&EPA
United States
Environmental Protection
Agency
                                         March 2013
                                    EPA/600/R-12/043B
               Nanomaterial Case Study:
  A Comparison of Multiwalled Carbon Nanotube
  and Decabromodiphenyl Ether Flame-Retardant
      Coatings Applied to Upholstery Textiles
                    (Draft for Peer Review)
                          March 2013
                           NOTICE
This document is a Peer Review Draft. This information is distributed solely for the purpose of
pre-dissemination peer review under applicable information quality guidelines. It has not been
formally disseminated by EPA. It does not represent and should not be construed to represent
any Agency determination or policy. It is being circulated for review of its technical accuracy
and science policy implications.
                 U.S. Environmental Protection Agency
                         Washington, DC

-------
                               Disclaimer
This document is distributed solely for the purpose of pre-dissemination peer review under
applicable information quality guidelines. It has not been formally disseminated by EPA. It does
not represent and should not be construed to represent any Agency determination or policy.
Mention of trade names or commercial products does not constitute endorsement or
recommendation for use.
March 2013                                 i                   DRAFT-DO NOT CITE OR QUOTE

-------
                          Table of  Contents
Table of Contents	ii

List of Tables	vii

List of Figures	xi

List of Text Boxes	xii

List of Additional Information Highlight Boxes	xiii

Key Terms	xiv

Abbreviations	xv

Authors, Contributors, and Reviewers	xix

Preface   	xxi

Executive Summary	xxiii

Chapter 1. Introduction to this Document	1-1
        1.1. Background	1-1
           1.1.1. Introduction to Comprehensive Environmental Assessment	1-3
           1.1.2. Purpose of this Document	1 -7
           1.1.3. How the CEA Framework and Process Were Applied	1 -8
           1.1.4. Selection of DecaBDE for Comparison	1-13
        1.2. Introduction to Flame Retardants in Textiles	1-15
           1.2.1. Standards for Textiles	1-16
           1.2.2. Flame-Retardant Materials as Solutions to Flammability	1-17
        1.3. DecaBDE and MWCNTs in Flame-Retardant Textiles	1-22
           1.3.1. Introduction to DecaBDE	1-22
           1.3.2. Introduction to MWCNTs	1-25
           1.3.3. MWCNTs as Alternative Flame-Retardant Materials in Upholstery Textiles	1-29

Chapter 2. Product Life Cycle	2-1
        2.1. Feedstocks	2-5
        2.2. Manufacturing	2-5
           2.2.1. Research and Development	2-5
           2.2.2. Material Synthesis	2-7
           2.2.3. Material Processing	2-12
           2.2.4. Product Manufacturing	2-15


March 2013                                    ii                    DRAFT-DO NOT CITE OR QUOTE

-------
         2.3. Storage and Distribution	2-20
         2.4. Use	2-20
            2.4.1. Life-Cycle Processes	2-21
            2.4.2. Potential Releases during the Use Stage	2-22
         2.5. Reuse, Recycling, and End of Life	2-24
            2.5.1. Reuse and Recycling	2-26
            2.5.2. Incineration	2-27
            2.5.3. Land-Filling	2-28
            2.5.4. Wastewater Treatment Plants	2-29

Chapter 3. 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	3-5
         3.3. Transport, Transformation, and Fate in Water and Sediment	3-7
            3.3.1. Surface Water and Sediment (Inland and Coastal)	3-7
            3.3.2. Ground Water	3-10
            3.3.3. Wastewater	3-11
         3.4. Transport, Transformation, and Fate in Soil	3-12
         3.5. Multimedia Models to Predict Environmental Fate and Transport	3-13

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. Absorption, Distribution, Metabolism and Excretion  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
March 2013                                      iii                     DRAFT-DO NOT CITE OR QUOTE

-------
            5.1.1. Quantitative Toxicity Assessment	5-8
            5.1.2. Systemic Toxicity	5-9
            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	5-18
            5.2.1. Aquatic Receptors	5-19
            5.2.2. Terrestrial Receptors	5-25
         5.3. Other Impacts	5-26
            5.3.1. Environmental Justice	5-27
            5.3.2. Energy Demand and Natural Resource Depletion	5-27
            5.3.3. Climate Change	5-31
            5.3.4. Economics	5-32

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	6-18
            6.3.3. Exposure Route and Dose (Kinetics)	6-25
            6.3.4. Impacts	6-34
         6.4. Moving From "Assessment" to "Management" in the CEA Process	6-42
         Case Study 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
March 2013                                       iv                     DRAFT-DO NOT CITE OR QUOTE

-------
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-8
Appendix C. Efficacy of Various Methods of Carbon Nanotube Purification	C-1
         Appendix C References	C-3
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-16
         Appendix E References	E-20
Appendix F. Toxicological and Ecological Effects	F-1
         F.1. Toxicological Effects	F-2
         F.2. Ecological Effects	F-31
         Appendix F References	F-50
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-5
         G.4. Exposure-Dose	G-7
         G.5. Potential Human Health, Ecological, and Other Impacts	G-12
         Appendix G References	G-16
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-59
March 2013                                     v                    DRAFT-DO NOT CITE OR QUOTE

-------
        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-7
        1.3. Full Comment Excerpts	1-31
        Appendix I References	I-74
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-6
        J.5. Transformation Map	J-9
        Appendix J References	J-13
March 2013                                     vi                    DRAFT-DO NOT CITE OR QUOTE

-------
                                     List of Tables
   Table 1-1.    Existing state regulatory initiatives for decaBDE	1-14
   Table 1-2.    Common textile fibers and degrees of flammability	1-16
   Table 1-3.    U.S. and international fire regulations for upholstery textiles.1	1-17
   Table 1-4.    Durability classifications of flame-retardant finishes	1-19
   Table 1-5.    Flame retardants summarized by chemical class, method of application, and mechanism of flame-
                   retardant action	1-22
   Table 1-6.    Major PBDE congeners	1-23
   Table 1-7.    Commercial formulations of PBDEs used as flame retardants	1-24
   Table 1-8.    Physical properties and chemical identity of decaBDE	1-25
   Table 1-9.    Physical properties and chemical identity of MWCNTs.1'2	1-27
   Table 1-10.   Overview of decaBDE and MWCNTs for flame-retardant textile application	1-30
   Table 1-11.   General qualitative comparisons of performance criteria for decaBDE and MWCNTs	1-31
   Table 1-12.   Flame test performance of decaBDE and MWCNTs	1-32
   Table 1-13.   Phy sicochemical properties of MWCNTs related to flame-retardant performance	1-33
   Table 2-1.    Summary of common CNT synthesis methods.1	2-8
   Table 2-2.    Current scale and projected growth in the CNT industry	2-9
   Table 2-3.    Potential  release scenarios during material synthesis	2-11
   Table 2-4.    Examples of functionalization of MWCNTs	2-14
   Table 2-5.    Potential  release scenarios during material processing of MWCNTs	2-15
   Table 2-6.    Potential  release scenarios during product manufacturing	2-19
   Table 2-7.    Potential  release scenarios during product use	2-24
   Table 2-8.    Potential  release scenarios during reuse and recycling	2-27
   Table 2-9.    Potential  release scenarios during incineration	2-28
   Table 2-10.   Potential release scenarios during land-filling	2-29
   Table 2-11.   Potential release scenarios during wastewater treatment	2-31
   Table 3 -1.    Summary of physicochemical properties that affect partitioning and fate of nanomaterials such as
                   multiwalled carbon nanotubes  (MWCNTs)	3-2
   Table 3-2.    Predicted environmental concentrations of CNTs using fate and transport modeling	3-15
   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.
                    	5-5
   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	5-24
   Table 5-4.    Estimated minimum energy requirements and process rates for synthesis of CNTs	5-28
March 2013
                                                  VII
DRAFT-DO NOT CITE OR QUOTE

-------
   Table 5-5.
   Table 5-6.
   Table 6-1.
   Table 6-2.
   Table 6-3.
   Table 6-4.
   Table 6-5.
   Table 6-6.
   Table 6-7.
   Table 6-8.
   Table 6-9.

   Table 6-10.
   Table 6-11.

   Table 6-12.

   Table 6-13.

   Table 6-14.
   Table A-1.

   Table A-2.
   Table B-l.
   Table B-2.
   Table C-l.
   Table D-l.
   Table D-2.
   Table D-3.
   Table D-4.
   Table E-l.
   Table E-2.
   Table E-3.
   Table E-4.
   Table E-5.
   Table E-6.
   Table E-7.
   Table E-8.
   Table F-l.
 Environmental assessment of production of 1 kilogram (kg) of carbon nanofibers.1	5-29
 Environmental assessment of production of one SWCNT polymer mesh	5-30
 Examples of risk management decisions	6-5
 Example questions for problem formulation	6-6
 Influential Factors options for all areas	6-8
 Research Identified by RTI Workshop Participants: MWCNT Material Processing	6-12
 Research Identified by RTI Workshop Participants: MWCNT Product Manufacturing	6-14
 Research Identified by RTI Workshop Participants: MWCNT Product Use	6-16
 Research Identified by RTI Workshop Participants: MWCNT Disposal / Recyling	6-18
 Research Identified by RTI Workshop Participants: MWCNT Mobility & Persistence in Air	6-20
 Research Identified by RTI Workshop Participants: MWCNT Mobility & Persistence in Waste Water.
   	6-23
Research Identified by RTI Workshop Participants: MWCNT Persistence in Sediment	6-25
Research Identified by RTI Workshop Participants: MWCNT Occupational Exposure via Inhalation. .
   	6-28
Research Identified by RTI Workshop Participants: Human Absorption, Metabolism
   and Excretion of MWCNTs	6-33
Research Identified by RTI Workshop Participants: Non-Cancer Human Health Impacts of MWCNTs.
   	6-35
Research Identified by RTI Workshop Participants: Societial Impacts of MWCNTs	6-42
 Nanomaterial application candidates and non-nano-enabled products for comparison in a case study.
    	A-5
 Factors for consideration in selecting a candidate for case study	A-6
Analytical techniques for detecting, measuring, and characterizing PBDEs	B-3
 Analytical techniques for detecting, measuring, and characterizing MWCNTs	B-4
 Purification methods for carbon nanotubes	C-2
Relevant studies of transformation (debromination) of BDE-209	D-l
Relevant studies of MWCNTs in aqueous media	D-4
Relevant studies of BDE-209 in soils and plants	D-7
Relevant studies of carbon nanotubes (CNTs)  in soils	D-8
BDE-209 concentrations in building dust	E-l
BDE-209 air concentrations in outdoor and  indoor air	E-3
BDE-209 concentrations in aquatic systems	E-5
BDE-209 concentrations in sewage effluent and sludge	E-8
BDE-209 concentration data in soil	E-9
Proxy data for estimating MWCNT concentrations in occupational air	E-ll
Measured concentrations of PBDEs in biota	E-16
Mean concentration of PBDEs in media/biota  in an aquatic ecosystem	E-19
 Select toxicokinetic studies for decaBDE	F-2
March 2013
                                                  VIII
                                                          DRAFT-DO NOT CITE OR QUOTE

-------
   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 inhalation studies for decaBDE	F-13
   Table F-6.    Select inhalation 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 MWCNT	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-34
   Table F-14.   Effects of exposure to MWCNTs in algae, macrophytes, and aquatic macroinvertebrates	F-38
   Table F-15.   Effects of exposure to MWCNTs via water on Ceriodaphnia dubia	F-39
   Table F-16.   Effects of exposure to MWCNTs on zebrafish and medaka embryos	F-40
   Table F-17.   Immune responses in rainbow trout (Oncorhynchus mykiss) head kidney cells following MWCNT
                   exposure	F-42
   Table F-18.   Effects of exposure to decaBDE in soil microbes, terrestrial invertebrates, and plants	F-43
   Table F-19.   Effects of exposure to MWCNTs in bacteria	F-45
   Table F-20.   Effects of exposure to MWCNTs on plants	F-49
   Table G-1.    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-12
   Table G-4.    Effects of decaBDE and MWCNTs on plants	G-14
   Table H-l.    Maj or PBDE 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-51
   Table H-6.    Biomagnification factors of select  PBDE congeners in an aquatic ecosystem	H-56
   Table 1-1.    Expert affiliations and area of expertise	1-3
   Table 1-2.    Public Commentators 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	1-7
   Table 1-5.    Relevant to Chapter 1 (including Preface and Executive Summary)	1-14
   Table 1-6.    Relevant to Chapter 2	1-17
March 2013
                                                   IX
DRAFT-DO NOT CITE OR QUOTE

-------
   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	1-53
   Table 1-14.   Comments received in response to the public comment period	1-64
March 2013
DRAFT-DO NOT CITE OR QUOTE

-------
                             List of Figures
Figure 1-1. Comprehensive environmental assessment framework	1-4
Figure 1-2. Comprehensive environmental assessment process	1-6
Figure 1-3. Detailed CEA framework used for the collective judgment prioritization process	1-12
Figure 1-4. Durability of additive flame retardants	1-20
Figure 1-5. The combustion process	1-21
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.1	2-4
Figure 3-1. Variability in MWCNT chemistry and implications in terms of life cycle, exposure, and risk.b
       	3-4
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: Decabromodiphenyl ether	J-4
Figure J-2. Physicochemical properties map: Multiwalled carbon nanotubes	J-5
Figure J-3. Transport map: Decabromodiphenyl ether	J-7
Figure J-4. Transport map: Multiwalled carbon nanotubes	J-8
Figure J-5. Transformation map: Decabromodiphenyl ether	J-ll
Figure J-6. Transformation map: Multiwalled carbon nanotubes	J-12
March 2013                                        xi                      DRAFT-DO NOT CITE OR QUOTE

-------
                       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	1-26
Text Box 4-1. Detecting, Measuring, and Characterizing MWCNTs	4-3
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
March 2013                                     xii                    DRAFT-DO NOT CITE OR QUOTE

-------
           List  of Additional Information

                               Highlight Boxes
1. Factors influencing flame retardant selection	1-18
2. MWCNTs are not widely used in flame-retardant textiles	1-28
3. MWCNTs are likely used in combination with other chemicals in flame-retardant applications	1-32
4. MWCNT release from various product matrices informs predictions of release from textiles	2-19
5. Impact of MWCNT release into wastewater treatment plants	2-30
6. Transformation throughout the product life cycle	3-2
7. Properties of the MWCNT formulation impact environmental release and transformation	3-3
8. Multimedia modeling of MWCNT environmental transport	3-16
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	5-7
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 non-cancer impacts of MWCNTs	5-18
17. Toxicity to benthic invertebrates	5-22
Gl. Uptake and Absorption in Aquatic Foodwebs	G-9
G2. Toxicity to Terrestrial Invertebrates	G-13
March 2013                                   xiii                   DRAFT-DO NOT CITE OR QUOTE

-------
                  Key Terms
Upholstery textiles
Flame retardant
DecaBDE
BDE-209
Multiwalled carbon
nanotube
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
March 2013
                          XIV
DRAFT-DO NOT CITE OR QUOTE

-------
              Abbreviations  and  Acronyms
A549
ACC
ACGIH

ADME
Al
APA
APR I
ASTM

ATSDR

B6C3Fi
BAF(s)
BAF/BCF(s)

BALB/c
BCF(s)
BDE
BDE-209
BDL
BfFV
BFR(s)
BHI
BMF
Br
BSA
BSEF

BSI
bw
C
°C
«C
C57BL/6
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
British Standards Institution
body weight
Carbon
degrees in Celsius
radiolabeled carbon
mouse strain
CA TB                 California (Bureau of Home
                      Furnishings and Thermal Insulation)
                      Technical Bulletin
Ca                    Calcium
Cal/EPA                California EPA
CalRecyle              California Department of Resources
                      Recycling and Recovery
Ce                    Cesium
CEA                   Comprehensive Environmental
                      Assessment
CFR                   Code of Federal Regulations
CINAHL                Cumulative Index to Nursing and Allied
                      Health Literature
Cl                     Chlorine
CLF                   Conservation Law Foundation
cm3                   cubic centimeters
CNF(s)                 carbon nanofiber(s)
CNQ                   could not quantify
CNT(s)                 carbon nanotube(s)
Co                    Cobalt
C02                   carbon dioxide
COOH-MWCNT          carboxylated MWCNT
CPC                   condensation particle counters
CPTC                 Consumer Product Testing Company
CVD                   chemical vapor deposition
decaBDE               decabrominated diphenyl ether
Dl                     deionized (water)
diBDE                 dibrominated diphenyl ether
DIN                   Deutsches Institut fur Normung
                      (Germany)
DLS                   dynamic light scattering
DNA                   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
March 2013
                                                           xv
                                                                 DRAFT-DO NOT CITE OR QUOTE

-------
ECNI
ECNI-MS

EEA
EEB
EEC
EI99
ENM(s)
EPA
EROD

E-RRF

EU
F344
FBCVD

Fe
FLE
FLM
FMVSS

FTIR

g, mg, |jg, ng, pg; kg

GC
GC/HRTOFMS

GD
GHG
GLP
GLRI
GPIIb/llla
gp'
GSI
GSRI
GWERD

heptaBDE
hexaBDE
HHPC-6
HiPCO®

HMVEC
hpf
electron capture negative ionization
electron capture negative ionization-
mass spectrometry
electron capture negative ionization
European Environmental Bureau
European Economic Community
Eco Indicator 1999 (method)
engineered nanoscale material(s)
U.S.  Environmental Protection Agency
ethoxyresorunfin-0-deethylase
enzyme
Element / Risk-Relevance-Factor (Pair
in CEA Framework)
European Union
rat strain
fluidized bed chemical vapor
deposition
iron
forelimb emergence
fluorescence microscopy
Federal Motor Vehicle Safety
Standards
Fourier transform infrared
spectroscopy
gram, milligram, microgram,
nanogram, picogram; kilogram
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
guaninephosphoribosyl-transferase
gonadosomatic index
Gulf South Research Institute
Ground Water and Ecosystems
Restoration Division of NRMRL
heptabrominated diphenyl ether
hexabrominated diphenyl ether
Hand-held airborne particle counter
a high pressure carbon monoxide
synthesis process
human microvascular endothelial cells
hours post fertilization
HR
HRMS
HSDB
IARC

ICF

ICL
ICP-MS

ICR
ID
IF
INEL
10
I PCS

IPEN

IRDC

IRIS

ISO

kg
Koc

Kow
L
L, mL
LC
LC/MS-MS

LCso
LCA
LDso
LDH
LOAEL
LOEC
LOEL
LOI
LRT
LSRI
M
M, mm, pm
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
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
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 106
meter, millimeter, micrometer
March 2013
                                                                  XVI
                                                                          DRAFT-DO NOT CITE OR QUOTE

-------
m2
m3
Mo
MRL
mRNA
MS
MWCNT(s)
MWCNT-OH
MWCNT-NH2
MWCNT-NHs*
MWNT
MWNT-COOH
n
NA
Na
NaCI
Nanomaterials
nCeo
NCC
NCCT

NCEA

NCSL(2011)

ND
NERL

NF
NFPA
NH2-MWCNT
IW-MWCNT
NHDF
Ni
NICNAS

NIEHS

NIH
NIOSH

NLM
square meters
cubic meters
magnesium
micronucleus
molybdenum
minimal risk level
messenger RNA
mass spectrometer
multiwalled carbon nanotube(s)
hydroxylated MWCNT
amine-functionalized MWCNT
ammonium-functionalized MWNT(s)
multiwalled nanotube(s)
carboxylated MWCNT
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)
National Institute of Health
National Institute for Occupational
Safety and Health
National Library of Medicine (NIH)
NMAM                  NIOSH Manual of Analytical Methods
NMRI                   mouse strain
NOAEL                  no-observed-adverse-effect level
NOE                    no observed effect
NOEC                   no observed effect concentration
NOEL                   no-observed-effect level
NOM                    natural organic matter
nonaBDE                nonabrominated diphenyl ether
NR                     not reported
NRC                    National Research Council
NRMRL                  National Risk Management Research
                        Laboratory (U.S. EPA, ORD)
NRMRL/GWERD          Ground Water and Ecosystems
                        Restoration Division of NRMRL
NSTC                   National Science and Technology
                        Council
NTP                    National Toxicology  Program
                        (NIEHS/NIH)
0                       oxygen
OAF                    overall assessment factor
OCSPP                  Office of Chemical Safety and
                        Pollution Prevention (U.S. EPA)
octaBDE                 octabrominated diphenyl ether
OD                     outer diameter; optical density
OECD                   Organisation for Economic Co-
                        operation and Development
OEL(s)                  occupational exposure limit(s)
OH-MWCNT              hydroxylated MWCNT
OPC                    optical particle counters
OPP                    Office of Pesticide Programs (U.S.
                        EPA)
ORD                    Office of Research and Development
                        (U.S. EPA)
ORISE                  Oak Ridge Institute for Science and
                        Education
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)
OW                     Office of Water (U.S. EPA)
P                       purity
p                       p-value, estimated probability, level of
                        statistical significance
p53v-                   mouse strain with impaired gene
                        stability
PAH(s)                  polycyclic aromatic hydrocarbon(s)
PBDD                   polybrominated dibenzo-p-dioxin
March 2013
                                                                XVII
                                                                        DRAFT-DO NOT CITE OR QUOTE

-------
PBDE
PBDF
PBS
PBZ
PCB(s)
PEC(s)

PEI
PEL
pentaBDE
PH
PINFA

PMMA
PMN
PND
POPs
ppb
ppm
PPM
ppt
R&D
RAW 264.7
REACH

REL
RfC
RfD
RNA
RRF

RTI
$
S
SA
SAFENANO
SD
SEM

Si
Si02
SNUR
polybrominated diphenyl ether
polybrominated dibenzofuran
phosphate buffered saline (solution)
personal breathing zone
polychlorinated biphenyl(s)
predicted environmental
concentration(s)
polyethyleneimine
permissible exposure limit
pentabrominated diphenyl ether
scale of acidity and alkalinity
Phosphorus, Inorganic and Nitrogen
Flame Retardants Association
polymethyl methacrylate
premanufacturing notice
postnatal day
persistent organic compounds
parts per billion
parts per million
Physicochemical Properties Map
parts per trillion
Research and Development
murine macrophage cell line
Registration, Evaluation, Authorisation
and Restrictions of Chemicals (EU)
recommended exposure limit
reference concentration
reference dose
ribonucleic acid
Risk Relevance Factor in CEA
Framework
Research Triangle Institute
cost in U.S. dollars
sulfur
surface area
Europe's Center of Excellence on
Nanotechnology Hazard and Risk,
based at the Institute of Occupational
Medicine
rat strain; standard deviation
scanning electron microscopy;
standard error of mean
silicone
silicone dioxoide
(TSCA) Significant New Use Rule
SOC(s)
SWCNT(s)
Ts
T4
TB
TB
TEM
tetraBDE
TfM
TGA
Ti
Ti02
TLV(s)
TOG
TOF
TpM
triBDE
TSCA
TWA
U.K.
U.SEPA
U.S.
UV
V
V79 cells

w/w
WHO
wt
XPS
Zn
synthetic organic compound(s)
single-walled carbon nanotube(s)
free tri-iodothyronine
free thyroxine
technical bulletin
total body
transmission electron microscopy
tetrabrominated diphenyl ether
Transformation Map
thermogravimetric analysis
titanium
titanium dioxide
threshold limit value(s)
total organic carbon
time of flight
Transport Map
tribrominated diphenyl ether
Toxic Substances Control Act
time weighted average
United  Kingdom
U.S. Environmental Protection Agency
United  States of America
ultraviolet
volume
lung fibroblast cell line from Chinese
hamster lung tissue
weight-for-weight measurement
World Health Organization
weight
x-ray photoelectron spectroscopy
March 2013
                                                                   XVIII
                                                                           DRAFT-DO NOT CITE OR QUOTE

-------
    Authors, Contributors,  and  Reviewers
EPA Project Leaders/Authors

Christina Powers, ORD/NCEA
Patricia Gillespie, ORD/NCEA
J. Michael Davis, ORD/NCEA


EPA Contributors

Ray Antonelli, ORISE
Jeff Gift, ORD/NCEA
Meredith Lassiter ORD/NCEA
Geniece Lehmann, ORD/NCEA
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


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
March 2013                              xix                DRAFT-DO NOT CITE OR QUOTE

-------
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, Ph.D., National Research Council Associate, US EPA/NRMRL/GWERD
March 2013                                    xx                   DRAFT-DO NOT CITE OR QUOTE

-------
                                            Preface
 1           This document is part of continuing efforts by the U.S. Environmental Protection Agency (EPA)
 2    to understand the scientific issues and information gaps associated with nanotechnology, consistent with
 3    recommendations in the U.S. EPA Nanotechnology White Paper (2007) and U.S. EPA Nanomaterial
 4    Research Strategy (2009). Although no national or international consensus definition for nanomaterials
 5    exists, a current working definition is a material having at least one dimension on the order of 1 to  100 nm
 6    (NSTC. 2011).
 7           Previous EPA documents similar to this one focused on nanoscale titanium dioxide used in
 8    drinking water treatment and in topical sunscreen (U.S. EPA. 2010d) and nanoscale silver used in
 9    disinfectant spray (U.S. EPA. 2010e). The nanomaterials considered in this document are multiwalled
10    carbon nanotubes (MWCNTs), as incorporated into flame-retardant coatings for upholstery textiles. This
11    document does not represent a risk assessment, nor is it intended to serve as a basis for near-term risk
12    management decisions on possible uses of MWCNTs. Rather, it is a case study, the external review draft
13    (U.S. EPA. 2012b) of which served as the starting point for identifying and prioritizing research gaps that,
14    if pursued, could inform future assessments and subsequent risk management decisions for MWCNTs in
15    this application. In revising the external review draft of the case study (U.S. EPA. 2012b) to create this
16    peer review draft, EPA streamlined the document to clearly reflect the identified research priorities and
17    input from public comments and expert stakeholders.
18           Like the previous case studies, this case study of MWCNTs is based on the comprehensive
19    environmental assessment (CEA) approach, which consists of both a framework and a process.
20    The organization of this document reflects the CEA framework, the principal elements of which are
21    described in Chapter 1 of this document and largely represented in Chapter 2 through Chapter 5.
22           This document also contains information about a traditional (i.e., "non-nanoenabled") product,
23    decabromodiphenyl ether flame-retardant upholstery coatings, against which the MWCNT flame-
24    retardant upholstery coating (i.e., the "nanoenabled" product) is compared. The primary purpose of
25    including a comparative element in the draft case study (U.S. EPA, 2012b) was to provide a more robust
26    database as a foundation from which to identify data gaps related to the nanoenabled product. Because it
27    has served its primary purpose, most of this comparative information has been moved to an appendix;
28    how the information about the traditional product might inform research planning for MWCNTs is
29    included in succinct textboxes in appropriate areas of the document.
      March 2013                                     xxi                    DRAFT-DO NOT CITE OR QUOTE

-------
 1           Following a general introduction to the materials and selected application in this case study in
 2    Chapter 1; Chapter 2 highlights stages of the product life cycle for the nanoenabled product. Chapter 2
 3    also identifies which stages in the product life cycle present opportunities for releases to the environment.
 4    Chapter 3 then provides information on the transport, transformation, and fate processes affecting the
 5    behavior of the nanomaterials, by-products, and transformation products in environmental compartments.
 6    Chapter 4 characterizes exposure, uptake, and dose for nanomaterials, by-products, and transformation
 7    products for different human populations and ecological receptors, after which Chapter 5 describes the
 8    human health, ecological, and other impacts related to those exposures.
 9           Collectively, these chapters represent the assembly of information across the vertical spectrum of
10    the CEA framework (Figure 1-1): as outlined in Chapter 1. however, this step is merely the first in the
11    CEA process (Figure 1-2). Next, a group of expert stakeholders representing a variety of technical
12    backgrounds and sectors used the draft case study document (U.S. EPA. 2012b) in a collective judgment
13    process to rate areas of the CEA framework in terms of importance for future risk assessments of
14    MWCNTs and their confidence in the data to support risk management decisions. Concurrently, the case
15    study was posted for public comment. Traditionally, the Agency has responded to expert feedback and
16    public comments by making in-text edits directly to case study documents as appropriate. In the current
17    case study, EPA not only addressed feedback and comments through in-text edits, but also placed greater
18    emphasis on highlighting the outcomes of the collective judgment step through incorporation on new
19    elements and highlighting key sections of the case study; by doing so EPA hopes to facilitate research to
20    support the compilation of new information in the CEA framework for future iterations of the approach.
21    Chapter 2 through Chapter 5 now reflect the areas of the framework expert stakeholders judged essential
22    for future risk assessment and management decisions. Areas identified as lower priorities are discussed in
23    appendices of the document. In addition, new information  identified through public or expert comments
24    on the draft (U.S. EPA. 2012b) is highlighted throughout this document in text boxes.
25           As described in more detail in Chapter 6. the identification of priority areas by a diverse group of
26    expert stakeholders is a key part of connecting research, risk assessment, and risk management for
27    MWCNTs (areas that were not identified as priorities are discussed in Appendix G). The next critical step
28    in this process is to engage the broader scientific community in implementing research in areas identified
29    as important to consider in future risk assessments but which lack sufficient data to support risk
30    management decisions for MWCNTs. Doing so will support the subsequent steps of the CEA process,
31    which involve a continued, iterative communication flow across the continuum of research, risk
32    assessment, and risk management.
      March 2013                                     xxii                     DRAFT-DO NOT CITE OR QUOTE

-------
                              Executive Summary
      Chapter 1: Introduction to this Document
 1           Background
 2           As part of an ongoing effort to identify research needs and data gaps in assessing the broad
 3    environmental implications of nanomaterials, this case study focuses on a specific nanomaterial in a
 4    particular application: multiwalled carbon nanotubes (MWCNTs) in flame-retardant coatings applied to
 5    upholstery textiles. The selection of this specific nanomaterial and particular application was made with
 6    input from representatives across the U.S. Environmental Protection Agency (EPA) and was based in part
 7    on its relevance to EPA programmatic interests and the similarity in the potential for release and exposure
 8    over the product life cycle compared to conventional flame-retardant materials that are being phased out
 9    of use.
10           Like previous case studies of nanoscale titanium dioxide and nanoscale silver, this case study is
11    built on the comprehensive environmental assessment (CEA) approach, which is both a framework and a
12    process. The CEA framework (Figure 1-1) starts with the inception of a material and encompasses
13    environmental fate, exposure-dose, and impacts associated with that material. The framework also
14    considers differences in environmental media and the physical, chemical, biological, and social conditions
15    in which the material occurs. Here, the framework is used to organize information about MWCNTs in the
16    case study systematically. This information does not represent a completed or even preliminary risk
17    assessment; rather, it is intended to inform research planning. The External Review Draft of the document
18    provided a basis for identifying and prioritizing data gaps and research needs for MWCNTs and other
19    nanomaterial assessments as part of the CEA process (Figure 1-2). Specifically, a group of expert
20    stakeholders representing diverse technical (e.g., toxicology, ecology, material science) and sector (e.g.,
21    industry, academia, government) perspectives engaged in a structured, collective judgment workshop
22    process such that each individual had equal input in identifying research priorities. To facilitate the
23    identification of key research gaps related to assessing MWCNTs in this application, the External Review
24    Draft case study provided a comparative perspective by also presenting information on a traditional flame
25    retardant, decabromodiphenyl ether (decaBDE). The prioritized research gaps that emerged are intended
26    to inform decision-makers in the EPA and the broader scientific community in developing research
27    agendas that support future risk assessment and risk management goals for MWCNTs. These Priority
28    Research Areas for MWCNTs are the primary focus of this revised document, with information on

      March 2013                                    xxiii                    DRAFT-DO NOT CITE OR QUOTE

-------
 1    decaBDE that supported identifying the priorities in the previous draft (U.S. EPA. 2012b). and are
 2    presented primarily in Appendix H. Background information on decaBDE, however, is provided in
 3    Chapter 1 to give the necessary context for reviewing the research priorities identified for MWCNTs. In
 4    addition, text boxes with the title "DecaBDE Can Inform MWCNT Assessment" are provided throughout
 5    the document to succinctly note how information on the  conventional material might inform research
 6    planning for MWCNTs. Information on MWCNTs that pertains to areas that were not prioritized for
 7    research is now located in Appendix G. Input on the External Review Draft case study from public and
 8    expert stakeholders also is highlighted throughout the document and is recorded in Appendix I.
 9           Given the purpose of the document, this case study does not purport to be a comprehensive
10    literature review; rather, available sources were incorporated specifically to support prioritizing and
11    subsequently planning research, as described above. As this case study involves an emerging technology,
12    some information, particularly regarding background or general concepts, was occasionally obtained from
13    non-peer-reviewed sources to supplement the published literature available. The most recent literature
14    search for this case study was conducted on May 11, 2012, using specific criteria to search the PubMed
15    database, Academic  Search Complete, Environment Complete, and CINAHL (Cumulative Index to
16    Nursing and Allied Health Literature) for records published since previous searches in November 2011
17    and January 2012. Search terms included carbon nanotube*, carbon nanofiber*,  CNT*, CNF*, MWNT*,
18    MWCNT*, and SWCNT*. Additional targeted literature searches were conducted on November 13, 2012,
19    using search terms specific to topic areas identified in public and expert comments.

20    Introduction to decaBDE and MWCNTflame-retardant textiles
21           Production and importation of decaBDE are currently being phased out in the United States as a
22    result of voluntary commitments within the industry and EPA actions in response to concerns regarding
23    potential human health and ecological impacts. As a result, a range of alternative flame-retardant
24    technologies, including nanotechnologies, is being evaluated as potential replacements for this
25    extensively used material. This document presents information on a potential alternative flame-retardant
26    technology, MWCNTs, in the context of the research priorities that could support future assessments of
27    this product. The primary purpose of this document is to inform research planning efforts for MWCNTs
28    across the scientific community. In doing so, the document supports a key objective of the CEA approach;
29    to link research, risk assessment, and risk management efforts iteratively.
30           In developing research plans for MWCNTs, understanding the considerations involved in their
31    potential use in flame-retardant textiles is informative. Many manufacturers incorporated flame-retardant
32    materials into textiles to comply with state, federal, and industry fire-safety standards (i.e., certain flame
33    test performance criteria that must be met). Once applied, flame retardants act to inhibit the combustion

      March 2013                                     xxiv                    DRAFT-DO NOT CITE OR QUOTE

-------
 1    process through a variety of physical or chemical means (e.g., producing inert gases that dilute the oxygen
 2    supply available to the flame, producing protective char barriers) (Section 1.2).
 3           Both decaBDE and MWCNTs can be mixed with binding agents and applied as coatings to
 4    increase the flame resistance of upholstery textiles. In this application, the two materials are both referred
 5    to as barrier technologies because they exhibit similar mechanisms of flame-retardant action: decaBDE
 6    forms a protective char barrier and MWCNTs form a network floccules layer (i.e., network of loosely
 7    bound MWCNT bundles). The similarity in potential applications for decaBDE and MWCNTs was a
 8    primary reason for including the comparison of the two materials as flame-retardant coatings in
 9    upholstery textiles in the External Review Draft of the case study, as the comparison informed the
10    identification of data gaps related to assessing possible risks and benefits associated with MWCNTs.
11    Moreover, the comparison of these materials highlighted MWCNT- and nano-specific factors that might
12    influence future research directions for nanomaterials and nanoenabled products. For example, unlike
13    with decaBDE, the physicochemical properties of MWCNTs are often intentionally altered during
14    synthesis; thus MWCNTs are not a single material with a defined set of characteristics, but rather a
15    variety of materials—often present as mixtures—with vastly different physicochemical characteristics.
16    Such variation in the physicochemical characteristics of MWCNTs presents challenges in describing the
17    releases, behavior, and effects of exposure to MWCNTs as a class of materials (Section 1.3). Importantly,
18    MWCNTs likely will be used in combination with other flame-retardant materials to provide sufficient
19    efficacy for the standards noted above (Section 1.2). In addition to introducing greater variability in
20    MWCNT behavior, exposure, and effects, the use of MWCNTs in combination with other materials raises
21    important implications for the potential use of MWCNTs in this application (Additional Information
22    Highlight Box 3).
      Chapter 2: Product Life Cycle
23           Little information is available on the commercial production and use of MWCNT flame-retardant
24    coatings, as few commercial-scale products currently exist. The manufacturing stages of MWCNT flame-
25    retardant textile coatings (Section 2.2). along with the use (Section 2.4) and reuse/recycling/end-of life
26    stages (Section 2.5). were identified as Priority Research Areas for upholstery textiles treated with
27    MWCNT flame retardants.
28           Based on the available data, releases of MWCNTs to the environment are expected to occur
29    throughout the life cycle of MWCNT flame-retardant upholstery textiles. The projected increase in
30    MWCNT production likely will result in increased environmental releases of MWCNTs from flame-
31    retardant textiles or other MWCNT products. Most MWCNTs released in the manufacturing stages are
32    anticipated to be in the free or bundled form (Footnote 11 in Chapter 2 explains this terminology), while


      March 2013                                     xxv                     DRAFT-DO NOT CITE OR QUOTE

-------
 1    most releases later in the life cycle are anticipated to be in the polymer or textile matrix-bound form.
 2    Upholstery textile products are expected to have a long lifespan and likely will be disposed of in
 3    municipal landfills or incineration facilities.
 4           Air and water releases of MWCNTs during manufacturing are expected to occur based on the
 5    activities performed in manufacturing stages of the product life cycle. Although release is particularly
 6    likely during mixing, handling, and equipment cleaning, releases are expected to be fairly well controlled
 7    when proper ventilation and environmental controls are in place. Air releases of MWCNTs have been
 8    measured during material synthesis but no data are available regarding release to water during
 9    manufacturing. Additionally, MWCNTs typically require purification and functionalization, which also
10    could result in releases due to chemical and physical processing methods (Section 2.2). Activities like
11    textile and furniture  processing might take place outside of closed systems and could result in
12    environmental releases of MWCNTs. Abrasion, washing, unintended use, and accidental exposure to high
13    heat or fire during the use stage could result in releases of MWCNTs (Sections 2.4 and 2.5).
14           No data are  currently available on the volume or potential release of MWCNTs in the use stage of
15    the flame-retardant upholstery textile product  life cycle. Based on decaBDE data, however, the potential
16    for release during this stage of the product life cycle could be relatively high. Similarly, no data currently
17    exist on the volume  or potential release of MWCNTs in upholstery textiles at end of life. Nevertheless,
18    the physical and chemical processes (e.g., shredding, milling, chemical treatment) used to recycle textiles
19    also could lead to releases of MWCNTs. Air releases from land-filling of MWCNT flame-retardant
20    upholstery also could occur due to mixing and compacting. In addition, release in leachate from landfills
21    is possible if the product or polymer matrix degrades. Although incineration at end of life presents the
22    potential for airborne release of MWCNTs and by-products, preliminary experimental data suggest that
23    MWCNTs will not be released to the environment when exposed to the sufficiently high temperatures of
24    municipal incinerators (Sections 2.4 and 2.5).  Incomplete incineration during other stages of the product
25    life cycle, however,  is one of the most likely airborne release scenarios for CNT textile coatings.
      Chapter 3: Transport, Transformation, and Fate
26           Although MWCNTs are incorporated into polymer matrices after the flame-retardant  production
27    stage, little information exists that describes the environmental behavior of these polymer matrices. As a
28    result, Chapter 3 focuses on the transport, transformation,  and fate of MWCNTs and not the polymer
29    matrices in which they are incorporated. Environmental transport, transformation, and fate of MWCNTs
30    in air, wastewater, and sediment were identified as Research Priority Areas. The environmental behavior
31    of MWCNTs is dictated by their physical and chemical properties—surface area, surface chemistry,
32    morphology (shape), solubility, presence or absence of functionalization and surface coatings  (e.g.,


      March 2013                                     xxvi                    DRAFT-DO NOT CITE OR QUOTE

-------
 1    engineered coatings or natural organic matter), and hydrophobicity. The nanostructured morphology,
 2    small size, and high surface area-to-volume ratio of MWCNTs can enhance chemical reactivity and
 3    propensity of MWCNTs to form bundles; single MWCNTs, as compared to bundles, will differ in their
 4    behavior in the environment (Section 3.1).
 5           Recent literature regarding the behavior of airborne MWCNTs is extremely limited, and
 6    dominant fate, transport, and transformation processes for MWCNTs in indoor and outdoor air are
 7    unknown. In aqueous media, such as wastewater, the hydrophobicity and van der Waals interactions of
 8    pure MWCNTs suggest they will bundle together or sorb to particles and be removed during the sewage
 9    treatment process, or settle out into sediment in receiving water bodies. Physicochemical characteristics
10    of the MWCNTs and environmental conditions, however, can alter this behavior. For example, the
11    presence of dissolved organic matter has been shown to debundle MWCNTs causing to them to remain in
12    solution. Similarly, surface coatings can affect the sorption behavior of MWCNTs in these systems and
13    influence their mobility, dispersion, and bioavailability in environmental media (Sections 3.2.
14    13, and 14).
15           Scientists have demonstrated the use of simple, deterministic models and more complex
16    probabilistic models to simulate movement of carbon nanotubes through, and predict environmental
17    concentrations in, environmental compartments. Differences in modeling approaches, model scale, and
18    model input data make comparisons across models for predicting environmental concentrations of CNTs
19    difficult. Nevertheless, a recent life-cycle-based analysis predicted the impacts of CNT synthesis in
20    aquatic systems by using output data from a single model of environmental concentrations. Nevertheless,
21    output data from  a single model predicting environmental concentrations were used in a recent life-cycle-
22    based analysis to predict the impacts of CNT synthesis in aquatic systems (Section 3.5).
      Chapter 4: Exposure-Dose
23           Several analytical challenges for nanomaterials combined with the lack of historical use of
24    MWCNTs in consumer products have so far prevented MWCNTs from being detected in ambient media,
25    which could  inform decisions related to potential exposures in human and ecological populations (Section
26    4.1). Human exposures to  MWCNTs  released throughout the flame-retardant textile coating life cycle are
27    expected to differ for workers, consumers, and the general public. Based on available information,
28    occupational and consumer exposures were identified as Priority Research Areas in the CEA collective
29    judgment workshop process for MWCNTs. Workers can be exposed to various forms of MWCNTs  (e.g.,
30    adsorbed to dust, as part of the polymer or textile matrix) via inhalation and ingestion of, and dermal
31    contact with, these substances during manufacturing, storage and distribution, and end-of-life activities. In
32    the workplace, the inhalation route is expected to represent the greatest potential for exposures, and


      March 2013                                      xxvii                    DRAFT-DO NOT CITE OR QUOTE

-------
 1    MWCNTs are expected to be in the participate phase when inhaled. Little is reported about consumer
 2    exposures to MWCNTs, especially those incorporated into flame-retardant textiles. Yet, based on
 3    activities expected to occur during use, repurposing, or reuse of upholstered products, consumers might
 4    be exposed to MWCNTs during each of these points in the product life cycle. The MWCNTs released
 5    from finished products also are expected to be in particulate form, generally adsorbed to dust or
 6    constituents of the polymer or textile matrix. The primary route of exposure (i.e., inhalation, ingestion, or
 7    dermal) for consumers is unknown.
 8           Developing exposure standards, guidelines, or recommendations for MWCNTs is complicated by
 9    the heterogeneity in MWCNT configurations and challenges measuring MWCNTs in occupational or
10    environmental settings. The National Institute for Occupational Safety and Health (NIOSH) established a
11    recommended exposure limit for elemental carbon, and several other occupational exposure limits have
12    been proposed by industry and international agencies (Section 4.2.5). In general, MWCNTs appear to be
13    biopersistent and might remain in the lung for several months after inhalation. Limited studies show that,
14    after oral exposure, most ingested MWCNTs are eliminated with no detectable metabolism or transport
15    into the blood. Distribution to the liver, lungs, and spleen, however, has been reported following
16    intravenous exposure (Section 4.2). Notably, the bioavailability, and thus dose, of MWCNTs likely will
17    be based on whether they are bound in a textile matrix, bundled, or free (Footnote 11. in Chapter 2
18    explains this terminology).
19           No evidence is currently available to determine whether portions of the population might
20    experience higher exposure levels to MWCNTs compared to the general population; however, the activity
21    of children and workers might increase total  exposure levels of MWCNTs relative to the general
22    population (Section 4.2).
23           Exposure and dose in ecological populations were not deemed Priority Research Areas for
24    MWCNTs in the CEA collective judgment workshop process, and thus information on these areas is now
25    located in Appendix G and Appendix H for MWCNTs and decaBDE, respectively. The anticipated
26    increase in MWCNT production (Section 2.2.2) along with increases in potential applications of the
27    material could lead to an increase in the number and type of exposures experienced by workers,
28    consumers, and ecological populations. These changes are expected to increase aggregate and cumulative
29    exposures to different formulations of MWCNTs, transformation products,  and by-products.
      Chapter 5: Potential Human  Health, Ecological, and Other Impacts
30           Expert stakeholders participating in the CEA collective judgment workshop process identified
31    human health impacts as a Priority Research Area for MWCNTs. Toxicology studies conducted on
32    animals are the only identified data on human health impacts of MWCNTs because no human data on


      March 2013                                    xxviii                    DRAFT-DO NOT CITE OR QUOTE

-------
 1    effects of MWCNT exposure exist. All routes of exposure were examined in this case study because each
 2    route (dermal, inhalation, and oral) offers potential for human exposures (Section 5.1). Toxicological
 3    effects from MWCNT exposure in animal models have been evaluated predominantly after dermal and
 4    inhalation exposures, rather than after oral exposure. Effects were generally localized and included
 5    irritation (skin and ocular), sensitization (respiratory), and inflammation (respiratory). In addition,
 6    MWCNTs altered immunological function after exposure via inhalation for 14 days or via a single
 7    intranasal injection. The carcinogenicity of MWCNTs following inhalation exposure has not been
 8    investigated; however, several studies using methods such as instillation indicate that some types of
 9    MWCNTs behave like asbestos, potentially inducing mesotheliomas, and might be more toxic than
10    asbestos (Section 5.1).
11           Expert stakeholders identified impacts in aquatic, but not terrestrial, biota as a Priority Research
12    Area. Considerations for the ecological impact of MWCNTs include the toxicity toward different species,
13    types of effects, and potential for bioaccumulation and biomagnification. More than 20 studies have
14    investigated the effects of MWCNTs on aquatic species or aquatic systems; those studies indicate  low
15    acute toxicity potential, with the effect level varying based on size and functionalization properties of the
16    MWCNTs. Chronic  studies show that MWCNTs can elicit immune responses and produce developmental
17    impacts (Section 5.2).
18           Other impacts, including economic or societal effects and alterations in environmental resources,
19    were identified as  a Priority Research Area by expert stakeholders. No empirical data exist relating
20    MWCNTs to other impacts, but the background literature on processes involved in manufacturing similar
21    materials (e.g., carbon nanofibers, single-walled carbon nanotubes) provides some basis for concern
22    regarding potential impacts of MWCNTs on energy demand, resource depletion, climate change, and
23    economics. These  related studies provide a plausible foundation for suggesting that MWCNT
24    manufacturing can be an energy-intensive process potentially causing the depletion of nonrenewable
25    natural resources like fossil fuels, and that the synthesis of MWCNTs can result in emissions of other
26    compounds causing adverse environmental effects (e.g., volatile organic compounds; Section 5.3).
      Chapter 6: Identifying  and Prioritizing Research Needs to Support Risk Assessment
      and Risk  Management
27           The External Review Draft of this document served as the foundation from which expert
28    stakeholders participating in the CEA process could identify key data gaps and determine research
29    priorities. The information presented in this revised document focuses on those priorities to inform
30    ongoing research planning for nanotechnology in the general scientific community and at EPA. Results of
31    these research efforts could subsequently support future assessments and risk management efforts for
      March 2013                                     xxix                    DRAFT-DO NOT CITE OR QUOTE

-------
 1    MWCNTs or other nanomaterials. Future evaluations of nanoenabled products, such as MWCNT in
 2    flame-retardant textile coatings, could involve the consideration of risk-related trade-offs, for example,
 3    thyroid health effects versus pulmonary health effects and environmental justice considerations versus
 4    energy costs. This document therefore strives to inform research planning efforts that would support
 5    conducting risk assessments that can inform risk management decisions about such trade-offs.
 6           The research priorities discussed in the case study were identified by a group of diverse expert
 7    stakeholders independently rating areas of the CEA framework based on two factors:
 8           •   Importance: how important an area is to consider in risk assessments of MWCNTs,
 9           •   Confidence: the availability and utility of current data to support risk management decisions
10               for MWCNTs.
11           For those areas they identified as "Important" to consider in future risk assessments of MWCNTs,
12    stakeholders were asked to rate the relative importance and confidence in data related to the relationship
13    of the area with risk factors that might be considered in risk assessment or risk management efforts for the
14    area. Areas that experts most commonly identified as being of high importance to risk assessment, and
15    were not confident in the data to support risk management decisions, are considered high priorities for
16    research. In contrast, areas rated as of high importance and for which experts had confidence in the data
17    might be of interest to decision-makers for evaluating risk management options for MWCNTs.
18           Most of the prioritized CEA framework areas were considered research priorities, including
19    release rates across the product life cycle; persistence and bioavailability in air, wastewater, or sediment,
20    and inhalation exposure in workers and consumers. Other areas identified as high Research Priority Areas
21    include absorption, metabolism, and excretion in humans, as well as impacts  on human health, aquatic
22    biota, and other considerations (i.e., economic, societal, environmental resources). For a subset of these
23    areas, experts identified potential risk managment decisions in the context of an example risk scenario for
24    that area and noted the type of assessment(s) that could inform those decisions. Specific research
25    questions to support such assessments also were identified, along with estimates of the financial and time
26    resources to carry out the research. Risk management decisions generally centered on choosing
27    appropriate control technologies or personal protective equipment, modifications to MWCNTs (e.g.,
28    reducing residence time in air by increasing aggregation potential), or limits on production and use of the
29    materials. Assessments to inform these and other types of risk management efforts included human health
30    risk assessments, cost benefit analyses, and life cycle assessments. Research areas to support such
31    assessments can be grouped into five general themes:  (1) the influence of MWCNT characteristics on
32    release from the product matrix;  (2) the influence of MWCNT characteristics and the product matrix both
33    on environmental transport and transformation, and on absorption across biological barriers (e.g.,
34    gastrointestinal tract); (3) development of analytical methods or tools to detect MWCNTs in complex

      March 2013                                      xxx                     DRAFT-DO NOT CITE OR QUOTE

-------
1    matrices and measure exposures; (4) human health impacts of MWCNTs and co-factors (e.g., solvents)
2    after acute and chronic exposures; and (5) improving public engagement in and understanding of potential
3    benefits and risks of nanotechnology.
4           The connection of specific questions within Priority Research Areas to the assessments and risk
5    management decisions they would subsequently support demonstrates the focus within the CEA approach
6    on linking communication across the continuum of research, risk assessment, and risk management.
7    Moreover, the specific questions are intended to provide more concrete support for strategic research
8    planning that informs future decision-making about MWCNTs.
     March 2013                                     xxxi                    DRAFT-DO NOT CITE OR QUOTE

-------
                                This page intentionally left blank.
March 2013                                     xxxii                    DRAFT-DO NOT CITE OR QUOTE

-------
        Chapter 1.  Introduction  to this  Document
      1.1. Background
 1          Nanoscale materials (nanomaterials) have been described as having at least one dimension
 2    approximately 1-100 nm (NSTC. 2011). Although this definition is not universally accepted and
 3    continues to evolve, 100 nm is typically used as an upper bound, and this working definition is used as the
 4    size standard in this case study. Engineered nanomaterials are intentionally synthesized at the nanoscale,
 5    rather than being produced as incidental by-products of combustion or a natural process such as erosion,
 6    to exploit the unique or novel properties that can arise from their small size. Like all emerging
 7    technologies, engineered nanomaterials offer the potential for both benefits and risks, the assessments of
 8    which depend on the availability of relevant data and other information.
 9          This document is part of an endeavor to identify what is known and, more importantly, what is
10    not known that could be of value in assessing the broad environmental implications of nanomaterials. As
11    a case study, this document presents information about a specific nanomaterial in a particular application.
12    It does not represent completed or even preliminary assessments; rather, the External Review Draft
13    provided a starting point in a process to identify and prioritize possible research directions to support
14    future risk assessments of nanomaterials. The prioritized research gaps that emerged are the focus of this
15    revised case study document. As with previous case studies, these research priorities are intended to
16    inform decision-makers in the U.S. Environmental Protection Agency (EPA) as well as the broader
17    scientific community in developing research agendas that support future risk assessment and risk
18    management goals. Such information is expected to be considered in the context of the particular focus,
19    budgetary constraints, ongoing research, and other considerations of any organization; however, as
20    discussed below, by using a holistic framework paired with input from a diverse group of expert
21    stakeholders, the priorities identified through the comprehensive environmental assessment (CEA)
22    approach employed in this case study can provide a unique perspective on research directions to support
23    future risk management goals.
24          The focus of this document is a specific application of a selected nanomaterial: the use of
25    engineered multiwalled carbon nanotubes (MWCNTs) as an agent in flame-retardant coatings on
      March 2013                                   1-1                  DRAFT-DO NOT CITE OR QUOTE

-------
 1    upholstery textiles.1 As described in detail in Appendix A. several candidate carbon-based nanomaterials
 2    and applications were identified as options for this case study using a systematic approach, and
 3    professional judgment was then applied to narrow down the selection to a single nanomaterial and
 4    application. First, candidate carbon-based nanomaterials were identified through initial strategic literature
 5    and Internet searches, news reports, and basic literature search statistics (e.g., number of total hits,
 6    number of hits in scientific databases). This approach provided an initial indication of overall data
 7    availability and research interest within the nanotechnology and scientific communities for several
 8    different broad groups of carbon-based nanomaterials (e.g., carbon nanotubes, carbon nanofibers,
 9    nanocrystalline cellulose). This group was further narrowed using a more judgment-based approach to
10    evaluating suitability, including consideration of the available data for multiple applications of each
11    nanomaterial. Finally, five feasible candidates of unique nanomaterial and application pairs—carbon
12    nanofibers in cement, MWCNTs in flame-retardant coatings, single-walled carbon nanotubes in textiles,
13    nanocrystalline cellulose in biodegradable packaging, and MWCNTs in rubber tires—were selected based
14    on additional professional judgment of suitability.
15            The process for selecting the material-application pair of MWCNT flame-retardant coatings for
16    upholstery textiles as a CEA case study involved individuals representing EPA program offices, regional
17    offices, and Office of Research and Development laboratories and centers. Individuals were appointed by
18    their organization within EPA to be involved with development of nanomaterial case study documents.
19    They were encouraged to share information on the five selected candidate carbon-based nanomaterials
20    and applications with colleagues in their organization and to represent the views of their organization in
21    voting for their preferences. The two candidates receiving the most votes were MWCNTs in flame-
22    retardant coatings and composites and SWCNTs in textiles. Rationale for selecting MWCNTs and
23    SWCNTs in each respective application included: relevance of both materials to Agency programs,
24    similarity in potential release and exposure over the product life cycle of textiles compared to existing
25    flame-retardant materials being phased out of use, greater availability of data compared to other candidate
26    applications, and potential for market expansion of CNTs (see Appendix A). Based on input that
27    MWCNTs were of greater interest (i.e., more widely produced than SWCNTs and might contain more
28    contaminants) and that an application involving textiles would be preferable,  a hybrid option was selected
29    as the topic of this case study: MWCNTs in flame-retardant coatings applied to upholstery textiles. This
30    selection does not imply that MWCNTs in flame-retardant coatings applied to textiles represents the
                ame 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.
      March 2013                                      1-2                   DRAFT-DO NOT CITE OR QUOTE

-------
 1    carbon-based nanomaterial and application with the largest current market share (see Section 1.3.2). but
 2    rather was based on the selection factors noted above.
 3           Using a similar selection process, EPA completed case studies of nanoscale titanium dioxide used
 4    for drinking water treatment and for topical sunscreen (U.S. EPA. 2010d) and nanoscale silver used as an
 5    agent in disinfectant spray products (U.S. EPA. 2010e). Unlike previous case studies, this case study
 6    incorporates  information about a traditional (i.e., "non-nanoenabled") flame-retardant product,
 7    decabromodiphenyl ether (decaBDE), against which the MWCNT flame-retardant coating (i.e., the
 8    "nanoenabled" product) can be compared (see Section 1.1.4). As discussed in greater detail below (see
 9    Section 1.1.3). the primary purpose of including comparative information on decaBDE was accomplished
10    with the External Review draft of this case study document (U.S. EPA. 2012b). and that information is
11    now primarily contained within Appendix H.
12           Part  of the rationale for compiling a series of nanomaterial case studies is that the properties
13    associated with different nanomaterials are often complex and vary considerably within, between, or
14    among specific types of nanomaterial groups, nanomaterials in general, and different applications of
15    nanomaterials. As a result, applying generalities could result in overlooking key characteristics or
16    information.  Focusing on a single example of an application of MWCNTs is not intended to represent all
17    ways in which this nanomaterial could be used or all issues that other applications might raise. By
18    considering this single application of MWCNTs, however, research directions can be identified that
19    would support future assessments of this material. Such information might be used more broadly as an
20    analog for other applications of MWCNTs or types of nanomaterials. For instance, research investigating
21    the influence of MWCNT surface treatment on potential release from flame-retardant textile coatings and
22    subsequent behavior in environmental media can also inform efforts to understand the influence of
23    surface treatment on the environmental behavior of MWCNTs in other applications.


      1.1.1.   Introduction to Comprehensive Environmental  Assessment
24           This case study of MWCNTs, like the previous case studies of nanoscale titanium dioxide (U.S.
25    EPA. 2010d) and nanoscale silver (U.S. EPA. 2010e). is built on the CEA approach, which consists of
26    both a framework and a process, the principal elements of which are illustrated in Figure 1-1 and Figure
27    1-2. respectively. The uppermost box of Figure 1-1 lists typical stages of a product life cycle: research
28    and development (R&D), feedstock processing, manufacturing, storage and distribution, use, and disposal
29    (which would include reuse or recycling, if applicable).
30           Although not considered a life-cycle stage in typical life cycle analyses, R&D is included in
31    business models of product value chains. Because of the relatively large portion of resources and

      March 2013                                    1-3                  DRAFT-DO NOT CITE OR QUOTE

-------
  1    information associated with this stage for emerging materials, such as nanomaterials, R&D is considered
  2    in the CEA framework. The actual volume of the material used in R&D is likely small but could represent
  3    a significant proportion of the total market, particularly during product development, given the limited
  4    number of full-scale commercial manufacturing efforts early in the life cycle for emerging materials. For
  5    these materials, processes in R&D lend insight to full-scale commercial processes and might constitute an
  6    important source of material release into the environment, as well as occupational exposures. Other CEA
  7    applications focusing on traditional or more mature materials or technologies might provide minimal or
  8    no information on the R&D portion of the product lifecycle, given that R&D would be less active.
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
                                             ENVIRONMENTAL
                                               CONDITIONS
        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
                                        . physical
                                                           Transport/Transformation/Fate
                                                          Primary and Secondary Substances
                                                                Exposure-Dose
                                                      Humans   Other Biota    Abiotic Resources'?
                                                                                  Source: (U.S. EPA, 2011 a]
between transport, transformation, and Figure 1 -1. Comprehensive environmental assessment
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
between a contaminant and a receptor,
whether living or nonliving. Living
                                                  framework.
                                      The CEA framework is used to systematically organize complex information 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 physical, chemical, biological, and
                                      social conditions in which the material event occurs. Details on these influential factors
                                      are thus included throughout the framework when possible.
       March 2013
                                                 1-4
                                                                        DRAFT-DO NOT CITE OR QUOTE

-------
 1    organisms include humans and other biota.2 Examples of nonliving, or abiotic, receptors include features
 2    of the natural landscape, structures such as buildings and statues, and painted surfaces of vehicles and
 3    other objects. Exposure can involve aggregate exposure across routes (e.g., inhalation, ingestion, dermal),
 4    cumulative exposure to multiple contaminants (both primary and secondary), and various spatiotemporal
 5    dimensions (e.g., activity patterns, diurnal and seasonal changes). Dose is the amount of a substance that
 6    enters an organism by crossing a biological barrier or which deposits on an inanimate object.
 7           As part of a chain of cause-effect events, dose links exposure with potential impacts of various
 8    types, as indicated in the last box of Figure 1-1. Human health effects might result when a certain
 9    delivered dose reaches a target cell or organ. In  an ecological context, effects might occur when a stressor
10    reaches a level sufficient to cause an adverse outcome in biotic or abiotic receptors. Impacts encompass
11    both qualitative hazards and quantitative exposure-response relationships and can extend to aesthetic
12    (e.g., alterations in visibility, taste, and odor), climate change, energy consumption, resource depletion,
13    socioeconomic, and other effects. Such effects are considered in the CEA framework, but their ultimate
14    inclusion would depend on whether the compiled  information indicates that such effects could reasonably
15    be expected to occur. As discussed below, the inclusion of such information in the CEA framework
16    should influence the selection of the technical experts for the next step of the CEA process.
17           Not reflected in Figure 1-1 is the role of analytical methods that make detecting, measuring, and
18    characterizing nanomaterials in the environment and in organisms possible. Characterizing a substance of
19    interest (e.g., determining its chemical identity,  reactivity, purity, and other properties) is fundamental to
20    the assessment of any  material. Thus, if adequate analytical techniques have not yet been developed or
21    need refinement, methods development must be included in research efforts to  inform future assessments.
22    For simplicity, such information is not included in this high-level view of the CEA framework. For the
23    purpose of this document, analytical methods for the materials in this case study are presented in detail in
24    Appendix B.
25           As previously mentioned, the CEA approach consists of both a framework and  a process.
26    Compiling the information described above into the CEA framework is the first step of the CEA process
27    (Figure 1-2).  Starting with the holistic perspective of the CEA framework facilitates identifying
28    information pertinent to consider for the material of focus,  which in turn supports problem formulation
29    and scoping for assessment purposes. Next, a collective judgment process is used to evaluate and
30    prioritize this information. Collective judgment, as applied in the CEA process to date, refers to a formal,
31    structured procedure enabling a range of participants to be  heard individually and to be  represented in a
32    transparent record of the collectively reached outcomes.  Collective judgment supports an essential feature
       The term biota is used throughout this document to refer to all living organisms other than humans.
      March 2013                                      1-5                    DRAFT-DO NOT CITE OR QUOTE

-------
  1
  2
  3
  4
  5
  6
  7
  8
  9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
of CEA: the inclusion of diverse
technical and stakeholder perspectives
to ensure that a holistic evaluation is
achieved (U.S. EPA. 2010h).
        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
encompasses the use of prioritized
information by research planners and
                 Compile Information
                  in CEA Framewor
                                   Develop Risk
                                 Management Plan
    Develop
  Research Plan
       I
Conduct Research
                                 Monitor, Evaluate
                                    Outcomes
                                         Source: (U.S. EPA. 2011 a)

Figure 1-2. Comprehensive environmental assessment
           process.
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.
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, 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
March 2013
        1-6
                        DRAFT-DO NOT CITE OR QUOTE

-------
 1    is responsive to current recommendations to further refine risk assessment and management approaches
 2    (see Chapter 6).
      1.1.2. Purpose of this Document
 3           This document has been revised from the External Review Draft that was used in the collective
 4    judgment step of the CEA process applied to MWCNTs (Figure 1-2). As a revised case study, it provides
 5    a basis for considering the outcomes of the collective judgment process to aid research planning that
 6    supports long-term assessment efforts; it does not, however, purport to present an exhaustive review of
 7    the literature. Furthermore, this case study is not an actual risk assessment and does not provide
 8    conclusions on potential ecological  or human health impacts related to MWCNTs. As discussed further
 9    below, this document is focused on  highlighting data gaps to inform risk assessment and risk management
10    processes related to MWCNTs, and, as such, it does not discuss benefits. A variety of potential economic,
11    social, and other benefits would likely need to be considered in future evaluations of risk-related trade-
12    offs for nanomaterials such as MWCNTs [e.g., Wang and Shapira (2012). Bonner (2011)1.
13           It must be emphasized that this case study has been developed without a specific regulatory or
14    policy objective in mind. Within the United States, regulatory decisions for nanomaterials may be made
15    by a number of federal agencies (e.g., EPA, Food and Drug Administration, Consumer Product Safety
16    Commission, Occupational Safety and Health Administration) under a variety of legislative frameworks
17    (e.g., Toxic Substances Control Act [TSCA], Federal Food Drug and Cosmetic Act). As discussed in
18    Chapter 6. TSCA is one of the statutes under which EPA  currently considers nanoscale substances.
19    Considerations of nanoscale substances under TSCA begin with a determination of whether the substance
20    is already included on the TSCA Chemical Substance Inventory based on whether the substance has the
21    same molecular identity as a substance listed on the Inventory (U.S. EPA. 2008c). Determinations of
22    whether nanoscale substances are new or existing substances are currently made on a case-by-case basis
23    (U.S. EPA. 2008c).
24           A variety of efforts are underway to increase the scientific body of knowledge such that
25    regulatory decisions through TSCA or other statutes could move beyond a case-by-case approach. These
26    include several research frameworks applicable to, or exclusively for, nanomaterials and intended to
27    support future assessments and subsequent risk management of these materials [e.g., (U.S. EPA. 2009).
28    (NRC. 2012). and (OECD. 2012). These frameworks and the CEA approach share a number of common
29    elements (e.g., focus on product life cycle, identifying environmental fate mechanisms and exposure
30    sources, importance of stakeholder engagement), yet as described above, the CEA approach incorporates
31    decision-support tools to engage stakeholders beyond that seen in other frameworks to date (Figure  1-2).

      March 2013                                     1-7                   DRAFT-DO NOT CITE OR QUOTE

-------
 1           This document presents information in the CEA framework for MWCNTs, as potentially used in
 2    flame-retardant coatings applied to upholstery textiles, in the context of the outcomes that emerged from
 3    engaging expert stakeholders in the CEA process for this material. This case study begins with a general
 4    overview in Chapter 1 of textiles and flame-retardant systems and where MWCNTs fit into that context,
 5    as well as detailed introductory information on decaBDE in the context of textiles and flame-retardant
 6    systems. Throughout the main body of this document (Chapter 2 through Chapter 5). the focus is on
 7    highlighting what is known and not known related to each portion of the CEA framework identified as a
 8    research priority in the CEA collective judgment step for MWCNTs in flame-retardant coatings. As such,
 9    in these chapters readers are referred to Appendix H for the detailed information regarding decaBDE at
10    each stage of the CEA framework; however, highlight-level information on decaBDE is presented in
11    select tables, figures, and text boxes (see Section 1.1.4 and Appendix D to provide a succinct comparison
12    between MWCNTs and decaBDE in this particular application with the intention that such comparisons
13    might inform MWCNT research planning.


      1.1.3. How the CEA Framework and Process Were Applied
14           An important aspect of the CEA approach is the ability to examine the relative risks and benefits
15    of, for example, different products or different formulation options, to aid in risk management decisions.
16    The particular comparison to focus on in an application of CEA would be guided by risk management
17    objectives. For example, MWCNT flame-retardant coatings applied to upholstery textiles might be
18    compared to conventional flame-retardant products, a different nanoenabled flame-retardant formulation,
19    a flame retardant not applied as a coating, or some other variable. Although several different options
20    could be of interest to  risk managers, considering every potential option in the present case study is not
21    feasible. Therefore, this document focuses solely on a comparison of MWCNTs and a traditional flame
22    retardant, decaBDE, as they might be used in flame-retardant coatings  for upholstery textiles, including
23    those used in homes and nonresidential areas such as public buildings and automobiles.
24           In the External Review Draft of this case study, the comparison between decaBDE and MWCNTs
25    provided: (1) a more robust database (i.e., that of a traditional product that has been relatively well
26    characterized) as a reference for identifying data gaps relating to a nanoenabled product; and (2) a context
27    for identifying key factors and data gaps related to assessing the risk-risk and risk-benefit trade-offs
28    between a nanoenabled product and a non-nanoenabled product. Although the specific characteristics,
29    exposure patterns, and effects associated with the use of MWCNTs and decaBDE  are  expected to differ
30    substantially, the data needed to inform risk assessment and risk management decision-making are
31    comparable; thus, the comparative framework was used to help determine whether relevant information

      March 2013                                     1-8                   DRAFT-DO NOT CITE OR QUOTE

-------
 1    (e.g., dominant exposure pathways, sensitive populations) is available and sufficient to inform future risk
 2    decision-making, and by extension, to identify key data gaps that could be pursued.
 3           The External Review Draft of this document represented the "Compile Information in CEA
 4    Framework" step of the CEA process (Figure 1-2). and thus supported the next step of the process:
 5    evaluating the data in the framework using a collective judgment technique to identify and prioritize
 6    information gaps about MWCNTs. The collective judgment prioritization technique used for this case
 7    study was funded by EPA and conducted independently by an EPA contractor, RTI International. Details
 8    related to the collective judgment method and its outcomes are described in a separate report prepared by
 9    RTI International (RTI. 2012). A summary of that process is described here with the outcomes discussed
10    in greater detail in Section 6.3.
11           In the collective judgment step of CEA applied to MWCNTs, selected experts representing
12    diverse sector (e.g., industry, academia, government) and technical backgrounds (e.g., toxicology,
13    ecology, material science) were first asked to read the External Review Draft of the case study. Next, they
14    were asked to consider what elements of the CEA framework were most important to understanding, and
15    therefore managing, the most significant risks associated with MWCNTs.
16           The experts identified important areas by independently rating areas of a more detailed view of
17    the  CEA framework (Figure 1-3). This detailed CEA framework illustrates discrete elements (blue boxes
18    in top left of Figure 1-3). or discrete pathways within the broad levels of the CEA framework (e.g.,
19    Product Life Cycle, Exposure in Figure 1-1). Each element is associated with "risk relevance factors"
20    (green boxes in top left of Figure 1-3). which might be considered in risk assessment or management
21    efforts of a material, such as MWCNTs.
22           Experts were asked to rate the importance of each element of the detailed CEA framework as
23    important, possibly important, or least important. If they rated the element important, they were then
24    asked to rate (1) the importance of each element-risk relevance factor pair (E-RRF) using the same scale,
25    and (2) their confidence in the availability and utility of current data for the E-RRF to  support risk
26    management decisions (as confident,  somewhat confident, or not confident).
      March 2013                                     1-9                   DRAFT-DO NOT CITE OR QUOTE

-------
 1           These ratings were collected in each of the three rounds of collective judgment prioritization used
 2    for this application of CEA:
 3               Round 1: Thirty-one selected participants entered their individual opinions on the E-RRF
 4               pairs in a spreadsheet and submitted the spreadsheet to a secure online platform (website);
 5               Round 2: Twenty-eight of the original 31 participants3 viewed the compiled opinions of the
 6               wider group through a series of bar charts and tables available via the website and were given
 7               the opportunity to re-enter their opinions;
 8               Round 3: A subset of participants (13) attended a structured workshop where they:
 9               a.  discussed their opinions in a structured collective judgment technique,
10               b.  finalized research priorities through a third round  of individually rating all E-RRFs and
11                   compiling these ratings,
12               c.  developed detailed research questions for a subset of those priorities.
13    The finalized priority areas determined in Round 3, part b of the technique, hereafter referred to as
14    "Priority Research Areas," are summarized in Figure 1-3 and discussed in greater detail in Section 6.3. As
15    discussed below, these outcomes were used to focus the information in this case study document.
16    How the Case Study  Was Streamlined to Emphasize Research Priorities
17           Compared to the External Review Draft (U.S. EPA. 2012b), this draft of the case study document
18    has been streamlined to clearly reflect the outcomes of the collective judgment step of the CEA process.
19    New text boxes have been embedded in the  document immediately following section headings that
20    correspond to elements of the detailed CEA framework (see Figure 1-3) to highlight the outcomes of the
21    RTI workshop (RTI. 2012) related to the E-RRFs discussed in that section of the case  study.
22           Boxes outlined in red with the title "Priority Research Area Highlight" (e.g.,  Section 2.2.2)
23    indicate that the E-RRFs discussed in that section were deemed to be priorities for continuing research by
24    participants in the RTI workshop, based on (1) high importance of that area to risk assessment and risk
25    management, and (2) low confidence in the utility and availability of the data on the topic. Boxes
26                    with the title "Unprioritized Research Area Highlight" (e.g., Section 2.1) indicate that
27    the E-RRFs discussed in that section were not identified by workshop participants as Research Priority
28    Areas (i.e., the most commonly selected rating was "possibly important" or "least important" rather than
29    "important"; therefore, the majority of participants did not rate the Importance and Confidence for those
30    E-RRFs). For these sections, all text relevant to decaBDE and MWCNTs was moved to Appendix H and
31    Appendix G. respectively, to focus the main body of the document on the priority research areas.
      3 Three participants from the first round of prioritization were unable to participate in the second round.
      March 2013                                     1-10                   DRAFT-DO NOT CITE OR QUOTE

-------
 1           In each "Priority Research Area" and "Unprioritized Research Area" highlight box, a graphic
 2    appears that summarizes information on how the 13 workshop participants individually rated the
 3    Importance of each element; and, for the subset of participants who stated the element was of highest
 4    importance, their Importance and Confidence ratings for each E-RRF. The collective Importance and
 5    Confidence for each E-RRF is expressed using an Importance/Confidence Matrix image, where the three
 6    Importance categories are shown on the Y-axis and the three Confidence categories are shown  on the
 7    X-axis, creating nine bins representing unique importance-confidence pairings (see lower left of Figure
 8    1-3). E-RRFs were assigned to a particular bin of the Importance/Confidence Matrix based on which
 9    ratings were most commonly selected by expert stakeholders for Importance and for Confidence.4'5 The
10    prioritization of the framework areas (i.e., E-RRFs) is therefore based on the most frequently selected
11    rating for each factor (Importance or Confidence), rather than on the most commonly selected
12    combination of Importance and Confidence for each E-RRF. E-RRFs in Unprioritized Research Areas
13    were not assigned to a particular bin since only a small subset of participants rated the Importance and
14    Confidence of the E-RRF.
15           In most instances, the most commonly agreed-upon Importance and Confidence ratings align with
16    the portion of the matrix with the largest number of stakeholders; however, in three instances, this is not
17    the case.  This lack of concordance reflects a difference in how individuals combined
18    Importance/Confidence ratings compared to the overall rating combination of all stakeholders.  In all
19    cases, the most commonly selected rating for Importance and the most commonly selected rating for
20    Confidence determines the placement of the E-RRF in the Importance/Confidence Matrix.
21           Finally, some case study sections present necessary supporting information for E-RRF pairs, but
22    do not directly discuss a specific E-RRF pair. These sections have been identified as "Neutral Research
23    Areas," and a small text box outlined in black with that title has been placed under the section heading.
24    For these cases, the text that originally appeared in the section remains, as it supports understanding of
25    other E-RRFs that are priorities for research.
      4In instances of a tie (i.e., six out of 13, or 46% of 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.
      5For example, if six out of 13 stakeholders rated an E-RRF "Important" and three out of 13 stakeholders rated the
      E-RRF "Possibly Important," the E-RRF was collectively rated "Important." Similarly, if four out of 13
      stakeholders, 31%, rated their confidence in an E-RRF as Not Confident and three out of 13, 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'V'Not Confident" bin of the matrix).

      March 2013                                     1-11                   DRAFT-DO NOT CITE OR QUOTE

-------
            r—
                               _7 8 —.

         SI
         a
         c
                          Somewhat
                          Confident
              Confidence that Current Data Can Support
                   Risk Management Decisions
                                                                            Product Life Cycle
                                                                  Environmental Transport, Transformation & Fate
                                                                             Exposure Route
                                                                             Dose (Kinetics)
                                                                      [ Absorption ]
                                                                               J
                          [ Terrestrial Biota
                           [ Absorption ]~
                             [ Distribution")
                               [ Metabolism ]
                                  Excretion
                                                                                Impacts
       Figure 1-3. Detailed CEA framework used for the collective judgment prioritization process.

       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 colored in represent areas collectively identified as of high importance
       for future assessments. Areas in red are those of highest priority for research since 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     How the Case Study Was Revised to Respond to Public and Peer Comments

 2             Additional changes were made to the case study document, and new elements were added in

 3     response to written feedback from 23 experts involved in the prioritization process, and in response to

 4     comments from several members of the public (see Appendix I for more detail). Consistent with the

 5     discussion above, these changes were implemented to emphasize research priorities identified through the

 6     RTI workshop process, and to improve the  scientific accuracy and rigor of the compiled information.

 1             First, as mentioned in Section 1.1.3 and discussed above, detailed information on decaBDE was

 8     moved to Appendix H and replaced with text boxes that include  highlights comparing information known

 9     about decaBDE to what is known about MWCNTs in Research Priority Areas. These highlights are meant

10     to illustrate how understanding the data on  decaBDE in flame-retardant upholstery textiles might help
       March 2013
1-12
DRAFT-DO NOT CITE OR QUOTE

-------
 1    guide research planning to elucidate potential risks of MWCNTs. These text boxes are consistently titled
 2    "DecaBDE Can Inform MWCNT Assessment" and are outlined in green (e.g., Section 2.2.2).
 3           Second, a series of "Additional Information Highlight Text Boxes," new figures, and new tables
 4    were embedded in the case study to draw attention to scientific concepts related to the priority areas that
 5    commenters felt were under-represented in the External Review Draft of the case  study. These elements
 6    were added to emphasize scientific topics that were included in the External Review Draft; but, were
 7    unclear or not clearly described, or to discuss a topic that was not included previously but is relevant to
 8    the topic and discussion. Additional Information Highlight Text Boxes, new tables, and new figures are
 9    outlined in blue (e.g..  Table 2-2).
10           Finally, Chapter 6 was expanded to include a final section (Section 6.3) that discusses the priority
11    research areas in more  detail. Section 6.3 builds on the red outlined "Priority Area Highlight" text boxes
12    described above, which are intended to briefly outline how participant ratings resulted in the area being
13    collectively identified as a priority. Examples of the rationale for prioritizing these areas are presented in
14    Section 6.3. along with factors that might be important to include in planning research for each area. In
15    addition, for some priority areas, commenters, workshop participants and targeted literature searches
16    identified relevant literature that had not been included in the External Review Draft of the case study.
17    This literature is discussed in Section 6.3 in the context of how it might influence  research planning for
18    the area. Finally, specific research questions identified by expert participants (or based on the available
19    literature)  are listed for each priority area.
      1.1.4. Selection of DecaBDE for Comparison
20           DecaBDE has been used widely in the textile industry to meet fire safety standards (see Section
21    1.2.1). Concern, however, is growing regarding the potential impacts of decaBDE on ecological and
22    human health. For example,  despite previous assumptions that decaBDE is relatively stable and inert in
23    the environment, recent studies have suggested that it can debrominate, or break down into lower-weight
24    congeners, which have been much more widely studied and are known to be highly toxic [(Environment
25    Canada. 2010: U.S. EPA. 2010b: Siddiqi et al.. 2003: Rahman etal.. 2001): see Section 3J_ and Text Box
26    H.3-1]. In response to these concerns regarding potential adverse impacts on human health and the
27    environment (see Chapter 5). limitations or bans on the use of decaBDE have been imposed recently both
28    in the United States and abroad. As summarized in Table  1-1. several states have begun to phase out or
29    restrict the use of decaBDE.
30           In December 2009, the two largest U.S. producers and the largest U.S. importer of decaBDE
31    announced voluntary commitments to phase out decaBDE in the United States by 2013 (U.S. EPA.

      March 2013                                     1-13                  DRAFT-DO NOT CITE OR QUOTE

-------
 1    2010a). As summarized in the EPA PBDE Action Plan dated December 30, 2009, several reports
 2    provided evidence for the human and environmental effects of this compound. A finding of "suggestive
 3    evidence of carcinogenic potential" was reported in the 2008 Toxicological Review ofDecaBDE (U.S.
 4    EPA. 2008b). Neurobehavioral effects also were identified in IRIS assessments for decaBDE and
 5    additional congeners (tetraBDE, pentaBDE, and hexaBDE).  Environmental hazards associated with
 6    PBDEs include persistence, potential for biomagnification, and breakdown of some PBDEs to more toxic
 7    congeners to produce effects at environmentally relevant concentrations (based on reports from
 8    Environment Canada and studies from other authors). Furthermore, in 2012, EPA initiated proposed
 9    amendments to (1) the Toxic Substances Control Act (TSCA) § 5(a)(2), a Significant New Use Rule
10    (SNUE), and (2) TSCA § 4, a Test Rule for decaBDE. The SNUR would require any entity planning to
11    manufacture or import decaBDE or articles to which decaBDE has been added to notify EPA at least 90
12    days in advance, which would provide the Agency an opportunity to review and evaluate data related to
13    the new use and to take action to limit or prohibit the new use if necessary. The Test Rule would require
14    laboratory studies to determine the effects that decaBDE has on human health and the environment (U.S.
15    EPA. 2012c).
      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 decaBDE      Vermont, Maryland, Maine, New York
      Source: National Conference of State Legislators (2011).
16            Although commercial MWCNT flame-retardant products are available, their presence is
17    relatively new, and they are by no means abundant on the market. Given the projected decline in
18    decaBDE use, as described above,  investigating these nanoenabled products as a potential emerging
19    alternative is relevant. The use of flame retardants in textiles is of interest to EPA and also aligns with the
20    needs of other organizations (e.g., Consumer Product Safety Commission, National Institute of Standards
21    and Technology). DecaBDE was chosen as the traditional flame-retardant product to compare to
22    MWCNTs due to its extensive use  since the 1970s and the robust scientific database available for it and
23    for the brominated flame retardant (BFR) family in general.
      March 2013                                     1-14                   DRAFT-DO NOT CITE OR QUOTE

-------
 1          The comparison of the larger body of information on decaBDE with the relatively small database
 2    for MWCNTs was intended to help pinpoint data gaps relating to this specific MWCNT product. For this
 3    reason, with the exception of Chapter 1, comparable information for decaBDE is included for each CEA
 4    framework area in Appendix H as a reference. Additionally, "Comparison Highlight Boxes" are included
 5    throughout the document to guide the reader to aspects of decaBDE that are particularly useful for
 6    drawing parallels to MWCNTs (see Appendix I). Because Chapter 1 is intended to provide relevant
 7    introductory information for both the traditional and nanoenabled product, it discusses relevant
 8    information on both MWCNTs and decaBDE while the rest of the document focuses solely on the
 9    "Priority Research Areas" relevant to MWCNTs.
      1.2. Introduction to Flame Retardants  in Textiles
10          Textiles and fabrics, which are networks of fibers composing flexible woven or nonwoven
11    materials, are flammable to varying degrees due to their ignitability and their potential to propagate flame
12    and produce burning droplets (PINFA. 2010). The behavior of various untreated textiles when exposed to
13    flame depends on the chemical composition of the raw materials. Table 1-2 lists several common
14    categories of textile fibers along with their flammability characteristics. The flammability of these fibers,
15    when incorporated in different textile products, has led to the development of numerous fire safety
16    standards (TINFA. 2010). as discussed in Section 1.2.1.
      March 2013                                   1-15                  DRAFT-DO NOT CITE OR QUOTE

-------
     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 flame1
     Viscose
Burns rapidly, similar to cotton
     Acetates
Burn heavily; can melt away from flame; form burning droplets2
     Acrylics
Burn rapidly; form burning droplets; produce dense black smoke
      Polyesters
      Polyolefins
      Polyamide
                        Burn slowly and hot;3 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
      1 Melting 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.
      2Burning droplets can form if the fiber melts slowly while in contact with the flame.
      3Burning hot refers to a high peak heat release rate.
      Source: PINFA (2010).
      1.2.1.  Standards for Textiles
1             Upholstery textiles, particularly those used outside of residential settings (e.g., in hospitals,
2     airports, airplanes, penal institutions, public transportation, office buildings), are subject to various state,
3     federal, and voluntary fire safety standards (see Table 1-3 for examples). Technical standards specify the
4     types of products to which standards apply, methodologies for conducting specific tests, measured
5     parameters of interest (e.g., time to ignition, heat release rate),  and performance criteria for each test and
6     product of interest (Illinois Environmental Protection Agency.  2007).
      March 2013
                                 1-16
DRAFT-DO NOT CITE OR QUOTE

-------
     Table 1 -3.     U.S. and international fire regulations for upholstery textiles.1
     Product Category
Standards
Description
     Automotive vehicle (bus and
     car) passenger compartments;
     curtains or blinds used in
     automotive vehicles
FMVSS302/DIN
75200/ISO 3795; DIN
50051
Specimen subjected to Bunsen burner flame for 15 seconds. The rate of
flame spread should be <101.6 mm/min (for a 245-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
16CFR1632(2000;
updated 2007); 16
CFR 1633 (2006); CA
TB 603 (2005); CATB
129;CATB121
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  49 CFR Part 238
      trains                        (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 test2 for furniture manufactured for use in public
buildings in California. Many other states have adopted TB 133.
     Cigarette testing of upholstered   Upholstered Furniture
     furniture fabric                 Action Council; CA TB
                                  116;NFPA701
                    Component standard. All upholstered furniture sold in California must
                    pass this flame test; applies to buildings under NFPA 701 code.
     1This 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.
     2Full-scale flame test refers to the use of a full piece of furniture or mockup (composite)
     Note: FMVS = 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 (2005): U.S. EPA (2012a); PINFA (2010).
      1.2.2.  Flame-Retardant Materials as Solutions to Flammability

1             The flammability of textiles and the standards described above have created a growing market

2     demand for technologies to increase flame resistance and meet fire safety regulations (Alaee. 2003). One

3     way to achieve this is through the use of flame-retardant materials, which are chemicals or other

4     manufactured components that have the quality of resisting or inhibiting the spread of fire. Even where

5     regulatory standards do not mandate flame resistance, market pressures and concerns about brand image

6     often cause manufacturers to incorporate flame-retardant materials into their products (Illinois

7     Environmental Protection Agency. 2007). In fact, the global market for flame-retardant materials is
      March 2013
                       1-17
                           DRAFT-DO NOT CITE OR QUOTE

-------
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
      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 1.2.2.2]. 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:
                                                      Factors influencing flame retardant selection
                                                   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:
        •   Flame test performance: a measure of the efficacy of the flame-retardant material; different
            measures are included in specific regulatory standards;6
        •   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).
      6For 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:BabrauskasandKrasny. 1985)1.
      March 2013
                                                1-18
                                                                             DRAFT-DO NOT CITE OR QUOTE

-------
 1            Another important aspect of performance is durability. Durability is a measure of the ability of a

 2    flame-retardant material to maintain an acceptable level of flame-retardant behavior throughout the

 3    lifetime of the textile as it undergoes abrasion, laundering, weathering, or other expected processes

 4    (PINFA. 2010; NRC. 2000). The durability standard required depends on the intended use of a textile

 5    product. Durability classifications for flame-retardant finishes in textiles are presented in Table 1-4.7 In

 6    some cases, an evaluation of durability is a component of the flame tests (e.g., both pre- and post-wash

 7    tests are required for some product uses).8 Importantly, these criteria pertain only to the performance of a

 8    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
      number of washes
Cyanamide and phosphoric acid,
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).
      1.2.2.2. Flame-Retardant Application Methods

 9            Two principal processes are used for incorporating flame-retardant materials into the textile

10    matrix: reactive and additive. Generally, flame retardants incorporated into the textile matrix using the

11    reactive process produce durable finishes; flame retardants simply added to the textile matrix produce

12    nondurable or semidurable finishes (U.S. EPA. 2005; Rahman et al.. 2001).

13            In the reactive process, flame-retardant materials are incorporated directly into polymeric

14    materials during the manufacturing process such that they are chemically (i.e., covalently) bound to the

15    raw materials of the final product (U.S. EPA. 2005; Rahman et al.. 2001). Direct incorporation also can
      '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.
      8The Federal Register specifies which textile types and products require flame resistance for up to a specific number
      of washes.
      March 2013
                   1-19
  DRAFT-DO NOT CITE OR QUOTE

-------
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
_Q
 ru
            Pad
                    Heat
                    Cure
      • Wash
                Dry
                                              A
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 (ftinctionalization) (PINFA. 2010; Laoutid et al.. 2009).
Flame-retardant materials produced by the reactive mechanism are often 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
                   Chemical
                   * Cure
        Fix/
       Wash
                                 Dry
            Back-
            Coat
Dry/
Cure
            Pad
                      Dry
                                             o
                                             ro
CTQ
U
o>
D"
                             Source: Adapted from NRC (2000).
Figure 1-4. Durability of additive flame retardants.
This general schematic of the "additive" application method of flame-
retardant materials for textiles demonstrates the steps in the additive flame-
retardant process that increase durability. The curing process can result in
cross-linking, thermal fixation, or ionic linkage between the flame-retardant
material and the fibers to increase durability.
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).
       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,
March 2013
  1-20
    DRAFT-DO NOT CITE OR QUOTE

-------
 1    leaching of flame-retardant material remains of greater concern for additive flame retardants than for
 2    reactive flame retardants (not shown in Figure 1-4) because the material is not covalently bound to the
 3    substrate (Rahman etal., 2001).
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
1.2.2.3. 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 etal.. 2001):
                                                       Source: Adapted from Alaee et al. (2003) and Laoutid (2009).
                                                     Figure 1-5. The combustion process.
                                                     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.
       •   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
      March 2013
                                              1-21
                                                                    DRAFT-DO NOT CITE OR QUOTE

-------
1    flame retardants are added to textiles, and the general mechanism of flame-retardant action for each class

2    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
                          modifications or grafted onto
                          polymer chains for reactive
                          application
                         Decreases thermal degradation;
                         reduces extinguishing time
     Phosphorous-based Organophosphorous,
                     Inorganic phosphates
                          Coatings; chemical bath
                         Protective coatings or layers; char
                         formation
     Nitrogen-based     Melamine,
                     Melamine salts
                          Intumescent coatings; back-
                          coatings; can be added to
                          polymer melt
                         Inert gas dilution (inhibits formation
                         of flammable gases); char formation
     Inorganic
Metal hydroxides,
Minerals
Fillers; back-coatings; can be
added to polymer melt
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 (2009).
     1.3.  DecaBDE and MWCNTs  in  Flame-Retardant Textiles

3           As noted previously, the purpose of this case study is to present available information that
4    supports research planning for conducting a comparative CEA in the future of a traditional flame
5    retardant (e.g., decaBDE) and a nanoenabled flame-retardant technology using MWCNTs, specifically in
6    upholstery textile coatings. The following sections provide a general overview of decaBDE and
7    MWCNTs, their use in textiles, and a brief comparison of observed flame-retardant action and efficacies.



     1.3.1. Introduction to DecaBDE

8           DecaBDE is part of a larger group of BFRs called polybrominated diphenyl ethers (PBDEs), a
9    group of 209 structurally similar BFRs that differ in the number and location of bromine atoms
     March 2013
                              1-22
                           DRAFT-DO NOT CITE OR QUOTE

-------
 1    (Table 1-6) (Rahman etal.. 2001; NRC. 2000). Although PBDEs are typically categorized into classes by
 2    number of bromine atoms (e.g., PBDE with two bromine atoms is a diBDE; ten bromine atoms is a
 3    decaBDE), a single class might contain several different PBDE congeners with the same number of
 4    bromine atoms in different locations (i.e., PBDE BFRs can have many isomers). As the only fully
 5    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-47,
BDE-85,
BDE-153
BDE-181
BDE-196
BDE-206
BDE-209
BDE-25, BDE-28, BDE-30, BDE-32, BDE-33, BDE-35, BDE-37
BDE-49, BDE-66, BDE-71, BDE-75, BDE-77
BDE-99, BDE-100, BDE-105, BDE-116, BDE-118, BDE-119, BDE-126, BDE-138, BDE-140
, BDE-154, BDE-155, BDE-166
, BDE-183, BDE-190
, BDE-197, BDE-203
, BDE-207, BDE-208

      Source: U.S. EPA (201 Ob).
 6           Commercial formulations of decaBDE (see Table 1-7) are generally 97-98% BDE-209 with less
 7    than 3% nonaBDE congeners present as impurities (Rahman et al.. 2001; NRC. 2000) (see Appendix B.
 8    Table B-l for analytical techniques used to distinguish PBDE congeners in samples). Although the terms
 9    decaBDE and BDE-209 often are used interchangeably, this case study primarily uses the term decaBDE
10    to refer generally to the flame-retardant formulation and BDE-209 to refer to the specific decaBDE
11    congener analyzed in scientific studies.
12           DecaBDE is the most widely used of the PBDEs and has been well studied. In 2001, decaBDE
13    use accounted for 83% of total PBDE production worldwide (U.S. EPA. 2010b); an estimated 10-20% of
14    decaBDE use is in the textile industry (Pure Strategies Inc.. 2005). At the end of 2004, both octa- and
15    pentaBDE were voluntarily withdrawn from the United States marketplace due to evidence of
16    environmental persistence and toxicity, which left decaBDE as the sole PBDE available for use in
17    commercial products in the United States (U.S. EPA, 201 Ob). As mentioned in Section 1.1.4. two U.S.
18    producers made a commitment on December 17, 2009, to phase out decaBDE in the  United States by
19    2013, due to similar concerns over environmental persistence and toxicity. Several standard

      March 2013                                    1-23                  DRAFT-DO NOT CITE OR QUOTE

-------
 1    physicochemical properties are used to describe traditional chemicals: melting point, boiling point,
 2    molecular weight, and others. Such values are presented for decaBDE in Table 1-8.

      Table 1 -7.    Commercial formulations of PBDEs used as flame retardants.
Name
Penta formulation1
Octa formulation
Deca formulation2
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%)

      1Trace amounts of additional congeners might be present in commercial formulations: <0.2% triBDE congeners.
      2Trace 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].
 3           DecaBDE can be applied to textiles by a variety of mechanisms, but this case study focuses on
 4    the application of decaBDE as a back-coating. This application method is used most frequently for
 5    decaBDE (Pure Strategies Inc.. 2005; NRC. 2000) and is most similar to the application method expected
 6    for MWCNTs used in textiles (see Section 1.3.2). The back-coating process usually involves mixing
 7    decaBDE with a copolymer or resin binder (Pure Strategies Inc.. 2005; NRC. 2000). DecaBDE combines
 8    the flame-retardant mechanism of most BFRs (i.e., releasing halogens during combustion to compete with
 9    the availability of oxygen for the flame) with formation of a protective char barrier (NRC. 2000) that
10    interferes with the spread of the flame and helps the material to self-extinguish (Pure Strategies Inc..
11    2005).
      March 2013
1-24
DRAFT-DO NOT CITE OR QUOTE

-------
Table 1 -8. Physical properties and chemical identity of decaBDE.

CASRN
Synonyms
Physical state
Melting point,
Boiling point
Vapor pressure
Henry's law constant
Density
Water solubility
Log Kow
Log Koc
Molecular weight
Chemical formula
Chemical structure
Physical property/chemical identity
1163-19-5
2,2',3,3',4,41,5,51,6,61-decaBDE; BDE-209; benzene, 1,1'-
oxybis[2,3,4,5,6,-pentabromo]-; decabromodiphenyl oxide;
decabromodiphenyl ether; decabromobiphenyl ether; ether,
bis(pentabromophenyl)
Solid
300-310 °C
Decomposes at >320 °C
4.63x10-6 Pa at 21 °C
1.93x1 0-s Latm/mol
0.04Pam3/molat25°C
3.0 grams/cm3
<0.1ug/Lat25°C
6.3-12.6
6.3
959.17
Ci2BnoO
Br Br
B^Y^Y"0^^^6'
Bf^X-S^B, Br-^s^8'
Br Br
Reference
NLM (2011)
NLM (2011):
ATSDR (2004)
Hardy (2002b)
ECB (2003)
ECB (2003)
Hardy (2002b)
Hardy (2002b):
Cetin and Odabasi (2005)
NRC (2000)
Hardy (2002b): ECB (2003)
Hardy (2002b)
Hardy (2002b)
NLM (2011): ECB (2003)
NLM (2011)

     Note: Kow = Octanol water partition coefficient, Koc = Soil organic carbon-water partition coefficient.

     1.3.2.  Introduction to MWCNTs
1           MWCNTs are carbon nanostructures composed of multiple concentrically nested graphene sheets
2    that look similar to nested rolls of chicken wire. Unlike many traditional chemicals, MWCNTs are not a
3    homogeneous group of molecules; many of the characteristics of MWCNTs can be intentionally or
4    unintentionally altered using different laboratory procedures, treatments, and synthesis methods (see
5    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-
6    1, altering the physicochemical properties of MWCNTs can alter their behavior during all stages of the
7    life cycle, in environmental compartments, and in humans and other biota. As a result, MWCNTs with
8    different physicochemical properties might produce different impacts downstream, but which
     March 2013
1-25
DRAFT-DO NOT CITE OR QUOTE

-------
1    physicochemical properties might drive these differences and to what degree are not fully understood.

2    MWCNTs are not a single homogeneous substance; ranges of values describing MWCNT

3    physicochemical properties are provided in Table 1-9 to illustrate the array of characteristics recorded for

4    MWCNTs in the literature (see Appendix B. Table B-2 for analytical techniques used to characterize

5    MWCNTs).
                 Text Box 1-1. Physicochemical Properties of Multiwalled Carbon Nanotubes (MWCNTs)
                  Affect Their Release, Behavior in the Environment, and Interaction with Biota
                                                        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

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 likely to be taken up by
                       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
      March 2013
                                                    1-26
                                                           DRAFT-DO NOT CITE OR QUOTE

-------
Table 1 -9.     Physical properties and chemical identity of MWCNTs.
                Physical property/chemical identity1'2
                                                                   Reference
 Physical state
Solid
 Morphology
Concentric cylinders
Johnston et al. (2010)
 Physical structure
                                                 Hirsch and Vostrowsky (2005)
 Purity,4 min. wt% C  Usually >90%; 7.5-40% reported for "multi-wall,
                    powdered cylinder cores" and "multi-wall, as
                    produced" by Sigma-Aldrich
                                                 See Appendix F study summaries; Sigma-Aldrich
                                                 (2012)
 Outer diameter
5-170 nm3
Aschberger et al. (2010): Li and Huang (2011):
Desai et al. (2012): He et al. (2012): Dawson et al.
(2011): Golovin et al. (2011): Lu et al. (2011b): Liu
et al. (2011a): Ji et al. (2011): Sigma-Aldrich
(2012)
 Length
20 nm-200 urn
Aschberger et al. (2010): He et al. (2012):
Aranberri et al. (2011): Golovin et al. (2011): Lu et
al. (2011b): Liu et al. (2011a): Ji et al. (2011):
Sigma-Aldrich (2012)
Aspect ratio
Chirality
Axial and radial
strength
Surface area
Bundle size
Surface
composition
Vapor pressure
Melting point
Stability
Density
Zeta potential, mV
Solubility in water
Up to 1,000
Varies; chiral angles, described by vectors (n, m),
produce different graphene sheet conformations (e.g.,
"zigzag" [m = 0], armchair [n = m]), and influence
other properties (e.g., mechanical, optical, electrical)
Axial: rigid; 10 times stronger than steel
Radial: flexible; can be bent up to 90 degrees
253-400 rrWgram
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)
2.1grams/mLat25°C
-23-0
Insoluble; functionalization treatments result in
different degrees of solubility
Cipiriano et al. (2007)
Gustavssonetal. (2011)
Gustavssonetal. (2011)
Aschberger et al. (2010); Aranberri et al. (2011):
Luetal. (201 1b):
Li and Huang (2011): Baitinger et al. (2011):
Johnston et al. (2010)

Sigma-Aldrich (2012)
Nanoshel (2011)
Sigma-Aldrich (2012)
Li and Huang (2011)
Lam et al. (2006): Johnston et al. (2010)
 1Values reported represent total ranges reported in literature.
 2Values 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.
 4lmpurities 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).
March 2013
                                    1-27
             DRAFT-DO NOT CITE OR QUOTE

-------
  1
  2
  3
  4
  5
  6
  7
  8
  9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
        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 etal.. 2005a). To date, at
least one MWCNT flame-retardant textile
coating is commercially available, but this
application does not appear to be widespread
(Nanocvl. 2009V Goncalves et al. (2012)
describe the incorporation of MWCNTs into
cotton or polyester using a process similar to
industrial dyeing methods. In addition, a patent
is currently 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).  Both the
work by Goncalves et al. (2012) and the recent
patent indicate that more commercial
applications of MWCNTs in textiles might be
       Additional Information Highlight Box 2:
    MWCNTs are not widely used in flame-retardant
                       textiles
  Use of MWCNTs in flame-retardant textiles is not currently a
  widespread application, nor is it anticipated to become one of
  the more widespread applications of MWCNTs in the near
  future. While global annual production capacity for MWCNTs
  has increased rapidly (Kohleretal., 2008) and is projected to
  reach 9,400 tons by 2015 (Innovative Research and Products
  Incorporated, 2011), most MWCNT production in the United
  States currently occurs in research labs or small-scale pilot
  manufacturing facilities (Schubauer-Berigan et al., 2011).
  Although one commercial MWCNT flame-retardant coating has
  been developed [(Luizi, 2009): Personal Communication:
  Nicolas Messin (Nanocyl). 3/2/2012],the proportion of total
  global MWCNT production used in textiles is anticipated to be
  very low (see Table 1-10 and Table 2-2). Future use of
  MWCNTs in flame-retardant 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), as well as the feasibility of
  large-scale production processes. Although researchers have
  already demonstrated the ability to incorporate MWCNTs into
  fabrics using a process that mirrors industrial dyeing methods
  (Goncalves etal., 2012), 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(Schnorrand Swager, 2011:  Kohleretal., 2008).
  Although greater production volumes of other MWCNT
  applications was one consideration in developing this case
  study, other factors, such as exposure potential, also were
  important (see Section 1.1).
available in the near future. 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: Bever. 2004: Kashiwagi et al.. 2004).
        The flame-retardant behavior of MWCNTs depends on the formation of a highly uniform,
network-structured layer of floccules, which are loosely bound MWCNT bundles, with no breaks or
cracks. The formation of the 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,
March 2013
1-28
                                                                               DRAFT-DO NOT CITE OR QUOTE

-------
 1    aspect ratio,9 and loading concentration (Cipiriano et al., 2007; Kashiwagi et al., 2007; Kashiwagi et al.,
 2    2005b; Kashiwagi et al., 2005a; Kashiwagi et al.. 2004). Cipiriano et al. (2007) were able to produce a
 3    more uniform floccule layer at lower concentration loadings, resulting in enhanced flame-retardant
 4    properties, by using MWCNTs with a higher aspect ratio. In nanoclay, the incorporation of
 5    organomodified montmorillonite, but not sodium-layered montmorillonite, stimulated char formation in
 6    polymer matrices, indicating that surface functional groups can be instrumental in flame-retardant action
 7    (Laoutid et al.. 2009).


      1.3.3. MWCNTs as Alternative Flame-Retardant Materials in Upholstery
      Textiles
 8           As the use of decaBDE begins to decline, cost-effective and feasible alternatives to replace this
 9    widely used flame retardant are being evaluated. MWCNTs are one of many possible alternatives to
10    replace decaBDE. Given the current, albeit limited, availability of a MWCNT product for textile
11    applications10 and the projected decline in cost (Sullivan. 2009). the use of such nanoenabled products is
12    likely to increase in the future as an emerging application. Table  1-10 provides a comparative summary of
13    decaBDE and MWCNTs, which illustrates several similarities in application method, flame-retardant
14    action, and relevant uses. Many parallels can also be drawn in the performance criteria (discussed in
15    Section 1.2.2.1) as shown in Table  1-11.
16           A summary of information available from actual flame tests for decaBDE and MWCNTs is
17    provided in Table 1-12. Note that information available for flame tests for both materials is not
18    standardized. Some information is qualitative, while other data are quantitative. In both cases, only a few
19    representative examples are described; in the absence of specific data on MWCNTs used in textiles
20    relevant to this case study, available information has been provided on a similar MWCNT flame-retardant
21    product used as a coating for a variety of materials. This product is used on foam and other structural
22    materials, but might prove useful for comparison to the current application because it generally has been
23    shown to increase flame resistance, increase the heat barrier and charring, and reduce the amount of
24    smoke created. Furthermore, the manufacturer advertises that their MWCNT flame-retardant coating is
25    appropriate for application to textiles (Mezzo. 2010). Table 1-13  describes properties of MWCNTs that
26    impact performance as a flame retardant.
      9This 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).
      10Personal Communication: Nicolas Messin (Nanocyl). 3/2/2012.
      March 2013                                    1-29                  DRAFT-DO NOT CITE OR QUOTE

-------
Table 1-10.   Overview of decaBDE and MWCNTs for flame-retardant textile application.
               MWCNTs
                                                  DecaBDE
Method of      Integrated by "melt blending" with polymer (Cipiriano et
incorporation    al., 2007: Kashiwagietal., 2005b: Kashiwagietal.,
into textile      2005a: Zhang and Horrocks, 2003): dispersed in resin
products        (e.g., silicon base) and applied as a coating (Nanocyl1)
               (Kohleretal., 2008): applied by "dyeing-like" method to
               fabrics (e.g., cotton) (Goncalves etal., 2012): "layer-
               by-layer coating" (foam applications) (Uddin and
               Nyden, 2011a: Davis and Kim, 2010)
                                                  Applied as a back-coating to textiles (NRG, 2000),
                                                  often with a binding agent such as latex (ECB, 2003),
                                                  or a copolymer (NRG, 2000)
Mechanism of   MWCNT network acts as a sealing or shielding agent
flame-          (i.e., a barrier) (Berger, 2007: Cipiriano etal., 2007:
retardant       Kashiwagi etal., 2007: Kashiwagi etal., 2005b:
action          Kashiwagietal., 2005a: Kashiwagietal., 2004)
                                                  Gas-phase radical quenching; creates a char barrier
                                                  (NRG, 2000)
Approximate
production
volume/
capacity2
From 2005 to 2009, global annual production capacity
increased from 294 tons (approximately 267 tonnes) to
more than 1,500 tons (approximately 1,361  tonnes)
(Kohleretal., 2008): projected to reach 9,400 tons
(approximately 8,528 tonnes) by 2015 (Innovative
Research and Products Incorporated, 2011): percent
of MWCNTs produced for textile use expected to be
very low
Worldwide demand in 2001 reported as 54,000-56,000
tonnes (Lawetal., 2006: Pure Strategies Inc., 2005):
>60,000 tonnes reported in 2007 (Illinois
Environmental  Protection Agency, 2007): 10-20% of
decaBDE produced is used in textiles3
Relevant use    Has been tested in "nanocomposites" with polyvinyl
in textile        acetate, and ethylene vinyl acetate (Kashiwagietal.,
applications     2005b: Kashiwagietal., 2005a: Kashiwagietal.,
               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 etal., 2012: Alimohammadi
               etal.. 2011)
                                                  Used in mattresses, draperies, commercial
                                                  upholstered furniture, and in transportation industry
                                                  fabrics (Pure Strategies Inc., 2005)
1See (Nanocvl. 2009: Sullivan. 2009).
2Note: not all of the production capacity is relevant for flame retardants or for use in upholstered textiles.
According to U.S. EPA (20|0b), 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.
March 2013
                                        1-30
                DRAFT-DO NOT CITE OR QUOTE

-------
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 (Cipirianoetal., 2007; Kashiwaqietal.,
                  2005b; Kashiwaqi etal., 2005a; Kashiwaqi etal.,
                  2004); increased limiting oxygen index (LOI)1
                  (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) (Grzybowski, 2009; Kashiwagietal., 2005b;
                  Kashiwagi etal., 2005a; Kashiwagi etal., 2004):
                  commercial formulations for textile use are reported
                  to be effective at 100 jig thickness2
                                                 Very efficient, can be used at relatively low
                                                 concentrations (Pure Strategies Inc., 2005: Rahman
                                                 etal., 2001): maximum of approximately 20% w/w
                                                 added as a back-coating (NRG, 2000):3 applied 10-
                                                 15% by weight to polymers in conjunction with resin
                                                 binder (U.S. EPA. 2010b)
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 textile  Can improve physical and mechanical properties
characteristics     (Siegfried, 2007: Hirsch and Vostrowsky, 2005:
                  Kashiwagietal., 2005b), as well as conductivity and
                  optical properties4 (Siegfried, 2007): fatigue
                  resistant; particle embedding can prevent cracks
                  (Grzybowski, 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;2 potential for nanotextiles to release
                  individual nanoparticles or clusters of nanoparticles
                  (Grellleretal.,2010)5
                                                 Semidurable (Rahman etal., 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.
2Personal Communication: Nicolas Messin, Global Sales and Marketing Manager for Thermosets (Nanocyl). 3/2/2012.
3Depends on the PBDE used (both decaBDE and hexaBDE referenced), the resin binder used, and the fabric to be treated.
"Optical properties of textiles include fluorescence or color-changing effects (theoretical; not necessarily relevant for MWCNTs in upholstery).
5Based 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.
March 2013
                                      1-31
                DRAFT-DO NOT CITE OR QUOTE

-------
                                  Additional Information Highlight Box 3:
         MWCNTs are likely used in combination with other chemicals in flame-retardant applications

  Although MWCNT applications demonstrate the ability to confer flame-retardant properties to a wide range of polymers and
  textiles (Goncalvesetal.,2012: Grzybowski, 2009: Mahy, 2009: Howlett, 2008), the physical barrier properties (i.e., the ability
  of MWCNTs  to produce a barrier to  seal or shield materials from  igniting) of MWCNTs  alone are 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) (Morgan, 2006: Bartholmai and Schartel, 2004). In many
  instances,  however, nanomaterials  can  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  to pass flammability tests in most applications, including upholstery textiles (Morgan, 2006:
  Schartel  et al., 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 (Maetal., 2011: Luand Wilkie, 2010: Beyer, 2005).
 Table 1-12.    Flame test performance of decaBDE and MWCNTs.
           Sample
Without Treatment
With Treatment
Source
MWCNTs1  Polyurethane
           foam
Burns quickly, dense smoke forms,
burning droplets fall
Spray coating forms "shell" to keep  Mahy (2009):
molten foam contained (no burning  Howlett (2008)
or flaming drops)
           Polyvinyl
           chloride
Burns easily, melts, structure destroyed   Does not melt, structure is retained  Mahy (2009
           Wires/Cables    Burns completely, releases dense
                          smoke and burning droplets, copper
                          core becomes exposed
                                     No burning droplets; low smoke
                                     density; copper core protected;
                                     passes UL94,2 IEC-332-33 tests
                                Mahy (2009
           Polypropylene   Heat release rate = 2,800 kW/m2
                                     1-2% addition = heat release rate   Gryzybowski (2009)
                                     of800kW/m2
           Cotton          Burned distance = 77 mm; burning time  Burned distance = 80 mm; burning  Goncalvesetal.
                          = 19 sec; burning rate = 243 mm/min     time = 21 sec; burning rate = 229   (2012)
                                                              mm/min

           Polyester       Burned distance = 66 mm; burning time  Burned distance = 66 mm; burning  Goncalvesetal.
                          = 26 sec; burning rate = 152 mm/min     time = 53 sec; burning rate = 75    (2012)
                                                              mm/min

DecaBDE   Sofa           Burns quickly (<5 minutes), temperature  Burns slowly, increases amount of  BSEF (2012)
                          increases from 20 °C to 800 °C         time to escape by up to 15 times
           50/50 polyester/ Sample burns completely
           cotton twill
Char length reduced (<6.3 in.)
                                                                     ICL Industrial
                                                                     Products (2010)
1Due 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.
Horizontal and vertical burning tests associated with American Society for Testing and Materials (ASTM) International standards.
3Large-scale flammability test for wire bundles under 20.5 kW flame.
 March 2013
                             1-32
                  DRAFT-DO NOT CITE OR QUOTE

-------
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 (Kashiwaqi et al., 2007; Kashiwaqi et al., 2005b; Kashiwaqi et al., 2005a; Kashiwaqi et
al.. 2004).
Commercially available MWCNTs for flame resistance are approximately 1 .5 urn in length [Nanocyl; (Howlett,
2008)1. Shorter MWCNTs (1-2 um) are more flame-retardant than lonqer MWCNTs (0.5-40 um) in polymer
blends (Pack etal., 2009).
MWCNTs that are "crushed" increase the time to ignition compared to uncrushed MWCNTs in polymer
nanocomposites (Laoutidetal., 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 (Laoutidetal., 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 (Mulejaetal., 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
(Goncalves etal. ,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 etal.. 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) (Sonq 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 (Cipirianoetal.,
2007).

