrylics Burn very slowly; tend to melt away from flame; might self-extinguish
Aramide
Does not burn; strong char formation
aMelting away from the flame refers to the burning characteristic where the fiber essentially melts more quickly than the flame can
spread, thereby removing the amount of fiber that is available to the flame to continue burning.
""Burning droplets can form if the fiber melts slowly while in contact with the flame.
°Burning hot refers to a high peak heat release rate.
Source: PINFA (2010).
1.2.1. Standards for Textiles
Upholstery textiles, particularly those used outside of residential settings (e.g., in hospitals,
airports, airplanes, penal institutions, public transportation, office buildings), are subject to various state,
federal, and voluntary fire safety standards (see Table 1-3 for examples). Technical standards specify the
types of products to which standards apply, methodologies for conducting specific tests, measured
parameters of interest (e.g., time to ignition, heat release rate), and performance criteria for each test and
product of interest (Illinois Environmental Protection Agency. 2007).
1-17
-------
Table 1-3. U.S. and international fire regulations for upholstery textiles.
Product Category3
Standards3
Description3
Automotive vehicle (bus and
car) passenger
compartments; curtains or
blinds used in automotive
vehicles
FMVSS 302/DIN Specimen subjected to Bunsen burner flame for 15 seconds.
75200/ISO 3795; The rate of flame spread should be <101.6 mm/min (for a 254-
DIN 50051 mm sample); requires test specimen to have a burning rate <100
mm/min (560-mm sample length) when subjected to a vertical
flame test.
Federal flammability
standard for mattresses and
mattress pads
16 CFR 1632 (2000;
updated 2007); 16
CFR 1633 (2006);
CATB 603 (2005);
CATB129; CATB
121
Cigarette test for ignition resistance sets requirements for testing
of prototype designs of mattresses and mattress pads (based on
CA TB 106). Open flame tests: the mattress set must not exceed
a peak heat release of 200 kW at any time during a 30-minute
test, and the total heat release for the first 10 minutes of the test
must not exceed 15 megajoules (25 megajoules in California).
Filling materials used in
upholstered furniture
CATB 117
Furniture that meets the CA TB 117 standard is less likely to
ignite rapidly, and if ignited, less likely to burn quickly or to
sustain burning.
Passenger equipment in
railroad trains
49 CFR Part 238
(2002); ISO 5658-2;
ISO 9705
Safety and flammability standards for components of fixed items
in passenger cars, seating upholstery, etc. Lateral flame spread
test with heat radiator and ignition flame: specimen 800 mm by
155 mm is measured for critical heat flux at extinguishment;
flame should not exceed 100 cm above the highest point of the
seat surface.
Seating furniture for use in
public occupancies
CATB 133
Requires full-scale flame test for furniture manufactured for use
in public buildings in California. Many other states have adopted
TB133.
Cigarette testing of
upholstered furniture fabric
Upholstered
Furniture Action
Council; CATB 116
Component standard. All upholstered furniture sold in California
must pass this flame test.
Draperies (vertical fabrics) NFPA 701
Applies to draperies in buildings under NFPA 701 code.
aThis list is not meant to be definitive or complete; some fire regulations are being re-evaluated and the contents of this table might
not be current.
""Full-scale flame test refers to the use of a full piece of furniture or mockup (composite)
Abbreviations: FMVSS = Federal Motor Vehicle Safety Standards and Regulations; DIN = Deutsches Institut fur Normung
(Germany); ISO = International Organization for Standardization; CFR = Code of Federal Regulations; CA TB = California Technical
Bulletin; NFPA = National Fire Protection Association
Sources: Lowell Center for Sustainable Production [Pure Strategies, Inc. (2005)1: PINFA (2010).
1-18
-------
1.2.2. Flame-Retardant Materials as Solutions to Flammability
The flammability of textiles and the standards described above have created a growing market
demand for technologies to increase flame resistance and meet fire safety regulations (Alaee. 2003). One
way to help meet this demand is through the use of flame-retardant materials, which are chemicals or
other manufactured components that have the quality of resisting or inhibiting the spread of fire. Even
where regulatory standards do not mandate flame resistance, market pressures and concerns about brand
image often cause manufacturers to incorporate flame-retardant materials into their products (Illinois
Environmental Protection Agency. 2007). In fact, the global market for flame-retardant materials is
expected to increase 4-5% by 2015 from the base market value of $3 billion in 2009 (Grzvbowski. 2009;
Sullivan. 2009).
The most commonly used flame-retardant materials are usually broadly categorized by chemical
structure (e.g., halogenated, phosphorous-based, nitrogen-based, inorganic). Each broad class represents
many possible flame-retardant compounds. Additionally, a variety of inert fillers (e.g., talc),
manufactured components (e.g., glass fibers and
microspheres), and more technologically
advanced solutions (e.g., advances in polymer
chemistry [see Section 1.2.2.21. nanotechnology)
offer flame-retardant properties to increase the
flame resistance of textiles (PINFA. 2010; U.S.
EPA. 2005: Zhang and Horrocks. 2003).
The standards and regulations do not
specify which flame-retardant materials, if any,
must be used in textiles. Thus, various industry
stakeholders must decide which flame-retardant
materials to use based on several key criteria.
1.2.2.1. Performance Criteria
Additional Information Highlight Box 1:
1. Factors influencing selection of flame retardants
Significant uncertainty surrounds which, if any, MWCNT
flame-retardant applications are most likely to be developed
for commercial use. The formulation of flame retardants is
largely dictated by performance criteria, including flame test
performance, efficiency, cost, and effect on textile
characteristics (see Section 1.2.2.1). Although this case study
discusses MWCNT flame-retardant coatings in textiles,
alternative flame-retardant products might better meet these
performance criteria than this selected application.
Consequently, these alternative applications might be more
prominent in the future than the application explored in this
case study. Additional Information Highlight Box 3 details
some of the challenges in developing MWCNT flame-
retardant applications that meet fire safety standards and
references some potential MWCNT flame-retardant
applications.
Performance criteria help determine which flame-retardant materials are appropriate for which
applications and provide a preliminary basis for stakeholders to compare these materials. Such
comparisons are also useful in considering what materials are suitable alternatives to existing
technologies. Some performance criteria proposed by EPA (U.S. EPA. 2005) include:
1-19
-------
• Flame test performance: a measure of the efficacy of the flame -retardant material; different
measures are included in specific regulatory standards;8
• Efficiency: the degree of flame-retardant action relative to the amount of material needed to
obtain the result;
• Cost: expense associated with raw materials and downstream production;
• Impacts on textile characteristics: effect on features that can alter a product's desirability to
consumers (e.g., enhanced strength, reduced aesthetic appeal).
Another important aspect of performance is durability. Durability is a measure of the ability of a
flame-retardant material to maintain an acceptable level of flame-retardant behavior throughout the
lifetime of the textile as it undergoes abrasion, laundering, weathering, or other expected processes
(PINFA. 2010; NRC. 2000). The durability standard required depends on the intended use of a textile
product. Durability classifications for flame-retardant finishes in textiles are presented in Table 1-4.9 In
some cases, an evaluation of durability is a component of the flame tests (e.g., both pre- and post-wash
tests are required for some product uses).10 Importantly, these criteria pertain only to the performance of a
specific flame-retardant material in a specific application.
Table 1-4. Durability classifications of flame-retardant finishes.
Durability Classification
Example Flame-Retardant Materials
Example Application in Textiles
Nondurable - not resistant to
washing
Boric acid, aluminum sulfate, ammonium
salts, phosphates, some halogenated
compounds
Mattresses, draperies, rarely
washed textiles
Semidurable - resistant to limited Cyanamide and phosphoric acid,
number of washes phosphorylation of cellulosic fibers, some
halogenated compounds
Tents, carpets, curtains (resistant
for up to 50 washings)
Durable - resistant to many
washes
Organic phosphorous compounds, some
brominated compounds
Clothing, other frequently washed
fabrics
Sources: BfFV (1998) and PINFA (2010).
8For example, the cigarette ignition test and vertical flame test measure aspects such as char length and afterglow of
a sample. Flame test parameters and standards are specific to products and end uses [see (Exponent 2010: ICL.
2010: Babrauskas and Krasny. 1985)1.
'Durability should not be confused with teachability, which refers to the percent removal of a flame retardant from
the textile matrix (NRC. 2000). Leachability is an important measure from a risk evaluation perspective, while
durability is more important from a product performance perspective.
10The Federal Register specifies which textile types and products require flame resistance for up to a specific
number of washes.
1-20
-------
1.2.2.2. Flatne-Retardant Application Methods
Two principal processes are used for incorporating flame-retardant materials into the textile
matrix: reactive and additive. Generally, flame retardants incorporated into the textile matrix using the
reactive process produce durable finishes; flame retardants simply added to the textile matrix produce
nondurable or semidurable finishes (U.S. EPA. 2005; Rahman et al.. 2001).
In the reactive process, flame-
retardant materials are incorporated directly
into polymeric materials during the
manufacturing process such that they are
chemically (i.e., covalently) bound to the raw
materials of the final product (U.S. EPA.
2005: Rahman etal.. 2001). Direct
incorporation also can be accomplished using
a chemical reaction between two monomers to
form a strong polymer chain or in a post-
reaction process such as chemical grafting
(functionalization) (PINFA. 2010; Laoutid et
<_)
'i_
_Q
ru _
LJ_
> Pad
^ Back- _
Coat
> Pad -
- Heat
Cure
Chemical
* Cure ~
^ Dry/
Cure
> Dry
-5
-3
>• Wash -
Fix/
* Wash ~
•*• Dry
•» Dry
/
\
3
ro
O>
i/i
3
CTQ
U
C
55
0"
rt-'
-<
Source: Adapted from NRC (2000).
Figure 1-4. Durability of additive flame retardants.
Note: This general schematic of the "additive" application method of
flame-retardant materials for textiles demonstrates the steps in the
al.. 2009). Flame-retardant materials produced additive flame-retardant process that increase durability. The curing
' v process can result in cross-linking, thermal fixation, or ionic linkage
by the reactive mechanism are often between the flame-retardant material and the fibers to increase
considered to be "inherently" flame resistant,
as is the case with a variety of polyester blend fabrics.
In the additive mechanism, flame-retardant materials are applied to the fibers, the finished textile,
or the finished product without the formation of chemical bonds and without a chemical reaction (Laoutid
et al.. 2009: U.S. EPA. 2005: Rahman et al.. 2001). In some cases, binding agents, resins, or copolymers
are used to increase the durability of the flame-retardant properties of the textile.
This case study focuses on the additive application of flame retardants to upholstery textiles. This
method can be distinguished further as illustrated by the schematic in Figure 1-4. which
shows the simple "pad/dry" technique and variations. In the pad/dry technique, the textile is immersed in
a bath of flame-retardant solution and then squeezed through rollers at a specific pressure to remove
excess solution. Back-coating describes several related application methods where a bonding resin
containing the flame retardant is spread and smoothed across the reverse surface of a textile using a knife
or blade (PINFA. 2010).
1-21
-------
Although additive flame retardants typically produce a nondurable finish (NRC. 2000). Figure
1-4 illustrates methods of addition that produce more durable finishes than the simpler pad/dry additive
techniques. Thermal or chemical curing, for example, allows for interaction between the flame-retardant
material and the fiber that results in a more durable finish than those produced using the simpler
techniques (PINFA. 2010). Curing provides the opportunity for cross-linking (polymerization of the
flame retardant onto the substrate), thermal fixation (deposition of the flame retardant within the fibers),
or ionic linkage (negatively charged complexes bind to positively charged groups). These processes
essentially "trap" the flame-retardant material within the polymer chains, producing a finish that is similar
to those produced by the reactive method (PINFA. 2010; NRC. 2000). After curing, the textile is
subjected to other processes (oxidation, neutralization, or washing) to remove by-products before the
material is dried. Although a more durable finish can be obtained with heat curing or chemical curing,
leaching of flame-retardant material remains of greater concern for additive flame retardants than for
reactive flame retardants (not shown in Figure 1-4) because the material is not covalently bound to the
substrate (Rahman et al., 2001).
1.2.2.1. Mechanism of Flame-Retardant Action
Once incorporated into the textile, flame-retardant
materials physically or chemically inhibit the combustion
process. Combustion occurs through a series of chemical
reactions including heating and ignition, volatilization, and
decomposition, which are self-propagating in the presence of
oxygen and a fuel source (U.S. EPA. 2005; Alaee. 2003) (see
Figure 1-5).
The mechanism of flame-retardant action can be
categorized generally as follows, although many flame
retardants actually inhibit the combustion process through a
combination of these mechanisms (U.S. EPA. 2005; Alaee.
2003: Rahman et al.. 2001):
Source: Adapted from Alaee et al. (2003) and
Laoutid (2009).
Figure 1-5. The combustion process.
Note: The combustion process consists of distinct
but overlapping reactions between a fuel source and
an oxidant in the presence of heat. Ignition and
volatilization in the presence of oxygen produce
additional heat, which propagates the cycle.
• Physical Dilution: The flame-retardant material (1) reduces the fuel content available for
combustion below the concentration needed to sustain flame propagation or (2) increases the
heat capacity of the product, which increases the amount of heat required for product ignition.
1-22
-------
• Chemical Interaction/Gas-Phase Radical Quenching: The flame-retardant material
thermally degrades and releases chemical radicals that are highly reactive with oxygen, which
reduces the amount of free oxygen available to supply the combustion process.
• Inert Gas Dilution: The flame-retardant material produces a large volume of
noncombustible gases that dilute the oxygen supply available to propagate the flame.
• Thermal Quenching: The flame-retardant material endothermically degrades, which
removes heat from the substrate and cools the material.
• Protective Coatings: The flame-retardant material forms a liquid or char coating that acts as
an insulation barrier to prevent heat transfer from the flame to unaffected areas of the
product.
These five processes act individually or in combination to increase the time to ignition, prevent
spread of the flame, or decrease extinguishing time (Alaee. 2003). Table 1-5 provides examples of each
major chemical class of flame retardant described in Section 1.2.2 along with a description of how the
flame retardants are added to textiles, and the general mechanism of flame-retardant action for each class
of flame retardant.
Table 1-5. Flame retardants summarized by chemical class, method of application, and mechanism
of flame-retardant action.
Chemical Class Examples
Application Method
Flame-Retardant Mechanism
Halogenated
Chlorinated
(polychlorinated biphenyls),
Brominated
(polybrominated diphenyl
ethers)
Variations of padding and
drying or back-coating
Gas-phase radical
quenching/chemical reaction to
slow the burning rate; also can
form a solid protective layer
Monomers and copolymers
(vinyl bromide),
Tetrabromobisphenol A
Combined with copolymeric Decreases thermal degradation;
modifications or grafted onto reduces extinguishing time
polymer chains for reactive
application
Phosphorous-
based
Nitrogen-based
Inorganic
Organophosphorous,
Inorganic phosphates
Melamine,
Melamine salts
Metal hydroxides,
Minerals
Coatings; chemical bath
Intumescent coatings; back-
coatings; can be added to
polymer melt
Fillers; back-coatings; can be
added to polymer melt
Protective coatings or layers; char
formation
Inert gas dilution (inhibits
formation of flammable gases);
char formation
Endothermic degradation/thermal
quenching or inert gas dilution;
forms protective layer; physical
dilution; thermal shielding
Sources: U.S. EPA (2005): BfFV (1998): NRC (2000): Xusen (2010): PINFA (2010): and Laoutid et al. (2009).
1-23
-------
1.3. DecaBDE and MWCNTs in Flame-Retardant Textiles
As noted previously, the purpose of this case study is to present available information that
supports research planning to inform a future comparative CEA of a traditional flame retardant
(e.g., decaBDE) and ananoenabled flame-retardant technology using MWCNTs, specifically in
upholstery textile coatings. Similarities across the product life cycle for different types of flame retardants
(e.g., manufacturing processes, use scenarios, and disposal procedures) can inform the identification of
pertinent MWCNT exposure scenarios for human and ecological receptors. Although the physicochemical
characteristics of MWCNTs differ in important ways from those of conventional materials, such as
decaBDE (see Text Box 1-1), the information gained from studying key parameters that influence
decaBDE impacts on ecological and human health (e.g., debromination, long-range transport) might
inform the identification of research priorities for MWCNTs. Although unique challenges exist in
understanding environmental transport, transformation and fate of engineered nanomaterials, including
MWCNTs (see Chapter 3. Additional Information Highlight Box 10). research on PBDEs, including
decaBDE, also has struggled with similar questions [e.g., availability of standard reference materials,
distinguishing among similar materials, necessity of using multiple analytic techniques in tandem (see
Appendix B)1. Strategies applied to address these issues for PBDEs could inform research planning for
MWCNTs. As discussed in Section 1.1.3. the primary purpose of including decaBDE for comparison was
accomplished when expert stakeholders reviewed the External Review Draft of this case study and
identified MWCNT research priorities; thus, information on decaBDE is presented primarily in Appendix
H. The following sections, however, provide a general overview of decaBDE and MWCNTs, their use in
textiles, and a brief comparison of observed flame-retardant action and efficacies so that readers can be
oriented to key considerations regarding the feasibility of MWCNTs as an alternative to decaBDE flame
retardants. Considerations relevant to other aspects of this comparison (e.g., production techniques,
exposure scenarios, environmental behavior) are described in subsequent chapters.
1.3.1. Introduction to DecaBDE
DecaBDE is part of a larger group of BFRs called polybrominated diphenyl ethers (PBDEs), a
group of 209 structurally similar BFRs that differ in the number and location of bromine atoms
(Table 1-6) (Rahman etal.. 2001: NRC. 2000). Although PBDEs are typically categorized into classes by
number of bromine atoms [e.g., PBDE with two bromine atoms is a dibromodiphenyl ether (diBDE); ten
bromine atoms is decaBDE], a single class might contain several different PBDE congeners with the same
1-24
-------
number of bromine atoms in different locations (i.e., PBDE BFRs can have many isomers). As the only
fully brominated PBDE, decaBDE is the exception, existing only as a single congener (BDE-209).
Table 1-6.
PBDE Class
DiBDE
TriBDE
TetraBDE
PentaBDE
HexaBDE
HeptaBDE
OctaBDE
NonaBDE
DecaBDE
Major PBDE congeners.
Congeners
BDE-7, BDE-8, BDE-11, BDE-12, BDE-13, BDE-15
BDE-17, BDE-25, BDE-28, BDE-30, BDE-32, BDE-33, BDE-35, BDE-37
BDE-47, BDE-49, BDE-66, BDE-71, BDE-75, BDE-77
BDE-85, BDE-99, BDE-100, BDE-105, BDE-116, BDE-11 8, BDE-11 9, BDE-126,
BDE-138, BDE-140
BDE-153, BDE-154, BDE-155, BDE-166
BDE-181, BDE-183, BDE-190
BDE-196, BDE-197, BDE-203
BDE-206, BDE-207, BDE-208
BDE-209
Source: U.S. EPA (201 Ob).
Commercial formulations of decaBDE (see Table 1-7) are generally 97-98% BDE-209 with less
than 3% nonaBDE congeners present as impurities (Rahman et al., 2001; NRC. 2000) (see Appendix B.
Table B-l for analytical techniques used to distinguish PBDE congeners in samples). Although the terms
decaBDE and BDE-209 often are used interchangeably, this case study primarily uses the term decaBDE
to refer generally to the flame-retardant formulation and BDE-209 to refer to the specific decaBDE
congener analyzed in scientific studies.
As discussed in Section 1.1, decaBDE is the most widely used of the PBDEs and has been well
studied; however, increased concerns regarding the potential impacts of decaBDE on ecological and
human health have resulted in several states beginning to phase out or restrict the use of decaBDE
(see Table 1-1).
In December 2009, the two largest U.S. producers and the largest U.S. importer of decaBDE
announced voluntary commitments to phase out decaBDE in the United States by 2013 (U.S. EPA.
2010a). As summarized in the EPA PBDE Action Plan dated December 30, 2009, several reports
provided evidence for the human and environmental effects of this compound (U.S. EPA. 2012c). A
finding of "suggestive evidence of carcinogenic potential" was reported in the 2008 Toxicological Review
1-25
-------
ofDecaBDE (U.S. EPA. 2008b). Neurobehavioral effects also were identified in IRIS (Integrated Risk
Information System) assessments for decaBDE and additional congeners (tetraBDE, pentaBDE, and
hexaBDE). Environmental hazards associated with PBDEs include persistence, potential for
biomagnification, and breakdown of some PBDEs to more toxic congeners to produce effects at
environmentally relevant concentrations (based on reports from Environment Canada and studies from
other authors, see Chapter 3. Chapter 5. and Appendix H). Furthermore, in 2012, EPA initiated proposed
amendments to (1) TSCA § 5(a)(2), a Significant New Use Rule (SNURX and (2) TSCA § 4, a Test Rule
for decaBDE. The SNUR would require any entity planning to manufacture or import decaBDE or
articles to which decaBDE has been added to notify EPA at least 90 days in advance, which would
provide the Agency an opportunity to review and evaluate data related to the new use and to take action to
limit or prohibit the new use if necessary. The Test Rule would require laboratory studies to determine the
effects that decaBDE has on human health and the environment (U.S. EPA. 2012c).
Table 1-7. Commercial formulations of PBDEs used as flame retardants.
Name
Penta formulation3
Octa formulation
Deca formulation13
"Trace amounts of additional
Congener Makeup and Percent Composition
Penta
Tetra
Hexa
Hexa
Hepta
Octa
Nona
Nona
Deca
BDE-99 (35-50%), BDE-100 (6-10%)
BDE-47 (25-37%)
BDE-153 (5-10%), BDE-154 (1-5%)
BDE-153 (5-10%), BDE-154 (1-5%)
BDE-183(40%)
BDE-197 (21%), BDE-203 (5-35%), BDE-196
BDE-208 (10%), BDE-207 (7%)
BDE-206 (2.2%), BDE-207 (0.24%), BDE 208
BDE-209 (>97%)
(8%)
(0.06%)
congeners might be present in commercial formulations: <0.2% triBDE congeners.
bTrace amounts of additional congeners might be present in commercial formulations: <0.003% heptaBDE congeners; <0.001 %
hexaBDE congeners; <0.002% pentaBDE congeners; <0.00003% tetraBDE congeners; <0.00001% triBDE congeners.
Source: U.S. EPA (201 Ob).
DecaBDE can be applied to textiles by a variety of mechanisms, but this case study focuses on
the application of decaBDE as a back-coating. This application method is used most frequently for
decaBDE (Pure Strategies Inc.. 2005; NRC. 2000) and is most similar to the application method expected
1-26
-------
for MWCNTs used in textiles (see Section 1.3.2). The back-coating process usually involves mixing
decaBDE with a copolymer or resin binder (Pure Strategies Inc.. 2005; NRC, 2000). DecaBDE combines
the flame-retardant mechanism of most BFRs (i.e., releasing halogens during combustion to compete with
the availability of oxygen for the flame) with formation of a protective char barrier (NRC. 2000) that
interferes with the spread of the flame and helps the material to self-extinguish (Pure Strategies Inc..
2005).
Table 1-8. Physical properties and chemical identity of decaBDE.
Physical property/chemical identity
Citation
CASRN
1163-19-5
Physical state
Solid
NLM (2011)
Synonyms 2,2',3,3',4,41,5,51,6,61-decaBDE; BDE-209; benzene, 1,1'- NLM (2011):
oxybis[2,3,4,5,6,-pentabromo]-; decabromodiphenyl oxide; ATSDR (2004)
decabromodiphenyl ether; decabromobiphenyl ether; ether,
bis(pentabromophenyl)
Hardy (2002b)
Melting point,
300-310°C
ECB (2003)
Boiling point
Decomposes at >320 °C
ECB (2003)
Vapor pressure
4.63 x 10~bPaat21 °C
Hardy (2002b)
Henry's law constant 1.93x10 L atm/mol
0.04 Pa m3/mol at 25 °C
Hardy (2002b);
Cetin and Odabasi (2005)
Density
3.0 grams/cm
NRC (2000)
Water solubility
.1 ug/Lat25°C
Hardy (2002b); ECB
(2003)
Log Kow
6.3-12.6
Hardy (2002b)
Log Koc
6.3
Hardy (2002b)
Molecular weight 959.17
NLM (2011): ECB (2003)
Chemical formula
NLM (2011)
Chemical structure
Abbreviations: Kow= Octanol water partition coefficient, Koc = Soil organic carbon-water partition coefficient.
1-27
-------
1.3.2. Introduction to MWCNTs
MWCNTs are carbon nanostructures composed of multiple concentrically nested graphene sheets
that look similar to nested rolls of chicken wire. Unlike many traditional chemicals, MWCNTs are not a
homogeneous group of molecules; many of the characteristics of MWCNTs can be intentionally or
unintentionally altered using different laboratory procedures, treatments, and synthesis methods (see
Sections 2.2 through 2.2.3 and Appendix G. Sections G.2.1 through G.2.2.2). As described in Text Box 1-
1, altering the physicochemical properties of MWCNTs can alter their behavior during all stages of the
life cycle, in environmental compartments, and in humans and other biota. As a result, MWCNTs with
different physicochemical properties might produce different impacts downstream, but which
physicochemical properties drive these differences and to what degree are not fully understood. Ranges of
values describing MWCNT physicochemical properties are provided in Table 1-9 to illustrate the array of
characteristics recorded for MWCNTs in the literature (see Appendix B. Table B-2 for analytical
techniques used to characterize MWCNTs).
1-28
-------
Important Physicochemical Properties of Nanomaterials:
•Size, including agglomeration/aggregation tendencies
• Morphology, including shape and crystal structure
•Surface area
•Chemical composition
• Surface chemistry and reactivity
•Solubilityand dispersion
• Conductive, magnetic, and optical properties
Text Box 1-1. Physicochemical Properties of Multiwalled Carbon Nanotubes (MWCNTs)
Affect Their Release, Behavior in the Environment, and Interaction with Biota
A substance's physicochemical characteristics largely
determine the environmental fate and transport and
potential for release, exposure, and impacts
associated with that substance. Yet, the key
characteristics that determine behavior differ between
conventional materials and engineered nanomaterials.
Driving characteristics for conventional materials
include boiling point, melting point, and density. These
are replaced at the nanoscale by size, surface area,
surface chemistry, and morphology. Altering just one
of these characteristics influences the behavior of nanomaterials. A single type of nanomaterial, such as MWCNTs, can be
created with dozens of variations of these properties. MWCNTs are generally nested tube-like structures with a very high
length-to-diameter ratio, but they can be engineered to have various lengths, surface coatings, and functionalizations.
Demonstrating and quantifying relationships between
individual characteristics and MWCNT behavior is
complicated due to the difficulty in altering only one
characteristic at a time. For example, oxidizing
MWCNTs can shorten them and make them more
straight (Johnston et al., 2010). Equally difficult is
assigning mechanisms of toxicity to the observed
effects. For instance, long, straight MWCNTs injected
under the skin of rats can produce more
inflammogenic effects than shorter bundles of
MWCNTs administered in the same manner
(Johnston et al., 2010), but whether the length of the
materials, their tendency to bundle and how tightly, or
all of these factors directly affect inflammation is
unclear. The complex relationships among
physicochemical properties, and between these
properties and the life cycles of nanomaterials, have
not been fully elucidated. The impacts of these
properties on nanomaterial behavior also have not
been analyzed adequately, particularly in terms of
understanding the potential environmental and health
effects of nanomaterials.
Single
physchem.
property
Acid treatments that purify
MWCNTs also introduce
structural defects.
Surface chemistry causes
MWCNTs to stabilize in a
solution in the presence of
nominal organic matter,
allowing dispersion and
mobility in aquatic
environments.
Shorter MWCNTs (<15n.m) are
more likelytobe taken upby
cells, all owing higher internal
cell exposure.
Functionalization of MWCNTs
by polystyrenetreatmentsup-
presses cytotoxic responses in
vitro and in vivo.
Other
physchem.
properties
Fate and
transport
1-29
-------
Table 1-9. Physical properties and chemical identity of MWCNTs.
Physical property/chemical identity3'
Citation
Physical state Solid
Morphology
Concentric cylinders
Johnston et al. (2010)
Physical structure
Hirsch and Vostrowsky (2005)
Purityd, min. wt%
C
Outer diameter
Length
Aspect ratio
Chirality
Usually >90%; 7.5-40% reported for "multi-
wall, powdered cylinder cores" and "multi-
wall, as produced" by Sigma-Aldrich
5-170 nmc
20 nm-200 urn
Up to 1,000
Varies; chiral angles, described by vectors (n,
See Appendix F study summaries; Sigma-Aldrich
(2012)
Aschberger et al. (2010): Li and Huang (2011):
Desai et al. (2012): He et al. (2012): Dawson et al.
(201 1 ); Golovin et al. (201 1 ); Lu et al. (201 1 b); Liu
et al. (201 1 a); Ji et al. (201 1 ); Sigma-Aldrich
(2012)
Aschberger et al. (2010): He et al. (2012):
Aranberri et al. (2011): Golovin et al. (2011): Lu et
al. (201 1b): Liu et al. (2011 a): Ji et al. (2011):
Sigma-Aldrich (2012)
Cipiriano et al. (2007)
Gustavsson et al. (2011)
m), produce different graphene sheet
conformations [e.g., "zigzag" (m = 0),
armchair (n = m)], and influence other
properties (e.g., mechanical, optical,
electrical)
Axial and radial
strength
Surface area
Bundle size
Surface
composition
Vapor pressure
Melting point
Stability
Density
Zeta potential, mV
Axial: rigid; 10 times stronger than steel
Radial: flexible; can be bent up to 90 degrees
253-400 m2/gram
0.9-100 urn
Pristine or modified with various functional
groups
No information available
3,652-3,697 °C
Stable up to 600 °C (CNTs)
2.1 grams/mL at 25 °C
-23-0
Gustavsson et al. (2011)
Aschberger et al. (2010): Aranberri et al. (2011):
Lu et al. (201 1b):
Li and Huang (2011): Baitinger et al.
Johnston et al. (2010)
Sigma-Aldrich (2012)
Nanoshel (2011)
Sigma-Aldrich (2012)
Li and Huang (2011)
(2011);
Solubility8 in water Insoluble; functionalization treatments result Lam et al. (2006): Johnston et al. (2010)
in different degrees of solubility
aValues reported represent total ranges reported in literature.
"Values are provided for MWCNTs where available, or values for single-walled CNTs or CNTs in general are provided, when not.
""Depending on the number of walls.
dlmpurities include cobalt, iron, nickel, and molybdenum (commonly used as catalysts in production; see Sections 2.2 through 2.2.3.
and Appendix G. Sections G.2.1 through G.2.2.2). Percent purity depends on purification methods (see Appendix C).
eSee Footnote 16 in Chapter 2.
1-30
-------
According to the scientific
literature, MWCNTs can act as flame
retardants in a variety of textiles, including
plastics, polymers, assorted fabrics, and
technical materials (Goncalves et al.. 2012;
Alimohammadi et al.. 2011; Binetruy and
Boussu. 2010: Kashiwagi et al.. 2005b:
Kashiwagi et al.. 2005a). To date, at least
one MWCNT flame-retardant textile
coating is commercially available, but this
application does not appear to be
widespread (Nanocvl. 2009) (see Additional
Information Highlight Box 2).
MWCNTs primarily inhibit flames
in a manner similar to that of one of the
mechanisms of decaBDE, that is, by
forming a protective layer that seals against
combustion. When formed by MWCNTs,
this protective char-like layer often is
referred to as a "network-structured layer"
that can act as a thermal shield (Laoutid et
al.. 2009; Cipiriano et al.. 2007; Kashiwagi
et al.. 2007: Kashiwagi et al.. 2005b:
Kashiwagi et al.. 2005a: Kashiwagi et al..
2004). reduce the peak heat release rate, and
increase thermal conductivity (Laoutid et
al.. 2009; Cipiriano et al.. 2007; Kashiwagi
et al.. 2007: Kashiwagi et al.. 2005b:
Kashiwagi et al.. 2005a: Beyer. 2004;
Kashiwagi et al.. 2004).
The flame-retardant behavior of
MWCNTs depends on the formation of a
highly uniform, network-structured layer of
Additional Information Highlight Box 2:
2. MWCNTs are not widely used as flame-retardants
in textiles
Although one commercial MWCNT flame-retardant coating has
been developed [(Luizi, 2009): Personal Communication: Nicolas
Messin (Nanocyl). 3/2/2012], MWCNTs currently are not widely
used as flame-retardants in textiles. Nevertheless, in recent years,
the potential use of MWCNTs as a flame-retardant additive has
been widely studied. Researchers have evaluated the feasibility of
integrating MWCNTs into materials such as polymers,
polyurethane, and epoxy resins for use in a multitude of
applications [e.g., (Yuetal., 2011: Leeetal., 2010b: Yuetal.,
2009)1. Results show that MWCNTs improve thermal stability and
provide a char layer for flame retardancy. In addition, research has
shown that nanocomposite materials containing MWCNTs perform
better compared to non-nanocomposite materials in terms of
mechanical or structural properties, such as tensile strength and
conductivity (Yuetal., 2011: Leeetal., 201 Ob: Yuetal., 2009:
Verdejoetal., 2008). Further, researchers have already
demonstrated the ability to incorporate MWCNTs into fabrics using
a process that mirrors industrial dyeing methods (Goncalvesetal.,
2012). Im et al. (2011) also illustrated that the integration of
MWCNT-AIOs fibers with polyurethane fibers, which are used in a
variety of products, including spandex/lycra textiles, improved
thermal oxidation stability. A patent is on record for a method to
incorporate MWCNTs into several different textile types (e.g.,
cotton, wool, silk, flax, nylon, polyester, acrylic) that includes
documentation of flame-retardant properties compared to raw
cotton (Alimohammadi et al., 2011). The work by Goncalves et al.
(2012), Im et al. (2011), and the recent patent indicate that more
commercial applications of MWCNTs in textiles might be available
in the near future.
Yet, future use of MWCNTs in flame-retardant coatings applied to
textiles will be determined by their ability to pass specific flame-
retardant regulatory tests (see Additional Information Highlight Box
1 Additional Information Highlight Box 3, and Table 1-12), and by
the feasibility of large-scale production processes. While global
annual production volumes for CNTs are estimated to range from
11 to 1,000 t/year (Piccinnoetal., 2012) and are projected to reach
9,400 tons by 2015 (Innovative Research and Products
Incorporated, 2011), the proportion of the total global MWCNTs
produced that are used in textiles is anticipated to be very low (see
Table 1-10 and Table 2-2 and Section 2.2.1.1). Other applications
of MWCNTs (e.g., electrodes, electronic components, filters and
membranes, sensors, cosmetics, molecular computing and data
storage, fuel cells) currently occupy a larger percentage of the
MWCNT market (Schnorr and Swager, 2011: Kohleretal., 2008).
Although greater production volumes for other MWCNT applications
were considered in developing this case study, other factors, such
as exposure potential, were also important (see Section 1.1).
floccules, which are loosely bound MWCNT bundles, with no breaks or cracks. The formation of the
1-31
-------
floccule layer, and in turn the flame-retardant behavior, varies according to a variety of factors, including
dispersion (which can be enhanced with surface treatments), size, shape, aspect ratio,11 and loading
concentration (Cipiriano et al.. 2007; Kashiwagi et al.. 2007; Kashiwagi et al.. 2005b; Kashiwagi et al..
2005a; Kashiwagi et al.. 2004). Cipiriano et al. (2007) were able to produce a more uniform floccule layer
at lower concentration loadings, resulting in enhanced flame-retardant properties, by using MWCNTs
with a higher aspect ratio. In nanoclay, the incorporation of organomodified montmorillonite, but not
sodium-layered montmorillonite, stimulated char formation in polymer matrices, indicating that surface
functional groups can be instrumental in flame-retardant action (Laoutid et al.. 2009).
1.3.3. MWCNTs as Alternative Flame-Retardant Materials in Upholstery
Textiles
As the use of decaBDE begins to decline, cost-effective and feasible alternatives to replace this
widely used flame retardant are being evaluated. MWCNTs are one of many possible alternatives to
replace decaBDE. Given the current, albeit limited, availability of an MWCNT product for textile
applications12 and the projected decline in cost (Sullivan. 2009). the use of such nanoenabled products is
likely to increase in the future as an emerging application. Table 1-10 provides a comparative summary of
decaBDE and MWCNTs, which illustrates several similarities in application method, flame-retardant
action, and relevant uses. Many parallels also can be drawn in the performance criteria (discussed in
Section 1.2.2.1). as shown in Table 1-11.
A summary of information available from actual flame tests for decaBDE and MWCNTs is
provided in Table 1-12. Note that information available for flame tests for both materials is not
standardized. Some information is qualitative, while other data are quantitative. In both cases, only a few
representative examples are described; in the absence of specific data on MWCNTs used in textiles
relevant to this case study, available information has been provided on a similar MWCNT flame-retardant
product used as a coating for a variety of materials. This product is used on foam and other structural
materials, but might prove useful for comparison to the current application because it generally has been
shown to increase flame resistance, increase the heat barrier and charring, and reduce the amount of
smoke created. Furthermore, the manufacturer advertises that their MWCNT flame-retardant coating is
1 lrThis dimension refers to the proportional relationship between the length and width of the nanotube; CNTs (carbon
nanotubes) typically are characterized as having large aspect ratios (i.e., greater length than width).
12Personal Communication: Nicolas Messin (Nanocyl). 3/2/2012.
1-32
-------
appropriate for application to textiles (Mezzo. 2010). Table 1-13 describes properties of MWCNTs that
impact performance as a flame retardant.
Table 1-10. Overview of decaBDE and MWCNTs for flame-retardant textile application.
MWCNTs
DecaBDE
Method of Integrated by "melt blending" with polymer
incorporation (Cipiriano et al., 2007: Kashiwaqi et al., 2005b:
into textile Kashiwaqi et al., 2005a: Zhang and Horrocks,
products 2003): dispersed in resin (e.g., silicon base) and
applied as a coating (Nanocyl3) (Kohler et al..
2008): applied by "dyeing-like" method to fabrics
(e.g., cotton) (Goncalves et al.. 2012): "laver-bv-
layer coating" (foam applications) (Uddin and
Myden, 2011 a: Davis and Kim, 2010).
Applied as a back-coating to textiles (NRC, 2000),
often with a binding agent such as latex (ECB,
2003), or a copolymer (NRC, 2000).
Mechanism MWCNT network acts as a sealing or shielding
of flame- agent (i.e., a barrier) (Berqer. 2007: Cipiriano et
retardant al.. 2007: Kashiwaqi et al.. 2007: Kashiwaqi et al..
action 2005b: Kashiwaqi et al.. 2005a: Kashiwaqi et al..
2004).
Gas-phase radical quenching; creates a char
barrier (NRC. 2000).
Approximate From 2005 to 2009, global annual production
production capacity increased from 294 tons (approximately
volume/ 267 tonnes) to more than 1,500 tons
capacity13 (approximately 1,361 tonnes); projected to reach
9,400 tons (approximately 8,528 tonnes) by 2015
(Innovative Research and Products Incorporated,
2011: Kohler etal., 2008): percent of MWCNTs
produced for textile use expected to be very low.
Worldwide demand in 2001 reported as 54,000-
56,000 tonnes (Law etal.. 2006: Pure Strategies
Inc.. 2005): >60,000 tonnes reported in 2007
(Illinois Environmental Protection Agency. 2007):
10-20% of decaBDE produced is used in textiles0
Relevant use Has been tested in "nanocomposites" with
in textile polyvinyl acetate, and ethylene vinyl acetate
applications (Kashiwaqi et al.. 2005b: Kashiwaqi et al.. 2005a:
Kashiwaqi et al., 2004): marketed as coating on
cables, metal, foam, and textiles (Nanocyl);
marketed as "additives" for flame-retardant
textiles or as industrial coatings for fabrics
(Siegfried, 2007): tested in various textiles
(e.g., cotton, wool, polyester, acrylic) with flame-
retardant action reported (Goncalves et al.. 2012:
Alimohammadi et al.. 2011).
Used in mattresses, draperies, commercial
upholstered furniture, and in transportation
industry fabrics (Pure Strategies Inc.. 2005).
aSee (Nanocvl. 2009: Sullivan. 2009).
bNote: not all of the production capacity is relevant for flame retardants or for use in upholstered textiles.
""According to U.S. EPA (201 Ob), production volumes for PBDEs are not readily available. The most recent industry reporting year
for market demand was 2001. At that time, 83% of all PBDE consumed worldwide was decaBDE.
1-33
-------
Table 1-11. General qualitative comparisons of performance criteria for decaBDE and MWCNTs.
MWCNTs
DecaBDE
Flame test Reduces peak heat release rate; formation of
performance network-structured protective layer; reduced rate
of pyrolysis (Cipiriano et al.. 2007: Kashiwaqi et
al.. 2005b: Kashiwaqi et al.. 2005a: Kashiwaqi et
al., 2004); increased limiting oxygen index (LOI)a
(Alimohammadi et al., 2011).
Allows textiles to comply with fire safety
standards in public places and public buildings.
Also used to comply with more stringent fire
safety requirements for home upholstered
furniture in countries such as Ireland, the United
Kingdom, and the United States (Pure Strategies
Inc., 2005).
Efficiency
Effective at very low concentrations (0.5-4% by
mass) (Grzvbowski. 2009: Kashiwaqi et al..
2005b: Kashiwaqi et al.. 2005a: Kashiwaqi et al..
Very efficient, can be used at relatively low
concentrations (Pure Strategies Inc.. 2005:
Rahman et al.. 2001): maximum of
2004): commercial formulations for textile use
are reported to be effective at 100 ug thickness13
approximately 20% w/w added as a back-coating
(NRC. 2000)°: applied 10-15% by weight to
polymers in conjunction with resin binder (U.S.
EPA, 201 Ob).
Cost
Relatively low cost of production ($100/kg)
compared to other nanocarbon products,
projected to decrease to $10-$20/kg (Sullivan,
2009): commercial prices vary with purity, size,
and functionalization (e.g., $7,000/kg for
functionalized, $2,000/kg for 95 wt% <8 nm;
$700/kg for 95 wt% >50 nm) (Cheap Tubes Inc..
2009).
Described as "cost effective" (Pure Strategies
Inc., 2005) and "relatively cheap" (Posner,
2004): decaBDE/antimony oxide mixture is
roughly $3.09/kg when used for draperies (Pure
Strategies Inc.. 2005).
Impacts on Can improve physical and mechanical properties
textile (Siegfried, 2007: Hirsch and Vostrowsky, 2005:
characteristics Kashiwagi et al., 2005b), as well as conductivity
and optical properties" (Siegfried, 2007): fatigue
resistant; particle embedding can prevent cracks
(Grzvbowski. 2009).
Must be applied to reverse side of fabric
because of negative effect on aesthetics.
Durability Commercial product has not been tested for
durability in laundering, but flame-retardant
performance is resistant to other chemical
treatments'3; potential for nanotextiles to release
individual nanoparticles or clusters of
nanoparticles (GreRler et al., 2010)e
Semidurable (Rahman et al.. 2001): resin
applied with decaBDE bonds to the fiber to
increase durability (Pure Strategies Inc.. 2005).
"Limiting Oxygen Index (LOI) is a measure of the minimum percent concentration of oxygen that will support combustion.
bPersonal Communication: Nicolas Messin, Global Sales and Marketing Manager for Thermosets (Nanocyl). 3/2/2012.
°Depends on the PBDE used (both decaBDE and hexaBDE referenced), the resin binder used, and the fabric to be treated.
dOptical properties of textiles include fluorescence or color-changing effects (theoretical; not necessarily relevant for MWCNTs in
upholstery).
eBased on the fact that textiles are known to lose 5-20% of their weight during use (abrasion, mechanical influence, washing, etc.);
authors note that textiles made from fibers with integrated nanoparticles are more likely to have longer lasting functionality
compared to those with nanoparticle surface coating or impregnation.
1-34
-------
Additional Information Highlight Box 3:
3. MWCNTs are likely used in combination with other chemicals in flame-retardant applications
Although MWCNTs demonstrate the ability to confer flame-retardant properties to a wide range of polymers and
textiles (Goncalvesetal., 2012: Grzybowski, 2009: Many 2009: Hewlett, 2008), nanomaterials such as MWCNTs are
likely not sufficient to pass flammability tests for composites and polymeric materials (e.g., UL 94, Limiting Oxygen
Index) (see Table 1-3 and Table 1-12 for example fire regulations and flame test performance, respectively) unless
combined with other flame-retardant agents (Morgan, 2006: Bartholmai and Schartel, 2004). A single study does
show, however, that, in at least one application (i.e., silicone-based foams), MWCNTs used alone could pass a
flammability test (Verdejo etal., 2008). In most instances nanomaterials, such as MWCNTs, likely would be combined
with other flame retardants to decrease flammability synergistically (Morgan, 2006). For example, Beyer (2006)
concluded that the addition of a microfiller (e.g., aluminum trihydrate) is essential to generate nanocomposites with
flame-retardant properties sufficient for industry and government standards. Isitman and Kaynak (2010) observed
similar synergies when carbon nanotubes were added to poly(methyl)methacrylate filled with an organophosphorus
flame retardant that acts through intumescence. Therefore, MWCNTs likely would be used in combination with
conventional fire retardants or other materials to pass flammability tests in most applications, including upholstery
textiles (Ullah and Ahmad, 2012: Im etal., 2011: Yu etal., 2011: Yu etal., 2009: Verdeio et al., 2008: Morgan, 2006:
Schartel etal., 2006: Bartholmai and Schartel, 2004). In support of this likelihood, additional examples of MWCNTs in
combination with traditional flame retardants are available in the literature (Ma etal., 2011: Luand Wilkie, 2010: Beyer,
1-35
-------
Table 1-12. Flame test performance of decaBDE and MWCNTs.
Sample
Without Treatment
With Treatment
Citation
MWCNTs3 Polyurethane Burns quickly, dense smoke forms, Spray coating forms "shell" to Mahy (2009):
foam burning droplets fall keep molten foam contained Hewlett (2008)
(no burning or flaming drops)
Polyvinyl Burns easily, melts, structure
chloride destroyed
Does not melt, structure is
retained
Mahy (2009)
Wires/Cables Burns completely, releases dense No burning droplets; low smoke Mahy (2009)
smoke and burning droplets, copper density; copper core protected;
core becomes exposed
passes UL94D, IEC-332-3ctests
Polypropylene Heat release rate = 2,800 kW/m 1-2% addition = heat release Gryzybowski
rate of 800 kW/m2 (2009)
Cotton Burned distance = 77 mm; burning
time = 19 sec; burning rate = 243
mm/min
Burned distance = 80 mm;
burning time = 21 sec; burning
rate = 229 mm/min
Goncalves et al.
(2012)
Polyester Burned distance = 66 mm; burning Burned distance = 66 mm; Goncalves et al.
time = 26 sec; burning rate = 152 burning time = 53 sec; burning (2012)
mm/min rate = 75 mm/min
DecaBDE Sofa
Burns quickly (<5 minutes), Burns slowly, increases amount BSEF (2012)
temperature increases from 20 °C to of time to escape by up to 15
800 °C times
50/50
polyester/
cotton twill
Sample burns completely
Char length reduced (<6.3 in.)
ICL Industrial
Products (2010)
aDue to the limited availability of information relevant to upholstery textiles, data for flame-retardant coatings for foam, polyvinyl
chloride, and other materials are included for additional context.
bHorizontal and vertical burning tests associated with American Society for Testing and Materials (ASTM) International standards.
large-scale flammability test for wire bundles under 20.5 kW flame.
1-36
-------
Table 1-13. Physicochemical properties of MWCNTs related to flame-retardant performance.
Property
Dispersion
Size
Functional
Groups and
Impurities
Aspect Ratio
Influence on Flame-Retardant Performance
Dispersion, which is influenced by surface chemistry, functionalization, or use of surfactants, plays
an important role in the flame-retardant properties of MWCNTs. For example, MWCNTs or SWCNTs
that were "well dispersed" in a polymer resulted in significantly reduced heat release rate compared
to "poorly dispersed" MWCNTs or SWCNTs (Kashiwaai et al., 2007: Kashiwaai et al., 2005b:
Kashiwaai et al., 2005a: Kashiwaai et al., 2004).
Commercially available MWCNTs for flame resistance are approximately 1.5 urn in length [Nanocyl;
(Howlett. 2008)1. Shorter MWCNTs (1-2 urn) are more flame-retardant than longer MWCNTs
(0.5-40 urn) in polymer blends (Packet al.. 2009).
MWCNTs that are "crushed" increase the time to ignition compared to uncrushed MWCNTs in
polymer nanocomposites (Laoutid et al., 2009).
MWCNTs coated with high density polyethylene are better dispersed and result in more
homogenous char formation compared to uncoated MWCNTs when producing nanocomposites
within an ethylene-vinyl acetate copolymer (Laoutid et al., 2009).
MWCNTs with triphenylphosphine functional groups have increased char production and flame
resistance compared to purified MWCNTs in experimental tests where the MWCNTs were not
incorporated into another material, such as polymer or foam (Muleia et al.. 2012).
Similar reduced heat release rates were reported for "crude" and "pure" MWCNTs in polymer
matrices, indicating that Co, Fe, and alumina contaminants on the crude MWCNTs do not interfere
with or enhance flame-retardant action (Beyer. 2004).
MWCNTs with higher acidity (highest quantity of surface oxygen-containing groups like carboxylic
acid, phenols, and carbonyls) have better flame performance than less acidic MWCNT treatments in
cotton and polyester (Conceives et al., 2012).
Chemical grafting of intumescent flame-retardant compounds onto the outer shell of MWCNTs
improves dispersion in polymer matrix and improves flame retardancy at lower mass loadings
relative to unmodified MWCNTs (Ma et al.. 2011).
Commercially available MWCNTs for flame resistance have an aspect ratio larger than 100
[Nanocyl; (Howlett. 2008)1.
CNTs significantly increase melt viscosity and reduce flammability of polymers due to high aspect
ratio and formation of jammed network (floccules layer) (Song et al.. 2012).
Increased aspect ratio improves flame retardancy (reduced heat release rate as evidenced by
increased storage modulus) at similar mass loadings (aspect ratio 49 versus 150) in polymer
matrices (Cipiriano et al., 2007).
1-37
-------
Table 1-13 (Continued): Physicochemical properties of MWCNTs related to flame-
retardant performance.
Property
Influence on Flame-Retardant Performance
Loading Loading concentrations of 0.1-0.5% are positively correlated with reduced heat release rate, as long
Concentration as the loading concentration does not cause the CNTs to be poorly dispersed (i.e., to agglomerate)
in polymers (Kashiwaqi et al.. 2005b: Kashiwaqi et al.. 2005a).
Loading concentrations of 0.5—4% MWCNT by mass show a shortened ignition delay time at 0.5%,
followed by an increased ignition delay time with increase in concentration up to about 1% and an
increase in peak heat release rate above 1% (loading; 2% and 4%). The ignition delay time and
peak heat release rate observed at 1% loading is reported to be due to the balance between the
effect of thermal conductivity and shielding performance of external radiant flux, although the authors
do not discuss why higher loading mass resulted in poorer performance in polymer matrices
(Kashiwaqi et al., 2004).
Mass loadings of 1-2% form solid protective layers with low mass loss rates; however, 4% loading of
the same MWCNTs have a higher peak mass loss rate. The authors attribute this to increased
thermal conductivity, which initially slows mass loss rate; but, once thermal energy accumulates in
the polymer sample, the mass loss increases more quickly (Cipiriano et al.. 2007).
Increased loading (from 100 ppm to 1,500 ppm) increases the LOI from 17.6% (raw cotton) to a max
of 23.8% (at 1,500 ppm), however concentrations of 250 ppm and above result in less dramatic
increases in LOI compared to the first 100 ppm (with an LOI of 22.2%) (Alimohammadi et al.. 2011).
1-38
-------
This page intentionally left blank.
-------
Chapter 2. Product Life Cycle
A product's life cycle encompasses all stages of its existence from "cradle to grave," starting with
the extraction of raw materials from the earth for the manufacture of the product and continuing
downstream until these materials are returned to the environment following disposal (U.S. EPA. 2006).
The components of the life cycle determine the potential for releases and possible impacts on human
health, ecological populations, and the environment (Som et al.. 2011). which can be evaluated
systematically within the framework of a comprehensive environmental assessment (CEA). Potential
environmental impacts of a product throughout its life cycle can be estimated using a life-cycle
assessment (LCA) approach, which involves four steps: goal definition and scope, inventory analysis,
impact analysis, and interpretation (U.S. EPA. 2006). The CEA approach incorporates information from
available LCAs in the "product life cycle" and "impacts" portions of the CEA framework to combine this
knowledge with other analyses or qualitative indicators related to transport, transformation, and fate,
exposure-dose, and additional impacts not considered in available LCAs. As discussed in Chapter 1. if a
plausible reason exists to include an impact in the CEA framework, information (qualitative or
quantitative) on that effect can be included from LCAs or other sources (if an LCA has not been
completed) to evaluate that particular impact.
A generalized depiction of the life cycle for multiwalled carbon nanotube (MWCNT) coatings
used to confer flame-retardant properties to upholstery textiles is presented along with comparable
information for decabromodiphenyl ether (decaBDE) in Figure 2-1. This figure breaks down the life cycle
of these materials into five main stages: (1) acquisition and processing of feedstocks, (2) manufacturing
(including research and development (R&D) processes), (3) storage and distribution, (4) use, and (5) end-
of-life processes (including disposal, reuse, and recycling). These stages correspond roughly to the four
primary life-cycle stages outlined by the United States Environmental Protection Agency (U.S. EPA.
2006). including raw materials acquisition, manufacturing, use/reuse/maintenance (with storage and
distribution discussed as a distinct stage in this case study), and recycle/waste management. As mentioned
in Chapter 1. R&D is included in the product life-cycle portion of the CEA framework, given its
importance regarding emerging materials such as MWCNTs. For such materials, R&D efforts can
elucidate potential risks associated with commercial-scale manufacturing. In fact, because it often takes
place when health and safety information is being developed for a material, R&D presents an ideal
opportunity to gather data on a product's potential impacts and to make design adjustments if appropriate.
2-1
-------
Life
Cycle
Release
Scenarios
DecaBDE
Release Form
CNT
Release Form1
Raw materials from which decaBDE
or MWCNTs are synthesized
c
a
a.
o
I
tt
45
Material Synthesis
Production of decaBDEsand
MWCNTs
Material Processing
Purification/modification of
decaBDE and MWCNTs
Product Manufacturing
Formulation and application of
flame retardant to textiles and
textile finishing and upholstering
_ I _
Material
Supplyof decaBDE and MWCNTS
Flame Retardant
Supplyof flame retardant
Product
Supplyof flame-retardant
upholstery textiles
Intended and unintended/
accidental use of flame-retardant
upholstery textiles
1
o
•a
Material
DisposalofdecaBDEand MWCNTS
Flame Retardant
Reuse, recycling, and disposal of
flame retardant
Product
Reuse, recycling, and disposal of
flame-retardant upholstery textile
Extraction
Processing
Synthesis
Recovery
Handling/packaging
Equipment cleaning
Accidents
Purification
Functionalization
Dispersal
Handling/packaging
Equipment cleaning
Accidents
Formulation
Application
Processing
Textile/furniture processini
Equipment cleaning
Accidents
Accidents
Free
Free
Free
Free
N/A
N/A
N/A
Free, Matrix-Bound
Matrix-Bound
Matrix-Bound
Matrix-Bound
Free, Matrix-Bound
Free, Matrix-Bound
Free, Bundled
Free, Bundled
Free, Bundled
Free, Bundled
Free, Bundled
Free, Bundled
Free, Bundled
Free, Bundled
Free, Bundled
Free, Bundled
Free, Bundled
Free, Bundled, Matrix-Bound
Matrix-Bound
Matrix-Bound
Matrix-Bound
Free, Bundled, Matrix-Bound
Free, Bundled, Matrix-Bound
Free, Matrix-Bound Free, Bundled, Matrix-Bound
Intended (sitting, cleaning, etc.)
U nintended/acci dents
Matrix-Bound
Free, Matrix-Bound
Matrix-Bound
Free, Matrix-Bound
Reuse/repurposing
Recycling
Incomplete incineration
Landfilling
Was tewater sludge
Matrix-Bound
Matrix-Bound
Matrix-Bound
Matrix-Bound
Matrix-Bound
Free, Matrix-Bound
Free, Matrix-Bound
Free, Matrix-Bound
Free, Matrix-Bound
Free, Matrix-Bound
Sources: Chaudhry et al. (2009): Kohler et al. (2008): Johnson et al. (2010): Zhou and Gong (2008): Som et al. (2011):
U.S. EPA (2005): Lassen et al. (1999): NRC (2000): Palm et al. (2002): Agrell et al. (2004): EU (EU. 2002).
Figure 2-1. Life-cycle stages, potential release scenarios, and forms of release for decaBDE and
MWCNTs in flame-retardant coatings applied to upholstery textiles.
1No data available that specifically describe release form of MWCNTs, thus, release forms of CNTs in general are discussed here.
Abbreviations: N/A = not applicable; CNT = carbon nanotube
2-2
-------
Similarly, as discussed below, differences between R&D activities and the commercial
manufacturing process (e.g., use of protective equipment, volume of material produced) could be
important considerations in mitigating potential risks to individuals involved in R&D versus commercial
manufacturing.
To conduct a comparative CEA, relevant information on life-cycle inventories from existing
LCAs would be incorporated into the product life cycle to characterize the inputs (e.g., raw materials,
solvents, processing reagents, energy) and outputs (e.g., emissions to air and water, coproducts)
associated with each material's manufacture. Impacts information from existing LCAs also would be
considered (see Chapter 5). Other LCA aspects also might apply, including using an appropriate
functional unit, which is a quantitative measure of a product's function or a process that facilitates
comparison (U.S. EPA. 2006). In the current case study, a functional unit might correspond to the degree
of flame retardancy conveyed by incorporation of a certain amount of MWCNTs. In general, for this case
study, data that specify appropriate functional units were not identified; the reader might, however,
consider how this aspect of existing or future LCAs could be incorporated into a future CEA when
evaluating data gaps and needs.
This chapter outlines important aspects of each of the five life-cycle stages outlined in Figure 2-1
for MWCNTs used in upholstery textiles. This chapter also includes descriptions of the important
environmental release scenarios for MWCNTs across the product life-cycle stages based on current
knowledge. A variety of release scenarios are possible throughout the life-cycle stages described in this
chapter. These are summarized in a series of tables that make qualitative comparisons between potential
MWCNT release and decaBDE release scenarios. While the release scenarios presented in these tables are
simplified due to a current lack of data, they help to highlight how a qualitative understanding of
decaBDE release might inform research planning focused on releases across the product life cycle of
MWCNT flame-retardant coatings in textiles. Figure 2-1 also outlines potential release scenarios for
MWCNT flame-retardant upholstery textile coatings throughout the life cycle along with potential forms
of the released substances (i.e., free, bundled, or matrix bound).
The term "free MWCNTs" refers to pure, unbound materials. The term "MWCNT bundles"
refers to clusters of MWCNTs loosely or tightly bound together.13'14'15 The terms "matrix-bound
13The term "bundle" is used to subsume aggregates, agglomerates, and other clusters of MWCNTs reported in the
supporting literature because of the inconsistency in usage and, more importantly, the frequent lack of adequate
information to determine which specific term might be more appropriately applied to a particular dispersion state
observed in a study or report. Where possible, this case study describes the relative characteristics of different
dispersion states (e.g., more loosely or tightly bundled, ropier or more entangled) and quantifies the differences
between these characteristics if this information is provided by the study authors. Footnotes continued next page.
2-3
-------
decaBDE" and "matrix-bound MWCNTs" refer to these materials as a part of a polymer matrix (e.g., the
flame-retardant formulation). Additionally, as illustrated in Figure 2-2. MWCNT formulations can be
altered at multiple stages of the product life cycle, meaning the formulation of MWCNTs released at
different stages of the product life-cycle can vary and may be further altered by environmental
transformations. As discussed in greater detail below, solvents and other reagents used during material
synthesis and processing and product manufacturing may also be released during the product life cycle.
While little information on the release of other substances during the product life cycle was identified, it
would be included, if available, in any future CEA applications on MWCNTs.
14MWCNTs may adsorb to dust particles (see Section 3.2). Although, as discussed in this chapter, adsorption to dust
could facilitate release (i.e., offer a transport vector) from a product matrix, the dust-MWCNT complex is not
considered a "release form" because dust was not included in the original product matrix.
15Other substances in the textile matrix (e.g., other flame-retardant materials) may be released along with MWCNTs,
but there is currently a lack of specific data on this topic. When available, information on the release of these
substances is discussed in this chapter.
2-4
-------
Product Storage, Transport
MWCNT
Physical/chemical changes that occur including: intentional
(e.g., productformulation) and unintentional (e.g.,
weathering) changes
Adapted from: Nowack et al. (2012
Figure 2-2. Variations in MWCNT formulations and functionalization along the product life-cycle.
Note: This figure illustrates how MWCNT formulations can be altered at multiple stages of the product life cycle. Alterations can
occur intentionally as part of product formulation, or unintentionally as the MVVCNTs move through the environment. Alterations
include both physical and chemical changes (e.g., the addition or removal of functional groups, changes in surface charge
reactivity, aggregation / agglomeration, physical shortening of tubes, or association with natural organic matter or contaminants).
Asa result of the multitude of changes that could occur as the MWCNT formulation moves through the product life cycle, risk
assessors and risk managers must consider not only the hazards associated with the original material, but also the hazards
associated with the various altered materials that could be traced back to the original material. In addition, the physicochemical
characteristics of the MWCNT formulation may differ depending on when release into the environment occurs (e.g., MWCNTs
released into the environment during the material manufacturing or processing, or the product manufacturing stage [blue arrow]
may differ from those released from products during use or disposal / recycling).
2-5
-------
2.1. Feedstocks
( Unprioritized
•_M£L T
pC
fl> V
i =s
m r
o
Q.
Three of
assessrre
consider
1.1. 3 for.
•
• •
• •
•
• •
• •
•
• •
a*
3/13 |£SS
iii _«
o
§--1
5/13
E-F
Research Area: Product Life Cycle
Volume
T
H
•
?,
E-RRF
Importance Rati
L M r
Release Rate
•
•
H M L H M L
IRF Confidence Rating E-RRF Confidence R
5/13
_ •- 1 participant s vote
Importance
~~ M - Medium: Possibly Important
H - High: Important
Confidence
L - Low: Not Confident
M - Medium: Somewhat Confident
H High' Confident
ating
13 RTI workshop participants (23%) identified the raw materials stage of the product life cycle as important to risk
3nt. Based on this information, this stage in the product life cycle was determined to be of lesser importance to
n a future risk assessment of MWCNTs, and all text relevant to these areas was moved to Appendix G. See Section
a detailed explanation of the prioritization process.
2.2. Manufacturing
The manufacturing stage for MWCNT flame-retardant upholstery can be viewed as a sequential
process involving synthesis, material processing (i.e., purification and modification), and product
manufacture (i.e., formulation of the flame-retardant mixture, application of the flame-retardant mixture
to textiles, and incorporation of the flame-retardant textile into consumer or commercial goods). R&D
also is included in this section, given the similarities to key aspects of synthesis, processing, and
manufacture.
2.2.1. Research and Development
Neutral Research Area: Product Life Cycle
The research and development stage of the product life cycle was not considered during the RTI collective judgment
prioritization process. This section of text is included in the main document because it supports an understanding of the priority
research areas of material synthesis and material processing (see Section 1.1.1 and Introduction to Chapter 2).
2-6
-------
2.2.1.1. Life-Cycle Processes
Research on MWCNTs and on flame-retardant coatings involving MWCNTs is principally
conducted in specialized laboratory environments. R&D activities are expected to be carried out by
individuals rather than automated mechanisms used in commercial-scale manufacture. The processes of
interest to researchers are similar to those used in commercial-scale manufacture of these materials:
synthesis, purification, modification, dispersion, incorporation into flame-retardant formulations, and
application to textiles. Current research efforts focus on synthesis and purification methods to improve
CNT quality and purity (Kohler et al.. 2008). The following sections (material synthesis, material
processing, and product manufacturing) provide detailed information on the processes of potential interest
for R&D.
2.2.1.2. Potential Releases during the R&D Stage
Release scenarios during the R&D stage are expected to be similar to release scenarios from
commercial synthesis described in the following sections, but the quantities released are anticipated to be
much smaller in the R&D stage. The quantities of MWCNTs handled in research laboratories are much
smaller than those handled in commercial-scale manufacturing facilities. Although R&D activities are
typically carried out in laboratories with specialized pollution control systems in place, including fume
hoods, ventilation systems, and environmental control systems, not all facilities have standardized
engineering controls. For example, these practices might not be in place for small start-up operations.
Given the experimental and somewhat unpredictable nature of R&D, releases from handling materials
during synthesis, processing and purification, storage, and analysis are possible.
As discussed in more detail in Section 4.1.2.4. multiple studies have collected particles and fibers
in workplace air to attempt to estimate MWCNT concentrations at the emission source, in area air, and in
the personal breathing zone of workers in small laboratories or research and development facilities
(Johnson etal.. 2010; Lee etal.. 2010a; Methner etal. 2010; Bello et al.. 2008; Han et al.. 2008). One
industry report measured airborne release of CNTs (0.25 ug/m3) in the R&D facilities for a company that
manufactures Thermocyl®, an MWCNT flame-retardant coating application (Luizi. 2009). No data were
found that describe how releases in academic labs compare with releases in commercial R&D labs,
though it is noted that estimates in Bello et al. (2008) were based on university laboratory settings.
Similarly, no data were identified quantifying potential releases of other substances (e.g., solvents,
reagents used in functionalization) during R&D; however, qualitative information on the types of
substances that might be released during CNT synthesis is discussed in the following section.
2-7
-------
2.2.2. Material Synthesis
Priority Research Area: Product Life Cycle
Material
Synthesis
1
o
0.
5/13
4/13
4/13
H M L
E-RRF Confidence Rating
H M L
E-RRF Confidence Rating
• - 1 participant's vote
Importance
L - Low: Least Important
M-Medium: Possibly Important
H - High: Important
Confidence
L - Low: Not Confident
M-Medium: Somewhat Confident
H - High: Confident
Five of 13 RTI workshop participants (38%) identified the material synthesis stage of the MWCNT-product life cycle as
important to risk assessment. These five participants were asked to rate the importance of volume and release rate at this
stage in the product life cycle as well as their confidence that the existing data on these subjects could support risk-
management decisions. Based on this information, release rate was deemed the highest priority for research in this area,
followed by volume. See Section 6.3.1.1 for more information on this priority area, and Section 1.1.3 for a detailed explanation
of the prioritization process.
DecaBDE Can Inform MWCNT Assessment
Releases of decaBDE and by-products or transformation products during material synthesis can be classified as incidental or
accidental. Larger production volumes can result in larger incidental release volumes and a greater potential for accidental
releases, and so the total production volume is important to consider (see Table 2-1 and Table 2-2 for overview of MWCNT
production techniques & volumes).
Incidental releases—such as release of decaBDE into the air as vapor from the reactor vessels is released or into wastewater
as equipment is cleaned—are minimal compared to the total volume of decaBDE produced, in part because large
manufacturing facilities typically use control mechanisms that limit release volume. One study found, for example, that
wastewater releases of decaBDE are unlikely to exceed 0.5 kg/ton if equipment is washed after every batch (EU, 2002).
Fugitive releases of decaBDE vapor from a reactor vessel have been estimated as 1.1 x 10~5 mg/ton, and releases from the
bagging of synthesized polybrominated diphenyl ethers (PBDEs) have been estimated as <70 grams/ton PBDE produced
[(EU. 2002): EEC (1993) as cited in EU (2002)1.
Accidental releases, such as leaks arising from faulty equipment or malfunctioning ventilation systems, can result in larger
environmental releases because control mechanisms like those for handling incidental releases are absent. Accidental
releases also might contain transformation products (e.g., polybrominated dibenzofurans (PBDFs) from high heat exposure to
decaBDE) that typically would not be included in incidental releases. Accidental releases, however, occur less often than
incidental releases.
Similar to decaBDE, MWCNT releases could be incidental or accidental. As shown in Table 2-3, decaBDE and MWCNTs can
be released to air during synthesis and purification, recovery, handling/packaging, cleaning, and accidents, while release to
water is generally limited to periods when equipment is being cleaned. Based on information for decaBDE, research planning
to inform future MWCNT risk assessments might consider the differences between the potential for, and implications of, both
incidental and accidental releases of MWCNTs during synthesis, including: Which is more frequent? Which would result in a
greater volume of compound entering the environment? Are systems in place to limit the occurrence of both? See Appendix H
for more information regarding release of decaBDE during material synthesis.
2-8
-------
2.2.2.1. Life-Cycle Processes
In 2010, chemical vapor deposition (CVD) synthesis produced approximately 83% of the global
supply of CNTs, followed by arc-discharge synthesis (12%) and laser-ablation synthesis (5%) (Patel.
2011) (see Table 1-10 for information on MWCNT global production capacity; see Additional
Information Highlight Box 2 for more discussion on CNT production levels). What proportion of
MWCNTs is synthesized using each method, however, is unclear. Table 2-1 summarizes some of the
performance characteristics of these three synthesis methods. Table 2-2 summarizes the current scale and
projected growth of the CNT manufacturing industry, with details on the percentage of companies using
each synthesis methods.
Table 2-1 . Summary of common CNT synthesis methods.
Characteristic3
Growth temperature
Production
Scalability
Product quality
CVDa
600-1,1 00 °C
Continuous
Scalable
Many structural defects
Arc discharge3
2,500-3,000 °C
Batch
Not currently scalable
Few structural defects
Laser ablation3
1,200°C
Batch
Not currently scalable
Few structural defects
Long tubes
Low crystallinityb
Short tubes
Carbon-containing metal
impurities
Diameter control
By-products
Over 45 side products,
including polycyclic aromatic
hydrocarbons and volatile
organic compounds
Black carbon and airborne
inorganic compounds
No Data
aNo data available that specifically describe MWCNTs.
bLow degree of structural ordering.
Abbreviations: CVD = Chemical vapor deposition
Sources: Li et al. (2010). Healy et al. (2008): Karthikeyan et al. (2009): Rafique and Iqbal (2011): Plata et al. (2009).
CVD synthesis takes place in two furnaces connected by a quartz tube (Healy et al.. 2008).
The catalyst mixture is heated before it is added to the furnace along with a carbon-containing gas
(Karthikeyan et al.. 2009; Healy et al.. 2008). CNTs are recovered once the furnaces cool to room
temperature (Karthikeyan et al.. 2009). MWCNTs can be grown on a substrate (e.g., Si/SiO2) or without a
substrate (Tsai et al.. 2009). CNTs produced using deposition substrates are recovered by automated or
manual mechanical removal (Kohler et al.. 2008): however, specific details regarding the method of
recovery were not identified.
2-9
-------
Table 2-2. Current scale and projected growth in the CNT industry.
Parameter
Number of Employees per Company3
Quantity of CNT Produced per Year
(kg/year)3
Employee Count
Year r
Projected
Industry
Growth
CNT Synthesis
Method Used
(% of
companies
using method)0
Employee Count Year 2
(% change from
Year1)b
Employee Count Year 3
(% change from
Year2)b
Total Percent Change
from Year 1 to Year 3b
CVD
Arch Discharge
Flame Combustion
Laser Ablation
Combined
Manufacture and
Pilot/Developmental Pilot/ Developmental
Manufacture Scale Scale Scale
2-100
0.2-2,500
172
196 (+14%)
214 (+9.2%)
+24%
62%
23%
15%
8%
1-30
0.1-300
20
43 (+115%)
62 (+44%)
+210%
NR
NR
NR
NR
1-130
0.1-2,800
192
239 (+24%)
276 (+15%)
+44%
NA
NA
NA
NA
aAt the time of the survey (Oct 2008 to May 2009) there were 61 companies manufacturing engineered carbonaceous nanomaterial
(ECN) or applying ECN in other manufacturing processes in the United States at full, pilot, or research scale with plans to scale up
within 5 years. Of these 61 companies, approximately 43 were CNT manufacturers. Approximately 59% of these 43 CNT companies
(i.e., 25) were at "full manufacturing scale," 11 % were pilot scale (i.e., 4), and 11 % were research and development scale with plans
to scale up (i.e., 4). The remaining ~ 18% (i.e., 10) did not participate so publicly available information regarding employee numbers
was used but data regarding production quantity were not.
bYear 1 was 2004 for nonparticipating companies (n = 5) and 2006 for participating companies (n = 26), therefore Year 2 was 2005
and 2007 and Year 3 was 2006 and 2008 for nonparticipating and participating companies, respectively.
°33% (i.e., 14) of manufacturers reported as "primary," 42% (i.e., 18) reported as "secondary," and 26% (i.e., 11) reported as both
primary and secondary manufacturers.
Abbreviations: NA = not applicable; NR = not reported
Adapted with permission of Wolters Kluwer Health, from Schubauer-Berigan et al. (2011).
Synthesis by arc discharge involves passing an electric current between two graphite electrodes
(Healy et al.. 2008) in the presence of an inert gas. The anode contains a hole filled with carbon powder
and a catalyst, and the electric current results in the vaporization of the graphite anode and subsequent
condensation on the cathode and the walls of the reaction vessel (Baddour and Briens. 2005). CNT
deposits form as black powder in the reaction vessel (Healy et al.. 2008). CNTs are generally recovered
2-10
-------
from a receptacle after arc-discharge synthesis (Kohler et al.. 2008). but whether recovery is usually a
manual or automated process is unclear.
Laser-ablation synthesis of MWCNTs involves vaporizing a metal-graphite composite block in
the presence of an inert gas and a catalyst (Karthikeyan et al.. 2009). The composite block is placed inside
an oven, a laser is pointed at the block, and argon gas is pumped parallel to the laser beam. As the laser
ablates the target at high temperatures, CNTs form and are carried by the gas flow onto a collector
(Karthikeyan et al.. 2009). See Sections 5.3.2 and 5.3.4 for information on the estimated energy
requirements and costs of CNT synthesis.
2.2.2.2. Potential Releases during the Material Synthesis Stage
The potential release of MWCNTs during synthesis primarily depends on the synthesis and
processing methods and the physical properties of the MWCNTs (e.g., size, bundling, density) (Kohler et
al.. 2008). Because few data were available that describe releases from commercial-scale manufacture of
MWCNTs, this section also relies on CNT release data from R&D facilities. Although releases of
MWCNTs could occur during the synthesis stage, evidence describing the likelihood and quantity of
release is mixed. In general, MWCNTs grown on substrates are likely to produce fewer airborne releases
than vapor-phase synthesis methods (Bello et al., 2009; Tsai et al., 2009). Mechanical removal (either
automated or manual) of CNTs from the substrate, however, can cause airborne release of CNTs (Kohler
et al., 2008). In one study, CVD synthesis, both with and without a substrate, resulted in a concentration
at the source of synthesis of more than 2 to 3 x 106 particles/cm3 that measured less than 560 nm in
diameter (Tsai et al.. 2009), but another study found no measurable airborne release of CNTs during
substrate-bound CVD growth of CNTs (Bello et al.. 2009). One study conducted in three commercial
facilities and four research laboratories concluded that during synthesis, processing, and product
manufacturing, nanoparticle releases occurred most frequently when opening the CVD vessel and when
preparing the catalysts (Lee etal.. 2010a). No data were found on potential releases from laser-ablation
synthesis of MWCNTs, but low levels of single-walled carbon nanotube (SWCNT) clusters were released
as aerosols during laser-ablation synthesis of SWCNTs in laboratory and field conditions (Maynard et al..
2004). This study intentionally agitated the SWCNTs, however, which would not be a normal component
of the synthesis stage.
Synthesis of MWCNTs might release even greater quantities of CNTs, by-products, and
feedstock materials than SWCNTs due to the larger quantities of precursor materials required for
synthesis (Tsai et al.. 2009). CNT powder generally comprises large bundles of CNTs and air release of
these bundles is likely to occur during operations involving agitation (e.g., scraping, shaking) (Fleury et
al.. 2011). but local exhaust ventilation substantially reduces airborne releases to the environment (Lee et
2-11
-------
al.. 2010a; Han et al.. 2008). Loose MWCNTs and the equipment used during synthesis are likely to be
cleaned up with a vacuum, which appears to be effective in reducing the airborne concentration of
nanoparticles (Lee et al.. 2010a). Vacuuming, rinsing, and changing dust filters and other cleaning and
maintenance activities also could result in subsequent release of MWCNTs to air or wastewater (Kohler et
al.. 2008V
MWCNTs released during synthesis can contain significant impurities (see Section 2.2.3.1). In
addition, by-products from materials used in synthesis can be released at multiple points during the
synthesis process (Plata et al.. 2009). Some by-products, such as phenol, can be formed from general
combustion processes used in CNT synthesis (Eckelman et al.. 2012). Air release of synthesis by-products
including polycyclic aromatic hydrocarbons and volatile organic compounds has been observed during
CVD synthesis of CNTs in the absence of engineering controls (Plata et al.. 2009): if employed, control
technologies would be expected to limit these releases.
Accidental releases could also occur during MWCNT synthesis. These accidental scenarios
include fugitive equipment leaks, malfunctioning ventilation systems, and exposure to fire and high heat.
MWCNTs will not necessarily be destroyed at high temperatures (i.e., those possible in accidental fires)
(Kohler et al.. 2008). which could lead to airborne release or creation of ash containing elevated levels of
MWCNTs (Chaudhry et al.. 2009). Additionally, damaged filters that collect MWCNTs could result in
airborne release during synthesis (Kohler et al.. 2008). Such accidental events, while unlikely, could
result in potentially large releases of MWCNTs to the environment.
Table 2-3 summarizes the anticipated potential release scenarios from the material synthesis stage
of MWCNTs. Information for decaBDE is provided for comparison, with more detailed information on
decaBDE available in Appendix H.
2-12
-------
Table 2-3. Potential release scenarios during material synthesis.
Information on release
Processes included in material
synthesis life-cycle stage
MWCNTs
DecaBDE
Synthesis and purification
Recovery of synthesized
substance
Occurs in closed vessel, but fugitive air Occurs in closed vessel, but fugitive air
emissions could occur; release emissions could occur
depends on synthesis method
Air release could occur during removal Air release could occur during bagging
from substrate and bagging
Handling/packaging
Air release could occur
Air release could occur
Equipment cleaning
Air and water release could occur
Air and water release could occur
Accidental releases (equipment
malfunction, etc.)
Air release could occur
Air release could occur
2.2.3. Material Processing
Priority Research Area: Product Life Cycle
Material
Processing
I1
1=
i«
« m
o
Q.
9/13
4/13
H M L
E-RRF Confidence Rating
Importance Rating
L M H
Release Rate
•
•
•
• •
•
• •
•
H M L
E-RRF Confidence Rating
• - 1 participant's vote
Importance
L - Low: Least Important
M — Medium: Possibly Important
H — High: Important
Confidence
L - Low: Not Confident
M — Medium: Somewhat Confident
H - High: Confident
0/13
Nine of 13 RTI workshop participants (69%) identified the material processing stage of the MWCNT-product life cycle as
important to risk assessment. These nine participants were asked to rate the importance of volume and release rate at this
stage in the product life cycle as well as their confidence that the existing data on these subjects could support risk-
management decisions. Based on this information, release rate was deemed the highest priority for research in this area,
followed by volume. See Section 6.3.1.2 for more information on this priority area and Section 1.1.3 for a detailed explanation
of the prioritization process.
DecaBDE Can Inform MWCNT Assessment
Although post-synthesis processing is an important part of the life cycle for MWCNTs, it does not occur for decaBDE and so is
not a consideration for assessment or risk management of decaBDE. Therefore, decaBDE does not provide useful information
that could be applicable to research planning to support future assessments of MWCNTs during material processing.
2-13
-------
2.2.3.1. Life-Cycle Processes
Material processing includes any modification of MWCNTs after synthesis and before
incorporation into a flame-retardant formulation. These modifications can include purification,
functionalization, and dispersal in solvents.
MWCNTs undergo physical and chemical processing before they are incorporated into flame-
retardant applications. Commercial MWCNTs that have not been purified can contain large amounts of
impurities, such as amorphous carbon, graphite, and encapsulated metallic particles (Hou et al.. 2008).
After synthesis, MWCNTs are typically purified using physical (e.g., flocculation, microfiltration,
centrifugation) or chemical (e.g., acid treatment) techniques (Hou et al.. 2008). Appendix C presents
various CNT purification methods and their efficacies in removing various classes of impurities.
By-products of purification techniques are expected to differ according to the technique used; one study
reported waste products of sodium hydroxide, ethanol, water, filtrate, and scrap membrane following
general purification of SWCNTs (Healv et al.. 2008).
One analysis found that samples of MWCNTs purified by the manufacturer contained metal
impurities ranging from 0.44 to 1.75 (wt%) (Ge et al.. 2011). After further purification with an acid
treatment, significant quantities of catalyst residues (e.g., cobalt, chromium, iron, manganese,
molybdenum, and nickel) remained (Ge et al., 2011). Additionally, trace noncatalyst impurities (mostly
transition metals such as cerium, gadolinium, holmium, etc.) were also found in purified CNTs, which the
authors theorize can form during production/post-purification processes or might be introduced as catalyst
synergists to improve the quality of synthesized CNTs (Ge etal. 2011). Additionally, the purification
process itself also can damage the CNTs by introducing structural defects (Gustavsson et al., 2011).
Before application, MWCNTs generally require surface runctionalization (Saeed. 2010).
Functionalization—the modification of materials by covalently or noncovalently attaching new molecular
components—can alter the physicochemical properties of MWCNTs dramatically (Kohler et al.. 2008;
Ma et al.. 2008; Hirsch and Vostrowsky. 2005). Often, the goal of ftinctionalization of MWCNTs is to
increase solubility,16 which facilities dispersion into solvents and polymers (Saeed. 2010). One example
of runctionalization consists of the covalent grafting of MWCNTs to traditional flame retardants (Maet
al.. 2008). Table 2-4 presents examples of MWCNT ftinctionalization. After ftinctionalization, MWCNTs
are often dispersed in water or organic solvents before they are incorporated into products (Saeed. 2010).
16While MWCNTs are not soluble materials in that they do not dissolve in solution, the term solubility is used to
refer to dispersibility in the case study. Both terms are used in the literature to describe MWCNTs.
2-14
-------
Most dispersion methods use dry mixing or liquid-phase sonication (agitation of particles with ultrasound
energy) to break up clumps and disperse MWCNTs in water or organic solvents.
2.2.3.2. Potential Releases from the Material Processing Stage
During the recovery, processing, handling, and packaging stages, CNTs are more likely to be
released as bundles from bulk powder than as individual CNTs (Kohler et al.. 2008) (see Table 2-3). In
general, releases resulting from liquid-phase processing of CNTs will be lower relative to those resulting
from dry handling, which can result in greater nanoparticle release (Kohler et al. 2008). What proportion
of MWCNT processing occurs in the liquid versus the dry phase, however, is unclear. Handling of
MWCNTs can cause airborne release of particles (Methner et al.. 2010). Dispersal of MWCNTs in
suspensions can reduce the likelihood of aerosolization (Johnson et al.. 2010). but mixing and sonicating
(common processes used to disperse MWCNTs in solution) might also result in airborne release of raw
and functionalized MWCNTs (Johnson etal.. 2010: Lee etal.. 2010a: Methner etal.. 2010).
Environmental control mechanisms likely would be in place to reduce environmental releases in facilities
that process MWCNTs (Fleury etal.. 2011: Methner et al.. 2010). Cleaning of processing equipment and
facilities can lead to release of MWCNTs to air or wastewater (Fleury etal.. 2011). MWCNTs released
during this stage might still contain some of the impurities listed in Section 2.2.3.1 and Appendix C.
2-15
-------
Table 2-4. Examples of functionalization of MWCNTs.
Functionalization
technique
Goal of functionalization
Additional reagents
Citation
Amidation - Formation of
carbon nanotube-acyl
amides
Creation of anchor groups for
further modification
thionyl chloride,
dicyclohexylcarbodiimide
Hirsch and
Vostrowsky (2005)
Fluorination
Solubility in polar solvents
elemental fluorine
Hirsch and
Vostrowsky (2005)
Chlorination
Solubility in polar solvents
chlorine gas
Hirsch and
Vostrowsky (2005)
Noncovalent exohedral
functionalization
Solubility in polar solvents
streptavidin
Hirsch and
Vostrowsky (2005)
Covalent grafting on Better dispersion in matrix;
intumescent flame retardant solubility and stability in polar
solvents; enhanced network
structure at very low nanotube
loading
poly(diaminodiphenyl methane Ma et al. (2008)
spirocyclic pentaerythritol
bisphosphonate)
Atom transfer radical
polymerization
In situ surface reversible
addition-fragmentation chain
transfer polymerization
Electrografting
Radiation polymerization
Liquid- and gas-phase
oxidization with thermal
treatment
Creation of anchor groups for
further modification
Solubility in polar solvents
Solubility in polar solvents
Solubility in polar solvents
Increased acidity or alkalinity;
improved flame-retardant
properties
styrene and methyl
methacrylate
styrene and N-
isopropylacrylamide
polyacrylonitrile
ethanol, poly(acrylic acid),
acrylic acid
nitric acid
Baskaran et al.
(2004)
Xu et al. (2007)
Petrov et al. (2004)
Chen et al. (2006)
Goncalves et al.
(2012)
Release of by-products from CNT processing also might occur. One modeling study notes the
potential for release of phenol from the production of nitric acid, a substance commonly used in CNT
purification (Eckelman et al.. 2012). No additional information was identified on the release of other
substances used during MWCNT purification. As discussed in Section 5.3.2. initial data suggest that the
impact of substances released during CNT purification is lower than that of those released during CNT
synthesis (Eckelman et al.. 2012).
The accidental release scenarios for MWCNTs during processing are similar to those in the
material synthesis stage (see Section 2.2.2.2). Additionally, spills of solutions containing dispersed
MWCNTs might occur during the processing stage, which could result in the release of MWCNTs to
2-16
-------
wastewater. Table 2-5 summarizes potential release scenarios from the material processing stage of
MWCNTs.
Table 2-5. Potential release scenarios during material processing of MWCNTs.
Processes included in material
processing life-cycle stage
Information on release
Purification
Air release possible due to physical purification methods such as
flocculation, microfiltration, centrifugation, etc. (see Appendix C)
Functionalization
Air and water release possible during functionalization reactions; release
depends on method of functionalization
Dispersion
Air release possible, especially during dry mixing
Handling/packaging
Air release possible during dry processes
Equipment cleaning
Air and water release possible
Accidental releases (spills, equipment
malfunction, etc.)
Water release possible from spills in liquid-phase reactions and once
MWCNTs are dispersed in solution; air release possible from fugitive
emissions
2.2.4. Product Manufacturing
Priority Research Area: Product Life Cycle
. roduct
Manufacturing
O)
O '
o
Q.
£.
• • •
• • *
• • •
• • •
12/13
1/13
( Release Rate
H M L
E-RRF Confidence Rating
H M L
E-RRF Confidence Rating
• - 1 participant's vote
Importance
L - Low:
M- Medium
H - High:
Confidence
L - Low:
M- Medium
H - High:
Least Important
Possibly Important
Important
Not Confident
Somewhat Confident
Confident
0/13
Twelve of 13 RTI workshop participants (92%) identified the product manufacturing stage of the MWCNT-product life cycle as
important to risk assessment. These 12 participants were asked to rate the importance of volume and release rate at this
stage in the product life cycle as well as their confidence that the existing data on these subjects could support risk-
management decisions. Based on this information, release rate was deemed the highest priority in this area, followed by
volume. See Section 6.3.1.3 for more information on this priority area and Section 1.1.3 for a detailed explanation of the
prioritization process.
2-17
-------
DecaBDE Can Inform MWCNT Assessment
The volume of decaBDE used in textiles depends on the type of fabric used (e.g., 30-40 grams/m3 in cotton to 70-80
grams/m3 for velour fabrics) (EU, 2002), which also influences the volume of impurities and compounds used in conjunction
with decaBDE (e.g., antimony trioxide).
The potential releases of decaBDE and other impurities during product manufacture are similar to those discussed in the
material synthesis stage (see the DecaBDE Comparison Box on Material Synthesis in Section 2.2.2), but with some notable
differences. First, any decaBDE released during product manufacture is likely to be matrix-bound. Second, the release rate of
decaBDE during the product manufacturing stage is smaller than the release rate during material synthesis. Formulation of
flame retardants generally occurs in closed systems with engineering controls that regulate temperature and pressure to
minimize potential releases; however, processes that occur before or after the substance enters the closed system could result
in environmental release. Indeed, one study found that environmental release was most likely to occur during the mixing of
decaBDE powder and cleaning operations of the flame-retardant formulation stage (EU, 2002).
Similar to observations with decaBDE, MWCNTs likely would be released in matrix-bound form in smaller quantities during the
product manufacturing stage compared to the material synthesis stage. As such, understanding the life cycle and releases of
decaBDE flame-retardant coatings can aid efforts to characterize the life cycle and potential releases of MWCNT flame-
retardant coatings more fully.
Considerations for MWCNT research planning to inform future risk assessments might include additional variables that are
unique to MWCNTs based on post-synthesis processing (e.g., purification by-products, surface functionalization) (see Section
2.2.3.1 and the DecaBDE Comparison Box on Material Processing). For example: What is the relative production volume of
MWCNT flame-retardant formulations with different physicochemical characteristics? Do different MWCNT chemistries,
application methods to textiles, and textile characteristics increase or decrease the volume used in textiles, or the rate of
release during product manufacturing? At what stages of product manufacturing are releases of free or matrix-bound
MWCNTs most likely to occur? See Appendix H for more information regarding release of decaBDE during product
manufacture.
In this section, product manufacturing for MWCNTs is described. This life-cycle stage is
considered to include the manufacture of flame-retardant formulations, the manufacture of textiles
containing MWCNT-based flame retardants, and the manufacture of end-use products containing flame-
retardant materials, such as furniture.
2.2.4.1. Life-Cycle Processes
CNTs can be dispersed in polymers by in situ polymerization or by using a twin-screw extruder (a
specialized machine using two screws to mix, compound, and react polymers) [Laxminarayana and Jalili
(2005) as cited in Kohler et al. (2008)1. One study described the manufacture of a "high heat"
acrylonitrile-butadiene-styrene polymer matrix filled with CNTs (Fleury etal. 2011). According to this
study, master batch granules (1-4 mm long) consisting of thermoplastic resin and 15% CNTs by weight
are mixed with the pure polymer matrix in injection molding and extrusion processes (where
thermoplastics are fed into a heated vessel and forced into a mold cavity where they cool) (Fleury et al..
2011). Little information is available regarding the ingredients and characteristics of MWCNT flame-
retardant formulations and possible by-products from their manufacture. Thermocyl®, an MWCNT flame
2-18
-------
retardant, includes silicone resins containing MWCNTs (Luizi. 2009). In another formulation, MWCNTs
have been used in place of ammonium polyphosphate, a traditional phosphorus-based flame retardant, as
a filler with polymethyl methacrylate (PMMA) and polyamide-6 (Motzkus et al.. 2012). General
components of MWCNT flame-retardant formulations include epoxies, polyesters, and vinylesters
(Alberding et al.. 2011).
MWCNTs can be applied to textiles as a flame-retardant coating by (1) soaking the textile or
(2) spray coating the surface of the textile (Luizi. 2009). The soaking application method for MWCNTs is
similar to that used to apply decaBDE flame-retardant coatings (see Figure 1-4). In a recent study,
researchers immersed cotton and polyester textiles in an MWCNT dispersion and maintained constant
motion to embed functionalized MWCNTs in the textiles (Goncalves et al.. 2012). The authors reported
acetic acid, sodium chloride, sodium carbonate, and sodium hydroxide as auxiliary reagents for this
embedding process (Goncalves et al.. 2012). An alternative immersion method involves soaking a textile
in a solution containing CNTs and then treating it with a crosslinking agent (similar to a binder), heating,
rinsing, and drying the textile (with or without heat) (Alimohammadi et al.. 2011). Uddin and Nyden
(2011 a) and Davis and Kim (2010) described a similar immersion method involving multiple treatments
with polymeric solutions to create a coating consisting of layers of polyacrylic acid/MWCNT-
polyethylenimine/ polyethylenimine. Flame retardant-treated materials also can be cured using ultraviolet
(UV) radiation (Luetal.. 201 la). Lee et al. (2010a) mentioned the spray application of CNTs in solution
to thinly coat wafers, but they did not describe the process in detail. The few laboratory-scale studies
investigating MWCNT flame retardants in textiles have reported MWCNT loadings ranging from 0.5 to
4% by mass (Grzvbowski. 2009; Kashiwagi et al., 2005b; Kashiwagi et al., 2005a; Kashiwagi et al.,
2004). which are about an order of magnitude lower than those for decaBDE. Thermocyl®, a commercial
MWCNT flame retardant, has been incorporated into polyethylene at a loading of 1% (Luizi. 2009). After
application, the MWCNT flame-retardant upholstery is cut, shaped, and glued or stapled to furniture.
2.2.4.2. Potential Releases during Product Manufacture
Environmental releases during the manufacture of MWCNT flame-retardant textiles can occur as
a result of the following activities: mixing, handling/packaging, application of the flame retardant to
textiles, textile processing/finishing, and accidents (Zhou and Gong. 2008). Few data are available that
describe releases from commercial-scale manufacture of MWCNT flame-retardant textiles. Therefore,
this section also relies on CNT release data from R&D facilities. Release of MWCNT bundles is possible
when nanotubes are blended with polymers to formulate the flame retardant; however, releases are
anticipated to be smaller for blending of master batches as opposed to blending of pure CNT powders
(Fleury et al.. 2011). The most critical phase for air and water releases during the formulation stage is the
2-19
-------
discharging and the cleaning of the mixing chamber (Fleury etal.. 2011). Release from the formulated
flame retardant is expected to be minimal, but packaging of the formulated flame retardant could result in
releases. Airborne releases of CNTs (1.45 ug CNT/m3) were measured in a facility that packages
Thermocyl®, an MWCNT flame-retardant coating (Luizi. 2009).
Application of MWCNT flame retardants to textiles also could lead to air or water releases.
Currently, how a particular production method used to incorporate MWCNT flame retardant in the textile
matrix might influence release potential and subsequent exposure potential is unknown. Any spray
application could lead to the potential airborne releases of matrix-bound MWCNTs if the application does
not occur in a closed environment. One study observed airborne release of both nanoparticles and fine
particles when spraying an MWCNT solution onto wafers as a coating (Lee etal.. 2010a). The dominant
particle size released during this study ranged from 50 to 110 nm and subsequent heating of the treated
wafers also led to the release of particles smaller than 30 nm (Lee etal.. 2010a). If the MWCNT flame-
retardant coating is applied by soaking the textile, water release of matrix-bound MWCNTs could occur
when the textile is rinsed. Additional cutting, sewing, shaping, stapling, and other textile finishing
processes could result in the airborne release office or matrix-bound MWCNTs through abrasion (Kohler
et al., 2008). One study found that grinding a nanocomposite containing CNTs created a substantial
amount of airborne particles made up of polymer fragments containing CNTs (Fleury etal.. 2011). How
applicable release from grinding of plastic nanocomposites is to release from textile applications,
however, is unclear. Airborne releases of CNTs (1 ug CNT/m3) have been measured in a facility
processing textiles treated with Thermocyl® (Luizi. 2009). Most manufacturing facilities, however, would
be expected to have controls in place to prevent or minimize airborne releases to the environment.
Equipment cleaning at any point in this life-cycle stage could lead to release of MWCNTs to wastewater.
Not enough is known about the other components of MWCNT flame-retardant formulations to
hypothesize about other substances that might be released as a part of the flame-retardant product matrix,
but the use of different substances in MWCNT flame-retardant formulations might lead to different
release characteristics.
2-20
-------
The accidental release scenarios for
MWCNTs during product manufacture are
similar to those in the material synthesis stage
(see Section 2.2.2.2). In the product
manufacturing stage, spills of MWCNT flame-
retardant formulation might also occur and
could result in the release of MWCNTs to
wastewater. Table 2-6 outlines potential release
scenarios from the product manufacturing stage
of MWCNT flame-retardant textiles along with
decaBDE for comparison.
Additional Information Highlight Box 4:
4. MWCNT release from various product matrices
informs predictions of release from textiles
Although little information describing release of MWCNTs from
flame-retardant textile coating applications was found in the
literature, information on similar applications might help inform
future assessments of release potential during product
manufacture. For example, Takaya et al. (2012) studied the
likelihood of MWCNT release from MWCNT-coated yarn during
the weaving process used to produce conductive fabric.
The authors concluded that, although the likelihood of release
of individual MWCNT fibers during the weaving process was
low, micron-sized yarn particles containing MWCNTs were
released into the air around the weaving loom. The authors
hypothesized that this release occurred as a result of the
mechanical forces applied to the yarn during weaving. These
findings suggest that the weaving processes used to produce
textiles coated with MWCNT flame retardant might lead to
similar releases of micron-sized textile particles containing
MWCNTs. The potential for release of MWCNTs or matrix-
bound MWCNTs from textiles during other stages of the
product life cycle (see Figure 2-1) is similarly unknown.
Wohlleben et al. (2013: 2011) evaluated release scenarios that
could occur during other stages of the product life cycle (e.g.,
weathering) for cement paste or thermoplastics (polyurethane
or polyoxymethylene) with MWCNTs incorporated into the
polymer. Results indicate that the majority of release occurs in
micron-size fragments containing MWCNTs, similar to what
was observed by Takaya et al. (2012) for release from
MWCNT-coated yarn. Wohelleben et al. (2013: 2011) also
found that (1) the type of product matrix (e.g., cement paste or
polyurethane) influences the size distribution and concentration
of released fragments, and (2) the diameter of released
fragments generally increased when MWCNTs were included
in the matrix compared to when they were not; however, the
number of airborne particles was generally similar regardless of
whether or not MWCNTs were incorporated. Whether and the
extent to which these data are applicable to MWCNT release
from textiles is unknown.
2-21
-------
Table 2-6. Potential release scenarios during product manufacturing.
Information on release
Processes included in product
manufacturing life-cycle stage
MWCNTs
DecaBDE
Formulation of flame retardant
Air release possible when mixing dry
MWCNTs into product formulation
Air release possible when mixing dry
decaBDE into product formulation
Handling/packaging
Air release possible
Air release possible
Flame retardant application to
textile
Air or water release possible
depending on application method; for
example, if substances are sprayed
onto textiles, release to air can occur
Air or water release possible
depending on application method
Thermal processing
Rinsing/drying
Equipment cleaning
Preliminary evidence on air release
due to high heat is mixed
Water release possible
Air and water release possible
Air release possible
Water release possible
Air and water release possible
Textile processing
Air release possible due to cutting,
sewing, shaping, and other finishing
processes
Air release possible due to cutting,
sewing, shaping, etc.
Furniture production
Air release possible due to cutting,
stapling, and other finishing processes
Air release possible due to cutting,
stapling, etc.
Accidental releases (e.g., spills,
equipment malfunction)
Air and water release possible
Air and water release possible
2-22
-------
2.3. Storage and Distribution
Unprioritized Research Area: Product Life Cycle
Product Storage,
Transport
gl
LJ
o
Q.
E
2/13
1/13
Release Rate
• - 1 participant's vote
Importance
L - Low:
Least Important
M-Medium: Possibly Important
H M L
E-RRF Confidence Rating
H M L
E-RRF Confidence Rating
H - High:
Confidence
L - Low:
M - Medium:
H - High:
Important
Not Confident
Somewhat Confident
Confident
10/13
Two of 13 RTI workshop participants (15%) identified the product storage and transport stage of the MWCNT product life cycle
as important to risk assessment. Based on this information, storage and transport were determined to be of lesser importance
to consider in a future risk assessment of MWCNTs, and all text relevant to these areas was moved to Appendix G. See
Section 1.1.3 for a detailed explanation of the prioritization process.
2.4. Use
Priority Research Area: Product Life Cycle
1
LJ r
O
a.
9/13
3/13
1/13
H M L
E-RRF Confidence Rating
H M L
E-RRF Confidence Rating
• — 1 participant's vote
Importance
L - Low:
M- Medium
H - High:
Confidence
L — Low:
M - Medium
H - High:
Least Important
Possibly Important
Important
Not Confident
Somewhat Confident
Confident
Nine of 13 RTI workshop participants (69%) identified the Use stage of the MWCNT-product life cycle as important to risk
assessment. These nine participants were asked to rate the importance of volume and release rate at this stage in the product
life cycle as well as their confidence that the existing data on these subjects could support risk-management decisions. Based
on this information, these areas were deemed a priority for research. See Section 6.3.1.4 for more information on this priority
area and Section 1.1.3 for a detailed explanation of the prioritization process.
2-23
-------
DecaBDE Can Inform MWCNT Assessment
Information on volume during the product use phase for decaBDE must be inferred from production volumes reported by
manufacturers; although production volumes are expected to decrease due to the withdrawal of decaBDE from the market
(see Section 1.3.1), its volume during the product use phase likely greatly exceeds that of MWCNTs due to both higher historic
production volumes (see Table 1-10) and larger volumes incorporated into textiles (see Table 1-11).
Studies measuring the concentration of BDE-209, the single isomer of decaBDE, in building dust and indoor/ambient air are
presented in Appendix Sections H.4.1.2 and E.1. The studies show relatively high concentrations of decaBDE in dust of
buildings and homes containing products treated with decaBDE. These concentrations are primarily attributed to the sorption
of decaBDE to particles and dust in these settings (see the DecaBDE Comparison Box on Transport, Transformation, and
Fate in Air in Section 3.2). DecaBDE-treated plastics are expected to have the greatest release potential during the use phase
of the product life cycle (Lassen et al., 1999): in textiles, the frequently used decaBDE/antimony trioxide flame-retardant
formulations result in improved durability (i.e., reduced likelihood of release) due to the copolymer resin that bonds to the
textile fibers [(Pure Strategies Inc., 2005): see Appendix H.2.4.21. Nevertheless, some estimates indicate that the principal
source of decaBDE release in wastewater is due to textile washing during the product use phase (EU, 2002). Regular use of
upholstered furniture is also expected to result in wear and tear or abrasion that could result in the release of small amounts of
free or matrix-bound decaBDE.
In general, decaBDE release scenarios during the product use phase are similar to those anticipated for MWCNTs due to the
similar application in question. For example, similar to decaBDE, MWCNTs are not expected to be released due to their
stability and lack of degradation. As the product matrix degrades during normal use or is washed, however, MWCNTs could be
released in either the free or matrix-bound form.
Based on information for decaBDE, research planning to inform future risk assessments of MWCNTs might consider: Whether
the use of MWCNTs in combination with other chemicals or materials (similar to decaBDE/antimony trioxide example) results
in increased or decreased rate of release from textiles. Will differences in physical-chemical properties and mechanism of
flame-retardant action increase or decrease likelihood of release during exposure to high heat or fire? How does aging or
weathering influence the potential release of MWCNTs from the textile matrix? What type of activities (e.g., washing) during
the product use phase might result in the most frequent or greatest environmental releases? See Appendix H for more
information regarding release of decaBDE during the use stage of the product life cycle.
2.4.1. Life-Cycle Processes
A wide variety of textiles contain flame-retardant coatings (see Section 1.2). Upholstery textiles
are expected to be used in public places where people of all ages will sit, lie, or walk on them. Some
unintended uses of upholstery textiles include outdoor use, repurposing for use in other products, burning
as kindling, or mouthing by children. Repurposing for use in other products and burning as kindling are
covered in Section 2.5 and Appendix Section H.2.5. In general, upholstery textiles are likely to have a
lifespan of at least 10 years (EU. 2002).
2-24
-------
2.4.2. Potential Releases during the Use Stage
Environmental releases from upholstery textiles coated with flame retardants are influenced by
(1) the potential use scenarios for the upholstery textiles and (2) the physicochemical properties of
MWCNTs. The anticipated long lifespan of upholstery textiles (>10 years) suggests that releases in this
stage could occur over several years (EU. 2002). Although no concentration data resulting from consumer
use are available for MWCNTs, the following characteristics of flame-retardant upholstery textiles are
expected to reduce releases MWCNTs (EU. 2002):
• Flame-retardant coatings must meet durability requirements to comply with regulations (see
Section 1.2.1):
• Flame retardant is often applied to the back of the fabric, minimizing wear and tear; and
• Upholstery textiles are unlikely to be washed frequently.
The integrity of the flame-retardant coating depends on the strength of the formulation that bonds
it to the textile surface (Sometal.. 2011: NRC. 2000). MWCNT flame-retardant textile coatings
considered in this case study are additive, suggesting that release from upholstery textiles could occur
during the use stage (see Section 1.2.2.2). While in general, CNTs are not likely to be released because
they are very stable and do not readily degrade, they could be released if the polymer matrix degrades
(Kohler et al.. 2008). or they could be released as a component of the polymer matrix. Factors that could
lead to MWCNT release from textiles include:
• Regular use of upholstered furniture (e.g., sitting, walking, lying) could abrade the textile
surface and release small amounts of free or matrix-bound MWCNTs either into the air or
onto the skin of users. However, early unpublished evidence presented at a public meeting
indicates that very small amounts of MWCNTs could be released as aerosols after subjecting
an MWCNT-polymer nanocomposite to simulated wear and tear tests (Uddin and Nyden.
201 Ib).
• Washing of textiles also could lead to water release of matrix-bound MWCNTs.
• Even though CNTs might be embedded in a matrix, depending on the production method, a
portion of a tube or a group of tubes could be left partially exposed in the final product.
Importantly, although most releases initially will be to the indoor environment, they could spread
outdoors through environmental transport mechanisms (see Chapter 3). In contrast, factors that suggest
minimal MWCNT release from textiles include:
• Most flame-retardant upholstery textiles will be used indoors, minimizing exposure to UV
light and weathering.
2-25
-------
• Upholstery textiles that are back-coated (a common application method) with MWCNT flame
retardant likely will not be subject to significant abrasion, washing, or UV light.
Although these processes also could result in release of MWCNTs to the air or to wastewater, less
degradation of upholstery textiles is expected for MWCNT-treated composites than for traditional
materials (see Section 1.3.3). For example, Nguyen et al. (2011) found that exposure of an epoxy
containing MWCNTs to UV radiation (295 to 400 nm) under conditions of controlled temperature (50 °C)
and humidity (75% relative humidity) can cause MWCNTs to form a dense network on the surface of
composites, which might minimize environmental release. The authors also found that the epoxy
containing MWCNTs degraded more slowly than unfilled epoxy or an epoxy containing another
nanoscale material (Nguyen et al.. 2011). A similar study found that when PMMA is filled with silane-
coated MWCNTs the amount of submicrometric airborne particles emitted decreased when the polymer is
exposed to fire compared to pristine PMMA (Motzkus et al.. 2012). The authors noted that the release of
airborne particles depends on a variety of factors, including the type of polymer matrix, the combustion
process, and the type of surface treatment (Motzkus et al.. 2012). A lack of data precludes a determination
of whether similar MWCNT releases could be expected to occur in textiles that contain them. Similarly,
no information was identified on whether or the extent to which other substances (e.g., surface coatings)
might be released with MWCNT flame-retardants from textiles during product use.
Unintended uses also could lead to the release of MWCNTs from flame-retardant textiles. Use of
flame-retardant upholstery textiles outdoors could lead to weathering, which could degrade the polymer
matrix resulting in a release. Mouthing by small children, pets, or rodents on flame-retardant textiles
could lead to release directly into the mouths of children, pets, or rodents if the back-coating is exposed
and the integrity of the fabric is compromised. Preliminary unpublished evidence presented at a public
meeting, however, suggests that few MWCNTs are released from a flame-retardant nanocomposite when
subjected to simulated chewing tests (Uddin and Nvden. 201 Ib). Accidental contact of flame-retardant
textiles with fire and high heat also could occur and would lead to possible airborne releases (see Section
2.2.2.2 for more details). No data were found, however, that describe the likelihood of this release from
this application. Table 2-7 outlines potential release scenarios from the use stage of MWCNT flame-
retardant textiles. Similar data on decaBDE are provided in the table for comparison; more detailed
information on decaBDE can be found in Appendix H.
2-26
-------
Table 2-7. Potential release scenarios during product use.
Information on release
Processes included in use
life-cycle stage
MWCNTs
DecaBDE
Intended use (e.g., sitting,
standing)
Release possible due to abrasion or other
physical/mechanical activities, resulting in
direct release to individual in contact with
the textile or to air as particle-bound
substance (e.g., due to abrasion, release
to dust)
Release possible due to abrasion or other
physical/mechanical activities, resulting in
direct release to individual in contact with
the textile or to air as particle-bound
substance (e.g., due to abrasion, release
to dust)
Cleaning
Water release possible, but infrequent Water release possible, but infrequent
Unintended use (outdoor
use)
Air release possible due to
weathering/degradation of the polymer;
outdoor use could result in release to
water or soil
Air release possible due to
weathering/degradation of the polymer;
outdoor use could result in release to
water or soil
Unintended use (mouthing)
Direct release to mouth likely if polymer
matrix surface is accessible and degraded
Direct release to mouth likely if polymer
matrix surface is accessible and degraded
Accidental releases
Preliminary evidence suggests that air
release due to high heat is unlikely
Air release possible due to exposure to
high heat or fire
2.5. Reuse, Recycling, and End of Life
Priority Research Area: Product Life Cycle
Disposal/
Recycling
O)
gl
-J
a> o>
E £E
* i
uj r
o
o.
10/13
3/13
0/13
Volume
•
•
•
•
•
E-RRF
Importance Rating
L M H
Release Rate
•
• •
•
•
• •
•
• •
H M L
E-RRF Confidence Rating
H M L
E-RRF Confidence Rating
• — 1 participant's vote
Importance
L - Low:
M — Medium:
H - High:
Confidence
L — Low:
M - Medium:
H - High:
Least Important
Possibly Important
Important
Not Confident
Somewhat Confident
Confident
Ten of 13 RTI workshop participants (77%) identified the disposal/recycling stage of the MWCNT-product life cycle as important
to risk assessment. These 10 participants were asked to rate the importance of volume and release rate at this stage in the
product life cycle, as well as their confidence that the existing data on these subjects could support risk-management decisions.
Based on this information, these areas were deemed a priority for research. See Section 6.3.1.5 for more information on this
priority area and Section 1.1.3 for a detailed explanation of the prioritization process.
2-27
-------
DecaBDE Can Inform MWCNT Assessment
No information specific to the volume of decaBDE, decaBDE flame-retardant formulations, or decaBDE-treated upholstery
textiles that are disposed of or recycled was identified. Nevertheless, volumes of decaBDE and decaBDE flame-retardant
formulations are expected to decrease as the product is phased out of use. Disposal, recycling, and reuse of decaBDE treated
textiles might continue for some time, however, given that upholstery textiles have a lifespan of at least 10 years (EU, 2002).
The disposal, recycling, and reuse volumes of MWCNTs, MWCNT flame-retardant formulations, and MWCNT flame-retardant
upholstery textiles are likely to be much smaller than for decaBDE (see Table 1-10 and Table 1-11), but the potential for
release of MWCNTs generally is expected to be similar to that of decaBDE during such disposal, recycling, reuse, and
repurposing of flame-retardant textiles.
Recycling
Although disposal or recycling of decaBDE and decaBDE flame-retardant formulations is expected to be minimal and only
occur when manufacturing facilities are cleaned, research from a plastic recycling plant suggests that some release of
decaBDE in the product matrix can occur during the recycling and disposal of plastics (Sjodin etal., 2001). Although releases
from recycling of upholstery textiles containing decaBDE flame retardant might be similar to those of plastics, the processing of
plastics differs from that of textiles.
Landfilling
Although no information quantifying the release of decaBDE from the landfilling of flame-retardant textiles was identified,
release to air, soil, or water is possible (Rahman et al., 2001: Lassen etal., 1999). Such releases are expected to be small,
however, based on key physicochemical properties, including low volatility and low leaching potential [Kim et al. (2006) as
cited in Wright et al. (2008) and Palm et al. (2002): see Table 2-10 and Table H-41.
l/l/asfei/rafer Treatment
The removal efficiency of decaBDE in wastewater treatment plants is not well characterized. DecaBDE, however, likely sorbs
onto particles during wastewater treatment that are then removed in sludge (Sometal., 2011: Lassen etal., 1999). Depositing
sludge in landfills or spreading it on agricultural soil is one of the most significant potential releases of decaBDE to soils
(Ciparisand Hale, 2005: EU, 2002: Lassen etal., 1999).
Incineration
Based on available data, release of decaBDE from municipal incinerators is expected to be limited due to the use of high
temperatures and other pollution control technologies (Kohleret al., 2008: Palm etal., 2002: Lassen etal., 1999). Incomplete
incineration outside of municipal incineration facilities, however, could result in airborne release of decaBDE and the formation
of PBDFs, polybrominated dibenzo-p-dioxins, polychlorinated dibenzo-p-dioxins, and nonhalogenated substances such as
polycyclic aromatic compounds (see Appendix H).
Based on decaBDE information, research planning to support future risk assessments of MWCNTs might consider: What
volume of MWCNTs, MWCNT flame-retardant formulations, and MWCNT-treated upholstery textiles is likely to be disposed of,
recycled, or reused? Under what incineration conditions could release occur? Could specific conditions be modified (e.g.,
incineration temperature) to increase or decrease the release of MWCNTs to air? Are MWCNTs likely to end up in sewage
sludge? See Appendix H for more information regarding the potential release of decaBDE during disposal or recycling.
The reuse, recycling, and end-of-life stage encompasses a variety of different transformation and
disposal processes for (1) MWCNTs, (2) MWCNT flame-retardant formulations, and (3) MWCNT flame-
retardant upholstery textiles. What the primary reuse, recycling, and end-of-life treatments are for
MWCNTs and MWCNT flame-retardant formulations are unclear. The reuse, recycling, and end-of-life
treatments for flame-retardant upholstery textiles containing MWCNTs, however, are expected to be
similar to those of traditional flame-retardant upholstery textiles (see Appendix H.2.5).
2-28
-------
2.5.1. Reuse and Recycling
2.5.1.1. Life-Cycle Processes
Reuse or recycling of MWCNTs or MWCNT flame-retardant formulations is unlikely. On the
other hand, textile waste often is recovered and reused or recycled (Kohler et al.. 2008): upholstered
furniture is sometimes reused, but is rarely recycled (CalRecycle. 2002). Upholstery could be donated to
charitable organizations and resold for residential use. Additionally, upholstery textiles could be
informally repurposed into clothing, blankets, and other textile products. Due to the difficulty of recycling
furniture and flame-retardant materials, flame-retardant furniture is typically land-filled (CalRecvcle.
2002; Lassen etal.. 1999). Of the small portion of upholstered furniture that is recycled, about 60% of the
material is recycled and 25-30% is composted (CalRecvcle. 2002). No data were found that describe the
proportion of other upholstery textiles (e.g., mattress ticking or curtains) that are typically recycled.
The main types of textile recycling processes are fiber-to-fiber recycling and polymer reduction
recycling. During the fiber-to-fiber process, textiles are shredded and blended with other fibers to create a
new mixture ready for spinning (Kohler et al., 2008). During the polymer reduction process, textiles are
cut and granulated to form pellets that are processed to break down the polymer to the molecular level to
be reused as raw material (Kohler et al., 2008). No data were found that described the prevalence of each
recycling process.
2.5.1.2. Potential Releases during the Reuse/Recycling Stage
Release of MWCNTs beyond releases described in the use stage is unlikely to occur during reuse
of flame-retardant upholstery textiles. Older textiles could release greater levels of MWCNTs, however,
due to increased degradation of the material. Informal repurposing of flame-retardant textiles likely would
require cutting and shredding, resulting in possible air release of MWCNTs. Airborne releases of
MWCNTs could occur during recycling of flame-retardant textiles. Recycling subjects textiles to a variety
of mechanical, thermal, and chemical treatments that could result in the airborne releases of additive
flame retardants from fibers (Kohler etal.. 2008). Recycling processes, such as shredding, milling, and
thermal processing, could lead to the airborne release of CNTs from upholstery textiles if carried out in
uncontrolled environments (Chaudhry et al.. 2009). Airborne releases during recycling of textiles likely
would be in the form of CNTs in a polymer matrix (Chaudhry et al.. 2009). Downey cling, the conversion
of waste materials into new materials of lesser quality and reduced functionality, could lead to cross-
contamination of other materials with CNTs, for example, if MWCNT-treated textiles were shredded and
mixed with other textiles for use as insulation (Chaudhry et al., 2009). Release of MWCNTs to water also
2-29
-------
could occur during chemical treatment and processing. Although release of MWCNTs is possible during
recycling of flame-retardant textiles, no data were found that indicate the likelihood of release from
recycling processes. Similarly, no information was identified on whether or the extent to which other
substances (e.g., surface coatings) might be released with MWCNT flame-retardants from textiles during
product reuse or recycling.
Table 2-8 outlines potential release scenarios from the reuse/recycling stage of MWCNT flame-
retardant textiles along with decaBDE for comparison.
Table 2-8. Potential release scenarios during reuse and recycling.
Information on release
Processes included in reuse/recycling
life-cycle stage MWCNTs DecaBDE
General reuse (product kept intact) Air release possible if textile Air release possible if textile
degrades degrades
Repurposing (product manipulated) Air release possible due to cutting, Air release possible due to cutting,
shredding, and other abrasive shredding, and other abrasive
processes processes
Recycling (product broken down) Air and water release possible due Air and water release possible due
to mechanical, thermal, and to mechanical, thermal, and
chemical treatment chemical treatment
2.5.2. Incineration
2.5.2.1. Life-Cycle Processes
The incineration of MWCNTs or MWCNT flame-retardant formulations is unlikely, but any
incineration likely would occur in a hazardous waste incinerator. Upholstery textiles treated with
MWCNT flame-retardant coatings might be sent to municipal incinerators for processing. Municipal
incinerators generally provide a well-controlled environment with pollution control mechanisms and
sufficiently high temperatures (850 °C) to destroy most materials (Kohler et al.. 2008). Processing in
municipal facilities is likely to result in complete incineration of the upholstery textiles. Alternatively,
upholstery textiles also might be incinerated in less well-controlled facilities or burned in open fires as a
rudimentary form of waste management or as kindling. These incineration methods are likely to result in
incomplete incineration of the upholstery textiles. No data were found that describe the prevalence of
incineration as a form of disposal for upholstery textiles or what proportion of incinerated textiles is
processed at well-controlled incineration facilities.
2-30
-------
2.5.2.2. Potential Releases during the Incineration Stage
Airborne releases of MWCNTs from well-controlled incineration are expected to be negligible.
MWCNTs are likely destroyed at the high temperatures used by municipal waste incinerators due to
oxidation of MWCNTs (Chaudhry et al.. 2009; Sobek and Bucheli. 2009). Yet, some studies demonstrate
that controlled incineration does not always achieve the maximum temperature and, in some cases where
the maximum temperature is achieved, clumps of waste are still yielded, indicating incomplete
incineration (Cataldo. 2002; Sui et al.. 2001). Any CNTs remaining following municipal incineration
could be expected to bind to other particles and be removed by the incinerator's filter (Kohler et al..
2008); however, incinerator removal efficiency for CNTs has not been studied (Som et al.. 2011).
Alternatively, incomplete incineration (e.g., open fires) of products containing CNTs could result in the
airborne release of CNTs in a polymer matrix (Chaudhry et al.. 2009). Yet, as noted in Section 2.4.2.
Nguyen et al. (2011) found that exposure to 50 °C in conjunction with UV radiation and 75% relative
humidity, caused MWCNTs to form a dense barrier on the surface of the polymer that prevented the
release of MWCNTs to the environment. No data were found that describe potential by-products of
incinerating upholstery textiles coated with MWCNT flame retardant in either municipal incinerators or
in incomplete incineration scenarios.
In summary, due to the high temperatures and pollution control mechanisms at municipal
incinerators, MWCNTs in flame-retardant textiles are expected to be destroyed during well-controlled
incineration. Preliminary evidence suggests that MWCNTs might not be released to the environment
during incomplete incineration. Table 2-9 outlines potential release scenarios from the incineration stage
of MWCNT flame-retardant textiles along with decaBDE for comparison.
Table 2-9. Potential release scenarios during incineration.
Information on release
Processes included in
incineration life-cycle stage MWCNTs DecaBDE
Complete incineration, Release unlikely Release unlikely
controlled
Incomplete incineration, Preliminary evidence suggests that Air release of decaBDE and harmful
uncontrolled air release is unlikely by-products likely; will likely reach
environment
2-31
-------
2.5.3. Land-Filling
2.5.3.1. General Processes
Land-filling of MWCNTs or MWCNT flame-retardant formulations is unlikely, except in the
case of floor sweepings from manufacturing facilities. Upholstered furniture and textiles generally are
disposed of in municipal landfills (Kohler et al. 2008). Remaining parts from recycled furniture, such as
cover cloth materials, also are sent to the landfill (CalRecvcle. 2002). Additionally, some textiles might
be disposed of in uncontrolled landfills or open dumping sites that have no pollution control mechanisms
in place. No data were found that describe the proportion of upholstery textiles disposed of in landfills or
any further processing that might occur at the landfill.
2.5.3.2. Potential Releases during the Land-filling Stage
Land-filling of MWCNT flame-retardant textiles could lead to water and air releases. Mechanical
land-filling processes (e.g., mixing and compacting) could lead to the airborne release of CNTs in a
polymer matrix (Chaudhry et al., 2009). Airborne release of CNTs after land-filling is complete, however,
is not likely (Chaudhry et al.. 2009). Degradation of the polymer matrix material in textiles could lead to
release of CNTs into leachate/soil because CNTs are very stable and do not readily degrade (Kohler et al.,
2008). No data were found, however, that identify MWCNTs or coproducts (e.g., surface coatings) in
land-fill leachate. Similarly, no data were identified that measure releases of MWCNTs from land-filling
flame-retardant textiles, but the physicochemical characteristics of these materials suggest that such
releases likely would be small. Table 2-10 outlines potential release scenarios from the land-filling stage
of MWCNT flame-retardant textiles along with decaBDE for comparison.
Table 2-10. Potential release scenarios during land-filling.
Information on release
Processes included in land-filling life-
cycle stage MWCNTs DecaBDE
Disposal Air release possible due to mixing Air release possible due to mixing
and compacting and compacting
Degradation No data exist; air and water release Air and water release unlikely
possible but unlikely
2-32
-------
2.5.4. Wastewater Treatment Plants
2.5.4.1. Life-Cycle Processes
The wastewater treatment process consists of filtering and treating wastewater to remove solids
and contaminants. Large facilities that manufacture MWCNTs and MWCNT flame retardants might
divert their wastewater to an on-site wastewater treatment plant. Alternatively, some wastewater from
these facilities might be directly processed by municipal wastewater treatment plants. Water releases of
MWCNTs that occur during the storage and distribution, use, and reuse/recycling/end-of-life stages also
would be treated in municipal wastewater treatment plants.
2.5.4.2. Potential Releases during the Wastewater Treatment Stage
Release of MWCNTs or MWCNT flame-retardant formulations into wastewater could occur
throughout the life cycle. Primary releases to wastewater during manufacturing stages are due to
equipment cleaning, formulation and application of the flame retardant, and accidental spills. Washing
processes (which can involve abrasion, detergents, and water), particularly in the product manufacturing
stages, are likely to result in the release of additive flame retardants from textiles to wastewater (Som et
al., 2011). Due to the physicochemical characteristics of MWCNT flame retardants (see Table 1-8 and
Table 1-9 in Section 1.3), MWCNTs are likely to sorb to particles during water treatment and be removed
in sludge (Som et al.,
2011). The potential
nonetheless exists for
releases from filter
backwash and other
wastewater treatment plant
equipment (EU. 2002V
Additionally, some of this
removed sludge is
deposited in landfills or
spread on agricultural soil
(EU. 2002: Lassen et al..
1999). However, the
potential release of
MWCNTs due to the
Additional Information Highlight Box 5:
5. Impact of MWCNT release into wastewater treatment plants
Petersen et al. (2011c) noted that most of the CNTs that are released into wastewater
likely come from the tailoring, finishing, use, and degradation of textiles containing CNTs,
or from research and development facilities. This hypothesis is based on current
knowledge of the types of products into which CNTs are generally incorporated (i.e.,
sealed materials or polymers) and the expectation that most of these products will be
disposed of in landfills (Petersen et al., 2011c). Though this may change if the use of
CNTs in other applications increases, the fraction of CNTs entering wastewater is
expected to be small and to result generally from the use of CNTs in textiles (Petersen et
al.,2011c).
Because some evidence suggests that MWCNTs impact microorganisms (for example,
see Appendix G.5.1.1 and Appendix F for data on impacts to soil microbes), important
considerations for wastewater treatment facilities include: How will CNTs be removed
from the aqueous phase, and what is the removal rate? How will MWCNTs impact
diverse bacterial communities in wastewater, which are responsible for pollutant and
contaminant removal in these systems? In a review of the available literature, Petersen et
al. (2011c) found that the chemistry of CNTs is modified in activated sludge, microbial
communities were often negatively impacted by the addition of CNTs, and that certain
environmental conditions (e.g., extracellular polymeric substances) could lessen the
negative impacts.
2-33
-------
spread of sludge on agricultural soil is unknown. The releases of MWCNTs from wastewater treatment
facilities are expected to be small, but release potential greatly depends on surface chemistry.
The removal efficiency of wastewater treatment plants is not well characterized for MWCNTs and the
spread of sewage sludge onto agricultural soil could represent a significant source of MWCNTs to soil.
See Section 3.3.3 for information regarding MWCNT removal efficiency of these wastewater treatment
plants. In addition, negative impacts on WWTP functionality due to MWCNT contamination and
subsequent antimicrobial activity are possible (see Appendix G.5.1.1.1 and Table F-8). Similarly, no
information was identified on whether or the extent to which other substances (e.g., surface coatings)
might be released from wastewater treatment plants from processes associated with the production, use or
disposal of MWCNT flame-retardants in textiles.
Table 2-11 outlines potential release scenarios from the wastewater treatment stage of MWCNT
flame-retardant textiles along with decaBDE for comparison.
Table 2-11. Potential release scenarios during wastewater treatment.
Information on release
Processes included in wastewater
treatment life-cycle stage
MWCNTs
DecaBDE
Release of effluent
Release unlikely due to sorption
behavior, but filter backwash could
lead to release
Release unlikely due to sorption
behavior, but filter backwash could
lead to release
Removal of sludge
No data exist, but release to soil
possible if sludge spread on
agricultural fields
Release to soil likely if sludge
spread on agricultural fields
2-34
-------
This page intentionally left blank
-------
Chapter 3. Transport, Transformation,
and Fate
Releases throughout the product life cycles of upholstery textile coatings containing multiwalled
carbon nanotube (MWCNT) flame retardant will, to some extent, lead to occurrence of primary and
secondary contaminants in air, soil, and aquatic media. Chapter 3 examines what might happen to these
substances after their release to the environment, including transport or transformation through chemical,
physical, and biological processes. Studies investigating the transport, transformation, and fate of
MWCNTs in the environment are summarized in Appendix D. No data on measured MWCNT
concentrations in environmental media were identified; however, modeling studies provide some
estimates (see Table 3-2).
MWCNTs can be released into the environment during the manufacturing, storage, distribution,
use, disposal, reuse, and recycling of upholstery textiles treated with flame retardants (see Chapter 2).
MWCNT flame-retardant formulations are used primarily as additives that are mixed with, not chemically
bound to, polymers in textile products (see Section 1.3). Because they are not chemically bound, these
substances can escape from the material and become a source of contamination to surrounding
environmental media (Moniruzzaman and Winev. 2006). Although some, if not most, releases after the
production stage are likely to be in the matrix-bound form, little information exists that describes the
environmental behavior of MWCNT-polymer complexes. As a result, this chapter focuses on the
transport, transformation, and fate of MWCNTs not embedded in a polymer matrix.
Section 3.1 provides a brief discussion of the chemical and physical characteristics and the
processes that influence behavior (e.g., mobility, persistence, and bioavailability) of MWCNTs in
environmental media. The sections that follow summarize the available information regarding their
behavior in indoor and outdoor air (Section 3.2). aquatic systems (Section 3.3). and terrestrial systems
(Section 3.4). A brief discussion of models that might be used for evaluating their fate and transport in
environmental media is provided in Section 3.5.
3-1
-------
Additional Information Highlight Box 6:
6. Transformation throughout the product life cycle
Engineered nanomaterials such as MWCNTs are unlikely to occur in the environment in their as-manufactured form (see
Figure 2-2). The intended use and the disposal of consumer products that contain MWCNTs, such as upholstery textiles, often
expose the product to a wide variety of environmental conditions that can alter the composite material and the behavior of
MWCNTs after release (Nowacketal., 2012). Nowack et al. (2012) discuss the various processes that can alter or transform
(e.g., photochemical transformation, oxidation, reduction, adsorption/desorption, combustion, abrasion) nanomaterials directly
or nanomaterials in products. These processes can change how MWCNTs aggregate, disperse, and interact with biota
(Nowack et al., 2012). Greater understanding of how these processes influence MWCNT flame-retardant coatings in
upholstery textiles could support future assessments of the material.
3.1. Physicochemical Factors Influencing Transport,
Transformation, and Fate
The environmental fate of MWCNTs will be dictated by their physical and chemical properties
(see Text Box 1-1 and Figure 3-1). These properties influence behavior, including mobility, persistence,
bioavailability, and likelihood for transformation in environmental media. A summary of key
physicochemical factors that might affect partitioning17 of MWCNTs and their fate in the environment is
provided in Table 3-1. Values for or descriptions of key physicochemical properties of MWCNTs (e.g.,
surface area, morphology, solubility) are provided in Table 1-9.
17MWCNTs are not molecules, and therefore are not governed by traditional molecular-based equilibrium
partitioning theory. Nevertheless, the term "partitioning" is used in the literature to refer to the association or
attachment of MWCNTs to different media, and thus is used throughout this document as well. For MWCNTs,
associations with media, particularly paniculate matter or organic phases, are governed by colloidal aggregation
theory and may be irreversible, unlike in traditional equilibrium partitioning.
3-2
-------
Table 3-1. Summary of physicochemical properties that affect partitioning and fate of
nanomaterials such as multiwalled carbon nanotubes (MWCNTs).
Physicochemical property How does this property affect chemical partitioning and fate?
Small size; single particle versus • Single particles versus bundles or clusters will differ in their mobility (and
cluster ultimate fate) in environmental media; generally, nanoparticle clusters are less
mobile in the environment than individual nanoparticles
Source: Ma-Hock et al. (2007)
High surface area-to-volume ratio • Large surface area enhances chemical reactivity and clustering
• Might cause other molecules to adhere and be transported with MWCNTs
Sources: Kohler et al. (2008): O'Driscoll et al. (2010)
Distinct morphology • Concentrically nested multiple graphene sheets, which frequently exhibit
"disturbed wall texture" and irregular shape, increases chemical reactivity
• Differences in morphology based on variations in synthesis
• Shape can affect the kinetics of deposition and transport in the environment;
depending on surface structure and shape, MWCNTs might exhibit different
reactivity
Sources: Kohler et al. (2008): Oberdorster et al. (2005)
Low water solubility; hydrophobic • Will result in poor dispersion
(potentially hpophihc) . prone to bundNng in the water co|umn and settling to sediments; though
functionalization and surface chemistry can alter partition coefficients and
rates
• Hydrophobic interactions play major role in adsorption of organic
contaminants
• Functionalization and presence of surface-active agents (surfactants,
dissolved organic matter) can improve their dispersion/increase solubility in
aqueous media
• Might interact with lipids in abiotic and biotic media; might be taken up by
microbial communities and plant roots
Sources: Helland et al. (2007): Christian et al. (2008): Klaper et al. (2010): Saeed
(2010): Kohler et al. (2008): Luoma (2008): Li et al. (2011b): Oberdorster et al.
(2006): Wu et al. (2006): Kennedy et al. (2008)
Carbon nanotubes (CNTs) are not dispersed by simple mixing because they tend to form bundles
through a van der Waals attraction among tubes. As shown in Table 2-4, MWCNTs, including those in
commercial products such as textiles, can be engineered to include charged functional groups to improve
their dispersion or to increase their solubility in aqueous media; the treated nanoscale materials
(nanomaterials) that remain dispersed tend to exhibit greater persistence in the environment (Klaper et al..
2010; Saeed. 2010; Kohler et al.. 2008: Luoma. 2008). CNTs in textiles might also be coated with a
surface coating, such as a polymer (Kohler et al.. 2008). These surface coatings could be degraded by
chemical or biological reactions, affecting persistence of the MWCNTs overtime in ways that depend on
both the presence of a coating and the type of coating used.
3-3
-------
Additional Information Highlight Box 7:
7. Properties of the MWCNT formulation impact environmental release and transformation
MWCNT physicochemical properties can vary substantially as a result of MWCNT purification (see Section 2.2.3.1 and
Appendix C), functionalization (see Section 2.2.3.1 and Table 2-4), and formulation of MWCNT flame-retardant products (see
Section 2.2.4.1). Because many MWCNT applications, including flame-retardant coatings in upholstery textiles, are still in
research and development phases, however, it is unclear what type of modifications during production will dominate in the
market.
The types of modification during production can alter product chemistry and thus influence MWCNT release (see Figure 2-2):
transformation, transport, and fate; exposure; and human health and environmental impacts (see Text Boxes Text Box 1-1 and
Text Box 5-1, Table 3-1, and Figure 3-1). For example, Nguyen et al. (2011) observed that exposure of an epoxy containing
MWCNTs to ultraviolet (UV) radiation can cause MWCNTs to form a dense network on the composite surface, which might
minimize environmental release (see Section 2.4.2). Although Nguyen et al. (2011) simply describe the MWCNTs used in their
study as "commercially available," understanding which modifications of MWCNTs likely dominate a particular application
market, and how such modifications can influence potential environmental releases throughout the product life cycle could
inform future assessment and risk management efforts. See Section 6.3.2 for additional discussion on how different properties
of MWCNTs and the environment can influence MWCNT behavior after release into the environment.
Environmental conditions (e.g., redox potential, pH, temperature, UV light, ionic strength,
characteristics of other contaminants present) are also likely to affect the behavior and environmental fate
(e.g., mobility, persistence, bioavailability) of MWCNTs (Toth etal.. 2011; Zhang etal.. 2011; Helland et
al.. 2007) (see Text Box 1-1 and Figure 3-1). The same is true for environmental processes such as
interactions with natural organic matter (NOM), which can alter the surface chemistry of the MWCNTs
(Petersen et al.. 201 Ic). Metals, such as lead, cadmium, and copper; hydrophobic organic chemicals and
other toxic organics (e.g., polycyclic aromatic hydrocarbons); and other pollutants (e.g., phenol,
dopamine) can sorb strongly to CNTs (Li etal.. 20 lib; Toth etal.. 2011; Cho et al.. 2008; Petersen et al..
2008; Chen et al.. 2007; Helland et al.. 2007). These associations might dictate CNT mobility and
bioavailability. Factors that have been shown to influence adsorption of organic contaminants to
MWCNTs include surface oxidation (adsorption capacity decreases with increasing oxygen content) and
pH (effects differ based on the contaminant) (Li etal.. 20 lib: Toth etal.. 2011; Cho et al.. 2008).
3-4
-------
Properties of the
Environment
Properties of
MWCNT Material
Properties of MWCNT
Agglomerates
Hydrosphere3
pH
Salts
Natural Organic
Material
Atmosphere
UVA/B Radiation
^^^^^m
Lithosphere
Weathering
Erosion
Interaction with
functional groups;
Adsorption6
Phototransfor (nation/
Photodegradation
Abrasion/Leaching
Surface chemistry
• functionalgroups
• level of oxidation
Influence fate and transport through
environmental media
End-of-Use
Environments
Incineration
Landfill
Wastewater
treatment
Mineralization,
Transformation, and
Release
Adapted from: Misra et al. (2012)
Figure 3-1. Variability in MWCNT chemistry and implications in terms of life cycle, exposure, and
risk.c
"Hydrosphere describes environmental media characterized by water (i.e., ground water, surface water). Lithosphere describes
media characterized by rock composition (i.e., soil, sediment).
bAdsorption to environmental compounds (e.g., hydrophobic organic contaminants and metals).
""Environmental transport and fate - and therefore exposure, impacts, and risk - are intimately tied to the interactions between
MWCNT formulation and environmental media conditions. As illustrated here, environmental conditions can result in
transformation of the MWCNTs, for example, exposure to UV radiation can result in phototransformation, which alters the surface
chemistry of the MWCNTs (Misra et al.. 2012). In turn, the properties of the MWCNTs dictate how the compound moves through
the environment and partitions across various environmental media. The result is a cycle of interactions between environmental
conditions and MWCNT properties that influence movement and bioavailability of MWCNTs in the environment.
3-5
-------
3.2. Transport, Transformation, and Fate in Air
Priority Research Area: Environmental Transport, Transformation, and Fate
o
0.
11/13
w
2/13
0/13
[ Bioavailability
•
I
•
•
•
• - 1 participant's vote
Importance
L - Low:
M - Medium
H - High:
Confidence
L - Low:
M - Medium
H - High:
Least Important
Possibly Important
Important
Not Confident
Somewhat Confident
Confident
TH M IT~ ML H M L
E-RRF Confidence Rating E-RRF Confidence Rating E-RRF Confidence Rating
Eleven of 13 RTI workshop participants (85%) identified environmental transport,
transformation, and fate in air as important to risk assessment. These 11 participants were
asked to rate the importance of MWCNT mobility, persistence, and bioavailability to
environmental fate in air, as well as their confidence that the existing data on these risk
relevance factors could support risk-management decisions. Based on this information, these
areas were deemed a priority for research. In the event of a tie (e.g., confidence ratings for
MWCNT mobility and bioavailability in air), the most conservative rating (i.e., higher
importance, lower confidence) was chosen. See Section 6.3.2.1 for more information on this
priority area and Section 1.1.3 for a detailed explanation of the prioritization process.
DecaBDE Can Inform MWCNT Assessment
In air, 99% of BDE-209, the single isomer of decaBDE, exists in the particulate phase, making its fate in air dependent on the
characteristics of the particles to which it adsorbs. In indoor air, the tendency to sorb to particles likely would lead to higher
concentrations of BDE-209 in house dust than in vapor in the air and to extended persistence (Kemmlein et al., 2003). In
outdoor air, evidence suggests long-range atmospheric transport of polybrominated diphenyl ethers (PBDEs) to remote
ecosystems, including the Arctic (de Witetal., 2010: Su etal., 2009: Agrell etal., 2004). This atmospheric deposition of BDE-
209 is thought to be a main source of the background contaminants in waters and soils (Vonderheide et al., 2008). Notably,
temperature changes and UV exposure influence BDE-209 concentrations; UV exposure results in lower brominated
compounds in the environment due to photolysis of BDE-209 (Shih and Wang, 2009).
Based on decaBDE information, the following questions might be considered in planning research to inform future MWCNT
risk assessments: Will differences in MWCNT physicochemical properties result in critical differences in mobility, persistence,
and bioavailability in air? Could specific surface modifications be made to MWCNTs in flame-retardant applications that might
increase or decrease the mobility, persistence, and bioavailability of MWCNTS in air? How do environmental conditions (e.g.,
temperature, UV) influence mobility, persistence, and bioavailability of MWCNTs in air? See Appendix H for more information
about BDE-209 fate, transport, and transformation in air.
3-6
-------
MWCNTs released from flame-retardant upholstery textile coatings could reach indoor and
outdoor air in several ways. For example:
• They can be released directly into ambient air during all stages of the product life cycle, as
previously described in Chapter 2 and can disperse through air away from the source of
release.
• They can become suspended in the surrounding indoor or outdoor air during multiple stages
of the product life cycle.
• They might remain suspended and be transported through the atmosphere or be deposited
onto surfaces. Particles that have been deposited on surfaces could become resuspended in
the air and redeposited elsewhere.
MWCNTs might distribute to indoor air and dust, and these sources could be a major contributor
to outdoor air concentrations. Several processes and factors could influence the behavior (e.g., mobility,
persistence, bioavailability) and ultimately the fate of airborne MWCNTs in indoor and outdoor
environments, including: (1) size—whether they are traveling as individual particles of varying sizes or as
larger bundles, (2) surface chemistry, (3) interactions with other airborne particles and chemical
compounds, (4) residence time in the air, and (5) distance traveled prior to deposition (Kohler et al., 2008;
U.S. EPA. 2007). The fate of airborne nanomaterials outdoors could be influenced by meteorological
factors, including wind, temperature, and precipitation (Navarro et al.. 2008).
Information in the recent literature regarding the behavior of airborne MWCNTs is limited.
No studies have examined transport mechanisms for MWCNTs in air. Yang et al. (2009) examined the
atmospheric aging of CNTs under normal ambient conditions (20 ± 0.5°C, relative humidity = 50 ± 1%)
and found that CNT surface area and pore volume [volume of space (holes) per gram nanotube; provided
in cmVgram] decreased overtime (up to 7-15 months) and coincided with decreases in surface oxygen of
the CNT as it aged. For MWCNTs, these conditions stabilized within 15-18 months. The total structural-
defect concentration also appeared to be lowered as the CNTs aged. The authors theorized that during
CNT aging under ambient conditions, oxygen leaves the surface of the CNTs. The structure then repairs
itself and becomes more thermodynamically stable with fixed values of surface area, pore volume, and
structural defects. Based on these results, the authors stated that CNT "physicochemical properties can be
characterized with reliability only after samples have sufficiently aged" (Yang et al.. 2009).
In the laboratory, Zhu et al. (2011) observed that MWCNTs exposed to air under ambient
temperature were slowly oxidized and shortened. The degree of oxidation increased overtime, and
MWCNTs were almost totally transformed into amorphous carbon after 15 days.
Complete transformation of pure, airborne CNTs to carbon dioxide could occur during
incineration at temperatures greater than 850°C and oxygen levels exceeding 21%. Under oxidative
3-7
-------
conditions, MWCNTs have been shown to burn off completely at 740°C (Som et al., 2011; Kohler et al.,
2008).
Section 4.1.2.4 describes MWCNT air concentration data collected in occupational settings.
No data were found on residential or consumer exposures to MWCNTs in air. Literature containing
concentrations of MWCNTs in outdoor air also was not found.
3.3. Transport, Transformation, and Fate in Water and
Sediment
MWCNTs released from the flame-retardant upholstery textile coating life cycle could enter
aquatic systems in several ways. For example:
• MWCNTs in ambient air subsequently could be deposited or washed out to aquatic systems.
• Erosion of contaminated soil could release MWCNTs to surface waters.
• Runoff flowing along the ground surface could transfer MWCNTs in contaminated soil to
nearby waterways.
• Wastewater effluents containing MWCNTs could be a source of contamination to receiving
water bodies near the discharge location.
• MWCNTs could leach from land-filled sewage sludge into subsoil and ground water and
migrate to surface water or sediment.
3.3.1. Surface Water and Sediment (Inland and Coastal)
The transport, transformation, and fate of MWCNTs in surface water (specifically, mobility,
persistence, and bioavailability) was not identified as a priority area by workshop participants during the
collective judgment step of the Comprehensive Environmental Assessment (CEA) process. However,
mobility, persistence, and bioavailability in sediment was determined to be a priority area. Due to the
limited available data, which overlaps between surface water and sediment, these topics are discussed
together.
3-8
-------
Priority Research Area: Environmental Transport, Transformation, and Fate
Sediment
Mobility
« i5
uj t:
o
Q.
E.
7/13
4/13
2/13
•— X
CO
CC
t
o
I-1
• - 1 participant's vote
Importance
L - Low: Least Important
M- Medium: Possibly Important
H - High: Important
Confidence
L - Low: Not Confident
M - Medium: Somewhat Confident
H - High: Confident
~H ML~~ H M L H M L
E-RRF Confidence Rating E-RRF Confidence Rating E-RRF Confidence Rating
Seven of 13 RTI workshop participants (54%) identified environmental transport,
transformation, and fate in sediment as important to risk assessment. These seven
participants were asked to rate the importance of MWCNT mobility, persistence, and
bioavailability to environmental fate in sediment, as well as their confidence that the existing
data on these risk relevance factors could support risk-management decisions. Based on
this information, these areas were deemed a priority for research. See Section 6.3.2.3 for
more information on this priority area and Section 1.1.3 for a detailed explanation of the
prioritization process.
Unprioritized Research Area: Environmental Transport, Transformation, and Fate
Mobility
Persistence
Bioavailability
u v
c o
— ~
u
o
0.
E.
2/13
E T-
Ml CO
t
o
8/13
o
I-1
re
*
o
M—L~~ ML ML
E-RRF Confidence Rating E-RRF Confidence Rating E-RRF Confidence Rating
3/13
• - 1 participant's vote
importance
L - Low: Least Important
M - Medium: Possibly Important
H - High: Important
Confidence
L -Low: Not Confident
M—Medium: Somewhat Confident
H - High: Confident
Two of 13 RTI workshop participants (15%) identified environmental transport,
transformation, and fate of MWCNTs in surface water as important to risk assessment.
Based on this information, MWCNTs in surface water were determined to be of lesser
importance to consider in a future risk assessments, and all text relevant to these areas
was moved to Appendix G. See Section 1.1.3 for a detailed explanation of the prioritization
process.
3-9
-------
The hydrophobicity and van
der Waals interactions of MWCNTs
imply they will partition (see
Footnote j/7 in Chapter 3) to the
participate phase when introduced to
aquatic systems (Kennedy et al.
2008). Pure MWCNTs are insoluble
in water and are prone to bundling
in the surface water column and
settling to sediments—making
benthic organisms potential vectors
for the transport of MWCNTs
through the food web (Christian et
al.. 2008) (see Section 43).
MWCNTs suspended in NOM
solutions have greater potential for
dispersion in natural waters
(ODriscolletal..201Q). Results of
the O'Driscoll et al. (2010) study
suggested that smaller diameter
DecaBDE Can Inform MWCNT Assessment
Water solubility and Kow of BDE-209 indicate that it will partition to the
particulate phase in water or bind strongly to sediments, making sediment a
sink for PBDEs. BDE-209 is the dominant PBDE congener in sediment
samples, although whether this is due to greater use of BDE-209 or its
resistance to environmental degradation is unclear. Despite relatively high
concentrations in sediment, the bioavailability of BDE-209 in sediment is
expected to be limited due to its strong hydrophobicity and large molecular
size (Liu et al., 2011b). Further, environmental transformations (e.g.,
debromination, See Appendix H, Text Box H.3-1) of BDE-209 likely influence
its concentration in sediment because lower brominated congeners are more
water soluble than higher brominated congeners, and are therefore more
mobile in the water column (Soderstrom et al., 2004: Watanabe and Sakai,
2003). Environmental conditions such as the amount of organic matter and
microbial or photolytic degradation processes also influence levels of PBDEs,
such as BDE-209, in sediment or surface waters.
As with decaBDE, physicochemical properties and environmental conditions
could be important to consider in planning research that informs future risk
assessments of MWCNTs. For example, will environmental processes act on
MWCNTs in a way that makes them more or less likely to partition to
sediments? Will MWCNTs that have been functionalized to increase
dispersibility in aqueous media be likely to sorb to particles in the
environment? Can specific physicochemical formulations of MWCNTs make
them more or less likely to partition to sediments even under environmental
conditions? What environmental conditions (e.g., organic matter content,
microbial community composition) influence mobility, persistence, and
bioavailability of MWCNTs in sediments? See Appendix H for more
information about BDE-209 fate, transport, and transformation in surface
water and sediment.
MWCNTs stay suspended in NOM solutions much longer than larger diameter MWCNTs. Dissolved
organic matter has been shown to debundle MWCNTs and induce conformational and electrostatic
stabilization of carbon-based nanomaterials under environmentally relevant conditions (Hyung and Kim.
2008: Wang et al.. 2008: Hvung et al.. 2007).
Changing conditions such as ionic strength and pH of an aqueous solution can influence sorption
behaviors of CNTs and subsequently the mobility, persistence, and bioavailability of these compounds in
water. Zhang et al. (2011) suggested that the overall effect of increasing ionic strength is that more
MWCNTs transfer from the aqueous phase and sorption increases. The authors also indicated that, in
general, decreasing the pH of aqueous solutions enhances MWCNT bundling. The presence of dissolved
organic matter, however, contributes to stabilization of MWCNTs in solution and suspended MWCNTs
become less sensitive to changes in ionic strength or solution pH (Zhang etal. 2011). Similarly, Hyung
and Kim (2008) determined that adsorption capacity is directly proportional to the aromatic carbon
content of the organic matter and the ionic strength of the solution and indirectly proportional to pH.
Adsorption strength is indirectly proportional to ionic strength and is not significantly changed by pH.
3-10
-------
Functionalization of MWCNTs can improve their dispersion or increase their solubility in
aqueous media, thereby increasing their mobility. As mentioned previously, engineered surface
modifications (e.g., functional groups and coatings) are used to improve CNT dispersion in aqueous
suspension. Column stability and settling experiments have shown pure MWCNTs settle and sink to
sediment as described previously. The presence of functional groups slows this settling, especially in
combination with NOM (Kennedy et al.. 2008; Hyung et al.. 2007).
As discussed in Section 3.1. MWCNTs can act as environmental adsorbates of metals,
hydrophobic organic compounds, and other toxic organics (Li etal. 20 lib: Cho et al.. 2008; Petersen et
al.. 2008; Chen et al.. 2007; Helland et al.. 2007). and the properties of the adsorbants might dictate
MWCNT mobility and dispersion in surface water.
Studies relevant to MWCNT fate and transport in aqueous media are summarized in Appendix D.
Table D-2. No data were found on concentrations of MWCNTs in surface water and sediment.
3.3.2. Ground Water
Unprioritized Research Area: Environmental Transport, Transformation, and Fate
^^ I*V*
o o
LJ
o
Q.
2/13
2/13
9/13
• - 1 participant's vote
Importance
L - Low: Least Important
M - Medium: Possibly Important
H - High: Important
Confidence
L - Low: Not Confident
M - Medium: Somewhat Confident
H - High: Confident
Mobility
Persistence
Bioavailability
01
E -r
o
Fffl
01
E -T
to
OL
o
w
H M L H M L ^ MIT
E-RRF Confidence Rating E-RRF Confidence Rating E-RRF Confidence Rating
Two of 13 RTI workshop participants (15%) identified environmental transport,
transformation, and fate of MWCNTs in ground water as important to risk assessment.
Based on this information, MWCNTs in ground water were determined to be of lesser
importance to consider in a future risk assessment of MWCNTs, and all text relevant to
these areas was moved to Appendix G. See Section 1.1.3 for a detailed explanation of the
prioritization process.
3-11
-------
3.3.3. Wastewater
Priority Research Area: Environmental Transport, Transformation, and Fate
gr
O 0)
E £
o
Q.
11/13
2/13
0/13
• - 1 participant's vote
Importance
L - Low: Least Important
M - Medium: Possibly Important
H — High: Important
Confidence
L - Low: Not Confident
M - Medium: Somewhat Confident
H - High: Confident
01
C -r
13
or
w
Mobility
H M L H M L ~~HM
E-RRF Confidence Rating E-RRF Confidence Rating E-RRF Confidence Rating
Eleven of 13 RTI workshop participants (85%) identified environmental transport,
transformation, and fate in wastewater as important to risk assessment. These 11
participants were asked to rate the importance of mobility, persistence, and bioavailability
to MWCNT environmental fate in wastewater, as well as their confidence that the existing
data on these risk relevance factors could support risk-management decisions. Based on
this information, these areas were deemed a priority for research. See Section 6.3.2.2 for
more information on this priority area and Section 1.1.3 for a detailed explanation of the
prioritization process.
Information in the recent literature
regarding the behavior of MWCNTs in
wastewater is limited. Because of their
hydrophobicity and tendency to form bundles,
however, pure MWCNTs likely would be
removed from the effluent by settling during the
sewage treatment process. Functionalized
MWCNTs could have improved dispersion and
increased solubility in wastewater, thereby
increasing their mobility and persistence in
wastewater effluents.
As discussed in Section 3.1. MWCNTs
can act as environmental adsorbates of metals,
hydrophobic organic compounds, and other
toxic organics (Li et al.. 20lib; Cho et al.. 2008;
Petersen et al.. 2008; Chen et al. 2007; Helland
et al.. 2007). and these adsorbants might dictate
DecaBDE Can Inform MWCNT Assessment
Because of its hydrophobicity, most (>99%) BDE-209 present
in wastewater sorbs to sediments, making sewage sludge a
major sink for this material (Ricklundetal., 2009: North, 2004).
As mentioned in Section 2.5, the application of sewage sludge
to agricultural fields is one of the most significant potential
releases of decaBDE to soils (Ciparis and Hale, 2005: Lassen
et al., 1999). Wastewater effluents, in contrast, could contain
lower brominated transformation products of BDE-209 and
thereby contaminate receiving water bodies or local aquatic
ecosystems near the discharge location (Peng et al., 2009:
Song etal., 2006).
Based on decaBDE information, considerations to include in
planning MWCNT research to support future assessments are:
Will MWCNT functionalization to increase dispersibility cause
them to remain dispersed in wastewater? What concentrations
of MWCNTs might end up in sludge applied to agricultural
soils? What is the bioavailability of MWCNTs in sewage sludge
and how does functionalization influence bioavailability? What
types of transformation products might partition in wastewater
sludge or effluents? See Appendix H for more information on
fate, transport, and transformation of BDE-209 in wastewater.
3-12
-------
CNT mobility and dispersion in wastewater. He et al. (2012) studied the behavior of MWCNTs stabilized
by humic acid during the coagulation-flocculation-sedimentation process of drinking water treatment and
found that humic acid-stabilized MWCNTs were effectively sequestered by this process.
Literature containing concentrations of MWCNTs in wastewater effluent or sludge was not
found.
3.4. Transport, Transformation, and Fate in Soil
Unprioritized Research Area: Environmental Transport, Transformation, and Fate
o o
E £
m
o
a.
3/13
5/13
5/13
Mobility
Persistence
Bioavailability
0>
E -T
15
u.K
?I =
w B
o
I-1
01
•I1
u.*
i?I=
UJ B
O
IJ
w
• - 1 participant's vote
Importance
L -Low: Least Important
M-Medium: Possibly Important
H —High: Important
Confidence
L - Low: Not Confident
M — Medium: Somewhat Confident
H - High: Confident
M |T~ H M L H M L
E-RRF Confidence Rating E-RRF Confidence Rating E-RRF Confidence Rating
Three of 13 RTI workshop participants (23%) identified environmental transport,
transformation, and fate of MWCNTs in soil as important to risk assessment. Based on
this information, MWCNTs in soil were determined to be of lesser importance to consider
in a future risk assessment of MWCNTs, and all text relevant to these areas was moved
to Appendix G. See Section 1.1.3 for a detailed explanation of the prioritization process.
3-13
-------
Unprioritized Research Area: Environmental Transport, Transformation, and Fate
Bioaccumulation
o
LLI -E
2/13
m
8/13
•e
o
r
E-
H M L
RRF Confidence Rating
• - 1 participanfsvote
Importance
L - Low: Least Important
M - Medium: Possibly Important
H - High: Important
Confidence
L - Low: Not Confident
M - Medium: Somewhat Confident
H - High: Confident
3/13
Two of 13 RTI workshop participants (15%) identified environmental transport, transformation, and fate of MWCNTs in biota as
important to risk assessment. Based on this information, MWCNTs in biota were determined to be of lesser importance to
consider in a future risk assessment of MWCNTs. In the draft document (U.S. EPA, 2012b) reviewed by RTI workshop
participants information pertinent to this area was primarily identified for plants in soil; since transport, transformation, and fate
in both biota and soil were not identified as priorities all relevant text was moved to Appendix G. See Section 1.1.3 for a detailed
explanation of the prioritization process.
3.5. Multimedia Models to Predict Environmental Fate and
Transport
| Neutral Research Area: Environmental Transport, Transformation, and Fate )
Multimedia models predicting environmental fate and transport of MWCNTs were not considered during the RTI collective
judgment prioritization process. This section of text, however, is included in the main document because it supports an
understanding of the priority research areas presented in this chapter.
Although empirical data on MWCNT concentrations in the environment are lacking, some
researchers have used modeling to simulate movement of CNTs through environmental compartments
and to derive predicted environmental concentrations (PECs). Mueller and Nowack (2008) used substance
flow analysis to model CNTs in air, soil, and water (not sediment) in Switzerland based on simplifying
assumptions. PECs were calculated for "realistic" scenarios and "high emission" scenarios and are
provided in Table 3-2.
Gottschalk et al. (2009) described a probabilistic material flow analysis framework to derive
probability distributions of PECs for engineered CNTs in soil, sludge-treated soil, air, surface water,
sediment, and sewage treatment plant effluent and sludge for the United States, Europe, and Switzerland
(see Table 3-2). As noted in Gottschalk et al. (2010). although both studies were designed to estimate
3-14
-------
PECs in environmental media, the two study designs were quite different and the methodologies used
varied considerably, making a direct comparison of PECs difficult. The differences included the model
type (deterministic versus probabilistic); model scale [Gottschalk et al. (2009) considered additional
environmental compartments (sediment and ground water), more flows associated with these additional
compartments, as well as production, manufacturing, and recycling processes]; model input data [newly
available model input data were used in Gottschalk et al. (2009)]: and amended categorization of the
products and allocation of the CNT mass to the product categories.
Another study employed the USEtox model to carry out a life cycle-based analysis of the aquatic
toxicity impacts associated with CNT synthesis (Eckelman et al.. 2012). This study used information on
the physicochemical properties of CNTs to model fate and transport of CNTs in freshwater systems under
"realistic" and "worst-case" or "conservative" scenarios, as described by the authors. Based on
information from Gottschalk et al. (2009). the realistic scenario assumed the fraction of CNTs removed
from the water column due to clustering and settling to be 90%. For the 10% remaining in the water
column, this scenario assumed a CNT exposure factor (defined by USEtox as the dissolved fraction of
CNTs in the water column) of 98%, with most of the remaining 2% partitioning to suspended solids. This
realistic scenario, which utilized Monte Carlo analysis, calculated a mean residence time in fresh water to
be on the order of days. The conservative scenario assumed an exposure factor in the water column of
100% and estimated the freshwater residence time for CNTs as 143 days. This study did not distinguish
between single-walled and multi walled CNTs.
3-15
-------
Table 3-2. Predicted environmental concentrations of CNTs using fate and transport modeling.
Medium
Air (|jg/m3)
Mueller and Nowack (2008)
1.5 x 10"3 (realistic scenario);
2.3 x 10"3(high emission scenario)
Gottschalk et al. (2009f
1 x irj6 (United States)
3x irj6 (Europe)
8x irj6 (Switzerland)
Soil (ug/kg or Aug/kg-yeara
1 x 10" (realistic scenario);
2x 10"2 (high emission scenario)
5.6 x irj4 (United States)
1.5x 10'3 (Europe)
1.9 x 10"3 (Switzerland)
Sludge-treated soil (Aug/kg-year)a
ND
3.1 x irj2 (United States)
7.4 x 10"2 (Europe)
ND (Switzerland)2
Surface water (ug/L)
5x10" (realistic scenario);
8 x 10"4 (high emission scenario)
1 x 10"6 (United States)
4 x 10"6 (Europe)
3x 10'6 (Switzerland)
Sediment (Aug/kg-year)a
ND
4.6 x 10"2 (United States)
2.4 x 10"1 (Europe)
2.3 x 10'1 (Switzerland)
Sewage treatment plant (STP)
effluent (ug/L)
ND
8.6 xiO'3 (United States)
1.5 xio~2 (Europe)
1.2x 10'2 (Switzerland)
STP sludge (mg/kg)
ND
6.8 x 10'2 (United States)
6.2 x 10'2 (Europe)
6.9 x 10"2 (Switzerland)
aFor Gottschalk et al. (2009). air, surface water, STP effluent, and STP sludge concentrations are modes (most frequent values)
from 2008. For soil, sludge-treated soil, and sediment, values are modes that represent annual increases in concentrations (Aug/kg-
year).
bln Switzerland, sewage sludge is not applied to soil.
Abbreviations: ND = no data
Sources: Reprinted (adapted) with permission of American Chemical Society [Mueller and Nowack (2008)1: and Gottschalk et al.
(2009),
Cullen et al. (2010) simulated subsurface mobility of MWCNTs compared with nanofullerenes
(nC6o) under a range of hydrologic and geological conditions (homogeneous and heterogeneous) using a
two-dimensional finite element model. In general, nanoparticles in systems with the same average
hydraulic properties were predicted to be less mobile if the systems were heterogeneous as opposed to
homogeneous. For the conditions evaluated, MWCNTs were predicted to be much more mobile compared
with nC6o because of two factors—nanoparticle shape and size. The smaller, spherical nC6o were more
efficiently collected on soil surfaces compared with the larger, cylindrical MWCNTs.
3-16
-------
Additional Information Highlight Box 8:
8. Multimedia modeling of MWCNT environmental transport
Multimedia modeling is challenged not only by a lack of empirical data on environmental concentrations of MWCNTs, but also
by a lack of knowledge regarding the predominant MWCNT fate and transport mechanisms in specific media and how
MWCNT physicochemical properties might impact these processes. For example, although classical filtration theory can model
particle transport in soil, additional mechanisms (e.g., deposition, straining) might play an important role in removal of non-
spherical, high-aspect colloidal particles, such as CNTs, from suspensions (Mattison etal., 2011). Classical filtration theory,
therefore, might not be applicable to CNTs due to the large aspect ratio and unique surface properties of these chemicals.
Mattison et al. (2011) tested the suitability of several parameters for describing MWCNT transport in subsurface environments,
including porous subsurface media. They concluded that a dual deposition model coupled with site blocking better described
MWCNT transport than traditional colloid filtration theory. In developing the dual deposition model, Mattison et al. (2011) used
column experiments in the lab and observed that initial mobility in the first pulse of MWCNTs (acid modified with an diameter
of 36 ± 11 nm, length of 540 ± 340 nm, and low level of metal impurities) that they evaluated was slower than a conservative
tracer (NaBr), although mobility increased relatively rapidly after initial breakthrough. During the second pulse, MWCNT
effluent in the sand-packed column reached a near-maximum concentration at the same time as the conservative tracer, and
then gradually rose to a maximum value more slowly than the tracer. Smaller grain size of sand or silt (e.g., 50-um versus
80-um grains) resulted in slower mobility of MWCNTs through media, while pore-water velocity and ionic strength had fairly
limited influence on MWCNT mobility (Mattison et al., 2011). Notably, normalized effluent concentrations of MWCNTs were
greater than 60% of the influent concentration. These results (Mattison etal., 2011) Similarly, O'Caroll et al. (2013) tested the
influence of diameter on MWCNT behavior in sand. They showed that smaller MWCNT's are actually less mobile in porous
media that larger MWCNTs, likely due to Brownian motion which results in more collisions between smaller-sized MWCNT's
and the porous media. The authors applied a transport model to their data that was based on colloid filtration theory and a site
blocking mechanism. These results (Mattison et al., 2011) and O'Caroll et al. (2013), differ from results with humic acid-
stabilized MWCNTs, which moved more quickly through media than the conservative tracer [(Wang et al., 2008) as cited in
Mattison et al. (2011)1. Several recent laboratory studies have shown that humic acids (representing NOM) in peat soil bind to
MWCNTs (Tian et al., 2012: Wang et al., 2011), which alters the surface functionalization and electronegative charge and
increases sorption of heavy metal ions (Tian etal., 2012). These alterations of surface functionalization could alter transport of
MWCNTs through soil. For example, Wang et al. (2008) observed that humic acid-coated CNTs are highly mobile in porous
media. In another study Zhang et al. (2012a) showed that increasing ionic strength results in MWCNT sorbing to soil minerals
in solution and promotes removal of MWCNTs from the aqueous phase. Results from these studies suggest that, although
empirical data on MWCNT fate and transport outside of the laboratory are lacking, modeling efforts can identify environmental
factors such as grain size and media composition that can alter the environmental behavior and physicochemical properties of
MWCNTs, which could affect their exposure and toxicity potential for human and ecological receptors.
3-17
-------
Chapter 4. Exposure-Dose
Releases of multiwalled carbon nanotubes (MWCNTs) to the indoor and outdoor environments
can occur at multiple stages of the product life cycle for flame-retardant upholstery textile coating
(Chapter 2). Subsequent transport, transformation, and fate processes dictate how MWCNTs distribute
through various environmental media once released (Chapter 3). Exposure describes the pathways
through which contact occurs between contaminants in the environment and living organisms and abiotic
receptors. Toxicokinetics [i.e., absorption, distribution, metabolism, excretion (ADME)] describes the
processes that relate exposure (or dosage) to the internal dose, which refers to the quantity of a chemical
or material that is taken up and absorbed by living organisms (U.S. EPA. 2010c).18
Section 4.1 introduces analytical techniques for identifying, characterizing, and measuring
MWCNTs in various matrices. The various metrics recommended for characterizing exposure and dose of
MWCNTs are also discussed, and available concentration data in various indoor and outdoor media are
presented. In the absence of data quantifying MWCNT exposures at the point of contact, measured
concentrations of MWCNTs in surrounding media can be used to estimate exposures using a scenario
evaluation approach. Section 4.2.1. Appendix G.4.1. and Appendix G.4.2 expand on the release scenarios
presented in Chapter 2 to discuss the potential human and ecological exposure pathways that link those
releases to receptors. No data were identified regarding relevant exposure pathways leading to impacts on
abiotic receptors,19 as a result, this comprehensive environmental assessment case study does not include
a discussion of exposure scenarios that would influence abiotic receptors. Although broad potential
impacts on society and the global environment are discussed in Section 5.3. exposure is either not
18The term "dose" is described generally by the United States Environmental Protection Agency (EPA) Integrated
Risk Information System (IRIS) as "[t]he amount of a substance available for interactions with metabolic processes
or biologically significant receptors after crossing the outer boundary of an organism." Several specific forms of
dose are also described by IRIS, but the definitions of these terms are not used consistently across the risk
assessment community. The following definitions of specific forms of dose are provided by IRIS: "The
POTENTIAL DOSE is the amount ingested, inhaled, or applied to the skin. The APPLIED DOSE is the amount
presented to an absorption barrier and available for absorption (although not necessarily having yet crossed the outer
boundary of the organism). The ABSORBED DOSE is the amount crossing a specific absorption barrier (e.g.,
The exchange boundaries of the skin, lung, and digestive tract) through uptake processes. INTERNAL DOSE is a
more general term denoting the amount absorbed without respect to specific absorption barriers or exchange
boundaries. The amount of the chemical available for interaction by any particular organ or cell is termed the
DELIVERED or BIOLOGICALLY EFFECTIVE DOSE for that organ or cell."
19The term "abiotic receptors" refers to nonliving entities such as buildings, statues or other structures, painted
surfaces of vehicles or other objects and features of the natural landscape.
4-1
-------
considered germane to the discussion of the impact (such as for economic impacts of manufacturing
MWCNTs) or the exposure characteristics related to the impact are already included in the general
discussion that follows (such as for higher potential exposure levels in certain populations related to
socioeconomic impacts).
The scenarios described in Sections 4.2.1 and Appendix G.4 describe the conditions under which
exposures might occur; this information can be used in combination with measured or modeled
concentrations in environmental media from Section 4.1 and exposure factors to estimate exposures.
Kinetic information then can be used to determine or estimate the internal dose that results from external
exposures. When available, point-of-contact measurements, administered dosages, tissue or body burdens,
scenario-specific exposure guidelines and recommendations are provided, and the toxicokinetics of
MWCNTs are described in Section 4.2.2 and Appendix G.4.2. Studies describing toxicokinetics of carbon
nanotubes (CNTs) in mammals are summarized in Appendix F. Finally, Section 4.4 discusses aggregate
exposures to MWCNTs from multiple sources and Section 4.5 discusses cumulative exposures to multiple
related stressors.
As described in Section 2.2.4. MWCNTs likely would be incorporated into a polymer or other
type of matrix in the flame-retardant formulation applied to upholstery textiles, and both the free and
matrix-bound forms might be released during the product life cycle. Very little data relevant to MWCNT
exposures, however, have been generated for the matrix-bound form of MWCNTs. This lack of data
necessitates a reliance on the existing data for free MWCNTs in the discussion throughout this chapter.
The extent to which exposure characteristics and dose implications differ between the free and matrix-
bound form of MWCNTs, however, is unknown at this time.
4-2
-------
Unprioritized Research Area: Exposure Route
Unprioritized Research Area: Dose (Kinetics)
o o
E £
* «
m r
o
0.
0/13
3/13
10/13
• - 1 participant's vote
Importance
L - Low: Least Important
M —Medium: Possibly Important
H - High: Important
Direct Contact
o o
c o
— e
U
O
Q.
E.
0/13
2/13
Absorption
11/13
No RTI workshop participants identified abiotic exposure or dose as important to risk
assessment. Based on this information, exposure of abiotic resources to MWCNTs was
determined to be of lesser importance to consider in a future risk assessment of MWCNTs.
As discussed above, abiotic exposure and dose were noted but not discussed in Chapter 4
due to insufficient data; as such this box is presented in the beginning of the chapter.
See Section 1.1.3 for a detailed explanation of the prioritization process.
4.1. Detection, Measurement, and Characterization
[ Neutral Research Area: Detection, Measurement, and Characterization |
Detection, measurement, and characterization of MWCNTs—including dose and exposure metrics and concentrations in
environmental media (outdoor air, aquatic and terrestrial ecosystems) and indoor environments (occupational, residential, and
nonresidential settings)—were not considered during the RTI collective judgment prioritization process. This section of text,
however, is included in the main document because it supports understanding of occupational exposure pathway scenarios
(see Section 4.2.1) and consumer exposure pathway scenarios (see Section 4.2.2), which were deemed priority research
areas.
Exposure scenario evaluation requires information on measured, modeled, or reasonably
estimated concentrations of a stressor in exposure media. As introduced in Chapter 1. MWCNTs represent
a group of compounds, encompassing substances that span a range of physicochemical characteristics and
properties. As a result, developing reliable analytical techniques for detecting, measuring, and
characterizing the full range and makeup of MWCNTs in environmental media can present challenges.
Text Box 4-1 provides a brief discussion of a few common analytical techniques and the general
challenges associated with them. Appendix B summarizes common analytical techniques and presents the
strengths and limitations of each technique.
4-3
-------
Configurations of MWCNTs vary depending on the type of material or substrate used in their
manufacture (see Section 2.2). Single analytical techniques used alone are generally not sufficient for
characterizing all of the properties of MWCNTs that can influence exposure; to characterize the presence
and form of MWCNTs in media adequately, multiple analytical methods must be used in tandem (see
Text Box 4-1 and Appendix B). Text Box 4-2 provides examples of the specific physicochemical
properties of MWCNTs that influence exposure, uptake, and dose.
Text Box 4-1. Detecting, Measuring, and Characterizing MWCNTs
Because MWCNTs tend to clump, multiple, orthogonal techniques are recommended to characterize MWCNTs adequately
in exposure media (Petersen and Henry, 2012). Analytical techniques for detecting, measuring, and characterizing
MWCNTs are summarized in Appendix B. Several detection and quantification techniques are available for MWCNTs in
aqueous media, but fewer are available for evaluating MWCNTs in other media. Due to the challenges associated with
detecting, measuring, and characterizing very small concentrations of highly reactive and polydispersed particles,
successful protocols for extracting MWCNTs from relevant matrices or media (e.g., textiles, polymers, body fluids) could
differ from those used to extract traditional compounds for exposure studies based on preliminary, unpublished evidence
presented at a public meeting (Uddin and Nyden, 2011a: Uddin and Nyden, 2011b). Additionally, many detection methods
rely on knowledge of properties of the material as produced, which can change dramatically during subsequent stages of
the product life cycle. Currently, the main application for detection methods is to verify concentrations of as-manufactured
MWCNTs during laboratory-based experiments.
Radioactive or isotopic labeling are precise quantification methods that work in any medium, but CNTs must be
radioactively or isotopically labeled prior to dispersal in environmental media, as well as purified to remove residual free
isotopes for this method to work (Wang et al., 2013: Petersen and Henry, 2012). In aqueous suspensions, the most
straightforward method for quantification is gravimetric assessments of suspended materials. The nominal concentration at
the start of the study must be known, however, because deviation from that nominal value is what is measured.
The potential for human exposure from upholstery textiles coated with flame retardants is
currently difficult to assess because data are not yet available on production of MWCNTs for flame-
retardant upholstery textiles (see Section 2.2.4). Furthermore, very little information is available on
whether or how MWCNTs might migrate from a product matrix; or on what methods could be used to
quantify exposure concentrations of matrix-bound MWCNTs and partially exposed MWCNTs (i.e., ends
of MWCNTs "sticking out" from the matrix). Understanding the behavior of the material requires
comparing potentially similar applications. For example, machining of CNT composites and the resultant
exposure could be compared to migration of MWCNTs out of textiles and the resultant exposure.
The reason for this is that machining of CNT composites could generate particles or fibers similar to those
generated from the wear and breakdown of MWCNT textiles, so these studies are useful from an
exposure assessment perspective. Estimating exposures during other product life-cycle stages also might
be possible by evaluating similar applications or alterations of materials containing MWCNTs. Such
4-4
-------
assumptions, however, could yield conclusions that are very different from the actual behavior and
exposure potential of the MWCNTs in flame-retardant upholstery textile applications.
Text Box 4-2. Specific Physicochemical Properties of MWCNTs Shown to Influence
Exposure, Kinetics, and Dose
As introduced in Text Box 1-1, the physicochemical characteristics of MWCNTs can be altered (both intentionally and
unintentionally) by using different methods, materials, and processing techniques under different ambient conditions, and these
characteristics can change further over the course of the life cycle. Several studies have explored how changes in individual
MWCNT characteristics can affect exposure, kinetics, and dose. The following physicochemical characteristics have been
identified as contributing to changes in the behavior of CNTs in vitro and in vivo [as summarized by Johnston et al. (2010)1.
Because most studies have focused on the inhalation and dermal routes of exposure, data are extremely limited for the oral
route.
Dispersion State. Individual CNTs often form larger bundles that range from tightly aggregated (i.e., thick and
ropy) to loosely agglomerated (i.e., tangled, like steel wool). Following inhalation, long, well-dispersed CNTs are
more likely to deposit deeper in the respiratory tract, where they might be taken up by cells via phagocytosis (i.e.,
engulfed by the cell membrane). Well-dispersed CNTs are more likely to translocate to other sites following
deposition in the lung, leading to a shift in CNT presence from the active airways to the interstitium and alveolar walls, and
ultimately to the circulatory system. Up to a certain size limit, bundles of CNTs tend to be more biopersistent in the lung than
well-dispersed CNTs, remaining in the conducting airways or entering macrophages at the point of contact. Studies of skin
cells (keratinocytes) have also demonstrated that dermal uptake of free MWCNTs can depend largely on the dispersion state
of the MWCNTs, with limited uptake in the absence of large bundles.
Morphology. Short (usually defined as <15 urn) CNTs are more readily taken up into cells, but they are also more
readily cleared (for example, via macrophages following inhalation). Longer CNTs, on the other hand, are more
persistent at the deposition site and might get "stuck" in the cell membrane, resulting in "frustrated" phagocytosis or
endocytosis. The length of CNTs might be more important than functionalization, as neutral and positively or
negatively charged CNTs are consistently internalized when CNT length remains <2 urn. CNTs engineered with rounded or
open ends will also exhibit different uptake mechanisms into cells. CNTs with rounded tips enter cells at a perpendicular angle,
whereas CNTs with open ends enter cells parallel to the cell surface (Shi etal., 2011). The ratio of MWCNT length to diameter
(i.e., aspect ratio), which can be a relevant exposure metric for fibers like asbestos, can also help predict the deposition sites
of CNTs in the respiratory tract, the internalization success of CNTs into cells, the speed at which uptake will occur, and the
potential for subsequent translocation. CNTs with higher aspect ratios, for example, are more likely to deposit deeper in the
lung, undergo frustrated internalization, take longer to achieve the ideal entry angle into cells, and are more likely to
translocate to pleura than CNTs with lower aspect ratios.
Surface Functionalization. CNTs often are treated or complexed with surfactants or other compounds designed
to functionalize them for a specific purpose (e.g., remain dispersed in water, interact with specific proteins). Some
surface functionalizations can also influence biopersistence of CNTs; for example, some CNTs functionalized to be
water soluble will be eliminated rapidly from the body, and CNTs functionalized to interact with specific proteins
might be more readily taken up by cells, including macrophages, which contribute to CNT clearance.
Contaminants. Several contaminants can be introduced during the manufacture of CNTs, including metals like
iron and nickel and various forms of carbon. Such impurities can "hitch a ride" with the CNTs to a biological surface
or interior of a cell, and contaminants that are not trapped within the central cavity of the CNT could be
bioavailable. Although purification processes often are used (with mixed success) to remove metal contaminants,
these processes tend to alter other characteristics of the CNTs, including length and morphology, which makes isolating the
effect of metal adsorbates on receptors difficult.
4-5
-------
4.1.1. Dose and Exposure Metrics
Dosages of free MWCNTs usually are expressed or quantified by mass concentration or by
particle/fiber count of individual MWCNTs or MWCNT bundles in a particular quantity of a medium
over an established period of time (Aschberger et al. 2010; Pauluhn. 2010a). Because exposures to
MWCNTs might involve only a small amount of mass but a large number of particles/fibers, the
appropriateness of traditional mass-based exposure and dose metrics for estimating and measuring
lexicologically relevant doses of engineered nanoscale materials (nanomaterials) is under debate
(Aschberger et al.. 2011).
Use of time-adjusted, mass-based
metrics historically has been the accepted
paradigm for quantifying exposure and dose for
most chemical substances. For example, mass
concentration has been used for more than 50
years as the metric for characterizing aerosol
exposures. Recent research has challenged the
ability of mass concentration to capture
appropriate nanomaterial dose-response
relationships, however, by illustrating that
airborne nanoscale particles—including both
engineered nanomaterials and nanoscale
particulate matter—can be more toxic than
larger airborne particles of the same
composition on a mass-for-mass basis (Maynard
and Aitken. 2007). Despite acknowledgment in
the scientific community that mass-based dose
metrics might not be appropriate for
nanomaterials, an alternative unifying metric for
characterizing dose has not yet been established,
and no single metric appears to be suitable for
all nanomaterials or exposure situations
(Pauluhn. 2010b: Mavnard and Aitken. 2007).
Additional Information Highlight Box 9:
9. Challenges related to MWCNT toxicokinetics
Although dosages of MWCNTs are often quantified using the
standard toxicological practice of mass concentration
measurements, this metric might not be appropriate for
nanomaterials such as MWCNT (Holgate, 2010). Many
nanoscale particles or fibers have greater toxicity on a mass-
for-mass basis than larger particles (Maynard and Aitken,
2007). Similarly, free MWCNTs could be more toxic than
bundled MWCNTs due to greater surface area-to-volume ratios
and kinetic differences that influence distribution of free versus
bundled MWCNTs (Johnston et al., 2010: Pauluhn, 2010a).
