DRAFT
DO NOT CITE OR QUOTE
E P A /600/ R~08/086 A
November 2008
External Review Draft
An Exposure Assessment of Polybrominated Diphenyl Ethers
NOTICE
THIS DOCUMENT IS A PRELIMINARY DRAFT. It has not been formally released by
the U.S. Environmental Protection Agency and should not at this stage be construed to
represent Agency Policy. It is being circulated for comment on its technical accuracy and
policy implications.
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC
-------
DISCLAIMER
This report 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.
11
-------
CONTENTS
LIST OF TABLES vi
LIST OF FIGURES viii
ABBREVIATIONS AND ACRONYMS ix
PREFACE x
AUTHORS AND REVIEWERS xi
EXECUTIVE SUMMARY xii
1.0. INTRODUCTION 1-1
REFERENCES 1-3
2.0 PRODUCTION, USE AND LIFECYCLE OF POLYBROMINATED DIPHENYL
ETHERS 2-1
2.1. INTRODUCTION 2-1
2.2. PRODUCTION 2-2
2.3. USES OF PBDEs 2-5
2.4. LIFECYCLE OF PBDE 2-7
2.4.1. Production Releases 2-7
2.4.2. PBDE Content of Consumer Products 2-8
2.4.3. Estimates of PBDE Releases to the Air Based on Chamber Testing 2-10
2.4.4. Estimates of PBDE Releases to Air from Consumer Products 2-12
2.4.5. Estimates of the Mass Flow of PBDEs Contained in Electronic and
Electrical Equipment (EEE) Waste 2-17
2.4.6. Estimates of the Mass Loading to Land and Water of PBDEs Present in
Sewage Treatment Plant Sewage Sludge and Effluent 2-21
2.5. AN EXAMPLE LIFECYCLE ANALYSIS OF DECABDE IN THE UNITED
STATES 2-25
2.6. SUMMARY 2-28
REFERENCES 2-30
3.0. ENVIRONMENTAL FATE OF POLYBROMINATED DIPHENYL ETHERS 3-1
3.1. INTRODUCTION 3-1
3.2. PHYSICAL/CHEMICAL PROPERTIES 3-1
3.2.1. Water Solubility (WS) 3-1
3.2.2. Octanol/Water Partition Coefficient (Kow) 3-2
3.2.3. Henry's Law Constant (H) 3-3
3.2.4. Vapor Pressure (VP) 3-3
3.2.5. Octanol/Air Partition Coefficient (K„a) 3-4
3.2.6. Gas-Particle Partitioning in Air 3-5
3.3. BIO ACCUMULATION, BIOCONCENTRATION, AND
BIOMAGNIFICATIONS OF PBDES IN THE AQUATIC ENVIRONMENT 3-6
3.4. BIOTIC AND ABIOTIC DEBROM1NATION AND TRANSFORMATION OF
PBDES 3-7
3.4.1. Microbial Degradation 3-8
3.4.2. In Vivo Metabolic Debromination in Animals 3-10
3.4.2,1 Evidence for Debromination in Fish 3-1 1
3.4.2.2. Evidence for Debromination in Rats 3-14
3.4.2.3. Evidence for Debromination in Birds 3-15
iii
-------
3.4.2.4. Evidence for Dehromination in House Cats 3-16
3.4.3. Abiotic Degradation of PBDEs 3-17
3.4.3.1. Photodegradation of PBDEs 3-17
3.4.3.2 Reaction with the Hydroxvl Radical 3-24
3.5. THERMAL DECOMPOSITION OF PBDE 3-24
3.6. PATTERNS OF ENVIRONMENTAL FATE OF PBDE 3-25
3.6.1. Fate of PBDEs in Air 3-26
3.6.2. Fate of PBDEs in Water 3-30
3.6.3. Fate of PBDEs in Soil 3-32
3.6.4. PBDEs in Sewage Treatment Plant Influent, Effluent and Sludge: A Major
Cause of PBDE Contamination in Surface Waters 3-33
3.7. BIO ACCUMULATION IN BIOTA 3-36
3.7.1. Bioaccumultation in the Aquatic Environment 3-36
3.7.2. Bioaccumultation in the Terrestrial Environment 3-38
3.7.2.1. Bioaccumulation in Birds 3-39
3.4.2.2. Bioaccumulation in Terrestrial Mammals 3-42
3.7.3. Bioaccumulation in Insects 3-44
3.8. ENVIRONMENTAL TIME TRENDS 3-45
3.8.1. Time Trends in Sediment Core Studies 3-46
3.8.2. Time Trends in Wildlife Tissue Samples 3-51
3.9. CONCLUSIONS 3-53
REFERENCES 3-55
4.0. ENVIRONMENTAL AND EXPOSURE MEDIA CONCENTRATIONS 4-1
4.1. INTRODUCTION 4-1
4.2. WATER AND SEDIMENT 4-1
4.3. SURFACE SOIL 4-6
4.4. INDOOR DUST 4-8
4.5 AIR CONCENTRATIONS 4-15
4.5.1. Outdoor Air Concentrations 4-15
4.5.2. Indoor Air and Simultaneous Indoor/Outdoor Monitoring 4-22
4.6. FISH CONCENTRATIONS " 4-27
4.6.1. Farmed Fish Concentrations 4-28
4.6.2. Fresh Water and Marine Fish 4-30
4.6.3. Fish From the Retail Marketplace 4-35
4.6.4. Observations from Fish Data 4-38
4.7. FOOD CONCENTRATIONS 4-39
4.8. ASSIGNING EXPOSURE MEDIA CONCENTRATIONS FOR EXPOSURE
ASSESSMENT PURPOSES 4-45
REFERENCES 4-51
5.0. HUMAN EXPOSURE 5-1
5.1. INTRODUCTION 5-1
5.2. BODY BURDEN DATA 5-1
5.2.1. Blood Data 5-3
5.2.2. Breast Milk Data 5-13
5.2.3. Adipose and Other Tissue 5-19
5.2.4. Selection of Representative Body Burden Profiles 5-21
IV
-------
5.3. STUDIES ON INTAKE, OR EXPOSURE, DOSE 5-22
5.4. ESTIMATES OF BACKGROUND INTAKES OF PBDES FOR ADULTS 5-32
5.5. CONVERTING ADULT INTAKE DOSE TO BODY BURDEN 5-35
5.6. EXPOSURE OF SPECIAL POPULATIONS TO PBDES 5-38
5.6.1. Impacts to Infants from Consumption of Breast Milk 5-38
5.6.2. Childhood Intakes 5-41
5.6.3. Body Burden Data to Characterize Occupational Exposures 5-42
5.6.4. Elevated Exposures at the High End of the General Population 5-43
5.7. UNCERTAINTY AND VARIABILITY IN ESTIMATING INTAKE DOSE
AND CONVERTING THAT DOSE TO A BODY BURDEN 5-47
5.8. OVERALL FINDINGS OF EXPOSURE OF AMERICANS TO PBDES 5-56
REFERENCES 5-59
v
-------
LIST OF TABLES
Table E-l. Representative exposure media concentrations xxv
Table E-2. Predicted doses and lipid-based concentrations of BDEs, with predicted
concentrations compared with lipid-based measurements in blood and milk xxvi
Table 1-L BDE congener numbers and chemical composition of the most commonly
studied BDE congeners 1-4
Table 2-1. Approximate BDE honiologue and congener-specific content of commercial
PBDE formulations 2-35
Table 2-2. Estimated environmental releases (MT) of BDE congeners from U.S.
decaBDE production facilities in 2004 2-36
Table 2-3. Mean concentration (ng/kg) ofpenta- and octaBDE in various flame retarded
polymers 2-36
Table 2-4. Mean concentrations of PBDEs (mg/kg) found in electrical and electronic
waste material at a recycling plant in Switzerland 2-36
Table 2-5. BDE congener concentrations (mg/kg) present in EEE waste components
sampled at a recycling facility 2-37
Table 2-6. Summary of BDE congener emissions (MT/yr) to air from a 10-yr service life
of fire retarded products. 2-38
Table 3-1. Estimated aqueous solubility values for PBDEs 3-66
Table 3-2. Estimated log Kow values for PBDEs 3-68
Table 3-3 Estimated Henry's Law constants (H) for PBDEs 3-70
Table 3-4. Estimated solid phase vapor pressures (PS) and sub-cooled liquid vapor
pressures (PL) of some PBDEs (Pascals at 25°C) 3-71
Table 3-5. Estimated octanol/air partition coefficients (log KOA) of PBDEs 3-73
Table 3-6. Calculated theoretical vapor-particle partitioning of PBDE congeners in
ambient air at 25°C (calculated using Equation 3-2) 3-74
Table 3-7. Estimated BAF and BMP values for various aquatic species 3-75
Table 3-8. The BDE congener distributions in influent, sludges and final effluent from
various surveys of sewage treatment plants in various countries 3-76
vi
-------
Table 4-1. C ongencr-spceific concentrations of PBDES in indoor dust in the United
States (units in ng/g dry weight) 4-59
Table 4.2. Outdoor and indoor congener-specific air concentrations of PBDEs in the
United States (units in pg/m3) 4-63
Table 4.3. Congener-specific fish concentrations of PBDEs for fish caught in the United
States (units in ng/g wet weight, wwt, or lipid weight, Iwt, in parenthesis if available) 4-72
Table 4.4. Congener-specific concentrations of PBDEs in food originating from the
United States (concentrations in ng/g wwt or specifically identified as ng/g lwt, if that
was how reported by the author and conversion to ng/g wwt was not possible) 4-75
Table 4.5. Exposure media concentrations (note: — means no data available; ND means
data available but not detected.) 4-79
Table 5.1. Blood lipid concentrations of PBDE congeners in the United States 5-68
Table 5.2. Breast milk concentrations of PBDE congeners in the United States 5-72
Table 5.3. Representative body burden levels of PBDEs in Americans 5-77
Table 5.4. Estimates of general population intake, or exposure, dose of total PBDEs 5-78
Table 5.5. Exposure pathways and factors for the PBDE intake dose estimate 5-79
Table 5.6. Congener-specific and total adult intake estimates of PBDEs (exposures in
units of ng/day) 5-81
Table 5.7. Pharmacokinetic parameters and predicted concentrations of BDEs compared
with measurements in blood and milk 5-82
Table 5.8. Pharmacokinetic parameters for modeling the body burden impacts to infants
via breast feeding, and then to children from food and household exposures 5-83
vii
-------
LIST OF FIGURES
Figure E-l. Approach for characterizing exposure to polybrominated diphenyl ethers in
this report xxvi
Figure 2-1, General structure of PBDEs 2-39
Figure 2-2. Lifceycle of PBDEs 2-40
Figure 2-3. TRI data showing environmental releases (kg) of decaBDE from primary
production facilities in 2006, and total environmental releases from 1997 ^ 2006 (US
EPA, 2008a) 2-41
Figure 2-4. Volatilization of BDE-47 from PBDE-treated consumer products in the U.K. .2-42
Figure 3-1. Time-trends of deposition flux of PBDEs (diamonds and solid line) and
PCBs (diamonds and dashed line) to the sediments in each of the Great Lakes 3-78
Figure 5.1. Approach for characterizing exposure to polybrominated diphenyl ethers in
this report 5-84
Figure 5.2. Modeled infant and childhood body burdens of PBDEs 5-85
Figure 5.3. The fraction of the maximum concentrations of PCDD/F TEQ and BDE 47
concentrations found at various percentiles within NHANES surveys of these two
contaminants in adults 5-86
V111
-------
ABBREVIATIONS AND ACRONYMS
(To be completed)
BDE brominated diphenyl ether
PBDE polybrominated diphenyl ether
IX
-------
PREFACE
The United States Environmental Protection Agency (U.S. EPA) has formed a
working group comprised of individuals from several program offices including the
Offices of Pesticides, Prevention, and Toxic Substances, the Office of Water, and the
Office of Research and Development, Office of Policy, Economics, and Innovation, to
study production, use, alternatives, environmental fate, exposure, and health effects of
polybrominated diphenyl ethers (PBDEs). This working group issued a project plan in
2006 that outlined projects in these areas. EPA reports regularly on progress in
completing the activities identified in the project plan, with the most recent status report
issued in March 2008. The Web site that describes this working group, including the
project plan, is http://www.epa.gov/oppt/pbde. This document addresses the exposure
assessment needs identified in that project plan. It provides a comprehensive assessment
of the exposure of Americans to this class of persistent organic pollutants. Individual
chapters in this document address: the production, use, and lifecyele of PBDEs;
environmental fate: environmental levels; and human exposure.
x
-------
AUTHORS AND REVIEWERS
The National Center for Environmental Assessment (NCEA), Office of Research
and Development was responsible for the preparation of this document.
AUTHORS
Matthew Lorber
Exposure and Risk Characterization Group
National Center for Environmental Assessment
U.S. Environmental Protection Agency
Washington, DC 20460
David Cleverly
Exposure and Risk Characterization Group
National Center for Environmental Assessment
U.S. Environmental Protection Agency
Washington, DC 20460
INTERNAL EPA REVIEWERS
Linda Bimbaum
Jeff Frithscn
Office of Research and Development
Daniel Axelrad
Greg Miller
Kceve Nachman
Office of Policy and Environmental Information
Bob Boethling
Lynn Delpire
Tala Henry
Office of Pollution, Prevention, and Toxic Substances
EXTERNAL PEER REV IEWERS
(To be arranged)
ACKNOWLEDGMENTS
(To be added)
XI
-------
EXECUTIVE SUMMARY
Polyhrominatcd diphenyl ethers, PBDEs, are a class of brominated flame
retardants that are added to plastics, polyurethane foam, textiles, and electronic
equipment to reduce the likelihood of ignition and to slow the bum rate should products
catch fire. Largely as a result of using fire-resistant materials in construction, consumer
goods, plastics and textiles, the incidence of fires has decreased 24% over the past 10
years. PBDEs are a major class of flame retardants contributing to this benefit.
PBDEs have a common structure of a brominated diphenyl ether molecule that
may have anywhere from 1 to 10 bromine atoms attached. Depending on the location
and number of bromine atoms, there are a possible 209 PBDE compounds; each are
termed congener and arc assigned a specific BDE number. PBDEs have been marketed
in three primary formulations: (1) the "penta" formulation, commercially known as DE-
71# and Bromkal 70-5D.E®; (2) the "octa" formulation, DE-79 I> and (3) the "deca"
formulation, DE-83R!& or Saytex 102E®. The formulations differ in their composition
of BDE congeners. The dominant congeners in pentaBDE (percent weight basis in
parenthesis) are BDEs 99 (35-50%), 47 (25-37%), 100 (6-10 %»), 153 (3-5%), and 154
(2-4%). The octa formulation, is comprised of BDEs 183 (40%), 197 (21%), 203 (5-8%),
196 (8%), 208 (10%), 207 (7%), 153 (5-10%), and 154 (1-5%). The deca formulation is
dominated by BDE 209 (97-99%), with the remainder being BDEs 207 and 208.
Studies have been conducted in laboratory animals to gain a better understanding
of the potential health risks of PBDEs. These studies have suggested potential concerns
about liver toxicity, thyroid toxicity, developmental and reproductive toxicity, and
developmental neurotoxicity. These findings raise particular concerns about the potential
risks to children. To date, there is no ev idence of carcinogenicity of any of the PBDEs
except decabromodiphenyl ether, BDE 209. In a review of toxicological studies as part
of a Toxicological Review for the U.S. EPA's Integrated Risk Information System
(IRIS), the U.S. EPA has found that the data supports a finding of "suggestive evidence
of carcinogenic potential" according to the 2005 U.S. EPA draft Guidelines for
Carcinogenic Risk Assessment.
Approximately 56,418 metric tons (MT) of PBDEs were produced worldwide in
2003, the latest reporting year, with between 40,000 and 67,000 MT/yr produced between
xn
-------
1999 and 2002. In 2001, decaBDE accounted for 83% of total worldwide production,
followed by pentaBDE (11%), and octaBDE (6%). Approximately 95% of the global
production of pentaBDE, 40% of octaBDE, and 44% of decaBDE were consumed in the
Americas (North, South, and Central America). The penta and octa formulations were
voluntarily withdrawn from the United States marketplace by their manufacturers at the
end of 2004, leaving only the deca formulation currently being marketed for use in
commercial products in the United States. The penta and octa formulations were banned
in Europe, thus leaving the deca formulation as the only currently used formulation in
Europe, too.
A limited litecycle analysis provides some indication of the magnitude of releases
of PBDEs to the outside and inside environments, based on estimates of releases from
PBDE production processes, from volatilization of PBDEs in products while in use, and
from the disposal of products containing PBDEs. The Toxics Release Inventory (TRI)
reports a total release of 33.5 metric tons (MT) of decaBDE to the air, land, and water in
2006. According to TRI, total environmental releases peaked in 1999, with a release of
53.9 MT, and stayed at similar levels through 2002. There was drop in releases in 2003,
to 36.3 MT, followed by an increase in 2004 of 44.8 MT, and then sequential declines in
2005 and 2006. Limited studies, including measurement of PBDE releases from products
in experimental air chambers and measurement of the concentrations of PBDEs in
consumer products, provide a starting point for estimating the releases from products in
use. The European Union used an empirical equation, which was developed to
approximate the volatilization rate of chemical plasticizers added to plastic films during
their product use, to estimate the annual release of PBDEs in Europe. According to their
calculations, the annual volatilization rate of all decaBDE, octaBDE, and pentaBDE from
consumer products, expressed as a percent of the total mass of these PBDEs introduced
into products during the year, is 0.038%, 0.054%, and 0.039%. respectively. Based on
use data, they estimated annual releases from volatilization to be 2.6 MT/yr of decaBDE,
1.4 MT/yr of octaBDE, and 4.3 MT/yr of pentaBDE. This same empirical approach was
used to estimate volatile releases from these products in the Americas. The estimates
include 7.5 MT/yr of decaBDE, 0.8 MT/yr of octaBDE, and 26.6 MT/yr of pentaBDE
(based on use data from 2001, when octaBDE and pentaBDE were still being marketed
-------
for use). These estimates can, at best, be considered initial estimates that are
unveri liable, but they give an order of magnitude indication of the releases of these flame
retardants from products in use.
After use, releases can also occur from disposal of products. An estimate of the
releases from disposal of electronic and electrical equipment (EEE) was made. In the
United States, an estimated 2.4 million MT (MMT) of EEE waste was generated in 2005.
Of this amount, 0.3 MMT was recovered for recycling, leaving 2.1 MMT that was either
disposed of in landfills or incinerated. Losses from PBDEs in recycled products were not
considered. Losses from landfills were also not considered because of the careful
conditions (liners, treated leachate) required by law for hazardous waste landfills in the
United States. Based on measured concentrations of decaBDE in EEE products and the
destruction efficiency from municipal solid waste incinerators with a high degree of
pollution control, releases of decaBDE totalled 0.046 MT/yr, Releases to water and land
from effluent and sludge from sewage treatment plants were also considered. Congener-
specific releases to land and water were estimated, and the total loading to water (sum of
28 congeners) was 12.8 MT/yr to land and 1.4 MT/yr to water.
Once in the environment, PBDEs persist for decades and have been detected
globally in air, soils, sediments, oceans, and wildlife. They are lipophilic and persistent
organic compounds having a strong propensity for bioaccumulation and
biomagnifications in the aquatic and terrestrial food webs. The atmospheric transport: and
surface deposition of PBDEs is the primary means of distributing PBDEs over long
geographical distances. Once released into the air, PBDEs partition between the vapor
and particle bound phases in the atmosphere in accordance with their respective vapor
pressures. Lower brominated PBDEs primarily exist in the vapor phase, while higher
brominated congeners are primarily adsorbed to atmospheric particles. Photolytic
degradation is the primary environmental metabolic pathway, although anaerobic
degradation has been shown to occur in sediments. These abiotic degradation processes
can result in denomination, which the stripping of bromine atoms from the
polybrominated diphenyl ether molecule. Soils and sediments are the ultimate sinks for
PBDEs released into the environment.
xiv
-------
Several studies have shown that the higher brominated BDE congeners can
undergo biotic debromination. Several recent studies have provided evidence of
microbial mediated reductive debromination of decaBDE and octaBDE under laboratory
conditions. For example, in one study, Snlfurospirillum multivorans bacterium incubated
with deca-BDE has induced reductive debromination of deca-BDE in vitro to yield octa-
BDE and hepta-BDE after a contact time of 2 months. The octa- and hepta-BDE did not
further debrominatc in the presence of the microbe—even after one year. The researcher
of this study concluded that the microorganism was specific to the degradation of deca-
BDE and is incapable of dcbrominating lower brominated PBDE compounds. In another
study, BDE 209 was debrominated in vitro to yield BDEs 206, 207, and 208 by contact
with anaerobic mcsophilic microorganisms indigenous to raw sewage sludge (microbial
species not identified). Methane was formed as a product of microbial respiration, and
the amount of BDE 209 decreased by 30% within 238 days.
Debromination in fish, birds, and mammals have provided evidence that this
process occurs in vivo. Laboratory studies of rainbow trout, lake trout, and caip,
involving fish food spiked with pure BDE 209, have clearly shown accumulation of
lower brominated BDE congeners not initially present in the feed. This evidence is
suggestive of metabolic synthesis of lower brominated congeners through debromination
of BDE 209. In another study, Sprague-Dawley rats were fed a commercial formulation
of decaBDE, which is 98.5% BDE 209. Evidence of metabolic debromination of BDE
209 to lower congeners was observed from an apparent 160% increase in the tissue
concentration of BDEs 197, 201, and 207 in the sacrificed rats as compared to levels in
the feed. Additional evidence for in vivo metabolic debromination was found for
chickens, starlings, and even house cats.
The study on house cats involved measuring the PBDE congener profile in cat fat
and then also in the serum of house cats consuming dry food only, canned wet food only,
and a combination of diy and wet cat food. The contamination of PBDEs in dry cat food
reflected the congener profile of decaBDE, with BDE 209 representing 83-93% of total
PBDE present in the food. Because of the high content of BDE 209 in dry cat food. BDE
209 dominated serum in cats only consuming the dry food. BDE 209 accounting for
4.2%, 21%, and 30% of serum PBDE levels in house cats consuming canned-, mixed-.
xv
-------
and dry-food, respectively. It was noted that BDE 207 was consistently present in serum
in significant concentrations of the dry food eaters as compared to the consumers of the
other food types. BDE 207 accounted for 4.5%, 9.8%. and 17% of the PBDE levels
detected in cats consuming canned-, mixed-, and dry-food-eaters, respectively. BDE 207
was present in the dry food eaters at approximately 50% of the total concentration of
BDE 209 which is uncharacteristic of the decaBDE congener profile (BDE 207 is
approximately 1% of BDE congeners present in decaBDE), and was not the pattern
observed in the wet food eaters. Moreover, the ratio of BDE 207 to BDE 209 in cat
serum was relatively constant in all dry cat food eaters. The authors regarded these data
as possible evidence for the metabolic debromination of BDE 209 to form BDE 207.
PBDEs have captured the attention of scientists and policymakers because levels
in the environment and humans have increased rapidly since these chemicals came into
use in the 1960s and 70s. Environmental time-trends can be observed from lake sediment
core studies and archived, animal tissue samples. The sediment cores show a
predominance for, and a stark rise in, BDE 209, while the animal tissue samples in
general show a predominance and rise in BDE 47. The rise in PBDE concentrations in
human blood and breast milk in North Americans (both from the United States and
Canada) throughout the 1990s into the 2000s, coupled with the fact that North American
body burdens exceed those of Europeans and others by factors of 10 or more, has served
to focus attention on North American exposures to PBDEs. The penta PBDE formulation
has garnered the most concern because it appears to be the major contributor to current
environmental, biota, and human body levels. Even with both the penta and octa
formulations having been withdrawn from the United States market, past use and possible
debromination of higher brominated congeners (eg., BDE 209) by photolytic or
biological mechanisms to form lower brominated congeners might result in the continued
presence of lower brominated congeners in humans and the environment.
Studies measuring the concentrations of the BDE congeners in environmental and
exposure media concentrations were compiled and summarized with the ultimate goal of
selecting representative BDE congener concentrations in exposure media (air, dust, food,
etc.) to which Americans are exposed. While the data were insufficient to derive
concentrations that could be considered statistically representative of the general US
XVJ
-------
population, they were deemed adequate for conducting the exposure modeling done in
this report. These exposure media concentrations were combined with exposure contact
rates (air breathed, food eaten, etc.) in order to estimate an overall intake dose.
Concentrations and contact rates were used to characterize background central tendency
exposures. Exposure media concentrations were either the straight average or geometric
mean of the concentrations found in the study selected to represent national background
conditions. Contact rates were arithmetic averages for adult populations as provided in
the U.S. EPA's Exposure Factors Handbook.
As lipophilic contaminants, PBDEs bioaccumulate in the lipids of organisms. In
humans, lipid concentrations are typically measured in. blood and breast milk, although
adipose tissue concentrations have also been measured. Measured human concentrations
of BDE congeners were compiled and representative background concentrations of BDE
congeners in human blood and milk of Americans were selected. Then, a simple
pharmacokinetic model was used to predict the background lipid concentrations of the
BDE congeners in humans using the background intake estimates. These predictions
were compared to the selected representative body burden concentrations. The model
predictions matched the measurements fairly well (as discussed below), suggesting
confidence in the exposure characterization. Figure E-l displays this study approach.
This exposure exercise focused on the congeners most often studied and found in
the environment. It is noted, however, that there is not a final selection of toxic or
otherwise most critical, congeners as there is, for example, with dioxin-like compounds.
Some studies focus on as few as four congeners while others measure over 15 congeners.
The total concentrations found in the individual studies (including for discussions below)
can mean the sum of different congeners. Many studies have focused on the penta
formulation congeners and not measured the critical deca congener, BDE 209. Reasons
cited for not including this congener include the historical predominance of the other
congeners and the analytical difficulties associated with measuring BDE 209. The
congeners selected for final modeling in this exercise include 28, 47, 99, 100, 138, 153,
154, 183, and 209. Tables E-l and E-2 provide the results of this exposure exercise.
Table E-l displays the final selection of congener concentrations in exposure
media. Very few studies were found that characterized surface water and surface soil.
xvi i
-------
Surface water concentrations, used as a surrogate for drinking water, were measured in
the San Francisco estuary. The total water concentration was 146 pg/L. A total of 33
surface soil measurements were taken in 15 states and measured for 30 BDE congeners in
the single study found for the United States. Concentrations of total BDEs averaged 103
ng/g dry weight (dwt), with a geometric mean concentration of 5,3 ng/g dwt. Average
soil concentrations were used in this exercise. Outdoor air was characterized by a study
by the California Air Resources Board (CARB). The CARB data of 84 samples were
taken in 2004 from 7 monitors on 12 dates from locations in the Bay Area and the South
Coast. While the profile with a 158 pg/nr1 average might be higher than other outdoor
profiles, it is still a reasonable representation of urban or suburban conditions. The
indoor environment has been a focus of study because of the use of PBDEs in consumer
products found in the home; a number of studies measuring PBDEs in indoor air and
house dust were found. The profile in indoor house dust originated from a study from
dust taken in 10 homes in 9 di fferent states in the western part of the country. The total
of 8,275 ng/g dwt in that representative profile compares with 5,811 ng/g dwt from a
study of 17 homes in Washington, DC area, and 9,271 ng/g dwt from 2 samples from a
computer lab in California. The geometric means of three locations (bedroom, living
room, and from a household vacuum) within 20 homes in Boston of 6,332, 13,882, and
4,213 ng/g, respectively also compare well to the 8,275 ng/g dwt estimate. The indoor air
profile came from this study of Boston homes in which the authors determined geometric
mean concentrations for 3 locations (personal, bedroom, living room). The average of
the 3 geometric means, 447 pg total BDEs/nr1, was used as the indoor air concentration.
Numerous studies on concentrations in food were summarized. A wealth of
studies on fish in the wild, raised in aquaculture, and purchased in the market place
showed a wide range in concentrations. Generally, concentrations in the wild were found
to be higher, and, in some studies, substantially higher than farm-raised or store-bought
fish. Essentially all of the representative food profiles, including the fish profiles,
originated from a single study sampling food from the retail market place in Texas.
While limited geographically, the market basket survey includes 62 samples, split
between samples of meat, dairy, and fish, taken from several supermarkets in Dallas.
Thirteen congeners, importantly including BDE 209, were measured. The average total
XVUl
-------
concentrations (not lipid weight based) used from this study include the following: 1.17
ng/g in finftsh, 0.13 ng/g in beef, 0.28 ng/g in pork, 0.36 ng/'g in poultry, 0.11 ng/g in
dairy, and 0.09 ng/g in eggs. A total concentration of 3.6 ng/g in shellfish originated
from the study in the San Francisco estuary, which also measured surface water
concentrations.
Body-burden studies were compiled and reviewed. Nearly all body-burden data is
from human blood and breast milk. Tabic E-2 shows the final selected profiles of BDE
congeners in blood and human breast milk of Americans. Body burdens of Americans
are higher than body burdens of individuals in other countries; these data suggest total
PBDE body burdens in the range of 30 to 100 ng/'g lipid weight (lwt) for Americans,
while body burdens of less than 10 ng/g lwt have been found for people in other
countries. Most of the data outside of the United States is from Europe. The study
selected as the representative blood profile came from the National Health and
Nutritional Examination Study (NHANES) from 2003/4. The geometric mean
concentration of all adults from this study wras 36 ng/g lwt. The predominant congener
found in body burden studies is BDE 47, explaining about 50% of the total concentration.
The next most abundant BDE congeners ate 99 and 153, both explaining the range of 10-
20% of total concentrations. Most of the studies have not measured BDE 209, including
the N1TANES study, but, when measured, it was found in about half the samples at low-
levels near 1-2 ng/g lwt, with the exception of the one case. In this case study, which
entailed a family of 4 including two parents and two young children, concentrations of
BDE 209 were above 100 ng/g lwt in the children. Low levels of BDE 209 have been
attributed to the rapid half-lite of 15 days in humans, and the higher levels in children
from this study were attributed to dust exposures in the house.
Unlike human blood data, there is no nationally representative breast milk study
that could be used to characterize concentrations of PBDEs. Of the several studies
evaluated, perhaps three studies could be used to represent background conditions. These
include one by the Environmental Working Group (EWG), which sampled 20 women
from around the United States; the Northwest Environment Watch (NEW) study, which
had a sample size of 40 including women residing in several states in the Northwest; and
perhaps a third study of 47 samples taken in Texas. The NEW study, which importantly
xix
-------
includes measurements of BDE 209, was used. The median total BDE concentration of
the 40 samples was 44.1 ng/g Iwt,
The first-order, single-compartment pharmacokinetic model used in this exercise
to convert intakes to body lipid concentrations requires the half-life elimination rates of
the BDE congeners and BDE absorption fractions (i.e., fractions of BDE intakes from
ingestion of dust, water, and foods, as well as inhaled BDEs, that arc absorbed in the
stomach or the lungs to accumulate in body lipids) as inputs. Only two studies were
found that provided human half-lives, with one deriving half-lives ranging from 2,9 to
11.0 yr for BDE congeners 47, 99, 100, 138, 153, and 154. The second reference derived
significantly smaller half-lives for the higher brominated congeners BDEs 183 and 209—
half-lives of 0.26 and 0.041 yr, respectively. The absorption for BDEs in dust ranged
from 0.04 ( for BDE 209 ) to 0.78 (for BDE 100). These were deri ved from a study of the
bioavailability of BDEs in dust when fed to rats. The absorption values for food and
water were near 0,90 for all of the congeners. The soil-dermal contact pathway required
a fraction of PBDE absorbed through the skin, set at 0.03 for all congeners, as well as
parameters reflecting adherence of soil contacted by the skin, contact events, and contact
surface areas.
Table E-2 provides the final results from this exposure assessment. As seen there,
the exposure pathways of dust ingestion and dust-dermal contact dominated total adult
exposure—explaining over 80% of the 540 ng/day total. It is also seen that the
congeners BDEs 47, 99, and 209 dominated the exposure intake: each explained about
27% of total intakes, with other congeners explaining the remaining intakes. The
predicted concentration of total BDEs of 35.9 ng/g lwt was similar to the NHANES
human blood measurement of 36. 4 ng/g lwt and the breast milk concentration of 44.1
ng/g lwt. Predictions appear reasonably close to measurements for 7 of 9 congeners. The
prediction of BDE 47 at 10.0 ng/g lwt did not appear to match the observed
measurements of 20.5 ng/g lwt in blood and 26.0 ng/g lwt in milk. The cause for the
underprediction of BDE, 47 is not known, but it could very easily be the assumed half-life
in humans. At 3.0 years, BDE-47 was eliminated nearly twice as fast as BDE 99, which
was assigned a half-life of 5.4 years. Had it been assigned an elimination half-life of 10
years, the prediction would jump to greater than 30 ng/g lwt, now twice as high as BDE
xx
-------
99, which is in line with measurements. On the other hand, BDE 99 was predicted at
14,6 ng/g hvt, which is higher than measurements of 5,0 (in blood) and 5.4 (in breast
milk) ng/g lwt. It is interesting that the calculated intake doses of BDEs 47 and 99 were
comparable, but that the observed body burdens of BDE 47 were so much higher than
BDE 99, In terms of modeling, BDE 99 was overpredicted while BDE 47 was
undcrpredieted. It was suggested that perhaps this result could be due to improper
assignment of elimination rates in the body, but also, this result could suggest
denomination of BDE 99 in the body. In other words, the modeling may correctly
determine body burdens due to intakes, but the modeling might not account for how these
body burdens could change over time due to internal metabolism. However, this is much
too speculative given current information on human and animal metabolism.
Exposures for key populations were also examined. These unique exposures
included occupational, childhood, infant via breast feeding, and unusually high exposures
at the high end of the general population. Findings from these examinations include:
1. Occupational: Limited studies from Sweden and from China suggest that
PBDE concentrations are elevated in occupational groups exposed to likely sources of
PBDEs. One study, which looked at incinerator workers in comparison to general
population exposures, did not find a difference. The only study of occupational
exposures in the United States found a significantly higher (p < 0.05) level of PBDEs in
workers in foam recycling facilities and individuals who installed carpet padding
manufactured from recycled foam. Specifically, the median total PBDE concentrations
(which didn't include BDE 209) in the foam workers, carpet installers, and control were
160, 178, and 19 ng/g lipid, respectively,
2. Children: Dose intakes for children were derived in a similar manner as
adults: exposure media concentrations were combined with contact rates for specific age
ranges of children. The total dose for the three age ranges of children modeled were 751
ng/day for the 1 -5 age range, 439 ng/day for the 6-11 range, and 536 ng/day for the 12-19
age range. On a body-weight basis, the doses are 50.1 ng/kg-day for ages 1-5 (assuming
15 kg bw), 14.6 ng/kg-d for 6-11 (30 kg), and 9.2 ng/kg-d for 12-19 (58 kg). The much
higher dose for the child age 1-5 was due to the doubling of soil/dust ingestion from 50
xxi
-------
nig day for all other age ranges to 100 mg/day. These child exposures compare to the
adult dose of 547 ng/day, or 7,8 ng/kg-day, assuming a 70 kg adult.
3. Infants Via Breast Milk: Intakes of PBDEs via breast milk were input into the
pharmacokinetic model to predict infant body burdens impacts. The key assumptions for
intake include 0.8 L/day breast milk ingestion, 4% fat in breast milk, and the 44.1 ng total
BDE/g lipid concentration in breast milk, as described earlier. With these assumptions,
the daily the total dose to infants is 1,411 ng/day. Assuming an average body weight of
10 kg for an infant during the months of breast-feeding, a dose is calculated as 141 ng/kg-
day. Two assumptions on elimination half-life were used in the pharmacokinetic
simulations: one a rapid half-life, on the order of weeks, and the second more typical of
the half-lives used for the lower brominated BDEs in adults, on the order of years. A
short half-life has been shown to be appropriate for dioxin-like compounds in infants,
while adult half-lives for dioxin have been measured in the order of years, which is why
these two half-lives were tested in infants. With the short half-life, lipid concentrations
were modeled to rise to over 200 ng/g lwt through age 5. to then drop gradually to below
100 ng/g lwt by age 19. When an adult-like half-life of 6 years was used in this infant
model, concentrations rose to nearly 200 ng/g lwt by age I, continued to rise to about 325
ng/g lwt by age 5, to then drop to below 100 ng/g lwt by age 19. Only one study was
available with which to compare these results; this study from a family in California was
earlier cited because of the finding of BDE 209 in the children. Total PBDE
concentrations in the 18-month old toddler rose to above 400 ng/g lwt, and concentrations
in the 5 year-old were above 200 ng/g lwt. At the same time, the adult concentrations in
the family were near 100 ng/g lwt. While insufficient to provide verification of either
modeling assumption on half-life in infants and children, when combined, this ancillary
data along with the pharmacokinetic modeling, support the conclusion that infant and
childhood body burdens could very possibly be significantly higher than that of adults.
4. Unusually High Exposures in the General Population: The 2003/2004
NHANES results show that the 95dl percentile is more than 10 times higher than the
median, and most surprising, the maximum found in the survey is 100 times higher than
the median. In contrast, the same information was sought on NH ANES results of dioxin-
like compounds, and it was found that the 95% is about 4 times the median, and the
xxn
-------
maximum found is only about 12 times the median. This same trend for PBDEs was
found in other blood and adipose tissue surveys, and also in indoor dust studies; that is,
that levels at the high end of the distribution appear substantially higher than the median
of the distribution. This suggests that unusually high body burdens in the general
population could result from exposures to unusually high dust concentrations.
While predictions of adult and even infant body burdens encouragingly are close
to observations, uncertainties exist in the intake dose estimates and the pharmacokinetic
modeling, starting from development of dose estimates based on limited environmental
measurements, to indoor contact rates with house dust, to the pharmacokinetic parameters
of absorption and elimination half-life. Contact rates for food/water ingestion and
inhalation are fairly well established, and the exposure media concentration summaries
suggest similarities among different studies. Therefore, the dose via food/water ingestion
and inhalation might be considered reasonably certain, for purposes of this discussion.
However, using the pharmacokinetic model, food/water ingestion and inhalation
explained less than 20% of the body burden. It was assumed that the remainder of the
exposures came from house dust through the pathways of ingestion and dermal contact.
Assignment of dust contact rates in combination with concentrations found in housedust
lead to exposures that do result in accurate predictions of body burdens. The claim is not
made that this "proves" the importance of indoor dust to overall general population
exposures. However, circumstantial evidence supporting this modeling are the high
concentrations found in United States house dust, particularly in comparison to house
dust concentrations from other countries. Specifically, house dust concentrations in
European studies were found to be lower than in the United States, by one order of
magnitude or more, and one hypothesis was that the difference in European and United
States body burdens (European body burdens are much lower than United States body
burdens) is due to exposure of Americans to high concentrations of BDEs in house dust.
Still, there was no "proof that contact with house dust explains the majority of body
burdens of Americans. Nonetheless, the overall weight of evidence of this exercise
supports the finding that the bulk of United States exposures occur in the indoor
environment through contact with house dust. The exercise suggests these exposures
xxiii
-------
account for between 80 and 90% of total exposures, with the remainder due primarily to
food ingestion. Nonetheless, more research is recommended to verify these findings and
better quantify the uncertainties that have been identified.
xxiv
-------
Table E-l. Representative exposure media concentrations
Exposure
Congener
Number
Media
28
47
99
100
138
153
154
183
209
Total
W ater, pg/1
3,3
42.7
27.6
7.2
0.3
3.9
2.9
4.4
42.3
146.1
Surface soil,
—
1.9
3.6
0.4
—
5.7
4.8
37.4
15.3
82.3
ng/g dwt
Indoor dust,
ng/g
ND
1857
2352
911
181
243
156
60
2394
8275
Outdoor air,
3
53
51
13
—
4
4
I
25
158
pg/nr
Indoor air,
27
177
79
16
—
5
7
—
121
447
pg/mJ
Shellfish, ng/g
ND
3.6
1.2
0.9
ND
ND
ND
ND
ND
5.7
wwt
Finfish, ng/g
0.03
0.60
0.17
0.13
0.001
0.02
0.05
0.002
0.09
1.17
wwt
Beef, ng/g wwt
0.02
0.05
0.04
0.006
0.0001
0.006
0.004
0.001
0.003
0.13
Pork, ng/g wwt
ND
0.08
0.12
0.015
0.001
0.02
0.01
0.009
0.02
0.28
Poultry, ng/g
0.0002
0.06
0.12
0.03
0.002
0.02
001
002
0.12
0.36
wwt
Dairy, ng/g wwt
0.0002
0.03
0.03
0.005
<0.0001
0.004
0.002
0.002
0.04
0.11
Eggs, ng/g wwt
0.0002
0.02
0.04
0.006
0.0001
0.004
0.003
0.0001
0.01
0.09
notes:
The "totals" may include congeners other than those listed and is greater than the sum of
the studied congeners; see Chapter 4 for more detail
— means no data available; ND means data available but not detected; wwt = wet weight
xxv
-------
Table E-2. Predicted doses and lipid-based concentrations ofBDEs, with predicted
concentrations compared with lipid-based measurements in blood and milk
Exposure Doses,
Pharmacokinetic
Results
Congener Number
TOTAL
28
47
99
100
138
153
154
183
209
I. Exposure Doses
soil ingestion,
ng/d
0.00
81.25
102.92
¦w 86
7.92
10.67
6.86
2.86
104.83
357.17
soil dermal
contact, ng/d
0.00
19.50
24.70
9.57
1.90
2.56
1.65
0.69
25.16
85.73
Inhalation, ng/d
0.32
2.15
1.00
0.21
0.00
0.06
0.09
0.00
1.45
5.28
Food & water
ingestion, ng/d
1.57
35.30
24.47
8.18
0.14
2.91
1.6
0.85
16.3
92.24
II. Pharmacokinetic Results
Predicted cone,
ng/g hvt
0.3
10.0
14,6
4.2
1.3
4.6
0.8
0.02
0.1
35.9
Observed blood,
ng/g hvt
1.2
20.5
5.0
3.9
NA
5.7
NA
NA
N'A
36.3
Observed milk,
ne/t! Iwt
1.7
26.0
5.4
5.2
NA
4.8
0.4
0.2
0.4
44.1
xxvi
-------
Figure E-l. Approach for characterizing exposure to polybroininated diphenyl ethers in
this report.
Compare
Compile data on
body burdens
Compile data on
exposure media
concentrations
Develop representative
congener profiles in
the media
Develop representative
body burden congener
profiles
Combine media concentrations
with exposure factors to
estimate intake dose
Convert intake dose to
Body burden using simple
1st order 1 -compartment
PK model
-------
Chapter 1 Introduction
Polybrominatcd diphcnyl ethers (PBDEs) arc a class of brominated flame
retardants that are added to plastics, poiyurethane foam, textiles, and electronic
equipment to reduce the likelihood of ignition and to slow the burn rate if the products do
catch fire. PBDEs have a common structure of a brominated diphenyl ether molecule
that may have anywhere from 1 to 10 bromine (Br) atoms attached. Depending on the
location and number of Br atoms, there are a possible 209 PBDE compounds, each
termed congeners. Each are assigned a specific BDE number (note: in this document,
the abbreviation PBDE will be used to denote the class of brominated flame retardants,
while BDE will be used in the context of PBDE congeners). For example, there are 42
tetrabromodiphenyl ether congeners (those with 4 bromine atoms), but only a few of
them, specifically BDE 47 and occasionally BDE 66, are found in the product
formulations and in environmental or exposure media (La Guardia, ct al., 2006). Table 1 -
1 shows the BDE congener number and chemical composition of the most commonly
studied BDE congeners.
PBDEs have been marketed in three primary formulations: the pentaBDE
formulation, commercially known as DE-71 and Bromkal 7(H5DE; the octaBDE
formulation, DE-79; and the decaBDE formulation, DE-83R. or Saytex 102E. The
formulations differ in their composition of BDE congeners. The penta formulation is
dominated (by weight) by penta congeners (50-62%) with secondary contributions by
tetra (24-38%) and hexa congeners (4-12%). The octa formulation is dominated by hepta
(45%) and octa congeners (33%), with secondary contributions by hexa (12%) and nona
(10%) congeners. The deca formulation is composed of essentially all BDE 209 (97-
99%, with 1-3% other, mainly nona, congeners), which is the congener with all ten Br
positions occupied. The penta and octa formulations were voluntarily withdrawn from
the United States (U.S.) marketplace by their manufacturers at the end of 2004, leaving
only the deca formulation for use in commercial products. The deca formula is also the
only currently used formulation in Europe because the penta and octa formulations were
banned. However, Sweden banned the use of the deca formulation in August of 2006, the
1-1
-------
ban to take effect January 1, 2007 (see:
http://www.emfacts.com/webIog/index.php?p=547).
PBDEs have captured the attention of scientists and policymakers because levels
in the environment and humans have increased rapidly since these chemicals came into
use. The rise in PBDE concentrations in blood and breast milk samples both from the
U.S. and Canada throughout the 1990s into the 2000s, coupled with the fact that North
American body burdens exceed those of Europeans and others by factors of 10 or more,
has served to focus attention on North American exposures to PBDEs. The penta PBDE
formulation has garnered the most concern because it appears to be the major contributor
to current environmental and human body levels. Even with both the penta and octa
formulations having been withdrawn from the U.S. market, past use and also the
possibility of denomination (loss of bromine atoms) of BDE 209 and other higher
brominated congeners by photolytic or biological mechanisms to form lower brominated
congeners could result in the continued presence of lower brominated congeners in the
environment.
Studies have been conducted in laboratory animals to gain a better understanding
of the potential health risks of PBDEs. These studies have suggested potential concerns
about liver toxicity, thyroid toxicity, developmental toxicity, and developmental
neurotoxicity—especially about potential risks in children. To date, there is no evidence
of carcinogenicity from exposure to any of the PBDEs except decabromodiphenyl ether,
BDE 209. In a review of toxicological studies as part of a Toxicological Review for the
U.S. Environmental Protection Agency (U.S. EPA)'s Integrated Risk Information System
(IRIS), the U.S. EPA (EPA, 2006) has found "suggestive evidence of carcinogenic
potential" for BDE 209, according to the 2005 EPA draft Guidelines for Carcinogenic
Risk Assessment (EPA, 2005).
The U.S. EPA has formed a working group comprised of individuals from several
program offices including the Offices of Pesticides, Prevention, and Toxic Substances,
the Office of Water, and of the Office of Research and Development, to study issues
surrounding polybrominated diphenyl ethers (see http://www.epa.gov/oppt/pbde). They
issued a project plan in 2006 that outlined projects to further study the toxicity,
environmental fate, and exposure to PBDE compounds. This document addresses the
1-2
-------
exposure assessment needs identified in that project plan and provides a comprehensive
assessment of the exposure of Americans to this class of persistent organic pollutants.
Subsequent chapters describe the historical use and composition of commercial
mixtures ofPBDEs (see Chapter 2), the environmental fate of PBDEs (see Chapter 3),
and environmental and exposure media concentrations of PBDEs (see Chapter 4). The
document concludes with a comprehensive exposure assessment addressing infant,
children, and adult exposures to PBDEs (see Chapter 5),
REFERENCES
EPA. 2006. Toxicological Review of Decabromodiphenyl Ether (BDE 209) (CASRN
1163-19-5) In Support of Summary Information on the Integrated Risk Information
System (IRIS). December, 2006. posted on, http://www.epa.mv/iris.
EPA. 2005, Guidelines for carcinogen risk assessment. Fed Regist 70:17765-18717.
Available online at http://vvvvw.epa.gov/cancerguidelines.
La Guardia, MJ, Hale RC, Harvey E. Detailed polybrominated diphenyl ether (PBDE)
congener composition of the widely used penta-, octa-, and deca-PBDE technical flame-
retardant mixtures. Env. Sci. and Tech. 2006: 40: 6247-6254.
1-3
-------
Table 1-1. BDE congener numbers and chemical composition of the most commonly
studied BDE congeners
BDE Congener
Number
Chemical
Formula
BDE Congener
Number
Chemical Formula
1. MonoBDE
VI HexaBDE
BDE 3
4-BDE
BDE 138
2,2',3,4,4',5'-BDE
II. BiBDE
BDE 140
2,2\3,4,4\6-BDE
BDE 7
2,4-BDE
BDE 153
2,2',4,4',5,5'-BDE
BDE 8
2,4'-BDE
BDE 154
2,2\4,4',5,6'-BDE
BDE 11
3,3'-BDE
BDE-155
2,2\4,4',6,6'-BDE
BDE 12
2,6-BDE
BDE 166
2,3,4,4',5,6-BDE
BDE 13
3,4'-BDE
VII. HeptaBDE
BDE 15
4,4'-BDE
BDE-181
2,2\3,4,4\5,6-BDE
III. TriBDE
BDE-183
2,2'.3,4,4',5',6-BDE
BDE 17
2,2',4-BDE
BDE-190
2,3,3',4,4',5,6-BDE
BDE 25
2,3',4-BDE
VIII. OctaBDE
BDE 28
2,4,4'-BDE
BDE-196
2,2',3,3',4,4',5',6-BDE
BDE 30
2,4,6-BDE
BDE-197
2,2',3,3',4,4',6,6'-BDE
BDE 32
2,4',6-BDE
BDE-203
2,2', 3,4,4\5,5',6-BDE
BDE 33
2',3,4-BDE
VIII. Nona BDE
BDE-35
3,3',4-BDE
BDE-206
2,2',3,3',4,4',5,5',6-BDE
BDE 37
3,4,4'-BDE
BDE-207
2,2\3,3',4,4\5,6,6'-BDE
IV. TetraBDE
BDE-208
2,2',3,3',4,5,5',6,6'-BDE
BDE 47
2,2'.4,4'-BDE
VIII. DecaBDE
BDE 49
2,2\4,5'-BDE
BDE-209
2,2',3,3',4,4',5,5',6,6'-BDE
BDE 66
2,3',4.4'-BDE
BDE 71
2,3',4',6-BDE
BDE 75
2,4,4',6-BDE
BDE-77
3,3',4,4'-BDE
V. PentaBDE
BDE 85
2,2\3,4,4'-BDE
BDE 99
2,2',4,4',5-BDE
BDE 100
2,2',4,4',6-BDE
BDE 105
2.3,3',4.4'-BDE
BDE 116
2,3,4,5,6-BDE
BDE 118
2,3',4,4',5-BDE
BDE 119
2.3',4,4'6-BDE
BDE 126
3,3',4,4',5-BDE
1-4
-------
2. PRODUCTION, USE, AND LIFECYCLE OF
POLYBROMINATED DIPHENYL ETHERS
2.1 INTRODUCTION
Each year, structural and vehicle fires result in several thousand deaths and
injuries and billions of dollars in property loss in the United States (U.S. Fire
Administration, 2004). Fire safety and prevention is a focal point of public policy.
Largely as a result of using fire-resistant materials in the construction of consumer goods,
plastics, and textiles, the incidence of fires has decreased 24% over the past 10 years
(U.S. Fire Administration, 2004). The use of brominated fire retardants in these materials
has played a major role in fire prevention (IPCS, 1994). Brominated fire retardants are in
widespread use and are now detected in the tissues of humans and wildlife, in soils, in
sediments, and in air. The ubiquitous presence of brominated fire retardants in the
environment has heightened concerns for adverse ecological and human health risks.
This chapter describes the production, use, and lifecycle of PBDEs. The information in
this chapter contributes to the understanding of the pathways leading to environmental
contamination.
PBDEs are a class of aromatic compounds intentionally manufactured to retard
the combustibility of treated materials. When fire occurs, the PBDE formulations utilize
gas-phase chemical reactions that interfere with the combustion process, thus delaying
ignition and inhibiting the spread of fire (D'Silva, 2004). These characteristics have
promoted the widespread use of PBDEs in textiles, flexible polyurethane foams used in
upholstery stuffing for furniture and car seats, electronic components, electrical
components, and plastics used in the casings of TVs, PCs and other electronic equipment.
The purpose of this chapter is to describe the production and uses of PBDEs, and
to present a basic lifecycle analysis of PBDEs involving estimates of environmental
release from production, product use, waste disposal, recycling of electronics and
electrical equipment (EEE) materials, and sewage treatment. Section 2.2 reviews the
production of commercial PBDE formulations. Section 2.3 surveys the specific uses of
PBDEs as fire retardants in a number of plastic resins and finished products. Section 2.4
2-1
-------
presents an initial lifecycle analysis that derives estimates of environmental releases of
specific BDE congeners based on production, the uses of PBDE-treated products, and the
disposal of PBDE-treated products once they become functionally obsolete. Because of
significant limitations in existing data, the lifecycle analysis presented in this chapter
should be regarded as a preliminary or initial assessment. That being said, the lifecycle
analysis does provide a platform from which distinct observations can be made regarding
the potential pathways of environmental releases of PBDE formulations and their BDE
congeners into the land, air and water of the United States from production through
disposal of PBDE treated materials.
2.2 PRODUCTION
Commercial production of PBDEs began in 1976 (IPCS, 1994). PBDEs have
been sold under various trade names but mainly consist of three commercial mixture
formulations: pentaBDE, octaBDE, and decaBDE. Each commercial formulation is
manufactured through the chemical reaction of bromine with diphenyl oxide and/or
diphenyl ether in the presence of an inorganic catalyst (e.g., Aid,;) (ATSDR, 2004). The
bromine amount and the time allotted for the chemical reaction control the extent of
bromination on the diphenyl ether molecule. The stepw ise addition of bromine causes
the formation of lower to higher chlorinated PDBE congeners until the total desired
amount of bromination is obtained. Figure 2.1 displays the general structure of the PBDE
compound. The molecular backbone consists of two phenyl rings interconnected by an
oxygen atom. There arc 10 positions whereby a bromine atom can substitute a hydrogen
atom on the molecule with the possibility of ten homologue groups identified by the
preface mono-, di-, tri-, tetra-, penta-, hexa-, hepta-, octa-, nona-, and decabrominated
diphenyl ether. Each congener is indicated by the positional numbering of the bromine
atom on the biphenyl rings. The congeners are assigned a number according to the
nomenclature of the International Union of Pure and Applied Chemistry (IUPAC).
Commercial formulations are mixtures of PBDE congeners with pentaBDE,
octaBDE and decaBDE having a bromine content of about 70.8%, 79%, and 83%,
respectively (European Union, 2001; 2002; 2003). Although 209 PBDE congeners are
theoretically possible, only a limited number have been detected in commercial flame
2-2
-------
retardant products. Table 2.1 is a compilation of the approximate BDE congener
compositions of the three commercial formulations. The dominant congeners in
pentaBDE (percent weight basis in parenthesis) are BDE 99 (35-50%), BDE 47 (25-
37%), BDE 100 (6-10 %), BDE 153 (3-5%), and BDE 154 (2-4%). On a homologue
basis, the pentaBDE formulation is dominated by pentaBDE (50-62%) and tetra-BDE
(24-38%)) homologues, with hexa-BDE amounting to only 4-12% of total mass.
The BDE congeners present in oetaBDE formulations are BDE 183 (40%), BDE
197 (21%), BDE 203 (5-35%), BDE 196 (8%), BDE 208 (10%), BDE 207 (7%), BDE
153 (5—10%), BDE 154 (1-5%), BDE 209 (not detected up to 3%), and BDE 190 (1%).
Hexa-, hepta-, octa-, and nona-BDE homologues dominate the oetaBDE commercial
formulations comprising 10-12%, 43-58%, 26-35%, and 8-14% of total BDE content,
respectively. With respect to decaBDE, 97-98% of the formulation is comprised of BDE
209, with the remainder being BDE 207 and BDE 208. BDE 209 is the dominant
congener, and it is a signature of this formulation.
As a result of mergers and acquisitions, PBDE-fire-retardant formulations are
produced by three chemical manufacturers worldwide: Albemarle, Chemtura and IC'L
Industrial Products. In the United States, all PBDE-fire-retardant products sold are
currently decaBDE formulations. Prior to 2005, the industry was more diverse. The
Dead Sea Bromine Group (DSBG), based in Israel, was a leading manufacturer of PBDE
fire retardants. ICL Industrial Products acquired the DSBG in 2005. Currently the major
producer of PBDE products in the United States is the Great Lakes Chemical
Corporation, which sold fire retardants under the Firemaster1M brand name. The Great
Lakes Chemical Corporation merged with Crompton Chemical to form Chemtura.
Chemtura produces flame retardant products comprised of decaBDE under the brand
names AZUB DB-40, AZUB DB-65, AZUB 2DA-65, and AZUB 3DA-65. Prior to 2005
the Great Lakes Chemical Corporation produced a pentaBDE based product under the
brand name DE-60F. DE-60F was replaced with Firemaster-550, a phosphorous-bromine
based fire retardant. Firemaster-550 is technically not enriched with PBDE, but the exact
active chemical ingredients remain proprietary. Albemarle produces a decaBDE
formulated fire retardant under the brand name SAYTEX 102E. Because of increased
international regulatory concern regarding the potential for human health and
2-3
-------
environmental effects, the PBDE industry voluntary ceased production of penta- and
octaBDE on December 31, 2004 (BSEF, 2006a).
Approximately 56,418 metric tons (MT) of PBDEs were consumed worldwide in
2003. Global consumption for prior years was 67,440 (2001), 67,225 (2000), and 40.000
MT (1999). Two-thirds of all PBDEs produced are consumed in the Americas and Asia.
In 2001, 83% of all PBDE consumed worldwide was decaBDE, followed by pentaBDE
(11 %), and octaBDE (6%). Within the Americas, the breakdown of the market demand
for deca-, penta-, and octaBDE in 2001 was 74%, 21.5%, and 4.5%, respectively. Prior
to 2004, which was when production of these PBDEs ceased; Americans used
approximately 95% of the global production of pentaBDE, 40% of octaBDE, and 44% of
decaBDE (BSEF, 2006a).
Concerns for environmental persistence, the potential for bioaccumulation into
ecological and human food chains, and the widespread occurrence of PBDEs in the
environment have precipitated regulatory and voluntary actions to further reduce the use
of PBDEs. The sole United States manufacturer of pentaBDE and octaBDE phased out
production of these products at the end of 2004. As well, several states enacted
individual legislation banning the use of pentaBDE, octaBDE and deca formulations as
summarized below.
• As of June 1, 2006 California and Maine prohibited the manufacture, distribution, and
processing of products containing pentaBDE and octaBDE flame retardants (CalEPA,
2006; NCEL, 2005).
• As of January 1, 2006, Hawaii, Illinois, and Michigan prohibited the manufacture,
processing, and distribution of products containing more than 0.1% of the PBDE
formulations pentaBDE and octaBDE (Illinois EPA, 2006; NCEL, 2005).
• Minnesota similarly prohibited the manufacture and uses of penta and octaBDE
effective January 1, 2008 (NCEL, 2007).
• Washington State banned the uses of penta and octaBDE and of decaBDE in
mattresses effective January 1, 2008 (Washington State, 2007). However, the
Washington State regulation does not prohibit the use of decaBDE in residential
upholstered furniture, televisions, or computers with electronic enclosures until
January 1, 2011.
-------
• Beginning October I, 2008, Maryland will prohibit the manufacture, processing, and
distribution of products containing more than 0.1% of the PBDE formulations of
pentaBDE and octaBDE (NCEL, 2007).
• As of August, 2004, New York prohibited the manufacture and processing, but not
the sale or use, of pentaBDE and octaBDE (NC'EL, 2005).
Internationally, other countries and governmental entities have banned the use of
penta. octa and decaBDE. The European Union (EU) banned the marketing and use of
pentaBDE and octaBDE effective February 6, 2003 with Directive 2003/11/EC of the
European Parliament and Counsel (BSEF, 2006a). At the time, the EU exempted
decaBDE from the ban. However, on April I, 2008 the European Court of Justice (the
EU's highest court) ruled that this decision to exempt decaBDE was flawed and ordered
the EU to ban the use of the decaBDE in electrical and electronic equipment (European
Court of Justice, 2008). Based on the directive of the court, the ban became effective on
July 1, 2008. The ban applies to all EU member countries. Australia banned the use and
import of pentaBDE effective March 6, 2007 (Australian Government, 2007). However,
it is expected that the use of recycled plastics containing PBDE formulations may
continue to incorporate these chemicals in the manufacture of new electronics and JT
units (Morf et al., 2005).
2.3. USES OF PBDEs
Due to the phase out by the manufacturer, as well as the state-specific bans noted
above, penta- and octaBDE are no longer used in products to inhibit flammability. For
the most part, decaBDE is not prohibited .from production and continues to be used. This
section reviews the past uses of pentaBDE and octaBDE, and the current uses of
decaBDE.
Prior to the end of production in 2004, approximately 95% of pentaBDE was used
as an additive fire retardant in flexible polyurethane foam (FPUF) materials (European
Union. 2001). The treated FPUF was used as seat cushioning and backing material for
domestic furniture; in bedding mattresses and in cushioning for automobile seats and
laminated automobile seat headrests (ATSDR, 2004). The remaining 5% use of
2-5
-------
pentaBDE was in the treatment of foam-based packaging materials and carpet padding.
Only 7.5% of the approximate 953,000 MT of FPUF produced annually in the United
States was treated with pentaBDE (ATSDR, 2004). Typically the pcntaBDE was mixed
with aromatic phosphate esters in a ratio of 3:1 prior to application to FPUF. Other past
uses of pentaBDE were in textile fabrics used in upholstery for furniture and automobile
seat covers; epoxy resins used as protective coatings to circuit boards; unsaturated
polyesters; paper laminates; flexible polyvinyl chloride used as electrical wire coatings;
rubber; paints and lacquers; rigid polyurethane foam; and adhesives. However,
pentaBDE was not added to acrylonitrile-butadiene-styrene (ABS)-based plastics used in
the manufacture of casings to television sets, computers, hairdryers, and automotive parts
(IPCS, 1994; ATSDR, 2004; Rahman et al„ 2001).
Mixed with antimony trioxide, octaBDE was primarily used as an additive flame
retardant to certain plastics and textiles. Approximately 95% of the use of octaBDE was
as an additive flame retardant in the production of ABS-based plasties, with the
remaining 5% used as an additive to high impact polystyrene (1I1PS), polybutylene
terephthalate (PBT), polyamide polymers, polycarbonate, nylon, polyolefin, and phenol-
formaldehyde resins (ATSDR, 2004; European Union, 2003; IPCS, 1994). OctaBDE
was typically added to ABS at a loading of 12-15% weight of the final product
(European Union, 2003). Materials that were treated with octaBDE include ABS plastics
in television sets, remote controls, facsimile machines, copiers, wire and cables, personal
computers (PC), PC monitors, keyboards, scanners, audio equipment, video equipment,
power adaptors, automobile parts, mobile phones, and kitchen appliances.
DecaBDE is used as a general purpose additive flame retardant to a wide array of
plastics having many product applications. HIPS, polyethylene (PE), polypropylene
(PP), PBT, and unsaturated polyesters (UPE) are common plastics treated with decaBDE
(A lace et a!., 2003; BSEF, 2006b). A major use of decaBDE in the United States is as an
additive flame retardant to HIPS. HIPS plastics are used in the manufacture of housings
and back panels to televisions, in casings to audio and video equipment, mobile phones,
remote controls, personal computers and PC monitors. The PE plastics are used in wire
and cables to electrical equipment. The PP-based plastics are used in communication
cables, capacitor films, building cables, pipes, stadium seats, lamp sockets and holders.
2-6
-------
and kitchen hoods. The PBT plastics arc used as connectors in electrical and electronic
equipment. UPE is used in building and construction materials as reinforced plastic
panels. DecaBDE is also added as a flame retardant to nylon, and to upholstery textiles
used in sofas, chairs, and office furniture. The decaBDE-treated nylon is used as
connectors in electrical and electronic equipment, circuit breakers, and coils. DecaBDE
is added to polymers in concentrations of 10-15%, by weight, and in conjunction with
antimony trioxide (Directorate-General Environment. 2005). A significant use of
decaBDE in the United States is the fire retardation of the back panels and the HIPS
plastic casings of television sets. In 2003, approximately 17,150 MT of decaBDE was
used in 28 million television sets sold in the United States (Pure Strategies, Inc, 2005).
2.4. LIFECYCLE OF PBDE
In general, the lifecyclc of PBDE, from production to eventual disposal, includes
activities that can cause the environmental releases of the chemical. It begins with the
production of PBDE commercial formulations, then proceeds to applications to textiles
and polymers, to polymer applications in consumer goods, to the use of consumer goods
treated with PBDE fire retardants, and to the final disposal of PBDE treated consumer
goods as products reach obsolescence. Figure 2-2 portrays the idealized lifecycle of
PBDE. Each element of the lifecycle can potentially release PBDEs to environment. The
scarcity of information with regard to all opportunities for environmental releases
restricts the lifecycle analysis presented in this chapter, limiting what can be said about
potential releases during PBDE production, the consumer use of PBDE-treated products,
and the disposal of PBDE-treated products.
2.4.1. Production Releases
The EPA's Toxic Release Inventory (TRI) provides data on environmental
releases from the production of decaBDE and its use to fire retard in the production of
other materials for 2006 (USEPA, 2008a; the latest reporting year). A total of 33.46 MT
of decaBDE was released to the air, land and water in 2006. Great Lakes Chemical, the
major United States producer of decaBDE, reported a total of 18.1 MT released to the
environment. Figure 2-3 displays a summary of fugitive, stack, surface water and land
treatment releases for 2006 (top chart on figure). PBDE production waste was also
2-7
-------
transferred to chemical waste landfills, bringing the total TRI reported data to 62.8 MT in
2006. However, the disposal of the PBDE waste in permitted hazardous waste landfills is
not considered as an environmental release, and, therefore, this production waste was not
added to Figure 2-3. Disposal in a permitted hazardous waste landfill is not considered to
be an "environmental release," because of the controls in place in RCRA-permitted
landfills are designed to prevent offsite migration and contamination of ground water
sources. Releases of decaBDE to the air from fugitive and stack emissions accounted for
the preponderance of environmental release, -89% of total releases, from the chemical
manufacture of decaBDE and its use in the production of other materials. The discharges
to surface water and the off-site land treatment of the waste accounted for 6 and 5%,
respectively. The bottom chart to Figure 2-3 displays total environmental releases of
decaBDE from production facilities from 1997 to 2006. It appears from the chart that
total environmental releases peaked in 1999 (at 53.9 MT) and remained essential the
same through 2002. There was a drop in environmental releases in 2003 (to 36.3 MT),
followed by an increase to 44.8 MT in 2004. Examination of total releases from 2004
through 2006 suggests that decaBDE releases decreased by about 25% over this time
interval (from 44.8 MT in 2004 to 33.4 MT in 2006). It is not known if this signals a
continued decrease in subsequent years of TRI reporting. The BDE congener distribution
in decaBDE provides a basis for assuming the environmental loading of BDE 209, BDE
207, and BDE 208 that may be associated with United States production in 2006. Table
2-2 displays the assumed BDE congener distribution in the environmental releases
associated with production of decaBDE in 2006. It should be noted that TRI data may
not represent all releases of PBDEs from all sources and use activities. Therefore, the
emissions of PBDE to air, water and land may possibly be underestimated (USEPA,
2008a).
2.4.2. PBDE Content of Consumer Products
The previous section reviewed current releases of decaBDE from primary
production facilities in the United States. It is believed, however, that the major
environmental rel eases occur as a function of the use and disposal of products containing
PBDEs. The extent and magnitude of such releases is largely unknown, although
-------
attempts have been made to quantify these releases. These quantifications are based on
the PBDE content of the product and mathematical modeling based on these
concentrations. This section reviews studies measuring the PBDE concentration in
consumer products. The next section presents a procedure to quantify releases from
consumer products.
An assessment of the penta and octaBDE content of various polymers was
conducted by the German EPA (Kcmmlein et al„ 2005). A total of eight samples, each of
acrylonitrile-butadiene-styrcne copolymer (ABS) resin, high-impact polystyrene (HIPS),
polyurethane hard foam (PUR-H), and epoxy resin (EP) were evaluated using the single-
ion monitoring and a gas chromatographic/mass spectrometer system. The polymers
were obtained from the manufacturers along with the percent addition of penta- and
octaBDE as flame retardants. ABS samples had been treated with 1% pctitaBDE and
2.95% octaBDE; PS samples with 2.96% octaBDE: and PUR-H and EP with 2%
pentaBDE. Table 2-3 shows the mean and standard deviation of the concentration of
penta and octaBDE detected in eight samples of each polymer.
A Swiss study published in 2003 provides a quantitative basis for assigning a
plausible PDBE content of final products made from lire retarded plastic resins (Morf, et
al., 2005). In this study, waste electronic products and components were sampled at an
electronics recycling facility in Bern. The sampled end products included television sets
(TV), video camcorders, radios, HiFI stereo systems, portable compact disk players,
mobile phones, standard telephones, toasters, and vacuum cleaners. Together, these
product lines represented about 90% of the total use of penta, octa, and decaBDE
employed as fire retardants in electrical and electronic appliances and equipment (Morf,
et al, 2005). Several samples of discarded material were taken from piles of copper
cable, printed circuit boards, TV housings and rear covers, and PC housings. The
samples from each product were composited, and then analyzed using either gas
chromatography or electron capture detection with mass spectrometry. The mean
concentrations (mg/kg) of the PBDEs found in each product are shown in Table 2-4.
In addition to the mean concentrations in electrical and electronic equipment
(EEE) waste, Morf et al. ( 2005 ) determined the abundance of select PBDE congeners in
the EEE materials. One sample of each of the following items were taken from the
2-9
-------
output piles of sorted materials at the recycling plant: a plastic PC screen housing, a
plastic TV-housing rear cover, a fine-grained plastic with particle size 2-5 and 5-10 mm
(general EEE waste), a fine-grained metal, a fine particulate, and a printed circuit board.
Samples were Soxhlet-extractcd and analyzed with HRGC/HRMS. Table 2-5
summarizes the results of the analysis of these single EEE waste material samples.
Because these were essentially grab samples, the results in Table 2-5 only serve as a
general indication of the potential congener distribution in the EEE waste.
The Consumer Products Safety Commission reported on the chemical fire
retardant content of polyurethane foams (PUFs) used in furniture in the mid-1990s
( Cobb, 1995). Seven PUF types were evaluated. Three samples of each of the foam
types were collected and analyzed by HRGC/HRMS. Only one PUF material contained
PBDE (PBDE species not identified). An average of 3% (range 2.9 - 3.2%) PBDE was
found in this one PUF. The other samples contained melamine and/or TDCP as fire
retardants. Hale et al. (2002) noted in a separate study that the pentaBDE content in PUF
can be as high as 30% by weight.
2.4.3. Estimates of PBDE Releases to the Air Based on Chamber Testing
PBDEs are basically fire retardant additives to materials and not bonded
chemically to the matrix. This means that there exists a potential for the PBDEs to
escape the matrix through the process of volatilization to the air. Thus the products
treated with PBDEs may off-gas PBDE within indoor air microenvironments while the
product is still in use. Volatilization is the most likely mechanism of release from the
product into the surrounding air.
Only one study was found which attempted to measure this possibility of release
under laboratory conditions (Kemmlein et al., 2003). Kemmlein et al. (2003) determined
PBDE emission rates by placing various products into an enclosed chamber, passing an
air stream over the products, and systematically sampling the chamber air over the
duration of the experiment. Test chambers consisted of volumes of either 0.02 m3 or 1
m\ and a temperature of 23°C was kept constant over the experiments. The products
tested represented a cross-section of materials commonly used in interior spaces:
insulation and assembly foams, information technology (IT) devices, upholstered
2-10
-------
furniture, upholstery polyurethanc foams, mattresses, and circuit boards. All materials
had been treated with PBDE as additive fire retardants. The treated materials were kept
in the enclosed chambers for a period of 100 days or longer in order to insure that steady-
state conditions were reached prior to sampling. Polyurcthane foam was used to
passively collect PBDE vapors that had escaped the material matrix. Additionally
chamber walls were rinsed with tiexane to collect PBDEs that had migrated to the walls
for a complete mass balance of PBDE's emitted from the test material.
As part of this chamber experiment, two PC workstations (A and B ) were tested
in I nr' emission test chambers under operational conditions (Kemmlein et a!., 2003).
Workstation A consisted of a PC monitor, a desktop computer, a keyboard, a mouse, and
a printer obtained from different manufacturers. Workstation B consisted of the same
array of computing devices, but the components were obtained from a single
manufacturer. Stations A and B were tested in the air chamber for 93 and 150 days,
respectively. In PC workstation A, BDEs 47, 100, 99, and 85 were detected in the air at
concentrations less than 0.3 ng/tn;. BDEs 47, 100, and 99 were detected in the chamber
air of PC workstation B at concentrations of 150, 28, and 61 ng nr. respectively.
Additionally trace amounts of BDE 153 were detected in both experiments. Testing
computer circuit boards in. the test chamber yielded the following results (expressed as an
emission rate in units of ng/unit/hr): BDE 17 = 0.6; BDE 28 = 1.9; BDE 47 = 14; BDE 66
- 0.4; BDE 100 =¦ 1.3; BDE 99 - 2.6; BDE 85 = 0.1; BDE 154 = 0.1 and BDE 153 -
0.04.
The PC equipment used in these tests was manufactured after 2000. To test the
hypothesis that older plastic casings may have emitted significantly more PBDEs,
Kemmlein et al. (2003) tested the back panel of a television casing manufactured before
1979. The back panel had been treated with oetaBDE, BDEs 28 (maximum
concentration 0.5 ng/m3), 47 (maximum concentration 8 ng/m3), 66 (maximum
concentration 0.24 ng/m3), 100 (maximum concentration 0.27 ng/nr), and 99 (maximum
concentration 0.84 ng/m"1) were detected in the air to the test chamber. The chamber
walls were rinsed with solvent, and the resulting solution was analyzed for the presence
of PBDE congeners. BDE 47 and 99 were detected in the rinse, corresponding to a
surface concentration of 568 ng/m2 and 514 ng/m2 of BDE 47 and BDE 99 respectively. It
2-11
-------
was concluded that a significant portion of the BDE congeners 47 and 99 were adsorbed
to the chamber wall; therefore, the reported air concentrations of these congeners were
likely underestimated. From the analysis of the older plastic back panel an observation
can be made that older treated products may continue to emit BDEs into the air after long
periods of time (i.e. 20 years or more).