March 2013
1-33
DRAFT-DO NOT CITE OR QUOTE

-------
Table 1-13, cont: 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 as the
Concentration loading concentration does not cause the CNTs to be poorly dispersed (i.e., to agglomerate) in polymers
              (Kashiwagi etal., 2005b: Kashiwagi etal., 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 (Kashiwagi etal., 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 (Cipirianoetal., 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 etal., 2011).
March 2013                                          1-34                     DRAFT-DO NOT CITE OR QUOTE

-------
This page intentionally left blank.

-------
                    Chapter 2.  Product Life Cycle
 1           A product's life cycle encompasses all stages of its existence from "cradle to grave," starting with
 2    the extraction of raw materials from the earth for the manufacture of the product and continuing
 3    downstream until these materials are returned to the environment following disposal (U.S. EPA, 2006).
 4    The components of the life cycle determine the potential for releases and possible impacts on human
 5    health, ecological populations, and the environment (Som et al.. 2011), which can be evaluated
 6    systematically within the framework of a comprehensive environmental assessment (CEA). Potential
 7    environmental impacts of a product throughout its life cycle can be estimated using  a life-cycle
 8    assessment (LCA) approach, which involves four steps: goal definition and scope, inventory analysis,
 9    impact analysis, and interpretation (U.S. EPA. 2006). The CEA approach incorporates information from
10    available LCAs in the "product life cycle" and "impacts" portions of the CEA framework to combine this
11    knowledge with other analyses or qualitative indicators related to transport, transformation, and fate,
12    exposure-dose, and additional impacts not considered in available LCAs. As discussed in Chapter 1. if a
13    plausible reason exists to include an impact in the CEA framework, information (qualitative or
14    quantitative) on that effect  can be included from LCAs or other sources (if an LCA has not been
15    completed) to evaluate that particular impact.
16           A generalized depiction of the life cycle for multiwalled carbon nanotube (MWCNT) coatings
17    used to confer flame-retardant properties to upholstery textiles is presented along with comparable
18    information for decabromodiphenyl ether (decaBDE) in Figure 2-1. This figure breaks down the life cycle
19    of these materials into five  main stages: (1) acquisition and processing of feedstocks, (2) manufacturing
20    (including research and development (R&D) processes), (3) storage and distribution, (4) use, and (5) end-
21    of-life processes (including disposal, reuse, and recycling). These stages correspond roughly to the four
22    primary life-cycle stages outlined by the U.S. Environmental Protection Agency (U.S. EPA. 2006).
23    including raw materials acquisition, manufacturing, use/reuse/maintenance (with storage and distribution
24    discussed as a distinct stage in this case study), and recycle/waste management. As mentioned in Chapter
25    i, R&D is included in the product life-cycle portion of the CEA framework, given its importance
26    regarding emerging materials such as MWCNTs. For such materials, R&D efforts can elucidate potential
27    risks associated with commercial-scale manufacturing. In fact, because it often takes place when health
28    and safety information is being developed for a material, R&D presents an ideal opportunity to gather
29    data on a product's potential impacts and to make design adjustments if appropriate.
      March 2013                                     2-1                  DRAFT-DO NOT CITE OR QUOTE

-------
                        life
                      Cycle
                                              Release
                                             Scenarios
                                                               DecaBDE
                                                             Release Form
                                                                                                       CNT
                                                                                                 Release Form1
       I
     Raw materials from which decaBDE
        or MWCNTs are synthesized
                         1
                                 Extraction
                                 Processing
                                                                           N/A
N/A
        01
        a.
       _o
        01
       Q
       •o
       I
        n
            Material Synthesis
        Production of decaBDEs and
                MWCNTs
                  Material Processing
               Purification/modification of
                 decaBDE and MWCNTs
          Product Manufacturing
       Formulation and application of
       flame retardantto textiles and
      textile finishing and upholstering
                         i
                                 Synthesis
                                 Recovery
                                 Handling/packaging
                                 Equipment cleaning
                                 Accidents

                                 Purification
                                 Functionalization
                                 Dispersal
                                 Handling/packaging
                                 Equipment cleaning
                                 Accidents

                                 Formulation
                                 Application
                                 Processing
                                 Textile/furniture processing
                                 Equipment cleaning
                                 Accidents
                                                                           N/A
                                                                           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
I
       5
       •a
        01
        5?
        o
       y,
          Material
Supply of decaBDE and MWCNTS
      Flame Retardant
   Supply of flame retardant
          Product
   Supply of flame-retardant
      upholstery textiles
                                       Accidents
                                                                    Free, Matrix-Bound    Free, Bundled, Matrix-Bound
                  i
                Intended and unintended/
             accidental use of flame-retardant
                   upholstery textiles
                                              Intended (sitting, cleaning, etc.)
                                              Unintended/accidents
                                                                    Matrix-Bound
                                                                    Free, Matrix-Bound
                                                                                  Matrix-Bound
                                                                                  Free, Matrix-Bound
                                              Reuse/repurposing
                                              Recycling
                                              incomplete incineration
                                              Landfilling
                                              Wastewater 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
                       Material
              Reuse, recycling, and disposal of
                 decaBDE and MWCNTS
       •2
       j          Flame Retardant
       "S     Reuse, recycling, and disposal of
       ~°           flame retardant
                       Product
              Reuse, recycling, and disposal of
             flame-retardant upholstery textile

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 etal. (1999]: NRG (2000]: Palm etal.  (2002]: Agrell etal. (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.
N/A = not applicable; CNT = carbon nanotube
March 2013
                                                      2-2
                                                                             DRAFT-DO NOT CITE OR QUOTE

-------
 1    Similarly, as discussed below, differences between R&D activities and the commercial manufacturing
 2    process (e.g., use of protective equipment, volume of material produced) could be important
 3    considerations in mitigating potential risks to individuals involved in R&D versus commercial
 4    manufacturing.
 5            To conduct a comparative CEA, relevant information on life-cycle inventories from existing
 6    LCAs would be incorporated into the product life cycle to characterize the inputs (e.g., raw materials,
 7    energy) and outputs (e.g., emissions to air and water, coproducts) associated with each material's
 8    manufacture. Impacts information from existing LCAs also would be considered (see Chapter 5). Other
 9    LCA aspects also might apply, including using an appropriate functional unit, which is a quantitative
10    measure of a product's function or a process that facilitates comparison (U.S. EPA. 2006). In the current
11    case study, a functional unit might correspond to the degree of flame retardancy conveyed by
12    incorporation of a certain amount of MWCNTs. In general, for this case study, data that specify
13    appropriate functional units were not identified; the reader might, however, consider how this aspect of
14    existing or future LCAs could be incorporated into a future CEA when evaluating data gaps and needs.
15    This chapter outlines important aspects of each  of the five life-cycle stages outlined in Figure 2-1 for
16    MWCNTs used in upholstery textiles. This chapter also includes descriptions of the important
17    environmental release scenarios for MWCNTs across the product life-cycle stages based on current
18    knowledge. A variety of release scenarios are possible throughout the life-cycle stages described in this
19    chapter. Figure 2-1 also outlines potential release scenarios for MWCNT flame-retardant upholstery
20    textile coatings throughout the life cycle along with potential forms of the released substances (i.e., free,
21    bundled, or matrix bound). The term free MWCNTs refers to pure, unbound materials. The term
22    MWCNT bundles refers to clusters of MWCNTs loosely or tightly bound together.11'12. The terms matrix-
23    bound decaBDE and MWCNTs refer to these materials as a part of a polymer matrix (e.g., the flame-
24    retardant formulation). Additionally, as illustrated in Figure 2-2. MWCNT formulations can be altered  at
25    multiple stages of the product life cycle, meaning the formulation of the MWCNT released at different
26    stages of the product life-cycle can vary.
      nThe 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.
      12MWCNTs, could adsorb to dust particles (see Section 3.2). Although the ways that adsorption to dust could
      facilitate release (i.e., offer a transport vector) from a product matrix are discussed in this chapter, the dust-MWCNT
      complex is not considered a "release form" because dust was not included in the original product matrix.
      March 2013                                       2-3                    DRAFT-DO NOT CITE OR QUOTE

-------
                              Raw Materials
                         Product Manufacturing
                       Product Storage, Transport
                           MWCNT
                           Physical/chemical changes that occur including: intentional
                           (e.g., product formulation) 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.1

1This 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 MWCNT moves through the environment. Alterations include both
physical and chemical changes, for example, 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. As a 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.
 March 2013
2-4
DRAFT-DO NOT CITE OR QUOTE

-------
     2.1. Feedstocks
                                 Unprioritized Research Area: Product Life Cycle
       0)  IV
       E  £
       ED
          o
          Q.
          E
                      3/13
5/13
                      5/13
                                                            Release Rate
                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
      Three out of 13 RTI workshop participants (23%) identified the raw materials stage of the product life cycle as important to risk
      assessment. Based on this information, this stage in the product life cycle 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.
     2.2. Manufacturing
1            The manufacturing stage for MWCNT flame-retardant upholstery can be viewed as a sequential

2    process involving synthesis, material processing (i.e., purification and modification), and product

3    manufacture (i.e., formulation of the flame-retardant mixture, application of the flame-retardant mixture

4    to textiles, and incorporation of the flame-retardant textile into consumer or commercial goods). R&D

5    also is included in this section, given the similarities to key aspects of synthesis, processing, and

6    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).
     March 2013
                                2-5
                     DRAFT-DO NOT CITE OR QUOTE

-------
      2.2.1.1. Life-Cycle Processes
 1            Research on MWCNTs and on flame-retardant coatings involving MWCNTs is principally
 2    conducted in specialized laboratory environments. Research and development (R&D) activities are
 3    expected to be carried out by individuals rather than automated mechanisms used in commercial-scale
 4    manufacture. The processes of interest to researchers are similar to those used in commercial-scale
 5    manufacture of these materials: synthesis, purification, modification, dispersion, incorporation into flame-
 6    retardant formulations, and application to textiles. Current research efforts focus on synthesis and
 7    purification methods to improve CNT quality and purity (Kohler et al.. 2008). The  following sections
 8    (material synthesis, material processing, and product manufacturing) provide detailed information on the
 9    processes of potential interest for R&D.

      2.2.1.2. Potential  Releases during the R&D Stage
10            Release scenarios during the R&D stage are expected to be similar to release scenarios from
11    commercial synthesis described in the following sections, but the quantities released are anticipated to be
12    much smaller in the R&D stage. The quantities of MWCNTs handled in research laboratories are much
13    smaller than those handled in commercial-scale manufacturing facilities. Although R&D activities are
14    typically carried out in laboratories with specialized pollution control systems in place, including fume
15    hoods, ventilation systems, and environmental control systems, not all facilities have standardized
16    engineering controls. For example, these practices might not be in place for small start-up operations.
17    Given the experimental and somewhat unpredictable nature of R&D, releases from handling materials
18    during synthesis, processing and purification, storage, and analysis are possible.
19            As discussed in more detail in Section 4.1.2.4, multiple studies have collected particles and fibers
20    in workplace air to attempt to estimate MWCNT concentrations at the emission source, in area air, and in
21    the personal breathing zone of workers in small laboratories or research and development facilities
22    (Johnson etal.. 2010; Lee etal. 2010; Methneretal.. 2010; Bello et al.. 2008; Han etal.. 2008). One
23    industry report measured airborne release of CNTs (0.25 ug/m3) in the R&D facilities for a company that
24    manufactures Thermocyl®, an MWCNT flame-retardant coating application (Luizi. 2009). No data were
25    found that describe how  releases in academic labs compare with releases in commercial R&D labs,
26    though it is noted that estimates in Bello et al. (2008) were based on university laboratory settings.
      March 2013                                     2-6                   DRAFT-DO NOT CITE OR QUOTE

-------
2.2.2.  Material Synthesis
                                  Priority Research Area: Product Life Cycle
         Material
         Synthesis
  o o
  E £
  « «
  EJ t:
     o
     0.
     E
                   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:
                         M- Medium
                         H - High:
                         Confidence
                         L - Low:
                         M- Medium
                         H - High:
Least Important
Possibly Important
Important

Not Confident
Somewhat Confident
Confident
 Five out 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.

 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 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  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.
March 2013
                                    2-7
                      DRAFT-DO NOT CITE OR QUOTE

-------
      2.2.2.1. Life-Cycle Processes
 1            In 2010, chemical vapor deposition (CVD) synthesis produced approximately 83% of the global
 2    supply of CNTs, followed by arc-discharge synthesis (12%) and laser-ablation synthesis (5%) (PateL
 3    2011) (see Table 1-10 for information on MWCNT global production capacity). What proportion of
 4    MWCNTs is synthesized using each method, however, is unclear. Table 2-1 summarizes some of the
 5    performance characteristics of these three synthesis methods. Table 2-2 summarizes the current scale and
 6    projected growth of the CNT manufacturing industry, with details on the percentage of companies using
 7    each synthesis methods.
      Table 2-1.    Summary of common CNT synthesis methods.1
Characteristic
Growth temperature
Production
Scalability
Product quality
CVD
600-1,1 00 °C
Continuous
Scalable
Many structural defects
Long tubes
Low crystallinity2
Arc discharge
2,500-3,000 °C
Batch
Not currently scalable
Few structural defects
Short tubes
Carbon-containing metal impurities
Laser ablation
1,200°C
Batch
Not currently scalable
Few structural defects
Diameter control
      By-products
Over 45 side products, including      Black carbon and airborne
polycyclic aromatic hydrocarbons and  inorganic compounds
volatile organic compounds
          No Data
      1No data available that specifically describe MWCNTs.
      2Low degree of structural ordering.
      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).
 8           CVD synthesis takes place in two furnaces connected by a quartz tube (Healy et al., 2008).
 9    The catalyst mixture is heated before it is added to the furnace along with a carbon-containing gas
10    (Karthikeyan et al.. 2009; Healy et al., 2008). CNTs are recovered once the furnaces cool to room
11    temperature (Karthikevan et al.. 2009). MWCNTs can be grown on a substrate (e.g., Si/SiO2) or without a
12    substrate (Tsai et al., 2009). CNTs produced using deposition substrates are recovered by automated or
13    manual mechanical removal (Kohler et al.. 2008); however, specific details regarding the method of
14    recovery were not identified.
      March 2013
                          2-8
DRAFT-DO NOT CITE OR QUOTE

-------
      Table 2-2.    Current scale and projected growth in the CNT industry.
                    Parameter
                                          Manufacture Scale
               Pilot/Developmental
                      Scale
             Combined Manufacture
                   and Pilot/
              Developmental Scale
      Number of Employees per Company1
                                                2-100
                      1-30
                     1-130
      Quantity of CNT Produced per Year
      (kg/year)1
                                              0.2-2,500
                     0.1-300
                    0.1-2,800
                       Employee Count Year 12
                                                 172
                       20
                      192
                       Employee Count Year 2
                       (% change from Year 1)2
      Projected Industry 	
      Growth            Employee Count Year 3
                       (% change from Year 2)2
                                              196 (+14%)
                   43 (+115%)
                   239 (+24%)
                                             214 (+9.2%)
                    62 (+44%)
                   276 (+15%)
                       Total Percent Change
                       from Year 1 to Year 32
                                                +24%
                     +210%
                     +44%
                       CVD
CNT Synthesis
Method Used     	
(% of companies   Arch Discharge
using method)3    	
62%
NR
NA
                                                      23%
                       NR
                      NA
                       Flame Combustion
                                                      15%
                                                                        NR
                                              NA
                       Laser Ablation
                                                 8%
                       NR
                       NA
      1At 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.
      2Year 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.
      333% (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.
      NA = not applicable; NR = not reported
      Adapted from Schubauer-Berigan et al. (2011).
1             Synthesis by arc discharge involves passing an electric current between two graphite electrodes

2     (Healy et al.. 2008) in the presence of an inert gas. The anode contains a hole filled with carbon powder

3     and a catalyst, and the electric current results in the vaporization of the graphite anode and subsequent

4     condensation on the cathode and the walls of the reaction vessel (Baddour and Briens. 2005). CNTs

5     deposits form as black powder in the reaction vessel (Healy etal.. 2008). CNTs are generally recovered

6     from a receptacle after arc-discharge synthesis (Kohler et al.. 2008), but whether recovery is usually a

7     manual or automated process is unclear.
      March 2013
                                                     2-9
                             DRAFT-DO NOT CITE OR QUOTE

-------
 1           Laser-ablation synthesis of MWCNTs involves vaporizing a metal-graphite composite block in
 2    the presence of an inert gas and a catalyst (Karthikevan et al.. 2009). The composite block is placed inside
 3    an oven, a laser is pointed at the block, and argon gas is pumped parallel to the laser beam. As the laser
 4    ablates the target at high temperatures, CNTs form and are carried by the gas flow onto a collector
 5    (Karthikevan et al.. 2009). See Sections 5.3.2 and 5.3.4 for information on the estimated energy
 6    requirements and costs of CNT synthesis.

      2.2.2.2.  Potential Releases during the Material Synthesis Stage
 7           The potential release of MWCNTs during synthesis primarily depends on the synthesis and
 8    processing methods and the physical properties of the MWCNTs (e.g., size, bundling, density) (Kohler et
 9    al.. 2008). Because few data were available that describe releases from commercial-scale manufacture of
10    MWCNTs, this section also relies on CNT release data from R&D facilities. Although releases of
11    MWCNTs could occur during the synthesis stage, evidence  describing the likelihood and quantity of
12    release is mixed. In general, MWCNTs grown on substrates are likely to produce fewer airborne releases
13    than vapor-phase synthesis methods (Bello et al.. 2009; Tsai et al.. 2009). Mechanical removal (either
14    automated or manual) of CNTs from the substrate, however, can cause airborne release of CNTs (Kohler
15    etal.. 2008). In one study, CVD synthesis, both with and without a substrate, resulted in a concentration
16    at the source of synthesis of more than 2 to 3  x 106 particles/cm3 that measured less than 560 nm in
17    diameter (Tsai et al., 2009). but another study found no measurable airborne release of CNTs during
18    substrate-bound CVD growth of CNTs (Bello etal., 2009). One study conducted in three commercial
19    facilities and four research laboratories concluded that during synthesis, processing, and product
20    manufacturing, nanoparticle releases occurred most frequently when opening the CVD vessel and when
21    preparing the catalysts (Lee etal.. 2010). No data were found on potential releases from laser-ablation
22    synthesis of MWCNTs, but low levels of SWCNT clusters were released as aerosols during laser-ablation
23    synthesis of SWCNTs in laboratory and field conditions (Maynard et al.. 2004). This study intentionally
24    agitated the SWCNTs, however, which would not be a normal component of the synthesis stage.
25           Synthesis of MWCNTs might release even greater quantities of CNTs, by-products, and
26    feedstock materials than SWCNTs due to the larger quantities of precursor materials required for
27    synthesis (Tsai et al.. 2009). CNT powder generally comprises large bundles of CNTs and air release of
28    these bundles is likely to occur during operations involving agitation (e.g., scraping, shaking) (Fleury et
29    al.. 2011). but local exhaust ventilation substantially reduces airborne releases to the environment (Lee et
30    al.. 2010; Han etal.. 2008). Loose MWCNTs and the equipment used during synthesis are likely to be
31    cleaned up with a vacuum, which appears to be effective in reducing the airborne concentration of
32    nanoparticles (Lee etal.. 2010). Vacuuming, rinsing, and changing dust filters and other cleaning and

      March 2013                                    2-10                   DRAFT-DO NOT CITE OR QUOTE

-------
 1    maintenance activities also could result in subsequent release of MWCNTs to air or wastewater (Kohler et
 2    al.. 2008).
 3            MWCNTs released during synthesis can contain significant impurities (see Section 2.2.3.1). In
 4    addition, by-products from materials used in synthesis can be released at multiple points during the
 5    synthesis process (Plata et al., 2009). Some by-products, such as phenol, can be formed from general
 6    combustion processes used in CNT synthesis (Eckelman et al.. 2012). Air release of synthesis by-products
 7    including polycyclic aromatic hydrocarbons and volatile organic compounds has been observed during
 8    CVD synthesis of CNTs in the absence of engineering controls (Plata et al.. 2009): if employed, control
 9    technologies would be expected to limit these releases.
10            Accidental releases could also occur during MWCNT synthesis. These accidental scenarios
11    include fugitive equipment leaks, malfunctioning ventilation systems, and exposure to fire and high heat.
12    MWCNTs will not necessarily be destroyed at high temperatures (i.e., those possible in accidental fires)
13    (Kohler et al.. 2008). which could lead to airborne release or creation of ash containing elevated levels of
14    MWCNTs (Chaudhrv et al.. 2009). Additionally, damaged filters that collect MWCNTs could result in
15    airborne release during synthesis (Kohler et al.. 2008). Such accidental events, while unlikely, could
16    result in potentially large releases of MWCNTs to the environment.
17            Table 2-3 summarizes the anticipated potential release scenarios from the material synthesis stage
18    of MWCNTs. Information for decaBDE is provided for comparison, with more detailed information on
19    decaBDE available in Appendix H.
      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
Occurs in closed vessel, but fugitive air
emissions could occur; release depends
on synthesis method
Occurs in closed vessel, but fugitive air
emissions could occur
      Recovery of synthesized substance
Air release could occur during removal
from substrate and bagging
Air release could occur during 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
      March 2013
                 2-11
     DRAFT-DO NOT CITE OR QUOTE

-------
      2.2.3. Material Processing
                                      Priority Research Area: Product Life Cycle
               Material
              Processing
           gz
         o o
         E £
         * M
        m t
           o
           Q.
9/13
4/13
                                                                 Release Rate
                  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 Confidei
                       H - High:    Confident
                       0/13
       Nine out 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.
      2.2.3.1.  Life-Cycle Processes

 1            Material processing includes any modification of MWCNTs after synthesis and before
 2    incorporation into a flame-retardant formulation. These modifications can include purification,
 3    functionalization, and dispersal in solvents.
 4            MWCNTs undergo physical and chemical processing before they are incorporated into flame-
 5    retardant applications. Commercial MWCNTs that have not been purified can contain large amounts of
 6    impurities, such as amorphous carbon, graphite, and encapsulated metallic particles (Houetal.. 2008).
 7    After synthesis, MWCNTs are typically purified using physical (e.g., flocculation, microfiltration,
 8    centrifugation) or chemical (e.g., acid treatment) techniques (Hou et al., 2008). Appendix C presents
 9    various CNT purification methods and their efficacies in removing various classes of impurities.
10    By-products of purification techniques are expected to differ according to the technique used; one study
      March 2013
                                2-12
                  DRAFT-DO NOT CITE OR QUOTE

-------
 1    reported waste products of sodium hydroxide, ethanol, water, filtrate, and scrap membrane following
 2    general purification of SWCNTs (Healy et al. 2008).
 3           One analysis found that samples of MWCNTs purified by the manufacturer contained metal
 4    impurities ranging from 0.44 to 1.75 (wt%) (Ge et al.. 2011). After further purification with an acid
 5    treatment, significant quantities of catalyst residues (e.g., cobalt, chromium, iron, manganese,
 6    molybdenum, and nickel) remained (Ge et al.. 2011). Additionally, trace noncatalyst impurities (mostly
 7    transition metals such as cerium, gadolinium, holmium, etc.) were also found in purified CNTs, which the
 8    authors theorize can form during production/post-purification processes or might be introduced as catalyst
 9    synergists to improve the quality of synthesized CNTs (Ge et al.. 2011). Additionally, the purification
10    process itself also can damage the CNTs by introducing structural defects (Gustavsson et al.. 2011).
11           Before application, MWCNTs generally require surface functionalization (Saeed. 2010).
12    Functionalization—the modification of materials by covalently or noncovalently attaching new molecular
13    components—can alter the physicochemical properties of MWCNTs dramatically (Kohler et al.. 2008;
14    Ma et al.. 2008; Hirsch and Vostrowsky. 2005). Often, the goal of functionalization of MWCNTs is to
15    increase solubility, which facilities dispersion into solvents and polymers (Saeed. 2010). One example of
16    functionalization consists of the covalent grafting of MWCNTs to traditional flame retardants (Maet al.,
17    2008). Table 2-4 presents examples of MWCNT functionalization. After functionalization, MWCNTs are
18    often dispersed in water or organic solvents before they are incorporated into products (Saeed. 2010).
19    Most dispersion methods use dry mixing or liquid-phase sonication (agitation of particles with ultrasound
20    energy) to break up clumps and disperse MWCNTs in water or organic solvents.

      2.2.3.2.  Potential Releases from the Material Processing Stage
21           During the recovery, processing, handling, and packaging stages, CNTs are more likely to be
22    released as bundles from bulk powder than as individual CNTs (Kohler et al.. 2008) (see Table 2-3). In
23    general, releases resulting from liquid-phase processing of CNTs will be lower relative to those resulting
24    from dry handling, which can result in greater nanoparticle release (Kohler et al.. 2008). What proportion
25    of MWCNT processing occurs in the liquid versus the dry phase, however, is unclear. Handling of
26    MWCNTs can cause airborne release of particles (Methner et al.. 2010). Dispersal of MWCNTs in
27    suspensions can reduce the likelihood of aerosolization (Johnson et al.. 2010). but mixing and sonicating
28    (common processes used to disperse MWCNTs in solution) might also result in airborne release of raw
29    and functionalized MWCNTs (Johnson etal.. 2010: Lee etal.. 2010: Methner etal.. 2010). Environmental
30    control mechanisms likely would be in place to reduce environmental releases in facilities that process
31    MWCNTs (Fleury et al.. 2011: Methner et al.. 2010).  Cleaning of processing equipment and facilities can
      March 2013                                     2-13                   DRAFT-DO NOT CITE OR QUOTE

-------
1    lead to release of MWCNTs to air or wastewater (Fleury et al., 2011). MWCNTs released during this

2    stage might still contain some of the impurities listed in Section 2.2.3.1 and Appendix C.
     Table 2-4.     Examples of functionalization of MWCNTs.
Functionalization technique
Amidation - Formation of carbon
nanotube-acyl amides
Fluorination
Chlorination
Noncovalent exohedral
functionalization
Goal of functionalization
Creation of anchor groups for
further modification
Solubility in polar solvents
Solubility in polar solvents
Solubility in polar solvents
Additional reagents
thionyl chloride,
dicyclohexylcarbodiimide
elemental fluorine
chlorine gas
streptavidin
Reference
Hirsch and Vostrowsky
(2005)
Hirsch and Vostrowsky
(2005)
Hirsch and Vostrowsky
(2005)
Hirsch and Vostrowsky
(2005)
     Covalent grafting on intumescent  Better dispersion in matrix;
     flame retardant                solubility and stability in polar
                                  solvents; enhanced network
                                  structure at very low nanotube
                                  loading
                             poly(diaminodiphenyl methane
                             spirocyclic pentaerythritol
                             bisphosphonate)
                             Maetal. (2008)
     Atom transfer radical
     polymerization
Creation of anchor groups for
further modification
styrene and methyl methacrylate  Baskaran et al. (2004)
      In situ surface reversible
      addition-fragmentation chain
      transfer polymerization
Solubility in polar solvents
styrene and N-
isopropylacrylamide
Xuetal. (2007)
      Electrografting
Solubility in polar solvents
polyacrylonitrile
Petrov et al. (2004)
      Radiation polymerization
Solubility in polar solvents
ethanol, poly(acrylic acid), acrylic  Chen et al. (2006)
acid
     Liquid- and gas-phase
     oxidization with thermal
     treatment
Increased acidity or alkalinity;
improved flame-retardant
properties
nitric acid
Goncalvesetal. (2012)
3             Release of by-products from CNT processing also might occur. One modeling study notes the

4    potential for release of phenol from the production of nitric acid, a substance commonly used in CNT

5    purification (Eckelman et al.. 2012). The release of other substances used during purification is also

6    possible.

7             The accidental release scenarios for MWCNTs during processing are similar to those in the

8    material synthesis stage  (see Section 2.2.2.2). Additionally, spills of solutions containing dispersed
      March 2013
                      2-14
                  DRAFT-DO NOT CITE OR QUOTE

-------
1     MWCNTs might occur during the processing stage, which could result in the release of MWCNTs to

2     wastewater. Table 2-5 summarizes potential release scenarios from the material processing stage of

3     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
               Product
            Manufacturing
o  o
   EO c
   C <
_  (0
u  r
   o
   0.
                        12/13
                        1/13
                        0/13
                                                                        M
                                      E-RRF Confidence Rating  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
       Twelve out 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.
      March 2013
                                                  2-15
DRAFT-DO NOT CITE OR QUOTE

-------
                                     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.
 1             In this section, product manufacturing for MWCNTs is described. This life-cycle stage is

 2    considered to include the manufacture of flame-retardant formulations, the manufacture of textiles

 3    containing MWCNT-based flame retardants, and the manufacture of end-use products containing flame-

 4    retardant materials, such as furniture.



      2.2.4.1. Life-Cycle Processes

 5             CNTs can be dispersed in polymers by in situ polymerization or by using a twin-screw extruder (a

 6    specialized machine using two screws to mix, compound, and react polymers) [Laxminarayana and Jalili

 7    (2005) as cited in Kohler et al. (2008)]. One study described the manufacture of a "high heat"

 8    acrylonitrile-butadiene-styrene polymer matrix filled with CNTs (Tleury et al.. 2011). According to this

 9    study, master batch granules (1-4 mm long) consisting of thermoplastic resin and 15% CNTs by weight

10    are mixed with the pure polymer matrix in injection molding and extrusion processes (where

11    thermoplastics are fed into a heated vessel and forced into a mold cavity where they cool)  (Fleury et al..

12    2011). Little information is available regarding the ingredients and characteristics of MWCNT flame-

13    retardant formulations and possible by-products from their manufacture. Thermocyl®, an MWCNT flame
       March 2013                                         2-16                    DRAFT-DO NOT CITE OR QUOTE

-------
 1    retardant, includes silicone resins containing MWCNTs (Luizi. 2009). In another formulation, MWCNTs
 2    have been used in place of ammonium polyphosphate, a traditional phosphorus-based flame retardant, as
 3    a filler with polymethyl methacrylate (PMMA) and polyamide-6 (Motzkus et al.. 2012). General
 4    components of MWCNT flame-retardant formulations include epoxies, polyesters, and vinylesters
 5    (Alberding et al.. 2011).
 6           MWCNTs can be applied to textiles as a flame-retardant coating by (1) soaking the textile or
 7    (2) spray coating the surface of the textile (Luizi. 2009). The soaking application method for MWCNTs is
 8    similar to that used to apply decaBDE flame-retardant coatings (see Figure 1-4). In a recent study,
 9    researchers immersed cotton and polyester textiles in an MWCNT dispersion and maintained constant
10    motion to embed functionalized MWCNTs in the textiles (Goncalves et al.. 2012). The authors reported
11    acetic acid, sodium chloride, sodium carbonate, and sodium hydroxide as auxiliary reagents for this
12    embedding process (Goncalves et al.. 2012). An alternative immersion method involves soaking a textile
13    in a solution containing CNTs and then treating it with a crosslinking agent (similar to a binder), heating,
14    rinsing, and drying the textile (with or without heat) (Alimohammadi et al.. 2011). Uddin and Nyden
15    (2011 a) and Davis and Kim (2010) described a similar immersion method involving multiple treatments
16    with polymeric solutions to create a coating consisting of layers of polyacrylic acid/MWCNT-
17    polyethylenimine/ polyethylenimine. Flame retardant-treated materials also can be cured using UV
18    radiation (Lu  et al.. 201 la). Lee et al. (2010) mentioned the spray application of CNTs in solution to
19    thinly coat wafers, but they did not describe the process in  detail.  The few laboratory-scale studies
20    investigating MWCNT flame retardants in textiles have reported MWCNT loadings ranging from 0.5 to
21    4% by mass (Grzvbowski. 2009; Kashiwagi et al..  2005b; Kashiwagi et al.. 2005a; Kashiwagi et al..
22    2004), which  are about an order of magnitude lower than those for decaBDE. Thermocyl®, a commercial
23    MWCNT flame retardant, has been incorporated into polyethylene at a loading of 1% (Luizi. 2009). After
24    application, the MWCNT flame-retardant upholstery is cut, shaped, and glued or stapled to furniture.

      2.2.4.2. Potential Releases during Product Manufacture
25           Environmental  releases during the manufacture of MWCNT flame-retardant textiles can occur as
26    a result of the following activities: mixing, handling/packaging, application of the flame retardant to
27    textiles, textile processing/finishing, and accidents (Zhou and Gong. 2008). Few data are available that
28    describe releases from commercial-scale manufacture of MWCNT flame-retardant textiles. Therefore,
29    this section also relies on CNT release data from R&D facilities. Release of MWCNT bundles is possible
30    when nanotubes are blended with polymers to formulate the flame retardant;  however, releases are
31    anticipated to be  smaller for blending of master batches as  opposed to blending of pure CNT powders
32    (Fleury et  al.. 2011). The most critical phase for air and water releases during the formulation stage is the

      March 2013                                      2-17                  DRAFT-DO NOT CITE OR QUOTE

-------
 1    discharging and the cleaning of the mixing chamber (Fleury et al., 2011). Release from the formulated
 2    flame retardant is expected to be minimal, but packaging of the formulated flame retardant could result in
 3    releases. Airborne releases of CNTs (1.45 ug CNT/m3) were measured in a facility that packages
 4    Thermocyl®, an MWCNT flame-retardant coating (Luizi. 2009).
 5           Application of MWCNT flame retardants to textiles also could lead to air or water releases.
 6    Currently, how a particular production method used to incorporate MWCNT flame retardant in the textile
 7    matrix might influence release potential and subsequent exposure potential is unknown. Any spray
 8    application could lead to the potential airborne releases of matrix-bound MWCNTs if the application does
 9    not occur in a closed environment. One study observed airborne release of both nanoparticles and fine
10    particles when spraying  an MWCNT solution onto wafers as a coating (Lee et al.. 2010). The dominant
11    particle size released during this study ranged from 50 to 110 nm and subsequent heating of the treated
12    wafers also led to the release of particles smaller than 30 nm (Lee et al.. 2010). Other data suggest that
13    thermal processing (i.e., exposure to high temperatures and UV) might not lead to air release of
14    MWCNTs from polymer matrices (Nguyen et al.. 2011). In an analysis of an epoxy containing MWCNT,
15    exposure to UV radiation and high temperatures (50 °C) degraded the polymer matrix, but the MWCNTs
16    formed a dense network on the composite surface and no evidence of particle release was observed
17    (Nguyen et al.. 2011). If the MWCNT flame-retardant coating is applied by soaking the textile, water
18    release of matrix-bound MWCNTs could occur when the textile is rinsed. Additional cutting, sewing,
19    shaping, stapling, and other textile finishing processes could result in the airborne release office or
20    matrix-bound MWCNTs through abrasion (Kohler et al., 2008). One study found that grinding a
21    nanocomposite containing CNTs created a substantial amount of airborne particles made up of polymer
22    fragments containing CNTs (Fleury et al.. 2011). How applicable release from grinding of plastic
23    nanocomposites is to release from textile applications, however, is unclear. Airborne releases of CNTs
24    (1 ug CNT/m3) have been measured in a facility processing textiles treated with Thermocyl® (Luizi.
25    2009). Most manufacturing facilities,  however, would be expected to have controls in place to prevent or
26    minimize airborne releases to the environment. Equipment cleaning at any point in this life-cycle stage
27    could lead to release of MWCNTs to wastewater. Not enough is known about the other components of
28    MWCNT flame-retardant formulations to hypothesize about other substances that might be released as a
29    part of the flame-retardant product matrix, but the use of different substances in MWCNT flame-retardant
30    formulations might lead to different release characteristics.
      March 2013                                    2-18                  DRAFT-DO NOT CITE OR QUOTE

-------
 1             The accidental release scenarios for

 2     MWCNTs during product manufacture are

 3     similar to those in the material synthesis stage

 4     (see Section 2.2.2.2). In the product

 5     manufacturing stage, spills of MWCNT flame-

 6     retardant formulation might also occur and

 7     could result in the release of MWCNTs to

 8     wastewater. Table 2-6 outlines potential release

 9     scenarios from the product manufacturing stage

10     of MWCNT flame -retardant textile s along with

11     decaBDE for comparison.
       Table 2-6.    Potential release scenarios
                  during product manufacturing.
                         Additional Information Highlight Box 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 MWCNT 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.
                                         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
Preliminary evidence on air release due to
high heat is mixed
Air release possible
Rinsing/drying
Equipment cleaning
Water release possible
Air and water 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
       March 2013
                  2-19
    DRAFT-DO NOT CITE OR QUOTE

-------
2.3.  Storage  and  Distribution
                              Unprioritized Research Area: Product Life Cycle
      Product Storage,
         Transport
  uj
     o
     0.
     E
                  2/13
                  1/13
                      MIT"         H    M    L
             E-RRF Confidence Rating  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
                  10/13
 Two out 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
    I1
  Cu
     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:    Least Important
                       M-Medium: Possibly Important
                       H - High:    Important
                       Confidence
                       L - Low:    Not Confident
                       M — Medium: Somewhat Confident
                       H - High:    Confident
 Nine out 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, 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.
March 2013
                                  2-20
                    DRAFT-DO NOT CITE OR QUOTE

-------
      2.4.1.  Life-Cycle Processes
1             A wide variety of textiles contain flame-retardant coatings (see Section 1.2). Upholstery textiles

2     are expected to be used in public places where people of all ages will sit, lie, or walk on them. Some

3     unintended uses of upholstery textiles include outdoor use, repurposing for use in other products, burning

4     as kindling, or mouthing by children. Repurposing for use in other products and burning as kindling are

5     covered in Section 2.5 and Appendix H.2.5. In general, upholstery textiles are likely to have a lifespan of

6     at least 10 years (EU. 2002).
                                     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  El 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.
      March 2013                                           2-21                      DRAFT-DO NOT CITE OR QUOTE

-------
      2.4.2. Potential Releases during the Use Stage
 1           Environmental releases are expected from upholstery textiles coated with flame retardants due to
 2    (1) the potential use scenarios for the upholstery textiles and (2) the physicochemical properties of
 3    MWCNTs. The anticipated long lifespan of upholstery textiles (>10 years) suggests that releases in this
 4    stage could occur over several years (EU. 2002). Although no concentration data resulting from consumer
 5    use are available for MWCNTs, the following characteristics of flame-retardant upholstery textiles are
 6    expected to reduce releases MWCNTs (EU. 2002):
 7           •  Flame-retardant coatings must meet durability requirements to comply with regulations (see
 8              Section 1.2.1):
 9           •  Flame retardant is often applied to the back of the fabric, minimizing wear and tear; and
10           •  Upholstery textiles are unlikely to be washed frequently.
11           The integrity of the flame-retardant coating depends on the strength of the formulation that bonds
12    it to the textile surface (Sometal.. 2011: NRC. 2000). MWCNT flame-retardant textile coatings
13    considered in this case study are additive, suggesting that release from upholstery textiles could occur
14    during the use stage (see Section 1.2.2.2). While in general, CNTs are not likely to be released because
15    they are very stable and do not readily degrade, they could be released if the polymer matrix degrades
16    (Kohler et al.. 2008). or they could be released as a component of the polymer matrix. Factors that could
17    lead to MWCNT release from textiles include:
18           •  Regular use of upholstered furniture (e.g., sitting, walking, lying) could abrade the textile
19              surface and release small amounts of free or matrix-bound MWCNTs either into the air or
20              onto the skin of users. However, early unpublished evidence presented at a public meeting
21              indicates that very small amounts of MWCNTs could be released as aerosols after subjecting
22              an MWCNT-polymer nanocomposite to simulated wear and tear tests (Uddin and Nyden.
23              201 Ib).
24           •  Washing of textiles also could lead to water release of matrix-bound MWCNTs.
25           •  Even though CNTs might be embedded  in a matrix, depending on the production method, a
26              portion of a tube or a group of tubes could be left partially exposed in the final product.
      March 2013                                    2-22                   DRAFT-DO NOT CITE OR QUOTE

-------
 1    Importantly, although most releases initially will be to the indoor environment, they could spread
 2    outdoors through environmental transport mechanisms (see Chapter 3). In contrast, factors that suggest
 3    minimal MWCNT release from textiles include:
 4           •   Most flame-retardant upholstery textiles will be used indoors, minimizing exposure to UV
 5               light and weathering.
 6           •   Upholstery textiles that are back-coated (a common application method) with MWCNT flame
 7               retardant likely will not be subject to significant abrasion, washing, or UV light.
 8    Although these processes also could result in release of MWCNTs to the air or to wastewater, less
 9    degradation of upholstery textiles is expected for MWCNT-treated composites than for traditional
10    materials (see Section 1.3.3). As mentioned in Section 2.2.4.2. Nguyen et al. (2011) found that exposure
11    of an epoxy containing MWCNTS to conditions of high heat and UV radiation can cause MWCNTs to
12    form a dense network on the surface of composites, which might minimize environmental release.
13    The authors also found that the epoxy containing MWCNTs degraded more slowly than unfilled epoxy or
14    an epoxy containing another nanoscale material (Nguyen et al., 2011). A similar study found that when
15    PMMA is filled with silane-coated MWCNTs the amount of submicrometric airborne particles emitted
16    decreased when the polymer is exposed to fire compared to pristine PMMA (Motzkus et al., 2012).
17    The authors noted that the release of airborne particles depends on a variety of factors, including the type
18    of polymer matrix, the combustion process, and the type of surface treatment (Motzkus et al., 2012).
19    A lack of data precludes a determination of whether similar MWCNT releases could be expected to occur
20    in textiles that contain them.
21           Unintended uses also could lead to the release of MWCNTs from flame-retardant textiles. Use of
22    flame-retardant upholstery textiles outdoors could lead to weathering, which could degrade the polymer
23    matrix resulting in a release. Mouthing by small children, pets, or rodents on flame-retardant textiles
24    could lead to release directly into the mouths of children, pets, or rodents if the back-coating is exposed
25    and the integrity of the fabric is compromised. Preliminary unpublished evidence presented at a public
26    meeting, however, suggests that few MWCNTs are released from a flame-retardant nanocomposite when
27    subjected to simulated chewing tests (Uddin and Nyden. 201 Ib). Accidental contact of flame-retardant
28    textiles with fire and high heat also could occur and would lead to possible airborne releases (see Section
29    2.2.2.2 for more details). No data were found, however, that describe the likelihood of this release from
30    this application. Table 2-7 outlines potential release scenarios from the use stage of MWCNT flame-
31    retardant textiles. Similar data on decaBDE are provided in the table for comparison; more detailed
32    information on decaBDE can be found in Appendix H.
      March 2013                                     2-23                   DRAFT-DO NOT CITE OR QUOTE

-------
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             Air release possible due to
                     weathering/degradation of the polymer;  weathering/degradation of the polymer;
                     outdoor use could result in release to    outdoor use could result in release to
                     water or soil                          water or soil
Unintended use (mouthing)
                     Direct release to mouth likely if polymer  Direct release to mouth likely if polymer
                     matrix surface is accessible and        matrix surface is accessible and
                     degraded                            degraded
Accidental releases
                     Preliminary evidence suggests that air   Air release possible due to exposure to
                     release due to high heat is unlikely      high heat or fire
2.5.  Reuse,  Recycling, and End  of  Life
                                  Priority Research Area: Product Life Cycle
         Disposal/
         Recycling
                 [     Volume     ]       [  Release Rate   ]
  o a>
  E £
  £> i
  m r
     o
     a.
     E
                   10/13
                             Of
                             III
3/13
                   0/13
Importance Rating
L M H



.-.I
•
•
A
•
•
E-RRF
Importance Rating
L M H


•
• •
•
•

• •
•
^_1

                   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
Ten out 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.
March 2013
                                   2-24
                      DRAFT-DO NOT CITE OR QUOTE

-------
                                     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 etal., 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.

       Wastewater 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
       (Ciparis and 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 (Kohleretal., 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,  PBDDs,  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 partition in sewage
       sludge? See Appendix H for more information regarding the potential release of decaBDE during disposal or recycling.
1             The reuse, recycling, and end-of-life stage encompasses a variety of different transformation and

2     disposal processes for (1) MWCNTs, (2) MWCNT flame-retardant formulations, and (3) MWCNT flame-

3     retardant upholstery textiles. What the primary reuse, recycling, and end-of-life treatments are for

4     MWCNTs and MWCNT flame-retardant formulations are unclear. The reuse, recycling, and end-of-life

5     treatments for flame-retardant upholstery textiles containing MWCNTs, however, are expected to be

6     similar to those of traditional flame-retardant upholstery textiles (see Appendix H.2.5).




      March 2013                                           2-25                     DRAFT-DO NOT CITE OR QUOTE

-------
      2.5.1. Reuse and Recycling

      2.5.1.1. Life-Cycle Processes
 1            Reuse or recycling of MWCNTs or MWCNT flame-retardant formulations is unlikely. On the
 2    other hand, textile waste often is recovered and reused or recycled (Kohler et al.. 2008): upholstered
 3    furniture is sometimes reused, but is rarely recycled (CalRecycle. 2002). Upholstery could be donated to
 4    charitable organizations and resold for residential use. Additionally, upholstery textiles could be
 5    informally repurposed into clothing, blankets, and other textile products. Due to the difficulty of recycling
 6    furniture and flame-retardant materials, flame-retardant furniture is typically land-filled (CalRecycle.
 7    2002; Lassen et al.. 1999). Of the small portion of upholstered furniture that is recycled, about 60% of the
 8    material is recycled and 25-30% is composted (CalRecycle. 2002). No data were found that describe the
 9    proportion of other upholstery textiles (e.g., mattress ticking or curtains) that are typically recycled.
10            The main types of textile recycling processes are fiber-to-fiber recycling and polymer reduction
11    recycling. During the fiber-to-fiber process, textiles are shredded and blended with other fibers to create a
12    new mixture ready for spinning (Kohler et al., 2008). During the polymer reduction process, textiles are
13    cut and granulated to form pellets that are processed to break down the polymer to the molecular level to
14    be reused as raw material (Kohler et al.. 2008). No data were found that described the prevalence of each
15    recycling process.

      2.5.1.2. Potential  Releases during the Reuse/Recycling Stage
16            Release of MWCNTs beyond releases described in the use stage is  unlikely to occur during reuse
17    of flame-retardant upholstery textiles. Older textiles could release greater levels of MWCNTs, however,
18    due to increased degradation of the material. Informal repurposing of flame-retardant textiles likely would
19    require cutting and shredding, resulting in possible air release of MWCNTs. Airborne releases of
20    MWCNTs could occur during recycling of flame-retardant textiles. Recycling subjects textiles to a variety
21    of mechanical, thermal, and chemical treatments that could result in the airborne releases of additive
22    flame retardants from fibers (Kohler et al.. 2008). Recycling processes, such as shredding, milling, and
23    thermal processing, could lead to the airborne release of CNTs from upholstery textiles if carried out in
24    uncontrolled environments (Chaudhry et al.. 2009). Airborne releases during recycling of textiles likely
25    would be in the  form of CNTs in a polymer matrix (Chaudhry et al.. 2009). Downcycling, the conversion
26    of waste materials into new materials of lesser quality and reduced functionality, could lead to cross-
27    contamination of other materials with CNTs, for example, if MWCNT-treated textiles were shredded and
28    mixed with other textiles for use as insulation (Chaudhry et al.. 2009). Release of MWCNTs to water also
      March 2013                                     2-26                   DRAFT-DO NOT CITE OR QUOTE

-------
 1    could occur during chemical treatment and processing. Although release of MWCNTs is possible during
 2    recycling of flame-retardant textiles, no data were found that indicate the likelihood of release from
 3    recycling processes.
 4            Table 2-8 outlines potential release scenarios from the reuse/recycling stage of MWCNT flame-
 5    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 degrades    Air release possible if textile degrades
      Repurposing (product manipulated)         Air release possible due to cutting,       Air release possible due to cutting,
                                          shredding, and other abrasive processes  shredding,  and other abrasive
                                                                           processes
      Recycling (product broken down)           Air and water release possible due to    Air and water release possible due to
                                          mechanical, thermal, and chemical       mechanical, thermal, and chemical
                                          treatment                          treatment
      2.5.2.  Incineration

      2.5.2.1. Life-Cycle Processes
 6            The incineration of MWCNTs or MWCNT flame-retardant formulations is unlikely, but any
 7    incineration likely would occur in a hazardous waste incinerator. Upholstery textiles treated with
 8    MWCNT flame-retardant coatings might be sent to municipal incinerators for processing. Municipal
 9    incinerators generally provide a well-controlled environment with pollution control mechanisms and
10    sufficiently high temperatures (850 °C) to destroy most materials (Kohler et al.. 2008). Processing in
11    municipal facilities is likely to result in complete incineration of the upholstery textiles. Alternatively,
12    upholstery textiles also might be incinerated in less well-controlled facilities or burned in open fires as a
13    rudimentary form of waste management  or as kindling. These incineration methods are likely to result in
14    incomplete incineration of the upholstery textiles. No data were found that describe the prevalence of
15    incineration as a form of disposal for upholstery textiles or what proportion  of incinerated textiles is
16    processed at well-controlled incineration facilities.
      March 2013                                       2-27                   DRAFT-DO NOT CITE OR QUOTE

-------
      2.5.2.2. Potential Releases during the Incineration Stage
 1           Airborne releases of MWCNTs from well-controlled incineration are expected to be negligible,
 2    but incomplete incineration (e.g., open fires) could lead to some airborne release. MWCNTs are likely to
 3    be destroyed at the high temperatures used by municipal waste incinerators (Chaudhry et al., 2009; Sobek
 4    and Bucheli. 2009). Any CNTs remaining following municipal incineration could be expected to bind to
 5    other particles and be removed by the incinerator's filter (Kohler et al.. 2008); however,  incinerator
 6    removal efficiency for CNTs has not been studied (Som et al.. 2011). Alternatively, incomplete
 7    incineration of products containing CNTs could result in the airborne release of CNTs in a polymer
 8    matrix (Chaudhry et al.. 2009). Nguyen et al. (2011) found that exposure to moderately high temperature
 9    (50 °C) and UV radiation, however, caused MWCNTs to form a dense barrier on the surface of the
10    polymer that prevented the release of MWCNTs to the environment. No data were found that describe
11    potential by-products of incinerating upholstery textiles coated with MWCNT flame retardant in either
12    municipal incinerators or in incomplete incineration scenarios.
13           In sum, due to the high temperatures and pollution control mechanisms at municipal incinerators,
14    MWCNTs in flame-retardant textiles are expected to be destroyed during well-controlled incineration.
15    Preliminary evidence suggests that MWCNTs might not be released to the environment during
16    incomplete incineration. Table 2-9 outlines potential release scenarios from the incineration stage of
17    MWCNT flame-retardant textiles along with decaBDE for comparison.
      Table 2-9.    Potential release scenarios during incineration.
Processes included in incineration
life-cycle stage
Complete incineration, controlled
Incomplete incineration, uncontrolled
Information on release
MWCNTs
Release unlikely
Preliminary evidence suggests
that air release is unlikely

DecaBDE
Release unlikely
Air release of decaBDE and harmful
by-products likely, likely will reach environment
18
19
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
      March 2013
                                             2-28
DRAFT-DO NOT CITE OR QUOTE

-------
 1    disposed of in municipal landfills (Kohler et al., 2008). Remaining parts from recycled furniture, such as
 2    cover cloth materials, also are sent to the landfill (CalRecvcle. 2002). Additionally, some textiles might
 3    be disposed of in uncontrolled landfills or open dumping sites that have no pollution control mechanisms
 4    in place. No data were found that describe the proportion of upholstery textiles disposed of in landfills or
 5    any further processing that might occur at the landfill.

      2.5.3.2. Potential Releases during the Land-filling  Stage
 6           Land-filling of MWCNT flame-retardant textiles could lead to water and air releases. Mechanical
 7    land-filling processes (e.g., mixing and compacting) could lead to the airborne release of CNTs in a
 8    polymer matrix (Chaudhry et al.. 2009). Airborne release of CNTs after land-filling is complete, however,
 9    is not likely (Chaudhry et al.. 2009). Degradation of the polymer matrix material in textiles could lead to
10    release of CNTs into leachate/soil because CNTs are very stable and do not readily degrade (Kohler et al..
11    2008). No  data were found, however, that identify MWCNTs in land-fill leachate. Few data were
12    identified that measure releases of MWCNTs from land-filling flame-retardant textiles, but the
13    physicochemical characteristics of these materials suggest that such releases likely would be  small. Table
14    2-10 outlines potential release scenarios from the land-filling stage of MWCNT flame-retardant textiles
15    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 and   Air release possible due to mixing and
                                  compacting                       compacting
      Degradation
                                  No data exist; air and water release     Air and water release unlikely
                                  possible but unlikely
16
17
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
      March 2013
                                              2-29
  DRAFT-DO NOT CITE OR QUOTE

-------
 1    divert their wastewater to an on-site wastewater treatment plant. Alternatively, some wastewater from
 2    these facilities might be directly processed by municipal wastewater treatment plants. Water releases of
 3    MWCNTs that occur during the storage and distribution, use, and reuse/recycling/end-of-life stages also
 4    would be treated in municipal wastewater treatment plants.
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
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. 2002). 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 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 MWCNTs and  the
spread of sewage sludge onto agricultural soil
    Additional Information Highlight Box 5:
          Impact of MWCNT release into
           wastewater treatment plants
Petersen et al. (2011b) noted that most of the CNTs that are
released into wastewater come from the tailoring, finishing,
use, and degradation of textiles containing CNTs, or from
research and development facilities. 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. (2011b)
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.
could represent a significant source of MWCNTs to soil. 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-18). See Section 3.3.3 for information regarding MWCNT removal
efficiency of these wastewater treatment plants.
      March 2013
                                               2-30
                   DRAFT-DO NOT CITE OR QUOTE

-------
1            Table 2-11 outlines potential release scenarios from the wastewater treatment stage of MWCNT

2    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        Release unlikely due to sorption
                                          behavior, but filter backwash could lead  behavior, but filter backwash could lead
                                          to release                          to release

     Removal of sludge                      No data exist, but release to soil        Release to soil likely if sludge spread
                                          possible if sludge spread on agricultural  on agricultural fields
                                          fields
     March 2013                                        2-31                    DRAFT-DO NOT CITE OR QUOTE

-------
                               This page intentionally left blank.
March 2013                                                         DRAFT-DO NOT CITE OR QUOTE

-------
           Chapter 3.  Transport, Transformation,
                                         and Fate
 1          Releases throughout the product life cycles of upholstery textile coatings containing multiwalled
 2    carbon nanotube (MWCNT) flame retardant will, to some extent, lead to occurrence of primary and
 3    secondary contaminants in air, soil, and aquatic media. Chapter 3 examines what might happen to these
 4    substances after their release to the environment, including transport or transformation through chemical,
 5    physical, and biological processes. Studies investigating the transport, transformation, and fate of
 6    MWCNTs in the environment are summarized in Appendix D. In general, information on environmental
 7    concentrations of MWCNTs was not found.
 8          MWCNTs can be released into the environment during the manufacturing, storage, distribution,
 9    use, disposal, reuse, and recycling of upholstery textiles treated with flame retardants (see Chapter 2).
10    MWCNT flame-retardant formulations are used primarily as additives that are mixed with, not chemically
11    bound to, polymers in textile products (see Section 1.3). Because they are not chemically bound, these
12    substances can escape from the material and become a source of contamination to surrounding
13    environmental media (Moniruzzaman and Winey. 2006). Although some, if not most, releases after the
14    production stage are likely to be in the matrix-bound form, little information exists that describes the
15    environmental behavior of MWCNT-polymer complexes. As a result, this chapter focuses on the
16    transport, transformation, and fate of MWCNTs not embedded in a polymer matrix.
17          Section 3.1 provides a brief discussion of the chemical and physical characteristics and the
18    processes that influence behavior (e.g., mobility, persistence, and bioavailability) of MWCNTs in
19    environmental media. The sections that follow summarize the available information regarding their
20    behavior in indoor and outdoor air (Section 3.2). aquatic systems (Section 3.3). and terrestrial systems
21    (Section 3.4). A brief discussion of models that might be used for evaluating their fate and transport in
22    environmental media is provided in Section 3.5.
      March 2013                                  3-1                  DRAFT-DO NOT CITE OR QUOTE

-------
                                    Additional Information Highlight Box 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 (Nowack et al., 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

1            The environmental fate of MWCNTs will be dictated by their physical and chemical properties
2    (see Text Box 1-1 and Figure 3-1). These properties influence behavior, including mobility, persistence,
3    bioavailability, and likelihood for transformation in environmental media. A summary of key
4    physicochemical factors that might affect partitioning of MWCNTs and their fate in the environment is
5    provided in Table 3-1. Values for or descriptions of key physicochemical properties of MWCNTs (e.g.,
6    surface area, morphology, solubility) are provided in Table 1-9.
     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
     cluster
                              • Single particles versus bundles or clusters will differ in their mobility (and ultimate fate) in
                                environmental media; generally, nanoparticle clusters are less mobile in the environment
                                than individual nanoparticles
                              Sources: Ma-Hock etal. (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
                              Source: Kohleretal. (2008): O'Driscoll etal. (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)
     March 2013
                                                  3-2
DRAFT-DO NOT CITE OR QUOTE

-------
     Table 3-1, cont.: 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?
     Low water solubility; hydrophobic
     (potentially lipophilic)
•   Will result in poor dispersion
•   Prone to bundling in the water column 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. (2011): Oberdorster et al. (2006): Wu et al.
(2006); Kennedy etal. (2008)
1            Carbon nanotubes (CNTs) are not dispersed by simple mixing because they tend to form bundles

2    through a van der Waals attraction among tubes. As shown in Table 2-4. MWCNTs, including those in

3    commercial products such as textiles, can be engineered to include charged functional groups to improve

4    their dispersion or to increase their solubility in aqueous media; the treated nanoscale materials

5    (nanomaterials) that remain dispersed tend to exhibit greater persistence in the environment (Klaper et al..

6    2010; Saeed. 2010; Kohler etal.. 2008: Luoma. 2008). CNTs in textiles might also be coated with a

7    surface coating, such as a polymer (Kohler et al.. 2008). These surface coatings could be degraded by

8    chemical or biological reactions, affecting persistence of the MWCNTs  over time in ways that depend on

9    both the presence of a coating and the type of coating used.
                                     Additional Information Highlight Box 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 high heat and UV radiation can cause MWCNTs to form a dense network on the composite surface, which might
      minimize environmental release (see Section 2.2.4.2). 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.
     March 2013
                    3-3
DRAFT-DO NOT CITE OR QUOTE

-------
        Properties of the
          Environment
                                          Properties of
                                       MWCNT Material
                                   Properties of MWCNT
                                       Agglomerates
          Hydrosphere3
         PH
         Salts
         Natural Organic
         Material
                     Interaction with
                    functional groups;
                       Adsorption
                            Phototransformation/
                             Photodegradation
 Atmosphere
UVA/B Radiation
           Lithosphere
         Weathering
         Erosion
Surface chemistry
 functional groups
 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.b

     Hydrosphere describes environmental media characterized by water (i.e., groundwater, surface water). Lithosphere describes media
     characterized by rock composition (i.e., soil, sediment).
     bEnvironmentol 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.
1            Environmental conditions (e.g., redox potential, pH, temperature, UV light, ionic strength, and

2    characteristics of other contaminants present) are also likely to affect the behavior and environmental fate

3    (e.g., mobility, persistence, bioavailability) of MWCNTs (Toth et al.. 2011; Zhang etal.. 2011; Helland et

4    al., 2007) (see Text Box 1-1 and Figure 3-1). The same is true for environmental processes such as

5    interactions with natural organic matter (NOM), which will alter the surface chemistry of the MWCNTs

6    (Petersen et al.. 201 la). Metals, such as lead, cadmium, and copper; hydrophobic organic chemicals and

7    other toxic organics (e.g., polycyclic aromatic hydrocarbons); and other pollutants (e.g., phenol,
     March 2013
                                              3-4
                              DRAFT-DO NOT CITE OR QUOTE

-------
1    dopamine) can sorb strongly to CNTs (Li et al.. 2011; Tothet al., 2011; Cho et al., 2008; Petersen et al..

2    2008; Chen et al.. 2007; Helland et al.. 2007). These associations might dictate CNT mobility and

3    bioavailability. Factors that have been shown to influence adsorption of organic contaminants to

4    MWCNTs include surface oxidation (adsorption capacity decreases with increasing oxygen content) and

5    pH (effects differ based on the contaminant) (Lietal.. 2011; Tothet al.. 2011; Cho et al.. 2008).
      3.2.  Transport, Transformation, and  Fate in Air
                   Priority Research Area: Environmental Transport, Transformation, and Fate
         I              I

        o
        0.
                                     01
                                     E -r
                                    13
                      11/13
                                  w
                      2/13
                      0/13
                         o>
                         E T
                        13
                                                          LLJ
                                                             o
                                                             I-1
                                                                                  LLI  CD
                                                                                     o
Bioavailability ]
•



•
•
* *

'
      • - 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"               ML                 ML
  E-RRF Confidence Rating  E-RRF Confidence Rating  E-RRF Confidence  Rating

Eleven out 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 PBDEs to remote ecosystems, including the Arctic (de Wit
      etal., 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.
      March 2013
                     3-5
DRAFT-DO NOT CITE OR QUOTE

-------
 1           MWCNTs released from flame-retardant upholstery textile coatings could reach indoor and
 2    outdoor air in several ways. For example:
 3           •   They can be released directly into ambient air during all stages of the product life cycle, as
 4               previously described in Chapter 2 and can disperse through air away from the source of
 5               release.
 6           •   They can become suspended in the surrounding indoor or outdoor air during multiple stages
 7               of the product life cycle.
 8           •   They might remain suspended and be transported through the atmosphere or be deposited
 9               onto surfaces. Particles that have been deposited on surfaces could become resuspended in
10               the air and redeposited elsewhere.
11           MWCNTs might distribute to indoor air and dust, and these sources could be a major contributor
12    to outdoor air concentrations. Several processes and factors could influence the behavior (e.g., mobility,
13    persistence, bioavailability) and ultimately the fate of airborne MWCNTs in indoor and outdoor
14    environments, including: (1) size—whether they are traveling as individual particles of varying sizes or as
15    larger bundles, (2) surface chemistry, (3) interactions with other airborne particles and chemical
16    compounds, (4) residence time in the air, and (5) distance traveled prior to deposition (Kohler et al., 2008;
17    U.S. EPA. 2007). The fate of airborne nanomaterials outdoors could be influenced by meteorological
18    factors, including wind, temperature, and precipitation (Navarro et al., 2008).
19           Information in the recent literature regarding the behavior of airborne MWCNTs is limited.
20    No studies have examined transport mechanisms for MWCNTs in air. Yang et al. (2009) examined the
21    atmospheric aging of CNTs under normal ambient conditions (20 ± 0.5°C, relative humidity = 50 ± 1%)
22    and found that CNT surface area and pore volume (volume of space [holes] per gram nanotube; provided
23    in cmVgram) decreased over time  (up to 7-15 months) and coincided with decreases in surface oxygen of
24    the CNT as it aged. For MWCNTs, these conditions stabilized within 15-18 months. The total structural-
25    defect concentration also appeared to be lowered as the CNTs aged. The authors theorized that during
26    CNT aging under ambient conditions, oxygen leaves the surface of the CNTs. The structure then repairs
27    itself and becomes more thermodynamically stable with fixed values of surface area, pore volume, and
28    structural defects. Based on these results, the authors stated that CNT "physicochemical properties can be
29    characterized with reliability only after samples have sufficiently aged" (Yang et al.. 2009).
30           In the laboratory, Zhu et al. (2011) observed that MWCNTs exposed to air under ambient
31    temperature were slowly oxidized and shortened. The degree of oxidation increased over time, and
32    MWCNTs were almost totally transformed into amorphous carbon after 15 days.
33           Complete transformation of pure, airborne CNTs to carbon dioxide could occur during
34    incineration at temperatures greater than 850°C and oxygen levels exceeding 21%. Under oxidative
      March 2013                                     3-6                   DRAFT-DO NOT CITE OR QUOTE

-------
 1    conditions, MWCNTs have been shown to burn off completely at 740°C (Som et al., 2011; Kohler et al.,
 2    2008).
 3          Section 4.1.2.4 describes MWCNT air concentration data collected in occupational settings.
 4    No data were found on residential or consumer exposures to MWCNTs in air. Literature containing
 5    concentrations of MWCNTs in outdoor air also was not found.
      3.3. Transport, Transformation, and Fate in Water and
            Sediment
 6          MWCNTs released from the flame-retardant upholstery textile coating life cycle could enter
 7    aquatic systems in several ways. For example:
 8          •   MWCNTs in ambient air subsequently could be deposited or washed out to aquatic systems.
 9          •   Erosion of contaminated soil could release MWCNTs to surface waters.
10          •   Runoff flowing along the ground surface could transfer MWCNTs in contaminated soil to
11              nearby waterways.
12          •   Wastewater effluents containing MWCNTs could be a source of contamination to receiving
13              water bodies near the discharge location.
14          •   MWCNTs could leach from land-filled sewage sludge into subsoil and ground water and
15              migrate to surface water or sediment.

      3.3.1. Surface Water and Sediment (Inland and Coastal)
16          The transport, transformation, and fate of MWCNTs in surface water (specifically, mobility,
17    persistence, and bioavailability) was not identified as a priority area by workshop participants during the
18    collective judgment step of the CEA process. However, mobility, persistence, and bioavailability in
19    sediment was determined to be a priority area. Due to the limited available data, which overlaps between
20    surface water and sediment, these topics are discussed together.
      March 2013                                   3-7                 DRAFT-DO NOT CITE OR QUOTE

-------
               Priority Research Area: Environmental Transport, Transformation, and Fate
        Sediment
 « B
 IU t
    o
    Q.
                  7/13
                 4/13
                  2/13
        H     M     L                  ML
    E-RRF Confidence Rating  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
 Seven out 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
        Surface
        Water
          Mobility
                                                                 Persistence
          Bioavailability
» B
u t
   o
   a
                 2/13
                                *
                              m
                8/13
                                                      LU
                          8 =
                          
-------
  1
  2
  3
  4
  5
  6
  7
  8
  9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
        The hydrophobicity and van
der Waals interactions of MWCNTs
imply they will partition to the
particulate 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 4J).
MWCNTs suspended in NOM
solutions  have  greater potential for
dispersion in natural waters
(ODriscolletaL 2010). Results of
the O'Driscoll  et al. (2010) study
suggested that  smaller diameter
MWCNTs stay suspended in NOM
         DecaBDE Can Inform MWCNT Assessment
Water solubility and K0w 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.
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 (Hvung and Kim. 2008; Wang et al., 2008;
Hyungetal.. 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 et al.. 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.
March 2013
          3-9
                                                                               DRAFT-DO NOT CITE OR QUOTE

-------
 1            Functionalization of MWCNTs can improve their dispersion or increase their solubility in
 2    aqueous media, thereby increasing their mobility. As mentioned previously, engineered surface
 3    modifications (e.g., functional groups and coatings) are used to improve CNT dispersion in aqueous
 4    suspension. Column stability and settling experiments have shown pure MWCNTs settle and sink to
 5    sediment as described previously. The presence of functional groups slows this settling, especially in
 6    combination with NOM (Kennedy et al.. 2008; Hyung et al.. 2007).
 7            As discussed in Section 3.1. MWCNTs can act as environmental adsorbates of metals,
 8    hydrophobic organic compounds, and other toxic organics (Li et al.. 2011; Cho et al.. 2008; Petersen et
 9    al.. 2008; Chen et al.. 2007; Helland et al.. 2007). and the properties of the adsorbants might dictate
10    MWCNT mobility and dispersion in surface water.
11            Studies relevant to MWCNT fate and transport in aqueous media are summarized in Appendix D.
12    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
             Ground
              Water
      si
         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
                                                                Persistence
                                                    Bioavailability
                                                                                O)
                                                                                = I
                                                                                (0
                                                                                o
       H    M     L          H    M    L         ^    M    IT
   E-RRF Confidence Rating  E-RRF Confidence Rating   E-RRF Confidence Rating
Two out 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.
      March 2013
                    3-10
DRAFT-DO NOT CITE OR QUOTE

-------
       3.3.3. Wastewater
                     Priority Research Area: Environmental Transport, Transformation, and Fate
        O 111
        E £
        * «
          o
          Q.
                       11/13
2/13
                       0/13
                                                                                          (  Bioavailability  ]
                                      D9
                                      1= -r
                                      18
                                    LLJ (0
                                      t
                                      o
                                                              {£.
                                       H
                                       o
                                                                                      D)
                                                                                    w
    H    M     L
E-RRF Confidence Rating
    H    M     L
E-RRF Confidence Rating
                                                                     H     M    IT
                                                                 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
               Eleven out 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.
 1             Information in the recent literature

 2    regarding the behavior of MWCNTs in

 3    wastewater is limited. Because of their

 4    hydrophobicity and tendency to form bundles,

 5    however, pure MWCNTs likely would be

 6    removed from the effluent by settling during the

 7    sewage treatment process. Functionalized

 8    MWCNTs could have improved dispersion and

 9    increased solubility in wastewater, thereby

10    increasing their mobility and persistence in

11    wastewater effluents.

12             As discussed in Section 3.1. MWCNTs

13    can act as environmental adsorbates of metals,

14    hydrophobic organic compounds, and other

15    toxic organics (Li etal.. 2011; Cho et al.. 2008;

16    Petersen et al.. 2008; Chen et al.. 2007; Helland

17    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 (Ricklund etal.,  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.
       March 2013
                                  3-11
                  DRAFT-DO NOT CITE OR QUOTE

-------
1    CNT mobility and dispersion in wastewater. He et al. (2012) studied the behavior of MWCNTs stabilized

2    by humic acid during the coagulation-flocculation-sedimentation process of drinking water treatment and

3    found that humic acid-stabilized MWCNTs were effectively sequestered by this process.

4           Literature containing concentrations of MWCNTs in wastewater effluent or sludge was not

5    found.
     3.4.  Transport, Transformation, and Fate in  Soil
               Unprioritized Research Area: Environmental Transport, Transformation, and Fate
        gl
        o
        a.
              3/13
                    5/13
                    5/13
                                                              Persistence
                                                                              Bioavailability
                                                                              o
                                  H    M    L
                              E-RRF Confidence Rating
    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
                                 Three out 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.
     March 2013
                                              3-12
                 DRAFT-DO NOT CITE OR QUOTE

-------
                Unprioritized Research Area: Environmental Transport, Transformation, and Fate
                                       Bioaccumulation
f1
      « ro

                    2/13
                    8/13
01
o
                                  ,

                                  I-
 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 out 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.
1           Although empirical data on MWCNT concentrations in the environment are lacking, some

2    researchers have used modeling to simulate movement of CNTs through environmental compartments

3    and to derive predicted environmental concentrations (PECs). Mueller and Nowack (2008) used substance

4    flow analysis to model CNTs in air, soil, and water (not sediment) in Switzerland based on simplifying

5    assumptions. PECs were calculated for "realistic" and "high exposure" scenarios and are provided in

6    Table 3-2.

7           Gottschalk et al. (2009) described a probabilistic material flow analysis framework to derive

8    probability distributions of PECs for engineered CNTs in soil, sludge-treated soil, air, surface water,
     March 2013
                  3-13
                DRAFT-DO NOT CITE OR QUOTE

-------
 1    sediment, and sewage treatment plant effluent and sludge for the United States, Europe, and Switzerland
 2    (see Table 3-2). As noted in Gottschalk et al. (2010). although both studies were designed to estimate
 3    PECs in environmental media, the two study designs were quite different and the methodologies used
 4    varied considerably, making a direct comparison of PECs difficult. The differences included the model
 5    type (deterministic versus probabilistic); model scale [Gottschalk et al. (2009) considered additional
 6    environmental compartments  (sediment and ground water), more flows associated with these additional
 7    compartments, as well as production, manufacturing, and recycling processes]; model input data [newly
 8    available model input data were used in Gottschalk et al. (2009)]: and amended categorization of the
 9    products and allocation of the CNT mass to the product categories.
10           Another study employed the USEtox model to carry out a life cycle-based analysis of the aquatic
11    toxicity impacts associated with CNT synthesis (Eckelman et al.. 2012). This study used information on
12    the physicochemical properties of CNTs to model fate and transport of CNTs in freshwater systems under
13    "realistic" and "worst-case" or "conservative" scenarios, as described by the authors. Based on
14    information from Gottschalk et al. (2009). the realistic scenario assumed the fraction of CNTs removed
15    from the water column due to clustering and settling to be 90%. For the 10% remaining in the water
16    column, this scenario assumed a CNT exposure factor (defined by USEtox as the dissolved fraction of
17    CNTs in the water column) of 98%, with most of the remaining 2% partitioning to suspended solids. This
18    realistic scenario, which utilized Monte Carlo analysis, calculated a mean residence time in fresh water to
19    be on the order of days. The conservative scenario assumed an exposure factor in the water column of
20    100% and estimated the freshwater residence time for CNTs as  143 days. This study did not distinguish
21    between single-walled and multiwalled CNTs.
      March 2013                                    3-14                   DRAFT-DO NOT CITE OR QUOTE

-------
     Table 3-2.    Predicted environmental concentrations of CNTs using fate and transport modeling.
Medium
Air (|jg/m3)
Soil
(|jg/kgorA|jg/kg-yeaij
Sludge- treated soil
(Aug/kg-year)
Surface water
(ug/L)
Sediment
(Aug/kg-year)
Sewage treatment plant (STP) effluent
(ug/L)
STP sludge
(mg/kg)
Mueller and Nowack (2008)
1.5x1 0-3 (realistic);
2.3* 1C-3 (high exposure)
1 x 10-2 (realistic);
2x 1C-2 (high exposure)
ND
5 xlQ-4 (realistic);
8* 10-4 (high exposure)
ND
ND
ND
Gottschalk et al. (2009)1
1 x 10-6 (United States)
3x10-6 (Europe)
8x10-6 (Switzerland)
5.6xlO-4 (United States)
1.5x1 0-3 (Europe)
1.9x1 0-3 (Switzerland)
3.1 x 10-2 (United States)
7.4x1 0-2 (Europe)
ND (Switzerland)2
1 x 10-6 (United States)
4x10-6 (Europe)
3x 1C-6 (Switzerland)
4.6x1 0-2 (United States)
2.4x 10-1 (Europe)
2.3x1 0-1 (Switzerland)
8.6x10-3 (United States)
1.5x10-2 (Europe)
1.2x1 0-2 (Switzerland)
6.8x1 0-2 (United States)
6.2x1 0-2 (Europe)
6.9x1 0-2 (Switzerland)
      1For 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).
      2ln Switzerland, sewage sludge is not applied to soil.
      ND = no data
1            Cullen et al. (2010) simulated subsurface mobility of MWCNTs compared with nanofullerenes
2    (nC6o) under a range of hydrologic and geological conditions  (homogeneous and heterogeneous) using a
3    two-dimensional finite element model. In general, nanoparticles in systems with the  same average
4    hydraulic properties were predicted to be less mobile if the systems were heterogeneous as opposed to
5    homogeneous. For the conditions evaluated, MWCNTs were predicted to be much more mobile compared
6    with nC6o because of two factors—nanoparticle shape and size. The smaller, spherical nC6o were more
7    efficiently collected on soil surfaces compared with the larger, cylindrical MWCNTs.
     March 2013
3-15
DRAFT-DO NOT CITE OR QUOTE

-------
                                  Additional Information Highlight Box 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 et al., 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 et al., 2011) 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 natural organic matter) in
 peat soil bind to MWCNTs (Tianetal., 2012: Wangetal., 2011), which alters the surface functionalization and electronegative
 charge and increases sorption of heavy metal ions (Tianetal., 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. 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.
March 2013                                            3-16                      DRAFT-DO NOT CITE OR QUOTE

-------
                        Chapter  4.  Exposure-Dose
 1           Releases of multiwalled carbon nanotubes (MWCNTs) to the indoor and outdoor environments
 2    can occur at multiple stages of the product life cycle for flame-retardant upholstery textile coating
 3    (Chapter 2). Subsequent transport, transformation, and fate processes dictate how MWCNTs distribute
 4    through various environmental media once released (Chapter 3). Exposure describes the pathways
 5    through which contact occurs between contaminants in the environment and living organisms and abiotic
 6    receptors. Toxicokinetics (i.e., absorption, distribution, metabolism, excretion [ADME]) describes the
 7    processes that relate exposure (or dosage) to the internal dose, which refers to the quantity of a chemical
 8    or material that is taken up and absorbed by living organisms (U.S. EPA. 2010c).13
 9           Section 4.1 introduces analytical techniques for identifying, characterizing, and measuring
10    MWCNTs in various matrices. The various metrics recommended for characterizing exposure and dose of
11    MWCNTs are also discussed, and available concentration data in various indoor and outdoor media are
12    presented. In the absence of data quantifying MWCNT exposures at the point of contact, measured
13    concentrations of MWCNTs in surrounding media can be used to estimate exposures using a scenario
14    evaluation approach. Section 4.2.1. Appendix G.4.1. and G.4.2 expand on the release scenarios presented
15    in Chapter 2 to discuss the potential human and ecological exposure pathways that link those releases to
16    receptors. No data were identified regarding  relevant exposure pathways leading to impacts on abiotic
17    resources (e.g., the manmade environment);  as a result, this comprehensive environmental assessment
18    case study does not include a discussion of exposure scenarios that would influence abiotic receptors.
19    Although broad potential impacts on society and the global environment are discussed in Section 5.3.
20    exposure is either not considered germane to the discussion of the impact (such as for economic impacts
21    of manufacturing MWCNTs) or the exposure characteristics related to the impact are already included in
      13The term "dose" is described generally by the U.S. 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."
      March 2013                                     4-1                   DRAFT-DO NOT CITE OR QUOTE

-------
 1    the general discussion that follows (such as for higher potential exposure levels in certain populations
 2    related to socioeconomic impacts).
 3            The scenarios described in Sections 4.2.1 and Appendix G.4 describe the conditions under which
 4    exposures might occur; this information can be used in combination with measured or modeled
 5    concentrations in environmental media from Section 4.1 and exposure factors to estimate exposures.
 6    Kinetic information then can be used to determine or estimate the internal dose that results from external
 7    exposures. When available, point-of-contact measurements, administered dosages, tissue or body burdens,
 8    scenario-specific exposure guidelines and recommendations are provided, and the toxicokinetics of
 9    MWCNTs are described in Section 4.2.2 and Appendix G.4.2. Studies describing toxicokinetics of carbon
10    nanotubes (CNTs) in mammals are summarized in Appendix F. Finally, Section 4.4 discusses aggregate
11    exposures to MWCNTs from multiple sources and Section 4.5 discusses cumulative exposures to multiple
12    related stressors.
13            As described in Section 2.2.4. MWCNTs likely would be incorporated into a polymer or other
14    type of matrix in the flame-retardant formulation applied to upholstery textiles, and both the free and
15    matrix-bound forms might be released during the product life cycle. Very little data relevant to MWCNT
16    exposures, however, have been generated for the matrix-bound form of MWCNTs. This lack of data
17    necessitates a reliance on the existing data for free MWCNTs in the discussion throughout this chapter.
18    The extent to which exposure characteristics and dose implications differ between the free and matrix-
19    bound form of MWCNTs, however, is unknown at this time.
         Unprioritized Research Area: Exposure Route
                        Unprioritized Research Area: Dose (Kinetics)
          OX
       uj tr
         o
         D.

•
• •
* *
• * *
• * *
* *

0/13 [ Direct Contact ]
3/13
10/13
                         01
                      uj tr
                         o
                         Q.
                         E
0/13
                                                                       2/13
                                                                                  Absorption
                                                                       11/13
      • - 1 participant's vote
      Importance
      L -Low:    Least Important
      M - Medium: Possibly Important
      H - High:    Important
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.
      March 2013
                    4-2
     DRAFT-DO NOT CITE OR QUOTE

-------
      4.1.  Detection,  Measurement, and  Characterization
                    Neutral Research Area: Environmental Transport, Transformation, and Fate
       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.
 1            Exposure scenario evaluation requires information on measured, modeled, or reasonably

 2    estimated concentrations of a stressor in exposure media. As introduced in Chapter 1. MWCNTs represent

 3    a group of compounds, encompassing substances that span a range of physicochemical characteristics and

 4    properties. As a result, developing reliable analytical techniques for detecting, measuring, and

 5    characterizing the full range  and makeup of MWCNTs in environmental media can present challenges.

 6    Text Box 4-1 provides a brief discussion of a few common analytical techniques and the general

 7    challenges associated with them. Appendix B summarizes common analytical techniques and presents the

 8    strengths and limitations of each technique.

 9            Configurations of MWCNTs vary depending on the type of material or substrate used in their

10    manufacture (see Section 2.2). Text Box 4-2 provides examples of the specific physicochemical

11    properties of MWCNTs that influence exposure, uptake, and dose. Single analytical techniques used alone

12    are generally not sufficient for characterizing all of the properties of MWCNTs that can influence

13    exposure; to characterize the presence and form of MWCNTs  in media adequately, multiple analytical

14    methods must be used in tandem (see  Text Box 4-1 and Appendix B).


                       Text Box 4-1. Detecting,  Measuring,  and Characterizing MWCNTs

       Because multiwalled carbon nanotubes (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 labeling is a precise  quantification method that works in any medium, but CNTs must be radioactively labeled
       prior to dispersal in environmental  media for this method to work. 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.
      March 2013                                         4-3                    DRAFT-DO NOT CITE OR QUOTE

-------
 1           The potential for human exposure from upholstery textiles coated with flame retardants is
 2    currently difficult to assess because data are not yet available on production of MWCNTs for flame-
 3    retardant upholstery textiles (see Section 2.2.4). Furthermore, very little information is available on
 4    whether or how MWCNTs might migrate out of a product matrix; or on what methods could be used to
 5    quantify exposure concentrations of matrix-bound MWCNTs and partially exposed MWCNTs (i.e., ends
 6    of MWCNTs "sticking out" from the matrix). Understanding the behavior of the material requires
 7    comparing potentially similar applications. For example, machining of CNT composites and the resultant
 8    exposure could be compared to migration of MWCNTs out of textiles and the resultant exposure.
 9    The reason for this is that machining of CNT composites could generate particles or fibers similar to those
10    generated from the wear and breakdown of MWCNT textiles, so these studies are useful from an
11    exposure assessment perspective. Estimating exposures during other product life-cycle stages also might
12    be possible by evaluating similar applications or alterations of materials containing MWCNTs. Such
13    assumptions, however, could yield conclusions that are very different from the actual behavior and
14    exposure potential of the MWCNTs in flame-retardant upholstery textile applications.
      March 2013                                     4-4                    DRAFT-DO NOT CITE OR QUOTE

-------
          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 multiwalled carbon nanotubes (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 (Shietal., 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.
March 2013
4-5
DRAFT-DO NOT CITE OR QUOTE

-------
4.1.1.  Dose and Exposure Metrics
  1
  2
  3
  4
  5
  6
  7
  8
  9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
        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. 201 Ob: Mavnard and Aitken. 2007).
      Additional Information Highlight Box 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).
March 2013
4-6
                                                                               DRAFT-DO NOT CITE OR QUOTE

-------
 1           Some research in animal models has shown that toxic effects of some MWCNTs do follow a
 2    more traditional dose-related curve based on the administered mass concentration of MWCNTs [for
 3    example, Ma-Hock (2009). as described in Section 5.1.3. and Asharani et al. (2008). as described in
 4    Section 5.2.1.21. indicating that mass can be considered an appropriate dose metric for some MWCNTs
 5    under certain exposure conditions. To what degree variations in other MWCNT characteristics (e.g.,
 6    length, width, bundling state) influence the appropriate application of mass-based metrics for MWCNTs
 7    is not well understood. CNTs might induce toxic effects beyond those expected based on mass
 8    concentration, however, which has been demonstrated in comparative tests of nanoparticulate carbon and
 9    quartz, commonly used indicator compounds (Donaldson et al.. 2006).
10           Alternative characteristics that have been considered as potentially relevant dose metrics for some
11    nanomaterials include particle size, surface area, surface chemistry, particle count per particle size, and
12    particle charge (Aschberger et al.. 2011; Maynard and Aitken. 2007). Although surface area has been
13    shown to be a better dose metric than mass in several rodent studies (Aschberger et al.. 2011; Sager and
14    Castranova. 2009; Tran et al.. 2000; Oberdorster.  1996). surface area has not been routinely measured or
15    recorded when examining occupational exposures (Aschberger et al.. 2011). Calculating surface area after
16    study completion is complicated by the differences in measurement techniques, the dynamic behavior of
17    MWCNTs (i.e., propensity to form bundles), and lack of thorough reporting.
18           Because the physical form of MWCNTs resembles fibers, other characteristics such as length,
19    diameter, aspect ratio, bundling state, and fiber count have been considered as characteristics potentially
20    relevant to quantifying potential exposures and doses of CNTs. Using fiber count as a dose metric can be
21    challenging, however, because MWCNT fibers generally are not uniform in size, and different sizes
22    might elicit different effects (see  Text Box 5-1). The diameters of MWCNTs in general can range from 10
23    to 200 nm (Hou et al.. 2008). and the lengths can vary widely, often by tens of microns (Donaldson et al..
24    2006).
25           Bundling also can be a relevant characteristic for considering dose-response relationships.
26    Researchers have noted that MWCNTs tend to form bundles, which then can combine into small
27    "clumps," some of which are nonrespirable (Pauluhn. 2010a). For inhalation exposures, these larger
28    clumps are therefore less toxic than free MWCNTs and MWCNT bundles in the respirable range, despite
29    the larger mass concentration. Characterization of inhalation exposure and subsequent dose based on
30    bundle size or aerodynamic diameter could therefore be more appropriate than characterization based on
31    mass or particle count. Bundling  of CNTs also could result in a toxic impact that is not observed with the
32    same mass of dispersed CNTs. For example, bundled CNTs could trigger an immune-system foreign-
33    body response because larger structures are potentially better recognized by macrophages (Johnston et al..
34    2010) (see Text Box 5-1). In vitro plant assays have shown that clustered MWCNTs fail to disperse

      March 2013                                     4-7                   DRAFT-DO NOT CITE OR QUOTE

-------
 1    throughout the culture, limiting exposure to a few cells (Tan and Fugetsu. 2007). Increasing the mass of
 2    clustered MWCNTs did not increase dispersion, rather this caused the clusters to grow larger and
 3    precipitate out of solution but did not cause a higher percentage of plant cells to be affected [Tan and
 4    Fugetsu (2007). described further in Appendix G.5.1.1.21.
 5           Aspect ratio, which refers to the ratio of a compound's length to diameter, has been shown by
 6    some to be an important characteristic for driving exposure and dose of CNTs (Kim et al.. 2011; Poland et
 7    al.. 2008). The fiber-like structure of CNTs can be considered similar to asbestos, causing many
 8    researchers to predict that the toxicity of CNTs will be driven by differences in aspect ratio, with CNTs
 9    having higher aspect ratios more frequently depositing deeper in the lungs and translocating to the pleura,
10    where mesothelioma, other cancers, and fibrosis (all effects of asbestos exposures) can occur
11    (Kim et al.. 2011) (see Additional Information Highlight Box 13). The utility of morphological
12    parameters like length, width, and aspect ratio for dose quantification is limited, however, by
13    inconsistencies in the literature regarding what constitutes "long" versus "short" or "high aspect ratio"
14    versus  "low aspect ratio." These distinctions are usually relative, based on the materials compared in an
15    individual study; specific incremental changes in length, width, or aspect ratio have not yet been
16    correlated to quantitative changes in  dose.
17            Since no single dose metric  has been identified to date as capable of accurately predicting the
18    toxicity of MWCNTs, consideration  of multiple characteristics together therefore has been proposed as a
19    potential alternative.  For example, aspect ratio and bundling state might need to be considered together.
20    Long, thin CNTs (i.e., those with higher aspect ratios) can penetrate deeply into airways, while bundled
21    CNTs are more likely to deposit in the upper airway. Particles in the upper airway can be removed though
22    mucociliary processes, whereas deposits in deeper regions are more likely to persist or translocate from
23    the lung to other tissues where they might shift the location of toxic effects  (Johnston et al.. 2010).
24           Measuring potential exposures outside of well-controlled experimental settings—for example, in
25    occupational  settings—introduces a different set of challenges that can only be addressed currently by
26    using multiple instruments and analytical techniques. As described in Text Box 4-1 and Appendix B;
27    however, many of the techniques needed for detection, measurement, and characterization of MWCNTs
28    are limited by inadequate levels of detection or restrictive measurement ranges for morphological
29    parameters. For example, a common method for counting fibers in workplace air, the National Institute
30    for Occupational Safety and Health (NIOSH) Manual of Analytical Methods (NMAM) 7400 (NIOSH.
31    1994).  does not detect fibers or bundles with diameters less than 0.25 um and does not differentiate
32    between MWCNTs and other fibers (Gustavsson et al.. 2011). Another method that can be used to
33    estimate MWCNT mass concentrations in workplace air is NMAM 5040 [highlighted in NIOSH (2010)1.
      March 2013                                      4-8                    DRAFT-DO NOT CITE OR QUOTE

-------
 1    but this method measures elemental carbon as a proxy for MWCNTs and does not automatically account
 2    for background contributions of other forms of particulate carbon (Birch. 2003).
 3           The metrics primarily used to determine potential exposure levels of MWCNTs in occupational
 4    air are total particle count by size fraction, total dust or particle mass, respirable14 dust or particle mass,
 5    inhalable15 dust or particle mass, total elemental carbon mass, and total or respirable fiber count
 6    (Xjustavsson et al.. 2011). The instruments required to capture these measurements include a suite of real-
 7    time, direct monitoring particle samplers and particle counters and a variety of area and personal air
 8    filters. The particles and fibers collected by these samplers and filters, however, are not limited to
 9    MWCNTs; instead, all particles or fibers within a certain size range are collected and counted, which
10    might lead to overestimation of exposure concentrations. This lack of specificity introduces a degree of
11    uncertainty that can be reduced only with adequate characterization of the samples collected on the filters
12    (Xjustavsson et al.. 2011). In general, characterization involves a form of electron microscopy to verify
13    morphological features and energy-dispersive spectroscopy to verify the chemical identity of the samples.
14    In summary, most estimates of workplace exposure are derived by relating real-time data on particle or
15    fiber counts with filter samples analyzed to determine particle or fiber mass, particle or fiber morphology,
16    and chemical composition; all analytical techniques involved in this multistep estimation of exposure
17    concentrations have analytical limitations that produce estimates with varying amounts of uncertainty
18    (Dahmetal..2011a).


      4.1.2. Concentrations in Environmental Media and  Indoor
              Environments
19           As described in the previous section, exposures can be estimated by combining knowledge of
20    concentrations in exposure media with assumptions about contact of humans, biota, or abiotic surfaces
21    with those media. The following sections describe the information available on concentrations of
22    MWCNTs and related substances in environmental media (i.e., air, water, soil).
      14The 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). The National
      Institute for Occupational Safety and Health considers particles with aerodynamic diameters >10 \im to be larger
      than respirable (Bartlev and Feldman. 1998).
        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.
      March 2013                                      4-9                    DRAFT-DO NOT CITE OR QUOTE

-------
                                     Additional Information Highlight Box 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., 2011b). 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):  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 etal., 2011b), in a
       way similar to differences between studies of toxicity outcomes with different types of MWCNTs.
      4.1.2.1. Outdoor Air

 1            A recent review of toxicity and exposure to CNTs indicates that ambient exposure to CNTs is

 2    possible, but very little outdoor environmental sampling data are available (Aschberger et al.. 2010).

 3    Dahm et al. (2011 a) measured background elemental carbon (inhalable fraction) outside CNT primary

 4    and secondary manufacturing facilities. Concentrations ranged from not detected (limit of detection 0.2 to

 5    0.5 ug elemental carbon/filter) to 0.76 ug/m at MWCNT manufacturing facilities. This information is of

 6    limited utility, however, because elemental carbon particles can be produced by many sources, and no

 7    electron microscopic analysis was conducted to determine if the particles collected included CNTs.

 8            Researchers have found MWCNTs in methane or propane flames from kitchen stoves, and

 9    automotive exhaust is thought to be a source of MWCNTs (Lagally et al., 2012; Aschberger et al.. 2010).

10    CNTs were found in the lung tissues of World Trade Center patients following the collapse of the

11    buildings on September 11, 2001, which indicates that dust  or smoke in urban environments can contain

12    CNTs (Aschberger etal..  2010).

13            As summarized in Table 3-2. two  life cycle-based models estimated flow of CNTs and other

14    nanomaterials from the products containing them to environmental compartments (Gottschalk et al.. 2009;

15    Mueller and Nowack. 2008). The predicted environmental concentrations (PECs) of CNTs in the air were

16    estimated by Mueller and Nowack (2008)  as 1.5 x  10"3and2.3 x 10"3 ug/m3 for the realistic exposure and
      March 2013                                        4-10                    DRAFT-DO NOT CITE OR QUOTE

-------
 1    high exposure scenarios, respectively; concentrations in airborne dust were not examined. Gottschalk et
 2    al. (2009) estimated a mode PEC of 1 x 10'6 ug/m3 CNTs in U.S. air for 2008. Differences between the
 3    models and the resulting estimates are discussed in Section 3.5.

      4.1.2.2. Aquatic Systems - Sediment and Surface Water
 4           No data were found on environmental concentrations of MWCNTs in aquatic environments, but
 5    as summarized in Table 3-2. two substance flow analyses have estimated PECs of CNTs in surface water
 6    (Gottschalk et al.. 2009; Mueller and Nowack. 2008). Mueller and Nowack (2008) estimated surface
 7    water CNT PECs of 5 x 10^ and 8 x  10"4 ug/L for the realistic and high exposure scenarios, respectively.
 8    Gottschalk et al. (2009) estimated not only a mode PEC of 1  x 10"6 ug/L for CNTs in surface water, but
 9    also an annual increase of 4.6 x  10"2 ug/kg-year to U.S. sediment.

      4.1.2.3. Terrestrial Systems - Soil
10           No data were found on environmental concentrations of MWCNTs in surface soil samples, but as
11    summarized in Table 3-2. two substance flow analyses have estimated PECs of CNTs in soil  (Gottschalk
12    et al.. 2009: Mueller and Nowack. 2008). Mueller and Nowack (2008) estimated soil CNT PECs of
13    1 x 10"2 and 2 x 10"2 ug/kg for the realistic and high exposure scenarios, respectively. Gottschalk et al.
14    (2009) estimated an annual increase of 5.6 x 10"4 ug/kg-year to U.S. soil.

      4.1.2.4. Occupational Settings-Air
15           No studies were found that measured MWCNT concentrations in air in facilities where textiles
16    containing MWCNTs are manufactured. Multiple studies have collected particles and fibers in workplace
17    air to attempt to estimate MWCNT concentrations at the emission source, in area air, and in the personal
18    breathing zone of workers in small laboratories or research and development facilities (Johnson et al..
19    2010; Lee etal.. 2010: Methner et al.. 2010: Bello et al.. 2008: Han et al.. 2008). as well as in larger pilot
20    plants and manufacturing and handling facilities (Dahm et al.. 201 la; Lee etal.. 2010; Takava et al..
21    2010). Some of these studies are discussed below, and Appendix E. Table E-6 presents  additional
22    information on particle, fiber, and MWCNT concentrations for the studies that reported quantitative
23    concentration values associated with MWCNTs.

      Concentrations in MWCNT Research Laboratories
24           Han et al. (2008)  measured concentrations of total particles in the area air and personal breathing
25    zones of workers in an MWCNT research facility. Two particle sizers were used to count particles  with

      March 2013                                     4-11                   DRAFT-DO NOT CITE OR QUOTE

-------
 1    sizes ranging from 14 to 630 nm and 0.5 to 20 um, respectively, and a portable aethalometer measured the
 2    mass of carbon black in the total particulate matter in the air. All fibers with aspect ratios greater than 3:1
 3    were collected on filters and analyzed using electron microscopy, and MWCNTs were distinguished from
 4    asbestos fibers using energy-dispersive spectroscopy. During the blending process, which creates a
 5    uniform size-distributed CNT powder, particulate matter concentrations in the air near the open blender
 6    ranged from 434.5 ug/m3 without exposure controls to no detection (limit of detection not reported) with
 7    exposure controls. The maximum MWCNT number concentration ranged between 172.9 (area air sample)
 8    and 193.6 (personal air sample) fibers per cm3 air during blending without exposure controls, and
 9    between 0.018 (personal air sample) and 0.05 (area air sample) fiber per cm3 air during blending with
10    exposure controls. During weighing and spraying, particulate matter concentrations ranged from 36.6
11    (area air sample) to 193.0 (personal air sample) ug/m3 without exposure controls and from below the level
12    of detection (area air sample; limit of detection not reported) to 30.9 ug/m3 (personal air sample) with
13    exposure controls. The maximum MWCNT number concentrations during weighing and spraying were
14    below detection (limit of detection not reported) in the absence of controls, and up to 1.997 fibers per cm3
15    air after controls (Han et al.. 2008). The reason for the increase in MWCNTs following implementation of
16    exposure controls (in this case, the control was "a simple fan") was not discussed. The maximum
17    MWCNT length observed was 1.5 um, which is smaller than the World Health Organization's minimum
18    length of 5 um for classification as a fiber (Aschberger et al.. 2010).
19           Using a suite of real-time particle sizers, particle counters, and filters with electron microscopy
20    and energy-dispersive spectroscopy analyses, Methner et al. (2010) measured and characterized
21    particulate matter in carbon-based nanomaterial research and development facilities. The highest particle
22    number concentration for particles ranging in size from 10 to 1,000 nm was measured when engineering
23    controls were turned off during the opening of an MWCNT growth chamber of a pulsed laser deposition
24    reactor. The maximum particle number concentration was 42,400 particles per cm3 in the absence  of
25    engineering controls, but when the same activity was performed in a sealed system with vacuum exhaust,
26    the particle number was reduced to 300 particles per cm3 (Methner et al.. 2010).
27           Johnson et al. (2010) used real-time particle sizers to count total particles per liter air for six size
28    cuts (300, 500, 1,000, 3,000, 5,000, and 10,000 nm) and per cubic centimeter air for the cumulative 10- to
29    1,000-nm size fraction of particles released to the air in a laboratory while two tasks were being
30    performed with raw and functionalized MWCNTs. The first task involved weighing MWCNTs and
31    transferring them to a beaker of stirring water, and the second task involved sonicating  a previously
32    mixed solution containing reconstituted water and 100 mg/L MWCNTs with 100 mg/L natural organic
33    matter. Filter samples also were collected at the emission source  (i.e., as close as possible to the
34    instruments used for each task) and in area air, and samples were analyzed using electron microscopy and

      March 2013                                     4-12                   DRAFT-DO  NOT CITE OR QUOTE

-------
 1    energy-dispersive spectroscopy. In general, particle number concentrations in the air were inversely
 2    proportional to particle size, with either zero or very few particles detected for the 5,000- and 10,000-nm
 3    size cuts for both types of MWCNTs used in both tasks. The maximum background-adjusted particle
 4    number concentrations for the raw MWCNTs occurred within the 300-nm size fraction; these
 5    concentrations were 123,403 particles/L air (above the upper limit of quantification) during weighing and
 6    transferring and 42,796 particles/L air during sonication (Johnson et al.. 2010). The particle number
 7    concentrations measured for the functionalized MWCNTs (MWCNT-OH) exhibited different trends from
 8    those for the raw MWCNTs. First, no additional 300-nm particles were detected above the background
 9    level during weighing and transferring. Second, particle counts were higher during sonication of
10    functionalized MWCNTs than during weighing and transferring, which is opposite of the trend observed
11    for the raw MWCNTs. Whereas the maximum background-adjusted particle number concentration was
12    3,065 particles/L air (500-nm size fraction) during weighing and transferring, the maximum concentration
13    was 144,623 particles/L air (above the limit of quantification;  300-nm size fraction) during sonication of
14    functionalized MWCNTs. The particle number concentrations measured for the cumulative 1-to 1,000-
15    nm size range were 1,576 and 2,776 particle/cm  air for the  raw MWCNTs and 676 and 726 particles/cm3
16    for the functionalized MWCNTs during weighing/transferring and sonicating, respectively. Johnson et al.
17    (2010) proposed that the cumulative measurements do not follow the same trends as the size cut
18    measurements because of the inclusion of particles smaller than 300-nm in the cumulative particle
19    counter.
20            Bello et al. (2008) used a real-time particle sizer and a particle counter to count total particles
21    with sizes ranging from 5.6 to 560 nm and 10 to 1,000 nm, respectively in the area air of a university
22    research laboratory. Personal air samples also were collected on filters and analyzed using scanning
23    electron microscopy and energy-dispersive spectroscopy. Bello et al. (2008) found that removal of
24    MWCNTs from the reactor furnace and detachment of MWCNTs from the nanotube growth substrate
25    during chemical vapor deposition (CVD) did not increase total airborne particle concentrations compared
26    to background. Additionally, no MWCNTs were observed in the personal air sample of a furnace operator
27    (Bello et al.. 2008).
28            In a later study, Bello et al. (2009) measured particulate matter, respirable particulate matter, and
29    respirable fibers in a laboratory during dry and wet machining of composite materials with and without
30    CNTs. Particle sizers were used to detect and count all particles with sizes ranging from 5 nm to 20 urn,
31    and particles and fibers in the respirable range were collected on filters near the source (i.e., 10 cm from
32    the machined composite) and in the breathing zone of the operator; filter samples were analyzed using
33    electron microscopy. Although the dry-cutting process did result in statistically significant increases in
34    airborne particles and fibers, no statistically significant differences were noted in the particle number,

      March 2013                                     4-13                   DRAFT-DO NOT CITE OR QUOTE

-------
 1    particle sizes, or total dust generated by dry cutting the composites with and without CNTs. Furthermore,
 2    analyses of filter samples revealed no single or bundled CNTs in the particles and fibers collected, and no
 3    CNTs were observed "sticking out" of the CNT composites (Bello et al., 2009).