The dose metric (e.g., mass, fiber number, surface area) is
particularly important for inhalation exposures. Specifically, free
MWCNTs (or those in smaller bundles) are in the respirable
range and therefore can interact with biological receptors, with
the potential to evoke toxic effects; however, nonrespirable
particles (of greater mass) will likely not interact with receptors,
and thus will not have the same toxic potential. In addition,
larger or bundled particles might be targeted more effectively
by macrophages as a part of the immune response (Johnston
et al., 2010: Kim et al., 2010). Research supports this
relationship: Increasing the mass of bundled MWCNTs causes
the bundles to grow larger instead of creating more bundles
(Tan and Fugetsu, 2007).
In addition to affecting the dose-response relationship,
MWCNT bundling makes it difficult to measure MWCNTs in
exposure media (to determine administered dose) and in
tissues (to determine absorbed dose and dose uptake) (Chen
et al., 2011: Ponti et al., 2010: Monteiro-Riviere and Inman,
2006). Currently available analytical techniques do not provide
sufficiently accurate results, so using multiple techniques to
characterize MWCNTs is recommended (see Text Box 4-1).
4-6
-------
Some research in animal models has shown that toxic effects of some MWCNTs do follow a
more traditional dose-related curve based on the administered mass concentration of MWCNTs [for
example, Ma-Hock et al. (2009). as described in Section 5.1.3. and Asharani et al. (2008). as described in
Section 5.2.1.2]. indicating that mass can be considered an appropriate dose metric for some MWCNTs
under certain exposure conditions. To what degree variations in other MWCNT characteristics (e.g.,
length, width, bundling state) influence the appropriate application of mass-based metrics for MWCNTs
is not well understood. CNTs might induce toxic effects beyond those expected based on mass
concentration, however, which has been demonstrated in comparative tests of nanoparticulate carbon and
quartz, commonly used indicator compounds (Donaldson et al.. 2006).
Alternative characteristics that have been considered as potentially relevant dose metrics for some
nanomaterials include particle size, surface area, surface chemistry, particle count per particle size, and
particle charge (Aschberger et al.. 2011; Mavnard and Aitken. 2007). Although surface area has been
shown to be a better dose metric than mass in several rodent studies (Aschberger et al.. 2011; Sager and
Castranova. 2009; Tran et al.. 2000; Oberdorster. 1996). surface area has not been routinely measured or
recorded when examining occupational exposures (Aschberger et al.. 2011). Calculating surface area after
study completion is complicated by the differences in measurement techniques, the dynamic behavior of
MWCNTs (i.e., propensity to form bundles), and lack of thorough reporting.
Because the physical form of MWCNTs resembles fibers, other characteristics such as length,
diameter, aspect ratio, bundling state, and fiber count have been considered as characteristics potentially
relevant to quantifying potential exposures and doses of CNTs. Using fiber count as a dose metric can be
challenging, however, because MWCNT fibers generally are not uniform in size, and different sizes
might elicit different effects (see Text Box 5-1). The diameters of MWCNTs in general can range from 10
to 200 nm (Hou et al.. 2008). and the lengths can vary widely, often by tens of microns (Donaldson et al..
2006).
Bundling also can be a relevant characteristic for considering dose-response relationships.
Researchers have noted that MWCNTs tend to form bundles, which then can combine into small
"clumps," some of which are nonrespirable (Pauluhn. 2010a). For inhalation exposures, these larger
clumps are therefore less toxic than free MWCNTs and MWCNT bundles in the respirable range, despite
the larger mass concentration. Characterization of inhalation exposure and subsequent dose based on
bundle size or aerodynamic diameter could therefore be more appropriate than characterization based on
mass or particle count. Bundling of CNTs also could result in a toxic impact that is not observed with the
same mass of dispersed CNTs. For example, bundled CNTs could trigger an immune-system foreign-
body response because larger structures are potentially better recognized by macrophages (Johnston et al..
2010) (see Text Box 5-1). In vitro plant assays have shown that clustered MWCNTs fail to disperse
4-7
-------
throughout the culture, limiting exposure to a few cells (Tan and Fugetsu. 2007). Increasing the mass of
clustered MWCNTs did not increase dispersion, rather this caused the clusters to grow larger and
precipitate out of solution but did not cause a higher percentage of plant cells to be affected [Tan and
Fugetsu (2007). described further in Appendix G.5.1.1.2].
Aspect ratio, which refers to the ratio of a compound's length to diameter, has been shown by
some to be an important characteristic for driving exposure and dose of CNTs (Kim etal.. 2011; Poland et
al.. 2008). The fiber-like structure of CNTs can be considered similar to asbestos, causing many
researchers to predict that the toxicity of CNTs will be driven by differences in aspect ratio, with CNTs
having higher aspect ratios more frequently depositing deeper in the lungs and translocating to the pleura,
where mesothelioma, other cancers, and fibrosis (all effects of asbestos exposures) can occur
(Kim et al.. 2011) (see Additional Information Highlight Box 13). The utility of morphological
parameters like length, width, and aspect ratio for dose quantification is limited, however, by
inconsistencies in the literature regarding what constitutes "long" versus "short" or "high aspect ratio"
versus "low aspect ratio." These distinctions are usually relative, based on the materials compared in an
individual study; specific incremental changes in length, width, or aspect ratio have not yet been
correlated to quantitative changes in dose.
Because no single dose metric has been identified to date as capable of accurately predicting the
toxicity of MWCNTs, consideration of multiple characteristics together therefore has been proposed as a
potential alternative. For example, aspect ratio and bundling state might need to be considered together.
Long, thin CNTs (i.e., those with higher aspect ratios) can penetrate deeply into airways, while bundled
CNTs are more likely to deposit in the upper airway. Particles in the upper airway can be removed though
mucociliary processes, whereas deposits in deeper regions are more likely to persist or translocate from
the lung to other tissues where they might shift the location of toxic effects (Johnston et al.. 2010).
Measuring potential exposures outside of well-controlled experimental settings—for example, in
occupational settings—introduces a different set of challenges that can only be addressed currently by
using multiple instruments and analytical techniques. As described in Text Box 4-1 and Appendix B;
however, many of the techniques needed for detection, measurement, and characterization of MWCNTs
are limited by inadequate levels of detection or restrictive measurement ranges for morphological
parameters. For example, a common method for counting fibers in workplace air, the National Institute
for Occupational Safety and Health (NIOSH) Manual of Analytical Methods (NMAM) 7400 (NIOSH.
1994). does not detect fibers or bundles with diameters less than 0.25 urn and does not differentiate
between MWCNTs and other fibers (Gustavsson et al.. 2011). Another method that can be used to
estimate MWCNT mass concentrations in workplace air is NMAM 5040 [highlighted in NIOSH (2010)1.
4-8
-------
but this method measures elemental carbon as a proxy for MWCNTs and does not automatically account
for background contributions of other forms of particulate carbon (Birch, 2003).
The metrics primarily used to determine potential exposure levels of MWCNTs in occupational
air are total particle count by size fraction, total dust or particle mass, respirable20 dust or particle mass,
inhalable21 dust or particle mass, total elemental carbon mass, and total or respirable fiber count
(Xjustavsson et al.. 2011). The instruments required to capture these measurements include a suite of real-
time, direct monitoring particle samplers and particle counters and a variety of area and personal air
filters. The particles and fibers collected by these samplers and filters, however, are not limited to
MWCNTs; instead, all particles or fibers within a certain size range are collected and counted, which
might lead to overestimation of exposure concentrations. This lack of specificity introduces a degree of
uncertainty that can be reduced only with adequate characterization of the samples collected on the filters
(Gustavsson et al.. 2011). In general, characterization involves a form of electron microscopy to verify
morphological features and energy-dispersive spectroscopy to verify the chemical identity of the samples.
In summary, most estimates of workplace exposure are derived by relating real-time data on particle or
fiber counts with filter samples analyzed to determine particle or fiber mass, particle or fiber morphology,
and chemical composition; all analytical techniques involved in this multistep estimation of exposure
concentrations have analytical limitations that produce estimates with varying amounts of uncertainty
(Dahmetal.. 201 la).
4.1.2. Concentrations in Environmental Media and Indoor
Environments
As described in the previous section, exposures can be estimated by combining knowledge of
concentrations in exposure media with assumptions about contact of humans, biota, or abiotic surfaces
with those media. The following sections describe the information available on concentrations of
MWCNTs and related substances in environmental media (i.e., air, water, soil).
20The respirable particulate fraction is generally defined as the "fraction of inhaled airborne particles that can
penetrate beyond the terminal bronchioles into the gas-exchange region of the lungs" (WHO. 1999). NIOSH
considers particles with aerodynamic diameters >10 ^m to be larger than respirable (Bartlev and Feldman. 1998).
21 The inhalable particulate fraction is generally defined as the "fraction of a dust cloud that can be breathed into the
nose or mouth" (WHO. 1999). which could include both particles with sizes within the respirable range and particles
that are larger than respirable.
4-9
-------
Additional Information Highlight Box 10:
10. Weaknesses of current analytical techniques
Although analytical techniques for identifying, quantifying, and characterizing MWCNTs are available, they often cannot
accurately characterize MWCNTs in complex environmental matrices for several reasons. First, traditional analytical
techniques were not developed for application to engineered nanomaterials (such as MWCNTs) at the low concentrations
likely to be found in environmental matrices (Petersen et al., 2011c). Second, the chemical transformations that MWCNTs
undergo during environmental transport might impede detection by standard analytical methods (von der Kammer et al.,
2012). Third, due to the colloidal associations of many engineered nanomaterials (including MWCNTs) in the environment,
their physicochemical properties depend on the environment in which they are found (von der Kammer et al., 2012). Finally,
the processes involved in isolating, observing, and quantifying engineered nanomaterials could alter the physicochemical
properties of the analyte of interest and introduce artifacts (von der Kammer et al., 2012). Appendix B describes some of the
specific disadvantages of several analytical techniques for identifying, quantifying, and characterizing MWCNTs in different
environmental matrices. Some additional analytical techniques used to detect CNTs (including both single-walled and
multiwalled) include: near infrared fluorescence spectroscopy in aquatic systems (Schierz et al., 2012: Rocha et al., 2011):
thermogravimetry in complex mixtures such as soot, coastal sediment, and biological macromolecules (Plata et al., 2012b):
programmed thermal analysis in surface water, tap water, wastewater, sediments, and various biological matrices (Doudrick et
al., 2012): isotopic labeling in biological samples (Wang et al., 2013): combined programmed thermal analysis/Raman
spectroscopy and thermal optical transmittance/reflectance in urban air (Doudrick et al., 2012): and microwave irradiation in
agricultural samples (Irin et al., 2012). Although some new analytical techniques are producing promising results, better
methods are needed to extract, clean up, separate, and store MWCNTs to improve efficiency, sensitivity, and specificity (von
der Kammer et al., 2012). Further, the limited number of studies that use the same analytical technique(s) impedes
comparison of results between studies (Petersen et al., 2011c), in a way similar to differences between studies of toxicity
outcomes with different types of MWCNTs.
4.1.2.1. Outdoor Air
A recent review of toxicity and exposure to CNTs indicates that ambient exposure to CNTs is
possible, but very little outdoor environmental sampling data are available (Aschberger et al.. 2010).
Dahm et al. (2011 a) measured background elemental carbon (inhalable fraction) outside CNT primary
and secondary manufacturing facilities. Concentrations ranged from not detected (limit of detection 0.2 to
0.5 ug elemental carbon/filter) to 0.76 ug/m at MWCNT manufacturing facilities. This information is of
limited utility, however, because elemental carbon particles can be produced by many sources, and no
electron microscopic analysis was conducted to determine if the particles collected included CNTs.
Researchers have found MWCNTs in methane or propane flames from kitchen stoves, and
automotive exhaust is thought to be a source of MWCNTs (Lagally etal. 2012; Aschberger et al.. 2010).
CNTs were found in the lung tissues of World Trade Center patients following the collapse of the
buildings on September 11, 2001 (Wu etal.. 2010).
As summarized in Table 3-2, two life cycle-based models estimated flow of CNTs and other
nanomaterials from the products containing them to environmental compartments (Gottschalk et al.. 2009;
Mueller and Nowack. 2008). The predicted environmental concentrations (PECs) of CNTs in the air were
estimated by Mueller and Nowack (2008) as 1.5 x 10"3 and 2.3 x 10"3 ug/m3 for the realistic exposure and
high exposure scenarios, respectively; concentrations in airborne dust were not examined. Gottschalk
4-10
-------
et al. (2009) estimated a mode PEC of 1 x 10'6 ug/m3 CNTs in United States air for 2008. Differences
between the models and the resulting estimates are discussed in Section 3.5.
4.1.2.2. Aquatic Systems - Sediment and Surface Water
No data were found on environmental concentrations of MWCNTs in aquatic environments, but
as summarized in Table 3-2. two substance flow analyses have estimated PECs of CNTs in surface water
(Gottschalk et al.. 2009; Mueller and Nowack. 2008). Mueller and Nowack (2008) estimated surface
water CNT PECs of 5 x 10"4and 8 x 10"4 ug/L for the realistic and high exposure scenarios, respectively.
Gottschalk et al. (2009) estimated not only a mode PEC of 1 x 10"6 ug/L for CNTs in surface water, but
also an annual increase of 4.6 x 10"2 ug/kg-year to United States sediment.
4.1.2.3. Terrestrial Systems - Soil
No data were found on environmental concentrations of MWCNTs in surface soil samples, but as
summarized in Table 3-2. two substance flow analyses have estimated PECs of CNTs in soil (Gottschalk
et al.. 2009: Mueller and Nowack. 2008). Mueller and Nowack (2008) estimated soil CNT PECs of
1 x 10"2and 2 x 10"2 ug/kg for the realistic and high exposure scenarios, respectively. Gottschalk et al.
(2009) estimated an annual increase of 5.6 x 10"4 ug/kg-year to United States soil.
4.1.2.4. Occupational Settings -Air
No studies were found that measured MWCNT concentrations in air in facilities where textiles
containing MWCNTs are manufactured. Multiple studies have collected particles and fibers in workplace
air to attempt to estimate MWCNT concentrations at the emission source, in area air, and in the personal
breathing zone of workers in small laboratories or research and development facilities (Johnson et al..
2010; Leeetal.. 2010a: Methner et al.. 2010: Bello et al.. 2008: Han et al.. 2008). as well as in larger pilot
plants and manufacturing and handling facilities (Dahm et al.. 201 la: Lee et al.. 2010a: Takava et al..
2010). Some of these studies are discussed below, and Appendix E. Table E-6 presents additional
information on particle, fiber, and MWCNT concentrations for the studies that reported quantitative
concentration values associated with MWCNTs.
Concentrations in MWCNT Research Laboratories
Han et al. (2008) measured concentrations of total particles in the area air and personal breathing
zones of workers in an MWCNT research facility. Two particle sizers were used to count particles with
sizes ranging from 14 to 630 nm and 0.5 to 20 um, respectively, and a portable aethalometer measured the
4-11
-------
mass of carbon black in the total particulate matter in the air. All fibers with aspect ratios greater than 3:1
were collected on filters and analyzed using electron microscopy, and MWCNTs were distinguished from
asbestos fibers using energy-dispersive spectroscopy. During the blending process, which creates a
uniform size-distributed CNT powder, particulate matter concentrations in the air near the open blender
ranged from 434.5 ug/m3 without exposure controls to no detection (limit of detection not reported) with
exposure controls. The maximum MWCNT number concentration ranged between 172.9 (area air sample)
and 193.6 (personal air sample) fibers per cm3 air during blending without exposure controls, and
between 0.018 (personal air sample) and 0.05 (area air sample) fiber per cm3 air during blending with
exposure controls. During weighing and spraying, particulate matter concentrations ranged from 36.6
(area air sample) to 193.0 (personal air sample) ug/m3 without exposure controls and from below the level
of detection (area air sample; limit of detection not reported) to 30.9 ug/m3 (personal air sample) with
exposure controls. The maximum MWCNT number concentrations during weighing and spraying were
below detection (limit of detection not reported) in the absence of controls, and up to 1.997 fibers per cm3
air after controls (Han et al., 2008). The reason for the increase in MWCNTs following implementation of
exposure controls (in this case, the control was "a simple fan") was not discussed. The maximum
MWCNT length observed was 1.5 um, which is smaller than the World Health Organization's minimum
length of 5 um for classification as a fiber (Aschberger et al., 2010).
Using a suite of real-time particle sizers, particle counters, and filters with electron microscopy
and energy-dispersive spectroscopy analyses, Methner et al. (2010) measured and characterized
particulate matter in carbon-based nanomaterial research and development facilities. The highest particle
number concentration for particles ranging in size from 10 to 1,000 nm was measured when engineering
controls were turned off during the opening of an MWCNT growth chamber of a pulsed laser deposition
reactor. The maximum particle number concentration was 42,400 particles per cm3 in the absence of
engineering controls, but when the same activity was performed in a sealed system with vacuum exhaust,
the particle number was reduced to 300 particles per cm3 (Methner et al.. 2010).
Johnson et al. (2010) used real-time particle sizers to count total particles per liter air for six size
cuts (300, 500, 1,000, 3,000, 5,000, and 10,000 nm) and per cubic centimeter air for the cumulative 10- to
1,000-nm size fraction of particles released to the air in a laboratory while two tasks were being
performed with raw and functionalized MWCNTs. The first task involved weighing MWCNTs and
transferring them to a beaker of stirring water, and the second task involved sonicating a previously
mixed solution containing reconstituted water and 100 mg/L MWCNTs with 100 mg/L natural organic
matter. Filter samples also were collected at the emission source (i.e., as close as possible to the
instruments used for each task) and in area air, and samples were analyzed using electron microscopy and
energy-dispersive spectroscopy. In general, particle number concentrations in the air were inversely
4-12
-------
proportional to particle size, with either zero or very few particles detected for the 5,000- and 10,000-nm
size cuts for both types of MWCNTs used in both tasks. The maximum background-adjusted particle
number concentrations for the raw MWCNTs occurred within the 300-nm size fraction; these
concentrations were 123,403 particles/L air (above the upper limit of quantification) during weighing and
transferring and 42,796 particles/L air during sonication (Johnson et al.. 2010). The particle number
concentrations measured for the functionalized MWCNTs (MWCNT-OH) exhibited different trends from
those for the raw MWCNTs. First, no additional 300-nm particles were detected above the background
level during weighing and transferring. Second, particle counts were higher during sonication of
functionalized MWCNTs than during weighing and transferring, which is opposite of the trend observed
for the raw MWCNTs. Whereas the maximum background-adjusted particle number concentration was
3,065 particles/L air (500-nm size fraction) during weighing and transferring, the maximum concentration
was 144,623 particles/L air (above the limit of quantification; 300-nm size fraction) during sonication of
functionalized MWCNTs. The particle number concentrations measured for the cumulative 1- to 1,000-
nm size range were 1,576 and 2,776 particle/cm air for the raw MWCNTs and 676 and 726 particles/cm3
for the functionalized MWCNTs during weighing/transferring and sonicating, respectively. Johnson et al.
(2010) proposed that the cumulative measurements do not follow the same trends as the size cut
measurements because of the inclusion of particles smaller than 300-nm in the cumulative particle
counter.
Bello et al. (2008) used a real-time particle sizer and a particle counter to count total particles
with sizes ranging from 5.6 to 560 nm and 10 to 1,000 nm, respectively in the area air of a university
research laboratory. Personal air samples also were collected on filters and analyzed using scanning
electron microscopy and energy-dispersive spectroscopy. Bello et al. (2008) found that removal of
MWCNTs from the reactor furnace and detachment of MWCNTs from the nanotube growth substrate
during chemical vapor deposition (CVD) did not increase total airborne particle concentrations compared
to background. Additionally, no MWCNTs were observed in the personal air sample of a furnace operator
(Bello et al.. 2008).
In a later study, Bello et al. (2009) measured particulate matter, respirable particulate matter, and
respirable fibers in a laboratory during dry and wet machining of composite materials with and without
CNTs. Particle sizers were used to detect and count all particles with sizes ranging from 5 nm to 20 urn,
and particles and fibers in the respirable range were collected on filters near the source (i.e., 10 cm from
the machined composite) and in the breathing zone of the operator; filter samples were analyzed using
electron microscopy. Although the dry-cutting process did result in statistically significant increases in
airborne particles and fibers, no statistically significant differences were noted in the particle number,
particle sizes, or total dust generated by dry cutting the composites with and without CNTs. Furthermore,
4-13
-------
analyses of filter samples revealed no single or bundled CNTs in the particles and fibers collected, and no
CNTs were observed "sticking out" of the CNT composites (Bello et al., 2009).
Concentrations in MWCNT Manufacturing and Packing Facilities
Dahm et al. (2011 a) used a series of filters to measure the inhalable size fractions of elemental
carbon and determine CNT and carbon nanofiber "structure" counts (defined as "single CNTs to large
agglomerates" viewed using electron microscopy) in the area air and personal breathing zones of six
pilot-scale CNT or carbon nanofiber primary and secondary manufacturing facilities. Sampling was
conducted while workers performed various tasks, including harvesting, sonicating, weighing, extruding,
manually transferring, and mixing MWCNTs; spray coating a product with an MWCNT solution; milling
MWCNT composites; and collecting and disposing of waste from MWCNT work areas. The elemental
carbon concentrations in the personal breathing zone samples generally were higher than the area air
samples. Inhalable elemental carbon concentrations in personal breathing zones ranged from 1.13 ug/m3
(sonicating, sieving, and spray coating) to 2.74 ug/m3 (harvesting) at the primary MWCNT manufacturing
facilities and from 0.8 ug/m3 (office work outside lab space) to 7.86 ug/m3 (extrusion, weighing, and
batch mixing) at the secondary MWCNT manufacturing facilities. The CNT structure counts in the
personal breathing zones ranged from 0.010 structure/cm3 (sonicating, sieving, and spray coating) to
0.399 structure/cm3 (harvesting) at the primary facilities, and from none observed (weighing, sonicating,
milling) to 0.242 structure/cm3 (extrusion, weighing, and batch mixing) at the primary and secondary
MWCNT manufacturing facilities (Dahm et al.. 201 la).
Inhalable elemental carbon concentrations in area air samples from Dahm et al. (2011 a) ranged
from not detected (sonicating, sieving, and spray coating; limits of detection ranged from 0.2 to 0.5 ug
elemental carbon/filter) to 4.62 ug/m3 (harvesting) and from not detected (weighing, sonicating, milling,
and mixing) to 1.01 ug/m3 (extrusion, weighing, and batch mixing) at the primary and secondary
MWCNT manufacturing facilities, respectively. The CNT structure counts ranged from none observed
(production and harvesting) to 0.134 structure/cm3 (harvesting at a different facility) and from none
observed (weighing, milling) to 0.008 structure/cm3 (extrusion, weighing, and batch mixing) in the area
air samples at the primary and secondary MWCNT manufacturing facilities, respectively (Dahm et al..
201 la).
Lee et al. (2010a) collected filter samples to measure respirable dust concentrations in the area air
and personal breathing zones of workers at three MWCNT manufacturing facilities and four research and
development laboratories throughout a normal workday. Fibers with aspect ratios greater than 3:1 were
collected on the filters and analyzed using electron microscopy, and MWCNTs were chemically identified
using energy-dispersive spectroscopy. A suite of particle sizers, differential mobility analyzers, and
4-14
-------
particle counters also was used to count particles with sizes ranging from 14 to 500 nm, and a dust
monitor was used to capture number concentrations of particle ranges from 0.25 to 32 urn in diameter.
A portable aethalometer measured the mass of carbon black in the total particulate matter in the air. Lee et
al. (2010a) generally found that the highest increases in particle number concentrations compared to
background were observed following the opening of the CVD chamber after MWCNT synthesis.
Increases in carbon black concentrations at this time were minimal, however, suggesting that most of the
particles released were more likely to be metal catalysts than MWCNTs. Furthermore, the authors
reported only one measurement of a detectable amount of MWCNTs on one filter from a single facility
(0.00312 tube/cm3); the study authors could not determine whether the lack of MWCNT detection
reflected a lack of MWCNTs in workplace air or flaws in the sampling process or analytical methods.
Total dust and respirable dust concentrations were measured in a study of two MWCNT packing
facilities, one of which was manually operated and the other automated [(Takaya et al.. 2010) English
translation available only for abstract]. Total dust concentrations in the area air, of both MWCNT packing
facilities, were approximately 240 ug/m3. Both total and respirable dust concentrations, however, were
substantially higher in the manual packing facility (total: 2,390 ug/m3; respirable: 390 ug/m3) than in the
automated packing facility (total: 290 ug/m3; respirable: 80 ug/m3).
4.1.2.5. Residential Settings-Air and Dust
No data were found on concentrations of MWCNTs in household air or dust.
4.1.2.6. Nonresidential Settings -Air and Dust
No data were found on concentrations of MWCNTs in nonresidential air or dust.
4.1.2.7. Transportation, Including Automobiles and Airplanes -Air and Dust
No data were found on concentrations of MWCNTs in air or dust in vehicles or aircraft.
4-15
-------
4.2. Human Exposure and Kinetics Leading to Dose
Limited data were found that measured or quantified human exposure to MWCNTs. Data on
concentrations of MWCNTs measured in media such as air, soil, or dust in various settings (described in
Section 4.1.2). however, can be used in conjunction with activity pattern and other exposure factor data
[such as those described in The Exposure Factors Handbook (U.S. EPA. 2011b)1 to inform estimates of
potential exposure through the various exposure pathways and scenario characteristics described in this
section.
The types of human exposure scenarios described here can be divided into four broad groups:
occupational, consumer, general public, and highly exposed populations. For the purposes of this case
study, occupational exposures include occupational exposures during synthesis, processing, or handling
of MWCNTs; manufacturing of flame retardants, application of the flame retardants to textiles, or textile
finishing and upholstering; storage of the MWCNTs, flame-retardant formulations, treated textiles, or
upholstered products; disposal of MWCNTs, flame-retardant formulations, treated textiles, or upholstered
products; and repurposing or recycling of treated upholstery textiles and end-user products (e.g.,
furniture). Consumer exposure scenarios include the intended or unavoidable use of treated upholstery
textiles in residential and nonresidential spaces, including on household or institutional/office furniture, in
vehicles, and in aircraft; unintended uses of treated upholstery textiles or end-use products such as reuse
or repurposing of furniture for something other than its original intended use; or recycling of upholstery
textiles for new uses. General public exposure includes primary exposure to members of the community
near manufacturing, disposal, or recycling facilities and secondary exposure to the general public through
environmental routes such as air, soil, or water. Highly exposed populations refers to exposure scenarios
that are expected to occur via similar pathways as outlined for consumers and the general public, but
where exposure levels are expected to be higher due to key differences in population characteristics such
as those described in The Child-Specific Exposure Factors Handbook (U.S. EPA. 2008a).
4-16
-------
4.2.1. Occupational Exposure Pathway Scenarios
Priority Research Area: Exposure Route
o v
E £
« B
uj r
o
Q.
13/13
0/13
0/13
• - 1 participant's vote
Importance
L - Low: Least Important
M - Medium: Possibly Important
H — High: Important
Confidence
L - Low: Not Confident
M - Medium: Somewhat Confident
H - High: Confident
H M L M H M L
E-RRF Confidence Rating E-RRF Confidence Rating E-RRF Confidence Rating
All 13 RTI workshop participants identified human occupational exposure as important to
risk assessment. The 13 participants were asked to rate the importance of ingestion,
inhalation, and dermal occupational exposures, as well as their confidence that the
existing data on these risk relevance factors could support risk-management decisions.
Based on this information, the inhalation exposure route was deemed the highest priority
for research in this area, followed by ingestion and dermal exposures. See Section 6.3.3.1
for more information on this priority area and Section 1.1.3 for a detailed explanation of the
prioritization process.
DecaBDE Can Inform MWCNT Assessment
The exposure routes through which workers might be exposed to decaBDE (i.e., inhalation, oral, and dermal) are similar
across the material synthesis, product manufacturing, storage, and disposal stages of the product life cycle. Certain exposure
routes, however, are more likely for occupations that are specific to certain life-cycle stages (e.g., working in material synthesis
facilities compared to working in recycling facilities). Industrial hygiene and personal protective equipment can reduce
exposure to decaBDE, although workers involved with some life cycle stages might be less likely to take these precautions.
No data are available, however, on the relative or estimated amounts of exposure at each life-cycle stage.
• Inhalation exposure to decaBDE, by-products, or dust containing the material suspended in air could occur during the
stages of material synthesis (e.g., handling decaBDE powders), product manufacturing (e.g., abrading treated textiles
during tailoring), storage (e.g., volatilized components of coating, dust), and disposal/reuse/recycling (e.g., abrading
or destroying textiles) (EU, 2002).
• Oral exposures could occur secondarily from inhaling decaBDE and then subsequently ingesting it or by ingesting
decaBDE that deposits from the air onto the skin, food, or food-contact surfaces during all life cycle stages. Dust
levels are expected to be higher in textile storage facilities, which could result in increased transport of decaBDE
adsorbed to dust and therefore increased potential for exposure during this stage.
• Dermal exposures could occur from decaBDE present in dust that deposits on skin or skin-contact surfaces at all life
cycle stages. During product manufacturing, the liquid flame-retardant coating could be spilled directly on skin; while
during disposal stages, physical contact with decaBDE flame-retardant coating on upholstery textiles also is possible.
Occupational exposures to decaBDE are generally expected to be similar to MWCNT exposure throughout the product life
cycle, given the similarity in application (see the DecaBDE Comparison Boxes in Section 2.4 and Section 2.5). Analysis of
decaBDE data thus helps identify important research questions that could inform future risk assessments of MWCNTs. For
example: Which stages of the MWCNT life cycle present the greatest occupational hazard to those working with MWCNTs and
MWCNT products? Is occupational risk of exposure greater by a particular route for each life-cycle stage? Are MWCNTs
expected to volatize in air or absorb to dust and deposit on surfaces in manufacturing or disposal facilities? Does modifying
physicochemical characteristics of MWCNTs (e.g., aspect ratio, surface functionalization) influence the dominant occupational
exposure routes? See Appendix H for more information on occupational exposure to decaBDE.
4-17
-------
Limited data were found to determine the extent of occupational exposures to MWCNTs during
the material synthesis, processing, and handling phases or to the flame-retardant product during
formulation, application, storage, and disposal phases. See Section 4.1.2.4 for MWCNT concentrations
measured in occupational settings, which could be applied with the exposure pathways and scenario
characteristics described below to estimate potential exposures through scenario evaluation.
Additional Information Highlight Box 11:
11. MWCNT dermal absorption
Although several studies highlight the potential for dermal exposure to MWCNTs throughout the life cycle of MWCNT flame-
retardant coatings in upholstery textiles (Uddin and Nyden, 2011b: Aschberger et al., 2010: Johnson et al., 2010: Lam et al.,
2006: Maynardetal., 2004), dermal penetration by MWCNTs has rarely been observed. Numerous studies on applications of
other engineered nanomaterials, such as sunscreen formulations containing nano-Ti02, have shown that although dermal
exposure might occur, dermal penetration is unlikely. With few exceptions (Sadriehetal., 2008: Kerteszetal., 2005: Menzel et
al., 2004), most dermal penetration studies have found clear evidence that nano-Ti02 in sunscreen formulations do not
penetrate beyond the stratum corneum or hair follicles, and it does not penetrate into living cells of healthy skin (Kiss et al.,
2008: Mavonetal., 2007: Pinheiroetal., 2007: Gamer etal., 2006: Lademannetal., 1999: Dussert and Gooris, 1997). In their
summary of evidence regarding the interaction of various nanoparticles with skin, Elder et al. (2009) concluded that dermal
absorption of nanoparticles does not appear to occur readily but can take place under certain conditions, especially when skin
is damaged. Although the behavior of nano-Ti02 and other nanoparticles cannot be extrapolated to MWCNTs, these findings
demonstrate that dermal exposure to engineered nanomaterials, even when the intended use is dermal application (e.g.,
sunscreen), does not necessarily lead to dermal penetration. Such evidence implies that the dermal toxicity of MWCNTs might
be driven by the lack of dermal penetration. Other studies have shown that dermal toxicity may be influenced by MWCNT
properties such as aggregation state. One study was identified that investigated dermal effects in vitro based on different
degrees of MWCNT dispersal. Vankoningsloo et al. (2010) found that MWCNTs were more cytotoxic when sonicated than
when agglomerated; however use of a dispersal agent in sonication reduced these effects. Data are lacking, however, on the
extent to which modifications to MWCNTs that occur during production (see Additional Information Highlight Box 7) or
transformation during subsequent stages of the product life cycle (see Additional Information Highlight Box 6) —which can
lead to different MWCNT aggregation states or other changes in physical properties —influence dermal absorption.
4.2.1.1. Synthesis, Processing, and Handling
As discussed in Section 2.2.2. synthesis of MWCNTs is achieved by one of three processes:
CVD, arc discharge, and laser ablation. Although many facilities use engineering controls (e.g., fume
hoods, closed production systems, high-efficiency particulate air-filtered vacuums) and require workers to
wear personal protective equipment (e.g., gloves, respirators, paper face masks, safety glasses, lab coats,
Tyvek clean suits) to minimize exposure to MWCNTs (Dahm et al., 201 la), not all facilities comply with
the General Safe Practices for Working with Engineered Nanomaterials in Research Laboratories, which
outlines the recommendations by NIOSH (2012). Many facilities do not employ the same level of
protective measures, and in many cases, the filtration technologies and personal protective equipment are
not appropriate for or sufficiently protective against exposures to nanomaterials (Dahm etal.. 201 la).
Therefore, exposure during handling and other operations might still occur, for example, when a reaction
chamber is opened to recover MWCNTs; while extracting, weighing, or manually transporting materials;
4-18
-------
or during maintenance and cleaning of equipment (Dahm et al., 201 la; Fleury etal. 2011; Aschberger et
al., 2010). Workers performing each operation are expected to be exposed to peak concentrations of
MWCNTs for only a short time while carrying out MWCNT handling tasks, but multiple production
cycles might occur within a day, resulting in several opportunities for short-duration, acute exposures
throughout the workday (Dahm et al.. 201 la; Lee et al.. 2010a). Most occupational exposure studies to
date have examined these short-duration, task-specific exposures instead of full-shift exposures, and task-
specific exposures have been evaluated only for a limited set of handling operations (see Section 4.1.2.4).
In general, MWCNTs observed in air and settled on surfaces in occupational environments during
synthesis, processing, and handling are in bundled form, but exposure to single MWCNTs is possible
(NIOSH. 2010). The pathways through which workers might be exposed to MWCNTs and MWCNT
bundles during synthesis, processing, and handling scenarios are described below:
• Inhalation. Handling dry powder might be the activity most likely to lead to inhalation
exposures during production. Dahm et al. (2011 a) and Johnson et al. (2010) observed that
workers handling dry powder often turned off vents, hoods, fans and other engineering
controls to avoid disturbing and dispersing MWCNTs. Handling processes such as weighing,
blending, transfer to containers, or maintenance also could result in inhalation exposure
(Dahmetal.. 201 la; Fleurvet al.. 2011; Aschberger etal.. 2010).
• Oral. Secondary oral exposures might occur if inhaled MWCNTs or MWCNTs that deposit
on the skin, food, or food-contact surfaces are subsequently ingested.
• Dermal. Particles generated during manufacturing and processing of CNTs can settle on the
skin of workers if proper personal protective equipment is not worn (Lam et al., 2006).
A study evaluating occupational exposure to CNTs during synthesis, processing, and
handling estimated (using adsorbed metals as proxy) that, on average, 0.2 to 6 mg of single-
walled CNTs (SWCNTs) are deposited on the gloves covering each hand of workers during
routine operations. Although the cotton gloves worn by workers could have adsorbed more
CNTs than bare skin or latex, the study illustrates that dermal exposure to CNTs could occur
in laboratory settings (Maynard et al.. 2004).
4.2.1.2. Formulation of Flame Retardant, Application to Textiles, Upholstering
No data were found on occupational exposures to MWCNTs during formulation of the flame
retardant, application of the flame retardantto textiles, or textile finishing and upholstering. Furthermore,
information on the processes for preparing MWCNT flame retardants and for applying them to textiles
are lacking. To confer the desired flame-retardant properties of MWCNTs to the textile product, however,
MWCNTs must be well dispersed in a polymer medium. To promote dispersion, MWCNTs are
sometimes ground or pulverized, which could lead to the release of single MWCNTs or bundles.
MWCNTs also might be mixed or sonicated, which could generate airborne water droplets, or mists,
4-19
-------
containing nanomaterials that then can be inhaled or deposited on surfaces (Fleury et al.. 2011;
Aschberger et al.. 2010; Johnson et al.. 2010).
As described in Section 2.2.4.1. one industry representative reported that textiles can be
immersed or spray coated with MWCNT flame retardants. The immersion method could result in worker
exposures, and exposures are expected to be highest during equipment handling and cleaning. Should
MWCNT flame retardants be sprayed onto textiles, however, MWCNTs in the wet polymer matrix might
be released as mists, which can occur even with nonvolatile liquids (U.S. EPA. 2005).
Exposures to MWCNTs also might occur when the treated upholstery textile is machined, drilled
(Aschberger et al.. 2010). or otherwise abraded during the textile finishing or upholstering processes. Wet
machining and dry machining of advanced nanomaterial composite systems were evaluated for generation
of respirable CNTs. Wet-cutting methods were not found to produce exposures significantly different
from background, while dry-cutting methods created statistically significant quantities of nanoscale and
fine particles and fibers composed of the composite material (i.e., no single or bundled CNTs were
observed in the samples) (Bello et al.. 2009).
Equipment cleaning can be a key contributor to work exposure during product manufacture,
application, and upholstering. One study identified equipment cleaning as one of the most important
occupational exposure scenarios because it often requires workers to be in direct contact with molten
polymers and residues containing CNTs (Fleury et al.. 2011).
The pathways through which workers might be exposed to single MWCNTs or MWCNT bundles
during general formulation of the flame retardant, application of the flame retardantto the textile, and
textile finishing and upholstering scenarios are expected to be comparable to those described in Section
4.2.1.1 on exposures during synthesis, processing, and handling. Additional considerations pertaining to
exposures to MWCNTs in combination with polymer ingredients, textile fibers or scraps, or other product
constituents during these scenarios are described below:
• Inhalation. Spray coating textiles with MWCNT flame retardants could result in inhalation
exposures to mists containing MWCNTs embedded in a liquid polymer mixture. Dry-cutting
MWCNT-treated textiles during tailoring and upholstering could lead to inhalation of fine
and ultrafine particles comprising MWCNT-polymer composites and textile dusts.
• Oral. Secondary oral exposures might occur if inhaled MWCNTs in mists and particulate
form or the MWCNT mists or particles that deposit on the skin, food, or food-contact
surfaces are subsequently ingested.
• Dermal. MWCNTs in mists and particulate form generated during product manufacturing
can land on the skin of workers if proper personal protective equipment is not worn
(Aschberger et al.. 2010; Johnson et al.. 2010; Lam et al.. 2006). The liquid flame-retardant
coating also can be spilled directly onto the skin.
4-20
-------
4.2.1.3. Storage of MWCNTs, Flame-Retardant Formulations, Treated Textiles, and
Upholstered Products
As described in Appendix G.2.2, MWCNTs and the flame-retardant formulations to which they
are added are expected to be stored in sealed receptacles that would limit potential for worker exposures
to these materials during storage. Defective packaging and accidental spills or releases, however, could
lead to rare exposures during storage operations.
Although no information was identified regarding procedures for storing treated upholstery
textiles, these products are likely packaged to protect them from exposure to elements like water and light
that could damage their aesthetics. Such packaging also is expected to limit exposures of workers to the
flame-retardant coatings. Once the textiles have been applied as upholstery to end-use products, these
products also are expected to be enclosed in protective packaging. Some surfaces of bulkier products
(e.g., furniture), however, might remain uncovered, which could lead to worker exposures during storage
operations, or exposures might occur during application and removal of packaging materials to and from
the product. Dust also can accumulate in storage facilities that frequently store textiles and textile
products, and MWCNTs that escape from the product matrix could sorb to dust particles. Ventilation
technologies and other contamination-prevention strategies like those used by manufacturing facilities are
not expected to be in place in storage facilities. Dust that has settled on surfaces in storage facilities can
be disturbed by worker operations, resuspended, and transported to other locations.
Although MWCNTs are not expected to be highly volatile, off-gassing of more volatile
components of the treated textiles might occur during storage of treated textiles or upholstered products.
Furthermore, due to the additive nature of MWCNT flame retardants, covalent bonding between the flame
retardant and the textile does not occur, suggesting that flame-retardant coatings that are loosely attached
to the textile surface might slough off during storage or handling. Because MWCNT flame retardants are
generally added to the back of the textile, however, the likelihood of this detachment seems low.
No data were found on occupational exposures to MWCNTs during storage throughout the
product life cycle of flame-retardant upholstery textile coating. The pathways through which workers
might be exposed to MWCNTs alone or MWCNTs in combination with polymer ingredients, textile
fibers or scraps, dusts, or other product constituents during storage of MWCNTs and MWCNT flame-
retardant formulations are expected to be comparable to those described in Sections 4.2.1.1 and 4.2.1.2
(exposures during synthesis, processing, and handling and during formulation of the flame retardant,
application to textiles, and upholstering). The pathways through which workers might be exposed to
MWCNT bundles or MWCNTs adsorbed to dust during storage of treated textiles and upholstered
products are expected to be inhalation, oral, and dermal. Workers could inhale volatile components of the
flame-retardant coating or MWCNTs adsorbed to dust in storage facilities, particularly facilities that are
4-21
-------
not well ventilated. Higher levels of dust in textile storage facilities could lead to increased transport of
MWCNTs adsorbed to dust. This could result in oral exposures to MWCNTs in dust transported to break
rooms, homes (via clothes), and other locations where MWCNTs adsorbed to dust can be unintentionally
ingested while eating or due to hand-to-mouth activity. Additionally, MWCNTs adsorbed to dust could be
resuspended by worker activities and deposit on the skin of workers if proper personal protective
equipment is not worn.
4.2.1.4. Disposal and Recycling of MWCNTs, Flame-Retardant Formulations, Treated
Textiles, and Upholstered Products
As described in Section 2.5. large-scale disposal, recycling, and reuse of MWCNTs and the
flame-retardant formulations to which they are added are unlikely, but containers used to store these
products might enter the waste stream, and workers at disposal and recycling facilities could be exposed
to product residues remaining in these containers.
Disposal and recycling of treated textiles and upholstered products, however, is prevalent. Mixing
and compacting of waste for land-filling; cleaning, shredding, blending, melting, and spinning scrap
textiles for recycling; and incomplete incineration of treated upholstery textiles all could result in
exposure of workers to MWCNTs (Chaudhry et al.. 2009). primarily in combination with other product
constituents and dusts.
No data were found on occupational exposures to MWCNTs during disposal and recycling
throughout the product life cycle of flame-retardant upholstery textile coating. The pathways through
which workers might be exposed to MWCNTs during general disposal and recycling of MWCNTs and
flame-retardant formulations are expected to be comparable to those described in Section 4.2.1.1
(exposures during synthesis, processing, and handling); worker exposure pathways for MWCNTs in
combination with polymer ingredients, textile fibers or scraps, or other product constituents during
disposal and recycling of treated textiles and upholstered products are expected to be similar to those
described in Section 4.2.1.2 (exposures during formulation of the flame retardant, application to textiles,
and upholstering) and Section 4.2.1.3 (exposure during storage and distribution) for these products.
4-22
-------
4.2.2. Consumer Exposure Pathway Scenarios
Priority Research Area: Exposure Route
Human.
Consumer
o o
C O i
£ C '
LU
O
Q.
£.
9/13
4/13
0/13
• - 1 participant's vote
Importance
L - Low: Least Important
M —Medium: Possibly Important
H —High: Important
Confidence
L - Low: Not Confident
M — Medium: Somewhat Confident
H -High: Confident
~R M IT" H M L H M L
E-RRF Confidence Rating E-RRF Confidence Rating E-RRF Confidence Rating
Nine of 13 RTI workshop participants (69%) identified exposure to consumers as important
to risk assessment. These nine participants were asked to rate the importance of ingestion,
inhalation, and dermal consumer exposures as well as their confidence that the existing data
on these risk relevance factors could support risk-management decisions. Based this
information, the inhalation route was the highest priority within this area, while ingestion and
dermal exposures were lower priorities. In the event of a tie (e.g., importance ratings for
dermal exposure routes), the most conservative rating was chosen.
See Section 6.3.3.2 for more information on this priority area and Section 1.1.3 for a detailed
explanation of the prioritization process.
DecaBDE Can Inform MWCNT Assessment
As with occupational exposures (see the DecaBDE Comparison Box in Section 4.2.1), studies of decaBDE provide insight into
possible routes of consumer exposure pathways to MWCNT flame-retardant coatings. During intended use, decaBDE present
in upholstered products in residential, commercial, and public settings can be disturbed and released as dust. Research
attributes most polybrominated diphenyl ether (PBDE) intake in toddlers, children, teenagers, and adults to household dust
from combined oral and dermal exposure (Johnson-Restrepo and Kannan, 2009). Unintended uses (e.g., repurposing of
treated upholstery textiles for clothing) or accidental releases (e.g., mouthing of textiles by children or animals, fire, or high
heat) are also possible consumer exposure scenarios; however, no information was identified that directly addresses these
potential exposures to decaBDE.
The extent of consumer exposure to decaBDE during intended or unintended use varies based on: differences in living space
size and time spent indoors or in rooms where exposure to decaBDE is more likely (Allen et al., 2008b), the presence of new
furniture (Rose et al., 2010), and the matrix or textile to which the flame-retardant chemical is applied. Similarly, weathering
processes [e.g., ultraviolet (UV) exposure, abrasion] differ across settings relevant to consumers, such as indoor/outdoor use
in homes and in airplanes or cars, and thus influence exposure differences between each type of setting. In planning research
to inform future MWCNT risk assessments, considerations might include: Do consumer exposure pathways differ between
intended and unintended consumer use? Are nonresidential settings (e.g., aircraft and automobiles) likely to contribute to
consumer exposures? Are the MWCNT textile matrices or surface treatments likely to lead to MWCNT in suspended particles
during intended use? What unintended uses of MWCNT flame-retardant products might pose a risk to consumers? See
Appendix H for more information on human consumer exposure routes for decaBDE.
4-23
-------
Additional Information Highlight Box 12:
12. Predicted dominant exposure routes for MWCNTs
Differences in production practices used to incorporate MWCNTs into fire retardants likely will result in different exposure
scenarios for consumers, by influencing the magnitude and form of the material to which consumers are exposed (Motzkus et
al., 2012) (see Section 1.2.2.2 and Additional Information Highlight Box 7). While little information is available on the
magnitude or form of nanomaterials released from product matrices (e.g., textiles), existing data from studies with other types
of nanomaterials suggests that release occurs as agglomerates of nanomaterials with other substances in the product matrix
(Nowack et al., 2012). Based on the physicochemical properties of MWCNTs (see Section 3.1) and information on BDE-209
(see Appendix H.3.1) MWCNTs and associated substances likely would be released from upholstery textiles in the particulate
phase. Yet, the primary exposure pathway(s) for consumers will vary depending on the types of upholstered products into
which the MWCNTs are incorporated. No studies were found in the literature that evaluated the potential for human exposure
to free or matrix-bound MWCNTs from any consumer product; however, because releases of MWCNTs in the particulate
phase are expected to be similar to BDE-209 releases, inferences can be drawn from PBDE data to provide indications of
primary exposure pathways for upholstery textiles in different scenarios [see the DecaBDE Comparison Box above, (in this
Section 4.2.2)1. Based on information from decaBDE and MWCNTs, predominant routes of exposure in consumer populations
are likely to be oral and dermal in residential and nonresidential spaces and in transportation vehicles (e.g., automobiles and
aircraft). During unintended uses, inhalation and dermal exposures are the likely primary exposure routes, with secondary oral
exposure as the secondary exposure route. See Section 4.2.2 for more information on each consumer exposure scenario for
MWCNTs in flame-retardant coatings applied to upholstery textiles.
No studies were found that evaluated the potential for consumer exposure to MWCNTs from any
consumer product. As a result, probable consumer exposure pathways and scenario characteristics for
exposure to free MWCNTs, bundled MWCNTs, and MWCNTs in combination with the polymer matrix,
textile fibers or scraps, or other product constituents cannot be differentiated at this time. As discussed in
Chapter 2. different MWCNT flame-retardant production processes are expected to result in differences in
release rates and release forms, which in turn will affect the magnitude of exposure during consumer use
and the form of the material to which consumers are exposed (Motzkus et al.. 2012). Based on the
physicochemical properties of MWCNTs, the assumption that MWCNTs and associated substances
released from consumer products will be present in the particulate phase is reasonable.
4.2.2.1. Intended Use- Upholstered Products in Residential Spaces
Although flame-retardant upholstery textiles typically are used in nonresidential settings (see
Section 4.2.2.2). some residential upholstered products, particularly mattresses, are known to contain
flame retardants, and other upholstered furniture products, like couches, sometimes might be treated with
flame retardants (Rose et al.. 2010). As introduced in Section 2.4. upholstered products are expected to be
used for many years, and contact with the textile might be frequent and prolonged, which could introduce
substantial wear and tear to the textile product. In addition, upholstery in residential spaces might
frequently be exposed to cleaning products, sweat, food, and other substances that could affect the
properties of the textile and the flame-retardant coating.
4-24
-------
The pathways through which consumers might be exposed in residential settings to MWCNTs
during general consumer use scenarios for end products upholstered with MWCNT flame-retardant
coatings are described below:
• Inhalation. Chronic inhalation of MWCNTs in combination with other product constituents
and dust could occur following release from upholstered products over time (due to wear and
tear from anticipated use, aging of materials, abrasion, UV light, water, cleaning chemicals,
among other factors; see Section 2.4.2). MWCNTs could settle onto surfaces, where they
might be disturbed and re-entrained, after which they could be inhaled by residents. Whether
inhalation is a primary route of consumer exposure for MWCNTs, particularly when
embedded in a polymer matrix, is unknown.
• Oral. MWCNTs in combination with other product constituents and dust could be ingested
after settling on food and food-contact surfaces or following hand-to-mouth activity. Whether
ingestion is a primary route of consumer exposure for MWCNTs, particularly when
embedded in a polymer matrix, is unknown. Preliminary, unpublished studies presented at a
public meeting indicate, however, that MWCNTs could be released from flame-retardant
barrier fabrics and polyurethane foams in very small amounts during normal wear and tear
(Uddin and Nyden. 201 Ib): these MWCNTs could settle onto food, food-contact surfaces, or
other surfaces where children could be exposed via hand-to-mouth activity during use.
• Dermal. Dermal exposure to MWCNTs in combination with other product constituents and
dust might occur while touching the textile surface (particularly if the portion of the textile
that has been treated with the flame-retardant coating is exposed) or touching surfaces upon
which particles have settled. Whether dermal uptake is a primary route of consumer exposure
for MWCNTs, particularly when embedded in a polymer matrix, is unknown. Preliminary,
unpublished studies indicate, however, that MWCNTs could be released from flame-retardant
barrier fabrics and polyurethane foams in very small amounts during normal wear and tear;
these MWCNTs could contact skin directly during use (Uddin and Nyden. 201 Ib).
4.2.2.2. Intended Use- Upholstered Products in Nonresidential Spaces
Due to regulations requiring that upholstery textiles used in nonresidential settings pass flame-
retardancy tests (see Table 1-3). many upholstery textiles in public, commercial, and institutional settings
are treated with flame retardants. The characteristics of the different settings in which these products are
used can vary considerably. For example, flame-retardant upholstery textiles might be used in seating for
airports and other transportation hubs and in waiting rooms, office buildings, penal institutions, and other
nonresidential spaces that can range from very small to very large and where consumers might spend
varying amounts of time. Some scenarios for nonresidential exposures are not likely to differ from those
expected from residential exposures, but a few key differences do exist. For example:
• Exposures to flame-retardant upholstery coatings in public spaces might be unavoidable.
Although consumers have some control over which products they bring into their home,
consumers have no control over the products they encounter in public spaces.
4-25
-------
• Some nonresidential exposures might occur over long periods of time and for extended
intervals (e.g., sitting in the same office chair every day over the course of several work
years), while some might occur infrequently and for short periods of time (e.g., sitting in
seating at the airport waiting for a flight).
• Products in public spaces might experience higher activity levels, more frequent cleaning,
and less care to the textile surface, all of which could damage or weaken the textile matrix
and influence releases and exposures.
With the exception of these potential differences in exposure settings and activity patterns, the
pathways and scenarios through which consumers might be exposed in nonresidential settings to
MWCNTs during general consumer use scenarios for end products upholstered with MWCNT flame-
retardant coatings are not expected to differ from those described previously in Section 4.2.2.1 on
exposures from intended use of upholstered products in residential spaces.
4.2.2.3. Intended Use -Aircraft and Automobile Upholstery
Flame-retardant upholstery can be used for seating, draperies, carpets, and other textiles in
passenger cars and public and private transportation.
The pathways through which consumers might be exposed in vehicles (including airplanes) to
MWCNTs during general consumer use scenarios for end products upholstered with MWCNT flame-
retardant coatings are described below:
• Inhalation. Inhalation of MWCNTs adsorbed to dust from worn or abraded automobile
upholstery is expected to occur. The recirculation of air in aircraft cabins also might affect
exposure to MWCNTs, if filters do not adequately remove these particles.
• Oral. Secondary oral exposures might occur if inhaled MWCNTs or MWCNTs that deposit
on the skin are subsequently ingested.
• Dermal. Dermal exposures to MWCNTs are expected to occur, particularly when skin
touches the treated part of the textile directly. Dermal exposure also can occur when particles
in the air settle on the skin. Different exposure characteristics or scenarios (e.g., children
sitting in safety seats) might influence whether dermal exposure occurs, or influence the
extent to which exposure occurs through this pathway.
4-26
-------
4.2.2.4. Unintended Use, Repurposing, or Reuse of Treated Textiles and Upholstered
Products
As introduced in Section 2.4. unintended uses of upholstery textiles treated with MWCNT flame-
retardant coatings could include repurposing of treated upholstery textiles for clothing, building
insulation, other in-home or outdoor furnishings, bedding, or other purposes. The repurposing stages
could introduce occupational exposures similar to those discussed in Section 4.2.1.2 (exposures during
formulation of the flame retardant, application to textiles, and upholstering) and Section 4.2.1.3 (exposure
during storage and distribution), as products that are treated with flame-retardant coatings are broken
down and reprocessed into new products.
Although no information was identified that directly addresses potential consumer exposures
following unintended use or reuse of flame-retardant upholstery textiles, exposure pathways and scenarios
from other life-cycle stages are relevant here. Most reuse scenarios might differ little from those for
anticipated consumer uses, but a few key differences might occur, particularly when products are
repurposed for new uses or used in unintended ways. For example:
• Similar processes to those involved with product manufacture (e.g., cutting, sewing) and
storage of textiles also might be employed for repurposing treated textiles. In this scenario,
however, these processes are not expected to occur in an occupational setting, but in the home
or another private space, where no personal protective equipment is worn and limited control
technologies are used. These processes, as employed for repurposing textiles, however, are
not expected to occur as commonly or at the same scale as in a manufacturing facility.
• Older, more degraded textiles with weakened matrices might be handled directly and
subjected to abrasion, thereby releasing the product constituents in the vicinity of the
consumer conducting the repurposing.
• Although dermal contact with products used for their intended purpose (e.g., furniture
seating) might be limited by a clothing barrier between the consumer and the treated textile,
should flame-retardant upholstery textiles be repurposed into clothing, direct dermal contact
might occur repeatedly over long periods of time.
With the exception of these potential differences in exposure characteristics, the pathways and
scenarios through which consumers might be exposed to MWCNTs during repurposing, reuse, or
unintended use of treated textiles and upholstered products are not expected to differ from exposure
pathways associated with the cutting, tailoring other abrasive processes involved with product
manufacturing (Section 4.2.1.2); storage of textile products (Section 4.2.1.3); and consumer use in
residential and nonresidential spaces (Sections 4.2.2.1 and 4.2.2.2).
4-27
-------
4.2.3. General Public Exposure Pathway Scenarios through
Environmental Media
Unprioritized Research Area: Exposure Route
Population
01
V V
I |
m r
o
Q.
3/13
6/13
4/13
Dermal
• - 1 participant's vote
Importance
L - Low: Least Important
M - Medium: Possibly Important
H — High: Important
Confidence
L - Low: Not Confident
M - Medium: Somewhat Confident
H - High: Confident
M IT~ H M L H M L
E-RRF Confidence Rating E-RRF Confidence Rating E-RRF Confidence Rating
Three of 13 RTI workshop participants (23%) identified exposure to the general human
population as important to risk assessment. Based on this information, human exposure in
the general population was determined to be of lesser importance to consider in a future
risk assessment of MWCNTs, and all text relevant to these areas was moved to Appendix
G. See Section 1.1.3 for a detailed explanation of the prioritization process.
4.2.4. Highly Exposed Populations
Neutral Research Area: Exposure Route
Populations with high exposure to MWCNTs were not considered during the RTI collective judgment prioritization process.
This section of text, however, is included in the main document because it supports understanding of occupational exposure
pathway scenarios (see Section 4.2.1) and consumer exposure pathway scenarios (see Section 4.2.2), which were deemed
priority research areas.
Occupation could increase exposure to MWCNT relative to the general population. In
occupational settings, the primary exposure pathway for MWCNTs is likely to be inhalation. Consumer
exposure pathways might be similar to those identified for decaBDE, namely ingestion of household dust,
but MWCNTs are less likely to be released from the polymer matrix (see Section 2.4.2; see Appendix H
for detailed information regarding decaBDE). Given the lack of data on consumer exposure to MWCNTs,
whether the primary route of exposure for highly exposed populations would be different from that of
decaBDE is difficult to determine. Dust levels in the home can vary by socioeconomic status or the type
and condition of housing (see Section 5.3.1). In turn, disproportionate levels of exposure can occur in
specific populations, including low-income and low-educational-attainment populations. Additionally, for
4-28
-------
pollutants for which inhalation exposure due to proximity to primary pollution sources is of concern,
socioeconomic status has been associated with increased exposures, which also might be true for
MWCNTs. Race and ethnicity do not present specific physiological conditions to increase susceptibility
to exposure, but demographic factors such as socioeconomic and educational status might cause some
populations to experience disproportionate exposures. The possibility of increased exposure to MWCNTs
due to characteristics associated with low socioeconomic status has not yet been explored in the literature.
In general, children are more susceptible to increased inhalation exposures because of increased
ventilation rates per unit of body weight and increased oral exposures due to hand-to-mouth and chewing
(e.g., mouthing furniture or fabric) behaviors. The relevance of the inhalation and oral pathways for
MWCNT consumer exposures, however, is unknown. In addition, lack of data on whether MWCNTs, if
released from flame-retardant textiles, would partition to dust precludes a determination of whether
children might experience elevated oral exposures to MWCNTs similar to those observed for decaBDE.
4.2.5. Exposure Reference Values and Recommendations
Neutral Research Area: Exposure Route
Exposure reference values for MWCNTs and recommendations from agencies or organizations were not considered during
the RTI collective judgment prioritization process. This section of text, however, is included in the main document because it
supports an understanding of occupational exposure pathway scenarios (see Section 4.2.1) and consumer exposure pathway
scenarios (see Section 4.2.2), which were deemed priority research areas.
A variety of exposure standards, guidelines, or recommendations are developed by different
organizations with purview over specific portions of the population or situations during which exposure
might occur (e.g., occupational exposures, general population drinking water exposures). Available
information on these types of values for MWCNTs is presented below. Section 5.1.1 discusses how some
of these values inform quantitative toxicity assessments.
MWCNTs can have features of both nanoparticles and fibers, and regulations exist to control
particles and fibers in the workplace. MWCNTs can appear as clumps or ropes, which can be counted as
single fibers if they fit the definition of a fiber. If the rope is not within the World Health Organization's
definition of a fiber (greater than 5 um in length, with an aspect ratio greater than 3:1), however, it would
not be counted as a fiber under the current measurement system. Some MWCNTs could therefore be
missed using current fiber classification methods (Donaldson et al., 2006).
Recently, NIOSH conducted a risk analysis for CNTs to establish a guideline exposure level for
occupational workers (NIOSH, 2010). They estimated a working lifetime inhalation exposure of
4-29
-------
0.2-2 ug/m3 [8-hour time-weighted average (TWA)] associated with a 10% excess risk of early-stage
adverse lung effects (95% lower confidence limit estimates) based on two subchronic animal inhalation
studies (Pauluhn. 201 Ob; Ma-Hock et al.. 2009) (see Section 5J_ for human health effects). The NIOSH-
recommended exposure limit is 7 ug/m3 for elemental carbon (see Table 4-1) as an 8-hour TWA
respirable mass airborne concentration (NIOSH. 2010). NIOSH also recommends that workplace airborne
exposure to CNTs be measured by NIOSH NMAM 5040, which has an upper limit of quantitation of
7 ug/m3 (NIOSH. 2010). Specifically, the animal-data-based risk estimates indicate that workers could
have >10% excess risk of developing early-stage pulmonary fibrosis if exposed over a full working
lifetime at the upper limit of quantitation for NIOSH NMAM 5040 (NIOSH. 2010). Other recommended
occupational exposure limits (OELs) and general human health exposure limits for inhalation of
MWCNTs and related materials are shown in Table 4-1.
As discussed by Schulte et al. (2010). deriving OELs for MWCNTs and other nanomaterials is
complicated by the challenges associated with measuring workplace exposures (see Text Box 4-1).
coupled with the variation in configurations of physicochemical properties that can influence exposure
and toxicity (see Text Box 4-2). The heterogeneity in MWCNT configurations could necessitate
developing OELs specific to individual formulations of MWCNTs (Schulte etal.. 2010). Alternatively,
OELs could be developed for groups of nanomaterials based on composition or toxic mechanism,
although such classifications could exclude physicochemical characteristics that influence biological
activity (Schulte et al.. 2010).
4-30
-------
Table 4-1. Established inhalation reference values and recommendations applicable to MWCNTs.
OELa
Value3
Citation3
The Occupational Safety & Health Administration - permissible exposure
limit (PEL) for respirable fraction of synthetic graphite
5,000 ug/m
3 Lam et al. (2006)
A review of CNT toxicity - human inhalation no-effect levels for INELaCute
workers derived from acute and subchronic inhalation studies
with MWCNTs
INELchronic
150 ug/m Aschberger et al.
(2010)
1
and
Bayer Pharmaceuticals - estimated OEL for TWA (6 hours/day, 5
days/week, 13 weeks) exposure to Baytubes® based on a no-observed
adverse-effect level of 0.1 mg/m3 divided by an inter-species dose-time
adjustment factor of 2
Pauluhnetal. (201 Oa)
50 ug/nT
NIOSH - recommended exposure limit for elemental carbon as an 8-hour
TWA respirable mass airborne concentration
7 ug/m
NIOSH (2010)
Nanocyl - estimated OEL for an 8-hour TWA exposure to MWCNTs based
on applying an overall assessment factor of 40 to the lowest-observed-
adverse-effect level of 0.1 mg/m3 in Ma-Hock et al. (2009)
2.5 ug/m
Nanocyl (2009)
Japanese New Energy and Industrial Technology Development
Organization - Interim OEL for SWCNTs and MWCNTs
Nakanishi et al. (2011)
30 ug/m3 (as cited in Morimoto
et al. (2013)
British Standards Institute - benchmark exposure limit based on one-tenth 0.1 fiber/cm BSI (2007) as cited in
of the Institute's asbestos exposure limit
air
NIOSH (2010)
Abbreviations: TWA = time-weighted average
"Reference values and recommendations listed in this table are current as of March 2013. Readers are encouraged to check the
sources, for future revisions of this information.
4-31
-------
4.2.6. Toxicokinetics, Dose, and Body Burden
Priority Research Area: Dose (Kinetics)
o
Q.
12/13
1/13
0/13
Absorption
•
• •
•
• *
* •
•
• •
•
E-RRF
Importance Rating
Distribution
o:
o: o
iii <3
o
H M L
E-RRF Confidence Rating
H M L
E-RRF Confidence Rating
Metabolism ]
•
•
•
• •
•
' 1
• •!
•
•
H M L
E-RRF Confidence Rating
• - 1 participant's vote
Importance
L -Low: Least Important
M-Medium: Possibly Important
H - High: Important
Confidence
L - Low:
M- Medium:
H - High:
W
Not Confident
Somewhat Confident
Confident
H M L
E-RRF Confidence Rating
for a detailed explanation of the prioritization process.
Twelve of 13 RTI workshop participants (92%) identified
the kinetics of MWCNTs in humans as important to risk
assessment. These 12 participants were asked to rate
the importance of ADME of MWCNTs as well as their
confidence that the existing data on these risk relevance
factors could support risk-management decisions.
Based on this information, absorption, metabolism and
excretion were deemed the highest priorities within this
area, followed by distribution. See Section 6.3.3.3 for
more information on this priority area and Section 1.1.3
DecaBDE Can Inform MWCNT Assessment
The toxicokinetics of MWCNTs is neither well understood (see Sections 4.2.6.1 and 4.2.6.2, and Additional Information
Highlight Box 14), nor expected to be similar to that of decaBDE, but the toxicokinetic questions that have been asked about
BDE-209 can help inform research planning for MWCNTs. When planning research to inform future risk assessments of
MWCNTs, investigators might consider the differences between acute and chronic studies or species, as well as other findings
with decaBDE, for instance: How are MWCNTs absorbed, distributed, metabolized, and excreted across various exposure
routes and durations? Does the type of solvent or dispersant used influence toxicokinetics of MWCNTs? Do these processes
differ across species? Is bioaccumulation over an extended period of time possible? Do MWCNTs distribute to fetuses or
neonates following in utero or early life exposures? See Appendix H for more information regarding the toxicokinetic behavior
of decaBDE and BDE-209.
Toxicokinetics can be used to relate exposure and contact, such as those described in the
scenarios above, with uptake and dose. Specifically, toxicokinetics describes how a material is absorbed,
distributed, metabolized, and excreted in an organism. An understanding of the relationship between each
of these concepts, which are often referred to as ADME, leads to an understanding of the concentration,
or dose, of material that can reach—and potentially accumulate in—different tissues of the body.
4-32
-------
4.2.6.1. Absorption, Distribution, Metabolism, Excretion
This section contains information regarding the toxicokinetic behavior of MWCNTs when
administered to mammals. Information regarding birds and fish is not presented in this section because,
when extrapolating toxicokinetic data to humans, studies conducted with rodents (rat or mouse) or
nonrodent mammals (dog or monkey) are generally used. Additionally, the toxicokinetic behavior in
response to MWCNTs might differ among birds, fish, and mammals. See Section 4.3 and Appendix G.4.2
for toxicokinetic information relevant to ecological exposures. Studies examining the toxicokinetics of
MWCNTs in mammals are summarized in Appendix F.
As discussed in Section 4.2.2. humans might be exposed to free MWCNTs, bundled MWCNTs,
and MWCNTs in combination with a polymer matrix, textile fibers or scraps, or other product
constituents. The bioavailability (and therefore dose) of MWCNTs is expected to differ for MWCNTs in
different forms or bundling states (see Text Box 4-2). In general, CNTs, including MWCNTs, appear to
be biopersistent. After intratracheal administration, MWCNTs have been observed to deposit and persist
within the lung for up to several months (Elgrabli et al.. 2008b: Deng et al.. 2007). Macrophage-mediated
clearance of MWCNTs after exposure via inhalation (Elgrabli et al., 2008a) and translocation of some
types of CNTs into the pleura and subpleura (Porter et al.. 2010; Ryman-Rasmussen et al., 2009a) have
been demonstrated. After oral exposure, most MWCNTs (administered at 10 ug/mouse by gavage) were
evident within the feces and also remained within the stomach and small and large intestines, with no
detectable transport into the blood or obvious metabolism through 28 days (Deng et al., 2007). Because
only one study was identified that evaluated distribution after oral exposure, whether distribution is
possible to other organs in the body following inhalation, dermal, and oral exposures to MWCNTs is not
well understood.
Distribution of CNTs to various organs has been reported following intravenous exposure (Deng
et al.. 2007; Cherukuri et al.. 2006). with predominant localization within the liver, lungs, and spleen.
This pathway, however, is not likely relevant for the exposures of concern in this evaluation (i.e.,
MWCNTs used in flame-retardant coatings on upholstery textiles are unlikely to be intravenously
applied).
4.2.6.2. Internal Dose and Body Burden
Based on toxicokinetic studies with rats, inhaled MWCNTs can remain in the lung following
exposure for an extended period, up to six months (Aschberger et al.. 2010). These studies reported
qualitative data, however, and no studies were found that reported levels of MWCNTs in the lung. One
study did report MWCNTs in the subpleura of mice following a single inhalation exposure to 30 mg/m3;
4-33
-------
no MWCNTs were detected in the subpleura following instillations of a lower concentration (1 mg/m3)
(Ryman-Rasmussen et al.. 2009a). Another study reported MWCNTs with known length just under 4 urn
in the pleura of mice following a single aspiration of 10-80 ug (Porter etal.. 2010). This finding is
notable because the pathogenic mechanism of asbestos fibers in the mesothelioma disease process occurs
in the pleural cavity (Aschberger et al.. 2010).
No detectable amounts of MWCNTs were observed in the blood following oral exposure in mice,
but the MWCNTs did remain in the stomach and small and large intestines (Deng et al.. 2007): the
observed MWCNTs remained unchanged (Aschberger et al.. 2010). As discussed in Section 4.2.6.1.
MWCNTs were distributed to multiple organs following intravenous injection (Deng et al.. 2007;
Cherukuri et al.. 2006). but this exposure pathway is unlikely to be relevant for this case study.
4.3. Ecological Exposure and Kinetics Leading to Dose
4.3.1. Factors Impacting Ecological Exposure
Neutral Research Area: Exposure Route
Factors impacting ecological exposure (e.g., properties of the environmental media and physiological and behavioral
characteristics of aquatic and terrestrial organisms) were not considered during the RTI collective judgment prioritization
process. Since this section of text supports understanding the unprioritized areas of exposure in aquatic and terrestrial biota,
this text is now included in Appendix G.4.2.1. See Appendix H.4.3.1 for a similar discussion relevant to decaBDE.
4.3.2. ADME in Ecological Receptors
Neutral Research Area: Dose (Kinetics)
General factors influencing ADME processes in aquatic and terrestrial biota were not considered during the RTI collective
judgment prioritization process. Because this section of text supports understanding the unprioritized areas of dose (kinetics)
in aquatic and terrestrial biota, this text is now included in Appendix G.4.2.2. See Appendix H.4.3.2 for a similar discussion
relevant to decaBDE.
4-34
-------
4.3.3. Exposure Pathways and Toxicokinetics in Aquatic Systems
4.3.3.1. Exposure Pathways in Aquatic systems
Unprioritized Research Area: Exposure Route
gx
el
o o
II1
o
Q.
E ,
4/13
5/13
4/13
Direct Contact
• - 1 participant's vote
Importance
L - Low: Least Important
M-Medium: Possibly Important
H — High: Important
Confidence
L -Low: Not Confident
M-Medium: Somewhat Confident
H - High: Confident
~R M IT" H M L H M L
E-RRF Confidence Rating E-RRF Confidence Rating E-RRF Confidence Rating
Four of 13 RTI workshop participants (31%) identified exposure to aquatic biota as
important to risk assessment. Based on this information MWCNT exposure to aquatic
biota was determined to be of lesser importance to consider in a future risk assessment of
MWCNTs, and all text relevant to these areas was moved to Appendix G.
See Section 1.1.3 for a detailed explanation of the prioritization process.
4.3.3.2. Toxicokinetics and Body Burden in Aquatic Systems
Unprioritized Research Area: Dose (Kinetics)
Aquatic
Biota
« n
u r
o
Q.
5/13
6/13
Metabolism
~H M IT~ H M L H M L
E-RRF Confidence Rating E-RRF Confidence Rating E-RRF Confidence Rating
2/13
Excretion
• — 1 participant's vote
Importance
L — Low: Least Important
M — Medium: Possibly Important
H — High: Important
Confidence
L - Low: Not Confident
M —Medium: Somewhat Confident
H - High: Confident
3)
C •
.
ff a
P
O
H M L
E-RRF Confidence Rating
Six of 13 RTI workshop participants (46%) identified
kinetics of MWCNTs in aquatic biota as important to
risk assessment. Based on this information MWCNT
kinetics to aquatic biota was determined to be of lesser
importance to consider in a future risk assessment of
MWCNTs, and all text relevant to these areas was
moved to Appendix G. See Section 1.1.3 for a detailed
explanation of the prioritization process.
4-35
-------
4.3.4. Exposure Pathways and Toxicokinetics in Terrestrial Systems
4.3.4.1. Exposure Pathways in Terrestrial Systems
Unprioritized Research Area: Exposure Route
Terrestrial
LU
o
0.
I.
1/13
7/13
5/13
Ingestion
Inhalation
Direct Contact
a
•E I
LU
D)
u.*
ce i>
or g
LU «j
o
a:
UJ
• - 1 participant's vote
Importance
L -Low: Least Important
M-Medium: Possibly Important
H - High: Important
Confidence
L -Low: Not Confident
M—Medium: Somewhat Confident
H - High: Confident
M L~~ H M L ML
E-RRF Confidence Rating E-RRF Confidence Rating E-RRF Confidence Rating
One of 13 RTI workshop participants (8%) identified exposure to terrestrial biota as
important to risk assessment. Based on this information MWCNT exposure to terrestrial
biota was determined to be of lesser importance to consider in a future risk assessment of
MWCNTs, and all text relevant to these areas was moved to Appendix G. See
Section 1.1.3 for a detailed explanation of the prioritization process.
4.3.4.2. Toxicokinetics and Body Burden in Terrestrial Systems
Unprioritized Research Area: Dose (Kinetics)
Terrestrial
Biota
u
o
Q.
E
2/13
6/13
5/13
Metabolism
LU
H M L
E-RRF Confidence Rating
• - 1 participant's vote
Importance
L - Low: Least Important
M-Medium: Possibly Important
H — High: Important
Confidence
L - Low: Not Confident
M - Medium: Somewhat Confident
H - High: Confident
H M L
E-RRF Confidence Rating
H M L ~~HM IT
E-RRF Confidence Rating E-RRF Confidence Rating
Two of 13 RTI workshop participants (15%) identified
kinetics of MWCNTs in terrestrial biota as important to
risk assessment. Based on this information MWCNT
kinetics in terrestrial biota was determined to be of
lesser importance to consider in a future risk
assessment of MWCNTs, and all text relevant to these
areas was moved to Appendix G. See Section 1.1.3 for
a detailed explanation of the prioritization process.
4-36
-------
4.4. Aggregate Exposures
Neutral Research Area: Exposure Route
Aggregate exposure to MWCNTs across multiple exposure routes was not considered during the RTI collective judgment
prioritization process. This section of text, however, is included in the main document because it supports an understanding of
occupational exposure pathway scenarios (see Section 4.2.1) and consumer exposure pathway scenarios (see Section 4.2.2),
which were deemed priority research areas.
Assessing aggregate exposures involves characterizing exposures to a single chemical across
multiple exposure routes. Due to the range of applications for which MWCNTs can be used, release from
multiple products and subsequent exposure via multiple routes is anticipated.
As described in Text Box 4-2. the properties of MWCNTs are easily altered through manipulation
of material characteristics such as size, bundling affinity, and surface treatments, and the degree to which
these specific changes affect the overall exposure profile is unclear. Moreover, the composition of
MWCNT formulations can vary with differences in synthesis techniques and remaining impurities from
manufacturing stock. If small changes in MWCNT characteristics result in measurable changes in the
nature and extent of exposure, each MWCNT formulation might be considered a unique substance;
therefore, MWCNTs produced by different manufacturers using different techniques might introduce
discrete sets of aggregate exposures. No consensus has been reached on which physicochemical
characteristics drive changes in exposure potential or what magnitude of change to any specific
characteristic or property is necessary to elicit a measurable change in exposure.
In addition to different material designs, MWCNTs can be used in a wide range of possible
applications, including in coatings, electronics, adhesives, polymer composites, thermoplastics, and
others. MWCNT applications can then be used in textiles, aerospace, construction, sporting goods,
medical applications, and many other types of products (Aschberger et al.. 2010). Thus, the potential for
exposure to MWCNTs exists where humans interact with any of these products as producers or
consumers or when CNTs are released to environmental media.
Exposure to MWCNTs is likely to occur through inhalation of MWCNT bundles and MWCNTs
sorbed to dust produced during the manufacture and processing of MWCNTs and composites containing
MWCNTs. A secondary pathway is through dermal exposure, which could occur in occupational settings
from dust settling on work surfaces. For consumers, exposure could occur from the abrasion or wear of
products containing MWCNTs. The general public could be exposed to MWCNTs via drinking water,
contact with contaminated soil, ingestion and inhalation of household dust, dermal contact with surfaces
upon which MWCNTs and dust have settled, and other pathways as a result of their release from product
matrices. MWCNT releases from composite materials, such as those used in sporting goods, plastics,
4-37
-------
touchscreens, and batteries are expected to be minimal, if not negligible, during consumer use because
MWCNTs used for these products are bound in relatively strong matrices. End-of-life product
dismantling, land-filling, and incineration, however, might offer greater potential for release of
constituent materials to environmental compartments because many of the processes involved in end-of-
life practices are intended to break down the strong matrices in which the MWCNTs are embedded
(Aschberger et al.. 2010). What the implications of these releases will be on exposures and impacts to
human health, ecological receptors, and other receptors is not yet known.
The anticipated market trend for production of MWCNTs is strong growth in the near future,
especially as production costs drop and a wider variety of applications is discovered (Lam et al.. 2006).
Strong market growth and diverse applications could lead to a greater diversity and number of exposure
scenarios, thus increasing the aggregate exposure potential for MWCNTs.
4.5. Cumulative Exposures
Neutral Research Area: Exposure Route
Cumulative exposure to MWCNTs and substances produced or released as a result of the MWCNT product life-cycle across
multiple exposure routes was not considered during the RTI collective judgment prioritization process. This section of text,
however, is included in the main document because it supports understanding of occupational exposure pathway scenarios (see
Section 4.2.1) and consumer exposure pathway scenarios (see Section 4.2.2), which were deemed priority research areas.
As stated in The Exposure Factors Handbook (U.S. EPA. 201 Ib). "Cumulative exposure is
defined as the exposure to multiple agents or stressors via multiple routes." For the purpose of this case
study, the "multiple agents or stressors" considered to contribute to cumulative exposure include those
substances that are produced or released as a result of the product life cycle of MWCNT flame-retardant
upholstery textile coatings, facilitate uptake of MWCNTs into humans and biota, are taken up as a result
of MWCNT exposures, or induce effects in humans or biota through a comparable or synergistic mode of
action. As mentioned in Section 4.4. different characteristics of different MWCNT formulations could
result in the necessity to consider different formulations as unique stressors, in which case each
formulation might represent a contribution to cumulative exposures.
Depending on which feedstocks are used in the manufacturing process, by-products might differ;
therefore, coexposures to MWCNTs and other compounds might differ. Although the generation of
impurities is likely during the manufacturing process, MWCNTs are typically purified after synthesis
with varying degrees of success (see Appendix C). As described in Section 2.2.2.2. Plata et al. (2009)
observed production of 45 side-products of CVD synthesis of MWCNTs, including polycyclic aromatic
4-38
-------
hydrocarbons, methane, and volatile organic carbons. SWCNT production can result in by-products of
sodium hydroxide, ethanol, water, filtrate, and scrap membrane (Healy et al.. 2008), but whether these by-
products also will be generated by MWCNT synthesis is unclear.
Functionalization, which involves covalent attachment of submolecular components to the
MWCNTs, is required before MWCNTs can be dispersed into polymers or organic solvents.
Functionalization can involve several different reagents, depending on the process used, as listed in Table
2-3. Any of the MWCNT-containing textiles could include small amounts of the reagents. No data were
found, however, on the by-products or impurities in textiles treated with MWCNT flame-retardant
coatings.
CNTs released to the environment might bind or sequester pollutants in a form that is not
bioavailable, thus reducing the impact of other toxic substances. For example, when MWCNTs are added
to sewage sludge, seed germination and root growth increased, which could be because the MWCNTs
bound pollutants (e.g., heavy metals, organic compounds) present in the sludge (Oleszczuk et al.. 2011).
On the other hand, MWCNTs might facilitate transport of these pollutants through environmental
compartments, across biological boundaries, and into cells, where they could react with cell machinery
(Johnston etal. 2010).
Increasing production and market growth for products containing CNTs likely will lead to
increasing levels of CNTs, by-products, and related compounds in the environment, as well as an increase
in exposures. Due to the heterogeneous nature of MWCNTs, the various manufacturing processes used,
exposure to a wide variety of CNTs and by-products from many different sources is possible. In addition,
the environmental persistence of CNTs could lead to long-term exposures or consecutive exposures in
multiple receptors.
4-39
-------
Chapter 5. Potential Human Health,
Ecological, and Other Impacts
The final step of compiling information into the comprehensive environmental assessment (CEA)
framework is to link potential impacts to receptors with the information described in the previous
chapters: the product life cycle; transport, transformation, and fate; and exposure-dose. The CEA
framework includes information relevant to impacts on human health and ecological receptors, similar to
what might be investigated in traditional risk assessment processes, as well as other plausible impacts that
might be considered in life-cycle-focused assessments (e.g., socioeconomics, climate change, resource
depletion).
Section 5.1 discusses potential impacts of exposure to multiwalled carbon nanotubes (MWCNTs)
and related contaminants on human health. This section relies heavily on evidence from experimental
studies with laboratory animals, the results of which could be extrapolated to humans using methods
established for quantitative toxicity assessment. As discussed in Chapter 4. humans could be exposed to
MWCNTs from flame-retardant upholstery textiles through a variety of pathways, with the contaminants
reaching receptors as a result of dermal deposition, oral ingestion, or inhalation. This section discusses
potential health impacts observed in studies with laboratory animals exposed to MWCNTs by these
exposure routes; data are grouped to illustrate the types of impacts (e.g., pulmonary toxicity, skin
irritation, reproductive effects) and sub-grouped by exposure routes for each impact.
Section 5.2 discusses the potential impacts of environmental media contaminated with MWCNTs
on ecological health, which encompasses impacts at the organism, population, and ecosystem levels. This
section is approached from an ecosystem perspective (aquatic vs. terrestrial), and data on groups of
organisms within those ecosystems are summarized. The focus of the discussion of impacts on ecological
health is on identifying and comparing data on exposure levels that might cause significant mortality,
delayed growth or development, reproductive defects, or other impacts that could alter community
structure and potentially cause ecosystem collapse.
Finally, Section 5.3 discusses other plausible impacts resulting from the product life cycles of
MWCNTs in flame-retardant upholstery textiles. The section includes a consideration of the energy input
requirements for synthesis of the MWCNTs, the economic impacts related to the cost of material
production, and the potential for disproportionate impacts on populations with lower socioeconomic
status.
5-1
-------
As noted throughout this document, MWCNTs are not a single material, but rather a mixture of
materials with different physicochemical properties. For the purpose of this case study, however,
MWCNTs are generally regarded as a single class of materials. Text Box 5-1 provides introductory-level
detail on how changes in physicochemical properties might influence toxicity. Throughout this chapter,
where physicochemical properties can be related to particular outcomes (e.g., fiber length on inhalation
endpoints), these properties are described and their potential influences on effects are discussed.
Additionally, Appendix F presents detailed toxicokinetic and toxicological study summaries in which the
MWCNT characteristics and components of test designs that influence toxicological outcomes are
provided.
5.1. Human Health Effects
Priority Research Area: Impacts
o
Q.
:*:
13/13
0/13
0/13
Cancer
O)
I1
w
• - 1 participant's vote
Importance
L - Low: Least Important
M-Medium: Possibly Important
H - High: Important
Confidence
L - Low: Not Confident
M - Medium: Somewhat Confident
H - High: Confident
01
e
o
9- -i
j= I
Hi to
O
Reproductive/
Developmental
•c
o
|J |
' M ' L ' ""H M L~~ ~H M IT
E-RRF Confidence Rating E-RRF Confidence Rating E-RRF Confidence Rating
All 13 RTI workshop participants identified human impacts of MWCNT exposure as
important to risk assessment. The 13 participants were asked to rate the importance of
cancer, noncancer, and reproductive and developmental effects as well as their
confidence that the existing data on these subjects could support risk management
decisions. Based on this information, noncancer effects were deemed the highest priority
for research in this area, followed by cancer, then reproductive and developmental effects.
See Section 6.3.4 for more information on this priority area and Section 1.1.3 for a
detailed explanation of the prioritization process.
5-2
-------
Text Box 5-1. Specific Physicochemical Properties of Multiwalled Carbon Nanotubes
Shown to Influence Toxicity
As introduced in Text Box 1-1, the physicochemical properties and behavior of multiwalled carbon nanotubes (MWCNTs) vary.
Many physicochemical characteristics of MWCNTs are interrelated, making it difficult to isolate a single characteristic and
determine how it influences toxicity. For example, MWCNTs that have undergone a grinding process have been observed to
induce more inflammation than their unground counterparts, but it is unclear whether this response results from reduced fiber
length or reduced bundling, which is itself related to a change in surface properties introduced by the grinding process (Muller
et al., 2005). In addition, dispersion state (and the characteristics that influence it, such as morphology and functionalization)
appears to be a driving factor behind granuloma formation. Specifically, more highly bundled CNTs induce large intraluminal
granulomas localized in the bronchi, while ground (and more dispersed) CNTs induce granulomas in the interstitial tissue of
the alveolar spaces and interstitium. Nevertheless, general assumptions about CNT toxicity are rarely made without also
presenting several exceptions. The following physicochemical characteristics contribute to changes in the toxicity of CNTs in
vitro and in vivo [as summarized by Johnston et al. (2010)1.
Morphology. Aspects of morphology, such as the diameter, length, and bundled state of CNTs might influence
toxicity. Following peritoneal injection, long, relatively straight CNTs induce asbestos-like effects (i.e.,
mesotheliomas), whereas bundles of shorter CNTs do not. Longer MWCNTs also have been associated with a
greater inflammatory response than shorter MWCNTs following dermal exposure. Cells might attempt to take up
CNTs that are too long and straight to be fully engulfed, which results in high reactive oxygen species and pro-inflammatory
cytokine release due to frustrated phagocytosis. In contrast, shorter bundled CNTs that are more easily engulfed by cells
produce little inflammatory response. Most toxicological studies use shorter CNTs (typically only a few microns in length), and
many studies do not characterize the length of the CNTs used due to the tendency of CNTs to bundle. Thus, the role of CNT
length in toxicity is not well understood. Diameter also might play a role in toxicity: Studies have shown that thin MWCNTs
(diameters of 50 nm or less) are more toxic than thick MWCNTs (diameters of 70 nm or more) (Fenoglioetal., 2012: Nagai et
al., 2011). According to Kim et al. (2011), although aspect ratio (the ratio of the CNT diameter and length) did not impact the
direct genotoxicity of MWCNTs, it might impact other aspects of toxicity (e.g., oxidative stressor inflammation) that could
indirectly induce genotoxicity. Similarly, others have suggested that aspect ratio plays a large role in toxicity (e.g., structural
similarities with asbestos fibers) (Stella, 2011: Johnston etal., 2010: Pacurarietal., 2010: Saeed, 2010).
Surface Functionalization. Alterations of surface chemistry can both enhance and reduce toxicity. Some surface
modifications, such as nitrogen-doping, result in less toxicity than pure CNTs because the modification makes the
CNTs more biocompatible. Other surface modifications, such as oxidation, result in greater toxicity than pristine
CNTs presumably because the modification promotes better dispersion of the CNTs and greater interaction with
cells. Some surface modifications can result in altered CNT shape or form. For example, prolonged oxidation treatment results
in shorter and straighter CNTs with different surface chemistry compared to nontreated CNTs. In a study by Jain et al. (2011),
MWCNT toxicity depended on functionalization density; a higher density of surface carboxyl groups resulted in lower toxicity to
male Swiss mice over a 4-week period.
Contaminants. Higher levels of metal impurities in CNTs are associated with an increase in toxicity. For example,
cobalt and nickel catalysts that remain in trace concentrations after CNT purification are likely responsible for
delayed hatching observed in zebrafish exposed as embryos to CNTs. Other potential contaminants remaining
after the manufacturing process include iron, amorphous carbon, hydrocarbons, and endotoxins, all of which can
induce unique toxic impacts. Although unpurified CNTs generally induce higher levels of toxicity (e.g., higher levels of
cytotoxicity, morphological changes in cells, release of pro-inflammatory cytokines, glutathione depletion), many purification
processes result in structural changes to the CNTs, making the attribution of increases in toxicity to contaminant content alone
difficult. As such, excluding certain unavoidable contaminants that are integral to the life cycle of an MWCNT from
consideration in toxicity studies might not be appropriate, even though many existing studies use pristine materials (Johnston
etal.,2010).
Wall Number. Several studies have illustrated that single-walled CNTs (SWCNTs) are potentially more toxic than
MWCNTs (Inoue et al., 2008: Tian et al., 2006: Jia et al., 2005: Radomski et al., 2005: Warheit et al., 2004):
however, other physicochemical properties likely varied between the SWCNTs and MWCNTs used in these
studies, making conclusions on the impact of number of walls difficult to decipher (Johnston etal., 2010).
5-3
-------
This section discusses the potential human health effects resulting from exposures to MWCNTs.
As noted in Chapter 4. exposure to this material from aggregate sources is likely; no studies were found
that investigated impacts to human health that could be attributed specifically to exposure to MWCNTs
released during the life cycles of MWCNT flame-retardant upholstery textile coatings. Primary literature
on MWCNTs was identified for most endpoints discussed in the following sections. When primary
literature on MWCNTs was not identified, literature on single-walled carbon nanotubes (SWCNTs) was
considered.
Toxicology studies conducted on animals provide much of the information discussed in this
chapter because data from studies on humans are unavailable. Effects observed in animal studies are
typically extrapolated to humans when conducting quantitative toxicity assessments [e.g., when
calculating a reference dose (RfD) or reference concentration (RfC); see Section 4.2.51. Potential health
effects associated with all routes of exposure (dermal, inhalation, and oral) are presented in this section
because each is plausible for humans (see Chapter 4 for additional exposure scenario information). Table
5-1 provides an overview of the findings for human health effects of decaBDE and MWCNTs and is
followed by a detailed discussion of the available data for MWCNTs. Detailed discussion of the data for
decaBDE can be found in Appendix H.
Table 5-1. Summary of effects observed after dermal, oral, and inhalation exposure to
decaBDE and MWCNTs.
Route of
exposure
Observed effect
MWCNTs DecaBDE
Yes No No data Yes No No data
Dermal
Local effects
Other effects
Skin irritation
Skin sensitization
Ocular irritation
Acute/subacute toxicity
Subchronic toxicity
Chronic toxicity
X X
(weak3)
Xa X
X X
(weak3)
Xb X
X X
X X
5-4
-------
Table 5-1 (Continued): Summary of effects observed after dermal, oral, and inhalation exposure to
decaBDE and MWCNTs.
Route of
exposure
MWCNTs
DecaBDE
Observed effect
Yes
No No data
Yes
No
No data
Inhalation0
Local effects
Inflammation
X
Respiratory sensitization
Xd
X
Other effects Acute/subacute toxicity
Xe
X
Subchronic toxicity
Xe
X
Chronic toxicity
X
X
Immunotoxicity
Xr
X
Oral
Effects
Acute/subacute toxicity
X
Subchronic toxicity
X
Chronic toxicity
X
Reproductive/developmental
Xj
"Details can be found in Appendix F.1.2 (Table F-4) (Ema et al.. 2011).
bSee Section 6.3.4.1 for information on a study identified after the collective judgment prioritization process (ECHA. 2013). Note that
the data from ECHA (2013) are within the REACH dossier, and thus have not been peer reviewed, nor have they been reviewed or
verified by ECHA or any other authoritative body
°Only studies utilizing the inhalation route of exposure are included in this table. Studies utilizing other pulmonary administration
methods (e.g., intratracheal instillation) are described in Section 5.1.3 and in Appendix F.1.2 (Table F-6).
dlnflammation was found in three OECD-compliant studies (LOAEC = 0.1 mg/m3) (Pauluhn. 201 Ob: Ellinger-Ziegelbauer and
Pauluhn. 2009: Ma-Hock et al.. 2009): respiratory sensitization was found by Ryman-Rasmussen et al. (2009b) (single 6-hour
exposure to ~ 104 mg/m3). Details can be found in Appendix F.1.2 (Table F-6).
eDetails can be found in Appendix F.1.2 (Table F-6) (Pauluhn. 201 Ob: Ellinger-Ziegelbauer and Pauluhn. 2009: Li et al.. 2009: Ma-
Hock etal.. 2009).
'Systemic immunosuppression was reported by Mitchell et al. (2009: 2007) (LOAEC = 0.3 mg/m3). Details can be found in
Appendix F.1.2 (Table F-6).
Decreased thymus weight was reported by Lim et al. (2011 b) (LOAEL = 1,000 mg/kg-day for 14 days). Details can be found in
Appendix F.1.2 (Table F-8). See Section 6.3.4.1 for information on two studies identified after the collective judgment prioritization
process (ECHA. 2013). Note that the data from ECHA (2013) are within the REACH dossier, and thus have not been peer reviewed,
nor have they been reviewed or verified by ECHA or any other authoritative body.
hEffects included changes to thyroid and liver observed in subchronic and chronic oral studies (NTP. 1986: Norriset al.. 1975: Norris
etal.. 1973)
'Details can be found in Appendix F.1.2 (Table F-8) (Lim etal.. 2011b).
'Effects reported in response to neonatal exposure include changes in sperm parameters (Tseng etal.. 2006) and changes in
locomotor activity or altered expression of proteins in the central nervous system (Johansson et al.. 2008: Viberg et al.. 2008: Viberg
et al.. 2007: Viberg et al.. 2003).
5-5
-------
Additional Information Highlight Box 13:
13. MWCNT fibers resemble asbestos fibers
Similarities between the shapes of MWCNTs and asbestos fibers have raised concern regarding potential health impacts
related to inhalation of MWCNTs (Murphy etal., 2011). Asbestos inhalation can cause mesothelioma, which is directly linked
to the aspect ratio (length/diameter) of the fiber (Donaldson etal., 2010). Asbestos fibers in the parietal pleural portion of the
lung cause inflammation and fibrosis that are believed to lead to mesothelioma (Schinwald et al., 2012: Donaldson et al.,
2010). Some early-stage effects critical for the development of mesothelioma (e.g., inflammation and early fibrosis) have been
demonstrated in studies using MWCNTs (Morimoto etal., 2012: Murphy etal., 2012: Murphy etal., 2011). Although the lung
often can clear foreign fibers, the aspect ratio of some fibers (including some MWCNTs) might limit the effectiveness of
biological clearance mechanisms (Donaldson et al., 2010). Importantly, a study by Murphy et al. (2011) investigated
clearance of CNTs of varying lengths and demonstrated a deficit in clearance ability within the parietal pleura for long, but not
short, fibers. Furthermore, Poland et al. (2008) reported a mesothelial inflammatory response following intraperitoneal
injection of long MWCNTs (>15 urn), but not short MWCNTs. Muller et al. (2009) also did not report an increase in
mesothelioma in rats injected intraperitoneally with 20 mg of MWCNTs with an average length <1 urn. Schinwald et al. (2012)
investigated this mechanism further and reported a clear threshold where MWCNTs longer than 4 urn were pathogenic to the
pleura. Nagai et al. (2011) demonstrated that, in addition to length, diameter and rigidity are important factors in the potential
carcinogenicity of MWCNTs. Thin, rigid MWCNTs (diameter ~ 50 nm) pierced mesothelial cell membranes and produced
mesothelioma in rats injected intraperitoneally (10 mg total dose), while thick (diameter ~ 150 nm) or flexible MWCNTs
(diameter ~ 2-20 nm) did not. The results of mode-of-action investigations for MWCNTs would provide qualitative data to
better determine the appropriateness of the comparison to asbestos; however no quantitative data were identified. See
Sections 5.1.11 and 6.3.4.1 for additional discussion on potential carcinogenic effects of MWCNTs.
Most toxicological studies for MWCNTs exposed animals by the dermal or inhalation route (see
Section 4.2.2 and Additional Information Highlight Box 12 for discussion of why dermal and inhalation
might be primary exposure pathways for MWCNTs in occupational settings). Effects are predominantly
localized and include irritation (both skin and ocular), sensitization (respiratory), and inflammation
(respiratory). When determining the toxicity of MWCNTs for humans, the following factors should be
considered:
• Several studies have shown that MWCNTs might induce prominent pulmonary inflammation
(Tauluhn. 201 Ob: Ellinger-Ziegelbauer and Pauluhn. 2009; Ma-Hock et al.. 2009) (see
Section 5.1.3).
• At least some CNTs were found to contain a large proportion of metal catalyst (iron and
nickel), which contributes significantly to oxidative stress, as indicated by the formation of
free radicals and accumulation of peroxidative products, depletion of total antioxidant
reserve, and loss of cell viability (Shvedova et al., 2003). Transition metals such as iron can
be important in the toxicity of a variety of pathogenic dusts because of their ability to cause
oxidative stress (Pulskamp et al., 2007; Ohio et al., 1999; Donaldson et al.. 1996; Kennedy et
al.. 1989).
5-6
-------
Although most MWCNT inhalation studies have not reported systemic toxicity (i.e., toxic effects
resulting from absorption and distribution of a toxicant to a site distant from its entry point) (see Section
5.1.2.2). subacute inhalation exposure to MWCNTs altered immunological function (Mitchell et al.. 2009;
Mitchell et al.. 2007) (see Section 5.1.8). The carcinogenicity of MWCNTs following inhalation exposure
also has not been investigated. Several studies using instillation and intraperitoneal injection (Sakamoto et
al.. 2009; Poland et al.. 2008; Takagi et al.. 2008). however, have demonstrated that certain forms of
MWCNTs could behave in a manner similar to asbestos and induce mesotheliomas. Therefore, inhalation
of certain forms of MWCNTs could be a human health concern.
5.1.1. Quantitative Toxicity Assessment
Neutral Research Area: Impacts
Quantitative toxicity assessment and determination of health reference values were not considered during the RTI collective
judgment prioritization process. This section of text is included in the main document because it supports the connection
between specific research questions related to human health impacts (a priority research area) and subsequent human health
risk assessments. To develop human health risk assessments, data would be needed to support reference value derivation,
examples of which are described below.
In a quantitative toxicity assessment, toxicity data are used to derive reference values, such as an
RfD for oral exposure or RfC for inhalation exposure. Similar to an RfD (as defined in Section 4.2.5). an
RfC is an estimate of a continuous inhalation exposure for a given duration to the human population
(including susceptible subgroups) that is likely to be without an appreciable risk of adverse health effects
over a lifetime. Both values, an RfC and an RfD, are derived from a benchmark dose lower confidence
limit, no-observed-adverse-effect level (NOAEL), a lowest-observed-adverse-effect level (LOAEL), or
another suitable point of departure, with uncertainty/variability factors applied to reflect limitations of the
data used. Other types of reference values also can be derived for use with other exposure durations (e.g.,
acute or subchronic), more specific populations (e.g., healthy workers), or specific exposure contexts
(e.g., emergency response or occupational exposure; see Section 4.2.5).
EPA has not evaluated MWCNTs to derive an RfD or an RfC. The only health effect-based
reference value for MWCNTs derived by a government agency is the draft recommended exposure limit
(REL) proposed by the National Institute for Occupational Safety and Health (NIOSH) for CNTs
(NIOSH. 2010). as discussed in Section 4.2.5. Acute and subchronic human no-effect levels for inhalation
exposures to the general public also have been proposed in the open literature; derivation of these values
is discussed briefly in the sections that follow.
5-7
-------
5.1.1.1. Health Reference Values
As mentioned above and discussed in Section 4.2.5. NIOSH (2010) conducted a risk analysis for
CNTs and established a draft REL of 7 ug/m3 (the high estimate of the limit of quantification for NIOSH
Method 5040) for carbon. The REL is based on a working lifetime inhalation exposure of 0.2-2 ug/m3
(8-hour time-weighted average) associated with a 10% excess risk of early-stage adverse lung effects
(95% lower confidence limit estimates) and was derived based on two subchronic (90-day) animal
inhalation studies (Pauluhn. 201 Ob: Ma-Hock et al. 2009).
A chronic human no-effect level for MWCNT exposure to the general public of 0.25 ug/m3 has
been estimated by Aschberger et al. (2010) based on a LOAEL (Ma-Hock et al.. 2009) and aNOAEL
(Pauluhn. 201 Ob) of 1.0 ug/m3 from the same subchronic studies NIOSH (2010) used in their derivation
of an REL. Aschberger et al. (2010) also calculated a human no-effect level for acute occupational
exposure to MWCNTs of 150 ug/m3 from a LOAEL of 11 mg/m3 based on the absence of inflammatory
effects in a rat study (Ellinger-Ziegelbauer and Pauluhn. 2009). Both values were calculated based on
guidance provided by Registration, Evaluation, Authorisation and Restriction of Chemicals (REACH) for
chemical safety assessment (ECHA. 2008). Additional details regarding the derivation of these values are
available in Aschberger et al. (2010). Overall, the relatively limited and often conflicting database of
currently available toxicological values for MWCNTs suggests that these values are preliminary estimates
associated with a high degree of uncertainty.
The two subchronic animal inhalation studies (Pauluhn. 201 Ob; Ma-Hock et al.. 2009) used to
derive the draft NIOSH (2010) REL and the subchronic human no-effect level proposed by Aschberger et
al. (2010) are described in detail in Section 5.1.3. Table 4-1 lists all identified inhalation reference values
and recommendations applicable to MWCNTs.
5.1.2. Systemic Toxicity
5-8
-------
DecaBDE Can Inform MWCNT Assessment
For all routes of exposure, decaBDE exhibits low acute toxicity. Subchronic occupational exposure to decaBDE in humans,
however, has provided evidence of toxicity. In addition, evidence suggests hepatic and thyroid effects in rats after subchronic
oral exposure. Further, chronic exposure in test animals resulted in multiple noncancer thyroid and hepatic effects. Yet,
regardless of exposure duration, minimal pulmonary effects and no ocular or dermal irritation were observed in exposed test
animals. Research planning to inform future risk assessments of MWCNTs might consider how impacts or effects vary across
multiple exposure routes and durations of exposure. See Appendix H for more detailed information on the noncancer effects of
decaBDE.
5.1.2.1. Acute or Subacute
No systemic effects were reported in two acute nose-only inhalation studies in rodents. One study
in rats used a single 6-hour exposure to a concentration of 11,241 mg/m3 MWCNTs and included a 90-
day post-exposure observation period (Ellinger-Ziegelbauer and Pauluhn. 2009). The other study, in mice,
used a single 6-hour exposure to a concentration of-104 mg/m3 and included a 14-day post-exposure
observation period (Ryman-Rasmussen et al.. 2009b).
No systemic effects were reported by two subacute inhalation studies in rodents. In one, mice
were exposed to an MWCNT concentration of 32.61 mg/m3 for 6 hours/day for up to 15 days (Li et al..
2007). In the other study, rats were exposed to concentrations up to 0.94 mg/m3 for 6 hours/day for 5 days
with a 1-month post-exposure observation period (Kim et al.. 2012a). In contrast to both these studies,
systemic immunosuppression was observed in mice exposed to > 0.3 mg/m3 MWCNTs via whole-body
inhalation for 6 hours/day for up to 14 days (Mitchell et al., 2009; Mitchell et al., 2007) (see Section 5.1.8
for further discussion of immune system effects). Acute and subacute inhalation studies identified for
MWCNTs, including key studies, are presented in Appendix F.1.2.
Limited data were identified on the acute or subacute toxicity of MWCNTs following oral or
dermal exposure (see Section 5.1.5 for discussion of skin irritation). Based on the Organisation for
Economic Co-operation and Development (OECD) TC 423, MWCNTs tested by Pauluhn (201 Ob) were
not acutely toxic (oral LD50 > 5,000 mg/kg body weight). One acute oral toxicity study on SWCNTs was
identified in which single doses of 1,000 mg/kg body weight of three different types of SWCNTs
(raw: 1 nm x 1-2 um, 25% Fe; purified: 1 nm x 1-2 urn, <4% Fe; ultrashort: 1 nm x 20-80 nm,
<1.5% Fe) were administered to mice (Kolosnjaj-Tabi et al.. 2010). No signs of toxicity (e.g., reduced
survival, delayed growth, behavioral abnormalities, clinical chemistry changes) were observed for any
type of SWCNT. Absolute and relative thymus weights, however, were decreased in female rats exposed
by gavage to 1,000 mg/kg-day MWCNTs for 14 days (Lim etal.. 20 lib). Neither an acute nor subacute
dermal exposure study was identified for MWCNTs.
5-9
-------
A single intraperitoneal injection of > 4 mg/kg MWCNTs decreased body weight in female mice
9 days after exposure (Fujitani et al., 2012). Furthermore, mice intraperitoneally injected with 0.25, 0.5,
or 0.75 mg/kg-day of purified, carboxylated MWCNTs for 5 days experienced decreased body weight
gain and increased markers of oxidative stress and hepatotoxicity (increased reactive oxygen species in
liver, enhanced activity of liver enzymes such as serum aminotransferases and alkaline phosphatases)
(Patlollaetal.. 2011) (see Table F-8 in Appendix F).
5.1.2.2. Subchronic
No data were identified on the subchronic systemic toxicity of MWCNTs following oral or
dermal exposure. No systemic effects were reported in two 13-week inhalation studies in rats. One study
used head-nose exposure at concentrations up to 2.5 mg/m3 for 6 hours/day and 5 days/week (Ma-Hock et
al.. 2009). The other study used nose-only exposure at concentrations up to 6 mg/m3 for 6 hours/day and 5
days/week, including a 6-month post-exposure observation period (Pauluhn. 201 Ob). Furthermore, no
systemic effects were reported by a 60-day inhalation study in mice exposed at a concentration of
32.61 mg/m3 for 6 hours/day (Li et al.. 2009). See Section 5.1.3 for discussion of pulmonary effects
following subchronic exposure.
5.1.2.3. Chronic
No data were identified on the chronic systemic toxicity of MWCNTs following exposure by any
route.
5.1.3. Pulmonary Toxicity
Material characteristics and study details associated with the acute and subchronic inhalation
studies identified for MWCNTs are presented in Appendix F.I.2, Table F-6. No chronic inhalation studies
were identified for MWCNTs. Toxicological responses generally have been consistent across studies
administering MWCNTs into the lungs (by intratracheal instillation, aspiration, or inhalation), with
exposed animals exhibiting pulmonary inflammation and fibrosis. Although intratracheal instillation
studies for MWCNTs report effects similar to those reported for inhalation studies (e.g., pulmonary
inflammation and fibrosis), the comparability of intratracheal instillation to the inhalation exposure route
is questionable (Morimoto et al.. 2012; Oberdorster. 2010). Animal exposures to MWCNTs via
intratracheal instillation (Park et al.. 2009; Muller et al.. 2008a: Mulleretal.. 2008b: Muller et al.. 2005)
generally have resulted in more severe effects than those observed for inhalation (Pauluhn. 201 Ob:
5-10
-------
Additional Information Highlight Box 14:
14. Applying traditional in vivo inhalation study design models to MWCNTs
Available evidence suggests that MWCNT inhalation exposure might cause human health effects (e.g., fibrosis, inflammation)
(see Sections 4.2.1, 4.2.2 and 5.1.3): however, the lack of chronic data, choice of exposure route, and differences in study
design could impede the use of existing data in future risk assessments and the subsequent risk management of potential
MWCNT human health effects (see Section 5.1).
Traditional particle exposure studies might not be relevant to human exposures because of known species differences and
exposure methods that have been shown to alter health effects (Nikula et al., 2001: Osier and Oberdorster, 1997: Paxton,
1995). Several studies have attempted to improve on approaches to allow for better translation to human health effects by
evaluating variables in experimental design that influence toxicity outcomes and by developing alternative methods to test
MWCNT toxicity. For example, Wako et al. (2010) demonstrated in a rat intratracheal instillation model that the methods used
to prepare the MWCNT suspensions necessary for intratracheal instillation can alter cellular responses in the lung and
suggested a primary method for similar, future MWCNT studies. Other suggested improvements to MWCNT toxicity testing
design have included alternative exposure methods in laboratory animals (as described above), cellular exposure models, and
specific methods for MWCNT sample preparation (Liu etal., 2012: Coccini etal., 2010: Alfaro-Moreno et al., 2008: Geysetal.,
2007: Osier and Oberdorster, 1997). For example, optimization of an in vitro cell culture model of respiratory epithelium has
improved the growth of a tight cell monolayer that could be used as a respiratory translocation model for nanomaterials (Geys
et al., 2007). Coccini et al. (2010) tested the effect of varying degrees of functionalization on the MWCNTs, which were
observed to alter the water solubility, dispersibility, and agglomeration, within two-cell culture models of cytotoxicity.
Ellinger-Ziegelbauer and Pauluhn. 2009; Li et al.. 2009; Li et al.. 2007; Mitchell et al.. 2007). During
instillation, MWCNTs are administered as a bolus dose, which could overpower clearance mechanisms
that might be adequate if exposure were to occur more slowly and gradually, as during inhalation. Also,
instilled MWCNTs administered as a bolus might bundle differently compared to inhaled MWCNTs,
which can produce artifactual granulomatous lesions (Muller et al.. 2005).
Pulmonary inflammation was observed in one acute (Ellinger-Ziegelbauer and Pauluhn. 2009)
and two subchronic (Pauluhn, 201 Ob; Ma-Hock et al.. 2009) OECD-compliant animal inhalation studies,
with granulomas observed in both subchronic studies. In the Ma-Hock et al. (2009) study, an exposure-
related increase in the incidence of granulomatous inflammation in the lung and lung-associated lymph
nodes was observed in exposed rats (head-nose exposure). At the mid and high exposures (0.5 and
2.5 mg/m3, respectively), increased lung weights, pronounced multifocal granulomatous inflammation,
diffuse histiocytic and neutrophilic inflammation, and intra-alveolar lipoproteinosis were observed in lung
and lung-associated lymph nodes. Similar effects were observed in a study by Pauluhn (201 Ob), in which
Wistar rats were exposed (nose only) to MWCNTs (0.1, 0.4, 1.5, 6 mg/m3); inflammatory changes in the
distal nasal cavities were observed at all but the lowest concentration. Additionally, exposure-related
lesions of the upper respiratory tract (e.g., goblet cell hyperplasia or metaplasia, eosinophilic globules,
and focal turbinate remodeling) and lower respiratory tract (e.g., inflammatory changes in the
bronchioloalveolar region and increased interstitial collagen staining) were observed at higher
concentrations. According to a study conducted by Kim et al. (20L2a), pulmonary DNA damage is
5-11
-------
initiated after only a few days of exposure; a significant increase in DNA damage was measured by
Comet assay in lung cells from rats immediately following 5 days of exposure (6 hours per day) to
0.94 mg/m3 MWCNTs. DNA damage remained detectable 1 month after the last exposure.
5.1.4. Eye Irritation
All in vivo eye irritation studies considered are presented in Appendix F.I.2 (Table F-4). In a key
ocular irritation study conducted with rabbits, one of two types of MWCNTs administered via instillation
to the conjunctival sac resulted in conjunctival redness and blood vessel hyperemia at 1 hour, but not at
24 hours (Emaetal.. 2011). Differences in purity, diameter, and surface area (see Appendix F. Table F-4)
between the two types of MWCNTs tested might explain why effects were observed with only one of the
two types.
5.1.5. Skin Irritation
Material characteristics and study details associated with the in vivo dermal studies considered
for MWCNTs are presented in Appendix F.1.2.
Based on the information available, even the most irritating of several MWCNTs tested appears
to be only a very weak skin irritant when tested on healthy, intact skin (Emaetal.. 2011; Kishore et al..
2009). The available studies were performed in accordance with accepted standard practices for
conventional skin irritation tests, which use healthy, intact skin exposed for 4 hours under semioccluded
conditions. No data were identified for dermabraded or damaged skin.
5.1.6. Reproductive Effects
DecaBDE Can Inform MWCNT Assessment
Most studies report that decaBDE does not induce reproductive or developmental effects when administered at high levels
during gestation or adulthood. Some research does show reproductive and neurodevelopmental effects, however, when
neonates are dosed directly. Even though mechanistic differences in potential reproductive or developmental effects of
decaBDE and MWCNTs are likely, data on decaBDE can inform research planning for future MWCNT risk assessments.
Specifically, decaBDE data provide an example of how exposure at different stages of development can influence health
outcomes, which can help guide research planning for materials such as MWCNTs. Based on decaBDE data, questions that
researchers might consider for MWCNTs include the following: Do reproductive and developmental effects of MWCNTs differ
when administered during different developmental stages? Does direct dosing of offspring with MWCNTs influence health
effects compared to maternal dosing? See Appendix H for more details regarding the reproductive and developmental effects
of decaBDE.
5-12
-------
No in vivo reproductive studies using highly relevant exposure routes were initially identified for
MWCNTs. See Section 6.3.4.2 for information on two studies (Lim et al., 201 la; Lim et al., 20lib; Bai et
al., 2010) that were identified after the collective judgment prioritization process.
5.1.7. Developmental Effects
One developmental study was identified in which MWCNTs were administered via gavage to
pregnant Sprague-Dawley rat dams at doses of 0, 40, 200, and 1,000 mg/kg-day on gestation days (GD)
GD6-GD19; dams were sacrificed on GD20 (Lim et al.. 20lib). Minimal maternal toxicity was observed
at 1,000 mg/kg-day; a dose-dependent decrease in thymus weight was observed, but no effects were
observed on maternal body weight, food consumption, and oxidant-antioxidant balance in the liver.
No differences in gestation index, fetal death, fetal and placental weights, or sex ratio were observed as a
result of MWCNT exposure. Therefore, the embryo-fetal NOAEL was 1,000 mg/kg-day.
When MWCNTs were administered via intratracheal spray to pregnant ICR mice in a single dose
of 0, 3, 4, or 5 mg/kg on GD9, teratogenic effects were observed in the two highest dose groups (Fujitani
et al.. 2012). Fetal examinations performed on GDIS showed external and skeletal malformations such as
short or absent tails, limb reduction deformities, fused ribs and vertebral bodies, and hypophalangia (see
Table F-8 in Appendix F). In the same study, intraperitoneal injection of a single dose of 0, 2, 3, 4, or 5
mg/kg MWCNTs to pregnant mice on GD9 resulted in similar malformations at all dose levels.
5.1.8. Immune System Effects
Inhalation studies examining the immune system after exposure to MWCNTs are summarized in
Appendix F.1.2. In one notable study, systemic immunosuppression (characterized by T-cell-dependent
antibody response to sheep erythrocytes and T-cell proliferative ability in presence of mitogen), was
observed in mice exposed to 0.3, 1, or 5 mg/m3 MWCNTs via inhalation for 14 days; no significant lung
inflammation or lung tissue damage, however, was observed (Mitchell et al.. 2007).22 The
immunosuppressive mechanism could involve a signal originating in the lungs that activates
cyclooxygenase enzymes in the spleen (Mitchell et al., 2009).
Although an OECD-compliant skin sensitization study did not identify MWCNTs as dermal
sensitizers (Emaetal.. 2011). they have been identified as respiratory sensitizers in several studies. Park
This study was questioned after publication primarily because an image in the study suggests that nanofibers, not
nanotubes, were used.
5-13
-------
et al. (2009) observed a potential allergic response in mice following intratracheal instillation of 50 mg/kg
MWCNTs. The authors noted that the significantly increased immunoglobulin E concentrations coupled
with pro-inflammatory responses likely resulted from B-cell activation by IL-10. Similarly, Inoue et al.
(2009) demonstrated a potential allergic response in mice following six weekly intratracheal instillations
of 50 ug MWCNTs. These authors reported a significant increase in the number of total immune cells
(including macrophages, neutrophils, eosinophils, and lymphocytes) in the bronchiolar lavage fluid,
concurrent with infiltration of eosinophils, neutrophils, and mononuclear cells in the lung. They also
noted a significant induction of goblet cell hyperplasia in the bronchial epithelial tissue, indicating that
MWCNTs exacerbate metaplasia in the presence or absence of other allergens. In addition, MWCNTs
were shown to promote allergic immune responses to ovalbumin (OVA) and to elicit an acute
inflammatory response in the lung in the absence of OVA in female BALB/cAnCrl mice following
intranasal exposure at < 400 ug/mouse (Nvgaard et al.. 2009). Of note, the OECD currently has no
guidelines for standardized respiratory sensitization studies, and the studies identified above exposed
animals by intratracheal instillation or intranasal exposure, not inhalation.
One study that used inhalation as the route of exposure was identified. Ryman-Rasmussen et al.
(2009b) exposed normal and OVA-sensitized mice to 100 mg/m3 MWCNT aerosol for 6 hours
(approximately 10 mg/kg MWCNT total) and found that airway fibrosis occurred differentially in OVA-
sensitized mice versus nonsensitized mice. The authors concluded that airway fibrosis occurs as a result
of MWCNT inhalation with preexisting inflammation, suggesting that individuals with preexisting
allergic inflammation are susceptible to the pulmonary effects of MWCNTs.
Immune effects from oral exposure to MWCNTs have been suggested by one study. As described
in Section 5.1.7. Lim (20lib) observed a decrease in thymus weight in rat dams following administration
of MWCNTs at the highest dose tested (1,000 mg/kg-day).
5.1.9. In Vitro Data
Some in vitro methods are routinely used in regulatory risk assessment (e.g., Ames test, dermal
absorption assays, assays of skin and eye irritation). In response to increasing pressure to move away
from toxicological testing in animals, alternative methods are under development (Adler et al.. 2011). At
the present time, however, data from vitro methods are more commonly used for making qualitative
judgments on the toxic potential of stressors and elucidating mechanisms of toxicity.
5-14
-------
Multiple in vitro studies were identified for MWCNTs; a few were selected for discussion here
and for inclusion in Table F-9 in Appendix F to highlight the major themes and important concepts
covered in the literature [see Section 6.3.4.1 for additional studies identified after the collective judgment
prioritization process (Pacurari et al.. 2012)].
One in vitro study on ocular effects of
MWCNTs was identified. The study produced
negative results (i.e., no irritation was observed)
when two sizes of MWCNTs were evaluated
using hen's egg test chorioallantoic membrane
with white leghorn chicken eggs (Kishore et al..
2009).
Negative results also were observed in an
in vitro dermal irritation study by Kishore et al.
(2009). in which two sizes of MWCNTs were
evaluated using the three-dimensional human
epidermis model with human skin cells.
The relevance of dermal in vitro studies to
occupational, consumer, and general public
exposures, however, depends on the ability of
CNTs to penetrate the stratum corneum barrier in
vivo, which is unknown at this time (Monteiro-
Riviere and Inman. 2006).
Radomski et al. (2005) found that
MWCNTs could promote platelet aggregation.
One possible mechanism for thrombus
development following MWCNT exposure
observed by the study authors was a change in the
abundance of GPIIb/IIIa (glycoprotein integrin
receptor), which triggers platelet adhesion. This
study is an example of in vitro data that could
identify data gaps in in vivo data.
Other in vitro studies have found that
Additional Information Highlight Box 15:
15. Developing NexGen-style models for
MWCNT toxicity testing
The toxicokinetics and potential toxicity of MWCNTs are
influenced by an array of factors aside from chemical
composition, including particle agglomeration, fiber length,
and functionalization (Coccini et al., 2010: Liu et al., 2010:
Wako et al., 2010). Given the many ways that these factors
can be combined to generate different types of MWCNTs, in
vitro methods might prove invaluable to quickly evaluate the
potential for these factors to influence MWCNT toxicity and
to support future MWCNT assessment efforts.
For example, Gasser et al. (2012) used an in vitro cellular
model to show that the ability of MWCNTs to cause oxidative
stress, cytokine/chemokine release, and apoptosis was more
related to pre-coating of the MWCNTs with pulmonary
surfactants than to functionalization of the MWCNTs. Thus,
understanding surfactant coating properties might be an
important factor for assessing MWCNT toxicity. Similarly, Liu
et al. (2012) recently used in vitro models to evaluate length-
dependent cytotoxicity of MWCNTs in immune and epithelial
cancer cell lines.
Application of an integrated testing strategy has been
suggested for nanomaterials as it provides a framework for
prioritization, screening, and targeted testing, using multiple
approaches and existing data to address the health effect
concerns regarding exposure to nanoparticles (U.S. EPA,
2009). Several different mechanism-based high-throughput
in vitro methods can be used predictively as screening tools
and then validated using a limited number of in vivo studies
(Nel et al., 2013). Such approaches must be carefully
developed and validated specifically for MWCNTs, as there
is always uncertainty involved in extrapolating in vitro results
to in vivo effects. For nanomaterials, factors such as
agglomeration, aggregation and surface reactivity present
challenges for in vitro testing beyond those of concern for
other types of materials (Hartung and Sabbioni, 2011).
Several efforts are currently underway to develop alternative
testing methods for nanomaterials and improve in vitro to in
vivo extrapolation approaches for these materials [e.g.,
MARINA (IOM, 2013), ENPRA (IOM, 2012),
(http://www.marina-fp7.eu/: http://www.enpra.eu/)1.
MWCNTs induce proinflammatory effects, generate reactive oxygen species and oxidative stress, inhibit
phagocytosis, and induce apoptosis, as reported in a review article (Donaldson et al.. 2006). Similarly
[y, in
5-15
-------
human T cells, MWCNTs were found to decrease cell viability and increase apoptosis in a dose- and
time-dependent manner at concentrations between 40 and 400 ug/mL (Bottini et al., 2006). The study
authors noted that the level of toxicity was significantly greater for oxidized MWCNTs compared to their
pristine counterparts (Bottini et al.. 2006).
5.1.10. Genotoxicity/Mutagenicity
Standard and modified in vitro genotoxicity tests have been conducted to investigate the
genotoxic potential of CNTs. Their genotoxic potential remains uncertain at this time, however, as
available tests have shown contradictory results, which might be due to differences in composition and
physicochemical characteristics of the CNTs (see Table F-9 in Appendix F).
MWCNTs were not found to be mutagenic when evaluated with the Ames test23 using
Salmonella typhimurium with and without metabolic activation (Di Sotto et al., 2009; Wirnitzer et al.,
2009). Bacterial mutagenicity-based assays might not be suitable for detecting genotoxicity induced by
nanoscale materials, however, because prokaryotes cannot perform endocytosis, and the nanoscale
materials might not be able to diffuse across the bacterial cell wall. This lack of uptake could lead to
false-negative results (Singh et al.. 2009). Purified MWCNTs also did not show genotoxic activity in
several assays at different dose levels and in different test systems, including micronucleus and sister-
chromatid exchange assays of human lymphocyte cells (Szendi and Varga. 2008) and a chromosome
aberration assay of Chinese hamster lung cells (Asakura et al.. 2010). The MWCNTs used in these tests
were predominantly high purity and contained minimal metal impurities such as metal catalysts, which
are included in the commercial MWCNT preparation as a result of the synthesis process. Such impurities
could influence the genotoxic potential of MWCNTs.
Other micronucleus assays have reported significant MWCNT-induced increases in micronuclei
of rat lung epithelial cells, Chinese hamster lung cells, and human lung carcinoma A549 cells (Kato et al..
2013; Asakura et al.. 2010; Muller et al.. 2008a). MWCNTs also have acted as clastogenic and aneugenic
agents simultaneously in human blood cells (Cveticanin et al.. 2010). Using the murine macrophage cell
line RAW 264.7, Migliore et al. (2010) observed a significantly increased MWCNT dose-related
percentage of DNA in comet tails in a Comet assay and a significant cytotoxic effect in a Trypan blue
assay. Cavallo et al. (2012) also reported an MWCNT concentration-dependent, statistically significant
induction of direct DNA damage in human lung epithelial A549 cells evidenced by percentage of DNA in
23The Ames test is a bacterial reverse mutation assay, designed to determine mutagenicity of the test compound.
5-16
-------
comet tails in a Comet assay that corresponded with reduced cell viability; the authors noted, however,
that oxidative DNA damage was not statistically significant. Patlolla et al. (201 Ob; 2010a) observed a
statistically significant, dose-dependent increase in the percentage of DNA in comet tails in a Comet
assay in normal human dermal fibroblast cells.
DecaBDE Can Inform MWCNT Assessment
Research indicates that decaBDE does not induce
genotoxicity. DecaBDE does, however, appear in studies of
carcinogenicity to target the liver and thyroid (see Appendix
H). Similar to many chemicals though, important differences
have been observed between species, gender, and target
organs. Based on observations with decaBDE, research
planning to inform future assessments of MWCNT
carcinogenicity might consider: Does exposure to MWCNTs
induce genotoxicity? Is evidence of MWCNT carcinogenicity
reproducible in multiple species, genders, and sites? See
Appendix H for more information about the carcinogenic
impacts of decaBDE on humans, including details on the
NTP study (1986).
5.1.11. Carcinogenicity
Carcinogenicity studies considered for
MWCNTs are presented in Table F-ll of
Appendix F. Currently, the carcinogenic potential
of MWCNTs is unknown. No studies have
investigated carcinogenicity of MWCNTs
following oral or inhalation exposures. Several
studies indicate, however, that certain types of
MWCNTs (e.g., those with high aspect ratios)
behave in a manner similar to asbestos, inducing
mesotheliomas when administered using methods
such as injection (e.g., intrascrotal,
intraperitoneal) (Sakamoto et al.. 2009; Poland et al.. 2008; Takagi et al.. 2008). Therefore, MWCNTs
could be carcinogenic should they reach the mesothelium after inhalation exposure. Similarities between
asbestos and MWCNTs in terms of aspect ratio and biopersistence in the lung, along with a tendency for
MWCNTs to bundle, could explain the observations of mesotheliomas (Schulte et al.. 2010; Sakamoto et
al.. 2009; Takagi et al.. 2008) (see Additional Information Highlight Box 13 for more discussion).
However, Muller et al. (2009). found that a single 20-mg injection of MWCNTs did not produce
mesotheliomas in male Wistar rats observed for 24 months post-exposure. Similarly, Varga and Szendi
(2010) found that peritoneal injection of 10 mg MWCNTs did not result in development of
mesotheliomas in F344 rats examined at 12 months post-exposure, but did result in a granulomatous
reaction.
Currently, the mechanism by which MWCNTs reach and persist in the pleura, including retention
time and the importance of factors such as fiber length or bundle size, is not understood well enough to
determine whether inhalation of MWCNTs could result in mesothelioma (see Section 6.3.4.1 for
additional studies on the potential carcinogenic mechanisms of MWCNTs that were identified after the
collective judgment prioritization process).
5-17
-------
The carcinogenic potential of MWCNTs also might be influenced by the presence of metal
contaminants (e.g., iron or nickel). These contaminants could play a role in carcinogenicity by
accelerating the generation of reactive oxygen species (Johnston et al., 2010).
Additional Information Highlight Box 16:
16. Mechanisms of toxicity for cancer and noncancer impacts of MWCNTs
MWCNTs have been reported to cause pathogenic or carcinogenic effects in several standard toxicity studies, which indicates that
human health impacts might occur as a result of MWCNT exposures (Delormeetal., 2012: Murphy etal., 2011: Donaldson etal.,
2010: Ryman-Rasmussen et al., 2009b). To date, data indicate that pathogenic or carcinogenic effects could arise from several
mechanisms, including increased cellular permeability through reactive oxygen species production and effects on the actin
filament system, as observed in human microvascular endothelial cells (Pacurari et al., 2012). Muller and Nowack (2008)
investigated the genotoxic potential of MWCNT using both an in vivo and in vitro rat model and showed that MWCNT exposure led
to the formation of micronuclei resulting from chromosomal alterations in the lung epithelia.
As described in Additional Information Highlight Box 13, several studies have shown the potential for MWCNTs to act like
asbestos, causing thoracic inflammation, (Morimoto et al., 2012: Murphy etal., 2012: Murphy etal., 2011), early fibrosis (Murphy
et al., 2012: Murphy et al., 2011), and fibrosis (Ryman-Rasmussen et al., 2009b). Additionally, MWCNTs have been shown to
penetrate into the alveolar region of the lung (Delorme etal., 2012) and to cause inflammation due to accumulation of alveolar
macrophages (Schinwaldetal., 2012). Although these biological events have been shown to lead to mesothelioma (Donaldson et
al., 2010), the current science on MWCNTs is not yet sufficient to determine if these mechanisms are responsible for the observed
effects. Continuing to build a greater understanding of the mechanisms of toxicity leading to potential health impacts is important
for future MWCNT risk assessment efforts.
5.1.12. Susceptible Populations
Sacks et al. (2011) defined susceptibility as "individual- and population-level characteristics that
increase the risk of health effects in a population, including, but not limited to, genetic background, birth
outcomes (e.g., low birth weight, birth defects), race, sex, life stage, lifestyle (e.g., smoking status,
nutrition), preexisting disease, socioeconomic status (e.g., educational attainment, reduced access to
health care), and characteristics that may modify exposure ... (e.g., time spent outdoors)." In this section,
populations susceptible to MWCNT impacts based on characteristics such as age, genetic background,
and disease are considered. Characteristics that could modify exposure and increase susceptibility are
discussed in Section 4.2.4; for a discussion on impacts related to socioeconomic status, see Section 5.3.
No information was identified regarding MWCNTs and susceptible populations. Because
MWCNTs appear to induce inflammatory and fibrotic effects (see Sections 5.1.3 and 5.1.8). however,
individuals with existing pulmonary disease and children with higher respiration rates could be
susceptible populations.
5-18
-------
5.2. Ecological Effects
This section presents a summary of data on the potential ecological impacts of environmental
contamination with MWCNTs. Specific information from the studies reviewed for this case study can be
found in Appendix F.2. Considerations for ecological impact include the absolute and relative toxicity of
MWCNTs in different species, types of effects across species, and other factors such as bioaccumulation
and biomagnification potential (see Appendix G.4.2). More than 20 studies have investigated the effects
of MWCNTs on aquatic species or aquatic systems. Short-term exposure studies indicate low acute
toxicity potential of MWCNTs, with the effect level varying based on size and functionalization
properties of the MWCNTs. Chronic studies show that MWCNTs can elicit immune responses and
produce developmental impacts in aquatic species. The terrestrial ecosystem studies focus on
agriculturally relevant plants and soil microbes. Available evidence suggests that MWCNTs exhibit some
antimicrobial effects. Conflicting data exist on whether MWCNTs are detrimental or beneficial to growth
and seed germination in terrestrial plants. In both aquatic and terrestrial ecosystems, studies are
predominantly laboratory-based experiments on single species. Therefore, limited research on the broad
ecological impact of MWCNT exposure is available, which might be the result of limited data on its
presence in the environment. As mentioned in Appendix G.4.2. ecological receptors can be exposed to
MWCNTs attached to textile fibers, embedded in polymers, or sorbed to other particles, all of which are
more likely to occur in the environment than the pristine compound. Studies examining exposure to larger
textile scraps, polymer particles, and other heterogeneous compounds containing MWCNTs, however, are
lacking. The results of laboratory studies using pristine compounds must therefore be considered,
recognizing that results might not translate directly into real-world exposure scenarios.
5-19
-------
5.2.1. Aquatic Receptors
Priority Research Area: Impacts
Aquatic
Biota
c
o o
C £ c
LJ r
o
0.
E.
7/13
5/13
H M L
E-RRF Confidence Rating
1/13
• - 1 participant's vote
Importance
L - Low:
M- Medium
H - High:
Confidence
L - Low:
M- Medium
H - High:
Least Important
Possibly Important
Important
Not Confident
Somewhat Confident
Confident
Tl IT
E-RRF Confidence Rating
area, followed by survival and reproductive effects. See Section
Section 1.1.3 for a detailed explanation of the prioritization process.
H M L H M L
E-RRF Confidence Rating E-RRF Confidence Rating
Seven of 13 RTI workshop participants (54%) identified
impacts of MWCNT exposure on aquatic biota as
important to risk assessment. These seven participants
were asked to rate the importance of survival,
developmental, reproductive, and other sublethal
endpoints to aquatic biota impacts, as well as their
confidence that the existing data on these subjects
could support risk management decisions. Based on
this information, developmental and other sublethal
endpoints were the highest priorities for research in this
6.3.4.3 for more information on this priority area and
DecaBDE Can Inform MWCNT Assessment
Bioaccumulation and bioavailability of polybrominated diphenyl ethers (PBDEs) in sediment and in the water column influence
PBDE toxicity to marine and freshwater invertebrates and algae. Deca-, octa-, and pentaBDE are not acutely toxic to fish up to
the limits of water solubility, but sublethal doses to decaBDE produced some endocrine effects in aquatic species.
Histopathological alterations in thyroid gland shapes and decreases in thyroid hormone expression in tail tissue during
metamorphosis were observed in African clawed frog tadpoles exposed to decaBDE, and these sublethal effects can influence
tadpole growth and development (Qinetal., 2010). Decreased thyroid hormone expression also was observed in Chinese rare
minnow and lake trout after exposure to decaBDE via water and diet, respectively (Li et al., 2011: Tomy et al., 2004).
The relevance of studies exposing fish to decaBDE is complicated because decaBDE in the environment is often transformed
into other congeners via biotic debromination. The impact of debromination on PBDE bioaccumulation, bioavailability, and
effects in aquatic biota can be compared to the impacts of functionalization, surface treatment, and aggregation of MWCNTs.
Research planning to inform future risk assessment of MWCNTs could consider: What is the impact of transformation,
partitioning, and formation of mixtures when considering toxicity data? See Appendix H for more details regarding the effects
of decaBDE on aquatic biota.
Limited information was identified on the toxicity of MWCNTs to algae and aquatic plants (see
Section 5.2.1.1). Only a few studies have investigated toxicity of MWCNTs to aquatic invertebrates and
aquatic vertebrates (see Sections 5.2.1.1 and 5.2.1.2). Although these studies provide information for
acute effects, they vary with regard to endpoints, doses, functionalization, and other material
characteristics (see Text Box 5-1). Little information was identified regarding toxicity to benthic
5-20
-------
invertebrates (see Additional Information Highlight Box 17). and most MWCNTs released to the aquatic
environment are expected to accumulate in the benthic environment (see Sections 3.1 and 3.3). A limited
amount of information was identified on the effects on ecosystems of chronic MWCNT exposure.
As discussed in Sections 3.1 and 3.3. MWCNTs have low water solubility and are expected to
partition to sediment where they might be available primarily to benthic organisms. Physicochemical
properties of MWCNTs suggest potential for bioaccumulation in aquatic systems (Helland et al.. 2007).
but no studies were identified on this topic. The potential impact of bioaccumulation of MWCNTs in
aquatic systems is therefore uncertain.
5.2.1.1. Algae, Aquatic Plants, and Aquatic Invertebrates
Table 5-2 provides a comparison of key reference values identified for the effects of decaBDE
and MWCNTs on algae, aquatic plants, and aquatic invertebrates. More detailed information on decaBDE
is available in Appendix H. Table F-12. Table F-14. and Table F-15 in Appendix F summarize details of
the MWCNT studies identified and reviewed for this section.
Table 5-2. Effects of decaBDE and MWCNTs on aquatic receptors: Algae, plants, and invertebrates.
Organism
Sediment
oligochaetes
Algae
Zebra
mussels
Macrophytes
MWCNTs
Effect
Effect level Citation
ND
Growth
inhibition
LOEC
1 mg/L Wei et al.
(2010)
ND
Positive effects on Velzeboer et
recolonization and community al. (2011)
structure after 3 months of
exposure in sediment
DecaBDE
Effect
Acute NOEC
28-day NOEC
Growth
inhibition,
96-hr EC50
DNA damage
Effect level3
>5,000 mg/kg
>3,841 mg/kg
>1 mg/L
0.1 to10ug/L
Citation
Hardy (2002a)
ACC (2001 a. b)b
Hardy (2002a)
Riva et al.
(2QQ7)
ND
5-21
-------
Table 5-2 (Continued): Effects of decaBDE and MWCNTs on aquatic receptors: Algae, plants,
and invertebrates.
Organism
Sediment/
Benthic
organisms
Water fleas
MWCNTs
Effect
Leptocheirus
plumulosus:
LCso
Hyalella
azteca:
LCso
Ceriodaphnia
dubia:
acute LCso
C. dubia:
Subchronic
growth
inhibition ECso
3-generation
reproduction
EC50
C. dubia: LC/s
Effect level
68 grams/kg
>264 grams/kg
2-1 00 mg/L
50.9 mg/L
4-1 7 mg/L
26 mg/L
(un-derivatized
MWCNTs)
Citation
Kennedy et al.
(2008)
Kennedy et al.
(2008)
Li and Huang
(2011)
Kennedy et al.
(2008)
Li and Huang
(2011)
Kennedy et al.
(2009)
DecaBDE
Effect Effect level3 Citation
ND
Daphnia BDE-209: ND; Wildlife
magna: pentaBDE: International
21-day LOEC 9.8 ug/L report
(growth) submitted to
Chemical
Manufacturers
Association
(Drottar and
Krueqer, 1998)b
21-day EC50 BDE-209: ND; Wildlife
(survival, pentaBDE: International
reproduction) 14 ug/L report
submitted to
Chemical
Manufacturers
Association
(Drottar and
Krueqer, 1998)b
"Concentration in media (water [units: ug/L or mg /L] or sediment [units: mg/kg]).
bAs cited in Environment Canada (2006).
Abbreviations: ND = No data identified, NOEC = No-observed-effect concentration, EC50 = Median effective concentration.
5-22
-------
Additional Information Highlight Box 17:
17. Toxicity to benthic invertebrates
The complexity of ecological systems combined with the wide variety of MWCNT chemistries can make determining how
parameters of the material (e.g., surface charge, aspect ratio) and the environment (e.g., pH, UV light) influence toxicity in
aquatic receptors difficult. For example, Mwangi et al. (2012) recently conducted 14-day toxicity tests in several benthic
species, including an amphipod (Hyalella azteca), a midge (Chironomus dilutes), a mussel (Villosa iris), and an oligochaete
(Lumbriculus variegates). The benthic invertebrates were exposed to sonicated or nonsonicated MWCNTs from two different
sources and to a nitric acid-modified MWCNT. The amphipod H. azteca was also exposed to nonsonicated MWCNTs with or
without the addition of EDTA or a nickel solution. Results showed decreased survival (in all species except L variegates),
growth (evaluated in V. iris only), and biomass (not evaluated in V. iris) of the test organisms. The authors concluded that acid
treatment, sonication, and type and source of MWCNT influenced survival, growth, and biomass in a species-dependent
manner. The authors also reported that the nickel solubilized from MWCNTs and the MWCNTs themselves contributed to
toxicity. See Section 6.3.4.3 for more discussion from recently identified evidence on factors influencing toxicity to aquatic
receptors and the lack of mechanistic data. These studies highlight the need for better characterization of adsorption,
distribution, metabolism, and excretion of nanomaterials in aquatic biota. Such data could help address current gaps in
understanding about bioavailability, potential interaction with other contaminants, and other factors that might determine or
influence MWCNT toxicity in aquatic receptors.
The effects of MWCNTs on marine algae, sediment macrophytes, and water-dwelling
invertebrates have not been extensively studied; only four published studies were initially identified (Li
and Huang. 2011: Velzeboer etal.. 2011: Wei etal.. 2010: Kennedy et al.. 2008). Because MWCNTs are
likely to partition to sediment, benthic organisms are expected to be primary receptors (Christian et al..
2008). Initially, no studies that investigated the effects of MWCNTs on benthic invertebrates were
identified. Some studies were identified, however, after the collective judgment prioritization process
(See Additional Information Highlight Box 17 and Section 6.3.4.3).
In a study on macrophyte growth, experimental plots were cleared of all living organisms and
MWCNTs were added to the sediment. After 3 months, researchers observed that the density of
macrophytes that had recolonized the plots was positively correlated with MWCNT levels (Velzeboer et
al.. 2011). This result was counter to the authors' initial hypothesis based on previous laboratory
experiments that macrophyte growth and species composition would be negatively affected by MWCNTs,
indicating a level of complexity in community-level effects of MWCNTs in real-environment situations
that is not well understood (Velzeboer et al.. 2011).
Unicellular green algae (Dunaliella tertiolectd) exposed to carboxylated MWCNTs in sea water
did not exhibit inhibited growth until concentrations reached 1 mg/L and above (Wei et al.. 2010).
Growth lagged up to 23 days behind the control, and exponential growth rates were reduced by 35% when
exposure was 10 mg/L, indicating mid-exponential growth phase cytotoxicity at high exposures (Wei et
al..201Q).
Two studies provided a wide variety of data for Ceriodaphnia dubia, a species of water flea (Li
and Huang. 2011: Kennedy et al.. 2008). The MWCNT studies describe an acute median lethal
5-23
-------
concentration in the mg/L range, the variation of which might be due to differences in functionalization
treatment and diameter size of the MWCNTs (Li and Huang. 2011) (see Table F-15 in Appendix F for
study-specific details and Text Box 5-1 for discussion of how physicochemical properties affect toxicity).
Subchronic and chronic growth and reproduction tests show that MWCNTs that had been treated to
increase dispersal and limit bundling (a common treatment in MWCNTs—see Section 2.2.3.1) are not
likely to cause significant, population-level effects until high doses (1- to 100-mg/L range) are reached
(Li and Huang. 2011). Even though MWCNTs are expected to partition to sediment, functionalization and
suspension in natural organic matter could improve dispersion and solubility of MWCNTs in aqueous
media and might increase the exposure levels of MWCNTs to water-dwelling aquatic organisms
(ODriscoll et al.. 2010: Kennedy et al.. 2008).
5.2.1.2. Aquatic Vertebrates
Table 5-3 provides a comparison of key reference values identified for the effects of decaBDE
and MWCNTs on aquatic vertebrates. Additional information on decaBDE is available in Appendix H.
Table F-13. Table F-16. and Table F-17. in Appendix F summarize details of the MWCNT studies
identified and reviewed for this section.
Four studies were identified that investigated the effects of MWCNTs on fish; three of these
studies used common laboratory species—zebrafish (Danio rerio) and Japanese medaka (Oryzias latipes)
—while one used rainbow trout (Oncorhynchus mykiss).
In one zebrafish study, embryos acutely exposed to MWCNTs showed dose-dependent increased
mortality rates, reduced blood circulation, and delayed development (hatching), and developmental
defects (bent notochord) starting at 60 ug/mL (Asharani et al.. 2008). The other zebrafish study (Cheng et
al.. 2009) examined embryos following a smaller yet more direct exposure—a single microinjection of
MWCNTs—and observed changes in enzyme expression signifying an immune response. Although no
increase in mortality or developmental defects was observed in the exposed zebrafish through adulthood,
survival in the second generation was significantly decreased. Similarly, Kim et al. (2012b) observed a
statistically significant increase in heart abnormalities, absence of swim bladders, caudal fin
malformation, and pericardial and peritoneal edemas in Japanese medaka embryos following 4 days of
continual exposure to 2,000 ug/L functionalized MWCNTs in the water. Increased mortality also
occurred in response to exposures to 1,500 and 2,000 ug/L, and exposure to 1,500 ug/L resulted in a
hatching delay.
Klaper et al. (2010) further investigated the possibility of an immune response in fish with an in
vitro study of rainbow trout. In this study, MWCNTs did not elicit an antiviral response at sublethal doses
5-24
-------
up to 10 ug/mL, regardless of various types of functionalization. Investigators did, however, observe
expression of IL-lbeta, evident of macrophage stimulation, at 5 and 10 ug/mL (Klaper et al., 2010).
Table 5-3. Effects of decaBDE and MWCNTs on aquatic receptors: aquatic vertebrates.
Organism
Exposure
Effect
Effect level
Citation
MWCNTs
Zebrafish
Japanese
medaka
Water
Water
Water
Microinjection
Microinjection
Water
Reduced blood circulation
Developmental effects
Increased mortality LOAEL
Developmental effects NOEL
Second-generation reduced survival
Developmental effects LOAEL
70 ug/mL
60 ug/mL
60 ug/mL
>2 ng/embryo
2 ng/embryo
1,500 ug/mL
Asharani et al. (2008)
Asharani et al. (2008)
Asharani et al. (2008)
Cheng et al. (2009)
Cheng et al. (2009)
Kim et al. (2012b)
DecaBDE
African clawed
frog
Lake trout
Rainbow trout
Lake whitefish
Chinese rare
minnow
Water
Diet, chronic
Diet, chronic
NR
Diet, chronic
Water, chronic
Water, chronic
Water, chronic
Thyroid effects LOAEL
Decreased thyroid hormones
LOAEL
Increased liver weight LOAEL
Vitellogenin production
Decreased growth LOAEL
Decreased growth LOAEL
Spermatogenesis inhibition LOAEL
Upregulation of thyroid hormones
1 ng/L
2.5 ng/gram
7.5 ng/kg
NR
2 ug/gram
10 ug/L
10ug/L
Variably occurred
at 0.1 -10 ug/L
Qin et al. (2010)
Tomy et al. (2004)
Kierkegaard et al. (1999)
Nakari and Pesala (2005)
Kuo et al. (2010)
Li et al. (2011)
Li et al. (2011)
Li et al. (2011)
Abbreviations: NR = Not reported, NOAEL = No-observed-adverse-effect level, LOAEL = Lowest-observed-adverse-effect level
5-25
-------
5.2.2. Terrestrial Receptors
Unprioritized Research Area: Impacts
Terrestri,
0,1
i C
UJ «
o
Important
Not Confident
Somewhat Confident
Confident
o
or
£S
t
o
Environmental
Resources
~R J\7T~ L ^H M L H M L
E-RRF Confidence Rating E-RRF Confidence Rating E-RRF Confidence Rating
Five of 13 RTI workshop participants (38%) identified other impacts of MWCNT exposure
as important to risk assessment. These five participants were asked to rate the
importance of economic, societal, and environmental resources impacts as well as their
confidence that the existing data on these subjects could support risk management
decisions. Based on this information, these areas were deemed a priority for research.
See Section 6.3.4.4 for more information on this priority area and Section 1.1.3 for a
detailed explanation of the prioritization process.
5-26
-------
As stated in Chapter 1, the CEA framework considers not only human and ecological health
impacts, but also aesthetic, environmental, social, legal, ethical, and economic impacts. Such impacts
might be associated with impacts on specific socioeconomic sectors (e.g., disparate impacts on
environmental justice communities), the environment as a whole (e.g., climate change, depletion of
natural resources, energy demand), or the built environment (e.g., damage to building facades).
Apart from the impacts discussed in Sections 5.1 and 5.2. the only other impacts considered in
this case study are those for which a plausible premise can be developed to support assumptions that a
discernible impact might occur as a result of the life cycle of MWCNTs flame-retardant upholstery textile
coatings. Data from MWCNTs on each of these impacts would be required for a thorough comparison;
however, in all cases where other impacts were identified as being of concern for either decaBDE or
MWCNTs, no data were available for the comparison material. For example, for decaBDE, empirical data
have revealed a correlation between decaBDE body burdens and socioeconomic status, indicating that
effects having environmental justice implications are plausible for decaBDE. No such empirical data exist
relating MWCNTs to other impacts, but the background literature on processes involved in manufacturing
similar materials [e.g., carbon nanofibers (CNFs) and SWCNTs] provides some basis for concerns
regarding potential impacts of MWCNTs on energy demand, resource depletion, climate change, and
economics.
5-27
-------
5.3.1. Environmental Justice
Environmental justice is defined by
EPA as the "fair treatment and meaningful
involvement of all people regardless of race,
color, national origin, or income with respect to
the development, implementation, and
enforcement of environmental laws, regulations,
and policies."24 The goal of environmental
justice is to give all people ".. .the same degree
of protection from environmental and health
hazards and equal access to the decision-making
process..." As a result, environmental justice
DecaBDE Can Inform MWCNT Assessment
Releases of decaBDE throughout the life cycle of a flame-
retardant upholstery coating product could have greater
impacts on racial and ethnic minorities and populations of lower
socioeconomic status, as studies have found higher body
burdens of BDE-209, the single isomer of decaBDE, among
these populations. The causal pathway of this disproportionate
exposure is not well understood, but certain characteristics of
living environments (e.g., smaller size, poorer ventilation, older
age) might contribute to increased exposure levels for lower
income families or individuals. Research planning to inform the
societal impacts of MWCNTs in future risk assessments could
consider: Are certain populations at greater risk of negative
MWCNT impacts because of social, economic, or cultural
differences? See Appendix H for more information about
unintended societal impacts of decaBDE.
impacts include those in which a particular group or geographic area experiences a disproportionate share
of the impacts associated with an environmental contaminant.
No information was identified that examined the relationship between MWCNT exposures and
socioeconomic status.
5.3.2. Energy Demand and Natural Resource Depletion
No information was identified that examined impacts on energy demand and natural resource
depletion associated with the production of MWCNT flame-retardant textile coatings. Various studies,
however, have calculated a large range of minimum energy requirements for synthesis of CNTs (Khanna
et al., 2008; Cipiriano et al., 2007; Smalley et al., 2007). Table 5-4 presents an overview of some
estimated minimum energy requirements and process rates for CNT synthesis.
DecaBDE Can Inform MWCNT Assessment
No information was identified that examines impacts on energy demand and natural resource depletion associated with the
production of decaBDE flame-retardant textile coatings. Therefore, decaBDE does not provide information on environmental
resource demands that could be applicable to research for future risk assessment of MWCNTs.
U.S. EPA Compliance and Enforcement. Environmental Justice, http://www.epa.gov/environmentaljustice/
5ibid
5-28
-------
The differences in energy requirements are largely attributable to different synthesis processes,
different process rates, different feedstocks, and process improvements as synthesis of CNTs has been
optimized (Gutowski et al., 2010; Kushnir and Sanden. 2008). Additionally, these energy requirements
represent only the minimum for synthesis and do not consider the energy required for purification,
additional infrastructure (e.g., equipment needed to regulate environmental conditions during synthesis
and processing), and other related processes. Including these additional energy requirements, Gutowski et
al. (2010) estimated that CNTs could be one of the most energy-intensive materials of all time. Although
information regarding the energy requirements for MWCNT synthesis is limited, such energy
requirements likely also would be sizeable and span a large range as synthesis processes are continually
optimized.
One environmental impact assessment examined the water inputs required for two methods of
continuous synthesis of SWCNTs via chemical vapor deposition (CVD) (Kolosnjaj-Tabi et al.. 2010).
These results indicated that production of SWCNTs can require significant amounts of water (almost
65,000 kg/hour at a manufacturing rate of about 595 kg/hour of SWCNTs). The relationship between
water requirements for SWCNT synthesis and MWCNT synthesis, however, is unclear.
A life-cycle assessment of CNF production via CVD calculated potential impacts on
acidification, eutrophication, and ozone layer depletion (Khanna et al., 2008). The results of this analysis
are presented in Table 5-5. The authors found that CNF production of both methane-based and ethylene-
based CNFs has minor impacts on acidification, eutrophication, and ozone layer depletion. This study did
not incorporate CNF emissions into its calculations, however, due to a lack of data on fate, transport, and
impacts of CNFs (Khanna et al.. 2008). The authors noted that the lack of models to predict endpoint
effects of some emissions (e.g., CNF emissions) renders these calculations uncertain. Plata et al. (2009)
found that release of gases such as methane, volatile organic compounds, and poly cyclic aromatic
hydrocarbons from MWCNT synthesis is possible. Methane release likely would have a negligible impact
on local air pollution and ozone depletion compared to existing methane sources; however, release of
volatile organic compounds such as 1,3-butadiene and benzene, could be significant on a local scale
(Plata etal.. 2009).
5-29
-------
Table 5-4. Estimated minimum energy requirements and process rates for synthesis of CNTs.
Synthesis
process Material
Vapor- CNF (methane-
Grown based)
CNF CNF (ethylene-
Process based)
Arc SWCNT
Ablation
HiPCO® SWCNT
HiPCO® SWCNT
Floating
Catalyst
CVDb
MWCNT
Laser
Ablation"
MWCNT, SWCNT
Synthesis energy
Process rate requirements
(kg/hr) (J/kg)
1.30X10'2 3.13 x 109
1.80X10'2 2.22 x 109
8.10 xlO'5 8.73 x 1010
4.50 x 10'4 2.41 x 1010
4.50 xlO'4 3.18 x 1010
2.95 x 108
(thermal
baseline)
Mr?
1.87x 108
(electricity
baseline)
2.11 x 108
(thermal
baseline)
Mr?
9.4 x 109
(electricity
baseline)
Estimated
energy per
hour (J/hr) Citation
4.07 x 107 Khanna et al. (2008); as
cited by Gutowski et al.
A 00 * 107 (2010)
Healy et al. (2008) and
7 07xi n6 Isaacs et al. (2008); as
cited by Gutowski et al.
(2010)
Smalley et al. (2007); as
1 .08 x 1 o7 cited by Gutowski et al.
(2010)
Healy et al. (2008) and
7 Isaacs et al. (2008); as
cited by Gutowski et al.
(2010)
MD Kushnir and Sanden
NR (2QQ8)C
MD Kushnir and Sanden
NR (2QQ8)C
aBenzene gas feedstock.
bGraphite feedstock.
°Authors report both baseline (shown above) and "efficient" estimates (not shown).
Abbreviations: CVD = chemical vapor deposition; CNF = carbon nanofiber; HiPCO1
process; NR = not reported.
Source: Reprinted with permission of IEEE, Gutowski et al. (2010): and Kushnir and
= a high pressure carbon monoxide synthesis
Sanden (2008).
5-30
-------
Table 5-5. Environmental assessment of production of 1 kilogram (kg) of carbon nanofibers.
Impact13
Impact category3
Acidification potential
Eutrophication potential
Ozone layer depletion potential
Methane-based
CNF
5.5
0.4
2.8 x 10'5
Ethylene-based
CNF
4.0
0.3
2.8 x 10'5
Unit
Kg SO2 Equivalent
Kg PO4 Equivalent
Kg CFC-11 Equivalent
Environmental impacts of production of 1 kg of carbon nanofibers (CNFs) using chemical vapor deposition calculated by SimaPro*
Eco-lndicator 1999 (EI99) method, hierarchist perspective (long-term; substances included if there is consensus regarding their
effect; damages avoidable by good management; fossil fuels assumed not easily substituted).
""Normalized and weighted impacts.
Source: Reprinted with permission of John Wiley and Sons; Khanna et al. (2008).
Eckelman et al. (2012) developed a life-cycle framework to compare the impact on ecological
(and specifically aquatic) organisms of CNT production versus CNT releases to environmental media
during the product life cycle. This assessment used existing data and a recently established consensus
model for life-cycle impact assessments, USEtox, to estimate ecotoxicity from emissions during CNT
production and CNT releases during product use and disposal for "realistic" and "worst case" scenarios.
They calculated the potentially affected fraction of aquatic organisms per unit mass of CNTs released and
"comparative toxic units for ecosystems" for different methods of synthesis and projected scale-up
results. The theoretical framework was useful for comparing the relative impacts of different synthesis
methods, what proportion of potential ecotoxicity is due to the synthesis process compared to required
purification methods, and also made projections based on future increased scale of production.
The authors concluded that the greatest ecotoxicity impacts do not result from release of CNTs during the
product life cycle or from unused reagents or synthesis products during production, but rather from the
emission of metals due to the combustion of fossil fuels necessary to generate electricity for CNT
synthesis or production of various inputs. Notably, authors needed to make modifications to the USEtox
model to apply it to CNTs because USEtox was developed for organic chemicals. The extent to which
USEtox, even with modifications, is suitable for evaluating CNTs is thus an important consideration
when reviewing the results of this study.
Another study analyzed a broad range of environmental impacts from the production of one
SWCNT polymer mesh (a transistor/electromagnetic interference-shielding application) (Dahlben and
Isaacs. 2009). The study examined both the energy requirements for raw material extraction and
manufacturing and emissions from these processes (excluding SWCNT emissions due to the current lack
5-31
-------
of consensus on its effects). The authors found that manufacturing this CNT application could damage
ecosystems and resource quality, as shown in Table 5-6. Damage to ecosystems (due to
acidification/eutrophication and land use) was expressed as the loss of species over a certain area in a
given time. Damage to resource quality was expressed as the surplus energy needed for future extractions
of minerals and fossil fuels (due to the resources needed to extract these materials for SWCNT polymer
mesh production).
Table 5-6. Environmental assessment of production of one SWCNT polymer mesh.
Impact category3 Impact13 Unit
Acidification/Eutrophication 1.229x-|0"2 PDF*m2yr
Land use 4.440 x iry3 PDF*m2yr
Minerals 1.117x-|0"3 MJ Surplus
Fossil fuels 7.531 x 10"1 MJ Surplus
"Environmental impacts of production of one SWCNT polymer mesh by high-pressure carbon monoxide synthesis (a form of
chemical vapor deposition) calculated by SimaPro® Eco-lndicator 1999 (EI99) method, hierarchist perspective (long-term;
substances included if there is consensus regarding their effect; damages avoidable by good management; fossil fuels assumed not
easily substituted).
bNormalized and weighted impacts.
Abbreviations: PDF*m2yr = potentially disappeared fraction per area (m2) per year;
MJ Surplus = Additional megajoules of energy required for future extraction of the resource.
Source: Dahlben and Isaacs (2009).
The authors found that production of SWCNT polymer mesh generated larger fossil-fuel impacts
relative to other measured environmental impacts. The authors also found that fossil-fuel impacts were
dominated by processes requiring energy-intensive equipment (e.g., furnace for synthesis, wet bench for
cleaning, and spinner for coating). Although the impact measures were reported to be low for all
categories, they represent the resource impacts of producing a single SWCNT polymer mesh. The level of
aggregation required to compile these metrics, however, coupled with the lack of models to predict
endpoint effects of some emissions (e.g., from nanotubes), render these metrics highly uncertain (Khanna
et al.. 2008). Additionally, this study did not include impacts of SWCNT emissions in its overall
calculations of environmental impact. How the environmental impacts of SWCNT polymer mesh
production differ from the impacts of flame-retardant textiles using MWCNTs or other alternative
materials, such as decaBDE, is unclear.
5-32
-------
5.3.3. Climate Change
No information was identified that directly examined climate change impacts due to MWCNT
flame-retardant coatings. Empirical data suggest, however, that release of greenhouse gases such as
methane, volatile organic compounds, and polycyclic aromatic hydrocarbons from MWCNT synthesis is
possible (Plata. 2009; Plata et al.. 2009). The authors concluded, however, that for commercial-scale
production, the contribution of MWCNT synthesis to atmospheric methane will be negligible compared
to existing methane sources, and that volatile organic compound emissions might be significant only on
the local scale (Plata etal.. 2009). Singh et al. (2009) calculated emissions of 4 kg CO2/kg SWCNT from
one method of CVD synthesis. Altering CNT synthesis methods might minimize formation of these or
other hazardous by-products. For example, Plata (2009) found that by identifying select thermally
generated compounds correlated with CNT growth rate, such compounds could be delivered to the
catalyst without thermal treatment and thereby eliminate the need to heat reactant gases.
On the other hand, a life-cycle assessment of CNF production calculated that manufacture of 1 kg
of methane-based CNFs equals at least 700 kg of CO2 equivalents, and 1 kg of ethylene-based CNFs
equals at least 400 kg of CO2 equivalents (Khanna et al., 2008). In other words, the authors calculated that
production of 1 kg of methane-based CNFs is equivalent to CO2 emissions from 78.5 gallons of gasoline
consumed, while production of 1 kg of ethylene-based CNFs is equivalent to CO2 emissions from
44.8 gallons of gasoline consumed.26 The relationship between climate change effects and synthesis of
SWCNTs, CNFs, and MWCNTs is unclear.
5.3.4. Economics
No information was identified that
calculated the cost of manufacturing MWCNTs
or MWCNT flame-retardant textiles. Isaacs et
al. (2010) estimated, however, that the cost of
manufacturing 1 gram of SWCNTs by arc
discharge, CVD, and HiPCO® (a type of CVD
commonly used to manufacture SWCNTs) is roughly $1,906, $1,706, and $485, respectively. These
estimates include all materials, labor, and equipment necessary for synthesis, dispersion, filtration,
inspection, and packaging of SWCNTs. MWCNTs are generally thought to be less expensive to produce
DecaBDE Can Inform MWCNT Assessment
No information was identified that calculated the cost of
manufacturing decaBDE or decaBDE flame-retardant textiles.
Therefore, decaBDE does not provide economic impact
information that could be applicable to research for future risk
assessment of MWCNTs.
26EPA GHG Calculator available at http://www.epa.gov/cleanenergv/energy-resources/calculator.html.
5-33
-------
than SWCNTs, and further optimization of MWCNT manufacturing is likely to decrease manufacturing
costs further.
5-34
-------
Chapter 6. Identifying and Prioritizing
Research Needs to Support Risk
Assessment and Risk Management
6.1. Context for Identifying and Prioritizing Research
Previous chapters in this case study represent the assembly of information through the vertical
continuum of the comprehensive environmental assessment (CEA) framework (see Figure 1-1) as
introduced in Chapter 1; however, this step is merely the first in the CEA process (see Figure 1-2).
The second step is for a diverse group of expert stakeholders to consider the information compiled in the
framework in the context of their own knowledge of multiwalled carbon nanotubes (MWCNTs) and
flame-retardant materials. This second step serves to identify and prioritize research needs for future risk
assessment efforts that inform risk management practices for MWCNTs in flame-retardant textile
coatings. The outcomes of this step are the focus of the current document and are intended to support
subsequent efforts within the CEA process to facilitate an iterative communication flow across the
horizontal spectrum of research, risk assessment, and risk management (see Figure 6-1).
Figure 6-1. Iterative communication flow in the CEA framework.
Recently, the National Research Council and others in the scientific community have made
several recommendations to improve risk assessment and risk management approaches.
Recommendations include calls for greater transparency and increased stakeholder engagement in
assessment efforts for evaluating options to mitigate the exposures or hazard(s) associated with an agent
(NRC. 2009). Greater transparency and broader stakeholder input promote informed evaluations of the
6-1
-------
various trade-offs between individual risk management options (NRC. 2009). More recently, the need to
consider longer term consequences of alternative options in a broader context has been recognized, which
would encompass social, environmental, and economic indicators (NRC. 2011). All of these
recommendations indicate the need for risk assessments to address cumulative effects from multiple
exposures to one or more stressors, vulnerability of susceptible populations, and potential for impacts
throughout the product life cycle (NRC. 2011. 2009). A shift toward this more holistic, systems-based
approach would provide more complete information to risk assessors to better inform risk managers in
making decisions that support long-term, sustainable management practices (NRC. 2011).
Many efforts are underway to gather information and develop approaches that support the
implementation of such recommendations [e.g. (Anastas. 2012; Lavoie etal.. 2010; Rossi etal.. 2006)1.
The CEA approach represents one such effort by recognizing that research supporting risk assessment
must be transparently planned and executed if risk assessments are to be used effectively and efficiently
for evaluating risk management options and understanding the longer term consequences of a broad scope
of complex information (e.g., cumulative risk, life-cycle analyses). The External Review Draft of this
document was used in the collective judgment step of the CEA process as part of an effort to plan such
research for MWCNTs in flame-retardant textile coatings. Specifically, a group of expert stakeholders
representing a variety of technical backgrounds (e.g., material characterization, human health effects,
exposure and dose) and sectors (e.g., industry, academia, nongovernmental organizations) used this
document as a starting point for identifying and prioritizing research needs to support assessments that
inform near-term risk management goals (see Section 1.1.3).
In planning research to support assessments, reviewing risk management goals is useful to
understand the types of analyses and assessments that would inform management efforts. Such a review is
consistent with the connections highlighted in the CEA process diagram (see Figure 1-2). Research
outcomes in the form of data from single studies or assessment reports [which have compiled information
across multiple studies (e.g., risk assessments, life-cycle assessments, meta-analyses)], feed back into the
CEA framework to provide additional information necessary for assessing risk-related trade-offs. This
additional information is subsequently used in developing adaptive risk management plans. Such future
evaluations of risk-related trade-offs could focus on informing one of a variety of risk management goals
for the application of MWCNTs in flame-retardant textile coatings. As discussed in Chapter 1. the
selection of any one goal for a comparative CEA would depend on the needs of risk managers at the time.
One particular scenario, however, based on what might occur for this specific nanoenabled product, is
described below as an example that might inform research planning for the research priorities identified in
this application of CEA. The use of this scenario is not meant to imply actual assessment or risk
management recommendations, but rather to illustrate the type of context in which the research priorities
6-2
-------
identified here could inform future evaluations of MWCNTs. Additional considerations and examples of
risk assessment and risk management decisions are discussed in Section 6.2.
As outlined in Chapter 1, the use of MWCNTs in flame-retardant textile coatings is not common,
although evidence suggests that they could be used more extensively in the future as conventional flame-
retardants such as decaBDE are phased out due to concerns surrounding environmental persistence and
human health effects (see Additional Information Highlight Box 1 and Section 1.1.3). Should a flame-
retardant textile coating containing MWCNTs be developed for use in the United States, the manufacturer
or importer likely would submit a premanufacturing notice (PMN) (U.S. EPA. 2008c). Information in a
PMN includes the identity of the chemical (i.e., name and structure), anticipated production volume, use
and disposal methods, human exposure estimates, and any readily available test data (U.S. EPA. 2010g).
Thus, one of the first risk management decisions for MWCNT flame-retardant textile coatings could be to
determine whether the material should be (1) produced without restriction or regulations, (2) imported,
produced, or used with limitations, or (3) prohibited from import, production, or use (U.S. EPA. 2010f).
The third outcome, prohibition, could result from several determinations, including insufficient
information on potential impacts of the material (U.S. EPA, 2010f).
To support this and other risk management decisions about the use of MWCNTs in flame-
retardant textile coatings, information must be readily available to decision-makers to enable a considered
determination within the relevant time constraints. Ideally, the research objectives identified and
prioritized in this CEA application would provide such information within 3 to 5 years of initiating the
research. Notably, completion of all research is not feasible within this time frame due to practical
constraints. Thus, the collective judgment step of the CEA approach emphasizes the prioritization of
information gaps (see Figure 1-2).
6.2. Identification and Prioritization of Research Needs
With the above context for identifying and prioritizing research needs to support future
assessments and risk management decisions for MWCNTs, the lingering question is: How? As discussed
in the previous section and in Chapter 1, a group of expert stakeholders representing diverse technical
(e.g., human health effects, ecological effects, material characterization) and sector (e.g., industry,
academia, government) perspectives participated in a structured collective judgment process that
supported equal representation of each individual's input. Specifically, participants used a more detailed
view of the CEA framework (see Figure 1-3) to consider each element of the framework (e.g., stages of
the product life cycle or spatial zones in the environment) in relation to potentially relevant risk factors
(e.g., mobility in air) associated with the life cycle of an MWCNT flame-retardant textile coating product.
6-3
-------
Stakeholders used an online software-based tool that supported their consideration of each area of the
CEA framework remotely. They used the tool in a series of exercises to qualify the (1) importance of
CEA framework components for future assessments, and (2) current state of the science of the
components for supporting risk management initiatives (see Figure 1-3). These exercises were intended to
actively engage stakeholders in identifying critical research needs based on each stakeholder's
(1) perception of the information presented in the case study, (2) individual experience, and (3) expert
opinion. Based on previous applications of CEA, this approach is thought to be an improvement over one
that would simply ask stakeholders to review and prioritize a predetermined list of data gaps, which could
exclude important data gaps a priori. This collective judgment exercise, coupled with a subsequent
structured, face-to-face workshop, provided a set of specific research needs (see Section 6.3).
As discussed in the previous section, planning and conducting research to fill the priority areas
discussed below (see Section 6.3) are facilitated by having an understanding of assessment and risk
management approaches. To help facilitate the transition from reviewing the case study to identifying and
prioritizing specific research needs, experts participating in the collective judgment exercise were
encouraged to consider the first step of conducting an assessment—problem formulation. Considering
problem formulation can similarly help plan research to fill the gaps that experts identified as priorities
(see Section 6.3). Placing a greater focus on problem formulation is recognized as an important step
toward improving risk assessment (NRC. 2009). but the preceding step—anticipating what information to
research to support problem formulation—has received less attention. The CEA approach places greater
emphasis on problem formulation during the research planning stage by engaging stakeholders to
prioritize areas for research managers to consider in allocating research resources. In doing so, the
approach ensures that (1) major factors important to stakeholders are included early in the planning
process, and (2) risk assessors and managers have the information they need to develop risk assessments
and management plans that include those considerations. This approach is consistent with recent National
Academy of Science recommendations (NRC. 2011. 2009). Similarly, the inclusion of long-term, broad
environmental impacts in the CEA framework supports identifying research gaps in these areas, which are
increasingly recognized as essential to assessments but often are more difficult to incorporate because of
insufficient data or knowledge on how to include such information (NRC. 2011).
Thus, in planning research to support future assessments and risk management efforts, having a
clear understanding of problem formulation is essential. This early step in the risk assessment process
establishes the goals, scope, focus, and potential options to consider in decision-making (NRC. 2009; Van
Leeuwen et al.. 1998). The problem formulation step helps establish the type of analyses, and associated
uncertainty and variability, that will be useful to a risk manager in making a decision about the material,
chemical, or technology of focus (NRC. 2009; Van Leeuwen et al.. 1998). Then, it is important to
6-4
-------
establish the type of research useful for the analyses that inform a risk manager about a particular
material, such as MWCNTs. Notably, planning for a variety of types of risk management decisions, as
shown in Table 6-1, is necessary. Although this particular CEA case study was not developed with a
specific risk management objective in mind, the identification of key research gaps within the CEA
framework can inform research planning that supports multiple assessment objectives identified during
the problem formulation stage of each respective assessment. These future assessments then can supply
new information in the CEA framework for MWCNTs, which could subsequently be evaluated in
collective judgment prioritization of risk-related trade-offs to inform specific risk management decisions.
Table 6-1. Examples of risk management decisions.
Area of Decision-Making
Directed At
Example Decisions
Product environmental health
and safety
• New chemicals
• Existing chemicals
• Biotechnology
• Pre-manufacturing notices
• Pesticide re-evaluations
• Permits to release genetically modified
organisms
Site management
• Risk avoidance
• Risk mitigation
• Site location
• Accidental releases
• Cleanup of hazardous waste landfills
• Degree of contamination, presence of
endangered species
Natural Resource Use
• Habitat integrity
• Species introductions
• Land use (e.g., road construction, mining,
agriculture, logging)
• Integrated pest management
Adapted with permission of Society of Environmental Toxicology and Chemistry (SETAC) Press; Van Leeuwen et al. (1998).
The External Review Draft of this document used the CEA framework to lay out the technical
aspects related to MWCNTs in flame-retardant textile coatings so that expert stakeholders could
determine (1) what types of risk management decisions are likely to be needed and (2) what information
would support assessments that inform those decisions. Stakeholder input was incorporated into this final
version of the case study document to emphasize areas of the CEA framework that stakeholders felt were
research priorities for risk assessment and risk management. The questions in Table 6-2 are examples of
those that could be asked during the problem formulation phase of an assessment to evaluate whether data
are available and useful (e.g., consider relevant endpoints, relevant exposure routes, doses and timing,
acceptable levels of uncertainty in assays selected, data variability) to support each component of an
assessment. The answers to these questions can help formulate an appropriate assessment approach or
facilitate the identification of additional data gaps that must be filled before the assessment can proceed.
6-5
-------
In the context of this case study, the extent to which these questions, along with other questions identified
by stakeholders, could be answered indicates whether specific research initiatives should be pursued in
certain areas of the CEA framework.
Table 6-2. Example questions for problem formulation.
General Areas
Specific Questions
What are the characteristics of the
stressor of concern?
What are the characteristics of the
exposure setting?
What are the characteristics of the
exposed populations?
What are the assessment
endpoints?
Is the stressor of concern chemical, physical, or biological?
What are the physicochemical characteristics of the stressor?
What are the locations and quantities of releases of the stressor to different
media?
What are the known concentrations of the stressor in different media?
What processes move the stressor through the environment?
How does the stressor change as it moves through the environment?
What is the spatial scale over which exposures to the stressor are likely to
occur?
Which individuals, populations, or population segments are expected to be
exposed?
Which species and trophic-level relationships are present in exposed
ecosystems?
What are the probable exposure routes and pathways for the population(s) of
interest?
Is exposure to the stressor expected to occur only during a single event or will
exposures be episodic or continuous?
What is the time scale over which exposures to the stressor are likely to occur?
What adverse effects have been observed in the population(s) of interest?
What are the most sensitive species and measured endpoints?
What processes affect the behavior of the stressor within the receptor?
How does the stressor change as it moves through the receptor?
What biological mechanisms are involved in the formation of adverse effects?
What social conditions or impacts might result from the stressor?
What economic conditions or impacts might result from the stressor?
What natural resources might be affected and how?
What ecosystem services might be altered and how?
6-6
-------
6.3. Research Priorities Identified through Collective
Judgment in this Application of CEA
The collective judgment process to engage expert stakeholders, which was summarized in Section
6.2 and in Chapter 1. is described in detail in a separate report, which was prepared by the contractor that
independently conducted the EPA-funded workshop (RTI International. 2012). The areas of the CEA
framework that experts most commonly identified as "important to consider in a risk assessment"
(presented in Chapter 2 through Chapter 5) are discussed here in the context of supporting research
planning and future assessment and risk management of MWCNTs in flame-retardant textile coatings.
For each area, why experts rated certain research areas as priorities is first discussed. When
participants individually rated an area's importance and their confidence that the current data could
support risk management, they had the option of completing a checklist of "influential factors." These
factors (Table 6-3) represent various aspects of MWCNT science that could play a role in determining,
inducing, or otherwise influencing the potential risks associated with the particular area in question and
are listed after the bulleted rationales provided by participants in each section. As this part of the
prioritization process was voluntary, not all participants chose to identify influential factors and omission
of an influential factor does not necessarily imply that the factor is unimportant. Nevertheless,
identification of the influential factors for each area provides some additional insight as to why a
particular area might be a priority, and what particular aspects of that area might warrant further research.
Next, other relevant literature, which was not included in previous drafts of the case study, is
discussed in terms of whether that literature might improve confidence that the data could support risk
management decisions, or how the data might support research planning. This literature was not included
in previous drafts of the document either because it was published after the document was made available,
or it was not identified in literature searches completed during document development. Expert reviewers
or public commenters identified the references during document reviews, and thus the references are
included in the sections below to build on information presented in Chapter 2 through Chapter 5. Finally,
for those areas that expert stakeholders discussed at the workshop, the potential risk scenario that they
identified for the area is outlined along with the types of risk management decisions they noted might be
made to mitigate or avoid the potential risk. The type(s) of assessments that could inform these types of
risk management decisions that participants noted are laid out, followed by a table detailing the key
research questions or areas that experts identified along with their estimates of the financial and time
6-7
-------
resources needed to carry out the research.27 For those Priority Research Areas not discussed at the
workshop due to time constraints, potential research directions are listed in the text. Priority Research
Areas are presented below in the order in which they appear in the CEA framework and in Chapter 2
through Chapter 5 of this document.
Table 6-3. "Influential Factors" options for all areas.
Option:
Methods,
Techniques
Engineered Nano
Material
Characteristics
| Surrounding
Media
Physical
Conditions
Chemical
Conditions
Biological
Conditions
Social Conditions
Description:
Analytical techniques, control technologies, MWCNT processing methods,
MWCNT purity, MWCNT synthesis methods, personal protective equipment, other
(specify other)
Adsorption/desorption ability, aggregation/agglomeration state, applied coatings,
biodegradability, catalytic activity, charge, conductive or magnetic properties,
crystalline phase, lipophilicity, matrix bound vs. free form, morphology,
persistence, redox potential, size/size distribution, specific surface area, structural
formula/molecular structure, surface chemistry, water solubility/dispersibility, other
(specify other)
Air, ground water, sediment, soil, surface water, wastewater, other (specify other)
Flow regime, light availability, soil porosity, soil/sediment fractionation,
temperature, wind, other (specify other)
Conductivity, dispersing agents, dissolved oxygen content, exposure to sunlight,
heavy metals in environment, ionic strength in environment, ligand concentrations
in environment, natural organic matter, other contaminants in environment, pH,
protein concentration in environment, salinity, surfactant (in lab) other (specify
other)
ADME, bioaccumulation, biomagnification, microbial communities, organism
health, species/individual developmental behavior, species/individual feeding
behavior, species/individual reproductive behavior, other (specify other)
Acute exposure, chronic exposure, exposure route, geographic location, habitat
structure, human activity, individual activity level, life stage, occupation,
subchronic exposure, susceptible populations/individuals, other (specify other)
6.3.1. Product Life Cycle
Workshop participants considered multiple areas of the product life cycle to be Priority Research
Areas. Material synthesis and processing and product manufacturing were noted to be particularly
important by workshop participants, in part because these areas of the CEA framework are instrumental in
Experts participating in the independently conducted RTI workshop generated time and financial estimates. The
estimates are based on participants' experience with planning and conducting scientific research; they are included
here to provide a general indication of the level of resources that would support answering the identified priority
research questions.
6-8
-------
determining all other parts of the CEA framework, including fate and transport in various environmental
media, bioavailability and potential exposure, and ecological and human toxicological impacts. In a recent
publication, Nowack et al. (2012) similarly note the importance of characterizing engineered
nanomaterials throughout the life cycle: How the starting material is intentionally modified through
functionalization (see Figure 2-2) and unintentionally modified by environmental factors (see Figure 3-1)
determines how it will behave in the environment, how organisms can be exposed, and the extent to
which it will adversely impact ecological and human receptors [(Nowack et al.. 2012): see Additional
Information Highlight Box 61. Although understanding chemical alterations to nanomaterials is essential
for evaluating potential release and toxicity, the properties of the product matrix and how nanomaterials
are incorporated into a matrix are equally important to understand (Nowack et al.. 2012). Nowack et al.
(2012) also note that a key research question is whether different modifications to nanomaterials and
incorporation into different product matrices will have increased or decreased reactivity or toxicity
relative to their pristine counterparts. Also, available data on the emissions of nanomaterials from
products and releases to the environment are scarce, particularly under realistic conditions (Nowack et al..
2012). This lack of data is due in part to the absence of robust analytical techniques and instrumentation
for accurately detecting and quantifying both emissions and environmental concentrations of
nanomaterials (Nowack et al.. 2012) (see Additional Information Highlight Box 10).