Hazratian and Harrad (2006) observed an apparent association between the ages
of PCs used in an office and the indoor air concentrations of PBDEs. They measured
PBDE levels in the indoor air of an office when an older and newer PC was used. There
were three experimentally conditions: 1) Only a PC built in 1998 was used; 2) A PC built
in 1998 and a PC built in 2003 were equally used by employees, and, 3) Only the PC
built in 2003 was used by the office employees. In condition (1), the total PBDE
concentration in indoor office air was approximately 431 pg/nr. In condition (2), the
indoor total PBDE concentration decreased to 253 pg/m3. When only the newer PC was
used, the total PBDE indoor air concentrations decreased further to 81 pg/nr.
A study of office buildings confirmed a qualitative association between the
presence of PBDE treated products and subsequent levels of PBDEs in indoor air (Harrad
et at., 2003). Indoor air with the highest PBDE levels were in rooms in office buildings
equipped with numerous desktop personal computers (12 - 16 per room), and numerous
PUF-containing chairs (11 - 22 per room). By comparison, the lowest indoor air PBDE
concentrations occurred in domestic environments that had no PUF-containing furniture
(Harrad et al, 2003).
Rigid polyurethanc foam (PUR-R) treated with decaBDE showed no emission of
PBDE congeners after 168 day residence time in a test chamber (Kemmlein et al., 2003).
PUR-R was manufactured for furniture upholstery stuffing. Thus the investigators
reported no detectable brominated organic compounds within the test chamber. Direct
analysis of the material showed detectable congeners of deca and nonaBDE in the
product.
2.4.4. Estimates of PBDE Releases to Air from Consumer Products
Attempts have been made to estimate the annual amount of PBDE that may be
volatilized to the air during the service life of a treated product. The EU has estimated
2-12
-------
the theoretical amount of decaBDE, oc-taBDE, and pentaBDE that could have volatilized
into the atmosphere over the European continent during the usage and product life of fire-
retardant products (European Union, 2001, 2002, 2003). Equation 1 gives an
approximate estimate of the percentage of PBDE that may volatilize over the product life:
Percentage loss due to volatilization — (l. 1 x 10" )x V x jV(%) (l)
Where:
Vp - vapor pressure of the PBDE flame retardant, mmHg at 21 °C
N = service life of the flame retarded product in yrs (assumed to be 10
yrs by the European Union).
Equation I was initially developed to approximate the volatilization rate of
chemical plasticizers added to plastic films during their product use (European Union,
2002). The EU concluded that the equation should be applicable to the estimation of
PBDEi loss by volatilization from a solid matrix because the equation emphasizes the
vapor pressure of the compound as the controlling factor. Multiplication of the annual
tonnage of PBDE treated materials by the annual percent loss rate yields a rough but
plausible estimate of PBDE releases to the air (kg/yr) from the treated products. In
review of the EU approach, Prevedouoro et al. ( 2004) commented that utilizing Equation
1 should be appropriate for estimating the outgassing of PBDEs under ambient
temperatures from solid matrices where they have been used as additives.
Assuming a vapor pressure of 3.47E-08 mmHg (at 21 C) and a product life of 10
yrs, a volatilization loss rate of decaBDE from the product was calculated by EU as
0.38% over 10 years (or 0.038% per year). Approximately 6,710 MT/yr of decaBDE was
used in plastics for all EU countries combined. Based on this quantity, the EU estimated
the total losses of decaBDE to the air during the service life of the treated plastic products
to be 2.55 MT/year in Europe (6,710 * 0.00038; European Union, 2002). Using a vapor
pressure of 4.9E-08 mmHg for octaBDE, the EU estimated a volatilization rate of 0.54%
over 10 years (or 0.054% per year) (European Union, 2002). Assuming a 1994
consumption figure of 2,550 MT/yr, the EU estimated that approximately 1.38 MT/yr of
octaBDE volatilized from product usage throughout Europe (European Union, 2002).
PentaBDE was assumed to have a vapor pressure of 3.5E-07 mmHg, and this yielded an
2-13
-------
estimated annual loss rate of 0.39% over a 10 year product life (or 0.039% per year)
(European Union, 2001). The EU used a value of 1,100 MT/yr pentaBDE in
polyurethane foam products to calculate an annual release rate of 4.3 MT/yr pentaBDE
for Europe as a whole (European Union, 2001).
It should be noted that the product service life of 10 years was an arbitrary
assumption by the EU, and, when combined with the average amount of PBDE treated
materials in use each year, is thought to give a reasonable, but unverifiable, result
(Prevedouoro ct al., 2004).
Prevedouoro et al. (2004) estimated the volatilization flux of BDE 47 from
products treated with pentaBDE and consumed in the United Kingdom. Using equation I
and applying a high and low BDE 47 consumption rate (74 and 65 MT in the year 2000,
respectively), Prevedouoro ct al. (2004) calculated the annual mass flux of BDE 47 to air
from the use of treated consumer products. Figure 2-4 shows these results. The peak in
the air emissions of BDE 47 occurred in 1997 for both the high and low consumption
values. BDE 47 emissions in 1997 were calculated to be 31 and 22.5 MT for the high
and low consumption rates, respectively.
This chapter uses the EU estimation technique as a basis for calculating possible
total volatilization of deeaBDE, octaBDE, and pentaBDE from treated products in the
Americas and Asia as well. This analysis assumes the following:
1. The market demand of the PBDE formulation for 2001 (the latest year
for which statistics are available) is used as a surrogate for inferring
the amount used to fire retard products
2. The product life is 10 years
3. Equation 1 is used to estimate the percent of PBDE volatilized from
the treated product
In 2001, the market demand for deeaBDE was approximately 24,500 MT in the
Americas (BSEF, 2006a). The Americas includes all countries in North, Central, and
South America. Roughly 80% of deeaBDE is used to fire-retardant hard plastics (Pure
Strategies, Inc., 2005), with the remaining 20% used in textiles. Therefore, with a total
2-14
-------
demand of 24,500 MT, it is estimated that the amount of deeaBDE incorporated into new
plastic products in the Americas, in 2001, is 19,600 MT. A volatilization loss rate of
0.038%/yr, or 0.38% over 10 years, would result in approximately 7.5 MT decaBDE/yr
volatilized from these new products made from the tire retardant plastics. Assuming no
additional products were put into the market during that period, this means that
approximately 75 MT of deeaBDE would have volatilized from plastic products over a
10-year time period in the Americas. There is no further breakdown of deeaBDE use by
individual country comprising North, Central and South America. If it is assumed that
the United States accounts for 80% of the total use of deeaBDE produced in 2001, then
15,680 MT may have been used in the United States. It is possible that 60 MT of
deeaBDE may volatilize over 10 yrs (6 MT/yr) lroni plastic products in use in the United
States. Total deeaBDE market demand in Asia in 2001 was 23,000 MT (BSEF, 2006a).
Asia includes the countries of China, India, and Japan. Applying the same assumptions
as in the estimate for the Americas would result in a volatilization of about 7.0 MT
decaBDE/yr (70 MT over 10 years) from plastic products used in Asia. The estimated
global market demand for deeaBDE in 2001 was 56,100 MT (BSEF, 2006a). An
assumption that 80% of deeaBDE is used in plastic products translates to an estimate of a
global volatilization loss of 17 MT decaBDE/yr or 170 MT over 10 years.
OctaBDE has a higher vapor pressure than deeaBDE; hence the rate of
volatilization from treated plastic materials will be greater. Applying Equation 1, and
assuming a vapor pressure of 4.9E-08 mmHg at 21°C (European Union, 2003), a
calculation of annual loss due to volatilization can be made using the 2001 market
demand for octaBDE. OctaBDE was primarily used as a flame retardant additive to ABS
polymer used in PC casings and monitors (ATSDR, 2004). In the European Union, 95%
of the use was to fire retard ABS plastic (European Union, 2003 ), with the remaining 5%
to be used in HIPS, PBT and polyamide polymers. This assumption is used to calculate
the rate of loss from ABS plastic. In 2001, approximately 1,500 MT were used in the
Americas and 1,500 MT/yr also used in Asia (BSEF, 2006a). In Europe the usage figure
was 610 MT. From equation 1, the loss from volatilization of octaBDE during the
service life of an ABS plastic product is approximately 0.54% over 10 years (or 0.054%
per year). It is estimated that Asia and the Americas might have each emitted 0.77 MT/yr
2-15
-------
(7.7 MT over 10 years) of oetaBDE into the air from volatilization from ABS plastic
products, while, in Europe, approximately 0.31 MT/yr (3.1 MT in 10 years) may have
been released. On a global scale, approx imately 1.9 MT/yr of oetaBDE (19 MT in 10
years) may have been emitted from plastics.
PentaBDE was mostly used as an additive fire retardant in flexible polyurethane
foam. The vapor pressure of pcntaBDE (3.7E-07 mmHg at 21°C), while low, is sufficient
to induce a low rate of volatilization during the life of foam stuffing in furniture and car
seats (European Union, 2001). Equation 1 predicts that the rate of loss due to the
volatilization of pentaBDE during the 10-year service life of the treated flexible PUF
material to be 3.9% (or 0.39% per year) (European Union, 2001). It is assumed that
approximately 96% of pentaBDE was used as a fire retardant in flexible polyurethane
foam, with the remainder used in rigid polyurethane elastomers for instrument casings
(European Union, 2001). The annual mass flux from the flexible PUF to the air can be
estimated using the market demand statistics for 2001 (BSEF, 2006b). These statistics
show total usage in the Americas, Europe, and Asia to be 6,816; 144; and 144 MT,
respectively. Annual emission estimates, assuming volatilization losses of 0.039%/yr, are
26.56, 0.56, and 0.56 MT/yr, (265.6, 5.6 and 5.6 MT ovcrlO years) for the Americas,
Europe, and Asia, respectively.
This analysis represents a picture of possible annual emissions from the
volatilization of PBDEs from treated products based on a 10-year product life. The
analysis does not take into account the introduction of new products treated with PBDEs
during this timeframe, but it is consistent with the approach taken by the EU in evaluating
potential human health risks associated with exposures to commercial PBDE
formulations (European Union, 2003). In this context, the EU has suggested that the
estimate of amount volatil ized from products is likely an underestimate because the
estimates are based on a single year's product use information. Therefore, the EU has
indicated that the actual amounts released to air could be one order of magnitude higher
than what is predicted by Equation 1 (European Union, 2001, 2002, 2003). In order to
compensate for this potential underestimation due to considering only one year of use, the
EU increased their calculation of air releases of PBDEs from products in European
countries by a factor of 10. A similar factor of 10 could be applied to estimates made
2-16
-------
here for the Americas, but given the uncertainties in the procedure (the validity of the
empirical equation, the availability of use information before 1999, etc) as well as the fact
that pentaBDE and oetaBDE were taken off the market in 2004 while other new products
could come into the market in future years, no adjustments are made here. In other
words, an annual estimate of loss which is really 1/10 the loss over 10 full years of loss
from a single year of use (2001 to be precise), will be assumed to generally represent any
year - no additional losses are assumed based on introduction of new product each year.
The above calculations presented estimates of releases of penta-, octa- and deca-
BDE formulations from the use of treated plastics. The BDE congener specific emissions
from plastic products are estimated by multiplying these releases by the congener
distributions in the various commercial PBDE formulations (as shown in Table 2-1).
Table 2-6 summarizes the estimated emissions of specific congeners based on the
analysis presented above. The total global release of BDE congeners (as apposed to the
total global release of PBDE formulations) from plastic consumer products is
approximately 4,65 MT/yr (46.5 MT over 10 yrs). Of this total, the dominant congeners
are BDEs 209 released at 16,67 MT/yr or 35.9% of the total for all congeners, 99 at
28.4%, 47 at 18.7%, 100 at 4.8%, and 153 at 2.7%, with all other congeners released at
less than 2% of the total.. The dominance of the congeners BDE 209, 99, 47 and 100
indicates the historical use of deca and pentaBDE in PBDE-treated products. A major
caveat to this analysis is that the calculated congener profile assumes the same congener
distribution as in the PBDE commercial formulation (see Table 2-1), and the congener
mix is volatilized with the commercial PBDE product.
2.4.5. Estimates of the Mass Flow of PBDEs Contained in Electronic and Electrical
Equipment (EEE) Waste
'Once PBDE-treated materials have reached their functional life they are discarded
in landfills, in incinerators, or they are recycled. This section describes the mass-flow
analysis of environmental releases of PBDEs associated with the disposal of PBDE-
treated EEE waste comprised of products that have reached their end-life. There is a
paucity of information on the amount of PBDE-treated products that may be discarded in
any given year. There exists some information on the amount of electronic and electrical
2-17
-------
equipment waste contained in municipal solid waste in the United States (USEPA, 2006)
In the United States, an estimated 2.4 million metric tons (MMT) of EEE waste were
generated in 2005 (USEPA, 2006) and incorporated into municipal solid waste. Of this
amount, approximately 300,000 MT (or 12.5%) of selected consumer electronics were
recovered for recycling. Selected consumer electronics subject to recycling in the United
States include products such as televisions, VCRs, DVD players, video cameras, stereo
systems, telephones, cell phones, hand-held electronic devices, personal computers,
laptop computers, printers, fax and copy machines (USEPA, 2006). Most, but not all, of
these products contain PBDE as a fire retardant (IPCS, 1994), Morf et al (2005) analyzed
the EEE waste at an EEE waste recycling facility in Switzerland and found an average
concentration of 510 ±35 mg/kg and 530 ±30 mg/kg of deeaBDE and oetaBDE,
respectively. In deriving these mean concentrations in the EEE waste, Morf et al. (2005)
specifically studied what they termed "small" electronic waste, including, "small
household appliances (e.g., toasters and vacuum cleaners), office and communication
appliances (e.g., personal computers and monitors, printers, phones, and fax and
photocopy machines), entertainment electronics (e.g., television (TV) sets, videos,
camcorders, radios, HiFis, and portable compact disk (CD players), and small size E&E
equipment (e.g., plugs and mobile phones)." They justified this selection of EEE waste
as the EEE which contained the bulk of brominated flame retardant use in EEE, and
analyzed the portions of these products most likely to contain PBDEs, such as electric
circuit boards and TV housings. The results from their tests are shown in Table 2.4 and
2.5, and were discussed earlier. Based on a mass balance of the total material input,
including parts measured for PBDEs and parts not measured for PBDEs, they calculated
total product concentrations, and the average of these total product concentrations are
given as 510 and 530 mg/kg for deeaBDE and oetaBDE, respectively. For purposes of
calculating deeaBDE and oetaBDE content in EEE waste in the United States, the total
product concentrations of deeaBDE and oetaBDE found in the Morf et al (2005) study
are assumed to be representative of EEE waste in the United States (as defined by
USEPA, 2006). Although not perfect, the composition of EEE waste described and
analyzed by Morf et al (2005) is similar enough to the composition of EEE waste in the
United States (USEPA, 2006) to permit a rough estimate of the amounts of deeaBDE and
2-18
-------
octaBDE that may in present in EEE waste in the United States. Using these data, it is
estimated that 1,224 MT dccaBDE and 1,272 MT octaBDE are contained in 2.4 MMT of
EEE waste generated in 2005 (510 mg/kg deca * 2.4MMT and 530 mg/kg octa*2.4
MMT). The recycled portion of the EEE waste is estimated to contain about 153 MT
decaBDE and 159 MT octaBDE (3E+05 MT recycled EEE waste * 510 and 530 mg/kg
for deca and octaBDE, respectively).
Not all of the EEE waste is recycled in the United States. The remaining 2.1
MMT of EEE material in MSW is either disposed of in landfills or municipal waste
incinerators. Of the amount of total EEE waste that is not recycled, roughly 20% is
combusted in MSW incinerators and 80% is landfilled (USEPA, 2006). This would
mean that 0.42 MMT of EEE waste is incinerated and 1.68 MMT is sent to landfills.
Using the mean concentration of deca BDE and octaBDE in EEE waste t Mori" ct al, 2005)
that may have been incinerated or landfilled in 2005 yields the following estimates:
• 230 MT decaBDE in EEE waste incinerated in 2005.
• 239 MT octaBDE in EEE waste incinerated in 2005.
• 857 MT decaBDE in EEE waste landfilled in 2005.
• 890 MT octaBDE in EEE waste landfilled in 2005.
PBDE in landfill leachate: While it is reasonable to assume that some leaching
of BDE congeners from landfilling of treated products may occur, only limited
information could be found in the literature that could support any estimation of this as an
environmental release. For example, Osaka ct al. (2004) evaluated the untreated and
treated leachate at seven landfills in Japan for the presence of BDE congeners. BDEs in
the raw leachate were detected in the following ranges of concentration (pg/L) across the
seven landfills: BDE 47 (not detected (ND) - 2,200); BDE 28 (ND - 970); BDE 66 (ND
- 3,200); BDE 99 (ND - 1,800); BDE 153 (ND - 27); BDE 154 (ND - 1,200). In the
treated landfill leachate, no PBDE congeners were detected, indicating the effectiveness
of the leachate treatment process. Odusanya et al (2008) reported on the PBDE
distribution in the leachate from five landfills in South Africa. BDE 28, BDE 47, BDE
66, BDE 71, BDE 75, and BDE 77 were regularly detected in raw landfill leachate
samples collected from all the landfill sites. No BDE-209 could be detected in any of the
2-19
-------
raw leachate samples. The range of concentrations of total PBDEs (pg/L) detected in
each landfill were as follows: Landfill 1 (ND-2670); landfill 2 (ND-6638); landfill 3
(ND-7230); landfill 4 (41-4009), and landfill 5 (90-9793). Individual congeners across
all five landfills ranged in the following concentrations (pg/L): BDE 28 (100-3,333);
BDE 47 (1,469- 9,793); BDE 66 iM) 4,020); BDE 71 (1,667- 9,459); BDE 75 <743—
7,426); BDE 77 (ND-4,257); BDE 85 (ND 1,240); BDE 99 (ND- 5,191); BDE 100
(ND- 2,162); BDE 119 (ND- 5,392); BDE 153 (ND- 875); BDE 154 (ND- 2,176); BDE
183 (ND- 263).
Odusanya et al (2008) and Osaka et al. (2004) give an indication of the ranges in
concentration of BDE congeners that may be present in untreated landfill leachate. It
should be noted, however, that these studies seem to indicate that PBDE congeners are
not expected to be detected in treated landfill leachate. In the United States, MSW
landfills are required by federal regulation to collect and treat landfill leachate, and to
continuously monitor groundwater for an indication of leachate migration from the
landfill (Code of Federal Regulations (40 CFR Part 258) Subtitle D of RCRA: Criteria
for Municipal Solid Waste Landfills (MSWLFs)). Therefore, no attempt is made here to
estimate the amount of PBDE potentially released into the United States environment
from the migration of landfill leachate.
PBDE in incineration emissions: MSW incinerators in the United States have not
been characterized for their potential stack emissions of PBDEs. PBDE congeners
detected in the stack emissions to MSW incinerators is usually a consequence of not
completely destroying PBDEs present in the waste during combustion (Sakai et al, 2006).
MSW incinerators typically operate with a combustion efficiency of about 98% (Yang et
al 2007), which means that 98% of the PBDE content of the waste is expected to be
destroyed during combustion. Previously it was estimated that approximately 230 and
239 MT of deca- and octaBDE may be contained in the EEE waste subject to
incineration. These amounts of deca and octaBDE present in the EEE waste should be
destroyed by about 98% leaving only 4.6 and 4.8 MT of deca and octaBDE subject to
stack emissions. In the United States, strict regulations have imposed highly effective air
pollution control devices on MSW incinerators (Federal Register, 1995). The application
of dry scrubbers combined with fabric filters on large MSW incinerators has generally
2-20
-------
reduced the concentrations of semivolatile organics present in the combustion gases
leaving the furnace by an additional 99% prior to emissions from the stack (Federal
Register, 1995). The previously calculated uncontrolled emissions of deca and octaBDE
would be further reduced by 99% with these air pollution control devices, leaving only
0.046 and 0.048 MT of decaBDE and octaBDE in the stack emissions from MSW
incinerators in the United States (4.6 MT decaBDE * (1 -0.99) and 4.8 MT octaBDE *
(1-0.99).
Incineration of MSW contaminated with PBDE also forms polybrominated
dibenzodioxins and dibenzofurans (PBDD and PBDF) in the combustion gases. PBDE: is
a direct precursor to PBDD/PBDF formation within thermal systems (Weber and Kuch,
2003). The mechanism for this is thought to consist of the infra-molecular elimination of
bromine (Bri) and/or hydrogen bromide (HBr) (Weber and Kuch, 2003). These
denomination reactions are enhanced at temperatures above 5(H) (' (Weber and Kuch,
2003), and the cleavage of the Bi j or HBr from the molecule leads to ring closure to form
PBDD and PBDF (Ebert and Bahadir, 2003). The kinetics of the conversion of PBDE to
PBDD and PBDF appears to be more favorable for the lower brominated diphenyl ethers
(e.g., pentaBDE) and less favorable for the decaBDE (Weber and Kuch, 2003). The
destruction efficiency of the PBDE contained in the MSW by controlled incineration
ranges from 90 to 99.9%, leaving enough PBDE for the emission of PBDE from the stack
and for the formation of PBDD and PBDF (Weber and Kuch, 2003). Weber and Kuch
(2003) and Ebert and Bahadir (2003) do not provide information on the efficiency of
thermolyticaUy converting the mass concentrations of PBDEs in the waste combusted to
the mass concentrations of PBDD and PBDF. formed in incinerator stack emissions.
Therefore, no attempt is made here to estimate stack air releases of PBDDs and PBDFs
during the incineration of MSW.
2.4.6. Estimates of the Mass Loading to Land and Water of PBDEs Present in
Sewage Treatment Plant Sewage Sludge and Effluent
Chapter 3 to this report gives an overview of the levels of PBDEs in sewage
treatment plant (STP) influent, effluent, and sewage sludge. Table 3-8 to Chapter 3
summarizes these data from various surveys of sewage treatment plants in various
2-21
-------
countries. Several references in the Table 3-8 apply to STPs operating in the United
States. However, the Palo Alto, California STP studied by North (2004) was chosen
from the Table to approximately represent current municipal sewage treatment processes
and practices in the United States. Because this was a careful mass balance study of the
distribution of PBDEs in sewage sludge and the effluent, it is used here to calculate
PBDE mass loadings to land and water for STPs in the United States. The focus of these
calculations is the mean BDE congener mass loadings in effluent discharges to surface
water, and in the sewage sludge that is applied to land as a soil amendment and fertilizer.
There are 16,519 STPs operating within the United States (USEPA, 2008b),
These STPs have a combined daily sewage flow rate of approximately 1.3E+11 L/d
(NAS, 2002). Approximately 93% of total daily sewage flow undergoes secondary
and/or tertiary treatment before the wastewater is discharged into surface water (USEPA,
2008b), which basically matches up with the treatment processes at the Palo Alto STP
studied by North (2004). Applying the factor of 93% to the total daily sewage flow from
all STPs combined gives an estimated flow rate of 1.2E+05 L/d for STPs with secondary
and tertiary treatment. The secondary and tertiary treatment processes remove solids
from the wastewater, concentrates the solids, and generates sewage sludge.
Approximately 6.3E+06 MT dw (metric tons, dry weight) of sewage sludge are used or
disposed of annually in the United States, of which 3.8E+06 MT dw (60%) is land
applied or commercially distributed as fertilizer (NAS, 2002; USEPA, 1999). The
remaining sludge is disposed of in the following manner: 1.1 E+06 MT dw are land tilled
(17%), 1.4E+06 MT dw are incinerated (22%) and 6.3E+04 MT dw (1%) have
miscellaneous uses such as daily landfill cover (NAS, 2002; USEPA, 1999).
North (2004) found that just five BDE congeners accounted for approximately 86
- 90% of total PBDEs detected in effluent and sewage sludge, respectively. These five
congeners are BDE 47, BDE 99, BDE 153, BDE 154, and BDE 209. The mean
concentrations (j.ig/kg dw) of these dominant congeners in sewage sludge (Table 3-8,
Chapter 3) were: BDE 47 = 757; BDE 99 = 944; BDE 153 - 88; BDE 154 = 68, and
BDE 209 = 1,183. The sum of 23 BDE congeners detected in the sludge was 3,381
|ig/kg. In the wastewater effluent discharged into surface water after secondary and
tertiary treatment, the mean concentration of the congeners were 10.5, 11.2, 0.98, 0.78,
2-22
-------
and 1.73 ng/L for BDE 47, BDE 99, BDE 153, BDE 154. and BDE 209, respectively.
The sum of 28 BDE congeners detected in STP effluent at the Palo Alto STP was 29,02
ng/L (North, 2004). BDE 209 dominated PBDE concentration in sewage sludge (at
approximately 35% by vvt), however, BDE congeners 47 and 99 accounted for about 36%
and 39% of the total mass of the five main congeners present in treated wastewater
effluent. BDE 209 represented only 6% of the sum of the five congeners present in the
STP effluent. These analyses suggest that sewage sludge is a major sink for BDE 209,
since its concentration was high in sludge but not in effluent.
Sewage sludge is also incinerated at STPs. No inventories of PBDE emissions in
the stack gases of sewage sludge incinerators could be found in the literature. North
(2004) did not stack test the incinerator for PBDEs at the multiple hearth sewage sludge
incinerator at the Palo Alto STP under the assumption that PBDEs in the incoming sludge
would be destroyed within the incineration system by >96%, leaving all PBDE congeners
below the limit of detection. However, PBDDs and PBDFs were sampled in the
incinerator emissions with the assumption that PBDEs entrained in the combustion gas
would be thermolytically converted to PBDDs/'PBDFs. Only one homologue group of
PBDFs was detected in the emissions to the incinerator, triBDF. North (2004) calculated
the mass loading of triBDF to be 2.8E-07 kg/yr based on a mean concentration, of 82
jig/m3 triBDF in the stack emissions.
The following mass loading calculations of BDE congeners from STPs to the
United States environment are made based on mean concentrations of BDE congeners
detected in the sludge and effluent at the Palo Alto STP.
• Estimated annual loading of BDE congeners to the land from the land
application of sewage sludge in the Unite States:
General equation:
Mass loading PBDE to land (MT / yr) = mean concentration PBDE in sewage sludge
x mass sludge applied to land/yr (3)
BDE 47 hading =. 757
7 V "g StJE 47 I • 10
3.S x 10 *6 MT shklgi-
BDE 47 ituuim
kg sludge
2.88 A/7" vr to land
vr
2-2^
-------
99 loading
SDE 99 /¦ >¦;
BDE 153 loading
BDE 153 hading
944
its; UDE W ! > 10 " k-4 3.8 x 10 "" ,W s/«,A,v
3 ,59 ,OT ¦' ir /« land
X8
II.V BDE 153
I - 10 ' kg 3. S x 10 ,f' A rr sludge
kg ¦hi.i.'r
0.33 MT rr so hmd
/4\ W BDE 154 1x10 '"kg 3.8 ~ 10" ,WT Mv
(•*} BDE 154 bading •=¦ oH x x —
/eg sludge
BDE 154 hmJing = 0.26 A/T / \r to land
(5) BDE 209 /,„«//»€ 209
A'g sludge
BDE 209 loading: — 4.5 .W / jr /w laml
1 x 10 •* A-tr 3,8 x 10 "• MT sludge
"8
(6)ZBDE „ ;i Ww#
,381
wjT 2 BDE 1x10"' % 3,8 x 10 Aff i/n^
kg sludge
I BDE ,,nt = 12 .8 A/r / vr to land
»S
Estimated annual loading of BDE congeners to surface waters from STP effluent
mi 11 ¦ ¦ ¦ - "" — Jim...— - . . - -
in the United States:
General equation:
Mass loading PBDE to water - mean concentration PBDE in STP effluent x total STP effluent/yr (4)
m
ng BDE 47 1 x 10 g kg 1 .3 x !0 ' " L iotaI effluent 365 d
BDE 47 loading
10 .5
L effluent ng 1x10" g
BDE 47 loading 498 kg yr (0.498 MT / yr ) to water
d
(2)
BDE 99 lauding
11
ng BDE 99
kg
L. effluent ng 1 > 10
BDE 99 loading - 53 J kg/ yr (0.531 MT / yr ) to woter
1.3x10'' L total effluent
d
365 d
<3)
BDE 153 loading
BDE 153 loading
0.98
ng BDE 153
1 x 10 "g
L effluent ng 1 x 10
46.5 kg/yr (0,0465 MT/yr')n, water
1.3x10 ' l. total effluent 365 d
d vr
(4)
BDE 154 loading
BDE 154 loading
0 .78
tig BDE 154
1 x 10
kg
L effluent
57 kg/yr (0.037 MT/
ng 1 x 10
I to water
1 .3 x 10 1 11 E total effluent 36,5 d
X —
J yr
2-24
-------
<5)
aBDE 2l» I \ 10 < it; 1 .3 x 10 " 1 L limit effluent 3t>5 d
BDE 20^ loaditii: - 1 . 73 - •< • —- : •< -
L effluent ng 1 x 10 ' o J yr
BDE im hihhno 82 .1 kg < yr (0.0821 MT ¦ >r ) m wafer
(6)
„ , , ,4 m y BDE | X 10 " i; In l.i* 10 L Mat effluent 365 J
Y BDF, ,s InuJimt - 2l> .1)2 — >. ; x - x
i-~* " " " /., effluent n\> 1 x 10 ' g d .vr
y BDE „ Idling * 1.380 kg / yr (I .38 MT / yr ) m water
2.5. AN EXAMPLE LIFECYCLE ANALYSIS OF DECABDE IN THE UNITED
STATES
This section presents a simplified lifecycle analysis of decaBDE used in plastics
and. electronic and electrical equipment (EEE) in the United States. This lifecycle
analysis is focused on the production, use, and disposal of plastics and EEE treated with
decaBDE. It is noted that all quantities generated in this example are highly uncertain.
This example is hypothetical and only meant to demonstrate considerations for, or a basic
approach to, a lifecycle analysis. For this example, data and procedures described in
previous sections are used.
/. Estimated annual decaBDE demand: 15,680 MT
In 2001 the market demand for decaBDE in the Americas was approximately
24,500 MT (BSEF, 2006a). If it is assumed that market demand is equivalent to annual
production, and 80% of the market demand in the Americas was in the United States,
then an estimated 15,680 MT decaBDE is annually produced in the United States.
2. Estimated annual releases of decaBDE to air, land and water from the production
of decaBDE and decaBDE treated materials: 33.46 MT in 2006 (29.87 MT to air; 1.63
MT to land and 1.96 MT to surface water).
This was the estimate provided for decaBDE production as reported to TRI for the
year 2006 (TJSEPA, 2008a) for releases to air, water, and as waste used in land fanning
(land application of production sludges). The year 2006 is the latest reporting year as of
the date of this analysis. A total of 29.9 MT of decaBDE was released to the air, 0,65
MT released to land, and 1.95 MT discharged to surface water in 2006.
-------
3. Estimate of amount of decaBDE volatilized from in use plastic products: 6 MT
decuBDE/yr (60 MT over 10 years of product use).
Applying procedures developed by the European Union (European Union, 2002)
for estimating the volatile release of decaBDE from treated plastic products gives an
estimated annual release of decaBDE of about 7.5 MT/yr in the Americas. Assuming
80% of decaBDE consumed in the Americas is by the United States gives an estimate of
about 6 MT decaBDE released into the air annually (or 60 MT over a 10 year product
lifetime) from plastic products currently in use in. the United States.
4. An estimated 1,224 MT of decaBDE may be contained in electronic and electrical
equipment (EEE) waste generated annually. Of this amount of decaBDE in EEE
waste, 857 MT decaBDE is landfilled; 214 MT decaBDE in EEE waste is incinerated
and 153 MT decaBDE is contained in recycled EEE materials.
Approximately 2.4 MMT of electronic and electrical equipment (EEE) waste was
generated in the United States in 2005 (USEPA, 2006). Aboutl2.5% or 300,000 MT of
the EEE waste was recycled in 2005. Of the EEE waste not recycled, 20% (0.42 MT) is
incinerated, and 80% (1.68 MT) is landfilled. The EEE waste consists of TVs, VCRs,
DVD players, video cameras, stereo systems and components, telephones, cell phones,
hand-held electronic devices, electronic game devices, personal computers, Wi-Fi
devices, laptop computers, LCD displays, printers, and scanners, Morf et al (2005)
estimated that EEE waste may contain 510 mg decaBDE per kg EEE waste. These
contaminant levels were used to estimate the amount of decaBDE present in the EEE
waste.
5. Estimated annual emissions of decaBDE to the air from the disposal of EEE waste
in landfills and from incineration; 0.043 MT/yr from incineration.
This calculation assumes the following: MSW incinerators achieving 98%
destruction efficiency of the organics in the waste (Yang et al, 2007), and a 99%
additional control of PBDE stack emissions is achieved with advanced air pollution
control technology (Federal Register, 2005). With these assumptions, it is estimated that
0.043 MT/yr of the 214 MT of decaBDE in the EEE waste is emitted from the stacks of
MSW incinerators in the United States.
2-26
-------
DecaBDE in EEE waste disposed of in landfills could still volatilize or leach, but
no estimates of those avenues of loss could be made. It should be stated that the
volatilization route is limited for landfills by the regulatory requirement that landfill
operators apply daily cover material over the buried waste under the federal regulations
of Subtitle D of the Resource Recovery and Conservation Act (RCRA). Although
information was found regarding concentrations of various PBDE congeners in landfill
leachate, no decaBDE was detected in these studies.
6. Amount of decaBDE volatilized to air from recycled EEE. No estimate is made.
The estimated amount of decaBDE that may have volatilized from the use of
treated products is described in item (3) above. It is presently not known to what extent
volatilization of decaBDE may continue to occur from the products comprising recycled
EEE materials. Morf et al. (2005) suggested that if volatilization docs occur, then it is
negligible. However, no attempt to estimate volatilization from EEE waste is made here.
7. Amount of decaBDE released to land and released to water associated with Sewage
Treatment Plant (STP) effluent and the disposal of sewage sludge: 0.082 MT are
annually released into surface waters; 4.5 MT released to land from the land
application of sewage sludge.
In the United States, there are over 16,000 STPs in operation that treat roughly
1.3E+11 L sewage/day. Approximately 93% of these facilities have advanced
wastewater treatment processes to include secondary and tertiary treatment. The focus of
estimating mass loadings of PBDEs to land from the land application of sewage sludge,
and the effluent to surface waters in the United States. The assessment depended on the
PBDE congener distribution discerned in the sludge and effluent of an advanced STP in
Palo Alto, California that received and treated industrial and domestic wastewater (North,
2004). DecaBDE (represented as BDE 209) represented 35% by wt of the total mass of
PBDE detected in the sewage sludge, but only 6% of total PBDEs in the STP effluent. In
the United States, approximately 3.8E-10 MT of sewage sludge generated by advanced
STPs are annually applied to land. Assuming a mean concentration of 1,183 fig
decaBDE kg sludge from the North (2004) study gives an estimate of 4.5 MT decaBDE
2-27
-------
applied to land each year. In the United States, approximately 4.75E+13 L of STP
effluent are discharged into United States waters each year from advanced STPs. North
(2004) determined a mean dccaBDE concentration in the STP effluent to be 1.73 ng/L.
With this mean effluent concentration, it is estimated that 82.1 kg (0.082 MT) decaBDE
are annually released into surface waters from advanced STPs.
2.6. SUMMARY
This chapter has presented information on the production, uses, and life cycle of
PBDE commercial formulations. It should be noted that the production figures are highly
uncertain, and no actual amounts of PBDEs used in the United States for the treatment of
plastics, textiles, and flexible polyethylene foam could be found. The industry usually
aggregates data on a global scale and statistics are available for the Americas, Asia, and
Europe (to include the countries of the EU). The industry breakdown indicates that
approximately 56.418 MT of PBDEs were consumed worldwide in 2001 (the latest
reporting year) of which 83% was decaBDE. The industry voluntarily ceased production
of penta- and octaBDE in December, 2004. Today, only decaBDE is being produced in
the United States, and the EU countries have banned the use of decaBDE effective April
1, 2008. It is estimated that 15,680 MT decaBDE is currently consumed in the United
States. Eighty percent of the decaBDE produced in the United States is used as an
additive fire retardant to rigid plastics used in casings to TVs, personal computers, LCD
screens and other electronic and electrical equipment (EEE). The remaining 20% are
used in textiles such as fabrics used in car seats and other upholstery fabrics, but not in
clothing.
The current and past use of PBDE-treated products can result in environmental
releases of PBDEs. The aim of this chapter has been to evaluate the amount of PBDEs
that could potentially be released from the life cycle of PBDEs, including the release
during production, the release from products in use, the release from disposal and
recycling of products containing PBDEs, and the releases from the land appl ication of
sewage sludge and the effluent surface water discharges from sewage treatment plants
that treat industrial and domestic sewage. While there are numerous data gaps in
determining these quantities for all PBDEs, this chapter has reviewed available
2-28
-------
information on these pathways and highlighted where key information is lacking in an
effort to comprehensively evaluate the life cycle of releases from production to use to end
disposal. Focusing on decaBDE as an example of lifecycle analysis, the following
releases to air, water and land in the United States are estimated, limited by the
availability of data;
TOTAL RELEASES TO AIR: 36 MT/yr
Production: 29.9 MT/yr
Product use: 6 MT/yr
Disposal: 0.043 MT/yr (only from EEE waste incineration)
Sewage treatment: No estimate, because sewage sludge incinerators have
not been adequately surveyed for decaBDE emissions.
TOTAL RELEASES TO LAND: 5.1 MT/yr
Production: 0.65 MT/yr (land application of sludges)
Product use: No estimate
Disposal: No estimate. Landfilling is not considered an
environmental release, because MSW disposal in landfills
is strictly regulated to prevent releases.
Recycling: No estimate.
Sewage treatment: 4.5 MT/yr (land application of sewage sludge)
TOTAL RELEASES TO SURFACE WATER: 2. M I
Production: 1.95 MT/yr
Product use: No estimate
Disposal: No estimate, because MSW disposal in landfills is strictly
regulated to prevent formation of leachate. DecaBDE not
detected in leachate from the few studies that could be
found.
Recycling: No estimate
Sewage treatment: 0.08 MT/yr in effluent discharged into surface water
from the sewage treatment plants.
TOTAL RESEASES TO THE ENVIRONMENT: 43 MT/yr
-------
From the llfecycle analysis of decaBDE, releases to the environment from
production, product use, disposal, EEE materials recycling and sewage treatment plant
sludge and effluent total 43,18 MT/yr. From this analysis, approximately 83.33% of total
environmental releases are to the air; 4.71% of total releases are to surface water, and
11,95% of total releases are to land (Note: the numbers in this paragraph include up to 4
significant figures. This is not to imply precision in the estimates, but is presented in this
manner so that the sum adds to 100%). Sewage sludge appears to be a major sink for
decaBDE. It should be noted that municipal solid waste landfills arc major
environmental reservoirs for decaBDE, and while this analysis presumes there are no
environmental releases to air or water, landfills in the United States contain about 857
MT/yr decaBDE as EEE waste.
REFERENCES,
Alaee, M.; Arias, P; Sjodin, A; Bergman, A. (2003), An overview of commercially used
brominated flame retardants, their applications, their use patterns in different
countries/regions and possible modes of release. Environ International 29:683-689.
ATSDR (Agency for Toxic Substances and Disease Registry) (2004). Toxicological
profile for Polybrominated biphcnyls and Polybrominated diphenyl ethers. U.S.
Department of Health and Human Services, Atlanta, GA. September.
Australian Government (2007), Prohibition on introduction of pentabromodiphenyl ether.
Newsletter. March, 2007. Available:
http://mcnas.enewsletter.com.aU/link/id/4Of3d45d9f63472723c/page.html#9864505938b
7bd8046ff.
BSEF. 2006a. An Introduction to Bromine, Bromine Science
and Environmental Forum, Brussels . Available: http://www,bscfsite.
eom/docs/bromine.pdf [accessed 30 May 2006],
BSEF. 2006b BFRs: applications for brominated flame retardants. Bromine Science and
Environmental Forum, Brussels. Available: http://www.bsefsite.com/docs/bromine.pdf
[accessed 30 May 2006].
CalEPA (State of California Environmental Protection Agency ) (2006). Polybrominated
diphenyl ethers: recommendations to reduce exposure in California. A report of the
CalEPA PBDE Workgroup, Sacramento, CA. February.
Code of Federal Regulations (40 CFR Part 258) Subtitle D of RCRA: Criteria for
Municipal Solid Waste Landfills (MSWLFs)).
2-30
-------
Cobb, D. (1995). Analysis of FR chemicals added to foams, fabric, batting, loose till and
barriers. Memorandum from David Cobb to Dale R. Ray, Project Manager, Upholstered
Furniture, Consumer Products Safety Commission, May 10, 2005.
Directorate-General Environment (2005). RoHS substances (Hg, Pb, CrfVT), Cd, PBB
and PBDE) in electrical and electronic equipment in Belgium. Federal Public Service
Health, Food Chain Safety and Environment, Brussels, Belgium.
D'Silva, K. ( 2004) Brominated organic micropollutants - igniting the flame retardant
issue. Critical Reviews in Environmental Science and Teehnol 34: 34-141.
Ebert. J; Bahadir, M (2003) Formation of PBDD/F from flame-retarded plastic materials
under thermal stress. Environment International 29:711 - 716.
Federal Register ( 2005) Standards of Performance for New Stationary Sources and
Emission Guidelines for Existing Sources: Municipal Waste Combustors. Federal
Register 60 (243): 65387-65436.
European Union (2001). Risk assessment report: diphenyl ether, pentabromo derivative
(pentabromodiphenyl ether). European Chemicals Bureau, Luxenbourg. ISBN 92-894-
0479-5.
European Union (2002). Risk assessment report: bis(pentabromodiphenyl) ether).
European Chemicals Bureau. Luxenbourg, Belgium.
European Union (2003). Risk assessment report: diphenyl ether, octabromo derivative
(octabromodiphenyl ether). European Chemicals Bureau, Luxenbourg.
European Court of Justice (2008 ). Case C-14/06, Denmark vs European Union to exempt
decaBDE from regulation. Available: http://curia.europa.eu/jurisp/cgi-
bin/form.pl?lang=en&newform=niewform&jurcdj=jurcdj&docj=docj&docav=docav&doc
som=docsom&typeord=ALLTYP&numaff=&ddatefs=;&mdatefs~&ydatefs~&ddatefe=-&
mdatefe=&ydatefe=&uomusuel=&domaine=&mots=decabde&resmax=100&Submit=Su
bmit [accessed 1 June 2008],
Hale, RC; La Guardia, MJ; Harvey,E; Mainor, TM (2002). Potential role of fire retardant-
treated pol yurcthane foam as a source of brominated diphenyl ethers to the US
environment. Chemosphere 46: 729-735.
Harrad, S; Wijesekera, R; Hunter, S; Halliwell, C; Baker, R(2003). Preliminary
assessment of U.K. human dietary and inhalation exposure to polybrominated dipheny l
ethers. Environ. Sci. Teehnol. 38: 2345-2350.
2-31
-------
Hazratian, S. and Harrad, S. (2006) Causes of variability in concentrations of
polychlorinated biphenyls and polybrominated diphenyl ethers in indoor air. Environ.
Sci. Tcchnol. 40: 7584-7589.
IPCS (International Program on Chemical Safety) (1994). Environmental Health Criteria
162: Brominated Diphenyl Ethers. World Health Organization, Geneva, Switzerland.
Illinois EPA (2006). DacaBDE study: A review of available scientific research. A report
to the general assembly and the governor in response to Public Act 94-100. Illinois
Environmental Protection Agency. January.
Kemmlein, S; Bergmami, M; Jann, O. (2005). Standard measurement method for the
determination of polybrominated flame retardants (pentabromodiphenylether,
octabromodiphenylether) in products. Federal Institute for Materials Research and
Testing (BAM), Federal Ministry of Environment, Berlin, Federal Republic of Germany.
Research Report.UBA-FB 000839/e.
Kemmlein, S; Bergmann, M; Jann, O. (2003). Emission of flame retardants from
consumer products and building Materials. Federal Institute for Materials Research and
Testing (BAM), Federal Ministry of Environment, Berlin, Federal Republic of Germany.
Report No. UBA-FB.
La Guardia, M.I.; Hale, R.C, Harvey, E. (2006). Detailed polybrominated diphenyl ether
(PBDE) congener composition of w idely used penta-, octa- and deca-PBDE technical
flame-retardant mixtures. Environmental Science and Technol. 40, 6247-6254.
Morf, L.S., Tremp, J., Gloor, R., Huber, Y., Stengele, M., Zennegg, M. (2005)
Brominated flame retardants in waste electrical and electronics equipment: substance
flows in a recycling plant. Environ Sci and Technol 39: 8691-8699.
NAS (2002) Biosolids Applied to Land: Advancing Standards and Practices. Board on
Environmental Studies and Toxicology, National Research Council, the National
Academy of Science, The National Academies Press, Washington, DC.
NCEL (National Caucus of Environmental Legislators) (2005). Enacted and introduced
PBDE legislation 2005. NCEL, Bethesda, MD. (internet site: http://www.ncel.net/).
NCEL(National Caucus of Environmental Legislators) (2007). Enacted and introduced
PBDE legislation. NCEL, Bethesda, MD. (internet site: http://www.ncel.net/).
North, K.D.(2004). Tracking polybrominated diphenyl ether releases in a wastewater
treatment plant effluent, Palo Alto, California. Environ. Sci. Technol.38 (17): 4484 -
4488.
2-32
-------
Odusanya, D.O; Okoukwo, J.O; Botha, B (2008), Polybrominatcd diphcnyl ethers
(PBDEs) in leachates from, selected landfill sites in South Africa. Chetnospherc, In Press.
Available online 10 April 2008.
Osako, M; Kim, Y; Sakai, S (2004) Leaching of brominated flame retardants in leachate
from landfills in Japan. Chemosphcre 57: 571-1579.
Palm, A.; Brorstrom-Lunnden, E.; Breivik, K. (2004). Transport and fate of
polybrominatcd diphenyl ethers in the Baltic and Arctic regions. Jointly prepard by the
Swedish Environmental Research Institute and the Norwegian Institute for Air Research
for the Nordic Council of Ministers, Copenhagen, Denmark. TemaNord 2004:554.
Peele, C. (2004). Washington State polybrominatcd diphenyl ether (PBDE) chemical
action plan: Interim Plan. Prepared by the Washington State Dept. of Health and the
Washington State Dept. of Ecology, Olympia, W A. Publication No. 04-03-056.
Prevedouoro, K; Jones, K.C.: Sweetman, A.J. (2004). Estimation of the production,
consumption, and atmospheric emissions of pentabrominated diphenyl ether in Europe
between 1970 and 2000. Environmental Science and Tcchnol. 38(12):3224-3231.
Pure Strategies, Inc. (2005). Decabromodiphenylether: An investigation of non-halogen
substitutes in electronic enclosure and textile applications. Prepared for the The Lowell
Center for Sustainable Production, University of Massachusetts Lowell, MA
Rahman, F; Langford, K.H; Scrimshar, M.D.,; Lester, J.N (2001). Polybrominatcd
diphenyl ether ( PBDE) flame retardants. The Science of the Total Environment 275:1-
17.
Sakai, S; Hirai.Y; Aizawa, H; Ota, S; Muroishi,Y (2006). Emission inventory of deca-
brominated diphenyl ether (DBDE) in Japan. J Mater Cycles Waste Manag (8): 56-62.
USEPA (1999) Biosolids Generation, Use, and Disposal in the United States. U.S.
Environmental Protection Agency, Municipal and Industrial Solid Waste Division
Office of Solid Waste, Washington, DC. EPA530-R-99-009.
t SI: PA (2006) Municipal Solid Waste Generation, Recycling, and Disposal in the United
States: Facts and Figures for 2005. U.S. Environmental Protection Agency, Office of
Solid Waste, Washington, DC. EPA530-R-06-011.
USEPA (2008a) TRI 2006 Public Data Release e-Rcport. Available on-line at
http://www.epa.gov/tri/tridata/tri04/ereport/2004eReport.pdf.
USEPA (2008b) Clean Watersheds Needs Survey 2004 Report to Congress. Office of
Water, U.S. Environmental Protection Agency, Washington, DC. January, 2008.
2-33
-------
U.S. Fire Administration (2004). Fire in the United States. 1992-2001. Thirteenth edition.
U.S. Department of Homeland Security. Washington DC. Report number FA-
286.October.
Washington State (2007).House Bill 1024 -Phasing out the use of polybrominated
diphcnyl ethers. Available on-line at:
http://apps.leg.wa.gov/billinfo/summary.aspx?bill==1024&year=2007#doeumcnts.
Weber, R; Kuch, B (2003) Relevance of BFRs and thermal conditions on the formation
pathways of brominated and brominated chlorinated dibenzodioxins and dibenzofurans.
Environment International 29: 699- 710.
Yang, Y B ; Sharifi, V N; Swithenbank, J (2007) Numerical simulation of Municipal
Solid Waste incineration in a moving-grate furnace and the effect of waste moisture
content. Progress in Computational Fluid Dynamics. 7 (5):261 - 273.
2-34
-------
Table 2.1. Approximate BDE homologue and congencr-spccific content of
commercial PBDE formulations
Congener
Identity
Commercial Formulations of Polybrominated Diphenyl
Ether Flame Retardants
PentaBDE
(Mass %
Composition)
OctaBDE
(Mass %
Composition)
DecaBDE
(Mass %
Composition)
BDE 17
<0.1%
BDE 28
0.2%
Total TriBDE
0-1%
0.00001%
BDE 47
25 - 37%
0.00003%
BDE 66
< 1%
Total TetraBDE
24 -38%
0.00003%
BDE 85
2%
BDE 99
35% - 50%
0.002%
BDE 100
6- 10%
Total PentaBDE
50 - 62%
0.002%
BDE 138
0.5%
BDE 153
3 - 5%
5- 10%
0.001%
BDE 154
2 - 4%
1 5%
Total HexaBDE
4-12%
l'i 12%
0.001%
BDE 183
40%
BDE 190
1%
Total HeptaBDE
43 -58%
0.003%
BDE 196
8%
BDE 197
21%
BDE 203
5- 35%
Total OctaBDE
26 - 35%
BDE 207
7%
2.1%
BDE 208
10%
0.1%
Total NonaBDE
8-14" o
2 - 3%
BDE 209
0 3%
97.8%
This table was composed from information contained in the following references: European
Union (2001); European Union (2003); European Union (2002); Kemmlein, et al (2005); Peele
(2004); Main Dept of Health (2005); Palm, et al, (2004); IPCS (1994); Peltala and Yla-Mononen
(2000); La Guardia et al (2006).
2-35
-------
Table 2-2. Estimated environmental releases (MT) of BDE congeners from
United States decaBDE production facilities in 2006
Congener
Stack Emissions
Fugitive
Emissions
Land
Application
Surface
Water
BDE 209
27.18
2.03
1.59
1.91
BDE 207
0.58
0.04
0.03
0.04
BDE 208
0.03
0.002
0.002
0.002
Calculations based on BDE congener distribution in decaBDE from Tabic 2-1.
Table 2-3. Mean concentration (ng/kg) of penta- and octaBDE in various flame
retarded polymers
Polymer
PentaBDE
OctaBDE
ABS
1.187 (±0.058)
0.528 (±0.035)
HIPS
0
1.057 (±0.061)
PUR-H
1.414 (± 0.053)
0
EP
1.414 (±0.148)
0
Source: Kemmtein et al, (2005). ABS ~ acrylonitrile-bntadiene-styrene-copolymer; HIPs - high-
impact polystyrene; PUR-H = polyurethane hard foam; EP = epoxy resin
Table 2-4. Mean concentrations of FBDEs (mg/kg) found in electrical and
electronic waste material at a recycling plant in Switzerland
Product
PentaBDE
OctaBDE
DecaBDE
Copper cable
25 (± 10)
100 (± 150)
170 (± 110)
Printed circuit boards
17 (± 7)
10 (± 1)
27 (± 19)
TV housings (wood)
10 (± 4)
10 (±4)
20 (± 30)
TV/PC housings
(plastic)
50 (±3)
7500 (± 600)
4800 (± 400)
TV housing rear covers
50 (± 20)
7700 (±3600)
13000 (±9000)
Source: Morf et al. (2005).
2-36
-------
Table 2-5. BDE congener concentrations (mg/kg) present in EEE waste
components sampled at a recycling facility in 2002
BDE
Congener
PC Screen
Housings
Television
Housings
Rear Covers
Fine
Particulates
(Dust)
Fine-Grained
Plastics
{5-10 mm)
Printed
Circuit
Boards
mg/kg
mg/kg
mg/kg
mg/kg
mg/kg
BDE 28
0.32
0.32
6.5
0.47
1.3
BDE 47
3.0
2.2
39
5.1
9.6
BDE 99
6.5
3,8
52
7.5
12
BDE 100
0.65
0.28
4.4
0.78
0.72
BDE 153
598
450
31
110
4.0
BDE 154
66
38
6.9
12
1.1
BDE 183
3800
3900
150
690
12
BDE 209
7,300
13,000
760
2,500
89
PentaBDE
13
7.9
120
17
28
OctaBDE
11,000
11,000
420
2,000
34
DecaBDE
7,300
13,000
760
2,500
89
Source: Morfet al. (2005).
2-37
-------
Table 2-6. Summary of estimated BDE congener emissions (MT/yr) to air from a
10-yr service life of fire retarded plastic products'1
BDE Congener
Europe
Americas
Asia
Global
BDE 17
0.001
0.027
0.001
0.027
BDE 28
0.001
0.053
0.001
0.053
BDE 47
0.174
8.241
0.174
8.705
BDE 66
0.006
0.266
0.006
0.281
BDE 85
0.011
0.532
0.011
0.562
BDE 99
0.264
12.494
0.264
13.198
BDE 100
0.045
2.127
0.045
2.246
BDE 138
0.003
0.133
0.003
0.14
BDE 153
0.046
1.121
0.08
1.269
BDE 154
0.026
0.821
0.04
0.901
BDE 183
0.125
0.308
0.308
0.778
BDE 190
0.003
0.008
0.008
0.019
BDE 196
0.025
0.062
0.062
0.156
BDE 197
0.066
0.162
0.162
0.408
BDE 203
0.063
0.154
0.154
0.389
BDE 207
0.07
0.21
0.201
0.495
BDE 208
0.034
0.084
0.084
0.211
BDE 209
2.257
7.273
6.829
16.672
TOTALS
3.22
34.08
8.43
46.51
a/This analysis assumes the market demand for 2001 from BSEF. 2006a.
2-38
-------
Figure 2-1. General structure of polybrominated diphenyl ethers.
0
where in - 11 = 1-10
2-39
-------
Figure 2-2. Lifecycle of polybrominated diphenyl ethers.
Textile
Formulation
Production
Resin
Master
Compounding
Production
batching
Textile
Backcoafine
Injection
Molding
Furniture
Production
Component
Assembly
Original Equipment
Manufacturer
Retail
Consumer
Landfill
Recycling
Incineration
2-40
-------
Figure 2-3. TR] data showing environmental releases (kg) of deuiBDH from
primary production facilities in 2006, and total environmental releases from 1997
2006 (US EPA. 2008a)
Total TR1 Releases 1997 - 2006
2-41
-------
Figure 2-4. Volatilization of BDE 47 from PBDE-treated consumer products in
the U.K.
3MO04
• BOe 47 high * 833E *t kw
3:
3 OOC •
•
* *
2.50E«-04
~ ~
JS
p! 006-04 :
s
A
|l.50e»M :
~ .*
~ *
$
* «
«3
i.oaE-o*
0 CK5F. * 00 « »« « *»«H*-i
~ ^
~ ' ^
*$*
—
S96S 1*H5 1978
!9»3 1685 W ?99S 2000
2005
Illustration from Prevedouoro et al. (2004). Assumes an average
Product life of 10 yrs. High and low estimates correspond to range of predicted
consumption rates for the year 2000 (74 and 65 MX, respectively).
2-42
-------
3. ENVIRONMENTAL FATE OF POLYBROMINATED
DIPHENYL ETHERS
3.1, INTRODUCTION
This chapter reviews the physical and chemical properties important to
understanding the environmental fate of PBDEs. Included in this chapter are brief
descriptions on how the chemicals move and partition among the environmental media
once released into the open environment. The physical and chemical properties control
tendencies for PBDEs to move into air, soil, water, and sediments, and to exchange
among environmental compartments. They also indicate the physical form and phase of
the chemical present in air and water. The physical and chemical properties influence the
extent to which biotic and abiotic processes may transform or degrade PBDEs in the
environment. Riocon cent rat ion properties indicate the relative propensities of BDf:
congeners to bioaccumulate and biomagnify in biota and ecological food chains. Overall,
the properties most important for understanding the environmental behavior of PBDEs
are water solubility (WS), octanol/water partition coefficient (k.,..,). octanol/air partition
coefficient (KqaX the Henry's Law constant (H), and vapor pressure (VP).
3.2 PHYSICAL/CHEMICAL PROPERTIES
The following subsection is a general overview of the physical and chemical
properties of the various PBDE congeners. Although there is the possibility of 209 BDE
congeners, this section focuses on most of the congeners that comprise the deca-, octa-,
and penta-BDE commercial formulations as fire retardants.
3.2.1. Water Solubility (WS)
The chemical parameter of WS describes how readily a chemical compound
(referred to as the solute) dissolves in water at a given temperature. This is one of the
most important parameters in environmental chemistry. Highly soluble chemicals arc
easily and quickly distributed within surface water and groundwater. A typical method of
estimating the aqueous solubility is by adding an excess amount of the pure chemical to
water until equilibrium is achieved and the maximum concentration in water is reached.
3-1
-------
Water solubility is typically measured in units of milligram solute per liter of water
(mg/L), and at a standard reference temperature (20°-25°C.). The higher the value, the
more soluble the chemical is in water. Generally chemicals are considered to have a low,
medium, or high aqueous solubility if their WS is < 0.1, 0.1-10, and >10 mg/L (measured
at 25°C), respectively (NAS, 2001). Table 3-1 summarizes the WTS of commercial PBDE
formulations and for some specific PBDE congeners. PBDEs generally have low WS.
Organic contaminants with low solubility do not readily dissolve in surface water
(relatively hydrophobic) and also typically have high log octanol water partition
coefficients, which suggest a high absorption capacity for organic carbon in soils and
sediments.
3.2.2. Octanol/Water Partition Coefficient (Kow)
Another useful chemical property for indicating how a chemical moves between
the aqueous phases and into sediments and biota is the octanol/water partition coefficient,
or K„w. The Kow value has become an important parameter in the prediction and
understanding of the fate of organic chemicals in the aquatic and terrestrial environments.
This property is derived through laboratory experiment and quantitative structure activity
relationships to related chemicals, and it is defined as the ratio of the concentration of a
contaminant in //-octanol (normal octanol) over the concentration of the same
contaminant in water. The //-octanol is intended to generally represent all organic
substances. As noted previously, it is related to water solubility in that the higher the
chemical water solubility, the lower the propensity for bioaccumulation (the lower the
value for Kmv). It is expressed as a unitless value. In the experimental derivation of the
Kow, octanol is an organic solvent used as a surrogate for organic matter. Although
dimensionless, the KoW coefficient is usually expressed as the logarithm, base 10, of the
ratio value (i.e., log Kow). In general, organic chemicals with an experimental log Kow
coefficient equal to or greater than 5.0 have the property of being very hydrophobic,
being tightly absorbed to organic matter, and possessing a high tendency to
bioaccumulate. Table 3-2 summarizes log Kmv values for brominated diphenyl ethers.
All PBDEs tested have high log Kow coefficients, which indicate that they have a high
tendency for bioaccumulation.
3-2
-------
3.2.3, Henry's Law Constant (H)
Organic contaminants can transfer from water bodies into the air and from air
back into water bodies. Henry's Law constant (H) is an air-water partition coefficient,
and it is a measure of the chemical's equilibrium distribution between air and water at a
specified temperature. In general, H is derived from the ratio of vapor pressure to the
chemical's aqueous solubility. Knowledge of H is essential to understanding the
direction and mass flux of contaminants transferring from water to air. The rate of
volatilization from water to air, and the scavenging of the gaseous phase of the
contaminant in air by precipitation (i.e., wet deposition) are governed by H. Usually H is
expressed as atm-nrVmol. Volatilization becomes an important transfer mechanism when
the computed H is between 10"' and 10"? atm m mot (Ritter et aL, 1995). Chemicals
with H values greater than 10~3 atm-nrVmol rapidly volatilize into air. PBDEs are low-
volatile organic chemicals with H at 1()"4 atm-nrVmol or less, with lower H at higher
degrees of bromination. Table 3-3 is a summary of estimated H for PBDEs.
3.2.4. Vapor Pressure (VP)
Volatilization of a chemical and its presence in air is driven by the VP of the
chemical. VP also controls the phase distribution of a chemical in air (e.g., the
proportion that exists in the vapor and particle-bound phases). The VP is a measure of
the force per unit area (i.e., pressure) exerted by a chemical in vapor phase while in
equilibrium with its liquid or solid phase at a specified temperature. VP is usually
expressed in units of Pascals (Pa) or millimeter of Hg (mmHg). In general, volatile
organic compounds (VQC) have a solid phase vapor pressure > 10 Pa at an atmospheric
temperature of 25° C (Olsen and Neilson, 2001). The semi volatile organic compounds
tend to have solid phase vapor pressures < 1 Pa at 25° C. PBDEs are semivolatile organic
compounds. Vapor pressures of PBDE decrease with increasing molecular weight and
degree of bromination. In a theoretical context, the subcooled liquid VPs of PBDE
compounds best represents their tendency to partition between the vapor and particle
bound phases in air. Because of this, the scientific literature on chemical and physical
3-3
-------
properties has primarily reported values for Pl and not for P$. However, Ps of the BDE
congeners can be calculated from the Pt, using equation 3-1 (Paasivirta et al., 1999).
log Ps = log Pl + (ASf ?R)x{ (1 -Tm/T)i 2.3026 ) (3-1)
Where:
Ps - solid phase vapor pressure. Fa
f'i = subcooled liquid vapor pressure. Pa
ASf = entropy of fusion, ~ 56 J kf'moL1 (Wania and Dugani, 2003)
R - ideal gas constant, 8.3143 J K'niof1
Tm = melting point temperature, K
T - reference temperature, K
Table 3-4 summarizes the calculated solid phase (Ps) and subcooled liquid vapor
pressures ( Pi) of some BDE congeners.
3.2.5. Octanol/Air Partition Coefficient (Koa)
The octanol/air partition coefficient, log Koa is the ratio of the concentration of
the chemical in air versus the concentration of the chemical in octanol when the octanol-
air system is at equilibrium (Harner and Shoeib, 2002). The value of Koa is
dimensionless. As with VP, log Kqa is temperature dependent. For every 10°C decrease
in atmospheric temperature there is a corresponding 2- to 3-fold increase in partitioning
of semivolatile organics to the organic phase (Harner and Shoeib, 2002). In broad terms,
the log Koa is suggestive of the environmental cycling of semivolatile compounds
between the air and organic phases such as soil particles, air particles, and vegetation.
The underlying assumption to the log Kqa is that n-octanol is a good surrogate for
absorption to all organic carbon. The greater the Log Kqa (e.g. ~ 10) the stronger is the
propensity of the atmospheric BDE congener to absorb to the organic content of soils and
vegetation (Wania, et al, 2002). It is, therefore, a relative indicator of chemical mobility,
and the tendency to exchange from the atmosphere to the surface. Log Kqa values from
6-11 indicate that atmospheric PBDEs strongly absorb into forest canopies and other
-------
vegetative biomass (Su et al., 2007). Table 3-5 summarizes the octanol/air partition
coefficients (log KoA) of PBDE compounds.
3.2.6. Gas-Particle Partitioning In Air
Semivolatile organic compounds present in ambient air partition between the
gaseous phase and the particle-bound phase. The physics of this behavior is controlled
by the VP of the chemical, the temperature of the surrounding air, and the availability of
airborne particulate matter. In tonus of environmental fate, the importance of this
phenomenon is that the aerosol-bound portion of the contaminant is subject to be
transported through the atmosphere over large geographical distances. The wet and dry
deposition of the contaminated particles is the most significant mechanism for removal of
the particle-bound phase of PBDEs from the air. The gas phase is a good predictor of the
air-to-leaf transfer of the semivolatile organic compound and of the possibility for
deposition into leafy vegetation, which is integral to the terrestrial food chain. Thus the
vapor-phase portion of the contaminant may be more significant in terms of human
exposures by way of the dietary pathway.
The fraction of the semivolatile compound that is particle bound (<£) can be
calculated from the subcooled liquid VP using the Junge-Pankow model as indicated in
equation 3-2 (Su et al., 2006).
= (c© )/(PL +(c-©)) (3-2)
Where:
if) - fraction of the compound adsorbed to aerosol particles
PL = saturation subcooled liquid phase VP, Pa
0 -- the particle surface area per unit volume of air, cm" aerosol/cnf of air
c =a constant, 17.2 Pa-cm
Table 3-6 provides the particle bound and vapor phases of the PBDE commercial
formulations and BDE congeners calculated from equation 3-2 and assuming the
subcooled liquid VP in Table 3-4 and an ambient air temperature of 25°C. The value for
0 in the equation assumes the particle surface area per unit volume air of 1.51-06
-------
enr/cm3 typical of aerosols in background urban air (Whitby, 1978). The predicted
particle-bound phase of the PBDE increases with decreasing VP and increasing number
of bromine atoms on the molecule.
By these calculations, >80% of the mono-, di-, tri-, and tetra- and <15% of the
penta-, hexa-, hepta-, octa-, and deca-BDE congeners may be present in the vapor phase
at an ambient air temperature of 25°C. Greater than 85% of the penta-dcca congeners are
expected to be associated with airborne particles. As seen in Table 3-6, 72% of BDE 85
is particle-bound, and this appears to be an anomalous calculation due to the fact that
Tittlemier et al. (2002) may have overestimated the subcooled liquid vapor pressure at
25°C. About 50-60% ofpcnta-BDE congeners are predicted be present in the vapor
phase. These predicted BDE phase distributions in ambient air agree with the
observations of Strandberg et al. (2001) for distributions in air over the Great Lakes.
Strandberg et al. (2001) found that at 20°C, about 80% of the tetrabromo homologu.es are
in the gas phase and about 70% of the hexabromo homologues are associated with the
particle phase. In a study by Chen et al. (2006), the estimated phase distribution of
PBDE congeners measured in urban air of southern China was as follows: BDE 28 (97%
vapor phase); BDE 47 (80% vapor phase); BDE 66 (77% vapor phase); and BDE 99
through BDE 154 ranged from 50-85% in the particle-bound phase. No ambient air
temperature was reported.
3.3. BIOACCUMULATION, BIOCONCENTRATION, AND
BIOMAGNIFICATION OF PBDES IN THE AQUATIC ENVIRONMENT
Bioaccumulation describes a process whereby an organism acquires a body
burden of a chemical in relation to contact through all possible pathways of exposure (i.e.
dietary absorption, transport across the respiratory surface, dermal absorption, and
inhalation; Gobas and Morrison. 2000). Bioaccumulation for an aquatic organism occurs
from contact of the organism with a chemical contaminant in the water column, the
sediments, and through the organism's food chain. The parameter that is often used to
model this entire process is the Bioaccumulation Factor (log BAF). This contrasts the
Bioconcentration Factor (BCF), which has been used to measure the accumulation from
the water column only. In the case of lipophilic and hydrophobic chemicals,
3-6
-------
bioaccumulation into fish becomes important in terms of human fish consumption. The
log BAF is derived as a logarithm of the ratio of the concentration of a chemical in the
tissue of an aquatic organism over the concentration of the chemical in water, in a real
aquatic setting. A laboratory-derived BCF entails an experimental setup where the
accumulation from the water column only is determined from the concentration of the
contaminant in the water tank divided by the concentration in the fish tissue. The log
BAF is expressed in units of liters per kilogram of tissue, and can be normalized to apply
to wet weight, dry weight, or percent lipid of the organism. The biomagnification. factor
(BMF) is the ratio of the concentration of the chemical in an organism to the
concentration of the chemical in the diet of the organism. The BMF is an indication of
increases in concentration of the chemical as it moves up trophic levels in the aquatic
ecosystem.
Only a few studies could be located in the scientific literature that has estimated
BAFs for PBDE congeners. Table 3-7 summarizes BAF and BCF for various aquatic
species.
The United Nations Environmental Program (UNEP) established a screening level
for assuming high potential for bioaccumulation of the contaminants into aquatic species
(WWF, 2005). The criteria for a high potential for bioaccumulation is that the BCF or
BAF in aquatic species for the chemical is greater than 5,000 (log BAF = 3.7). With the
exception of BDE 85, it appears that the tri-, tetra-, and penta-BDE congeners exceed
these criteria, and, therefore, have a high potential for bioaeeumulating in aquatic
organisms. BDE 209 has a very low potential for bioaeeumulating within the aquatic
food web.
3.4. BIO TIC AND ABIOTIC DEB RO Ml N ATI O N AND TRANSFORMATION
OF PBDEs
Certain biotic and abiotic processes can transform PBDEs in the environment.
The processes most important to the breakdown of PBDEs include biodegradation,
biotransformation, and photolysis. Biodegradation involves the breakdown of PBDEs by
aerobic and anaerobic microorganisms into smaller compounds. The microbial
organisms transform the contaminants through metabolic or enzymatic processes.
3-7
-------
Biotransformation is the conversion of the chemical structure of the PBDEs through
metabolic pathways. Similar to biodegradation, the reaction is catalyzed by enzymes, but
the process occurs in vivo in animals. Photolysis involves the breakdown of PBDEs by
the action and the energy of sunlight. All of these environmental fate processes can
involve the stripping of bromine atoms from the molecule, a process referred to as
'denomination,' With denomination, the higher brominated BDE congeners can
breakdown to form lower brominated species. This section will focus on each of these
degradation pathways. It should be stressed that there is only limited scientific
information supporting the action of these pathways in degrading, transforming, and
debrominating PBDEs.
3,4,1. Microbial Degradation of PBDEs
In theory, microbial communities can degrade and transform organic
contaminants present in soils and sediments. This occurs because the microorganisms
use the contaminants as a source of carbon (NAS, 1993). Carbon is a building block in
the development of new cells during reproduction and growth. In addition, the microbes
can extract electrons from the organic contaminant to obtain energy. The microorganism
gains energy by breaking chemical bonds and transferring electrons away from the
contaminant (NAS, 1993).
One variation of microbial degradation of organic compounds is reductive
dehalogenation. In this process, the microbes catalyze reactions that promote the
replacement of a halogen atom with a hydrogen atom on the organic compound. This is
the primary microbial degradation pathway for higher molecular weight PBDEs.
Although the higher brominated congeners are debrominated aerobically, the lower
molecular weight congeners that are products of this process may be further
debrominated by aerobic bacterial degradation via oxidative dehalogenation (Kim et al
2007).
Several recent studies have provided evidence of microbial mediated reductive
denomination of decaBDE and octaBDE under laboratory conditions. Research is
currently focused on anaerobic microbial strains that have demonstrated the ability to
dehalogenate organic compounds in vitro. In particular, Dehalococcoides ethenogenes
3-8
-------
195, Dehalococcoides sp, strain BAV1, and Sulfurospirillum rrmltivorans can
dechlorinate a broad range of chlorinated compounds including trichlorocthylene (TCE),
pcrchlorocthylcne (PERC), chlorobenzcnes, PCBs, and PCDDs (Fennell ct al., 2004; Wu
et al., 2002), and, therefore, are being investigated for their potential to debrominate
PBDEs {He et al., 2006).
Sulfurospirillum multivorans bacterium incubated with deca-BDE has induced
reductive debromination of deca-BDE in vitro to yield octa-BDE and hepta-BDE after a
contact time of 2 months (He et al., 2006). The octa- and hepta-BDE did not farther
debrominate in the presence of the microbe—even after one year. Thus the
microorganism was specific to the degradation of deca-BDE and is incapable of
denominating lower brominated PBDE compounds. The organism normally has an
affinity for TCE. but when deca-BDE was dissolved in TCE and exposed to ,s\
multivaomas, reductive denomination did not occur. However TCE was completely
dechlorinated to c/s-DCE.
Dehalococcoides ethenogenes 195 bacterium was experimentally tried in an
attempt to debrominate octa-BDE (He et al., 2006). D. ethenogenes of the 195 strain
contains reductive dehalogcnase (RD) genes and is the only bacterium known to
reduetively dechlorinate tetraehloroethene and trichloroethane to ethane (Seshadri et al.
2005). D. ethenogenes strain 195 was initially isolated from an anaerobic sewage
treatment plant (STP) digester containing sewage sludge (Seshadri, ct al. 2005). When
incubated for 6 months with octaBDE dissolved in TCE, the bacterium showed marked
debromination of octaBDE to yield penta, hexa, and heptaBDE congeners (He et al.,
2006). However, when incubated with octaBDE without the TCE solvent, no
debromination occurred. The authors combined a number of Dehalococcoides species in
a product they called ANAS 195 and incubated it with octa-BDE dissolved in TCE for 12
months. One hundred thirty nMols octaBDE comprised of BDE congeners 153, 183, 196
203, 207, and 208 added to the culture yielded 11.5 nMoles of combined BDE congeners
47, 49, 99, and 154.