      Concentrations in MWCNT Manufacturing and Packing Facilities
 4           Dahm et al. (2011 a) used a series of filters to measure the inhalable size fractions of elemental
 5    carbon and determine CNT and carbon nanofiber "structure" counts (defined as "single CNTs to large
 6    agglomerates" viewed using electron microscopy) in the area air and personal breathing zones of six
 7    pilot-scale CNT or carbon nanofiber primary and secondary manufacturing facilities. Sampling was
 8    conducted while workers performed various tasks, including harvesting, sonicating, weighing, extruding,
 9    manually transferring, and mixing MWCNTs; spray coating a product with an MWCNT solution; milling
10    MWCNT composites; and collecting and disposing of waste from MWCNT work areas. The elemental
11    carbon concentrations in the personal breathing zone samples generally were higher than the area air
12    samples. Inhalable elemental carbon concentrations in personal breathing zones ranged from 1.13 ug/m3
13    (sonicating, sieving, and spray coating) to 2.74 ug/m3 (harvesting) at the primary MWCNT manufacturing
14    facilities and from 0.8 ug/m3 (office work outside lab space) to 7.86 ug/m3 (extrusion, weighing, and
15    batch mixing) at the secondary MWCNT manufacturing facilities. The CNT structure counts in the
16    personal breathing zones ranged from 0.010 structure/cm3 (sonicating, sieving, and spray coating) to
17    0.399 structure/cm3 (harvesting) at the primary facilities, and from none observed (weighing, sonicating,
18    milling) to 0.242 structure/cm3 (extrusion, weighing, and batch mixing) at the primary and secondary
19    MWCNT manufacturing facilities (Dahm et al.. 201 la).
20           Inhalable elemental carbon concentrations in area air samples from Dahm et al. (2011 a) ranged
21    from not detected (sonicating, sieving, and spray coating; limits of detection ranged from 0.2 to 0.5  ug
22    elemental carbon/filter) to 4.62  ug/m3 (harvesting) and from not detected (weighing, sonicating, milling,
23    and mixing) to 1.01 ug/m3 (extrusion, weighing, and batch mixing) at the primary and secondary
24    MWCNT manufacturing facilities, respectively. The CNT structure counts ranged from none observed
25    (production and harvesting) to 0.134 structure/cm3 (harvesting at a different facility) and from none
26    observed (weighing, milling) to 0.008 structure/cm3 (extrusion, weighing, and batch mixing) in the area
27    air samples at the primary and secondary MWCNT manufacturing facilities, respectively (Dahm et al..
28    201 la).
29           Lee et al.  (2010) collected filter samples to measure respirable dust concentrations in the area air
30    and personal breathing zones of workers at three MWCNT manufacturing facilities and four research and
31    development laboratories throughout a normal workday. Fibers with aspect ratios greater than 3:1 were
32    collected on the filters and analyzed using electron microscopy, and MWCNTs were chemically identified

      March 2013                                      4-14                   DRAFT-DO NOT CITE OR QUOTE

-------
 1    using energy-dispersive spectroscopy. A suite of particle sizers, differential mobility analyzers, and
 2    particle counters also was used to count particles with sizes ranging from 14 to 500 nm, and a dust
 3    monitor was used to capture number concentrations of particle ranges from 0.25 to 32 um in diameter.
 4    A portable aethalometer measured the mass of carbon black in the total particulate matter in the air. Lee et
 5    al. (2010) generally found that the highest increases in particle number concentrations compared to
 6    background were observed following the opening of the chemical vapor disposition (CVD) chamber after
 7    MWCNT synthesis. Increases in carbon black concentrations at this time were minimal, however,
 8    suggesting that most of the particles released were more likely to be metal catalysts than MWCNTs.
 9    Furthermore, the authors reported only one measurement of a detectable amount of MWCNTs on one
10    filter from a single facility (0.00312 tube/cm ); the study authors could not determine whether the lack of
11    MWCNT detection reflected a lack of MWCNTs in workplace air or flaws in the sampling process or
12    analytical methods.
13           Total dust and respirable dust concentrations were measured in a study of two MWCNT packing
14    facilities, one of which was manually operated and the other automated [(Takava et al., 2010) English
15    translation available only for abstract].  Total dust concentrations in the area air, of both MWCNT packing
16    facilities, were approximately 240 ug/m3. Both total and respirable dust concentrations, however, were
17    substantially higher in the manual packing facility (total: 2,390 ug/m3; respirable: 390 ug/m3) than in the
18    automated packing facility (total: 290 ug/m3; respirable: 80 ug/m3).

      4.1.2.5. Residential Settings-Air and Dust
19           No data were found on concentrations of MWCNTs in household air or dust.

      4.1.2.6. Nonresidential Settings - Air and Dust
20           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
21           No data were found on concentrations of MWCNTs in air or dust in vehicles or aircraft.
      March 2013                                    4-15                   DRAFT-DO NOT CITE OR QUOTE

-------
      4.2.  Human  Exposure and Kinetics Leading to Dose
 1           Limited data were found that measured or quantified human exposure to MWCNTs. Data on
 2    concentrations of MWCNTs measured in media such as air, soil, or dust in various settings (described in
 3    Section 4.1.2). however, can be used in conjunction with activity pattern and other exposure factor data
 4    [such as those described in The Exposure Factors Handbook (U.S. EPA. 2011b)1 to inform estimates of
 5    potential exposure through the various exposure pathways and scenario characteristics described in this
 6    section.
 7           The types of human exposure  scenarios described here can be divided into four broad groups:
 8    occupational, consumer, general public, and highly exposed populations. For the purposes of this case
 9    study, occupational exposures include occupational  exposures during synthesis, processing, or handling
10    of MWCNTs; manufacturing of flame retardants, application of the flame retardants to textiles, or textile
11    finishing and upholstering; storage of the MWCNTs, flame-retardant formulations, treated textiles, or
12    upholstered products; disposal of MWCNTs, flame-retardant formulations, treated textiles, or upholstered
13    products; and repurposing or recycling of treated upholstery textiles and end-user products (e.g.,
14    furniture). Consumer exposure scenarios include the intended or unavoidable use of treated upholstery
15    textiles in residential and nonresidential spaces, including on household or institutional/office furniture, in
16    vehicles, and in aircraft; unintended uses of treated upholstery textiles or end-use products such as reuse
17    or repurposing of furniture for something other than its original intended use; or recycling of upholstery
18    textiles for new uses. General public exposure includes primary exposure to members of the community
19    near manufacturing, disposal, or recycling facilities  and secondary exposure to the  general public through
20    environmental routes such as air, soil, or water. Highly exposed populations refers  to exposure scenarios
21    that are expected to occur via similar pathways as outlined for consumers and the general public, but
22    where exposure levels are expected to be higher due to key differences in population characteristics such
23    as those described in The Child-Specific Exposure Factors Handbook (U.S. EPA. 2008a).
      March 2013                                     4-16                  DRAFT-DO NOT CITE OR QUOTE

-------
4.2.1. Occupational Exposure Pathway Scenarios
                                  Priority Research Area: Exposure Route
      Occupational
   gr
 V 
-------
  1            Limited data were found to determine the extent of occupational exposures to MWCNTs during
  2    the material synthesis, processing, and handling phases or to the flame-retardant product during
  3    formulation, application, storage, and disposal phases. See Section 4.1.2.4 for MWCNT concentrations
  4    measured in occupational settings, which could be applied with the exposure pathways and scenario
  5    characteristics described below to estimate potential exposures through scenario evaluation.
                                     Additional Information Highlight Box 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. 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) influence dermal absorption.
      4.2.1.1. Synthesis, Processing, and Handling
 6            As discussed in Section 2.2.2. synthesis of MWCNTs is achieved by one of three processes:
 7    CVD, arc discharge, and laser ablation. Although many facilities use engineering controls (e.g., fume
 8    hoods, closed production systems, high-efficiency particulate air-filtered vacuums) and require workers to
 9    wear personal protective equipment (e.g., gloves, respirators, paper face masks, safety glasses, lab coats,
10    Tyvek clean suits) to minimize exposure to MWCNTs (Dahm etal.. 2011 a), not all facilities comply with
11    the General Safe Practices for Working with Engineered Nanomaterials in Research Laboratories, which
12    outlines the recommendations by NIOSH (2012). Many facilities do not employ the same level of
13    protective measures, and in many cases, the filtration technologies and personal protective equipment are
14    not appropriate for or sufficiently protective against exposures to nanomaterials (Dahm etal.. 201 la).
15    Therefore, exposure during handling and other operations might still occur, for example, when a reaction
16    chamber is opened to recover MWCNTs; while extracting, weighing, or manually transporting materials;
17    or during maintenance and cleaning of equipment (Dahm etal.. 201 la: Fleury etal.. 2011; Aschberger et
18    al.. 2010). Workers performing each operation are expected to be exposed to peak concentrations of
19    MWCNTs for only a short time while carrying out MWCNT handling tasks, but multiple production
20    cycles might occur within a day, resulting in several opportunities for short-duration, acute exposures

      March 2013                                       4-18                    DRAFT-DO NOT CITE OR QUOTE

-------
 1    throughout the workday (Dahm et al., 201 la; Lee et al., 2010). Most occupational exposure studies to
 2    date have examined these short-duration, task-specific exposures instead of full-shift exposures, and task-
 3    specific exposures have been evaluated only for a limited set of handling operations (see Section 4.1.2.4).
 4    In general, MWCNTs observed in air and settled on surfaces in occupational environments during
 5    synthesis, processing, and handling are in bundled form, but exposure to single MWCNTs is possible
 6    (NIOSH. 2010). The pathways through which workers might be exposed to MWCNTs and MWCNT
 7    bundles during synthesis, processing, and handling scenarios are described below:
 8           •   Inhalation. Handling dry powder might be the activity most likely to lead to inhalation
 9               exposures during production. Dahm et al. (201 la) and Johnson et al. (2010) observed that
10               workers handling dry powder often turned off vents, hoods, fans and other engineering
11               controls to avoid disturbing and dispersing MWCNTs. Handling processes such as weighing,
12               blending, transfer to containers, or maintenance also could result in inhalation exposure
13               (Dahm et al.. 201 la:  Fleury et al.. 2011; Aschberger et al.. 2010).
14           •   Oral. Secondary oral exposures might occur if inhaled MWCNTs or MWCNTs that deposit
15               on the skin, food, or food-contact surfaces are subsequently ingested.
16           •   Dermal. Particles generated during manufacturing and processing of CNTs can settle on the
17               skin of workers if proper personal protective equipment is not worn (Lam et al.. 2006).
18               A study evaluating occupational exposure to CNTs during synthesis, processing, and
19               handling estimated (using adsorbed metals as proxy) that, on average, 0.2 to 6 mg of single-
20               walled CNTs are deposited on the gloves  covering each hand of workers during routine
21               operations. Although the cotton gloves worn by workers could have adsorbed more CNTs
22               than bare skin or latex, the study illustrates that dermal exposure to  CNTs could occur in
23               laboratory settings (Maynard et al., 2004).

      4.2.1.2. Formulation of Flame Retardant, Application to Textiles, Upholstering
24           No data were found on occupational exposures to MWCNTs during formulation of the flame
25    retardant, application of the flame retardant to textiles, or textile finishing and upholstering. Furthermore,
26    information on the processes for preparing MWCNT flame retardants and for applying them to textiles
27    are lacking. To confer the desired flame-retardant properties of MWCNTs to the textile product, however,
28    MWCNTs must be well dispersed in a polymer medium. To promote dispersion, MWCNTs are
29    sometimes ground or pulverized, which could lead to the release of single MWCNTs or bundles.
30    MWCNTs also might be mixed or sonicated, which could generate airborne water droplets, or mists,
31    containing nanomaterials that then can be inhaled or deposited on surfaces (Fleury et al., 2011;
32    Aschberger etal.. 2010; Johnson  etal.. 2010).
33           As described in Section 2.2.4.1. one industry representative reported that textiles can be
34    immersed or spray coated with MWCNT flame retardants. The immersion method could result in worker
35    exposures, and exposures are expected to be highest during equipment handling and cleaning. Should
      March 2013                                    4-19                   DRAFT-DO NOT CITE OR QUOTE

-------
 1    MWCNT flame retardants be sprayed onto textiles, however, MWCNTs in the wet polymer matrix might
 2    be released as mists, which can occur even with nonvolatile liquids (U.S. EPA. 2005).
 3           Exposures to MWCNTs also might occur when the treated upholstery textile is machined, drilled
 4    (Aschberger et al., 2010). or otherwise abraded during the textile finishing or upholstering processes. Wet
 5    machining and dry machining of advanced nanomaterial composite systems were evaluated for generation
 6    of respirable CNTs. Wet-cutting methods were not found to produce exposures significantly different
 7    from background, while dry-cutting methods created statistically significant quantities of nanoscale and
 8    fine particles and fibers composed of the composite material (i.e., no single or bundled CNTs were
 9    observed in the samples) (Bello et al.. 2009).
10           Equipment cleaning can be a key contributor to work exposure during product manufacture,
11    application, and upholstering. One study identified equipment cleaning as one of the most  important
12    occupational exposure  scenarios because it often requires workers to be in direct contact with molten
13    polymers and  residues  containing CNTs (Tleury et al.. 2011).
14           The pathways through which workers might be exposed to single MWCNTs or MWCNT bundles
15    during general formulation of the flame  retardant, application of the flame retardant to the  textile, and
16    textile finishing and upholstering scenarios are expected to be comparable to those described in Section
17    4.2.1.1 on exposures during synthesis, processing, and handling. Additional considerations pertaining to
18    exposures to MWCNTs in combination with polymer ingredients,  textile fibers or scraps, or other product
19    constituents during these scenarios are described below:
20           •   Inhalation. Spray coating textiles with MWCNT flame retardants could result in inhalation
21               exposures to mists containing MWCNTs embedded in a liquid polymer mixture. Dry-cutting
22               MWCNT-treated textiles during tailoring and upholstering could lead to inhalation of fine
23               and ultrafine particles comprising MWCNT-polymer composites and textile dusts.
24           •   Oral. Secondary oral exposures might occur if inhaled MWCNTs in mists and particulate
25               form or the MWCNT mists  or particles that deposit on the skin, food, or food-contact
26               surfaces are subsequently ingested
27           •   Dermal. MWCNTs in mists and particulate form generated during product manufacturing
28               can land on the skin of workers if proper personal protective equipment is not  worn
29               (Aschberger et al.. 2010; Johnson et al.. 2010; Lam et al.. 2006). The liquid flame-retardant
30               coating also can be spilled directly onto the skin.

      4.2.1.3. Storage of MWCNTs, Flame-Retardant Formulations, Treated Textiles, and
              Upholstered Products
31           As described in Appendix G.2.2. MWCNTs and the flame-retardant formulations to which they
32    are added are expected to be stored in sealed receptacles that would limit potential for worker exposures
      March 2013                                    4-20                   DRAFT-DO NOT CITE OR QUOTE

-------
 1    to these materials during storage. Defective packaging and accidental spills or releases, however, could
 2    lead to rare exposures during storage operations.
 3            Although no information was identified regarding procedures for storing treated upholstery
 4    textiles, these products are likely packaged to protect them from exposure to elements like water and light
 5    that could damage their aesthetics. Such packaging also is expected to limit exposures of workers to the
 6    flame-retardant coatings. Once the textiles have been applied as upholstery to end-use products, these
 7    products also are expected to be enclosed in protective packaging. Some surfaces of bulkier products
 8    (e.g., furniture), however, might remain uncovered, which could lead to worker exposures during storage
 9    operations, or exposures might occur during application and removal of packaging materials to and from
10    the product. Dust also can accumulate in storage facilities that frequently store textiles and textile
11    products, and MWCNTs that escape from the product matrix could sorb to dust particles. Ventilation
12    technologies and other contamination-prevention strategies like those used by manufacturing facilities are
13    not expected to be in place in storage facilities. Dust that has settled on surfaces in storage facilities can
14    be disturbed by worker operations, resuspended, and transported to other locations.
15            Although MWCNTs are not expected to be highly volatile, off-gassing of more volatile
16    components of the treated textiles might occur during storage of treated textiles or upholstered products.
17    Furthermore,  due to the additive nature of MWCNT flame retardants, covalent bonding between the flame
18    retardant and  the textile does not occur, suggesting that flame-retardant coatings that are loosely attached
19    to the textile surface might slough off during storage or handling. Because MWCNT flame retardants are
20    generally added to the back of the textile, however, the likelihood of this detachment seems low.
21            No data were found on occupational exposures to MWCNTs during storage throughout the
22    product life cycle of flame-retardant upholstery textile coating. The pathways through which workers
23    might be exposed to MWCNTs alone or MWCNTs in combination with polymer ingredients, textile
24    fibers or scraps, dusts, or other product constituents during storage of MWCNTs and MWCNT flame-
25    retardant formulations are expected to be comparable to those described in Sections 4.2.1.1 and 4.2.1.2
26    (exposures during synthesis, processing, and handling and during formulation of the flame retardant,
27    application to textiles, and upholstering). The pathways through which workers might be exposed to
28    MWCNT bundles or MWCNTs adsorbed to dust during storage of treated textiles and upholstered
29    products are expected to be inhalation, oral, and dermal. Workers could inhale volatile components of the
30    flame-retardant coating or MWCNTs adsorbed to dust in storage facilities, particularly facilities that are
31    not well ventilated. Higher levels of dust in textile storage facilities could lead to increased transport of
32    MWCNTs adsorbed to dust. This could result in oral exposures to MWCNTs in dust transported to break
33    rooms, homes (via clothes), and other locations where  MWCNTs adsorbed to dust can be unintentionally
34    ingested while eating or due to hand-to-mouth activity. Additionally, MWCNTs adsorbed to dust could be

      March 2013                                     4-21                  DRAFT-DO NOT CITE OR QUOTE

-------
 1    resuspended by worker activities and deposit on the skin of workers if proper personal protective
 2    equipment is not worn.

      4.2.1.4. Disposal and Recycling of MWCNTs, Flame-Retardant Formulations, Treated
               Textiles, and Upholstered Products
 3           As described in Section 2.5. large-scale disposal, recycling, and reuse of MWCNTs and the
 4    flame-retardant formulations to which they are added are unlikely, but containers used to store these
 5    products might enter the waste stream, and workers at disposal and recycling facilities could be exposed
 6    to product residues remaining in these containers.
 7           Disposal and recycling of treated textiles and upholstered products, however, is prevalent. Mixing
 8    and compacting of waste for land-filling; cleaning, shredding, blending, melting, and spinning scrap
 9    textiles for recycling; and incomplete incineration of treated upholstery textiles all could result in
10    exposure of workers to MWCNTs (Chaudhry et al.. 2009). primarily in combination with other product
11    constituents and dusts.
12           No data were found on occupational exposures to MWCNTs during disposal and recycling
13    throughout the product life cycle of flame-retardant upholstery textile coating. The pathways through
14    which workers might be exposed to MWCNTs during general disposal and recycling of MWCNTs and
15    flame-retardant formulations are expected to be comparable to those described in Section 4.2.1.1
16    (exposures during synthesis, processing, and handling); worker exposure pathways for MWCNTs in
17    combination with polymer ingredients, textile fibers or scraps, or other product constituents during
18    disposal and recycling of treated textiles and upholstered products are expected to be similar to those
19    described in Section 4.2.1.2 (exposures during formulation of the flame retardant, application to textiles,
20    and upholstering) and Section 4.2.1.3 (exposure during storage and distribution) for these products.
      March 2013                                    4-22                  DRAFT-DO NOT CITE OR QUOTE

-------
4.2.2. Consumer Exposure Pathway Scenarios
                                  Priority Research Area: Exposure Route
       Consumer
   gr
 V 
-------
 1            No studies were found that evaluated the potential for consumer exposure to MWCNTs from any

 2    consumer product. As a result, probable consumer exposure pathways and scenario characteristics for

 3    exposure to free MWCNTs, bundled MWCNTs, and MWCNTs in combination with the polymer matrix,

 4    textile fibers or scraps, or other product constituents cannot be differentiated at this time. As discussed in

 5    Chapter 2, different MWCNT flame-retardant production processes are expected to result in differences in

 6    release rates and release forms, which in turn will affect the magnitude of exposure during consumer use

 7    and the form of the material to which consumers are exposed (Motzkus et al.. 2012). Based on the

 8    physicochemical properties of MWCNTs, the  assumption that MWCNTs and associated substances

 9    released from consumer products will be present in the particulate phase is reasonable.
                                     Additional Information Highlight Box 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).  Based on the physicochemical properties of
       MWCNTs, MWCNTs and associated substances likely would be released from upholstery textiles in the particulate phase, in a
       manner similar to BDE-209 (Nowacketal., 2012) (see Section 3J. and Appendix H.3.1). 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 4.2.21). 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.
      4.2.2.1.  Intended Use - Upholstered Products in Residential Spaces

10            Although flame-retardant upholstery textiles typically are used in nonresidential settings (see

11    Section 4.2.2.2), some residential upholstered products, particularly mattresses, are known to contain

12    flame retardants, and other upholstered furniture products, like couches, sometimes might be treated with

13    flame retardants (Rose et al., 2010). As introduced in Section 2.4, upholstered products are expected to be

14    used for many years, and contact with the textile might be frequent and prolonged, which could introduce

15    substantial wear and tear to the textile product.  In addition, upholstery in residential spaces might

16    frequently be exposed to cleaning products, sweat, food, and other substances that could affect the

17    properties of the textile and the flame-retardant coating.
      March 2013                                       4-24                    DRAFT-DO NOT CITE OR QUOTE

-------
 1           The pathways through which consumers might be exposed in residential settings to MWCNTs

 2    during general consumer use scenarios for end products upholstered with MWCNT flame-retardant

 3    coatings are described below:

 4           •   Inhalation. Chronic inhalation of MWCNTs in combination with other product constituents
 5               and dust could occur following release from upholstered products over time (due to wear and
 6               tear from anticipated use, aging of materials, abrasion, UV light, water, cleaning chemicals,
 7               among other factors; see Section 2.4.2). MWCNTs could settle onto surfaces, where they
 8               might be disturbed and re-entrained, after which they could be inhaled by residents. Whether
 9               inhalation is a primary route of consumer exposure for MWCNTs, particularly when
10               embedded in a polymer matrix, is unknown.

11           •   Oral. MWCNTs in combination with other product constituents and dust could be ingested
12               after settling on food and food-contact surfaces or following hand-to-mouth activity. Whether
13               ingestion is a primary  route of consumer exposure for MWCNTs, particularly when
14               embedded in a polymer matrix, is unknown. Preliminary, unpublished studies presented at a
15               public meeting indicate, however, that MWCNTs could be released from flame-retardant
16               barrier fabrics and polyurethane foams in very small amounts during normal wear and tear
17               (Uddin and Nyden.  201 Ib); these MWCNTs could settle onto food, food-contact surfaces, or
18               other surfaces where children could be exposed via hand-to-mouth activity during use.

19           •   Dermal. Dermal exposure to MWCNTs in combination with other product constituents  and
20               dust might occur while touching the textile surface (particularly if the  portion of the textile
21               that has been treated with the flame-retardant coating is exposed) or touching surfaces upon
22               which particles have settled. Whether dermal uptake is a primary route of consumer exposure
23               for MWCNTs, particularly when embedded in a polymer matrix, is unknown. Preliminary,
24               unpublished studies indicate, however, that MWCNTs could be released from flame-retardant
25               barrier fabrics and polyurethane foams in very small amounts during normal wear and tear;
26               these MWCNTs could contact skin directly during use (Uddin and Nyden. 201 Ib).


      4.2.2.2.  Intended Use  - Upholstered Products in Nonresidential Spaces

27           Due to regulations requiring that upholstery textiles used in nonresidential settings pass flame-

28    retardancy tests (see Table 1-3). many upholstery textiles in public, commercial, and institutional settings

29    are treated with flame retardants. The characteristics of the different settings in which these products are

30    used can vary considerably. For example, flame-retardant upholstery textiles might be used in seating for

31    airports and other transportation hubs and in waiting rooms, office buildings, penal institutions, and other

32    nonresidential spaces that can range from very small to very large and where consumers might spend

33    varying amounts of time. Some  scenarios for nonresidential exposures are not likely to differ from those

34    expected from residential exposures, but a few key differences do exist. For example:

35           •   Exposures to flame-retardant upholstery coatings in public spaces might be unavoidable.
36               Although consumers have  some control over which products they bring into their home,
37               consumers have no  control over the products they encounter in public spaces.
      March 2013                                    4-25                   DRAFT-DO NOT CITE OR QUOTE

-------
 1           •   Some nonresidential exposures might occur over long periods of time and for extended
 2               intervals (e.g., silting in the same office chair every day over the course of several work
 3               years), while some might occur infrequently and for short periods of time (e.g., sitting in
 4               seating at the airport waiting for a flight).

 5           •   Products in public spaces might experience higher activity levels, more frequent cleaning,
 6               and less care to the textile surface, all of which could damage or weaken the textile matrix
 7               and influence releases and exposures.

 8           With the exception of these potential differences in exposure settings and activity patterns, the

 9    pathways and scenarios through which consumers might be exposed in nonresidential settings to

10    MWCNTs during general consumer use scenarios for end products upholstered with MWCNT flame-

11    retardant coatings are not expected to differ from those described previously in Section 4.2.2.1 on

12    exposures from intended use of upholstered products in residential spaces.


      4.2.2.3.  Intended Use -Aircraft and Automobile Upholstery

13           Flame-retardant upholstery can be used for seating, draperies, carpets, and other textiles in

14    passenger cars and public and private transportation.

15           The pathways through which consumers might be exposed in vehicles (including airplanes) to

16    MWCNTs during general consumer use scenarios for end products upholstered with MWCNT flame-

17    retardant coatings are described below:

18           •   Inhalation. Inhalation of MWCNTs adsorbed to dust from worn or abraded automobile
19               upholstery is expected to occur. The recirculation of air in aircraft cabins also might affect
20               exposure to MWCNTs, if filters do not adequately remove these particles.

21           •   Oral. Secondary oral exposures might occur if inhaled MWCNTs or MWCNTs that deposit
22               on the skin are subsequently ingested.

23           •   Dermal. Dermal exposures to MWCNTs are expected to occur, particularly when skin
24               touches the treated part of the textile directly. Dermal exposure also can occur when particles
25               in the air settle on the skin. Different exposure characteristics or scenarios (e.g., children
26               sitting in safety seats) might influence whether dermal exposure occurs, or influence the
27               extent to which exposure occurs through this pathway.
      March 2013                                     4-26                  DRAFT-DO NOT CITE OR QUOTE

-------
      4.2.2.4. Unintended Use, Repurposing, or Reuse of Treated Textiles and Upholstered
               Products
 1            As introduced in Section 2.4. unintended uses of upholstery textiles treated with MWCNT flame-
 2    retardant coatings could include repurposing of treated upholstery textiles for clothing, building
 3    insulation, other in-home or outdoor furnishings, bedding, or other purposes. The repurposing stages
 4    could introduce occupational exposures similar to those discussed in Section 4.2.1.2 (exposures during
 5    formulation of the flame retardant, application to textiles, and upholstering) and Section 4.2.1.3 (exposure
 6    during storage and distribution), as products that are treated with flame-retardant coatings are broken
 7    down and reprocessed into new products.
 8            Although no information was identified that directly addresses potential consumer exposures
 9    following unintended use or reuse of flame-retardant upholstery textiles, exposure pathways and scenarios
10    from other life-cycle stages are relevant here. Most reuse scenarios might differ little from those for
11    anticipated consumer uses, but a few key differences might occur, particularly when products are
12    repurposed for new uses or used in unintended ways. For example:
13            •   Similar processes to those involved with product manufacture (e.g., cutting, sewing) and
14               storage of textiles also might be employed for repurposing treated textiles. In this scenario,
15               however, these processes are not expected to occur in an occupational setting, but in the home
16               or another private space, where no personal protective equipment is worn and limited control
17               technologies are used. These processes, as employed for repurposing textiles, however, are
18               not expected to occur as commonly or at the same scale as in a manufacturing facility.
19            •   Older, more degraded textiles with weakened matrices might be handled directly and
20               subjected to abrasion, thereby releasing the product constituents in the vicinity of the
21               consumer conducting the repurposing.
22            •   Although dermal contact with products used for their intended purpose (e.g., furniture
23               seating) might be limited by a clothing barrier between the consumer and the treated textile,
24               should flame-retardant upholstery textiles be repurposed into clothing, direct dermal contact
25               might occur repeatedly over long periods of time.
26            With the exception of these potential differences in exposure characteristics, the pathways and
27    scenarios through which consumers might be exposed to MWCNTs during repurposing, reuse, or
28    unintended use of treated textiles and upholstered products are not expected to differ from exposure
29    pathways  associated with the cutting, tailoring other abrasive processes involved with product
30    manufacturing (Section 4.2.1.2); storage of textile products (Section 4.2.1.3); and consumer use in
31    residential and nonresidential  spaces (Sections 4.2.2.1 and 4.2.2.2).
      March 2013                                     4-27                  DRAFT-DO NOT CITE OR QUOTE

-------
     4.2.3. General Public Exposure Pathway Scenarios through
             Environmental Media
                                Unprioritized Research Area: Exposure Route
         I             1
     £ re
     LU 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
     ~H    M    L~~              ML           H    M    L
  E-RRF Confidence Rating  E-RRF Confidence Rating  E-RRF Confidence Rating

Three out  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.
1           Occupation could increase exposure to MWCNT relative to the general population. In

2    occupational settings, the primary exposure pathway for MWCNTs is likely to be inhalation. Consumer

3    exposure pathways might be similar to those identified for decaBDE, namely ingestion of household dust,

4    but MWCNTs are less likely to be released from the polymer matrix (see Section 2.4.2; see Appendix H

5    for detailed information regarding decaBDE). Given the lack of data on consumer exposure to MWCNTs,

6    whether the primary route of exposure for highly exposed populations would be different from that of

7    decaBDE is difficult to determine. Dust levels in the home can vary by socioeconomic status or the type

8    and condition of housing (see Section 5.3.1). In turn, disproportionate levels of exposure can occur in

9    specific populations, including low-income and low-educational-attainment populations. Additionally, for
     March 2013
                  4-28
DRAFT-DO NOT CITE OR QUOTE

-------
 1    pollutants for which inhalation exposure due to proximity to primary pollution sources is of concern,
 2    socioeconomic status has been associated with increased exposures, which also might be true for
 3    MWCNTs. Race and ethnicity do not present specific physiological conditions to increase susceptibility
 4    to exposure, but demographic factors such as socioeconomic and educational status might cause some
 5    populations to experience disproportionate exposures. The possibility of increased exposure to MWCNTs
 6    due to characteristics associated with low socioeconomic status has not yet been explored in the literature.
 7           In general, children are more susceptible to increased inhalation exposures because of increased
 8    ventilation rates per unit of body weight and increased oral exposures due to hand-to-mouth and chewing
 9    (e.g., mouthing furniture or fabric) behaviors. The relevance of the inhalation and oral pathways for
10    MWCNT consumer exposures, however, is unknown. In addition, lack of data on whether MWCNTs, if
11    released from flame-retardant textiles, would partition to dust precludes a determination of whether
12    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.
13           A variety of exposure standards, guidelines, or recommendations are developed by different
14    organizations with purview over specific portions of the population or situations during which exposure
15    might occur (e.g., occupational exposures, general population drinking water exposures). Available
16    information on these types of values for MWCNTs is presented below. Section 5.1.1 discusses how some
17    of these values inform quantitative toxicity assessments.
18           MWCNTs can have features of both nanoparticles and fibers, and regulations exist to control
19    particles and fibers in the workplace. MWCNTs can appear as clumps or ropes, which can be counted as
20    single fibers if they fit the definition of a fiber. If the rope is not within the World Health Organization's
21    definition of a fiber (greater than 5 um in length, with an aspect ratio greater than 3:1), however, it would
22    not be counted as a fiber under the current measurement system. Some MWCNTs could therefore be
23    missed using current fiber classification methods (Donaldson et al. 2006).
24           Recently, NIOSH conducted a risk analysis for CNTs to establish a guideline exposure level for
25    occupational workers (NIOSH. 2010). They estimated a working lifetime inhalation exposure of
26    0.2-2 ug/m3 (8-hour time-weighted average) associated with a 10% excess risk of early-stage adverse

      March 2013                                     4-29                   DRAFT-DO NOT CITE OR QUOTE

-------
 1    lung effects (95% lower confidence limit estimates) based on two subchronic animal inhalation studies
 2    (Pauluhn. 201 Ob: Ma-Hock et al.. 2009) (see Section 5J_ for human health effects). The NIOSH-
 3    recommended exposure limit is 7 ug/m3 for elemental carbon (see Table 4-1) as an 8-hour time-weighted
 4    average respirable mass airborne concentration (NIOSH. 2010). NIOSH also recommends that workplace
 5    airborne exposure to CNTs be measured by NIOSH NMAM 5040, which has an upper limit of
 6    quantitation of 7 ug/m3 (NIOSH. 2010). Specifically, the animal-data-based risk estimates indicate that
 7    workers could have >10% excess risk of developing early-stage pulmonary fibrosis if exposed over a full
 8    working lifetime at the upper limit of quantitation for NIOSH NMAM 5040 (NIOSH. 2010). Other
 9    recommended occupational exposure limits (OELs) and general human health exposure limits for
10    inhalation of MWCNTs and related materials are shown in Table 4-1.
11           As discussed by Schulte et al. (2010). deriving OELs for MWCNTs and other nanomaterials is
12    complicated by the challenges associated with measuring workplace exposures (see Text Box 4-1).
13    coupled with the variation in configurations of physicochemical properties that can influence exposure
14    and toxicity (see Text Box 4-2). The heterogeneity in MWCNT configurations could necessitate
15    developing OELs specific to individual formulations of MWCNTs (Schulte et al.. 2010).  Alternatively,
16    OELs could be developed for groups of nanomaterials based on composition or toxic mechanism,
17    although such classifications could exclude physicochemical characteristics that influence biological
18    activity (Schulte etal. 2010).
      March 2013                                    4-30                  DRAFT-DO NOT CITE OR QUOTE

-------
Table 4-1.    Established inhalation reference values and recommendations applicable to MWCNTs.
OEL
The Occupational Safety & Health Administration - permissible exposure limit
(PEL) for respirable fraction of synthetic graphite
A review of CNT toxicity - human inhalation no-effect levels for INELacute
woikeis deiived fiom acute and subchionic inhalation studies with
MWCNTs |NELchronic
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
NIOSH - recommended exposure limit for elemental carbon as an 8-hour TWA
respirable mass airborne concentration
Value
5,000 jig/m3
150 (ig/m3
1 (ig/m3 and
2 jig/m3
50 jig/m3
7 (ig/m3
Reference
Lam etal. (2006)
Aschberger et al. (2010)

Pauluhnetal. (2010a)
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/m3
                                  Nanocyl (2009)
Japanese New Energy and Industrial Technology Development Organization -
Interim OEL for MWCNTs based on unpublished data by unpublished study
                                  Kobayashietal.
                    3.0 jig/kg-day   (unpublished) as cited in
                                  NIOSH (2010)
British Standards Institute - benchmark exposure limit based on one-tenth of the
Institute's asbestos exposure limit
                   01 fiber/cm3 air  BSI (2QQ7) as cited in
                   u.l Tiber/cm  air  N|QSH
TWA = time-weighted average
March 2013
4-31
DRAFT-DO NOT CITE OR QUOTE

-------
     4.2.6. Toxicokinetics,  Dose, and Body Burden
                                       Priority Research Area: Dose (Kinetics)
         gl
       V 
                   E -r
                   IS
Least Important
Possibly Important
Important

Not Confident
Somewhat Confident
Confident
HI
  o
  IJ
                                                          Twelve out 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  absorption,
                                                          distribution, metabolism, and excretion 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 for a detailed explanation of the prioritization
process.
              M     L
    E-RRF Confidence Rating
                                   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.
1            Toxicokinetics can be used to relate exposure and contact, such as those described in the

2    scenarios above, with uptake and dose.  Specifically, toxicokinetics describes how a material is absorbed,

3    distributed, metabolized, and excreted in an organism. An understanding of the relationship between each

4    of these concepts, which are often referred to as ADME, leads to an understanding of the concentration,

5    or dose, of material that can reach—and potentially accumulate in—different tissues of the body.
     March 2013
                                      4-32
                                             DRAFT-DO NOT CITE OR QUOTE

-------
      4.2.6.1. Absorption, Distribution, Metabolism, Excretion
 1           This section contains information regarding the toxicokinetic behavior of MWCNTs when
 2    administered to mammals. Information regarding birds and fish is not presented in this section because,
 3    when extrapolating toxicokinetic data to humans, studies conducted with rodents (rat or mouse) or
 4    nonrodent mammals (dog or monkey) are generally used. Additionally, the toxicokinetic behavior in
 5    response to MWCNTs might differ among birds, fish, and mammals.  See Section 4.3 and Appendix G.4.2
 6    for toxicokinetic information relevant to ecological exposures. Studies examining the toxicokinetics of
 7    MWCNTs in mammals are summarized in Appendix F.
 8           As discussed in Section 4.2.2. humans might be exposed to free MWCNTs, bundled MWCNTs,
 9    and MWCNTs in combination with a polymer matrix, textile fibers or scraps, or other product
10    constituents. The bioavailability (and therefore dose) of MWCNTs is  expected to differ for MWCNTs in
11    different forms or bundling states (see Text Box 4-2). In general, CNTs, including MWCNTs, appear to
12    be biopersistent. After intratracheal administration, MWCNTs have been observed to deposit and persist
13    within the lung for up to several months (Elgrabli et al.. 2008b: Deng et al.. 2007). Macrophage-mediated
14    clearance of MWCNTs after exposure via inhalation (Elgrabli  et al.. 2008a) and translocation of some
15    types of CNTs into the pleura and subpleura (Porter et al.. 2010;  Ryman-Rasmussen et al.. 2009a) have
16    been demonstrated. After oral exposure, most MWCNTs (administered at 10 ug/mouse by gavage) were
17    evident within the feces and also remained within the stomach and small and large intestines, with no
18    detectable transport into the blood or obvious metabolism through 28  days (Deng et al.. 2007). Because
19    only one study was identified that evaluated distribution after oral exposure, whether distribution is
20    possible to other organs in the body following inhalation, dermal, and oral exposures to MWCNTs is not
21    well understood.
22           Distribution of CNTs to various organs has been reported following intravenous exposure (Deng
23    et al.. 2007; Cherukuri et al.. 2006). with predominant localization within the liver, lungs, and spleen.
24    This pathway, however, is not likely relevant for the exposures of concern in this evaluation (i.e.,
25    MWCNTs used in flame-retardant coatings on upholstery textiles are  unlikely to be intravenously
26    applied).

      4.2.6.2. Internal Dose and  Body Burden
27           Based on toxicokinetic studies with rats, inhaled MWCNTs can remain in the lung following
28    exposure for an extended period, up to six months (Aschberger et al..  2010). These studies reported
29    qualitative data, however, and no studies were found that reported levels of MWCNTs in the lung.  One
30    study did report MWCNTs in the subpleura of mice following  a single inhalation exposure to 30 mg/m3;

      March 2013                                    4-33                   DRAFT-DO NOT CITE OR QUOTE

-------
 1    no MWCNTs were detected in the subpleura following instillations of a lower concentration (1 mg/m3)
 2    (Ryman-Rasmussen et al.. 2009a). Another study reported MWCNTs with known length just under 4 um
 3    in the pleura of mice following a single aspiration of 10-80 ug (Porter et al.. 2010). This finding is
 4    notable because the pathogenic mechanism of asbestos fibers in the mesothelioma disease process occurs
 5    in the pleural cavity (Aschberger et al.. 2010).
 6           No detectable amounts of MWCNTs were observed in the blood following  oral exposure in mice,
 7    but the MWCNTs did remain in the stomach and small and large intestines (Deng et al.. 2007): the
 8    observed MWCNTs remained unchanged (Aschberger et al.. 2010). As discussed in Section 4.2.6.1.
 9    MWCNTs were distributed to multiple organs following intravenous injection (Deng et al.. 2007;
10    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. Absorption, Distribution, Metabolism and Excretion in
      Ecological  Receptors
                                   Neutral Research Area: Dose (Kinetics)
       General factors influencing absorption, distribution, metabolism and excretion (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.
      March 2013                                     4-34                   DRAFT-DO NOT CITE OR QUOTE

-------
4.3.3.  Exposure  Pathways  in Aquatic Systems
                             Un prioritized Research Area: Exposure Route
        Aquatic
         Biota
   01
LJ r
   o
   CL
4/13
                5/13
                4/13
                                                             Inhalation
                                                                   Direct Contact
                                                                             o>
                                          g|s
                                          m 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
                    ~R    M     C~~         H    M    L           H    M    L
                E-RRF Confidence Rating   E-RRF Confidence Rating   E-RRF Confidence Rating

              Four out 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.
Toxicokinetics and Body Burden in Aquatic Systems
                              Unprioritized Research Area: Dose (Kinetics)
        Aquatic
          Biota
  c*
  o> a>
  E £
  » i
 in r
    o
    Q.
                 5/13
 6/13
                                                              Distribution
                 2/13
    H     M    L
E-RRF Confidence Rating
  I     Excretion    I
                                             H    M    L           H     M    L
                                         E-RRF Confidence Rating  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
                               en
                o
                      H    M     L
                 E-RRF Confidence Rating
                         Six  out  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.
March 2013
                                  4-35
                                        DRAFT-DO NOT CITE OR QUOTE

-------
4.3.4.  Exposure Pathways in Terrestrial Systems
                             Unprioritized Research Area: Exposure Route
   01
 a, o
 E £
 » ro
 LJ t
   O
   Q.
   E
1/13
7/13
                5/13
                                                                                   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
                    ~H	M     C~~         H     M    L            H    M     L
                E-RRF Confidence Rating   E-RRF Confidence Rating  E-RRF Confidence Rating

              One out 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.
Toxicokinetics and Body Burden in Terrestrial Systems
                              Unprioritized Research Area: Dose (Kinetics)
    o
    a.
    I
                 2/13
                 6/13
                                                                                     Metabolism
                      H    M    L           H     M    L          ~~HM    IT
                 E-RRF Confidence Rating   E-RRF Confidence Rating  E-RRF Confidence Rating
                 5/13
                        Excretion
  • - 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
               •- I
               13
                o
                I-1
                      H    M    L
                 E-RRF Confidence Rating
Two  out  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.
March 2013
                                  4-36
               DRAFT-DO NOT CITE OR QUOTE

-------
      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.
 1           Assessing aggregate exposures involves characterizing exposures to a single chemical across
 2    multiple exposure routes. Due to the range of applications for which MWCNTs can be used, release from
 3    multiple products and subsequent exposure via multiple routes is anticipated.
 4           As described in Text Box 4-2. the properties of MWCNTs are easily altered through manipulation
 5    of material characteristics such as size, bundling affinity,  and surface treatments, and the degree to which
 6    these specific changes affect the overall exposure profile  is unclear. Moreover, the composition of
 7    MWCNT formulations can vary with differences in synthesis techniques and remaining impurities from
 8    manufacturing stock. If small changes in MWCNT characteristics result in measurable changes in the
 9    nature and extent of exposure, each MWCNT formulation might be considered a unique substance;
10    therefore, MWCNTs produced by different manufacturers using different techniques might introduce
11    discrete sets of aggregate exposures. No consensus has been reached on which physicochemical
12    characteristics drive changes in exposure potential or what magnitude of change to any specific
13    characteristic or property is necessary to elicit a measurable change in exposure.
14           In addition to different material designs, MWCNTs can be used in a wide range of possible
15    applications, including in coatings, electronics, adhesives, polymer composites, thermoplastics, and
16    others. MWCNT applications can then be used in textiles, aerospace, construction, sporting goods,
17    medical applications, and many other types of products (Aschberger et al.. 2010). Thus, the potential for
18    exposure to MWCNTs exists where humans interact with any of these products as producers or
19    consumers or when CNTs are released to environmental media.
20           Exposure to MWCNTs is likely to occur through inhalation of MWCNT bundles and MWCNTs
21    sorbed to dust produced during the manufacture  and processing of MWCNTs and composites containing
22    MWCNTs. A secondary pathway is through dermal exposure, which could occur in occupational settings
23    from dust settling on work surfaces. For consumers, exposure could occur from the abrasion or wear of
24    products containing MWCNTs. The general public could be exposed to MWCNTs via drinking water,
25    contact with contaminated soil, ingestion and inhalation of household dust, dermal contact with surfaces
26    upon which MWCNTs and dust have settled, and other pathways as a result of their release from product
27    matrices. MWCNT releases from composite materials, such as those used in sporting goods, plastics,

      March 2013                                     4-37                   DRAFT-DO NOT CITE OR QUOTE

-------
 1    touchscreens, and batteries are expected to be minimal, if not negligible, during consumer use because
 2    MWCNTs used for these products are bound in relatively strong matrices. End-of-life product
 3    dismantling, land-filling, and incineration, however, might offer greater potential for release of
 4    constituent materials to environmental compartments because many of the processes involved in end-of-
 5    life practices are intended to break down the strong matrices in which the MWCNTs are embedded
 6    (Aschberger et al.. 2010). What the implications of these releases will be on exposures and impacts to
 7    human health, ecological receptors, and other receptors is not yet known.
 8    The anticipated market trend for production of MWCNTs is strong growth in the near future, especially as
 9    production costs drop and a wider variety of applications is discovered (Lametal.. 2006). Strong market
10    growth and diverse applications could lead to a greater diversity and number of exposure scenarios, thus
11    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.
12           As stated in The Exposure Factors Handbook (U.S. EPA. 201 Ib). "Cumulative exposure is
13    defined as the exposure to multiple agents or stressors via multiple routes." For the purpose of this case
14    study, the "multiple agents or stressors" considered to contribute to cumulative exposure include those
15    substances that are produced or released as a result of the product life cycle of MWCNT flame-retardant
16    upholstery textile coatings, facilitate uptake of MWCNTs into humans and biota, are taken up as a result
17    of MWCNT exposures, or induce effects in humans or biota through a comparable or synergistic mode of
18    action. As mentioned in Section 4.4. different characteristics of different MWCNT formulations could
19    result in the necessity to consider different formulations as unique stressors, in which case each
20    formulation might represent a contribution to cumulative exposures.
21           Depending on which feedstocks are  used in the manufacturing process, by-products might differ;
22    therefore, coexposures to MWCNTs and other compounds might differ. Although the generation of
23    impurities is likely during the manufacturing process, MWCNTs are typically purified after synthesis
24    with varying degrees of success (see Appendix C). As described in Section 2.2.2.2. Plata et al. (2009)
25    observed production of 45 side-products of CVD synthesis of MWCNTs, including polycyclic aromatic
      March 2013                                     4-38                   DRAFT-DO NOT CITE OR QUOTE

-------
 1    hydrocarbons, methane, and volatile organic carbons. Single-walled carbon nanotube production can
 2    result in by-products of sodium hydroxide, ethanol, water, filtrate, and scrap membrane (Healy et al.,
 3    2008). but whether these by-products also will be generated by MWCNT synthesis is unclear.
 4           Functionalization, which involves covalent attachment of submolecular components to the
 5    MWCNTs, is required before MWCNTs can be dispersed into polymers or organic solvents.
 6    Functionalization can involve several different reagents, depending on the  process used, as listed in Table
 7    2-3. Any of the MWCNT-containing textiles could include small amounts  of the reagents. No data were
 8    found, however, on the by-products or impurities in textiles treated with MWCNT flame-retardant
 9    coatings.
10           CNTs released to the environment might bind or sequester pollutants in a form that is not
11    bioavailable, thus reducing the impact of other toxic substances. For example, when MWCNTs are added
12    to sewage  sludge, seed germination and root growth increased, which could be because the MWCNTs
13    bound pollutants (e.g., heavy metals, organic compounds) present in the sludge (Oleszczuk et al.. 2011).
14    On the other hand, MWCNTs might facilitate transport of these pollutants  through environmental
15    compartments, across biological boundaries, and into cells, where they could react with cell machinery
16    (Johnston etal.. 2010).
17           Increasing production and market growth for products containing CNTs likely will lead to
18    increasing levels of CNTs, by-products, and related compounds in the environment, as well as an increase
19    in exposures. Due to the heterogeneous nature of MWCNTs, the various manufacturing processes used,
20    exposure to a wide variety of CNTs and by-products from many different sources is possible.  In addition,
21    the environmental persistence of CNTs could lead to long-term exposures or consecutive exposures in
22    multiple receptors.
      March 2013                                     4-39                  DRAFT-DO NOT CITE OR QUOTE

-------
                               This page intentionally left blank.
March 2013                                                         DRAFT-DO NOT CITE OR QUOTE

-------
              Chapter 5.  Potential Human  Health,
                   Ecological, and Other  Impacts
 1          The final step of compiling information into the comprehensive environmental assessment (CEA)
 2    framework is to link the information described in the previous chapters on the product life cycle;
 3    transport, transformation, and fate; and exposure-dose with potential impacts to receptors. The CEA
 4    framework includes information relevant to impacts on human health and ecological receptors, similar to
 5    what might be investigated in traditional risk assessment processes, as well as other plausible impacts that
 6    might be considered in life-cycle-focused assessments (e.g., socioeconomics, climate change, resource
 7    depletion).
 8          Section 5.1 discusses potential impacts of exposure to multiwalled carbon nanotubes (MWCNTs)
 9    and related contaminants on human health. This section relies heavily on evidence from experimental
10    studies with laboratory animals, the results of which could be extrapolated to humans using methods
11    established for quantitative toxicity assessment. As discussed in Chapter 4. humans could be exposed to
12    MWCNTs from flame-retardant upholstery textiles through a variety of pathways, with the contaminants
13    reaching receptors through dermal deposition, oral ingestion, or inhalation. This section discusses
14    potential health impacts observed in studies with laboratory animals exposed to MWCNTs by these
15    exposure routes; data are grouped to illustrate the types of impacts (e.g., pulmonary toxicity, skin
16    irritation, reproductive effects) and sub-grouped by exposure routes for each impact.
17          Section 5.2 discusses the potential impacts of environmental media contaminated with MWCNTs
18    on ecological health, which encompasses impacts at the organism, population, and ecosystem levels. This
19    section is approached from an ecosystem perspective (aquatic vs. terrestrial), and data on groups of
20    organisms within those ecosystems are summarized. The focus of the discussion of impacts  on ecological
21    health is on identifying and comparing data on exposure levels that might cause significant mortality,
22    delayed growth or development, reproductive defects, or other impacts that could alter community
23    structure and potentially cause ecosystem collapse.
24          Finally, Section 5.3 discusses other plausible impacts resulting from the product life cycles of
25    MWCNTs in flame-retardant upholstery textiles. The section includes a consideration of the energy input
26    requirements for synthesis of the MWCNTs, the economic impacts related to the cost of material
27    production, and the potential for disproportionate impacts on populations with lower socioeconomic
28    status.

      March 2013                                   5-1                  DRAFT-DO NOT CITE OR QUOTE

-------
1            As noted throughout this document, MWCNTs are not a single material, but rather a mixture of

2    materials with different physicochemical properties. For the purpose of this case study, however,

3    MWCNTs are generally regarded as a single class of materials. Text Box 5-1 provides introductory-level

4    detail on how changes in physicochemical properties might influence toxicity. Throughout this chapter,

5    where physicochemical properties can be  related to effects on particular outcomes (e.g., fiber length on

6    inhalation endpoints), these properties are described and their potential influences on effects are

7    discussed. Additionally, Appendix F presents detailed toxicokinetic and toxicological study summaries in

8    which the MWCNT characteristics and components of test designs that influence toxicological outcomes

9    are provided.
     5.1. Human Health  Effects
                                        Priority Research Area: Impacts
     sl
        o
        Q.
              :*:
                     13/13
                    0/13
                    0/13
                                            ?!=
                                            Ml CO
                                              O
                                                                                      Reproductive/
                                                                                      Developmental
      • - 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~              ML           H    M     L
  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, non-cancer, 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, non-cancer 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.
     March 2013
                    5-2
DRAFT-DO NOT CITE OR QUOTE

-------
            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, ground MWCNTs, which 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 might also play a role in toxicity; studies have shown that thin MWCNTs
(diameters of 50 nm or less) are less toxic than thick MWCNTs (diameters of 70 nm or more) (Fenoglio etal., 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: Pacurari etal., 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 an MWCNT's
life cycle 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).
March 2013
5-3
                                                                                DRAFT-DO NOT CITE OR QUOTE

-------
 1           This section discusses the potential human health effects resulting from exposures to MWCNTs.
 2    As noted in Chapter 4. exposure to this material from aggregate sources is likely; no studies were found
 3    that investigated impacts to human health that could be attributed specifically to exposure to MWCNTs
 4    released during the life cycles of MWCNT flame-retardant upholstery textile coatings. Primary literature
 5    on MWCNTs was identified for most endpoints discussed in the following sections. When primary
 6    literature was not identified on MWCNTs, literature on single-walled carbon nanotubes (SWCNTs) was
 7    considered.
 8           Toxicology studies conducted on animals provide much of the information discussed in this
 9    chapter because data from studies on humans are unavailable. Effects observed in animal studies are
10    typically extrapolated to humans when conducting quantitative toxicity assessments (e.g., when
11    calculating a reference dose [RfD] or reference concentration  [RfC]; see Section 4.2.5). Potential health
12    effects associated with all routes of exposure (dermal, inhalation, and oral) are presented in this section
13    because each is plausible for humans (see Chapter 4 for additional exposure scenario information). Table
14    5-1 provides an overview of the findings for human health effects of decaBDE and MWCNTs and is
15    followed by a detailed discussion of the available data for MWCNTs. Detailed discussion of the  data for
16    decaBDE can be found in Appendix H.
      March 2013                                     5-4                   DRAFT-DO NOT CITE OR QUOTE

-------
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 toxicity
Subchronic toxicity
Chronic toxicity
X(weaki) X
X X
X (weak2) X
X X
X X
X X
Inhalation
Local effects
Other effects
Inflammation
Respiratory sensitization
Acute toxicity
Subchronic toxicity
Chronic toxicity
Immunotoxicity
X3 X
X3 X
X X
X X
X X
X4 X
March 2013
5-5
DRAFT-DO NOT CITE OR QUOTE

-------
Table 5-1, cont.: Summary of effects observed after dermal, oral, and inhalation exposure to
            decaBDE and MWCNTs.
Route of n. . „ .
Observed effect
exposure
MWCNTs
Yes No No data
DecaBDE
Yes No

No data
Oral
 Effects            Acute toxicity
                   Subchronic toxicity                                       X          X5

                   Chronic toxicity                                           X          X5

                   Reproductive/developmental                          X (weak)7      X6

 1Exposure 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) when animals
 were exposed to 1 % Nikkiso-MWCNTs in an OECD 406-compliant study (Emaetal., 2011).
 2Conjunctival redness and blood vessel hyperemia at 1 hour, but not at 24 hours (Emaetal.. 2011).
 Inflammation was found in three OECD-compliant studies (Pauluhn, 2010b: Ellinger-Ziegelbauer and Pauluhn, 2009: Ma-Hock etal., 2009):
 respiratory sensitization was found in Park  et al. (2009).
 4lmmunosuppressive results were found in  Mitchell et al. (2009:2007). and  Nygaard et al. (2009).
 5Effects included changes to thyroid and liver observed in subchronic and chronic oral studies (NTP, 1986: Morris etal., 1975: Morris etal.,
 1973):
 6Effecte 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 etal..2008: Viberq et al.. 2008: Viberq etal.. 2007: Viberq
 etal.. 2003).
 7Effects reported to prenatal exposure (via  intraperitoneal injection) included external and skeletal malformations (Fujitani etal.,2012): an
 oral study did not report any developmental effects (Limetal.,  2011b).
March 2013                                               5-6                        DRAFT-DO NOT CITE OR QUOTE

-------
  1

  2

  3

  4

  5

  6
10
11
12
13
14
15
16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34
        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:

        •    Numerous in vitro and in vivo studies have shown that carbon nanotubes (CNTs) (both
            SWCNTs and MWCNTs) might induce prominent pulmonary inflammation (Pauluhn. 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 a 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 etal.. 1999; Donaldson et al.. 1996; Kennedy et
            al.. 1989).
Although no subchronic inhalation studies were

identified for systemic toxicity (i.e., toxic effects

resulting from absorption and distribution of a

toxicant at a site distant from its entry point)

(see Section 5.1.2.2), acute 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 has

not been investigated. However, several studies

using methods such as instillation and

intraperitoneal injection (Sakamoto  et al.. 2009;

Poland et al.. 2008: Takagi et al.. 2008) 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 of

human health concern.
    Additional Information Highlight Box 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 et al.,  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 pleura
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
(Murphy et al., 2012: Murphy et al.,  2011). Nevertheless, pre-
existing inflammation  might be needed for the generation of
fibrosis (Ryman-Rasmussen et al., 2009b). 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. The results
of   mode-of-action investigations  for   MWCNTs  provide
qualitative data to better determine the appropriateness of the
comparison to asbestos; however no quantitative data were
identified.
       March 2013
                                                 5-7
                      DRAFT-DO NOT CITE OR QUOTE

-------
      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.
 1            In a quantitative toxicity assessment, appropriate toxicity information is collected and evaluated.
 2    These data are used to derive reference values, such as an RfD for oral exposure or RfC for inhalation
 3    exposure. Similar to an RfD (as defined in Section 4.2.5). an RfC is an estimate of a continuous inhalation
 4    exposure for a given duration to the human population (including susceptible subgroups) that is likely to
 5    be without an appreciable risk of adverse health effects over a lifetime. Both values, an RfC and an RfD,
 6    are derived from a benchmark dose lower confidence limit, no-observed-adverse-effect level (NOAEL), a
 7    lowest-observed-adverse-effect level (LOAEL), or another suitable point of departure, with
 8    uncertainty/variability factors applied to reflect limitations of the data used. Other types of reference
 9    values also can be derived for use with other exposure durations (e.g., acute or subchronic), more specific
10    populations (e.g., healthy workers), or specific exposure contexts (e.g., emergency response or
11    occupational exposure; see Section 4.2.5).
12            EPA has not evaluated MWCNTs to derive an RfD  or an RfC. The only health effect based
13    reference value for MWCNTs derived by a government agency is the draft recommended exposure limit
14    (REL) proposed by the National Institute for Occupational Safety and Health (NIOSH) for CNTs
15    (NIOSH. 2010) as discussed in Section 4.2.5. Acute and subchronic human no-effect levels for inhalation
16    exposures to the general public also have been proposed in the open literature; derivation of these values
17    is briefly discussed in the sections that follow.

            5.1.1.1. Health Reference Values
18            To date, EPA has not evaluated MWCNTs to establish an RfD or RfC. As discussed in Section
19    4.2.5. NIOSH (2010) conducted a risk analysis for CNTs and established a draft REL of 7 ug/m3 (the high
20    estimate of the limit of quantification for NIOSH Method 5040) for carbon. The REL is based on a
21    working lifetime inhalation exposure of 0.2-2  ug/m3 (8-hour time-weighted average) associated with a
22    10% excess risk of early-stage adverse lung effects (95% lower confidence limit estimates) and was
23    derived  using two subchronic (90-day) animal  inhalation studies (Pauluhn. 201 Ob: Ma-Hock et al.. 2009).
      March 2013                                      5-8                    DRAFT-DO NOT CITE OR QUOTE

-------
 1           Although not derived by a government agency, a subchronic human occupational no-effect level
 2    for MWCNT exposure to the general public of 0.25 ug/m3 has been estimated by Aschberger et al. (2010)
 3    based on a LOAEL (Ma-Hock et al.. 2009) and a NOAEL (Pauluhn. 201 Ob) of 1.0 ug/ m3 from the same
 4    subchronic studies NIOSH (2010) used in their derivation of an REL. Aschberger et al. (2010) also
 5    calculated a human no-effect level for acute exposure to MWCNTs of 150 ug/m3 from a LOAEL of
 6    11 mg/m3 based on the absence of inflammatory effects in a rat study (Ellinger-Ziegelbauer and Pauluhn.
 7    2009). Although both values were calculated based on guidance provided by Registration, Evaluation,
 8    Authorisation and Restrictions of Chemicals (REACH) for chemical safety assessment (ECHA. 2008).
 9    the relatively limited and often conflicting database of currently available toxicological values for
10    MWCNTs suggests that these values are preliminary estimates associated with a high degree of
11    uncertainty.
12           The two subchronic animal inhalation studies (Pauluhn. 201 Ob: Ma-Hock et al.. 2009) used to
13    derive the draft NIOSH (2010) REL and the subchronic human no-effect level proposed by Aschberger et
14    al. (2010) are described in detail in Section 5.1.3.
      5.1.2. Systemic Toxicity
                                  DecaBDE Can Inform MWCNT Assessment
       For all routes of exposure, decaBDE exhibits low acute toxicity. Subchronic occupational exposure to decaBDE in humans,
       however, has led to 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 non-cancer 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 might vary across
       multiple exposure routes and durations of exposure. See Appendix H for more detailed information on the non-cancer effects
       of decaBDE.
             5.1.2.1. Acute
15           No data were identified on the acute systemic toxicity of MWCNTs following inhalation
16    exposure (see Section 5.1.8 for Immune System Effects). Acute inhalation studies identified for
17    MWCNTs, including key studies, are presented in Appendix F.1.2.
18           Limited data were identified on the acute toxicity of MWCNTs following oral or dermal exposure
19    (see Section 5.1.5 for discussion of skin irritation). Based on the OECD TC 423, MWCNTs tested by
20    Pauluhn (201 Ob) were not acutely toxic (oral LD50 > 5,000 mg/kg body weight). An acute dermal
21    exposure study was not identified for MWCNTs. One acute oral toxicity study on SWCNTs was
22    identified in which single doses of 1,000 mg/kg body weight of three different types of SWCNTs

      March 2013                                      5-9                    DRAFT-DO NOT CITE OR QUOTE

-------
 1    (raw:  1 nm x 1-2 um, 25% Fe; purified: 1 nm x 1-2 um, <4% Fe; ultrashort: 1 nm x 20-80 nm,
 2    <1.5% Fe) were administered to mice (Kolosnjaj-Tabi et al.. 2010). No signs of toxicity (e.g., reduced
 3    survival, delayed growth, behavioral abnormalities, clinical chemistry changes) were observed for any
 4    typeofSWCNT.
 5            One acute intraperitoneal toxicity study was identified for MWCNTs. Mice injected with 0.25,
 6    0.5, or 0.75 mg/kg-day of purified, carboxylated MWCNTs for 5 days experienced decreased body weight
 7    gain, and increased markers of oxidative stress and hepatotoxicity (increased reactive oxygen species in
 8    liver, enhanced activity of liver enzymes such as serum aminotransferases and alkaline phosphatases)
 9    (Tatlollaetal..20in (see Table F-8 in Appendix F).

            5.1.2.2. Subchronic
10            No data were identified on the subchronic systemic toxicity of MWCNTs following exposure by
11    any route. See Section 5.1.3 for discussion of pulmonary effects and Section 5.1.8 for discussion of
12    immune system effects following subchronic exposure.

            5.1.2.3. Chronic
13            No data were identified on the chronic systemic toxicity of MWCNTs following exposure by any
14    route.
      5.1.3. Pulmonary Toxicity
15           Material characteristics and study details associated with the acute and subchronic inhalation
16    studies identified for MWCNTs are presented in Appendix F.1.2. Table F-6. No chronic inhalation studies
17    were identified for MWCNTs. Toxicological responses generally have been consistent across studies
18    administering MWCNTs into the lungs (by intratracheal instillation, aspiration, or inhalation), with
19    exposed animals exhibiting pulmonary inflammation and fibrosis. Results reported for animals exposed to
20    MWCNTs via intratracheal instillation (Park et al.. 2009; Muller et al.. 2008a: Mulleretal.. 2008b:
21    Muller et al.. 2005). however, generally resulted in more severe effects than those observed for inhalation
22    (Pauluhn. 201 Ob: Ellinger-Ziegelbauer and Pauluhn. 2009: Li et al.. 2009: Li et al.. 2007: Mitchell et al..
23    2007).
      March 2013                                    5-10                  DRAFT-DO NOT CITE OR QUOTE

-------
                                     Additional Information Highlight Box 14:
                        Applying traditional in vivo inhalation study design models to MWCNTs
       Available evidence suggests that MWCNT inhalation exposure might occur and cause subsequent 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 (Geys et al., 2007) (Liu etal., 2012: Coccini et al., 2010: Alfaro-Moreno et
       al., 2008: 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 etal., 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.
 1            Pulmonary inflammation was observed in one acute (Ellinger-Ziegelbauer and Pauluhn. 2009)
 2    and two subchronic (Pauluhn. 201 Ob; Ma-Hock et al., 2009) Organisation for Economic Co-operation and
 3    Development (OECD)-compliant animal inhalation studies, with granulomas observed in both subchronic
 4    studies. In the study conducted by Ma-Hock et al. (2009), an exposure-related increase in the incidence of
 5    granulomatous inflammation in the lung and lung-associated lymph nodes was observed in exposed rats
 6    (head-nose exposure). At the mid and high exposures (0.5 and 2.5 mg/m3, respectively), increased lung
 7    weights, pronounced multifocal granulomatous inflammation, diffuse histiocytic and neutrophilic
 8    inflammation, and intra-alveolar lipoproteinosis were observed in lung and lung-associated lymph nodes.
 9    Similar effects were observed in a study by Pauluhn (201 Ob), in which Wistar rats were exposed (nose
10    only) to MWCNTs (0.1, 0.4, 1.5, 6 mg/m3); inflammatory changes in the distal nasal cavities were
11    observed at all but the lowest concentration. Additionally, exposure-related lesions of the upper
12    respiratory tract (e.g., goblet cell hyperplasia or metaplasia, eosinophilic globules, and focal turbinate
13    remodeling) and lower respiratory tract (e.g., inflammatory changes in the bronchioloalveolar region and
14    increased interstitial collagen staining) were observed at higher concentrations. According to a study
15    conducted by Kim et al. (2012). pulmonary DNA damage is initiated after only a few days of exposure; a
16    significant increase in DNA damage was measured by Comet assay in lung cells from rats immediately
17    following 5 days of exposure (6 hours per day) to 0.94 mg/m3 MWCNTs. DNA damage remained
18    detectable one month after the last exposure.
      March 2013                                        5-11                     DRAFT-DO NOT CITE OR QUOTE

-------
      5.1.4. Eye Irritation
 1           All in vivo eye irritation studies considered are presented in Appendix F.I.2 (Table F-4). In a key
 2    ocular irritation study conducted with rabbits, one of two types of MWCNTs administered via instillation
 3    to the conjunctival sac resulted in conjunctival redness and blood vessel hyperemia at 1 hour, but not at
 4    24 hours (Emaetal.. 2011). Differences in purity, diameter, and surface area (see Appendix F. Table F-4)
 5    between the two types of MWCNTs tested could explain why effects were observed with only one of the
 6    two types.


      5.1.5. Skin Irritation
 7           Material characteristics and study details associated with the in vivo dermal studies considered
 8    for MWCNTs are presented in Appendix F.1.2.
 9           Based on the information available, even the most irritating of several MWCNTs tested appears
10    to be only a very weak skin irritant when tested on healthy, intact skin (Emaet al., 2011; Kishore et al.,
11    2009). The available studies were performed in accordance with accepted standard practices for
12    conventional  skin irritation tests, which use healthy, intact skin exposed for 4 hours under semioccluded
13    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. The data on decaBDE raise important questions that should also be considered for research
       planning to inform future MWCNT risk assessment, for example: Do reproductive and developmental effects of MWCNTs differ
       when administered during different developmental periods? See Appendix H for more details regarding the reproductive and
       developmental effects of decaBDE.
14           No in vivo reproductive studies using highly relevant exposure routes were initially identified for
15    MWCNTs. See Section 6.3.4.2 for information on two studies (Lim et al., 201 la; Lim et al., 20 lib; Bai et
16    al.. 2010) that were identified after the collective judgment prioritization process.
      March 2013                                      5-12                   DRAFT-DO NOT CITE OR QUOTE

-------
      5.1.7. Developmental Effects
 1           One developmental study was identified in which MWCNTs were administered via gavage to
 2    pregnant Sprague-Dawley rat dams at doses of 0, 40, 200, and 1,000 mg/kg-day on gestation days (GD)
 3    GD6-GD19; dams were sacrificed on GD20 (Lim et al.. 20lib). Minimal maternal toxicity was observed
 4    at 1,000 mg/kg-day; a dose-dependent decrease in thymus weight was observed, but no effects were
 5    observed on maternal body weight, food consumption, and oxidant-antioxidant balance in the liver.
 6    No differences in gestation index, fetal death, fetal and placental weights, or sex ratio were observed as a
 7    result of MWCNT exposure. Therefore, the embryo-fetal NOAEL was 1,000 mg/kg-day.
 8           When MWCNTs were administered via intraperitoneal or intratracheal injection to pregnant ICR
 9    mice in a single dose of 0, 2, 3, 4, or 5 mg/kg on GD9, teratogenic effects were observed (Tujitani et al..
10    2012). Fetal examinations performed on GDIS showed external and skeletal malformations such as short
11    or absent tails, cleft palate, limb reduction deformities, fused ribs and vertebral bodies, and
12    hypo/hyperphalangia (see Table F-8 in Appendix F).
      5.1.8. Immune System  Effects
13           Inhalation studies examining the immune system after exposure to MWCNTs are summarized in
14    Appendix F.I.2. In one notable study, systemic immunosuppression (characterized by T-cell dependent
15    antibody response to sheep erythrocytes and T-cell proliferative ability in presence of mitogen), was
16    observed in mice exposed to 0.3, 1, or 5 mg/m3 MWCNTs via inhalation for 14 days, however no
17    significant lung inflammation or lung tissue damage was observed (Mitchell et al.. 2007).16 The
18    immunosuppressive mechanism could involve a signal originating in the lungs that activates
19    cyclooxygenase enzymes in the spleen (Mitchell et al.. 2009). In addition, an acute inflammatory response
20    was observed in female BALB/cAnCrl mice following administration of MWCNTs via single intranasal
21    injection at 200 or 400 ug/mouse (Nygaard et al.. 2009).
22           MWCNTs have been identified as respiratory sensitizers in several studies. Park et al. (2009)
23    observed a potential allergic response in mice following intratracheal instillation of 50 mg/kg MWCNTs.
24    The authors noted that the significantly increased immunoglobulin E concentrations coupled with pro-
      16This study was questioned after publication primarily because an image in the study indicated that nanofibers, not
      nanotubes, were used. According to Lison and Muller (2008), although Mitchell et al. (2007) might be correct in
      their assertion that the results of these early installation experiments show immune effects in the lungs, these studies
      are "probably of little relevance to assess the hazard of MWCNT because they could not be reproduced upon
      inhalation exposure." They maintain that conclusions about immunological effects cannot be reached by Mitchell et
      al. (2007).
      March 2013                                     5-13                  DRAFT-DO NOT CITE OR QUOTE

-------
 1    inflammatory responses likely resulted from B-cell activation by IL-10. Similarly, Inoue et al. (2009)
 2    demonstrated a potential allergic response in mice following six weekly intratracheal instillations of 50 ug
 3    MWCNTs. These authors reported a significant increase in the number of total immune cells (including
 4    macrophages, neutrophils, eosinophils, and lymphocytes) in the bronchiolar lavage fluid, concurrent with
 5    infiltration of eosinophils, neutrophils, and mononuclear cells in the lung. They also noted a significant
 6    induction of goblet cell hyperplasia in the bronchial epithelial tissue, indicating that MWCNTs exacerbate
 7    metaplasia in the presence or absence of other allergens. Of note, the Organisation for Economic
 8    Co-operation and Development currently has no guidelines for standardized respiratory sensitization
 9    studies, and the studies identified above used intratracheal instillation, not inhalation, as the route of
10    exposure.
11            One study that used inhalation as the route of exposure was identified. Ryman-Rasmussen et al.
12    (2009b) exposed mice with allergic asthma to 100 mg/m3 MWCNT aerosol for 6 hours (approximately 10
13    mg/kg MWCNT total) and found that airway fibrosis occurred differentially in ovalbumin-sensitized mice
14    versus nonsensitized mice. The authors concluded that airway fibrosis occurs as a result of MWCNT
15    inhalation with preexisting inflammation, suggesting that individuals with preexisting allergic
16    inflammation are susceptible.
17            Immune effects from oral exposure to MWCNTs have been suggested by one study. As described
18    in Section 5.1.7, Lim (20lib) observed a decrease in thymus weight in rat dams following administration
19    of MWCNTs at the highest dose tested (1,000 mg/kg-day).


      5.1.9.  In Vitro Data
20            In vitro data can be used to make judgments on the toxic potential of stressors, but the relevance
21    of in vitro data to predicting toxicological responses of "real-world" exposures is not always clear.
      March 2013                                     5-14                   DRAFT-DO NOT CITE OR QUOTE

-------
  1
  2
  3
  4
  5
  6
  7
  8
  9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
        Multiple in vitro studies were identified
for MWCNTs; a few were selected for discussion
here and inclusion in Table F-9 in Appendix F to
highlight the major themes and important concepts
covered in the literature.
        One in vitro study was identified on
ocular effects of MWCNTs. 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
    O        OO
with white leghorn chicken eggs (Ki shore 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) investigated
       Additional Information Highlight Box 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).  A number  of   different mechanism-based  high-
    throughput in vitro methods can be used predictively as
    screening tools and then validated using a limited amount of
    in vivo studies (Nel et al.,  In  Press).  These  types  of
    approaches  for characterizing effects of MWCNTs must be
    carefully developed and interpreted,  as there is uncertainty
    involved in extrapolating  in vitro results to in vivo effects.
effects of MWCNTs on platelet aggregation and 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 potentially identify data
gaps in in vivo data.
        Other in vitro studies have found that 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, in human T cells, MWCNTs were found to decrease
cell viability and increase  apoptosis in a dose- and time-dependent manner at concentrations between 40
March 2013
5-15
                                                                               DRAFT-DO NOT CITE OR QUOTE

-------
 1    and 400 ug/mL (Bottini et al., 2006). The study authors noted that the level of toxicity was significantly
 2    greater for oxidized MWCNTs compared to their pristine counterparts (Bottini et al., 2006).
      5.1.10.  Genotoxicity/Mutagenicity
 3           Standard and modified in vitro genotoxicity tests have been conducted to investigate the
 4    genotoxic potential of CNTs. However, their genotoxic potential is uncertain at this time, as available
 5    tests have shown contradictory results, which might be due to differences in composition and
 6    physicochemical characteristics of the CNTs (see Table F-9 in Appendix F).
 7           MWCNTs were not found to be mutagenic when evaluated with the Ames test17 using
 8    Salmonella typhimurium with and without metabolic activation (Di Sotto et al.. 2009; Wirnitzer et al..
 9    2009). However, bacterial mutagenicity-based assays might not be suitable for detecting genotoxicity
10    induced by nanoscale materials because prokaryotes cannot perform endocytosis, and the nanoscale
11    materials might not be able to diffuse across the bacterial cell wall. This lack of uptake could lead to
12    false-negative results (Singh et al., 2009). Purified MWCNTs also did not  show genotoxic activity in
13    several other assays at different dose levels and in different test systems, including micronucleus and
14    sister-chromatid exchange assays of human lymphocyte cells (Szendi and Varga. 2008) and a
15    chromosome aberration assay of Chinese hamster lung cells (Asakura et al., 2010). The MWCNTs used in
16    these tests were predominantly high purity and contained minimal metal impurities such as metal
17    catalysts, which are included in the commercial MWCNT preparation as a result of the synthesis process.
18    Such impurities could influence the  genotoxic potential of MWCNTs.
19           Other micronucleus assays have reported significant MWCNT-induced increases in micronuclei
20    of rat lung epithelial cells, Chinese hamster lung cells, and human lung carcinoma A549 cells (Asakura et
21    al.. 2010; Muller et al.. 2008a: Kato et al.. In Press). MWCNTs also have acted as clastogenic and
22    aneugenic agents simultaneously in  human blood cells (Cveticanin et al.. 2010). Using the murine
23    macrophage cell line RAW 264.7, Migliore et al. (2010) observed a significantly increased MWCNT
24    dose-related percentage of DNA in comet tails in a Comet  assay and a significant cytotoxic effect in a
25    Trypan blue assay. Cavallo et al. (2012) also reported a MWCNT concentration-dependent, statistically
26    significant induction of direct DNA  damage in human lung epithelial A549 cells evidenced by percentage
27    of DNA in comet tails in a Comet assay that corresponded with reduced cell viability; however, the
28    authors noted that oxidative DNA damage was not statistically significant. Patlolla et al. (201 Ob; 2010a)
      17The Ames test is a bacterial reverse mutation assay, designed to determine mutagenicity of the test compound.
      March 2013                                     5-16                   DRAFT-DO NOT CITE OR QUOTE

-------
 1    observed a statistically significant, dose-dependent increase in the percentage of DNA in comet tails in a
 2    Comet assay in normal human dermal fibroblast cells.
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
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 forms of
MWCNTs 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
  DecaBDE Can Inform MWCNT Assessment
Research  indicates  that  decaBDE  does  not  induce
genotoxicity. DecaBDE does appear, however, to target the
liver and thyroid in studies of Carcinogenicity (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).
al.. 2008; Takagi etal.. 2008). Therefore, the lung could be another target of MWCNT Carcinogenicity.
       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 postexposure. Similarly, Varga and Szendi
(2010) found that peritoneal injection of 10 mg MWCNTs did not result in development of
mesotheliomas in F-344 rats examined at 12 months postexposure, but did result in a granulomatous
reaction. Intratracheal instillation studies are useful for evaluating respiratory toxicity for particles, such
as MWCNTs, because they produce results that are qualitatively similar to those from inhalation studies
for endpoints such as pulmonary inflammation and fibrosis; however, such studies also have limitations
because treatment with bundled particles can produce artifactual granulomatous lesions (Muller et al..
2005). One possible reason for the mesotheliomas observed in some studies is that MWCNTs are more
cohesive than asbestos; consequently, MWCNTs bundle easily into granules after instillation or injection
into animals (Schulte etal.. 2010; Sakamoto et al.. 2009; Takagi et al.. 2008). 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.
       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).
      March 2013
                                               5-17
                   DRAFT-DO NOT CITE OR QUOTE

-------
                                   Additional Information Highlight Box 16:
                      Mechanisms of toxicity for cancer and non-cancer 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 a human microvascular endothelial cells (Pacurari etal., 2012). Muller et al. (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, early fibrosis (Murphy etal., 2012: Murphy etal., 2011), and fibrosis (Ryman-Rasmussen
    etal., 2009b). Additionally, MWCNTs have been shown to penetrate into the alveolar region of the lung (Delormeetal., 2012) and
    to cause inflammation due to accumulation of alveolar macrophages (Schinwaldetal.). These biological events have been shown
    to lead to mesothelioma (Donaldson et al., 2010): however, 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
 1            Sacks et al. (2011) defined susceptibility as "individual- and population-level characteristics that
 2    increase the risk of health effects in a population, including, but not limited to, genetic background, birth
 3    outcomes (e.g., low birth weight, birth defects), race, sex, life stage, lifestyle (e.g., smoking status,
 4    nutrition), preexisting disease, socioeconomic status (e.g., educational attainment, reduced access to
 5    health care), and characteristics that may modify exposure ... (e.g., time spent outdoors)." In this section,
 6    populations susceptible to MWCNT impacts based on characteristics such as age, genetic background,
 7    and disease are considered. Characteristics that could modify exposure and increase susceptibility are
 8    discussed in Section 4.2.4; for a discussion on impacts related to socioeconomic status, see Section 5.3.
 9            No information was identified regarding MWCNTs and susceptible populations. Because
10    MWCNTs appear to induce inflammatory and fibrotic effects (see Sections 5.1.3 and 5.1.8). however,
11    individuals with existing  pulmonary disease and children with higher respiration rates could be
12    susceptible populations.
      5.2.  Ecological Effects
13            This section presents a summary of data on the potential ecological impacts of environmental
14    contamination with MWCNTs. Specific information from the studies reviewed for this case study can be
15    found in Appendix F.2. Considerations for ecological impact include the absolute and relative toxicity of

      March 2013                                       5-18                    DRAFT-DO NOT CITE OR QUOTE

-------
 1    MWCNTs and other factors such as bioaccumulation and biomagnification potential (see Appendix
 2    G.4.2). For aquatic ecosystems, little information was identified for MWCNTs; conversely, much
 3    information was identified on the potential effects of MWCNTs in terrestrial ecosystems. The terrestrial
 4    ecosystem studies focus on agriculturally relevant plants and soil microbes. In both aquatic and terrestrial
 5    ecosystems, studies are predominantly laboratory-based experiments on single species. Therefore, limited
 6    research on the broad ecological impact of MWCNT exposure is available, which might be the result of
 7    limited data on its presence in the environment. As mentioned in Appendix G.4.2. ecological receptors
 8    can be exposed to MWCNTs attached to textile fibers, embedded in polymers, or sorbed to other
 9    particles, all of which are more likely to occur in the environment than exposure to the pristine
10    compound. Studies examining exposure to larger textile scraps, polymer particles, and other
11    heterogeneous compounds containing MWCNTs, however, are lacking. The results of laboratory studies
12    using pristine compounds must therefore be considered, recognizing that results might not  translate
13    directly into real-world exposure scenarios.
      5.2.1. Aquatic Receptors
                                          Priority Research Area: Impacts
                       J
        E £
        « w
        u r
          o
          0.
          E
                       7/13
5/13
                       1/13
                                                                   Reproductive
                          M
                                           H
                                                            01
                                                            E -T
                                                                M
                                                                                 w
H    M
                E-RRF Confidence Rating  E-RRF Confidence Rating  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
            w
              o
              I-1
                                                        Seven out 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  the this information,  developmental and
                                                        other sublethal endpoints were the highest priorities for
research in this area, followed by survival and reproductive effects. See Section 6.3.4.3 for more information on this priority
area and Section 1.1.3 for a detailed explanation of the prioritization process.
                         M    IT
                E-RRF Confidence Rating
      March 2013
                                5-19
                                                                       DRAFT-DO NOT CITE OR QUOTE

-------
                                   DecaBDE Can Inform MWCNT Assessment
       Bioaccumulation and bioavailability of 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 (Qin et
       al., 2010). Decreased thyroid hormone expression also was observed in Chinese rare minnow and lake trout after exposure to
       decaBDE via water and diet, respectively (Tomy et al., 2004: Li et al., In Press). 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.
 1            Limited information was identified on the toxicity of MWCNTs to algae and aquatic plants (see
 2    Section 5.2.1.1). Only a few studies have investigated toxicity of MWCNTs to aquatic invertebrates and
 3    aquatic vertebrates (see Sections 5.2.1.1 and 5.2.1.2). Although these studies provide information for
 4    acute effects, they vary with regard to endpoints, doses, functionalization, and other material
 5    characteristics (see Text Box 5-1). Little information was identified regarding toxicity to benthic
 6    invertebrates (see Additional Information Highlight Box 17). and most MWCNTs released to the aquatic
 7    environment are expected to accumulate in the benthic environment  (see Sections 3.1 and 3.3). A limited
 8    amount of information was identified on ecosystem effects due to chronic MWCNT exposure.
 9            As discussed in Sections 3.1 and 3.3. MWCNTs have low water solubility and are expected to
10    partition to sediment where they might be available primarily to benthic organisms. Physicochemical
11    properties of MWCNTs suggest potential for bioaccumulation in aquatic systems (Helland et al.. 2007).
12    but no studies were identified on this topic. The potential impact of bioaccumulation of MWCNTs  in
13    aquatic systems is therefore uncertain.

             5.2.1.1. Algae, Aquatic Plants, and Aquatic Invertebrates
14            Table 5-2 provides a comparison of key reference values identified for the effects of decaBDE
15    and MWCNTs on algae, aquatic plants, and aquatic invertebrates. More detailed information on decaBDE
16    is available in Appendix H. Table F-12. Table F-14.  and Table F-15  in Appendix F summarize details of
17    the MWCNT studies identified and reviewed for this section.
      March 2013                                        5-20                    DRAFT-DO NOT CITE OR QUOTE

-------
Table 5-2.    Effects of decaBDE and MWCNTs on aquatic receptors: Algae, plants,
         and invertebrates.
Organism
Sediment
oligochaetes
Algae
Zebra
mussels
Macrophytes
Sediment/
Benthic
organisms
MWCNTs
Effect Effect level
Citation
ND
Growth 1 mg/L
inhibition LOEC
Wei et al.
(2010)
ND
Positive effects on recolonization
and community structure after 3
months of exposure in sediment
Leptocheirus 68 grams/kg
plumulosus:
LCso
Hyalella >264 grams/kg
azteca:
LCso
Velzeboer et al.
(2011)
Kennedy et al.
(2008)
Kennedy et al.
(2008)
DecaBDE
Effect Effect level1 Citation
Acute NOEC >5,000 mg/kg Hardy (2002a)
28-day NOEC >3,841 mg/kg ACC (2001a. b)2
Growth >1 mg/L Hardy (2002a)
inhibition,
96-hr ECso
DNA damage 0.1 to 10 ug/L Riva et al.
(2007)
ND
ND
March 2013
5-21
DRAFT-DO NOT CITE OR QUOTE

-------
Table 5-2, cont.: Effects of decaBDE and MWCNTs on aquatic receptors: Algae, plants, and
           invertebrates.
Organism
Water fleas
MWCNTs
Effect
Ceriodaphnia
dubia:
acute LCso
C. dubia:
Subchronic
growth
inhibition ECso
3-generation
reproduction
ECso
C.dubia: LC/s
Effect level
2-1 00 mg/L
50.9 mg/L
4-1 7 mg/L
26 mg/L
(un-derivatized
MWCNTs)
Citation
Li and Huang
(2011)
Kennedy et al.
(2008)
Li and Huang
(2011)
Kennedy et al.
(2009)
DecaBDE
Effect
Daphnia magna:
21-day LOEC
(growth)
21-dayECso
(survival,
reproduction)
Effect level1
BDE-209: ND;
pentaBDE:
9.8 ug/L
BDE-209: ND;
pentaBDE:
14 ug/L
Citation
Wildlife
International
report submitted
to Chemical
Manufacturers
Association
(Drottar and
Krueger, 1998)2
Wildlife
International
report submitted
to Chemical
Manufacturers
Association
(Drottar and
Krueger, 1998)2
 1Concentration in media (water [units: ug/L or mg /L] or sediment [units: mg/kg]).
 2As cited in Environment Canada (2006).
 ND = No data identified, NOEC = No-observed-effect concentration, ECso = Median effective concentration
                                 Additional Information Highlight Box 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 (V/7/osa 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.
March 2013
5-22
DRAFT-DO NOT CITE OR QUOTE

-------
 1           The effects of MWCNTs on marine algae, sediment macrophytes, and water-dwelling
 2    invertebrates have not been extensively studied; only four published studies were initially identified (Li
 3    and Huang. 2011; Velzeboer etal.. 2011; Wei etal.. 2010; Kennedy et al.. 2008). Because MWCNTs are
 4    likely to partition to sediment, benthic organisms are expected to be primary receptors (Christian et al..
 5    2008). Initially, no studies that investigated the effects of MWCNTs on benthic invertebrates, however,
 6    were identified. See Additional Information Highlight Box 17 and Section 6.3.4.3 for information on
 7    studies that were  identified after the collective judgment prioritization process.
 8           In a study on macrophyte growth, experimental plots were cleared of all living organisms and
 9    MWCNTs were added to the sediment. After three months, researchers observed that the density of
10    macrophytes that had recolonized the plots was positively correlated with MWCNT levels (Velzeboer et
11    al.. 2011). This result was counter to the authors' initial hypothesis based on previous laboratory
12    experiments that  macrophyte growth and species composition would be negatively affected by MWCNTs,
13    indicating a level of complexity in community4evel effects of MWCNTs in real-environment situations
14    that is not well understood (Velzeboer et al.. 2011).
15           Unicellular green algae (Dunaliella tertiolectd) exposed to carboxylated MWCNTs in sea water
16    did not exhibit inhibited growth until concentrations reached 1 mg/L and above (Wei etal.. 2010).
17    Growth lagged up to 23 days behind the control, and exponential growth rates were reduced by 35% when
18    exposure was 10  mg/L, indicating mid-exponential growth phase cytotoxicity at high exposures (Wei et
19    al.. 2010).
20           Two studies provided a wide variety of data for Ceriodaphnia dubia, a species of water flea (Li
21    and Huang. 2011; Kennedy et al.. 2008). The MWCNT studies describe an acute median lethal
22    concentration in the mg/L range, the variation of which might be due to differences in functionalization
23    treatment and diameter size of the MWCNTs (Li and Huang. 2011) (see Table F-15 in Appendix F for
24    study-specific details and Text Box 5-1 for discussion of how physicochemical properties affect toxicity).
25    Subchronic and chronic growth and reproduction tests show that MWCNTs that had been treated to
26    increase dispersal and limit bundling (a common treatment in MWCNTs—see Section 2.2.3.1) are not
27    likely to cause significant, population-level effects until high doses (1- to 100-mg/L range) are reached
28    (Li and Huang. 2011). Even though MWCNTs  are expected to partition to sediment, functionalization and
29    suspension in natural organic matter could improve dispersion and solubility of MWCNTs in aqueous
30    media and might  increase the exposure levels of MWCNTs to water-dwelling aquatic organisms
31    (ODriscoll et al..  2010: Kennedy et al.. 2008).
      March 2013                                     5-23                   DRAFT-DO NOT CITE OR QUOTE

-------
            5.2.1.2. Aquatic Vertebrates

1            Table 5-3 provides a comparison of key reference values identified for the effects of decaBDE

2    and MWCNTs on aquatic vertebrates. Additional information on decaBDE is available in Appendix H.

3    Table F-13, Table F-16. and Table F-17, in Appendix F summarize details of the MWCNT studies

4    identified and reviewed for this section.
     Table 5-3.    Effects of decaBDE and MWCNTs on aquatic receptors: aquatic vertebrates.
         Organism
   Exposure
             Effect
   Effect level
        Citation
       MWCNTs
      Zebrafish
Water
Reduced blood circulation
70 ug/mL
Asharani et al. (2008)
                     Water
               Developmental effects
                                60 ug/mL
                 Asharani et al. (2008)
                     Water
               Increased mortality LOAEL
                                60 ug/mL
                 Asharani et al. (2008)
                     Microinjection    Developmental effects NOEL
                                               >2 ng/embryo      Cheng et al. (2009)
                     Microinjection    Second-generation reduced survival    2 ng/embryo
                                                                Cheng et al. (2009)
       Japanese
       medaka
Water
Developmental effects LOAEL
1,500ug/mL
Kim etal. (2012)
       DecaBDE
      African clawed   Water
      frog
               Thyroid effects LOAEL
                                 1ng/L
                 Qin etal. (2010)
       Lake trout
Diet, chronic     Decreased thyroid hormones LOAEL   2.5 ng/gram
                                                 Tomy et al. (2004)
       Rainbow trout   Diet, chronic     Increased liver weight LOAEL
                                               7.5 ng/kg
                                                 Kierkegaard etal. (1999
                     NR
               Vitellogenin production
                                 NR
                 Nakari and Pesala (2005)
       Lake whitefish   Diet, chronic     Decreased growth LOAEL
                                               2 ug/gram
                                                 Kuoetal. (2010)
       Chinese rare
       minnow
Water, chronic    Decreased growth LOAEL
                                 10ug/L
                     Water, chronic   Spermatogenesis inhibition LOAEL     10ug/L
                     Water, chronic   Upregulation of thyroid hormones
                                               Variably occurred   Li et al. (In Press)
                                               at 0.1-10 ug/L
       NR = Not reported, NOAEL = No-observed-adverse-effect level, LOAEL = Lowest-observed-adverse-effect level


5            Four studies were identified that investigated the effects of MWCNTs on fish; three of these

6    studies used common laboratory species—zebrafish (Danio rerio) and Japanese medaka (Oryzias latipes),

7    while one used rainbow trout (Oncorhynchus mykiss).
     March 2013
                                  5-24
                                          DRAFT-DO NOT CITE OR QUOTE

-------
 1            In one zebrafish study, embryos acutely exposed to MWCNTs showed dose-dependent increased
 2    mortality rates, reduced blood circulation, and delayed development (hatching), and developmental
 3    defects (bent notochord) starting at 60 ug/mL (Asharani et al., 2008). The other zebrafish study (Cheng et
 4    al., 2009) examined embryos following a smaller yet more direct exposure—a single microinjection of
 5    MWCNTs—and observed changes in enzyme expression signifying an immune response. Although no
 6    increase in mortality or developmental defects were observed in the exposed zebrafish through adulthood,
 7    survival in the second generation was significantly decreased. Similarly, Kim et al. (2012) observed a
 8    statistically significant increase in heart abnormalities, absence of swim bladders, caudal fin
 9    malformation, and pericardial and peritoneal edemas in Japanese medaka embryos following 4 days of
10    continual exposure to 2,000 ug/L functionalized MWCNTs in the water. Increased mortality also
11    occurred in response to exposures to 1,500 and 2,000 ug/L, and exposure to 1,500 ug/L resulted in a
12    hatching delay.
13            Klaper et al. (2010) further investigated the possibility of an immune response in fish with an in
14    vitro study of rainbow trout. In this study, MWCNTs did not elicit an antiviral response at sublethal doses
15    up to  10 ug/mL; regardless of various types of functionalization.  Investigators did, however, observe
16    expression of IL-lbeta, evident of macrophage stimulation, at 5 and 10 ug/mL (Klaper et al., 2010).
      5.2.2. Terrestrial Receptors
                                     Unprioritized Research Area: Impacts
            Terrestrial
              Biota
      si
         o
         a.
                     1/13
                     7/13
                     5/13
        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
  01
  E -r
  to
W
       Other Sublethal
          Endpoints
    H    M    L
E-RRF Confidence Rating
    H    M    L
E-RRF Confidence Rating
        H    M    L
    E-RRF Confidence Rating
 One out of 13 RTI workshop participants (8%) identified
 impacts of MWCNT  exposure to terrestrial biota  as
 important to risk assessment. Based on this information,
 impacts in terrestrial  biota 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.
      March 2013
                     5-25
                 DRAFT-DO NOT CITE OR QUOTE

-------
      5.3.  Other Impacts
                                    Priority Research Area: Exposure Route
           1
       « w
      uj tr
         o
         Q.
         E
5/13
                     4/13
                     4/13
                                   a
                                   •B
                                   a
                                   o
                                   I
                                                         D)
                                                         •c
                                                         o
                                                          -3 I
                                                          o:
   S
                                                          t
                                                          o
                                                          I-1
                                                                                    Environmental
                                                                                      Resources
       • - 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          ~H    M    L           H    M    L
               E-RRF Confidence Rating  E-RRF Confidence Rating   E-RRF Confidence Rating
              Five out 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 the 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.
 1           As stated in Chapter 1. the CEA framework considers not only human and ecological health
 2    impacts, but also aesthetic, environmental, social, legal, ethical, and economic impacts. Such impacts
 3    might be associated with impacts on specific socioeconomic sectors (e.g., disparate impacts on
 4    environmental justice communities), the environment as a whole (e.g., climate change, depletion of
 5    natural resources, energy demand), or the built environment (e.g., damage to building facades).
 6           Apart from the impacts discussed in Sections 5.1 and 5.2. the only other impacts considered in
 7    this case  study are those for which a plausible premise can be developed to support assumptions that a
 8    discernible impact might occur as a result of the life cycle of MWCNTs flame-retardant upholstery textile
 9    coatings. Data from MWCNTs on each of these  impacts would be required for a thorough comparison;
10    however, in all cases where other impacts were identified as being of concern for either decaBDE or
11    MWCNTs, no data were available for the comparison material. For example, for decaBDE, empirical data
12    have revealed a correlation between decaBDE body burdens and socioeconomic status, indicating that
13    effects having environmental justice implications are plausible for decaBDE. No such empirical data exist
14    relating MWCNTs to other impacts, but the background literature on processes involved in manufacturing
15    similar materials (e.g., carbon nanofibers [CNFs] and SWCNTs) provides some basis for concerns
16    regarding potential impacts of MWCNTs on energy demand, resource depletion, climate change, and
17    economics.
      March 2013
                                5-26
DRAFT-DO NOT CITE OR QUOTE

-------
      5.3.1.  Environmental Justice
 i
 2
 O
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
        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."18 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..."19 As aresult, environmental justice
impacts include those in which a particular
   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, poor 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.
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
                                  DecaBDE Can Inform MWCNT Assessment
       No information was identified that examined 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.
17            No information was identified that examined impacts on energy demand and natural resource
18    depletion associated with the production of MWCNT flame-retardant textile coatings. Various studies,
19    however, have calculated a large range of minimum energy requirements for synthesis of CNTs (Khanna
20    et al.. 2008; Cipiriano et al.. 2007; Smalley et al.. 2007). Table 5-4 presents an overview of some
21    estimated minimum energy requirements and process rates for CNT synthesis.
22            The differences in energy requirements are largely attributable to different synthesis processes,
23    different process rates, different feedstocks, and process improvements as synthesis of CNTs has been
       U.S. EPA Compliance and Enforcement. Environmental Justice, http://www.epa.gov/environmentaljustice/
       9ibid
      March 2013
                                               5-27
                     DRAFT-DO NOT CITE OR QUOTE

-------
 1    optimized (Gutowski et al., 2010; Kushnir and Sanden, 2008). Additionally, these energy requirements
 2    represent only the minimum for synthesis and do not consider the energy required for purification,
 3    additional infrastructure (e.g., equipment needed to regulate environmental conditions during synthesis
 4    and processing), and other related processes. Including these additional energy requirements, Gutowski et
 5    al. (2010) estimated that CNTs could be one of the most energy-intensive materials of all time. Although
 6    information regarding the energy requirements for MWCNT synthesis is limited, such energy
 7    requirements likely also would be sizeable and span a large range as synthesis processes are  continually
 8    optimized.
 9            One environmental impact assessment examined the water inputs required for two methods of
10    continuous synthesis of SWCNTs via chemical vapor deposition (CVD) (Kolosnjaj-Tabi et al.. 2010).
11    These results indicated that production of SWCNTs can require significant amounts of water (almost
12    65,000 kg/hour at a manufacturing  rate of about 595 kg/hour of SWCNTs). The relationship  between
13    water requirements for SWCNT synthesis and MWCNT synthesis, however, is unclear.
Table 5-4.
Synthesis
process
CVD
Arc discharge
HiPCO®
HiPCO®
Floating
catalyst CVD1
Laser
ablation2
Estimated minimum energy requirements and process rates for synthesis of CNTs.
Material
CNF (methane-based)
CNF (ethylene-based)
SWCNT
SWCNT
SWCNT
MWCNT
MWCNT
Process rate
(kg/hr)
1.30x10-2
1.80x10-2
8.10x10-5
4.50 x 10-4
4.50 x 10-4
NR
NR
Synthesis energy
requirements
(J/kg)
3.13 x109
2.22 x 109
8.73x101°
2.41 x 1010
3.18x101°
2.95x108 (thermal)
1.87xio« (electric)
2.11 x 1Q8 (thermal)
9.4 xlO9 (electric)
Estimated
energy per
hour (J/hr)
4.07 x 107
3.96 x 107
7.07 x 1Q6
1.08x10?
1.43x10?
NR
NR
Reference
Khanna et al. (2008)
Heal vetal. (2008)
Smallevetal. (2007)
Heal vetal. (2007)
Kushnir and Sanden (2008)
Kushnir and Sanden
(2008)3
      1 Benzene gas feedstock.
      2Graphite feedstock.
      3Authors report both baseline (shown above) and "efficient" estimates (not shown).
      CVD = chemical vapor deposition; CNF = carbon nanofiber; HiPCO® = a high pressure carbon monoxide synthesis process; NR = not
      reported.
      Source: Gutowski etal. (2010).
      March 2013
5-28
DRAFT-DO NOT CITE OR QUOTE

-------
 1            A life-cycle assessment of CNF production via CVD calculated potential impacts on
 2    acidification, eutrophication, and ozone layer depletion (Khanna et al., 2008). The results of this analysis
 3    are presented in Table 5-5. The authors found that CNF production of both methane-based and ethylene-
 4    based CNFs has minor impacts on acidification, eutrophication, and ozone layer depletion. This study did
 5    not incorporate CNF emissions into its calculations, however, due to a lack of data on fate, transport, and
 6    impacts of CNFs (Khanna et al.. 2008). The authors noted that the lack of models to predict endpoint
 7    effects of some emissions (e.g., CNF emissions) renders these calculations uncertain. Plata et al. (2009)
 8    found that release of gases such as methane, volatile organic compounds, and polycyclic aromatic
 9    hydrocarbons from MWCNT synthesis is possible. Methane release would likely have a negligible impact
10    on local air pollution and ozone depletion compared to existing methane sources; however, release of
11    volatile organic compounds such as 1,3-butadiene and benzene, could be  significant on a local scale
12    (Plata etal.. 2009).
      Table 5-5.    Environmental assessment of production of 1 kilogram (kg) of carbon nanofibers.1

                                                         Impact2
      Impact category                     	           Unit
                                          Methane-based CNF  Ethylene-based CNF
      Acidification                                  5.5                4.0               Kg S02 Equivalent
      Eutrophication                                 4.0                3.0               Kg P04 Equivalent
      Ozone layer depletion potential                 2.8 x 1Q-5           2.8 x 1Q-5          Kg CFC-11 Equivalent
      Environmental impacts of production of 1 kg of carbon nanofibers (CNFs) using chemical vapor deposition calculated by SimaPro® Eco-
      Indicator 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).
      formalized and weighted impacts.
      Source: Khanna etal. (2008).

13            Eckelman et al. (2012) developed a life-cycle framework to compare the impact on ecological
14    (and specifically aquatic) organisms of CNT production versus CNT releases to environmental media
15    during the product life cycle. This assessment used existing data and a recently established consensus
16    model for life-cycle impact assessments, USEtox, to estimate ecotoxicity from emissions during CNT
17    production and CNT releases during product use and disposal for "realistic" and "worst case" scenarios.
18    They calculated the potentially affected fraction of aquatic organisms per unit mass of CNTs released and
19    "comparative toxic units for ecosystems" for different methods of synthesis and proj ected scale-up
20    results. The theoretical framework was useful  for comparing the  relative impacts of different synthesis
21    methods, what proportion of potential ecotoxicity is due to the synthesis process compared to required
      March 2013                                       5-29                    DRAFT-DO NOT CITE OR QUOTE

-------
  1    purification methods, and also made projections based on future increased scale of production.
 2    The authors concluded that the greatest ecotoxicity impacts do not result from release of CNTs during the
 3    product life cycle or from unused reagents or synthesis products during production, but rather from the
 4    emission of metals due to the combustion of fossil fuels necessary to generate electricity for CNT
 5    synthesis or production of various inputs.
 6            Another study analyzed a broad range of environmental impacts from the production of one
 7    SWCNT polymer mesh (a transistor/electromagnetic interference-shielding application) (Dahlben and
 8    Isaacs. 2009). The study examined both the  energy requirements for raw material extraction and
 9    manufacturing and emissions from these processes (excluding SWCNT  emissions due to the current lack
10    of consensus on its effects). The authors found that manufacturing this CNT application could damage
11    ecosystems and resource quality, as shown in Table 5-6. Damage to ecosystems (due to
12    acidification/eutrophication and land use) was expressed as the loss of species over a certain area in a
13    given time. Damage to resource quality was expressed as the surplus energy needed for future extractions
14    of minerals and fossil fuels (due to the resources needed to extract these materials for SWCNT polymer
15    me sh production).
      Table 5-6.    Environmental assessment of production of one SWCNT polymer mesh.
      Impact category1                                     Impact2                Unit
      Acidification/Eutrophication                             1.229x10-2               PDF*m2yr
      Land use                                           4.440 x 1Q-3               PDF*m2yr
      Minerals                                           1.117xlQ-3               MJ Surplus
      Fossil fuels                                         7.531x10-1               MJ Surplus
      1 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).
      formalized and weighted impacts.
      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).

16            The authors found that production  of SWCNT polymer mesh generated larger fossil fuel impacts
17    relative to other measured environmental impacts. The authors also found that fossil fuel impacts were
18    dominated by processes requiring energy-intensive equipment (e.g., furnace for synthesis, wet bench for
19    cleaning, and  spinner for coating). Although the impact measures were reported to be low for all
20    categories, they represent the resource impacts of producing a single SWCNT polymer mesh. The level of
21    aggregation required to compile these metrics, however, coupled with the lack of models to predict
      March 2013                                       5-30                    DRAFT-DO NOT CITE OR QUOTE

-------
 1    endpoint effects of some emissions (e.g., from nanotubes), render these metrics highly uncertain (Khanna
 2    etal.. 2008). Additionally, this study did not include impacts of SWCNT emissions in its overall
 3    calculations of environmental impact. How the environmental impacts of SWCNT polymer mesh
 4    production differ from the impacts of flame-retardant textiles using MWCNTs or other alternative
 5    materials, such as decaBDE, is unclear.
      5.3.3. Climate Change
 6           No information was identified that directly examined climate change impacts due to MWCNT
 7    flame-retardant coatings. Empirical data suggest, however, that release of greenhouse gases such as
 8    methane, volatile organic compounds, and polycyclic aromatic hydrocarbons from MWCNT synthesis is
 9    possible (Plata. 2009; Plata et al.. 2009). The authors concluded, however, that for commercial-scale
10    production the contribution of MWCNT synthesis to atmospheric methane will be negligible compared to
11    existing methane sources, and that volatile organic compound emissions might be significant only on the
12    local scale (Plata et al.. 2009). Singh et al. (2009) calculated emissions of 4 kg CO2/kg SWCNT from one
13    method of CVD synthesis. Altering CNT synthesis methods might minimize formation of these or other
14    hazardous by-products. For example, Plata (2009) found that by identifying select thermally generated
15    compounds correlated with CNT growth rate, such compounds could be delivered to the catalyst without
16    thermal treatment and thereby eliminate the need to heat reactant gases.
17           On the other hand, a life-cycle assessment of CNF production calculated that manufacture of 1 kg
18    of methane-based CNFs equals at least 700 kg of CO2 equivalents, and 1 kg of ethylene-based CNFs
19    equals at least 400 kg of CO2 equivalents (Khanna et al.. 2008). In other words, the authors calculated that
20    production of 1 kg of methane-based CNFs is equivalent to CO2 emissions from 78.5 gallons of gasoline
21    consumed, while production of 1 kg of ethylene-based CNFs is equivalent to CO2 emissions from
22    44.8 gallons of gasoline consumed.20 The relationship between climate change effects due to synthesis of
23    SWCNTs, CNFs, and MWCNTs is unclear.
      20EPA GHG Calculator available at http://www.epa.gov/cleanenergv/energv-resources/calculator.html.
      March 2013                                    5-31                   DRAFT-DO NOT CITE OR QUOTE

-------
5.3.4. Economics
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
       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
     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.
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
than SWCNTs, and further optimization of MWCNT manufacturing is likely to decrease manufacturing
costs further.
March 2013
5-32
                                                                         DRAFT-DO NOT CITE OR QUOTE

-------
          Chapter  6.  Identifying and  Prioritizing
              Research  Needs to  Support Risk
            Assessment  and  Risk Management
 i
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
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).
                                         Risk
                                      Assessment
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
various trade-offs between individual risk management options (NRC. 2009). More recently, the need to
     March 2013
                                     6-1
DRAFT-DO NOT CITE OR QUOTE

-------
 1    consider longer term consequences of alternative options in a broader context has been recognized, which
 2    would encompass social, environmental, and economic indicators (NRC. 2011). All of these
 3    recommendations indicate the need for risk assessments to address cumulative effects from multiple
 4    exposures to one or more stressors, vulnerability of susceptible populations, and potential for impacts
 5    throughout the product life cycle (NRC. 2011. 2009). A shift toward this more holistic, systems-based
 6    approach would provide more complete information to risk assessors to better inform risk managers in
 7    making decisions that support long-term, sustainable management practices (NRC. 2011).
 8           Many efforts are underway to gather information and develop approaches that support the
 9    implementation of such recommendations [e.g.,  (Anastas. 2012; Lavoie etal.. 2010; Rossi et al.. 2006)].
10    The CEA approach represents one such effort by recognizing that research supporting risk assessment
11    must be transparently planned and executed if risk assessments are to be used effectively and efficiently
12    for evaluating risk management options and understanding the longer term consequences of a broad scope
13    of complex information (e.g., cumulative risk, life-cycle analyses). The External Review Draft of this
14    document was used in the collective judgment step of the CEA process as part of an effort to plan such
15    research for MWCNTs in flame-retardant textile coatings. Specifically, a group of expert stakeholders
16    representing a variety of technical backgrounds (e.g., analytical chemistry, toxicology, polymer science)
17    and sectors (e.g., industry, academia, nongovernmental organizations) used this document as a starting
18    point for identifying and prioritizing research needs to support assessments that inform near-term risk
19    management goals (see  Section 1.1.3).
20           In planning research to support assessments, it is useful to review risk management goals to
21    understand the types of analyses and assessments that would inform management efforts. Such a review is
22    consistent with the connections highlighted in the CEA process diagram (see Figure 1-2). Research
23    outcomes in the form of data from single studies or assessment reports [which have compiled information
24    across multiple studies (e.g., risk assessments, life-cycle assessments, meta-analyses)], feed back into the
25    CEA framework to provide additional information necessary for assessing risk-related trade-offs. This
26    additional information is subsequently used in developing adaptive risk management plans. Such future
27    evaluations of risk-related trade-offs could focus on informing one of a variety of risk management goals
28    for the application of MWCNTs in flame-retardant textile coatings. As discussed in Chapter 1. the
29    selection of any one goal for a comparative CEA would depend on the needs of risk managers at the time.
30    One particular scenario, however, based on what might occur for this specific nanoenabled product, is
31    described below as an example that might inform research planning for the research priorities identified in
32    this application of CEA. The use of this scenario is not meant to imply actual assessment or risk
33    management recommendations, but rather to illustrate the type  of context in which  the research priorities
      March 2013                                     6-2                   DRAFT-DO NOT CITE OR QUOTE

-------
 1    identified here could inform future evaluations of MWCNTs. Additional considerations and examples of
 2    risk assessment and risk management decisions are discussed in Section 6.2.
 3           As outlined in Chapter 1, the use of MWCNTs in flame-retardant textile coatings is not common,
 4    although evidence suggests that they could be used more extensively in the future as conventional flame-
 5    retardants such as decaBDE are phased out due to concerns surrounding environmental persistence and
 6    human health effects (see Additional Information Highlight Box land Section 1.1.3). Should a flame-
 7    retardant textile coating containing MWCNTs be developed for use in the United States, the manufacturer
 8    or importer likely would submit a premanufacturing notice (PMN) (U.S. EPA. 2008c). Information in a
 9    PMN includes the identity of the chemical (i.e., name and structure), anticipated production volume, use
10    and disposal methods, human exposure estimates, and any readily available test data (U.S. EPA. 2010g).
11    Thus, one of the first risk management decisions for MWCNT flame-retardant textile coatings could be to
12    determine whether the material should be (1) produced without restriction or regulations, (2) imported,
13    produced, or used with limitations, or (3) prohibited from import, production, or use (U.S. EPA. 2010f).
14    The third outcome, prohibition, could result from several determinations, including insufficient
15    information on potential impacts of the material (U.S. EPA. 201 Of).
16           To support this and other risk management decisions about the use of MWCNTs in flame-
17    retardant textile coatings, information must be readily available to decision-makers to enable a considered
18    determination within the relevant time constraints. Ideally, the research objectives identified and
19    prioritized in this CEA application would provide such information within 3 to 5 years of initiating the
20    research. Notably, completion of all research is not feasible within this time frame due to practical
21    constraints. Thus, the collective judgment step of the CEA approach emphasizes the prioritization of
22    information gaps (see Figure 1-2).