In each product life cycle stage in the CEA framework, "volume" and "release rate" are risk
relevance factors that might be considered in risk assessments and risk management efforts of a material.
Although these might be considered separately during future assessment and risk management efforts
they are discussed together for each area below. In general, workshop participants rated release rate more
highly than volume in terms of importance to consider in future risk assessments of MWCNTs. Experts
noted during the workshop that this is partially due to the fact that although production volume is
important for understanding the potential scale of impact (see Table 2-2 for information on current scale
and projected growth of MWCNTs production), release rate is what will ultimately determine the extent
of exposure. In other words, a large production volume might not cause concern if release rate is
relatively small, but a smaller production volume could still cause concern if release rate is particularly
high. Given that very little information is available for either topic and that each product life cycle area
has such a large influence on all other CEA areas, volume and release rate are particularly important to
consider in future risk assessment and risk management of MWCNTs.
6.3.1.1. Material Synthesis: Volume and Release Rate
The workshop participants identified MWCNT material synthesis as important to risk assessment.
Risk relevance factors that might be considered in future assessment or risk management efforts for this
6-9
-------
or other stages of the product life cycle include volume and release rate. Participants most commonly
identified both of these factors as important to risk assessment. Participants were generally not confident
or only somewhat confident in the ability of data on each factor to support risk management decisions
(see Priority Research Area Highlight Box in Section 2.2.2). Below are examples of experts' rationale for
rating these areas as research priorities.
• Material synthesis—Volume: Reporting from material synthesis is voluntary at this
point, so little is known about MWCNT synthesis volume.
• Material synthesis—Release rate: No universal reporting mechanism is available, so
little is known about MWCNT release rates. In addition, there is uncertainty about
manufacturing factors that could have ecological consequences (e.g., surfactants, not just
active ingredients, should be tested to make experiments more realistic to real-world
scenarios.)
The overarching influential factors identified by multiple experts for both volume and release rate
characteristics included a variety of methods and techniques (e.g., control technologies, personal
protective equipment, MWCNT synthesis and purity); engineered nanomaterial (ENM) characteristics;
factors associated with the surrounding media; and physical (e.g., flow regime, temperature, wind),
chemical, and social conditions (e.g., acute, subchronic, and chronic exposure).
As discussed in Section 2.2.1. chemical vapor deposition, arc discharge, and laser ablation are
three commonly used methods for mass production of CNTs. Yet, one newly identified reference points
out that progress has been made on "controllable growth" CNT synthesis methods (Huang etal.. 2012).
These controlled synthesis methods are used to develop CNTs with specific physicochemical properties
(e.g., defined wall number and diameter, high-purity); however, they are not yet at the point of large-scale
application (Huang et al.. 2012). No studies were identified regarding the volume of MWCNTs produced
for use in flame-retardant textiles or the release rate during synthesis of MWCNTs for flame-retardant
formulations (see Table 2-2 for general information on production volume).
No studies were identified in reviews of the draft document regarding the volume of MWCNTs
produced for use in flame-retardant textiles or the release rate during synthesis of MWCNTs for flame-
retardant formulations (see Table 2-2 and Additional Information Highlight Box 2 for general information
on production volume).
As noted in Section 2.2.2.2. several potential release scenarios during material synthesis are
plausible, including recovery of the synthesized substance, handling/packaging, equipment cleaning, and
accidental release (e.g., fugitive leaks, equipment malfunction, malfunctioning ventilation systems,
exposure to fire and heat). In particular, during handling and mixing of raw materials, CNTs might
become airborne under conditions where powder is being handled, weighed, or mixed, although this will
largely depend on the synthesis methods and use of control technologies (SAFENANO. 2012): see
6-10
-------
Sections 2.2.2.2 and 6.3.3.1). In one study for example, Dahm et al. (2011 a) found that some facilities
typically handling large quantities (40 grams to 1 kg of MWCNTs handled per day in powder form and
aqueous form) still had measured releases above the NIOSH recommended exposure limit (REL) of
7 ug/m3 of concern, despite the use of enclosed processes or other forms of control technologies (see
Section 4.1.2.4). The authors note, however, that their measurements were collected as inhalable mass
concentration of elemental carbon while the NIOSH recommended exposure limit is based on the
respirable mass concentration of elemental carbon, suggesting that measurements at primary facilities
might not be appropriately characterized in reference to this standard (see Section 6.3.3.1 for information
regarding occupational exposure and use of control technologies).
Although time did not allow for expert stakeholders participating in the CEA collective judgment
workshop on MWCNTs to develop specific research questions for material synthesis, the existing
information described above and in Chapter 2 point to several research areas that could support future risk
assessment and risk management efforts, including:
• What is the median volume of MWCNTs produced at manufacturing facilities?
• What volume of metal catalysts and support materials (e.g., aluminum, silica) are required for
manufacturing MWCNTs in current production facilities?
• What volume of water is used by MWCNT production facilities to clean equipment?
• Does release rate vary during MWCNT synthesis using chemical vapor deposition, fluidized bed
chemical vapor deposition, arc discharge or other methods?
• Are metal catalyst or support materials released during MWCNT synthesis?
6.3.1.2. Material Processing: Volume and Release Rate
The workshop participants identified MWCNT material processing as important to risk
assessment, and subsequently processing volume and release rate were both most commonly identified as
important risk relevance factors. Participants generally were not confident or only somewhat confident in
the ability of data on each factor to support risk management decisions (see Priority Research Area
Highlight Box in Section 2.2.3). Below are examples of experts' rationale for rating these areas as
research priorities.
• Material processing—Volume: While processing volume information may be available
to manufacturers and EPA regulators, and volume figures for decaBDE could be used to
calculate potential worst case release scenarios, little information is available for
scientists and the general public. Depending on the application of the product, there is
great opportunity for exposure.
6-11
-------
• Material processing—Release rate: There is not enough data on releases from material
processing, yet it is critical to determine release rate and exposure. Fabric coating
operations, for example, are resulting in releases to the environment, so there is a need to
understand the form of release (e.g., matrix bound, aggregate, etc.) and develop better
analytical methods for carbon nanotube quantification.
The overarching influential factors identified by multiple experts for both material processing volume and
release rate characteristics included: methods and techniques (e.g., control technologies and MWCNT
purity); ENM characteristics, factors associated with the surrounding media (e.g., air, sediment, soil,
wastewater); and physical, chemical, biological, and social conditions.
No additional data were identified for this area in revising the draft document; to extend on
available data for MWCNT material processing, however, research planning efforts might consider input
from the CEA collective judgment workshop. Expert stakeholders discussed release rate during material
processing to identify potential risk scenarios that might occur during this stage of the product life cycle
along with specific research questions that, if pursued, might inform future assessment and risk
management efforts in this area (RTI International 2012). Based on those discussions, potential risk
scenarios include: (1) release into the air that results in worker exposure and (2) MWCNT release into
sewage treatment plants that results in environmental exposures. To mitigate or avoid these potential risk
scenarios, experts noted that risk management decisions could include the use of controls to minimize
MWCNT concentrations in occupational air or wastewater. Experts noted that quantifying release rates in
these scenarios could inform the selection of appropriate control technologies. This type of quantification
might be carried out as part of an environmental or occupational assessment, according to experts. To
support such measurements, the research questions listed in Table 6-4 were developed by expert
participants along with an estimate of the resources and time to carry out the research.
6-12
-------
Table 6-4. Research identified by RTI workshop participants: MWCNT material processing.
Estimated Estimated Time
Research Finances ($) Frame
What is occupational exposure at current MWCNT processing facilities?
[Evaluate exposure to workers at manufacturing facilities to identify key steps in 100000 1 year
exposure. Evaluate control or personal protective equipment measures for reducing '
exposure].3
What is release rate in waste water from current MWCNT processing facilities?
[Evaluate release rates based on different technologies for: synthesis, filtration, 100,000 1 year
remediation.!
Develop Method (instrument) to characterize and quantify in waste liquid for
monitoring.
What is the best method to capture/destroy CNT in waste liquid?
Can CNT synthesis techniques reduce potential releases through control of
initial raw CNT form?
What air handling technologies can be used to reduce occupational exposure?
500,000
300,000
300,000
200,000
5 years
3 years
3 years
2 years
aNote: Information in brackets [ ] denotes details extracted from responses experts provided along with the question or
research area.
6.3.1.3. Product Manufacturing: Volume and Release Rate
The workshop participants identified MWCNT product manufacturing as important to risk
assessment, and subsequently volume and release rate at this stage were most commonly identified as
important risk relevance factors. Participants were generally not confident or only somewhat confident
that data could support risk management decisions related to either volume or release rate at this stage of
the product life cycle (see Priority Research Area Highlight Box in Section 2.2.4). Below are examples of
experts' rationale28 for rating these areas as research priorities.
• Product manufacturing—Volume: No universal reporting mechanism is currently available
to capture product manufacturing volume of MWCNTs, thus the magnitude for potential for
release is not known (we currently assume decaBDE-product volumes to calculate worst case
scenarios). While confidential business information may be known to manufacturers and
available to EPA regulators, little information is available for scientists and the general
public. Development of methods to quantify and detect ENMs is needed.
Note that rationales are taken directly from participants' responses in the RTI workshop process.
6-13
-------
• Product manufacturing—Release rate: There has not been enough study of manufacturing
release rates, which are critical for determining exposure. Belter analytical methods are
needed for carbon nanotube quantification. There is a disconnect between EHS research and
real-world ENM toxicity, characterizations, etc.
The overarching influential factors identified by multiple experts for both product manufacturing volume
and manufacturing release rate characteristics included: methods techniques (e.g., processing methods,
personal protective equipment); ENM characteristics; factors associated with the surrounding media; and
physical, chemical, biological and social conditions.
Of all the stages in the product life cycle, workshop participants placed the most importance on
product manufacturing, with 12 of 13 participants (92 percent) identifying it as being important to
consider in future risk assessments of MWCNTs. No additional studies were identified regarding the
volume of MWCNT flame-retardant textiles produced, although Additional Information Highlight Box 2
notes that the scale is likely relatively small compared to other MWCNT products (e.g., electronics).
Two new sources were identified on CNT release during product manufacturing of textiles in
general. SAFENANO (2012) noted that when fibers are coated with CNTs, a suspension containing
CNTs is likely to be applied to a textile either through dipping or spraying. Dipping activities would be
less likely to result in airborne CNTs and therefore have less potential for exposure than spray
applications. During spinning of fibers or weaving, SAFENANO (2012) noted that CNTs could be
released due to the high-energy mechanical abrasion of the filaments as they are drawn, spun, and wound,
but that released CNTs are not likely to be "free" in this scenario. This premise is also supported by the
results of Takaya et al. (2012). Yarn coated with MWCNTs was released during the weaving process
without evidence of free MWCNTs in the respirable dust (see Section 6.3.3.1 for more information
regarding occupational exposure and use of control technologies in the industry).
To build on available data for product manufacturing of MWCNT flame- retardant textiles,
research planning efforts might consider input from the CEA collective judgment workshop. Expert
stakeholders participating in the workshop discussed release rate during product manufacturing to identify
potential risk scenarios that might occur during this stage of the product life cycle along with specific
research questions that, if pursued, might inform future assessment and risk management efforts in this
area (RTI International. 2012). Similar to the material processing product life cycle stage, for product
manufacturing, workshop participants noted the possibility of release to air and subsequent worker
exposure, as well as MWCNT release to sewage treatment plants that could result in environmental
exposures (see Section 6.3.2.2 for discussion on the Priority Research Area "environmental transport,
transformation and fate of MWCNTs in wastewater"). The research questions related to these potential
6-14
-------
scenarios developed by expert participants are shown in Table 6-5. along with an estimate of the
resources and time to carry out the research.
Table 6-5. Research identified by RTI workshop participants: MWCNT product manufacturing.
Estimated Estimated Time
Research Finances ($) Frame
What is the step in manufacturing that presents most risk of release to the
environment?
[Evaluate potential release rates (e.g., grams/day) based on the manufacturing 100,000 1 year
technique(s) used for: functionalization, dispersion, coating. Evaluate how
manufacturing techniques influence released particle characteristics.]3
What is the step in manufacturing that presents most risk of occupational
exposure?
[Evaluate potential release rate (e.g., mass / m3, surface area/ m3, number/m3) to -| QQ QQQ -| vea|-
occupational air based on manufacturing technique during: functionalization,
dispersion, coating. Evaluate how manufacturing technique influences released particle
characteristics.]
How does MWCNT functionalization affect the filtration efficiency and size „__ __,, „
distribution? 100<000 2years
How does the dispersion technique affect the filtration efficiency and size . __ nm „
distribution? 100'000 2years
"Note: Information in brackets [ ] denotes details extracted from responses experts provided along with the question or research
area.
6.3.1.4. Use: Volume and Release Rate
The workshop participants identified MWCNT use as important to risk assessment, and
subsequently use volume and release rate were most commonly identified as important risk relevance
factors. Participants generally were not confident or only somewhat confident in the ability of data on
each factor to support risk management decisions (see Priority Research Area Highlight Box in Section
2.4). Below are examples of experts' rationale for rating these areas as research priorities.
• Use—Volume: No universal reporting mechanism is currently available to capture volume of
use of MWCNTs, thus the magnitude for potential for release is not known (we currently
assume decaBDE-product volumes to calculate worst case scenarios). Although confidential
business information may be known to manufacturers and available to EPA regulators, little
information is available for scientists and the general public.
• Use—Release rate: Better analytical methods are needed for carbon nanotube quantification,
particularly for measuring consumer exposure to releases from furniture and other products,
as well as quantifying and detecting ENMs in the environment.
6-15
-------
The overarching influential factors identified by multiple experts for both use volume and use release rate
characteristics included: methods and techniques (e.g., MWCNT processing), ENM characteristics (e.g.,
applied coatings, matrix bound vs. free, morphology, surface chemistry), factors associated with the
surrounding media, and chemical, biological and social conditions (e.g., chronic exposure, life stage).
No additional studies were identified regarding the volume of MWCNTs in flame-retardant
textiles in use, as this application is not widespread (see Additional Information Highlight Box 2). In the
absence of data, predictions must be made using known information about decaBDE flame-retardant
textiles and textiles in general. As discussed in Chapter 2, relatively little degradation of textiles
containing MWCNT flame-retardant coatings is expected during use; however, releases that do occur are
primarily expected to be caused by the degradation of the product matrix. For example, garments
typically have 10 percent weight loss over the course of the life cycle due to washing, ironing,
weathering, thermal degradation, and wear and tear (SAFENANO. 2012; Chaudhry et al.. 2009). Other
nonclothing textiles have been reported to lose between 5 percent and 20 percent of their weight during
normal use due to washing, exposure to heat, aging, and abrasion (GreBler et al., 2010). Very little
information is available regarding the likelihood of MWCNTs being released from textiles during normal
use, particularly in the specific application of flame-retardant textiles, but some evidence exists that
MWCNTs could become airborne after the textile is burned and the char residue is mechanically
disturbed (Uddin and Nyden. 201 la; Nydenet al.. 2010) or during washing (Goncalves et al., 2012).
Experts at the workshop noted that the release rate during product use was a particularly
important area for further investigation and in need of better analytical techniques due to its direct
influence on human exposure (RTI International. 2012). Similar to material processing and product
manufacturing, workshop participants also noted that potential risk might arise in this product life cycle
stage from release of MWCNTs in indoor air or the environment (e.g., wastewater treatment), resulting in
exposure to consumers (see Sections 6.3.3.2 and 6.3.2.2 for discussion on Priority Research Areas:
Consumer Exposure and Wastewater, respectively). For this product life cycle stage, they noted that
regulations pertaining to the type of matrix used with MWCNT flame retardants might be used to avoid or
mitigate potential risk. Based on the experts' input, human health and ecological risk assessments could
inform this type of risk management decision. Research questions that experts identified to support
conducting such assessments are listed in Table 6-6. along with estimates of the resources and time to
carry out the research.
6-16
-------
Table 6-6. Research identified by RTI workshop participants: MWCNT product use.
Estimated Estimated Time
Research Finances ($) Frame
How does particle functionalization and matrix affect aging and release to air?
((use accelerated weathering test), measure, quantify (number and 300,000 2-3 years
concentration), and characterize (size distribution) [MWCNTs]).
How does particle functionalization and matrix affect release in washing
MWCNT textile products? ([use mini washing machines], measure, quantify 300,000 2-3 years
[number and concentration], and characterize [size distribution] [MWCNTs]).
6.3.1.5. Disposal/ Recycling: Volume and Release Rate
The workshop participants identified MWCNT disposal and recycling as important to risk
assessment. Subsequently, the risk relevance factors of MWCNT disposal and recycling volume and
release rate were both most commonly identified as important. Participants generally were not confident
or only somewhat confident that the data on each factor could support risk management decisions (see
Priority Research Area Highlight Box in Section 2.5). Below are examples of experts' rationale for rating
areas as research priorities.
• Disposal/recycling—Volume: It is unclear to what extent products containing MWCNTs
will be recycled. This information is not available yet and may not be until the products are
on the market. The potential for release from disposal and recycling operations and processes
is unknown. Currently estimates are based on decaBDE data. There is currently no system
like a national registry, however this step will be less critical for exposure and risk
assessment than previous steps.
• Disposal/recycling—Release rate: Releases to air from incineration and to water from
landfills are possible. Better analytical methods are needed for MWCNT quantification.
Release rate will depend on specific recycling methods, which are still evolving, and there is
the potential for secondary products to emerge. Understanding release rate is critical to
estimating exposures. Large volumes of waste may accumulate and increase the potential for
a release event. Currently, no system like a national registry exists. However, this step will be
less critical for exposure/risk assessment than previous steps. Research is needed on how
companies can dispose of ENMs in an environmentally conscious way.
The overarching influential factors identified by multiple experts for both disposal and recycling
volume and use release rate characteristics included methods and techniques; ENM characteristics; factors
associated with the surrounding media (e.g., air, wastewater, surface water); and physical, chemical,
biological and social conditions.
6-17
-------
As noted above, one reason experts rated disposal and recycling as important is that very little
information is available about the release from textiles during these processes. Similarly, no information
is available on the by-products, metals, and other waste products that could result during these processes.
Workshop participants noted that MWCNTs could be released into the environment during disposal and
recycling, and that potential release could lead to exposures (e.g., release to wastewater that ends up in the
environment; release to air resulting in potential occupational exposure).
New information from additional literature or references already included in previous drafts of
the document suggests the following on recycling or disposal of textiles in general. During recycling,
textiles undergo various mechanical, thermal, and chemical treatments that could result in CNT release
from the product matrix (SAFENANO. 2012). Industrial textiles are often reclaimed and recycled in
specialized facilities while other types of textiles might be shredded and repurposed as a part of the
recycling process (e.g., polyester is often shredded, granulated into pellets, and processed to recover the
monomers and re-polymerized and processed by extruding, melting, spinning, etc., into new fibers)
(Chaudhry et al.. 2009). How MWCNT-treated flame-retardant upholstery would be recycled, and how
the MWCNTs might be released from the product matrix as a result of the recycling process, is unclear.
During disposal, textiles are expected to be either sent to a landfill or incinerated (SAFENANO.
2012; Chaudhry et al., 2009). Although no information regarding the current or predicted volume of
disposed MWCNT textiles was available, general quantities of disposed textiles were reported in
Chaudhry et al. (2009). For example, approximately 5.5 kg of textiles per person per year are disposed of
and burned in Switzerland; in the UK, 75 percent of "fashion textiles" are landfilled. This information
suggests that the likelihood of disposal of MWCNT flame-retardant textiles through landfilling or
incineration will vary by country, but could be relatively high. The likelihood of CNTs being released
from the product matrix during these processes partially depends on which disposal method is used. For
example, uncontrolled incineration could result in CNT release if the resulting char is mechanically
disturbed (SAFENANO. 2012: Uddin and Nvden. 201 la: Nydenetal.. 2010). whereas under controlled
conditions (>850°C), CNTs are likely to be destroyed and not released (SAFENANO. 2012; Nvden et al.,
2010). No data are currently available, however, to corroborate these predictions.
In building on existing literature discussed above and in Chapter 2. research planning efforts
might consider research questions identified by expert stakeholders participating in the CEA collective
judgment workshop for MWCNTs (RTI International. 2012). Based on their input, potential risk
scenarios for the disposal/recycling stage of the product life cycle might include (1) unknown or relatively
large volumes of materials at the end of the product life cycle (e.g., incineration, recycling, reuse, litter,
landfill disposal), (2) release during recycling for reuse (e.g., industrial shredding), resulting in release to
the environment (e.g., wastewater) and subsequent environmental exposure, (3) occupational exposure
6-18
-------
from MWCNTs released in air during recycling, and (4) release during sludge application to land that
results in exposure to humans or other biota on the surface. To mitigate or avoid these potential risk
scenarios, experts noted that risk managers might use a reclaim system to return upholstery and
upholstered products to manufacturers at the end of the product life, or limit MWCNT production and
use. Further, risk managers might need to evaluate the use of controls in occupational settings or
regulation of sludge waste disposal. Life cycle analysis, and occupational exposure and environmental
risk assessments, could inform these types of risk management decisions. Research questions that
participants identified to inform such assessments are listed in Table 6-7. along with estimates of the
resources and time to carry out the research.
Table 6-7. Research identified by RTI workshop participants: MWCNT disposal / recycling.
Estimated Finances Estimated
Research ($) Time Frame
How much volume of CNTs is used in upholstery?
How much volume of CNTs is lost from upholstery during life span? -mnnnn -i
iuu,uuu i year
How much volume of CNTs is lost via destruction (e.g., burning), recycling,
reuse, litter, or disposal in landfill?
What is the airborne release rate of MWCNTs during shredding (e.g., form, ._.-. _,,.-. .
size distribution, number, & mass concentration)? ' "
Survey the nanotechnology industry and municipal sewage treatment plants
to gather mass of sludge/year applied to land:
[Is sufficient sludge being applied that uptake from plants, or exposure to '
farmers is possible?]3
aNote: Information in brackets [ ] denotes details extracted from responses experts provided along with the question or research
area.
6.3.2. Environmental Transport, Transformation, & Fate
The sections below discuss areas within environmental transport, transformation, and fate that
were considered to be Priority Research Areas by workshop participants. Relevant to this priority area,
OECD (2012) noted that research needs to inform risk assessment include more robust data regarding
how nanomaterials move through different environmental and biological media, particularly in relation to
variation in physical-chemical properties (i.e., morphology, surface chemistry, size, functionalization).
OECD (2012) also identified mechanisms of bioaccumulation and predictive models for bioaccumulation
of nanomaterials as important to informing risk assessment, particularly because evidence suggests that
traditional relationships between octanol water partition coefficient (Kow) and bioaccumulation or
6-19
-------
bioconcentration factors might not be applicable to carbon nanotubes (OECD. 2012; Petersen et al.,
201 Ic).
6.3.2.1. Air: Mobility, Persistence, and Bioavailability
The workshop participants identified MWCNT air issues as important to risk assessment, and
subsequently the risk relevance factors of MWCNT mobility, persistence, and bioavailability were most
commonly identified as important. Participants were generally not confident or only somewhat confident
that data on each factor could support risk management decisions (see Priority Research Area Highlight
Box in Section 3.2). Below are examples of experts' rationale for rating these areas as research priorities.
• Mobility in air: Mobility in air is a primary route of exposure and existing data are
insufficient.
• Persistence in air: MWCNTs may persist in air, yet data are currently insufficient. More
data are needed on persistence and degradation of carbon nanotubes in air and the availability
of MWCNTs in air for inhalation by humans and ecological receptors. Indirect effects of
decaBDE need to be considered.29
• Bioavailability in air: While absorption across epithelial tissues has not been observed in
other organisms, data on the bioavailability of MWCNTs are currently insufficient.
Participants at the workshop who identified fate and transport in air to be important also
voluntarily listed influential factors to include in developing research plans for this area, such as:
analytical techniques; MWCNT characteristics (e.g., aggregation state, persistence, surface chemistry);
and a variety of physical and chemical conditions.
To date, very little information is available regarding the mobility, persistence, and bioavailability
of MWCNTs in air. From an ecological toxicology perspective, Petersen et al. (20 lie) note that, although
a few studies estimate the release of CNTs to air (and other environmental media) and the potential risks
to ecological receptors, such models are limited by the lack of analytical techniques needed to detect and
quantify CNTs in environmental matrices accurately.
The issue of MWCNT bioavailability and characterization in air also is relevant to human health,
as workshop participants noted. Laboratory studies have often been hindered by the inability to generate
aerosolized MWCNT particles. Ahn et al. (2011) have used heat and sonication prior to atomization to
generate untangled MWCNTs in aerosol without the use of surfactants. These untangled MWCNTs could
then be used in in vivo toxicity models, unlike those often produced in previous studies that also had
29Assumed to mean that indirect effects of decaBDE can inform research planning for MWCNTs.
6-20
-------
tangled or clumped structures. Although no data were identified on the mobility, persistence, or
bioavailability of MWCNTs in air under normal environmental conditions, these studies might provide
insight as to which physicochemical properties of MWCNT increase and decrease aerosolization
potential. Aersolization potential could in turn influence mobility, persistence, or bioavailability.
Information gaps persist, however, regarding the influence of environmentally relevant conditions on
aerosolization potential and other aspects of MWCNT behavior in air.
In extending upon existing data for mobility, persistence and bioavailability of MWCNTs in air,
research planning efforts might consider input from the CEA collective judgment workshop on
MWCNTs. Expert stakeholders participating in the workshop discussed persistence and mobility in air to
identify potential risk scenarios that might occur in this environmental spatial zone, along with specific
research questions that, if pursued, might inform future assessment and risk management efforts in this
area (RTI International 2012). Based on their input, potential risk might arise from the persistence of
CNTs released to occupational or ambient air resulting in longer residence times that increase the
probability of exposure. Similarly, mobility dictates the extent of potential MWCNT exposure in ambient
air, and thus can influence risk. To mitigate or avoid potential exposure in occupational or ambient air,
experts noted that risk mangers might reduce MWCNT residence time in air (e.g., by increasing
aggregation potential or decreasing mobility and retainment) or limit MWCNT production and use. To
inform such risk management decisions, experts suggested conducting human health risk assessments,
exposure assessments quantifying MWCNTs in air, and cost benefit analyses. To support such types of
assessments, expert stakeholders recommended the research areas listed in Table 6-8. along with
estimates of the resources and time to carry out the research.
6-21
-------
Table 6-8. Research identified by RTI workshop participants: MWCNT mobility & persistence in air.
Research
Develop a model to predict atmospheric residence time as a function of CNT
particle characteristics (QSAR).
Determine CNT properties and meteorological properties that increase
aggregation rate and decrease residence time.
Develop new methods or instruments to improve CNT quantification in air
(determine number or mass of CNT/m3).
Apply conventional Benefit/Cost Analysis Procedures [to determine the
economic consequences of limiting MWCNT product and use]3.
Develop model to predict extent of mobility as a function of CNT particle
characteristics (QSAR) for near-field and long-distance transport.
Alter CNT properties or meteorological properties to increase aggregation and
decrease mobility.
Estimated
Finances ($)
500, 000
>1 million
2 million
200,000
1 million
>1 million
Estimated Time
Frame
3 years
5 years
3 years
1 year
3 years
5 years
"Note: Information in brackets [ ] denotes details extracted from responses experts provided along with the question or research
area.
6.3.2.2. Wastewater: Mobility, Persistence, and Bioavailability
The workshop participants identified the environmental transport, transformation, and fate of
MWCNTs in wastewater as important to risk assessment. Subsequently, the risk relevance factors of
mobility, persistence, and bioavailability in wastewater were most commonly identified as important to
risk assessment. Participants were generally not confident or only somewhat confident that the data on
each factor could support risk management decisions (see Priority Research Area Highlight Box in
Section 3.3.3). Below are some examples of experts' rationale for rating these areas as research priorities.
• Mobility in wastewater: The extent to which MWCNTs move in wastewater determines
which environmental compartments will be exposed (e.g., water, soil, sediment). Wastewater
is the most likely route into the environment; more research is needed on how ENMs get out
of the wastewater stream and how to prevent this.
• Persistence in wastewater: MWCNTs might be persistent in wastewater and the potential
for MWCNT transformation and subsequent effects of transformation are unknown.
• Bioavailability in wastewater: MWCNTs are potentially persistent and likely to interact
with activated sludge given results from studies with other organisms.
The overarching influential factors identified by multiple experts for all three characteristics
included analytical techniques; control technologies; MWCNT purity; a variety of MWNCT
characteristics (including but not limited to aggregation/agglomeration state, applied coatings, size
6-22
-------
distribution, surface chemistry, water solubility/dispersibility); factors associated with the surrounding
media (particularly wastewater and sediment); chemical conditions (ionic strength, natural organic matter,
other contaminants in the environment, and salinity); and biological conditions (microbial communities in
the environment).
In the introduction to a recent special Environmental Science and Technology issue on
Transformations of Nanomaterials in the Environment, Plata et al. (2012a) outlined some of the
challenges involved in characterizing nanomaterials (in particular, CNTs) in water and sediment and some
promising analytical techniques that would help increase understanding of MWCNT mobility,
persistence, and bioavailability in wastewater. Some of the advances in analytical methods to detect
MWCNTs in aqueous systems, such as wastewater, are mentioned in Additional Information Highlight
Box 10. Research on analytical methods serves as a basis to begin to assess MWCNT mobility,
persistence, and bioavailability in wastewater (and in other environmental matrices).
Recent studies of MWCNT transport in water-saturated porous media might help inform
understanding of MWCNT mobility in wastewater. Wang et al. (2012) observed that 75 percent of
functionalized MWCNTs delivered through a water-saturated sand system were detected in effluent.
The authors also concluded that MWCNTs longer than 8 urn were more likely to be deposited than
shorter MWCNTs. Although wastewater systems are unique, they often involve the use of filters such as
sand. These results suggest that functionalized MWCNTs might be mobile during wastewater treatment
processes, with longer MWCNTs exhibiting less mobility than shorter MWCNTs.
With respect to the potential transformation of MWCNTs in wastewater, a recent paper by
Nowack et al. (2012) concluded that released CNTs could enter wastewater (and other environmental
compartments) where they could be transformed by photochemistry, oxidation, adsorption of natural
organic matter and other organic colloids, biotransformation, and continued abrasive forces (Nowack et
al.. 2012). The authors also noted that transformations could change CNT aggregation, dispersibility, and
interaction with biota in the environment (Nowack et al.. 2012). These conclusions highlight the
importance of studying not only MWCNTs, but also potential environmental transformations of
MWCNTs to understand the dynamics driving mobility, persistence, and bioavailability in wastewater.
To that end, expert stakeholders at the workshop discussed MWCNT persistence and mobility in
wastewater to identify the type of risks that might arise in this area of the environment and develop
specific research questions that could support future assessment and risk management efforts of the
material (RTI International 2012). They noted that MWCNTs could be released in either pulse industrial
discharges to sewers, or in semicontinuous loadings from industrial, commercial, and residential
wastewater with flame-retardant materials. To mitigate or avoid potential risks from these scenarios,
experts identified several types of decisions risk managers might consider: including using pretreatment
6-23
-------
controls to prevent MWCNT discharge from industrial facilities or regulating efficiencies of control
technologies to mitigate MWCNT release into the environment. Information to support making these
types of risk management decisions might include analytical measurements made during ecological risk
assessments or evaluations completed during life cycle analyses.
According to expert participants, research that could inform these and other types of assessments
in this area might include those listed in Table 6-9. along with resource and time estimates to carry out the
research.
Table 6-9. Research identified by RTI workshop participants: MWCNT mobility & persistence
in wastewater.
Estimated Estimated Time
Research Finances ($) Frame
How does the degree of functionalization and changes in wastewater treatment
processes (e.g., activated sludge, disinfection processes) affect the rate of
transformation?
400,000
3 years
[Evaluate the rate of transformation of MWCNTs alone and in a product matrix.]'
How to extract and characterize MWCNTs from suspended and fixed biomass or
treated effluent with minimal modifications to surface group, functionalization, 400,000 3 years
impregnated metals, and coatings.
What are the transformation byproducts from MWCNT and flame-resistant fibers?
To what extent does MWCNT surface properties and incorporation into fibers
affect distribution of MWCNTs between treated effluent and biosolids for different
wastewater treatment plant configurations?
[Evaluate using batch or OECD experiments using pilot tests with two or more MWCNT
materials. Distribution coefficients for some nanomaterials are available which could be
used with existing WWTP models to crudely predict MWCNT removals.]
Develop extraction and/or analytical techniques to quantify MWCNTs, of diverse
origin, at environmentally relevant levels in raw sewage, treated effluent and
biosolids.
[Detection limits of several methods exist, and may be relevant to apply given current
acute toxicity test results.]
300,000
250,000
600,000 with
pilot plant
300,000
600,000
3 years
2.5 years
4 years with
pilot plant
3 years
4 years for
new methods
aNote: Information in brackets [ ] denotes details extracted from responses experts provided along with the question or research
area.
6.3.2.3. Sediment: Mobility, Persistence, and Bioavailability
The workshop participants identified environmental transport, transformation, and fate of
MWCNTs in sediment as important to risk assessment. Subsequently, the risk relevance factors of
6-24
-------
mobility, persistence, and bioavailability in sediment were most commonly identified as important to risk
assessment. Participants were generally not confident or only somewhat confident that the data on each
factor could support risk management decisions (see Priority Research Area Highlight Box in Section
3.3.1). Below are some examples of experts' rationale for rating these areas as research priorities.
• Mobility in sediment: Sediment is a likely ultimate repository for MWCNTs and the extent
to which MWCNTs can be redistributed to the water column through sediment disruption is
unclear.
• Persistence in sediment: Degradation of MWCNTs is likely to be slow, if it occurs at all.
More research is needed on transformations in the environment and interactions between
microbes and soil.
• Bioavailability in sediment: MWCNTs are potentially persistent, however uptake was not
observed in several recent studies.
The overarching influential factors identified by multiple experts for all three characteristics of MWCNTs
in sediment included a variety of MWNCT characteristics (e.g., aggregation/agglomeration state and
surface chemistry) and chemical conditions (particularly natural organic matter).
In a recent review paper, Schwyzer et al. (2012) investigated the colloidal stabilities of 10
different MWCNT formulations under various environmental conditions (humic acid, natural organic
matter, calcium, etc.). Authors found that sedimentation dominates MWCNT behavior in the water
column, with most MWCNTs (8 of 10 MWCNT types tested) coming out of suspension after 5 days
(Schwyzer et al.. 2012). Yet, the authors also found that the presence of humic acid resulted in wide
variability in the fractions of MWCNTs in suspension (4-88 percent) (Schwyzer et al.. 2012). Both
oxygen content and the diameter of CNTs also influenced the ability of CNTs to stay in suspension
(Schwyzer et al.. 2012). Similarly, Petersen et al. (20lie) noted that MWCNTs sorb more readily to
sediments in seawater, but tend to stay in the water column in aquatic systems with high concentrations of
dissolved organic matter (DOM). The authors concluded that future studies of the subsurface mobility of
CNTs (including mobility in sediment) should examine a larger range of porous media size, mineralogy,
aqueous chemistry (including DOM), and natural soils (e.g., clays, silts, peats). The authors also stressed
that the influence of CNT functionalization and surface properties on transport are critical research areas.
The potential for MWCNTs to sorb or settle to sediments suggests that understanding MWCNT
behavior in sediment is important. Some workshop participants stated that MWCNTs are likely to persist
in sediment. Although previous studies indicated that carboxylated SWCNTs (but not pristine SWCNTs)
can be transformed by soil enzymes (Allen et al.. 2009; Allen et al.. 2008a), suggesting that similar
processes could occur in sediment, more updated and sediment-specific research should be conducted to
confirm that MWCNTs could biodegrade in sediment.
6-25
-------
Another aspect of MWCNT behavior in sediments is the potential to impact benthic organisms. In
a recent paper examining the effect of MWCNTs on bioaccumulation of poly cyclic aromatic
hydrocarbons (PAHs) by Chironomus plumosus larvae in sediment, Shen et al. (2012) concluded that
MWCNT-associated PAHs might have been absorbed by larvae; they hypothesized that CNTs could
increase the exposure risk of PAHs to benthic organisms due to their unique structure. These results
suggest that uptake of MWCNT by organisms could occur in sediment. Previous research, however, has
found no substantial uptake of CNTs by sediment-dwelling earthworms (Petersen et al.. 2008) or soil-
dwelling earthworms (Petersen et al.. 201 Ib: Petersen et al.. 2009). These differing conclusions support
the designation of MWCNT bioavailability in sediment as a research priority.
Although time did not allow for expert stakeholders at the workshop to discuss bioavailability in
sediment specifically, they did consider persistence in sediment (RTI International. 2012). Based on this
discussion, the relative persistence of MWCNTs could influence the potential risk associated with
instances when continuous deposition and burial of fiber or polymers containing MWCNTs from multiple
sources occur (e.g., air, wastewater discharge, release from products, storm water). To mitigate or avoid
potential risks influenced by the persistence of MWCNTs in sediment, experts noted that risk managers
might consider regulating the efficiencies of control technologies that can minimize MWCNT
concentrations in the environment, or regulating loadings of CNTs in fabrics. Although specific
assessments that could inform such risk management decisions were not identified by experts, such
information might include measurements of MWCNTs in environmental compartments in ecological risks
assessments and measurements to quantify release of MWCNTs from textile production. Experts
identified the research directions listed in Table 6-10. which might support conducting these and other
types of assessments, along with estimates of the resources and time to carry out the research.
Table 6-10. Research identified by RTI workshop participants: MWCNT persistence in sediment.
Research
Estimated
Finances ($)
Estimated Time
Frame
How does the degree of functionalization or changes in sediment affect the rate
of transformation?
QQQ CQQ QQQ o_c
' ' "
Is there long term persistence? Are MWCNT released from fabrics and
sediments?
[Evaluate both the rate of release from fibers as well as the transformation of MWCNT
alone and combined in a matrix.]3
How to extract and characterize MWCNT from sediment with minimal
modifications to surface group, functionalization, impregnated metals, and 400,000 3 years
coatings.
6-26
-------
Estimated Estimated Time
Research Finances ($) Frame
What are the transformation byproducts from MWCNT and flame-resistant
fibers?
.... ., , . ... t .. • . . t. . ,. t,. Not provided Not provided
What byproducts could be formed during degradation processes and would this
be impacted by the MWCNT concentration in the fabrics?
aNote: Information in brackets [ ] denotes details extracted from responses experts provided along with the question or research
area.
6.3.3. Exposure Route and Dose (Kinetics)
The following exposure routes and dose/kinetic topics were considered to be Priority Research
Areas according to workshop participants.
6.3.3.1. Exposure Route- Human Occupational: Ingestion, Inhalation, Dermal
Human occupational exposure was identified as important to risk assessment by the workshop
participants. Risk relevance factors that might be considered in risk assessment or management of
occupational exposure include ingestion, inhalation, and dermal routes of exposure. Although experts
strongly agreed that occupational exposure is important, the way they rated the importance of each risk
relevance factor varied. Overall, they were not confident or only somewhat confident that the current data
could support risk management decisions for each factor (see Priority Research Area Highlight Box in
Section 4.2.1). Below are some examples of why experts rated these areas with varying levels of
importance and confidence.
• Human occupational exposure—Ingestion: Experts do not see much concern from
ingestion of MCs30 based on data; there will likely be good controls to limit exposure via
ingestion. From the exposure amount perspective, the relative amount of CNT intake in the
form of ingestion would be much lower than inhalation in occupational settings and there are
some studies suggesting luminal surfaces of GI tracts are resistant to the passage of CNTs.
• Human occupational exposure—Inhalation: Inhalation is thought to be a likely first type
and some say the most important type of human occupational exposure; inhalation exposure
must be controlled as it can cause effects, including indirect effects in children. Some animal
studies have been conducted but exposure routes and administration techniques have been
criticized.
30Word choice of expert; assumed to mean multiwalled carbon nanotubes.
6-27
-------
• Human occupational exposure—Dermal: Dermal exposure is less likely an issue based on
the known physicochemical properties of CNTs. A low dermal absorption rate is expected
and there would likely be controls to limit dermal exposure, however some studies in insects
have shown some systematic effects (however other routes, inhalation and ingestion, were not
completely blocked).
The overarching influential factors identified by multiple experts for human occupational
exposure characteristics included methods and techniques (e.g., processing and synthesis methods,
personal protective equipment); ENM characteristics (e.g., applied coatings, morphology, persistence);
factors associated with the surrounding media (e.g., air); and physical, chemical, biological, and social
conditions (e.g., chronic exposure, occupation).
As discussed in Section 4.2.1.4. MWCNTs are found in facilities ranging from research
laboratories and production plants to those where they are processed, used, disposed, and recycled. The
limited literature available suggests potential for worker exposure in at least some of these types of
facilities [(Dahm et al.. 201 la: Johnson et al.. 2010; Lee et al.. 2010a: Han et al.. 2008): see Sections
4.1.2.4. 4.2.1. and 4.2.5)1. The extent of worker exposure to MWCNTs, however, is not well understood.
No published literature was available on investigations of occupational exposure to MWCNTs in flame-
retardant textiles specifically, and no more recent literature regarding exposure potential was identified
after the CEA workshop process. OECD (2012) noted that research needed to inform risk assessment
includes obtaining robust data on the exposure of workers at all stages of the life cycle from material
synthesis to disposal and recycling facilities.
Although no information was identified specific to MWCNTs in flame-retardant textiles,
literature pertaining to similar applications is available. For example, Takaya et al. (2012) reported that
occupational workers could be exposed to respirable particles of yarn coated with MWCNTs during the
weaving process of the production of a conductive fabric. The authors concluded that the mechanical
force of weaving with the MWCNT-coated yarn was sufficient to break the coating and release yarn
fibers still embedded with MWCNT, but weaving is unlikely to provide sufficient mechanical action to
release individual MWCNTs from the coating layer [(Takaya et al.. 2012): see Additional Information
Highlight Box 41. Similar exposure scenarios can be envisioned with MWCNTs in flame-retardant
upholstery textiles, but how use of MWCNTs as coatings on upholstery textiles compared to yarns would
affect the release and subsequent exposure is unclear.
Several publications by Schubauer-Berigan et al. (2011) (see Table 2-2) and Dahm et al. (20lib:
201 la) examined the current industry of engineered carbonaceous nanomaterial to better characterize
potential occupational exposure and use of engineering controls. Although little information was provided
specific to MWCNT and no information was provided specific to flame-retardant textiles, the data
indicate that many companies use various engineering controls (including local exhaust ventilation, high-
6-28
-------
efficiency participate air filters, enclosed production processes, and safety cabinets) for production,
laboratory procedures, and research and development operations (Dahm et al., 201 Ib). Many companies
also use health and safety training, good hygiene practices, and other practices or administrative
procedures to reduce occupational exposure. Nearly one in four companies surveyed,however, either did
not report using respiratory protection or were using an ineffective form of respiratory protection (Dahm
et al.. 20lib). Similarly, one in seven companies not reporting respiratory protection stated that such
protection was not needed because operations were fully enclosed; however, the authors note that NIOSH
recently recommended the use of respirators even when processes are enclosed if measurement data
suggest that the nanomaterial release is not well controlled (Dahm et al.. 20 lib). Results of these studies
suggest that, although many companies use several forms of protective measures, measures can still be
improved to mitigate potential occupational exposure. Recommended exposure limits for different
formulations of MWCNTs and evaluation criteria for assessing the release possibilities of individual
MWCNTs are greatly needed. In 2010, NIOSH suggested an 8-hour recommended exposure limit of
7 ug/m3 for carbon nanotubes and nanofibers but that value has not yet been finalized (NIOSH. 2010).
Similarly, OECD is currently developing standard test methods for MWCNT, but this process is
complicated by the lack of an accepted "representative" MWCNT that could be broadly applicable to
other MWCNT formulations (Takavaet al.. 2012).
To develop a better understanding of human occupational exposures, research planning efforts
could incorporate input from the CEA collective judgment workshop on MWCNTs. Expert stakeholders
participating in the workshop discussed inhalation exposure in occupational settings to identify the type
of risks that might arise and to develop specific research questions that could support future assessment
and risk management efforts of the material (RTI International. 2012). Based on their discussion, risks
could arise due to inhalation of the material (pristine MWCNTs, functionalized MWCNTs, or otherwise
modified MWCNTs) during any part of the manufacturing process. To mitigate or avoid potential risk
associated with occupational inhalation exposures, experts noted that risk managers might consider using
engineering controls, personal protective equipment, or if necessary, banning the material outright.
Experts noted that information from assessments that might inform such risk management decisions
include a no-observed-effect level or lowest-observed-effect level for the relevant material and other data
from occupational exposure assessments. To conduct such assessments, experts recommended the
research areas, along with financial and time estimates, listed in Table 6-11.
6-29
-------
Table 6-11. Research identified by RTI workshop participants: MWCNT occupational exposure
via inhalation.
Research
Estimated
Finances ($)
Estimated Time Frame
Acute and chronic rodent bioassay studies after inhalation
exposure at relevant doses of well-characterized material.
3 years
2 million <2 year rodent bioassaV &
an additional year for setup
and analysis)
Analytical and rodent studies to examine effect of co-factors
(e.g., solvents, resins) on particles size, deposition, translocation,
and removal.
3 years
2 million <2 year rodent bioassaV &
an additional year for setup
and analysis)
6.3.3.2. Exposure Route- Human Consumer: Ingestion, Inhalation, Dermal
Human consumer exposure was identified as important to future MWCNT risk assessment by the
workshop participants. Risk relevance factors that might be considered in risk assessments or
management efforts in this area include ingestion, inhalation, and dermal routes of exposure. Participants
rated the importance of each of these risk relevance factors differently, but were generally not confident
or only somewhat confident that the data on each route could support risk management decisions (see
Priority Research Area Highlight Box in Section 4.2.2). Below are examples of why experts rated these
topics at varying levels of importance and confidence.
• Human consumer exposure—Ingestion: Currently there is not much concern from
ingestion based on data; however, more analytics are needed to determine risk of human
consumer exposure via ingestion.
• Human consumer exposure—Inhalation: The risk of human consumer exposure via
inhalation depends on release rate; if it is high enough it could become an issue. There is a
need to understand persistence and how upholstery exposure affects consumers.
• Human consumer exposure—Dermal: The risk of human consumer exposure via a dermal
route depends on dermal absorption rates. Need to understand potential exposure to children,
especially potential impacts from crawling around on floor.
The overarching influential factors identified by multiple experts for all three exposure routes included:
analytical techniques; control technologies; MWCNT purity; personal protective equipment; a variety of
MWNCT characteristics (e.g., aggregation/agglomeration state, applied coatings, persistence, size
distribution, surface chemistry, water solubility/dispersibility); factors associated with the surrounding
media (particularly air and wastewater); chemical conditions (dispersing agents, ionic strength, salinity,
6-30
-------
other contaminants in the environment, surfactants in a lab study); and social conditions (exposure
duration, human activity, life stage, susceptible populations).
Similarly, OECD (2012) noted that one research need to inform risk assessment is improved
characterization of consumer exposure, including the concentration in and release from consumer
products. Although the likelihood of MWCNT release from polymers, textiles, and other product matrices
during normal product use is not well understood, one additional study was identified by experts
reviewing the draft document. Nyden et al. (2010) provides some information on the potential release of
nanoparticulate additives during incineration of polyurethane foam, which is relevant in the context of
MWCNT use as flame-retardant additives because behavior during incineration is an important
consideration for potential consumer exposure. This study specifically investigated the release of carbon
nanofibers (CNFs), but identifies mechanisms of release that might be applicable to MWCNTs. In a
controlled, well-ventilated flame test, the authors found evidence of CNFs in the char on the surface of
the foam after burning but not in the smoke, indicating that CNFs were not released into the air during
this process. The researchers noted that CNF release to the environment still might be possible under
different combustion conditions (e.g., under-ventilated conditions that might decrease the destructive
forces of the flame on the CNFs). The authors also identified released submicron particles when the char
residue was mechanically disturbed, but could not definitively attribute the particles to the CNFs (Nyden
et al.. 2010). A later publication by the same authors suggested that the aerosolized particles from the
disturbed char were likely CNF bundles partially encapsulated in a thin layer of charred polyurethane
foam after spectroscopic measurements (Uddin and Nyden. 201 la). The authors concluded that the major
potential hazard for CNF exposure during well-ventilated combustion resulted from the disturbance of
residual char rather than from aerosolized CNF in smoke (Uddin and Nyden. 201 la).
Many of the available MWCNTs studies involve polymer matrices in which the MWCNTs are
embedded, but in this specific application of flame-retardant textiles, MWCNTs likely would be
incorporated through dipping, dyeing, thermal fixation, and other methods (as described in Chapter 1).
Nevertheless, data show that, in addition to the polymer matrix and the application technique, the
chemical properties of MWCNTs are important in determining the likelihood that MWCNTs leave the
product matrix during washing. Goncalves et al. (2012) investigated the extent to which acidic or basic
MWCNTs were "washed out" of polyester and cotton textiles. The results indicated that the more acidic
MWCNTs were less likely to be removed from the matrix, particularly in the polyester textile. Although
removal from the textile under realistic conditions (e.g., the washing machine) might not present much
opportunity for consumer exposure given that upholstery textiles are unlikely to washed frequently, these
data do suggest that acidic MWCNTs might be preferable to incorporate in flame-retardant textiles to
reduce potential release (see Section 2.4.2).
6-31
-------
Although expert stakeholders participating in the CEA collective judgment workshop did not
focus on consumer exposure due to time constraints, the results from recently identified literature do point
to several questions that currently are not addressed by the available literature. For example, if MWCNTs
can be removed from the textile matrix during washing, how frequently are MWCNT flame-retardant
textiles likely laundered or hand washed? Could consumers be dermally exposed to MWCNT particles
from handling the laundered textiles or be exposed to airborne MWCNTs after drying of the textiles?
How do differences in surface chemistry and textile type influence this exposure potential? How do
differences in cleaning practices (e.g., machine washing versus hand washing versus spot cleaning of
upholstery) influence potential exposure? To what extent do other activities (e.g., chewing on textiles)
lead to consumer exposures?
6.3.3.3. Dose (Kinetics)- Human: Absorption, Distribution, Metabolism, and
Excretion (ADME)
Human dose/kinetics was identified as important to future MWCNT risk assessment by workshop
participants. Factors that risk assessors or managers might consider in this area include absorption,
distribution, metabolism, and excretion. Expert workshop participants identified each factor as important,
with the exception of distribution, which they identified as possibly important. Participants were generally
not confident or only somewhat confident in the ability of data to support risk management decisions
related to each factor (see Priority Research Area Highlight Box in Section 4.2.6). Below are examples of
experts' rationale for rating these areas as research priorities.
• Dose (Kinetics)—Human Absorption: Except for the inhalation route, absorption needs to
occur to have an effect; need to know if ENMs are bioavailable. Concern is significantly
reduced if oral or dermal absorption does not occur after exposure; data are currently
insufficient.
• Dose (Kinetics)—Human Distribution: Distribution within tissues and specific organs
could be important to determine risk; data are currently insufficient.
• Dose (Kinetics)—Human Metabolism: Data are currently insufficient.
• Dose (Kinetics)—Human Excretion: Data are currently insufficient.
As noted above, at least one participant recognized that this area is a priority for research because
currently available data on all four components of human toxicokinetics are insufficient. Importantly,
concerns for risk through routes other than inhalation would be significantly reduced if oral and dermal
absorption were determined not to occur. Experts developing specific research questions related to
MWCNT absorption noted that although little evidence is available demonstrating dermal absorption (via
abraded skin), additional work should be considered due to the potential for high exposure, especially in
6-32
-------
children. Additionally, distribution within tissues and organs, in conjunction with data on mode of action
and toxicity, could play a key role in increasing or limiting risk.
Experts agreed that methods/techniques were influential to understanding toxicokinetics of
MWCNTs, including analytical techniques, MWCNT purification methods, processing and synthesis
methods, and control technologies and personal protective equipment. Several characteristics of the
MWCNT formulation also were deemed by multiple experts as influential factors for all toxicokinetic
components, including aggregation/agglomeration state, matrix bound vs. free form, persistence, size/size
distribution, and surface chemistry. Several experts also noted the importance of applied coatings,
morphology, and adsorption/desorption ability for adsorption and distribution, as well as the importance
of water solubility/dispersibility for distribution and elimination. Finally, several experts noted the
importance of social conditions on toxicokinetics: Exposure length (acute, subchronic, chronic), exposure
route, human activity, life stage, and individual susceptibility all can influence the human toxicokinetic
processes for MWCNTs.
Generally, inhaled nanomaterials—including CNTs—are assumed to be more toxic than
conventional-sized materials, in part because they can be inhaled more deeply into the lung (resulting in
longer residence times and greater particle-cell interactions), have a smaller size distribution (so are more
readily internalized by individual cells and more readily migrate through the body), and have large
surface area-to-mass ratios resulting in more reactivity (Bakand et al., 2012). Absorption, distribution,
metabolism, and excretion are key processes that determine how the administered or received dose of
MWCNTs differs from the internal dose that reaches a target organ or tissue. Toxicokinetic processes thus
play a key role in determining toxic potential.
Nevertheless, little is currently known about the specifics of absorption, distribution, metabolism,
and excretion of MWCNTs (see Section 4.2.6 and Additional Information Highlight Box 9). Wang et al.
(2013) reviewed several studies that elucidated adsorption and distribution (but not metabolism or
excretion) of carbon nanomaterials using isotopic labeling. More information specific to MWCNTs,
however, is still needed to understand the effects of morphology, dispersion state, and functionalization in
ADME. One proposed approach is to investigate carbon nanomaterials with different isotopes, labeling
the nanomaterial and its functional groups to trace the distribution and metabolism of both the parent
material and its derivatives (Wang etal. 2013).
In addition, pathways of CNT uptake into cells, intercellular trafficking, and distribution are not
well characterized, despite multiple studies on the subject (Al-Jamal et al., 2011). Toxicokinetics could
also vary with different MWCNT formulations or cell type. For example, Al-Jamal et al. (2011) recently
showed that MWCNTs functionalized with an ammonium group (NH3+) could be internalized by human
lung epithelial cells by three different mechanisms (membrane wrapping, direct membrane translocation,
6-33
-------
and clustering within vesicular compartment), whereas Zhang et al. (2012b) showed that uptake of
oxidized MWCNTs by human epithelial cervical cancer cells occurred through nonspecific cellular
uptake. Jain et al. (2011) found that biodistribution of acid-oxidized MWCNTs (which contain surface
carboxyl groups) in mice depended on the density of functionalization, and this physicochemical
characteristic particularly influenced clearance of MWCNTs from reticuloendothelial systems such as
liver, spleen, and lungs. Shorter, more oxidized MWCNTs (therefore with more functionalization) quickly
partitioned to the kidney but were rapidly excreted through the renal system; longer, less functionalized
MWCNTs and pristine MWCNTs preferentially accumulated in the liver rather than the kidney and were
more likely to be excreted in the feces through biliary pathways. The authors suggest that this
functionalization-dependent distribution between organs and excretion patterns might also explain why
pristine and less-oxidized MWCNTs do not demonstrate nephrotoxicity in subchronic studies.
A greater understanding of physicochemical properties influencing internal dose (i.e., particle
kinetics in biological systems) and therefore biopersistence and bioaccumulation also was noted by
OECD (2012) as particularly important research needed to inform risk assessment given that this
information is useful in interpreting toxicological results. Similarly, additional research on appropriate
dose metrics is needed to inform risk assessment based on evidence suggesting that particle surface area
or number concentration, rather than standard mass concentration, might be more appropriate for
nanomaterials (OECD. 2012).
To continue building on existing literature for human toxicokinetics summarized in Chapter 4. as
well as the newly identified sources discussed above, research planning efforts might consider input from
the CEA collective judgment workshop on MWCNTs. Experts participating in the workshop discussed
human absorption, metabolism, and excretion to identify potential risk scenarios and specific research
questions that could inform future assessment and risk management efforts (RTI International. 2012).
Based on their discussion, potential risks related to MWCNT toxicokinetics might include the potential
for absorption in the lungs or gastrointestinal tract, degradation into a more toxic metabolite, and
bioaccumulation of the material due to lack of excretion. To mitigate or avoid potential risks related to
MWCNT toxicokinetics in humans, experts noted that risk managers might consider implementing
appropriate control technologies to minimize exposures, minimizing the absorption potential of
MWCNTs, or limiting MWCNT production or use. Information to inform such risk management efforts
could come from human health risk assessment, occupational exposure assessment, and a cost-benefit
analysis. To support conducting such assessments, experts recommended the areas of research listed in
Table 6-12, along with estimates of the resources and time to carry out the research.
6-34
-------
Table 6-12. Research identified by RTI workshop participants: Human absorption, metabolism,
and excretion of MWCNTs.
Research
Estimated
Finances ($)
Estimated Time
Frame
Determine particle properties that influence extent and rate of absorption
across mammalian lung epithelial tissue, Gl luminal epithelia, and dermal
layers.
Quantify extent and rate of absorption across mammalian lung epithelial tissue,
Gl luminal epithelia, and dermal layers (if answer is yes, maximize particle
properties that decrease absorption while maintaining beneficial uses).
[Examining all three absorption processes increases the potential for discovery of
unique interactions among systems.]3
5 million
5 years
Develop analytical techniques for measuring the original MWCNT or
metabolites in cells.
[Evaluate the degradability of the relevant MWCNT material compared to original
MWCNT. Determine of the half-life of relevant MWCNT material in biological systems.]
275,000
2 years
Measuring the original MWCNT or metabolites in tissues after whole body
inhalation exposures.
500,000
2 years
Perform experiments in rodents after exposure to determine fate and clearance
of MWCNT.:
[Compare of the fate of relevant MWCNT material compared to original MWCNT in
rodents. Measure of the half-life of MWCNT material in biological systems.]
500,000
3 years
Develop tracer methodology to detect excretion by-products of the relevant
MWCNT material to enable:
[Quantify levels of by-products in the body.]
300,000
2 years
"Note: Information in brackets [ ] denotes details extracted from responses experts provided along with the question or research
area.
6.3.4. Impacts
The following impact areas were considered to be Priority Research Areas according to workshop
participants.
6.3.4.1. Human: Cancer and Noncancer
Experts in the workshop generally rated human health impacts as important to consider in future
risk assessments of MWCNTs. Participants most commonly identified human cancer as an important
impact and generally were somewhat confident in the availability and utility of current data to support
risk management decisions; however, those who identified human noncancer as an important impact were
generally not confident in the current data. Listed below are examples of experts' reasons for choosing
these topics as Priority Research Areas.
6-35
-------
• Human Cancer: Data are currently insufficient. There is a lack of cancer studies done on
inhalation exposure, particularly occupational inhalation exposure.
• Human Noncancer: Data are currently insufficient.
Those who chose to note influential factors generally agreed that MWCNT characteristics (e.g., surface
coatings, size, morphology) and social conditions (e.g., exposure duration, exposure route, occupation,
individual or population susceptibility) were of particular importance for future assessments of human
cancer and noncancer effects.
OECD (2012) similarly identified topics pertaining to human impacts as important for risk
assessment. These topics include the identification of toxicological endpoints specific to nanoparticles to
ensure that risk assessors identify all appropriate biological responses potentially leading to adverse
outcomes. Better characterization of mode of action in mammalian systems and interspecies variation also
were identified as important research to inform risk assessment.
Safe Work Australia recently commissioned the National Industrial Chemicals Notification and
Assessment Scheme (NICNAS) to conduct a health hazard assessment and hazard classification on
SWCNTs and MWCNTs (NICNAS, 2012). Information considered in this assessment included studies
published from January 2007 through June 2010. Classification of each material was based on two
existing systems: (1) the Globally Harmonized System of Classification and Labelling of Chemicals, and
(2) the Approved Criteria for Classifying Hazardous Substances. In October 2012, the NICNAS classified
MWCNTs as hazardous for repeated or prolonged inhalation exposure and carcinogenicity. In contrast,
MWCNTs were not classified as hazardous for the following endpoints: (1) acute oral toxicity, (2) acute
dermal toxicity, (3) specific target organ toxicity after a single exposure, (4) irritation to respiratory
system, (5) skin irritation and sensitization, (6) eye irritation. Data were insufficient to classify MWCNTs
for the following endpoints: (1) acute inhalation toxicity, (2) respiratory sensitization, (3) genotoxicity/
mutagenicity, (4) reproductive/developmental effects.
In addition to effects described in Section 5.1, MWCNTs have been shown to generate reactive
oxygen species, increase cell permeability in human microvascular endothelial cells (HMVEC), promote
cell migration in HMVEC (Pacurari et al., 2012). and cause inflammation (see Additional Information
Highlight Box 16). MWCNTs also have been shown to demonstrate subpleural deposition and pleural
translocation (Mercer et al.. 2010; Rvman-Rasmussen et al.. 2009a). which has important implications for
carcinogenicity. In particular, as noted in Additional Information Highlight Box 13. MWCNTs have
structural similarities to asbestos, raising concern over the potential for asbestos-like effects (e.g.,
mesotheliomas).
In a review of the available literature, Donaldson et al. (2010) concluded that long MWCNT
fibers are retained in the stomata of the parietal pleura, which is normally responsible for particle
6-36
-------
clearance; therefore, this site is where inflammation and pathogenic effects are manifested. To elucidate
the molecular mechanism of pathogenic action, Murphy et al. (2012) investigated the pro-inflammatory
response of mesothelial cells and macrophages. The results indicated that CNTs indirectly resulted in an
increase in cytokine release from mesothelial cells, as a result of frustrated phagocytosis of the
macrophages. The authors concluded that the response in the pleura is first initiated by the macrophages,
which in turn stimulate a pro-inflammatory response from the adjacent mesothelial cells (Murphy et al..
2012).
A reviewer identified a REACH dossier describing acute and subacute oral exposure studies and
an acute dermal exposure study that had not been described in the External or Peer Review Drafts of this
document (ECHA. 2013). Note that the data within the REACH dossier have not been peer reviewed, nor
have they been reviewed or verified by ECHA or any other authoritative body. Nevertheless, in the acute
oral exposure study, MWCNTs were administered via gavage to female Sprague-Dawley rats as a single
dose of 0.5, 1, 5, 50, or 100 mg/kg; observations were made 1 day later. The design of the subacute oral
exposure study was similar to that of the acute study except that doses of 0.05 or 0.5 mg/kg were
administered 5 days/week for 28 days. The results of both studies were similar. No changes in
hematology or serum biochemistry were observed. No gross pathological changes; no changes in liver,
kidney, or spleen weights; and no sign of toxicity to the stomach, the colon, the small intestine, the
spleen, the pancreas, or the kidneys were evident. The only reported effect was the presence of
inflammatory granulomatous changes in the interlobular space in the liver (acute study only) and in the
hepatic parenchyma (both acute and subacute studies) at all dose levels. Despite these histological
findings in the liver, no steatosis, cholestasis, or hepatocellular insufficiency was observed in either study.
As reported by the acute dermal exposure study described in the REACH dossier (ECHA. 2013). Wistar
rats demonstrated no clinical signs, no effects on body weight, and no gross pathological effects in 14
days after a 24-hour semi-occluded dermal exposure to 2,000 mg/kg MWCNTs.
To extend upon the existing research described above and in Chapter 5. research planning efforts
related to carcinogenic effects in humans could consider how data from existing studies inform additional
research on this topic. For instance, future risk assessments would benefit from 24-month carcinogenicity
studies that utilized a human-relevant route of exposure. The use of intracheal and intraperitoneal
injections do not replicate human-relevant routes of exposure. In addition, research planning efforts
related to non-carcinogenic effects might consider input from the CEA collective judgment workshop.
Experts in the CEA Workshop discussed that potential risks of noncancer effects result from chronic or
nonchronic exposures to coated or functionalized MWCNT in upholstery textiles. To mitigate or avoid
potential risks related to noncancer effects in humans, experts noted that risk managers might work to
develop exposure limits for workers or other subpopulations, and if necessary consider banning the
6-37
-------
material; however, assessments would be necessary to inform this or other types of risk management
decisions. Experts did not specify the types of assessments that would inform these decisions but they did
note that reference values should be identified for materials in consumer or occupational exposure
scenarios. Such information could be developed in a human health hazard assessment. To support this
type of assessment, experts recommended the research areas listed in Table 6-13. along with estimates of
the resources and time to carry out the research.
Table 6-13. Research identified by RTI workshop participants: Noncancer human health impacts
of MWCNTs.
Research
Conduct acute and chronic rodent bioassay studies after inhalation exposure at
relevant doses using well-characterized material.
Perform experiments to test impacts of exposure on immune compromised
individuals.
Estimated Estimated Time
Finances ($) Frame
3 years
(2 year rodent
2 million bi0af*ay&,
an additional
year for setup
and analysis)
1 million 3 years
6.3.4.2. Human: Reproductive/Developmental
The workshop participants who rated human health impacts as important most commonly rated
reproductive and developmental effects as possibly important. They were generally not confident in the
availability and utility of current data to support risk management decisions related to these effects (see
Priority Research Area Highlight Box in Section 5.1). Listed below are examples of experts' rationale for
rating this area as a research priority.
• Human Health Impacts—Reproductive/ Developmental: Data are currently insufficient.
There is concern about the long-term effects of ENMs due to large effects on reproduction
shown in animal populations.
Experts who chose to note influential factors commonly selected the following considerations to
take into account in planning research on human reproductive or developmental impacts: MWCNT
purity; applied coatings; persistence; surface chemistry; matrix bound versus free form; morphology;
exposure route; and life stage.
6-38
-------
Previous drafts of this document did not contain information regarding reproductive toxicity of
MWCNTs, but two studies investigating developmental toxicity in rodents were described (Fujitani et al..
2012; Lim etal. 20lib).
An expert involved in the collective judgment prioritization process identified one study of
MWCNT reproductive toxicity in mice (Bai et al.. 2010). and a targeted literature search identified a
study of developmental toxicity in rats (Lim etal.. 201 la). The findings presented by Lim et al. (2011 a)
appear to be from the same group of experiments described by Lim et al. (20 lib) and discussed in
Section 5.1.7; the conclusions of the two reports are the same: No differences in gestation index, fetal
death, fetal and placental weights, or sex ratio were observed as a result of maternal MWCNT exposure at
1,000 mg/kg-day. Thus, the study by Lim et al. (201 la) is not described in further detail here.
Bai et al. (2010) conducted a reproductive toxicity assay using intravenous injection of water-
soluble amine and carboxylate-functionalized MWCNTs in male mice (single injection or 5 doses over
13 days of 5 mg/kg). Results indicated that MWCNTs accumulated in the testes, generated oxidative
stress, and reduced the thickness of the seminiferous epithelium (authors reported that this damage was
reversible) without producing any significant effects on sperm parameters, sex hormones, fertility,
pregnancy rate, or delivery success of female mice mated with treated males (Bai etal.. 2010). Although
it used an exposure method that is not typically considered to be relevant, this study is useful in that it
indicates that if MWCNTs were absorbed into the bloodstream via inhalation or oral exposure they would
not likely cause male reproductive effects. Therefore, use of resources to investigate reproductive and
developmental toxicity further might be better allocated toward characterizing female reproductive
endpoints or developmental effects during various critical windows of development.
Due to time constraints, expert stakeholders participating in the CEA collective judgment
workshop did not discuss potential human reproductive or developmental effects; however, the limited
available data suggest several remaining research questions that might inform future risk assessment and
management efforts of MWCNTs. For instance, does gestational exposure to MWCNTs result in
behavioral or other subtle neurodevelopmental effects in offspring observed through adolescence? Does
comparing different formulations of MWCNTs within the same experiment indicate certain
physicochemical characteristics that influence reproductive or developmental effects?
6-39
-------
6.3.4.3. Aquatic Biota: Survival, Developmental, Reproductive, and Other
Sublethal
Workshop participants identified impacts on aquatic biota as important to future MWCNT risk
assessment. Risk relevance factors that might be considered in risk assessments and management efforts
for this element are survival, developmental, reproductive, and other sublethal effects. Although experets
varied somewhat in how they rated the importance of the "aquatic impacts" category and each risk
relevance factor, there was overall strong agreement among those who found aquatic impacts important
that they were not confident or only somewhat confident in the availability and utility of the current data
to support risk management decisions (see Priority Research Area Highlight Box in Section 5.1). Listed
below are examples of experts' rationale for why these topics were chosen as Priority Research Areas.
• Impacts to Aquatic Biota—Survival: Data does not indicate acute toxicity
• Impacts to Aquatic Biota—Developmental: As seen with endocrine disrupting chemicals,
aquatic species are very sensitive and low exposures can lead to developmental effects. A few
studies could go a long way toward understanding chronic effects.
• Impacts to Aquatic Biota—Reproductive: As seen with endocrine disrupting chemicals,
aquatic species are very sensitive and low exposures can lead to reproductive effects.
• Impacts to Aquatic Biota—Other Sublethal Effects: These effects are not typically studied
until environmental problems occur, need to give adequate attention.
The most commonly selected influential factors identified across all three areas included the
following: analytical techniques; a variety of MWCNT characteristics (e.g., adsorption/desorption ability,
aggregation/agglomeration state, lipophilicity, persistence, redox potential, surface chemistry); physical
and chemical conditions associated with the surrounding media (particularly ground water, sediment,
surface water, and wastewater); biological conditions (ADME, bioaccumulation, biomagnification,
microbial communities, organism health, developmental behavior, feeding behavior, reproductive
behavior); exposure route; habitat structure; and geographic location. The discussion below provides
more detailed information relating to aquatic toxicity, particularly in terms of aspects relevant to these
identified influential factors. This information was gathered from literature identified after the workshop.
OECD (2012) similarly noted that a better understanding of the absorption, distribution,
metabolism, and excretion of nanomaterials in ecological receptors is needed to evaluate the utility and
appropriateness of standard ecotoxicological assessments for nanomaterials. OECD (2012) also
recommended research on environmentally relevant conditions that could influence MWCNT toxicity.
As described in a recent review by Petersen et al. (20lie), the aquatic toxicity of CNTs has been
investigated in a variety of organisms including fish, algae, daphnia, copepods, amphibians (larvae),
protozoans, and bacteria. Toxicity depends on the bioavailability of MWCNTs to aquatic organisms (see
6-40
-------
Additional Information Highlight Box Gl). which is thought to be influenced by specific surface
chemistry and runctionalization; conflicting evidence, however, makes this relationship difficult to
elucidate. Petersen et al. (2010) reported that acid-treated MWCNTs, which were expected to be more
bioavailable than pure MWCNTs, did not actually accumulate at greater rates in the oligochaete
Lumbriculus variegatus. Similarly, different surface coatings and charges had no apparent impact on
accumulation or elimination of MWCNTs in Daphnia magna (Petersen et al.. 201 la), but a clear increase
in MWCNT toxicity to Ceriodaphnia dubia was observed with the addition of positively charged
functional groups (Kennedy et al.. 2009). Yet, an earlier study did show a dose- and time-dependent
increase in uptake of MWCNTs in a eukaryotic unicellular protozoan, Stylonchia mytilus (Zhu et al..
2006). After 5 days of exposure to 50 (ig/mL, authors report that S. mytilus redistributed micron-size
granules of MWCNTS into the water column. Authors suggests that the micron-size granules would
likely settle to the bottom of the water column, thus influencing distribution of MWCNTs in the water
column. Similar to other studies with aquatic invertebrates, the authors observed decreased viability and
growth in S. mytilus with increasing concentrations of MWCNTs greater than 1 (ig/mL (see
Section 5.2.1.1).
Another potentially important factor is the presence of metallic impurities. Mwangi et al. (2012)
found conflicting evidence: The removal of such impurities by acid pre-treatment decreased the lethal
effects of MWCNTs to mussels (Villosa iris), midges (Chironomus dilutus), and amphipods (Hyalella
aztecd) but did not mitigate biomass reduction for H. azteca, C. dilutus, and L. variegates, indicating that
metal impurities are responsible for some, but not all, of the effects observed (see Additional Information
Highlight Box 17). Abiotic factors, such as pH and natural organic matter (NOM), are also expected to
influence the bioavailability and toxicity of MWCNTs; however, Edgington et al. (2010) found no impact
on toxicity to D. magna with a range of dissolved organic carbon content in NOM (although some
variation was reported based on the source of NOM). Clearly, MWCNTs do not conform to classic
theoretical predictions of chemical behavior in aquatic media, which is further complicated by incomplete
information or conflicting data.
Another important factor in aquatic toxicity of MWCNTs is their potential interaction with other
contaminants. For example, Shen et al. (2012) showed that the presence of MWCNTs in sediment or soil
can complicate the toxicity of a system by altering the bioavailability of hydrophobic organic
contaminants like PAHs.
Available evidence indicates that MWCNTs are not absorbed in the gut. For example, Edgington
et al. (2010) observed MWCNTs in the gut of D. magna without evidence of nanotube uptake into the
microvilli. Similarly, several studies conducted by Mouchet et al. (2011; 2010; 2008) showed double-
walled carbon nanotubes (DWCNT) in the gut lumen but not in the blood, liver, or interstitial cells of
6-41
-------
Xenopus laevis larvae, and there was no evidence that DWCNT passed across the intestinal barrier.
Because current evidence shows a lack of CNT absorption across epithelial membranes, any observed
toxic effects are likely due to interactions with epithelial surfaces. For example, Oncorhynchus mykiss
exposed to SWCNTs showed signs consistent with impaired gill function (e.g., dilation of blood vessels
in the brain, other signs of oxidative stress), suggesting that SWCNT accumulation on the gill surface
results in impaired respiration (Petersen et al.. 201 Ic: Smith et al.. 2007). Indeed, authors observed
elevated mucous secretion with SWCNT deposits associated with mucoproteins in the gills. Although
SWCNTs were observed in the gut of exposed fish, likely as a result of drinking water with SWCNTs, no
histological changes were observed, with the exception of increased lipid peroxidation at one time point
in the 6-week experiment (Tetersen et al.. 201 Ic: Smith et al.. 2007). Similarly, Mwangi et al. (2012)
found no evidence of MWCNT penetration through cell membranes in C. dilutus and//, azteca.
Nevertheless, authors observed MWCNT accumulation in the guts of these organisms, which they
speculated, along with lack of depuration after transfer into clean water, contributed to decreased survival
and biomass (Mwangi et al., 2012). Evidence supporting the possibility of MWCNT accumulation in the
guts of aquatic organisms is found in studies in water fleas showing that elimination of MWCNTs from
the guts of C dubia (Kennedy et al., 2008) and D. magna (Petersen et al., 201 la; Petersen et al.. 201 Ic)
was possible only with the addition of algae as a food source. Edgington et al. (2010) also attributed the
toxicity of MWCNTs in D. magna to blockage in the gut. MWCNT aggregation behavior also has been
shown to be an influential factor in toxicity as evidenced by increased lethality in C. dubia with greater
aggregation (Kennedy et al., 2009; Kennedy et al., 2008). which also might support the influence of
digestive tract blockage. Together, these studies demonstrate the potential of MWCNTs to have toxic
impacts on growth and survival of benthic invertebrates and other aquatic biota; however, the mechanism
of that toxicity and influence of environmental factors remains unclear.
Due to time constraints, expert stakeholders participating in the CEA collective judgment
workshop did not discuss potential impacts in aquatic biota; however, based on the literature described
above, remaining questions related to potential impacts in aquatic receptors include the following:
• What modifications to MWCNTs might decrease interaction of the material with epithelial
surfaces in aquatic biota?
• Are there population level effects in aquatic biota?
6.3.4.4. Other: Economic, Societal, and Environmental Resources
Workshop participants identified "other impacts" as important to future MWCNT risk
assessment. Economic, societal, and environmental resource considerations are included as risk relevance
factors within this element of the CEA framework. Although experts varied somewhat in how they rated
6-42
-------
the importance of the "other impacts" category and each risk relevance factor, they generally strongly
agreed that other impacts is important and they were "not confident" that the current data could support
risk management decisions. Listed below are examples of why experts chose these topics as Priority
Research Areas.
• Other Impacts—Economic: There may be consequences of nanotechnology that are not yet
known.
• Other Impacts—Societal: There may be consequences of nanotechnology that are not yet
known.
• Other Impacts—Environmental Resources: There may be consequences of
nanotechnology that are not yet known.
Although the influential factors identified within each area (i.e., economic, societal,
environmental) varied somewhat, the influential factors identified for all three areas overlapped greatly,
including analytical techniques, control technologies, geographic location, human activity, occupation,
subchronic exposure, and susceptible populations. Additional influential factors selected for at least one
but not all three areas included acute exposure, chronic exposure, and life stage.
Newly identified information on potential economic, societal, and environmental impacts reveals
several ongoing efforts to balance societal needs with potential risks. For instance, according to Safe
Work Australia's Human Health Hazard Assessment and Classification of Carbon Nanotubes (NICNAS.
2012). carbon nanotubes have garnered much attention in recent years due to unique physical and
chemical properties that show promise for a wide variety of advanced applications across many diverse
fields. Unique challenges to risk communicators result from the uncertainty surrounding these new
applications and the implication that the rapid innovation has for developing a wide variety of MWCNT
formulations. Risk assessors and risk communicators need to strike a difficult balance between mitigating
potential risk and unintended consequences of novel technology (Priest 2012; Siegrist et al., 2011)
without stifling much needed scientific innovation to meet the demands of a growing global economy, the
global population, and the ever-shrinking availability of natural resources (Klaine et al., 2012; OECD.
2009). Nanotechnology has been praised for its potential to offer solutions for many of today's
environmental concerns, including pollution, drinking water filtration, climate change, and energy
efficiency, while also fueling economic growth by promoting new technologically advanced industries.
As a rapidly growing and evolving field, however, much uncertainty exists regarding potential trade-offs
and unintended consequences (TPEN. 2012; Siegrist etal.. 2011; OECD. 2009; Sass. 2007). For this
reason, emphasis has been placed on interdisciplinary participation and collaboration, including
government, stakeholders, researchers, academics, and the public, throughout all stages of the risk
assessment process to understand fully and effectively communicate potential risks while continuing to
6-43
-------
invest in technologically advanced solutions to everyday problems and to manage public perception
(Klaineetal. 2012; Siegristetal.. 2011; OECD. 2009; Sass. 2007).
The OECD Working Party on Manufactured Nanomaterials (OECD. 2012) noted that the ability
to evaluate the economic impacts of nanotechnology depends on the development of valuation models
specific to nanotechnology. Two valuation models described in the OECD report are the Defra model,
which is based on a comparative valuation of an existing product and a nanoenabled "replacement," and
the STAR METRICS approach, which uses an input/output method to analyze data between industries
(OECD. 2012). The Defra model offers some useful insight for this particular case study because a non-
nanoenabled product, decaBDE, has already been identified for comparison. Although no information
was identified regarding the energy and resource demands for decaBDE, some economic assumptions
could be made for MWCNT in flame-retardant textiles based on how decaBDE diffused and performed in
the market. For example, as noted in Table 1-10. global demand for decaBDE was quite high between
2001 and 2007, driven in part by its use in textile applications for flame retardancy. Because flammability
performance and ability to meet stringent regulations is important for a variety of industries with textile
applications (see Section 1.2.1). the possibility exists for the demand for MWCNT in this application to
increase as decaBDE or other flame retardants are phased out of use (see Additional Information
Highlight Box 2). Demand could increase, particularly when MWCNTs are used in combination with
other flame-retardant materials (see Additional Information Highlight Box 3). As discussed in Chapter 1.
considerations related to efficacy and production scale remain to be seen for MWCNT flame-retardant
coatings (see Additional Information Highlight Box 1 and Additional Information Highlight Box 2).
Although two valuation models are available to begin to inform economic impacts of MWCNTs, they are
limited because the development and commercial success of products containing MWCNTs face unique
challenges compared to non-nanoenabled products (OECD. 2012). For example, the research and
development stage for nanotechnologies often requires a very high level of investment and is
accompanied by a long lag period before any potential payoff is realized. With such a large number of
MWCNT-enabled products in this stage and relatively few products that have moved past it, estimating
when and how that payoff might take place is difficult. Similarly, additional risks are associated with
consumer perception or acceptance of nanoenabled products.
Regarding environmental impacts, according to a recent review article by Kim and Fthenakis
(2012). although manufacturing ENMs (e.g., MWCNTs) might require more energy than non-nanoscale
materials, the amount of ENMs used in products might be much lower than non-nanoscale materials,
which could result in lower energy requirements in the long term. As suggested in Chapter 5. Kim and
Fthenakis (2012) state that the potential impact of ENMs, such as MWCNTs, on energy use and
greenhouse gas emissions warrants further evaluation.
6-44
-------
Considerations raised in the literature on potential economic and environmental impacts suggest
several remaining research questions, including the following:
• What is the relative impact on environmental resources (e.g., water, energy) of MWCNT
production compared to other flame-retardant materials (non-nanoenabled or nanoenabled)?
• What data (e.g., production volumes, release rates across the product life cycle) on MWCNTs
in flame-retardant textiles can be incorporated into existing life cycle inventory databases?
What modifications to existing databases might inform research specific to MWCNTs (e.g.,
inclusion of physicochemical characteristics such as aspect ratio)?
• What modifications to existing tools used life cycle assessments (e.g., USEtox) would
improve efforts to evaluate environmental impacts from MWCNTs in flame-retardant textile
coatings?
• What are the economic implications of producing MWCNTs and MWCNT products (e.g.,
jobs created, infrastructure development)?
Workshop participants discussed societal impacts within this category to identify a potential risk
scenario and to develop research questions that could inform future assessment and risk management
efforts for MWCNTs (RTI International 2012). They noted that unintended consequences of MWCNTs
could arise, which might lead to concerns in government and industry, fear in the public, and
abandonment of future applications of the material. To manage or avoid these risks, experts suggested
that risk managers might consider efforts "to build capacity and enable informed consent," or ban the
material if necessary. They further suggested a socioeconomic assessment to inform such risk
management decisions. To support this type of assessment, experts recommended the research areas listed
in Table 6-14. along with estimates of the resources and time to carry out the research.
Table 6-14. Research identified by RTI workshop participants: Societal impacts of MWCNTs.
Estimated Estimated Time
Research Finances ($) Frame
Evaluate the capacity of the institutions to meaningfully engage the public on 50,000 0.5 year
nanotechnology.
Characterize the public's understanding of the benefits and risks of 150,000 1.5 years
nanotechnology and their potential for participating in decision-making.
Capacity building to improve understanding of benefits and risks of
nanotechnology.
Development of a more effective systems approach to examine interrelated ,-00 „„„ „
consequences (good and bad) of new technologies. ' years
Development of new methods of facilitating communication amongst
stakeholders on complex issues like nanotechnology.
6-45
-------
6.4. Moving From "Assessment" to "Management" in the
CEA Process
Compiling information in the CEA framework and conducting the subsequent collective
judgment and prioritization process represent the steps of the "assessment" phase of the CEA process (see
Figure 1-2). The "management" stage of the process involves moving the results of these steps into
research plans or risk management plans. The outcome of identifying and prioritizing information gaps in
the collective judgment step of CEA applied to MWCNTs is a list of topics in the CEA framework rated
on the basis of each topic's importance to risk assessment efforts and the level of confidence the experts
have in the information available for that topic to support risk management decisions. Areas collectively
deemed of high importance to risk assessment, but least understood based on available information, were
high priorities for developing specific research questions in the final stage of the prioritization process, a
face-to-face workshop. If these research questions are pursued, the knowledge gained could support a
variety of risk assessments and other analyses that then could be used to update the CEA framework, and
subsequently carry out a collective judgment prioritization of risk-related trade-offs to inform risk
management decisions.
Such research fits within the overall research paradigm at EPA, the principles of which include
sustainability, systems thinking, integrated transdisciplinary research, and relevant, responsive, and rapid
research (Anastas, 2012). Within this research paradigm, known as the "Path Forward," are six national
research programs (Anastas. 2012). Research specific to nanomaterials falls within the Chemical Safety
for Sustainability program. This Program is oriented toward evaluating and potentially addressing some
of the priority research areas identified through the CEA process for MWCNTs. Notably, given the
integrated, transdisciplinary nature of CEA, some of the identified research objectives might best be
addressed by other government agencies, academic institutions, or others in the scientific community. For
instance, the priorities identified in the CEA process for MWCNTs could inform several agencies (e.g.,
National Science Foundation, National Institute of Environmental Health Sciences) in the development of
funding opportunities (e.g., Science to Achieve Results [STAR] grants, requests for proposals) for
nanomaterial research. The research priorities identified in the CEA process are thus intended not only to
inform EPA research, but also to serve as a resource for the broader scientific community.
6-46
-------
References
ACC (American Chemistry Council). (2001a). Decabromodiphenyl ether: A prolonged sediment toxicity test
with Lumbriculus variegatus using spiked sediment with 2% total organic carbon. Wildlife International, Ltd.
ACC (American Chemistry Council). (200Ib). Decabromodiphenyl ether: A prolonged sediment toxicity test
with Lumbriculus variegatus using spiked sediment with 5% total organic carbon. Wildlife International, Ltd.
Adler. S: Basketter. D: Creton. S: Pelkonen. O: vanBenthem. J: Zuang. V: Andersen. KE: Angers-Loustau. A:
Aptula. A: Bal-Price. A: Benfenati. E: Bernauer. U: Bessems. J: Bois. FY: Boobis. A: Brandon. E: Bremer.
S: Broschard. T: Casati. S: Coecke. S: Corvi. R: Cronin. M: Daston. G: Dekant. W: Felter. S: Grignard. E:
Gundert-Remy. U: Heinonen. T: Kimber. I: Kleinjans. J: Komulainen. H: Kreiling. R: Kreysa. J: Leite. SB:
Loizou. G: Maxwell. G: Mazzatorta. P: Munn. S: Pfuhler. S: Phrakonkham. P: Piersma. A: Poth. A: Prieto. P:
Repetto. G: Rogiers. V: Schoeters. G: Schwarz. M: Serafimova. R: Tahti. H: Testai E: van Delft J: van
Loveren. H: Vinken. M: Worth. A: Zaldivar. JM. (2011). Alternative (non-animal) methods for cosmetics
testing: current status and future prospects-2010 [Review]. Arch Toxicol 85: 367-485.
http://dx.doi.org/10.1007/s00204-011-0693-2
Agrell C: ter Schure. AFH: Sveder. J: Bokenstrand. A: Larsson. P: Zegers. BN. (2004). Polybrominated
diphenyl ethers (PBDES) at a solid waste incineration plant I: Atmospheric concentrations. Atmos Environ
38: 5139-5148. http://dx.doi.0rg/10.1016/i.atmosenv.2004.05.024
Ahn. KH: Kim. SM: Yu. U. (2011). Multi-Walled Carbon Nanotube (MWCNT) Dispersion and Aerosolization
with Hot Water Atomization without Addition of Any Surfactant. 2: 65-69.
http://dx.doi.Org/10.5491/SHAW.2011.2.l.65
Al-Jamal. KT: Nerl. H: Miiller. KH: Ali-Boucetta. H: Li. S: Haynes. PD: Jinschek. JR: Prato. M: Bianco. A:
Kostarelos. K: Porter. AE. (2011). Cellular uptake mechanisms of functionalised multi-walled carbon
nanotubesby 3D electron tomography imaging. Nanoscale 3: 2627-2635.
http://dx.doi.org/10.1039/clnrl0080g
Alaee. M. (2003). An overview of commercially used brominated flame retardants, their applications, their use
patterns in different countries/regions and possible modes of release [Review]. Environ Int 29: 683-689.
http://dx.doi.org/10.1016/s0160-4120(03)00121-l
Alberding. MR: Maleckj HC: Shah. TK: Adcock. DJ. (2011). Flame-resistant composite materials and articles
containing carbon nanotube-infused fiber materials. (U.S. Patent Application No. 12/968,187). Washington,
DC: U.S. Patent and Trademark Office.
Alfaro-Moreno. E: Nawrot TS: Vanaudenaerde. BM: Hoylaerts. MF: Vanoirbeek. JA: Nemery. B: Hoet PH.
(2008). Co-cultures of multiple cell types mimic pulmonary cell communication in response to urban PM10.
EurRespirJ32: 1184-1194. http://dx.doi.org/10.1183/09031936.00044008
Alimohammadi F: Parvinzadeh. M: Shamei. A. (2011). Carbon nanotube embedded textiles. (U.S. Patent
Application No. 13/052,059). Washington, DC: U.S. Patent and Trademark Office.
Allen. BL: Kichambare. PD: Gou. P: Vlasova. II: Kapralov. AA: Konduru. N: Kagan. VE: Star. A. (2008a).
Biodegradation of single-walled carbon nanotubes through enzymatic catalysis. Nano Lett 8: 3899-3903.
http://dx.doi.org/10.1021/nl802315h
Allen. BL: Kotchev. GP: Chen. Y: Yanamala. NV: Klein-Seetharaman. J: Kagan. VE: Star. A. (2009).
Mechanistic investigations of horseradish peroxidase-catalyzed degradation of single-walled carbon
nanotubes. J Am Chem Soc 131: 17194-17205. http://dx.doi.org/10.1021/ia9083623
Allen. JG: McClean. MD: Stapleton. HM: Webster. TF. (2008b). Critical factors in assessing exposure to PBDEs
via house dust. Environ Int 34: 1085-1091. http://dx.doi.0rg/10.1016/i.envint.2008.03.006
R-1
-------
Anastas. PT. (2012). Fundamental changes to EPA's research enterprise: the path forward. Environ Sci Technol
46: 580-586. http://dx.doi.org/10.1021/es203881e
Aranberri. I: German. L: Matellanes. L: Suarez. MJ: Abascal E: Iturrondobeitia. M: Ballestero. J. (2011).
Investigation on flame retardancy and rheological and thermomechanical characterisation of multiwall
carbon nanotube reinforced nanocomposites. Plastics, Rubber and Composites 40: 133-138.
http://dx.doi.org/10.1179/1743289811xl2988633927998
Asakura. M: Sasaki. T: Sugiyama. T: Takava. M: Koda. S: Nagano. K: Arito. H: Fukushima. S. (2010).
Genotoxicity and cytotoxicity of multi-wall carbon nanotubes in cultured Chinese hamster lung cells in
comparison with chrysotile A fibers. J Occup Health 52: 155-166.
Aschberger. K: Johnston. HJ: Stone. V: Aitken. RJ: Hankin. SM: Peters. SA: Tran. CL: Christensen. FM. (2010).
Review of carbon nanotubes toxicity and expo sure Appraisal of human health risk assessment based on open
literature [Review]. Crit Rev Toxicol 40: 759790. http://dx.doi.org/10.3109/10408444.2010.506638
Aschberger. K: Micheletti. C: Sokull-Kluttgen. B: Christensen. FM. (2011). Analysis of currently available data
for characterising the risk of engineered nanomaterials to the environment and human health-lessons learned
from four case studies. Environ Int 37: 1143-1156. http://dx.doi.0rg/10.1016/i.envint.2011.02.005
Asharani. PV: Serina. NGB: Nurmawati. MH: Wu. YL: Gong. Z: Valiyaveettil S. (2008). Impact of multi-
walled carbon nanotubes on aquatic species. J Nanosci Nanotechnol 8: 3603-3609.
http://dx.doi.org/10.1166/inn.2008.432
ATSDR (Agency for Toxic Substances and Disease Registry). (2004). Toxicological profile for polybrominated
biphenyls and polybrominated diphenyl ethers [ATSDR Tox Profile]. Atlanta, GA: Agency for Toxic
Substances & Disease Registry, http://www.atsdr.cdc.gov/toxprofiles/tp68.pdf
Babrauskas. V: Krasny. J. (1985). Fire behavior of upholstered furniture. Gaithersburg, MD: National
Engineering Laboratory. http://fire.nist.gov/bfrlpubs/fire85/PDF/f85003.pdf
Baddour. CE: Briens. C. (2005). Carbon nanotube synthesis: A review. Int J Chem React Eng 3: 1-20.
http://dx.doi.org/10.2202/1542-6580.1279
Bai. Y: Zhang. Y: Zhang. J: Mu. Q: Zhang. W: Butch. ER: Snyder. SE: Yan. B. (2010). Repeated
administrations of carbon nanotubes in male mice cause reversible testis damage without affecting fertility.
Nat Nanotechnol 5: 683-689. http://dx.doi.org/10.1038/nnano.2010.153
Baitinger. EM: Vekesser. NA: Kovalev. IN: Sinitsyn. AA: Tsvgankov. IA: Ryabkov. YI: Viktorov. W. (2011).
Structure of multiwalled carbon nanotubes grown by chemical vapor deposition. Inorg Mater 47: 251-254.
http://dx.doi.org/10.1134/s0020168511030058
Bakand. S: Haves. A: Dechsakulthorn. F. (2012). Nanoparticles: a review of particle toxicology following
inhalation exposure. Inhal Toxicol 24.
Bartholmai. M: Schartel B. (2004). Layered silicate polymer nanocomposites: new approach or illusion for fire
retardancy? Investigations of the potentials and the tasks using a model system. Polym Advan Technol 15:
355-364. http://dx.doi.org/10.1002/pat.483
Bartlev. D: Feldman. R. (1998). NIOSH manual of analytical methods (NMAM), Fourth Edition (pp. 1-6).
(METHOD 0600). Bartley, D; Feldman, R.
Baskaran. D: Mays. JW: Bratcher. MS. (2004). Polymer-grafted multiwalled carbon nanotubes through surface-
initiated polymerization. Angew Chem Int Ed Engl 43: 2138-2142. http://dx.doi.org/10.1002/anie.200353329
Bello. D: Hart. J: Ahn. K: Hallock. M: Yamamoto. N: Garcia. E: Ellenbecker. M: Wardle. B. (2008). Particle
exposure levels during CVD growth and subsequent handling of vertically-aligned carbon nanotube films.
Carbon 46: 974-981.
Bello. D: Wardle. B: Yamamoto. N: deVilloria. R: Garcia. E: Hart. A: Ahn. K: Ellenbecker. M: Hallock. M.
(2009). Exposure to nanoscale particles and fibers during machining of hybrid advanced composites
containing carbon nanotubes. J Nanopart Res 11: 231-249. http://dx.doi.org/10.1007/sll051-008-9499-4
R-2
-------
Berger. M. (2007). Flame-retardant materials with more nanotechnology and less toxic chemicals. Honolulu, HI:
Nanowerk LLC. http://www.nanowerk.com/spotlight/spotid=2445.php
Beyer. G. (2004). Carbon nanotubes as a new class of flame retardants for polymers. In Proceedings of the 52nd
IWCS/Focus. Eatontown, NJ: International Wire and Cable Symposium.
Beyer. G. (2005). Filler blend of carbon nanotubes and organoclays with improved char as a new flame retardant
system for polymers and cable applications. Fire and Materials 29: 61-69. http://dx.doi.org/10.1002/fam.866
Beyer. G. (2006). Flame retardancy of nanocomposites based on organoclays and carbon nanotubes with
aluminium trihydrate. Polym Advan Technol 17: 218-225. http://dx.doi.org/10.1002/pat.696
Binetruv. C: Boussu. F. (2010). Recent advances in textile composites: Proceedings of the 10th International
Conference on Textile Composites. In C Binetruy; F Boussu (Eds.). Lancaster, PA: DEStech Publications,
Inc.
Birch. M. (2003). Diesel paniculate matter: As elemental carbon. In NIOSH Manual of Analytical Methods
(NMAM), Fourth Edition (pp. 1-5). (5040). Birch, M.
Bottini. M: Bruckner. S: Nika. K: Bottini. N: Bellucci S: Magrini. A: Bergamaschi. A: Mustelin. T. (2006).
Multi-walled carbon nanotubes induce T lymphocyte apoptosis. Toxicol Lett 160: 121-126.
BSEF (Bromine Science and Environmental Forum). (2012). Fire risk of upholstered sofas.
http://www.bsef.com/fire-safetv-benefits/consumer-safetv/upholstered-sofa
BSI (British Standards Institution). (2007). Nanotechnologies - Part 2: Guide to safe handling and disposal of
manufactured nanomaterials. (PD 6699-2:2007). London, UK.
Bundesministerium fur Frauenangelegenheiten. (1998). Health aspects of flame retardants in textiles. Germany:
Bundesministerium fur Frauenangelegenheiten und Verbraucherschutz.
http://www.verbraucherrat.at/download/flamehealth.pdf
CalRecvcle (California Department of Resources Recycling and Recovery). (2002). Examples of product reuse
and recycling. Sacramento, CA.
Cataldo. F. (2002). A study on the thermal stability to 1000C of various carbon allotropes and carbonaceous
matter both under nitrogen and in air. Fullerenes, Nanotubes, and Carbon Nanostructures 10: 293-311.
http://dx.doi.org/10.1081/FST-120016451
Cavallo. D: Fanizza. C: Ursini. CL: Casciardi. S: Paba. E: Ciervo. A: Fresegna. AM: Maiello. R: Marcelloni.
AM: Buresti G: Tombolini. F: Bellucci. S: lavicoli. S. (2012). Multi-walled carbon nanotubes induce
cytotoxicity and genotoxicity in human lung epithelial cells. J Appl Toxicol 32: 454-464.
http://dx.doi.org/10.1002/iat.2711
Cetin. B: Odabasi M. (2005). Measurement of Henry's law constants of seven polybrominated diphenyl ether
(PBDE) congeners as a function of temperature. Atmos Environ 39: 5273-5280.
http://dx.doi.0rg/10.1016/i.atmosenv.2005.05.029
Chaudhry. Q: Aitken. R: Hankin. S: Donaldson. K: Olsen. S: Boxall A: Kinloch. I: Friedrichs. S. (2009).
Nanolifecycle : A lifecycle assessment study of the route and extent of human exposure via inhalation for
commercially available products and applications containing carbon nanotubes. York, United Kingdom:
Food and Environment Research Agency.
Cheap Tubes Inc. (2009). Multi walled carbon nanotubes prices, http://www.cheaptubesinc.com/carbon-
nanotubes-prices.htm#Multi Walled Nanotubes Prices
Chen. S: Wu. G: Liu. Y: Long. D. (2006). Preparation of poly (aery lie acid) grafted multiwalled carbon
nanotubes by a two-step irradiation technique. Macromolecules 39: 330-334.
http://dx.doi.org/10.1021/ma0520500
Chen. W: Duan. L: Zhu. D. (2007). Adsorption of polar and nonpolar organic chemicals to carbon nanotubes.
Environ Sci Technol 41: 8295-8300. http://dx.doi.org/10.1021/es071230h
R-3
-------
Chen. Z: Yadghar. AM: Zhao. L: Mi. Z. (2011). A review of environmental effects and management of
nanomaterials. Toxicol Environ Chem 93: 1227-1250. http://dx.doi.org/10.1080/02772248.2011.580579
Cheng. J: Chan. CM: Veca. LM: Poon. WL: Chan. PK: Ou. L: Sun. YP: Cheng. SH. (2009). Acute and long-
term effects after single loading of functionalized multi-walled carbon nanotubes into zebrafish (Danio rerio).
Toxicol Appl Pharmacol 235: 216-225. http://dx.doi.0rg/10.1016/j.taap.2008.12.006
Cherukurl P: Gannon. CJ: Leeuw. TK: Schmidt. HK: Smallev. RE: Curlev. SA: Weisman. RB. (2006).
Mammalian pharmacokinetics of carbon nanotubes using intrinsic near-infrared fluorescence. PNAS 103:
18882-18886. http://dx.doi.org/10.1073/pnas.06Q9265103
Cho. HH: Smith. BA: Wnuk. JD: Fairbrother. DH: Ball WP. (2008). Influence of surface oxides on the
adsorption of naphthalene onto multiwalled carbon nanotubes. Environ Sci Technol 42: 2899-2905.
http://dx.doi.org/10.1021/es702363e
Christian. P: Von der Kammer. F: Baalousha. M: Hofmann. T. (2008). Nanoparticles: Structure, properties,
preparation and behaviour in environmental media [Review]. Ecotoxicology 17: 326-343.
http://dx.doi.org/10.1007/sl0646-008-0213-l
Ciparis. S: Hale. RC. (2005). Bioavailability of polybrominated diphenyl ether flame retardants inbiosolids and
spiked sediment to the aquatic Oligochaete, Lumbriculus variegatus. Environ Toxicol Chem 24: 916-925.
http://dx.doi.0rg/10.1897/04-179r.l
Cipiriano. BH: Kashiwagi. T: Raghavan. SR: Yang. Y: Grulke. EA: Yamamoto. K: Shields. JR: Douglas. JF.
(2007). Effects of aspect ratio of MWNT on the flammability properties of polymer nanocomposites.
Polymer (Guildf) 48: 6086-6096. http://dx.doi.0rg/10.1016/i.polvmer.2007.07.070
Coccini. T: Roda. E: Sarigiannis. DA: Mustarelli. P: Quartarone. E: Profumo. A: Manzo. L. (2010). Effects of
water-soluble functionalized multi-walled carbon nanotubes examined by different cytotoxicity methods in
human astrocyte D384 and lung A549 cells. Toxicology 269: 41-53.
http://dx.doi.0rg/10.1016/i.tox.2010.01.005
Cullen. E: O'Carroll. D: Yanful. EK: Sleep. B. (2010). Simulation of the subsurface mobility of carbon
nanoparticles at the field scale. Advances in Water Resources 33: 361-371.
http://dx.doi.0rg/10.1016/i.advwatres.2009.12.001
Cveticanin. J: Joksic. G: Leskovac. A: Petrovic. S: Sobot AV: Neskovic. O. (2010). Using carbon nanotubes to
induce micronuclei and double strand breaks of the DNA in human cells. Nanotechnology 21: 1-7.
http://dx.doi.Org/10.1088/0957-4484/21/l/015102
Dahlben. LJ: Isaacs. JA. (2009). Environmental assessment of manufacturing with carbon nanotubes. In A
Cozzi; T Ohji (Eds.), Environmental issues and waste management technologies in the materials and nuclear
industries XII (pp. 243-253). Hoboken, NJ: John Wiley and Sons.
Dahm. MM: Evans. DE: Schubauer-Berigan. MK: Birch. ME: Fernback. JE. (201 la). Occupational Exposure
Assessment in Carbon Nanotube and Nanofiber Primary and Secondary Manufacturers. Ann Occup Hyg -: 1-
15. http://dx.doi.org/10.1093/annhyg/merllO
Dahm. MM: Yencken. MS: Schubauer-Berigan. MK. (201 Ib). Exposure control strategies in the carbonaceous
nanomaterial industry. J Occup Environ Med 53: S68-S73.
http://dx.doi.org/10.1097/JOM.Ob013e31821bld3b
Davis. R: Kim. YS. (2010). Fabrication, characterization, and flammability testing of multiwalled carbon
nanotube layer-by-layer coated polyurethane foam. (NIST Technical Note 1676). Gaithersburg, MD:
National Institute of Standards and Technology.
Dawson. P: Duenas. JA: Boyle. MG: Dohertv. MD: Bell SEJ: Kern. AM: Martin. OJF: Teh. AS: Teo. KBK:
Milne. WL (2011). Combined antenna and localized plasmon resonance in raman scattering from random
arrays of silver-coated, vertically aligned multiwalled carbon nanotubes. Nano Lett 11:365-371.
http://dx.doi.org/10.1021/nll02838w
R-4
-------
de Wit CA: Herzke. D: Vorkamp. K. (2010). Brominated flame retardants in the Arctic environment ~ trends
and new candidates [Review]. Sci Total Environ 408:2885-2918.
http://dx.doi.0rg/10.1016/i.scitotenv.2009.08.037
Delorme. M: Muro. Y: Aral T: Banas. D: Frame. SR: Reed. K: Warheit D. (2012). Ninety-day inhalation
toxicity study with a vapor grown carbon nanofiber in rats. Toxicol Sci 128: 449-460.
http://dx.doi.org/10.1093/toxsci/kfsl72
Deng. X: Jia. G: Wang. H: Sun. H: Wang. X: Yang. S: Wang. T: Liu. Y. (2007). Translocation and fate of multi-
walled carbon nanotubes in vivo. Carbon 45: 1419-1424.
Desai C: Addo Ntim. S: Mitra. S. (2012). Antisolvent precipitation of hydrophobic functionalized multiwall
carbon nanotubes in an aqueous environment. J Colloid Interface Sci 368: 115-120.
http://dx.doi.0rg/10.1016/i.jcis.2011.ll.019
Pi Sotto. A: Chiaretti. M: Carru. GA: Bellucci S: Mazzanti. G. (2009). Multi-walled carbon nanotubes: Lack of
mutagenic activity in the bacterial reverse mutation assay. Toxicol Lett 184: 192-197.
http://dx.doi.0rg/10.1016/i.toxlet.2008.ll.007
Donaldson. K: Aitken. R: Tran. L: Stone. V: Puffin. R: Forrest. G: Alexander. A. (2006). Carbon nanotubes: A
review of their properties in relation to pulmonary toxicology and workplace safety [Review]. Toxicol Sci
92: 5-22. http://dx.doi.org/10.1093/toxsci/kfi 130
Donaldson. K: Beswick. PH: Gilmour. PS. (1996). Free radical activity associated with the surface of particles:
A unifying factor in determining biological activity? Toxicol Lett 88: 293-298.
Donaldson. K: Murphy. FA: Puffin. R: Poland. CA. (2010). Asbestos, carbon nanotubes and the pleural
mesothelium: A review and the hypothesis regarding the role of long fibre retention in the parietal pleura,
inflammation and mesothelioma [Review]. Part Fibre Toxicol 7: 42. http://dx.doi.org/10.1186/1743-8977-7-5
Doudrick. K: Herckes. P: Westerhoff. P. (2012). Detection of Carbon Nanotubes in Environmental Matrices
Using Programmed Thermal Analysis. Environ Sci Technol 46: 12246-12253.
http://dx.doi.org/10.1021/es300804f
Drottar. K: Krueger. H. (1998). Pentabromodiphenyl oxide (PeBDPO): A flow-through life-cycle toxicity test
with the cladoceran (Daphnia magna). (439A-109). Easton, MD: Wildlife International Ltd.
Dussert. AS: Gooris. E. (1997). Characterisation of the mineral content of a physical sunscreen emulsion and its
distribution onto human stratum corneum. Int J Cosmet Sci 19: 119-129. http://dx.doi.org/10.1111/i. 1467-
2494.1997.tb00175.x
ECB (European Chemicals Bureau). (2003). Bis(pentabromophenyl) ether: Summary risk assessment report.
(Special Publication 1.02.78). Luxembourg: Office for Official Publications of the European Communities.
http://echa.europa.eu/documents/10162/Od251331-31ea-4963-aebO-f97ea886d594
ECHA (European Chemicals Agency). (2008). Guidance on information requirements and chemicals safety
assessment. Helsinki, Finland, http://echa.europa.eu/web/guest/guidance-documents/guidance-on-
information-requirements-and-chemical-safetv-assessment
ECHA (European Chemicals Agency). (2013). Multi-Walled Carbon Nanotubes (MWCNT), synthetic graphite
in tubular shape. http://apps.echa.europa.eu/registered/data/dossiers/DISS-b281dlaO-c6d8-5dcf-e044-
00144f67d031/DISS-b281dlaO-c6d8-5dcf-e044-00144f67d031 DISS-b281dlaO-c6d8-5dcf-e044-
00144f67d031.html
Eckelman. MJ: Mauter. MS: Isaacs. JA: Elimelech. M. (2012). New perspectives on nanomaterial aquatic
ecotoxicity: production impacts exceed direct exposure impacts for carbon nanotoubes. Environ Sci Technol
46: 2902-2910. http://dx.doi.org/10.1021/es203409a
Edgington. AJ: Roberts. AP: Taylor. LM: Alloy. MM: Reppert. J: Rao. AM: Mao. J: Klaine. SJ. (2010). The
influence of natural organic matter on the toxicity of multiwalled carbon nanotubes. Environ Toxicol Chem
29: 2511-2518. http://dx.doi.org/10.1002/etc.309
R-5
-------
Elder. A: Lynch. I: Grieger. K: Chan-Remillard. S: Gatti A: Gnewuch. H: Kenawv. E: Korenstein. R:
Kuhlbusch. T: Linker. F: Matias. S: Monteiro-Riviere. N: Pinto. VRS: Rudnitsky. R: Savolainen. K:
Shvedova. A. (2009). Human health risks of engineered nanomaterials. In Nanoparticles: Risks and Benefits.
The Netherlands: Springer Netherlands.
Elgrabli D: Abella-Gallart S: Robidel F: Rogerieux. F: Boczkowskj J: Lacroix. G. (2008a). Induction of
apoptosis and absence of inflammation in rat lung after intratracheal instillation of multiwalled carbon
nanotubes. Toxicology 253: 131-136. http://dx.doi.0rg/10.1016/j.tox.2008.09.004
Elgrabli. D: Floriani. M: Abella-Gallart. S: Meunier. L: Gamez. C: Delalain. P: Rogerieux. F: Boczkowski. J:
Lacroix. G. (2008b). Biodistribution and clearance of instilled carbon nanotubes in rat lung. Part Fibre
Toxicol 5: 20. http://dx.doi.org/10.1186/1743-8977-5-20
Ellinger-Ziegelbauer. H: Pauluhn. J. (2009). Pulmonary toxicity of multi-walled carbon nanotubes (Baytubes)
relative to alpha-quartz following a single 6h inhalation exposure of rats and a 3 months post-exposure
period. Toxicology 266: 16-29. http://dx.doi.0rg/10.1016/i.tox.2009.10.007
Ema. M: Matsuda. A: Kobavashi. N: Nava. M: Nakanishi. J. (2011). Evaluation of dermal and eye irritation and
skin sensitization due to carbon nanotubes. Regul Toxicol Pharmacol 61: 276-281.
http://dx.doi.0rg/10.1016/i.vrtph.2011.08.007
Environment Canada. (2006). Canadian Environmental Protection Act, 1999: Ecological screening assessment
report on polybrominated diphenyl ethers (PBDEs).
Environment Canada. (2010). Ecological state of the science report on decabromodiphenyl ether (decaBDE).
http://www.ec.gc.ca/lcpe-cepa/default.asp?lang=En&n=B901A9EB&offset=l&toc=show
EU (European Union). (2002). European Union risk assessment report: Bis(pentabromophenyl) ether. In 1st
Priority List (pp. 1-282). (EUR 20402 EN). Luxembourg: Office for Official Publications of the European
Communities, http://echa.europa.eu/documents/10162/da9bc4c4-8e5b-4562-964c-5b4cf59d2432
Exponent (Exponent Engineering and Scientific Consulting). (2010). Fire & flammability testing.
http://www.exponent.com/fire flammabilitv testing/
Fenoglio. I: Aldieri. E: Gazzano. E: Cesano. F: Colonna. M: Scarano. D: Mazzucco. G: Attanasio. A: Yakoub.
Y: Lison. D: Fubini. B. (2012). Thickness of multiwalled carbon nanotubes affects their lung toxicity. Chern
Res Toxicol 25: 74-82. http://dx.doi.org/10.1021/tx200255h
Fleury. D: Bomfim. JAS: Vignes. A: Girard. C: Metz. S: Munoz. F: Rmili. B: Ustache. A: Guiot A: Bouillard.
JX. (2011). Identification of the main exposure scenarios in the production of CNT-polymer nanocomposites
by melt-moulding process. J Clean Prod, http://dx.doi.0rg/10.1016/i.jclepro.2011.ll.009
Fujitani. T: Ohyama. K: Hirose. A: Nishimura. T: Nakae. D: Ogata. A. (2012). Teratogenicity of multi-wall
carbon nanotube (MWCNT) in ICR mice. J Toxicol Sci 37: 81-89.
Gamer. AO: Leibold. E: Van Ravenzwaav. B. (2006). The in vitro absorption of microfine zinc oxide and
titanium dioxide through porcine skin. Toxicol In Vitro 20: 301-307.
http://dx.doi.0rg/10.1016/i.tiv.2005.08.008
Gasser. M: Wick. P: Clift MJ: Blank. F: Diener. L: Yaa B: Gehr. P: Krug. HF: Rothen-Rutishauser. B. (2012).
Pulmonary surfactant coating of multi-walled carbon nanotubes (MWCNTs) influences their oxidative and
pro-inflammatory potential in vitro. Part Fibre Toxicol 9: 1-13. http://dx.doi.org/10.1186/1743-8977-9-17
Ge. C: Li. W: Li. Y: Li. B: Du. J: Oiu. Y: Liu. Y: Gao. Y: Chai. Z: Chen. C. (2011). Significance and systematic
analysis of metallic impurities of carbon nanotubes produced by different manufacturers. J Nanosci
Nanotechnol 11: 2389-2397. http://dx.doi.org/10.1166/inn.2011.3520
Gevs. J: Nemerv. B: Hoet PH. (2007). Optimisation of culture conditions to develop an in vitro pulmonary
permeability model. Toxicol In Vitro 21: 1215-1219. http://dx.doi.0rg/10.1016/i.tiv.2007.05.012
Ghio. AJ: Stonehuerner. J: Dailev. LA: Carter. JD. (1999). Metals associated with both the water-soluble and
insoluble fractions of an ambient air pollution particle catalyze an oxidative stress. Inhal Toxicol 11:37-49.
R-6
-------
Golovin. YI: Golovin. DY: Shuklinov. AV: Stolyarov. RA: Vasvukov. VM. (2011). Electrodeposition of nickel
nanoparticles onto multiwalled carbon nanotubes. TechPhys Lett 37: 253-255.
http://dx.doi.org/10.1134/sl063785011030217
Goncalves. AG: Jarrais. B: Pereira. C: Morgado. J: Freire. C: Pereira. MFR. (2012). Functionalization of textiles
with multi-walled carbon nanotubes by a novel dyeing-like process. Journal of Materials Science 47: 5263-
5275. http://dx.doi.org/10.1007/sl0853-012-6412-4
Gottschalk. F: Sonderer. T: Scholz. RW: Nowack. B. (2009). Modeled environmental concentrations of
engineered nanomaterials (TiO2, ZnO, Ag, CNT, Fullerenes) for different regions. Environ Sci Technol 43:
9216-9222. http://dx.doi.org/10.1021/es9015553
Gottschalk. F: Sonderer. T: Scholz. RW: Nowack. B. (2010). Possibilities and limitations of modeling
environmental exposure to engineered nanomaterials by probabilistic material flow analysis. Environ Toxicol
Chem29: 1036-1048. http://dx.doi.org/10.1002/etc.135
Grefiler. S: Simko. M: Gazso. A: Fiedeler. U: Nentwich. M. (2010). Nano-Textiles. NanoTrust Dossiers 15: 1-5.
Grzybowski. K. (2009). Potential application of Carbon Nanotubes (CNTs) as a flame retardant additive -
technical and market analysis. Grzybowski, K. http://www.frost.com/prod/servlet/market-insight-
print.pag?docid= 159773 930
Gustavsson. P: Hedmer. M: Rissler. J. (2011). Carbon nanotubes: Exposure, toxicology and protective measures
in the work environment. Swedish Work Environment Authority.
https://lup.lub.lu.se/luur/download?func=downloadFile&recordOId=2295102&fileOId=2300375
Gutowskj TG: Liow. JYH: Sekulic. DP. (2010). Minimum exergy requirements for the manufacturing of carbon
nanotubes. In Proceedings of the 2010 IEEE international symposium on sustainable systems and technology.
Piscataway, NJ: IEEE. http://dx.doi.org/10.1109/ISSST.2010.5507687
Han. JH: Lee. EJ: Lee. JH: So. KP: Lee. YH: Bae. GN: Lee. SB: Ji. JH: Cho. MH: Yu. IJ. (2008). Monitoring
multiwalled carbon nanotube exposure in carbon nanotube research facility. Inhal Toxicol 20: 741-749.
http://dx.doi.org/10.1080/08958370801942238
Hardy. M. (2002a). The toxicology of the three commercial polybrominated diphenyl oxide (ether) flame
retardants. Chemosphere 46: 757-777. http://dx.doi.org/10.1016/s0045-6535(01)00240-5
Hardy. ML. (2002b). A comparison of the properties of the major commercial PBDPO/PBDE product to those of
major PBB andPCB products. Chemosphere 46: 717-728.
Hartung. T: Sabbioni. E. (2011). Alternative in vitro assays in nanomaterial toxicology. Wiley Interdiscip Rev
Nanomed Nanobiotechnol 3: 545573. http://dx.doi.org/10.1002/wnan. 153
He. M: Zhou. R: Guo. X. (2012). Behavior of stabilized multiwalled carbon nanotubes in a FeC13 coagulation
system and the structure characteristics of the produced floes. J Colloid Interface Sci 366: 173-178.
http://dx.doi.0rg/10.1016/i.icis.2011.09.059
Healy. ML: Dahlben. LJ: Isaacs. JA. (2008). Environmental assessment of single-walled carbon nanotube
processes. J IndEcol 12: 376-393. http://dx.doi.org/10.1111/i. 1530-9290.2008.00058.X
Helland. A: Wick. P: Koehler. A: Schmid. K: Som. C. (2007). Reviewing the environmental and human health
knowledge base of carbon nanotubes [Review]. Environ Health Perspect 115:1125-1131.
Hirsch. A: Vostrowsky. O. (2005). Functionalization of carbon nanotubes. Top Curr Chem 245: 193-237.
http://dx.doi.org/10.1007/b98169
Holgate. ST. (2010). Exposure, uptake, distribution and toxicity of nanomaterials in humans [Review]. Journal
of Biomedical Nanotechnology 6: 1-19. http://dx.doi.org/10.1166/jbn.201Q. 1098
Hou. PX: Liu. C: Cheng. HM. (2008). Purification of carbon nanotubes. Carbon 46: 2003-2025.
http://dx.doi.0rg/10.1016/i.carbon.2008.09.009
Howlett M. (2008). Nanocyl: Carbon nanotubes specialist. Presentation presented at.
R-7
-------
Huang. J: Zhang. Q: Zhao. M: Wei. F. (2012). A review of the large-scale production of carbon nanotubes: The
practice of nanoscale process engineering. Chinese Science Bulletin 57: 157-166.
http://dx.doi.org/10.1007/sll434-011-4879-z
Hyung. H: Fortner. JD: Hughes. JB: Kim. JH. (2007). Natural organic matter stabilizes carbon nanotubes in the
aqueous phase. Environ Sci Technol 41: 179-184. http://dx.doi.org/10.1021/es061817g
Hyung. H: Kim. JH. (2008). Natural organic matter (NOM) adsorption to multi-walled carbon nanotubes: Effect
of NOM characteristics and water quality parameters. Environ Sci Technol 42: 4416-4421.
http://dx.doi.org/10.1021/es702916h
ICL (ICL Industrial Products). (2010). Flame retardant systems for textiles: General application data sheet for
textile coating applications [Fact Sheet].
http://www.iclfr.com/brome^rome.nsf/viewAllBvUNID/B438A248158C3DB4C22577ED0044C095/$file/h
overetFR svstems.pdf
Illinois Environmental Protection Agency. (2007). Report on alternatives to the flame retardant DecaBDE:
Evaluation of toxicity, availability, affordability, and fire safety issues. Illinois.
Im. JS: Bai. BC: Bae. TS: In. SJ: Lee. YS. (2011). Improved anti-oxidation properties of electrospun
polyurethane nanofibers achieved by oxyfluorinated multi-walled carbon nanotubes and aluminum
hydroxide. Mater ChemPhys 126: 685-692. http://dx.doi.0rg/10.1016/j.matchemphys.2010.12.061
Innovative Research and Products Incorporated. (2011). Production and applications of carbon nanotubes,
carbon nanofibers, fullerenes, graphene and nanodiamonds: A global technology survey and market analysis.
http://www.innoresearch.net/report_summary.aspx?id=77&pg=53 l&rcd=ET-l 13&pd=2/l/2011
Inoue. K: Koike. E: Yanagisawa. R: Hirano. S: Nishikawa. M: Takano. H. (2009). Effects of multi-walled
carbon nanotubes on a murine allergic airway inflammation model. Toxicol Appl Pharmacol 237: 306-316.
http://dx.doi.0rg/10.1016/i.taap.2009.04.003
Inoue. K: Takano. H: Koike. E: Yanagisawa. R: Sakurai M: Tasaka. S: Ishizaka. A: Shimada. A. (2008). Effects
of pulmonary exposure to carbon nanotubes on lung and systemic inflammation with coagulatory disturbance
induced by lipopolysaccharide in mice. Exp Biol Med 233: 1583-1590. http://dx.doi.org/10.3181/0805-RM-
179
IOM (Institute of Medicine). (2012). ENPRA: Risk assessment of engineered nanoparticles. Available online at
http://www.enpra.eu/
IOM (Institute of Medicine). (2013). Marina: Managing risks of nanomaterials. Available online at
http://www.marina-fp7.eu/
IPEN. (2012). Nano techno logy and the environment: A mismatch between claims and reality. European
Environmental Bureau. http://www.eeb.org/documents/090713-OECD-environmental-Brief.pdf
Irin. F: Shrestha. B: Cafias. JE: Saed. MA: Green. MJ. (2012). Detection of carbon nanotubes in biological
samples through microwave-induced heating. Carbon 50: 4441-4449.
http://dx.doi.0rg/10.1016/i.carbon.2012.05.022
Isaacs. JA: Tanwanj A: Healy. ML: Dahlben. LJ. (2010). Economic assessment of single-walled carbon
nanotube processes. J Nanopart Res 12: 551-562. http://dx.doi.org/10.1007/sll051-009-9673-3
Isitman. NA: Kaynak. C. (2010). Nanoclay and carbon nanotubes as potential synergists of an organophosphorus
flame-retardant inpoly(methyl methacrylate). Polym Degrad Stabil 95: 1523-1532.
http://dx.doi.0rg/10.1016/i.polvmdegradstab.2010.06.013
Jain. S: Thakare. VS: Das. M: Godugu. C: Jain. AK: Mathur. R: Chuttani. K: Mishra. AK. (2011). Toxicity of
multiwalled carbon nanotubes with end defects critically depends on their functionalization density. Chem
Res Toxicol 24: 2028-2039. http://dx.doi.org/10.1021/tx2003728
Ji P: Yang. M: Feng. W. (2011). Functionalization of multiwalled carbon nanotubes with amphiphilic
poly(aspartic acid). AIChE J 57: 772-777. http://dx.doi.org/10.1002/aic. 12306
R-8
-------
Jia. G: Wang. H: Yan. L: Wang. X: Pel. R: Yan. T: Zhao. Y: Quo. X. (2005). Cytotoxicity of carbon
nanomaterials: Single-wall nanotube, multi-wall nanotube, and fullerene. Environ Sci Technol 39: 1378-
1383.
Johansson. N: Viberg. H: Fredriksson. A: Eriksson. P. (2008). Neonatal exposure to deca-brominated diphenyl
ether (PBDE 209) causes dose-response changes in spontaneous behaviour and cholinergic susceptibility in
adult mice. Neurotoxicology 29: 911-919. http://dx.doi.0rg/10.1016/i.neuro.2008.09.008
Johnson-Restrepo. B: Kannan. K. (2009). An assessment of sources and pathways of human exposure to
polybrominated diphenyl ethers in the United States. Chemosphere 76: 542-548.
http://dx.doi.0rg/10.1016/i.chemosphere.2009.02.068
Johnson. PR: Methner. MM: Kennedy. AJ: Steevens. JA. (2010). Potential for occupational exposure to
engineered carbon-based nanomaterials in environmental laboratory studies. Environ Health Perspect 118:
49-54. http://dx.doi.org/10.1289/ehp.0901076
Johnston. HJ: Hutchison. GR: Christensen. FM: Peters. S: Hankin. S: Aschberger. K: Stone. V. (2010). A critical
review of the biological mechanisms underlying the in vivo and in vitro toxicity of carbon nanotubes: The
contribution of physico-chemical characteristics [Review]. Nanotoxicology 4: 207-246.
http://dx.doi.org/10.3109/17435390903569639
Karthikevan. S: Mahalingam. P: Karthik. M. (2009). Large scale synthesis of carbon nanotubes [Review]. E-
Journal of Chemistry 6: 1-12.
Kashiwagi T: Du. F: Douglas. JF: Winev. KI: Harris. RH: Shields. JR. (2005a). Nanoparticle networks reduce
the flammability of polymer nanocomposites. Nat Mater 4: 928-933. http://dx.doi.org/10.1038/nmatl502
Kashiwagi. T: Du. F: Winev. KI: Groth. KM: Shields. JR: Bellaver. SP: Kim. H: Douglas. JF. (2005b).
Flammability properties of polymer nanocomposites with single-walled carbon nanotubes: Effects of
nanotube dispersion and concentration. Polymer (Guildf) 46: 471-481.
http://dx.doi.0rg/10.1016/i.polvmer.2004.10.087
Kashiwagi. T: Fagan. J: Douglas. JF: Yamamoto. K: Heckert. AN: Leigh. SD: Obrzut J: Du. F: Lin-Gibson. S:
Mu. M: Winev. KI: Haggenmueller. R. (2007). Relationship between dispersion metric and properties of
PMMA/SWNT nanocomposites. Polymer (Guildf) 48: 4855-4866.
http://dx.doi.0rg/10.1016/i.polvmer.2007.06.015
Kashiwagi. T: Grulke. E: Hilding. J: Groth. K: Harris. R: Butler. K: Shields. J: Kharchenko. S: Douglas. J.
(2004). Thermal and flammability properties of polypropylene/carbon nanotube nanocomposites. Polymer
(Guildf) 45: 4227-4239. http://dx.doi.0rg/10.1016/i.polvmer.2004.03.088
Kato. T: Totsuka. Y: Ishino. K: Matsumoto. Y: Tada. Y: Nakae. D: Goto. S: Masuda. S: Ogo. S: Kawanishi. M:
Yagi. T: Matsuda. T: Watanabe. M: Wakabavashi. K. (2013). Genotoxicity of multi-walled carbon nanotubes
in both in vitro and in vivo assay systems. Nanotoxicology 7: 452-461.
http://dx.doi.org/10.3109/17435390.2012.674571
Kemmlein. S: Hahn. O: Jann. O. (2003). Emissions of organophosphate and brominated flame retardants from
selected consumer products and building materials. Atmos Environ 37: 5485-5493.
http://dx.doi.0rg/10.1016/i.atmosenv.2003.09.025
Kennedy. AJ: Gunter. JC: Chappell MA: Goss. JD: Hull MS: Kirgan. RA: Steevens. JA. (2009). Influence of
nanotube preparation in aquatic bioassays. Environ Toxicol Chem 28: 1930-1938.
http://dx.doi.0rg/10.1897/09-024.l
Kennedy. AJ: Hull MS: Steevens. JA: Dontsova. KM: Chappell MA: Gunter. JC: Weiss. CA. Jr. (2008).
Factors influencing the partitioning and toxicity of nanotubes in the aquatic environment. Environ Toxicol
Chem27: 1932-1941. http://dx.doi.0rg/10.1897/07-624.l
Kennedy. TP: Dodson. R: Rao. NV: Ky. H: Hopkins. C: Baser. M: Tolley. E: Hoidal JR. (1989). Dusts causing
pneumoconiosis generate OH and produce hemolysis by acting as Fenton catalysts. Arch Biochem Biophys
269: 359-364.
R-9
-------
Kertesz. Z: Szikszai. Z: Gontier. E: Moretto. P: Surleve-Bazeiille. JE: Kiss. B: Juhasz. I: Hunyadi. J: AZ. K.
(2005). Nuclear microprobe study of TiO2-penetration in the epidermis of human skin xenografts. Nucl
Instrum Methods Phys Res B 231: 280-285. http://dx.doi.0rg/10.1016/i.nimb.2005.01.071
Khanna. V: Bakshi. BR: Lee. LJ. (2008). Carbon nanofiber production - Life cycle energy consumption and
environmental impact. J Ind Ecol 12: 394-410. http://dx.doi.org/10.1111/i. 1530-9290.2008.00052.X
Kierkegaard. A: Balk. L: Tjarnlund. U: De wit. CA: Jansson. B. (1999). Dietary uptake and biological effects of
Decabromodiphenyl Ether in Rainbow Trout (Oncorhynchus mykiss ). Environ Sci Technol 33: 1612-1617.
http://dx.doi.org/10.1021/es9807082
Kim. HC: Fthenakis. V. (2012). Life Cycle Energy and Climate Change Implications of Nanotechnologies. J Ind
Ecolno-no. http://dx.doi.0rg/10.llll/j.1530-9290.2012.00538.x
Kim. JE: Lim. HT: Minai-Tehrani. A: Kwon. JT: Shin. JY: Woo. CG: Choi. M: Baek. J: Jeong. DH: Ha. YC:
Chae. CH: Song. KS: Ahn. KH: Lee. JH: Sung. HJ: Yu. IJ: Beck. GR: Cho. MH. (2010). Toxicity and
clearance of intratracheally administered multiwalled carbon nanotubes from murine lung. J Toxicol Environ
Health A 73: 1530-1543. http://dx.doi.org/10.1080/15287394.2010.511578
Kim. JS: Lee. K: Lee. YH: Cho. HS: Kim. KH: Choi KH: Lee. SH: Song. KS: Kang. CS: Yu. IJ. (2011). Aspect
ratio has no effect on genotoxicity of multi-wall carbon nanotubes. Arch Toxicol 85: 775-786.
http://dx.doi.org/10.1007/s00204-010-0574-0
Kim. JS: Sung. JH: Song. KS: Lee. JH: Kim. SM: Lee. GH: An. KH: Lee. JS: Shin. JH: Park. JD: Yu. IJ.
(2012a). Persistent DNA damage measured by Comet assay of Sprague-Dawley rat lung cells after five days
of inhalation exposure and 1 month post-exposure to dispersed multi-wall carbon nanotubes (MWCNTs)
generated by new MWCNT aerosol generation system. Toxicol Sci 128: 439-448.
http://dx.doi.org/10.1093/toxsci/kfsl61
Kim. KT: Jang. MH: Kim. JY: Xing. B: Tanguav. RL: Lee. BG: Kim. SD. (2012b). Embryonic toxicity changes
of organic nanomaterials in the presence of natural organic matter. Sci Total Environ 426: 423-429.
http://dx.doi.0rg/10.1016/i.scitotenv.2012.03.050
Kim. YJ: Osako. M: Sakai. S. (2006). Leaching characteristics of polybrominated diphenyl ethers (PBDEs) from
flame-retardant plastics. Chemosphere 65: 506-513. http://dx.doi.0rg/10.1016/i.chemosphere.2006.01.019
Kishore. AS: Surekha. P: Murthy. PB. (2009). Assessment of the dermal and ocular irritation potential of multi-
walled carbon nanotubes by using in vitro and in vivo methods. ToxicolLett 191: 268-274.
http://dx.doi.0rg/10.1016/i.toxlet.2009.09.007
Kiss. B: Biro. T: Czifra. G: Toth. BI: Kertesz. Z: Szikszai. Z: Kiss. AZ: Juhasz. I: Zouboulis. CC: Hunyadi. J.
(2008). Investigation of micronized titanium dioxide penetration in human skin xenografts and its effect on
cellular functions of human skin-derived cells. Exp Dermatol 17: 659-667. http://dx.doi.org/10.1111/i. 1600-
0625.2007.00683.x
Klaine. SJ: Koelmans. AA: Home. N: Carlev. S: Handy. RD: Kapustka. L: Nowack. B: von der Kammer. F.
(2012). Paradigms to assess the environmental impact of manufactured nanomaterials. Environ Toxicol
Chem31: 3-14. http://dx.doi.org/10.1002/etc.733
Klaper. R: Arndt D: Setvowati K: Chen. J: Goetz. F. (2010). Functionalization impacts the effects of carbon
nanotubes on the immune system of rainbow trout, Oncorhynchus mykiss. Aquat Toxicol 100: 211-217'.
http://dx.doi.0rg/10.1016/i.aquatox.2010.07.023
Kohler. A: Som. C: Helland. A: Gottschalk. F. (2008). Studying the potential release of carbon nanotubes
throughout the application life cycle. J Clean Prod 16: 927-937.
http://dx.doi.0rg/10.1016/i.jclepro.2007.04.007
Kolosniai-Tabi. J: Hartman. KB: Boudjemaa. S: Ananta. JS: Morgan! G: Szwarc. H: Wilson. LJ: Moussa. F.
(2010). In vivo behavior of large doses of ultrashort and full-length single-walled carbon nanotubes after oral
and intraperitoneal administration to Swiss mice. ACS Nano 4: 1481-1492.
http://dx.doi.org/10.1021/nn901573w
R-10
-------
Kuo. YM: Sepulveda. MS: Sutton. TM: Ochoa-Acuna. HG: Muir. AM: Miller. B: Hua. I. (2010).
Bioaccumulation and biotransformation of decabromodiphenyl ether and effects on daily growth in juvenile
lake whitefish (Coregonus clupeaformis). Ecotoxicology 19: 751-760. http://dx.doi.org/10.1007/sl0646-009-
0451-x
Kushnir. D: Sanden. BA. (2008). Energy requirements of carbon nanoparticle production. J Ind Ecol 12: 360-
375. http://dx.doi.org/10.1111/i. 1530-9290.2008.00057.X
Lademann. J: Weigmann. H: Rickmever. C: Barthelmes. H: Schaefer. H: Mueller. G: Sterry. W. (1999).
Penetration of titanium dioxide microparticles in a sunscreen formulation into the horny layer and the
follicular orifice. Skin Pharmacol Physiol 12: 247-256. http://dx.doi.org/10.1159/000066249
Lagallv. CD: Reynolds. CCO: Grieshop. AP: Kandlikar. M: Rogak. SN. (2012). Carbon Nanotube and Fullerene
Emissions from Spark-Ignited Engines. Aerosol Sci Technol 46: 156-164.
http://dx.doi.org/10.1080/02786826.2011.617399
Lam. CW: James. JT: Mccluskev. R: Arepalli S: Hunter. RL. (2006). A review of carbon nanotube toxicity and
assessment of potential occupational and environmental health risks [Review]. Crit Rev Toxicol 36: 189-217.
http://dx.doi.org/10.1080/10408440600570233
Laoutid. F: Bonnaud. L: Alexandra. M: Lopez-Cuesta. JM: Dubois. P. (2009). New prospects in flame retardant
polymer materials: From fundamentals to nanocomposites. Mater Sci Eng R 63: 100-125.
http://dx.doi.0rg/10.1016/i.mser.2008.09.002
Lassen. C: Lokke. S: Andersen. LI. (1999). Brominated flame retardantssubstance flow analysis and assessment
of alternatives. Environmental Project.
Lavoie. ET: Heine. LG: Holder. H: Rossi. MS: Lee. RE: Connor. EA: Vrabel MA: Difiore. DM: Davies. CL.
(2010). Chemical alternatives assessment: enabling substitution to safer chemicals. Environ Sci Technol 44:
9244-9249. http://dx.doi.org/10.1021/esl015789
Law. RJ: Allchin. CR: de Boer. J: Covaci. A: Herzke. D: Lepom. P: Morris. S: Tronczynskj J: de Wit CA.
(2006). Levels and trends of brominated flame retardants in the European environment. Chemosphere 64:
187-208. http://dx.doi.0rg/10.1016/i.chemosphere.2005.12.007
Laxminaravana. K: Mill N. (2005). Functional nanotube-based textiles: Pathway to next generation fabrics with
enhanced sensing capabilities. Text Res J 75: 670-680. http://dx.doi.org/10.1177/0040517505059330
Lee. JH: Lee. SB: Bae. GN: Jeon. KS: Yoon. JU: Ji. JH: Sung. JH: Lee. BG: Yang. JS: Kim. HY: Kang. CS: Yu.
II (2010a). Exposure assessment of carbon nanotube manufacturing workplaces. Inhal Toxicol 22: 369-381.
http://dx.doi.org/10.3109/08958370903367359
Lee. SK: Bai BC: Im. JS: In. SJ: Lee. YS. (201 Ob). Flame retardant epoxy complex produced by addition of
montmorillonite and carbon nanotube. J Ind Eng Chem 16: 891-895.
http://dx.doi.0rg/10.1016/i.jiec.2010.09.014
Li. JG: Li. ON: Xu. JY: Cai. XO: Liu. RL: Li. YJ: Ma. JF: Li. WX. (2009). The pulmonary toxicity of multi-wall
carbon nanotubes in mice 30 and 60 days after inhalation exposure. J Nanosci Nanotechnol 9: 1384-1387.
Li. JG: Li. WX: Xu. JY: Cai. XQ: Liu. RL: Li. YJ: Z. QF: Li. QN. (2007). Comparative study of pathological
lesions induced by multi-walled carbon nanotubes in lungs of mice by intratracheal instillation and
inhalation. Environ Toxicol 22: 415-421. http://dx.doi.org/10.1002/tox.20270
Li. M: Huang. CP. (2011). The responses of Ceriodaphnia dubia toward multi-walled carbon nanotubes: Effect
of physicalchemical treatment. Carbon 49: 1672-1679. http://dx.doi.0rg/10.1016/i.carbon.2010.12.052
Li W: Zhu. L: Zha. J: Wang. Z. (201 la). Effects of decabromodiphenyl ether (BDE-209) on mRNA
transcription of thyroid hormone pathway and spermatogenesis associated genes in Chinese rare minnow
(Gobiocypris rams). Environ Toxicol. http://dx.doi.org/10.1002/tox.20767
Li. X: Zhao. H: Quan. X: Chen. S: Zhang. Y: Yu. H. (201 Ib). Adsorption of ionizable organic contaminants on
multi-walled carbon nanotubes with different oxygen contents. J Hazard Mater 186: 407-415.
http://dx.doi.0rg/10.1016/i.ihazmat.2010.ll.012
R-11
-------
Li. Y: Lu. D: Wong. CP. (2010). Carbon nanotubes (CNTs). In Electrical conductive adhesives with
nanotechnologies. New York, NY: Springer, http://dx.doi.org/10.1007/978-0-387-88783-8 2
Lim. JH: Kim. SH: Lee. 1C: Moon. C: Shin. DH: Kim. HC: Kim. JC. (201 la). Evaluation of Maternal Toxicity
in Rats Exposed to Multi-Wall Carbon Nanotubes during Pregnancy. 26: e2011006.
http://dx.doi.org/10.5620/eht.2011.26.e2011006
Lim. JH: Kim. SH: Shin. IS: Park. NH: Moon. C: Kang. SS: Park. SC: Kim. JC. (201 Ib). Maternal exposure to
multi-wall carbon nanotubes does not induce embryo-fetal developmental toxicity in rats. Birth Defects Res
B Dev Reprod Toxicol 92: 69-76. http://dx.doi.org/10.1002/bdrb.2Q283
Liu. D: Wang. L: Wang. Z: Cuschieri. A. (2012). Different cellular response mechanisms contribute to the
length-dependent cytotoxicity of multi-walled carbon nanotubes. Nanoscale Res Lett 7: 361.
http://dx.doi.org/10.1186/1556-276X-7-361
Liu. L: Fang. Z: Gu. A: Guo. Z. (201 la). Lubrication effect of the paraffin oil filled with functionalized
multiwalled carbon nanotubes for bismaleimide resin. Tribology Letters 42: 59-65.
http://dx.doi.org/10.1007/sll249-011-9749-Y
Liu. M: Tian. S: Chen. P: Zhu. L. (201 Ib). Predicting the bioavailability of sediment-associated polybrominated
diphenyl ethers using a 45-d sequential Tenax extraction. Chemosphere 85: 424-431.
http://dx.doi.0rg/10.1016/i.chemosphere.2011.07.069
Liu. X: Hurt. RH: Kane. AB. (2010). Biodurability of Single-Walled Carbon Nanotubes Depends on Surface
Functionalization. Carbon 48: 1961-1969. http://dx.doi.0rg/10.1016/i.carbon.2010.02.002
Lu. H: Song. L: Hu. Y. (201 la). A review on flame retardant technology in China. Part II: Flame retardant
polymeric nanocomposites and coatings. Polym Advan Technol 22: 379-394.
http://dx.doi.org/10.1002/pat.1891
Lu. H: Wilkie. C. (2010). Synergistic effect of carbon nanotubes and decabromodiphenyl oxide/Sb2O3 in
improving the flame retardancy of polystyrene. Polym Degrad Stabil 95.
http://dx.doi.0rg/10.1016/i.polvmdegradstab.2009.12.011.
Lu. Y: Shen. Q: Dai. Z. (20lib). Multiwalled carbon nanotubes as sorbent for online solid-phase extraction of
resveratrol in red wines prior to fused-core CIS-based ultrahigh-performance liquid chromatographytandem
mass spectrometry quantification. J Agric Food Chem 59: 70-77. http://dx.doi.org/10.1021/jfl03374f
Luizi. F. (2009). Responsible care and nanomaterials: Case study nanocyl. Presentation presented at European
Responsible Care Conference, October 21-23, 2009, Prague, Czech Republic.
Luoma. SN. (2008). Silver nanotechnologies and the environment: Old problems or new challenges.
Washington, DC: Project on Emerging Nanotechnologies.
Ma-Hock. L: Gamer. AO: Landsiedel R: Leibold. E: Frechen. T: Sens. B: Linsenbuehler. M: Van Ravenzwaav.
B_. (2007). Generation and characterization of test atmospheres with nanomaterials. Inhal Toxicol 19: 833-
848. http://dx.doi.org/10.1080/08958370701479190
Ma-Hock. L: Treumann. S: Strauss. V: Brill S: Luizi. F: Mertler. M: Wiench. K: Gamer. AO: van Ravenzwaav.
B: Landsiedel R. (2009). Inhalation toxicity of multiwall carbon nanotubes in rats exposed for 3 months.
Toxicol Sci 112: 468-481. http://dx.doi.org/10.1093/toxsci/kfpl46
Ma. H: Song. P: Fang. Z. (2011). Flame retarded polymer nanocomposites: Development, trend and future
perspective. Science China Chemistry 54: 302-313. http://dx.doi.org/10.1007/sll426-010-4196-4
Ma. HY: Tong. LF: Xu. ZB: Fang. ZP. (2008). Functionalizing carbon nanotubes by grafting on intumescent
flame retardant: Nanocomposite synthesis, morphology, rheology, and flammability. Adv Funct Mater 18:
414-421. http://dx.doi.org/10.1002/adfm.200700677
Mahy. M. (2009). Silicones and carbon nanotubes: From antistatic to fire barrier and fouling release coatings.
Nanocyl. http://www.materialsdav.be/Documents/Materialsday09 nanocyl.pdf
R-12
-------
Mattison. NT: O'Carroll. DM: Rowe. RK: Petersen. EJ. (2011). Impact of porous media grain size on the
transport of multiwalled carbon nanotubes. Environ Sci Technol 45: 9765-9775.
http://dx.doi.org/10.1021/es2017076
Mavon. A: Miguel C: Lejeune. O: Payre. B: Moretto. P. (2007). In vitro percutaneous absorption and in vivo
stratum corneum distribution of an organic and a mineral sunscreen. Skin Pharmacol Physiol 20: 10-20.
http://dx.doi.org/10.1159/000096167
Maynard. AD: Aitken. RJ. (2007). Assessing exposure to airborne nanomaterials: Current abilities and future
requirements. Nanotoxicology 1: 26-41. http://dx.doi.org/10.1080/17435390701314720
Maynard. AD: Baron. PA: Folev. M: Shvedova. AA: Kisin. ER: Castranova. V. (2004). Exposure to carbon
nanotube material: Aerosol release during the handling of unrefined single-walled carbon nanotube material.
J Toxicol Environ Health A 67: 87-107. http://dx.doi.org/10.1080/15287390490253688
Menzel F: Reinert T: Vogt J: Butz. T. (2004). Investigations of percutaneous uptake of ultrafine T1O2 particles
at the high energy ion nanoprobe LIPSION. Nucl Instrum Methods Phys Res B 219-220: 82-86.
http://dx.doi.0rg/10.1016/i.nimb.2004.01.032
Mercer. RR: Hubbs. AF: Scabillonj JF: Wang. L: Battelli. LA: Schwegler-Berry. D: Castranova. V: Porter. DW.
(2010). Distribution and persistence of pleura! penetrations by multi-walled carbon nanotubes. Part Fibre
Toxicol 7: 28. http://dx.doi.org/10.1186/1743-8977-7-28
Methner. M: Hodson. L: Dames. A: Geraci C. (2010). Nanoparticle emission assessment technique (NEAT) for
the identification and measurement of potential inhalation exposure to engineered nanomaterials - Part B:
Results from 12 field studies. J Occup Environ Hyg 7: 163-176.
http://dx.doi.org/10.1080/15459620903508066
Mezzo. L. (2010). Latest development of carbon nano tubes nanocomposites. Presentation presented at Materia
Nova, May 18, 2010, Mons, Belgium.
Migliore. L: Saracino. D: Bonelli. A: Colognato. R: D'Errico. MR: Magrini. A: Bergamaschj A: Bergamaschi.
K (2010). Carbon nanotubes induce oxidative DNA damage in RAW 264.7 cells. Environ Mol Mutagen 51:
294-303. http://dx.doi.org/10.1002/em.20545
Misra. SK: Dybowska. A: Berhanu. D: Luoma. SN: Valsami-Jones. E. (2012). The complexity of nanoparticle
dissolution and its importance in nanotoxicological studies. Sci Total Environ 438: 225-232.
http://dx.doi.0rg/10.1016/i.scitotenv.2012.08.066
Mitchell LA: Gao. J: Vander Wai. R: Gigliotti. A: Burchiel SW: Mcdonald. JD. (2007). Pulmonary and
systemic immune response to inhaled multiwalled carbon nanotubes. Toxicol Sci 100: 203-214.
http://dx.doi.org/10.1093/toxsci/kfml96
Mitchell. LA: Lauer. FT: Burchiel. SW: McDonald. JD. (2009). Mechanisms for how inhaled multiwalled
carbon nanotubes suppress systemic immune function in mice. Nat Nanotechnol 4: 451-456.
http://dx.doi.org/10.1038/nnano.2009.151
Moniruzzaman. M: Winev. KI. (2006). Polymer nanocomposites containing carbon nanotubes. Macromolecules
39: 5194-5205. http://dx.doi.org/10.1021/ma060733p
Monteiro-Riviere. NA: Inman. AO. (2006). Challenges for assessing carbon nanomaterial toxicity to the skin.
Carbon44: 1070-1078. http://dx.doi.0rg/10.1016/j.carbon.2005.ll.004
Morgan. AB. (2006). Flame retarded polymer layered silicate nanocomposites: A review of commercial and
open literature systems. Polym Advan Technol 17: 206-217. http://dx.doi.org/10.1002/pat.685
Morimoto. Y: Hirohashi. M: Ogamj A: Oyabu. T: Myojo. T: Todoroki. M: Yamamoto. M: Hashiba. M:
Mizuguchi. Y: Lee. BW: Kuroda. E: Shimada. M: Wang. WN: Yamamoto. K: Fujita. K: Endoh. S: Uchida.
K: Kobavashj N: Mizuno. K: Inada. M: Tao. H: Nakazato. T: Nakanishi. J: Tanaka. I. (2012). Pulmonary
toxicity of well-dispersed multi-wall carbon nanotubes following inhalation and intratracheal instillation.
Nanotoxicology 6: 587-599. http://dx.doi.org/10.3109/17435390.2011.594912
Morimoto. Y: Horie. M: Kobavashi. N: Shinohara. N: Shimada. M. (2013). Inhalation Toxicity Assessment of
Carbon-Based Nanoparticles. Ace ChemRes 46: 770-781. http://dx.doi.org/10.1021/ar200311b
R-13
-------
Motzkus. C: Chivas-Joly. C: Guillaume. E: Ducourtieux. S: Saragoza. L: Lesenechal D: Mace. T: Lopez-Cuesta.
JM: Longuet C. (2012). Aerosols emitted by the combustion of polymers containing nanoparticles. J
NanopartRes 14: 687. http://dx.doi.org/10.1007/sl 1051-011-0687-2
Mouchet F: Landois. P: Datsvuk. V: Puech. P: Pinelli. E: Flahaut E: Gauthier. L. (2011). International
amphibian micronucleus standardized procedure (ISO 21427-1) for in vivo evaluation of double-walled
carbon nanotubes toxicity and genotoxicity in water. Environ Toxicol 26: 136-145.
http://dx.doi.org/10.1002/tox.20537
Mouchet F: Landois. P: Puech. P: Pinelli. E: Flahaut. E: Gauthier. L. (2010). Carbon nanotube ecotoxicity in
amphibians: assessment of multiwalled carbon nanotubes and comparison with double-walled carbon
nanotubes. Nanomed 5: 963-974. http://dx.doi.org/10.2217/nnm.10.60
Mouchet F: Landois. P: Sarremejean. E: Bernard. G: Puech. P: Pinelli. E: Flahaut E: Gauthier. L. (2008).
Characterisation and in vivo ecotoxicity evaluation of double-wall carbon nanotubes in larvae of the
amphibianXenopus laevis. Aquat Toxicol 87: 127-137. http://dx.doi.0rg/10.1016/i.aquatox.2008.01.011
Mueller. NC: Nowack. B. (2008). Exposure modeling of engineered nanoparticles in the environment. Environ
Sci Technol 42: 4447-4453. http://dx.doi.org/10.1021/es7029637
Muleja. AA: Mbianda. XY: Krause. RW: Pillav. K. (2012). Synthesis, characterization and thermal
decomposition behaviour of triphenylphosphine-linked multiwalled carbon nanotubes. Carbon 50: 2741-
2751. http://dx.doi.0rg/10.1016/i.carbon.2012.02.033
Muller. J: Decordier. I: Hoet PH: Lombaert N: Thomassen. L: Huaux. F: Lison. D: Kirsch-Volders. M. (2008a).
Clastogenic and aneugenic effects of multi-wall carbon nanotubes in epithelial cells. Carcinogenesis 29: 427-
433. http://dx.doi.org/10.1093/carcin/bgm243
Muller. J: Delos. M: Panin. N: Rabolli. V: Huaux. F: Lison. D. (2009). Absence of carcinogenic response to
multiwall carbon nanotubes in a 2-year bioassay in the peritoneal cavity of the rat. Toxicol Sci 110: 442-448.
http://dx.doi.org/10.1093/toxsci/kfplOO
Muller. J: Huaux. F: Fonseca. A: Nagv. JB: Moreau. N: Delos. M: Raymundo-Pinero. E: Beguin. F: Kirsch-
Volders. M: Fenoglio. I: Fubinj B: Lison. D. (2008b). Structural defects play a major role in the acute lung
toxicity of multiwall carbon nanotubes: Toxicological aspects. ChemRes Toxicol 21: 1698-1705.
http://dx.doi.org/10.1021/tx800101p
Muller. J: Huaux. F: Moreau. N: Misson. P: Heilier. JF: Delos. M: Arras. M: Fonseca. A: Nagv. JB: Lison. D.
(2005). Respiratory toxicity of multi-wall carbon nanotubes. Toxicol Appl Pharmacol 207: 221-231.
http://dx.doi.0rg/10.1016/i.taap.2005.01.008
Murphy. FA: Poland. CA: Puffin. R: Al-Jamal KT: Ali-Boucetta. H: Nunes. A: Byrne. F: Prina-Mello. A:
Volkov. Y: Li. S: Mather. SJ: Bianco. A: Prato. M: MacNee. W: Wallace. WA: Kostarelos. K: Donaldson. K.
(2011). Length-dependent retention of carbon nanotubes in the pleural space of mice initiates sustained
inflammation and progressive fibrosis on the parietal pleura. Am J Pathol 178: 2587-2600.
http://dx.doi.0rg/10.1016/i.aipath.2011.02.040
Murphy. FA: Schinwald. A: Poland. CA: Donaldson. K. (2012). The mechanism of pleural inflammation by long
carbon nanotubes: interaction of long fibres with macrophages stimulates them to amplify pro-inflammatory
responses in mesothelial cells. Part Fibre Toxicol 9: 8. http://dx.doi.org/10.1186/1743-8977-9-8
Mwangi JN: Wang. N: Ingersoll CG: Hardestv. DK: Brunson. EL: Li. H: Deng. B. (2012). Toxicity of carbon
nanotubes to freshwater aquatic invertebrates. Environ Toxicol Chem 31: 1823-1830.
http://dx.doi.org/10.1002/eto. 1888
Nagai. H: Okazakj Y: Chew. SH: Misawa. N: Yamashita. Y: Akatsuka. S: Ishihara. T: Yamashita. K:
Yoshikawa. Y: Yasui H: Jiang. L: Ohara. H: Takahashj T: Ichihara. G: Kostarelos. K: Miyata. Y:
Shinohara. H: Toyokuni. S. (2011). Diameter and rigidity of multiwalled carbon nanotubes are critical factors
in mesothelial injury and carcinogenesis. PNAS 108: E1330-E1338.
http://dx.doi.org/10.1073/pnas. 1110013108
R-14
-------
Nakanishi. J: Ogura. I: Kishimoto. A: Gamo. M: Ema. M: M. N: Shinohara. N: Kobavashi. N. (2011). Risk
assessment of manufactured nanomaterials: Carbon nanotubes (CNT). (NEDO project #P06041 (Research
and Development of Nanoparticle Characterization Methods)). Kawasaki City, Japan: NEDO.
http://www.aist-riss.jp/main/modules/product/nano rad.html?ml lang=en
Nakari. T: Pessala. P. (2005). In vitro estrogenicity of polybrominated flame retardants. Aquat Toxicol 74: 272-
279. http://dx.doi.org/10.1016/i.aquatox.2005.06.004
Nanocvl. (2009). THERMOCYL. http://www.nanocvl.com/en/Products-Solutions/Products/THERMOCYL.
Nanoshel. (2011). Technology, http://www.nanoshel.com/nanotechnology.php
Navarro. E: Baun. A: Behra. R: Hartmann. NB: Filser. J: Miao. AJ: Quigg. A: Santschi. PH: Sigg. L. (2008).
Environmental behavior and ecotoxicity of engineered nanoparticles to algae, plants, and fungi [Review].
Ecotoxicology 17: 372-386. http://dx.doi.org/10.1007/sl0646-008-0214-0
NCSL (National Conference of State Legislators). (2011). State regulation of PBDEs. Washington, DC: National
Conference of State Legislatures, http://www.ncsl.org/issues-research/environment-and-natural-
resources/state-regulation-of-pbdes.aspx
Nel A: Xia. T: Meng. H: Wang. X: Lin. S: Ji Z: Zhang. H. (2013). Nanomaterial Toxicity Testing in the 21st
Century: Use of a Predictive Toxicological Approach and High-Throughput Screening. Ace Chem Res 46:
607-621. http://dx.doi.org/10.1021/ar300022h
Nguyen. T: Pellegrin. B: Bernard. C: Gu. X: Gorham. JM: Stutzman. P: Stanley. D: Shapiro. A: Byrd. E:
Hettenhouser. R: Chin. J. (2011). Fate of nanoparticles during life cycle of polymer nanocomposites. J Phys
Conf Ser 304: 012060. http://dx.doi.Org/10.1088/1742-6596/304/l/012060
NICNAS (National Industrial Chemicals Notification and Assessment Scheme). (2012). Human health hazard
assessment and classification of carbon nanotubes. Canberra ACT, Australia: Safe Work Australia.
http://www.safeworkaustralia.gov.au/sites/swa/about/publications/pages/human-health-hazard-assessment-
and-classification-of-carbon-nanotubes
Nikula. KJ: Vallyathan. V: Green. FHY: Hahn. FF. (2001). Influence of exposure concentration or dose on the
distribution of paniculate material in rat and human lungs. Environ Health Perspect 109: 311-318.
NIOSH (National Institute for Occupational Safety and Health). (1994). Method 7400: asbestos and other fibers
by PCM. In NIOSH manual of analytical methods (4 ed.). Cincinnati, OH: U.S.: Department of Health and
Human Services, Centers for Disease Control and Prevention, National Institute for Occupational Safety and
Health (NIOSH). http://www.cdc.gov/niosh/nmam/pdfs/7400.pdf
NIOSH (National Institute for Occupational Safety and Health). (2010). NIOSH Current Intelligence Bulletin:
Occupational exposure to carbon nanotubes and nanofibers (pp. 149). Washington, DC.
http://www.cdc.gov/niosh/docket/archive/pdfs/NIOSH-161-A/0161-A-110110-DraftDocument.pdf
NIOSH (National Institute for Occupational Safety and Health). (2012). General safe practices for working with
engineered nanomaterials in research laboratories. U.S. Department of Health and Human Services.
NLM (National Institutes of Health, National Library of Medicine). (2011). ChemlDPlus [Database].
Norris. JM: Ehrmantraut JW: Gibbons. CL: Kociba. PJ: Schwetz. BA: Rose. JO: Humiston. CG: Jewett GL:
Crummett WB: Gehring. PJ: Tirsell JB: Brosier. JS. (1973). Toxicological and environmental factors
involved in the selection of decabromodiphenyl oxide as a fire retardant chemical. Appl Polym Symp 22:
195-219.
Norris. JM: Kociba. RJ: Schwetz. BA: Rose. JO: Humiston. CG: Jewett GL: Gehring. PJ: Mailhes. JB. (1975).
Toxicology of octabromobiphenyl and decabromodiphenyl oxide. Environ Health Perspect 11: 153-161.
North. KD. (2004). Tracking polybrominated diphenyl ether releases in a wastewater treatment plant effluent,
Palo Alto, California. Environ Sci Technol 38: 4484-4488. http://dx.doi.org/10.1021/es049627y
Nowack. B: Ranville. JF: Diamond. S: Gallego-Urrea. JA: Metcalfe. C: Rose. J: Home. N: Koelmans. AA:
Klaine. SJ. (2012). Potential scenarios for nanomaterial release and subsequent alteration in the environment
[Review]. Environ Toxicol Chem 31: 5059. http://dx.doi.org/10.1002/etc.726
R-15
-------
NRC (National Research Council). (2000). Toxicological risks of selected flame-retardant chemicals.
Washington, DC: National Academy Press.
NRC (National Research Council). (2009). Science and decisions: Advancing risk assessment. Washington, DC:
National Academies Press, http://www.nap.edu/catalog/12209.html
NRC (National Research Council). (2011). Sustainability and the U.S. EPA: Committee on incorporating
sustainability in the U.S. Environmental Protection Agency. Washington, DC: The National Academies
Press, http://www.nap.edu/catalog.php7record_id=13152
NRC (National Research Council). (2012). A research strategy for environmental, health, and safety aspects of
engineered nanomaterials. Washington, DC: National Academies Press.
http://www.nap.edu/catalog.php7record id=13347
NSTC (National Science and Technology Council). (2011 a). National Nanotechnology Initiative:
Environmental, health, and safety research strategy. Arlington, VA: National Nanotechnology Initiative.
http://www.nano.gov/node/681
NSTC (National Science and Technology Council). (20lib). National nanotechnology initiative: Strategic plan.
Arlington, VA: National Nanotechnology Initiative, http://www.nano.gov/node/581
NTP (National Toxicology Program). (1986). NTP toxicology and carcinogenesis studies of Decabromodiphenyl
Oxide (CAS No. 1163-19-5) inF344/N rats andB6C3Fl mice (Feed studies) (pp. 1-242). (309). Research
Triangle Park, NC: U.S. Department of Health and Human Services.
Nyden. MR: Zammarano. M: Harris. RH: Kramer. R: Uddin. NM: Marsh. ND. (2010). Characterizing particle
emissions from burning polymer nanocomposites. In 21th BCC Conference on Flame Retardation (pp. 717-
719). National Institute of Standards and Technology - Nanotech 2010.
http ://materiales .azc .uam. mx/area/Ingenieria_Materiales/investigaci%C3%B 3 n/2261204/cuan%20calif/Cuan
%20TechConWo2010/CD/Nanotech2010/pdf/1350.pdf
Nygaard. UC: Hansen. JS: Samuelsen. M: Alberg. T: Marioara. CD: Levik. M. (2009). Single-walled and multi-
walled carbon nanotubes promote allergic immune responses in mice. Toxicol Sci 109: 113-123.
http://dx.doi.org/10.1093/toxsci/kfp057
O'Carroll. DM: Liu. X: Mattison. NT: Petersen. EJ. (2013). Impact of diameter on carbon nanotube transport in
sand. J Colloid Interface Sci 390: 96-104. http://dx.doi.0rg/10.1016/i.jcis.2012.09.034
Oberdorster. E: Zhu. S: Blicklev. TM: McClellan-Green. P: Haasch. ML. (2006). Ecotoxicology of carbon-based
engineered nanoparticles: Effects of fullerene (C60) on aquatic organisms. Carbon 44: 1112-1120.
Oberdorster. G. (1996). Significance of particle parameters in the evaluation of exposure-dose-response
relationships of inhaled particles [Review]. Inhal Toxicol 8 Supplement: 73-89.
http://dx.doi.org/10.1080/02726359608906690
Oberdorster. G. (2010). Safety assessment for nanotechnology and nanomedicine: concepts of nanotoxicology. J
InternMed 267: 89-105. http://dx.doi.0rg/10.llll/i.1365-2796.2009.02187.x
Oberdorster. G: Maynard. A: Donaldson. K: Castranova. V: Fitzpatrick. J: Ausman. K: Carter. J: Karn. B:
Krevling. W: Lai. D: Olin. S: Monteiro-Riviere. N: Warheit D: Yang. H. (2005). Principles for
characterizing the potential human health effects from exposure to nanomaterials: Elements of a screening
strategy. Part Fibre Toxicol 2:1-35. http://dx.doi.org/10.1186/1743-8977-2-8
ODriscoll NJ: Messier. T: Robertson. MD: Murimboh. J. (2010). Suspension of multi-walled carbon nanotubes
(CNTs) in freshwaters: Examining the effect of CNT size. Water Air Soil Pollut 208: 235-241.
http://dx.doi.org/10.1007/sll270-009-0162-l
OECD (Organisation for Economic Co-operation and Development). (2009). OECD conference on potential
environmental benefits of nanotechnology: Fostering safe innovation-led growth. Paris, France. July 14-17,
2009.
http://www.oecd.org/fr/innovation/oecdconferenceonpotentialenvironmentalbenefitsofnanotechnologvfosteri
ngsafeinnovation-ledgrowth.htm
R-16
-------
OECD (Organisation for Economic Co-operation and Development). (2012). Important issues on risk assessment
of manufactured nanomaterials. In Series on the Safety of Manufactured Nanomaterials, No 3 3.
(ENV/JM/MONO(2012)8). Paris.
http://search.oecd.org/officialdocuments/displavdocumentpdf/?cote=env/im/mono(2012)8&doclanguage=en
Oleszczuk. P: Josko. I: Xing. B. (2011). The toxicity to plants of the sewage sludges containing multiwalled
carbon nanotubes. J Hazard Mater 186: 436-442. http://dx.doi.0rg/10.1016/i.jhazmat.2010.ll.028
Osier. M: Oberdorster. G. (1997). Intratracheal inhalation vs intratracheal instillation: differences in particle
effects. Toxicol Sci 40: 220-227.
Pack. S: Kashiwagi T: Stemp. D: Koo. J: Si. M: Sokolov. JC: Rafailovich. MH. (2009). Segregation of carbon
nanotubes/organoclays rendering polymer blends self-extinguishing. Macromolecules 42: 6698-6709.
http://dx.doi.org/10.1021/ma900966k
Pacurari. M: Castranova. V: Vallyathan. V. (2010). Single- and multi-wall carbon nanotubes versus asbestos:
Are the carbon nanotubes a new health risk to humans [Review]. J Toxicol Environ Health A 73: 378-395.
http://dx.doi.org/10.1080/15287390903486527
Pacurari. M: Qian. Y: Fu. W: Schwegler-Berry. D: Ding. M: Castranova. V: Quo. NL. (2012). Cell permeability,
migration, and reactive oxygen species induced by multiwalled carbon nanotubes in human microvascular
endothelial cells. J Toxicol Environ Health A 75: 112-128. http://dx.doi.org/10.1080/15287394.2011.615110
Palm. A: Cousins. IT: Mackav. D: Tysklind. M: Metcalfe. C: Alaee. M. (2002). Assessing the environmental fate
of chemicals of emerging concern: A case study of the polybrominated diphenyl ethers. Environ Pollut 117:
195-213.
Park. EJ: Cho. WS: Jeong. J: Yi. J: Choi. K: Park. K. (2009). Pro-inflammatory and potential allergic responses
resulting from B cell activation in mice treated with multi-walled carbon nanotubes by intratracheal
instillation. Toxicology 259: 113-121. http://dx.doi.0rg/10.1016/i.tox.2009.02.009
Patel V. (2011). Global carbon nanotubes market - industry beckons. In Nanotech Insights. Honolulu, HI:
Nanowerk. http://www.nanowerk.com/spotlight/spotid=231 IS.php
Patlolla. A: Knighten. B: Tchounwou. P. (2010a). Multi-walled carbon nanotubes induce cytotoxicity,
genotoxicity and apoptosis in normal human dermal fibroblast cells. Ethn Dis 20: SI-65-72.
Patlolla. A: Patlolla. B: Tchounwou. P. (2010b). Evaluation of cell viability, DNA damage, and cell death in
normal human dermal fibroblast cells induced by functionalized multiwalled carbon nanotube. Mol Cell
Biochem 338: 225-232. http://dx.doi.org/10.1007/sll010-009-0356-2
Patlolla. AK: Berry. A: Tchounwou. PB. (2011). Study of hepatotoxicity and oxidative stress in male Swiss-
Webster mice exposed to functionalized multi-walled carbon nanotubes. Mol Cell Biochem 358: 189-199.
http://dx.doi.org/10.1007/sll010-011-0934-y
Pauluhn. J. (2010a). Multi-walled carbon nanotubes (Baytubes): Approach for derivation of occupational
exposure limit. Regul Toxicol Pharmacol 57: 78-89. http://dx.doi.0rg/10.1016/i.yrtph.2009.12.012
Pauluhn. J. (2010b). Subchronic 13-week inhalation exposure of rats to multiwalled carbon nanotubes: Toxic
effects are determined by density of agglomerate structures, not fibrillar structures. Toxicol Sci 113: 226-242.
http://dx.doi.org/10.1093/toxsci/kfp247
Paxton. JW. (1995). The allometric approach for interspecies scaling of pharmacokinetics and toxicity of anti-
cancer drugs. Clin Exp Pharmacol Physiol 22: 851-854.
Peng. X: Tang. C: Yu. Y: Tan. J: Huang. Q: Wu. J: Chen. S: Mai. B. (2009). Concentrations, transport, fate, and
releases of polybrominated diphenyl ethers in sewage treatment plants in the Pearl River Delta, South China.
Environ Int 35: 303-309. http://dx.doi.0rg/10.1016/i.envint.2008.07.021
Petersen. EJ: Henry. TB. (2012). Methodological considerations for testing the ecotoxicity of carbon nanotubes
and fullerenes: Review [Review]. Environ Toxicol Chem 31: 60-72. http://dx.doi.org/10.1002/etc.710
Petersen. EJ: Huang. Q: Weber. J. r.. W. J. (2008). Ecological uptake and depuration of carbon nanotubes by
Lumbriculus variegatus. Environ Health Perspect 116: 496-500.
R-17
-------
Petersen. EJ: Huang. Q: Weber. WJ. (2010). Relevance of octanol-water distribution measurements to the
potential ecological uptake of multi-walled carbon nanotubes. Environ Toxicol Chem 29: 1106-1112.
http://dx.doi.org/10.1002/eto. 149
Petersen. EJ: Pinto. RA: Landrum. PF: Weber. WJ. (2009). Influence of carbon nanotubes on pyrene
bioaccumulation from contaminated soils by earthworms. Environ Sci Technol 43:4181-4187.
http://dx.doi.org/10.1021/es803023a
Petersen. EJ: Pinto. RA: Mai. DJ: Landrum. PF: Weber. WJ. Jr. (2011 a). Influence of polyethyleneimine
graftings of multi-walled carbon nanotubes on their accumulation and elimination by and toxicity to Daphnia
magna. Environ Sci Technol 45: 1133-1138. http://dx.doi.org/10.1021/esl030239
Petersen. EJ: Pinto. RA: Zhang. L: Huang. Q: Landrum. PF: Weber. WJ. (20lib). Effects of polyethyleneimine-
mediated functionalization of multi-walled carbon nanotubes on earthworm bioaccumulation and sorption by
soils. Environ Sci Technol 45: 3718-3724. http://dx.doi.org/10.1021/esl03004r
Petersen. EJ: Zhang. L: Mattison. NT: O'Carroll DM: Whelton. AJ: Uddin. N: Nguyen. T: Huang. Q: Henry.
TB: Holbrook. RD: Chen. KL. (201 Ic). Potential release pathways, environmental fate, and ecological risks
of carbon nanotubes [Review]. Environ Sci Technol 45: 9837-9856. http://dx.doi.org/10.1021/es201579y
Petrov. P: Lou. X: Pagnoulle. C: Jerome. C: Calberg. C: Jerome. R. (2004). Functionalization of multi-walled
carbon nanotubes by electrografting of Poly aery lonitrile. Macromol Rapid Comm 25: 987-990.
http://dx.doi.org/10.1002/marc.200400055
Piccinno. F: Gottschalk. F: Seeger. S: Nowack. B. (2012). Industrial production quantities and uses often
engineered nanomaterials in Europe and the world. J Nanopart Res 14: 1-11.
http://dx.doi.org/10.1007/sll051-012-1109-9
PINFA (Phosphorus, Inorganic and Nitrogen Flame Retardants Association). (2010). Innovative and sustainable
flame retardants in transportation. Brussels, Belgium.
Pinheiro. T: Pallon. J: Alves. LC: Verissimo. A: Filipe. P: Silva. JN: Silva. R. (2007). The influence of
corneocyte structure on the interpretation of permeation profiles of nanoparticles across skin. Nucl Instrum
Methods Phys Res B 260: 119-123. http://dx.doi.0rg/10.1016/i.nimb.2007.02.014
Plata. PL. (2009) Carbon nanotube synthesis and detection: Limiting the environmental impact of novel
technologies. (Doctoral Dissertation). Massachusetts Institute of Technology, Cambridge, MA. Retrieved
from http://hdl.handle.net/1912/2970
Plata. PL: Ferguson. PL: Westerhoff. P. (2012a). Express it in numbers: efforts to quantify engineered
nanoparticles in environmental matrices advance. Environ Sci Technol 46: 12243-12245.
http://dx.doi.org/10.1021/es302789n
Plata. PL: Hart. AJ: Reddv. CM: Gschwend. PM. (2009). Early evaluation of potential environmental impacts of
carbon nanotube synthesis by chemical vapor deposition. Environ Sci Technol 43: 8367-8373.
http://dx.doi.org/10.1021/es901626p
Plata. PL: Reddv. CM: Gschwend. PM. (2012b). Thermogravimetry-mass spectrometry for carbon nanotube
detection in complex mixtures. Environ Sci Technol 46: 12254-12261. http://dx.doi.org/10.1021/es203198x
Poland. CA: Puffin. R: Kinloch. I: Maynard. A: Wallace. WAH: Seaton. A: Stone. V: Brown. S: MacNee. W:
Donaldson. K. (2008). Carbon nanotubes introduced into the abdominal cavity of mice show asbestos-like
pathogenicity in a pilot study. Nat Nanotechnol 3: 423-428. http://dx.doi.org/10.1038/nnano.20Q8.111
Ponti. J: Colognato. R: Rauscher. H: Gioria. S: Broggi F: Franchini. F: Pascual C: Giudetti. G: Rossi. F. (2010).
Colony Forming Efficiency and microscopy analysis of multi-wall carbon nanotubes cell interaction. Toxicol
Lett 197: 29-37. http://dx.doi.0rg/10.1016/i.toxlet.2010.04.018
Porter. DW: Hubbs. AF: Mercer. RR: Wu. N: Wolfarth. MG: Sriram. K: Leonard. S: Battelli. L: Schwegler-
Berry. D: Friend. S: Andrew. M: Chen. BT: Tsuruoka. S: Endo. M: Castranova. V. (2010). Mouse pulmonary
dose- and time course-responses induced by exposure to multi-walled carbon nanotubes. Toxicology 269:
136-147. http://dx.doi.0rg/10.1016/i.tox.2009.10.017
R-18
-------
Posner. S. (2004). Survey and technical assessment of alternatives to decabromodiphenyl ether (decaBDE) in
textile applications. Sundbyberg, Sweden: Swedish Chemicals Inspectorate.
Priest S. (2012). Opinion: No objections to nano? Scientist.
Pulskamp. K: Diabate. S: Krug. HF. (2007). Carbon nanotubes show no sign of acute toxicity but induce
intracellular reactive oxygen species in dependence on contaminants. Toxicol Lett 168: 58-74.
http://dx.doi.0rg/10.1016/i.toxlet.2006.ll.001
Pure Strategies Inc. (2005). Decabromodiphenylether: An investigation of non-halogen substitutes in electronic
enclosure and textile applications. Lowell, MA: Lowell Center for Sustainable Production.
Oin. X: Xia. X: Yang. Z: Yaa S: Zhao. Y: Wei. R: Li. Y: Tian. M: Zhao. X: Oin. Z: Xu. X. (2010). Thyroid
disruption by technical decabromodiphenyl ether (DE-83R) at low concentrations in Xenopus laevis. J
Environ Sci 22: 744-751.
Radomskj A: Jurasz. P: Alonso-Escalano. D: Drews. M: Morandi. M: Malinski. T: Radomskj MW. (2005).
Nanoparticle-induced platelet aggregation and vascular thrombosis. Br JPharmacol 146: 882-893.
Rafique. MMA: Iqbal. J. (2011). Production of carbon nanotubes by different routes A review. UN 1: 29-34.
http://dx.doi.org/10.4236/ieas.2011.12004
Rahman. F: Langford. KH: Scrimshaw. MD: Lester. JN. (2001). Polybrominated diphenyl ether (PBDE) flame
retardants [Review]. Sci Total Environ 275: 1-17. http://dx.doi.org/10.1016/80048-9697(0D00852-X
Ricklund. N: Kierkegaard. A: McLachlan. MS: Wahlberg. C. (2009). Mass balance of decabromodiphenyl
ethane and decabromodiphenyl ether in a WWTP. Chemosphere 74: 389-394.
http://dx.doi.0rg/10.1016/i.chemosphere.2008.09.054
Riva. C: Binelli. A: Cognj D: Provini. A. (2007). Evaluation of DNA damage induced by decabromodiphenyl
ether (BDE-209) in hemocytes of Dreissena polymorpha using the comet and micronucleus assays. Environ
Mol Mutagen 48: 735-743. http://dx.doi.org/10.1002/em.20353
Rocha. JD: Bachilo. SM: Ghosh. S: Arepalli. S: Weisman. RB. (2011). Efficient spectrofluorimetric analysis of
single-walled carbon nanotube samples. Anal Chem 83: 7431-7437. http://dx.doi.org/10.1021/ac2014788
Rose. M: Bennett. DH: Bergman. A: Fangstrom. B: Pessah. IN: Hertz-Picciotto. I. (2010). PBDEs in 2-5 year-
old children from California and associations with diet and indoor environment. Environ Sci Technol 44:
2648-2653. http://dx.doi.org/10.1021/es903240g
Rossi. M: Tickner. J: Geiser. K. (2006). Alternatives assessment framework of the Lowell Center for Sustainable
Production. Lowell, MA: University of Massachusetts Lowell.
http://www.chemicalspolicv.org/downloads/FinalAltsAssess06.pdf
RTI International. (2012). Nanomaterial case study workshop process: Identifying and prioritizing research for
multiwalled carbon nanotubes: Summary report-final. Research Triangle Park, NC: U.S. Environmental
Protection Agency.
Ryman-Rasmussen. JP: Cesta. MF: Brodv. AR: Shipley-Phillips. JK: Everitt JL Tewksbury. EW: Moss. OR:
Wong. BA: Dodd. DE: Andersen. ME: Bonner. JC. (2009a). Inhaled carbon nanotubes reach the subpleural
tissue in mice. Nat Nanotechnol 4: 747-751. http://dx.doi.org/10.1038/nnano.2009.305
Ryman-Rasmussen. JP: Tewksbury. EW: Moss. OR: Cesta. MF: Wong. BA: Bonner. JC. (2009b). Inhaled
multiwalled carbon nanotubes potentiate airway fibrosis in murine allergic asthma. Am J Respir Cell Mol
Biol 40: 349-358. http://dx.doi.org/10.1165/rcmb.2008-0276OC
Sacks. JD: Stanek. LW: Luben. TJ: Johns. DO: Buckley. BJ: Brown. JS: Ross. M. (2011). Particulate-matter
induced health effects: Who is susceptible? [Review]. Environ Health Perspect 119: 446-454.
http://dx.doi.org/10.1289/ehp. 1002255
Sadrieh. N: Wokovich. AM: Gopee. NV: Siitonen. PH: Cozart. CR: Howard. PC: Doub. WH: Buhse. LF. (2008).
Analysis of dermal penetration of titanium dioxide (TiO2) from sunscreen formulations containing micro-
and nano-scale particles of TiO2 [Abstract]. Toxicologist 102: 210.
R-19
-------
Saeed. K. (2010). Review on properties, dispersion and toxicology of carbon nanotubes. J Chem Soc Pakistan
32: 559-564.
SAFENANO. (2012). Textiles and carbon nanotubes: Helping business with risk, regulation, and responsibility.
Edinburgh: Institute of Occupational Medicine, Europe's Centre of Excellence on Nanotechnology Hazard
and Risk. http://www.safenano.Org/Portals/3/SN Content/Documents/DOWNLOADS/LCA%20-
%20Textiles_v2011 Europe.pdf
Sager. TM: Castranova. V. (2009). Surface area of particle administered versus mass in determining the
pulmonary toxicity of ultrafine and fine carbon black: Comparison to ultrafine titanium dioxide. Part Fibre
Toxicol 6:15. http://dx.doi.org/10.1186/1743-8977-6-15
Sakamoto. Y: Nakae. D: Fukumori. N: Tavama. K: Maekawa. A: Imai. K: Hirose. A: Nishimura. T: Ohashi. N:
Ogata. A. (2009). Induction of mesothelioma by a single intrascrotal administration of multi-wall carbon
nanotube in intact male Fischer 344 rats. J Toxicol Sci 34: 65-76.