When normalized on a cell-count basis, the ANAS 195 culture had a rate of
debromination of octaBDE that was twice the rate of D. ethenogenes 195. An additional
strain of Dehalococcoides species, Dehalococcoides sp strain BAV1, is the only known
3-9
-------
microorganism that dechlorinates the lower chlorinated organic compounds (e.g.,
dichloroethane, to ethane; Krajmalnik-Browrs et al.. 2004). and, therefore. He ct al.
(2006) investigated this species for the potential to further debrominate tetra- and penta-
BDE congeners. The experiment showed that bromines in the 2 (ortho) and 4 (para)
positions on the PBDE molecule (e.g., BDE 99 and 47) were the most resistant to
microbial reductive denomination.
In another study, Gereck et al. (2005) debrominated BDE 209 in vitro to yield
BDE s 206, 207. and 208 by contact with anaerobic mesophilic microorganisms
indigenous to raw sewage sludge (microbial species not identified). Methane was formed
as a product of microbial respiration, and the amount of BDE 209 decreased by 30%
within 238 days. This disappearance rate of BDE 209 was statistically significant (p
-0.037), and it corresponded to a pseudo first-order degradation rate constant of IxlO"3
per day, which is equivalent to a half life of approximately 690 days. Denomination of
BDE 209 was evident in the subsequent formation of nonaBDE and a number of
unresolved octa-BDE congeners. The removal of bromine atoms in the para and mefa
positions on the BDE 209 molecule resulted in the formation of BDE 208 and BDE 207,
respecti vely. These products were formed at concentrations that were 5% of the initial
concentration of BDE 209. Although BDE 206 was observed in the experiment, it could
not unambiguously be concluded that it was a product of microbial degradation of BDE
209.
Rayne et al. (2003a) demonstrated the in vitro anaerobic microbial reductive
denomination of BDE 15. A non-differentiated anaerobic bacterium from the in situ
remediation of contaminants in a polluted river system was colonized by passing
contaminated water through a bioreactor. Reductive debromination of BDE 15 occurred
in vitro at hydraulic retention times of 3.4 and 6.8 hours in a fixed-film, plug-flow
biological reactor. The products formed were BDE 3 (4-monoBDE) and non-substituted
diphenyl ether.
3-10
-------
3.4.2, In Vivo Metabolic Debromination in Animals
Once absorbed into an organism, metabolic processes can break down PBDEs.
There is a growing body of scientific evidence to suggest that certain freshwater and
marine fish species and marine mammals are capable of metabolicaliy denominating
PBDE congeners in vivo. This section is an overview of available information on
debromination in wildlife.
3.4.2.1. Evidence for Debromination in Fish
Certain fish species have the capacity to debrominate PBDE congeners in vivo.
This involves the removal of bromine atoms in the para and meta positions of a higher
molecular weight PBDE congener to form lower brominated PBDE compounds.
Debromination appears to mainly transpire in the gastrointestinal tract and fish liver, and
catalyzed by one of three possible pathways: the deiodinase (DI) thyroid hormone
regulating enzymes (Stapleton et al., 2006a; Tomy et al., 2004), endogenous microbial
activity in the gut (Stapleton et al., 2004), and the enzymes of the microsomal
monooxygenase system (Stapleton et al.5 2004). To date, no one single pathway has been
experimentally elucidated. It is not known whether all fish have the metabolic capacity
to biotransform PB DEs, nor is it known whether the rate of debromination PBDEs varies
among fish species.
Laboratory studies of rainbow trout, lake trout, and carp, involving fish food
spiked with pure BDE 209, have clearly shown accumulation of lower brominated BDE
congeners not initially present in the feed. This evidence is suggestive of metabolic
synthesis of lower brominated congeners through debromination of BDE 209 (Stapleton
et al., 2004, 2006a, 2006b; Tomy et al., 2004; Kierkegaard et al., 1999).
The suggestion of metabolic debromination of decaBDE in fish was initially
observed in the late 1990's in an in vivo rainbow trout study (Kierkegaard et al., 1999).
In this study, rainbow trout were fed fish food spiked with 7.5-10 trig of decaBDE/kg of
body weight/day for 16, 49, or 120 days. Muscle tissues and livers from exposed fish and
controls were analyzed separately for PBDE. Pre ferential accumulation of PBDE in the
liver was observed over the 120-day-study period. Due to a lack of pure analytical
3-11
-------
standards for the BDE congeners, it was not possible to identify specific BDE congener
present in the fish; only analysis ofhomologue groups was possible. However, hexaBDE
and the first cluting octaBDE congeners were identified as possible products of metabolic
debromination of decaBDE, because these congeners were absent in both the spiked fish
food and the control fish. This early suggestion of the potential for debromination of
PBDEs to occur in fish prompted further in vivo studies.
Tomy et al. <2004) investigated the capacity of juvenile lake trout to debrominate
a broad range of BDE congeners. The fish were exposed in aquaria to known amounts of
thirteen BDE congeners spiked onto fish food. The BDE congeners included 28, 47, 66,
77, 85, 99, 100, 138, 153, 154, 183, 190, and 209. The duration of dietary exposure to
the spiked food was for 56 d, followed by 112 d of depuration. Five fish were sacrificed
and tissues were sampled on days 0, 7, 14, 28, and 56 of the period of feeding and on
days 7, 14, 28, 56, and 112 of the depuration period. The strongest evidence from this
study for die debromination of higher molecular weight BDE congeners in lake trout
comes from the comparison of patterns of BDE congeners present in untreated fish
tissues to congeners present in the spiked fish food. The presence of one penta (an
unidentified congener) and two hexaBDEs (BDE 140 and an unidentified hexaBDE) in
the fish tissues served as an indicator of the occurrence of in vivo debrominaization,
because these BDE congeners were absent in both the fish food and in the tissues from
the control group. It was proposed that BDE 140 was directly formed from the metabolic
debromination of BDE 209. Evidence for this is that the bromine substitution pattern of
BDE 140 could not have been derived from the substitution pattern of the other two
higher brominated congeners tested (e.g., BDE 190 or BDE 183).
Common carp that were fed a diet spiked with either BDE 99 or BDE 183 for 62
days showed significant in vivo debrominazation of both BDE congeners (Stapleton et
al, 2004). Six experimental and two control groups (12 fish to a group) were fed spiked
or clean fish pellets, respectively. One fish from each tank was sampled on days 0, 5, 10,
20, 30, 44, 62, and during 37 days of depuration following dietary exposure to BDEs.
The gastrointestinal tissues and liver was dissected from the fish and analyzed for BDEs
as was the remaining homogenized fish sample. BDE 99 was debrominated within carp
tissues (gastrointestinal tract and liver) to BDE 47 at a rate of 9.5% ± 0.8% of the BDE
3-12
-------
99 dietary concentrations. It is interesting to note that BDE 28 was detected in low
concentrations in carp tissues from exposure days 20 to 62, suggesting that BDE 47 was
further denominated, although this was not proven. BDE 183 was debrominated to BDE
154 and another unidentified hcxaBDE congener. The rate of biotransformation to
hexaBDE congeners was estimated to be 14.2% ± 5% of the BDE 183 dietary
concentrations. The site of denomination was the intestinal tract. The authors
postulated that the metabolic denomination of BDEs 99 and 183 could be mediated by
either intestinal microflora, and/or by endogenous enzyme systems such as the hepatic
monooxygenase enzymes.
Stapleton et al. (2006a) provided additional confirmational evidence of metabolic
debromination of BDE 209 in fish. Sixty juvenile rainbow trout were randomly placed in
aquaria holding 15 fish each (with one aquarium containing a control group). The
exposed fish were fed fishmeal spiked with 939 ± 14 ng BDE 209/g wet weight for a
period of 5 months, which is equivalent to a dose of 9.4 jig BDE 209/kg body
weight/day. After 5 months of exposure, the fish were sacrificed and fresh blood was
drawn from the dorsal aorta. Livers were removed and analyzed for PBDE congeners
and assayed for microsomal activity. Approximately 401 ±68 ng BDE 209/g (wet
weight) PBDEs accumulated in the fish liver during the experiment, indicating that the
liver was the primary accumulating tissue. In the pooled serum, collected prior to the
experiment, it was determined that BDE 209 averaged <2.4 ng/g serum, whereas during
the experiment, the concentration in pooled blood ranged from 26-40 ng/g serum.
Steady-state serum concentrations were achieved in the last two months of the
experiment. Denomination was observed with the appearance of lower brominated
congeners that increased in concentration in the fish tissues throughout the exposure
period. Three nonaBDE, six octaBDE, and four heptaBDE congeners were present in the
exposed fish tissues that were not present in the spike food mixture. High resolution
GC/MS confirmed the identification of the following BDE congeners in the fish tissue:
188, 201, 202, 207, and 208. While BDE 207 and 208 initially increased in tissue
concentration, concentrations decreased over the last 2 months of the exposure period.
The authors speculated that the BDE 207 and 208 continued to be metabolized, albeit, the
study did not identify specific metabolites. A mass balance by Stapleton et al. (2006a)
3-13
-------
indicated that debromination of BDE 209 to less brommated congeners occurred in fish
liver and the metabolites were subsequently transported throughout the body by means of
the circulatory system.
In summary, four major fish feeding studies have clearly demonstrated the
potential for in vivo debromination of PBDEs. The higher brominated BDE 209 and the
decaBDE commercial formulation have been successfully debrominated through a
metabolic pathway in fish. The exact mechanism remains unknown. Current theory
suggests that enzymes involved in thyroid hormone catabolism, as well as hepatic
enzymes, may play a role in the biotransformation of the higher brominated species to
lower brominated congeners. It is not clear whether all fish species, both marine and
freshwater, have the innate propensity for biolransforming PBDE, but debromination of
PBDEs has been observed in juvenile rainbow trout, juvenile lake trout, and common
carp (all of which are freshwater species). It seems that bromine atoms in the meta
positions, positions 3 and 3", on the BDE molecule are most easily removed.
3.4.2.2. Evidence for Debrominaiion in the Rat
In vivo debromination of PBDEs has been observed in a rat feeding study. Huwe
and Smith (2007) found evidence of the metabolic debrominazation of BDE 209 in male
Sprague-Dawley rats fed a commercial formulation of deca-BDE (98.5% BDE 209).
Other congeners detected in the formulation included nonaBDEs, octa-BDEs, and a trace
of BDE 183. Deca-BDE was mixed in com oil to a concentration of 18.9pg of BDE
209/rnL of oil. Eighteen rats were fed an oral, daily dose of 3.8 jig in 200 jiL of oil/rat
(equivalent to 0.3 j-tg/g/d of the total diet) for a period of 21 days. A control group was
fed the standard rat diet over the same period. Experimental rats were sacrificed in
groups of three on days 0, 3, 7, 1.0, 14, and 21 following the cessation of dosing with
BDE 209. Prior to sacri fice, daily samples of urine and feces were collected from all rats.
After sacrifice, samples of the blood (plasma), the liver, the gastrointestinal (GI) tract,
and the remaining carcass were collected from each rat, and homogenized prior to sample
analysis by an isotope dilution GC/MS method. BDE 209, nonaBDE, and octaBDE
congeners were found to accumulate in the rat liver of the dosed animals at amounts that
were 2-3 times higher than in other tissues. Evidence of metabolic debromination of
3-14
-------
BDE 209 to lower congeners was observed from an apparent 160% increase in the tissue
concentration of BDEs 197, 201, and 207 as compared to levels in the feed. Huwe and
Smith (2007) postulated that the possible formation of BDEs 197 and 207 resulted from
removal of bromine atoms from the met a positions on the BDE 209 molecule. The
formation of BDE 201 was postulated to occur from the debromination along para and
meta positions of the BDE 209 molecule.
The rat may serve as an indication of the possibility that metabolic debromination
of BDE congeners occurs generally in mammals including humans. However, this
remains to be proven.
3.4.2.3. Evidence for Debromination in Birds
There is ev idence for the metabolic debromination of higher brominated PBDEs
in chickens (Pirard and Pauw, 2007) and starlings (Van den Steen et al., 2007). Pirard
and Pauw (2007) fed seven Sexaline hens with a diet containing 3.4 mg/kg penta-BDE
formulation (De-71, Great Lakes Chemicals) for 14 weeks. Egg samples and daily
excreta samples were collected during the experiment. At the end of 14 weeks, the hens
were sacrificed and samples of fat and liver were taken for chemical analysis. With
regard to tissue and egg distributions, 3,030, 3,711, and 2,826 ng/g lipid adjusted total
PBDEs were detected in the liver, adipose tissue, and eggs, respectively. Pirard and
Pauw (2007) derived BCFs for select BDE congeners by dividing the congener
concentrations detected in abdominal chicken fat to the congener concentrations in the
chicken feed. The estimated BCFs were as follows: BDE 47 = 0.7; BDE 100 = 1.8; BDE
99 = 0.6; BDE 154 = 2.2; BDE 153 = 2.0; and BDE 183 = 1.0. The authors investigated
absorption/excretion percentages of BDE congeners excreted by the second week of
dosing. Higher proportions of lower brominated compounds (i.e., BDEs 47, 100, and
99), were found in chicken excreta compared to hexa- and hepta-BDEs. Pirard and Pauw
(2007) postulated that because it was unlikely that the high amount of BDE 47 found in
excreta came from the fraction unabsorbed in the intestinal tract, the excess BDE 47 in
chicken excreta wras evidence of formation from the reductive metabolic debromination
of congeners BDE 99 and BDE 100. The authors further speculated that BDE 153 could
be debrominated to form BDE 99 and BDE 154 debrominated to form BDEs 99 and 100.
3-15
-------
Van den Steen et al. (2007) studied starlings for the bioaccumulation and tissue
distribution of BDE 209, Four adult male starlings were housed in a large outdoor aviary
and exposed to a solution of BDE 209 in peanut oil through a silastic tube implant. The
exposed group received an implantation dose of 46.8 |ag BDE 209/day for 76 days, and a
control group (n=3) received an implant filled with unfortified peanut oil over the same
time period. During the exposure period, 300 (,iL of blood was taken from each bird
every 3-7 days. Following the exposure period, the experimental birds were euthanized
and the pectoral muscle and bird liver were excised for analysis of BDE 209. It was
found from analyzing the silastic tubes that only 50% of the total dose of BDE 209
diffused from the silastic tube implant into the bird. The half-life of BDE 209 in the
blood of the starlings was estimated to be 13 days (95% confidence interval: 11 to 18
days). BDE 209 accumulated in muscle tissue at a rate 2-fold higher than in liver
explained by the higher metabolic activity of the liver. In addition to BDE 209, other
BDE congeners were detected in bird tissues. The detection of substantial amounts of
BDE congeners 208, 207,206, 197, 196, and 183 in liver and muscle was evidence of
metabolic debromination of deca-BDE in avian tissues (Van den Steen et al., 2007).
3.4.2.4. Evidence for Debromination in House Cats
Dye et al (2007) reported on the possible metabolic debromination of PBDE in
house cats. The purpose of this study was to investigate the relationship between the
incidence of feline hyperthyroidism in cats and the dietary ingestion of dry cat food
contaminated with PBDE in addition to ingestion of PBDE contaminated house dust.
The congeners BDE 47, 99, 207, and 209 were most frequently detected in blood serum
of 23 exposed cats, and were related to dietary their intake. The cats consuming only
canned-wet cat food (mostly comprised of fish)
BDE 207 > BDE 47 > BDE 99. The remaining cats that consumed both food types
(n=l 1) exhibited a mix of BDE congeners in serum with no one congener dominating the
others.
3-16
-------
The evidence for possible metabolic denomination of BDE 209 in house cats
stemmed from a comparison of the BDE congener distributions present in serum to the
BDE congeners profiles in the dry and wet cat food. The contamination of PBDEs in dry
cat food reflected the congener profile of decaBDE, with BDE: 209 representing 83-93%
of total PBDE present in the feed. Because of the high content of BDE 209 in dry cat
food, BDE 209 dominated serum in cats only consuming the dry food. BDE 209
accounting for 4.2%, 21%, and 30% of serum PBDE levels in house cats consuming
canned-, mixed-, and dry-food, respectively. It was noted that BDE 207 was consistently
present in scrum in significant concentrations of the dry food eaters as compared to the
consumers of the other food types. BDE 207 accounted for 4.5%, 9.8%, and 17% of the
PBDE levels detected in cats consuming canned-, mixed-, and dry-food-eaters,
respectively. BDE 207 was present in the dry food eaters at approximately 50% of the
total concentration of BDE 209 which is uncharacteristic of the decaBDE congener
profile (BDE 207 is approximately 1% of BDE congeners present in decaBDE), and was
not the pattern observed in the wet food eaters. Moreover, the ratio of BDE 207 to BDE
209 in cat serum was relatively constant in all dry cat food eaters. The authors regarded
these data as possible evidence for the metabolic denomination of BDE 209 to form
BDE 207. These data are only suggestive of the in vivo denomination in house cats, but
if confirmed through replicate studies, would generally imply the possibility for
metabolic debromination of BDEs in humans.
3.4.3. Abiotic Degradation of PBDEs
Abiotic degradation predominantly occurs in the atmosphere and on soil surfaces.
The energy of sunlight can degrade PBDEs in air and soils via photolysis, and the
presence of the hydroxy! radical in air can deplete some PBDEs present in the
atmosphere. The following is a brief review of these processes.
3.4.3.1. Photogradation of PBDEs
Several studies have shown that higher brominatcd BDE congeners can
photodegrade to form lower brominated congeners as photochemical byproducts. The
photodegradation is defined as the photochemical transformation of a molecule into
3-17
-------
lower molecular weight fragments, usually in an oxidation process (IUPAC, 1996). This
term is widely used in the destruction (oxidation) of pollutants by UV-based processes;
e.g., the absorption of photons present in wavelengths found in sunlight, i.e., ultraviolet
(UV) radiation. Photodegradation of PBDE occurs from the removal of a bromine atom
on the PBDE molecule, thus transforming the higher molecular weight BDE to lower
brominated congeners.
Fang et al (2008) reported on the experimental photodegration of BDE 28, BDE
47, BDE 99. BDE 100, BDE 153 and BDE 183 dissolved inhcxane. Known
concentrations of the BDE congeners (approximately 150 ng/mL) present in solvent were
exposed to >290 nm UV wavelengths in a photo reactor consisting of a 500 W mercury
lamp, quartz tubes, Pyrex glass tubes and a water cooling system.. The experiments were
performed in triplicate and a control was used that consisted of BDE congener solutions
placed in total darkness. Photodegradation occurred for all studied congeners, and
generally followed pseudo-first-order kinetics. The photodegradation half lives of the
BDE congeners discerned from this study were: BDE 28 = 4.97 h; BDE 47 - 2.53 h;
BDE 99 = 0.32 h; BDE 100 = 6.46 h; BDE 153 = 0.29 h; and BDE 183 = 0.26 h.
Generally the higher brominated BDE congeners degraded at a faster rate than the lower
brominated congeners. It was concluded that the main decomposition mechanism
induced by photolysis was reductive debromination to form lower brominated BDE
congeners. Fang et al. (2008) observed the following photodegradation products:
• BDE 28 (triBDE) photodegraded to form BDE 8 and BDE 15 (both
diBDEs); the diBDEs further degraded to form the monoBDE congener
BDE 1 and BDE 3;
• BDE 47 (tetraBDE) photodegraded to form two tri-brominated species
(BDE 17 and 28), which in turn further are debrominated to form di-
brominated BDE congeners
• BDE 4, 8 and 15, debrominated to form monoBDE congeners BDE 1 and
BDE 3;
• BDE 99 (pentaBDE) photodegraded to form the tetraBDE congeners 66,
49 and 47 (tetraBDE), which subsequently undergo further photolysis.
3-18
-------
• BDE 100 (pentaBDE) photodegraded to form BDE 75 and BDE 47 (both
tctraBDEs).
• BDE 153 (hexaBDE) photodegraded to form BDE 99, 101 and 117
(pcntaBDEs).
Kajiwara et al (2007) reported on experimentally photodegrading BDE 209
contained in flame retarded plast ic. The purpose of their study was to invest igate the
potential for the photolytic debromination of the brominated fire rctardants (BFR)
decaBDE and deeabromodiphenyl ethane (DBDPE) in treated plastics. Four different fire
retarded plastic samples were evaluated including pure high impact polystyrene (HIPS)
not treated with BFR, HIPS treated with decaBDE (0.15% wt/wt), HIPS treated with
DBDPE (about 0.10% wt/wt), and used TV casing made from HIPS. It was noted that
DBDPE is a chemical substitute for decaBDE, and has a chemical structure similar to
BDE 209. The plastic samples were dissolved in toluene and shaken overnight to be
completely mixed. The toluene was evaporated by air drying in a dark room. After
drying, solid materials including TV casing were pulverized in liquid nitrogen chamber.
The pulverized plastic samples were initially passed through a 300 micron (pm) sieve and
then a 106 pm sieve. The plastic powder collected between two sieves (106 pm-300 pm)
was used in the UV irradiation experiment. Aliquots of the plastic powder (0.3 g) were
placed in sealed quartz tubes and subsequently exposed to natural sunlight at time
intervals of 0, 7, 14, 28, 56, and 112 days. Controls (tubes of the same powdered plastic
samples) were placed in a temperature controlled dark room. Following each test period,
samples were extracted and then analyzed using high resolution gas chromatography
coupled with high resolution mass spectrometry in the ion monitoring mode. Photolysis
of BDE 209 in the decaBDE treated HIPS samples was observed after one week of
exposure to sunlight. The BDE 209 concentration in the plastic decreased and hepta-,
oct a-, and nona-BDE congeners were were formed as products of photolysis. The
mechanism of the photodegradation of BDE 209 was postulated to be by the process of
eliminating bromine atoms on the molecule. Approximately 80% of the BDE 209 had
degraded after 112 days of exposure to sunlight, but only 5% of the BDE 209 mass was
converted to lower brominated BDE congeners, suggesting the formation of unknown
products of photolysis. Tri though octa polybrominated dibenzofurans (PBDFs) appeared
3-19
-------
in the irradiated samples suggesting that the photolysis of BDE 209 formed PBDFs, but
not in sufficient quantities to elose the mass balance of the products formed. No clear
pattern of photodegradation was observed with the other plastic samples. No loss of
BDE 209 was observed in the control samples. This study demonstrated the
photodegradation of BDE 209 in treated HIPS by natural sunlight.
Staplcton and Dodder (2008) experimentally photodegraded decaBDE present in
house dust by exposure to natural sunlight. Two different house dust materials were
obtained from the National Institute of Standards and Technology (N1ST) in Maryland.
The first material was typical indoor house dust known as Standard Reference Material
2585 (SRM 2585). During chemical analysis it was found to be contaminated with a
variety of BDE congeners ranging from tri- to decabromo-substituted congeners. The
second dust sample was identified as SRM 2583, and was a standard reference material
used in the analysis of metals and organics in dust. The SRM 2583 dust sample was
preeleaned to be free of PBDE contamination and then spiked with 1.298 g of a solution
of BDE 209 dissolved in toluene. The sample was air dried to evaporate the toluene,
resulting in a concentration of 2,180 ng BDE 209 /g dry weight (dwt) in the SRM 2583
dust sample. The SRM 2585 dust sample was not preeleaned nor spiked with BDE 209.
Aliquots of both SRM dust samples were placed into two 4.5 mL methylacrylate
chambers (referred to as UV cuvettes) that are routinely used for measuring UV
absorbance of test materials. The UV cuvettes containing the dust samples were exposed
to natural sunlight outdoors in a tray lined with aluminum foil. Except during periods of
precipitation, the cuvettes were placed outside daily, Monday through Friday, from
approximately 9:00 AM to 4:00 PM until 200-h cumulative exposure to sunlight was
achieved. Samples were transferred to a laboratory where they were extracted and
analyzed for PBDE using high resolution gas chromatograph coupled with high
resolution mass spectrometry operated in the electron capture chemical ionization mode.
The average intensity of sunlight over the duration of the experiment was 545 watts per
square meter (W/irr), with a range of 61 to 929 W/m\ The average of outdoor
temperature was 26.9°C and a range from 18.7 to 32°C. Photodegradation of BDE 209
was observed in both SRM dust samples after 100 h exposure to natural sunlight. The
first-order BDE 209 photodegradation rates were calculated as 2.3 E-3 and 1.7 E-3 per
3-20
-------
hour in the spiked (SRM 2583) and natural (SRM 2585) dust samples, respectively. The
oetabrominated congeners, BDE 201 and BDE 202, were observed in both reference dust
samples to be products formed from the debromination of BDE 209 induced by
photodegradation. In the SRM 2583 (spiked) sample, additional congeners were formed:
BDE 183, BDE 197, BDE 202, BDE 203, BDE 206, BDE 207 and BDE 208. From these
data, the authors calculated a mass balance based on the photodegradation of BDE 209 in
the spiked dust samples. After 200 h exposure to natural sunlight, the initial
concentration of BDE 209 was observed to have decreased by approximately 38%. The
authors concl uded that 35% of the decrease in the concentration of BDE 209 as due to
debromination and the subsequent formation of lower brominated BDE congeners and
3% of the decrease in concentration was postulated to have been caused by the
volatilization of BDE 209.
Rayne et al. (2006) photodegraded BDE 153 in acetonitrile, distilled water, and
seawater at ultraviolet radiation wavelengths (UV radiation) of 302 nm (nanometers).
BDE 153 dissolved in acetonitrile and irradiated for 5 mill formed three primary
photodegradation products: (1) penta-BDE isomers 99, 101, and 118 (20% yield); (2) the
brominated dibenzofuran congener 1,2,4,7,8-PeBDF (30% yield); and (3) three non-
speciated tetrabrominated 2-hydroxybiphenyls (20% yield). Continued irradiation up to
1 hr caused the photo degradation of the penta-BDE congeners to form the tetra-BDE
isomers BDE 47, 49, 66, and 77. The irradiation of BDE 153 in distilled and/or seawater
produced the same tetra-BDE isomer byproducts, albeit, less efficiently than acetonitrile.
Rayne et al. (2006) were unable to conclude firmly whether or not the photodegradation
of PBDEs in aquatic systems under natural conditions is a viable process.
Bezares-Cruz et al. (2004) decomposed BDE 209 to form lower brominated BDE
congeners through natural sunlight at wavelengths of 300, 305.5, 311.4, 317.6, 325.4,
332.4, and 368 nm. Deca-BDE formulation (approximately 97% BDE 209) was
dissolved in hexane to create three solutions of BDE 209 ranging in concentrations from
6.92E-06" to 6.14E-06 micro molar (ju.M). Control samples were prepared in the same
manner but kept in darkness. Samples were exposed to sunlight on clear days in the
summer and fall of 2003, BDE 209 dissolved in hexane and exposed to sunlight
photodegraded within minutes. After 30-45 minutes of exposure to mid-afternoon
3-21
-------
sunlight on October 23 and July 2 of 2003, the BDE 209 concentration was reduced to
approximately 5% and 1% of the initial concentration, respectively, corresponding to a
pseudo-first-order reaction rate of 1.11 x I0"'7s and 1.86 x 10 ; s. The higher reaction rate
in July was due to the increased intensity of solar flux as compared with October. There
was no evidence of degradation of B I) E 209 in the control solutions not exposed to
sunlight. The solar irradiation of BDE 209 dissolved in hexane catalyzed the reductive
debromination of the congener. Forty-three PBDE congeners of various bromine
substitutions were formed during different times of exposures to sunlight. After 5
minutes of solar irradiation, the disappearance of BDE 209 was matched by the initially
rapid formation of nonaBDE congeners. The other congeners (oetaBDEs, heptaBDEs,
and hcxaBDEs) accumulate successively over the 60-min-exposure period. OetaBDEs
are transformed to heptaBDEs that are then transformed to hcxaBDEs. With respect to
the possibility of the photodegradation of BDE 209 in natural waters, Bezares-Cruz et al
(2004) postulated that the photochemical reaction would expected to be somewhat
attenuated by sorption of BDE 209 onto colloidal particles in the water column, and by
the light attenuation properties of humic materials in aquatic systems. Furthermore, the
presence of hydrogen donors necessary in invoke the reaction would likely be at lower
concentrations in natural waters as compared to hexane.
Soderstrom et ai (2004) reported on the experimental photodegradation of BDE
209 in various matrices, including toluene, silica gel, sand, soil and sediment. UV-
exposure experiments were conducted both in the laboratory with artificial UV-light (all
matrices) and under natural conditions with outdoor sunlight (sand, soil, sediment). To
begin the experiment, a 0.5 gram sample of each of the matrices was placed into Pyrex-
tubes and fortified with a solution of 10.5 ng/pL decaBDE dissolved in toluene). The
toluene was then allowed to evaporate while the samples were kept in the dark. For the
exposure to artificial light, the samples were placed in an apparatus consisting of four
mercury UV lamps equipped with filters to mimic the sunlight spectra in the UV range of
300-400 nm. The irradiance intensity from the UV lamps was estimated to have been 1.6
mW/cm*. The exposure of soil and sediment were extended to an additional 121 and 244
hours. For the conditions of natural sunlight, the Pyrex tubes containing fortified samples
of each matrix were placed in a tray and put on the roof of the laboratory during the
3-22
-------
month of July, 2007. Maximum UV-irradiance from the sun at mid-day was measured to
have been 2.3 mW/cnf. Exposure durations of the samples ranged from 0 to 96 hours.
All samples were transferred to a laboratory, extraeted and analyzed using
HRGC/HRMS. Sodersttom et al (2004) found that BDE 209 photolytically degraded
from exposure to both artificial and natural sunlight. With exposure to artificial light,
photodegradation was more rapid and complete when BDE 209 was associated with
toluene or silica gel than with sand or sediment. BDE 209 in toluene or silica gel
degraded to 1% of the initial BDE 209 concentration over 8 h when exposed to artificial
light, but degraded to about 21 and 57% over 96 hrs when associated with sand and
sediment. By comparison, natural sunlight degraded BDE 209 in sand and sediment to
about 36% and 43% over a 96 h exposure. The half-life of BDE 209 varied by matrix
and artificial or natural sunlight. The half-life BDE 209 in toluene or silica exposed to
artificial light was less than 0.25 h. By comparison, BDE 209 in sand, sediment and soil
had estimated half lives of 12 h, 40 - 60 h, and 150 - 200 h, respectively, when exposed to
artificial light. For exposure to natural sunlight, the half lives of BDE 209 in sand and
sediments were 37 h and 80 h, respectively. Soderstrom et al (2004) reported that the
photoytic debromination of BDE 209 formed lower brominated BDEs in both artificial
and natural sunlight conditions. These included: BDE 47 (silica gel only), BDE 100
(toluene only), BDE 119 (toluene, sand, sediments and soil). BDE 99 (toluene and silica
gel), BDE 154 (all matrices), BDE 153 (toluene, sand, sediments and soil), BDE 140 (all
matrices except soil), BDE 128 (all matrices), BDE 183 (all matrices), and BDEs 206,
207 and 208 (all matrices).
Sanchez-Prado et al. (2005) investigated the photodegradation of pentaPBDE at
two UV irradiation intensities. The pentaBDE technical formulation was dissolved in
cyclohexane to a concentration of 10 ugmL. The penta-BDE mixture was comprised of
the BDE congeners 47 (4.1 fig/ml), 85 (0.1 |ug/ml), 99 (1.2 |u.g/ niL), 100 (4.1 |ug/'mL),
153 (0.23 j.ig/mL), and 154 (0.34 ng/mL). A 5 ml aliquot of the solution was adsorbed to
100 jam polydimethylsiloxane (PDMS) fibers and placed in a laboratory photo reactor
equipped with two low-pressure mercury lamps (8-10 W, 254 nm) for up to 1 hour.
Equal control samples were stored in complete darkness. Chemical analyses showed no
degradation of the BDE mixture occurred in the controls. Debromination of the BDE
3-23
-------
congeners occurred over both intensities of UV irradiation. The reaction rate was
observed to be dependent on the degree of bromination (e.g., increasing rate of
photodegradation with increasing number of bromine atoms on the molecule). The
reaction rate was observed to be independent of the irradiation intensity. Approximately
21 BDE congeners were observed as the degradation products from the irradiation of the
mixture of BDEs 47, 85, 99, 100, 153, and 154. Based on sequential formation of these
products, Sanchez-Prado et al. (2005) postulated that the principle photodegradation
pathway was reductive denomination,
3.4.3.2. Reaction with the Hydroxy! Radical
Tropospheric reactions with the hydroxyl radical have been observed to be an
important atmospheric degradation pathway for halogenated hydrocarbons. The OH"
radical is generally formed from the photolysis of ozone (O;,) to form a single oxygen
atom that subsequently combines with hydrogen. No direct measurements of
atmospheric degradation rates of PBDEs from interaction with the OH" radical could be
located in the literature. However, atmospheric half-lives have been estimated from
quantitative structural activity relationships. Assuming an OH" radical concentration of 5
x l O^ hydroxyl radicals per cm3, the atmospheric half-lives of penta-, octa-, and deca-
bromodiphenyl ether homologues have been estimated to be 29, 140, and 476 days,
respectively (ATSDR, 2004). A degradation rate constant for BDE 99 has been
estimated to be 1.27 x 10"12 cm' molecule s using the same assumption of OH" radical
concentration (European Union, 2001). The ATSDR concluded that atmospheric
degradation from reaction with the OH radical is likely to be an insignificant atmospheric
loss mechanism (ATSDR, 2004).
3.5. THERMAL DECOMPOSITION OF PBDE
PBDEs are added to plastics, textiles, and other materials in order to inhibit
combustibility and delay the spread of fire (see Chapter 2 for more detail). Basically the
PBDE additives interfere with the combustion process by forming bromine gas which, in
turn, displaces the O2 necessary to sustain the oxidation reactions. In addition, bromine
radicals are formed which interfere with the chain reactions of hydrogen and hydroxyl
3-24
-------
radicals in the fire. The duality of these processes severely retards and constrains thermal
oxidation.
The combustion of PBDE treated materials can lead to the formation and emission
of polybrontinated dibenzofurans (PBDFs) in the smoke (Weber and Kuch, 2003). It is
believed that the molecular structure is suitable for PBDEs to be precursor compounds to
PBDFs formation. Debromination of BDE 209 occurs at temperatures of about SOOT,
leading to the formation of brominated biphenyls. Further combustion of brominated
biphenyls causes PBDFs to be formed within the combustion gases (Webber and Kuch,
2003). Rupp and Metzger (2005) formed both PBDFs and PBDDs as byproducts from
the thermolysis of BDE 47 and BDE 153 in quartz ampoules heated to temperatures
varying from 250-500oC for 5-10 minutes. OctaBDF (octabrominated dibenzofuran)
and heptaBDF congeners can be formed and emitted during the extrusion of HIPS plastic
treated with decabromodiphcnyl ether/antimony (antimony (III) oxide at 275°C (Luljk et
al„ 1992). The yields of PBDFs showed a significant increase as a function of the
number of extrusion cycles. Lower brominated diphenyl ethers were present in the
emissions from the extrusion process confirming that the debromination of decaBDE and
subsequent exchange of bromine atoms with hydrogen was a necessary step to the
thermolytic formation of PBDFs. In a series of experiments of combusting deca-, penta-,
and, octaBDEs in quartz ampoules at 510-630°C, Buser (1987) formed PBDFs and
PBDDs from the intramolecular cyclization reactions involving the attack of an oxygen
atom on the diphenyl ether molecule. The initial steps to these reactions were the
debromination of the parent compound followed by hydrogen substitution.
Polybrominated benzenes and polybrominatcd phenols were also formed as combustion
byproducts.
3.6. PATTERNS OF ENVIRONMENTAL FATE OF PBDE
Commercial Octa and PentaBDE consist of mixtures of BDE congeners.
PentaBDE predominately contains BDE 47, BDE 99, BDE 100, BDE 153, and BDE 154.
and commercial octaBDE mainly contains hepta-, octa-, and hexaBDE congeners.
Commercial decaBDE are almost entirely BDE 209 with a small amount of nonaBDE.
The BDE congeners are dispersed into the open environment whenever the commercial
3-25
-------
formulations arc released from their manufacture, use, and disposal. The environmental
fate of PBDE congeners is dictated by their physical and chemical properties and their
propensity for biotic and abiotic transformation. These characteristics were reviewed in
previous sections to this chapter (see section 3.2 of this Chapter). Chapter 2 reviewed the
use and life-cycle of PBDE formulations, and presented estimates of environmental
releases. The purpose of this section is to describe, in general terms, the fate of PBDEs
once they are released into the environment.
PBDEs are mainly discharged into the air from production, use and recycling of
PBDE treated plastics, electronics, computers, textiles, and polyurcthane foam, (less than
5%) PBDEs are discharged into surface waters from industrial activity and sewage
treatment plants (STPs). The land disposal of sewage sludge and industrial sludges also
contributes to environmental loadings.
Dispersion of BDE congeners in the environment is governed by their respective
physical and chemical properties. Air and water are primary transport media, while soils
and sediments are environmental sinks. In general, BDE congeners are highly
hydrophobic and lipophilic compounds that have low water solubility and low vapor
pressures. The atmosphere and marine currents can transport BDEs over relatively long
distances (> 1,000 km). Evidence for this comes from the presence of PBDEs in the
tissues of deep ocean-dwelling whales and other marine mammals far from
anthropogenic sources activities. With the exception of BDE 209, BDE congeners
bioaccumulate into terrestrial and aquatic food webs. The body burdens of BDE
congeners in a wide variety of avian species, fish, and aquatic and terrestrial mammals
indigenous to geographical areas ranging from the equator to the poles also substantiate
BDE's propensity for long-range transport. The following subsections discuss the fate of
PBDEs in air, water, and soil.
3.6.1. Fate of PBDEs in Air
The air compartment is the primary medium of dispersing PBDEs over large
geographical areas. PBDE behavior in air is that of other semivolatile compounds (i.e.,
BDE congeners partition between the vapor phase and particle-bound phase in
accordance with their respective vapor pressures (sec Table 3-6). At standard
3-26
-------
temperature and pressure, the mono through tri brominated compounds are primarily in
vapor phase; the tetra and penta congeners split between vapor and particle bound phases
but predominate in the vapor phase; the hexa and oeta brominated BDEs are mostly
attached to atmospheric particles; and BDE 209 is exclusively adsorbed to atmospheric
particles. Experimental evidence suggests that the lower brominated BDEs, those that
predominate in the vapor phase, can be degraded in the atmosphere through reaction with
the hydroxyl radical as well as through photolytic chemistry. Wania and Dugani (2003)
investigated the propensity for long-range atmospheric transport through the atmosphere
with the application of four long-range atmospheric transport models. Wania and Dugani
(2003) concluded that the tetra- and penta-BDE congeners have a greater tendency to be
transported long distances than lower and higher brominated species. The characteristic
travel distances of different PBDEs were calculated from the air models. The
characteristic travel distance is defined as the horizontal distance traveled by a parcel of
air whereby 63% of the initial air concentration of PBDE is depleted (Wania and Dugani,
2003). The models showed that the characteristic travel distance ranged from 483 - 1,113
km for tetraBDE; 608 - 1,349 km for pentaBDEs; and 480 - 735 km for decaBDE
(Wania and Dugani, 2003). In theory,
• Vapor phase mono, di, and tri brominated compounds have the highest
tendency for long-range transport in the atmosphere because they are not
susceptible to particle surface deposition.
• The tetra and penta brominated congeners more equally partition between
vapor and particle-bound phases and have characteristic travel distances that
are somewhat less than the vapor-phase compounds.
• This bi-phased distribution in air increases the atmospheric removal of the
moderately brominated compounds because of the partial adsorption to
atmospheric particles and subsequent wet and dry deposition to the surface.
BDE 209 is almost entirely associated with atmospheric particles and, therefore, its long-
range transport potential is theoretically diminished by the deposition processes.
However, adsorption of BDE 209 to atmospheric aerosol particles leaves the possibility
that a small fraction of the BDE 209 contaminated aerosol can be transported over long
distances, which is dependent only on the fate of the aerosol (Wania and Dugani, 2003;
3-27
-------
Gouin ct al., 2006. There exist significant gaps in the understanding of how aerosols are
dispersed regionally and globally. This hinders the possibility of making definitive
conclusions on the long-range transport potential of BDE 209. Aerosol particles may
mimic the behavior of gases making them less prone to surface deposition; this may
explain the measurement of low air concentrations of BDE 209 in the Arctic whenever
the prevailing winds emanate from urban areas in Alaska and Canada ( Wang et al.,
2005). Others have observed the routine presence of the lighter, more volatile PBDE
congeners (i.e., BDE 28, 47, and 99) in air over remote oceanic regions, and BDE 209
has only been occasionally detected (Wurl et al, 2006a). It has been noted (see
subsection 3.4.3.1 above) that BDE 209 is labile to photodegradation by natural sunlight.
Photodegradation results in the loss of bromine atoms on the BDE 209 molecule to form
lower brominated BDE congeners as products of photolysis, and this may further explain
the low potential for long range transport of BDE 209. Photodegradation is likely a
major atmospheric process for degrading and transforming higher brominated PBDEs to
less brominated BDE congeners, and experiments have clearly shown that photodegration
of PBDEs is induced by the spectra of UV radiation under natural conditions.
Generally BDE 47, 99, and 209 dominate the congener pattern when detected in
urban air. The densely settled urban areas are geographically dispersed sources of the
BDE air burden, and a strong negative atmospheric concentration gradient from the urban
center out to rural and remote areas is often observed in air monitoring studies (Gouin et
al, 2005). This implies that urban areas contribute to background air concentrations of
PBDE in rural and remote settings, and should be considered as major area-wide sources
of PBDEs present in ambient air. Chapter 4, Section 4.5 discusses concentrations of
individual congeners and total BDEs in indoor and outdoor air in the United States and
abroad.
In addition to the physical processes of wet and dry surface deposition, and
photochemical reactions and degradation/debromination, the leafy surfaces of deciduous
forest canopies can significantly deplete PBDEs in the atmosphere via air-to-leaf transfer
(Su et al.. 2007). The uptake into leaves transfers the BDE congeners from the air to the
terrestrial ecosystem when the leaves drop to the surface. This process is seen as being
both vapor-phase and particle-bound interception by the leaf during the growing
3-28
-------
season—with the vapor phase more easily absorbed. The more volatile, low molecular
weight PBDE compounds are deposited primarily through gaseous diffusion and achieve
a partitioning equilibrium between the air and the leaf. The less volatile, heavier PBDE
compounds are scavenged by the leaf via dry particle deposition caused by particle
impaction and diffusion (Horstman and McLachlan, 1998). These processes of
vegetative atmospheric depletion of semivolatile compounds are also observed with
coniferous forests, grasses, and green leafy crops (Horstman and McLachlan, 1998).
Therefore dense forests, open grasslands, and productive farmlands functionally reduce
the air concentration of BDEs as contaminated air masses move over these areas. The
octanol/air partition coefficient (log Kqa) of B DE congeners (see Table 3-5) is an
important indicator of this propensity for uptake into leafy biomass with all congeners
having a log Kqa > 6. The atmospheric scavenging by forest canopies, grasses and other
leafy biomass can partially explain the seasonal variability of atmospheric BDEs. Long-
term monitoring studies indicate higher concentrations of BDEs in air during the cooler
winter months, and lower concentrations over the spring and summer (Gouin et aL
2005). An early springtime "bud-burst" effect of reducing the air concentrations of BDEs
at the onset of spring has been documented to occur (Gouin et al., 2005). In the spring
and summer, greater diurnal variability in air concentrations of BDEs is observed during
the warmer months, and, conversely, stable air concentrations of BDEs are observed
during the cooler winter months. The relative stability of BDEs over rural and remote
areas during the winter can be explained by diffusion of BDEs from urban centers
induced by a negative concentration gradient, whereas the variability in the spring and
summer reflects vegetative influences on the atmospheric concentrations (Gouin et at.,
2005).
Large bodies of water also have the capacity to influence the atmospheric
concentrations of BDEs by means of the air-water exchange of the contaminant. The rate
of exchange of BDEs is dependent on the state of equilibrium between air and water
concentrations. If the ratio of the BDE fugacity in water is greater than the BDE fugacity
in air (in the vapor phase), then there is a tendency for the BDE to volatilize from the
water into the air. Likewise, if the BDE vapor phase fugacity in air is greater than the
dissolved phase in water, then there is a tendency for the transfer of BDE from air into
3-29
-------
water. Equilibrium is achieved when the water-air fugacity ratio ~ 1 in which case there
is no air-water exchange of BDEs. Phytoplankton biomass on the surface of oceans plays
a role in influencing the fugacity gradient (Jaward et al., 2004). The BDEs in the
dissolved phase in the water column will be drawn to partition to the carbonaceous
biomass due to their high KoW (see Table 3-2). The dissolved phase concentration is
decreased, which, in turn, may induce diffusion from air to force the air-to-water
exchange. In this case, the capacity of an ocean to absorb BDEs from the atmosphere is
indirectly controlled by the density of surface biomass. BDEs have been measured in the
dissolved phase and in suspended particulates to seawater (Wurl ct al., 2006b). In coastal
waters off China, concentrations of total PBDE in the dissolved phase ranged between
40.2 and 228.2 pg/L, and between 8.1 and 69.1 pg/L in the suspended particulate matter
(Wurl et al., 2006b). In the San Francisco estuary BDE 47, 99, and 209 were found to be
the most abundant congeners detected in the dissolved phase (Oros ct al., 2005), whereas
Wurl et al. (2006b) found BDE 28, 47, and 100 dominated total PBDEs in the open ocean
with BDE 209 at detected only at trace levels. In the freshwater of Lake Ontario
approximately 60% of the total PBDE was composed of BDEs 47 and 99, with BDE 100,
153, and 154 congeners each contributing approximately 5 to 8% of the total
(Environment Canada, 2006). Section 4.2 provides further discussions on sediment
concentrations of PBDEs.
3.6.2. Fate of PBDEs in Water
Once encompassed within surface waters as a result of atmospheric deposition
and/or the direct discharge from anthropogenic source activities, the PBDEs partition
between the water column and the sediments in proportion to their physical-chemical
properties. The benthic sediments are a primary sink for PBDEs. PBDEs then can
bioaccumulate up the aquatic food web beginning with benthic organisms and ending
with predators at the top of the food chain, e.g., piscivorous fish, birds, and terrestrial
mammals.
In general, PBDE concentrations arc highest in sediment samples collected
downstream of the following: industrial/urban areas, outfalls to sewage treatment plants,
and urban locations without heavy industries. The lowest PBDE concentrations are
3-30
-------
generally found in sediments collected at remote and agricultural areas. BDE 209
appears to dominate the congener profile of aquatic sediments, however most congeners
have not been detected. Section 4.2 includes a complete description of sediment studies
that bear out these trends.
Only two studies could be located that evaluated the phase distribution of BDE
congeners in water. In Lake Michigan, Streets et al. (2006) found dissolved phase BDE
congeners 47, 99, 100, and 66 in concentrations ranging from 0.13 to 10 pg/L (for
individual congeners). Three congeners were detected in the particle-bound phase:
BDEs 47, 99, and 100 were found at concentrations ranging from 0.18 to 1.4 pg/L. BDE
209 was not evaluated in this study. Wurl et al. (2006a) evaluated the phase distribution
of BDEs 28, 47, 99, 100, 153, 156, 183, and 209 in the sea-surface microlayer (SML) and
subsurface seavvater (SSW) from locations off the coast of Hong Kong. The SML of the
seawater is defined as a 100-jim thick boundary layer between the atmosphere and the
ocean surface that is comprised of naturally occurring organic matter and micrometer-
sized suspended particles. The subsurface is the layer below the SML down to a depth of
1 m. Over all samples, total PBDE concentrations ranged from 11.3 to 62.3 pg/L in the
dissolved phase and from 26.2 to 32.5 pg/L in the particle-bound phase. BDE 209 was
detected at trace levels in the dissolved phase at all sampling locations. Only the BDE
congeners 28 and 47 were detected in the particle-bound phase of the SML. Below the
SML in the SSW, BDEs 28 and 47 were detected at three sampling locations. BDE 99
was the only other congener detected in the dissolved phase of the subsurface layer.
Once in the sediments or in the suspended organic matter in the water column, the
PBDEs bioaccumulate into the ecological food chains beginning with the benthie
organisms and continuing up to the top predators. The BDE congeners have a high
capacity for bioaccumulation and biomagnifications in biota as indicated by their
relatively high Kow factors (see Table 3-2). Contamination of fish tissue with BDE
congeners exposes fish-eating fish, piscivorous birds, terrestrial animals, and humans to
BDEs via the dietary pathway. There is suggestive evidence that fish are able to
transform highly brominated PBDE congeners {in vivo) to lower brominated PBDE
congeners through the process of metabolic debromination (see section 3.4.2.1 .for further
discussion). It has been suggested that metabolic debromination may cause the formation
3-31
-------
of BDE 47 and BDE 99. Piscivorous raptors bioaccumulatc PBDEs through their
habitual consumption of contaminated fish. These birds may also have the capacity for
metabolic debromination of higher brominated BDE congeners to lower brominated
congeners, although the evidence for this is highly suggestive, and stems primarily from
the study of chickens and starlings. PBDE congeners have been detected in marine
mammals such as whales, seals, and porpoises, indicating their exposures to PBDEs from
their diet. Bioaccumulation within a broad range of animals is an indication of the
consequences of widespread PBDE contamination in the aquatic and marine trophic
networks (see section 3.7.1. Bioaccumulation in the Aquatic Environment, below, for a
more detailed discussion).
3.6.3. Fate of PBDEs in Soil
Soil is a major sink and environmental reservoir for PBDEs. Atmospheric PBDEs
exchange between the air and the soil compartments by means of gas and particle
depositional processes. The atmospheric gas-particlc partitioning of BDE congeners and
the type of biomass covering the soil surface has an affect on the flux of PBDEs from air
to soil. There are apparent differences in the magnitude of soil concentrations of BDE
congeners in soils overlaid by coniferous woodlands, deciduous woodlands, and grass.
When the soils are covered by vegetation, especially forests, the scavenging of
semivolatile organic chemicals from the atmosphere is enhanced, which, in turn,
increases their depositional flux to the terrestrial surface relative to deposition to bare soil
(MacLeod, 2003). In consideration of gas-particle partitioning and ground cover, Palma
et al. (2002) estimated the environmental media partitioning efficiency of air releases of
unsubstituted diphenyl ether (DE), and BDEs 47, 99, and 209 from PBDE sources.
Fugacity modeling suggested that approximately 68% of DE and 98% of BDEs 47, 99,
and 209 of the air emission will partition into the soil compartment at equilibrium. There
is suggestive evidence that PBDEs, especially BDE 209, in soils can be degraded by the
microbial reductive debromination as well as the photolytic debromination in soils,
although the viability of these processes in soils is not presently well understood (see
section 3.4.3 for a detailed discussion).
-> -> -)
-------
On a geographical and regional scale, the PBDE concentrations in soils usually
reflect a gradient (high to low) from the central city out to rural areas consistent with
atmospheric measurements (Harrad and Hunter, 2006). Therefore, urban areas are
regional sources of PBDEs to soil via the air to soil exchange. Cetin and Odabasi (2007)
found a strong relationship between the BDE congener profile in soil and air (r2 - 0.13 to
0.79, p < 0.01), thus further supporting the assumption of a close link between air and
soil. From this evidence it is concluded that the atmosphere is a major transport media
for the PBDEs detected in soils.
Some soil studies of PBDEs show a dominance of congeners 47, 99, 100, 153,
and 154 (Hassanin et al., 2004), but other soil data shows that BDE 209 dominants total
soil PBDE concentration (Cetin and Odabasi, 2007), especially in urban areas. BDE 209
also has been shown to dominate the sediment profile. Dominance of BDE 209 reflects
the change from the general global use of the commercial penta- and octa-BDE
formulations to deca-BDE, as well as high sorption and persistence of BDE 209. Further
discussions of PBDEs in surface soils arc found in Section 4.3.
3.6.4. PBDEs in Sewage Treatment Plant Influent, Effluent, and Sludge: A Cause of
BDE Contamination in Surface Waters
Sewage treatment plant (STP) operations are likely a significant source of PBDEs
to surface water leading to local contamination of the freshwater and coastal marine
environments. The STP receives wastewater from homes, businesses, and, in many
cases, industries, which subjects the wastewater to different degrees of treatment before
the treated effluent is discharged into surface waters. Much of the STP sewage sludge
generated by the treatment of wastewater is disposed of on land, which, in turn, can lead
to water pollution through soil erosion into surface waters.
North (2004) sampled influent, effluent, and sewage sludge for the presence of 41
BDE congeners at a STP in Palo Alto, CA. The STP employs tertiary treatment methods
and processes approximately 95E+06 L/day of wastewater generated by residents (60%),
industries (10%), and commercial businesses and institutions (30%). The Palo Alto STP
discharges treated effluent into to the San Francisco Estuary; therefore the STP is likely a
source of local PBDE contamination to the water, sediments, and biota. Results showed
3-33
-------
that more than 90% of the total PBDE concentrations in the STP effluent discharged to
surface waters after tertiary treatment were comprised of BDE congeners 47, 99, 100,
153, and 209. BDE 99 dominated the congener profile. The rank order and mean
concentrations of BDE congeners in the STP effluent were BDE 99 (11,200 pg/L) > BDE
47 (10.467 pg/L) > BDE 100 (1,983 pg/L) > BDE 209 (1.730 pg/L) > BDE 153 (983
pg/L). With respect to the sewage sludge from the STP tertiary process, BDEs 47, 99,
and 209 represented approximately 85% of the total concentration of BDEs detected.
BDE 209 dominated the profile; it represented about 35%) of total BDEs. The rank order
and mean concentrations of BDE congeners in the STP sewage sludge (jag/kg dwt) were
BDE 209 (1,183) > BDE 99 (944) > BDE 47 (757). From a mass balance perspective.
North (2004) estimated that 96%> of the PBDEs that enter the STP are adsorbed to sludge,
and 4% is deposited into surface water with the wastewater effluent. The high percent
adsorption of the BDE congeners to the STP sludge can be attributed to their high log
Kow , which is > 5.0 for all congeners (see Section 3.2.2). It is not known whether BDE
209 in sewage sludge photolytically or microbially degrades to form lower brominated
congeners.
Knoth et al. (2007) investigated the distribution of BDEs 28, 47, 99, 153, 154,
183, and 209 in sewage sludge samples from 11 STPs in Germany. Thirty-nine sewage
sludge samples from different stages of the wastewater treatment process (primary
sludge, secondary excess sludge, and dewatered digested sludge) were collected from
March 2002 to June 2003. BDE 209 dominated the PBDE distribution in all STP
sludges. BDE 209 concentrations in sludges ranged from 97.1 to 2,217 ng/'g dwt with a
mean of 429 ng/'g dwt. The sum of BDE congeners 28, 47, 99, 153, 154, 183 ranged
from 12.5 to 288 ng/'g dwt (mean 126 ng/g dwt). The BDE congener profile remained
rather static from one sludge type to another. Knoth et al. (2007) speculated that this may
provide evidence for the biotransformation of BDE 209 to lower BDE congeners. With
over half the sewage sludge applied to land, Knoth et al. (2007) estimated that 150
kg/acre of pentaBDE plus octaBDE, and 350 kg/acre of decaBDE were applied to land in
2001 from the land farming of contaminated sewage sludge in Germany.
Wang et al. (2007) analyzed sewage sludge samples from 31 STPs in 26 cities in
China for the distribution of BDE congeners 17, 28, 47, 66, 71, 85, 99, 100, 138, 153,
3-34
-------
154, 183. and 209. The concentrations of the sum of all congeners excluding BDE 209
ranged from 6.2 to 57 ng/g dwt, with a mean and median concentration of 19.6 and 16.0
ng/g dwt. respectively. As with the studies discussed above, BDE 209 was the dominant
congener in most of the sludge samples. The percentage of BDE 209 concentration to
total PBDE concentration in Chinese sewage sludges averaged 55%—with a median of
69%. BDE 209 concentrations in sewage sludge ranged from non-detect to 1,109 ng/g
dwt (mean—70.8 ng/g dwt; median—25.5 ng/g dwt). BDE 209 was not detected in four
sludge samples. In more than 80% of the sludge samples, BDE 209 was less than 100
ng/g dwt. Other dominant BDE congeners in the sludges were BDE 47 (mean 24% of
total BDE concentration), BDE 99 (mean 22%), and BDE 183 (mean 13%). Wang et al.
(2007) regressed the data and determined correlations with BDE congener pairs. A
significant and positive correlation (r - 0.814, p < 0.001) was between BDEs-47 and 99.
This association implied significant contamination of the sludge from the pcntaBDE
commercial formulation.
Song et al. (2006) investigated the fate, partitioning, and mass loading of BDEs
throughout the STP process from influent to sewage sludge to effluent. Three sets of
samples were taken from various segments of the sewage treatment process at an STP in
Windsor, Ontario, Canada over 3 days spaced over a 6-week period between the end of
March and early May, 2004. The BDE congeners 28, 47, 71, 99, 100, 138, 153, 154. and
183 were evaluated in the study. BDE 209 was not included in the list of analytes. The
congeners of commercial pentaBDE formulation (BDEs 47, 99, 100, 153, and 154) were
detected in all samples and at all stages of the STP process. On average, approximately
83% of total BDEs detected in the STP were comprised of BDEs 47 and 99. BDEs 47
and 99 were found to be associated with the colloidal suspension of particles and the
dissolved phase of organic matter in the wastewater influent. In the STP process, the
wastewater being treated has a relatively short (12 hours) hydraulic residence time. The
authors speculated that the short residence time combined with the low Henry's Law
constant, and large estimated aqueous degradation half-lives would mean that loss of both
BDEs 47 and 99 by volatilization or degradation during treatment would be negligible.
Further, the high octanol/water partitioning coefficients of BDE 47 and BDE 99 (log Kw
> 5) drives the compounds to partition to the wastewater solids, and BDE 47 and -99 are
3-35
-------
removed from the wastewater with the removal of solids during the treatment process.
This, in turn, enriches the sewage treatment plant sludges with BDEs, but reduces their
loading with the effluent. The mass balance calculations of PBDEs at this 61 million L/d
STP were as follows. Of the 10,560 mg/d total PBDEs entering the STP, approximately
9,609.6 mg/d (91%) ends up in the STP sludges (primary settling and activated sludge),
and 950.4 mg/d (9%) are discharged into the surface water with the treated final effluent.
Table 3-8 summarizes the BDE congener distributions in influent, sludges, and
final effluent from various surveys of sewage treatment plants in various countries.
Generally, North America has higher total BDE concentrations in STP sludge than
Europe, with relatively high concentrations of BDE 209. However, BDE 209 was not
routinely evaluated in European sludge studies.
3.7. BIOACCUMLATION IN BIOTA
3.7.1. Bioaccumulation in the Aquatic Environment
The hydrophobic and lipophilic properties of BDE congeners cause aquatic
organisms to bioaccumulate BDEs with exposures within their food web. up the food
chain s. Despite the recent prevalence in the use of the decaBDE commercial
formulation, which principally contains BDE 209, the lower brominaled BDE congeners
are most prevalent in the tissues of aquatic organisms. As discussed in previous sections,
the BDE congeners enter a watcrbody from atmospheric deposition, air-to-water transfer,
or direct discharge from industries and sewage treatment plants. Once BDEs enter the
aquatic system they partition between the water column and sediments according to their
physical-chemical properties. Moreover, within the water column, BDEs partition
between the dissolved phase and the particle-bound phase. Benthic invertebrates and
mollusks living and feeding directly on contaminated sediments acquire a body burden of
BDEs that is then passed up the food chain from bottom-feeding fish to top piscivorous
fish, raptors, reptiles, and mammals. Mollusks internally absorb BDEs from the water
column through filtering of the water. Fish acquire BDEs through the dietary pathway,
but also from the water passing through their gills. In this regard, trophic level
bioaccumulation in marine and freshwater biota is similar to the pattern of
bioaccumulation of other classes of lipophilic and persistent compounds such as the
3-36
-------
PCBs (Evcnsct et al., 2005; Bragigand ct at, 2006; Eljarrat ct al., 2004), i.e. the
concentration of PBDEs from one trophic level to another seems to biomagnify between
organisms. Bragigand et al. (2006) illustrated this point in the study of trophic level
transfers ofBDE congeners in an aquatic ecosystem in estuaries of France. They found
that the bottom-feeder bivalves and worms had the lowest concentrations of BDE 47,
while the water column feeder eels and soles had the highest concentrations. High
concentrations were also found in flounder and mid-range concentrations in shrimp.
BDEs 47, 66, 99, 100, 153, and 154 have been detected in multiple species of
marine and freshwater organisms. In most cases, BDE 209 has not been included in the
list of analytcs, and, therefore, it is often suggested that BDE 209 cannot be found in
aquatic biota. However, BDE 209 has been detected in hardhead catfish, Atlantic
stingrays, sharp-nosed sharks, and bull sharks off the Florida Coast (Johnson-Restrepo et
al., 2005; in roach fish in a Baltic Sea estuary (Burreau et al., 2004), and in Mysid shrimp
off the Dutch coast (Verslyckea et al., 2005).
For the most part, total PBDEs are highest in top predator marine mammals such
as 6,500 ng/g lipid wt (lwt) in whitebeaked dolphin off the Dutch coast (Soni et al.,
1998). However, water filtering marine mollusks appear to accumulate BDEs at the
highest levels observed in any field survey (e.g., an average total PBDE concentration of
13,502 ng/g lwt in clams from the San Francisco Bay estuary [Ores et al., 2005]).
Another observation that can be made from existing studies is that marine fish have
higher overall body burdens of total PBDEs as compared to freshwater fish. Marine fish
species tend to have total mean PBDE body burdens in a range of < 1.0 to 1,600 ng/g lwt
(Antarctic Rockcod and Florida bull shark, respectively), whereas freshwater fish
typically have body burdens from 1.0 to 300 ng/'g hvt (Detroit River bigmouth buffalo
fish and Lake Michigan lake trout, respectively) (Corsolini et al., 2006; Johnson-
Restretpo et al., 2006; Valters et al., 2005; Streets et al., 2006).
In most fish (ocean and freshwater) and marine mammals, BDE 47 is the major
congener contributing >30% to total body burden of PBDEs, The congener distribution
in tissues of aquatic biota usual ly follow the order BDE 47 > BDE 99 > BDE 100 > BDE
154 > BDE 153 > BDE 49 > BDE 28. In the few studies where BDE 209 was measured
in fish tissue (ocean fish), BDE 209 has ranged from < 1 % up to 88% of total PBDE
3-37
-------
body burden. Because of small sample size, these studies are insufficient to indicate the
general contribution BDE 209 makes to PBDE in fish tissue.
This review of PBDEs detected in fish and marine mammals illustrates the fact
that PBDEs are transported long distances, far away from their original source. PBDEs
are detected in deep ocean water marine fish and mammals and in marine ecosystems
ranging from the equator to the poles. This provides additional evidence for the long-
range transport of PBDEs. Details of specific studies measuring BDE congeners in
freshwater, marine, farmed, and store-bought fish, including a tabular summary of
congener-specific concentrations, are provided in Section 4,6.
3,7.2. Bioaceumulation in the Terrestrial Environment
Field studies of the bioaceumulation of BDEs in terrestrial environment have
included fish-eating mammals, birds of prey, and carnivorous mammals in the wild.
These animals are top predators in the food chain and acquire a body burden of BDEs
from trophic level exposures in a similar manner as aquatic organisms. The following
provides a summary of bioaceumulation of PBDEs in terrestrial animals from trophic
level exposures.
3.7.2.1. Bioaceumulation in Birds
Birds of prey are ideal candidates to study the bioaceumulation of PBDEs because
they are carnivores, typically secondary and tertiary consumers in terrestrial and aquatic
food chains. Naert et al. (2007) studied the body burden distribution of BDEs in brain
and adipose tissue samples of buzzards (Buteo huteo), sparrow hawks (Accipiter nisus),
cormorants (Phalacrocorax carbo sinensis), and blackbirds (Turdus merula) from field
samples collected at different locations in Switzerland between 2003 and 2005. Lower
concentrations of BDE congeners were detected in the brain of avian species as compared
to adipose tissue. The median total PBDE concentrations (sum of BDEs 28, 47, 99, 100,
153, 154, and 183) in brain ranged from below the detection limit in blackbirds to 14 ng/'g
wet weight (wwt) in sparrow hawks. The median concentrations of total PBDEs in
adipose tissue ranged from below the detection limit, in blackbirds, to 709 ng/g wwt in
spaiTow hawks. The authors postulated that the difference in median concentrations
3-38
-------
between the two tissues (brain and adipose) is attributed to the blood-brain barrier
protecting against the accumulation of BDEs in the brain. In this study, the sparrow
hawk had the overall highest median body burden of total PBDEs of all the species:
790.2 ng/g wwt verses the common buzzard {34,55 ng/g wwt), blackbird (0.82 ng/g wwt),
and cormorant (98.76 ng/g wwt). BDE 99 was the dominant congener detected in the
sparrow hawk accounting for approximately 40% of total PBDEs. BDE 47 body burden
was about 26% of total PBDEs. The sparrow hawk is a specialist feeder, and consumes
finches, sparrows, and wood pigeons. The cormorant (having the next highest mean
PBDE body burden) nests in swampy areas or near large bodies of water such as rivers
and lakes. Its diet consists mainly of fish. BDE 47 contributed approximately 42% of the
total body burden of PCDEs in the cormorant. In. the common buzzard, the most
prevalent congeners detected were BDEs 153 (29% of total), 99 (23%), and 47 (22%).
The common buzzard preys mainly on small mammals (mice, voles, rabbits, squirrels,
rats, moles). The blackbird, on the other hand, feeds principally on seeds and insects, and
BDE 47 was the only congener detected in the adipose tissue of this species.
Voorspoels et al. (2006a) investigated the occurrence and distribution of PBDEs
in sparrow hawks, common buzzards, and owls of Belgium; the highest mean
concentration of total PBDEs was measured in li ver of the sparrow hawk (i.e., 4,900 ng/g
Iwt), followed by the liver of the common buzzard (480 ng/g lwt), and owls (250 ng/g
hvt). Congeners can be ordered according to their relative contribution to the total PBDE
content as follows for buzzards: 153 > 47 > 99 > 183 > 100 > 154; for sparrowhawks: 99
> 47 >153 > 100 >183 > 154; for owls 153 > 99 > 47 > 183 >100 > 154. BDE 209 was
present in 6 out of 44 liver samples and 19 out of 25 serum samples of sparrow hawks,
for both common buzzards and owls combined (Voorspoels et al., 2006a). In 2/2 samples
of sparrow hawk liver, BDE 209 was detected at a mean concentration of 17 ng/g lwt;
and 26 ng/g iwt in serum. In buzzards, BDE 209 was detected only at a frequency of
3/29 (10%) in li ver (mean = 79 ng/g lwt). However, in the serum of buzzards the
frequency of detection was a surprising 16 out of 20 samples (80%). Chen et al. (2007)
also confirmed the presence of BDE 209 in liver of 3 buzzards from Beijing, China
(mean = 71 ng/g lwt). BDE 209 was the congener present at the highest concentration—a
surprising 48% of total PBDEs.
3-39
-------
Later, Voorspoels ct al. (2007) again evaluated sparrow hawks, common
buzzards, and passerines for body burdens of BDE congeners. The authors found similar
results: the highest levels of total PBDEs were detected in the liver of sparrow hawks, at
9,506 ng/g Iwt. The liver of the common buzzard had a mean concentration of 727 ng/g
Iwt, which were considerably lower than liver of sparrow hawks. The PBDE
concentrations in adipose tissues and eggs of passerines averaged 160 and 220 ng/g Iwt,
respectively.
Because the passerine is a favorite prey of the sparrow hawk, Voorspoels et al.
(2007) was able to observe the biomagnification potential of BDE congeners in a simple
trophic system. The biomagnifications potential (BMP) is derived as the ratio of lipid
normalized median BDE congener concentrations in the same tissues of both predator
and prey (Voorspoels et al., 2007). A ratio >1 would indicate a potential for
biomagnifications between species. Voorspoels et al. (2007) calculated the mean BMP of
BDE congeners when passed on through the dietary pathway from passerine to sparrow
hawk. The rank of BMP from high to low was BDE 183 = 29; BDE 100 - 25: BDE 154
= 24; BDE 153 - 21; BDE 99 = 20; BDE 47 = 10; and BDE 28 = 4. The mean BMP for
the sum of these BDE congeners was 17. It is apparent that the BMP increases from
lower brominated to higher brominated BDE congeners within this simple terrestrial food
chain. This trend generally tracks the log Kow of the BDE congeners (see Table 3-2). Eor
example, the log Kmv of BDE 183 is approximately 8.3, which is considerably higher than
the log Kowof BDE 47, which is 6.0.
Chen et al. (2007a) reported on the distribution of BDE congeners in birds of prey
(i.e., the common kestrel, the sparrow hawk, the Japanese sparrow hawk, the little owl,
the scops owl, and the common buzzard) that were collected in the vicinity of Beijing,
China. The sum of the mean PBDE congener concentrations was highest in the common
kestrel. Chen et al. (2007a) indicated that these concentrations are among the highest
levels in birds reported in the open literature: muscle = 12,300 ng/g Iwt, liver 12,200
ng/g Iwt, and pooled kidney = 5,340 ng/g Iwt. Kestrels dwell in the Beijing area year
round, establishing nests and foraging in urban fringe or urban centers. This, in
combination with their dietary habits (a preference for small mammals), suggests that the
dietary pathway may have caused the high concentrations through biomagnification.
3-40
-------
Congener profiles of BDE concentrations in bird tissues indicated a dominance of the
more highly brominated congeners. Profiles in the muscle and liver of Chinese sparrow
hawks, for example, showed that BDE 153 > 99 > 47 > 183 > 154 > 209 >207. This
differed from the BDE profile in the tissues of Japanese sparrow hawks—99 > 153 > 47
> 209 > 207 > 183. In the liver of scops owl, long-eared owl, upland buzzard and
common buzzard, the congener profile was dominated by BDE 209 (followed by BDE 99
and 153). Chen et al. ( 2007a) attributed the dominance of the higher brominated
congeners to the heavy use of commercial octa and deca formulations.
She et al (2008) recently reported on the distribution of PBDEs in piscivorous and
omnivorous bird eggs obtained from a San Francisco Bay estuary in California, and from
Gray's Harbor, Washington. A total of 169 eggs were collected from the following
species, combined: Caspian tern (Sterna caspia); Forster's tern (Sterna forsterii);
California Least torn (Sterna antillarum brownie); Clapper rail (Ra/ltis longirostris
obsoletus). All tern species were fish eating birds, and the clapper rail was omnivorous.
The pattern of the five major PBDE congeners detected in the bird eggs (i.e., BDE
47>BDE 99>BDE 10()>BDE>153> BDE 154) were consistent with the BDE congener
patterns detected in fish. BDE 47 was the most abundant congener and represented
approximately 60% of total PBDEs detected in bird eggs. BDE 209 was not detected in
any egg sample. Total PBDEs in bird eggs ranged from about 1,080 to 63,300 ng/g Iwt.
Median concentrations of PBDEs in Caspian tern eggs for 2000-2003 were 2,410, 4,730,
3,720 and 2880 ng/g lwt, respectively, in Forster's terns were 1,820, 4,380, 5,460 and
3,600 ng/g lwt, respectively, and in California Least terns for 2001 and 2002 were 5,060
and 5,170 ng/g lwt, respectively. In contrast, median PBDEs concentration in California
Clapper rail eggs for 2001 was 379 ng/g lwt. The rails consumed mostly invertebrate
species. The authors concluded that the dietary consumption of PBDE contaminated fish
was responsible for the accumulation of PBDE congener in the eggs offish eating terns.
3-41
-------
3.7,2,2. Bioaccumulation in Terrestrial Mammals
Only a very limited number of terrestrial mammals have been studied for their
body burden of PBDEs. Many mammals are top predators and, therefore, would be
expected to bioaccumulate and biomagnify PBDEs from the food web.
Voorspoels et al. (2006b) studied the red fox for PBDEi contamination and food
ehain bioaccumulation. The red fox is a top terrestrial predator that mainly consumes
voles, rabbits, squirrels, and mice as prey. To the extent the prey becomes contaminated
with PBDEs, the fox may be exposed to PBDEs that have biomagnified after moving
through the trophic system. Voorspoels et al. (2006b) sampled tissue for the occurrence
and distribution of PBDEs as a body burden in 33 red foxes indigenous to Belgium and
then examined biomagnification of BDE congeners from prey to predator. The median
sum of PBDE congeners measured in the foxes were considered low, ranging between
2,2 and 3.4 ng/g lipid weight: in adipose tissue, liver, and muscle. BDE 209 was detected
at a frequency of 40% in liver, 21 % in muscle, and 15% in adipose tissue samples. In fox
liver, the BDE 209 congener, on average, constituted 70% of total PBDE concentration—
a finding inconsistent with other top predator species. BDE 153 was the most frequently
detected congener in all tissues (96%-l00% of all samples), whereas BDE 47 was
detected in 33, 40, and 100%) of adipose, liver, and muscle tissues, respectively.
Voorspoels et al. (2006b) did not observe any evidence of metabolic debromination in the
fox. The authors observed that total PBDE concentrations were lower in fox tissues as
compared to voles and mice (the main diet of foxes). Voorspoels et al. (2006b)
speculated that this may be due to the high capacity of the fox to metabolize lower
brominated BDEs in vivo, similar to what has been observed in the grizzly bear
(Christensen, et al, 2005). The authors postulated that this high metabolic activity might
be related to the fact that no biomagnification of BDE congeners between prey and
predator was observed in the fox (Voorspoels et al., 2007). The study of PBDE
distributions in foxes showed that BDE 209 does, however, bioaccumulate in terrestrial
top predators such as the red fox ( Voorspoels et al. 2006b).