      6.2. Identification and  Prioritization of  Research Needs
23           With the above context for identifying and prioritizing research needs to support future
24    assessments and risk management decisions for MWCNT, the lingering question is: How? As discussed
25    in the previous section and in Chapter 1, a group of expert stakeholders representing diverse technical
26    (e.g., toxicology, ecology, material science) and sector (e.g, industry, academia, government) perspectives
27    participated in a structured collective judgment process that supported equal representation of each
28    individual's input. Specifically, participants used a more detailed view of the CEA framework (see Figure
29    1-3) to consider each element of the framework (e.g.,  stages of the product life cycle or spatial zones in
30    the environment) in relation to potentially relevant risk factors (e.g., mobility in air) associated with the
31    life cycle of a MWCNT flame-retardant textile coating product.  Stakeholders used an online software-

      March 2013                                     6-3                   DRAFT-DO NOT CITE OR QUOTE

-------
 1    based tool that supported their consideration of each area of the CEA framework remotely. They used the
 2    tool in a series of exercises to qualify the: (1) importance of CEA framework components for future
 3    assessments, and (2) current state of the science of the components for supporting risk management
 4    initiatives (see Figure 1-3). These exercises were intended to actively engage stakeholders in identifying
 5    critical research needs based on each stakeholder's: (1) perception of the information presented in the
 6    case study, (2) individual experience, and (3) expert opinion. Based on previous applications of CEA, this
 7    approach is thought to be an improvement over one that would simply ask stakeholders to review and
 8    prioritize a predetermined list of data gaps, which could exclude important data gaps a priori.  This
 9    collective judgment exercise, coupled with a subsequent structured, face-to-face workshop, provided a set
10    of specific research needs (see Section  6.3).
11           As discussed in the previous section, planning and conducting research to fill the priority areas
12    discussed below (see Section 6.3) is facilitated by having an understanding of assessment and risk
13    management approaches. To help facilitate the transition from reviewing the case study to identifying and
14    prioritizing specific research needs, experts participating in the collective judgment exercise were
15    encouraged to consider the first step of conducting an assessment—problem formulation. Considering
16    problem formulation can similarly help plan research to fill the gaps that experts identified as priorities
17    (see  Section 6.3). Placing a greater focus on problem formulation is recognized as an important step
18    toward improving risk assessment (NRC. 2009). but the preceding step—anticipating what information to
19    research to support problem formulation—has  received less attention. The CEA approach places greater
20    emphasis on problem formulation during the research planning stage by engaging stakeholders to
21    prioritize areas for research mangers to consider in allocating research resources. In doing so, it ensures
22    that (1) major factors important to stakeholders are included early in the planning process, and (2) risk
23    assessors and managers have the information they need to develop risk assessments and management
24    plans that include those considerations. This approach is consistent with recent National Academy of
25    Science recommendations (NRC. 2011. 2009). Similarly, the inclusion of long-term, broad environmental
26    impacts in the CEA framework supports identifying research gaps in these areas, which are increasingly
27    recognized as essential to assessments but are often more difficult to incorporate because of insufficient
28    data or knowledge on how to include such information (NRC. 2011).
29           Thus, in planning research to support future assessments and risk management efforts, it is
30    essential to have a clear understanding  of problem formulation. This early step  in the risk assessment
31    process establishes the goals, scope, focus, and potential options to consider in  decision-making (NRC.
32    2009; Van Leeuwen et al..  1998). The problem formulation step helps establish the type of analyses, and
33    associated uncertainty and variability, which will be useful to a risk manager in making a decision about
34    the material, chemical, or technology of focus (NRC. 2009; Van Leeuwen et al.. 1998). It is important

      March 2013                                      6-4                    DRAFT-DO NOT CITE OR QUOTE

-------
 1    then to establish the type of research that will be useful for the analyses that inform a risk manager about
 2    a particular material, such as MWCNT. Notably, it is necessary to plan for a variety of types of risk
 3    management decisions, as shown in Table 6-1. Although this particular CEA case study was not
 4    developed with a specific risk management objective in mind, the identification of key research gaps
 5    within the CEA framework can inform research planning that supports multiple assessment objectives
 6    identified during the problem formulation stage of each respective assessment. These future assessments
 7    can then supply new information in the CEA framework for MWCNTs, which could subsequently be
 8    evaluated in collective judgment prioritization of risk-related trade-offs to inform specific risk
 9    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 from Van Leeuwen et al. (1998).
10            The External Review Draft of this document used the CEA framework to lay out the technical
11    aspects related to MWCNTs in flame-retardant textile coatings so that expert stakeholders could
12    determine (1) what types of risk management decisions are likely to be needed and (2) what information
13    would support assessments that inform those decisions. Stakeholder input was incorporated into this final
14    version of the case study document to emphasize areas of the CEA framework that stakeholders felt were
15    research priorities for risk assessment and risk management. The questions in Table 6-2 are examples of
16    those that could be asked during the problem formulation phase of an assessment to evaluate whether data
17    are available and useful (e.g., consider relevant endpoints, relevant exposure routes, doses and timing,
18    acceptable levels of uncertainty in assays selected, data variability) to support each component of an
19    assessment. The answers to these questions can help formulate an appropriate assessment approach or
20    facilitate the identification of additional data gaps that must be filled before the assessment can proceed.
      March 2013
                    6-5
              DRAFT-DO NOT CITE OR QUOTE

-------
1     In the context of this case study, the extent to which these questions, along with other questions identified
2     by stakeholders, could be answered indicates whether specific research initiatives should be pursued in
3     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?
                                    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?
      What are the characteristics of the
      exposed populations?
                                   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 are the assessment endpoints?   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?
      March 2013
                                                      6-6
DRAFT-DO NOT CITE OR QUOTE

-------
      6.3.  Research  Priorities  Identified through Collective
            Judgment in this Application of CEA
 1           The collective judgment process to engage expert stakeholders, which was summarized in Section
 2    6.2 and in Chapter 1. is described in detail in a separate report, which was prepared by the contractor that
 3    independently conducted the EPA-funded workshop (RTI. 2012). The areas of the CEA framework that
 4    experts most commonly identified as "important to consider in a risk assessment" (presented in Chapter 2
 5    through Chapter 5) are discussed here in the context of supporting research planning and future
 6    assessment and risk management of MWCNTs in flame-retardant textile coatings.
 7           For each area, the reasons why the experts rated certain research areas as priorities, is first
 8    discussed. When participants individually rated the importance of an area and their confidence that the
 9    current data could support risk management, they had the option of completing a checklist of "influential
10    factors." These factors (Table 6-3) represent various aspects of MWCNT science that could play a role in
11    determining, inducing, or otherwise influencing the potential risks associated with the particular area in
12    question and are listed after the rationales provided by participants. As this portion of the prioritization
13    process was voluntary, not all participants chose to identify influential factors and omission of an
14    influential factor does not necessarily imply that the  factor is unimportant. Nevertheless, identification of
15    the influential factors for each area provides some additional insight as to why a particular area might be a
16    priority, and what particular aspects of that area might warrant further research.
17           Next, other relevant literature, which was not included in the External  Review Draft of the case
18    study, is discussed in terms of whether that literature might improve confidence in the ability of the data
19    to support risk management decisions, or how it might support research planning. Finally, for those areas
20    that expert stakeholders  discussed at the workshop, the potential risk scenario that they identified for the
21    area is outlined along with the types of risk management decisions they noted might be made to mitigate
22    or avoid the potential risk. The type(s) of assessments that could inform these types of risk management
23    decisions that participants noted are laid out, followed by a table detailing the key research questions or
24    areas that experts identified along with their estimates of the financial and time resources needed to carry
25    out the research.21 For those Research Priority Areas not discussed at the workshop due to time
26    constraints, potential research directions are listed in the text. Research Priority Areas are presented below
      21 Note that time and financial estimates were generated by experts participating in the independently conducted RTI
      workshop. They are based on participants' experience with planning and conducting scientific research and are
      included here with the intention of providing a general indication of the level of resources that would support
      carrying out the identified priority research questions.
      March 2013                                     6-7                   DRAFT-DO NOT CITE OR QUOTE

-------
 1    in the order in which they appear in the CEA framework and in Chapter 2 through Chapter 5 of this
 2    document.
      Table 6-3.   Influential Factors options for all areas.
Methods, Techniques
Engineered Nano
Material
Characteristics
Surrounding Media
Physical Conditions
Chemical Conditions
Biological Conditions
Social Conditions
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, groundwater, 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
 3           Multiple areas of the product life cycle were considered to be Priority Research Areas according
 4    to workshop participants. Material synthesis and processing as well as product manufacturing were noted
 5    to be particularly important by workshop participants, in part because these areas of the CEA framework
 6    play a key role in determining all other parts of the CEA framework, including fate and transport in
 7    various environmental media, bioavailability and potential exposure, as well as ecological and human
 8    toxicological impacts. In a recent publication, Nowack et al.  (2012) similarly note the importance of
 9    characterizing engineered nanomaterials throughout the life cycle: how the starting material is
10    intentionally modified through functionalization (see Figure  2-2) and unintentionally modified by
11    environmental factors (see Figure 3-1) determines how it will behave in the environment, how organisms
12    can be exposed, and the extent to which it will adversely impact ecological and human receptors
13    [(Nowack et al.. 2012): see Additional Information Highlight Box 6]. While the chemical alterations to
14    nanomaterials are important for potential release and toxicity, the properties of the product matrix and the
15    manner in which nanomaterials are incorporated into a matrix are equally important (Nowack et al..
      March 2013
6-8
DRAFT-DO NOT CITE OR QUOTE

-------
 1    2012). Nowack et al. (2012) also note that a key research question is whether different modifications to
 2    nanomaterials and incorporation into different product matrices will have increased or decreased
 3    reactivity or toxicity relative to their pristine counterparts. There is also a lack of available data on the
 4    emissions of nanomaterials from products and releases to the environment, particularly under realistic
 5    conditions (Nowack et al., 2012). This is due in part to the absence of robust analytical techniques and
 6    instrumentation for accurately detecting and quantifying both emissions and environmental concentrations
 7    of nanomaterials (Nowack et al.. 2012) (see Additional Information Highlight Box 10).
 8           In each product life cycle stage in the CEA framework, "volume" and "release rate" are risk
 9    relevance factors that might be considered in risk assessments and risk management efforts of a material.
10    While these might be considered separately during future assessment and risk management efforts they
11    are discussed together for each area below. In general, workshop participants rated release rate more
12    highly than volume in terms of importance to consider in future risk assessments of MWCNTs. Experts
13    noted during the workshop that this is partially due to the fact that while production volume is important
14    for understanding the potential scale of impact (see Table 2-2 for information on current scale and
15    projected growth of MWCNTs production), release rate is what will ultimately determine the extent of
16    exposure. In other words, a large production volume might not cause  concern if release rate is relatively
17    small, but a smaller production volume could still cause concern if release rate is particularly high. Given
18    that very little information is  available for either topic and that each product life cycle area has such a
19    large influence on all other CEA areas, volume and release rate are particularly important to consider in
20    future risk assessment and risk management of MWCNTs.

      6.3.1.1. Material  Synthesis: Volume and  Release Rate
21           The workshop participants identified MWCNT material synthesis as important to risk assessment.
22    Risk relevance factors that might be considered in future assessment or risk management efforts for this
23    or other stages of the product life  cycle include volume and release rate. Participants most commonly
24    identified both of these factors as  important to risk assessment. Participants were generally not confident
25    or only somewhat confident in the ability of data on each factor to support risk management decisions
26    (see Priority Research Area Highlight Box in Section 2.2.2). Below are examples of experts' rationale  for
27    rating these areas as research priorities.
28               •   Material synthesis—Volume: Reporting from material synthesis is voluntary at this
29                   point, so little is known about MWCNT synthesis volume.
30               •   Material synthesis—Release rate: No universal reporting mechanism is  available, so
31                   little is known about MWCNT release rates. In addition, there is currently uncertainty
32                   about manufacturing factors that could have ecological consequences (e.g., surfactants,


      March 2013                                      6-9                    DRAFT-DO NOT CITE OR QUOTE

-------
 1                   not just active ingredients, should be tested to make experiments more realistic to real-
 2                   world scenarios.)
 3    The overarching influential factors identified by multiple experts for both volume and release rate
 4    characteristics included a variety of methods and techniques (e.g., control technologies, personal
 5    protective equipment, MWCNT synthesis and purity), ENM characteristics, factors associated with the
 6    surrounding media, as well as physical (e.g., flow regime, temperature, wind), chemical, and social
 7    conditions (e.g., acute, subchronic, and chronic exposure).
 8           No studies were identified regarding the volume of MWCNTs produced for use in flame-
 9    retardant textiles or the release rate during synthesis of MWCNTs for flame-retardant formulations (see
10    Table 2-2 for general information on production volume). As noted in Section 2.2.2.2. there are several
11    potential release scenarios during material synthesis, including recovery of the synthesized substance,
12    handling/packaging, equipment cleaning, and accidental release (e.g., fugitive leaks, equipment
13    malfunction, malfunctioning ventilation systems, exposure to fire and heat). In particular, during handling
14    and mixing of raw materials,  CNTs might become airborne under conditions where powder is being
15    handled, weighed, or mixed, although this will largely depend on the synthesis methods and use of control
16    technologies (SAFENANO. 2012): see Sections 2.2.2.2 and 6.3.3.1). In one study for example, Dahm et
17    al. (2011 a) found that some facilities typically handling large quantities (40 grams to  1 kg of MWCNTs
18    handled per day in powder form and aqueous form), still had measured releases above the NIOSH REL of
19    7 ug/m3 of concern, despite the use of enclosed processes or other forms of control technologies.
20    The authors note,  however, that their measurements were collected as inhalable mass concentration of
21    elemental carbon while the NIOSH REL is based on the respirable mass concentration of elemental
22    carbon, suggesting that measurements at primary facilities might not be appropriately characterized in
23    reference to this standard (see Section 6.3.3.1 for information regarding occupational exposure and use of
24    control technologies).
25           Though time did not allow for expert stakeholders participating in the CEA collective judgment
26    workshop on  MWCNTs  to develop specific research questions for material synthesis, the existing
27    information described above and in Chapter 2 point to several research areas that could  support future risk
28    assessment and risk management efforts, including:
29    •  What is the  median volume of MWCNT produced at manufacturing facilities?
30    •  What volume  of metal catalysts and support materials (e.g., aluminum, silica) are required for
31       manufacturing MWCNTs in current production facilities?
32    •  What volume  of water is used by MWCNT production facilities to clean equipment?
33    •  Does release rate vary during MWCNT synthesis using chemical vapor deposition, fluidized bed
34       chemical  vapor deposition, arc discharge or other methods?

      March 2013                                     6-10                   DRAFT-DO NOT CITE OR QUOTE

-------
 1    •   Are metal catalyst or support materials released during MWCNT synthesis?

      6.3.1.2. Material Processing: Volume and  Release Rate
 2            The workshop participants identified MWCNT material processing as important to risk
 3    assessment, and subsequently processing volume and release rate were both most commonly identified as
 4    important risk relevance factors. Participants were generally not confident or only somewhat confident in
 5    the ability of data on each factor to support risk management decisions (see Priority Research Area
 6    Highlight Box in Section 2.2.3). Below are examples of experts' rationale for rating these areas as
 7    research priorities.
 8               •   Material processing—Volume: While processing volume information may be available
 9                   to manufacturers and EPA regulators, and volume figures for decaBDE could be used to
10                   calculate potential worst case release scenarios, little information is available for
11                   scientists and the general public. Depending on the application of the product, there is
12                   great opportunity for exposure.
13               •   Material processing-Release rate: There is not enough data on releases from material
14                   processing, yet it is critical to determine release rate and exposure. Fabric coating
15                   operations, for example, are resulting in releases to the environment, so there is a need to
16                   understand the form of release (e.g., matrix bound, aggregate, etc.) and develop better
17                   analytical methods for carbon nanotube quantification.
18    The overarching influential factors identified by multiple experts for both material processing volume and
19    release rate characteristics included: methods and techniques (e.g., control technologies and MWCNT
20    purity), ENM characteristics, factors associated with the surrounding media (e.g., air, sediment, soil,
21    wastewater), as well as physical, chemical, biological and social conditions.
22            No additional data were identified for this area in revising the External Review Draft; however to
23    extend upon available data for MWCNT material processing, research planning efforts might consider
24    input from the CEA collective judgment workshop. Expert stakeholders discussed release rate during
25    material processing to identify potential risk scenarios that might occur during this stage of the product
26    life cycle along with specific research questions that, if pursued, might inform future assessment and risk
27    management efforts in this area (RTI. 2012). Based on those discussions, potential risk scenarios include:
28    (1) release into the air that results in worker exposure, and (2) MWCNT release into sewage treatment
29    plants (STP) that results in environmental exposures. To mitigate or avoid these potential risk scenarios
30    experts noted that risk management decisions could include the use of controls to minimize MWCNT
31    concentrations in occupational air or waste water. Experts noted that quantifying release rates in these
32    scenarios could inform the selection of appropriate control technologies. This type of quantification might
33    be carried out as part of an environmental or occupational assessment, according to experts. To support
      March 2013                                     6-11                   DRAFT-DO NOT CITE OR QUOTE

-------
 1    these types of measurements the following research questions were developed by expert participants
 2    along with an estimate of the resources and time to carry out the research:
      Table 64.     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           ino 000           1
      exposure. Evaluate control or personal protective equipment measures for reducing          '               ^ear
      exposure].1
      What is release rate in wastewater 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
      1Note: 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
 3            The workshop participants identified MWCNT product manufacturing as important to risk
 4    assessment, and subsequently volume and release rate at this stage were most commonly identified as
 5    important risk relevance factors. Participants were generally not confident or only somewhat confident in
 6    the ability of data to support risk management decisions related to either volume or release rate at this
 7    stage of the product life cycle (see Priority Research Area Highlight Box in Section 2.2.4). Below are
 8    examples of experts' rationale22 for rating these areas as research priorities.
 9            •   Product manufacturing—Volume: No universal reporting mechanism is currently available
10               to capture product manufacturing volume  of MWCNTs, thus the magnitude for potential for
11                release is not known (we currently assume decaBDE-product volumes to calculate worst case
12                scenarios). While confidential business information may be known to manufacturers and
13                available to EPA regulators, little information is available for scientists and the general
14               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.
      March 2013                                       6-12                   DRAFT-DO NOT CITE OR QUOTE

-------
 1           •   Product manufacturing—Release rate: There has not been enough study of manufacturing
 2               release rates, which are critical for determining exposure. Better analytical methods are
 3               needed for carbon nanotube quantification. There is a disconnect between EHS research and
 4               real-world ENM toxicity, characterizations, etc.
 5    The overarching influential factors identified by multiple experts for both product manufacturing volume
 6    and manufacturing release rate characteristics included: methods techniques (e.g., processing methods,
 7    personal protective equipment), ENM characteristics, factors associated with the surrounding media, as
 8    well as physical, chemical, biological and social conditions.
 9           Of all the  stages in the product life cycle, workshop participants placed the most importance on
10    product manufacturing, with 12 of 13 participants (92%) identifying it as being important to consider in
11    future risk assessments of MWCNTs. No studies were identified regarding the volume of MWCNT flame
12    retardant textiles produced, although Additional Information Highlight Box 2 notes that the scale is
13    relatively small and is primarily associated with the research and development stage.
14           Chaudhry et al. (2009) provides the most relevant information available regarding the material
15    synthesis (see Section 6.3.1) and product manufacturing stages of CNT textiles. The authors note that the
16    production of such textiles involves the use of fibers, yarn, fabrics, or finished products where the CNTs
17    will be embedded in the matrix via melting, mixing, co-extruding, granulating, or fixing/curing. Further,
18    CNTs could be released during these stages when powders are handled or CNTs are dispersed in
19    polymers (Chaudhry et al.. 2009); see Section 2.2.3.2. Similarly, SAFENANO (2012) noted that when
20    fibers are coated with CNTs, a suspension containing CNTs is likely to be applied to a textile either
21    through dipping or spraying. Dipping activities would be less likely to result in airborne CNTs and
22    therefore have less potential for exposure than spray applications. During spinning of fibers or weaving,
23    SAFENANO (2012) noted that CNTs could be released due to the high-energy mechanical abrasion of
24    the filaments as they are drawn, spun, and wound, but that released CNTs are not likely to be "free" in
25    this scenario. This premise is also supported by the  results of Takaya et al. (2012). Yarn coated with
26    MWCNTs was released during the weaving process without evidence of free MWCNTs in the respirable
27    dust (see Section 6.3.3.1 for more information regarding occupational exposure and use of control
28    technologies in the industry).
29           To build upon available data for product manufacturing of MWCNT flame retardant textiles,
30    research planning  efforts might consider input from the CEA collective judgment workshop. Expert
31    stakeholders participating in the workshop discussed release rate during product manufacturing to identify
32    potential risk scenarios that might occur during this stage of the product life cycle along with specific
33    research questions that, if pursued, might inform future assessment and risk management efforts in this
34    area (RTI. 2012). Similar to the material processing product life cycle stage, for product manufacturing,
3 5    workshop participants noted the possibility of release to air and subsequent worker exposure, as well as

      March 2013                                      6-13                   DRAFT-DO NOT CITE OR QUOTE

-------
      MWCNT release to sewage treatment plants that could result in environmental exposures (see Section

      6.3.2.2 for discussion on the Research Priority of the environmental transport, transformation and fate of

      MWCNTs in waste water). The following research questions related to these potential scenarios were

 4    developed by expert participants, 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.]1

      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
      occupational air based on manufacturing technique during: functionalization,              100,000           1 year
      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 distribution?      100,000           2 years

      1Note: 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

 6            The workshop participants identified MWCNT use as important to risk assessment, and

 7    subsequently use volume and release rate were most commonly identified as important risk relevance

 8    factors. Participants were generally not  confident or only somewhat confident in the ability of data on

 9    each factor to support risk management decisions (see Priority Research Area Highlight Box in Section

10    2.4). Below are examples of experts' rationale for rating these areas as research priorities.

11            •   Use—Volume: No universal reporting mechanism is currently available to capture volume of
12                use of MWCNTs, thus the magnitude for potential for release is not known (we currently
13                assume decaBDE-product volumes to calculate worst case scenarios). While confidential
14                business information may be known to manufacturers and available to EPA regulators, little
15                information is available for scientists and the general public.
      March 2013                                       6-14                    DRAFT-DO NOT CITE OR QUOTE

-------
 1           •   Use—Release rate: Better analytical methods are needed for carbon nanotube quantification,
 2               particularly for measuring consumer exposure to releases from furniture and other products,
 3               as well as quantifying and detecting ENMs in the environment.
 4    The overarching influential factors identified by multiple experts for both use volume and use release rate
 5    characteristics included: methods and techniques (e.g., MWCNT processing), ENM characteristics (e.g.,
 6    applied coatings, matrix bound vs. free, morphology, surface chemistry), factors associated with the
 7    surrounding media, as well as chemical, biological and social conditions (e.g., chronic exposure, life
 8    stage).
 9           No studies were identified regarding the volume of MWCNT use in flame-retardant textiles, as
10    this application is not widespread (see Additional Information Highlight Box 2). In the absence of data,
11    predictions must be made using known information about decaBDE flame-retardant textiles and textiles
12    in general. Release during use is primarily expected to be caused by the degradation of the product
13    matrix. For example, garments typically have  10% weight loss over the course of the life cycle due to
14    washing, ironing,  weathering, thermal degradation, and wear and tear (SAFENANO. 2012; Chaudhry et
15    al.. 2009). Other non-clothing textiles have been reported to lose between 5% and 20% of their weight
16    during normal use due to washing, exposure to heat, aging, and abrasion (GreBler et al., 2010). Very little
17    information is available regarding the likelihood of MWCNTs being released from textiles during normal
18    use, particularly in the specific application of flame-retardant textiles, but some evidence exists that
19    MWCNTs could become  airborne after the textile is burned and the char residue is mechanically
20    disturbed (Uddin and Nyden. 201 la: Nydenetal.. 2010) or during washing (Goncalves et al.. 2012).
21           Experts at the workshop noted that the release rate during product use was a particularly
22    important area for further investigation and in need of better analytical techniques due to its direct
23    influence on human exposure (RTI. 2012). Similar to material processing and product manufacturing,
24    workshop participants also noted that potential risk might arise in this product life cycle stage from
25    release of MWCNTs in indoor air or the environment (e.g., wastewater treatment), resulting in exposure
26    to consumers (see Sections 6.3.3.2 and 6.3.2.2 for discussion on Priority Research Areas: Consumer
27    Exposure and Wastewater, respectively). For this product life cycle stage they noted that regulations
28    pertaining to the type of matrix used with MWCNT flame retardants might be used to avoid or mitigate
29    potential risk. Based on their input, human health and ecological risk assessments could inform this type
30    of risk management decision. Research questions that experts identified to support conducting these types
31    of assessments, along with estimates of the resources and time to carry out the research, include:
      March 2013                                     6-15                   DRAFT-DO NOT CITE OR QUOTE

-------
      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 quantities (number and concentration) and         300,000         2-3years
      characterize (size distribution)?

      How does particle functionalization and matrix affect release in washing MWCNT
      textile products (use mini washing machines, measure quantities (number and            300,000         2-3 years
      concentration) and characterize (size distribution))?
      6.3.1.5. Disposal/ Recycling: Volume and Release Rate

 1            The workshop participants identified MWCNT disposal and recycling as important to risk

 2    assessment. Subsequently, the risk relevance factors of MWCNT disposal and recycling volume and

 3    release rate were both most commonly identified as important. Participants were generally not confident

 4    or only somewhat confident in the ability of data on each factor to support risk management decisions

 5    (see Priority Research Area Highlight Box in Section 2.5). Below are examples of experts' rationale for

 6    rating areas as research priorities.

 7            •  Disposal/recycling—Volume: It is unclear to what extent products containing MWCNTs will
 8               be recycled. This information is not available yet and may not be until the products are on the
 9               market. The potential for release from disposal and recycling operations and processes is
10               unknown. Currently estimates are based on decaBDE data. There is currently no system like a
11               national registry, however this step will be less critical for exposure and risk assessment than
12               previous steps.

13            •  Disposal/recycling—Release rate: Releases to air from incineration  and to water from
14               landfills are possible. Better analytical methods are needed for MWCNT quantification.
15               Release rate will  depend on specific recycling methods, which are still evolving, and there is
16               the potential for secondary products to emerge. Understanding release rate is critical to
17               estimating exposures. Large volumes of waste may accumulate and increase the potential for
18               a release event. Currently, no system like a national  registry exists. However, this step will be
19               less critical for exposure/risk assessment than previous steps. Research is needed on how
20               companies can dispose of ENMs in an environmentally conscious way.

21    The overarching influential factors identified by multiple experts for both  disposal and recycling volume

22    and use release rate characteristics included: methods and techniques, ENM characteristics, factors

23    associated with the surrounding media (e.g., air, wastewater, surface water), as well as physical, chemical,

24    biological and social conditions.
      March 2013                                      6-16                   DRAFT-DO NOT CITE OR QUOTE

-------
 1           As noted above, one reason experts rated disposal and recycling as important is that very little
 2    information is available about the release from textiles during these processes. Similarly, no information
 3    is available on the byproducts, metals, and other waste products that could result during these processes.
 4    Workshop participants noted that MWCNTs could be released into the environment during disposal and
 5    recycling, and that potential release could lead to exposures (e.g., release to wastewater that ends up in the
 6    environment; release to air resulting in potential occupational exposure).
 7           During recycling, textiles undergo various mechanical, thermal, and chemical treatments that
 8    could result in CNT release from the product matrix (SAFENANO. 2012). Industrial textiles are often
 9    reclaimed and recycled in specialized facilities while other types of textiles might be shredded and
10    repurposed as a part of the recycling process (e.g., polyester is often shredded, granulated into pellets, and
11    processed to recover the monomers and re-polymerized and processed by extruding, melting, spinning,
12    etc., into new fibers) (Chaudhry et al.. 2009). How MWCNT-treated flame retardant upholstery would be
13    recycled, and how the MWCNTs might be released from the product matrix as a result of the recycling
14    process, is unclear.
15           During disposal, textiles are expected either to be sent to a landfill or incinerated (SAFENANO.
16    2012; Chaudhry et al., 2009). Although no information regarding the current or predicted volume of
17    disposed MWCNT textiles was available, general quantities of disposed textiles were reported in
18    Chaudhry et al. (2009). For example, approximately 5.5 kg of textiles  per person per year are disposed of
19    and burned in Switzerland; in the UK, 75% of "fashion textiles" are landfilled. This information suggests
20    that the likelihood of disposal  of MWCNT flame-retardant textiles through landfill or incineration will
21    vary by country, but could be relatively high. The likelihood of CNTs  being released from the product
22    matrix during these processes partially depends on which disposal method is used. For example,
23    uncontrolled incineration could result in CNT release if the resulting char is mechanically disturbed
24    (SAFENANO. 2012: Uddin and Nvden. 201 la: Nydenetal.. 2010). whereas under controlled conditions
25    (>850°C), CNTs are likely to be destroyed and not released (SAFENANO. 2012: Nydenetal.. 2010).
26    However, no data are currently available to corroborate these predictions.
27           In building upon existing literature discussed above and in Chapter 2. research planning efforts
28    might consider research questions identified by expert stakeholders participating in the CEA collective
29    judgment workshop for MWCNTs (RTI. 2012). Based on their input, potential risk scenarios for the
30    disposal/ recycling stage of the product life cycle might include: (1) unknown or relatively large volumes
31    of materials at the end of life the product life cycle (e.g., incineration,  recycling, reuse, litter, landfill
32    disposal), (2) release during recycling for reuse (e.g., industrial shredding), resulting in release to the
33    environment (e.g., waste water) and subsequent environmental exposure, (3) occupational exposure from
34    MWCNTs released in air during recycling, and (4) release during sludge application to land that results in

      March 2013                                     6-17                  DRAFT-DO NOT CITE OR QUOTE

-------
 1    exposure to humans or other biota on the surface. To mitigate or avoid these potential risk scenarios,
 2    experts noted that risk managers might use a reclaim system to return upholstery and upholstered products
 3    to manufacturers at the end of the product life, or limit MWCNT production and use. Further, risk
 4    managers might need to evaluate the use of controls in occupational settings or regulation of sludge waste
 5    disposal. Life cycle analysis, and occupational exposure and/or environmental risk assessments could
 6    inform these types of risk management decisions. Research questions that participants identified to inform
 7    these types of assessments, along with estimates of the resources and time to carry out the research,
 8    include:
      Table 6-7.    Research Identified by RTI Workshop Participants: MWCNT Disposal / Recyling.

                                                                       Estimated Finances  Estimated Time
      Research                                                                 ($)             Frame
      How much volume of CNTs is used in upholstery?
      How much volume of CNTs is lost from upholstery during life span?                        mn 000           1
      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,        im nm           .
      number & mass concentration)?                                                ' uu'uuu            'year
      Survey the nanotechnology industry and municipal STP to gather mass of sludge/year
      applied to land: Is sufficient sludge being applied that uptake from plants, or exposure to          50,000            1 year
      farmers is possible?
      6.3.2. Environmental Transport, Transformation, & Fate
 9           The sections below discuss areas within environmental transport, transformation, and fate that
10    were considered to be Priority Research Areas by workshop participants. Relevant to this priority area,
11    OECD (2012) noted that research needed to inform risk assessment includes more robust data regarding
12    how nanomaterials move through different environmental and biological media, particularly in relation to
13    variation in physical-chemical properties (i.e., morphology, surface chemistry, size, functionalization).
14    OECD (2012) also identified mechanisms of bioaccumulation and predictive models for bioaccumulation
15    of nanomaterials as important to informing risk assessment, particularly because evidence suggests that
16    traditional relationships between octanol water partition coefficient (Kow) and bioaccumulation or
17    bioconcentration factors (BAF/BCFs) may not be applicable to carbon nanotubes (OECD. 2012; Petersen
18    etal..2011b).
      March 2013                                      6-18                   DRAFT-DO NOT CITE OR QUOTE

-------
      6.3.2.1. Air: Mobility, Persistence, and Bioavailability
 1            The workshop participants identified MWCNT air issues as important to risk assessment, and
 2    subsequently the risk relevance factors of MWCNT mobility, persistence, and bioavailability were most
 3    commonly identified as important. Participants were generally not confident or only somewhat confident
 4    in the ability of data on each factor to support risk management decisions (see Priority Research Area
 5    Highlight Box in Section 3.2). Below are examples of experts' rationale for rating these areas as research
 6    priorities.
 7            •  Mobility in air: Mobility in air is a primary route of exposure and existing data are
 8               insufficient.
 9            •  Persistence in air: MWCNTs may persist in air, yet data are currently insufficient. More
10               data are needed on persistence and degradation of carbon nanotubes in air and the availability
11               of MWCNTs in air for inhalation by humans and ecological receptors. Indirect effects of
12               decaBDE need to be considered.23
13            •  Bioavailability in air: While absorption across epithelial tissues has not been observed in
14               other organisms, data on the bioavailability of MWCNTs are  currently insufficient.
15            Participants at the workshop who identified fate and transport in air to be important also
16    voluntarily listed influential factors to include in developing research plans for this area, such as:
17    analytical techniques, MWCNT characteristics (e.g., aggregation state, persistence, surface chemistry),
18    and a variety of physical and chemical conditions.
19            To date, there is very little information available regarding the mobility, persistence, and
20    bioavailability of MWCNTs  in air. From an ecological toxicology perspective, Petersen et al. (201 Ib)
21    note that while a few studies estimate the release of CNTs to air (and other environmental media), as well
22    as the potential risks to ecological receptors, such models are limited by the lack of analytical techniques
23    needed to accurately detect and quantify CNTs in environmental matrices.
24            The issue of MWCNT bioavailability and characterization in air is also pertinent to human health,
25    as was noted by the workshop participants. Laboratory studies have often been hindered by the inability
26    to generate aerosolized MWCNT particles. Ahn et al. (2011) have used heat and sonication prior to
27    atomization to generate untangled MWCNTs in aerosol without the use of surfactants. These untangled
28    MWCNTs could then be used in in vivo toxicity models, unlike those often produced in previous studies
29    that also had tangled or clumped structures.  While no data was identified on the mobility, persistence, or
30    bioavailability of MWCNTs  in air under normal environmental conditions, these studies may provide
31    insight as to which physicochemical properties of MWCNT increase and decrease aerosolization
      23 Assumed to mean that indirect effects of decaBDE can inform research planning for MWCNTs.
      March 2013                                      6-19                   DRAFT-DO NOT CITE OR QUOTE

-------
 1    potential. Aersolization potential could in turn influence mobility, persistence, or bioavailability.
 2    Information gaps persist, however, regarding the influence of environmentally relevant conditions on
 3    aerosolization potential and other aspects of MWCNT behavior in air.
 4           In extending upon existing data for mobility, persistence and bioavailability of MWCNTs in air,
 5    research planning efforts might consider input from the CEA collective judgment workshop on
 6    MWCNTs. Expert stakeholders participating in the workshop discussed persistence and mobility in air to
 7    identify potential risk scenarios that might occur in this environmental spatial zone, along with specific
 8    research questions that, if pursued, might inform future assessment and risk management efforts in this
 9    area (RTI. 2012). Based on their input, potential risk might arise from the persistence of CNTs released to
10    occupational or ambient air resulting in longer residence times that increase the probability of exposure.
11    Similarly, mobility dictates the extent of potential MWCNT exposure in ambient air, and thus can
12    influence risk. To mitigate or avoid potential exposure in occupational or ambient air, experts noted that
13    risk mangers might reduce MWCNT residence time in air (e.g., by increasing aggregation potential or
14    decreasing mobility and retainment), or limit MWCNT production and use. To inform these types of risk
15    management decisions, experts suggested carrying out human health risk assessments, exposure
16    assessments quantifying MWCNTs in air, and cost benefit analyses. To support conducting these types  of
17    assessments expert stakeholders recommended the following research areas, along with estimates of the
18    resources and time to carry out the research:
      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.
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
      March 2013                                     6-20                   DRAFT-DO NOT CITE OR QUOTE

-------
      6.3.2.2. Wastewater: Mobility, Persistence, and Bioavailability
 1            The workshop participants identified the environmental transport, transformation, and fate of
 2    MWCNTs in wastewater as important to risk assessment. Subsequently, the risk relevance factors of
 3    mobility, persistence, and bioavailability in wastewater were each most commonly identified as important
 4    to risk assessment. Participants were generally not confident or only somewhat confident in the ability of
 5    data on each factor to support risk management decisions (see Priority Research Area Highlight Box in
 6    Section  3.3.3). Below are some examples of experts' rationale for rating these areas as research priorities.
 7            •   Mobility in wastewater: The extent to which MWCNTs move in wastewater determines
 8               which environmental compartments will be exposed (e.g., water, soil, sediment). Waste water
 9               is the most likely route into the environment; more research is needed on how ENMs get out
10               of the waste water stream and how to prevent this.
11            •   Persistence in wastewater: MWCNTs might be persistent in wastewater and the potential
12               for MWCNT transformation and subsequent effects of transformation are unknown.
13            •   Bioavailability in wastewater: MWCNTs are potentially persistent and likely to interact
14               with activated sludge given results from studies with  other organisms.
15    The overarching influential factors identified by multiple experts  for all three characteristics included:
16    analytical techniques, control technologies, MWCNT purity, a variety of MWNCT characteristics
17    (including but not limited to aggregation/agglomeration state, applied coatings, size distribution, surface
18    chemistry, water solubility/dispersibility), factors associated with the surrounding media (particularly
19    wastewater and sediment), chemical conditions (ionic strength, natural organic matter, other contaminants
20    in the environment, and salinity), and biological conditions (microbial communities in the environment).
21            In the introduction to a recent special Environmental Science and Technology issue on
22    Transformations of Nanomaterials in the Environment, Plata et al. (2012a) outlined some of the
23    challenges involved in characterizing nanomaterials (in particular, CNTs) in water and sediment and some
24    promising analytical techniques that would help increase understanding of MWCNT mobility,
25    persistence, and bioavailability in wastewater. Some of the advances in analytical methods to detect
26    MWCNTs in aqueous systems, such as wastewater, are mentioned in Additional Information Highlight
27    Box 10. Research on analytical methods serves as a basis to begin to assess MWCNT mobility,
28    persistence, and bioavailability in wastewater (and in other environmental matrices).
29            Recent studies of MWCNT transport in water-saturated porous media might help inform
30    understanding of MWCNT mobility in wastewater. Wang et al. (2012) observed that 75% of
31    functionalized MWCNTs delivered through a water-saturated sand system were detected in effluent.
32    The authors also concluded that MWCNTs with lengths greater than 8 um were more likely to be
33    deposited than smaller MWCNTs. Although wastewater systems  are unique, they often involve the use of
      March 2013                                     6-21                  DRAFT-DO NOT CITE OR QUOTE

-------
 1    filters such as sand. These results suggest that functionalized MWCNTs might be mobile during
 2    wastewater treatment processes, with longer MWCNTs exhibiting less mobility than shorter MWCNTs.
 3           With respect to the potential transformation of MWCNTs in wastewater, a recent paper by
 4    Nowack et al. (2012) concluded that released CNTs could enter wastewater (and other environmental
 5    compartments) where they could be transformed by photochemistry, oxidation, adsorption of natural
 6    organic matter and other organic colloids, biotransformation, and continued abrasive forces (Nowack et
 7    al.. 2012). The authors also noted that transformations could change CNT aggregation, dispersibility, and
 8    interaction with biota in the environment (Nowack et al.. 2012). These conclusions highlight the
 9    importance of studying not only MWCNTs, but also potential environmental transformations of
10    MWCNTs to understand the dynamics driving mobility, persistence, and bioavailability in wastewater.
11           To that end, expert stakeholders at the workshop discussed MWCNT persistence and mobility in
12    wastewater to identify the type of risks that might arise in this area of the environment and develop
13    specific research questions that could support future assessment and risk management efforts of the
14    material (RTI. 2012). They noted that MWCNTs could be released in either pulse  industrial  discharges or
15    sewers, or in semi-continuous loadings from industrial, commercial, and residential wastewater with
16    flame-retardant materials. To mitigate or avoid potential risks from these scenarios, experts identified
17    several types of decisions risk managers might consider, including: utilizing pretreatment controls to
18    prevent MWCNT discharge from industrial facilities, or regulating efficiencies of control technologies to
19    mitigate MWCNT release into the environment. Information to support making these types of risk
20    management decisions might include analytical measurements made during ecological risk assessments,
21    or evaluations completed during life cycle analyses.
22           According to expert participants, research that could inform these and other types of assessments
23    in this area might include the  following, along with resource and time estimates to carryout the research:
      March 2013                                     6-22                   DRAFT-DO NOT CITE OR QUOTE

-------
     Table 6-9.    Research Identified by RTI Workshop Participants: MWCNT Mobility & Persistence
               in Waste Water.
Research
Estimated
Finances ($)
Estimated Time
Frame
     How does the degree of functionalization and changes in wastewater treatment
     processes (e.g., activated sludge, disinfection processes) affect the rate of
     transformation?
     [Evaluate the rate of transformation of MWCNTs alone and in a product matrix.]
                         400,000
                 3 years
How to extract and characterize MWCNTs from suspended and fixed biomass or treated
effluent with minimal modifications to surface group, functionalization, 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.]


400,000

300,000

250,000

600,000 with
pilot plant

300 000


600 000


3 years

3 years

2.5 years

4 years with pilot
plant

3 years


4 years for new
methods
     1Note: 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
1            The workshop participants identified environmental transport, transformation and fate of
2    MWCNTs as important to risk assessment. Subsequently, the risk relevance factors of mobility,
3    persistence, and bioavailability in sediment were each most commonly identified as important to risk
4    assessment. Participants were generally not confident or only somewhat confident in the ability of data on
5    each factor to support risk management decisions (see Priority Research Area Highlight Box in Section
6    3.3.1). Below are some examples of experts' rationale for rating these areas as research priorities.
     March 2013
6-23
DRAFT-DO NOT CITE OR QUOTE

-------
 1           •   Mobility in sediment: Sediment is a likely ultimate repository for MWCNTs and the extent
 2               to which MWCNTs can be redistributed to the water column through sediment disruption is
 3               unclear.
 4           •   Persistence in sediment: Degradation of MWCNTs is likely to be slow, if it occurs at all.
 5               More research is needed on transformations in the environment and interactions between
 6               microbes and soil.
 7           •   Unavailability in sediment: MWCNTs are potentially persistent, however uptake was not
 8               observed in several recent studies.
 9    The overarching influential factors identified by multiple experts for all three characteristics of MWCNTs
10    in sediment included: a variety of MWNCT characteristics (e.g., aggregation/agglomeration state and
11    surface chemistry) and chemical conditions (particularly natural organic matter).
12           In a recent review paper, Petersen et al. (20 lib) noted that MWCNTs sorb more readily to
13    sediments in seawater, but will tend to stay in the water column in aquatic systems with high
14    concentrations of dissolved organic matter (DOM). The authors concluded that future studies of the
15    subsurface mobility of CNTs (including mobility in sediment) should examine a larger range of porous
16    media size, mineralogy, aqueous chemistry (including DOM), and natural soils (e.g., clays, silts, peats).
17    The authors also stressed that the influence of CNT functionalization and surface properties on transport
18    are critical research areas.
19           Some workshop participants stated that MWCNTs are likely to persist in sediment. While
20    previous studies indicated that carboxylated SWCNTs (but not pristine SWCNTs) can be transformed by
21    soil enzymes (Allen et al., 2009; Allen et al.. 2008a). which suggest that similar processes could occur in
22    sediment, more  updated and sediment-specific research needs to be conducted to confirm that MWCNTs
23    could biodegrade in sediment.
24           In a recent paper examining the effect of MWCNTs on bioaccumulation of polycyclic aromatic
25    hydrocarbons (PAHs) by Chironomusplumosus larvae in sediment, Shen et al. (2012) concluded that
26    MWCNT-associated PAHs may have been absorbed by larvae and hypothesized that CNTs could
27    increase the exposure risk of PAHs to benthic organisms due to their unique structure. These results
28    suggest that uptake of MWCNT by organisms could potentially occur in sediment. Some previous
29    research, however, has not found substantial uptake of CNTs by sediment-dwelling earthworms (Petersen
30    et al.. 2008) or soil-dwelling earthworms (Petersen et al.. 201 la: Petersen et al.. 2009). These differing
31    conclusions support the designation of MWCNT bioavailability in sediment as a research priority.
32           While time did not allow for expert stakeholders at the workshop to specifically  discuss
33    bioavailability in sediment, they did consider persistence in sediment (RTI. 2012). From this discussion,
34    the relative persistence of MWCNTs could influence the potential risk associated with instances when
35    there is continuous deposition and burial of fiber or polymers  containing MWCNTs  from multiple sources
      March 2013                                     6-24                   DRAFT-DO NOT CITE OR QUOTE

-------
 1    (e.g., air, waste water discharge, release from products, storm water). To mitigate or avoid potential risks
 2    influenced by the persistence of MWCNTs in sediment, experts noted that risk managers might consider
 3    regulating the efficiencies of control technologies that can minimize MWCNT concentrations in the
 4    environment, or regulating loadings of CNTs in fabrics. While specific assessments that could inform
 5    these types of risk management decisions were not identified by experts, such information might include:
 6    measurements of MWCNTs in environmental compartments in ecological risks assessments, as well as
 7    measurements to quantify release of MWCNTs from textile production. Experts identified the following
 8    research directions, which might support conducting these and other types of assessments,along with
 9    estimates of the resources and time to carry out the research:
      Table 6-10.   Research Identified by RTI Workshop Participants: MWCNT Persistence in Sediment.

                                                                           Estimated     Estimated Time
      Research                                                             Finances ($)        Frame
      How does the degree of functionalization or changes in sediment affect the rate of
      transformation?
      Is there long term persistence? Are MWCNT released from fabrics and sediments?     300,000-600,000     3-5 years
      [Evaluate both the rate of release from fibers as well as the transformation of MWCNT
      alone and combined in a matrix.]1
      How to extract and characterize MWCNT from sediment with minimal modifications to      .-,, -,,,,          „
        ,        ........        , ,    , .     ,   ,.                       400,000          3 years
      surface group, functionalization, impregnated  metals, and coatings.
      What are the transformation byproducts from  MWCNT and flame resistant fibers?
      What byproducts could be formed during degradation processes and would this be      Not provided      Not provided
      impacted by the MWCNT concentration  in the fabrics?
      1Note: 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)
10           The following exposure routes and dose/kinetic topics were considered to be Priority Research
11    Areas according to workshop participants.

      6.3.3.1. Exposure Route- Human Occupational: Ingestion, Inhalation, Dermal
12           Human occupational exposure was identified as important to risk assessment by the workshop
13    participants. Risk relevance factors that might be considered in risk assessment or management of
14    occupational exposure include: ingestion, inhalation, and dermal routes of exposure. Although experts
15    strongly agreed that occupational exposure is important, the way they rated the importance of each risk

      March 2013                                      6-25                   DRAFT-DO NOT CITE OR QUOTE

-------
 1    relevance factor, varied. Overall, they were not confident or only somewhat confident that the current data
 2    could support risk management decisions for each factor (see Priority Research Area Highlight Box in
 3    Section 4.2.1). Below are some examples of why experts' rated these areas with varying levels of
 4    importance and confidence.
 5           •   Human occupational exposure—Ingestion: Experts do not see much concern from
 6               ingestion of MCs24 based on data; there will likely be good controls to limit exposure via
 7               ingestion. From the exposure amount perspective, the relative amount of CNT intake in the
 8               form of ingestion would be much lower than inhalation in occupational settings and there are
 9               some studies presenting our luminal surfaces of GI tracks are resistant to the passage of
10               CNTs.
11           •   Human occupational exposure—Inhalation:  Inhalation is thought to be a likely first type
12               and some say the most important type of human occupational exposure; inhalation exposure
13               must be controlled as it can cause effects, including indirect effects in children.  Some animal
14               studies have been conducted but exposure routes and administration techniques have been
15               criticized.
16           •   Human occupational exposure—Dermal: Dermal exposure is less likely  an issue based on
17               the known physicochemical properties of CNTs. A low dermal absorption rate is expected
18               and there would likely be controls to limit dermal exposure, however some  studies in insects
19               have shown some systematic effects (however other routes, inhalation and ingestion, were not
20               completely blocked).
21    The overarching influential factors identified by multiple experts for human occupational exposure
22    characteristics included: methods and techniques (e.g., processing and synthesis methods, personal
23    protective equipment), ENM characteristics (e.g., applied coatings, morphology, persistence), factors
24    associated with the surrounding media (e.g., air), as well as physical, chemical, biological and social
25    conditions (e.g., chronic exposure, occupation).
26           MWCNTs are found in facilities ranging from research laboratories and production  plants to
27    those where they are processed, used, disposed, and recycled, and the limited literature available suggests
28    potential for worker exposure in at least some of these types of facilities [(Dahm et al.. 201 la: Johnson et
29    al.. 2010; Lee etal.. 2010: Han et al.. 2008): see Sections 4.1.2.4. 4.2.1. and 4.2.5)1. The extent of worker
30    exposure to MWCNTs, however, is not well understood. No published literature was available
31    investigating occupational exposure to MWCNTs in flame-retardant textiles specifically and no more
32    recent literature regarding exposure potential was identified after the CEA workshop process. OECD
33    (2012) noted that research needed to inform risk assessment includes obtaining robust data on the
34    exposure of workers at all stages of the life cycle from material synthesis to disposal and recycling
35    facilities.
      24Word choice of expert; assumed to mean multiwalled carbon nanotubes
      March 2013                                     6-26                   DRAFT-DO NOT CITE OR QUOTE

-------
 1           While no information was identified specific to MWCNTs in flame retardant textiles, literature
 2    pertaining to similar applications is available. For example, Takaya et al. (2012) reported that
 3    occupational workers could be exposed to respirable particles of yarn coated with MWCNTs during the
 4    weaving process of the production of a conductive fabric. The authors concluded that the mechanical
 5    force of weaving with the MWCNT-coated yarn was sufficient to break the coating and release yarn
 6    fibers still embedded with MWCNT, but is not likely to provide sufficient mechanical action to release
 7    individual MWCNTs from the coating layer [(Takaya et al.. 2012): see Additional Information Highlight
 8    Box 4]. Similar exposure scenarios can be imagined with MWCNTs  in flame retardant upholstery textiles,
 9    but how use of MWCNTs as coatings on upholstery textiles compared to yarns would affect the release
10    and subsequent exposure is unclear.
11           Several publications by Schubauer-Berigan et al. (2011) (see Table 2-2) and Dahm et al. (20lib:
12    201 la) examined the current industry of engineered carbonaceous nanomaterial to better characterize
13    potential occupational exposure and use of engineering controls. While little information was provided
14    specific to MWCNT and no information was provided specific to flame-retardant textiles, the data
15    indicate that many companies employ various  engineering controls (including use  of LEV, HEPA filters,
16    enclosed production processes and safety cabinets) for production, laboratory procedures, and research
17    and development operations (Dahm etal.. 201 Ib). Many companies also use health and safety training,
18    good hygiene practices, and other practice or administrative methods to reduce occupational exposure.
19    However, nearly one in four companies surveyed either did not report using respiratory protection or were
20    using an ineffective form of respiratory protection (Dahm et al.. 20 lib). Similarly, one in seven
21    companies not reporting respiratory protection stated that such protection was not needed because
22    operations were fully enclosed; however, the authors note that NIOSH recently recommended the  use of
23    respirators even when processes are enclosed if measurement data suggest that the nanomaterial release is
24    not well controlled (Dahm  et  al.. 20 lib). Results of these studies suggest that while many companies
25    employ several forms of protective measures, there is still room for improvement to better characterize
26    and mitigate potential occupational exposure. There is a great need for recommended exposure limits for
27    different formulations of MWCNTs and evaluation criteria for assessing the release possibilities of
28    individual MWCNTs. In 2010 NIOSH suggested an 8-hour REL of 7 ug/m3 for carbon nanotubes  and
29    nanofibers but that value has not yet been finalized (NIOSH. 2010). Similarly, OECD is currently
30    developing standard test methods for MWCNT, but this process is complicated by the lack of an accepted
31    "representative" MWCNT that could be broadly applicable to other MWCNT formulations (Takaya et al..
32    2012).
33           To develop a better understanding of human occupational exposures, research planning efforts
34    could incorporate input from the CEA collective judgment workshop on MWCNTs.  Expert stakeholders

      March 2013                                      6-27                   DRAFT-DO NOT CITE OR QUOTE

-------
 1    participating in the workshop discussed inhalation exposure in occupational settings to identify the type
 2    of risks that might arise in this area of the environment and develop specific research questions that could
 3    support future assessment and risk management efforts of the material (RTI. 2012). Based on their
 4    discussion, potential risks could arise due to inhalation of the material (pristine MWCNTs, functionalized
 5    MWCNTs, or otherwise modified MWCNTs) during any part of the manufacturing process. To mitigate
 6    or avoid potential risk associated  with occupational inhalation exposures, experts noted that risk managers
 7    might consider the use of engineering controls, personal protective equipment, or if necessary, banning
 8    the material outright. Experts noted that information from assessments that might inform these types of
 9    risk management decisions include a NOEL or LOEL for the relevant material and other data from
10    occupational exposure assessments. To carry out these types of assessments, experts recommended the
11    following research areas, along with financial and time estimates:
      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.
                                                                  2 million
                         3 years
                         (2 year rodent bioassay &
                         an additional year for setup
                         and anlaysis)
      Analytical and rodent studies to examine effect of co-factors (e.g.,
      solvents, resins) on particles size, deposition, translocation, and removal.
                                                                  2 million
                         3 years
                         (2 year rodent bioassay &
                         an additional year for setup
                         and anlaysis)
      6.3.3.2. Exposure Route - Human Consumer: Ingestion, Inhalation, Dermal
12           Human Consumer exposure was identified as important to risk assessment by the workshop
13    participants. Risk relevance factors that might be considered in risk assessments or management efforts in
14    this area include: ingestion, inhalation, and dermal routes of exposure. Participants rated the importance
15    of each of these risk relevance factors differently, but were generally not confident or only somewhat
16    confident in the ability of data on each route to support risk management decisions (see Priority Research
17    Area Highlight Box in Section 4.2.2). Below are examples of why experts rated these topics at varying
18    levels of importance and confidence.
      March 2013
6-28
DRAFT-DO NOT CITE OR QUOTE

-------
 1           •   Human consumer exposure—Ingestion: Currently there is not much concern from
 2               ingestion based on data; however, more analytics are needed to determine risk of human
 3               consumer exposure via ingestion.
 4           •   Human consumer exposure—Inhalation: The risk of human consumer exposure via
 5               inhalation depends on release rate; if it is high enough it could become an issue. There is a
 6               need to understand persistence and how upholstery exposure affects consumers.
 7           •   Human consumer exposure—Dermal: The risk of human consumer exposure via a dermal
 8               route depends on dermal absorption rates. Need to understand potential exposure to children,
 9               especially potential impacts from crawling around on floor.
10    The overarching influential factors identified by multiple experts for all three exposure routes included:
11    analytical techniques, control technologies, MWCNT purity, personal protective equipment, a variety of
12    MWNCT characteristics (e.g., aggregation/agglomeration state, applied coatings, persistence, size
13    distribution, surface chemistry, water solubility/dispersibility), factors associated with the surrounding
14    media (particularly air and wastewater), chemical conditions (dispersing agents, ionic strength, salinity,
15    other contaminants in the environment, surfactants in a lab study), and social conditions (exposure
16    duration, human activity, life stage, susceptible populations).
17           Similarly, OECD (2012) noted that research needs to inform risk assessment include improved
18    characterization of consumer exposure, including the concentration in and release from consumer
19    products. Although the likelihood of MWCNT release from polymers, textiles, and other product matrices
20    during normal product use is not well understood, Nyden et al. (2010) provides some information on the
21    potential release of nanoparticulate additives during incineration of polyurethane foam (PUF)—which is
22    relevant in the context of MWCNT use as flame-retardant additives because behavior during incineration
23    is an important consideration for potential consumer exposure. This study specifically investigated the
24    release of carbon nanofibers (CNFs), but identifies mechanisms of release that might be applicable to
25    MWCNTs. In a controlled well-ventilated flame test, the authors  found evidence of CNFs in the char on
26    the surface  of the foam after burning but not in the smoke content, indicating that CNFs were not released
27    into the air  during this process. The researchers noted that CNF release to the environment might still be
28    possible under different combustion conditions (e.g., under-ventilated conditions that might decrease the
29    destructive  forces of the flame on the CNFs). The authors also identified released submicron particles
30    when the char residue was mechanically disturbed, but could not  definitively attribute the particles to the
31    CNFs (Nyden et al.. 2010). A later publication by the same authors suggested that the aerosolized
32    particles from the disturbed char were likely CNF bundles partially encapsulated in a thin layer of charred
33    PUF after spectroscopic measurements (Uddin and Nyden. 201 la). The authors concluded that the major
34    potential hazard for CNF exposure during well-ventilated combustion resulted from the disturbance of
35    residual char rather than from aerosolized CNF in smoke (Uddin  and Nyden. 201 la).
      March 2013                                     6-29                   DRAFT-DO NOT CITE OR QUOTE

-------
 1           Many of the available MWCNTs studies involve polymer matrices in which the MWCNTs are
 2    embedded, but in this specific application of flame-retardant textiles, MWCNTs likely would be
 3    incorporated through dipping, dyeing, thermal fixation, and other methods (as described in Chapter 1).
 4    Nevertheless data show that, in addition to the polymer matrix and the application technique, the chemical
 5    properties of MWCNTs will be important in determining the likelihood that MWCNTs leave the product
 6    matrix during washing. Goncalves et al. (2012) investigated the extent to which acidic or basic MWCNTs
 7    were "washed out" of polyester and cotton textiles. The results indicated that the more acidic MWCNTs
 8    were less likely to be removed from the matrix, particularly in the polyester textile. Although removal
 9    from the textile under realistic conditions (e.g., the washing machine) might not present much opportunity
10    for consumer exposure given that upholstery textiles are unlikely to washed frequently, these data do
11    suggest that acidic MWCNTs might be preferable to incorporate in flame-retardant textiles to reduce
12    potential release (see Section 2.4.2).
13           While expert stakeholders participating in the CEA collective judgment workshop did not focus
14    on consumer exposure due to time constraints, the results from recently identified literature do point to
15    several questions that are currently not addressed by the available literature. For example, if MWCNTs
16    can be removed from the textile matrix during washing, how frequently are MWCNT flame-retardant
17    textiles likely laundered or hand washed? Could consumers be dermally exposed to MWCNT particles
18    from handling the laundered textiles or be exposed to airborne MWCNTs after drying of the textiles?
19    How do differences in surface chemistry and textile type influence this exposure potential? How do
20    differences in cleaning practices (e.g., machine washing versus hand washing versus spot cleaning of
21    upholstery) influence potential exposure? To what extent do other activities (e.g., chewing on textiles)
22    lead to consumer exposures?

      6.3.3.3.  Dose (Kinetics) - Human: Absorption, Distribution, Metabolism, and Excretion
23           Human dose/kinetics was identified as important to risk assessment by workshop participants.
24    Factors that risk assessors or managers might consider in this area include absorption, distribution,
25    metabolism, and excretion. Expert workshop participants identified each factor as important, with the
26    exception of distribution, which they identified as possibly important. Participants were generally not
27    confident or only somewhat confident in the ability of data to support risk management decisions related
28    to each factor (see Priority Research Area Highlight Box in Section 4.2.6). Below are examples of
29    experts' rationale for rating these areas  as research priorities.
      March 2013                                     6-30                   DRAFT-DO NOT CITE OR QUOTE

-------
 1            •   Dose (Kinetics)—Human Absorption: Except for the inhalation route, absorption needs to
 2               occur to have an effect; need to know if ENMs are bioavailable. Concern is significantly
 3               reduced if oral or dermal absorption does not occur after exposure; data are currently
 4               insufficient.
 5            •   Dose (Kinetics)—Human Distribution: Distribution within tissues and specific organs
 6               could be important to determine risk; data are currently insufficient.
 7            •   Dose (Kinetics)—Human Metabolism: Data are currently insufficient.
 8            •   Dose (Kinetics)—Human Excretion: Data are currently insufficient.
 9    As note above, at least one participant recognized that this area is a priority for research because currently
10    available data on all four components of human toxicokinetics are insufficient. Importantly, concerns for
11    risk through routes other than inhalation would be significantly reduced if it was determined that oral and
12    dermal absorption does not occur. Experts developing specific research questions related to MWCNT
13    absorption noted that although little evidence is available demonstrating dermal absorption (via abraded
14    skin), additional work should be considered due to the potential for high exposure, especially in children.
15    Additionally, distribution within tissues and organ, in conjunction with data on mode of action and
16    toxicity, could play a key role in increasing or limiting risk.
17            There was agreement among experts that methods/techniques were influential to understanding
18    toxicokinetics of MWCNTs, including analytical techniques, MWCNT purification methods, processing
19    and synthesis methods, and control technologies and personal protective equipment. Several
20    characteristics of the MWCNT formulation were also deemed by multiple experts as influential factors for
21    all toxicokinetic components, including aggregation/agglomeration state, matrix bound vs. free form,
22    persistence, size/size distribution, and surface chemistry. Several experts also noted the importance of
23    applied coatings, morphology, and  adsorption/desorption ability for adsorption and distribution, as well as
24    the importance of water solubility/dispersibility for distribution  and elimination. Finally, several experts
25    noted the importance of social conditions on toxicokinetics: Exposure length (acute, subchronic, chronic),
26    exposure route, human activity, life stage, and individual susceptibility can all influence the human
27    toxicokinetic processes for MWCNTs.
28            Generally, inhaled nanomaterials—including CNTs—are assumed to be more toxic than
29    conventional-sized materials, in part because they can be inhaled more deeply into the lung (resulting in
30    longer residence times and greater particle-cell interactions), have a smaller size distribution (so are more
31    readily internalized by individual cells and more easily migrate through the body), and have large  surface
32    area-to-mass ratios resulting in more reactivity (Bakand et al., 2012). Absorption, distribution,
33    metabolism, and excretion are key processes that determine how the administered or received dose of
34    MWCNTs differs from the internal dose that reaches a target organ or tissue. Toxicokinetic processes thus
35    play a key role in determining toxic potential.

      March 2013                                     6-31                   DRAFT-DO NOT CITE OR QUOTE

-------
 1           Yet, little is currently known about the specifics of absorption, distribution, metabolism, and
 2    excretion of MWCNTs (see Section 4.2.6 and Additional Information Highlight Box 9). Pathways of
 3    CNT uptake into cells, intercellular trafficking, and distribution are not well characterized despite
 4    multiple studies on the subject (Al-Jamal et al.. 2011). Additionally, toxicokinetics could vary with
 5    different MWCNT formulations or cell type. For example, Al-Jamal et al. (2011) recently showed that
 6    MWCNTs functionalized with an ammonium group and positive charge (NH3+) could be internalized by
 7    human lung epithelial cells by three different mechanisms (membrane wrapping, direct membrane
 8    translocation, and clustering within vesicular compartment), whereas Zhang et al. (2012) showed that
 9    uptake of oxidized MWCNTs by human epithelial cervical cancer cells occurred through nonspecific
10    cellular uptake. Jain et al. (2011) found that biodistribution of acid-oxidized MWCNTs (which contain
11    surface carboxyl groups) in mice was dependent on the  density of functionalization, and this
12    physicochemical characteristic particularly influenced clearance of MWCNTs from reticuloendothelial
13    systems such as liver, spleen, and lungs. MWCNTs with shorter lengths and higher degrees of oxidation
14    (therefore greater density of functionalization) quickly partitioned to the kidney but were rapidly excreted
15    through the renal system; longer, less functionalized MWCNTs and pristine MWCNTs preferentially
16    accumulated in the liver rather than the kidney and were more likely to be excreted in the feces through
17    biliary pathways. The authors suggest that this functionalization-dependent distribution between organs
18    and excretion patterns might also explain why pristine and less-oxidized MWCNTs do not demonstrate
19    nephrotoxicity in subchronic studies.
20           A greater understanding of physicochemical properties influencing internal dose (i.e., particle
21    kinetics in biological systems) and therefore biopersistence and bioaccumulation was also noted by
22    OECD (2012) as particularly important research needed to inform risk assessment since this information
23    is useful in interpreting toxicological results. Similarly,  additional research on appropriate dose metrics is
24    needed to inform risk assessment based on evidence suggesting that particle surface area or number
25    concentration, rather than standard mass concentration,  may be more appropriate for nanomaterials
26    (OECD. 2012).
27           To continue building upon existing literature for human toxicokinetics summarized in Chapter 4.
28    as well as the newly identified the sources discussed above, research planning efforts might consider input
29    from the CEA collective judgment workshop on MWCNTs. Experts participating in the workshop
30    discussed human absorption, metabolism, and excretion to identify potential risk scenarios and specific
31    research questions that could inform future assessment and risk management efforts (RTI. 2012). Based
32    on their discussion, potential risks related to MWCNT toxicokinetics might include the potential for
33    absorption in the lungs or GI tract, degradation into a more toxic metabolite, and bioaccumulation of the
34    material due to lack of excretion. To mitigate or avoid potential risks related to MWCNT toxicokinetics in

      March 2013                                     6-32                   DRAFT-DO NOT CITE OR QUOTE

-------
1    humans, experts noted that risk managers might consider: implementing appropriate control technologies

2    to minimize exposures, minimizing the absorption potential of MWCNTs, or limiting MWCNT

3    production or use. Information to inform these types of risk management efforts could come from: human

4    health risk assessment, occupational exposure assessment, and a cost benefit analysis. To support

5    conducting these and other types of assessments, experts recommended the following areas of research,

6    along with estimates of the resources and time to carry out the research:
     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         5 million          5 years
     that decrease absorption while maintaining beneficial uses).
     [Examining all three absorption processes increases the potential for discovery of
     unique interactions among systems.]1

     Develop analytical techniques for measuring the original MWCNT or metabolites in
     cells.
     [Evaluate the degradability of the relevant MWCNT material compared to original           275,000          2 years
     MWCNT. Determine of the half-life of relevant MWCNT material in biological
     systems.]

     Measuring the original MWCNT or metabolites in tissues after whole body inhalation        „........          „
     exposures.                                                                   suu.uuu          i 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.]

     Develop tracer methodology to detect excretion by-products of the relevant MWCNT
     material to enable:                                                             300,000          2 years
     [Quantify levels of by-products in the body.]

     1Note: Information in brackets [ ] denotes details extracted from responses experts provided along with the question or research area.
     March 2013                                         6-33                    DRAFT-DO NOT CITE OR QUOTE

-------
      6.3.4. Impacts
 1           The following impact areas were considered to be Priority Research Areas according to workshop
 2    participants.

      6.3.4.1.  Human: Cancer and Non-Cancer
 3           Experts in the workshop generally rated human health impacts as important to consider in future
 4    risk assessments of MWCNTs. Participants most commonly identified human cancer impacts as
 5    important and generally somewhat confident in the availability and utility of current data to support risk
 6    management decisions; however, those who identified human non-cancer impacts as important were
 7    generally not confident in the current data. Listed below are examples of experts' reasons for choosing
 8    these topics as Research Priority Areas.
 9           •   Human Cancer: Data are currently insufficient. There is a lack of cancer studies done on
10               inhalation exposure, particularly occupational inhalation exposure.
11           •   Human Non-cancer: Data are currently insufficient.
12    Those who chose to note influential factors generally agreed that MWCNT characteristics (e.g., surface
13    coatings, size, morphology) and social conditions (e.g., exposure duration, exposure route, occupation,
14    individual or population susceptibility) were of particular importance for future assessments of human
15    cancer and non-cancer effects.
16           OECD (2012) similarly identified topics pertaining to human impacts as important for risk
17    assessment. These topics include the identification of toxicological endpoints specific to nanoparticles to
18    ensure that risk assessors identify all appropriate biological responses potentially leading to adverse
19    outcomes. Better characterization of mode of action in mammalian systems and interspecies variation
20    were also identified as important research to inform risk assessment.
21           In addition to effects described in Section 5.1, MWCNTs have been shown to generate reactive
22    oxygen species, increase cell permeability in human microvascular endothelial cells (HMVEC), promote
23    cell migration in HMVEC (Pacurari et al., 2012), and cause inflammation (see Additional Information
24    Highlight Box 16). MWCNTs have also been shown to demonstrate subpleural deposition and pleural
25    translocation (Mercer et al., 2010; Rvman-Rasmussen et al., 2009a), which has important implications for
26    carcinogenicity. In particular, as noted in Additional Information Highlight Box 13. MWCNTs have
27    structural similarities to asbestos, raising concern over the potential for asbestos-like effects (e.g.,
28    mesotheliomas).
29           In a review of the available literature, Donaldson et al. (2010) concluded that long MWCNT
30    fibers are retained in the stomata of the parietal pleura, which is normally responsible for particle

      March 2013                                     6-34                  DRAFT-DO NOT CITE OR QUOTE

-------
 1    clearance; therefore, this is the site of inflammation and pathogenic effects. Schinwald et al. (2012)
 2    investigated this mechanism further and reported a clear threshold where MWCNTs greater than 4 (im
 3    were pathogenic to the pleura. To elucidate the molecular mechanism of pathogenic action, Murphy et al.
 4    (2012) investigated the pro-inflammatory response of mesothelial cells and macrophages. The results
 5    indicated that CNTs indirectly resulted in an increase in cytokine release from mesothelial cells, as a
 6    result of frustrated phagocytosis of the macrophages. The authors concluded that the response in the
 7    pleura is first initiated by the macrophages, which in turn stimulate a pro-inflammatory response from the
 8    adjacent mesothelial cells (Murphy et al.. 2012).
 9            To extend upon the existing research described above and in Chapter 5. research planning efforts
10    might consider input related to human health effects from the CEA collective judgment workshop. While
11    time did not allow for experts participating in the workshop to discuss carcinogenic effects in humans,
12    they did develop specific research questions to further investigate non-cancer effects in humans. Based on
13    their discussion, potential risks of non-cancer effects result from chronic or non-chronic exposures to
14    coated or functionalized  MWCNT in upholstery textiles. To mitigate or avoid potential risks related to
15    non-cancer effects in humans, experts noted that risk managers might consider banning the material
16    outright; however, assessments would be necessary to inform this or other types of risk management
17    decisions. Experts did not specify the types of assessments that would inform these decisions, but they did
18    note that a NOEL  or LOEL should be identified for materials in consumer or occupational exposure
19    scenarios. Such information could be used in a human health hazard assessment. To support this type of
20    assessment, experts recommended the following research areas, along with estimates of the resources and
21    time to carry out the research:
      Table 6-13.   Research Identified by RTI Workshop Participants: Non-Cancer 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
Finances ($)
2 million
1 million
Estimated Time
Frame
3 years
3 years
      March 2013                                     6-35                   DRAFT-DO NOT CITE OR QUOTE

-------
      6.3.4.2. Human: Reproductive/Developmental
 1            The workshop participants who rated human health impacts as important most commonly rated
 2    reproductive and developmental effects as possibly important. They were generally not confident in the
 3    availability and utility of current data to support risk management decisions related to these effects (see
 4    Priority Research Area Highlight Box in Section 5.1). Listed below are examples of experts' rationale for
 5    rating this area as a research priority.
 6            •   Human Health Impacts—Reproductive/ Developmental: Data are currently insufficient.
 7               There is concern about the long-term effects of ENMs due to large effects on reproduction
 8               shown in animal populations.
 9            Experts who chose to note influential factors commonly selected the following considerations to
10    take into account in planning research on human reproductive or developmental impacts: MWCNT purity,
11    applied coatings, persistence, surface chemistry, matrix bound versus free form, morphology, exposure
12    route, and life stage.
13            The External Review Draft of this document did not contain information regarding reproductive
14    toxicity of MWCNTs, but two studies investigating developmental toxicity in rodents were described
15    (Fujitani et al., 2012; Lim et al., 201 Ib). An expert involved in the collective judgment prioritization
16    process identified, one study of MWCNT reproductive toxicity in mice (Bai et al., 2010) and a targeted
17    literature search identified a study of developmental toxicity in rats (Lim et al., 201 la). The findings
18    presented by Lim et al. (2011 a) appear to be from the same group of experiments described by Lim et al.
19    (20 lib) and discussed in Section 5.1.7; the conclusions of the two reports are the same: no differences in
20    gestation index, fetal death, fetal and placental weights, or sex ratio were observed as a result of maternal
21    MWCNT exposure at 1,000 mg/kg-day. Thus, the study by Lim et al. (2011 a) is not described in further
22    detail here.
23            Bai et al. (2010) conducted a reproductive toxicity assay using intravenous injection of water-
24    soluble amine and carboxylate-functionalized MWCNTs in male mice (single injection or 5 doses over
25    13 days of 5 mg/kg). Results indicated that MWCNTs accumulated in the testes, generated oxidative
26    stress, and reduced the thickness of the seminiferous epithelium (authors reported that this damage was
27    reversible) without producing any significant effects on sperm parameters, sex hormones, fertility,
28    pregnancy rate, or delivery success of female mice mated with treated males (Bai et al.. 2010). Although
29    this study used an exposure method that is not typically considered to be relevant, it is useful in that it
30    indicates that if MWCNTs were absorbed into the bloodstream via inhalation or oral exposure they would
31    not likely cause male reproductive effects. Therefore, use of resources to further investigate reproductive
32    and developmental toxicity might be better allocated toward characterizing female  reproductive endpoints
33    or developmental effects during various critical windows of development.

      March 2013                                     6-36                   DRAFT-DO NOT CITE OR QUOTE

-------
 1           Due to time constraints, expert stakeholders participating in the CEA collective judgment
 2    workshop did not discuss potential human reproductive or developmental effects; however, the limited
 3    available data suggest several remaining research questions that might inform future risk assessment and
 4    management efforts of MWCNTs. For instance, does gestational exposure to MWCNTs result in
 5    behavioral or other subtle neurodevelopmental effects in offspring observed through adolescence? Does
 6    comparing different formulations of MWCNTs within the same experiment indicate certain
 7    physicochemical characteristics that influence reproductive or developmental effects?

      6.3.4.3.  Aquatic Biota: Survival, Developmental,  Reproductive, and Other Sublethal
 8           Workshop participants identified impacts on aquatic biota as important to risk assessment. Risk
 9    relevance factors that might be considered in risk assessments and management efforts for this element
10    are: survival, developmental, reproductive, and other sublethal effects. Although there was some variation
11    in the way experts rated the importance of the "aquatic impacts" category and each risk relevance factor,
12    there was overall strong agreement among those who found aquatic impacts important that they were "not
13    confident" or only "somewhat confident" in the availability and utility of the current data to support risk
14    management decisions (see Priority Research Area Highlight Box in Section 5.1). Listed below are
15    examples of experts' rationale for why these topics were chosen as Research Priority Areas.
16           •   Impacts to Aquatic Biota—Survival: Data does not indicate acute toxicity
17           •   Impacts to Aquatic Biota—Developmental: As seen with endocrine disrupting chemicals,
18               aquatic species are very sensitive and low exposures can lead to developmental effects. A few
19               studies could go a long way toward understanding chronic effects.
20           •   Impacts to Aquatic Biota—Reproductive: As seen with endocrine disrupting chemicals,
21               aquatic species are very sensitive and low exposures can lead to reproductive effects.
22           •   Impacts to Aquatic Biota—Other Sublethal Effects: These effects are not typically studied
23               until environmental problems occur, need to give adequate attention.
24           The most commonly selected influential factors identified across all three areas included the
25    following: analytical techniques; a variety of MWCNT characteristics (e.g., adsorption/desorption ability,
26    aggregation/agglomeration state, lipophilicity, persistence, redox potential, surface  chemistry); physical
27    and chemical conditions associated with the surrounding media (particularly groundwater, sediment,
28    surface water, and wastewater); biological conditions (ADME, bioaccumulation, biomagnification,
29    microbial communities, organism health, developmental behavior, feeding behavior, reproductive
30    behavior); exposure route; habitat structure; and geographic location. The discussion below provides
31    more detailed information relating to aquatic toxicity, particularly in terms of aspects relevant to these
      March 2013                                     6-37                   DRAFT-DO NOT CITE OR QUOTE

-------
 1    identified influential factors. This information was gathered from literature identified after the workshop
 2    process.
 3           OECD (2012) similarly noted that a better understanding of the absorption, distribution,
 4    metabolism, and excretion of nanomaterials in ecological receptors is needed to evaluate the utility and
 5    appropriateness of standard ecotoxicological assessments for nanomaterials. OECD (2012) also
 6    recommended research of environmentally relevant conditions that could influence MWCNT toxicity.
 7           As described in a recent review by Petersen et al. (201 Ib). the aquatic toxicity of CNTs has been
 8    investigated in a variety of organisms including fish, algae, daphnia, copepods, amphibians (larvae),
 9    protozoans, and bacteria. Toxicity depends on the bioavailability of MWCNTs to aquatic organisms (see
10    Additional Information Highlight Box Gl). which is thought to be influenced by specific surface
11    chemistry and functionalization; but, conflicting evidence makes this relationship difficult to elucidate.
12    Petersen et al. (2010) reported that acid-treated MWCNTs, which were expected to be more bioavailable
13    than pure MWCNTs, did not actually accumulate at greater rates in the oligochaete L. variegatus.
14    Similarly, different surface coatings and charges had no apparent impact on accumulation or elimination
15    of MWCNTs in D. magna (Petersen et al.. 201 la; Petersen et al.. 201 Ib). but a clear increase in MWCNT
16    toxicity to C. dubia was observed with the addition of positively charged functional groups (Kennedy et
17    al.. 2009).
18           Another potentially important factor is the presence of metallic impurities. Mwangi et al. (2012)
19    found conflicting evidence: The removal of such impurities by acid pre-treatment decreased the lethal
20    effects of MWCNTs to mussels (V. iris), midges (C.  dilutes), and amphipods (H. azteca) but did not
21    mitigate biomass reduction for H. azteca, C. dilutes,  and L. variegates, indicating that metal impurities are
22    responsible for some, but not all, of the effects observed (see Additional Information Highlight Box 17).
23    Abiotic factors, such as pH and natural organic matter (NOM), are also expected to influence the
24    bioavailability and toxicity of MWCNTs; however, Edgington et al. (2010) found no impact on toxicity to
25    D. magna with a range of dissolved organic carbon content in NOM (although some variation was
26    reported based on the source of NOM). Clearly, MWCNTs do not conform to classic theoretical
27    predictions of chemical behavior in aquatic media, which is further complicated by incomplete
28    information or conflicting data.
29           Another important factor in aquatic toxicity of MWCNTs is their potential interaction with  other
30    contaminants. For example, Shen et al. (2012) showed that the presence of MWCNTs in sediment or soil
31    can complicate the toxicity of a system by altering the bioavailability of hydrophobic organic
32    contaminants like polycyclic aromatic hydrocarbons.
33           Available evidence indicates that MWCNTs are not absorbed in the gut.  For example, Edgington
34    et al. (2010) observed MWCNTs in the gut of D. magna without evidence of nanotube uptake into the

      March 2013                                     6-38                   DRAFT-DO NOT CITE OR  QUOTE

-------
 1    microvilli. Similarly, several studies conducted by Mouchet et al. (2011; 2010; 2008) showed DWCNT in
 2    the gut lumen but not in the blood, liver, or interstitial cells of X. laevis larvae, and there was no evidence
 3    that DWCNT passed across the intestinal barrier. Because current evidence shows a lack of CNT
 4    absorption across epithelial membranes, any observed toxic effects are likely due to interactions with
 5    epithelial surfaces. For example, O. mykiss exposed to SWCNTs showed signs consistent with impaired
 6    gill function (e.g., dilation of blood vessels in the brain, other signs of oxidative stress), suggesting that
 7    SWCNT accumulation on the gill surface results in impaired respiration (Petersen et al.. 201 Ib; Smith et
 8    al.. 2007). Indeed, authors observed elevated mucous secretion with SWCNT deposits associated with
 9    mucoproteins in the  gills. Although SWCNTs were observed in the gut of exposed fish, likely as a result
10    of drinking water with SWCNTs, no histological changes after were observed, with the exception of
11    increased lipid peroxidation at one time point in the 6-week experiment (Petersen et al.. 201 Ib; Smith et
12    al.. 2007). Similarly, Mwangi et al. (2012)  found no evidence of MWCNT penetration through  cell
13    membranes in C. dilutus and H. azteca. Nevertheless, authors observed MWCNT accumulation in the gut
14    of these organisms, which they speculated, along with lack of depuration after transfer into clean water,
15    contributed to decreased survival  and biomass (Mwangi et al.. 2012).  Evidence supporting the possibility
16    of MWCNT accumulation in the gut of aquatic organisms is found in studies in water fleas showing that
17    elimination of MWCNTs from the gut of C dubia (Kennedy et al., 2008) and D. magna (Petersen et al..
18    201 la; Petersen et al.. 2009) was possible only with the addition of algae as a food source. Edgington et
19    al. (2010) also attributed the toxicity of MWCNTs in D. magna to blockage in the gut. MWCNT
20    aggregation behavior has also been shown  to be an influential factor in toxicity as evidenced by increased
21    lethality in C. dubia with greater aggregation (Kennedy et al.. 2009; Kennedy et al.. 2008). which might
22    also support the influence of digestive tract blockage. Together, these studies demonstrate the potential of
23    MWCNTs to have toxic impacts on growth and survival of benthic invertebrates and other aquatic biota;
24    however, the mechanism of that toxicity and influence of environmental factors remains unclear.
25           Due to time  constraints, expert stakeholders participating in the CEA collective judgment
26    workshop did not discuss potential impacts in aquatic biota; however, based on the literature  described
27    above, remaining questions related to potential impacts in aquatic receptors include the following:
28           •   What modifications to MWCNTs might decrease interaction of the material with epithelial
29               surfaces in aquatic biota?
30           •   Are there population level effects in aquatic biota?

      6.3.4.4. Other: Economic, Societal, and Environmental Resources
31           Workshop participants identified "other impacts" as important to risk assessment. Economic,
32    societal, and environmental resource considerations are all included as risk relevance factors  within this

      March 2013                                     6-39                   DRAFT-DO NOT CITE OR QUOTE

-------
 1    element of the CEA framework. Although there was some variation in the way experts rated the
 2    importance of the "other impacts" category and each risk relevance factor, there was generally strong
 3    agreement among the experts who found other impacts to be important that they were "not confident" in
 4    the ability of the current data to support risk management decisions. Listed below are examples of reasons
 5    that experts chose these topics as Research Priority Areas.
 6           •   Other Impacts—Economic: There may be consequences of nanotechnology that are not yet
 7               known.
 8           •   Other Impacts—Societal: There may be consequences of nanotechnology that are not yet
 9               known.
10           •   Other Impacts—Environmental Resources: There may be consequences of
11               nanotechnology that are not yet known.
12           While there was some variation in the influential factors identified within each area (i.e.,
13    economic, societal, environmental), there was a great deal of overlap in influential factors identified for
14    all three areas, including analytical techniques, control technologies, geographic location, human activity,
15    occupation, subchronic exposure, and susceptible populations. Additional influential factors selected for
16    at least one but not all three areas included acute exposure, chronic exposure, and life stage.
17           Newly identified information on potential economic, societal, and environmental impacts reveals
18    several ongoing efforts to balance societal needs with potential risks. For instance, according to Safe
19    Work Australia's Human Health Hazard Assessment and Classification of Carbon Nanotubes (NICNAS,
20    2012). carbon nanotubes have garnered much attention in recent years due to unique physical and
21    chemical properties that show promise for a wide variety of advanced applications across many diverse
22    fields. Unique challenges to risk communicators result from the uncertainty surrounding these new
23    applications and the implication that the rapid innovation has for development of a wide variety of
24    MWCNT formulations. Risk assessors and risk communicators need to strike a difficult balance between
25    mitigating potential risk and unintended consequences of novel technology (Priest 2012; Siegrist et al.,
26    2011) without stifling much needed scientific innovation to meet the demands of a growing global
27    economy, the global population, and the ever-shrinking availability of natural resources (Klaine et al.,
28    2012; OECD. 2009). Nanotechnology has been praised for its potential to offer solutions for many of
29    today's environmental concerns,  including pollution, drinking water filtration, climate change, and energy
30    efficiency, while also fueling economic growth by promoting new technologically advanced industries.
31    However, as a rapidly growing and evolving field, there is a great deal of uncertainty regarding potential
32    trade-offs and unintended consequences (TPEN. 2012: Siegrist etal.. 2011: OECD. 2009: Sass. 2007V For
33    this reason, emphasis has been placed on interdisciplinary participation and collaboration, including
34    government, stakeholders, researchers, academics, and the public, throughout all stages of the risk
      March 2013                                     6-40                  DRAFT-DO NOT CITE OR QUOTE

-------
 1    assessment process to fully understand and effectively communicate potential risks while continuing to
 2    invest in technologically advanced solutions to everyday problems and to manage public perception
 3    OClaine et al.. 2012; Siegrist et al.. 2011; OECD. 2009; Sass. 2007).
 4           The OECD Working Party on Nanotechnology (OECD. 2012) noted that the ability to evaluate
 5    the economic impacts of nanotechnology depends on the development of valuation models specific to
 6    nanotechnology. Two valuation models described in the OECD report are the Defra model, which is
 7    based on a comparative valuation of an existing product and a nanoenabled "replacement," and the STAR
 8    METRICS approach, which uses an input/output method to analyze data between industries (OECD.
 9    2012). The Defra model offers some useful insight for this particular case study because a non-
10    nanoenabled product,  decaBDE, has already been identified for comparison. Although no information
11    was identified regarding the energy and resource demands for decaBDE, some economic assumptions
12    could be made for MWCNT in flame-retardant textiles based on how decaBDE diffused and performed in
13    the market. For example, as noted in Table 1-10. global demand for decaBDE was quite high between
14    2001 and 2007, driven in part by its use in textile applications for flame retardancy. Because flammability
15    performance and ability to meet stringent regulations is important for a variety of industries with textile
16    applications (see Section 1.2.1). the possibility exists for the demand for MWCNT in this application to
17    be on the same scale (i.e., similar percentage of total production). As discussed in Chapter 1.
18    considerations related to efficacy and production scale remain to be seen for MWCNT flame-retardant
19    coatings (see Additional Information Highlight Box 1 and Additional  Information Highlight Box 2).
20    Although two valuation models are available to begin to inform economic impacts of MWCNTs, these are
21    limited because the development and commercial success of products  containing MWCNTs face unique
22    challenges compared to non-nanoenabled products (OECD. 2012). For example, the research and
23    development stage for nanotechnologies often requires a very high level of investment and is
24    accompanied by a long lag period before any potential payoff is realized. With such a large number of
25    MWCNT-enabled products in this stage and relatively few products that have moved past it, estimating
26    when and how that payoff might take place is difficult. Similarly, additional risks are associated with
27    consumer perception or acceptance of nanoenabled products.
28           Considerations raised in the literature on potential economic, societal, and environmental impacts
29    suggest several remaining research questions, including the following:
30           •  What is the relative impact on environmental resources (e.g., water, energy) of MWCNT
31              production compared to other  flame-retardant materials (non-nanoenabled or nanoenabled)?
32           •  What are the economic implications of producing MWCNTs and MWCNT products (e.g.,
33              jobs created, infrastructure development)?
      March 2013                                    6-41                  DRAFT-DO NOT CITE OR QUOTE

-------
 1           •  What lessons from GMOs and other emerging technologies can be applied when
 2              communicating the potential risks and benefits of MWCNTs?
 3           Workshop participants discussed societal impacts within this category to identify a potential risk
 4    scenario and to develop research questions that could inform future assessment and risk management
 5    efforts for MWCNTs (RTI. 2012). They noted that unintended consequences of MWCNTs could arise,
 6    which might lead to concerns in government and industry, fear in the public, and abandonment of future
 7    applications of the material. To manage or avoid these risks, experts suggested that risk managers might
 8    consider efforts "to build capacity and enable informed consent," or ban the material if necessary. They
 9    further suggested a socioeconomic assessment to inform these types of risk management decisions. To
10    support this type of assessment, experts recommended the following research areas, along with estimates
11    of the resources and time to carry out the research:
      Table 6-14.  Research Identified by RTI Workshop Participants: Societial 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 nanotechnology       150,000         1.5 year
      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
      consequences (good and bad) of new technologies.                              500,000         2 year
      Development of new methods of facilitating communication amongst stakeholders on
      complex issues like nanotechnology.
      6.4. Moving  From "Assessment" to "Management" in the
      CEA  Process
12           Compiling information in the CEA framework and the subsequent collective judgment and
13    prioritization process represent the steps of the "assessment" phase of the CEA process (see Figure 1-2).
14    The "management" stage of the process involves moving the results of these steps into research plans or
15    risk management plans. The outcome of identifying and prioritizing information gaps in the collective
16    judgment step of CEA applied to MWCNTs is a list of topics in the CEA framework rated on the basis of
17    each topic's importance to risk assessment efforts and the level of confidence the experts have in the
18    information available for that topic to support risk management decisions. Areas collectively deemed of

      March 2013                                     6-42                   DRAFT-DO NOT CITE OR QUOTE

-------
 1    high importance to risk assessment, but least understood based on available information, were high
 2    priorities for developing specific research questions in the final stage of the prioritization process, a face-
 3    to-face workshop. If these research questions are pursued, the knowledge gained could support a variety
 4    of risk assessments and other analyses that could then be used to update the CEA framework, and
 5    subsequently carry out a collective judgment prioritization of risk-related trade-offs to inform risk
 6    management decisions.
 7           Such research fits within the overall  research paradigm at EPA, the principles of which include
 8    sustainability; systems thinking; integrated transdisciplinary research; and relevant, responsive, and rapid
 9    research (Anastas. 2012). Within this research paradigm, known as the "Path Forward," are six national
10    research programs (Anastas. 2012). Research specific to nanomaterials falls within the Chemical Safety
11    for Sustainability program, which has several themes (e.g., Life Cycle Considerations, Systems Models),
12    and thus this Program is oriented toward addressing some of the priority research areas identified through
13    the CEA process for MWCNTs. For example, questions related to physicochemical properties might be
14    investigated under the Program's theme on inherent chemical properties. Yet, the integrated
15    transdisciplinary nature of CEA and the Path Forward suggest that some research questions could be
16    addressed by multiple national programs working collaboratively. Notably, some of the identified
17    research objectives might best be addressed by other government agencies, academic institutions, or
18    others in the scientific community. The research priorities identified in the CEA process are thus intended
19    not only to inform EPA research, but also to  serve as a resource for the broader scientific community.
      March 2013                                     6-43                   DRAFT-DO NOT CITE OR QUOTE

-------
References
   ACC (American Chemistry Council). (200la). 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.
   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. IJ. (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
      nanotubes by  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.
      htto://dx.doi.org/10.1016/s0160-4120(03)00121-l

   Alberding. MR: Malecki. 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

   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.
March 2013                                       R-1                    DRAFT-DO NOT CITE OR QUOTE

-------
   Aschberger. K: Johnston. HJ: Stone. V: Aitken. RJ: Hankin. SM: Peters. SA: Trail. CL: Christensen. FM. (2010).
      Review of carbon nanotubes toxicity and exposure Appraisal of human health risk assessment based on open
      literature [Review]. CritRev 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/j.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

   Babrauskas. V: Krasny. J. (1985). Fire behavior of upholstered furniture. Gaithersburg, MD: National
      Engineering Laboratory. http://fire.mst.govMrlpubs/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

   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.
March 2013                                        R-2                    DRAFT-DO NOT CITE OR QUOTE

-------
   Birch. M. (2003). Diesel paniculate matter: As elemental carbon. In NIOSH Manual of Analytical Methods
      (NMAM), Fourth Edition (pp. 1-5). (5040). Birch, M.

   Bonner. JC. (2011). Carbon nanotubes as delivery systems for respiratory disease: do the dangers outweigh the
      potential benefits? [Review]. Expert Review of Respiratory Medicine 5: 779-787.
      http://dx.doi.org/10.1586/ers.ll.72

   Bottinj M: Bruckner. S: Nika. K: Bottinj N: Bellucci S: Magrini. A: Bergamaschj 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.

   BfFV; 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.

   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.
      http://www.qsa.man.dtu. dk/English/Research/Reports.aspx?lg=showcommon&id=265562

   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

   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: 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/i.taap.2008.12.006

   Cherukuri. 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
March 2013                                       R-3                    DRAFT-DO NOT CITE OR QUOTE

-------
   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

   Coccinj 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/annhvg/merllO

   Dahm. MM: Yencken. MS: Schubauer-Berigan. MK. (20 lib). 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. WI. (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

   de Wit. CA: Herzke. D: Vorkamp. K. (2010). Brominated flame retardants in the Arctic environment ~ trends
      and new candidates  [Review]. Sci Total Environ In Press, Corrected Proof: 2885-2918.
      http://dx.doi.0rg/10.1016/i.scitotenv.2009.08.037

   Delorme. M: Muro. Y: Arai 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). Translation 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.icis.2011.ll.019
March 2013                                        R-4                    DRAFT-DO NOT CITE OR QUOTE

-------
   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/kfj 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 pleura!
      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/10.1046/i. 1467-
      2494.1997.171707.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://esis.jrc.ec.europa.eu/doc/existing-chemicals/risk  assessment/SUMMARY/decasumO 13.pdf

   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

   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
      Chem29: 2511-2518. http://dx.doi.org/10.1002/etc.309

   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/i.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/j.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
March 2013                                        R-5                    DRAFT-DO NOT CITE OR QUOTE

-------
   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.ee.gc.ca/lcpe-cepa/default.asp?lang=En&n=B901A9EB&offset=l&toc=show

   EU (European Union). (2002). European risk assessment report for bis(pentabromophenyl) ether.
      http://esis.jrc.ec.europa.eu/doc/existing-
      chemicals/risk assessment/REPORT/decabromodiphenyletherreportO 13 .pdf

   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. Chem
      Res Toxicol 25: 74-82. http://dx.doi.org/10.1021/tx200255h

   Fleurv. D: Bomfim. JAS: Vignes. A: Girard. C: Metz. S: Mufioz. 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

   Fourches. D: Pu. D: Tropsha. A.  (2011). Exploring quantitative nano structure-activity relationships (QNAR)
      modeling as a tool for predicting biological effects of manufactured nanoparticles [Review]. Comb Chem
      High Throughput Screen 14: 217-225. http://dx.doi.org/10.2174/138620711794728743

   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.

   Gasser. M: Wick. P: Clift MJ: Blank. F: Diener. L: Yan. 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: 17-27. 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

   Ohio. 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.

   Golovin. YI: Golovin. DY: Shuklinov. AV: Stolyarov. RA: Vasvukov. VM. (2011). Electrodeposition of nickel
      nanoparticles onto multiwalled carbon nanotubes. Tech Phys 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
      Chem 29: 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.
March 2013                                        R-6                    DRAFT-DO NOT CITE OR QUOTE

-------
   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=159773930

   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

   Gutowski. 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

   Haa 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.

   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.

   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.

   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^rome^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.

   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
March 2013                                       R-7                    DRAFT-DO NOT CITE OR QUOTE

-------
   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

   IPEN. (2012). Nanotechnology 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: Canas. 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: Tanwani. 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). PolymDegrad 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

   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-
      Journalof 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
March 2013                                        R-8                    DRAFT-DO NOT CITE OR QUOTE

-------
   Kashiwagi T: Pagan. 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.polymer.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.polymer.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. (In Press) Genotoxicity of multi-walled carbon
      nanotubes in both in vitro and in vivo assay systems. Nanotoxicology.
      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
      Chem 27: 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.

   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
      InstrumMethods 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 JE: Lim. HT: Minai-Tehrani. A: Kwoa JT: Shia 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: Aa KH:  Lee. JS: Shia JH: Park. JD: Yu. IJ. (2012).
      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. (2012). 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
March 2013                                       R-9                    DRAFT-DO NOT CITE OR QUOTE

-------
   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

   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. Ill 1/j. 1600-
      0625.2007.00683.x

   Klaine. SJ: Koelmans. AA: Home. N: Parley. 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

   Kolosnjai-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
   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: Alexandre. 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 retardants sub stance 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: Covad 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: Jalili 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
March 2013                                       R-10                    DRAFT-DO NOT CITE OR QUOTE

-------
   Lee. JH: Lee. SB: Bae. GN: Jeoa KS: Yoon. JU: Ji. JH: Sung. JH: Lee. BG: Lee. JH: Yang. JS: Kim. HY: Kang.
      CS: Yu. IJ. (2010). Exposure assessment of carbon nanotube manufacturing workplaces. Inhal Toxicol 22:
      369-381. http://dx.doi.org/10.3109/08958370903367359

   Li. JG: Li. QN: 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. Carbon49:  1672-1679. http://dx.doi.0rg/10.1016/i.carbon.2010.12.052

   Li. W: Zhu. L: Zha. J: Wang. Z. (In Press) 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. (2011). 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

   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: Kim. SH: 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: Shia IS: Park. NH: Mooa C: Kang. SS: Kim SH: Park. SC: Kim. JC. (20lib). 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

   Lison. D: Muller. J. (2008). Lung and systemic responses to carbon nanotubes  (CNT) in mice  [Letter]. Toxicol
      Sci 101:  179-180; author reply 181-172. http://dx.doi.org/10.1093/toxsci/kfm249

   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.  (201 Ib). 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/ifl03374f
March 2013                                       R-11                    DRAFT-DO NOT CITE OR QUOTE

-------
   Luizj 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

   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.

   Maynard. AD: Aitken. PJ. (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 ultrafme TiO2 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:  Scabilloni. 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: Bergamaschj
      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.2Q545

   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
March 2013                                      R-12                    DRAFT-DO NOT CITE OR QUOTE

-------
   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/i.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

   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/sll051-011-0687-2

   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).
      Clastogenie 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: Ravmundo-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. Chem Res 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
March 2013                                       R-13                     DRAFT-DO NOT CITE OR QUOTE

-------
   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/etc. 1888

   Nagai H: Okazaki. Y: Chew. SH: Misawa. N: Yamashita. Y: Akatsuka. S: Ishihara. T: Yamashita. K:
      Yoshikawa. Y: Yasui. H: Jiang. L: Ohara. H: Takahashi. 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

   Nakari. T: Pessala. P. (2005). In vitro  estrogenicity of polybrominated flame retardants. Aquat Toxicol 74: 272-
      279. http://dx.doi.0rg/10.1016/i.aquatox.2005.06.004

   Nanocvl. (2009). THERMOCYL. http://www.nanocyl.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. (In Press) Nanomaterial Toxicity Testing in the 21st
      Century: Use of a Predictive Toxicological Approach and High-Throughput Screening. Ace Chem Res.
      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: VaUyathan. 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/review/docketl61A/pdfs/carbonNanotubeCIB PublicReviewOfDraft.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. RJ: 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.
March 2013                                      R-14                    DRAFT-DO NOT CITE OR QUOTE

-------
   Norris. JM: Kociba. RJ: Schwetz. BA: Rose. JQ: Humiston. CG: Jewell 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 Chem31: 5059. http://dx.doi.org/10.1002/etc.726

   NRC (National Research Council). (2000). Toxicological risks of selected flame-retardant chemicals.
      Washington, DC: National Academy Press.

   NRC (National Research Council). (2007). Toxicity testing in the 21st century: A vision and a strategy.
      Washington, DC: National Academies Press. http://www.nap.edu/catalog/l 1970.html

   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). 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 and B6C3F1 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: Lovik. 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

   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: 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
March 2013                                      R-15                    DRAFT-DO NOT CITE OR QUOTE

-------
   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/env/securitedesproduitschimiquesetbiosecurite/securitedesnanomateriauxmanufacture
      s/oecdconferenceonpotentialenvironmentalbenefitsofnanotechnologvfosteringsafeinnovation-ledgrowth.htm

   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 33.
      (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: Vallyathaa  V. (2010). Single- and multi-wall carbon nanotubes versus asbestos:
      Are the carbon nanotubes anew 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: Sl-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.vrtph.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.
      EnvironInt 35: 303-309. http://dx.doi.0rg/10.1016/i.envint.2008.07.021
March 2013                                       R-16                    DRAFT-DO NOT CITE OR QUOTE

-------
   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. http://dx.doi.org/10.1289/ehp. 10883

   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/etc. 149

   Petersen. EJ: Pinto. RA: Landrum PF: Weber. WJ. (2009). Influence of carbon nanotubes onpyrene
      bioaccumulation from contaminated soils by earthworms. Environ Sci Technol 43: 4181-4187.
      http://dx.doi.org/10.1021/es803023a

   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. (20lib). 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 Polyacrylonitrile. Macromol Rapid Comm 25: 987-990.
      http://dx.doi.org/10.1002/marc.200400055

   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. Ill

   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: Worfarth. 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

   Posner. S. (2004).  Survey and technical assessment of alternatives to decabromodiphenyl ether (decaBDE) in
      textile applications. Sundbyberg, Sweden: Swedish Chemicals Inspectorate.
March 2013                                       R-17                     DRAFT-DO NOT CITE OR QUOTE

-------
   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
      intraceHular 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.
   Qia 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 J Pharmacol 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: Cogni. D: Provini. A. (2007). Evaluation of DNA damage induced by decabromodiphenyl
      ether (BDE-209) in hemocytes of Dreissena polymorpha using the comet and micro nucleus assays. Environ
      Mol Mutagen48: 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: Fangstrora 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 (Research Triangle Institute). (2012). Nanomaterial case study workshop process: Identifying and
      prioritizing research for multiwalled carbon nanotubes. Summary report-final. Research Triangle Park, NC:
      RTI International, prepared for U.S. Environmental Protection Agency.
      http://cfpub.epa. gov/ncea/nano/recordisplav.cfm?deid=244011

   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.
   Saeed. K. (2010). Review on properties, dispersion and toxicology of carbon nanotubes. J Chem Soc Pakistan
      32: 559-564.
March 2013                                      R-18                    DRAFT-DO NOT CITE OR QUOTE

-------
   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_v201 l_Europe.pdf

   Sager. TM: Castranova. V. (2009). Surface area of particle administered versus mass in determining the
      pulmonary toxicity of ultrafme and fine carbon black: Comparison to ultrafme 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: Ill-Ill, 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 pleura!
      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

   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.0rg/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.

   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. InF 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(201 D.pdf
March 2013                                      R-19                    DRAFT-DO NOT CITE OR QUOTE

-------
   Sigma-Aldrich. (2012). Nanotubes and Mlerenes. 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/i.biomaterials.2009.04.009

   Sjodin. A: Carlsson. H: Thuresson. K: Sjorin. 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.

   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 ChemRes 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/j.atmosenv.2009.08.034

   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.

   Takagi. A: Kitajima. 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.

   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 Eiseigaku Zasshi 52: 182-188.
March 2013                                      R-20                    DRAFT-DO NOT CITE OR QUOTE

-------
   Tan. XM: Fugetsu. B. (2007). Multi-walled carbon nanotubes interact with cultured rice cells: Evidence of a
      self-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.jcis.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.

   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 [EPAReport].  (EPA-742/R-05-002A). http://www.epa.gov/dfe/pubs/flameret/altrep-v 1/altrepv 1 -f 1 c.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. http://www.epa.gov/nrmrl/lcaccess/pdfs/600r06060.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

   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) [EPAReport]. (635-R-07-008F).

   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-inventorypaper2008.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 fmal.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
March 2013                                      R-21                    DRAFT-DO NOT CITE OR QUOTE

-------
   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 [EPAReport], 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)
      [EPAReport]. (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

   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.

   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).

   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: 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/i.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 inBDNF, 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
March 2013                                      R-22                    DRAFT-DO NOT CITE OR QUOTE

-------
   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
      Environ400: 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. J: Shapira. P. (2012). Partnering with universities: a good choice for nanotechnology start-up firms?
      Small Business Economics 38: 197-215. http://dx.doi.org/10.1007/sl 1187-009-9248-9

   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.

   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

   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. Y: Hudson. JS: Lu. Q: Moore. JM: Mount AS: Rao. AM: Alexov. E: Ke. PC. (2006). Coating single-walled
      carbon nanotubes with phospholipids [Letter]. JPhys ChemB  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 eahtml
   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
March 2013                                       R-23                    DRAFT-DO NOT CITE OR QUOTE

-------
   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: 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

   Zhang. T: Du. Z: Zou. W: Li. H: Zhang. C. (2012). 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.polvmdegradstab.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
March 2013                                       R-24                    DRAFT-DO NOT CITE OR QUOTE

-------
Appendix A: Case Study Candidate
 Evaluation and Selection Process

-------
       Appendix  A. Case Study  Candidate Evaluation
                                          and Selection  Process

 1          Appendix A describes the process of evaluating and selecting carbon-based nanomaterials in
 2   specific applications as candidates on which to focus in the current case study. Specifically, it outlines the
 3   process by which the candidate carbon-based nanomaterial applications were identified and highlights
 4   factors relevant to the suitability of each candidate for a Nanomaterial Case Study. This information was
 5   used to support an informed selection of the nanomaterial and application for this case study by U.S.
 6   Environmental Protection Agency (EPA) program offices, labs, and centers in the Office of Research and
 7   Development, and regional offices during September 2011.