Sass. J. (2007). Nanotechnology's invisible threat: Small science, big consequences. Natural Resources Defense
Council, http://www.nrdc.org/health/science/nano/nano.pdf
Schartel. B: Bartholmai. M: Knoll. U. (2006). Some comments on the main fire retardancy mechanisms in
polymer nanocomposites. Polym Advan Technol 17: 772-777'. http://dx.doi.org/10.1002/pat.792
Schierz. A: Parks. AN: Washburn. KM: Chandler. GT: Ferguson. PL. (2012). Characterization and quantitative
analysis of single-walled carbon nanotubes in the aquatic environment using near-infrared fluorescence
spectroscopy. Environ Sci Technol 46: 12262-12271. http://dx.doi.org/10.1021/es301856a
Schinwald. A: Murphy. F: Prina-Mello. A: Poland. C: Byrne. F: Glass. J: Dickerson. J: Schultz. D: Movia. D:
Jeffree. C: Macnee. W: Donaldson. K. (2012). The threshold length for fibre-induced acute pleural
inflammation: Shedding light on the early events in asbestos-induced mesothelioma. Toxicol Sci 128: 461-
470. http://dx.doi.org/10.1093/toxsci/kfsl71
Schnorr. JM: Swager. TM. (2011). Emerging Applications of Carbon Nanotubes. Chem Mater 23: 646-657.
http://dx.doi.org/10.1021/cml02406h
Schubauer-Berigan. MK: Dahm. MM: Yencken. MS. (2011). Engineered carbonaceous nanomaterials
manufacturers in the United States. J Occup Environ Med 53: S62-S67.
http://dx.doi.org/10.1097/JOM.Ob013e31821ble2c
Schulte. PA: Murashov. V: Zumwalde. R: Kuempel. ED: Geraci. CL. (2010). Occupational exposure limits for
nanomaterials: state of the art. J Nanopart Res 12: 1971-1987. http://dx.doi.org/10.1007/sll051-010-0008-l
Schwyzer. I: Kaegi R: Sigg. L: Smaida. R: Magrez. A: Nowack. B. (2012). Long-term colloidal stability of 10
carbon nanotube types in the absence/presence of humic acid and calcium. Environ Pollut 169: 64-73.
http://dx.doi.0rg/10.1016/i.envpol.2012.05.004
Shen. M: Xia. X: Wang. F: Zhang. P: Zhao. X. (2012). Influences of multiwalled carbon nanotubes and plant
residue chars on bioaccumulation of polycyclic aromatic hydrocarbons by Chironomus plumosus larvae in
sediment. Environ Toxicol Chem 31: 202-209. http://dx.doi.org/10.1002/etc.722
Shi. X: von Dem Bussche. A: Hurt. RH: Kane. AB: Gao. H. (2011). Cell entry of one-dimensional nanomaterials
occurs by tip recognition and rotation [Letter]. Nat Nanotechnol 6: 714-719.
http://dx.doi.org/10.1038/nnano.2011.151
Shih. YH: Wang. CK. (2009). Photolytic degradation of polybromodiphenyl ethers under UV-lamp and solar
irradiations. J Hazard Mater 165: 34-38. http://dx.doi.org/10.1016/i.jhazmat.2008.09.103
Shvedova. AA: Castranova. V: Kisin. ER: Schwegler-Berry. D: Murray. AR: Gandelsman. VZ: Maynard. A:
Baron. P. (2003). Exposure to carbon nanotube material: assessment of nanotube cytotoxicity using human
keratinocyte cells. J Toxicol Environ Health A 66: 1909-1926. http://dx.doi.org/10.1080/713853956
Siddiqi MA: Laessig. RH: Reed. KD. (2003). Polybrominated dihenyl ethers (PBDEs): New pollutants- old
diseases. ClinMedRes 1: 281-290.
R-20
-------
Siegfried. B. (2007). NanoTextiles: Functions, nanoparticles and commercial applications. Dtibendorf,
Switzerland: Empa Materials Science & Technology.
Siegrist M: Nowack. B: Kastenholz. H. (2011). Environmental considerations of and societal reactions to
nanotechnology in the food sector. In F Kampers (Ed.), Nanotechnology in the Agri-Food Sector:
Implications for the Future (pp. 209-223): Wiley-VCH Verlag GmbH & Co.
http://publicationslist.org/data/nowack/ref-l 13/Siegrist%20(2011).pdf
Sigma-Aldrich. (2012). Nanotubes and fullerenes. http://www.sigmaaldrich.com/materials-
science/nanomaterials/nanotubes-fullerenes.html
Singh. N: Manshian. B: Jenkins. GJ: Griffiths. SM: Williams. PM: Maffeis. TG: Wright. CJ: Doak. SH. (2009).
NanoGenotoxicology: the DNA damaging potential of engineered nanomaterials [Review]. Biomaterials 30:
3891-3914. http://dx.doi.0rg/10.1016/j.biomaterials.2009.04.009
Sjodin. A: Carlsson. H: Thuresson. K: Sjolin. S: Bergman. A: Ostman. C. (2001). Flame retardants in indoor air
at an electronics recycling plant and at other work environments. Environ Sci Technol 35: 448-454.
http://dx.doi.org/10.1021/es000077n
Smallev. RE: Smith. KA: Colbert. DT: Nikolaev. P: Bronikowski. MJ: Bradley. RK: Rohmund. F. (2007).
Single-wall carbon nanotubes from high pressure CO. (U.S. Patent No. 7,204,970). Washington, DC: U.S.
Patent and Trademark Office.
Smith. CJ: Shaw. BJ: Handy. RD. (2007). Toxicity of single walled carbon nanotubes to rainbow trout,
(Oncorhynchus mykiss): respiratory toxicity, organ pathologies, and other physiological effects. Aquat
Toxicol 82: 94-109. http://dx.doi.0rg/10.1016/j.aquatox.2007.02.003
Sobek. A: Bucheli. TD. (2009). Testing the resistance of single- and multi-walled carbon nanotubes to
chemothermal oxidation used to isolate soots from environmental samples. Environ Pollut 157: 1065-1071.
http://dx.doi.0rg/10.1016/i.envpol.2008.09.004
Soderstrom. G: Sellstrom. U: De wit CA: Tysklind. M. (2004). Photolytic denomination of decabromodiphenyl
ether (BDE 209). Environ Sci Technol 38: 127-132. http://dx.doi.org/10.1021/es034682c
Som. C: Wick. P: Krug. H: Nowack. B. (2011). Environmental and health effects of nanomaterials in
nanotextiles and facade coatings [Review]. Environ Int 37: 1131-1142.
http://dx.doi.0rg/10.1016/i.envint.2011.02.013
Song. M: Chu. S: Letcher. RJ: Seth. R. (2006). Fate, partitioning, and mass loading of polybrominated diphenyl
ethers (PBDEs) during the treatment processing of municipal sewage. Environ Sci Technol 40: 6241-6246.
Song. P: Yu. Y: Zhang. T: Fu. S: Fang. Z: Wu. 0. (2012). Permeability, Viscoelasticity, and Flammability
Performances and Their Relationship to Polymer Nanocomposites. Ind Eng Chem Res 51: 7255-7263.
http://dx.doi.org/10.1021/ie300311a
Stella. GM. (2011). Carbon nanotubes and pleural damage: Perspectives of nanosafety in the light of asbestos
experience [Review]. Biointerphases 6: P1-P17. http://dx.doi.0rg/10.1116/l.3582324
Su. Y: Hung. H: Brice. KA: Su. K: Alexandrou. N: Blanchard. P: Chan. E: Sverko. E: Fellin. P. (2009). Air
concentrations of polybrominated diphenyl ethers (PBDEs) in 20022004 at a rural site in the Great Lakes.
Atmos Environ43: 6230-6237. http://dx.doi.0rg/10.1016/i.atmosenv.2009.08.034
Sui YC: Acosta. PR: Gonzalez-Leon. JA: Bermudez. A: Feuchtwanger. J: Cui BZ: Flores. JO: Saniger. JM.
(2001). Structure, Thermal Stability, and Deformation of Multibranched Carbon Nanotubes Synthesized by
CVD in the AAO Template. J Phys ChemB 105: 1523-1527. http://dx.doi.org/10.1021/ip002408o
Sullivan. FA. (2009). Carbon nanotubes: Endless opportunities. Sullivan, FA.
http://www.frost.com/prod/servlet/press-release.pag?docid=159277406.
Szendi K: Varga. C. (2008). Lack of genotoxicity of carbon nanotubes in a pilot study. Anticancer Res 28: 349-
352.
R-21
-------
Takagi A: Kitaiima. S: Kanno. J: Hirose. A: Nishimura. T: Fukumori. N: Ogata. A: Ohashj N. (2008). Induction
of mesothelioma in p53+/- mouse by intraperitoneal application of multi-wall carbon nanotube. J Toxicol Sci
33: 105-116.
Takava. M: Ono-Ogasawara. M: Shinohara. Y: Kubota. H: Tsuruoka. S: Koda. S. (2012). Evaluation of exposure
risk in the weaving process of MWCNT-coated yarn with real-time particle concentration measurements and
characterization of dust particles. Ind Health 50: 147-155. http://dx.doi.org/10.2486/indhealth.MS1312
Takava. M: Serita. F: Ono-Ogasawara. M: Shinohara. Y: Saito. H: Koda. S. (2010). [Airborne particles in a
multi-wall carbon nanotube production plant: observation of particle emission and personal exposure 1:
Measurement in the packing process]. Sangyo EiseigakuZasshi52: 182-188.
Tan. XM: Fugetsu. B. (2007). Multi-walled carbon nanotubes interact with cultured rice cells: Evidence of a
serf-defense response. Journal of Biomedical Nanotechnology 3: 285-288.
http://dx.doi.org/10.1166/ibn.2007.035
Tian. F: Cui. D: Schwarz. H: Estrada. GG: Kobavashi. H. (2006). Cytotoxicity of single-wall carbon nanotubes
on human fibroblasts. Toxicol In Vitro 20: 1202-1212.
Tian. X: Li. T: Yang. K: Xu. Y: Lu. H: Lin. D. (2012). Effect of humic acids on physicochemical property and
Cd(II) sorption of multiwalled carbon nanotubes. Chemosphere 89: 1316-1322.
http://dx.doi.0rg/10.1016/i.chemosphere.2012.05.082
Tomy. GT: Palace. VP: Halldorson. T: Braekevelt E: Danell R: Wautier. K: Evans. B: Brinkworth. L: Fisk. AT.
(2004). Bioaccumulation, biotransformation, and biochemical effects of brominated diphenyl ethers in
juvenile Lake Trout ( Salvelinus namaycush). Environ Sci Technol 38: 1496-1504.
http://dx.doi.org/10.1021/es035070v
Toth. A: Torocsik. A: Tombacz. E: Olah. E: Heggen. M: Li. C: Klumpp. E: Geissler. E: Laszlo. K. (2011).
Interaction of phenol and dopamine with commercial MWCNTs. J Colloid Interface Sci 364: 469-475.
http://dx.doi.0rg/10.1016/i.icis.2011.08.044
Tran. CL: Buchanan. D: Cullen. RT: Searl. A: Jones. AD: Donaldson. K. (2000). Inhalation of poorly soluble
particles II. Influence of particle surface area on inflammation and clearance. Inhal Toxicol 12: 1113-1126.
http://dx.doi.org/10.1080/08958370050166796
Tsai. SJ: Hofmann. M: Hallock. M: Ada. E: Kong. J: Ellenbecker. M. (2009). Characterization and evaluation of
nanoparticle release during the synthesis of single-walled and multiwalled carbon nanotubes by chemical
vapor deposition. Environ Sci Technol 43: 6017-6023.
Tseng. LH: Lee. CW: Pan. MH: Tsai. SS: Li MH: Chen. JR: Lav. JJ: Hsu. PC. (2006). Postnatal exposure of the
male mouse to 2,2',3,3',4,4',5,5',6,6'-decabrominated diphenyl ether: Decreased epididymal sperm functions
without alterations in DNA content and histology in testis. Toxicology 224: 33-43.
http://dx.doi.0rg/10.1016/i.tox.2006.04.003
U.S. EPA (U.S. Environmental Protection Agency). (2005). Furniture flame retardancy partnership:
Environmental profiles of chemical flame-retardant alternatives for low-density polyurethane foam: Volume
1 [EPA Report]. (EPA-742/R-05-002A). http://www.epa.gov/dfe/pubs/flameret/altrep-vl/altrepvl-flc.pdf
U.S. EPA (U.S. Environmental Protection Agency). (2006). Life cycle assessment: Principles and practice [EPA
Report]. (EPA/600/R-06/060). Cincinnati, OH.
U.S. EPA (U.S. Environmental Protection Agency). (2007). Nanotechnology white paper [EPA Report]. (EPA
100/B-07/001). Washington, DC. http://www.epa.gov/osa/pdfs/nanotech/epa-nanotechnologv-whitepaper-
0207.pdf
U.S. EPA (U.S. Environmental Protection Agency). (2008a). Child-specific exposure factors handbook [EPA
Report]. (EPA/600/R-06/096F). Washington, DC.
http ://cfpub. epa. gov/ncea/cfm/recordisplav. cfm?deid= 199243
U.S. EPA (U.S. Environmental Protection Agency). (2008b). Toxicological review of decabromodiphenyl ether
(CAS No. 1163-19-5) [EPA Report]. (635-R-07-008F).
R-22
-------
U.S. EPA (U.S. Environmental Protection Agency). (2008c). TSCA inventory status of nanoscale substances
general approach [EPA Report], http://www.epa.gov/oppt/nano/nmsp-inventorvpaper2008.pdf
U.S. EPA (U.S. Environmental Protection Agency). (2009). Nanomaterial research strategy (final report) [EPA
Report]. (EPA/620/K-09/011). Washington, DC.
http://www.epa.gov/nanoscience/files/nanotech research strategy final.pdf
U.S. EPA (U.S. Environmental Protection Agency). (2010a). DecaBDE phase-out initiative [EPA Report].
http://www.epa.gov/oppt/existingchemicals/pubs/actionplans/deccadbe.html
U.S. EPA (U.S. Environmental Protection Agency). (2010b). An exposure assessment of polybrominated
diphenyl ethers [EPA Report]. (EPA/600/R-08/086F). Washington, DC.
http ://cfpub. epa. gov/ncea/cfm/recordisplav. cfm?deid=210404
U.S. EPA (U.S. Environmental Protection Agency). (2010c). IRIS site help and tools: Glossary, acronyms and
abbreviations. Available online at http://www.epa.gov/ncea/iris/help gloss.htm (accessed August 23, 2010).
U.S. EPA (U.S. Environmental Protection Agency). (2010d). Nanomaterial case studies: Nanoscale titanium
dioxide in water treatment and in topical sunscreen (final) [EPA Report]. (EPA/600/R-09/057F). Research
Triangle Park, NC. http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=230972
U.S. EPA (U.S. Environmental Protection Agency). (2010e). Nanomaterial case study: Nanoscale silver in
disinfectant spray (external review draft) [EPA Report]. (EPA/600/R-10/081). Washington, DC.
U.S. EPA (U.S. Environmental Protection Agency). (20101). New chemicals program meetings and review
process [EPA Report], http://www.epa.gov/opptintr/newchems/pubs/process.htm
U.S. EPA (U.S. Environmental Protection Agency). (2010g). What information must be submitted about new
chemical substances? [EPA Report], http://www.epa.gov/oppt/newchems/pubs/whatinfo.htm
U.S. EPA (U.S. Environmental Protection Agency). (2010h). Workshop summary for the EPA Board of
Scientific Counselors [EPA Report]. (EPA/600/R-10/042). Research Triangle Park, NC.
U.S. EPA (U.S. Environmental Protection Agency). (201 la). Comprehensive environmental assessment: A
meta-assessment approach to increase effectiveness of risk management and research planning [Fact Sheet].
Research Triangle Park, NC: http://www.epa.gov/nanoscience/files/CEAPrecis.pdf. U.S. Environmental
Protection Agency. http://www.epa.gov/nanoscience/files/CEAPrecis.pdf
U.S. EPA (U.S. Environmental Protection Agency). (201 Ib). Exposure factors handbook 2011 edition (final)
[EPA Report]. (EPA/600/R-09/052F). http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=236252
U.S. EPA (U.S. Environmental Protection Agency). (2012a). Certain polybrominated diphenylethers; Significant
new use rule and test rule [EPA Report]. http://www.gpo.gov/fdsvs/pkg/FR-2012-04-02/pdf/2012-7195.pdf
U.S. EPA (U.S. Environmental Protection Agency). (2012b). Nanomaterial case study: a comparison of
multiwalled carbon nanotube and decabromodiphenyl ether flame-retardant coatings applied to upholstery
textiles (draft) [EPA Report]. (EPA/600/R-12/043A). RTF, NC.
U.S. EPA (U.S. Environmental Protection Agency). (2012c). Polybrominated diphenyl ethers (PBDEs) action
plan summary [EPA Report], http://www.epa.gov/oppt/existingchemicals/pubs/actionplans/pbde.html
U.S.GAP (U.S. Government Accountability Office ). (2013). Chemical assessments: An agencywide strategy
may help EPA address unmet needs for integrated risk information system assessments. In Report to the
Chairman, Committee on Environment and Public Works, US Senate. (GAO-13-369).
http ://www. gao. gov/products/GAO-13-369
Uddin. N: Nyden. M. (201 la). Characterization of nanoparticle release from burning polymer nanocomposites.
NSTI-Nanotech 2011, Boston, MA.
Uddin. NM: Nyden. MR. (201 Ib). Characterizing particle release from polymer nanocomposites. Presentation
presented at Nanotech Conference & Expo 2011, June 13-16, 2011, Boston, MA.
R-23
-------
Ullah. S: Ahmad. F. (2012). Enhancing the char resistant of expandable graphite based intumescent fire retardant
coatings by using multi-wall carbon nano tubes for structural steel. Solid State Phenomena 185: 90-93.
http://dx.doi.org/10.4028/www.scientific.net/SSP.185.90
VanLeeuwen. C: Biddinger. G: Gess. D: Moore. D: Natan. T: Winkelmann. D. (1998). Problem formulation. In
G Biddinger (Ed.), Ecological Risk Assessment Decision-Support System: A Conceptual Design (pp. 7-14).
Vankoningsloo. S: Piret JP: Saout C: Noel. F: Mejia. J: Zouboulis. CC: Delhalle. J: Lucas. S: Toussaint O.
(2010). Cytotoxicity of multi-walled carbon nanotubes in three skin cellular models: effects of sonication,
dispersive agents and corneous layer of reconstructed epidermis. Nanotoxicology 4: 84-97.
http://dx.doi.org/10.3109/17435390903428869
Varga. C: Szendi K. (2010). Carbon nanotubes induce granulomas but not mesotheliomas. In Vivo 24: 153-156.
Velzeboer. I: Kupryianchyk. D: Peeters. ETH. M: Koelmans. AA. (2011). Community effects of carbon
nanotubes in aquatic sediments. Environ Int 37: 1126-1130. http://dx.doi.0rg/10.1016/i.envint.2011.01.016
Verdejo. R: Barroso-Buians. F: Rodriguez-Perez. MA: Antonio de saia. J: Arroyo. M: Lopez-Manchado. MA.
(2008). Carbon nanotubes provide serf-extinguishing grade to silicone-based foams. J Mater Chem 18: 3933.
http://dx.doi.org/10.1039/b805943h
Viberg. H: Fredriksson. A: Eriksson. P. (2007). Changes in spontaneous behaviour and altered response to
nicotine in the adult rat, after neonatal exposure to the brominated flame retardant, decabrominated diphenyl
ether (PBDE 209). Neurotoxicology 28: 136-142. http://dx.doi.0rg/10.1016/j.neuro.2006.08.006
Viberg. H: Fredriksson. A: Jakobsson. E: Orn. U: Eriksson. P. (2003). Neurobehavioral derangements in adult
mice receiving decabrominated diphenyl ether (PBDE 209) during a defined period of neonatal brain
development. Toxicol Sci 76: 112-120. http://dx.doi.org/10.1093/toxsci/kfg210
Viberg. H: Mundv. W: Eriksson. P. (2008). Neonatal exposure to decabrominated diphenyl ether (PBDE 209)
results in changes in BDNF, CaMKII and GAP-43, biochemical substrates of neuronal survival, growth, and
synaptogenesis. Neurotoxicology 29: 152-159. http://dx.doi.0rg/10.1016/i.neuro.2007.10.007
von der Kammer. F: Ferguson. PL: Holden. PA: Masion. A: Rogers. KR: Klaine. SJ: Koelmans. AA: Home. N:
Unrine. JM. (2012). Analysis of engineered nanomaterials in complex matrices (environment and biota):
general considerations and conceptual case studies [Review]. Environ Toxicol Chem 31:32-49.
http://dx.doi.org/10.1002/etc.723
Vonderheide. AP: Mueller. KE: Meija. J: Welsh. GL. (2008). Polybrominated diphenyl ethers: Causes for
concern and knowledge gaps regarding environmental distribution, fate and toxicity [Review]. Sci Total
Environ 400: 425-436. http://dx.doi.0rg/10.1016/i.scitotenv.2008.05.003
Wako. K: Kotani. Y: Hirose. A: Doi T: Hamada. S. (2010). Effects of preparation methods for multi-wall
carbon nanotube (MWCNT) suspensions on MWCNT induced rat pulmonary toxicity. J Toxicol Sci 35: 437-
446.
Wang. H: Yang. ST: Cao. A: Liu. Y. (2013). Quantification of carbon nanomaterials in vivo. Ace Chem Res 46:
750-760. http://dx.doi.org/10.1021/ar200335i
Wang. X: Lu. J: Xing. B. (2008). Sorption of organic contaminants by carbon nanotubes: Influence of adsorbed
organic matter. Environ Sci Technol 42: 3207-3212. http://dx.doi.org/10.1021/es702971g
Wang. X: Shu. L: Wang. Y: Xu. B: Bai. Y: Tao. S: Xing. B. (2011). Sorption of peat humic acids to multi-
walled carbon nanotubes. Environ Sci Technol 45: 9276-9283. http://dx.doi.org/10.1021/es202258q
Wang. Y: Kim. JH: Baek. JB: Miller. GW: Pennell. KD. (2012). Transport behavior of functionalized multi-wall
carbon nanotubes in water-saturated quartz sand as a function of tube length. Water Res 46: 4521-4531.
http://dx.doi.0rg/10.1016/i.watres.2012.05.036
Warheit. DB: Laurence. BR: Reed. KL: Roach. DH: Reynolds. GAM: Webb. TR. (2004). Comparative
pulmonary toxicity assessment of single-wall carbon nanotubes in rats. Toxicol Sci 77: 117-125.
Watanabe. I: Sakai. S. (2003). Environmental release and behavior of brominated flame retardants [Review].
Environ Int 29: 665-682.
R-24
-------
Wei. L: Thakkar. M: Chen. Y: Ntim. SA: Mitra. S: Zhang. X. (2010). Cytotoxicity effects of water dispersible
oxidized multiwalled carbon nanotubes on marine alga, Dunaliella tertiolecta. Aquat Toxicol 100: 194-201.
http://dx.doi.0rg/10.1016/i.aquatox.2010.07.001
WHO. (1999). Hazard prevention and control in the work environment: Airborne dust (pp. 1-96).
(WHO/SDE/OEH/99.14). WHO.
http://www.who.int/occupational health/publications/en/oehairbornedust3.pdf
Wirnitzer. U: Herbold. B: Voetz. M: Ragot J. (2009). Studies on the in vitro genotoxicity of baytubes,
agglomerates of engineered multi-walled carbon-nanotubes (MWCNT). Toxicol Lett 186: 160-165.
http://dx.doi.0rg/10.1016/i.toxlet.2008.ll.024
Wohlleben. W: Brill. S: Meier. MW: Mertler. M: Cox. G: Hirth. S: von Vacano. B: Strauss. V: Treumann. S:
Wiench. K: Ma-Hock. L: Landsiedel R. (2011). On the lifecycle of nanocomposites: Comparing released
fragments and their in-vivo hazards from three release mechanisms and four nanocomposites. Small 7: 2384-
2395. http://dx.doi.org/10.1002/smll.201002054
Wohlleben. W: Meier. MW: Vogel. S: Landsiedel. R: Cox. G: Hirth. S: Tomovic. Z. (2013). Elastic CNT-
polyurethane nanocomposite: synthesis, performance and assessment of fragments released during use.
Nanoscale 5: 369-380. http://dx.doi.org/10.1039/c2nr32711b
Wright HE: Zhang. Q: Mihelcic. JR. (2008). Integrating economic inputoutput life cycle assessment with risk
assessment for a screening-level analysis. Int J Life Cycle Assess 13: 412-420.
http://dx.doi.org/10.1007/sll367-008-0006-z
Wu. M: Gordon. RE: Herbert. R: Padilla. M: Moline. J: Mendelson. D: Litle. V: Travis. WD: Gil. J. (2010). Case
report: Lung disease in World Trade Center responders exposed to dust and smoke: carbon nanotubes found
in the lungs of World Trade Center patients and dust samples. Environ Health Perspect 118: 499-504.
http://dx.doi.org/10.1289/ehp.0901159
Wu. Y: Hudson. JS: Lu. Q: Moore. JM: Mount AS: Rao. AM: Alexov. E: Ke. PC. (2006). Coating single-walled
carbon nanotubes with phospholipids [Letter]. J Phys Chem B 110: 2475-2478.
http://dx.doi.org/10.1021/ipQ57252c
Xu. G: Wu. WT: Wang. Y: Pang. W: Zhu. Q: Wang. P. (2007). Functionalized carbon nanotubes with
polystyrene-block-poly (N-isopropylacrylamide) by in situ RAFT polymerization. Nanotechnology 18:
145606. http://dx.doi.org/10.1088/0957-4484/18/14/145606
Xusen. (2010). Non-Halogenated smoke-suppressing fire retardants: Products [Fact Sheet].
http://www.xusen.com/pages/products01 1 en.html
Yang. L: Kim. P: Meyer. HM: Agnihotri. S. (2009). Aging of nanocarbons in ambient conditions: probable
metastability of carbon nanotubes. J Colloid Interface Sci 338: 128-134.
http://dx.doi.0rg/10.1016/i.icis.2009.06.017
Yu. H: Liu. J: Wang. Z: Jiang. Z: Tang. T. (2009). Combination of Carbon Nanotubes with Ni2O3 for
Simultaneously Improving the Flame Retardancy and Mechanical Properties of Polyethylene. J Phys Chem C
113: 13092-13097. http://dx.doi.org/10.1021/jp902081e
Yu. H: Liu. J: Wen. X: Jiang. Z: Wang. Y: Wang. L: Zheng. J: Fu. S: Tang. T. (2011). Charing polymer wrapped
carbon nanotubes for simultaneously improving the flame retardancy and mechanical properties of epoxy
resin. Polymer (Guildf) 52: 4891-4898. http://dx.doi.0rg/http://dx.doi.org/10.1016/i.polymer.2011.08.013
Zhang. L: Petersen. EJ: Huang. Q. (2011). Phase distribution of 14C-labeled multiwalled carbon nanotubes in
aqueous systems containing model solids: Peat. Environ Sci Technol 45: 1356-1362.
http://dx.doi.org/10.1021/esl026097
Zhang. L: Petersen. EJ: Zhang. W: Chen. Y: Cabrera. M: Huang. Q. (2012a). Interactions of 14C-labeled multi-
walled carbon nanotubes with soil minerals in water. Environ Pollut 166: 75-81.
http://dx.doi.0rg/10.1016/i.envpol.2012.03.008
Zhang. S: Horrocks. AR. (2003). A review of flame retardant polypropylene fibres. Progress in Polymer Science
28: 1517-1538. http://dx.doi.0rg/10.1016/i.progpolvmsci.2003.09.001
R-25
-------
Zhang. T: Du. Z: Zou. W: Li. H: Zhang, C. (2012b). The flame retardancy of blob-like multi-walled carbon
nanotubes/silica nanospheres hybrids in poly (methyl methacrylate). PolymDegrad Stabil 97: 1716-1723.
http://dx.doi.0rg/10.1016/i.polymdegradstab.2012.06.014
Zhou. FL: Gong. RH. (2008). Manufacturing technologies of polymeric nanofibres and nanofibre yarns. Polymer
International 57: 837-845. http://dx.doi.org/10.1002/pi.2395
Zhu. SM: Li. JQ: Pang. JH: Zhang. HS: Liang. ER. (2011). Stability of multi-wall carbon nanotubes in air
[Abstract]. Carbon49: 354-355. http://dx.doi.0rg/10.1016/i.carbon.2010.08.022
Zhu. Y: Zhao. Q: Li. Y: Cai X: Li W. (2006). The interaction and toxicity of multi-walled carbon nanotubes
with Stylonychia mytilus. J Nanosci Nanotechnol 6: 1357-1364. http://dx.doi.org/10.1166/jnn.2006.194
R-26
-------
Appendix A. Case Study Candidate
Evaluation and Selection Process
Appendix A describes the process of evaluating and selecting carbon-based nanomaterials in
specific applications as candidates on which to focus in the current case study. Specifically, it outlines the
process by which the candidate carbon-based nanomaterial applications were identified and highlights
factors re levant to the suitability of each candidate for a Nanomaterial Case Study. This information was
used to support an informed selection of the nanomaterial and application for this case study by U.S.
Environmental Protection Agency (EPA) program offices, labs, and centers in the Office of Research and
Development, and regional offices during September 2011.
A.1. Background
As discussed in Chapter 1, the EPA (2007) Nanotechnology White Paper called for the use of
nanomaterial case studies and multidisciplinary expert workshops as a means to inform research planning
to support the risk assessment process for nanomaterials. In response to the recommendations of the
Nanotechnology White Paper, EPA has been developing case studies of selected nanomaterials in specific
applications, including Nanoscale Titanium Dioxide in Water Treatment and Topical Sunscreen (U.S.
EPA. 2010) and Nanoscale Silver in Disinfectant Spray (U.S. EPA. 2012). To continue this series of case
studies, five carbon-based nanomaterial applications were identified as candidates for the next
Nanomaterial Case Study (presented in alphabetical order by material and application):
• Carbon nanofibers in cement/concrete,
• Carbon nanotubes (multiwalled) in flame-retardant coatings and composites,
• Carbon nanotubes (multiwalled) in rubber tires,
• Carbon nanotubes (single-walled) in textiles, and
• Nanocrystalline cellulose in biodegradable packaging.
This appendix provides further detail on the process by which the candidate carbon-based
nanomaterial applications were identified, briefly summarizes the state of the science for each of the
identified candidates, and highlights factors re levant to the suitability of each candidate for a
Nanomaterial Case Study. The following criteria were used as guides in judging the candidates:
• "Nano-ness" of the material (i.e., whether the material is intentionally engineered at the
nanoscale and has properties that distinguish it from conventional forms of the material);
A-1
-------
• Potential for exposure throughout the product life cycle (in humans, both occupational and
general public, as well as in other biota);
• Availability of data (whether directly related or inferred from other materials/products);
• Feasibility of comparing the nanoenabled application to a non-nanoenabled application; and
• Relevance to EPA programs.
Despite the wealth of information on carbon-based nanomaterials, EPA had previously
encountered difficulties in identifying an appropriate carbon-based nanomaterial application as a
candidate for a case study. These difficulties arose in part due to the following factors:
• Few data are available on the actual commercial use of carbon-based nanomaterials in
products.
• Little to no information is available on release of carbon-based nanomaterials from
applications during normal use.
• Small loadings (<5% by weight or volume) of carbon-based nanomaterials are generally
required to confer desirable properties.
• Compared to other types of nanomaterials, the applications for carbon-based nanomaterials
appear to have a smaller exposure potential outside of occupational and disposal scenarios.
Although these challenges are still present, the recent increase in carbon-based nanomaterial
research has produced new data, and new applications have been proposed that offer greater potential for
widespread exposure. As a result, the selection of a carbon-based nanomaterial application for the next in
the Nanomaterial Case Study series appears to be feasible.
A.2. Candidate Identification Process
The process by which the five nanomaterial application candidates were identified is summarized
in Figure A-l. As noted in Section 1.1, the process began with a systematic approach to the identification
of nanomaterials and applications for investigation. This preliminary phase of the process utilized
strategic literature and Internet searches to identify supporting scientific literature, relevant news reports,
and nanomaterial information aggregation websites. Certain basic literature search statistics (e.g., number
of total hits, number of hits in scientific databases) were then evaluated to provide a preliminary metric of
interest within the nanotechnology community for each candidate nanomaterial/application combination
and general data availability. After the "long list" of material/application combinations was identified and
narrowed down using this systematic approach, a more judgment-based approach was used to evaluate the
suitability of a "medium list" of potential candidates and identify the most feasible candidates for a case
A-2
-------
study. In this phase of the process, a list of suitability questions was used to step through the life cycle of
a specific nanoenabled product and evaluate the characteristics of that product that might affect release,
exposure, environmental fate, and impact on humans, ecological receptors, and the environment.
A professional judgment as to whether a case study feasibly could be conducted for each candidate then
was determined based on the answers to the suitability questions, and a "short list" of five feasible
candidates was developed based on professional judgment of suitability.
A-3
-------
PROCESS
RESULT
General literature search: Scientific
literature, manufacturer websites,
nanotechnology information aggregation
websites, expert opinion.
Approximately 30 carbon-based
nanomaterials grouped into 5 main
categories: carbon nanofibers, carbon
nanotubes, fullerenes, graphene, and
cellulose nanomaterials.
Broad literature search: Variations of
search terms from five nanomaterial
categories in combination with the terms
"use," "application," and "product."
"Long list": General and specific
applications of carbon-based
nanomaterials in each of the five
nanomaterial categories.
411 database searches: DIALOG "All
Science" database universe searched using
general nanomaterial category/general
application combinations.
"Medium list": Nanomaterial
category/genera I application combinations
(e.g., CNTs in electronics) that received a
high number of hits in the DIALOG
databases.
Specific literature searches and high-level
review: EBSCO search and review of
scientific literature published since 2009
for each nanomaterial category/general
application combination.
Specific associations: Specific applications
associated with specific materials (e.g.,
SWCNTs in conductive films) in each
general category were selected for further
investigation.
Cursory review of available information:
Scientific literature, Web resources, expert
consultation on specific
nanomaterial/application combinations.
Exposure potential and data availability
assessed using life-cycle questionnaire.
"Shortlist": Specific
nanomaterial/specific application
combinations that offer sufficient
potential for exposure and adequate
overall data availability.
Identification of non-nano material for
comparison: Review of literature and
Web resources for non-nanoenabled
products for comparison to nanoenabled
applications.
Final candidate list
Figure A-1. Candidate identification process.
A-4
-------
A.3. Nanomaterial Application Candidates
The candidate identification process yielded five nanomaterial applications for consideration as
the subject of anew case study. The nanomaterial application candidates and proposed non-nanoenabled
products for comparison are presented in Table A-l below. As discussed in Chapter 1, the non-
nanoenabled product is intended to provide a frame of reference against which the ecological, human
health, and other implications of the selected nanomaterial application can be compared.
Table A-1. Nanomaterial application candidates and non-nanoenabled products for comparison in
a case study.
Nanomaterial
Application
Proposed Non-Nanoenabled Product for
Comparison
Carbon nanofiber (CNF)
Cement/concrete (CNF-
reinforced)
Steel- or glass-fiber reinforced cement/concrete
Multiwalled carbon
nanotubes (MWCNTs)
Flame-retardant coatings and
composites
Polybrominated diphenyl ether (PBDE) flame-retardant
materials
MWCNT
Rubber tires (MWCNT as
filler)
Carbon black and silica filler in tires
Single-walled carbon
nanotubes (SWCNTs)
Textiles (SWCNT-
reinforced/impregnated)
Traditional textiles reinforced with carbon fiber polymer
composites
Nanocrystalline cellulose
(NCC)
Biodegradable packaging
(NCC-polymer composite)
Packaging containing polylactic acid (PLA) resin alone
A.4. Summary of Factors Affecting Suitability
Key considerations of the suitability of each candidate for development into a case study were
summarized in a suitability chart (Table A-2). Based on the initial findings of this analysis (which should
not be considered exhaustive or comprehensive), four of five applications appeared to still be in the
research stage and unavailable for the commercial market in the United States or internationally.
Although carbon nanotubes have been incorporated into military textiles, carbon nanotube flame-retardant
coating is the only application identified as being currently available on the consumer market.
A-5
-------
Table A-2. Factors for consideration in selecting a candidate for case study.
Selection Factors
Candidates
0)
£
o
c
o
o
0)
O
LL
•z.
0
*- w
C 0)
•o 55
re Q.
o3 £
K o
^
re jg
l- oi
Z c
O *j
§ 0
^ 0
0)
jZ
0)
£1
3
1-
"2.
O
§
^
w
.2
•5
0)
1-
1-
2
O
^
(0
0)
f*
re
re
O)
0) m
•c ?
!§•!
CQ re
O re
"Z. Q.
1 = High D = Medium D = Low
Nanomaterial production volume
Nanomaterial manufacturing processes
Nanomaterial characterization
Nanomaterial release from application
Nanomaterial fate and transport in the environment
Human exposure to nanomaterial
Ecological exposure to nanomaterial
Human health effects of nanomaterial
Ecological effects of nanomaterial
Nanoproduct data (i.e., nanomaterial application data)
Conventional (i.e., non-nano) product data for comparison
•
•
n
n
n
n
n
n
D
•
D
n
D
D
D
•
D
D
•
D
n
D
D
D
•
D
n
•
D
D
n
D
D
•
D
n
D
D
n
n
n
n
n
D
n
£1
re
re
Q
Data Availability Qualifiers.
High (•) = The database appears to contain a large quantity of good quality, diverse literature on the topic.
Medium ( ) = The database contains some literature specific to the topic, but substantial data gaps were observed.
Low (D) = Little to no information was identified on the topic.
O = No
> = Yes
O = Possibly
Is there presumptive potential for exposure: During manufacturing?
During use?
At end of life?
Will extrapolation from other applications/materials be necessary?
Is the nanomaterial currently produced in quantities >5 tons/year?
Is material production expected to increase in the near future?
Will the nanomaterial be used in the product at levels >5% by weight?
Is the product already on the market?
Is there evidence that the nanomaterial is hazardous: To humans?
To eco receptors?
O
•
0
•
•
•
o
o
•
o
o
o
0
•
•
•
o
•
•
o
o
o
0
•
•
•
•
o
•
o
o
o
0
o
o
o
o
o
•
o
o
•
0
o
0
•
o
o
o
0
o
0)
£1
re
.*;
=
V)
Answers to Suitability Questions.
Yes (•) = Initial findings suggest that an affirmative answer can be given with relatively high confidence.
Possibly (O) = Initial findings from the preliminary literature review were conflicting.
No (0) = Initial findings suggest that a negative answer can be given with relatively high confidence.
A-6
-------
A.5. U.S. EPA Program Involvement in Final Selection
As noted in Chapter 1. the selection of which of the five nano-carbon product candidates to use in
this case study document involved representatives from EPA program offices, labs and centers within the
Office of Research and Development, and regional offices. To facilitate distributing and discussing
information relevant to the selection, an internal online forum was developed using a commercially
available product (www.IdeaScale.com). The forum included brief introductory material on this
nanomaterial case study series, instructions on using the website to nominate nano-carbon product
candidates, links to tables summarizing life-cycle information on five candidate nano-carbon products
(i.e., the applications in Table A-2), and a more detailed report summarizing the state of the science for
each candidate.
A link to the forum was sent to EPA representatives along with a request to share the link with
colleagues in their organization. Representatives and others in the Agency could then use the forum to
discuss the candidates informally and nominate candidates for selection. Representatives were asked to
submit a formal vote that reflected input from their colleagues through the IdeaScale forum and other
communication channels they wished to use, as well as consideration of their own knowledge, the
information provided on the forum, and the consideration of the criteria listed in Section A.I.
The candidates receiving the most votes were SWCNTs in textiles and MWCNTs in flame-
retardant coatings and composites; thus, a hybrid option (MWCNTs in flame-retardant coatings applied to
textiles) was selected. This choice reflected comments that, although the textile application was
preferable, MWCNTs were perhaps of greater interest based on indications that they will contain a higher
level of contaminants and are currently more widely produced.
Appendix A References
U.S. EPA (U.S. Environmental Protection Agency). (2007). Nanotechnology white paper [EPA Report]. (EPA
100/B-07/001). Washington, DC. http://www.epa.gov/osa/pdfs/nanotech/epa-nanotechnology-whitepaper-
0207.pdf
U.S. EPA (U.S. Environmental Protection Agency). (2010). Nanomaterial case studies: Nanoscale titanium
dioxide in water treatment and in topical sunscreen (final) [EPA Report]. (EPA/600/R-09/057F). Research
Triangle Park, NC. http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=230972
U.S. EPA (U.S. Environmental Protection Agency). (2012). Nanomaterial case study: Nanoscale silver in
disinfectant spray (final report) [EPA Report]. (EPA/600/R-10/081F). Washington, DC.
A-7
-------
Appendix B. Analytical Methods for
Detecting, Measuring, and Characterizing
BDE-209 and Multiwalled
Carbon Nanotubes
This appendix provides a brief overview of some of the available techniques to detect, quantify,
and characterize polybrominated diphenyl ethers (PBDEs; specifically BDE-209 [decaBDE]) and
multiwalled carbon nanotubes (MWCNTs) in laboratory, biological, and environmental settings, along
with current challenges to making such measurements. This information is not intended to be exhaustive
in reporting every applicable method and associated challenges or to be comprehensive in describing
available methods; rather, it is a summary of relatively common or known methods for characterizing
BDE-209 and MWCNTs based on information available at the time this case study was developed.
B.1. Measuring and Characterizing PBDEs and MWCNTs
Accurately measuring BDE-209 or MWCNTs in relevant biological or environmental media is
critical for evaluating any potential impacts of either material on human health, ecological populations, or
environmental resources (Alcock et al.. 2011; Lehman et al.. 2011). The choice of which measurement
technique to use for either BDE-209 or MWCNT samples will ultimately involve a consideration of trade-
offs related to cost, time, selectivity, and sensitivity (Alcock et al., 2011; Lehman et al.. 2011; Stapleton.
2006). In evaluating which analytical technique(s) to use, having an understanding of the challenges
related to quantifying and characterizing BDE-209 and MWCNTs is useful. For both materials, multiple
techniques might be required to characterize all of the physicochemical properties of interest in a single
sample (e.g., molecular composition, purity, shape, surface charge) (Alcock et al.. 2011; Lehman et al.,
2011). Moreover, the training of personnel and the capital cost associated with some of these tools can
impede the analysis of materials by multiple laboratories (Alcock et al.. 2011). In addition, the
standardization and validation of methods, availability of material standards, and the consistent reporting
of material characteristics in peer-reviewed literature have proven difficult for both BDE-209 and
MWCNTs (Alcock et al.. 2011; Lehman et al.. 2011). Distinguishing the sample from background
concentrations or other materials of similar composition (e.g., nonaBDE congeners versus decaBDE,
B-1
-------
single-walled CNTs versus MWCNTs) is also a challenge for both materials (Lehman etal.. 2011;
Stapleton. 2006).
For BDE-209, gas chromatography coupled with mass spectrometry is the most prevalently used
method; however, several variations in this approach exist (Stapleton. 2006). Even small differences in
analytical techniques can result in significant variation in results (Alcock et al.. 2011; Stapleton. 2006).
Recent efforts to standardize methods and develop techniques that minimize material degradation have
improved interlaboratory variation, but continue to be the subject of study, particularly for measuring
samples in complex milieus (Stapleton. 2006). Detection and characterization of BDE-209 has proven
more difficult than lower brominated compounds due in part to degradation at high temperatures and with
ultraviolet light exposure (Stapleton. 2006).
For MWCNTs, transmission electron microscopy is generally used to characterize structural
properties of the material, which is the first step in differentiating between MWCNTs, single-walled
carbon nanotubes (SWCNTs), or other materials; however, using this tool in tandem with others is
necessary to characterize the material (Lehman et al.. 2011) more completely. Challenges related to
characterizing and quantifying MWCNTs include their propensity to agglomerate or otherwise transform
(e.g., surface oxidize) during the process of production, purification, or exposure, as well as interference
from experimental artifacts (e.g., metal catalysts used in material production) (Petersen and Henry. 2012;
Lehman etal.. 2011). The challenge of combining multiple techniques (e.g., gas chromatography and
mass spectrometry) for BDE-209 analyses is amplified for MWCNTs in that a multitude of measurements
and sampling techniques are generally required to fully characterize nanomaterials (Lehman et al.. 2011).
B.2. Summary Tables
The tables below highlight techniques for detecting, measuring, and characterizing PBDEs and
MWCNTs. Table B-l briefly outlines advantages and disadvantages of individual approaches to gas
chromatography and spectrometry, as well as a few alternative techniques that are available for studying
PBDEs. Table B-2 provides a brief overview of available methods to characterize a range of MWCNT
properties. More detail on each approach can be found in the references listed at the end of this appendix,
particularly the recent review by Lehman et al. (2011).
B-2
-------
Table B-1. Analytical techniques for detecting, measuring, and characterizing PBDEs.
Citation(s)
Stapleton (2006)
Stapleton (2006)
Stapleton (2006)
La Guardia et al.
(2006)
Stapleton (2006)
La Guardia et al.
(2006)
Stapleton (2006)
Stapleton (2006)
Stapleton (2006)
Technique3
Atmospheric pressure
photoionization (APPI)-
coupled liquid
chromatography
(LC)/MS-MSb'c
Gas chromatography
(GC)/electron capture
detection0
GC/electron capture
negative ionization
mass spectrometry
(ECNI-MS)a'b'c
GC/electron ionization
(El) MSa'b'c
GC/high resolution
mass spectrometry
(HRMS)b'c
GC/HR time of flight
(TOP) MSa'b'c
On-column injection
GCa'c
Application(s)
• Determination of
congener ratios in
environmental and
biological media
• Determination of
congener ratios in
environmental media
• Determination of
congener ratios in
environmental and
biological media
• Determination of
congener ratios in
environmental and
biological media
• Determination of
congener ratios in
environmental and
biological media
• Determination of
congener ratios in
environmental media
• Separation and
detection of PBDE
congeners
• Determination of
molecular weight
Advantages
• Relatively soft
ionization technique
compared to
electrospray
ionization
• Inexpensive
• Ability to detect
halogenated organic
compounds
• Low limit of detection
• Selectivity
• Selectivity
• Sensitivity
• Can detect relatively
high molecular weight
analytes
• Spectral data can be
obtained over a wide
mass range with little
sacrifice in sensitivity
• Necessary equipment
commonly present in
laboratories
• Precise
discrimination,
particularly of BDE-
209
Disadvantages
• Limited
chromatographi
c resolution
relative to gas
chromatograph
y
• Relative
imprecision
compared to
GC/ECNI
• Selectivity
• Interference
may occur with
methoxylated
PBDEs
• Necessary
equipment is
not commonly
found in
laboratories
• Expensive
• Expensive
• Low sample
concentration
required for
accuracy
• Small injection
volume
• Sample must
be free of
impurities
B-3
-------
Table B-1 (Continued): Analytical techniques for detecting, measuring, and characterizing PBDEs.
Citation(s)
Stapleton (2006)
Technique3
Programmable
temperature
vaporization (PTV)
injection GCb'c'd
Application(s)
• Separation and
detection of PBDE
congeners
• Determination of
molecular weight
Advantages
• Relatively large
injection volume
compared to other
GC techniques
• Separation can be
performed on human
serum
Disadvantages
• Requires
significant
optimization to
perform
separation
Stapleton (2006) Split/splitless injection
gas chromatography
(GC)M
• Separation and
detection of PBDE
congeners
• Determination of
molecular weight
• Can be used on
environmental
samples
• Necessary equipment
commonly present in
laboratories
• Injection
volume must be
small
High injection
temperature
techniques listed in alphabetical order.
bUsed for chemical (in vitro) analysis as reported in reference document.
°Used to analyze in vivo samples as reported in reference document.
dUsed to analyze environmental samples as reported in reference document.
Table B-2. Analytical techniques for detecting, measuring, and characterizing MWCNTs.
Citation(s)
Technique3
Application(s)
Advantages
Disadvantages
Petersen and
Henry (2012)
Lehman et al.
(2011)
Johnston et al.
(2010)
Petersen and
Henry (2012)
Petersen and
Henry (2012)
Petersen and
Henry (2012)
Atomic force
microscopy13
Centrifugationb
Chemothermal
oxidationd
(at 375 °C)
Cryotransmission
electron
microscopy
(CEM)b
• Size (diameter and
length) and shape
• Dispersion in solution
• Length distribution
• Size of nanoparticle
aggregates
• Measurement of
MWCNT concentration
in environmental
samples
• Properties of MWCNTs
in aqueous phase
• Centrifugation
equipment
commonly present in
laboratories
• Allows for
quantitative
determination of
MWCNT
concentration
_
• Limited to samples in
aqueous phase
• Dispersion difficult;
requires extensive
sonication
• Accuracy may be
affected by dispersion
• Inaccurate
• Limited to samples in
aqueous phase
B-4
-------
Table B-2 (Continued): Analytical techniques for detecting, measuring, and characterizing
MWCNTs.
Citation(s)
Petersen and
Henry (2012)
Lehman et al.
(2011)
Petersen and
Henry (2012)
Lehman et al.
(2011)
Lehman et al.
(2011)
Johnston et al.
(2010)
Petersen and
Henry (2012)
Ayrault (2000)
Petersen and
Henry (2012)
Revel and
Ayrault (2000)
Petersen and
Henry (2012)
Johnston et al.
(2010)
Technique3
Fluorescence
microscopy
(FLM)c'd
Fourier transform
infrared
spectroscopy
(FTIR)b
Dynamic light
scattering (DLS)b
Gas pycnometryb
ICP mass
spectrometry
(ICP-MS)c'd
Instrumental
neutron activation
analysis'3^
Light
microscopyb'c'd
Application(s)
• Detection of MWCNTs
in environmental media
or tissue from biological
specimens
• Nanoparticle
functionalization
• CNT orientation (parallel
or perpendicular to
beam)
• Size of aggregates in
aqueous phase
• Density
• Metal concentration in
environmental samples
• Metal concentrations
• Identification of large
MWCNT aggregates
Advantages
Can detect single
MWCNT
Reliable detector of
carboxylic acids
• Useful for detecting
changes in MWCNT
size at various
points during
synthesis or
experimentation
• Can be used to
determine both bulk
and skeletal
densities
• Can be used to
study health effects
of MWCNT
exposure by
detecting changes in
protein expression
or structure
• Faster than
instrumental neutron
activation analysis
• Can be more
accurate than ICP-
MS
• Necessary
equipment is
common in
laboratories
Disadvantages
• Necessary equipment
is not common in
laboratories
• Sample preparation
may result in water
contamination or
altered surface
functionalization
• Estimation of size by
DLS based on
spherical molecular
structure and cannot
be used for absolute
calculation of
aggregate size
• Sample must be
powder
• Necessary equipment
is not commonly
found in laboratories
• Safety risk associated
with radioactivity
• Lower sample
throughput than ICP-
MS
• Requires equipment
not commonly found
in laboratories
• Technique provides
qualitative, non-
specific information
• Detection limited to
large aggregates
B-5
-------
Table B-2 (Continued): Analytical techniques for detecting, measuring, and characterizing
MWCNTs.
Citation(s)
Technique3
Application(s)
Advantages
Disadvantages
Lehman et al.
(2011)
N2 gas
adsorption13
Surface area
determination
• Permanent
quadrupole inhibits N2
adsorption to some
substrates
• Model of N2
adsorption based on
homogeneity across
adsorption surface
• May not be suitable
for characterization of
ecotoxicity
Lehman et al.
(2011)
Johnston et al.
(2010)
Lehman et al.
(2011)
Petersen and
Henry (2012)
Johnston et al.
(2010)
Petersen and
Henry (2012)
Johnston et al.
(2010)
Lehman et al.
(2011)
Johnston et al.
(2010)
Lehman et al.
(2011)
Petersen and
Henry (2012)
Nitrogen and
phosphorous
dopingb'c
Optical density
(UV-vis
absorbance)b'c'd
Radioactive
labelingc'd
Raman
spectroscopyb'c'd
Scanning electron
microscopy
(SEM)b
Thermal optical
transmittanced
• Structural defects
• MWCNT concentration
in solution
• Detection of MWCNTs
in environmental media
• Analysis of MWNCNT
purity
• Detection of defects in
MWNCT structure
• Tube alignment
• Tube diameter
• Surface morphology
• Surface purity
• Loss of mass at various
temperatures
• Can detect non-
carbon atoms
present in MWCNTs
• Can differentiate
between pentagonal
and hexagonal
structure
• Necessary
equipment is
common in
laboratories
• Quantitative
• Versatile (can be
used in many forms
of environmental
media)
• Relatively high
resolution
information about
structure
• Repeatable
• Samples can contain
dissolved
environmental
material
• Interaction between
donor molecules and
pentagonal and
hexagonal structures
on nanoparticle
surface have not yet
been quantified
• Results sensitive to
presence of other
compounds in
solution
• Expensive
• Inherent danger of
radioactivity
• Complex
interpretation of
spectra for MWCNTs
• Does not provide
information on
internal morphology
• Only useful for
MWCNTs in aqueous
phase
• Necessary equipment
is uncommon in
laboratories
B-6
-------
Table B-2 (Continued): Analytical techniques for detecting, measuring, and characterizing
MWCNTs.
Citation(s)
Lehman et al.
(2011)
Lehman et al.
(2011)
Petersen and
Henry (2012)
Johnston et al.
(2010)
Technique3 Application(s) Advantages
Thermogravimetric • MWCNT purity analysis • Necessary
analysis (TGA)b equipment is
common in
laboratories
Transmission • Surface morphology • Provides high
Electron . CrystaNinity resolution
Microscopy information about
(TEM)b'Cl nanotube structure
Disadvantages
• Sample size
requirements may be
large for certain
applications (3-10
mg)
• Multiple
measurements
needed to ensure
accuracy of data
• Difficult sample
preparation;
preparation may
damage sample
• requires expert
personnel
• Images susceptible to
excessive beam
exposure
• Difficult to analyze
large volumes in
timely manner
Lehman et al. X-ray diffraction
(2011)
Skeletal density
Cannot be used to
determine bulk
density
Lehman et al. X-ray
(2011)
Echlin (1998)
microanalysisb'c
• Purity
• Narrowing of
incident beam allows
greater resolution
Lehman et al.
(2011)
Petersen and
Henry (2012)
X-ray
photoelectron
spectroscopy
(XPS)b
• Surface chemical
composition
• Presence of functional
groups
• May be inaccurate
without fluorine
tagging
techniques listed in alphabetical order.
bUsed for chemical (in vitro) analysis as reported in reference document.
°Used to analyze in vivo samples as reported in reference document.
dUsed to analyze environmental samples as reported in reference document.
Appendix B References
Alcock. RE: MacGillivrav. BH: Busby. JS. (2011). Understanding the mismatch between the demands of risk
assessment and practice of scientists—the case of Deca-BDE [Review]. Environ Int 37: 216-225.
http://dx.doi.0rg/10.1016/i.envint.2010.06.002
B-7
-------
Echlin. P. (1998). Low-voltage energy-dispersive X-ray microanalysis of bulk biological materials. Microsc
Microanal 4: 577-584. http://dx.doi.org/10.1017/S1431927698980552
Johnston. HJ: Hutchison. GR: Christensen. FM: Peters. S: Hankin. S: Aschberger. K: Stone. V. (2010). A critical
review of the biological mechanisms underlying the in vivo and in vitro toxicity of carbon nanotubes: The
contribution of physico-chemical characteristics [Review]. Nanotoxicology 4: 207-246.
http://dx.doi.org/10.3109/17435390903569639
La Guardia. MJ: Hale. RC: Harvey. E. (2006). Detailed polybrominated diphenyl ether (PBDE) congener
composition of the widely used penta-, octa-, and deca-PBDE technical flame-retardant mixtures. Environ
Sci Technol 40: 6247-6254. http://dx.doi.org/10.1021/es060630m
Lehman. JH: Terrones. M: Mansfield. E: Hurst. KE: Meunier. V. (2011). Evaluating the characteristics of
multiwall carbon nanotubes. 49: 2581-2602. http://dx.doi.0rg/10.1016/j.carbon.2011.03.028
Petersen. EJ: Henry. TB. (2012). Methodological considerations for testing the ecotoxicity of carbon nanotubes
and fullerenes: Review [Review]. Environ Toxicol Chem 31: 60-72. http://dx.doi.org/10.1002/etc.710
Revel G: Ayrault S. (2000). Comparative use of INAA and ICP-MS methods for environmental studies. Journal
of Radioanal Chem 244: 73-80. http://dx.doi.Org/10.1023/A:1006723022439
Stapleton. HM. (2006). Instrumental methods and challenges in quantifying polybrominated diphenyl ethers in
environmental extracts: A review [Review]. Anal Bioanal Chem 386: 807-817.
http://dx.doi.org/10.1007/s00216-006-0400-v
B-8
-------
Appendix C. Efficacy of Various Methods
of Carbon Nanotube Purification
Appendix C provides a comparative overview of various purifications methods for carbon
nanotubes reported in the literature, and as summarized by Hou et al. (2008). This information is
organized to demonstrate the relative effectiveness of each method at removing each of the specified
carbonaceous or metallic impurities. Each method is ranked as effective, partially effective, or not
effective (/= effective; O= partially effective; *= not effective) for removing each impurity specified in
Table C-l below.
C-1
-------
Table C-1. Purification methods for carbon nanotubes.
Carbonaceous impurities
Yield
Purification methods (wt%)
Chemical A;,. o QC
.. Air -2-35
methods
CI2, H20, HCI -15
Gas phasp
H20, Ar, 02 -30
O2, C2H2F4, SF6 25-48
HNO3 -30-50
H2O2, HCI 10-75
Mixture of acid or KMnO4 30-75
Microwave in inorganic acid 10-60
Electro- A|kg|j Qf gcjd so|ution ^80
chemical
Amorphous
carbon / CNT
(carbon
nanoparticle)
CT
D
D
D
D
D
D
0
0
Graphite
Particles
0
0
0
0
0
0
o
0
0
Carbon
impurities
adhering to
CNT walls
O
0
0
o
0
0
0
0
0
Soluble
carbon in
some organic
solutions
(e.g., CS2,
toluene)
0
0
0
0
0
0
0
0
0
Metallic impurities
Metal
wrapped by Metal
Exposed polyhedral encapsulated
metal carbon in CNTs
.
.
.
D D
D 0
D 0
D D
D 0
• 0
o
0
0
0
0
DO
C-2
-------
Table C-1 (Continued): Purification methods for carbon nanotubes.
Carbonaceous impurities
Purification methods
Physical
methods
Multistep
methods
Filtration
Centrifugation
Solubilization with functional groups
High temperature annealing
Other physical techniques to remove metal
particles
Chromatography, electrophoresis, FFFb
HIDE0, wet grinding, filtration, oxidation,
sonication, centrifugation
Filtration/magnetic filtration, oxidation, annealing
Sonication in H2O2, HNO3/HF/SDS, filtration
High temperature annealing extraction
Yield
(wt%)
-30-84
-10-40
-17-50
-70-90
-10-NR
NR
-2%
-9-20%
-25
-90%
Amorphous
carbon / CNT
(carbon
nanoparticle)
O
0
0
0
0
n
n
n
n
0
Graphite
Particles
O
0
0
0
0
n
n
0
0
0
Carbon
impurities
adhering to
CNT walls
0
0
0
0
0
0
0
n
0
n
Soluble
carbon in
some organic
solutions
(e.g., CS2,
toluene)
n
n
n
0
0
0
n
n
n
n
Metallic impurities
Exposed
metal
O
0
0
n
n
0
n
n
n
n
Metal
wrapped by Metal
polyhedral encapsulated
carbon in CNTs
O
0
0
n
n
0
0
n
0
n
0
0
0
n
n
0
0
0
0
n
aD= effective (•= effective with further HCI treatment); O= partially removed (•= partially removed with further HCI treatment); 0= not effective
bField-flow fractionation.
°Hydrothermally initiated dynamic extraction.
Source: Reprinted (adapted) with permission of Elsevier; [Hou et al. (2008)1.
C-3
-------
Appendix C References
Hou. PX: Liu. C: Cheng. HM. (2008). Purification of carbon nanotubes. Carbon 46: 2003-2025.
http://dx.doi.0rg/10.1016/i.carbon.2008.09.009
C-4
-------
Appendix D. Study Summaries on the
Transport, Transformation, and Fate of
BDE-209 and MWCNTs in Environmental
Systems
The following tables provide details from studies related to fate, transport, or transformation of
decabromodiphenyl ether [(decaBDE) specifically the single isomer of decaBDE, BDE-209] or
multiwalled carbon nanotubes (MWCNTs) in environmental media. Study information that provides
support for the transformation (debromination) of BDE-209 in environmental media is included in Table
D-l. Table D-2 provides study summaries related to the behavior of MWCNTs in aqueous media. Table
D-3 and Table D-4 present information from studies conducted in terrestrial ecosystems for BDE-209 and
MWCNTs, respectively.
Table D-1. Relevant studies of transformation (debromination) of BDE-209.
Citation Relevant Study Information3
Biotic Debromination
Deng et al. • Used aerobic bacterium Lysinibacillus fusiformis strain DB-1 to determine its capability to
(2011) debrominate BDE-209 in sediments of the Lianjiang River, China; DB-1 is an indigenous
bacterium in PBDE-contaminated sediments.
• DB-1 efficiently transformed BDE-209 to lower-brominated BDEs using lactate, pyruvate, and
acetate as carbon sources, and this debromination was an aerobic (oxygen-consuming)
process.
He et al. (2006) • Studied degradation of BDE-209 in various microbial cultures.
• Debromination of BDE-209 was observed with anaerobic bacteria including Sulfurospirillum
multivorans and Dehaloccoides species.
Huang et al. • Studied behavior of BDE-209 in soil-plant system; transportation of BDE-209 within plants was
(2010) examined using six plant species.
• OH-metabolized and debrominated products of BDE-209 were measured in plants and soil;
higher proportions of penta- through diBDE congeners in plant tissues than in the soil suggest
either further debromination of PBDEs within plants or lower PBDEs are more readily taken up
by plants; significant negative correlation between residual BDE-209 concentration and soil
microbial biomass, suggesting microbial metabolism and degradation of BDE-209.
D-1
-------
Table D-1 (Continued): Relevant studies of transformation (debromination) of BDE-209.
Citation
Relevant Study Information3
Tokarz et al.
(2008)
Studied reductive debromination in an anaerobic sediment microcosm experiment.
BDE-209 debrominated slowly; its half-life ranged from 6 to 50 years, and averaged about 14
years; formation of nona-, octa-, hepta-, and hexaBDEs and 9 new congeners, including lower
brominated congeners that are greater environmental concern (more bioavailable);
experiments suggested anaerobic reductive debromination of BDE-209; authors noted
competing influences of hydrophobicity and reactivity that can significantly retard rate of
debromination.
Wang et al. • Examined microbial degradation of BDE-209 in the rhizosphere of ryegrass using arbuscular
(2011) mycorrhizae.
• 12 lower brominated congeners were detected in soil samples and 24 were detected in plant
samples; there was evidence of debromination in soil and within plants based on higher
proportion of di-through hepta-BDEs.
Abiotic Debromination (Photolysis)
Ahn et al. (2006)
Characterized photodegradation of BDE-209 adsorbed on clay, metal oxides, and sediment
using sunlight and artificial UV light.
• Photodegradation rates were likely dependent on the chemical and physical properties of the
sorbent; however, degradation product distribution was not believed to be dependent on
sorbent type.
• Enhanced photolytic transformation rates observed when BDE-209 was adsorbed to clay
minerals; no significant BDE-209 degradation of BDE-209 observed on metal oxides; very slow
degradation rates of BDE-209 sorbed to carbon-rich sediment; results indicated stepwise
(sequential) debromination.
• Suggested that organic matter has inhibitory effect on photodegradation possibly by shielding
BDE-209 from the light or by satisfying excited states of the BDE-209 before they can form
products.
An et al. (2008) • Characterized photolytic activity on BDE-209 degradation using a TiC>2 photocatalyst.
• TiC>2 was an effective photocatalyst for degrading BDE-209; results suggested sequential
debromination occurred in the formation of products during BDE-209 transformation; products
were hexa-, penta, and tetraBDEs.
Bezares-Cruz
(2004)
• Examined BDE-209 photochemical transformation using a hexane solvent and solar light.
• Reaction rate of photolysis was dependent on solar intensity and what the BDE-209 was
adsorbed to; 43 PBDEs were detected, including BDE-47.
Christiansson et • Examined BDE-209 photochemical transformation in the laboratory using various solvents and
al. (2009) UV irradiation.
• BDE-202 was identified as a marker of BDE-209 photolysis; PBDEs accounted for about 90%
of products formed (primarily heptaBDEs to nonaBDEs); Poly brominated dibenzofurans
(PBDFs) accounted for approximately 10% of products formed; authors noted the formation of
BDE-183 and BDE-153 as important environmental congeners.
Hua et al. (2003)
Raff and Hites
(2007)
• Measured photochemical reactions of BDE-209 on artificial surfaces—quartz glass, silica
particles, humic acid-coated silica particles—using UV light and natural sunlight.
• Transformation occurred more slowly using sunlight irradiation; presence of humic acid slowed
transformation; tetraBDE and pentaBDE were not found at detectable levels.
• Examined the role of photolysis in the atmospheric removal of BDE-209.
• Determined that photolysis is minor removal process; removal of particle-bound BDE-209 more
likely due to wet (primarily) and dry deposition.
D-2
-------
Table D-1 (Continued): Relevant studies of transformation (debromination) of BDE-209.
Citation
Relevant Study Information3
Schenker et al. • Used multimedia model that incorporated photolysis to predict fate of BDE-209 in
(2008) environmental compartments.
• Model estimated that about 13% of pentaBDE and 2% of tetraBDE in the environment occurs
from degradation of BDE-209; model-predicted degradation in the atmosphere (e.g., by
photolysis) represents 45% of BDE-209 loss; loss to deposition estimated at 30%.
Shih and Wang
(2009)
Examined solar and UV-lamp degradation of BDE-209.
Observed that photodegradation of BDE-209 was not affected by initial BDE-209
concentrations; photodegradation rate increases with increasing light intensity;
photodegradation of higher brominated congeners faster than for lower brominated congeners;
photodegradation of BDE-209 is a sequential dehalogenation mechanism with stepwise
bromine losses.
Soderstrom et • Studied photodegradation of BDE-209 in toluene, on silica gel, and in sand, soil, sediment
al. (2004) using artificial UV light and natural outdoor sunlight.
• Debromination rates were strongly dependent on matrix type; half-lives were shorter using
artificial matrices (<15 min); longer half-lives observed on more complex natural matrices (40-
200 hours) ; no matrix-related or light intensity-related differences in the debromination pattern
of the BDE congeners formed; formation of lower brominated BDEs (nona-hexaBDEs)
occurred, including BDE-154 and BDE-183; PBDFs were also formed.
Stapleton and
Dodder (2008)
Studied photodegradation of BDE-209 in house dust exposed to natural sunlight.
|nitja| BDE.209 concentration decreased by about 38%, 35% of which was believed to be due
to debromination.
"Additional information obtained from U.S. EPA (2010).
D-3
-------
Table D-2. Relevant studies of MWCNTs in aqueous media.
Citation
Relevant Study Information
Chae et al. • Studied photochemical reactivity of CNT aggregates and compared with other fullerene
(2011) nanoparticles.
• After photosensitization from ultraviolet irradiation, the primary mode of oxidation of chemical
compounds by CNTs is singlet oxygen production; this reactivity appeared to be correlated with
the surface area of colloidal aggregates in solution.
Chappell et al. • Examined the mechanism by which humic substances stabilize MWCNT dispersions in
(2009) aqueous media.
• Adding humic substances to MWCNTs in solution enhanced stability, decreased particle
diameter, and decreased polydispersivity; presence of surfactive domains in the structure of the
humic substances directly impacts CNT dispersal in solution.
Christian et al. • Studied aggregation of nanoparticles and effects of humic acid and cations on CNT stability.
(2008) t Cations, in particular divalent cations (e.g., Ca2+ and Mg2+), were found to reduce the stability of
CNT with or without NOM surface coating.
Desai et al. • Investigated antisolvent precipitation of functionalized MWCNTs and aggregation behavior in
/2012) the aqueous media.
• Organic-soluble MWCNTs functionalized to be hydrophobic by addition of octadecylamine were
shown to form stable dispersions in water/solvent systems even after antisolvent precipitation
via aggregation; stability was shown to be long term, and particle aggregation increased with
the addition of electrolytes.
Han et al. (2008) • Investigated the influence of clay minerals on the stability of surfactant-facilitated MWCNTs.
• Solutions of MWCNTs facilitated by three surfactants reacted differently to addition of two
minerals—kaolinite and montmorillonite; stability of the solutions after mineral addition
depended on the surfactant and the mineral; two mechanisms by which minerals were shown to
affect the stability of MWCNT solution were by mineral adsorption to surfactants and bridging
between mineral and MWCNTs by surfactant.
He et al. (2012) • Studied the behavior of stabilized MWCNTs in a ferric chloride coagulation system and the
structure characteristics of the produced floes.
• MWCNTs stabilized by humic acid were effectively removed from solution by coagulation after
application of relatively large amounts of ferric chloride.
Holbrooket al.
(2010)
Hyung et al.
(2007)
• Examined surface water constituents that affect MWCNT coagulation.
• Higher influent concentrations of kaolin and alginate increased MWCNT removal by
coagulation; higher concentrations of NOM reduced MWCNT removal by coagulation.
• Studied the aqueous stability of MWCNTs in the presence of NOM.
• For the same initial MWCNT concentrations, suspended MWCNT concentrations were
considerably higher in solutions of synthetic modeled Suwannee River NOM and actual river
water than in solutions of sodium dodecyl sulfate, a common surfactant; the mechanism for
CNT-NOM interactions are dependent on the characteristics of the MWCNTs and the NOM.
D-4
-------
Table D-2 (Continued): Relevant studies of MWCNTs in aqueous media.
Citation
Relevant Study Information
Hyung and Kim • Investigated the effect of NOM characteristics and water quality parameters on NOM adsorption
(2008) to MWCNTs.
• Adsorption capacity was directly proportional to NOM aromatic carbon content and the ionic
strength of the solution; adsorption capacity was indirectly proportional to pH; adsorption
strength was indirectly proportional to ionic strength and not significantly changed by pH.
Kennedy et al. • Investigated factors that influence the partitioning of CNTs (raw versus functionalized [either
(2008) engineered or natural]) in the aquatic environment.
• Pure CNTs had limited potential for aqueous transport; instead, aggregation and adsorption to
sediment particles; no ionic strength influences on aggregate size; hydrophobicity of CNTs likely
increased affinity for particles and enhanced aggregation; aqueous destabilization by van der
Waals attractions and rapid sedimentation; sedimentation rate might accelerate with increased
concentration; dispersion was enhanced by surface modifications (engineered or NOM)—
increased residency time in surface water, aggregate size/structure changes.
Kennedy et al. • Examined influence of surface modifications and various dispersal methods on MWCNT fate
(2009) and toxicity.
• Dissolved organic matter, humic acid, and fulvic acid were shown to be dispersing agents of
MWCNTs; humic acid was a more effective dispersant than fulvic acid; sonication treatment of
MWCNTs was shown to increase fragmentation of the particles relative to magnetic stirring;
functionalization and laboratory methods of dispersal of MWCNTs affect their behavior in
aqueous solutions in the presence of NOM.
Kummerer et al. • Investigated biodegradability of functionalized and nonfunctionalized MWCNTs in aqueous
(2011) media.
• MWCNTs were not biodegradable under the conditions tested. Surface modification resulted in
better solubility, but not better biodegradability.
Lin et al. (2009) • Studied the influence of solution pH and ionic strength on the interaction between tannic acid-
facilitated MWCNTs of various diameters.
• Suspension of MWCNTs in tannic acid solution greatly improved with tannic acid concentration
until a plateau concentration was reached; suspension was greatest for particles of 40 nm
diameter, followed by 60 nm, 20 nm, 100 nm, and 10 nm, respectively; MWCNTs stabilized in
tannic acid were stable at pH > 5, and precipitated at pH < 5; presence of ions Na+, Mg2+, Ca2+,
and La3+ caused tannic acid-stabilized MWCNTs to aggregate, in a manner exponentially
correlated to ionic valence.
Lin et al. (2010) • Described stabilities of MWCNTs in forms of particulate aggregates and surfactant-facilitated
suspensions in various fresh surface waters.
• Nonfunctionalized MWCNTs could not stabilize in eight samples of natural surface waters by
shaking, but stabilized in one sample with high dissolved organic content after sonication;
nonfunctionalized MWCNTs did not stabilize in one surface water sample that also had a high
NOM content, suggesting that other characteristics of the sample affected MWCNT
stabilization; MWCNTs stabilized with CTAB surfactant were destabilized in all surface water
samples, TX100- and SDBS- facilitated MWCNT suspensions remained stable in all eight
surface water samples; addition of cations to solutions were shown to destabilize surfactant-
facilitated MWCNT suspensions.
Liu et al. (2009) • Examined the mobility of MWCNTs in porous media using column experiments.
• At low flow rates similar to those found in natural subsurface aqueous environments, OH- and
COOH-functionalized MWCNTs were retained in porous media to a significant extent; at high
flow rates, OH- and COOH-functionalized MWCNTs were very mobile; a medium with a large
number of small pores was shown to retain the functionalized MWCNTs better than media with
fewer wider pores.
D-5
-------
Table D-2 (Continued): Relevant studies of MWCNTs in aqueous media.
Citation
Relevant Study Information
Petersen et al. • Studied uptake in sediment spiked with MWCNTs by sediment-burrowing Lumbriculus
(2008) variegatus.
• Study showed that CNTs did not readily absorb into organism tissues; sizes of MWCNTs could
have been a factor in the lack of absorption by organisms.
Saleh et al. • Examined aggregation kinetics of MWCNTs in aquatic media with varying solution pH and salt
(2008) concentrations and presence of organic matter.
• Increasing monovalent and divalent salt concentration in aqueous solution and increasing
solution pH from acidic to basic in aqueous solution reduced aggregation of sonicated
MWCNTs; addition of humic acid to solution also reduced aggregation rate and enhanced
sonicated MWCNT stability; these results show that sonicated MWCNTs are relatively stable in
solution chemistries with electrolyte and pH levels typical of natural aquatic environments.
Wang et al.
(2009)
• Investigated sorption of humic acid and aromatic compounds by MWCNTs.
• With increasing concentration of humic acid in solution, MWCNTs increasingly sorbed humic
acid until a plateau was reached; maximum humic acid sorption capacity of MWCNTs
depended on TL-TL interactions, surface area of MWCNTs, and dispersion of MWCNTs; sorption
of hydrophobic organic compounds decreased with increasing humic acid concentrations,
suggesting that sorption of hydrophobic organic compounds would be suppressed in the
presence of NOM.
Zhang et al. • Investigated NOM, pH, and ionic strength effects on adsorption of SOCs by MWCNTs in natural
(2010) waters.
• NOM showed a more significant effect on sorption of SOCs by MWCNTs than pH or ionic
strength of solution, which had negligible impacts on SOC sorption; surface functionalization of
MWCNTs with hydroxyl and carboxyl groups slightly suppressed the effects of NOM on SOC
sorption by MWCNTs, and this suppression decreased with increasing hydrophobicity of the
SOC.
Zhang et al. • Examined interactions (phase distribution) between MWCNTs and aqueous systems containing
(2011) peat under various conditions (ionic strength and pH).
• Presence of DOM greatly increased the stability of MWCNTs in aqueous solution in a way
similar to surfactant stabilization; solid peat did not adsorb MWCNTs except with the increasing
concentration of sodium cations.
Abbreviations: CNT = carbon nanotube; MWCNT = multiwalled carbon nanotube; CTAB = cetyl trimethyl ammonium bromide; NOM
= natural organic matter, DOM = dissolved organic matter; SDBS = sodium dodecyl benzene sulfonate; SOC = synthetic organic
chemicals
D-6
-------
Table D-3. Relevant studies of BDE-209 in soils and plants.
Citation
Relevant Study Information
Soils
Li et al. (2010) • Collected wet and dry particle deposition samples at the urban sites of Guangzhou and Hong Kong,
South China.
• Depositional fluxes of BDE-209 ranged from 273 to 6,000 (mean 2,220) ng/m2-day in Guangzhou and
from 29.1 to 1,100 (mean 259) ng/m2-day in Hong Kong.
• BDE-209 was most abundant PBDE congener; distinct seasonal patterns were observed—higher
depositional fluxes during winter; lower fluxes during the summer; seasonal variation associated with
local usage and meteorological factors.
Liu et al. (2011) • Studied the effects of BDE-209 on soil microbial activities and function using soil enzymatic activity
analysis.
• Bacterial counts were suppressed as BDE-209 concentration increased; BDE-209 inhibited microbial
diversity and altered soil microbial community structure.
Yu et al. (2010) • Studied the effects of humic acids (HA) and microorganisms on the migration of BDE-209 in soils
using soil enzymatic activity analysis.
• Distribution of BDE-209 in the colloidal fraction related to the HA and microorganism concentration; HA
acted as surface modifier and microorganisms acted as biosurfactants; BDE-209 transported by soil
colloids along with water currents especially in the presence of HA and microorganisms.
Zhu et al. (2010) • Studied the response of bacterial communities in soils spiked with BDE-209; soil microbial activities
and composition were affected by BDE-209.
• BDE-209, although expected to be of low bioavailability, had an adverse impact on the structure and
function of the soil microbial community and microbial processes; high doses of BDE-209 were toxic,
inhibiting growth for some microorganisms.
Zou et al. (2007) • Investigated the distribution and fate of BDE-209 in soils, Pearl River Delta, China.
• Concentrations of BDE-209 generally decreased with increasing soil depth; BDE-209 in soil was
significantly correlated with total organic carbon levels; sorption of BDE-209 on organic matter
influences its distribution, transportation, and fate in the environment.
D-7
-------
Table D-3 (Continued): Relevant studies of BDE-209 in soils and plants.
Citation
Relevant Study Information
Plants
Huang etal. (2010
• Described the uptake, translocation, and metabolism of BDE-209 in six plant species—ryegrass,
alfalfa, pumpkin, summer squash, maize, and radish.
• Accumulation of BDE-209 occurred in the roots and shoots of all plants.
• Root lipid content was positively correlated with BDE-209 uptake.
• Translocation factor (ConcentrationShoot/Concentrationroot) of BDE-209 was inversely related to BDE-
209 concentration in the roots, suggesting root lipids restrict translocation of BDE-209 from roots to
shoots because of its partitioning to root lipids.
Salamova and Hites
(2010)
Evaluated air samples and tree bark for levels of PBDEs.
• BDE-209 concentration in tree bark was strongly correlated with concentrations of these compounds in
the air and precipitation; highest air and tree bark concentrations occurred at urban sites.
Vrkoslavova et al. • Studied the ability of plants (tobacco and nightshade) to accumulate and translocate PBDEs from
(2010) contaminated sewage sludge.
• BDE-209 was accumulated via roots into tobacco tissue at 116.8 ng/gram dry wt; BDE-209 was not
detected in nightshade; PBDEs detected in aboveground plant biomass provided evidence of
translocation by plants.
Table D-4. Relevant studies of carbon nanotubes (CNTs) in soils.
Citation
Relevant Study Information
Jaisi and
Elimelech (2009)
• Investigated the transport behavior of functionalized single-walled CNTs in columns of natural
soil.
• Single-walled CNT mobility in soils is likely limited because of its irregular shape, large aspect
ratio, and bundled (aggregated) state—these properties would promote soil retention.
• Natural soil environments that are more heterogeneous and contain "open soil structures" could
promote CNT mobility in soil. Dissolved organic molecules in soil pore water could also
enhance the colloidal stability of CNTs and increase their mobility.
Petersen et al. • Studied the effects of modifying 14C-labeled MWCNTs with polyethyleneimine surface
(2011) coatings—making them more stable in solution and modifying surface charges. Tested
MWCNT sorption by soils and uptake and elimination behaviors by earthworms.
• Nearly linear sorption isotherms for regular MWCNTs and nonlinear isotherms for modified
MWCNTs, indicating that the PEI coatings influenced MWCNT interactions with soils; little
difference in sorption results among the different soils tested; soil type might not be as
important as the MWCNT characteristics in predicting soil sorption behaviors.
D-8
-------
Appendix D References
Ahn. M -Y: Filley. TR: JafVert. CT: Nies. L: Hua. I: Bezares-Cruz. J. (2006). Photodegradation of
decabromodiphenyl ether adsorbed onto clay minerals, metal oxides, and sediment. Environ Sci Technol 40:
215-220. http://dx.doi.org/10.1021/es051415t
An. T: Chen. J: Li. G: Ding. X: Sheng. G: Fu. J: Mai. B: O'shea. KE. (2008). Characterization and the
photocatalytic activity of TiO2 immobilized hydrophobic montmorillonite photocatalysts. Catalysis Today
139: 69-76. http://dx.doi.0rg/10.1016/i.cattod.2008.08.024
Bezares-Cruz. J: JafVert. CT: Hua. I. (2004). Solar photodecomposition of decabromodiphenyl ether: products
and quantum yield. Environ Sci Technol 38: 4149-4156.
Chae. SR: Watanabe. Y: Wiesner. MR. (2011). Comparative photochemical reactivity of spherical and tubular
fullerene nanoparticles in water under ultraviolet (UV) irradiation. Water Res 45: 308-314.
http://dx.doi.0rg/10.1016/i.watres.2010.07.067
Chappell MA: George. AJ: Dontsova. KM: Porter. BE: Price. CL: Zhou. P: Morikawa. E: Kennedy. AJ:
Steevens. JA. (2009). Surfactive stabilization of multi-walled carbon nanotube dispersions with dissolved
humic substances. EnvironPollut 157: 1081-1087. http://dx.doi.0rg/10.1016/i.envpol.2008.09.039
Christian. P: Von der Kammer. F: Baalousha. M: Hofmann. T. (2008). Nanoparticles: Structure, properties,
preparation and behaviour in environmental media [Review]. Ecotoxicology 17: 326-343.
http://dx.doi.org/10.1007/sl0646-008-0213-l
Christiansson. A: Eriksson. J: Teclechiel D: Bergman. A. (2009). Identification and quantification of products
formed via photolysis of decabromodiphenyl ether. Environ Sci Pollut Res Int 16: 312-321.
http://dx.doi.org/10.1007/sll356-009-0150-4
Deng. D: Guo. J: Sun. G: Chen. X: Qiu. M: Xu. M. (2011). Aerobic denomination of deca-BDE: Isolation and
characterization of an indigenous isolate from a PBDE contaminated sediment. Int Biodeterior
Biodegradation65: 465-469. http://dx.doi.0rg/10.1016/j.ibiod.2011.01.008
Desai C: Addo Ntim. S: Mitra. S. (2012). Antisolvent precipitation of hydrophobic functionalized multiwall
carbon nanotubes in an aqueous environment. J Colloid Interface Sci 368: 115-120.
http://dx.doi.0rg/10.1016/i.jcis.2011.ll.019
Han. Z: Zhang. F: Lin. D: Xing. B. (2008). Clay Minerals Affect the Stability of Surfactant-Facilitated Carbon
Nanotube Suspensions. Environ Sci Technol 42: 6869-6875. http://dx.doi.org/10.1021/es801150j
He. J: Robrock. KR: Alvarez-cohen. L. (2006). Microbial Reductive Debromination of Polybrominated Diphenyl
Ethers (PBDEs). Environ Sci Technol 40: 4429-4434. http://dx.doi.org/10.1021/es052508d
He. M: Zhou. R: Guo. X. (2012). Behavior of stabilized multiwalled carbon nanotubes in a FeC13 coagulation
system and the structure characteristics of the produced floes. J Colloid Interface Sci 366: 173-178.
http://dx.doi.0rg/10.1016/i.icis.2011.09.059
Holbrook. RD: Kline. CN: Filliben. JJ. (2010). Impact of source water quality on multiwall carbon nanotube
coagulation. Environ Sci Technol 44: 1386-1391. http://dx.doi.org/10.1021/es902946j
Hua. I: Kang. N: JafVert. CT: Fabrega-duque. JR. (2003). Heterogeneous photochemical reactions of
decabromodiphenyl ether. Environ Toxicol Chem 22: 798-804. http://dx.doi.org/10.1002/etc.5620220418
Huang. H: Zhang. S: Christie. P: Wang. S: Xie. M. (2010). Behavior of decabromodiphenyl ether (BDE-209) in
the soil-plant system: Uptake, translocation, and metabolism in plants and dissipation in soil. Environ Sci
Technol 44: 663-667. http://dx.doi.org/10.1021/es901860r
Hyung. H: Fortner. JD: Hughes. JB: Kim. JH. (2007). Natural organic matter stabilizes carbon nanotubes in the
aqueous phase. Environ Sci Technol 41: 179-184. http://dx.doi.org/10.1021/es061817g
D-9
-------
Hyung. H: Kim. JH. (2008). Natural organic matter (NOM) adsorption to multi-walled carbon nanotubes: Effect
of NOM characteristics and water quality parameters. Environ Sci Technol 42: 4416-4421.
http://dx.doi.org/10.1021/es702916h
Jaisi DP: Elimelech. M. (2009). Single-walled carbon nanotubes exhibit limited transport in soil columns.
Environ Sci Technol 43: 9161-9166. http://dx.doi.org/10.1021/es901927y
Kennedy. AJ: Gunter. JC: Chappell MA: Goss. JD: Hull MS: Kirgan. RA: Steevens. JA. (2009). Influence of
nanotube preparation in aquatic bioassays. Environ Toxicol Chem 28: 1930-1938.
http://dx.doi.0rg/10.1897/09-024.l
Kennedy. AJ: Hull MS: Steevens. JA: Dontsova. KM: Chappell MA: Gunter. JC: Weiss. CA. Jr. (2008).
Factors influencing the partitioning and toxicity of nanotubes in the aquatic environment. Environ Toxicol
Chem27: 1932-1941. http://dx.doi.0rg/10.1897/07-624.l
Kummerer. K: Menz. J: Schubert. T: Thielemans. W. (2011). Biodegradability of organic nanoparticles in the
aqueous environment. Chemosphere 82: 1387-1392. http://dx.doi.0rg/10.1016/i.chemosphere.2010.ll.069
Li. J: Zhang. G: Xu. Y: Liu. X: Li. XD. (2010). Dry and wet particle deposition of polybrominated diphenyl
ethers (PBDEs) in Guangzhou and Hong Kong, South China. J Environ Monit 12: 1730-1736.
http://dx.doi.org/10.1039/c001526a
Lin. D: Liu. N: Yang. K: Xing. B: Wu. F. (2010). Different stabilities of multiwalled carbon nanotubes in fresh
surface water samples. EnvironPollut 158: 1270-1274. http://dx.doi.0rg/10.1016/i.envpol.2010.01.020
Lin. D: Liu. N: Yang. K: Zhu. L: Xu. Y: Xing. B. (2009). The effect of ionic strength and pH on the stability of
tannic acid-facilitated carbon nanotube suspensions. Carbon 47: 2875-2882.
http://dx.doi.0rg/10.1016/i.carbon.2009.06.036
Liu. L: Zhu. W: Xiao. L: Yang. L. (2011). Effect of decabromodiphenyl ether (BDE 209) and dibromodiphenyl
ether (BDE 15) on soil microbial activity and bacterial community composition. J Hazard Mater 186: 883-
890. http://dx.doi.0rg/10.1016/i.jhazmat.2010.ll.079
Liu. X: O'Carroll DM: Petersen. EJ: Huang. Q: Anderson. CL. (2009). Mobility of multiwalled carbon
nanotubes in porous media. Environ Sci Technol 43: 8153-8158. http://dx.doi.org/10.1021/es901340d
Petersen. EJ: Huang. Q: Weber. J. r.. W. J. (2008). Ecological uptake and depuration of carbon nanotubes by
Lumbriculus variegatus. Environ Health Perspect 116: 496-500.
Petersen. EJ: Pinto. RA: Zhang. L: Huang. Q: Landrum. PF: Weber. WJ. (2011). Effects of polyethyleneimine-
mediated functionalization of multi-walled carbon nanotubes on earthworm bioaccumulation and sorption by
soils. Environ Sci Technol 45: 3718-3724. http://dx.doi.org/10.1021/esl03004r
Raff. JD: Hites. RA. (2007). Deposition versus photochemical removal of PBDEs from Lake Superior air.
Environ Sci Technol 41: 6725-6731. http://dx.doi.org/10.1021/es070789e
Salamova. A: Hites. RA. (2010). Evaluation of tree bark as a passive atmospheric sampler for flame retardants,
PCBs, and organochlorine pesticides. Environ Sci Technol 44: 6196-6201.
http://dx.doi.org/10.1021/esl01599h
Saleh. NB: Pfefferle. LD: Elimelech. M. (2008). Aggregation kinetics of multiwalled carbon nanotubes in
aquatic systems: Measurements and environmental implications. Environ Sci Technol 42: 7963-7969.
http://dx.doi.org/10.1021/es801251c
Schenker. U: Soltermann. F: Scheringer. M: Hungerbiihler. K. (2008). Modeling the Environmental Fate of
Polybrominated Diphenyl Ethers (PBDEs): The Importance of Photolysis for the Formation of Lighter
PBDEs. Environ Sci Technol 42: 9244-9249. http://dx.doi.org/10.1021/es801042n
Shih. YH: Wang. CK. (2009). Photolytic degradation of polybromodiphenyl ethers under UV-lamp and solar
irradiations. J Hazard Mater 165: 34-38. http://dx.doi.org/10.1016/j.jhazmat.2008.09.103
Soderstrom. G: Sellstrom. U: De wit CA: Tysklind. M. (2004). Photolytic debromination of decabromodiphenyl
ether (BDE 209). Environ Sci Technol 38: 127-132. http://dx.doi.org/10.1021/es034682c
D-10
-------
Stapleton. HM: Dodder. NG. (2008). Photodegradation of decabromodiphenyl ether in house dust by natural
sunlight. Environ Toxicol Chem 27: 306-312. http://dx.doi.Org/10.1897/07-301R.l
Tokarz. JA: Ahn. MY: Leng. J: Fillev. TR: Nies. L. (2008). Reductive debromination of polybrominated
diphenyl ethers in anaerobic sediment and a biomimetic system. Environ Sci Technol 42: 1157-1164.
U.S. EPA (U.S. Environmental Protection Agency). (2010). An exposure assessment of polybrominated
diphenyl ethers [EPA Report]. (EPA/600/R-08/086F). Washington, DC.
http ://cfpub. epa. gov/ncea/cfm/recordisplav. cfm?deid=210404
Vrkoslavova. J: Demnerova. K: Mackova. M: Zemanova. T: Macek. T: Haislova. J: Pulkrabova. J: Hradkova. P:
Stiborova. H. (2010). Absorption and translocation of polybrominated diphenyl ethers (PBDEs) by plants
from contaminated sewage sludge. Chemosphere 81: 381-386.
http://dx.doi.0rg/10.1016/i.chemosphere.2010.07.010
Wang. S: Zhang. S: Huang. H: Christie. P. (2011). Behavior of decabromodiphenyl ether (BDE-209) in soil:
Effects of rhizosphere and mycorrhizal colonization of ryegrass roots. Environ Pollut 159: 749-753.
http://dx.doi.0rg/10.1016/i.envpol.2010.ll.035
Wang. X: Tao. S: Xing. B. (2009). Sorption and competition of aromatic compounds and humic acid on
multiwalled carbon nanotubes. Environ Sci Technol 43: 6214-6219.
Yu. S: Zou. P: Zhu. W: Yang. L: Xiao. L: Jiang. L: Wang. X: Wu. J: Yuan. Y. (2010). Effects of humic acids
and microorganisms on decabromodiphenyl ether, 4,4-dibromodiphenyl ether and anthracene transportation
in soil. Science China Chemistry 53: 950-968. http://dx.doi.org/10.1007/sll426-010-0126-8
Zhang. L: Petersen. EJ: Huang. Q. (2011). Phase distribution of 14C-labeled multiwalled carbon nanotubes in
aqueous systems containing model solids: Peat. Environ Sci Technol 45: 1356-1362.
http://dx.doi.org/10.1021/esl026097
Zhang. S: Shao. T: Bekaroglu. SS: Karanfil T. (2010). Adsorption of synthetic organic chemicals by carbon
nanotubes: Effects of background solution chemistry. Water Res 44: 2067-2074.
http://dx.doi.0rg/10.1016/i.watres.2009.12.017
Zhu. W: Liu. L: Zou. P: Xiao. L: Yang. L. (2010). Effect of decabromodiphenyl ether (BDE 209) on soil
microbial activity and bacterial community composition. World J Microbiol Biotechnol 26: 1891-1899.
http://dx.doi.org/10.1007/sll274-010-0371-l
Zou. M. -Y: Ran. Y: Gong. J: Mai. B. -X: Zeng. EY. (2007). Polybrominated diphenyl ethers in watershed soils
of the Pearl River Delta, China: Occurrence, inventory, and fate. Environ Sci Technol 41: 8262-8267.
http://dx.doi.org/10.1021/es071956d
D-11
-------
Appendix E. Environmental Contaminant
Concentrations
Appendix E provides information available from the literature regarding reported environmental
concentrations of decaBDE (BDE-209) and multiwalled carbon nanotubes (MWCNTs) in environmental
media (dust, indoor and outdoor air, aquatic systems, sewage, and soil) (Section E.I), and biota (Section
E.2). Literature was identified primarily using review articles published in the past two years. Targeted
literature searches were carried out as needed.
E.1. Concentrations in Environmental Media
The following tables provide details from studies that measured BDE-209 in dust, air, water,
sediment, soil, and sewage effluent/sludge. No data were identified on MWCNT concentrations in
environmental media.
Table E-1. BDE-209
concentrations in building dust.
Citation Relevant study info
BDE-209 levels (ng/gram)
United States
Stapleton et al.
(2005)
Sharp and Lunder
(2004) as cited in
U.S. EPA (2010)
• Washington, DC
• Dust samples from 16 homes
• Throughout United States
• 10 homes
Mean (dry wt): 2,090
Mean (dry wt): 2,394
Sjodin et al. (2008)
Atlanta, GA
Dust in vacuum cleaner bags analyzed from 10 homes
BDE-209 was the dominant congener in dust samples
Median (range):
2,000(120-21,000)
Charles et al. (2005)
• Computer labs, CA
• 2 carpet dust samples
• BDE-209 was the dominant congener in carpet dust
Mean: 5,180
Schecter et al.
(2005)
Dallas, TX
9 vacuum samples
BDE-209 was the dominant congener in 7 samples
Mean (Median) (dry wt):
8,567 (665)
E-1
-------
Table E-1 (Continued): BDE-209 concentrations in building dust.
Citation
Relevant study info
BDE-209 levels (ng/gram)
Allen et al. (2008)
Boston, MA
20 homes; 3 locations/home: living room, bedroom,
vacuum
One sample contained highest concentration of BDE-209
in house dust reported to date (527,000 ng/gram)
Geometric means:
4,502 (main living area);
1,703 (bedroom);
1,811 (vacuum)
Wu et al. (2007)
[levels reported by
U.S. EPA (201 0)1
Harrad et al.
(2008b)
Johnson-Restrepo
and Kannan (2009)
Batterman et al.
(2010)
Watkins et al.
(2011)
• Boston, MA area
• 46 women; 11 samples above detection limits
• Amarillo, Austin, TX carpet dust from 17 homes
• Albany, NY
• Vacuum dust from 12 homes
• Southeast Michigan (Ann Arbor area)
• Vacuum dust from 10 office buildings
• Boston, MA area
• Vacuum dust from 8 office buildings
Median: ND; 9,020
Mean (geometric mean):
1,600(1,300)
Mean (median) (dry wt):
2,810(903)
Mean: 6,930
Geometric Mean: 4,204
International
Muenhor et al.
(2010)
Harrad et al.
(2008b)
Harrad et al.
(2008a)
• Thailand: 5 electronic/ electrical waste storage facilities
• Dust, 25 samples
• Canada: carpet dust from 7 homes
• U.K.: carpet dust from 16 homes
• Concentrations of BDE-209 in 2 UK samples were highest
recorded to date in a domestic (or office) indoor dust
sample (520,000 and 100,000 ng/gram)
• Birmingham, UK
• 30 homes, 18 offices, 20 cars
• BDE-209 concentrations (ng/gram) in three samples were
highest to date at 2,600,000 (car), 2,200,000 (home),
1,400, 000 (home)
Mean: 33,000
Mean (geometric mean):
670 (590) (Canada);
45,000 (3,800) (U.K.)
Mean (median):
260,000 (8,100) (homes);
30,000 (6,200) (offices);
410,000 (100,000) (cars)
Ma et al. (2009) • Taizhou, China
• 5 dust samples from electronic waste recycling workshop
floor
• BDE-209 accounted for major proportion of total PBDEs
in dust
Sjodin et al. (2008)
Household dust in vacuum cleaner bags from 10 homes
in each country: 6 cities in Germany, 2 cities in Australia,
and 1 city in the United Kingdom (total n=30)
BDE-209 dominant congener
Mean (range) (dry wt):
29,800 (5,560-80,600)
Median (range):
63 (<6-410) (Germany);
730 (23-13,000) (Australia);
10,000 (910-54,000) (U.K.)
Note: Additional information obtained from U.S. EPA (2010): Abbreviations: dry wt = Dry weight; ND = Not detected
E-2
-------
Table E-2. BDE-209 air concentrations in outdoor and indoor air.
Citation
Relevant study info
BDE-209 levels
United States
Hoh and Hites
(2005) as cited in
U.S. EPA (2010)
Outdoor sampling at 5 locations; sampling every 12
days (August 2003-January 2004)
Chicago (urban), remote locations in Michigan and
Louisiana, agricultural site in Arkansas, and small
college town of Bloomington, Indiana
BDE-209 dominant congener at all sites
Means (pg/m )
60.1 (Chicago)
1.4 (Michigan)
2.6 (Louisiana)
9.0 (Arkansas)
2.2 (Indiana)
Hoh et al. (2005)
• Outdoor sampling at 5 locations; sampling every 12
days (September 2002-December 2003 or
throughout 2003 [Chicago only])
• Chicago (urban), remote locations in Michigan and
Louisiana, agricultural site in Arkansas, and small
college town of Bloomington, Indiana
Values at different sampling dates
(pg/m3)
15, 16, 17, 65 (Chicago)
9.7, 12 (Louisiana)
20, 22 (Arkansas)
0.2, 7.3 (Indiana)
Strandberg et al.
(2001)
• Outdoor sampling at 4 locations; 4 samples/year at
each location, May-October (1997-1999)
• 1 urban (Chicago), 1 remote (Michigan), 2 rural
(Michigan, New York)
Means (pg/m )
0.3 (Chicago)
ND (rural/remote sites)
Charles et al. (2005)
• Indoor and outdoor sampling at industrial and office
sites, outdoors at UC Davis (2004)
• Control - outdoors at UC Davis
• Indoors at computer facility measured concentrations
with computers on and off. Concentrations higher
when computers turned on compared to when
computers turned off.
Mean (range) (pg/m )
10.6 (4.44-17.8) (control)
58 (50.2-65.3) (indoors, computer
facility)
140-11,400 (range for outdoors
surrounding electronics recycling
facility)
79,700-833,000 (range for
indoors at electronics recycling
facility)
45.5-1,940 (range for outdoors at
auto shredder facility)
CADAMP
TfCal/EPA. 2006):
as cited in U.S. EPA
(2010)
7 outdoor sampling sites in California: 4 Bay Area
sites, 3 South Coast sites (2003-2004)
6 monthly samples in 2003; 12 monthly samples in
2004
25 pg/m
E-3
-------
Table E-2 (Continued): BDE-209 air concentrations in outdoor and indoor air.
Citation
Relevant study info
BDE-209 levels
Allen et al. (2007)
• Indoor air at 20 urban residences, Boston, MA area
(January-March 2006)
• Personal air (within 30 cm of breathing zone),
bedroom, and main living area
• Total personal air concentrations for BDE-209 was
significantly higher than bedroom and main living
room concentrations
• Inhalation may account for up to 22% of the total
BDE-209 exposure in United States adults.
Geometric means (pg/m
173.6 (personal air)
94.8 (bedroom)
94.2 (living room)
Johnson-Restrepo
and Kannan(2009)
Indoor air in 12 homes, Albany, NY (December ND (ng/m
2007-January 2008)
Salamova and Hites
(2011)
Vapor, particle, and precipitation samples collected at
2 urban sites, 1 rural site, 2 remote sites) around the
Great Lakes (2005-2009; part of the Integrated
Atmospheric Deposition Network)
Statistical analysis indicated that levels of BDE-209
have not changed between 2005 and 2009
Means (pg/m )
Vapor: 3.4, 1.8 (urban); 0.7 (rural);
0.5, 0.8 (remote)
Particle: 13, 56 (urban); 1.9
(rural); 1.3, 2.5 (remote)
Precipitation (ng/L): 2.1, 4.1
(urban); 0.6 (rural); 0.4, 0.5
(remote)
Batterman et al. • Airborne particulate matter and vapor samples
(2010) collected at 10 office buildings in southeast Michigan
(Ann Arbor area)
Means and medians were all
below the limit of detection
International
Su et al. (2007) • Air samples collected in the Canadian High Arctic
(Alert, Nunavut); PBDEs quantified in 104 samples
(2002-2004)
• Lack of seasonality effects for BDE-209; BDE-209
likely particle-bound and experiences LRT
Mean (range) (pg/m
1.6(0.091-9.8)
Chang et al. (2009)
• Characterized airborne exposure of students to BDE-
209 and other PBDEs inside and outside a computer
classroom with 61 computers, southern Taiwan
college
• BDE-209 was one of the five highest indoor
concentrations
• Mean BDE-209 concentration outdoors significantly
higher than the mean in indoor air
Means (pg/m )
23.0 (inside classroom)
53.3 (outside, open space in front
of teacher building)
Agrell et al. (2004)
Atmospheric concentrations (gaseous and
particulate) of BDE-209 measured at solid waste
incineration plant in Sweden
Particulate concentrations at MSW significantly
higher
Medians (pg/m )
10.4 (MSW)
6.5 (reference site)
E-4
-------
Table E-2 (Continued): BDE-209 air concentrations in outdoor and indoor air.
Citation
Relevant study info
BDE-209 levels
Gouin et al. (2006)
Examined particle bound air transport of BDE-209,
Southern Ontario, Canada (2002)
Nearly all BDE-209 sorbed to aerosol particles
LRT of BDE-209 might be controlled by transport
characteristics of aerosols to which they sorb
Mean (range) (pg/m
19(ND-105)
Note: Additional information obtained from U.S. EPA (2010).
Abbreviations: LRT = Long range transport; MSW = Municipal solid waste; ND = Not detected.
E-5
-------
Table E-3. BDE-209 concentrations in aquatic systems.
Citation
Relevant study info
BDE-209 levels
United States
Oros et al. (2005)
• San Francisco estuary (2002)
• 48 sediment samples; 33 water samples
Range (pg/L)
ND-191 (surface water)
ND (sediment)
Yun et al. (2008)
• Saginaw River Watershed, Michigan (2004)
• 53 surficial sediment samples
• BDE-209 was the predominant congener (79% and
90% of the total PBDE in the Shiawassee and
Saginaw Rivers, respectively)
Means (ng/gram dry wt)
2.28 (Shiawassee River)
4.76 (Saginaw River)
1.98 (Saginaw Bay)
Song et al. (2005b;
2005a: 2004)
• Great Lakes
• 16 total sediment sampling stations
Range (ng/gram): 4.3-242
(surficial sediment)
Raff and Hites
(2004)
• Mississippi and tributaries
• Suspended sediment samples from 31 sites (2002-
2003)
• BDE-209 was the dominant congener (96.8% of total
concentration)
Range of 15 PBDEs (ng/gram dry
wt): 29-1,548
Ashley et al. (2006)
• Delaware River
• 4 sediment samples
• BDE-209 was the dominant congener (49% of total
concentration)
Range (ng/gram dry wt): 0.16-
14.79
Dodder et al. (2002)
• Lake Hadley, Indiana
• 4 surficial sediment samples
• BDE-209 was the dominant congener
Range (ng/gram dry wt): 19-36
La Guardia et al.
(2007)
• Downstream of WWTP of plastics manufacturer,
North Carolina
• 8 sediment sample locations downstream of outfall
(2002 and 2005); 2 sludge samples, 1 for each year
• BDE-209 was the dominant congener in sediment
(>89% of total concentration)
Range
2002 SD: 300-3,150 ng/gram
2005 SD: 181-2,390 ng/gram
International
Toms et al. (2006)
as cited in U.S. EPA
(2010)
Estuarine, freshwater, marine sediments, Australia
90 sediment samples from remote and industrial
areas (2002-2003 and 2005)
BDE-209 was the dominant congener in 86% of
samples
Mean (range) (ng/gram dry wt)
4.7(ND-60.9) (all PBDEs)
E-6
-------
Table E-3 (Continued): BDE-209 concentrations in aquatic systems.
Citation
Relevant study info
BDE-209 levels
Christensen and • Danish marine coastal areas, freshwater lakes, river
Platz (2001) (2000)
• BDE-209 was the dominant congener in marine and
freshwater sediments
• Highest BDEs detected in urban sediments
Eljarrat et al. (2005)
Coastal areas, Spain
13 marine sediment samples
BDE-209 was the dominant congener (50-99% of
total concentration)
Range (ng/gram dry wt)
<0.9-3.9 (marine)
<1.3-8.1 (freshwater)
Range (ng/gram dry wt)
2.46-132.10
Eljarrat et al. (2007)
• Spanish River Vero, samples collected up- and
downstream from an industrial park (2004, 2005)
• 6 sediment and 3 effluent samples
• Maximum BDE-209 in sediment downstream of
industrial park (that includes textile industry)
Maximum (ng/gram dry wt)
5,395 (2004)
12,459 (2005) (collected 5 meters
downstream of textile industry
effluent discharge)
Qiu et al. (2007)
Lake Ontario
Sediment core study
Mean (ng/gram dry wt):
14 (surficial)
Zhu and Hites
(2005)
• Lake Michigan and Lake Erie
• Sediment core study
• BDE-209 was the dominant congener in both
sediment cores (95-99% of total concentration)
Surface concentrations (ng/gram)
315 (Lake Michigan)
39 (Lake Erie)
De Boer et al.
(2003)
• Various locations, The Netherlands
• Collected 44 SPM samples at 18 locations
• Collected 22 sediment samples at 17 locations
• SPM identified as an important carrier for BDE-209 in
aquatic environment. Maximum of 4,600 mg/kg dry wt
likely related to spills from textile industries; maximum
of 510 mg/kg dry wt in sediment at same location of
maximum SPM
Median (Range) (ug/kg dry wt)
71 (<9-4,600) (SPM)
22 (<4-510) (sediment)
Eljarrat et al. (2004)
[also reported in
Law et al. (2006b)
review article]
• BDE-209 determined in 5 riverine and 8 marine
sediments, Spain
Range (ng/kg dry wt)
2.06-39.89 (river)
2.95-132.11 (marine)
Sawal et al. (2004)
[also reported in
Law et al. (2006b)
review article]
• BDE-209 determined in 29 surface sediment from
River Elbe, Germany and Czech Republic
• BDE-209 represented 80% of total BDEs
Range (ug/kg dry wt):
0.5-17.4
From Law et al.
(2006b) review
article
• BDE-209 determined in sediments from Lake Mjosa,
Norway (Schlabach et al., 2004)
• In some parts of the lake, BDE-209 represented 50-
90% of total BDEs
Range total BDE (ug/kg dry wt):
0.6-27
Voorspoels et al.
(2004) [also
reported in Law et
Analyzed sediments from Belgian North Sea, Western
Scheldt Estuary
BDE209 was detected in 83% of samples from the
Maximum (ng/kg dry wt): 1,200
(at estuary)
E-7
-------
Table E-3 (Continued): BDE-209 concentrations in aquatic systems.
Citation
Relevant study info
BDE-209 levels
al. (2006b) review
article]
Belgian North Sea and in 100% of samples from the
Scheldt Estuary
Ricklund et al.
(2010)
Measured levels of BDE-209 in 11 lake sediment
samples and 7 marine sediment samples in Sweden
No known point sources of BDE-209 exist; presence
in sediments was presumed to be evidence of long-
range atmospheric transport and deposition
Range (ng/gram dry wt)
0.48-11 (lake)
1.0-88 (marine)
Mai et al. (2005) • Examined 66 surface sediment samples from the
Pearl River Delta and South China Sea, China
• Sources of PBDEs in the area: waste discharges from
urban centers; regional growth of electronic
manufacturing
• BDE-209 dominated congener compositions in
sediments; PBDE composition analysis provided
possible evidence of debromination of BDE-209
Range (ng/gram dry wt): 0.4-7,340
Chen et al. (2007b)
Examined 3 sediment cores from the Pearl River
Estuary, South China
Increased BDE-209 flux in the upper sediment cores
attributed to rapid regional growth of electronics and
other industry
Range (ng/gram):13.5-30.3
Guzzella et al. • PBDEs measured in sediment cores (2005) from Lake Range (ng/gram dry wt): 1.6-15.3
(2008) Maggiore and tributary grab samples, Italy and
Switzerland
• BDE-209 was the dominate congener (>95% of total
PBDEs)
• Increase in BDE-209 attributed to textile industries
Zhao et al. (2011)
Measured concentrations of PBDEs in sediments of
the Daliao River Estuary, China
BDE209 was the dominating congener in all samples
Intrusion of sea waters accelerated deposition of the
colloid-associated PBDEs; significantly negative
correlations observed between PBDE concentration
and both pH and salinity in bottom waters; higher river
flow in the flood season (summer) accelerated
transport of PBDEs to the ocean; TOC and PBDE
distributions indicated that TOC controlled
distributions of PBDEs in sediments of the estuary
Range of all PBDEs (ng/gram dry
wt): 0.13-1.98
(BDE-209 levels stated to be
about 1 order of magnitude higher)
Note: Additional information obtained from U.S. EPA (2010)
Abbreviations: dry wt = Dry weight; ND = Not detected; SPM = Suspended particulate matter; TOC = Total organic carbon;
VWVTP = Wastewater treatment plant
E-8
-------
Table E-4. BDE-209 concentrations in sewage effluent and sludge.
BDE-209 levels
Citation
Relevant study information
STP effluent
Sewage sludge
United States
Hale et al. (2001)
Mid-Atlantic biosolids
Northeast biosolids
Gulf biosolids
West biosolids
Ranges (ug/kg
dry wt)
84.8-1,460
1,940-4,890
368 (single site)
340-450
Hale et al. (2003)
• Lake Superior watershed communities
• Lake Michigan watershed communities
Mean
(ug/gram dry wt)
510
466
North (2004) • Samples analyzed for 41 BDE congeners in CA 1,730 (pg/L)
• STP discharges effluent into San Francisco
estuary
• In sludge, BDE-209 was 35% of total BDEs
• Estimated that 96% of PBDEs that enter the
STP adsorb to sludge; 4% in effluent
La Guardia et al.
(2007)
• Downstream of WWTP of plastics
manufacturer, North Carolina
• 2 sludge samples, one taken in 2002 and the
other in 2005
• BDE-209 was the dominant congener in sludge
Mean
(ug/kg dry wt)
1,183
58,800 ug/kg
drywt(2002
measurement)
37,400 ug/gram dry
wt (2005
measurement)
U.S. EPA (2009)
• National Sewage Sludge Survey
• Evaluated 74 STPs in 35 states (2006-2007)
• Nationally, BDE-209 was the dominant
congener
Mean (ug/kg
dry wt) 2,181
E-9
-------
Table E-4 (Continued): BDE-209 concentrations in sewage effluent and sludge.
BDE-209 levels
Citation
Relevant study information
STP effluent
Sewage sludge
International
De Boer et al.
(2003)
• Various locations, The Netherlands
• Collected 13 sewage treatment plant (STP)
influent/effluent samples at 9 locations
(measured filtering out particulate matter); 3
sludge samples
Median (range)
(ug/kg dry wt)
24 (<0.5-330)
(influent)
350(310-920)
(effluent)
<180, 190, 8.6
(ug/kg dry wt)
Knoth et al. (2007) • Sewage sludge from 11 STPs in Germany
(2002-2003)
• BDE-209 was the dominant congener in
sludges; no PBDEs with fewer than 7 bromines
observed
• Estimated 350 kg/acre BDE-209 applied to land
in Germany in 2001
Mean (ng/g dry wt)
429
Wang et al. (2007)
Clarke et al. (2008)
Kupper et al.
(2008)
Eljarrat et al.
(2QQ7)
Ricklund et al.
(2009)
• Sewage sludge from 31 STPs in 26 cities in -
China
• BDE-209 was dominant congener in most
samples
• Australian sewage sludge survey; 16 WWTPs
(2006)
• Presented urban mean, rural mean, and overall
mean of BDE-209 in sludge samples
• Switzerland, monitoring network -
• 16 WWTPs
• Spanish River Vero, samples collected up- and 1,170 ng/L (2005
downstream from an industrial park (2004, effluent maximum)
2005)
• 6 sediment and 3 effluent samples
• Stockholm, Sweden -
• WWTP (2006, 2007)
Mean (ng/gram
dry wt) 68. 5
Mean (ug/kg dry
wt) 880 (urban);
490 (rural)
720 (overall)
Mean (ug/kg dry
wt)310
800 Mean
(ng/gram dry wt)
Note: Additional information obtained from U.S. EPA (2010)
Abbreviations: dry wt = Dry weight; STP = sewage treatment plant; WWTP = Wastewater treatment plant.
E-10
-------
Table E-5. BDE-209 concentration data in soil.
Citation
Relevant study info
BDE-209 levels
United States
Offenberg et al.
(2006) as cited in
U.S. EPA (2010)
• 33 surface soil samples, 15 states
• BDE-209 detected in 24/33 samples
Mean (ng/gram dry wt): 15.3
Yun et al. (2008)
Saginaw River Watershed, Michigan
26 floodplain surface soil samples (2004)
Mean (ng/gram dry wt)
10.8 (Shiawassee River)
2.77 (Saginaw River)
0.6 (Saginaw Bay)
International
Sellstrom et al.
(2005)
• 5 sites in Sweden
• Evaluated sewage sludge amended soils and
Range (ng/gram dry wt)
0.028-2,220
earthworms
Luo et al. (2009)
Southern China
Analyzed road and farmland soils from e-
waste recycling region
BDE-209 contributions averaged 84% in
samples from the e-waste region higher than
97% in samples from the industrial and
reference sites
Mean (range) (ng/gram dry wt)
19.7 (rural farmland soil)
59.8 (farmland soil near industrial)
E-waste region
1,539.3 (69.1-6,319.6) (road soil)
32.2 (farmland soil near dismantling
workshop)
29.9 (farmland soil near open burning site)
Zou et al. (2007)
• Pearl River, China
• 33 surface soil samples; 3 point source
samples
Mean (range) (ng/gram dry wt)
13.8 (2.38-66.6) (SS)
70.5 (25.7-102)(PS contaminated)
Note: Additional information obtained from U.S. EPA (2010)
Abbreviations: dry wt = Dry weight; SS = Surface soil; PS = Point source
E-11
-------
Table E-6. Proxy data for estimating MWCNT concentrations in occupational air.
Citation
Relevant study info
Proxy data and CNT counts
United States
Bello et al. (2008)
Personal breathing zone and area air sampling and
real-time monitoring in a CNT research laboratory
synthesizing and handling CNTs
Fast mobility particle sizer measured number
concentration for particles sized 5.6-560 nm.
Personal breathing zone and area air samples near
the emission source were collected and analyzed for
respirable dust and respirable fiber concentrations,
and electron microscopy characterized particles and
fibers on filters.
No increase in total particle
number concentration or in
particle number in any size range
compared to background.
No individual or bundled CNTs
detected.
Bello et al. (2009)
Personal breathing zone and area air sampling and
real-time monitoring during machining of carbon,
alumina, CNT-carbon, and CNT-alumina composites
at a research laboratory
Fast mobility particle sizer and aerodynamic particle
sizer measured number concentrations of particles
sized 5.6-560 nm and 0.5-20 urn, respectively, and
condensation particle counter counted all particles 10
nm-1 urn. Total dust mass was measured in real
time using TSI Dust Trak® Personal breathing zone
and area air samples near the emission source were
collected and analyzed for respirable dust and
respirable fiber concentrations, and electron
microscopy characterized particles and fibers on
filters
No engineering controls were employed
Dry cutting of all composites produced significant
numbers of nanoscale particles, and particle sizes
were similar for all composites.
The thinnest CNT-alumina composite released fewer
nanoscale particle than the other composites during
dry cutting
No discernible difference between the number of
respirable particles and fibers produced during dry
cutting of CNT-composites versus base composites
No individual CNT structures or
bundles were observed in the
samples.
No CNT structures or bundles
were observed in the composite
particle dust.
E-12
-------
Table E-6 (Continued): Proxy data for estimating MWCNT concentrations in occupational air.
Citation
Relevant study info
Proxy data and CNT counts
Dahm et al. (2011)
Task based area air, full-shift personal breathing
zone (PBZ), and outdoor background sampling at 6
CNT/CNF primary and secondary manufacturing
facilities
PBZ samples collected for inhalable fraction and area
air samples collected for inhalable and respirable
fractions of elemental carbon mass. Electron
microscopy characterized CNT structures (both
single CNTs and bundles) on filters
PBZ samples collected during dry powder handling
tasks at two secondary MWCNT facilities exceeded
the National Institute of Occupational Safety and
Health recommended exposure limit (7 ug/m3
elemental carbon) in the presence of controls
CNT/CNF structure were identified on filters at all
sites and correlation between filter mass and CNT
structure count was statistically significant (p = 0.01)
after exclusion of single outlier value
Elemental carbon mass (ua/m )
and CNT structure count at
primary MWCNT facilities
Outdoor background: not detected
MWCNT production and
harvesting: 1.6-2.74 (PBZ),0.49-
4.62 (area: inhalable), not
detected to 0.78 (area:
respirable), 0.090-0.399
CNTs/cm3 (PBZ), 0.026-0.134
CNTs/cm3 (area: inhalable)
MWCNT sonication, sieving, and
spray coating: 1.13 (PBZ), not
detected (area: inhalable), not
detected to 0.7 (area:
respirable),0.010 CNTs/cm3
(PBZ),0.002 CNTs/cm3 (area:
inhalable)
Elemental carbon mass (ug/m )
and CNT structure count at
secondary MWCNT facilities
Outdoor background: not detected
Office work and waste collection:
0.8-1.06 [PBZ),0.001-0.214
CNTs/cm3 (PBZ)
Weighing, mixing, sonication,
extruding, transferring MWCNTs:
not detected to 7.86 (PBZ),not
detected to 1.01 (area: inhalable),
not detected to 2.76 (area:
respirablej, not detected to 0.242
CNTs/cm3 (PBZ), not detected to
0.008 CNTs/cm (area: inhalable)
Milling MWCNT composite:
not detected
E-13
-------
Table E-6 (Continued): Proxy data for estimating MWCNT concentrations in occupational air.
Citation
Relevant study info
Proxy data and CNT counts
Johnson et al. • Area air sampling and real-time monitoring of
(2010) MWCNT and hydroxylated MWCNT (MWCNT-OH)
emissions during weighing, transferring, and
sonicating with water and natural organic matter
• HHPC-6 particle counter measured particle number
per liter air for 6 size cuts: 300, 500, 1,000, 3,000,
5,000, and 10,000 nm. Condensation particle counter
measured total particle numberslO-1,000 nm.
Electron microscopy characterized MWCNT
structures on filters.
• Area air sample collected prior to tasks was used as
background concentration and subtracted from
samples taken during performance of tasks
Adjusted number concentration
(particles/L)
Raw MWCNT weighing,
transferring, and mixing without
ventilation: 123,403 (300 nm:
above limit of quantitation), 34,446
(500 nm), 4,338 (1,000 nm), 50
(3,000 nm), 0 (5,000 and 10,000
nm).
MWCNT-OH weighing,
transferring, and mixing without
ventilation: 0 (300 and 10,000
nm), 3,065 (500 nm), 1,699 (1,000
nm), 280 (3,000 nm), 4 (5,000
nm).
Raw MWCNT sonication: 42,796
(300 nm), 23,777 (500 nm), 2,184
(1,000nm), 86 (3,000 nm), 0
(5,000 and 10,000 nm).
MWCNT-OH sonication: 144,623
(300 nm: above limit of
quantitation), 65,402 (500 nm),
6,205 (1,000 nm), 0 (3,000, 5,000,
and 10,000 nm).
Total adjusted number
concentration 10-1.000 nm
(particles/cm )
Raw MWCNT weighing,
transferring, and mixing without
ventilation: 1,576
MWCNT-OH weighing,
transferring, and mixing without
ventilation: 676
Raw MWCNT sonication: 2,776
MWCNT-OH sonication: 726
E-14
-------
Table E-6 (Continued): Proxy data for estimating MWCNT concentrations in occupational air.
Citation
Relevant study info
Proxy data and CNT counts
Methner et al. • Area air sampling and real-time monitoring of 2
(2010) MWCNT research and development laboratories
during specific handling tasks
• Condensation particle counters (CPC) counted
particles sized 10-1,000 nm and optical particle
counters (OPC) counted particle sized 300-500 nm
and 500-1,000 nm. Electron microscopy
characterized MWCNT structures on filters and
energy-dispersive X-ray analysis confirmed chemical
identity.
• MWCNT structures on microscopy grids were not
quantified, but were detected in samples taken during
weighing and sonication of both raw and
functionalized MWCNTs. No MWCNT structures
were observed on filters sampling background. Filters
were not analyzed by electron microscopy for
samples taken during opening of the growth
chamber.
• Measured particle number concentrations are
background adjusted
CPC particle number
concentrations (particles/cm3):
10-1.000 nm fraction
Opening MWCNT growth
chamber: 300 (with exhaust),
42,400 (without exhaust)
Handling raw MWCNTs: 1,480-
1,580 (weighing); 2,200-2,800
(sonicating)
Handling functionalized MWCNTs:
680 (weighing); 730 (sonicating)
OPC particle number
concentrations (particles/L): 300-
500 nm fraction: 500-1.000 nm
fraction
Opening MWCNT growth
chamber: 0; 0 (with exhaust), 350;
400 (without exhaust)
Handling raw MWCNTs:
53,1,000-123,400 (above limit of
quantitation); 3,900-34,400
(weighing); 23,900-42,800;
6,500-23,800 (sonicating)
Handling functionalized MWCNTs:
0; 3,100 (weighing); 144,600
(above limit of quantitation);
65,400 (sonicating)
International
Han et al. (2008)
• Personal and area air sampling and real-time aerosol
monitoring conducted at MWCNT research facility
• Scanning mobility particle sizer with ultrafine
condensation particle counter and aerodynamic
particle sizer monitored particle size distribution 14-
630 nm and 0.5-20 urn, respectively, and
aethalometer characterized mass exposure to carbon
black. Electron microscopy characterized MWCNT
structures on filters and energy-dispersive X-ray
analysis confirmed chemical identity.
• Exposure controls included installation of a fan,
cleaning, and equipment rearrangement (i.e.,
isolation)
• No values exceeded ACGIH TLVs or Korean Ministry
of Labor OELs for carbon black or particles not
otherwise specified, but fiber counts exceeded limits
for asbestos and other fiber or tube-like materials
Total dust concentration (ua/m )
No control measures: 210-430
With control measures: not
detected
MWCNT counts (MWCNTs/cm3)
No control measures: 172.9-
193.6
With control measures: 0.018-
0.05
E-15
-------
Table E-6 (Continued): Proxy data for estimating MWCNT concentrations in occupational air.
Citation
Relevant study info
Proxy data and CNT counts
Lee et al. (2010)
• Personal air sampling, area sampling, and real-time
aerosol monitoring conducted at 7 MWCNT handling
facilities (3 manufacturing plants, 4 research
laboratories)
• Scanning mobility particle sizer, dust monitor, and
aethalometer characterized particle number, size
distribution, and mass exposures. Electron
microscopy characterized MWCNT structures on
filters.
• No values exceeded American Conference of
Governmental Industrial Hygienists (ACGIH)
threshold limit values (TLVs) or Korean Ministry of
Labor occupational exposure levels (OELs) for
carbon black, particles not otherwise specified, or
asbestos.
• Nanoscale particles most often released during
opening of chemical vapor deposition (CVD) reactor
and catalyst preparation.
• Nanoscale particles assumed to be primarily metal
catalysts, not MWCNTs
3.
One filter sample detected
MWCNTs at 0.00312 tubes/cm
all others were non-detects
Total suspended particulate
matter
Personal air: 7.8 - 320 ug/m3
Area air: 12.6-187 ug/m3
Particle number count for mode
particle sizes (particles/ cm ) - by
task
Catalyst preparation (mode
diameter 20-30nm): 18,600-
75,000
CVD opening (mode diameter 20
or 50 nm): 6,974-16,857
Other operations (no mode
diameters reported): 5,276-6,399
Takaya et al.(2010)
(English translation
available only for
abstract)
Real-time aerosol monitoring and personal air
sampling in two MWCNT packing facilities
(automated packing versus manual packing)
Nanoscale and submicron-/micron-scale particles
measured using scanning mobility particle sizer and
optical particle counter, respectively
Submicron scale particles (not nanoparticles)
released during bagging
Airborne dust (both facilities):
240 ug/m3
Personal air at manual facility:
2,390 ug/m3 (total dust); 390
ug/m3 (respirable dust)
Personal air at automated facility:
290 ug/m3 (total dust); 80 ug/rrf
(respirable dust)
E-16
-------
E.2. Concentrations in Biota
The following tables provide details from studies that measured BDE-209 in different types of
biota. No data were identified on MWCNT concentrations in biota.
Table E-7. Measured concentrations of PBDEs in biota.
Species
Location
Year
PBDE
burden Common
Tissue (ng/gram)a congener(s)
Source
Terrestrial birds
Peregrine
falcon
Chesapeake Bay 1993-2002 Eggs Median: 201b BDE-153: 26% Potter et al. (2009)
CT, MA, ME, NH,
VT
1996, -
1999-2006
Median: 440"
BDE-153,
BDE-99
Chen et al. (2008)
Common
blackbird
Switzerland
2003-2005 Brain
BDL
NA
Adipose BDL
NA
Naert et al. (2007) as
•cited in U.S. EPA
(2010)
TB
0.82
BDE-47: 100%
Sparrow hawk Switzerland
2003-2005 Brain
14
NR
NR
Naert et al. (2007) as
•cited in U.S. EPA
(2010)
Belgium
NR
TB 790.2
NR Liver Mean: 4,900
Median:
1,300
Brain Mean: 1,200
Median: 360
Adipose Mean: 1,900
NR Liver Mean: 9,500
BDE-99: 40%
Top 3: BDE-99,
BDE-47,
BDE-1 53
NR
NR
Top 4: BDE-99,
BDE-47,
BDE-1 00,
BDE-1 53
Voorspoels et al.
(2006b)
Voorspoels et al.
(2006b)
Voorspoels et al.
(2006b)
Voorspoels et al.
(2007) as cited in
U.S. EPA (2010)
E-17
-------
Table E-7 (Continued): Measured concentrations of PBDEs in biota.
Species
Common
buzzard
Common
buzzard
Cormorant
Owls
Common
kestrel
Passerines
Location
Switzerland
Belgium
Beijing, China
NR
Switzerland
Belgium
Beijing, China
NR
Year Tissue
2003-2005 TB
NR Liver
NR Liver
NR Liver
2003-2005 TB
NR NR
NR Muscle
Liver
Kidney
NR Adipose
Eggs
PBDE
burden
(ng/gram)a
34.55
Mean: 480
Median: 70
148
Mean: 720
98.76
250
Mean:
12,300
Mean:
12,200
Mean: 5,340
160
220
Common
congener(s)
BDE-153:29%
BDE-99: 23%
BDE-47: 22%
Top 3: BDE-
153, BDE-47,
BDE-99
BDE-209: -43%
NR
BDE-47: 42%
Top 3: BDE-
153, BDE-99,
BDE-47
NR
NR
NR
NR
NR
Source
Naert et al. (2007) as
cited in U.S. EPA
(2010)
Voorspoels et al.
(2006b)
Chen et al. (2007a)
Voorspoels et al.
(2007) as cited in U.S.
EPA (2010)
Naert et al. (2007) as
cited in U.S. EPA
(2010)
Voorspoels et al.
(2006b)
Chen et al. (2007a)
Voorspoels et al.
• (2007) as cited in U.S.
EPA (2010)
Marine birds
Herring gull
Fulmar
Murre
Heron
Great Lakes
Northern Canada
Northern Canada
British Columbia
1981-2000 Eggs
1975-1998 Eggs
1975-1998 Eggs
1983-2000 Eggs
9.4-1,544
0.212-2.37
0.442-2.93
1,308-288
NR
NR
NR
NR
Norstrom et al. (2002)
Wakeford et al. (2002)
as cited in U.S. EPA
(2010)
Wakeford et al. (2002)
as cited in U.S. EPA
(2010)
Wakeford et al. (2002)
as cited in U.S. EPA
(2010)
E-18
-------
Table E-7 (Continued): Measured concentrations of PBDEs in biota.
Species
Location
Year
PBDE
burden
Tissue (ng/gram)a
Common
congener(s) Source
Mammals
Red fox
Belgium
NR Adipose, Median
liver, range:
muscle 2.2-3.4
Liver- BDE-209: Voorspoels et al.
70% (2006a)
Ringed seals, Canadian Arctic NR
female
Blubber Mean: 25.8 Tetra, pentaBDE Alaee et al. (1999)
Ringed seals, Canadian Arctic NR
male
Arctic
Arctic
1981
2000
Blubber Mean: 50.0
Tetra,
pentaBDE
0.6
Tetra,
pentaBDE
6.0
Tetra,
pentaBDE
Alaee et al. (1999)
Ikonomou et al. (2002)
Ikonomou et al. (2002)
Beluga whales, Canadian Arctic NR
female
St. Lawrence
estuary
NR
Blubber Mean: 81.2
665
Tetra,
pentaBDE
NR
Alaee et al. (1999)
Lebeuf et al. (2001) as
cited in U.S. EPA
(2010)
Beluga whales, Canadian Arctic NR
male
St. Lawrence
estuary
NR
Blubber Mean: 160
466
Tetra,
pentaBDE
NR
Alaee et al. (1999)
Lebeuf et al. (2001) as
cited in U.S. EPA
(2010)
Baffin Island
1982 Blubber 2
1997
15
Tri to hexaBDE
Stern and Ikonomou
. (2000)
Tri to hexaBDE
Harbor seals San Francisco Bay 1989-1998 Blubber Range: Tetra, penta,
88-8,325 hexaBDE
She et al. (2002)
Harbor
porpoise
Vancouver
NR
Blubber 2,269
TetraBDE:
>50%
Ikonomou et al. (2000)
Fish
Lake trout Lake Ontario
Lake Erie
1997
Lake Superior
Lake Huron
NR 434
117
392
251
NR
Luross et al. (2002)
E-19
-------
Table E-7 (Continued): Measured concentrations of PBDEs in biota.
Species
Rainbow trout
Mountain
whitefish
Largescale
sucker
Carp
Lower trophic
PBDE
burden Common
Location Year Tissue (ng/gram)a congener(s) Source
Spokane River, 1999 NR 297d NR
WA
Spokane River, 1999 NR 1,250d NR
WA
Columbia River, 1992-2000 Muscle Mean range NR
British Columbia 4.5-19.1
Spokane River, 1999 NR 105d NR
WA
Virginia 1998-1999 NR 1,140d NR
levels
Johnson
(2001)
Johnson
(2001)
Rayne et
Johnson
(2001)
Johnson
(2001)
and Olson
and Olson
al. (2003)
and Olson
and Olson
Caddisflies Pyrenees Mountains, NR
Spain
Larva TB Mean range: NR
0.65-13.00°
Pupa TB Mean range
9.32- 27°
Bartrons et al. (2007)
Midges Pyrenees Mountains, NR
Spain
Larva TB Mean range: NR
0-13.07
Pupa TB Mean range
3.9-5.2°
Bartrons et al. (2007)
Abbreviations: BDE-47: tetraBDE; BDE-99: pentaBDE; BDE-153: hexaBDE; BDE-209: decaBDE; BDL = Below detection level; TB =
Total body; NR = Not reported; NA = Not applicable.
ang/gram lipid weight, unless otherwise specified.
bUnits = ng/gram wet weight.
°Units = ng/gram dry weight.
°Measurement for a single fish only.
E-20
-------
Table E-8. Mean concentration of PBDEs in media/biota in an aquatic ecosystem.
Media
Water (pg/L)
Sediment (ng/gram dry
wt)
Biota level 1 (mg/gram
Mussels
Biota level 2 (mg/gram
Zooplankton
White fish
Biota level 3 (mg/gram
Emerald shiner
Goldeye
White sucker
Biota level 4 (mg/gram
Total PBDEs BDE-47 BDE-99 BDE-100 BDE-153
47.01 16.98 9.01 1.89 1.02
1.31 0.12 0.15 0.03 0.06
lipid wt)
127.32 21.11 26.41 5.7 8.13
lipid wt)
61.57 11.71 17.79 4.89 5.81
11.1 1.82 1.48 0.6 0.43
lipid wt)
151.07 60.78 2.15 10.65 7.25
381.26 83.84 116.15 24 31.98
150.03 71.4 1.22 16.49 12.55
lipid wt)
BDE-209
-------
Allen. JG: Webster. TF: McClean. MD: Stapleton. HM: Nelson. JW. (2007). Personal exposure to
Polybrominated Diphenyl Ethers (PBDEs) in residential indoor air. Environ Sci Technol 41: 4574-4579.
http://dx.doi.org/10.1021/es0703170
Ashley. J: Libero. D: Halscheid. E: Zaoudeh. L: Stapleton. H. (2006). Polybrominated diphenyl ethers (PBDEs)
in American eels from the Delaware River, USA. Partnership for the Delaware Estuary.
http://www.delawareestuarv.org
Bartrons. M: Grimalt. JO: Catalan. J. (2007). Concentration changes of organochlorine compounds and
polybromodiphenyl ethers during metamorphosis of aquatic insects. Environ Sci Technol 41: 6137-6141.
http://dx.doi.org/10.1021/es0703271
Batterman. S: Godwin. C: Chernyak. S: Jia. C: Charles. S. (2010). Brominated flame retardants in offices in
Michigan, USA. Environ Int 36: 548-556. http://dx.doi.0rg/10.1016/j.envint.2010.04.008
Bello. D: Hart. J: Ahn. K: Hallock. M: Yamamoto. N: Garcia. E: Ellenbecker. M: Wardle. B. (2008). Particle
exposure levels during CVD growth and subsequent handling of vertically-aligned carbon nanotube films.
Carbon 46: 974-981.
Bello. D: Wardle. B: Yamamoto. N: deVilloria. R: Garcia. E: Hart. A: Ahn. K: Ellenbecker. M: Hallock. M.
(2009). Exposure to nanoscale particles and fibers during machining of hybrid advanced composites
containing carbon nanotubes. J Nanopart Res 11: 231-249. http://dx.doi.org/10.1007/sll051-008-9499-4
Cal/EPA (California Environmental Protection Agency). (2006). California polybromintaed diphenyl ethers
(PBDEs) program. Available online at http://www.arb.ca.gov/toxics/pbde.htm
Chang. FH: Yang. CR: Tsai CY: Lin. WC. (2009). Airborne polybrominated diphenyl ethers in a computer
classroom of college in Taiwan. Iranian Journal of Environmental Health Science and Engineering 6: 121-
130.
Charles. MJ: Groskova. D: Cahill. TM. (2005). Near-source ambient air monitoring of polybrominated diphenyl
ethers. Davis, CA: California Air Resources Board.
http://www.arb.ca.gov/toxics/pbde%20final%20report%202005.pdf
Chen. D: La Guardia. MJ: Harvey. E: Amaral M: Wohlfort. K: Hale. RC. (2008). Polybrominated diphenyl
ethers in peregrine falcon (Falco peregrinus) eggs from the northeastern U.S. Environ Sci Technol 42: 7594-
7600. http://dx.doi.org/10.1021/es8010749
Chen. D: Mai. B: Song. J: Sun. Q: Luo. Y: Luo. X: Zeng. EY: Hale. RC. (2007a). Polybrominated diphenyl
ethers in birds of prey from Northern China. Environ Sci Technol 41: 1828-1833.
http://dx.doi.org/10.1021/es062045r
Chen. SJ: Luo. XJ: Lin. Z: Luo. Y: Li. KG: Peng. XZ: Mai. BX: Ran. Y: Zeng. EY. (2007b). Time trends of
polybrominated diphenyl ethers in sediment cores from the Pearl River Estuary, South China. Environ Sci
Technol 41: 5595-5600. http://dx.doi.org/10.1021/es070351e
Christensen. JH: Platz. J. (2001). Screening of polybrominated diphenyl ethers in blue mussels, marine and
freshwater sediments in Denmark. J Environ Monit 3: 543-547.
Clarke. B: Porter. N: Symons. R: Marriott P: Ades. P: Stevenson. G: Blackbeard. J. (2008). Polybrominated
diphenyl ethers and polybrominated biphenyls in Australian sewage sludge. Chemosphere 73: 980-989.
http://dx.doi.0rg/10.1016/i.chemosphere.2008.06.034
Dahm. MM: Evans. DE: Schubauer-Berigan. MK: Birch. ME: Fernback. JE. (2011). Occupational Exposure
Assessment in Carbon Nanotube and Nanofiber Primary and Secondary Manufacturers. Ann Occup Hyg -: 1-
15. http://dx.doi.org/10.1093/annhvg/merllO
de Boer. J: Wester. PG: van der Horst A: Leonards. PEG. (2003). Polybrominated diphenyl ethers in influents,
suspended paniculate matter, sediments, sewage treatment plant and effluents and biota from the
Netherlands. Environ Pollut 122: 63-74. http://dx.doi.org/10.1016/s0269-7491(02)00280-4
Dodder. NG: Strandberg. B: Kites. RA. (2002). Concentrations and spatial variations of polybrominated
diphenyl ethers and several organochlorine compounds in fishes from the northeastern United States. Environ
Sci Technol 36: 146-151. http://dx.doi.org/10.1021/es010947g
E-22
-------
Eljarrat E: de la Cal A: Barcelo. D. (2004). Determination of decabromodiphenyl ether in sediments using
selective pressurized liquid extraction followed by GC-NCI-MS. Anal Bioanal Chem 378: 610-614.
http://dx.doi.org/10.1007/s00216-003-2220-7
Eljarrat. E: De La Cal A: Larrazabal D: Fabrellas. B: Fernandez-Alba. AR. : Borrull F.: Marce. RM: Barcelo.
IX (2005). Occurrence of polybrominated diphenylethers, polychlorinated dibenzo-p-dioxins, dibenzofurans
and biphenyls in coastal sediments from Spain. Environ Pollut 136: 493-501.
Eljarrat. E: Labandeira. A: Marsh. G: Raldua. D: Barcelo. D. (2007). Decabrominated diphenyl ether in river fish
and sediment samples collected downstream an industrial park. Chemosphere 69: 1278-1286.
http://dx.doi.0rg/10.1016/i.chemosphere.2007.05.052
Gouin. T: Thomas. GO: Chaemfa. C: Harner. T: Mackav. D: Jones. KG. (2006). Concentrations of
decabromodiphenyl ether in air from Southern Ontario: Implications for particle-bound transport.
Chemosphere 64: 256-261. http://dx.doi.0rg/10.1016/i.chemosphere.2005.12.071
Guzzella. L: Roscioli C: Binelli. A. (2008). Contamination by polybrominated diphenyl ethers of sediments
from the Lake Maggiore basin (Italy and Switzerland). Chemosphere 73: 1684-1691.
http://dx.doi.0rg/10.1016/i.chemosphere.2008.06.073
Hale. RC: Alaee. M: Manchester-Neesvig. JB: Stapleton. HM: Ikonomou. MG. (2003). Polybrominated diphenyl
ether flame retardants in the North American environment [Review]. Environ Int 29: 771-779.
http://dx.doi.org/10.1016/S0160-4120(03)00113-2
Hale. RC: La Guardia. MJ: Harvey. EP: Gaylor. MO: Mainor. TM: Duff. WH. (2001). Flame retardants.
Persistent pollutants inland-applied sludges. Nature 412: 140-141. http://dx.doi.org/10.1038/35084130
Han. JH: Lee. EJ: Lee. JH: So. KP: Lee. YH: Bae. GN: Lee. SB: Ji. JH: Cho. MH: Yu. IJ. (2008). Monitoring
multiwalled carbon nanotube exposure in carbon nanotube research facility. Inhal Toxicol 20: 741-749.
http://dx.doi.org/10.1080/08958370801942238
Harrad. S: Ibarra. C: Abdallah. MA. -E: Boon. R: Neels. H: Covad A. (2008a). Concentrations of brominated
flame retardants in dust from United Kingdom cars, homes, and offices: causes of variability and
implications for human exposure. Environ Int 34: 1170-1175. http://dx.doi.0rg/10.1016/i.envint.2008.05.001
Harrad. S: Ibarra. C: Diamond. M: Melymuk. L: Robson. M: Douwes. J: Roosens. L: Dirtu. AC: Covad A.
(2008b). Polybrominated diphenyl ethers in domestic indoor dust from Canada, New Zealand, United
Kingdom and United States. Environ Int 34: 232-238. http://dx.doi.0rg/10.1016/j.envint.2007.08.008
Hoh. E: Kites. RA. (2005). Brominated flame retardants in the atmosphere of the East-Central United States.
Environ Sci Technol 39: 7794-7802. http://dx.doi.org/10.1021/es050718k
Hoh. E: Zhu. L: Kites. RA. (2005). Novel flame retardants, l,2-bis(2,4,6-tribromophenoxy)ethane and 2,3,4,5,6-
pentabromoethylbenzene, in United States' environmental samples. Environ Sci Technol 39: 2472-2477.
http://dx.doi.org/10.1021/es048508f
Ikonomou. MG: Fischer. M: He. T: Addison. RF: Smith. T. (2000). Congener patterns, spatial and temporal
trends of polybrominated diphenyl ethers in biota samples from the Canadian west coast and the Northwest
Territories. Organohalogen Compounds 47: 77-80.
Ikonomou. MG: Rayne. S: Addison. RF. (2002). Exponential increases of the brominated flame retardants,
polybrominated diphenyl ethers, in the Canadian Arctic from 1981 to 2000. Environ Sci Technol 36: 1886-
1892. http://dx.doi.org/10.1021/es011401x
Johnson-Restrepo. B: Kannan. K. (2009). An assessment of sources and pathways of human exposure to
polybrominated diphenyl ethers in the United States. Chemosphere 76: 542-548.
http://dx.doi.0rg/10.1016/i.chemosphere.2009.02.068
Johnson. A: Olson. N. (2001). Analysis and occurrence of polybrominated diphenyl ethers in Washington state
freshwater fish. Arch Environ Contam Toxicol 41: 339-344. http://dx.doi.org/10.1007/s002440010257
Johnson. PR: Methner. MM: Kennedy. AJ: Steevens. JA. (2010). Potential for occupational exposure to
engineered carbon-based nanomaterials in environmental laboratory studies. Environ Health Perspect 118:
49-54. http://dx.doi.org/10.1289/ehp.0901076
E-23
-------
Knoth. W: Mann. W: Meyer. R: Nebhuth. J. (2007). Polybrominated diphenyl ether in sewage sludge in
Germany. Chemosphere 67: 1831-1837. http://dx.doi.0rg/10.1016/j.chemosphere.2006.05.113
Kupper. T: de Alencastro. LF: Gatsigazi. R: Furrer. R: Grandjean. D: Tarradellas. J. (2008). Concentrations and
specific loads of brominated flame retardants in sewage sludge. Chemosphere 71: 1173-1180.
http://dx.doi.0rg/10.1016/i.chemosphere.2007.10.019
La Guardia. MJ: Hale. RC: Harvey. E. (2007). Evidence of debromination of Decabromodiphenyl Ether (BDE-
209) in biota from a wastewater receiving stream. Environ Sci Technol 41: 6663 -6670.
http://dx.doi.org/10.1021/es070728g
Law. K: Halldorson. T: Danell R: Stern. G: Gewurtz. S: Alaee. M: Marvin. C: Whittle. M: Tomy. G. (2006a).
Bioaccumulation and trophic transfer of some brominated flame retardants in a Lake Winnipeg (Canada)
food web. Environ Toxicol Chem 25: 2177-2186.
Law. RJ: Allchin. CR: de Boer. J: Covaci. A: Herzke. D: Lepom. P: Morris. S: Tronczynski. J: de Wit CA.
(2006b). Levels and trends of brominated flame retardants in the European environment. Chemosphere 64:
187-208. http://dx.doi.0rg/10.1016/i.chemosphere.2005.12.007
Lebeuf. M: Love. K: Trottier. S. (2001). Polybrominated diphenyl ethers in beluga whales (Delphinapterus
leucus) from the St. Lawrence Estuary, Canada: Levels and temporal trends. In Abstracts of the 2nd
International Workshop on Brominated Flame Retardants, BFR 2001, May 1416. Stockholm, Sweden:
Stockholm University.
Lee. JH: Lee. SB: Bae. GN: Jeon. KS: Yoon. JU: Ji. JH: Sung. JH: Lee. BG: Yang. JS: Kim. HY: Kang. CS: Yu.
II (2010). Exposure assessment of carbon nanotube manufacturing workplaces. Inhal Toxicol 22: 369-381.
http://dx.doi.org/10.3109/08958370903367359
Luo. Y: Luo. XJ: Lin. Z: Chen. SJ: Liu. J: Mai. BX: Yang. ZY. (2009). Polybrominated diphenyl ethers in road
and farmland soils from an e-waste recycling region in Southern China: Concentrations, source profiles, and
potential dispersion and deposition. Sci Total Environ 407: 1105-1113.
http://dx.doi.0rg/10.1016/i.scitotenv.2008.10.044
Luross. JM: Alaee. M: Sergeant DB: Cannon. CM: Whittle. DM: Solomon. KR: Muir. DC. (2002). Spatial
distribution of polybrominated diphenyl ethers and polybrominated biphenyls in lake trout from the
Laurentian Great Lakes. Chemosphere 46: 665-672.
Ma. J: Addink. R: Yun. S: Cheng. J: Wang. W: Kannan. K. (2009). Polybrominated dibenzo-p-dioxins/
dibenzofurans and polybrominated diphenyl ethers in soil, vegetation, workshop-floor dust, and electronic
shredder residue from an electronic waste recycling facility and in soils from a chemical industrial complex
in eastern China. Environ Sci Technol 43: 7350-7356.
Mai. B: Chen. S: Luo: Chen: Yang: Sheng: Peng: Fu: Zeng. EY. (2005). Distribution of Polybrominated
Diphenyl Ethers in Sediments of the Pearl River Delta and Adjacent South China Sea. Environ Sci Technol
39: 3521-3527. http://dx.doi.org/10.1021/es048083x
Methner. M: Hodson. L: Dames. A: Geraci. C. (2010). Nanoparticle emission assessment technique (NEAT) for
the identification and measurement of potential inhalation exposure to engineered nanomaterials - Part B:
Results from 12 field studies. J Occup Environ Hyg 7: 163-176.
http://dx.doi.org/10.1080/15459620903508066
Muenhor. D: Harrad. S: All N: Covaci. A. (2010). Brominated flame retardants (BFRs) in air and dust from
electronic waste storage facilities in Thailand. Environ Int 36: 690-698.
http://dx.doi.0rg/10.1016/i.envint.2010.05.002
Naert. C: Van Peteghem. C: Kupper. J: Jenni. L: Naegeli. H. (2007). Distribution of polychlorinated biphenyls
and polybrominated diphenyl ethers in birds of prey from Switzerland. Chemosphere 68: 977-987.
http://dx.doi.0rg/10.1016/i.chemosphere.2007.01.009
Norstrom. RJ: Simon. M: Moisev. J: Wakeford. B: Weseloh. DV. (2002). Geographical distribution (2000) and
temporal trends (1981-2000) of brominated diphenyl ethers in Great Lakes hewing gull eggs. Environ Sci
Technol 36: 4783-4789. http://dx.doi.org/10.1021/es025831e
E-24
-------
North. KD. (2004). Tracking polybrominated diphenyl ether releases in a wastewater treatment plant effluent,
Palo Alto, California. Environ Sci Technol 38: 4484-4488. http://dx.doi.org/10.1021/es049627y
Offenberg. JH: Stapleton. HM: Strynar. MJ: Lindstrom. AB. (2006). Polybrominated diphenyl ethers in U.S.
soils. Dioxin 2006, August 21-25, 2006, Oslo, Norway.
Pros. PR: Hoover. D: Rodigari. F: Crane. D: Sericano. J. (2005). Levels and distribution of polybrominated
diphenyl ethers in water, surface sediments, and bivalves from the San Francisco Estuary. Environ Sci
Technol 39: 33-41. http://dx.doi.org/10.1021/es048905q
Potter. KE: Watts. BD: La Guardia. MJ: Harvey. EP: Hale. RC. (2009). Polybrominated diphenyl ether flame
retardants in Chesapeake Bay region, USA, peregrine falcon (Falco peregrinus) eggs: urban/rural trends.
Environ Toxicol Chem 28: 973-981. http://dx.doi.0rg/10.1897/08-350.l
Qiu. X: Marvin. CH: Kites. RA. (2007). Dechlorane plus and other flame retardants in a sediment core from
Lake Ontario. Environ Sci Technol 41: 6014-6019. http://dx.doi.org/10.1021/es070810b
Raff. J: Kites. R. (2004). Polybrominated diphenyl ethers in Mississippi River suspended sediment.
Organohalogen Compounds 66: 3673-3677.
Rayne. S: Ikonomou. MG: Antcliffe. B. (2003). Rapidly increasing polybrominated diphenyl ether
concentrations in the Columbia River system from 1992 to 2000. Environ Sci Technol 37: 2847-2854.
http://dx.doi.org/10.1021/es0340073
Ricklund. N: Kierkegaard. A: McLachlan. MS. (2010). Levels and potential sources of decabromodiphenyl
ethane (DBDPE) and decabromodiphenyl ether (DecaBDE) in lake and marine sediments in Sweden.
Environ Sci Technol 44: 1987-1991. http://dx.doi.org/10.1021/es903701q
Ricklund. N: Kierkegaard. A: McLachlan. MS: Wahlberg. C. (2009). Mass balance of decabromodiphenyl
ethane and decabromodiphenyl ether in a WWTP. Chemosphere 74: 389-394.
http://dx.doi.0rg/10.1016/i.chemosphere.2008.09.054
Salamova. A: Kites. RA. (2011). Discontinued and alternative brominated flame retardants in the atmosphere
and precipitation from the great lakes basin. Environ Sci Technol 45: 8698-8706.
http://dx.doi.org/10.1021/es2020378
Sawal. G: Stachel B: Lepom. P. (2004). Polybrominated diphenyl ethers in sediments from the River Elbe,
Germany. In Proceedings of the Third International Workshop on Brominated Flame Retardants.
http://www.bfr2010.com/abst/2004/Individual%20Papers/BFR2004%20Abstract%20032%20Sawal.pdf
Schecter. A: Papke. O: Joseph. JE: Tung. KG. (2005). Polybrominated diphenyl ethers (PBDEs) in U.S.
computers and domestic carpet vacuuming: possible sources of human exposure. J Toxicol Environ Health A
68: 501-513. http://dx.doi.org/10.1080/15287390590909715
Schlabach. M: Field. E: Gundersen. H: Mariussen. E: Kjellberg. G: Breivik. E. (2004). Pollution of Lake Mjesa
by brominated flame retardants. 66: 3779-3785.
Sellstrom. U: de Wit. CA: Lundgren. N: Tysklind. M. (2005). Effect of sewage-sludge application on
concentrations of higher-brominated Diphenyl Ethers in soils and earthworms. Environ Sci Technol 39:
9064-9070. http://dx.doi.org/10.1021/es051190m
Sharp. R: Lunder. S. (2004). In the dust: Toxic fire retardants in American homes. Washington, DC:
Environmental Working Group, http://www.ewg.org/research/dust
She. J: Petreas. M: Winkler. J: Visita. P: Mckinnev. M: Kopec. D. (2002). PBDEs in the San Francisco Bay
Area: measurements in harbor seal blubber and human breast adipose tissue. Chemosphere 46: 697-707.
Sjodin. A: Papke. O: Mcgahee. E: Focant JF: Jones. RS: Pless-Mullolj T: Toms. LM: Herrmann. T: Miiller. J:
Needham LL: Patterson. DG. (2008). Concentration of polybrominated diphenyl ethers (PBDEs) in
household dust from various countries. Chemosphere 73: S131-S136.
http://dx.doi.0rg/10.1016/i.chemosphere.2007.08.075
E-25
-------
Song. W: Ford. JC: Li. A: Mills. WJ: Buckley. PR: Rockne. KJ. (2004). Polybrominated diphenyl ethers in the
sediments of the Great Lakes. 1. Lake Superior. Environ Sci Technol 38: 3286-3293.
http://dx.doi.org/10.1021/es035297q
Song. WL: Ford. JC: Li. A: Sturchio. NC: Rockne. KJ: Buckley. PR: Mills. WJ. (2005a). Polybrominated
diphenyl ethers in the sediments of the Great Lakes. 3. Lakes Ontario and Erie. Environ Sci Technol 39:
5600-5605. http://dx.doi.org/10.1021/es050631z
Song. WL: Li. A: Ford. JC: Sturchio. NC: Rockne. KJ: Buckley. PR: Mills. WJ. (2005b). Polybrominated
diphenyl ethers in the sediments of the great lakes. 2. Lakes Michigan and Huron. Environ Sci Technol 39:
3474.3479. http://dx.doi.org/10.1021/es048291p
Stapleton. HM: Dodder. NG: Offenberg. JH: Schantz. MM: Wise. SA. (2005). Polybrominated diphenyl ethers
in house dust and clothes dryer lint. Environ Sci Technol 39: 925-931.
Stern. GA: Ikonomou. M. (2000). Temporal trends of polybrominated diphenyl ethers in SE Baffin beluga:
Increasing evidence of long range atmospheric transport. Organohalogen Compounds 47: 8184.
Strandberg. B: Dodder. NG: Basu. I: Kites. RA. (2001). Concentrations and spatial variations of polybrominated
diphenyl ethers and other organohalogen compounds in Great Lakes air. Environ Sci Technol 35: 1078-1083.
http://dx.doi.org/10.1021/es001819f
Su. Y: Hung. H: Sverko. E: Fellin. P: Li. H. (2007). Multi-year measurements of polybrominated diphenyl ethers
(PBDEs) in the Arctic atmosphere. Atmos Environ 41: 8725-8735.
http://dx.doi.0rg/10.1016/i.atmosenv.2007.07.032
Takava. M: Serita. F: Ono-Ogasawara. M: Shinohara. Y: Saito. H: Koda. S. (2010). [Airborne particles in a
multi-wall carbon nanotube production plant: observation of particle emission and personal exposure 1:
Measurement in the packing process]. Sangyo EiseigakuZasshi52: 182-188.
Toms. L: Mueller. J: Mortimer. M: Symons. R: Stevenson. G: Gaus. C. (2006). Assessment of concentrations of
polybrominated diphenyl ether flame retardants in aquatic environments in Australia. Australian Government
Department of the Environment and Heritage.
http://www.environment.gov.au/settlements/publications/chemicals/bfr/aquatic.html
U.S. EPA (U.S. Environmental Protection Agency). (2009). Targeted national sewage sludge survey statistical
analysis report [EPA Report]. (EPA-822-R-08-018). Washington, DC.
U.S. EPA (U.S. Environmental Protection Agency). (2010). An exposure assessment of polybrominated
diphenyl ethers [EPA Report]. (EPA/600/R-08/086F). Washington, DC.
http ://cfpub. epa. gov/ncea/cfm/recordisplav. cfm?deid=210404
Voorspoels. S: Covaci. A: Jaspers. VLB: Neels. H: Schepens. P. (2007). Biomagnification of PBDEs in three
small terrestrial food chains. Environ Sci Technol 41: 411-416. http://dx.doi.org/10.1021/es061408k
Voorspoels. S: Covaci. A: Lepom. P: Escutenaire. S: Schepens. P. (2006a). Remarkable findings concerning
PBDEs in the terrestrial top-predator red fox (Vulpes vulpes). Environ Sci Technol 40: 2937-2943.
http://dx.doi.org/10.1021/es060081k
Voorspoels. S: Covaci. A: Lepom. P: Jaspers. VL: Schepens. P. (2006b). Levels and distribution of
polybrominated diphenyl ethers in various tissues of birds of prey. Environ Pollut 144: 218-227.
http://dx.doi.0rg/10.1016/i.envpol.2005.12.027
Voorspoels. S: Covaci. A: Schepens. P. (2004). PBDEs in sediments from a polluted area in Europe: The
Belgian North Sea, the Western Scheldt Estuary and tributaries. In Proceedings of the Third International
Workshop on Brominated Flame Retardants.
http://www.bfr2010.com/abst/2004/Individual%20Papers/BFR2004%20Abstract%20027%20Voorspoels.pdf
Wakeford. BJ: Simon. MJ: Elliott JE: Braune. BM. (2002). Analysis of polybrominated diphenyl ethers (BDEs)
in wildlife tissues Canadian Wildlife Service contributions. In Abstracts of the 4th Annual Workshop on
Brominated Flame Retardants in the Environment, June 1718. Burlington, Ontario: Canada Centre for Inland
Waters.
E-26
-------
Wang. Y: Zhang. Q: Lv. J: Li. A: Liu. H: Li. G: Jiang. G. (2007). Polybrominated diphenyl ethers and
organochlorine pesticides in sewage sludge of wastewater treatment plants in China. Chemosphere 68: 1683-
1691. http://dx.doi.0rg/10.1016/i.chemosphere.2007.03.060
Watkins. DJ: Mcclean. MD: Fraser. AJ: Weinberg. J: Stapleton. HM: Sjodin. A: Webster. TF. (2011). Exposure
to PBDEs in the office environment: evaluating the relationships between dust, handwipes, and serum.
Environ Health Perspect 119: 1247-1252. http://dx.doi.org/10.1289/ehp.1003271
Wu. N: Webster. T: Hermann. T: Paepke. O: Tickner. J: Hale. R: Harvey. E: La Guardia. M: Jacobs. E. (2007).
Associations of PBDE levels in breast milk with diet and indoor dust concentrations. Organohalogen
Compounds 67: 654-656.
Yun. SH: Addink. R: McCabe. JM: Ostaszewskj A: Mackenzie-Taylor. D: Taylor. AB: Kannan. K. (2008).
Polybrominated diphenyl ethers and polybrominated biphenyls in sediment and floodplain soils of the
Saginaw River watershed, Michigan, USA. Arch Environ Contam Toxicol 55: 1-10.
http://dx.doi.org/10.1007/s00244-007-9084-3
Zhao. X: Zhang. H: Ni. Y: Lu. X: Zhang. X: Su. F: Fan. J: Guan. D: Chen. J. (2011). Polybrominated diphenyl
ethers in sediments of the Daliao River Estuary, China: levels, distribution and their influencing factors.
Chemosphere 82: 1262-1267. http://dx.doi.0rg/10.1016/j.chemosphere.2010.12.032
Zhu. LY: Kites. RA. (2005). Brominated flame retardants in sediment cores from lakes michigan and erie.
Environ Sci Technol 39: 3488-3494. http://dx.doi.org/10.1021/es048240s
Zou. M. -Y: Ran. Y: Gong. J: Mai. B. -X: Zeng. EY. (2007). Polybrominated diphenyl ethers in watershed soils
of the Pearl River Delta, China: Occurrence, inventory, and fate. Environ Sci Technol 41: 8262-8267.
http://dx.doi.org/10.1021/es071956d
E-27
-------
Appendix F. Toxicological and
Ecological Effects
Appendix F comprises tables summarizing the material characteristics, study design, and results
of select toxicokinetic and toxicological studies for BDE-209 (a specific congener of decabrominated
diphenyl ether) and multiwalled carbon nanotubes (MWCNTs). Much of the relevant toxicological data
for BDE-209 has been summarized in reviews, and many of the toxicological endpoints of concern have
been identified and used by regulatory agencies to establish reference values for the protection of human
health and the environment (see Chapter 5). Because the BDE-209 studies have been described in detail
in many previous reviews, a relatively small subset of the BDE-209 studies discussed in Chapter 5 of this
document are summarized in the tables of this appendix. Select BDE-209 study summary tables are
generally provided only for general comparison to MWCNT study summary tables or for the primary
exposure route of concern (oral).
By comparison, the toxicokinetics and toxicological effects of MWCNTs are not well understood
and only one draft reference value has been established for the protection of human health (see Chapter
5). Moreover, as emphasized throughout this document, variations in certain physicochemical
characteristics of MWCNTs are likely to affect their behavior in biological systems and impacts to
humans and biota. Appendix F therefore provides summary tables for most of the MWCNT toxicokinetic
and toxicological studies referenced in Chapter 5 of this case study. These tables supply more detailed
information on material characteristics, study design, and observed effects than was presented in the text.
Appendix F provides information on the absorption, distribution, metabolism, and elimination of
BDE-209 and MWCNTs from identified toxicokinetic studies (Section F.I.I). In addition, summary
tables present information from select in vivo studies reporting effects other than carcinogenicity using
the dermal, ocular, inhalation, and oral routes of exposure (Section F.I.2); in vitro studies including those
investigating genotoxicity and mutagenicity (Section F.I.3); and carcinogenicity studies (Section F.I.4)
for BDE-209 and MWCNTs.
Appendix F also provides summary tables of data from studies that investigated effects of
BDE-209, other polybrominated diphenyl ethers (PBDEs), and MWCNTs on different types of biota.
Data from studies in aquatic ecosystems are summarized in Sections F.2.1 and F.2.2; data from terrestrial
ecosystems are summarized in Sections F.2.3 and F.2.4.
Literature was identified primarily using review articles published in the past two years. Targeted
literature searches were carried out as needed.
F-1
-------
F.1. Toxicological Effects
F.1.1. Toxicokinetic Studies
Table F-1. Select toxicokinetic studies for decaBDE.
Route of
exposure
Oral
(diet)
i.v.
Oral
(diet)
Description
of decaBDE
Unlabeled decaBDE
(92% pure) and
[U-14C] decaBDE
(98.9% pure)
[U-14C] decaBDE
(98.9% pure)
Unlabeled decaBDE
(92% pure) and
[U-14C] decaBDE
(98.9% pure)
Species
Fischer 344 rat
(male)
Fischer 344 rat
(male)
Fischer 344 rat
(male)
Exposure
duration
12 days
(sacrificed 24,
48, 72 hours
after exposure
to [U-14C] on
day 8)
72 hours
12 days
(sacrificed
72 hours after
exposure to
[U-14C] on
Doses tested
0.
4.
1.
0.
0.
0.
2.
,0277%,
80% diet
07 mg/kg
,025,
,0509,
250, 0.487,
49, 4.99
Effects observed
Excretion results: urine 0.004-0.012%, feces 82.5-
86.4% (recovery not related to dose); tissue
recovery: 0.109% in liver, 0.248% in muscle, 0.136% in
skin (other smaller quantities reported); for all tissues the
maximum percent (%) in organs and tissues was
reported in the low-dose group; for both doses percent
(%) of dose remaining in the gut contents and gut tissues
decreased with time after exposure
Excretion results: urine 0.129%, feces 70.0%; tissue
recovery: 4.27% in liver, 5.063% in Gl, 12.9% in muscle,
7.25% in skin, 2.99% in fat (other smaller quantities
reported)
Recovery of radiolabeled decaBDE in feces ranged from
91.3-101% of the amount ingested; recovery was not
related to dose; liver weights of rats increased as dose
increased
Citation
el Dareer
(1987)
el Dareer
(1987)
el Dareer
(1987)
etal.
etal.
etal.
% diet
day 8)
F-2
-------
Table F-1 (Continued): Select toxicokinetic studies for decaBDE.
Route of
exposure
Oral
(diet)
Oral (diet)
Oral
Oral
(diet)
Oral
Description
of decaBDE
[U-14C] decaBDE
(97.9-99.2% pure)
diluted with unlabeled
decaBDE
[U-14C] decaBDE
(97.9-99.2% pure)
diluted with unlabeled
decaBDE
[U-14C] decaBDE
(assumed to be77.4%
pure based on
reference description)
>99.8% pure
77.4% decaBDE,
21.8% nonaBDE, and
0.8% octaBDE
Exposure
Species duration
Rat >8 days
(7 days
unlabeled,
1 day labeled,
then returned
to unlabeled
diet for
remainder of
holding period)
Rat >8 days
(7 days
unlabeled,
1 day labeled,
then returned
to unlabeled
diet for
remainder of
holding period)
Sprague- 16 days
Dawley rat
(male and
female)
Pregnant Wistar 96 hours
rat (female) (gestation days
16-19)
Rat NR
Doses tested Effects observed
250-50,000 Excretion results: feces 61%, urine 0.1%
ppm
250-50,000 Excretion results: urine 0.01%, feces >99% in 72 hours
ppm
1 mg/kg Excretion results: urine <1.0%, feces 90.6% (day 1),
>8.4% (day 2), >99% (at 48 hours); tissue recovery:
limited absorption to Gl at 1, 3, and 16 days; 0.06% in
spleen, 0.01% in adrenals (no others reported) at 16
days
2.61 mg/kg- >19% recovered in tissues; efficient absorption reported;
day highest residue concentrations in endocrine glands and
the liver; most of recovered product was unchanged
decaBDE with 9-27% biotransformation products (nona-
and octaBDEs) in tissues and 14% in fetuses; main
metabolic pathways are debromination and oxidation
NR Slight accumulation occurring very slowly over time in
adipose tissue
Citation
NTP (1986)
NTP (1986)
Morris et al.
(1975)
Riu et al. (2008)
Great Lakes
(1976) and IRDC
(1976, 1977) as
cited in NRC
(2000)
F-3
-------
Table F-1 (Continued): Select toxicokinetic studies for decaBDE.
Route of Description
exposure of decaBDE
Oral 77.4% pure
(diet)
Oral >98% pure, specific
(gavage) activity 17.5 Ci/mol
Oral 98.5% pure
(diet)
Oral NR
(diet)
Oral NR
Species
Rat
Sprague-
Dawley rat
(male)
Sprague-
Dawley rat
(male)
Lactating cow
(female)
Sprague-
Dawley rat
(male)
Exposure
duration
2 years
3, 7 days
21 days
followed by
21 -day
withdrawal
period
3 months
90 days
Doses tested
0; 0.01; 0.1; 1
mg/kg-day
3 u mo I/kg,
15 Ci/mol,
1 mL/kg
volume
0.3 ug/gram
of diet
Naturally
contaminated
diet (not
measured)
1 00 mg/kg
bw-day
Effects observed
No increase in the kidney, muscle, or serum
>10% absorbed; 90% excreted in feces (65%
metabolites); 10% excretion in bile (mostly metabolites)
After 21 days 5% of decaBDE was measured as BDE-
209 (<4% in feces); nona- and octaBDEs were also
present; BDE-209 was highest in the liver, followed by
the Gl track; several lower congeners were present at
higher concentrations than could be attributed directly to
dose impurities as the result of debromination
BDE-209 was dominant congener in all tissue samples
except milk (milk concentrations were generally low);
dominant output route was feces; congener profiles in
adipose tissue and feed differed; BDE-207, BDE-196,
BDE-197, and BDE-182 accumulated to a greater extent
in the fat compared to their isomers suggesting
metabolic debromination of BDE-209; indicates that
meat may be a more important human exposure route to
higher brominated BDEs than dairy products
Preferential accumulation of BDE-209 in the liver; BDE-
209 induced hepatotoxicity (indicated by serum clinical
chemistry data for AST, ALP, T-CHO, HDL-C, Cr, and
TEA); significantly increased CYP2B1 expression in
mRNA; metabolites of BDE-183, 196, 197, 202, 203,
206, 207, and 208 were all found in kidney and liver
tissues (207 most prominent)
Citation
Dow (1994) as
cited in NRC
(2000)
Morck et al.
(2003)
Huwe and Smith
(2007)
Kierkegaard et al.
(2007)
Wang et al.
(2010)
F-4
-------
Table F-1 (Continued): Select toxicokinetic studies for decaBDE.
Route of Description
exposure of decaBDE
In vitro 98 ±1% pure
(incubation)
Oral 98% pure
Gavage Unlabeled BDE-209
or i.v. (>98% pure)
Liver BDE-209
microsomal
depletion
Species
Human
hepatocytes;
2 cryopreserved
(1 male,
1 female),
1 fresh (male)
Sprague-
Dawley rat
(female)
Sprague-
Dawley rat
(male)
Harbor seal
(Phoca vitulina)
and sperm
whale (Physeter
catodon)
Exposure
duration
48-hour
exposure for
cryopreserved
cell cultures;
1 dose per
24 hours for
three days for
fresh
hepatocyte
cultures
GD7toPND4
1, 3,6,24,48,
72, 96, 120, or
144 hours
NR
Doses tested Effects observed
10 nmol/well No hydroxylated or debrominated metabolites observed;
Up-regulation of genes encoding for cytochrome P450
monooxygenase (CYP) 1A2, CYP3A4, deiodinase type
1, and glutathione S-transferase M1
5 umol/kg Increased accumulation with time in maternal blood,
placenta, fetuses, and neonates; more BDE-209 found in
neonate whole-body samples obtained during lactation
than fetal whole-body samples during pregnancy;
increased nonaBDE in maternal blood and placenta over
time; slight changes observed for octaBDEs in maternal
blood and placenta; significant decrease observed in the
fetuses or neonates for BDE-196 and 198/203
2 umol/mL Bioavailability calculated to be >26%; 13 metabolites
were identified in the plasma (octa- nona-, and
hexaBDEs) at concentrations 4 times higher than the
parent compound on days 3 and 7; BDE-209 was rapidly
distributed to well perfused tissues (e.g., liver)
31 ug/mL No detectable depletion of parent BDE-209; lack of
microsomal depletion consistent with persistent and
accumulative nature of BDE-209
Citation
Stapleton et al.
(2009)
Cai et al. (2011)
Sandholm et al.
(2003)
de Boer et al.
(1998,2000)as
cited in Hakkand
Letcher (2003)
F-5
-------
Table F-1 (Continued): Select toxicokinetic studies for decaBDE.
Route of
exposure
Oral
Oral (diet)
Description
of decaBDE
DecaBDE (BDE-209)
BDE-209 with
detectable amounts
of nona- and
octaBDEs
Exposure
Species duration
Sprague- 72 hours
Dawley rat
(male)
Rainbow trout 16, 49, 120
(Oncorhynchus days
mykiss) (male
and female)
Doses tested Effects observed
3 mg/kg Excretion results: urine <0.05%, feces: >90%, bile: 9.5%;
tissue recovery: 0.9 in liver, 3.5 in Gl, 0.7 in muscle
(other smaller quantities reported); rats metabolized
BDE-209 to fecal metabolites (including debrominated
mono-OH- and ortho-MeO-OH-BDEs) via oxidative
debromination
1.7-10 Low uptake efficiency; elevated levels in liver and
mg/kg-day muscle (20-40 times greater in liver, 560 ±210 ng/gram
fresh wt to 870 ± 220 ng/gram fresh wt from day 16 to
day 120 for liver and 10 ± 3.2 ng/gram fresh wt to
38 ± 14 ng/gram fresh wt from day 16 to day 120 for
muscle), decreased upon depuration; metabolites
detected in liver and muscle tissues, not all metabolites
decreased with depuration
Citation
Morck and
Klasson-Wehler
(2001) as cited in
Hakk and Letcher
(2003)
Kierkegaard et al.
(1999)
Abbreviations: Gl = Gastrointestinal tract; NR = Not reported
F-6
-------
Table F-2. Select toxicokinetic studies for MWCNTs.
Route of Diameter Length
exposure (nm) (urn)
Surface
area
(m2/gram) Purity
Species
Exposure duration
Doses
tested Effects observed Citation
Absorption
Intratracheal 10-20 0.01-
instillation; 0.6
i.v.; and
gavage
Inhalation 20-50 0.5-2
(intratracheal
instillation)
Inhalation Mean: Median:
(aspiration) 49 ± 13.4 3.86
Oral 10-20 0.01-
(gavage) °-6
NR >95%
NR (% w/w):
0.53 Ni, 0.08
S, <0.02 Mg,
<0.01 Na,
<0.01 V
NR 0.78%;
0.41%
sodium,
0.32% iron
NR >95%
Kunming
mouse
(male)
Sprague-
Dawley rat
(male)
C57BL/6
mouse
(male)
Kunming
mouse
(male)
Single exposure;
28-day observation
Single exposure;
1-, 7-, 30-, 90-, and
180-day observations
and 6-month MWCNT
elimination
observation
Single exposure;
1-, 7-, 28- and 56-day
observations
Single exposure;
12-hour observation
10ug 20% of administered dose Deng et al.
remained within the lung at (2007)
28 days (intratracheal instillation);
80% accumulate in the liver and
remain at 28 days (i.v.); only
levels measured in stomach,
large and small intestines, 74%
directly excreted (gavage)
1,10,100 MWCNTs did not significantly Elgrabli et al.
ug/rat cross the pulmonary barrier; (2008b)
MWCNTs were evident within the
lungs at 6 months
10,20, MWCNTs reached the pleura and Porter et al.
40, 80 ug induced pleural inflammation at (2010)
56 days
10ug Majority of MWCNTs evident in Deng et al.
feces, stomach, and small and (2007)
large intestines; no detectable
transport into the blood;
MWCNTs remained unchanged
suggesting biopersistence if not
excreted
F-7
-------
Table F-2 (Continued): Select toxicokinetic studies for MWCNTs.
Route of
exposure
Diameter Length
(nm) (Mm)
Surface
area
(m2/gram)
Purity
Species Exposure duration
Doses
tested
Effects observed
Citation
Distribution
Inhalation 10-50 <0.1-10 NR NR
Inhalation NR 0.9- 197 NR
(intratracheal 0.15
instillation)
C57BL6
mouse
(male)
Wistar
albino rat
(male)
Single 6-hour
exposure; 14-week
observation
Single exposure;
24 hour, 1 week,
1- and 3-month
observations
1, 30
mg/m3
0.2, 1, 5
mg/kg
MWCNTs reached the subpleura;
nanotubes were embedded in
subpleural wall and within
subpleural macrophages
MWCNTs translocated from the
lung to liver and kidney (not to
the heart) at 1 month
Ryman-
Rasmussen et
al. (2009a)
Reddy et al.
(2010)
Elimination
Inhalation 20-50 0.5-2 NR
(intratracheal
instillation)
(% w/w):
0.53 Ni, 0.08
S, <0.02 Mg,
<0.01 Na,
<0.01 V
Sprague-
Dawley rat
(male)
Single exposure; 1-,
7-, 30-, 90-, and
180-day observations
and 6-month MWCNT
elimination
observation
1, 10, 100
ug/rat
Following phagocytosis of the
MWCNTs, the macrophages
underwent apoptosis, with no
inflammatory response or other
physiological and histological
pathology
Elgrabli et
(2008a)
al.
Abbreviations: NR = Not reported
F-8
-------
F.1.2. In Vivo Studies (Excluding Carcinogenicity Studies)
Table F-3. Select dermal and ocular studies for decaBDE.
Route of
exposure
Description
of decaBDE Species
Exposure
duration
Doses
tested
Effects observed Citation
Irritation
Dermal
Dermal
Dry solid New Zealand
albino rabbit
NR Rabbit
24 hours, 3 days,
2 weeks
Single
administration 24
hours; 14 days
observation
500 mg,
reported by
NRC (2000)
200, 2,000
mg/kg
No dermal response in intact skin; Norris et al. (1975) [also reported
no indication of bromacne in NTP (1986); Norris et al. (1973);
Dow (1972) and IRDC (1974) as
cited in NRC (2000)1
Slight erythematous and edematous IRDC (1974) and Great Lakes
response in abraded skin (1977) as cited in NRC (2000):
Norris et al. (1975)
Sensitization
Dermal
Systemic
Dermal
Homogenous 5% Human
suspension in
petrolatum; 77.4%
decaBDE, 21.8%
nonaBDE, 0.8%
octaBDE
effects
NR Rabbit
3 times per week
for 3 weeks
Single
administration 24
hours observation
NR
200, 2,000
mg/kg
No skin sensitization response Norris et al. (1975) [also reported
in NTP (1986); Norris et al. (1973):
Dow (1972) as cited in NRC
(2000)]
No treatment-related effects in body IRDC (1974) and Great Lakes
weight gain or survival (1977) as cited in NRC (2000)
F-9
-------
Table F-3 (Continued): Select dermal and ocular studies for decaBDE.
Route of Description
exposure of decaBDE
Species
Exposure
duration
Doses
tested
Effects observed
Citation
Ocular
Ocular Saytex102
Rabbit
Single application 100 mg
No primary eye irritation
Pharmakon (1981) as cited in NRC
(2000)
Ocular Dry solid
New Zealand Single application 100 mg per Transient irritation of conjunctival Norris et al. (1975) [also reported
albino rabbit eye membranes in washed and in NTP (1986): Norris et al. (1973):
unwashed eyes (not sustained past IRDC (1974) Dow (1972) as cited
24 hours) in NRC (2000)1
Abbreviations: NR = Not reported
F-10
-------
Table F-4. Select dermal and ocular studies for MWCNTs.
1%Nikkiso- 44 NR 69
MWCNTs
2% Mitsui 60 NR 23
product of
MWCNTs
MWCNTs Inner: 3-8, 5-9 10-15
outer: 140
±30
MWCNTs Inner: 2-6, 0.1-10 30-45
outer: 10-
15
176 ppmGa, 80
ppm Al, 53 ppm
Fe, 16 ppm Cd,
0.5 ppm Li
3,600 ppm Fe,
14 ppmCr, 6
ppm Bi, 4 ppm
Ni
NR
NR
Kbl:New
Zealand
white rabbit
(male)
Kbl:New
Zealand
white rabbit
(male)
New Zealand
white rabbit
(female)
New Zealand
white rabbit
(female)
4-hour exposure; 1-, 0.5 gram
24-, 48-, and
72-hour observation
4-hour exposure; 1-, 0.5 gram
24-, 48-, and
72-hour observation
4-hour under semi- 0.5 gram
occlusive
conditions; 96-hour
observation
4-hour under semi- 0.5 gram
occlusive
conditions; 96-hour
observation
Exposure resulted in a
primary irritation index
(Pll) of 0.6 (calculated
by mean dermal
response score at 24
hours + mean dermal
response score at 48
hours + mean dermal
response score at 72
hours divided by 3; a
Pll score greater than
5 is considered
positive)
No erythema or edema
was observed
No erythema or edema
at 72 hours
No erythema or edema
at 72 hours
Emaetal. (2011)
(OECD 404
compliant; not GLP
compliant)
Emaetal. (2011)
(OECD 404
compliant; not GLP
compliant)
Kishore et al. (2009)
(OECD 404
compliant)
Kishore et al. (2009)
(OECD 404
compliant)
F-11
-------
Table F-4 (Continued): Select dermal and ocular studies for MWCNTs.