Christensen et al, (2005) reported on the influence of diet on the BDE congener
distribution and other persistent organic contaminants in the tissues of 12 grizzly bears (6
3-42
-------
each from coastal and interior areas) in British Columbia, Canada, Dietary consumption
of meat was estimated to be 0 to 19% and 13 to 61% of total diet for interior and coastal-
maritime bears, respectively. Maritime bears mostly consumed salmon as the primary
meat source, whereas the meat source of interior bears was more varied. The remaining
diet of all bears consisted of vegetation. With the exception of PBDEs, the persistent
organic chemicals (POPs) were higher in the tissues of the bears that consumed meat.
However, there was no significant difference in total PBDE tissue burden or congener
distribution between the two groups of bears. Total PBDEs dominated total POP
concentration in interior bears but not in maritime bears. The ranking by higher-to-lower
total contaminant concentration of POPs in the tissues of interior and maritime bears
showed the following pattern respectively: PBDEs >PCBs > HCB > HCH > CHL >
DDT and PCBs > CHL> HCB > DDT > PBDEs > HCH. Both bear groups showed a
marked difference in the mean BDE congener profile. Maritime fish-eating bears
displayed a BDE congener pattern dominated by BDE 47 followed by 209 > 99 > 100 >
153. In the interior bears the tissues were dominated by BDE 209 followed by 206 > 47
> 207 > 208. The predominance of the lighter congener BDE 47 in the maritime bears
suggests that PBDE exposure mainly occurs through the fish consumption pathway. It
was postulated that the fish acquire PBDE contamination as a result of the long-range
atmospheric transport and deposition into the marine food web. The authors speculated
that the dominance of BDE 209 in the interior bears, whose diet was richer in vegetation
than the maritime bears, may indicate the possibility of the PBDE exposures via the
pathway of air-to-plant-to bear. In this paradigm the vegetation accumulates PBDEs
from local source air emissions of decaBDE and passes BDE 209 onto the bear through
its diet. This notion was supported by the observations that BDE 209 is not detectable in
fish, and the higher BDE congeners had negative bioaecumulation slopes in the tissue of
bears indicating a preferential exposure to local sources through their consumption of
terrestrial vegetation (Christensen et al., 2005).
The polar bear diet is chiefly comprised of ringed seal. Wolkers et al. (2004)
studied the cod-seal-polar bear bioaecumulation pattern of PBDEs in an Arctic food web
and found a food chain pattern consistent with what might be expected for the BDE 47
congener: BDE 47 concentrations in polar bear were great than those in seal, which in
-------
turn were greater than those in polar cod. From the 22 brominatcd compounds measured,
only BDEs 47, 85, 99, 100, and 154 were detected in polar cod. In addition to these
congeners, the tissues of ringed seals contained BDE 66, Concentrations of PBDEs in
ringed seals were overall higher than in cod. In particular, BDE 47 and 99 were
approximately one order of magnitude higher in seals than in cod. PBDE 47 comprised
more than 90% of the total PBDEs present in ringed seals. Only BDE 47 was detected in
polar bears. Male and female polar bears were found to have a body burden of BDE 47
that was 1.5 and 3 times the body burden of seals, respecti vely.
3.7.3. Bioaccumulation in Insects
Insects may acquire a body burden of PBDEs through contact with contaminated
environmental media or through the dietary pathway. Only a very limited number of
species of aquatic insects, caddisflies (triehoptera), and midges (diptera), have been
investigated for the occurrence and distribution of PBDE in insect tissues (Bartrons et. al
2007). Most insect species undergo a number of developmental transformations from
larva and pupa to adult, and they begin their life cycle in aquatic sediments. Insects are at
the beginning of the food chain in most aquatic ecosystems. It has been noted that pupae
emerge during periods of high ftsh activity and are an attractive source of food to many
fish species (Bartrons et. al 2007). Therefore, in terms of movement through trophic
systems and biomagnifying from lower to higher predators, insects can be viewed as an
important beginning to contamination of PBDEs and other persistent organic pollutants in
aquatic organisms.
Bartrons et al. (2007) obtained 22 samples of larvae and pupae of four variants of
caddisflies and midges from two, high-altitude mountain lakes in the Pyrenees Mountains
of Spain. Samples of the larvae and pupae of triehoptera belonged to the
polycentropodidae and limnephilidae families, and samples of diptera belonged to the
chironomidae and ceratopogonidae families, thus producing samples from a total of four
distinct insect types. Total concentrations of PBDEs in insect larva ranged from 0.65-
1.68 ng/g dwt and 0-13.07 ng/g dwt for caddisflies and midges, respectively. Total
PBDEs in insect pupa ranged from 5.17-9.32 ng/g dwt in caddisflies and 3.91-27.38 ng/g
dwt in midges. In general, pupae contained significantly higher concentrations of total
3-44
-------
PBDE than larvae of the same taxonomic group. BIDE 209 was detected in the larva and
pupa of the limnephilidae and polycentropodidae families ofcaddisflies but not in the
ceratopogonidac and chironomidae families of midges flies. Caddisfly pupa of the
polycentropodidae family had the highest overall level of BDE 209 (4.93 ng/g dwt).
BDE 47 was the most consistently detected congener in larva and pupa of midges and
caddisflies, with the exception of midge larva of the chironomidae family where it was
not detected.
3.8. ENVIRONMENTAL TIME TRENDS
Sediment cores, archived vegetation and biological tissue samples have been used
to infer time trends of the levels of PBDEs in the environment. The dated sediment cores
and archived vegetation provide a clear record of the varying PBDE concentration over
the decades. Whale blubber, bird eggs, and bird fat have been studied at different time
intervals and can be used to evaluate the time trends of PBDE levels in biological tissues.
The following is a limited review of these studies; further detail can be found in the
studies themselves. These studies all show a similar trend: that PBDEs were absent in
the environment until their introduction as flame retardant products in the 1970s. Their
presence in the environment increased throughout the remainder the 20th century into the
21century.
3.8.1. Time Trends from Sediment Core Studies
Sediment core studies have been used to study 20th century temporal trends in
environmental levels of such contaminants as PCBs, dioxins. and PBDEs. The approach
is to take several cores usually within a lake (which is quiescent unlike moving water
bodies such as rivers), vertically , and date sectionssectioning the cores, and dating the
slices using radiotracers. Once a ctionsection is dated, then a measurement of the
contaminant in that segment provides an indication of relative environmental levels
during that time period. Further, knowing the deposition rate of sediments in that
particular water body, the total loading into that water body can be estimated. This
approach is useful for persistent contaminants that predominantly partition to benthic
sediments. These characteristics are true for persistent and bioaccumulative toxics
3-45
-------
(PBTs) such as PBDEs. This section provides an overview of sediment core studies
undertaken for PBDEs around the world and what has been learned from them.
Li et al. (2006) reported on the chronology of deposition flux of PBDEs into the
sediments of all five Great Lakes (Lake Michigan, Lake Huron, Lake Superior, Lake
Ontario, and Lake Erie) and three inland seepage lakes. Their work was also chronicled
in three literature articles (Song et al., 2005a, b; 2004). Twenty-two sediment cores were
collected in 2001 and 2002 and horizontally sectioned into a total of 247 samples (175
from the Great Lakes and 48 from the inland lakes). Analytical results are separately
reported as the sum of nine PBDE (SgBDEs) congeners (BDE 28, 47, 66, 85, 99,100,
153, 154, and 183) and BDE 209. The range of mean surface sediment concentrations in
all the Great Lakes was 1.4-5.6 ng/g dwt and 10.5-226.6 ng/g dvvt for the L.BDEs and
BDE 209, respectively. BDE 209 dominated the total PBDE concentrations in all the
lake sediments. Figure 3-1 graphically displays the temporal trends of deposition fluxes
of total PBDE and PCBsto the sediments in each of the Great Lakes as determined by Li
et al. (2006). In general, PBDE levels began to rise between 1920 and 1950. There is a
striking increase in the deposition of total PBDE to the lake sediments from 1970's to
2002. The temporal trends in the dated sediment cores from this study suggest that the
increase in PBDE input to the Great Lakes tend to be first order. From these data Li et al.
(2006) calculated concentration doubling times t (_•) ranging from 9 year to 43 years for
I9BDEs and from 7 to > 70 years for BDE: 209. There is no evidence of any recent
decline in PBDE loadings to the Great Lake sediments. Li et al. (2006) also estimated the
annual PBDE loading rates to the surface sediments of the Great Lakes by multiplying
the mean surface 11 ux times the total surface area of the lakes. According to their
calculations, the Great Lakes received approximately 0.17 tons of L,BDI.s and 4.4 tons
of BDE 209 in 2002, primarily from atmospheric deposition. Total deposition of BDE
209 to the Great Lakes was estimated to be over 25 times the deposition of the sum of
BDE 28, 47, 66, 85, 99,100, 153, 154, and 183 combined.
Qiu et al. (2007) evaluated the distribution of flame retardants Dechlorane Plus
(DP), 1,2-bis-(2,4,6-tribromophenoxy)ethane (TBE), and PBDEs in dated sediment layers
in a sediment core study of Lake Ontario. They measured 20 BDE congeners, including
BDE 209. In the sediment surface corresponding to 2004, Qiu et al. (2007) detected £ 3.
3-46
-------
yPBDE congeners (i.e., tri thru hepta BDE congeners 28, 47, 49, 99, 100, 116, 153. 154,
181, and 183) at a concentration of 2.8 ng/g dwt. The concentration of BDE 209 was
approximately 15 ng/g dwt in the surficial layer, which was approximately 5 times higher
than the sum of the other BDE congeners. The authors postulated that the sedimentary
record of PBDEs in Lake Ontario reflected atmospheric deposition as the primary
transport mechanism to the lake. BDE 209 was initially detected in Lake Ontario
sediments around 1980 and underwent a dramatic increase from 1990 to 2000. E 3.7PBDE
congeners were initially detected around 1955, gradually increased in concentration
between 1960 and 1990, and sharply increased in concentration from 1990 through 2000.
In 2004, Zhu and Kites (2005) conducted core studies for 18 BDE congeners from
two study sites; Lakes Michigan and Erie. Like the other Great Lakes studies, BDE 209
predominated, making up 95-99% of the concentration. The concentration of BDE 209 is
315 and 39 ng/g dwt in Lakes Michigan and Erie, respectively, while all other congeners
measured 2.6 and i. 1 ng/g dwt in Lakes Michigan and Erie, respectively. This article
plots the rise in concentration, showing how BDE 209 was about 10 ng/g dwt in 1960 in
Lake Michigan and rose to above 300 ng/g dwt by the early 2000s. The sum of the others
similarly started well below 1.0 ng/g in 1960 to end up at 2.6 by the 2000s. Similar
trends were seen in Lake Erie, with 209 not showing up until 1980 at all at 1 ng/g dwt to
rise to 38 by the 2000s.
Marvin et al. (2007) studied the temporal trends of PBDEs in archived freeze-
dried samples of suspended sediment taken from the water column of the Niagara River
feeding into Lake Ontario. The archive contained samples that had been collected from
1980 to 2002. The Niagara River flows into Lake Ontario, and there has been a history
of heavy industry along its river banks. A total of 16 BDE congeners were evaluated in
the samples. PBDE concentrations in the suspended sediments significantly increased
between the time periods 1980 and 1988. However, the most current samples indicate a
decline in total PBDE concentration in suspended sediments since 2001. This is different
than the sediment core studies above, which show increases—at least until the last dating
time at 2004. This difference could be attributed to the differences in the type of samples
taken (suspended particles verses dated sediment core layers) and the fact that BDE 209
was not specifically analyzed in the samples. Another factor that may contribute to this
3-47
-------
difference is that Marvin et al. (2007) indicated that the primary inputs to the Niagara
River were industrial point sources in the watershed, whereas inputs of PBDEs to the
Great Lake sediments were primarily from atmospheric deposition,
Chen et al. (2007b; with earlier discussions of the partial data set also in Mai et
al., 2005) examined the temporal trends of 3 dated sediment cores obtained from the
Pearl River delta in China. A total of 17 congeners were measured in these cores.
Generally, the total concentrations increased from 1975 to the late 1980's and early
1990's. BDE 209, in particular, displayed an exponential increase in concentration in the
sediments from the time period 1990 to 1995 and the year 2005. Chen et al. (2007b)
surmised that the sediments reflect the dominant use of the pentaBDE formulation prior
to 1990, and a subsequent dominance of the use of decaBDE from 1990 onward.
Deposition flux of BDE congeners not including BDE 209 ranged from 5.8-106.2
ng/cirf, with an average of 56.0 ng/crrf, while the deposition flux of BDE 209 was higher
ranging from 172.8 to 563 ng/enr. Chen et al. (2007b) estimated the total deposition flux
of total congeners, not including BDE 209 and BDE 209, to the Peal River delta in 2005,
to be 2.1 and 29.7 MT. respectively.
Mmh et al. (2007) analyzed the occurrence and distribution of 11 BDE congeners,
including BDE 209, in three dated sediment cores of Tokyo Bay. The investigators
reported results in terms of the sum of BDE 3 through BDE 207 (ZPBDEs) and BDE 209,
separately. There were differences in the chronology of mass concentration of PBDEs
from one sediment core to another. In addition there was a clear concentration gradient
of SPBDEs and BDE 209 in the surface sediments, with the sediments at the mouth of
Tokyo Bay having the highest concentrations. The authors indicated that the apparent
concentration gradient clearly demonstrated that populated areas such as the cities of
Tokyo and Yokohama are major emission sources of PBDEs to the bay. Core dating data
suggested that the SPBDEs concentrations consistently increased beginning in 1945 and
reached a maximum in 1988, with concentrations at about 3 ng/g dwt or less at that
maximum. There appeared to a slight decrease from 1988 to the surface sediments
representative of the year 2000. BDE 209 appeared in the cores in the 1960s, reaching
maximums between 20 and 80 ng/g dwt in 2000, the last year of dat ing. The authors
explained that the apparent decease in the XPBDEs concentration in the mid 1990s may
3-48
-------
reflect Japan's phase-out and reduction in the general use of penta- and oetaBDE
commercial formulations in the early 1990s, and similarly, that the continual rise of BDE
209 reflects ongoing and increasing use of decaBDE.
Stern et al. (2005) reported the results of a sediment core study of a remote
northern Arctic lake on Devon Island in Canada, A number of persistent organic
pollutants, including PBDEs, were investigated in order to observe the temporal trends of
contaminant loadings. Four sediment cores were collected in 1999 and archived in a
freezer. Core #2 was analyzed for the concentration of PBDEs. Maximum deposition of
total PBDEs occurred in the most recent surface sediments with a depositional flux
estimated at 28.5 ng/nr/yr. BDE 47 was the most abundant congener followed by BDE
99 and BDE 100. These three congeners represented about 80% of total PBDEs
determined in the sediment core slices. The authors speculated that this congener pattern
may be indicative of anaerobic microbial decomposition of more highly broroinated
PBDEs in the sediments, although there was no direct proof this had occurred (Stern et
al., 2005). The authors concluded that the contamination of the remote Arctic lake with
PBDEs was the result of the long-range transport and deposition into the lake, because
there were no local sources of these contaminants.
Evenset et al. (2007) undertook a sediment core study to examine the historical
deposition of PBDEs and other persistent organic pollutants into the sediments of Lake
Ellasjoen, a remote lake on an island in the central Barents Sea of the Norwegian Arctic.
Four replicate sediment cores were collected in April, 2001 from a depth of 34 m and
measured for 10 BDE congeners (not including BDE 209). PBDEs could only be
detected in the upper 4 cm of the sediment core, which corresponds to a time range
encompassing the 1940's through 2001. Of the ten BDEs, only BDEs 28, 47, 99, 100,
and 153 were detected. They were first detected in core segments corresponding to 1953,
at levels of 0.1 ng/g dwt and less, and they continuously rise in concentration to 1994,
with the maximum individual congener concentration of 0.45 ng/g dwt. This dominant
congener was BDE 47, followed by BDE 99, 28, and 100. BDE 153 was only found in
the surface sediments to the lake. The primary route of entry of BDE congeners was
assumed to be from atmospheric deposition, thus indicating the long-range transport of
PBDEs (Evens et al., 2007).
3-49
-------
Zegcrs et al. (2003) summarizxd various studies on the chronology of PBDE
concentrations in dated sediment cores collected in Europe, The sediment core samples
were obtained from the Oslofjord river in Norway; from the marine sediments of the
Wadden Sea off the coast of The Netherlands; from Lake Woserin in the state of
Mecklenburg-Vorpommern, Germany; and from the Kim me ridge clay formation in the
United Kingdom, a marine formation from the Jurassic period. Pb-210 and Cs-137 (one
or both) isotopes were used to mark distinct years in the age of the sediment layers. The
findings of all of these studies are similar to all of the previously described core studies;
Zeger at al. (2003) showed rising concentrations from about the 1970s to the present,
with dominance of BDE 209, particularly in the later years. From the Drammenfjord
River in Norway, BDE 47 initially appeared in the sediments in 1975 at 1,2 ng/g dwt and
steadily increased to a level of 5.8 ng/g dwt in 1999. Similarly, BDE 209 also first
appeared in the sediments in 1975, but it increased at a sharper rate than BDE 47. The
surface sediment concentration of BDE 209 is 80-fold higher than in 1975, rising from
1.3 to 105 ng/g dwt. The concentration pattern from high to low suggests that BDE
209>BDE 99>BDE 47>BDE 100. In Lake Woserin in Germany, the oldest sediment
layer was dated to 1628, and the most current was 1997. BDE congeners initially
appeared in 1973, continued to gradually increase, and reached a peak in 1994, with
slight decreases seen for all congeners from 1994 to 1997. BDE 99 was the most
abundant congener in the sediments, and increased 3-fold in concentration from 1973 to
1994 to a high of 11.3 ng/g dwt. Unlike other study results, BDE 209 did not dominate
the profile. Rather, it had concentrations similar to BDEs 99 and 47, reaching a high of
10.7 ng/g dwt in 1994. Clay layers from the Kimmeridge Clay Formation (Blackstone-
Band) in southern England, dated from the Jurassic period 100,000 to 150,000 year ago,
showed no detectable PBDE congeners. BDE congeners initially appeared in 1965 in
cores from the western Wadden Sea off the coast of the Netherlands, although the dating
resolution of decades was not sufficient to determine if PBDEs were present from 1945 to
1965. All detectable BDE congeners increased in concentration from their initial
appearance until 1989. From 1989 to 1995, all congeners slightly decreased in sediment
concentration. BDE 209 was the most abundant congener in the sediments from 1978
through 1995, reaching a high concentration of 380 ng/g dwt. In 1995 BDEs 47 and 99
3-50
-------
reached high concentrations—about 20 ng/g dwt. Concentrations of 12 other congeners
were mostly non-detects (at detection limits of about 1 -3 ng/g dwt).
3.8.2. Time Trends from Aquatic Wildlife Samples
The body burden of PBDEs and other persistent organic pollutants in whales,
seals, and different fish species have been measured over discreet time intervals. This
provides a basis for assessing the changes in relative body burdens in wildlife with the
passage of time. Like sediment core studies, these studies show a rise of BDE congeners
through the 1990s into the 2000s. However, very different than sediment core studies,
these show a predominance of BDE 47 and a virtual absence of BDE 209. The absence
of BDE 209 is also true with human blood and tissue sampling (see Chapter 5).
Lebeuf et al (2004) studied the levels and temporal trends of PBDEs in the
blubber of 54 stranded adult beluga whales from the St. Lawrence Estuary in Quebec,
Canada. The beaching of whales on the shores of the St. Lawrence Estuary had occurred
in the time period 1988 through 1999, and samples were collected during these years and
frozen for future analysis. The total PBDE concentration (including BDEs 28, 47, 71, 77,
99, 100. 153. 154, 155, and 183) ranged from about 20 to almost 1000 ng/g wwt in
samples from 54 whales. The data suggests an exponential increase of PBDE congeners
from 1988 to 1999. The authors estimate the doubling times of BDE congeners to range
between 2 and 9 years for all congeners (Lebeuf et al., 2004).
Kajiwara et al. (2004) used archived fur seal adipose tissue samples to investigate
the relative time-trends of PBDEs in coastal waters of Japan. Ten fat samples had been
collected over a period between 1972 and 1998 and stored in the Environmental
Specimen Bank for Global Monitoring at Ehime University. Kajiwara et al. (2004)
analyzed tissue samples for the presence of PCBs, DDT, and BDEs (3, 15, 28, 47, 99,
153, 154, 183, and 209). The sum of the PBDE congeners ranged in concentration from a
low of 0.33 ng/g Iwt in the yearl972 up to 100 ng/g Iwt in the year 1994. BDE 47 was the
most abundant congener of the total PBDEs in all samples analyzed. No BDE 209 was
detected in the fur seal fat samples despite the fact that the decaBDE formulation
constituted 67% and 100% of total commercial PBDE usage in Japan in 1985 and 2000,
respectively.
3-51
-------
Batterman et al. (2007) reported on the time-trends of PBDEs in archived frozen
tissues samples front rainbow smelt, walleye, and lake trout obtained from the Great
Lakes over the time period 1979-2005. All fish species had been collected from sites in
Lakes Erie, Huron, Michigan, Ontario and Superior every other year as part of a
monitoring program conducted by the U.S. Environmental Protection Agency (U.S.
EPA). Only BDE congeners 47, 99, 100, and 153 were investigated in this study due to
their usual pattern of dominance in fish tissue. The total PBDE concentrations in trout at
each of the lakes increased exponentially and rapidly over the period from 1979-1980 to
the mid- 1990s. Concentrations were always highest for BDE 47, rising to as high as 50 -
> 100 ng/g wwt in the latest samples in all lakes, while concentrations of the other three
congeners were between 5 and 10 times lower than BDE 47, rising to no more than 15
ng/g wwt. Doubling times were calculated for all species and lakes. They ranged from
about 2 to over 20 years. The relatively stable congener pattern in all lakes across all fish
species (i.e., BDE 47 > PBE 100 > PBE 99 > PBE 153) suggests that atmospheric
deposition was the primary reason for the presence of PBDEs in the Great Lakes. The
authors claim that there is no evidence to suggest any declines in PBDE concentrations in
the Great Lakes region.
Rayne et al. (2003b) reported on the time trends of PBDEs in tissues of 41
mountain whitefish and 6 sucker fish in the Columbia River system in southeastern
British Columbia, Canada. Eleven congeners were measured, but not BDEs 183 and 209.
From 1992 to 2000 the XPBDL congeners in whitefish increased 11.8-fold and 6.5-fold in
fish caught near the towns of Genelle and Beaver Creek along the Columbia River,
respectively. The authors calculated a short doubling time of total PBDEs of 1.6 years
between the years 1995 and 2000 in whitefish caught near Genelle. At the confluence of
Beaver Creek and the Columbia River (25 miles further downstream from Genelle and 9
miles downstream of a secondary metal smelting operation), the PBDE concentrations
increased over 6-fold from 1992 to 2000—from 4.5 ng/g wwt in 1992 to 29.2 ng/g wwt
in 2000. In whitefish caught on the Slocan River in an unpopulated pristine area that was
not directly impacted by urban or industrial activities, PBDE concentrations were 0.9
ng/g wwt in 1996, which is about 20-50 times lower than those at other testing locations
3-52
-------
near towns and possible sources of PBDEs. Based on concentration differences in
species, the authors speculated that PBDEs hioaccumulated less in sucker than whitetlsh.
3.9. CONCLUSIONS
This chapter focused on the environmental fate of PBDEs. Sections have
discussed fate properties, movement and transformations in the environment, temporal
trends in the environment, and also evidence for metabolic transformations, specifically
denomination, in animals and humans. The following conclusions are made;
1. PBDE congeners are lipophilic and persistent organic compounds having a strong
propensity for bioaccumulation and hiomagnideations in the aquatic and
terrestrial food webs.
2. PBDE congeners are ubiquitous environmental contaminants and are detected
globally in air, soils, sediments, oceans, and wildlife.
3. The atmospheric transport and surface deposition of PBDEs is the primary means
of distributing PBDEs over long geographical distances.
4. The detection of low molecular weight BDE congeners in Arctic air suggests
long-range atmospheric transport of these contaminants from industrialized
countries.
5. Once released into the air, PBDEs partition between the vapor and particle bound
phases in the atmosphere in accordance with their respective VPs. The BDE
congeners with 1-4 bromines atoms primarily exist in the vapor phase; BDE
congeners with 5-6 bromine atoms more equally partition between the vapor
phase and the particle bound phase, and BDE congeners having >6 bromine atoms
are primarily adsorbed to atmospheric particles.
6. Photochemical reaction with the hydroxy I radical appears to be an insignificant
atmospheric degradation pathway for vapor phase BDE congeners.
7. High molecular weight BDE congeners exhibit a propensity for the breakdown
and decomposition both in soils and air by UV light (i.e., photolysis). Photo-
degradation in air and soils can debrominate the higher BDE congeners to form
-------
lower brominatcd BDE congeners, and may be a significant degradation pathway
in the environment.
8. Ocean currents also play a significant role in globally distributing PBDE
contamination in aquatic food webs and in terrestrial food webs connected to the
aquatic environment.
9. Higher brominatcd PBDE congeners can undergo metabolic debromination in
fish, mammals, and birds to form lower brominated congeners.
10. PBDEs can undergo reductive debromination by anaerobic microorganisms.
11. Soils and sediments are environmental sinks for PBDEs.
12. There exists good evidence that microbial anaerobic degradation naturally occurs
in sediments and soils. However, the degradation rate has yet to be determined.
Also, there is evidence of photolysis of PBDE in soils and suspended sediments,
although again degradation rates arc unavailable.
13. Although BDE 209 dominates in soils and sediments in North America, BDE 47
usually dominates in fish tissues. BDE 209 has a Sow potential for
bioaccumulation, and once absorbed, may be easily excreted. In the human and
other animals (see Chapter 5). BDE 209 has a half-life on the order of weeks
whereas other BDE congeners, including BDE 47, have half-lives on the order of
years.
14. Since the 1970s, sediment core samples show predominance and a starker rise of
BDE 209, while the animal tissue samples show predominance and rise in BDE
47.
REFERENCES
ATSDR (Agency for Toxic Substances and Disease Registry) (2004). Toxicological
profile for Polybrominated biphenyls and Polybrominated diphenyl ethers. U.S.
Department of Health and Human Services, Atlanta, GA. September.
Bartrons, M.; Grimalt. J. O.; Catalan, J. (2007). Concentration changes of organochlorine
compounds and polybromodiphenyl ethers during metamorphosis of aquatic insects.
Environ. Sci. Technol.41( 17): 6137-6141.
3-54
-------
Battcrmana, S; Chcrnyaka, S; Gwynna, E; Cantonwinea, D; Jiaa, C; Begnochcb, L;
Hiekey, J.P. (2007). Trends of brominated diphenyl ethers in fresh and archived Great
Lakes fish (1979-2005). Chcmosphere 69 (3): 444-457.
Bezarcs-Cruz, J; Jafvert, C.T: Hua, I. (2004). Solar photodccomposition of
dccabromodiphcnyl ether: Products and quantum yield. Environ. Sci. Technol.38:4149-
4156.
Boon, J.P: Lewis, W.E; Tjocn-A-Choy, M.R; Allchin, C.R; Law, R. J; de Boer, J; ten
HaUers-Tjabbes, C.C; Zegers, B.N (2002). Levels of polybrominatcd diphenyl ether
(PBDE) flame retardants in animals representing different trophic levels of the North Sea
food web. Environ. Sci. Teehnol. 36; 4025-4032.
Bragigand, V; Amiard-Triquet, C; Parlier, E; Boury, P; Marchand, P; El Hourch, M.
(2006). Influence of biological and ecological factors on the bioac-cumulation of
polybrominated diphenyl ethers in aquatic food webs from French estuaries. Science of
the Total Environment 368: 615-626.
Burreau, S: Zcbiihr, Y; Broman, D; Tshaq, R (2004). Biomagnification of polychlorinated
biphenyls (PCBs) and polybrominated diphenyl ethers (PBDEs) studied in pike (Esox
lucius), perch (Percafluviatilis) and roach (Rutilus rutilus) from the Baltic Sea.
Chemosphere 55: 1043-1052.
Buser, H.R. (1987). Polybrominated dibenzofurans and dibenzo-p -dioxins: Thermal
reaction products of polybrominated diphenyl ether flame retardants. Environ. Sci.
Teehnol 20: 404-408.
C "clin. 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.
(."eiin. B and Odabasi, M. (2007). Particle-phase dry deposition and air-soil gas-exchange
of polybrominated diphenyl ethers I PBDEs i in Izmir, Turkey. Environ. Sci. Teehnol. 41
(14): 4986-4992.
Chen, D; Mai, B; Song, J; Sun, Q; Luo, Y; Luo, X; Zeng, E.Y; Hale, R.C (2007a).
Polybrominated diphenyl ethers in birds of prey from Northern China. Environ. Sci.
Teehnol. 41 (6): 1828-1833.
Chen, S.-J.; Luo, X.-J.; Lin, Z.; Luo, Y.; Li, K.-C.; Peng, X.-Z.; Mai, B.-X.: Ran, Y.;
Zeng, E. Y.(2007b) Time trends of polybrominatcd diphenyl ethers in sediment cores
from the Pearl River Estuary, South China. Environ. Sci. Teehnol. 41 (16):5595 -5600.
Chen, Lai-Guo; Mai, Bi-Xian; Bi, Xin-Hui; Chen, She-Jun; Wang, Xin-Ming; Ran, Y;
Luo, Xiao-Jun; Sheng Guo-Ying;, Fu, Jia-Mo; Zeng, E.Y. (2006). Concentration levels,
compositional profiles, and gas-partiele partitioning of polybrominated diphenyl ethers in
the atmosphere of an urban city in south China. Environ. Sci. Teehnol. (40): 1190-1196.
3-55
-------
Christenscn, J. R,; MacDuffee, M.; MacDoanld R. W.; Whiticar, M.; Ross, P. (2005).
Persistent organic pollutants in British Columbia grizzly bears: Consequence of divergent
diets. Environ. Sci. Technol 39: 6952-6960.
Corsolini. S; Covaci, A; Ademollo, N; Focardi, S; Schepens, P (2006). Occurrence of
organochlorine pesticides (OCPs) and their enantiomeric signatures, and concentrations
of polyhrominated diphenyl ethers (PBDEs) in the Adelie penguin food web, Antarctica.
Environ. Pollution 140:371-382.
Dodder, N.G; Strandberg, B; Elites, R.A.(2002). Concentrations and spatial variations of
polyhrominated diphenyl ethers and several organochlorine compounds in fishes from the
northeastern United States. Environ. Sci. Technol. 36:146-151.
Dye, i. A.; Venier, ML; Zhu, L.; Ward, C. R.; Hites, R. A.; Birnbaum, L. S. (2007)
Elevated PBDE levels in pet cats: Sentinels for humans? Environ. Sci. Technol.41(18):
6350-6356.
Eljarrat, E; de la Cal, A; Raldua. D; Duran. C; Barcelo, D. (2004). Occurrence and
bioavailability of polyhrominated diphenyl ethers and hexabromocyclododecane in
sediment and fish from the Cinca River, a tributary of the Ebro River (Spain). Environ.
Sci. Technol. 38: 2603-2608.
Environment Canada (2006) Ecological Screening Assessment Report on
Polyhrominated Diphenyl Ethers ( PBDEs). June.
European Union (2003). Risk assessment report: diphenyl ether, oetabromo derivative
(oetabromodiphenyl ether). European Chemicals Bureau, Luxembourg,
European Union (2002). Risk assessment report: bis(pentabromodiphenyl) ether).
European Chemicals Bureau, Luxenbourg, Belgium.
European Union (2001). Risk assessment report: diphenyl ether, pentabromo derivative
(pentabromodiphenyl ether). European Chemicals Bureau, Luxenbourg. ISBN 92-894-
0479-5.
Evenset, A; Christensena, G. N,; Kallenbom, R (2005). Selected chlorobornanes,
polychlorinated naphthalenes and brominated flame retardants in Bjornoya (Bear Island)
freshwater biota. Environ Pollution 136: 419-430.
Evenset, A; Christenscn, G.N; Carroll, J; Zaborska, A; Berger, U; Herzke, D; Gregor, D.
(2007) Historical trends in persistent organic pollutants and metals recorded in sediment
from Lake Ellasjoen, Bjornoya, Norwegian Arctic. Environ. Pollut. 146:196-205.
3-56
-------
Fang, L; Huang, J; Yu, G; Wang. L (2008) Photochemical degradation of six
polybrominated diphenyl ether congeners under ultraviolet irradiation in hexane.
Chcmosphere 71: 258-267.
Fennell, D. E.; Nijenhuis, I.; Wilson, S. F.; Zinder, S. H.;Ha"ggbIom, M. M. (2004)
Dehalococcoicks ethcnogene.s strain 195 reductively dechlorinates diverse chlorinated
aromatic pollutants. Environ. Sci. Technol., 38, 2075-2081.
Gerecke, AC; Hart maim, PC; Hecb, NV; Kohler, HP E.; Giger, W; Schmid, P; Zcnncgg,
M; Kohler, V! (2005) Environ. Sci. Technol. 39(4): 1078-1083.
Gobas, F. and Morrison, H.A. (2000). Bioconcentration and biomagnifications in the
aquatic environment. In: Handbook of Property Estimation Methods for Chemicals,
Hoc-tilling, R.S. and Mackay, D., editors. Lewis Publishers, NY, NY.
Gouin, T; Haraer, T; Daly, G.L; Wania, F; Mackay, D; Jones, K.C. (2005) Variability
of concentrations of polybrominated diphenyl ethers and polychlorinated biphenyls in
air:Implications for monitoring, modeling and control. Atmos. Environ. 39: 151-166.
Gouin, T; Thomas, GO; Chaemfa, C; Harner, T ; Mackay, D; Jones, KC (2006)
Concentrations of decabromodiphenyl ether in air from Southern Ontario: Implications
for particle-bound transport. Chcmosphere 64: 256-261.
Gustalsson, K.; Bjork, M.; Burreau,S.; Gilek, M (1999) Bioaccumulation kinetics of
brominated flame retardants (Polybrominated diphenyl ethers) in blue mussels (Mytilus
edulis). Environ. Toxicol. Chem (18): 1218-1224.
Harner, T; Shoeib, M. (2002) Measurements of Octanol-Air Partition Coefficients (KOA)
for polybrominated diphenyl ethers (PBDEs): Predicting partitioning in the environment.
J. Chem. Eng. Data (47): 228-232.
Harrad, S. and Hunter, S. ( 2006). Concentrations of polybrominated diphenyl ethers in air
and soil on a rural-urban transect across a major UK conurbation. Environ. Sci, Technol.
40(15): 4548-4553.
Hassanin, A.; Breivik, K.; Meijer, S. N.; Steinnes, E,; Thomas, G. O.; Jones, K, C.
(2004). PBDEs in Esuropean background soils: levels and factors controlling their
distribution. Environ. Sci. Technol. 38(3):738-745.
He, J.; Robrock, K. R.; Alvarez-Cohen, L. (2006) microbial reductive debromination of
polybrominated diphenyl ethers (PBDEs). Environ. Sci. Technol.40(14); 4429-4434.
Horstmann, M; Mclachlan, MS (1998) Atmospheric deposition of semivolatile organic
compounds to two forest canopies. Atmosph Environ 32 (10): 1799"11809.
3-57
-------
Huwe, J. K.: Smith, D. J. (2007). Accumulation, whole-body depletion, and
debromination of decabromodiphenyl ether in male Sprague-Dawley rats following
dietary exposure. Environ. Sci. Technol. 41(7): 2371-2377.
IUPAC (International Union of Pure And Applied Chemistry) (1996) Glossary of terms
used in photochemistry (IUPAC Recommendations 1996): Pure and Applied Chemistry
68, (12): 2223-2286.
Jaward, F.M; Barber, J.L; Booij, K; Dachs, J; Lohmann, R; Jones, K.C. (2004) Evidence
for dynamic air-water coupling and cycling of persistent organic pollutants over the open
Atlantic ocean. Environ. Sci. Technol., 38: 2617 -2625.
Johnson-Restrepo, B; Kannan, K; Addink, R; Adams, DH (2005) Polybrominated
diphenyl ethers and polychlorinated biphenyls in a marine food web of coastal Florida.
Environ. Sci. Technol. 39: 8243-8250.
Kajiwara, N; Noma, Y; Takigami, H (2007) Photolytic debromination of
decabromodiphenyl ether (decaBDE) and ethane (DBDPE) in flame-retarded plastics.
Presented at Dioxin 2007 International Symposium, Tokyo, Japan, September 2-7.
Published in proceedings: Organohalogen Compounds 69: 924™<927.
Kajiwara. N; Ueno, D; Takahashi, A; Baba, N; Tanabe, S.(2004). Polybrominated
diphenyl ethers and organochlorines in archived northern fur seal samples from the
paci fic coast of Japan, 1972-1998. Environ. Sci. Technol. 38 (14): 3804 -3809.
Kierkegaard, A; Balk, L; Tjamlund, U; de Wit, C A.; Jansson, B (1999) Dietary uptake
and biological effects of decabromodiphenyl ether in rainbow trout (Oncorhynchus
mykiss). Sci. Technol.,33:1612-1617.
Kim, Y-M; Nam, I-H; Murugcsan, K; Schmidt, S; Crowley, D.E; Chang, Y-S (2007)
Biodegradation of diphenyl ether and transformation of selected brominated congeners by
Sphingomonas sp. PH-07. Appl Microbiol Biotcchnol 77:187-194.
Knoth, W; Mann,W; Meyer, R; Nebhuth, J.(2007). Polybrominated diphenyl ether in
sewage sludge in Germany. Chemosphere 67: 1831-1837.
Krajmalnik-Brown, R; Holscher, T; Thomson, I.N.; Saunders, F.; Ritalahti. KM; Loftier,
FE (2004) Genetic identification of a putative vinyl chloride reductase in
Dehalococcoides sp. strain BAVl. Appl Environ Microbiol. 70(10): 6347-6351.
Kuramochi, H; Maeda, K; Kawamoto, K (2004). Physicochemical properties of selected
polybrominated diphenylethers and comparison with some brominated aromatics and
PCDDs. Organohalogen Compounds 66: 2420-2425.
3-58
-------
Lebeuf. M.; Goutcux, B.; Measures, L.; Trotticr, S.(2004).Levels and temporal trends
(1988-1999) of polybrominated diphenyl ethers in beluga whales (delphinapterus leueas)
from the St. Lawrence Estuary, Canada Environ. Sci. Technol. 38(11):2971 -2977.
Li, A; Rocknc, K.J; Sturchio, N; Mills. W.J; Song, W; J.Ford; Buckley. D.R. (2006).
Chronology of PBDE Air Deposition In The Great Lakes From Sedimentary Records.
Submitted to Great Lake Atmospheric Deposition Program, Office Air and Radiation
Division, USEPA Region V, Chicago, IL.
Luijk ,R; Govers , HJ; Nelissen, L (1992) Formation of polybrominated dibenzofurans
during extrusion of high-impact polystyrene/dccabromodiphenylether/
Antymony (111) oxide. Environ Sci Technol 26; 219 -2198.
Mai, B; Chen S;, Luo, X; Chen, L; Yang, Q; Shcng, G; Peng, P; Fu, J; Zeng, E.Y.(2005)
Distribution of polybrominated diphenyl ethers in sediments of the Pearl River delta and
adjacent South China Sea. Environ. Sci. Technol. 39 (10); 3521 -3527.
MacLeod, M. (2003 ). On the influence of forests on the overall fate of semivolatile
organic contaminants Environmental Energy Technologies Division, Lawrence Berkeley
National Laboratory, Berkeley, CA. LBNL - 52156.
Marsh, G; Hu, J; Jakobsson, E; Rahm, S; Bergman, A (1999). Synthesis and
characterization of 32 polybrominated diphenyl ethers. Environ. Sci. Technol.33; 3033-
3037.
Marvin, C; Williams. D; Kuntz, K. Klawunn, P; Backus, S; Kolic, T; Lucaciu, C;
MacPherson, K; Reiner, E. (2007) Temporal trends in polychlorinated dibenzo-p-dioxins
and dibenzofurans, dioxin-like PCBs, and polybrominated diphenyl ethers in Niagara
river suspended sediments. Chemosphere 67:1808-1815.
Minh, N.H; Isobe, T; Ueno, D; Matsumoto, K;Mine, M; Kajiwara, N; Takahashi, S;
Tanabe, S.(2007). Spatial distribution and vertical profile of polybrominated diphenyl
ethers and hexabromocyclododecanes in sediment core from Tokyo Bay, Japan. Environ.
Pollut. 148:409-417.
Naert, C; Van Peteghem, C; J; Kupper, J: Naegeli, L.J.H. (2007). Distribution of
polychlorinated biphenyls and polybrominated diphenyl ethers in birds of prey from
Switzerland. Chemosphere 68: 977-987.
NAS (National Academy of Science) (2001). Classifying Drinking Water Contaminants
for Regulator}' Consideration. Committee on Drinking Water Contaminants, National
Research Council. National Academy Press, Washington DC.ISBN: 0-309-07408-8.
NAS (National Academy of Science) (1993). In situ bioremediation: When does it work?
Commission on Engineering and Technical Systems, National Research Council.
National Academy Press. Washington DC.
3-59
-------
North, K.D.(2004). Tracking polybrominated diphenyl ether releases in a wastewater
treatment plant effluent, Palo Alto, California. Environ. Sci. Technol.38 (17): 4484 -
4488.
Oberg, K; Warman, K; Obcrg, T.(2002). Distribution and levels of brominated flame
retardants in sewage sludge. Chemosphere 48: 805-809.
Olsen, E and Nielsen, F.( —001). Predicting vapour pressures of organic compounds from
their chemical structure for classification according to the VOC Directive and Risk
Assessment in General. Molecules 6:370-389.
Oros, D. R; 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.
Paasivirta, J; Sinkkonen, S; Mikkelson, P; Rantio, T; Wania, F (1999). Estimation of
vapor pressures, solubilities and Henry's law constants of selected persistent organic
pollutants as functions of temperature. Chemosphere, 39 (5): 811-832.
Palma, A; Cousins, I.T; Mackay, D; Tysklinda, M: Metcalfeb, C; Alaeec, M (2002)
Assessing the environmental fate of chemicals of emerging concern: A case study of the
polybrominated diphenyl ethers. Envron Pollut (117): 195-213.
Pirard, C; Pauw, E.D. (2007). Absorption, disposition and excretion of polybrominated
diphenyl ethers (PBDEs) in chicken. Chemosphere 66(2): 320-325.
Qiu, X; Marvin, C.H; Hites, R.A.(2007). Dechlorane Plus and other flame retardants in a
sediment core from Lake Ontario. Environ. Sci. Technol. 41: 6014-6019.
Ramu, K; Kajiwara, N; Lam, P.K.S; Jefferson, T.A; Zhou, K.; Tanabe, S. (2006).
Temporal variation and biomagnification of organohalogen compounds in finless
porpoises (Neophocaena phocaenoides) from the South China Sea. Environmental
Pollution 144(2): 516-523.
Rayne, S; Wan,P; lkonomou, M (2006) Photochemistry of a major commercial
polybrominated diphenyl ether flame retardant congcner:2,2',4,4',5,5'hexabromodiphenyl
ether (BDE153). Environ. International 32:575-585.
Rayne,S: Ikonomoub, MG;, Whale, MD (2003a). Anaerobic microbial and
photochemical degradation of 4, 4'-dibromodiphenyl ether. Water Research 37:551-56.
Rayne, S; lkonomou. M.G; Antcliffe, B (2003b) Rapidly increasing polybrominated
diphenyl ether concentrations in the Columbia River System from 1992 to 2000. Environ.
Sci. Technol. 37(13):2847-2854.
3-60
-------
Ritter, L; Solomon, K.R; Forget. J.( 1995). A review of selected persistent organic
pollutants. International Program on Chemical Safety (IPCS). World Health
Organization. December. Report number PCS/95.39.1
Rupp, S; Metzger, JW (2005) B ro m i n a I ed- c h I or i n a ted diphenyl ethers formed
by thermolysis of polybrominated diphenyl ethers at low temperatures. Chemosphere 60:
1644-1651.
Sanchez-Prado, L; Gonzalez-Barreiro, C; Lores, M; Llompart, M; Garcia-Jares, C; Cela,
R. (2005). Photochemical studies of a polybrominated diphenyl ethers (PBDES) technical
mixture by solid phase microextraction (SPME). Chemosphere 60:922-928.
Seshadri, S, et al. (2005) Genome sequence of the PCE-dechlorinating bacterium
Dehalococcoicies ethenogenes. Science 307(5706): 105-108.
She, J; Holden, A; Adelsbach, TL; Tanner, M; Schvvarzbach, SE; Yce, JL; Hooper, K
(2008). Concentrations and time trends of polybrominated diphenyl ethers (PBDEs) and
polychlorinated biphenyls (PCBs) in aquatic bird eggs from San Francisco Bay. CA
2000-2003. Chemosphere, In Press, Corrected Proof, Available online 7 May 2008.
Soderstrom, G; Sellstrom, U; de Wit, CA; Tysklind, M (2004) Photolytic debromination
of decabromodiphenyl ether (BDE 209). Environ. Sci. Technol 38: 127132.
Song, M; Chu, S; Letcher. R.J; Seth, R. (2006) Fate, Partitioning, and mass loading of
polybrominated diphenyl ethers (PBDEs) during the treatment processing of municipal
sewage. Environ. Sci. Technol. 40 (20): 6241 -6246.
Song W, A Li, JC Ford, NC Sturchio, KJ Rockne. DR Buckley, and WJ Mills. 2005a.
Polybrominated diphenyl ethers in the sediments of the Great Lakes. 2. Lake Michigan
and Huron. Env Sci and Tech 39: 3474-3479
Song W, JC Ford, A Li, NC Sturchio, DJ Rockne, DR Buckley and WJ Mills. 2005b.
Polybrominated diphenyl ethers in the sediments of the Great Lakes 3. Lakes Ontario and
Erie. Env Sci and Tech 39: 5600-5605.
Song W, JC Ford, A Li. WJ Mills, DR Buckley, and KJ Rockne. 2004. Polybrominated
diphenyl ethers in the sediments of Great Lakes. 1. Lake Superior. Env Sci and Tech
38: 3268-3293
Soni, B.G; Philp, A.R; Foster. R.G; Knox, B.E. (1998). Do flame retardants
threaten ocean life? Nature 394:28-29.
Stapleton, HM; Baker, JE (2003). Comparing polybrominated diphenyl ether and
polychlorinated biphenyl bioaccumulation in a food web in Grand Traverse Bay, Lake
Michigan. Arch. Environ. Contam. Toxicol. 2:227-234.
3-61
-------
Staplcton, HM; Letcher, RJ, Baker, JE (2004). Denomination of polybrominated
diphenyl ether congeners BDE 99 and BDE 183 in the intestinal tract of the common carp
(Cyprinus carpio). Environ. Sci. Technol. 38: 1054-1061.
Stapleton, H. M.; Brazil, B.; Holbrook, R. D.; Mitchelmore, C. L.; Benedict, R.;
Konstantinov, A.; Potter, D. (2006a). In vivo and in vitro denomination of
decabromodiphenyl ether (BDE 209) by juvenile rainbow trout and common carp.
Environ. Sci. Technol; 40{ 15); 4653-4658.
Stapleton, H.M.; Dodder, N.G; Kuckhck, J.R; Reddy, C.M; Schantz, M.M;
Becker, P.R; Gulland , F; Porter, B.J; Wise, S.A.(2006b). Determination of HBCD,
PBDEs and MeO-BDEs in California sea lions (Zalophus californianus) stranded
between 1993 and 2003. Marine Pollution Bulletin 52:522-531.
Stapleton, HM. and Dodder, NG (2008) Photodegradation of decabromodiphenyl ether in
house dust by natural sunlight. Environ Toxicol and Chem 27 (2): 306-312.
Stern, G.A.; Braekevelt, E; Helm, P.A; Bidleman, T.F; Oiitridgec, P.M; Lockhart,W.L;
McNeeley, R; Rosenberg, B; Ikonomou, M.G; Hamilton, P; Tomy, G.T; Wilkinson, P.
(2005) Modern and historical fluxes of halogenated organic contaminants to a lake in the
Canadian arctic, as determined from annually laminated sediment cores. Science of the
Total Environment 342: 223- 243.
Strandberg, B; Dodder, N.G; Basu, I; Hites, R.A. (2001). Concentrations and spatial
variations of polybrominated diphenyl ethers and other organohalogen compounds in
Great Lakes air. Environ. Sci. Technol. (35): 1078-1083.
Streets, S.S ; Henderson, S.A; Stoner, A.D ; Carlson, D.L; Simcik, M.F; Swackhamer,
D.L. (2006). Partitioning and bioaccumulation of PBDEs and PCBs in Lake Michigan.
Environ. Sci. Technol. (40): 7263-7269.
Su, Y.; Wania, F.; FTarner, T.: Lei, Y. D (2007) Deposition of polybrominated diphenyl
ethers, polychlorinated biphenyls, and polycyclic aromatic hydrocarbons to a boreal
deciduous forest. Environ. Sci. Technol. 41(2): 534-540.
Su, Y; Lei,Y.D; Wania, F; Shoeib, M; Harner, T (2006) Regressing Gas/Particle
partitioning data for polycyclic aromatic hydrocarbons. Environ. Sci. Technol. 40
(11):3558 -3564.
Tittlemier, S.A; Halldorson, T; Stern, G.A; Tomy, G.T (2002). Vapor pressures, aqueous
solubility, and Henry's law constants of some brominated fire retardants. Environ Toxicol
and Chem 21, #9: 1804-1810.
Tomy, GT; Palace, VP; Halldorson, T; Braekevelt, E; Danell, R;Wautier,K; Evans, B;
Brinkworth,L; Fisk, AT (2004). Bioaccumulation, biotransformation, and biochemical
3-62
-------
effects of brominated diphenyl ethers in juvenile lake trout (Salvelimts namayeush).
Environ. Sci. Teehnol 38: 1496 -1504.
Ueno, D; Kajiwara, N; Tanaka, H; Subramanian, A; Fillmann, G; Lam, P.K.S; Zheng,
G.J; Muchitar, M; Razak, H; Prudente, M; Chung, Kyu-Hyuck; Tanabe, S (2004).
Global pollution monitoring of polybrominated diphenyl ethers using Skipjack tuna as a
bioindicator. Environ. Sci. Teehnol. 38: 2312-2316.
VaIters. K; Li, H; Alaee, M; D'Sa, I; Marsh, G; Bergman, A; Letcher, R.J (2005).
Polybrominated diphenyl ethers and hydroxylated and methoxylated brominated and
chlorinated analogues in the plasma of fish from the Detroit River. Environ. Sci.
Teehnol., 39 (15)15612 -5619.
Van den Stecn, E; Covaci, A; Jaspers, V.L.B; Dauwc, T; Voorspoels, S; Ecns. M;
Pinxten, R. (2007). Accumulation, tissue-specific distribution and debromination of
decabromodiphenyl ether (BDE 209) in European starlings (Sturnus vulgaris).
Environmental Pollution, In Press. Available on-line January 16, 2007 at:
www.scicnccdirect.com.
Verslyckea, T.A; Vethaakb, A.D; Arijsa, K; Janssen, C.R. (2005). Flame retardants,
surfactants and organotins in sediment and mysid shrimp of the Scheldt estuary (The
Netherlands). Environmental Pollution 136:19-31.
Voorspoels, S; Covaci, A. Lepom, P; Jaspers, V.L.B; Schepens, P. (2006a). Levels and
distribution of polybrominated diphenyl ethers in various tissues of birds of prey.
Environmental Pollution 144: 218-227.
Voorspoels, S; Covaci, A; Lepom, P; Escutenaire, S; Schepens, P.(2006b) Remarkable
findings concerning PBDEs in the terrestrial top-predator red fox (Vulpes vulpes).
Environ. Sci. Teehnol. 40(9):293 7-2943.
Voorspoels, S; Covaci, A; Jaspers, V.L.B; Ncels, H; Schepens, P. (2007)
Biomagnification of PBDEs in three small terrestrial food chains. Environ. Sci. Teehnol.
41 (2),'411—416.
Wang. X-M; Ding, X; Mai, B; Xie. Z; Xiang, C; Sun. L; Sheng, G ; Fu, J; Zeng. EY
(2005) Polybrominated diphenyl ethers in airborne particulates collected during a
research expedition from the Bohai Sea to the Arctic. Environ. Sci. Teehnol. 39 (20):
7803 -7809.
Wania, F; Dugani. C.B (2003) Assessing the long-range transport potential of
polybrominated diphenyl ethers: a comparison of four multimedia models. Environ
Toxicol and Chem 22:1252-1261.
3-63
-------
Wania, F.; Lei, Y.D.; Harner, T, (2002) Estimating octanol-air partition coefficients of
nonpolar semivolatile organic compounds from gas chromatographic retention times.
Anal. ('hem. 74 (14): 3476 - 3483.
Wang, Y; Zhang, Q Lv, J; Li, A; Liu, H; Li, G; Jiang. G (2007) Polybrominated diphenyl
ethers and organochlorinc pesticides in sewage sludge of wastewater treatment plants in
China. Chemosphere 68 (9): 1683-1691.
Webber, R and Kuch, B. (2003) Relevance of BFRs and thermal conditions on the
formation pathways of brominated and brominated - chlorinated dibenzodioxinsand
dibenzofurans. Environ International 29:699-710.
Whitby, K.T. (1978). The physical properties of sulfur aerosols. Atmospheric
Environment 12: 135-159.
Wolkers. H.; van Bavel, B.; Derocher, A. E.; Wiig, O.; Kovacs, K. M.; Lydcrsen, C.;
Lindstrom, G. (2004). Congener-specific accumulation and food chain transfer of
polybrominated diphenyl ethers in two arctic food chains. Environ. Sci. Technol. 38(6):
1667-1674.
Wong, A; Lei, Y. D; Alaee, M; Wania, F (2001) Vapor pressures of the polybrominated
diphenyl ethers. J. Chem. Eng. Data (46): 239-242.
Wu, Q.; Watts, J. E. M.; Sowers, K. R.; May, H. D. (2002) Identification
of a bacterium that specifically catalyzes the reductive dechlorination
of polychlorinated biphenyls with doubly flanked chlorines. Appl. Environ. Microbiol.
68:807-812.
Wuii, O; Potter, J.R; C.D; Obbard, J.P.(2006a). Polybrominated diphenyl ethers (PBDFV)
over the open Indian Ocean. Atmos. Environ. 40: 5558-5565.
Wurl,0; Kwan Sing Lam, P; Obbard, J.P. (2006b) Occurrence and distribution of
polybrominated diphenyl ethers (PBDEs) in the dissolved and suspended phases
of the sea-surface microlayer and sea water in Flong Kong, China. Chemosphere 65:
1660-1666.
WWF (World Wildlife Fund) (2005). Stockholm Convention: "New POPs": Screening
additional POPs candidates. April.
Zegcrs, B.N., Lewis, W.A., Booij, K., Smittcnberg, R.H., Boer, W,, de Boer, J., Boon,
J.P.(2003). Levels of polybrominated diphenyl ether flame retardants in sediment cores
from Western Europe. Environ. Sci. Technol. 37: 3803-3807.
Zhu LY, RA Hites. 2005. Brominated flame retardants in sediment cores from Lakes
Michigan and Erie. Env Sci and Tech 39: 3488-3494.
3-64
-------
Table 3-1. Estimated water solubility values for PBDEs.
Brominated diphenyl
ether
Aqueous Solubility
(mg/L (al 250 C)
Reference
Solubility score
Deca-BDE
(Commercial)
< .001
European Union, 2002
Low
Oeta-BDE
(Commercial)
0.005
0.002
European Union (2003)
ATSDR (2004)
Low
Penta-BDE
(Commercial)
0.013
European Union (2001)
Low
BDE 11
0.088
Palma et al. (2002)
Low
BDE 15
0.13
ATSDR (2004)
Low/Medium
Borderline
BDE 17
0.026
Palma et al. (2002)
Low-
BDE 18
0.026
Palma et al. (2002)
Low
BDE 28
0.07
ATSDR (2004)
Low-
BDE 32
0.026
Palma et al. (2002)
Low
BDE 35
0.005
Palma et al, (2002)
Low
BDE 37
0.005
Palma et al. (2002)
Low
BDE 39
0.005
Palma et al. (2002)
Low-
BDE 47
0.001--0.002
Palma et al. (2002)
ATSDR (2004)
Low
BDE 66
0.018
ATSDR (2004)
Low
BDE 77
0.006
ATSDR (2004)
Low
BDE 85
9 x 10 s x lu '
0.006
Palma et al. (2002)
ATSDR (2004)
Low
BDE 99
9 x 10"7—2.4 x 10^
0.009
Palma et al. (2002)
ATSDR (2004)
Low
BDE 100
0.04
ATSDR (2004)
Low
BDE 116
9 X 10 N \ 10 '
Palma et al. (2002)
Low
BDE 119
9 x 10 s\ 10
Palma et al. (2002)
Low
BDE 128
4.15 x I0 '1
Palma et al. (2002)
Low
BDE 138
0.001
ATSDR (2004)
Low
BDE 153
0.001
0.0009
ATSDR (2004)
Tittlemier et al. (2002 )
Low
BDE 154
0.001
0.0009
ATSDR (2004)
Tittlemier et al. (2002)
Low-
BDE 172
2.16.x 10 7
Palma et al. (2002)
Low
3-65
-------
BDE 176
2.16 x ur
Palma et al. (2002)
Low
BDE 181
2.16 x 1(F
Palma et al. (2002)
Low
BDE 183
0.002
0.0015
ATSDR (2004)
Tittlemier et al. (2002)
Low
BDE 185
2.16 x lit
Palma et al. (2002)
Low
BDE 190
2.16 x ur
Palma et al. (2002)
Low
Solubility score is as follows: low - relatively insoluble; medium - somewhat soluble; high - very soluble
3-66
-------
Table 3-2. Estimated octanol water partition coefficients (log Kow) values for PBDEs.
Brorainated diphenyl ether
Log KflW coefficient
Reference
Deca-BDE (Commercial)
6.27
6.27
\i s|)R. 2004
European Union (2003)
Octa-BDE (Commercial)
6.29
6.29
ATSDR. 2004
European Union (2003)
Penta-BDE (Commercial)
6.64-6.97
6.57
ATSDR. 2004
European Union (2001)
BDE 11
5.83
Palma et al. ,2002
BDE 15
5.86
5.74
5.55
Kuramochi et al., 2004
A TSDR. 2004
Tittlemier et al.(2002)
BDE 17
5.52-5.88
Palma et al. (2002)
BDE 18
5.52-5.88
5.74
Palma et al. (2002)
ATSDR, 2004
BDE 28
5.94
5.98
ATSDR (2004)
Tittlemier et al.(2002)
BDE 32
5.52-5.88
Palma et al. (2002)
BDE 35
5.52-6.72
Palma et al. (2002)
BDE 37
5.52-6.72
Palma et al. (2002)
BDE 39
5.52-6.72
Palma et: al. (2002)
BDE 47
6.01-6.77
6.81
6.48
6.55
Palma et al. (2002)
ATSDR (2004)
Kuramochi et al. (2004)
Tittlemier et al. (2002)
BDE 66
6.73
Tittlemier et al. (2002)
BDE 77
6.73
Tittlemier et al. (2002)
BDE 85
6.57-7.66
7.03
Palma et al. (2002)
Tittlemier et al. (2002)
BDE 99
6.53-7.66
7.32
7.21
7.13
Palma et al. (2002)
ATSDR (2004)
Kuramochi et al. (2004)
Tittlemier et al. (2002 )
BDE 100
7.24
6.86
ATSDR (2004)
Tittlemier et al. (2002)
BDE 116
6.71-7.66
Palma et al. (2002)
BDE 119
6.71-7.66
Palma et al. (2002)
BDE 128
7.39- 8.55
Palma et al. (2002)
BDE 138
7.91
Tittlemier et al. (2002)
3-67
-------
BDE 153
7.9
7.83
7.62
ATSDR (2004)
Kuramoehi et al. (2004)
Tittlemier et al. (2002)
BDE 154
7.82
7.39
ATSDR (2004)
Tittlemier et al. (2002)
BDE 172
9.44
Palm a et al. (2002)
BDE 172
9.44
Paltna et al. (2002)
BDE 1 76
9.44
Palma et al. (2002)
BDE 181
9.44
Palma et al. (2002)
BDE 183
8.27
ATSDR (2004)
BDE 185
9.44
Palma et al. (2002)
BDE 190
9.44
8.36
Palma et al. (2002)
Tittlemier et al. (2002)
3-68
-------
Table 3-3. Estimated Henry's Law constants (H) for PBDEs
Brominated diphenvl ether
25"C)
Reference
DecaBDE (Commercial)
1.20E-08
3.95E-07
ATSDR, 2004
Cetin et al. (2005)
OctaBDE (Commercial)
7.5E-08 2.6E-07
ATSDR, 2004
PentaBD E (Commere ial )
3.5E-06-1.2E-05
ATSDR. 2004
BDE 15
2.07E-04
2.I0E-04
ATSDR, 2004
Tittlemier et al.
(2002)
BDE 28
5.03E-05
7.80E-05
5.10E-05
ATSDR {2004)
Cetin et al. (2005)
Tittlemier et al.
(2002)
BDE 47
1.48E-05
8.39E-06
ATSDR (2004)
Cetin et al. (2005)
BDE 66
4.93 E-06
ATSDR (2004)
BDE 77
1.18E-05
ATSDR (2004)
BDE 85
1.09E-06
ATSDR (2004)
BDE 99
2.27E-06
5.92E-06
ATSDR (2004)
Cetin et al. (2005)
BDE 100
6.81 E-07
2.37E-06
ATSDR (2004)
Cetin et al. (2005)
BDE 153
6.61 E-07
4.34E-06
ATSDR (2004)
Cetin et al. (2005)
BDE 154
2.37E-06
7.90E-07
ATSDR (2004)
Cetin et al. (2005)
BDE 183
7.3E-08
ATSDR (2004)
BDE 209
3.95E-07
Cetin et al. (2005 )
3-69
-------
Table 3-4. Estimated solid phase vapor pressures (Ps) and subcooled liquid vapor
pressures (PiJ of some PBDEs (Pascals at 25°C)
Brominated di phenyl
ether
T
m
(K)
T
m
Reference
Ps
Pi
Pl
Reference
Deca-BDE
(Commercial)
573.15
European
Union
9.28F.-09
4.63E-06
European union (2002)
Oeta-BDE
(Commercial)
473.15
European
Union
1.26E-07
6.59E-06
European union (2003)
Penta-BDB
(Commercial)
475.15
European
Union
8.60E-07
4.69E-05
European union (2001)
BDE 1
NA
0.163
Wong et al. (2001)
BDE 2
NA
0.128
Wong et al. (2001)
BDE 3
NA
2.59E-01
Tittlemier et al. (2002)
BDE 7
NA
1.68E-02
Wong et al. (2001)
BDE 8
NA
1.37E-02
Wong et al. (2001)
BDE 10
NA
2.77E-02
Wong et al. (2001)
BDE 12
NA
1.19E-02
Wong et al. (2001)
BDE 15
329.15
Marsh et al.
(1999)
8.59E-03
4.88E-03
1.73E-02
9.84E-03
Tittlemier et al. (2002)
Wong et al. (200!)
BDE 28
337.15
Marsh et al.
(1999)
9.07E-04
6.51E-04
2.19E-03
1.57E-03
Tittlemier et al, (2002)
Wania et al. (2003)
BDE 30
358,15
Marsh et al.
(1999)
1.18E-03
4.56E-03
Wong et al. (2001)
BDE 32
350.40
Palm et al.
(2002)
6.91 E-04
2.25E-03
Wong et al. (2001)
BDE 33
NA
1.49E-03
E49E-03
Wong et al. (2001)
BDE 35
413.01
Palm et al.
(2002)
E04E-04
1.39E-03
Wong et al. (2001)
BDE 37
321.65
Palm et al.
(2002)
6.00E-04
1.02E-03
Wong et al. (2001)
BDE 47
353.65
Palm et al.
(2002)
7.42E-05
5.52E-05
2.50E-04
1.86E-04
Wong etal. (2001)
Tittlemier et al. (2002)
BDE 66
NA
2.38E-04
1.22E-04
Wong et al. (2001)
Tittlemier et al. (2002)
BDE 69
NA
4.1 Mil; 1)4
Wong et al. (2.001)
BDE 75
408.15
Marsh et al.
(1999)
4.10E-05
4.92E-04
Wong et al, (2001)
BDE 77
368.15
Marsh et al.
(1999)
3.21E-05
1.40E-05
1.56E-04
6.79E-05
Wong et al. (2001)
Tittlemier et al. (2002)
BDE 82
NA
4.80E-05
6.47E-05
Wong et al. (2001)
3-70
-------
BDE 85
396.45
Palm et al.
(2002)
1.07E-06
9.86E-06
Tittlemier et al. (2002)
BDE 99
365.45
Palm et al.
(2002)
1.49E-05
3.85E-06
7.94E-06
6.82E-05
1.76E-05
3.63E-05
Wong et al. (2001)
Tittlemier et al. (2002)
Wania et al. (2003)
BDE 100
371.15
Marsh et al.
(1999)
5.50E-06
7.07E-06
2.86E-05
3.68E-05
Tittlemier et al. (2002)
Wania et al. (2003)
BDE 115
NA
3.02E-05
Wong et al. (2001)
BDE 138
NA
1.58E-06
Tittlemier et al. (2002)
BDE 153
456.15
Palm et al.
(2002)
1.63E-07
5.80E-06
5.80E-06
2.09E-06
Wong et al. (2001)
Tittlemier et al. (2002)
BDE 154
416.15
Marsh et al.
(1999)
2.64E-07
3.80E-06
Tittlemier et al. (2002)
BDE 183
NA
4.68E-07
Tittlemier et al. (2002)
BDE 190
470.4
Palm et al.
(2002)
I.85E-08
5.76E-09
9.05E-07
2.82E-07
Wong et al. (2001)
Tittlemier et al. (2002)
Note: P,s was calculated from Pl using equation 3-1 and assuming the melting point
temperature (Tm) in the table. NA = not available in the literature.
3-71
-------
Table 3-5. Estimated octanol/air partition coefficients (log Koa) of PBDEs
Brominated Diphenyl Ether
LogK„3
<•« 25° C)
Reference
BDE 1
7.24
Wania et at (2002)
BDE 2
7.36
Wania et al, (2002)
BDE 7
8.37
Wania et al. (2002)
BDE 8
8.47
Wania et al. (2002)
BDE 10
8.12
Wania et al. (2002)
BDE 12
8,55
Wania et al. (2002)
BDE 13
8.57
Wania et al, (2002)
BDE 15
8.64
Wania et al. (2002)
BDE 17
9.30
Harner and Shoeib (2002)
BDE 21
9.49
Wania et al (2002)
BDE 28
9.50
Harner and Shoeib (2002)
BDE 30
9.02
Wania et al. (2002)
BDE 32
9.28
Wania et al. (2002)
BDE 35
9.48
Wania et al. (2002)
BDE 37
9.68
Wania et al. (2002)
BDE 47
10.53
10.34
Harner and Shoeib (2002)
Wania et al. (2002)
BDE 66
10.82
10.49
Harner and Shoeib (2002)
Wania et al. (2002)
BDE 69
10.23
Wania et al. (2002)
BDE 75
10.13
Wania et al. (2002)
BDE 77
10,87
10.70
Harner and Shoeib (2002)
Wania el at. (2002)
BDE 82
11.14
Wania et al. (2002)
BDE 84
1 1.52
Wania et al (2002 )
BDE 85
11.66
Harner and Shoeib (2002)
BDE 99
11.31
11.28
Harner and Shoeib (2002)
Wania etal. (2002)
BDE 100
11.13
Harner and Shoeib (2002)
BDE 126
11.97
Harner and Shoeib (2002)
BDE 153
11.82
12.15
Harner and Shoeib (2002)
Wania et al, (2002)
BDE 154
11.92
Harner and Shoeib (2002)
BDE 156
11.97
Harner and Shoeib (2002)
BDE 183
11.96
Harner and Shoeib (2002)
3-72
-------
Table 3-6. Calculated theoretical vapor-particle partitioning of PBDE congeners in
ambient air at 25°C (calculated using equation 3-2).
Brominated Diphenyl
Ether
Vapor phase, %
Particle phase, %
BDE 1
100
0
BDE 2
100
0
BDE 3
100
0
BDE 7
100
0
BDE 8
100
0
BDE 10
100
0
BDE 12
100
0
BDE 15
100
0
BDE 28
99
1
BDE 30
99
1
BDE 32
99
1
BDE 33
98
2
BDE 35
98
9
BDE 37
98
-i
BDE 47
90
10
BDE 66
87
13
BDE 69
94
6
BDE 75
95
5
BDE 77
81
19
BDE 82
71
29
BDE 85
28
72
BDE 99
61
39
BDE 100
56
44
BDE 115
54
46
BDE 138
6
94
BDE 153
7
93
BDE 154
13
87
BDE 183
T
98
BDE 190
1
99
3-73
-------
Table 3-7. Estimated BAF and BMF values for various aquatic species.
Broininated Diphenyl
Species
Factor
Reference
Ether
B10ACCUMULATION FACTORS (BAF)
BDE 28
Lake Trout
7.6
T'omv et al. (2004)
BDE 47
Lake Trout
7.3
Streets et al. (2006)
Blue Mussels
6.1
Gustafsson et al (1999)
BDE 66
Lake Trout
7,3
Streets et al. (2006)
BDE 85
Lake Trout
2.3
Tomy et al. (2004)
BDE 99
Lake Trout
6.7
Sheas ct al. (2006)
Blue Mussels
6.1
(iust ilsson et al (1999)
BDE 10
Lake Trout
7.5
streets et al. (2006)
BIOMAGNIF1CATJON FACTORS (BMF)
BDE 15
Porpoise
1.6-2.4
Ramu et al. (2006)
BDE 28
Lake Trout
7.6
Tomy et al. (2004)
Porpoise
1.6 2.4
Ramu et al. (2006)
BDE 47
Lake Trout
2.1
Tomy et al. (2004)
Coho Salmon
3.2
Stapleton and Baker, (2003)
Porpoise
1.5
Ramu et al. (2006)
BDE 66
Luke Trout
7.8
Tomy et al. (2004)
BDE 99
Lake Trout
6.6
Tomy et al. (2004)
Porpoise
1.8-2.4
Ramu et al. (2006)
BDE 100
Lake Trout
6.5
Tomy et al, (2004)
Poipoise
1.7-2.4
Ramu et al. (2006)
BDE 138
Lake Trout
3.2 8.7
Tomy et al. (2004)
BDE 153
Lake Trout
9.4
Tomy et al. (2004)
Coho Salmon
4.0
Stapleton et al. (2003)
Poipoise
1.6-2.2
Ramu et al. (2006)
BDE 154
Lake Trout
13.3
Tomy et al. (2004)
Porpoise
1.4-2.2
Ramu et al. (2006)
BDE 183
Lake Trout
3.9
Tomy et al. (2004)
Porpoise
0.8-2.2
Ramu et al. (2006)
BDE 190
Lake Trout
1.6-5.1
Tomy et al. (2004)
BDE 209
Lake Trout
0.3
Tomy et al, (2004)
3-74
-------
Table 3-8, The BDE congener distributions in influent, sludges and final effluent from
various surveys of sewage treatment plants in various countries.