      A.1. Background
 8          As discussed in Chapter 1, the EPA (2007) Nanotechnology White Paper called for the use of
 9   nanomaterial case studies and multidisciplinary expert workshops as a means to inform research planning
10   to support the risk assessment process for nanomaterials. In response to the recommendations of the
11   Nanotechnology White Paper, EPA has been developing case studies of selected nanomaterials in specific
12   applications, including Nanoscale Titanium Dioxide in Water Treatment and Topical Sunscreen (U.S.
13   EPA. 2010c) and Nanoscale Silver in Disinfectant Spray (U.S. EPA. 2012b). To continue this series of
14   case studies, five carbon-based nanomaterial applications were identified as candidates for the next
15   Nanomaterial Case Study (presented in alphabetical order by material and application):
16          •   Carbon nanofibers in cement/concrete
17          •   Carbon nanotubes (multi-walled) in flame-retardant coatings and composites
18          •   Carbon nanotubes (multi-walled) in rubber tires
19          •   Carbon nanotubes (single-walled) in textiles
20          •  Nanocrystalline cellulose in biodegradable packaging
      March 2013                                 A-1                  DRAFT - DO NOT CITE OR QUOTE

-------
 1           This appendix provides further detail on the process by which the candidate carbon-based
 2    nanomaterial applications were identified, briefly summarizes the state of the science for each of the
 3    identified candidates, and highlights factors relevant to the suitability of each candidate for a
 4    Nanomaterial Case Study. The following criteria were used as guides in judging the candidates:
 5           •   "Nano-ness" of the material (i.e., whether the material is intentionally engineered at the
 6               nanoscale and has properties that distinguish it from conventional forms of the material);
 7           •   Potential for exposure throughout the product life cycle (in humans, both occupational and
 8               general public, as well as in other biota);
 9           •   Availability of data (whether directly related or inferred from other materials/products);
10           •   Feasibility of comparing the nano-enabled application to a non-nano-enabled application; and
11           •   Relevance to EPA programs.
12           Despite the wealth of information on carbon-based nanomaterials, EPA had previously
13    encountered difficulties in identifying an appropriate carbon-based nanomaterial application as a
14    candidate for a case study. These difficulties arose in part due to the following factors:
15           •   Few data are available on the actual commercial use of carbon-based nanomaterials in
16               products.
17           •   Little to no information is available on release of carbon-based nanomaterials  from
18               applications during normal use.
19           •   Small loadings (<5% by weight or volume) of carbon-based nanomaterials are generally
20               required to confer desirable properties.
21           •   Compared to other types of nanomaterials, the applications for carbon-based nanomaterials
22               appear to have a smaller exposure potential outside of occupational and disposal scenarios.
23           Although these challenges are still present, the recent increase in carbon-based nanomaterial
24    research has produced new data, and new applications have been proposed that offer greater potential for
25    widespread exposure. As a result, the selection of a carbon-based nanomaterial application for the next in
26    the Nanomaterial Case Study series appears to be feasible.
       A.2. Candidate Identification  Process
27           The process by which the five nanomaterial application candidates were identified is summarized
28    in Figure A-l. As noted in Section 1.1, the process began with a systematic approach to the identification
29    of nanomaterials and applications for investigation. This preliminary phase of the process utilized
30    strategic literature and Internet searches to identify supporting scientific literature, relevant news reports,
31    and nanomaterial information aggregation websites. Certain basic literature search statistics (e.g., number

      March 2013                                     A-2                   DRAFT - DO NOT CITE OR QUOTE

-------
 1    of total hits, number of hits in scientific databases) were then evaluated to provide a preliminary metric of
 2    interest within the nanotechnology community for each candidate nanomaterial/application combination
 3    and general data availability. After the "long list" of material/application combinations was identified and
 4    narrowed down using this systematic approach, a more judgment-based approach was used to evaluate the
 5    suitability of a "medium list" of potential candidates and identify the most feasible candidates for a case
 6    study. In this phase of the process, a list of suitability questions was used to step through the life cycle of
 7    a specific nano-enabled product and evaluate the characteristics of that product that might affect release,
 8    exposure, environmental fate, and impact on humans, ecological receptors, and the environment.
 9    A professional judgment as to whether a case study feasibly could be conducted for each candidate then
10    was determined based on the answers to the suitability questions, and a "short list" of five feasible
11    candidates was developed based on professional judgment of suitability.
      March 2013                                      A-3                    DRAFT - DO NOT CITE OR QUOTE

-------
               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/general 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.
March 2013
A-4
DRAFT - DO NOT CITE OR QUOTE

-------
       A.3.  Nanomaterial Application Candidates
 1           The candidate identification process yielded five nanomaterial applications for consideration as

 2    the subject of a new case study. The nanomaterial application candidates and proposed non-nanoenabled

 3    products for comparison are presented in Table A-l below. As discussed in Chapter 1. the non-

 4    nanoenabled product is intended to provide a frame of reference against which the ecological, human

 5    health, and other implications of the selected nanomaterial application can be compared.
      Table A-1.   Nanomaterial application candidates and non-nano-enabled products for comparison
               in a case study.
      Nanomaterial
Application
Proposed Non-Nano-Enabled Product for Comparison
      Carbon nanofiber (CNF)
Cement/concrete (CNF-
reinforced)
Steel- or glass-fiber reinforced cement/concrete
        Iti-walled 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 CNTs
      (SWCNTs)
Textiles (SWCNT-
reinforced/impregnated)
Traditional textiles reinforced with carbon fiber polymer
composites
      Nanocrystalline cellulose
      (NCC)
Biodegradable packaging (NCC-  Packaging containing polylactic acid (PLA) resin alone
polymer composite)
       A.4.  Summary of Factors Affecting Suitability

 6           Key considerations of the suitability of each candidate for development into a case study were

 7    summarized in a suitability chart (Table A-2). Based on the initial findings of this analysis (which should

 8    not be considered exhaustive or comprehensive), four of five applications appeared to still be in the

 9    research stage and unavailable for the commercial market in the United States or internationally.

10    Although carbon nanotubes have been incorporated into military textiles, carbon nanotube flame-retardant

11    coating is the only application identified as being currently available on the consumer market.
      March 2013
                      A-5
                   DRAFT - DO NOT CITE OR QUOTE

-------
Table A-2.     Factors for consideration in selecting a candidate for case study.





Selection Factors








Candidates

0)
£
o
c
0
0

c
E
0}
o

2
0
*- <"
•§ w
i- o
re Q.
o E
(£. o
E "o
— re
ul w

2 c
0 £
§ 8
^ 0


£
i—
i»

3
°^

z
o
§
s




M
0)

X
l~
H

o

(0


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
•
o
•
•
•
0
o
•
o
o
o
o
•
•
•
0
•
•
o
o
o
o
•
•
•
•
o
•
o
o
o
o
o
o
0
o
o
•
o
o
•
o
o
0
•
o
o
o
0
  c
  0
  0)
  a
 V)
 Answers to Suitability Questions.
 Yes (•) = Initial findings suggest that an affirmative answer can be given with relatively high confidence.
 Possibly ( ) = Initial findings from the preliminary literature review were conflicting.
 No (o) = Initial findings suggest that a negative answer can be given with relatively high confidence.
March 2013
A-6
DRAFT - DO NOT CITE OR QUOTE

-------
       A.5.  U.S.  EPA Program Involvement in  Final Selection
 1           As noted in Chapter 1. the selection of which of the five nano-carbon product candidates to use in
 2    this case study document involved representatives from EPA program offices, labs and centers within the
 3    Office of Research and Development, and regional offices. To facilitate distributing and discussing
 4    information relevant to the selection, an internal online forum was developed using a commercially
 5    available product (www.IdeaScale.com). The forum included brief introductory material on this
 6    nanomaterial case study series, instructions on using the website to nominate nano-carbon product
 7    candidates, links to tables summarizing life-cycle information on five candidate nano-carbon products
 8    (i.e., the applications in Table A-2),  and a more detailed report summarizing the state of the science for
 9    each candidate.
10           A link to the forum was sent to EPA representatives along with a request to share the link with
11    colleagues in their organization. Representatives and others in the Agency could then use the forum to
12    discuss the candidates  informally and nominate candidates for selection. Representatives were asked to
13    submit a formal vote that reflected input from their colleagues through the IdeaScale forum and other
14    communication channels they wished to use, as well as consideration of their own knowledge, the
15    information provided on the forum, and the consideration of the criteria listed in Section A.I.
16           The candidates receiving the most votes were  SWCNTs in textiles and MWCNTs in flame-
17    retardant coatings and  composites; thus, a hybrid option (MWCNTs in flame-retardant coatings applied to
18    textiles) was selected. This choice reflected comments that, although the textile application was
19    preferable, MWCNTs were perhaps  of greater interest based on indications that they will contain a higher
20    level of contaminants and are currently more widely produced.
21

      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). (2010c). 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). (2012b). Nanomaterial case study: Nanoscale silver in
         disinfectant spray (final report) [EPA Report]. (EPA/600/R-10/081F). Washington, DC.
      March 2013                                     A-7                    DRAFT - DO NOT CITE OR QUOTE

-------
This page intentionally left blank.

-------
   Appendix B. Analytical Methods for
 Detecting, Measuring, and Characterizing
     BDE-209 and Multiwalled Carbon
               Nanotubes
March 2013                         DRAFT - DO NOT CITE OR QUOTE

-------
     Appendix  B.  Analytical Methods for Detecting,
                               Measuring,  and  Characterizing
                               BDE-209  and  Multiwalled
                               Carbon  Nanotubes

 1          This appendix provides a brief overview of some of the available techniques to detect, quantify,
 2   and characterize polybrominated diphenyl ethers (PBDEs; specifically BDE-209 [decaBDE]) and
 3   multiwalled carbon nanotubes (MWCNTs) in laboratory, biological, and environmental settings, along
 4   with current challenges to making such measurements. This information is not intended to be exhaustive
 5   in reporting every applicable method and associated challenges or to be comprehensive in describing
 6   available methods; rather, it is a summary of relatively common or known methods for characterizing
 7   BDE-209 and MWCNTs based on information available at the time this case study was developed.


      B.1.  Measuring and  Characterizing PBDEs and MWCNTs
 8          Accurately measuring BDE-209 or MWCNTs in relevant biological or environmental media is
 9   critical for evaluating any potential impacts of either material on human health, ecological populations, or
10   environmental resources (Alcock et al.. 2011; Lehman et al.. 2011). The choice of which measurement
11   technique to use for either BDE-209 or MWCNT samples will ultimately involve a consideration of trade-
12   offs related to cost, time, selectivity, and sensitivity (Alcock et al., 2011; Lehman et al., 2011; Stapleton.
13   2006). In evaluating which analytical technique(s) to use, having an understanding of the challenges
14   related to quantifying and characterizing BDE-209 and MWCNTs is useful. For both materials, multiple
15   techniques might be required to characterize all of the physicochemical properties of interest in a single
16   sample (e.g., molecular composition, purity, shape, surface charge) (Alcock et al.. 2011; Lehman et al..
17   2011). Moreover, the training of personnel and the capital cost associated with some of these tools can
18   impede the analysis of materials by multiple laboratories  (Alcock et al.. 2011). In addition, the
19   standardization and validation of methods, availability of material standards, and the consistent reporting
20   of material characteristics in peer-reviewed literature have proven difficult for both BDE-209 and
21   MWCNTs (Alcock et al.. 2011; Lehman et al.. 2011). Distinguishing the sample from background
22   concentrations or other materials of similar composition (e.g., nonaBDE congeners versus decaBDE,
23   single-walled CNTs versus MWCNTs) is also a challenge for both materials (Lehman et al.. 2011;
24   Stapleton. 2006).
     March 2013                                B-1                DRAFT - DO NOT CITE OR QUOTE

-------
 1           For BDE-209, gas chromatography coupled with mass spectrometry is the most prevalently used
 2    method; however, a number of variations in this approach exist (Stapleton. 2006). Even small differences
 3    in analytical techniques can result in significant variation in results (Alcock et al., 2011; Stapleton, 2006).
 4    Recent efforts to standardize methods and develop techniques that minimize material degradation have
 5    improved interlaboratory variation, but continue to be the subject of study, particularly for measuring
 6    samples in complex milieus (Stapleton. 2006). Detection and characterization of BDE-209 has proven
 7    more difficult than lower brominated compounds due in part to degradation at high temperatures and with
 8    ultraviolet light exposure (Stapleton. 2006).
 9           For MWCNTs, transmission electron microscopy is generally used to characterize structural
10    properties of the material, which is the first step in differentiating between MWCNTs, SWCNTs, or other
11    materials; however, using this tool in tandem with others is necessary to characterize the material
12    (Lehman et al.. 2011) more completely. Challenges related to characterizing and quantifying MWCNTs
13    include their propensity to agglomerate or otherwise transform (e.g., surface oxidize) during the process
14    of production, purification, or exposure, as well as interference from experimental artifacts (e.g., metal
15    catalysts used in material production) (Petersen and Henry. 2012; Lehman etal.. 2011). The challenge of
16    combining multiple techniques (e.g., gas chromatography and mass spectrometry) for BDE-209 analyses
17    is amplified for MWCNTs in that a multitude of measurements and sampling techniques are generally
18    required to fully characterize nanomaterials (Lehman et al.. 2011).
       B.2. Summary Tables
19           The tables below highlight techniques for detecting, measuring, and characterizing PBDEs and
20    MWCNTs. Table B-l briefly outlines advantages and disadvantages of individual approaches to gas
21    chromatography and spectrometry, as well as a few alternative techniques that are available for studying
22    PBDEs. Table B-2 provides a brief overview of available methods to characterize a range of MWCNT
23    properties. More detail on each approach can be found in the references listed at the end of this appendix,
24    particularly the recent review by Lehman et al. (2011).
      March 2013                                    B-2                   DRAFT - DO NOT CITE OR QUOTE

-------
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)
Technique1
Atmospheric pressure
photoionization (APPI)-
coupled liquid
chromatography
(LC)/MS-MS2>3
Gas chromatography
(GC)/electron capture
detection3
GC/electron capture
negative ionization mass
spectrometry (ECNI-
MS)1>2>3
GC/electron ionization
(El) MS1>2>3
GC/high resolution mass
spectrometry (HRMS)2>3
GC/HR time of flight
(TOP) MS1.2.3
On-column injection
GC1>3
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
chromatographic
resolution
relative to gas
chromatography
• 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
March 2013
B-3
DRAFT - DO NOT CITE OR QUOTE

-------
Table B-1, cont: Analytical techniques for detecting, measuring, and characterizing PBDEs.
Citation(s)
Stapleton (2006)
Stapleton (2006)
Technique1
Programmable
temperature vaporization
(PTV) injection GC2>3>4
Split/splitless injection
gas chromatography
(GC)2>4
Application(s)
• Separation and detection
of PBDE congeners
• Determination of
molecular weight
• 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
• Can be used on
environmental samples
• Necessary equipment
commonly present in
laboratories
Disadvantages
• Requires
significant
optimization to
perform
separation
• Injection volume
must be small
• High injection
temperature
  techniques listed in alphabetical order
  2Used for chemical (in vitro) analysis as reported in reference document
  3Used to analyze in vivo samples as reported in reference document
  4Used to analyze environmental samples as reported in reference document
Table B-2.   Analytical techniques for detecting, measuring, and characterizing MWCNTs.
Citation(s)
Petersen and
Henry (2012)
Lehman etal.
(2011)
Johnston etal.
(2010)
Petersen and
Henry (2012)
Petersen and
Henry (2012)
Petersen and
Henry (2012)
Petersen and
Henry (2012)
Technique1
Atomic force
microscopy2
Centrifugation2
Chemothermal
oxidation4
(at375°C)
Cryotransmission
electron microscopy
(CEM)2
Fluorescence
microscopy (FLM)3<4
Application(s)
• 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
• Detection of MWCNT in
environmental media or
tissue from biological
specimens
Advantages

• Centrifugation
equipment commonly
present in laboratories
• Allows for quantitative
determination of
MWCNT concentration

• Can detect single
MWCNTs
Disadvantages
• Limited to samples in
aqueous phase
• Dispersion difficult;
requires extensive
sonication
• Accuracy may be
affected by dispersion
• Inaccurate
• Limited to samples in
aqueous phase
• Necessary equipment is
not common in
laboratories
March 2013
B-4
DRAFT - DO NOT CITE OR QUOTE

-------
Table B-2, cont: Analytical techniques for detecting, measuring, and characterizing MWCNTs.
Citation(s)
Lehman etal.
(2011)
Petersen and
Henry (2012)
Lehman etal.
(2011)
Lehman etal.
(2011)
Johnston etal.
(2010)
Petersen and
Henry (2012)
Revel and
Avrault(2000)
Petersen and
Henry (2012)
Revel and
Avrault(2000)
Petersen and
Henry (2012)
Johnston etal.
(2010)
Technique1
Fourier transform
infrared spectroscopy
(FTIR)2
Dynamic light
scattering (DLS)2
Gas pycnometry2
ICP mass
spectrometry (ICP-
MS)3>4
Instrumental neutron
activation analysis2'3'4
Light microscopy2.3.4
Application(s)
• 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
• 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
• 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
March 2013
B-5
DRAFT - DO NOT CITE OR QUOTE

-------
Table B-2, cont: Analytical techniques for detecting, measuring, and characterizing MWCNTs.
Citation(s)
Technique1
Application(s)
Advantages
Disadvantages
Lehman etal.     N2 gas adsorption2'4   • Surface area
(2011)                               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 etal.
(2011)
Johnston etal.
(2010)

Lehman etal.
(2011)
Petersen and
Henry (2012)
Johnston etal.
(2010)
Petersen and
Henry (2012)
Johnston etal.
(2010)

Lehman etal.
(2011)
Johnston etal.
(2010)


Lehman etal.
(2011)

Nitrogen and
phosphorous
doping2'3


Optical density (UV-
vis absorbance)2.3.4




Radioactive
labeling3'4


Raman
spectroscopy2'3'4




Scanning electron
microscopy (SEM)2

• 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

• Can detect non-carbon
atoms present in
MWCNT
• 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


• 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 MWCNT




• Does not provide
information on internal
morphology
March 2013
                                   B-6
                                        DRAFT - DO NOT CITE OR QUOTE

-------
Table B-2, cont: Analytical techniques for detecting, measuring, and characterizing MWCNTs.
Citation(s)      Technique1
                    Application(s)
                          Advantages
                       Disadvantages
Petersen and
Henry (2012)
Thermal optical
transmittance4
• Loss of mass at various
  temperatures
• Samples can contain   •  Only useful for
  dissolved                MWCNTs in aqueous
  environmental material     phase

                       •  Necessary equipment is
                          uncommon in
                          laboratories
Lehman etal.
(2011)
Thermogravi metric
analysis (TGA)2
  MWCNT purity analysis
• Necessary equipment
  is common in
                                                                laboratories
Sample size
requirements may be
large for certain
applications (3-10 mg)
Multiple measurements
needed to ensure
accuracy of data
Lehman etal.
(2011)
Petersen and
Henry (2012)
Johnston etal.
(2010)

Transmission Electron
Microscopy (TEM)2>3>4





• Surface morphology
• Crystallinity





• Provides high
resolution information
about nanotube
structure



• 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 etal.    X-ray diffraction2       •  Skeletal density
(2011)
                                                                      •  Cannot be used to
                                                                         determine bulk density
Lehman et al.    X-ray microanalysis
(2011)
Echlin (1998)
                                2<3
                       Purity
                          •  Narrowing of incident
                             beam allows greater
                             resolution
Lehman etal.    X-ray photoelectron    •  Surface chemical
(2011)          spectroscopy (XPS)2     composition
Petersen and                         .  Presence of functional
Henry (2012)                           groups
                                                                      •  May be inaccurate
                                                                         without fluorine tagging
techniques listed in alphabetical order
2Used for chemical (in vitro) analysis as reported in reference document
3Used to analyze in vivo samples as reported in reference document
4Used to analyze environmental samples as reported in reference document
March 2013
                                     B-7
                                          DRAFT - DO NOT CITE OR QUOTE

-------
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

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: Christensea 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 Chem31: 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 Chem244: 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-Y
March 2013                                        B-8                     DRAFT - DO NOT CITE OR QUOTE

-------
                                This page intentionally left blank.
March 2013                                     B-1                   DRAFT - DO NOT CITE OR QUOTE

-------
 Appendix C. Efficacy of Various Methods
     of Carbon Nanotube Purification
March 2013                          DRAFT - DO NOT CITE OR QUOTE

-------
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.
       March 2013                      C-1          DRAFT-DO NOT CITE OR QUOTE

-------
Table C-1.    Purification methods for carbon nanotubes.
Carbonaceous impurities
Purification methods
Chemical Gas phase Air (Plus HCI)
methods CI2, H20, HCI
H20, Ar, 02 (Plus HCI)
02, C2H2F4, SF6
Liquid phase HMOs
H202, HCI
Mixture of acid or KMn04
Microwave in inorganic acid
Electrochemical Alkali or acid solution (Plus HCI)
Physical Filtration
methods Centrifugation
Solubilization with functional groups
High temperature annealing
Other physical techniques
Chromatography, electrophoresis, FFF2
Multistep HIDE3, wet grinding, filtration, oxidation, sonication,
methods centrifugation
Filtration/magnetic filtration, oxidation, annealing
Sonication in H202, HNOs/HF/SDS, filtration
High temperature annealing extraction
Yield Amorphous
(wt%) carbon Graphite
-2-35
-15
-30
25-48
-30-50
10-75
30-75
10-60
-80
-30-84
-10-40
-17-50
-70-90
-10-
?
-2
-9-20
-25
-90
1*^= effective; O= partially effective; *= not effective
2Field-flow fractionation
3Hydrothermally initiated dynamic extraction
Source: Hou etal. (2008). Purification of carbon nanotubes. Carbon 46: 2003-2025.
Si
s
s
s
s
s
s
o
o
o
o
o
X
X
s

s
s
o
http://dx.doi.org/10.
X
X
X
X
X
X
O
X
X
O
O
O
X
X
s

X
O
O
Carbon Soluble
impurity on carbon in
walls solution
O x
O *
O *
O *
O *
O x
O x
O x
X X
X •/
X •/
X •/
X X
X X
X X
0 S
S S
O S
S S
Metallic impurities
Metal covered
Exposed in polyhedral
metal carbon
^ O
s o
s o
s s
s o
s o
s s
s o
•/ X
o o
o o
x O
/- /-
/- /-
o o
/• o
/- /-
^ o

Metal
encapsulated
byCNT
O
O
O
o
o
o
o
o
o
X
x
x
S
•/
x
O
O
O

1016/i.carbon.2008.09.009.
                         March 2013
C-2
DRAFT-DO NOT CITE OR QUOTE

-------
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
March 2013                                                           DRAFT-DO NOT CITE OR QUOTE

-------
  Appendix D: Study Summaries on the
  Transport, Transformation, and Fate of
Decabromodiphenyl Ether and Multiwalled
          Carbon Nanotubes in
         Environmental Systems
March 2013                        DRAFT-DO NOT CITE OR QUOTE

-------
    Appendix  D.  Study Summaries on  the

                               Transport,  Transformation, and

                               Fate of BDE-209 and  MWCNTs in

                               Environmental Systems


1          The following tables provide details from studies related to fate, transport, or transformation of

2   decabromodiphenyl ether [(decaBDE) specifically the single isomer of decaBDE, BDE-209] or

3   multiwalled carbon nanotubes (MWCNTs) in environmental media. Study information that provides

4   support for the transformation (debromination) of BDE-209 in environmental media is included in Table

5   D-l. Table D-2 provides study summaries related to the behavior of MWCNTs in aqueous media. Table

6   D-3 and Table D-4 present information from studies conducted in terrestrial ecosystems for BDE-209 and

7   MWCNTs, respectively.




    Table D-1.   Relevant studies of transformation  (debromination) of BDE-209.


    Citation        Relevant Study Information1

    Biotic Debromination

    Deng et al. (2011)  • Used aerobic bacterium Lysinibacillus fusiformis strain DB-1 to determine its capability to 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 examined
    (2010)           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.
     March 2013                                  D-1                  DRAFT-DO NOT CITE OR QUOTE

-------
Table D-1, cont: Relevant studies of transformation (debromination) of BDE-209.
Citation
                  Relevant Study Information1
Biotic Debromination
Tokarzetal.
                  • 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. (2011)  • Examined microbial degradation of BDE-209 in the rhizosphere of ryegrass using arbuscular
                    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 Ti02 photocatalyst.

                  • Ti02 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      • Examined BDE-209 photochemical transformation using a  hexane solvent and solar light.
(2004)
                  • 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 UV
al. (2009)            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)    • 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.
March 2013
                                                       D-2
DRAFT-DO NOT CITE OR QUOTE

-------
Table D-1, cont. Relevant studies of transformation (debromination) of BDE-209.
Citation
                  Relevant Study Information1
 Abiotic Debromination (Photolysis)
Raff and Hites
                  • 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.
Shih and Wang
Schenker et al.     • Used multimedia model that incorporated photolysis to predict fate of BDE-209 in 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%.

                  • 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.

                  • Studied photodegradation of BDE-209 in toluene, on  silica gel, and in sand, soil, sediment 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.

                  • Studied photodegradation of BDE-209 in house dust  exposed to natural sunlight.

                  • Initial BDE-209 concentration decreased by about 38%,  35% of which believed to be due to
                    debromination.

Additional information obtained from U.S. EPA (201 Oa).
Soderstrometal.
Stapleton and
Dodder (2008)
March 2013
                                                      D-3
DRAFT-DO NOT CITE OR QUOTE

-------
Table D-2.    Relevant studies of MWCNTs in aqueous media.
 Citation
                  Relevant Study Information
 Chae et al. (2011)   • Studied photochemical reactivity of CNT aggregates and compared with other fullerene 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.
 Chappelletal.
 Christian etal.
                  • Examined the mechanism by which humic substances stabilize MWCNT dispersions in 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.

                  • Studied aggregation of nanoparticles and effects of humic acid and cations on CNT stability.

                  • 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 the
(2Q12)               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. (2008b)   • 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.
 Holbrooketal.
 (2010)
 Hyungetal.
 (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.
March 2013
                                                      D-4
DRAFT-DO NOT CITE OR QUOTE

-------
Table D-2, cont: Relevant studies of MWCNTs in aqueous media.
 Citation
                  Relevant Study Information
 Hyung and Kim
 Kennedy etal.
                  •  Investigated the effect of NOM characteristics and water quality parameters on NOM adsorption 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.

                  •  Investigated factors that influence the partitioning of CNTs (raw versus functionalized [either 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 and
(2009)                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.
 Kummereretal.
 (2011)
                  •  Investigated biodegradability of functionalized and nonfunctionalized MWCNTs in aqueous media.

                  •  MWCNTs were not biodegradable under the conditions tested. Surface modification resulted in better
                     solubility, but not better biodegradability.

                  •  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.
March 2013
                                                      D-5
DRAFT-DO NOT CITE OR QUOTE

-------
Table D-2, cont:  Relevant studies of MWCNTs in aqueous media.
 Citation
                 Relevant Study Information
 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.

 Petersen et al.     •  Studied ecological uptake in sediment spiked with MWCNTs by sediment-burrowing Lumbriculus
 (2008a)              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. (2008)  •  Examined aggregation kinetics of MWCNTs in aquatic media with varying solution pH and salt
                     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 TT-TT
                     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.
 (2010)
                  • Investigated NOM, pH, and ionic strength effects on adsorption of SOCs by MWCNTs in natural 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.
(2011 a)
                  •  Examined interactions (phase distribution) between MWCNTs and aqueous systems containing 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.
 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
March 2013
                                                     D-6
DRAFT-DO NOT CITE OR QUOTE

-------
 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 a)    •  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.
 March 2013                                             D-7                       DRAFT-DO NOT CITE OR QUOTE

-------
Table D-3, cont: Relevant studies of BDE-209 in soils and plants.
 Citation
                  Relevant Study Information
 Plants
 Huang et al.        • Described the uptake, translocation, and metabolism of BDE-209 in six plant species—ryegrass, alfalfa,
 (2010)               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 coatings—
(2011 a)              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.
March 2013
                                                      D-8
DRAFT-DO NOT CITE OR QUOTE

-------
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 Biodegradation
   65: 465-469. http://dx.doi.0rg/10.1016/i.ibiod.2011.01.008

Desai. C: Addo Ntira 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.icis.2011.ll.019

Han. Z: Zhang. F: Lin. D: Xing. B. (2008b). 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.jcis.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
March 2013                                        D-9                     DRAFT-DO NOT CITE OR QUOTE

-------
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.org/10.1897/09-
   024.1

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 Chem 27:
   1932-1941. http://dx.doi.org/10.1897/07-624.1

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. Environ Pollut 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. (2009b). 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. (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.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. (2008a). Ecological uptake and depuration of carbon nanotubes by
   Lumbriculus variegatus. Environ Health Perspect 116: 496-500. http://dx.doi.org/10.1289/ehp. 10883

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

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.0rg/10.1016/i.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
March 2013                                         D-10                     DRAFT-DO NOT CITE OR QUOTE

-------
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. 1

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.

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

Vrkoslavova. J: Demnerova. K: Mackova. M: Zemanova. T: Macek. T: Hajslova. 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. (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. 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 RiverDelta, China: Occurrence, inventory, and fate. Environ Sci Technol 41: 8262-8267.
   http://dx.doi.org/10.1021/es071956d
March 2013                                        D-11                    DRAFT-DO NOT CITE OR QUOTE

-------
This page intentionally left blank.

-------
       Appendix E: Environmental
       Contaminant Concentrations
March 2013                         DRAFT-DO NOT CITE OR QUOTE

-------
    Appendix  E.  Environmental  Contaminant

                               Concentrations

1          Appendix E provides information available from the literature regarding reported environmental
2   concentrations of decaBDE (BDE-209) and multiwalled carbon nanotubes (MWCNTs) in environmental
3   media (dust, indoor and outdoor air, aquatic systems, sewage, and soil) (Section E.I), and biota (Section
4   E.2). Literature was identified primarily using review articles published in the past two years. Targeted
5   literature searches were carried out as needed.

     E.1. Concentrations in  Environmental Media

6          The following tables provide details from studies that measured BDE-209 in dust, air, water,
7   sediment, soil, and sewage effluent/sludge. No data were identified on MWCNT concentrations in
8   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)
 • Washington, DC
 • Dust samples from 16 homes
                                                               Mean (dry wt): 2,090
     Sharp and Lunder
     (2004) as cited in U.S.
     EPA(2010a)
 • Throughout United States
 • 10 homes
                                                               Mean (dry wt): 2,394
     Siodinetal. (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 etal. (2005)
 • Computer labs, CA
 • 2 carpet dust samples
                        BDE-209 was the dominant congener in carpet dust
                                                               Mean: 5,180
Schecteretal. (2005)
                      • Dallas, TX
                      • 9 vacuum samples
                      • BDE-209 was the dominant congener in 7 samples
                                               Mean (Median) (dry wt):
                                               8,567 (665)
    March 2013
                         E-1
                                                            DRAFT-DO NOT CITE OR QUOTE

-------
Table E-1, cont: BDE-209 concentrations in building dust.
Citation
Relevant study info
BDE-209 levels
United States
Allen etal. (2008)
Wuetal. (2007)
[levels reported by
U.S. EPA (2010a)]
Harradetal. (2008b)
Johnson-Restrepo
and Kannan (2009)
Battermanetal.
(2010)
Watkins et al. (2011)
• 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)
• Boston, MA area
• 46 women; 1 1 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
Geometric means:
4,502 (main living area);
1,703 (bedroom);
1,811 (vacuum)
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)
• Thailand: 5 electronic/ electrical waste storage facilities
• Dust, 25 samples
   Mean: 33,000
 Harradetal. (20Q8b)
• 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)
   Mean (geometric mean):
   670 (590) (Canada);
   45,000 (3,800) (U.K.)
 Harradetal. (2008a)
• 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 (median):
   260,000 (8,100) (homes);
   30,000 (6,200) (offices);
   410,000 (100,000) (cars)
 Ma etal. (2009)
• Taizhou, China
• 5 dust samples from electronic waste recycling workshop floor
• BDE-209 accounted for major proportion of total PBDEs in dust
   Mean (range) (dry wt): 29,800
   (5,560-80,600)
 Sjodin etal. (2008)      • Household dust in vacuum cleaner bags from 10 homes in each   Median (range):
                         country: 6 cities in Germany, 2 cities in Australia, and 1 city in     63 (<6-410) (Germany);
                         the United Kingdom (total n=30)                              730 (23-13,000) (Australia);
                       • BDE-209 dominant congener                                10,000(910-54,000) (U.K.)

Note: Additional information obtained from U.S. EPA (201 Oa): Acronyms: dry wt = Dry weight; ND = Not detected
March 2013
                              E-2
DRAFT-DO NOT CITE OR QUOTE

-------
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
(2010a)
                       • Outdoor sampling at 5 locations; sampling every 12 days    Means (pg/m3)
                         (August 2003-January 2004)                            60.1 (Chicago)
                       • Chicago (urban), remote locations in Michigan and         1.4 (Michigan)
                         Louisiana, agricultural site in Arkansas, and small college   2.6 (Louisiana)
                         town of Bloomington, Indiana                            9.0 (Arkansas)
                       • BDE-209 dominant congener at all sites                  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.        • Outdoor sampling at 4 locations; 4 samples/year at each
 (2001)                   location, May-October (1997-1999)
                       • 1 urban (Chicago), 1 remote (Michigan), 2 rural (Michigan,
                         New York)
 Charles etal. (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.
 Means (pg/m3)
 0.3 (Chicago)
 ND (rural/remote sites)


 Mean (range) (pg/m3)
 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 [(Gal/EPA,
 2006); as cited in U.S.
 EPA(2010a)
                      • 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/m3
 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 U.S. adults.
                                                                              Geometric means (pg/m3)
                                                                              173.6 (personal air)
                                                                              94.8 (bedroom)
                                                                              94.2 (living room)
March 2013
                                                      E-3
  DRAFT-DO NOT CITE OR QUOTE

-------
Table E-2, cont: BDE-209 concentrations in outdoor and indoor air.
 Citation
                     Relevant study info
                                                                             BDE-209 levels
 United States
Johnson-Restrepo
and Kannan (2009)
                       • Indoor air in 12 homes, Albany, NY (December 2007-       ND (ng/m3)
                         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/m3)
                                                                              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 collected at
 (2010)                   10 office buildings in southeast Michigan (Ann Arbor area)

 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
Chang etal. (2009)
                                                                              Means and medians were all below
                                                                              the limit of detection
                                                                              Mean (range) (pg/m3)
                                                                              1.6(0.091-9.8)
                         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/m3)
                                                                              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/m3)
                                                                              10.4 (MSW)
                                                                              6.5 (reference site)
 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/m3)
                                                                              19(ND-105)
Note: Additional information obtained from U.S. EPA (201 Oa]
LRT = Long range transport; MSW = Municipal solid waste; ND = Not detected
March 2013
                                                      E-4
                                                                                DRAFT-DO NOT CITE OR QUOTE

-------
Table E-3.     BDE-209 concentrations in aquatic systems.
 Citation
Relevant study info
                                                                          BDE-209 levels
 United States
 Orosetal. (2005)
 • San Francisco estuary (2002)
 • 48 sediment samples; 33 water samples
                                                                           Range (pg/L)
                                                                            ND-191 (surface water)
                                                                            ND (sediment)
Hun Yunetal. (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)
 Songetal. (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
                                                                           Range (ng/gram dry wt): 0.16-14.79
                      •  BDE-209 was the dominant congener (49% of total
                         concentration)
 Dodder etal. (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.        •  Downstream of WWTP of plastics manufacturer, North
 (2007)                   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)

 International
                                                       Range
                                                       2002 SD: 300-3,150 ng/gram
                                                       2005 SD: 181-2,390 ng/gram
 Toms etal. (2006) as
 cited in U.S. EPA
 (2010a)
 •  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)
 Christensen and Platz
 (2001)
 •  Danish marine coastal areas, freshwater lakes, river
    (2000)
 •  BDE-209 was the dominant congener in marine and
    freshwater sediments
 •  Highest BDEs detected in urban sediments
                                                                           Range (ng/gram dry wt)
                                                                           <0.9-3.9 (marine)
                                                                           < 1.3-8.1 (freshwater)
March 2013
                                E-5
                                                                             DRAFT-DO NOT CITE OR QUOTE

-------
Table E-3, cont: BDE-209 concentrations in aquatic systems.
 Citation
Relevant study info
                                                                             BDE-209 levels
 International
 Eljarratetal. (2005)
 •  Coastal areas, Spain
 •  13 marine sediment samples
                                                                              Range (ng/gram dry wt)
                                                                              2.46-132.10
                       • BDE-209 was the dominant congener (50-99% of total
                         concentration)
 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.
   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)
 Eljarratetal. (2004)
 [also reported in Law
 etal.(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)
Sawaletal. (2004)
[also reported in Law
etal. (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 (Schlabachetal., 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
 Voorspoelsetal.
 (2004) [also reported
 inLawetal. (2006b)
 review article]
 • Analyzed sediments from Belgian North Sea, Western
   Scheldt Estuary
 • BDE209 was detected in 83% of samples from the Belgian
   North Sea and in 100% of samples from the Scheldt Estuary
                                                                              Maximum (ng/kg dry wt): 1,200 (at
                                                                              estuary)
March 2013
                                E-6
                                                                              DRAFT-DO NOT CITE OR QUOTE

-------
Table E-3, cont: BDE-209 concentrations in aquatic systems.
 Citation
                      Relevant study info
                                                                            BDE-209 levels
 International
 Ricklundetal. (2010)
                         Measured levels of BDE-209 in 11 lake sediment samples   Range (ng/gram dry wt)
                         and 7 marine sediment samples in Sweden                0.48-11 (lake)
                         No known point sources of BDE-209 exist; presence in      1.0-88 (marine)
                         sediments was presumed to be evidence of long-range
                         atmospheric transport and deposition
                                                                              Range (ng/gram dry wt): 0.4-7,340
 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
                       • Examined 3 sediment cores from the Pearl River Estuary,    Range (ng/gram): 13.5-30.3
                         South China
                       • Increased BDE-209 flux in the upper sediment cores
                         attributed to rapid regional growth of electronics and other
                         industry
Guzzellaetal. (2008)
                       • PBDEs measured in sediment cores (2005) from Lake
                         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
                                                                              Range (ng/gram dry wt): 1.6-15.3
                                                                              Range of all PBDEs (ng/gram dry wt):
                                                                              0.13-1.98
                                                                              (BDE-209 levels stated to be about
                                                                              1 order of magnitude higher)
 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

Note: Additional information obtained from U.S. EPA (201 Oa]
dry wt = Dry weight; ND = Not detected; SPM = Suspended particulate matter; TOC = Total organic carbon; WWTP = Waste water treatment
plant
March 2013
                                                      E-7
                                                                              DRAFT-DO NOT CITE OR QUOTE

-------
Table E-4.     BDE-209 concentrations in sewage effluent and sludge.
 Citation
Relevant study info
                                                                                 BDE-209 levels
STP effluent
                                                                                          Sewage sludge
 United States
 Hale etal. (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 etal. (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
                     • 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
                                                 1,730(pg/L)
                       adsorb to sludge; 4% in effluent
                Mean
                (ug/kg dry wt)
                1,183
 La Guardia (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
                                                                    58,800 ug/kg dry wt
                                                                    (2002 measurement)
                                                                    37,400 ug/gram dry
                                                                    wt (2005
                                                                    measurement)
 U.S. EPA
 • 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
 International
De Boer etal. (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
March 2013
                                E-8
    DRAFT-DO NOT CITE OR QUOTE

-------
Table E-4, cont: BDE-209 concentrations in sewage effluent and sludge.
                                                                               BDE-209 levels
 Citation
                   Relevant study info
                                                                      STP effluent
Sewage sludge
 International
 Wang etal. (2007)
                    • Sewage sludge from 31 STPs in 26 cities in China
                    • BDE-209 was dominant congener in most samples
                                                                                      Mean (ng/gram
                                                                                      dry wt) 68.5
Clarke etal. (2008)
                     • Australian sewage sludge survey; 16 WWTPs (2006)
                     • Presented urban mean, rural mean, and overall
                      mean of BDE-209 in sludge samples
                                                                                      Mean (ug/kg dry wt)
                                                                                      880 (urban);
                                                                                      490 (rural)
                                                                                      720 (overall)
Kupper et al. (20J
Eljarratetal. (20C
Ricklundetal.
(2009)
)8) • Switzerland, monitoring network
• 16 WWTPs
17) • Spanish River Vero, samples collected up- and 1,170 ng/L (2005
downstream from an industrial park (2004, 2005) effluent maximum)
• 6 sediment and 3 effluent samples
• Stockholm, Sweden
• WWTP (2006, 2007)
Mean (ug/kg dry wt)
310
"
800 Mean
(ng/gram dry wt)
Note: Additional information obtained from U.S. EPA (201 Oa]
dry wt = Dry weight; STP = sewage treatment plant; WWTP = Waste water treatment plant
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 (2010a)
                    •  33 surface soil samples, 15 states
                    •  BDE-209 detected in 24/33 samples
                                                                 Mean (ng/gram dry wt): 15.3
Yunetal. (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)
March 2013
                                                   E-9
                                                                           DRAFT-DO NOT CITE OR QUOTE

-------
Table E-5, cont: BDE-209 concentration data in soil.
 Citation
Relevant study info
BDE-209 levels
 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 (201 Oa]
dry wt = Dry weight; SS = Surface soil; PS = Point source
March 2013
                                 E-10
           DRAFT-DO NOT CITE OR QUOTE

-------
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 jim, respectively, and condensation
                         particle counter counted all particles 10 nm-1  jim. 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
March 2013
                                E-11
  DRAFT-DO NOT CITE OR QUOTE

-------
Table E-6, cont: Proxy data for estimating MWCNT concentrations in occupational air.
 Citation
Relevant study info
Proxy data and CNT counts
 United States
 Dahmetal. (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 jig/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 (iig/m3) 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 (iig/m3) 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: respirable), not
 detected to 0.242 CNTs/cm3 (PBZ),
 not detected to 0.008 CNTs/cm3
 (area: inhalable)
 Milling MWCNT composite:
 not detected
March 2013
                                E-12
  DRAFT-DO NOT CITE OR QUOTE

-------
Table E-6, cont: Proxy data for estimating MWCNT concentrations in occupational air.
 Citation
Relevant study info
Proxy data and CNT counts
 United States
 Johnson et al. (2010)
 • Area air sampling and real-time monitoring of 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
 (1,000 nm), 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,OOP nm (particles/cm3)
 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
March 2013
                               E-13
  DRAFT-DO NOT CITE OR QUOTE

-------
Table E-6, cont: Proxy data for estimating MWCNT concentrations in occupational air.
 Citation
Relevant study info
Proxy data and CNT counts
 United States
 Methneretal. (2010)
 • Area air sampling and real-time monitoring of 2 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,OOP 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
                       • 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 (iig/m3)
                                                        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
March 2013
                                E-14
  DRAFT-DO NOT CITE OR QUOTE

-------
Table E-6, cont: Proxy data for estimating MWCNT concentrations in occupational air.
 Citation
Relevant study info
Proxy data and CNT counts
 International
 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
                                                         One filter sample detected MWCNTs
                                                         at 0.00312 tubes/cm3; all others were
                                                         non-detects

                                                         Total suspended particulate matter
                                                         Personal air: 7.8 - 320 jig/m3
                                                         Area air: 12.6-187 jig/m3

                                                         Particle number count for mode
                                                         particle sizes (particles/ cm3) - 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
 Takayaetal.(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
 jig/m3
 Personal air at manual facility: 2,390
 jig/m3 (total dust); 390 jig/m3
 (respirable dust)
 Personal air at automated facility: 290
 jig/m3 (total dust); 80 jig/m3
 (respirable dust)
March 2013
                                E-15
  DRAFT-DO NOT CITE OR QUOTE

-------
 E.2. Concentrations in Biota
1 The following tables provide details from studies that measured BDE-209 in different types of
2 biota. No data were identified on MWCNT concentrations in biota.
Table E-7. Measured concentrations of
Species Location Year
PBDEs in
Tissue
biota.
PBDE
burden
(ng/gram)1

Common
congener(s)

Source
Terrestrial birds
Peregrine falcon Chesapeake Bay 1993-2002
CT, MA, ME, NH, VT 1996,
1999-2006
Common Switzerland 2003-2005
blackbird
Sparrow hawk Switzerland 2003-2005
Belgium NR


NR NR
Common Switzerland 2003-2005
buzzard
Belgium NR
Beijing, China NR
Eggs

Brain
Adipose
TB
Brain
Adipose
TB
Liver
Brain
Adipose
Liver
TB
Liver
Liver
Median: 2012
Median: 4402
BDL
BDL
0.82
14
709
790.2
Mean: 4,900
Median: 1,300
Mean: 1,200
Median: 360
Mean: 1,900
Mean: 9,500
34.55
Mean: 480
Median: 70
148
BDE-153: 26%
BDE-153, BDE-
99
NA
NA
BDE-47: 100%
NR
NR
BDE-99: 40%
Top 3: BDE-99,
BDE-47, BDE-
153
NR
NR
Top 4: BDE-99,
BDE-47, BDE-
100, BDE-153
BDE-153: 29%
BDE-99: 23%
BDE-47: 22%
Top 3: BDE-153,
BDE-47, BDE-99
BDE-209: -43%
Potter etal. (2009)
Chen et al. (2008)
Naert et al. (2007) as
- cited in U.S. EPA
(2010a)
Naert etal. (2007) as
cited in U.S. EPA
(2010a)
Voorspoels et al.
(2006b)
Voorspoels et al.
(2006b)
Voorspoels et al.
(2006b)
Voorspoels et al. (2007)
as cited in U.S. EPA
(2010a)
Naert et al. (2007) as
cited in U.S. EPA
(2010a)
Voorspoels et al.
(2006b)
Chen et al. (2007a)
March 2013
E-16
DRAFT-DO NOT CITE OR QUOTE

-------
Table E-7, cont: Measured concentrations of PBDEs in biota.
Species Location
Year Tissue
PBDE
burden
(ng/gram)1
Common
congener(s)
Source
Terrestrial birds
Common NR
buzzard
Cormorant Switzerland
Owls Belgium
Common kestrel Beijing, China
Passerines NR
NR Liver
2003-2005 TB
NR NR
NR Muscle
Liver
Kidney
NR Adipose
Eggs
Mean: 720
98.76
250
Mean: 12,300
Mean: 12,200
Mean: 5,340
160
220
NR
BDE-47: 42%
Top 3: BDE-
153, BDE-99,
BDE-47
NR
NR
NR
NR
NR
Voorspoels et al. (2007) as
cited in U.S. EPA (2010a)
Naertetal. (2007) as cited
in U.S. EPA (2010a)
Voorspoels etal. (2006b)
Chenetal. (2007a)
Voorspoels et al. (2007) as
cited in U.S. EPA(2010a)
Marine birds
Herring gull Great Lakes
Fulmar Northern Canada
Murre Northern Canada
Heron 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 etal. (2002)

Wakeford et al. (2002) as
cited in U.S. EPA (2010a)
Wakefordetal. (2002) as
cited in U.S. EPA (2010a)
Wakefordetal. (2002) as
cited in U.S. EPA (2010a)
Mammals
Red fox Belgium
Ringed seals, Canadian Arctic
female
Ringed seals, Canadian Arctic
male
Arctic
Arctic
NR Adipose,
liver,
muscle
NR Blubber
NR Blubber
1981
2000
Median range:
2.2-3.4
Mean: 25.8
Mean: 50.0
0.6
6.0
Liver- BDE-209:
70%
Tetra, pentaBDE
Tetra, pentaBDE
Tetra, pentaBDE
Tetra, pentaBDE
Voorspoels et al.
(2006a)
Alaee etal. (1999)
Alaee etal. (1999)
Ikonomou etal. (2002)
Ikonomou etal. (2002)
March 2013
E-17
DRAFT-DO NOT CITE OR QUOTE

-------
Table E-7, cont: Measured concentrations of PBDEs in biota.
Species
Location
   Year
             PBDE
             burden
 Tissue    (ng/gram)1
                Common
               congener(s)
                      Source
Mammals
Beluga whales,   Canadian Arctic      NR
female        	
               St. Lawrence estuary NR
                             Blubber    Mean: 81.2    Tetra, pentaBDE  Alaeeetal. (1999)
                                       665
                                 NR
                                      Lebeuf et al. (2001) as
                                      cited in U.S. EPA
                                      (2010a)
Beluga whales,   Canadian Arctic      NR
male          	
               St. Lawrence estuary NR
               Baffin Island
                   1982
                                 1997
                             Blubber    Mean: 160     Tetra, pentaBDE  Alaeeetal. (1999)
                                       466
                                 NR
                                      Lebeuf et al. (2001) as
                                      cited in U.S. EPA
                                      (2010a)
           Blubber   2
                                       15
                       Tri to hexaBDE   Stern and Ikonomou
                                 Tri to hexaBDE
Harbor seals     San Francisco Bay   1989-1998  Blubber    Range:        Tetra, penta,     She etal. (2002)
                                                      88-8,325      hexaBDE
Harbor porpoise  Vancouver
                  NR
           Blubber   2,269
                       TetraBDE: >50%  Ikonomou et al. (2000)
Fish
Lake trout
Lake Ontario
1997
NR
434
NR
Lurossetal. (2002)

Rainbow trout
Mountain
whitefish
Largescale
sucker
Carp
Lake Erie
Lake Superior
Lake Huron
Spokane River, WA
Spokane River, WA
Columbia River,
British Columbia
Spokane River, WA
Virginia



1999 NR
1999 NR
1992-2000 Muscle
1999 NR
1998-1999 NR
117
392
251
2974
1.2504
Mean range
4.5-19.1
1054
1.1404



NR
NR
NR
NR
NR

Johnson and Olson
(2001)
Johnson and Olson
(2001)
Rayne etal. (2003)
Johnson and Olson
(2001)
Johnson and Olson
(2001)
March 2013
                                   E-18
                                        DRAFT-DO NOT CITE OR QUOTE

-------
Table E-7, cont: Measured concentrations of PBDEs in biota.
Species
Lower trophic
Caddisflies
Midges
Location Year
levels
Pyrenees Mountains, NR
Spain
Pyrenees Mountains, NR
Spain
Tissue

Larva TB
Pupa TB
Larva TB
Pupa TB
PBDE
burden Common
(ng/gram)1 congener(s)

Mean range: NR
0.65-13.003
Mean range
9.32- 273
Mean range: NR
0-13.07
Mean range
S.9-5.23
Source

Bartronsetal. (2007)
Bartronsetal. (2007)
1 ng/gram lipid weight, unless otherwise specified.
2 Units = ng/gram wet weight
3 Units = ng/gram dry weight
4 Measurement for a single fish only
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
Table E-8. Mean concentration of PBDEs in media/biota in an aquatic ecosystem.
Media Total PBDEs
Water (pg/L) 47.01
Sediment (ng/gram dry wt) 1 .31
BDE-47 BDE-99 BDE-100 BDE-153 BDE-209
16.98 9.01 1.89 1.02 
-------
Table E-8, cont: Mean concentration of PBDEs in media/biota in an aquatic ecosystem.

 Media                      Total PBDEs     BDE-47     BDE-99     BDE-100     BDE-153      BDE-209
Biota level 4 (mg/gram lipid wt)
Walleye
Burbot
54.39
240.32
16.21
44.37
2.56
20.48
2.34
10.49
1.98
12.12
24.72
98.68
 
-------
   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 fromNorthern 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/annhyg/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

   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
March 2013                                        E-21                    DRAFT-DO NOT CITE OR QUOTE

-------
   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

   Haa JH: Lee. EJ: Lee. JH: So. KP: Lee. YH: Bae. GN: Lee. SB: Ji. JH: Cho. MH: Yu. IJ. (2008a). 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/i.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. ArchEnviron Contain 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

   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/i.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: Covad 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.
March 2013                                        E-22                    DRAFT-DO NOT CITE OR QUOTE

-------
   Lee. JH: Lee. SB: Bae. GN: Jeoa KS: Yoon. JU: Ji. JH: Sung. JH: Lee. BG: Lee. JH: Yang. JS: Kim. HY: Kang.
      CS: Yu. IJ. (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: Ali. 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

   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

   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
March 2013                                         E-23                    DRAFT-DO NOT CITE OR QUOTE

-------
   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/reports/inthedust

   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-Mulloli. 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

   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/es05063 Iz

   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:  Hites. 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
March 2013                                        E-24                    DRAFT-DO NOT CITE OR QUOTE

-------
   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 Eiseigaku Zasshi 52: 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/settiements/publications/chemicals^fr/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). (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

   Voorspoels. S: Covaci. A: Jaspers. VLB: Neels. H: Schepens. P. (2007). Biomagnification of PBDEs in three
      small terrestrial food chains. Environ SciTechnol 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.

   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: 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

   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/i.chemosphere.2010.12.032

   Zhu. LY: Hites. 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
March 2013                                        E-25                    DRAFT-DO NOT CITE OR QUOTE

-------
   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
March 2013                                       E-26                    DRAFT-DO NOT CITE OR QUOTE

-------
      Appendix F. Toxicological and
             Ecological Effects
March 2013                            DRAFT-DO NOT CITE OR QUOTE

-------
      Appendix F.  lexicological  and
                                  Ecological  Effects
 1           Appendix F comprises tables summarizing the material characteristics, study design, and results
 2    of select toxicokinetic and toxicological studies for BDE-209 and MWCNTs. Much of the relevant
 3    toxicological data for BDE-209 has been summarized in reviews, and many of the toxicological endpoints
 4    of concern have been identified and used by regulatory agencies to establish reference values for the
 5    protection of human health and the environment (see Chapter 5). Because the BDE-209 studies have been
 6    described in detail in many previous reviews, a relatively small subset of the BDE-209 studies discussed
 7    in Chapter 5 of this document are summarized in the tables of this appendix. Select BDE-209 study
 8    summary tables are generally provided only for general comparison to MWCNT study summary tables or
 9    for the primary exposure route of concern (oral).
10           By comparison, the toxicokinetics and toxicological effects of MWCNTs are not well understood
11    and only one draft reference value has been established for the  protection of human health (see Chapter
12    5). Moreover, as emphasized throughout this document, variations in certain physicochemical
13    characteristics of MWCNTs are likely to affect their behavior in biological systems and impacts to
14    humans and biota. Appendix F therefore provides summary tables for most of the MWCNT toxicokinetic
15    and toxicological studies referenced in Chapter 5 of this case study. These tables supply more detailed
16    information on material characteristics, study design, and observed effects than was presented in the text.
17           Appendix F provides information on the absorption, distribution, metabolism, and elimination of
18    BDE-209 and MWCNTs from identified toxicokinetic studies (Section F.I.I). In addition, summary
19    tables present information from select in vivo studies reporting effects other than carcinogenicity using
20    the dermal, ocular, inhalation, and oral routes of exposure (Section F.I.2); in vitro studies including those
21    investigating genotoxicity and mutagenicity (Section F.I.3); and carcinogenicity studies (Section F.I.4)
22    for BDE-209 and MWCNTs.
23           Appendix F also provides summary tables of data from studies that  investigated effects of
24    BDE-209, other PBDEs, and MWCNTs on different types of biota. Data from studies in aquatic
25    ecosystems are summarized in Sections F.2.1 and F.2.2; data from terrestrial ecosystems are summarized
26    in Sections F.2.3 and F.2.4.
27           Literature was identified primarily using review articles published in the past two years. Targeted
28    literature searches were carried out as needed.
29
      March 2013                                    F-1                  DRAFT-DO NOT CITE OR QUOTE

-------
 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 day
8)


Doses
tested
0.0277%,
4.





1.


0.
0.
0.
2.
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 et al.
(1987)





el Dareer et al.
(1987)

el Dareer et al.
(1987)


% diet




March 2013
F-2
DRAFT-DO NOT CITE OR QUOTE

-------
Table F-1, cont: Select toxicokinetic studies for decaBDE.
Route of
exposure
Oral
(diet)





Oral (diet)





Oral
Oral
(diet)
Description of
decaBDE
[U-14C] decaBDE
(97.9-99.2% pure)
diluted with unlabeled
decaBDE




[U-i«C] 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
Exposure Doses
Species duration tested
Rat >8 days 250-50,000
(7 days ppm
unlabeled,
1 day labeled,
then returned
to unlabeled
diet for
remainder of
holding
period)
Rat >8 days 250-50,000
(7 days ppm
unlabeled,
1 day labeled,
then returned
to unlabeled
diet for
remainder of
holding
period)
Sprague- 16 days 1 mg/kg
Dawley rat
(male and
female)
Pregnant 96 hours 2.61 mg/kg-
Wistar rat (gestation day
(female) days 16-19)

Excretion results:





Excretion results:





Effects
observed
feces 61%, urine 0.1%





urine 0.01 %, feces >99% in 72 hours





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
>19% recovered in tissues; efficient absorption reported; 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
Citation
NTP (1986)





NTP (1986)





Morris etal.
(19Z5)
Riu etal.
(2008)

March 2013
F-3
DRAFT-DO NOT CITE OR QUOTE

-------
Table F-1, cont: Select toxicokinetic studies for decaBDE.
Route of
exposure
Oral
Oral
(diet)
Oral
(gavage)
Oral
(diet)
Oral
(diet)
Oral
Description of
decaBDE Species
77.4% decaBDE, Rat
21.8%nonaBDE,
and 0.8% octaBDE
77.4% pure Rat
>98% pure, Sprague-Dawley
specific rat (male)
activity 17.5Ci/mol
98.5% pure Sprague-
Dawley rat
(male)
NR Lactating cow
(female)
NR Sprague-
Dawley rat
(male)
Exposure
duration
NR
2 years
3, 7 days
21 days
followed by
21 -day
withdrawal
period
3 months
90 days
Doses
tested
NR
0; 0.01; 0.1; 1
mg/kg-day
3 umol/kg,
15Ci/mol,
1 ml/kg
volume
0.3 ug/gram
of diet
Naturally
contaminated
diet (not
measured)
100 mg/kg
bw-day
Effects
observed
Slight accumulation occurring very slowly over time in adipose tissue
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 TBA); 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
Great Lakes
(1976) and
IRDC(1976,
1977) as cited
in NRC (2000)
Dow (1994) as
cited in NRC
(2000)
Morketal.
(2003)
Huwe and
Smith (2007)
Kierkegaard et
al. (2007)
Wangetal.
(2010a)
March 2013
F-4
DRAFT-DO NOT CITE OR QUOTE

-------
Table F-1,
Route of
exposure
In vitro
(incubation)







Oral






Gavage
ori.v.


Liver
microsomal
depletion


cont: Select toxicokinetic studies for decaBDE.
Description of
decaBDE Species
98 ± 1 % pure Human
hepatocytes; 2
cryopreserved (1
male, 1 female), 1
fresh (male)




98% pure Sprague-
Dawley rat
(female)




Unlabeled BDE- Sprague-
209 (>98% Dawley rat
pure) (male)

BDE-209 Harbor seal
(Phoca vitulina)
and sperm whale
(Physeter catodon)

Exposure Doses
duration tested
48-hour 10nmol/well
exposure for
cryopreserved
cell cultures; 1
dose per 24
hours for three
days for fresh
hepatocyte
cultures
GD7 to PND4 5 umol/kg






1,3,6,24,48, 2umol/mL
72, 96, 120, or
144 hours

NR 31 ug/mL




Effects
observed
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






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
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)
No detectable depletion of parent BDE-209; lack of microsomal depletion
consistent with persistent and accumulative nature of BDE-209




Citation
Stapletonetal.
(2009)







Caietal.
(2011)





Sandholm et
al. (2003)


de Boer et al.
(1998, 2000)
as cited in
Hakk and
Letcher (2003)
March 2013
F-5
DRAFT-DO NOT CITE OR QUOTE

-------
Table F-1, cont: 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


Species
Sprague-
Dawley rat
(male)

Rainbow trout
(Oncorhynchus
mykiss) (male and
female)


Exposure
duration
72 hours



16, 49, 120
days




Doses
tested
3 mg/kg



1.7-10
mg/kg-day




Effects
observed
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
Low uptake efficiency; elevated levels in liver and 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)




 Gl = Gastrointestinal tract; NR = Not reported
March 2013
F-6
DRAFT-DO NOT CITE OR QUOTE

-------
Table F-2.    Select toxicokinetic studies for MWCNTs.
Route of Diameter Length
exposure (nm) (urn)
Surface
area
(m2/gram) Purity
Species
Exposure
duration
Doses Effects
tested observed
Citation
Absorption
Intratracheal 10-20 0.01-0.6
instillation;
i.v.; and
gavage
Inhalation 20-50 0.5-2
(intratracheal
instillation)
Inhalation Mean: Median:
(aspiration) 49 ±13.4 3.86
Oral 10-20 0.01-0.6
(gavage)
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
10 ug 20% of administered dose remained
within the lung at 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 , 1 0, 1 00 MWCNTs did not significantly cross
ug/rat the pulmonary barrier; MWCNTs
were evident within the lungs at 6
months
1 0, 20, 40, MWCNTs reached the pleura and
80 ug induced pleural inflammation at 56
days
10 ug Majority of MWCNTs evident in
feces, stomach, and small and large
intestines; no detectable transport
Dengetal.
(2007)
Elgrablietal.
(2008b)

Porter et al.
(2010)
Dengetal.
(2007)
                                                                                           into the blood; MWCNTs remained
                                                                                           unchanged suggesting
                                                                                           biopersistence if not excreted
March 2013
F-7
DRAFT-DO NOT CITE OR QUOTE

-------
Table F-2, cont: Select toxicokinetic studies for MWCNTs.
                                Surface
Route of    Diameter  Length      area
exposure      (nm)      (urn)    (rrWgram)     Purity
                                                           Species
Exposure
duration
Doses
tested
 Effects
observed
                                                                                                                 Citation
 Distribution
Inhalation 10-50 <0.1-10 NR NR



Inhalation NR 0.9-0.15 197 NR
(intratracheal
instillation)
C57BL6
mouse
(male)

Wistar
albino rat
(male)
Single 6-hour exposure;
14-week observation


Single exposure; 24
hour, 1 week, 1-and3-
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)

Reddyetal.
(2010)

 Elimination
 Inhalation     20-50
 (intratracheal
 instillation)
0.5-2    NR         (%w/w):0.53  Sprague-
                     Ni, 0.08 S,     Dawley rat
                     <0.02 Mg,     (male)
                     <0.01 Na,
                     0.01 V
                                                                   Single exposure; 1-, 7-,  1,10,100   Following phagocytosis of the
                                                                   30-, 90-, and 180-day   ug/rat      MWCNTs, the macrophages
                                                                   observations and
                                                                   6-month MWCNT
                                                          Elgrablietal.
                                                                     elimination observation
                          underwent apoptosis, with no
                          inflammatory response or other
                          physiological and histological
                          pathology
 NR = Not reported
March 2013
                            F-8
                                                                                                                   DRAFT-DO NOT CITE OR QUOTE

-------
F.1.2. In Vivo Studies (Excluding Carcinogenicity Studies)
Table F-3.   Select dermal and ocular studies for decaBDE.
Route of Description
exposure of decaBDE Species
Exposure Doses
duration tested
Effects
observed
Citation
Irritation
Dermal Dry solid New Zealand
albino rabbit
Dermal NR Rabbit
24 hours, 3 days, 2 500 mg,
weeks reported by
NRC (2000)
Single 200, 2000
administration 24 mg/kg
hours; 14 days
observation
No dermal response in intact skin; no
indication of bromacne
Slight erythematous and edematous
response in abraded skin
Norris et al. (1975) [also reported in
NTP (1986): Norrisetal. (1973):
Dow (1972) and IRDC (1974) as cited
in NRC (2000)1
IRDC (1974) and Great Lakes (1977)
as cited in NRC (2000): Norris et al.
(1975)
Sensitization
Dermal Homogenous 5% Human
suspension in
petrolatum; 77.4%
decaBDE, 21.8%
nonaBDE, 0.8%
octaBDE
3 times per week for NR
3 weeks
No skin sensitization response
Norris et al. (1975) [also reported in
NTP (1986); Norrisetal. (1973);
Dow (1972) as cited in NRC (2000)1
Systemic effects
Dermal NR Rabbit
Single 200, 2,000
administration 24 mg/kg
hours observation
No treatment-related effects in body
weight gain or survival
IRDC (1974) and Great Lakes (1977)
as cited in NRC (2000)
March 2013
F-9
DRAFT-DO NOT CITE OR QUOTE

-------
Table F-3. cont.:, Select dermal and ocular studies for decaBDE.
  Route of
 exposure
Description
of decaBDE
Species
Exposure
duration
Doses
tested
 Effects
observed
Citation
 Ocular
 Ocular      Saytex 102
                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 in
                               albino rabbit                          eye          membranes in washed and unwashed    NTP (1986): Norris et al. (1973): IRDC
                                                                                eyes (not sustained past 24 hours)       (1974) Dow (1972) as cited in NRC
                                                                                                                   (2000)1

NR = Not reported
March 2013
                                    F-10
                                                                                 DRAFT-DO NOT CITE OR QUOTE

-------
Table F-4. Select dermal and ocular studies for MWCNTs.
Test Diameter
substance (nm)
Surface
Length area
(urn) (nWgram) Purity
Species
Exposure
duration
Doses Effects
tested observed
Citation
Skin irritation
1%Nikkiso- 44
MWCNTs
2% Mitsui 60
product of
MWCNTs
MWCNTs Inner: 3-8,
outer: 140
±30
MWCNTs Inner: 2-6,
outer: 10-
15
NR 69 176ppmGa, 80
ppm Al, 53 ppm
Fe, 16ppmCd,
0.5 ppm Li
NR 23 3,600 ppm Fe, 14
ppm Cr, 6 ppm Bi,
4 ppm Ni
5-9 10-15 NR
0.1-10 30-45 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-,
24-, 48-, and 72-hour
observation
4-hour exposure; 1-,
24-, 48-, and 72-hour
observation
4-hour under semi-
occlusive conditions;
96-hour observation
4-hour under semi-
occlusive conditions;
96-hour observation
0.5 gram Exposure resulted in a
primary irritation index of
0.6
0.5 gram No erythema or edema
was observed
0.5 gram No erythema or edema
at 72 hours
0.5 gram No erythema or edema
at 72 hours
Ema et al. (2011)
(OECD 404 compliant;
not GLP compliant)
Ema et al. (2011)
(OECD 404 compliant;
not GLP compliant)
Kishoreetal. (2009)
(OECD 404 compliant)
Kishoreetal. (2009)
(OECD 404 compliant)
March 2013
F-11
DRAFT-DO NOT CITE OR QUOTE

-------
Table F-4, cont: Select dermal and ocular studies for MWCNTs
Test Diameter Length
substance (nm) (urn)
Surface
area
(nWgram) Purity
Species
Exposure
duration
Doses
tested
Effects
observed
Citation
Skin sensitization
1%Nikkiso- 44 NR
MWCNTs
2% Mitsui 60 NR
product of
MWCNTs
69 176ppmGa, 80
ppm Al, 53 ppm
Fe, 16ppmCd,
0.5 ppm Li
23 3, 600 ppm Fe, 14
ppm Cr, 6 ppm Bi,
4 ppm Ni
Sic: Hartley
guinea pig
(male)
Sic: 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 NR
MWCNTs
2% Mitsui 60 NR
product of
MWCNTs
69 176 ppm Ga, 80
ppm Al, 53 ppm
Fe, 16ppmCd,
0.5 ppm Li
23 3,600 ppm Fe, 14
ppm Cr, 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
Emaetal. (2011)
(OECD 405 compliant)
Emaetal. (2011)
(OECD 405 compliant)
 NR = Not reported
March 2013
F-12
DRAFT-DO NOT CITE OR QUOTE

-------
Table F-5.    Select inhalation studies for decaBDE.
Route of
exposure
Inhalation
(Intratracheal
injection)
Inhalation
Description
of decaBDE
77.4% purity
decabromodiphenyl
oxide (DBDPO) dust,
respirable size
DE-83, 97% purity
aerosolized dust
Species
Sprague-Dawley
rat (male)
Spartan rat (male
and female)
Exposure
duration
3,10,30,91,365,
416, 556 days
1 hour (observed
for 14 days)
Doses
tested
20 mg/mL rat
serum
2 or
48.2 mg/L 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
Citation
Dow Chemical Co. (1990b)
Great Lakes Chemical Corporation
(1994)andlRDC(1974)
                                                                                   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.
March 2013
F-13
DRAFT-DO NOT CITE OR QUOTE

-------
Table F-6. Select inhalation studies for MWCNTs.


Route of Diameter
exposure (nm)
Surface
area
Length (nWgram
(urn) ) Purity Species


Exposure Doses
duration tested


Effects
observed


Citation
Acute or subacute
Inhalation 1,900-
(nose only) 2,900

Inhalation 50
(intratracheal
instillation)
Inhalation 50
Inhalation 10-20
(whole body)

Inhalation 10-20
(whole body)

~<1 253 98.6% (bulk) Wistarrat
and 99.1% (male)
(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 1 1, 241 mg/m3
for 6 hours; 7 -,28-,
90-day observation

Single exposure; 1 .7 mg/kg
8-, 16-, 24-day
observation
6 hours/day; 32.61 mg/m3
5, 10, 15 days
6 hours/day; 0.3, 1 , 5.3 mg/m3
7, 14 days

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)
Lietal.
(2007)
Lietal.
(2007)
Mitchell et
al. (2007)

Mitchell et
al. (2009)

March 2013
F-14
DRAFT-DO NOT CITE OR QUOTE

-------
Table F-6, cont: Select inhalation studies for MWCNTs.
Route of Diameter
exposure (nm)
Length
(Mm)
Surface
area
(rrWgram
) Purity Species
Exposure
duration
Doses
tested
Effects
observed
Citation
Acute or subacute
Inhalation NR
(intratracheal
instillation)
Inhalation 11.3
(intratracheal
instillation)
Inhalation NR
(intratracheal
instillation)
Intranasal 15.04 ±
injection 0.47
Inhalation 11-170
(intratracheal
instillation)
NR
0.7
NR
0.5-200
(reported
by
supplier)
5-9
NR NR SD rat
(female)
NR 98%; traces of Wistarrat
Co and Fe (female)
catalysts
NR NR Wistar rat
(female)
139.7 >90% carbon BALB/cAnNCr
(as reported I mouse
by supplier) (female)
12.83 >90% carbon ICR mouse
(male)
Single exposure;
3-, 15-, 28-, 60-day
observation
Single exposure;
3-day observation
Single exposure;
3- and 60-day
observation
3 days
Single exposure;
1-, 3-, 7-, 14-day
observation
2.2, 8.9, 22.2
mg/kg
0.5, 2, 5 mg/rat
2 mg/rat
200, 400
ug/mouse (with 10
ug OVA per
injection and 10ug
OVA booster given
at 2 1,22, and 23
days)
5, 20, 50 mg/kg
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 IgE in serum and
inflammatory cells in BALF
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
Mulleretal.
(2005)
Mulleretal.
(2008a)

Mulleretal.
(2008b)

Nygaard et
al. (2009)

Park et al.
(2009)
March 2013
F-15
DRAFT-DO NOT CITE OR QUOTE

-------
Table F-6, cont: Select inhalation studies for MWCNTs.
Surface
area
Route of Diameter Length (nWgram
exposure (nm) (urn) ) Purity
Inhalation 30-50 0.3-50 109.29 >94%
(nose-only)







Inhalation 10-15 -20 NR 95%








Species
C57BL/6
mouse (male)







Sprague-
Dawley rat
(male)






Exposure
duration
Single exposure; 6
hours; 1-, 14-day
observation






6 hours/day;
5 days; 1 month
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
ofTGF-|31andlL-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 etal.
(2012)





Subchronic
Inhalation 50 10 280 >95%


Inhalation 5-15,500- 0.1-1 250-300 90%
(head-nose) 1,300,
1,300-
2,000/900-
1,500,700-
800
Kunming
mouse
(female)
Wistar rat
(male and
female)



6 hours/day;
30, 60 days;

6 hours/day;
13 weeks




32.61 mg/m3


0.1,0.4, 2.5 mg/m3





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)


March 2013
F-16
DRAFT-DO NOT CITE OR QUOTE

-------
Table F-6, cont: Select inhalation studies for MWCNTs.

Route of
exposure

Diameter
(nm)
Surface
area
Length (nWgram
(Mm) )


Purity


Species

Exposure
duration

Doses
tested

Effects
observed


Citation
 Subchronic
Inhalation 67
(intratracheal
instillation)



Inhalation 1,900-
(nose only) 2,900



3-30 26 99.79%





~<1 253 98.6% (bulk)
and 99.1%
(micronized)


I CR mouse
(male)




Wistar rat 6 hours/day;
(male and 5 days/week;
female) 13 weeks


25, 50 ug/week/
mouse (both with
and without 1 ug
OVA/2 wk)


0.1,0.4,1.5,6
mg/m3



Increased total cells in BALF, Inoue et al.
infiltration of inflammatory leukocytes (2009)
in airways, induction of goblet cell
hyperplasia in both groups,
enhanced response in sensitized
group
Sustained pulmonary inflammation at Pauluhn
>1 .5 mg/m3; granulomas and (2010)
alveolar hyperplasia at >6 mg/m3; no (OECD 413
systemic toxicity; 0. 1 mg/m3 was compliant)
NOAEC
NR = Not reported
March 2013
F-17
DRAFT-DO NOT CITE OR QUOTE

-------
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 Morris etal. (1973)1
Subchronic
NR 97-99% pure
Intragastric >98%pure
Oral 94-98% pure
(diet)
Oral 94-98% pure
(diet)
Rat (male and
female)
Wistar rat
(female)
F344/N rat (male
and female)
B6C3Fi mouse
(male and
female)
28 days
7-28 days
14 days
14 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,
1 00,000 ppm
0, 50,00, 10,000,
20,000, 50,000,
100,000 ppm
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
Great Lakes (1976);
IRDC (1976) as cited in
NRC (2000)

Bruchajzer etal. (2010)
NTP (1986)
NTP (1986)
March 2013
F-18
DRAFT-DO NOT CITE OR QUOTE

-------
Table F-7,
Route of
exposure
cont: Select oral
Description of
decaBDE
and intragastric
Species
studies for decaBDE.
Exposure Doses
duration tested

Effects
observed

Citation
Subchronic
Oral
(gavage)
Oral
(gavage)
Oral
(diet)
Oral
(diet)
Oral
(diet)
97% pure
97% pure
77.4% decaBDE,
21.8%nonaBDE,
0.8% octaBDE
94-98% pure
94-98% pure
Wistar rat
(male)
Wistar rat
(female)
Sprague-
Dawley rat
(male)
F344/N rat (male
and female)
B6C3Fi mouse
(male and
female)
28 days 0,1.9,3.8,7.5,15,30,
60 mg/kg
28 days 0,1.9,3.8,7.5,15,30,
60 mg/kg
30 days 0, 8, 80, 800 mg/kg-
day
13 weeks 0, 3,100; 6,200;
12,500; 25,000,
50,000 ppm
13 weeks 0,3,100,6,200,
12,500, 25,000,
50,000 ppm
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)
Decreased activity of P450c17 (BMDL 0.18 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
Van der Ven et
al. (2008)
Van der Ven et
al. (2008)
Norris et al.
(1975)
NTP(1986)
NTP(1986)
Chronic
Oral
(diet)
Oral
(diet)
77.4% decaBDE,
21.8%nonaBDE,
0.8% octaBDE
94-98% pure
Sprague-
Dawley rat (male
and female)
F344/N rat
(male)
2 years 0, 0.01, 0.1, 1.0 mg
2 years 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, 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)
Norris et al.
(1975)
NTPQ986)
March 2013
F-19
DRAFT-DO NOT CITE OR QUOTE

-------
Table F-7,
Route of
exposure
cont: Select oral and intragastric
Description of
decaBDE Species
studies for
Exposure
duration
decaBDE.
Doses
tested

Effects
observed

Citation
Chronic
Oral
(diet)
Oral
(diet)
Oral
(diet)
Oral
(diet)
Oral
(diet)
Oral
(diet)
Oral (metal
gastric
tube)
94-98% pure F344/N rat
(female)
94-98% pure B6C3Fi mouse
(male)
94-98% pure B6C3Fi mouse
(female)
94-98% pure F344/N rat (male)
94-98% pure F344/N rat
(female)
77.4% pure Sprague-Dawley
rat (male and
female)
>99% NMRI mice
(male)
2 years
2 years
2 years
2 years
2 years
2 years
single dose
given at age
3, 10, or 19
days
0, 2,500, 50,000 ppm
0, 2,500, 50,000 ppm
0, 2,500, 50,000 ppm
1,120, 2,240 mg/kg
(adjusted)
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, or
20.1 mg/kg-day for 10
day old mice
Increased incidence of neoplastic lesions (nodules in the liver
of high-dose group) and nonneoplastic lesions; degeneration
of the eye in low dose group
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; enzyme induction; significant
increases in hepatic CYP1A mRNA, CYP2B mRNA, CYP1A1
protein, and 7-pentoxyresorufin 0-dealkylase activity;
increased 7-ethoxyresofurin 0-deethylase activity
Increased 7-ethoxyresofurin 0-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 or PND19
NTP(1986)
NTP(1986)
NTP(1986)
NTP(1986)
NTP(1986)
Morris (1973):
Dow (1994) and
Kociba et al.
(1975) as cited
in NRC (2000)
Vibergetal.
(2003)
March 2013
F-20
DRAFT-DO NOT CITE OR QUOTE

-------
Table F-8.     Select intubation and injection studies for MWCNTs.

Route of
exposure

Diameter Length
(nm) (um)
Surface
area
(m2/
gram)


Purity

Exposure Doses
Species duration tested

Effects
observed Citation
 Acute/Subacute
 intra-
 peritoneal
 injection
15-30
15-20
>95%;         Swiss-        daily, 5 days       0.25,0.5,0.75
functionalized   Webster mice                   mg/kg-day
(2-7% COOH)
                                                Dose-related increase in ROS 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
 (gavage)
10-15
 -20    NR
-95%
Sprague-
Dawley rat
(pregnant
dams)
Single dose on
GD6-GD19
of 0, 40, 200, and
1,000 mg/kg-day
Dose-dependent decrease in
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
Limetal.
(2011)
March 2013
                                      F-21
                                                                                            DRAFT-DO NOT CITE OR QUOTE

-------
Table F-8, cont: Select intubation and injection studies for MWCNTs.

Route of
exposure

Diameter
(nm)

Length
(Mm)
Surface
area
(m2/gram)


Purity

Exposure Doses
Species duration tested

Effects
observed Citation
 Developmental Studies
 Intraperitoneal  Width distribution from 70-170 nm with the         ICRmice      Single dose on
               greatest frequency occurring at 90-110 nm, length  (pregnant      GD9; fetuses
               distribution between 1 -19 urn with the greatest    dams)         examined on
               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, 2, 3, 4, or 5      Dose-dependent decreased maternal
                                      mg/kg-bw         body weight; increased number of
                                                        resorptions, decreased 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)
 Intratracheal    Width distribution from 70-170 nm with the         ICRmice      Single dose on
 injection       greatest frequency occurring at 90-110 nm, length  (pregnant      GD9; fetuses
               distribution between 1 -19 urn with the greatest    dams)         examined on
               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 5 mg/kg-
                                      bw)
Decreased final body weight of 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)
March 2013
F-22
             DRAFT-DO NOT CITE OR QUOTE

-------
F.1.3.  Genotoxicity,  Mutagenicity, and Other In Vitro Studies
Table F-9.    Select genotoxicity, mutagenicity, and other in vitro studies for MWCNTs.
   Assay
               Surface
                 area
Diameter   Length    (m2/
  (nm)    (urn)   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
110-170 5-9
100 to 0.2-1
>150
(bi modal
distribution)
130 >90%;<0.1%Fe;
residual amorphous
carbon; other metal
contaminants
NR >95%; no free
amorphous carbon
Salmonella
typhimurium
strains TA 98, TA
100; Escherichia
co//' strain
WP2uvrA
Salmonella strains
TA1535, TA100,
TA1537, TA98,
TA102
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
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
DiSottoetal.
(2009)

Wirnitzer et al.
(2009)
(OECD471)
March 2013
                             F-23
                                                       DRAFT-DO NOT CITE OR QUOTE

-------
Table F-9, cont: Select genotoxicity, mutagenicity, and other in vitro studies for MWCNTs.
Assay
Diameter
(nm)
Surface
area
Length (m2/
(urn) gram) Purity
Species
Doses
tested
Effects
observed
Citation
DMA damage and unscheduled DMA synthesis
Chromosome
aberration
Comet assay
Comet assay
Comet assay
Sister chromatid
exchange;
micronucleus
assay
Chromosome
aberration
88 ±5
110-170
20-40
15-30
10-30
100 to
>150
(bimodal
distribution)
5 ±4.5 NR NR
5-9 22 >98%;<0.1%Fe;
~1.5%Ni;other
metal catalysts
0.5-200 NR 93.37%
15-20 NR >95%
1-2 NR 95-98%
0.2-1 NR >95%;nofree
amorphous carbon
Chinese hamster
lung cells
Murine
macrophage cell
line RAW 264.7
human lung
epithelial A549
cells
normal human
dermal fibroblast
cells (NHDF)
Human
lymphocyte cells
V79 cells
0.078,0.31,1.4,
5.0, 20, 80 ug/mL, -
S9
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
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
Significantly higher percent (%) DNA in
comet tails for doses >1 ug/mL; significant
dose-related effect overall
Significantly higher perrcent (%) 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
Asakuraetal.
(2010)
Migliore et al.
(2010)
Cavallo et al.
(2012)
Patlolla et
al.(2010b;
2010a).
Szendi and
Varga (2008)
Wirnitzer et al.
(2009)
(OECD 473)
March 2013
F-24
DRAFT-DO NOT CITE OR QUOTE

-------
Table F-9, cont: Select genotoxicity, mutagenicity, and other in vitro studies for MWCNTs.
Assay
Diameter
(nm)
Surface
area
Length (m2/
(um) gram)
Purity
Species
Doses Effects
tested observed
Citation
DMA damage and unscheduled DMA synthesis
Sister chromatid
exchange
Comet assay (in
vivo intra-
tracheal
instillation; 3
hours)
DMA adduct
assay (in vivo
intra-tracheal
instillation; 3,
24, 72, 168
hours)
DMA damage
via Western blot
>80% of
particles
70-110,
Gaussian
peak at 90
>80% of
particles
70-110,
Gaussian
peak at 90
>80% of
particles
70-110,
Gaussian
peak at 90
NR
>70% NR
of
particle
s1-4,
peak: 2
>70% NR
of
particle
s1-4,
peak: 2
>70% NR
of
particle
s1-4,
peak: 2
NR NR
3,500, 470, and 20
ppm iron, sulfur,
and chlorine
contaminants
3,500, 470, and 20
ppm iron, sulfur,
and chlorine
contaminants
3,500, 470, and 20
ppm iron, sulfur,
and chlorine
contaminants
NR
Chinese hamster
ovary AA8 eel Is
Male ICR mice (6
weeks old); lung
cells
Male ICR mice (6
weeks old); lung
DNA
Mouse embryonic
stem cells
0.1, 1.0, 2.0 ug/mL Significant increase in sister chromatid
exchange frequency at all doses;
approximately 3-fold increase over
controls at 1.0 ug/mL
Single doses of 0.05 Significant, dose-dependent increase in
or 0.2 mg/animal DNA damage observed by significant
increases in DNA tail moment and
percentage of DNA in the tail compared to
controls.
Single dose of 0.2 Three (out of four analyzed) DNA adducts
mg/animal 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.
100 ug/mL 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 of XRCC4
Katoetal. (In
Press): CNT
characteristics
in Sakamoto et
al (2009)

Katoetal. (In
Press): CNT
characteristics
in Sakamoto et
al (2009)

Katoetal. (In
Press): CNT
characteristics
in Sakamoto et
al (2009)

Zhuetal.
(2QQ7)
March 2013
F-25
DRAFT-DO NOT CITE OR QUOTE

-------
Table F-9, cont: Select
Assay
Diameter
(nm)
genotoxicity, mutagenicity, and other in vitro studies for MWCNTs.
Length
(Mm)
Surface
area
(m2/
gram) Purity
Species
Doses
tested
Effects
observed
Citation
Mutation
Micronucleus
assay
Mutation at
hgprt locus
Micronucleus
assay
Micronucleus
assay
Mutation assay
(in vivo intra-
tracheal
instillation; 8 -
12 weeks)
Micronucleus
assay
88 ±5
88 ±5
20-40
>80% of
particles
70-110,
Gaussian
peak: 90
>80% of
particles
70-110,
Gaussian
peak at 90
110-170
5 ±4.5
5 ±4.5
1-5
>70%
of
particle
s1-4,
peak: 2
>70%
of
particle
s1-4,
peak: 2
5-9
NR NR
NR NR
NR >99%wt
NR 3,500, 470, and
20 ppm iron,
sulfur, and
chlorine
contaminants
NR 3,500, 470, and
20 ppm iron,
sulfur, and
chlorine
contaminants
22 >98%;<0.1%Fe;
~1.5%Ni;other
metal catalysts
Chinese hamster
lung cells
Chinese hamster
lung cells
Human blood cells
Human lung
carcinoma A549 calls
Male guanine
phosphoribosyl-
transferase (gpf)
mice (9 weeks old)
Murine macrophage
cell line RAW 264.7
0.02,0.078,0.31,
1.3, 5.0ug/mL, -S9
6.3, 12.5, 25, 50,
100ug/mL, -S9
0.25-1 50 uL/5 ml
total cell culture
volume, -S9
20, 100,200ug/mL
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
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)
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
Asakuraetal.
(2010)
Asakuraetal.
(2010)
Cveticanin et
al. (2010)
Katoetal. (In
Press): CNT
characteristics
in Sakamoto et
al (2009)

Katoetal. (In
Press): CNT
characteristics
in Sakamoto et
al (2009)

Migliore et al.
(2010)
March 2013
F-26
DRAFT-DO NOT CITE OR QUOTE

-------
Table F-9, cont: Select
Diameter
Assay (nm)
genotoxicity, mutagenicity, and other in vitro studies for MWCNTs.
Surface
Length area (m2/
(urn) gram) Purity Species
Doses
tested
Effects
observed
Citation
Mutation
Micronucleus 11.3
assay
platelet NR
aggregation
(Aggro-Link
data reduction
system;
Chronolog)
T-cell viability; 20-40
Trypan Blue
exclusion assay
0.7 NR 98%; traces of Co Rat lung epithelial
and Fe cells
NR NR NR human platelet cells
1-5 NR 95% Jurkat T lymphocyte
(leukemia) cells from
healthy human blood
donors
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
1,10ng/cell(40,
400 ug/mL); Cell
aliquots collected at
0, 24, 48, 72, 96,
Significant increase in micronuclei
concentration-dependent increase in
platelet aggregation; activation of
GPIIb/llla
time-dependent decrease in the viability of
Jurkat T leukemia cells; increased number
of cells staining with annexin V indicating
increased apoptosis
Mulleretal.
(2QQ8a)
Radomski et
al. (2005)
Bottini et al.
(2006)

                                                                                 120 hours post-
                                                                                 exposure; stained
                                                                                 for 5 min with
                                                                                 Trypan Blue to
                                                                                 determine cell
                                                                                 proliferation and
                                                                                 percentage of
                                                                                 apoptotic Jurkat or
                                                                                 peripheral blood
                                                                                 lymphocytes (PBL)
                                                                                 determined
                                                                                 using annexin V-
                                                                                 FITC
NR = Not reported
March 2013
F-27
DRAFT-DO NOT CITE OR QUOTE

-------
F.1.4. Carcinogenicity Studies
Table F-10.   Select carcinogenicity studies for decaBDE.
   Route of     Description of                     Exposure         Doses
  exposure       decaBDE         Species        duration          tested
                                                                                     Effects
                                                                                    observed
                                                                                  Citation
 Oral
94-98% pure     F344/N rat (male)
2 years    1,120, 2,240 mg/kg-
          day (adjusted)
Some evidence of carcinogenicity; increased
incidences of neoplastic nodules of the liver (low
dose 7/50, high dose 15/49, control 1/50)
NTPM986
 Oral         94-98% pure      F344/N rat
                              (female)
                                    2 years    1,120, 2,550 mg/kg-
                                              day (adjusted)
                              Some evidence of carcinogenicity; increased
                              incidences of neoplastic nodules of the liver (low
                              dose 3/49, high dose 9/50, control 1/50)
                                             NTPM986
 Oral         94-98% pure      B6C3Fi mouse
                              (male)
                                    2 years    25,000, 50,000 ppm
                               Equivocal evidence of carcinogenicity; increased
                               incidences of hepatocellular adenomas or
                               carcinomas (combined) in both dose groups
                                             NTPM986
 Oral         94-98% pure      B6C3Fi mouse
                              (female)
                                    2 years    25,000, 50,000 ppm    No evidence of carcinogenicity
                                                                           NTPM986
 Oral         77.4% decaBDE,   Sprague-Dawley
             21.8% nonaBDE,   rat (male and
             0.8% octaBDE     female)
                                    2 years    0,0.01,0.1,1.0
                                              mg/kg-day
                               No alterations in appearance, behavior, bodyweight,  Kociba et al. (1975) as
                               feed consumption, hematologic analyses, urinalysis,  reported in NTP (1986)
                               clinical chemistry, organ weights, survival, or tumor   and NRC (2000)
                               incidence
March 2013
                                      F-28
                                                                   DRAFT-DO NOT CITE OR QUOTE

-------
Table F-11.   Select carcinogenicity studies for MWCNT.


Test Diameter Length
substance (nm) (urn)
MWCNTs 11. 3 ±3.9 -0.7
with
structural
defects



MWCNTs 11. 3 ±3.9 -0.7
without
structural
defects
Short 14.84± 1-5
MWCNTs 0.50



Short 10.40 ± 5-20
MWCNTs 0.32


Long 165.02 ± Maxim
MWCNTs 4.68 urn 56



Surface
area
(m2/
gram) Purity
299 1.97%AI,
0.49% Fe,
0.48% Co




190 0.37% Al,
<0.01%Fe,
<0.01%Co

NR 7.9% Fe,
5.1%Cu,
9.7% Ni,
5.5% Zn,
3.7% Co
NR (ug/gram)
13.4Fe,
1 Cu, 5 Ni,
7.5 Zn
NR (ug/gram)
37.3 Fe,
1.2Cu,
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 Effects
tested observed
2, 20 2 mesotheliomas at low dose (1 at 20 months and 1 at
mg/rat 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
2, 20 No tumors observed at low dose; 3 mesotheliomas (at
mg/rat 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
100 ug/mL No significant effects




100 ug/mL No significant effects



100 ug/mL Inflammatory responses; formation of granulomas on
peritoneal surface of the diaphragm; foreign body giant
cells comparable to long-fiber amosite asbestos





Citation
Mulleret
al. (2009)





Mulleret
al. (2009)


Poland et
al. (2008)



Poland et
al. (2008)


Poland et
al. (2008)



March 2013
F-29
DRAFT-DO NOT CITE OR QUOTE

-------
Table F-11, cont.: Select carcinogenicity studies for MWCNTs.


Test
substance
MWCNTs










MWCNTs
MWCNTs






Diameter
(nm)
70-110,
peak at 90




10-30





>80% of
particles
70-110,
Gaussian
peak at 90


Length
(Mm)
72.5%
within
1-4,
peak at
2

1-2





>70%
of
particle
s1-4,
peak: 2
Surfac
earea
(m2/
gram) Purity
NR (ppm) 3,500
Fe, 470 S, 20
Cl, <5 Br, <40
F


NR 95-98%





NR 3,500, 470,
and 20 ppm
iron, sulfur,
and chlorine
contaminants



Species
p53 (+/-)
mouse
(male)



F344 rat





F344 rat
(male)





Exposure
duration
Single injection,
25-week
observation



Single
exposure; 12-
month
observation


single
intrascrotal
injection;
observed at 52
weeks


Doses Effects
tested observed
3 mg/mL 100% mortality by week 25; mesothelioma incidence
14/16 (87.5%, 1 1 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 mg/rat Granulomatous reactions of foreign body type with
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)


 NR = Not reported
March 2013
F-30
DRAFT-DO NOT CITE OR QUOTE

-------
 F.2. Ecological Effects
F.2.1. Effects of DecaBDE and Other PBDEs on Aquatic Receptors
TableF-12.
Effects of exposure to PBDEs in aquatic invertebrates.
Life Number/ Test type/ Test Doses
stage group duration substance Medium tested

Effect
Effect dose

Notes
Algae1
NR
NR
NR Semi-chronic/ Commercial NR NR
96 hours decaBDE
NR Semi-chronic/ Commercial NR NR
96 hours pentaBDE
ECso >1 mg/L
NOEC >water solubility
Marine algae; review article, study-
specific details were not provided
Freshwater algae; review article, study-
specific details were not provided
Zebra mussels (Dreissena polymorpha)2
Post-
spawn
Comet assay/
48, 96, 168 T , . . ,
. ' Technical grade
150 micronucleus decaBDE, 98% Water 0.1, 2, 10 ug/L
(MN) assay/ pure
48, 96 hours
Mortality, hemocyte NOE
viability
DNA damage, All tested doses
chromosomal
aberrations
None
Increasing trend over time @ 0.1, 2 ug/L;
damage increased until 96hr then
decreased at 168 hours @10 ug/L
Freshwater oligochaete (Lumbriculus variegates)3
NR
k.n Chronic/28 Commercia 0 , . .....
NR , , Dni_ Sediment NR
days decaBDE
ECso >5,000 mg/kg
NOEC >5,000 mg/kg
Review article, endpoints not specified
Review article, endpoints not specified
March 2013
F-31
DRAFT-DO NOT CITE OR QUOTE

-------
Table F-12, cont.: Effects of exposure to PBDEs in aquatic invertebrates.
Life
stage
Number/ Test type/
group duration
Freshwater oligochaete (Lumbriculus
NR
Adult
Adult
NR Chronic/28
days
on Chronic/28
oU .
days
80 Chronic/28
days
Test
substance
variegates)3
Commercial
pentaBDE
55%pentaBDE,
36% tetraBDE,
8.5% hexaBDE
Mixture, 97%
decaBDE
Doses
Medium tested

Sediment NR
c , , 0,3.1,6.3,13,
Sediment ' ' „
25, 50 mg/kg
Sediment 0, 313, 625,
1,250,2,500,
5,000 mg/kg
Effect

ECso
ECso, survival,
reproduction
LOEC, survival,
reproduction
NOEC, survival,
reproduction
Effect
dose

>50 mg/kg
>50 mg/kg
6.3 mg/kg
>5,000 mg/kg
(mean measured
3,841 mg/kg)
Notes

Review article, endpoints not specified
Review article, study-specific details
were not provided
Review article, study-specific details
were not provided
Review article, doses listed are minimal
measured cone.
Water fleas (Daphnia magna)4
24 hours
old at
test start
24 hours
old at
test start
40 Acute/96 hours
40 Chronic/21 days
55%pentaBDE,
34% tetraBDE,
12% hexaBDE
55%pentaBDE,
34% tetraBDE,
12% hexaBDE
NR 0,1.4,2.6,5.3,
9.8, 20 ug/L
... , 0,1.4,2.6,5.3,
Water 9.8, 20 ug/L
ECso mortality,
immobility
ECso mortality,
immobility
ECso reproduction
LOEC mortality,
immobility
LOEC growth
17 ug/L
14 ug/L
14 ug/L
20 ug/L
9.8 ug/L
Review article, study-specific details
were not provided
Review article, study-specific details
were not provided
Review article, study-specific details
were not provided
Review article, study-specific details
were not provided
Review article, study-specific details
were not provided
March 2013
F-32
DRAFT-DO NOT CITE OR QUOTE

-------
Table F-12, cont.: Effects of exposure to PBDEs in aquatic invertebrates.
Life
stage
Number/
group
Test type/
duration
Test Doses
substance Medium tested

Effect
Effect
dose



Notes


Water fleas (Daphnia magna)4


24 hours
old at
test start


NR

NR

NR



20


NR

NR

NR



Chronic/21


Chronic/21



days


days

Acute/48 hours

Chronic/21


days


42% heptaBDE,
36%octaBDE, 0,0.13,0.25,
14%nonaBDE, Water 0.54,0.83,1.7
6% hexaBDE, ug/L
2% decaBDE

Commercial NR NR
octaBDE
Commercial NR NR
pentaBDE
Commercial NR NR
pentaBDE
NOEC survival,
reproduction,
growth
ECso, LOEC
survival,
reproduction,
growth
ECso, NOEC
>1.7ug/L


>2.0 ug/L


>water solubility


- Review article,
were not provii


Review article,


study-specific
Jed


study-specific


details


details
were not provided
ECso
14 ug/L
Review article,
study-specific
details
were not provided
LOEC

9.8 ug/L

Review article,

endpoints not specified


1 Source: Hardy (2002a)
2Source: Riva et al. (2007)
3Source: Rows 1-3: Hardy (2002a); rows 4-6: Environment Canada (2006]
"Source: Rows 1-7: Environment Canada (2006): rows 8-10: Hardy (2002a)
ECso = Median effective dose; LOEC = Lowest observed effect concentration; NR: Not reported; NOE: No observed effects at tested doses; NOEC = Maximum no observed effect concentration
March 2013
F-33
DRAFT-DO NOT CITE OR QUOTE

-------
Table F-13.   Effects of exposure to PBDEs in fish and frogs.
Test
Duration substance
Juvenile
16 days
49 days
120 days
96 hours
NR
72 hours
Juvenile
30 days
Medium
Doses
tested
Effect
Effect
dose
Notes
rainbow trout (Oncorhynchus mykissy

Dow FR-300-BA2

Commercial
pentaBDE
PentaBDE,
tetraBDE mix
Tetra-, penta-, and
octaBDE
Diet
NR
Injection
Cell
culture
7.5-10
mg/kg-day3
NR
NR
0-264 ug/l
Increased liver weight, increased blood
lactate concentrations, decreased
lymphocyte count
LCso
Egg mortality
Vitellogenin production
NOE
NOE
7.5 mg/kg-day
>water
solubility
>12ug/egg
10-50 ug/l
None
Saw effects in group after 71 days of depuration
(non-exposure), indicating potential delayed chronic
effects
None
Review article, study-specific details were not
provided
Review article, study-specific details were not
provided
In vitro hepatocyte assay; intensity of response
increased with increasing number of bromine atoms
lake whitefish (Coregonus clupeaformis)4
DecaBDE;
97.5-99.25% pure
Diet
0,0.1,1,2
ug/gram
Negative growth effects
2 ug/gram
Otolith increment widths narrowed starting day 1
March 2013
F-34
DRAFT-DO NOT CITE OR QUOTE

-------
Table F-13, cont.: Effects of exposure to PBDEs in fish and frogs.
              Test                 Doses
 Duration   substance   Medium    tested
                                                Effect
                                                           Effect
                                                           dose
                                                                          Notes
Juvenile lake trout (Salvelinus namaycush)5
                                            Mortality, whole body growth rate
                                                                     NOE        No effects observed during exposure period and up to 112 days
                                                                                 depuration
                                            Phase I EROD activity in liver         NOE
                                            microsomes
                                                                                 Measured on days 14, 56 of uptake and 14, 56 of depuration
„  ,      DecaBDE,    n. ,
56 days   ^n...        Diet
          >96% pure
                                  Liver somatic index changes (liver
                                  weight •*• whole fish weight x 100)
                      ng/gram
                                                                               NOE        No effects observed during exposure period and up to 112 days
                                                                                           depuration
                     Decreased free thyroxine (T4)
                     concentrations
                                   2.5, 25      Level significantly decreased at both doses (greater decrease at 25
                                   ng/gram     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
                                            Decreased free tri-iodothyronine (Ts)    2.5
                                            concentrations                      ng/gram
                                                                                 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"
Adult Chinese rare minnow (Gobiocypris rarus)6
                                            Mortality, malformations
                                                                     NOE
                                                                    None
                                            Decreased body length,
                                            gonadosomatic index (GSI) changes
                                                                     10 ug/L      GSI = Gonad weight •*• whole fish weight x 100
21 days
DecaBDE,
>99% pure
Water
0.01, 0.1, 1,
 0 ug/L
        of spermatogenesis,
reduction of spermatocytes
10 ug/L      Males only
                                            mRNA levels of thyroid hormone
                                            related genes
                                                                     Variable at   Upregulation of n/'s at all doses; tr-a in all male groups but not female;
                                                                     all doses    ttr in all female groups but not male; cf/'o2 in females at 0.01 ug/L,
                                                                                 females + males at 0.1 ug/L.
March 2013
                                           F-35
                                                                                                  DRAFT-DO NOT CITE OR QUOTE

-------
Table F-13, cont.: Effects of exposure to PBDEs in fish and frogs.
              Test                  Doses
 Duration   substance    Medium     tested
                             Effect
                                               Effect
                                               dose
                                Notes
Adult Chinese rare minnow (Gobiocypris rarus)6
21 days    DecaBDE,    Water
          >99% pure
0.01, 0.1,   Liver degeneration, hepatocyte swelling
1,10ug/L
                                            10ug/L      Females only
Unspecified/general fish7
48 hours   Commercial   NR
          decaBDE
NR
LCso
>500 mg/L   Review article, study-specific details were not
            provided
48 hours   Commercial   NR
          octaBDE
NR
LCso
>500 mg/L   Review article, study-specific details were not
            provided
48 hours   Commercial   NR
          pentaBDE
NR
LCso
>500 mg/L   Review article based value on data for Oryzias latipes,
            details were not provided
NR       Commercial   Diet
          pentaBDE
NR
Reproduction (spawning success)
NOE        Review article based value on data for Gasterpsteis
            aculeatus, details were not provided
March 2013
                 F-36
                                                                        DRAFT-DO NOT CITE OR QUOTE

-------
Table F-13, cont.: Effects of exposure to PBDEs in fish and frogs.
               Test
 Duration   substance
Medium
Doses
tested
Effect
Effect
dose
Notes
Tadpoles (Xenopus /aew's)8
                                                 Physical malformation, abnormal behavior,
                                                 increased mortality
                                                                      NOE
                                                                      None
                                                 Delayed time to metamorphosis
                                                                      1,000 ng/L    Statistically significant at 1,000 ng/L; concentration-
                                                                                   dependent trend of metamorphic delay in all groups
 51 days   DE-83R      Water
1, IU, IUU,
1,000 ng/L 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
                                                 Decrease in TR-J3-A mRNA expression (thyroid
                                                 hormone) in tail tissue
                                                                        I doses
                                                                      None
1Source: Rows 1-3: Kierkegaard etal. (1999): row 4: Hardy (2002a); row 5, Nakari and Pesala (2005); individuals per group not reported
Commercial mixture contains 77.4% decaBDE, 21.8% nonaBDE, 0.8% octaBDE as reported in Hardy (2002a)
30ne dose administered, dose was minimum 7.5 mg/kg-day, maximum 10 mg/kg-day
4Source: Kuo et al. (2010): 75 individuals per test group
5Source: Tomy et al. (2004): 70 individuals per test group
6Source: Li et al. (2011): individuals per group not reported
7Source: Hardy (2002a) (Review article)
8Source: Qinetal. (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 = Sin row 4
EROD = Ethoxyresorunfin-0-deethylase enzyme; LCso = 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
March 2013
                              F-37
                                                                                        DRAFT-DO NOT CITE OR QUOTE

-------
F.2.2. Effects  of MWCNTs on Aquatic Receptors
Table F-14.   Effects of exposure to MWCNTs in algae, macrophytes, and aquatic macroinvertebrates.
Organism
Test
substance
Dimensions
Test
duration
Medium
Doses
tested
Effect
Effect
dose
Notes
 Unicellular green algae
 (Dunaliella fert/'o/ecfa)1
Carboxylated   OD: 20-30 nm  96 hours   Water
MWCNT       L: 50 urn
0.1, 0.5,1,  Inhibited growth  LOEL 1 mg/L,    Growth lagged up to 23 days at
2.5, 5,10                   NOEL 0.5 mg/L  10 mg/L, 36% reduction in
mg/L                                     exponential growth rate
 Aquatic macrophytes
 (Chara, Elodea nuttallii,
 Potamogeton obtusifulius,
 Glycera, Alisma
 plantagoaquatica)2
Pure MWCNT   OD: 20-30 nm  3 months   Stream bed   0.002,0.2,   Changes in      LOEL 0.002     Density increased compared to
              ID: 5-10 nm              sediment     2 grams/kg  macrophyte      grams/kg        control at all levels; significant at
              L:10-30um                                      density                        0.002 and 2 grams/kg
Macroinvertebrates
(Gastropoda, Crustacea,
Oligochaeta, Hirundinea,
Bivalvia, Arachnida, Diptera)2
Amphipods
(Leptocheirus plumulosus)3
Amphipods
(Hyalella azteca)3
Pure MWCNT OD: 20-30 nm 3 months
ID: 5-10 nm
L: 10-30 urn
Pure MWCNT OD: 10-30 nm 10 days
L: 10-30 urn
Pure MWCNT OD: 10-30 nm 10 days
L: 10-30 urn
Stream bed 0.002, 0.2,
sediment 2 grams/kg
Sediment 4, 10, 33,
99, 300
grams/kg
Sediment 3, 9, 29,
87, 264
grams/kg
Recolonization
rates
Biodiversity
Death
Death
LOEL 0.002
grams/kg
NOEL >2
grams/kg
LOEL 99
grams/kg, LCso
68 grams/kg
LOEL 264
grams/kg
Dose-dependent increase in
recolonization (# of taxa and
individuals) compared to control
No observed effect
Mortality 30 ±10% at 99
grams/kg,
30 ± 0% at 300 grams/kg
Mortality 53 ± 25%
1Source: Wei et al. (2010): microwave assisted acid oxidation was used to carboxylate pristine nonfunctionalized MWCNT resulting in functionalized MWCNTs with 7.61%carboxylation, 1% residual
cobalt by weight, elemental composition 948:51:1 carbon:oxygen:cobalt
2Source: Velzeboer et al. (2011): MWCNTs nonfunctionalized, purity 95% wt
3Source: Kennedy et al. (2008): authors note that mortality increased as particle size decreased
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
March 2013
                       F-38
                                      DRAFT-DO NOT CITE OR QUOTE

-------
TableF-15.
Effects of exposure to MWCNTs via water on
Individuals Outer
per test diameter
group (nm)3
Functionalization
Concentrations
tested (mg/L)
Ceriodaphnia dubia
Effect

Effect dose
Acute (24 hours)1
20
20
20
20
20
20
20
20
20
Subchronic (48
NR
NR
NR
>12
>12
10-20
30-40
50-70
10-20
30-40
50-70
10-20
30-40
50-70
hours)2
10-30
20-30
20-30
50-70
50-70
NF
NF
NF
Ozone-treated
Ozone-treated
Ozone-treated
Ultrasound-probe treated
Ultrasound-probe treated
Ultrasound-probe treated

NF
MWCNT-OH
MWCNT-COOH
Ozone-treated
Ultrasound-probe treated
1-200
1-200
1-200
1-200
1-200
1-200
1-200
1-200
1-200

25.1,39.5,59.6
120.2
88.9
5, 10, 20, 50, 100
1,2,3,5,10
LCso
LCso
LCso
LCso
LC5o
LCso
LCso
LCso
LCso

ECso
mortality increase
mortality increase
Body length decrease4
Body length decrease4
17 mg/L
8 mg/L
20 mg/L
100 mg/L
100 mg/L
100 mg/L
8 mg/L
7 mg/L
2 mg/L

50.9 mg/L
No effects
No effects
LOEC 5 mg/L; 73% of
control at 100 mg/L
NOEC 1 mg/L; LOEL 2
mg/L; 75% of control at
10 mg/L
March 2013
F-39
DRAFT-DO NOT CITE OR QUOTE

-------
Table F-15, cont.: Effects of exposure to MWCNTs via water on Ceriodaphnia dubia.
Individuals Outer
per test diameter
group (nm)3
Functionalization
Concentrations
tested (mg/L)
Effect
Effect dose
Chronic (8 days)1
>8 50-70
>8 50-70
Ozone-treated
Ultrasound-probe
treated
5 concentrations, 0.5 to >30
5 concentrations, 0.5 to >30
ECso decreased
reproduction5
ECso decreased
reproduction5
17 mg/L
4 mg/L
1 Source: Li and Huang (2011)
2Source: Rows 1-3: Kennedy et al. (2008): rows 4-5: Li and Huang (2011)
30ther properties not reported in Li and Huang (2011), Kennedy et al. (2008) reported length 10-30 urn and purity 95%
"Dose-response growth assay
5Three generation reproductive test
ECso = Median effective concentration; LCso = 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
Table F-16.    Effects of exposure to MWCNTs on zebrafish and medaka embryos.
Number
per
group
Zebrafish,
75
60
Outer
diameter Test Exposure
(nm) substance medium
72 hours post fertilization1
Microinjectio
30^0 MWCNT nat8-cell
stage
30^0 MWCNT Water
Doses
tested

2.5, 5,
10, 20,
30, 40,
50, 60,
70, 100,
200, 300
ug/mL
2.5, 5,
10, 20,
30, 40,
50, 60,
70, 100,
200, 300
ug/mL
Effect

Heart rate
decrease
Reduced blood
circulation
Delayed hatching
Increased mortality
Effect
dose

NOEC 10 ug/mL,
LOEC 20 ug/mL
NOEC 60 ug/mL,
LOEC 70 ug/mL
NOEC 50 ug/mL3,
LOEC 60 ug/mL
NOEC 50 ug/mL,
LOEC 60 ug/mL
Notes

Measured at 24, 48, 72
hpf; dose-dependent
drop at 48 and 72 hpf
Observed at 24, 48,
and 72 hpf
8%at60ug/mL, 60%
at100mg/mL
10%at60ug/mL, 80%
at100ug/mL, 97% at
200 ug/mL
March 2013
F-40
DRAFT-DO NOT CITE OR QUOTE

-------
Table F-16, cont.: 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, 96 hours post injection2
NR
„„. Microinjectio
19'9 MWCNT ntat1-ce"
stage
2ng/
embryo
Developmental
defects
Immune response
NOE
2ng
Through adult stage
Observed at 24 hpf
Medaka, 96 hours continual exposure3
12




NR oxidized Water
MWCNT



500,
1,000,
1,500,
2,000
ug/mL
Increased mortality NOEC 1,000 ug/mL, MWCNTs were
LOEC 1 ,500 ug/mL functionalized by acid
treatment


Medaka, 10 days following 4 days of continual exposure4
12




12




NR oxidized Water
MWCNT



NR oxidized Water
MWCNT



500, Increased
1,000, malformations
1,500,
2,000
ug/mL
500, Hatching delay
1,000,
1,500,
2,000
ug/mL
NOEd.OOOug/mL,
LOEC 1, 500 ug/mL



NOEC 1 ,000 ug/mL,
LOEC 1, 500 ug/mL



MWCNTs were
functionalized by acid
treatment


MWCNTs were
functionalized by acid
treatment


Zebrafish, 56 days post injection2
NR
19.9 BSA-
MWCNT
Microinjectio 2ng/
n at 1-cell embryo
stage
Reduced survival 2 ng
of 2nd generation
Measured at day 14 of
2nd generation
lifecycle
1Source: Asharani et al. (2008)
2Source: Cheng etal. (2009)
3Source: Kim et al. (2012]
"Authors state NOEC of 40 ug/mL and LOEC of 60 ug/mL but do not provide the effect level at 50 ug/mL.
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
March 2013
F-41
 DRAFT-DO NOT CITE OR QUOTE

-------
Table F-17.   Immune responses in rainbow trout (Oncorhynchus mykiss) head kidney cells
           following MWCNT exposure.