Test Diameter
substance (nm)
Surface
Length area
(urn) (m2/gram) Purity
Species
Exposureduration
Doses
tested
Effects observed
Citation
Skin sensitization
1%Nikkiso- 44
MWCNTs
2% Mitsui 60
product of
MWCNTs
NR 69 176ppmGa, 80
ppm Al, 53 ppm
Fe, 16 ppm Cd,
0.5 ppm Li
NR 23 3,600 ppm Fe,
14 ppmCr, 6
ppm Bi, 4 ppm
Ni
Sic: Hartley
guinea pig
(male)
Slc:Hartley
guinea pig
(male)
3 doses; 6-hour
challenge was
conducted on day
28
3 doses; 6-hour
challenge was
conducted on day
28
0.4 gram
paste
0.4 gram
paste
No sensitization
observed
No sensitization
observed
Emaetal. (2011)
(OECD 406
compliant; Buehler
method)
Emaetal. (2011)
(OECD 406
compliant; Buehler
method)
Ocular irritation
1%Nikkiso- 44
MWCNTs
2% Mitsui 60
product of
MWCNTs
NR 69 176 ppm Ga, 80
ppm Al, 53 ppm
Fe, 16 ppm Cd,
0.5 ppm Li
NR 23 3,600 ppm Fe,
14 ppmCr, 6
ppm Bi, 4 ppm
Ni
Kbl:New
Zealand white
rabbit (male)
Kbl:New
Zealand white
rabbit (male)
Single exposure to
left eye; right eye
served as control
Single exposure to
left eye; right eye
served as control
0.1%,
0.25%
(0.1 mL)
1%
(0.1 mL)
Conjunctival redness
and blood vessel
hyperemia at 1 hour,
not at 24 hours
No eye irritation
observed
Ema et al. (2011)
(OECD 405
compliant)
Ema et al. (2011)
(OECD 405
compliant)
Abbreviations: NR = Not reported
F-12
-------
Table F-5. Select pulmonary exposure studies for decaBDE.
Route of
exposure
Intratracheal
injection
Inhalation
Description
of decaBDE
77.4% purity
decabromodiphenyl
oxide (DBDPO)
dust, respirable
size
DE-83, 97% purity
aerosolized dust
Exposure Doses
Species duration tested
Sprague-Dawley 3, 10, 30, 91, 20 mg/mL
rat (male) 365,416,556 rat serum
days
Spartan rat 1 hour (observed 2 or
(male and for 14 days) 48.2 mg/L
female) air
Effects observed
Slightly enlarged thoracic lymph nodes in 3/10
rats on Days 10 and 30; scattered focal
aggregates of alveolar macrophages on Days
10 and 556
1 instance of marked to slight respiratory
difficulty, 1 instance of ocular porphyrin
discharge at 2 mg/L dose level before Day 13;
eye squint, changes in motor activity (first
decreased, then increased), respiratory
difficulty, ocular porphyrin discharge at 48.2
mg/L dose level before Day 13. All rats normal
on Days 13 and 14.
Citation
Dow Chemical Co.
(1990)
Great Lakes
Chemical Corporation
(1994) and IRDC
(1974)
F-13
-------
Table F-6. Select pulmonary exposure studies for MWCNTs.
Route of Diameter
exposure (nm)
Surface
Length area
(urn) (m2/gram) Purity Species
Exposure
duration Doses tested
Effects observed
Citation
Acute or subacute
Inhalation 1,900-
(nose only) 2,900
Intratracheal 50
instillation
Inhalation 50
Inhalation 10-20
(whole
body)
Inhalation 10-20
(whole
body)
~<1 253 98.6% Wistarrat
(bulk) and (male)
99.1%
(micronized)
10 280 >95% Kunming
mouse
(female)
10 280 >95% Kunming
mouse
(female)
5-15 100 0.5% Ni and C57BL/6
Fe mouse (male)
5-15 100 0.5% Ni and C57BL/6
Fe mouse (male)
Single exposure 11,241mg/m3
for 6 hours;
7-,28-, 90-day
observation
Single 1.7mg/kg
exposure; 8-,
16-, 24-day
observation
6 hours/day; 32.61 mg/m3
5, 10, 15 days
6 hours/day; 0.3, 1, 5.3
7, 14 days mg/m3
6 hours/day; 0.3, 1 mg/m3
14 days
Deregulation of genes
(inflammation, oxidative stress,
and fibroses) at 241 mg/m3; mild
reversible inflammation and no
fibroses at 1 1 mg/m3 (LOAEC)
Inflammation of lining of bronchi
at 24 days; severe destruction
of alveolar netted structure
around CNT clumps
Thickening of alveolar wall, but
alveolar structure remained
No local pulmonary effects; non-
monotonic systemic immune
suppression
Systemic immune suppression,
not due to systemic uptake of
MWCNTs, but release of
immune suppressing signals
from lung
Ellinger-
Ziegelbauer
and
Pauluhn
(2009)
(OECD 403
compliant)
Li et al.
(2007)
Li et al.
(2QQ7)
Mitchell et
al. (2007)
Mitchell et
al. (2009)
F-14
-------
Table F-6 (Continued): Select pulmonary exposure studies for MWCNTs.
Route of Diameter
exposure (nm)
Intratracheal NR
instillation
Intratracheal 11.3
instillation
Intratracheal NR
instillation)
Intranasal 15.04±
0.47
Intratracheal 11-170
instillation
Length Surface Purity Species Exposure
(urn) area duration
(m2/gram)
NR NR NR SDrat Single
(female) exposure; 3-,
15-,28-,60-day
observation
0.7 NR 98%; traces Wi star rat Single
of Co and (female) exposure; 3-day
Fe catalysts observation
NR NR NR Wistarrat Single
(female) exposure;
3- and 60-day
observation
0.5-200 139.7 >90% BALB/cAnNCrl 3 days
(reported carbon (as mouse
by reported by (female)
supplier) supplier)
5-9 12.83 >90% carbon ICR mouse Single
(male) exposure; 1-,
3-, 7-, 14-day
observation
Doses
tested
2.2, 8.9, 22.2
mg/kg
0.5, 2, 5 mg/rat
2 mg/rat
-33.25, -66.5,
-133 ug/mouse
per day (with
10 ug OVA per
injection and
1 Dug OVA
booster given at
21,22, and
23 days)
5, 20, 50 mg/kg
Effects
observed
Inflammation and fibrosis;
granulomas with ground
MWCNTs
Significant dose-dependent
increase in micronucleated
pneumocytes
Toxicity of CNT mediated by
defective sites in carbon
framework; significant differences
between ground MWCNTs not
heated, heated to 600°C, and to
2,400°C for both short and long-
term response
Increased OVA-specific serum
IgE and lgG1 at all doses;
increased OVA-specific serum
lgG2a at medium and high
doses; increased BALF
eosinophils, neutrophils,
macrophages, MCP-1 and TNFa
and increased mediastinal lymph
node cell number at the medium
dose; in a separate experiment,
at all doses, BALF neutrophil
number was increased 24 hours
after a single exposure to
MWCNTs in the absence of OVA
Increase in immune cells and
granulomas; increase in
inflammatory cytokines (IL-1,
TNF-a, IL-6, IL-4, IL-5, IL-10,
IL-12, IFN-v) and IgE; distribution
of B cells in spleen
Citation
Muller et
al. (2005)
Muller et
al. (2008a)
Muller et
al. (2008b)
Nygaard et
al. (2009)
Park et al.
(2009)
F-15
-------
Table F-6 (Continued): Select pulmonary exposure studies for MWCNTs.
Surface
Route of Diameter Length area
exposure (nm) (urn) (m2/gram) Purity
Inhalation 30-50 0.3-50 109.29 >94%
(nose-only)
Inhalation 10-15 -20 NR 95%
Exposure
Species duration
C57BL/6 Single
mouse (male) exposure; 6
hours; 1-, 14-
day observation
Sprague- 6 hours/day;
Dawley rat 5 days; 1 month
(male) observation
Doses tested
-10 mg/kg;
concentration in
air was
103.6±8.34
mg/m3 (both with
and without 20
ug OVA injection
given 14 and 7
days before
exposure)
0.1, 0.34, 0.94
mg/m3
Effects observed
OVA sensitized group: significant
airway fibrosis at 14 days,
Elevated PDGF-AA and TGF- (31
at day 1 but not day 14;
increased IL-5 mRNA levels
Not sensitized group: elevated
PDGF-AA, but not increased
levels of TGF- (3 land IL-13
Pulmonary DMA damage
initiated; a Comet assay
performed on lung cells showed
a significant increase in DMA
damage for high dose compared
to controls immediately and
1 month following the last
exposure
Citation
Ryman-
Rasmussen
etal.
(2009b)
Kim et al.
(2012a)
Subchronic
Inhalation
Inhalation
(head-nose)
50 10 280 >95%
5-15,500-0.1-1 250-300 90%
1,300,
1,300-
2,000/900-
1,500,
700-800
Kunming
mouse
(female)
Wistar rat
(male and
female)
6 hours/day; 32.61 mg/m3
30, 60 days;
6 hours/day; 0.1,0.4,2.5
5 days/week; mg/m3
13 weeks
No obvious toxicity at 30 days;
severe pulmonary toxicity at
60 days
Minimal granulomatous
inflammation in lung at
0.1 mg/m3 (LOEC); significant
granulomatous inflammation
<0.5 mg/m3; no systemic toxicity;
no pulmonary fibrosis
Li et al.
(2009)
Ma-Hock et
al. (2009)
(OECD413
compliant)
F-16
-------
Table F-6 (Continued): Select pulmonary exposure studies for MWCNTs.
Route of Diameter
exposure (nm)
Intratracheal 67
instillation
Inhalation 1,900-
(nose only) 2,900
Surface
Length area
(urn) (m2/gram) Purity
3-30 26 99.79%
~< 1 253 98.6%
(bulk)
and 99.1%
(micronized)
Species
ICR mouse
(male)
Wistar rat
(male and
female)
Exposure
duration Doses tested
6 weeks 25, 50 ug/week/
mouse (both
with and without
1 ug OVA/2 wk)
6 hours/day; 0.1, 0.4, 1.5,
5 days/week; 6 mg/m3
13 weeks
Effects observed
Increased total cells in BALF,
infiltration of inflammatory
leukocytes in airways, induction
of goblet cell hyperplasia in both
groups, enhanced response in
sensitized group
Sustained pulmonary
inflammation at > 1.5 mg/m3;
granulomas and alveolar
hyperplasia at > 6 mg/m3;
no systemic toxicity; 0.1 mg/m3
was NOAEC
Citation
Inoue et al.
(2009)
Pauluhn
(2010)
(OECD413
compliant)
Developmental Studies
Intratracheal Width distribution from 70-170 nm with the ICR mice Single dose on
spray greatest frequency occurring at 90-110 nm, (pregnant GD9; fetuses
length distribution between 1-19 urn with the dams) examined on
greatest frequency occurring between 1-5 urn GD18
[reported to be identical to those described by
Takagi et al. (2008) and Sakamoto et al. (2009)1
0, 3, 4, or Decreased final body weight of
5 mg/kg-bw) dams and decreased body
weight of live fetuses in 5mg/kg
group; external and skeletal
malformations (e.g., limb
deformities, fused ribs) observed
more frequently (ratio of litter with
malformed fetuses and percent
[%] incidence) in two highest
dose groups
Fujitani et
al. (2012)
Abbreviations: NR = Not reported
F-17
-------
Table F-7. Select oral and intragastric studies for decaBDE.
Route of Description of
exposure decaBDE
Species
Exposure
duration
Doses tested
Effects observed
Citation
Acute
Oral NR
(gavage)
Intragastric 77.4% decaBDE,
intubation 21.8% nonaBDE,
0.8% octaBDE
Rat
Sprague-
Dawley rat
(female)
Single dose
Single dose
(acute)
5,000 mg/kg
126, 252, 500,
1,000, 2,000 mg/kg
(10% corn oil
suspension)
No clinical signs, toxicity, or death
No indications of toxicity; no detectable
pathological changes
LSRI(1984); Great
Lakes (1984); IRDC
(1974) as cited in NRC
(2000)
Norrisetal. (1975) [also
in Norris et al. (1973)1
Subchronic
NR 97-99% pure
Intragastric >98% pure
Oral 94-98% pure
(diet)
Oral 94-98% pure
(diet)
Oral 97% pure
(gavage)
Rat (male and
female)
Wistar rat
(female)
F344/N rat
(male and
female)
B6C3F-I mouse
(male and
female)
Wistar rat
(male)
28 days
7-28 days
14 days
14 days
28 days
0, 7.4,
75 mg/kg-day
0, 10, 100, 1,000
mg/kg-day
0, 5,000, 10,000,
20,000, 50,000,
100,000 ppm
0, 50,00, 10,000,
20,000, 50,000,
100,000 ppm
0, 1.9, 3.8, 7.5, 15,
30, 60 mg/kg
No histology in liver or thyroid (NOAEL 74
mg/kg-day)
2-fold induction of CYP1A and CYP2B at 10-
1,000 mg/kg-day (not dose-dependent)
No treatment-related clinical signs or gross
pathologic effects
No treatment-related clinical signs or gross
pathologic effects
Increased weight of seminal vesicle/coagulation
gland (BMDL 0.2 mg/kg-day); increased
expression of hepatic CYP1 A and CYP2B
(BMDL 0.5-0.7 mg/kg-day)
Great Lakes (1976);
IRDC (1976) as cited in
NRC (2000)
Bruchajzer et al. (2010)
NTP (1986)
NTP (1986)
Van der Ven et al.
(2008)
F-18
-------
Table F-7 (Continued): Select oral and intragastric studies for decaBDE.
Route of
exposure
Oral
(gavage)
Oral
(diet)
Oral
(diet)
Oral
(diet)
Description of
decaBDE
97% pure
77.4% decaBDE,
21.8%nonaBDE,
0.8% octaBDE
94-98% pure
94-98% pure
Exposure
Species duration
Wistar rat 28 days
(female)
Sprague- 30 days
Dawley rat
(male)
F344/N rat 13 weeks
(male and
female)
B6C3F-I mouse 13 weeks
(male and
female)
Doses tested
0, 1.9, 3.8, 7.5, 15,
30, 60 mg/kg
0, 8, 80, 800 mg/kg-
day
0, 3,100; 6,200;
12,500; 25,000,
50,000 ppm
0, 3,100,6,200,
12,500, 25,000,
50,000 ppm
Effects observed
Decreased activity of P450c1 7 (BMDL0.18
mg/kg-day)
Decrease in packed cell volume and total red
blood cell count in highest dose group; enlarged
livers in mid- and high-dose groups; liver and
kidney lesions at 800 mg/kg-day; thyroid
hyperplasia at 80 mg/kg-day
No treatment-related clinical signs or gross or
microscopic pathologic effects
No treatment-related clinical signs or gross or
microscopic pathologic effects
Citation
Van der Ven et al.
(2008)
Norris et al.
NTP (1986)
NTP (1986)
(1975)
Chronic
Oral
(diet)
Oral
(diet)
77.4% decaBDE,
21.8% nonaBDE,
0.8% octaBDE
94-98% pure
Sprague- 2 years
Dawley rat
(male and
female)
F344/N rat 2 years
(male)
0, 0.01, 0.1, 1.0 mg
0, 2,500, 50,000
ppm
No differences observed in hematology or
urinalysis at 1 year; no other results reported
(report published before completion of study)
Increased incidence of neoplastic lesions
(nodules in the liver; acinar cell adenomas,
Norris et al.
NTP (1986)
(1975)
sarcoma of the spleen, hepatocellular
adenomas and carcinomas, thyroid gland
follicular cell adenomas or carcinomas) and
nonneoplastic lesions (thrombosis and
degeneration of the liver; fibrosis of the spleen,
lymphoid hyperplasia)
Oral 94-98% pure
(diet)
F344/N rat
(female)
2 years 0,2,500,50,000
ppm
Increased incidence of neoplastic lesions
(nodules in the liver of high-dose group) and
nonneoplastic lesions; degeneration of the eye
in low dose group
NTP (1986)
F-19
-------
Table F-7 (Continued): Select oral and intragastric studies for decaBDE.
Route of
exposure
Oral
(diet)
Oral
(diet)
Oral
(diet)
Description of
decaBDE
94-98% pure
94-98% pure
94-98% pure
Species
B6C3F-I mouse
(male)
B6C3F-I mouse
(female)
F344/N rat
(male)
Exposure
duration
2 years
2 years
2 years
Doses tested
0, 2,500, 50,000
ppm
0, 2,500, 50,000
ppm
1,120, 2,240 mg/kg
(adjusted)
Effects observed
Dose-dependent increase in thyroid follicular
cell hyperplasia; centrilobular hypertrophy
(indicated by enlarged hepatocytes with frothy
vacuolated cytoplasm)
Increased incidence of nonneoplastic lesions
Increased incidences of thrombosis and
degeneration of the liver in high-dose group;
Citation
NTP (1986)
NTP (1986)
NTP (1986)
enzyme induction; significant increases in
hepatic CYP1A mRNA, CYP2B mRNA,
CYP1A1 protein, and 7-pentoxyresorufin O-
dealkylase activity; increased 7-ethoxyresofurin
O-deethylase activity
Oral 94-98% pure
(diet)
Oral 77.4% pure
(diet)
Oral (metal >99%
gastric
tube)
F344/N rat
(female)
Sprague-
Dawley rat
(male and
female)
NMRI mice
(male)
2 years
2 years
single dose
given at
age 3, 10,
or 19 days
1,200, 2,550 mg/kg
(adjusted)
0, 0.01, 0.1, 1
mg/kg-day
2.22, 20.1
mg/kg-day for 3 and
19 day old mice; 0,
1.34, 13.4, or20.1
mg/kg-day for 10
day old mice
Increased 7-ethoxyresofurin O-deethylase
activity
No histology; NOAEL 1 mg/kg-day
Statistically significant changes in spontaneous
behavior variables (increased activity for
locomotion, rearing, total activity) at 2, 4, and 6
months at highest dose when exposed on
PND3 (developmental effects), but not PND10
orPND19
NTP (1986)
Norris (1973); Dow
(1994) and Kociba et al.
(1975) as cited in NRC
(2000)
Viberg et al. (2003)
F-20
-------
Table F-8. Select intubation and injection studies for MWCNTs.
Surface
area
Route of Diameter Length (m2/
exposure (nm) (Mm) gram)
Purity
Species
Exposure
duration
Doses tested
Effects observed
Citation
Acute/Subacute
intra- 15-30 15-20
peritoneal
injection
>95%; Swiss- daily, 5 days
functionalized Webster
(2-7% mice
COOH)
0.25, 0.5, 0.75 Dose-related increase in ROS
mg/kg-day level in liver homogenate at all
doses; increase in LHPs in liver
homogenate and ALT in serum at
medium and high dose; increase
in serum ALP at high dose; non-
statistically significant, dose-
dependent increase in AST/GOT
at all doses.
Patlolla et
al. (2011)
Developmental Studies
Oral 10-15 -20 NR -95%
(gavage)
Sprague- Daily dose on
Dawleyrat GD6-GD19
(pregnant
dams)
of 0, 40, 200, and Dose-dependent decrease in
1,000 mg/kg-day absolute and relative thymus
weight and increase in
malondialdehyde concentration
(maternal effects); no other
treatment-related maternal or fetal
(developmental) effects were
reported; 1,000 mg/kg-day was
the embryo-fetal NOAEL
Lim et al.
(2011)
F-21
-------
Table F-8 (Continued): Select intubation and injection studies for MWCNTs.
Route of
exposure
Diameter
(nm)
Length
(Mm)
Surface
area
(m2/
gram) Purity
Exposure
Species duration Doses tested Effects observed
Citation
Intraperitoneal Width distribution from 70-170 nm with the
greatest frequency occurring at 90-110 nm,
length distribution between 1 -19 urn with the
greatest frequency occurring between 1 - 5
urn
[reported to be identical to those described by
Takagi et al. (2008) and Sakamoto et al.
(2009)1
ICR mice Single dose on 0,2, 3, 4, or 5 Dose-dependent decreased
(pregnant GD9; fetuses mg/kg-bw maternal body weight; increased
dams) examined on number of resorptions, decreased
GD18 number of live fetuses per litter in
the two highest dose groups;
external and skeletal
malformations (e.g., cleft palate,
limb deformities, hypo/
hyperphalangia) observed more
frequently (ratio of litter with
malformed fetuses and percent
(%) incidence at all dose levels
Fujitani et
al. (2012)
F-22
-------
F.1.3. Genotoxicity, Mutagenicity, and Other In Vitro Studies
Table F-9. Select genotoxicity, mutagenicity, and other in vitro studies for MWCNTs.
Assay
Diameter Length
(nm) (Mm)
Surface
area
(m2/
gram) Purity
Species
Doses tested
Effects observed
Citation
Cytotoxicity
Trypan blue
test
110-170 5-9
22 >98%; <0.1%Fe;
~1.5%Ni; other
metal catalysts
Murine
macrophage cell
line RAW 264.7
0.01, 0.1, 1, 10,
100 ug/mL
Significant cytotoxic effect at 10 and
100 ug/mL
Migliore et al.
(2010)
Bacterial mutation
Ames assay
Ames assay
DNA damage
Chromosome
aberration
110-170 5-9
100 to 0.2-1
>150
(bimodal
distribution)
and unscheduled DNA
88 ±5 5 ±4.5
130 >90%; <0.1%Fe;
residual
amorphous
carbon; other
metal
contaminants
NR >95%; no free
amorphous
carbon
synthesis
NR NR
Salmonella
typhimurium
strains TA 98, TA
100; Escherichia
coll strain
WP2uvrA
Salmonella
strains TA1535,
TA100, TA1537,
TA98, TA102
Chinese hamster
lung cells
0.01, 0.05, 0.13,
0.23, 0.46, 1.26,
2.30, 4.60, 9.0
ug/plate, ±S9
50, 158, 500,
1,581, 5,000
ug/plate, ±S9
0.078, 0.31, 1.4,
5.0, 20, 80 ug/mL,
-S9
No significant increase in number of
revertant colony with or without
metabolic activation
Not mutagenic and bacteriotoxic up to
5,000 ug/plate with or without
metabolic activation
Formation of polyploidy with no
structural chromosome aberration at
> 5.0 ug/mL for 24-hour test and > 1 .3
ug/mL for 48-hour test
Di Sotto et al.
(2009)
Wirnitzer et al.
(2009)
(OECD471)
Asakura et al.
(2010)
F-23
-------
Table F-9 (Continued): Select genotoxicity, mutagenicity, and other in vitro studies for MWCNTs.
Assay
Comet assay
Comet assay
Comet assay
Sister
chromatid
exchange;
micronucleus
assay
Chromosome
aberration
Sister
chromatid
exchange
Comet assay
(in vivo intra-
tracheal
instillation; 3
hours)
Diameter
(nm)
110-170
20-40
15-30
10-30
100 to
>150
(bimodal
distribution)
>80% of
particles
70-110,
Gaussian
peak at 90
>80% of
particles
70-110,
Gaussian
peak at 90
Length
(Mm)
5-9
0.5-200
15-20
1-2
0.2-1
>70% of
particles
1-4,
peak: 2
>70% of
particles
1-4,
peak: 2
Surface
area
(m2/
gram) Purity
22 >98%; <0.1%Fe;
~1.5%Ni; other
metal catalysts
NR 93.37%
NR > 95%
NR 95-98%
NR >95%; no free
amorphous
carbon
NR 3,500, 470, and
20 ppm iron,
sulfur, and
chlorine
contaminants
NR 3,500, 470, and
20 ppm iron,
sulfur, and
chlorine
contaminants
Species
Murine
macrophage cell
line RAW 264.7
human lung
epithelial A549
cells
normal human
dermal fibroblast
cells (NHDF)
Human
lymphocyte cells
V79 cells
Chinese hamster
ovary AA8 cells
Male ICR mice (6
weeks old); lung
cells
Doses tested
0.01, 0.1, 1, 10,
100 ug/mL
5, 10,40, 100
ug/mL
40, 200, 400
ug/mL
1 mg/mL, -S9
2.5, 5, 10 ug/ml_,
±S9
0.1, 1.0,2.0
ug/mL
Single doses of
0.05 or 0.2
mg/animal
Effects observed
Significantly higher percent (%) DNA
in comet tails for doses > 1 ug/ml_;
significant dose-related effect overall
Significantly higher percent (%) DNA
in comet tails concentration-
dependent for 10, 40 ug/ml_ at 2 hours
post exposure and 5, 10, 100 ug/mL 4
hours post exposure; corresponds
with reduced cell viability
Statistically significant, dose-
dependent increase in percent (%)
DNA in comet tails
No significant cytotoxic effects
No cytotoxic or clastogenic effects
detected with or without metabolic
activation
Significant increase in sister chromatid
exchange frequency at all doses;
approximately 3-fold increase over
controls at 1.0 ug/ml_
Significant, dose-dependent increase
in DNA damage observed by
significant increases in DNA tail
moment and percentage of DNA in the
tail compared to controls.
Citation
Migliore et al.
(2010)
Cavallo et al.
(2012)
Patlolla et
al.(2010b;
2010a).
Szendi and
Varga (2008)
Wirnitzer et al.
(2009)
(OECD 473)
Kato et al.
(2013): CNT
characteristics
in Sakamoto
et al. (2009)
Kato et al.
(2013): CNT
characteristics
in Sakamoto
et al. (2009)
F-24
-------
Table F-9 (Continued): Select genotoxicity, mutagenicity, and other in vitro studies for MWCNTs.
Assay
DMA adduct
assay (in vivo
intra-tracheal
instillation; 3,
24, 72, 168
hours)
DMA damage
via Western
blot
Diameter
(nm)
>80% of
particles
70-110,
Gaussian
peak at 90
NR
Surface
area
Length (m2/
(urn) gram)
>70% of NR
particles
1-4,
peak: 2
NR NR
Purity
3,500, 470, and
20 ppm iron,
sulfur, and
chlorine
contaminants
NR
Species
Male ICR mice (6
weeks old); lung
DNA
Mouse
embryonic stem
cells
Doses tested
Single dose of 0.2
mg/animal
100 ug/mL
Effects observed
Three (of four analyzed) DNA adducts
related to oxidative stress and lipid
peroxidation significantly increased
(relative to controls) in a time
dependent manner up to 72 hours; a
significant smaller significant increase
relative to controls was observed at
168 hours.
Increased expression of two isoforms
of base excision repair protein 8-
oxoguanine-DNA glycosylase 1
(OGG1), double strand break repair
protein Rad 51; phosphorylation of
H2AX histone at serine 139; SUMO
modification ofXRCC4
Citation
Kato et al.
(2013); CNT
characteristics
in Sakamoto
et al. (2009)
Zhu et al.
(2007)
Mutation
Micronucleus
assay
Mutation at
hgprt locus
Micronucleus
assay
Micronucleus
assay
88 ±5
88 ±5
20-40
>80% of
particles
70-110,
Gaussian
peak: 90
5 ±4.5 NR
5 ±4.5 NR
1-5 NR
>70% of NR
particles
1-4,
peak: 2
NR
NR
>99% wt
3,500, 470, and
20 ppm iron,
sulfur, and
chlorine
contaminants
Chinese hamster
lung cells
Chinese hamster
lung cells
Human blood
cells
Human lung
carcinoma A549
calls
0.02, 0.078, 0.31,
1.3, 5.0 ug/ml_, -
S9
6.3, 12.5, 25, 50,
100 ug/ml_, -S9
0.25-150 ul_/5 ml_
total cell culture
volume, -S9
20, 100,200
ug/mL
Increased bi-and multi-nucleated cells
at > 0.31 ug/mL; no micronucleus
induction
Negative hgprt mutagenicity at all
doses
MWCNTs acted as clastogen and
aneugen agents simultaneously
6 hour treatment at 20 ug/mL inhibited
cell growth to around 70% of control
levels. Significant, dose-dependent
increase in frequency of
micronucleated cells at all doses (up
to 8.6% at 200 ug/ml_)
Asakura et al.
(2010)
Asakura et al.
(2010)
Cveticanin et
al. (2010)
Kato et al.
(2013): CNT
characteristics
in Sakamoto
et al. (2009)
F-25
-------
Table F-9 (Continued): Select genotoxicity, mutagenicity, and other in vitro studies for MWCNTs.
Assay
Mutation
assay (in vivo
intra-tracheal
instillation; 8 -
12 weeks)
Micronucleus
assay
Micronucleus
assay
Platelet
aggregation
(Aggro-Link
data reduction
system;
Chronolog)
Diameter
(nm)
>80% of
particles
70-110,
Gaussian
peak at 90
110-170
11.3
NR
Surface
area
Length (m2/
(urn) gram)
>70% of NR
particles
1-4,
peak: 2
5-9 22
0.7 NR
NR NR
Purity
3,500, 470, and
20 ppm iron,
sulfur, and
chlorine
contaminants
>98%; <0.1%Fe;
~1.5%Ni; other
metal catalysts
98%; traces of Co
and Fe
NR
Species
Male guanine
phosphoribosyl-
transferase (gpt)
mice (9 weeks
old)
Murine
macrophage cell
line RAW 264.7
Rat lung
epithelial cells
human platelet
cells
Doses tested
One, two, or four
single doses of
0.2 mg/animal
given once, two
weeks apart, and
every week,
respectively
0.01, 0.1, 1, 10,
100 ug/ml_, -S9
10,25, 50, 100,
150 ug/ml_, -S9
0.2-300 ug/ml_;
platelets isolated
and resuspended
in Tyrode's
solution;
aggregation was
studied for 8 min
post-addition of
MWCNTs
Effects observed
No increase in gpt mutant frequencies
following single or double dose, but
significant increase (approximately 2-
fold over control) after 4 doses.
No increases observed in SPi-
(sensitive to P2 interference) mutation
frequencies.
Significantly more micronucleated
cells for doses > 1 ug/mL
Significant increase in micronuclei
concentration-dependent increase in
platelet aggregation; activation of
GPIIb/llla
Citation
Kato et al.
(2013); CNT
characteristics
in Sakamoto
et al. (2009)
Migliore et al.
(2010)
Muller et al.
(2008a)
Radomski et
al. (2005)
F-26
-------
Table F-9 (Continued): Select genotoxicity, mutagenicity, and other in vitro studies for MWCNTs.
Surface
area
Diameter Length (m2/
Assay (nm) (urn) gram) Purity
T-cell viability; 20-40 1-5 NR 95%
Trypan Blue
exclusion
assay
Species
Jurkat T
lymphocyte
(leukemia) cells
from healthy
human blood
donors
Doses tested Effects observed Citation
1, 10 ng/cell (40, time-dependent decrease in the Bottini et al.
400 ug/ml_); Cell viability of Jurkat T leukemia cells; (2006)
aliquots collected increased number of cells staining
at 0, 24, 48, 72, with annexin V indicating increased
96, 120 hours apoptosis
post-exposure;
stained for 5 min
with Trypan Blue
to determine cell
proliferation and
percentage of
apoptotic Jurkat
or peripheral
blood
lymphocytes
(PEL) determined
using annexin V-
FITC
Abbreviations: NR = Not reported
F-27
-------
F.1.4. Carcinogenicity Studies
Table F-10.
Route of
exposure
Oral
Oral
Oral
Oral
Oral
Select carcinogenicity studies for decaBDE.
Description of
decaBDE
94-98% pure
94-98% pure
94-98% pure
94-98% pure
77.4%
decaBDE,
21.8%
nonaBDE,
0.8% octaBDE
Exposure
Species duration
F344/N rat 2 years
(male)
F344/N rat 2 years
(female)
B6C3F-: mouse 2 years
(male)
B6C3F-: mouse 2 years
(female)
Sprague-Dawley 2 years
rat (male and
female)
Doses tested
1,120,
2,240 mg/kg-day
(adjusted)
1,120,
2,550 mg/kg-day
(adjusted)
25,000,
50,000 ppm
25,000,
50,000 ppm
0, 0.01, 0.1,
1.0 mg/kg-day
Effects observed
Some evidence of carcinogenicity; increased
incidences of neoplastic nodules of the liver
(low dose 7/50, high dose 15/49, control 1/50)
Some evidence of carcinogenicity; increased
incidences of neoplastic nodules of the liver
(low dose 3/49, high dose 9/50, control 1/50)
Equivocal evidence of carcinogenicity;
increased incidences of hepatocellular
adenomas or carcinomas (combined) in both
dose groups
No evidence of carcinogenicity
No alterations in appearance, behavior,
bodyweight, feed consumption, hematologic
analyses, urinalysis, clinical chemistry, organ
weights, survival, or tumor incidence
Citation
NTP (1986)
NTP (1986)
NTP (1986)
NTP (1986)
Kocibaetal. (1975) as
reported in NTP (1986)
and NRC (2000)
F-28
-------
Table F-11. Select carcinogenicity studies for MWCNTs.
Surface
area
Test
substance
MWCNTs
with
structural
defects
MWCNTs
without
structural
defects
Short
MWCNTs
Short
MWCNTs
Long
MWCNTs
Diameter Length (m2/
(nm) (Mm) gram)
11.3± -0.7 299
3.9
11. 3± -0.7 190
3.9
14.84 ± 1-5 NR
0.50
10.40± 5-20 NR
0.32
165.02 ± Maximum NR
4.68 56
Purity
1.97%AI,
0.49% Fe,
0.48% Co
0.37%AI,
<0.01%Fe,
<0.01%Co
7.9% Fe,
5.1%Cu,
9.7% Ni,
5.5% Zn,
3.7% Co
(ug/gram)
13.4 Fe,
1 Cu, 5Ni,
7.5 Zn
(ug/gram)
37.3 Fe,
1.2 Cu,
6.2 Ni,
3.4 Co
Species
Wistar rat
(male)
Wistar rat
(male)
C57BL/6
mouse
(female)
C57BL/6
mouse
(female)
C57BL/6
mouse
(female)
Exposure
duration
Single
injection,
24-month
observation
Single
injection,
24— month
observation
Single
injection,
24-hour, 7-day
observations
Single
injection,
24-hour, 7-day
observations
Single
injection,
24-hour, 7-day
observations
Doses
tested
2,20
mg/rat
2,20
mg/rat
100
ug/mL
100
ug/mL
100
ug/mL
Effects observed
2 mesotheliomas at low dose (1 at 20 months and
1 at terminal sacrifice); 1 other peritoneal tumor at
low dose at 16.6 months; no mesotheliomas at high
dose; 3 other peritoneal tumors (1 lipoma at 13.8
months, 1 angiosarcoma and 1 liposarcoma at
terminal sacrifice) at high dose; no significant
difference in body weight or survival rates
No tumors observed at low dose; 3 mesotheliomas
(at 10.7, 18.9, and 19.8 months) and 3 lipomas
(at terminal sacrifice) at high dose; no significant
difference in body weight or survival rates
No significant effects
No significant effects
Inflammatory responses; formation of granulomas on
peritoneal surface of the diaphragm; foreign body
giant cells comparable to long-fiber amosite asbestos
Citation
Muller et
al. (2009)
Muller et
al. (2009)
Poland et
al. (2008)
Poland et
al. (2008)
Poland et
al. (2008)
F-29
-------
Table F-11 (Continued): Select carcinogenicity studies for MWCNTs.
Surface
Test
substance
MWCNTs
MWCNTs
MWCNTs
Diameter
(nm)
70-110,
peak at
90
10-30
>80% of
particles
70-110,
Gaussian
peak at
90
area
Length (m2/
(urn) gram)
72.5% NR
within 1-
4, peak at
2
1-2 NR
>70% of NR
particles
1-4,
peak: 2
Purity Species
(ppm) 3,500 p53 (+/-)
Fe, 470 S, mouse
20 Cl, <5 Br, (male)
<40F
95-98% F344 rat
3,500, 470, F344 rat
and 20 ppm (male)
iron, sulfur,
and chlorine
contaminants
Exposure
duration
Single
injection,
25-week
observation
Single
exposure; 12-
month
observation
single
intrascrotal
injection;
observed at
52 weeks
Doses
tested Effects observed
3 mg/mL 1 00% mortality by week 25; mesothelioma incidence
14/16 (87.5%, 11 as cause of death, 3 incidental);
moderate to severe fibrous peritoneal adhesion with
slight ascites; fibrous peritoneal thickening with
black-colored depositions; high incidence of
macroscopic peritoneal tumors; peritoneal
mesothelial lesions
1 0 Granulomatous reactions of foreign body type with
mg/rat multinucleated giant cells in liver; abdominal cavities
dispersed carbon on the adjacent organs (omentum,
peritoneum), resulting in partial expansion of the
gastric wall and residual carbon in peritoneal
envelope of liver; no signs of mesothelioma
1 mg/kg 85.7% mortality by 37-40 weeks; intraperitoneally
disseminated hypertrophic mesothelioma (cause of
death in 6/7 animals); nodular and papillary lesions
of mesothelioma; mesotheliomas invasive to
adjacent organs and metastasized into pleura
Citation
Takagi et
al. (2008)
Varga
and
Szendi
(2010)
Sakamoto
etal.
(2009)
Abbreviations: NR = Not reported
F-30
-------
F.2. Ecological Effects
F.2.1. Effects of DecaBDE and Other PBDEs on Aquatic Receptors
Table
Life
stage
F-12. Effects of exposure
N u mber / Test type /
group duration
to PBDEs in aquatic invertebrates.
Test
substance Medium Doses tested
Effect
Effect dose Notes
Algae3
NR
NR
Zebra
Post-
spawn
NR Semi-chronic/
96 hours
NR Semi-chronic/
96 hours
Commercial NR NR
decaBDE
Commercial NR NR
pentaBDE
EC50
NOEC
>1 mg/L Marine algae; review article, study-
specific details were not provided
Hardy (2002)
>water solubility Freshwater algae; review article,
study-specific details were not
provided
Hardy (2002)
mussels (Dreissena polymorphaf
150 Comet assay/
48, 96, 168
hours;
micronucleus
(MN) assay/
48, 96 hours
Technical Water 0.1, 2, 10 ug/L
grade
decaBDE, 98%
pure
Mortality,
hemocyte viability
DNA damage,
chromosomal
aberrations
NOE None
All tested doses Increasing trend overtime @0.1, 2
ug/L; damage increased until 96hr
then decreased at 168 hours @10
Riva et al. (2007).
F-31
-------
Table F-12 (Continued): Effects of exposure to PBDEs in aquatic invertebrates.
Life
stage
Number/ Test type/
group duration
Test
substance Medium Doses tested
Effect
Effect dose
Notes
Freshwater oligochaete (Lumbriculus variegatesf
NR
NR
Adult
Adult
NR Chronic/28
days
NR Chronic/28
days
80 Chronic/28
days
80 Chronic/28
days
Commercial Sediment NR
decaBDE
Commercial Sediment NR
pentaBDE
55%pentaBDE, Sediment 0,3.1,6.3,13,
36% tetraBDE, 25, 50 mg/kg
8.5% hexaBDE
Mixture, 97% Sediment 0,313,625,
decaBDE 1,250,2,500,
5,000 mg/kg
EC50
NOEL
EC50
EC5o, survival,
reproduction
LOEC, survival,
reproduction
NOEC, survival,
reproduction
>5,000 mg/kg
>5,000 mg/kg
>50 mg/kg
>50 mg/kg
6.3 mg/kg
>5,000 mg/kg
(mean
measured 3,841
Review article, endpoints not
specified (Row 1)c
Hardy (2002)
Review article, endpoints not
specified (Row 2)c
Hardy (2002)
Review article, endpoints not
specified (Row 3)c
Hardy (2002)
Review article, study-specific details
were not provided (Row 4)c
Environment Canada (2006).
Review article, study-specific details
were not provided (Row 5)c
Environment Canada (2006).
Review article, doses listed are
minimal measured concentration
(Row6)c
mg/kg) Environment Canada (2006).
F-32
-------
Table F-12 (Continued): Effects of exposure to PBDEs in aquatic invertebrates.
Life Number/ Test type/
stage group duration
Test
substance Medium
Doses tested Effect Effect dose
Notes
Water fleas (Daphnia magnaf
24 40 Acute/96 hours
hours
old at
test start
24 40 Chronic/21
hours days
old at
test start
24 20 Chronic/21
hours days
old at
test start
55%pentaBDE, NR
34% tetraBDE,
12%hexaBDE
55%pentaBDE, Water
34% tetraBDE,
12%hexaBDE
42% Water
heptaBDE,
36% octaBDE,
14% nonaBDE,
6% hexaBDE,
2% decaBDE
0,1.4,2.6,5.3, EC50 mortality, 17 ug/L
9.8, 20 ug/L immobility
0,1.4,2.6,5.3, EC50 mortality, 14 ug/L
9.8, 20 ug/L immobility
EC5o reproduction 14 ug/L
LOEC mortality, 20 ug/L
immobility
LOEC growth 9.8 ug/L
0, 0.13, 0.25, NOEC survival, >1.7 ug/L
0.54, 0.83, 1.7 reproduction,
ug/L growth
EC50, LOEC >2.0 ug/L
survival,
reproduction,
growth
Review article, study-specific details
were not provided (Row 1)d
Environment Canada (2006)
Review article, study-specific details
were not provided (Row 2)d
Environment Canada (2006)
Review article, study-specific details
were not provided (Row 3)d
Environment Canada (2006)
Review article, study-specific details
were not provided (Row 4)d
Environment Canada (2006)
Review article, study-specific details
were not provided (Row 5)d
Environment Canada (2006)
Review article, study-specific details
were not provided (Row 6)d
Environment Canada (2006)
Review article, study-specific details
were not provided (Row 7)d
Environment Canada (2006)
F-33
-------
Table F-12 (Continued): Effects of exposure to PBDEs in aquatic invertebrates.
Life
stage
NR
NR
NR
Number/
group
NR
NR
NR
Test type /
duration
Chronic/
21 days
Acute/48 hours
Chronic/21
days
Test
substance Medium Doses tested Effect
Commercial NR NR EC5o, NOEC
octaBDE
Commercial NR NR EC50
pentaBDE
Commercial NR NR LOEC
pentaBDE
Effect dose Notes
>water solubility Review article, study-specific details
were not provided (Row 8)d
Hardy (2002)
14 ug/L Review article, study-specific details
were not provided (Row 9)d
Hardy (2002)
9.8 ug/L Review article, endpoints not
specified (Row 10)
Hardy (2002)
Abbreviations: EC50 = Median effective dose; LOEC = Lowest observed effect concentration; NR: Not reported; NOE: No observed effects at tested doses;
NOEC = Maximum no observed effect concentration.
"Source: Hardy (2002).
bSource: Riva et al. (2007).
°Source: Rows 1-3: Hardy (2002): rows 4-6: Environment Canada (2006).
dSource: Rows 1-7: Environment Canada (2006): rows 8-10: Hardy (2002).
F-34
-------
Table F-13. Effects of exposure to PBDEs in fish and frogs.
Doses
Duration Test substance Medium tested Effect
Effect dose
Notes
Juvenile rainbow trout (Oncorhynchus mykiss)3
16 days Dow FR-300-BAb Diet 7.5-10 Increased liver weight, increased
mg/kg-dayc blood lactate concentrations,
49 days
120 days
96 hours Commercial NR NR LC50
pentaBDE
NR PentaBDE, Injection NR Egg mortality
tetraBDE mix
NOE
NOE
7.5
mg/kg-day
> water
solubility
> 12 ug/egg
None (Row 1)a
Kierkegaard et al. (1999)
Saw effects in group after 71 days of depuration
(non-exposure), indicating potential delayed
chronic effects (Row 2)a
Kierkegaard et al. (1999)
None (Row3)a
Kierkegaard et al. (1999)
Review article, study-specific
provided (Row4)a
Hardy (2002)
Review article, study-specific
provided (Row 5)a
Hardy (2002)
details were not
details were not
72 hours Tetra-, penta-, and Cell 0-264 ug/L Vitellogenin production
octaBDE culture
10-40 ug/L In vitro hepatocyte assay; intensity of response
increased with increasing number of bromine
atoms (Row 6)a
Nakari and Pessala (2005)
Juvenile lake whitefish (Coregonus clupeaformis)
30 days DecaBDE; Diet
97.5-99.25% pure
0,0.1,1,2 Negative growth effects
M9/gram
2 ug/gram Otolith increment widths narrowed starting
day 1
Kuo et al. (2010)
F-35
-------
Table F-13 (Continued): Effects of exposure to PBDEs in fish and frogs.
Duration Test substance
Doses
Medium tested
Effect
Effect dose Notes
Juvenile lake trout (Salvelinus namaycushf
56 days DecaBDE,
>96% pure
Diet
0, 2.5, 25
ng/gram
Mortality, whole body growth rate
NOE No effects observed during exposure period
and up to 112 days depuration
Tomy et al. (2004)
Phase I EROD activity in liver
microsomes
NOE Measured on days 14, 56 of uptake and 14, 56
of depuration
Tomy et al. (2004)
Liver somatic index changes (liver
weight •*• whole fish weight x 100)
NOE No effects observed during exposure period
and up to 112 days depuration
Tomy et al. (2004)
Decreased free thyroxine (T4) 2.5, 25
concentrations ng/gram
Level significantly decreased at both doses
(greater decrease at 25 ng/gram) at 56 days, no
effects at earlier time points; levels remain
decreased in high dose but not low dose after
112 days depuration
Tomy et al. (2004)
Decreased free tri-iodothyronine (T3) 2.5 ng/gram
concentrations
Level significantly decreased at 2.5 ng/gram at
56 days, no effects at earlier time points; no
effects at 25 ng/gram at any time; author states
"no consistent differences related to PBDE
exposure level were evident"
Tomy et al. (2004)
F-36
-------
Table F-13 (Continued): Effects of exposure to PBDEs in fish and frogs.
Duration Test substance
Adult Chinese rare minnow
21 days DecaBDE,
> 99% pure
Adult Chinese rare minnow
21 days DecaBDE, > 99%
pure
Doses
Medium tested
(Gobiocypris rarusf
Water 0.01,0.1,1,
10 ug/L
(Gobiocypris rarusf
Water 0.01, 0.1, 1,
10 ug/L
Effect
Mortality, malformations
Decreased body length,
gonadosomatic index (GSI) changes
Inhibition of spermatogenesis,
reduction of spermatocytes
mRNA levels of thyroid hormone
related genes
Liver degeneration, hepatocyte
swelling
Effect dose Notes
NOE None
Lietal.(2011)
1 0 ug/L GSI = Gonad weight •*• whole fish
Lietal.(2011)
10 ug/L Males only
Lietal.(2011)
weight x 100
Variable at all Upregulation of nis at all doses; tr-a in all male
doses groups but not female; ttr in all female groups
but not male; dio2 in females at 0.01 ug/L,
females + males at 0.1 ug/L.
Lietal.(2011)
10 ug/L Females only
Lietal.(2011)
Unspecified/general fish9
48 hours Commercial
decaBDE
NR
NR
LC5
>500 mg/L Review article, study-specific details were not
provided
Hardy (2002)
48 hours Commercial
octaBDE
NR
NR
LC
50
>500 mg/L Review article, study-specific details were not
provided
Hardy (2002)
48 hours Commercial
pentaBDE
NR
NR
LC50
> 500 mg/L Review article based value on data for Oryzias
latipes, details were not provided
Hardy (2002)
F-37
-------
Table F-13 (Continued): Effects of exposure to PBDEs in fish and frogs.
Duration
NR
Tadpoles
51 days
Test substance Medium
Commercial Diet
pentaBDE
(Xenopus laevisf
DE-83R Water
Doses
tested
NR
1, 10, 100,
1,OOOng/L
Effect
Reproduction (spawning success)
Physical malformation, abnormal
behavior, increased mortality
Effect dose Notes
NOE Review article based value on data for
Gasterpsteis aculeatus, details were not provided
Hardy (2002)
NOE None (Row 1)h
Qinetal. (2010)
FLE, stage 57/58); n = 70
Delayed time to metamorphosis
1,000 ng/L Statistically significant at 1,000 ng/L;
concentration-dependent trend of metamorphic
delay in all groups (Row 2)h
Qin et al. (2010)
FLE, stage 57/58); n = 70
Histological alterations in thyroid
glands (epithelial cell height, follicle
size, colloid depletion, colloid
vacuolation)
All doses Statistically significant increase in mean epithelial
cell height at 100 and 1,000 ng/L; multilayer
follicular epithelial cells at all doses (Row 3)h
Qin et al. (2010)
FLE, stage 57/58); n = 10-20
Decrease in TR-p-A mRNA expression All doses
(thyroid hormone) in tail tissue
None (Row 4)h
Qin et al. (2010)
FLE, stage 57/58); n = 8
Abbreviations: EROD = Ethoxyresorunfin-O-deethylase enzyme; LC50 = Median lethal concentration; LOEC = Lowest observed effect concentration; NOE: No observed effects at
tested doses; NOEC = Maximum no observed effect concentration tested; NR: Not reported.
"Source: Rows 1-3: Kierkegaard et al. (1999): rows 4-5: Hardy (2002): row 6, Nakari and Pessala (2005): individuals per group not reported.
""Commercial mixture contains 77.4% decaBDE, 21.8% nonaBDE, 0.8% octaBDE as reported in Hardy (2002).
°One dose administered, dose was minimum 7.5 mg/kg-day, maximum 10 mg/kg-day.
dSource: Kuo et al. (2010): 75 individuals per test group.
eSource: Tomy et al. (2004): 70 individuals per test group.
'Source: Li et al.(2011): individuals per group not reported.
9Source: Hardy (2002) (Review article).
hSource: Qin et al. (2010): free swimming larvae 5 days post-fertilization (stage 46/47) through forelimbs emergence (FLE, stage 57/58); n = 70 in rows 1-2; n = 10-12 in row 3,
n = 8 in row 4.
F-38
-------
F.2.2. Effects of MWCNTs on Aquatic Receptors
Table F-14. Effects of exposure to MWCNTs in algae, macrophytes, and aquatic macro!nvertebrates.
Organism
Unicellular green algae
(Dunaliella tertiolectaf
Aquatic macrophytes
(Chara sp., Elodea nuttallii,
Potamogeton obtusifulius,
Glycera sp., Alisma
Test
substance
Carboxylated
MWCNTs
Pure
MWCNTs
Dimensions
OD:
20-30 nm
L: 50 urn
OD:
20-30 nm
ID: 5-10 nm
L: 10-30 urn
Test
duration
96 hours
3 months
Medium
Water
Stream bed
sediment
Doses
tested
0.1, 0.5,
1,2.5, 5,
1 0 mg/L
0.002,
0.2,2
grams/kg
Effect
Inhibited
growth
Changes in
macrophyte
density
Effect
dose
LOEL 1 mg/L,
NOEL
0.5 mg/L
LOEL 0.002
grams/kg
Notes
Growth lagged up to 23 days
at 10 mg/L, 36% reduction in
exponential growth rate
Density increased compared
to control at all levels;
significant at 0.002 and 2
grams/kg
Macroinvertebrates
(Gastropoda, Crustacea,
Oligochaeta, Hirundinea,
Bivalvia, Arachnida,
Diptera)b
Pure
MWCNTs
OD:
20-30 nm
ID: 5-10 nm
L: 10-30 urn
3 months Stream bed 0.002, Recolonization LOEL 0.002 Dose-dependent increase in
sediment 0.2,2 rates grams/kg recolonization (#of taxa and
grams/kg individuals) compared to
control
Biodiversity
NOEL
>2 grams/kg
No observed effect
F-39
-------
Table F-14 (Continued): Effects of exposure to MWCNTs in algae, macrophytes, and aquatic macroinvertebrates.
Organism
Amphipods
(Leptocheirus plumulosus)0
Amphipods
(Hyalella azteca)0
Test
substance
Pure
MWCNTs
Pure
MWCNTs
Dimensions
OD:
10-30 nm
L: 10-30 urn
OD:
10-30 nm
L: 10-30 urn
Test
duration Medium
10 days Sediment
10 days Sediment
Doses
tested Effect
4, 10, 33, Survival
99, 300
grams/kg
3, 9, 29, Survival
87, 264
grams/kg
Effect
dose
LOEL 99
grams/kg,
LC50 68
grams/kg
LOEL
264 grams/kg
Notes
Survival 30 ± 10% at
99 grams/kg,
30 ± 0% at 300 grams/kg
Survival 53 ± 25% at
264 grams/kg
Abbreviations: L = Length; LCso = Median lethal concentration; ID = Inner diameter; LOEL = Lowest observed effect level; NOEL = Maximum no observed effect
level; NR: Not reported; OD= Outer diameter.
aSource: Wei et al. (2010); microwave assisted acid oxidation was used to carboxylate pristine nonfunctionalized MWCNTs resulting in functionalized MWCNTs
with 7.61% carboxylation, 1% residual cobalt by weight, elemental composition 948:51:1 carbon:oxygen:cobalt.
bSource: Velzeboer et al. (2011): MWCNTs nonfunctionalized, purity 95% wt.
cSource: Kennedy et al. (2008): authors note that mortality increased as particle size decreased.
F-40
-------
Table F-15. Effects of exposure to MWCNTs via water on Ceriodaphnia dubia.
Individuals
per test
group
Outer
diameter
(nm)c
Functionalization
Concentration
s tested
(mg/L)
Effect
Effect dose
Acute (24 hours)3
20
20
20
20
20
20
20
20
20
10-20
30-40
50-70
10-20
30-40
50-70
10-20
30-40
50-70
NF
NF
NF
Ozone-treated
Ozone-treated
Ozone-treated
Ultrasound-probe
treated
Ultrasound-probe
treated
Ultrasound-probe
treated
1-200
1-200
1-200
1-200
1-200
1-200
1-200
1-200
1-200
LC50
LC50
LC50
LC50
LC50
LC50
LC50
LC50
LC50
17 mg/L; Li and Huang (2011).
8 mg/L; Li and Huang (2011).
20 mg/L; Li and Huang (2011).
100 mg/L; Li and Huang (2011).
100 mg/L; Li and Huang (2011).
100 mg/L; Li and Huang (2011).
8 mg/L; Li and Huang (2011).
7 mg/L; Li and Huang (2011).
2 mg/L; Li and Huang (2011).
Subchronic (48 hours)0
NR
NR
NR
>12
>12
Chronic (8
10-30
20-30
20-30
50-70
50-70
days)3
NF
MWCNT-OH
MWCNT-COOH
Ozone-treated
Ultrasound-probe
treated
25.1, 39.5, 59.6
120.2
88.9
5, 10, 20, 50,
100
1,2, 3, 5, 10
EC50
Mortality increase
Mortality increase
Body length
decreased
Body length
decreased
50.9 mg/L (Row 1)b;
Kennedy et al. (2008)
No effects (Row 2)b;
Kennedy et al. (2008)
No effects (Row 3)b;
Kennedy et al. (2008)
LOEC 5 mg/L; 73% of control at
100 mg/L (Row 4)b;
Li and Huang (2011).
NOEC 1 mg/L;
LOEL 2 mg/L; 75%. of control at
10 mg/L (Row 5)b;
Li and Huang (2011)
>8 50-70 Ozone-treated 5 concentrations, ECso decreased
0.5to>30 reproduction8
17 mg/L;
Li and Huang (2011)
>8 50-70 Ultrasound-probe 5 concentrations, ECso decreased 4 mg/L;
treated 0.5to>30 reproduction8 Li and Huang (2011)
Abbreviations: EC50 = Median effective concentration; LC50 = Median lethal concentration; LOEC = Lowest observed effect
concentration; NF= Not Functionalized; NOE: No observed effects at tested doses; NOEC = Maximum no observed effect
concentration tested; NR = Not reported.
"Source: Li and Huang (2011).
bSource: Rows 1-3: Kennedy et al. (2008): rows 4-5: Li and Huang (2011).
°Other properties not reported in Li and Huang (2011). Kennedy et al. (2008) reported length 10-30 urn and purity 95%.
dDose-response growth assay.
eThree generation reproductive test.
F-41
-------
Table F-16. Effects of exposure to MWCNTs on zebrafish and medaka embryos.
Number Outer
per diameter Test Exposure Doses
group (nm) substance medium tested Effect
Effect dose
Notes
Zebrafish, 72 hours post fertilization3
75 30-40 MWCNTs Microinjection 2.5, 5,
at 8-cell
stage
Heart rate
10, 20, decrease
30, 40,
50, 60,
70,
100,
200,
300
|jg/mL
NOEC 10|jg/mL,
LOEC 20 |jg/mL
Measured at 24, 48,
72 hpf; dose-
dependent drop at
48 and 72 hpf
Reduced blood
circulation
NOEC 60 ug/mL,
LOEC 70 ug/ml_
Observed at 24, 48,
and 72 hpf
60 30-40 MWCNTs Water 2.5, 5,
10,20,
30, 40,
50, 60,
70,
100,
200,
300
ug/ml_
Delayed
hatching
Increased
mortality
NOEC
40 ug/ml_3,
LOEC 60 ug/mL
NOEC 40 ug/mL,
LOEC 60 ug/mL
10%at60ug/mL,
80% at 100ug/mL
8% at 60 ug/mL,
60% at 1 00 mg/mL
97% at 200 ug/mL
Zebrafish, 96 hours post injection
NR
19.9
BSA- Microinjection 2 ng/ Developmental
MWCNTs at 1-cell stage embryo defects
NOE
Immune
response
2 ng
Through adult stage
Observed at 24 hpf
Medaka, 96 hours continual exposure0
12 NR oxidized Water
MWCNTs
500, Increased
1,000, mortality
1,500,
2,000
ug/mL
NOEC 1,000
ug/mL,
LOEC 1,500
ug/mL
MWCNTs were
functionalized by
acid treatment
Medaka, 10 days following 4 days of continual exposure
12
NR
oxidized
MWCNTs
Water
500,
1,000,
1,500,
2,000
ug/mL
Increased
malformations
NOEC
1,000 ug/mL,
LOEC
1,500 ug/mL
MWCNTs were
functionalized by
acid treatment
F-42
-------
Table F-16 (Continued): Effects of exposure to MWCNTs on zebrafish and medaka embryos.
Number
per
group
12
Zebrafish
Outer
diameter Test Exposure
(nm) substance medium
NR oxidized Water
MWCNTs
, 56 days post injection13
Doses
tested Effect
500, Hatching delay
1,000,
1,500,
2,000
ug/mL
Effect dose
NOEC
1,000 |jg/mL,
LOEC
1,500 |jg/ml_
Notes
MWCNTs were
functionalized by
acid treatment
NR 19.9 BSA- Microinjection 2ng/ Reduced 2 ng
MWCNTs at 1-cell stage embryo survival of 2nd
generation
Measured at day 14
of 2nd generation
lifecycle
Abbreviations: BSA = bovine serum albumin; hpf = hours post fertilization; LOEC = Lowest observed effect concentration; NOE: No
observed effects at tested doses; NOEC = Maximum no observed effect concentration tested; NR: Not reported.
"Source: Asharani et al. (2008).
"Source: Cheng et al. (2009).
°Source: Kim et al. (2012b).
dAuthors state NOEC of 40 ug/mL and LOEC of 60 ug/mL but do not provide the effect level at 50 ug/mL.
Table F-17. Immune responses in rainbow trout (Oncorhynchus mykiss) head kidney cells following
MWCNT exposure.
Endpoint
Antiviral response
Macrophage
stimulation
Cytotoxicity
Measured as:
IFNa expression observed after 6 hours
incubation
IL-1b expression observed after 24 hours
incubation
Measured after 24 hours incubation
Pure
MWCNTs3
NOE
LOEL
5 ug/mL
NOE
Anionic MWCNTs3
NOE
LOEL 0.1 ug/mL
(dose-dependent)
NOE
Note: Testing protocol: head kidney cells of adult fish were collect and incubated 5 days prior to experiment; doses tested: 0.1, 0.5,
1, 5, and 10 ug/mL MWCNTs diluted in water solution; 8 replicates for antiviral response/ macrophage stimulation, duplicate for
cytotoxicity.
Abbreviations: LOEL = Lowest observed effect level; NOE: No observed effects at tested.
aPure MWCNTs: >95% pure; Anionic MWCNTs: functionalized with sulfonate groups, centrifuged and ultrafiltered to purify. Both had
diameter 10-20nm, length 1-2 urn.
Source: Klaper et al. (2010)
F-43
-------
F.2.3. Effects of DecaBDE and Other PBDEs on Terrestrial Receptors
Table F-18. Effects of exposure to decaBDE in soil microbes, terrestrial invertebrates, and plants.
Doses
Test Test tested (mg/kg
duration substance dry weight soil) Effect
Effect
dose (mg/kg) Notes
Soil microbes
180 days DecaBDE 1,10,100 Altered community
structure
Cytotoxicity
All doses Control shows increase in Shannon-Weaver index at each checkpoint
from 15 days to 180 days; index number decreased compared to control
(continues to increase over time) at all doses beginning at -90 days,
indicating less diversity overtime
Liu et al. (2011):
100
Total bacteria count was -50%; dose-dependent decrease observed at
all doses; alpha, beta, gamma-proteobacteria groups were decreased to
74.7-84.7% at 100 mg/kg; decreases not observed at lower doses
Liu et al. (2011);
180 days DecaBDE 1,10,100
Changes in alkaline
phosphatase (APA)
enzyme activity
All doses Increased activity from 60-120 days at 10, 100 mg/kg then decreased
activity (inhibition) at 180 days; increased activity through full study at 1
mg/kg
Liu et al. (2011):
Changes in urease enzyme All doses
activity
Increased activity through 150 days at 1 mg/kg; activity increased at 10
mg/kg from 15 days to 120 days, and at 100 mg/kg at every checkpoint
except 150 days
Liu et al. (2011):
Nitrifying bacteria"
4 weeks DecaBDE 15.1,230,2,274 Change in behavior
NOE
Measured by change in nitrate/nitrite content
Sverdrup et al. (2006):
F-44
-------
Table F-18 (Continued): Effects of exposure to decaBDE in soil microbes, terrestrial invertebrates, and plants.
Doses
Test Test tested (mg/kg
duration substance dry weight soil) Effect
Effect
dose (mg/kg) Notes
Red clover (Trifolium pratensef
21 days DecaBDE 15.1,230,2,274 Early life growth
NOE
Measured by mean fresh weight of seedlings per soil plot for 15-17 days
post seedling emergence
Sverdrup et al. (2006)
Corn (Zea mays)
21 days 55% 62.5, 125,250, Germination NOE
pentaBDE, 500, 1,000
36%
tetraBDE, |_OEL reduced shoot height 250
8.6%
hexaBDE
None
Environment Canada (2006):
None
Environment Canada (2006):
LOEL reduced shoot height 62.5
None
Environment Canada (2006):
Soil invertebrate (Enchytraeus crypticusf
21 days DecaBDE 15.1,230,2,274 Reproductive effects
NOE
Measured as number of juveniles per soil plot
Sverdrup et al. (2006):
F-45
-------
Table F-18 (Continued): Effects of exposure to decaBDE in soil microbes, terrestrial invertebrates, and plants.
Doses
Test Test tested (mg/kg
duration substance dry weight soil) Effect
Effect
dose (mg/kg) Notes
Earthworms (Eisenia fetida)f
7 days DecaBDE 0.01,0.1,1,5, Hydroxyl free radical
10, 50, 100 generation
All doses Dose-dependent increase starting at lowest dose, reached level 2x
control at highest dose (Row 1)f
Xie et al. (2011)
56 days DecaBDE, 320,668,1,240, Reproductive effects
98% pure 2,480,4,910
NOE
Also observed at 28 days (no effects seen) (Row 2)f
Environment Canada (2006)
56 days Commercial 84.9,166,361, Mortality
octaBDE 698,1,470
NOE
None (Row3)f
Environment Canada (2006)
Reproductive effects
NOE
None (Row4)f
Environment Canada (2006)
Abbreviations: NOE = No observed effects.
"Source: Liu et al. (2011): 3 replicates per group
bSource: Sverdrup et al. (2006): number of replicates per group not reported
°Source: Sverdrup et al. (2006): 20 seeds per test group
dSource: Environment Canada (2006): 40 seeds per test group
eSource: Sverdrup et al. (2006): 40 adult worms per test group
'Source: Row 1: Xie et al. (2011). 60 worms per test group; row 2-4: Environment Canada (2006). 80 worms per test group
F-46
-------
F.2.4. Effects of MWCNTs on Terrestrial Receptors
Table F-19.
Test
substance
General soil
MWCNTs
Effects of exposure to MWCNTs in bacteria.
Test
Properties duration
microorganisms3
OD: 15.1 nm, 11 days
L: 10-20|jm,
SA:
237.1 m2/gram,
V: 0.86 cm3/gram,
P: 96%
20 days
Exposure protocol
MWCNTs suspended
in water, suspension
applied to soil
MWCNTs suspended
in water, suspension
applied to soil
Doses
Medium tested
Soil 50, 500,
5,000
ug/gram
Soil 50, 500,
5,000
ug/gram
Effect Effect dose Notes
LOEL decreased 500 ug/gram Indicated by enzyme
microbial activity presence; nonsignificant
tendency to be repressed at
500 ug/gram; activity
decreased by 34.2-60.5% at
5,000 ug/gram.
Chung et al. (2011).
LOEL decreased 5,000 C decreased by 36.9-43.4%,
biomass ug/gram N decreased n by 27.8-30.4%.
Chung et al. (2011).
Gram negative Escherichia coli°
MWCNTs
MWCNT-Fe
OD: 44.0 nm, 24 hours
L: 1.5 pm,
SA: 42 m2/gram,
0.08%wt Fe
OD: 44.0 nm, 24 hours
L: 1.5 urn,
SA: 42 m2/gram,
4.24%wt Fe
Exposure to aqueous
suspensions of
MWCNTs at room
temp, gentle stirring;
strain: MG1655
Exposure to aqueous
suspensions of
MWCNTs at room
temp, gentle stirring;
strain: MG1655
Water 10, 100
mg/mL
Water 10, 100
mg/mL
LOEL cytotoxicity 1 00 mg/mL ~ 50% loss in viability (Row 1 )b
Simon-Deckers et al. (2009)
LOEL cytotoxicity 1 00 mg/mL ~ 60% loss in viability (Row 2)b
Simon-Deckers et al. (2009)
F-47
-------
Table F-19 (Continued): Effects of exposure to MWCNTs in bacteria.
Test
substance
MWCNTs
Test
Properties duration
OD: 1 hour
17.4 ± 6.1 nm,
L: 77 ±31 |jm
Doses
Exposure protocol Medium tested
Incubation exposure to Filter in CNQ
MWCNT-coated filter culture
in 0.9% NaCI solution
Effect Effect dose
cytotoxicity Effect seen
Notes
-32% inactivated cells
(Row 3)b
Kang et al. (2009),
Gram negative Escherichia coli°
MWCNTs
OD: 30 nm, 1 hour
L: 70 urn,
0.62%wt
metal catalysts
5x107cells/mL Cell 5 ug/mL
incubated with culture
MWCNTs in saline for
1 hour at 37 °C; strain:
K12
Incubation exposure to Filter in CNQ
MWCNT-coated filter culture
in 0.9% NaCI solution;
strain: K12
Cytotoxicity 5 ug/mL
Cytotoxicity Effect seen
Reduced Effect seen
metabolic activity
~3x reduction in viability
compared to controls(Row 4)b
Kang et al. (2008)
~3.8x reduction in viability
compared to controls (Row 5)b
Kang et al. (2008)
30% metabolic activity
compared to 74% in control
(Row6)b
Kang et al. (2008)
Metallic-pollutant resistant Cupriavididus metallidurans CH34C
MWCNTs
MWCNT-Fe
OD: 44.0 nm, 24 hours
L: 1.5 urn,
SA: 42 m2/gram,
0.08%wt Fe
OD: 44.0 nm, 24 hours
L: 1.5 urn,
SA: 42 m2/gram,
4.24%wt Fe
Exposure to aqueous Water 10,100
suspensions of mg/mL
MWCNTs at room
temp, gentle stirring
Exposure to aqueous Water 10,100
suspensions of mg/mL
MWCNTs at room
temp, gentle stirring
NOEL NOE
cytotoxicity
NOEL NOE
cytotoxicity
No loss in viability
Simon-Deckers et al. (2009)
No loss in viability
Simon-Deckers et al. (2009)
F-48
-------
Table F-19 (Continued): Effects of exposure to MWCNTs in bacteria.
Test Test
substance Properties duration
Exposure protocol
Doses
Medium tested Effect
Effect dose Notes
Gram negative Pseudomonas aeruginosa"
MWCNTs OD: 1 hour
17.4 ± 6.1 nm,
L: 77 ±31 |jm
Incubation exposure to
MWCNT-coated filter
in 0.9% NaCI solution
Filter in CNQ Cytotoxicity
culture
Effect seen -25% inactivated cells
Kang et al. (2009)
Gram positive Staphylococcus epidermidisd
MWCNTs OD: 1 hour
17.4 ± 6.1 nm,
L: 77 ±31 urn
Incubation exposure to
MWCNT-coated filter
in 0.9% NaCI solution
Filter in CNQ Cytotoxicity
culture
Effect seen -50% inactivated cells
Kang et al. (2009)
Gram positive Bacillus subtilise
MWCNTs OD: 1 hour
17.4 ± 6.1 nm,
L: 77 ±31 urn
OH-MWCNT OD: 15-30 nm, 1 hour
L: 1-5 urn
COOH- OD: 15-30 nm, 1 hour
MWCNT L1_5um
NH2- OD:15-30nm, 1 hour
MWCNT L ^5 um
Incubation in 0.9%
NaCI solution
Cells suspended in
1 ml_ of solution
Cells suspended in
1 ml_ of solution
Cells suspended in
1 ml_ of solution
Filter in CNQ Cytotoxicity
culture
Water, 100ug/mL Delayed growth
culture
Water, 100ug/mL Delayed growth
culture
Water, 100ug/mL Delayed growth
culture
NOE None (Row 1)e
Kang et al. (2009)
CNQ NOE in solutions of Dl water,
PBS, BHI, or 0.9% NaCI
(Row2)e
Arias and Yang (2009).
NOE NOE in solutions of Dl water,
PBS, BHI, or 0.9% NaCI
(Row 3)e
Arias and Yang (2009),
NOE NOE in solutions of Dl water,
PBS, BHI, or 0.9% NaCI
(Row4)e
Arias and Yang (2009),
F-49
-------
Table F-19 (Continued): Effects of exposure to MWCNTs in bacteria.
Test
substance Properties
Test
duration Exposure protocol
Doses
Medium tested
Effect
Effect dose Notes
Gram negative Salmonella typhimuriurrf
OH-MWCNT
COOH-
MWCNT
NH2-MWCNT
OD: 15-30 nm 1 hour
L: 1-5 pm
OD: 15-30 nm 1 hour
L: 1-5 urn
OD: 15-30 nm 1 hour
L: 1-5 urn
Cells suspended in
1 ml_ of water
Cells suspended in
1 ml_ of solution
Cells suspended in
1 ml_ of solution
Cells suspended in
1 ml_ of solution
Water
Culture
Water,
culture
water,
culture
100,170, Delayed growth NOE
290, 375,
500 ug/mL
100,170, Delayed growth NOE
290, 375,
500 ug/mL
100,170, Delayed growth NOE
290, 375,
500 ug/mL
100,170, Delayed growth NOE
290, 375,
500 ug/mL
Nonsignificant reductions in
viability at all doses
Arias and Yang (2009)
NOE up to 500 ug/L in BHI
broth, PBS, or 0.9% NaCI
Arias and Yang (2009)
NOE up to 500 ug/L in Dl
water, BHI broth, PBS, or
0.9% NaCI
Arias and Yang (2009)
NOE up to 500 ug/L in water,
BHI broth, PBS, or 0.9% NaCI
Arias and Yang (2009)
Abbreviations: BHI = Brain heart infusion broth; COOH-MWCNT= Carboxylated, acid treated with COOH groups derived from the surface of CNTs by acid treatment; Dl = deionized
water; NH2-MWCNT=Functionalized with NH2 groups by activation of carboxyl moieties with thionyl chloride and subsequent reaction with CH3(CH2)i6CH2-NH2; CNQ= Could not
quantify; L = Length; LOEL = Lowest observed effect level; NOE = No observed effect; NOEL = Maximum no observed effects level; ND = Not determined; OD = Outer diameter;
P = Purity; PBS = Phosphate buffers saline; SA = Surface area; V = Volume.
"Source: Chung et al. (2011).
bSource: Rows 1-2 Simon-Deckers et al. (2009). created with aerosol-assisted catalytic chemical vapor deposition using Fe as catalyst, heat purified; row 3, Kang et al. (2009). purified
with hydrochloric acid then dispersed in 0.1 ug/mL ethanol sonicated for 10 min in a bath sonicator; rows 4-6: Kang et al. (2008).
°Source: Simon-Deckers et al. (2009). details same as footnote b for this source.
dSource: Kang et al. (2009). details same as footnote b for this source.
eSource: Row 1: Kang et al. (2009). details same as footnote b for this source; rows 2-4: Arias and Yang (2009). no details provided
'Source: Arias and Yang (2009). no details provided; OH-MWCNT: Functionalized with OH groups derived directly from the surface of CNTs.
F-50
-------
Table F-20. Effects of exposure to MWCNTs on plants.
MWCNT
properties
Exposure
protocol
Plant Doses tested
Effect
dose
Notes
Decreased germination rates3
SA: 126
m2/gram,
OD: 10-20 nm,
L: 1-2 pm
SA: 73 m2/gram,
OD: 40-60 nm,
L: NR
SA: 357
m2/gram,
OD: <10 nm,
L: NR
SA: NR,
OD:
110-170 nm,
L: 5-9 pm
30 seeds in
water, 5 days
10 seeds in
sewage sludge,
3 days
10 seeds in
sewage sludge,
3 days
60 seeds in
water, 4 days
Rapeseed 2,000 mg/L
Radish 2,000 mg/L
Ryegrass 2,000 mg/L
Corn 2,000 mg/L
Lettuce 2,000 mg/L
Cucumber 2,000 mg/L
Garden 0.01, 0.1,
cress 0.5% wt
Garden 0.01, 0.1,
cress 0.5% wt
Mustard 10, 20, 40 ug/mL
Urad 10, 20, 40 ug/mL
bean
NOE
NOE
NOE
NOE
NOE
NOE
LOEL
0.01%
LOEL
0.1%
NOE
NOE
Non-significant decrease (Row 1)a
Lin and Xing (2007)
Non-significant decrease (Row 2)a
Lin and Xing (2007)
None (Row3)a
Lin and Xing (2007)
Non-significant decrease (Row 4)a
Lin and Xing (2007)
No observed effect (Row 5)a
Lin and Xing (2007)
Non-significant decrease (Row 6)a
Lin and Xing (2007)
50-70% inhibition (compared to 10% in
control) (Row 7)a
Oleszczuk et al. (2011)
60% inhibition at 0.1%, 40% inhibition
at 0.5% (compared to 10% in control)
(Row 8)a
Oleszczuk et al. (2011)
None (Row9)a
Ghodake et al. (2010)
None (Row 10)a
Ghodake et al. (2010)
Changes in root growth3
>95% pure,
SA: 126
m2/gram,
OD: 10-20 nm,
L: 1-2 urn
30 seeds in
water, 5 days
Rapeseed 2,000 mg/L
Radish 2,000 mg/L
NOE
NOE
None (Row 11)a
Lin and Xing (2007)
None (Row 12)a
Lin and Xing (2007)
Ryegrass 2,000 mg/L
NOE Non-significant total root growth
decrease; root length increase
(Row13)a
Lin and Xing (2007)
F-51
-------
Table F-20 (Continued): Effects of exposure to MWCNTs on plants.
MWCNT
properties
>95% pure,
SA: 126
m2/gram,
OD: 10-20 nm,
L: 1-2 |jm
(continued)
Exposure
protocol
30 seeds in
water, 5 days
(continued)
Plant
Corn
Lettuce
Cucumber
Doses tested
20,00 mg/L
2,000 mg/L
2,000 mg/L
Effect
dose
NOE
NOE
NOE
Notes
None (Row 14)a
Lin and Xing (2007)
None (Row 15)a
Lin and Xing (2007)
None (Row 16)a
Lin and Xing (2007)
Changes in root growth3
SA: 357
m2/gram,
OD: <10 nm,
L: NR
SA: 73 m2/gram,
OD: 40-60 nm,
L: NR
SA: NR,
OD: 110-170
nm,
L: 5-9 urn
10 seeds in
sewage sludge,
3 days
10 seeds in
sewage sludge,
3 days
60 seeds in
water, 4 days
Garden
cress
Garden
cress
Mustard
Urad
bean
0.01,0.1, LOEL
0.5% wt 0.01%
0.01,0.1, NOE
0.5% wt
10, 20, 40ug/mL LOEL
10
ug/mL
10, 20, 40ug/mL NOE
Root length inhibition -30% greater than
control at all doses (Row 17)a
Oleszczuk et al. (2011)
None (Row 18)a
Oleszczuk et al. (2011)
138% increase in root length at
10 ug/mL,
202% increase at 20 ug/mL,
135% increase at 40 ug/mL
(Row19)a
Ghodake et al. (2010)
None (Row 20)a
Ghodake et al. (2010)
Cytotoxicity
SA: 250-300
m2/gram,
OD: 9.5 nm,
L: 1.5 urn
T87 cells in
suspension,
7 days (in
exponential
growth phase
on day 3)
Thale
cress
10 mg/L
LOEL Began on day 2; dose-dependent
10 mg/L inhibition observed; poor linearity of
curves; more severe with fine MWCNT
agglomerates than loose MWCNT
agglomerates
Lin et al. (2009)
Abbreviations: L = Length; LOEL = Lowest observed effect level; NOE = No observed effect; NR = Not reported; OD = Outer
diameter; SA = Surface area; V = Volume.
"Source: Rows 1-6 and 11-16: Lin and Xing (2007)-MWCNT purity >95%;
Rows 7-8 and 17-18: Oleszczuk et al. (2011)- MWCNT purity >95%;
Rows 9-10 and 19-20: Ghodake et al. (2010)- MWCNT purity 90%.
bSource: Lin et al. (2009). MWCNT carbon purity 90%; MWCNT properties reported for loose agglomerates, fine agglomerates had
same average diameter but other dimensions not analyzed.
F-52
-------
Appendix F References
Arias. LR: Yang. L. (2009). Inactivation of bacterial pathogens by carbon nanotubes in suspensions. Langmuir
25: 3003-3012. http://dx.doi.org/10.1021/la802769m
Asakura. M: Sasaki. T: Sugiyama. T: Takava. M: Koda. S: Nagano. K: Arito. H: Fukushima. S. (2010).
Genotoxicity and cytotoxicity of multi-wall carbon nanotubes in cultured Chinese hamster lung cells in
comparison with chrysotile A fibers. J Occup Health 52: 155-166.
Asharani PV: Serina. NGB: Nurmawati MH: Wu. YL: Gong. Z: Valiyaveettil S. (2008). Impact of multi-
walled carbon nanotubes on aquatic species. J Nanosci Nanotechnol 8: 3603-3609.
http://dx.doi.org/10.1166/inn.2008.432
Bottini M: Bruckner. S: Nika. K: Bottinj N: Bellucci S: Magrinj A: Bergamaschj A: Mustelin. T. (2006).
Multi-walled carbon nanotubes induce T lymphocyte apoptosis. Toxicol Lett 160: 121-126.
Bruchaizer. E: Frydrych. B: Sporny. S: Szymanska. JA. (2010). Toxicity of penta- and decabromodiphenyl
ethers after repeated administration to rats: a comparative study. Arch Toxicol 84: 287-299.
http://dx.doi.org/10.1007/s00204-009-0495-Y
Cai. Y: Zhang. W: Hu. J: Sheng. G: Chen. D: Fu. J. (2011). Characterization of maternal transfer of
decabromodiphenyl ether (BDE-209) administered to pregnant Sprague-Dawley rats. Reprod Toxicol 31:
106-110. http://dx.doi.0rg/10.1016/i.reprotox.2010.08.005
Cavallo. D: Fanizza. C: Ursini. CL: Casciardi S: Paba. E: Ciervo. A: Fresegna. AM: Maiello. R: Marcelloni.
AM: Buresti G: Tombolini. F: Bellucci. S: lavicoli. S. (2012). Multi-walled carbon nanotubes induce
cytotoxicity and genotoxicity in human lung epithelial cells. J Appl Toxicol 32: 454-464.
http://dx.doi.org/10.1002/iat.2711
Cheng. J: Chan. CM: Veca. LM: Poon. WL: Chan. PK: Qu. L: Sun. YP: Cheng. SH. (2009). Acute and long-
term effects after single loading of functionalized multi-walled carbon nanotubes into zebrafish (Danio rerio).
Toxicol Appl Pharmacol 235: 216-225. http://dx.doi.0rg/10.1016/j.taap.2008.12.006
Chung. H: Son. Y: Yoon. TK: Kim. S: Kim. W. (2011). The effect of multi-walled carbon nanotubes on soil
microbial activity. Ecotoxicol Environ Saf 74: 569-575. http://dx.doi.0rg/10.1016/i.ecoenv.2011.01.004
Cveticanin. J: Joksic. G: Leskovac. A: Petrovic. S: Sobot AV: Neskovic. O. (2010). Using carbon nanotubes to
induce micronuclei and double strand breaks of the DNA in human cells. Nanotechnology 21: 1-7.
http://dx.doi.Org/10.1088/0957-4484/21/l/015102
Deng. X: Jia. G: Wang. H: Sun. H: Wang. X: Yang. S: Wang. T: Liu. Y. (2007). Translocation and fate of multi-
walled carbon nanotubes in vivo. Carbon 45: 1419-1424.
Pi Sotto. A: Chiaretti. M: Carru. GA: Bellucci. S: Mazzanti. G. (2009). Multi-walled carbon nanotubes: Lack of
mutagenic activity in the bacterial reverse mutation assay. Toxicol Lett 184: 192-197.
http://dx.doi.0rg/10.1016/i.toxlet.2008.ll.007
Dow Chemical Co (Dow Chemical Company). (1990). Pulmonary clearance and tissue response following a
single intratracheal injection of decabromodiphenyl oxide (DBDPO) dust in male rats with attachment &
cover letter dated 030890 [TSCA Submission]. (86900000194). Baton Rouge, LA: Ethyl Corporation.
http://www.ntis.gov/search/product.aspx? ABBR=OTS0522264
el Dareer. SM: Kalin. JR: Tillery. KF: Hill. PL. (1987). Disposition of decabromobiphenyl ether in rats dosed
intravenously or by feeding. J Toxicol Environ Health 22: 405-415.
http://dx.doi.org/10.1080/15287398709531082
Elgrabli D: Abella-Gallart S: Robidel F: Rogerieux. F: Boczkowskj J: Lacroix. G. (2008a). Induction of
apoptosis and absence of inflammation in rat lung after intratracheal instillation of multiwalled carbon
nanotubes. Toxicology 253: 131-136. http://dx.doi.0rg/10.1016/j.tox.2008.09.004
F-53
-------
Elgrabli. D: Floriani. M: Abella-Gallart S: Meunier. L: Gamez. C: Delalain. P: Rogerieux. F: Boczkowskj J:
Lacroix. G. (2008b). Biodistribution and clearance of instilled carbon nanotubes in rat lung. Part Fibre
Toxicol 5: 20. http://dx.doi.org/10.1186/1743-8977-5-20
Ellinger-Ziegelbauer. H: Pauluhn. J. (2009). Pulmonary toxicity of multi-walled carbon nanotubes (Baytubes)
relative to alpha-quartz following a single 6h inhalation exposure of rats and a 3 months post-exposure
period. Toxicology 266: 16-29. http://dx.doi.0rg/10.1016/i.tox.2009.10.007
Ema. M: Matsuda. A: Kobavashj N: Nava. M: Nakanishi. J. (2011). Evaluation of dermal and eye irritation and
skin sensitization due to carbon nanotubes. Regul Toxicol Pharmacol 61: 276-281.
http://dx.doi.0rg/10.1016/i.vrtph.2011.08.007
Environment Canada. (2006). Canadian Environmental Protection Act, 1999: Ecological screening assessment
report on polybrominated diphenyl ethers (PBDEs).
Fujitani. T: Ohyama. K: Hirose. A: Nishimura. T: Nakae. D: Ogata. A. (2012). Teratogenicity of multi-wall
carbon nanotube (MWCNT) in ICR mice. J Toxicol Sci 37: 81-89.
Ghodake. G: Seo. YD: Park. D: Lee. PS. (2010). Phytotoxicity of carbon nanotubes assessed by Brassica Juncea
and Phaseolus Mungo. Journal of Nanoelectronics and Optoelectronics 5: 157-160.
http://dx.doi.org/10.1166/ino.2010.1084
Great Lakes Chemical Corporation. (1994). Initial submission: Letter from Great Lakes Chemical to USEPA Re:
Tetrabromobisphenol A, Pentabromoethylbenzene, Decabromodiphenyl Ether & Dibromopropyl Acrylate
with attachments dated 011184 [TSCA Submission]. (FYI-OTS-0794-1105). West Lafayette, IN.
http://www.ntis.gov/search/product.aspx?ABBR=OTS0001105
Hakk. H: Letcher. RJ. (2003). Metabolism in the toxicokinetics and fate of brominated flame retardants~a
review [Review]. Environ Int 29: 801-828. http://dx.doi.org/10.1016/80160-4120(03)00109-0
Hardy. M. (2002). The toxicology of the three commercial polybrominated diphenyl oxide (ether) flame
retardants. Chemosphere 46: 757-777. http://dx.doi.org/10.1016/s0045-6535(01)00240-5
Huwe. JK: Smith. DJ. (2007). Accumulation, whole-body depletion, and debromination of decabromodiphenyl
ether in male Sprague-Dawley rats following dietary exposure. Environ Sci Technol 41: 2371-2377.
Inoue. K: Koike. E: Yanagisawa. R: Hirano. S: Nishikawa. M: Takano. H. (2009). Effects of multi-walled
carbon nanotubes on a murine allergic airway inflammation model. Toxicol Appl Pharmacol 237: 306-316.
http://dx.doi.0rg/10.1016/i.taap.2009.04.003
IRDC (International Research & Development Corporation). (1974). Acute toxicity studies in rats and rabbits
with test data and cover letter dated 03-08-90. (86900000327). Great Lakes Chemical Corporation.
http://www.ntis.gov/search/product.aspx? ABBR=OTS0523319
Kang. S: Herzberg. M: Rodrigues. DF: Elimelech. M. (2008). Antibacterial effects of carbon nanotubes: Size
does matter. Langmuir 24: 6409-6413. http://dx.doi.org/10.1021/la800951v
Kang. S: Mauter. MS: Elimelech. M. (2009). Microbial cytotoxicity of carbon-based nanomaterials: Implications
for river water and wastewater effluent. Environ Sci Technol 43: 2648-2653.
http://dx.doi.org/10.1021/es8031506
Kato. T: Totsuka. Y: Ishino. K: Matsumoto. Y: Tada. Y: Nakae. D: Goto. S: Masuda. S: Ogo. S: Kawanishj M:
Yagi T: Matsuda. T: Watanabe. M: Wakabavashj K. (2013). Genotoxicity of multi-walled carbon nanotubes
in both in vitro and in vivo assay systems. Nanotoxicology 7: 452-461.
http://dx.doi.org/10.3109/17435390.2012.674571
Kennedy. AJ: Hull. MS: Steevens. JA: Dontsova. KM: Chappell. MA: Gunter. JC: Weiss. CA. Jr. (2008).
Factors influencing the partitioning and toxicity of nanotubes in the aquatic environment. Environ Toxicol
Chem27: 1932-1941. http://dx.doi.0rg/10.1897/07-624.l
Kierkegaard. A: Asplund. L: de Wit. CA: McLachlan. MS: Thomas. GO: Sweetman. AJ: Jones. KG. (2007). Fate
of higher brominated PBDEs in lactating cows. Environ Sci Technol 41: 417-423.
F-54
-------
Kierkegaard. A: Balk. L: Tjarnlund. U: De wit. CA: Jansson. B. (1999). Dietary uptake and biological effects of
Decabromodiphenyl Ether in Rainbow Trout (Oncorhynchus mykiss ). Environ Sci Technol 33: 1612-1617.
http://dx.doi.org/10.1021/es9807082
Kim. JS: Sung. JH: Song. KS: Lee. JH: Kim. SM: Lee. GH: An. KH: Lee. JS: Shin. JH: Park. JD: Yu. IJ.
(2012a). Persistent DNA damage measured by Comet assay of Sprague-Dawley rat lung cells after five days
of inhalation exposure and 1 month post-exposure to dispersed multi-wall carbon nanotubes (MWCNTs)
generated by new MWCNT aerosol generation system. Toxicol Sci 128: 439-448.
http://dx.doi.org/10.1093/toxsci/kfsl61
Kim. KT: Jang. MH: Kim. JY: Xing. B: Tanguav. RL: Lee. BG: Kim. SD. (2012b). Embryonic toxicity changes
of organic nanomaterials in the presence of natural organic matter. Sci Total Environ 426: 423-429.
http://dx.doi.0rg/10.1016/i.scitotenv.2012.03.050
Kishore. AS: Surekha. P: Murthy. PB. (2009). Assessment of the dermal and ocular irritation potential of multi-
walled carbon nanotubes by using in vitro and in vivo methods. Toxicol Lett 191: 268-274.
http://dx.doi.0rg/10.1016/i.toxlet.2009.09.007
Klaper. R: Arndt D: Setvowati K: Chen. J: Goetz. F. (2010). Functionalization impacts the effects of carbon
nanotubes on the immune system of rainbow trout, Oncorhynchus mykiss. Aquat Toxicol 100: 211-217.
http://dx.doi.0rg/10.1016/i.aquatox.2010.07.023
Kuo. YM: Sepulveda. MS: Sutton. TM: Ochoa-Acuna. HG: Muir. AM: Miller. B: Hua. I. (2010).
Bioaccumulation and biotransformation of decabromodiphenyl ether and effects on daily growth in juvenile
lake whitefish (Coregonus clupeaformis). Ecotoxicology 19: 751-760. http://dx.doi.org/10.1007/sl0646-009-
0451-x
Li. JG: Li. ON: Xu. JY: Cai XQ: Liu. RL: Li. YJ: Ma. JF: Li. WX. (2009). The pulmonary toxicity of multi-wall
carbon nanotubes in mice 30 and 60 days after inhalation exposure. J Nanosci Nanotechnol 9: 1384-1387.
Li. JG: Li. WX: Xu. JY: Cai. XO: Liu. RL: Li. YJ: Z. OF: Li. ON. (2007). Comparative study of pathological
lesions induced by multi-walled carbon nanotubes in lungs of mice by intratracheal instillation and
inhalation. Environ Toxicol 22: 415-421. http://dx.doi.org/10.1002/tox.20270
Li. M: Huang. CP. (2011). The responses of Ceriodaphnia dubia toward multi-walled carbon nanotubes: Effect
of physicalchemical treatment. Carbon 49: 1672-1679. http://dx.doi.0rg/10.1016/i.carbon.2010.12.052
Li W: Zhu. L: Zha. J: Wang. Z. (2011). Effects of decabromodiphenyl ether (BDE-209) on mRNA transcription
of thyroid hormone pathway and spermatogenesis associated genes in Chinese rare minnow (Gobiocypris
rams). Environ Toxicol. http://dx.doi.org/10.1002/tox.20767
Lim. JH: Kim. SH: Shin. IS: Park. NH: Moon. C: Kang. SS: Park. SC: Kim. JC. (2011). Maternal exposure to
multi-wall carbon nanotubes does not induce embryo-fetal developmental toxicity in rats. Birth Defects Res
B Dev Reprod Toxicol 92: 69-76. http://dx.doi.org/10.1002/bdrb.20283
Lin. C: Fugetsu. B: Su. Y: Watari. F. (2009). Studies on toxicity of multi-walled carbon nanotubes on
Arabidopsis T87 suspension cells. J Hazard Mater 170: 578-583.
http://dx.doi.0rg/10.1016/i.jhazmat.2009.05.025
Lin. D: Xing. B. (2007). Phytotoxicity of nanoparticles: inhibition of seed germination and root growth. Environ
Pollut 150: 243-250. http://dx.doi.0rg/10.1016/i.envpol.2007.01.016
Liu. L: Zhu. W: Xiao. L: Yang. L. (2011). Effect of decabromodiphenyl ether (BDE 209) and dibromodiphenyl
ether (BDE 15) on soil microbial activity and bacterial community composition. J Hazard Mater 186: 883-
890. http://dx.doi.0rg/10.1016/i.jhazmat.2010.ll.079
LSRI (Life Science Research Israel). (1984). Acute oral range finding study in the rat. (LSRI Schedule No.
DSB/052/HBCD). Beer Sheva, Israel: Dead Sea Bromine Company.
Ma-Hock. L: Treumann. S: Strauss. V: Brill S: Luizi. F: Mertler. M: Wiench. K: Gamer. AO: van Ravenzwaav.
B: Landsiedel R. (2009). Inhalation toxicity of multiwall carbon nanotubes in rats exposed for 3 months.
Toxicol Sci 112: 468-481. http://dx.doi.org/10.1093/toxsci/kfpl46
F-55
-------
Migliore. L: Saracino. D: Bonelli. A: Colognato. R: D'Errico. MR: Magrini. A: Bergamaschi. A: Bergamaschi
K (2010). Carbon nanotubes induce oxidative DNA damage in RAW 264.7 cells. Environ Mol Mutagen 51:
294-303. http://dx.doi.org/10.1002/em.20545
Mitchell LA: Gao. J: Vander Wai R: Gigliotti. A: Burchiel SW: Mcdonald. JD. (2007). Pulmonary and
systemic immune response to inhaled multiwalled carbon nanotubes. Toxicol Sci 100: 203-214.
http://dx.doi.org/10.1093/toxsci/kfml96
Mitchell. LA: Lauer. FT: Burchiel. SW: McDonald. JD. (2009). Mechanisms for how inhaled multiwalled
carbon nanotubes suppress systemic immune function in mice. Nat Nanotechnol 4: 451-456.
http://dx.doi.org/10.1038/nnano.2009.151
Morck. A: Hakk. H: Orn. U: Klasson Wehler. E. (2003). Decabromodiphenyl ether in the rat: Absorption,
distribution, metabolism, and excretion. Drug Metab Dispos 31: 900-907.
http://dx.doi.0rg/10.1124/dmd.31.7.900
Muller. J: Decordier. I: Hoet PH: Lombaert. N: Thomassen. L: Huaux. F: Lison. D: Kirsch-Volders. M. (2008a).
Clastogenic and aneugenic effects of multi-wall carbon nanotubes in epithelial cells. Carcinogenesis 29: 427-
433. http://dx.doi.org/10.1093/carcin/bgm243
Muller. J: Delos. M: Panin. N: Rabolli. V: Huaux. F: Lison. D. (2009). Absence of carcinogenic response to
multiwall carbon nanotubes in a 2-year bioassay in the peritoneal cavity of the rat. Toxicol Sci 110: 442-448.
http://dx.doi.org/10.1093/toxsci/kfplOO
Muller. J: Huaux. F: Fonseca. A: Nagv. JB: Moreau. N: Delos. M: Raymundo-Pinero. E: Beguin. F: Kirsch-
Volders. M: Fenoglio. I: Fubini. B: Lison. D. (2008b). Structural defects play a major role in the acute lung
toxicity of multiwall carbon nanotubes: lexicological aspects. ChemRes Toxicol 21: 1698-1705.
http://dx.doi.org/10.1021/tx800101p
Muller. J: Huaux. F: Moreau. N: Misson. P: Heilier. JF: Delos. M: Arras. M: Fonseca. A: Nagv. JB: Lison. D.
(2005). Respiratory toxicity of multi-wall carbon nanotubes. Toxicol Appl Pharmacol 207: 221-231.
http://dx.doi.0rg/10.1016/i.taap.2005.01.008
Nakari. T: Pessala. P. (2005). In vitro estrogenicity of polybrominated flame retardants. Aquat Toxicol 74: 272-
279. http://dx.doi.org/10.1016/i.aquatox.2005.06.004
Norris. JM: Ehrmantraut JW: Gibbons. CL: Kociba. RJ: Schwetz. BA: Rose. JQ: Humiston. CG: Jewett. GL:
Crummett. WB: Gehring. PJ: Tirsell JB: Brosier. JS. (1973). Toxicological and environmental factors
involved in the selection of decabromodiphenyl oxide as a fire retardant chemical. Appl Polym Symp 22:
195-219.
Norris. JM: Kociba. RJ: Schwetz. BA: Rose. JQ: Humiston. CG: Jewett GL: Gehring. PJ: Mailhes. JB. (1975).
Toxicology of octabromobiphenyl and decabromodiphenyl oxide. Environ Health Perspect 11: 153-161.
NRC (National Research Council). (2000). Toxicological risks of selected flame-retardant chemicals.
Washington, DC: National Academy Press.
NTP (National Toxicology Program). (1986). NTP toxicology and carcinogenesis studies of Decabromodiphenyl
Oxide (CAS No. 1163-19-5) inF344/N rats andB6C3Fl mice (Feed studies) (pp. 1-242). (309). Research
Triangle Park, NC: U.S. Department of Health and Human Services.
Nygaard. UC: Hansen. JS: Samuelsen. M: Alberg. T: Marioara. CD: Levik. M. (2009). Single-walled and multi-
walled carbon nanotubes promote allergic immune responses in mice. Toxicol Sci 109: 113-123.
http://dx.doi.org/10.1093/toxsci/kfp057
Oleszczuk. P: Josko. I: Xing. B. (2011). The toxicity to plants of the sewage sludges containing multiwalled
carbon nanotubes. J Hazard Mater 186: 436-442. http://dx.doi.0rg/10.1016/i.jhazmat.2010.ll.028
Park. EJ: Cho. WS: Jeong. J: Yi. J: Choi. K: Park. K. (2009). Pro-inflammatory and potential allergic responses
resulting from B cell activation in mice treated with multi-walled carbon nanotubes by intratracheal
instillation. Toxicology 259: 113-121. http://dx.doi.0rg/10.1016/i.tox.2009.02.009
Patlolla. A: Knighten. B: Tchounwou. P. (2010a). Multi-walled carbon nanotubes induce cytotoxicity,
genotoxicity and apoptosis in normal human dermal fibroblast cells. Ethn Dis 20: SI-65-72.
F-56
-------
Patlolla. A: Patlolla. B: Tchounwou. P. (2010b). Evaluation of cell viability, DNA damage, and cell death in
normal human dermal fibroblast cells induced by functionalized multiwalled carbon nanotube. Mol Cell
Biochem 338: 225-232. http://dx.doi.org/10.1007/sll010-009-0356-2
Patlolla. AK: Berry. A: Tchounwou. PB. (2011). Study of hepatotoxicity and oxidative stress in male Swiss-
Webster mice exposed to functionalized multi-walled carbon nanotubes. Mol Cell Biochem 358: 189-199.
http://dx.doi.org/10.1007/sll010-011-0934-v
Pauluhn. J. (2010). Subchronic 13-week inhalation exposure of rats to multiwalled carbon nanotubes: Toxic
effects are determined by density of agglomerate structures, not fibrillar structures. Toxicol Sci 113: 226-242.
http://dx.doi.org/10.1093/toxsci/kfp247
Poland. CA: Puffin. R: Kinloch. I: Maynard. A: Wallace. WAH: Seaton. A: Stone. V: Brown. S: MacNee. W:
Donaldson. K. (2008). Carbon nanotubes introduced into the abdominal cavity of mice show asbestos-like
pathogenicity in a pilot study. Nat Nanotechnol 3: 423-428. http://dx.doi.org/10.1038/nnano.20Q8.111
Porter. DW: Hubbs. AF: Mercer. RR: Wu. N: Wolfarth. MG: Sriram. K: Leonard. S: Battelli. L: Schwegler-
Berry. D: Friend. S: Andrew. M: Chen. BT: Tsuruoka. S: Endo. M: Castranova. V. (2010). Mouse pulmonary
dose- and time course-responses induced by exposure to multi-walled carbon nanotubes. Toxicology 269:
136-147. http://dx.doi.0rg/10.1016/i.tox.2009.10.017
Qin. X: Xia. X: Yang. Z: Yan. S: Zhao. Y: Wei. R: Li. Y: Tian. M: Zhao. X: Qin. Z: Xu. X. (2010). Thyroid
disruption by technical decabromodiphenyl ether (DE-83R) at low concentrations in Xenopus laevis. J
Environ Sci 22: 744-751.
Radomski. A: Jurasz. P: Alonso-Escalano. D: Drews. M: Morandi. M: Malinski. T: Radomski. MW. (2005).
Nanoparticle-induced platelet aggregation and vascular thrombosis. Br JPharmacol 146: 882-893.
Reddv. AR: Krishna. PR: Reddv. YN: Himabindu. V. (2010). Translocation and extra pulmonary toxicities of
multi wall carbon nanotubes in rats. Toxicol Mech Meth 20: 267-272.
http://dx.doi.org/10.3109/15376516.2010.484077
Riu. A: Cravedi. JP: Debrauwer. L: Garcia. A: Canlet C: Jouanin. I: Zalko. D. (2008). Disposition and metabolic
profiling of [14C]-decabromodiphenyl ether in pregnant Wistar rats. Environ Int 34: 318-329.
http://dx.doi.0rg/10.1016/i.envint.2007.03.007
Riva. C: Binelli. A: Cogni. D: Provini. A. (2007). Evaluation of DNA damage induced by decabromodiphenyl
ether (BDE-209) in hemocytes of Dreissena polymorpha using the comet and micronucleus assays. Environ
Mol Mutagen48: 735-743. http://dx.doi.org/10.1002/em.20353
Ryman-Rasmussen. JP: Cesta. MF: Brodv. AR: Shipley-Phillips. JK: Everitt JI: Tewksbury. EW: Moss. OR:
Wong. BA: Dodd. DE: Andersen. ME: Bonner. JC. (2009a). Inhaled carbon nanotubes reach the subpleural
tissue in mice. Nat Nanotechnol 4: 747-751. http://dx.doi.org/10.1038/nnano.2009.305
Ryman-Rasmussen. JP: Tewksbury. EW: Moss. OR: Cesta. MF: Wong. BA: Bonner. JC. (2009b). Inhaled
multiwalled carbon nanotubes potentiate airway fibrosis in murine allergic asthma. Am J Respir Cell Mol
Biol 40: 349-358. http://dx.doi.org/10.1165/rcmb.2008-0276OC
Sakamoto. Y: Nakae. D: Fukumori. N: Tavama. K: Maekawa. A: Imai. K: Hirose. A: Nishimura. T: Ohashi. N:
Ogata. A. (2009). Induction of mesothelioma by a single intrascrotal administration of multi-wall carbon
nanotube in intact male Fischer 344 rats. J Toxicol Sci 34: 65-76.
Sandholm. A: Emanuelsson. BM: Wehler. EK. (2003). Bioavailability and half-life of decabromodiphenyl ether
(BDE-209) in rat. Xenobiotica 33: 1149-1158. http://dx.doi.org/10.1080/00498250310001609156
Simon-Deckers. A: Loo. S: Mayne-L'hermite. M: Herlin-Boime. N: Menguy. N: Reynaud. C: Gouget B:
Carriere. M. (2009). Size-, composition- and shape-dependent toxicological impact of metal oxide
nanoparticles and carbon nanotubes toward bacteria. Environ Sci Technol 43: 8423-8429.
http://dx.doi.org/10.1021/es9016975
Stapleton. HM: Kelly. SM: Pel R: Letcher. RJ: Gunsch. C. (2009). Metabolism of polybrominated diphenyl
ethers (PBDEs) by human hepatocytes in vitro. Environ Health Perspect 117: 197-202.
F-57
-------
Sverdrup. LE: Hartnik. T: Mariussen. E: Jensen. J. (2006). Toxicity of three halogenated flame retardants to
nitrifying bacteria, red clover (Trifolium pratense), and a soil invertebrate (Enchytraeus crypticus).
Chemosphere 64: 96-103. http://dx.doi.0rg/10.1016/i.chemosphere.2005.ll.056
Szendi K: Varga. C. (2008). Lack of genotoxicity of carbon nanotubes in a pilot study. Anticancer Res 28: 349-
352.
Takagi. A: Kitaiima. S: Kanno. J: Hirose. A: Nishimura. T: Fukumori. N: Ogata. A: Ohashi. N. (2008). Induction
of mesothelioma in p53+/- mouse by intraperitoneal application of multi-wall carbon nanotube. J Toxicol Sci
33: 105-116.
Tomy. GT: Palace. VP: Halldorson. T: Braekevelt E: Danell R: Wautier. K: Evans. B: Brinkworth. L: Fisk. AT.
(2004). Bioaccumulation, biotransformation, and biochemical effects of brominated diphenyl ethers in
juvenile Lake Trout ( Salvelinus namaycush). Environ Sci Technol 38: 1496-1504.
http://dx.doi.org/10.1021/es035070v
Van der Ven. LT: van de Kuil. T: Leonards. PE: Slob. W: Canton. RF: Germer. S: Visser. TJ: Litens. S:
Hakansson. H: Schrenk. D: van den Berg. M: Piersma. AH: Vos. JG: Opperhuizen. A. (2008). A 28-day oral
dose toxicity study in Wistar rats enhanced to detect endocrine effects of decabromodiphenyl ether
(decaBDE). Toxicol Lett 179: 6-14. http://dx.doi.0rg/10.1016/j.toxlet.2008.03.003
Varga. C: Szendi. K. (2010). Carbon nanotubes induce granulomas but not mesotheliomas. In Vivo 24: 153-156.
Velzeboer. I: Kupryianchyk. D: Peeters. ETH. M: Koelmans. AA. (2011). Community effects of carbon
nanotubes in aquatic sediments. Environ Int 37: 1126-1130. http://dx.doi.0rg/10.1016/i.envint.2011.01.016
Viberg. H: Fredriksson. A: Jakobsson. E: Orn. U: Eriksson. P. (2003). Neurobehavioral derangements in adult
mice receiving decabrominated diphenyl ether (PBDE 209) during a defined period of neonatal brain
development. Toxicol Sci 76: 112-120. http://dx.doi.org/10.1093/toxsci/kfg210
Wang. F: Wang. J: Dai. J: Hu. G: Luo. X: Mai. B. (2010). Comparative tissue distribution, biotransformation and
associated biological effects by decabromodiphenyl ethane and decabrominated diphenyl ether in male rats
after a 90-day oral exposure study. Environ Sci Technol 44: 5655-5660. http://dx.doi.org/10.1021/esl01158e
Wei. L: Thakkar. M: Chen. Y: Ntim. SA: Mitra. S: Zhang. X. (2010). Cytotoxicity effects of water dispersible
oxidized multiwalled carbon nanotubes on marine alga, Dunaliella tertiolecta. Aquat Toxicol 100: 194-201.
http://dx.doi.0rg/10.1016/i.aquatox.2010.07.001
Wirnitzer. U: Herbold. B: Voetz. M: Ragot J. (2009). Studies on the in vitro genotoxicity of baytubes,
agglomerates of engineered multi-walled carbon-nanotubes (MWCNT). Toxicol Lett 186: 160-165.
http://dx.doi.0rg/10.1016/i.toxlet.2008.ll.024
Xie. X: Wu. Y: Zhu. M: Zhang. YK: Wang. X. (2011). Hydroxyl radical generation and oxidative stress in
earthworms (Eisenia fetida) exposed to decabromodiphenyl ether (BDE-209). Ecotoxicology 20: 993-999.
http://dx.doi.org/10.1007/sl0646-011-0645-x
Zhu. L: Chang. DW: Dai. L: Hong. Y. (2007). DNA damage induced by multiwalled carbon nanotubes in mouse
embryonic stem cells. Nano Lett 7: 3592-3597. http://dx.doi.org/10.1021/nl071303v
F-58
-------
Appendix G. Unprioritized Areas of the
CEA Framework for MWCNTs
G.1. Introduction to this Appendix
As described in Chapter 1. the Final case study document has been streamlined to clearly reflect
the outcomes of the collective judgment step of the Comprehensive Environmental Assessment (CEA)
process. Certain topics within the CEA framework (Figure 1-3) were designated priorities for research,
based on high importance for risk assessment and low confidence that the current data could support risk
management decisions. Information related to these research priority areas is located in Chapter 2 through
Chapter 5 of the document. All detailed information on multiwalled carbon nanotubes (MWCNTs) that
was not identified as a priority research area during the collective judgment step of the CEA process has
been moved to this appendix. The designation of certain areas as "unprioritized" is not intended to imply
that the topics are unimportant or that continued research is not needed; it simply implies that the topic
was determined to be of lesser importance for risk assessment.
G.2. Product Life Cycle
G.2.1. Feedstocks
G.2.1.1. Life-Cycle Processes
A wide variety of hydrocarbons and catalysts are used to synthesize MWCNTs. The raw
materials required for MWCNT synthesis include a precursor carbon material, an inert gas, and metal
catalysts, with other specific materials depending on the particular synthetic pathway used (as described
in Section 2.2) (Moisala et al., 2003). Support materials such as aluminum, manganese oxide, or silica are
also used during synthesis of MWCNTs (Gustavsson et al. 2011).
G-1
-------
Table G-1. Percent yields for agglomerated growth of SWCNTs and MWCNTs using various
synthesis methods, processing temperatures, and catalysts.
SWCNTs3
Method of
Synthesis
CVD
CVD
CVD
CVD
CVD
CVD
CVD
CVD
CVD
CVD
Processing
Catalyst Temperature
Used (°C) Yield (%)
Fe(Mo)AI2O3 900 0.1-10
Co/Mo/SiO2 600-800 0.33-1.8
Fe/MgO 900 5.2
Fe/Co/MgO 1,000 5.5-7.6
Fe/MgO 850 8-20
Fe/MgO 900 11
Fe/Mg/AI-LDH 900 17.6
Fe/Mo/AI2O3 850 20-60
Fe(CO)5 1,200 25-44
Fe/Mo/MgO 800 550
MWCNTs3
Method
of
Synthesis
FBCVD
FBCVD
CVD
FBCVD
CVD
FBCVD
FBCVD
CVD
CVD
CVD
CVD
CVD
FBCVD
FBCVD
CVD
CVD
CVD
CVD
FBCVD
CVD
CVD
FBCVD
CVD
CVD
Catalyst
Used
Fe/AI2O3
NiAI2O3
Co/Mo/AI2O3
Ni/SiO2
Co/W/MgO
Fe/SiO2
Fe/AI2O3
Ni/Mo/MgO
Fe/Co/AI2O3
Co/Mo/MgO
Fe/SiO2
Mo/MgO
Fe/Mo/MgO
Fe(Ni)/AI2O3
Ni/Mg/AI-LDH
Ni/SiO2
Ni/MgO
Co/AI-LDH
Fe/Mo/AI2O3
Co/Mo/AI2O3
Co/AI-LDH
Fe/CO/CaCO3
Ni/Fe/AI2O3
Co/Mn/Zn/AI
Processing
Temperature
(°C)
500-700
650-800
700
450-850
1,000
550-1,050
550-750
1,000
700
1,000
650-800
900
600-1,000
700-850
700
680
600
700
850
700
850
600-850
600
650
Yield (%)
1-20
2-17
2-25
2-145
4-47
10-50
10-70
10-100
14-56
16
30-116
33.4
66-400
70-300
109-254
124^26
166^80
188
274
280^80
560-625
1,100
6,000
17,900
"Agglomerated growth differs from vertical and horizontal growth; however, no yield data were provided for vertical or horizontal
growth of CNTs.
Abbreviations: CVD = chemical vapor deposition; FBCVD = fluidized bed chemical vapor deposition; LDH = layered double
hydroxide.
Source: Reprinted with permission of John Wiley and Sons; Zhang et al. (2011).
G-2
-------
No quantitative data were found on the total magnitude of feedstocks that are or might be used in
commercial synthesis of MWCNTs. Table G-L however, lists the amount of inputs required to synthesize
1 gram of single-walled carbon nanotubes (SWCNTs) using two of the common forms of carbon
nanotube (CNT) synthesis (both of which are discussed in Section 2.2.2): chemical vapor deposition
(CVD) and fluidized bed chemical vapor deposition (FBCVD). According to Healy et al. (2008). SWCNT
synthesis requires large quantities of feedstocks, or inputs, compared to outputs. These inputs were
calculated, however, assuming very low synthesis reaction yield (2.95%-4.50%), or mass of CNTs
divided by the mass of carbon fed into the system. Current synthesis reaction yields can range from 1% to
17,900% depending on synthesis method, choice of catalyst, and organizational structure (i.e., vertically
aligned, agglomerated, horizontally aligned; the largest synthesis yields are obtained from agglomerated
MWCNT growth) (Zhang et al.. 2011). Process optimization has led to dramatically improved yields over
the past few years (Zhang et al.. 2011). SWCNT reaction yields also have been reported to be typically
much lower than MWCNT reaction yields (Zhang et al.. 2011) (see Table G-l).31
Limited information suggests that MWCNT synthesis requires more precursor material than
SWCNT synthesis (Tsai et al.. 2009); however, no information regarding the mass of inputs of precursor
materials and catalysts was identified for MWCNTs.
G.2.1.2. Potential Releases during the Feedstock Extraction Stage
Release of MWCNTs would not occur during this initial phase of the life cycle—the feedstock
extraction stage—given that synthesis does not occur until the next stage. Hazardous raw materials,
however, could be released during the extraction and processing of feedstock materials. Release of CNTs
also could occur if reactors are not cleaned between runs. The specific raw materials that could be
released depends on the method of production, but likely would include catalyst metals and carbon
precursor materials. No data quantifying the potential volume of releases during feedstock extraction for
MWCNTs were identified.
31 An alternative method for estimating magnitude of MWCNT feedstock inputs is to consider magnitude of
feedstock inputs for carbon-carbon composites. The processes to produce carbon-carbon composites and MWCNTs
are the same, with the exception that producing MWCNTs requires the use of metal catalysts [Personal
Communication: K. Lafdi (University of Dayton). 11/16/12].
G-3
-------
G.2.2. Storage and Distribution
G.2.2.1. Life-Cycle Processes
The storage and distribution stage involves the handling and transport of (1) MWCNTs,
(2) MWCNT flame-retardant formulations, and (3) MWCNT flame-retardant upholstery textiles.
The principal method of transport for these materials is not known, but likely would be by truck, train, or
cargo ship.
• MWCNTs, if not immediately incorporated into a flame-retardant formulation, likely would
be stored at the site of synthesis/processing in sealed receptacles until they are incorporated
into flame-retardant formulations or transported to the sites where flame-retardant
manufacture would occur.
• MWCNT flame-retardant formulations are typically stored at manufacturing plants in
drums, tanks, or more permanent storage vessels until they are packaged and sent to textile
manufacturers (U.S. EPA. 2005).
• MWCNT flame-retardant upholstery textiles and end-use products likely also would be
stored at the site of manufacture (or an intermediate storage site) and then transported to retail
locations.
G.2.2.2. Potential Releases during Storage and Distribution
Storage and distribution of (1) MWCNTs, (2) MWCNT flame-retardant formulations, and
(3) MWCNT flame-retardant upholstery textiles could result in the following releases to the environment,
but all release scenarios are unlikely.
• Releases of MWCNTs prior to incorporation in flame-retardant formulations are likely to be
negligible. Release would be due primarily to accidents, as the MWCNTs would be stored in
sealed receptacles after synthesis. Exposure of the receptacles to high heat or fire could lead
to the airborne release of MWCNTs (see Section 2.2.2.2 for more details).
• Releases of flame-retardant formulations could result in releases of MWCNTs to the
environment (U.S. EPA. 2005). The possible scenarios for release of MWCNT flame-
retardant formulations during storage and distribution include damage to containers holding
the flame-retardant formulation, leakage resulting from mishandling of containers, or faulty
or improper stacking of cartons in transport vehicles. If the containers are sealed properly and
not damaged during transport, releases of product prior to application might be limited to
spills.
• Releases of flame-retardant upholstery textiles could result from accidental exposure to
high heat or fire, off-gassing of volatile components, and infestation with pests. Exposure of
the flame-retardant upholstery textiles to high heat or fire during storage and distribution
could lead to the degradation of the polymer matrix and subsequent airborne release of
MWCNTs (both free and matrix bound) (see Sections 2.2.2.2 and 2.4.2 for more details).
G-4
-------
• Off-gassing of the volatile components of MWCNT flame retardants also could occur in
poorly ventilated areas that experience high temperatures (e.g., storage units, warehouses).
MWCNTs per se, however, are not highly volatile (see Table 1-9 and Table 3-1). Infestation
of textile or furniture storage facilities with rodents or other pests also could lead to the
release of MWCNT flame-retardant materials to the environment.
Table G-2 outlines potential release scenarios from the storage and distribution stage of
(1) MWCNTs, (2) MWCNT flame-retardant formulations, and (3) MWCNT flame-retardant upholstery
textiles. Parallel potential release scenarios for decabromodiphenyl ether(decaBDE) are provided in the
table for comparative purposes; more detailed information on release scenarios for decaBDE is provided
in Appendix H.
Table G-2. Potential release scenarios during storage and distribution.
Information on release
Processes included in storage
and distribution life-cycle stage
DecaBDE
MWCNTs
Storage/transport of raw materials
(decaBDE and MWCNTs)
Release unlikely if properly
stored
Release unlikely if properly stored
Accidental releases of raw
materials (decaBDE and
MWCNTs)
Air release possible due to
storage container defects
Air release possible due to
storage container defects
Storage/transport of flame-
retardant formulation
Release unlikely if properly
stored
Release unlikely if properly stored
Accidental releases of flame-
retardant formulation
Water release possible due to
spills from mishandling or faulty
packaging
Water release possible due to
spills from mishandling or faulty
packaging
Storage/transport of treated
textiles
Small air release possible if
properly stored
Small air release possible if
properly stored
Accidental releases of treated
textiles
Air release possible due to
exposure to high heat, pest
infestation, etc.
Release possible due to pest
infestation; preliminary evidence
suggests that release due to high
heat is unlikely
G-5
-------
G.3. Transport, Transformation, and Fate
G.3.1. Transport, Transformation, and Fate in Water and Sediment
G.3.1.1. Surface Water and Sediment (Inland and Coastal)
Although the transport, transformation, and fate of MWCNTs in surface water was not identified
as a priority area, the transport, transformation, and fate of MWCNTs in sediment was. Because of the
limited available data, which overlaps between surface water and sediment, however, these topics are
discussed together in the main body of the document (Section 3.3.1) and therefore the surface water
discussion was not extracted and presented here as a separate area.
G.3.1.2. Ground Water
MWCNTs in soil could leach into subsoil and ground water and migrate to surface water;
however, no data were found on concentrations of MWCNTs in ground water (see Table 3-2 for estimates
from modeling studies).
G.3.2. Transport, Transformation, and Fate in Soil
MWCNTs released from textile products can enter terrestrial ecosystems and be transported in
several ways. Early reviews speculated that the propensity of MWCNTs to adsorb to soil surfaces could
make them less mobile (Borm et al.. 2006; Wiesner et al.. 2006). Recently, researchers showed that
MWCNTs modified with surface coatings to enhance their aqueous stability or change their surface
charge behave in the environment differently than pure MWCNTs (Petersen et al.. 201 la).
Petersen et al. (2011 a) examined sorption profiles of pure MWCNTs and MWCNTs
functionalized with a polyethyleneimine surface coating and determined that sorption isotherms for pure
MWCNTs were nearly linear, whereas isotherms for modified MWCNTs were nonlinear, indicating that
surface coating can influence MWCNT interactions with soils. The authors also suggested that MWCNT
characteristics (such as presence of surface coating) are better predictors of sorption behavior than soil
type (and organic carbon content).
Properties of the soil environment (e.g., soil type, soil organic matter, pH, ionic strength, presence
of other pollutants) also could affect particle transport. General information on how those properties
affect nanoparticles (not specific to MWCNTs) is available in the literature (Navarre et al.. 2008; U.S.
EPA. 2007).
G-6
-------
If MWCNTs are present in soils, plant roots could interact with those associated with soil
material and in soil pore water (Navarre et al., 2008). Plants could also be exposed to MWCNTs in air
and water. Airborne MWCNTs could attach to leaves and other aerial parts of plants and be translocated
to different tissues of the plant, in which case plants also might act as transfer vectors for MWCNTs in the
food chain. Additionally, bioaccumulation might be possible for carbon-based nanomaterials (Navarre et
al.. 2008).
Studies relevant to the fate and transport of CNTs in soil are provided in Appendix D. Table D-4.
Studies that examined MWCNT uptake, translocation, and transformation in plants were not found.
Literature that presents soil concentrations of MWCNTs has not been identified, although estimates from
modeling studies are available (see Table 3-2).
G.4. Exposure-Dose
G.4.1. Human Exposure and Kinetics Leading to Dose
G.4.1.1. General Public Exposure Pathway Scenarios through Environmental Media
No information was found on exposure to MWCNTs in the general public from environmental
media (e.g., air, water, soil). See Section 4.1.2 for model estimates of MWCNT concentrations in
environmental media that could be used with the exposure pathway and scenario characteristics below to
estimate potential exposures.
G.4.1.1.1 Outdoor Air
Releases of MWCNTs to outdoor air throughout the product life cycle of the flame-retardant
textile coatings are possible (see Chapter 2 and Section G.2). Once MWCNTs are released to air, they
might sorb or attach, depending on the surface coating and functionalization, to particulate matter and be
subject to long-range transport to areas distant from their source (see Section 3.2). Although this
phenomenon has not been observed for MWCNTs, it has been observed for other compounds and no
evidence yet exists to preclude the possibility that it would occur for MWCNTs. No data are available on
MWCNT concentrations in ambient air; however, modeling studies provide some estimates (see Table
3-2) and general public exposure pathways could be similar to those observed for particulate-phase
decaBDE (see Appendix H).
G-7
-------
Other product constituents of flame-retardant textiles (e.g., pieces of the polymer matrix or the
textile fabric) also can be released. As discussed in Section 3.2. the physicochemical properties of
MWCNTs released to air might change over time as a result of aging, which could result in exposure of
the general public to different MWCNTs than those that were first synthesized or released.
G.4.1.1.2 Water
Releases of MWCNTs and other product constituents to wastewater and ambient water bodies are
possible throughout the product life cycle of flame-retardant textile coatings (see Section G.2). Once
released to water, MWCNTs are expected to sorb to particulate matter in the water column or to
sediments, which might limit their mobility (see Section 3.3). This behavior implies that MWCNTs also
primarily will be removed to sludge during wastewater treatment. No data are available on MWCNT
concentrations in surface waters (see Section 4.1.2.2); however, modeling studies provide some estimates
(see Table 3-2) and general public exposure pathways could be similar to those observed for particulate-
phase decaBDE (see Appendix H). MWCNT surface functionalization, however, might affect stability of
free MWCNTs in water and efficacy of water treatment methods in removal of MWCNTs (see Section
3.3.3). which could result in more or less exposure to MWCNTs in surface and drinking water, depending
on the type of functionalization.
G.4.1.1.3 Soil
Releases to ambient air and water throughout the product life cycle of flame-retardant textile
coatings will result in deposition of MWCNT particles and other product constituents (see Section G.2) to
soil. Once deposited, MWCNTs are expected to sorb strongly to soil, which might limit their mobility
(see Section 3.4). No data are available on MWCNT concentrations in surface soils (see Section 4.1.2.3);
however, modeling studies provide some estimates (see Table 3-2) and general public exposure pathways
could be similar to those observed for particulate-phase decaBDE (see Appendix H).
G.4.2. Ecological Exposure and Kinetics Leading to Dose
G.4.2.1. Factors Impacting Ecological Exposure
In biota, potential exposure routes for MWCNTs include ingestion, inhalation, or direct contact.
The potential for exposure via each route along with subsequent uptake and dose depends on several
factors, including properties of the environmental media and physiological and behavioral characteristics
of aquatic and terrestrial organisms. These factors can, in turn, influence the bioavailability of MWCNTs.
Go
-o
-------
As discussed in Chapter 3 and Section H.3. the physicochemical properties of MWCNTs dictate their
partitioning in the environment. This partitioning drives the exposure potentials for water-dwelling,
sediment-dwelling, and terrestrial organisms. For example, CNTs without functionalizing surfactants are
hydrophobic and will interact with other CNTs and organic matter in aquatic systems, resulting in stable
suspensions and bundling followed by sedimentation (Koelmans et al.. 2009; Hyung et al.. 2007). Stable
suspensions and settling allow for exposure of both water-column and benthic organisms to MWCNTs in
aquatic systems (Velzeboer et al.. 2011).
Properties of the environmental media also can influence exposure potential for MWCNTs by
affecting bioavailability and MWCNT form. For example, the presence of dissolved organic matter in an
aquatic system can cause MWCNTs to debundle. Bacterial studies have shown that debundling of
MWCNTs can result in greater cytotoxicity (Kang et al.. 2009). Changes in properties such as ionic
strength or the pH of a solution might influence sorption behaviors of CNTs (Petersen et al.. 201 la).
which could differentially alter exposure levels of benthic and water-column organisms.
G.4.2.2. Absorption, Distribution, Metabolism and Excretion in Ecological Receptors
As discussed in Section 4.2.6. an understanding of absorption, distribution, metabolism and
excretion (ADME) processes can be used to relate exposure concentrations to the concentration, or dose,
of material that reaches the tissues of an organism. Elucidation of organism-specific ADME processes can
help explain observations of high body burdens that were not predicted based on environmental fate and
partitioning alone. ADME processes influence whether and for how long a material is retained in a tissue
(i.e., whether the material will bioaccumulate) and how such retention rates might differ among trophic
levels (i.e., whether concentrations of the material will biomagnify in a food web). Bioaccumulation and
biomagnification have been shown to influence ecological exposures and might similarly influence uptake
of, and exposure of ecological receptors to, MWCNTs.
Ecological receptors are likely to be exposed to MWCNTs through treated products or scraps and
debris from products generated during end-of-life stages of the product life cycle (see Section 2.5).
The materials released during these processes can contain components other than the contaminant of
concern (e.g., textile material, glue, composite ingredients). As discussed at the beginning of Chapter 4
and throughout Section 4.2. studies are lacking on the matrix-bound state of MWCNTs and how exposure
characteristics and dose implications differ for free versus matrix-bound forms. Like the situation with
human exposures discussed in Section 4.2. exposure considerations for ecological receptors are informed
by data on MWCNTs not embedded in a polymer matrix or associated with other product ingredients
(e.g., textile fibers, coating ingredients). No data are currently available regarding leachability or
environmental release of free MWCNTs from their source products in the environment.
G-9
-------
Additional Information Highlight Box G1:
G1. Uptake and Absorption in Aquatic Foodwebs
The bioavailability of MWCNTs in aquatic systems is greatly influenced by the extent of uptake and absorption across
epithelial barriers of aquatic organisms. Uptake and absorption are in turn influenced by the aggregation or dispersal state of
MWCNTs (see Section 3.1). Evidence to date for a variety of aquatic species does not indicate absorption of MWCNTs or
SWCNTs across epithelial membranes (Petersen et al., 2011b). This evidence includes studies using surface-modified
MWCNTs to enhance bioavailability by altering the octanol-water distribution behavior, yet greater bioaccumulation was not
observed (Petersen et al., 2010). Similarly, MWCNTs surface modified with polyethyleneimine to increase their stability in
solution did not result in increased bioaccumulation (Petersen et al., 2011 a). As discussed in Section 6.3, understanding the
likelihood of uptake and absorption across epithelial barriers informs the development of risk assessments and subsequent
risk management decisions for MWCNTs in aquatic environments.
G.4.2.3. Exposure Pathways in Aquatic Systems
Information on ecological uptake pathways for MWCNTs in aquatic environments is limited, but
existing studies indicate that some water-dwelling organisms can take up MWCNTs stabilized in organic
matter via absorption in the gut (Kennedy et al.. 2008). Functionalization also could affect uptake by
aquatic organisms by altering the binding between the MWCNTs and body tissues (Li and Huang. 2011).
G.4.2.3.1. Toxicokinetics and Body Burden in Aquatic Systems
Limited information is available on MWCNT ADME and body burdens in aquatic organisms. As
mentioned in Section 3.3. CNTs are likely to attract lipophilic molecules in aqueous media (Wu et al..
2006). and association of MWCNTs with lipophilic molecules could affect uptake in aquatic ecosystems.
One study exposed a species of water flea (Ceriodaphnia dubid) to MWCNTs stabilized in suspended
natural organic matter and demonstrated that carbon materials can be present in the gut (Kennedy et al..
2008). suggesting that some water-dwelling organisms can take up MWCNTs. Surface functionalization
by lipophilic molecules in the natural environment could further affect uptake by aquatic organisms by
altering the binding between particles and body tissues (Li and Huang. 2011).
A few studies have shown that MWCNTs can be taken up by aquatic invertebrates, but are not
bioaccumulated over time (Petersen et al.. 201 la). For example, Peterson et al. (2010) determined tissue
concentrations of MWCNTs in the freshwater sediment blackworm (Lumbriculus variegatus) exposed to
MWCNTs via soil for 30 days. The authors calculated biota-sediment accumulation factors between 0.1
and 1, indicating that retention of MWCNTs by this species is approximately one-tenth the concentration
in the sediment (Petersen et al.. 2010). Although this suggests that MWCNTs will not continue to build
up in the tissues of some aquatic invertebrates over time, the small concentrations in these species might
be better retained by larger predator species, leading to net accumulation in those species through dietary
sources. Additionally, suggestions have been made that current methods for measuring bioaccumulation
G-10
-------
and calculating bioconcentration factors are not sufficient for nanomaterials (Handy etal.. 2012). These
methods rely on an evenly dispersed aqueous solution of the compound that achieves a steady-state
concentration between external media and biological tissues, which is potentially incompatible with the
dynamic behavior of nanomaterials in environmental media and the challenges associated with dispersion
of MWCNTs in particular. Further, traditional understanding of bioaccumulation assumes that the
processes of uptake and elimination follow we 11-characterized kinetics and diffusive flux models, which
are based on underlying biological mechanisms of solute transporter channels. Pathways of uptake and
elimination for nanomaterials, including MWCNTs, are not well understood, and the degree to which the
bioaccumulation pathways might differ from those of conventional materials is unclear
(Handy etal.. 2012V
Despite the lack of studies directly investigating uptake, absorption efficiency, and
bioaccumulation of MWCNTs in aquatic food webs, the high persistence and hydrophobicity of
MWCNTs are characteristics generally associated with bioaccumulative substances (Petersen et al., 2010;
Helland et al., 2007). Based on these characteristics alone, MWCNTs are expected to accumulate in
aquatic food webs under some conditions. Which additional material, environmental, or biological
characteristics determine whether and to what degree bioaccumulation occurs are unknown
(Handy etal.. 2012).
G.4.2.4. Exposure Pathways in Terrestrial Systems
Limited information is available regarding exposure pathways and ecological uptake of
MWCNTs in terrestrial environments. As also discussed in Section 3.2, limited evidence exists that
airborne MWCNTs can quickly (within approximately two weeks) transform to amorphous carbon (Zhu
et al.. 2011). thus limiting exposures to terrestrial organisms. Other data suggest that MWCNTs might
stabilize in ambient conditions, however, which would serve to increase exposures (Yang et al.. 2009). If
MWCNTs are present in soils, plant roots could interact with those in soil or pore water
(Navarro et al.. 2008).
G.4.2.4.1. Toxicokinetics and Body Burden in Terrestrial Systems
Limited information is available on MWCNT ADME and body burdens in terrestrial organisms.
MWCNTs present in soils could be absorbed or consumed by biota; MWCNTs taken up by plant roots
and plant tissues also could be consumed. Few studies have attempted to measure tissue concentrations of
MWCNTs in biota. One laboratory study was identified that determined tissue concentrations of
MWCNTs in earthworms (Eisenia foetida) exposed to MWCNTs via soil for 30 days. The authors
calculated biomagnification factors between 0.01 and 0.1, indicating that tissue concentrations of
G-11
-------
MWCNTs in this species is approximately l/100th to l/10th the concentration in the sediment (Petersen
et al.. 2010). In another study using 14C-labeled pure MWCNTs and MWCNTs with various
polyethyleneimine surface coatings, Petersen et al. (2011 a) assessed the extent to which modified
MWCNTs concentrate in earthworms. Results indicated that surface coating did not significantly affect
MWCNT uptake or elimination rates over a 28-day period. The bioaccumulation factor remained less
than 0.12 throughout the study regardless of MWCNT type (purified or modified with surface coatings),
indicating that accumulation of MWCNTs from soil by earthworms is low (Petersen et al.. 201 la). As
discussed in Section G.4.2.1. past studies have speculated that the high persistence and hydrophobicity of
MWCNTs are characteristics generally associated with bioaccumulative substances; however, recent
studies have shown that MWCNTs do not behave like other bioaccumulative substances because altering
the octanol-water distribution behavior does not change bioaccumulation factor values (Petersen et al..
2010; Helland et al.. 2007). Also, the complexity of food web interactions that cross aquatic and terrestrial
systems makes determining the source of MWCNTs in terrestrial food webs difficult.
G.5. Potential Human Health, Ecological, and Other Impacts
G.5.1. Ecological Effects
G.5.1.1. Terrestrial Receptors
Compared to other groups of organisms, a large amount of data was identified regarding toxicity
of MWCNT to soil microbes and plants (see Sections G.5.1.1.1 and G.5.1.1.2). No information was
identified for toxicity to terrestrial vertebrates (see Section G.5.1.1.3). but some assumptions can be made
for mammals based on toxicity studies intended for human health purposes presented in Section 5.1.
G-12
-------
Table G-3. Effects of decaBDE and MWCNTs on soil microbes and invertebrates.
Organism
Soil
microbes
Invertebrate
worms
DecaBDE
Effect
Acute NOEL
Chronic
cytotoxicity LCso
Acute NOEL
(Enchytraeus
crypticus)
Chronic NOEL
(Eisenia fetida)
Oxidative stress
(E. fetida)
Effect level
>2,274
mg/kg
(6 months)
1 00 mg/kg
>2,274
mg/kg
>4,910
mg/kg
0.1-10
mg/kg
Citation
Sverdrup et al.
(2006)
Liu et al.
(2011)
Sverdrup et al.
(2006)
ACC (2001) as
cited in
Environment
Canada (2006)
Xie et al.
(2011)
MWCNTs
Effect
Acute NOEL
(Cupriavidus
metallidurans)
Acute LD50
(Escherichia
CO//)
Chronic
cytotoxicity
NOEL; LOAEL
ND
ND
ND
Effect level
>100 mg/L
100 mg/mL
500 ug/gram
5,000
ug/gram
ND
ND
ND
Citation
Simon-
Deckers et al.
(2009)
Simon-
Deckers et al.
(2009)
Chung et al.
(2011)
~
-
Abbreviations: NOEL = No-observed-effect level, LD50 = Median lethal dose, LOAEL = Lowest-observed-effect level, ND = No data
identified
G.5.1.1.1 Soil Microbes and Terrestrial Invertebrates
Changes in soil microbial activity result in changes to nutrient cycling; therefore, studying the
impact of contaminants on soil microbes can provide insight on how those contaminants might affect
ecosystem function (Chung etal.. 2011). Similarly, effects on terrestrial invertebrates, such as worms, can
influence health and fertility of a soil ecosystem (Xie et al.. 2011). Table G-3 presents key toxicity values
identified for the effects of MWCNTs on soil microbes and invertebrates. Parallel data for decaBDE are
provided in the table for comparative purposes; more detailed information on decaBDE can be found in
Appendix H and Appendix F. Table F-18. Appendix F (Table F-19) summarizes details of the MWCNT
studies identified and reviewed for this section.
G-13
-------
Multiple studies have shown that CNTs
exhibit antimicrobial activity, suggesting that
release of MWCNTs into soils might adversely
affect soil microcosms. This possibility, however,
has not yet been investigated outside of controlled
lab experiments (Chung etal.. 2011). A short,
1-hour exposure to low doses of MWCNTs (e.g.,
5 ug/mL [5 ppm]) resulted in mortality rates of
20-50% in Escherichia coll, Pseudomonas
aeruginosa, and Staphylococcus epidermidis cell
cultures, which are 1.5-5 times higher than
background mortality levels (Kang et al.. 2009;
Kang et al., 2008). An increase in exposure level
or duration, however, does not drastically increase
cytotoxicity, and species-specific responses vary.
For example, exposure to 100 mg/mL (100,000
ppm) MWCNTs for 24 hours caused 50-60%
cytotoxicity in E. coll, yet had no effect on
Cupriavidus metallidurans, a more environmentally relevant bacterium (Simon-Deckers et al., 2009). In a
chronic duration study (Chung et al.. 2011). the authors showed that addition of MWCNTs at a high
concentration of 5 mg/gram soil (5,000 ppm) resulted in an average of 34.2-60.5% decrease in microbial
activity over 11 days; lower levels of MWCNTs (500 and 50 ug/gram [ppm]) did not cause significant
cytotoxicity.
G.5.1.1.2 Terrestrial Plants
Table G-4 describes key toxicity values identified for the effects MWCNTs on terrestrial plants.
Parallel data for decaBDE are provided in the table for comparative purposes; more detailed information
on decaBDE can be found in Appendix H and Appendix F, Table F-18. Appendix F (Table F-20)
summarizes details of the MWCNT studies identified and reviewed for this section.
Additional Information Highlight Box G2:
G2. Toxicity to Terrestrial Invertebrates
The impact of CNTs on terrestrial invertebrate reproduction,
development, and survival has been studied in earthworms.
Scott-Fordsmand et al. (2008) found that reproduction (i.e.,
cocoon production) of earthworms (Eisenia veneta) was
affected by double-walled carbon nanotubes (DWCNT)
administered in food at concentrations greater than 37 mg
DWCNT/kg food. The authors found no effect of DWCNTs
on earthworm hatchability or survival at up to 495 mg
DWCNT/kg food. Uptake, bioaccumulation, and
depuration—important considerations for predicting
toxicity—have also been studied in earthworms. Petersen et
al. (2008) assessed uptake and depuration behaviors of
MWCNTs in earthworms, determining bioaccumulation
factors that indicated a lack of both absorption and
equilibrium partitioning to tissues. Furthermore, absorption
in earthworms was limited whether the MWCNTs were
pristine or coated with polyethylene!mine (PEI), with little
apparent difference in uptake among different types of
MWCNTs (Petersen et al., 2011a). Although these studies
show limited uptake or absorption, which hint toward limited
target tissue-level exposure and limited toxicity, no studies
have been identified that were specifically focused on
toxicity to terrestrial invertebrates.
G-14
-------
Table G-4. Effects of decaBDE and MWCNTs on plants.
Endpoint
Germination
NOEL
Germination
LOAEL
Root growth
NOAEL
Root growth
LOAEL
Shoot height
LOAEL
DecaBDE
Organism Effect level Citation
Corn Great Lakes
Chemical
Corporation (2000)
Red >2,274 mg/kg Sverdrup et al.
clover (2006)
ND ND
ND ND
ND ND
Corn penta/tetraBDE Great Lakes
mix: 250 mg/kg Chemical
Corporation (2000)
MWCNTs
Organism
Corn, rapeseed,
radish, ryegrass,
lettuce, cucumber
Brown mustard,
blackgram
Garden cress
Garden cress
Corn, rapeseed,
radish, ryegrass,
lettuce, cucumber
Thale cress
Garden cress
ND
Effect level
>2,000 mg/L
>40 ug/mL
0.01% w/w
0.1% w/w
>2,000 mg/L
>10 mg/L
0.01% w/w
ND
Citation
Lin and Xing
(2QQ7)
Ghodake et al.
(2010)
Oleszczuk et al.
(2011)
Oleszczuk et al.
(2011)
Lin and Xing
(2QQ7)
Lin et al. (2009)
Oleszczuk et al.
(2011)
-
Abbreviations: ND = No data identified, w/w = weight-for-weight measurement, NOEL = No-observed-effect level, NOAEL = No-
observed-adverse-effect level, LOAEL = Lowest observed adverse effect level
Both beneficial and detrimental effects of nanoparticle exposures have been reported for plants.
For example, Khodakovskaya et al. (2011) have demonstrated positive effects on seed germination and
plant growth from MWCNT exposure. Other studies with MWCNTs, however, have implied that
exposure to high levels could have negative effects on seed germination and plant growth, as several
studies show trends and a few show statistically significant impacts. For example, Lin and Xing (2007)
showed that exposure to MWCNTs with diameters 10-20 nm at a concentration of 2,000 mg/L caused no
significant differences in germination rates or root length for six different agriculturally relevant plant
species, although a nonstatistically significant decrease in germination was observed in four of the species
(Lin and Xing. 2007). Conversely, Oleszczuk et al. (2011) found that MWCNTs added to sewage sludge32
Sewage sludge samples, collected from four municipal industrial sewage treatment plants, were analyzed and
reported to contain heavy metals (e.g., Pb, Cr, Cd, Cu, Ni) and polycyclic aromatic hydrocarbons.
G-15
-------
at concentrations of 0.01, 0.1, and 0.5% weight-for-weight significantly inhibited garden cress (Lepidium
sativum) seed germination. The authors observed diameter-dependent responses, as root growth was
inhibited at all three concentrations for the smaller diameter MWCNTs but was not affected at any
concentration for the larger diameter MWCNTs (Oleszczuk et al.. 2011) (see Text Box 5-1). Finally, no
physical injury to cell morphology was observed in thale cress (Arabidopsis thaliana) cell suspensions
exposed to 10 mg/L MWCNTs, but significant loss in cell viability and growth and chlorophyll inhibition
were observed after 7 days of exposure. Cytotoxicity was more severe following exposure to fine, small
bundles than to loose, large bundles of MWCNTs (Lin et al.. 2009). indicating that dispersion state could
play a role in toxicity (see Text Box 5-1 and Appendix F. Table F-20).
A study by Tan and Fugetsu (2007) provides some insight on the mechanism through which
MWCNT exposure affects plant growth and the ecological relevance of the trend described above.
Cultures of rice cells in an embryonic growth stage formed large associations with MWCNT; the cells
that interacted with the MWCNTs experienced high cell death. Only some cells within the culture
associated with the MWCNTs, however, and clumps formed by this initial subset of the cells in the
culture continued to attract other MWCNTs, forming larger associations over the course of the 4-day
exposure period. Cells that did not form these associations with MWCNTs were not adversely affected by
the MWCNTs as exposure continued. The authors stated that their results illustrate how some plants
might be able to tolerate low levels of MWCNTs without major population-level effects due to a self-
defense response (Tan and Fugetsu, 2007).
Ghosh et al. (2011) illustrated clastogenicity in Allium cepa (onion) bulbs exposed to 0, 10, 20,
and 50 ug/mL MWCNTs using traditional cell culture tests. Chromosomal aberrations, DNA cross-
linking, and induction of apoptosis led authors to conclude that MWCNTs might have a significant
impact on genomic activities of plants.
G.5.1.1.3 Terrestrial Vertebrates
The impacts of oral exposure to MWCNTs on mammals are uncertain (see Section 5.1). Testing
performed in mammals for relevance to humans has focused on inhalation exposure routes; acute
inhalation studies have found that MWCNTs or associated contaminants can induce oxidative stress,
pulmonary inflammation, and fibrosis. MWCNTs might cause slight skin and eye irritation. No studies
were identified that specifically investigated the ecological effects of MWCNTs on terrestrial vertebrates.
G-16
-------
Appendix G References
ACC (American Chemistry Council). (2001). Effect of decabromodiphenyl oxide (DBDPO) on the survival and
reproduction of the earthworm, Eisenia fetida. (Study No. 46540). Columbia, MO: ABC Laboratories, Inc.
Borm. PJA: Robbins. D: Haubold. S: Kuhlbusch. T: Fissan. H: Donaldson. K: Schins. R: Stone. V: Kreyling. W:
Lademann. J: Krutmann. J: Warheit DB: Oberdorster. E. (2006). The potential risks of nanomaterials: A
review carried out for ECETOC [Review]. Part Fibre Toxicol 3:1-35. http://dx.doi.org/10.1186/1743-8977-
3-11
Chung. H: Son. Y: Yoon. TK: Kim. S: Kim. W. (2011). The effect of multi-walled carbon nanotubes on soil
microbial activity. Ecotoxicol Environ Saf 74: 569-575. http://dx.doi.0rg/10.1016/i.ecoenv.2011.01.004
Environment Canada. (2006). Canadian Environmental Protection Act, 1999: Ecological screening assessment
report on polybrominated diphenyl ethers (PBDEs).
Ghodake. G: Seo. YD: Park. D: Lee. PS. (2010). Phytotoxicity of carbon nanotubes assessed by Brassica Juncea
and Phaseolus Mungo. Journal of Nanoelectronics and Optoelectronics 5: 157-160.
http://dx.doi.org/10.1166/ino.2010.1084
Ghosh. M: Chakrabortv. A: Bandvopadhyav. M: Mukherjee. A. (2011). Multi-walled carbon nanotubes
(MWCNT): induction of DNA damage in plant and mammalian cells. J Hazard Mater 197: 327-336.
http://dx.doi.0rg/10.1016/i.jhazmat.2011.09.090
Great Lakes Chemical Corporation. (2000). Pentabromodiphenyl oxide (PeBDPO): A toxicity test to determine
the effects of the test substance on seedling emergence of six species of plants. Easton, MD: Wildlife
International, Ltd.
Gustavsson. P: Hedmer. M: Rissler. J. (2011). Carbon nanotubes: Exposure, toxicology and protective measures
in the work environment. Swedish Work Environment Authority.
https://lup.lub.lu.se/luur/download?func=downloadFile&recordOId=2295102&fileOId=2300375
Handy. RD: Cornells. G: Fernandes. T: Tsvusko. O: Decho. A: Sabo-Attwood. T: Metcalfe. C: Steevens. JA:
Klaine. SJ: Koelmans. AA: Home. N. (2012). Ecotoxicity test methods for engineered nanomaterials:
practical experiences and recommendations from the bench [Review]. Environ Toxicol Chem 31: 15-31.
http://dx.doi.org/10.1002/etc.706
Healy. ML: Dahlben. LJ: Isaacs. JA. (2008). Environmental assessment of single-walled carbon nanotube
processes. JIndEcol 12: 376-393. http://dx.doi.org/10.1111/i. 1530-9290.2008.00058.X
Helland. A: Wick. P: Koehler. A: Schmid. K: Som. C. (2007). Reviewing the environmental and human health
knowledge base of carbon nanotubes [Review]. Environ Health Perspect 115:1125-1131.
Hyung. H: Fortner. JD: Hughes. JB: Kim. JH. (2007). Natural organic matter stabilizes carbon nanotubes in the
aqueous phase. Environ Sci Technol 41: 179-184. http://dx.doi.org/10.1021/es061817g
Kang. S: Herzberg. M: Rodrigues. DF: Elimelech. M. (2008). Antibacterial effects of carbon nanotubes: Size
does matter. Langmuir 24: 6409-6413. http://dx.doi.org/10.1021/la800951v
Kang. S: Mauter. MS: Elimelech. M. (2009). Microbial cytotoxicity of carbon-based nanomaterials: Implications
for river water and wastewater effluent. Environ Sci Technol 43: 2648-2653.
http://dx.doi.org/10.1021/es8031506
Kennedy. AJ: Hull MS: Steevens. JA: Dontsova. KM: Chappell MA: Gunter. JC: Weiss. CA. Jr. (2008).
Factors influencing the partitioning and toxicity of nanotubes in the aquatic environment. Environ Toxicol
Chem27: 1932-1941. http://dx.doi.0rg/10.1897/07-624.l
Khodakovskava. MV: de Silva. K: Nedosekin. DA: Dervishj E: Biris. AS: Shashkov. EV: Galanzha. El: Zharov.
VP. (2011). Complex genetic, photothermal, and photoacoustic analysis of nanoparticle-plant interactions.
PNAS 108: 1028-1033. http://dx.doi.org/10.1073/pnas.1008856108
G-17
-------
Koelmans. AA: Nowack. B: Wiesner. MR. (2009). Comparison of manufactured and black carbon nanoparticle
concentrations in aquatic sediments. Environ Pollut 157: 1110-1116.
http://dx.doi.0rg/10.1016/i.envpol.2008.09.006
Li. M: Huang. CP. (2011). The responses of Ceriodaphnia dubia toward multi-walled carbon nanotubes: Effect
of physicalchemical treatment. Carbon 49: 1672-1679. http://dx.doi.0rg/10.1016/i.carbon.2010.12.052
Lin. C: Fugetsu. B: Su. Y: Watari. F. (2009). Studies on toxicity of multi-walled carbon nanotubes on
Arabidopsis T87 suspension cells. J Hazard Mater 170: 578-583.
http://dx.doi.0rg/10.1016/i.jhazmat.2009.05.025
Lin. D: Xing. B. (2007). Phytotoxicity of nanoparticles: inhibition of seed germination and root growth. Environ
Pollut 150: 243-250. http://dx.doi.0rg/10.1016/i.envpol.2007.01.016
Liu. L: Zhu. W: Xiao. L: Yang. L. (2011). Effect of decabromodiphenyl ether (BDE 209) and dibromodiphenyl
ether (BDE 15) on soil microbial activity and bacterial community composition. J Hazard Mater 186: 883-
890. http://dx.doi.0rg/10.1016/i.ihazmat.2010.ll.079
Moisala. A: Nasibulin. AG: Kauppinen. EL (2003). The role of metal nanoparticles in the catalytic production of
single-walled carbon nanotubesa review. J Phys Condens Matter 15: S3011-S3035.
http://dx.doi.org/10.1088/0953-8984/15/42/003
Navarro. E: Baun. A: Behra. R: Hartmann. NB: Filser. J: Miao. AJ: Quigg. A: Santschi. PH: Sigg. L. (2008).
Environmental behavior and ecotoxicity of engineered nanoparticles to algae, plants, and fungi [Review].
Ecotoxicology 17: 372-386. http://dx.doi.org/10.1007/sl0646-008-0214-0
Oleszczuk. P: Josko. I: Xing. B. (2011). The toxicity to plants of the sewage sludges containing multiwalled
carbon nanotubes. J Hazard Mater 186: 436-442. http://dx.doi.0rg/10.1016/i.jhazmat.2010.ll.028
Petersen. EJ: Huang. Q: Weber. WJ. (2008). Bioaccumulation of radio-labeled carbon nanotubes by Eisenia
foetida. Environ Sci Technol 42: 3090-3095. http://dx.doi.org/10.1021/es071366f
Petersen. EJ: Huang. Q: Weber. WJ. (2010). Relevance of octanol-water distribution measurements to the
potential ecological uptake of multi-walled carbon nanotubes. Environ Toxicol Chem 29: 1106-1112.
http://dx.doi.org/10.1002/eto. 149
Petersen. EJ: Pinto. RA: Zhang. L: Huang. Q: Landrum. PF: Weber. WJ. (201 la). Effects of polyethyleneimine-
mediated functionalization of multi-walled carbon nanotubes on earthworm bioaccumulation and sorption by
soils. Environ Sci Technol 45: 3718-3724. http://dx.doi.org/10.1021/esl03004r
Petersen. EJ: Zhang. L: Mattison. NT: O'Carroll. DM: Whelton. AJ: Uddin. N: Nguyen. T: Huang. Q: Henry.
TB: Holbrook. RD: Chen. KL. (201 Ib). Potential release pathways, environmental fate, and ecological risks
of carbon nanotubes [Review]. Environ Sci Technol 45: 9837-9856. http://dx.doi.org/10.1021/es201579y
Scott-Fordsmand. JJ: Krogh. PH: Lead. JR. (2008). Nanomaterials in ecotoxicology. Integr Environ Assess
Manag 4: 126-128. http://dx.doi.org/10.1002/ieam.5630040116
Simon-Deckers. A: Loo. S: Mavne-L'hermite. M: Herlin-Boime. N: Menguy. N: Reynaud. C: Gouget B:
Carriere. M. (2009). Size-, composition- and shape-dependent toxicological impact of metal oxide
nanoparticles and carbon nanotubes toward bacteria. Environ Sci Technol 43: 8423-8429.
http://dx.doi.org/10.1021/es9016975
Sverdrup. LE: Hartnik. T: Mariussen. E: Jensen. J. (2006). Toxicity of three halogenated flame retardants to
nitrifying bacteria, red clover (Trifolium pratense), and a soil invertebrate (Enchytraeus crypticus).
Chemosphere 64: 96-103. http://dx.doi.0rg/10.1016/i.chemosphere.2005.ll.056
Tan. XM: Fugetsu. B. (2007). Multi-walled carbon nanotubes interact with cultured rice cells: Evidence of a
serf-defense response. Journal of Biomedical Nanotechnology 3: 285-288.
http://dx.doi.org/10.1166/ibn.2007.035
Tsai SJ: Hofmann. M: Hallock. M: Ada. E: Kong. J: Ellenbecker. M. (2009). Characterization and evaluation of
nanoparticle release during the synthesis of single-walled and multiwalled carbon nanotubes by chemical
vapor deposition. Environ Sci Technol 43: 6017-6023.
G-18
-------
U.S. EPA (U.S. Environmental Protection Agency). (2005). Furniture flame retardancy partnership:
Environmental profiles of chemical flame-retardant alternatives for low-density polyurethane foam: Volume
1 [EPA Report]. (EPA-742/R-05-002A). http://www.epa.gov/dfe/pubs/flameret/altrep-vl/altrepvl-flc.pdf
U.S. EPA (U.S. Environmental Protection Agency). (2007). Nanotechnology white paper [EPA Report]. (EPA
100/B-07/001). Washington, DC. http://www.epa.gov/osa/pdfs/nanotech/epa-nanotechnology-whitepaper-
0207.pdf
Velzeboer. I: Kupryianchyk. D: Peeters. ETH. M: Koelmans. AA. (2011). Community effects of carbon
nanotubes in aquatic sediments. Environ Int 37: 1126-1130. http://dx.doi.0rg/10.1016/i.envint.2011.01.016
Wiesner. MR: Lowry. GV: Alvarez. P: Dionysiou. D: Biswas. P. (2006). Assessing the risks of manufactured
nanomaterials. Environ Sci Technol 14: 4336-4345. http://dx.doi.org/10.1021/es062726m
Wu. Y: Hudson. JS: Lu. Q: Moore. JM: Mount AS: Rao. AM: Alexov. E: Ke. PC. (2006). Coating single-walled
carbon nanotubes with phospholipids [Letter]. J Phys Chem B 110: 2475-2478.
http://dx.doi.org/10.1021/ipQ57252c
Xie. X: Wu. Y: Zhu. M: Zhang. YK: Wang. X. (2011). Hydroxyl radical generation and oxidative stress in
earthworms (Eisenia fetida) exposed to decabromodiphenyl ether (BDE-209). Ecotoxicology 20: 993-999.
http://dx.doi.org/10.1007/sl0646-011-0645-x
Yang. L: Kim. P: Meyer. HM: Agnihotri. S. (2009). Aging of nanocarbons in ambient conditions: probable
metastability of carbon nanotubes. J Colloid Interface Sci 338: 128-134.
http://dx.doi.0rg/10.1016/i.icis.2009.06.017
Zhang. Q: Huang. JQ: Zhao. MQ: Qian. WZ: Wei. F. (2011). Carbon nanotube mass production: principles and
processes [Review]. ChemSusChem4: 864-889. http://dx.doi.org/10.1002/cssc.201100177
Zhu. SM: Li. JQ: Pang. JH: Zhang. HS: Liang. ER. (2011). Stability of multi-wall carbon nanotubes in air
[Abstract]. Carbon 49: 354-355. http://dx.doi.0rg/10.1016/i.carbon.2010.08.022
G-19
-------
Appendix H. Compilation of CEA
Framework Data for DecaBDE
Chapter 2 through Chapter 5. and Appendix G. present information on multiwalled carbon
nanotubes (MWCNTs) in a comprehensive environmental assessment (CEA) framework. This appendix
contains detailed, parallel information on decabromodiphenyl ether (decaBDE). As noted in Chapter 1,
Chapter 2 through Chapter 5, and Appendix G. also contain text boxes that highlight information about
decaBDE and tables and figures with side-by-side comparisons of decaBDE and MWCNT data. These
elements provide a highlight-level comparison between the two compounds as used in flame-retardant
textiles to illustrate key concepts that might be helpful to risk assessors evaluating MWCNTs.
Supplemental details about decaBDE are provided in this appendix, to provide more in-depth data for
comparison for each CEA framework element. With the exception of Section H.I, the section numbers of
this appendix are parallel to corresponding sections of the MWCNT CEA framework presented in
Chapter 2 through Chapters. Section H.I provides an introduction to decaBDE, which is identical to the
introduction provided in Section 1.3.1 of the main text. It is repeated here to remind readers of the
introductory details regarding physical and chemical properties of decaBDE that set the stage for
understanding the remainder of this appendix.
H.1. Introduction to DecaBDE
DecaBDE is part of a larger group of brominated flame retardants (BFRs) called polybrominated
diphenyl ethers (PBDEs), a group of 209 structurally similar BFRs that differ in the number and location
of bromine atoms (Table G-l) (Rahman etal.. 2001: NRC. 2000). Although PBDEs are typically
categorized into classes by number of bromine atoms (e.g., PBDE with two bromine atoms is a
dibrominated diphenyl ether [diBDE]; ten bromine atoms is a decaBDE), a single class might contain
several different PBDE congeners with the same number of bromine atoms in different locations (i.e.,
PBDE BFRs can have many isomers). As the only fully brominated PBDE, decaBDE is the exception,
existing only as a single congener (BDE-209).
H-1
-------
Table H-1. Major PBDE congeners.
PBDE Class Congeners
DiBDE BDE-7, BDE-8, BDE-11, BDE-12, BDE-13, BDE-15
TriBDE BDE-17, BDE-25, BDE-28, BDE-30, BDE-32, BDE-33, BDE-35, BDE-37
TetraBDE BDE-47, BDE-49, BDE-66, BDE-71, BDE-75, BDE-77
PentaBDE BDE-85, BDE-99, BDE-100, BDE-105, BDE-116, BDE-118, BDE-119, BDE-126, BDE-138, BDE-140
HexaBDE BDE-153, BDE-154, BDE-155, BDE-166
HeptaBDE BDE-181, BDE-183, BDE-190
OctaBDE BDE-196, BDE-197, BDE-203
NonaBDE BDE-206, BDE-207, BDE-208
DecaBDE BDE-209
Source: U.S. EPA (201 Oa).
Commercial formulations of decaBDE (see Table H-2) are generally 97-98% BDE-209 with less
than 3% nonabrominated diphenyl ether (nonaBDE) congeners present as impurities (Rahman et al.,
2001; NRC. 2000) (see Appendix B. Table B-l for analytical techniques used to distinguish PBDE
congeners in samples). Although the terms decaBDE and BDE-209 often are used interchangeably, this
case study primarily uses the term decaBDE to refer generally to the flame-retardant formulation and
BDE-209 to refer to the specific decaBDE congener analyzed in scientific studies.
DecaBDE is the most widely used of the PBDEs and has been well studied. In 2001, decaBDE
use accounted for 83% of total PBDE production worldwide (U.S. EPA. 2010a): an estimated 10-20% of
decaBDE use is in the textile industry (Pure Strategies Inc.. 2005). At the end of 2004, both octa- and
pentaBDE were voluntarily withdrawn from the U.S. marketplace due to evidence of environmental
persistence and toxicity, which left decaBDE as the sole PBDE available for use in commercial products
in the United States (U.S. EPA. 2010a). Several standard physicochemical properties are used to describe
traditional chemicals: melting point, boiling point, molecular weight, and others. Such values are
presented for decaBDE in Table H-3.
H-2
-------
Table H-2. Commercial formulations of PBDEs used as flame retardants.
Name
Penta formulation3
Octa formulation
Deca formulation13
Congener Makeup and Percent Composition
Tetra
Penta
Hexa
Hexa
Hepta
Octa
Nona
Nona
Deca
BDE-47 (25-37%)
BDE-99 (35-50%), BDE-100 (6-10%)
BDE-153 (5-10%), BDE-154 (1-5%)
BDE-153 (5-10%), BDE-154 (1-5%)
BDE-183(40%)
BDE-197 (21%), BDE-203 (5-35%), BDE-196
BDE-208 (10%), BDE-207 (7%)
BDE-206 (2.2%), BDE-207 (0.24%), BDE-208
BDE-209 (>97%)
(8%)
(0.06%)
aTrace amounts of additional congeners might be present in commercial formulations: <0.2% triBDE congeners.
bTrace amounts of additional congeners might be present in commercial formulations: <0.003% heptaBDE congeners; <0.001 %
hexaBDE congeners; <0.002% pentaBDE congeners; <0.00003% tetraBDE congeners; <0.00001% triBDE congeners.
Source: U.S. EPA (201 Oa).
DecaBDE can be applied to textiles by a variety of mechanisms, but this case study focuses on
the application of decaBDE as a back-coating. This application method is used most frequently for
decaBDE (Pure Strategies Inc.. 2005; NRC. 2000) and is most similar to the application method expected
for MWCNTs used in textiles (see Section 1.3.2). The back-coating process usually involves mixing
decaBDE with a copolymer or resin binder to comply with fire safety standards (Pure Strategies Inc..
2005; NRC. 2000). DecaBDE combines the flame-retardant mechanism of most BFRs (releasing
halogens during combustion to compete with the availability of oxygen for the flame) with formation of a
protective char barrier (NRC. 2000) that interferes with the spread of the flame and helps the material to
self-extinguish (Pure Strategies Inc.. 2005).
H-3
-------
Table H-3. Physical properties and chemical identity of decaBDE.
Physical Property/Chemical Identity Reference
CASRN 1163-19-5
NLM (2011)
Synonyms 2,2',3,3',4,41,5,51,6,61-decaBDE; BDE-209; benzene, NLM (2011):
1,1'-oxybis[2,3,4,5,6,-pentabromo]-; decabromodiphenyl oxide; ATSDR (2004)
decabromodiphenyl ether; decabromobiphenyl ether; ether,
bis(pentabromophenyl)
Physical state Solid
Melting point, 300-310 °C
Hardy (2002b)
ECB (2003)
Boiling point Decomposes at >320 °C ECB (2003)
Vapor pressure 4.63 x 10~6 Pa at 21 °C Hardy (2002b)
Henry's law 1.93 x 10~8 L atm/mol Hardy (2002b):
constant 0.04 Pa m3/mol at 25 °C Cetin and Odabasi (2005)
Density 3.0 grams/cm3 NRC (2000)
Water solubility <0.1 ug/L at 25 °C Hardy (2002b);
ECB (2003)
Log Kow 6.3-12.6
Log Koc 6.3
Molecular weight 959.17
Chemical formula Ci2Bri0O
Chemical structure Br
Bry^^
Hardy (2002b)
Hardy (2002b)
NLM (2011); ECB (2003)
NLM (2011)
x°Y\-Br
Br Br
Br
Abbreviations: Kow = Octanol/water partition coefficient, Koc = Soil organic carbon/water partition coefficient.
H.2. Product Life Cycle
A product's life cycle encompasses all stages of its existence from "cradle to grave," starting with
the extraction of raw materials from the earth for the manufacture of the product and continuing
downstream until these materials are returned to the environment following disposal (U.S. EPA. 2006).
The components of the life cycle determine the potential for releases and possible impacts on human
health, ecological populations, and the environment (Som et al.. 2011). which can be evaluated
H-4
-------
systematically within the CEA framework. Potential environmental impacts of a product throughout its
life cycle can be estimated using a life-cycle assessment (LCA) approach, which involves four steps: goal
definition and scope, inventory analysis, impact analysis, and interpretation (U.S. EPA. 2006). The CEA
approach incorporates information from available LCAs in the "product life cycle" and "impacts"
portions of the CEA framework and combines this knowledge with other analyses or qualitative indicators
related to transport, transformation, and fate, exposure-dose, and additional impacts not considered in
available LCAs. As discussed in Chapter 1. if a plausible reason exists to include an impact in the CEA
framework, qualitative or quantitative information on that effect can be included from LCAs or other
sources (if an LCA has not been completed) to evaluate that particular impact.
A generalized depiction of the life cycle for decaBDE and MWCNT coatings used to confer
flame-retardant properties to upholstery textiles is presented in Chapter 2. in Figure 2-1. That figure
illustrates the life cycle of these materials as five main stages: (1) acquisition and processing of
feedstocks; (2) manufacturing, including research and development (R&D) processes; (3) storage and
distribution; (4) use; and (5) end-of-life processes (including disposal, reuse, and recycling). These stages
correspond roughly to the four primary life-cycle stages the U.S. Environmental Protection Agency
(EPA) (U.S. EPA. 2006) outlines: (1) raw materials acquisition, (2) manufacturing,
(3) use/reuse/maintenance (with storage and distribution discussed as a distinct stage in this case study),
and (4) recycle/waste management. As mentioned in Chapter 1, R&D is included in the product life-cycle
portion of the CEA framework, given its importance regarding emerging materials such as MWCNTs. For
such materials, R&D efforts can elucidate potential risks associated with commercial-scale
manufacturing. In fact, because it often takes place when health and safety information is being developed
for a material, R&D presents an ideal opportunity to gather data on a product's potential impacts and to
make design adjustments if appropriate. Similarly, as discussed below, differences between R&D
activities and the commercial manufacturing process (e.g., use of protective equipment, volume of
material produced) could be important considerations in mitigating potential risks to individuals involved
in R&D versus commercial manufacturing.
To conduct a comparative CEA, relevant information on life-cycle inventories from existing
LCAs would be incorporated into the product life cycle to characterize the inputs (e.g., raw materials,
energy) and outputs (e.g., emissions to air and water, co-products) associated with each material's
manufacture. Impacts information from existing LCAs also would be considered (see Section H.5). Other
LCA aspects also might apply, including using an appropriate functional unit, which is a quantitative
measure of a product's function or a process that facilitates comparison (U.S. EPA. 2006). In the current
case study, a functional unit might correspond to the degree of flame retardancy conveyed by
incorporation of a certain amount of decaBDE. In general, for this case study, data that specify
H-5
-------
appropriate functional units were not identified; the reader might, however, consider how this aspect of
existing or future LCAs could be incorporated into a future CEA when evaluating data gaps and needs.
This section outlines important aspects of each of the five life-cycle stages outlined in Figure 2-1
(in Chapter 2) for decaBDE used in upholstery textiles. This section also includes descriptions of the
important environmental release scenarios for decaBDE and MWCNTs across the product life-cycle
stages based on current knowledge. A variety of release scenarios are possible throughout the life-cycle
stages described in this appendix. Figure 2-1 (in Chapter 2) also outlines potential release scenarios for
decaBDE flame-retardant upholstery textile coatings throughout the life cycle along with potential forms
of the released substances (i.e., free, bundled, or matrix bound). The term "free decaBDE" refers to pure,
unbound materials. The term "matrix-bound decaBDE" refers to materials that are part of a polymer
matrix (e.g., the flame-retardant formulation).
H.2.1. Feedstocks
H.2.1.1. Life-Cycle Processes
The raw materials used in commercial synthesis of decaBDE are phenol, bromine, and a catalyst
(e.g., aluminum bromide or iron) (IPCS. 1994). Phenol is produced from cumene, which is obtained
primarily from the distillation or other processing of petroleum products (Mahapatra. 2010). Commercial
production of bromine involves the drying of brine, typically obtained from sea water (Kesner. 2005).
No data were found on the energy and resource demands of raw material extraction for synthesis of
decaBDE.
H.2.1.2. Potential Releases during the Feedstock Extraction Stage
Release of decaBDE would not occur during this initial phase of the life cycle given that its
synthesis does not occur until the stage that follows extraction. Release of hazardous raw materials,
however, could occur during the extraction and processing of feedstock materials. Release of decaBDE
also could occur if reactors are not cleaned between runs. Specific materials that could be released include
petroleum-based chemicals. No data were identified quantifying the potential volume of releases during
feedstock extraction for decaBDE.
H-6
-------
H.2.2. Manufacturing
The manufacturing stage for decaBDE flame-retardant upholstery can be viewed as a sequential
process involving synthesis of decaBDE, material processing, and product manufacture (i.e., formulation
of the flame-retardant mixture, application of the flame-retardant mixture to textiles, and incorporation of
the flame-retardant textile into consumer or commercial goods). Discussion of R&D also is included in
this section, given the similarities to key aspects of synthesis, processing, and manufacture.
H.2.2.1. Research and Development
H.2.2.1.1 Life-Cycle Processes
Research on decaBDE and on flame-retardant coatings involving decaBDE is principally
conducted in specialized laboratory environments. R&D activities are expected to be carried out by
individuals rather than automated mechanisms used in commercial-scale manufacture. The processes of
interest to researchers are similar to those used in commercial-scale manufacture of these materials:
synthesis, purification, modification, dispersion, incorporation into flame-retardant formulations, and
application to textiles. Substantially less R&D related to decaBDE flame retardants is expected to occur at
the present time compared to MWCNT flame retardants, given that decaBDE flame-retardant
technologies are more mature and the use of decaBDE is decreasing or being phased out due to health and
ecological concerns. The following sections (material synthesis, material processing, and product
manufacturing) provide detailed information on the processes of potential interest for R&D.
H.2.2.1.2 Potential Releases during the R&D Stage
Release scenarios during the R&D stage are expected to be similar to release scenarios from
commercial synthesis described in the following sections, but the quantities released are anticipated to be
much smaller in the R&D stage. The quantities of decaBDE handled in research laboratories are much
smaller than those handled in commercial-scale manufacturing facilities. Although R&D activities are
typically carried out in laboratories with specialized pollution control systems in place, including fume
hoods, ventilation systems, and environmental control systems, not all facilities have standardized
engineering controls. For example, these practices might not be in place for small start-up operations.
Given the experimental and somewhat unpredictable nature of R&D, releases from handling of materials
during synthesis, processing and purification, storage, and analysis are possible.
No information was found in the literature that describes release of decaBDE from R&D
facilities. No data were found that describe how releases in academic labs compare with releases in
commercial R&D labs.
H-7
-------
H.2.2.2. Material Synthesis
H.2.2.2.1 Life-Cycle Processes
Commercial synthesis of decaBDE involves conversion of phenol to diphenyl ether via the
Williamson ether synthesis [Kirk Othmer (2005) as cited in Wright et al. (2008)]. Diphenyl ether is then
brominated in the presence of a catalyst (generally, aluminum bromide or iron) to produce commercial
decaBDE (EU. 2002; IPCS. 1994). Commercially, decaBDE is synthesized in a batch process in enclosed
vessels during both the reaction and the subsequent drying process (IPCS. 1994). DecaBDE powder is
collected in bags during the recovery phase following the synthesis process (EU. 2002). Commercial
formulations of decaBDE typically contain decaBDE, 97-98% weight-for-weight measurement, and other
PBDEs (primarily nonaBDE), 0.3-3.0% weight-for-weight measurement (IPCS. 1994). No information
was found on by-products of decaBDE synthesis.
H.2.2.2.2 Potential Releases during the Material Synthesis Stage
Synthesis of decaBDE could result in releases to air or water (U.S. EPA. 2005a). Fugitive
releases of decaBDE vapor from a reactor vessel have been estimated as 1.1 * 10"5 mg/ton, and release
from the bagging of synthesized PBDEs have been estimated as <70 grams/ton PBDE produced [(EU.
2002); EEC (1993) as cited in EU (2002)]. Airborne releases of decaBDE particles likely would sorb to
dust (see Section 3.2). but loose dust likely would be vacuumed and the area would be washed with water,
reducing airborne particles (EU. 2002). The main source of water release of decaBDE during the
synthesis stage would be due to cleaning of equipment and floors after synthesis. One study found,
however, that wastewater releases of decaBDE are unlikely to exceed 0.5 kg/ton if equipment is washed
after every batch (EU. 2002). Releases directly to skin could occur through handling of bags containing
solid decaBDE (U.S. EPA. 2005a). Large manufacturing facilities, however, likely would have exhaust
ventilation in place to minimize air release into the general environment. Engineering controls that
regulate temperature and pressure to minimize the potential for release also would likely be in place (U.S.
EPA. 2005a). As a result, air and water releases of decaBDE to the environment during the synthesis
stage are not expected to be large.
Accidental releases through fugitive equipment leaks, malfunctioning ventilation systems, and
exposure to fire or high heat could occur at all stages of manufacturing (material synthesis, material
processing, and product manufacture). A fugitive equipment leak or ventilation malfunction could lead to
the airborne releases of decaBDE, as could exposure to fire or high heat. If exposed to high heat,
decaBDE can form polybrominated dibenzofurans (PBDFs), polychlorinated dibenzo-p-dioxins, and
nonhalogenated substances such as polycyclic aromatic compounds, which could be released into the
H-8
-------
environment (EU. 2002). Such accidental events could result in larger releases of decaBDE to the
environment than normal release scenarios due to the lack of control mechanisms compared to those in
place to mitigate anticipated releases. Table 2-3 in Chapter 2 summarizes the anticipated potential release
scenarios from the material synthesis stage of decaBDE and MWCNTs.
H.2.2.3. Material Processing
H.2.2.3.1 Life-Cycle Processes
Material processing includes any modification of decaBDE after synthesis and before
incorporation into a flame-retardant formulation. These modifications can include purification,
functionalization, and dispersal in solvents. After synthesis, decaBDE does not require further processing
before incorporation into the flame-retardant formulation.
H.2.2.3.2 Potential Releases from the Material Processing Stage
No release scenarios for decaBDE are summarized here because the activities specified for this
stage are not anticipated to occur for decaBDE.
H.2.2.4. Product Manufacturing
In this section, product manufacturing for decaBDE is described. This life-cycle stage is
considered to include the manufacture of flame-retardant formulations, the manufacture of textiles
containing decaBDE-based flame retardants, and the manufacture of end-use products containing flame-
retardant materials, such as furniture.
H.2.2.4.1 Life-Cycle Processes
In a typical decaBDE flame-retardant formulation, decaBDE and antimony trioxide (a synergist
used to enhance the activity of decaBDE) are first mixed as a dispersion in water (EU. 2002). This mix is
stored in tanks and then piped directly into a closed vessel (EU. 2002). The decaBDE-antimony trioxide-
water dispersion is added to emulsion polymers and mixed in this closed vessel to formulate the flame
retardant (EU. 2002). One analysis found trace amounts of polybrominated dibenzo-p-dioxins (PBDDs)
and PBDFs as impurities in commercial mixtures of decaBDE (Ren etal.. 2011). No data were found that
describe other characteristics of decaBDE flame-retardant formulations or the by-products of their
manufacture.
During application, the decaBDE flame-retardant formulation is typically back-coated, or applied
as a resin to the reverse surface of textiles along with a binding agent, such as latex or a copolymer (see
H-9
-------
Section 1.2.2.2). Due to the high efficiency of decaBDE flame retardants, they can be used in
formulations with low loadings compared to other brominated flame retardants (Pure Strategies Inc.,
2005). Typical loadings of decaBDE in textiles range from 30 to 40% by dry weight of the dry coating
with different loadings applied to different types of fabrics (30-40 grams/m3 in cotton to 70-80 grams/m3
for velour fabrics) (EU. 2002). After application, the decaBDE flame-retardant upholstery textile is cut,
shaped, and glued or stapled to furniture.
H.2.2.4.2 Potential Releases during Product Manufacture
Release scenarios for product manufacturing are likely to be similar to those in the material
synthesis and processing stages, but release amounts are probably lower (U.S. EPA. 2005a; EU, 2002).
Additionally, releases from this stage likely will not be decaBDE, but rather decaBDE in a polymer
matrix. As discussed, decaBDE generally is synthesized as a powder and then mixed into solution when
the flame retardant is formulated, minimizing releases of decaBDE to dust (U.S. EPA. 2005a).
Nevertheless, manufacture of decaBDE flame-retardant coatings could release vapors if mixing and
handling of raw decaBDE occurs in an open system (U.S. EPA. 2005a). One study found that
environmental release was most likely to occur during the mixing of decaBDE powder and cleaning
operations of the flame-retardant formulation stage (EU. 2002). Formulation of flame retardants,
however, generally occurs in closed systems with engineering controls that regulate temperature and
pressure to minimize potential releases (U.S. EPA. 2005a: EU. 2002). Releases of decaBDE in this stage
also can contain the impurities listed in Section H.2.2.4.1.
The application of decaBDE flame-retardant coatings to upholstery textiles could result in the
release of aerosolized decaBDE due to thermal processing, but release would occur only if the
manufacturing plant does not have engineering controls in place to prevent such releases. Cutting, sewing,
shaping, stapling, and other textile finishing processes could result in the airborne release office
decaBDE or decaBDE in a polymer matrix through abrasion. Equipment cleaning also could lead to the
release of decaBDE in wastewater during the processing stages of product manufacture.
The accidental release scenarios for decaBDE during product manufacture are similar to those in
the material synthesis stage (see Section H.2.2.2.2). Additionally, in this stage, spills could lead to release
of decaBDE flame-retardant formulations in wastewater. Volatilization is unlikely due to the low
volatility of decaBDE (see Table H-3). Table 2-6 in Chapter 2 outlines potential release scenarios from
the product manufacturing stage of decaBDE and MWCNT flame-retardant textiles.
H-10
-------
H.2.3. Storage and Distribution
H.2.3.1. Life-Cycle Processes
The storage and distribution stage concerns the handling and transport of (1) decaBDE,
(2) decaBDE flame-retardant formulations, and (3) decaBDE flame-retardant upholstery textiles.
The main method of transport for these materials is not known, but likely would be by truck, train, or
cargo ship.
DecaBDE, if not immediately incorporated into a flame-retardant formulation, likely would be
stored at the site of synthesis/processing in sealed receptacles until it is incorporated into flame-retardant
formulations or transported to sites where manufacture of the flame retardant occurs.
DecaBDE flame-retardant formulations are typically stored at manufacturing plants in drums,
tanks, or more permanent storage vessels until they are packaged and sent to textile manufacturers (U.S.
EPA. 2005a).
DecaBDE flame-retardant upholstery textiles and end-use products likely also would be stored
at the site of manufacture (or an intermediate storage site) and then transported to retail locations.
H.2.3.2. Potential Releases during Storage and Distribution
Storage and distribution of (1) decaBDE, (2) decaBDE flame-retardant formulations, and
(3) flame-retardant textiles could result in releases to the environment, but all release scenarios are
unlikely.
Releases of decaBDE separate from flame-retardant formulations are likely to be negligible.
Release would be due primarily to accidents, as the materials would be stored in sealed receptacles after
synthesis. Exposure of the receptacles to high heat or fire could lead to the airborne release of decaBDE
(see Section H.2.2.2.2 for more details).
Releases from flame-retardant formulations could result in releases of decaBDE to the
environment (U.S. EPA. 2005a). The possible scenarios for release of decaBDE flame-retardant
formulations during storage and distribution include damage to containers holding the flame-retardant
formulation, leakage resulting from mishandling of containers, or faulty or improper stacking of cartons
in transport vehicles. If the containers are sealed properly and not damaged during transport, releases of
product prior to application might be limited to spills.