PBDE Congener
STP Influent
STP Sludge
STP
Location
Reference
(Br substitution)
(ng/L)
(fig/kg dwt)
Final Effluent
(ng/L)
BDE 1
NA
ND
ND
Palo Alto, CA
North (2004)
BDE 2
NA
ND
ND
Palo Alto, CA
North (2004)
BDE 3
NA
ND
ND
Palo Alto, CA
North (2004)
BDE 7
NA
ND
0.016
Palo Alto. CA
North (2004)
BDE 8
NA
ND
0.0042
Palo Alto. CA
North (2004)
BDE 10
NA
ND
ND
Palo Alto, CA
North (2004)
BDE 12
NA
ND
ND
Palo Alto, CA
North (2004)
BDE 13
NA
ND
ND
Palo Alto, CA
North (2004)
BDE 15
NA
0.62
0.008
Palo Alto, CA
North (2004)
BDE 17
NA
5.7
0.19
Palo Alto, CA
North (2004)
BDE 25
NA
0.62
0.0099
Palo Alto, CA
North (2004)
BDE 28
NA
13
0.266
Palo Alto, CA
North (2004)
1.3
22
ND
Ontario, CAN
Song et al. (2006)
BDE 30
NA
ND
ND
Palo Alto, CA
North (2004)
BDE 32
NA
ND
ND
Palo Alto, CA
North (2004)
BDE 35
NA
ND
0.005
Palo Alto, CA
North (2004)
BDE 37
NA
0.24
0.0038
Palo Alto, CA
North (2004)
BDE 47
NA
757
10.5
Palo Alto, CA
North (2004)
102
1,819
14
Ontario, CAN
Song et al (2006
NA
2.77
NA
Germany (11 STPs)
Knoth et al (2007)
NA
7.0*
NA
Sweden <22 STPs)
Oberg et al.(2002)
BDE 49
NA
18
0.266
Palo Alto, CA
North (2004)
BDE 66
NA
21
0.217
Palo Alto, CA
North (2004)
BDE 71
NA
2.8
0.043
Palo Alto, CA
North (2004)
BDE 75
NA
1.0
0.018
Palo Alto, CA
North (2004)
BDE 77
NA
ND
ND
Palo Alto, CA
North (2004)
BDE 85
NA
34
0.352
Palo Alto, CA
North (2004)
NA
0.42*
NA
Sweden (22 STPs)
Oberg et al. (2002)
BDE 99
NA
944
11.2
Palo Alto, CA
North (2004)
121
2,004
16.0
Ontario, CAN
Song et al (2006
NA
137.19
NA
Germany (1 1 STPs)
Knoth et al (2007)
NA
10*
NA
Sweden (22 STPs)
Oberg et al. (2002)
BDE 106
NA
165
2.0
Palo Alto, CA
North(2004)
19
289
2.8
Ontario, CAN
Song et al (2006)
NA
154.45
NA
Germany (11 STPs)
Knoth et al (2007)
NA
1.7*
NA
Sweden (22 STPs)
Oberg e< al. (2002)
BDE 105
NA
ND
ND
Palo Alto, CA
North (2004)
BDE 116
NA
ND
ND
Palo Alto, CA
North (2004)
BDE 119
NA
ND
0.014
Palo Alto, CA
North (2004)
BDE 126
NA
ND
ND
Palo Alto, CA
North (2004)
BDE 138
NA
7.7
0.096
Palo Alto, CA
North (2004)
1.0
26.1
ND
Ontario, CAN
Song et al (2006)
NA
ND
NA
Sweden (22 STPs)
Oberg et al. (2002)
BDE 140
NA
2.7
0.031
Palo Alto, CA
North (2004)
BDE 153
NA
88
0.98
Palo Alto, CA
North (2004)
11
193
1.60
Ontario, CAN
Song et al. (2006)
NA
27.31
NA
Germany (11 STPs)
Knoth et al. (2007)
3-75
-------
PBDE Congener
STP Influent
STP Sludge
STP
Location
Reference
(Br substitution)
(ng/L)
(jig/kg dwt)
Final Effluent
(ng/L)
NA
0.86*
NA
Sweden (22 STPs)
Oherg et al. (2002)
BDE 154
NA
68
0.776
Palo Alto, CA
North (2004)
7.6
120
ND
Ontario, CAN
Song et al. (2006)
NA
18.51
NA
Germany (11 STPs)
Knoth et al. (2007)
NA
0.72*
NA
Sweden (22 STPs)
Oberg et al.(2002)
BDE 155
NA
7.1
0.073
Palo Alto, CA
North (2004)
BDE 181
NA
ND
ND
Palo Alto, CA
North (2004)
BDE 183
NA
10
0.080
Palo Alto, CA
North (2004)
1.7
34
ND
Ontario, CAN
Song et al. (2006)
NA
13.66
NA
Germany (11 STPs)
Knoth et al. (2007)
BDE 190
NA
1.0
0.0039
Palo Alto, CA
North (2004)
BDE 206
NA
16
0.041
Palo Alto, C'A
North (2004)
BDE 207
NA
22
0.095
Palo Alto, CA
North (2004)
BDE 208
NA
1 1
0.051
Palo Alto, CA
North (2004)
BDE 209
NA
1,183
1.73
Palo Alto, CA
North (2004)
NA
363.46
NA
Germany (11 STPs)
Knoth et al. (2007)
NA
11
NA
Sweden (22 STPs)
Knoth et al. (2007)
NA
3.381
29
Palo Alto, C'A
North (2004)
Total 1151)1 s
265
4,324
36
Ontario, CAN
Song et al. (2006)
NA
987.21
NA
Germany (11 STPs)
Knoth et al. (2007)
NA - Not analyzed for; ND - not detected; *median concentration as opposed to mean.
3-76
-------
Figure 3-1. Time-trends of deposition flux of PBDEs (blue solid diamonds ) and PCBs
(red open diamonds) to the sediments in each of the Great Lakes, (left x axis shows
PBDE levels and right x axis shows PCB levels)
re
a>
L_
o
a
0 008 -
0.004 -
r 0 07
0.050 -
SU12 f
- tO
/
- 0,06
- 0.05
0 040 -
0 1
- 0 04
0 0'30 -
¦¦ t
- 0 03
0 020 -
—, ,
-0.02
-0.01
- fi rm
0 010 -
0.000 «
y.; u
0.160
ON4D
070 -
0.140
i.'.UbO
0.1 JO
3 050
0 100
0.U4U
o .uao
3 030
o.o&a
0. U2U
U U4U
J 010
0.U20
ij ijnij
0.000
Notes: SLf 12 = 12 samples Lake Superior; LIV118 = 18 samples Lake Michigan; IILi 12 = 12 samples
Lake Huron; ON 48 = 48 samples Lake Ontario. The X axis shows the year dated by Pb-210 in the
sediment core.
Source: Li et a!., 2006
3-77
-------
Chapter 4 ENVIRONMENTAL AND EXPOSURE MEDIA
CONCENTRATIONS
4.1. INTRODUCTION
This chapter summarizes the concentrations of individual congeners and total
PBDEs in environmental and exposure media. The emphasis is on data from the United
States, although key data sets from other countries will be described as a contrast and a
supplement. In addition, emphasis will be placed on the measurements for BDE 209, as
this deca congener is essentially the sole component of the deca formulation that is still
currently being produced and marketed in the United States. Much of the literature has
focused on the profile of congeners ("profile'" is the term used in this chapter to describe
the suite of congeners) associated with the penta formulation (BDEs 47, 99, 100, 153, and
154), and, subsequently, there is a paucity of data for the primary markers of the octa
formulation, BDE 183 (about 40% of the octa formulation), and the deca formulation,
BDE 209 (about 98%). When possible, information on the possible debromination of
BDE 209 will be provided. The sections on air, indoor dust, fish, and food include a
comprehensive table of reported measurements of individual congeners in the United
States; sections on water, sediment, and surface soil do not include such tables because
there is little or no United States data on these media in the literature. The chapter
concludes with development of a tabular assignment of BDE congener concentrations for
each exposure media. These assignments will be used in the next chapter, which
determines exposure dose of BDEs based on exposure media concentrations and contact
rates.
4.2. WATER AND SEDIMENT
Water has been rarely sampled for PBDEs in America, and it is questionable
whether available monitoring can be considered representative of drinking water. The
San Francisco (SF) Estuary Regional Monitoring Program for Trace Substances sampled
water, surface sediments, and bivalves (oysters, mussels, clams) in the SF Estuary for 22
BDE congeners, including BDE 209 (Oros et al., 2005). A total of 33 water samples
were taken, with total PBDEs ranging from 3-513 pg/L, and a mean concentration of 146
4-1
-------
pg/L. The region of the bay which had the highest concentrations, the Lower South Bay,
receives 26% of the Estuary's total publicly owned treatment works (POTW) wastewater
effluents and only 10% of the Estuary's freshwater inflow. The most abundant congeners
were 47, 99, and 209. When 209 was not reported as "Q" (outside QA limits, detected
but not reported), it was positive and of a comparable magnitude as 47, though mostly at
a slightly higher concentration. BDE 47 concentrations ranged from about 17 to greater
than 60 pg/L, with one high sample at 123 pg/L; BDE 99 was next highest with
quantified concentrations (about 1/3 of the samples were outside quality limits) ranging
from about 11 to 35 pg/L, with one high level at 90 pg/L . BDEs 17, 100, and 28/33
(congeners notated as "28/33" translate to congeners which have co-eluted and arc
analyzed as though they were one congener) were quantified at levels less than 5 pg/L,
and others not below detection or very infrequently detected at levels less than 5 pg/L. It
was found that the PBDEs were predominantly associated with the sediment fraction of
the water column.
A second study, by Johnson et al. (2006), sampled for PBDEs in water and fish of
Washington state rivers and lakes. Results from 15 samples taken in 7 rivers and 3 lakes
in 2005 and 2006 were available. BDEs measured included 47, 49. 66, 71, 99, 100, 138,
153, 154, 183, 184, 190, and 209, although BDE 209 was never detected. Total
concentrations ranged from 1 to 926 pg/L, although only two samples were above 100
pg/L, with one of them at 146 pg/L. The average was 91 pg/L, but there was a much
lower median of 16 pg/L. The most abundant congeners found, and at the highest
concentrations, were 47, 99, and 100.
Finally. Streets et al. (2006) measured PBDEs in the dissolved and particle-bound
phases in Lake Michigan and found BDE congeners 47, 99, 100, and 66 were in the
dissolved phase in concentrations ranging from 0.13 to 10 pg/L (for individual
congeners). The quantified particle-phase concentrations for BDEs 47, 99, and 100
ranged from 0.18 to 1.4 pg/L. BDE 209 was not evaluated in this study.
In summary, only three studies measuring surface water concentrations in the
United States could be found. In one study, total concentrations in the SF estuary ranged
from 3 to 500+ pg/L, dominated by BDE congeners 47, 99, and 209. In the second study,
BDE 209 could not be quantified, and total concentrations of the lower brominated
4-2
-------
congeners ranged widely from I to 916 pg/L, although only 2 samples were above 100
pg/'L, and one of them was at 146 pg/L. The mean and median in that study were 91 and
16 pg/L, respectively. The third study did not measure BDE 209, and, in that study, the
authors separated out the dissolved and particle phase concentrations. In total, only the
sum of BDEs 47, 66, 99, and 100 ranged up to 10 pg/L.
Much of the data on PBDEs in sediments of water bodies were taken in the
context of sediment core studies, whose purpose is to elaborate on temporal trends of
PBDEs in the environment (see Section 3.8.1). This section provides an overview of
studies just focusing on surface sediments and on the surficial sediment concentrations in
the sediment core studies.
Briefly, the sediment core studies showed BDE 209 to dominate surficial
sediment profiles. The average concentrations of BDE congeners, not including BDE
209, in these studies were generally less than 10 ng/g, while BDE 209 concentrations
ranged anywhere from 10 ng/g to 300+ ng/'g. The maximum total BDE concentrations
found exceeded 1,000 ng/g in some cases and they were dominated by BDE 209, which
comprised over 90% of the total mass measured.
The study of the SF Estuary included both water and sediment sampling (Oros et
al., 2005). A total of 48 sediment samples were taken and positive detections were only
noted for BDEs 47, 99, 183, 204, and 205. It is noteworthy that BDE 209 was not found
with a detection limit of 1.5 ng/g. This contrasts all other sediment studies where BDE
209 dominated the profile. It is also the only study found which provided concentrations
of BDEs 204 and 205. Total concentrations ranged from ND (not detected) to 212 ng/g
dry weight (dwt) basis. BDE 47 was detected 42% of the time, with a range of detections
of 1.1 to 100 ng/g dwt, with an average of 12 ng/g dwt; BDE 99 was detected 77% of the
time with a range of 0.3 to 71 and a mean of 5 ng/g dwt; BDEs 183, 204, and 205 were
detected once, twice, and once respectively.
Raff and Hites (2004) collected suspended sediment samples from 26 sites along
the Mississippi River and five of its major tributaries during July/August 2002 and March
2003. A total of 15 congeners were measured; individual congeners were not identified
nor were individual congener concentrations provided. Total concentrations ranged from
31-1,548 ng/g dwt, with an average of 327 ng/g dwt. Consistent with the Great Lakes
4-3
-------
sediment cores described in Chapter 3, BDE 209 was the overwhelmingly dominant
congener. On average, it comprised 96.8% of the total concentration. Congeners 47 and
99 were the only other congeners accouting for concentrations in excess of 1% of the
total PBDE concentrations—1.16 and 1.26%, respectively. Based on concentrations near
the mouth of the river, combined with records on suspended sediment concentrations and
outflows, the authors estimated that 8 tons/yr of PBDEs arc discharged into the Gulf of
Mexico. Other interesting trends include higher concentrations in the spring, attributed to
high runoff, and evidence of debromination, as two sites contained higher concentrations
of the nona congener BDE 206, as compared to BDE 209.
Ashley et al. (2006) analyzed four sediment samples, along with samples from
eels taken from the Delaware River (eel data summarized below). The sediment samples
were collected in 2002 as part of an earlier polychlorinated biphenyl (PCB) study, and
they were reanalyzed in 2006 for PBDEs. While the total concentrations in these samples
were lower than some of the studies above—between 0.7 and 21.7 ng/g dry weight—the
proportions were similar. BDE 209 explained about 50% of the concentration, while
BDEs 99 and 47 were next, accounting for 15 and 14%, respectively.
Four surficial sediment samples were taken in Lake Hadley in Indiana (Dodder et
al.. 2002). Total BDEs, including 47, 99, 100, 153, 154, and 209, ranged from 24 to 71
ng/g dwt, dominated by BDE 209 at 19 to 33 ng/g dwt. Other than a measurement of 22
ng/g dwt for BDE 99, all other measurements of the other congeners were near or less
than 5 ng/g dwt.
Hale et al. (2002) studied the environmental impacts of a polyurethane foam
manufacturing facility in the United States mid-Atlantic Region which had ceased
production in 1997; sampling occurred in 2001. The interior of the facility was tested, as
was soil adjacent to the facility, sediment in a stream leaving the facility, and sediment
and bluefish in a pond about 250 meters from the facility. Total concentrations of PBDEs
(including BDEs 47, 99, and 100; 153 and 154 were measured but not detected) in
surface sediments in the stream leaving the facility were 17.2 ng/g dwt and 132 ng/g dwt,
with one sample having all non-detccts. In the pond, total concentrations in surface
sediments were 0.5 and ND ng/g in two pond sediment samples These concentrations are
4-4
-------
similar to measurements described at other settings, suggesting that the foam production
facility did not result in noteworthy impacts to nearby aquatic settings.
Hale et al (2001) also measured PBDE (along with PCB) concentrations in fish
and sediment samples from two large Virginia watersheds. The study sites included the
larges bodies of freshwater in Virginia; they were not selected based on concerns from a
particular source of contamination, although the general predominance of furniture
manufacturing facilities was noted. Total PBDE concentrations, including BDEs 47, 99,
100, 153, and 154, ranged from non-dctect to 52.3 ng/g dwt in 17 samples. BDE 49 was
also measured, but not detected in sediment samples. The concentrations were dominated
by BDE 47, explaining about 53% of total concentrations, with BDE 99 second at about
35% of total, and the remaining concentration due to BDE 100 (7%), 153 (<5%), and 154
(<5%).
Toms et al. (2006) undertook a comprehensive assessment of the fate and
distribution of 26 BDE congeners in the aquatic environment of Australia in 2003-2004.
An aim of this study was to determine the background concentrations and congener
compositions of BDEs in estuarine, freshwater, and marine sediments. Ninety sediment
samples were analyzed from locations representing various land uses ranging from
remote to industrial. A total of 25 BDE congeners were detected in samples from 35 of
46 sites (76%), and total PBDE concentrations ranged from non-detect to 60.9 ng/g dw
with an overall mean (± SD) and median of 4.7 ± 12.6 and 0.3 ng/g dw, respectively.
BDE-209 dominated the congener distribution in 86% of the sediment samples.
Christcnscn and Platz (2001) sampled sediment from Danish marine coastal areas,
freshwater lakes, and a river in 2000. The congeners measured included BDEs 47, 99,
100, 153, and 209, and the total concentrations ranged from 0.06-24.7 and 0.07-10.6
ng/g dw in marine and freshwater sediment, respectively. BDE-209 dominated the
congener profile in marine and freshwater sediments with a median concentration of 3,35
ng/g dw and 2.05 ng/g dw, respectively. The rank order of BDE congeners by
concentration in sediment was 209 > 99 > 47 > 100 > 153. The highest concentrations of
BDEs were detected in sediments in harbors and lakes located in urban areas. Because
PBDEs were never produced in Denmark, the authors postulated that PBDEs entered the
-------
Danish environment primarily by long-range transport. The authors suggested that
additional local sources could be evaporation and leaching from PBDE-treated products.
In a study by Eljarrat et al. (2005), 13 marine sediment samples were collected
from three coastal areas of Spain in 2002. The sediments were analyzed for 40 BDE
congeners, but only 12 were detected in the sediments: BDEs 28, 33, 47, 66, 77, 100, 99,
1 18, 154, 153, 183, and 209. Total PBDE levels ranged from 2.7 to 134 ng/g dwt in
coastal marine sediments, with highest levels in sediments off the coast of Barcelona. All
the sediment samples were dominated by BDE-209, which constituted between 50 and
99% of the total PBDE contamination. The usual congener profile found in the coastal
marine sediments was 209 > 47 > 99 > 100 > 153.
Rayne et al. (2003) measured surficial sediment in the Columbia River System in
southeastern British Columbia and found total BDEs (for mono- through hexabrominated
BDEs) ranged from 3.8 to 90.9 ng/g oc (organic carbon). Given that organic carbon is
around 1% total dwt in water body sediments, this suggests concentrations less than 1
ng/g dwt. They did not measure for BDE 209.
Water body sediment studies in North America included one in the San Francisco
Estuary, one in Virginia watersheds, one in a pond near a closed polyurethane foam
manufacturing factiliy, one in British Columbia, and several core studies describing
trends in the Great Lakes (the trend studies are reviewed in Chapter 3). Overall, total
concentrations were mostly less than 20 ng/g dwt but ranged as high as 300 ng/g dwt in
Lakes Michigan and Erie. With one exception, the study in the SF Bay Estuary (Oros et
al., 2005), over 90% of the total concentration was BDE 209 when it was measured. In
the SF Bay, BDE 209 was not detected in sediments at all (detection limit sufficiently
low at 1.5 ng/g dw), and BDE 47 dominated the profile.
4.3. SURFACE SOIL
Only one systematic study could be found that looked at PBDEs in surface soils in
the United States (Offenberg et al., 2006) in predominantly suburban, background,
settings. A total of 33 surface soil samples were taken in 15 states and measured for 30
BDE congeners. Concentrations of total BDEs av eraged 103 ng/g and had a geometric
mean concentration of 5.3 ng/g dry weight dwt, and a range of 0.09 to 1200 ng/g dwt.
4-6
-------
BDE 47 was detected in 31 of 33 samples averaging 1.9 ng/g dwt over the entire data set
(ND = 0). BDE 99 was observed in 30 samples and averaged 3.6 ng/g dwt. BDE 209
was found in 24 samples and averaged 15.3 ng/g dwt. The highest coneentrations were
found for BDE 183, but it was only found in 3 samples at eoncentrations ranging from
121 to 562 ng/g dwt, so that the survey-wide average was 37.4 ng/g dwt.
Hale et al (2002) measured surface soil near a polyurethane foam production
facility which had been closed a few years earlier than sampling. They quantified
concentrations of BDEs 47, 99, and 100 (BDEs 153 and 154 measured for but not
detected) in 3 samples with total concentrations of ND (no congeners detected), 13.6, and
76.0 ng/g dwt. In these two samples, BDEs 47 and 99 dominated explaining near half the
total concentrations each.
One European study looked at background soils, and one study was found
evaluating soil concentrations near an electronics recycling facility in China. Hassanin et
al. (2004) reported on sampling of 66 surface samples (0-5 cm) and for 38 of these, a
paired subsurface sample (5-10 cm), taken from grassland and woodland areas in the
United Kingdom and Norway. These samples were collected in 1998, and 20 BDEs were
measured for, although BDE 209 was not measured. The congeners most routinely
detected (85-100% of the time) in the samples were 47, 99, 100, 153, 154, and 183.
Major trends include the following: United Kingdom concentrations were higher than
Norway concentrations, woodland concentrations were higher than grassland
concentrations, and subsurface concentrations were significantly lower than surface soil
concentrations. The total concentration of all BDEs ranged from 0.065 to 12.0 ng/g dwt.
The results for grassland/United Kingdom, woodland/United Kingdom, and
woodland/Norway were presented in terms of medians, min/max, and percent detected.
The selected range of median concentrations for these three groupings of surface soils,
were (pg/g dwt, with percent detected in parenthesis): BDE 17 (52-75%) - 27, 28, 35;
BDE 28 (14-76%) - 17, 21, 29; BDE 47 (100%) - 61, 490, 250; BDE 99 (92-95%) -
280, 900, 360; BDE 100 (90-95%) - 36, 110, 58; BDE 153 (54-95%) - 72, 210, 51;
BDE 154 (90-100%) - 22, 100, 42; BDE 183 (54-100%) - 26, 70, 25. The authors state
that the percentage found in soil mirrored the technical penta product, Bromkal 70-5DE,
with BDE 47 averaging 21% of total PBDE, 99 averaging 40%, 100 averaging 6%, 153
4-7
-------
averaging 8.7%, and 154 averaging 4.4%. A major shortcoming of this surface soil
study, however, was that BDE 209 was not measured. Given high measurements of BDE
209 in sediments, its lack of measurement in this broad ranging surface soil sampling
study is unfortunate.
Cai and Zang (2005) characterized soil concentrations near an electronics
recycling facility in China. For the six soil samples, the congeners measured and their
average concentrations were as follows: BDE 3 presented as ND (3 is a mono-BDE);
BDE 15 presented as 0.76 ng/g dwt (15 is a di-BDE); BDE 28—5.01 (tri-BDE); BDE
47—217; BDE 99—552; BDE 139—35; BDE 153—74; BDE 154—49; and BDE 183—
12.3. The total concentration is 945 ng/g dwt. The authors state that the isomer pattern of
47, 99, 139, 153, and 154 is consistent with the penta-BDE formulation. Again, no
measurements of BDE 209 were made in this study.
In summary, the sparse literature suggests background soil concentrations of total
BDEs might average above 100 ng/g dwt, although a central tendency estimate
(geometric mean or median) might be lower at near 5 ng/g dwt. Total concentrations in
soils at industrial sites, which might suggest a more direct source, could range as high as
1,000 ng/g dwt. The congener pattern noted in the studies on background soils is similar
to the penta-BDE formulation. However, BDE 209 has mostly not been sampled
frequently in soil studies; in the one study where it was measured, it was found in 24 of
33 samples, averaging 15.3 ng/g dwt.
4.4. INDOOR DUST
House dust was the focus of several United States studies because of the concern
for indoor exposures in residences. Data was also available in one study on
concentrations in a computer laboratory. As will be seen, indoor dust from both homes
and places of work are dominated by BDEs 47, 99, and 209. Table 4.1 provides
congener-specific concentrations for indoor dust from the literature.
House dust and dryer lint samples were collected from 16 homes in the
Washington, DC area and 1 from Charleston, SC. The samples were analyzed for 22
individual PBDE congeners (Stapleton, et al., 2005). Total concentration ranged from
780 ng/g dwt to 30,000 ng/g dwt, with a mean total of 5,900 ng/g dwt. The dominant
4-8
-------
congeners were the ones associated with the penta and decaBDE commercial mixtures.
BDEs 47, 99, and 209. The mean concentration of these three congeners was 1,220;
1,700; and 2,090 ng/g dwt, respectively. No correlations were found in the study with
year of construction, type of flooring (hardwood vs. carpet), or the number of televisons
and PCs (and hours of computer use per week). However, an inverse relationship was
found with the area of the home and the contribution of BDE 209 to the total PBDE
concentration in dust. Clothes dryer lint, examined in five of the homes, showed
concentrations ranging from 480 to 3,080 ng/g dwt. The two house dust samples with the
highest concentration had the commercial pentaBDE commercial ration of 0.6 for BDE
47/BDE 99.
Ten women from a prior Environmental Working Group (EWG) study on breast
milk (of a total of 20 women in the earlier study ) collected samples of dust from their
home (Sharp and Lunder, 2004). The 10 samples were taken in CA (2 samples), TX, CO,
DC, MI, WA, OR, FL, and MT. A total of 13 congeners were sampled; detection levels
ranging from 50 ng/g dwt for the lower brominated congeners to 90 ng/g dwt for the mid-
brominated congeners and finally 400 ng/g dwt for BDE 209. One of the 10 individuals
had very high measurements, at a total of 41,203 ng/g dwt for all congeners; the next
highest was 16,366 ng/g dwt. This individual had used her vacuum to clean up
polyurethane foam residues when she removed carpet padding, two mattress pads, and an
uncovered foam cushion from her home. This study did not find the archtypal penta BDE
commercial blend of 0.6 for BDE 47/BDE 99 in all cases. In half of the samples, a
significantly higher amount of BDE 47 was found. However, the overall averages of the
three highest congeners, BDEs 47, 99, and 209, were similar to the findings for homes in
the Washington, DC area: BDE 47 was found at an average of 1,847 ng/g dwt, BDE 99
at 2,352 ng/g dwt, and BDE 209 at 2,394 ng/g dwt. Concentrations in the dust were not
correlated to the number of electronic appliances or computers, foam furniture, or recent
remodeling.
BDE concentrations in dust were determined from 10 vacuum bag samples from
Atlanta, GA and from 10 vacuum bag samples from Germany, Australia, and the United
Kingdom fSjodin, et al., 2004) for a total of 40 samples. Only the United States and
German samples, and only BDEs 47, 99, 100, 153, 154, 183, and 209 were available in
4-9
-------
Sjodin ct al. (2004). The United States data were substantially higher than the German
data: the range of total PBDEs in Germany ranged from 17-550 ng/g dwt (median - 74)
while the United States data ranged from 530-29,000 ng/g dwt (4,200 median). The
authors claimed that the pentaBDE pattern found in these samples was similar to that in
the products; specifically, that BDE 99 is similar in concentration to that of BDE 47.
This contention is different than made by others, who note that the ratio of BDE 47 to
BDE 99 in commercial penta formulations is about 0.60. Median concentrations (mean
concentrations were not available) of the top three congeners, 47, 99, and 209, were 430,
880, and 2,000 ng/g dwt. VVenning et al. (2006) collected vacuum bag dust and air
conditioner filter dust from homes in Northern California and Wellington, New Zealand,
but the average congener-specific concentrations they presented were for all samples, not
distinguishing by geography or whether it was vacuum or air conditioner generated dust.
The average total BDE concentration over 13 samples, including 10 vacuum and 3 air
conditioner samples, was 13,570 ng/g dwt, with BDE 209 comprising 9,052 ng/g dwt,
and BDEs 99 and 47 comprising 2,140 and 1,120 ng/g dwt respectively. The BDE 209
results were skewed by a business air conditioner filter showing 84,500 ng/g dwt, from a
total of 96,300 ng/g dwt, so the general representativeness of these samples must be
questioned.
Indoor air and dust were sampled in 120 homes in Cape Cod, MA, in two rounds
of 60 samples each starting in 99 and ending in 2001 (Rudel et al., 2003). A total of 89
homes were sampled for PBDEs in dust. Only 3 PBDEs (47, 99, and 100) were analyzed,
and the reporting limits were high—400 ng/g dwt for BDEs 47 and 99 and 300 ng/g dwt
for BDE 100; the data were not tremendously informative. BDE 47 was found in about
50%of the samples with a maximum of 9,860 ng/g dwt; BDE 99 was found also about
50% of the time with a maximum of 22,500 ng/g dwt; and BDE 100 was found 20% of
the time at a maximum of 3,400 ng/g dwt.
Four computer wipe and 9 vacuum bag bulk dust samples were taken in Dallas
Texas in 2004 (Sehecter et al., 2005; date of sampling not given but presumed to be in
2004 as paper was submitted in July, 2004). The computer wipe samples were in units of
ng/'100 cm2, and because no other data are available in these units to compare with,
results from this study are not provided in Table 4.1. Briefly, concentrations ranged from
4-10
-------
77 to 1,536 ng/100 cm', with BDE 209 explaining over 90% in the two samples of PC
monitor screens, and about 53% in both PC casing samples. In these PC casing samples,
BDEs 99 and 47 were next in predominance, accounting for 25 and 11%, respectively. In
9 bulk dust samples, the total ranged from 705 to 65,777 ng/g dwt, with a median of
2,507 and a mean of 12,136 ng/g dwt. BDE 209 was the prominent congener in 7 of 9
samples (BDE 99 was the dominant in the other 2 samples), explaining 95% of the
concentration in the highest sample (65,777 ng/g dwt of 69,283 ng/g dwt total) and 66-
87% in the other samples. In one sample, however, BDE 209 explained less than 1% of
the 30,368 ng/g dwt total, with BDE 99 explaining 13,841 (46%) and BDE 47 explaining
10,538 ng/g dwt (35%).
Allen et al. (2008) reports on BDEs in house dust samples that were part of a
study looking at the correlations between indoor air, dust wipe samples, and bulk dust
samples in 20 homes in the urban setting of Boston, MA. They took three bulk samples
out of each home and developed geometric mean congener-specific concentrations for the
three common locations: the main living room, the bedroom, and a sample from a home
vacuum (location undetermined). Like other studies, BDE 209 dominated all locations,
with the geometric means over these three locations being 4702, 1866, and 1811 ng/g
dwt, respectively. BDE 99 was second highest at 2460, 1170, and 536, respectively, and
BDE 47 was the third most prevalent at 1865, 837, and 338, respectively. The geometric
mean of the total concentrations (including 17, 28/33, 47, 49, 66, 75, 85/155, 99, 100,
138, 153, 154, 183, 196, 197, 203, 206, 207, 208, and 209) of the three locations were
13732, 6255, and 4,269 ng/g dwt, respectively. They found a high correlation between
dust wipes and bulk dust samples. They did find a correlation between air and dust
concentrations of the penta formulation congeners (BDEs 17 up to BDE 154), but did not
find a correlation between BDE 209 air and dust concentrations. They found the highest
concentration of BDE 209 - 527,000 ng/g dwt - and the highest concentration of total
PBDEs - 544,000 ng/g dwt - that has been reported in the literature.
In another study of the Boston, MA, area, VVu et al. (2007) sampled 11 houses and
measured for 17, 28, 47, 66, 85, 99, 100, 138, 153, 154, 183, and 209. This was a study
in which womens' breast milk in these homes were also sampled (results reported in
Chapter 5). Results were log normal with a geometric mean of total BDE concentration
4-11
-------
of 1,910 ng/g dwt. BDE 99 and 47 were quantified nearly all the lime, with median
concentrations of 1010 and 670 ng/g dwt, respectively. BDE 209 was quantified in 5 of
11 samples, with quantified concentrations ranging between 1360 and 9020 ng/g dwt.
The median total concentration was 1910 ng/g dwt. There was a "high con-elation"
between dust levels and breast milk levels, as well as with reported dietary habits,
particularly the consumption of dairy products.
Harrad et al (2008) sampled between 10 and 28 homes in each of these locations;
AmariHo/Austin in Texas; Birmingham, UK; Toronto, Canada; and Wellington, New
Zealand. They measured for BDEs 28, 47, 99, 100, 153, 154, 183, and 209. The most
interesting finding was that the profile of BDE congeners found in the different locations.
The UK site had the highest BDE 209 concentrations, with an average of 45,000, a
median of 2,800, and a geometric mean of 3,800 ng/'g dwt (n=T6), compared to the results
from Texas which had an average of 1,600, a median of 1,300, and a geometric mean of
1,300 ng/g dwt (n=17). The two highest concentrations of BDE 209 in the UK, at
520.000 and 100,000 ng/g dwt, are the highest for this congener ever reported in dust
(Harrad et al 2008). The concentrations of the tri-hexa congeners (BDEs 28 thru 154)
were the highest for the Texas dust. For example, the geometric mean for the sum of
these six congeners from the Texas sites was 1,800 compared to 52 from the UK sites.
Due to the high BDE 209 findings, the overall total concentations were highest for the
UK. The geometric mean total concentrations for the UK, US, and Canadian sites were:
4,500; 3,600; and 1,200 ng/g dwt. Totals were not provided for New Zealand because the
BDE 183 and 209 congeners were not measured. Concentrations of the tri-hexa
congeners totaled 92 ng/'g dwt, which was comparable to the tri-hexa totals for the UK.
The California Air Resources Board (CARB) has conducted a substantial amount
of sampling for brominated flame retardants in air and dust, both outside near industrial
sites and in urban and background locations, and inside at workplaces and industrial sites
(see http://www.arb.ca.gov/toxics/pbde.htni). Two samples of carpet dust in an indoor
computer training facility totaled 10,200 and 4,800 ng/g dwt (CARB, 2005). Congeners
in the dust that were not in the air that was sampled in this facility included BDEs 203,
183, 155, 138, and 77 (203 and 183 are congeners in the octa formulation). The
4-12
-------
dominant congener was BDE 209, at 7,560 and 2,800 ng/g dwt, followed by BDE 99 at
856 and 695 ng/g dwt and BDE 47 at 502 and 411 ng/g dwt.
Sampling of indoor dust has also occurred in foreign countries, including Europe
and Kuwait. Fabrellas et al. (2005) report sampling done by "Euroconsumers
Organization" and "CIEMAT-POPs Group," in which dust was collected from 100
vacuum cleaner bags from Spain (34 bags), Belgium (32 bags), Portugal (22 bags), and
Italy (12 bags) and measured for BDEs. Homologue group concentrations from mono- to
liepta-BDE were determined; Levels of detections ranged from 0.006-0.220 ng/g dwt
(BDE 3) to 0.55-20.9 ng/g dwt (for BDE 209). Results suggest Italy has the highest total
concentrations, at 581 ng/g dwt, followed by Portugal (354 ng/g dwt), Spain (238 ng/g
dwt), and Belgium (190 ng/g dwt), however, it is not known if these mean values are
significantly different The highest value from Spain was sucked out from the inside of a
computer. Urban samples were found to be higher than rural, and BDE 209 was found to
comprise greater than 60% of total PBDE in all cases. The ratio of BDE 47 to BDE 99,
which is 0.5—0.7 in 24% of the samples, is the same as the commercial penta
formulation, DE-71, which has ratio of 0.6, suggesting these samples were dominated by
this commercial mixture. Harrad et al. (2006) measured BDEs in indoor dust at 8 homes
sampled in 2005 in the UK. They reported an average of 215 ng/g total (maximum of
625 ng/g total), which included BDEs 28, 47. 49. 66, 85. 99, 100, 153, and 154 (209 not
measured). The ratio of 47:99 was about 0.5. The study that Harrad et al. (2008) later
published on different locations around the world (United Kingdom, United States, New
Zealand, and Canada) included this earlier data. Karlsson et al, (2007) report on
measurements of PBDEs in air, housedust, and blood from five households. The average
total BDE (including 13 BDEs) was about 690 ng/g dwt, dominated by BDE 209 that was
490 ng/g dwt, followed by BDEs 99 and 47, at 99 and 51 ng/g dwt, respectively. Gevao
et al. (2005) report on simultaneous air and dust from homes in Kuwait in 2005. PUF
samplers were used to measure air samples and bulk dust samples were obtained from
vacuum bags in 17 homes in Kuwait between Feb 29 and April 11, 2004. Individual
congeners measured included BDEs 28. 47, 100, 99, 85, 154, 153, 183 (209 not
measured), although individual concentrations were not provided. Total PBDE dust
concentrations ranged from 0.2-24 ng/g dwt (geometric mean 9 ng/g dwt) and were log
4-13
-------
normally distributed. BDEs 47 and 99 seemed to track well with what the authors
claimed was a commercial penta mixture; however, the 47/99 ratio of this mixture was
not 0.6, but rather close to 1.0. House dust correlated with indoor air —when one was
high, the other was also.
Recent data from Europe show very high levels of BDE 209. Household dust and
lint samples were collected in Scotland, Northern England, and Germany (Pless-Mulloli,
et al., 2006) and compared with other published data from the United States, although the
comparison did not include more recent data such as that from Allen et al (2008). The
United Kingdom had the highest concentrations, though they were dominated by BDE
209, while Germany's concentrations were much lower. The United Kingdom dust
samples had a mean of 11,325 ng/*g, and a median of 3.933 ng/g and a maximum of
54,858 ng/g. The UK samples were dominated by BDE 209, which had an average
concentration of 11,233 ng/g (of the 11,325 ng/g total). The median of United States dust
measurements were similar to the median of the United Kingdom's measurements at
4,200 ng/g, but BDE 209 was not nearly as dominant, having a median of only 2,000
ng/g. As discussed earlier, the data in Allen et al (2008) from homes in the Boston, MA,
area had higher BDE 209 concentrations than these measurements summarized in Pless-
Mulloli et al. (2006).
In summary, the most common indoor congeners include those associated with
the penta formulations, BDEs 47 and 99, and the single congener most associated with
the deca formulation, BDE 209. Concentrations of total BDEs ranged from the low 100s
to well over 10,000 ng/g dwt. Most authors reported that the results suggested the
presence of the penta formulation, based on the ratio of the two congeners, BDEs 47 and
99. However, there was not uniform agreement on the ratio: some claimed that
comparabl e amounts of the two congeners suggest the presence of the commercial
product, while others claim that the penta formulation translates to a consistent 47/99
ratio of about 0.6. When quantified, BDE-209 dominated the profile above 47 and 99,
with concentrations generally in the 2,000-10,000 ng/g dwt range, although with some
outlier concentrations above 50,000 ng/g dwt. Dust concentrations of BDEs appear much
higher in United States's dust, well into the range of 1,000s of ng/g dwt for total
concentrations, compared with limited measurements in Europe, which appear to be
4-14
-------
limited to the 100s of ng/g dwt (except for data from the United Kingdom showing
substantial concentrations of BDE 209), and, in one study from Kuwait, with the highest
measurement just above 100 ng/g dwt.
4.5. AIR CONCENTRATIONS
Efforts to monitor PBDEs in outdoor air have been more extensive than efforts to
monitor indoor air in the United States, despite evidence that shows indoor air to have
generally higher concentrations of BDEs than outdoor air. Outdoor measurements have
been taken in California, in the Great Lakes Region, and in several other states including
New York, Louisiana, Indiana, Maryland, and other states. Only one study focusing on
indoor air in residential settings, in the Boston MA area, could be found. Sites in Europe
and Asia suggest lower concentrations as compared to the United States, although one
study from China reports much higher concentrations. One study reported on air
concentrations within cars in Greece. Section 4.5.1 describes efforts to monitor outdoor
air while Section 4.5.2 describes indoor-air studies. Table 4.2 provides congener-specific
concentrations for both outdoor and indoor air in the United States.
4.5.1. Outdoor Air Concentrations
CARB established the California Ambient Dioxin Air Monitoring Network. This
Network, comprised of 11 sites, began sampling in 2002, and the data for the years 2002-
2004 are posted at http://www.arb.ca.gov/pub/dioxin/cadamp.php. PBDE data was taken
from 7 of these 11 sites, 4 from the Bay Area and 3 from the South Coast, beginning in
2003. There were 6 monthly samples in 2003, and 12 monthly samples in 2004.
Individual site data and site/statewide averages are available from the Web site. Twelve
congeners were measured, including 17, 28, 47, 65, 88, 85, 99, 100, 153, 154, 183, and
209. BDEs 47 and 99 had similar 12-month averages in 2004, of 53 and 51 pg/m3,
respectively, with BDE 209 coming in third at 25 pg/'nr. BDE 100 averaged 13 pg/m3,
and others ranged from 0.04 to 4.0 pg/m3. The total average total concentration for 2004
was 160 pg/m3.
CARB (2005) also sponsored a research monitoring effort that included indoor
and outdoor sampling at industrial sites, office sites, and outdoors at the University of
4-15
-------
California at Davis (UC Davis). The UC Davis sampling was conducted to evaluate the
effectiveness of two different active outdoor samplers: one included a filter (for capture
of particle-bound PBDEs), followed by a PUF (for vapor-phase PBDEs), and the second
sampler included a filter followed by an XAD-2 resin. Four sampling units, two of each,
took two samples per day on two consecutive days, 3/17/2004 and 3/18/2004, providing a
total of 8 samples. Statistical analysis suggested the two measured comparably, except
that the filter/PUF setup could not measure BDE 209 due to analytical difficulties. A
total of 33 congeners were measured, making it the most substantial air study found in the
literature. Total BDE concentration outdoors in this test averaged 38 pg/m3 for the
filter/PUF setup and 93 pg/m3 for the filter/XAD setup where BDE 209 made up 10
pg/m3 of that total. BDEs 47 and 99 dominated the profiles, averaging 34.6 and 12.5
pg/m3, respectively, between the two sampling units.
Hoh and Hites (2005) took air samples every 12 days at 5 locations from Lake
Michigan to the Gulf of Mexico, between 8/2002 and 12/2003, covering most (but not
all) months of the year. One site was urban, Chicago, two sites were in remote locations
in Michigan and Louisiana, one site was in an agricultural region, and one was in the
university city of Bloomington, Indiana. Although numerous congener data were
evaluated, individual data was only available for 47, 99, 100, and 209 (these 4
contributed about 80% of the total PBDE). The highest congener individually was 47,
averaging between 7 and 17 pg/m3 over all sites, while the other congeners mostly
averaged under 7 pg/m3. The key exception was BDE 209, which was important in
Chicago, averaging 60 pg/m^ while it was 9 pg/m3 or less in other sites. Overall, the total
PBDE concentration at the Chicago site was the highest, averaging 100 pg/nr; the
concentrations at the other four sites were comparable, averaging under 30 pg/m3.
Strandberg et al. (2001) monitored BDEs at four sampling sites include downtown
Chicago (urban site), Sleeping Bear Dunes, MI (rural site), Sturgeon Point, NY (rural),
and Eagle Harbor, MI (remote). Four samples per year (taken May through October) for
each of the four sites and for three years (1997, 1998, and 1999) resulted in a total of 48
samples. BDEs included 47, 99, 100, 153, 154, 190, and 209. The average total BDE for
Chicago was about 50 pg/m3 over 3 years, while it was about 5-15 pg/m3 for the other
three sites. BDE 209 was not detected at the 3 rural/remote sites, and detected at levels
4-16
-------
only at 0,3 pg/m3 in Chicago, This is in stark contrast to the monitoring of Hoh and I lues
(2005), which showed an average of 60 pg/nr in 2002/2003. This could likely be due to
the fact that Strandberg et al. (2001) did their measurements between 1997 and 1999,
prior to the time when the penta and octa formulations were taken off the United States
market, and the market became dominated by the deea formulation (which contains 97%
BDE 209). Also, Strandberg et al. (2001) note analytical difficulties with BDE 209
(degradation during analysis, low recovery) and discusses other possibilities for low
findings in air (low mobility from source, etc); Hoh and Hites (2005) notes much higher
BDE 209 in their data and attributes it to greater usage of decaBDE, not analytical,
problems. Gas/particle partitioning suggests 80% of BDE 47 was in the vapor phase, 55-
65% in the vapor phase for BDEs 99 and 100, and only about 30% in vapor phase for
BDEs 153 and 154.
This sampling network reported on by Strandberg et al. (2001) and Hoh and Hites
(2005) continued to be monitored during 2005 and 2006, and PBDE results from that
effort are reported on by Venier and Hites (2008). The total PBDE concentrations were
highest at the "urban" sites of Chicago and Cleveland, with mean concentrations of 65
and 87 pg/m3, respectively, with the higher concentrations at Cleveland due to the
presence of higher concentrations of BDE 209 in several samples. The mean
concentrations at the "rural" sites of Sturgeon Point and Sleeping Bear Dunes were 9.2
and 8.1 pg/m3, respectively, and the mean concentration at the "remote" site of Eagle
Harbor was 5.8 pg/m3. The authors also studied the time trends of BDE 47, 99, and 209
from data starting in 2003, and observed that BDEs 47 and 99 were declining rapidly, but
that BDE 209 was not declining at any of the 5 sites.
Three sites within the Cheasapeake Bay area were sampled for BDEs as part of a
larger effort at air quality monitoring in the region (Goel et al., 2006). A total of 240
samples were taken between 2001 and 2003, using high volume samplers to measure both
particle and gas phases for BDE 47, 99, 100, and 154. BDEs were detected in 75% of the
samples, but detection frequency was highest at Lewes, 98%, compared to Dover, 77%,
and Horn Point, 52%. Gaseous phase concentrations were presented and were highest at
Lewes, with geometric mean concentrations over the 3-year period of 174 pg/m3,
followed by Dover and Horn Point at 19 and 10 pg/nr, respectively. It is not clear why
4-17
-------
particle phase concentrations were not presented in Goel et al. (2006). except that they
were much less frequently found (<5% of samples from Horn Point and Dover), but at
40-50% of samples at Lewes. BDE 47 was the highest found in all three sites, from
about twice as high as 99 and 100, in Horn Point and Dover (10-20 pg/nr for BDE 47
compared to 5-8 pg/nr for BDEs 99 and 100 as geometric means), to about 7-10 times as
high at Lewes (175 pg/m3 for BDE 47 compared to 17-26 pg/nr for BDEs 99 and 100).
BDE 154 was essentially not detected in either gaseous or particle phases. The finding at
Lewes was attributed to use of spray irrigation of municipal wastewater near the
sampling site.
One of the CARB sites involved outdoor sampling near a possible source of
PBDF: release—near an autoshredder. On this site, there were 4 samplers, including 2
upwind and 2 downwind, sampling for 3 days, resulting in 12 samples. There were no
field activities during the first day of sampling, and the results were lower that day when
compared to days afterward—when the facility was operating. Another key observation
was that samples were much higher downwind compared to upwind. However, even the
upwind sites were higher than the UC Davis site discussed earlier, suggesting that there
was offgasing from previously deposited BDEs from the nearby source. The highest
measurement was for BDE 209, averaging over 2,400 pg/nr'1 for 4 sampling dates. Only
one other sample had at least one reading above 100 pg/m1, and that was for BDE 99,
which had the second highest readings overall, with an average of 165 pg/m3 over the 9
days. BDE 47 had the next highest, averaging 64 pg/m3.
A second CARB site was near another source: an electronics recycling plant.
Three days of outdoor sampling with 4 samplers (2 in front and 2 in back) for a total of
12 samples, measured the impact of the release of PBDEs from electronics within this
recycling facility. Samplers located in front, near the loading dock, which was open most
of the time, had higher concentrations as compared to the back two samplers. BDE 209
was the highest congener measured, averaging 2,764 pg/m3, with the next highest BDE
183 averaging 116, and then BDE 47 averaging 70 pg/m3. The interesting thing about
BDE 209, which exists predominantly in the particle form, is that the concentrations were
40-fold higher in the front samplers, near the open doors to the indoor recycling facility.
This suggests that near a source (such as near the front doors in a recycling facility where
4-18
-------
BDE 209 could be released), BDE 209 concentrations can be extremely elevated, but far
from a source (in the back of the building), concentrations are much lower.
Another study evaluating outdoor air impacts near a source occurred in the United
Kingdom. Air concentrations before and after a major bonfire were analyzed for both
PAHs and PBDEs in November of 2000 (Farrar, et al., 2004). "Bonfire Festival" in the
United Kingdom occurs every year on November 5. Three samplers were set up in the
garden of a home in a residential area. They were not in the immediate vicinity of any
public bonfires, but there were bonfires known to be occurring in the general area. Daily
samples were taken from November 1 to November 13, and the evidence clearly showed
a rise in concentration on the 5th, with background levels on the 4th and before and from
the 6th onward. A total of 21 BDE congeners were measured, although BDE 209 was not
measured. Background concentrations were only about 4 pg/m3, with quantifiable
measurements of BDEs 47 (2 pg/m3), BDE 99 (1.5 pg/m3), and BDE 100 (0.5 pg/m3).
This clean air was attributed to air originating over the Atlantic Ocean. Concentrations
reached 95 pg/nr on November 5, with the highest concentrations of BDEs 99 (14
pg/m3), BDE 153 (13 pg/nr), BDE 154 (10 pg/m3), BDE 166 (10 pg/m3), BDE 47 (8
pg/m-1), BDE 49 (8 pg/m3), and similar concentrations of 4 pg/nr for BDEs 66, 85, 100,
181, and 190. Likely sources of the PBDEs were the burning of discarded clothing or
furniture, with temperatures not hot enough to destroy the PBDEs. No explanations were
provided for the predominance of BDE 153 and 154, which is different from the more
typical predominance of BDEs 47 and 99 (as seen in background air).
Wilford et al. (2008) sampled air at a semi-rural site in northwest England in
April to May of 2004. Their analysis focused almost exclusively on characterizing the
particulate phase of BDEs, with a particular emphasis on BDE 209 and other higher
brominated congeners. They did measure for the vapor phase using a polyurethane foam
(PUF) sampler for a 7-day period, and found very low levels of BDE 209 in a limited
number of samples. More importantly, they stated that the vapor phase was dominated
by tri- to hexa-BDE congeners, and that these congeners were found at vapor-phase
concentrations at least two orders of magnitude higher than in the particle-phase.
However, they did not provide any results for these vapor-phase concentrations, or for
total (vapor+particle) phase concentrations. Their mean and median concentration of
4-19
-------
total BDEs in the particle phase over 28 samples was 41 and 18 pg/m3, respectively. The
mean and median BDE 209 concentrations were 20 and 13 pg/m \ respectively, showing
the dominance of this congener in the particle-phase profile. In tact, the sum of hepta
thru deca-eongeners, or more specifically, the sum of BDEs 183, 196, 197, 206, 207. 208,
and 209, explained over 90% of the total profile over all samples. Concentrations of the
lower brominatcd congeners in the particle phase were mostly less than 1 pg/nr\ BDE
183 was found in 10 of the 28 samples, at a mean of 4.6 pg/m\ at a high of 92 pg/m'.
Jaward et al. (2004, 2005) have provided a comprehensive sampling of outdoor
air in both Asia (Jaward et al., 2005) and Europe (Jaward ct al., 2004). They used
polyurethane foam (PUF) disk samplers in a passive mode, meaning they leave the PUF
exposed to outdoor air to capture PBDEs for a period of 6 weeks, then analyze the PUFs.
Two drawbacks from this approach are that it purports to capture only the vapor phase
BDEs (it was stated that the PUF captures some particles, but the amount on particles was
hard to ascertain), and, secondly, that BDE 209 was not analyzed in their studies. In their
Asian study, Jaward et al. (2005) employed PUF samplers simultaneously at 77 sites and
measured the air between Sep. 21 and Nov 16, 2004. Rural and urban sites were sampled
in China (32 samples), Japan (20), South Korea (15), and Singapore (10). BDEs
measured include 17, 28, 32, 47, 49, 75, 99, and 100. Total BDEs ranged from 0.1 to
tens of pg/m3, which the authors note is consistent with measurements for the remote
coast of Ireland and rural/semirural England. However, some samples in China measured
up to 340 pg/m3 in an industrial city known to manufacture electronics, which is similar
to values reported in urban United Kingodm where PBDE usage is known to be high.
Mean or median concentrations were not provided. Only ranges were given and, as such,
trends between congeners cannot be described. With a DL of 0,13 pg/m3, maximums
found include (in units of pg/m3) the following: BDE 17: 35 in China, <1.7 otherwise:
BDE 28: 130 in China, 6, 52, and 2.6 otherwise; BDE 32: 13 in China, <1.2 otherwise;
BDE 47: 78 in China, <10 otherwise; BDE 49: 48 in China, <3 otherwise, BDE 75: 13,
12, 19, 1.2 in four locations; BDE 99: 50 in China, <10 otherwise; BDE 100: 5.5 in
China, <2.3 otherwise. It was stated that 47 and 99 dominated the profile, contributing
around 75% to the overall BDE burden. A similar approach was taken in Europe: 71
PUF samplers were deployed in 22 European countries, including 25 in urban locations
4-20
-------
and 46 in rural/remote locations. BDEs measured and the results, percent detected in
parentheses, and ranges, pg/nv\ are as follows: BDE 28 (82%): <0.5-30; BDE 47 (55%):
<8-80; BDE 49 (30): <0.5-12; BDE 75 (54%): <0.5-3; BDE 99 (45%): <10-120; BDE
100 (41%), <2-20; BDE 153 (55%): <0.7-15; BDE 154 (44%,): <0.8-10. Generally there
was over a 700 fold ratio between high and low total measurements, with the highest
being in urban locations in the United Kingdom and the lowest in remote areas of
Iceland, Ireland, Norway, and Sweden. Generally low levels were found in Eastern
Europe. As in Asia, BDEs 47 and 99 constituted about 75% of the total PBDEs.
A study measuring BDEs in China suggested much higher concentrations
compared to the measurements of Jaward et al. (2005) discussed above. Chen et al.
( 2006) reported on results from 32 pairs of samples collected from four sites—2
industrial, 1 urban, 1 background, in China during June 15-30, 2004. High volume
PUF/GFF PS-1 samplers were used and detection limits were 0.14-0.58 pg/mJ for BDEs
28, 47, 66, 85, 99, 100, 138, 153, 154, and 183, and was 14.3 pg/m3 for 209. BDE 209
dominated the profile, averaging 4,200, 750, 264, and 478 pg/m' for the two industrial,
urban, and background setting, while the sum of the other PBDEs correspondingly
averaged 3,673, 230, 89, and 105 pg/m3. The results appeared to be correlated to wind
patterns: when blowing in from the west where there were several big electronic markets
(where computers are assembled and dismantled), BDE 209 used in circuitry was very
dominant; whereas when wind blew in from the southeast, directly from industrial areas,
the sum of the other BDEs dominated. Congener patterns in one of the industrial sites,
the urban sites, and the background sites suggests nearby use of the commercial
formulation, with the high readings of BDE 209, and the nearby use of the commercial
penta formulation, with BDEs 47 and 99 comprising greater than 50% of the total.
In summary, the major trends seen from the air monitoring studies include: 1)
outdoor concentrations measured in the United States tended to be in the range of 20-200
pg/nr for total BDEs; 2) the profile was dominated by BDEs 47 and 99, suggesting a
penta formulation influence; however, when BDE 209 was measured , it was seen to have
concentrations equal or greater than 47 and 99; one study in Chicago had BDE 209 at
concentrations 6 times higher than 47 or 99, and a study in China also showed a
dominance of BDE 209 in air profiles of industrial, urban, and background settings; 3)
4-21
-------
concentrations were found to be higher in industrial/urban sett ings as compared to
rural/background settings. This is consistent with the expectation that concentrations
should be higher nearer to where BDEs are released from their use in commercial
products - such as near a source burning such products (like the UK bonfire described
above) or shredding/recycling such products {like the autoshredding and recycling sites
studied by CARB and described above); and 4) comparisons with measurements in
Europe and Asia are hard to make-—one study from China suggests airborne
concentrations well into the thousands, while other measurements in Europe and Asia
appear more in line with those in the United States with total BDEs mostly under 100
pg/m3.
4.5.2, Indoor Air and Simultaneous Indoor/Outdoor Monitoring
The primary sources of PBDEs to which the general population is exposed arc the
products in which they are used. This would suggest that indoor exposures would be of
primary importance for this class of PBTs, and, specifically, indoor air might contain
higher concentrations of PBDEs as compared to outdoor air. Because the primary
sources tend to be indoors, several researchers have focused their efforts in the indoor
environment, and their efforts include simultaneous indoor/outdoor air measurements. In
all these studies, the indoor measurements are higher than nearby outdoor measurements,
often by factors of 10 or more. There is some evidence that combustion of products
containing PBDEs release the PBDEs, but the evidence for this was primarily an open-air
bonfire, not a controlled waste combustion process (see discussion above on the United
Kingdom bonfire study by Farrar et al., 2004). There is also evidence that combustion of
products containing PBDEs can result in formation of polybrominated dibenzo-jw-dioxins
and dibenzofurans (PBDD/Fs), and monobromo-polyehlorinated dibenzo-/?-dioxins and
dibenzofurans (MoBPXDD/Fs). Hayakawa et al (2004) found that the levels of
MoBPXDD/Fs correlated positively with that of polychlorinated dibenzo-/>dioxins and
dibenzofurans (PCDD/Fs) in Japanese ambient air testing, and cited other laboratory
testing showing concurrent formation of PCDD/Fs and MoBPXDD/Fs (Sakai et al,
2001). They also showed a correlation between PBDD/Fs and some key PBDEs, citing
other work showing formation for PBDFs from combustion of materials containing
4-22
-------
decaBDE (Lenoir ct al, 1994). So, while the combustion of PBDEs appears to produce
other brominated compounds, unlike dioxins, PBDEs are not formed themselves from
combustion.
Unfortunately, United States researchers have not measured indoor air as much as
indoor dust, or as much as researchers from Canada and abroad. Only two studies have
been located which measured indoor air: The first was an indoor air study in which dust
were sampled in 20 urban residences in the Boston, MA. with the air data reported in
Allen et al. (2007). At each home, one personal and two area samples were taken over
the course of a week, presumably during 2005 (the date of sampling was not provided but
the data was reported in 2006). BDEs include 17, 28/33, 47, 49, 66, 85/155, 99, 100,
153, 154, and 209. The area samples were taken in the bedroom and living room, and the
personal air sample was taken near the breathing zone in the bedroom. Participants
turned on the samplers when returning in the evening, and turned them off when leaving
to work. The sum of the geometric mean concentrations of repotted congeners over the
20 residences were 605, 392, and 366 pg/nr for the personal, bedroom, and living room
air. BDE 47 had the highest concentration, with geometric means of 227, 158, and 145
pg/nr for the same three locations, with BDE 209 second highest at 174, 95, and 94
pg/nr, and BDE 99 third highest at 1 11, 67, and 60 pg/nr'.
The second United States study found for indoor air measurement of PBDEs was
conducted by CARJB (2005). Air monitoring was conducted indoors within a computer
training facility in a public office building, within and around an electronics recycling
facility, and outside of an automotive shredding/metal recycling facility (the recycling
and shredding sites were also described in the previous section) on the UC Davis campus.
This was the most comprehensive single data set that could be found in the literature,
with these four sites and the measurement of 33 congeners. The octa- and nona-BDEs
were not measured, however, for most of the samples because there was analytical
intereference. The sampling for the indoor training facility included sampling when the
computers were on versus q//'and also included carpet dust samples. The dominant
congener found in the air was BDE 47, followed by BDEs 99, 100, and 28. In all cases,
BDEs were higher in air when the computers were on versus off, but the most meaningful
difference was for BDE 209. when considered in terms of the relative difference between
4-23
-------
off and on, rather than on the magnitude of the concentration alone. Specifically, BDE
209 increased from 2 pg/nv' when the computers were off to concentrations from 56-74
pg/nr when they were on in one set of tests and from 18, when off to 47-56 pg/nr, when
on, in the second test. The largest absolute concentration difference was for BDE 47,
which was 21.3 and 112 pg/m3 in the two tests (starting from concentrations in the 800
pg/nr range in the off condition and then increasing by 213 and 112 pg/m3 when the
computers were turned on), followed by BDEs 99, 100, and then 209. The average
concentrations of total BDEs, on (3 measurements) averaged with off (I measurement) on
the 2 days of measurement was 1,550 and 2,010 pg/nr, with BDE 209 averaging 50 and
65 pg/nr1 for the two days.
A second site studied by ( ARB was an electronics recycling facility. Outdoor
measurements from 2 samplers in front and 2 in back suggested concentrations in the
range of 2,000-3,000 pg/nr1, dominated by BDE 209. Generally, sampling indoors
showed substantially higher BDE concentrations compared to outdoors at this facility.
Two samplers sampled 3 days consecutively in 2004 for a total of 6 samples.
Concentrations ranged from 316,000 to 833,000 pg/nr. The overwhelmingly dominant
congener was BDE 209, ranging from 79,000 to 833,000 pg/nr, with an average over
423,000 pg/m3. The next highest congeners were BDE 183, averaging 23,813 pg/m3, 153
averaging over 5,600 pg/m"\ 154 averaging over 3,400 pg/nr', and then 99, 47, and 49, all
about at 1,770 pg/m3. Measurements of the filter and XAD-2 separately showed
comparable amounts in vapor (XAD-2) as compared to particle (filter) for the tri-BDEs,
about 10 times more particle than vapor for the tetra-BDEs, and then essentially none in
the vapor phase for penta-, hexa-. hepta-, and marginal amounts for deca BDE congeners.
Only the sampling within the computer training facility showing total
concentrations ranging from 1,500-2,000 pg/nr' might represent non-occupational
concentrations that could apply to the general public. However, these concentrations are
high when compared to other indoor monitoring studies conducted in Canada and abroad.
Shoeib et al. (2004) reported on measurements from 10 indoor and 3 outdoor
samplers in Toronto, Canada, using a traditional high-volume, two-phase air PS-1 air
sampler. Samples were collected in Nov/Dec of 2001 and then again in March of 2003.
BDEs congeners included 17, 28/33, 47, 85, 99, 100, 153, and 154. Total concentrations
4-24
-------
outdoors were 39 and 48 pg/nr (2 locations, 3 samples) and indoors were 410, 358, 490,
2,088, 381, and 76 pg/m3 (6 locations, 7 samples), leading to an average ratio difference
of 15 between indoor and outdoor samples, BDE 47 represented about 46% of the
sample, followed by 99 at 25%. Both the Junge-Pankow model (Pankow, 1987) and the
model based on the octanol air partition coefficient, the Koa model (Shoeib and Harrier,
2002) very well predicted the gas/particle partitioning that was observed in the high-
volume sampler. Observed particle phase percentages for the congeners, derived from
the indoor sampler, from high to low arc as follows: BDE 183-82%, BDE 154-82%,
BDE 153-81%, BDE 85-75%, BDE 99-62%, BDE 66-24%, 47-20%, BDEs 28/33-4%,
and BDE 17-3%.
Wilford et al. (2004) sampled air using a PUF disk as a passive sampler in 74
homes and at 7 outdoor sites during the winter of 2002/3 in Ottowa, Canada. Indoor air
concentrations of PBDEs were log-normally distributed with a geometric mean of 120
pg/nr (high of 3,600 pg/m3), which is approximately 50 times higher than outdoor air
concentrations, <0.1-4.4 pg/m1. Congeners measured include 17, 28, 47, 66, 71, 85, 99,
100, 153, and 154. The highest mean concentration was 47 at 160 pg/nr, followed by 99
at 42 pg/nr', 28 at 24 pg/nr\ and so on. Wilford et al. (2004) stated that the indoor
passive samplers are sampling mainly the gas phase with only a small contribution from
particulates. It was found that the technical formulations tend to be enriched with heavier
congeners, 99 and 100, while the indoor air was dominated by the lighter more volatile
congeners, 47 and 28.
Limited studies abroad (United Kingdom and Kuwait) showed indoor
concentrations of total BDEs similar to these Canadian studies suggesting a range of 20-
200 pg/nr total. However, like the Canadian studies, measurements were not made for
BDE 209. Hazrati and Harrad (2005) passively sampled 12 homes, 10 offices, and 1
private car in the United Kingdom for a period of one year, for sa?npling events which
took between 4 and 6 weeks. BDE congeners sampled include 28, 47, 49, 66, 85, 99,
100, 153, and 154. Concentrations ranged between 5 and 1,418 pg/nr\ with a mean of
148.2 and median of 38.4 pg/m". In comparison with outdoor air from other studies, the
authors suggested a 20-fold difference between indoor and outdoor air. PUF samplers
were used to measure air samples and bulk dust samples were obtained from vacuum
4-25
-------
bags in 17 homes in Kuwait between February 29, 2004 and April 11, 2004 (Cievao et al.,
2005). Individual congener concentrations included 28, 47, 100, 99, 85, 154, 153, and
183, Air concentrations ranged from 2,5-385 pg/nr, with a geometric mean of 10 pg/m3.
BDE 47 was the most abundant congener representing, on average, 51 % of the total
PBDE concentration measured. The next most abundant congener, BDE 99, represented
about 28% of the total .
Harrad et al. (2004) presented BDE congener data on 47, 99, 100, 153, and 154
for indoor and outdoor air, and meat and vegan diets. The purpose of these
measurements was to estimate daily exposure to BDEs via inhalation and diet. Indoor air
concentrations were much higher than outdoor air concentrations; BDEs 153 and 154
were comparable in both environments, but 47, 99, and 100 were each over 100 times
higher indoors than outdoors. The mean total concentrations of BDEs in air in various
environments include the following: 21 pg/m3 in outdoor air, 525 pg/m3 in domestic
indoor air, and 2,788 pg/m3 in workplace environments.
Harrad ct al. (2006) collected indoor air samples at 92 locations in the United
Kingdom, including 31 homes, 33 offices, 25 cars, and 3 public locations (post office,
coffee shop, and supermarket) using PUFs. They measured BDEs 28, 47, 49, 66, 85, 99,
100, 153, and 154. The overall average of all measurements was 273 pg/nr' total, with a
median of 47 pg/m3 and a maximum of 8,180 pg/m3 found in cars. Cars had the highest
average concentration, although that was skewed by three measurements greater than
2,000 pg/m3; the highest concentration in the non-car environments was 1,416 pg/m3 and
the average excluding the cars was 110 pg/m3. Like other studies, the major contributors
to these concentrations were BDEs 47 and 99.
Karlsson et al. (2007) report on measurements of PBDEs in indoor air, housedust,
and blood from five households, although air was not sampled in one of the households.
Unlike some of these other studies, BDE 209 was measured. This congener was found in
1 of 4 samples, at 257 pg/mJ (DL = 173 pg/m3), BDE 47 was found in 3 of 4 samples,
ranging from 126 to 171 pg/m3. The only other two congeners quantified in air were
BDEs 28 and 66, and they were found in 7 of 8 measurements ranging from 6 to 28
pg/m3.
4-26
-------
Mandalakis et al. (2008) measured the levels of PBDEs inside automobile cabins
in Greece. They measured the indoor air of 31 cars, taking 41 samples from low-volume
samplers which were on for 48 hours, both when the car was in use and when not in use.
The median concentration of total PBDEs was 201 pg/m3, dominated by BDE 209,
explaining about half the total concentration. Seventeen other congeners were measured,
with the next two predominant congeners being BDEs 47 and 99. The maximum found
was over an order of magnitude higher than this median at 2644 pg/nr.
In summary, only one study could be located which measured indoor air
concentrations of BDEs in the United States in residences rather than in occupational
settings. This was the study of urban residences in the Boston, MA, area, where the
average total concentration was in the range of 200-500 pg/m"' total. The only other
indoor measurements came from within a computer lab, and total concentrations were in
the range of 1.500-2,000 pg/m3, with BDE 209 in the range of 50-70 pg/m3. This was
higher than indoor concentrations measured in Canada and abroad, which were mostly
less than 200 pg/nr1 total, although BDE 209 was most often not measured in studies in
Canada or abroad. Indoor industrial/occupational concentrations were substantially
higher, with measurements in the hundreds of thousands of pg/m3. In United States,
Canadian, and studies abroad, indoor concentrations were found to be higher than
outdoor concentrations in simultaneous measurements by factors of 10 to 100. BDE 47
appears to dominate indoor air concentrations, encompassing about one-third the total
congener concentration when BDE 209 was measured in one study (Allen, et al., 2007),
and about half the total concentration in other studies when it was not measured. In the
one study where BDE 209 was measured in the indoor environment, it was the second
highest concentration, followed by BDE 99. In one study of PBDE concentrations inside
cars, BDE 209 did dominate the profile explaining about half the total concentration.
4.6. FISH CONCENTRATIONS
This section reviews the data on fish, including fish caught in the wild, fanned
fish, and fish samples from market basket surveys. While emphasis is on the United
States studies, noteworthy studies from Canada and abroad are included as well. Like
much of the data on other environmental levels, these studies often did not include BDE
4-27
-------
183 and 209, the primary markers for the octa and deca formulations, respectively.
Another issue for evaluation and comparison offish studies is that some results are
reported on a wet weight of whole tissue and others on a lipid-basis. Like other PBTs,
PBDEs bioaccumulate in lipids of animals, so it would be useful to have all data reported
on both a wet weight and a lipid-basis. While the author's reported results are provided
below, lipid- or wet-based concentrations are concurrently provided in parenthesis when
possible. Table 4.3 provides congener-specific fish concentrations for fish caught in the
United States.
4.6.1. Farmed Fish Concentrations
A total of 70 farmed and wild salmon were collected from wholesale and retail
outlets in Maine in August 2003 and May 2004 (Shaw et al., 2005). They represented
salmon from three regions: two farms in eastern Maine, three in eastern Canada, and one
in Norway. Samples were composited so that the Maine sample results were displayed
for the two farms. Results were provided for 9 congeners including 28, 47, 66, 85, 99,
100, 138, 153, and 154. Samples were analyzed with skin on and skin off to see if that
made a difference, and results suggested essentially no difference. There was also not a
correlation between lipid content and fresh weight concentrations, which is counter-
intuitive. It was found that levels, in the neighborhood of 1 ng/g total wet weight (wwt)
basis, were 4-5 times lower than PBDE concentrations reported in farmed salmon from
British Columbia and northwestern Europe, but comparable to levels found in farmed
salmon from Chile.
Jacobs et al. (2002) measured PCBs, DDT, and PBDEs in farmed and wild
European Atlantic salmon, aquaculture feeds, and fish oils used to supplement the feed.
Seven British salmon samples, 5 additional salmon samples (two from Ireland, and 3
purchased from a Belgian market), 8 salmon feeds (from 4 different Scottish sources), 5
fish oils, and 1 vegetable oil were analyzed. BDE congeners analyzed include 28, 71, 47,
75, 66, 100, 99, 1.53, and 154. Total BDEs ranged from 1.1-85.2 ng/g lipid weight (lwt)
in 13 salmon samples (average = 33,8), with the highest found in a wild salmon sample.
BDE 47 predominated, averaging about 53% of total. The levels of BDE in feed ranged
4-28
-------
from 8.1 to 23.9 ng/g Iwt for 8 feed samples, and the range in fish oil was ND to 12.7
ng/g lwt. BDE 47 similarly dominated the feed and fish oil samples.
The most comprehensive study on BDEs in farmed fish was conducted by Hitcs et
al. (2004). PBDEs were measured in about 700 farmed and wild salmon collected from
around the world, including from Maine and Washington, at the AXYS Analytical Lab in
Sidney, BC. PBDE congeners 1, 2, 3, 7, 8, 10, 1 1, 12, 13, 15, 17, 25, 28, 30, 32, 33, 35,
37, 47, 49, 66, 71, 75, 77, 85, 99, 100, 105, 116, 119, 126, 138, 140, 153, 154, 155, 166,
191,183, 190, 206, 207, and 208 were found. BDE 209 was looked for but not detected
with a detection limit of 0.1 ng/g wwt (other congeners had DLs at 0,001 to 0.01 ng/g
wwt). The concentrations of total PBDEs in these samples ranged up to 10 ng/g wwt. In
order of magnitude, the samples suggested the following: Europe farmed salmon > North
America fanned salmon > Chile farmed salmon > wild salmon. Farmed salmon ranged
from around 0.3 up to nearly 10 ng/g wwt, while the bulk of wild salmon w ere less than
0.3 ng/g wwt. In terms of wild salmon species, Chinook was the highest, with an average
of 2.26 ng/g wwt for the 9 samples as compared to 0.13 ng/g wwt for the 36 others. In
terms of composition. BDE 47 dominates (around 47%), followed by BDE 99 (18%),
BDE 100 (10%), BDEs 28/33 and 49 (5%), with other congeners under 5%. Although
not found in the fanned salmon, BDE 209 was found in the salmon feed and at about
15% of total concentration. Total PBDE concentrations in 13 feed samples ranged from
0.5 to 10.9 ng/g wwt.
Hayward et al. (2006) collected 18 samples of farmed and wild caught fish from
supermarkets in Maryland or Washington, DC in 2004, and an additional 4 samples were
collected in 2001, including one from North Carolina. There were 6 farm-raised salmon,
6 wild caught salmon, 5 bluefish, and 5 rockfish among the 22 samples. They were
analyzed for 22 BDEs, including 209. but only BDEs 28, 47, 49, 99, 100, 153, and 154
could be found routinely. Individual congener results were graphed and not provided in a
table, so they were not added to Table 4.3. However, they appear consistent with other
measurements described here and in Table 4.3. Wild bluefish had the highest total PBDE
concentration averaging 15.1 ng/g wwt (n =¦ 5), followed by wild rockfish at 5.4 ng/g wwt
(n = 5), fanned salmon at 1.0 ng/g wwt (n = 6), and Alaskan salmon (King, Coho,
Sockeye) at 0.4 ng/g wwt (n = 6). BDE 47 had the highest concentrations, explaining
4-29
-------
55-70% of the total, followed by 99 and 100, both of which took turns being second
behind 47 in different samples, BDE 153 was not found in wild salmon and only found
at 1-3% of total in other species, and BDE 183 was found in only one sample.
4.6.2, Fresh Water and Marine Fish
Oros et al. (2005) measured 22 BDE congeners but detected only BDEs 47, 99,
and 100 in bivalves (clams, mussels, and oysters) in composited samples originating from
16 locations, at wet weight concentrations with an average of 5.7 ng/g wwt, ranging from
2-13 ng/g wwt (855—13,502 ng/g Iwt) in the San Francisco Estuary. The BDE congener
found most abundantly was 47; it is found in all samples above its MDL and comprising
about 50-70% of the total concentration, while 99 and 100 were found in about 50%of the
samples.
Manchester-Neesvig et al. (2001) sampled 21 coho and Chinook salmon from
Lake Michigan tributaries in 1996 and analyzed them for 6 PBDEs: 47. 66, 99, 1.00, 153,
and 154. These samples have among the highest fish (and terrestrial animal)
concentrations in the literature, ranging from 45 to 148 ng/g wwt (773 to 8,120 ng/g lwt),
with an average of 80.1 ng/g wwt (2,440 ng/g lwt). BDE 47 dominated the
concentration, comprising 56% of the total, followed by BDE 99 at 19%, BDE 100 at
12%, BDE 154 at 6.6%, and BDE 153 at 3.6%. The concentrations of PCBs tracked very
well with PBDEs, suggesting that PBDEs have been part of Lake Michigan for many
years, like PCBs.
A more comprehensive evaluation of Great Lakes fish was conducted by Zhu and
Hites (2004), Lake trout collected between 1980 and 2000 from Lakes Superior,
Michigan, Huron, and Ontario, and walleye from Lake Erie, were analyzed for 15
PBDEs, including 17, 28, 47, 49, 66, 71, 85, 99, 100, 138, 153, 154, 183, 190, and 209.
The levels of BDE 209 found in all fish samples was in the undetectable range of 3.6 ng/g
lwt, thus it was deemed to be ND in all samples and not reported. Total PBDEs (sum of
BDEs 47, 99, 100, 153, and 154 only) rose from under 15 ng/g lwt in 1980 to the range of
58-180 ng/g lwt in 1990, and finally in the range of 400-1,400 ng/g lwt in 2000. The
authors concluded that the trends suggest a doubling time of 3-4 years. The authors
discussed some spatial trends, but mostly the concentrations in the lakes over time were
4-30
-------
comparable. The summary information is the average congener concentration over all
lakes for the most recent year, 2000. There was a systematic change of congener
distribution over time. More of the octa-BDE product, which was high in 153 and 154.
was used in the 1980s, and there was a higher proportion of these two in the 1984 fish as
compared to the 1996 fish. The ratio of penta-BDE use divided by octa-BDE in the mid-
1980s was 0.7 (more octa used), but by the mid-1990s, there was a shift and the ratio was
now 2 (twice as much penta used). Subsequently, fish in the latter 1990s were dominated
by BDE 47.