       Endpoint                       Measured as                  Pure MWCNT1       Anionic MWCNT1

Antiviral response        IFNa expression observed after 6 hours incubation  NOE          NOE

Macrophage stimulation   IL-1b expression observed after 24 hours          LOEL5ug/mL  LOEL0.1 ug/mL (dose-
                       incubation                                                  dependent)

Cytotoxicity             Measured after 24 hours incubation               NOE          NOE
1Pure MWCNTs: >95%pure; Anionic MWCNTs: functionalized with sulfonate groups, centrifuged and ultrafiltered to purity. Both had diameter
10-20nm, length 1-2 urn
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
LOEL = Lowest observed effect level; NOE: No observed effects at tested
Source: Klaperetal. (2010)
March 2013                                          F-42                     DRAFT-DO NOT CITE OR QUOTE

-------
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                             Effect
  duration   substance  dry weight soil)          Effect          dose (mg/kg)                                  Notes

 Soil microbes1

                                      Altered community structure    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 over time
  180 days   DecaBDE   1,10,100       	
                                      Cytotoxicity                 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

                                      Changes in alkaline           All doses    Increased activity from 60-120 days at 10,100  mg/kg then decreased activity
                                      phosphatase (APA) enzyme               (inhibition) at 180 days; increased activity through full study at 1 mg/kg
 180 days    DecaBDE   1,10,100        activity	
                                      Changes in urease enzyme    All doses    Increased activity through 150 days at 1 mg/kg; activity increased at 10 mg/kg from
                                      activity                               15 days to 120 days, and at 100 mg/kg at every checkpoint except 150  days

 Nitrifying bacteria2

 4 weeks     DecaBDE   15.1,230,2,274   Change in behavior           NOE        Measured by change in nitrate/nitrite content

 Red clover (Trifolium pratense)3

 21 days     DecaBDE   15.1,230,2,274   Early life growth             NOE        Measuredby mean fresh weight of seedlings per soil plot for 15-17 days post
                                                                           seedling emergence
March 2013                                         F-43                                                            DRAFT-DO NOT CITE OR QUOTE

-------
Table F-18, cont.: 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
Corn (Zea mays)4
55%
pentaBDE,
36% 62.5, 125, 250,
y tetraBDE, 500,1,000
8.6%
hexaBDE
Germination
LOEL reduced shoot
height
LOEL reduced shoot
height
NOE None
250 None
62.5 None
Soil invertebrate (Enchytraeus crypticus)5
21 days DecaBDE 15.1,230,2,274
Reproductive effects
NOE Measured as number of juveniles per soil plot
Earthworms (Eisenia fetida)6
/days DecaBDE 0.01,0.1,1,5,10,
50, 100
56 days DecaBDE, 320,668,1,240,
98% pure 2,480,4,910
(-R . Commercial 84.9,166,361,
DbaayS octaBDE 698,1,470
Hydroxyl free radical
generation
Reproductive effects
Mortality
Reproductive effects
All doses Dose-dependent increase starting at lowest dose, reached level 2x control at
highest dose
NOE Also observed at 28 days (no effects seen)
NOE None
NOE None
1Source: Liu et al. (2011a]: 3 replicates per group
2Source: Sverdrup et al. (2006): number of replicates per group not reported
3Source: Sverdrup et al. (2006): 20 seeds per test group
4Source: Environment Canada (2006): 40 seeds per test group
5Source: Sverdrup et al. (2006): 40 adult worms per test group
6Source: Row 1: Xie et al. (2011], 60 worms per test group; row 2-4'. Environment Canada (2006], 80 worms per test group
NOE = No observed effects
March 2013
F-44
DRAFT-DO NOT CITE OR QUOTE

-------
F.2.4. Effects of MWCNTs on Terrestrial Receptors
Table F-19. Effects of exposure to MWCNTs in bacteria.
Test
substance
General soil
MWCNT
Properties
microorganisms1
OD: 15.1 nm,
L: 10-20 jjm,
SA: 237.1 m2/gram,
V: 0.86 cm3/gram,
P: 96%
Test Exposure
duration protocol

1 1 days MWCNTs suspended in
water, suspension
applied to soil
20 days 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 presence;
microbial activity nonsignificant tendency to be
repressed at 500 ug/gram; activity
decreased by 34.2-60.5% at
5,000 ug/gram
LOEL decreased 5,000 ug/gram C decreased by 36.9-43.4%, N
biomass decreased n by 27.8-30.4%
Gram negative Escherichia coli2
MWCNT
MWCNT-Fe
MWCNT
OD: 44.0 nm,
L: 1.5 |jm,
SA: 42 m2/gram,
0.08%wt Fe
OD: 44.0 nm,
L: 1.5|jm,
SA: 42 m2/gram,
4.24%wt Fe
OD: 17.4 ± 6.1 nm,
L: 77 ± 31 |jm
24 hours Exposure to aqueous
suspensions of MWCNT
at room temp, gentle
stirring; strain: MG1655
24 hours Exposure to aqueous
suspensions of MWCNT
at room temp, gentle
stirring; strain: MG1655
1 hour Incubation exposure to
MWCNT-coated filter in
0.9% NaCI solution
Water 10, 100
mg/mL
Water 10, 100
mg/mL
Filter in CNQ
culture
LOEL cytotoxicity 100mg/mL ~ 50% loss in viability
LOEL cytotoxicity 100mg/mL ~ 60% loss in viability
cytotoxicity Effect seen -32% inactivated cells
March 2013
F-45
DRAFT-DO NOT CITE OR QUOTE

-------
Table F-19, cont.: Effects of exposure to
Test Test
substance Properties duration
M WC NTs in bacteria.
Exposure Doses Effect
protocol Medium tested Effect dose

Notes
Gram negative Escherichia coli2
OD: 30 nm,
MWCNT |: C70°0/M™' t , 1 hour
0.62%wt metal
catalysts
5x107cells/mL incubated Cell 5ug/mL Cytotoxicity 5ug/mL
with MWCNTs in saline culture
for 1 hour at 37 °C;
strain: K12
Incubation exposure to Filter in CNQ Cytotoxicity Effect seen
MWCNT-coated filter in culture
.9% NaU solution,
strain: K1 2 Reduced metabolic Effect seen
activity
~3x reduction in viability
compared to controls
~3.8x reduction in viability
compared to controls
30% metabolic activity compared
to 74% in control
Metallic-pollutant resistant Cupriavididus metallidurans CH343
MWCNT OD: 44.0 nm, 24 hours
L: 1.5 urn,
SA: 42 m2/gram,
0.08%wt Fe
MWCNT-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 NOEL Cytotoxicity NOE
suspensions of MWCNT mg/mL
at room temp, gentle
stirring
Exposure to aqueous Water 10,100 NOEL Cytotoxicity NOE
suspensions of MWCNT mg/mL
at room temp, gentle
stirring
NOE
NOE
Gram negative Pseudomonas aeruginosa4
MWCNT OD: 17.4 ± 6.1 nm, 1 hour
L: 77 ± 31 urn
Incubation exposure to Filter in CNQ Cytotoxicity Effect seen
MWCNT-coated filter in culture
0.9% NaCI solution
-25% inactivated cells
March 2013
F-46
DRAFT-DO NOT CITE OR QUOTE

-------
Table F-19, cont.: Effects of exposure to MWCNTs in bacteria.
Test
substance Properties
Test
duration
Exposure Doses
protocol Medium tested
Effect
Effect dose Notes
Gram positive Staphylococcus epidermidis4
MWCNT OD: 17.4 ± 6.1 nm,
L: 77 ± 31 |jm
1 hour
Incubation exposure to Filter in culture CNQ
MWCNT-coated filter in
0.9% NaCI solution
Cytotoxicity Effect seen -50% inactivated cells
Gram positive Bacillus subtilis*
MWCNT OD: 17.4 ± 6.1 nm,
L: 77 ± 31 |jm
OH-MWCNT OD: 15-30 nm,
L: 1-5 |jm
COOH- OD: 15-30 nm,
MWCNT L: 1-5 urn
NH2-MWCNT OD: 15-30 nm,
L: 1-5 |jm
1 hour
1 hour
1 hour
1 hour
Incubation in 0.9% NaCI Filter in culture CNQ
solution
Cells suspended in 1 ml Water, culture 100ug/mL
of solution
Cells suspended in 1 ml Water, culture 100ug/mL
of solution
Cells suspended in 1 ml Water, culture 100ug/mL
of solution
Cytotoxicity NOE None
Delayed growth CNQ NOE in solutions of Dl
water, PBS, BHI, or 0.9%
NaCI
Delayed growth NOE NOE in solutions of Dl
water, PBS, BHI, or 0.9%
NaCI
Delayed growth NOE NOE in solutions of Dl
water, PBS, BHI, or 0.9%
NaCI
March 2013
F-47
DRAFT-DO NOT CITE OR QUOTE

-------
Table F-19, cont.: Effects of exposure to MWCNTs in bacteria.
Test
substance
Properties
Test
duration
Exposure
protocol
Medium
Doses
tested
Effect
Effect
dose
Notes

Gram negative Salmonella typhimuriunf
OH-
MWCNT
COOH-
MWCNT
NH2-
MWCNT
OD: 15-30nm
L: 1-5 |jm
OD: 15-30nm
L: 1-5 |jm
OD: 15-30nm
L: 1-5 |jm
1 hour
1 hour
1 hour
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,
375,
100,
375,
100,
375,
100,
375,
170,
500
170,
500
170,
500
170,
500
,290,
ug/mL
,290,
ug/mL
,290,
ug/mL
,290,
ug/mL
Delayed growth
Delayed growth
Delayed growth
Delayed growth
NOE
NOE
NOE
NOE
Nonsignificant reductions in
viability at all doses
NOE up to 500 ug/L in BHI
broth, PBS, or 0.9% NaCI
NOE up to 500 ug/L in
water, BHI broth, PBS,
0.9% NaCI
NOE up to 500 ug/L in
BHI broth, PBS, or 0.9
Dl
, or
water,
% NaCI
1Source: Chung etal. (2011)
2Source: 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)
3Source: Simon-Deckers et al. (2009), details same as footnote b
4Source: Kang et al. (2009). details same as footnote b
5Source: Row 1: Kang et al. (2009), details same as footnote b; rows 2-4: Arias and Yang (2009), no details provided
6Source: Arias and Yang (2009), no details provided; OH-MWCNT: Functionalized with OH groups derived directly from the surface of CNTs
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
March 2013
F-48
DRAFT-DO NOT CITE OR QUOTE

-------
Table F-20.   Effects of exposure to MWCNTs on plants.
MWCNT
properties
Exposure
protocol
Plant
Doses
tested
Effect
dose
Notes
Decreased germination rates1
SA: 126 m2/gram,
L: 1-2 |jm
SA: 73 m2/gram,
OD: 40-60 nm,
LNR
SA: 357 m2/gram,
OD:<10nm,
LNR
SA: NR,
OD:1 10-170 nm,
L: 5-9 |jm
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
Radish
Ryegrass
Corn
Lettuce
Cucumber
Garden
cress
Garden
cress
Mustard
Urad bean
2,000 mg/L
2,000 mg/L
2,000 mg/L
2,000 mg/L
2,000 mg/L
2,000 mg/L
0.01, 0.1, 0.5% wt
0.01, 0.1, 0.5% wt
10, 20, 40ug/mL
10,20,40ug/mL
NOE
NOE
NOE
NOE
NOE
NOE
LOEL
0.01%
LOEL
0.1%
NOE
NOE
Non-significant decrease
Non-significant decrease
None
Non-significant decrease
No observed effect
Non-significant decrease
50-70% inhibition (compared to 10% in
control)
60% inhibition at 0.1%, 40% inhibition at
0.5% (compared to 10% in control)
None
None
Changes in root growth1
>95% pure,
SA: 126
m2/gram,
OD: 10-20 nm,
L- 1 2 urn

30 seeds in water,
5 days
Rapeseed
Radish
Ryegrass
Corn
Lettuce
Cucumber
2,000 mg/L
2,000 mg/L
2,000 mg/L
20,00 mg/L
2,000 mg/L
2,000 mg/L
NOE
NOE
NOE
NOE
NOE
NOE
None
None
Non-significant total root growth decrease;
root length increase
None
None
None
March 2013
F-49
DRAFT-DO NOT CITE OR QUOTE

-------
Table F-20, cont.: Effects of exposure to MWCNTs on plants.
    MWCNT
   properties
Exposure
 protocol
Plant
Doses
tested
Effect
dose
Notes
Changes in root growth1
SA: 357          10 seeds in       Garden
m2/gram,         sewage sludge, 3  cress
OD: <10 nm, L:    days
NR
                       0.01, 0.1, 0.5% wt   LOEL     Root length inhibition-30% greater than
                                        0.01 %    control at all doses
SA: 73 rrWgram,   10 seeds in       Garden
OD: 40-60 nm,    sewage sludge, 3  cress
L: NR           days
                       0.01, 0.1, 0.5% wt   NOE
                                  None
SA: NR,
00:110-170
nm,
L: 5-9 urn
60 seeds in water, Mustard 10, 20, 40ug/mL
4 days
Uradbean 10, 20, 40ug/mL
LOEL10
ug/mL
NOE
138% increase in root length at 10 ug/mL,
202% increase at 20 ug/mL, 135% increase
at 40 ug/mL
None
Cytotoxicity2
SA: 250-300
m2/gram,
OD:9.5nm,
L: 1.5 urn
T87 cells in Thale 10mg/L
suspension, 7 cress
days (in
exponential
growth phase on
day 3)
LOEL 10
mg/L
Began on day 2; dose-dependent inhibition
observed; poor linearity of curves; more
severe with fine MWCNT agglomerates than
loose MWCNT agglomerates
1Source: Rows 1-6: Lin and Xing (2007)- MWCNT purity >95%; rows 7-8: Oleszczuk et al. (2011)- MWCNT purity >95%; rows 9-
10: Ghodake et al. (2010)- MWCNT purity 90%
2Source: Lin et al. (2009a).  MWCNT carbon purity 90%; MWCNT properties reported for loose agglomerates, fine agglomerates had
same average diameter but other dimensions not analyzed
L = Length; LOEL = Lowest observed effect level; NOE = No observed effect; NR = Not reported; OD = Outer diameter; SA =
Surface area; V = Volume
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/jnn.2008.432

   Bottinj M: Bruckner. S: Nika. K: Bottinj N: Bellucci S: Magrinj A: Bergamaschi. A: Mustelin. T. (2006).
      Multi-walled carbon nanotubes induce T lymphocyte apoptosis. Toxicol Lett 160: 121-126.
March 2013
                               F-50
                                        DRAFT-DO NOT CITE OR QUOTE

-------
   Bruchaizer. E: Frydrych. B: Sporny. S: Szymariska. 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: Tombolinj 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: Ou. L: Sua 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/i.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). Translation 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). (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

   el Dareer. SM: Kalia 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/i.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/j.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).

   Fujitanj 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.
March 2013                                         F-51                     DRAFT-DO NOT CITE OR QUOTE

-------
   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=OTSOOO 1105

   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. (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

   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. Langmuir24: 6409-6413.  http://dx.doi.org/10.102l/la80095 Iv

   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: Kawanishi. M:
      Yagi. T: Matsuda. T: Watanabe. M: Wakabavashi. K. (In Press) Genotoxicity of multi-walled carbon
      nanotubes in both in vitro and in vivo assay systems. Nanotoxicology.
      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.

   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: Aa KH: Lee. JS: Shia JH: Park. JD: Yu.  IJ. (2012).
      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.  (2012). 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
March 2013                                        F-52                    DRAFT-DO NOT CITE OR QUOTE

-------
   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. 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. 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. Carbon49: 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: Kim SH: 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. (2009a). 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/j.envpol.2007.01.016

   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.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 See 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

   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.2Q545

   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. DrugMetab 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).
      Clastogenie and aneugenic effects of multi-wall carbon nanotubes in epithelial cells. Carcinogenesis 29: 427-
      433. http://dx.doi.org/10.1093/carcin/bgm243
March 2013                                        F-53                    DRAFT-DO NOT CITE OR QUOTE

-------
   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: Toxicological aspects. Chem Res 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.0rg/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 and B6C3F1 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: Sl-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-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
March 2013                                         F-54                     DRAFT-DO NOT CITE OR QUOTE

-------
   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: Yaa 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 J Pharmacol 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: 261-212.
      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 micro nucleus 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. Xenobiotica33: 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: Pei. R: Letcher. RJ: Gunsch. C. (2009). Metabolism of polybrominated diphenyl
      ethers (PBDEs)by human hepatocytes in vitro. Environ Health Perspect 117: 197-202.
      http://dx.doi.org/10.1289/ehp.11807

   Sverdrup. LE: Hartnik. T: Mariussen. E: Jensea 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: Kitajima. 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
March 2013                                        F-55                    DRAFT-DO NOT CITE OR QUOTE

-------
   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: Wang. J: 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

   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
March 2013                                         F-56                     DRAFT-DO NOT CITE OR QUOTE

-------
             Appendix G.
Unprioritized Areas of the CEA Framework
             for MWCNTs
March 2013                           DO NOT CITE OR QUOTE

-------
        Appendix G.  Unprioritized Areas of the  CEA
                                       Framework for MWCNTs


     G.1.  Introduction to this Appendix
 1          As described in Chapter 1. the Peer Review Draft of the case study has been streamlined to
 2   clearly reflect the outcomes of the collective judgment step of the CEA process. Certain topics within the
 3   CEA framework (Figure 1-3) were designated priorities for research, based on high importance for risk
 4   assessment and low confidence that the current data could support risk management decisions. All
 5   detailed information on multiwalled carbon nanotubes (MWCNTs) that was not identified as a priority
 6   research area during the collective judgment step of the CEA process has been moved to this appendix.
 7   The designation of certain areas as "unprioritized" is not intended to imply that the topics are unimportant
 8   or that continued research is not needed; it simply implies that the topic was determined to be of lesser
 9   importance for risk assessment.


     G.2.  Product Life Cycle

     G.2.1. Feedstocks

     G.2.1.1. Life-Cycle Processes
10          A wide variety of hydrocarbons and catalysts are used to synthesize MWCNTs. The raw
11   materials required for MWCNT synthesis include a precursor carbon material, an inert gas, and metal
12   catalysts, with other specific materials depending on the particular synthetic pathway used (as described
13   in Section 2.2) (Moisala et al., 2003). Support materials such as aluminum, manganese oxide, or silica are
14   also used during synthesis of MWCNTs (Gustavsson et al.. 2011).
15          No quantitative data were found on the total magnitude of feedstocks that are or might be used in
16   commercial synthesis of MWCNTs. Table G-l. however, lists the amount of inputs required to synthesize
17   1 gram of single-walled carbon nanotubes (SWCNTs) using two of the common forms of carbon
18   nanotube (CNT) synthesis (both of which are discussed in Section 2.2.2): chemical vapor deposition
19   (CVD) and fluidized bed chemical vapor deposition (FBCVD).  According to Healy et al. (2008). SWCNT
     March 2013                                G-1                  DRAFT-DO NOT CITE OR QUOTE

-------
Table G-1.    Percent yields for agglomerated growth of SWCNTs and MWCNTs using various
           synthesis methods, processing temperatures, and catalysts.
SWCNTs1
Method of Processing
Synthesis Catalyst Used Temperature (°C) Yield (%)
CVD Fe(Mo)AI203 900 0.1-10
CVD Co/Mo/Si02 600-800 0.33-1.8
CVD Fe/MgO 900 5.2
CVD Fe/Co/MgO 1,000 5.5-7.6
CVD Fe/MgO 850 8-20
CVD Fe/MgO 900 11
CVD Fe/Mg/AI-LDH 900 17.6
CVD Fe/Mo/AI203 850 20-60
CVD Fe(CO)5 1,200 25-44
CVD Fe/Mo/MgO 800 55

MWCNTs1
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/AI203
NiAI203
Co/Mo/AI203
Ni/Si02
Co/W/MgO
Fe/Si02
Fe/AI203
Ni/Mo/MgO
Fe/Co/AI203
Co/Mo/MgO
Fe/Si02
Mo/MgO
Fe/Mo/MgO
Fe(Ni)/AI203
Ni/Mg/AI-LDH
Ni/Si02
Ni/MgO
Co/AI-LDH
Fe/Mo/AI203
Co/Mo/AI203
Co/AI-LDH
Fe/CO/CaC03
Ni/Fe/AI203
Co/Mn/Zn/AI
Processing
Temperature
500-700
650-800
700
450-850
1,000
550-1,050
550-750
1,000
700
1,000
650-800
900
600-10,00
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
124-426
166-480
188
274
280-480
560-625
1,100
6,000
17,900
Source: Zhang etal. (201Mb).
1Agglomerated growth differs from vertical and horizontal growth; however, no yield data were provided for vertical or horizontal growth of
CNTs.
CVD = chemical vapor deposition; FBCVD = fluidized bed chemical vapor deposition; LDH = layered double hydroxide
March 2013
G-2
DRAFT-DO NOT CITE OR QUOTE

-------
 1    synthesis requires large quantities of feedstocks, or inputs, compared to outputs. These inputs were
 2    calculated, however, assuming very low synthesis reaction yield (2.95%-4.50%), or mass of CNTs
 3    divided by the mass of carbon fed into the system. Current synthesis reaction yields can range from 1% to
 4    17,900% depending on synthesis method, choice of catalyst, and organizational structure (i.e., vertically
 5    aligned, agglomerated, horizontally aligned; the largest synthesis yields are obtained from agglomerated
 6    MWCNT growth) (Zhang et al.. 201 Ib). Process optimization has led to dramatically improved yields
 7    over the past few years (Zhang et al.. 20 lib). SWCNT reaction yields also have been reported to be
 8    typically much lower than MWCNT reaction yields (Zhang  et al.. 201 Ib) (see Table G-l).25
 9           Limited information suggests that MWCNT synthesis requires more precursor material than
10    SWCNT synthesis (Tsai et al.. 2009); however, no information regarding the mass of inputs of precursor
11    materials and catalysts was identified for MWCNTs.

      G.2.1.2. Potential Releases during the  Feedstock Extraction Stage
12           Release of MWCNTs would not occur during this initial phase of the life cycle—the feedstock
13    extraction stage—given that synthesis does not occur until the next stage. Hazardous raw materials,
14    however, could be released during the extraction and processing of feedstock materials. Release of CNTs
15    also could occur if reactors are not cleaned between runs. The specific raw materials that could be
16    released depends on the method of production; but likely would  include catalyst metals and carbon
17    precursor materials. No data quantifying the potential volume  of releases during feedstock extraction for
18    MWCNTs were identified.
      G.2.2.  Storage and Distribution

      G.2.2.1. Life-Cycle Processes
19           The storage and distribution stage involves the handling and transport of (1) MWCNTs,
20    (2) MWCNT flame-retardant formulations, and (3) MWCNT flame-retardant upholstery textiles.
21    The principal method of transport for these materials is not known, but likely would be by truck, train, or
22    cargo ship.
      25An 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].
      March 2013                                     G-3                    DRAFT-DO NOT CITE OR QUOTE

-------
 1           •   MWCNTs, if not immediately incorporated into a flame-retardant formulation, likely would
 2               be stored at the site of synthesis/processing in sealed receptacles until they are incorporated
 3               into flame-retardant formulations or transported to the sites where flame-retardant
 4               manufacture would occur.

 5           •   MWCNT flame-retardant formulations are typically stored at manufacturing plants in
 6               drums, tanks, or more permanent storage vessels until they are packaged and sent to textile
 7               manufacturers (U.S. EPA. 2005a).

 8           •   MWCNT flame-retardant upholstery textiles and end-use products likely also would be
 9               stored at the site of manufacture  (or an intermediate storage site) and then transported to retail
10               locations.


      G.2.2.2. Potential Releases during Storage and Distribution

11           Storage and distribution of (1) MWCNTs, (2) MWCNT flame-retardant formulations, and

12    (3) MWCNT flame-retardant upholstery textiles could result in the following releases to the environment,

13    but all release scenarios are unlikely.

14           •   Releases of MWCNTs prior to incorporation in flame-retardant formulations are likely to be
15               negligible. Release would be due primarily to accidents, as the MWCNTs would be stored in
16               sealed receptacles after synthesis. Exposure of the receptacles to high heat or fire could lead
17               to the airborne release of MWCNTs (see Section 2.2.2.2 for more details).

18           •   Releases of flame-retardant formulations could result in releases of MWCNTs to the
19               environment (U.S. EPA. 2005a). The possible scenarios for release of MWCNT flame-
20               retardant formulations during storage and distribution include damage to containers holding
21               the flame-retardant formulation,  leakage resulting from mishandling of containers, or faulty
22               or improper stacking of cartons in transport vehicles. If the containers are sealed properly and
23               not damaged during transport, releases of product prior to application might be limited to
24               spills.

25           •   Releases of flame-retardant upholstery textiles could result from accidental exposure to
26               high heat or fire, off-gassing of volatile components, and infestation with pests. Exposure of
27               the flame-retardant upholstery textiles to high heat or fire during storage and distribution
28               could lead to the degradation of the polymer matrix and subsequent airborne release of
29               MWCNTs (both free and matrix  bound) (see  Sections 2.2.2.2 and 2.4.2 for more  details).

30           •   Off-gassing of the volatile components of MWCNT flame retardants also  could occur in
31               poorly ventilated areas that experience high temperatures  (e.g., storage units, warehouses).
32               MWCNTs per se, however, are not highly volatile (see Table 1-9 and Table 3-1). Infestation
33               of textile or furniture storage facilities with rodents or other pests also could lead to the
34               release of MWCNT flame-retardant materials to the environment.
35           Table G-2 outlines potential release scenarios from the storage and distribution stage of

36    (1) MWCNTs, (2) MWCNT flame-retardant  formulations, and (3) MWCNT flame-retardant upholstery

37    textiles. Parallel potential release scenarios for  decaBDE are provided in the table for comparative

38    purposes; more detailed information on release  scenarios  for decaBDE is provided in Appendix H
      March 2013                                     G-4                    DRAFT-DO NOT CITE OR QUOTE

-------
     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)

           Accidental releases of raw materials
           (decaBDE and MWCNTs)

           Storage/transport of flame-retardant
           formulation

           Accidental releases of flame-retardant
           formulation

           Storage/transport of treated textiles
Release unlikely if properly stored     Release unlikely if properly stored
Air release possible due to storage
container defects
Air release possible due to storage
container defects
Release unlikely if properly stored     Release unlikely if properly stored
           Accidental releases of treated textiles
Water release possible due to spills
from mishandling or faulty packaging

Small air release possible if properly
stored

Air release possible due to exposure
to high heat, pest infestation, etc.
Water release possible due to spills
from mishandling or faulty packaging

Small air release possible if properly
stored

Release possible due to pest
infestation; preliminary evidence
suggests that release due to high heat
is unlikely
      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)

1            Although the transport, transformation, and fate of MWCNTs in surface water was not identified
2    as a priority area, the transport, transformation, and fate of MWCNTs in sediment was.  Because of the
3    limited available data, which overlaps between surface water and sediment, however, these topics are
4    discussed together in the main body of the document (Section 3.3.1) and therefore the surface water
5    discussion was not extracted and presented here as a separate area.
     G.3.1.2. Ground Water

6           MWCNTs in soil could leach into subsoil and ground water and migrate to surface water;
7    however, no data were found on concentrations of MWCNTs in ground water.
     March 2013
           G-5
    DRAFT-DO NOT CITE OR QUOTE

-------
      G.3.2.  Transport, Transformation, and Fate in Soil
 1           MWCNTs released from textile products can enter terrestrial ecosystems and be transported in
 2    several ways. Early reviews speculated that the propensity of MWCNTs to adsorb to soil surfaces could
 3    make them less mobile (Borm et al.. 2006; Wiesner et al.. 2006). Recently, researchers showed that
 4    MWCNTs modified with surface coatings to enhance their aqueous stability or change their surface
 5    charge behave in the environment differently than pure MWCNTs (Petersen et al.. 201 la).
 6           Petersen et al. (2011 a) examined sorption profiles of pure MWCNTs and MWCNTs
 7    functionalized with a polyethyleneimine surface  coating and determined that sorption isotherms for pure
 8    MWCNTs were nearly linear, whereas isotherms for modified MWCNTs were nonlinear, indicating that
 9    surface coating can influence MWCNT interactions with soils. The authors also suggested that MWCNT
10    characteristics (such as presence of surface coating) are better predictors of sorption behavior than soil
11    type (and organic carbon content).
12           Properties of the soil  environment (e.g.,  soil type, soil organic matter, pH, ionic strength, presence
13    of other pollutants) also could affect particle transport. General information on how those properties
14    affect nanoparticles (not specific to MWCNTs) is available in the literature (Navarre et al.. 2008; U.S.
15    EPA. 2007).
16           If MWCNTs are present in soils, plant roots could interact with those associated with soil
17    material and in soil pore water (Navarre et al.. 2008). Plants could also be exposed to MWCNTs in air
18    and water. Airborne  MWCNTs could attach to leaves and other aerial parts of plants and be translocated
19    to different tissues of the plant, in which case plants also might act as transfer vectors for MWCNTs in the
20    food chain. Additionally, bioaccumulation might be possible for carbon-based nanomaterials (Navarro et
21    al.. 2008).
22           Studies relevant to the fate and transport of CNTs in soil are provided in Appendix D. Table D-4.
23    Studies that examined MWCNT uptake, translocation, and transformation in plants  were not found.
24    Literature that presents soil concentrations of MWCNTs has not been identified.
      March 2013                                     G-6                    DRAFT-DO NOT CITE OR QUOTE

-------
      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
 1           No information was found on exposure to MWCNTs in the general public from environmental
 2    media (e.g., air, water, soil). See Section 4.1.2 for model estimates of MWCNT concentrations in
 3    environmental media that could be used with the exposure pathway and scenario characteristics below to
 4    estimate potential exposures.

 5    G.4.1.1.1 Outdoor Air
 6           Releases of MWCNTs to outdoor air throughout the product life  cycle of the flame-retardant
 7    textile coatings are possible (see Chapter 2 and Section G.2). Once MWCNTs are released to air, they
 8    might sorb or attach, depending on the surface coating and functionalization, to particulate matter and be
 9    subject to long-range transport to areas distant from their source (see Section 3.2).  Although this
10    phenomenon has not been observed for MWCNTs, it has been observed for other compounds and no
11    evidence yet exists to preclude the possibility that it would occur for MWCNTs. No data are available on
12    MWCNT concentrations in ambient air, but general public exposure pathways could be similar to those
13    observed for particulate-phase decaBDE (see Appendix H).
14           Other product constituents of flame-retardant textiles (e.g., pieces of the polymer matrix or the
15    textile fabric) also can be released. As discussed in Section 3.2. the physicochemical properties of
16    MWCNTs released to air might change over time as a result of aging, which could result in exposure of
17    the general public to different MWCNTs than those that were first synthesized or released.

18    G.4.1.1.2 Water
19           Releases of MWCNTs and other product constituents to wastewater and ambient water bodies are
20    possible throughout the product life cycle of flame-retardant textile coatings (see Section G.2). Once
21    released to water, MWCNTs are expected to sorb to particulate matter in the water column or to
22    sediments, which might limit their mobility (see Section 3.3). This behavior implies that MWCNTs also
23    primarily will be removed to sludge during wastewater treatment. No data are available on MWCNT
24    concentrations in surface waters (see Section 4.1.2.2). but general public  exposure  pathways could be
25    similar to those observed for particulate-phase decaBDE (see Appendix H). MWCNT surface
26    functionalization, however, might affect stability office MWCNTs in water and efficacy of water
      March 2013                                    G-7                    DRAFT-DO NOT CITE OR QUOTE

-------
 1    treatment methods in removal of MWCNTs (see Section 3.3.3), which could result in more or less
 2    exposure to MWCNTs in surface and drinking water, depending on the type of functionalization.

 3    G.4.1.1.3 Soil
 4           Releases to ambient air and water throughout the product life cycle of flame-retardant textile
 5    coatings will result in deposition of MWCNT particles and other product constituents (see Section G.2) to
 6    soil. Once deposited, MWCNTs are expected to sorb strongly to soil, which might limit their mobility
 7    (see Section 3.4). No data are available on MWCNT concentrations in surface soils (see Section 4.1.2.3).
 8    but general  public exposure pathways could be similar to those observed for particulate-phase decaBDE
 9    (see Appendix H).


      G.4.2.  Ecological Exposure and Kinetics Leading to Dose

      G.4.2.1. Factors Impacting Ecological Exposure
10           In biota, potential exposure routes for MWCNTs include ingestion, inhalation, or direct contact.
11    The potential for exposure via each route along with subsequent uptake and dose depends on several
12    factors, including properties of the environmental media and physiological and behavioral characteristics
13    of aquatic and terrestrial organisms. These factors can, in turn, influence the bioavailability of MWCNTs.
14    As discussed in Chapter 3 and Section H.3. the physicochemical properties of MWCNTs dictate their
15    partitioning in the environment. This partitioning drives the exposure potentials for water-dwelling,
16    sediment-dwelling, and terrestrial organisms. For example, CNTs without functionalizing surfactants are
17    hydrophobic and will  interact with other CNTs and organic matter in aquatic systems, resulting in stable
18    suspensions and bundling followed by sedimentation (Koelmans et al.. 2009; Hyung et al.. 2007). Stable
19    suspensions and settling allow for exposure of both water-column and benthic organisms to MWCNTs in
20    aquatic systems (Velzeboer et al.. 2011).
21           Properties of the environmental media also can influence exposure  potential for MWCNTs by
22    affecting bioavailability and MWCNT form.  For example, the presence of dissolved organic matter in an
23    aquatic system can cause MWCNTs to debundle. Bacterial studies have shown that debundling of
24    MWCNTs can result in greater cytotoxicity (Kang et al.. 2009). Changes in properties such as ionic
25    strength or the pH of a solution might influence sorption behaviors of CNTs (Petersen et al.. 201 la).
26    which could differentially alter exposure levels of benthic and water-column organisms.
      March 2013                                    G-8                    DRAFT-DO NOT CITE OR QUOTE

-------
      G.4.2.2. Absorption, Distribution, Metabolism and Excretion in Ecological Receptors
 1            As discussed in Section 4.2.6. an understanding of absorption, distribution, metabolism and
 2    excretion (ADME) processes can be used to relate exposure concentrations to the concentration, or dose,
 3    of material that reaches the tissues of an organism. Elucidation of organism-specific ADME processes can
 4    help explain observations of high body burdens that were not predicted based on environmental fate and
 5    partitioning alone. ADME processes influence whether and for how long a material is retained in a tissue
 6    (i.e., whether the material will bioaccumulate) and how such retention rates might differ among trophic
 7    levels (i.e., whether concentrations of the material will biomagnify in a food web). Bioaccumulation and
 8    biomagnification have been shown to influence ecological exposures and might similarly influence uptake
 9    of, and exposure of ecological receptors to, MWCNTs.
10            Ecological receptors are likely to be exposed to MWCNTs through treated products or scraps and
11    debris from products generated during end-of-life stages of the product life cycle (see Section 2.5).
12    The materials released during these processes can contain components other than the contaminant of
13    concern (e.g., textile material, glue, composite ingredients). As discussed at the beginning of Chapter 4
14    and throughout Section 4.2. studies are lacking on the matrix-bound state of MWCNTs and how exposure
15    characteristics and dose implications differ for free versus matrix-bound forms. Like the situation with
16    human exposures discussed in Section 4.2. exposure considerations for ecological receptors are informed
17    by data on MWCNTs not embedded in a polymer matrix or associated with other product ingredients
18    (e.g., textile fibers, coating ingredients). No data are currently available regarding leachability or
19    environmental release of free MWCNTs from their source products in the environment.
                                   Additional Information Highlight Box 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
20            Information on ecological uptake pathways for MWCNTs in aquatic environments is limited, but
21    existing studies indicate that some water-dwelling organisms can take up MWCNTs stabilized in organic
      March 2013                                       G-9                     DRAFT-DO NOT CITE OR QUOTE

-------
 1    matter via absorption in the gut (Kennedy et al.. 2008). Functionalization also could affect uptake by
 2    aquatic organisms by altering the binding between the MWCNTs and body tissues (Li and Huang. 2011).

      Toxicokinetics and Body Burden in Aquatic Systems
 3           Limited information is available on MWCNT ADME and body burdens in aquatic organisms. As
 4    mentioned in Section 3.3. CNTs are likely to attract lipophilic molecules in aqueous media (Wu et al..
 5    2006). and association of MWCNTs with lipophilic molecules could  affect uptake in aquatic ecosystems.
 6    One study exposed a species of water flea (Ceriodaphnia dubid) to MWCNTs stabilized in suspended
 7    natural organic matter and demonstrated that carbon materials can be present in the gut (Kennedy et al..
 8    2008). suggesting that some water-dwelling organisms can take up MWCNTs. Surface functionalization
 9    by lipophilic molecules  in the natural environment could further affect uptake by aquatic organisms by
10    altering the binding between particles and body tissues (Li and Huang. 2011).
11           A few studies have shown that MWCNTs can be taken up by aquatic invertebrates, but are not
12    bioaccumulated over time (Petersen et al.. 201 la). For example,  Peterson et al. (2010) determined tissue
13    concentrations of MWCNTs in the freshwater sediment blackworm (Lumbriculus variegatus) exposed to
14    MWCNTs via soil for 30 days. The authors calculated biota-sediment accumulation factors between 0.1
15    and 1, indicating that retention of MWCNTs by this species is approximately one-tenth the concentration
16    in the sediment (Petersen et al.. 2010). Although this suggests that MWCNTs will not continue to build
17    up in the tissues of some aquatic invertebrates over time, the small concentrations in these species might
18    be better retained by larger predator species, leading to net accumulation in those species through dietary
19    sources. Additionally, suggestions have been made that current methods for measuring bioaccumulation
20    and calculating bioconcentration factors are not sufficient for nanomaterials (Handy et al.. 2012). These
21    methods rely on an evenly dispersed aqueous solution of the compound that achieves a steady-state
22    concentration between external media and biological tissues, which is potentially incompatible with the
23    dynamic behavior of nanomaterials in environmental media and  the challenges associated with dispersion
24    of MWCNTs in particular. Further, traditional understanding of bioaccumulation assumes that the
25    processes of uptake and elimination follow well-characterized kinetics and diffusive flux models, which
26    are based on underlying biological mechanisms of solute transporter channels. Pathways of uptake and
27    elimination for nanomaterials, including MWCNTs, are not well understood, and the degree to which the
28    bioaccumulation pathways might differ from those of conventional materials is unclear
29    (Handy etal.. 2012).
30           Despite the lack of studies directly investigating uptake, absorption efficiency, and
31    bioaccumulation of MWCNTs in aquatic food webs, the high persistence and hydrophobicity of
32    MWCNTs are characteristics generally associated with bioaccumulative substances (Petersen et al.. 2010;

      March 2013                                    G-10                   DRAFT-DO NOT CITE OR QUOTE

-------
 1    Helland et al.. 2007). Based on these characteristics alone, MWCNTs are expected to accumulate in
 2    aquatic food webs under some conditions. Which additional material, environmental, or biological
 3    characteristics determine whether and to what degree bioaccumulation occurs are unknown
 4    (Handy etal.. 2012).

      G.4.2.4. Exposure Pathways in Terrestrial Systems
 5           Limited information is available regarding exposure pathways and ecological uptake of
 6    MWCNTs in terrestrial environments. As also discussed in Section 3.2. limited evidence exists that
 7    airborne MWCNTs can quickly (within approximately two weeks) transform to amorphous carbon (Zhu
 8    etal.. 2011). thus limiting exposures to terrestrial organisms. Other data suggest that MWCNTs might
 9    stabilize in ambient conditions, however, which would serve to increase exposures (Yang etal.. 2009). If
10    MWCNTs are present in soils, plant roots could interact with those in soil or pore water
11     (Navarre et al.. 2008).

      Toxicokinetics and Body Burden in Terrestrial Systems
12           Limited information is available on MWCNT ADME and body burdens in terrestrial organisms.
13    MWCNTs present in soils could be absorbed or consumed by biota; MWCNTs taken up by plant roots
14    and plant tissues also could be consumed. Few studies have attempted to measure tissue concentrations of
15    MWCNTs in biota.  One laboratory study was identified that determined tissue concentrations of
16    MWCNTs in earthworms (Eiseniafoetida) exposed to MWCNTs via soil for 30 days. The authors
17    calculated biomagnification factors between 0.01 and 0.1, indicating that tissue concentrations of
18    MWCNTs in this species is approximately l/100th to l/10th the concentration in the sediment (Petersen
19    etal.. 2010). In another study using 14C-labeled pure MWCNTs and MWCNTs with various
20    polyethyleneimine surface coatings, Petersen et al. (2011 a) assessed the  extent to which modified
21    MWCNTs concentrate in earthworms. Results indicated that surface  coating did not significantly affect
22    MWCNT uptake or elimination rates over a 28-day period. The bioaccumulation factor remained less
23    than 0.12 throughout the study regardless of MWCNT type (purified or modified with surface coatings),
24    indicating that accumulation of MWCNTs from soil by earthworms is low (Petersen et al.. 201 la). As
25    discussed in Section G.4.2.1. past studies have  speculated that the high persistence and hydrophobicity of
26    MWCNTs are characteristics generally associated with bioaccumulative substances; however, recent
27    studies have shown that MWCNTs do not behave like other bioaccumulative substances because altering
28    the octanol-water distribution behavior does not change bioaccumulation factor values (Petersen et al..
29    2010; Helland et al.. 2007). Also, the complexity of food  web interactions that cross aquatic and terrestrial
30    systems makes determining the source of MWCNTs in terrestrial food webs difficult.

      March 2013                                     G-11                   DRAFT-DO NOT CITE OR QUOTE

-------
    G.5. Potential  Human  Health, Ecological, and Other Impacts

    G.5.1. Ecological Effects

    G.5.1.1. Terrestrial Receptors
1          Compared to other groups of organisms, a large amount of data was identified regarding toxicity
2   of MWCNT to soil microbes and plants (see Sections G.5.1.1.1 and G.5.1.1.2). No information was
3   identified for toxicity to terrestrial vertebrates (see Section G.5.1.1.3), but some assumptions can be made
4   for mammals based on toxicity studies intended for human health purposes presented in Section 5.1.
    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
(Enc/iyfraeus
crypticus)
Chronic NOEL
(Eisenia fetida)
Oxidative stress
(E. fetida)
Effect level
>2,274
mg/kg
(6 months)
100 mg/kg
>2,274
mg/kg
>4,910
mg/kg
0.1-10
mg/kg
Citation
Sverdrupetal.
(2006)
Liuetal.
(20 11 a)
Sverdrupetal.
(2006)
ACC (2001) as
cited in
Environment
Canada (2006)
Xieetal. (2011)
MWCNTs
Effect
Acute NOEL
(Cupriavidus
metallidurans)
Acute LDso
(Escherichia coli)
Chronic
cytotoxicity
NOEL; LOAEL
Effect level
>100 mg/L
100 mg/mL
500 ug/gram
5,000
ug/gram
Citation
Simon- Deckers
etal. (2009)
Simon- Deckers
etal. (2009)
Chung etal.
(2011)
ND
     NOEL = No-observed-effect level, LDso = Median lethal dose, LOAEL = Lowest-observed-effect level, ND = No data identified
    March 2013
G-12
DRAFT-DO NOT CITE OR QUOTE

-------
G.5.1.1.1 Soil Microbes and Terrestrial Invertebrates
  1
  2
  3
  4
  5
  6
  7
  8
  9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
        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 et al.. 2011). Similarly, effects on terrestrial invertebrates, such as worms, can
influence health and fertility of a soil ecosystem (Xie etal. 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.
        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 coli, 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. coli, yet had no effect on
       Additional Information Highlight Box 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. (2008b) 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 polyethyleneimine  (PEI), with little
    apparent  difference in  uptake among different  types  of
    MWCNTs (Petersen et al.,  2011a).  While 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.
Cupriovidus metallidumns, 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.
March 2013
G-13
                                                                               DRAFT-DO NOT CITE OR QUOTE

-------
 1    G.5.1.1.2 Terrestrial Plants
 2           Table G-4 describes key toxicity values identified for the effects MWCNTs on terrestrial plants.
 3    Parallel data for decaBDE are provided in the table for comparative purposes; more detailed information
 4    on decaBDE can be found in Appendix H and Appendix F. Table F-18. Appendix F (Table F-20)
 5    summarizes details of the MWCNT studies identified and reviewed for this section.
      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 (2000a)
Red clover >2,274 mg/kg Sverdrup et al.
(2006)

ND
ND
ND
Corn penta/tetraBDE Great Lakes
mix: 250 mg/kg Chemical
Corporation (2000a)
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
Effect level
>2,000 mg/L
>40 ug/mL
0.01 %w/w
0.1%w/w
>2,000 mg/L
>10mg/L
0.01 %w/w
Citation
Lin and Xing
(2QQ7)
Ghodake et al.
(2010)
Oleszczuketal.
(20H)
Oleszczuketal.
(20H)
Lin and Xing
(2QQ7)
Linetal. (2009a)

Oleszczuketal.
(20H)
ND
       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
 6    Both beneficial and detrimental effects of nanoparticle exposures have been reported for plants. For
 7    example, Khodakovskaya et al. (2011)) have demonstrated positive effects on seed germination and plant
 8    growth from MWCNT exposure. Other studies with MWCNTs, however, have implied that exposure to
 9    high levels could have negative effects on seed germination and plant growth, as several studies show
10    trends and a few show statistically significant impacts. For example, Lin and Xing (2007) showed that
11    exposure to MWCNTs with diameters  10-20 nm at a concentration of 2,000 mg/L caused no significant
      March 2013
G-14
DRAFT-DO NOT CITE OR QUOTE

-------
 1    differences in germination rates or root length for six different agriculturally relevant plant species,
 2    although a nonstatistically significant decrease in germination was observed in four of the species (Lin
 3    and Xing, 2007). Conversely, Oleszczuk et al. (2011) found that MWCNTs added to sewage sludge26 at
 4    concentrations of 0.01, 0.1, and 0.5% weight-for-weight significantly inhibited garden cress (Lepidium
 5    sativum) seed germination. The authors observed diameter-dependent responses, as root growth was
 6    inhibited at all three concentrations for the smaller diameter MWCNTs but was not affected at any
 7    concentration for the larger diameter MWCNTs (Oleszczuk et al.. 2011) (see Text Box 5-1). Finally, no
 8    physical injury to cell morphology was observed in thale cress (Arabidopsis thaliana) cell suspensions
 9    exposed to 10 mg/L MWCNTs, but significant loss in cell viability and growth and chlorophyll inhibition
10    were observed after 7 days of exposure.  Cytotoxicity was more severe following exposure to fine, small
11    bundles than to loose, large bundles of MWCNTs (Lin et al..  2009a). indicating that dispersion state could
12    play a role in toxicity (see Text Box 5-1 and Appendix F. Table F-20).
13            A study by Tan and Fugetsu (2007) provides some insight on the mechanism through which
14    MWCNT exposure affects plant growth and the ecological relevance of the trend described above.
15    Cultures of rice cells in an embryonic growth stage formed large associations with MWCNT; the cells
16    that interacted with the MWCNTs experienced high cell death.  Only some cells within the culture
17    associated with the MWCNTs, however, and clumps formed  by this initial subset of the cells in the
18    culture continued to attract other MWCNTs, forming larger associations over the course of the 4-day
19    exposure period. Cells that did not form these associations with MWCNTs were not adversely affected  by
20    the MWCNTs as exposure continued. The authors stated that their results illustrate how some plants
21    might be able to tolerate low levels of MWCNTs without major population-level effects due to a self-
22    defense response (Tan and Fugetsu. 2007).
23            Ghosh et al. (2011) illustrated clastogenicity mAllium cepa (onion) bulbs exposed to 0, 10, 20,
24    and 50 ug/mL MWCNTs using traditional cell culture tests. Chromosomal aberrations, DNA cross-
25    linking, and induction of apoptosis led authors to conclude that MWCNTs might have a significant
26    impact on genomic activities of plants.

27    G.5.1.1.3 Terrestrial Vertebrates
28            The impacts of oral  exposure to MWCNTs on mammals are uncertain (see Section 5.1). Testing
29    performed in mammals for relevance to humans has focused on inhalation exposure routes; acute
30    inhalation studies have found that MWCNTs or associated contaminants can induce oxidative stress,
      26Sewage 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.
      March 2013                                     G-15                    DRAFT-DO NOT CITE OR QUOTE

-------
1    pulmonary inflammation, and fibrosis. MWCNTs might cause slight skin and eye irritation. No studies

2    were identified that specifically investigated the ecological effects of MWCNTs on terrestrial vertebrates.
             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-ll

     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. (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.

     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: Metoalfe. 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 Chem31: 15-31.
        http://dx.doi.org/10.1002/etc.7Q6

     Healy. ML: Dahlben. LJ: Isaacs. JA. (2008). Environmental assessment of single-walled carbon nanotube processes.
        J Ind Ecol 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: Former. 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. Langmuir24: 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 Chem 27:
        1932-1941. http://dx.doi.org/10.1897/07-624.1
     March 2013                                        G-16                     DRAFT-DO NOT CITE OR QUOTE

-------
Khodakovskava. MV: de Silva. K: Nedosekin. DA: Dervishi. 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

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. Carbon49: 1672-1679. http://dx.doi.0rg/10.1016/i.carbon.2010.12.052

Lin. C: Fugetsu. B: Su. Y: Watari. F. (2009a). 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. (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.jhazmat.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: Santschj 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.org/10.1016/i.jhazmat.2010.11.028

Petersen. EJ: Huang. Q: Weber. WJ. (2008b). 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/etc. 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

Petersea EJ: Zhang. L: Mattison. NT:  O'Carroll DM: Wheltoa AJ: Uddin. N: Nguyen. T: Huang. Q: Henry. TB:
   Holbrook. RD: Chen. KL. (20lib). 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: 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

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 self-
   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.
March 2013                                        G-17                     DRAFT-DO NOT CITE OR QUOTE

-------
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). (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

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]. JPhys ChemB 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.jcis.2009.06.017

Zhang. Q: Huang. JQ: Zhao. MQ: Qian. WZ: Wei. F. (20lib). 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].  Carbon49:  354-355. http://dx.doi.0rg/10.1016/i.carbon.2010.08.022
March 2013                                        G-18                      DRAFT-DO NOT CITE OR QUOTE

-------
This page intentionally left blank.

-------
       Appendix H. Compilation of
    CEA Framework Data for DecaBDE
March 2013                        DRAFT - DO NOT CITE OR QUOTE

-------
     Appendix H.  Compilation of CEA Framework
                               Data for  DecaBDE

 1          Chapter 2 through Chapter 5. and Appendix G. present information on multiwalled carbon
 2   nanotubes (MWCNTs) in a comprehensive environmental assessment (CEA) framework. This appendix
 3   contains detailed, parallel information on decabromodiphenyl ether (decaBDE). As noted in Chapter 1,
 4   Chapter 2 through Chapter 5. and Appendix G. also contain text boxes that highlight information about
 5   decaBDE and tables and figures with side-by-side comparisons of decaBDE and MWCNT data. These
 6   elements provide a highlight-level comparison between the two compounds as used in flame-retardant
 7   textiles to illustrate key concepts that might be helpful to risk assessors evaluating MWCNTs.
 8   Supplemental details about decaBDE are provided in this appendix, to provide more in-depth data for
 9   comparison for each CEA framework element. With the exception of Section H.I, the section numbers of
10   this appendix are parallel to corresponding sections of the MWCNT CEA framework presented in
11   Chapter 2 through Chapters. Section H.I provides an  introduction to decaBDE, which is identical to the
12   introduction provided in Section 1.3.1ofthe main text. It is repeated here to remind readers of the
13   introductory details regarding physical and chemical properties of decaBDE that set the stage for
14   understanding the remainder of this appendix.


      H.1. Introduction  to DecaBDE
15          DecaBDE is part of a larger group of brominated flame retardants (BFRs) called polybrominated
16   diphenyl ethers (PBDEs), a group of 209 structurally similar BFRs that differ in the number and location
17   of bromine atoms (Table G-l) (Rahman etal.. 2001: NRC. 2000). Although PBDEs are typically
18   categorized into classes by number of bromine atoms (e.g., PBDE with two bromine atoms is a diBDE;
19   ten bromine atoms is a decaBDE), a single class might contain several different PBDE congeners with the
20   same number of bromine atoms in different locations (i.e., PBDE BFRs can have many isomers). As the
21   only fully brominated PBDE, decaBDE is the exception, existing only as a single congener (BDE-209).
     March 2013                                 H-1                 DRAFT - DO NOT CITE OR QUOTE

-------
      Table H-1.    Major PBDE congeners.
PBDE Class
DiBDE
TriBDE
TetraBDE
PentaBDE
HexaBDE
HeptaBDE
OctaBDE
NonaBDE
DecaBDE
Congeners
BDE-7, BDE-8, BDE-11, BDE-12, BDE-13, BDE-15
BDE-17,
BDE-47,
BDE-85,
BDE-153
BDE-181
BDE-196
BDE-206
BDE-209
BDE-25, BDE-28, BDE-30, BDE-32, BDE-33, BDE-35, BDE-37
BDE-49, BDE-66, BDE-71, BDE-75, BDE-77
BDE-99, BDE-100, BDE-105, BDE-116, BDE-118, BDE-119, BDE-126, BDE-138, BDE-140
, BDE-154, BDE-155, BDE-166
, BDE-183, BDE-190
, BDE-197, BDE-203
, BDE-207, BDE-208

      Source: U.S. EPA (201 Oa].
 1    Commercial formulations of decaBDE (see Table H-2) are generally 97-98% BDE-209 with less than 3%
 2    nonaBDE congeners present as impurities (Rahman et al., 2001; NRC. 2000) (see Appendix B. Table B-l
 3    for analytical techniques used to distinguish PBDE congeners in samples). Although the terms decaBDE
 4    and BDE-209 often are used interchangeably, this case study primarily uses the term decaBDE to refer
 5    generally to the flame-retardant formulation and BDE-209 to refer to the specific decaBDE congener
 6    analyzed in scientific studies.
 7           DecaBDE is the  most widely used of the PBDEs and has been well studied. In 2001, decaBDE
 8    use accounted for 83% of total PBDE production worldwide (U.S. EPA. 2010a): an estimated 10-20% of
 9    decaBDE use is in the textile industry (Pure Strategies Inc.. 2005). At the end of 2004, both octa- and
10    pentaBDE were voluntarily withdrawn from the U.S. marketplace due to evidence of environmental
11    persistence and toxicity,  which left decaBDE as the sole PBDE available for use in commercial products
12    in the United States (U.S. EPA. 2010a). Several standard physicochemical properties are used to describe
13    traditional chemicals: melting point, boiling point, molecular weight, and others. Such values are
14    presented for decaBDE in Table H-3.
      March 2013                                     H-2                  DRAFT - DO NOT CITE OR QUOTE

-------
     Table H-2.   Commercial formulations of PBDEs used as flame retardants.
Name
Penta formulation1
Octa formulation
Deca formulation2
Congener
Tetra
Penta
Hexa
Hexa
Hepta
Octa
Nona
Nona
Deca
Makeup and Percent Composition
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%)

     1Trace amounts of additional congeners might be present in commercial formulations: <0.2% triBDE congeners.
     2Trace 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).
1            DecaBDE can be applied to textiles by a variety of mechanisms, but this case study focuses on
2    the application of decaBDE as a back-coating. This application method is used most frequently for
3    decaBDE (Pure Strategies Inc.. 2005; NRC. 2000) and is most similar to the application method expected
4    for MWCNTs used in textiles (see Section 1.3.2). The back-coating process usually involves mixing
5    decaBDE with a copolymer or resin binder to comply with fire  safety standards (Pure Strategies Inc..
6    2005; NRC. 2000). DecaBDE combines the flame-retardant mechanism of most BFRs (releasing
7    halogens during combustion to compete with the availability of oxygen for the flame) with formation of a
8    protective char barrier (NRC. 2000) that interferes with the spread of the flame and helps the material to
9    self-extinguish (Pure Strategies Inc.. 2005).
     March 2013                                       H-3                    DRAFT - DO NOT CITE OR QUOTE

-------
     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',4,4',5,5',6,6'-decaBDE; BDE-209; benzene, NLM (2011): ATSDR (2004)
1,1'-oxybis[2,3,4,5,6,-pentabromo]-; decabromodiphenyl oxide;
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 1Q-6 Pa at 21 °C Hardy (2002b)
Henry's law constant 1 .93 x 1Q-8 L atm/mol Hardy (2002b);
0.04 Pa rrWmol at 25 °C Cetin and Odabasi
Density 3.0 grams/cm3
NRC (2000)
(2005)

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 Ci2BrioO
Chemical structure !r
Hardy (2002b)
Hardy (2002b)
NLM (2011): ECB
NLM (2011)


2003)

Br
     Note: Kow = Octanol/water partition coefficient, Koc = Soil organic carbon/water partition coefficient.
      H.2. Product Life Cycle
1    A product's life cycle encompasses all stages of its existence from "cradle to grave," starting with the
2    extraction of raw materials from the earth for the manufacture of the product and continuing downstream
3    until these materials are returned to the environment following disposal (U.S. EPA, 2006).
4    The components of the life cycle determine the potential for releases and possible impacts on human
5    health, ecological populations, and the environment (Som etal.. 2011). which can be evaluated
6    systematically within the CEA framework. Potential environmental impacts of a product throughout its
7    life cycle can be estimated using a life-cycle assessment (LCA) approach, which involves four steps: goal
8    definition and scope, inventory analysis, impact analysis, and interpretation (U.S. EPA. 2006). The CEA
     March 2013
H-4
DRAFT - DO NOT CITE OR QUOTE

-------
 1    approach incorporates information from available LCAs in the "product life cycle" and "impacts"
 2    portions of the CEA framework and combines this knowledge with other analyses or qualitative indicators
 3    related to transport, transformation, and fate, exposure-dose, and additional impacts not considered in
 4    available LCAs. As discussed in Chapter 1, if a plausible reason exists to include an impact in the CEA
 5    framework, qualitative or quantitative information on that effect can be included from LCAs or other
 6    sources (if an LCA has not been completed) to evaluate that particular impact.
 7           A generalized depiction of the life cycle for decaBDE and MWCNT coatings used to confer
 8    flame-retardant properties to upholstery textiles is presented in Chapter 2. in Figure 2-1. That figure
 9    illustrates the life cycle of these materials as five main  stages: (1) acquisition and processing of
10    feedstocks; (2) manufacturing, including research and development (R&D) processes; (3) storage and
11    distribution; (4) use; and (5) end-of4ife processes (including disposal, reuse, and recycling). These stages
12    correspond roughly to the four primary life-cycle stages the U.S. Environmental Protection Agency (U.S.
13    EPA. 2006) outlines: (1) raw materials acquisition, (2)  manufacturing, (3) use/reuse/maintenance (with
14    storage and distribution discussed as a distinct stage in this case study), and (4) recycle/waste
15    management. As mentioned in Chapter 1, R&D is included in the product life-cycle portion of the CEA
16    framework, given its importance regarding emerging materials such as MWCNTs. For such materials,
17    R&D efforts can elucidate potential risks associated with commercial-scale manufacturing. In fact,
18    because it often takes place when health and safety information is being developed for a material, R&D
19    presents an ideal opportunity to gather data on a product's potential impacts and to make design
20    adjustments if appropriate. Similarly, as discussed below, differences between R&D activities and the
21    commercial manufacturing process (e.g., use of protective equipment, volume of material produced) could
22    be important considerations in mitigating potential risks to individuals involved in R&D versus
23    commercial manufacturing.
24           To conduct a comparative  CEA, relevant information on life-cycle inventories from existing
25    LCAs would be incorporated into the product life cycle to characterize the inputs (e.g., raw materials,
26    energy) and outputs  (e.g., emissions to air and water, co-products) associated with each material's
27    manufacture. Impacts information  from existing LCAs also would be considered (see Section H.5). Other
28    LCA aspects also might apply, including using an appropriate functional unit, which is a quantitative
29    measure of a product's function or a process that facilitates comparison (U.S. EPA. 2006). In the current
30    case study, a functional unit might correspond to the degree of flame retardancy conveyed by
31    incorporation of a certain amount of decaBDE. In general, for this case study, data that specify
32    appropriate functional units were not identified; the reader might, however, consider how this aspect of
33    existing or future LCAs could be incorporated into a future CEA when evaluating data gaps and needs.
      March 2013                                     H-5                   DRAFT - DO NOT CITE OR QUOTE

-------
 1           This section outlines important aspects of each of the five life-cycle stages outlined in Figure 2-1
 2    (in Chapter 2) for decaBDE used in upholstery textiles. This section also includes descriptions of the
 3    important environmental release scenarios for decaBDE and MWCNTs across the product life-cycle
 4    stages based on current knowledge. A variety of release scenarios are possible throughout the life-cycle
 5    stages described in this appendix. Figure 2-1 (in Chapter 2) also outlines potential release scenarios for
 6    decaBDE flame-retardant upholstery textile coatings throughout the life cycle along with potential forms
 7    of the released substances (i.e., free, bundled, or matrix bound). The term "free decaBDE" refers to pure,
 8    unbound materials. The term "matrix-bound decaBDE" refers to materials that are part of a polymer
 9    matrix (e.g., the flame-retardant formulation).


      H.2.1. Feedstocks

      H.2.1.1.  Life-Cycle Processes
10           The raw materials used in commercial synthesis of decaBDE are phenol, bromine, and a catalyst
11    (e.g., aluminum bromide or iron) (IPCS, 1994). Phenol is produced from cumene, which is obtained
12    primarily from the distillation or other processing of petroleum products (Mahapatra. 2010). Commercial
13    production of bromine involves the drying of brine, typically obtained from sea water (Kesner. 2005).
14    No data were found on the energy and resource demands of raw material extraction for synthesis of
15    decaBDE.

      H.2.1.2.  Potential Releases during the Feedstock Extraction Stage
16           Release of decaBDE would not occur during this  initial phase of the life cycle given that its
17    synthesis does not occur until the stage that follows extraction. Release of hazardous raw materials,
18    however, could occur during the extraction and processing of feedstock materials. Release of decaBDE
19    also could occur if reactors are not cleaned between runs. Specific materials that could be released include
20    petroleum-based chemicals. No data were identified quantifying the potential volume of releases during
21    feedstock extraction for decaBDE.
      H.2.2. Manufacturing
22           The manufacturing stage for decaBDE flame-retardant upholstery can be viewed as a sequential
23    process involving synthesis of decaBDE, material processing, and product manufacture (i.e., formulation
24    of the flame-retardant mixture, application of the flame-retardant mixture to textiles, and incorporation of


      March 2013                                     H-6                  DRAFT - DO NOT CITE OR QUOTE

-------
 1    the flame-retardant textile into consumer or commercial goods). Discussion of R&D also is included in
 2    this section, given the similarities to key aspects of synthesis, processing, and manufacture.

      H.2.2.1. Research and Development
 3    H.2.2.1.1 Life-Cycle Processes
 4           Research on decaBDE and on flame-retardant coatings involving decaBDE is principally
 5    conducted in specialized laboratory environments. R&D activities are expected to be carried out by
 6    individuals rather than automated mechanisms used in commercial-scale manufacture. The processes of
 7    interest to researchers are similar to those used in commercial-scale manufacture of these materials:
 8    synthesis, purification, modification, dispersion, incorporation into flame-retardant formulations, and
 9    application to textiles.  Substantially less R&D related to decaBDE flame retardants is expected to occur at
10    the present time compared to MWCNT flame retardants, given that decaBDE flame-retardant
11    technologies are more mature and the use of decaBDE is decreasing or being phased out due to health and
12    ecological concerns. The following sections (material synthesis, material processing, and product
13    manufacturing) provide detailed information on the processes of potential interest for R&D.

14    H.2.2.1.2 Potential Releases during the R&D Stage
15           Release scenarios during the R&D stage are expected to be similar to release scenarios from
16    commercial synthesis described in the following sections, but the quantities released are anticipated  to be
17    much smaller in the R&D stage. The quantities of decaBDE handled in research laboratories are much
18    smaller than those handled in commercial-scale manufacturing facilities. Although R&D activities are
19    typically carried out in laboratories with specialized pollution control systems in place, including fume
20    hoods, ventilation systems, and environmental control systems, not all facilities have standardized
21    engineering controls. For example, these practices might not be in place for small start-up operations.
22    Given the experimental and somewhat unpredictable nature of R&D, releases from handling of materials
23    during synthesis, processing and purification, storage, and analysis are possible.
24           No information was found  in the literature that describes release of decaBDE from R&D
25    facilities. No data were found that describe how releases in academic labs compare with releases in
26    commercial R&D labs.
      March 2013                                     H-7                  DRAFT - DO NOT CITE OR QUOTE

-------
      H.2.2.2. Material Synthesis
 1    H.2.2.2.1 Life-Cycle Processes
 2           Commercial synthesis of decaBDE involves conversion of phenol to diphenyl ether via the
 3    Williamson ether synthesis [Kirk Othmer (2005) as cited in Wright et al. (2008)]. Diphenyl ether is then
 4    brominated in the presence of a catalyst (generally, aluminum bromide or iron) to produce commercial
 5    decaBDE (EU. 2002; IPCS. 1994). Commercially, decaBDE is synthesized in a batch process in enclosed
 6    vessels during both the reaction and the subsequent drying process (IPCS. 1994). DecaBDE powder is
 7    collected in bags during the recovery phase following the synthesis process (EU. 2002). Commercial
 8    formulations of decaBDE typically contain decaBDE, 97-98% weight-for-weight measurement, and other
 9    PBDEs (primarily nonaBDE), 0.3-3.0% weight-for-weight measurement (IPCS. 1994). No information
10    was found on by-products  of decaBDE synthesis.

11    H.2.2.2.2 Potential Releases during the Material Synthesis Stage
12           Synthesis of decaBDE could result in releases to air or water (U.S. EPA. 2005a). Fugitive
13    releases of decaBDE vapor from a reactor vessel have been estimated as 1.1 * 10"5 mg/ton, and release
14    from the bagging of synthesized PBDEs have been estimated as <70 grams/ton PBDE produced [(EU.
15    2002); EEC (1993) as cited in EU (2002)]. Airborne releases of decaBDE particles likely would sorb to
16    dust (see Section 3.2). but  loose dust likely would be vacuumed and the area would be washed with water,
17    reducing airborne particles (EU. 2002). The main source of water release of decaBDE during the
18    synthesis stage would be due to cleaning of equipment and floors after synthesis. One study found,
19    however, that wastewater releases of decaBDE are unlikely to exceed 0.5 kg/ton if equipment is washed
20    after every batch (EU. 2002). Releases directly to skin could occur through handling of bags containing
21    solid decaBDE (U.S. EPA. 2005a). Large manufacturing facilities, however, likely would have exhaust
22    ventilation in place to  minimize air release into the general environment. Engineering controls that
23    regulate temperature and pressure to minimize the potential for release also would likely be in place (U.S.
24    EPA. 2005a). As a result, air and water releases of decaBDE to the environment during the synthesis
25    stage are not expected to be large.
26           Accidental releases through fugitive equipment leaks, malfunctioning ventilation systems, and
27    exposure to fire or high heat could occur at all stages of manufacturing (material synthesis, material
28    processing, and product manufacture). A fugitive equipment leak or ventilation malfunction could lead to
29    the airborne releases of decaBDE, as could exposure to fire or high heat. If exposed to high heat,
30    decaBDE can form polybrominated dibenzofurans (PBDFs), polychlorinated dibenzo-p-dioxins, and
31    nonhalogenated substances such as polycyclic aromatic compounds, which could be released into the


      March 2013                                    H-8                   DRAFT - DO NOT CITE OR QUOTE

-------
 1    environment (EU, 2002). Such accidental events could result in larger releases of decaBDE to the
 2    environment than normal release scenarios due to the lack of control mechanisms compared to those in
 3    place to mitigate anticipated releases. Table 2-3 in Chapter 2 summarizes the anticipated potential release
 4    scenarios from the material synthesis stage of decaBDE and MWCNTs.

      H.2.2.3. Material Processing
 5    H.2.2.3.1  Life-Cycle Processes
 6           Material processing includes any modification of decaBDE after synthesis and before
 7    incorporation into a flame-retardant formulation. These modifications can include purification,
 8    functionalization, and dispersal in solvents. After synthesis, decaBDE does not require further processing
 9    before incorporation into the flame-retardant formulation.

10    H.2.2.3.2 Potential Releases from the Material Processing Stage
11           No release scenarios for decaBDE are summarized here because the activities specified for this
12    stage are not anticipated to occur for decaBDE.

      H.2.2.4. Product Manufacturing
13           In this section, product manufacturing for decaBDE is described. This life-cycle stage is
14    considered to include the manufacture of flame-retardant formulations, the manufacture of textiles
15    containing decaBDE-based flame retardants, and the manufacture of end-use products containing flame-
16    retardant materials, such as furniture.

17    H.2.2.4.1  Life-Cycle Processes
18           In a typical decaBDE flame-retardant formulation, decaBDE  and antimony trioxide (a synergist
19    used to enhance the activity of decaBDE) are first mixed as a dispersion in water (EU. 2002). This mix is
20    stored in tanks and then piped directly into a closed vessel (EU. 2002). The decaBDE-antimony trioxide-
21    water dispersion is added to emulsion polymers and mixed in this closed vessel to formulate the flame
22    retardant (EU. 2002). One analysis found trace amounts of polybrominated dibenzo-p-dioxins (PBDDs)
23    and PBDFs as impurities in commercial mixtures of decaBDE (Ren et al., 2011). No data were found that
24    describe other characteristics of decaBDE flame-retardant formulations or the by-products of their
25    manufacture.
26           During application, the decaBDE flame-retardant formulation is typically back-coated, or applied
27    as a resin to the  reverse surface of textiles along with a binding agent, such as latex or a copolymer (see

      March 2013                                     H-9                   DRAFT - DO NOT CITE OR QUOTE

-------
 1    Section 1.2.2.2). Due to the high efficiency of decaBDE flame retardants, they can be used in
 2    formulations with low loadings compared to other brominated flame retardants (Pure Strategies Inc..
 3    2005). Typical loadings of decaBDE in textiles range from 30 to 40% by dry weight of the dry coating
 4    with different loadings applied to different types of fabrics (30-40 grams/m3 in cotton to 70-80 grams/m3
 5    for velour fabrics) (EU. 2002). After application, the decaBDE flame-retardant upholstery textile is cut,
 6    shaped, and glued or stapled to furniture.

 7    H.2.2.4.2 Potential  Releases during Product Manufacture
 8           Release scenarios for product manufacturing are likely to be similar to those in the material
 9    synthesis and processing stages, but release amounts are probably lower (U.S. EPA. 2005a; EU. 2002).
10    Additionally, releases from this stage likely will not be decaBDE, but rather decaBDE in a polymer
11    matrix. As discussed, decaBDE generally is synthesized as a powder and then mixed into solution when
12    the flame retardant is formulated, minimizing releases of decaBDE to dust (U.S. EPA. 2005a).
13    Nevertheless, manufacture of decaBDE flame-retardant coatings could release vapors if mixing and
14    handling of raw decaBDE occurs in an open system (U.S.  EPA. 2005a). One study found that
15    environmental release was most likely to occur during the mixing of decaBDE powder and cleaning
16    operations of the flame-retardant formulation stage (EU. 2002). Formulation of flame retardants,
17    however, generally  occurs in closed systems with engineering controls that regulate temperature and
18    pressure to minimize potential releases (U.S. EPA. 2005a: EU. 2002). Releases of decaBDE in this stage
19    also can contain the impurities listed in Section H.2.2.4.1.
20           The application of decaBDE flame-retardant coatings to upholstery textiles could result in the
21    release of aerosolized decaBDE due to thermal processing, but release would occur only if the
22    manufacturing plant does not have engineering controls in place to  prevent such releases. Cutting, sewing,
23    shaping, stapling, and other textile finishing processes could result  in the airborne release office
24    decaBDE or decaBDE in a polymer matrix through abrasion. Equipment cleaning also could lead to the
25    release of decaBDE in wastewater during the processing stages of product manufacture.
26           The accidental release scenarios for decaBDE during product manufacture are similar to those in
27    the material synthesis stage (see Section H.2.2.2.2). Additionally, in this stage, spills could lead to release
28    of decaBDE flame-retardant formulations in wastewater. Volatilization is unlikely due to the low
29    volatility of decaBDE (see Table H-3). Table 2-6 in Chapter 2 outlines potential release scenarios from
30    the product manufacturing stage of decaBDE and MWCNT flame-retardant textiles.
      March 2013                                    H-10                  DRAFT - DO NOT CITE OR QUOTE

-------
      H.2.3. Storage and Distribution

      H.2.3.1. Life-Cycle Processes
 1           The storage and distribution stage concerns the handling and transport of (1) decaBDE,
 2    (2) decaBDE flame-retardant formulations, and (3) decaBDE flame-retardant upholstery textiles.
 3    The main method of transport for these materials is not known, but likely would be by truck, train, or
 4    cargo ship.
 5           DecaBDE, if not immediately incorporated into a flame-retardant formulation, likely would be
 6    stored at the site of synthesis/processing in sealed receptacles until it is incorporated into flame-retardant
 7    formulations or transported to sites where manufacture of the flame retardant occurs.
 8           DecaBDE flame-retardant formulations are typically stored at manufacturing plants in drums,
 9    tanks, or more permanent storage vessels until they are packaged and sent to textile manufacturers (U.S.
10    EPA. 2005a).
11           DecaBDE flame-retardant upholstery textiles and end-use products likely also would be stored
12    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
13           Storage and  distribution of (1) decaBDE, (2) decaBDE flame-retardant formulations, and
14    (3) flame-retardant textiles could result in releases to the environment, but all release scenarios are
15    unlikely.
16           Releases of decaBDE separate from flame-retardant formulations are likely to be negligible.
17    Release would be due primarily to accidents, as the materials would be stored in sealed  receptacles after
18    synthesis. Exposure of the receptacles to high heat or fire could lead to the airborne release of decaBDE
19    (see Section H.2.2.2.2 for more details).
20           Releases from flame-retardant formulations could result in releases of decaBDE to the
21    environment (U.S. EPA. 2005a). The possible scenarios for release of decaBDE flame-retardant
22    formulations during storage and distribution include damage to containers holding the flame-retardant
23    formulation, leakage resulting from mishandling of containers, or faulty or improper stacking of cartons
24    in transport vehicles. If the containers are sealed properly and not damaged during transport, releases of
25    product prior to application might be limited to spills.
26           Releases from flame-retardant upholstery textiles could result from accidental  exposure to
27    high heat or fire, off-gassing of volatile components, and infestation with pests. Exposure  of the flame-
28    retardant upholstery textiles to high heat or fire during storage and distribution could lead to the
      March 2013                                    H-11                   DRAFT - DO NOT CITE OR QUOTE

-------
 1    degradation of the polymer matrix and subsequent airborne release of decaBDE (both free and matrix
 2    bound) (see Sections H.2.2.2.2 and H.2.4 for more details). Off-gassing of the volatile components of
 3    decaBDE flame retardants also could occur in poorly ventilated areas that experience high temperatures
 4    (e.g., storage units, warehouses). DecaBDE itself, however, is not highly volatile (see Table 1-8 and Table
 5    1-9. both in Chapter 1; and Table 3-1, in Chapter 3). Infestation of textile or furniture storage facilities
 6    with rodents or other pests also could lead to the release of decaBDE flame=retardant materials to the
 7    environment.
 8           Table 2-7 in Chapter 2 outlines potential release scenarios from the storage and distribution stage
 9    of (1) decaBDE and MWCNTs, (2) decaBDE and MWCNT flame-retardant formulations, and
10    (3) decaBDE and MWCNT flame-retardant upholstery textiles.


      H.2.4. Use

      H.2.4.1. Life-Cycle Processes
11           A wide variety of textiles contain flame-retardant coatings (see Section 1.2). Upholstery textiles
12    are expected to be used in public places where people of all ages will sit, lie, or walk on them. Some
13    unintended uses of upholstery textiles include outdoor use, repurposing for use in other products, burning
14    as kindling, or mouthing by children. Repurposing for use in other products and burning as kindling are
15    covered in Section H.2.5. In general, upholstery textiles are likely to have a lifespan of at least 10 years
16    (EU. 2002).

      H.2.4.2. Potential Releases during the Use Stage
17           Environmental releases from upholstery textiles coated with flame retardants are expected due to
18    (1) the potential use scenarios for the upholstery textiles and (2) the physicochemical properties of
19    decaBDE.  The anticipated long lifespan of upholstery textiles  (>10 years) suggests that releases in this
20    stage could occur over several years (EU. 2002). Indeed, environmental  concentrations of decaBDE in
21    buildings with products containing decaBDE can be high, especially in dust (see Sections H.4.1.2.5 and
22    H.4.1.2.6). One of the most important pathways for these high environmental concentrations is the
23    airborne release of decaBDE sorbed to dust in the environment (see Section H.3.2). The following
24    characteristics of flame-retardant upholstery textiles, however, are expected to reduce releases of
25    decaBDE (EU. 2002):
26           •   Flame-retardant coatings must meet durability requirements to comply with regulations
27               (see Section 1.2.1):

      March 2013                                     H-12                  DRAFT - DO NOT CITE OR QUOTE

-------
 1           •   Flame retardant often is applied to the back of the fabric, minimizing wear and tear; and
 2           •   Upholstery textiles are unlikely to be washed frequently.
 3           The integrity of the flame-retardant coating depends on the strength of the formulation that bonds
 4    it to the textile surface (Som et al., 2011; NRC, 2000). The decaBDE flame-retardant textile coatings
 5    considered in this case study are additive, suggesting that release from upholstery textiles could occur
 6    during the use stage. DecaBDE/antimony trioxide flame-retardant formulations are considered relatively
 7    durable, however, due to the copolymer resin that bonds to the textile fibers (Pure Strategies Inc.. 2005).
 8    Even if migration of decaBDE through the polymer were to occur, it would be expected to be very slow
 9    due to the high molecular weight of decaBDE (Lassen et al.. 1999). In a substance flow analysis of
10    plastics containing decaBDE, however, Lassen et al. (1999) found that release of decaBDE was expected
11    to be greatest during the use stage.
12           Regular use of upholstered furniture (e.g., sitting, walking, lying) could abrade the textile surface
13    and release small amounts of free or matrix-bound decaBDE either into the air or onto the skin of users.
14    Washing of textiles also could lead to water release  of matrix-bound decaBDE. By some estimates, the
15    principal source of decaBDE release in wastewater is due to textile washing (EU. 2002). Most flame-
16    retardant upholstery textiles will be used indoors, minimizing exposure to UV light and weathering.
17    Upholstery textiles that are back-coated with decaBDE flame retardant likely will not be subject to
18    significant abrasion, washing, or UV light. Additionally, most releases of decaBDE initially will be to the
19    indoor environment, but they could spread outdoors through environmental transport mechanisms (see
20    Section H.3) (U.S. EPA.  2010a; Lassen et al.. 1999). These  processes could result in airborne release of
21    decaBDE or releases in wastewater (U.S.  EPA. 2010a).
22           Unintended uses also could lead to the release of decaBDE from flame-retardant textiles. Use of
23    flame-retardant upholstery textiles outdoors could lead to weathering, which could degrade the polymer
24    matrix and release decaBDE. Mouthing by small children, pets, or rodents on flame-retardant textiles
25    could lead to the release of decaBDE directly into children's, pets', or rodents' mouths if the back-coating
26    is exposed and the integrity of the fabric is compromised. Accidental contact of flame-retardant textiles
27    with fire and high heat also could occur and could lead to airborne releases of decaBDE (see Section
28    H.2.2.2.2] for more details). No data were found, however, that describe the likelihood of these releases
29    from this application. Table 2-8 in Chapter 2 outlines potential release scenarios from the use stage of
30    decaBDE and MWCNT flame-retardant textiles.
      March 2013                                     H-13                  DRAFT - DO NOT CITE OR QUOTE

-------
      H.2.5.  Reuse, Recycling, and End of Life
 1           The reuse, recycling, and end-of-life stage encompasses a variety of different transformation and
 2    disposal processes for (1) decaBDE, (2) decaBDE flame-retardant formulations, and (3) decaBDE flame-
 3    retardant upholstery textiles. What the primary reuse, recycling, and end-of-life treatments are for
 4    decaBDE and decaBDE flame-retardant formulations are unclear.

      H.2.5.1. Reuse and Recycling
 5    H.2.5.1.1 Life-Cycle Processes
 6           Reuse or recycling of decaBDE or decaBDE flame-retardant formulations is unlikely. On the
 7    other hand, textile waste  often is recovered and reused or recycled (Kohler et al., 2008); upholstered
 8    furniture is sometimes reused, but is rarely recycled (CalRecycle. 2002). Upholstery could be donated to
 9    charitable organizations and resold for residential use. Additionally, upholstery textiles could be
10    informally repurposed into clothing, blankets, and other textile products. Due to the difficulty of recycling
11    furniture and flame-retardant materials, flame-retardant furniture is  typically land-filled (CalRecycle.
12    2002; Lassenet al., 1999). Of the small portion of upholstered furniture that is recycled, about 60% of the
13    material is recycled and 25-30% is composted (CalRecycle, 2002).  No data were found that describe the
14    proportion of other upholstery textiles (e.g., mattress ticking or curtains) that are typically recycled.
15           The main types of textile recycling processes are fiber-to-fiber recycling and polymer reduction
16    recycling. During the fiber-to-fiber process, textiles are shredded and blended with other fibers to create a
17    new mixture ready for spinning (Kohler et al.. 2008). During the polymer reduction process, textiles are
18    cut and granulated to form pellets that are processed to break down  the polymer to the molecular level to
19    be reused as raw material (Kohler et al.. 2008). No data were found that describe the prevalence of each
20    recycling process.

21    H.2.5.1.2 Potential Releases during the Reuse/Recycling Stage
22           Release of decaBDE beyond releases described in the use stage is unlikely to occur during reuse
23    of flame-retardant upholstery textiles. Older textiles could release greater levels of decaBDE, however,
24    due to increased degradation of the material. Informal repurposing of flame-retardant textiles likely would
25    require cutting and shredding, resulting in possible air release of decaBDE. Airborne releases of decaBDE
26    could occur during recycling of flame-retardant textiles. Recycling subjects textiles to a variety  of
27    mechanical, thermal,  and chemical treatments that could result in the airborne releases of additive flame
28    retardants from fibers (Kohler etal.. 2008). One analysis found airborne releases of decaBDE at a plastic
29    recycling plant with the highest concentrations of airborne particles measured near the shredder (Sjodin et

      March 2013                                     H-14                   DRAFT - DO NOT CITE OR QUOTE

-------
 1    al., 2001). Although releases from recycling of upholstery textiles containing decaBDE flame retardant
 2    might be similar to those of plastics, the processing of plastics is likely to differ from that of textiles.
 3    Release of decaBDE to water also could occur during chemical treatment and processing. Although
 4    release of decaBDE is possible during recycling of flame-retardant textiles, no data were found that
 5    indicate the likelihood of release from recycling processes.
 6            Table 2-9 in Chapter 2 outlines potential release scenarios from the reuse/recycling stage of
 7    decaBDE and MWCNT flame-retardant textiles.

      H.2.5.2. Incineration
 8    H.2.5.2.1 Life-Cycle Processes
 9            The incineration of decaBDE or decaBDE flame-retardant formulations is unlikely, but any
10    incineration likely would occur in a hazardous waste incinerator. Upholstery textiles treated with
11    decaBDE flame-retardant coatings might be sent to municipal incinerators for processing. Municipal
12    incinerators  generally provide a well-controlled environment with pollution control mechanisms and
13    sufficiently high temperatures (850 °C) to destroy most materials (Kohler et al.. 2008). Processing in
14    municipal facilities is likely to  result in complete incineration of the upholstery textiles. Alternatively,
15    upholstery textiles also might be incinerated in less we 11-controlled facilities or burned in open fires as a
16    rudimentary form of waste management or as kindling. These incineration methods are likely to result in
17    incomplete incineration of the upholstery textiles. No data were found that describe the prevalence of
18    incineration as a form of disposal for upholstery textiles or what proportion of incinerated textiles is
19    processed at we 11-controlled incineration facilities.

20    H.2.5.2.2 Potential Releases during the Incineration Stage
21            Airborne releases of decaBDE from well-controlled incineration are expected to be negligible,
22    but incomplete incineration (e.g., open fires) could lead to some airborne release. Little empirical data
23    exist that describe the prevalence of decaBDE in incinerator residues, but decaBDE is expected to be
24    destroyed by the high-temperature incineration used at most municipal incineration facilities (Palm et al..
25    2002; Lassen et al.. 1999). These temperatures also are  sufficiently high to prevent the formation of
26    PBDFs and PBDDs during incineration of decaBDE  (see Section H.2.2.2.2). In one study of atmospheric
27    concentrations of PBDEs near solid-waste incinerators, the authors found that incineration facilities do
28    not give rise to a substantial proportion of atmospheric releases of decaBDE (Agrell etal.. 2004).
29    Incomplete incineration, however, could lead to the airborne release of decaBDE and the formation of
30    PBDFs, PBDDs, polychlorinated dibenzo-p-dioxins,  and nonhalogenated substances such as polycyclic
      March 2013                                      H-15                   DRAFT - DO NOT CITE OR QUOTE

-------
 1    aromatic compounds (see Section H.2.2.2.2). Current pollution control technologies for municipal
 2    incinerators are expected to effectively filter these emissions and prevent their release to the environment
 3    (EU. 2002).
 4           Due to the high temperatures and pollution control mechanisms at municipal incinerators,
 5    decaBDE in flame-retardant textiles are expected to be destroyed during well-controlled incineration.
 6    Incineration by open flame in uncontrolled environments, however, might lead to airborne releases of
 7    decaBDE and harmful by-products.
 8           Table 2-10 in Chapter 2 outlines potential release scenarios from the incineration stage of
 9    decaBDE and MWCNT flame-retardant textiles.

      H.2.5.3. Land-Filling
10    H.2.5.3.1 General Processes
11           Land-filling of decaBDE or decaBDE flame-retardant formulations is unlikely, except in the case
12    of floor sweepings from manufacturing facilities. Upholstered furniture and textiles generally are
13    disposed of in municipal landfills  (Kohler et al.. 2008). Remaining parts from recycled furniture, such as
14    cover cloth materials, also are  sent to the landfill (CalRecycle. 2002). Additionally, some textiles might
15    be disposed of in uncontrolled landfills or open dumping sites that have no pollution control mechanisms
16    in place. No data were found that describe the proportion of upholstery textiles disposed of in landfills or
17    any further processing that might occur at the landfill.

18    H.2.5.3.2 Potential Releases during the Land-filling Stage
19           Land-filling of decaBDE flame-retardant textiles could lead to water and air releases due to
20    mechanical processes such as mixing and compacting. DecaBDE also could leach from land-filled textiles
21    and migrate into the  underlying soil  or ground water (Rahman et al.. 2001; Lassen et al.. 1999). however,
22    no evidence of decaBDE in land-fill leachate has been found [Kim et al. (2006) as cited in Wright et al.
23    (2008)]. Additionally, this release scenario is unlikely due to the low leaching potential of decaBDE (see
24    Table 1-8). Flame retardants containing decaBDE could volatilize to the atmosphere over time (Rahman
25    et al.. 2001). but volatilization of decaBDE is expected to be negligible due to low volatility of decaBDE
26    (Palm et al.. 2002) (see  Table 3-1  in Chapter 3).
27           Few data were identified that measure releases of decaBDE from land-filling of flame-retardant
28    textiles, but the physicochemical characteristics of these materials suggest that such releases likely would
29    be small. Table 2-11 in  Chapter 2  outlines potential  release scenarios from the land-filling stage of
30    decaBDE and MWCNT flame-retardant textiles.
      March 2013                                     H-16                  DRAFT - DO NOT CITE OR QUOTE

-------
      H.2.5.4. Wastewater Treatment Plants
 1    H.2.5.4.1 Life-Cycle Processes
 2           The wastewater treatment process consists of filtering and treating wastewater to remove solids
 3    and contaminants. Large facilities that manufacture decaBDE and decaBDE flame retardants might divert
 4    their wastewater to an on-site wastewater treatment plant. Alternatively, some wastewater from these
 5    facilities might be directly processed by municipal wastewater treatment plants. Water releases of
 6    decaBDE that occur during the storage and distribution, use, and reuse/recycling/end-of-life stages also
 7    would be treated in municipal wastewater treatment plants.

 8    H.2.5.4.2 Potential Releases during the Wastewater Treatment Stage
 9           Release of decaBDE or decaBDE flame-retardant formulations into wastewater could occur
10    throughout the life cycle. Primary releases to wastewater during manufacturing stages are due to
11    equipment cleaning, formulation and application of the flame retardant, and accidental spills. Washing
12    processes (which can involve abrasion, detergents, and water), particularly in the product manufacturing
13    stages, are likely to result in the release of additive flame retardants from textiles to wastewater (Som et
14    al., 2011). Due to the physicochemical characteristics of decaBDE flame retardants (see Table 1-8 and
15    Table 1-9 in Section 1.3), the material is likely to sorb onto particles during water treatment and be
16    removed in sludge (Somet al., 2011; Lassenet al., 1999). The potential nonetheless exists for releases
17    from filter backwash and other wastewater treatment plant equipment (EU. 2002). Additionally, some of
18    this removed sludge  is deposited in landfills or spread on agricultural soil (EU, 2002; Lassenet al.. 1999).
19    This activity represents one of the most significant potential releases to soil of decaBDE flame-retardant
20    coatings (Ciparis and Hale. 2005; Lassen et al.. 1999). The releases of decaBDE from wastewater
21    treatment facilities are expected to be small. The removal efficiency of wastewater treatment plants is not
22    well characterized for decaBDE and the spread of sewage sludge onto  agricultural soil could represent a
23    significant source of decaBDE to soil. See Section H.3.3.3 for information regarding decaBDE removal
24    efficiency of wastewater treatment plants.
25           Table 2-11 in Chapter 2 outlines potential release scenarios from the wastewater treatment stage
26    of decaBDE and MWCNT flame-retardant textiles.
      March 2013                                    H-17                  DRAFT - DO NOT CITE OR QUOTE

-------
       H.3.  Transport, Transformation, and Fate
 1           Releases throughout the product life cycles of upholstery textile coatings containing decaBDE
 2    flame retardant will, to some extent, lead to occurrence of primary and secondary contaminants in air,
 3    soil, and aquatic media. Chapter 3 examines what might happen to these substances after their release to
 4    the environment, including transport or transformation through chemical, physical, and biological
 5    processes. Studies investigating the transport, transformation, and fate of decaBDE in the environment are
 6    summarized in Appendix D. and concentrations of BDE-209, the single isomer of deca-substituted BDE,
 7    in environmental compartments are provided in Appendix E.
 8           DecaBDE can be released into the environment during the manufacturing, storage, distribution,
 9    use, disposal, reuse, and recycling of upholstery textiles treated with flame retardants (see Chapter 2).
10    DecaBDE flame-retardant formulations are used primarily as additives that are mixed with, not
11    chemically bound to, polymers in textile products. Because they are not chemically bound, these
12    substances can escape from the material and become a source of contamination to surrounding
13    environmental media (Yu et al.. 2010; Vonderheide etal.. 2008; Moniruzzaman and Winey. 2006; Song
14    et al.. 2006; Soderstrom et al.. 2004). Although some, if not most, releases after the production stage are
15    likely to be in the matrix-bound form, little information exists that describes the environmental behavior
16    of decaBDE-polymer complexes. As a result, this section focuses on the transport, transformation, and
17    fate of decaBDE not embedded in a polymer matrix.
18           Section H.3.1 provides a brief discussion of the chemical and physical characteristics and the
19    processes that influence behavior (e.g., mobility, persistence, bioavailability) of decaBDE in
20    environmental media. The sections that follow summarize the available information regarding the
21    behavior of each substance in indoor and outdoor air (Section H.3.2). aquatic systems (Section H.3.3).
22    and terrestrial systems (Section H.3.4). A brief discussion of models that might be used for evaluating the
23    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
24           The environmental fate of BDE-209 will be dictated by its chemical and physical properties and
25    its propensity for biotic and abiotic transformation. BDE-209 could transform physically, chemically, or
26    biologically once released to the environment, leading to substances that present a very different hazard
27    than the hazard of the untransformed material originally released. BDE-209 has been shown to
28    biologically and photolytically debrominate (lose a bromine atom) to form lower brominated congeners
      March 2013                                    H-18                  DRAFT - DO NOT CITE OR QUOTE

-------
 1    that are more readily bioavailable [U.S. EPA (2010a); Vonderheide et al. (2008); Song et al. (2006);
 2    Watanabe and Sakai (2003); Darnerud et al. (2001); see Text Box H.3-1 and additional sources in
 3    Appendix D, Table D-11. Because the chemical properties associated with transformation products of
 4    decaBDE influence their transport, transformation, and fate in the environment, degradation processes of
 5    decaBDE are introduced in this section. A summary of key physicochemical factors that are likely to
 6    affect partitioning27 and fate of BDE-209 and related PBDEs in the environment is presented in Table
 7    H-4. Values for key physicochemical properties of BDE-209 are provided in Section H.I (see Table H-3).
 8           Biotic debromination is the breakdown of BDE-209 into lower brominated compounds by aerobic
 9    and anaerobic microorganisms. Biotic transformation processes for BDE-209 that occur in soil, sediment,
10    or sewage sludge have been described in recent literature (see Appendix D. Table D-l). These processes
11    result in dehalogenation through microbe catalysis reactions that stimulate the replacement of a halogen
12    atom (e.g., bromine, chlorine, fluorine) with a hydrogen atom (Kuivikko et al.. 2010; Kim et al.. 2007).
13           Photolysis or photodegradation is a chemical (abiotic) process by which molecules are broken
14    down through the  absorption of light. PBDEs are vulnerable to photolysis, which induces reductive
15    debromination causing higher brominated congeners like BDE-209 to photodegrade to form lower
16    brominated congeners. These lower brominated congeners are potentially more stable and bioavailable in
17    the environment due to lower molecular weight and a lower octanol/water partition coefficient (Kow)
18    (Soderstrom et al., 2004). PBDFs also have been identified as photolysis products of BDE-209. Sunlight
19    could degrade BDE-209 in air, surficial soils, water, and surficial sediments via photolysis (Christiansson
20    et al., 2009;  Soderstrom et al., 2004). This and other abiotic transformation processes for BDE-209 have
21    been demonstrated in recent literature (see Appendix D. Table D-l).
22           Transformation also can occur with elevated temperatures (e.g., incineration, fire); thermal
23    breakdown products of PBDEs include polybrominated, polychlorinated, and mixed
24    brominated/chlorinated dibenzo-p-dioxins and dibenzofurans, and are similar to polychlorinated dibenzo-
25    p-dioxins/polychlorinated dibenzofurans in their persistence and toxicity (Watanabe and Sakai. 2003;
26    Darnerud etal.. 2001; Rahman etal.. 2001).
27    DecaBDE formulations used in textile and other products contain the fully brominated congener, which is
28    less mobile in the  environment than lower brominated congeners, probably due to low volatility, water
29    solubility, and bioaccumulation, and the high propensity to adsorb to sediments. The lower brominated
30    compounds are generally more volatile, water soluble, and bioaccumulative compared with higher
31    brominated compounds (Watanabe and Sakai. 2003) and are believed to be structurally analogous to
      27Partitioning 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. 2010a).
      March 2013                                     H-19                   DRAFT - DO NOT CITE OR QUOTE

-------
1     polychlorinated biphenyls (PCBs), so their chemical properties, persistence, and behavior in the

2     environment, are expected to follow similar patterns. PBDEs are expected to be more vulnerable to

3     environmental degradation than PCBs, however, because their carbon-bromine bonds are weaker than the

4     carbon-chlorine bonds of PCBs (Shih and Wang. 2009; Watanabe and Sakai. 2003; Rahman etal.. 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
              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
                                                  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).
degradation by engineered nanoparticles (Vonderheide et al., 2008). Biotic degradation pathways include debromination by
both aerobic and anaerobic microbes found in soil (Wang etal., 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.
      March 2013
                                                    H-20
 DRAFT - DO NOT CITE OR QUOTE

-------
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 bioavai lability 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 (Ahn et
                            • 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 (2010a); Watanabe and Sakai (2003)

High octanol/water partition    •  PBDEs partition between water and sediment based on solubility and K0w
coefficient (Kow)              .  Less mobility, strongly sorbed to soils, sediments, sludge
                            •  Not easily distributed within surface water and ground water
                            Sources: U.S. EPA (2010a); 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 (2010a); 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)

PBDE = polybrominated diphenyl ether; BDE-209 = single  isomer of decabrominated diphenyl ether
March 2013
                                                      H-21
DRAFT - DO NOT CITE OR QUOTE

-------
      H.3.2. Transport, Transformation, and Fate in Air
 1           BDE-209 released from the flame-retardant upholstery textile coatings life cycle could reach
 2    indoor and outdoor air in several ways. For example:
 3           •   BDE-209 can be released directly into ambient air during all stages of the product life cycle,
 4               as previously described in Section H.2. BDE-209 that remains in the particle phase can
 5               disperse through air away from the source of release.
 6           •   Particulate BDE-209 can become suspended in the surrounding indoor or outdoor air during
 7               multiple stages of the product life cycle.
 8           •   Particulate BDE-209 might remain suspended and be transported through the atmosphere or
 9               deposited onto surfaces. Particles that have been deposited on surfaces could become
10               resuspended in the air and redeposited elsewhere.
11           If released indoors, BDE-209 can distribute indoors to air, dust, vacuums, and air filter systems. If
12    the source of BDE-209 is inside a building, levels of BDE-209 in indoor air can be much higher than in
13    outdoor air (Hale et al., 2006). Indoor air sources also could contribute to outdoor air concentrations.
14    Variability in the indoor air and dust levels of BDE-209 can be influenced by indoor/outdoor exchange
15    rates, building ventilation rates, and the number and age of all PBDE-treated products (e.g., electronics,
16    mattresses, draperies, furniture) present in the building (Hazrati and Harrad. 2006).
17           As mentioned earlier, BDE-209 has very low  vapor pressure and a high octanol/water partition
18    coefficient and is therefore more likely to be transported on particles in the air than as a vapor (see Table
19    H-4) (Breivik et al.. 2006). Approximately 99% of BDE-209 in ambient air is expected to be present in
20    the particle phase (U.S. EPA. 2010a). Particles could remain suspended in air or deposit on surfaces with
21    the potential for resuspension. Because nonvolatile compounds like BDE-209 tend to sorb to particles,
22    they likely would be concentrated in household dust and could experience an extended indoor lifetime
23    (Kemmlein et al.. 2003).
24           When attached to particles, BDE-209 can have an extended residence time and persistence in the
25    atmosphere. Longer residence time in the atmosphere  allows more time for the particles to be mobilized
26    by wind and other forces and makes long-range atmospheric transport (LRT) in the atmosphere possible
27    (Gouin et al.. 2006; Wania and Dugani. 2003). Evidence exists for LRT of PBDEs to remote ecosystems,
28    including the Arctic (de Wit etal., 2010;  Su et al., 2009; Agrell et al., 2004); some literature suggests that
29    PBDEs in remote Arctic regions originated in urban areas in North America (Breivik et al., 2006).
30           LRT has been considered by some to be the reason for PBDE occurrence in rural and more
31    remote sites. Some evidence has shown LRT of BDE-209 adsorbed to airborne particulate matter at
32    regional and global scales (Vonderheide et al.. 2008).  although other researchers have suggested that the
      March 2013                                    H-22                  DRAFT - DO NOT CITE OR QUOTE

-------
 1    potential for LRT of BDE-209 is low. Atmospheric deposition of BDE-209 and other PBDEs is thought
 2    to be a main source of these contaminants in background waters and soils (Vonderheide et al., 2008).
 3           Understanding of BDE-209 LRT behavior is limited. As stated previously, BDE-209 in the air is
 4    primarily attached to particles (U.S. EPA. 2010a: Su et al.. 2009). Efforts by Mueller and Nowack (2008)
 5    to model atmospheric transport of engineered nanoparticles can be used to infer that materials, such as
 6    BDE-209, adsorbed to particles will eventually deposit or wash out (wet deposition) in aquatic or
 7    terrestrial systems. In turn, the fate of BDE-209 in air is likely dictated by the characteristics of the
 8    particles to which it adsorbs.
 9           Plants exposed to BDE-209 from air can accumulate BDE-209 and act as transfer vectors in the
10    food chain. Airborne BDE-209-laden particles could attach to leaves and other aboveground parts of
11    plants and translocate to different tissues of the plant. Salamova and Kites (2010) evaluated PBDE levels
12    in air samples and tree bark and determined that BDE-209 concentrations in tree bark were strongly
13    correlated with concentrations in the air and in precipitation. The highest air and tree bark concentrations
14    occurred at urban sites.
15           Breakdown and transformation of BDE-209 in the air by photolysis also can occur (see Section
16    H.3.1 and Text Box H.3-1). and studies that evaluated this phenomenon in air are presented in Appendix
17    D, Table D-l. Temperature variability could be an important factor that explains seasonal patterns of
18    BDE-209 burdens in air because changes in temperature affect gas-solid partitioning coefficients and
19    subsequently the transfer and retention of BDE-209 in air (Vonderheide et al.,  2008).
20           Section H.4.1.2 and Section E.I of Appendix E provide summaries of studies with BDE-209
21    concentration data in building dust and indoor/ambient air.