Releases from flame-retardant upholstery textiles could result from accidental exposure to
high heat or fire, off-gassing of volatile components, and infestation with pests. Exposure of the flame-
retardant upholstery textiles to high heat or fire during storage and distribution could lead to the
H-11
-------
degradation of the polymer matrix and subsequent airborne release of decaBDE (both free and matrix
bound) (see Sections H.2.2.2.2 and H.2.4 for more details). Off-gassing of the volatile components of
decaBDE flame retardants also could occur in poorly ventilated areas that experience high temperatures
(e.g., storage units, warehouses). DecaBDE itself, however, is not highly volatile (see Table 1-8 and Table
1-9. both in Chapter 1; and Table 3-1. in Chapter 3). Infestation of textile or furniture storage facilities
with rodents or other pests also could lead to the release of decaBDE flame-retardant materials to the
environment.
Table 2-7 in Chapter 2 outlines potential release scenarios from the storage and distribution stage
of (1) decaBDE and MWCNTs, (2) decaBDE and MWCNT flame-retardant formulations, and
(3) decaBDE and MWCNT flame-retardant upholstery textiles.
H.2.4. Use
H.2.4.1. Life-Cycle Processes
A wide variety of textiles contain flame-retardant coatings (see Section 1.2). Upholstery textiles
are expected to be used in public places where people of all ages will sit, lie, or walk on them. Some
unintended uses of upholstery textiles include outdoor use, repurposing for use in other products, burning
as kindling, or mouthing by children. Repurposing for use in other products and burning as kindling are
covered in Section H.2.5. In general, upholstery textiles are likely to have a lifespan of at least 10 years
(EU. 2002).
H.2.4.2. Potential Releases during the Use Stage
Environmental releases from upholstery textiles coated with flame retardants are expected due to
(1) the potential use scenarios for the upholstery textiles and (2) the physicochemical properties of
decaBDE. The anticipated long lifespan of upholstery textiles (>10 years) suggests that releases in this
stage could occur over several years (EU. 2002). Indeed, environmental concentrations of decaBDE in
buildings with products containing decaBDE can be high, especially in dust (see Sections H.4.1.2.5 and
H.4.1.2.6). One of the most important pathways for these high environmental concentrations is the
airborne release of decaBDE sorbed to dust in the environment (see Section H.3.2). The following
characteristics of flame-retardant upholstery textiles, however, are expected to reduce releases of
decaBDE (EU. 2002):
• Flame-retardant coatings must meet durability requirements to comply with regulations
(see Section 1.2.1):
H-12
-------
• Flame retardant often is applied to the back of the fabric, minimizing wear and tear; and
• Upholstery textiles are unlikely to be washed frequently.
The integrity of the flame-retardant coating depends on the strength of the formulation that bonds
it to the textile surface (Som et al.. 2011; NRC. 2000). The decaBDE flame-retardant textile coatings
considered in this case study are additive, suggesting that release from upholstery textiles could occur
during the use stage. DecaBDE/antimony trioxide flame-retardant formulations are considered relatively
durable, however, due to the copolymer resin that bonds to the textile fibers (Pure Strategies Inc.. 2005).
Even if migration of decaBDE through the polymer were to occur, it would be expected to be very slow
due to the high molecular weight of decaBDE (Lassen etal. 1999). In a substance flow analysis of
plastics containing decaBDE, however, Lassen et al. (1999) found that release of decaBDE was expected
to be greatest during the use stage.
Regular use of upholstered furniture (e.g., sitting, walking, lying) could abrade the textile surface
and release small amounts of free or matrix-bound decaBDE either into the air or onto the skin of users.
Washing of textiles also could lead to water release of matrix-bound decaBDE. By some estimates, the
principal source of decaBDE release in wastewater is due to textile washing (EU. 2002). Most flame-
retardant upholstery textiles will be used indoors, minimizing exposure to UV light and weathering.
Upholstery textiles that are back-coated with decaBDE flame retardant likely will not be subject to
significant abrasion, washing, or UV light. Additionally, most releases of decaBDE initially will be to the
indoor environment, but they could spread outdoors through environmental transport mechanisms (see
Section H.3) (U.S. EPA. 2010a: Lassen etal. 1999). These processes could result in airborne release of
decaBDE or releases in wastewater (U.S. EPA. 2010a).
Unintended uses also could lead to the release of decaBDE from flame-retardant textiles. Use of
flame-retardant upholstery textiles outdoors could lead to weathering, which could degrade the polymer
matrix and release decaBDE. Mouthing by small children, pets, or rodents on flame-retardant textiles
could lead to the release of decaBDE directly into children's, pets', or rodents' mouths if the back-coating
is exposed and the integrity of the fabric is compromised. Accidental contact of flame-retardant textiles
with fire and high heat also could occur and could lead to airborne releases of decaBDE (see Section
H.2.2.2.21 for more details). No data were found, however, that describe the likelihood of these releases
from this application. Table 2-8 in Chapter 2 outlines potential release scenarios from the use stage of
decaBDE and MWCNT flame-retardant textiles.
H-13
-------
H.2.5. Reuse, Recycling, and End of Life
The reuse, recycling, and end-of-life stage encompasses a variety of different transformation and
disposal processes for (1) decaBDE, (2) decaBDE flame-retardant formulations, and (3) decaBDE flame-
retardant upholstery textiles. What the primary reuse, recycling, and end-of-life treatments are for
decaBDE and decaBDE flame-retardant formulations are unclear.
H.2.5.1. Reuse and Recycling
H.2.5.1.1 Life-Cycle Processes
Reuse or recycling of decaBDE or decaBDE flame-retardant formulations is unlikely. On the
other hand, textile waste often is recovered and reused or recycled (Kohleretal.. 2008); upholstered
furniture is sometimes reused, but is rarely recycled (CalRecycle. 2002). Upholstery could be donated to
charitable organizations and resold for residential use. Additionally, upholstery textiles could be
informally repurposed into clothing, blankets, and other textile products. Due to the difficulty of recycling
furniture and flame-retardant materials, flame-retardant furniture is typically land-filled (CalRecycle.
2002; Lassen etal.. 1999). Of the small portion of upholstered furniture that is recycled, about 60% of the
material is recycled and 25-30% is composted (CalRecycle. 2002). No data were found that describe the
proportion of other upholstery textiles (e.g., mattress ticking or curtains) that are typically recycled.
The main types of textile recycling processes are fiber-to-fiber recycling and polymer reduction
recycling. During the fiber-to-fiber process, textiles are shredded and blended with other fibers to create a
new mixture ready for spinning (Kohler et al.. 2008). During the polymer reduction process, textiles are
cut and granulated to form pellets that are processed to break down the polymer to the molecular level to
be reused as raw material (Kohler et al.. 2008). No data were found that describe the prevalence of each
recycling process.
H.2.5.1.2 Potential Releases during the Reuse/Recycling Stage
Release of decaBDE beyond releases described in the use stage is unlikely to occur during reuse
of flame-retardant upholstery textiles. Older textiles could release greater levels of decaBDE, however,
due to increased degradation of the material. Informal repurposing of flame-retardant textiles likely would
require cutting and shredding, resulting in possible air release of decaBDE. Airborne releases of decaBDE
could occur during recycling of flame-retardant textiles. Recycling subjects textiles to a variety of
mechanical, thermal, and chemical treatments that could result in the airborne releases of additive flame
retardants from fibers (Kohler et al.. 2008). One analysis found airborne releases of decaBDE at a plastic
recycling plant with the highest concentrations of airborne particles measured near the shredder (Sjodin et
H-14
-------
al.. 2001). Although releases from recycling of upholstery textiles containing decaBDE flame retardant
might be similar to those of plastics, the processing of plastics is likely to differ from that of textiles.
Release of decaBDE to water also could occur during chemical treatment and processing. Although
release of decaBDE is possible during recycling of flame-retardant textiles, no data were found that
indicate the likelihood of release from recycling processes.
Table 2-9 in Chapter 2 outlines potential release scenarios from the reuse/recycling stage of
decaBDE and MWCNT flame-retardant textiles.
H.2.5.2. Incineration
H.2.5.2.1 Life-Cycle Processes
The incineration of decaBDE or decaBDE flame-retardant formulations is unlikely, but any
incineration likely would occur in a hazardous waste incinerator. Upholstery textiles treated with
decaBDE flame-retardant coatings might be sent to municipal incinerators for processing. Municipal
incinerators generally provide a well-controlled environment with pollution control mechanisms and
sufficiently high temperatures (850 °C) to destroy most materials (Kohler et al.. 2008). Processing in
municipal facilities is likely to result in complete incineration of the upholstery textiles. Alternatively,
upholstery textiles also might be incinerated in less well-controlled facilities or burned in open fires as a
rudimentary form of waste management or as kindling. These incineration methods are likely to result in
incomplete incineration of the upholstery textiles. No data were found that describe the prevalence of
incineration as a form of disposal for upholstery textiles or what proportion of incinerated textiles is
processed at well-controlled incineration facilities.
H.2.5.2.2 Potential Releases during the Incineration Stage
Airborne releases of decaBDE from we 11-controlled incineration are expected to be negligible,
but incomplete incineration (e.g., open fires) could lead to some airborne release. Little empirical data
exist that describe the prevalence of decaBDE in incinerator residues, but decaBDE is expected to be
destroyed by the high-temperature incineration used at most municipal incineration facilities (Palm et al..
2002; Lassen et al.. 1999). These temperatures also are sufficiently high to prevent the formation of
PBDFs and PBDDs during incineration of decaBDE (see Section H.2.2.2.2). In one study of atmospheric
concentrations of PBDEs near solid-waste incinerators, the authors found that incineration facilities do
not give rise to a substantial proportion of atmospheric releases of decaBDE (Agrell et al.. 2004).
Incomplete incineration, however, could lead to the airborne release of decaBDE and the formation of
PBDFs, PBDDs, polychlorinated dibenzo-p-dioxins, and nonhalogenated substances such as polycyclic
H-15
-------
aromatic compounds (see Section H.2.2.2.2). Current pollution control technologies for municipal
incinerators are expected to effectively filter these emissions and prevent their release to the environment
(EU. 2002).
Due to the high temperatures and pollution control mechanisms at municipal incinerators,
decaBDE in flame-retardant textiles are expected to be destroyed during well-controlled incineration.
Incineration by open flame in uncontrolled environments, however, might lead to airborne releases of
decaBDE and harmful by-products.
Table 2-10 in Chapter 2 outlines potential release scenarios from the incineration stage of
decaBDE and MWCNT flame-retardant textiles.
H.2.5.3. Land-Filling
H.2.5.3.1 General Processes
Land-filling of decaBDE or decaBDE flame-retardant formulations is unlikely, except in the case
of floor sweepings from manufacturing facilities. Upholstered furniture and textiles generally are
disposed of in municipal landfills (Kohler et al.. 2008). Remaining parts from recycled furniture, such as
cover cloth materials, also are sent to the landfill (CalRecycle. 2002). Additionally, some textiles might
be disposed of in uncontrolled landfills or open dumping sites that have no pollution control mechanisms
in place. No data were found that describe the proportion of upholstery textiles disposed of in landfills or
any further processing that might occur at the landfill.
H.2.5.3.2 Potential Releases during the Land-filling Stage
Land-filling of decaBDE flame-retardant textiles could lead to water and air releases due to
mechanical processes such as mixing and compacting. DecaBDE also could leach from land-filled textiles
and migrate into the underlying soil or ground water (Rahman et al.. 2001; Lassenetal.. 1999). however,
no evidence of decaBDE in land-fill leachate has been found [Kim et al. (2006) as cited in Wright et al.
(2008)1. Additionally, this release scenario is unlikely due to the low leaching potential of decaBDE (see
Table 1-8). Flame retardants containing decaBDE could volatilize to the atmosphere overtime (Rahman
et al.. 2001). but volatilization of decaBDE is expected to be negligible due to low volatility of decaBDE
(Palm et al.. 2002) (see Table 3-1 in Chapter 3).
Few data were identified that measure releases of decaBDE from land-filling of flame-retardant
textiles, but the physicochemical characteristics of these materials suggest that such releases likely would
be small. Table 2-11 in Chapter 2 outlines potential release scenarios from the land-filling stage of
decaBDE and MWCNT flame-retardant textiles.
H-16
-------
H.2.5.4. Wastewater Treatment Plants
H.2.5.4.1 Life-Cycle Processes
The wastewater treatment process consists of filtering and treating wastewater to remove solids
and contaminants. Large facilities that manufacture decaBDE and decaBDE flame retardants might divert
their wastewater to an on-site wastewater treatment plant. Alternatively, some wastewater from these
facilities might be directly processed by municipal wastewater treatment plants. Water releases of
decaBDE that occur during the storage and distribution, use, and reuse/recycling/end-of-life stages also
would be treated in municipal wastewater treatment plants.
H.2.5.4.2 Potential Releases during the Wastewater Treatment Stage
Release of decaBDE or decaBDE flame-retardant formulations into wastewater could occur
throughout the life cycle. Primary releases to wastewater during manufacturing stages are due to
equipment cleaning, formulation and application of the flame retardant, and accidental spills. Washing
processes (which can involve abrasion, detergents, and water), particularly in the product manufacturing
stages, are likely to result in the release of additive flame retardants from textiles to wastewater (Som et
al., 2011). Due to the physicochemical characteristics of decaBDE flame retardants (see Table 1-8 and
Table 1-9 in Section 1.3), the material is likely to sorb onto particles during water treatment and be
removed in sludge (Som et al., 2011; Lassen et al., 1999). The potential nonetheless exists for releases
from filter backwash and other wastewater treatment plant equipment (EU. 2002). Additionally, some of
this removed sludge is deposited in landfills or spread on agricultural soil (EU. 2002; Lassen et al.. 1999).
This activity represents one of the most significant potential releases to soil of decaBDE flame-retardant
coatings (Ciparis and Hale. 2005; Lassen et al.. 1999). The releases of decaBDE from wastewater
treatment facilities are expected to be small. The removal efficiency of wastewater treatment plants is not
well characterized for decaBDE and the spread of sewage sludge onto agricultural soil could represent a
significant source of decaBDE to soil. See Section H.3.3.3 for information regarding decaBDE removal
efficiency of wastewater treatment plants.
Table 2-11 in Chapter 2 outlines potential release scenarios from the wastewater treatment stage
of decaBDE and MWCNT flame-retardant textiles.
H-17
-------
H.3. Transport, Transformation, and Fate
Releases throughout the product life cycles of upholstery textile coatings containing decaBDE
flame retardant will, to some extent, lead to occurrence of primary and secondary contaminants in air,
soil, and aquatic media. Chapter 3 examines what might happen to these substances after their release to
the environment, including transport or transformation through chemical, physical, and biological
processes. Studies investigating the transport, transformation, and fate of decaBDE in the environment are
summarized in Appendix D. and concentrations of BDE-209, the single isomer of deca-substituted BDE,
in environmental compartments are provided in Appendix E.
DecaBDE can be released into the environment during the manufacturing, storage, distribution,
use, disposal, reuse, and recycling of upholstery textiles treated with flame retardants (see Chapter 2).
DecaBDE flame-retardant formulations are used primarily as additives that are mixed with, not
chemically bound to, polymers in textile products. Because they are not chemically bound, these
substances can escape from the material and become a source of contamination to surrounding
environmental media (Yu et al.. 2010; Vonderheide et al.. 2008; Moniruzzaman and Winey. 2006; Song
et al.. 2006; Soderstrom et al.. 2004). Although some, if not most, releases after the production stage are
likely to be in the matrix-bound form, little information exists that describes the environmental behavior
of decaBDE-polymer complexes. As a result, this section focuses on the transport, transformation, and
fate of decaBDE not embedded in a polymer matrix.
Section H.3.1 provides a brief discussion of the chemical and physical characteristics and the
processes that influence behavior (e.g., mobility, persistence, bioavailability) of decaBDE in
environmental media. The sections that follow summarize the available information regarding the
behavior of each substance in indoor and outdoor air (Section H.3.2). aquatic systems (Section H.3.3).
and terrestrial systems (Section H.3.4). A brief discussion of models that might be used for evaluating the
fate and transport of these substances in environmental media is provided in Section H.3.5.
H.3.1. Physicochemical Factors Influencing Transport,
Transformation, and Fate
The environmental fate of BDE-209 will be dictated by its chemical and physical properties and
its propensity for biotic and abiotic transformation. BDE-209 could transform physically, chemically, or
biologically once released to the environment, leading to substances that present a very different hazard
than the hazard of the untransformed material originally released. BDE-209 has been shown to
biologically and photolytically debrominate (lose a bromine atom) to form lower brominated congeners
H-18
-------
that are more readily bioavailable [U.S. EPA (2010a); Vonderheide et al. (2008): Song et al. (2006):
Watanabe and Sakai (2003): Darnerud et al. (2001): see Text Box H.3-1 and additional sources in
Appendix D. Table D-11. Because the chemical properties associated with transformation products of
decaBDE influence their transport, transformation, and fate in the environment, degradation processes of
decaBDE are introduced in this section. A summary of key physicochemical factors that are likely to
affect partitioning33 and fate of BDE-209 and related PBDEs in the environment is presented in Table
H-4. Values for key physicochemical properties of BDE-209 are provided in Section H.I (see Table H-3).
Biotic debromination is the breakdown of BDE-209 into lower brominated compounds by aerobic
and anaerobic microorganisms. Biotic transformation processes for BDE-209 that occur in soil, sediment,
or sewage sludge have been described in recent literature (see Appendix D. Table D-l). These processes
result in dehalogenation through microbe catalysis reactions that stimulate the replacement of a halogen
atom (e.g., bromine, chlorine, fluorine) with a hydrogen atom (Kuivikko et al.. 2010: Kim et al.. 2007).
Photolysis or photodegradation is a chemical (abiotic) process by which molecules are broken
down through the absorption of light. PBDEs are vulnerable to photolysis, which induces reductive
debromination causing higher brominated congeners like BDE-209 to photodegrade to form lower
brominated congeners. These lower brominated congeners are potentially more stable and bioavailable in
the environment due to lower molecular weight and a lower octanol/water partition coefficient (Kow)
(Soderstrom et al.. 2004). PBDFs also have been identified as photolysis products of BDE-209. Sunlight
could degrade BDE-209 in air, surficial soils, water, and surficial sediments via photolysis (Christiansson
et al.. 2009: Soderstrom et al.. 2004). This and other abiotic transformation processes for BDE-209 have
been demonstrated in recent literature (see Appendix D, Table D-l).
Transformation also can occur with elevated temperatures (e.g., incineration, fire); thermal
breakdown products of PBDEs include polybrominated, poly chlorinated, and mixed
brominated/chlorinated dibenzo-p-dioxins and dibenzofurans, and are similar to polychlorinated dibenzo-
p-dioxins/polychlorinated dibenzofurans in their persistence and toxicity (Watanabe and Sakai. 2003:
Darnerud et al.. 2001: Rahman et al.. 2001).
DecaBDE formulations used in textile and other products contain the fully brominated congener, which is
less mobile in the environment than lower brominated congeners, probably due to low volatility, water
solubility, and bioaccumulation, and the high propensity to adsorb to sediments. The lower brominated
compounds are generally more volatile, water soluble, and bioaccumulative compared with higher
33Partitioning refers to the potential for a chemical or other substance to move from one environmental medium to
another (e.g., air, water, sediment) and the tendency to accumulate in one particular medium over another (U.S.
EPA. 20Wa).
H-19
-------
brominated compounds (Watanabe and Sakai. 2003) and are believed to be structurally analogous to
polychlorinated biphenyls (PCBs), so their chemical properties, persistence, and behavior in the
environment, are expected to follow similar patterns. PBDEs are expected to be more vulnerable to
environmental degradation than PCBs, however, because their carbon-bromine bonds are weaker than the
carbon-chlorine bonds of PCBs (Shih and Wang. 2009; Watanabe and Sakai. 2003: Rahman et al.. 2001).
Text Box H.3-1. BDE-209 Undergoes Biotic and Abiotic Debromination
Higher brominated PBDE congeners like BDE-209 (the principal constituent in decaBDE) have lower bioaccumulation
potential, water solubility, and volatility, and therefore have been considered relatively safe (Watanabe and Sakai, 2003).
Lower brominated congeners (including degradation products of BDE-209), however, are predicted to be more volatile,
water soluble, and bioaccumulative than the higher brominated congeners, and these degradation products are therefore
likely to be more bioavailable in the environment than BDE-209. Soils and sediments are major sinks for higher brominated
compounds, and other pathways are relatively minor (see the illustration below; the thickness of the arrows is an indication
of the strength of the pathway). The potential significance of other pathways increases for the lower brominated congeners.
Understanding the fate and potential
toxicity of BDE-209 requires
understanding the various degra-
dation processes that dictate its
persistence in the environment.
Debromination of BDE-209 in the
environment affects potential trans-
port and accumulation through the
food web (Huang et al., 2010) and
therefore potential for human
exposure through the diet. Several
biotic and abiotic processes of
BDE-209 degradation have been
demonstrated in air, water, soil, and
sediments (Vonderheide et al.,
bioaccumulation
adsorption to
sediment and
particulate matter
adsorption to
sediment and
particulate matter
Although higher brominated congeners primarily adsorb to solids, lower
brominated congeners can readily volatilize to air, dissolve in water and
interstitial spaces in soil, bioaccumulate in biota, and adsorb to solids.
As a result, the fate of the parent compound BDE-209 is expected to differ
substantially from that of its lower brominated transformation products.
Adapted from Watanabe and Sakai (2003).
Photolytic degradation by
solar rays and UV light is a
significant abiotic process of BDE
degradation and has been studied
more than other processes. Other
abiotic processes include geo-
chemical degradation by metal
oxides found in certain soils and
degradation by engineered nanoparticles (Vonderheide et al., 2008). Biotic degradation pathways include debromination by
both aerobic and anaerobic microbes found in soil (Wangetal., 2011: Huang etal., 2010: Lee and He, 2010) and possible
further degradation in plants (Huang etal., 2010). These processes proceed in a stepwise fashion, removing one bromine
atom at a time, and can therefore contribute to the levels of lower brominated congeners from nona- to tetra- and even
diBDEs in the environment. BDE-47 (tetraBDE) and BDE-99 (pentaBDE) are the congeners most often detected in humans
and biota globally (Tokarz et al., 2008: Vonderheide et al., 2008), and debromination of BDE-209 might be an
environmental source of these congeners, which are known to be toxic and face restrictions on their production (Ross et
al., 2009). In accordance with the comprehensive environmental assessment approach, some consideration of the
exposure, uptake, and effects of lower brominated congeners produced during the natural debromination of BDE-209 is
appropriate. Several studies have described the debromination of BDE-209 (often in the laboratory) and identified ranges of
the lower brominated congeners produced; however, the specific congener profiles that are expected to occur in natural
environmental systems following BDE-209 debromination have not been clearly identified.
H-20
-------
Table H-4. Summary of physicochemical properties that affect partitioning and fate of BDE-209.
Physicochemical
property
How does this property affect chemical partitioning and fate?
High molecular weight
• Limits bioavailability to biota
Sources: Kierkegaard et al. (2004): Huang et al. (2010)
Low water solubility;
hydrophobic
• Expected to associate primarily with organic (carbon-rich) particles in soil, sediment,
sewage sludge; percent total organic carbon likely plays a major role in transport and
distribution of BDE-209 in these media
• Less mobility, strongly sorbed to solid particles such as soils, sediments, and sewage
sludge
• Expected to adsorb to particles by van der Waals forces and hydrophobic attraction
(Ann etal.. 2006)
• Reduced bioavailability to some aquatic biota
Sources: Vonderheide et al. (2008): Qui et al. (2010): U.S. EPA (2010a): Hua et al.
(2003): Mikula and Svobodova (2006): Rahman et al. (2001): Zhu et al. (2010): Yu et al.
(2010)
Low volatility; low vapor
pressure
> Vapor pressures of PBDEs are inversely related to both molecular weight (i.e., as
molecular weight increases, vapor pressure decreases) and degree of bromination
• Partitioning for BDE-209 is expected to be 1% vapor phase and 99% associated with
airborne particles; as the degree of bromination increases, likelihood increases for
BDE congeners to partition to the particle phase in air
• Lower brominated PBDEs could be expected to be present primarily in the vapor
phase and be more susceptible to long-range transport; however, BDE-209 could
sorb to aerosol particles and can experience long-range transport
Sources: U.S. EPA (201 Oa): Watanabe and Sakai (2003)
High octanol/water • PBDEs partition between water and sediment based on solubility and Kow
partition coefficient (Kow) . Less mobj|jty] Str0ngly sorbed to soils, sediments, sludge
• Not easily distributed within surface water and ground water
Sources: U.S. EPA (201 Oa): Vonderheide et al. (2008)
Low Henry's law constant
(KH)
High octanol/air partition
coefficients (Koa)
• PBDEs partition between water and air based on KH; lower KH at higher degrees of
bromination; BDE-209 not expected to readily volatilize to air
Source: U.S. EPA (201 Oa)
• Stronger propensity to adsorb to the organic matter in soils and vegetation; indicator
of chemical mobility in the atmosphere; tendency for atmospheric BDE-209 to deposit
on forest canopies and other vegetative biomass
Sources: U.S. EPA (201 Oa): Wania et al. (2002): Vonderheide et al. (2008)
High lipophilicity • Dissolves in fats, oils, lipids, and nonpolar solvents (e.g., hexane, toluene)
Sources: Mikula and Svobodova (2006): Rahman et al. (2001): Vonderheide et al.
(2008)
• Greater tendency to bioaccumulate and biomagnify in the food chain
Sources: Rahman et al. (2001): Vonderheide et al. (2008)
Abbreviations: PBDE = polybrominated diphenyl ether; BDE-209 = single isomer of decabrominated diphenyl ether
H-21
-------
H.3.2. Transport, Transformation, and Fate in Air
BDE-209 released from the flame-retardant upholstery textile coatings life cycle could reach
indoor and outdoor air in several ways. For example:
• BDE-209 can be released directly into ambient air during all stages of the product life cycle,
as previously described in Section H.2. BDE-209 that remains in the particle phase can
disperse through air away from the source of release.
• Particulate BDE-209 can become suspended in the surrounding indoor or outdoor air during
multiple stages of the product life cycle.
• Particulate BDE-209 might remain suspended and be transported through the atmosphere or
deposited onto surfaces. Particles that have been deposited on surfaces could become
resuspended in the air and redeposited elsewhere.
If released indoors, BDE-209 can distribute indoors to air, dust, vacuums, and air filter systems. If
the source of BDE-209 is inside a building, levels of BDE-209 in indoor air can be much higher than in
outdoor air (Hale et al.. 2006). Indoor air sources also could contribute to outdoor air concentrations.
Variability in the indoor air and dust levels of BDE-209 can be influenced by indoor/outdoor exchange
rates, building ventilation rates, and the number and age of all PBDE-treated products (e.g., electronics,
mattresses, draperies, furniture) present in the building (Hazrati and Harrad. 2006).
As mentioned earlier, BDE-209 has very low vapor pressure and a high octanol/water partition
coefficient and is therefore more likely to be transported on particles in the air than as a vapor (see Table
H-4) (Breivik et al.. 2006). Approximately 99% of BDE-209 in ambient air is expected to be present in
the particle phase (U.S. EPA. 2010a). Particles could remain suspended in air or deposit on surfaces with
the potential for resuspension. Because nonvolatile compounds like BDE-209 tend to sorb to particles,
they likely would be concentrated in household dust and could experience an extended indoor lifetime
(Kemmlein et al.. 2003).
When attached to particles, BDE-209 can have an extended residence time and persistence in the
atmosphere. Longer residence time in the atmosphere allows more time for the particles to be mobilized
by wind and other forces and makes long-range atmospheric transport (LRT) in the atmosphere possible
(Gouin et al.. 2006; Wania and Dugani. 2003). Evidence exists for LRT of PBDEs to remote ecosystems,
including the Arctic (de Witetal.. 2010; Su et al.. 2009; Agrell et al.. 2004); some literature suggests that
PBDEs in remote Arctic regions originated in urban areas in North America (Breivik et al.. 2006).
LRT has been considered by some to be the reason for PBDE occurrence in rural and more
remote sites. Some evidence has shown LRT of BDE-209 adsorbed to airborne particulate matter at
regional and global scales (Vonderheide et al.. 2008). although other researchers have suggested that the
H-22
-------
potential for LRT of BDE-209 is low. Atmospheric deposition of BDE-209 and other PBDEs is thought
to be a main source of these contaminants in background waters and soils (Vonderheide et al., 2008).
Understanding of BDE-209 LRT behavior is limited. As stated previously, BDE-209 in the air is
primarily attached to particles (U.S. EPA. 2010a: Su et al. 2009V Efforts by Mueller and Nowack (2008)
to model atmospheric transport of engineered nanoparticles can be used to infer that materials, such as
BDE-209, adsorbed to particles will eventually deposit or wash out (wet deposition) in aquatic or
terrestrial systems. In turn, the fate of BDE-209 in air is likely dictated by the characteristics of the
particles to which it adsorbs.
Plants exposed to BDE-209 from air can accumulate BDE-209 and act as transfer vectors in the
food chain. Airborne BDE-209-laden particles could attach to leaves and other aboveground parts of
plants and translocate to different tissues of the plant. Salamova and Kites (2010) evaluated PBDE levels
in air samples and tree bark and determined that BDE-209 concentrations in tree bark were strongly
correlated with concentrations in the air and in precipitation. The highest air and tree bark concentrations
occurred at urban sites.
Breakdown and transformation of BDE-209 in the air by photolysis also can occur (see Section
H.3.1 and Text Box H.3-1). and studies that evaluated this phenomenon in air are presented in Appendix
D, Table D-l. Temperature variability could be an important factor that explains seasonal patterns of
BDE-209 burdens in air because changes in temperature affect gas-solid partitioning coefficients and
subsequently the transfer and retention of BDE-209 in air (Vonderheide et al., 2008).
Section H.4.1.2 and Section E.I of Appendix E provide summaries of studies with BDE-209
concentration data in building dust and indoor/ambient air.
H.3.3. Transport, Transformation, and Fate in Water and Sediment
BDE-209 released from the flame-retardant upholstery textile coating life cycle could enter
aquatic systems in several ways. For example:
• BDE-209 in ambient air subsequently could be deposited or washed out to aquatic systems.
• Erosion of contaminated soil could release BDE-209 to surface waters.
• Runoff flowing along the ground surface could transfer BDE-209 in contaminated soil to
nearby waterways.
• Wastewater effluents containing BDE-209 could be a source of contamination to receiving
water bodies near the discharge location.
• BDE-209 could leach from land-filled sewage sludge into subsoil and ground water and
migrate to surface water or sediment.
H-23
-------
H.3.3.1. Surface Water and Sediment (Inland and Coastal)
Water solubility and Kow are important physicochemical factors for predicting behavior of
BDE-209 in the aqueous phase—and these parameters predict that BDE-209 will partition to the
particulate phase (e.g., sorb to suspended organic matter) in water or bind strongly to sediments (U.S.
EPA. 2010a; Hale et al., 2006; Watanabe and Sakai. 2003). The lower brominated congeners are more
water soluble than the higher brominated congeners and are expected to be more mobile in water
fWatanabe and Sakai. 2003).
Sediment is both a sink and a reservoir for PBDEs such as BDE-209. In general, BDE-209 is the
dominant congener in sediment samples (Tokarz et al.. 2008); notably higher concentrations in urban and
industrial areas and near outfalls of wastewater treatment plants have been measured (U.S. EPA. 2010a)
(see Appendix E). Whether the higher concentration of BDE-209 in sediment samples is due to greater
BDE-209 use or less environmental degradation compared with other congeners is unknown.
The physicochemical properties of BDE-209 and the characteristics of sediment both affect the
bioavailability of BDE-209 in sediment. The bioavailability of BDE-209 in sediment is expected to be
limited due to its strong hydrophobicity and large molecular size (Liu et al.. 20lib). Due to its tendency
to sink to sediments, BDE-209 might be bioavailable to benthic organisms, but generally not to water-
column-dwelling organisms. Benthic organisms might, therefore, act as vectors for the transport of
decaBDE through the food web (see Section H.4.3).
The amount of organic matter in the sediment is an important factor controlling the partitioning of
BDE-209 in sediments (Liuetal.. 20lib). In a study conducted by Liu et al. (20lib), the authors
concluded that desorption of BDE-209 in sediment was more difficult as contaminant-sorbent interaction
time increased, likely because of entrapment of BDE-209 molecules in the micropores of organic matter.
Total desorption also decreased with increased total organic carbon in the sediments.
Physicochemical properties of PBDEs and the characteristics of sediment might not serve as
perfect predictors of levels in surface waters due to microbial or photolytic degradation processes that
could reduce concentrations of the material. Elevated quantities of hydroxylated PBDEs (a possible
oxidation product of PBDEs) have been measured in surface waters near sewage treatment plants
(Vonderheide et al.. 2008). Anaerobic microbial reductive debromination is potentially a driving
transformation process in sediment (see Section H.3.1 and Appendix D. Table D-l): the process,
however, can be very slow.
Appendix E provides a summary of studies with BDE-209 concentration data in surface water
and sediment. Most identified studies focus on sediment concentrations of BDE-209.
H-24
-------
H.3.3.2. Ground Water
BDE-209 present in soil could leach into subsoil and ground water. Based on its physicochemical
properties, BDE-209 does not dissolve in water, but could sorb to suspended organic matter in the ground
water plume and migrate to surface water (U.S. EPA. 2010a). No data were found on concentrations of
BDE-209 in ground water.
H.3.3.3. Wastewater
As introduced in Section H.2. manufacturers, homes, and public buildings could be significant
sources of BDE-209 to wastewater treatment plants and municipal sewage treatment facilities (Hale et al..
2006). Because of its hydrophobicity and tendency to partition to solids, most BDE-209 in wastewater
would be expected to sorb to settling solids (i.e., will not remain in effluent) during the sewage treatment
process (Ricklund et al.. 2009: North. 2004).
BDE-209 partitions strongly to particulate matter, so sewage sludge is expected to be a major sink
for BDE-209, although effluent is likely dominated by the more water soluble lower brominated
congeners. Sewage sludge applied to agricultural fields could be a source of BDE-209 to soils (Huang et
al..201Q: Vrkoslavovaetal.. 2010: U.S. EPA. 2009: Vonderheide et al.. 2008: Knoth et al.. 2007: Hale et
al.. 2006: Law et al.. 2006b: Sellstrom et al.. 2005). Runoff along the surface of the ground then could
transfer BDE-209 in the sewage sludge to nearby terrestrial systems or waterways. Sludge contaminated
with BDE-209 could be disposed of in landfills (and possibly leach to subsoils and ground water) or be
incinerated (introducing emissions to the atmosphere).
Wastewater effluents, although not dominated by the higher brominated PBDEs, might be a
source of BDE-209 and transformation product contamination to receiving water bodies of local aquatic
ecosystems near the discharge location (Peng et al., 2009: Song et al.. 2006). Wastewater irrigation for
farmlands could be a source of PBDEs in agricultural soils. The mobility of BDE-209 in the receiving soil
could be enhanced if it sorbs to dissolved organic matter in the irrigation water (Wang et al.. 2010b).
Recent studies that present concentrations of BDE-209 in wastewater effluent and sludge are
presented in Section H.4.1.2.2 and Section E.I of Appendix E.
H-25
-------
H.3.4. Transport, Transformation, and Fate in Soil
BDE-209 released from the flame-retardant upholstery textile coatings life cycle could enter
terrestrial ecosystems in several ways:
• BDE-209 in ambient air subsequently could be deposited on soil and plants. Some particles
that deposit on soil or plants might experience secondary transport via wind and become
resuspended into ambient air and redeposited into nearby terrestrial ecosystems.
• Runoff flowing along the ground surface could transfer BDE-209 in contaminated soil to
nearby terrestrial ecosystems.
• Disposal products containing BDE-209 could be deposited in solid waste landfills. Sewage
sludge containing BDE-209 also might be land-filled. BDE-209 in land-filled waste could
leach into subsoils.
Similar to sediments, soils are a major sink for PBDEs, including BDE-209. BDE-209 released
from the flame-retardant upholstery textile coating life cycle can enter terrestrial ecosystems and
distribute in soil and plants.
As with sediment, water solubility (hydrophobicity) and Kow are important physicochemical
factors for predicting behavior of BDE-209 in soil, and these parameters predict that BDE-209 will
associate primarily with organic (carbon-rich) particles in soil and experience limited mobility (U.S. EPA.
2010a: Yu et al.. 2010; Zhu et al.. 2010). In addition, percent total organic carbon likely plays a major
role in BDE-209 transport and distribution in soil (Zou et al.. 2007).
Plants exposed to BDE-209 from soil also can accumulate BDE-209 and potentially transform it
to lower brominated compounds. Debromination in this context has been investigated only in the soil-
plant system, however, not in plants alone; the possibility therefore remains that debromination might
occur exclusively in the soil, after which plants take up the transformation products (Huang etal. 2010).
If present in soils, BDE-209 could contact plant roots and partition to root lipids due to its high
lipophilicity and thereby transport into plant tissues (Huang et al.. 2010; Vrkoslavova et al.. 2010). Huang
et al. (2010) studied the uptake, translocation, and metabolism of BDE-209 in six plant species and found
that root lipid content was positively correlated with BDE-209 uptake in those species. BDE-209
accumulated in the roots and shoots of all plants. The translocation factor, measured as the
Concentrationshoot/Concentrationroot of BDE-209, however, was inversely related to BDE-209
concentration in the roots, suggesting root lipids restrict translocation of BDE-209 from roots to shoots
because of its partitioning to root lipids. In contrast to Huang et al. (2010). the study conducted by
Vrkoslavova et al. (2010) provided evidence of translocation of BDE-209 and other PBDEs in plants from
the root lipids of tobacco plants to other plant tissues. Concentrations in soil will further depend on the
amount of vegetative cover, which also could scavenge BDE-209 (U.S. EPA. 2010a).
H-26
-------
Evidence of BDE-209 debromination (transformation) by soil microorganisms and photolysis in
surface soils was presented earlier (see Section H.3.1 and Appendix D, Table D-l). Other studies relevant
to the fate and transport of BDE-209 in soil and plants are provided in Appendix D. Table D-3. Evidence
that BDE-209 affects soil microbial community structure and function is also available; some of these
studies are also discussed in Section H.5.2.2.1 and Section F.2.3 of Appendix F.
Recent studies that present soil concentrations of BDE-209 have been identified; some of these
data are presented in H.4.1.2.3 and Section E.I of Appendix E.
H.3.5. Multimedia Models to Predict Environmental Fate and Transport
Multimedia models to predict environmental fate and transport of BDE-209 and PBDE congener
profiles in environmental media have been used in recent studies. Breivik et al. (2006) used a multimedia
fate and transport model to provide further understanding of how temporal variability and forest cover
help control LRT distance for BDE-209 and determined that the fate of atmospheric BDE-209 is likely to
be controlled by deposition. Results of their study suggested that the variability of precipitation and the
occurrence of periods without precipitation and with strong winds can impact LRT of BDE-209. They
determined that the "forest filter effect" might also be important to the LRT of BDE-209; scenarios with a
forest yielded lower estimates of air travel distance than scenarios without a forest. Gouin et al. (2005)
used a multimedia mass-balance model to assess the importance of seasonal variability (including snow
pack, temperature, forest canopy) on concentrations of PBDEs in air, although results were not specific to
BDE-209. The authors concluded that PBDE concentrations experience a "spring-pulse" due to particle-
bound deposition of PBDEs in the snow pack during winter, followed by transfer of PBDEs to the soil
surface following snow melt and volatilization back into the atmosphere as springtime temperatures
increase. Emerging spring foliage then takes up PBDEs, decreasing atmospheric concentrations during the
summer months and inhibiting LRT.
Bogdal et al. (2010) used PBDE measurement data for the Lake Thun catchment area,
Switzerland, in air, lake water, lake sediment, and tributary water, and combined results from a
multimedia fate model that used site-specific environmental parameters from the lake catchment to
predict PBDE congener patterns in water and sediment. They incorporated measured loadings of PBDEs
in air, tributaries, and wastewater into their model. The authors reported that their model successfully
predicted general PBDE congener patterns in water and sediment, but that the model tended to
underestimate concentrations in water and overestimate concentrations in sediment relative to measured
concentrations. The parameter driving this discrepancy appeared to be partitioning of PBDEs between the
aqueous dissolved phase and suspended particulate matter phase in the water column.
H-27
-------
H.4. Exposure-Dose
Releases of decaBDE to the indoor and outdoor environments can occur at multiple stages of the
product life cycle for flame-retardant upholstery textile coating (see Section H.2). and subsequent
transport, transformation, and fate processes dictate how decaBDE distributes through various
environmental media once released (see Section H.3). Exposure describes the pathways through which
contact occurs between contaminants in the environment and living organisms and abiotic receptors.
Toxicokinetics (i.e., ADME) describes the processes that relate exposure (or dosage) to the internal dose,
which refers to the quantity of a chemical or material that is taken up and absorbed by living organisms
(U.S. EPA. 2010b).34
Section H.4.1 introduces analytical techniques for identifying, characterizing, and measuring
decaBDE in various matrices. The various metrics recommended for characterizing exposure and dose of
decaBDE are also discussed, and available concentration data in various indoor and outdoor media are
presented. In the absence of data quantifying decaBDE exposures at the point of contact, measured
concentrations of decaBDE in surrounding media can be used to estimate exposures using a scenario
evaluation approach. Sections H.4.2 and H.4.3 expand on the release scenarios presented in Section H.2
to discuss the potential human and ecological exposure pathways that link those releases to receptors.
No data were identified regarding relevant exposure pathways leading to impacts on abiotic resources
(e.g., the manmade environment); as a result, this CEA case study does not include a discussion of
exposure scenarios that would influence abiotic receptors. Although broad potential impacts on society
and the global environment are discussed in Section H.5.3. exposure is either not considered germane to
the discussion of the impact (such as for economic impacts of manufacturing) or the exposure
characteristics related to the impact are already included in the general discussion that follows (higher
potential exposures levels related to such as for socioeconomic impacts).
34The term "dose" is described generally by the EPA Integrated Risk Information System (IRIS) as "[t]he amount of
a substance available for interactions with metabolic processes or biologically significant receptors after crossing the
outer boundary of an organism." Several specific forms of dose are also described by IRIS, but the definitions of
these terms are not used consistently across the risk assessment community. The following definitions of specific
forms of dose are provided by IRIS: "The POTENTIAL DOSE is the amount ingested, inhaled, or applied to the
skin. The APPLIED DOSE is the amount presented to an absorption barrier and available for absorption (although
not necessarily having yet crossed the outer boundary of the organism). The ABSORBED DOSE is the amount
crossing a specific absorption barrier (e.g., The exchange boundaries of the skin, lung, and digestive tract) through
uptake processes. INTERNAL DOSE is a more general term denoting the amount absorbed without respect to
specific absorption barriers or exchange boundaries. The amount of the chemical available for interaction by any
particular organ or cell is termed the DELIVERED or BIOLOGICALLY EFFECTIVE DOSE for that organ or cell."
H-28
-------
Probable exposure scenarios throughout the flame-retardant upholstery textile coating life cycle
are identified for workers, consumers, the general public, and highly exposed populations in
Section H.4.2 and for aquatic and terrestrial biota in Section H.4.3. These scenarios describe the
conditions under which exposures might occur; this information can be used in combination with
measured or modeled concentrations in environmental media from Section H.4.1 and exposure factors to
estimate exposures. Kinetic information then can be used to determine or estimate the internal dose that
results from external exposures. When available, point-of-contact measurements, administered dosages,
tissue or body burdens, and scenario-specific exposure guidelines and recommendations are provided, and
the toxicokinetics of decaBDE are described. Finally, Section H.4.4 discusses aggregate exposures to
decaBDE from multiple sources and Section HAS discusses cumulative exposures to multiple related
stressors. Measured concentrations of decaBDE in environmental media are provided in Appendix E. and
studies describing toxicokinetics of PBDEs in mammals are summarized in Appendix F.
As described in Section H.2.2.4. decaBDE is expected to be incorporated into a polymer or other
type of matrix in the flame-retardant formulation applied to upholstery textiles, and both the free and
matrix-bound form might be released during the product life cycle. Very little data relevant to BDE-209
exposures, however, have been generated for the matrix-bound form. This lack of data necessitates a
reliance on the existing data for free BDE-209 in the discussion throughout this section. The extent to
which exposure characteristics and dose implications differ between the free and matrix-bound forms of
BDE-209, however, is unknown at this time.
H.4.1. Detection, Measurement, and Characterization
Exposure scenario evaluation requires information on measured, modeled, or reasonably
estimated concentrations of a stressor in exposure media. As introduced in Section H.I. PBDEs represent
a group of compounds encompassing substances that span a range of physicochemical characteristics and
properties. As a result, developing reliable analytical techniques for detecting, measuring, and
characterizing the full range and makeup of PBDEs in environmental media can present challenges. Text
BoxH.4-1 provides an abbreviated discussion of a few common analytical techniques and the general
challenges associated with them. Appendix B summarizes common analytical techniques for each
substance, and presents the strengths and limitations of each technique.
The diversity of known PBDE congeners makes the characterization of exposure to a single
congener such as decaBDE difficult, as described in Text BoxH.4-1. Identifying the specific source of
decaBDE in environmental samples is even more difficult. DecaBDE is present in dust in many different
environments, but the source of the dust could be from manufacturing, wear, or breakdown of many
H-29
-------
consumer products, including electronics, foams, and polymer textile coatings on furniture such as
couches and mattresses. Some research has been conducted using forensic microscopy to characterize the
source and transfer mechanisms of BDE-209 in indoor environments, but the results are qualitative
(Lagalante et al.. 2011; Webster et al.. 2009). PBDEs are persistent pollutants, so quantifying whether
exposure and body burdens of the lower brominated PBDEs are the result of weathering and breakdown
of BDE-209, or from other PBDEs such as pentaBDE and octaBDE [widely used until recently
(Stapleton. 2006)1. is difficult.
H-30
-------
Text Box H.4-1. Detecting, Measuring, and Characterizing PBDEs
Mass spectrometry is used most frequently to detect polybrominated diphenyl ethers (PBDEs), and gas chromatography
(GC) is employed to measure and characterize PBDE mixtures. High-resolution mass spectrometry is the most selective
method, but is also expensive and labor intensive. Low-resolution mass spectrometry is insensitive to congeners with more
than six bromines (like BDE-209), but optimization of aspects like
reagent gas, source temperature, and source pressure can
increase sensitivity.
Accurate characterization in environmental samples has proven
more problematic for BDE-209 than for lesser brominated
congeners. How a GC system is set up can contribute to
variation in measurements. For example, use of higher
temperatures and longer column lengths can lead to degradation
of higher brominated congeners and quicker evaporation of
higher molecular weight congeners.
3000-
2000-
1000-
0 -
s 8 s
I
£ 5
HI til *
I
. i .
\
^-_
i i
3
I
1
"V i
I
!
PBDESTD
3
I
Su-.
%
1
f
Distinguishing between certain PCBs and PBDEs is difficult with
most systems, which can lead to imprecise measurements.
Additionally, not all PBDE congeners are commercially available
for use as standards (approximately 160 of the 209 congeners
are currently available). Determining whether unidentified peaks
in chromatograms (see gas chromatograph output to right) are Example Gas Chromatograph Output for a PBDE Mixture with
due to these nonstandardized PBDE congeners or to other
compounds can be challenging (Stapleton, 2006).
Retention Time (min)
Multiple Unidentified Peaks (Stapleton. 2006)
H.4.1.1. Dose and Exposure Metrics
Environmental concentrations of decaBDE are commonly quantified in terms of mass (e.g.,
picograms, nanograms) per volume (e.g., m3, L) for air and water measurements or mass per weight (e.g.,
grams dry weight, grams wet weight) of soil, sediment, or tissue samples (Frederiksen et al.. 2009).
Human exposure to decaBDE and subsequent dose has been quantified by detection of BDE-209 in the
serum, breast milk, adipose tissue, and hair of humans (Darnerud et al.. 2001).
H.4.1.2. Concentrations in Environmental Media and Indoor Environments
As described in the previous section, exposures can be estimated by combining knowledge of
concentrations in exposure media with assumptions about contact of humans, biota, or abiotic surfaces
with those media. The following sections describe the information available on concentrations of
BDE-209 and related substances in environmental media (i.e., air, water, soil).
H-31
-------
H.4.1.2.1. Outdoor Air
Research has shown that BDE-209 comprises between 6 and 31% of total PBDE concentrations
in outdoor air (Frederiksen et al.. 2009). Outdoor air generally has lower concentrations of total PBDEs,
which can be one or two orders of magnitude lower than in indoor air (see Appendix E). This lower
concentration is partly due to the dilution factor of outdoor environments, but also occurs because indoor
environments contain more sources of PBDEs such as electronics, furniture, plastics, and coatings in
enclosed spaces (Paso et al.. 2010).
Mean levels of BDE-209 measured in outdoor air from locations throughout the United States
range from 1.4 to 60.1 pg/m3 (U.S. EPA. 2010a). and BDE-209 has been detected in the particulate phase
in air near point sources (ATSDR, 2004). Sampling locations included rural, agricultural, and urban
locations. The highest level detected was 65 pg/m3 in urban Chicago, and the lowest level was 0.2 pg/m3
at an agricultural site in Indiana (Hoh et al., 2005). International studies have reported mean BDE-209
levels ranging from 1.6 to 53.3 pg/m3 (Chang et al., 2009). The lowest environmental concentration
reported in the literature was 0.091 pg/m3 in the Canadian High Arctic (Su et al., 2007). and the highest
level reported was 105 pg/m3 in Southern Ontario, Canada, nearly all of which was sorbed to aerosol
particles (Gouin et al.. 2006). Information from additional studies of concentrations in air is available in
Appendix E. Table E-2.
H.4.1.2.2. Aquatic Systems - Sediment and Surface Water
PBDEs are hydrophobic and therefore are not detected in large concentrations in the water
column (ATSDR, 2004). In aquatic systems (including both water and sediment), BDE-209 is the
predominant PBDE congener detected (49% to nearly 97% of the total PBDEs), with most detected in
sediment (ATSDR, 2004). BDE-209 has been observed in both sediments and surface waters in multiple
locations in the United States, from California to Delaware and Wisconsin to Mississippi (Ashley et al..
2006; Raff and Hites. 2004). Surface water in the San Francisco Bay estuary of California was found to
contain BDE-209 at concentrations below the limit of detection (reported as 20-200 pg/L for individual
congeners) to 191.0 pg/L water, with all but one of the detectable concentrations ranging between 12.2
and 87.8 pg/L (Pros et al.. 2005). Levels of BDE-209 observed in sediment ranged from below the level
of detection (reported as 0.1 to 1.5 ng/gram dry weight) to 3,150,000 ng/gram in sediment downstream of
a wastewater treatment plant for a plastics manufacturer in North Carolina (La Guardia et al.. 2007). In
another study, the highest total PBDE concentration in suspended sediment from the Mississippi River
and five tributaries was 1,548 ng/gram, with BDE-209 accounting for an average of 96.8% of the total
observed PBDE concentration at the 31 sampling sites (Raff and Hites. 2004). Appendix E. Table E-3
provides additional information on these studies.
H-32
-------
As detailed above, detected levels of BDE-209 in surface water are many times lower than those
in sediment, which is consistent with assumptions based on the physical properties and environmental fate
of BDE-209 discussed in Section H.3. The highest level of BDE-209 detected in surface water was
191 pg/L, equivalent to 191 parts per quadrillion, whereas the highest amount in sediment was
3,150,000 ng/gram, equivalent to 3,150 parts per million (ppm). These two measurements differ by a
factor of approximately 165 million.
Mean levels of BDE-209 in sewage sludge in the United States have varied widely, ranging from
84.8 to 58,800 ng/gram dry weight (La Guardia et al.. 2007: Hale etal.. 2001). In these studies, BDE-209
was the most frequent PBDE congener detected. International studies have shown levels ranging from
68.5 to 880 ng/gram dry weight (Clarke et al.. 2008: Wang et al.. 2007). A mean effluent BDE-209
concentration of 1,730 pg/L was reported from a sewage treatment plant in California (North. 2004).
Internationally, effluent concentrations have ranged from 310 to 1,170 ng/L (Eljarrat et al.. 2007: de Boer
et al., 2003). See Appendix E. Table E-4 for additional information on these studies.
H.4.1.2.3. Terrestrial Systems - Soil
Mean levels of BDE-209 observed in U.S. surface soils in two studies ranged from 0.6 to
15.3 ng/gram dry weight lYOffenberg et al.. 2006) as cited in U.S. EPA QOlOa); (Yun etal.. 2008)1.
The range of concentrations in international studies was 0.028-2,220 ng/gram dry weight (Sellstrom et
al.. 2005). Both the lowest and highest levels of BDE-209 detected were from a study in Sweden.
The only other international studies examined BDE-209 levels in Chinese soil; these measurements
ranged from 2.38 to 6,319.6 ng/gram dry weight (Luo et al.. 2009: Zou et al.. 2007). Appendix E. Table
E-5 presents additional information on these studies.
H.4.1.2.4. Occupational Settings -Air
No studies were found that measured levels of BDE-209 in air in facilities where textiles
containing BDE-209 are manufactured. Outdoor air at an automobile shredding facility, however, was
found to have BDE-209 levels ranging from 45.5 to 1,940 pg/m3 (Charles et al.. 2005). Given that
BDE-209 is used in automobile upholstery, the textiles in the automobiles could have contributed to the
observed levels. BDE-209 levels have been measured in indoor occupational settings, including
e-recycling facilities, circuit-board assembly halls, and computer facilities. Reported levels in the air
ranged from a median of 220 pg/m3 in a circuit-board assembly hall to a high of 833,000 pg/m3 in an
e-recycling facility (Trederiksen et al.. 2009: Charles et al.. 2005). See Appendix E. Table E-2 for a
summary of additional studies reporting indoor and outdoor air monitoring results.
H-33
-------
H.4.1.2.5. Residential Settings -Air and Dust
Two U.S.-based studies evaluated residential indoor air levels of BDE-209. Levels of BDE-209 in
the air ranged from below the limit of detection (limit of detection not reported) to 94 pg/m3 for the living
room and 173.6 pg/m3 for personal air (i.e., breathing zone) (Allen et al.. 2007). A study in Sweden
detected BDE-209 at 257 pg/m3 in the living room air in one of five household samples, while all other
samples in the study were below the limit of detection (173 pg/m3) (Petersen and Henry. 2012).
Appendix E. Table E-2 provides a summary of additional indoor air studies.
Levels of BDE-209 in household dust are orders of magnitude higher than in other matrices by
weight, and BDE-209 is the main PBDE contaminant in household dust (Paso etal.. 2010). BDE-209 has
been found in household dust in U.S. studies at median levels ranging from 665 to 2,000 ng/gram dry
weight (Frederiksen et al., 2009). In international studies, median levels of BDE-209 in house dust ranged
from 60 ng/gram dry weight (Germany) to 7,100 ng/gram dry weight (United Kingdom) (Frederiksen et
al., 2009). See Appendix E. Table E-l for additional studies that report observed levels of BDE-209 in
household dust.
H.4.1.2.6. Nonresidential Settings -Air and Dust
BDE-209 is one of the main PBDE contaminants in office dust (Watkins et al.. 2011; Batterman
et al.. 2010; Harrad et al.. 2008). One study of U.S. office buildings found a mean concentration of
6,930 ng/gram BDE-209 in office dust, but half of the sample sites had concentrations of 1 ng/gram or
concentrations below the limit of detection (limit of detection not reported) (Batterman et al.. 2010).
Another study of dust in U.S. offices reported an average BDE-209 concentration of 4,204 ng/gram
(geometric mean) with a range of concentrations between 912 and 106,204 ng/gram among sample sites
(Watkins et al.. 2011). In a study conducted in the United Kingdom, Harrad et al. (2008) reported a
median concentration of 6,200 ng/gram BDE-209 in office dust. See Appendix E. Table E-l for study
summaries that report observed levels of BDE-209 in office dust.
One study conducted in U.S. office buildings examined BDE-209 concentrations in airborne
particulate matter and vapor; concentrations were all below the limit of detection, which was not reported
(Batterman et al.. 2010). Appendix E. Table E-2 provides a summary of this study.
H.4.1.2.7. Transportation, Including Automobiles and Airplanes—Air and Dust
As discussed in Section H.4.2.2.3 below, flame-retardant upholstery can be used in seats and
other textiles in transportation vehicles such as automobiles and airplanes. No studies were identified that
reported levels of BDE-209 in automobile or aircraft air, but several studies have investigated levels of
BDE-209 in automobile or aircraft dust (Lagalante et al., 2009; Christiansson et al., 2008; Harrad et al..
H-34
-------
2008). Less time is generally spent in automobiles than indoors, but levels of BDE-209 in automobile
dust are about 20 times higher than in household dust (Lagalante et al., 2009). In one study, the median
level of BDE-209 in passenger cars was estimated as 8.12 ug/gram dust (82% of the total PBDE
concentration in dust); personal automobiles generally had lower levels of decaBDE in dust than dealer
vehicles (Lagalante et al.. 2009). Other studies have reported median BDE-209 levels in dust from cars as
high as 100 ug/gram, with a highest individual sample of 2,600 ug/gram (Harrad et al.. 2008) (see
Appendix E. Table E-l for more information). Levels of PBDE congeners in the vehicles were not
statistically significantly different by vehicle manufacturer, model year, country of manufacture, seat
type, or the presence of heated seats (Lagalante et al.. 2009).
A study evaluating dust in aircraft during 20 international flights observed BDE-209
concentrations ranging from below the limit of detection (value not reported) to 189,882 ng/gram, with a
median level of 17,262 ng/gram (Christiansson et al.. 2008). No other studies were found that evaluated
levels of BDE-209 in air or dust of aircraft.
H.4.2. Human Exposure and Kinetics Leading to Dose
Limited data were found that measured or quantified human exposure to BDE-209. Data on
concentrations of BDE-209 measured in media such as air, soil, or dust in various settings (described in
Section H.4.1.2). however, can be used in conjunction with activity pattern and other exposure factor data
[such as those described in The Exposure Factors Handbook (U.S. EPA. 2011)] to inform estimates of
potential exposure through the various exposure pathways and scenario characteristics described in this
section.
The types of human exposure scenarios described here can be divided into four broad groups:
occupational, consumer, general public, and highly exposed populations. For the purposes of this case
study, occupational exposures include occupational exposures during synthesis, processing, or handling
of decaBDE; manufacturing of flame retardants, application of the flame retardants to textiles, or textile
finishing and upholstering; storage of the decaBDE, flame-retardant formulations, treated textiles, or
upholstered products; disposal of decaBDE, flame-retardant formulations, treated textiles, or upholstered
products; and repurposing or recycling of treated upholstery textiles and end-user products (e.g.,
furniture). Consumer exposure scenarios include the intended or unavoidable use of treated upholstery
textiles in residential and nonresidential spaces, including on household or institutional/office furniture, in
vehicles, and in aircraft; unintended uses of treated upholstery textiles or end-use products such as reuse
or repurposing of furniture for something other than its original intended use; or recycling of upholstery
textiles for new uses. General public exposure includes primary exposure to members of the community
H-35
-------
near manufacturing, disposal, or recycling facilities and secondary exposure to the general public through
environmental routes such as air, soil, or water. Highly exposed refers to exposure scenarios that are
expected to occur via similar pathways as outlined for consumers and the general public, but where
exposure levels are expected to be higher due to key differences in population characteristics such as
those described in The Child-Specific Exposure Factors Handbook (U.S. EPA. 2008a).
H.4.2.1. Occupational Exposure Pathway Scenarios
Limited data were found to determine the extent of occupational exposures to BDE-209 during
the material synthesis, processing, and handling phases or to the flame-retardant product during
formulation, application, storage, and disposal phases. See Section H.4.1.2.4 for BDE-209 concentrations
measured in occupational settings, which could be applied with the exposure pathways and scenario
characteristics described below to estimate potential exposures through scenario evaluation.
H.4.2.1.1. Synthesis, Processing, and Handling
As discussed in Section H.2.2.2. BDE-209 synthesis involves conversion of phenol to diphenyl
ether, followed by bromination in the presence of a catalyst, typically aluminum bromide or iron (WHO.
1994). The synthesis and drying processes are carried out in enclosed vessels, so under normal
circumstances exposure is unlikely to occur during this process. After synthesis, decaBDE powders are
removed from the chamber and bagged. As discussed in Section H.2.2.2.2). the low vapor pressure of
decaBDE results in negligible exposures to decaBDE as a vapor during synthesis or bagging, but
exposures to decaBDE adsorbed to dust could occur (EU. 2002). DecaBDE adsorbed to dust is expected
to settle quickly on surfaces in the occupational environment; no data were found to determine the extent
of exposures to decaBDE and decaBDE adsorbed to dust during the synthesis, processing, and handling
phases. The pathways through which workers might be exposed to decaBDE and decaBDE adsorbed to
dust during general synthesis, processing, and handling scenarios are described below:
• Inhalation. Bagging and other handling of decaBDE powders might be the activities most
likely to lead to exposures. Aerosol particles in the inhalable size range could be inhaled by
workers if respirators are not worn.
• Oral. Secondary oral exposures might occur if inhaled decaBDE or decaBDE that deposits
on the skin, food, or food-contact surfaces are subsequently ingested.
• Dermal. DecaBDE might settle on the skin if proper personal protective equipment is not
worn.
H-36
-------
H.4.2.1.2. Formulation of Flame Retardant, Application to Textiles, Upholstering
As discussed in Section H.2.2.4. decaBDE powder is mixed with other ingredients to create a
paint-like flame-retardant product, after which the flame retardant is back-coated onto a textile intended
for use as upholstery. No data were found on the extent of occupational exposures to decaBDE during
formulation of the flame retardant, application of the flame retardant to textiles, or textile finishing and
upholstering. As with exposures during the previous life-cycle stages, exposures to decaBDE vapors
during the product manufacturing stages are expected to be low due to the low vapor pressure of this
congener; however, exposures to decaBDE adsorbed to dust or attached to the product matrix could
occur.
The first step in the product manufacturing chain is compounding (i.e., mixing) decaBDE powder
with antimony trioxide in water, which typically occurs under local exhaust ventilation (EU. 2002).
The mixture is then added, through a closed system, to the emulsion polymers in a sealed mixing vessel.
Use of ventilation controls and a closed system in generating the flame-retardant coating is expected to
greatly reduce the chance of occupational exposures during these stages under normal circumstances.
Exposures might still occur to decaBDE adsorbed to dust, however, when the decaBDE powder is
emptied into the mixer. Exposures also could occur during transfer of materials; equipment cleaning,
maintenance, and repair; and as the result of accidental spills or releases (EU. 2002).
The potential for occupational exposure also exists when the flame-retardant polymer mixture is
applied to the textile as a resin back-coating. DecaBDE flame retardant is typically not added manually,
and exposures are expected to be greatest during handling and cleaning of coating equipment (EU. 2002).
but accidental spills and releases also might occur. Occupational exposures to decaBDE adsorbed to dust,
in the polymer matrix, or attached to textile fibers or scraps might occur as a result of cutting, sewing, and
otherwise abrading the decaBDE-treated upholstery textile product during textile finishing and
application to a consumer end-use product.
The pathways through which workers might be exposed to decaBDE during general formulation
of the flame retardant, application of the flame retardant to the textile, and textile finishing and
upholstering scenarios are expected to be comparable to those described in Section H.4.2.1.1 on
exposures during synthesis, processing, and handling. Additional considerations pertaining to exposures
to decaBDE in combination with polymer ingredients, textile fibers or scraps, or other product
constituents during these scenarios are described below:
• Inhalation. Abrading textiles during tailoring and upholstering could lead to inhalation of
decaBDE, other product ingredients, and textile dusts.
H-37
-------
• Oral. Secondary oral exposures might occur if inhaled decaBDE and associated product
constituents or decaBDE particles that deposit on the skin, food, or food-contact surfaces are
subsequently ingested.
• Dermal. DecaBDE and associated product constituents generated during product
manufacturing can land on the skin of workers if proper personal protective equipment is not
worn. The liquid flame-retardant coating also can be spilled directly onto the skin.
H.4.2.1.3. Storage of DecaBDE, Flame-Retardant Formulations, Treated Textiles, and Upholstered
Products
As described in Section G.2.2. decaBDE and the flame-retardant formulations to which it is
added are expected to be stored in sealed receptacles that would limit potential for worker exposures to
these materials during storage. Defective packaging and accidental spills or releases, however, could lead
to rare exposures during storage operations.
Although no information was identified regarding procedures for storing treated upholstery
textiles, these products are likely packaged to protect them from exposure to elements like water and light
that could damage their aesthetics. Such packaging also is expected to limit exposures of workers to the
flame-retardant coatings. Once the textiles have been applied as upholstery to end-use products, these
products also are expected to be enclosed in protective packaging. Some surfaces of bulkier products
(e.g., furniture), however, might remain uncovered, which could lead to worker exposures during storage
operations, or exposures might occur during application and removal of packaging materials to and from
the product. Dust also can accumulate in storage facilities that frequently store textiles and textile
products, and decaBDE that escapes from the product matrix could sorb to dust particles. Ventilation
technologies and other contamination-prevention strategies like those manufacturing facilities use are not
expected to be in place in storage facilities. Dust that has settled on surfaces in storage facilities can be
disturbed by worker operations, resuspended, and transported to other locations.
Although decaBDE is not expected to be highly volatile, off gassing of more volatile components
of the treated textiles might occur during storage of treated textiles or upholstered products. Furthermore,
due to the additive nature of decaBDE flame retardants, covalent bonding between the flame retardant and
the textile does not occur, suggesting that flame-retardant coatings that are loosely attached to the textile
surface might slough off during storage or handling. Because decaBDE flame retardants are generally
added to the back of the textile, however, the likelihood of this detachment seems low.
No data were found on occupational exposures to decaBDE during storage throughout the
product life cycle of flame-retardant upholstery textile coating. A study examining residential exposures
to a range of PBDEs, however, did identify age of furniture as one of the drivers of exposure, with higher
BDE-209 body burdens in children aged 2-5 years correlating with newer furniture (e.g., couches,
H-38
-------
mattresses) (Rose et al.. 2010). Whether furniture is currently treated more often with decaBDE flame
retardants than previously or whether the age of the furniture influences the rate of decaBDE release (with
greater amounts released from newer furniture), however, is unclear, which limits the applicability of this
finding to the refinement of realistic exposure scenarios.
The pathways through which workers might be exposed to decaBDE alone or decaBDE adsorbed
to dust during storage of decaBDE and decaBDE flame-retardant formulations are expected to be
comparable to those described in Sections H.4.2.1.1 and H.4.2.1.2 on exposures during synthesis,
processing, and handling and during formulation of the flame retardant, application to textiles, and
upholstering. Additional considerations pertaining to exposures to decaBDE alone or in combination with
polymer ingredients, textile fibers or scraps, or other product constituents during general treated textile of
upholstered product storage scenarios are described below:
• Inhalation. Workers could inhale volatile components of the flame-retardant coating or
decaBDE adsorbed to dust in storage facilities, particularly facilities that are not well
ventilated. Furthermore, decaBDE adsorbed to dust could be resuspended in the air by worker
activities, and subsequently inhaled. PBDE exposures have not been measured at textile
storage facilities, but they have been measured at electronic waste storage facilities.
The median estimated inhalation exposure to BDE-99 (the highest measured PBDE in air) in
male workers was 0.0011 ng/kg body weight (bw) per day (Muenhor et al.. 2010). Although
BDE-209 exposures are likely to differ due to lower volatility and greater propensity to
adsorb to particles, debromination of BDE-209 could result in worker exposures to lower
brominated congeners.
• Oral. Higher levels of dust in textile storage facilities could lead to increased transport of
decaBDE adsorbed to dust. This could result in oral exposures to decaBDE in dust
transported to break rooms, homes (via clothes), and other locations where decaBDE
adsorbed to dust can be unintentionally ingested while eating or due to hand-to-mouth
activity. In electronic waste storage facilities, the median exposure to BDE-209 via dust
ingestion was 2.89 ng/kg-bw-day for average ingestion scenarios and 7.2 ng/kg-bw-day for
high-end ingestion scenarios (Muenhor et al.. 2010).
• Dermal. DecaBDE adsorbed to dust could be resuspended by worker activities and deposit
on the skin of workers if proper personal protective equipment is not worn.
H.4.2.1.4. Disposal and Recycling of DecaBDE, MWCNTs, Flame-retardant Formulations, Treated
Textiles, and Upholstered Products
As described in Section H.2.5. large-scale disposal, recycling, and reuse of decaBDE, and the
flame-retardant formulations to which it is added are unlikely, but containers used to store these products
might enter the waste stream, and workers at disposal and recycling facilities could be exposed to product
residues remaining in these containers.
H-39
-------
Disposal and recycling of treated textiles and upholstered products, however, is prevalent. Mixing
and compacting of waste for land-filling; cleaning, shredding, blending, melting, and spinning scrap
textiles for recycling; and incomplete incineration of treated upholstery textiles all could result in
exposure of workers (Chaudhry et al.. 2009) to decaBDE, primarily in combination with other product
constituents and dusts.
No data were found on the extent of occupational exposures to decaBDE during disposal,
recycling, and reuse throughout the decaBDE flame-retardant upholstery textile coating product life cycle.
As discussed in Section H.4.1.2.4. BDE-209 was detected in the air outside of an automobile shredding
facility (Charles et al.. 2005). which suggests that exposure to BDE-209 during end-of-life operations can
occur.
The pathways through which workers might be exposed to decaBDE during general disposal and
recycling of decaBDE powder and flame-retardant formulations are expected to be comparable to those
described in Section H.4.2.1.1 (exposures during synthesis, processing, and handling), and worker
exposure pathways for decaBDE in combination with polymer ingredients, textile fibers or scraps, or
other product constituents during disposal and recycling treated textiles and upholstered product are
expected to be similar to those described in Sections H.4.2.1.2 (exposures during formulation of the flame
retardant, application to textiles, and upholstering) and H.4.2.1.3 (exposure during storage and
distribution) for these products. Additional considerations pertaining to exposures to decaBDE alone or in
combination with polymer ingredients, textile fibers or scraps, or other product constituents during treated
textile or upholstered product disposal and recycling scenarios are described below:
• Inhalation. Workers operating machines that abrade or destroy textile materials, those
handling these products, and other workers in the vicinity of operations that agitate or abrade
textile materials can inhale decaBDE adsorbed to dust and other product constituents, as
observed by Sjodin et al. (2001) at an e-waste recycling plant where electronics are stored,
dismantled, and shredded. Workers at incineration facilities also might inhale small particles
comprising decaBDE and other substances in the incinerator as well as polybrominated
dibenzofurans (PBDFs) and polybrominated dibenzo-p-dioxins (PBDDs) (see Section
H.2.5.2.2) if treated textiles and upholstered products are not incinerated at sufficiently high
temperatures.
• Oral. No additional considerations.
• Dermal. Workers at disposal and recycling facilities might come into physical contact with
the decaBDE flame-retardant coating on an upholstery textile during the process of moving or
handling products. DecaBDE could migrate directly to skin of workers if proper personal
protective equipment is not worn.
H-40
-------
H.4.2.2. Consumer Exposure Pathway Scenarios
BDE-209 is expected to be released from consumer products in the particulate phase. See
Sections H.4.1.2.5. H.4.1.2.6. and H.4.1.2.7 for data on concentrations of BDE-209 measured in
residential, nonresidential, and general public settings, which could be applied with the exposure
pathways and scenario characteristics described below to estimate potential exposures to BDE-209
through a scenario evaluation approach.
H.4.2.2.1. Intended Use - Upholstered Products in Residential Spaces
As discussed in Section H.4.1.2.5. PBDE concentration in house dust tends to be higher than in
other matrices, and BDE-209 is the dominant congener in house dust. Although flame-retardant
upholstery textiles typically are used in nonresidential settings (see Section H.4.2.2.2). some residential
upholstered products, particularly mattresses, are known to contain decaBDE, and other upholstered
furniture products, like couches, sometimes might be treated with decaBDE (Rose et al.. 2010). One study
has shown that body burdens of BDE-209 in children are positively associated with presence of new
furniture, but are not associated with presence or use of electronics that often contain decaBDE (Rose et
al.. 2010). This finding suggests that decaBDE use in residential upholstery does contribute to overall
decaBDE exposures related to the use of decaBDE in flame-retardant upholstery textiles. Higher body
burdens of BDE-209 also have been associated with smaller living spaces, and higher concentrations of
decaBDE have been measured in the main living area of the house than in the bedroom (Allen et al..
2008). indicating that variations in decaBDE exposures can be expected due to variations in housing
characteristics and human behavior patterns (i.e., time spent by individuals in different rooms or outside
the house). Furthermore, decaBDE released from products is suspected to debrominate to some degree to
lower brominated congeners in residential settings (Allen et al.. 2008). and will therefore lead to
exposures to PBDEs other than decaBDE.
As introduced in Section H.2.4. upholstered products are expected to be used for many years, and
contact with the textile might be frequent and prolonged, which could cause substantial wear and tear on
the textile product. In addition, upholstery in residential spaces might frequently be exposed to cleaning
products, sweat, food, and other substances that could affect the properties of the textile and the flame-
retardant coating.
The pathways through which consumers might be exposed in residential settings to decaBDE
during general consumer use scenarios for end products upholstered with decaBDE are described below:
H-41
-------
• Inhalation. Chronic inhalation of particles of decaBDE in combination with other product
constituents and dust could occur following release from upholstered products over time (due
to wear and tear from anticipated use, aging of materials, abrasion, UV light, water, cleaning
chemicals, among other factors; see Section H.2.4.2). Particulate decaBDE could settle onto
surfaces, where it might be disturbed and re-entrained, after which it could be inhaled by
residents. Inhalation is not expected to be a primary route of exposure for decaBDE (Johnson-
Restrepo and Kannan, 2009; Allen et al.. 2008). however, because the contribution of inhaled
dust particles is expected to be minimal due to a lack of correlation between concentrations of
decaBDE in dust and in air (Allen et al.. 2008); see below.
• Oral. DecaBDE in combination with other product constituents and dust could be ingested
after settling on food and food-contact surfaces or following hand-to-mouth activity.
Ingestion of household dusts is hypothesized to be a major exposure pathway for PBDEs
(Allen et al.. 2008). Johnson-Restrepo and Kannan (2009) reported that most PBDE intake in
toddlers, children, teenagers, and adults was attributed to the oral route of exposure via
ingestion of household dust (56-77% attributed to combined oral and dermal exposure).
• Dermal. Dermal exposure to decaBDE in combination with other product constituents and
dust might occur while touching the textile surface (particularly if the portion of the textile
that has been treated with the flame-retardant coating is exposed) or touching surfaces upon
which particles have settled (Frederiksen et al.. 2009). Lorber (2008) estimated that dermal
exposure to PBDE compounds in household dust could be a significant contributor (estimated
at 16%) to the body burden of PBDEs in adults. The estimated contribution of BDE-209 to
body burden from dermal contact exposure was 25.2 ng per day of a total 85.9 ng total
PBDEs per day from that exposure route (Lorber. 2008). Johnson-Restrepo and Kannan
(2009) similarly reported that the dermal route was a primary route of exposure for PBDEs
(second to the oral route of exposure for contribution of human intake).
H.4.2.2.2. Intended Use - Upholstered Products in Nonresidential Spaces
Due to regulations requiring that upholstery textiles used in nonresidential settings pass flame-
retardancy tests (see Table H-3). many upholstery textiles in public, commercial, and institutional settings
are treated with decaBDE. The characteristics of the different settings in which these products are used
can vary considerably. For example, flame-retardant upholstery textiles might be used in seating for
airports and other transportation hubs and in waiting rooms, office buildings, penal institutions, and other
nonresidential spaces that can range from very small to very large and where consumers might spend
varying amounts of time. Some scenarios for nonresidential exposures are not likely to differ from those
expected from residential exposures, but a few key differences do exist:
• Exposures to flame-retardant upholstery coatings in public spaces might be unavoidable.
Although consumers have some control over which products they bring into their home, they
have little control over the products they encounter in public spaces.
H-42
-------
• Some nonresidential exposures might occur over long periods of time and for extended
intervals (e.g., sitting in the same office chair every day over the course of several work
years), although some might occur infrequently and for short periods of time (e.g., sitting in
seating at the airport waiting for a flight).
• Products in public spaces might experience higher activity levels, more frequent cleaning,
and less care to the textile surface, all of which could damage or weaken the textile matrix
and influence releases and exposures.
With the exception of these potential differences in exposure settings and activity patterns, the
pathways and scenarios through which consumers might be exposed in nonresidential settings to
decaBDE during general consumer use scenarios for end products upholstered with decaBDE flame-
retardant coatings are not expected to differ from those described previously in Section H.4.2.2.1 on
exposures from intended use of upholstered products in residential spaces.
H.4.2.2.3. Intended Use -Aircraft and Automobile Upholstery
Flame-retardant upholstery can be used for seating, draperies, carpets, and other textiles in
passenger cars and public and private transportation. Aircraft and automobile passengers, and those
working in these environments (e.g., cab drivers, flight attendants), could be exposed to higher levels of
BDE-209 due to the higher concentrations of PBDEs in dust in those environments as compared to home
environments (Lagalante et al.. 2011; Christiansson et al.. 2008). Dust generated by abrasion of treated
upholstery fabric is the most likely pathway for BDE-209 exposure in automobiles (Lagalante et al..
2011). No association was found, however, between time spent in automobiles and plasma PBDE levels
in children aged 2-5 years in California (Rose etal. 2010); these children spent an average 7.2 hours per
week (range 0-20 hours) in the car.
Photodegradation of BDE-209 is low in cars because automobile glass blocks UVB radiation,
which is the region of the spectrum most strongly absorbed by BDE-209. BDE-209 adsorbed to sodium
sulfate does photodegrade in automobiles, however, and has a half-life of approximately 19 days.
The congeners BDE-47 and BDE-99 are environmentally and lexicologically relevant products of BDE-
209 debromination (see Text Box H.3-1). but their presence in automobile dust is mainly from
volatilization and weathering of products containing pentaBDE, rather than photodegradation of BDE-209
and other higher brominated congeners (Lagalante et al., 2011).
The pathways through which consumers might be exposed in vehicles (including airplanes) to
decaBDE during general consumer use scenarios for end products upholstered with decaBDE flame-
retardant coatings are described below:
H-43
-------
• Inhalation. Inhalation of decaBDE adsorbed to dust from worn or abraded automobile
upholstery is expected to occur. Inhalation exposure to photodegradation products of
decaBDE could occur in automobiles if automobile textiles contain decaBDE adsorbed to
sodium sulfate. The recirculation of air in aircraft cabins also might affect exposure to
particulate decaBDE if filters do not adequately remove these particles.
• Oral. Secondary oral exposures might occur if inhaled particulate decaBDE or particulate
decaBDE that deposits on the skin is subsequently ingested.
• Dermal. Dermal exposures to decaBDE or decaBDE photodegradates (due to worn or
abraded automobile upholstery) are expected to occur, particularly when skin touches the
treated part of the textile directly. Dermal exposure also can occur when particles in the air
settle on the skin. Different exposure characteristics or scenarios (e.g., children sitting in
safety seats) might influence whether dermal exposure occurs, or influence the extent to
which exposure occurs through this pathway.
H.4.2.2.4. Unintended Use, Repurposing, or Reuse of Treated Textiles and Upholstered Products
As introduced in Section H.2.4. unintended uses of upholstery textiles treated with decaBDE
flame-retardant coatings could include repurposing of treated upholstery textiles for clothing, building
insulation, other in-home or outdoor furnishings, bedding, or other purposes. The repurposing stages
could introduce occupational exposures similar to those discussed in Section H.4.2.1.2 (exposures during
formulation of the flame retardant, application to textiles, and upholstering) and Section H.4.2.1.3
(exposure during storage and distribution), as products that are treated with flame-retardant coatings are
broken down and reprocessed into new products.
Although no information was identified that directly addresses potential consumer exposures
following unintended use or reuse of flame-retardant upholstery textiles, exposure pathways and scenarios
from other life-cycle stages are relevant here. Most reuse scenarios might differ little from those for
anticipated consumer uses, but a few key differences might occur, particularly when products are
repurposed for new uses or used in unintended ways:
• Processes similar to those involved with product manufacture (e.g., cutting, sewing) and
storage of textiles also might be employed for repurposing treated textiles. In this scenario,
however, these processes are not expected to occur in an occupational setting, but in the home
or another private space, where no personal protective equipment is worn and limited control
technologies are used. These processes, as employed for repurposing textiles, however, are
not expected to occur as commonly or at the same scale as in a manufacturing facility.
• Older, more degraded textiles with weakened matrices might be handled directly and be
subjected to abrasion, thereby releasing the product constituents in the vicinity of the
consumer conducting the repurposing.
H-44
-------
• Although dermal contact with products used for their intended purpose (e.g., furniture
seating) might be limited by a clothing barrier between the consumer and the treated textile,
should flame-retardant upholstery textiles be repurposed into clothing, direct dermal contact
might occur repeatedly over long periods of time.
With the exception of these potential differences in exposure characteristics, the pathways and
scenarios through which consumers might be exposed to decaBDE during repurposing, reuse, or
unintended use of treated textiles and upholstered products are not expected to differ from exposure
pathways associated with the cutting, tailoring, or other abrasive processes involved with product
manufacturing (Section H.4.2.1.2); storage of textile products (Section H.4.2.1.3); and consumer use in
residential and nonresidential spaces (Sections H.4.2.2.1 and H.4.2.2.2).
H.4.2.3. General Public Exposure Pathway Scenarios through Environmental Media
No information was found on exposure to decaBDE in the general public from environmental
media (e.g., air, water, soil). See Section H.4.1.2 for concentrations of BDE-209 in environmental media
that could be used with the exposure pathway and scenario characteristics below to estimate potential
exposures.
H.4.2.3.1. Outdoor Air
Releases of decaBDE to outdoor air throughout the product life cycle of the flame-retardant
textile coatings are possible (see Section H.2). Once released to air, decaBDE can sorb to particulate
matter and experience long-range transport to areas distant from its source (see Section H.3.2). As
summarized in Section H.4.1.2.1. decaBDE has been measured in outdoor air at concentrations much
lower than those measured indoors. Nonetheless, general public exposures to decaBDE adsorbed to
particulate matter in ambient air are expected to occur, with the primary route being inhalation.
Other product constituents of flame-retardant textiles (e.g., pieces of the polymer matrix or the
textile fabric) also can be released, and in the case of decaBDE, combustion by-products (e.g., PBDDs,
PBDFs) and lower brominated transformation products are expected to be present in air as a result of the
flame-retardant textile coating life cycle.
H.4.2.3.2. Water
Releases of decaBDE and other product constituents to wastewater and ambient water bodies
throughout the product life cycle of flame-retardant textile coatings are possible (see Section H.2). Once
released to water, decaBDE is expected to sorb to particulate matter in the water column or to sediments,
H-45
-------
which might limit their mobility (see Section H.3.3). This behavior implies that decaBDE also primarily
will be removed to sludge during wastewater treatment.
As summarized in Section H.4.1.2.2. decaBDE has been measured in surface waters at low
concentrations [below levels of detection to 191.0 pg/L (Pros et al.. 2005)]. As a result, general public
exposures to decaBDE and its transformation products in water are expected to occur, with the primary
routes being dermal (through bathing and swimming) and oral (drinking and incidental ingestion during
bathing and swimming).
H.4.2.3.3. Soil
Releases to ambient air and water throughout the product life cycle of flame-retardant textile
coatings will result in deposition of particles of decaBDE and other product constituents (see Section H.2)
to soil. Once deposited, decaBDE is expected to sorb strongly to soil, which might limit mobility (see
Section H.3.4).
As summarized in Section H.4.1.2.3. decaBDE has been measured in surface soils and is
expected to be present in sludge applied to agricultural soils. As a result, general public exposures to
decaBDE and its transformation products in soils are expected to occur, with the primary routes being
dermal and oral (although in incidental amounts). Furthermore, decaBDE has been shown to translocate
from soil to plant tissues, suggesting that decaBDE can enter the food web, and dietary oral exposures
also might occur.
H.4.2.4. Highly Exposed Populations
This section discusses characteristics of individuals and populations that might result in increased
exposure (relative to the general population) to decaBDE released during the life cycle of flame-retardant
upholstery textile coating.
The primary exposure pathway for decaBDE is likely to be ingestion of household dust, and dust
levels in the home can vary by socioeconomic status or the type and condition of housing (see Section
H.5.3.1). Disproportionate levels of exposure can occur in specific populations, including low-income and
low-educational-attainment populations. Although race and ethnicity have not been shown to be
associated with specific physiological conditions that increase susceptibility to exposure, demographic
factors such as socioeconomic and educational status could cause some populations to bear a
disproportionate level of the exposure burden.
Children are likely to experience higher exposures than the general population. Data suggest that
breast-fed infants are potentially exposed to BDE-209 through their mother's milk [i.e., worldwide,
median detected levels of BDE-209 in breast milk range from 0.1 to 2.9 ng/gram liquid weight
H-46
-------
(Frederiksen et al., 2009); maximum concentration of 7 breast milk samples containing decaBDE in the
United States was 8.24 ng/gram lipid (ATSDR. 2004)1. Young children also take in more household dust
than adults, with estimates for children at 100-200 mg/day compared to 50 mg/day for adults (U.S. EPA.
2008a). Increased hand-to-mouth activity contributes to increased exposures in children. Occupation also
could increase exposure relative to the general population, primarily for workers involved in manufacture
of decaBDE or flame retardants containing decaBDE, or textile products treated with decaBDE.
H.4.2.5. Exposure Reference Values and Recommendations
A variety of exposure standards, guidelines, or recommendations are developed by different
organizations with purview over specific portions of the population or situations during which exposure
might occur (e.g., occupational exposures, general population drinking water exposures). Available
information on these types of values for decaBDE is presented below. Section H.5.1.1 discusses how
some of these values inform quantitative toxicity assessments.
As of January 2011, no national-level environmental or occupational health standards had been
established for decaBDE (POD. 2011). EPA has derived a reference dose (RfD) for decaBDE, based on
developmental neurobehavioral effects, of 0.007 mg/kg-day (U.S. EPA. 2008b) (see Section H.5.1).
An RfD is an estimate (taking into account uncertainty) of the daily exposure to the human population,
including sensitive populations, that is "likely to be without an appreciable risk of deleterious effects
during a lifetime" (U.S. EPA. 2008b). EPA also has derived a cancer slope factor for decaBDE based on
neoplastic nodules or carcinomas (combined) in the liver of treated male rats, of 7 x 10~3 per mg/kg-day.
A cancer slope factor is a plausible upper bound on the estimate of risk per mg/kg-day of oral exposure
(U.S. EPA. 2008b). The Agency for Toxic Substances and Disease Registry (ATSDR) has developed a
minimal risk level (MRL) for decaBDE for intermediate duration (15-365 days) oral exposure of
10 mg/kg-day (ATSDR. 2004) based on Hardy et al. (2002). MRL values are estimates of the daily
exposure to a hazardous chemical that is likely to be without appreciable risks of noncancer health effects
over a specific duration of exposure.35 MRLs are intended as screening levels, rather than clean-up or
action levels for any agency. Differences in exposure duration (chronic lifetime versus intermediate) and
the key study used to derive the estimate contributed to the several-orders-of-magnitude difference
between the MRL and RfD for decaBDE.
35ATSDR - Minimal Risk Levels: http://www.atsdr.cdc.gov/mrls/index.asp.
H-47
-------
H.4.2.6. Toxicokinetics, Dose, and Body Burden
Toxicokinetics can be used to relate exposure and contact, such as those described in the
scenarios above, with uptake and dose. Specifically, toxicokinetics describes how a material is absorbed,
distributed, metabolized, and excreted in an organism. An understanding of the relationship between each
of these concepts, which are often referred to as AD ME, leads to an understanding of the concentration,
or dose, of material that can reach—and potentially accumulate in—different tissues of the body.
H.4.2.6.1. Absorption, Distribution, Metabolism, Excretion
This section contains information regarding the toxicokinetic behavior of decaBDE when
administered to mammals. Information regarding birds and fish is not presented in this section because,
when extrapolating toxicokinetic data to humans, studies conducted with rodents (rat or mouse) or
nonrodent mammals (dog or monkey) are generally used. Additionally, differences among species have
been noted in numerous studies of decaBDE, and the toxicokinetic behavior in response to decaBDE
differs among birds, fish, and mammals. For example, fish generally debrominate decaBDE to pentaBDE
congeners, although mammals debrominate decaBDE, to a lesser degree, to heptaBDE congeners. See
Section H.4.3 for toxicokinetic information relevant to ecological exposures.
Early toxicokinetic studies (el Dareer et al., 1987; NTP. 1986; Norris et al., 1975; Norris et al.,
1973) were conducted on decaBDE shortly after it was developed as a flame retardant. These studies
demonstrated that decaBDE is poorly absorbed [0.3-1.5%; (NTP. 1986)] from the gastrointestinal tract in
rats following oral exposure and eliminated in the feces as the parent congener without prior metabolism.
Essentially no elimination occurred through the urine, and more than 99% of the dose was recovered in
feces by 48 hours, indicating a lack of accumulation in tissues (Norris etal.. 1975). The half-life of
decaBDE is relatively short; the serum half-life was reported as 15 days in a human study where workers
were exposed to BDE-209 (U.S. EPA. 2010a).
More recent studies, although in general agreement, have reported higher absorption rates that
might be due to the solvent used to administer decaBDE. In general, BDE-209 is not expected to
accumulate in terrestrial organisms. Many studies, however, have reported levels of BDE-209 in humans
(breast milk, serum, and umbilical cord blood), food items (dairy, eggs, infant formula), and biota (fish,
shellfish), indicating that some absorption and accumulation occur overtime (Frederiksen et al., 2009).
Given the high trophic levels of the organisms where accumulation has been observed (humans, predatory
fish, and piscivorous birds), biomagnification appears to occur in these receptors (Environment Canada,
2010; U.S. EPA. 2010a). Bioaccumulation and biomagnification are discussed further in Section G.4.2.1.
Because of the low absorption of BDE-209, blood and tissue levels following acute (short-term)
oral exposures are typically low and represent a small fraction of the total dose. More than 66% of the
H-48
-------
parent compound was excreted in the feces of rats following oral exposures to BDE-209 (Riu et al., 2008;
Morck et al., 2003). The same experiments showed that the highest concentrations of BDE-209 were
found in plasma and blood-rich tissues such as liver, kidney, adrenal glands, ovaries, heart, and the
intestinal wall following a single oral exposure (Morck et al.. 2003) or 4-day gavage exposure (Riu et al..
2008; Morck et al.. 2003). In these acute studies, BDE-209 was not readily distributed to adipose tissue.
In contrast, evidence shows that BDE-209 can accumulate in adipose tissue following chronic
oral exposure. Studies by Norris et al. [1974; 1975, as cited in Hardy et al. (2009)] exposed Sprague-
Dawley rats to a commercial product called FR-300-BA in the diet at 0.01, 0.1, and 1.0 mg/kg-day for
3, 6, or 12 months. FR-300-BA comprised 77.4% BDE-209, 21.8% nonaBDE, and 0.8% octaBDE. After
6 months of treatment, bromine concentrations (measured by neutron activation analysis) in adipose tissue
were higher in treated rats (~3 ug/gram) than in controls (~1 ug/gram), but after 12 months the bromine
levels in adipose tissue were similar to controls. Bromine did not accumulate in other tissues such as liver,
kidney, and serum. Norris et al. [1974; 1975, as cited in (Hardy et al., 2009)1 also followed the
elimination of bromine from male Sprague-Dawley rats that were dosed with FR-300-BA in the diet for
90 days at 1.0 mg/kg-day and subsequently fed a control diet. After 10 days on a control diet,
concentrations of bromine in the liver were similar to controls, but concentrations in adipose tissue were
higher (-2.5 to 4 ug/gram) than controls (-0-2 ug/gram). Another study showed a time- and dose-
dependent increase of bromine levels in adipose tissue indicating accumulation following dietary
exposure to decabromodiphenyl oxide (a synonym for decaBDE) at 0.01, 0.1, or 1 mg/kg-day, for up to 2
years (Kocibaet al.. 1994). Kociba et al. (1994) also reported that bromine content was not increased
compared to controls in the kidney, muscle, or serum of rats in the same study.
Absorbed decaBDE is metabolized in the liver and a minor fraction of the parent compound is
metabolized to lower PBDE congeners, such as tetraBDE and pentaBDE (see Text Box H.3-1). Morck et
al. (2003) reported that decaBDE was the predominant substance detected in the liver metabolites, with
trace levels of nonaBDE. Similarly, a minor fraction of decaBDE (less than 3%) was debrominated to
lower BDE congeners in a feeding study that exposed male Sprague-Dawley rats to DE-83R (98.5%
decaBDE) (Huwe and Smith. 2007).
Metabolism of decaBDE to lower brominated congeners by oxidative debromination is indicated
by some evidence in studies of rats and fish, but the mechanisms and location of metabolic processes are
not well characterized due to limited availability of toxicokinetic data (Hakk and Letcher. 2003).
Metabolism of decaBDE also differs among species. In lactating cows, Kierkegaard et al. (2007)
suggested that decaBDE debrominates to hepta-, octa-, and nonaBDEs. Octa- and nonaBDEs were found
in liver and kidney of rats fed 100 mg/kg-day BDE-209 for 3 months (Wang et al.. 2010a). In contrast, in
H-49
-------
vitro studies of human hepatocytes have shown evidence of low or no metabolism, possibly because of
low entry of BDE-209 into cells under the experimental conditions used (Stapleton et al.. 2009).
Based on available studies (Huwe and Smith. 2007; Kierkegaard et al.. 2007; Hakk and Letcher.
2003; Morck et al.. 2003; Sandholm et al.. 2003). the following pathways for debromination of decaBDE
can be deduced for mammals:
1. Deiodinase enzymes can debrominate decaBDE to nona-, octa-, and heptaBDEs.
2. Debrominated neutral metabolites can undergo hydroxylation to potentially form phenols
or catechols, possibly via an arene oxide, which could involve the action of cytochrome P450
enzymes.
a. The formed hydroxylated BDEs can compete with thyroxine for binding to a
thyroxine transport protein present in blood serum.
b. The catechols then are methylated, potentially by the action of catechol-O-
methyltransferase, to form guaiacols.
c. The guaiacol metabolites further oxidize to highly reactive quinones, which bind to
cellular macromolecules.
d. The reactive intermediates are subject to rapid conjugation via Phase II metabolic
processes, leading to water-soluble metabolites that are excreted via bile and feces,
as observed in conventional and cannulated rats.
Two toxicokinetic studies of fetal rats were identified. In a study by Riu et al. (2008).
radiolabeled 14C-BDE-209 (99.8% pure, dissolved in peanut oil) was administered orally to pregnant rats
on Gestation Days (GD) GD16-GD19 (2 mg/kg-day). The toxicokinetic results were similar to those
noted previously in this section. Approximately 72% of the dose was found in the feces and the digestive
tract contents, although 0.1% was excreted in the urine. The remainder of the dose was distributed in
various tissues, with 6.5% in the liver and 5.3% in the digestive tract contents. All other tissues contained
less than 1% of the administered dose. The fetuses (sum for the whole litter) contained 0.43% of the dose.
In a recent study by Cai et al. (2011). BDE-209 and its metabolites were detected in the placenta and
milk, and eventually in the fetuses or neonates when BDE-209 (prepared in peanut oil) was administered
to pregnant Sprague-Dawley rats from GD7 to PND4. In the same study, detectable amounts of
nonaBDEs (BDE-206, 207, 208) and octaBDEs (BDE-196, 197/204, 198/203) were observed in the dosed
rats. The predominant debrominated metabolites of BDE-209 detected in fetuses were nonaBDEs (BDE-
208, 207, 206). The level of BDE-206 in the fetal or pup bodies was significantly lower on GD21 and
PND4 than on GD7. The octaBDEs BDE-196, BDE-198, and BDE-203 were observed in fetuses and
pups, but were minor debromination metabolites of BDE-209.
No animal studies have been identified that evaluate decaBDE ADME upon inhalation and
dermal exposures.
H-50
-------
H.4.2.6.2. Internal Dose and Body Burden
Levels of decaBDE in human tissues have been reported in several occupational studies and in
studies of the general public. A study in workers at an electronics dismantling plant in Sweden evaluated
levels of five PBDEs in serum samples from plant workers. The mean concentration of BDE-209 in that
study was as high as 5 ng/gram lipid (Darnerud et al.. 2001). Total PBDE serum levels in hospital
cleaners, computer clerks, and electronics dismantlers were 3, 4, and 26 ng/gram lipid, respectively
(Darnerud et al., 2001). As discussed in the previous section, decaBDE can accumulate in adipose tissue
over time with chronic exposure (Hardy et al.. 2009). DecaBDE also might biomagnify in the food web
from lower trophic levels to higher trophic levels. DecaBDE can debrominate to lower PBDE congeners
in the body, which are more bioaccumulative than decaBDE (Yogui and Sericano. 2009).
As shown in Table H-5, the median level of BDE-209 in the serum reported in one U.S. study
was less than 0.96 ng/gram liquid weight. International studies reported median serum levels ranging
from 0.77 to 18.5 ng/gram liquid weight. The highest median levels reported were in men aged 40-50
years in Norway and Sweden, and the lowest median levels were from maternal serum in a study in the
Faroe Islands (Trederiksen et al., 2009).
BDE-209 has been detected in breast milk in American women at a measured mean concentration
of 0.92 ng/gram liquid weight (see Table H-5). A study that evaluated the breast milk of women in the
Pacific Northwest region of Canada and the United States observed a median level of 0.43 ng/gram liquid
weight.
Worldwide, median detected levels of BDE-209 in breast milk ranged from 0.1 to 2.9 ng/gram
liquid weight (Trederiksen et al.. 2009). Decline in the use of decaBDE flame retardants containing BDE-
209 are expected to result in a decline in breast milk concentrations over time.
Breastfeeding infants, and even infants who are fed infant formula, are likely to be exposed to
BDE-209 through consumption of breast milk and formula. Levels of BDE-209, and levels of total
PBDEs, are higher in American samples of breast milk than levels found in infant formula. The levels in
both breast milk and formula, however, are far lower than the amount detected in household dust
(Frederiksen et al., 2009).
H-51
-------
Table H-5. Median tissue concentration ranges (in ng/gram liquid weight) for three polybrominated
diphenyl ether congeners in humans.
Country
Breast Milk
United States
International
BDE-47
7.69-27.8a
0.03-27.8
BDE-99
1.46-5.7
0.02-5.36
BDE-209
0.92b
0.1-2.9
Adipose Tissue
United States
International
Blood (Serum)
United States
International
Cord Blood
United States
International
Placenta
United States
International
29.3
0.52-2.3
0.63-46
0.25-4.55
13.6-25
0.98-3.8
NR
0.25-0.77
10.3
0.236-1.4
0.32-13
0.09-1.94
4.3-7.1
0.07-4.3
NR
0.12-0.41
NR
NR
<0.96
1.1-18.5
Below detection0
2.2
NR
1.0
aHigh level observed in a joint United States/Canada study.
bMean concentration reported.
°Detection limits not reported.
Abbreviations: NR = Not reported.
Source: Frederiksen et al. (2009).
HAS. Ecological Exposure and Kinetics Leading to Dose
H.4.3.1. Factors Impacting Ecological Exposure
In biota, potential exposure routes for decaBDE include ingestion, inhalation, or direct contact.
The potential for exposure via each route along with subsequent uptake and dose depends on several
H-52
-------
factors, including properties of the environmental media and physiological and behavioral characteristics
of aquatic and terrestrial organisms. These factors can, in turn, influence the bioavailability of decaBDE.
As discussed in Section H.3. the physicochemical properties of BDE-209 dictate partitioning into the
environment. This partitioning drives the exposure potentials for water-dwelling, sediment-dwelling, and
terrestrial organisms. For example, BDE-209 preferentially binds to soils and sediment when released to
the environment (Hale et al.. 2006) and likely will be present only in limited quantities in surface water or
ground water (see Section H.4.1.2.2) (U.S. EPA. 2010a). Sediment-dwelling organisms are therefore key
ecological receptors of BDE-209.
H.4.3.2. Absorption, Distribution, Metabolism and Excretion in Ecological Receptors
As discussed in Section H.4.2.6. an understanding of ADME processes can be used to relate
exposure concentrations to the concentration, or dose, of material that reaches the tissues of an organism.
Elucidation of organism-specific ADME processes can help explain observations of high body burdens
that were not predicted based on environmental fate and partitioning alone. ADME processes influence
whether and for how long a material is retained in a tissue (i.e., whether the material will bioaccumulate)
and how such retention rates might differ among trophic levels (i.e., whether concentrations of the
material will biomagnify in a food web).
Bioaccumulation and biomagnification have been shown to influence ecological exposures for
decaBDE, as described further in Sections G.4.2.3 and H.4.3.4. Bioaccumulation is the process by which
an organism takes a chemical into the body through all exposure routes and dilutes the chemical through
excretion, metabolism, and growth, but accumulates a net "body burden" of the chemical (Environment
Canada. 2010; U.S. EPA. 2010a). Biomagnification is the process by which a chemical increases in
concentration in tissues as it moves up trophic levels in an ecosystem (U.S. EPA. 2010a).
Bioaccumulation factors, the ratio of the chemical contaminant in the tissue of the biota (from dietary
exposure and uptake directly from media) to chemical contaminant in the medium, and biomagnification
factors (BMFs), the ratio of the chemical concentration in an organism's tissue to the concentration of the
same chemical in the tissues of its diet, are used as measures of persistence and potential for impacts as a
chemical moves through an ecosystem (U.S. EPA. 2010a). A substance is considered bioaccumulative
when it has a high bioaccumulation factor, generally greater than 5,000 (U.S. EPA. Final Rule 40 Code of
Federal Regulations 372). Bioaccumulation studies that show body burdens in organisms in remote
locations far from a direct, nondietary exposure source (e.g., water, air), such as those in Greenland
peregrine falcons or Florida coastal sharks, are indicative of trophic biomagnification through the food
web (Environment Canada. 2010). Biomagnification can be modeled using fugacity-based dynamic fate
H-53
-------
models that consider environmental conditions, ecosystem properties, and food-web dynamics (Lim and
Lastoskie. 2011).
Ecological receptors are likely to be exposed to decaBDE through treated products or scraps and
debris from products generated during end-of-life stages of the product life cycle (see Section H.2).
The materials released during these processes can contain components other than the contaminant of
concern (e.g., textile material, glue, composite ingredients). As discussed at the beginning of this section
(H.4) and throughout Section H.4.2. studies are lacking on the matrix-bound state of these compounds
and how exposure characteristics and dose implications differ for the free and matrix-bound forms. As for
the discussion of human exposures in Section H.4.2. exposure considerations for ecological receptors are
informed by data on BDE-209 not embedded in a polymer matrix or associated with other product
ingredients (e.g., textile fibers, coating ingredients). Field studies have found raw PBDEs in
environmental media, which indicates that the compounds can leach from the product matrix (see
Appendix E).
H.4.3.3. Exposure Pathways in Aquatic Systems
PBDEs primarily transition to the sediment in aquatic ecosystems (Mikula and Svobodova.
2006); as a result, benthic organisms might take up decaBDE via absorption or ingestion of sediment.
Secondary exposure via movement through the food web results in greater body burdens of PBDEs in
predatory fish than in herbivorous or omnivorous fish (Mikula and Svobodova. 2006). In fish, uptake of
PBDE from the water column via gills is limited by the large molecular size of PBDEs, but dietary uptake
efficiencies have been shown to range from 40 to 92% (Mikula and Svobodova. 2006). Due to the
tendency for PBDEs to partition into sediment and the inefficiency of uptake via gills, the more likely
route of exposure for fish is secondary exposure due to bioaccumulation and biomagnification in the food
web.
H.4.3.3.1. Toxicokinetics and Body Burden in Aquatic Systems
Limited information is available on the mechanisms of BDE-209 ADME in aquatic organisms;
most studies to date have measured concentrations of PBDEs in tissues of aquatic organisms to estimate
body burdens. BDE-209 and the lower brominated congeners BDE-206, BDE-207, and BDE-208 were
experimentally shown to accumulate in the liver of juvenile lake whitefish (Coregonus clupectformis) as a
result of exposure to BDE-209 via the diet (Kuo et al., 2010). Fish exposed to 2 ug/gram diet BDE-209
for 30 days accumulated a mean of 5.80 nmol/gram lipid in the liver, compared to 0.208 nmol/gram lipid
for the rest of the body and 0.183 nmol/gram lipid in the liver of control fish. Liver concentrations of
H-54
-------
BDE-206, BDE-207, and BDE-208 also were higher compared to control, although concentrations were
less than 0.01 nmol/gram lipid for each congener (Kuo et al.. 2010).
Similarly, juvenile lake trout (Salvelinus namaycush) exposed to various PBDE congeners for
56 days accumulated measurable PBDE concentrations, and depuration half-lives ranged from 26 to
346 days (Tomy et al.. 2004). When BDE-209 was present in the diet at 3.4 ng/gram dry weight, the
uptake rate constant was calculated as 132 grams/day. When the concentration of BDE-209 in the diet
was raised to 27.5 ng/gram dry weight, the uptake rate constant was much lower—6.1 grams/day. Study
authors determined a half-life for BDE-209 of 26 ± 5 days, and a BMP of 0.3 (Tomv et al.. 2004). BMFs
of other congeners ranged from 1.6 to 45.9.
Body burdens of PBDEs in aquatic organisms have been studied in top predators such as
piscivorous birds of prey and top-level fish and in lower level organisms like insects and crabs
(Environment Canada. 2010; U.S. EPA. 2010a: Environment Canada. 2006) (see Section E.2 of Appendix
E). Bottom feeders and bivalves often have the lowest PBDE body burdens, eels and higher level fish like
sole and flounder have the highest body burdens, and shrimp have mid-range body burdens (U.S. EPA.
2010a). Studies have shown accumulation of BDE-209 (exceeding 100 ng/gram) in top predators,
including sharks in coastal Florida and marine mammals such as harbor porpoise and white-beaked
dolphin (Environment Canada. 2010). Although BDE-209 likely only accumulates at low levels in lower
trophic-level organisms, biomagnification can lead to relatively greater concentrations in higher trophic
levels.
Bioaccumulation rates of BDE-209 and other PBDE congeners are affected by significant
biotransformation and debromination (Tomy et al., 2004). so determining the level or pattern of PBDE
uptake from the environment and accumulation in biota is difficult. Laboratory-based studies that control
the exposure rates and measure tissue concentrations allow for calculations of uptake rates, depuration
rates, and BMFs. For example, in a study by Kierkegaard et al. (1999). juvenile rainbow trout
(Oncorhynchus mykiss) were exposed to technical-grade decaBDE via diet for 120 days, and then
observed for 71 days post-exposure. The level of decaBDE in the diet ranged from 7.5 to 10 mg/kg body
weight per day. Study authors determined that the total uptake in muscle was between 0.02 and 0.13% of
the exposure level. Fish tissue concentrations of BDE-209 increased over the course of the exposure
period, reaching 38 ng/gram fresh weight in muscle and 870 ng/gram fresh weight in the liver at Day 120.
Concentrations declined during the 71-day depuration period to 9.5 ng/gram fresh weight in muscle tissue
and 30 ng/gram fresh weight in the liver (Kierkegaard et al.. 1999).
Bioavailability and bioaccumulation of BDE-209 are limited by the high molecular weight of
BDE-209 and its strong sorption to soils and sediments (Kierkegaard et al.. 2004). In general, PBDEs can
bioaccumulate and biomagnify in the aquatic food web (Agrell et al.. 2004). Their propensity to
H-55
-------
bioaccumulate and biomagnify depends in large part on their level of bromination. Highly brominated
congeners like BDE-209 have a tendency to sink into aquatic sediments and are a minor congener found
in aquatic biota. Benthic sediments are a major sink for PBDEs, but BDE-209 does not appear to be
readily available to benthic organisms for uptake, although some movement through the food web does
occur (Ciparis and Hale. 2005). To what extent BDE-209 can be transformed in the environment to lower
brominated congeners like BDE-47 and BDE-99 is uncertain; BDE-47 and BDE-99 have been identified
as the congeners frequently found in biota (Watanabe and Sakai. 2003).
Studies analyzing tissue levels and body burdens of PBDEs in organisms having various roles in
a specific ecosystem and food web best illustrate biomagnification, as they quantify the body burdens of
different organisms and relate these to food web relationships. In a study by Law et al. (2006a), authors
illustrated biomagnification of brominated flame retardants (including PBDEs) in Lake Winnipeg, Canada
by determining trophic structure, assessing trophic transfer, and quantifying the magnitude of
biomagnification. Samples of water, sediment, plankton, mussels, and six fish species were collected over
a four-year period and analyzed for whole-body (in invertebrates) or muscle-tissue (in vertebrates)
concentrations of contaminants (see Appendix E. Table E-8). As Table H-6 shows, biomagnification of
various PBDE congeners in individual species predator-prey relationships ranged from very positive
(BDE-209 concentration in emerald shiner was 33 times higher than in zooplankton) to negative (BDE-99
concentration in emerald shiner was 10 times lower than in zooplankton) (Law et al.. 2006a). The general
trend illustrates, however, that higher level predators generally have higher body burdens of PBDEs than
lower level prey. The authors determined a trophic magnification factor (which represents the average
predator-prey transfer through a food web, as opposed to a BMP, which represents a transfer for a single
predator-prey relationship) of 3.7 for total PBDEs in the system, and congener-specific trophic
magnification factors of 5.2 for BDE-47 (tetraBDE), 1.5 for BDE-99 and 3.0 for BDE-100 (pentaBDEs),
and 10.4 for BDE-209 (decaBDE) (Law et al.. 2006a).
H.4.3.4. Exposure Pathways in Terrestrial Systems
Although PBDEs are detected commonly in terrestrial ecosystems, exposure levels are higher for
terrestrial organisms with diets that consist of animals from the aquatic system than for herbivorous
organisms or organisms with diets consisting of animals from the terrestrial system (Mikula and
Svobodova. 2006). As discussed in Section H.3.2. LRT can result in exposure to terrestrial organisms far
from the initial source of release (de Witetal. 2010; Su et al.. 2009; Breivik et al.. 2006; Agrell et al..
2004).
H-56
-------
H.4.3.4.1. Toxicokinetics and Body Burden in Terrestrial Systems
Table H-6.
Predator
Biomagnification factors of select PBDE congeners in an aquatic ecosystem.
Prey BDE-47 BDE-99
System: zooplankton -> emerald shiner -> walleye, burbot3
Walleye
Burbot
Emerald shiner 0.3 1.2
Emerald shiner 0.7 9.5
Emerald shiner Zooplankton 5.2 0.1
System: zooplankton, mussels -> white sucker -> walleye3
Walleye
White sucker
White sucker
White sucker 0.2 2.1
Zooplankton 6.1 0.1
Mussels 3.4 0.1
System: zooplankton, mussels -> goldeye -> walleye3
Walleye
Goldeye
Goldeye
Goldeye 0.2 0.1
Zooplankton 7.2 6.5
Mussels 4 4.4
System: zooplankton -> whitefish -> walleye3
Walleye
White fish
White fish
Whitefish 8.9 1.7
Emerald shiner 0.1 0.7
Zooplankton 0.2 0.1
BDE-100 BDE-153 BDE-209
0.2 0.3 0.6
1 1.7 2.4
2.2 1.2 33
0.1 0.2 2
3.4 2.2 9.9
2.9 1.5 0.2
0.1 0.1 0.6
4.9 5.5 34
4.2 3.9 0.8
3.9 4.6 6.8
0.1 0.1 0.1
0.1 0.1 2.9
"Trophic levels: mussel -> zooplankton, whitefish -> goldeye, emerald shiner, white sucker -> burbot, walleye.
Note: Biomagnification factor (BMP) is the lipid-corrected BDE concentration in predators / lipid-corrected BDE concentration in
prey. BMP >1 indicates concentration in predator higher than in prey; BMP <1 indicates concentration in prey higher than in
predator. Gray shading highlights where BMP >1, indicating that biomagnification has occurred as one moves up trophic levels.
Source: Adapted with permission of John Wiley and Sons; Law et al. (2006a).
H-57
-------
Limited information is available on the mechanisms of BDE-209 ADME in terrestrial organisms;
most studies to date have measured concentrations of PBDEs in specific tissues of terrestrial organisms to
estimate body burdens. Body burdens of PBDEs in terrestrial organisms have been studied in top
predators such as piscivorous and carnivorous mammals and birds of prey. DecaBDE studies have shown
high accumulation (exceeding 100 ng/gram) in the liver and muscle tissues of top predators, including
kestrel and sparrowhawk in China, the United Kingdom, and Sweden; peregrine falcon in the United
Kingdom, Sweden, and Greenland; and buzzard and red fox in Belgium (Environment Canada. 2010).
An EPA (20K)a) review reported total PBDE concentrations in tissues of predatory birds ranged from
below detection limits in some tissue types to greater than 12,000 ng/gram weight in some muscle and
liver tissues (see Appendix E. Table E-7).
As previously discussed, the high molecular weight and strong sorption of BDE-209 to soils and
sediments would suggest that bioavailability and bioaccumulation are limited in terrestrial systems. Body
burdens of BDE-209, however, have been identified in some—but not all—terrestrial organisms,
suggesting that biomagnification does sometimes occur. The complexity of food web interactions that
cross aquatic and terrestrial systems makes it challenging to determine whether the source of BDE-209 in
terrestrial food webs stems from contaminated abiotic media in the aquatic environment (e.g., sediments,
interstitial waters), the terrestrial environment (e.g., soils, pore water), or a combination of both.
H.4.4. Aggregate Exposures
Assessing aggregate exposures involves characterizing exposures to a single chemical across
multiple exposure routes. Due to the range of applications for which decaBDE can be used, release from
multiple products and subsequent exposure via multiple routes is anticipated.
BDE-209 and the other PBDEs are ubiquitous in the environment (Paso et al.. 2010).
The average daily intake of PBDEs from various routes has been evaluated, and inhalation of dust
provides the highest contribution to body burden of PBDEs. As discussed in Section H.4.1.2.5. levels of
BDE-209 in household dust are orders of magnitude higher than in other matrices, by weight. BDE-209 is
also the main PBDE contaminant in household dust (Paso et al.. 2010). Data are not available on the
relative contribution to household dust of BDE-209 from textile sources.
Measurable levels of BDE-209 are found in various types of food worldwide, including milk,
fish, shellfish, eggs, beef, chicken, cheese, butter, and other dairy products. The highest concentrations in
food have been reported for cod liver, with fish generally making up the highest dietary source of
BDE-209 (Paso etal.. 2010; Frederiksen et al.. 2009).
H-58
-------
As a result of exposure from various sources, BDE-209 has been detected in breast milk, serum
samples, umbilical cord blood, and the placenta of humans (Paso et al., 2010; Frederiksen et al., 2009).
Ingestion of food, ingestion of dust, inhalation of dust, and dermal contact with soil and dust are the
known pathways by which humans are primarily exposed to BDE-209. One review estimates that
exposure to BDE-209 from ingestion and dermal contact with soil and dust represents more than 29% of
total PBDE exposure from these exposure routes, and that exposure through those routes accounts for
82% of total PBDE exposure from all routes (Lorber. 2008).
HAS. Cumulative Exposures
As stated in The Exposure Factors Handbook (U.S. EPA, 2011). "Cumulative exposure is defined
as the exposure to multiple agents or stressors via multiple routes." For the purpose of this case study, the
"multiple agents or stressors" considered to contribute to cumulative exposure include those substances
that are produced or released as a result of the product life cycles of decaBDE flame-retardant upholstery
textile coatings, facilitate uptake of decaBDE into humans and biota, are taken up as a result of decaBDE
exposures, or induce effects in humans or biota through a comparable or synergistic mode of action.
As discussed in Section H.4.4 on aggregate exposure, PBDEs are ubiquitous in the environment
due to their widespread use and physicochemical characteristics. The lower brominated congeners can be
metabolites of higher congeners such as BDE-209, and subsequent exposure to lower congeners is likely
when BDE-209 is released to the environment. Like BDE-209, many of the lower brominated congeners
such as BDE-47 and BDE-99 can bioaccumulate, and the lower brominated congeners generally are more
toxic than BDE-209. Exposure to the lower brominated metabolites of BDE-209 is also likely following
environmental degradation or aging of composites or textiles containing BDE-209 (Lagalante et al.. 2011;
Christiansson et al.. 2008).
Manufactured textiles treated with decaBDE could include impurities from the synthesis process
such as PBDDs and PBDFs (Ren et al.. 2011). Breakdown or aging of flame-retardant textiles could
contribute trace amounts of these pollutants to the environment. In addition, disposal or incineration of
these textiles might generate more impurities.
Synergistic and antagonistic reactions have been observed in composites. Antimony compounds,
which are typically used in the formulation of flame retardants containing decaBDE, tend to act
synergistically with halogenated flame retardants to produce highly corrosive hydrogen chloride gas or
hydrogen bromide gas, for example (Textile Exchange, 2012). These exposures likely would be limited to
manufacturing activities. The decomposition or incineration of manufactured textiles, however, might
release other gases that could cause synergistic reactions.
H-59
-------
H.5. Potential Human Health, Ecological, and Other Impacts
The final step of compiling information into the CEA framework is to link the information
described in the previous chapters on the product life cycle; transport, transformation, and fate; and
exposure-dose with potential impacts to receptors. The CEA framework includes information relevant to
impacts on human health and ecological receptors, similar to what might be investigated in traditional risk
assessment processes, as well as other plausible impacts that might be considered in life-cycle-focused
assessments (e.g., socioeconomics, climate change, resource depletion).
Section H.5.1 discusses potential impacts of exposure to decaBDE, and related contaminants on
human health. This section relies heavily on evidence from experimental studies with laboratory animals,
the results of which could be extrapolated to humans using established quantitative toxicity assessment
techniques. As discussed in Section H.4. humans could be exposed to decaBDE or related contaminants
from flame-retardant upholstery textiles through a variety of pathways, reaching receptors through dermal
deposition, oral ingestion, or inhalation of these contaminants. This section discusses potential health
impacts from these exposure routes; data are grouped to illustrate the types of impacts (e.g., pulmonary
toxicity, skin irritation, reproductive effects) observed in studies with laboratory animals exposed to
decaBDE and sub-grouped by exposure routes for each impact.
Section H.5.2 discusses the potential impacts of environmental media contaminated with
decaBDE on ecological health, which encompasses impacts on the organism, population, and ecosystem
levels. This section is therefore approached from an ecosystem perspective (aquatic vs. terrestrial), and
data on groups of organisms within those ecosystems are summarized. The discussion of impacts to
ecological health focuses on identifying and comparing data on exposure levels that might cause
significant mortality, delayed growth or development, reproductive defects, or other impacts that could
alter community structure and potentially cause ecosystem collapse.
Finally, Section H.5.3 discusses other plausible impacts resulting from the product life cycles of
decaBDE in flame-retardant upholstery textiles. The section includes a consideration of the energy input
requirements for synthesis of decaBDE, the economic impacts related to the cost of material production,
and the potential for disproportionate impacts on populations with lower socioeconomic status.
H.5.1. Human Health Effects
This section discusses the potential human health effects resulting from exposures to decaBDE.
As noted in Section H.4. exposure to decaBDE from aggregate sources is likely; no studies were found
that investigate impacts to human health that can be attributed directly to exposure to decaBDE or related
H-60
-------
compounds released during the life cycles of decaBDE flame-retardant upholstery textile coatings.
Toxicology studies presented for decaBDE generally were conducted using BDE-209. As discussed in
Section H.3 (see Text Box H.3-1). environmental degradation and debromination of decaBDE results in
contamination of media with lower PBDE congeners. These lower PBDEs have toxicological relevance,
as they are more bioavailable than decaBDE and potentially more toxic; toxicity of PBDEs generally
decreases with increased number of bromine atoms (Rahman et al. 2001).
Toxicology studies conducted on animals comprise much of the information discussed in this
chapter because studies on humans in the literature are limited. Effects observed in animal studies are
typically extrapolated to humans when conducting quantitative toxicity assessments (e.g., when
calculating an RfD or RfC; see Section H.4.2.5). Potential health effects associated with all routes of
exposure (dermal, inhalation, and oral) are presented in this section because each is plausible for humans
(see Section H.4 for additional exposure scenario information).
Dermal and oral exposures to decaBDE in dust seem to be the primary routes of exposure for
consumer populations (see Section H.4.2.2). Because of higher levels of decaBDE contamination in dust
and on other particles, the oral and dermal routes might also be expected to be prominent for general
public exposures (see Section G.4.1.1). Available data for decaBDE indicate that the inhalation exposure
route appears to dominate for workers (see Section H.4.2.1). Although inhalation is a possible route of
exposure, especially for workers, many inhalation toxicology studies identified were conducted by
administering the test material (decaBDE) via intratracheal instillation and pharyngeal aspiration; these
routes of administration require an invasive delivery of chemicals or particles and are not as
physiologically relevant for risk assessment purposes, but could provide biological information useful for
qualitative, mode-of-action determinations.
Available information on these exposure routes in experimental animal studies is grouped by the
main types of health impacts observed in the literature, namely, in vivo and in vitro data on systemic
toxicity, pulmonary toxicity, eye irritation, skin irritation, reproductive effects, developmental effects,
immune system effects, genotoxicity/mutagenicity, carcinogenicity, and susceptible populations.
Toxicology studies were reviewed and determined to be key if the following criteria were met:
• appropriate species and test system were used,
• appropriate dose levels were used,
• route of exposure was appropriate for humans,
• control groups were appropriate, and
• the study was consistent with standard principles and practices.
H-61
-------
In some cases, multiple studies investigating the same endpoint were available, and the most
robust study or the study that most closely aligned with current guidelines for toxicity testing was chosen
as key. In other cases, no studies were available that met all the criteria provided above; available studies
were then summarized with deficiencies noted in the text. If the study was considered key, a written
summary was included in the appropriate section of this appendix. Key studies and supporting non-key
studies are summarized in Table F-3 through Table F-ll in Appendix F. Because a large amount of
published data is available for decaBDE, key studies presented in text are primarily those summarized by
reviews or agency reports, and only a representative subset of studies are included in Appendix F.
The paragraph that follows (see Table 5-1 in Chapter 5) provides an overview of the findings for
human health effects of decaBDE, after which a detailed discussion of the available data is presented.
Most toxicological studies for decaBDE involve the oral route of exposure (see Section H.4.2.2
for discussion on why the oral route appears to be a primary exposure pathway for decaBDE), with
thyroid and liver changes observed in rats and mice in subchronic and chronic studies (NTP, 1986; Norris
etal.. 1975). Several studies (Johansson et al.. 2008; Viberg et al.. 2008; Viberg et al.. 2007; Tseng et al.
2006; Viberg et al., 2003) also reported effects of neonatal exposure, including changes in sperm
parameters (Tseng et al., 2006) and changes in locomotor activity or altered expression of proteins in the
central nervous system (Johansson et al., 2008; Viberg et al., 2008; Viberg et al., 2007; Viberg et al..
2003). EPA calculated an RfD of 0.007 mg/kg-day in 2008 (U.S. EPA. 2008b) based on the
developmental neurobehavioral effects observed in the Viberg et al. (2003) study (see Section H.5.1.1.1
for details on RfD derivation).
With regard to carcinogenicity, the National Toxicology Program (NTP) stated that there was
"some evidence of carcinogenicity" for male and female rats based on significantly increased incidences
of neoplastic nodules of the liver, and "equivocal evidence of carcinogenicity" for male mice based on a
significantly increased incidence of hepatocellular tumors in only the low-dose group and nonstatistically
significant increases in thyroid follicular cell tumors in both dose groups (NTP. 1986). Additionally, the
International Agency for Research on Cancer determined that decaBDE is not classifiable as a human
carcinogen (Group 3) based on limited evidence in animals (IARC. 1998). In 2008, EPA used the
descriptor "suggestive evidence of carcinogenic potential" for decaBDE (U.S. EPA. 2008b) under
relevant guidelines (U.S. EPA. 2005b) (see Section 5.1.11).
H.5.1.1. Quantitative Toxicity Assessment
In a quantitative toxicity assessment, appropriate toxicity information is collected and evaluated.
These data then are used to derive toxicity values, such as an RfD for oral exposure or a reference
concentration (RfC) for inhalation exposure. Similar to an RfD (as defined in Section H.4.2.5). an RfC is
H-62
-------
an estimate of a continuous inhalation exposure for a given duration to the human population (including
susceptible subgroups) that is likely to be without an appreciable risk of adverse health effects over a
lifetime. Both values, an RfC and an RfD, are derived from a benchmark dose lower confidence limit, no-
observed-adverse-effect level (NOAEL), a lowest-observed-adverse-effect level (LOAEL), or another
suitable point of departure, with uncertainty/variability factors applied to reflect limitations of the data
used. Other types of toxicity values also can be derived to provide exposure limit values for other
exposure durations (e.g., acute or subchronic), more specific populations (e.g., healthy workers), or
specific exposure contexts (e.g., emergency response or occupational exposure; see Section H.4.2.5).
The sections that follow discuss the derivation of an RfD for decaBDE; due to limited data, an RfC for
decaBDE has not been determined.
H.5.1.1.1. Health Reference Values
As mentioned in Section H.4.2.5. EPA (2008b) calculated an RfD of 0.007 mg/kg-day, based on
developmental neurobehavioral effects observed in the Viberg et al. (2003) study (see Section H.5.1.7).
The NOAEL of 2.22 mg/kg from this study was used as the point of departure. A total uncertainty factor
of 300 was applied to account for interspecies differences (10*), intraspecies differences (10*), and
dosing duration (3x). As mentioned previously, due to the limited toxicity data available, an RfC for
decaBDE has not been determined.
H.5.1.2. Systemic Toxicity
H.5.1.2.1. Acute
Acute toxicity studies for all routes of exposure were identified for decaBDE. Results indicate
that, for all routes of exposure, decaBDE exhibits low acute toxicity. No mortality occurred after a 1-hour
inhalation exposure in rats (5 animals/sex/group) to 200 mg/L BDE-209; no gross pathological changes
were observed during the 2-week observation period (CPTC. 1978).
The low acute oral toxicity of decaBDE (Zhou etal. 2001; Kierkegaard et al.. 1999; Norris et al..
1973) might be due in part to poor gastrointestinal absorption. No clinical signs of toxicity or death were
observed when a single dose (up to 5,000 mg/kg) of BDE-209 was administered to rats via gavage (Great
Lakes Chemical Corporation. 1994; IRDC. 1974).
The low acute dermal toxicity of decaBDE is presumed based on lack of treatment-related effects
in rabbits following single administrations of 200 or 2,000 mg/kg BDE-209 to clipped intact skin for
24 hours (14-day observation period) (Great Lakes Chemical Corporation. 2000b: IRDC. 1974).
H-63
-------
H.5.1.2.2. Subchronic
In humans, an increase in primary hyperthyroidism and a significant reduction in calf sensory and
fibula motor nerve velocities were observed in workers exposed to decaBDE during manufacturing
[(Bahn et al. (1980) as cited in NTP (1986): Bialik (1982), as cited in HSDB (2011)1: whether these
effects are due to decaBDE or polybrominated biphenyls is unclear, however, because only
polybrominated biphenyls were detected in blood.
Numerous subchronic oral studies were identified for decaBDE, and all studies considered,
including those summarized below, are presented in F.I.2 in Appendix F. DecaBDE-related thyroid and
liver changes were observed in male rats when administered a lower purity (77.4%) form of decaBDE
(Norris etal.. 1975: Norris et al.. 1973). Effects included thyroid hyperplasia, increased liver weight, and
hepatic centrilobular cytoplasmic enlargement and vacuolation (Norris et al., 1975: Norris et al., 1973).
H.5.1.2.3. Chronic
Numerous chronic oral studies were identified for decaBDE, and all studies considered, including
those summarized below, are presented in Section F.I.2 in Appendix F. In a chronic study conducted by
NTP (1986). a dose-dependent increase in thyroid follicular cell hyperplasia was observed in male mice
fed BDE-209 (purity 94-97%) in the diet for 103 weeks (NTP. 1986): these effects were not observed in
female mice or female and male rats that were similarly exposed to BDE-209. Centrilobular hypertrophy
(consisting of enlarged hepatocytes with frothy vacuolated cytoplasm) also was observed in male mice,
but not in female mice or in male and female rats (NTP. 1986). Incidences of thrombosis and
degeneration of the liver were increased in male rats at the LOAEL (2,240 mg/kg-day), but not at
1,120 mg/kg-day (NOAEL); these hepatic effects were not observed in female rats or in mice of either
sex. The NTP (1986) studies were considered for the basis of the EPA (2008b) quantitative cancer
assessment (see Section 5.1.11). Observed changes in liver weight and hepatocytomegaly might have
been due to enzyme induction, as supported by recent studies conducted by Van der Ven et al. (2008) and
Bruchajzer et al. (2010) (see Table F-7 in Appendix F).
H.5.1.3. Pulmonary Toxicity
When BDE-209 was administered to rats via a single intratracheal injection of 20 mg BDE-209
dust (purity 77.4%) suspended in rat serum (Dow Chemical Co. 1990b). minimal histopathological
changes (scattered focal aggregates of alveolar macrophages) occurred, consistent with retention of large
dust particles that would not normally reach the lungs during inhalation. Rats exposed to 2,000 or
48,000 mg/m3 BDE-209 for 1 hour exhibited dyspnea at both dose levels (Great Lakes Chemical
H-64
-------
Corporation. 1994; IRDC. 1974); all animals survived until study termination and were normal at the end
of the 14-day observation period (see Table F-5 in Appendix F).
H.5.1.4. Eye Irritation
All in vivo eye irritation studies considered are presented in Section F.1.2 (Table F-3). Key
studies are summarized below.
DecaBDE does not appear to be an eye irritant. Ocular exposure to dry solid decaBDE caused
transient conjunctival irritation in washed and unwashed rabbit eyes when 100 mg of decaBDE was
administered via instillation to the conjunctival sac (NRC. 2000; IRDC. 1974) [Effects in some rabbits
included very slight conjunctival redness and chemosis, and slight or moderate discharge (Great Lakes
Chemical Corporation. 1994)1. Investigators concluded that the effects were not serious enough to be
considered primary eye irritation (Norris et al.. 1975; IRDC. 1974). Pharmakon (1994) similarly reported
that decaBDE (Saytex 102) did not cause primary eye irritation when instilled once (100 mg/eye) into the
eyes of rabbits. Rats exposed to 2,000 or 48,000 mg/m3 BDE-209 dust in the ambient air for 1 hour,
however, exhibited ocular porphyrin discharge at both dose levels, and eye squint at the high
concentration (IRDC. 1974); all animals survived until study termination and were normal at the end of
the 14-day observation period.
H.5.1.5. Skin Irritation
The material characteristics and study details associated with the in vivo dermal studies
considered for decaBDE and MWCNTs are presented in Section F.1.2 in Appendix F
DecaBDE does not appear to be a skin irritant based on observations from a human skin irritation
study (Dow Chemical Co. 1990a; Norris et al., 1975; Norris et al., 1973). a skin irritation study in rabbits
(Norris et al.. 1975; IRDC. 1974; Norris et al.. 1973). and an acne-genesis study in rabbits (Pharmakon
Research International. 1994). Dermal studies were conducted with BDE-209.
H.5.1.6. Reproductive Effects
In general, studies found that decaBDE was not a reproductive toxicant at doses up to and
exceeding 1,000 mg/kg-day (Tseng et al.. 2008; Hardy et al.. 2002; Dow Chemical Co. 1990c; NTP.
1986). Van der Ven et al. (2008) reported significant, decaBDE dose-related changes in epididymis and
seminal vesicle weight for male rats and decreased activity of CYP17, a key enzyme in the androgen
synthesis pathway, for female rats administered 1.9-60 mg/kg by oral gavage. No corresponding
histopathological changes, sperm counts, or morphology of epididymal sperm, however, were observed.
H-65
-------
Based on these results, the authors concluded that BDE-209 might represent a hazard to
reproductive health.
H.5.1.7. Developmental Effects
Several studies reported no developmental effects for decaBDE at doses up to and exceeding
1,000 mg/kg-day (Hardy et al. 2009; Tseng et al.. 2008; Hardy et al.. 2002; Dow Chemical Co. 1990c).
A significant increase in CYP450 activity in adult male CD-I mouse offspring was noted in the study
conducted by Tseng et al. (2008) at doses of 1,500 mg/kg-day; however, this dose level exceeds the
current dose limit (e.g., 1,000 mg/kg-day) recommended by international toxicity testing guidance
documents (OECD. 2007; U.S. EPA. 1998). Another study found significant increases in numbers of rat
litters with subcutaneous edema and delayed ossification of skull bones at 1,000 mg/kg-day (Norris et al..
1975); dams for this study were administered BDE-209 (77.4% containing 21.8% nonabromodiphenyl
oxide and 0.8% octabromodiphenyl oxide) via gavage at dose levels of 0, 10; 100; or 1,000 mg/kg-day.
Consequently, the NOAEL and LOAEL for fetal effects in this study were 100 and 1,000 mg/kg-day,
respectively; the NOAEL for maternal effects was 1,000 mg/kg-day.
In contrast, several studies did observe adverse effects when neonatal mice or rats were exposed
orally to decaBDE at lower doses (Johansson et al.. 2008; Viberg et al.. 2008; Viberg et al., 2007; Tseng
et al.. 2006; Viberg et al.. 2003). Effects of neonatal exposure included changes in sperm parameters
(Tseng et al.. 2006) and changes in spontaneous behavior or altered expression of proteins in the central
nervous system (Johansson et al.. 2008; Viberg et al.. 2008; Viberg et al.. 2007; Viberg et al.. 2003). Rice
et al. (2007) orally exposed male and female mouse pups to decaBDE at doses of 0, 6, or 20 mg/kg-day
from postnatal days PND2 through PND15. Treatment-related effects occurred only in the high-dose
group and included a reduction in palpebral reflex on PND14, a reduction in forelimb grip in males on
PND16, a change in the slope of the linear trend for serum T4 in males on PND21, and a change in the
linear slope of motor activity on PND70. Of the available studies, Viberg et al. (2003) was selected for
the derivation of the RfD (see Section H.5.1.1.1).
H.5.1.8. Immune System Effects
No immunology studies were identified for decaBDE.
H-66
-------
H.5.1.9. In Vitro Data
In vitro data can be used to make judgments on the toxic potential of stressors, but the relevance
of in vitro data to predicting toxicological responses of "real-world" exposures is not always clear.
No in vitro data were identified for decaBDE.
H.5.1.10. Genotoxicity/Mutagenicity
DecaBDE does not appear to be genotoxic and generally did not induce (1) gene mutations in
bacteria (Salmonella typhimurium) in Ames assays (Chemical Manufacturers Association, 1998; GSRI.
1990: Huntingdon Life Sciences. 1990: NTP. 1986: Haworth et al.. 1983: Litton Bionetics. 1976).
(2) gene mutations in mouse L5178Y lymphoma cells, (3) chromosomal aberrations in mouse bone
marrow cells, or (4) sister-chromatid exchanges or cell transformation in Chinese hamster ovary cells
(Myhr et al.. 1990: McGregor et al.. 1988). These findings are consistent with those reported by EPA
(U.S. EPA. 2008b). Because decaBDE has consistently failed to produce genotoxic or mutagenic
responses, the aforementioned studies are not summarized in Appendix F.
H.5.1.11. Carcinogenicity
Carcinogenicity studies considered for decaBDE are presented in Table F-10 of Appendix F.
The target organs for decaBDE Carcinogenicity appear to be the liver and thyroid; decaBDE was not,
however, included on the most recent U.S. NTP list of carcinogens.36
Information on the Carcinogenicity of decaBDE is available from three chronic feeding studies in
rodents (Kociba et al.. 1994: NTP. 1986: Kocibaetal.. 1975). In the NTP study (1986). a treatment-
related increase in liver neoplastic nodules was observed in low- and high-dose male rats (7/50 and 15/49,
respectively, compared to 1/50 in controls) and high-dose female rats (9/50 compared to 1/50 and 3/49 in
control and low-dose groups, respectively). F344/N rats were fed BDE-209 (94-98% pure) at dietary
concentrations of 0, 25,000, or 50,000 ppm for 103 weeks (equivalent to 0, 1,120, and 2,240 mg/kg-day in
male rats; 0, 1,200, and 2,550 mg/kg-day in female rats). The increase in liver neoplastic nodules was not
accompanied by an increase in hepatocellular carcinomas in rats. Hepatocellular adenomas or carcinomas
(combined), however, were observed in low- and high-dose male mice (8/50 controls, 22/50 low-dose
mice, 18/50 high-dose mice). Male mice also exhibited a marginal increase in thyroid gland follicular cell
36The U.S. NTP 12th Report on Carcinogens (released June 2011) is available at
http://ntp.niehs.nih.gov/?obiectid=03C9AF75-ElBF-FF40-DBA9EC0928DF8B15
H-67
-------
adenomas or carcinomas (combined) at the low and high doses (0/50 controls, 4/50 low-dose mice, 3/50
high-dose mice). The possible significance of this finding was strengthened by increased incidences of
follicular cell hyperplasia in the male mice (2/50 controls, 10/50 low-dose mice, 19/50 high-dose mice),
but was weakened by increased mortality in control animals. Based on these results, the NTP (1986)
study concluded that there was "some evidence of carcinogenicity" for male rats in the low-dose group
and both male and female rats in the high-dose group based on significantly increased incidences of
neoplastic nodules of the liver, and "equivocal evidence of carcinogenicity" for male mice based on a
significantly increased incidence of hepatocellular tumors in only the low-dose group and nonstatistically
significant increases of thyroid follicular cell tumors in both dose groups. Although the International
Agency for Research on Cancer (1998) reports that decaBDE is not classifiable as a human carcinogen
(Group 3) based on limited evidence in animals, EPA, under the Guidelines for Carcinogen Risk
Assessment (U.S. EPA. 2005b). determined that the descriptor "suggestive evidence of carcinogenic
potential" is appropriate for decaBDE (U.S. EPA. 2008b) based on the data from NTP (1986)
demonstrating evidence of carcinogenicity in more than one species, sex, and site.
H.5.1.12. Susceptible Populations
Sacks et al. (2011) defined susceptibility as "individual- and population-level characteristics that
increase the risk of health effects in a population, including, but not limited to, genetic background, birth
outcomes (e.g., low birth weight, birth defects), race, sex, life stage, lifestyle (e.g., smoking status,
nutrition), preexisting disease, socioeconomic status (e.g., educational attainment, reduced access to
health care), and characteristics that may modify exposure ... (e.g., time spent outdoors)." In this section,
populations susceptible to decaBDE impacts based on characteristics such as age, genetic background,
and disease are considered. Characteristics that could modify exposure and increase susceptibility were
discussed previously in Section H.4.2.4; for a discussion on impacts related to socioeconomic status, see
Section H.5.3.
Results regarding developmental neurotoxicity studies are conflicting. Whether young children
comprise a sensitive (i.e., more susceptible) population is therefore unclear. A few animal studies have
indicated that BDE-209 might cause developmental neurotoxicity, affecting motor and cognitive domains;
however, in discussing susceptible populations for decaBDE, EPA noted that differences in the effects of
decaBDE on neurodevelopment are unclear and whether other targets (thyroid and liver) are more
sensitive in children is unknown (U.S. EPA. 2008b). Disposition studies using pregnant rats indicate that
fetuses are less exposed to decaBDE than mothers. Fetuses (whole litter) contained only 0.43% of the
dose in a study by Riu et al. (2008). In a study by Inoue et al. (2006). higher brominated congeners like
decaBDE transferred from blood to milk to a lesser degree than did lower brominated congeners. In a
H-68
-------
study by Fukata et al. (2005). BDE-209 was not detected in umbilical cord tissue, but was found at
23 ng/gram lipid weight in umbilical cord serum and 10 ng/gram lipid weight in maternal serum.
Exposure to decaBDE in infants, however, appears to be greater than in adults. Additional information
regarding populations that are susceptible to greater levels of exposure can be found in Section H.4.
No sources indicating that specific genetic polymorphisms increase susceptibility were identified.
H.5.2. Ecological Effects
This section presents a summary of data on the potential ecological impacts of environmental
contamination with decaBDE. Specific information from the studies reviewed for this case study can be
found in Section F.2 in Appendix F. Considerations for ecological impact include the absolute and
relative toxicity of the decaBDE and other factors such as bioaccumulation and biomagnification potential
(see Section G.4.2.1). For aquatic ecosystems, much information was available for decaBDE primarily
because decaBDE has been studied extensively in aquatic vertebrates. Conversely, little information was
identified on the potential effects of decaBDE in terrestrial ecosystems. The terrestrial ecosystem studies
focus on agriculturally relevant plants and soil microbes. In both aquatic and terrestrial ecosystems,
studies are predominantly laboratory-based experiments on single species. Few studies address how
exposure and uptake of PBDEs relate to ecological health and effects in the field (Vonderheide et al..
2008). Some field studies show correlations between PBDE exposure, reproductive behavior, and
immunosuppression, but ecological consequences and potential population-level impacts of
environmental PBDE contamination in general, and decaBDE in particular, remain uncertain
(Vonderheide et al.. 2008). As mentioned in Section G.4.2.1. ecological receptors can be exposed to
decaBDE attached to textile fibers, embedded in polymers, or sorbed to other particles, all of which are
more likely to occur in the environment than exposure to the pristine compound. Studies examining
exposure to larger textile scraps, polymer particles, and other heterogeneous compounds containing
decaBDE, however, are lacking. The results of laboratory studies using pristine compounds must
therefore be considered, recognizing that results might not translate directly into real-world exposure
scenarios.
As discussed in Section H.3 (see Text Box H.3-1). environmental degradation and debromination
of decaBDE result in contamination of media with lower PBDE congeners, which are generally more
bioavailable and more toxic than decaBDE. Also as stated previously in Section H.3. debromination of
decaBDE is expected to contribute significantly to the environmental presence of BDE-47, atetraBDE,
and BDE-100 and BDE-99, which are both pentaBDEs, among other congeners (Gandhi et al.. 2011;
Ross et al.. 2009). In this section, information is presented on the acute toxicity of decaBDE and other
H-69
-------
PBDEs, as appropriate. Considerations for negative ecological impacts from continual long-term
exposures in an ecosystem are also discussed.
H.5.2.1. Aquatic Receptors
DecaBDE has been shown not to be acutely toxic to species offish or marine algae studied to
date (Hardy, 2002a). Few studies were identified regarding the acute or chronic toxicity of decaBDE to
marine or freshwater algae and benthic invertebrates (see Section H.5.2.1.1). No information was
identified regarding toxic effects on aquatic plants or water-dwelling invertebrates (see Section
H.5.2.1.1): some information was identified on toxicity to aquatic vertebrates (see Section H.5.2.1.2).
DecaBDE is not expected to be chronically toxic to aquatic organisms at environmentally
relevant concentrations due to its physicochemical properties, specifically high molecular weight and low
water solubility (Hardy. 2002a) (see Section H.I). These properties suggest that decaBDE accumulation
directly from water into biota is unlikely; a more likely route of exposure is dietary (Gandhi et al.. 2011).
But as noted previously, the factors affecting bioavailability of decaBDE are not well understood.
Comparatively, however, lower PBDEs such as pentaBDEs are known to have high potential for
bioaccumulation (U.S. EPA. 2010a). Because these congeners are transformation products of decaBDE,
their chronic toxicity is considered.
H.5.2.1.1. Algae, Aquatic Plants, and Aquatic Invertebrates
The paragraphs that follow describe literature identified for the effects of decaBDE on algae,
aquatic plants, and aquatic invertebrates. Table F-12. Table F-14. and Table F-15 in Appendix F
summarize details of the studies identified and reviewed for this section.
No studies were identified that investigated the effects of decaBDE on algae or on aquatic plants.
A single study on water-dwelling aquatic invertebrates was identified involving freshwater bivalve zebra
mussels (Dreissenapolymorpha) exposed to technical-grade decaBDE at sublethal levels of 0.1, 2, or
10 ug/L. This study showed DNA damage that increased as levels of decaBDE exposure increased,
indicating potential for genotoxicity (Rivaet al.. 2007). A review by Hardy (2002a) stated that decaBDE
was nontoxic to marine algae and sediment oligochaetes, but no details on the derivation of these
conclusions were provided. The review also investigates toxicity of octaBDE and pentaBDE, concluding
that octaBDE is neither acutely toxic nor chronically toxic to a species of water flea, and pentaBDE is not
acutely toxic to algae, up to the limit of their water solubility (Hardy. 2002a). Details on endpoints
observed were not provided.
Environment Canada (2006) reviewed ecotoxicity studies for multiple PBDE mixtures. They
report high (>5,000 mg/kg) no-observed-effect levels and median (>50 mg/kg) effective concentration
H-70
-------
values for chronic survival and reproduction effects for a freshwater oligochaete (Lumbriculus variegatus)
exposed to sediments that contained a mixture of 55% pentaBDE and 36% tetraBDE as well as a mixture
containing 97% decaBDE. For water fleas (Daphnia magna), Environment Canada (2006) reported
toxicity values in the low ug/L range for survival, growth, and reproduction following chronic exposure
to a commercial pentaBDE mixture [(Drottar and Krueger (1998) as cited in Environment Canada
(2006)]. Some water fleas are therefore more sensitive to PBDEs than oligochaete worms, but as
mentioned in Section H.3.3. decaBDE is not likely to remain in the water column; instead, it partitions to
sediment, where benthic invertebrates are expected to be exposed.
H.5.2.1.2. Aquatic Vertebrates
Table 5-3 in Chapter 5 describes key toxicity values identified for the effects of decaBDE and
MWCNTs on aquatic vertebrates. Table F-13, Table F-16, and Table F-17 in Appendix F summarize
details of the studies identified and reviewed for this section.
A review by Hardy (2002a) reports on acute toxicity of decaBDE, octaBDE, and pentaBDE,
stating that all three congeners have a fish 48-hour median lethal concentration of greater than 500 mg/L,
indicating that the congeners are not acutely toxic to fish up to the limit of their water solubility.
No effects on egg mortality were observed in rainbow trout (Oncorhynchus mykiss) at doses up to 12 ug
pentaBDE per egg, and no effects on reproduction or spawning success were observed in three-spined
stickleback (Gasterosteus aculeatus) exposed to pentaBDE (Hardy. 2002a).
Tests conducted on frogs often measure low-dose, chronic thyroid disruption, because
metamorphic development from tadpole to frog is controlled by thyroid hormones (Qin et al.. 2010).
African clawed frog (Xenopus laevis) tadpoles (Table 5-3 in Chapter 5) exposed to decaBDE in their
water at sublethal doses ranging from 1 to 1,000 ng/L experienced histopathological alterations in thyroid
gland cell shapes and decreases in thyroid hormone expression in tail tissue during metamorphosis at all
tested doses. Additionally, researchers observed a concentration-dependent trend of delay in time to
metamorphosis with a statistically significant delay at 1,000 ng/L (Qin etal. 2010).
Endocrine effects also have been studied in Chinese rare minnow (Gobiocypris rarus) and lake
trout (Salvelinus namaycush) chronically exposed to decaBDE via water and diet, respectively ely (Li et
al., 2011; Tomy et al., 2004). Expression of thyroid hormone-related genes was variably affected in both
studies; indicating the potential for chronic endocrine disruption but not elucidating a mechanism for
those effects or a clear effect level. Chronic toxicity of decaBDE in fish is complicated by biotic
debromination of decaBDE, which can result in bioaccumulation of octa-, hepta-, hexa-, and pentaBDE
congeners (Gandhi et al., 2011; Stapleton et al.. 2004) (as discussed in Section H.4.2.6.1). Most
informative, therefore, are studies of multiple PBDEs or PBDE mixtures. In one such study, induction of
H-71
-------
vitellogenin production was observed in hepatocyte cell cultures of rainbow trout (Oncorhynchus mykiss)
exposed to PBDE mixtures (Nakari and Pessala. 2005).This estrogenic response raises concerns for
population dynamic impacts due to endocrine disruption (Mikula and Svobodova. 2006).
H.5.2.2. Terrestrial Receptors
A limited amount of information was found regarding toxicity of decaBDE to soil microbes,
plants, and terrestrial invertebrates (see Sections H.5.2.2.1 and H.5.2.2.2). No information was identified
regarding toxicity to terrestrial vertebrates (see Section H.5.2.2.3): nevertheless, some assumptions can be
made for mammals based on toxicity studies intended for human health purposes presented in Section
H.5.1. Many studies of decaBDE in terrestrial ecosystems have focused on bioaccumulation and
biomagnification; important considerations for potential ecological hazard (see Section H.4.3.4). These
studies, however, did not investigate occurrence of toxic effects.
H.5.2.2.1. Soil Microbes and Terrestrial Invertebrates
Changes in soil microbial activity result in changes to nutrient cycling; therefore, studying the
impact of contaminants on soil microbes can provide insight on how those contaminants might affect
ecosystem function (Chung etal., 2011). Similarly, effects on terrestrial invertebrates, such as worms, can
influence health and fertility of a soil ecosystem (Xie et al., 2011). Table F-18 and Table F-19 in
Appendix F summarize details of the studies identified and reviewed for this section.
Two studies were identified that investigated the toxicity of decaBDE to soil microbes. Although
Sverdrup et al. (2006) showed no effects on nitrifying ability of bacteria following exposure to decaBDE
at levels up to 2,274 mg/kg in soil, Liu et al. (2011 a) found that microbial cytotoxicity significantly
increased at doses 10-fold lower (100 mg/kg), and community structure was altered following long-term
exposure to decaBDE. The rate of community diversity increase over time was significantly slower from
Day 90 through the last day of the study (Day 180) when soil contained 1-100 mg/kg decaBDE. After six
months, the total bacterial count in the soil containing 100 mg/kg decaBDE was approximately half that
of the control plot. Treatment soil microcosms were dominated by Pseudomonas, Bacillus, and
uncultured bacteria types, and had significantly reduced cell counts for alpha, beta, and gamma type
proteobacteria and the Cytophaga-Flavobacterium-Bacteroides group (Liu et al.. 201 la).
Studies of oligochaete worms \Enchytraeus crypticus (a soil worm) and Eiseniafetida
(earthworms)] showed that survival, reproductive behavior, and number of offspring are not affected by
long-term exposure to decaBDE in soil in the grams/kg range [Sverdrup et al. (2006); ACC (2001) as
cited in Environment Canada (2006)1. Earthworms, however, experienced a sublethal, dose-dependent
increase in hydroxyl radical generation and subsequent oxidative stress after 1 week of exposure to 0.1-
H-72
-------
10 mg/kg decaBDE (Xie etal.. 2011). Oxidative stress in earthworms is considered a biomarker
indicative of potential for greater impacts of soil contaminants within terrestrial ecosystems (Xie et al.,
2011).
H.5.2.2.2. Terrestrial Plants
Table F-18 and Table F-20 in Appendix F summarize details of the studies identified and
reviewed for this section.
Few studies were identified that investigated effects of PBDEs on plants; those reviewed found
no adverse effects at environmentally relevant concentrations. No effects on seedling emergence were
observed in red clover (Trifolium pretense) exposed to decaBDE or corn (Zea mays) exposed to a PBDE
mixture (55% pentaBDE and 36% tetraBDE) at levels in the grams/kg range (Sverdrup et al. 2006; Great
Lakes Chemical Corporation, 2000a). In corn, mean shoot height was unaffected at concentrations up to
125 mg/kg, but was significantly reduced at 250 mg/kg and above [Great Lakes Chemical Corporation
(2000a) as cited in Environment Canada (2006)1. As discussed in Section H.3.4. soils are a major sink for
PBDEs in terrestrial systems, and uptake by plants is possible. Recent measured concentrations in soil
have been in the ng/gram (0.001 mg/kg) range (see Table E-5 in Appendix E).
H.5.2.2.3. Terrestrial Vertebrates
No studies were identified that specifically investigated the effects of decaBDE on terrestrial
vertebrates outside the laboratory setting. Results from extensive testing performed in mammals for
toxicological relevance to humans are reported in Section H.5.1.
Chronic effects of PBDEs at environmentally relevant exposure concentrations are a possibility in
terrestrial vertebrates, primarily due to assumed ecological impacts associated with high biomagnification
rates, as discussed previously in Section G.4.2.1.
H.5.3. Other Impacts
As stated in Section H.I, the CEA framework considers not only human and ecological health
impacts, but also aesthetic, environmental, social, legal, ethical, and economic impacts. Such impacts
might be associated with impacts on specific socioeconomic sectors (e.g., disparate impacts on
environmental justice communities), the environment as a whole (e.g., climate change, depletion of
natural resources, energy demand), or the built environment (e.g., damage to building facades).
Apart from the impacts discussed in Sections H.5.1 and H.5.2. the only other impacts considered
in this case study are those for which a plausible premise can be developed to support assumptions that a
H-73
-------
discernible impact might occur as a result of the life cycles of decaBDE flame-retardant upholstery textile
coatings. Empirical data have revealed a correlation between decaBDE body burdens and socioeconomic
status, indicating that effects having environmental justice implications are plausible for decaBDE.
H.5.3.1. Environmental Justice
Environmental justice is defined by EPA as the "fair treatment and meaningful involvement of all
people regardless of race, color, national origin, or income with respect to the development,
implementation, and enforcement of environmental laws, regulations, and policies."37 The goal of
environmental justice is to give all people "...the same degree of protection from environmental and
health hazards and equal access to the decision-making process..."38 As a result, environmental justice
impacts include those in which a particular group or geographic area experiences a disproportionate share
of the impacts associated with an environmental contaminant.
Releases of decaBDE throughout the life cycle of a flame-retardant upholstery coating product
could disproportionately impact certain communities. In a review by Zota et al. (2010). findings from
several recent studies suggest that racial and ethnic minorities and populations having lower
socioeconomic status (i.e., low income, low educational attainment) experience disproportionate
exposures to PBDEs. For example, Rose et al. (2010) observed that body burdens of BDE-209, among
other congeners, were significantly higher in children aged 2-5 years born to mothers of lower
educational attainment compared to those born to mothers achieving a college degree or higher.
The causal pathway connecting low socioeconomic status to elevated PBDE exposure is not well
understood, but Zota et al. (2010) hypothesized that furniture quality and the characteristics of the living
spaces (e.g., size, ventilation, age), which populations of lower socioeconomic status might occupy,
contribute to elevated exposure to PBDEs. Indeed, Rose et al. (2010) demonstrated that higher maternal
education attainment is correlated with larger living spaces, and in turn, children living in larger homes
had lower body burdens of BDE-209. Similarly, Stapleton et al. (2012) found that variation in PBDE
serum concentrations in children could be explained by handwipe levels, house dust levels, father's
education, breast feeding duration, age, and gender (different factors associated with different PBDE
congeners).
37U.S. EPA Compliance and Enforcement. Environmental Justice, http://www.epa.gov/environmentaljustice/
38ibid
H-74
-------
H.5.3.2. Energy Demand and Natural Resource Depletion
No information was identified that examined impacts on energy demand and natural resource
depletion associated with the production of decaBDE flame-retardant textile coatings.
H.5.3.3. Climate Change
No information was identified that examined climate change impacts due to decaBDE flame-
retardant textile coatings.
H.5.3.4. Economics
No information was identified that calculated the cost of manufacturing decaBDE or decaBDE
flame-retardant textiles.
Appendix H References
ACC (American Chemistry Council). (2001). Effect of decabromodiphenyl oxide (DBDPO) on the survival and
reproduction of the earthworm, Eisenia fetida. (Study No. 46540). Columbia, MO: ABC Laboratories, Inc.
Agrell. C: ter Schure. AFH: Sveder. J: Bokenstrand. A: Larsson. P: Zegers. BN. (2004). Polybrominated
diphenyl ethers (PBDES) at a solid waste incineration plant I: Atmospheric concentrations. Atmos Environ
38: 5139-5148. http://dx.doi.0rg/10.1016/i.atmosenv.2004.05.024
Ahn. M. -Y: Fillev. TR: JafVert. CT: Nies. L: Hua. I: Bezares-Cruz. J. (2006). Photodegradation of
decabromodiphenyl ether adsorbed onto clay minerals, metal oxides, and sediment. Environ Sci Technol 40:
215-220. http://dx.doi.org/10.1021/es051415t
Allen. JG: McClean. MD: Stapleton. HM: Webster. TF. (2008). Critical factors in assessing exposure to PBDEs
via house dust. Environ Int 34: 1085-1091. http://dx.doi.0rg/10.1016/i.envint.2008.03.006
Allen. JG: Webster. TF: McClean. MD: Stapleton. HM: Nelson. JW. (2007). Personal exposure to
Polybrominated Diphenyl Ethers (PBDEs) in residential indoor air. Environ Sci Technol 41: 4574-4579.
http://dx.doi.org/10.1021/es0703170
Ashley. J: Libero. D: Halscheid. E: Zaoudeh. L: Stapleton. H. (2006). Polybrominated diphenyl ethers (PBDEs)
in American eels from the Delaware River, USA. Partnership for the Delaware Estuary.
http://www.delawareestuary.org
ATSDR (Agency for Toxic Substances and Disease Registry). (2004). Toxicological profile for polybrominated
biphenyls and polybrominated diphenyl ethers [ATSDR Tox Profile]. Atlanta, GA: Agency for Toxic
Substances & Disease Registry, http://www.atsdr.cdc.gov/toxprofiles/tp68.pdf
Batterman. S: Godwin. C: Chernyak. S: Jia. C: Charles. S. (2010). Brominated flame retardants in offices in
Michigan, USA. Environ Int 36: 548-556. http://dx.doi.0rg/10.1016/j.envint.2010.04.008
Bogdal C: Scheringer. M: Schmid. P: Blauenstein. M: Kohler. M: Hungerbiihler. K. (2010). Levels, fluxes and
time trends of persistent organic pollutants in Lake Thun, Switzerland: Combining trace analysis and
multimedia modeling. Sci Total Environ 408: 3654-3663. http://dx.doi.0rg/10.1016/i.scitotenv.2010.04.038
H-75
-------
Breivik. K: Wania. F: Muir. DC: Alaee. M: Backus. S: Pacepavicius. G. (2006). Empirical and modeling
evidence of the long-range atmospheric transport of decabromodiphenyl ether. Environ Sci Technol 40:
4612-4618.
Bruchaizer. E: Frydrych. B: Sporny. S: Szymanska. JA. (2010). Toxicity of penta- and decabromodiphenyl
ethers after repeated administration to rats: a comparative study. Arch Toxicol 84: 287-299.
http://dx.doi.org/10.1007/s00204-009-0495-v
Cai Y: Zhang. W: Hu. J: Sheng. G: Chen. D: Fu. J. (2011). Characterization of maternal transfer of
decabromodiphenyl ether (BDE-209) administered to pregnant Sprague-Dawley rats. Reprod Toxicol 31:
106-110. http://dx.doi.0rg/10.1016/i.reprotox.2010.08.005
CalRecvcle (California Department of Resources Recycling and Recovery). (2002). Examples of product reuse
and recycling. Sacramento, CA.
Cetin. B: Odabasi M. (2005). Measurement of Henry's law constants of seven polybrominated diphenyl ether
(PBDE) congeners as a function of temperature. Atmos Environ 39: 5273-5280.
http://dx.doi.0rg/10.1016/i.atmosenv.2005.05.029
Chang. FH: Yang. CR: Tsai CY: Lin. WC. (2009). Airborne polybrominated diphenyl ethers in a computer
classroom of college in Taiwan. Iranian Journal of Environmental Health Science and Engineering 6: 121-
130.
Charles. MJ: Groskova. D: Cahill TM. (2005). Near-source ambient air monitoring of polybrominated diphenyl
ethers. Davis, CA: California Air Resources Board.
http://www.arb.ca.gov/toxics/pbde%20final%20report%202005.pdf
Chaudhrv. Q: Aitken. R: Hankin. S: Donaldson. K: Olsen. S: Boxall A: Kinloch. I: Friedrichs. S. (2009).
Nanolifecycle : A lifecycle assessment study of the route and extent of human exposure via inhalation for
commercially available products and applications containing carbon nanotubes. York, United Kingdom:
Food and Environment Research Agency.
Chemical Manufacturers Association. (1998). Bacterial reverse mutation assay of decabromo diphenyl oxide,
with cover letter dated 09/14/1998: Final report. Washington, DC.
http://www.ntis.gov/search/product.aspx?ABBR=OTS0559516
Christiansson. A: Eriksson. J: Teclechiel D: Bergman. A. (2009). Identification and quantification of products
formed via photolysis of decabromodiphenyl ether. Environ SciPollutRes Int 16: 312-321.
http://dx.doi.org/10.1007/sll356-009-0150-4
Christiansson. A: Hovander. L: Athanassiadis. I: Jakobsson. K: Bergman. A. (2008). Polybrominated diphenyl
ethers in aircraft cabins~a source of human exposure? Chemosphere 73: 1654-1660.
http://dx.doi.0rg/10.1016/i.chemosphere.2008.07.071
Chung. H: Son. Y: Yoon. TK: Kim. S: Kim. W. (2011). The effect of multi-walled carbon nanotubes on soil
microbial activity. Ecotoxicol Environ Saf 74: 569-575. http://dx.doi.0rg/10.1016/i.ecoenv.2011.01.004
Ciparis. S: Hale. RC. (2005). Bioavailability of polybrominated diphenyl ether flame retardants inbiosolids and
spiked sediment to the aquatic Oligochaete, Lumbriculus variegatus. Environ Toxicol Chem 24: 916-925.
http://dx.doi.0rg/10.1897/04-179r.l
Clarke. B: Porter. N: Symons. R: Marriott P: Ades. P: Stevenson. G: Blackbeard. J. (2008). Polybrominated
diphenyl ethers and polybrominated biphenyls in Australian sewage sludge. Chemosphere 73: 980-989.
http://dx.doi.0rg/10.1016/i.chemosphere.2008.06.034
CPTC (Consumer Product Testing Company, Inc.). (1978). A dermal LD50 study in albino rabbits and an
inhalation LD50 study in albino rats. Fairfield, NJ.
Darnerud. PO: Eriksen. GS: Johannesson. T: Larsen. PB: Viluksela. M. (2001). Polybrominated diphenyl ethers:
Occurrence, dietary exposure, and toxicology [Review]. Environ Health Perspect 109: 49-68.
http://dx.doi.org/10.2307/3434846
H-76
-------
Paso. AP: Fatokj OS: Odendaal JP: Okonkwo. JO. (2010). A review on sources of brominated flame retardants
and routes of human exposure with emphasis on polybrominated diphenyl ethers. Environ Rev 18: 239-254.
http://dx.doi.org/10.1139/alO-010
de Boer. J: Wester. PG: van der Horst A: Leonards. PEG. (2003). Polybrominated diphenyl ethers in influents,
suspended paniculate matter, sediments, sewage treatment plant and effluents and biota from the
Netherlands. Environ Pollut 122: 63-74. http://dx.doi.org/10.1016/s0269-7491(02)00280-4
de Wit CA: Herzke. D: Vorkamp. K. (2010). Brominated flame retardants in the Arctic environment ~ trends
and new candidates [Review]. Sci Total Environ 408:2885-2918.
http://dx.doi.0rg/10.1016/i.scitotenv.2009.08.037
POD (U.S. Department of Defense). (2011). Chemical & material emerging risk alert: Decabromodiphenyl ether
(DecaBDE). (Risk Alert # 02-11). Washington, DC.
Dow Chemical Co (Dow Chemical Company). (1990a). Human skin sensitization study on: FR 300BA
(decabromodiphenyl oxide (IBT #F2366) with cover letter dated 030890 [TSCA Submission].
(86900000182). Baton Rouge, LA: Ethyl Corporation.
http://www.ntis.gov/search/product.aspx? ABBR=OTS0522253
Dow Chemical Co (Dow Chemical Company). (1990b). Pulmonary clearance and tissue response following a
single intratracheal injection of decabromodiphenyl oxide (DBDPO) dust in male rats with attachment &
cover letter dated 030890 [TSCA Submission]. (86900000194). Baton Rouge, LA: Ethyl Corporation.
http://www.ntis.gov/search/product.aspx? ABBR=OTS0522264
Dow Chemical Co (Dow Chemical Company). (1990c). Results of a reproduction study in rats maintained on
diets containing decabromodiphenyl oxide with attachment and cover letter dated 030890 [TSCA
Submission]. (86900000181). Baton Rouge, LA: Ethyl Corporation.
http://www.ntis.gov/search/product.aspx?ABBR=OTS0522252
Drottar. K: Krueger. H. (1998). Pentabromodiphenyl oxide (PeBDPO): A flow-through life-cycle toxicity test
with the cladoceran (Daphnia magna). (439A-109). Easton, MD: Wildlife International Ltd.
ECB (European Chemicals Bureau). (2003). Bis(pentabromophenyl) ether: Summary risk assessment report.
(Special Publication 1.02.78). Luxembourg: Office for Official Publications of the European Communities.
http://echa.europa.eu/documents/10162/Od251331-31ea-4963-aebO-f97ea886d594
el Dareer. SM: Kalin. JR: Tillery. KF: Hill. PL. (1987). Disposition of decabromobiphenyl ether in rats dosed
intravenously or by feeding. J Toxicol Environ Health 22: 405-415.
http://dx.doi.org/10.1080/15287398709531082
Eljarrat E: Labandeira. A: Marsh. G: Raldua. D: Barcelo. D. (2007). Decabrominated diphenyl ether in river fish
and sediment samples collected downstream an industrial park. Chemosphere 69: 1278-1286.
http://dx.doi.0rg/10.1016/i.chemosphere.2007.05.052
Environment Canada. (2006). Canadian Environmental Protection Act, 1999: Ecological screening assessment
report on polybrominated diphenyl ethers (PBDEs).
Environment Canada. (2010). Ecological state of the science report on decabromodiphenyl ether (decaBDE).
http://www.ec.gc.ca/lcpe-cepa/default.asp?lang=En&n=B901A9EB&offset=l&toc=show
EU (European Union). (2002). European Union risk assessment report: Bis(pentabromophenyl) ether. In 1st
Priority List (pp. 1-282). (EUR 20402 EN). Luxembourg: Office for Official Publications of the European
Communities, http://echa.europa.eu/documents/10162/da9bc4c4-8e5b-4562-964c-5b4cf59d2432
Frederiksen. M: Vorkamp. K: Thomsen. M: Knudsen. LE. (2009). Human internal and external exposure to
PBDEs - A review of levels and sources [Review]. Int J Hyg Environ Health 212: 109-134.
http://dx.doi.0rg/10.1016/i.ijheh.2008.04.005
Fukata. H: Omori-Inoue. M: Osada. H: al.. e. (2005). Current status of maternal and fetal exposure to brominated
flame retardants, PCBs and dioxins in Japan. Organohalogen Compounds 67: 16171619.
H-77
-------
Gandhi. N: Bhavsar. SP: Gewurtz. SB: Tomy. GT. (2011). Can biotransformation of BDE-209 in lake trout
cause bioaccumulation of more toxic, lower-brominated PBDEs (BDE-47, -99) over the long term? Environ
Int 37: 170-177. http://dx.doi.0rg/10.1016/i.envint.2010.08.013
Gouin. T: Harner. T: Daly. GL: Wania. F: Mackav. D: Jones. KG. (2005). Variability of concentrations of
polybrominated diphenyl ethers and polychlorinated biphenyls in air: Implications for monitoring, modeling
and control. Atmos Environ 39: 151-166. http://dx.doi.0rg/10.1016/i.atmosenv.2004.09.022
Gouin. T: Thomas. GO: Chaemfa. C: Harner. T: Mackav. D: Jones. KG. (2006). Concentrations of
decabromodiphenyl ether in air from Southern Ontario: Implications for particle-bound transport.
Chemosphere 64: 256-261. http://dx.doi.0rg/10.1016/i.chemosphere.2005.12.071
Great Lakes Chemical Corporation. (1994). Initial submission: Letter from Great Lakes Chemical to USEPA Re:
Tetrabromobisphenol A, Pentabromoethylbenzene, Decabromodiphenyl Ether & Dibromopropyl Aery late
with attachments dated 011184 [TSCA Submission]. (FYI-OTS-0794-1105). West Lafayette, IN.
http://www.ntis.gov/search/product.aspx?ABBR=OTS0001105
Great Lakes Chemical Corporation. (2000a). Pentabromodiphenyl oxide (PeBDPO): A toxicity test to determine
the effects of the test substance on seedling emergence of six species of plants. Easton, MD: Wildlife
International, Ltd.
Great Lakes Chemical Corporation. (2000b). Product and data information on decabromodiphenyl oxide,
octabromodiphenyl oxide and pentabromodiphenyl oxide with attachments. (44-8227036). West Lafayette,
IN. http://www.ntis.gov/search/product.aspx?ABBR=OTS0525626
GSRI (Gulf South Research Institute). (1990). Mutagenecity testing of HFO 102 with cover letter dated 030890
[TSCA Submission]. (86900000196). Baton Rouge, LA: Ethyl Corporation.
http://www.ntis.gov/search/product.aspx? ABBR=OTS0522266
Hakk. H: Letcher. RJ. (2003). Metabolism in the toxicokinetics and fate of brominated flame retardants~a
review [Review]. Environ Int 29: 801-828. http://dx.doi.org/10.1016/80160-4120(03)00109-0
Hale. RC: La Guardia. MJ: Harvey. E: Gaylor. MO: Mainor. TM. (2006). Brominated flame retardant
concentrations and trends in abiotic media. Chemosphere 64: 181-186.
http://dx.doi.0rg/10.1016/i.chemosphere.2005.12.006
Hale. RC: La Guardia. MJ: Harvey. EP: Gaylor. MO: Mainor. TM: Duff. WH. (2001). Flame retardants.
Persistent pollutants inland-applied sludges. Nature 412: 140-141. http://dx.doi.org/10.1038/35084130
Hardy. M. (2002a). The toxicology of the three commercial polybrominated diphenyl oxide (ether) flame
retardants. Chemosphere 46: 757-777. http://dx.doi.org/10.1016/s0045-6535(01)00240-5
Hardy. ML. (2002b). A comparison of the properties of the major commercial PBDPO/PBDE product to those of
major PBB andPCB products. Chemosphere 46: 717-728.
Hardy. ML: Banasik. M: Stedeford. T. (2009). Toxicology and human health assessment of decabromodiphenyl
ether [Review]. CritRev Toxicol 39: 1-44.
Hardy. ML: Schroeder. R: Biesemeier. J: Manor. O. (2002). Prenatal oral (gavage) developmental toxicity study
of decabromodiphenyl oxide in rats. Int J Toxicol 21: 83-91. http://dx.doi.org/10.1080/10915810252866051
Harrad. S: Ibarra. C: Abdallah. MA. -E: Boon. R: Neels. H: Covad A. (2008). Concentrations of brominated
flame retardants in dust from United Kingdom cars, homes, and offices: causes of variability and
implications for human exposure. Environ Int 34: 1170-1175. http://dx.doi.0rg/10.1016/i.envint.2008.05.001
Haworth. S: Lawlor. T: Mortelmans. K: Speck. W: Zeiger. E. (1983). Salmonella mutagenicity test results for
250 chemicals. EnvironMutagen 5: 3-142. http://dx.doi.org/10.1002/em.2860050703
Hazrati. S: Harrad. S. (2006). Causes of variability in concentrations of polychlorinated biphenyls and
polybrominated diphenyl ethers in indoor air. Environ Sci Technol 40: 7584-7589.
Hoh. E: Zhu. L: Kites. RA. (2005). Novel flame retardants, l,2-bis(2,4,6-tribromophenoxy)ethane and 2,3,4,5,6-
pentabromoethylbenzene, in United States' environmental samples. Environ Sci Technol 39: 2472-2477.
http://dx.doi.org/10.1021/es048508f
H-78
-------
HSDB (Hazardous Substances Data Bank). (2011). Decabromobiphenyl ether (CASRN: 1163-19-5).
Washington, DC: National Library of Medicine, http://toxnet.nlm.nih.gov/cgi-
bin/sis/search/a?dbs+hsdb: (@,term+(@POCNO+2911
Hua. I: Kang. N: Jafvert CT: Fabrega-duque. JR. (2003). Heterogeneous photochemical reactions of
decabromodiphenyl ether. Environ Toxicol Chem 22: 798-804. http://dx.doi.org/10.1002/etc.5620220418
Huang. H: Zhang. S: Christie. P: Wang. S: Xie. M. (2010). Behavior of decabromodiphenyl ether (BDE-209) in
the soil-plant system: Uptake, translocation, and metabolism in plants and dissipation in soil. Environ Sci
Technol 44: 663-667. http://dx.doi.org/10.1021/es901860r
Huntingdon Life Sciences. (1990). Ames metabolic activation test to assess the potential mutagenic effect of
bromkal 82-0-de with cover letter dated 031290 [TSCA Submission]. (86900000366). Parsippany, NJ: BASF
Corporation, http://www.ntis.gov/search/product.aspx?ABBR=OTS0522929
Huwe. JK: Smith. DJ. (2007). Accumulation, whole-body depletion, and debromination of decabromodiphenyl
ether in male Sprague-Dawley rats following dietary exposure. Environ Sci Technol 41: 2371-2377.
IARC (International Agency for Research on Cancer). (1998). Some flame retardants and textile chemicals, and
exposures in the textile manufacturing industry. Lyon, France.
Inoue. K: Harada. K: Takenaka. K: Uehara. S: Kono. M: Shimizu. T: Takasuga. T: Senthilkumar. K: Yamashita.
F: Koizumi. A. (2006). Levels and concentration ratios of poly chlorinated biphenyls and polybrominated
diphenyl ethers in serum and breast milk in Japanese mothers. Environ Health Perspect 114:1179-1185.
IPCS (International Programme on Chemical Safety). (1994). Brominated diphenyl ethers. Geneva, Switzerland:
World Health Organization, http://www.inchem.org/documents/ehc/ehc/ehcl62.htm
IRDC (International Research & Development Corporation). (1974). Acute toxicity studies in rats and rabbits
with test data and cover letter dated 03-08-90. (86900000327). Great Lakes Chemical Corporation.
http://www.ntis.gov/search/product.aspx?ABBR=OTS0523319
Johansson. N: Viberg. H: Fredriksson. A: Eriksson. P. (2008). Neonatal exposure to deca-brominated diphenyl
ether (PBDE 209) causes dose-response changes in spontaneous behaviour and cholinergic susceptibility in
adult mice. Neurotoxicology 29: 911-919. http://dx.doi.0rg/10.1016/i.neuro.2008.09.008
Johnson-Restrepo. B: Kannan. K. (2009). An assessment of sources and pathways of human exposure to
polybrominated diphenyl ethers in the United States. Chemosphere 76: 542-548.
http://dx.doi.0rg/10.1016/i.chemosphere.2009.02.068
Kemmlein. S: Hahn. O: Jann. O. (2003). Emissions of organophosphate and brominated flame retardants from
selected consumer products and building materials. Atmos Environ 37: 5485-5493.
http://dx.doi.0rg/10.1016/i.atmosenv.2003.09.025
Kesner. M. (2005). How is bromine produced? Kesner, M. http://www.weizmann.ac.il/sci-
tea/Brombook/pdf/chapter3 .pdf
Kierkegaard. A: Asplund. L: de Wit. CA: McLachlan. MS: Thomas. GO: Sweetman. AJ: Jones. KG. (2007). Fate
of higher brominated PBDEs in lactating cows. Environ Sci Technol 41: 417-423.
Kierkegaard. A: Balk. L: Tjarnlund. U: De wit. CA: Jansson. B. (1999). Dietary uptake and biological effects of
Decabromodiphenyl Ether in Rainbow Trout (Oncorhynchus mykiss ). Environ Sci Technol 33: 1612-1617.
http://dx.doi.org/10.1021/es9807082
Kierkegaard. A: Bjorklund. J: Friden. U. (2004). Identification of the flame retardant decabromodiphenyl ethane
in the environment. Environ Sci Technol 38: 3247-3253.
Kim. YJ: Osako. M: Sakai S. (2006). Leaching characteristics of polybrominated diphenyl ethers (PBDEs) from
flame-retardant plastics. Chemosphere 65: 506-513. http://dx.doi.0rg/10.1016/i.chemosphere.2006.01.019
Kim. YM: Nam. IH: Murugesan. K: Schmidt S: Crowlev. DE: Chang. YS. (2007). Biodegradation of diphenyl
ether and transformation of selected brominated congeners by Sphingomonas sp. PH-07. Appl Microbiol
Biotechnol 77: 187-194. http://dx.doi.org/10.1007/s00253-007-1129-z
H-79
-------
Knoth. W: Mann. W: Meyer. R: Nebhuth. J. (2007). Polybrominated diphenyl ether in sewage sludge in
Germany. Chemosphere 67: 1831-1837. http://dx.doi.0rg/10.1016/j.chemosphere.2006.05.113
Kociba. R: Frauson. L: Humiston. C: Norris. J: Wade. C: Lisowe. R: Quast. J: Jersey. C: Jewett G. (1975).
Results of a two-year dietary feeding study with decabromodiphenyl Oxide (DBDPO) in rats. Journal of Fire
and Flammability 2: 267285.
Kociba. RJ: Frauson. LO: Humiston. CC: Norris. JM: Wade. CE: Lisowe. RW: Quast. JF: Jersey. CC: Jewett.
GL. (1994). Initial submission: Results of a two-year dietary feeding study with decabromodiphenyl oxide
(DBDPO) in rats [TSCA Submission]. Midland, MI: Dow Chemical Company.
http://www.ntis.gov/search/product.aspx? ABBR=OTS0001103
Kohler. A: Som. C: Helland. A: Gottschalk. F. (2008). Studying the potential release of carbon nanotubes
throughout the application life cycle. J Clean Prod 16: 927-937.
http://dx.doi.0rg/10.1016/i.iclepro.2007.04.007
Kuivikko. M: Sorsa. K: Kukkonen. JV: Akkanen. J: Kotiaho. T: Vahatalo. AV. (2010). Partitioning of tetra- and
pentabromo diphenyl ether and benzo[a]pyrene among water and dissolved and paniculate organic carbon
along a salinity gradient in coastal waters. Environ Toxicol Chem 29: 2443-2449.
http://dx.doi.org/10.1002/etc.308
Kuo. YM: Sepulveda. MS: Sutton. TM: Ochoa-Acuna. HG: Muir. AM: Miller. B: Hua. I. (2010).