Battcrman et al. (2007) also conducted a comprehensive temporal study of
archived and fresh fish from the Great Lakes. Chapter 3 summarized this study as it
measured numerous archived and freshly caught fish from all of the Great Lakes from
1979 to the present. Batterman et al. (2007) focused only on BDEs 47, 99, 100, and 153,
and found consistent increases with doubling times in the range of 2-4 years. The total
concentrations (sum of the four congeners) in trout ranged from about 20 to over 100
ng/g fw in the last samples taken, between 2000 and 2005. Concentrations were a bit
lower in the last samples of smelt taken, between 3 and 81 ng/g fw. BDE 47 dominated
the total, contributing about 70% of total concentration, with BDEs 99 and 100
contributing about 13% each.
Although the focus was on analog compounds (which are defined as compounds
structurally analogous to PBDEs, such as hydroxylatcd PBDEs, or OH-PBDEs), 6 BDEs
including 47. 99, 100, 153, 154. and 183 were measured in the plasma of fish from the
Detroit River (Valters et al., 2005). Composited plasma samples (blood was eentrituged
to separate red blood cells from plasma) were obtained from 13 fish species. Total
PBDEs ranged from 0.16 to > 21.0 ng/g wwt (in channel catfish; all other samples less
than 10.0 ng/g wwt). Given that plasma was 0.6-2.7% lipids, lipid concentrations would
be much higher, at over 200 ng/g lwt. BDE 47 was the dominant congener (over 60% in
most samples), followed by 99 and 100. which were of similar magnitude; then 153 and
154 at lower and sometimes non-detect values. BDE 183 was only detected in channel
catfish at 0.805 ng/g wwt; BDE 47 was 11.5 ng/g wwt in this catfish.
A total of 63 samples of fish were collected from Washington State rivers and
lakes by the Washington State Department of Ecology in 2005 and 2006 (Johnson et al,
4-31
-------
2006). These samples were measured for 13 congeners including 47, 49, 66, 71, 99, 100,
138, 153, 154, 183, 184, 191, and 209. Results for total ranged from ND to 1,059 ng/g
wwt, with a median of 2.8 ng/g wwt and a mean of 35 ng/g wwt (derived assuming ND =
0). The most frequently detected congener was BDE 47; it was detected 84% of the time
with a median concentration of 1.5 ng/g wwt. It accounted for 68% of the total
concentration found, with BDE 99 accounting for 16% as the second highest
concentration. Other congeners detected frequently equal BDEs 100, 154, 153, and 99,
detected 51, 49, 40, and 38% of the time. BDE 209 was not detected, but it had a high
detection limit ranging from 1-6 ng/g wwt (all other detection limits were less than 0.5
ng/g wwt). The Spokane River was unambiguously the most impacted river, with 3
samples taken measuring 76, 417, and the survey maximum, of 1,059 ng/g wwt. Not
ironically, this river also had the highest water concentrations in this same survey, with
two measurements above 100 pg/L at 146 and 926 pg/L, with all other water
measurements well under 100 pg/L. Johnson et al. (2006) report on an earlier study by
the Washington State Department of Ecology on fish in the Spokane River, and they
similarly found high concentrations, ranging from 30 to 1,222 ng/g wwt in 47 samples
(some of which were composites).
Fish from three locations in the Savannah River were sampled in 2005 and
analyzed for 12 BDEs (Sajwan et al.. 2006). Individual congener results were not
provided, and results suggested generally higher levels of BDEs, similar to other fresh
water bodies, with totals at 10 to >300 ng/g hvt (which would equal about 1/10 as much
on a fresh weight basis). The interesting finding from this study was that the
predominant congener was identified as BDE 30 in all three locations, a congener which
not looked for in any other fish study. The second most predominant congeners were the
familiar BDEs 47 and 99.
Ashley et al. (2006) analyzed samples of eel and sediment from the Delaware
River (sediment samples described above). The eel samples were collected in 1998 as
part of an earlier effort focusing on PCBs. Eel was selected as a good bioindicator of the
quality of the water body since eels have a small range of habitat throughout the water
body where they live. Like other aquatic biota samples from freshwater systems, the
PBDE concentrations in these eels were high, with totals ranging from 1 to 408 ng/g wwt,
4-32
-------
with an average of 86 ng/g wwt over 17 samples. The predominant congener was BDE
47, explaining 56% of the concentrations, BDE 100 was the next most predominant
congeners at 30%, and other congeners contributed 3% or less. BDE 209 was not
detected in the samples, although the detection limit was not provided.
Hate et al (2001) measured BDEs 47, 49, 99, 100, 153, and 154 in 70 samples of
fish from the Roanoake River, the Dan River, and the Hyco River, which are within two
large watersheds in Virginia. Overvall statistics of findings were not provided. It was
noted that BDE 47 was quantified (at over 5 ng/g Iwt) in 88% of samples, with over half
the fish samples having concentrations greater than 100 ng/g lwt, and with 16 samples
containing BDE 47 at over 1000 ng/g lwt. The highest total concentration was 47,900
ng/g lwt. Overall, BDE 47 dominated the profiles, explaining between 40 and 75% for
the five different species tested, with BDEs 99 and 100 explaining in the range of 5 -
20% of concentrations. BDE 49 was also measured and identified, generally explaining
less than 5% of total PBDE concentrations in the sampled fish.
Marine fish sampled off the coast of Florida were analyzed for 12 PBDEs by
Johnson-Restrepo et al. (2005). Specifically, a total of 88 specimens from nine species of
marine fishes and twro species of dolphins were analyzed for BDEs 28, 47, 66, 85, 99,
100, 138, 153, 154, 183, 203, and 209, The overall mean concentration of total PBDEs
were 10-fold greater than the levels reported for Arctic marine fish such as polar cod
from Stortjorden, and for several fish species from southern Greenland. On the other
hand, the concentrations were one order of magnitude lower than what was reported for
lake trout in the Great Lakes. Total concentrations in teleost fishes ranged from 8 ng/g
lwt to 88 ng/g lwt, with a mean of 43 ng/g lwt. Mean concentrations of total BDEs in
sharks and dolphins were higher, ranging from 37.8 ng/g lwt to 1,630 ng/g lwt. BDE 47
was overwhelmingly found at the highest concentrations in teloest fish and dolphins (10s
of ng/g Kvt in teleost, but 100s of ng/g lwt in dolphins), but BDE 209 was the highest
found in sharks (16-778 ng/g lwt). Other than for sharks, BDE 209 was found at an
average of 0.5 ng/g lwt with several non-detects. Johnson-Restrepo et al. (2005) also did
a temporal study, evaluating 9 samples of dolphin blubber collected between 1991 and
1996, with 6 analogous samples collected between 2001 and 2004. The authors similarly
also compared bull shark samples collected in 1993 and 1994 with those collected in
4-33
-------
2002 and 2004. A clear elevation was seen in both PCB and PBDE concentrations; total
PBDEs in the dolphins rose from 363 to 1,190 ng/g Iwt, and total PBDEs in sharks rose
from 78 to 1,630 ng/g lwt.
Rayne et al, (2003) studied the temporal trend of increasing BDE concentrations
in fish from the Columbia River System in southeastern British Columbia. The
concentrations of 33 individual mono- through hcxabrominated BDE congeners (which
includes congeners up to BDE 156) were measured in mountain whitefish, largeseale
suckers, and surfkial sediments from several locations on the Columbia and Kootenay
River systems in southeastern British Columbia, Canada. A total of 41 Whitefish
samples were obtained from the period of 1992 to 2000, specifically 1992, 1994 and
1995, 1998, and 2000. Eleven sediment and 6 sucker samples were taken in 2001 and
2000, respectively. Total PBDE concentrations in whitefish, obtained at two locations,
increased by factors of 11.8 and 6.5, respectively, over the period from 1992 and 2000:
they went from 6.1 ng/g wwt (avg) in 1992 to 19.1 ng/g wwt in 1994 and 1995 to 71.8
ng/g wwt in 2000. The 6.5 factor increase at the other location was seen from the starting
concentration of 4.5 ng/g wwt to 29.2 ng/g wwt. At a remote site, total PBDE was 0.9
ng/g wwt in the whitefish, which the authors attribute to domestic wastewater draining
directly into the river at the other two sites. Suckers sampled in 2000 had lower
concentrations than whitefish at 5.0 ng/g wwt. The sediment congener pattern was
somewhat different than the fish. The primary congeners in the sediment and fish were
47, 99, and 100, but BDE 47 was the major congener in sediments (46-63% total),
followed by 99 (23-39%), and 100 (6-8%), while for whitefish, 99 was the dominant
congener, followed by BDEs 47 and 100. This is contrary to other literature, which
showed similar dominance of 47 in both sediment and fish.
Vives et al. (2004) collected fish from eleven high mountain lakes in Europe and
one in Greenland. The importance of these lakes being high is that the only way BDEs
could have reached these aquatic ecosystems is by long range transport. Liver and
muscle tissue of trout (brown trout, brook trout, and arctic char) were sampled. The
authors were unclear as to which BDEs were measured: they stated that all congeners
were identified using external standards comprised of 39 individual congeners up to BDE
190 (not including 209), but they only present results for BDEs 28, 33, 47, 99, 100, 153,
4-34
-------
and 154, while not providing any information on other congeners including that they
were looked for but not detected. Results from 55 trout specimens were as follows: 0.1-
1.3 and 0.07-0.77 ng/g wwt in liver and muscle, respectively (2.4-40.0 and 2.9-41.0 ng/g
Iwt). BDE 47 was the highest congener found, followed by 99, 100, 153, 154, and 28.
This disparity was much more apparent in liver as compared to muscle. There was an age
relationship found in the data: older fish had significantly higher BDE concentrations.
Concentration increases of between 4 and 12 times were found between 1- and 21-year-
old individual fish samples.
Peng et al. (2005) collected 60 tissue samples from 6 rivers and 3 estuaries in
2003 in China. BI)F congeners 28, 47, 99, 100, 153, 154, and 183 were measured.
Results showed river-average results ranged from 25 to 152 ng/g Iwt, and 31 to 281 ng/g
Iwt from the estuaries. The fish species were not identified. In all rivers, BDE 47 is the
predominant congener found, exceeding other congeners by factors of 3 or more and
comprising over 50% of the total concentration. In most cases, BDE 154 was the next
highest, followed by BDEs 99 and 100. The one exception was an estuary, where BDE
154 accounted for 101 ng/g iwt of a total of 281 ng/g Iwt, followed by 47 at 92 ng/g Iwt.
In all other locations, BDE 154 was 10 or less ng/g Iwt.
4.6.3. Fish From the Retail Marketplace
The fish data that might be considered the most relevant for estimating exposure
to the general population is the retail market basket sampling described in this section.
Also, this was essentially the only data on fish in the literature that reported
measurements of BDEs 183 and 209. Fish were either the primary target of the sampling
or included among a variety of food products.
Schecter et al. (2006a; derived from data in Schecter et al., 2004, with additional
data and analyses) reports the results of a comprehensive market basket survey entailing
62 samples that encompassed meat products, dairy products, and fish. BDEs 17, 28, 47,
66, 77, 85, 99, 100, 138, 153, 154, 183, and 209 were reported, although the average
results reported, which were based on the assumption of Vz DL, could be misleading due
to high detection limits, coupled with a high frequency of "nd" (not detected) or "na" (not
analyzed). For example, BDE 17 was only detected in 4 of 18 samples, at detection
4-35
-------
limits ranging from 0.0001 to 0.0007 ng/g wwt. However, the mean was reported as
0.0008 ng/g wwt, due to several samples with detection limits above 0.002 ng/g wwt.
The detection limit issue was even more apparent with BDE 209. For example, the mean
concentration for BDE 209 in fish was 0.1 ng/g wwt, but it was not detected in 14 of 24
samples, and the high mean was driven by two samples at 1.27 and 0.68 ng/g wwt; all
other positive detections were less than 0.025 ng/g wwt. The detection limits on BDE
209 were highly variable, ranging from 0.011 to 0.17 ng/g wwt. The data regarding the
more commonly found BDEs at higher concentrations (congeners 47, 99, 100, 153, and
154) must be assumed to be valid. A second key limiter of this data set is that all samples
were purchased from supermarkets in Dallas, TX, so it is unclear to what extent this data
represents national trends. A large proportion of the food supply is national in scope,
such as non-perishable or frozen food items, but others might be locally produced (fruit
and vegetables) or perhaps regionally produced (animal food products). Still, this data
set represents the most comprehensive survey of terrestrial animal food products in the
United States analyzed for PBDEs. The top four BDE congeners found in fish were BDE
47 at 0.6 ng/g wwt, BDE 99 at 0.17 ng/g wwt, BDE 100 at 0.13 ng/g wwt. and BDE 209
at 0.098 ng/g wwt. The mean concentration of BDE 183, the primary marker for the octa
BDE formulation, was 0.002 ng/g wwt, and the mean total concentration over all 24 fish
samples was 1.12 ng/g wwt.
Fish, meat, and fowl products were purchased in December 2003 and February
2004 from 3 different food markets in Sacremento and El Dorado Hills in Northern
California (Luksemburg ct al. 2004). A total of 31 different BDE congeners were
measured, although no congener-specific data were provided. Homologue BDE groups
were measured, and tables were provided with data for those. The total concentrations
found in fish ranged from 0.09 to 4.9 ng/g wwt.
The Norwegian Institute of Public Health in Oslo, Norway conducted an
intcrcomparison laboratory study on the measurement of PBTs including PCDD/Fs,
PCBs, and PBDEs (Haug et al., 2005). A total of 73 laboratories participated but only 21
laboratories reported back concentrations of 7 PBDEs in samples of chicken, trout, and
palm oil. The 7 BDEs were 28, 47, 99, 100, 153, 154, 183, and 209. Lake trout had the
highest concentrations by a large margin: mean concentrations were 0.6; 95.2; 78.6;
4-36
i
-------
39.2; 9.9; 12.0; 0.02; and 0.06 ng/g wwt for the 8 congeners, respectively, totaling 236
ng/g wwt. These results appear anomalous, not only because the concentration were so
much higher than anything else in the literature (total PBDEs have been found in the
single digit ng/g wwt or less, not in the hundreds of ng/g wwt), but also because their
reported concentrations in chicken and palm oil were up to 4 orders of magnitude lower,
at concentrations less than 0.0! ng/g wwt.
Tittlcmeier et al. (2004) purchased 122 fish and shellfish from retail stores in 3
Canadian cities (Vancouver, Halifax, and Toronto) in the winter of 2002 and analyzed
them, for 18 BDE congeners. The fish types include salmon, trout, tilapia, Arctic char,
mussels, oysters, shrimp, and crab, and the BDEs include 15, 17, 28, 47, 66, 71, 75, 77,
85,99, 100, 119, 126, 138, 153, 154, 183, and 190. Total BDEs per fish group ranged
from 0.02 in shrimp to 1.6 ng/g wwt in trout. Salmon was second with a geometric mean
concentration of 1.5 ng/g wwt Trout and salmon had the highest lipid content at 8 and
11%, respectively, partially explaining the high amount of BDEs in them. The third
highest lipid content, at 7.9%, was found in char, and not unexpectedly, char had the third
highest total BDE, at 0.6 ng/g wwt. BDE 47 was the most predominant congener,
explaining 48% of the concentration, with BDE 99 contributing 24%, Following these
two were BDEs 100, 28, 153, 154, and 183. Fanned samples showed higher
concentrations than wild samples, particularly for salmon.
Ohta et al. (2002) evaluated the concentration of PBDEs in breast milk and food
products, including several fish species, in Japan. The PBDEs evaluated included 28, 47,
99, 153. and 154. Concentrations in 20 fish samples (4 young yellowtail, 4 mackeral 2
natural yellowtail, 3 salon, 3 yellow tuna, and 3 short-necked clams) ranged between 0.02
and 1.65 ng/g wwt in the edible tissues of the fish, with the highest in yellow-fin tuna.
The dominant congeners were BDE-47—at over 50% in the samples—and BDE-99— at
about 20%. Questionnaires on food consumption were given to the women, and a strong
correlation was found between consumption of fish and breast milk concentration.
Specifically, in a "high" group offish consumers (n = 5), the average breast milk
concentration of BDEs were 1,724 pg/g Iwt, and this was significantly higher than the
concentration of BDEs in the "low" group offish consumers (n = 3), which was 774 pg/g
lwt.
4-37
-------
Pirard et al. (2005) presented results from salmon, whole trout, and Spanish
mussels purchased from a Belgian supermarket. Total concentrations were 8.19, 2.69,
and 10,22 ng/g Iwt for mussel, trout, and salmon, respectively. BDE 47 dominated the
results, comprising over 60% for the three samples, followed by 99 (in mussel and trout)
or 100 (in salmon).
Mcng ct al. (2007) collected samples of 13 fish species from local fish markets
and supermarkets from 11 fishery-producing regions in Guangdong Province, China.
These include freshwater farmed fish, seawatcr farmed fish, and wild marine fish. Eleven
congeners were measured, including BDE 209. The median and mean of the total of 10
BDEs, not including 209, were 0.16 and 0.23 ng/g wwt, respectively. BDE 209 was
found in only 14 of the 390 fish samples, ranging from < 0.1 to 0.57 ng/g wwt, although
the detection limit of BDE 209 was the highest, at 0.1 ng/g wwt, compared to 0.001-
0.003 ng/g wwt detection limits of other congeners.
4.6.4. Observations from Fish Data
While this review falls short of a comprehensive review of the literature on
brominated flame rctardants in fish, certainly it should be clear that sampling for fish is of
critical importance, both in the context of sampling fish as food in retail markets and fish
farms, and sampling in open aquatic settings as a marker for understanding and tracking
the status of PBDEs in the environment. Unfortunately, most of this sampling did not
include the higher brominated marker congeners 183 and 209. The very limited sampling
for BDE 183 suggested insignificant concentrations, but the market basket sampling by
Schecter et al. (2006a) and the marine environment sampling by Johnson-Restrepo et al.
(2005) suggested that BDE 209 can be significant and sometimes the highest congener
found in fish. Most often, however, BDE 47 dominated the profile, explaining over 50%
of the concentration found, with BDE 99 the second highest found explaining around
25%. The ratio of BDE 47 to BDE 99 was about 2.0 in these studies. Generally, total
BDE concentrations were highest in open water environments (lakes, rivers, oceans) in
contrast to fanned fish or fish obtained from marketplaces. In the wild environment,
concentrations ranged above 1,000 ng/g wwt but most often were well above 10 ng/g
wwt averaging between 10 and 100 ng/g wwt. Concentrations in. farmed and market fish
4-38
-------
were generally lower, in the neighborhood of 1-5 ng/g wet weight basis. It is not clear
why store-bought fish might be lower than wild eaught fish, except possibly that the
focus of wild caught fish is from locations historically known to be impacted by
contaminants such as dioxins, PCBs, or PBDEs, such as the Great Lakes. Like sediment
cores (described in the previous chapter) and body burden measurements (described in
the next chapter), temporal sampling of fish has provided evidence of the rise of these
compounds in environmental matrices throughout the 1990s into the 21^ century.
4.7. FOOD CONCENTRATIONS
United States data on food is highlighted by three market basket surveys: one in
Texas (Schecter et al., 2006a), one in California (Luksemburg et al., 2004), and one
sampling from several states in the United States
-------
Fish, meat, and fowl products were purchased in December 2003 and February
2004 from three different food markets in Sacremento and El Dorado Hills in Northern
California (Luksemburg et al., 2004). Thirty-one different BDE congeners were
measured, although no congener-specific data were provided, Homologue BDE groups
were measured, and tables were provided with data for those. The total PBDE
concentrations (sum of homologue contrations provided) were higher in fish (from 0.09
to 4.9 ng/g wwt) and fowl (0.09 to 2.5 ng/g wwt) than in beef and deer meat products
(from 0.1 to 0.4 ng/g wwt). It was stated that the highest concentrations of individual
congeners were 47, 99, and 100, although decaBDE, which is the single congener BDE
209, most often was the highest homologue group found,
A non-statistical sampling of market basket meat and poultry items from large
supermarkets was undertaken in 2004, in these states: FL, VA, CT, PA, ND, MT, OR,
NM and AZ (Huwe et al, 2005). A total of 65 meat samples including hamburger (n -
11), bacon (n =11), chicken fat (n = 22), pork fat (n = 11), and beef fat (n -- 10) were
measured for BDEs 28/33, 47, 85, 99, 100, 153, 154, 183. Beef had the lowest amounts
of BDEs (0.25 ng/g Iwt for beef fat and 0.67 ng/g Iwt for hamburger), while chicken and
pork had the highest concentrations— 2.96 ng/g Iwt and 2.62 ng/g Iwt, respectively.
These concentrations were driven up somewhat by a few high samples: One pork and
two chicken samples had total BDEs greater than 15.0 ng/g Iwt. BDE 99 dominated all
food types, thus, accounting for 39%, 36%, 46%, 40%, and 44% for hamburger, bacon,
chicken fat, pork fat, and beef fat respectively. BDE 47 was second most dominant,
explaining 27%, 27%, 27%, 41%, and 28%, respectively. The ratio of BDE-47 to BDE-
99 averaged 0.78, which is similar to the BDE 47/99 ratio of 0.6 found in a-penta
formulation. This contrasts a BDE 47/99 ratio of sometimes greater than 2, which was
identified in several studies in fish as described in Section 4.6.5. As will be described in
the next chapter, humans also have more BDE 47 congeners than 99, but about this same
factor of 2.0,
Huwe et al. (2003) earlier conducted research-oriented studies on BDEs in
poultry. Chicken fat samples from three chicken production sites known to have chickens
"contaminated" by consumption of animal feed with ball clay which had high levels of
dioxins, and two other chicken fat samples from an uncontaminated production site were
4-40
-------
analyzed for BDEs. The congeners analyzed include the following: 17/25, 33, 47, 66,
100, 99. 154, 85, 153, 140, 138, 183, and 209. Four other unidentified congeners were
noted, hut they are not included in the summary. The samples from two of the
"contaminated" sites did not appear to have BDE levels different than the
noncontaminated production site. So, for calculation of averages, the nine samples from
these three locations were averaged, and the four samples from the contaminated site,
where levels were higher, was analyzed separately. The authors conducted a PBDE
analysis of the stored contaminated feed but did not find elevated levels of PBDEs (data
not provided). This suggested to them that feed was not the cause of high levels of
PBDEs in the one site. The authors did state that in the one production site where BDE
levels were higher, a factory producing penta-BDE formulations was located in the same
city, which may have explained the higher levels found there. The average total BDE
concentration of 9 "uncontaminated" chicken was 8.3 ng/g Iwt. and the average of 4
"contaminated" chicken was 24.8 ng/g Iwt. BDE 99 was the highest congener; it
averages 3.1 ng/g Iwt uncontaminated and 10.9 ng/g Iwt contaminated. BGE congener 47
averaged 2.3 ng/g Iwt uncontaminated and 6.9 ng/g Hvt contaminated. BDE 209 was
generally low in all samples, with actually lower concentrations in the contaminated
versus uncontaminated sites; the average overall 13 samples was 1.1 ng/g Iwt.
Studies of PBDEs in food abroad mostly originate from Europe, and results are
comparable to America in that total PBDEs tend to be less than 1 ng/g, dominated by
BDEs 47 and 99, with sparse data on BDE 209. As noted below, however, two studies
have recently documented high levels of BDE 209 in food samples. In the United
Kingdom study, BDE 209 was found at the highest concentration of all congeners in
nearly every sample, and in the study from Spain. BDE 209 was found at the highest
level in several food products. In both studies, high concentrations of BDE 209 were 10
times or more higher than other congeners.
A recent study commissioned by WWF (formerly World Wildlikc Fund, now just
WWF), conducted by the Netherlands Organization for Applied Scientific Research
(TNO) (Peters, 2006), measured 30 BDEs in 26 food products from different countries in
Europe. The food products ranged from honey in the United Kingdom, to salami in Italy,
to pork chops in Poland. BDEs were found in 19 of 26 products: a low total of 0.15 ng/g
4-41
-------
found in honey to a high total of 1.3 ng/g wwt in minced beef. BDE 209 was looked for
but not found, although it cannot be expected to be found with a high detection limit of
5.0 ng/g wwt. Like other studies, BDEs 47 was found most frequently; it was found in 17
of 26 samples at an average positive concentration of 0.49 ng/g, Interestingly, BDE .32,
not measured in other studies, was found next most frequently, in 12 of 26 samples, with
an average positive concentration of 0.07 ng/g wwt. BDE 99 was found in 8 of 26
samples at an average positive concentration of 0.12 ng/g.
Another European study included a market basket survey of BDEs, including
BDE 209, from Belgium (Voorspoels et al., 2007), although BDE 209 was not quantified
in any sample. Concentrations generally were all less than 0.1 ng/g wwt, with fish being
the highest generally, and only 2 of 7 food samples contained BDE concentrations greater
than 1.0 ng/g wwt, at 1.0 and 1.6 ng/g wwt. Butter was the highest of the non-fish
samples, at 0.8 ng/g wwt. The eight meat samples were all under 0.2 ng/g wwt, as were
fast food and eggs. Cheese was slightly higher at 0.22 ng/g wwt.
As noted, two recent surveys showed high levels of BDE 209, one from United
Kingdom < IS A. 2006) and one from Spain (Gomara et al., 2006). In the dietary survey
conducted by the Food Standards Agency of the United Kingdom, composite samples
representing 19 food groups were analyzed for a suite of 17 BDE congeners, including
BDEs 183 and 209. The meat products concentrations of BDE 209 was exceedingly high
at 3.64 ng/g wwt, and in fact, BDE 209 had the highest in all but 2 composite food group
samples, although concentrations other than the meat concentration appear more in line
with other values in the literature, at less than 0.5 ng/g wwt. For example, BDE 209 was
found at 0.29 in "fats and oils" and the next highest congener concentrations were BDEs
49 and 99, both reported at 0.08 ng/g wwt. Although summary statistics were not
supplied, generally BDEs 47 and 99 were similar at concentrations between 0.01 and 0.10
ng/g, consistent with other surveys, and other congeners were present but at lower
concentrations. Gomara et al. (2006) collected 104 Spanish food samples randomly from
local supermarkets all over Spain from 2003 to 2005. The samples encompassed 21
types of food, including milk and dairy products, eggs, sea fish (tuna, sardine, and
others), meat and meat products, vegetable oil, and shellfish. A total of 15 BDEs were
measured, including for the first time the higher brominated BDE congeners 184, 191,
4-42
-------
196, and 197, in addition to 183 and 209, The highest total median concentration was
found in fish, 189 pg/g vvwt, followed by oils at 119 pg/g wwt, meats at 76 pg/g wwt,
shellfish at 76 pg/g wwt, eggs at 74 pg/g wwt and dairy at 66 pg/g wwt. BDE 209 was
the predominant congener in oil and egg samples, at 25 and 37 pg/g wwt, respectively,
but it was also found at significant levels in all other food products with medians in dairy
products at 4 pg/g wwt, in meats at 11 pg/g wwt, in fish at 5 pg/g fwt and shellfish at 7
pg/g fwt. Otherwise, BDEs 47 and 99 were the predominant congeners in dairy products
(11 and 9 pg/g wwt, respectively), meats {17 and 15 pg/g wwt), fish (115 and 15 pg/g
wwt), and shellfish (11 and 5 pg/g wwt).
Earlier in 2003, Bocio et al. (2003) evaluated dietary exposure of individuals in
Spain to PBDEs. A total of 54 samples were developed as composites of numerous food
types, including vegetables, tubers, pulses (peas, beans, and lentils), cereals, fruits, fish,
and shellfish, meat and meat products (pork, chicken, beef, lamb) eggs, milk, diary, fats
and oils. Samples were analyzed for total homologue groups, so no individual congeners
were measured. Because of this, the "total" concentrations here probably should not be
compared with totals^from other studies which measured and reported on individual
congeners. The highest concentration of total PBDEs was found in oils and fats ( 0.6 ng/g
wwt), followed by fish and shellfish (0.3 ng/g wwt), meat and meat products (0.1 ng/g
wwt), and eggs (0.06 ng/g wwt), with essentially none for vegetables and grains. A
predominance of the tetra and penta homologues, followed by hex a congeners, was found
in the samples.
As described in the section above on fish, the Norwegian Institute of Public
Health in Oslo, Norway conducted an intercomparison laboratory study on the
measurement of PBTs including PCDD/f s, PCBs, and PBDEs (Haug et al., 2005). It was
noted that fish concentrations were unusually high, totaling 236 ng/g wwt. Results for
the other congeners: 28, 47, 99, 100, 153, 154, 183, and 209, were more in line with
other values from around with world. Chicken levels of these seven were 0.0007, 0.02,
0.02, 0.007, 0.005, 0,003, 0.003, and 0.07 ng/g wwt, thus totaling 0.14 ng/g wwt, and
palm oil levels were 0.005, 0.02, 0.04, 0.02, 0.008, 0.006, 0.007, and 0.39 ng/g wwt, thus
totaling 0.49 ng/g wwt. It is noted that BDE 209 was the highest in these chicken and
palm oil samples, while it was among the lowest in the fish samples reported earlier.
4-43
-------
Irish animal and vegetative food products were surveyed by Tlustos et al. (2005)
for the presence of PBDEs. A total of 65 samples, most of them representing pooled
samples of 10 or more, were analyzed for PBDEs 17, 28, 47, 47, 66, 71, 77, 85, 99, 100,
119, 126, 138, 153, 154, and 183, Only total concentrations were reported, with no
discussion on congener distribution in the food products. Food products included:
carcass fat of bovine, avian (including duck), ovine, and porcine; dairy products included
butter, cheddar cheese, soft cheese, processed cheese, dairy spread, and yogurt; liver of
bovine, avian (chicken and turkey), ovine, and porcine; soup; cereals; fruit; vegetables;
vegetable/animal fat; and vegetable oil. Concentrations of total PBDEs ranged narrowly
from 0.85 to 1.49 ng/'g Iwt in all the terrestrial animal food products, and 0.17-0.34 ng/g
vvwt in the cereals, fruit, vegetables, and vegetable oil.
Levels of 47, 99, 100, 153, and 154 were measured in 33 food items between
2002 and 2004 in Norway (Knutscn et al., 2005). Levels ranged from 0.1-0.5 ng/g Iwt
for most terrestrial animal food products, but there were high measurements in eggs, 3.85
ng/g Iwt, margarine (with 5% fish oil) at 3.08 ng/g Iwt, liver paste a 1.17 ng/g Iwt, and
pork liver at 1.08 ng/g Iwt. Levels were much higher in fish, with wet weight
concentrations for the majority of fish types above 1.0 ng/g wwt, with a high of 9.73 ng/g
wwl for cod liver oil.
Harrad et al. (2004) presented data on BDEs 47, 99, 100, 153, and 154 for indoor
and outdoor air, and meat and vegan diets. The purpose of these measurements was to
estimate daily exposure to BDEs via inhalation and diet. The total PBDE concentration
in the composited samples of vegan and omnivorous diets were 0.15 and 0.18 ng/g dwt,
respectively. Similar to air, BDE congeners 47, 99, and 100 were higher in meat diets as
compared to vegan diets—BDEs 153 and 154 were very similar. Meat diet samples
averaged 0.07, 0.07, and 0.02 ng/g dwt for BDEs 47, 99, and 100, while the analogous
concentrations for vegan samples was 0.05. 0.06, and 0.01 ng/g dwt.
Schecter et al. (2006b) looked at the changes in PBDE levels when beef, lamb,
and fish were cooked. While they found a reduction in the amount of total BDEs in the
samples, they did not provide concentrations before and after cooking, so it cannot be
confirmed if, in fact, concentrations decreased. They did show how the percent lipid was
reduced by cooking (by fat dripping away), but concentrations are estimated as the mass
4-44
-------
of BDEs divided by the mass of lipids (or whole weight), and before and after masses
were not provided. Still, the reported amount of total BDEs decreased by over 60% in
beef and lamb, by over 50% in catfish, and by smaller amounts (-10%) in trout and
salmon. This certainly suggests that concentrations also would have been reduced, and it
would have been helpful if the authors had provided concentrations in their article.
While their recommendation that exposures consider cooked foods is reasonable, most,
often exposure via food is calculated by concentrations provided in measurements from
food products that are uncooked due to lack of information on concentrations in cooked
foods.
In summary, total BDE concentrations in terrestrial food products seem to be
lower (less than 1 ng/'g wwt) than fish BDE concentrations (1-5 ng/g wwt). Like for fish,
a paucity of data on BDE 209 makes it hard to generalize, although when sampled for, it
appeared to be present at levels comparable to those for BDE 47 and 99—the congeners
generally found at the highest concentration. BDE 99 was typically found at the highest
concentration, about twice as high as BDE 47. This contrasts the relationship between
these two congeners for fish, where BDE 47 is about twice as high as 99. In very limited
sampling, concentrations appear to be much lower in food products of vegetative origin
(such as cereals, fruits and vegetables) as compared to terrestrial animal food products.
This is to be expected, as these organic PBTs tend to bioaccumulate in fat of animals.
4.8. ASSIGNING EXPOSURE MEDIA CONCENTRATIONS FOR EXPOSURE
ASSESSMENT PURPOSES
The challenges with assigning values for individual congeners in exposure media
are many:
1) There is no consistent set of measured congeners, as in the case of the 17 toxic dioxin
and furan congeners. While the California Air Resources Board studies in air entailed 33
congeners (CARB, 2005, 2006), most studies only measured a handful of congeners.
Moreover, the majority of past studies have not measured the deca congener, BDE 209,
which remains the primary congener in currently produced and marketed product. The
second congener for which limited data exist is BDE 183. The octa formulation of
PBDEs contains about 44% hepta congeners, the most of any homologue group, and
4-45
-------
BDE 183 is the dominant stable congener within that homologue group. Its presence in
an environmental matrix could only have occurred via denomination of higher
brominated congeners or because of the presence of the octa formulation. For this
reason, BDE 183 is generally considered a primary marker for the possible presence of
the octa BDE formulation.
2) There are no statistically designed surveys which would provide a rigorous estimate of
exposure media concentrations to which the general population of the United States is
exposed. Most studies are non-statistical targeted surveys, conducted within a small
geographic area, and driven by a limited budget.
3) Key exposure matrices remain sparsely studied, including indoor air, outdoor soil, and
animal food products of terrestrial origin (meat, dairy, and eggs). Food concentration
data have only been obtained in the context of a limited number of retail market basket
surveys. While these have merit, their coverage is limited: only one in three had any
dairy samples, and the one study which did obtain samples from several states (Huwe et
al.« 2005) did not include measurements for BDE 209.
Therefore, it is not possible to derive media concentrations which are statistically
representative of general population exposures. Nonetheless, reasonable assumptions can
be made for exposure media concentrations to use in making the exposure estimates
(presented in Chapter 5). Decisions needed to be made on which congeners to include in
this derivation, and which studies to rely upon. The key congeners that have been
measured include the primary congeners of the penta formulation, including BDEs 47,
99, 100, 153, and 154. BDEs 183 and 209 remain of interest for their representation of
the octa and deca formulations, as described above. The U.S. EPA PBDE project plan
(see http://www.epa.gov/oppt/pbde) identifies BDE 28 as a triBDE of interest, BDE 85 as
a tetraBDE of interest, BDE 197 as an oetaBDE of interest, and BDE 206 as nonaBDE of
interest, so these will be included. Finally, many studies have also included the triBDE
17, the tetraBDE 66, and the hexaBDE 85 in their measurements; so these final congeners
will be added, bringing the total to 14 congeners for which profiles will be derived.
There are limited concentrations available for other congeners, as displayed in the media-
specific tables earlier in this chapter, but only these 14 will be assigned values used in
exposure calculations in the next chapter.
4-46
-------
Determining final concentrations to represent the general United States population
exposure is not straightforward. Some key considerations for this compilation include
the following:
1) The studies should come from the United States, or perhaps Canada. Only when North
American studies are unavailable will European or other foreign studies be used.
2) Occupational data, while of interest, does not represent general population exposure.
For indoor, as well as outdoor, exposures, it should be clear that the data were not taken
in the vicinity of known sources, such as recycling facilities, autoshredding facilities,
manufacturing facilities, and the like.
3) Studies with a full suite of congeners, and, in particular, BDEs 183 and 209, are
preferable to studies which have the limited set of BDEs associated with the penta
formulation—BDEs 47, 99, 100, 153, and 154.
4) Attempts to average congener-specific data across studies should be done with caution,
if at all. For example, it may be preferable to use data from one geographic area, as long
as it is appropriately background and not occupational, if that data has a full congener
suite For example, relying on an entire set of urban air data taken by the California Air
Resources Board, including the standard BDEs 47, 99, etc, as well as critical BDEs 183
and 209, would be preferable to only using the BDE 183 and 209 data from CARB, while
averaging the CARB data on other congeners with data from other urban or rural
locations in the United States.
With that as backdrop. Table 4.5 contains the final derived profiles for water,
surface soil, indoor dust, indoor and outdoor air, and categories of food products, for use
in the exposure calculations in the next chapter. It should be noted that when averages
were required that were derived from individual sample data, it was assumed that ND - 0
for these calculations. This was done because often the listed detection limit, such as
those from Schecter et al. (2006a), were much higher than the detected quantifications
(see Section 4.6 for more detail and examples on the detection limit issue in the Schecter
et al, 2006a effort). Following now is a brief justification for each of the media:
1. Drinking Water: Of the three studies found for surface water (none were found for
ground water) in the United States, the results from the San Francisco Estuaiy (Oros et
4-47
-------
al., 2005) were used to represent drinking water exposures. Concentrations of total
PBDEs were similar in the three studies, near or below 100 pg/L, but the study of the San
Francisco Estuary included BDE 209, The range of total PBDE concentrations ranging
from 3 to 513 pg/L, with a mean of 146.2 pg/L. The mean concentration of BDE 209 in
18 measurements was 42.3 pg/L.
2. Surface soil: The only systematic study of surface soil concentrations not associated
with an industrial or other contaminated source for United States was from Offenberg et
al. (2006), and results from this study will be used in this exercise. The average
concentration over the 14 congeners of this assessment was 82 ng/g dry. The only other
data on background soils came from Hassanin et al. (2004). It represented European soils
and had a much lower total PBDE concentration of only about 1 ng/g dwt.
3. Indoor dust: The Environmental Working Group (Sharp and Lunder, 2004) data set
includes samples provided by 10 women who earlier participated in a breast milk
sampling program. The concentrations of the congeners track well with other data, as
seen in Table 4.2. Because this data originated from 9 different states, it was judged that
this might be the most representative data set, although the concentrations may be a bit
high. The total of 8,275 ng/g dwt for the 14 congeners compares with 5,811 ng/g dwt
from the 17 homes in Washington, DC area (Stapleton, et al., 2005); the 9,271 ng/g dwt
from 2 samples from a computer lab in California (CARB, 2005); and the geometric
means of three locations (living room, bedroom, and from a household vacuum) within
20 homes in Boston of 13732, 6255, and 4,269 ng/g dwt, ng/g dwt, respectively (Allen et
al., 2008).
4. Outdoor air: The CARB data of 84 samples taken in 2004 from 7 monitors on 12
dates from locations in the Bay Area and the South Coast were used for the profile of
outside air. While the profile at 158 pg/nr might be higher than the profiles measured by
Hoh and Hitcs (2005) or Strandberg et al. (2001), in fact the congener-specific
measurements made in these two studies in urban areas are similar to the California
measurements. For example, the CARB data included a measurement of 25 pg/nr1 for
4-48
-------
BDE 209, while Hoh and Hites (2005) measured 60,1 pg/m' as an average of 28 samples
in 2002/2003. Strandberg et al. (200!) measured 0,30 pg/nr for BDE 209, but his
measurements pertain to 1997-99, which was before the prominent use of the deca
formulation. Strandberg et al, (2001) did measure concentrations of 33 and 16 pg/nr for
BDEs 47 and 99, respectively, which compares well with Hoh and Hites (2005),
measuring 17 and 7 pg/m3 for these two congeners, and with the CARB data showing 53
and 51 pg/m"' for these two congeners, respectively. Therefore, while possibly a bit high,
it would appear that the CARB data captures current urban conditions, and because 84
measurements of 10 of the 14 desired congeners were available (more than any other
study), these data were used.
5. Indoor air: The only indoor air measurements in the United States that could be
representative of general population exposures were taken in 20 urban residences in
Boston, MA (Allen et al., 2007). The geometric mean concentrations were presented for
three locations (personal, bedroom, living room), and the average of the 3 geometric
means were used as the representative indoor air concentrations. The sum of the
geometric mean concentrations for reported congeners for the three locations equaled 605
pg/m3 for the "personal" samples, which were taken near the breathing zone in the
bedroom, 392 pg/m3 in the bedroom, and 366 pg/nr1 in the living room.
6. Shellfish: The only shellfish data available were the data on clams, oysters, and
mussels from the San Francisco Estuary (Oros et al, 2005), so this data was used for the
profile. While measurements were made for all 14 congeners, it was stated that non-
detccts were found for all but BDEs 47, 99, and 100. The wwt total average
concentration was 5.7 ng/g wwt, with a range of 2-13 ng/g wwt
7. Finfish: The retail market place data from Schcctcr et al. (2006a) contained data
thought to be most representative for use in this exposure assessment, so it will be used
here. A total of 24 samples including tuna, salmon, shark, trout, catfish, and herring,
were taken. Measurements and quanti fications were made for 12 of the 14 congeners,
and the total concentration was 1.17 ng/g wwt. There were some substantially higher
4-49
-------
measurements taken from fish in the Great Lakes, including findings by Manchcster-
Neesvig et al, (2001) on 21 coho and Chinook salmon samples showing an average of 80
ng/g wwt, or the temporal study by Zhu and Hites (2004) showing an average of 120 ng/g
wwt for lake trout in Lakes Superior, Michigan, Huron, and Ontario. However, other
retail market surveys, such as the one in California, show a range of 0.04 to 4.9 ng/g wwt
in fish, and the one in Canada, including 122 fish and shellfish, shows a range of 0,02
ng/g wwt (in shrimp) to 1.6 ng/g wwt (in trout). Other data on farmed fish and fish from
abroad, reviewed in Section 4.5, similarly showed concentrations mostly below 10 ng/g
wwt and near the value of 1.17 ng/g wwt measured by Sehecter et al. (2006a).
8. Beef: The retail market place data from Scheeter et al. (2006a) will be used for beef
congener data. The total of 0,13 ng/g wwt, the average of 3 beef samples (2 ground, I
tenderloin), is comparable to the other major retail market basket surveys: Huwe et al.'s
(2005) beef samples (n = 11) averaged 0.42 ng/g Ivvt (roughly 0.08 ng/g wwt) for the five
major congeners (47, 99, 100, 153, 154), and Luksemburg et al.'s (2004) beef samples (n
- 4) from Northern California averaged 0.15 ng/g wwt total.
9. Pork: The retail market place data from Scheeter et al. (2006a) will be used for pork
congener data. The total of 0,28 ng/g wwt, the average of 7 pork samples (3 bacon, 1
pork, 2 pork sausage, I ground pork), is comparable to the results of Huwe et al. (2005),
whose pork samples including 11 bacon and 11 pork fat samples, showed a total of 1.59
ng/g Iwt for the five major BDE congeners of 47, 99, 100, 153, and 154, Assuming about
a 15% lipid weight of pork, this translates to 0,24 ng/g wwt
10. Poultry: The retail market place data from Scheeter et al. (2006a) will be used for
poultry congener data. The total of 0.36 ng/g wwt, the average of 3 poultry samples
(chicken breast, ground chicken, ground turkey) is comparable to the findings of
Luksemburg et al. (2004). Luksemburg et al. (2004) found an average of 0.41 ng/g wwt
for 7 poultry samples (4 chicken and 3 turkey samples) and Huwe et al. (2005) found an
average of 2.78 ng/g Iwt in 22 chicken fat samples. Assuming 15% fat in whole-weight
chicken, this translates to about 0.42 ng/g wwt. Interestingly, the duck sample from
4-50
-------
Luksemburg at 2.5 ng/g wwt, and the duck sample from Schecter et al. (2006a) at 1.3
ng/g ww i. were the highest of the poultry samples and these were both not included in the
displayed average. This high concentration occurred because duck is very fat, listed as
75% lipid in Schecter et al. (2006a). The chicken and turkey samples were listed at
between 5 and 11% lipid.
11. Dairy: The retail market place data from Schecter et al. (2006a) will be used for
dairy congener data. A total of 15 samples w ere available, including various cheeses,
cow/goat milk, yogurt, ice cream, and infant formula. The average whole-weight
concentration was 0.11 ng/g wwt.
12. Eggs: Schecter ct al. (2006a) reports on the average of 6 food egg samples. Unlike
the meat samples used for this assessment, individual sample data were not presented, so
that the presentation of average congener concentrations assuming ND - 'A DL in
Schecter et al. (2006a) was used here. The average whole-weight concentration was 0.09
ng/g wwt, the lowest of the terrestrial food products, but it was comparable to the whole-
weight concentrations of the other studies, which ranged up to 0.36 ng/g wwt.
REFERENCES
Allen JG, MD McClean, HM Stapleton, JW Nelson, TF Webster. 2007. Personal
exposure to polybrominated diphenyl ethers (PBDEs) in residential indoor air. Env. Sci.
Tech. 41:4574-4579.
Allen JG, MD McClean, HM Stapleton, TF Webster. 2008. Critical factors in assessing
exposure to PBDEs via house dust. Env. Int.. article-in-press. Available May 15, 2008.
Ashley J, D Libero, E Halscheid, L Zaoudeh, H Stapleton. Polybrominated diphenyl
ethers (PBDEs) in American Ells from the Delaware River, USA. Unpublished study
sponsored by the Partnership for the Delaware Estuary, Inc, and posted on their website,
http://www:delawareestuary.org. Accessed Oct 26, 2006.
Batterman, S, S Chernyak, E (iwynn. D Cantonwine, C Jia, L Begnoche, J P Hiekey,
2007, Trends of brominated diphenyl ethers in fresh and archived Great Lakes fish
(1979-2005). Chemosphere 69: 444-457.
4-51
-------
Bocio, L. JM Llobct, J Domingo, J Corbella, A Teixido, and C Casas. 2003.
Polvbrominated Diphenyl Ethers (PBDEs) in Foodstuffs: Human Exposure through the
Diet J. Agric, Food Chem. 2003(51): 3191-3195.
CADAMP. 2006. California Ambient Dioxin Air Monitoring Network web page for
polybrominated diphenyl ethers: http://www.arb.ca.gov/toxies/pbde.htm. Accessed May
11. 2006, for descriptions and data on PBDE monitoring in California.
Cai Z, G Jiang. 2005. Levels of polybrominated diphenyl ethers in soils from an e-waste
recycling site. Organohalogen Compounds 67: 548-550.
CARB. 2005. California Air Resources Board. Near-source ambient air monitoring of
polybrominated diphenyl ethers. Project #01-407, prepared by: MJ Charles, D Ciroskova,
and TM Cahill, Department of Environmental Toxicology, One Shields Ave, University
of California, Davis, Davis, CA 95616.
Chen L, B Mai, X Bi, S Chen, X Wang, Y Ran, X Luo, G Sheng, I Fu, and E Zeng.
2006. Concentration levels, composition profiles, and gas-particle partitioning of
polybrominated diphenyl ethers in the atmosphere of an urban city in South China. Env
Sci and fech 40: 1190-1196.
Christensen, J.H., J Platz. 2001. Screening of polybrominated diphenyl ethers in blue
mussels, marine and freshwater sediments in Denmark. J, of Env. Men, 3: 543-547.
Dodder NG, B Strandberg, and RA Hites. 2002. Concentrations and spatial variations of
polybrominated diphenyl ethers and several organochlorine compounds in fishes from the
Northeastern United States. Env. Sci Tech 36: 146-151.
Eljarrat, E.; de la Cal, A; Lazzazabal, D; Fabrellas, B; Fernandez-Alba, A.R; Borrull, F;
Marce, R.M; Barcelo, D. 2005. Occurrence of polybrominated diphenyl ethers,
polychlorinated dibenzo-p-dioxins, dibenzofurans and biphenyls in coastal sediments
from Spain. Environ Pollution 136: 493-501.
EPA. 2003. Exposure and Human Health Reassessment of 2,3,7,8-Tetraehlorodibenzo-p-
Dioxin (TCDD) and Related Compounds. United States Environmental Protection
Agency, Office of Research and Development, National Center for Environmental
Assessment. NAS Review Draft. December, 2003. EPA/600/P-00/001 C(a-f). Available
at, http:/Avwwr.epa.gov/ncea/dioxin.htm
Fabrellas, B, A Martinez, B Ramos, ML Ruiz, I Navaroo, A Torre. 2005. Results of an
European survey based on PBDEs analysis in household dust. Organohalogen
Compounds 67: 452- 454
Farrar NJ, KEC Smith, RGM Lee, GO Thomas, AJ Sweetman, and KC Jones. 2004.
Atmospheric emissions of polybrominated diphenyl ethers and other persistent organic
4-52
-------
pollutants during a major anthropogenic combustion event. Env Sci and Tech 38:1681 -
1685.
FSA, 2006. Brominated Chemicals: UK Dietary Estimates. Published by Food
Standards Agency, UK, Information Sheet 10/06. published at,
http://www.food.gov.uk/science/survcillancc/fsisbranch2006/fsisl006. Accessed Sep 28,
2006.
Gomara B. L Herrero, MJ Gonzalez. 2006. Survey of polybrominated diphenyl ether
levels in Spanish commercial foodstuffs. Env. Sci. and Tech 40: 7541-7547.
Gevao B, L Ali, A Al-Omair, M Helaleh. 2005. Non-dietary human exposure to
polybrominated diphenyl ethers in Kuwait. Grganohalogens, Volume 67: 1526-1529.
Goel A, LL McConnelf, A Torrents, JR Seudlark, S Simonich. 2006. Spray irrigation of
treated municipal wastewater as a potential source of atmospheric PBDEs. Env Sc i and
Tech 40:2142-2148.
Hale RC, M J La Guardia, EP Harvey, TM Mainor, WH Duff, MO Gaylor. 2001.
Polybrominated diphenyl ether flame retardants in Virginia freshwater fishes (USA).
Env. Sci. Tech 35: 4585-4591.
Hale RC, MJ La Guardia, E Harvey, TM Mainor. 2002. Potential role of tire retardant-
treated polyurethane foam as a source of brominated diphenyl ethers to the US
environment. Chemosphere 46: 729-735.
Harrad S, C Ibarra, M Diamond, L Mclymuk, M Robson, J Douwes, L Rooscns, AC
Dirtu, A Covad. 2008. Polybrominated diphenyl ethers in domestic indoor dust from
Canada, New Zealand, United Kingdom and United States. Env. hit 34: 232-238.
Harrad S, S Hazrati, C Ibarra. 2006. Concentrations of polychlorinated biphenyls in
indoor air and polybrominated diphenyl ethers in indoor air and dust in Birmingham,
United Kingdom: implications for human exposure. Env. Sci and Tech. 40: 4633-4638.
Harrad S, R Wijcsckera, S Hunger, C Halliwell, R Baker. 2004. Preliminary assessment
of U.K. human dietary and inhalation exposure to polybrominated diphenyl ethers. Env.
Sci. and Tech. 38: 2345-2350.
Hassanin A, K Breivik, SN Meijer, E Steinnes, GO Thomas, KC Jones. 2004. PBDEs in
European Background Soils: Levels and Factors Controlling Their Distribution. Env Sci
and Tech. 38:738-745.
Hayakawa K, H Takatsuki, I Watanabc, S Sakai. 2004. Polybrominated diphenyl ethers
(PBDEs), polybrominated dibenzo-p-dioxins/dibenzofurans (PBDD/Fs) and monobromo-
polychlorinated dibenzo-p-dioxins/dibenzofurans (MoBPXDD/Fs) in the atmosphere and
bulk deposition in Kyoto, Japan. Chemosphere 57: 343-356.
4-53
-------
Hazrati S, S Harrad 2005. Implications of passive sampling derived concentrations of
airborne PCBs and PBDEs in urban indoor microenvironments. Organohalogcn
Compounds 66: 1033-1036,
Hang L, C Thomsen, G Becher. 2005. The quality of PBDE analysis in food - results
from an interlaboratory comparison study. Organohalogen Compounds 67: 622-625.
Hayward D, J Wont, AJ Krynitsky. 2006. Polybrominated cliphenyl ethers and
polychlorinated biphenyls in commercially wild caught and farm-raised fish fillets in the
United States. In press, in Environmental Research.
Hites RA, JA Foran, SJ Schwager, BA Knuth, MC Hammilton, and DO Carpenter. 2004.
Global assessment of polybrominated diphenyl ethers in farmed and wild salmon. Env
Sci and Tech 19:4945-4949.
Hoh E, RA Hites. 2005. Brominated flame retardants in the atmosphere of the east-
central United States. Env Sci and Tech 39:7794-7802.
Huwe J, GL Larsen. 2005. Polychlorinated dioxins, furans, and biphenyls, and
polybrominated diphenyl ethers in a U.S. meat market basket and estimates of dietary
intake. Env Sci and Tech 39:5606-5611
Huwe JK, M Lorentzsen, K Thuresson, A Bergman. 2003. Analysis of mono- to deca-
brominatcd diphenyl ethers in chickens at the part per billion level. Chemosphere 46:
635-640.
Jacobs. MN, A Covaci, P. Schepens. 2002. Investigation of selected persistent organic
pollutants in fanned atlantic salmon (salmo salar), salmon aquaculture feed, and fish oil
components of the feed. Env Sci and Tech 36:2797-2805.
Jaward FM, NJ Farrar, T Harncr, AJ Sweetman, KC Jones. 2004. Passive air sampling
of PCBs, PBDEs, and organochlorine pesticides across Europe. Env Sci and Tech 38:
34-41.
Jaward FM, G Zhang. JJ Nam, AJ Sweetman, JP Qbbard, Y Kobara, KC Jones. 2005
Passive air sampling of polychlorinated biphenyls, organochlorine compounds, and
polybrominated diphenyl ethers across Asia. Env Sci and Tech 39:8638-8645.
Johnson A, K Seiders, C Deligcannis, K Kinney, P Sandvik, B Era-Miller, D Alkire,
2006. PBDE Flame Retardants in Washington Rivers and Lakes: Concentrations in Fish
and Water, 2005-2006. Washington Ecology Section, Environmental Assessment
Program, Washington State Department of Ecology, Olympia, Washington 98504-7710.
Publication No. 06-03-027. August, 2006.
4-54
-------
Johnson-Restrcpo B, K Kanna. R Addink, DH Adams. 2005. Polybrominatcd diphenyl
ethers and polychlorinated biphenyls in a marine foodweb of Coastal Florida . Ertv Sci
and Tech. 39:8243-8250.
Karlsson M, A Julander, B van Bavei, L Hardell. 2007. Levels of brominated flame
retardants in blood in relation to levels in household and dust. Env. Int. 33: 62-69,
Knutsen H, C Bergsten, C Thomsen, A Sletta, G Becher, J Alexander, HM Meltzer.
2005. Preliminary assessment of PBDE exposure from food in Norway. Organohalogen
Compounds 67: 1624- 1627.
Luksemburg W, R Wenning. M Maier, A Patterson, S Braithwaite, 2004.
Polybrominatcd diphenyl ethers (PBDE) and polychlorinated dibenzo-p-dioxins
(PCDD/F) and biphenyls (PCB) in fish, beef, and fowl purchased in food markets in
Northern California USA. Organohalogen Compounds 66: 3932-3937.
Manehcster-Neesvig JB, K Valters, and WC Sonzogni. 2001. Comparison of
polybrominated diphenyl ethers (PBDEs) and polychlorinated biphenyls (PCBs) in Lake
Michigan salmonids. Env Sci. and Tech 35: 1072-1077.
Mandalakis M. EG Stephanou, Y Horii, K Kannan. 2008. Emerging contaminants in car
interiors: evaluating the impact of airborne PBDEs and PBDD/Fs. Env. Sci and Tech.
Advance on-line, published on-line 3/26/08.
Meng, X-Z, EY Zcng, l.PYu. Y Guo, BX Mai. 2007. Assessment of human exposure to
polybrominated diphenyl ethers in China via fish consumption and inhalation. Env. Sci
and Tech 41: 4882-4887.
Offenberg JH, HM Stapleton, MJ Stryner, AB Lindstrom, 2006. Polybrominated
diphenyl ethers in US soils. Presented at Dioxin 2006, held Aug 21-25 in Oslo, Norway.
Summary statistics provided separately by JH Offenberg to MN Lorber by personal
communication, September 2006.
Ohta S, D Ishizuka, H Nishimura, T Nakao, O Aozasa, Y Shimidzu, F Ochiai, T Kida, M
Nishi, H Miyata. 2002. Comparison of polybrominated diphenyl ethers in fish,
vegetables, and meats and levels in human milk of nursing women in Japan.
Chemosphere 46:689-696
Oros DR, D Hoover, F Rodigari, D Crane, J Sericano. 2005. Levels and distribution of
polybrominated diphenyl ethers in water, surface sediments, and bivalves from the San
Francisco Estuary. Env Sci and Tech 39:33-41.
Pankow, J. F. Atmos. Environ. 1987, 21, 1405-1409.
Peng JH, HK Yak, CW Huang, JC Chung. 2005. Polybrominated diphenyl ethers
(PBDEs) in fish samples from rivers and estuaries. Organohalogen 67: 551-553.
4-55
-------
Peters, RJB. 2006, Man-Made Chemicals in Food Products, Prepared by TNO,
Netherlands Organization for Applied Scientific Research Research, Laan van Westenenk
501, P.O. Box 342, 7300 AH Apiedoorn, the Netherlands, TNO-Report, 2006-A-
R0095/B version 2. Prepared for, G. Watson, WWF-UK, Panka House, Weyside Park,
Godalming, Surrey GU7 1XR, United Kingdom, Available at,
h ttp: // www .pa nda.org/epo/ch a in.
Pirard G, G Eppe, AC Massart, ED Pauw, JF Focant. 2005. Evaluation of GC-MS/MS
for determination of PBDEs in fish and shellfish samples. Organohalogen Compounds
67:171-174.
Pless-Mulloli T, A Scehctcr, B Schilling, O Papke. 2006, Levels of PBDE in household
dust and lint in the UK, Germany and the USA. Presented at Dioxin 2006, held Aug 21-
25 in Oslo, Norway. Abstract available for download at, http://www.dioxin2006.org.
Raff J, R Hites. 2004. Polybrominated Diphenyl Ethers in Mississippi River Suspended
Sediment. Organohalogen Compounds 66: 3673-3677.
Rayne S, MG Ikonomou, R Antcliffe. 2003. Rapidly increasing polybrominated either
concentrations in the Columbia river system from 1992 to 2000. Env Sci and Tech 37:
2847-2854.
Rudel, RA, DE Camann, ID Spengler, LR Kom, JG Brody. 2003. Phthlatcs,
alkylphcnols, pesticides, polybrominated diphenyl ethers, and other endocrine-disrupting
compounds in indoor air and dust. Env Sci and Tech 37: 4543-4553.
Sajawan KS, S Nune, J Richardson, K Senthilkumar, BG Loganathan. 2006.
Contamination profiles of persistent organochlorines and polybrominated diphenyl ethers
in fish from coastal waters off Savannah, GA, USA. Presented at Dioxin 2006, held Aug
21-25 in Oslo, Norway. Abstract available for download at, http://wvvw.dioxin2006.org.
Schecter A, O Papke, TR Harris, J Olson, KC Tung, A Musumba, L Birnbaum. 2006a.
Polybrominated Diphenyl Ether (PBDE) levels in an expanded market basket survey of
U.S. food and estimated PBDE dietary intake by age and sex. Env Health Persp
114:1515-1520.
Schecter A. O Papke, KC Tung, T Brown, and A Musumba. 2006b. Changes in
polybrominated diphenyl ether (PBDE) levels in cooked foods. Tox and Env Chem 88:
201-211.
Schecter A, O Papke, JE Joseph, KC Tung. 2005. Polybrominated diphenyl ethers
(PBDEs) in U.S. computers and domestic carpet vacuuming: possible sources of human
exposure. J of Tox. And Env. Health, Part A, 68:501-513.
4-56
-------
Schecter A, O Papkc, K Tung, D Staskal, and L Birnbuam. 2004. Polybrominatcd
dipheny] ethers contamination of United States food. Env Sci and Tech 38:5306-5311.
Sharp, R, S. Lunder. 2004. In the Dust. Toxic Fire Retardants in American Homes.
Published by Environmental Working Group, report posted at
http://wvw.cwg.org/rcports/inthcdust/indcx.php.
Shaw S, A Bourakovsky, D Brenner, DO Carpenter, L Tao, K Kanna, C Hong. 2005.
Polybrominatcd Diphenyl Ethers (PBDEs) in farmed salmon (Salmo salar) from Maine
and Eastern Canada. Organohalogen Compounds 67: 644-646.
Shoeib M, T Harner, M Ikonomou, K Kannan. 2004. [ndoor and outdoor air
concentrations and phase partitioning of perfluoroalkyl sulfonamides and polybrominatcd
diphenyl ethers. Env Sci and Tech 38:1313-1320.
Shoeib M, T Harner. 2002. Using measured oetanol-air partition coefficients to explain
environmental partitioning of organochlorine pesticides. Environ. Toxicol. Chem. 21:
984-990.
Sjodin A, O Papke, E McGahee, R Jones, JF Focant, T Pless-Mulloli, LM Toms, R
Wang, Y Zhang, L Needham, T Hermrmann, D Patterson. 2004. Concentration of
polybrominatcd diphenyl ethers (PBDEs) in house hold dust from various countries -
inhalation a potential route of human exposure. Organohalogen 66: 3770-3775
Stapleton, HM, NG Dodder, JH Offenbcrt, MM Schantz, and SA Wise. 2005
Polybrominatcd diphenyl ethers in house dust and clothes dryer lint, Env Sci and Tech
39:925-931.
Strandberg B, NG Dodder, I Basu, and RA Hites. 2001. Concentrations and spatial
variations of polybrominatcd diphenyl ethers and other organohalogen compounds in
Great Lakes Air. Env Sci and Tech 35: 1078-1083.
Summer S. Streets SS, Henderson SA, Stoner AD, Carlson DL, Siracik MF,
Svvackhamcr DL. 2006. Partit ioning and Bioaccumuiation of PBDEs and PCBs in Lake
Michigan. Environ. Sci. Techno!. 40: 7263 -7269.
Tittlemeier SA, D Forsyth, K Breakwell, V Verigin, JJ Ryan, S Hayward, 2004.
Polybrominatcd diphenyl cithers in retail fish and shellfish samples purchased from
Canadian markets. J. Agr Food Chem 52: 7740-7745.
Tlustos, C, I Pratt, S White, A Fernandes, M Rose. 2005. Investigation into the levels of
PCDD/Fs, PCBs, and PBDEs in Irish Produce. Organohalogen Compounds 65: 1474 -
1477.
Toms, L; Mueller, J; Mortimer, M; Symons, R; Stevenson, G; Gaus, C. (2006)
Assessment of concentrations of polybrominatcd diphenyl ether flame retardants in
4-57
-------
aquatic environments in Australia. Australian Government Department of the
Environment and Heritage,Canberra.
Valters K, H Li, M Alaee, I D'sa, G Marsh, A Bergman, RJ Letcher. 2005.
Polybrominated diphenyl ethers and hydroxylated and methoxylated brominated and
chlorinated analogues in the plasma of fish from the Detroit River. Env Sei and Tech 39:
5612-5619.
Vcnier M, RA Hites. 2008. Flame retardants in the atmosphere near the Great Lakes.
Env. Sci. and Tech. 42: 4745-4751.
Vives I, JO Grimalt, S lacorte, M. Guillamon, D Barcel, BO Rosseland. 2004.
Polybromodiphenyl ether flame retardants in fish from lakes in European high mounts
and Greenland. Env Sci and Tech 38: 2338-2344.
Voorspoeis S, A Covaci, 11 Neels, P. Schepens, 2007, Dietary PBDE intake: A market-
basket study in Belgium. Env. Intern. 33: 93-97.
Wilford BH, T Harncr, J Zhu, M Shoeib, KC Jones. 2004. Passive sampling survey of
polybrominated diphenyl ether flame retardants in indoor and outdoor air in Ottowa,
Canada: Implications for sources and exposure. Env Sci and Tech 38:5312-5318.
Wilford BH, GO Thomas, KC Jones, B Davison, DK Hurst. 2008. Decabromodiphenyl
ether (deca-BDE) commercial mixture components, and other PBDEs, in airborne
particles at a UK site. Env. Int 34: 412-419.
Wu N, T Webster, T Hermann, O Paepke, J Tiekner, R Hale, E Harvey, M La Guardia, E
Jacobs. 2005. Associations of PBDE levels in breast milk with diet and indoor dust
concentrations. Organohalogen Compounds 67: 654-656.
Zhu LY, RA Hites. 2004. Temporal trends and spatial distributions of brominated flame
retardants in archived fishes from the Great Lakes. Env Sci and Tech 38:2779-2784.
4-58
-------
Table 4.1. Congener-specific concentrations of PBDEs in indoor dust in the United
States (units in ng/g dry weight)
Congener 1 Concentration, 1 Comment 1 Citation
J ng/g dwt 1 1
BiBDE
15
11
Mean 10 homes throughout US, but 9 ND, 1
at 109
EWG, 2004
TriBDE
17
9
Mean 17 homes, Wash DC
Stapleton, et aL 2005
5.5
Carpet dust (n™2) front computer labs in CA
CARB. 2005
2.4,6.4
Median, mean (n=-9) from vacuum samples in
Dallas. TX, in 2004
Scheeter et aL, 2005
1.4, 0.6, 0.4
Geo. mean of 3 loe/home from 20 homes in
Boston, MA area; liv rm, bdrm, and vac
Allen et al. 2008
25
2
Carpet dust (is 2) from computer labs in CA
CARB, 2005
28
21
Mean 17 homes, Wash DC, also listed as
congener 33
Stapleton, et al., 2005
ND (50)
Mean 10 homes throughout US
EWG, 2004
ITD"
Carpet dust (n = 2) from computer labs in CA
CARB, 2005
3, 20.3
Median, mean (n--9) from vacuum samples in
Dallas, TX, in 2004
Schecter et al., 2005
25, 14
Mean, geometric mean from homes in
Amarillo/Austin. Tex from 2006 (n=20)
llarradetal, 2008
30
ND
Carpet dust (n=2) from computer tabs in CA
CARB. 2005
32
Hnd~
Carpet dust (n=2) from computer labs in CA
CARB. 2005
33
ND (50)
Mean 10 homes throughout United States
EWG, 2004
28/33
16.3. 10.5,6.4
Geo. mean of 3 toc/home from 20 homes in
Boston, MA area; liv nn, bdrm, and vac
Allen etal. 2008
35
ND
Carpet dust (n=2)from computer labs in C'A
CARB, 2005
37
ND
Carpet dust
-------
20.3
Carpet dust (n~2) from computer labs in CA
( ARB. 2005
5.5. 26.1
Median, mean (n=9) from vacuum samples in
Dallas, TX, in 2004
Schecter et al., 2005
17.2,15.3,6.0
Geo. mean of 3 loc/home from 20 homes in
Boston, MA area; liv rm, bdrm, and vac
Allen et al. 2008
ND, 293.0
Median, max, from 11 homes in the Boston,
MA area
Wu et al. 2007
71
ND
Mean 17 homes. Wash DC
Stapleton, et al., 2005
ND
Carpet dust (n=2) from computer labs in CA
( ARB. 2005
75
ND
Carpet dust (n=2) from computer labs in CA
( ARB. 2005
9,3, 5.3. 3.6
Geo. mean of 3 loc/home from 20 homes in
Boston, MA area: liv rm. bdrm, and vac
Allen et al. 2008
77
0.8
Carpet dust (n=2) from computer labs in CA
< ARB. 2005
0.1.0.1
Median, mean (n=9) from vacuum samples in
Dallas, TX, in 2004
Schecter et al, 2005
PenfaBDE
85
83
Mean 17 homes. Wash DC
Stapleton, et al., 2005
100
Mean 10 homes throughout United States, but
8 ND, two at 453 and 544
EWG, 2004
51
Carpet dust (n~2) from computer labs in CA
t ARB. 2005
28.5, 96,4
Median, mean (n=9) from vacuum samples in
Dallas, TX, in 2004
Schecter et al., 2005
ND, 787
Median, max, from 11 homes in the Boston,
MA area
Wu et al. 2007
85/155
124.0, 51.8, 19.2
Geo. mean of 3 loc/home from 20 homes in
Boston. MA area; liv rm. bdrm. and vac
Allen etal. 2008
99
1,700
Mean 17 homes. Wash DC
Stapleton, et al., 2005
ND (400) io
22,500
Range of 89 homes sampled in Cape Cod,
VIA; 55% detected
Rudel et al. 2003
2.352
Mean 10 homes throughout United States
EWG, 2004
880 (69-3,700)
Median; range from 10 homes in Atlanta
Sjodin et al., 2004.
776
Carpet dust (n=2) from computer labs in CA
CARB. 2005
612, 2,295
Median, mean (n-9) from vacuum samples in
Dallas, TX, in 2004
Schecter et al, 2005
2460, 1170, 536
Geo, mean of 3 loc/home from 20 homes in
Boston. MA area; liv rm, bdrm, and vac
Allen etal. 2008
1014,14979
Median, max, from 11 homes in the Boston,
MA area
Wu et al. 2007
1400, 840
Mean, geometric mean from homes in
Amarillo/Aiistin, Tex from 2006 (ti =20)
Harrad et al. 2008
100
274
Mean 17 homes, Wash DC
Stapleton, et al.. 2005
ND (300) to
3,400
Range of 89 homes sampled in Cape Cod,
MA; 20% detected
Rudel et al. 2003
911
Mean 10 homes throughout United States
EWG, 2004
150(<15-660)
Median; range from 10 homes in Atlanta
Sjodin, et al, 2004
135
Carpet dust (n=2) from computer labs in CA
CARE. 2005
103,429
Median, mean (n=9) from vacuum samples in
Dallas, TX, in 2004
Schecter et al, 2005
436. 204, 77
Geo. mean of 3 loc/home from 20 homes in
Boston, MA area; liv rm, bdrm, and vac
Allen etal. 2008
174, 2776
Median, max, from 11 homes in the Boston,
MA area
Wu et al. 2007
240. 160
Mean, geometric mean from homes in
Amarillo/Austin, Tex from 2006 (n=20)
Harrad et al. 2008
4-60
-------
1 16
229
Carpet dust (n~-2) from computer labs in CA
CARB, 2005
_ .
70
Carpet dust (n=2) from computer labs in C'A
CARB, 2005
119
12
Carpet dust (n=2) from computer labs in CA
CARB. 2005
126
ND
Carpet dust (n=2) from computer labs in C'A
CARB. 2005
HexaBDE
138
17
Mean 17 homes, Wash DC
Stapleton, et al., 2005
181
Mean 10 homes throughout US, but 8 ND, 1
at 1,668
EWG, 2004
15
Carpet dust (n=2) from computer labs in CA
CARB, 2005
8. 23
Median, mean (ii=9) from vacuum samples in
Dallas, TX, in 2004
Scheeter et al. 2005
20.9. 12.1, 5.2
Geo. mean of 3 loc/home from 20 homes in
Boston, MA area; liv mi. fadrm, and vac
Allen et al. 2008
ND, 77
Median, ng/g, from 11 homes in the Boston,
MA area
Wu et al. 2007
153
181
Mean 17 homes. Wash DC
Stapleton. et al.. 2005
243
Mean 10 homes throughout US, but 7 ND, 1
at 1,510
EWG, 2004
140(5-650)
144
Carpet dust (n~;2) from computer labs in CA
CARB, 2005
61, 199
Median, mean (n~9) from vacuum samples in
Dallas. TX, in 2004
Scheeter et al. 2005
234.4. 124.2,
47.0
Geo. mean of 3 loc/home from 20 homes in
Boston, MA area; liv rm, bdrm, and vac
Allen et al. 2008
107, 563
Median, max, from 11 homes in the Boston.