      H.3.3. Transport, Transformation,  and Fate in Water and Sediment
22           BDE-209 released from the flame-retardant upholstery textile coating  life cycle could enter
23    aquatic systems in several ways. For example:
24           •   BDE-209 in ambient air subsequently could be deposited or washed out to aquatic systems.
25           •   Erosion of contaminated soil could release BDE-209 to surface waters.
26           •   Runoff flowing along the ground surface could transfer BDE-209 in contaminated soil to
27               nearby waterways.
28           •   Wastewater effluents containing BDE-209 could be a source of contamination to receiving
29               water bodies near the discharge location.
30           •   BDE-209 could leach from land-filled sewage sludge into subsoil and ground water and
31               migrate to surface water or sediment.
      March 2013                                     H-23                  DRAFT - DO NOT CITE OR QUOTE

-------
      H.3.3.1. Surface Water and Sediment (Inland and Coastal)
 1           Water solubility and Kow are important physicochemical factors for predicting behavior of
 2    BDE-209 in the aqueous phase—and these parameters predict that BDE-209 will partition to the
 3    particulate phase (e.g., sorb to suspended organic matter) in water or bind strongly to sediments (U.S.
 4    EPA, 2010a; Hale et al., 2006; Watanabe and Sakai, 2003). The lower brominated congeners are more
 5    water soluble than the higher brominated congeners and are expected to be more mobile in water
 6    (Watanabe and Sakai. 2003).
 7           Sediment is both a sink and a reservoir for PBDEs such as BDE-209. In general, BDE-209 is the
 8    dominant congener in sediment samples (Tokarz et al.. 2008); notably higher concentrations in urban and
 9    industrial areas and near outfalls of wastewater treatment plants have been measured (U.S. EPA. 2010a)
10    (see Appendix E). Whether the higher concentration of BDE-209 in sediment samples is due to greater
11    BDE-209 use or less environmental degradation compared with other congeners is unknown.
12    The physicochemical properties of BDE-209 and the characteristics of sediment both affect the
13    bioavailability of BDE-209 in sediment. The bioavailability of BDE-209 in sediment is expected to be
14    limited due to its strong hydrophobicity and large molecular size (Liu et al.. 201 Ib). Due to its tendency
15    to sink to sediments, BDE-209 might be bioavailable to benthic organisms, but generally not to water-
16    column-dwelling organisms. Benthic organisms might, therefore, act as vectors for the transport of
17    decaBDE through the food web (see Section H.4.3).
18           The amount of organic matter in the sediment is an important factor controlling the partitioning of
19    BDE-209 in sediments (Liu et al., 20lib). In a study conducted by Liu et al. (20lib), the authors
20    concluded that desorption  of BDE-209 in sediment was more  difficult as contaminant-sorbent interaction
21    time increased, likely because of entrapment of BDE-209 molecules in the micropores of organic matter.
22    Total desorption also decreased with increased total organic carbon in the sediments.
23           Physicochemical properties of PBDEs and the characteristics of sediment might not serve as
24    perfect predictors of levels in surface waters due to microbial  or photolytic degradation processes that
25    could reduce concentrations of the material. Elevated quantities of hydroxylated PBDEs (a possible
26    oxidation product of PBDEs) have been measured in surface waters near sewage treatment plants
27    (Vonderheide et al.. 2008). Anaerobic microbial reductive debromination is potentially a driving
28    transformation process in sediment (see Section H.3.1 and Appendix D. Table D-l): the process,
29    however, can be very slow.
30           Appendix E provides a summary of studies with BDE-209 concentration data in surface water
31    and sediment. Most identified studies focus on sediment concentrations of BDE-209.
      March 2013                                    H-24                   DRAFT - DO NOT CITE OR QUOTE

-------
      H.3.3.2. Ground Water
 1           BDE-209 present in soil could leach into subsoil and ground water. Based on its physicochemical
 2    properties, BDE-209 does not dissolve in water, but could sorb to suspended organic matter in the ground
 3    water plume and migrate to surface water (U.S. EPA. 2010a). No data were found on concentrations of
 4    BDE-209 in ground water.

      H.3.3.3. Wastewater
 5           As introduced in Section H.2. manufacturers, homes, and public buildings could be significant
 6    sources of BDE-209 to wastewater treatment plants and municipal sewage treatment facilities (Hale et al.,
 7    2006). Because of its hydrophobicity and tendency to partition to solids, most BDE-209 in wastewater
 8    would be expected to sorb to settling solids (i.e., will not remain in effluent) during the sewage treatment
 9    process (Ricklund et al.. 2009: North. 2004).
10           BDE-209 partitions strongly to particulate matter, so sewage sludge is expected to be a major sink
11    for BDE-209, while effluent is likely dominated by the more water soluble lower brominated congeners.
12    Sewage sludge applied to agricultural fields could be a source of BDE-209 to soils (Huang etal.. 2010;
13    Vrkoslavova et al.. 2010: U.S. EPA. 2009: Vonderheide et al.. 2008: Knoth et al.. 2007: Hale et al.. 2006:
14    Law et al.. 2006b: Sellstrom et al.. 2005). Runoff along the surface of the ground then could transfer
15    BDE-209 in the sewage sludge to nearby terrestrial systems or waterways. Sludge contaminated with
16    BDE-209 could be disposed of in landfills (and possibly leach to subsoils and ground water) or be
17    incinerated  (introducing emissions to the atmosphere).
18           Wastewater effluents, although not dominated by the higher brominated PBDEs, might be a
19    source of BDE-209 and transformation product contamination to receiving water bodies of local aquatic
20    ecosystems near the discharge location (Peng et al.. 2009;  Song etal.. 2006). Wastewater irrigation for
21    farmlands could be  a source of PBDEs in agricultural soils. The mobility of BDE-209 in the receiving soil
22    could be enhanced if it sorbs to dissolved organic matter in the irrigation water (Wang etal.. 2010b).
23           Recent studies that present concentrations of BDE-209 in wastewater effluent and sludge are
24    presented in Section H.4.1.2.2 and Section E.I of Appendix E.
      March 2013                                    H-25                   DRAFT - DO NOT CITE OR QUOTE

-------
      H.3.4.  Transport, Transformation, and Fate  in Soil
 1           BDE-209 released from the flame-retardant upholstery textile coatings life cycle could enter
 2    terrestrial ecosystems in several ways:
 3           •   BDE-209 in ambient air subsequently could be deposited on soil and plants. Some particles
 4               that deposit on soil or plants might experience secondary transport via wind and become
 5               resuspended into ambient air and redeposited into nearby terrestrial ecosystems.
 6           •   Runoff flowing along the ground surface could transfer BDE-209 in contaminated soil to
 7               nearby terrestrial ecosystems.
 8           •   Disposal products  containing BDE-209 could be deposited in solid waste landfills. Sewage
 9               sludge containing BDE-209 also might be land-filled. BDE-209 in land-filled waste could
10               leach into subsoils.
11           Similar to sediments, soils are a major sink for PBDEs, including BDE-209. BDE-209 released
12    from the flame-retardant upholstery textile coating life cycle can enter terrestrial ecosystems and
13    distribute in soil and plants.
14           As with sediment, water solubility (hydrophobicity) and Kow are important physicochemical
15    factors for predicting behavior of BDE-209 in soil, and these parameters predict that BDE-209 will
16    associate primarily with organic (carbon-rich) particles in soil and experience limited mobility (U.S. EPA.
17    2010a: Yu et al..  2010; Zhuetal..  2010). In addition, percent total organic carbon likely plays a major
18    role in BDE-209  transport and distribution in soil (Zou et al.. 2007).
19           Plants exposed to BDE-209 from soil also can accumulate  BDE-209 and potentially transform it
20    to lower brominated compounds. Debromination in this context has been investigated only in the soil-
21    plant system, however, not in plants alone; the possibility therefore remains that debromination might
22    occur exclusively in the soil, after which plants take up the transformation products (Huang et al.. 2010).
23    If present in soils, BDE-209 could contact plant roots and partition to root lipids due to its high
24    lipophilicity and thereby transport into plant tissues (Huang  et al.. 2010; Vrkoslavova et al.. 2010). Huang
25    et al. (2010) studied the uptake, translocation, and metabolism of BDE-209 in six plant species and found
26    that root lipid content was positively correlated with BDE-209 uptake in those species.  BDE-209
27    accumulated in the roots and shoots of all plants. The translocation factor, measured as the
28    Concentrationshoot/Concentrationroot of BDE-209, however, was inversely related to BDE-209
29    concentration in the  roots, suggesting root lipids restrict translocation of BDE-209 from roots to shoots
30    because of its partitioning to root lipids. In contrast to Huang et al.  (2010). the study conducted by
31    Vrkoslavova et al. (2010) provided evidence of translocation of BDE-209 and other PBDEs in plants from
32    the root lipids of tobacco plants to other plant tissues. Concentrations in soil will further depend on the
33    amount of vegetative cover, which also could scavenge BDE-209 (U.S. EPA. 2010a).
      March 2013                                     H-26                  DRAFT - DO NOT CITE OR QUOTE

-------
 1           Evidence of BDE-209 debromination (transformation) by soil microorganisms and photolysis in
 2    surface soils was presented earlier (see Section H.3.1 and Appendix D. Table D-l). Other studies relevant
 3    to the fate and transport of BDE-209 in soil and plants are provided in Appendix D. Table D-3. Evidence
 4    that BDE-209 affects soil microbial community structure and function is also available; some of these
 5    studies are also discussed in Section H.5.2.2.1  and Section F.2.3 of Appendix F.
 6           Recent studies that present soil concentrations of BDE-209 have been identified; some of these
 7    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
 8           Multimedia models to predict environmental fate and transport of BDE-209 and PBDE congener
 9    profiles in environmental media have been used in recent studies. Breivik et al. (2006) used a multimedia
10    fate and transport model to provide further understanding of how temporal variability and forest cover
11    help control LRT distance for BDE-209 and determined that the fate of atmospheric BDE-209 is likely to
12    be controlled by deposition. Results of their study suggested that the variability of precipitation and the
13    occurrence of periods without precipitation and with strong winds can impact LRT of BDE-209. They
14    determined that the "forest filter effect" might also be important to the LRT of BDE-209; scenarios with a
15    forest yielded lower estimates of air travel distance than scenarios without a forest. Gouin et al. (2005)
16    used a multimedia mass-balance model to assess the importance of seasonal variability (including snow
17    pack, temperature, forest canopy) on concentrations  of PBDEs in air, although results were not specific to
18    BDE-209. The authors concluded that PBDE concentrations experience a "spring-pulse" due to particle-
19    bound deposition of PBDEs in the snow pack during winter, followed by transfer of PBDEs to the soil
20    surface following snow melt and volatilization back into the atmosphere as springtime temperatures
21    increase. Emerging spring foliage then takes up PBDEs, decreasing atmospheric concentrations during the
22    summer months and inhibiting LRT.
23           Bogdal et al. (2010) used PBDE measurement data for the Lake Thun catchment area,
24    Switzerland, in air, lake water, lake  sediment, and tributary water, and combined results from a
25    multimedia fate model that used site-specific environmental parameters from the lake catchment to
26    predict PBDE congener patterns in water and sediment. They incorporated measured loadings of PBDEs
27    in air, tributaries, and wastewater into their model. The authors reported that their model successfully
28    predicted general PBDE congener patterns in water and sediment, but that the model tended to
29    underestimate concentrations in water and overestimate concentrations in sediment relative to measured
30    concentrations. The parameter driving this  discrepancy appeared to be partitioning of PBDEs between the
31    aqueous dissolved phase and suspended particulate matter phase in the water column.

      March 2013                                    H-27                  DRAFT - DO NOT CITE OR QUOTE

-------
       H.4.  Exposure-Dose
 1            Releases of decaBDE to the indoor and outdoor environments can occur at multiple stages of the
 2    product life cycle for flame-retardant upholstery textile coating (see Section H.2). and subsequent
 3    transport, transformation, and fate processes dictate how decaBDE distributes through various
 4    environmental media once released (see Section H.3). Exposure describes the pathways through which
 5    contact occurs between contaminants in the environment and living organisms and abiotic receptors.
 6    Toxicokinetics (i.e., ADME) describes the processes that relate exposure (or dosage) to the internal dose,
 7    which refers to the quantity of a chemical or material that is taken up and absorbed by living organisms
 8    (U.S. EPA. 201 Ob).28
 9            Section H.4.1 introduces analytical techniques for identifying, characterizing, and measuring
10    decaBDE in various matrices. The various metrics recommended for characterizing exposure and dose of
11    decaBDE are also discussed, and available concentration data in various indoor and outdoor media are
12    presented. In the absence of data  quantifying decaBDE exposures at the point of contact, measured
13    concentrations of decaBDE in surrounding media can be used to estimate exposures using a scenario
14    evaluation approach. Sections H.4.2 and H.4.3 expand on the release scenarios presented in Section H.2
15    to discuss the potential human and ecological exposure pathways that link those releases to receptors.
16    No data were identified regarding relevant exposure pathways leading to impacts on abiotic resources
17    (e.g., the manmade environment); as a result, this CEA case study does not include a discussion of
18    exposure scenarios that would influence abiotic receptors. Although broad potential impacts on society
19    and the global environment are discussed in Section H.5.3. exposure is either not considered germane to
20    the discussion of the impact (such as for economic impacts of manufacturing) or the exposure
21    characteristics related to the impact are already included in the general discussion that follows (higher
22    potential exposures levels related to such as for socioeconomic impacts).
23            Probable exposure scenarios throughout the flame-retardant upholstery textile coating life cycle
24    are identified for workers, consumers, the general public, and highly exposed populations in
      28The 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."
      March 2013                                      H-28                   DRAFT - DO NOT CITE OR QUOTE

-------
 1    Section H.4.2 and for aquatic and terrestrial biota in Section H.4.3. These scenarios describe the
 2    conditions under which exposures might occur; this information can be used in combination with
 3    measured or modeled concentrations in environmental media from Section H.4.1 and exposure factors to
 4    estimate exposures. Kinetic information then can be used to determine or estimate the internal dose that
 5    results from external exposures. When available, point-of-contact measurements, administered dosages,
 6    tissue or body burdens, and scenario-specific exposure guidelines and recommendations are provided, and
 7    the toxicokinetics of decaBDE are described. Finally, Section H.4.4 discusses aggregate exposures to
 8    decaBDE from multiple sources and Section H.4.5 discusses cumulative exposures to multiple related
 9    stressors. Measured concentrations of decaBDE in environmental media are provided in Appendix E. and
10    studies describing toxicokinetics of PBDEs in mammals are summarized in Appendix F.
11           As described in Section H.2.2.4. decaBDE is expected to be incorporated into a polymer or other
12    type of matrix in the flame-retardant formulation applied to upholstery textiles, and both the free and
13    matrix-bound form might be released during the product life cycle. Very little data relevant to BDE-209
14    exposures, however, have been generated for the matrix-bound form. This lack of data necessitates a
15    reliance on the existing data for free BDE-209 in the discussion throughout this section. The extent to
16    which exposure characteristics and dose implications differ between the free and matrix-bound forms of
17    BDE-209,  however, is unknown at this time.


      H.4.1. Detection, Measurement, and Characterization
18           Exposure scenario evaluation requires information on measured, modeled, or reasonably
19    estimated concentrations of a stressor in exposure media. As introduced in Section H.I. PBDEs represent
20    a group of compounds encompassing substances that span a range of physicochemical characteristics and
21    properties. As a result, developing reliable analytical techniques for detecting, measuring, and
22    characterizing the full range and makeup of PBDEs in environmental media can present challenges. Text
23    BoxH.4-1 provides an abbreviated discussion of a few common analytical techniques and the general
24    challenges associated with them. Appendix B summarizes common analytical techniques for each
25    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 Box H.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
      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

      March 2013                                     H-29                  DRAFT - DO NOT CITE OR QUOTE

-------
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.
March 2013                                    H-30                  DRAFT - DO NOT CITE OR QUOTE

-------
                                                         3000 •
                                                         20K1 -
                                                                                          PBDESTD
                          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.
                                                         (000
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
due to these  nonstandardized PBDE  congeners or to  other
compounds can be challenging (Stapleton, 2006).
                                                                             Retention Time (min)

                                                            Example Gas Chromatograph Output for a PBDE Mixture with
                                                            Multiple Unidentified Peaks (Stapleton, 2006]
1
2
3
4
5
HALL 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).
March 2013
                                                       H-31
DRAFT - DO NOT CITE OR QUOTE

-------
      H.4.1.2. Concentrations in Environmental Media and Indoor Environments
 1           As described in the previous section, exposures can be estimated by combining knowledge of
 2    concentrations in exposure media with assumptions about contact of humans, biota, or abiotic surfaces
 3    with those media. The following sections describe the information available on concentrations of
 4    BDE-209 and related substances in environmental media (i.e., air, water, soil).

 5    H.4.1.2.1. Outdoor Air
 6           Research has shown that BDE-209 comprises between 6 and 31% of total PBDE concentrations
 7    in outdoor air (Trederiksen et al.. 2009). Outdoor air generally has lower concentrations of total PBDEs,
 8    which can be one or two orders of magnitude lower than in indoor air (see Appendix E). This lower
 9    concentration is partly due to the dilution factor of outdoor environments, but also occurs because indoor
10    environments contain more sources of PBDEs such as electronics, furniture, plastics, and coatings in
11    enclosed spaces (Paso et al., 2010).
12           Mean levels of BDE-209 measured in outdoor air from locations throughout the United States
13    range from 1.4 to 60.1 pg/m3 (U.S. EPA, 2010a). and BDE-209 has been detected in the particulate phase
14    in air near point sources (ATSDR, 2004). Sampling locations included rural, agricultural, and urban
15    locations. The highest level detected was 65 pg/m3 in urban Chicago, and the lowest level was 0.2 pg/m3
16    at an agricultural site in Indiana (Hohet al.. 2005). International studies have reported mean BDE-209
17    levels ranging from 1.6 to 53.3 pg/m3 (Chang et al., 2009). The lowest environmental concentration
18    reported in the literature was 0.091 pg/m3 in the Canadian High Arctic (Su et al., 2007). and the highest
19    level reported was  105 pg/m3 in Southern Ontario, Canada, nearly all of which was sorbed to aerosol
20    particles (Gouin et al.. 2006). Information from additional studies of concentrations in air is available in
21    Appendix E. Table E-2.

22    H.4.1.2.2. Aquatic Systems - Sediment and Surface Water
23           PBDEs are hydrophobic and therefore are not detected in large concentrations in the water
24    column (ATSDR. 2004). In aquatic systems (including both water and sediment), BDE-209 is the
25    predominant PBDE congener detected (49% to nearly 97% of the total PBDEs), with most detected in
26    sediment (ATSDR. 2004). BDE-209 has been observed in both sediments and surface waters in multiple
27    locations in the United States, from California to Delaware and Wisconsin to Mississippi (Ashley et al..
28    2006; Raff and Hites. 2004). Surface water in the San Francisco Bay estuary of California was found to
29    contain BDE-209 at concentrations below the limit of detection (reported as 20-200 pg/L for individual
30    congeners) to 191.0 pg/L water, with all but one of the detectable concentrations ranging between 12.2
31    and 87.8 pg/L (Pros et al.. 2005). Levels of BDE-209 observed in sediment ranged from below the level

      March 2013                                    H-32                  DRAFT - DO NOT CITE OR QUOTE

-------
 1    of detection (reported as 0.1 to 1.5 ng/gram dry weight) to 3,150,000 ng/gram in sediment downstream of
 2    a waste water treatment plant for a plastics manufacturer in North Carolina (La Guardia et al., 2007). In
 3    another study, the highest total PBDE concentration in suspended sediment from the Mississippi River
 4    and five tributaries was 1,548 ng/gram, with BDE-209 accounting for an average of 96.8% of the total
 5    observed PBDE concentration at the 31 sampling sites (Raff and Hites. 2004). Appendix E, Table E-3
 6    provides additional information on these studies.
 7           As detailed above, detected levels of BDE-209 in surface water are many times lower than those
 8    in sediment, which is consistent with assumptions based on the physical properties and environmental fate
 9    of BDE-209 discussed in Section H.3. The highest level of BDE-209 detected in surface water was
10    191 pg/L, equivalent to 191 parts per quadrillion, whereas the highest amount in sediment was
11    3,150,000 ng/gram, equivalent to 3,150 parts per million (ppm). These two measurements differ by a
12    factor of approximately 165 million.
13           Mean levels of BDE-209 in sewage sludge in the United States have varied widely, ranging from
14    84.8 to 58,800 ng/gram dry weight (La Guardia et al.. 2007: Hale etal.. 2001). In these studies, BDE-209
15    was the most frequent PBDE congener detected. International studies have shown levels ranging from
16    68.5 to 880 ng/gram dry weight (Clarke et al.. 2008; Wang et al.. 2007). A mean effluent BDE-209
17    concentration of 1,730 pg/L was reported from a sewage treatment plant in California (North. 2004).
18    Internationally, effluent concentrations have ranged from 310 to 1,170 ng/L (Eljarrat et al., 2007; de Boer
19    etal.. 2003). See Appendix E. Table E-4 for additional information on these studies.

20    H.4.1.2.3. Terrestrial Systems - Soil
21           Mean levels of BDE-209 observed in U.S. surface soils in two studies ranged from 0.6 to
22    15.3 ng/gram dry weight  [(Offenberg et al.. 2006) as cited in U.S. EPA (2010a): (Yun etal.. 2008)1.
23    The range of concentrations in international studies was 0.028-2,220 ng/gram dry weight (Sellstrom et
24    al.. 2005). Both the lowest and highest levels of BDE-209 detected were from a study in Sweden.
25    The only other international studies examined BDE-209 levels in Chinese soil; these measurements
26    ranged from 2.38 to 6,319.6 ng/gram dry weight (Luo et al.. 2009; Zou etal. 2007). Appendix E. Table
27    E-5 presents additional information on these studies.

28    H.4.1.2.4. Occupational Settings - Air
29           No studies were found that measured levels of BDE-209 in air in facilities where textiles
30    containing BDE-209 are manufactured. Outdoor air at an automobile shredding facility, however, was
31    found to have BDE-209 levels ranging from 45.5 to 1,940 pg/m3 (Charles etal.. 2005). Given that
32    BDE-209 is used in automobile upholstery, the textiles in the automobiles could have contributed to the

      March 2013                                    H-33                 DRAFT - DO NOT CITE OR QUOTE

-------
 1    observed levels. BDE-209 levels have been measured in indoor occupational settings, including
 2    e-recycling facilities, circuit-board assembly halls, and computer facilities. Reported levels in the air
 3    ranged from a median of 220 pg/m3 in a circuit-board assembly hall to a high of 833,000 pg/m3 in an
 4    e-recycling facility (Frederiksen et al., 2009; Charles et al., 2005). See Appendix E. Table E-2 for a
 5    summary of additional studies reporting indoor and outdoor air monitoring results.

 6    H.4.1.2.5. Residential Settings - Air and Dust
 7           Two U.S.-based studies evaluated residential indoor air levels of BDE-209. Levels of BDE-209 in
 8    the air ranged from below the limit of detection (limit of detection not reported) to 94 pg/m3 for the living
 9    room and 173.6 pg/m3 for personal air (i.e., breathing zone) (Allen et al., 2007). A study in Sweden
10    detected BDE-209 at 257 pg/m3 in the living room air in one of five household samples, while all other
11    samples in the study were below the limit of detection (173 pg/m3) (Petersen and Henry. 2012). Appendix
12    E, Table E-2 provides a summary of additional indoor air studies.
13           Levels of BDE-209 in household dust are orders of magnitude higher than in other matrices by
14    weight, and BDE-209 is the main PBDE contaminant in household dust (Paso et al.. 2010). BDE-209 has
15    been found in household dust in U.S. studies at median levels ranging from 665 to 2,000 ng/gram dry
16    weight (Frederiksen et al., 2009). In international studies, median levels of BDE-209 in house dust ranged
17    from 60 ng/gram dry weight (Germany) to  7,100 ng/gram dry weight (United Kingdom) (Frederiksen et
18    al.. 2009). See Appendix E. Table E-l for additional studies that report observed levels of BDE-209 in
19    household dust.

20    H.4.1.2.6. Nonresidential Settings - Air and Dust
21           BDE-209 is  one of the main PBDE contaminants in office dust (Watkins et al.. 2011; Batterman
22    etal., 2010; Harrad et al., 2008a). One study of U.S. office buildings found a mean concentration of
23    6,930 ng/gram BDE-209 in office dust,  but half of the sample sites had concentrations of 1 ng/gram or
24    concentrations below the limit of detection (limit of detection not reported) (Batterman et al., 2010).
25    Another study of dust in U.S. offices reported an average BDE-209 concentration of 4,204 ng/gram
26    (geometric  mean) with a range of concentrations between 912 and  106,204 ng/gram among sample sites
27    (Watkins et al., 2011). In a study conducted in the United Kingdom, Harrad et al. (2008a) reported a
28    median concentration of 6,200 ng/gram BDE-209 in office dust. See Appendix E. Table E-l  for study
29    summaries  that report observed levels of BDE-209 in office dust.
30           One study conducted in U.S.  office buildings examined BDE-209 concentrations in airborne
31    particulate matter and vapor; concentrations were all below the limit of detection, which was not reported
32    (Batterman et al.. 2010). Appendix E. Table E-2 provides a summary of this study.

      March 2013                                     H-34                   DRAFT - DO NOT CITE OR QUOTE

-------
 1    H.4.1.2.7. Transportation, Including Automobiles and Airplanes—Air and Dust
 2           As discussed in Section H.4.2.2.3 below, flame-retardant upholstery can be used in seats and
 3    other textiles in transportation vehicles such as automobiles and airplanes. No studies were identified that
 4    reported levels of BDE-209 in automobile or aircraft air, but several studies have investigated levels of
 5    BDE-209 in automobile or aircraft dust (Lagalante et al.. 2009; Christiansson et al.. 2008; Harrad et al..
 6    2008a). Less time is generally spent in automobiles than indoors, but levels of BDE-209 in automobile
 7    dust are about 20 times higher than in household dust (Lagalante et al.. 2009). In one study, the median
 8    level of BDE-209 in passenger cars was estimated as 8.12 ug/gram dust (82% of the total PBDE
 9    concentration in dust); personal automobiles generally had lower levels of decaBDE in dust than dealer
10    vehicles (Lagalante et al.. 2009). Other studies have reported median BDE-209 levels in dust from cars as
11    high as 100 ug/gram, with a highest individual sample of 2,600 ug/gram (Harrad et al., 2008a) (see
12    Appendix E. Table E-l for more information). Levels of PBDE congeners in the vehicles were not
13    statistically significantly different by vehicle manufacturer, model year, country of manufacture,  seat
14    type, or the presence of heated seats (Lagalante et al., 2009).
15           A study evaluating dust in aircraft during 20 international flights observed BDE-209
16    concentrations  ranging from below the limit of detection (value not reported) to 189,882 ng/gram, with a
17    median level of 17,262 ng/gram (Christiansson et al., 2008). No other studies were found that evaluated
18    levels of BDE-209 in air or dust of aircraft.


      H.4.2. Human Exposure and Kinetics  Leading to Dose
19           Limited data were found that measured or quantified human exposure to BDE-209. Data on
20    concentrations  of BDE-209 measured in media such as air, soil, or dust in various settings (described in
21    Section H.4.1.2). however, can be used in conjunction with activity pattern and other exposure factor data
22    [such as those described in The Exposure Factors Handbook (U.S. EPA. 2011)1 to inform estimates of
23    potential exposure through the various exposure pathways and scenario characteristics described in this
24    section.
25           The types of human exposure scenarios described here can be divided into four broad groups:
26    occupational, consumer, general public, and highly exposed populations. For the purposes of this case
27    study, occupational exposures include occupational exposures during synthesis, processing, or handling
28    of decaBDE; manufacturing of flame retardants, application of the flame retardants to textiles, or textile
29    finishing and upholstering; storage of the decaBDE, flame-retardant formulations, treated textiles, or
30    upholstered products; disposal of decaBDE, flame-retardant formulations, treated textiles, or upholstered
31    products; and repurposing or recycling of treated upholstery textiles and end-user products (e.g.,

      March 2013                                     H-35                  DRAFT - DO NOT CITE OR QUOTE

-------
 1    furniture). Consumer exposure scenarios include the intended or unavoidable use of treated upholstery
 2    textiles in residential and nonresidential spaces, including on household or institutional/office furniture, in
 3    vehicles, and in aircraft; unintended uses of treated upholstery textiles or end-use products such as reuse
 4    or repurposing of furniture for something other than its original intended use; or recycling of upholstery
 5    textiles for new uses. General public exposure includes primary exposure to members of the community
 6    near manufacturing, disposal, or recycling facilities and secondary exposure to the general public through
 7    environmental routes such as air, soil, or water. Highly exposed refers to exposure scenarios that are
 8    expected to occur via similar pathways as outlined for consumers  and the general public, but where
 9    exposure levels are expected to be higher due to key differences in population characteristics such as
10    those described in The Child-Specific Exposure Factors Handbook (U.S. EPA. 2008a).

      H.4.2.1. Occupational Exposure Pathway Scenarios
11           Limited data were found to determine the extent of occupational exposures to BDE-209 during
12    the material synthesis, processing, and handling phases or to the flame-retardant product during
13    formulation, application, storage, and disposal phases. See Section H.4.1.2.4 for BDE-209 concentrations
14    measured in occupational settings, which could be applied with the exposure pathways and scenario
15    characteristics described below to estimate potential exposures through scenario evaluation.

16    H.4.2.1.1. Synthesis, Processing, and Handling
17           As discussed in Section H.2.2.2.  BDE-209 synthesis involves conversion of phenol to diphenyl
18    ether, followed by bromination in the presence of a catalyst, typically aluminum bromide or iron (WHO.
19    1994). The synthesis and drying processes are carried out in enclosed vessels, so under normal
20    circumstances exposure is unlikely to occur during this process. After synthesis, decaBDE powders are
21    removed from the chamber and bagged. As discussed in Section H.2.2.2.2). the low vapor pressure of
22    decaBDE results in negligible exposures  to decaBDE as a vapor during synthesis or bagging, but
23    exposures to decaBDE adsorbed to dust could occur (EU. 2002). DecaBDE adsorbed to dust is expected
24    to settle quickly on surfaces in the occupational environment; no data were found to determine the extent
25    of exposures to decaBDE and decaBDE adsorbed to dust during the synthesis, processing,  and handling
26    phases. The pathways through which workers might be exposed to decaBDE and decaBDE adsorbed to
27    dust during general synthesis, processing, and handling scenarios  are described below:
28           •   Inhalation. Bagging and other handling of decaBDE  powders might be the activities most
29               likely to lead to exposures. Aerosol particles in the inhalable size range could be inhaled by
3 0               workers if respirators are not worn.
      March 2013                                    H-36                  DRAFT - DO NOT CITE OR QUOTE

-------
 1           •   Oral. Secondary oral exposures might occur if inhaled decaBDE or decaBDE that deposits
 2               on the skin, food, or food-contact surfaces are subsequently ingested.
 3           •   Dermal. DecaBDE might settle on the skin if proper personal protective equipment is not
 4               worn.
 5    H.4.2.1.2. Formulation of Flame Retardant, Application to Textiles, Upholstering
 6           As discussed in Section H.2.2.4, decaBDE powder is mixed with other ingredients to create a
 7    paint-like flame-retardant product, after which the flame retardant is back-coated onto a textile intended
 8    for use as upholstery. No data were found on the extent of occupational exposures to decaBDE during
 9    formulation of the flame retardant, application of the flame retardant to textiles, or textile finishing and
10    upholstering. As with exposures during the previous life-cycle stages, exposures to decaBDE vapors
11    during the product manufacturing stages are expected to be low due to the low vapor pressure of this
12    congener; however, exposures to decaBDE adsorbed to dust or attached to the product matrix could
13    occur.
14           The first step in the product manufacturing chain is compounding (i.e., mixing) decaBDE powder
15    with antimony trioxide in water, which typically occurs under local exhaust ventilation (EU. 2002).
16    The mixture is then added, through a closed system, to the emulsion polymers in a sealed mixing vessel.
17    Use of ventilation controls and a closed system in generating the flame-retardant coating is expected to
18    greatly reduce the chance of occupational exposures during these stages under normal circumstances.
19    Exposures might still occur to decaBDE adsorbed to dust, however, when the decaBDE powder is
20    emptied into the mixer. Exposures also could occur during transfer of materials; equipment cleaning,
21    maintenance,  and repair; and as the result of accidental spills or releases (EU. 2002).
22           The potential for occupational exposure also exists when the flame-retardant polymer mixture is
23    applied to the textile as a resin back-coating. DecaBDE flame retardant is typically not added manually,
24    and exposures are expected to be greatest during handling and cleaning of coating equipment (EU. 2002).
25    but accidental spills and releases also might occur.  Occupational exposures to decaBDE adsorbed to dust,
26    in the polymer matrix, or attached to textile fibers or scraps might occur as a result of cutting, sewing, and
27    otherwise abrading the decaBDE-treated upholstery textile product during textile finishing and
28    application to a consumer end-use product.
29           The pathways through which workers might be exposed to decaBDE during general formulation
30    of the flame retardant, application of the flame retardant to the textile, and textile finishing and
31    upholstering scenarios are expected to be comparable to those described in Section H.4.2.1.1 on
32    exposures during synthesis, processing, and handling. Additional considerations pertaining to exposures
33    to decaBDE in combination with polymer ingredients, textile fibers or scraps, or other product
34    constituents during these scenarios  are described below:

      March 2013                                      H-37                  DRAFT - DO NOT CITE OR QUOTE

-------
 1           •   Inhalation. Abrading textiles during tailoring and upholstering could lead to inhalation of
 2               decaBDE, other product ingredients, and textile dusts.
 3           •   Oral. Secondary oral exposures might occur if inhaled decaBDE and associated product
 4               constituents or decaBDE particles that deposit on the skin, food, or food-contact surfaces are
 5               subsequently ingested.
 6           •   Dermal. DecaBDE and associated product constituents generated during product
 7               manufacturing can land on the skin of workers if proper personal protective equipment is not
 8               worn. The liquid flame-retardant coating also can be spilled directly onto the skin.
 9    H.4.2.1.3. Storage of DecaBDE, Flame-Retardant Formulations, Treated Textiles, and Upholstered
10    Products
11           As described in Section G.2.2, decaBDE and the flame-retardant formulations to which it is
12    added are expected to be stored in sealed receptacles that would limit potential for worker exposures to
13    these materials during  storage. Defective packaging and accidental spills or releases, however, could lead
14    to rare exposures during storage operations.
15           Although no information was identified regarding procedures for storing treated upholstery
16    textiles, these products are likely packaged to protect them from exposure to elements like water and light
17    that could damage their aesthetics. Such packaging also is expected to limit exposures of workers to the
18    flame-retardant coatings. Once the textiles have been applied as upholstery to end-use products, these
19    products also are expected to be enclosed in protective packaging. Some surfaces of bulkier products
20    (e.g., furniture), however, might remain uncovered, which could lead to worker exposures during storage
21    operations, or exposures might occur during application and removal of packaging materials to and from
22    the product. Dust also  can accumulate in storage facilities that frequently store textiles and textile
23    products, and decaBDE that escapes from the product matrix could sorb to dust particles. Ventilation
24    technologies and other contamination-prevention strategies like those manufacturing facilities use are not
25    expected to be in place in storage facilities.  Dust that has settled on surfaces in storage facilities can be
26    disturbed by worker operations, resuspended, and transported to other locations.
27           Although decaBDE is not expected to be highly volatile, off gassing of more volatile components
28    of the treated textiles might occur during storage of treated textiles or upholstered products. Furthermore,
29    due to the additive nature of decaBDE flame retardants, covalent bonding between the flame retardant and
30    the textile does not occur, suggesting that flame-retardant coatings that are loosely attached to the textile
31    surface might slough off during storage or handling. Because decaBDE flame retardants are generally
32    added to the back of the textile, however, the likelihood of this detachment seems low.
33           No data were found on occupational exposures to decaBDE during storage throughout the
34    product life cycle of flame-retardant upholstery textile coating. A study examining residential exposures
35    to a range of PBDEs, however, did identify  age of furniture as one of the drivers of exposure, with higher
      March 2013                                     H-38                  DRAFT - DO NOT CITE OR QUOTE

-------
 1    BDE-209 body burdens in children aged 2-5 years correlating with newer furniture (e.g., couches,

 2    mattresses) (Rose etal.. 2010). Whether furniture is currently treated more often with decaBDE flame

 3    retardants than previously or whether the age of the furniture influences the rate of decaBDE release (with

 4    greater amounts released from newer furniture), however, is unclear, which limits the applicability of this

 5    finding to the refinement of realistic exposure scenarios.

 6           The pathways through which workers might be exposed to decaBDE alone or decaBDE adsorbed

 7    to dust during storage of decaBDE and decaBDE flame-retardant formulations are expected to be

 8    comparable to those described in Sections H.4.2.1.1 and H.4.2.1.2 on exposures during synthesis,
 9    processing, and handling and during formulation of the flame retardant, application to textiles, and

10    upholstering. Additional considerations pertaining to exposures to decaBDE alone or in combination with

11    polymer ingredients, textile fibers or scraps, or  other product constituents during general treated textile of

12    upholstered product storage scenarios are described below:

13           •   Inhalation. Workers could inhale volatile components of the flame-retardant coating or
14               decaBDE adsorbed to dust in storage facilities, particularly facilities that are not well
15               ventilated. Furthermore, decaBDE  adsorbed to dust could be resuspended in the air by worker
16               activities, and subsequently inhaled. PBDE exposures have not been measured at textile
17               storage facilities, but they have been measured at electronic waste storage facilities.
18               The median estimated inhalation exposure to BDE-99 (the highest measured PBDE in air) in
19               male workers was 0.0011 ng/kg body weight (bw) per day (Muenhor et al., 2010). Although
20               BDE-209 exposures are likely to differ due to lower volatility and greater propensity to
21               adsorb to particles,  debromination of BDE-209 could result in worker exposures to lower
22               brominated congeners.

23           •   Oral. Higher levels of dust in textile storage facilities could lead to increased transport of
24               decaBDE adsorbed to dust. This could result in oral exposures to decaBDE in dust
25               transported to break rooms, homes  (via clothes), and other locations where decaBDE
26               adsorbed to dust can be unintentionally ingested while eating or due to hand-to-mouth
27               activity. In electronic  waste storage facilities, the median exposure to BDE-209 via dust
28               ingestion was 2.89 ng/kg-bw-day for average ingestion scenarios and 7.2 ng/kg-bw-day  for
29               high-end ingestion scenarios (Muenhor et al.. 2010).

30           •   Dermal. DecaBDE adsorbed to dust could be resuspended by worker activities and deposit
31               on the skin of workers if proper personal protective equipment is not worn.

32    H.4.2.1.4. Disposal and Recycling of DecaBDE, MWCNTs,  Flame-retardant Formulations, Treated
33    Textiles, and Upholstered Products

34           As described in Section H.2.5. large-scale disposal, recycling, and reuse of decaBDE, and the

35    flame-retardant formulations to  which it is added are unlikely, but containers used to store these products

36    might enter the waste stream, and  workers  at disposal and recycling facilities could be exposed to product

37    residues remaining in these containers.

38           Disposal and recycling of treated textiles and upholstered products, however, is prevalent. Mixing
39    and compacting of waste for land-filling; cleaning, shredding, blending, melting, and spinning scrap


      March 2013                                      H-39                   DRAFT - DO NOT CITE OR QUOTE

-------
 1    textiles for recycling; and incomplete incineration of treated upholstery textiles all could result in
 2    exposure of workers (Chaudhry et al.. 2009) to decaBDE, primarily in combination with other product
 3    constituents and dusts.
 4           No data were found on the extent of occupational exposures to decaBDE during disposal,
 5    recycling, and reuse throughout the decaBDE flame-retardant upholstery textile coating product life cycle.
 6    As discussed in Section H.4.1.2.4. BDE-209 was detected in the air outside of an automobile shredding
 7    facility (Charles et al.. 2005). which suggests that exposure to BDE-209 during end-of-life operations can
 8    occur.
 9           The pathways through which workers might be exposed to decaBDE during general disposal and
10    recycling of decaBDE powder and flame-retardant formulations are expected to be comparable to those
11    described in Section H.4.2.1.1 (exposures during synthesis, processing, and handling), and worker
12    exposure pathways for decaBDE in combination with polymer ingredients, textile fibers or scraps, or
13    other product constituents during disposal and recycling treated textiles and upholstered product are
14    expected to be similar to those described in Sections H.4.2.1.2 (exposures during formulation of the flame
15    retardant, application to textiles, and upholstering) and H.4.2.1.3  (exposure during storage and
16    distribution) for these products. Additional considerations pertaining to exposures to decaBDE alone or in
17    combination with polymer ingredients, textile fibers or scraps, or other product constituents during treated
18    textile or upholstered product disposal and recycling scenarios are described below:
19           •   Inhalation. Workers operating machines  that abrade or destroy textile materials, those
20               handling these products, and other workers in the vicinity of operations that agitate or abrade
21               textile materials can inhale decaBDE adsorbed to dust and other product constituents, as
22               observed by Sjodin et al. (2001) at an e-waste recycling  plant where electronics are stored,
23               dismantled, and shredded. Workers at incineration facilities also might inhale small particles
24               comprising decaBDE and other substances in the  incinerator as well as polybrominated
25               dibenzofurans (PBDFs) and polybrominated dibenzo-p-dioxins (PBDDs) (see Section
26               H.2.5.2.2) if treated textiles and upholstered products are not incinerated at sufficiently high
27               temperatures.
28           •   Oral. No additional considerations.
29           •   Dermal. Workers at disposal and recycling facilities might come into physical contact with
30               the decaBDE flame-retardant coating on an upholstery textile during the process of moving or
31               handling products. DecaBDE could migrate directly to skin of workers if proper personal
32               protective equipment is not worn.

      H.4.2.2. Consumer Exposure Pathway Scenarios
33           BDE-209 is expected to be released from consumer products in the particulate phase. See
34    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
35    residential, nonresidential, and general public settings, which  could be applied with the exposure
      March 2013                                     H-40                   DRAFT - DO NOT CITE OR QUOTE

-------
 1    pathways and scenario characteristics described below to estimate potential exposures to BDE-209
 2    through a scenario evaluation approach.

 3    H.4.2.2.1. Intended Use - Upholstered Products in Residential Spaces
 4           As discussed in Section H.4.1.2.5. PBDE concentration in house dust tends to be higher than in
 5    other matrices, and BDE-209 is the dominant congener in house dust. Although flame-retardant
 6    upholstery textiles typically are used in nonresidential settings (see Section H.4.2.2.2). some residential
 7    upholstered products, particularly mattresses, are known to contain decaBDE, and other upholstered
 8    furniture products, like couches, sometimes might be treated with decaBDE (Rose et al.. 2010). One study
 9    has shown that body burdens of BDE-209 in children are positively associated with presence of new
10    furniture, but are not associated with presence or use of electronics that often contain  decaBDE (Rose et
11    al.. 2010). This finding suggests that decaBDE use in residential upholstery does contribute to overall
12    decaBDE exposures related to the use of decaBDE in flame-retardant upholstery textiles. Higher body
13    burdens of BDE-209 also have been associated with smaller living spaces, and higher concentrations of
14    decaBDE have been measured in the main living area of the house than in the bedroom (Allen et al..
15    2008). indicating that variations in decaBDE exposures can be expected due to variations in housing
16    characteristics and human behavior patterns (i.e., time spent by individuals in different rooms or outside
17    the house). Furthermore, decaBDE released from products is suspected to debrominate to some degree to
18    lower brominated congeners in residential settings (Allen et al.. 2008). and will therefore lead to
19    exposures to PBDEs other than decaBDE.
20           As introduced in Section H.2.4. upholstered products are expected to be used for many years, and
21    contact with the textile might be frequent and prolonged, which could cause substantial wear and tear on
22    the textile product. In addition, upholstery in residential spaces might frequently be exposed to cleaning
23    products, sweat, food, and other substances that could affect the properties of the textile and the flame-
24    retardant coating.
25           The pathways through which consumers might be  exposed in residential settings to decaBDE
26    during general consumer use scenarios for end products upholstered with decaBDE are described below:
27           •   Inhalation. Chronic inhalation of particles of decaBDE in combination with other product
28               constituents and dust could occur following release from upholstered products over time (due
29               to wear and tear from anticipated use, aging of materials, abrasion, UV light, water, cleaning
30               chemicals, among other factors; see Section H.2.4.2). Particulate decaBDE could settle onto
31               surfaces, where it might be disturbed and re-entrained, after which it could be inhaled by
32               residents. Inhalation is not expected to be a primary route of exposure for decaBDE (Johnson-
33               Restrepo and Kannan. 2009; Allen et al.. 2008). however, because the contribution of inhaled
34               dust particles is expected to be minimal due to a lack of correlation between concentrations of
35               decaBDE in dust and in air (Allen et al..  2008): see below.
      March 2013                                     H-41                  DRAFT - DO NOT CITE OR QUOTE

-------
 1           •   Oral. DecaBDE in combination with other product constituents and dust could be ingested
 2               after settling on food and food-contact surfaces or following hand-to-mouth activity.
 3               Ingestion of household dusts is hypothesized to be  a major exposure pathway for PBDEs
 4               (Allen et al.. 2008). Johnson-Restrepo and Kannan (2009) reported that most PBDE intake in
 5               toddlers, children, teenagers, and adults was attributed to the oral route of exposure via
 6               ingestion of household dust (56-77% attributed to combined oral and dermal exposure).

 7           •   Dermal. Dermal exposure to decaBDE in combination with other product constituents and
 8               dust might occur while touching the textile surface (particularly if the portion of the textile
 9               that has been treated with the flame-retardant coating is exposed) or touching surfaces upon
10               which particles have settled (Frederiksen et al., 2009). Lorber (2008) estimated that dermal
11               exposure to PBDE compounds in household dust could be a significant contributor (estimated
12               at 16%) to the body burden of PBDEs in adults. The estimated contribution of BDE-209 to
13               body burden from dermal contact exposure was 25.2 ng per day of a total 85.9 ng total
14               PBDEs per day from that exposure route (Lorber. 2008). Johnson-Restrepo and Kannan
15               (2009) similarly reported that the dermal route was a primary route of exposure for PBDEs
16               (second to the oral route of exposure for contribution of human intake).

17    H.4.2.2.2. Intended  Use - Upholstered Products in Nonresidential Spaces

18           Due to regulations requiring that upholstery textiles used in nonresidential settings pass flame-

19    retardancy tests (see Table H-3). many upholstery textiles in public, commercial, and institutional settings

20    are treated with decaBDE. The characteristics of the different settings in which these products are used

21    can vary considerably. For example, flame-retardant upholstery textiles  might be used in seating for

22    airports and other transportation hubs and in waiting rooms, office buildings, penal institutions, and other

23    nonresidential spaces that can range from very small to very large and where consumers might spend

24    varying amounts of time. Some scenarios for nonresidential exposures are not likely to differ from those

25    expected from residential exposures, but a  few key differences  do exist:

26           •   Exposures to flame-retardant upholstery coatings in public spaces might be unavoidable.
27               Although consumers have some control over which products they bring into their home, they
28               have little control over the products they encounter in public spaces.

29           •   Some nonresidential exposures might occur over long periods of time and for extended
30               intervals (e.g., silting in the same office chair every day over the course of several work
31               years), while some might occur infrequently and for short periods of time (e.g., sitting in
32               seating at the airport waiting for a flight).

33           •   Products in public spaces might experience higher  activity levels, more frequent cleaning,
34               and less care to the textile surface, all of which could damage  or weaken the textile matrix
35               and influence releases and exposures.

36           With the exception of these potential differences in exposure settings and activity patterns, the

37    pathways and scenarios through which consumers might be exposed in nonresidential settings to

38    decaBDE during general consumer use  scenarios for end products upholstered with decaBDE flame-

39    retardant coatings are not expected to differ from those described previously in Section H.4.2.2.1 on

40    exposures from intended use of upholstered products in residential spaces.


      March 2013                                     H-42                   DRAFT - DO NOT CITE OR QUOTE

-------
 1    H.4.2.2.3. Intended Use - Aircraft and Automobile Upholstery
 2           Flame-retardant upholstery can be used for seating, draperies, carpets, and other textiles in
 3    passenger cars and public and private transportation. Aircraft and automobile passengers, and those
 4    working in these environments (e.g., cab drivers, flight attendants), could be exposed to higher levels of
 5    BDE-209 due to the higher concentrations of PBDEs in dust in those environments as compared to home
 6    environments (Lagalante et al.. 2011; Christiansson et al.. 2008). Dust generated by abrasion of treated
 7    upholstery fabric is the most likely pathway for BDE-209 exposure in automobiles (Lagalante et al..
 8    2011). No association was found, however, between time spent in automobiles and plasma PBDE levels
 9    in children aged 2-5 years in California (Rose etal.. 2010); these children spent an average 7.2 hours per
10    week (range 0-20 hours) in the car.
11           Photodegradation of BDE-209 is low in cars because automobile glass blocks UVB radiation,
12    which is the region of the spectrum most strongly absorbed by BDE-209. BDE-209 adsorbed to sodium
13    sulfate does photodegrade in automobiles, however, and has a half-life of approximately 19 days.
14    The congeners BDE-47 and BDE-99 are environmentally and lexicologically relevant products of BDE-
15    209 debromination (see Text Box H.3-1). but their presence in automobile dust is mainly from
16    volatilization and weathering of products containing pentaBDE, rather than photodegradation of BDE-209
17    and other higher brominated congeners (Lagalante et al., 2011).
18           The pathways through which consumers might be exposed in vehicles (including airplanes) to
19    decaBDE during general consumer use scenarios for end products upholstered with decaBDE flame-
20    retardant coatings are described below:
21           •   Inhalation. Inhalation of decaBDE adsorbed to dust from worn or abraded automobile
22               upholstery is expected to occur. Inhalation exposure to photodegradation products of
23               decaBDE could  occur in automobiles if automobile textiles contain decaBDE adsorbed to
24               sodium sulfate. The recirculation of air in aircraft cabins also might affect exposure to
25               particulate decaBDE if filters do not adequately remove these particles.
26           •   Oral. Secondary oral exposures might occur if inhaled particulate decaBDE or particulate
27               decaBDE that deposits on the skin is subsequently ingested.
28           •   Dermal. Dermal exposures to decaBDE or decaBDE photodegradates (due to worn or
29               abraded automobile upholstery) are expected to occur, particularly when skin touches the
30               treated part of the textile directly. Dermal exposure also can occur when particles in the air
31               settle on the skin. Different exposure characteristics or scenarios (e.g., children silting in
32               safety seats) might influence whether dermal exposure occurs, or influence the extent to
33               which exposure  occurs through this pathway.
34    H.4.2.2.4. Unintended Use,  Repurposing, or Reuse of Treated Textiles and Upholstered Products
35           As introduced in Section H.2.4. unintended uses of upholstery textiles treated with decaBDE
36    flame-retardant coatings could include repurposing of treated upholstery textiles for clothing, building
      March 2013                                    H-43                  DRAFT - DO NOT CITE OR QUOTE

-------
 1    insulation, other in-home or outdoor furnishings, bedding, or other purposes. The repurposing stages
 2    could introduce occupational exposures similar to those discussed in Section H.4.2.1.2 (exposures during
 3    formulation of the flame retardant, application to textiles, and upholstering) and Section H.4.2.1.3
 4    (exposure during storage and distribution), as products that are treated with flame-retardant coatings are
 5    broken down and reprocessed into new products.
 6           Although no information was identified that directly addresses potential consumer exposures
 7    following unintended use or reuse of flame-retardant upholstery textiles, exposure pathways and scenarios
 8    from other life-cycle stages are relevant here. Most reuse scenarios might differ little from those for
 9    anticipated consumer uses, but a few key differences might occur, particularly when products are
10    repurposed for new uses or used in unintended ways:
11           •   Processes similar to those involved with product manufacture (e.g., cutting, sewing) and
12               storage of textiles also might be employed for repurposing treated textiles. In this scenario,
13               however, these processes are not expected to occur in an occupational setting, but in the home
14               or another private space, where no personal protective equipment is worn and limited control
15               technologies are used. These processes, as employed for repurposing textiles, however, are
16               not expected to occur as commonly or at the same scale as in a manufacturing facility.
17           •   Older, more degraded textiles with weakened matrices might be handled directly and be
18               subjected to abrasion, thereby releasing the product constituents in the vicinity of the
19               consumer conducting the repurposing.
20           •   Although dermal contact with products used for their intended purpose (e.g., furniture
21               seating) might be limited by a clothing barrier between the consumer and the treated textile,
22               should flame-retardant upholstery textiles be repurposed into clothing, direct dermal contact
23               might occur repeatedly over long periods of time.
24           With the exception of these potential differences in exposure characteristics, the pathways and
25    scenarios through which consumers might be exposed to decaBDE during repurposing, reuse, or
26    unintended use of treated textiles and upholstered products are not expected to differ from exposure
27    pathways associated with the cutting, tailoring, or other abrasive processes involved with product
28    manufacturing (Section H.4.2.1.2); storage of textile products (Section H.4.2.1.3); and consumer use in
29    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
30           No information was found on exposure to decaBDE in the general public from environmental
31    media (e.g., air, water, soil). See Section H.4.1.2 for concentrations of BDE-209 in environmental media
32    that could be used with the exposure pathway and scenario characteristics below to estimate potential
33    exposures.
      March 2013                                     H-44                   DRAFT - DO NOT CITE OR QUOTE

-------
 1    H.4.2.3.1. Outdoor Air
 2           Releases of decaBDE to outdoor air throughout the product life cycle of the flame -retardant
 3    textile coatings are possible (see Section H.2). Once released to air, decaBDE can sorb to particulate
 4    matter and experience long-range transport to areas distant from its source (see Section H.3.2). As
 5    summarized in Section H.4.1.2.1. decaBDE has been measured in outdoor air at concentrations much
 6    lower than those measured indoors. Nonetheless, general public exposures to decaBDE adsorbed to
 7    particulate matter in ambient air are expected to occur, with the primary route being inhalation.
 8           Other product constituents of flame-retardant textiles (e.g., pieces of the polymer matrix or the
 9    textile fabric) also can be released, and in the case of decaBDE, combustion by-products (e.g., PBDDs,
10    PBDFs) and lower brominated transformation products are expected to be present in air as a result of the
11    flame-retardant textile coating life cycle.

12    H.4.2.3.2. Water
13           Releases of decaBDE and other product constituents to wastewater and ambient water bodies
14    throughout the product life cycle of flame-retardant textile coatings are possible (see Section H.2). Once
15    released to water, decaBDE is expected to sorb to particulate matter in the water column or to sediments,
16    which might limit their mobility (see Section H.3.3). This behavior implies that decaBDE also primarily
17    will be removed to sludge during wastewater treatment.
18           As summarized in Section H.4.1.2.2. decaBDE has been measured  in surface waters at low
19    concentrations [below levels of detection to 191.0 pg/L (Pros et al.. 2005)]. As a result, general public
20    exposures to decaBDE and its transformation products in water are expected to occur, with the primary
21    routes being dermal (through bathing and swimming) and oral  (drinking and incidental ingestion during
22    bathing and swimming).

23    H.4.2.3.3. Soil
24           Releases to ambient air and water throughout the product life cycle of flame-retardant textile
25    coatings will result in deposition of particles  of decaBDE and other product constituents  (see Section H.2)
26    to soil. Once deposited, decaBDE is expected to sorb strongly to soil, which might limit mobility (see
27    Section H. 3.4).
28           As summarized in Section H.4.1.2.3. decaBDE has been measured  in surface soils and is
29    expected to be  present in sludge applied to agricultural soils. As a result, general public exposures to
30    decaBDE and its transformation products in soils are expected  to occur, with the primary routes being
31    dermal and oral (although in incidental amounts). Furthermore, decaBDE has been shown to translocate
      March 2013                                     H-45                   DRAFT - DO NOT CITE OR QUOTE

-------
 1    from soil to plant tissues, suggesting that decaBDE can enter the food web, and dietary oral exposures
 2    also might occur.

      H.4.2.4. Highly Exposed Populations
 3           This section discusses characteristics of individuals and populations that might result in increased
 4    exposure (relative to the general population) to decaBDE released during the life cycle of flame-retardant
 5    upholstery textile coating.
 6           The primary exposure pathway for decaBDE is likely to be ingestion of household dust, and dust
 7    levels in the home can vary by socioeconomic status or the type and condition of housing (see Section
 8    H.5.3.1). Disproportionate levels of exposure can occur in specific populations, including low-income and
 9    low-educational-attainment populations. Although race and ethnicity have not been shown to be
10    associated with specific physiological conditions that increase susceptibility to exposure, demographic
11    factors such as socioeconomic and educational status could cause some populations to bear a
12    disproportionate level of the exposure burden.
13           Children are likely to experience higher exposures than the general population. Data suggest that
14    breast-fed infants are potentially exposed to BDE-209 through their mother's milk [i.e., worldwide,
15    median detected levels of BDE-209 in breast milk range from 0.1 to 2.9 ng/gram liquid weight
16    (Frederiksen et al.. 2009); maximum concentration of 7 breast milk samples containing decaBDE in the
17    United States was 8.24 ng/gram lipid (ATSDR. 2004)1. Young children also take in more household dust
18    than adults, with estimates for children at 100-200 mg/day compared to 50 mg/day for adults (U.S. EPA,
19    2008a). Increased hand-to-mouth activity contributes to increased exposures in children. Occupation also
20    could increase exposure relative to the general population, primarily for workers involved in manufacture
21    of decaBDE or flame retardants containing decaBDE, or textile products treated with decaBDE.

      H.4.2.5. Exposure Reference Values and Recommendations
22           A variety of exposure standards, guidelines, or recommendations are developed by different
23    organizations with purview over specific portions of the population or situations during which exposure
24    might occur (e.g.,  occupational exposures, general population drinking water exposures). Available
25    information on these types of values for decaBDE is presented below. Section H.5.1.1 discusses how
26    some of these values inform quantitative toxicity assessments.
27           As of January 2011, no national-level environmental or occupational health standards had been
28    established for decaBDE (POD. 2011). EPA has derived a reference dose (RfD) for decaBDE, based on
29    developmental neurobehavioral effects, of 0.007  mg/kg-day (U.S. EPA. 2008b) (see Section H.5.1).


      March 2013                                    H-46                  DRAFT - DO NOT CITE OR QUOTE

-------
 1    An RfD is an estimate (taking into account uncertainty) of the daily exposure to the human population,
 2    including sensitive populations, that is "likely to be without an appreciable risk of deleterious effects
 3    during a lifetime" (U.S. EPA. 2008b). EPA also has derived a cancer slope factor for decaBDE based on
 4    neoplastic nodules or carcinomas (combined) in the  liver of treated male rats, of 7 x 1CT3 per mg/kg-day.
 5    A cancer slope factor is a plausible upper bound on the estimate of risk per mg/kg-day of oral exposure
 6    (U.S. EPA. 2008b). The Agency for Toxic Substances and Disease Registry has developed a minimal risk
 7    level (MRL) for decaBDE for intermediate duration (15-365 days) oral exposure  of 10 mg/kg-day
 8    (ATSDR. 2004) based on Hardy et al. (2002). MRL values are estimates of the daily exposure to a
 9    hazardous chemical that is likely to be without appreciable risks of non-cancer health effects over a
10    specific duration of exposure.29 MRLs are intended as screening levels, rather than clean-up or action
11    levels for any agency. Differences in exposure duration (chronic lifetime versus intermediate) and the key
12    study used to derive the estimate contributed to the several-orders-of-magnitude difference between the
13    MRL and RfD  for decaBDE.

      H.4.2.6. Toxicokinetics, Dose, and Body Burden
14           Toxicokinetics can be used to relate exposure and contact, such as those described in the
15    scenarios above, with uptake and dose. Specifically, toxicokinetics describes how a material  is absorbed,
16    distributed, metabolized, and excreted in an organism. An understanding of the relationship between each
17    of these concepts, which are often referred to as ADME, leads to an understanding of the concentration,
18    or dose, of material that can reach—and potentially accumulate in—different tissues of the body.

19    H.4.2.6.1. Absorption, Distribution, Metabolism, Excretion
20           This section contains information regarding the toxicokinetic behavior of decaBDE when
21    administered to mammals. Information regarding birds and fish is not presented in this  section because,
22    when extrapolating toxicokinetic data to humans, studies conducted with rodents  (rat or mouse) or
23    nonrodent mammals (dog or monkey) are generally used. Additionally, differences among species have
24    been noted in numerous studies of decaBDE, and the toxicokinetic behavior in response to decaBDE
25    differs among birds, fish, and mammals. For example, fish generally debrominate decaBDE to pentaBDE
26    congeners, while mammals debrominate decaBDE, to a lesser degree, to heptaBDE congeners. See
27    Section H.4.3 for toxicokinetic information relevant to ecological exposures.
28           Early toxicokinetic studies (el Dareer et al.,  1987; NTP. 1986; NorrisetaL 1975; Norris et al..
29    1973) were conducted on decaBDE shortly after it was developed as a flame retardant.  These studies
      29ATSDR - Minimal Risk Levels: http://www.atsdr.cdc.gov/mrls/index.asp.
      March 2013                                    H-47                  DRAFT - DO NOT CITE OR QUOTE

-------
 1    demonstrated that decaBDE is poorly absorbed [0.3-1.5%; (NTP. 1986)1 from the gastrointestinal tract in
 2    rats following oral exposure and eliminated in the feces as the parent congener without prior metabolism.
 3    Essentially no elimination occurred through the urine, and more than 99% of the dose was recovered in
 4    feces by 48 hours, indicating a lack of accumulation in tissues (Norris et al.. 1975). The half-life of
 5    decaBDE is relatively short; the serum half-life was reported as 15 days in a human study where workers
 6    were exposed to BDE-209 (U.S. EPA. 2010a).
 7           More recent studies, although in general agreement, have reported higher absorption rates that
 8    might be due to the solvent used to administer decaBDE. In general, BDE-209 is not expected to
 9    accumulate in terrestrial organisms. Many studies, however, have reported levels of BDE-209 in humans
10    (breast milk, serum, and umbilical cord blood), food items (dairy, eggs, infant formula), and biota (fish,
11    shellfish), indicating that some absorption and accumulation occur over time (Frederiksen et al.. 2009).
12    Given the high trophic levels of the organisms where accumulation has been observed (humans, predatory
13    fish, and piscivorous birds), biomagnification appears to occur in these receptors (Environment Canada.
14    2010; U.S. EPA. 2010a). Bioaccumulation and biomagnification are discussed further in Section G.4.2.1.
15           Because  of the low absorption of BDE-209, blood and tissue levels following acute (short-term)
16    oral exposures are typically low and represent a small fraction of the total dose. More than 66% of the
17    parent compound was excreted in the feces of rats following oral exposures to BDE-209 (Riu et al.. 2008;
18    Morck et al., 2003). The same experiments showed that the highest concentrations of BDE-209 were
19    found in plasma and blood-rich tissues such as liver, kidney, adrenal glands, ovaries, heart, and the
20    intestinal wall following a single oral exposure (Morck et al., 2003) or 4-day gavage exposure (Riu et al..
21    2008; Morck et al..  2003). In these acute studies, BDE-209 was not readily distributed to adipose tissue.
22           In contrast, evidence shows that BDE-209 can accumulate in adipose tissue following chronic
23    oral exposure. Studies by Norris et al. [1974;  1975, as cited in Hardy et al. (2009)1 exposed Sprague-
24    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
25    3, 6, or 12 months. FR-300-BA  comprised 77.4% BDE-209, 21.8% nonaBDE, and 0.8% octaBDE. After
26    6 months of treatment, bromine  concentrations (measured by neutron activation analysis) in adipose tissue
27    were higher in treated rats (~3 ug/gram) than in controls (~1 ug/gram), but after 12 months the bromine
28    levels in adipose tissue were similar to controls. Bromine did not accumulate in other tissues such as liver,
29    kidney, and serum. Norris et al.  [1974; 1975, as cited in (Hardy et al.. 2009)] also followed the
30    elimination of bromine from male Sprague-Dawley rats that were dosed with FR-300-BA in the diet for
31    90 days at 1.0 mg/kg-day and subsequently fed a control diet. After 10 days on a control diet,
32    concentrations of bromine in the liver were similar to controls, but concentrations in adipose tissue were
33    higher (~2.5 to 4 ug/gram) than  controls (~0-2 ug/gram). Another study showed a time- and dose-
34    dependent increase  of bromine levels in adipose tissue indicating accumulation following dietary

      March 2013                                     H-48                  DRAFT - DO NOT CITE OR QUOTE

-------
 1    exposure to decabromodiphenyl oxide (a synonym for decaBDE) at 0.01, 0.1, or 1 mg/kg-day, for up to 2
 2    years (Kociba et al.. 1994). Kociba et al. (1994) also reported that bromine content was not increased
 3    compared to controls in the kidney, muscle, or serum of rats in the same study.
 4           Absorbed decaBDE is metabolized in the liver and a minor fraction of the parent compound is
 5    metabolized to lower PBDE congeners, such as tetraBDE and pentaBDE (see Text BoxH.3-1). Morck et
 6    al. (2003) reported that decaBDE was the predominant substance detected in the liver metabolites, with
 7    trace levels of nonaBDE. Similarly, a minor fraction of decaBDE (less than 3%) was debrominated to
 8    lower BDE congeners in a feeding study that exposed male Sprague-Dawley rats to DE-83R (98.5%
 9    decaBDE) (Huwe and Smith. 2007).
10           Metabolism of decaBDE to lower brominated congeners by oxidative debromination is indicated
11    by some evidence in studies of rats and fish, but the mechanisms and location of metabolic processes are
12    not well characterized due to limited availability of toxicokinetic data (Hakk and Letcher. 2003).
13    Metabolism of decaBDE also differs among species. In lactating cows, Kierkegaard et al. (2007)
14    suggested that decaBDE debrominates to hepta-, octa-, and nonaBDEs. Octa- and nonaBDEs were found
15    in liver and kidney of rats fed 100 mg/kg-day BDE-209 for 3 months (Wang et al.. 2010a). In contrast, in
16    vitro studies of human hepatocytes have shown evidence of low or no metabolism, possibly because of
17    low entry of BDE-209 into cells under the experimental conditions used (Stapleton et al.. 2009).
18           Based on available studies (Huwe and  Smith. 2007; Kierkegaard et al., 2007;  Hakk and Letcher.
19    2003; Morck et al.. 2003; Sandholm et al.. 2003). the following pathways for debromination of decaBDE
20    can be deduced for mammals:
21               1.  Deiodinase enzymes can debrominate decaBDE to nona-, octa-, and heptaBDEs.
22               2.  Debrominated neutral metabolites can undergo hydroxylation to potentially form phenols
23               or catechols, possibly via an arene oxide, which could involve the action of cytochrome P450
24               enzymes.
25               a.  The formed hydroxylated BDEs can compete with thyroxine for binding to a thyroxine
26                  transport protein present in blood serum.
27               b.  The catechols then are methylated, potentially by the action of catechol-O-
28                  methyltransferase, to form guaiacols.
29               c.  The guaiacol metabolites further oxidize to highly reactive quinones,  which bind to
30                  cellular macromolecules.
31               d.  The reactive intermediates are subject to rapid conjugation via Phase  II metabolic
32                  processes, leading to water-soluble metabolites that are excreted via bile and feces, as
33                  observed in conventional and cannulated rats.
34    Two toxicokinetic studies of fetal rats were identified. In a study by Riu et al. (2008).  radiolabeled
35    14C-BDE-209 (99.8% pure, dissolved in peanut oil) was administered orally to pregnant rats on Gestation

      March 2013                                     H-49                 DRAFT - DO NOT CITE OR QUOTE

-------
 1    Days (GD) GD16-GD19 (2 mg/kg-day). The toxicokinetic results were similar to those noted previously
 2    in this section. Approximately 72% of the dose was found in the feces and the digestive tract contents,
 3    while 0.1% was excreted in the urine. The remainder of the dose was distributed in various tissues, with
 4    6.5% in the liver and 5.3% in the digestive tract contents. All other tissues contained less than 1% of the
 5    administered dose. The fetuses (sum for the whole litter) contained 0.43% of the dose. In a recent study
 6    by Cai et al. (2011). BDE-209 and its metabolites were detected in the placenta and milk, and eventually
 7    in the fetuses or neonates when BDE-209 (prepared in peanut oil) was administered to pregnant Sprague-
 8    Dawley rats from GD7 to PND4. In the same study, detectable amounts of nonaBDEs (BDE-206, 207,
 9    208) and octaBDEs (BDE-196, 197/204, 198/203) were observed in the dosed rats. The predominant
10    debrominated metabolites of BDE-209 detected in fetuses were nonaBDEs (BDE-208, 207, 206).
11    The level of BDE-206 in the fetal or pup bodies was significantly lower on GD21 and PND4 than on
12    GD7. The octaBDEs BDE-196, BDE-198, and BDE-203 were observed in fetuses and pups, but were
13    minor debromination metabolites of BDE-209.
14           No animal studies have been identified that evaluate decaBDE ADME upon inhalation and
15    dermal exposures.

16    H.4.2.6.2. Internal Dose and Body Burden
17           Levels  of decaBDE in human tissues have been reported in several occupational studies and in
18    studies of the general public. A study in workers at an  electronics dismantling plant in Sweden evaluated
19    levels of five PBDEs in serum samples  from plant workers. The mean concentration of BDE-209 in that
20    study was as high as 5 ng/gram lipid (Darnerud et al.. 2001). Total PBDE serum levels in hospital
21    cleaners, computer clerks, and electronics dismantlers  were 3, 4, and 26 ng/gram lipid, respectively
22    (Darnerud et al.. 2001). As discussed in the previous section, decaBDE can accumulate in  adipose tissue
23    over time with chronic exposure (Hardy et al.. 2009). DecaBDE also might biomagnify in the food web
24    from lower trophic levels to higher trophic levels. DecaBDE can debrominate to lower PBDE congeners
25    in the body, which are more bioaccumulative than decaBDE (Yogui and Sericano. 2009).
26           As shown in Table H-5. the median level of BDE-209 in the serum reported in one U.S. study
27    was less than 0.96 ng/gram liquid weight. International studies reported median serum levels ranging
28    from 0.77 to 18.5  ng/gram liquid weight. The highest median levels reported were in men  aged 40-50
29    years in Norway and Sweden, and the lowest median levels were from maternal serum in a study in the
30    Faroe Islands (Frederiksen et al.. 2009).
31           BDE-209 has been detected in breast milk in American women at a measured mean concentration
32    of 0.92 ng/gram liquid weight (see Table H-5). A study that evaluated the breast milk of women in the
      March 2013                                    H-50                  DRAFT - DO NOT CITE OR QUOTE

-------
1    Pacific Northwest region of Canada and the United States observed a median level of 0.43 ng/gram liquid
2    weight.
     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
Adipose Tissue
United States
International
Blood (Serum)
United States
International
Cord Blood
United States
International
Placenta
United States
International
BDE-47

7.69-27.81
0.03-27.8

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
BDE-99

1.46-5.7
0.02-5.36

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
BDE-209

0.922
0.1-2.9

NR
NR

<0.96
1.1-18.5

Below detection3
2.2

NR
1.0
















     1High level observed in a joint United States/Canada study.
     2Mean concentration reported.
     Detection limits not reported.
     NR = Not reported.
     Source: Frederiksen etal. (2009).
3    Worldwide, median detected levels of BDE-209 in breast milk ranged from 0.1 to 2.9 ng/gram liquid
4    weight (Frederiksen et al.. 2009). Decline in the use of decaBDE flame retardants containing BDE-209
5    are expected to result in a decline in breast milk concentrations over time.
6            Breastfeeding infants, and even infants who are fed infant formula, are likely to be exposed to
7    BDE-209 through consumption of breast milk and formula. Levels of BDE-209, and levels of total
8    PBDEs, are higher in American samples of breast milk than levels found in infant formula. The levels in
     March 2013
H-51
DRAFT - DO NOT CITE OR QUOTE

-------
 1    both breast milk and formula, however, are far lower than the amount detected in household dust
 2    (Frederiksenetal.. 2009).


      HAS. Ecological Exposure and  Kinetics Leading to Dose

      H.4.3.1. Factors Impacting Ecological Exposure
 3           In biota, potential exposure routes for decaBDE include ingestion, inhalation, or direct contact.
 4    The potential for exposure via each route along with subsequent uptake and dose depends on several
 5    factors, including properties of the environmental media and physiological and behavioral characteristics
 6    of aquatic and terrestrial organisms. These factors can, in turn, influence the bioavailability of decaBDE.
 7    As discussed in Section H.3. the physicochemical properties of BDE-209 dictate partitioning into the
 8    environment. This partitioning drives the exposure potentials for water-dwelling, sediment-dwelling, and
 9    terrestrial organisms. For example, BDE-209 preferentially binds to soils and sediment when released to
10    the environment (Hale et al.. 2006) and likely will be present only in limited quantities in surface water or
11    ground water (see Section H.4.1.2.2) (U.S. EPA. 2010a). Sediment-dwelling organisms are therefore key
12    ecological receptors of BDE-209.

      H.4.3.2. Absorption, Distribution, Metabolism and Excretion in  Ecological  Receptors
13           As discussed in Section H.4.2.6. an understanding of ADME processes can be used to relate
14    exposure concentrations to the concentration, or dose, of material that reaches the tissues of an organism.
15    Elucidation of organism-specific ADME processes can help explain observations of high body burdens
16    that were not predicted based on environmental fate and partitioning alone. ADME processes  influence
17    whether and for how long a material is retained in a tissue (i.e., whether the material will bioaccumulate)
18    and how such retention rates might differ among trophic levels (i.e., whether concentrations of the
19    material will biomagnify in a food web).
20           Bioaccumulation and biomagnification have been shown to influence ecological exposures for
21    decaBDE, as described further in Sections G.4.2.3 and H.4.3.4. Bioaccumulation is the process by which
22    an organism takes a chemical into the body through all exposure routes and dilutes the chemical through
23    excretion, metabolism, and growth, but accumulates a net "body burden" of the chemical (Environment
24    Canada. 2010; U.S. EPA. 2010a). Biomagnification is the process by which a chemical increases in
25    concentration in tissues as it moves up trophic levels in an ecosystem (U.S. EPA. 2010a).
26    Bioaccumulation factors, the ratio of the chemical contaminant in the tissue of the biota (from dietary
27    exposure and uptake directly from media) to chemical contaminant in the medium, and biomagnification

      March 2013                                    H-52                 DRAFT - DO NOT CITE OR QUOTE

-------
 1    factors (BMFs), the ratio of the chemical concentration in an organism's tissue to the concentration of the
 2    same chemical in the tissues of its diet, are used as measures of persistence and potential for impacts as a
 3    chemical moves through an ecosystem (U.S. EPA, 2010a). A substance is considered bioaccumulative
 4    when it has a high bioaccumulation factor, generally greater than 5,000 (U.S. EPA. Final Rule 40 Code of
 5    Federal Regulations 372). Bioaccumulation studies that show body burdens in organisms in remote
 6    locations far from a direct, nondietary exposure source (e.g., water, air), such as those in Greenland
 7    peregrine falcons or Florida coastal sharks, are indicative of trophic biomagnification through the food
 8    web (Environment Canada. 2010). Biomagnification can be modeled using fugacity-based dynamic fate
 9    models that consider environmental conditions, ecosystem properties, and food-web dynamics (Lim and
10    Lastoskie. 2011).
11           Ecological receptors are likely to be exposed to decaBDE through treated products or scraps and
12    debris from products generated during end-of-life stages of the product life cycle (see  Section H.2).
13    The materials released during these processes can contain components other than the contaminant of
14    concern (e.g., textile material, glue, composite ingredients). As discussed at the beginning  of this section
15    (H.4) and throughout Section H.4.2. studies are lacking on the matrix-bound state of these compounds
16    and how exposure characteristics and dose implications differ for the free and matrix-bound forms. As for
17    the discussion of human exposures in Section H.4.2, exposure considerations for ecological receptors are
18    informed by data on BDE-209 not embedded in a polymer matrix or associated with other product
19    ingredients (e.g., textile fibers, coating ingredients). Field studies have found raw PBDEs in
20    environmental media, which indicates that the compounds can leach from the product matrix (see
21    Appendix E).

      HAS.3. Exposure Pathways in Aquatic Systems
22    PBDEs primarily transition to the sediment in aquatic ecosystems (Mikula and Svobodova. 2006); as a
23    result, benthic organisms might take up decaBDE via absorption or ingestion of sediment.  Secondary
24    exposure via movement through the food web results in greater body burdens of PBDEs in predatory fish
25    than in herbivorous or omnivorous fish (Mikula and  Svobodova. 2006). In fish, uptake of PBDE from the
26    water column via gills is limited by the large molecular size of PBDEs, but dietary uptake  efficiencies
27    have been shown to range from 40 to 92% (Mikula and Svobodova. 2006). Due to the tendency for
28    PBDEs to partition into sediment and the inefficiency of uptake via gills, the more likely route of
29    exposure for fish  is  secondary exposure due to bioaccumulation and biomagnification in the food web.
      March 2013                                    H-53                  DRAFT - DO NOT CITE OR QUOTE

-------
      Toxicokinetics and Body Burden in Aquatic Systems
 1           Limited information is available on the mechanisms of BDE-209 ADME in aquatic organisms;
 2    most studies to date have measured concentrations of PBDEs in tissues of aquatic organisms to estimate
 3    body burdens. BDE-209 and the lower brominated congeners BDE-206, BDE-207, and BDE-208 were
 4    experimentally shown to accumulate in the liver of juvenile lake whitefish (Coregonus clupeaformis) as a
 5    result of exposure to BDE-209 via the diet (Kuo et al.. 2010). Fish exposed to 2 ug/gram diet BDE-209
 6    for 30 days accumulated a mean of 5.80 nmol/gram lipid in the liver, compared to 0.208 nmol/gram lipid
 7    for the rest of the body and 0.183 nmol/gram lipid in the liver of control fish. Liver concentrations of
 8    BDE-206, BDE-207, and BDE-208 also were higher compared to control, although concentrations were
 9    less than 0.01 nmol/gram lipid for each congener (Kuo et al., 2010).
10           Similarly, juvenile lake trout (Salvelinus namaycush) exposed to various  PBDE congeners for
11    56 days accumulated measurable PBDE concentrations, and depuration half-lives ranged from 26 to
12    346 days (Tomy et al.. 2004). When BDE-209 was present in the diet at 3.4 ng/gram dry weight, the
13    uptake rate constant was calculated as 132 grams/day. When the concentration of BDE-209 in the diet
14    was raised to 27.5 ng/gram dry weight, the uptake rate constant was much lower—6.1 grams/day. Study
15    authors determined a half-life for BDE-209 of 26 ± 5 days, and a BMP of 0.3 (Tomv et al.. 2004). BMFs
16    of other congeners ranged from 1.6 to 45.9.
17           Body burdens of PBDEs in aquatic organisms have been studied in top predators such as
18    piscivorous birds of prey and top-level fish and in lower level organisms like insects and crabs
19    (Environment Canada. 2010; U.S. EPA. 2010a: Environment Canada. 2006) (see  Section E.2 of Appendix
20    E). Bottom feeders and bivalves often have the lowest PBDE body burdens, eels and higher level fish like
21    sole and flounder have the highest body burdens, and shrimp have mid-range body burdens (U.S. EPA.
22    2010a). Studies have shown accumulation of BDE-209 (exceeding  100 ng/gram)  in top predators,
23    including sharks in coastal Florida and marine mammals such as harbor  porpoise  and white-beaked
24    dolphin (Environment Canada. 2010). Although BDE-209 likely only accumulates at low levels in lower
25    trophic-level organisms, biomagnification can lead to relatively greater concentrations in higher trophic
26    levels.
27           Bioaccumulation rates of BDE-209 and other PBDE congeners are affected by significant
28    biotransformation and debromination (Tomy et al., 2004). so determining the level or  pattern of PBDE
29    uptake from the environment and accumulation in biota is difficult. Laboratory-based  studies that control
30    the exposure rates and measure tissue concentrations allow for calculations of uptake rates, depuration
31    rates, and BMFs. For example, in a study by Kierkegaard et al. (1999). juvenile rainbow trout
32    (Oncorhynchus mykiss) were exposed to technical-grade decaBDE via diet for 120 days, and then
33    observed for 71 days postexposure. The level of decaBDE in the diet ranged from 7.5  to 10 mg/kg body

      March 2013                                    H-54                  DRAFT - DO NOT CITE OR QUOTE

-------
 1    weight per day. Study authors determined that the total uptake in muscle was between 0.02 and 0.13% of
 2    the exposure level. Fish tissue concentrations of BDE-209 increased over the course of the exposure
 3    period, reaching 38 ng/gram fresh weight in muscle and 870 ng/gram fresh weight in the liver at Day 120.
 4    Concentrations declined during the 71-day depuration period to 9.5 ng/gram fresh weight in muscle tissue
 5    and 30 ng/gram fresh weight in the liver (Kierkegaard et al., 1999).
 6           Bioavailability and bioaccumulation of BDE-209 are limited by the high molecular weight of
 7    BDE-209 and its strong sorption to soils and sediments (Kierkegaard et al.. 2004). In general, PBDEs can
 8    bioaccumulate and biomagnify in the aquatic food web (Agrell et al.. 2004). Their propensity to
 9    bioaccumulate and biomagnify depends in large part on their level of bromination. Highly brominated
10    congeners like BDE-209 have a tendency to sink into aquatic sediments and are a minor congener found
11    in aquatic biota. Benthic sediments are a major sink for PBDEs, but BDE-209 does not appear to be
12    readily available to benthic organisms for uptake, although some movement through the food web does
13    occur (Ciparis and Hale. 2005). To what extent  BDE-209 can be transformed in the environment to lower
14    brominated congeners like BDE-47 and BDE-99 is uncertain; BDE-47 and BDE-99 have  been identified
15    as the congeners frequently found in biota (Watanabe and Sakai. 2003).
16           Studies analyzing tissue levels and body burdens of PBDEs in organisms having various roles in
17    a specific ecosystem and food web best illustrate biomagnification, as they quantify the body burdens of
18    different organisms and relate these to food web relationships. In a study by Law et al. (2006a), authors
19    illustrated biomagnification of brominated flame retardants (including PBDEs) in Lake Winnipeg, Canada
20    by determining trophic structure, assessing trophic transfer, and quantifying the magnitude of
21    biomagnification. Samples of water, sediment, plankton, mussels, and six fish species were collected over
22    a four-year period and analyzed for whole-body (in invertebrates) or muscle-tissue (in vertebrates)
23    concentrations of contaminants (see Appendix E. Table E-8). As Table H-6 shows, biomagnification of
24    various PBDE congeners in individual species predator-prey relationships ranged from very positive
25    (BDE-209 concentration in emerald shiner was  33 times higher than in zooplankton)  to negative (BDE-99
26    concentration in emerald shiner was 10 times lower than in zooplankton) (Law et al.. 2006a). The general
27    trend illustrates, however, that higher level predators generally have higher body burdens of PBDEs than
28    lower level prey. The authors determined a trophic magnification factor (which represents the average
29    predator-prey transfer through a food web, as opposed to a BMP, which represents a transfer for a single
3 0    predator-prey relationship) of 3.7 for total PBDEs in the system, and congener-specific trophic
31    magnification factors of 5.2 for BDE-47 (tetraBDE), 1.5 for BDE-99 and 3.0 for BDE-100 (pentaBDEs),
32    and 10.4 for BDE-209 (decaBDE) (Law etal.. 2006a).
      March 2013                                    H-55                  DRAFT - DO NOT CITE OR QUOTE

-------
     H.4.3.4. Exposure Pathways in Terrestrial Systems
1            Although PBDEs are detected commonly in terrestrial ecosystems, exposure levels are higher for
2    terrestrial organisms with diets that consist of animals from the aquatic system than for herbivorous
3    organisms or organisms with diets consisting of animals from the terrestrial system (Mikula and
4    Svobodova. 2006). As discussed in Section H.3.2. LRT can result in exposure to terrestrial organisms far
5    from the initial source of release (de Wit et al., 2010; Su et al., 2009; Breivik et al., 2006; Agrell et al.,
6    2004).

     Toxicokinetics and Body Burden in Terrestrial Systems
     Table H-6.    Biomagnification factors of select PBDE congeners in an aquatic ecosystem.
Predator
System:
Walleye
Burbot
Emerald
System:
Walleye
zooplankton


shiner
zooplankton

White sucker
White sucker
Prey
-> emerald shiner
Emerald shiner
Emerald shiner
Zooplankton
BDE-47
-> walleye,
0
0
5
burbot1
.3
.7
.2
BDE-99

1,
9,
0,

.2
.5
.1
BDE-100

0.2
1
2.2
BDE-153

0.
1.
1.

,3
,7
2
BDE-209

0.6
2.4
33
, mussels -> white sucker -> walleye1
White sucker
Zooplankton
Mussels
0
6
3
.2
.1
.4
2
0,
0,
.1
.1
.1
0.1
3.4
2.9
0.
2.
1.
2
2
,5
2
9.9
0.2









     1Trophic 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: Law etal. (2Q06a).
     March 2013                                        H-56                    DRAFT - DO NOT CITE OR QUOTE

-------
Table H-6, cont: Biomagnification factors of select PBDE congeners in an aquatic ecosystem.
Predator
System: zooplankton
Walleye
Goldeye
Goldeye
System: zooplankton
Walleye
White fish
White fish
Prey
, mussels -> goldeye ->
Goldeye
Zooplankton
Mussels
-> whitefish -> walleye1
White fish
Emerald shiner
Zooplankton
BDE-47
BDE-99
BDE-100
BDE-153
BDE-209
walleye1
0.
7.
2
2
4

8.
0.
0.

,9
,1
2
0,
6,
.1
.5
4.4

1
0,
0,

.7
.7
.1
0.1
4.9
4.2

3.9
0.1
0.1
0.
5.
3.

4.
0.
0.
1
,5
,9

,6
,1
,1
0.
,6


34
0.

6.
0.
2.
8

8
,1
,9





      1Trophic 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: Lawetal. (2006a).
 1            Limited information is available on the mechanisms of BDE-209 ADME in terrestrial organisms;
 2    most studies to date have measured concentrations of PBDEs in specific tissues of terrestrial organisms to
 3    estimate body burdens. Body burdens of PBDEs in terrestrial organisms have been  studied in top
 4    predators such as piscivorous and carnivorous mammals and birds of prey. DecaBDE studies have shown
 5    high accumulation (exceeding 100 ng/gram) in the liver and muscle tissues of top predators, including
 6    kestrel and sparrowhawk in China, the United Kingdom, and Sweden; peregrine falcon in the United
 7    Kingdom, Sweden, and Greenland;  and buzzard and red fox in Belgium (Environment Canada. 2010).
 8    An EPA (2010a) review reported total PBDE concentrations in tissues of predatory birds ranged from
 9    below detection limits in some tissue types to greater than 12,000 ng/gram weight in some muscle and
10    liver tissues (see Appendix E.  Table E-7).
11            As previously discussed, the high molecular weight and strong sorption of BDE-209 to soils and
12    sediments would suggest that bioavailability and bioaccumulation are limited in terrestrial systems. Body
13    burdens of BDE-209, however, have been identified in some—but not all—terrestrial organisms,
14    suggesting that biomagnification does sometimes occur. The complexity of food web interactions that
15    cross aquatic and terrestrial systems makes it challenging to determine whether the  source of BDE-209 in
16    terrestrial food webs stems from contaminated abiotic media in the aquatic environment (e.g., sediments,
17    interstitial waters), the terrestrial environment (e.g., soils, pore water), or a combination of both.
      March 2013                                       H-57                   DRAFT - DO NOT CITE OR QUOTE

-------
      H.4.4. Aggregate Exposures
 1           Assessing aggregate exposures involves characterizing exposures to a single chemical across
 2    multiple exposure routes. Due to the range of applications for which decaBDE can be used, release from
 3    multiple products and subsequent exposure via multiple routes is anticipated.
 4           BDE-209 and the other PBDEs are ubiquitous in the environment (Paso et al.. 2010).
 5    The average daily intake of PBDEs from various routes has been evaluated, and inhalation of dust
 6    provides the highest contribution to body burden of PBDEs. As discussed in Section H.4.1.2.5. levels of
 7    BDE-209 in household dust are orders of magnitude higher than in other matrices, by weight. BDE-209 is
 8    also the main PBDE contaminant in household dust (Paso et al.. 2010). Data are not available on the
 9    relative contribution to household dust of BDE-209 from textile sources.
10           Measurable levels of BDE-209 are found in various types of food worldwide, including milk,
11    fish, shellfish, eggs, beef, chicken, cheese, butter, and other dairy products. The highest concentrations in
12    food have been reported for cod liver, with fish generally making up the highest dietary source of
13    BDE-209 (Paso etal.. 2010; Frederiksen et al..  2009).
14           As a result of exposure from various sources, BDE-209 has been detected in breast milk, serum
15    samples, umbilical cord blood, and the placenta of humans (Paso et al., 2010; Frederiksen et al.. 2009).
16    Ingestion of food, ingestion of dust, inhalation of dust, and dermal contact with soil and dust are the
17    known pathways by which humans are primarily exposed to BDE-209. One review estimates that
18    exposure to BDE-209 from ingestion and dermal contact with soil and dust represents more than 29% of
19    total PBDE exposure from  these exposure routes, and that exposure through those routes accounts for
20    82% of total PBDE exposure from all routes (Lorber. 2008).
      HAS. Cumulative Exposures
21           As stated in The Exposure Factors Handbook (U.S. EPA. 2011). "Cumulative exposure is defined
22    as the exposure to multiple agents or stressors via multiple routes." For the purpose of this case study, the
23    "multiple agents or stressors" considered to contribute to cumulative exposure include those substances
24    that are produced or released as a result of the product life cycles of decaBDE flame-retardant upholstery
25    textile coatings, facilitate uptake of decaBDE into humans and biota, are taken up as a result of decaBDE
26    exposures, or induce effects in humans or biota through a comparable or synergistic mode of action.
27           As discussed in Section H.4.4 on aggregate exposure, PBDEs are ubiquitous in the environment
28    due to their widespread use and physicochemical characteristics. The lower brominated congeners can be
29    metabolites of higher congeners such as BDE-209, and subsequent exposure to lower congeners is likely
      March 2013                                    H-58                 DRAFT - DO NOT CITE OR QUOTE

-------
 1    when BDE-209 is released to the environment. Like BDE-209, many of the lower brominated congeners
 2    such as BDE-47 and BDE-99 can bioaccumulate, and the lower brominated congeners generally are more
 3    toxic than BDE-209. Exposure to the lower brominated metabolites of BDE-209 is also likely following
 4    environmental degradation or aging of composites or textiles containing BDE-209 (Lagalante et al., 2011;
 5    Christiansson et al., 2008).
 6           Manufactured textiles treated with decaBDE could include impurities from the synthesis process
 7    such as PBDDs and PBDFs (Ren etal.. 2011). Breakdown or aging of flame-retardant textiles could
 8    contribute trace amounts of these pollutants to the environment. In addition, disposal or incineration of
 9    these textiles might generate more impurities.
10           Synergistic and antagonistic reactions have been observed in composites. Antimony compounds,
11    which are typically used in the formulation of flame retardants containing decaBDE, tend to act
12    synergistically with halogenated flame retardants to produce highly corrosive hydrogen chloride gas or
13    hydrogen bromide gas, for example (Textile Exchange. 2012). These exposures likely would be limited to
14    manufacturing activities. The decomposition or incineration of manufactured textiles, however, might
15    release other gases that could cause synergistic reactions.


       H.5. Potential  Human  Health, Ecological, and Other
       Impacts
16           The final step of compiling information into the CEA framework is to link the information
17    described in the previous chapters on the product life cycle; transport, transformation, and fate; and
18    exposure-dose with potential impacts to receptors. The CEA framework includes information relevant to
19    impacts on human health and ecological receptors, similar to what might be investigated in traditional risk
20    assessment processes, as well as other plausible impacts that might be considered in life-cycle-focused
21    assessments (e.g., socioeconomics, climate change, resource depletion).
22           Section H.5.1 discusses potential impacts of exposure to decaBDE, and related contaminants on
23    human health. This section relies heavily on evidence from experimental studies with laboratory animals,
24    the results of which could be extrapolated to humans using established quantitative toxicity assessment
25    techniques. As discussed in Section H.4. humans could be exposed to decaBDE or related contaminants
26    from flame-retardant upholstery textiles through a variety of pathways, reaching receptors through dermal
27    deposition, oral ingestion, or inhalation of these contaminants. This section discusses potential health
28    impacts from these exposure routes; data are grouped to illustrate the types of impacts (e.g., pulmonary
      March 2013                                    H-59                  DRAFT - DO NOT CITE OR QUOTE

-------
 1    toxicity, skin irritation, reproductive effects) observed in studies with laboratory animals exposed to
 2    decaBDE and sub-grouped by exposure routes for each impact.
 3           Section H.5.2 discusses the potential impacts of environmental media contaminated with
 4    decaBDE on ecological health, which encompasses impacts on the organism, population, and ecosystem
 5    levels. This section is therefore approached from an ecosystem perspective (aquatic vs. terrestrial), and
 6    data on groups of organisms within those ecosystems are summarized. The discussion of impacts to
 7    ecological health focuses on identifying and comparing data on exposure levels that might cause
 8    significant mortality, delayed growth or development, reproductive defects, or other impacts that could
 9    alter community structure and potentially cause ecosystem collapse.
10           Finally, Section H.5.3 discusses other plausible impacts resulting from the product life cycles of
11    decaBDE in flame-retardant upholstery textiles. The section includes a consideration  of the energy input
12    requirements for synthesis of decaBDE, the economic impacts related to the cost of material production,
13    and the potential for disproportionate impacts on populations with lower socioeconomic  status.


      H.5.1. Human Health Effects
14           This section discusses the potential human health effects resulting from exposures to decaBDE.
15    As noted in Section H.4. exposure to decaBDE from aggregate sources is likely; no studies were found
16    that investigate impacts to human health that can be attributed  directly to exposure to decaBDE or related
17    compounds released during the life cycles of decaBDE flame-retardant upholstery textile coatings.
18    Toxicology studies presented for decaBDE generally were conducted using BDE-209. As discussed in
19    Section H.3 (see Text Box H.3-1). environmental degradation and debromination of decaBDE results in
20    contamination of media with lower PBDE congeners. These lower PBDEs  have toxicological relevance,
21    as they are more bioavailable than decaBDE and potentially more toxic; toxicity of PBDEs generally
22    decreases with increased number of bromine atoms (Rahman et al.. 2001).
23           Toxicology studies conducted on animals comprise much of the information discussed in this
24    chapter because studies on humans  in the literature are limited. Effects observed in animal studies are
25    typically extrapolated to humans when conducting quantitative toxicity assessments (e.g., when
26    calculating an RfD or RfC; see Section H.4.2.5). Potential health effects associated with  all routes of
27    exposure (dermal, inhalation, and oral) are presented in this section because each is plausible for humans
28    (see Section H.4 for additional exposure scenario information).
29           Dermal and oral exposures to decaBDE in dust seem to be the primary routes of exposure for
30    consumer populations (see Section H.4.2.2). Because of higher levels of decaBDE contamination in dust
31    and on other particles, the oral and dermal routes might also be expected to be prominent for general

      March 2013                                     H-60                  DRAFT - DO NOT CITE OR QUOTE

-------
 1    public exposures (see Section G.4.1.1). Available data for decaBDE indicate that the inhalation exposure
 2    route appears to dominate for workers (see Section H.4.2.1). Although inhalation is a possible route of
 3    exposure, especially for workers, many inhalation toxicology studies identified were conducted by
 4    administering the test material (decaBDE) via intratracheal instillation and pharyngeal aspiration; these
 5    routes of administration require an invasive delivery of chemicals or particles and are not as
 6    physiologically relevant for risk assessment purposes, but could provide biological information useful for
 7    qualitative, mode-of-action determinations.
 8           Available information on these exposure routes in experimental animal studies is grouped by the
 9    main types of health impacts observed in the literature, namely, in vivo and in vitro data on systemic
10    toxicity, pulmonary toxicity, eye irritation, skin irritation, reproductive effects, developmental effects,
11    immune system effects, genotoxicity/mutagenicity, carcinogenicity, and susceptible populations.
12    Toxicology studies were reviewed and determined to be key if the following criteria were met:
13           •   appropriate species and test system were used,
14           •   appropriate dose levels were used,
15           •   route of exposure was appropriate for humans,
16           •   control groups were appropriate, and
17           •   the study was consistent with standard  principles and practices.
18           In some cases, multiple studies investigating the same endpoint were available, and the most
19    robust study or the study that most closely aligned with current guidelines for toxicity testing was chosen
20    as key. In other cases, no studies were available that met all the criteria provided above; available studies
21    were then summarized with deficiencies noted in the text. If the study was considered key, a written
22    summary was included in the appropriate section of this appendix. Key studies and supporting non-key
23    studies are  summarized in Table F-3 through Table F-ll in Appendix F. Because a large amount of
24    published data is available for decaBDE, key studies presented in text are primarily those summarized by
25    reviews or agency reports, and only a representative subset of studies are included in Appendix F.
26           The paragraph that follows (see Table 5-1 in Chapter 5) provides an overview of the findings for
27    human health effects of decaBDE, after which a detailed discussion of the available data is presented.
28           Most toxicological studies for decaBDE involve the oral route of exposure (see Section H.4.2.2
29    for discussion on why the oral route appears to be a primary exposure pathway for decaBDE), with
30    thyroid and liver changes observed in rats and mice in subchronic and chronic studies (NTP. 1986; Norris
31    etal.. 1975). Several studies (Johansson et al.. 2008; Viberg et al. 2008; Viberg et al.. 2007; Tseng et al..
32    2006; Viberg et al.. 2003) also reported effects of neonatal exposure, including changes in sperm
33    parameters (Tseng et al.. 2006) and changes in locomotor activity or altered expression of proteins in the


      March 2013                                      H-61                  DRAFT - DO NOT CITE OR QUOTE

-------
 1    central nervous system (Johansson et al., 2008; Viberg et al., 2008; Viberg et al., 2007; Viberg et al.,
 2    2003). EPA calculated an RfD of 0.007 mg/kg-day in 2008 (U.S. EPA. 2008b) based on the
 3    developmental neurobehavioral effects observed in the Viberg et al. (2003) study (see Section H.5.1.1.1
 4    for details on RfD derivation).
 5           With regard to carcinogenicity, the National Toxicology Program (NTP) stated that there was
 6    "some evidence of carcinogenicity" for male and female rats based on significantly increased incidences
 7    of neoplastic nodules of the liver, and "equivocal evidence of carcinogenicity" for male mice based on a
 8    significantly increased incidence of hepatocellular tumors in only the low-dose group and nonstatistically
 9    significant increases in thyroid follicular cell tumors in both dose groups (NTP. 1986). Additionally, the
10    International Agency for Research on Cancer determined that decaBDE is not classifiable as a human
11    carcinogen (Group 3) based on limited evidence in animals (IARC. 1998).  In 2008, EPA used the
12    descriptor "suggestive evidence of carcinogenic potential" for decaBDE (U.S. EPA. 2008b) under
13    relevant guidelines (U.S. EPA. 2005b) (see Section 5.1.11).

      H.5.1.1. Quantitative Toxicity Assessment
14           In a quantitative toxicity assessment, appropriate toxicity information is collected and evaluated.
15    These data then are used to derive toxicity values, such as an RfD for oral exposure or RfC for inhalation
16    exposure.  Similar to an RfD (as defined in Section H.4.2.5). an RfC is an estimate of a continuous
17    inhalation exposure for a given duration to the human population (including susceptible subgroups) that is
18    likely to be without an appreciable risk of adverse health effects over a lifetime. Both values, an RfC and
19    an RfD, are derived from a benchmark dose lower confidence limit, no-observed-adverse-effect level
20    (NOAEL), a lowest-observed-adverse-effect level (LOAEL), or another suitable point of departure, with
21    uncertainty/variability factors applied to reflect limitations of the data used. Other types of toxicity values
22    also can be derived to provide exposure limit values for other exposure durations (e.g., acute or
23    subchronic), more specific populations (e.g., healthy workers), or specific exposure contexts (e.g.,
24    emergency response or occupational exposure; see Section H.4.2.5). The sections that follow discuss the
25    derivation of an RfD for decaBDE; due to limited data, an RfC for decaBDE has not been determined.

26    H.5.1.1.1.  Health Reference Values
27           As mentioned in Section H.4.2.5. EPA (2008b) calculated an RfD of 0.007 mg/kg-day, based on
28    developmental neurobehavioral effects observed in the Viberg et al. (2003) study (see Section H.5.1.7).
29    The NOAEL of 2.22 mg/kg from this  study was used as the point of departure. A total uncertainty factor
30    of 300 was applied to account for interspecies differences (10*), intraspecies differences (10*), and
      March 2013                                     H-62                   DRAFT - DO NOT CITE OR QUOTE

-------
 1    dosing duration (3 x). As mentioned previously, due to the limited toxicity data available, an RfC for
 2    decaBDE has not been determined.

      H.5.1.2. Systemic Toxicity
 3    H.5.1.2.1. Acute
 4           Acute toxicity  studies for all routes of exposure were identified for decaBDE. Results indicate
 5    that, for all routes of exposure, decaBDE exhibits low acute toxicity. No mortality occurred after a 1-hour
 6    inhalation exposure in rats (5 animals/sex/group) to 200 mg/L BDE-209; no gross pathological changes
 7    were observed during the 2-week observation period (CPTC. 1978).
 8           The low acute  oral toxicity of decaBDE (Zhou et al.. 2001; Kierkegaard et al.. 1999; Norris et al..
 9    1973) might be due in part to poor gastrointestinal absorption. No clinical signs of toxicity or death were
10    observed when a single dose (up to 5,000 mg/kg) of BDE-209 was administered to rats via gavage (Great
11    Lakes Chemical Corporation. 1994; IRDC. 1974).
12           The low acute  dermal toxicity of decaBDE is presumed based on lack of treatment-related effects
13    in rabbits following single administrations of 200 or 2000 mg/kg BDE-209 to clipped intact skin for
14    24 hours (14-day observation period) (Great Lakes Chemical Corporation, 2000b; IRDC. 1974).

15    H.5.1.2.2. Subchronic
16           In humans, an increase in primary hyperthyroidism and a significant reduction in calf sensory and
17    fibula motor nerve velocities were observed in workers exposed to decaBDE during manufacturing
18    [(Bahn et al. (1980) as  cited in NTP (1986): Bialik (1982), as cited in HSDB (2011)1: whether these
19    effects are due to decaBDE or polybrominated biphenyls is unclear, however, because only
20    polybrominated biphenyls were detected in blood.
21           Numerous subchronic oral  studies were identified for decaBDE, and all studies considered,
22    including those summarized  below, are presented in F.I.2 in Appendix F. DecaBDE-related thyroid and
23    liver changes were observed in male rats when administered a lower purity (77.4%)  form of decaBDE
24    (Norris et al..  1975: Norris et al.. 1973). Effects included thyroid hyperplasia, increased liver weight, and
25    hepatic centrilobular cytoplasmic enlargement and vacuolation (Norris et al.. 1975: Norris et al.. 1973).

26    H.5.1.2.3. Chronic
27           Numerous chronic oral studies were identified for decaBDE, and all studies  considered, including
28    those summarized below, are presented in Section F.I.2 in Appendix F. In a chronic study conducted by
29    NTP (1986). a dose-dependent increase in thyroid follicular cell hyperplasia was observed in male mice
      March 2013                                    H-63                  DRAFT - DO NOT CITE OR QUOTE

-------
 1    fed BDE-209 (purity 94-97%) in the diet for 103 weeks (NTP. 1986); these effects were not observed in
 2    female mice or female and male rats that were similarly exposed to BDE-209. Centrilobular hypertrophy
 3    (consisting of enlarged hepatocytes with frothy vacuolated cytoplasm) also was observed in male mice,
 4    but not in female mice or in male and female rats (NTP. 1986). Incidences of thrombosis and
 5    degeneration of the liver were increased in male rats at the LOAEL (2,240 mg/kg-day), but not at
 6    1,120 mg/kg-day (NOAEL); these hepatic effects were not observed in female rats or in mice of either
 7    sex. The NTP (1986) studies were considered for the basis of the EPA (2008b) quantitative cancer
 8    assessment (see Section H.5.1.11). Observed changes in liver weight and hepatocytomegaly might have
 9    been due to enzyme induction, as supported by recent studies conducted by Van der Ven et al. (2008) and
10    Bruchajzer et al. (2010) (see Table F-7 in Appendix F).

      H.5.1.3. Pulmonary Toxicity
11           When BDE-209 was administered to rats via a single intratracheal injection of 20 mg BDE-209
12    dust (purity 77.4%) suspended in rat serum (Dow Chemical Co. 1990b). minimal histopathological
13    changes (scattered focal aggregates of alveolar macrophages) occurred, consistent with retention of large
14    dust particles that would not normally reach the lungs during inhalation. Rats exposed to 2,000 or
15    48,000 mg/m3 BDE-209 for 1 hour exhibited dyspnea at both dose levels (Great Lakes Chemical
16    Corporation, 1994; IRDC. 1974); all animals survived until study termination and were normal at the end
17    of the 14-day observation period (see Table F-5 in Appendix F).

      H.5.1.4. Eye  Irritation
18           All in vivo eye irritation studies considered are presented in Section F.I.2 (Table F-3). Key
19    studies are summarized below.
20           DecaBDE does not appear to be an eye irritant. Ocular exposure to dry solid decaBDE caused
21    transient conjunctival irritation in washed and unwashed rabbit eyes when 100 mg of decaBDE was
22    administered via instillation to the conjunctival sac (NRC. 2000; IRDC. 1974)  [Effects in some rabbits
23    included very slight conjunctival redness and chemosis and slight or moderate discharge (Great Lakes
24    Chemical Corporation. 1994)1. Investigators concluded that the effects were not serious  enough to be
25    considered primary eye irritation (Norris et al.. 1975; IRDC. 1974). Pharmakon (1994) similarly reported
26    that decaBDE (Saytex 102) did not cause primary eye irritation when instilled once (100 mg/eye) into the
27    eyes of rabbits. Rats exposed to 2,000 or 48,000 mg/m3 BDE-209 dust in the ambient air for 1 hour,
28    however, exhibited ocular porphyrin discharge  at both dose levels, and eye squint at the  high
      March 2013                                    H-64                  DRAFT - DO NOT CITE OR QUOTE

-------
 1    concentration (IRDC. 1974); all animals survived until study termination and were normal at the end of
 2    the 14-day observation period.

      H.5.1.5. Skin Irritation
 3           The material characteristics and study details associated with the in vivo dermal studies
 4    considered for decaBDE and MWCNTs are presented in Section F.I.2. in Appendix F
 5           DecaBDE does not appear to be a skin irritant based on observations from a human skin irritation
 6    study (Dow Chemical Co. 1990a; Norris et al.. 1975; Norris et al.. 1973). a skin irritation study in rabbits
 7    (Norris et al.. 1975; IRDC. 1974; Norris et al.. 1973). and an acne-genesis study in rabbits (Pharmakon
 8    Research International. 1994). Dermal studies were conducted with BDE-209.

      H.5.1.6. Reproductive Effects
 9           In general, studies found that decaBDE was not a reproductive toxicant at doses up to and
10    exceeding 1,000 mg/kg-day (Tseng et al.. 2008; Hardy et al.. 2002; Dow Chemical Co. 1990c; NTP.
11    1986). Van der Ven et al. (2008) reported significant, decaBDE dose-related changes in epididymis and
12    seminal vesicle weight for male rats and decreased activity of CYP17, a key enzyme in the androgen
13    synthesis pathway, for female rats administered 1.9-60 mg/kg by oral gavage. No corresponding
14    histopathological changes, sperm counts, or morphology of epididymal sperm, however, were observed.
15    Based on these results, the authors concluded that BDE-209 might represent a hazard to
16    reproductive health.

      H.5.1.7. Developmental Effects
17           Several studies reported no developmental effects for decaBDE at doses up to and exceeding
18    1,000 mg/kg-day (Hardy et al.. 2009; Tseng et al.. 2008; Hardy et al.. 2002; Dow Chemical Co. 1990c).
19    A significant increase in CYP450 activity in adult male CD-I mouse offspring was noted in the study
20    conducted by Tseng et al. (2008) at doses of 1,500 mg/kg-day; this dose  level, however, exceeds the
21    current dose  limit (e.g., 1,000 mg/kg-day) recommended by international toxicity testing guidance
22    documents (OECD. 2007; U.S. EPA. 1998). Another study found significant increases in numbers of rat
23    litters with subcutaneous edema and delayed ossification of skull bones at 1,000 mg/kg-day (Norris et  al..
24    1975); dams  for this study were administered BDE-209 (77.4% containing 21.8% nonabromodiphenyl
25    oxide and 0.8% octabromodiphenyl oxide) via gavage at dose levels of 0, 10; 100; or 1,000 mg/kg-day.
26    Consequently, the NOAEL and LOAEL for fetal effects in this study were  100 and 1,000 mg/kg-day,
27    respectively; the NOAEL for maternal effects  was 1,000 mg/kg-day.