Bioaccumulation and biotransformation of decabromodiphenyl ether and effects on daily growth in juvenile
lake whitefish (Coregonus clupeaformis). Ecotoxicology 19: 751-760. http://dx.doi.org/10.1007/sl0646-009-
0451-x
La Guardia. MJ: Hale. RC: Harvey. E. (2007). Evidence of denomination of Decabromodiphenyl Ether (BDE-
209) in biota from a wastewater receiving stream. Environ Sci Technol 41: 6663 -6670.
http://dx.doi.org/10.1021/es070728g
Lagalante. AF: Oswald. TD: Calvosa. FC. (2009). Polybrominated diphenyl ether (PBDE) levels in dust from
previously owned automobiles at United States dealerships. Environ Int 35: 539-544.
http://dx.doi.0rg/10.1016/i.envint.2008.09.011
Lagalante. AF: Shedden. CS: Greenbacker. PW. (2011). Levels of polybrominated diphenyl ethers (PBDEs) in
dust from personal automobiles in conjunction with studies on the photochemical degradation of
decabromodiphenyl ether (BDE-209). Environ Int 37: 899-906.
http://dx.doi.0rg/10.1016/i.envint.2011.03.007
Lassen. C: Lekke. S: Andersen. LI. (1999). Brominated flame retardantssubstance flow analysis and assessment
of alternatives. Environmental Project.
Law. K: Halldorson. T: Danell R: Stern. G: Gewurtz. S: Alaee. M: Marvin. C: Whittle. M: Tomy. G. (2006a).
Bioaccumulation and trophic transfer of some brominated flame retardants in a Lake Winnipeg (Canada)
food web. Environ Toxicol Chem 25: 2177-2186.
Law. RJ: Allchin. CR: de Boer. J: Covaci. A: Herzke. D: Lepom. P: Morris. S: Tronczynski. J: de Wit CA.
(2006b). Levels and trends of brominated flame retardants in the European environment. Chemosphere 64:
187-208. http://dx.doi.0rg/10.1016/i.chemosphere.2005.12.007
Lee. LK: He. J. (2010). Reductive debromination of polybrominated diphenyl ethers by anaerobic bacteria from
soils and sediments. Appl Environ Microbiol 76: 794-802. http://dx.doi.org/10.1128/AEM.01872-09
Li W: Zhu. L: Zha. J: Wang. Z. (2011). Effects of decabromodiphenyl ether (BDE-209) on mRNA transcription
of thyroid hormone pathway and spermatogenesis associated genes in Chinese rare minnow (Gobiocypris
rams). Environ Toxicol. http://dx.doi.org/10.1002/tox.20767
Lim. DH: Lastoskie. CM. (2011). A dynamic multimedia environmental and bioaccumulation model for
brominated flame retardants in Lake Huron and Lake Erie, USA. Environ Toxicol Chem 30: 1018-1025.
http://dx.doi.org/10.1002/etc.482
H-80
-------
Litton Bionetics. (1976). Mutegnicity evaluation of compound 277-10 (final) with test data and cover letter
[TSCA Submission]. (8690000332). West Lafayette, IN: Great Lakes Chemical Corporation.
http://www.ntis.gov/search/product.aspx? ABBR=OTS0523324
Liu. L: Zhu. W: Xiao. L: Yang. L. (201 la). Effect of decabromodiphenyl ether (BDE 209) and dibromodiphenyl
ether (BDE 15) on soil microbial activity and bacterial community composition. J Hazard Mater 186: 883-
890. http://dx.doi.0rg/10.1016/i.ihazmat.2010.ll.079
Liu. M: Tian. S: Chen. P: Zhu. L. (201 Ib). Predicting the bioavailability of sediment-associated polybrominated
diphenyl ethers using a 45-d sequential Tenax extraction. Chemosphere 85: 424-431.
http://dx.doi.0rg/10.1016/i.chemosphere.2011.07.069
Lorber. M. (2008). Exposure of Americans to polybrominated diphenyl ethers [Review]. J Expo Sci Environ
Epidemiol 18: 2-19. http://dx.doi.org/10.1038/sj.jes.7500572
Luo. Y: Luo. XJ: Lin. Z: Chen. SJ: Liu. J: Mai. BX: Yang. ZY. (2009). Polybrominated diphenyl ethers in road
and farmland soils from an e-waste recycling region in Southern China: Concentrations, source profiles, and
potential dispersion and deposition. Sci Total Environ 407: 1105-1113.
http://dx.doi.0rg/10.1016/i.scitotenv.2008.10.044
Mahapatra. N.. (2010) Design and simulation of cumene plant using aspen plus. (National Institute of
Technology Rourkela, Rourkela, Odisha, India. Retrieved from
http://ethesis.nitrkl.ac.in/1746/l/nirlipt ethesis.pdf
McGregor. DB: Brown. A: Cattanach. P: Edwards. I: McBride. D: Riach. C: Caspary. WJ. (1988). Responses of
the L5178Y tk+/tk- mouse lymphoma cell forward mutation assay: III. 72 coded chemicals. Environ Mol
Mutagen 12: 85-154. http://dx.doi.org/10.1002/em.286Q120111
Mikula. P: Svobodova. Z. (2006). Brominated flame retardants in the environment: Their sources and effects (A
review). Acta Veterinaria Brno 75: 587-599. http://dx.doi.org/10.2754/avb200675040587
Moniruzzaman. M: Winev. KI. (2006). Polymer nanocomposites containing carbon nanotubes. Macromolecules
39: 5194-5205. http://dx.doi.org/10.1021/ma060733p
Morck. A: Hakk. H: Orn. U: Klasson Wehler. E. (2003). Decabromodiphenyl ether in the rat: Absorption,
distribution, metabolism, and excretion. Drug Metab Dispos 31: 900-907.
http://dx.doi.0rg/10.1124/dmd.31.7.900
Mueller. NC: Nowack. B. (2008). Exposure modeling of engineered nanoparticles in the environment. Environ
Sci Technol 42: 4447-4453. http://dx.doi.org/10.1021/es7029637
Muenhor. D: Harrad. S: All N: Covad A. (2010). Brominated flame retardants (BFRs) in air and dust from
electronic waste storage facilities in Thailand. Environ Int 36: 690-698.
http://dx.doi.0rg/10.1016/i.envint.2010.05.002
Myhr. B: McGregor. D: Bowers. L: Riach. C: Brown. AG: Edwards. I: McBride. D: Martin. R: Caspary. WJ.
(1990). L5178Y mouse lymphoma cell mutation assay results with 41 compounds. Environ Mol Mutagen 16:
138-167. http://dx.doi.org/10.1002/em.2850160506
Nakari. T: Pessala. P. (2005). In vitro estrogenicity of polybrominated flame retardants. Aquat Toxicol 74: 272-
279. http://dx.doi.org/10.1016/i.aquatox.2005.06.004
NLM (National Institutes of Health, National Library of Medicine). (2011). ChemlDPlus [Database].
Norris. JM: Ehrmantraut JW: Gibbons. CL: Kociba. PJ: Schwetz. BA: Rose. JQ: Humiston. CG: Jewett. GL:
Crummett WB: Gehring. PJ: Tirsell. JB: Brosier. JS. (1973). Toxicological and environmental factors
involved in the selection of decabromodiphenyl oxide as a fire retardant chemical. Appl Polym Symp 22:
195-219.
Norris. JM: Kociba. RJ: Schwetz. BA: Rose. JQ: Humiston. CG: Jewett GL: Gehring. PJ: Mailhes. JB. (1975).
Toxicology of octabromobiphenyl and decabromodiphenyl oxide. Environ Health Perspect 11: 153-161.
North. KD. (2004). Tracking polybrominated diphenyl ether releases in a wastewater treatment plant effluent,
Palo Alto, California. Environ Sci Technol 38: 4484-4488. http://dx.doi.org/10.1021/es049627y
H-81
-------
NRC (National Research Council). (2000). Toxicological risks of selected flame-retardant chemicals.
Washington, DC: National Academy Press.
NTP (National Toxicology Program). (1986). NTP toxicology and carcinogenesis studies of Decabromodiphenyl
Oxide (CAS No. 1163-19-5) inF344/N rats andB6C3Fl mice (Feed studies) (pp. 1-242). (309). Research
Triangle Park, NC: U.S. Department of Health and Human Services.
OECD (Organisation for Economic Co-operation and Development). (2007). OECD guideline for the testing of
chemicals developmental neurotoxicity study. Paris, France: Organization for Economic Co-operation and
Development. http://www.oecdbookshop.org/oecd/displav.asp?K=5L4FG25MNKXS&DS=Test-No.-426-
Developmental-Neurotoxicitv-Studv
Offenberg. JH: Stapleton. HM: Strynar. MJ: Lindstrom. AB. (2006). Polybrominated diphenyl ethers in U.S.
soils. Dioxin 2006, August 21-25, 2006, Oslo, Norway.
Pros. PR: Hoover. D: Rodigari. F: Crane. D: Sericano. J. (2005). Levels and distribution of polybrominated
diphenyl ethers in water, surface sediments, and bivalves from the San Francisco Estuary. Environ Sci
Technol 39: 33-41. http://dx.doi.org/10.1021/es048905q
Palm. A: Cousins. IT: Mackav. D: Tysklind. M: Metcalfe. C: Alaee. M. (2002). Assessing the environmental fate
of chemicals of emerging concern: A case study of the polybrominated diphenyl ethers. Environ Pollut 117:
195-213.
Peng. X: Tang. C: Yu. Y: Tan. J: Huang. Q: Wu. J: Chen. S: Mai. B. (2009). Concentrations, transport, fate, and
releases of polybrominated diphenyl ethers in sewage treatment plants in the Pearl River Delta, South China.
Environ Int 35: 303-309. http://dx.doi.0rg/10.1016/i.envint.2008.07.021
Petersen. EJ: Henry. TB. (2012). Methodological considerations for testing the ecotoxicity of carbon nanotubes
and fullerenes: Review [Review]. Environ Toxicol Chem 31: 60-72. http://dx.doi.org/10.1002/etc.710
Pharmakon Research International (Pharmakon Research International, Inc.). (1994). Initial submission: Assay
of comedogenicity in the rabbit ear with Saytex 102 and attached Saytex 102 technical bulletin [TSCA
Submission]. (FYI-OTS-0794-1095). Richmond, VA: Ethyl Corporation.
http://www.ntis.gov/search/product.aspx?ABBR=OTS0001095
Pure Strategies Inc. (2005). Decabromodiphenylether: An investigation of non-halogen substitutes in electronic
enclosure and textile applications. Lowell, MA: Lowell Center for Sustainable Production.
Qin. X: Xia. X: Yang. Z: Yan. S: Zhao. Y: Wei. R: Li. Y: Tian. M: Zhao. X: Qin. Z: Xu. X. (2010). Thyroid
disruption by technical decabromodiphenyl ether (DE-83R) at low concentrations in Xenopus laevis. J
Environ Sci 22: 744-751.
Qiu. YW: Zhang. G: Guo. LL: Zheng. GJ: Cai SO. (2010). Bioaccumulation and historical deposition of
polybrominated diphenyl ethers (PBDEs) in Deep Bay, South China. Mar Environ Res 70: 219-226.
http://dx.doi.0rg/10.1016/i.marenvres.2010.05.004
Raff. J: Hites. R. (2004). Polybrominated diphenyl ethers in Mississippi River suspended sediment.
Organohalogen Compounds 66: 3673-3677.
Rahman. F: Langford. KH: Scrimshaw. MD: Lester. JN. (2001). Polybrominated diphenyl ether (PBDE) flame
retardants [Review]. Sci Total Environ 275: 1-17. http://dx.doi.org/10.1016/80048-9697(0D00852-X
Ren. M: Peng. P: Cai. Y: Chen. D: Zhou. L: Chen. P: Hu. J. (2011). PBDD/F impurities in some commercial
deca-BDE. Environ Pollut 159: 1375-1380. http://dx.doi.0rg/10.1016/i.envpol.2011.01.004
Rice. DC: Reeve. EA: Herlihy. A: Zoeller. RT: Thompson. WD: Markowski. VP. (2007). Developmental delays
and locomotor activity in the C57BL6/J mouse following neonatal exposure to the fully-brominated PBDE,
decabromodiphenyl ether. Neurotoxicol Teratol 29: 511-520. http://dx.doi.0rg/10.1016/i.ntt.2007.03.061
Ricklund. N: Kierkegaard. A: McLachlan. MS: Wahlberg. C. (2009). Mass balance of decabromodiphenyl
ethane and decabromodiphenyl ether in a WWTP. Chemosphere 74: 389-394.
http://dx.doi.0rg/10.1016/i.chemosphere.2008.09.054
H-82
-------
Riu. A: Cravedi JP: Debrauwer. L: Garcia. A: Canlet C: Jouanin. I: Zalko. D. (2008). Disposition and metabolic
profiling of [14C]-decabromodiphenyl ether in pregnant Wistar rats. Environ Int 34: 318-329.
http://dx.doi.0rg/10.1016/i.envint.2007.03.007
Riva. C: Binelli. A: Cognj D: Provini. A. (2007). Evaluation of DNA damage induced by decabromodiphenyl
ether (BDE-209) in hemocytes of Dreissena polymorpha using the comet and micronucleus assays. Environ
Mol Mutagen48: 735-743. http://dx.doi.org/10.1002/em.20353
Rose. M: Bennett. DH: Bergman. A: Fangstrom. B: Pessah. IN: Hertz-Picciotto. I. (2010). PBDEs in 2-5 year-
old children from California and associations with diet and indoor environment. Environ Sci Technol 44:
2648-2653. http://dx.doi.org/10.1021/es903240g
Ross. PS: Couillard. CM: Ikonomou. MG: Johannessen. SC: Lebeuf. M: Macdonald. RW: Tomy. GT. (2009).
Large and growing environmental reservoirs of Deca-BDE present an emerging health risk for fish and
marine mammals. Mar Pollut Bull 58: 7-10. http://dx.doi.0rg/10.1016/i.marpolbul.2008.09.002
Sacks. JD: Stanek. LW: Luben. TJ: Johns. DO: Buckley. BJ: Brown. JS: Ross. M. (2011). Particulate-matter
induced health effects: Who is susceptible? [Review]. Environ Health Perspect 119: 446-454.
http://dx.doi.org/10.1289/ehp. 1002255
Salamova. A: Kites. RA. (2010). Evaluation of tree bark as a passive atmospheric sampler for flame retardants,
PCBs, and organochlorine pesticides. Environ Sci Technol 44: 6196-6201.
http://dx.doi.org/10.1021/esl01599h
Sandholm. A: Emanuelsson. BM: Wehler. EK. (2003). Bioavailability and half-life of decabromodiphenyl ether
(BDE-209) in rat. Xenobiotica 33: 1149-1158. http://dx.doi.org/10.1080/00498250310001609156
Sellstrom. U: de Wit. CA: Lundgren. N: Tysklind. M. (2005). Effect of sewage-sludge application on
concentrations of higher-brominated Diphenyl Ethers in soils and earthworms. Environ Sci Technol 39:
9064-9070. http://dx.doi.org/10.1021/es051190m
Shih. YH: Wang. CK. (2009). Photolytic degradation of polybromodiphenyl ethers under UV-lamp and solar
irradiations. J Hazard Mater 165: 34-38. http://dx.doi.0rg/10.1016/j.jhazmat.2008.09.103
Sjodin. A: Carlsson. H: Thuresson. K: Sjolin. S: Bergman. A: Ostman. C. (2001). Flame retardants in indoor air
at an electronics recycling plant and at other work environments. Environ Sci Technol 35: 448-454.
http://dx.doi.org/10.1021/es000077n
Soderstrom. G: Sellstrom. U: De wit CA: Tysklind. M. (2004). Photolytic debromination of decabromodiphenyl
ether (BDE 209). Environ Sci Technol 38: 127-132. http://dx.doi.org/10.1021/es034682c
Som. C: Wick. P: Krug. H: Nowack. B. (2011). Environmental and health effects of nanomaterials in
nanotextiles and facade coatings [Review]. Environ Int 37: 1131-1142.
http://dx.doi.0rg/10.1016/i.envint.2011.02.013
Song. M: Chu. S: Letcher. RJ: Seth. R. (2006). Fate, partitioning, and mass loading of polybrominated diphenyl
ethers (PBDEs) during the treatment processing of municipal sewage. Environ Sci Technol 40: 6241-6246.
Stapleton. HM. (2006). Instrumental methods and challenges in quantifying polybrominated diphenyl ethers in
environmental extracts: A review [Review]. Anal Bioanal Chem 386: 807-817.
http://dx.doi.org/10.1007/s00216-006-0400-y
Stapleton. HM: Alaee. M: Letcher. RJ: Baker. JE. (2004). Debromination of the flame retardant
decabromodiphenyl ether by juvenile carp (Cyprinus carpio) following dietary exposure. Environ Sci
Technol 38: 112-119.
Stapleton. HM: Eagle. S: Sjodin. A: Webster. TF. (2012). Serum PBDEs in a North Carolina toddler cohort:
Associations with hand wipes, house dust and socioeconomic variables. Environ Health Perspect 120: 1049-
1054. http://dx.doi.org/10.1289/ehp. 1104802
Stapleton. HM: Kelly. SM: Pei. R: Letcher. RJ: Gunsch. C. (2009). Metabolism of polybrominated diphenyl
ethers (PBDEs) by human hepatocytes in vitro. Environ Health Perspect 117: 197-202.
H-83
-------
Su. Y: Hung. H: Brice. KA: Su. K: Alexandrou. N: Blanchard. P: Chan. E: Sverko. E: Fellin. P. (2009). Air
concentrations of polybrominated diphenyl ethers (PBDEs) in 20022004 at a rural site in the Great Lakes.
Atmos Environ43: 6230-6237. http://dx.doi.0rg/10.1016/i.atmosenv.2009.08.034
Su. Y: Hung. H: Sverko. E: Fellin. P: Li. H. (2007). Multi-year measurements of polybrominated diphenyl ethers
(PBDEs) in the Arctic atmosphere. Atmos Environ 41: 8725-8735.
http://dx.doi.0rg/10.1016/i.atmosenv.2007.07.032
Sverdrup. LE: Hartnik. T: Mariussen. E: Jensen. J. (2006). Toxicity of three halogenated flame retardants to
nitrifying bacteria, red clover (Trifolium pratense), and a soil invertebrate (Enchytraeus crypticus).
Chemosphere 64: 96-103. http://dx.doi.0rg/10.1016/i.chemosphere.2005.ll.056
Textile Exchange. (2012). Flame retardants: Textile finishes for flame resistant fabrics.
http://www.teonline.com/knowledge-centre/flame-retardants.html
Tokarz. JA: Ahn. MY: Leng. J: Fillev. TR: Nies. L. (2008). Reductive denomination of polybrominated
diphenyl ethers in anaerobic sediment and a biomimetic system. Environ Sci Technol 42: 1157-1164.
Tomy. GT: Palace. VP: Halldorson. T: Braekevelt. E: Danell. R: Wautier. K: Evans. B: Brinkworth. L: Fisk. AT.
(2004). Bioaccumulation, biotransformation, and biochemical effects of brominated diphenyl ethers in
juvenile Lake Trout ( Salvelinus namaycush). Environ Sci Technol 38: 1496-1504.
http://dx.doi.org/10.1021/es035070v
Tseng. LH: Lee. CW: Pan. MH: Tsai. SS: Li MH: Chen. JR: Lav. JJ: Hsu. PC. (2006). Postnatal exposure of the
male mouse to 2,2',3,3',4,4',5,5',6,6'-decabrominated diphenyl ether: Decreased epididymal sperm functions
without alterations in DNA content and histology in testis. Toxicology 224: 33-43.
http://dx.doi.0rg/10.1016/i.tox.2006.04.003
Tseng. LH: Li MH: Tsai SS: Lee. CW: Pan. MH: Yao. WJ: Hsu. PC. (2008). Developmental exposure to
decabromodiphenyl ether (PBDE 209): Effects on thyroid hormone and hepatic enzyme activity in male
mouse offspring. Chemosphere 70: 640-647. http://dx.doi.0rg/10.1016/j.chemosphere.2007.06.078
U.S. EPA (U.S. Environmental Protection Agency). (1998). Health effects test guidelines OPPTS 870.3700
prenatal developmental toxicity study [EPA Report]. (EPA 712C98207). Washington, DC.
U.S. EPA (U.S. Environmental Protection Agency). (2005a). Furniture flame retardancy partnership:
Environmental profiles of chemical flame-retardant alternatives for low-density polyurethane foam: Volume
1 [EPA Report]. (EPA-742/R-05-002A). http://www.epa.gov/dfe/pubs/flameret/altrep-vl/altrepvl-flc.pdf
U.S. EPA (U.S. Environmental Protection Agency). (2005b). Guidelines for carcinogen risk assessment [EPA
Report]. (EPA/630/P-03/001F). Washington, DC. http://www.epa.gov/cancerguidelines/
U.S. EPA (U.S. Environmental Protection Agency). (2006). Life cycle assessment: Principles and practice [EPA
Report]. (EPA/600/R-06/060). Cincinnati, OH.
U.S. EPA (U.S. Environmental Protection Agency). (2008a). Child-specific exposure factors handbook [EPA
Report]. (EPA/600/R-06/096F). Washington, DC.
http ://cfpub. epa. gov/ncea/cfm/recordisplav. cfm?deid= 199243
U.S. EPA (U.S. Environmental Protection Agency). (2008b). Toxicological review of decabromodiphenyl ether
(CAS No. 1163-19-5) [EPA Report]. (635-R-07-008F).
U.S. EPA (U.S. Environmental Protection Agency). (2009). Targeted national sewage sludge survey statistical
analysis report [EPA Report]. (EPA-822-R-08-018). Washington, DC.
U.S. EPA (U.S. Environmental Protection Agency). (2010a). An exposure assessment of polybrominated
diphenyl ethers [EPA Report]. (EPA/600/R-08/086F). Washington, DC.
http ://cfpub. epa. gov/ncea/cfm/recordisplav. cfm?deid=210404
U.S. EPA (U.S. Environmental Protection Agency). (2010b). IRIS site help and tools: Glossary, acronyms and
abbreviations. Available online at http://www.epa.gov/ncea/iris/help gloss.htm (accessed August 23, 2010).
U.S. EPA (U.S. Environmental Protection Agency). (2011). Exposure factors handbook 2011 edition (final)
[EPAReport]. (EPA/600/R-09/052F). http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=236252
H-84
-------
Van der Ven. LT: van de Kuil T: Leonards. PE: Slob. W: Canton. RF: Germer. S: Visser. TJ: Litens. S:
Hakansson. H: Schrenk. D: van den Berg. M: Piersma. AH: Vos. JG: Opperhuizen. A. (2008). A 28-day oral
dose toxicity study in Wistar rats enhanced to detect endocrine effects of decabromodiphenyl ether
(decaBDE). Toxicol Lett 179: 6-14. http://dx.doi.0rg/10.1016/j.toxlet.2008.03.003
Viberg. H: Fredriksson. A: Eriksson. P. (2007). Changes in spontaneous behaviour and altered response to
nicotine in the adult rat, after neonatal exposure to the brominated flame retardant, decabrominated diphenyl
ether (PBDE 209). Neurotoxicology 28: 136-142. http://dx.doi.0rg/10.1016/j.neuro.2006.08.006
Viberg. H: Fredriksson. A: Jakobsson. E: Orn. U: Eriksson. P. (2003). Neurobehavioral derangements in adult
mice receiving decabrominated diphenyl ether (PBDE 209) during a defined period of neonatal brain
development. Toxicol Sci 76: 112-120. http://dx.doi.org/10.1093/toxsci/kfg210
Viberg. H: Mundv. W: Eriksson. P. (2008). Neonatal exposure to decabrominated diphenyl ether (PBDE 209)
results in changes in BDNF, CaMKII and GAP-43, biochemical substrates of neuronal survival, growth, and
synaptogenesis. Neurotoxicology 29: 152-159. http://dx.doi.0rg/10.1016/i.neuro.2007.10.007
Vonderheide. AP: Mueller. KE: Meija. J: Welsh. GL. (2008). Polybrominated diphenyl ethers: Causes for
concern and knowledge gaps regarding environmental distribution, fate and toxicity [Review]. Sci Total
Environ 400: 425-436. http://dx.doi.0rg/10.1016/i.scitotenv.2008.05.003
Vrkoslavova. J: Demnerova. K: Mackova. M: Zemanova. T: Macek. T: Haislova. J: Pulkrabova. J: Hradkova. P:
Stiborova. H. (2010). Absorption and translocation of polybrominated diphenyl ethers (PBDEs) by plants
from contaminated sewage sludge. Chemosphere 81: 381-386.
http://dx.doi.0rg/10.1016/i.chemosphere.2010.07.010
Wang. F: Wang. J: Dai. J: Hu. G: Luo. X: Mai. B. (2010a). Comparative tissue distribution, biotransformation
and associated biological effects by decabromodiphenyl ethane and decabrominated diphenyl ether in male
rats after a 90-day oral exposure study. Environ Sci Technol 44: 5655-5660.
http://dx.doi.org/10.1021/esl01158e
Wang. S: Zhang. S: Huang. H: Christie. P. (2011). Behavior of decabromodiphenyl ether (BDE-209) in soil:
Effects of rhizosphere and mycorrhizal colonization of ryegrass roots. Environ Pollut 159: 749-753.
http://dx.doi.0rg/10.1016/i.envpol.2010.ll.035
Wang. T: Wang. Y: Fu. J: Wang. P: Li. Y: Zhang. Q: Jiang. G. (201 Ob). Characteristic accumulation and soil
penetration of polychlorinated biphenyls and polybrominated diphenyl ethers in wastewater irrigated
farmlands. Chemosphere 81: 1045-1051. http://dx.doi.0rg/10.1016/i.chemosphere.2010.07.045
Wang. Y: Zhang. Q: Lv. J: Li. A: Liu. H: Li. G: Jiang. G. (2007). Polybrominated diphenyl ethers and
organochlorine pesticides in sewage sludge of wastewater treatment plants in China. Chemosphere 68: 1683-
1691. http://dx.doi.0rg/10.1016/i.chemosphere.2007.03.060
Wania. F: Dugani. CB. (2003). Assessing the long-range transport potential of polybrominated diphenyl ethers: a
comparison of four multimedia models. Environ Toxicol Chem 22: 1252-1261.
Wania. F: Lei. YD: Harner. T. (2002). Estimating octanol-air partition coefficients of nonpolar semivolatile
organic compounds from gas chromatographic retention times. Anal Chem 74: 3476-3483.
Watanabe. I: Sakai. S. (2003). Environmental release and behavior of brominated flame retardants [Review].
Environ Int 29: 665-682.
Watkins. DJ: Mcclean. MD: Fraser. AJ: Weinberg. J: Stapleton. HM: Sjodin. A: Webster. TF. (2011). Exposure
to PBDEs in the office environment: evaluating the relationships between dust, handwipes, and serum.
Environ Health Perspect 119: 1247-1252. http://dx.doi.org/10.1289/ehp.1003271
Webster. TF: Harrad. S: Millette. JR: Holbrook. RD: Davis. JM: Stapleton. HM: Allen. JG: Mcclean. MD:
Ibarra. C: Abdallah. MA. -E: Covaci. A. (2009). Identifying transfer mechanisms and sources of
decabromodiphenyl ether (BDE 209) in indoor environments using environmental forensic microscopy.
Environ Sci Technol 43: 3067-3072. http://dx.doi.org/10.1021/es803139w
WHO (World Health Organization). (1994). Decabromodiphenyl ether. In Brominated Diphenylethers. Geneva.
http://www.inchem.org/documents/ehc/ehc/ehcl62.htm
H-85
-------
Wright HE: Zhang, 0: Mihelcic. JR. (2008). Integrating economic inputoutput life cycle assessment with risk
assessment for a screening-level analysis. Int J Life Cycle Assess 13: 412-420.
http://dx.doi.org/10.1007/sll367-008-0006-z
Xie. X: Wu. Y: Zhu. M: Zhang. YK: Wang. X. (2011). Hydroxyl radical generation and oxidative stress in
earthworms (Eisenia fetida) exposed to decabromodiphenyl ether (BDE-209). Ecotoxicology 20: 993-999.
http://dx.doi.org/10.1007/sl0646-011-0645-x
Yogui. GT: Sericano. JL. (2009). Polybrominated diphenyl ether flame retardants in the U.S. marine
environment: a review [Review]. Environ Int 35: 655-666. http://dx.doi.0rg/10.1016/i.envint.2008.ll.001
Yu. S: Zou. P: Zhu. W: Yang. L: Xiao. L: Jiang. L: Wang. X: Wu. J: Yuan. Y. (2010). Effects of humic acids
and microorganisms on decabromodiphenyl ether, 4,4-dibromodiphenyl ether and anthracene transportation
in soil. Science China Chemistry 53: 950-968. http://dx.doi.org/10.1007/sll426-010-0126-8
Yun. SH: Addink. R: McCabe. JM: Ostaszewski. A: Mackenzie-Taylor. D: Taylor. AB: Kannan. K. (2008).
Polybrominated diphenyl ethers and polybrominated biphenyls in sediment and floodplain soils of the
Saginaw River watershed, Michigan, USA. Arch Environ Contam Toxicol 55: 1-10.
http://dx.doi.org/10.1007/s00244-007-9084-3
Zhou. T: Ross. DG: DeVito. MJ: Crofton. KM. (2001). Effects of short-term in vivo exposure to polybrominated
diphenyl ethers on thyroid hormones and hepatic enzyme activities in weanling rats. Toxicol Sci 61: 76-82.
Zhu. W: Liu. L: Zou. P: Xiao. L: Yang. L. (2010). Effect of decabromodiphenyl ether (BDE 209) on soil
microbial activity and bacterial community composition. World J Microbiol Biotechnol 26: 1891-1899.
http://dx.doi.org/10.1007/sll274-010-0371-l
Zota. AR: Adamkiewicz. G: Morello-Frosch. RA. (2010). Are PBDEs an environmental equity concern?
Exposure disparities by socioeconomic status. Environ Sci Technol 44: 5691-5692.
http://dx.doi.org/10.1021/esl01723d
Zou. M. -Y: Ran. Y: Gong. J: Mai. B. -X: Zeng. EY. (2007). Polybrominated diphenyl ethers in watershed soils
of the Pearl River Delta, China: Occurrence, inventory, and fate. Environ Sci Technol 41: 8262-8267.
http://dx.doi.org/10.1021/es071956d
H-86
-------
Appendix I. External Review Draft
Comments and Agency Responses
1.1. Background
The External Review Draft of this case study (U.S. EPA. 2012) served as the starting point for
identifying and prioritizing research gaps that, if pursued, could inform future assessments and
subsequent risk management decisions for multiwalled carbon nanotubes (MWCNTs) in flame-retardant
upholstery textiles or similar materials and applications. As discussed in more detail in Chapter 1. the
draft was the basis for the collective judgment step of the comprehensive environmental assessment
(CEA) process (see Figure 1-2). in which experts read the case study document, participated in an online
data prioritization exercise, and (for a subset of experts) attended a workshop. The collective judgment
step resulted in the identification of elements and risk relevant factors of the CEA framework as priority
areas for future assessment or research due to data gaps and importance to risk management. These areas
are hereafter referred to as "Priority Research Areas." Some of these areas are most relevant to
individuals who plan research. These areas are those that the experts rated as important to consider in risk
assessments but in which they were not confident the available data could support risk management
decisions. Other priority areas are more relevant to individuals who develop assessments or are
responsible for risk management efforts (e.g., researching which type of risk management plan would be
most suitable given current information); those areas are the ones experts rated as important to consider in
risk assessments and in which they had greater confidence that available data might support risk
management decisions (see Section 1.1.3). Notably, in applying the CEA approach to MWCNTs, the
majority of priority areas that emerged are most pertinent to research planning rather than developing
assessments. Throughout the revision process for the case study, efforts were made to streamline the
document so that it would clearly reflect each priority that emerged from the CEA collective judgment
step, input from public comments, and the opinions of expert stakeholders involved in prioritizing the
research gaps.
This appendix documents how the External Review Draft of the case study was revised to:
(1) respond to public comments and input from experts participating in the collective judgment step of the
CEA process (Figure 1-2). and (2) reflect the priorities identified through the CEA collective
judgment step.
1-1
-------
1.1.1. Appendix I Development Process
As part of the collective judgment step of the CEA process (Figure 1-2) described in
Section 1.1.3. 23 experts provided written responses to the following charge questions:
1. Do you know of additional, specific studies on MWCNTs that should be included in the case
study to help identify data gaps that are important to support future assessment and risk
management efforts for MWCNTs in flame-retardant textile coatings?
1. Is the science accurately conveyed throughout the document? If not, please list any areas that
need improvement and provide specific comments in the text to highlight areas that should be
refined.
2. Does the comparison of decabromodiphenyl ether (decaBDE) and MWCNTs in the case
study document help to identify research gaps to support future assessments and risk
management decisions for MWCNTs? If not, please briefly explain.
3. Do you have any specific comments on how this document could be improved?
One of the 23 experts also provided a PDF copy of the case study with free-form comments (i.e.,
not specific to a charge question) linked to specific regions of text. In addition, four members of the
public provided input during the public comment period announced in a July 2, 2012 Federal Register
Notice.39 Finally, an interagency commenter provided input on the draft document during the public
comment period. Affiliations of the experts are provided in Table 1-1; affiliations of the interagency and
public commenters are provided in Table 1-2.
All charge question responses, public comments, and expert free-form comments received on the
External Review Draft were tracked with the aid of an Excel-based comment tracking sheet. Longer
responses and public comments were broken down into distinct, individual thoughts and assigned unique
comment numbers. Each unique comment was assigned to the most relevant portion of the detailed CEA
framework (Figure 1-3). which was also used for the online collective judgment prioritization exercises of
the CEA process.40
Comments then were assigned "themes" based on recurring topics so that similar comments
related to the same CEA framework area could be grouped and comments with similar concepts that
crossed multiple CEA framework areas also could be grouped. Themes were determined progressively
and the list of themes was reconsidered and revised as more comments were considered, to develop a
39http://www.gpo.gov/fdsYS/pkg/FR-2012-07-02/html/2012-16137.htm
40Note that the CEA framework does not include the introductory and background topics from Chapter 1 of the case
study, such as flame-retardant regulations or the choice of nanomaterial and application. Comments that referred to
topics outside of the framework areas were assigned to a numbered section of the actual case study as opposed a
CEA framework area. Specific chapter assignments were also used in place of CEA framework area in cases of
highly specific comments that referred to a specific line of text within the case study.
1-2
-------
limited number of themes that created broad categories. Multiple themes were often, but not always,
applied to a single comment to either increase the degree of specificity of the theme (e.g., a general
comment stating that the comparison between decaBDE and MWCNTs was not useful compared to a
comment that specified why the comparison was not useful) or to account for multiple unrelated themes in
the same comment (e.g., a comment noting that information on release rates and exposure are data gaps
and noted the need for better analytical techniques). The themes provided a flexible and inclusive method
for grouping similar comments.
Finally, Agency responses were drafted for each comment using consistent language where
possible to connect comments from multiple commenters that expressed the same basic ideas and themes.
Responses were focused on applying one of the five main categories of action taken, as described in detail
in Chapter 1 and summarized in Table 1-3. In some cases, more than one response category was
appropriate for the same comment or group of comments. This is particularly true for comments that
applied to broader themes or CEA framework areas. The "Agency Response" includes references to
multiple categories (e.g., addition of Information Highlight Boxes as well as in-text edits) as necessary to
describe the complete actions that were taken in response to each comment or group of comments.
Table 1-1. Expert affiliations and area of expertise.
Expertise Area Sector Affiliation
Ecological Effects Academic Institutions and Centers
Exposure & Dose Government
Ecological Effects Government
Human Health Effects Government
Policy Nongovernmental Organization
Material Characterization Academic Institutions and Centers
Exposure and Dose Government
Exposure and Dose Academic Institutions and Centers
Material Characterization Government
Human Health Effects Independent Consulting
Policy Academic Institutions and Centers
Material Characterization Government
Human Health Effects Independent Consulting
I-3
-------
Table 1-1 (Continued): Expert affiliations and area of expertise.
Expertise Area
Sector Affiliation
Environmental Fate and Transport
Academic Institutions and Centers
Manufacturing
Industry
Ecological Effects
Academic Institutions and Centers
Policy
Government
Exposure and Dose
Industry
Environmental Fate and Transport
Government
Ecological Effects
Independent Consulting
Exposure and Dose
Academic Institutions and Centers
Risk Assessment
Industry
Risk Assessment
Government
Note: Order of expertise area and affiliation is not associated with author identification numbers in Table 1-11 and Table 1-12. to
preserve anonymity of experts.
Table I-2. Public commenters' affiliations and area of expertise.
Commentator ID Expertise Area
Sector Affiliation
IA1
Not Identified
Government
P1
Not Identified
Industry
P2
Not Identified
Industry
P3
Not Identified
Government
P4
Not Identified
Nongovernmental Organization
I-4
-------
Table 1-3. Agency response categories.
Response
Category Purpose
Actions Taken
Notes
1 Emphasize
outcomes of
collective
judgment step to
support MWCNT
research planning
in this application
ofCEA
Added Priority
Research Area
Highlight Boxes
Moved unprioritized
sections to Appendix
Text boxes were embedded into the case study to indicate
the outcomes of the collective judgment process, outlined
in red for priority areas and outlined in qray for
unprioritized areas.
For areas of the detailed CEA framework that experts
determined were of lesser importance to consider in a future
risk assessment of MWCNTs, all text was moved to
Appendix G to help focus the revised document on the most
important information.
Added final Priority
Focus Section 6.3
A final section of the case study was added that discusses
the "Priority" areas in detail. The section includes further
details on workshop participants' rationale for designating
the area a priority, and notes additional relevant literature
not previously included in the case study, but identified by
commenters, workshop participants, or targeted literature
searches, which might inform future research in the priority
area
Update
discussion of
Priority Research
Areas to reflect
current state of
the science and
full range of topics
Added "Information
Highlight Text Boxes"
Text boxes were embedded in the case study to draw
attention to scientific concepts related to priority areas that
commenters felt were under-represented or incompletely
represented in the External Review Draft. Boxes, outlined
in blue, highlighted literature not previously included in the
case study that was provided by the commenters and
workshop participants, or identified through a targeted
literature search.
Added new figures and
tables
New figures and tables were added to draw attention to
scientific concepts previously under-represented or
incompletely characterized in the case study (due to, for
example, insufficient data at the time the case study was
written). These new figures and tables are clearly described
as new in the caption and where possible have been
outlined in blue.
Improve accuracy
Text edits to pre-
existing text
Some specific revisions were made to text that appeared in
the External Review Draft version of the draft case study in
order to clarify and improve the accuracy of statements.
I-5
-------
Table 1-3 (Continued): Agency response categories.
Response
Category Purpose
Actions Taken
Notes
Streamline
document to
support MWCNT
research planning
in this application
ofCEA
Moved decaBDE
discussions to
appendix
Detailed information on decaBDE that originally appeared in
the main body of the document was moved to Appendix H.
Added "DecaBDE A series of DecaBDE Comparison Text Boxes was added to
Comparison Boxes" to the body of the document (outlined in green) to illustrate
main text how information from decaBDE could be used to inform
MWCNT research planning in priority areas; these boxes
refer the reader to Appendix H for more detailed
information.
No action
necessary
No action
Many comments provided informative feedback that the
Agency appreciates, but did not require any edits or
changes to the body of the case study document. This was
either due to the fact that the comment pertained to an area
of the CEA framework that was not identified as a priority
area by workshop participants or did not suggest specific
action to be taken.
1-6
-------
1.1.2. Appendix I Organization
This appendix contains three types of comment tables:
1. Table 1-4. Table 1-5. Table 1-6. Table 1-7. Table 1-8. and Table 1-9: Response tables for
comments that were addressed in the body of the document based on the first four response
types described in Table 1-2 (this includes some cases where the comment was acknowledged
but no major action was taken because the related CEA area was not identified as important
by the collective judgment process), organized by relevant chapters of the case study
document;
2. Table 1-11: Response table for comments that were not specifically addressed through
revisions or edits to the main body of the case study because the comment did not suggest
that any action needed to be taken; and
3. Table 1-12. Table 1-13. and Table 1-14: Look-up tables for the comment IDs listed in the
response tables, organized by the way in which the comment was submitted (i.e., expert
charge question responses, expert free-form comments, and public comments). Original
commenter text excerpts are provided, along with the commenter and comment ID numbers,
and the theme or themes applied to each comment. Note that although commenter ID
numbers are provided in Table 1-2 above for public commentors, they are not included in
Table 1-1. to preserve anonymity of expert input.
As noted previously, comments were grouped according to similar themes, so several unique
comments are presented as relevant to a single Agency response, as indicated by the ID numbers in the far
right column of the response tables. The comments have been collectively summarized to provide readers
with a relatively quick overview of the common theme among the comments and to illustrate how the
group of comments relate to the action taken by the Agency. As noted above, the original individual
comments and assigned themes are provided in the final tables of this appendix (Table 1-12. Table 1-13.
and Table I-14V
1-7
-------
1.2. Responses to Comments
1.2.1. Addressed Comments
Table I-4. Relevant to the general case study or multiple sections of the case study.
Related to CEA
Framework
Area
Comment Theme
Response
Related to
Response Global
Category Comment ID
General Five comments generally noted
the utility of highlighting or
emphasizing Priority Research
Areas and data gaps in the case
study.
A series of text boxes was added throughout the document to highlight Priority
Areas (i.e., red outlined text boxes) for research, along with those areas
distributed elsewhere in the Importance / Confidence Matrix during the collective
judgment step of the CEA process (see Section 1.1.3). In addition, a final
section (Section 6.3) was added that discusses the outcomes of the collective
judgment prioritization process in more detail. Future applications of the CEA
approach will strive to include more figures and tables as appropriate to
summarize and highlight research priorities or data gaps.
1,2
39,61, 103,
120, 140
General Thirteen comments did not find the
comparison with decaBDE useful
overall (for example, due to
differences in physicochemical
properties or because it added
length to the document)
In revising the document, effort was made to clarify the purpose of comparing
MWCNT and decaBDE. The primary purpose of providing decaBDE information,
to inform research gap identification for MWCNT, was carried out in the
collective judgment step with the External Review Draft of the case study; as
such, decaBDE information is now primarily in Appendix H. In addition, efforts
were made to succinctly highlight how understanding decaBDE data might
inform research planning for MWCNT priority topics by adding "DecaBDE
Comparison Text Boxes."
27, 28, 29,
34,61,62,
68, 90, 98,
99, 112, 113,
203
General One comment found the Discussion of decaBDE was moved to Appendix H to provide continuity in the
comparison with decaBDE to be presentation of information on decaBDE and MWCNT; efforts were made to
useful, but suggested that succinctly highlight how understanding decaBDE data might inform research
segregating the information, rather planning for MWCNT priority topics by adding "DecaBDE Comparison Text
than alternating back and forth, Boxes."
would improve the document.
73
-------
Table 1-4 (Continued): Relevant to the general case study or multiple sections of the case study.
Related to CEA
Framework
Area Comment Theme
Response
Related to
Response Global
Category Comment ID
General One comment expressed concern
that rather than highlighting
research needs and data gaps for
MWCNT, comparison with
decaBDE could make MWCNT
appear to be relatively "good" or
"safe."
The document was reviewed to check for any biased statements related to the
use of decaBDE or MWCNTs. No instances of bias were found; however, the
purpose of including the comparison in the document was clarified in response
to this and other comments on the utility of comparing decaBDE and MWCNT in
research planning efforts (see Section 1.1.3). Further, detailed discussion
regarding decaBDE was moved to Appendix H to provide greater focus on
MWCNTs.
2, 3,4
34
General Eight comments recommended
that the CEA case study put
greater emphasis on in vitro
assessment of toxicity for
nanomaterials rather than
traditional in vivo assessments,
which would align well with the
Agency's Nanomaterial Research
Strategy, Strategic Plan for
Evaluating the Toxicity of
Chemicals, and NexGen.
As noted in Chapter 1, the case study was developed without a particular
regulatory or policy objective in mind and is not intended to establish or evaluate
specific testing protocols; however, discussion was added to Chapter 1 to clarify
the purpose of the case study and its relationship to other research frameworks,
including the Nanomaterial Research Strategy, EPA's Strategic Plan for
Evaluating the Toxicity of Chemicals, and OECD (2012) that were developed to
inform decision-making for nanomaterials to provide greater context for the
Agency's direction for future risk assessment efforts. Further, because "Impacts:
Human (Cancer, Non Cancer, and reproductive/developmental)" was identified
as a Priority Research Area by workshop participants, information regarding
NexGen and integrated testing strategy (ITS) was included in Additional
Information Highlight Box 15, which describes the trend toward developing
innovative biologically/toxicologically relevant in vitro models.
2, 3
254, 256,
257, 258,
292, 293,
294, 297,
298, 299
General One comment suggested focusing
the case study more on recently
published data than on review
papers published prior to 2010.
In writing and revising the document, efforts were made to include current,
accurate, and validated information.
160
I-9
-------
Table 1-4 (Continued): Relevant to the general case study or multiple sections of the case study.
Related to CEA
Framework
Area
Comment Theme
Response
Related to
Response Global
Category Comment ID
General Two comments noted the use of
non-peer reviewed sources to
inform the case study and
expressed concern that this
publication would offer validity to
"gray" literature. Another
questioned the use of a specific
reference that only represented a
report abstract.
As nanomaterials are new, emerging compounds, research in the field is rapid
and ever-progressing. The literature chosen for inclusion in this case study was
intended to thoroughly illustrate the state of the science and depth of research in
the field at the time the case study was written. Inclusion of literature in the case
study should not be interpreted as validation of that literature, or establishment
of particular piece of literature as key and substantial in the field. Information in
the CEA framework (i.e., the case study document) is intended to be iteratively
updated through the CEA process; key concepts presented in this case study
are subject to scientific challenges as the emerging field of nanotechnology
progresses. The Agency acknowledges the fact that the accuracy of all scientific
publications is ultimately determined through the process of repetition, or lack
thereof, by other researchers in the field and that the concepts presented in this
case study are subject to scientific re-evaluation as the emerging field of
nanotechnology progresses. A statement was added to Section 1.1.2 regarding
the date of the last literature review and associated search terms, as well as
clarifying the use of unpublished literature to supplement peer-reviewed
literature when appropriate.
12, 13, 165
General Two comments noted the need for
not only presenting the most
important information, but also
discussing conflicting evidence to
present a more balanced
discussion.
Generally, Information Highlight Text Boxes and Priority Research Area
Highlight Boxes have been added to the document to help highlight data gaps
and, where applicable, to clearly present conflicting evidence (e.g., no effect
versus effect findings). As these comments were general, no specific actions
were taken to address these comments.
1,2
125, 22
General One comment noted the need for
incorporating value of information
analysis into the CEA.
The Agency appreciates the feedback. Value of Information (VOI) was
incorporated into the workshop breakout group exercises for the areas identified
as research priorities. Future applications of the CEA approach will attempt to
incorporate greater VOI analysis during earlier steps in the process.
79
Chapter 1 / One comment questioned whether The purpose of the draft case study document and its relationship to any future
Chapter 6 the case study could be used to regulatory decisions for MWCNT was clarified in Chapter 1 (see Section 1.3)
accomplish more than its intended and expanded upon in Chapter 6 (see Section 6.3).
purpose of identifying and
prioritizing research gaps.
35
1-10
-------
Table 1-4 (Continued): Relevant to the general case study or multiple sections of the case study.
Related to CEA
Framework
Area Comment Theme
Chapter 1 / One comment suggested an
Chapter 6 OECD (2012) publication
Response
The purpose of the draft case study document and its relationship to any future
reaulatorv decisions for MWCNT was clarified in Chapter 1 (see Section 1.3)
Related to
Response Global
Category Comment ID
3 154, 146
regarding important issues in the and expanded upon in Chapter 6 (see Section 6.3). In particular, the OECD
risk assessment of nanomaterials. (2012) publication was used to highlight research needs and data gaps in the
Another suggested this reference assessment of nanomaterials in Section 6.3.
to describe a predictive
toxicological paradigm for the
assessment of nanomaterials.
Product life
cycle: Raw
Materials,
Synthesis
AND
Chapter 1 /
Chapter 4
Four comments noted the need for
more information regarding
material synthesis (e.g.,
functionalization, handling). One of
these comments additionally noted
the need to focus on chemistry
and functionalization specific to
use in flame retardants.
"Product Life Cycle: Material Synthesis, Material Processing, and Product
Manufacturing" was identified by workshop participants as a Priority Research
Area; therefore, additional information was incorporated into the relevant Priority
Research Area Highlight Box and Section 6.3.1.1. Additionally, Figure 2-1 and
Additional Information Highlight Box 6 were added to clarify the distinction(s)
between as-manufactured versus modified MWCNTs by incorporation into
products or transformation in the environment. This topic is also discussed in
Chapter 4. Additionally, Table 1-13 was added to Chapter 1 regarding
functionalization and chemistry of MWCNT specific to use in flame retardants.
1,2, 3
6,213,214,
216
Product life
cycle: Product
manufacturing;
AND
Exposure
Route:
Human
Occupational,
Human
Consumer
Eleven references were provided
along with a comment that these
studies indicate MWCNT alone will
not be useful as a flame retardant
unless they are combined with
other chemicals/materials to
achieve flame-retardant
performance.
The suggested literature was incorporated into the relevant Additional
Information Highlight Box 3 to reflect that MWCNT are likely to be used in
combination with other chemicals/materials to achieve flame-retardant
performance needs. Additionally, Priority Research Area Highlight Boxes and
Sections 6.3.1.3, 6.3.3.1, and 6.3.3.2 were included since "Product Life Cycle:
Product Manufacturing", "Exposure Route: Occupational", and "Exposure Route:
Consumer" were identified by workshop participants as Priority Research Areas.
1,2
40,41,42,
43, 44, 45,
46, 47, 48,
49, 50, 51
Product Life
Cycle: Use,
Release Rate
AND
Exposure
Route: Human
One comment suggested
identifying data gaps clearly and
prioritizing accordingly using the
specific example of release of
CNT from different media.
Additional Information Highlight Box 4 was added to Section 2.2.4.2 to highlight
release of MWCNT from different product matrices, and Additional Information
Highlight Box 12 was added to Section 4.2.2 to highlight the exposure routes
that seem most likely or most dominant of all the potential routes. Additionally,
because "Product Life Cycle: Use: Release Rate" and "Exposure Route:
Human" were identified by workshop participants as Priority Research Areas,
this comment was considered and addressed through the addition of Priority
Research Area Highlight Boxes and Sections 6.3.1.4 and 6.3.3.
1,2
114
1-11
-------
Table 1-4 (Continued): Relevant to the general case study or multiple sections of the case study.
Related to CEA
Framework
Area Comment Theme
Response
Related to
Response Global
Category Comment ID
Product life
cycle: Material
Synthesis and
Processing
AND
Exposure
Route: Human:
Occupational
Three comments noted the lack of
information regarding material
synthesis and potential for human
occupational exposure.
The comments provided details
and references (or names of
individuals to contact) for more
information regarding the current
scale, outlook, manufacturing
processes, and potential exposure
in MWCNT industry.
Suggested literature was reviewed and incorporated to enhance the discussion
of materials synthesis (by adding Table 2-2 to summarize estimated growth in
the industry, quantities produced, etc., and adding Additional Information
Highlight Box 2 in Chapter 1 regarding the fact that currently MWCNTs are not
widely used in flame-retardant textiles). This was also addressed by the addition
of a Priority Research Area Highlight Boxes and Sections 6.3.1.1 and 6.3.3.1
given that "Product Life Cycle: Material Synthesis" and "Exposure Route:
Human: Occupational" were identified as Priority Research Areas in the
workshop. The suggested individuals were contacted; one responded. Their
input was incorporated as a footnote in Section G.2.1.1.
1,2,3 54,214,216
Product Life
Cycle: Product
manufacturing
AND
Exposure
Route: Human:
Occupational
Two comments suggested
literature on potential release and
exposure in occupational
environments that was already
incorporated into the case study.
Although the provided references were already included in the case study, this
comment was considered and addressed by adding greater emphasis to
potential release and exposure during product manufacturing in a new
discussion in Section 6.3. In addition, Priority Research Area Highlight Boxes
and Sections 6.3.1.3 and 6.3.1.1 were added for "Life Cycle: Product
manufacturing" and "Exposure: Human: Occupational" as they were identified by
workshop participants as Priority Research Areas.
1, 3
144, 224
Dose: Human:
Absorption
AND
Impacts:
Human:
Cancer,
Noncancer
Three comments noted additional
references pertaining to the mode
of action of MWCNTs in the lungs,
sub-pleural deposition and pleural
translocation in the lung, as well
as structural similarities to other
particles of concern (i.e.,
asbestos), and potential
similarities to other particles of
toxicological concern. References
were also provided.
Additional Information Highlight Box 13 was added to discuss the comparison of
MWCNTs and asbestos. Additionally, the text was revised regarding the
toxicological concern from inhalation of MWCNTs. Finally, because "Dose:
Human: Absorption" and "Impacts: Human Cancer and Noncancer" was
identified by workshop participants as a Priority Research Area, Priority
Research Area Highlight Boxes were added and the topics are now discussed in
Sections 6.3.3.3 and 6.3.4.1.
1,2,3 64,122,135
Chapter 4 / One comment questioned whether
Chapter 6 information from two sources
regarding decaBDE use in
cars/aircraft was accurate.
The cited literature was reviewed and accuracy of the statement was verified.
Additional sources corroborating the statement were added.
229
1-12
-------
Table 1-4 (Continued): Relevant to the general case study or multiple sections of the case study.
Related to CEA
Framework
Area
Comment Theme
Response
Related to
Response Global
Category Comment ID
Chapter 4 /
Chapter 6
One comment questioned whether Information regarding the use of MWCNTs in combination with other
2, 3
244
information in reports regarding
MWCNT flame-retardant action
was accurately conveyed in that
the flame-retardant effects of
MWCNT alone are not sufficient
for regulatory standards.
chemicals/materials was included in Additional Information Highlight Box 3.
Additionally, uncertainty of MWCNT use in flame retardants due to small scale
(mostly R&D) and flame-retardant selection has been added in Additional
Information Highlight Box 2.
Section 4.2.5 / Several additional references were
Section 5.1.7 provided to improve the discussion
of decaBDE in the case study.
Information on decaBDE is now primarily in Appendix H: however, the provided
references were reviewed and incorporated into the text as appropriate.
2,23
Multiple Two comments recommended
DecaBDE additional information and sources
Sections relevant to decaBDE, including
physicochemical properties and
toxicokinetics, suggesting some
sources in the case study were
outdated (a specific example was
water solubility). Another noted
that additional information might
be available regarding resource
demands of decaBDE synthesis.
Basic information regarding decaBDE was reviewed for accuracy and edited as
needed. The document currently defines the water solubility of decaBDE as <0.1
ug/L in Table 1-8: no changes to this table were made as a result of this
comment. A targeted literature search was performed but no new information on
energy and resource demands of raw material extraction for synthesis of
decaBDE was found so no changes were made. "Product Life Cycle: Material
Synthesis" was identified by workshop participants as a Priority Research Area
for MWCNTs but decaBDE was not the focus of this document.
1,24, 82
1-13
-------
Table 1-5. Relevant to Chapter 1 (including Preface and Executive Summary).
Related to CEA
Framework
Area
Comment Theme
Response
Related to
Response Global
Category Comment ID
Preface One comment suggested that the
characterization of the amount of
literature available on uptake of
MWCNTs in aquatic organisms
was incorrect.
The statement was revised and clarified.
162
Executive One comment suggested there
Summary/ may be more recently published
Chapter 1 literature relevant to MWCNTs.
Text in the Executive Summary and in Section 1.1.2 was revised to clarify that
the original literature search was conducted in November 2011. A second,
limited literature search was conducted in May 2012 with the goal of capturing
literature published since November 2011. Additional targeted literature
searches were performed in November 2012 to address data gaps in priority
areas identified by workshop participants. The case study; however, is not
intended to be a comprehensive literature review. New literature was added only
if it enhanced discussions on key priority areas or data gap sections identified in
the collective judgment step of the CEA process. For example, although an
article published in 2012 corroborating details from two previously published
studies already included in the case study might not add to efforts to plan
research for MWCNTs, literature showing a new or conflicting finding would
inform research planning that supports future decision-making.
129
Chapter 1 Twenty four comments noted that
MWCNT in flame retardant might
not be the primary application
available to consumers, and
questioned whether a larger scale
application might have been a
more appropriate selection. One
comment questioned using a
single manufacturer of MWCNT
flame-retardant product as
evidence of the application's
feasibility.
Efforts will be made to include greater consideration of the current market share
of a particular application in future applications of the CEA approach. In addition,
greater detail related to information considered in the selection process, which
was originally in Appendix A, is now also included in Chapter 1. New text
describing the selection rationale is also included in Chapter 1. Additional details
regarding the use of flame-retardant MWCNT products were provided through
personal communication with Nanocyl, publicly available Nanocyl promotional
materials, and several other publications, as indicated in the text. Finally,
additional citations were included that suggest more manufacturers will enter the
market soon.
I, 30, 75, 53,
192, 193,
195, 199,
202, 204,
205, 206,
207, 209,
211,227,
230, 234,
238, 239,
240, 242,
243, 248,
250,251,
252, 253
1-14
-------
Table 1-5 (Continued): Relevant to Chapter 1 (including Preface and Executive Summary).
Related to CEA
Framework
Area
Comment Theme
Response
Response Related to
Category Global
Comment ID
Chapter 1 Two comments noted that use of
MWCNTs in flame retardant will be
determined by their ability to pass
specific flame-retardant regulatory
tests, which will in turn dictate the
feasibility of a large-scale
application.
Additional Information Highlight Box 2 has been added to Section 1.3.2 to clarify
that the use of MWCNTs in flame-retardant textiles is at an exploratory stage.
Although research and development efforts have shown the material's capability
as a flame retardant, many considerations will determine whether MWCNT use
in flame-retardant textiles will grow. Additional Information Highlight Box 1 was
added to clarify the factors influencing selection of flame-retardant materials. In
addition, Table 1-3 lists some regulatory standards for flame-retardant textiles,
including California Technical Bulletins 116, 117, and 133. Table 1-11 and Table
1-12 provide details on the comparative flame-retardant performance of
decaBDE and MWCNTs. Additionally, Table 1-13 has been added to clarify the
physicochemical properties of MWCNTs that are related to flame-retardant
performance.
210,211
Section 1.2.2 Two comments suggested
highlighting the factors influencing
selection of flame retardant and
the uncertainty involved in
MWCNT success in this
application compared to other
applications. One of these also
suggested that alternative
applications of MWCNTs will be
more prominent in the future than
the selected application.
Additional Information Highlight Box 1 was added to clarify the factors
influencing manufacturers' selection of flame-retardant materials. Another
Information Highlight Box was added to describe the uncertainty surrounding
which applications MWCNTs will be used in (Additional Information Highlight
Box 2). In addition, greater detail related to information considered in the
selection process, which was formerly in Appendix A, is now also included in
Chapter 1. New text describing the selection rationale is also included in
Chapter 1.
2, 3
56, 59
Chapter 1 Two comments suggested a more
detailed discussion on the
distinction between different
chemistries, surface
functionalizations, and classes of
MWCNTs rather than more
general discussion would improve
the document. Another comment
requested that the term
"functionalization" be defined.
A goal of the case study is to provide information on an emerging
nanomaterial—MWCNTs—illustrated through that material's use in a specific
application—flame-retardant textiles. Given the relative dearth of information on
specific formulations of MWCNTs used in flame-retardant coatings, or other
applications, information on a variety of MWCNT formulations is included in the
case study to provide a greater scope of details on the nanomaterial in general.
Additional Information Highlight Box 7 and Figure 3-1 have been added to
highlight variations in MWCNT formulations and functionalization, and how
these variations affect the product life-cycle. Table 1-13 was added describing
the chemistry/functionalization described in literature specific to use as a flame
retardant. Existing Text Box 1-1, Text Box 4-2, Text Box 5-1, and Table 2-4
contain generalized information regarding functionalization and how
physicochemical properties of MWCNTs affect release, environmental fate,
exposure, and toxicity. Functionalization is defined in Section 2.2.3.1.
2, 3
7, 9, 11
1-15
-------
Table 1-5 (Continued): Relevant to Chapter 1 (including Preface and Executive Summary).
Related to CEA Related to
Framework Response Global
Area Comment Theme Response Category Comment ID
Chapter 1 One comment noted that chirality The Agency appreciates the suggestion but the source used for information on 5 163
might not be a consideration for chirality, Gustavsson et al. (2011) mentions specifically the variation for
MWCNTs. A reference was MWCNTs. The literature provided by the commenter is for double-walled carbon
provided. nanotubes and was therefore not included in the case study.
1-16
-------
Table 1-6. Relevant to Chapter 2.
Related to CEA
Framework
Area Comment Theme
Response
Related to
Response Global
Category Comment ID
Product Life Two comments suggested
Cycle: Raw additional literature might be
Materials; available to fill in data gaps
Product Life pertaining to MWCNT feedstocks
Cycle: Material and replace the surrogate SWCNT
Synthesis information currently in the case
study.
"Product Life Cycle: Material Synthesis" was identified by workshop participants
as a Priority Research Area. A Priority Research Area Highlight Box and Section
6.3.1.1 were added; a targeted literature search was performed to inform the
sections. Data for SWCNTs were used to describe feedstock stages of the
product life-cycle because no data on MWCNT feedstock were identified at the
time. Because "Product Life Cycle: Raw Materials" was not identified as a
priority research gap, no additional focus was given to improving this portion of
the case study with new information. Healy et al. (2008), however, was
deemphasized to focus on just presenting the more recent information regarding
SWCNT feedstocks by Zhang et al. (2011b).
1,2, 3
83,214
Section 2.1.1 Two comments questioned the
logic regarding synthesis yields of
CNTs.
The reported values were verified in the literature.
84, 85
Chapter 2 Three comments identified
additional information that might
be available from NIOSH and
Albermale regarding material
synthesis, processing, and
potential release of decaBDE.
The purpose of including comparative information for decaBDE was to better
inform the collective judgment process by helping highlight data gaps and
research needs for MWCNTs; as noted in Section 1.1.2. detailed information on
decaBDE has been moved to Appendix H and replaced with succinct
comparison highlight text boxes to focus the Peer Review Draft on only the most
important and relevant information for the Priority Research Areas for MWCNTs
as identified by the workshop experts. As such, no additional information was
incorporated to enhance the discussion of decaBDE in the case study.
212,217,218
Chapter 2 One comment noted that not all
laboratories, particularly small
R&D facilities or older facilities,
have the same engineering
controls.
This text was revised and clarified. Additionally, "Product Life Cycle: Material
Synthesis" and "Exposure: Human: Occupational" were identified as Priority
Research Areas and this comment was considered and addressed through the
addition of Priority Research Area Highlight Boxes and Sections 6.3.1.1 and
6.3.3.1.
1, 3
216
1-17
-------
Table 1-6 (Continued): Relevant to Chapter 2.
Related to CEA
Framework
Area
Comment Theme
Response
Related to
Response Global
Category Comment ID
Section 2.5.4 Three comments noted a data gap
regarding the impact of MWCNTs
on wastewater treatment plants,
particularly in terms of suggested
evidence of antimicrobial activity.
Two of these also suggested
literature.
Additional consideration of the impact of MWCNT contamination on functionality
of wastewater treatment plants has been added in Additional Information
Highlight Box 5. Readers are referred to Section G.5.1.1.and Appendix Table
F-18 for details on the toxicity of MWCNTs to microbes. This comment was also
considered in developing text for Section 6.3.2.2 given that "Env TT&F: Waste
Water" was identified by workshop participants as a Priority Research Area.
1
18, 19, 72
Chapter 2 One comment noted a step during
Feedstock Extraction (cleaning
between reactor runs) that might
have been overlooked in
discussion of release rates during
this process.
Text was revised to clarify this point.
215
Product Life
Cycle: Disposal/
Recycling
One comment noted the possibility
of MWCNT being both bound or
free during end-of-life stages.
"Product Life Cycle: Disposal/Recycling" was identified by workshop participants
as a Priority Research Area, and this comment was considered and addressed
through the addition of a Priority Research Area Highlight Box and Section
6.3.1.5. Additionally, Figure 2-1 has been updated to illustrate the potential for
MWCNTs to become "free" during end-of-life stages of the life-cycle process.
1, 3
81
1-18
-------
Table 1-7. Relevant to Chapter 3.
Related to CEA
Framework
Area
Env TT&F
(Analytical
Techniques)
Comment Theme
Two comments provided
references pertaining to analytical
techniques for MWCNT in
environmental media and impacts
of release.
Response
Analytical techniques were discussed by workshop participants as related to
several research priorities within environmental transport, transformation, and
fate. As such, the suggested literature was incorporated into the related sections
of Section 6.3.2 or in Additional Information Hiahliaht Box 10, which was added
to describe the weaknesses of current analytical techniques as relevant.
Response
Category
1,2
Related to
Global
Comment ID
128, 139
Env TT&F One comment noted a data gap
regarding environmental
decomposition and transformation
of MWCNTs
Several aspects of environmental transport, transformation, and fate were
identified by workshop participants as Priority Research Areas. Priority
Research Area Highlight Boxes and Section 6.3.2 have been included.
The current lack of data noted by this comment is highlighted in the discussion
of these priority areas in Section 6.3.2.
201
Chapter 1 or 3 Eight comments discussed
MWCNT release, transformation,
and fate in the environment in
terms of potential variation in
polymer chemistry; references
were also provided.
Additional Information Highlight Box 6 and Additional Information Highlight Box 7
were added to Chapter 3 to address these issues; the provided literature was
included where relevant. In addition, release and processes that influence
release in various media were identified by workshop participants as a Priority
Research Area and thus discussed in Section 6.3.2.
127,219,
220,221,
222, 225,
226, 228
EnvTF&T: Soil
Three comments noted an
instance where the authors of a
reference appeared to be
speculating rather than stating
definitive evidence.
The statements were revised and clarified.
168, 169, 170
1-19
-------
Table 1-7 (Continued): Relevant to Chapters.
Related to CEA
Framework
Area Comment Theme
Response
Related to
Response Global
Category Comment ID
Env TT&F One comment discussed the
influence of environmental
parameters on the aggregation
process and toxicity of MWCNT.
Two references were provided.
As noted by workshop participants, uncertainty exists concerning how properties
of the environmental media influence MWCNT fate and transport, and
conversely, how MWCNT properties influence fate, transport, and toxicity. New
figures have been added to the main text of the document (Figure 2-1 and
Figure 3-1) showing variations in MWCNTs formulations and functionalization
throughout the lifecycle, and how various environmental properties can affect
MWCNT transport and fate. Additionally, readers are reminded of existing tables
in Appendix D that summarize studies on behavior of MWCNTs in various
environmental media; several of these studies identify environmental conditions
that influence MWCNT behavior. Text Box 1-1. Text Box 4-2. Text Box 5-1. and
Table 2-4 contain generalized information regarding how physicochemical
properties of MWCNTs affect release, environmental fate, exposure, and
toxicity.
2, 3
21,65,66
Env TT&F: One comment suggested literature
Soil/Sediment: pertaining to soil/sediment
Bioavailability bioavailability of MWCNT
"Env TT&F: Sediment: bioavailability" was identified by workshop participants as
a Priority Research Area. A Priority Research Area Highlight Box and Section
6.3.2.3 were added on this topic. The suggested literature was reviewed and
incorporated into Section 6.3.4.3.
161
Env TT&F: Air One comment suggested literature
(Analytical relevant to analytical techniques
Techniques) for urban air.
The suggested reference was included in Additional Information Highlight Box
10, which discusses analytical techniques for detecting MWCNTs in urban air.
92
Env TT&F One comment noted the use of The distinction between hydrophobicity and lipophilicity was identified as an 3 164
lipophilicity and hydrophobicity important research topic by workshop participants. Text was clarified as
was not always clear and that the necessary throughout the document.
sources used might be outdated.
Chapter 3 One comment suggested that
discussion of different
mechanisms of soil transport, for
example colloid-facilitated
transport, physical straining, and
site blocking, be included in the
case study.
Additional Information Highlight Box 8. which discusses transport mechanisms,
was added to the document.
22
Abbreviations: Env TT&F = Environmental Transport, Transformation, and Fate.
I-20
-------
Table 1-8. Relevant to Chapter 4.
Related to CEA
Framework
Area
Comment Theme
Response
Related to
Response Global
Category Comment ID
Exposure One comment noted occupational
Route: Human: exposure is a data gap for
Occupational MWCNTs; another suggested a
reference relevant to this topic.
"Exposure: Human: Occupational" was identified by workshop participants as a
Priority Research Area. A Priority Research Area Highlight Box and Section
6.3.3.1 were added for MWCNT occupational exposure, and the suggested
literature was incorporated.
1
38, 152
Exposure One comment noted a reference
Route: Human- to elucidate potential consumer
Consumer exposure to MWCNTs.
"Exposure: Human: Consumer" was identified by workshop participants as a
Priority Research Area, as denoted by a Priority Research Area Highlight Box
on this topic. The suggested literature was incorporated into Section 6.3.3.2.
197
Exposure
Route: Human
One commenter asked if there
was additional supporting
evidence for the findings of
Aschberger et al. (2010).
The statement attributed to Aschberger et al. (2010) was corroborated by an
additional source that is now included in the case study document.
223
Section 4.2.1 One comment noted that the
discussion of dermal absorption
might not reflect the most recent
and accurate science; further, the
comment indicated that discussion
can be drawn from information
available in previous case studies.
Information on dermal absorption was reviewed for accuracy. In addition,
because this area was identified as a priority research gap, discussion on
dermal absorption of nanomaterials from past case studies on Nano-Ag and
Nano-TiO2 was included in Additional Information Highlight Box 11 along with
information (as available) from a targeted literature search to present the most
accurate and current information on dermal absorption.
1,2, 3
110
Dose (Kinetics) One comment noted challenges
related to understanding
toxicokinetics of nanomaterials
due to limitations in traditional
analytical techniques. Another
suggested a reference regarding
interspecies scaling of
pharmacokinetics.
"Dose (Kinetics): Human" was identified by workshop participants as a Priority
Research Area; as such, information regarding limitations in standard "mass
concentration" metrics and analytical techniques was added to Additional
Information Highlight Box 9 to support research planning that involves
extrapolating kinetic and toxicity results from laboratory animals to humans.
1,2
260, 262
1-21
-------
Table 1-8 (Continued): Relevant to Chapter 4.
Related to CEA
Framework
Area
Comment Theme
Response
Response Related to
Category Global
Comment ID
Section 4.3
Dose (Kinetics):
Aquatic Biota—
Absorption
AND
Env TT&F:
Biota:
Bioaccumulation
Four comments noted that more
recent literature suggests that
MWCNTs do not appear to be
absorbed across the gut lining or
into other tissues, and therefore
are not expected to act in a similar
manner to bioaccumulative
substances. References were also
provided.
The provided literature has been added to Additional Information Highlight Box
G1 in Appendix G to further discuss uptake and absorption in aquatic food
webs.
2 174, 175,
177, 172
Dose: Human: Three comments suggested
Absorption literature to enhance the
discussion about absorption of
MWCNTs in humans.
The suggested literature was incorporated into Section 6.3.3.3. In addition,
"Dose: Human: Absorption" was included in a Priority Research Area Highlight
Box as it was identified by workshop participants as a Priority Research Area.
1 63,64,153
Chapter 4 One comment recommended the
ATSDR (2004) toxicological profile
for PBDEs as a reference.
The ATSDR document was included in several instances; however one of the
aims of this version of the case study was to reduce the focus on decaBDE and
instead shift focus toward highlighting only those aspects of decaBDE that could
be particularly useful to risk assessment of MWCNTs. Therefore, rather than
including an exhaustive review of the potential exposure scenarios and relevant
information for decaBDE by delving further into the ATSDR Toxicological Profile
for PBDEs, emphasis was placed on condensing decaBDE information.
157
Dose (Kinetics): One comment noted that a
Terrestrial Biota: reference describing
Absorption lipophilicity/accumulation potential
in roots was speculative rather
than definitive, and two references
were provided regarding the
absorption and impacts of
MWCNTs on terrestrial biota.
The text was reviewed and removed as appropriate. Although the Agency
appreciates the additional literature, it was not incorporated into the body of the
document because the area of "Impacts: Terrestrial Biota" was not identified by
workshop participants as a Priority Research Area.
3 150, 151, 176
Section 4.3 Two comments noted potential
errors in transcription from primary
sources.
The statements were revised and clarified.
173, 177
I-22
-------
Table 1-8 (Continued): Relevant to Chapter 4.
Related to CEA
Framework
Area Comment Theme
Response
Related to
Response Global
Category Comment ID
Text Box 4-1 One comment noted that there are
currently no mass-spectrometry
techniques for MWCNTs, and the
methodology described in the case
study measures only metal
concentrations still associated with
MWCNTs.
The statement was revised and clarified.
171
Dose: Aquatic One comment noted that the
Biota: discussion of potential absorption
Absorption of carbon materials in the gut was
not clear and further suggested
additional literature where
absorption across the gut tract
was not observed.
The statement was revised and clarified to accurately convey the science.
The provided literature has been added to the discussion in Appendix G.
Because uptake vs. absorption is a key issue that impacts toxicity, Additional
Information Highlight Box G1 also was included in Appendix G.
172
I-23
-------
Table 1-9. Relevant to Chapter 5.
Related to CEA
Framework
Area
Comment Theme
Response
Response
Category
Related to
Global
Comment ID
Chapter 5 One comment suggested an
alternative definition of "fibers" with
additional references.
A general discussion of aspect ratios for CNTs and comparisons to asbestos
has been added in Additional Information Highlight Box 13: this box includes one
of the references suggested by the commenter. The second reference pertained
to carbon nanofibers and was therefore not included. The discussion regarding
the 3:1 aspect ratio for fibers as defined by the World Health Organization was
not revised, however, because it represents the current measuring classification.
111
Chapter 5 One comment provided a
reference to replace SWCNT
surrogate data regarding acute
oral toxicity of MWCNTs.
The document was updated to reflect the provided literature. Data on SWCNTs
was included only when data on MWCNTs was extremely limited and when the
SWCNT data is highly relevant; SWCNT is clearly identified as such.
117
Section 5.1.5 One comment suggested
additional literature to support the
discussion related to dermal
sensitization of MWCNTs.
The discussion in Section 5.1.5 was revised to better reflect the scientific results
of the suggested literature. In addition, Pauluhn (2010) was reviewed and
included in the discussion. Table 5-1 already indicated that skin sensitization is
negative, so no additional revisions were made.
121
Section 5.1.6 One comment suggested an
additional reference on
reproductive toxicity that had
originally been omitted due to the
less relevant route of exposure.
Text was revised to clarify that no key studies were identified for reproductive
effects related to MWCNT exposure. Although the use of a less relevant
exposure route (i.e., intravenous injection) in the study by Bai et al. (2010)
precluded its use as a key study, the area of "Impacts: Human: Reproductive/
Developmental" was identified as a Priority Research Area, and thus this study
has been noted in Section 6.3.4.2.
115
Human Impacts: Four comments reiterated the
Noncancer influence of physicochemical
properties on toxicokinetics and
toxicity. One comment suggested
carboxylated CNTs are more likely
to degrade and therefore are less
likely to induce profibrogenic
effects, as colloid-facilitated
transport is an important factor.
"Impacts: Human" was identified by workshop participants as a Priority
Research Area. Additional information was incorporated into Section 6.3.4.
Additionally, Text Box 5-1 describes influences of physicochemical properties on
MWCNT toxicity. The Agency reviewed the suggested literature and determined
that making conclusions regarding whether dispersion state/carboxylated
MWCNTs influence fibrosis based on this study is difficult, as metal
concentration varied with degree of carboxylation. Therefore, it was not added to
the document.
1,2
261,263,
264, 265, 10
I-24
-------
Table 1-9 (Continued): Relevant to Chapters.
Related to CEA
Framework
Area Comment Theme
Response
Related to
Response Global
Category Comment ID
Impacts: Human Seventeen comments noted
significant advancements in
complex in vitro models that have
resulted in reproducible effects
and showing good concordance
with in vivo data. Comments
suggested such approaches would
prove useful for MVVCNT
assessment in the future.
"Impacts: Human" was identified by workshop participants as a Priority
Research Area. Additional Information Highlight Box 15 was added to highlight
the trend toward developing innovative biologically/toxicologically relevant in
vitro models.
275, 276,
277, 278,
279, 280,
281,282,
283, 284,
285, 286,
287, 288,
289, 290, 291
Section 5.2.2.1 Two comments recommended
references pertaining to
accumulation in terrestrial biota.
The suggested literature was reviewed and incorporated as deemed relevant
into Additional Information Highlight Box G2 in Appendix G, which discusses
toxicity to terrestrial invertebrates.
182, 183
Impacts: Human One comment noted the results
found in intratracheal instillation
studies in mice corroborated those
reported via inhalation.
Four studies on intratracheal instillation are currently included in Section 5.1.3
along with a comment that these studies show similar endpoints as the
inhalation studies. In addition, seven alternative exposure route pulmonary
studies are included in Appendix Table F-6. Additionally, Additional Information
Highlight Box 14 was added, which discusses inhalation study designs for
MWCNTs.
37
Impacts: Human Ten comments questioned the
applicability of traditional in vivo
toxicological (particularly
inhalation) models for MWCNTs;
several references were provided
noting potential
weaknesses/confounding factors.
"Impacts: Human" was identified by workshop participants as a Priority
Research Area. References and discussion regarding the
complexity/confounding nature of MWCNTs in traditional in vivo models were
included in Additional Information Highlight Box 14. Additionally, Additional
Information Highlight Box 15 was added regarding the trend toward the
development of innovative toxicologically/biologically relevant in vitro models.
1,2, 3
259, 266,
267, 268,
269, 270,
271,272,
273, 274
Impacts: One comment noted respiratory
Human: sensitization as an example of
Noncancer where conflicting data exists, but
the consensus of most evidence
should be highlighted.
Throughout the document, information highlight text boxes have been added to
the document to help highlight data gaps and present a balanced representation
of evidence of a finding, or lack thereof. Specific to this comment, "Impacts:
Human: Noncancer" was identified by workshop participants as a Priority
Research Area, which corresponds with the identification in this comment of
respiratory sensitization as an example of an area with conflicting data.
Noncancer respiratory effects, including inflammatory changes and immune
responses characteristic of respiratory sensitization, are discussed in Section
5.1.3 and Additional Information Highlight Box 13 further discusses available
evidence on health impacts from MWCNT inhalation exposures.
125
I-25
-------
Table 1-9 (Continued): Relevant to Chapters.
Related to CEA
Framework
Area Comment Theme
Response
Related to
Response Global
Category Comment ID
Impacts:
Aquatic Biota
Table 5-2, Table
5-4, Table 5-5
Seven comments suggested
additional literature regarding the
impacts of MWCNTs to benthic
invertebrates, soil/sediment
organisms, or other aquatic biota.
Impacts: Aquatic Biota" was identified by workshop participants as a Priority
Research Area. The topic was thus highlighted in a Priority Research Area
Highlight Box and discussed in Section 6.3.4.3. Additional Information Highlight
Box 17 was added regarding toxicity to benthic and aquatic invertebrates in
Section 5.2.1.1, incorporating provided literature where applicable. Additionally,
Table 5-2 was updated to include the rest of the suggested literature (literature
was split up into each of these action elements to avoid redundancy). Specific to
comment 178, Shen et al. (2012) was not included in the added text because
the primary focus involved the bioaccumulation of polycyclic aromatic
hydrocarbons.
1,2, 3
126, 138,
178, 180,
181, 182, 183
Section 5.2.1
One comment identified an
improperly cited statement.
The text was revised.
179
I-26
-------
Table 1-10. Relevant to Chapter 6 and Appendices.
Related to CEA
Framework
Area
Comment Theme
Response
Related to
Response Global
Category Comment ID
Chapter 6 One comment suggested the usefulness of
highlighting the information leading to voluntary
phase-out of decaBDE.
Text was revised in Section 1.1.3 to further emphasis the type
of information considered in decisions surrounding the voluntary
phase-out of decaBDE.
74
Impacts: Three comments identified additional literature
Human: pertaining to SWCNT that might be useful as
Cancer, surrogate or supportive data for MWCNTs. One
Noncancer comment stated that major studies were presented
already, and that the additional SWCNT studies
would not be relevant.
Studies on SWCNTs were not the focus of this case study.
Some information on SWCNTs, however, is provided in the
various appendix tables when information on MWCNTs did not
exist, especially if this area was considered a Priority Research
Area.
96, 116, 118,
119
Appendix J One commenter felt summary tables and figures
contained too much information to be useful as
summary or highlight text.
Efforts to clearly convey information in summary tables and
figures is part of ongoing work to refine the CEA approach. See
Appendix J for an example of current pilot-stage efforts to use
knowledge maps to convey more clearly information in tables
and text.
149
I-27
-------
Table 1-11. Comments that required no action.
Related to CEA
Framework Area
Comment Theme
Agency Response
Related to Comment ID
General
One comment recommended that inferences be more clearly
stated as such.
The Agency appreciates the feedback; however, effort was
made in the original draft to distinguish inferences (i.e.,
"could occur") from documented information (i.e., "has been
shown to occur"). This decision was made explicitly to avoid
implying that any inferences were supported by documented
scientific data; rather, such inferences were drawn from
literature that, in the absence of concrete evidence specific
to MWCNTs, suggested the likelihood of certain statements
occurring. As such, no action was taken in response to this
comment.
109
General
One commenter stated that the science was general
comprehensive but some areas could be improved. They did not
indicate which areas.
In revising the document, efforts were made to improve key
areas identified as Priority Research Areas; however, no
specific revisions have been made in regards to this
comment.
120
General
14 commenters stated that the science was accurately conveyed
in the case study document.
The Agency appreciates the feedback.
26, 52, 67, 147, 32, 58, 77,
97, 101, 105, 130, 134, 155,
189
General
20 commenters stated that the comparison of MWCNT to
decaBDE was useful. Commenters defended this position with a
variety of reasons, such as the comparison helped to identify
research gaps, helped the reader understand the process of
regulating a chemical and identifying knowledge gaps, or
illustrating the manufacturing process and exposure scenarios for
flame-retardant upholstery.
The Agency appreciates the feedback.
33,55,60,123,124,141,
184, 196, 235, 69, 78, 86,
94, 102, 106, 131, 136, 148,
156, 190
I-28
-------
Table 1-11 (Continued): Comments that required no action.
Related to CEA
Framework Area
Comment Theme
Agency Response
Related to Comment ID
Chapter 1 One commenter stated that the science was accurately portrayed
however the Chapter on characterization was not complete.
No specific revisions were identified to address this
comment; however, efforts were made throughout the
document to present complete information.
89
Product Life
Cycle:
Release Rate /
Impacts
One commenter stated that the science was accurately portrayed
however some issues of terminology could be clarified.
No specific revisions were identified to address this
comment; however, efforts were made throughout the
document to use consistent terminology, particularly in
relation to specific types of MWCNT emissions and
toxicological effects, when such information was available.
93
Env TT&F: Two comments suggested additional literature pertaining to
Soil Env TT&F in wastewater and soil that were already included in the
case study.
Zhang et al. (2011 a) was already included in Section 3.3.1
(transport, transformation, and fate n surface water and
sediment). Holbrook et al. (2010) was already included in
Appendix Table D-2, which summarizes studies relevant to
MWCNT fate and transport in aqueous media, including
sediment. The two studies were not added to Section 3.4
because Env TT&F: Soil was not identified by workshop
participants as a Priority Research Area.
166,167
Env TT&F: One commenter called out a section of the preface which they felt
Soil: was not fully representative of the body of science. A second
bioavailability commenter asked for more detail about surface chemistry of
MWCNTs in the preface.
No revisions were made to the preface because the preface
is not intended to include this level of detail.
161,200
Env TT&F:
Soil / Impacts:
Other
Two commenters provided literature already included in the case
study or specific to SWCNTs.
The Agency appreciates the suggestion. However, the
literature was not incorporated into the Case Study because
the references are not specific to MWCNT and the behavior
of metal oxide nanomaterials is not considered appropriate
to predict the effects of MWCNT.
71,144
Env TT&F One comment provided a reference pertaining to MWCNT in
(Aqueous media) aqueous media.
The suggested reference was previously included in the
case study in Section 3.3.1.
143
None
16 comments were received that did not require agency action. The Agency appreciates the feedback.
3,4,20,194,198,208,231,
232,236,237,241,245,
249, 255, 296
I-29
-------
Table 1-11 (Continued): Comments that required no action.
Related to CEA
Framework Area Comment Theme
Agency Response
Related to Comment ID
None
12 commenters stated that there were no additional literature they The Agency appreciates the feedback.
were aware of that should be included in the case study.
25, 31, 57, 76, 80, 88, 100,
104, 108, 133, 159, 188
None
10 commenters stated that they did not have any comments on The Agency appreciates the feedback.
improving the document, or that their comments for improvements
were incorporated into the other charge question responses.
5, 95, 1 87, 70, 87, 1 07, 1 32,
142, 158, 191
I-30
-------
1.3. Full Comment Excerpts
Table 1-12. Comments received in response to the expert charge questions.
Comment Author Charge
ID ID Question Comment Text Excerpt Themes
25 E1 1 I think this report was thorough and included all available literature on MWCNTs. No additional literature
26 E1 2 I believe that the science was presented in an objective and clear manner. Science accurate
27 E1 3 Further, I believe this comparison made things confusing, since it is impossible to compare the fate Comparison NOT
and transport and effects of two types of compounds that are completely different in their chemistry useful; decaBDE was
and physical properties. So what has been found for flame retardants in relation to their movement in bad comparison
the environment, half-life, uptake by biota, and ultimately effects, is rather meaningless when applied choice; fundamentally
to MWCNTs. different compounds
don't overlap
28 E1 3 No, this comparison was not useful. I am not sure why this was done, but have the impression that it Comparison NOT
was used to "beef" up the document since so little information exists on MWCNTs. useful; added length
29 E1 4 Remove the flame retardant vs. MWCNTs comparison. As already mentioned, this comparison Comparison NOT
doesn't help when trying to elucidate research gaps as related to MWCNTs. useful; highlighted data
gaps
30 E1 4 Also, the use of MWCNTs as flame retardants has not been materialized, so why solely focus on this Choice of nanomaterial
particular application? and application;
application not
currently in use
31 E2 1 At this point in time, I must admit that I cannot think of any studies that might help and I find the Draft No additional literature
Case Study Document to be very comprehensive
32 E2 2 To the best of my knowledge the science is accurately conveyed Science accurate
33 E2 3 I do think that the comparison makes sense that it is good to have facts and data on decaBDE to Comparison WAS
measure information up against. useful; highlighted data
gaps
1-31
-------
Table 1-12 (Continued): Comments received in response to the expert charge questions.
Comment
ID
Author Charge
ID Question
Comment Text Excerpt
Themes
34 E2 3 My only concern is that decaBDE is a problematic substance and bans and limitations have been put Comparison NOT
on its use and hence comparing MWCNTs with such a problematic substance would inevitably make useful; decaBDE was
MWCNTs look good bad comparison choice
35
E2
Somehow I feel that the purpose of "identifying and prioritizing research" is too modest a scope for all
the work that has been put into drafting this comprehensive report. Identification of the most
appropriate manner in which to regulate MWCNTs, it would be in line with effort put into this.
CEA framework/
methodology; purpose
of CEA
36
E3
See response to question # 2. In general, the draft document cites most of the relevant published
literature on MWCNT that would permit the identification of research gaps.
No additional literature
37
E3
The review of the health data on MWCNT described in Chapter 5 and summarized in Table 5-1,
doesn't completely capture all the relevant evidence regarding adverse pulmonary effects from
exposure to MWCNT. Studies with mice and rats exposed to MWCNT by pharyngeal aspiration and
by intratracheal administration support findings of pulmonary irritation, granulomas, alveolar septal
thickening, and pulmonary fibrosis that were consistent with effects observed in inhalation studies.
Highlight data gap;
additional
considerations for
section
38
E3
40
E4
In the absence of occupational exposure information on decaBDE and MWCNT it's not possible to
determine the potential extent of exposure or exposure characteristics (e.g., physical and chemical
state of MWCNT, other contaminants). The absence of such data makes it's difficult to assess the
potential hazard risk.
Highlight data gap
39 E3 4 See comment to # 3 regarding focus of document. It would have been informative if the authors of the Highlight data gap;
document would have listed at the end of each Chapter their perspective as to research data gaps. suggestion for
improvement
"Synergistic effect of carbon nanotubes and decabromodiphenyl oxide/Sb2O3 in improving the flame
retardancy of polystyrene" Lu, H.; Wilkie, C. A. Polym. Degrad. Stab. 2010, 95, 564-571.
Choice of nanomaterial
and application;
alternative products/
formulations
41
E4
"Role of Surface Interactions in the Synergizing Polymer/Clay Flame Retardant Properties" Pack, S.; Choice of nanomaterial
Kashiwagi, T.; Cao, C.; Korach, C. S.; Lewin, M.; Rafailovich, M. H. Macromolecules 2010, 43, 5338- and application;
5351. alternative products/
formulations
I-32
-------
Table 1-12 (Continued): Comments received in response to the expert charge questions.
Comment Author Charge
ID ID Question Comment Text Excerpt Themes
42 E4 1 "Segregation of Carbon Nanotubes/Organoclays Rendering Polymer Blends Self-Extinguishing" Pack, Choice of nanomaterial
S.; Kashiwagi, T.; Stemp, D.; Koo, J.; Si, M.; Sokolov, J. C.; Rafailovich, M. H. Macromolecules 2009, and application;
42, 6698-6709. alternative products/
formulations
43 E4 1 "Nanoclay and carbon nanotubes as potential synergists of an organophosphorus flame retardant in Choice of nanomaterial
poly(methyl methacrylate)" Isitman, Nihat AN; Kaynak, Cevdet Polym. Degrad. Stab. 2010, 95, 1523- and application;
1532 alternative products/
formulations
44 E4 1 "Layered silicate polymer nanocomposites: new approach or illusion for fire retardancy? Investigations General information on
of the potentials and the tasks using a model system" Bartholmai, M.; Schartel, B. Polymers for flame retardancy of
Advanced Technologies 2004, 15, 355-364. nanocomposites
45 E4 1 "Filler blend of carbon nanotubes and organoclays with improved char as a new flame retardant Choice of nanomaterial
system for polymers and cable applications" Beyer, G. Fire Mater. 2005, 29, 61-69. and application;
alternative products/
formulations
46 E4 1 "Flame retardancy of nanocomposites based on organoclays and carbon nanotubes with aluminum Choice of nanomaterial
trihydrate" Beyer, G. Polym. Adv. Technol. 2006, 17, 218-225. and application;
alternative products/
formulations
47 E4 1 "Fire behavior of polyamide 6/multiwall carbon nanotube nanocomposites" Schartel, B.; Potschke, P.; General information on
Knoll, U.; Abdel-Goad, M. European Polymer Journal 2005, 41, 1061-1070. flame retardancy of
nanocomposites
48 E4 1 "Some comments on the main fire retardancy mechanisms in polymer nanocomposites" Schartel, B.; General information on
Bartholmai, M.; Knoll, U. Polym. Adv. Technol. 2006, 17, 772-777. flame retardancy of
nanocomposites
49 E4 1 "Flame retarded polymer layered silicate nanocomposites: a review of commercial and open literature General information on
systems" Morgan, A. B. Polym. Adv. Technol. 2006, 17, 206-217. flame retardancy of
nanocomposites
50 E4 1 "Flammability reduction of flexible polyurethane foams via carbon nanofiber network formation" General information on
Zammarano, M.; Kramer, R. H.; Harris, R.; Ohlemiller, T. J.; Shields, J. R.; Rahatekar, S. S.; Lacerda, flame retardancy of
S.; Gilman, J. W. Polym. Adv. Technol. 2008, 19, 588-595. nanocomposites
I-33
-------
Table 1-12 (Continued): Comments received in response to the expert charge questions.
Comment
ID
Author Charge
ID Question
Comment Text Excerpt
Themes
51
E4 1
While the papers by Schartel above do not always look at nanotubes, they help make it clear why
nanotubes alone are not enough to obtain a passing flammability result in most regulatory tests.
Therefore, nanotubes alone will not be useful as a flame-retardant product unless they are combined
with other flame retardants, which is a very different product than that outlined in this draft document.
Choice of nanomaterial
and application;
alternative products/
formulations;
application not
currently in use
52
E4
In regards to the science being accurately conveyed in the document, the answer is yes and no. Yes
in that the MWCNT chemistry and known toxicity and its complexity is accurately captured, ...
Science accurate
53
E4
In regards to the science being accurately conveyed in the document, the answer is yes and no ... but
no in that the application for MWCNT which would drive its use is very wrong. The driving factors for
flame-retardant solution use (why an industry would use the technology) are completely missing, and
therefore the choice of MWCNT in textile back-coatings is unfortunately incorrect. MWCNT will be
used in greater quantities in other applications, not in the one selected in this draft case study.
Choice of nanomaterial
and application
Application not
currently in use
54
E3
Most, if not all, of MWCNT use in the United States is in research laboratories or small scale pilot
manufacturing processes [Schubauer-Berigan et al. 2011: Engineered carbonaceous nanomaterials
manufacturers in the United States: workforce size, characteristics, and feasibility of epidemiologic
studies. J Occup Environ Med 53 (Suppl 6):S62-S67]. The draft document could have just focused on
the toxicology data (e.g., in vitro, in vivo) with MWCNT (and other CNTs) which would have provided
an equally sufficient amount of information to permit a determination of research needed to support
future assessments and risk management decisions.
Choice of nanomaterial
and application;
application not
currently in use
55
E4 3
For the comparison of MWCNT to decaBDE, the document does and does not identify the research
gaps. It correctly represents the research gaps in regards to nanotube toxicity, exposure and
release....
Comparison WAS
useful; impacts;
exposure
56
E4 4
For specific comments on how to improve the document, the document needs to explain better why
certain flame retardants are chosen for an application and why CNT will not be used, but something
else may be - or better yet, what applications are more likely to have a high percentage of using
MWCNT in their application, and what the risk factors are associated with that application. MWCNT
are far more likely to be used in polymer composites than in textiles for furniture, and that is a
completely different exposure model, especially if most of the composites enter into aerospace use.
Choice of nanomaterial
and application
;suggestion for
improvement
57
58
E5 1
No, I am unaware of any other studies.
No additional literature
E5 2
Yes, the science is conveyed accurately throughout the document.
Science accurate
I-34
-------
Table 1-12 (Continued): Comments received in response to the expert charge questions.
Comment
ID
Author Charge
ID Question
Comment Text Excerpt
Themes
59
E4
For the comparison of MWCNT to decaBDE, the document does and does not identify the research
gaps ... does not correctly show what applications that MWCNT will be used in, nor does it show the
uncertainty around which applications MWCNT will be used in that would drive studies like this.
Comparison NOT
useful; choice of
nanomaterial and
application; application
not currently in use
60
E5
Yes and no, yes because it is very nice to have the comparison because it is much easier to see the
data gaps for MWCNT when compared to a well-studied material.
Comparison WAS
useful; highlighted data
gaps
61
E5
The document was very thorough and well written, however, for volunteers to read, comprehend, and
analyze all the data gaps it was quite time consuming, especially with tight timelines. Certain sections
of the document were a little verbose and could be shortened. ... Perhaps some of the data in text
could be put into a tabular form to reduce the length of the document but allow the reader to visually
see where the data gaps are.
Added length
62
E5
As mentioned in the previous response, the document did a wonderful job pointing out the data gaps Comparison NOT
for MWCNTs but it seemed that the document focused more on decaBDE due to the excess of useful; highlighted data
information available for the chemical. The decaBDE data could be distilled down a bit to shorten the gaps; decaBDE focus
document.
63
E6 1
"Evaluation of the interactions between multiwalled carbon nanotubes and CACO-2 cells," by Clark, Additional
KAetal. DOI 10.1080/15287394.2011.589105 (relevant to Chapter 4 and/or 5) considerations for
section
64 E6 1 "Cell permeability, migration, and reactive oxygen species induced by multiwalled carbon nanotubes in Additional
human microvascular endothelial cells," by Pacurari, M et al. DOI 10.1080/15287394.2012.625549 considerations for
(relevant to Chapter 4 and 5) section
65
E6 1
"Impact of Porous Media Grain Size on the Transport of Multi-walled Carbon Nanotubes," by Mattison, Additional
NJetal, DOI 10.1021/es2017076 (relevant to Chapters 2 and 3) considerations for
section
66
E6
"Sorption of Peat Humic Acids to Multi-Walled Carbon Nanotubes," by Wang, XL et al. DOI
10.1021/es202258q (relevant to Chapter 3)
Additional
considerations for
section
67
E6
I was impressed by the thoroughness and accuracy of the science presented in the document.
Science accurate
I-35
-------
Table 1-12 (Continued): Comments received in response to the expert charge questions.
Comment
ID
Author Charge
ID Question
Comment Text Excerpt
Themes
68 E5 3 Yes and no ... No, because I felt like the information about decaBDE started to overtake the document Comparison NOT
because there are not significant amounts of information available for MWCNTs for all the scenarios useful; decaBDE focus
covered in the case study.
69
E6 3
I was initially skeptical about the ability of decaBDE to help identify the research gaps in MWCNTs
due to the unique properties of nanomaterials; however, I grew to think it was an excellent model for
comparison as went through the process.
Comparison WAS
useful; fundamentally
different compounds
don't overlap
70
E6 4
I don't have any specific comments on improving the document.
71
E7 1
I am not aware of additional studies specifically on MWCNT, but rather two papers that demonstrate
the potential issues that future research needs to address: 2. Soybean susceptibility to manufactured
nanomaterials with evidence for food quality and soil fertility interruption
John H. Priester, Yuan Ge, Randall E. Mielke, Allison M. Horst, Shelly Cole Moritz, Katherine
Espinosa, Jeff Gelb, Sharon L. Walker, Roger M. Nisbet, Youn-Joo An, Joshua
P. Schimel, Reid G. Palmer, Jose A. Hernandez-Viezcas, Lijuan Zhao, Jorge L. Gardea-Torresdey,
and Patricia A. Holden. PNAS Plus: Soybean susceptibility to manufactured nanomaterials with
evidence for food quality and soil fertility interruption. PNAS, August 20, 2012 DOI:
10.1073/pnas. 1205431109
Recent literature
published; additional
considerations for
section
Abstract Based on previously published hydroponic plant, planktonic bacterial, and soil microbial
community research, manufactured nanomaterial (MNM) environmental buildup could profoundly alter
soil-based food crop quality and yield. However, thus far, no single study has at once examined the
full implications, as no studies have involved growing plants to full maturity in MNM-contaminated field
soil. We have done so for soybean, a major global commodity crop, using farm soil amended with two
high-production metal oxide MNMs (nano-CeO2 and -ZnO). The results provide a clear, but
unfortunate, view of what could arise over the long term: (i) for nano-ZnO, component metal was taken
up and distributed throughout edible plant tissues; (ii) for nano-CeO2, plant growth and yield
diminished, but also (iii) nitrogen fixation—a major ecosystem service of leguminous crops—was shut
down at high nano-CeO2 concentration. Juxtaposed against widespread land application of
wastewater treatment biosolids to food crops, these findings forewarn of agriculturally associated
human and environmental risks from the accelerating use of MNMs.
I-36
-------
Table 1-12 (Continued): Comments received in response to the expert charge questions.
Comment
ID
Author Charge
ID Question
Comment Text Excerpt
Themes
72
E7
I am not aware of additional studies specifically on MWCNT, but rather two papers that demonstrate
the potential issues that future research needs to address 1) The use of CNT as treatment
technologies for contaminants in wastewater treatment applications. While CNT's may be effective in
removing the contaminants, without further study, the CNTs (or MWCNT) could potentially damage
the waste water treatment capability or as noted in the second paper, become part of the sewage
sludge that is then applied to agricultural crops. Even with relatively non-toxic materials such as nano
ZnO, some crop damage was demonstrated.
Recent literature
published; additional
considerations for
section
Carbon nanotubes -the promising adsorbent in Wastewater treatment 2007 J. Phys.: Conf. Ser. 61
698 (http://iopscience.iop.Org/1742-6596/61/1/140 ) Y. H. Li,1* Y. M. Zhao,1 W. B. Hu,1 I. Ahmad, 1 Y.
Q. Zhu,1 X. J. Peng,2 Z. K. Luan2
Abstract. Carbon materials are a class of significant and widely used engineering adsorbent. As a new
member of the carbon family, carbon nanotubes have exhibited great potentials in applications as
composite reinforcements, field emitters for flat panel display, sensors, energy storage and energy
conversion devices, and catalysts support phases, because of their extraordinary mechanical,
electrical, thermal and structural properties. In particular, the large specific surface areas, as well as
the high chemical and thermal stabilities, make carbon nanotubes an attractive adsorbent in
wastewater treatment. The adsorption properties of the carbon nanotubes to a series of toxic agents,
such as lead, cadmium and 1,2-dichlorobenzene have been studied and the results show that carbon
nanotubes are excellent and effective adsorbent for eliminating these harmful media in water.
The effects of the morphologies and the surface status on the carbon nanotube adsorption capacities
are also discussed.
73 E7 2 The document seems to be very well done, thorough, and comprehensive. The organization of the Comparison WAS
document—alternating between decaBDE and MWCNT is discordant and a challenge to read when useful; fundamentally
trying to identify the issues for MWCNT-polymers as flame retardants. These two materials are not different compounds
related, except to provide a function. Having the information for decaBDE available is valuable to don't overlap
understand what the high-level issues are and what to be concerned about for the MWCNT-polymers,
but segregating the information would have made reading and referencing the document easier.
74
E7
See response to Q2. The information presented about decaBDE was interesting, but it would have
been helpful if the information was highlighted that led to the voluntary agreement to stop using
decaBDE. This would have highlighted what the issues were for deca and to make sure they were
avoided for MWCNT-polymer flame retardants.
Comparison NOT
useful; suggestion for
improvement
I-37
-------
Table 1-12 (Continued): Comments received in response to the expert charge questions.
Comment
ID
Author Charge
ID Question
Comment Text Excerpt
Themes
75
E7
Not the document per se, but rather the choice of the material for this CEA evaluation. The information
about MWCNT-polymer flame retardants is so uncertain and so little known about what the
characteristics are of the MWCNT-polymers that may be used as flame retardants that the exercise
has not elucidated unique approaches, but rather what should be done for any new material that will
have wide distribution and application. Thus, so much of the priority research is recommended in the
Confidence Matrix in the "Red" box—important and not confident. A study of MWCNT alone may be
more productive and provide a better research agenda.
Choice of nanomaterial
and application
76
E8
no
No additional literature
77
78
E8
yes
Science accurate
E8
Yes, and the comparative assessment is helpful.
Comparison WAS
useful
79
E8
82
E9
Link (at least conceptually) to Value of Information analysis.
Highlight data gap;
suggestion for
improvement
80
81
E9
E9
1
2
No, I don't.
In Fig 2-1 at End of Life MWCNT Release Form is only considered as Matrix Bound, but if they might
be Free at every other life cycle stage, including In Use, they may also become Free at End of Life.
No additional literature
-
In the first paragraph of 2.1.1 it is declared that "No data were found on the energy and resource
demands of raw material extraction for synthesis of decaBDE." It seems that this information should
be obtainable.
83
E9
In the 2nd paragraph of 2.1.1., LCA inputs from SWCNT are used, but it occurs to me that just as a
very specific application of MWCNTs is selected to focus on for the CEA, so should the data inputs
such as energy used in LCA and cost be kept consistent with the material in question. The low yield/
high cost of SWCNTs is not relevant. Focus on the synthesis method used for the current commercial
source of MWCNTs for flame-retardant upholstery coatings or similar methods that may be used for
that purpose.
84
E9
In the 2nd paragraph of 2.1.1: ...yields of CNT synthesis reactions are listed as ranging from 1% to
17,900%. How is a yield of >100% physically possible?
I-38
-------
Table 1-12 (Continued): Comments received in response to the expert charge questions.
Comment Author Charge
ID ID Question Comment Text Excerpt
Themes
85
E9 2
I find the claim in the 2nd paragraph of 2.1.1 that"... synthesis of MWCNTs requires larger amounts of -
precursor material than SWCNTs" to be dubious. The fact that MWCNTs are much cheaper to
produce than SWCNTs would not support the claim that they are lower yield.
86
E9 3
Yes. With the same application and similar concerns regarding types of risks, the decaBDE is a
valuable existing case study to inform evaluation of MWCNT.
Comparison WAS
useful
87
E9 4
no comment
88
89
E10 1
I don't know of any additional studies that would be appropriate.
No additional literature
E10 2
I thought that the science was accurately portrayed. Although, the Chapter on characterization was
not complete.
Science accurate
90
E10
The materials are quite different in mode of action and potential concerns from an environmental
standpoint. Possibly, comparison of MWCNT to existing treatments would be quite useful to
understand the requirements of the application. However, I didn't find the detailed comparison of the
two materials illuminating because of the quite different properties. From a practical standpoint, the
additional information on decaBDE created a very thick document. (I like to work from paper and it
was hard to carry around.)
Comparison NOT
useful; added length;
fundamentally different
compounds don't
overlap
91
E10
A technical summary would have been useful.
Suggestion for
improvement
92
E11
There are several new ES&T articles on CNT analysis - here is one Doudrick et al. (DOI:
10.1021/es300804f)
Detection/measuremen
t analytics; additional
considerations for
section
93
E11
Yes. The question is really around terminology. A major issue is emission rates and toxicity. I
understand the life-cycle perspective, but I think it is really important to get at these emission rates
and toxicity.
Science accurate;
purpose of CEA
94
E11
Yes
Comparison WAS
useful
I-39
-------
Table 1-12 (Continued): Comments received in response to the expert charge questions.
Comment Author Charge
ID ID Question Comment Text Excerpt
Themes
95
E11
The framework is fine for pre-workshop. There are rapidly developing sciences around this issue
which should be included as references.
Recent literature
published; purpose of
CEA; suggestion for
improvement
96
97
98
99
100
101
102
103
104
105
106
E12
E12
E12
E12
E13
E13
E13
E13
E14
E14
E14
1
2
3
4
1
2
3
4
1
2
3
No, I believe the major studies were presented. There are some studies of SWCNT, but they have
such different properties that I don't think they would be useful.
Yes, the information on MWCNT is done well.
No, it is more of a distraction than anything. If the intent were to consider synergistic interactions, that
might have been helpful. However, that was not the case. There is virtually nothing in the discussion
of decaBDE that transfers to MWCNT. The only thing that the inclusion did was confuse the issues.
Had all of the extraneous material been eliminated, the document would have been much smaller and
mostly would have exposed the huge gaps in what is known about MWCNTs
See above - eliminate all of the decaBDE discussion.
No.
Yes.
Yes.
The document is long and detailed. It will be best to have a short pare that summarizing the main
conclusion part at the beginning or the end of the document.
I don't know of any additional studies.
Yes.
Yes.
No additional literature
Science accurate
Comparison NOT
useful; highlighted data
gaps; fundamentally
different compounds
don't overlap
Comparison NOT
useful; decaBDE focus
No additional literature
Science accurate
Comparison WAS
useful
Added length;
suggestion for
improvement
No additional literature
Science accurate
Comparison WAS
useful
107
E14
No.
I-40
-------
Table 1-12 (Continued): Comments received in response to the expert charge questions.
Comment
ID
108
109
Author
ID
E15
E15
Charge
Question
1
2
Comment Text Excerpt
a) No. I am not aware of any published or unpublished data.
I find the document heavily weighted toward supposition, e.g., exp
Themes
No additional literature
osure "could" occur; MWCNT Suggestio
n for
110
E15
"could" be released. This language leaves the reader believing that such phenomena have already improvement
been documented. This is incorrect. Examples: Page 72, Iines14, 19, 20, 27 are examples of "...could
occur..." It would be more appropriate to indicate that while potential exposures have been identified,
none have been supported by actual data.
There are occasional instances where outdated references have been cited when more recent
information speaks otherwise. For example, page 120, line 16 refers to a paper by Maynard (2004)
that speaks about the potential dermal exposure to CNTs. Numerous studies on the sunscreen
nanomaterials, TiO2 and ZnO, have shown that while dermal exposure might occur, dermal
penetration is unlikely. Therefore, dermal uptake should be updated to reflect the more recent
information.
111
E15 2
Page 135, line 4 cites the WHO definition of a fiber, referring to the aspect ratio of 3:1. However, data
for small particles such as CNTs suggests that the aspect ratio of 20:1 is more appropriate (DeLorme
et al, 2012; Schinwald et al, 2012).
112
E15 3
a) Not really. The comparison of a molecule to a particle does not help - each has its own properties
that drive the exposure assessment and risk management procedures. The physical properties dictate
exposure and risk management. For example, exposure to a semi-volatile molecule is different than
exposure to a particle that can agglomerate and grow in size when it is suspended in air. It would be
better to focus solely on MWCNTs and the properties that dictate toxicity, exposure, etc.
Comparison NOT
useful; decaBDE was
bad comparison
choice; fundamentally
different compounds
don't overlap
113
E15 4
Remove the comparison to decaBDE. This substance can be introduced as a justification for the
industry to look at other substances that may provide flame retardation, but it does not provide a
framework for the assessment of MWCNTs.
Comparison NOT
useful
114
E15 4
b) Describing potential exposure scenarios is important, but should be followed by either identification Highlight data gap;
of data gaps and research ideas, or prioritizing those scenarios using the available information. For suggestion for
example, can information about release of CNTs from other media be used to give a likelihood of improvement
release of CNTs from textiles? Using release of other substances is inappropriate.
1-41
-------
Table 1-12 (Continued): Comments received in response to the expert charge questions.
Comment
ID
Author Charge
ID Question
Comment Text Excerpt
Themes
115
E16
Reproductive Toxicity
The majority of additional studies that should be included are listed in the next section. However in
terms of the reproductive effects where the report states there are no studies, the authors could report
on the study by Bai et al. (2010). They found that intravenous injection of functionalized MWCNT
resulted in accumulation in the testes, oxidative stress and decreased the thickness of the
seminiferous epithelium in the testis at day 15 without effecting the quality or quantity of sperm
production. By 60 days the damage was repaired and the treatment did not affect the pregnancy rate
and delivery success of female mice that mated with the treated male mice when compared with
controls.
Additional
considerations for
section
116
E16
118
E16
Genotoxicity/ Carcinogenicity
Within the sections on genotoxicity and carcinogenicity, the report also could consider the work of by
Sargent at al. (2011) where the authors looked at mitotic spindle aberrations in an epithelial cell line at
concentrations anticipated in exposed workers, specifically at doses equivalent to 20 weeks of
exposure at the Permissible Exposure Limit for particulates not otherwise regulated (Sargent et al.
2011). The study showed significant disruption of the mitotic spindle at occupationally relevant doses
and concluded that the increased proliferation in carbon nanotube-exposed cells indicates a greater
potential to pass the genetic damage to daughter cells. Based on these findings and the knowledge
that disruption of the centrosome is common in many solid tumors such as lung cancer and that
aneuploidy is an early event in the progression of many cancers (suggesting a role in both
tumorigenesis and tumor progression), the authors suggested that caution should be used in the
handling and processing of carbon nanotubes (Sargent et al. 2011).
Human carcinogenicity
mechanisms for
SWCNTs
117 E16 1 Oral Toxicity Additional
The report states there "no data was identified on the acute toxicity of MWCNT following oral or considerations for
dermal exposure" however within the derivation of an OEL for Baytubes, Pauluhn (2010) reported that section
based on the OECD TG 423 (Acute Oral Toxicity - Acute Toxic Class Method) that the MWCNTs'
tested (Baytubes) were not acutely toxic with an LD5o-oral of>5000 mg/kg bw.
In a later study, the authors looked at these effects in relation to other particles and found that at equal Additional
dose, diesel exhaust particles generated larger levels of 8-oxodG in rat liver than carbon black did and considerations for
exposure to fullerenes C60 and SWCNT were the least potent (M0ller et al. 2012). Based on these section
interesting findings, the authors noted that the extent of translocation from the gut is largely
unresolved but should be investigated further.
I-42
-------
Table 1-12 (Continued): Comments received in response to the expert charge questions.
Comment
ID
Author Charge
ID Question
Comment Text Excerpt
Themes
119
E16 1
As well as the Kolosnjaj-Tabi et al. 2010 study using SWCNT mentioned in the decaBDE MWCNT
report, Folkmann et al. (2009) investigated oxidative DMA damage in outlying tissues, specifically the
liver and the lung after intragastric administration of SWCNT at a low dose of 0.064 or 0.64 mg/kg
body weight. They noted that oral exposure to low doses of SWCNT was associated with elevated
levels of markers of oxidative DNA damage, specifically 8-oxo-7,8-dihydro-2'-deoxyguanosine (8-
oxodG) in the liver and lung.
Additional
considerations for
section
120
E16 2
Overall the science is portrayed comprehensively however there are a few incidences which could
have been improved.
Science accurate
121
E16 2
One area where the scientific interpretation could be improved is in relation to sensitization as there
are actually 2 reports in the literature examining dermal sensitization, both of which appear negative
(Pauluhn et al. 2010, Ema et al. 2011) and this should be reflected in Table 5.1.
122
E16 2
In terms of the mechanisms by which MWCNT may reach and persist in the pleura, this is dealt with
very sparsely. At very least the authors should have reported the findings of Ryman-Rasmussen et al.
2009 Mercer et al. 2010, showing sub-pleural deposition in the case of the former but most crucially
pleural translocation was demonstrated in the case of the latter after lung instillation. The argument
surrounding this has also been laid out by Donaldson et al. (2010) with evidence of the basis of
pleural retention from Murphy et al. (2011 and 2012) and most recently by Schinwald et al. (2012).
This whole section could do with more consideration.
123
E16 3
I have found the comparison with decaBDE very useful not least as it reminds us all that standard
chemicals also suffer from the same problem of insufficient information and that it is not just new
nano-materials. Therefore it is good as it helps us understand that it is not realistic to want 100%
clarity across every area of a risk assessment as almost no substance has this and instead we must
focus on what gaps must be filled and which would be nice to fill but are not a priority.
Comparison WAS
useful; highlighted data
gaps
124
E16 3
In addition, I found the information regarding the release of decaBDE from textiles and its
concentration in dusts found in office buildings particularly in drawing conclusions about CNT release.
Comparison WAS
useful
125
E16 4
An issue with documents as all-encompassing as this is that it is certainly impossible to include
everything so often the focus is on the most important studies and results. However I would like to
have seen more depth and discussion around the hazard data within the reports and summery
evaluations. For instance, there is always conflicting data within science but it is sometimes helpful to
summarize on balance what the majority of studies shows (specifically giving negative and positive
data equal footing). An example of this is in terms of respiratory sensitization where I would suggest
most studies point toward exacerbatory effects of MWCNT when given in conjunction with OVA rather
than direct sensitization - note this is relatively common amongst particles including diesel soot.
I-43
-------
Table 1-12 (Continued): Comments received in response to the expert charge questions.
Comment
ID
Author Charge
ID Question
Comment Text Excerpt
Themes
126
E17
Edgington AJ, Roberts AP, Taylor LM, Alloy MM, Reppert J, Rao, AM, Mao J, Klaine SJ. 2010.
The Influence of natural Organic Matter on the Toxicity of Multiwalled Carbon Nanotubes. Environ.
Toxicol. Chem. 29(11): 2511-2518.
Additional
considerations for
section
127
E17
Nowack B, Ranville JF, Diamond S, Gallego-Urrea J, Metcalfe C; Rose J, Home N, Koelmans AA,
Klaine SJ. 2012. Nanoparticle Release, Aging and Transformation in the Environment. Environmental
Toxicology and Chemistry 31(1): 50-59.
MWCNT physical/
chemical properties
128
E17
132
E17 4
von der Kammer F, Ferguson PL, Holden PA, Masion A, Rogers K, Klaine SJ, Koelmans AA, Home Detection/measuremen
N, Unrine JM. 2012. Analysis of Nanomaterials in Complex Matrices (Environment and Biota): General t analytics
Considerations and Conceptual Case Studies. Environmental Toxicology and Chemistry 31(1): 32-49.
129
130
131
E17
E17
E17
1
2
3
A quick Web of Science lit review for the past year reveals a few more.
Yes
Yes
Recent literature
published
Science accurate
Comparison WAS
useful
From an ecotox and environmental fate perspective this is a very good presentation. As this is my first
experience with this process, I found the integration of the lifecycle assessment information and
concepts very helpful in my comprehension of how this differs from a traditional risk assessment.
Praise for case study
133
134
135
E18
E18
E18
1
2
2
No.
Overall I believe the science is accurately conveyed throughout the document.
However, in my opinion some of the studies are underplayed in their demonstration of toxicolc
No additional literature
Science accurate
>gical
concern. I would suggest a greater emphasis on toxicological concern for studies showing MWCNT
could induce mesothelioma and behave in a similar manner to asbestos. These inhalation studies are
strong indicators of potential toxicological concern. The document merely states that the inhalation
route of exposure for MWCNT "might be of toxicological concern." (see Section 5-7, line 8 and 9).
136
E18
Yes.
Comparison WAS
useful
137
E18 4
No.
I-44
-------
Table 1-12 (Continued): Comments received in response to the expert charge questions.
Comment
ID
Author Charge
ID Question
Comment Text Excerpt
Themes
138 E19 1 Below are two recent studies investigating the impact of MWCNT releases within the environment that Recent literature
may be of interest for determining effects and proper characterization techniques for MWCNT then published
they enter complex media.
1- Mwangi, J. N., Wang, N., Ingersoll, C. G., Hardesty, D. K., Brunson, E. L, Li, H. and Deng, B.
(2012), Toxicity of carbon nanotubes to freshwater aquatic invertebrates. Environmental Toxicology
and Chemistry, 31: 1823-1830.
139
E19 1
141
E19
Below are two recent studies investigating the impact of MWCNT releases within the environment that
may be of interest for determining effects and proper characterization techniques for MWCNT then
they enter complex media.
2-Nowack, B., Ranville, J. F., Diamond, S., Gallego-Urrea, J. A., Metcalfe, C., Rose, J., Home, N.,
Koelmans, A. A. and Klaine, S. J. (2012), Potential scenarios for nanomaterial release and
subsequent alteration in the environment. Environmental Toxicology and Chemistry, 31: 50-59.
Recent literature
published
140 E19 2 The science and information is accurately communicated in the document. The use of tables and flow Science accurate;
charts is helpful and illustrative. In future documents; it may also be beneficial to include more suggestion for
representative graphs and photographs from the literature to effectively portray the results generated, improvement
The comparison of decaBDE and MWCNT is helpful in assisting reviewers with understanding the
process typically utilized for regulating a more traditional chemical by helping to identify knowledge
gaps in the process. Also, the ability to compare some of the pitfalls and phasing out of decaBDE
allows researchers to make more informed decisions about how MWCNT can be objectively
investigated and incorporated into the flame retardants.
Comparison WAS
useful; highlighted data
gaps
142
E19 4
The document is well organized.
143
E20 1
Natural Organic Matter Stabilizes Carbon Nanotubes in the Aqueous Phase
Hoon Hyung, John D. Fortner, Joseph B. Hughes, and Jae-Hong Kim
144
E20 1
A Review of Carbon Nanotube Toxicity and Assessment of Potential Occupational and Environmental -
Health Risks
Chiu-wing Lam, John T. James, Richard McCluskey, Sivaram Arepalli, Robert L. Hunter
145
E20 1
Utilization of selected area electron diffraction patterns for characterization of air submicron particulate —
matter collected by a thermophoretic precipitator. John J Bang, Elizabeth A Trillo, Lawrence E Murr
146
E20 1
A Predictive Toxicological Paradigm for the Safety Assessment of Nanomaterials. Huan Meng, Tian
Xia, Saji George, and Andre E. Nel
I-45
-------
Table 1-12 (Continued): Comments received in response to the expert charge questions.
Comment Author Charge
ID ID Question Comment Text Excerpt Themes
147 E20 2 To my knowledge, the document conveys scientific aspect of our limited knowledge and related issues Science accurate
about MWNT in fairly reasonable way.
148 E20 3 It is believed that the comparison helps reviewers understand the underlying issues in MWCNT Comparison WAS
applications and identify research gaps for exposure assessment and risk management decisions useful; highlighted data
gaps
149 E20 4 In some tables and figures, the authors seemed to include too much information that can make
reviewers miss the main point for each table/figure. This could be a subjective opinion.
150 E21 1 1) Larue, C.A, Pinault, M.B, Czarny, B. C, Georgin, D.D, Jaillard, D.E, Bendiab, N.F, Mayne-
L'Hermite, M.B, Taran, F.D, Dive, V.C, Carriere, M, Quantitative evaluation of multi-walled carbon
nanotube uptake in wheat and rapeseed, (2012) Journal of Hazardous Materials, 227-228, pp. 155-
163.
151 E21 1 2) Wang, X., Han, H., Liu, X., Gu, X., Chen, K., Lu, D. Multi-walled carbon nanotubes can enhance
root elongation of wheat (Triticum aestivum) plants, (2012) Journal of Nanoparticle Research, 14 (6),
art. no. 841,
152 E21 1 3) Evaluation of exposure risk in the weaving process of MWCNT-coated yarn with real-timeparticle
concentration measurements and characterization of dust particles (2012) Industrial Health, 50 (2), pp.
147-155
153 E21 1 4) Clark, K.A., O'Driscoll, C., Cooke, C.A., Smith, B.A., Wepasnick, K., Fairbrother, D.H., Lees, P.S.J., -
Bressler, J.P., Evaluation of the interactions between multiwalled carbon nanotubes and caco-2 cells
(2012) Journal of Toxicology and Environmental Health - Part A: Current Issues, 75 (1), pp. 25-35.
154 E21 1 5) I found the OECD recent document (March 2012 on Nano Risk Assessment useful: Important
Issues on Risk Assessment of Manufactured Nanomaterials Series on the Safety of Manufactured
Nanomaterials, No. 33, at
http://search.oecd.org/officialdocuments/displaydocumentpdf/?cote=env/jm/mono(2012)8&doclanguag
e=en
155 E21 2 Yes it is well done. Science accurate
156 E21 3 Yes. Comparison WAS
useful
I-46
-------
Table 1-12 (Continued): Comments received in response to the expert charge questions.
Comment
ID
Author Charge
ID Question
Comment Text Excerpt
Themes
157
E21
Another good reference for decaBDE is the ASTDR Toxicological Profile for Polybrominated Biphenyls
and Polybrominated Diphenyl Ethers at: http://vwvw.atsdr.cdc.qov/toxprofiles/tp68.pdf
It is from 2004 but it shows that there is substantial exposure data in addition to what is mentioned in
the case study document.
DecaBDE data needs
improvement
158
E21
No
159
E22 1
I answer this question more fully in combination with number 2 where I identify places where I don't
think the science is accurately conveyed in this case through the omission of relevant references.
Specific references that I describe in response to question 2 are the following
160
One overall comment is for the authors of this report to focus more on data that has been published
rather than speculations from review papers published prior to much reason being conducted (i.e.,
before 2008). The science is at a stage when prior expectations need to be evaluated based on
currently available evidence rather than rely upon early speculations. Many of the early speculations
have been proven to be wrong and are now outdated.
Recent literature
published; suggestion
for improvement
161
E22 2
Page xxi, lines 1-2 - Some published papers indicate that MWCNTS function similarly to hard carbons Highlight data gap;
decreasing MWCNT availability to organisms in soils and sediments (i.e., Petersen, E. J. et al. additional
Environ. Sci. Technol. 2009, 43(11), 4181-4187.; Shen, et al. Environ. Toxicol. Chem. 2012, 31 (1), considerations for
202-209.). section
162
E22
Pagexxii, lines 19-20-Actually, numerous studies, probably >20, have been conducted on the
effects and uptake of MWCNTs to aquatic organisms.
Highlight data gap;
suggestion for
improvement
163
E22
Table 1-9 - Chirality only is relevant for SWCNTs and the inner tube of DWCNTs (Yang, S. W.; Parks, Recent literature
A. N.; Saba, S. A.; Ferguson, P. L; Liu, J., Photoluminescence from Inner Walls in Double-Walled published; additional
Carbon Nanotubes: Some Do, Some Do Not. Nano Letters 2011, 11, (10), 4405-4410.). considerations for
section
164 E22 2
165 E22 2
Table 3-2 - It is unclear why the authors mean by "potential lipophilicity" and how this would differ
from "hydrophobic"; hydrophobic sorbents would typically adsorb lipids in addition to organic
contaminants. Those sources are all extremely old, so this is probably just speculation from before
many experimental measurements were taken. I suggest deleting this section of the table.
Page 3-8, lines 31-33 - This reference is highly questionable since it only refers to an abstract.
Suggestion for
improvement
Suggestion for
improvement
I-47
-------
Table 1-12 (Continued): Comments received in response to the expert charge questions.
Comment Author Charge
ID ID Question Comment Text Excerpt
Themes
166
E22
Page 3-12, line 29-The Holbrook et al. 2010 reference is also highly relevant here.
Additional
considerations for
section; suggestion for
improvement
167
Page 3-14 - The Zhang et al. 2011 reference is also relevant in this section since it discusses sorption
of MWCNTs onto peat, a soil component.
Additional
considerations for
section; suggestion for
improvement
168 E22 2 Page 3-14, line 17-The review papers cited here about MWCNT adsorption onto soil surfaces were Suggestion for
published before sorption studies had been conducted. These articles are reviews that probably just improvement
contained speculation along these lines, so this sentence could be deleted.
169
E22
Page 3-14, line 28-29-Again, those authors did not know how those properties would affect
nanoparticle sorption, but they probably speculated. The wording should be changed to reflect this or
else this sentence just deleted.
Additional
considerations for
section; suggestion for
improvement
170
E22
Page 3-15, lines 1-2-This sentence is just speculation and could be deleted.
Suggestion for
improvement
171
Text box 4-1 - There are not mass spectrometry techniques for characterizing MWCNTs, nor is that
provided in Appendix B other than ICP-MS which only measures metal concentrations still associated
with the MWCNTs.
172
E22
174
E22
Page 4-48, line 5 - It is unclear what the authors mean to indicate by stating that absorption of carbon
materials can occur in the gut. In the study cited, the authors showed light microscope pictures of
carbon nanotubes in the gut tract, not absorption across the gut tract. Another study by Edgington et
al. (Edgington, A. J. et al. Environ. Toxicol. Chem. 2010, 29(11), 2511-2518.) did not find MWCNT
absorption across the gut tract.
Recent literature
published; additional
considerations for
section
173 E22 2 Page 4-49, line 6 - These values were BSAF values, not BAF values. This sentence should be edited Suggestion for
accordingly. improvement
Page 49, lines 18-20 - This sentence is entirely speculation. In studies of CNT uptake by a wide range Recent literature
of organisms (excluding plants), absorption into tissues is consistently minimal (see Table 2 of published; highlighted
Petersen et al. 2011, ES&T pages 9837-9856). data gaps
I-48
-------
Table 1-12 (Continued): Comments received in response to the expert charge questions.
Comment
ID
Author Charge
ID Question
Comment Text Excerpt
Themes
175
E22 2
177
180
E22 2
Page 4-49, line 21-27 - What these statements do not include is the lack of CNT absorption across
gut linings in a wide variety of organisms. Also, changing the octanol-water partitioning behaviors of
MWCNTs did not change their uptake, so MWCNTs' apparent lipophilicity and hydrophobicity do not
appear to impact bioaccumulation (Petersen et al., 2010, Environ. Toxicol. Chem., pages 1106-1112).
Thus, the expectation in lines 24-25 goes against the data trend for MWCNT absorption in a range of
studies (see previous comment for the relevant citation). A critical step not mentioned is that the
compound has to be absorbed across the microvilli which did not appear to occur for MWCNTs in any
study.
MWCNT physical/
chemical properties;
suggestion for
improvement
176 E22 2 Page 4-50, lines 9-10-This paper was published before data was collected on this topic and thus this Suggestion for
speculation can be deleted. Moreover, being lipophilic is insufficient for expectations of accumulation improvement
in root lipids because transport into the cells and roots would need to occur first.
Page 4-51, lines 10-13-The study cited (Petersen 2010) actually shows that MWCNTs probably do
not behave similar to other bioaccumulative substances because changing the octanol-water
distribution behavior did not change BAF values. The purpose of that study was to investigate whether
MWCNTs behaved similar to bioaccumulative substances such as hydrophobic organic chemicals and
the findings suggested that they did not apparently as a result of a lack of absorption across the gut
tract and into the organism. The Helland 2007 review is an outdated paper from 2007 that just
speculates about this topic.
MWCNT physical/
chemical properties
178 E22 2 Page 5-22 lines 22-23-Actually, there are numerous studies on benthic invertebrates (i.e., Kennedy Recent literature
2008 cited on page R-12 and Shen et al. 2012 and Kennedy et al. 2009 cited in response to question published
1). A list of studies on CNT toxicity to soil and sediment organisms is available as Table 3 in Petersen
et al. 2011 ES&T pages 9837-9856.
179 E22 2 Page 5-22, lines 27-28-The study cited has nothing to do with bioaccumulation in aquatic systems. Suggestion for
Numerous papers have been conducted on exactly this topic and some of these studies were improvement
described in the previous Chapter.
Table 5 -2 - This table missing numerous important citations for MWCNTs as described in previous
comments.
181
Page 5-23, lines 18-20 - The authors are missing a ton of studies on the effects of MWCNTs to
aquatic organisms. Some have been described above and many others are cited in Chapter 4 of this
document and in the review paper by Petersen et al. 2011 ES&T pages 9837-9856. There are
probably >20 references overall on this topic.
Highlight data gap;
suggestion for
improvement
I-49
-------
Table 1-12 (Continued): Comments received in response to the expert charge questions.
Comment
ID
Author Charge
ID Question
Comment Text Excerpt
Themes
182
E22 2
Table 5-4 and discussion on page 5-29 - This table is missing the important reference for Scott-
Fordsmand et al. (2008 Ecotox. Environ. Saf. 616-619.). Many papers testing accumulation of
MWCNTs also have information about weight change, lipid content change and acute toxicity even
though those endpoints were not the purpose of the study (Petersen et al. Environ. Sci. Technol.
2011, 45 (8), 3718-3724.; Petersen, E. J. etal. Environ. Sci. Technol. 2008, 42 (8), 3090-3095.; and
Petersen etal. Environ. Sci. Technol. 2009, 43 (11), 4181-4187).
Recent literature
published
183
E22 2
Page 5-31, lines 3-5 - There have actually been papers showing positive effects from MWCNT
exposure such as Khodavorosky et al. 2009, 2011, so it is unnecessary to start the discussion about
nanoparticles in general or describe nano-AI2O3 plant findings.
Recent literature
published; suggestion
for improvement
184
E22 3
Having the decaBDE as a comparison was helpful in some ways and a hindrance in others. The ways
that this comparison were helpful is that it gave me an idea of the amount of information available for a
compound that has been studied for a much longer time period than MWCNTs. Otherwise, it would
have just been comparing the research available for MWCNTs against our perception of what a good
amount of data would have been for the risk assessment.
Comparison WAS
useful
185
However, this comparison was limiting in that it was unclear how big of a deal the data gaps that still
do exist for decaBDE actually were for risk assessors. I would have been curious, and perhaps this is
not information that could have been shared, about how limited risk assessors felt about the current
state of information and their ability to make a judgment about decaBDE.
Comparison NOT
useful; suggestion for
improvement
186
Another limitation of including this comparison is that I had to more carefully skim through the
document when I was doing the hazard ranking to focus solely on the MWCNT data. I read through
the decaBDE information the first time I read the document but mostly skipped it subsequent times
when doing the prioritization exercise.
Comparison NOT
useful; added length
187
Most of my comments were addressed in response to the other questions. I think it is a good
document overall, but there are definitely places where it could be improved. The huge controversy in
the news about whether BDE actually effectively worked in the ways it was described to in this
document as a result of the Chicago Tribune series on flame retardants
(http://media.apps.chicagotribune.com/flames/index.html) made me wonder the fire research on
BDEs and the ways they supposedly worked was accurate.
188
E23 1
No
No additional literature
189
E23 2
Yes
Science accurate
I-50
-------
Table 1-12 (Continued): Comments received in response to the expert charge questions.
Comment Author Charge
ID ID Question Comment Text Excerpt Themes
190 E23 3 Yes, the comparison of decaBDE and MWCNT in the case study document is useful in identifying Comparison WAS
research gaps. useful
191 E23 4 No
1-51
-------
Table 1-13. Free-form comments received from experts.
Author
Comment ID ID
Context (of External
Review Draft)
Comment Text Excerpt
Themes
192
E4-F Preface, pg xiii lines
12-13
But is this application really relevant?? MWCNT will not be used in textile coatings Choice of nanomaterial and
alone since they are unable to pass the tests by themselves. If they cannot pass the application; choice
regulatory test, they will never be used in the application. I do not think this is a
good application to study MWCNT release into the environment as MWCNT are
more likely to be used in electronics and aerospace rather than commodity
consumer goods.
constricted science of
MWCNTs
193
E4-F Preface, pg xvi lines
21-22
One unknown - can MWCNT really be mixed with binding agents, applied to
textiles, AND pass the necessary regulatory tests? If they cannot, then they won't
be used and this study may be misled by focusing on textiles when potential
exposure could be in other applications not looked at by this CEA.
Choice of nanomaterial and
application; choice
constricted science of
MWCNTs
194
E4-F Preface, pg xvi lines
31-33
Agreed - composition of MWCNT mixture can be a major factor to consider when
studying these materials. Purity of the MWCNT particle, the other components of
the raw material (amorphous carbon, catalysts, etc.) will likely all have an effect on
toxicity and human exposure issues.
195
196
197
198
199
E4-F
E4-F
E4-F
E4-F
E4-F
Preface, pg wii
lines 3-4
Preface, pg wii
lines 7-8
Preface, pg xviii
lines 8-9
Preface, pg xix
lines 5-8
Preface, pg xix
lines 25-26
Do any commercial products exist? I think the answer is no, and possibly for good
reason (MWCNT alone does not work)
Agree - manufacturing equipment won't be different for treating fabrics, and so any
way of applying decaBDE will be used to apply MWCNT.
Need to look at Nyden paper from Interflam 2010 - MWCNT in ash from cone
calorimeter fires can be shaken lose and made airborne.
Seems that environmental chemistry of MWCNT is an unknown that needs to be
researched, regardless of its use in textile back-coatings or not
This is because MWCNT use is way too new in most cases - its use in consumer
goods is limited. Need to gather information on what MWCNT is actually used in
today to determine realistic environmental exposure routes.
Choice of nanomaterial and
application
Comparison WAS useful
Additional references
Highlight data gap
Choice of nanomaterial and
application; choice
constricted science of
MWCNTs
I-52
-------
Table 1-13 (Continued): Free-form comments received from experts.
Author
Comment ID ID
Context (of External
Review Draft)
Comment Text Excerpt
Themes
200
201
E4-F
E4-F
preface, pg xx
lines 12-13
preface, pg xxi
lines 12-13
But with which type of MWCNT? What surface chemistry for these materials?
Seems that environmental decomposition/transformation of MWCNT is unknown
and should be studied. Do MWCNT degrade into something else that is bioactive?
Or are they environmentally stable (persistent)?
MWCNT physical/ chemical
properties
Highlight data gap
202
E4-F Chapter 1, pg 1-2
lines 3-7
Are any of these really commercially viable? Perhaps at the time of the start of this Choice of nanomaterial and
study to proactively look for systems of concern this made sense, but I think this application; choice
original premise needs to be revisited. Are these really the PRIME applications of constricted science of
MWCNT in the future where emissions could are from, or are there applications MWCNTs
missed in this original assessment more likely to be the ones that should be looked
at?
203
E4-F Chapter 1, pg 1-2
lines 23-26
Yes and no. Yes for other MWCNT coatings on other objects (paints for example)
but no for polymer applications where the MWCNT is well embedded in the final
part (polymer composites, injection molded plastic parts) - unless said parts are
deliberately abraded/reground, in which case textile abrasion data may be relevant
to MWCNT in composite and injection molded applications.
Choice of nanomaterial and
application; alternative
products/ formulations
204
E4-F Chapter 1, pg 1-6
lines 21-22
Given this point, really important then that some application survey research be
done to make sure that the key uses of MWCNT are truly captured rather than
assuming the textile application will be the key one.
Choice of nanomaterial and
application; choice
constricted science of
MWCNTs
205
E4-F Chapter 1, pg 1-7
line 1
Agree - you can't look at everything, but if you end up looking at an application that
can NEVER occur, are you really spending your time wisely? I think looking at
MWCNT coming off of aerospace composites that are ground/sanded/drilled is a far
better use of time and volumes there will likely be higher than in the textile
application, where it is not yet even proven that MWCNT will even work.
Choice of nanomaterial and
application; choice
constricted science of
MWCNTs
206
E4-F Chapter 1, pg 1-8
lines 11-13
Not necessarily. You're assuming that they will work as flame retardants in all
applications, something that has yet to be proven. Yes, they do show flame-
retardant effects, but not enough to pass required regulatory tests. If they don't
pass the regulatory fire tests, they will never be used - ever.
Choice of nanomaterial and
application
207
E4-F Chapter 1, pg 1-11
lines 6-17
Very true, but if something doesn't work, then the performance criteria ensures that
this particular flame retardant is NEVER used in this specific application.
Choice of nanomaterial and
application
I-53
-------
Table 1-13 (Continued): Free-form comments received from experts.
Author
Comment ID ID
Context (of External
Review Draft)
Comment Text Excerpt
Themes
208
E4-F Chapter 1, pg 1-15
Table 1-5
MWCNT can also be melt compounded into plastic, but not a lot of data out there
saying that this nanocomposite can actually be spun into fiber for lowered
flammability textile strands/fibers.
Choice of nanomaterial and
application
209
E4-F Chapter 1, pg 1-21
lines 3-4
This is the key issue. It is unknown if MWCNT will EVER be used in this application
to provide fire safety to textiles or furniture. Therefore - it this really a proper focus
of study? Sure - look at MWCNT emissions for use in products, but I am not
convinced that MWCNT will ever be used to provide fire protection of fabrics.
Choice of nanomaterial and
application; choice
constricted science of
MWCNTs
210
E4-F Chapter 1, pg 1-23
lines 1-2
Flawed premise here. Just because it shows these flammability effects does not
mean it can be used in future textiles. Nanocomposites alone rarely pass regulatory
tests by themselves, and if MWCNT does not pass the regulatory tests, then it will
never be used. Therefore need to conduct research on IF MWCNT can actually be
used in this application or not.
Choice of nanomaterial and
application; alternative
products/ formulations
211
E4-F Chapter 1, pg 1-24
Table 1-12
Note that the data does not say if it passed a particular test or not. It needs to be
validated that MWCNT CAN actually bring FR performance in a PU foam
application. If it cannot, it will never be used and its potential environmental
exposure routes change completely.
Choice of nanomaterial and
application; alternative
products/ formulations
212
E4-F Chapter 2, pg 2-1
Figure 2
This is incorrect for decaBDE in regards to handling/packaging, equipment
cleaning, and accidents - where there are known releases of decaBDE. See
Albemarle for details - they have known data showing releases will occur from this
part of the process.
DecaBDE data needs
improvement; Additional
references
213
E4-F Chapter 2, pg 2-4
lines 11-12
You probably won't easily get this information from the MWCNT manufacturers, but,
it's clear you need this data. How MWCNT and SWCNT are made can be very
different - different metal catalysts, different impurities, and different CNT length, all
of which are important to your study.
Highlights research need
214
E4-F Chapter 2, pg 2-4
lines 22-23
There are much better references out there. Please talk with some US based CNT
synthesis researchers for details. JM Tour - Rice University, K Lafdi - University of
Dayton - either one of these two professors can quickly bring you up to speed with
more accurate information.
Additional references
215
E4-F Chapter 2, pg 2-5
lines 2-3
Well ... maybe. It depends on if the reactors for making the CNTs need to be
cleaned out between runs. If they do need to be cleaned out, there could be CNT or
decaBDE release at this point. More likely that CNT release would occur if the
reactor is not fully cleaned out before synthesis.
Product life cycle
I-54
-------
Table 1-13 (Continued): Free-form comments received from experts.
Author
Comment ID ID
Context (of External
Review Draft)
Comment Text Excerpt
Themes
216
E4-F Chapter 2, pg 2-6
lines 22-23
Don't make this assumption. Some R&D labs have no clue how to safely handle
nanoparticles, and releases are quite common if hoods and engineering controls
are limited, as they often are in older research buildings at established universities.
Product life cycle
217
E4-F Chapter 2, pg 2-7
lines 6-7
There is a report from work done at the University of Dayton Research Institute in
cooperation with NIOSH. Let me see about getting you the data - or connecting you
to the NIOSH researchers who generated the data back in 2008.
Additional references
218
E4-F Chapter 2, pg 2-14
Table 2-5
Again, information incorrect here from Handling/packaging, equipment cleaning,
accidental releases. See Albemarle for details.
Additional references
219
E4-F Chapter 3, pg 3-1
lines 16-18
For MWCNT-polymer complex, there are two types to consider - those bound
through intermolecular entanglement, and those where the MWCNT is covalently
bound to the polymer. They will behave quite differently in the final environment and
will likely also have different emission schemes into the environment. DecaBDE can
only have intermolecular entanglement (van der Waals forces) and therefore can
escape/migrate from polymers easier than MWCNT.
MWCNT chemistry and
nature of bond with
polymer/product
220
E4-F Chapter 3, pg 3-6
lines 1-2
All of this will depend upon how the MWCNT interacts with the polymer. If it is
covalently bound to the polymer, it may not ever be in single or bundle form. If just
mixed into the polymer matrix, then yes, it will likely migrate into the environment
via the same pathways as decaBDE.
MWCNT bond with polymer
not adequately described for
implications in
release/exposure; MWCNT
chemistry and nature of
bond with polymer/product
221
E4-F Chapter 3, pg 3-8
lines 11-12
Assuming MWCNT actually end up in this application (which I think I've made clear
I severely doubt), then the length of the tubes and how they interact with the
polymer is a whole area of study that is very situation and MWCNT specific.
Something that is chemically bound to the polymer may not come out at all as the
textile back-coating degrades unless the polymer is biodegraded away upon
ingestion/environmental exposure. And that final form of the MWCNT when the
polymer is finally worn/decomposed/digested away may be very different than the
starting form of the MWCNT. So if you're really going to look at this, I think you
need to actually do a couch mockup and put it through the paces and properly
measure what happens as a function of MWCNT purity, length, and surface
chemistry as I am quite confident the results will be very different for each different
type of MWCNT.
MWCNT chemistry and
nature of bond with
polymer/product
I-55
-------
Table 1-13 (Continued): Free-form comments received from experts.
Author
Comment ID ID
Context (of External
Review Draft)
Comment Text Excerpt
Themes
222
E4-F Chapters, pg 3-10
lines 24-25
Don't make this assumption - see my comments above about the surface chemistry
of the MWCNT. If they are designed to chemically react with the textile
back-coating, their final solubility when they escape into the environment may be
very different than what is shown here.
223
E4-F Chapter 4, pg 4-10
lines 17-18
Very interesting ... I did not know this. Have these results been verified?
224
E4-F Chapter 4, pg 4-13
lines 1-2
May still be able to use the NIOSH report here (Methner 2010). Not relevant for
textile back-coating, but still useful to see where nanotube release and exposure
could realistically occur in a manufacturing environment.
Additional references
225
E4-F Chapter 4, pg 4-25
lines 22-23
A lot here will depend upon what chemistry and process is used to get the MWCNT
into the final product. If the CNT are chemically bound to polymer while being
applied, primary exposure may be to CNT + Polymer particulates, not CNT directly.
If CNT is not chemically bound to polymer, then using the decaBDE models is
appropriate.
MWCNT physical/ chemical
properties; Nature of bond
with polymer/product
226
E4-F Chapter 4, pg 4-26
lines 1-2
Again, see comments on the chemistry of the MWCNT and its interaction with the
polymer, as this will dictate how they may be emitted during post-product use
handling (grinding, incineration, conversion to scrap, etc.) Very likely the textiles
would be treated the same way as textiles are handled today at the end of their
lifetime, but how the MWCNT interacts with the polymer (covalent vs. van der
Waals) will dictate potential exposure as a function of recycle/end-of-use operation.
MWCNT physical/ chemical
properties; Nature of bond
with polymer/product
227
E4-F Chapter 4, pg 4-26
lines 14-15
This indicates that validation of MWCNT exposure route via particular application
needs to be studied as a high priority item. Going back to specific MWCNT
chemistry, how that MWCNT will affect potential release just as much as end-use
application, and so I think first some time is needed to ensure what applications are
really likely to use MWCNTs so that one finds the right applications to be concerned
about, and doesn't miss any that may not be obvious at first glance.
Choice of nanomaterial and
application; Highlighted data
gaps; research applications
of particular concern and
widespread use/ choice
constricted science of
MWCNTs
I-56
-------
Table 1-13 (Continued): Free-form comments received from experts.
Author
Comment ID ID
Context (of External
Review Draft)
Comment Text Excerpt
Themes
228
E4-F Chapter 4, pg 4-26
lines 20-21
I'm not sure about this assumption. If the MWCNT is covalently bound to the
polymer, then it cannot be released in particle form. Rather, it would be released in
a "polymer + MWCNT" agglomerate which likely has its own unique exposure
issues. MWCNT that is not covalently bound to the polymer may indeed come out
of the polymer and then agglomerate, but there are not studies out that that show
what MWCNTs will do if they have no adjacent MWCNTs to agglomerate with. Will
they prefer to agglomerate with household dust, or will they agglomerate with
soils/minerals preferentially? Completely unknown and it should be looked at.
MWCNT physical/ chemical
properties; Highlighted data
gaps; Nature of bond with
polymer/product
229
E4-F Chapter 4, pg 4-30
lines 10-11
Deca is not used in automotive seat cushions, or really in any part of the car. So I'm
not sure the data in the two reports here is accurate. Deca is used in aircraft
though, but not in the seat cushions or fabrics. Rather it is used in other parts of the
aircraft - mostly non-structural fascia, and other plastic parts. So exposure from
airplanes and cars will be different - as it will be with MWCNTs. Just because Deca
is used in an application does not mean that MWCNT will be used as a replacement
technology. MWCNT will only be used IF it yields a pass of the specific material
flammability regulation.
DecaBDE data needs
improvement
230
E4-F Chapter 4, pg 4-30
lines 28-29
Again, I don't think this is correct since Deca is not used in cars today. Given the
very easy flammability test for automobiles, MWCNT will not be used to provide fire
protection because there are far cheaper solutions out there.
DecaBDE data needs
improvement
231
E4-F Chapter 4, pg 4-30
lines 31-32
This however, I do agree with. I think studying what exactly has built up in aircraft
over the years would be a study worth looking into.
232
E4-F Chapter 4, pg 4-32
lines 12-13
So this would be worthwhile to determine.
Highlight data gap
233
E4-F Chapter 4, pg 4-32
lines 22-23
MWCNT stability to environmental conditions would also be a worthwhile study to
investigate - not just for environmental resins, but for product durability as well. This
would make a great "dual use" study.
Highlight data gap
234
E4-F Chapter 4, pg 4-34
line 1
I think once you figure out what MWCNT will actually be used in, then this question Choice of nanomaterial and
starts to get answered and you'll quickly figure out through demographic studies application; choice
who is most likely to be a part of a high-exposure population. If MWCNTs get constricted science of
mostly used in higher-end performance applications (example - used a lot today in MWCNTs
high end golf clubs) then you may find that the highly exposed population is actually
the affluent, and not lower income or children.
I-57
-------
Table 1-13 (Continued): Free-form comments received from experts.
Author
Comment ID ID
Context (of External
Review Draft)
Comment Text Excerpt
Themes
235
236
E4-F Chapter 4, pg 4-43
(section 4.3 heading)
E4-F Chapter 5, pg 5-1
lines 12-13
Regardless of application - once MWCNT gets into the environment then all of this Comparison WAS useful
is relevant and I think is well established science. Perhaps there are some
unknowns specific to MWCNTs, but in general, I would say that using decaBDE as
the model, or any pollutant for that matter, is a very reasonable and well thought out
place to start. My comment here applies to section 4.4 and 4.5 as well.
Assuming of course that MWCNT gets used in upholstery textiles. Again, I doubt
they will, but if they actually do, then yes, this is correct.
Choice of nanomaterial and
application; choice
constricted science of
MWCNTs
237
E4-F Chapter 5, pg 5-2
(section 5.1 heading)
I somehow doubt you'll ever be able to definitively get data on this for MWCNT or
any other material since human testing is banned. You may be able to gather this
information from post-mortem studies, but otherwise I don't think this area should
be studied due to the practical issues of being unable to exposure humans to these
chemicals or MWCNT deliberately. However, funded research into mimics of
human health effects with artificially grown tissue or perhaps really strong modeling
software would be worthwhile to pursue. This would be good dual-use research in
that it would benefit both environmental and pharmaceutical/medicine fields.
Highlight data gap
238
E4-F Chapter 5, pg 5-20
section heading
5.1.12
See my comments above regarding application driving who will likely be a high
exposure population. Until you know the actual applications that MWCNT will likely
be used in, you really cannot extrapolate Deca data to MWCNT population
exposure. If indeed MWCNT gets out of multiple future household projects and
ends up in household dust then maybe you can make the correlation. Otherwise I
think you have to wait until you get the application identified, which again supports
that research into the most likely applications of MWCNT technology is a high
priority research item that needs to be answered.
Choice of nanomaterial and
application; choice
constricted science of
MWCNTs
239
E4-F Chapter 5, pg 5-21
section heading 5.2
Agree that MWCNT reaction with environment is unknown and should be studied.
cannot comment on the priority though - that would have to be determined based
upon the known applications and likely release of MWCNT to the environment.
While there are lots of unknowns here, I think they are low priority to assess until
you address what MWCNT will be used in, and in what form, so you can see the
vectors of release to the environment/population.
I Choice of nanomaterial and
application; choice
constricted science of
MWCNTs
I-58
-------
Table 1-13 (Continued): Free-form comments received from experts.
Author
Comment ID ID
Context (of External
Review Draft)
Comment Text Excerpt
Themes
240
E4-F Chapter 5, pg 5-34
lines 3-4
See my comment about the affluent likely to be the group most likely to be exposed
to MWCNT first. I don't know if that is really an issue of environmental justice or not,
but I suspect that the rich will be initially exposed to MWCNT at higher levels than
anyone else initially.
Choice of nanomaterial and
application; choice
constricted science of
MWCNTs
241
E4-F Chapter 5, pg 5-34
lines 2-25
Very interesting observation. A good life cycle analysis of MWCNT synthesis as a
function of the different manufacturing routes would be useful to research and
develop. Likely a good dual-use research project that could be supported by
industry so that they reduce their production costs.
Highlight data gap
242
E4-F Chapter 5, pg 5-38
line 21 (MWCTN
section)
These current costs are why MWCNT will not be used in textiles anytime soon.
Prices have greatly dropped in recent years (please contact NanoCyl in Belgium to
get an idea of how much - we're talking hundreds of dollars per kilogram these
days) and NanoCyl calculates prices dropping even further, but not likely to be at
the levels to be a cost effective flame retardant for consumer goods, especially
since MWCNT alone has been unable, to date, to pass any regulatory fire tests. If it
cannot pass the test, it will never be used. More likely, MWCNT will be used in
other goods requiring improved electrical, thermal, and mechanical properties
(structural composites, electronics) and so economic impacts could be quite
different in these areas - with positive effects in generating light weight composites
that yield fuel savings over metal on aircraft today.
All of this is why I think a really good LCA for MWCNT use, manufacture, and total
impacts is needed.
Choice of nanomaterial and
application; choice
constricted science of
MWCNTs; other applications
should have been
considered
243
E4-F Chapter 6, pg 6-2
lines 3-4
If in the end, effort is spent really focusing on MWCNT as if it was going to be used
in flame-retardant textiles, then I think this entire program will be wasted effort with
maybe a few gems of good information that come out of it. I strongly recommend
spending some time up front making sure the applications for MWCNT are correct
(what will they REALLY be used in with a high probability of occurring) and then
other aspects of this study come into play. I do feel MWCNT release into the
environment should be looked at in a proactive manner, but it needs to be done
with an application that makes sense so that the research effort is focused and
targeted and yields information that practically addresses potential MWCNT release
and exposure.
Choice of nanomaterial and
application; choice
constricted science of
MWCNTs
244
E4-F Chapter 6, pg 6-2
lines 23-28
These papers DO NOT suggest this. The results of the papers have unfortunately
been misinterpreted. MWCNT does have flame-retardant effects, but not enough of
a flame-retardant effect to be used in this application.
Choice of nanomaterial and
application
I-59
-------
Table 1-13 (Continued): Free-form comments received from experts.
Author
Comment ID ID
Context (of External
Review Draft)
Comment Text Excerpt
Themes
245
E4-F Chapter 6, pg 6-5
lines 11-12
Assuming of course enough members of the group know how to accurately assess
the niche information about material flammability and what drives the selection of a
particular chemical or nanoparticle into an application. I would not assume that this
information is adequately covered by the community in the CEA since flame
retardant research is a niche field of applied engineering. Hopefully there are other
material fire scientists in this process. If I'm the only one, then I am concerned that
my one vote/voice in the CEA process will be missed and again, this effort will be
spent on something which may yield some good information, but fails to address in
a proactive manner potential MWCNT releases and how they will affect our society
and environment.
CEA framework/
methodology; Choice of
nanomaterial and application
246
E4-F Appendix A, pg A-1
lines 13-14
Why not electronics? Sporting goods?
Choice of nanomaterial and
application; alternative
products/ formulations
247
E4-F Appendix A, pg A-1
line 17
Aerospace composites - very likely
Choice of nanomaterial and
application; alternative
products/ formulations
248
249
E4-F Appendix A, pg A-2
line 15
E4-F Appendix A, pg A-2
lines 17-18
There is more than you think - again, talk with NanoCyl in Belgium. They are mostly Choice of nanomaterial and
used in polymers requiring enhanced electrical, thermal, and mechanical properties application; alternative
that can justify the significant increase in cost - so higher end applications, not products/ formulations;
general consumer goods. Additional references
Agree on this - and this is DEFINITELY worth looking into.
Highlight data gap
250
E4-F Appendix A, pg A-5
Figure A-1
"medium list"
I would argue is that this is where you went wrong in your selection study. Given all Choice of nanomaterial and
the deca release from E-waste, MWCNT release from electronics seems the next application; alternative
most logical step for study as it is in high use, and consumer electronics are being products/ formulations
disposed of at an ever increasing rate.
I-60
-------
Table 1-13 (Continued): Free-form comments received from experts.
Author
Comment ID ID
Context (of External
Review Draft)
Comment Text Excerpt
Themes
251
E4-F Appendix A, pg A-6
Table A-1 (SWCNT
comparison)
This is far more likely to go into commercial production with in the next decade
when compared to MWCNTs for textiles.
Choice of nanomaterial and
application; alternative
products/ formulations
252
E4-F Appendix A, pg A-7
Table A-2
(on the market)
This is incorrect. I know NanoCyl claimed it, but no one is actually buying it and
using it, so this assumption is incorrect. I strongly disagree that you can make this
statement with high confidence given what drives material selection in the flame
retardant field.
Choice of nanomaterial and
application
253
E4-F Appendix A, pg A-8
lines 15-16
Now that I see this ... I see where the problem really lies. The compromise hybrid
solution does not reflect reality since it misses the realistic drivers which determine
what flame retardants will be used in a specific application. I would propose you go
back and look at SWCNTs in textiles or MWCNTs in flame-retardant composites
(only when the MWCNTs are combined with other flame retardants though) and
pick one, rather than pursuing this hybrid.
Choice of nanomaterial and
application; choice
constricted science of
MWCNTs
1-61
-------
Table 1-14. Comments received in response to the public comment period.
Comment ID Author ID Comment Text Excerpt
Themes
P1
"... there is significant information on decaBDE which is not included ... example ... cited water solubility of
20-30 ug/L as reported in the 1994 WHO IPCS document... was based on the 1970's published work of
Morris et al., and reflects that of the then commercial product which was approximately 77%
decabromodiphenyl ether... That... does not reflect the commercial product (>=97% decabromodiphenyl
ether) that has been in use for about 20 years. The water solubility of decaBDE is considerably lower than
20-30 ug/L. The measured water solubility, determined in a guideline/GLP-compliant study in 1997, of the
>=97% decaBDE product is <0.1 ug/L"
DecaBDE data needs
improvement; p-chem
properties
P1
"I would like to make you aware of critical papers on decaBDE
emails due to size."
I will provide these papers in separate
DecaBDE data needs
improvement; impacts
P2
"I wish to express appreciation for the EPA's methodical efforts on the three case studies. It provides a
point of constancy across a broad and widely dispersed community."
Praise for Case Study;
nanomaterial series
P2
"Comparison allows reader to consider incremental risk ... The range of issues considered in a CEA leads
the reader to assume that every question must be resolved. The comparison brings perspective and some
surprise on how public knowledge for a registered substance may still have many gaps worthy of further
examination."
Praise for Case Study;
comparison element
P2
"The Agency continues to grapple with using the CEA to formulate and prioritize research goals. With the
three case studies, this is illustrated by the changing format for the external review: from a large, diverse
meeting for TiO2 to a much smaller one for Ag and, now, to a two day mix of public and closed meetings,
the latter incorporating an RTI-led methodology."
CEA framework/
methodology; purpose
of CEA
P2
"With the MWCNT case study, the flame-retardant formulation imposes constraints on MWCNT chemistry,
where the crux is surface functionalization. While decaBDE is comparable to other brominated substances
when preparing a flame-retardant formulation, MWCNT chemistry is not as robust and should not be
generalized to be any MWCNT (PEI functionalized, pristine, carboxylated, and so on). ... By being drawn to
a specific application, flame retardants in textiles, the reader expects the CEA to focus consideration on the
relevant MWCNT chemistry, which is not clear in the draft. For this case study, two categories of MWCNTs
are involved; one manufactured and one surface modified."
Choice of nanomaterial
and application; choice
constricted science of
MWCNTs
I-62
-------
Table 1-14 (Continued): Comments received in response to the public comment period.
Comment ID Author ID Comment Text Excerpt
Themes
P2
"The tension between description (draft case study) and evaluation (identifying meaningful, constructive
research goals) occurs at the point of selecting a specific application when examining a class of materials.
Does one focus on the particulars of the example, as I have done, or does one use the particulars as an
introduction to the generalized chemistry, as the three case studies do?"
CEA framework/
methodology; purpose
of CEA
P2
"This specific application is not well established commercially, which means that there are few who can
speak from their "own knowledge of multiwalled carbon nanotubes (MWCNT) and flame-retardant
materials" (page 6-1). The external group is being asked to craft a context while addressing an
encyclopedic listing of facts and literature citations."
Choice of nanomaterial
and application; choice
constricted science of
MWCNTs
P2
"the textile case involves two forms of MWCNT, the commercial one and the modification found in the fire
retardant formulation. ... MWCNT kinds (classes, groupings, sub-categories) are not visible in the draft,
which is written to generalize MWCNT chemistry in order to encompass the many stages of a CEA."
Choice of nanomaterial
and application; choice
constricted science of
MWCNTs
10
P2
"Recent studies have demonstrated that carboxylated CNTs are less likely to induce profibrogenic effects
(Reference for MWCNTs: Wang et al., 2011, ACS Nano 5(12):9772-9787) and are more likely to degrade
(Reference forSWCNTs: Liu et al., 2010, Carbon 48:1961-1969). ... These references are not in the draft
and would bolster the discussion surrounding Jain's article (page 5-3). As with Magic Nano and aerosol
spray exposure, colloid facilitated transport is a significant factor."
MWCNT physical/
chemical properties
11
P2
"Also, a discussion on the meaning of functionalizing would be appropriate."
Material Processing;
define terminology
12
P2
"... the difficult theme of scientific relevance. I note that the earlier case studies cited presentations to
society meetings and unpublished papers. The current draft is more circumspect."
CEA framework/
methodology;
references/ literature
13
P2
"Yet, an EPA-sponsored document inherently provides an imprimatur or implies a relevancy to any articles
cited. Unfortunately, there have been several mis-steps in the literature, especially for carbonaceous
materials, e.g., fullerenes (Henry et. al, Env. Health Perspectives, 2007, 115(7): 1059-1065) and carbon
nanotubes (Jakubek et al, Biomaterials 2009, 30:6351-6357). And, it may take a considerable time for
such mis-steps to be undone, e.g., ~3 years for dispersing fullerenes in THF. This means that the CEA
methodology, in taking a broad view of product chemistry at all points along the product life cycle and in
attempting scientific currency, encourages the citing of ephemeral literature, meaning articles that are
within the time scale of the literature's self-correcting dynamics. Insubstantial questions result placing the
review panel in an awkward position."
CEA framework/
methodology;
references/ literature
I-63
-------
Table 1-14 (Continued): Comments received in response to the public comment period.
Comment ID Author ID Comment Text Excerpt Themes
14 P2 "Obviously, the colleagues at the October meetings face challenges. Supplementing the life cycle approach CEA framework/
with concepts from the NNI's EHS plan or the principles from the recent NRC report on EHS are options, methodology; other
but utilizing the Agency's approach to SNURs and SNUNs is perhaps more pertinent." frameworks
15 P2 "As a possible contribution to this dialog, examining the MWCNT case history using SNUR (material CEA framework/
submitted for a PMN) and SNUN (extension of a SNUR to new applications/use patterns) perspectives, methodology; questions
leads to questions on MWCNT categories and their relationship to properties. to consider for risk
A: Categorization of MWCNTs: assessment
1. Are categories of MWCNTs recognizable, as argued above?
2. Do the many papers, including past mis-steps, point to MWCNT groupings (sufficient equivalency for
risk assessment purposes)?
3. Can biological outcomes be combined with physicochemical characterization in defining an MWCNT
sub-category?
4. Is there sufficient equivalency among the carboxylated MWCNT studies to identify it as an MWCNT
group that requires a separate PMN (SNUR) and a separate CEA?
5. Does the flame retardant specific case require a modified MWCNT that nevertheless remains part of
the unmodified MWCNT group?
B: Categorization of Properties:
1. Can MWCNT properties be divided into intrinsic and context dependent categories?
2. Are biological responses from in vitro or in vivo testing better described as dependent outcomes than
as independent properties?
3. Do MWCNTs display the chemical-particle duality of nanoscale-Ag?
4. Is biopersistence a context-dependent property of the use pattern such that durability in non-lung
fluids can dominate risk assessment for some MWCNT modifications?
5. What are the CEA instances of characterization (defined stages in the life cycle that should be
common to all MWCNT materials, e.g., "as manufactured") that are useful for comparisons and
informatics?
6. Can the relevance of the case study's cited literature be informed by these considerations?"
16 P2 "In summary, it may be that an MWCNT- CEA is not well suited for setting research priorities when being CEA framework/
constrained to a linear raw material-to landfill sequence, while also incorporating the recent scientific methodology; other
literature. The virus life cycle, for example, starts with a virus entering the cell, follows virus replication and frameworks
cell exit, but does not normally include dispersal in the environment. Using SNUR/SNUN administrative
concepts, or perhaps those of other templates, in combination with the CEA would provide a useful
perspective and focus when setting priorities."
17 P2 References (The commenter provided references). References
I-64
-------
Table 1-14 (Continued): Comments received in response to the public comment period.
Comment ID Author ID Comment Text Excerpt
Themes
18
P3
"The EPA's "Nanomaterial Case Study: Comparison of Multiwalled Carbon Nanotube and
Decabromodiphenyl Ether Flame-Retardant Coatings Applied to Upholstery Textiles (DRAFT)" does not
adequately address the potential impacts to wastewater treatment plants (WWTPs) from the introduction of
multiwalled carbon nanotubes (MWCNTs)."
Other impacts;
Additional
considerations for
section
19
P3
"The study identifies the potential release of MWCNTs to WWTPs during the manufacturing, storage and
distribution, use, and reuse/ recycling/ end-of-life stages (2-29). The study's executive summary states that
laboratory-based studies of MWCNTs established acute antimicrobial activity at low exposure
concentrations (xxii). Many WWTPs use biological treatment extensively in their processes. MWCNTs
behavior and interaction with WWTP microbiology may be influenced by: size, morphology, surface area,
chemical composition; surface chemistry and reactivity; solubility and dispersion; and conductive,
magnetic, and optical properties (1-19). It is essentially unknown how MWCNTs will behave in WWTPs or
what their potential impacts on biological treatment processes may be. It is strongly recommended that
evaluation of potential impacts to WWTPs be incorporated into the EPA's comprehensive environmental
assessment (CEA) framework during the impact assessment stage whenever the potential for the
introduction of a contaminant to a WWTP during its life cycle is identified."
Other impacts;
Additional
considerations for
section
23
P1
"...please seethe following web address for decaBDE's EU risk assessment completed in 2002:
http://esis.irc.ec.europa.eu/doc/risk assessment/REPORT/decabromodiphenyletherreportOI 3.pdf. That
document discusses decaBDE's use in textiles, and has an up to date (as of 2002) discussion of
decaBDE's toxicology."
DecaBDE data needs
improvement; impacts
24
P1
"For information on decaBDE's potential for absorption and metabolism, please see Hardy et al. 2009
(Critical Reviews in Toxicology) and the Biesemeier et al. papers I sent last week. After 20 years studying
decaBDE, my opinion is that publications over the last decade claiming substantial absorption or
metabolism of decaBDE are incorrect and are due to faulty methodology."
DecaBDE data needs
improvement;
toxicokinetics
254
P4
Companies manufacturing MWCNTs should be required to assess potential risks using an integrated
testing scheme that relies on best current practices in the field while actively reducing reliance on animal-
based testing methods.
CEA framework/
methodology; Future of
risk assessment
255
P4
We recognize EPA's recent efforts to encourage manufacturers to test nanomaterials using nanomaterial-
specific, high-throughput, analytical and in vitro methods rather than relying on animal-based methods that
have not proven reliable for this purpose.
In vitro methods
I-65
-------
Table 1-14 (Continued): Comments received in response to the public comment period.
Comment ID Author ID Comment Text Excerpt
Themes
256
P4
We note that EPA's recent Nanomaterial Testing Strategy outlines a preference for using analytical
methods for in-depth characterization followed by assessment of toxicity using in vitro methods. Within the
Strategy, animal tests are considered to be a final tier of toxicity testing and are to be considered on a
case-by-case basis. We encourage EPA to maintain this position as data needs for multi-walled carbon
nanotubes continue to be assessed, rather than resort to the de facto prescription of unproven animal-
based testing.
In vitro methods; Future
of risk assessment
257
P4
259
P4
Consistent with EPA's 2009 Strategic Plan for Evaluating the Toxicity of Chemicals recommendations
contained in the National Academy of Science's Toxicity Testing in the 21 st Century: A Vision and a
Strategy should be followed for any testing of nanomaterials. This report states, "[T]oxicity testing is
approaching a scientific pivot point... It is poised to take advantage of the revolutions in biology and
biotechnology. Advances in toxicogenomics, bioinformatics, systems biology, epigenetics, and
computational toxicology could transform toxicity testing from a system based on whole-animal testing to
one founded primarily on in vitro methods..." The field of nanotechnology is in a position to take full
advantage of these new approaches, and we expect that EPA will be involved with continued development
and validation of new, primarily non-animal methods for assessment of nanomaterials.
Future of risk
assessment
258 P4 We request that EPA make clear to registrants that the test methods required will be decided on a case-by- In vitro methods
case basis with preference given to in vitro nanomaterial-specific methods.
In addition to the problems with extrapolating information from animal studies to humans for conventional
chemicals, nanomaterials possess unique physical and toxicological properties that render animal testing
even more problematic. Well known confounding issues include: variations in responses to chemicals in
different species and strains of animals, variations in target organs and tissue effects in different species
and strains, as well as different toxic thresholds between species including humans
Highlight data gap;
Analytical techniques
260
P4
Additionally, because mass concentration is not sufficient for comparison of nanomaterials of the same
chemical composition and because number concentration is affected by the subjective exclusion
parameters as well as by the analytical parameters used, toxicokinetics is then even more problematic for
nanomaterials. Further, most laboratories are not even capable of making the measurements required for
nanomaterial-related toxicokinetics, thereby making these studies even more suspect.
Toxicokinetics;
Analytical techniques
261
P4
As EPA notes, many factors influence the toxicity profile of MWCNTs. Critical factors include variations in
aggregation/agglomeration, bundling, fiber length, fiber width, surface functionalization, surface coating,
heavy metal contaminants, and wall number. Each factor listed (as well as others yet to be identified or
studied) is capable of completely changing the toxicity profile and the manner in which the MWCNT is
absorbed and distributed in the body.
p-chem properties
262
P4
Paxton, JW. The allometric approach for interspecies scaling of pharmacokinetics and toxicity of anti-
cancer drugs. Clin. Exp. Pharmacol. Physiol. 1995; 22: 851-854.
Toxicokinetics
I-66
-------
Table 1-14 (Continued): Comments received in response to the public comment period.
Comment ID Author ID Comment Text Excerpt
Themes
263
P4
266
P4
Wako, K., et al. Effects of preparation methods for multi-wall carbon nanotube (MWCNT) suspensions on
MWCNT induced rat pulmonary toxicity. J Toxicol Sci. 2010 Aug; 35(4):437-46.
Analytical techniques
264 P4 Liu, D et al. Different cellular response mechanisms contribute to the length-dependent cytotoxicity of multi- Impacts; In vitro
walled carbon nanotubes. Nanoscale Res Lett. 2012 Jul 2;7(1):361. methods
265 P4 Coccini, T. et al., Toxicology. Effects of water-soluble functionalized multi-walled carbon nanotubes Impacts; In vitro
examined by different cytotoxicity methods in human astrocyte D384 and lung A549 cells. Toxicology. 2010 methods
Feb 28, 269(1 ):41-53.
EPA also notes that many studies result in data that may be in conflict with existing data. This conundrum
is due in part to the extreme heterogeneity of MWCNTs and also because animal-based studies have
repeatedly failed to give reproducible results. Because of the infinite number of toxicity profiles from the
unending number of modifications that can be made on each type of MWCNT, it is impossible to conduct
conclusive in vivo studies for any class of nanomaterials, and MWCNTs are not an exception.
In vitro methods;
Highlight data gap
267
P4
While EPA describes inhalation toxicity as one of the main exposure scenarios of concern, it should be
noted that inhalation toxicity testing using animals results in a myriad of practical and scientific problems.
Inhalation toxicity testing using rats has proven to be expensive and technically challenging with respect to
delivering an appropriate dose of nanomaterials.
Highlight data gap;
Analytical techniques
268
P4
More importantly, biological relevance is unlikely with rat-based tests. Issues relating to breathing mode,
physiology, relative sizes of nerve bulbs and the different rate of particle clearance of rats compared to
humans all point to important anatomical and physiological differences that preclude clear data
extrapolation between species.
Toxicokinetics;
Analytical techniques
269
P4
Warheit, D. Nanoparticles Health Impacts? Nanomaterials Today. 2004;7: 32-35.
Impacts
270
P4
Nikula, KJ, et al. Influence of exposure concentration or dose on the distribution of particulate material in
rat and human lungs. Environmental Health Perspectives. 2001; 109(4): 311-318.
Analytical techniques
271
P4
Intratracheal instillation and laryngeal aspiration are the least costly in vivo methods and therefore are the
most commonly used methods for assaying the pulmonary toxicity of nanomaterials. However, scientific
concerns related to these methods are expressed by the European Commission (EC): "[t]his mode of
exposure is not physiological...the lung surface receives particles contained in a liquid, which is likely to
affect the defense systems of the lung." Thus the EC clarifies that often, the reaction is to the liquid bolus
highly concentrated with nanomaterials, and the response often has little to do with the actual toxicity of a
given nanoparticle. The dose, dose rate, and dispersive abilities of this method are often criticized as
lacking relevance.
Highlight data gap
I-67
-------
Table 1-14 (Continued): Comments received in response to the public comment period.
Comment ID Author ID Comment Text Excerpt
Themes
272
P4
Scientific Committee on Emerging and Newly-Identified Health Risks. Opinion on the Appropriateness of
the Risk Assessment Methodology in Accordance with the Technical Guidance Documents for New and
Existing Substances for Assessing the Risks of Nanomaterials. 2007.
Impacts; Future of risk
assessment
273
P4
Osier, M and Oberdorster, G. Intratracheal inhalation vs Intratracheal Instillation: Differences in Particle
Effects. Fundamental and Applied Toxicology. 1997; 40, 220-227.
Analytical techniques
274
P4
Laryngeal aspiration has problems similar to those described for intratracheal instillation, including both a
high dose and high dose rate. The EC also notes that laryngeal aspiration also results in "unusually high
doses to the bronchioles and the induction of alveolar inflammation." The EC goes further to specify that,
"neither [intratracheal instillation or laryngeal aspiration] can be used to determine NOEL.
Analytical techniques;
Future of risk
assessment; Highlight
data gap
275
P4
Because of the problems associated with these (and other) in vivo tests, there is a concerted effort to begin
using human cell-based co-cultures to assay potential toxicity for this exposure route. In vitro models using
human cell co-cultures have proven to be informative and will help to move inhalation toxicity testing from
studies on rats toward methods that are relevant to humans.
In vitro methods; Future
of risk assessment
276
P4
Gasser, M. et al., Pulmonary surfactant coating of multi-walled carbon nanotubes (MWCNTs) influences
their oxidative and pro-inflammatory potential in vitro. Part Fibre Toxicol. 2012 May 24;9(1):17.
In vitro methods;
Highlight data gap
277
P4
Geys J, Nemery B, Hoet PH. Optimisation of culture conditions to develop an in vitro pulmonary
permeability model. Toxicol In Vitro. 2007; 21(7): 1215-9.
In vitro methods;
Analytical techniques
278
P4
Bur M, Rothen-Rutishauser B, Huwer H, Lehr CM. A novel cell compatible impingement system to study in
vitro drug absorption from dry powder aerosol formulations. Eur J Pharm Biopharm. 2008 Aug 17.
In vitro methods;
Highlight data gap
279
P4
Human -relevant in vitro methods: VitroCell has published on a variety of experiment types, including in
vitro repeat exposure, in vitro exposure using 3D organotypic exposure to cigarette smoke (as an
alternative to rat inhalation), exposure to complex mixtures and sequential exposure to pollutants, as well
as reports on prevalidation studies on the toxic effects of inhalable substances.
In vitro methods;
Highlight data gap
280
P4
http://www.vitrocell.com/index.php?Nav_Nummer=8&
In vitro methods;
Highlight data gap
281
P4
Human -relevant in vitro methods: Companies such as MatTek have tested nanomaterials in concert with
corporations such as Proctor and Gamble and have achieved in vitro-in vivo data concordance using the
MatTek EpiAirway three-dimensional tissue constructs.
In vitro methods;
Highlight data gap
282
P4
http://www.mattek.com/pages/nanoparticles/
In vitro methods;
Highlight data gap
-------
Table 1-14 (Continued): Comments received in response to the public comment period.
Comment ID Author ID
283 P4
284 P4
Comment Text Excerpt
In vitro models using cell cultures and co-cultures have proven to be informative.
Geys J, Nemery B, Hoet PH. Optimisation of culture conditions to develop an in vitro pulmonary
Themes
In vitro methods
In vitro methods;
permeability model. Toxicol In Vitro. 2007; 21(7): 1215-9.
Highlight data gap
285
P4
Bur M, Rothen-Rutishauser B, Huwer H, Lehr CM. A novel cell compatible impingement system to study in
vitro drug absorption from dry powder aerosol formulations. Eur J Pharm Biopharm. 2008 Aug 17.
In vitro methods;
Highlight data gap
286
P4
Rothen- Rutishauser et al. developed a triple co-culture comprised of epithelial cells, macrophages, and
dendritic cells which the author states, "simulates the most important barrier functions of the epithelial
airway." Measurement of cellular responses to MWCNTs, including reactive oxygen species, release of
tumor necrosis factor, and apoptosis illustrate the effectiveness of pre-coating MWCNTs to decrease
toxicity. This model system has been compared to in vivo results with good data concordance.
In vitro methods;
Highlight data gap
287
P4
Gasser, M. et al., Pulmonary surfactant coating of multi-walled carbon nanotubes (MWCNTs) influences
their oxidative and pro-inflammatory potential in vitro. Part Fibre Toxicol. 2012 May 24;9(1):17.
In vitro methods;
Highlight data gap
288
P4
Alfaro-Moreno et al. found that bicultures and tricultures of human lung cells released granulocyte colony-
stimulating factor (G-CSF), macrophage inflammatory protein (MIP)- 1beta, interleukin (IL)-1beta, IL-6,
tumor necrosis factor alpha, and MIP-1 alpha. The authors go on to state that these effects are consistent
with those systemic effects described for particulate matter and correspond to inflammation, endothelial
dysfunction, and bone marrow cell mobilization.
In vitro methods;
Highlight data gap
289
P4
Alfaro-Moreno, E. et al. Co-cultures of multiple cell types mimic pulmonary cell communication in response
to urban PM10. Eur RespirJ. 2008; 32:1184-1194.
In vitro methods
290
P4
Cavallo, et al. made use of A549 (human lung epithelial) cells to study the mode of toxicity caused by
MWCNTs. Early cytotoxic and genotoxic effects were observed (including membrane damage, surface
morphological changes, and direct DNA damage). These data contribute to understanding the mechanism
by which MWCNTs may induce toxic effects.
In vitro methods;
Highlight data gap
291
P4
Cavallo, D. et al., J Appl Toxicol. Jan 23. Multi-walled carbon nanotubes induce cytotoxicity and
genotoxicity in human lung epithelial cells. J Appl Toxicol. 2012 Jun; 32(6):454-64.
In vitro methods
292
P4
Because nanomaterials differ from traditional chemicals and have proven difficult to test using some of the
outdated animal-based methods used for traditional chemicals, it is critical to completely and accurately
characterize nanomaterials, as is described by EPA's Nanomaterial Testing Strategy, and to then apply in
vitro and in silico methods within an integrated testing strategy (ITS). ITS take into account existing data to
design a rational, chemical-specific testing strategy to satisfy regulatory needs without relying primarily on
animal testing.
Future of risk
assessment
-------
Table 1-14 (Continued): Comments received in response to the public comment period.
Comment ID Author ID Comment Text Excerpt
Themes
293
P4
We suggest that MWCNTs (as well as other nanomaterials) be tested using NexGen Respiratory Toxicity Future of risk
Model developed by EPA's Office of Research and Development. This model system takes advantage of assessment; in vitro
both an air-liquid-interface in vitro cell-based construct coupled with omics-based mechanistic pathway and methods
biomarker identification.
294
P4
http://nas-sites.org/emergingscience/files/2012/06/Devlin.pdf
Future of risk
assessment
295
P4
In the event that additional toxicity data is requested, we urge EPA to require manufacturers to use high-
throughput methods that have been specifically designed for MWCNTs in order to reduce reliance on
animal-based testing, as described above.
Future of risk
assessment; in vitro
methods
296
P4
We look forward to seeing progress made toward the replacement of animal-based testing methods and
encourages EPA to continue working toward this goal.