MA area
Wu et al. 2007
240, 120
Mean, geometric mean from homes in
Amarilio/Austin, Tex from 2006 (n 20)
Harrad et al. 2008
154
156
Mean 17 homes. Wash DC
Stapleton et al. 2005
156
Mean 10 homes throughout US, but 8 ND, 1
at 1,050
EWG, 2004
77(<7-260)
Median; range from 10 homes in Atlanta
Sjodin, et al., 2004
95
Carpet dust (n~2) from computer labs in CA
CARB, 2005
54, 189
Median, mean (n=9) from vacuum samples in
Dallas, TX, in 2004
Scheeter et al, 2005
182.8.94.4,35.0
Geo, mean of 3 loc/home from 20 homes in
Boston, MA area; liv rm, bdrm, and vac
Allen etal. 2008
93. 455
Median, ng/g, from 11 homes in the Boston,
MA area
Wu et al. 2007
240, 100
Mean, geometric mean from homes in
Amarillo/Austin, Tex from 2006 (n=20)
Harrad et al. 2008
155
8
Carpet dust (n=2) from computer labs in CA
CARB, 2005
156
ND
Mean 17 homes. Wash DC
Stapleton, et al.. 2005
166
ND
Carpet dust (n=2) from computer labs in CA
CARB. 2005
HeptaBDE
181
ND
Carpet dust (n=2) from computer labs in CA
CARB, 2005
183
31
Mean 17 homes. Wash DC
Stapleton. et al., 2005
60
Mean 10 homes throughout United States, but
9 ND, 1 at 604
EWG. 2004
73 (<8-4.000)
Median; range from 10 homes in Atlanta
Sjodin et al., 2004
130
Carpet dust (n=2) from computer labs in C'A
CARB, 2005
18.6. 19.3
Median, mean (n=9) from vacuum samples in
Dallas, TX, in 2004
Scheeter et al.. 2005
27.9, 32.9, 15.1
Geo. mean of 3 loc/home from 20 homes in
Allen etal. 2008
4-61
-------
Boston, MA area; liv rm, bdrm, and vac
28, 16
Mean, geometric mean from homes in
Amarillo/Austin, Tex from 2006 (ir=l7)
Harradetal. 2008
184
ND
Mean 17 homes. Wash DC
Stapleton, et al,, 2005
190
5
Mean 17 homes. Wash DC
Stapleton, et al,, 2005
24
Carpet dust (n==2) from computer labs in CA
CARB, 2005
191
ND
Mean 17 homes. Wash DC
Stapleton, et al., 2005
OetaBDE
1%
15
Mean 17 homes. Wash DC
Stapleton, et al., 2005
3.6, 2,6. 3.9
Geo. mean of 3 loc/home from 20 homes in
Boston, MA area: liv rm, bdrm. and vac
Allen et al. 2008
197
17
Mean 17 homes, Wash DC
Stapleton, et al.. 2005
2,7, 3.3, 5.6
Geo, mean of 3 loc/home from 20 homes in
Boston, MA area; liv mi. bdrm, and vac
Allen et al. 2008
203
109
Carpet dust (n=2) from computer labs in CA
CARB, 2005
3.6, 3.6.4,9
Geo. mean of 3 loc/home from 20 homes in
Boston, MA area; liv rm, bdrm, and vac
Allen et al. 2008
NonaBDE
206
51
Mean 17 homes, Wash DC
Stapleton. et al.. 2005
76.3,48.1,40.5
Geo. mean of 3 loc/home from 20 homes in
Boston, MA area; liv rm. bdrm. and vac
Allen et al. 2008
207
30
Mean 17 homes. Wash DC
Stapleton, et al., 2005
45.9, 25.3, 26.6
Geo. mean of 3 loc/home from 20 homes in
Boston, MA area; liv rm, bdrm, and vac
Allen et al. 2008
208
35
Mean 17 homes, Wash DC
Stapleton, et al.. 2005
35.6, 17.5, 29.4
Geo. mean of 3 loc/home from 20 homes in
Boston, MA area; liv rm, bdrm. and vac
Allen etal. 2008
DecaBDE
209
2.090
Mean 17 homes. Wash DC
Stapleton, et al., 2005
2,394
Mean 10 homes throughout United States
EWG. 2004
2000{120-
21.000)
Median; range from 10 homes in Atlanta
Sjodin et al„ 2004
7500
Carpet dust (n=2) from computer labs in CA
CARB, 2005
665. 8567
Median, mean (rr=9) from vacuum samples in
Dallas, TX, in 2004
Schecter et al,, 2005
4502, 1703, 1811
Geo. mean of 3 loc/home from 20 homes in
Boston, MA area; liv rm, bdrm, and vac
Allen etal. 2008
ND, 9020
Median, ng/g, from 11 homes in the Boston,
MA area
Wu et al. 2007
1600, 1300
Mean, geometric mean from homes in
Ainarillo/Austm, Tex from 2006 (n=17)
Harrad et al. 2008
4-62
-------
Table 4.2. Outdoor and indoor congener-specific air concentrations of PBDEs in the
United States (units in pg/m3)
Congener
Concentration,
pg/nr
Comment
Citation
TriBDE
17
3,6
Outdoors at UC/Davis, 2004; average of 2 days
of filter/PUF and filter/XAD samplers
CARB, 2005
46
Indoors in computer facility, average of 2 days,
8 measurements, 6 with computers on, 2 with
computers off, in CA
CARB. 2005
1.5-5,6
Range, n~12 (3 days, 4 samplers) outdoors
surrounding electronics recycling facility in CA
CARB. 2005
35-106
Range, n~6 (3 days, 2 samplers) indoors in
electronics recycling facility in CA
CARB, 2005
0.3-5.8
Range. n=9 (3 days, 3 samplers) outdoors in
auto shredder facility in CA
CARB, 2005
1.7
N=84 (12 sampling dates, 7 monitors); average
for 2004 in Bav Area and South Coast of CA
CAD A MP, 2006
7.6, 8.1, 7.0
Geo. mean of 20 homes, 3 loc/home in Boston,
MA area; "personal", bedroom, and living room
Allen et al. (2007);
25
2.3
Outdoors at UC/Davis, 2004; average of 2 days
of filter/PUF and filter/XAD samplers
CARB, 2005
29
Indoors in computer facility, average of 2 days,
8 measurements. 6 with computers on, 2 with
computers off, in CA
CARB. 2005
0.4-4.7
Range, n=12 (3 days, 4 samplers) outdoors
surrounding electronics recycling facility in C'A
CARB, 2005
16-103
Range. n=6 (3 days, 2 samplers) indoors in
electronics recycling facility in CA
CARB, 2005
0.5-3.1
Range, n=9 (3 days, 3 samplers) outdoors in
auto shredder facility in CA
CARB. 2005
28
99
Indoors in computer facility, average of 2 days,
8 measurements, 6 with computers on, 2 with
computers off, in CA
CARB, 2005
2.9-17.8
Range, n 12 (3 days, 4 samplers) outdoors
surrounding electronics recycling facility in C'A
CARB, 2005
102-372
Range, n=6 (3 days, 2 samplers) indoors in
electronics recycling facility in CA
CARB, 2005
0.4-7.4
Range, n=9 (3 days, 3 samplers) outdoors in
auto shredder facility in CA
CARB, 2005
3.0
N=84 (12 sampling dates. 7 monitors): average
for 2004 in Bav Area and South Coast of CA
CADAMP, 2006
28 + 33
4.0
Outdoors at UC/Davis. 2004; average of 2 days
of filter/PUF and filter/XAD samplers
CARB, 2005
29.6, 27.3. 25.4
Geo. mean of 20 homes, 3 loc/home in Boston,
MA area; "personal", bedroom, and living room
Allen et ai, (2007)
30
0.2
Outdoors at UC/Davis. 2004; average of 2 days
of filter/PUF and filter/XAD samplers
CARB, 2005
ND
Indoors in computer facility, average of 2 days,
8 measurements, 6 with computers on, 2 with
computers off, in CA
CARB, 2005
ND
Range, n=!2 (3 days, 4 samplers) outdoors
CARB, 2005
4-63
-------
surrounding electronics recycling facility in CA
ND
Range, n=6 (3 clays. 2 samplers) indoors in
electronics recycling facility in CA
CARB, 2005
ND-0,9
Range, n=9 (3 days, 3 samplers) outdoors in
auto shredder facility in CA
CARB, 2005
32
ND
Outdoors at UC/Davis, 2004; average of 2 days
of filler/PUF and filter/XAD samplers
CARB, 2005
ND
Indoors in computer facility, average of 2 days.
8 measurements, 6 with computers on, 2 with
computers off, in CA
CARB, 2005
ND
Range, n=I2 (3 days, 4 samplers) outdoors
surrounding electronics recycling facility in CA
CARB, 2005
ND
Range, n=6 (3 days, 2 samplers) indoors in
electronics recycling facility in CA
CARB, 2005
ND-0.9
Range, n~9 (3 days, 3 samplers) outdoors in
auto shredder facility in CA
CARB, 2005
33
ND
Indoors in computer facility, average of 2 days,
8 measurements, 6 with computers on, 2 with
computers off, in CA
CARB, 2005
MM.4
Range, n=12 (3 days, 4 samplers) outdoors
surrounding electronics recycling facility in CA
CARB, 2005
ND
Range, n~6 (3 days, 2 samplers) indoors in
electronics recycling facility in CA
CARB, 2005
0.7-35
Range. n=9 (3 days, 3 samplers) outdoors in
auto shredder facility in CA
CARB. 2005
35
0,4
Outdoors at UC/Davis, 2004; average of 2 days
of t iller I'l i and filter/XAD samplers
CARB, 2005
ND
Indoors in computer facility, average of 2 days,
8 measurements, 6 with computers on, 2 with
computers off, in CA
CARB, 2005
N D-1.1
Range, n~12 (3 days, 4 samplers) outdoors
surrounding electronics recycling facility in CA
CARB, 2005
ND-18.1
Range, n~6 (3 days, 2 samplers) indoors in
electronics recycling facility in CA
CARB, 2005
ND
Range. n=9 (3 days, 3 samplers) outdoors in
auto shredder facility in CA
CARB, 2005
TetraBDE
37
2.3
Outdoors at UC/Davis, 2004; average of 2 days
of filter/PUF and filter/XAD samplers
CARB, 2005
ND
Indoors in computer facility, average of 2 days,
8 measurements. 6 with computers on, 2 with
computers off, in CA
rCARBT2005
1.3-8.4
Range. n=12 (3 days, 4 samplers) outdoors
surrounding electronics recycling facility in CA
CARB, 2005
ND-468
Range, n=6 (3 days, 2 samplers! indoors in
electronics recycling facility in CA
CARB, 2005
0.6-2.4
Range, n=9 (3 days, 3 samplers) outdoors in
auto shredder facility in CA
CARB, 2005
47
6.2
Michigan (n=35), 2002/3
Hoh & Hites, 2005
17.4
Chicago
-------
33
Chicago, n=l2, 97-99
Strandberg et al 2001
175 (671)
Lewes, Del Marva, MD (n=95) geom. mean
(max), gaseous phase only
Goel et al„ 2006
9.7(26)
Horn Point. Del Mam, MD (n=98) geom.
mean (max), gaseous phase only
Goel et al., 2006
17(52)
Dover, Del Marva, MD (n=47) geom. mean
(max), gaseous phase only
Goel et al., 2006
34,5
Outdoors at UC/Davis, 2004; average of 2 days
of fiiter/PUF and filter/XAD samplers
CARB. 2005
1,065
Indoors in computer facility, average of 2 days,
8 measurements, 6 with computers on, 2 with
computers off, in C'A
CARB, 2005
30-128
Range, n=12 (3 days, 4 samplers) outdoors
surrounding electronics recycling facility in CA
CARB, 2005
604-2,850
Range, n~6 (3 days, 2 samplers) indoors in
electronics recycling facility in CA; avg = 1,772
CARB, 2005
30-88
Range, u " (3 days, 3 samplers) outdoors in
auto shredder facility in CA
CARB, 2005
53,0
N-84 (12 sampling dates. 7 monitors); average
for 2004 in Bav Area and South Coast of CA
CADAMP, 2006
226,8, 157,9,
Geo. mean of 20 homes, 3 loc/home in Boston,
Allen et al. (2007)
145.1
MA area; ''personal", bedroom, and living room
49
1.2
Outdoors at UC/Davis, 2004; average of 2 days
of fiiter/PUF and filter/XAD samplers
CARB, 2005
59
Indoors in computer facility, average of 2 days,
8 measurements, 6 with computers on, 2 with
computers off, in CA
CARB, 2005
4,6-36,7
Range, n--12(3 days, 4 samplers) outdoors
surrounding electronics recycling facility in CA
CARB, 2005
449-2860
Range, n=6 (3 days. 2 samplers) indoors in
electronics recycling facility in CA; avg = 1764
CARB, 2005
2.7-10,5
Range, n=9 (3 days, 3 samplers) outdoors in
auto shredder facility in CA
CARB, 2005
49
9.1, 6.0, 7.2
Geo. mean of 20 homes, 3 loc/home in Boston.
MA area; "personal", bedroom, and living room
Allen et al. (2007)
65
2.0
N=84 (12 sampling dates, 7 monitors); average
for 2004 in Bay Area and South Coast of CA
CADAMP, 2006
66
1.6
Outdoors at UC/Davis, 2004; average of 2 days
of fiiter/PUF and filter VXD samplers
CARB, 2005
29
Indoors in computer facility, average of 2 days,
8 measurements. 6 with computers on, 2 with
computers off, in CA
CARB, 2005
2.9-10
Range, n 12 (3 days, 4 samplers) outdoors
surrounding electronics recycling facility in CA
CARB, 2005
103-750
Range. n=6 (3 days, 2 samplers) indoors in
electronics recycling facility in CA
CARB, 2005
2.3-5.1
Range, n=9 (3 days, 3 samplers) outdoors in
auto shredder facility in CA
CARB, 2005
3.7, 3.5, 3.5
Geo. mean of 20 homes, 3 loc/home in Boston,
MA area; "personal", bedroom, and living room
Allen, et al. (2007)
71
1.7
Outdoors at UC/Davis, 2004; average of 2 days
of fiiter/PUF and filter/XAD samplers
CARB, 2005
ND
Indoors in computer facility, average of 2 days,
8 measurements, 6 with computers on, 2 with
CARB, 2005
4-65
-------
computers off, in CA
ND
Range. n=12 (3 days, 4 samplers) outdoors
surrounding electronics recycling facility in CA
CARB, 2005
ND
Range, n=6 (3 days, 2 samplers) indoors in
electronics recycling facility in CA
CARB, 2005
ND
Range, n=9 (3 days, 3 samplers) outdoors in
auto shredder facility in CA
CARB, 2005
75
0.4
Outdoors at UC/Davis, 2004; average of 2 days
of filter/PUF and filter/XAD samplers
CARB, 2005
6.6
Indoors in computer facility, average of 2 days,
8 measurements, 6 with computers on, 2 with
computers off, in CA
CARB. 2005
ND
Range, o—12 (3 days, 4 samplers) outdoors
surrounding electronics recycling facility in CA
CARB, 2005
ND-I94
Range, n=6 (3 days, 2 samplers) indoors in
electronics recycling facility in CA
CARB, 2005
0.5-2.4
Range, n=9 (3 days, 3 samplers) outdoors in
auto shredder facility in CA
CARB, 2005
77
ND
Indoors in computer facility, average of 2 days,
8 measurements, 6 with computers on, 2 with
computers off, in CA
CARB, 2005
ND
Range, n 12 (3 days, 4 samplers) outdoors
surrounding electronics recycling facility in CA
CARB, 2005
22.1-75.5
Range, n-~6 (3 days, 2 samplers) indoors in
electronics recycling facility in CA
CARB, 2005
ND-2,4
Range, n=9 (3 days, 3 samplers) outdoors in
auto shredder facility in CA
CARB, 2005
2.0
N~84(12 sampling dates. 7 monitors); average
for 2004 in Bay Area and South Coast of CA
CADAMP, 2006
PentaBDE
85
I.J
Outdoors at UC/Davis, 2004; average of 2 days
of filter/PUF and filter/XAD samplers
CARB, 2005
13
Indoors in computer facility, average of 2 days,
8 measurements, 6 with computers on, 2 with
computers off. in CA
CARB, 2005
ND-6
Range, o 12 (3 days, 4 samplers) outdoors
surrounding electronics recycling facility in CA
CARB, 2005
87-284
Range, n=6 (3 days, 2 samplers) indoors in
electronics recycling facility in CA
CARB, 2005
2.1-9.0
Range, n=9 (3 days, 3 samplers) outdoors in
auto shredder facility in CA
CARB, 2005
2.3
N=84 (12 sampling dates, 7 monitors); average
for 2004 in Bay Area and South Coast of CA
CADAMP, 2006
3.8.2.7,2.5
Geo. mean of 20 homes, 3 loc/home in Boston,
MA area; "personal", bedroom, and living room
Allen, et al. (2007)
99
5.1
Michigan (n=35), 2002/3
Hoh & Mites, 2005
7.4
Chicago (n=28), 2002/3
Hoh & Hites, 2005
5.1
Indiana (n=38), 2002/3
Hoh & Hites, 2005
5.4
Arkansas (n=30), 2002/3
Hoh & Hites, 2005
3.0
Louisiana (n=26), 2002/3
Hoh & Hites, 2005
3.4
3 Rural/remote sites in MI, NY (n=36), 97-99
Strandberg et al 2001
16
Chicago, n=12, 97-99
Strandberg et al 2001
26 (178)
Lewes, Del Marva, MD (n=95) geom. mean
(max), gaseous phase only
Goel et al., 2006
4-66
-------
5.3 (26)
Horn Point, Del Marva. MD (n=98) geom.
mean (max), gaseous phase only
Goel et al, 2006
7.7(17)
Dover, Del Marva, MD (n=47) geom. mean
(max), gaseous phase only
Goel et al., 2006
12.5
Outdoors at UCVDavis, 2004: average of 2 days
of filter/PUF and filter/XAD samplers
CARB, 2005
239
Indoors in computer facility, average of 2 days.
8 measurements, 6 with computers on, 2 with
computers off, in CA
CARB. 2005
9-85
Range, n=12 (3 days, 4 samplers) outdoors
surrounding electronics recycling facility in CA
CARB, 2005
402-3210
Range, n=6 (3 days, 2 samplers) indoors in
electronics recycling facility in CA; avg=l,771
CARB, 2005
21-111
Range, n=9 (3 days, 3 samplers) outdoors in
auto shredder facility in CA
CARB, 2005
51.0
N=~84 (12 sampling dates, 7 monitors): average
for 2004 in Bay Area and South Coast of CA
CADAMP, 2006
110.8, 66.9,
60.3
Geo. mean of 20 homes, 3 loc/home in Boston,
MA area: "personal", bedroom, and living room
Allen et al. (2007)
100
1.1
Michigan (n=35), 2002/3
Hoh & Hites, 2005
1.8
Chicago (n~28), 2002/3
Hob & Hites, 2005
1.0
Indiana (n=38). 2002/3
Hoh & Hites, 2005
1.1
Arkansas (n~30), 2002/3
Hoh & Hites, 2005
0.7
Louisiana (n=26), 2002/3
Hoh & Hites, 2005
0.5
3 Rural/remote sites in Ml. NY (n=36), 97-99
Strandberg et al 2001
2.0
Chicago, tr=12, 97-99
Strandberg et al 2001
17(73)
Lewes, Del Marva, MD (n=95) geom. mean
(max), gaseous phase only
Goel et al., 2006
5.4 (5.4)
Horn Point. Del Marva. MD (n=98) geom.
mean (max), gaseous phase only
Goel et al., 2006
5.3 (5.3)
Dover, Del Marva, MD (n=47) geom. mean
(max), gaseous phase only
Goel et al.. 2006
4.8
Outdoors at UC/Davis, 2004: average of 2 days
of filter/PUF and filter/XAD samplers
CARB, 2005
100
Indoors in computer facility, average of 2 days,
8 measurements. 6 with computers on, 2 with
computers off, in CA
CARB, 2005
3-17
Range, n= 12 (3 days, 4 samplers) outdoors
surrounding electronics recycling facility in C'A
CARB, 2005
69-448
Range, n=6 (3 days, 2 samplers) indoors in
electronics recycling facility in CA
CARB, 2005
5.2-21.5
Range, n=9 (3 days, 3 samplers) outdoors in
auto shredder facility in CA
CARB. 2005
13.0
N=84 (12 sampling dates, 7 monitors); average
for 2004 in Bay Area and South Coast of CA
CADAMP, 2006
22.2, 14.4, 12.0
Geo. mean of 20 homes, 3 loc/home in Boston,
MA area; "personal."' bedroom, and living room
Allen et al. (2007)
116
0.2
Outdoors at UC/Davis, 2004; average of 2 days
of filter/PUF and filter/XAD samplers
CARB, 2005
14.7
Indoors in computer facility, average of 2 days.
8 measurements. 6 with computers on, 2 with
computers off, in CA
CARB. 2005
ND
Range. n=12 (3 days, 4 samplers) outdoors
surrounding electronics recycling facility in CA
g ¦ ^ ^ '
CARB, 2005
4-67
-------
ND
Range, n=6 (3 days, 2 samplers) indoors in
electronics recycling facility in C'A
CARB, 2005
ND
N=84 (12 sampling dates, 7 monitors); average
for 2004 in Bav Area and South Coast of CA
CADAMP, 2006
118
0.3
Outdoors at UC/Davis, 2004; average of 2 days
offilter/PUF and filter/XAD samplers
CARB, 2005
ND
Indoors in computer facility, average of 2 days,
8 measurements, 6 with computers on. 2 with
computers off, in CA
CARB. 2005
ND-5.3
Range, n--l 2 (3 days, 4 samplers) outdoors
surrounding electronics recycling facility in CA
CARB, 2005
67-343
Range, n=6 (3 days, 2 samplers) indoors in
electronics recycling facility in CA
CARB, 2005
1.1-3.9
Range, n=9 (3 days, 3 samplers) outdoors in
auto shredder facility in CA
CARB, 2005
119
1.4
Outdoors at UC/Davis, 2004; average of 2 days
offilter/PUF and filter/XAD samplers
CARB, 2005
ND
Indoors in computer facility, average of 2 days,
8 measurements, 6 with computers on, 2 with
computers off, in CA
CARB, 2005
ND
Range, n=i2 (3 days, 4 samplers) outdoors
surrounding electronics recycling facility in CA
CARB, 2005
ND
Range, n~6 (3 days, 2 samplers) indoors in
electronics recycling facility in CA
CARB, 2005
ND
Range, u "> (3 days, 3 samplers) outdoors in
auto shredder facility in CA
CARB, 2005
126
ND
Outdoors at UC/Davis, 2004; average of 2 days
of filter/PDF and filter/XAD samplers
CARB. 2005
ND
Indoors in computer facility, average of 2 days,
8 measurements, 6 with computers on, 2 with
computers off, in CA
CARB, 2005
ND
Range, n= l2 (3 days, 4 samplers) outdoors
surrounding electronics recycling facility in CA
CARB, 2005
ND
Range. n=6 (3 days, 2 samplers) indoors in
electronics recycling facility in CA
CARB, 2005
ND
Range, n=9 (3 days, 3 samplers) outdoors in
auto shredder facility in CA
CARB, 2005
HexaBDE
138
0.9
Outdoors at UC/Davis, 2004; average of 2 days
of filter/PUF and filter/XAD samplers
CARB. 2005
ND
Indoors in computer facility, average of 2 days,
8 measurements, 6 with computers on. 2 with
computers off, in CA
CARB, 2005
ND-6,8
Range, n 12 (3 days, 4 samplers) outdoors
surrounding electronics recycling facility in CA
CARB, 2005
95-346
Range, n=6 (3 days. 2 samplers) indoors in
electronics recycling facility in CA
CARB. 2005
1.3-3.9
Range, n=9 (3 days, 3 samplers) outdoors in
auto shredder facility in CA
CARB, 2005
153
0.20
3 Rural/remote sites in MI, NY (n=36), 97-99
Strandberg et al.,
2001
0.53
Chicago, n= 12, 97-99
Strandberg et al,
2001
2.0
Outdoors at UC/Davis, 2004; average of 2 days j CARB. 2005
4-68
-------
of filter/PUF and filter/XAD samplers
11
Indoors in computer facility, average of 2 days,
8 measurements, 6 with computers on, 2 with
computers off, in C'A
CARB, 2005
3-150
Range. n=12 (3 days, 4 samplers) outdoors
surrounding electronics recycling facility in CA
(. ARB. 2005
1,120-8,900
Range, n=6 (3 days, 2 samplers) indoors in
electronics recycling facility in CA; avg = 5,623
CARB, 2005
11.3-33.2
Range, ir=9 <3 days, 3 samplers) outdoors in
auto shredder facility in CA
CARB, 2005
3.9
N=84 (12 sampling dates, 7 monitors); average
for 2004 in Bav Area and South Coast of CA
CADAMP, 2006
8.6, 4.0, 3.5
Geo. mean of 20 homes, 3 loc/home in Boston,
MA area; "personal", bedroom, and living room
Allen et al. (2007)
154
0.12
3 Rural/remote sites in Ml, NY (n=36). 97-99
Strandberg et al.,
2001
041
Chicago, n=12, 97-99
Strandberg et al,,
2001
ND
For Lewes, Horn Point, Del Marva, MD (n ~
240), gaseous phase only
Goel et al., 2006
2.8
Outdoors at UC/Davis. 2004; average of 2 days
of filter/PUF and filter/XAD samplers
CARB, 2005
13
Indoors in computer facility, average of 2 days,
8 measurements, 6 with computers on, 2 with
computers off. in CA
CARB, 2005
1.5-86.7
Range, n=i2 (3 days, 4 samplers) outdoors
surrounding electronics recycling facility in CA
CARB, 2005
1230-5260
Range. n=6 (3 days, 2 samplers) indoors in
electronics recycling facility in CA; avg = 3455
CARB, 2005
3.8-36.9
Range, n=9 (3 days. 3 samplers) outdoors in
auto shredder facility in CA
CARB, 2005
4.0
N=84 (12 sampling dates, 7 monitors); average
for 2004 in Bay Area and South Coast of CA
CADAMP, 2006
9.1,6.1,5.2
Geo. mean of 20 homes, 3 loc/home in Boston,
MA area; "personal," bedroom, and living room
Allen et al. (2007)
155
ND
Outdoors at UC/Davis. 2004; average of 2 days
of filter/PUF and filter/XAD samplers
CARB, 2005
ND
Indoors in computer facility, average of 2 days,
8 measurements, 6 with computers on, 2 with
computers off, in CA
CARB, 2005
ND
Range, n=12 (3 days, 4 samplers) outdoors
surrounding electronics recycling facility in CA
CARB. 2005
ND
Range, n=6 (3 days, 2 samplers) indoors in
electronics recycling facility in CA
CARB, 2005
ND-0.6
Range, n=9 (3 days. 3 samplers) outdoors in
auto shredder facility in CA
CARB, 2005
166
ND
Outdoors at UC/Davis, 2004; average of 2 days
of filter/PUF and filter/XAD samplers
CARB, 2005
ND
Indoors in computer facility, average of 2 days,
8 measurements, 6 with computers on, 2 with
computers off. in CA
CARB, 2005
ND-2.5
Range. n=12 (3 days, 4 samplers) outdoors
surrounding electronics recycling facility in CA
CARB, 2005
35-463
Range, n=6 (3 days, 2 samplers) indoors in
CARB. 2005
4-69
-------
electronics recycling facility in CA
ND
Range, n~9 (3 days. 3 samplers) outdoors in
auto shredder facility in CA
CARB, 2005
HeptaBDE
181
ND
Outdoors at UODavis, 2004; average of 2 days
offilter/PUF and filter \.\l) samplers
CARB, 2005
ND
Indoors in computer facility, average of 2 days,
8 measurements, 6 with computers on, 2 with
computers off, in CA
CARB, 2005
ND
Range, n=12 (3 days, 4 samplers) outdoors
surrounding electronics recycling facility in CA
CARB, 2005
ND
Range, n=6 (3 days, 2 samplers) indoors in
electronics recycling facility in CA
CARB, 2005
ND
Range, n=9 (3 days, 3 samplers) outdoors in
auto shredder facility in CA
CARB, 2005
183
1.4
Outdoors at UC/Davis, 2004; average of 2 days
of filter/PUF and filter/XAD samplers
CARB, 2005
ND
Indoors in computer facility, average of 2 days,
8 measurements, 6 with computers on, 2 with
computers off, in CA
CARB, 2005
6-456
Range, n=!2 (3 days, 4 samplers) outdoors
surrounding electronics recycling facility in CA
CARB, 2005
5610-36700
Range, n-: 6 (3 days, 2 samplers) indoors in
electronics recycling facility in CA; avg —
23813
CARB. 2005
4.0-32.6
Range, n=r9 (3 days, 3 samplers) outdoors in
auto shredder facility in CA
CARB, 2005
1.4
N=~84 (12 sampling dates, 7 monitors): average
for 2004 in Bay Area and South Coast of CA
CAD AMP, 2006
190
<0.06 (DL)
3 Rural/remote sites in Ml, NY (n~36), 97-99
Strandberg et al,
2001
<0.06 (DL)
Chicago, n=l2, 97-99
Strandberg et al.,
2001
ND
Outdoors at UC/Davis, 2004; average of 2 days
of filter/PUF and tllter/XAD samplers
CARB, 2005
ND
indoors in computer facility, average of 2 days,
8 measurements. 6 with computers on, 2 with
computers off, in CA
CARB. 2005
ND-16
Range, n=12 (3 days, 4 samplers) outdoors
surrounding electronics recycling facility in CA
CARB, 2005
288-1300
Range, n=6 (3 days, 2 samplers) indoors in
electronics recycling facility in CA
CARB, 2005
ND-3.8
Range. n=9 (3 days, 3 samplers) outdoors in
auto shredder facility in CA
CARB, 2005
Nona 151) 11
203
ND
Outdoors at UC/Davis, 2004; average of 2 days
of filter/PUF and filter/XAD samplers
CARB, 2005
ND
Indoors in computer facility, average of 2 days.
8 measurements. 6 with computers on, 2 with
computers off, in CA
CARB, 2005
DeeaBDE
209
1.4
Michigan (n=35), 2002/3
I Ioh & Kites. 2005
60.1
Chicago (n=28). 2002/3
I Ioh & Mites. 2005
2.2
Indiana (n=38), 2002/3
Hoh & Ilites. 2005
4-70
-------
9.0
Arkansas (n=30), 2002/3
Hoh & Mites, 2005
2.6
Louisiana (n=26), 2002/3
Hoh & llites, 2005
<0.10 (DL)
3 Rural/remote sites in MI. NY (n=36). 97-99
Strandberg et al„
2001
0.30
Chicago, n=12. 97-99
Strandberg et al,
2001
10.6
Outdoors at UC'/Davis. 2004; average of 2 days
of filter/PUF and filter/XAD samplers
CARB, 2005
58
indoors in computer facility, average of 2 days.
8 measurements. 6 with computers on, 2 with
computers off, in CA
CARB, 2005
140-11400
Range, n= 12 (3 days, 4 samplers) outdoors
surrounding electronics recycling facility in
CA; mean = 2764
CARB. 2005
79700-833000
Range, n=6 (3 days, 2 samplers) indoors in
electronics recycling facility in CA; avg =
423.466
CARB, 2005
123-] 940
Range, n~9 (3 days, 3 samplers) outdoors in
auto shredder facility in CA; avg = 2,403
CARB, 2005
25.0
N=84 (12 sampling dates, 7 monitors); average
for 2004 in Bay Area and South Coast of CA
CAD AMP, 2006
174, 95, 94
Geo. mean of 20 homes, 3 loc/home in Boston.
MA area; "personal," bedroom, and living room
Allen et al. (2007)
4-71
-------
Table 4.3. C ongener-specific fish concentrations of PBDEs for fish caught in the United
States (units in ng/g wet weight, wwt. or lipid weight. Iwt, in parenthesis if available)
Congener
Concentration,
ng/g wwt {or Iwt)
Comment
Citation
TriBDE
17
0.011
N =24; consumer fish including tuna, salmon,
shark, trout, catfish, herring
Schecter et at, 2006
0.10(0.77 Iwt)
Time/spatial trend study for lake trout in
several Great Lakes; average for 2000 only
Zhu & Hites, 2004
28
0.01
Average of 2 farm salmon farm composited
from Maine; eyeball estimate from graph
Shaw et al., 2005
0.026
N~24; consumer fish including tuna, salmon,
shark, trout, catfish, herring
Schecter et al., 2006
1.67(10.21 Iwt)
Time/spatial trend study for lake trout in
several Great Lakes; average for 2000 only
Zhu & Hites, 2004
1.2 (5.4 Iwt)
Average of 88 specimens, 9 species of marine
fish, 2 species of dolphin, off Florida coast
Johnson-Restrepo et
al., 2005
TetraBDE
47
0.60
N=24; consumer fish including tuna, salmon,
shark, trout, catfish, herring
Schecter et al., 2006
3.6
N=17 locations in SF Estuary; bivalve (clam,
oyster, mussel) average
Oros et al„ 2005
0.43
Average of 2 farm salmon farm composited
from Maine; eyeball estimate from graph
Shaw et al., 2005
52.1
Average of 21 salmon samples from Lake
Michigan
Manchester-Nees v ig
et al.. 2001
63.72 (391.12 Iwt)
Time/spatial trend study for lake trout in
several Great Lakes; average for 2000 only
Zhu & Hites, 2004
2—110
Time/spatial trend study for lake trout/smelt
in Great Lakes; range for last samples. ~ 2003
Batterman et al..
2007
46.0 (488.3 Iwt)
Average of 88 specimens. 9 species of marine
fish, 2 species of dolphin, off Florida coast
Johnson-Restrepo et
al., 2005
1.5, 22, 84%
N=63; median, mean, frequency detection in
Wash State survey in rivers and lakes
Johnson et al., 2006
49
4.78 (35.35 Iwt)
Time/spatial trend study for lake trout in
several Great Lakes; average for 2000 only
Zhu & Hites, 2004
ND, 1.3.33%
N=60; median, mean, frequency detection in
Wash State survey in rivers and lakes
Johnson et al.. 2006
66
1.82 (11.29 Iwt)
Time/spatial trend study for lake trout in
several Great Lakes; average for 2000 only
Zhu & Hites, 2004
0.021
N=24: consumer fish including tuna, salmon,
shark, trout, catfish, herring
Schecter et al., 2006
1.7
Average of 21 salmon samples from Lake
Michigan
Manchester-Neesvig,
et al.. 2001
0.3 (1.8 Iwt)
Average of 88 specimens, 9 species of marine
fish, 2 species of dolphin, off Florida coast
Johnson-Restrepo et
al., 2005
ND. 1.0, 19%
N=27; median, mean, frequency detection in
Wash State survey in rivers and lakes
Johnson et al,, 2006
7)
ND
Time/spatial trend study for lake trout in
several Great Lakes; average for 2000 only
Zhu & Hites, 2004
ND, <0.5, 3%
N—63; median, mean, frequency detection in
Wash State survey in rivers and lakes
Johnson et al., 2006
4-72
-------
77
0,001
N=24; consumer fish including tuna, salmon,
shark, trout, catfish, herring; mostly na or nd,
with one 3.6 herring
Schecter et al 2006
PentaBDE
85
0.67 (4.31 Iwt)
Time/spatial trend study for lake trout in
several Great Lakes; average for 2000 only
Zhit & llites, 2004
0.004
N=24: consumer fish including tuna, salmon,
shark, trout, cattish, herring; mostly na or nd;
one catfish at 41.6
Schecter et al 2006
0.03
Average of 2 farm salmon farm composited
from Maine: eyeball estimate from graph
Shaw et at., 2005
0.1 (0.9 Iwt)
Average of 88 specimens, 9 species of marine
fish. 2 species of dolphin, off Florida coast
Jobnson-Restrepo et
al 2005
88
0.03
Average of 2 farm salmon farm composited
from Maine; eyeball estimate from graph
Shaw et al., 2005
99
22.96 (142.20 Iwt)
Time/spatial trend study for lake trout in
several Great Lakes; average for 2000 only
Zhii & llites, 2004
0.4 - 15
Time/spatial trend study for trout/smelt in
Great Lakes; range for last sampled, -2003
Batterman et al,
2007
0.17
N~24; consumer fish including tuna, salmon,
shark, trout, catfish, herring
Schecter et al., 2006
0.10
Average of 2 farm salmon farm composited
from Maine; eyeball estimate from graph
Shaw et al., 2005
9.3
Average of 21 salmon samples from Lake
Michigan
Manchester-Neesvig.
et al., 2001
1.2
N=I7 locations in SI Estuary; bivalve (clam,
oyster, mussel) average
Oros et al 2005
5.5 {23.1 hvt)
Average of 88 specimens, 9 species of marine
fish. 2 species of dolphin, off Florida coast
Johnson-Restrepo et
al 2005
ND, 17.0, 38%
N=63; median, mean, frequency detection in
Wash State survey in rivers and lakes
Johnson et al, 2006
TOO
17.50(109.51 Iwt)
Time/spatial trend study for lake trout in
several Great Lakes; average for 2000 only
Zhu & l lites, 2004
0.5 - 15
Time/spatial trend study for trout/smelt in
Great Lakes; range for last sampled, - 2003
Batterman et al.,
2007
0.13
N=24; consumer fish including tuna, salmon,
shark, trout, cattish, herring
Schecter et al 2006
0.08
Average of 2 farm salmon farm composited
from Maine; eyeball estimate from graph
Shaw et al., 2005
9,7
Average of 21 salmon samples from Lake
Michigan
Manchester-Neesvig,
etal., 2001
0.9
N=17 locations in SF Estuary: bivalve (clam,
oyster, mussel) average
Oros et al 2005
14.0(55.4)
Average of 88 specimens, 9 species of marine
fish, 2 species of dolphin, o ff Florida coast
Johnson-Restrepo et
al., 2005
1.0. 5.1, 51%
N=63; median, mean, frequency detection in
Wash State survey in rivers and lakes
Johnson et al., 2006
HexaBDE
138
ND
Time/spatial trend study for lake trout in
several Great Lakes; average for 2000 only
Zhu & llites. 2004
0,001
N=24; consumer fish including tuna, salmon,
shark, trout, catfish, herring: mostly nd except
2 catfish at 5.1 and 7.9
Schecter et al 2006
0.05
Average of 2 farm salmon farm composited
Shaw et al., 2005
4-73
-------
from Maine; eyeball estimate from graph
•-0.001
Average of 88 specimens, 9 species of marine
fish. 2 species of dolphin, off Florida eoast
Johnson-Restrep et al
2005
ND, <0.9, 2%
N^63; median, mean, frequency detection in
Wash State survey in rivers and lakes
Johnson et al., 2006
153
0.021
N=24; consumer fish including tuna, salmon,
shark, trout, catfish, herring;
Schecter et al 2006
4.43 (27.48 Iwt)
Time/spatial trend study for lake trout in
several Great Lakes: average for 2000 only
Zhu & llites, 2004
0.1 -3.6
Time/spatial trend study for trout/smelt in
Great Lakes; range for last sampled, - 2003
Batterman et al.
(2007)
3.8(17.0 Iwt)
Average of 88 specimens, 9 species of marine
fish, 2 species of dolphin, off Florida coast
Johnson-Restrepo et
al 2005
ND. 1.1,40%
N=63; median, mean, frequency detection in
Wash State survev in rivers and lakes
Johnson et al., 2006
154
0.049
N=24: consumer fish including tuna, salmon,
shark, trout, catfish, herring
Schecter et al 2006
7.63 (46.84 Iwt)
Time/spatial trend study for lake trout in
several Great Lakes; average for 2000 only-
Zhu & Hites, 2004
9.2 (2.6 Iwt)
Average of 88 specimens, 9 species of marine
fish, 2 species of dolphin, off Florida coast
Johnson-Restrepo et
al 2005
0.48. 0.88. 49%
N—63; median, mean, frequency detection in
Wash State survev in rivers and lakes
Johnson et al.. 2006
HeptaBDE
1 S3
0.002
N~24: consumer fish including tuna, salmon,
shark, trout, catfish, herring
Schecter et al 2006
0.13 (0.73 Iwt)
Time,'spatial trend study for lake trout in
several Great Lakes; average for 2000 only
Zhu & Hites, 2004
1.1 (3.1 I2t)
Average of 88 specimens, 9 species of marine
fish, 2 species of dolphin, off Florida coast
Johnson-Restrepo et
al 2005
ND, <0.9, 3%
N=63; median, mean, frequency detection in
Wash State survey in rivers and lakes
Johnson et al., 2006
184
ND, <0,9. 2%
N=60; median, mean, frequency detection in
Wash State survey in rivers and lakes
Johnson et al., 2006
190
ND
Time/spatial trend study for lake trout in
several Great Lakes; average for 2000 only
Zhu & Hites. 2004
191
ND, <0,9, 0%
N =60; median, mean, frequency detection in
Wash State survey in rivers and lakes
Johnson et al, 2006
NonaBDE
203 j 1.3 (4,3 iwt)
Average of 88 specimens, 9 species of marine 1 Johnson-Restrepo et
fish, 2 species of dolphin, off Florida coast | al 2005
DecaBDE
209
0.092
N--24; consumer fish including tuna, salmon,
shark, trout, catfish, herring: majority nd but
one catfish at 1269
Schecter et al 2006
0.5 (96.6 Iwt)
A verage of 88 specimens, 9 species of marine
fish, 2 species of dolphin, off Florida coast
Johnson-Restrepo et
al 2005
ND, <5.3, 6%
N=63; median, mean, frequency detection in
Wash State survev in rivers and lakes
Johnson et al.. 2006
4-74
-------
Table 4.4, Congener-specific concentrations of PBDEs in food originating from the
United States {concentrations in ng/g wwt or specifically identified as ng/g Iwt, if that
was how reported by the author and conversion to ng/g wwt was not possible).
Congener
Concentration,
ng/g wwt or Iwt
Food type; comment
Citation
TriBDE
17
0.0008
N=18 meat (lipid = 26.3%) including pork,
chicken, beef; dominated by NDs at DL = 0.7
Seheeter et al., 2006
0.2s-)
N=15 dairy (lipid = 10.2%) including cheese,
milk, infant formula, yogurt, ice cream
Schecter et al., 2006
17/25
0.008
9 chicken fat ("90% lipid) samples from three
production non-contaminated sites
Huwe, etal., 2003
0.008
4 chicken fat (90% lipid) samples from a
contaminated production site
Huwe, et al., 2003
28
0.005
N-~18 meat (lipid = 26.3%) including pork,
chicken, beef; dominated by one beef at 59.7
Schecter et al., 2006
0.79
N= 15 dairy (lipid * 10.2%) including cheese,
milk, infant formula, yogurt, ice cream
Schecter et al., 2006
33
0.017
9 chicken fat (90% lipid) samples from three
production non-contaminated sites
Huwe, et al., 2003
0.028
4 chicken fat (90% lipid) samples from a
contaminated production site
Huwe. etal., 2003
28/33
0.004 Iwt
Market basket hamburger (n= 11), 10 locations
around the country, sampled in 2004
1 luwe & Larsen 2005
0.004 Iwt
Market basket bacon (n=l 1), 10 locations
around the country, sampled in 2005
Huwe & Larsen 2005
0.003 hvt
Market basket chicken fat (n=22), 10
locations around the country, sampled in 2005
Huwe & Larsen 2005
0,007 Iwt
Market basket pork fat (n= 11), 10 locations
around the country, sampled in 2005
Huwe & Larsen 2005
0.002 Iwt
Market basket beef fat us M0. 10 locations
around the country, sampled in 2004
Huwe & Larsen 2005
FetraBDE
47
2.26
9 chicken fat ("90% lipid) samples from three
production non-contaminated sites
Huwe. et al., 2003
6.87
4 chicken fat (90% lipid) samples from a
contaminated production site
Huwe, et al., 2003
0.093
N=18 meat ( lipid = 26.3%) including pork,
chicken, beef;
Schecter et al., 2006
0.032
N=I5 dairy (lipid = 10.2%) including cheese,
milk, infant formula, yogurt, ice cream
Schecter et al. 2006
0.18 Iwt
Market basket hamburger (n=l 1), 10 locations
around the country, sampled in 2004
Huwe & Larsen 2005
0.23 Iwt
Market basket bacon (n=l 1), 10 locations
around the country, sampled in 2005
Huwe & Larsen 2005
0.81 Iwt
Market basket chicken tat(n=22). 10
locations around the country, sampled in 2005
Huwe & Larsen 2005
1.07 Iwt
Market basket pork fat (n=l 1). 10 locations
around the country, sampled in 2005
Huwe & Larsen 2005
0.07 Iwt
Market basket beef fat (n= 10), 10 locations
around the country, sampled in 2004
Huwe & Larsen 2005
4-75
-------
66
0.00
9 chicken fat (90% lipid) samples from three
production non-contaminated sites
lluwe, et al,, 2003
0.018
4 chicken fat (90% lipid) samples from a
contaminated production site
Huvve, et al., 2003
1.19
N=18 meat (lipid = 26.3%) including pork,
chicken, beef;
Schecter et al, 2006
0.61
N= 15 dairy (lipid = 10.2%) including cheese,
milk, infant formula, yogurt, ice cream
Seheeter et al.. 2006
77
0.83
N=18 meat (lipid = 26.3%) including pork,
chicken, beef; mostly ND or NA
Schecter et al., 2006
0.10
NM5 dairy (lipid = 10.2%) including cheese,
milk, infant formula, yogurt, ice cream; all nd
or na
Schecter et al., 2006
PcntaBDE
85
0.11
9 chicken fat (90% lipid) samples from three
production non-contaminated sites
lluwe, et al, 2003
0.52
4 chicken fat (90% lipid) samples from a
contaminated production site
Huwe, et al., 2003
4,93
N~l 8 meat (lipid = 26.3%) including pork,
chicken, beef;
Schecter et al„ 2006
1.08
N~15 dairy (lipid -- 10.2%) including cheese,
milk, infant formula, yogurt, ice cream; most
nd or na except one cheese at 5.52
Schecter et al., 2006
0.022 hvt
Market basket hamburger (n=l 1), 10 locations
around the count 17, sampled in 2004
l luwe & Larsen, 2005
0.023 Iwt
Market basket bacon (11= 11 ), 10 locations
around the country, sampled in 2005
Huwe & Larsen, 2005
0.047 hvt
Market basket chicken fat (n=22), 10
locations around the country, sampled in 2005
Huwe & Larsen, 2005
0.034 Kvt
Market basket pork fat
-------
contaminated production site
0.023
N=~18 meat (lipid = 26.3%) including pork,
chicken, beef;
Schecter et al„ 2006
0.005
N=15 dairy (lipid = 10.2%) including cheese,
milk, infant formula, yogurt, ice cream
Seheeter et al., 2006
0.042 lwt
Market basket hamburger (n~l IK 10 locations
around the country, sampled in 2004
1 luwe & Larsen. 2005
0.041 lwt
Market basket bacon (n—11), 10 locations
around the country, sampled in 2005
lluwe & Larsen, 2005
0,28 lwt
Market basket chicken fat (n=22), 10
locations around the country, sampled in 2005
litiwe & Larsen, 2005
0.18 lwt
Market basket pork fat (n=l 1), 10 locations
around the country, sampled in 2005
lluwe & Larsen, 2005
0.02 ng/g lwt
Market basket beef fat (n=10). 10 locations
around the country, sampled in 2004
lluwe & Larsen, 2005
HexaBDE
138
0.02
9 chicken fat (90% lipid) samples from three
production non-contaminated sites
Huwe, et al.. 2003
0.16
4 chicken fat (90% lipid) samples from a
contaminated production site
Huwe, et al., 2003
0.002
N= 18 meat (lipid — 26.3%) including pork,
chicken, beef;
Schecter et al., 2006
0.0003
N= 15 dairy (lipid « 10.2%) including cheese,
milk, infant formula, yogurt, ice cream; most
nd or na with high DL
Schecter et al., 2006
140
0.003
9 chicken fat (90% lipid) samples from three
production non-contaminated sites
Huwe, et al., 2003
0.015
4 chicken fat (90% lipid) samples from a
contaminated production site
Huwe, et al, 2003
153
060
9 chicken fat (90% lipid) samples from three
production non-contaminated sites
Huwe, et al., 2003
2.90
4 chicken fat (90% lipid) samples from a
contaminated production site
Huwe, et al., 2003
0.021
N= 18 meat (lipid = 26.3%) including pork,
chicken, beef;
Schecter et al., 2006
0,004
N=15 dairy (lipid = 10.2%) including cheese,
milk, infant formula, yogurt, ice cream
Schecter et al., 2006
0.10 lwt
Market basket hamburger (n=I 1). 10 locations
around the country, sampled in 2004
Huwe & Larsen. 2005
0.078 Kvt
Market basket bacon (n= 11), 10 locations
around the country, sampled in 2005
Huwe & Larsen. 2005
0.22 lwt
Market basket chicken fat (11=22), 10
locations around the country, sampled in 2005
Huwe & Larsen, 2005
0.12 lwt
Market basket pork tat (n=l 1), 10 locations
around the country, sampled in 2005
Huwe & Larsen, 2005
0.019 lwt
Market basket beef fat (n= 10), 10 locations
around the country, sampled in 2004
Huwe & Larsen, 2005
154
0.18
9 chicken fat (90% lipid) samples from three
production non-contaminated sites
Huwe, et al., 2003
0.55
4 chicken fat (90% lipid) samples from a
contaminated production site
Huwe, et al,, 2003
0.014
N=18 meat (lipid = 26.3%) including pork,
chicken, beef;
Schecter et al.. 2006
0.002
N--15 dairy (lipid = 10,2%) including cheese,
Schecter et al.. 2006
4-77
-------
milk, infant formula, yogurt, ice cream
0.029 Kvt
Market basket hamburger (n=l 1), 10 locations
around the country, sampled in 2004
lluwe & Larsen, 2005
0,042 Kvt
Market basket bacon (n~l 1). 10 locations
around the country, sampled in 2005
Uuwe & Larsen, 2005
0.088 Kvt
Market basket chicken fat (n=22), 10
locations around the country, sampled in 2005
Huwe & Larsen, 2005
0.087 Iwt
Market basket pork fat (n=I 1), 10 locations
around the country, sampled in 2005
l luwe & Larsen, 2005
0.011 Iwt
Market basket beef fat (n=I0), 10 locations
around the country, sampled in 2004
lluwe & Larsen, 2005
HeptaBDE
183
0.19
9 chicken fat (90% lipid) samples from three
production non-contaminated sites
l luwe, et al„ 2003
0.34
4 chicken fat (90% lipid) samples from a
contaminated production site
l luwe, et al„ 2003
10.1
N = 18 meat (lipid = 26.3%) including pork,
chicken, beef;
Schecter et al, 2006
1.86
N~15 dairy (lipid = 10.2%) including cheese,
milk, infant formula, yogurt, ice cream
Schecter et al., 2006
0.029 hvt
Market basket hamburger (n~ 1 1), 10 locations
around the country, sampled in 2004
l luwe & Larsen, 2005
0.12 Iwt
Market basket bacon (n=l 1), 10 locations
around the country, sampled in 2005
Uuwe & Larsen, 2005
0.13 Kvt
Market basket chicken fin (n~22), 10
locations around the country, sampled in 2005
Uuwe & Larsen, 2005
0,084 Kvt
Market basket pork fat (n=l 1), 10 locations
around the country, sampled in 2005
1 luwe & Larsen, 2005
0.012 Iwt
Market basket beef fat (n= 10), 10 locations
around the country, sampled in 2004
Huwe & Larsen, 2005
DecaBDE
209
1.24
9 chicken fat (90% lipid) samples from three
production non-contaminated sites
Huwe, et al., 2003
0.72
4 chicken fat (90% lipid) samples from a
contaminated production site
Huwe, et al., 2003
0,053
N= 18 meat (lipid = 26.3%) including pork,
chicken, beef; dominated by I turkey at 245
Schecter et al., 2006
0.041
N = 15 dairy (lipid =' 10,2%)) including cheese,
milk, infant formula, yogurt, ice cream;
dominated by 1 sample at 481; others at < 20
Schecter et al., 2006
4-78
-------
Table 4.5. Exposure media concentrations (note:
— means no data available; ND means data available but not detected.)
Expusure
Media
17
28
47
66
85
99
1U0
138
153
154
183
197
206
209
Total
Referenee; n
Water, pg/1
6.6
3.3
42.7
2.2
1.3
27.6
7.2
0.3
3.9
2.9
4.4
0.1
1.4
42.3
146.1
Oros et al.
(2005); n=33a
Surface
soil, ng/g
dwt
1.9
<0.1
<0.1
3.6
0,4
5,7
4.8
37.4
12.4
0.8
15.3
82,3
Offenberg et al.
(2006)
Indoor
dust, na'"g
—
ND
1857
21
100
2352
911
Tsi
243
156
60
—
—
2394
l~~8275
EWG (2004);
mean 10 homes
Outdoor
air, pg/m3
¦->
3
53
—
2
51
13
—
4
4
1
—
—
25
158
CARB (2005);
n=84
Indoor air,
pg/mJ
8
27
177
4
3
79
16
5
7
121
447
Allen et al.
(2006); Webster
(2006); n==8b
Shellfish,
ng/g wwt
ND
ND
3.6
ND
ND
1.2
0.9
ND
ND
ND
ND
ND
ND
ND
5,7
Oros et al.
(2005); n=l T
Finfish,
ng/g wwt
0.01
0.03
0.60
0.02
0.004
0.17
0.13
0.001
0.02
0.05
0,002
—
—
0.09
1.17
Schecter et al.
(2006); n=24
Beef, ng/g
wwt
0.0005
0.02
0.05
0.0002
0.0006
0.04
0.006
0.0001
0.006
0.004
0.001
—
—
0.003
0.13
Schecter et al.
(2006); n=3
Pork, ng/g
wwt
0.0001
ND
0.08
ND
0.005
0.12
0.015
0,001
0.02
0.01
0.009
—
—
0.02
0.28
Schecter et al.
(2006); n=6
Poultry,
ng/g wwt
0.0001
0.0002
0.06
0.0003
ND
0.12
0.03
0.002
0.02
001
002
—
—
0.12
0.36
Schecter et al.
(2006); n=3
Dairy, ng/g
wwt
<1)001
0.0002
0.03
0.0003
0.0008
0.03
0.005
<0.0001
0.004
0.002
0.002
—
—
0.04
0.11
Schecter et al.
(2006); n 15
Eggs, ng/g
wwt
0.0001
0.0002
0.02
0.0002
0.002
0.04
0.006
0.0001
0.004
0.003
0.0001
—
—
0.01
0.09
Schecter et al.
(2006); n-6
a: for Oros et al, (2005), concentrations reported as "Q" meaning "detected, but not reportable because outside QA limits" not counted in averaging; "bdl"
meaning below detection limit counted as 0; b: Allen et al. (2006) and Webster (2006) sampled 3 locations within 20 urban residences, and presented geometric
means for the three locations over all 20 homes. Results presented are the average of the 3 geometric mean values; c: Oros et al. (2005) claims that only BDEs
47, 99, and 100 were detected in sampling of clams, oysters, and mussels - that others were measured but not detected - from SF Estuary.
4-79
-------
Chapter 5 HUMAN EXPOSURE
5.1. INTRODUCTION
The rise in PBDEs in breast milk in the United States throughout the 1990s in the
2000s, coupled with the finding that American, breast milk concentrations exceed those
of European women by factors of 10 or more, has served to focus attention on United
States exposures to PBDEs. The first section of this chapter reviews the data on body
burdens of PBDEs, with a focus on data from the United States. Robust data are
available on blood and breast milk, while limited data are available on other matrices—
including adipose tissue and liver. The section on body burdens concludes with
assignment of representative PBDE congener background profiles in blood and mother's
milk. The next section reviews estimates of dose of PBDEs currently available in the
literature. This is followed by development of an estimate of background dose using the
environmental media concentration profiles developed in Chapter 4, in combination with
exposure contact rates. The dose estimate made here is compared to the literature
estimates of dose. The next section of the chapter attempts to use a simple
pharmacokinetic framework to see if the dose estimates can explain body burdens of the
individual congeners. Figure 5-1 depicts this approach. In combination with mother's
milk concentrations, the same PK model was used to evaluate the impact of breast-
feeding on the PBDE body burden of infants. The chapter concludes with a series of
findings from this examination of United States exposures to PBDEs.
5.2. BODY BURDEN DATA
Data from around the world on PBDE body burdens, with an emphasis on data
from the United States are reviewed in this section. Subsections on blood and breast milk
conclude with a table displaying individual congener data; data on adipose and other
tissues are too sparse to warrant a table of raw data. This section on body burden data
concludes with a table suggesting representative profiles in blood and mother's milk for
selected congeners.
5-1
-------
5.2.1. Blood Data
Twelve studies that contained data on PBDEs in blood in the United States were
located. Table 5-1 shows congener-specific data from these.
The most statistically rigorous and expansive study of background exposures to
PBDEs is a recent analysis of 2003/4 NHANES data by Sjodin ct al. (2008).
Unfortunately, BDE 209 was not measured in this NHANES study. A total of 2,040
serum samples from individuals 12 years of age and older were analyzed for 10 BDE
congeners, including BDEs 17, 28, 47, 66, 85, 99, 100, 153, 154, and 183. The geometric
mean concentrations over the entire population were, in ng/g lipid weight (Iwt),
descending order: 20.5 for BDE 47, 5.7 for BDE 153, 5.0 for BDE 99. 3.9 for BDE 100,
and 1.2 for BDE 28. Geometric means were not provided for the other BDE congeners
because they were detected at less than 60%, including the low frequencies of 5, 15, 21,
and 23% for BDEs 17, 183, 66, and 85, respectively. The sum of these geometric means
was 36.3 ng/g lwt. The 95lh percentile for the sum of PBD Es was 291 ng/g hvt, and the
maximum found was 3,680 ng/g lwt, with BDE 47 at 2,350 ng/g hvt for this individual.
A statistically significant relationship between age and concentration was found for
BDEs 28, 47, 99, 100, and 153. Specifically, the highest concentrations were found in
the age group 12-19 years, with lower concentrations for 20-39, and 40-59, but then a
rise for the category >60 years. For example, the geometric mean concentration of BDE
47 for these four age categories, respectively, was 28.2, 21.5, and 1 7.7, and then a rise to
19.7 ng/g lwt. Although not statistically significant, the geometric mean concentrations
for males were higher than females for BDEs 47, 99, 100, 153, and 154. Although BDE
47 was found most frequently and at the highest concentrations for the survey as a whole,
10.5% of participants had BDE 153 concentrations that were higher than BDE 47. Race
was identified as a statistically significant factor for BDE 99; the geometric mean for
non-hispanie blacks was the highest at 6.2 ng/g lwt, Mexican Americans were at 5.9 ng/g
lwt, and non-hispanic whites were at 4.7 ng/g lwt.
Other than NHANES, the most comprehensive study of PBDEs in blood was
conducted by Sehecter et al. (2005). Although limited geographically, this study
contained two pooled samples of 100 individuals each and 39 samples of blood from
individuals in Mississippi and New York. Also, the study included an archived blood
5-2
-------
sample from 1973. which was from a pool of 100 individuals from Dallas, Texas, All
samples were analyzed for 13 BDEs including 17, 28, 47, 66, 77, 85, 99, 100, 138, 153,
154, 183, and 209, All congeners were non-detects for the 1973 sample. The 2 pooled
samples collected in 2003, one serum and one whole blood, both also n = 100 as noted,
were from discarded samples from the University ofTexas Southwestern Medical Center
in Dallas. The 39 individuals sampled in 2003 included 29 from Mississippi and 10 from
New York. The results from the three groups of samples from 2003 were fairly similar:
total concentrations of the 13 congeners were 61.8 ng/g Ivvt for the serum pool, 79.7 ng/'g
Iwt. for the whole blood pool, and 52.6 ng/g hvt as the mean for the 39 individuals. The
congener-specific trends were similar as well: BDE 47 dominated the profile by
encompassing between 44 and 53% of the total, with BDE 99 and 153 comprising about
the same amounts - BDE 99 was between 14 and 21% and 153 was between 11 and 20%.
From the individual samples, an interesting trend was that women had higher
concentrations than men: the range and mean of total BDE in 22 men were 4.6-192.8
ng/g Iwt and 25.1 ng/'g Iwt., and for 17 women, the range and mean of total BDE were
5.6-365.5 ng/g hvt and 74.1 ng/g Iwt. Although women had a higher level of PBDEs in
their blood than men in this study, Schecter et al (2005) stated that their results were not
statistically significant. BDE 209 was found at low levels in the pooled blood, 1.4 ng/'g
Iwt, and in the individual samples, it averaged 1.7 ng/g Iwt with non-detects in 19 of 39
individual samples.
Sjodin et al. (2004a) conducted a more rigorous temporal evaluation by collecting
samples that represented different time frames from the mid-1980s until the early 2000s.
Specifically, scram pools were collected in the southeastern United States from a blood
bank in Memphis, TN, representing years 1985-1997, and 2002, and serum pools were
collected in Seattle, WA, representing years 1999-2002. The pooled samples were
clustered to represent the following time periods: 1985-89 (n=9), 1990-94 (n=14), 1995-
99 (n=10), and 2000-02 (n-7), and the samples were measured for BDEs 47, 85, 99, 100,
153, and 154. A clear trend was seen with total concentrations rising from 9.6 ng/g Iwt in
the 1985-89 time frame, to 48 ng/g Iwt in 1990-94, 71 ng/g Iwt. in 1995-99, and 61 ng/g
Iwt in 2000-02. BDE 47 comprised 55-65% of total in the 4 time frames. In contrast to
BDEs, BB-153 (a marker for polybrominated biphenyls) and CB-153 (a marker for
polychlorinated biphenyls) decreased in the samples over time.
5-3
-------
Sjodin et al. (2001) provided another data set from the latter part of the 1980s
showing similarly low concentrations of PBDEs. Twelve samples from United States
donors who provided blood at a commercial blood collection facility in the state of
Illinois in 1988, and then stored at -70°C, were retrieved for analysis. Seven BDEs were
quantified: 47, 99, 100, 153, 182, 203, and 209. Three unidentified octa BDE and three
unidentified nona BDEs were noted. The data was reported in units of pmol/g Iw,
converted to ng/g lwt by multiplying by the congener's molecular weigh) in pg mole, and
then pg converted to ng by a multiplication of 0.001 ng/pg. Although total concentrations
were not provided—only median and range of concentrations for individual congeners
were provided—an estimate of a median, total concentration developed as a sum of the
medians of individual congeners was 2.7 ng/g lwt. This is much lower than the 50-80
ng/g lwt total found by Schecter et al. (2005) from samples taken in 2003. Of interest is
the finding of BDE 209 in 5 of 12 samples, with positives ranging from 1.5 - 33.6 ng/g
lwt. It was stated that BDE 209 has a short half-life in humans, 6.8 days, so the presence
here suggests continual exposure near the time of blood sample collection. Although not
stated by the authors, its presence could also be laboratory contamination Also of note is
that the congener most consistently found of the other 6 was BDE 153, found at a range
of 0.1 - 2.0 in all 12 samples.
Adipose tissue and serum was sampled from two disparate cohorts of women who
were sampled in the late 1990s, and it was compared with a third cohort of women
sampled between 1959 and 1967 (Petreas et al, 2003). One set of 32 adipose tissue
samples were from women undergoing surgery for breast cancer between 1996 and 98.
The second set was serum from a group of 50 Laotian women of reproductive age living
in the SF Bay area, taken in the 1997-99 time frame. The final set was serum from a
study of pregnant women enrolled in a case-control study of cryptorchidism and
hypospadias as part of the Child Health and Development Studies (CHDS), taken
between 1959 and 1967. Only BDE 47 could be quantified in this study, so it was the
only congener measured and reported. Like the Schecter et al. ( 2005 ) finding of mostly
non-detects in blood sampled from 1973, the entire set of 420 scrum samples from 1959-
1967 did not have any detections of BDE 47 at its high quantitation limit of 10 ng/g lwt.
The mean and median from the adipose samples was 29.9 and 16.5 ng/g lwt, respectively,
with 100% quantifiable measurements. The reproductive study revealed a mean and
-------
median of 50,6 and 10 ng/g Iwt with 48% quantified. There was no relationship between
BDE 47 concentrations and age in both the adipose tissue and reproductive studies. This
contrasts their finding of an increase of PCB 153 (which was also measured in these
samples) with age in the adipose and reproductive studies.
Ninety-three anglers who had sufficient blood volume and who completed a fish
consumption questionnaire were sampled between 2001 and 2003 (Moriand ct al., 2005).
The urban anglers were from New York: and New Jersey. Analysis was conducted for
BDEs 47, 85, 99, 100, 153.154, and 183. BDE 209 could not be reported because of
high background contamination during the processing of the unknown samples. Fish
eaters were categorized as "none" (eating no locally caught fish) or "any" (eating some
locally caught fish), arid "any" was further subcategorized based on amount of fish meals
per month of locally caught fish. The highest congener found was BDE 47, at a
geometric mean of 13.3 ng/g Iwt, followed by BDE 99, at 3.2 ng/g lwt, BDE 153
similarly at 3.2 ng/g lwt, and BDE 100 at 2.7 ng/g lwt. All other congeners were not
detected or very infrequently detected with geometric means less than 1 ng/g lwt.
Although total concentrations found were not discussed, a sum of the geometric means of
the congeners was 24.5 ng/g lwt. The straight mean might be higher because of the
presence of a few very high concentrations. BDE 47 was found at a high of 1,388 ng/g
lwt, and the high BDE 99 concentration was 546 ng/g lwt, for example. There were
moderate, but statistically insignificant, increases in BDE concentrations from no local
fish intake to > 1 meal/week.
Focant ct al. (2004) describe a unique methodology for analysis of PBDEs and
other PBTs. This method uses comprehensive two-dimensional gas chromatography and
isotope dilution time-of-flight mass spectrometry (GC/GC-IDTOFMS) for the
simultaneous measurement of selected polychlorinated biphenyls (PCBs), organochlorine
pesticides (OCPs), and brominated flame retardants. Unlike classic GC/MS, this method
evaluates all contaminants simultaneously with one injection into the GC column. Three
milk samples and one blood sample were analyzed by Focant et al. (2004). The blood
sample was from a pooled sample collected from 15 individuals in 2002 in three cities:
Philadelphia, Memphis, and Miami. BDEs 17, 28, 47, 66, 85, 99, 100, 153, and 154 were
analyzed. The blood sample results were expressed in pg/g fwt from blood serum, and
were only presented on a graph. The results were estimated from the graph, and
-------
converted to ng/g Kvt, assuming 0.65% lipid in the serum. The total BDE from the one
sample was 52.7 ng/g lwt, dominated by BDE 47 at 28.1 ng/g Iwt, with comparable
contributions by 99, 100, and 153 at 9.2, 6.9, and 6.2 ng/g lwt, respectively.
Twelve paired samples of maternal and cord blood were obtained from a hospital
in Indianapolis during Aug-Dec, 2001, and analyzed for BDEs 47, 99, 100, 153, 154, and
183 (Mazdi et al., 2003). Results for maternal and cord blood were essentially identical:
the range and median of total PBDE for mother's blood was 15 to 480 ng/g lwt and 37
ng/g lwt, respectively, and the corresponding range and median for infant blood was 14 to
460 ng/g lwt and 39 ng/g Kvt, respectively. BDE 47 accounted for 53-64% of total
PBDEs; BDEs 99, 100, and 153 each contributed 10-15% or total. BDE 154 and 183
were found rarely and at low levels. There was no age or BMI relationship with total
BDEs. The authors claimed that the blood concentrations were 20-100 times higher than
a similar population of Swedish mothers and children.
Wolff et al. (2005) conducted a study of exposures among mothers who were
pregnant near the World Trade Center (WTC) Site on September 11, 2001. The study
involved a complex evaluation of exposures, including measurement of key persistent
contaminants including polycyclic aromatic hydrocarbons, polychlorinated biphenyls,
dioxins, and PBDEs in blood. The authors did not find an association of PBDEs and
measures of potential exposure to WTC contaminants, and they generally found low
levels of PBDEs. Of 100 mothers, they found median levels for BDEs 28, 47, 99, 100,
and 153 at 0.65, 9.7, 1.5, 1.8, and 1.8 ng/g lwt respectively.
Bradman et al. (2006) reports on a sampling of blood from pregnant Latina
women, primarily from Mexico, living in Salinas Valley, California. The serum
specimens were collected between September 1999 and January 2001. Seven BDEs were
measured, including 47, 85, 99, 100, 153, 154, and 183. The resulting concentrations
were fairly low, although not as low as the Wolff et al. (2005) study on pregnant women
during the World Trade Center attacks. The median total from the 24 women was 21
ng/g Kvt—with a high of 320 and a low of 5.3 ng/g lwt. Like other studies, the dominant
congener was BDE 47. The authors show that the levels are highest among women who
have spent less than 5 years in the United States, compared to women who have spent
more than 5 years in the United States.
-------
Fischer ct al, (2006) present a case study of PBDEs in a family that showed
somewhat elevated levels in the parents but higher levels in one child and still higher
levels in the toddler of the family. Samples were collected from a family of 4, including
35 and 37 year-old parents, a 5 year-old daughter, and an 18-month old son in September
and December of 2004. The sum of BDEs 47, 99, 100,153, and 154 in the parents ranged
between 64 and 147 ng/g lwt in the two sampling dates, and BDE 209 contributed a
relatively small addition at between 2 and 23 ng/g lwt. The story was much different
with the children. The 5 year-old daughter had concentrations of 237, 239/249 ng/g lwt
(the last two were duplicates of the same December sample) of the 5 congeners for the
September and then December samples but a disparate range of 143 ng/g lwt of BDE 209
in the September sample and 9/12 ng/g lwt in the December sample (duplicates). The
toddler had the highest concentrations of all: 418 and 488/476 ng/g lwt for the five
congeners and 233 and 19/26 ng/g lwt of BDE 209 in the September and December
samples, respectively. The authors discounted laboratory error and attributed the higher
concentrations in the children to exposure to house dust. While the authors have
discounted laboratory error, it would appear that a decline by an order of magnitude in
both the toddler and infant is substantial, and could be due to some difference in the two
laboratories. A decline of this magnitude is plausible because of the short half-life of
BDE 209 in humans - it has been quantified on the order of 15 days, while the half-lives
of lower brominated congeners has been quantified on the order of years. Whether due to
differences in laboratories or differences in exposures, BDE 209 was quantified by both
laboratories and this in itself is worthy of reporting. The higher levels of the other
congeners in the toddler were attributed to his consumption of breast milk, although
Fisher et al (2006) suggest that it might also be due to exposure to house dust. Of the
non-209 congeners, the typical trend of seeing BDE 47 at the highest concentrations was
true for all participants and sampling dates; BDE 153 was second most prevalent, ranging
from one-thrid to one-half of the concentration of BDE 47.
Schecter et al. (2006a) measured the concentrations of 12 BDEs (17, 28, 47, 66,
85. 99, 100, 138, 153, 154, 183, and 209) in the blood of 8 vegans (no animal food
products including dairy). They found a range of BDE concentrations of 12.4 to 127 ng/g
lwt total, with a median of 23.9 and a mean of 53.3 ng/g lwt. BDE 47 was the most
abundant congener, with the highest mean concentration of 23 ng/g lwt. The second
5-7
-------
highest was BDE 153 at 14 ng/g lvvt. BDEs 99 and 100 totaled about 6 ng/g lwt. BDE
209 was not detected, although detection limits were relatively high at between 2 and 7
ng/g lwt. The authors characterize these findings as lower, but not substantially lower,
than other studies of BDEis in the blood of Americans, and they suggest that this could be
the result of not consuming food of animal origin. They note that for dioxins and other
POPs, foods of animal origin have been attributed as the main source of exposure, and
because the concentrations of BDEs in vegans, who have been so for a minimum of 5
years, is not that much lower than other populations, exposures other than food may be
important for this class of compounds.
These data suggest a range of approximately 30-100 ng/g lwt total BDEs is
representative of blood in the general population of Americans in the 2000s, with
occasional measurements in the 100s and 1,000s of ng/g lwt. Sampling of cord and
maternal blood, as well as mother's milk, in Canada suggests somewhat lower levels.
Ryan and Oostdam (2004) collected small amounts of individual maternal and cord blood
samples from the North West Territories of Canada that were part of a "Northern
Contaminants Program", composited them, and measured the composites for BDE
congeners including 28, 47, 85, 99, 100, 153, 154, and 183. These samples pertained to
years 1994 to 1999. In ten maternal composite samples, the mean total PBDE
concentration was 23.3 ng/g lwt, with a range of 13.1 to 46.5 ng/g lwt, and the cord blood
averaged 12 ng/g lwt (range or individual composite samples not provided). In the
mother's blood plasma, BDE 47 comprised about 40% of total whereas in cord blood,
BDE 47 comprised about 77%.
In contrast to the North American studies above, most studies from Europe, Asia,
and elsewhere suggest concentrations of total BDEs in blood less than 10 ng/g lwt.
Studies below include the countries of Sweden, Norway, New Zealand, the United
Kingdom, Spain, the Faroe Islands, Japan, and Nicaragua.
Blood from 37 Swedish men were sampled in 1991 and 2001, and measured for
PBDEs 28, 47, 99, 100, 128, 154, 153, 183, 196, 197, 203, 206, 207, 208, and 209
(Jakobbson et al., 2005). An additional 10 men were sampled in 1988 and 2002. These
men were specifically selected to represent levels of fish consumption, so they do not
necessarily represent a cross-section of the average populat ion. The median and range of
total concentrations from 1991 and 2001 was 11 (3.3-59) and 14 (4.2-57), respectively,
-------
expressed in units of pmol/'g Iwt. Because individual results were not provided, these
totals could not be converted to ng/g hvt. However, the conversion would result in
concentrations that are somewhere between about one-fourth and three-fourths the listed
concentrations as the conversion factors for individual congeners range from 0.486
[ng/g]/[pmoI/g] for BDE 47 to 0.959 [ng/g]/[pmol/g] for BDE 209, BDE 47 was the
dominant congener in 1991, which explains about 19% of total concentrations. BDEs
154 and 209 explained 12% and BDE 153 accounted for 8%. BDE 153, however, was
the dominant congener in 2001; 21% of the mean. BDE 47 explained 11%, and BDEs
209 and 154 accounted for 10% and 8% respectively. The finding of an increase in BDE
153, and the meaningful contribution of BDE 209 are noteworthy in this study. The
authors note a decline of other POPs like I B-153, p,p'-DDE and hexachlorobenzene
between 1991 and 2001 of between 30-50%, while no decrease and a small increase was
noted for BDEs. A second study in Sweden (Karlsson et al, 2007) looked at levels in air,
dust, and blood from individuals in five households. Concentrations of the tri-hexa
brominated congeners (BDEs 28, 58, 66, 99, 100, 153, and 154) were near the detection
limit for 4 of the 5 individuals, with concentrations below 10 ng/g lwt for all
measurements and the tri-hexa total at less than 15 ng/g lwt for all individuals. BDE 209
was detected in 4 of 5 individuals, at concentrations ranging from about 9,4 to 17,4 ng/g
lwt. One individual also had quantified measurements of BDEs 197/204 (co-eluting).