      March 2013                                     H-65                  DRAFT - DO NOT CITE OR QUOTE

-------
 1           In contrast, several studies did observe adverse effects when neonatal mice or rats were exposed
 2    orally to decaBDE at lower doses (Johansson et al.. 2008; Viberg et al.. 2008; Viberg et al.. 2007; Tseng
 3    et al.. 2006; Viberg et al.. 2003). Effects of neonatal exposure included changes in sperm parameters
 4    (Tseng et al.. 2006) and changes in spontaneous behavior or altered expression of proteins in the central
 5    nervous system (Johansson et al.. 2008; Viberg et al.. 2008; Viberg et al.. 2007; Viberg et al.. 2003). Rice
 6    et al. (2007) orally exposed male and female mouse pups to decaBDE at doses of 0, 6, or 20 mg/kg-day
 7    from postnatal days PND2 through PND15. Treatment-related effects occurred only in the high-dose
 8    group and included a reduction in palpebral reflex on PND14, a reduction in forelimb grip in males on
 9    PND16, a change in the slope of the linear trend for serum T4 in males on PND21, and a change in the
10    linear slope of motor activity on PND70. Of the available studies, Viberg et al.  (2003) was selected for
11    the derivation of the RfD (see Section H.5.1.1.1).

      H.5.1.8. Immune System Effects
12           No immunology studies were identified for decaBDE.

      H.5.1.9. In Vitro  Data
13           In vitro data can be used to make judgments on the toxic potential of stressors, but the relevance
14    of in vitro data to predicting toxicological responses of "real-world" exposures  is not always clear.
15           No in vitro  data were identified for decaBDE.

      H.5.1.10. Genotoxicity/Mutagenicity
16           DecaBDE does not appear to be genotoxic and generally did not induce (1) gene mutations in
17    bacteria (Salmonella typhimurium) in Ames assays (Chemical Manufacturers Association. 1998; GSRI.
18    1990; Huntingdon Life Sciences. 1990; NTP. 1986; Haworth et al.. 1983; Litton Bionetics.  1976).
19    (2) gene mutations in mouse  L5178Y lymphoma cells, (3) chromosomal aberrations in mouse bone
20    marrow cells, or (4) sister-chromatid exchanges or cell transformation in Chinese hamster ovary cells
21    (Myhr et al., 1990; McGregor et al.. 1988). These findings are consistent with those reported by EPA
22    (U.S. EPA. 2008b). Because  decaBDE has consistently failed to produce  genotoxic or mutagenic
23    responses, the aforementioned studies are not summarized in Appendix F.
      March 2013                                    H-66                  DRAFT - DO NOT CITE OR QUOTE

-------
      H.5.1.11. Carcinogenicity
 1           Carcinogenicity studies considered for decaBDE are presented in Table F-10 of Appendix F.
 2    The target organs for decaBDE Carcinogenicity appear to be the liver and thyroid; decaBDE was not,
 3    however, included on the most recent U.S. NTP list of carcinogens.30
 4           Information on the Carcinogenicity of decaBDE is available from three chronic feeding studies in
 5    rodents (KocibaetaL 1994; NTP. 1986; Kocibaetal.. 1975). In the NTP study (1986). a treatment-
 6    related increase in liver neoplastic nodules was observed in low- and high-dose male rats (7/50 and 15/49,
 7    respectively, compared to 1/50 in controls) and high-dose female rats (9/50 compared to 1/50 and 3/49 in
 8    control and low-dose groups, respectively). F344/N rats were fed BDE-209 (94-98% pure) at dietary
 9    concentrations of 0, 25,000, or 50,000 ppm for 103 weeks (equivalent to 0, 1,120, and 2,240 mg/kg-day in
10    male rats; 0, 1,200, and 2,550 mg/kg-day in female rats). The increase in liver neoplastic nodules was not
11    accompanied by an increase in hepatocellular carcinomas in rats. Hepatocellular adenomas or carcinomas
12    (combined), however, were observed in low- and high-dose male mice  (8/50 controls, 22/50 low-dose
13    mice, 18/50 high-dose mice). Male mice also exhibited a marginal increase in thyroid gland follicular cell
14    adenomas or carcinomas (combined) at the low and high doses (0/50 controls, 4/50 low-dose mice, 3/50
15    high-dose mice). The possible significance of this finding was strengthened by increased incidences of
16    follicular cell hyperplasia in the male mice (2/50 controls, 10/50 low-dose mice, 19/50 high-dose mice),
17    but was weakened by increased mortality in control animals. Based on these results, the NTP (1986)
18    study concluded that there was "some evidence of Carcinogenicity" for male rats in the low-dose group
19    and both male and female rats in the high-dose group based on significantly increased incidences of
20    neoplastic nodules of the liver, and "equivocal evidence of Carcinogenicity" for male mice based on a
21    significantly increased incidence of hepatocellular tumors in only the low-dose group and nonstatistically
22    significant increases of thyroid follicular cell tumors in both dose groups. Although the International
23    Agency for Research on Cancer (1998) reports that decaBDE is not classifiable as a human carcinogen
24    (Group 3) based on limited evidence in animals, EPA, under the Guidelines for Carcinogen Risk
25    Assessment (U.S. EPA. 2005b). determined that the descriptor "suggestive evidence of carcinogenic
26    potential" is appropriate for decaBDE (U.S. EPA. 2008b) based on the  data from NTP (1986)
27    demonstrating evidence of Carcinogenicity in more than one species, sex, and site.
      30The U.S. NTP 12th Report on Carcinogens (released June 2011) is available at
      http://ntp.niehs.nih. gov/?obiectid=03C9AF75-ElBF-FF40-DBA9EC0928DF8B15

      March 2013                                     H-67                  DRAFT - DO NOT CITE OR QUOTE

-------
      H.5.1.12. Susceptible Populations
 1            Sacks et al. (2011) defined susceptibility as "individual- and population-level characteristics that
 2    increase the risk of health effects in a population, including, but not limited to, genetic background, birth
 3    outcomes (e.g., low birth weight, birth defects), race, sex, life stage, lifestyle (e.g., smoking status,
 4    nutrition), preexisting disease, socioeconomic status (e.g., educational attainment, reduced access to
 5    health care), and characteristics that may modify exposure ...  (e.g., time spent outdoors)." In this section,
 6    populations susceptible to decaBDE impacts based on characteristics  such as age, genetic background,
 7    and disease are considered. Characteristics that could modify exposure and increase susceptibility were
 8    discussed previously in Section H.4.2.4; for a discussion on impacts related to socioeconomic status, see
 9    Section H.5.3.
10            Results regarding developmental neurotoxicity studies are conflicting. Whether young children
11    comprise a sensitive (i.e., more susceptible) population is therefore unclear. A few animal studies have
12    indicated that BDE-209 might cause developmental neurotoxicity, affecting motor and cognitive domains;
13    however, in discussing susceptible populations for decaBDE, EPA noted that differences in the effects of
14    decaBDE on neurodevelopment are unclear and whether other targets (thyroid and liver) are more
15    sensitive in children is unknown (U.S. EPA. 2008b). Disposition studies using pregnant rats indicate that
16    fetuses are less exposed to decaBDE than mothers. Fetuses (whole litter) contained only 0.43% of the
17    dose in a study by Riu et al. (2008). In a study by Inoue et al. (2006),  higher brominated congeners like
18    decaBDE transferred from blood to milk to a lesser degree than did lower brominated congeners. In a
19    study by Fukata et al. (2005), BDE-209 was not detected in umbilical cord tissue, but was found at
20    23 ng/gram lipid weight in umbilical cord serum and 10 ng/gram lipid weight in maternal serum.
21    Exposure to decaBDE in infants, however, appears to be greater than  in adults. Additional information
22    regarding populations that are susceptible to greater levels of exposure can be found in Section H.4.
23    No sources indicating that specific genetic polymorphisms increase susceptibility were identified.
      H.5.2.  Ecological Effects
24           This section presents a summary of data on the potential ecological impacts of environmental
25    contamination with decaBDE. Specific information from the studies reviewed for this case study can be
26    found in Section F.2 in Appendix F. Considerations for ecological impact include the absolute and
27    relative toxicity of the decaBDE and other factors such as bioaccumulation and biomagnification potential
28    (see Section G.4.2.1). For aquatic ecosystems, much information was available for decaBDE primarily
29    because decaBDE has been studied extensively in aquatic vertebrates. Conversely,  little information was
      March 2013                                     H-68                   DRAFT - DO NOT CITE OR QUOTE

-------
 1    identified on the potential effects of decaBDE in terrestrial ecosystems. The terrestrial ecosystem studies
 2    focus on agriculturally relevant plants and soil microbes. In both aquatic and terrestrial ecosystems,
 3    studies are predominantly laboratory-based experiments on single species. Few studies address how
 4    exposure and uptake of PBDEs relate to ecological health and effects in the field (Vonderheide et al.,
 5    2008).  Some field studies show correlations between PBDE exposure, reproductive behavior, and
 6    immunosuppression, but ecological consequences and potential population-level impacts of
 7    environmental PBDE contamination in general, and decaBDE in particular, remain uncertain
 8    (Vonderheide et al.. 2008). As mentioned in Section G.4.2.1, ecological receptors can be exposed to
 9    decaBDE attached to textile fibers, embedded in polymers, or sorbed to other particles, all of which are
10    more likely to occur in the environment than exposure to the pristine compound. Studies examining
11    exposure to larger textile scraps, polymer particles, and other heterogeneous compounds containing
12    decaBDE, however, are lacking. The results of laboratory studies using pristine compounds must
13    therefore be considered, recognizing that results might not translate directly into real-world exposure
14    scenarios.
15           As discussed in Section H.3 (see Text BoxH.3-1). environmental degradation and debromination
16    of decaBDE result in contamination of media with lower PBDE congeners, which are generally more
17    bioavailable and more toxic than decaBDE. Also as stated previously in Section H.3. debromination of
18    decaBDE is expected to contribute significantly to the environmental presence of BDE-47, a tetraBDE,
19    and BDE-100 and BDE-99, which are both pentaBDEs, among other congeners (Gandhi et al., 2011;
20    Ross et al., 2009). In this section, information is presented on the acute toxicity of decaBDE and other
21    PBDEs, as appropriate. Considerations for negative ecological impacts from continual long-term
22    exposures in an ecosystem are also discussed.

      H.5.2.1. Aquatic Receptors
23           DecaBDE has been  shown not to be acutely toxic to species of fish or marine algae studied to
24    date (Hardy. 2002a). Few studies were identified regarding the acute or chronic toxicity of decaBDE to
25    marine or freshwater algae and benthic invertebrates (see  Section H.5.2.1.1). No information was
26    identified regarding toxic effects on aquatic plants or water-dwelling invertebrates (see Section
27    H.5.2.1.1): some information was identified on toxicity to aquatic vertebrates (see Section H.5.2.1.2).
28           DecaBDE is not expected to be chronically toxic to aquatic organisms at environmentally
29    relevant concentrations due to its physicochemical properties, specifically high molecular weight and low
30    water solubility (Hardy. 2002a) (see Section H.I). These properties suggest that decaBDE accumulation
31    directly from water into biota is unlikely; a more likely route of exposure is dietary (Gandhi et al.. 2011).
      March 2013                                     H-69                  DRAFT - DO NOT CITE OR QUOTE

-------
 1    But as noted previously, the factors affecting bioavailability of decaBDE are not well understood.
 2    Comparatively, however, lower PBDEs such as pentaBDEs are known to have high potential for
 3    bioaccumulation (U.S. EPA. 2010a). Because these congeners are transformation products of decaBDE,
 4    their chronic toxicity is considered.

 5    H.5.2.1.1. Algae, Aquatic Plants, and Aquatic Invertebrates
 6           The paragraphs that follow describe literature identified for the effects of decaBDE on algae,
 7    aquatic plants, and aquatic invertebrates. Table F-12. Table F-14. and Table F-15 in Appendix F
 8    summarize details of the studies identified and reviewed for this section.
 9           No studies were identified that investigated the effects of decaBDE on algae or on aquatic plants.
10    A single study on water-dwelling aquatic invertebrates was identified involving freshwater bivalve zebra
11    mussels (Dreissenapolymorphd) exposed to technical-grade decaBDE at sublethal levels of 0.1, 2, or
12    10 ug/L. This study showed DNA damage that increased as levels of decaBDE exposure increased,
13    indicating potential for genotoxicity (Riva et al.. 2007). A review by Hardy (2002a) stated that decaBDE
14    was nontoxic to marine algae and sediment oligochaetes, but no details on the derivation of these
15    conclusions were provided. The review also investigates toxicity of octaBDE and pentaBDE, concluding
16    that octaBDE is neither acutely toxic nor chronically toxic to a species of water flea, and pentaBDE is not
17    acutely toxic to algae, up to the limit of their water solubility (Hardy. 2002a). Details on endpoints
18    observed were not provided.
19           Environment Canada (2006) reviewed ecotoxicity studies for multiple PBDE mixtures. They
20    report high (>5,000 mg/kg) no-observed-effect levels and median (>50 mg/kg) effective concentration
21    values for chronic survival and reproduction effects for a freshwater oligochaete (Lumbriculus variegatus)
22    exposed to sediments that contained a mixture of 55% pentaBDE and 36% tetraBDE as well as a mixture
23    containing 97% decaBDE. For water fleas (Daphnia magna), Environment Canada (2006) reported
24    toxicity values in the low ug/L range for survival, growth, and reproduction following chronic exposure
25    to a commercial pentaBDE mixture [(Drottar and Krueger (1998) as cited in Environment Canada
26    (2006)]. Some water fleas are therefore more sensitive to PBDEs than oligochaete worms, but as
27    mentioned in Section H.3.3. decaBDE is not likely to remain in the water column; instead, it partitions to
28    sediment, where benthic invertebrates are expected to be exposed.

29    H.5.2.1.2. Aquatic Vertebrates
30           Table 5-3 in Chapter 5 describes key toxicity values identified for the effects of decaBDE and
31    MWCNTs on aquatic vertebrates. Table F-13.  Table F-16. and Table F-17  in Appendix F summarize
32    details of the studies identified and reviewed for this section.

      March 2013                                     H-70                  DRAFT - DO NOT CITE OR QUOTE

-------
 1           A review by Hardy (2002a) reports on acute toxicity of decaBDE, octaBDE, and pentaBDE,
 2    stating that all three congeners have a fish 48-hr median lethal concentration of greater than 500 mg/L,
 3    indicating that the congeners are not acutely toxic to fish up to the limit of their water solubility.
 4    No effects on egg mortality were observed in rainbow trout (Oncorhynchus mykiss) at doses up to 12 ug
 5    pentaBDE per egg, and no effects on reproduction or spawning success were observed in three-spined
 6    stickleback (Gasterosteus aculeatus) exposed to pentaBDE (Hardy. 2002a).
 7           Tests conducted on frogs often measure low-dose, chronic thyroid disruption, because
 8    metamorphic development from tadpole to frog is controlled by thyroid hormones (Qin et al.. 2010).
 9    African clawed frog (Xenopus laevis) tadpoles (Table 5-3 in Chapter 5) exposed to decaBDE in their
10    water at sublethal doses ranging from 1 to 1,000 ng/L experienced histopathological alterations in thyroid
11    gland cell shapes and decreases in thyroid hormone expression in tail tissue during metamorphosis at all
12    tested doses. Additionally, researchers observed a concentration-dependent trend of delay in time to
13    metamorphosis with a statistically significant delay at 1,000 ng/L (Qin etal.. 2010).
14           Endocrine effects also have been studied in Chinese rare minnow (Gobiocypris rams) and lake
15    trout (Salvelinus namaycush) chronically exposed to decaBDE via water and diet, respectively (Li et al.;
16    Tomy et al.. 2004). Expression of thyroid hormone-related genes was variably affected in both studies;
17    indicating the potential for chronic endocrine disruption but not elucidating a mechanism for those effects
18    or a clear effect level. Chronic toxicity of decaBDE in fish is complicated by biotic debromination of
19    decaBDE, which can result in bioaccumulation of octa-, hepta-, hexa-, and pentaBDE congeners (Gandhi
20    etal., 2011; Stapleton et al., 2004) (as discussed in Section H.4.2.6.1). Most informative, therefore, are
21    studies of multiple PBDEs or PBDE mixtures. In one such study, induction of vitellogenin production
22    was observed in hepatocyte cell cultures of rainbow trout (Oncorhynchus mykiss) exposed to PBDE
23    mixtures (Nakari and Pessala. 2005).This estrogenic response raises concerns for population dynamic
24    impacts due to endocrine disruption (Mikula and Svobodova.  2006).

      H.5.2.2. Terrestrial Receptors
25           A limited amount of information was found regarding toxicity of decaBDE to  soil microbes,
26    plants, and terrestrial invertebrates (see Sections H.5.2.2.1 and H.5.2.2.2). No information was identified
27    regarding toxicity to terrestrial vertebrates (see Section H.5.2.2.3); nevertheless, some  assumptions can be
28    made for mammals based on toxicity studies intended for human health purposes presented in Section
29    H.5.1. Many studies of decaBDE in terrestrial ecosystems have focused on bioaccumulation and
30    biomagnification; important considerations for potential ecological hazard (see Section H.4.3.4). These
31    studies, however, did not investigate occurrence of toxic effects.
      March 2013                                     H-71                   DRAFT - DO NOT CITE OR QUOTE

-------
 1    H.5.2.2.1. Soil Microbes and Terrestrial Invertebrates
 2           Changes in soil microbial activity result in changes to nutrient cycling; therefore, studying the
 3    impact of contaminants on soil microbes can provide insight on how those contaminants might affect
 4    ecosystem function (Chung et al.. 2011). Similarly, effects on terrestrial invertebrates, such as worms, can
 5    influence health and fertility of a soil ecosystem (Xie etal. 2011). Table F-18 and Table F-19 in
 6    Appendix F summarize details of the studies identified and reviewed for this section.
 7           Two studies were identified that investigated the toxicity of decaBDE to soil microbes. Although
 8    Sverdrup et al. (2006) showed no effects on nitrifying ability of bacteria following exposure to decaBDE
 9    at levels up to 2,274 mg/kg in soil, Liu et al. (201 la) found that microbial cytotoxicity significantly
10    increased at doses 10-fold lower (100 mg/kg), and  community structure was altered following long-term
11    exposure to decaBDE. The rate of community diversity increase overtime was significantly slower from
12    Day 90 through the last day of the study (Day 180) when soil contained 1-100 mg/kg decaBDE. After six
13    months, the total bacterial count in the soil containing 100  mg/kg decaBDE was approximately half that
14    of the control plot. Treatment soil microcosms were dominated by Pseudomonas, Bacillus, and
15    uncultured bacteria types, and had significantly reduced cell counts for alpha, beta, and gamma type
16    proteobacteria and the Cytophaga-Flavobacterium-Bacteroides group (Liu etal., 201 la).
17           Studies of oligochaete worms \Enchytraeus crypticus (a soil worm) and Eiseniafetida
18    (earthworms)] showed that survival, reproductive behavior, and number of offspring are not affected by
19    long-term exposure to decaBDE in soil in the grams/kg range [Sverdrup et al. (2006); ACC (2001) as
20    cited in Environment Canada (2006)]. Earthworms, however, experienced a sublethal, dose-dependent
21    increase in hydroxyl radical generation and subsequent oxidative stress after 1 week of exposure to 0.1-
22    10 mg/kg decaBDE (Xie etal.. 2011). Oxidative stress in earthworms is considered a biomarker
23    indicative of potential for greater impacts of soil contaminants within terrestrial ecosystems (Xie et al..
24    2011).

25    H.5.2.2.2. Terrestrial Plants
26           Table F-18 and Table F-20 in Appendix  F summarize details of the studies identified and
27    reviewed for this section.
28           Few studies were identified that investigated effects of PBDEs on plants; those reviewed found
29    no adverse effects at environmentally relevant concentrations. No effects on seedling emergence were
30    observed in red clover (Trifolium pretense} exposed to decaBDE or corn (Zea mays) exposed to a PBDE
31    mixture (55% pentaBDE and 36% tetraBDE) at levels in the grams/kg range (Sverdrup et al.. 2006; Great
32    Lakes Chemical Corporation. 2000a). In corn, mean shoot  height was unaffected at concentrations up to
33    125 mg/kg, but was significantly reduced at 250 mg/kg and above  [Great Lakes Chemical  Corporation

      March 2013                                      H-72                   DRAFT - DO NOT CITE OR QUOTE

-------
 1    (2000a) as cited in Environment Canada (2006)1. As discussed in Section H.3.4, soils are a major sink for
 2    PBDEs in terrestrial systems, and uptake by plants is possible. Recent measured concentrations in soil
 3    have been in the ng/gram (0.001 mg/kg) range (see Table E-5 in Appendix E).

 4    H.5.2.2.3. Terrestrial Vertebrates
 5           No studies were identified that specifically investigated the effects of decaBDE on terrestrial
 6    vertebrates  outside the laboratory setting. Results from extensive testing performed in mammals for
 7    toxicological relevance to humans are reported in Section H.5.1.
 8           Chronic effects of PBDEs at environmentally relevant exposure concentrations are a possibility in
 9    terrestrial vertebrates, primarily due to assumed ecological impacts associated with high biomagnification
10    rates, as discussed previously in Section G.4.2.1.
      H.5.3.  Other Impacts
11           As stated in Section H.I, the CEA framework considers not only human and ecological health
12    impacts, but also aesthetic, environmental, social, legal, ethical, and economic impacts. Such impacts
13    might be associated with impacts on specific socioeconomic sectors (e.g., disparate impacts on
14    environmental justice communities), the environment as a whole (e.g., climate change, depletion of
15    natural resources, energy demand), or the built environment (e.g., damage to building facades).
16           Apart from the impacts discussed in Sections H.5.1 and H.5.2. the only other impacts considered
17    in this case study are those for which a plausible premise can be developed to support assumptions that a
18    discernible impact might occur as a result of the life cycles of decaBDE flame-retardant upholstery textile
19    coatings. Empirical data have revealed a correlation between decaBDE body burdens and socioeconomic
20    status, indicating that effects having environmental justice implications are plausible for decaBDE.

      H.5.3.1. Environmental Justice
21           Environmental justice is defined by EPA as the "fair treatment and meaningful involvement of all
22    people regardless of race, color, national origin, or income with respect to the development,
23    implementation, and enforcement of environmental laws, regulations, and policies."31 The goal  of
24    environmental justice is to give all people ".. .the same degree of protection from environmental and
      31U.S. EPA Compliance and Enforcement. Environmental Justice, http://www.epa.gov/environmentaljustice/
      March 2013                                     H-73                  DRAFT - DO NOT CITE OR QUOTE

-------
 1    health hazards and equal access to the decision-making process... "32 As a result, environmental justice
 2    impacts include those in which a particular group or geographic area experiences a disproportionate share
 3    of the impacts associated with an environmental contaminant.
 4           Releases of decaBDE throughout the life cycle of a flame-retardant upholstery coating product
 5    could disproportionately impact certain communities. In a review by Zota et al. (2010). findings from
 6    several recent studies suggest that racial and ethnic minorities and populations having lower
 7    socioeconomic status (i.e., low income, low educational attainment) experience disproportionate
 8    exposures to PBDEs. For example, Rose et al. (2010) observed that body burdens of BDE-209, among
 9    other congeners, were significantly higher in children aged 2-5 years born to mothers of lower
10    educational attainment compared to those born to mothers achieving a college degree or higher.
11           The causal pathway connecting low socioeconomic status to elevated PBDE exposure is not well
12    understood, but Zota et al. (2010) hypothesized that furniture quality and the characteristics of the living
13    spaces (e.g., size, ventilation, age), which populations of lower socioeconomic status might occupy,
14    contribute to elevated exposure to PBDEs. Indeed, Rose et al. (2010) demonstrated that higher maternal
15    education attainment is correlated with larger living spaces, and in turn, children living in larger homes
16    had lower body burdens of BDE-209. Similarly, Stapleton et al. (2012) found that variation in PBDE
17    serum concentrations in children could be explained by handwipe  levels, house dust levels, father's
18    education, breast feeding duration, age, and gender (different factors associated with different PBDE
19    congeners).

      H.5.3.2. Energy  Demand and Natural Resource Depletion
20           No  information was identified that examined  impacts on energy demand and natural resource
21    depletion associated with the production  of decaBDE flame-retardant textile coatings.

      H.5.3.3. Climate Change
22           No  information was identified that examined  climate change impacts due to decaBDE flame-
23    retardant textile coatings.
      32ibid
      March 2013                                    H-74                  DRAFT - DO NOT CITE OR QUOTE

-------
     H.5.3.4. Economics

1            No information was identified that calculated the cost of manufacturing decaBDE or decaBDE
2    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

        Ann. 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/i.envint.2010.04.008

        Bogdal C: Scheringer. M: Schmid. P: Blauenstein. M: Kohler. M: Hungerbuhler. 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

        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.

        Bruchajzer. 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
     March 2013                                       H-75                    DRAFT - DO NOT CITE OR QUOTE

-------
   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.
      http://www.qsa.man.dtu. dk/English/Research/Reports.aspx?lg=showcommon&id=265562

   Chemical Manufacturers Association. [Wagner, V.O. and Klug, M.L.] (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 Sci Pollut Res 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

   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 In Press, Corrected Proof: 2885-2918.
      http://dx.doi.0rg/10.1016/i.scitotenv.2009.08.037
March 2013                                        H-76                    DRAFT - DO NOT CITE OR QUOTE

-------
   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://esis.jrc.ec.europa.eu/doc/existing-chemicals/risk assessment/SUMMARY/decasumO 13.pdf

   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.ee.gc.ca/lcpe-cepa/default.asp?lang=En&n=B901A9EB&offset=l&toc=show

   EU (European Union). (2002). European risk assessment report for bis(pentabromophenyl) ether.
      http://esis.jrc.ec.europa.eu/doc/existing-
      chemicals/risk  assessment/REPORT/decabromodiphenyletherreportO 13 .pdf

   Frederiksen. M: Vorkamp. K: Thomsen. M: Knudsen. LE. (2009). Human internal and external exposure to
      PBDEs -A review of levels and sources [Review]. IntJHyg Environ Health 212: 109-134.
      http://dx.doi.0rg/10.1016/i.iiheh.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.

   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. KC. (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. KC. (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
March 2013                                       H-77                   DRAFT - DO NOT CITE OR QUOTE

-------
   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=OTSOOO 1105

   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. http://dx.doi.org/10.1080/10408440902837967

   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.  (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

   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

   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: (g),term+@DOCNO+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
March 2013                                        H-78                   DRAFT - DO NOT CITE OR QUOTE

-------
   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 polychlorinated 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 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

   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/i.chemosphere.2006.05.113

   Kociba. R: Frauson. L: Humiston. C: Norris. J: Wade. C: Lisowe. R: Quasi 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. PJ: 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=OTSOOO 1103
March 2013                                        H-79                    DRAFT - DO NOT CITE OR QUOTE

-------
   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

   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 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

   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 retardants sub stance 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: Covad A: Herzke. D: Lepom. P: Morris. S: Tronczynskj 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

   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/si.ies.7500572
March 2013                                        H-80                    DRAFT - DO NOT CITE OR QUOTE

-------
   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

   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.2860120111

   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. DrugMetab 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: Ah. 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

   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.0rg/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. 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.

   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

   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 and B6C3F1 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
March 2013                                        H-81                    DRAFT - DO NOT CITE OR QUOTE

-------
   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.
      EnvironInt 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.

   Oia 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.

   Qiu. YW: Zhang. G: Guo. LL: Zheng. GJ: Cai SQ. (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: Kites. 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: Markowskj 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

   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 micro nucleus 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
March 2013                                        H-82                    DRAFT - DO NOT CITE OR QUOTE

-------
   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. Xenobiotica33: 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/i.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 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.

   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

   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.
      http://dx.doi.org/10.1289/ehp. 1104802

   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.
      http://dx.doi.org/10.1289/ehp.11807

   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/j.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: Ann. 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.
March 2013                                         H-83                    DRAFT - DO NOT CITE OR QUOTE

-------
   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/i.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 [EPAReport]. (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. http://www.epa.gov/nrmrl/lcaccess/pdfs/600r06060.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) [EPAReport]. (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

   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/i.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  inBDNF, 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
March 2013                                        H-84                    DRAFT - DO NOT CITE OR QUOTE

-------
   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
      Environ400: 425-436. http://dx.doi.0rg/10.1016/i.scitotenv.2008.05.003

   Vrkoslavova. J: Demnerova. K: Mackova. M: Zemanova. T: Macek. T: Hajslova. 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: Wang. J: 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. (2010b). 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

   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

   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
March 2013                                         H-85                    DRAFT - DO NOT CITE OR QUOTE

-------
   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
March 2013                                        H-86                   DRAFT - DO NOT CITE OR QUOTE

-------
      Appendix I. External Review Draft
      Comments and Agency Responses
2

3
  March 2013                         DRAFT-DO NOT CITE OR QUOTE

-------
        Appendix  I.  External Review Draft Comments
                                          and Agency  Responses


       1.1.  Background
 1          The External Review Draft of this case study (U.S. EPA. 2012a) served as the starting point for
 2    identifying and prioritizing research gaps that, if pursued, could inform future assessments and
 3    subsequent risk management decisions for multiwalled carbon nanotubes (MWCNTs) in flame-retardant
 4    upholstery textiles or similar materials and applications. As discussed in more detail in Chapter 1, the
 5    draft was the basis for the collective judgment step of the comprehensive environmental assessment
 6    (CEA) process (see Figure 1-2). in which experts read the case  study document, participated in an online
 7    data prioritization exercise, and (for a subset of experts) attended a workshop. The collective judgment
 8    step resulted in the identification of elements and risk relevant factors of the CEA framework  as priority
 9    areas for future assessment or research due to data gaps and importance to risk management. These areas
10    are hereafter referred to as "Priority Research Areas." Some of these areas are most relevant to
11    individuals who plan research. These areas are those that the experts rated as important to consider in risk
12    assessments but in which they were not confident the available  data could support risk management
13    decisions. Other priority areas are more relevant to individuals  who develop assessments or are
14    responsible for risk management efforts (e.g., researching which type of risk management plan would be
15    most suitable given current information); those areas are the ones experts rated as important to consider in
16    risk assessments and in which they had greater confidence that  available data might support risk
17    management decisions (see Section  1.1.3). Notably, in applying the CEA approach to MWCNTs, the
18    majority of priority areas that emerged are most pertinent to research planning rather than developing
19    assessments. Throughout the revision process for the case study, efforts were made to streamline the
20    document so that it would clearly reflect each priority that emerged from the CEA collective judgment
21    step, input from public comments, and the opinions of expert stakeholders involved in prioritizing the
22    research gaps.
23          This appendix documents how the External Review Draft of the case study was revised to:
24    (1) respond to public comments and input from experts participating in the collective judgment step of the
25    CEA process (Figure 1-2). and (2) reflect the priorities identified through the CEA collective
26    judgment step.
      March 2013                                  1-1                 DRAFT - DO NOT CITE OR QUOTE

-------
      1.1.1. Appendix Development Process
 1           As part of the collective judgment step of the CEA process (Figure 1-2) described in
 2    Section 1.1.3. 23 experts provided written responses to the following charge questions:
 3           1.   Do you know of additional, specific studies on MWCNTs that should be included in the case
 4               study to help identify data gaps that are important to support future assessment and risk
 5               management efforts for MWCNTs in flame-retardant textile coatings?
 6           1.   Is the science accurately conveyed throughout the document? If not, please list any areas that
 7               need improvement and provide specific comments in the text to highlight areas that should be
 8               refined.
 9           2.   Does the comparison of decaBDE and MWCNTs in the case study document help to identify
10               research gaps to support future assessments and risk management decisions for MWCNTs? If
11               not, please briefly explain.
12           3.   Do you have any specific comments on how this document could be improved?
13           One of the 23 experts also provided a PDF copy of the case study with free-form comments (i.e.,
14    not specific to a charge question) linked to specific regions of text.  In addition, four members of the
15    public provided input during the public comment period announced in a July 2, 2012 Federal Register
16    Notice.33 Finally, an interagency commenter provided input on the  draft document during the public
17    comment period. Affiliations of the experts are provided in Table 1-1; affiliations of the interagency and
18    public commenters are provided in Table 1-2.
19           All charge question responses, public comments, and expert free-form comments received on the
20    External Review Draft were tracked with the aid of an Excel-based comment tracking sheet. Longer
21    responses and public comments were broken down into distinct, individual thoughts and assigned unique
22    comment numbers. Each unique comment was assigned to the most relevant portion of the detailed CEA
23    framework (Figure 1-3), which was also used for the online collective judgment prioritization exercises of
24    the CEA process.34
25           Comments then were assigned "themes" based on recurring topics so that similar comments
26    related to the same CEA framework area could be grouped and comments with similar concepts that
27    crossed multiple CEA framework areas also could be grouped. Themes were determined progressively
28    and the list of themes  was reconsidered and revised as more comments were considered, to develop a
29    limited number of themes that created broad categories. Multiple themes were often, but not always,
      33http://www.gpo.gov/fdsvs/pkg/FR-2012-07-02/html/2012-16137.htm
      34Note 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.
      March 2013                                     I-2                   DRAFT - DO NOT CITE OR QUOTE

-------
1     applied to a single comment to either increase the degree of specificity of the theme (e.g., a general
2     comment stating that the comparison between decaBDE and MWCNTs was not useful compared to a
3     comment that specified why the comparison was not useful) or to account for multiple unrelated themes in
4     the same comment (e.g., a comment noting that information on release rates and exposure are data gaps
5     and noted the need for better analytical techniques). The themes  provided a flexible and inclusive  method
6     for grouping similar 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
      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.
      March 2013
     I-3
DRAFT - DO NOT CITE OR QUOTE

-------
      Table 1-2.  Public Commentators 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

 1           Finally, Agency responses were drafted for each comment using consistent language where

 2    possible to connect comments from multiple commenters that expressed the same basic ideas and themes.

 3    Responses were focused on applying one of the five main categories of action taken, as described in detail

 4    in Chapter 1 and summarized in Table 1-3. In some cases, more than one response category was

 5    appropriate for the same comment or group of comments. This is particularly true for comments that

 6    applied to broader themes or CEA framework areas. The "Agency Response" includes references to

 7    multiple categories (e.g., addition of Information Highlight Boxes as well as in-text edits) as necessary to

 8    describe the complete actions that were taken in response to each comment or group of comments.
      1.1.2. Appendix Organization
 9           This appendix contains three types of comment tables:

10           1.   Table 1-4. Table 1-5. Table 1-6. Table 1-7. Table 1-8. and Table 1-9: Response tables for
11               comments that were addressed in the body of the document based on the first four response
12               types described in Table 1-2 (this includes some cases where the comment was acknowledged
13               but no major action was taken because the related CEA area was not identified as important
14               by the collective judgment process), organized by relevant chapters of the case study
15               document;
16           2.   Table 1-11: Response table for comments that were not specifically addressed through
17               revisions or edits to the main body of the case study because the comment did not suggest
18               that any action needed to be taken; and
19           3.   Table 1-12. Table 1-13. and Table 1-14: Look-up tables for the comment IDs listed in the
20               response tables, organized by the way in which the  comment was submitted (i.e., expert
21               charge question responses, expert free-form comments, and public comments). Original
22               commenter text excerpts are provided, along with the commenter and comment ID numbers,
23               and the theme or themes applied to each comment. Note that while commenter ID numbers
24               are provided in Table 1-2 above for public commentators, they are not included in Table 1-1.
25               to preserve anonymity of expert input.
      March 2013                                     I-4                   DRAFT - DO NOT CITE OR QUOTE

-------
1             As noted previously, comments were grouped according to similar themes, so several unique

2     comments are presented as relevant to a single Agency response, as indicated by the ID numbers in the far

3     right column of the response tables. The comments have been collectively summarized to provide readers

4     with a relatively quick overview of the common theme among the comments and to illustrate how the

5     group of comments relate to the action taken by the Agency. As noted above, the original individual

6     comments and assigned themes are provided in the final tables of this appendix (Table 1-12. Table 1-13.

7     and Table 1-14).
      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 of CEA
             Added Priority Research
             Area Highlight Boxes
              Moved unprioritized
              sections to Appendix
                                   Added final Priority
                                   Focus Section 6.3
                  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 gray 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.

                  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"
                                   Added new figures and
                                   tables
                  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.

                  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.
      March 2013
                                     I-5
                                           DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-3, cont.:. Agency response categories
 Response
 Category
Purpose
Actions Taken
Notes
            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.
            Streamline
            document to
            support MWCNT
            research planning
            in this application of
            CEA
              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
              Comparison Boxes" to
              main text
                   A series of DecaBDE Comparison Text Boxes was added to the
                   body of the document (outlined in green) to illustrate 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.
March 2013
                                     I-6
                                            DRAFT - DO NOT CITE OR QUOTE

-------
 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
Response
Category
 Related to
  Global
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 comparison
with decaBDE to be useful, but
suggested that segregating the
information, rather than alternating
back and forth, would improve the
document.
Discussion of decaBDE was moved to Appendix H to provide continuity in the
presentation of information on decaBDE and MWCNT; efforts were made to succinctly
highlight how understanding decaBDE data might inform research planning for MWCNT
priority topics by adding "DecaBDE Comparison Text Boxes."
               73
                    March 2013
                                                       I-7
                                             DRAFT - DO NOT CITE OR QUOTE

-------
Table 14, cont: Relevant to the general case study or multiple sections of the case study.
 Related to CEA
Framework Area
         Comment Theme
                                 Response
Response
Category
  Related to
   Global
 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 1.1.4). 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
                     March 2013
                                                                                   DRAFT - DO NOT CITE OR QUOTE

-------
Table 14, cont: Relevant to the general case study or multiple sections of the case study.
 Related to CEA
Framework Area
         Comment Theme
                                  Response
Response
Category
 Related to
   Global
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/
Chapter 6
One comment questioned whether the
case study could be used to
accomplish more than its intended
purpose of identifying and prioritizing
research gaps.
The purpose of the draft case study document and its relationship to any future
regulatory decisions for MWCNT was clarified in Chapter 1 (see Section 1.3) and
expanded upon in Chapter 6 (see Section 6.3).
                 35
                      March 2013
                                                            I-9
                                                  DRAFT - DO NOT CITE OR QUOTE

-------
Table 14, cont: Relevant to the general case study or multiple sections of the case study.
 Related to CEA
Framework Area
         Comment Theme
                                  Response
Response
Category
  Related to
   Global
 Comment ID
Chapter 1/
Chapter 6
One comment suggested an OECD
(2012) publication regarding important
issues in the risk assessment of
nanomaterials. Another suggested
this reference to describe a predictive
toxicological paradigm for the
assessment of nanomaterials.
The purpose of the draft case study document and its relationship to any future
regulatory decisions for MWCNT was clarified in Chapter 1 (see Section 1.3) and
expanded upon in Chapter 6 (see Section 6.3). In particular, the OECD (2012)
publication was used to highlight research needs and data gaps in the assessment of
nanomaterials in Section 6.3.
              154,146
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-
  1,2,3
 6,213,214,
     216
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.
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
   1,2
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.
40,41,42,43,
44, 45, 46, 47,
48, 49, 50, 51
                      March 2013
                                                          1-10
                                                 DRAFT - DO NOT CITE OR QUOTE

-------
Table 14, cont: Relevant to the general case study or multiple sections of the case study.
 Related to CEA
Framework Area
         Comment Theme
                                  Response
Response
Category
 Related to
   Global
Comment ID
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
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
                      March 2013
                                                          1-11
                                                 DRAFT - DO NOT CITE OR QUOTE

-------
Table 14, cont: Relevant to the general case study or multiple sections of the case study.
Related to CEA
Framework Area
Dose:
Human:
Absorption
AND
Impacts:
Human:
Cancer,
Non-Cancer
Chapter 4/
Chapter 6
Chapter 4/
Chapter 6
Comment Theme
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.
One comment questioned whether
information from two sources
regarding decaBDE use in
cars/aircraft was accurate.
One comment questioned whether
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.
Response
Additional Information Hiqhliqht 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 Non-cancer" 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 .

The cited literature was reviewed and accuracy of the statement was verified. Additional
sources corroborating the statement were added.
Information regarding the use of MWCNTs in combination with other
chemicals/materials was included in Additional Information Hiqhliqht 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 Hiqhliqht Box 2.
Response
Category
1,2,3
3
2,3
Related to
Global
Comment ID
64, 122, 135
229
244
                 March 2013
1-12
DRAFT - DO NOT CITE OR QUOTE

-------
Table 14, cont: Relevant to the general case study or multiple sections of the case study.
Related to CEA
Framework Area
Section 4.2.51
Section 5. 1.7
Multiple DecaBDE
Sections
Comment Theme
Several additional references were
provided to improve the discussion of
decaBDE in the case study.
Two comments recommended
additional information and sources
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.
Response
Information on decaBDE is now primarily in Appendix H; however, the provided
references were reviewed and incorporated into the text as appropriate.
Basic information regarding decaBDE was reviewed for accuracy and edited as needed.
The document currently defines the water solubility of decaBDE as <0.1 UQ/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.
Response
Category
3
3
Related to
Global
Comment ID
2,23
1,24,82
                 March 2013
1-13
DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-5.     Relevant to Chapter 1 (including Preface and Executive Summary).
 Related to CEA
Framework Area
         Comment Theme
                                  Response
Response
Category
 Related to
   Global
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
Summary/
Chapter 1
One comment suggested there may
be more recently published 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 fire 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 fire-
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.
            8, 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
                      March 2013
                                                          1-14
                                                 DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-5, cont: Relevant to Chapter 1 (including Preface and Executive Summary).
 Related to CEA
Framework Area
         Comment Theme
                                  Response
Response
Category
 Related to
   Global
Comment ID
Chapter 1
Two comments noted that use of
MWCNTs in fire retardant will be
determined by their ability to pass
specific fire-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 fire-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 fire 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 fire-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
                      March 2013
                                                           1-15
                                                 DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-5, cont: Relevant to Chapter 1 (including Preface and Executive Summary).
Related to CEA
Framework Area
Chapter 1
Chapter 1
Comment Theme
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.
One comment noted that chirality
might not be a consideration for
MWCNTs. A reference was provided.
Response
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 fire 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.

The Agency appreciates the suggestion but the source used for information on chirality,
Gustavsson et al. (2011) mentions specifically the variation for MWCNTs. The literature
provided by the commenter is for double-walled carbon nanotubes and was therefore
not included in the case study.
Response
Category
2,3
5
Related to
Global
Comment ID
7,9,11
163
                 March 2013
1-16
DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-6.     Relevant to Chapter 2.
 Related to CEA
Framework Area
         Comment Theme
                                  Response
Response
Category
 Related to
   Global
Comment ID
Product Life
Cycle: Raw
Materials; Product
Life Cycle:
Material
Synthesis
Two comments suggested additional
literature might be available to fill in
data gaps pertaining to MWCNT
feedstocks and replace the surrogate
SWCNT 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 (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
                      March 2013
                                                           1-17
                                                 DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-6, cont: Relevant to Chapter 2.
 Related to CEA
Framework Area
         Comment Theme
                                  Response
Response
Category
 Related to
   Global
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"
   1,3
     81
                                                    during end-of-life stages of the life-cycle process.
                     March 2013
                                                          1-18
                                                 DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-7.    Relevant to Chapter 3.
Related to CEA
Framework Area
Env. TT&F
(Analytical
Techniques)
Env TT&F
Chapter 1 or 3
Env TF&T: Soil
Comment Theme
Two comments provided references
pertaining to analytical techniques for
MWCNT in environmental media and
impacts of release.
One comment noted a data gap
regarding environmental
decomposition and transformation of
MWCNTs
Eight comments discussed MWCNT
release, transformation, and fate in
the environment in terms of potential
variation in polymer chemistry;
references were also provided.
Three comments noted an instance
where the authors of a reference
appeared to be speculating rather
than stating definitive evidence.
Response
Analytical techniques were discussed by workshop participants as related to several
research priorities within environmental transport, transformation, and fate. As such, the
suaaested literature was incorporated into the related sections of Section 6.3.2 or in
Additional Information Hiqhliqht Box 10, which was added to describe the weaknesses
of current analytical techniques as relevant.
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 hiqhliqhted in the discussion of these priority areas in Section 6.3.2.

Additional Information Hiqhliqht Box 6 and Additional Information Hiqhliqht 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.

The statements were revised and clarified.
Response
Category
1,2
1
2
3
Related to
Global
Comment ID
128, 139
201
127, 219, 220,
221,222,225,
226, 228
168, 169, 170
                  March 2013
1-19
DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-7, cont: Relevant to Chapter 3.
 Related to CEA
Framework Area
         Comment Theme
                                  Response
Response
Category
 Related to
   Global
Comment ID
EnvTT&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
   2,3
 21,65,66
                                                     physicochemical properties of MWCNTs affect release, environmental fate, exposure,
                                                     and toxicity.
EnvTT&F:
Soil/Sediment:
Bioavailability
One comment suggested literature
pertaining to soil/sediment 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
(Analytical
Techniques)
One comment suggested literature
relevant to analytical 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
EnvTT&F
One comment noted the use of lipophilicity
and hydrophobicity was not always clear
and that the sources used might be
outdated.
The distinction between hydrophobicity and lipophilicity was identified as an important
research topic by workshop participants. Text was clarified as necessary throughout the
document.
                164
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
Note: Env TT&F = Environmental Transport, Transformation, and Fate
                      March 2013
                                                           I-20
                                                  DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-8.     Relevant to Chapter 4.
 Related to CEA
Framework Area
         Comment Theme
                                  Response
Response
Category
 Related to
   Global
Comment ID
Exposure Route:
Human:
Occupational
One comment noted occupational
exposure is a data gap for 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.
              38,152
Exposure Route:
Human-Consumer
One comment noted a reference to
elucidate potential 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-Ti02 was included in
Additional Information Highlight Box 11 along with information (as available) from a
  1,2,3
    110
                                                    targeted literature search to present the most accurate and current information on
                                                    dermal absorption.
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
                      March 2013
                                                           1-21
                                                 DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-8, cont: Relevant to Chapter 4.
 Related to CEA
Framework Area
         Comment Theme
                                 Response
Response
Category
 Related to
   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.
           174,175,177,
                172
Dose: Human:
Absorption
Three comments suggested 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.
             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):
Terrestrial Biota:
Absorption
One comment noted that a reference
describing 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.
            150,151,176
                     March 2013
                                                          I-22
                                                DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-8, cont: Relevant to Chapter 4.
 Related to CEA
Framework Area
         Comment Theme
                                 Response
Response
Category
 Related to
   Global
Comment ID
Section 4.3
Two comments noted potential errors
in transcription from primary sources.
The statements were revised and clarified.
              173,177
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
Biota: Absorption
One comment noted that the
discussion of potential 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
                     March 2013
                                                         I-23
                                                DRAFT - DO NOT CITE OR QUOTE

-------
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 et al. (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:
Non-cancer
Four comments reiterated the
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
                      March 2013
                                                           I-24
                                                 DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-9, cont: Relevant to Chapter 5.
Related to CEA
Framework Area
Impacts: Human
Section 5.2.2.1
Impacts: Human
Impacts: Human
Comment Theme
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
MWCNT assessment in the future.
Two comments recommended
references pertaining to accumulation
in terrestrial biota.
One comment noted the results found
in intratracheal instillation studies in
mice corroborated those reported via
inhalation.
Ten comments questioned the
applicability of traditional in vivo
toxicological (particularly inhalation)
models for MWCNTs; several
references were provided noting
potential weaknesses/confounding
factors.
Response
"Impacts: Human" was identified by workshop participants as a Priority Research Area.
Additional Information Hiqhliqht Box 15 was added to hiqhliqht the trend toward
developing innovative biologically/toxicologically relevant in vitro models.
The suggested literature was reviewed and incorporated as deemed relevant into
Additional Information Hiqhliqht Box G2 in Appendix G, which discusses toxicity to
terrestrial invertebrates.
Four studies on intratracheal instillation are currently included in Section 5.1 .3 alonq 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 Hiqhliqht Box 14 was added, which
discusses inhalation study designs for MWCNTs.
"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 Hiqhliqht Box 14.
Additionally, Additional Information Hiqhliqht Box 15 was added regarding the trend
toward the development of innovative toxicologically/biologically relevant in vitro models.
Response
Category
2
2
2
1,2,3
Related to
Global
Comment ID
275, 276, 277,
278, 279, 280,
281,282,283,
284, 285, 286,
287, 288, 289,
290, 291
182, 183
37
259, 266, 267,
268, 269, 270,
271,272,273,
274
                  March 2013
I-25
DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-9, cont: Relevant to Chapter 5.
 Related to CEA
Framework Area
         Comment Theme
                                  Response
Response
Category
 Related to
   Global
Comment ID
Impacts: Human:
Non-Cancer
One comment noted respiratory
sensitization as an example of 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: Non-cancer"
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. Non-cancer 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
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
                      March 2013
                                                           I-26
                                                 DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-10.   Relevant to Chapter 6 and appendices.
Related to CEA
Framework Area
Chapter 6
Impacts:
Human:
Cancer,
Non-cancer
Appendix J
Comment Theme
One comment suggested the usefulness of highlighting
the information leading to voluntary phase-out of
decaBDE.
Three comments identified additional literature
pertaining to SWCNT that might be useful as surrogate
or supportive data for MWCNTs. One comment stated
that major studies were presented already, and that the
additional SWCNT studies would not be relevant.
One commenter felt summary tables and figures
contained too much information to be useful as
summary or highlight text.
Response
Text was revised in Section 1 .1 .4 to further emphasis the type of
information considered in decisions surrounding the voluntary phase-
out of decaBDE.
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.
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
more clearly convey information in tables and text.
Response
Category
3
3
3
Related to
Global
Comment ID
74
96,116,118,
119
149
                 March 2013
I-27
DRAFT - DO NOT CITE OR QUOTE

-------
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.
                     March 2013
                                                         I-28
                        DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-11, cont.: 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
EnvTT&F:Soil
Two comments suggested additional literature pertaining to
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
EnvTT&RSoil:
bioavailability
One commenter called out a section of the preface which they felt
was not fully representative of the body of science. A second
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
                     March 2013
                                                         I-29
                        DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-11, cont.: Comments that required no action.
Related to CEA
Framework Area
EnvTT&F
(Aqueous media)
None
None
None
Comment Theme
One comment provided a reference pertaining to MWCNT in
aqueous media.
16 comments were received that did not require agency action.
12 commenters stated that there were no additional literature they
were aware of that should be included in the case study.
10 commenters stated that they did not have any comments on
improving the document, or that their comments for improvements
were incorporated into the other charge question responses.
Agency Response
The suggested reference was previously included in the
case study in Section 3.3. 1 .

The Agency appreciates the feedback.
The Agency appreciates the feedback.
The Agency appreciates the feedback.
Related to Comment ID
143
3,4,20,194,198,208,231,
232,236,237,241,245,
249, 255, 296
25,31,57,76,80,88,100,
104, 108, 133, 159, 188
5, 95, 187, 70, 87, 107, 132,
142, 158, 191
                 March 2013
I-30
DRAFT - DO NOT CITE OR QUOTE

-------
 1.3. Full Comment Excerpts
Table 1-12.   Comments received in response to the expert charge questions.
Comment
ID
25
26
27
28
29
30
31
32
33
34
Author
ID
E1
E1
E1
E1
E1
E1
E2
E2
E2
E2
Charge
Question
1
2
3
3
4
4
1
2
3
3
Comment Text Excerpt
I think this report was thorough and included all available literature on MWCNTs.
I believe that the science was presented in an objective and clear manner.
Further, I believe this comparison made things confusing, since it is impossible to compare the fate and
transport and effects of two types of compounds that are completely different in their chemistry and
physical properties. So what has been found for flame retardants in relation to their movement in the
environment, half-life, uptake by biota, and ultimately effects, is rather meaningless when applied to
MWCNTs.
No, this comparison was not useful. I am not sure why this was done, but have the impression that it was
used to "beef" up the document since so little information exists on MWCNTs.
Remove the flame retardant vs. MWCNTs comparison. As already mentioned, this comparison doesn't
help when trying to elucidate research gaps as related to MWCNTs.
Also, the use of MWCNTs as flame retardants has not been materialized, so why solely focus on this
particular application?
At this point in time, I must admit that I cannot think of any studies that might help and I find the Draft Case
Study Document to be very comprehensive
To the best of my knowledge the science is accurately conveyed
I do think that the comparison makes sense that it is good to have facts and data on decaBDE to measure
information up against.
My only concern is that decaBDE is a problematic substance and bans and limitations have been put on
its use and hence comparing MWCNTs with such a problematic substance would inevitably make
MWCNTs look good
Themes
No additional literature
Science accurate
Comparison NOT useful;
decaBDE was bad comparison
choice; fundamentally different
compounds don't overlap
Comparison NOT useful; added
length
Comparison NOT useful;
highlighted data gaps
Choice of nanomaterial and
application; application not
currently in use
No additional literature
Science accurate
Comparison WAS useful;
highlighted data gaps
Comparison NOT useful;
decaBDE was bad comparison
choice
              March 2013
1-31
DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-12, cont: Comments received in response to the expert charge questions.
Comment
ID
35
36
37
38
39
40
41
42
Author
ID
E2
E3
E3
E3
E3
E4
E4
E4
Charge
Question
4
1
2
3
4
1
1
1
Comment Text Excerpt
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.
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.
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.
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.
See comment to # 3 regarding focus of document. It would have been informative if the authors of the
document would have listed at the end of each Chapter their perspective as to research data gaps.
"Synergistic effect of carbon nanotubes and decabromodiphenyl oxide/Sb203 in improving the flame
retardancy of polystyrene" Lu, H.; Wilkie, C. A. Polym. Degrad. Stab. 2010, 95, 564-571.
"Role of Surface Interactions in the Synergizing Polymer/Clay Flame Retardant Properties" Pack, S.;
Kashiwagi, T.; Cao, C.; Korach, C. S.; Lewin, M.; Rafailovich, M. H. Macromolecules2010, 43, 5338-5351.
"Segregation of Carbon Nanotubes/Organoclays Rendering Polymer Blends Self-Extinguishing" Pack, S.;
Kashiwagi, T.; Stemp, D.; Koo, J.; Si, M.; Sokolov, J. C.; Rafailovich, M. H. Macromolecules 2009, 42,
6698-6709.
Themes
CEA framework/ methodology;
purpose of CEA
No additional literature
Highlight data gap; additional
considerations for section
Highlight data gap
Highlight data gap; suggestion for
improvement
Choice of nanomaterial and
application; alternative products/
formulations
Choice of nanomaterial and
application; alternative products/
formulations
Choice of nanomaterial and
application; alternative products/
formulations
                 March 2013
I-32
DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-12, cont: Comments received in response to the expert charge questions.
Comment
ID
43
44
45
46
47
48
49
50
51
Author
ID
E4
E4
E4
E4
E4
E4
E4
E4
E4
Charge
Question
1
1
1
1
1
1
1
1
1
Comment Text Excerpt
"Nanoclay and carbon nanotubes as potential synergists of an organophosphorus flame retardant in
poly(methyl methacrylate)" Isitman, NihatAli; Kaynak, Cevdet Polym. Degrad. Stab. 2010, 95, 1523-
1532
"Layered silicate polymer nanocomposites: new approach or illusion for fire retardancy? Investigations of
the potentials and the tasks using a model system" Bartholmai, M.; Schartel, B. Polymers for Advanced
Technologies 2004, 15, 355-364.
"Filler blend of carbon nanotubes and organoclays with improved char as a new flame retardant system for
polymers and cable applications" Beyer, G. Fire Mater. 2005, 29, 61-69.
"Flame retardancy of nanocomposites based on organoclays and carbon nanotubes with aluminum
trihydrate" Beyer, G. Polym. Adv. Technol. 2006, 17, 218-225.
"Fire behaviour of polyamide 6/multiwall carbon nanotube nanocomposites" Schartel, B.; Potschke, P.;
Knoll, U.; Abdel-Goad, M. European Polymer Journal 2005, 41, 1061-1070.
"Some comments on the main fire retardancy mechanisms in polymer nanocomposites" Schartel, B.;
Bartholmai, M.; Knoll, U. Polym. Adv. Technol. 2006, 17, 772-777.
"Flame retarded polymer layered silicate nanocomposites: a review of commercial and open literature
systems" Morgan, A. B. Polym. Adv. Technol. 2006, 17, 206-217.
"Flammability reduction of flexible polyurethane foams via carbon nanofiber network formation"
Zammarano, M.; Kramer, R. H.; Harris, R.; Ohlemiller, T. J.; Shields, J. R.; Rahatekar, S. S.; Lacerda, S.;
Gilman, J. W. Polym. Adv. Technol. 2008, 19, 588-595.
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.
Themes
Choice of nanomaterial and
application; alternative products/
formulations
General information on flame
retardancy of nanocomposites
Choice of nanomaterial and
application; alternative products/
formulations
Choice of nanomaterial and
application; alternative products/
formulations
General information on flame
retardancy of nanocomposites
General information on flame
retardancy of nanocomposites
General information on flame
retardancy of nanocomposites
General information on flame
retardancy of nanocomposites
Choice of nanomaterial and
application; alternative products/
formulations; application not
currently in use
                 March 2013
I-33
DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-12, cont: Comments received in response to the expert charge questions.
Comment
ID
52
53
54
55
56
57
58
59
Author
ID
E4
E4
E3
E4
E4
E5
E5
E4
Charge
Question
2
2
3
3
4
1
2
3
Comment Text Excerpt
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, ...
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.
Most, if not all, of MWCNT use in the United States is in research laboratories or small scale pilot
manufacturing processes [Schubauer-Berigan etal. 2011: Engineered carbonaceous nanomaterials
manufacturers in the United States: workforce size, characteristics, and feasibility of epidemiologic
studies. J Occup Environ Med53(Suppl6):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.
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....
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.
No, I am unaware of any other studies.
Yes, the science is conveyed accurately throughout the document.
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.
Themes
Science accurate
Choice of nanomaterial and
application ;application not
currently in use
Choice of nanomaterial and
application; application not
currently in use
Comparison WAS useful; impacts;
exposure
Choice of nanomaterial and
application ;suggestion for
improvement
No additional literature
Science accurate
Comparison NOT useful; choice
of nanomaterial and application;
application not currently in use
                 March 2013
I-34
DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-12, cont: Comments received in response to the expert charge questions.
Comment
ID
60
61
62
63
64
65
66
67
68
69
Author
ID
E5
E5
E5
E6
E6
E6
E6
E6
E5
E6
Charge
Question
3
4
4
1
1
1
1
2
3
3
Comment Text Excerpt
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.
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.
As mentioned in the previous response, the document did a wonderful job pointing out the data gaps for
MWCNTs but it seemed that the document focused more on decaBDE due to the excess of information
available for the chemical. The decaBDE data could be distilled down a bit to shorten the document.
"Evaluation of the interactions between multiwalled carbon nanotubes and CACO-2 cells," by Clark, KA et
al. DO1 10.1080/15287394.2011.589105 (relevant to Chapter4 and/or 5)
"Cell permeability, migration, and reactive oxygen species induced by multiwalled carbon nanotubes in
human microvascular endothelial cells," by Pacurari, M etal. DO1 10.1080/15287394.2012.625549
(relevant to Chapter 4 and 5)
"Impact of Porous Media Grain Size on the Transport of Multi-walled Carbon Nanotubes," by Mattison, NJ
et al, DOI 10.1021/es2017076 (relevant to Chapters 2 and 3)
"Sorption of Peat Humic Acids to Multi-Walled Carbon Nanotubes," by Wang, XL et al. DOI
10.1021/es202258q (relevant to Chapter 3)
I was impressed by the thoroughness and accuracy of the science presented in the document.
Yes and no ... No, because I felt like the information about decaBDE started to overtake the document
because there are not significant amounts of information available for MWCNTs for all the scenarios
covered in the case study.
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.
Themes
Comparison WAS useful;
highlighted data gaps
Added length
Comparison NOT useful;
highlighted data gaps; decaBDE
focus
Additional considerations for
section
Additional considerations for
section
Additional considerations for
section
Additional considerations for
section
Science accurate
Comparison NOT useful;
decaBDE focus
Comparison WAS useful;
fundamentally different
compounds don't overlap
                 March 2013
I-35
DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-12, cont: Comments received in response to the expert charge questions.
 Comment
    ID
Author
  ID
 Charge
Question
Comment Text Excerpt
Themes
    70
  E6
          I don't have any specific comments on improving the document.
    71
  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: 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
          +
          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-Ce02 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-Ce02, plant growth and yield diminished, but also (iii) nitrogen
          fixation—a major ecosystem service of leguminous crops—was shut down at high nano-Ce02
          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.
                                                          Recent literature published;
                                                          additional considerations for
                                                          section
                     March 2013
                                                               I-36
                                                                               DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-12, cont: Comments received in response to the expert charge questions.
 Comment
    ID
Author
  ID
 Charge
Question
                                   Comment Text Excerpt
           Themes
    72
  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 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.

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,1X. J. Peng,2Z. 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.
Recent literature published;
additional considerations for
section
    73
  E7
          The document seems to be very well done, thorough, and comprehensive. The organization of the
          document—alternating between decaBDE and MWCNT is discordant and a challenge to read when trying
          to identify the issues for MWCNT-polymers as flame retardants. These two materials are not related,
          except to provide a function. Having the information for decaBDE available is valuable to 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.
                                                                                             Comparison WAS useful;
                                                                                             fundamentally different
                                                                                             compounds don't overlap
                     March 2013
                                                               I-37
                                                                               DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-12, cont: Comments received in response to the expert charge questions.
Comment
ID
74
75
76
77
78
79
80
81
82
Author
ID
E7
E7
E8
E8
E8
E8
E9
E9
E9
Charge
Question
3
4
1
2
3
4
1
2
2
Comment Text Excerpt
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.
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.
no
yes
Yes, and the comparative assessment is helpful.
Link (at least conceptually) to Value of Information analysis.
No, I don't.
In Fig2-1 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.
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.
Themes
Comparison NOT useful;
suggestion for improvement
Choice of nanomaterial and
application
No additional literature
Science accurate
Comparison WAS useful
Highlight data gap; suggestion for
improvement
No additional literature
-
-
                 March 2013
I-38
DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-12, cont: Comments received in response to the expert charge questions.
Comment
ID
83
84
85
86
87
88
89
90
91
Author
ID
E9
E9
E9
E9
E9
E10
E10
E10
E10
Charge
Question
2
2
2
3
4
1
2
3
4
Comment Text Excerpt
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.
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 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.
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.
no comment
I don't know of any additional studies that would be appropriate.
I thought that the science was accurately portrayed. Although, the Chapter on characterization was not
complete.
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.)
A technical summary would have been useful.
Themes

-
—
Comparison WAS useful
-
No additional literature
Science accurate
Comparison NOT useful; added
length; fundamentally different
compounds don't overlap
Suggestion for improvement
                 March 2013
I-39
DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-12, cont: Comments received in response to the expert charge questions.
Comment
ID
92
93
94
95
96
97
98
99
100
101
102
Author
ID
E11
E11
E11
E11
E12
E12
E12
E12
E13
E13
E13
Charge
Question
1
2
3
4
1
2
3
4
1
2
3
Comment Text Excerpt
There are several new ES&T articles on CNT analysis - here is one Doudrick et al. (DOI:
10.1021/es300804f)
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.
yes
The framework is fine for pre-workshop. There are rapidly developing sciences around this issue which
should be included as references.
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.
Themes
Detection/measurement analytics;
additional considerations for
section
Science accurate; purpose of
CEA
Comparison WAS useful
Recent literature published;
purpose of CEA; suggestion for
improvement
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
                 March 2013
I-40
DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-12, cont: Comments received in response to the expert charge questions.
Comment
ID
103
104
105
106
107
108
109
110
111
Author
ID
E13
E14
E14
E14
E14
E15
E15
E15
E15
Charge
Question
4
1
2
3
4
1
2
2
2
Comment Text Excerpt
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.
a) No. I am not aware of any published or unpublished data.
I find the document heavily weighted toward supposition, e.g., exposure "could" occur; MWCNT "could" be
released. This language leaves the reader believing that such phenomena have already 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, Ti02 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.
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; Schinwaldetal, 2012).
Themes
Added length; suggestion for
improvement
No additional literature
Science accurate
Comparison WAS useful
-
No additional literature
Suggestion for improvement

--
                 March 2013
1-41
DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-12, cont: Comments received in response to the expert charge questions.
 Comment
    ID
Author
  ID
 Charge
Question
Comment Text Excerpt
Themes
    112
  E15
          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
           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
          b) Describing potential exposure scenarios is important, but should be followed by either identification of
          data gaps and research ideas, or prioritizing  those scenarios using the available information. For example,
          can information about release of CNTs from  other media be used to give a likelihood of release of CNTs
          from textiles? Using release of other substances is inappropriate.
                                                          Highlight data gap; suggestion for
                                                          improvement
    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
                     March 2013
                                                               I-42
                                                                               DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-12, cont: Comments received in response to the expert charge questions.
 Comment
    ID
Author
  ID
 Charge
Question
                                    Comment Text Excerpt
           Themes
    116
  E16
    1
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
          Oral Toxicity
          The report states there "no data was identified on the acute toxicity of MWCNT following oral or dermal
          exposure" however within the derivation of an DEL for Baytubes, Pauluhn (2010) reported that based on
          the OECD TG 423 (Acute Oral Toxicity - Acute Toxic Class Method) that the MWCNTs' tested (Baytubes)
          were not acutely toxic with an LDso-oral of>5000 mg/kg bw.
                                                                                              Additional considerations for
                                                                                              section
    118
  E16
          In a later study, the authors looked at these effects in relation to other particles and found that at equal
          dose, diesel exhaust particles generated larger levels of 8-oxodG in rat liver than carbon black did and
          exposure to fullerenes C60 and SWCNT were the least potent (M0ller et al 2012). Based on these
          interesting findings, the authors noted that the extent of translocation from the gut is largely unresolved but
          should be investigated further.
                                                                                              Additional considerations for
                                                                                              section
    119
  E16
          As well as the Kolosnjaj-Tabi et al 2010 study using SWCNT mentioned in the decaBDE MWCNT report,
          Folkmann et al (2009) investigated oxidative DNA 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
                      March 2013
                                                               I-43
                                                                                DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-12, cont: Comments received in response to the expert charge questions.
 Comment
    ID
Author
  ID
 Charge
Question
Comment Text Excerpt
Themes
    120
  E16
          Overall the science is portrayed comprehensively however there are a few incidences which could have
          been improved.
                                                          Science accurate
    121
  E16
          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
          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
          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
          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
          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.
                      March 2013
                                                               I-44
                                                                                DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-12, cont: Comments received in response to the expert charge questions.
Comment
ID
126
127
128
129
130
131
132
133
134
135
136
Author
ID
E17
E17
E17
E17
E17
E17
E17
E18
E18
E18
E18
Charge
Question
1
1
1
1
2
3
4
1
2
2
3
Comment Text Excerpt
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): 251 1-2518.
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.
von der Kammer F, Ferguson PL, Holden PA, Masion A, Rogers K, Klaine SJ, Koelmans AA, Home N,
Unrine JM. 2012. Analysis of Nanomaterials in Complex Matrices (Environment and Biota): General
Considerations and Conceptual Case Studies. Environmental Toxicology and Chemistry 31(1): 32-49.
A quick Web of Science lit review for the past year reveals a few more.
Yes
Yes
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.
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 toxicological
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).
Yes.
Themes
Additional considerations for
section
MWCNT physical/ chemical
properties
Detection/measurement analytics
Recent literature published
Science accurate
Comparison WAS useful
Praise for case study
No additional literature
Science accurate

Comparison WAS useful
                 March 2013
I-45
DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-12, cont: Comments received in response to the expert charge questions.
Comment
ID
137
138
139
140
141
142
143
144
Author
ID
E18
E19
E19
E19
E19
E19
E20
E20
Charge
Question
4
1
1
2
3
4
1
1
Comment Text Excerpt
No.
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. 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.
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.
The science and information is accurately communicated in the document. The use of tables and flow
charts is helpful and illustrative. In future documents; it may also be beneficial to include more
representative graphs and photographs from the literature to effectively portray the results generated.
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.
The document is well organized.
Natural Organic Matter Stabilizes Carbon Nanotubes in the Aqueous Phase
Hoon Hyung, John D. Fortner, Joseph B. Hughes, and Jae-Hong Kim
A Review of Carbon Nanotube Toxicity and Assessmentof Potential Occupational and Environmental
Health Risks
Chiu-wing Lam, John T. James, Richard McCluskey, Sivaram Arepalli, Robert L. Hunter
Themes
-
Recent literature published
Recent literature published
Science accurate; suggestion for
improvement
Comparison WAS useful;
highlighted data gaps
-
-
—
                 March 2013
I-46
DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-12, cont: Comments received in response to the expert charge questions.
Comment
ID
145
146
147
148
149
150
151
152
153
Author
ID
E20
E20
E20
E20
E20
E21
E21
E21
E21
Charge
Question
1
1
2
3
4
1
1
1
1
Comment Text Excerpt
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
A Predictive Toxicological Paradigm for the Safety Assessment of Nanomaterials. Huan Meng, Tian Xia,
Saji George, and Andre E. Nel
To my knowledge, the document conveys scientific aspect of our limited knowledge and related issues
about MWNT in fairly reasonable way.
It is believed that the comparison helps reviewers understand the underlying issues in MWCNT
applications and identify research gaps for exposure assessment and risk management decisions
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.
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.
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,
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
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.
Themes
-
-
Science accurate
Comparison WAS useful;
highlighted data gaps
-
~
~
~
~
                 March 2013
I-47
DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-12, cont: Comments received in response to the expert charge questions.
Comment
ID
154
155
156
157
158
159
160
161
162
Author
ID
E21
E21
E21
E21
E21
E22
E22
E22
E22
Charge
Question
1
2
3
3
4
1
2
2
2
Comment Text Excerpt
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.orq/officialdocuments/displavdocumentpdf /?cote=env/im/mono(2012)8&doclanquaqe=en

Yes it is well done.
Yes.
Another good reference for decaBDE is the ASTDR Toxicological Profile for Polybrominated Biphenyls and
Polybrominated Diphenyl Ethers at: http://www.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.
no
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
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.
Page xxi, lines 1-2 - Some published papers indicate that MWCNTS function similarly to hard carbons
decreasing MWCNT availability to organisms in soils and sediments (i.e., Petersen, E. J. et al. Environ. Sci.
Technol. 2009, 43 (11), 4181-4187.; Shen, etal. Environ. Toxicol. Chem. 2012, 31 (1), 202-209.).
Page xxii, lines 19-20 -Actually, numerous studies, probably >20, have been conducted on the effects and
uptake of MWCNTs to aquatic organisms.
Themes

Science accurate
Comparison WAS useful
DecaBDE data needs
improvement
-
~
Recent literature published;
suggestion for improvement
Highlight data gap; additional
considerations for section
Highlight data gap; suggestion for
improvement
                 March 2013
I-48
DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-12, cont: Comments received in response to the expert charge questions.
Comment
ID
163
164
165
166
167
168
169
170
171
Author
ID
E22
E22
E22
E22
E22
E22
E22
E22
E22
Charge
Question
2
2
2
2
2
2
2
2
2
Comment Text Excerpt
Table 1-9 - Chirality only is relevant for SWCNTs and the inner tube of DWCNTs (Yang, S. W.; Parks, A.
N.; Saba, S. A.; Ferguson, P. L; Liu, J., Photoluminescence from Inner Walls in Double-Walled Carbon
Nanotubes: Some Do, Some Do Not. Nano Letters 2011, 11, (10), 4405-4410.).
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.
Page 3-12, line 29 - The Holbrook et al. 2010 reference is also highly relevant here.
Page 3-14 - The Zhang et al. 201 1 reference is also relevant in this section since it discusses sorption of
MWCNTs onto peat, a soil component.
Page 3-14, line 17 - The review papers cited here about MWCNT adsorption onto soil surfaces were
published before sorption studies had been conducted. These articles are reviews that probably just
contained speculation along these lines, so this sentence could be deleted.
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.
Page 3-15, lines 1-2 - This sentence is just speculation and could be deleted.
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.
Themes
Recent literature published;
additional considerations for
section
Suggestion for improvement
Suggestion for improvement
Additional considerations for
section; suggestion for
improvement
Additional considerations for
section; suggestion for
improvement
Suggestion for improvement
Additional considerations for
section; suggestion for
improvement
Suggestion for improvement
~
                 March 2013
I-49
DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-12, cont: Comments received in response to the expert charge questions.
 Comment
    ID
Author
  ID
 Charge
Question
Comment Text Excerpt
Themes
    172
  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. etal. 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
          Page 4-49, line 6 - These values were BSAF values, not BAF values. This sentence should be edited
          accordingly.
                                                          Suggestion for improvement
    174
  E22
          Page 49, lines 18-20 - This sentence is entirely speculation. In studies of CNT uptake by a wide range of
          organisms (excluding plants), absorption into tissues is consistently minimal (see Table 2 of Petersen et
          al. 2011.ES&T pages 9837-9856).
                                                          Recent literature published;
                                                          highlighted data gaps
    175
  E22
          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
          impactbioaccumulation (Petersen etal., 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
          Page 4-50, lines 9-10 - This paper was published before data was collected on this topic and thus this
          speculation can be deleted. Moreover, being lipophilic is insufficient for expectations of accumulation in
          root lipids because transport into the cells and roots would need to occur first.
                                                          Suggestion for improvement
    177
  E22
          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
                     March 2013
                                                               I-50
                                                                               DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-12, cont: Comments received in response to the expert charge questions.
 Comment
    ID
Author
  ID
 Charge
Question
Comment Text Excerpt
Themes
    178
  E22
          Page 5-22 lines 22-23 - Actually, there are numerous studies on benthic invertebrates (i.e., Kennedy 2008
          cited on page R-12 and Shen et al. 2012 and Kennedy et al. 2009 cited in response to question 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.
                                                         Recent literature published
    179
  E22
          Page 5-22, lines 27-28 - The study cited has nothing to do with bioaccumulation in aquatic systems.
          Numerous papers have been conducted on exactly this topic and some of these studies were described in
          the previous Chapter.
                                                         Suggestion for improvement
    180
  E22
          Table 5 -2 - This table missing numerous important citations for MWCNTs as described in previous
          comments.
    181
  E22
          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
    182
  E22
          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
          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-AI203 plant findings.
                                                         Recent literature published;
                                                         suggestion for improvement
    184
  E22
          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
                     March 2013
                                                              1-51
                                                                              DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-12, cont: Comments received in response to the expert charge questions.
Comment
ID
185
186
187
188
189
190
191
Author
ID
E22
E22
E22
E23
E23
E23
E23
Charge
Question
3
3
4
1
2
3
4
Comment Text Excerpt
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.
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.
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.
No
Yes
Yes, the comparison of decaBDE and MWCNT in the case study document is useful in identifying
research gaps.
No
Themes
Comparison NOT useful;
suggestion for improvement
Comparison NOT useful; added
length

No additional literature
Science accurate
Comparison WAS useful
-
                 March 2013
I-52
DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-13.   Free-form comments received from experts.
Comment ID
192
193
194
195
196
197
198
199
200
Author
ID
E4-F
E4-F
E4-F
E4-F
E4-F
E4-F
E4-F
E4-F
E4-F
Context (of External
Review Draft)
Preface, pgxiii lines 12-
13
Preface, pgxvi lines 21-
22
Preface, pgxvi lines 31-
33
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
preface, pgxx lines 12-
13
Comment Text Excerpt
But is this application really relevant?? MWCNT will not be used in textile coatings alone
since they are unable to pass the tests by themselves. If they cannot pass the 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.
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.
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.
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.
But with which type of MWCNT? What surface chemistry for these materials?
Themes
Choice of nanomaterial and
application; choice constricted
science of MWCNTs
Choice of nanomaterial and
application; choice constricted
science of MWCNTs

Choice of nanomaterial and
application
Comparison WAS useful
Additional references
Highlight data gap
Choice of nanomaterial and
application; choice constricted
science of MWCNTs
MWCNT physical/ chemical
properties
                 March 2013
I-53
DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-13, cont: Free-form comments received from experts.
Comment ID
Author
  ID
 Context (of External
    Review Draft)
                             Comment Text Excerpt
          Themes
    201
 E4-F
preface, pgxxi lines 12-
13
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)?
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 study to
proactively look for systems of concern this made sense, but I think this original premise
needs to be revisited. Are these really the PRIME applications of MWCNT in the future
where emissions could are from, or are there applications missed in this original
assessment more likely to be the ones that should be looked at?
Choice of nanomaterial and
application; choice constricted
science of MWCNTs
    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 lines
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
                     March 2013
                                                             I-54
                                                                             DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-13, cont: Free-form comments received from experts.
Comment ID
Author
  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, pg2-1 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
                     March 2013
                                                             I-55
                                                                             DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-13, cont: Free-form comments received from experts.
Comment ID
Author
  ID
 Context (of External
    Review Draft)
                             Comment Text Excerpt
          Themes
    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
      . 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
    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
Chapters, pg 3-1
Iines16-18
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
                     March 2013
                                                             I-56
                                                                             DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-13, cont: Free-form comments received from experts.
Comment ID
221
222
223
224
225
Author
ID
E4-F
E4-F
E4-F
E4-F
E4-F
Context (of External
Review Draft)
Chapter 3, pg 3-8 lines
11-12
Chapters, pg 3-10 lines
24-25
Chapter 4, pg 4-10 lines
17-18
Chapter 4, pg 4-13 lines
1-2
Chapter 4, pg 4-25 lines
22-23
Comment Text Excerpt
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.
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.
Very interesting ... I did not know this. Have these results been verified?
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.
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.
Themes
MWCNT chemistry and nature
of bond with polymer/product

-
Additional references
MWCNT physical/ chemical
properties; Nature of bond with
polymer/product
                 March 2013
I-57
DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-13, cont: Free-form comments received from experts.
Comment ID
Author
  ID
 Context (of External
    Review Draft)
                             Comment Text Excerpt
          Themes
    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
    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
facia 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
                     March 2013
                                                             I-58
                                                                             DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-13, cont: Free-form comments received from experts.
Comment ID
230
231
232
233
234
235
236
Author
ID
E4-F
E4-F
E4-F
E4-F
E4-F
E4-F
E4-F
Context (of External
Review Draft)
Chapter 4, pg 4-30 lines
28-29
Chapter 4, pg 4-30 lines
31-32
Chapter 4, pg 4-32
Iines12-13
Chapter 4, pg 4-32 lines
22-23
Chapter 4, pg 4-34 line
1
Chapter 4, pg 4-43
(section 4.3 heading)
Chapters, pg5-1 lines
12-13
Comment Text Excerpt
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.
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.
So this would be worthwhile to determine.
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.
I think once you figure out what MWCNT will actually be used in, then this question starts to
get answered and you'll quickly figure out through demographic studies who is most likely to
be a part of a high-exposure population. If MWCNTs get mostly used in higher-end
performance applications (example - used a lot today in high end golf clubs) then you may
find that the highly exposed population is actually the affluent, and not lower income or
children.
Regardless of application - once MWCNT gets into the environment then all of this 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.
Themes
DecaBDE data needs
improvement
-
Highlight data gap
Highlight data gap
Choice of nanomaterial and
application; choice constricted
science of MWCNTs
Comparison WAS useful
Choice of nanomaterial and
application; choice constricted
science of MWCNTs
                 March 2013
I-59
DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-13, cont: Free-form comments received from experts.
Comment ID
Author
  ID
 Context (of External
    Review Draft)
                             Comment Text Excerpt
          Themes
    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
Chapters, pg 5-21
section heading 5.2
Agree that MWCNT reaction with environment is unknown and should be studied. I 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.
Choice of nanomaterial and
application; choice constricted
science of MWCNTs
    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
24-25
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
                     March 2013
                                                             I-60
                                                                             DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-13, cont: Free-form comments received from experts.
Comment ID
Author
  ID
 Context (of External
    Review Draft)
                             Comment Text Excerpt
          Themes
    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
                     March 2013
                                                             1-61
                                                                             DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-13, cont: Free-form comments received from experts.
Comment ID
245
246
247
248
249
250
Author
ID
E4-F
E4-F
E4-F
E4-F
E4-F
E4-F
Context (of External
Review Draft)
Chapter 6, pg 6-5
lines'! 1-12
Appendix A, pg A-1
lines 13-14
Appendix A, pg A-1 line
17
Appendix A, pg A-2 line
15
Appendix A, pg A-2
lines 17-18
Appendix A, pg A-5
Figure A-1 "medium list"
Comment Text Excerpt
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.
Why not electronics? Sporting goods?
Aerospace composites - very likely
There is more than you think - again, talk with NanoCyl in Belgium. They are mostly used in
polymers requiring enhanced electrical, thermal, and mechanical properties that can justify
the significant increase in cost - so higher end applications, not general consumer goods.
Agree on this - and this is DEFINITELY worth looking into.
I would argue is that this is where you went wrong in your selection study. Given all the deca
release from E-waste, MWCNT release from electronics seems the next most logical step
for study as it is in high use, and consumer electronics are being disposed of at an ever
increasing rate.
Themes
CEA framework/ methodology;
Choice of nanomaterial and
application
Choice of nanomaterial and
application; alternative
products/ formulations
Choice of nanomaterial and
application; alternative
products/ formulations
Choice of nanomaterial and
application; alternative
products/ formulations;
Additional references
Highlight data gap
Choice of nanomaterial and
application; alternative
products/ formulations
                 March 2013
I-62
DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-13, cont: Free-form comments received from experts.
Comment ID
251
252
253
Author
ID
E4-F
E4-F
E4-F
Context (of External
Review Draft)
Appendix A, pg A-6
Table A-1 (SWCNT
comparison)
Appendix A, pg A-7
Table A-2
(on the market)
Appendix A, pg A-8
lines 15-16
Comment Text Excerpt
This is far more likely to go into commercial production with in the next decade when
compared to MWCNTs for textiles.
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.
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.
Themes
Choice of nanomaterial and
application; alternative
products/ formulations
Choice of nanomaterial and
application
Choice of nanomaterial and
application; choice constricted
science of MWCNTs
                 March 2013
I-63
DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-14.   Comments received in response to the public comment period.
Comment ID
1
2
3
4
5
6
7
Author ID
P1
P1
P2
P2
P2
P2
P2
Comment Text Excerpt
"... there is significant information on decaBDE which is not included ... example ... cited water solubility of 20-30 pg/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 pg/L. The measured water solubility, determined in a guideline/GLP-compliant
study in 1997, of the >=97% decaBDE product is <0.1 pg/L."
"I would like to make you aware of critical papers on decaBDE ... I will provide these papers in separate emails due to
size."
"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."
"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."
"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 Ti02 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."
"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."
"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?"
Themes
DecaBDE data needs
improvement; p-chem
properties
DecaBDE data needs
improvement; impacts
Praise for Case Study;
nanomaterial series
Praise for Case Study;
comparison element
CEA framework/
methodology; purpose of
CEA
Choice of nanomaterial
and application; choice
constricted science of
MWCNTs
CEA framework/
methodology; purpose of
CEA
                March 2013
I-64
DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-14, cont: Comments received in response to the public comment period.
Comment ID
8
9
10
11
12
13
14
Author ID
P2
P2
P2
P2
P2
P2
P2
Comment Text Excerpt
"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."
"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."
"Recent studies have demonstrated that carboxylated CNTs are less likely to induce profibrogenic effects (Reference for
MWCNTs: Wang et al., 201 1 , ACS Nano 5(12):9772-9787) and are more likely to degrade (Reference for SWCNTs: 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."
"Also, a discussion on the meaning of functionalizing would be appropriate."
"... 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."
'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, 1 15(7): 1059-1 065) 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."
"Obviously, the colleagues at the October meetings face challenges. Supplementing the life cycle approach with concepts
from the NNI's EHS plan or the principles from the recent NRC report on EHS are options, but utilizing the Agency's
approach to SNURs and SNUNs is perhaps more pertinent. "
Themes
Choice of nanomaterial
and application; choice
constricted science of
MWCNTs
Choice of nanomaterial
and application; choice
constricted science of
MWCNTs
MWCNT physical/
chemical properties
Material Processing;
define terminology
CEA framework/
methodology; references/
literature
CEA framework/
methodology; references/
literature
CEA framework/
methodology; other
frameworks
                March 2013
I-65
DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-14, cont: Comments received in response to the public comment period.
Comment ID
15
16
17
Author ID
P2
P2
P2
Comment Text Excerpt
"As a possible contribution to this dialog, examining the MWCNT case history using SNUR (material submitted for a PMN)
and SNUN (extension of a SNUR to new applications/use patterns) perspectives, leads to questions on MWCNT
categories and their relationship to properties.
A: Categorization of MWCNTs:
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?"
"In summary, it may be that an MWCNT- CEA is not well suited for setting research priorities when being constrained to a
linear raw material-to landfill sequence, while also incorporating the recent scientific literature. The virus life cycle, for
example, starts with a virus entering the cell, follows virus replication and 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."
References
Themes
CEA framework/
methodology; questions to
consider for risk
assessment
CEA framework/
methodology; other
frameworks
References
                March 2013
I-66
DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-14, cont: Comments received in response to the public comment period.
Comment ID
18
19
23
24
254
255
Author ID
P3
P3
P1
P1
P4
P4
Comment Text Excerpt
"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)."
"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."
"...please see the following web address for decaBDE's EU risk assessment completed in 2002:
http://esisjrc.ec.europa.eu/doc/risk assessment/REPORT/decabromodiphenyletherreport013.pdf . That document
discusses decaBDE's use in textiles, and has an up to date (as of 2002) discussion of decaBDE's toxicology. "
"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."
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.
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.
Themes
Other impacts; Additional
considerations for section
Other impacts; Additional
considerations for section
DecaBDE data needs
improvement; impacts
DecaBDE data needs
improvement;
toxicokinetics
CEA framework/
methodology; Future of
risk assessment
In vitro methods
                March 2013
I-67
DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-14, cont: 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
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 21st 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-case basis with
preference given to in vitro nanomaterial-specific methods.
In vitro methods
    259
   P4
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
                      March 2013
                                                                I-68
                                                                               DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-14, cont: Comments received in response to the public comment period.
Comment ID
261
262
263
264
265
266
267
268
269
Author ID
P4
P4
P4
P4
P4
P4
P4
P4
P4
Comment Text Excerpt
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.
Paxton, JW. The allometric approach for interspecies scaling of pharmacokinetics and toxicity of anti-cancer drugs. Clin.
Exp. Pharmacol. Physiol. 1995; 22: 851-854.
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.
Liu, D et al. Different cellular response mechanisms contribute to the length-dependent cytotoxicity of multi-walled carbon
nanotubes. Nanoscale Res Lett. 2012 Jul 2;7(1):361.
Coccini, T. et al., Toxicology. Effects of water-soluble functionalized multi-walled carbon nanotubes examined by different
cytotoxicity methods in human astrocyte D384 and lung A549 cells. Toxicology. 2010 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.
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.
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.
Warheit, D. Nanoparticles Health Impacts? Nanomaterials Today. 2004;7: 32-35.
Themes
p-chem properties
Toxicokinetics
Analytical techniques
Impacts; In vitro methods
Impacts; In vitro methods
In vitro methods; Highlight
data gap
Highlight data gap;
Analytical techniques
Toxicokinetics; Analytical
techniques
Impacts
                March 2013
I-69
DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-14, cont: Comments received in response to the public comment period.
Comment ID
270
271
272
273
274
275
276
277
Author ID
P4
P4
P4
P4
P4
P4
P4
P4
Comment Text Excerpt
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.
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.
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.
Osier, M and Oberdorster, G. Intratracheal inhalation vs Intratracheal Instillation: Differences in Particle Effects.
Fundamental and Applied Toxicology. 1997; 40, 220-227.
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.
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.
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.
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.
Themes
Analytical techniques
Highlight data gap
Impacts; Future of risk
assessment
Analytical techniques
Analytical techniques;
Future of risk assessment;
Highlight data gap
In vitro methods; Future of
risk assessment
In vitro methods; Highlight
data gap
In vitro methods; Analytical
techniques
                March 2013
I-70
DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-14, cont: Comments received in response to the public comment period.
Comment ID
278
279
280
281
282
283
284
285
286
Author ID
P4
P4
P4
P4
P4
P4
P4
P4
P4
Comment Text Excerpt
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.
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.
http://www.vitrocell.com/index.php7Nav Nummer=8&

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.
http://www.mattek.com/paqes/nanoparticles/

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 permeability model.
Toxicol In Vitro. 2007; 21(7): 1215-9.
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.
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.
Themes
In vitro methods; Highlight
data gap
In vitro methods; Highlight
data gap
In vitro methods; Highlight
data gap
In vitro methods; Highlight
data gap
In vitro methods; Highlight
data gap
In vitro methods
In vitro methods; Highlight
data gap
In vitro methods; Highlight
data gap
In vitro methods; Highlight
data gap
                March 2013
1-71
DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-14, cont: Comments received in response to the public comment period.
Comment ID
287
288
289
290
291
292
293
294
Author ID
P4
P4
P4
P4
P4
P4
P4
P4
Comment Text Excerpt
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.
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.
Alfaro-Moreno, E. et al. Co-cultures of multiple cell types mimic pulmonary cell communication in response to urban PM10.
EurRespirJ.2008;32:1184-1194.
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
DMA damage). These data contribute to understanding the mechanism by which MWCNTs may induce toxic effects.
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.
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.
We suggest that MWCNTs (as well as other nanomaterials) be tested using NexGen Respiratory Toxicity Model developed
by EPA's Office of Research and Development. This model system takes advantage of both an air-liquid-interface in vitro
cell-based construct coupled with omics-based mechanistic pathway and biomarker identification.
http://nas-sites.orq/emerqinqscience/files/2012/06/Devlin.pdf

Themes
In vitro methods; Highlight
data gap
In vitro methods; Highlight
data gap
In vitro methods
In vitro methods; Highlight
data gap
In vitro methods
Future of risk assessment
Future of risk assessment;
in vitro methods
Future of risk assessment
                March 2013
I-72
DRAFT - DO NOT CITE OR QUOTE

-------
Table 1-14, cont: Comments received in response to the public comment period.
Comment ID
295
296
297
298
299
Author ID
P4
P4
P4
P4
P4
Comment Text Excerpt
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.
We look forward to seeing progress made toward the replacement of animal-based testing methods and encourages EPA
to continue working toward this goal.
Environmental Protection Agency, "Nanomaterial Research Strategy," 2009.
Environmental Protection Agency. 2009. The U.S. Environmental Protection Agency's Strategic Plan for Evaluating the
Toxicity of Chemicals. Office of the Science Advisor, Science Policy Council, U.S. EPA
(http://www.epa.qov/osa/spc/toxicitytestinq/docs/toxtest strateqy 032309.pdf

National Research Council, "Toxicity Testing in the Twenty-First Century: A Vision and a Strategy," report of the
Committee on Toxicity and Assessment of Environmental Agents, June 2007.
Themes
Future of risk assessment;
in vitro methods
Future of risk assessment;
In vitro methods
Future of risk assessment
Future of risk assessment
Future of risk assessment
                March 2013
I-73
DRAFT - DO NOT CITE OR QUOTE

-------
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 exposure Appraisal of human health risk assessment based on open
   literature [Review]. CritRev Toxicol 40: 759790. http://dx.doi.org/10.3109/10408444.2010.506638

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 nano materials. In Series on the Safety of Manufactured Nano materials, No 33.
   (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 polycyclic 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). (2012a). 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. (20lib). Carbon nanotube mass production: principles and
   processes [Review]. ChemSusChem4: 864-889. http://dx.doi.org/10.1002/cssc.201100177
March 2013                                          I-74                    DRAFT-DO NOT CITE OR QUOTE

-------
This page intentionally left blank.

-------
    Appendix J. Knowledge Map Pilot:
 Environmental Transport, Transformation
                and Fate
March 2013                         DRAFT - DO NOT CITE OR QUOTE

-------
     Appendix J.  Knowledge Map  Pilot:
                              Environmental Transport,
                              Transformation  and  Fate

 1         Appendix J presents work from a pilot project to continue refining the CEA approach. The main
 2   objective of this work is to develop methods for visually and concisely representing information in the
 3   CEA Framework. Specifically, Chapter 3. "Environmental Transport, Transformation and Fate" in the
 4   External Review Draft of this case study document (Chapter 3. Section G.3. and Section H.3 in the
 5   current Peer Review Draft) was selected to develop a proof of concept for a particular method to visually
 6   convey information. Here, this method is briefly described followed by the  products to date from this
 7   pilot.


      J.1. Knowledge Maps
 8         Knowledge maps are used in a variety of fields (e.g., organization management, journalism) to
 9   visually represent concepts and how they relate to one another (Novak and  Canas. 2008; Kim et al.,
10   2003). Given their wide applicability and utility in representing complex concepts, knowledge maps were
11   selected for this pilot to visually convey information in the CEA framework. A knowledge map is
12   composed of a network of nodes (usually presented as common shapes) connected by edges (depicted as
13   lines or arrows). Nodes usually represent concepts, whereas edges show relationships between the
14   concepts.


      J.2. Transport, Transformation and Fate Knowledge Maps:
      Overview
15         The knowledge maps developed in this pilot are three conceptually linked portions of one map
16   displaying environmental transport, transformation, and fate for either decabromodiphenyl ether
17   (decaBDE) or multiwalled carbon nanotubes (MWCNT). Showing all of the information related to
18   environmental transport, transformation, and fate on one map would result in a great amount of detail in a
19   small amount of space, and thus for readability the information pertinent to each process is displayed in
20   three distinct maps. The Physicochemical Properties Map (PPM) focuses on the material itself, providing
     March 2013                              J-1                DRAFT-DO NOT CITE OR QUOTE

-------
 1    a sense of the basic properties of the material, which might in turn influence how the material partitions in
 2    the environment after release from a product life cycle stage (e.g., product manufacturing, use,
 3    disposal/recycling). The Transport Map (TpM) portrays how the material might subsequently move
 4    between environmental zones (i.e., aquatic, atmospheric, terrestrial) based on available information. In
 5    addition, it provides a sense of which zone(s) are sinks for the material, and thus would likely have higher
 6    concentrations of the material than others would. Once the material makes it to a zone, the
 7    Transformation Map (TfM) depicts how the material might be transformed due to the combination of
 8    material and environmental properties within a given zone.
 9           Note that because the maps are conceptually a single unit, symbols mean the same thing across all
10    the maps. For example, black edges with a solid arrowhead always represent movement whenever they
11    appear in any of the maps. Similarly, the environmental zones look the same in all three maps.
12           Each type of map (i.e., PPM, TpM, TfM) is described in greater detail below, along with
13    a brief comparison of the decaBDE and MWCNT maps in each category.
       J.3.  Physicochemical Properties Maps
14           The PPM35 shows the properties of a material likely to influence its behavior in the environment,
15    the corresponding effects on environmental behavior, and finally how resulting behaviors influence
16    movement to environmental zones.
17           The PPM has three  banks of nodes. In the top bank, material properties are represented in
18    rectangular nodes. The nodes are color-coded based on information in Table 3-2 and Table H-3; nodes are
19    dark red if the value is high and light red if the value is low36, whereas light green means that the property
20    can vary according to the specific sample of the material (e.g., MWCNT with one type of surface coating
21    versus another) and light purple indicates a property which may or may not occur based on environmental
22    conditions.
23           The middle bank of nodes contains effects in ovals. Like the properties, they are color coded so
24    that high effects are dark yellow, while low effects are light yellow. Where variability exists, such that the
25    effect can be high or low, a dark yellow and a light yellow circle have been placed in a box that represents
26    that effect. A white oval with a dashed border indicates an effect that is predicted to occur based on
      35 Note: These maps are based primarily on Table 3-1 (for MWCNT) and Table H-3 (for decaBDE), and although
      details from other parts of Chapter 3, Section G3, and Section H.3 are included where relevant, the PPMs represent a
      translation of information in tabular form to figures in this pilot work.
      36Note: Scaling the shading of red to present values more specific than "low" or "high" was determined to be too
      confusing for this representation.
      March 2013                                   J-2                   DRAFT-DO NOT CITE OR QUOTE

-------
 1    material properties but has not yet been observed. The effect nodes are connected to the property nodes
 2    via green edges with diamond arrowheads; these edges indicate a simple causal relation, i.e., that a
 3    particular property leads to a particular effect. Green edges that connect to boxed effect nodes indicate
 4    that the given property might lead to a high or low effect, given varying circumstances.
 5           The lower bank of nodes  contains the environmental zones, which are more fully described in the
 6    transport map. Black edges connect effect nodes to zone nodes to indicate that the effect facilitates
 7    transport to that zone. Levels of transport are differentiated, with high transport represented by thick, solid
 8    lines and low transport presented  with thin, dashed lines.
 9           Overall, the map can be read as a three-part sentence that follows the edges from property to
10    effect, to zone. For example, "the  low water solubility of decaBDE leads to high sorption to organic
11    matter, which causes a high level  of transport to the soil."
12           The decaBDE and MWCNT PPMs are consistent with each other in their use of symbols;
13    however, there are notable differences between the two maps. For example, the PPM for MWCNT
14    includes more types of nodes (e.g., light green property nodes and boxed effect nodes) because MWCNT
15    properties can vary based on size, shape, surfactant, and other conditions. In addition, while all property
16    nodes in decaBDE are separated by whitespace, property nodes that are two poles of the same property
17    (like "Size: single" and "Size: cluster") are placed next to each other on the MWCNT map, again due to
18    the variability associated with the material.
      March 2013                                    J-3                  DRAFT-DO NOT CITE OR QUOTE

-------

Henry's law
constant
/


Vapor
pressure
| Molecular
weight

//

Water
solubility



Koa

Phy
X
'sicochemical Properties
Kow


Lipophilicity
/

                                                                                                                 Legend
                                                                                                          Properties
                                                                          Effects
                                                                                                               Edge Types
                                                                                                                 Leads To
                                                                                                                  High
                                                                                                               Transport To
                                                                                                               	>
                                                                                                                  Low
                                                                                                               Transport To
Figure J-1. Physicochemical properties map: Decabromodiphenyl ether.
March 2013
J-4
DRAFT-DO NOT CITE OR QUOTE

-------
                                                             Physicochemical Properties
                                                                 Legend
                                                                                                     Properties
                                                                       Effects
                                                                                                                ^redictedx
                                                                                                                \ Effect J
                                                                                                          Edge Types
                                                                                                           Leads To
                                                                                                            High
                                                                                                         Transport To

                                                                                                            Low
                                                                                                         Transport To
Figure J-2. Physicochemical properties map: Multiwalled carbon nanotubes.
March 2013
J-5
DRAFT-DO NOT CITE OR QUOTE

-------
       J.4.  Transport Map
 1           The TpM begins with the release of a material in a product life cycle stage (see PPM). From there
 2    the map shows how the material might travel between or within spatial zones, the environmental
 3    properties that affect material transport, and some processes that describe the movement.
 4           The TpM has only one bank of nodes, containing the environmental zones. There are three main
 5    zones: atmospheric (grey), terrestrial (brown), and aquatic (blue). Within each of these zones there are
 6    more specific sub zones (e.g., waste water, surface water, sediment within the aquatic zone). Zones that
 7    are sinks for the material, based on information available in Chapter 3. Section G.3. and Section H.3 of
 8    the document, have a thick black border. When no data were found on the concentration of the material in
 9    that zone, the border is dotted.
10           Transport is represented by solid or dotted black lines leading from one zone to another. The bold
11    solid lines represent movement described as "most likely to occur," based on information available in
12    Chapter 3, Section G.3, and Section H.3. Dotted lines thus represent data inferred from other materials or
13    likely due to chemical properties, but for which explicit measurements have not been taken (e.g.,
14    MWCNT leeching from soil to groundwater). For instance, in the decaBDE TpM a dotted line is used to
15    depict decaBDE transport from surface water to sediment is most likely to occur based on low water
16    solubility and other material properties (see Figure J-3).
17           For many of the transport processes, the text in Chapter 3. Section G.3. and Section H.3 used a
18    term to describe the process (such as deposition, leaching, or runoff), and those terms are  listed as a label
19    on the appropriate edge. Any environmental properties of the spatial zones that were cited in the chapter
20    as affecting material transport are represented by numbers to highlight the influence these factors might
21    have on the extent, rate, or direction of transport.
22           Overall, the maps can be read as a sentence that follows the edges from release to a spatial zone
23    and then to another spatial zone. For example, "decaBDE is released to the atmospheric zone, where is
24    can be transported through deposition to waste water, with the extent of deposition influenced by the
25    amount of total organic carbon (TOC) in the air, precipitation, wind, and temperature."
26           When comparing the decaBDE and MWCNT TpMs, it is clear that the spatial zones and
27    environmental properties are the same; however, as expected, there are differences in the specific types of
28    transport that occur and the environmental properties that influence them. More notably, there are no
29    known sinks for MWCNT, whereas there are three for decaBDE (soil, sediment, and sewage sludge). For
30    MWCNT there is also much less  data on concentrations in the zones and therefore many more dotted
31    lines than in the decaBDE map.

      March 2013                                   J-6                   DRAFT-DO NOT CITE OR QUOTE

-------
                                                               decaBDE
                                                                release
                                                                                   Surface Water
                                                                                      Deposition (1,4, 5,6)
                                                                                            Discharge
                                                                                 (1) lsorption(1,10)
                                                                                                    Aquatic
                                                    Waste Water
                                                               (D
                                                                                     Sediment    | Sewage Sludge)
                                                                            Leaching (1)

                                                                                Runoff (1,4)
                                                                          Agricultural Usage & Landfills (1)
                                                                  Incineration


Zone Concentration
: No Data ;

Data Available

I Known Sink I


Legend

Edge Types
Movement >

Strong Movement ^






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


Figure J-3. Transport map: Decabromodiphenyl ether.
March 2013
J-7
DRAFT-DO NOT CITE OR QUOTE

-------
                                                         MWCNT
                                                         release
                               Deposition (4,5,6,10)
                                                                                     Leaching
	 y 	 ....
Atmospheric
Indoor Air Outdoor Air



	 y. 	 y 	
Terrestrial
Biota








: Waste Water :...=£
    :  Residential :
Rural
                                               Animals  :  : :    Soil


Commercial


(10)

i Urban
















Plants ':<••••'•
. .""K"""::::. •

(4,5,6,10)





Er

     	>•
     Runoff
                                                                                                     Aquatic
        :      Settling (2)
        ;      i(178>:    Sediment   :
	y.	1...    .'.'.V.'.V.V.V.V.'.V.'.V.V."
: Surface Water *••-:  Groundwater  :
''	A	:    '"A	:
     Biota

: ;  Animals

: j   Plants
                                                                       Erosion (1,2,7,8)  :
                                                                                  Leaching (1J,8)
                                                Deposition (4,5,6,10)
Figure J-4. Transport map: Multiwalled carbon nanotubes.
March 2013
                    J-8
DRAFT-DO NOT CITE OR QUOTE

-------
       J.5.  Transformation  Map
 1           Each TfM begins with the material entering spatial zones, either from release in a product life
 2    cycle stage (see PPM), or movement between spatial zones (see TpM). From there the map shows
 3    transformations that occur within those zones, the environmental properties that affect the transformation,
 4    and the resulting compounds.
 5           There are three banks of nodes in the TfMs. The top bank represents the spatial zones (carried
 6    over from the previous two maps and including known sinks). The second bank of nodes represents the
 7    transformations that are associated with the spatial zones. Transformations are grouped into three main
 8    categories: chemical, biological, and physical, which are represented by yellow (photo, thermal and
 9    geochemical degredation), red (sorption), or green (biological degredation) parallelograms, respectively.
10    Many of the  spatial sub-zones are associated with specific transformations; these instances are denoted by
11    a red edge linking the sub-zone with the particular transformation (e.g., plants within "terrestrial biota"
12    are associated with biological degradation in the decaBDE map). An edge to the main "Transformations"
13    box implies an association with all of the transformations within that box (e.g., Soil within "terrestrial" is
14    associated with biological, photo, and geochemical degradation in the decaBDE map). When the text in
15    Chapter 3. Section G.3. and Section H.3 cites reason to infer data that has not been explicitly measured or
16    proven, it is represented by a dashed line throughout the map.
17           As in the TpM, available information from the text on the environmental properties that might
18    influence transformation is represented by numbers next to the relevant edge connecting the zone and the
19    transformation.
20           The third bank of nodes is nested inside of the transformations and represents the compounds that
21    result from the transformation that occurred. A black line with an open circle is used as the edge
22    connecting each transformation to the resulting compound. The resulting compounds are then represented
23    by solid black shapes, such as a triangle to represent polybrominated/-chlorinated dibenzofurans in the
24    decaBDE map.
25           The spatial zones, general transformation types, and environmental properties are the same in the
26    legend of both the  decaBDE and MWCNT TfMs; however, there are differences between the maps for
27    each material in the specific transformations that occur and the environmental properties that influence
28    them. Notably, there are fewer types of transformation products associated with MWCNTs compared
29    with decaBDE, although this may be due to a lack of data. Similar to the transport map, there are more
30    dashed edges in the MWCNT map than the decaBDE map, representing greater uncertainty in the
31    particular type  of transformations that might occur for MWCNT. Overall, the maps can be read as a three-

      March 2013                                   J-9                   DRAFT-DO NOT CITE OR QUOTE

-------
1    part sentence that follows the edges from spatial zone to transformation to resulting compound. For
2    example, "decaBDE can be transported to surface water in the aquatic zone, which is associated with
3    photodegradation, the extent of which is influenced by quantity of UV light and results in
4    polybrominated/-chlorinated dibenzo furans."
5
     March 2013                                    J-10                  DRAFT-DO NOT CITE OR QUOTE

-------
                        Atmospheric
Indoor Air
Residential

Commercial



Outdoor Air
Rural

Urban

                     Atmospheric Transformations
                   Photo    ! Thermal
                 degradation// degradation
     Biological  /  /  Photo   / fieochemical
    Degredation/ /Degredation/ / Degredation
Legend

Transformation Types
/Chemical /

/Biological/
M^BHV
flUiljyi^W
Edge Types
Movement ^
Associated with
Results in „

| Known dBDE Sink 1

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

Resulting Compounds
decabromodiphenyl polybrominated/-chlorinated
ether bound complex dibenzofurans
lower brominated polybrominated/-chlorinated
congeners dibenzo-p-dioxins
+ A

 Figure J-5. Transformation map: Decabromodiphenyl ether.
March 2013
J-11
DRAFT-DO NOT CITE OR QUOTE

-------
                          Atmospheric

                      Indoor Air    j ;  Outdoor Air

                     ; Residential I  i I  j   Rural  ;

                                 : ;  !   Urban  ;

                                 I
                                 I
                                   Terrestrial

                                  Biota

                                 Animals  ;  i   Soil

                                  Plants   !
                                                                                     Aquatic

                                                                               Biota            i ; Ground Water i
                                                                   :  :   Plants   i:   Animals   i  i :'"''
                                                                   :  :	       ...j :	       ....i  : ; Waste Water ;
                                                              iSurface Water; i   Sediment   ;   i Sewage Sludge ;
                      Atmospheric Transformations
                         i   i   Surface  11  Surface
                 Oxidation/   /_,__„_,_.;__//d
degredation / / degredation
6,10
                                                           Terrestrial Transformations
  Surface
degredation
                                             I I  Surface  7
                                            /  /degredation/
                                                                        degi
                                                                         /
                                                                         l o
                                                                          Aquatic Transformations
                                                                             Surface
                                                                                        Surface
degradation // degradation
                                                                                                   Oxidation
                                                                                10 Q
                                                                                                                      10
                                                                                                  6 j)   (j)1Q
Legend

Transformation Types
/Chemical /

/Biological/
jHSf



Edge Types
Movement w
Associated with
Suspecte dAssociation
Results in ^
Suspected Results^




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



Resulting Compounds
Multiwall carbon
nanotubes MWCNT
• •
Amorphous Bundled
Carbon MWCNTs
• •


Figure J-6. Transformation map: Multiwalled carbon nanotubes.
March 2013
                       J-12
                                                     DRAFT-DO NOT CITE OR QUOTE

-------
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
March 2013                                     J-13                   DRAFT-DO NOT CITE OR QUOTE

-------