Future of risk
assessment; In vitro
methods
297
P4
Environmental Protection Agency, "Nanomaterial Research Strategy," 2009.
Future of risk
assessment
298 P4 Environmental Protection Agency. 2009. The U.S. Environmental Protection Agency's Strategic Plan for Future of risk
Evaluating the Toxicity of Chemicals. Office of the Science Advisor, Science Policy Council, U.S. EPA assessment
(http://www.epa. gov/osa/spc/toxicitytesting/docs/toxtest_strategy_032309.pdf
299
P4
National Research Council, "Toxicity Testing in the Twenty-First Century: A Vision and a Strategy," report Future of risk
of the Committee on Toxicity and Assessment of Environmental Agents, June 2007. assessment
I-70
-------
Appendix I References
Aschberger. K: Johnston. HJ: Stone. V: Aitken. RJ: Hankin. SM: Peters. SA: Iran. CL: Christensen. FM. (2010).
Review of carbon nanotubes toxicity and exposureAppraisal of human health risk assessment based on open
literature [Review]. Crit Rev Toxicol 40: 759790. http://dx.doi.org/10.3109/10408444.2010.506638
ATSDR (Agency for Toxic Substances and Disease Registry). (2004). Toxicological profile for polybrominated
biphenyls and polybrominated diphenyl ethers [ATSDR Tox Profile]. Atlanta, GA: Agency for Toxic
Substances & Disease Registry, http://www.atsdr.cdc.gov/toxprofiles/tp68.pdf
Bai Y: Zhang. Y: Zhang. J: Mu. Q: Zhang. W: Butch. ER: Snyder. SE: Yan. B. (2010). Repeated
administrations of carbon nanotubes in male mice cause reversible testis damage without affecting fertility.
Nat Nanotechnol 5: 683-689. http://dx.doi.org/10.1038/nnano.2010.153
Gustavsson. P: Hedmer. M: Rissler. J. (2011). Carbon nanotubes: Exposure, toxicology and protective measures
in the work environment. Swedish Work Environment Authority.
https://lup.lub.lu.se/luur/download?func=downloadFile&recordOId=2295102&fileOId=2300375
Healy. ML: Dahlben. LJ: Isaacs. JA. (2008). Environmental assessment of single-walled carbon nanotube
processes. J IndEcol 12: 376-393. http://dx.doi.org/10.1111/i. 1530-9290.2008.00058.X
Holbrook. RD: Kline. CN: Filliben. JJ. (2010). Impact of source water quality on multiwall carbon nanotube
coagulation. Environ Sci Technol 44: 1386-1391. http://dx.doi.org/10.1021/es902946j
OECD (Organisation for Economic Co-operation and Development). (2012). Important issues on risk assessment
of manufactured nanomaterials. In Series on the Safety of Manufactured Nanomaterials, No 3 3.
(ENV/JM/MONO(2012)8). Paris.
http://search.oecd.org/officialdocuments/displavdocumentpdf/?cote=env/im/mono(2012)8&doclanguage=en
Pauluhn. J. (2010). Subchronic 13-week inhalation exposure of rats to multiwalled carbon nanotubes: Toxic
effects are determined by density of agglomerate structures, not fibrillar structures. Toxicol Sci 113: 226-242.
http://dx.doi.org/10.1093/toxsci/kfp247
Shen. M: Xia. X: Wang. F: Zhang. P: Zhao. X. (2012). Influences of multiwalled carbon nanotubes and plant
residue chars on bioaccumulation of poly cyclic aromatic hydrocarbons by Chironomus plumosus larvae in
sediment. Environ Toxicol Chem 31: 202-209. http://dx.doi.org/10.1002/etc.722
U.S. EPA (U.S. Environmental Protection Agency). (2012). Nanomaterial case study: a comparison of
multiwalled carbon nanotube and decabromodiphenyl ether flame-retardant coatings applied to upholstery
textiles (draft) [EPA Report]. (EPA/600/R-12/043A). RTF, NC.
Zhang. L: Petersen. EJ: Huang. Q. (201 la). Phase distribution of 14C-labeled multiwalled carbon nanotubes in
aqueous systems containing model solids: Peat. Environ Sci Technol 45: 1356-1362.
http://dx.doi.org/10.1021/esl026097
Zhang. Q: Huang. JQ: Zhao. MQ: Qian. WZ: Wei. F. (201 Ib). Carbon nanotube mass production: principles and
processes [Review]. ChemSusChem4: 864-889. http://dx.doi.org/10.1002/cssc.201100177
1-71
-------
Appendix J. Knowledge Map Pilot:
Environmental Transport, Transformation,
and Fate
Appendix J presents work from a pilot project to continue refining the comprehensive
environmental assessment (CEA) approach. The main objective of this work is to develop methods for
visually and concisely representing information in the CEA framework. Specifically, Chapter 3.
"Environmental Transport, Transformation, and Fate" in the External Review Draft of this case study
document (Chapter 3. Section G.3. and Section H.3 in the final document) was selected to develop a proof
of concept for a particular method to convey information visually to decision-makers (e.g., research
planners, risk managers). This method is briefly described here, followed by a presentation of the
products to date from this pilot.
J.1. Knowledge Maps
Knowledge maps (KMs) are used in a variety of fields (e.g., organization management,
journalism) to present concepts and how they relate to one another visually (Novak and Canas. 2008; Kim
et al., 2003). Given their wide applicability and utility in representing complex concepts, knowledge maps
were selected for this pilot to convey information in the CEA framework visually. A KM comprises a
network of nodes (usually presented as common shapes) connected by edges (depicted as lines or arrows).
Nodes usually represent concepts, whereas edges show relationships between the concepts.
J.2. Transport, Transformation, and Fate Knowledge Maps:
Overview
The KMs developed in this pilot are three conceptually linked portions of one map displaying
environmental transport, transformation, and fate for either decabromodiphenyl ether (decaBDE) or
multiwalled carbon nanotubes (MWCNTs). Showing all of the information related to environmental
transport, transformation, and fate on one map would result in excessive detail in a small amount of space
and thus, for readability, the information pertinent to each process is displayed in three distinct maps.
The Physicochemical Properties Map (PPM) focuses on the material itself, providing a sense of the basic
J-1
-------
properties of the material, which might in turn influence how the material partitions in the environment
after release from a product life cycle stage (e.g., product manufacturing, use, disposal/recycling).
The Transport Map (TpM) portrays how the material might subsequently move between environmental
zones (i.e., aquatic, atmospheric, terrestrial) based on available information. In addition, it provides a
sense of which zone(s) are sinks for the material, and thus would likely have higher concentrations of the
material than others. Once the material makes it to a zone, the Transformation Map (TfM) depicts how it
might be transformed due to the combination of material and environmental properties within a given
zone.
Note that because the maps are conceptually a single unit, symbols mean the same thing across all
the maps. For example, black edges with solid arrowheads represent movement whenever they appear in
any map. Similarly, the environmental zones are visually similar in all three maps.
Each type of map (i.e., PPM, TpM, TfM) is described in greater detail below, along with a brief
comparison of the decaBDE and MWCNT maps in each category.
J.3. Physicochemical Properties Maps
The PPM41 shows the properties of a material likely to influence its behavior in the environment,
the corresponding effects on environmental behavior, and finally how resulting behaviors influence
movement to environmental zones.
The PPM has three banks of nodes. In the top bank, material properties are represented in
rectangular nodes. The nodes are color-coded based on information in Table 3-1 and Table H-3; nodes are
dark red if the value is high and light red if the value is low,42 whereas light green means that the property
can vary according to the specific sample of the material (e.g., MWCNTs with one type of surface coating
versus another).
The middle bank of nodes contains effects in ovals. Like the properties, they are color-coded so
that large effects are dark yellow, while small effects are light yellow. Where variability exists, such that
the effect can be high or low, both a dark yellow circle and a light yellow circle have been placed in the
box that represents that effect. A white oval with a dashed border indicates an effect that is predicted to
41Note: These maps are based primarily on Table 3-1 (for MWCNTs) and Table H-3 (for decaBDE), and although
details from other parts of Chapters. Section G.3. and Section H.3 are included where relevant, the PPMs represent
a translation of information in tabular form to figures in this pilot work.
42Note: Scaling the shading of red to present values more specific than "low" or "high" was determined to be too
confusing for this representation.
J-2
-------
occur based on material properties but has not yet been observed. The effect nodes are connected to the
property nodes via green edges with diamond arrowheads; these edges indicate a simple causal
relationship, that is, that a particular property leads to a particular effect. Green edges that connect to
boxed effect nodes indicate that the given property might lead to a large or small effect, given varying
circumstances.
The lower bank of nodes contains the environmental zones, which are more fully described in the
transport map. Black edges connect effect nodes to zone nodes to indicate that the effect facilitates
transport to that zone. Levels of transport are differentiated, with high transport represented by thick, solid
lines and low transport represented with thin, dashed lines. In all three types of maps (i.e., PPM, TpM,
TfM), a dotted line around a spatial zone node means that no data were identified from experimental
measurements on the presence or absence of the compound in the zone.
Overall, the map can be read as a three-part sentence that follows the edges from property to
effect, to zone. For example, "the low water solubility43 of decaBDE leads to high sorption to organic
matter, which causes a high level of transport to the soil."
Although the decaBDE and MWCNT PPMs are consistent with each other in their use of symbols
the two maps have notable differences. For example, the PPM for MWCNTs includes more types of
nodes (e.g., light green property nodes and boxed effect nodes) because MWCNT properties can vary
based on size, shape, surfactant, and other conditions. In addition, although all property nodes in
decaBDE are separated by white space, property nodes that are two poles of the same property (like
"Size: single" and "Size: cluster") are placed next to each other on the MWCNT map, again due to the
variability associated with the material.
43See Footnote 16 in Chapter 2 for information on the use of the term "solubility" in this document.
J-3
-------
Physicochemical Properties
Properties
Low Value
Effects
S^LwT\
/"TSgiTN
\ Fffar-t )
Edge Types
Laads To
'
High
Transport To
Low
Transport To
Zone Concentration
\ No Data i
Data Available
1 Known Sink
Figure J-1. Physicochemical properties map for decabromodiphenyl ether.
J-4
-------
Physicocnemical Properties
Atmospheric
Indoor Air
Residential :
Commercial
: Outdoor Air
: : Rural :
: : Urban :
Tern
Soil
istrial
*••.'...
Biota
: Animals :
i Plants :
Aquatic
Surface
Water
Sediment
Waste
Water
Sewage
Sludge
Ground
Water :
Biota
Plants :
Animals :
Figure J-2. Physicochemical properties map for multiwalled carbon nanotubes.
J.4. Transport Map
The TpM begins with the release of a material in a product life cycle stage (see PPM). From
there, the map shows how the material might travel between or within spatial zones, the environmental
properties that affect material transport, and some processes that describe the movement.
The TpM has only one bank of nodes, containing the environmental zones. Within each of three
main zones—atmospheric (grey), terrestrial (brown), and aquatic (blue)—are more specific subzones
J-5
-------
(e.g., wastewater, surface water, sediment within the aquatic zone). Zones that are sinks for the material,
based on information available in Chapter 3. Section G.3, and Section H.3 of the document, have a thick
black border. Dotted borders indicate that no data on the concentration of the material in that zone were
found.
Transport is represented by solid or dotted black lines leading from one zone to another. The bold
solid lines represent movement described as "most likely to occur," based on information available in
Chapter 3. Section G.3. and Section H.3. Dotted lines thus represent data inferred from other materials or
likely due to chemical properties, but for which explicit measurements have not been taken (e.g.,
MWCNTs leeching from soil to ground water). For instance, in the decaBDE TpM, a dotted line is used
to depict decaBDE transport from surface water to sediment is most likely to occur based on low water
solubility and other material properties (see Figure J-6).
For many of the transport processes, the text in Chapter 3. Section G.3. and Section H.3 used a
term to describe the process (such as deposition, leaching, or runoff), and those terms are listed as a label
on the appropriate edge. Any environmental properties of the spatial zones that were cited in the chapter
as affecting material transport are represented by numbers to highlight the influence these factors might
have on the extent, rate, or direction of transport.
Overall, the maps can be read as a sentence that follows the edges from release to a spatial zone
and then to another spatial zone. For example, "decaBDE is released to the atmospheric zone, where it
can be transported through deposition to wastewater, with the extent of deposition influenced by the
amount of total organic carbon (TOC) in the air, precipitation, wind, and temperature."
In comparing the decaBDE and MWCNTs TpMs, the spatial zones and environmental properties
clearly are the same; however, as expected, the specific types of transport that occur and the
environmental properties that influence them differ. More notably, MWCNTs have no known sinks,
whereas decaBDE has three (soil, sediment, and sewage sludge). For MWCNTs fewer data on
concentrations in the zones are available and therefore many more dotted lines appear on the map than on
the decaBDE map.
J-6
-------
decaBDE
release
Sediment | |Sewage Sludge)
Zone Concentration
: No Data |
Data Available
Known Sink |
Legend
Edge Types
Movemenl ^
Strong Movement.
Environmental Properties
HOC 6. Temp
2. Metal Contenl 7. Ionic Sirength
3 UV Light 8 pH
5. Wind 10. Time
Figure J-3. Transport map for decabromodiphenyl ether.
J-7
-------
^^^^^^^H I MWCNT 1
^^^mumj^n^^i^ re|ease ^j^jjjjjgjjjujjj^
Deposilion (4,5,6,10) grosim
! i v ;
Leaching
Deposition (4,5,6.10)
^^^^^^^^^1 Legend
Edge Types Environmental Properties Zone Concentration
Suspected Movement ' Toc 6. Temp : MO Data |
Known Movement j 3 uu Llghl 8. pH Data Available
Strong Movement «• Pr«lp 9. Redox Potential
5. Wind 10. Tim*
T
Aquatic
<•• Waste Water Biota I
! Animals | '-.
Sewage Sludge
A
Figure J-4. Transport map for multiwalled carbon nanotubes.
J.5. Transformation Map
Each TfM begins with the material entering spatial zones, either from release in a product life
cycle stage (see PPM) or movement between spatial zones (see TpM). From there, the map shows
transformations that occur within those zones, the environmental properties that affect the transformation,
and the resulting compounds.
Three banks of nodes occur in the TfMs. The top bank represents the spatial zones (carried over
from the previous two maps and including known sinks). The second bank represents the transformations
that are associated with the spatial zones. Transformations are grouped into three main categories:
chemical, biological, and physical, which are represented by yellow (photo, thermal, and geochemical
degradation), red (sorption), or green (biological degradation) parallelograms, respectively. Many of the
spatial subzones are associated with specific transformations; these instances are denoted by a red edge
linking the subzone with the particular transformation (e.g., plants within "terrestrial biota" are associated
with biological degradation in the decaBDE map). An edge to the main "Transformations" box implies an
association with all of the transformations within that box (e.g., soil within "terrestrial" is associated with
J-8
-------
biological, photo, and geochemical degradation in the decaBDE map). When the text in Chapter 3.
Section G.3. and Section H.3 cites reason to infer data that have not been explicitly measured or proven, it
is represented by a dashed line throughout the map. As in the TpM, available information from the text on
the environmental properties that might influence transformation is represented by numbers next to the
relevant edge connecting the zone and the transformation.
The third bank of nodes is nested inside the transformations and represents the compounds that
result from the transformation that occurred. A black line with an open circle is used as the edge
connecting each transformation to the resulting compound. The resulting compounds are then represented
by solid black shapes, such as a triangle to represent polybrominated/-chlorinated dibenzofurans in the
decaBDE map.
The spatial zones, general transformation types, and environmental properties are the same in the
legend of both the decaBDE and MWCNT TfMs; however, the maps differ with regard to material in the
specific transformations that occur and the environmental properties that influence them. Notably, fewer
types of transformation products are associated with MWCNTs compared with decaBDE, although this
might be due to a lack of data. One of the MWCNT transformation products (MWCNT*) refers to
MWCNTs that are slightly altered (e.g., changes in surface charge), yet are still accurately described as
MWCNTs. Similarly, "bundled MWCNTs" refers to aggregated or agglomerated MWCNTs (see
Footnote H for more discussion on the term "bundle"). In addition, as in the transport map, more dashed
edges occur in the MWCNT map than in the decaBDE map, representing greater uncertainty in the
particular type of transformations that might occur for MWCNTs. Overall, the maps can be read as a
three-part sentence that follows the edges from spatial zone to transformation to resulting compound. For
example, "decaBDE can be transported to surface water in the aquatic zone, which is associated with
photodegradation, the extent of which is influenced by quantity of UV light and results in
polybrominated/-chlorinated dibenzo furans."
J-9
-------
Photo / Geochemlcal/
Degradation / Degradation/
Legend
Transformation Types
/Chemical/
/ Biological/
Edge Types
Movement
Associated wi
Results in Q
Environmental Properties
t.TOC 6. Temp
2. Metel Content 7. Ionic Strength
3. JVLJgN 8 pH
4.Precp 9. Reiox Potental
5. Wind 10. Time
t\
Zone Concentration
No Data :
Data Available
| Known Sink 1
Resulting Compounds
decabromodiphenyl polybrominated/-chlorinated
ether bound complex dibenzofurans
lower bromirtated polybrominated/-chlorinated
congeners dibenzo-p-dioxins
Figure J-5. Transformation map for decabromodiphenyl ether.
J-10
-------
Atmospheric
Indoor Air
Residential :
Commercial
: : Outdoor Air :
: : Rural : :
: : Urban : ;
Atmospheric Transformations
/Oxidation// Surff,e /ASu*c' /
/ // dearadation/ / dearadation/
• 1
?6,10 '•(-
•°""
t
Terrestrial
: Biota
• : Animals Soil
•
; : Plants
i
f
Aquatic
Terrestrial Transformations
/ Surface // Surface /
degradation; / degradation/
io''Q o'10
•
Biota ;
: Plants : : Animals ; :
Surface Water : Ground Water:
IT
Aquatic
Waste Water
Sediment
Sewage Sludge
Transformations
IJ.IJ.JI.IJ / . Sulface /ASurff(e //Oxidation/
•p«u"*fi" / degradation, , degradation// "_/
;rij -v
'"10
Legend
Transformation Types
/Chemical /
/Biological /
Zone Concentration
: No Data :
Data Available
or modeled
Environmental Properties
1.TOC 6. Temp
2. Metal Content 7. Ionic Strength
3. UV Light 8.pH
4. Precip 9. Redox Potential
5. Wind 10. Time
Edge Types
Movement >
Associated with
Suspected Association
Results in
Suspected Results
Resulting Compounds
Multiwall carbon
nanotubes MWCNT*
(MWCNT) (slightly altered)
• •
Amorphous Bundled
Carbon MWCNTs
• •
Figure J-6. Transformation map for multiwalled carbon nanotubes.
J.6. Future Applications of Knowledge Maps in CEA
The maps depicted in the previous sections serve as a template for visually representing available
information on other chemicals in future CEA documents. The maps thus meet the primary objective of
this proof of concept work to convey information visually that was previously described in text or tables.
Given the amount of information contained in the chapter selected for this pilot project, three maps were
generated to avoid overwhelming individuals using the maps to review available information. Yet,
common nodes and symbols connect all three maps to maintain continuity.
J-11
-------
Beyond supporting a more visual representation of available information, the KMs were
generated with the intent of facilitating the identification of data gaps. Dotted lines are used in the maps to
show clear contrasts in data availability between the two compounds when users compare maps for each
compound. Although data gaps are apparent by this comparison, using these maps to distinguish which
data gaps are most important or relevant to study is difficult. Although the maps can show whether data
are available, in their present form, they have no way of presenting the quality or quantity of data
available. Maps using a weight-of-evidence approach that presents that kind of information have been
produced for more narrowly defined contexts [see (Wiedemann et al.. 2011)1. but those results portrayed
much more detail than could be effectively shown in maps of this granularity.
A third objective of this pilot project was to determine whether the maps could be used as
templates for generating maps of other chemicals or materials. This third objective was achieved by using
the decaBDE maps to generate maps for MWCNTs. As noted above, several modifications were required
to convey information on MWCNTs using the decaBDE templates, but the modifications were relatively
easy to execute. That these same maps could serve as templates for use in future CEA case studies on
different chemicals is therefore likely.
Future efforts to use KMs in applications of CEA could focus on two objectives. The first
objective would be identifying and testing software tools other than the one used here. Although the
software used in this pilot proved useful for an initial proof of concept, it does not have good options for
automatically generating new maps via a template; thus, manual drawing of nodes and edges to match the
new data would be required. A second objective for future work could include how to represent
quantitative data in the maps. Quantitative data could include values for physicochemical properties,
ranges of concentrations of a compound in a spatial zone, or rates of transport. Such information can be
difficult to represent succinctly and visually because it can vary widely among studies, samples, and
geographic locations. Future work thus would include decisions on what data to include and how to
represent them without overcomplicating the maps. Although future efforts to expand on this pilot project
will be useful to realize the full utility of KMs in applications of CEA, this initial work demonstrates the
benefits of visually representing available information in CEA case studies through KMs.
J-12
-------
Appendix J References
Kim. S: Suh. E: Hwang. H. (2003). Building the knowledge map: An industrial case study. Journal of Knowledge
Management 7: 34-45. http://dx.doi.org/10.1108/13673270310477270
Novak. JD: Cafias. AJ. (2008). The theory underlying concept maps and how to construct and use them. (Technical
report IHMC CmapTools 2006-01 Rev 01-2008). Pensacola, FL: Florida Institute for Human and Machine
Cognition, http://cmap.ihmc.us/publications/researchpapers/theorvcmaps/theorvunderlvingconceptmaps.htm
Wiedemann. P: Schiitz. H: Spangenberg. A: Krug. HF. (2011). Evidence maps: communicating risk assessments in
societal controversies: the case of engineered nanoparticles. Risk Anal 31: 1770-1783.
http://dx.doi.0rg/10.llll/i.1539-6924.2011.01725.x
J-13
-------
Appendix K. Peer Review Draft Comments
and Agency Responses
K.1. Background
This appendix documents how the Peer Review Draft of the case study document was revised to
respond to comments from experts participating in an independent Letter Peer Review of the document.
A U.S. Environmental Protection Agency (EPA) contractor, Versar International, independently selected
five experts in fields related to life cycle and risk assessment of nanomaterials (see Table K-l) (Versar.
2013). Each expert provided written general observations and specific observations in addition to
responses to the following charge questions:
Completeness and Accuracy
1. Is the science accurately conveyed throughout the document? If not, (1) please list any areas that
need improvement and (2) provide specific comments/revisions that will more accurately convey
the science.
Clarity and Use of Priority Areas to Support Research Planning
2. Are the priority areas identified by previous experts that reviewed the MWNCT [multiwalled
carbon nanotube] Case Study ERD [External Review Draft] and participated in the workshop
process clearly highlighted throughout the document?
Throughout the 2013 PRD [Peer Review Draft] report, areas of the CEA framework are
highlighted if they were identified as priority areas (i.e., areas important to consider in future
risk assessments ofMWCNTs but generally lacking confidence in the current state of the science
to support risk management decisions) by previous experts that reviewed the 2012 MWNCT Case
Study External Review Draft (ERD) and participated in a workshop process, "Nanomaterial
Case Studies Workshop Process: Identifying and Prioritizing Research for Multiwalled Carbon
Nanotubes. "
3. For the priority areas that have been highlighted, do you know of any additional existing studies
specific to MWCNTs that should be included in the case study to help develop research plans that
subsequently support future assessment and risk management efforts for MWCNTs in flame-
retardant textile coatings?
4. Does the content of the document support research planning to inform future assessment and risk
management efforts for MWCNTs? If not, please provide specific revisions that would improve
the utility of the document for research planning to support future assessment and risk
management efforts ofMWCNTs.
K-1
-------
5. Is the organization of the document logical, and does it facilitate comprehension and
understanding? If not, please provide specific organizational revisions that would improve the
comprehension and understanding.
Responses to Public and Peer Comments on the MWCNT Case Study ERD
6. Are comments from peer and public stakeholders who reviewed the MWNCT Case Study ERD
transparently conveyed and clearly summarized (i.e., original comments making up comment
themes can be easily identified, comment themes are readily identifiable and understandable) in
Appendix I? If not, please provide specific revisions to more transparently and clearly convey
comments.
7. Are EPA responses to comments transparently and clearly conveyed (i.e., original comments
making up comment themes can be easily identified, comment themes are readily identifiable and
understandable) in Appendix I? If not, please provide specific revisions to more transparently and
clearly convey EPA responses.
8. In revising the case study document, is it evident that EPA was responsive to input from
stakeholders and made appropriate changes to the document? If not, please state any specific
instance(s) in which EPA did not respond appropriately and how the response could be revised to
respond appropriately.
Potential Utility of Knowledge Maps
9. Is the rationale and selected approach for the pilot project described in Appendix J clear? If not,
please indicate specific points that could be clarified.
Appendix J presents the results of a pilot project to visually convey information presented in CEA
case study documents more concisely.
10. As a whole, do the Transport, Transformation, and Fate Knowledge Maps presented in Appendix
J facilitate the identification of knowledge gaps for MWCNTs? If not, please provide specific
revisions to the approach that could improve the identification of knowledge gaps for MWCNTs.
11. Do each of the maps (i.e., Physicochemical Properties, Transport, and Transformation Maps for
decaBDE and MWCNT) in Appendix J accurately and clearly present information conveyed in
Chapter 3 and related appendices on Transport, Transformation, and Fate? If not, please list
specific revisions to improve the accuracy or clarity of the maps.
K-2
-------
Table K-1. Expert affiliations and area of expertise.
Expert ID
Expertise Areas
Sector Affiliation
E24-PR
Transport, Transformation, and Fate of MWCNTs
Ecological Effects of MWCNTs
Carbon Nanotube Expertise
Academia
E25-PR
Life Cycle Stages for MWCNTs
Exposure, Dose, and Translocation of Nanomaterials
Carbon Nanotube Expertise
Independent Consulting
E26-PR
Life Cycle Stages for MWCNTs
Carbon Nanotube Expertise
Nongovernmental Organization
E27-PR
Analytical Methods for Characterizing Nanomaterials Government
Exposure, Dose, and Translocation of Nanomaterials
Carbon Nanotube Expertise
DecaBDE Expertise
E28-PR
Analytical Methods for Characterizing Nanomaterials Academia
Life Cycle Stages for MWCNTs
Transport, Transformation, and Fate of MWCNTs
Ecological Effects of MWCNTs
Carbon Nanotube Expertise
Note: Expertise areas are as defined by the EPA contractor conducting the independent letter peer review. Sector affiliation is based
on information that the reviewers submitted to the EPA contractor.
K.1.1. Appendix Development Process
The procedure used to develop Appendix I also was used to develop this appendix (see Section
1.1.1). Briefly, all responses to charge questions and specific observations on the Peer Review Draft were
tracked with the aid of an Excel-based comment tracking sheet. General observations were included in the
comment tracking sheet when they included distinct or more detailed information than that which
reviewers included in charge question responses. All comments were grouped by themes that developed
iteratively during the process of reviewing comments and developing revisions. Agency responses were
then drafted for each comment using consistent language to connect, where possible, comments from
multiple commenters that expressed the same basic ideas and themes. Responses were focused on
applying one of the three main categories of action taken (see Table K-2). Two additional response
categories were used to denote when action was not taken either because reviewers did not request any
action (e.g., comments expressing approval of the document) or because the Agency determined that
K-3
-------
action was not appropriate or feasible. In some cases, more than one response category was appropriate
for the same comment or group of comments. The "Agency Response" includes references to multiple
categories (e.g., document organization and improve clarity when text was moved and added) as
necessary to describe the complete actions that were taken in response to each comment or group of
comments. In instances when reviewers suggested revisions that the Agency did not pursue, a rationale
for not taking action is included in the Agency Response.
Table K-2. Agency response categories.
Response
Category Purpose
Actions Taken
Notes
1
Improve
document
nrnanbatinn
Increased numbering
and indexing
Subsections that were
are now numbered.
not
included
in the
Table
of Contents
Moved text to improve
clarity or overall
document flow
Text was moved to clarify key concepts or figures, primarily
in Chapter 1 and Appendix J.
Improve accuracy Made text edits to
and clarity pre-existing text
Some specific revisions were made to text that appeared in
the Peer Review Draft version of the case study to clarify
and improve the accuracy of statements.
Improve scientific Suggested references References were added when:
completeness
were reviewed and
incorporated into the
final document
1. They were specific to MWCNTs in flame-retardant
textiles, and
2. They would add significant value to researchers
carrying out research in the priority research areas
identified by RTI workshop participants.
No revision
No action was taken
Some comments suggested revisions that either
contradicted other reviewer comments, were not feasible to
carry out in the current document, or were not appropriate
to carry out after careful review of the comment. In each
instance, the Agency provides an explanation for why no
action was taken in response to these comments.
No revision
requested
No action was taken
Many comments provided informative feedback that the
Agency appreciates, but did not require any edits or
changes to the body of the case study document.
K-4
-------
K.1.2. Appendix K Organization
This appendix contains four types of comment tables:
1. Actions Taken: Table K-3. Table K-4. Table K-5. Table K-6.Table K-7. Table K-8. Table
K-9, Table K-10. These tables include responses and summaries of comments that were
addressed in the body of the document or appendices based on the first three response types
described in Table K-2 (this includes some cases where the comment was acknowledged but
no major action was taken for reasons described in the Agency Response). Tables in this
group are organized by relevant chapters of the case study document. Comments are grouped
in tables based on where the revision was made in the document, and listed chronologically.
2. No Revision: Table K-ll. This table includes responses and summaries of comments that
were not specifically addressed through revisions or edits to the main body of the case study
because the Agency determined that no action was appropriate.
3. No Revisions Requested: Table K-12. This table includes responses and summaries of
comments that did not suggest action needed to be taken.
4. Look-up table: Table K-13. This table is a look-up table for the comment identification (ID)
numbers listed in the response tables. Excerpts of the original text that reviewers provided are
included, along with the commenter and comment ID numbers, and the theme or themes
applied to each comment. Note that commenter ID numbers correspond to those listed in
Table K-l, above.
As noted previously, comments were grouped according to similar themes, so several unique
comments are presented as relevant to a single Agency response, as indicated by the ID numbers in the far
right column of the response tables. The comments have been collectively summarized to provide readers
with a relatively quick overview of the common theme among the comments and to illustrate how the
groups of comments relate to the action taken by the Agency. As noted above, the original individual
comments and assigned themes are provided in the final table of this appendix (Table K-13).
K-5
-------
K.2. Responses to Comments
K.2.1. Addressed Comments
Table K-3. Revisions relevant to the general case study or multiple sections of the case study.
Location in Peer
Review Draft that
Reviewer
Referenced
Comment Theme
Response
Related
to Global
Response Comment
Category ID
Title
One reviewer suggested
changing the title of the
document to reflect the fact
that most information on
decaBDE is now in an
appendix.
The Agency appreciates the reviewer's feedback. The title was revised as
appropriate.
429
All
One reviewer stated
generally that the scientific
completeness of the case
study might be improved by
considering references
provided by the reviewer.
The Agency appreciates the reviewer's feedback regarding specific literature that
could be informative to include. These references were reviewed and
incorporated as appropriate into the final case study document.
322, 340,
341, 353
All
One reviewer suggested
improving numbering and
indexing throughout the
document.
The Agency appreciates the reviewer's feedback. Although the existing page
numbering system was maintained as an indexing system for both the electronic
and hard copy of the case study, the Agency revised the document to include
additional indexing of sub-sections. In addition, text in Section 1.1.3.3 was revised
to clarify the placement of textboxes throughout the document.
1,4
335, 343,
345, 369,
370, 379,
393
All
Four reviewers noted that the
document could be more
succinct or pointed out
specific instances of
redundancy.
The Agency appreciates the reviewers' feedback. In revising the Peer Review
draft, redundant or verbose passages were removed when possible; however,
because some readers will review only portions of the document, some repetition
of content is deemed necessary to convey key concepts.
2,4
349, 364,
365, 422,
457, 536
K-6
-------
Table K-3 (Continued): Revisions relevant to the general case study or multiple sections of the case study.
Location in Peer
Review Draft that
Reviewer
Referenced
Chapters 1 and 2
Chapter 3
(Table 3-1,
Pages 3-9 and
3-10)
General Chapter 3
Chapter 3
(Page 3-1, lines 6
and 7; Additional
Information
Highlight Box 6);
Appendix G
(Page G-7,
lines 12 and 23
and Page G-8,
line 7)
Chapters 2, 4, and
6
Comment Theme
Two reviewers provided
additional references or
pointed to citations already
included in the document to
improve text discussing
MWCNT production or use
volumes.
One reviewer suggested
revisions to improve the
accuracy of text describing
MWCNT behavior in the
environment.
Two reviewers suggested
additional literature to
consider on environmental
fate.
One reviewer pointed out an
inconsistency with how data
on MWCNT environmental
concentrations are discussed
in the text.
Several comments
suggested minor copy edits
in the main text.
Response
The Agency appreciates the reviewers' suggestions for additional information to
include in the case study related to MWCNT production volumes. No additional
information was identified specific to the production or use of MWCNTs as flame-
retardants in textiles; however, one of the suggested references contained
information on global and European production volumes of MWCNTs in general.
This information was added to Additional Information Highlight Box 2 in Chapter 1.
and a cross-reference to this discussion is now included in Section 2.2.2. 1 .
A footnote was added to Chapter 2 (see Footnote 16) to clarify the use of the
terms solubility and dispersibility. Similarly, a footnote (see Footnote 17) was
added to Chapter 3 to clarify the use of the term partitioninq.
The Agency appreciates the reviewers' suggestions. The Schwyzer et al. (2012)
reference was reviewed and incorporated into Section 6.3.2.3. A link to Section
6.3.2 was added to Additional Information Hiqhliqht Box 7 in Chapter 3. The
O'Carroll et al. (2013) and Zhang et al. (2012) references were added to
Additional Information Hiqhliqht Box 8. The Liu et al. (2009) reference was already
in the case study (Cited in Table D-2).
The availability of modeled, but not measured, data on MWCNT environmental
concentrations was clarified in Chapter 3 and Appendix G. A specific reference to
modeled values in Table 3-2 was added in Appendix G.
The Agency appreciates the feedback and revised the document as appropriate.
Related
to Global
Response Comment
Category ID
„ , 325, 376,
^' d 439
2 526, 528
3 330, 452
383, 384,
392
315, 316,
2 317,318,
319, 320
K-7
-------
Table K-3 (Continued): Revisions relevant to the general case study or multiple sections of the case study.
Location in Peer Related
Review Draft that to Global
Reviewer Response Comment
Referenced Comment Theme Response Category ID
Additional Two reviewers noted These instances were reviewed and research gaps were clarified as appropriate
Information instances where knowledge in Additional Information Highlight Box 17 and Chapter 6.
Highlight Box 17; gaps could be more clearly „ ,..-.,- .__
Chapter 6 stated in Chapter 6 or an
Additional Information
Highlight Box.
K-8
-------
Table K-4. Revisions relevant to Chapter 1 (including Preface and Executive Summary).
Location in Peer
Review Draft that
Reviewer
Referenced
Comment Theme
Response
Related
to Global
Response Comment
Category ID
All
One reviewer suggested
clarifying the document
structure in terms of how
appendices and text boxes are
used throughout the document.
Section 1.1.3 was revised to convey more clearly where information is located
throughout the document, including (1) what information is contained in
appendices, (2) why this information was deemed appropriate for inclusion in
appendices rather than the main text, and (3) the different types of text boxes.
Any future efforts to develop CEA case studies will strive to improve how
information in the document is organized.
339, 344,
346, 347
General Chapter 1
One reviewer suggested
several review articles that
could be incorporated into the
document to add to discussion
on the environmental fate and
characterization of engineered
nanomaterials. The reviewer
suggested these articles in
reference to another comment
on clarifying the discussion
about environmental fate in the
comparison of decaBDE and
MWCNTs.
The Agency appreciates the reviewer's recommendations for additional
literature to consider. These references were reviewed in developing the final
case study; as review articles, however, they contain very little information
specific to MWCNTs, and topics covered in the references are discussed in
several places in the case study (e.g., Text Box 1-1 and Text Box 4-1, Figure
2-2 and Figure 3-1, Appendix B). Therefore, these references were not
incorporated into the final case study, but Chapter 1 was restructured to convey
more clearly how information on decaBDE could inform considerations related
to environmental fate and characterization of MWCNTs in Section 1.3.
323
General Chapter 1
One reviewer suggested
several articles that could be
incorporated into the document
related to the use of MWCNTs
in flame-retardants.
The Agency appreciates the reviewer's suggestion of additional literature to
incorporate into the case study. These references were reviewed, and Yu et al.
(2009), Yu et al. (2011), Lee et al. (2010), Verdejo et al. 2008), Im et al. (2011),
and Ullah and Ahmad (2012) were added to Additional Information Highlight
Box 2 and Additional Information Highlight Box 3 because they were most
relevant to the use of MWCNTs in flame-retardant textiles.
324
Section 1.1 One reviewer suggested
shortening Section 1.1 by
moving information to an
appendix.
The Agency appreciates the reviewer's feedback. Text in Section 1.1 was
moved to later sections of Chapter 1 to clarify the information and focus on key
concepts in this section.
504
K-9
-------
Table K-4 (Continued): Revisions relevant to Chapter 1 (including Preface and Executive Summary).
Location in Peer
Review Draft that
Reviewer
Referenced
Comment Theme
Response
Related
to Global
Response Comment
Category ID
Page 1-11 One reviewer requested
lines 15-17 clarification of the instances in
which the most commonly
agreed-on Importance and
Confidence ratings do not align
with the part of the matrix with
the largest number of
stakeholders due to a
difference in how individuals
combined
Importance/Confidence ratings
compared to the overall rating
combination of all stakeholders.
The Agency appreciates the reviewer's feedback; although this statement is
intended to emphasize how priorities were identified, rather than to highlight
specific instances. A list of the three instances was added to the document as a
footnote
(1) Env. T.T.F: Wastewater- Bioavailability,
(2) Env. T.T.F.: Sediment - Mobility, and
(3) Dose (Kinetics: Human - Distribution) (see Footnote 7).
423
Figure 1-3 Two reviewers commented on
Figure 1-3. One noted that the
upper left was not clear, while
the other noted that the figure
as a whole provided a very
good overview of the collective
judgment results.
The Agency appreciates the reviewers' feedback. The upper left of Figure 1-3 is
small so that the graphic will on the page but its intent is only to convey how the
detailed framework on the right of the figure appeared before the collective
judgment process. Nevertheless, the size of the figure was increased as much
as possible to improve the clarity of the top left. In future CEA case studies,
attempts will be made to improve how results of the collective judgment
process are conveyed.
2, 5
371, 397
Section 1.1.4 Two reviewers suggested
revising Chapter 1 to clarify the
selection, and utility, of
comparing decaBDE with
MWCNTs.
The Agency appreciates the reviewers' feedback. Chapter 1 is now structured
to convey more clearly why decaBDE was selected as a comparison for
MWCNT flame-retardants (see Section 1.1.3.1) and why this comparison can
be informative despite differences in physicochemical properties between the
two materials (see Section 1.3).
1,2
332, 333,
354, 372,
430
Table 1-8, Table
1-9, Text Boxes;
Page 1-14,
line 16; Additional
Information
Highlight Box 1
Two reviewers suggested
revisions to improve scientific
clarity in Chapter 1 by adding
citations or discussing the
selection of citations.
The Agency appreciates the reviewers' feedback and took the following actions:
(1) Section 1.1.2 now includes a description of the method used to identify and
select references for inclusion in the document, including the use of gray or
non-peer-reviewed sources; (2) A cross-reference was added to Section 1.1.3.1
to direct readers to the citations (in Section 1.3.2) in support of the statement
that commercial MWCNT flame-retardant products are available; (3) The use of
review articles as references in tables and text boxes was evaluated and
maintained because they provide accurate references for readers. In revising
the Peer Review Document to a final case study, efforts were made to ensure
that text boxes included references in a consistent manner. In addition, a
sentence was added to Section 1.1.3 to explain that text boxes contain
references for specific values or studies, but not for general statements that are
discussed further in the main text.
355, 373,
436, 437,
438
K-10
-------
Table K-4 (Continued): Revisions relevant to Chapter 1 (including Preface and Executive Summary).
Location in Peer
Review Draft that
Reviewer
Referenced
Comment Theme
Response
Related
to Global
Response Comment
Category ID
Section 1.1.3 One reviewer noted that the
number of experts in the RTI
workshop is too small for
statistical evaluation and the
expertise areas of the
participants are not discussed.
The Agency appreciates the reviewer's feedback; greater detail on the
expertise qualifications of RTI workshop participants is available in the RTI
Workshop Summary Report (RTI International, 2012). To avoid redundancy with
this resource, detailed information on RTI participants' expertise was not added
to the report; however, the general description of expertise areas was
expanded and a reference to Table 1-1. which lists the expertise and sector
affiliation of the participants involved in reviewing the External Review Draft, is
now included in Section 1.1.3. The number of workshop participants is also
addressed in the RTI report. We agree with the reviewer's point that the number
of participants was not large enough for statistical evaluation and thus have not
carried out statistics on the collective judgment outcomes.
337
Section 6.3 Two reviewers suggested that
the rationales RTI workshop
participants provided for priority
or unprioritized areas could be
made more prominent in the
document.
The Agency appreciates the reviewers' feedback and has added a statement to
Section 1.1.3 referring readers to Chapter 6 for more details on the identified
research priorities, including participants' rationales. Although no information
specific to rationales for unprioritized areas is included in the document,
readers can refer to the RTI Workshop Summary Report for additional detail on
any rationale that workshop participants provided. For future CEA case studies,
efforts will be made to improve how information related to collective judgment
outcomes is conveyed in the document.
338, 450
K-11
-------
Table K-5. Revisions relevant to Chapter 2.
Location in Peer
Review Draft
that Reviewer
Referenced
Comment Theme
Response
Related
to Global
Response Comment
Category ID
General
Chapter 2
One reviewer suggested
revisions and an additional
reference to clarify the
discussion on MWCNT
incineration.
The Agency reviewed the suggested references and the existing text on
incineration of MWCNTs in Section 2.5.2.2. The text was revised to clarify the
information and the suggested references were incorporated, as appropriate.
407
Tables comparing
MWCNT to
DecaBDE
One reviewer suggested
adding a qualifier for tables
comparing decaBDE and
MWCNTs, or removing them
completely.
The Agency appreciates the reviewer's feedback. These tables were reviewed
in developing the final case study document but were retained in the
document to provide readers with a basis for understanding potential MWCNT
release from textile matrices in the context of conventional materials like
decaBDE. New text was added to Chapter 2 to qualify these tables as the
reviewer suggested.
525
Figures 2-1 and
2-2
Two reviewers suggested
revisions to improve the
accuracy or clarity of figures in
Chapter 2.
The Agency appreciates the reviewers' suggestions. These figures were
reviewed and Figure 2-1 was revised to remove "reuse and recycling" in
reference to decaBDE and MWCNTs. Figure 2-2 was revised to show more
clearly the potential for direct release into the environment during material
synthesis and processing and product manufacturing. The figure legend for
Figure 2-2 also was revised to represent this point more clearly.
378, 425
Page 2-14,
lines 5-6;
Page 2-22,
lines 1-3;
Sections 2.2 and
2.3; Figure 2-1
Three reviewers suggested
revisions to improve the
scientific clarity or accuracy of
text discussing MWCNT
release from polymers and
factors considered in a
lifecycle analysis (LCA) of
MWCNTs.
The Agency appreciates the reviewers' feedback. Text was revised in Section
2.1 to include more discussion of the factors important in LCA (e.g.,
feedstocks, co-products, by-products). Text throughout Chapter 2 also was
revised to clarify the discussions of potential by-product release during
MWCNT purification. Finally, text in Section 2.4.1 was revised to clarify factors
influencing environmental releases during the product use stage.
380, 381,
400, 523
K-12
-------
Table K-5 (Continued): Revisions relevant to Chapter 2.
Location in Peer
Review Draft
that Reviewer
Referenced
Comment Theme
Response
Related
to Global
Response Comment
Category ID
Sections 2.4 and
2.5
One reviewer suggested
additional literature to consider
on release of MWCNTs from
polymers.
Two of the suggested references (Wohelleben et al. 2011, 2013) were
incorporated into Additional Information Highlight Box 4. which discusses how
data on MWCNT release from other product matrices could inform
understanding of MWCNT release from textiles. The remaining suggested
references were not incorporated because they were not specific to MWCNTs
or were not publically available.
451
Additional
Information
Highlight Box 5
One reviewer stated that the
text discussing CNTs in
wastewater was speculative.
This text was reviewed and revised as appropriate to convey information on
CNTs in wastewater clearly and accurately.
524
Table K-6. Revisions relevant to Chapter 3.
Location in
Peer Review
Draft that
Reviewer
Referenced
Comment Theme
Response
Related
to Global
Response Comment
Category ID
Figure 3-1 One reviewer suggested a
revision to improve the scientific
accuracy of Figure 3-1.
The figure was revised with a footnote to include the information the
reviewer suggested.
527
K-13
-------
Table K-7. Revisions relevant to Chapter 4.
Location in Peer
Review Draft
that Reviewer
Referenced
Page 4-1, line 17;
Page 4-2
Comment Theme
One reviewer suggested
explaining "abiotic receptors"
more clearly.
Footnote 19 was added to the
"abiotic receptors."
Response
Introduction of Chapter 4 to clarify the term
Response
Category
2
Related
to Global
Comment
ID
473
Text Box 4.1 One reviewer suggested adding a
review article on in vivo
quantification of CNTs to Text Box
4-1.
An overview of in vivo quantification methods, focusing on isotopic labeling
and tracing methods has been added to Text Box 4-1, with a brief discussion
of the advantages, applicable situations, and limits of the methods.
474
Page 4-10,
lines 10-12;
Additional
Information
Box 12
Two reviewers suggested minor
revisions to clarify references in
Chapter 4.
The text was reviewed and revised to convey more clearly information from
each reference.
386, 532
Additional
Information
Highlight Box 11
One reviewer suggested an
additional reference related to in
vitro data on dermal absorption of
MWCNTs and MWCNT inhalation
exposure.
The Agency appreciates the reviewer's suggestion. The reference was
reviewed and incorporated into Additional Information Highlight Box 11.
476
Table 4-1
One reviewer suggested revisions
to improve the accuracy of Table
4-1.
Table 4-1 has been revised to reflect the correct occupational exposure limit
(OEL) value of 30 mg/m3 and the reference has been changed to Nakanishi,
J., Ed. (2011) as cited in Morimoto et al. (2012). Also, the table now indicates
that the OEL applies to both SWCNTs and MWCNTs.
475
K-14
-------
Table K-8. Revisions relevant to Chapter 5.
Location in
Peer Review
Draft that
Reviewer
Referenced
Comment Theme
Response
Related
to Global
Response Comment
Category ID
Page 4-20, One reviewer suggested an
lines 20-22 additional reference related to
MWCNT inhalation exposure.
The Agency appreciates the reviewer's suggestion. The reference was
reviewed and incorporated into Additional Information Highlight Box 13 and
Additional Information Highlight Box 16. as well as Section 5.1.3.
477
General One reviewer suggested adding to
Chapter 5 the description of one study on
environmental impacts of CNTs.
The Agency reviewed the study highlighted by the reviewer and clarified the
discussion in the case study accordingly.
404
Section 5.1 One reviewer suggested revising
the description of subchronic and
chronic data from rodent studies.
The reviewer provided additional
studies to consider on this topic
and noted available data from a
study already included in the case
study.
The Agency appreciates the reviewer's feedback. The text in Section 5.1 was
revised to reflect the data related to subchronic systemic observations
accurately.
442, 482
Page 5-7, One reviewer noted that a
lines 7-9 reference on SWCNT
inflammatory effects should be
added to a statement about
inflammatory effects of MWCNTs
and SWCNTs.
The Agency appreciates the reviewer's suggestion. This study was not added
to the case study because information specific to MWCNTs is available on this
topic; the text was revised, however, to reflect the specific focus on MWCNTs
in the case study document.
481
Table 5-1 Two reviewers suggested
revisions to improve the accuracy
or utility of Table 5-1. One of the
reviewers included a new
reference that could be used to fill
in data gaps identified in Table 5-1
and corresponding text.
Table 5-1 was reviewed for accuracy and completeness. Corrections were
made to the table to improve its accuracy and to make it more consistent with
the text corresponding to Table 5-1. Footnotes were added to the table to refer
readers to Appendix F.1.2 for more information about specific studies and to
Section 6.3.4.1 for more information about the newly identified studies. In
addition, information from a reference one reviewer suggested was
incorporated into Section 6.3.4.1.
2, 3
409, 426,
439, 480
K-15
-------
Table K-8 (Continued): Revisions relevant to Chapter 5.
Location in
Peer Review
Draft that
Reviewer
Referenced
Comment Theme
Response
Related
to Global
Response Comment
Category ID
Page 5-12
DecaBDE
Comparison Box
One reviewer questioned the utility
of drawing parallels between
reproductive effects of decaBDE
and MWCNTs.
The Agency appreciates the reviewer's feedback. Text in the DecaBDE
Comparison Box on reproductive and developmental effects was revised to
clarify how information on DecaBDE might inform research planning for future
MWCNT risk assessment.
533
Page 5-7; One reviewer suggested
Page 5-17 improving the clarity of the
discussion on MWCNTs and
asbestos-like behavior.
The text was reviewed for clarity and accuracy. The suggested references
were incorporated, as appropriate, into Additional Information Highlight Box 13.
2, 3
483, 486,
487, 488
Page 5-9
lines 1-14;
Page 5-13
line 17
One reviewer suggested minor
revisions to improve the clarity of
text in Chapters.
The Agency appreciates the reviewer's suggestions. The text was revised to
improve clarity. In one instance, instead of removing the text as the reviewer
suggested, the description of the derivation of the reference value was
improved.
484, 485,
540
Page 5-14 One reviewer suggested clarifying
the use of in vitro data in
regulatory risk assessment and
suggested additional references to
consider on alternative methods
for nanomaterial toxicity testing.
The introduction to Section 5.1.9 on in vitro studies used to evaluate human
health effects was revised to reflect the growing use of in vitro studies in
regulatory risk assessment and acknowledge current work to incorporate in
vitro studies in evaluations of nanomaterials. The recommended references
were reviewed and incorporated as appropriate into Section 5.1.9 or Additional
Information Highlight Box 15 as part of the discussion on Next Generation style
models for MWCNT toxicity testing.
2, 3
443, 444,
445, 446
Page 5-18 One reviewer suggested clarifying
lines 2-3 the characterization of literature
on ecotoxicity of MWCNTs in
aquatic and terrestrial
ecosystems. The reviewer also
provided two additional citations to
consider.
The Agency appreciates the reviewer's feedback. The characterization of
available literature was revised as appropriate in Section 5.2 on Ecological
Effects. The studies on effects in terrestrial organisms are listed in this
appendix as a reference for readers, but were not incorporated into the
document because RTI workshop participants identified "Impacts: Terrestrial
Biota" as a lower research priority.
388
Page 5-17 One reviewer suggested
line 20 correction of a statement that
intratracheal instillation studies are
qualitatively similar to inhalation
studies and provided references to
support this correction.
The Agency appreciates the reviewer's suggestion. The specified text was
corrected and moved to Section 5.1.3 where this topic is first discussed. In
addition, the suggested references were incorporated with the revised text.
489
K-16
-------
Table K-8 (Continued): Revisions relevant to Chapter 5.
Location in
Peer Review
Draft that
Reviewer
Referenced
Comment Theme
Response
Response
Category
Related
to Global
Comment
ID
Page 6-34 One reviewer noted that
information in Chapter 6 might be
better suited for incorporation in
Chapter 5.
Chapter6 contains references that were not included in previous drafts of the
document either because they were published after the document was made
available, or they were not identified in literature searches carried out during
document development. Chapter 6 expands on the previous chapters by
presenting additional information that could be useful to researchers in
planning research in priority areas identified by RTI workshop participants.
Although this organization was maintained in the final case study, a cross-
reference was added to Chapter 5 to remind readers that additional
information is available in Chapter 6.
491
K-17
-------
Table K-9. Revisions relevant to Chapter 6.
Location in Peer
Review Draft
that Reviewer
Referenced
Comment Theme
Response
Related to
Global
Response Comment
Category ID
All
One reviewer suggested additional
literature on a recent assessment of
CNTs to consider incorporating in
the document.
The Agency appreciates the reviewer's suggestion. New text discussing this
reference is now included in Section 6.3.4.1.
441
General One reviewer suggested several
Chapter 2 review articles related to MWCNT
synthesis.
The Agency appreciates the reviewer's suggestions. These references were
reviewed and Huang et al. (2012) was incorporated in Section 6.3.1.1. The
other references were not incorporated because they (1) were not specific to
MWCNTs, (2) contained information already in the case study, or (3) discussed
one synthesis method in depth, which is not a goal of the case study document.
326
General Three reviewers commented on the
Chapter 5 climate change discussion in
Chapters. One reviewer suggested
removing the section due to
insufficient science, while another
suggested revising to state that
MWCNT production and climate
change clearly are connected. A
third reviewer expressed
appreciation for the inclusion of
climate change and other impacts
(e.g., energy, economics).
Although data are not clear regarding whether or to what extent MWCNTs will
impact climate change, RTI workshop participants identified "Other Impacts" as
a research priority; thus, the information is included in the case study
document. The lack of available data suggests that this topic should, in fact, be
highlighted as an area of needed research. To further this discussion, a new
reference was added to Section 6.3.4.4 that discusses potential climate change
impacts from engineered nanomaterials, such as MWCNTs.
389, 537,
538
General Two reviewers suggested additional
Chapter 5 literature related to MWCNT
bioaccumulation and impacts in
aquatic receptors.
The Agency appreciates the reviewers' suggestions. The Zhu et al. (2006)
reference was added to Chapter 6. In addition, text in Section 6.3.4.3 (Aquatic
Biota) was revised to reference one of the suggested papers (Peterson et al.
2011) accurately, which was already discussed in the text. The remaining
references were reviewed and were not incorporated because (1) they were not
specific to MWCNTs, (2) they stated similar findings to references already in the
case study, or (3) they focus on an area that RTI workshop participants did not
identify as a research priority (bioaccumulation in biota).
2, 3,4
331,453,
454
K-18
-------
Table K-9 (Continued): Revisions relevant to Chapter 6.
Location in Peer
Review Draft
that Reviewer
Referenced
Comment Theme
Response
Related to
Global
Response Comment
Category ID
Page 6-35 One reviewer pointed out an
line 18 instance in Chapter 6 where the
science was unclear or inaccurate.
Text in Chapter 6 was revised to improve the accuracy of this point.
492
Table K-10. Revisions relevant to Appendices.
Location in Peer
Review Draft that
Reviewer
Referenced
Comment Theme
Response
Related to
Global
Response Comment
Category ID
Appendix F,
Page F-13 through
F-17
One reviewer suggested a
revision to improve the scientific
accuracy of Appendix F.
The Agency appreciates the reviewer's suggestion and replaced "inhalation"
with "pulmonary exposure" when describing intratracheal instillation and
intranasal injection studies in Appendix F.
490
Appendices G and
H
One reviewer suggested making
the structure of Appendix G and
Appendix H more similar.
The Agency appreciates the reviewer's feedback. As its title indicates,
Appendix G contains information on MWCNTs in unprioritized areas of the
CEA framework. As such, the structure does not correspond to the structure of
Appendix H, which contains information on decaBDE for all areas of the CEA
framework. Although no change was made to the structure of Appendix G, text
in Section GJ_was revised to clarify that information on MWCNTs in prioritized
areas of the CEA framework is included in the main text of the document.
334
Appendix I,
Table I-3
One reviewer suggested
improving the clarity of Appendix I
by moving a table.
Table I-3 was moved as the reviewer suggested.
478
K-19
-------
Table K-10 (Continued): Revisions relevant to Appendices.
Location in Peer
Review Draft that
Reviewer
Referenced
Comment Theme
Response
Related to
Global
Response Comment
Category ID
Appendix J
Two reviewers suggested
revisions to the structure of
Appendix J to clarify the
objectives of Knowledge Maps
(KMs) or the maps themselves.
The Agency appreciates the reviewer's feedback; text describing each figure
(Figures J-1 to J-6) is now located more closely to the appropriate figure. In
addition, a new final section in the appendix now discusses the extent to which
the project outcomes meet the identified objectives.
1,2
358, 362,
415
Appendix J
One reviewer stated that the KMs
are useful for identifying
knowledge gaps provided they
use up-to-date data.
The Agency appreciates the reviewers' feedback. In finalizing the case study
document, the maps were reviewed to ensure they incorporate the most
current information included in the case study (see Chapter 1 for details on
how literature was selected for the document).
359
Appendix J
Three reviewers suggested
revising the KMs to clarify the
data available.
The KMs now use solid lines to denote "data available or modeled" and
include solid lines around compartments where there is a modeled
concentration available for MWCNTs. In addition new text was added to the
end of the appendix on the extent to which weight of evidence can be
addressed using the KM approach.
360, 416,
470
Appendix J,
Figures J-1 and J-2
Three reviewers suggested
revisions to improve the MWCNT
Physicochemical Properties KM.
The Agency appreciates the reviewers' feedback and revised the map where
appropriate. In addition, a footnote is now included in Appendix J to refer
readers to Chapter 2 regarding the use of the term "solubility" in the document.
417,418,
420, 465,
466, 517
Appendix J,
Page J-2, line 18;
Figures J-1, J-5, and
J-6
Two reviewers suggested minor
revisions to improve the accuracy
of Appendix J.
The Agency appreciates the feedback and revised the appendix as
appropriate.
419, 535
Appendix J, Two reviewers suggested
Figure J-6 revisions to clarify Figure J-6.
Figure J-6 was revised to clarify the different types of MWCNTs depicted and a
reference was added to Footnote 13 in Chapter 2 for an explanation of the
term bundled.
421,468
K-20
-------
Table K-11. Comments that required no action.
Location in Peer
Review Draft that
Reviewer
Referenced
Comment Theme
Response
Related
to Global
Comment
ID
All
One reviewer noted that MWCNTs cannot
be used alone as flame retardants in
textiles and suggested including a
representative compound that is used in
combination with CNTs as part of the
comparison in the case study.
The Agency appreciates the reviewer's feedback. Given the uncertainty
surrounding which flame retardants could be used in conjunction with MWCNTs
in textiles or other applications, a particular compound was not selected to
incorporate into the comparison throughout the document. Additional
Information Highlight Box 3. however, discusses the potential use of other flame
retardants in conjunction with MWCNTs. Efforts also were made to incorporate
additional references relevant to this discussion based on suggestions from
other reviewer comments (see Table K-4. comment #324).
431, 539
Page 4-1, line 17;
Page 4-2
One reviewer suggested moving the
"Unprioritized Research Area" text box for
"abiotic receptors" to an appendix.
The Agency appreciates the reviewer's feedback; however, the text box
depicting how RTI workshop participants rated Exposure and Dose (Kinetics) in
abiotic receptors is included in the main text to convey how this area was
identified as an unprioritized research area. The inclusion criteria for this text
box are the same as the inclusion criteria for all text boxes depicting how RTI
workshop participants rated unprioritized research areas.
479
All
One reviewer suggested incorporating
information from a risk assessment of
CNTs completed in Japan.
The Agency appreciates the suggestion; however, this report was not
incorporated because it presents information already in the case study and
would not add significant value to MWCNT research planning. For instance, the
OEL developed in this study is listed in Table 4-1 as it is included in another
reference suggested by the reviewer.
440
All
One reviewer stated that the document
often does not mention areas for which no
information is available but which would be
necessary to complete an assessment.
The reviewer pointed to stability and
transformation in water and sediment as an
example.
The Agency appreciates the reviewer's feedback. Information relevant to
MWCNT stability and transformation in surface water is discussed in Section
3.3.1 and 6.3.2.2. Although RTI workshop participants did not identify transport,
transformation, and fate of MWCNTs in ground water as a research priority,
information on this topic is discussed in Appendix G so that the topic can be
evaluated further as lower priority research gaps are addressed. The types of
information necessary to complete an assessment are discussed in the context
of problem formulation in Section 6.2.
401
All
One reviewer suggested adding
summaries to the end of each chapter.
The Agency appreciates the reviewer's feedback. To avoid increasing the
document length and adding redundancy, no additional summaries were added;
however, the Executive Summary does provide an overview of information
presented in each chapter.
395
K-21
-------
Table K-11 (Continued): Comments that required no action.
Location in Peer
Review Draft that
Reviewer
Referenced
Comment Theme
Response
Related
to Global
Comment
ID
All
Three reviewers commented on the use of
cross-references in the document. One
reviewer suggested removing extraneous
cross-references from the main text, while
another suggested reducing cross-
references in appendices. A third reviewer
stated that the document was well-
structured and the use of cross-references
was appreciated.
The Agency appreciates the reviewers' feedback. The frequent use of cross-
references is intended to reduce redundancy and to inform readers who might
not read the entire case study from beginning to end; thus, cross-references
were maintained in the final document.
350, 394,
463
All
Two reviewers suggested expanding the
scope of the case study to include
SWCNTs, or suggested literature on
SWCNTs to consider adding to the case
study document.
The Agency appreciates the reviewers' feedback and will consider broadening
the scope of any future CEA case study documents to include other similar
materials. For the current case study, SWCNT information was only included in
areas for which no information on MWCNTs was identified in an effort to convey
clearly where data gaps exist specific to MWCNTs.
327, 495,
499, 502
Text Boxes; Two reviewers commented on the
Information Additional Information Highlight Boxes.
Highlight Boxes One did not fully agree with the use of
these boxes in the document. The other
reviewer stated that these text boxes work
well to convey additional information and
are generally relevant.
The Agency appreciates the reviewers' feedback regarding the utility of the
"Additional Information Highlight Boxes." These boxes were added to draw the
readers' attention to (1) how the CEA process can improve scientific quality of
the case study document, and (2) to call attention to new information related to
the priority research areas that the RTI workshop participants identified. In
developing any future CEA case studies, we will consider this input on the utility
of the "Additional Information Highlight Boxes."
399, 511,
529, 530,
534
Chapters 1 and 6
Two reviewers commented on the
description of CEA. One reviewer noted
that the approach is explained too much,
while another noted that the thorough
description provides a good introduction to
the document. The second reviewer also
suggested moving Sections 6.1 and 6.2
into Chapter 1.
The Agency appreciates the reviewers' feedback. The CEA approach is
described in greater depth than more well-established approaches (e.g., risk
assessment) because many readers might not be familiar with the components
of the approach and how they work. Future efforts to develop CEA case studies
will strive to refer readers to other resources for detailed explanations of the
CEA approach. The content in Sections 6.1 and 6.2 was maintained at the
beginning of Chapter 6 to provide context as readers move from information on
each area of the CEA framework in Chapters
2-5 to the translation of identified research priorities and to research planning.
In addition, moving this text would make Chapter 1 quite lengthy and would not
provide necessary context to readers who focus their review of the document
on Chapter 6.
366, 396,
410
K-22
-------
Table K-11 (Continued): Comments that required no action.
Location in Peer
Review Draft that
Reviewer
Referenced
Comment Theme
Response
Related
to Global
Comment
ID
Preface, One reviewer suggested using the
Page xxi, line 4 European definition of nanomaterials in the
case study document.
The Agency appreciates the reviewer's suggestion; however, the size range
used in the case study from the National Nanotechnology Initiative (NNI)
definition aligns with the European definition of the primary particle size range
for nanomaterials. Although the NNI and European definitions differ in terms of
whether other aspects (i.e., unique properties determined by size) should be
incorporated into the definition, the sole use of the NNI definition is maintained
in the case study to avoid confusing readers.
363
General Chapter 1
One reviewer suggested that information
on flame-retardants in Chapter 1 could be
moved to an appendix.
The Agency appreciates the reviewer's feedback; however, many of the details
regarding MWCNT flame-retardant use were added in response to comments
on the External Review Draft of the case study document. Thus, these details
have been retained in Chapter 1.
398
Chapter 1,
Table 1-5
Two reviewers suggested incorporating
additional information on MWCNT flame-
retardants into Table 1-5.
The Agency appreciates the reviewers' suggestions; however, information on
the incorporation of MWCNTs into flame-retardant textiles is included in Table
1-10. Table 1-5 focuses on incorporation methods for historically used flame
retardants. MWCNT flame retardants are not included in Table 1-5 because
they are not yet common in this application, although, as discussed in Section
1.3.3 and Additional Information Highlight Box 2. their use might increase in the
future.
424, 518
Chapter 1,
Page 1-25, line 12
One reviewer questioned the inclusion of a
particular reference for text discussing risk
and benefit comparisons of nanomaterials.
The reference was reviewed along with the text it supports in the document.
The reference is included in this part of the document because it is an example
of a discussion on the potential risks and benefits of CNTs in a particular
application.
368
Chapter 1,
Page 1-26, line 1
One reviewer suggested a revision to make
a statement more definitive in the text.
The Agency appreciates the reviewer's suggestion; however, the terms used in
the text were specifically selected to convey the state of the science clearly and
accurately, and thus no change was made.
374
Chaper 1,
Figure 1-1
One reviewer suggested explicitly including
ground water in Figure 1-1.
The Agency appreciates the reviewer's feedback; however, the focus of this
version of the case study is on priorities identified by experts in the RTI
workshop. Because the experts did not identify ground water as a priority
research area, additional literature on this topic has not been incorporated into
the main text of the document. Additional information on ground water is
included in Appendix G and Appendix H. The references the reviewer
suggested are also included in the Full Comment Excerpt Table (Table K-13) in
this appendix as a resource for readers. The high-level depiction of the CEA
framework (Figure 1-1) does not specifically include the ground water
compartment but the more detailed depiction of the framework (Figure 1-3)
does.
367
K-23
-------
Table K-11 (Continued): Comments that required no action.
Location in Peer
Review Draft that
Reviewer
Referenced
Comment Theme
Response
Related
to Global
Comment
ID
Chapter 1 ,
Text Box 1-1
One reviewer suggested adding another
review article as a reference to Text Box 1-
1.
The Agency appreciates the suggestion; however, the addition of another
review article was not deemed necessary to support information in this text box.
The reference is included elsewhere in the case study as a resource for
readers.
375
Chapter 1,
Table 1-9
One reviewer suggested broadening the
range of MWCNT purity in Table 1-9.
The Agency appreciates the suggestion regarding MWCNT purity. The phrase
"usually >90%" is based on study reports that are summarized in Appendix F.
Table 1-9 does report a wider range (7.5 -40%) for some MWCNTs to bring the
greater variation in purity for some MWCNTs to the readers' attention.
519
Chapter 1,
Table 1-9
One reviewer stated that chirality should
not be used to describe MWCNTs.
The Agency appreciates the suggestion regarding MWCNT chirality; however,
this statement was reviewed in revising the External Review Draft and found to
be accurate. Because MWCNTs have multiple layers of graphene, the angle
between carbon atoms in each sheet can vary between sheets, confirming that
a range of chiral angles in any one tube are possible.
520
Chapter 1,
Table 1-9
One reviewer noted that peer-reviewed
literature should be used rather than
product data sheets (i.e., Sigma-Aldrich
2012 citation) for statements about
MWCNT parameters.
The Agency appreciates the reviewer's feedback; however, in most instances,
this reference is provided in conjunction with references from literature to
provide readers with another source of information on MWCNTs. In the two
instances where the reference is provided alone, other sources of this
information (melting point and density) were not located. Further, EPA has no
reason to believe that the values the manufacturer reported for these
parameters are not representative of what would be incorporated into flame-
retardant coatings applied to upholstery textiles (e.g., a commercially available
MWCNT).
521
General Chapter 2
Two reviewers suggested additional
literature to consider related to
nanomaterial release from polymers.
The Agency appreciates the reviewers' suggestions. The suggested references
were reviewed but not incorporated into the final case study because they either
(1) are not specific to the potential release of MWCNTs coatings in textiles, or
(2) contained information similar to references already included in the case
study. See Additional Information Highlight Box 4 for related discussion.
328, 382
General Chapter 2
One reviewer suggested two additional
references related to disposal of MWCNT.
The Agency appreciates the reviewers' suggestions. These references were
reviewed and not incorporated because one did not contain information specific
to MWCNTs, and the other provided only modeled data, not measured data.
329
Chapter 2,
Figure 2-1
One reviewer suggested including textiles
that do not have flame retardants
incorporated in Figure 2-1.
The Agency appreciates the reviewer's suggestion but including the use of
MWCNTs in textiles for reasons other than as a flame-retardant would broaden
the scope of the document. As discussed in Chapter 1, the focus on a particular
application is a key aspect of the case study approach.
377
K-24
-------
Table K-11 (Continued): Comments that required no action.
Location in Peer
Review Draft that
Reviewer
Referenced
Comment Theme
Response
Related
to Global
Comment
ID
Chapter 2,
Figure 2-1
One reviewer stated that available
information suggests that MWCNTs would
almost always be released as homo- or
hetero-aggregates. Thus, truly "free"
MWCNTs are not likely to be released.
The Agency appreciates the reviewer's feedback. The use of the term "free"
was reviewed and found to be appropriate given that it was consistently used in
conjunction with "bundled" or "matrix bound" to denote that a small fraction of
MWCNTs might be "free" while the remaining fraction is in the other release
forms. Although environmental transformations could lead to MWCNT
aggregation or agglomeration, existing data are insufficient to state that no free
MWCNTs will be released.
522
Chapters 3 and 6
One reviewer noted that while although the
identified priority research areas for human
health align with those identified by other
groups; others identified terrestrial biota as
of higher priority.
The Agency appreciates the reviewer's feedback on the alignment of the
research priorities that RTI workshop participants identified with European
Union efforts. Although experts in the RTI workshop did not identify terrestrial
receptors as a research priority area, information pertaining to this area is still
included in Appendix G of the document as a reference for interested readers.
449
General Chapter 4
One reviewer suggested separating human
and environmental exposure scenarios and
data in Chapter 4.
The Agency appreciates the reviewer's feedback; human exposure scenarios
(Section 4.2) are described separately from environmental concentrations
(Section 4.1) and ecological exposure scenarios (Section 4.2).
459
Chapter 4, One reviewer suggested an additional
Section 4.1.2.2 reference related to environmental
concentrations of carbon black.
The Agency appreciates the reviewer's suggestion but this reference was not
included in the final document because it is specific to carbon black, rather than
MWCNTs. Including these data could make the data gap of measured
concentrations of MWCNTs in sediment less clear for readers.
385
Chapter 4, Section
4.3 heading
One reviewer questioned the use of the
terms environmental and ecological
systems throughout the case study.
The Agency appreciates the reviewer's feedback. The use of these terms
throughout the case study document was reviewed and checked for
consistency. "Ecological" has been used in the document to refer to receptors
while "environment" is used more broadly.
387
Chapter 4,
Text Box 4-1;
Additional
Information
Highlight Box 10
One reviewer suggested adding a more
explicit description of the limitations of
available analytical techniques to detect
MWCNTs in environmental media.
The Agency appreciates the reviewer's suggestion but no change was deemed
necessary as a result of this comment. The difficulties associated with
accurately characterizing MWCNTs in complex environmental matrices are
extensively discussed in the case study (e.g., Section 4.1. Text Box 4-1.
Additional Information Highlight Box 10). No reference is offered for the specific
detection limit offered by the reviewer. Further, the actual detection limits will
vary greatly according to the media and CNT, and will change as better
techniques are developed.
531
Chapter 4, One reviewer suggested adding a
Section 4.2.1 summary table of occupational exposure
scenarios and available exposure levels to
Chapter 4.
The Agency appreciates the reviewer's suggestion; however, a summary table
was not added to avoid redundancy with information presented in bullet form in
Chapter 4.
448
K-25
-------
Table K-11 (Continued): Comments that required no action.
Location in Peer
Review Draft that
Reviewer
Referenced Comment Theme
Response
Related
to Global
Comment
ID
General Chapter 6;
DecaBDE
Comparison Boxes;
Priority Research
Area Boxes
One reviewer suggested adding a table or
figure to Chapter 6 that captured all of the
questions included in DecaBDE
Comparison text boxes.
The Agency appreciates the reviewer's feedback; however, the research
questions developed by RTI workshop participants (Table 6-4 through Table
6-14) are intended to provide research managers and others in the scientific
community with a clear understanding of the research priorities that emerged
from the CEA collective judgment step. Although the questions listed at the end
of decaBDE comparison boxes can provide additional input to researchers as
they develop experiments relevant to the identified research priorities, including
a table listing these additional questions could detract from the clear priorities
RTI participants identified.
342
General Chapter 6
One reviewer suggested adding
recommendations in the document on how
to implement research in the identified
priority areas.
The Agency appreciates the reviewer's feedback. As stated in Chapter 1. the
goal of the case study document is to identify research priorities that could
inform future assessment and subsequent risk management efforts. Making
recommendations on how to implement research in priority areas is thus outside
the scope of the document. Rather, the document provides a mechanism to
share these priorities within EPA and with others in the research community.
Individuals in these organizations can then identify how best to implement any
research in their organizations or through appropriate funding mechanisms. The
tables throughout Section 6.3 contain specific research objectives and
estimates of resources to support implementing the research, which could
inform efforts to implement research throughout the scientific community.
455
General Chapter 6
One reviewer stated that the case study
should emphasize the use of existing data
and alternative methods.
The Agency appreciates the reviewer's feedback. The use of alternative
methods for MWCNT testing is discussed in Additional Information Highlight
Box 15. In addition, RTI workshop participants identified developing data from
quantitative structure activity relationship models as research to inform the
identified priority areas (see Section 6.3). Throughout Section 6.3 research
gaps identified by RTI workshop participants are discussed in the context of
building off of existing data summarized in the document.
456
General Chapter 6
One reviewer suggested going beyond
research gap identification by adding a
chapter or section with a more explicit
discussion of risk assessment.
The Agency appreciates the reviewer's feedback; however, the purpose of the
draft case study document and its relationship to any future regulatory decisions
for MWCNT is discussed in Chapter 1 (see Section 1.3) and expanded on in
Chapter 6 (see Section 6.3).
356
Chapter 6,
Section 6.3
One reviewer suggested adding a
reference to Section 6.3 related to
terrestrial receptors.
The Agency appreciates the reviewer bringing this reference to our attention;
however, this particular reference was not added to the document because it
focuses on terrestrial receptors, which were not identified as a priority area by
experts in the RTI workshop.
303
K-26
-------
Table K-11 (Continued): Comments that required no action.
Location in Peer
Review Draft that
Reviewer
Referenced
Comment Theme
Response
Related
to Global
Comment
ID
Chapter 6,
Page 6-1,
lines 11-16
One reviewer suggested including
recommendations on research gaps
related to industrial nanomaterial
production and application volumes.
The Agency appreciates the reviewer's suggestion; however, MWCNT
production volumes are already noted as a research gap in Section 6.3.1.1 and
Section 6.3.1.3 discusses research gaps that RTI workshop participants
identified related to MWCNT product manufacturing.
390
Chapter 6,
Table 6-7
One reviewer suggested adding greater
detail on how RTI workshop participants
estimated time and financial resources for
carrying out the research needed to
address the identified research questions.
The Agency appreciates the reviewer's feedback; Footnote 27 states that
experts in the RTI workshop relied on their own knowledge of and experience
with conducting research to reach estimates of time and financial resources
needed to address the identified research questions.
391
General Appendices
One reviewer suggested condensing the
appendices and moving relevant text into
the main text.
The Agency appreciates the reviewer's feedback; however, information was
placed in the appendices, rather than the main text of the document if it was
informative for those interested in details relevant to a particular topic but not
critical for understanding the key points of the topic. The appendices are
deemed essential to keep the main text of this document as concise as
possible; thus, no changes have been made in response to this comment.
348
General Appendix B
Two reviewers suggested references that
could be included in Appendix B.
The Agency appreciates the reviewer's feedback; however, as described in
Chapter 4 and Appendix B, this appendix is not intended to provide a
comprehensive list of all available analytical techniques. In addition, the focus of
Appendix B is providing an overview of strengths and limitations associated with
commonly used analytical techniques for MWCNT. Techniques that are
becoming available, but not yet common, or used with other materials, are not
included in Appendix B.
302, 331
Appendix E,
Table E-6
One reviewer questioned whether data are
available for environmental compartments
for MWCNTs.
The Agency appreciates the reviewer's inquiry; however, no measurements of
MWCNTs in environmental compartments were identified in literature reviewed
in developing the draft case study. No additional studies with environmental
concentrations of MWCNTs were identified by reviewers of the draft document.
As such, no data were added to Table E-6: however, Table 3-2 contains data
from modeling studies on MWCNTs.
493
General Appendix J
Two reviewers suggested moving the
Knowledge Maps (KMs) currently in
Appendix J to the main text of the
document.
The Agency appreciates the reviewers' feedback. KMs have been maintained
as an appendix, however, because they represent a pilot project and were not
part of experts' consideration of potential research gaps (for revisions related to
these comments, see Table K-10. Comment IDs 358, 362, 415).
472, 505,
515
General Appendix J
One reviewer suggested adding greater
detail to the terrestrial compartment in the
KMs.
The Agency appreciates the reviewer's feedback; however, the lack of data on
MWCNT concentrations or behavior in different soil types suggests that this
revision would not be meaningful or feasible at this time. This level of detail
could be incorporated into future applications of KMs in CEA.
361
K-27
-------
Table K-11 (Continued): Comments that required no action.
Location in Peer
Review Draft that
Reviewer
Referenced Comment Theme
Response
Related
to Global
Comment
ID
General Appendix J One reviewer suggested making the KMs The Agency appreciates the reviewer's feedback but adding more quantitative
more quantitative. detail to the KMs is not feasible given the current gaps in data and differences ,-,.. ,-,.,,
in metrics used between available studies. This suggestion will be considered in
any future use of KMs in CEA.
Appendix J, One reviewer suggested that sediment The Agency appreciates the feedback. Information in the document on
Figure J-4 could be denoted as a "sink" for MWCNTs MWCNTs in sediment was reviewed; however, no change was made to Figure .„
in Figure J-4. J-4 due to a lack of evidence to support denoting this spatial zone as an
MWCNT sink at this time.
K-28
-------
Table K-12. Comments for which no action was requested.
Location in Peer
Review Draft that
Reviewer
Referenced
Comment Theme
Response
Related to
Global
Comment
ID
All
One reviewer noted that they are working on a project relevant The Agency appreciates the reviewer's feedback.
to this case study, "Development of safe and eco-friendly flame-
retardant materials based on CNT co-additives for commodity
polymers."
433
All
All
All
Text Boxes
All
Priority Research
Area Text Boxes;
General Chapter 6
Five reviewers stated that document accurately and clearly
conveys the science.
Two reviewers stated that the document presents the most
relevant research findings for future research to inform MWCNT
risk management.
Three reviewers stated that the document supports research
planning to inform future risk assessments of MWCNTs.
One reviewer found the addition of text boxes in the document
helpful.
One reviewer noted that the use of "Environmental" in CEA
might convey to some readers that human health is not
considered in this approach.
Five reviewers stated that the document clearly presents
research priorities for research planning.
The Agency appreciates
The Agency appreciates
The Agency appreciates
The Agency appreciates
The Agency appreciates
take this perspective into
work with CEA.
The Agency appreciates
the reviewers' feedback.
the reviewers' feedback.
the reviewers' feedback.
the reviewer's feedback.
the reviewer's feedback and will
account in considering future
the reviewers' feedback.
300, 306,
321,402,
403, 428,
434, 435,
458, 494
427, 498
304, 408,
500, 501
497, 503
432
301, 341,
406, 447,
496
K-29
-------
Table K-12 (Continued): Comments for which no action was requested.
Location in Peer
Review Draft that
Reviewer
Referenced
Priority Research
Area Text Boxes
General Appendix I
General Appendix I
General Appendix I
General Appendix J
General Appendix J
General Appendix J
Comment Theme
One reviewer suggested developing a peer-review manuscript
to discuss research prioritization outcomes from the RTI
workshop in more depth.
Five reviewers stated that peer and public stakeholder
comments are clearly presented in the document.
Five reviewers stated that EPA was responsive to comments
and made appropriate changes.
Three reviewers stated that EPA responses to peer and public
comments were clearly conveyed.
Four reviewers found the rationale and approach for KM
development clear.
Three reviewers stated that KMs were useful for identifying
knowledge gaps and clearly conveyed relevant information for
decaBDE and MWCNTs.
Three reviewers stated that the KMs are accurate and clear.
Response
The Agency appreciates the reviewer's feedback. Several
peer-review manuscripts are currently in production to
discuss the research priorities identified in the matrices.
The Agency appreciates the reviewers' feedback.
The Agency appreciates the reviewers' feedback.
The Agency appreciates the reviewers' feedback.
The Agency appreciates the reviewers' feedback.
The Agency appreciates the reviewers' feedback.
The Agency appreciates the reviewers' feedback.
Related to
Global
Comment
ID
336
307, 351,
411,460,
461, 506
309, 310,
311, 352,
357,413,
462, 507,
509, 510
308, 412,
508
312,414,
464, 512
313,469,
513
314, 360,
471
K-30
-------
K.3. Full Comment Excerpts
Table K-13. Comments received in response to the expert charge questions.
Comment
ID
300
301
302
303
304
Author Charge
ID Question
E24- 1
PR
E24- 2
PR
E24- 3
PR
E24- 3
PR
E24- 4
PR
Comment Text Excerpt
(Note: Comment excerpts were edited to correct spelling and major grammatical and format issues [in
order to improve clarity], but generally were maintained in their original [as-submitted] form)
To the best of my knowledge, I believe the science is accurately conveyed throughout the
document. No revisions are needed. The extensive review of the science and document that
preceded this current review has resulted in a well written and accurate version of the final
document.
Yes, the Priority Research Areas are clearly identified throughout the document. Putting the
Priority Research Areas in text boxes really highlights the priority areas. In addition, the
discussion of each Priority Research Area in Ch. 6 helps to further emphasize the areas and
provides an overall succinct summary of the research needs related to MWCNTs in flame-
retardants.
There are two additional studies that could be considered for inclusion to support the Priority
Research Areas. 1. Irin et al. (2012) is listed in Highlight Box 10 but not listed in the tables in
Appendix B. 1. Irin et al. (2012) is listed in Highlight Box 10 but not listed in the tables in
Appendix B.
a. Irin, F, B Shrestha, JE Cafias, M Saed, and MJ Green. 2012. Detection of carbon nanotubes
in plant roots through microwave-induced heating. Carbon 50:4441-4449.
There are two additional studies that could be considered for inclusion to support the Priority
Research Areas. 2. Li et al (2013) could be added to pg. 6-24, line 30. This paper
demonstrated no substantial uptake of MWCNTs in earthworms in soil. Li, S, F Irin, FO Atore,
MJ Green, and JE Canas-Carrell. 2013. Determination of carbon nanotube bioaccumulation in
earthworms measured by a microwave-based detection technique. Science of the Total
Environment, In press, doi: 10. 1016/j.scitotenv.2012. 12.037.
The document definitely supports research planning to inform future risk assessments. The
questions outlined at the end of the green text boxes (DecaBDE Can Inform MWCNT
Assessment) are very helpful to outline research needs. In addition, the Priority Research Area
text boxes (and Unprioritized Research Areas) also helped to outline research needs. Chapter
6 is also instrumental in providing further discussion to inform research planning.
Themes
Praise for case study,
Science is accurate &
clear, No revisions
needed
Praise for case study,
Research priorities are
clear, No revisions
needed
Consider additional
literature, Detection/
measurement
techniques
Consider additional
literature, Terrestrial
receptors
Praise for case study,
Supports research
planning, No revisions
needed
K-31
-------
Table K-13 (Continued): Comments received in response to the expert charge questions.
Comment
ID
Author Charge
ID Question
Comment Text Excerpt
(Note: Comment excerpts were edited to correct spelling and major grammatical and format issues [in
order to improve clarity], but generally were maintained in their original [as-submitted] form)
Themes
305
E24- 4
PR
With the exception of the Impacts Priority Research Area for Aquatic Biota (p 5-19), the other
text boxes clearly define research gaps and questions to ask. After reading Section 6.3.4.3,
particular research needs are more apparent. This particular text box could be improved to
include what is discussed in Section 6.3.4.3.
Praise for case study,
Document organization,
Aquatic receptors
306
E24- 5
PR
The organization of the document is excellent. It flows in a logical manner and is easy to
understand, at least to someone with a toxicology background and that does nanotoxicology
research. While I cannot speak from the perspective of someone with less background, I
believe the document does a good job in defining key concepts or ideas for those perhaps not
as versed in toxicology or nanotoxicology.
Praise for case study,
Document organization,
No revisions needed
307
E24-
PR
Yes, peer and public stakeholder comments are clearly summarized and conveyed through the
use of Tables. After looking at Tables 1-12 through 1-14, it appears that comments are clearly
summarized in Tables I-4-I-9 and 1-11. No further revisions are necessary.
Praise for case study,
Peer & public comments
are clear, No revisions
needed
308
E24-
PR
Yes, EPA responses to comments are also transparently and clearly conveyed through the
use of Tables. EPA responses are detailed with specific sections, Priority Research Area
boxes, or Highlight Text Boxes where revisions were incorporated as suggested. In addition,
EPA responses were fully justified in the few cases where no change was made
Praise for case study,
EPA responses to
comments clear, No
revisions needed
309
E24-
PR
It is evident that EPA was highly responsive to input from comments. Here are a few things
that stood out regarding EPA responsiveness to comments: Some reviewers did not see the
utility of comparisons to decaBDE while others did see the utility of the comparison. As a
result, EPA moved the majority of the text related to decaBDE to an Appendix and then
inserted text boxes on how decaBDE data might inform future research for MWCNTs.
Praise for case study,
EPA was responsive to
comments & made
appropriate changes,
No revisions needed
310
E24- 8
PR
It is evident that EPA was highly responsive to input from comments. Here are a few things
that stood out regarding EPA responsiveness to comments:
Based on the comments, it seems that the External Draft was not as concise and focused as
this revised version. I found the revised version to be very easy to follow and the use of text
boxes and movement of most of the text related to decaBDE to an Appendix allowed this
version to seem more focused and important ideas or research priorities really stood out.
Praise for case study,
EPA was responsive to
comments & made
appropriate changes,
No revisions needed
311
E24-
PR
It is evident that EPA was highly responsive to input from comments. Here are a few things
that stood out regarding EPA responsiveness to comments: EPA also seemed to be very
responsive to comments related to the literature (or lack thereof) used in the document. In fact,
suggestions from reviewers led to the inclusion of the Additional Information Highlight Boxes
which really added to the document.
Praise for case study,
EPA was responsive to
comments & made
appropriate changes,
No revisions needed
K-32
-------
Table K-13 (Continued): Comments received in response to the expert charge questions.
Comment
ID
312
313
314
315
316
317
318
319
320
321
Author
ID
E24-
PR
E24-
PR
E24-
PR
E24-
PR
E24-
PR
E24-
PR
E24-
PR
E24-
PR
E24-
PR
E25-
PR
Charge
Question
9
10
11
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
1
Comment Text Excerpt
(Note: Comment excerpts were edited to correct spelling and major grammatical and format issues [in
order to improve clarity], but generally were maintained in their original [as-submitted] form)
Not being familiar with Knowledge Maps, I have no real context for comparison. However, the
pilot project was clearly described and easy to understand. After reading the text, it was easy
to interpret the maps. I did not find anything that needs further clarification. The maps appear
to concisely present data discussed in Ch. 3 and Appendix D.
The maps are extremely useful in conjunction with the references in Appendix D (and Ch. 3) to
identify potential data gaps to drive future research to assist in risk assessment/management.
The maps are easy to follow once one has read the explanation of how to read and interpret
the maps. These maps are extremely valuable, especially for those conducting research with
MWCNTs.
Yes, each of the maps (PPM, TpM, and TfM) accurately and clearly presented the information
presented both in Ch. 3 and Appendix D. In fact, the Knowledge Maps helped to concisely
present the information presented. See answers to Q9 and Q1 0 above as further support of
this answer.
It is awkward that Table 2-3 is mentioned before Table 2-1 and 2-2.
CNTs deposits should probably just be CNT deposits
The first sentence states that 9 out of 13 RTI workshop participants identified the MATERIAL
PROCESSING STAGE of the MWCNT However the figure indicates that 9/13
participants found USE and material processing is not mentioned and this figure is found under
the Heading 2.4 Use. I believe the first statement must have been cut and paste from
elsewhere.
It seems the heading for the Neutral Research Area is incorrect as it says - Neutral Research
Area: Environmental Transport, Transformation and Fate, which is the same title for the
Neutral Research Area in Ch. 3. I'm not sure what the title heading should be.
This research area pertains to consumer exposure and yet the 2nd sentence says that the 9
participants were asked to rate the importance of ingestion, inhalation and dermal
OCCUPATIONAL exposures. It should be CONSUMER exposures.
Should be "as noted above" instead of "as note above"
Yes, absolutely; my impression was that the authors accurately conveyed the science
considered. Several spot checks on the literature I'm well familiar with did not show any
inaccuracies.
Themes
Praise for case study,
Knowledge Maps, No
revisions needed
Praise for case study,
Knowledge Maps, No
revisions needed
Praise for case study,
Knowledge Maps, No
revisions needed
Copy edit, DecaBDE
Comparison Box
Copy edit
Copy edit
Copy edit
Copy edit
Copy edit
Praise for case study,
Science is accurate &
clear, No revisions
needed
K-33
-------
Table K-13 (Continued): Comments received in response to the expert charge questions.
Comment
ID
Author Charge
ID Question
Comment Text Excerpt
(Note: Comment excerpts were edited to correct spelling and major grammatical and format issues [in
order to improve clarity], but generally were maintained in their original [as-submitted] form)
Themes
322
£25-
PR
I've got two critical points: A) Completeness: I'm not saying that you have to take everything
into account, but I collected the literature that I thought missing and indicated where I would
place/consider such literature in the report. (See Comment #332-#334 for (B) Comment).
Consider additional
literature
323
E25-
PR
A) Literature that one should check for possible consideration (until November 2012):
A.1) Environmental fate (chapter 1.1.4, see point B just mentioned above)
Hassellov, M., Readman, J.W., Ranville, J.F., Tiede, K., 2008. Nanoparticle analysis and
characterization methodologies in environmental risk assessment of engineered nanoparticles.
Ecotoxicology 17, 344-361.
Klaine, S.J., Alvarez, P.J.J., Batley, G.E., Fernandes, T.F., Handy, R.D., Lyon, D.Y.,
Mahendra, S., Mclaughlin, M.J., Lead, J.R., 2008. Nanomaterials in the environment:
Behavior, fate, bioavailability, and effects. Environmental Toxicology & Chemistry 27, 1825-
1851
Tiede, K., Boxall, A.B.A., Tear, S.P., Lewis, J., David, H., Hassellov, M., 2008. Detection and
characterization of engineered nanoparticles in food and the environment. Food Additives and
Contaminants 25, 795-821.
Consider additional
literature, Scientific
completeness,
Environmental fate
324
E25-
PR
A) Literature that one should check for possible consideration (until November 2012):
A.2) MWCNTs use in flame-retardant textiles (Chapter 1.3.2, 1.3.3)
Fasfous, 1.1., Radwan, E.S., Dawoud, J.N., 2010. Kinetics, equilibrium and thermodynamics of
the sorption of tetrabromobisphenol A on multiwalled carbon nanotubes. Applied Surface
Science 256, 7246-7252.
Haiou, Y., Jie, L, Xin, W., Zhiwei, J., Yujie, W., Lu, W., Jun, Z., Shaoyun, F., Tao, T., 2011.
Charing polymer wrapped carbon nanotubes for simultaneously improving the flame
retardancy and mechanical properties of epoxy resin. Polymer 52, 4891-4898.
Haiou, Y., Jie, L., Zhe, W., Zhiwei, J., Tao, T., 2009. Combination of carbon nanotubes with Ni
2O 3 for simultaneously improving the flame retardancy and mechanical properties of
polyethylene. Journal of Physical Chemistry C 113, 13092-13097.
Im, J.S., Bai, B.C., Bae, T.-S., In, S.J., Lee, Y.-S., 2011. Improved anti-oxidation properties of
electrospun polyurethane nanofibers achieved by oxyfluorinated multi-walled carbon
nanotubes and aluminum hydroxide. Materials Chemistry and Physics 126, 685-692.
[continued on next page]
Consider additional
literature, Scientific
completeness, MWCNT
flame retardants
K-34
-------
Table K-13 (Continued): Comments received in response to the expert charge questions.
Comment
ID
Author Charge
ID Question
Comment Text Excerpt
(Note: Comment excerpts were edited to correct spelling and major grammatical and format issues [in
order to improve clarity], but generally were maintained in their original [as-submitted] form)
Themes
324
£25-
PR
[continued from previous page]:
Ji Sun, I., Byong Choi, B., Tae-Sung, B., Se Jin, I., Young-Seak, L., 2011. Improved anti-
oxidation properties of electrospun polyurethane nanofibers achieved by oxyfluorinated multi-
walled carbon nanotubes and aluminum hydroxide. Materials Chemistry and Physics 126, 685-
692.
Lee, S.K., Bai, B.C., Im, J.S., In, S.J., Lee, Y.-S., 2010. Flame retardant epoxy complex
produced by addition of montmorillonite and carbon nanotube. Journal of Industrial and
Engineering Chemistry 16, 891-895.
Qiu, J., Zhang, S.-h., Wang, G.-j., Gong, Y.-l., 2009. Surface modification and application of
multi-walled carbon nanotubes in fire-retardant coatings. New Carbon Materials 24, 344-348.
Ullah, S., Ahmad, F., 2012. Enhancing the Char Resistant of Expandable Graphite Based
Intumescent Fire Retardant Coatings by using Multi-wall Carbon Nano Tubes for Structural
Steel. Diffusion and Defect Data Part B (Solid State Phenomena) 185, 90-93.
Verdejo, R., Barroso-Bujans, F., Rodriguez-Perez, M.A., de Saja, J.A., Arroyo, M., Lopez-
Manchado, M.A., 2008. Carbon nanotubes provide self-extinguishing grade to silicone-based
foams. Journal of Materials Chemistry 18, 3933-3939.
Wang, J.-X., Jiang, D.-Q., Gu, Z.-Y., Yan, X.-P., 2006. Multiwalled carbon nanotubes coated
fibers for solid-phase microextraction of polybrominated diphenyl ethers in water and milk
samples before gas chromatography with electron-capture detection. Journal of
Chromatography A 1137, 8-14.
Outside the review time range:
Knight, C.C., Ip, F., Zeng, C., Zhang, C., Wang, B., 2013. A highly efficient fire-retardant
nanomaterial based on carbon nanotubes and magnesium hydroxide. Fire and Materials 37,
91-99.
[same as previous page]
325
E25-
PR
A) Literature that one should check for possible consideration (till November 2012): A.3)
Production volumes estimations of raw material (Chapter 2)
Piccinno F, Gottschalk F, Seeger S, Nowack B. Industrial Production Quantities and Uses of
Ten Engineered Nanomaterials in Europe and the World. J Nanopart Res. 2012; 14
Future Markets. The World Market for Carbon Nanotubes, Nanofibers, Fullerenes, and POSS.
Future Markets, Inc.; 2011.
Consider additional
literature, Scientific
completeness, MWCNT
production volumes
K-35
-------
Table K-13 (Continued): Comments received in response to the expert charge questions.
Comment
ID
Author Charge
ID Question
Comment Text Excerpt
(Note: Comment excerpts were edited to correct spelling and major grammatical and format issues [in
order to improve clarity], but generally were maintained in their original [as-submitted] form)
Themes
326
£25-
PR
A) Literature that one should check for possible consideration (till November 2012):
A.4) Material synthesis (Chapter 2.2.2), these are all review articles
Ying, L.S., Salleh, M.A.b.M., Yusoff, H.B.M., Rashid, S.B.A., Abd Razak, J.B., 2011.
Continuous production of carbon nanotubes -A review. Journal of Industrial and Engineering
Chemistry 17, 367-376.
Golnabi, H., 2012. Carbon nanotube research developments in terms of published papers and
patents, synthesis and production. Scientia Iranica 19, 2012-2022.
Huang, J., Zhang, Q., Zhao, M., Wei, F., 2102. A review of the large-scale production of
carbon nanotubes: The practice of nanoscale process engineering. Chinese Science Bulletin
57, 157-166.
Yanfeng, M., Bin, W., Yingpeng, W., Yi, H., Yongsheng, C., 2011. The production of
horizontally aligned single-walled carbon nanotubes. Carbon 49, 4098-4110.
Kumar, M., Ando, Y., 2010. Chemical Vapor Deposition of Carbon Nanotubes: A Review on
Growth Mechanism and Mass Production. Journal of Nanoscience and Nanotechnology 10,
3739-3758.
Consider additional
literature, Scientific
completeness, MWCNT
synthesis
327
E25-
PR
A) Literature that one should check for possible consideration (till November 2012):
A.5) Releases during the material synthesis/handling (Chapter 2.2.2.2
Ogura, I., Sakurai, H., Mizuno, K., Gamo, M., 2011. Release potential of single-wall carbon
nanotubes produced by super-growth method during manufacturing and handling. Journal of
Nanoparticle Research 13, 1265-1280.
Consider additional
literature, Scientific
completeness
328
E25-
PR
A) Literature that one should check for possible consideration (till November 2012):
A.6) Releases during use (Chapter 2.4.2.)
Schlagenhauf, L, Chu, B.T.T., Buha, J., Nueesch, F., Wang, J., 2012. Release of Carbon
Nanotubes from an Epoxy-Based Nanocomposite during an Abrasion Process. Environmental
Science & Technology 46, 7366-7372.
Liang, S., Chen, G., Peddle, J., Zhao, Y., 2012. Reversible dispersion and releasing of single-
walled carbon nanotubes by a stimuli-responsive TTFV-phenylacetylene polymer. Chemical
Communications 48, 3100-3102.
Consider additional
literature, Scientific
completeness, MWCNT
release from polymers
K-36
-------
Table K-13 (Continued): Comments received in response to the expert charge questions.
Comment
ID
Author Charge
ID Question
Comment Text Excerpt
(Note: Comment excerpts were edited to correct spelling and major grammatical and format issues [in
order to improve clarity], but generally were maintained in their original [as-submitted] form)
Themes
329
£25-
PR
A) Literature that one should check for possible consideration (till November 2012):
A.7) Waste handling, incineration (Chapter 2.5.2.)
Walser, T., Limbach, L.K., Brogioli, R., Erismann, E., Flamigni, L, Hattendorf, B., Juchli, M.,
Krumeich, F., Ludwig, C., Prikopsky, K., Rossier, M., Saner, D., Sigg, A., Hellweg, S.,
Guenther, D., Stark, W.J., 2012. Persistence of engineered nanoparticles in a municipal solid-
waste incineration plant. Nature Nanotechnology 7, 520-524.
Mueller NC, Buha J, Wang J, Ulrich A, Nowack B. Modeling the flows of engineered
nanomaterials during waste handling. Environmental Science: Processes & Impacts. 2013;
15(1): 251-9.
Consider additional
literature, Scientific
completeness, Waste
incineration
330
E25-
PR
A) Literature that one should check for possible consideration (till November 2012):
A.8) Fate in water (Chapter 3.3.)
Schwyzer I, Kaegi R, Sigg L, Smajda R, Magrez A, Nowack B. Long-term colloidal stability of
10 carbon nanotube types in the absence/presence of humic acid and calcium. Environmental
Pollution.2012; 169(0): 64-73.
Consider additional
literature, Scientific
completeness, Fate in
water
331
E25-
PR
A) Literature that one should check for possible consideration (till November 2012):
A.9) Ecotoxicological effects (Chapter 5.2.)
Zhu Y, Zhao Q, Li Y, Cai X, Li W. 2006. The interaction and toxicity of multi-walled
carbon nanotubes with Stylonychia mytilus. J. Nanosci. Nanotechnol. 6:1357-1364.
Ghafari P, St-Denis CH, Power ME, Jin X, Tsou V, Mandal HS, Bols NC, Tang XW. 2008.
Impact of carbon nanotubes on the ingestion and digestion of bacteria by ciliated protozoa.
Nature Nanotechnology 3:347-351.
[continued on next page]
Consider additional
literature, Scientific
completeness,
Ecological receptors
K-37
-------
Table K-13 (Continued): Comments received in response to the expert charge questions.
Comment
ID
Author Charge
ID Question
Comment Text Excerpt
(Note: Comment excerpts were edited to correct spelling and major grammatical and format issues [in
order to improve clarity], but generally were maintained in their original [as-submitted] form)
Themes
331
£25-
PR
[continued from previous page]
Schwab F, Bucheli TD, Lukhele LP, Magrez A, Nowack B, Sigg L, Knauer K. 2010. Are
Carbon Nanotube Effects on Green Algae Caused by Shading and Agglomeration?
Environmental Science & Technology 45:6136-6144.
Templeton RC, Ferguson PL, Washburn KM, Scrivens WA, Chandler GT. 2006. Lifecycle
effects of single-walled carbon nanotubes (SWNTs) on an estuarine meiobenthic copepod.
Environ. Sci. Technol. 40:7387-7393.
Roberts AP, Mount AS, Seda B, Souther J, Qiao R, Lin S, Ke PC, Rao AM, Klaine SJ.
2007. In vivo biomodification of lipid-coated carbon nanotubes by Daphnia magna.
Environ. Sci. Technol. 41:3025-3029.
Cheng JP, Flahaut E, Cheng SH. 2007. Effect of carbon nanotubes on developing
zebrafish (Danio rerio) embryos. Environ. Toxicol. Chem. 26:708-716.
A. 10) Analytical methods for detecting, measuring, and characterizing (Appendix B).
I know, fullerenes are not really CNT and the studies are not without controversy, but, since
there's nothing else for comparison/analogies?
Farre, M., Perez, S., Gajda-Schrantz, K., Osorio, V., Kantiani, L., Ginebreda, A., Barcelu, D.,
2010. First determination of C60 and C70 fullerenes and N-methylfulleropyrrolidine C60 on the
suspended material of wastewater effluents by liquid chromatography hybrid quadrupole linear
ion trap tandem mass spectrometry. Journal of Hydrology 383, 44-51.
Sanchis, J., Berrojalbiz, N., Caballero, G., Dachs, J., Farre, M., Barcelo, D., 2011. Occurrence
of Aerosol-Bound Fullerenes in the Mediterranean Sea Atmosphere. Environmental Science &
Technology 46, 1335-1343.
[same as previous page]
332
E25-
PR
I've got two critical points (see Comment #322 for first part (A) of comment):
B) In the context of the comparison of MWCNT to decaBDE (i.e., justification for such a
comparison), I would pay more attention to the complex issues of fate we are faced with when
assessing the environmental distribution of nanomaterials, as described several times
(Christian et al. 2008; Hassellov et al. 2008; Klaine et al. 2008; Tiede et al. 2008).
See below for further explanation of this point.
DecaBDE comparison,
Scientific clarity,
Environmental fate
K-38
-------
Table K-13 (Continued): Comments received in response to the expert charge questions.
Comment
ID
Author Charge
ID Question
Comment Text Excerpt
(Note: Comment excerpts were edited to correct spelling and major grammatical and format issues [in
order to improve clarity], but generally were maintained in their original [as-submitted] form)
Themes
333
£25-
PR
I've got two critical points (see Comment #322 for first part (A) of comment):
B) The pros and cons in comparing MWCNTto decaBDE have been comprehensively
mentioned in the first review process and I do not want to repeat this; the
responses/adaptations of the authors satisfy so far. However, I'm still missing a kind of
introductory (immediate) explanation also from a material property perspective (well placed at
the beginning—for example, in Chapter 1.1.4. Selection of DecaBDE for Comparison) as to
why the comparison of MWCNT to decaBDE can be useful. It is somehow incomplete to focus
on a comparison that exclusively helps identify research gaps by illustrating the manufacturing
process and exposure scenarios and by understanding the process of regulating a chemical,
etc.
Revise text to clarify,
DecaBDE comparison,
Selection of decaBDE
334
E25-
PR
I've got two critical points (see Comment #322 for first part (A) of comment):
B) In this context I also wonder if the "partner chapters" G and H should be chapters of the
appendices or moved to the main text. I guess you could build these two chapters in a more
congruent way or rather exactly with the same structure. This would highlight similarities,
differences in material properties, and data gaps. For example, in Chapter H.3.3.1, water
solubility and Kow are said to be important physicochemical factors for predicting the behavior
of BDE-209 in water. I agree, and of course the Kow is not applicable to non-soluble
MWCNTs. Now the equivalent chapter for MWCNTs in G is not really there; however,
discussion/literature on the crucial parameter "fate of carbon nanotubes" in water is not totally
missing (see also literature suggestions made). Regarding Table H-3: an equivalent table—
even if partially empty—for MWCNTs in Chapter G would be very interesting.
Document organization,
DecaBDE comparison,
Separation of non-
priority area information
to appendices
335
E25- 2
PR
Yes. Basically, priority research areas are highlighted very well in the report. The text boxes
throughout are well organized and color-coded: red refers to the so-called priority areas for
research, grey to unprioritized areas, and green boxes are used for the decaBDE-MWCNT
comparison. However, numbering and indexing for those boxes is missing, making it difficult to
read the report as a hard copy.
Priority Research Areas,
Text boxes, Document
organization
336
E25-
PR
The Importance/Confidence Matrices are clear and easy to understand. But I wonder if a more
detailed textual analysis of these matrices' based output would provide the basis for a
research paper to be extracted from this report.
Priority Research Areas,
Future work, No
revisions needed
337
E25-
PR
Unfortunately, having only 13 participants is not enough for statistical evaluation, and the
competences of the participants are not really discussed.
Priority Research Areas,
RTI participants
K-39
-------
Table K-13 (Continued): Comments received in response to the expert charge questions.
Comment
ID
338
339
340
341
342
343
344
Author
ID
E25-
PR
E25-
PR
E25-
PR
E25-
PR
£25-
PR
£25-
PR
£25-
PR
Charge
Question
2
2
3
4
4
5
5
Comment Text Excerpt
(Note: Comment excerpts were edited to correct spelling and major grammatical and format issues [in
order to improve clarity], but generally were maintained in their original [as-submitted] form)
Furthermore, I'm not sure whether it makes sense that the authors only present in Chapter 6
— section 6.3 with the participants' rationale for designating the areas shows the outcomes of
the prioritization process in more detail — the identification of these priority areas by expert
stakeholders that in the end represents the crucial part for connecting risk research,
assessment, and management. Since these outcomes have been used to focus the
information in the whole case study, should this information not be included as a key part of
the report already at the beginning?
Another point is that even if the additional information highlight boxes (with scientific concepts
related to the priority areas) as well as the other figures and tables refer to priority research
areas (box 6 and 7 refer e.g., explicitly to issues covering such areas), the organizational
relationship between the different boxes (figures and tables) seems to me not too apparent in
the report. A table description with reference to the areas, for example, for Table 6-3, and
directly appended to the table (and not only in the main text body), would help. The same kind
of intra-textual references/indications on the interconnections are also needed for the green
boxes, since these boxes reveal how data on decaBDE should be used for MWCNT research
planning that is focused on these priority areas.
Please see the literature suggestions made above under Question 1 that have been grouped
by the different research areas.
Yes, this goal is achieved. However, it will be important to evaluate the missing literature
(please see the comments above) to exclude that until now not considered (possibly
conflicting) results would inform differently research planning.
Second, at the end of the green boxes concrete research questions are listed in each case to
inform MWCNT research planning in the identified priority areas. I'm wondering if it would
make sense to collect and prioritize all these concrete questions/instructions for researchers in
a kind of summarizing or concluding table, box, or figure. The same question arises for all
conclusions made on research planning outside of the comparison of decaBDE to MWCNT.
As partially mentioned in my introductory comments, I would suggest the authors look at: -
numbering and indexing all chapters, boxes, tables, and figures
As partially mentioned in my introductory comments, I would suggest the authors look at: -
creating more transparent and consistent organization regarding the alternating between
decaBDE and MWCNT in the text and document structure, which at present seems discordant
Themes
Document organization,
RTI participants,
Chapter 1 content
Document organization,
Numbers index
document, Clarify text
box types
Consider additional
literature
Consider additional
literature, Research
priorities are clear
Document organization,
Clarify research gaps,
Collect & prioritize
research gaps
Document organization,
Number & index
document
Document organization,
DecaBDE comparison,
DecaBDE & MWCNT
text
K-40
-------
Table K-13 (Continued): Comments received in response to the expert charge questions.
Comment
ID
345
346
347
348
349
350
351
352
353
Author
ID
E25-
PR
E25-
PR
£25-
PR
£25-
PR
£25-
PR
£25-
PR
£25-
PR
£25-
PR
£25-
PR
Charge
Question
5
5
5
5
5
5
6
7
7
Comment Text Excerpt
(Note: Comment excerpts were edited to correct spelling and major grammatical and format issues [in
order to improve clarity], but generally were maintained in their original [as-submitted] form)
As partially mentioned in my introductory comments, I would suggest the authors look at: -
numbering and indexing all pages in the hard copy
As partially mentioned in my introductory comments, I would suggest the authors look at: -
declaring explicitly and at the beginning of this manuscript how the authors distinguish
between content for appendices E and F and Chapter 3; appendix F and Chapters 4 and 5;
appendices G, H, and Chapter 6; and appendix G and Chapters 2-5
As partially mentioned in my introductory comments, I would suggest the authors look at: -
explicitly declaring at the beginning of the manuscript the use and purpose of the different
kinds of boxes
As partially mentioned in my introductory comments, I would suggest the authors look at: -
condensing the annotations and possibly moving relevant annotated text (as proposed above)
into the main document
As partially mentioned in my introductory comments, I would suggest the authors look at: -
condensing lengthy passages
As partially mentioned in my introductory comments, I would suggest the authors look at: -
removing any extraneous cross-references in the text between the chapters, especially when
later chapters refer back to earlier ones.
Yes.
Yes, my impression is that the authors took the comments seriously and did a very good job
revising the manuscript and responding to the suggestions/comments.
Some remaining suggestions: Missing current literature: There are still some additional studies
on MWCNTs that possibly should be included (please see above comments).
Themes
Document organization,
Number & index
document, Electronic
call outs
Revise text to clarify,
Document organization,
Main text vs.
appendices
Document organization,
Text boxes, Clarify text
box types
Document organization,
DecaBDE comparison,
Move appendix text to
main text
Document length,
Reduce redundancy
Document organization,
Cross-references,
Praise for case study,
Peer& Public
comments are clear, No
revisions needed
Praise for case study,
EPA was responsive to
comments & made
appropriate changes,
No revisions needed
Consider additional
literature
K-41
-------
Table K-13 (Continued): Comments received in response to the expert charge questions.
Comment
ID
354
355
356
357
358
359
360
361
362
Author Charge
ID Question
E25- 7
PR
E25- 7
PR
E25- 7
PR
E25- 8
PR
E25- 9
PR
E25- 10
PR
E25- 1 1
PR
E25- 1 1
PR
E25- 1 1
PR
Comment Text Excerpt
(Note: Comment excerpts were edited to correct spelling and major grammatical and format issues [in
order to improve clarity], but generally were maintained in their original [as-submitted] form)
Some remaining suggestions: Critique on the comparison with decaBDE:
As mentioned above, this comparison needs some additional legitimation, and an introductory
explanation from a material property perspective, as to why the comparison of MWCNT to
decaBDE is meaningful.
Some remaining suggestions: Missing gray literature: I know the report cannot fully present an
exhaustive review of all the literature. However, I would tell the reader in more detail how peer
review, non-peer review, and gray literature, etc., was treated.
As suggested once, I also think that the authors could possibly add a final concluding chapter
or remarks by giving some prudent statements on risk assessment that go beyond the modest
purpose of identifying and prioritizing research gaps.
See please comments on Questions 6 and 7.
Yes, absolutely. However, at the end of the chapter, I'd like to see some discussion or
conclusion concerning the objective of this subproject that was to develop methods (for
visually and concisely presenting information) in CEA.
Yes, if the model input data classification (known, not available, and similar descriptions) are
used for all maps and if this classification reflects the newest evidence available. If not, this
leads to misleading conclusions and non-legitimate decisions in this context.
Basically, yes. These maps are based on Table 3-1 and Table H-3 and transform information
in tabular form into figures, like flow charts. Figure J-2 reflects table H-3; however, this table is
based on old and more non-quantitative data for MWCNTs — there is newer (partially
contradicting) evidence available, as listed above. I would adapt this figure by considering
these data or at least, for example, the modeling studies on MWCNT and environment that the
authors referred to in Table 3-2. The same applies to map J-4.
Terrestrial box: Is a distinction of natural, agricultural, and urban soils meaningful/ feasible?
I know the figures already contain a legend; however, I think that a clear map description
directly appended to the table would facilitate analysis and understanding.
Themes
Revise text to clarify,
DecaBDE comparison,
MWCNT
physicochemical
properties
Revise text to clarify,
Scientific clarity,
Selection of literature
Document scope/
content, Clarify
discussion in text,
Expand risk assessment
discussion
Praise for case study,
EPA was responsive to
comments & made
appropriate changes,
Scientific completeness
Knowledge Maps,
Clarify discussion in text
Knowledge Maps
Knowledge Maps,
Scientific completeness,
Revise figures
Knowledge Maps,
Revise figures
Knowledge Maps,
Clarify discussion in text
K-42
-------
Table K-13 (Continued): Comments received in response to the expert charge questions.
Comment
ID
363
364
365
366
367
368
369
370
371
Author
ID
E25-
PR
E25-
PR
£25-
PR
£25-
PR
£25-
PR
£25-
PR
£25-
PR
£25-
PR
£25-
PR
Charge
Question
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
Comment Text Excerpt
(Note: Comment excerpts were edited to correct spelling and major grammatical and format issues [in
order to improve clarity], but generally were maintained in their original [as-submitted] form)
Possibly also consider the European definition* of nanomaterials based on their primary
particle size. *EU. Commission Recommendation of 1 8 October 201 1 on the definition of
nanomaterials. Official Journal of the European Union; 2011. p. 38-40.
Pg. 22 "Although no national or international consensus definition for nanomaterials exists, a
current working definition is a material having at least one dimension on the order of 1 to 1 00
nm (NSTC, 2011)." Pg. 34 "Nanoscale materials (nanomaterials) have been described as
having at least one dimension 1 approximately 1-100 nm (NSTC, 2011)." I had several times-
when reading through the whole document -the impression of already read, this is only the
first of such passages. I have not inventoried them, please see also next comment and the
generic remark above on lengthy.
This passage (or identical content) is - as an example - given at least three times throughout
the document, please see pg. 34, pg. 22, (pg. 25), (pg. 40), pg. 484
The CEA framework and process (principal elements) are explained too intensively (and
mostly from a general perspective) throughout the manuscript, this wearies the reader.
The groundwater compartment is missing: I would not be too secure about the harmlessness
of engineered nanomaterial in groundwater, experts talk according to each other, there are
hardly any data on that compartment. I would discuss this problem. See e.g., Colvin, V.L.,
2003. The potential environmental impact of engineered nanomaterials. Nature Biotechnology
21, 1166-1170. ; U.S. House Committee on Science. Hearing on Societal Implications of
Nanotechnology, April, 2003. 108th Congress (House Committee on Science, Washington,
DC, 2003.
Why such a source (Boner, 201 1 ) on drug-delivery systems in this context? The other source
is more realistic, however, about partnering with universities as a choice for nanotechnology
start-up firms (Wang and Shapira, 2012).
Boxes outlined in red and gray are not numbered and not listed in any directory?
Why are these kind of chapters not numbered in the manuscript? How the Case Study Was
Revised to Respond to Public and Peer Comments
Upper left part of the figure is not clear and unreadable.
Themes
Consider additional
literature, Clarify
discussion in text,
Definition of nano
Document length,
Reduce redundancy
Document length,
Reduce redundancy
Document length, CEA,
Future work
Priority Research Areas,
Ground water
Revise text to clarify,
Scientific accuracy,
Check a reference
Document organization,
Number & index
document, Text boxes
Document organization,
Numbers index
document
Copy edit, Revise figure
for clarity, Figure 1-3
K-43
-------
Table K-13 (Continued): Comments received in response to the expert charge questions.
Comment
ID
372
373
374
375
376
377
378
379
380
381
Author
ID
E25-
PR
E25-
PR
E25-
PR
E25-
PR
£25-
PR
£25-
PR
£25-
PR
£25-
PR
£25-
PR
£25-
PR
Charge
Question
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
Comment Text Excerpt
(Note: Comment excerpts were edited to correct spelling and major grammatical and format issues [in
order to improve clarity], but generally were maintained in their original [as-submitted] form)
1.1.4. Selection of DecaBDE for Comparison; Title and text do not correlate, the text is
something else, it's more on impacts of decaBDE on ecological and human health.
Missing source after "Although commercial MWCNT flame-retardant products are available,
their presence is 16 relatively new, and they are by no means abundant on the market." and in
Highlight Box 1
Is Johnston et al., 2010 the only source for this whole box on physicochemical properties and
fate? Possibly consider:
Aschberger, K., Johnston, H.J., Stone, V., Aitken, R.J., Hankin, S.M., Peters, S.A.K., Tran,
C.L., Christensen, P.M., 2010. Review of carbon nanotubes toxicity and exposure-Appraisal of
human health risk assessment based on open literature. Critical Reviews in Toxicology 40,
759-790.
Kohleretal., 2008 do not really treat flame-retardant textiles. I would not refer to this study in
this box. However, Schubauer-Berigan et al., 2011, Nicolas Messin (Nanocyl), 2012 seem up
to date, can you extract more from them? I wonder if the authors could provide in the main text
some concrete values on the MWCNT use/production amounts for the U.S.?
1. Should the authors also consider textile manufacturing that is not restricted only to the
incorporation of the flame retardant substances into the textiles?
2. End of Life: Reuse of products containing decaBDE and MWCNTs is clear for me. But, I can
hardly imagine reuse, recycling and disposal of decaBDE and MWCNTs?
Numbering and indexing for Prioritized and Unprioritized Research Area Boxes is missing.
"The release of other substances used during purification is also possible." This sentence is
isolated and alone, are there sources or/and other examples of such substances that may be
released?
"Environmental releases are expected from upholstery textiles coated with flame retardants
due to 1 (1 ) the potential use scenarios for the upholstery textiles and (2) the physicochemical
properties of MWCNTs." Is this a meaningful sentence? Or should it be Environmental release
is influenced by...
Themes
Document organization,
DecaBDE comparison,
Selection of decaBDE
Revise text to clarify,
Scientific clarity, Add
literature citation
Revise text to clarify,
Clarify research gaps,
Minor Text edit
Revise text to clarify,
Scientific completeness,
MWCNT
physicochemical
properties
Revise text to clarify,
Scientific completeness,
MWCNT Production
Volumes
Document scope/
content, Figures, Figure
2-1 revision
Revise text to clarify,
Scientific accuracy,
Figure 2-1 revision
Document organization,
Text boxes
Revise text to clarify,
Scientific clarity, Add
literature citation
Revise text to clarify,
Scientific clarity,
MWCNT release from
polymers
K-44
-------
Table K-13 (Continued): Comments received in response to the expert charge questions.
Comment
ID
382
383
384
385
386
387
Author
ID
E25-
PR
E25-
PR
£25-
PR
£25-
PR
£25-
PR
£25-
PR
Charge
Question
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
Comment Text Excerpt
(Note: Comment excerpts were edited to correct spelling and major grammatical and format issues [in
order to improve clarity], but generally were maintained in their original [as-submitted] form)
I do not know any studies on MWCNT for such release, but I'm asking if you should mention
here available experimental/analytical release studies performed for other engineered
nanomaterials?
Benn, T.M., Westerhoff, P., 2008. Nanoparticle Silver Released into Water from Commercially
Available Sock Fabrics. Environmental Science & Technology 42, 4133-4139.
Geranio, L, Heuberger, M., Nowack, B., 2009. The Behavior of Silver Nanotextiles during
Washing. Environmental Science & Technology 43, 8113-8118.
Windier, L., Lorenz, C., von Goetz, N., Hungerbuehler, K., Amberg, M., Heuberger, M.P.,
Nowack, B., 2012. Release of titanium dioxide from textiles during washing. Environmental
Science & Technology 46, 8181-8188.
"In general, information on environmental concentrations of MWCNTs was not found." This
contradicts e.g., what is presented in Table 3-2.
"What concentrations ..." Should you refer to Table 3-2?
Perhaps you can also extract some values from: Koelmans, A.A., Nowack, B., Wiesner, M.R.,
2009. Comparison of manufactured and black carbon nanoparticle concentrations in aquatic
sediments. Environmental Pollution 157, 1110-1116.
...likely would be released from upholstery textiles in the particulate phase, in a manner similar
to BDE-209 (Nowack et al., 2012)" Is Nowack et al. 2012 the right source here?
4.3. Ecological Exposure... Ecological seems in the whole manuscript to be used synonym for
environmental, however, environment is used as well, is such use consistent?
Themes
Document scope/
content, Scientific
completeness, MWCNT
release from polymers
Revise text to clarify,
Revise for consistency,
MWCNT environmental
Concentrations
Revise text to clarify,
Scientific clarity,
MWCNT environmental
Concentrations
Consider additional
literature, Scientific
completeness, Aquatic
receptors
Revise text to clarify,
Scientific accuracy,
Check a reference
Revise text to clarify,
Scientific clarity,
Ecological receptors
K-45
-------
Table K-13 (Continued): Comments received in response to the expert charge questions.
Comment
ID
388
389
390
391
392
393
394
395
Author
ID
E25-
PR
E25-
PR
E25-
PR
£25-
PR
£25-
PR
£25-
PR
£26-
PR
£26-
PR
Charge
Question
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
General
Impression
General
Impression
Comment Text Excerpt
(Note: Comment excerpts were edited to correct spelling and major grammatical and format issues [in
order to improve clarity], but generally were maintained in their original [as-submitted] form)
"For aquatic ecosystems, little information was identified for MWCNTs; conversely, much
information was identified on the potential effects of MWCNTs in terrestrial ecosystems." Is it
not just the other way around? This seems wrong to me, I added above several
ecotoxicological literature on aquatic systems (see please comments above), please find here
a few terrestrial studies:
Scott-Fordsmand JJ, Krogh PH, Schaefer M, Johansen A. 2008. The toxicity testing of double-
walled nanotubes-contaminated food to Eisenia veneta earthworms. Ecotoxicology and
Environmental Safety 71:616-619.63;
Atieh MA, Fakhru'l-Razi A, Girun N, Chuah TG, El-Sadig M, Biak DRA. 2006. Effect of multi-
wall carbon nanotubes on the mechanical properties of natural rubber. Composite Structures
75.
5.3.3. Climate Change Is this chapter really necessary, I would not refer to climate dynamics,
this link is - as the authors said - totally unclear (unnecessary?) from a scientific perspective.
Why not adding also some recommendations regarding making industrial nanomaterial
production and application volumes transparent?
Could the authors give an insight into the estimation procedure/method of finances and time
frame?
"No data are available on MWCNT concentrations in ...These statements contradict to what is
presented in Chapter 3.5, e.g., Table 3-2
No numbering and indexing for titles: Toxico kinetics and Body Burden in Aquatic Systems;
Toxi co kinetics and Body Burden in Terrestrial Systems?
The document is well structured and cross-referencing links within the document are a highly
appreciated feature. Overall, the accuracy is good (to my knowledge).
Clarity of presentation sometimes suffers due to the length of the text; however, often the
information is also presented in an overview in tables which is very good. But it would also be
good to summarize after each chapter.
Themes
Check accuracy of
statement, Revise for
consistency, Aquatic
receptors
Document scope/
content, Climate change
Priority Research Areas,
Specific research
questions
Priority Research Areas,
Specific research
questions, RTI
workshop
Revise text to clarify,
Revise for consistency,
MWCNT environmental
concentrations
Document organization,
Numbers index
document
Praise for case study,
Cross-references
Document length,
Clarify discussion in
text, Chapter summaries
K-46
-------
Table K-13 (Continued): Comments received in response to the expert charge questions.
Comment
ID
396
397
398
399
400
401
402
403
Author
ID
E26-
PR
E26-
PR
E26-
PR
E26-
PR
£26-
PR
£26-
PR
£26-
PR
£26-
PR
Charge
Question
General
Impression
General
Impression
General
Impression
General
Impression
General
Impression
General
Impression
General
Impression
General
Impression
Comment Text Excerpt
(Note: Comment excerpts were edited to correct spelling and major grammatical and format issues [in
order to improve clarity], but generally were maintained in their original [as-submitted] form)
The whole process of the CEA is thoroughly described and provides a very good introduction
to the remainder of the document, as well as to the terminology used. Also, the previous steps
of the current document are adequately described and give the reader a good understanding
of why the document is built as it is. However, a good understanding of the framework, as
such, was for me better achieved after reading Chapters 6.1 and 6.2.
Figure 1-3 provides a very good overview of the results of the collective judgment prioritization
process.
The relatively long introduction and explanation of flame retardancy is not at all times very
relevant to CEA — I would reduce it to a few pages explaining the principles of MWCNT for
flame retardancy and providing the perspectives of potential volumes and thus releases — and
put the remainder in an appendix.
Additional highlight boxes work well in conveying the additional information and are in most
instances relevant.
It is a bit confusing that LCA is mixed up a bit with risk assessment — it is clear that e.g., the
feedstock may not be of importance for risk assessment of MWCNT, but in an LCA it can be of
importance due to other types of emissions, resource usage, etc. Likewise in the
manufacturing, there will be extensive use of solvents to refine the MWCNT or reagents for
functionalization — which are not taken into account in the CEA (except to a minor extent in
Chapter 4.5 "cumulative exposures") but would be in an LCA (as well as the waste material/by-
products from the refining). It is stressed several times, e.g., on page 5.1 lines 3-7, that the
CEA framework includes other impacts than toxicological — however, this is not at all well
communicated in the document as a whole. The whole Chapter 2 on the product life cycle
could have put a lot more weight on these issues rather than just being briefly mentioned in
Chapter 5. 3.
Areas in which there is no information are often not mentioned, i.e., it should be clearly stated
that in order to actually do an assessment we would need this information. An example is
Chapter 3.3 where nothing is mentioned about stability and transformation in water and
sediment.
The conclusions are generally sound.
Generally, I find that the science is accurately conveyed.
Themes
Document organization,
Clarify discussion in
text, Chapter 1 content
Praise for case study,
Figures, Figure 1-3
Document organization,
Move text to an
appendix, MWCNT
Flame-Retardants
Praise for case study,
Document organization,
Text boxes
Revise text to clarify,
Scientific clarity, LCA
Revise text to clarify,
Clarify research gaps
Praise for case study,
Science is accurate &
clear
Praise for case study,
Science is accurate &
clear
K-47
-------
Table K-13 (Continued): Comments received in response to the expert charge questions.
Comment
ID
404
537
405
406
407
408
409
410
Author
ID
E26-
PR
E26-
PR
£26-
PR
£26-
PR
£26-
PR
£26-
PR
£26-
PR
£26-
PR
Charge
Question
1
1
1
2
3
4
4
5
Comment Text Excerpt
(Note: Comment excerpts were edited to correct spelling and major grammatical and format issues [in
order to improve clarity], but generally were maintained in their original [as-submitted] form)
A few comments to Chapter 5 are: Page 5-29 - the review of Eckelman et al. (2012) lacks the
essential information that, when applying the USEtox framework, the authors make
adaptations in order to handle the CNTs. USEtox is developed for organic chemicals and will
probably not be suitable for CNTs. This could be a factor to consider in their results.
A few comments to Chapter 5 are: In Chapter 5.3.3, I don't necessarily think there is an
unclear relationship between climate change and production of MWCNT. The high CO2 in
Khanna et al. is due to the energy consumption and, considering the high energy consumption
listed earlier, there will be huge climate change impacts related to all production pathways.
Chapter 6.3.4.4 lacks reflections on the gaps in knowledge regarding broad environmental
considerations (life cycle assessments).
I find that the boxes presenting "prioritized" or "unprioritized" research areas are very
illustrative and very clearly highlight the opinions of previous experts.
I think that the incineration part (Chapter 2.5.2.2) is relying too much on anticipation that
MWCNT will be oxidized in incineration. A few papers I have (included at the bottom of this
review) do not clearly indicate that this is so. And even controlled incineration has clumps of
waste that are not fully incinerated or where the temperature does not reach maximum. (Sui,
YC et al. (2001). Structure, thermal stability, and deformation of multibranched carbon
nanotubes synthesized by CVD in the AAO template. J Phys Chem B 105(8): 1523-1527.;
Cataldo, F.( 2002). A study on the thermal stability to 1 000°C of various carbon allotropes and
carbonaceous matter both under nitrogen and in air. Fullerenes, Nanotubes and Carbon
Nanostructures. 10(4): 293-311.)
I find that the document is very supportive in future research planning. Chapter 6 makes a
good conclusion on the process.
In Chapter 5 it would be very helpful to have a table that better summarizes the studies
mentioned in the text. Table 5-1 is not very helpful (it almost says there are no data on
anything) since there are useful results from a range of studies. Tables 5-2 and 5-3 convey
much more useful information.
The overall structure is logical. However, as previously mentioned, I find that the information in
Chapters 6.1 and 6.2 would be better placed in Chapter 1 to give a better understanding of the
purpose of the document.
Themes
Revise text to clarify,
Scientific clarity, Add
detail from existing
reference
Praise for case study,
Scientific accuracy,
Climate change
Revise text to clarify,
Clarify research gaps,
LCA
Praise for case study,
Research priorities are
clear
Consider additional
literature, Scientific
accuracy, Waste
incineration
Praise for case study,
Research priorities are
clear, Supports
research planning
Revise text to clarify,
Clarify research gaps,
Table 5-1
Document organization,
Clarify discussion in
text, Chapter 1 content
K-48
-------
Table K-13 (Continued): Comments received in response to the expert charge questions.
Comment
ID
411
412
413
414
415
416
417
418
419
420
Author
ID
E26-
PR
E26-
PR
E26-
PR
£26-
PR
£26-
PR
£26-
PR
£26-
PR
£26-
PR
£26-
PR
£26-
PR
Charge
Question
6
7
8
9
9
9
10
11
Specific
Observations
Specific
Observations
Comment Text Excerpt
(Note: Comment excerpts were edited to correct spelling and major grammatical and format issues [in
order to improve clarity], but generally were maintained in their original [as-submitted] form)
They are comprehensively and transparently conveyed.
I find that the comments and responses have been thoroughly described in Appendix I. I
particularly like that the comments and responses have been grouped and thus provide a
better overview of how the EPA has responded to the comment. This makes it more
transparent and clear.
It is evident that EPA has been responsive to the comments.
The rationale is OK as described. I find that the knowledge maps provide very good visual
information.
I must admit that the text in itself didn't provide a good understanding of the structures of the
figures with nodes and edges until I saw the figures (but maybe that's my English that does not
suffice).
Furthermore, the text (and maybe the figures as well?) lacks good explanation of what is the
"weight of evidence" for the information conveyed regarding the I Physicochemical Properties
Map (PPM). There is a mix-up between Varying parameters' and 'unknowns' in my
understanding. And the purple "potential," does that represent lack of evidence or what? How
is it represented that there is no knowledge?
I think that the TfM and TpM clearly illustrate the limited knowledge on MWCNT environmental
fate. The same clarity is missing in the PPM. I don't have good ideas for how to improve this.
Figure J-2 seems to lack potential coating/functionalization and the consequences of this. But
apart from this, I believe they do summarize the information given in other parts of the
document.
It refers to Table 3-2 that has the environmental concentrations, whereas it should refer to
Table 3-1
Confusing that lipophilicity and Kow are separated — just use one of them.
Themes
Praise for case study,
Peer & public
comments are clear
Praise for case study,
Peer & public
comments are clear,
EPA responses to
comments are clear
Praise for case study,
EPA was responsive to
comments & made
appropriate changes
Knowledge Maps
Knowledge Maps,
Document organization
Knowledge Maps,
Clarify discussion in
text, Clarify knowledge
gaps
Knowledge Maps,
Clarify research gaps,
PPM KM
Knowledge Maps,
Clarify research gaps,
PPM KM
Copy edit
Knowledge Maps,
Revise figure for clarity,
PPM KM
K-49
-------
Table K-13 (Continued): Comments received in response to the expert charge questions.
Comment
ID
421
422
423
424
425
426
427
428
429
430
Author
ID
E26-
PR
E26-
PR
£26-
PR
£26-
PR
£26-
PR
£26-
PR
£27-
PR
£27-
PR
£27-
PR
£27-
PR
Charge
Question
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
General
Impression
General
Impression
General
Impression
General
Impression
Comment Text Excerpt
(Note: Comment excerpts were edited to correct spelling and major grammatical and format issues [in
order to improve clarity], but generally were maintained in their original [as-submitted] form)
An explanation for MWCNT is missing.
Repetition of sentence.
The reader is left curious about which "three instances" there was disagreement between
stakeholders — either refer to another place in the document where this is explained or place a
footnote explaining this.
The table should also indicate where MWCNT are placed — with application methods and
mechanisms.
The figure is a bit misleading since releases to the environment may also occur from the
material processing and product manufacturing. Would be better to have the "environment" to
the right and the application running through all life cycle stages.
You should double check that the crosses are still in the right place, e.g., it is strange that the
cross for reproductive/developmental is under no data when there are some studies??
The document provides a comprehensive overview on the use of MWCNTs as flame
retardants in textiles. It covers important issues such as life cycle, transport/fate, exposure, and
hazard, and gives recommendations for identifying and prioritizing research needs.
The document is very clear and interesting to read.
According to the title, a comparison to DecaBDE flame retardant coatings should be made.
This relates more to the previous version (July 2012 for public consultation) of the document,
where MWCNT was set in direct comparison to Deca BDE. In the current document, most of
the content on Deca BDE was moved to an Annex, which is appreciated as the document
focuses more on MWCNTs; however, a change in the title could be envisaged to better fit the
current scope.
I am not completely sure how much DecaBDE contributes to inform MWCNT assessments in
relation to the research priorities, as these are very different compounds with different
characteristics and behavior in the environment and in the human body. In my review, the
comparison should relate more to the (expected) risks and MWCNT could be used as an
alternative to the phased out DecaBDE. For this, a general overview with comparisons could
be presented.
Themes
Knowledge Maps,
Clarify discussion in text
Copy edit, Reduce
redundancy
Revise text to clarify,
RTI participants
Revise text to clarify,
Add detail from cited
study, Table 1-5
Revise text to clarify,
Revise figure for clarity,
Figure 2-2
Revise text to clarify,
Scientific accuracy,
Table 5-1
Praise for case study,
Document is
comprehensive
Praise for case study,
Science is accurate &
clear
Revise text to clarify,
Title
Revise text to clarify,
DecaBDE comparison,
MWCNT risks
K-50
-------
Table K-13 (Continued): Comments received in response to the expert charge questions.
Comment
ID
431
432
433
434
435
436
437
438
Author
ID
E27-
PR
E27-
PR
£27-
PR
£27-
PR
£27-
PR
£27-
PR
£27-
PR
£27-
PR
Charge
Question
General
Impression
General
Impression
General
Impression
1
1
1
1
1
Comment Text Excerpt
(Note: Comment excerpts were edited to correct spelling and major grammatical and format issues [in
order to improve clarity], but generally were maintained in their original [as-submitted] form)
In addition, MWCNT cannot be used alone in textiles (and other applications) as flame
retardants because they do not comply with the flammability test and instead are used with
other flame retardants to decrease the flammability synergistically. Therefore, in my view, it is
not adequate to directly compare MWCNT alone with DecaBDE. A representative compound
which is used in combination with CNTs (e.g., phosphonates) should be included in the
comparison.
Just a remark. 'Comprehensive Environmental Risk Assessment' - as a European, I was
wondering about the 'Environmental' as we usually speak about ' Human Health and
Environmental' and would not consider Human Health to be included in an Environmental Risk
Assessment.
Our group is currently taking part in the FP7 Project 'DEROCA' Development of safe and eco-
friendly flame retardant materials based on CNT co-additives for commodity polymers. Our
task is a comparative risk assessment to other flame retardant solutions. If you are interested
(we have just started and no results yet available), here is the website link:
http://www.deroca.eu/DEROCA_WEB/UK/Home.awp
The science is, in general, accurately and clearly conveyed throughout the document.
I consider that the data discussed in this document have been retrieved from the latest peer-
reviewed scientific papers and most accurate reviews.
However, there are some specific areas and references that can be extended and included.
References (often not the original references, but those of reviews. I can only refer to those
that I know very well):
Table 1-8: ECB 2003: EU Risk Assessment Report as reference for DecaBDE phys-chem
properties; and
However, there are some specific areas and references that can be extended and included.
References (often not the original references, but those of reviews). Table 1-9: Johnston and
Aschberger as references for MWCNT phys-chem properties. Please refer to original sources,
where these properties have been determined.
Be consistent with references. For example, in Text Boxes 5-1 (but also others), only part of
the text is referenced. Either give references for all statements or no references in text boxes
and only in the accompanying text.
Themes
Document scope/
content, DecaBDE
comparison, MWCNTs
flame-retardants
CEA
Document scope/
content, Future work,
MWCNTs flame-
retardants
Praise for case study,
Science is accurate &
clear
Praise for case study,
Science is accurate &
clear
Consider additional
literature, Scientific
completeness,
decaBDE data
Consider additional
literature, Scientific
clarity, Review articles
vs. original sources
Revise text to clarify,
Scientific clarity, Clarify
use of references
K-51
-------
Table K-13 (Continued): Comments received in response to the expert charge questions.
Comment
ID
Author Charge
ID Question
Comment Text Excerpt
(Note: Comment excerpts were edited to correct spelling and major grammatical and format issues [in
order to improve clarity], but generally were maintained in their original [as-submitted] form)
Themes
439
£27-
PR
Additional relevant information sources on MWCNTs could be considered:
There is a REACH registration dossier on MWCNT (for Baytubes and Nanocyl), parts of which
are publicly available. It contains, for example, information on TG GLP studies which were not
published in peer-reviewed papers. This information could be used to fill gaps in Table 5-1 and
the corresponding text. Link: http://apps.echa.europa.eu/registered/data/dossiers/DISS-
b281d1aO-c6d8-5dcf-e044-00144f67d031/DISS-b281d1aO-c6d8-5dcf-e044-
00144f67d031_DISS-b281d1aO-c6d8-5dcf-e044-00144f67d031.html; It also gives rough
information on manufacturing volume, as they were registered for the tonnage volume: 100-
1000 tons/year.
Consider additional
literature, Scientific
completeness, MWCNT
Production Volumes &
Table 5-1
440
E27-
PR
The Japanese Risk Assessment of Manufactures Nanomaterials - CNT (NEDO project):
Nakanishi, J., Ed. Risk assessment of manufactured nanomaterials: Carbon nanotube (CNT).
Final report issued on August 12, 2011. NEDO project (P06041) Research and development of
nanoparticle characterization methods. Link: http://www.aist-
riss.jp/main/modules/product/nano_rad.html?ml_lang=en
Consider additional
literature, Scientific
completeness, MWCNT
analytical methods
441
E27-
PR
Human Health Hazard Assessment and Classification of Carbon Nanotubes. National
Industrial Chemicals Notification and Assessment Scheme (NICNAS). Safe Work Australia.
October 2012. ISBN 978 0 642 78542 8 [Online PDF]. Link:
http://www.safeworkaustralia.gov.au/sites/swa/about/publications/pages/human-health-hazard-
assessment-and-classification-of-carbon-nanotubes
Consider additional
literature, Scientific
completeness,
Occupational exposure
442
E27-
PR
I disagree with the following statements in Chapter 5: 5-10: Subchronic - chronic systemic
toxicity; The way it is presented it seems that there are no subchronic/chronic studies (by any
route) available - or no studies testing systemic effects. (In vivo) - Studies are not made
separately for local or systemic effects, but they can be observed and/or reported. There are
two subchronic inhalation studies with MWCNTs available which have not reported systemic
effects. In addition, the Pauluhn study included a 6 M post-exposure observation period -this
could be added where the study is described.
Revise text to clarify,
Scientific accuracy,
Systemic toxicity
443
E27- 1
PR
I disagree with the following statements in Chapter 5: 5-14: Relevance of in vitro data; The
introductory sentence should be revised. There is a general requirement to replace animal
studies by alternative methods and some in vitro test are validated and regularly used in
regulatory risk assessment: mutagenicity (usually only in vivo confirmation of positive in vitro
results), dermal absorption, and skin and eye irritation.
Revise text to clarify,
Scientific accuracy,
Alternative methods
K-52
-------
Table K-13 (Continued): Comments received in response to the expert charge questions.
Comment
ID
Author Charge
ID Question
Comment Text Excerpt
(Note: Comment excerpts were edited to correct spelling and major grammatical and format issues [in
order to improve clarity], but generally were maintained in their original [as-submitted] form)
Themes
444
£27-
PR
A good overview on the suitability of in vitro methods (not validated for NM yet) for NM hazard
assessment are given in:
SCCS 2012: Guidance on the safety assessment of nanomaterials in cosmetics. Link:
http://ec.europa.eu/health/scientific_committees/consumer_safety/docs/sccs_s_005.pdf;
Hartung et al. 2011. Alternative in vitro assays in nanomaterial toxicology. Link:
http://onlinelibrary.wiley.com/doi/
10.1002/wnan. 153/abstract;jsessionid=DF763EC777E464BDF520FF2C46822C6C.d03t03
Consider additional
literature, Scientific
completeness,
Alternative methods
445
E27-
PR
Several EU framework projects are focusing on development of alternative methods for NM
toxicity testing, in vitro-in vivo extrapolation (e.g., ITS-Nano, MARINA, ENPRA).
Consider additional
literature, Scientific
completeness,
Alternative methods
446
E27-
PR
Further information on in vitro data and alternative methods to support toxicological studies can
be found in:
Adler, S., D. Basketter, et al. (2011). Alternative (non-animal) methods for cosmetics testing:
current status and future prospects-2010. Archives of toxicology 85(5): 367-485.;
Hartung, T., B. J. Blaauboer, et al. (2011). An expert consortium review of the EC-
commissioned report "alternative (Non-Animal) methods for cosmetics testing: current status
and future prospects-2010". ALTEX28(3): 183-209.
Consider additional
literature, Scientific
completeness,
Alternative methods
447
E27-
PR
Priority areas can be clearly identified as they are clearly highlighted throughout the document.
The presentation of the division of importance and confidence ranking (and associated colors)
is considered useful, especially the confidence because it plays an important role in the
regulatory decision making. For example, in Chapter 4, Section 4.2 Human Exposure and
Kinetics Leading to Dose, priority research areas identified inhalation exposure route as the
highest priority area, followed by ingestion and dermal exposure. This information is supported
by the literature presented.
Praise for case study,
Research Priorities are
clear
448
E27-
PR
In each subsection, exposure scenarios sorted by exposure route (inhalation, oral, and dermal)
are presented for the occupational scenarios. For clarity, a summary table of this information
highlighting the activities related to each kind of exposure with exposure levels data (if
available) is suggested for:
Synthesis, Processing, and Handling.
Formulation of Flame Retardant, Application to Textiles, Upholstering.
Storage of MWCNTs, Flame-Retardant Formulations, Treated Textiles, and Upholstered
Products.
Disposal and Recycling of MWCNTs, Flame-Retardant Formulations, Treated Textiles, and
Upholstered Products.
Document organization,
Clarify research gaps,
Add summary table
K-53
-------
Table K-13 (Continued): Comments received in response to the expert charge questions.
Comment
ID
Author Charge
ID Question
Comment Text Excerpt
(Note: Comment excerpts were edited to correct spelling and major grammatical and format issues [in
order to improve clarity], but generally were maintained in their original [as-submitted] form)
Themes
449
£27-
PR
Additional Comments:
The identified priority areas for Human Health are in accordance with the findings within the
ENRHES (http://ihcp.jrc.ec.europa.eu/whats-new/enhres-final-report) and the currently running
ITS-Nano FP7 project (http://www.its-nano.eu/). For the Environment, there is a different view
on the terrestrial compartment. It was identified as priority research area under ENRHES and
ITS-Nano. Sediment and soil are usually sinks for NM (e.g., see the paper: Aschberger.
Environment International 37 (2011) 1143-1156). Sedimented NM in wastewater, which was
also identified as priority, could end up in soil together with sludge.
Priority Research
Areas, Disagreement
with research priorities
identified by others
450
E27-
PR
In general, reasons for not prioritized areas could be better explained (e.g., terrestrial
environment).
Priority Research
Areas, RTI participants,
Rationale for non-
priority areas
451
E27-
PR
Please find some additional references below and in the Specific Observations section (not
exhaustive, as I didn't have time to check everything carefully): Studies on release of polymer
bound CNTs (it seems that not much information has been included in the report):
Wohlleben et al. 2011. On the lifecycle of nanocomposites: comparing release fragments and
their in vivo hazards from three release mechanisms and four nanocomposites. Small. No 16,
2384-2395.
Wohlleben et al. 2013. Elastic CNT-polyurethane nanocomposite: synthesis, performance, and
assessment of fragments released during use. Nanoscale 369-380
Stahlmecke et al. (IUTA Duisburg Germany) 2013. Investigations on the possible CNT release
at the end-of-Life of CNT-composites. Poster (not found on internet, but could be made
available to you).
Ogura et al. (AIST, TASC Japan) 2012. Potential release of carbon nanotubes from their
composites during grinding. Poster at Nanosafe 2012:
http://www.certh.gr/dat/A974408A/file.pdf
Movahedi, Alireza 2008. Risk assessment of nanoparticle release from the structure of
polymeric nanocomposites - Master Thesis University Goteburg:
http://libris.kb.se/bib/12060839
Consider additional
literature, Scientific
completeness, MWCNT
release from polymers
K-54
-------
Table K-13 (Continued): Comments received in response to the expert charge questions.
Comment
ID
Author Charge
ID Question
Comment Text Excerpt
(Note: Comment excerpts were edited to correct spelling and major grammatical and format issues [in
order to improve clarity], but generally were maintained in their original [as-submitted] form)
Themes
452
£27-
PR
Please find some additional references below and in the Specific Observations section (not
exhaustive, as I didn't have time to check everything carefully): Environmental Fate:
O'Carroll, D. M., Liu, X., Mattison, N. T., Petersen, E. J. 2013. Impact of size on carbon
nanotube transport in natural porous media. Journal of Colloid & Interface Science, 390(1), 96-
104.
Zhang, L, Petersen, E. J., Zhang, W., Chen, Y. S., Cabrera, M., Huang, Q. 2012. Phase
distribution of 14C-labeled multi-walled carbon nanotubes in aqueous systems containing
model solids: clay. Environmental Pollution. 166, 75-81.
Liu, X., O'Carroll, D.M., Petersen, E.J., Huang, Q., Anderson, L. 2009. Mobility of multi-walled
carbon nanotubes in porous media. Environmental Science & Technology. 43, (21), 8153-
8158.
Consider additional
literature, Scientific
completeness,
Environmental fate
453
E27-
PR
Please find some additional references below and in the Specific Observations section (not
exhaustive, as I didn't have time to check everything carefully):
Ecotoxicity:
Petersen, E. J., Pinto, R. A., Mai, D. J., Landrum, P. F., Weber, W. J., Jr. 2011. Influence of
Polyethyleneimine Graftings of Carbon Nanotubes on their Accumulation and Elimination by
and Toxicityto Daphnia magna. Environmental Science and Technology. 44, (3), 1133-1138.
Galloway, Tamara, et al. "Sublethal toxicity of nano-titanium dioxide and carbon nanotubes in a
sediment dwelling marine polychaetes." Environmental Pollution 158.5 (2010): 1748-1755.
(NB! SWCNT & marine sediment!!!). NOEC > 0.03 g/kg to Arenicola marina (lugworm)
Consider additional
literature, Scientific
completeness,
Ecological receptors
454
E27-
PR
Please find some additional references below and in the Specific Observations section (not
exhaustive, as I didn't have time to check everything carefully):
Bioaccumulation:
Petersen, E. J., Pinto, R. A., Mai, D. J., Landrum, P. F., Weber, W. J., Jr. 2011. Influence of
Polyethyleneimine Graftings of Carbon Nanotubes on their Accumulation and Elimination by
and Toxicityto Daphnia magna. Environmental Science and Technology. 44, (3), 1133-1138.
Petersen, E. J., Akkanen, J., Kukkonen, J. V. K., Weber, W. J., Jr. 2009. Biological Uptake and
Depuration of Carbon Nanotubes by Daphnia magna. Environmental Science & Technology.
43, (8), 2969-2975.
Consider additional
literature, Scientific
completeness,
Bioaccumulation
455
E27-
PR
The document is mainly focusing on the research areas but does not include recommendations
on how this could be achieved.
Document scope/
content, Research
implementation
K-55
-------
Table K-13 (Continued): Comments received in response to the expert charge questions.
Comment
ID
456
457
458
459
460
461
462
463
464
465
Author
ID
E27-
PR
E27-
PR
£27-
PR
£27-
PR
£27-
PR
£27-
PR
£27-
PR
£27-
PR
£27-
PR
£27-
PR
Charge
Question
4
5
5
5
6
7
8
8
9
9
Comment Text Excerpt
(Note: Comment excerpts were edited to correct spelling and major grammatical and format issues [in
order to improve clarity], but generally were maintained in their original [as-submitted] form)
In addition to increasing the knowledge about MWCNTs or NM in general, there should be also
focus on better exploitation of existing data, reading across and the use of alternative methods
(in vitro/QSAR).
The document organization is logical although sometimes it is repetitive and could be shorter in
some parts.
The extended executive summary relating to the different chapters gives a clear overview on
the content of the document and helps to focus on specific areas.
In my opinion, within Chapter 4, a better separation between human and environmental
exposure scenarios and data would be helpful. This separation is achieved in Chapter 5, in
which the information on potential human health and ecological impacts of CNTs are well
divided in two subsections (5.1. Human Health Effects and 5.2. Ecological Effects).
Comments are transparently addressed.
Comments are transparently and clearly conveyed. The double presentation of original
comments and summarized comments make it very transparent.
Yes, comments were either addressed in the document or justifications were given where not.
The reader friendliness has improved by focusing on MWCNTs and moving DecaBDE to an
Annex. While it is appreciated that the main document makes reference to the Annex,
references from the Annex to the main document should be avoided, as it makes it quite
confusing to read.
In principle, the rationale is clear and gives a very nice overview.
Some steps are not 100% comprehensible (e.g., purple color for potential value):
Figure J-2: I suggest putting lipophilicity together with water solubility as they are usually
antipods. Lipophilicity would then also influence dispersion (add arrow).
Themes
Document scope/
content, Research
priorities, Alternative
methods
Praise for case study,
Reduce redundancy,
Document organization
Praise for case study,
Science is accurate &
clear, Executive
summary
Document organization,
Human & ecological
exposure scenarios
Praise for case study,
Peer& public
comments are clear
Praise for case study,
Peer& public
comments are clear
Praise for case study,
EPA was responsive to
comments & made
appropriate changes
DecaBDE comparison,
Cross-references,
Appendices
Praise for case study,
Knowledge Maps
Knowledge Maps,
Revise figure for clarity,
lipophilicity
K-56
-------
Table K-13 (Continued): Comments received in response to the expert charge questions.
Comment
ID
466
467
468
469
470
471
472
473
474
Author
ID
E27-
PR
E27-
PR
£27-
PR
£27-
PR
£27-
PR
£27-
PR
£27-
PR
£27-
PR
£27-
PR
Charge
Question
9
9
9
10
10
11
11
Specific
Observations
Specific
Observations
Comment Text Excerpt
(Note: Comment excerpts were edited to correct spelling and major grammatical and format issues [in
order to improve clarity], but generally were maintained in their original [as-submitted] form)
Some steps are not 100% comprehensible (e.g., purple color for potential value):
Figure J-2: The arrow from bioaccumulation should not just go to plants, but to biota as it could
also impact terrestrial animals (earthworms); an additional arrow should go to aquatic biota.
Figure J-3 and J-4: It is interesting to notice that for DecaBDE all movements are known, but
there are no strong movements, whereas for MWCNTs there is only one known movement
which is considered strong. Is evidence not sufficient to consider sediment as a known sink?
Figure J-6: The differences between the different MWCNT types are not clear and should be
better explained. Circle and Octagon are both MWCNT - if it is the same, why not use the
same symbol. If different, please explain. Bundled MWCNT - only by sorption?
The Maps - with the dotted line - in principle help to identify knowledge gaps.
However, for MWCNTs, almost all lines are dotted so is there a knowledge gap in general?
Maybe you could differentiate between suspected pathways (something is known) and those
where there is no knowledge at all.
I didn't have time to read this chapter in detail - there is not much text there - but it seems that
the most important issues are captured.
I would suggest moving the maps to Chapter 3 so that there can be a direct link/comparison.
Otherwise, the maps are isolated without explanation (see comment above).
Explain 'abiotic receptors' better.
In this section more information on quantification of carbon nanomaterials in vivo could be
given. In 2011, a review was published focusing on isotopic labeling and tracing methods.
Please see reference:
Wang, H., S.-T. Yang, et al. (2012). "Quantification of Carbon Nanomaterials in Vivo."
Accounts of Chemical Research 46(3): 750-760.
Themes
Knowledge Maps,
Revise figure for
accuracy,
Bioaccumulation
Knowledge Maps,
Clarify research gaps,
Sediment as a sink
Knowledge Maps,
Revise figure for clarity,
MWCNT types
Knowledge Maps
Knowledge Maps,
Clarity of knowledge
gaps, Future work
Knowledge Maps,
Science is accurate &
clear
Knowledge Maps,
Document organization,
Move KM s to main text
Revise text to clarify,
Scientific clarity, Abiotic
receptors
Consider additional
literature, Scientific
completeness, In vivo
quantification
K-57
-------
Table K-13 (Continued): Comments received in response to the expert charge questions.
Comment
ID
Author Charge
ID Question
Comment Text Excerpt
(Note: Comment excerpts were edited to correct spelling and major grammatical and format issues [in
order to improve clarity], but generally were maintained in their original [as-submitted] form)
Themes
475
E27- Specific
PR Observations
According to the review :
Morimoto, Y., M. Horie, et al. (2012). "Inhalation Toxicity Assessment of Carbon-Based
Nanoparticles." Ace Chem Res.
This value is 30ug/m3 and was calculated in rats exposed to SWCNTs from two studies.
According to the authors, this value also applies to MWCNTs. This result is reported in:
Nakanishi, J., Ed. Risk assessment of manufactured nanomaterials: Carbon nanotube (CNT).
Final report issued on August 12, 2011. NEDO project (P06041) "Research and development
of nanoparticle characterization methods", and the data was calculated from two animal
studies:
1. Morimoto, Y., M. Hirohashi, et al. (2012). "Pulmonary toxicity of well-dispersed multi-wall
carbon nanotubes following inhalation and intratracheal instillation." Nanotoxicology 6(6): 587-
599.
2. Kobayashi, N.; Naya, M.; Mizuno, K.; Yamamoto, K.; Ema, M.; Nakanishi, J. Pulmonary and
systemic responses of highly pure and well-dispersed single-wall carbon nanotubes after
intratracheal instillation in rats. Inhalation Toxicol. 2011, 23, 814-824.
Consider additional
literature, Scientific
accuracy, Reference
values
476
E27-
PR
Specific
Observations
In vitro data on dermal absorption is provided in this paper:
Vankoningsloo, S., J. P. Piret, et al. (2010). "Cytotoxicity of multi-walled carbon nanotubes in
three skin cellular models: effects of sonication, dispersive agents, and corneous layer of
reconstructed epidermis." Nanotoxicology 4(1): 84-97.
This reference can be also added in Appendix F in Table F-2.
Consider additional
literature, Scientific
completeness, Dermal
absorption
477
E27-
PR
Specific
Observations
Suggested reference for inhalation study of MWCNT by the wet aerosolization method:
Morimoto, Y., M. Hirohashi, et al. (2012). "Pulmonary toxicity of well-dispersed multi-wall
carbon nanotubes following inhalation and intratracheal instillation." Nanotoxicology 6(6): 587-
599.
In the inhalation study, rats were exposed to 0.37 mg/m(3) aerosols of well-dispersed
MWCNTs (>70% of MWCNTs were individual fibers) for 4 weeks, and were sacrificed at 3
days, 1 month, and 3 months after the end of exposure. The inhalation exposures delivered
less amounts of MWCNTs into the lungs, and therefore less pulmonary inflammation
responses was observed, as compared to intratracheal instillation. The results of this study
show that well-dispersed MWCNT can produce pulmonary lesions, including inflammation.
Consider additional
literature, Scientific
completeness,
Occupational exposure
478
E27- Specific For clarity Table I-3 could be moved previous to the introduction to 1.1.2. Appendix
PR Observations Organization. (In page I-4, after line 8)
Document organization,
Table 1.3
479
E27-
PR
Specific
Observations
Response category 1; Actions Taken: Moved unprioritized sections to Appendix.
To comply with this action, box of unprioritized research area in page 4-2 to Appendix.
Document organization,
Text boxes, Abiotic
receptors
K-58
-------
Table K-13 (Continued): Comments received in response to the expert charge questions.
Comment
ID
480
481
482
483
484
485
486
487
Author
ID
E27-
PR
E27-
PR
E27-
PR
E27-
PR
E27-
PR
E27-
PR
E27-
PR
E27-
PR
Charge
Question
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
Comment Text Excerpt
(Note: Comment excerpts were edited to correct spelling and major grammatical and format issues [in
order to improve clarity], but generally were maintained in their original [as-submitted] form)
Please note that an acute dermal (TG 402 GLP) study is available from the REACH
registration dossier
Add references to all X where there are results (see link above)
http://apps.echa.europa.eu/registered/data/dossiers/DISS-b281d1aO-c6d8-5dcf-e044-
00144f67d031/DISS-b281d1aO-c6d8-5dcf-e044-00144f67d031 DISS-b281d1aO-c6d8-5dcf-
e044-00144f67d031.html
Add Shvedova 2008: Am J Physiol Lung Cell Mol Physiol 295:1552-565 as reference for
SWCNT inflammatory effects - currently you only list MWCNT authors.
5-7: Line 17 ...although no subchronic inhalation studies were identified for systemic toxicity ...
The way it is presented it seems that there are no subchronic/chronic studies (by any route)
available (see also table 5-1). -the fact is that no systemic effects were observed/described
from available studies. Subchronic: 2 Inhalation studies (Pauluhn, Ma-Hock), Chronic
(injection): Sakamoto 2009 (52 W), Muller 2009 (2y);
Add Muller 2009 and Poland 2008 study on page 5-7 where you describe that certain type of
MWCNTs could behave in a manner similar to asbestos - as Muller/Poland showed that short
MWCNTs did not induce mesotheliomas.
Health Reference Values: suggest adding reference to Table 4-1 which is more
comprehensive and includes more values (e.g., the Pauluhn and NEDO value).
I did not find a description of how the Aschberger values where derived in 5.1.3. Maybe delete
this.
These studies are not a proof for lung carcinogenicity - but rather a proof that if they reach a
tissue from where they cannot easily be removed (mesothelium) they can be carcinogenic.
The main concern raised by these studies is that if MWCNT would reach the mesothelium they
would act like asbestos. In the lung there would be clearance and the reaction could be
different. The risk of lung carcinogenicity is probably similar to other NP. Therefore I suggest
rephrasing; you could say:
'Another target organ of MWCNT could be the lung' (but there is no study yet showing that).
You could write certain types of CNTs (long rigid) induced mesotheliomas, with certain
diameter ... - see text in 6-34 which I suggest to move to Chapter 5.
Themes
Consider additional
literature, Scientific
completeness, Table 5-
1 (dermal absorption)
Consider additional
literature, Scientific
completeness, SWCNT
reference
Revise text to clarify,
Scientific accuracy,
Systemic toxicity
Consider additional
literature, Scientific
accuracy,
Mesotheliomas
Revise text to clarify,
Add a cross-link
Revise text to clarify,
Scientific accuracy,
Health reference values
Revise text to clarify,
Scientific accuracy,
Carcinogenicity
Revise text to clarify,
Scientific accuracy,
Carcinogenicity
K-59
-------
Table K-13 (Continued): Comments received in response to the expert charge questions.
Comment
ID
488
489
490
491
492
493
Author
ID
E27-
PR
E27-
PR
£27-
PR
£27-
PR
£27-
PR
£27-
PR
Charge
Question
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
Comment Text Excerpt
(Note: Comment excerpts were edited to correct spelling and major grammatical and format issues [in
order to improve clarity], but generally were maintained in their original [as-submitted] form)
In addition:
Nagai, H., Y. Okazaki, et al. (201 1 ). "Diameter and rigidity of multiwalled carbon nanotubes are
critical factors in mesothelial injury and carcinogenesis." Proceedings of the National Academy
of Sciences of the United States of America 108(49): E1 330-1 338.
In the section 5.1.11 on Carcinogenicity, I suggest you reference a paper from 2011 in which a
relationship between diameter and rigidity of MWCNT as key factor in mesothelial injury and
carcinogenesis. The authors tested the potential carcinogenesis of different MWCNT after i.p.
injection into rats. Higher frequency of malignant mesothelioma was observed when animals
were treated CNT with higher diameter and length. Carcinogenicity was independent of fiber
suspension characteristics as CNT samples contained same number of dispersed fibres and
lacked agglomerates. Specific genetic alterations associated with mesotheliomas are reported
in this study.
Disagree that results from intratracheal instillation studies are qualitatively similar to inhalation
studies. Due to the bolus effect the clearance in the lung is overwhelmed - therefore not only
quantitatively but also qualitatively different effects (Ref: Oberdoester2010, J Intern Med 267:
89-105).
See also Morimoto 2012 study as listed above.
Morimoto, Y., M. Hirohashi, et al. (2012). "Pulmonary toxicity of well-dispersed multi-wall
carbon nanotubes following inhalation and intratracheal instillation." Nanotoxicology 6(6): 587-
599.
Inhalation studies: suggest not to list intratracheal instillation (+ intranasal injection) studies as
'inhalation' - maybe use: 'pulmonary exposure'
6.4.3.1, 6.3.4.2; It seems that in this part newer results are presented than in Chapter 5 -
should this not be moved there? For example, page 5.17
You cannot identify a NOEL or LOEL for consumers - you can only determine a reference
value (which considers different assessment factors) - please rephrase
Is there only data for occupational air and no data for environmental compartments as for
Deca-BDE?
Themes
Consider additional
literature, Scientific
completeness,
Carcinogenicity
Consider additional
literature, Scientific
accuracy, Instillation vs.
inhalation exposure
Revise text to clarify,
Scientific accuracy,
Pulmonary exposure
Document organization
Revise text to clarify,
Scientific accuracy, RTI
workshop
Document scope/
content, Scientific
completeness, MWCNT
environmental
concentrations
K-60
-------
Table K-13 (Continued): Comments received in response to the expert charge questions.
Comment
ID
Author Charge
ID Question
Comment Text Excerpt
(Note: Comment excerpts were edited to correct spelling and major grammatical and format issues [in
order to improve clarity], but generally were maintained in their original [as-submitted] form)
Themes
494
E28-
PR
In general, I have found that the authors and the external reviewers have done an admirable
job of summarizing the latest research findings concerning the state of the science regarding
MWCNT fate, exposure, and effects. In particular, while there is broad consensus for certain
properties and behaviors of MWCNTs in the environment (e.g., their tendency to aggregate
and to associate with particulates in aqueous systems), others are more controversial (e.g.,
potential for bioaccumulation or ability to cross cell membranes). These areas of respective
clarity or lack thereof are very important to identify in a case study describing materials for
which fate/transport models based on physicochemical properties are of limited relevance. I
believe that this aspect was well described in the case study.
Praise for case study,
Science is accurate &
clear
495
E28- 1
PR
I do feel that the entire case study suffers from a relatively narrow focus on MWCNT. In many
cases, SWCNT may behave in a very similar fashion to MWCNT in environmental and
biological systems, and I feel that the case study would have benefited from inclusion of the
relevant research (e.g., fate and transport, analytical methods, exposure predictions) related to
SWCNT also. As noted in the document, SWCNT may also be used in textile applications and
will therefore experience many of the same routes-of-entry to the environment. Risk
assessment of the (likely) more plentiful MWCNT may be informed by research that is relevant
to BOTH classes of CNT, including the (likely) lower production-volume SWCNT.
Document scope/
content, Scientific
completeness, SWCNT
reference
496
E28-
PR
The priority areas are very clearly indicated in the document, starting with Figure 1-3 and
continuing with highlighted boxes within the main body of the text.
Praise for case study,
Research priorities are
clear, Figure 1-3
497
E28-
PR
In particular, I find the "color-coding" of the boxes within the document easy to follow and
helpful.
Praise for case study,
Document organization,
Text boxes
498
E28-
PR
After carefully reviewing the list of references, the descriptions of previous studies and their
conclusions, and performing independent literature searches, I believe that the case study has
accurately represented the most relevant research findings that can provide guidance for
future research supporting risk management of MWCNT. I do not know of any MWCNT-
specific studies that have been omitted from the case study.
Praise for case study,
Science is accurate &
clear, Scientific
completeness
499
E28-
PR
However, as I note above, I believe that the authors and external reviewers missed an
opportunity by explicitly excluding environmental fate, transport, and effects data for SWCNT
(especially with regard to aquatic fate and mechanisms of human/aquatic toxicity). SWCNT
may be expected to behave in very similar ways to MWCNT in the environment (based on
similarity of their physicochemical properties) and thus a more complete CEA might have been
conducted on "CNT in textiles", irrespective of their intended function in the textile (e.g., flame-
retardant vs. stain repellent). In fact, this approach would be quite defensible in light of the fact
(as indicated in the box on page 1-28) that MWCNT use as flame retardant coatings in textiles
is not yet widespread and is not anticipated to increase dramatically in the future.
Document scope/
content, Scientific
completeness, SWCNT
reference
K-61
-------
Table K-13 (Continued): Comments received in response to the expert charge questions.
Comment
ID
500
501
502
503
504
505
506
507
508
Author
ID
E28-
PR
E28-
PR
E28-
PR
E28-
PR
E28-
PR
E28-
PR
E28-
PR
E28-
PR
E28-
PR
Charge
Question
4
4
4
5
5
5
6
6
7
Comment Text Excerpt
(Note: Comment excerpts were edited to correct spelling and major grammatical and format issues [in
order to improve clarity], but generally were maintained in their original [as-submitted] form)
With the exception of some discontinuities in research priorities (e.g., the identification of
"ecological impacts of MWCNT in aquatic biota" as a priority area, while classifying both
exposure and dose of MWCNT in ecological populations as lower priority), I found that the
document was generally very well structured to guide future research in the area of
environmental and health implications of MWCNT.
One admirable aspect of this case study is that the environmental and (especially aquatic)
exposure/transport/effects characteristics discussed for MWCNT in Chapters 3-5 are relevant
for more than just the potential release from use in flame-retardants on textiles. Therefore, the
predictions, prioritization, and implications will be valid for MWCNT releases to the
environment under a variety of scenarios.
Conversely, I find that the weakest part of the document is Chapter 2 (Product Life Cycle),
since much of it is focused exclusively on the use of MWCNT in the very narrow application of
flame retardants in textiles.
I've reviewed both the Draft for External Review and the current (revised) Case Study, and I
find that the use of information boxes, DecaBDE/MWCNT-comparison boxes, and priority-area
boxes throughout the text has greatly improved the logical flow of the document.
I do, however, believe that Section 1.1 is unnecessarily verbose and might be better suited to
an appendix.
Also, Appendix J (Knowledge Maps) - if used at all - would be much more effective as part of
the main body of text, perhaps in Chapter 6.
Yes - I have almost no comments on this aspect of the case study. Appendix I is somewhat
difficult to follow, but it is well-organized and cross-referenced among reviewer comments,
agency responses, and document edits/amendments.
In general, the revisions were made in a very thoughtful and systematic manner.
As above - yes, the responses to comments are also very well-described in Appendix I. I have
no specific suggestions for improvement.
Themes
Praise for case study,
Supports research
planning
Praise for case study,
Supports research
planning, Information
applicable to multiple
MWCNTs applications
Document scope/
content, Scientific
completeness, Focus
on MWCNT as FR in
textiles
Praise for case study,
Document organization,
Text boxes
Document length,
Document organization,
Chapter 1 content
Knowledge Maps,
Document organization,
Move KMs to main text
Praise for case study,
Peer& public
comments are clear
Praise for case study,
EPA was responsive to
comments & made
appropriate changes
Praise for case study,
EPA was responsive to
comments & made
appropriate changes
K-62
-------
Table K-13 (Continued): Comments received in response to the expert charge questions.
Comment
ID
509
510
511
512
513
514
515
516
Author
ID
E28-
PR
E28-
PR
E28-
PR
E28-
PR
E28-
PR
E28-
PR
E28-
PR
E28-
PR
Charge
Question
8
8
8
9
10
10
10
11
Comment Text Excerpt
(Note: Comment excerpts were edited to correct spelling and major grammatical and format issues [in
order to improve clarity], but generally were maintained in their original [as-submitted] form)
Comparison of the case study document in its present form to the draft for external review,
together with a careful read of the stakeholder comments, indicates that EPA was very
responsive to input. It is evident that each particular comment was addressed and I have found
many cases in which the document was changed accordingly.
The two most significant changes made were to relocate the bulk of the DecaBDE discussion
to an appendix and the incorporation of informational and comparative "text boxes" in the main
document to highlight priority areas and identify areas in which relevant comparisons could be
made between MWCNT and DecaBDE properties and behavior. I believe that these were
appropriate and very useful changes that have improved the document considerably.
The only exception is that I do not agree fully with the EPA's approach in "Response Category
2", wherein new information or references brought to light during the external review were
added as separate, highlighted (in blue) information text boxes. I believe that the new
references and/or information should have been more seamlessly integrated with the existing
main text body. There may be cases where the information would have been best presented in
such "call-outs", but I would estimate that such cases would be the exception rather than the
rule, so most newly identified literature should have been added directly to the text. I will qualify
this statement by saying that it was somewhat difficult for me to track all of these additions
from comment to response to information box, so I'm unable to identify specific areas where
this could be corrected.
I do believe that the rationale for the pilot project is well-described. I have no problems with
that aspect of Appendix J. My more substantive comments are detailed below.
Overall, I think that the knowledge maps are useful for conveying general processes and
properties of MWCNT and DecaBDE relevant to the CEA as implemented.
However, I believe that these would be MUCH more useful if the modules and connectors were
made more quantitative. It should be possible to estimate (from literature) magnitudes of rates,
concentrations, and volumes relevant to the processes of transport, transformation, and
physicochemical apportioning in the environment for BOTH the MWCNT and DecaBDEs. This
would make the comparisons between the two candidate flame retardant materials much more
robust.
Also, if they are to be used, these knowledge maps should be placed in Chapter 3.
The three classes of knowledge maps do a reasonable job of summarizing (visualizing) the
connections among processes and the overall information conveyed in the case study, but
again - they should be made quantitative if they are to be of much use.
Themes
Praise for case study,
EPA was responsive to
comments & made
appropriate changes
Praise for case study,
EPA was responsive to
comments & made
appropriate changes,
Text boxes
Document organization,
Reviewer doesn't fully
agree with EPA
response, Text boxes
Knowledge Maps,
Rationale for KMs is
clear
Knowledge Maps
Knowledge Maps,
Revise figure for clarity,
Add quantitative
information to KMs
Knowledge Maps,
Document organization,
Move KMs to main text
Knowledge Maps, Add
quantitative information
to KMs
K-63
-------
Table K-13 (Continued): Comments received in response to the expert charge questions.
Comment
ID
517
518
519
520
521
522
523
524
Author
ID
E28-
PR
E28-
PR
E28-
PR
E28-
PR
E28-
PR
E28-
PR
E28-
PR
E28-
PR
Charge
Question
11
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
Comment Text Excerpt
(Note: Comment excerpts were edited to correct spelling and major grammatical and format issues [in
order to improve clarity], but generally were maintained in their original [as-submitted] form)
Also, in Figure J-2, "water solubility" with or without surfactants is not an appropriate term for
MWCNT. A more accurate term would be dispersibility.
There should be an entry in this table for MWCNT illustrating mechanism of flame retardancy.
Several inconsistencies or errors of usage appear in this table. First - it is misleading to state
the purity of MWCNT as "usually > 90%". This is much more variable (shown many times in the
literature). I suggest a broader and more defensible range be presented.
Also, Chirality is truly meaningless for MWCNT - these properties apply only to individual
"layers" of the CNTs or to SWCNT specifically.
Finally, citations of "Sigma-Aldrich (2012)" for melting-point, density, etc. are inappropriate.
Peer-reviewed literature should be cited for these parameters, not product data-sheets.
Based on the known behavior of MWCNT in nearly all media, it is very unlikely that any
scenario would result in release of truly "free" MWCNT to the environment. These materials will
most likely always exist in homo- or heteroaggregate form.
Also, based on information in "Additional Information Box 3", I would consider that release of
MIXED chemical/MWCNT flame retardant materials from textiles would be of great concern!
This is not mentioned in this section but should be addressed.
A reference is quoted stating that most CNTs in wastewater come from tailoring, finishing, use,
and degradation of textiles. This is completely speculative and there is no direct evidence of
this from analytical studies (even occurrence in wastewater of CNTs is not well-founded based
on solid analytical data). This statement should be removed.
Themes
Knowledge Maps,
Scientific accuracy,
Revise figures
Revise text to clarify,
Scientific clarity, Table
1-5
Revise text to clarify,
Scientific accuracy,
MWCNT purity
Revise text to clarify,
Scientific accuracy,
MWCNT chirality
Revise text to clarify,
Scientific accuracy,
MWCNT
physicochemical
properties
Revise text to clarify,
Scientific accuracy,
MWCNT release from
polymers
Revise text to clarify,
Scientific accuracy,
MWCNT release from
polymers
Revise text to clarify,
Scientific accuracy,
MWCNT release to
wastewater
K-64
-------
Table K-13 (Continued): Comments received in response to the expert charge questions.
Comment
ID
525
526
527
528
529
530
531
532
Author
ID
E28-
PR
E28-
PR
E28-
PR
E28-
PR
E28-
PR
E28-
PR
E28-
PR
E28-
PR
Charge
Question
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
Comment Text Excerpt
(Note: Comment excerpts were edited to correct spelling and major grammatical and format issues [in
order to improve clarity], but generally were maintained in their original [as-submitted] form)
I find these tables very uninformative. They are quite qualitative and in some cases are
completely speculative. The overall impression given to the reader is that nearly all processes
that apply to release of DecaBDE from textiles may also apply to MWCNT. While this might be
true in some cases, I believe that many of the scenarios and processes highlighted in these
comparative tables are far too simplistic. There are many physicochemical differences between
MWCNT and DecaBDE that would cause behavior in these release processes to differ
QUANTITATIVELY, which is very important. Consider removing or qualifying these tables, or
where possible: provide quantitative estimates of rates or volumes (as appropriate).
The term solubility is inappropriate when discussing nanoparticles like MWCNT in solution.
They do not dissolve in the sense of molecules. This should be changed.
This figure omits the important phenomenon of hydrophobic organic contaminant and metal
adsorption to MWCNT - this is a relevant "modification" that could affect fate/transport. It is
discussed in the text but should be added to the figure.
Discussion of "partitioning" of MWCNT to sediment is not thermodynamically correct. These
are not molecules and so their association with particulate matter (or organic phases) is not
governed by equilibrium partitioning theory (molecular-based). Instead, the association is
governed by colloidal homo- and heteroaggregation theory (e.g., DLVO), such that attachment
to particles may be irreversible or at least highly hysteretic. This section should be amended to
reflect this critical difference between nanoparticulate contaminants such as MWCNT and
molecular species such as DecaBDE.
I don't see any reason for this discussion to be "called-out" in an information box. It should be
included within the main text body (this is true for many of the blue-outlined "information
boxes").
The information in Information Box #10 should be merged into Text Box 4-1. The discussion,
information, and references are related to essentially the same general idea and topic:
Analytical methods for MWCNT are immature.
Within the combined text box, it should be stated EXPLICITLY that nearly all currently available
methods are unsuitable for detection, characterization, and quantitation of MWCNT at ppb (or
lower) concentrations in complex but important media such as soil or sediment.
The identification of CNT in the lung tissues of WTC victims was equivocal and I don't think this
statement should be so strong.
Themes
Revise text to clarify,
Scientific clarity,
MWCNT release from
polymers
Revise text to clarify,
Scientific accuracy,
MWCNT solubility
Revise text to clarify,
Scientific clarity, Revise
Figures
Revise text to clarify,
Scientific accuracy,
MWCNT association
with particulate matter
Document organization,
Text boxes, integrate
text box into main text
Document organization,
Text boxes, MWCNT
analytical methods
Revise text to clarify,
Scientific clarity,
MWCNT analytical
methods
Check accuracy of
statement, Scientific
accuracy, Exposure
K-65
-------
Table K-13 (Continued): Comments received in response to the expert charge questions.
Comment
ID
533
534
535
536
538
539
540
Author
ID
E28-
PR
E28-
PR
E28-
PR
E28-
PR
E27-
PR
E27-
PR
E27-
PR
Charge
Question
Specific
Observations
Specific
Observations
Specific
Observations
Specific
Observations
4
4
Specific
Observations
Comment Text Excerpt
(Note: Comment excerpts were edited to correct spelling and major grammatical and format issues [in
order to improve clarity], but generally were maintained in their original [as-submitted] form)
I'm not convinced that it is necessarily useful to draw parallels between reproductive toxic
effects of DecaBDE and MWCNT, since these are going to have VERY different mechanisms.
This information should be included in the main text (as mentioned before for other similar
boxes).
The word "degradation" is spelled incorrectly in the figures.
These sentences are redundant (repetitive).
The consideration of other impacts such as energy, climate change, economics etc., is
appreciated.
As mentioned above, it should be considered that MWCNT alone will not be sufficient to
replace DecaBDE.
Footnote 16: not clear what is meant to express: not relevant for MWCNT or for inhalation
exposure? - only the first sounds logic
Themes
DecaBDE comparison,
Scientific clarity,
Reproductive effects
Document organization,
Text boxes, Aquatic
receptors
Copy edit, Knowledge
Maps, Revise figures
Copy edit, Reduce
redundancy
Praise for case study,
Document scope/
content, Climate change
Document scope/
content, DecaBDE
comparison, MWCNTs
flame-retardants
Revise text to clarify,
Footnote 16
Appendix K References
Versar. (2013). External peer review of EPA's letter peer review draft report: Nanomaterial case study: A comparison of multiwalled carbon nanotube and
decabromodiphenyl ether flame-retardant coatings applied to upholstery textiles. (Contract No. EP-C-12-045, Task Order 10). Springfield, VA.
K-66
-------
United States
Environmental Protection
Agency
National Center for Environmental Assessment
Office of Research and Development
Washington, DC 20460
------- |