196, 206, and 207 at concentrations ranging from 3,5 to 9.7 ng/g lwt. While the levels
overall are generally low compared to levels found in the United States, they are
noteworthy like the other study from Sweden discussed here in that the profiles had
significant contributions from BDE 209 and even other nona and decaBDE congeners.
Pooled samples of about 20 individuals each from 5 hospitals in Norway (total
number of samples analyzed was not provided) were analyzed for 11 BDEs: 28, 37, 47,
85, 99, 100, 119, 138, 153, 154. and 183 (Thomsen et al., 2005a). Samples for the years
1977, 1982, 1988, 1991, 1994, 1997, 1998, 1999, 2000, 2001, 2002, and 2003 were
obtained. The sum of the seven most abundant congeners (28, 47, 99, 100, 153, 154, and
183) showed a concentration range of 0.5 to 5.0 ng/g lwt, with a clear trend of low
concentrations for the early years: 0,5 in 1977, 1.3 in 1982, etc, which consistently rose
to high levels between 3.6 to about 5.5 ng/g lwt between 1997 and 2003. A spike of
about 5 ng/g lwt in 1991 could not be explained. BDE 47 was found in the highest
5-9
-------
concentration except for one of the samples; the relative amount of BDE 153 appears to
be increasing over time. Results for BDE 209 were presented, but the article is not
consistent in its reporting of BDE 209 concentrations. The text suggests a median BDE
209 of 8.7 ng/'g lwt, and the figures in the article show a remarkably high measurement of
35 ng/g lwt in 2000. These levels appear so much higher than others that even the
authors suggest, "contamination of the sample cannot be totally excluded."
Harrad and Porter (2007) report on concentrations of BDEs including 47, 99, 100,
153, 154, and 183 in the blood of 23 individuals (10 males, 13 females age 20 to 64)
sampled in 2001 in Wellington, New Zealand. The mean concentration of total BDEs
was similar to other European studies at 7.17 ng/g lwt. Also similar to other studies was
the finding that BDE 47 concentrations dominated, accounting for over 50% of the
concentration. The second most found congener was BDE 153, explaining between 15
and 20% of concentrations. The authors claim the exposure was due to imported
consumer goods, because these products are not produced in New Zealand.
Thomas et al. (2006) measured BDEs in the blood of 154 volunteers in 13
locations in the United Kingdom in 2003. They measured 22 congeners including BDE
209. The median total BDE was 5.6 ng/g lwt. with a range of 0.6 to 420 ng/g lwt, and
only 5% greater than 30 ng/g lwt. BDE 209 was quantified in only 11 samples, with a
high of 240 ng/g lwt. although the detection limit was high at 15 ng/g lwt. BDEs 47, 99,
100, 153, 154, and 183 were regularly detected, similar to other studies, but the median
concentration, of BDE 153 was the highest at 1.7 ng/g lwt, followed by BDE 47 at 0.82
ng/g lwt. This is atypical because BDE 47 is most often the highest found in profiles.
Gomara el al (2007) report on the sampling of PBDEs in human umbilical cord
serum, maternal and paternal serum, placentas, and breast milk from individuals living in
two locations (Vallecas and Getafe) in Madrid, Spain. The sampling occurred between
October 2003 and May 2004, and involved 391 individual samples including 113 of
maternal serum, 104 of paternal serum, 92 of umbilical cord serum, 30 of placenta, and
52 of breast milk. Fifteen individual congeners were measured in all samples, including
BDE 209. The maternal, paternal, and umbilical cord serum samples had medians which
ranged narrowly between 9.7 and 17 ng total PBDE/g lwt. BDE 47 was the predominant
congener in the serum samples.
5-10
-------
Blood concentrations of pregnant Faroese (the Faroe Islands are between Shetland
and Iceland) women were determined from samples taken in 1994, and then their
children's blood was sampled and measured 7 years later, in 2002 (Fangslrom et al.,
2005). Fifty-seven mothers and 42 children were sampled, of which 41 were
mother/child pairs. BDEs 47, 99, 100, 153, 209, and 153/154 were measured.
Concentrations were low, with a median total concentration of just over 5 ng/g Iwt for
both mothers and children. The predominant congener for the mother was BDE 47 and
the co-eluting BDEs 153/154, accounting for 26% each. For children, the predominant
congener was 153, explaining about 46% of the total concentration. BDE 209 was
present in both mothers and children at low concentrations of 0,8 and 1.0 ng/g lwt,
respectively.
Samples of maternal blood plasma, cord blood plasma, and breast milk were taken
in 2000-2001 from 15 mothers living in Stockholm (Guvenuis ct al., 2003). They were
opportunistic samples from women between 28 and 38 years old. For 55% of the
women, this was their first child; for 33%> of the women, this was their second child; and
for 14%, this was their third. Ten BDEs were measured, including 17, 28, 47, 66, 85, 99,
100, 153, 154, and 183. The median and range from the three matrices, in ng/g Iwt, were
as follows: maternal blood—2.1 and (0.7-8,4); cord blood—1.7 and (0.5-4.3); breast
milk—2.1 and (0.6 -7.7). BDE 47 was the predominant BDE in all matrices (46-70%) in
breast milk, 31-61% in maternal blood, and 45-94% in cord blood), followed by BDEs
153, 99, and 100. BDE 47 correlated in maternal blood and cord blood, but the levels of
153. 99. and 100 were higher in maternal blood as compared to cord blood (BDE 153 at
0.56 ng/g lwt in maternal blood but only 0.17 ng/g lwt in cord blood, e.g.). The authors
suggested this might indicate that the higher brominated congeners do not pass thru the
placenta to the same extent as do the lower brominated congeners.
Fukata et al. (2005) measured 27 BDE congeners including the key ones (47, 99,
100, 153, 154, 183, and 209) in umbilical cord tissue, maternal blood serum, and cord
blood serum in Japan. Samples from eight volunteers were obtained and split into pool A
and pool B (maternal blood was only available from pool B since there was insufficient
volume). Umbilical cord samples were uniformly lower for all congeners as compared to
cord serum and maternal serum, which were similar to each other. Total PBDEs in the
two pools were as follows: 5 and 1.7 ng/g lwt in umbilical cord A and B respectively, 35
5-11
-------
and 18 ng/g lvvt in cord serum A and B. and 20 ng/g lwt in maternal serum from pool B,
BDE 209 was not detected in umbilical cord, but it was found at 23 and 10 ng/g lwt in
cord serum and 10 ng/g lwt in maternal serum. PCBs and CDD/Fs had a slightly
different trend. While umbilical cord was always lower than cord seaim (like PBDEs),
they were close in magnitude and in fact maternal serum was significantly higher than
either organ.
The one study which presented blood data outside of the United States with
concentrations comparable to those found in the United States was a study in Nicaragua
(Athanasiadou et al., 2008: results there are expressed in pmol/g lwt; results expressed in
ng/g lwt are found in an earlier publication: Faldt et al., 2005). Five pools of serum from
teenagers lived and/or worked near a waste disposal in Managua, Nicaragua, and then
four pools of serum from women in different settings (urban areas, fishing villages, etc.)
were analyzed (one analysis per pool). BDEs 47, 99, 100. 153, 183, 203, and 209 were
analyzed. The pool of teenagers who both worked at the disposal site and lived nearby
(no other pool was that exposed) had the highest concentrations, with a total over 600
ng/g lwt. The average of the other 8 pools was 38 ng/g lwt. BDE 47 was the most
prominent congener, contributing just under 50% of the total concentration, with BDE 99
second at about 20%, BDE 100 at 11 %, and so on. BDE 209 was present at equal levels
in the teenagers living near and working at the disposal site, and all other groups, at about
5 ng/g lwt.
in summary, the review of literature on PBDEs in blood has revealed these trends:
1) Total PBDEs in the general population in the United States appear to be in the range of
30 - 100 ng/g lwt, although the studies have included outliers at the low end and at the
high end. The most valid study of national trends, an evaluation of NHANES 2003/2004
(2040 serum samples, all greater than 12 years of age), had geometric mean
concentrations of key congeners 28, 47, 99, 100, 153, and 154 at 1.2,. 20.5, 5.0, 3.9, 5.7,
and 2.3 ng/g lwt. One study of pregnant women in New York associated with 9/11
evaluations showed levels below 10 ng/g lwt, while outliers of studies with reasonably
large sample size show concentrations in the 100s and even the 1000s of ng/g lwt;
5-12
-------
2) one case study from the United States (presumably Fischer et al. 2006?) on 4
individuals in one family showed much higher levels in the children as compared to the
adults, with exposure to house dust (because of elevated BDE 209 in the children) and
mother's milk suggested to explain the levels in the 100s in the children as compared to
around 100 ng/g l\vt for the parents. This was the only study found with children;
3) United States levels are higher than levels found in nearly every other study done
outside of the United States, with most non-United States data suggesting total PBDEs to
be less than 10 ng/g Iwt;
4) the predominant congener is BDE 47, explaining about 50% of the total concentration.
The second most found congeners are 99 and 153, both explaining in the range of 10-
20% of total concentrations.
5) Most of the studies have not measured BDE 209, but, when measured, it was found in
about half the samples at low levels near 1-2 ng/g lwt. The exception of the one case
study of the family of 4, which showed levels above 100 ng/g lwt in children; low levels
of BDE 209 have been attributed to the rapid half-life of 15 days in humans, and the
higher levels in children attributed to dust exposures in the house.
5.2.2. Breast Milk Data
While blood data suggested concentrations of total PBDEs in the range of 30-100
ng/g lwt, data on PBDEs in breast milk suggest possibly higher concentrations, with
medians or means in some studies in the United States above 100 ng/g lwt. Table 5-2
shows congener-specific milk concentrations of BDEs from studies in the United States.
Interestingly, in one blood study, analysis of results by individuals suggested that females
could have meaningfully higher concentrations than males. As described above, Scheeter
et al. (2005) found that, in 39 samples from 22 males and 17 females, the range and mean
in males were 4.6-192.8 ng/g lwt and 25.1 ng/g Iwt, and, for females, the range and mean
were 5.6-365.5 ng/g lwt and 74,1 ng/g lwt. However, Schecter et al (2005) does note that
these differences are not statistically significant.
-------
The Environmental Working Group
-------
concentration for the 23 samples of 0.9 ng/g lwt. Measurements were not correlated with
age or length of time in nursing.
Focant et al (2004) evaluated two different analytical methods on both blood and
milk samples. They analyzed three pooled samples of mother's milk: one pool from two
mothers in Denver, one pool from 10 samples collected in 2003 in California, and the
third pool from 10 individuals in North Carolina, also in 2003. BDEs 17, 28, 47, 66, 85,
99, 100, 153, and 154 were analyzed. The average total BDE from these three was 315
ng/g lwt, with BDE 47 the highest at 193 ng/g lwt (61% of total), followed by BDJ5 99 at
55 ng/g lwt (17%), and BDE 100 at 34 ng/g lwt (11 %). Results from the more traditional
GC-IDHRMS analysis were compared with the newer GCxGC-LD TOFMS, revealing
reasonable agreement. Because often the number of samples (derived from the different
pools) analyzed by GC-IDHRMS was higher than the newer method, the summary results
are from these analyses.
Two studies have been conducted in the Boston, Massachusets area. In one,
breast milk was collected from 46 women, with total BDE concentrations ranging from 4
to 263 ng/g lwt, with a median of 28 ng/g Kvt (Wu et al., 2007). Only one sample, the
highest at 263 ng/g lwt, wras higher than about 130 ng/g lwt. BDE 47 dominated most
samples, although BDE 153 predominated in three samples, and BDE 209 was above
detection limits in 11 samples. Questionnaire data suggested that concentrations in house
dust as well as the consumption of frozen dairy products provided the strongest
associations with log-transformed total PBDE in breast milk. The second study in the
Boston area was conducted by Johnson-Restrepo et al. (2007), who measured breast milk
in 38 volunteer donors between June and November 2004. They measured 17 congeners
including BDE 209, and found a median total concentration of 19.8 ng/g lipid, with a
range of 0.06 to 1910 ng/g lwt and a mean of 75 ng/g lwt. BDE 209 was not detected in
any sample, although the detection limit appeared high at 204 ng/g lwt. The most
abundant congener found was BDE 47, explaining about half the concentrations found.
A limited set of longitudinal data (I.e., data on changes over time) was available
for 3 women for BDEs 47 and 99 (Sjodin et al., 2005). These data originate from an
ongoing study a Pennsylvania State University College of Medicine, where a cohort of 30
participants who seek prenatal and pediatric care are being enlisted for a longitudinal
study of PBDEs, pesticides, and PCBs in milk. Contrary to expectations, levels of all of
5-15
-------
these contaminants suggested increases over time. BDE 47 increased from 30 to 40 ng/g
lwt in 2 of 3 women from postpartum day 40 to 120, and from 10 to about 15 ng/g lwt
from day 40 to day 60 in the other woman. BDE 99 increased from 50 to 100 ng/g lwt in
one participant from day 40 to day 120, increased slightly from approximately 5 to 6 ng/g
Iwt from day 40 to day 60 in another, and decreased from 8 to about 5 ng/g lwt from day
40 to day 90 in the third participant.
Ryan et al. (2006) reported on the sampling of human milk from two locations in
North America: in Hamilton, Ontario, Canada and in Austin, Texas, of the United States.
The Canadian milk samples included 34 that pertained to 2005, 13 that pertained to 2003,
14 that pertained to 2002, and 26 that pertained to 1992. Specific congeners were not
identified; only total PBDEs was reported. The lowest concentrations found were in
1992, with a median of 3.1 ng/g lwt total PBDEs. Concentrations rose to a median of 33
ng/g lwt in 2003, with another rise to a median of 39 ng/g lwt in 2003, and then a decline
to a median 20 ng/g lwt for 2005 samples. The Texas samples pertained to years 2004
and 2002, and the median concentrations for those two years was higher than these
Canadian samples at medians for 2004 and 2002 of 43 and 44 ng/g lwt
Like the blood data, the mother's milk data suggests much higher concentrations
in America compared to European and Asian countries. Extracted milk fat from the 3ui
round of the WHO-coordinated exposure study was evaluated for the presence of BDEs
(Kotz et al., 2005). As of the writing of the study, samples from 17 different locations (of
24 total) were used. Nine BDEs were quantified, including congeners 15, 28, 47, 77, 99,
100, 126, 153, and 183. The highest level by far was the level found in a sample from the
United States at 373.6 ng/g lwt; the second highest was 10,3 ng/g lwt from a sample from
Ireland. Hie predominant congener was BDE 47, at 63% in United States samples (233
ng/g Iwt), followed by BDE 99 at 16% (60 ng/g lwt) and BDE 100 at 11% (41 ng/g Iwt).
Schuhmachet* et al. (2007) report on concentrations of 15 BDE congeners
(although the specific congeners were not identified except to note that samples were
fortified with BDEs 28, 77, 99, 153, and 183; no information on whether BDE 209 was
included). Their study included 15 women, sampled in 2002, that lived in an urban area
(7-10 km from a hazardous waste incinerator; 7 women) and near an industrial zone (8
women). There did not appear to be a distinction in the two small groups, with means of
2.2 and 2.5 ng/g lwt in the urban and industrial zones, respectively.
5-16
-------
Thomsen et al. (2005b) sampled breast milk of 151 women representing the
northern, southwestern, and eastern parts of Norway. Samples were analyzed for BDEs
28, 37, 47, 85, 99, 100, 119, 138, 153, 154, and 183, The sum of the seven most
abundant BDE congeners (28, 47, 99, 100, 153, 154, and 183) ranged from 0.95 to 21.05
ng/g Iwt, with a median of 2.35 ng/g Iwt, which is comparable to other European
countries. BDE 47 was the most abundant; it had a median of about 1.15 ng/g Iwt. BDE
153 was the second most abundant: a median of about 0.50 ng/g Iwt. By use of a
multiple linear regression model, it was shown that there was a statistically significant
positive correlation with age (older women had higher concentrations) and a negative
correlation with parity (number of children) and education. The mean concentrations of
total BDEs for women <28, 28-31, and >32 was 3.24, 3.39, and 5.17 ng/g Iwt,
respectively. The mean concentration for women with one child was 4.33 ng/g Iwt.
Women with more than one child had a lower concentration: 3.64 ng/g Iwt. The authors
assert that this was the first study to show statistically significant correlations with age
and number of children.
Gomara et al (2007) report on the sampling of PBDEs in human umbilical cord
serum, maternal and paternal serum, placentas, and breast milk from individuals living in
two locations (Vallecas and Getafe) in Madrid, Spain. The sampling occurred between
October 2003 and May 2004, and involved 391 individual samples including 113 of
maternal serum, 104 of paternal serum, 92 of umbilical cord serum, 30 of placenta, and
52 of breast milk. Fifteen individual congeners were measured in all samples, including
BDE 209. Breast milk samples had median concentrations in the two locations of 6.1 and
5.5 ng total PBDE/g Iwt, which was a bit lower than the blood samples, which had
medians ranging from 9.7 to 17 ng total PBDE/g Iwt in the various blood matrices. BDE
209 dominated the breast milk samples, with medians of 2.8 and 2.9 ng/g Iwt (ranging as
high as 52 ng/g Iwt) in the two locations.
A total of 89 lactating mothers in four towns in Japan provided both serum and
milk samples for analysis of 13 BDEs, including BDEs 15, 28, 47, 99, 100, 153, 154,
183, 196, 197, 206, 207, and 209 (Inouc et al., 2006). The geometric means for the total
amounts of the 13 PDEs in human milk and serum was 1.56 and 2.89 ng/g Kvt,
respectively. BDE 209 was the predominant congener in serum, accounting for 38% of
the total amount of BDEs, but it was a minor component in milk, accounting for 8%. In
5-17
-------
milk, BDEs 47 and 153 were the major contributors, accounting for 28 and 23% of total
BDEs, respectively. Nursing duration was found to be correlated with PCB
concentrations in the infants, but not so for PBDEs. In contrast, geography was tied to
PBDEs, but not to PCBs.
Fangstrom et al. (2008) conducted a temporal study on Swedish mother's milk.
Fourteen pooled milk samples representing 1980 (116 mothers pooled), 1984/5 (102
mothers), several of the years between 88 and 2002 (20 mothers), 2003 (15 mothers), and
2004 (20 mothers) were sampled for BDEs 47, 77, 99, 100, 153, and 209. It was not
possible to quantify BDE 209 in milk, and the authors suggest this could be due to the
short half-life of this compound that has been noted for serum. Front the middle of the
1990s, the concentrations of the lower brominated BDE congeners (47, 99, and 100) are
decreasing, while 153 appears to be retaining its levels reached towards the latter 1990s.
BDE 47 was near or less than 0.5 ng/g Iwt prior to 1990, reached levels above 2.0 ng/g
lwt by mid-1990s, and dropped sequentially from 1.8 (in 2001) to 1.4 (2002) to 1.2
(2003) to 0.9 (2004). Very similar trends were seen for 99 and 100, with peaks between
0.5 and 0.8 in the mid-1990s, dropping to 0.3 ng/g lwt in the early 2000s. Meanwhile,
the highest concentration of BDE 153 till the 1999 sampling was the BDE 99 sample at
0.82. Afterwards, BDE 153 ranged from 0.7 to 1.3 between 2001 and 2004. This trend
of dropping lower chlorinated BDEs and increasing higher BDEs was seen in the United
States, Norway, the Netherlands and the Faroe Islands. This might be due to declines in
use of the pentaBDE formulation and/or the debromination of BDE 209 and other higher
brominated BDEs.
In a temporal study of organohalogen compounds, including PCBs, PCDD/Fs,
and PBDEs. in the breast milk of German women, Furst (2006) found an increase in the
mean concentration of PBDEs from a pooled sample (n=300) in 1992 to a sampling of 79
women in 2002. Specifically, the mean total concentration (including BDEs 28, 47, 66,
85, 99, 100, 153, 154, and 183) increased from 1.87 ng/g Kvt to 3.75 ng/g lwt. This
contrasted with the other organohalogen compounds studied, which showed reductions
during this time period.
Ohta et al. (2002) determined the concentration of PBDEs in breast milk of 12
primaparae women at one month after delivery in Japan. The PBDEs evaluated included
28, 47, 99, 153, and 154. Concentrations ranged between 0.7 and 2.8 ng/g lwt in the
5-18
-------
breast milk. Samples were also taken of numerous food products, including 20 fish
samples, spinach, potato, carrots, pork, beef, and chicken (see Chapter 4 for a summary
of the results). Questionnaires on food consumption were given to the women, and a
strong correlation was found between consumption of fish and milk concentration. In a
"high" group of fish consumers (n=5), the average breast milk concentration was 1.7 ng/g
Iwt, while the concentration in the "low" group of fish consumers (n-3) was 0.8 ng/g Iwt.
Toms et al. (2007) measured BDEs in pooled milk samples from Australia. A
total of 157 milk samples were collected between 2002 and 2003, and they were pooled
to create 17 regional samples. Eighteen congeners were measured; BDE 209 was not
measured. Total PBDE averaged 11.1 ng/g Iwt (median = 11.0 ng/g Iwt) with a narrow
range of 6.1-18.7 ng/g Iwt. BDE 47 dominated the profile, explaining over 50% of the
total, followed by 99, 100, and 153—all of which contributed between 10 and 20%.
In summary, data suggests concentrations of total BDEs in women's breast milk
in the United States exceeds that in blood, perhaps averaging near 100 ng/g Iwt, in
contrast to a range more like 30-100 ng Iwt in blood generally. Part of this trend could
be a gender issue, as one blood study (Schecter et al. 2005) suggested higher
concentrations in women than in men, although this difference, was not statistically
significant. Similar to blood data, concentrations in breast milk of United States women
exceeded that of women outside of the United States. Breast milk concentrations of total
PBDEs outside of the United States appears to mostly below 10 ng/g Iwt. As for blood,
BDE 209 was usually not measured or when measured, not found in most instances (only
one study found significant levels of BDE 209 in blood, and that was in a case study and
it was found in one child and one toddler). BDE 47 was the predominant congener,
followed by either BDEs 99 or 153.
5.2.3. Adipose and Other Tissue
Most of the body burden data originates from either blood or human milk. The
very limited data on adipose and other tissue is generally consistent with blood and milk
data - United States data are higher than data from other countries; BDE 47
predominates, and BDE 209 is either not measured for or not found when measured.
Johnson-Restrepo et al. (2005) measured BDEs in adipose tissue from 52
individuals who were undergoing liposuction during Oct, 2003-Oct, 2004 in New York
5-19
-------
City. BDEs sampled included 28, 30, 47, 85, 99, 100, 153, 154, and some unidentified
di, tri, and penta-BDEs. The occurrence of BDE 209 was examined qualitatively, but it
was not detected. Total BDEs ranged from 17.4 to 9,630 ng/g Iwt, with a median of 77.3
ng/g lwt and a mean of 399 ng/g lwt. The mean dropped to 141 ng/g lwt when two
outliers of 9.630 and 4,060 ng/g lwt were dropped. The authors claimed this was 10 to
100 times higher than concentrations reported for human adipose tissue collected from
several European countries. The data showed no correlation with age, which is consistent
with the theory that most: exposure was only in recent years. BDE 47 was the
predominant congener, explaining roughly 33% of the total BDEs (at a mean of 132 ng/g
lwt), followed by BDE 153 (23%; 91.8 ng/g lwt), BDE 99 (18%; 74.4 ng/g lwt), and
BDE 100 (17%; 67.7 ng/g lwt).
Data on 11 adipose tissue samples from 32 women in the industrial port town of
Porto Alegre, Brazil, showed a range of 0.73 to 3.69 ng/g lwt, with BDE 47 dominating
the profile (median = 0.52 ng/g Iwt), followed by BDE 99 (median = 0.34 ng/g lwt), BDE
153 (median = 0.19 ng/g lwt) and BDE 100 (median = 0.12 ng/g lwt). BDE 153 was
found in only 27% of samples with a median of 0.07 ng/g lwt (Kalantzi et al., 2005).
Naert et al. (2006) measured BDE congeners 28, 47, 99, 100, 153, 154, and 183 in
abdominal adipose tissue from 53 individuals (31 men, 22 women) who died from natural
causes or accidents in Belgium. The mean age of the individuals was 53 when they died,
and no information on diet or exposures was available. Consistent with blood data from
Europe, total BDE concentrations ranged between 1.23 and 57.2 ng/g lwt. with a median
of 5.32 ng/g lwt. The most predominant congener was BDE 153, at a median of 2.40
ng/g lwt, with all other congeners having medians under 1.00 ng/g lwt.
Schecter et al. (2007) measured levels in fetal liver tissue samples that were
obtained from 4 stillborn fetuses and 7 liveborn infants, all of whom died shortly after
birth and before any feeding. Thirteen congeners were evaluated, although BDE 209 was
not found in any samples. The mean total concentration was 19.5 ng/g lwt (assuming ND
= 0, at ND = Vz DL. the total went up to 23.1 ng/g lwt. owing mainly to an average of 3.5
ng/g lwt assuming ND ~ !4 DL for BDE 209). The congener profile was fairly similar to
blood and breast milk profiles, being dominated by BDE 47 at 55% of the total, BDE 99
at 23%, and BDE 100 and 153 at 10 and 6%, respectively. One sample was uniquely
high, 96 ng/g lwt, while all other samples were under 33 ng/g Kvt. This one high sample
5-20
-------
was from the infant who lived the longest, 7 days. The authors did not clarity why this
infant was characterized as having died after 7 days before any feeding could occur.
Gomara et al (2007) report on the sampling of PBDEs in human umbilical cord
serum, maternal and paternal serum, placentas, and breast milk from individuals living in
two locations (Vallecas and Getafe) in Madrid, Spain. The sampling occurred between
October 2003 and May 2004, and involved 391 individual samples including 113 of
maternal serum, 104 of paternal serum, 92 of umbilical cord serum, 30 of placenta, and
52 of breast milk. Fifteen individual congeners were measured in all samples, including
BDE 209, The concentrations were lowest in the placentas, with a median value of 1.9
ng total PBDEs/g lwt. This compared to medians in the breast milk around 6 ng total
PBDE/g lwt, and medians in the blood matrices ranging narrowly between 9.7 and 17.0
ng total PBDE/g lwt.
5.2.4. Selection of Representative Body Burden Profiles
In Chapter 4 (Section 4.8), exposure media concentrations were assigned for a
select group of PBDE congeners in order to conduct the exposure assessment.
Specifically, these assigned media concentrations were used to estimate exposure intake
dose (see Section 5.4 below.). Also, simple pharmacokinetic exercises were applied to
those intakes and used to predict body burden levels (see Section 5.5 below). In order to
evaluate the merit of that exercise, those body burden predictions were compared with
measured body burdens. This section will assign representative central tendency body
burdens for that purpose.
As also discussed in Chapter 4 in the context of exposure media concentrations,
assignment of these representative body burdens should have these characteristics: 1)
they should originate from the United States, or maybe Canada, as representative of
industrial North American patterns, 2) they should be representative of background and
not occupational exposures, and 3) preferably, they should originate from a single study
that has the ideal characteristics of having a large sample size, being from a diverse
geographic area, and containing the appropriate set of congeners. Table 5.3 has those
assignments, one set each for blood and mother's milk.
5-21
-------
The blood data are from Sjodin et al. (2008). Of all the United States blood data
available, this evaluation of NHANES 2003/4 is the most representative of recent,
national trends. Geometric mean concentrations, shown in Table 5.3, are provided for
BDEs 28, 47, 99, 100, and 153, but not for BDEs 17, 66, 85, 154, and 183, because these
were only quantified in 40% of the samples. Arithmetic means were not provided in
Sjodin et al. (2008) and medians might have been selected over geometric means as most
representative of central tendency, but the median of BDE 99 was provided as less than
detection limit, while the geometric mean was provided for BDE 99. There are five
studies on PBDEs on United States breast milk that could be used to represent
background conditions. These include two by public interest research groups—the
Environmental Working Group (EWG; Lunder and Sharp, 2004) and the Northwest
Environment Watch (NEW, 2004), two conducted in the Boston, MA area (Wu et al.,
2007; Jfohnson-Rcstrcpo et al., 2007), and a study of 47 samples taken in Texas by
Schecter et al. (2003). None of the studies have a large sample size. The EWG study
includes 20 samples from around the United States, so it has broad coverage, but it also
has the highest concentrations of any United States study on breast milk. The NEW
study represents several states in the Northwest, has a sample size of 40, and its
concentrations track well with EWG, but are consistently lower. The Schecter et al.
(2003) data on 47 samples are all from Texas. Their data also tracks with the two studies,
but it is the lowest of all the data and originates from one state only. The two studies
from Boston, MA, found the lowest concentrations, with median levels at 28 ng/g lwt
total (Wu et al.. 2007) and 19.8 ng/g lwt (Johnson-Restrepo et al, 2007). The NEW
median data from the Northwest will be used to characterize breast milk; it is displayed in
Table 5.3. It was chosen because it represented several states, it had as large a sample
size as others, and the concentrations were consistent among the US studies other than
the EWG study which had the highest concentrations. Mean concentrations were also
provided in the NEW study, but they were meaningfully higher than median
concentrations: the median total PBDE concentration was 50 ng/g lwt (this median
contained more congeners than the median of 44 ng/g lwt shown in Table 5.3) while the
mean total concentration was 97 ng/g lwt. Because geometric means were chosen to
represent blood concentrations, the most analogous statistical representation of breast
milk concentrations, median concentrations, were chosen.
5-22
-------
5.3. STUDIES ON INTAKE, OR EXPOSURE], DOSE
Several researchers who measured PBDEs in exposure media then went on to
estimate the exposure dose associated with that media. Their approach was simply to
associate the media concentration with a contact rate. For example, a given concentration
in dust times an amount of dust ingested per day provides an estimate of the daily dose
via dust ingestion. This was done most often for studies on house dust and food, but
researchers also considered inhalation exposures when measuring air. A second approach
used to estimate intake dose from house dust was to measure a surface loading (not a bulk
concentration) on the hands in units of mass/unit area, and then use an empirical model to
correlate that loading with an intake rate, in units of mass/time. Other researchers took a
more studied approach to evaluating exposure to PBDEs, using models, statistical
approaches, and other avenues to provide estimates and insights on exposure to PBDEs.
Table 5-4 provides estimates of these exposure doses.
As discussed in Chapter 4, studies have shown indoor dust to contain much higher
concentrations of PBDEs compared to outdoor soils. Subsequently, the "soil ingestion"
pathway has focused on indoor dust measurements. Stapleton et al. (2005, 2008b) have
conducted the most comprehensive dust pathway analyses with their studies on house
dust coupled with exposure modeling. Using estimates of inadvertent ingestion of dust
by young children (ages 1 -4), 0.02-0.2 g/day, they used their measurements of PBDEs in
dust to estimate a total ingestion of PBDEs range from 120 to 6,000 ng. day (Stapleton et
al., 2005). They list an adult exposure of 3.3 ng/'day, but this is based on a low estimate
of 0.56 mg/day of dust ingestion. This value was found in EPA's Exposure Factors
Handbook (EPA, 1997), where Hawley (1985) is cited for using a value of 0.56 mg/day
to characterize adult exposure to housedust from normal activities in the house (higher
exposures of over 100 mg/day resulted from "work in the attic"). Stapleton et al. (2008b)
later approached the dust ingestion pathway from a different angle. They measured
PBDEs in hands using sterile gauze pads soaked in isopropyl alcohol. They measured
PBDEs on the hands of 33 individuals residing in Durham, North Carolina, and found a
median total load per hand of 130 ng, or when normalized to surface area, a surface
loading of total PBDEs of 135 pg cm' . They measured 13 congeners including BDE 209,
which was found in 22 of the 33 samples, with a median total load of 25.5 ng, and a high
-------
of 270 ng. Using an empirical approach based on contact events per day, transfer
efficiency, the hand loadings, and the fraction of hand coming in contact with the mouth,
they estimated median exposures to the adult and child to be 154 ng/day and 1380
ng/day.
In contrast to the low dust ingestion rate of 0.56 nig day assumed by Stapleton et
al, (2005) in their earlier estimates of PBDE ingestion via dust, Sjodin et al. (2004b)
assumed an upper limit ingestion rate of 100 mg/day, and using their median
concentration of 4,200 ng/g in house dust from Atlanta (range of 530-29,000 ng/g), they
suggest that this pathway could add up to 400 ng/day of PBDE exposure. Such a total
would dwarf estimates of 40 to 100 ng/day from food ingestion they cite from the
literature. Harrad et al (2008) used "mean" and "high" dust ingestion rates of 20 and 50
mg/day to model adult ingestion of PBDEs based on measurements from homes in Texas.
Using arithmetic mean concentrations from dust samples (n-28), they estimate a "high"
total intake of 228 ng/day, which includes BDEs 28, 47, 99, 100, 153, 154, 183. and 209.
They also provide estimates for toddlers (6-24 months) assuming 50 mg/day as a "mean"
and 200 mg/day as a "high"ingestion rate. Again using average dust concentrations, a
high toddler intake estimate was 91 ng/day. They also provide estimates for toddlers and
adults in the UK, New Zealand, and Canada, based on dust sampling from those
countries. The most interesting finding was that UK dust was substantially higher in
BDE 209 compared to all other countries, and median intakes of BDE 209 were double
those from the US. Using data that originated from Kuwait, Gevao et al. (2005) used
standard exposure assumptions for dust ingestion for children (100 mg/day) and adults
(10 mg/day) and found that the mean ingestion of total PBDEs averaged 2.0 ng/day for
children, and 0.2 for adults. Harrad et al. (2006) used indoor dust measurements from 8
homes in the United Kingdom to estimate a possible range of adult dust ingestion
exposures from 0.9 to 22 ng/day total BDEs, and a range for toddlers of between 12 and
43 ng/day.
Similar to the relationship between outdoor soils and indoor dust, outdoor air
concentrations were typically lower than indoor air measurements, and, hence, literature
estimates of inhalation exposures to PBDEs have focused on indoor air concentrations.
Using data from passive indoor air in a Canadian study, Wilford et al. (2004) assumed
standard resting respiration rates and found that the median exposure via inhalation was
5-24
-------
1.9 ng/'day for females and 2.0 ng/day for males, which compares to a Canadian estimate
of 44 ng/day by dietary intake. Hazrati and Iiarrad (2005) used passive air measurements
in 12 homes, 10 offices, and 1 private car, and an inhalation rate of 20 m'Vday, to estimate
a mean daily intake via inhalation of 4.3 ng/day in the United Kingdom. Iiarrad et al.
(2006) used other air data to conclude that average inhalation intakes were 2 ng/day total
PBDEs and less in the United Kingdom. Using air concentration data from a study in
Kuwait, Gevao et al. (2005) estimated inhalation doses of 0.4 ng/day for adults and 0.2
ng/day for children (with inhalation rates of 20 m7day for adults and 8.3 nrVday for
children. However, Meng et al. (2007) estimated inhalation exposures to PBDEs using
data on outdoor urban air concentrations from China (see Chen et al., 2006). They found
median exposures doses to range from 2.7-9.2 ng/day. They note this is higher than the
estimates provided in Wilford et al. (2004) and Harrad et al. (2006), but they claim this is
because they included BDE 209 in their estimates unlike these other two studies, and this
congener dominated the air profiles. Mandalakis et al. (2008) studied air concentrations
in cars in Greece, and combining their measurements with inhalation rates and time in
cars per day, they estimated inahaltion of total PBDEs while driving to range between
0.0005 to 2.9 ng/day, with a median of 0.2 ng/day. This exposure was dominated by
BDE 209, explaining about half of all exposure. They also found that, despite the small
amount of time in the car, that this acti vity contributed 29% of overall daily inhalation
exposure.
Most direct dose estimates in the literature pertain to dietary dose, and they are
developed mostly by individuals who also measured food concentrations. Estimates of
dose originate from the United States and overseas, with United States studies
summarized here first. Schecter et al, (2006b) combined their measured average food
concentrations with food consumption rates to calculate intakes for various age ranges (2-
5, 6-11, 12-19—>— 60) and for males and females. Total intakes ranged from about 0.9-
1.5 ng/kg body weight/day for males/females above the age of 12. For ages 2-5 (males &
females), the intake was estimated at 2.7 ng/kg-d, and for ages 6-11 (males & females),
the intake was 1.8 ng/kg-d. Individual intakes between 47 and 99 were nearly identical
for males and females, ranging from 0.4 to 0.7 ng/kg-d after age 12, 0.9-1.5 ng/kg-d for
the two earlier age ranges. Huwe et al. (2005) estimated a dietary intake of PBDEs from
meats for a consumer of "lean meats" (5% lipids) was 0.3 ng/kg-d, while a "higher fat
5-25
-------
meals" consumer had an intake of 0.8 ng/kg-d. The estimates are based on an average
body weight, of 53 kg. Fish, meat, and fowl products were purchased in December 2003
and February 2004 from 3 different food markets in Sacremento and El Dorado Hills in.
Northern California (Luksemburg et al., 2004). Using average daily intake by adults and
children taken from the Exposure Factors Handbook (EPA, 1997), dose estimates were
provided with their data. Using the highest and lowest concentrations measured in wild
and farm-raised fish, the theoretical average daily intakes of PBDEs through fish
ingestion ranged between 0.1 and 1.0 ng/kg-day in children and between 0.02 and 1.0
ng/kg-day in adults. Assuming the highest and lowest concentrations measured in beef
and chicken products, theoretical average daily intakes of PBDEs through ingestion
ranged between 0.4 and 20 ng/kg-d in children and between 0.4 and 10 ng/kg-d in adults.
Ten studies were found providing dietary dose estimates in Europe and Asia.
Bocio et al. (2003) estimated that, dietary intake equaled 97.3 ng/day for total PBDEs for
adults in Spain, based on a total diet survey. This was based on homologue group
concentrations; a later survey of Spanish foods evaluating 15 individual congeners
arrived at an estimate of 38.5 ng/day for BDEs (Gomara et al., 2006). Schuhmacher et al.
(2007) used the food concentration data of Bocio et al. (2003), in combination with food
consumption rates for an urban and industrial area of Spain to calculate dietary intakes of
total BDEs of 72 (urban) and 63 (industrial) ng/day. Harrad et al, (2004) measured
PBDEs in duplicate diet samples from both a vegan and omnivorous diet. They
estimated a dietary exposure average 107 ng/day (median - 91 ng/day) using
consumption data from the survey, and although they did not estimate an exposure intake
for the vegans, they provided the omnivorous and vegan concentrations, and it is noted
that the vegan concentrations were about one-half the omnivorous concentrations. They
note this discrepancy is not as large as the discrepancy for other POPs, like dioxins,
which bioconcentrate substantially more in animal fat. They could not explain this trend
but noted other literature showing significant concentrations in vegetative food products.
Knutsen et al. (2005) combined concentrations from a market basket survey with a
comprehensive food consumption survey to estimate a mean daily exposure of 62.5
ng/day, a median exposure of 48.6 ng/day, and a 95% exposure of 149.0 ng/day for
Norway. Without fish, these numbers were about one-third as much, with mean at 20.0
and median at 19.2 ng/day. When a different survey more specific to fish types was used,
5-26
-------
the median rose to 74.2 ng/day, and then even higher when a recommended daily
additional intake of cod liver oil was assumed (which is recommended by the Norwegian
government as a healthy supplement)—median intake rose to 122.9 ng/day. A similar
combination of a comprehensive market basket consumption survey with composite food
samples measured for PBDEs was conducted by Bakker et al. (2008) for the Netherlands.
They found a median dietary intake of 0.79 ng/kg-day, with a 95% of 1.62 ng/kg-day,
dominated by dairy and fish at 39 and 28%, respectively. An estimate of 51 ng total
BDEs/day was derived for diet only for Swedish general population (Darnerud et al.,
2001). Concentrations in a market basket survey were combined with dietary intakes
from fish and fish products, meat, dairy, and fats/oils. Fish products contributed about
half of total. Meat, dairy, and fats/oils contributed about 15% each. Using mother's milk
concentration of 4.2 ng/g Iwt, they estimated an infant dose of 110 ng/day.
It is noted that BDE 209 was not included in any of these dietary intakes and
surveys; the standard suite of BDE congeners (47, 99, 100, 153, and 154) were included.
Three studies were found that included BDE 209 as well: one on fish consumption in
China, one in Belgium, and one was in the United Kingdom. Fish consumption intakes
were determined for a surveyed group of individuals in China, and the median dose
ranged from 1.7-12.9 ng total/day for several age ranges (Meng et al., 2007). While
including BDE 209, it was found in only 14 of 390 fish samples upon which the intake
estimates were based. This could have been due to the high detection limit, 0.1 ng/g wwt,
of this congener compared to detections limits of 0.001-0.003 ng/g wwt of 10 other
congeners measured. The one in Belgium did not include individual congener breakouts,
but it was stated that BDE 209 was never found above LOQ in any food sample. The
average total adult intake was only 35 ng/day, generally low but in line with other
European surveys which did not include BDE 209 (Voorspoels et al., 2007).
However, the presence or absence of BDE 209 could very well be a major issue
for these literature estimates of European exposures, as a dietary estimate based on a
survey which did include BDE 209 arrived at substantially different results. The United
Kingdom Food Surveillance Agency ( FSA, 2006) recently published results of a food
survey including BDE 209 and quite alarmingly found BDE 209 at the highest level of all
BDEs (as noted in Chapter 4 in the section on food concentrations). They estimated
exposure doses in conjunction with their food concentrations and found average intakes
5-27
-------
totaling 5,9 ng/kg-day, of which 4.5 ng/kg-day was due to BDE 209. These results either
are questionable themselves, or alternately, throw into question other European surveys
on "total" dietary dose of BDEs which have not measured BDE 209 in the food. It is
noted that, in a study on BDEs in dust in Europe, BDE 209 dominated substantially over
other congeners (Fabrellas et al, 2005), providing support to this finding in food, and
suggesting that much of the European literature on exposure to BDEs does not tell the
complete story by not considering BDE 209.
Jones-Otazo et al. (2005) used models, in combination with reported food
concentrations, to provide a comprehensive evaluation of exposure to PBDEs pertinent to
the Toronto urban environment. A regional, multi-media fate model termed Multi-media
Urban Model (MUM-Fate) was used to predict outdoor soil, outdoor air, indoor, and
residential dust concentrations of PBDEs. The results were combined with measured
concentrations from food and mother's milk to determine potential exposures of PBDEs
in this complex exercise. Exposure scenarios included the following: elevated indoor
sources, fish eater, occupational exposure (from an electronics recycling plant), and
exposures experienced by four younger age-ranged individuals (0-6 months), toddler (6
months-4 years), child (5-11 years), and teen (12-19 years). They modeled "total
PBDE," which included 47, 99, 100, 153, and 154. BDE 209 was excluded due to a lack
of data in most media. Their results suggested first that 100-422 g/day of PBDEs were
released into the 470 km2 modeled area; these emission rates resulted in modeled air
concentrations which matched measured concentrations. Their results suggest a range of
average daily intake from all sources to be 155 ng/day for the adult to 1,965 ng/day for
the infant (2 to 280 ng/kg bw). Nearly 100 ng/day of this adult exposure is modeled to
come from soil/dust exposures, mainly indoor dust ingestion (dermal and inhalation
minimal). For toddlers, over 90% of their daily intake of 264 ng/day comes from dust
ingestion. Following household dust, "dairy, meat, and eggs" exposure contributed 16%
of total exposures. Other exposure estimates include 227 ng/day for the fish eater and
2,190 ng/day for the occupational exposure. While the finding that house dust contributed
the most to human exposure for all scenarios, this has to be considered carefully in light
of the data input into the exercise, particularly the food concentrations. It is noted that
their assumption of total concentration for dairy, meat, and eggs was 101 pg/g wwt, while
Schecter et al. (2006b) found that total BDE ranged from 39 ppt wwt to 1426 ppt wwt in
-------
18 meat samples collected in Texas. Therefore, it seems quite possible that the food
concentrations were lower than would be assumed had the estimates been developed for
United States conditions.
McDonald (2005) developed a dose estimate starting from body burdens and
working backward using pharmacokinetic (PK) modeling. Examining 6 studies that
evaluated PBDE body burdens in individual women (serum, milk, and adipose tissue
studies), McDonald found that the median concentration was about 48 ng/g lwt, a mean
of 90 ng/g lwt, and a 95% of 302 ng/g lwt. These totals were the sum of congeners 47,
99, 100, 153, and 154. McDonald used a simple first-order, single compartment (lipid
compartment) PK model to determine the dose required to obtain congener specific body
burdens. He assumed congener-specific adsorptions ranging from 0.78-0.94. Congener-
specific half-lives were based on rat data and a correlation between rat and human half-
lives, and he came up with half-lives of 3 yrs for BDE 47, 5.4 yr for BDE 99, 2,9 yrs for
BDE 100, 11.7 yrs for BDE 153, and 5.8 yrs for BDE 154. The estimated total dose of
PBDEs (sum of 5 congeners) was 8.5, 16.0, and 53.6 ng/kg-d for the median, the mean,
and the 95% concentrations, respectively. He went further to evaluate this dose level in
terms of potential for health impact, based on rat testing. He found that rodent-to-human
body burden concentrations of concern were < 1 for alterations of male and female
reproductive organs in rats, < 10 for neurodevelopmental effects in mice, and <100 for
neurodevelopmental effects in rats. They also looked at other intake studies and noted
that an intake estimate of 0.4-11 ng/kg-day was derived in a forward manner in another
United States study on food intakes only.
Webster et al. (2005) conducted a monte carlo exercise on exposures to BDE 47
using concentrations of this congener in air, food, dust, and using traditional exposure
factors from the EPA's Exposure Factor's Handbook (EPA, 1997). They derived
estimates of exposure dose (in ng/kg-day) to infants, young children, and adults. Then,
using a simple 1-compartment, steady-state approach and assuming a 4-year half-life,
they estimated body burdens resulting from median doses and compared that to some
tissue data. From their monte carlo simulations, the mean dose to infants, children, and
adults were 123.9, 7,7, and 0,9 ng/kg-day, respectively. The infant dose was dominated
by breast milk ingestion, explaining 95% of dose. The children's dose was dominated by
the dust-related exposures of dermal (35%) and dust ingestion (36%), with diet being the
5-29
-------
third most important (28%). For the adult, diet dominated at (63%), while dust ingestion
(23%) and dermal (11%) comprised a significant portion of the rernainng exposures.
Several uncertainties were identified: absorption fractions, concentrations in exposure
media, and so on. Based on the median adult dose (0.8 ng/kg-day), the steady-state lipid
concentration was predicted to be 5 ng/g Iwt. Assuming breast milk lipid concentrations
were similar to overall body lipid concentrations, this was lower than the average BDE
concentration found in other United States studies, which the author claimed to average
about 20 ng/g Iwt. The authors suggested possible issues with exposure dose and PK
modeling assumptions (4-year half-life too short).
She et al. (2005) used the ratio of BDE 99 to BDE 47 (99/47) as a way of
understanding the possible sources of exposure to humans. The study evaluated data
from numerous studies which provided data to develop ratios in the penta formulation, in
house dust, and in mother's milk, and made observations in each. For the penta
formulation. BDEs 99 and 47 are present at about the same amount, 40% each, with BDE
100 at 7%. The ratio of 99/47 in United States house dust ranged from 0.5 to 2.0, with an
average ratio of about 1, which would be consistent with the penta BDE formulation with
use information. This suggests the widespread use of penta in North America. The
variation in the ratio could be due to denomination and different rates of volatilization
between the congeners, but overall. United States dust appears consistent with
expectations. United Kingdom dust showed high levels of BDE 209, which is consistent
with the deca formulation. Mother's milk, on the other hand, showed a 99/47 ratio of 0.2
to 0.3 in three studies, with the highest levels in milk to have a ratio more like 0.6. Their
explanations and discussions focused on these points: 1) BDE 47 is more
bioaccumulative than BDE 99, so that alone could lead to a downward shift in the ratio
(i.e., more prominence of the more bioaccumulative BDE 47 in mother's milk), 2) this
bioaccumulation trend extends not only to direct dust exposures by the mother, but
maybe more importantly to exposures by animals and then exposures by humans by
animal food products, 3) this latter point suggests that on average, a 0.2-0.3 ratio would
suggest a predominant pathway by food, and 4) the higher mother's milk concentrations,
when the ratio was more like 0.6. might suggest that dust is an important pathway as well
(because dust is a direct exposure with one bioaccumulation step, while food is a
secondary pathway).
5-30
-------
The highest estimates of exposure dose, on a body weight, are for infants via
breast milk. Scheeter ct al, (2005) calculated infant intakes to be 307 ng/kg-d, dominated
by BDEs 47 at 169 ng/kg-d, followed by BDE 99 at 49 ng/kg-d, and BDE 100 at 36
ng/kg-d. As noted above, Webster et al. (2005) calculated an intake of 123.9 ng/kg-day
for BDE 47 alone, and Jones-Otazo et al. (2005) modeled an intake of 280 ng/kg-day for
total BDEs for the infant ages 0-6 months. The lowest intake estimates were for breast-
feeding infants in China, where Meng et al. (2007) calculated a median intake of total
BDEs to be about 6-7 ng/kg-day (ages 0 - 1 year).
In summary, some of the key observations from exposure and exposure dose
estimation include:
1) on a body weight basis, adult exposures for the dietary pathway alone in the United
States appear to be estimated in the range of 0.5-2.0 ng/kg-day;
2) although no United States studies have estimated inhalation exposures, estimates for
the UK, Canada, and one study in Kuwait suggest much lower exposures in the range of
0,4 - > 4.0 ng/day (on a body weight basis assuming a 70 kg adult, 4 ng/day = 0.06
ng/kg-day);
.3) adult exposures to dust have been estimated to be higher than this range, perhaps more
like 3.0 - 6.0 ng/kg-day; these estimates are made based on vacuum dust sampling
combined with dust ingestion rates of 100 mg/day;
4) by these two findings, the suggestion is that dust would dominate adult exposures, and
in fact, the majority of studies looking at multiple pathways conclude that dust exposures
may dominate. However, there is some uncertainty in the literature on this point. One
study using monte carlo techniques found that adult exposures to BDE 47 were
dominated by diet, 63%, as compared to dust, 23%, and dermal contact with dust, 11%
(this was the only study found which looked at dermal impacts). Another study assumed
indoor dust ingestion was < 1 mg/day, so that this pathway resulted only in a BDE intake
of about 3 ng/day;
5-31
-------
5) there seems to be uniform agreement that child and toddler exposures are higher than
adult on a body weight basis, > 7 ng/kg-day, and are dominated by dust exposures;
6) infant exposures are dominated by breast milk ingestion, with body weight-based
exposures above 100 ng/kg-day, with one estimate over 300 ng/kg-day,
5.4. ESTIMATES OF BACKGROUND INTAKES OF PBDES FOR ADULTS
The procedures to estimate individual intakes of PBDEs for adults were
developed in a manner similar to that done in the USEPA Dioxin Reassessment (EPA,
2003), with one important addition, as will be described shortly. Intakes are a function of
contact rates in combination with exposure media concentrations. As in the Dioxin
Reassessment, intakes are defined as the amounts of contaminants crossing the body
boundary but not yet absorbed into the blood stream. For this reason, absorption
fractions are not applied to inhalation and ingestion intakes; they are later applied to
estimate body burden in the use of the pharmacokinetic model. An absorption traction
was, however, used for the dermal contact pathway, as this will result in the estimation of
an amount crossing the skin boundary. This amount is considered "absorbed" and no
further absorption fraction is required for this dermal dose. Table 4-5 provides the
exposure media concentrations that were developed Chapter 4, and Table 5-5 in this
chapter provides the contact rates, and for dermal exposure, the full dermal contact
algorithm.
The important addition to the procedures originally laid out for dioxin-like
compounds, applied here to PBDEs, relates to the importance of the indoor versus the
outdoor environment. The Dioxin Reassessment used air concentrations from outdoor
ambient air measurements for the inhalation pathway, and measurements of dioxins in
background soils for soil ingestion and soil dermal contact. In the case of dioxins, the
primary sources are emissions from combustion sources into the open environment, with
subsequent accumulation in outdoor soils and, of primary importance to dioxin exposure,
in the terrestrial and aquatic food chains. In contrast, the primary cause for PBDE
exposures are their use in commercial products that are part of the indoor environment
(PC circuitry, foam cushions, etc), and, as described in Chapter 4, indoor air and indoor
dust concentrations of PBDEs are orders of magnitude higher than outdoor air and soil.
5-32
-------
Therefore, the use of outdoor measurements in air and soil does not appear appropriate
for inhalation and soil/dust pathways for PBDEs, Sjodin et al. (2004b) and Staplcton et
al, (2005, 2008b) recognized the importance of exposure to indoor dust when calculating
dust ingestion intakes in the hundreds to thousands of ng/day total PBDEs for adults and
children. Their intake calculations used "soil ingestion" contact rates, applying them in
total to their indoor dust measurements, as though the entire contact from "soil" is, in
fact, from "indoor dust." In contrast to their calculations of dust exposure intakes in the
100s of ng/day, their diet intakes (described in the previous section) were always under
100 ng/day total PBDEs, This shows the importance of exposures to indoor dust and
indoor air.
The approach taken in this assessment is to estimate a weighted average
concentration of "dust/soil" and air, which, in theory, considers what portion of total soil
ingcstion/dermal contact and inhalation comes from indoor dust and indoor air. The
surrogate used to estimate this portion will be "time spent indoors." The Exposure
Factors Handbook (EPA, 1977) provides tables on hours/day spent indoors, and for
adults the recommended number of hours per day is 21, which is 87.5% (or, expressed as
a fraction, 0.875) of the time. Therefore, a weighted average concentration, C[avg], of
"soil/dust" and air that will be used in the adult soil ingestion, soil dermal contact, and
inhalation pathways is C(indoor dust/indoor air) * 0.875 + Qoutdoor soil/outdoor air) *
0.125. This approach simplistically assumes that exposures are proportional to time
indoors versus outdoors, which could be totally incorrect for soil pathways if in fact, the
actual exposures to dust/soil all were to take place outdoors. In contrast, this is a
reasonable approach for inhalation, for obvious reasons. According to the Exposure
Factors Handbook, children under 11 years of age spend 19 hours per day indoors, so
their fraction indoors will be 0.792. Children ages 12 and higher spend the adult number
of 21 hours per day indoors.
Table 5-5 provides all of the exposure parameters and a brief description of the
pathways. Table 5-6 provides the final adult intake estimates for the 13 PBDEs for which
environmental media concentrations were derived in Chapter 4. The total dose to adults
was 547 ng/day, or on a body weight basis assuming a 70-kg adult, 7.8 ng/kg-day. The
following are key observations from Table 5-6, and some accompanying discussion:
5-33
-------
1) Predominance of the Dust Pathways: Given the procedures and parameters in
Table 5-5, coupled with the media concentrations provided in Table 4-5, it would appear
that the pathways of dust ingestion and dust dermal contact overwhelm the exposure of
adults to PBDEs. Dust ingestion and dermal contact together account for 82% of the total
exposure. The estimate of 363 ng day of PBDEs via dust ingestion is consistent with
Sjodin et al. (2004b), who calculated an intake of 400 ng/day assuming a dust ingestion
rate of 1.00 mg/day and comparable PBDE dust concentrations as were used in this
exercise. It is higher than the median estimate of 154 ng/day for adults estimated by
Stapleton et al. (2008b) based on measurements of PBDEs on hands and then an
empirical model to calculate intakes. The estimate of the PBDE intake by dust ingestion
is linearly related to the dust ingestion rate and the PBDE concentration. Therefore, use
of a lower rate of dust ingestion, such as the 0.56 mg/day used by Stapleton et al (2005),
would lead to a much lower ingestion of PBDE by the dust ingestion pathway. Further
discussions on the uncertainty of this pathway are provided in Section 5,6 below.
2) Prevalence, or Lack Thereof, of Food Exposures; After dust exposures, about 18%
of exposures are due to inhalation and food/water ingestion. In contrast to this finding,
the literature has an abundance of dietary surveys of PBDEs and discussions on exposure
via food consumption. A signi ficant portion of this literature deals with food exposures
in Europe and Asia. Studies on indoor dust concentrations for locations outside of the
United States tend to show lower concentrations of PBDEs (except for BDE 209; see
discussions in Chapter 4), and so for Europe and Asia, food very likely dominates overall
exposure. For the United States, the literature is now beginning to recognize the
importance of the dust pathway and much more work on dust exposures has occurred
since about 2005 (see summaries in Section 5,3).
3) Distribution of Dose Among BDE Congeners: The percentage of total dose
attributed to BDEs 47, 99, and 209 are about equal at 25%, 28%, and 27%, respectively,
followed by BDE 100 at 11%, for a total of 91% among those four congeners. Needless
to say, dust-related exposures dominated for the individual congeners, but mostly for
BDE 209, wher#dust ingestion and dust dermal contact explained 88% of the dose, while
it explained 68% of exposure to BDE 47 and 82% to BDE 99. Exposures to BDEs 138,
153, and 154 were all low at between 2-3% of total. BDE 183, generally considered to be
a marker for the presence of the octa PBDE formulation, although it could be present in
5-34
-------
environmental media as a result of denomination of higher brominated BDEs such as
BDE 209, was a small contributor to overall dose, at 1%. Although at least some
measurements were made of all the congeners on Table 5-6, it would appear that from a
dose perspective, perhaps BDE congeners 17, 28, 66, 197, and 206 can be neglected.
However, from a body burden perspective, BDE 28 makes up between 2-4% of total
body burden, while BDE 138, which makes up 2% of the dose, is virtually absent in body
tissues. This exemplifies the importance of the interplay between dose and body burden
when understanding and quantifying exposure to this class of compounds.
5.5. CONVERTING ADULT INTAKE DOSE TO BODY BURDEN
The intake doses in the previous section will be converted to body burdens in this
section. Assuming first-order kinetics and that PBDEs accumulate in body lipids, the
equation for the change in lipid concentrations over time is
t5CBDE/'5t = (Dud; (11fABSi;,,| ) BI.(t) k*( B. „ (I) (2)
where
C'bde is the congener-specific lipid-based concentration over time (ng/g Iwt)
Dbde is the daily dose of BDE (ng/day)
ABSbdb is the congener-specific and route-specific absorption fraction
BL(t) is the body lipid mass over time (g)
k is the first-order elimination rate of the congener in the body (day"1).
As presented here, k is assumed to be a constant, but it too could vary over time. The
solution to this partial differential equation is
CbdeO) - Cbdi:(0) * e'"U) + [ (Dbdb (t) * ABSbde) ' BL(t) ] * [(1 - c~kt)/k] (3)
where
Cbde(0) is the initial body burden at time 0
5-35
-------
Assuming a constant BL and a constant dose over time, the steady state lipid
concentration (ie,, when t approaches infinity) is easily calculated as:
Cbde = (Dbde * ABSBDE)/(k * BL) (4)
Equation (4) is used here to estimate an adult body burden, using the congener-
specific adult doses provided in Table 5-6. It is assumed that bodies are 25% lipid,
leading to a BL value of 17,500 g. Estimates of absorption for other POPs (dioxins and
PCBs) when ingested on soil were in the range of 0.30 to 0.70, and it was noted that
dioxins might be more bioavailable on house dust as compared to soil (Paustenbach et al.,
1997, 2006; ATSDR, 2004). Huwe et al (2008) studied the retention and excretion of
BDE congeners administered to male rats in com oil and household dust (MST reference
material). The rats were dosed at a "high" dose of 6 pg/kg-day or a "low" dose of 1
pg/kg-day. They were fed this amount for 21 days and then killed 24 hrs after the last
feeding. Fifteen BDE congeners were measured in adipose tissue and liver, and feces
collected during the experiment were also measured to provide a mass balance. By
calculating the amounts "retained" from the dose in the body by this mass balance, they
could surmise absorption by this amount retained. The retention amounts for the high
dose group exposed to PBDEs in dust, expressed as a fraction of total exposure, ranged
from 0.04 (for BDE 209) to 0.78 for BDE 100. For key congeners BDE 47, 99, 138, 153,
154, and 183, the fractions retained were 0.69, 0.44, 0.67, 0,73, 0.19, and 0.48.
respectively. Results were similar for corn oil and for the lower exposure amounts.
Although not ideal because it is not human data (although human data are understandably
rare) and it is not corroborated else, these values will be used in (he model ing of this
study for congener-specific values of ABS. McDonald (2005) assumed absorption
fractions of 0.78 to 0.94 for BDEs 47, 99, 100, 153, and 154 based on experiments with
rats, and the values he used will be used for these congeners for all other pathways in this
study, including inhalation and water/food ingestion. There is no additional absorption
assumed for dermal contact since an absorption fraction was already included in the dose
estimates for that pathway. McDonald (2005) also cited Geyer et al. (2004) to assign
half-lives of 2.9 to 11.0 yr to this same group of congeners, and these values will be used
here. Thuresson et al. (2006) provided estimates of 0.26 yr (94 d) for BDE 183 and 0.041
5-36
-------
yr (15 d) for BDE 209, and these will be used here. The first-order elimination rates,
keen. are easily calculated as 0.693/hl (where hi ® half-life). For lack of better
information, average values of absorption and half-life, 0.90 and 6 years, respectively,
will be assigned to congeners BDE 28 and 138, two other congeners which are present in
exposure media and in measurements of blood or milk. BDE congeners 17, 66, 85, 197,
and 206 will be neglected in this example as they are rarely found in exposure media or
body measurements. The final set of doses, pharmacokinetic parameters, predicted lipid-
based concentrations using these parameters in Equation (3) above, and then the observed
concentrations of the 9 PBDE congeners arc shown in Tabic 5-7.
Overall, the total concentration was predicted at 35.9 ng/'g lwt, while it was
observed at 36.3 ng/g lwt in blood and 44.1 ng/'g Kvt in milk. Predictions appear
reasonably close to measurements for 6 of the 8 congeners (for which there are
predictions and at least one observation). The prediction of BDE 47 at 10.0 ng/g lwt does
not appear to match well with the observed measurements of 20.5 ng/g hvt in blood and
26.0 ng/g lwt in milk. Conversely, the prediction of BDE 99 is higher than both
measurements, with a prediction of 14.6 ng/g lwt, while it was found at 5.0 and 5.4 ng/g
lwt in blood and milk, respectively. Otherwise. BDE 100 was predicted at 4.2 ng/g lwt
while it was found at 3.9 and 5,2 ng/g lwt. BDE 153 was predicted at 4.6 ng/g lwt, while
it was found at 5.7 and 4.8 ng/g lwt.
Obviously there are uncertainties with use of this pharmacokinetic model and the
assignment of parameters; these are discussed below in Section 5.6. Even with
uncertainties, one can make some valid and important observations about exposure of
Americans to PBDEs using this framework. For one, indoor exposures of soil/dust
ingestion, dermal contact, and inhalation dominate total exposures. While this might be a
valid observation based on intake estimates alone, it is strengthened significantly by this
simple pharmacokinetic exercise. If one calculates body burdens of these nine congeners
based on water/food ingestion alone (the dose is estimated at 92 ng/day or 1.3 ng/kg-day
with a body weight of 70 kg), the body burden is modeled to be 10 ng/g lwt, which is
much less than the observed medians of 36 and 44 ng/g lwt in blood and breast milk.
With 23.8 ng/g lwt predicted to occur from soil/dust ingestion, dermal contact, and
inhalation, the suggestion is that 70% or more of the United States body burden could be
explained by non-food exposures occurring in the indoor environment.
5-37
-------
A second observation is that while there may be significant exposure to BDE 209,
it does not appear to be showing up in large concentrations in adults, although it might be
a different story for children, as discussed below. The calculated intake dose is 27% of
the total, equal essentially to intakes of BDEs 47 and 99 (calculated at 25 and 28%,
respectively). However, because of a rapid half-life in the body (15 days), tissue levels
are predicted to be as low as 0.05 ng/g Iwt, which is numerically close to the observed 0.4
ng/g Iwt in breast niilk (some blood studies other than NHANES quantified BDE 209 in
blood, also, at around 0.1 ng/g Iwt). Said another way, concentrations of BDE 209 are
much lower in both observed measurements and predicted concentrations than the 5-30
ng/g Iwt found and predicted for BDEs 47 and 99.
However, to observe that BDE 209 has not been quantified at concentrations
approaching that of BDEs 47 and 99 in blood lipid may not be fully informative. The
presence of lower brominatcd BDE congeners in blood could have resulted from
debromination of BDE 209. Also, it may be that BDE 209 is not as lipophilic as other
BDEs. Measuring BDE 209 in extracted blood lipids might underestimate its presence in
blood because some BDE 209 may be present in the unanalyzed portion of a blood
sample.
5.6. EXPOSURE OF SPECIAL POPULATIONS OF INTEREST TO PBDES
This section discusses exposures of special populations to PBDEs, and develops
intake and body burden estimates for some of the populations. Speci fically, this section
provides an overview of body burden studies on occupational populations, it derives
intake estimates and resulting body burden estimates for infants, it derives intake
estimates for children within specific age ranges, and it discusses an important
observation that there appears to be a proportion of individuals at the high end of the
general population who are experiences significantly higher exposures than the remaining
general population.
5.6.1 Impacts to Infants from Consumption of Breast Milk
Using the profile of PBDEs in mother's milk, the dose to the infant was modeled
as follows:
D = c * f * IR (1)
5-38
-------
where D is the ingested dose of PBDEs (rig/day)
C is the concentration in milk fat (iig/g Iwt)
f is the fraction of fat in breast milk
IR is the ingestion rate of breast milk (g whole weight/d)
The dose term. D, can easily be converted to a body-weight-based dose term by
dividing by infant body weight, BW. The rate of ingestion of mother's milk and the
fraction of fat in the mother's milk were assumed to be constant over the duration of
breast-feeding. Smith {1987) reported that studies in Britain and Houston found that the
breast milk ingestion rate for 7- to 8-month-old infants ranged from 677 to 922 mL/d and
723 to 751 mL/d, respectively, and that breast milk ingestion rates remain relatively
constant over an infant's life. Smith (1987) also assumed that mother's milk has a 4% fat
content. These assumptions were adopted for the purposes of the modeling exercise
described here: IR = 800 g/d (which assumes 1 L milk weighs 1 kg) and f =• 0.04. Given
the total PBDE concentration of 44.1 ng/g Iwt in mother's milk (see Table 5-7), the total
dose to infants is 1,411 ng/'day. Assuming an average body weight of 10 kg for an infant
during the months of breast-feeding, a dose is calculated as 141 ng/kg-day. This is about
half the intake as the 307 ng/kg-day estimated by Schecter et al. (2005).
Infant impacts to breast milk and children's body burdens were handled in a
different manner than adult body burdens. The approach mirrors what was done for
dioxins by Lorber and Phillips (2002), who modeled the impact of dioxm-like
compounds in infants resulting from consumption of breast milk. The procedures in
Lorber and Phillips (2002), as applied to PBDEs instead of dioxin toxic equivalents
(TEQs), include the following: 1) total PBDEs were modeled instead of individual
congeners; one half-life (variable as noted below) and absorption (constant at 0.80) were
used to characterize this surrogate measure of exposure; 2) the dynamic solution to
Equation 2 above, shown in Equation 3, was used to be able to characterize changes in
dose, elimination half-life, body lipid tractions, and body weight over time, and 3) the
elimination halt-life for total PBDEs in infants will be more rapid than had been assumed
for individual congeners for adults.
Lorber and Phillips (2002) cited the pharmacokinetic modeling work of Kreuzer
et al. (1997) in their assignment of the overall elimination rate for dioxin TEQs from
5-39
-------
infancy into childhood. Kreuzer et al. (1997) found that the overall elimination rate of
2,3,7,8-TCDD in infants was driven by the non-metabolic process of fecal elimination.
Specifically, the overall elimination half-life of 2,3,7,8-TCDD was modeled to be about
0.4 yrs at birth, compared to 5 years or more at adulthood, because of the magnitude of
lipid loss in fecal elimination in infants. Since 2,3,7,8-TCDD accumulates in lipids (as
do the PBDEs), the rapid loss of lipids via fecal elimination results in a comparable rapid
loss in 2,3,7,8-TCDD. Lorber and Phillips (2002) assumed this intial, rapid half-life at
birth would rise to a half-life of 5 years by 18 years of age. They verified their approach
by showing how infant body burdens of TEQs were predicted to be much closer to
measurements in the literature when assuming this rapid half-life as compared to
assuming the longer half-life that would be appropriate for adults. Similar data are not
available for PBDEs, but given the similar lipophilicsty of PBDEs and CDD/Fs, this
approach seems reasonable. Specifically, this same half-life profile used in Lorber and
Phillips (2002) will be used here: a rapid half-life at birth rising to an overall,
representative half-life of 6 years by age 11. Drawing on information in the Exposure,
Factors Handbook (EPA, 1997), Lorber and Phillips (2002) also assigned temporally-
varying body lipid contents and body weights. Absorption of PBDEs will be assumed to
be 80% (absorption fraction = 0.8), and the initial body burden at birth will be 37 ng/'g
Iwt, similar to the adult total body burden from blood measurements. Derivation of
intake dose was provided above, and it was 1,411 ng total PBDE/day for one year of
breast feeding, followed by child intakes from 1 year on (see Section 5.6.2 below). Table
5-8 shows the final PK parameters, including dose estimates, from birth until age 19.
Figure 5-2 shows the final results of this exercise, including the impact on body burden
assuming the overall half-life of 6 years.
With the assumption of more rapid elimination earlier in life, the infant body
burden rises to about 125 ng/g Iwt at age 1 and continues to rise to above 200 ng/g iwt
through age 5. At this point the concentration begins to drop, ultimately decreasing to
100 ng/g Iwt by age 19. If total BDEs had an overall half-life of 6 yrs from birth on, than
the body burden would rise to near 200 ng/g Iwt by age 1, and continue to rise to about
325 ng/g Iwt by age 5, only then to slowly dissipate to levels below 100 ng/g Iwt by age
19.
5-40
-------
The validity of these predictions cannot be easily verified because of the lack of
data in the literature. However, there is one study described above in Section 4.2 on
blood levels of four individuals within a family in California. Fischer et al. (2006)
present data on the two parents, a 35 and 37 year old, a 5-year-old daughter, and an 18-
month-old son in September and December of 2004. The sum of BDEs 47, 99, 100,153,
154, and 209 in the parents ranged between 64 and 147 ng/g lwt in the two sampling
dates, but the concentrations in the children were much higher. The 5 year-old daughter
had concentrations of 237, 239/249 ng/g lwt (the last two were duplicates of the same
December sample) of the 5 congeners for the September and then December samples.
The toddler had the highest concentrations of all: 418 and 488/476 for the five congeners.
Also of interest were very high initial concentrations of BDE 209, which dropped
significantly in both the 5 year old and the toddler: the 5 year old had concentrations of
143 ng/g lwt of BDE 209 in the September sample and 9/12 ng/g lwt in the December
sample. The toddler had 233 and 19/26 ng/g lwt in the September and December
samples, respectively. Discounting laboratory error, the authors attribute the higher
concentrations in the children to exposure to house dust, and the drop in 209 levels
between the September arid December samples to the short half-life of BDE 209 in
humans. While the authors have discounted laboratory error, it would appear that a
decline by an order of magnitude in both the toddler and infant is substantial, and could
be due to some difference in the two laboratories. However, a decline of this magnitude
is plausible because of the short half-life of BDE 209 in humans, as noted by the authors.
The higher levels of the other congeners in the toddler were attributed by Fisher to
consumption o f breast milk, although it could also be due to exposure to house dust.
These high measurements in children support the simple PK modeling done here, but are
not sufficient to verify it. Body burden measurements in children remain an uncertainty,
although the modeling and this single family provide some evidence that body burdens in
children are higher than adults, with the suggestion that they could be very high. Further,
the analysis here suggests that these higher levels could persist through childhood into
early adulthood.
5.6.2. Childhood Intakes
5-41
-------
The total dose to adults was 547 ng/day, or 7.8 ng/kg-d (assuming an adult body
weight of 70 kg). Using the intake rates provided in Table 5,5 and the exposure media
concentrations developed in Chapter 4 (shown in Table 4.5), the total dose for the three
age ranges of children were as follows: 751 ng/day for the 1-5 age range, 439 ng/day for
the 6-11 range, and 536 ng/day for the 12-19 age range. On a body weight basis, the
doses are 50.1 ng/kg-d for ages 1-5 (assuming 15 kg bw), 14.6 ng/kg-d for 6-11 (30 kg),
and 9.2 ng/kg-d for 12-19 (58 kg). The much higher dose for the child age 1-5 was due
to the doubling of soil/dust ingestion from 50 mg/day to 100 mg day. Otherwise, the
trends as elaborated above for adults in Section 5.4, such as the predominance of the soil
ingestion and dermal contact pathways, were similar for children.
5.6.3. Body Burden Data to Characterize Occupational Exposures
Limited studies from Sweden and from China suggest that PBDE concentrations
are elevated in occupational groups exposed to likely sources of PBDEs. One study,
which looked at incinerator workers in comparison to general population exposures, did
not find a difference. The only study of occupational exposures in the United States
found a significantly higher (p < 0.05) level of PBDEs in workers in foam recycling
facilities and individuals who installed carpet padding manufactured from recycled foam.
This study from the United States (Stapleton et al., 2008a) included 12 foam
workers from two foam recycling facilities (one in Maryland and one in California), 3
carpet layers who worked in association with the California facility, and 5 control group
individuals comprised of spouses and clerical workers from the facilities. The median
total PBDE concentrations (comprised of BDEs 17, 28, 47, 66, 99, 100, 153, 154, and
183) in the foam workers, carpet installers, and control were 160, 178, and 19 ng/g lipid,
respectively. The body burdens of the workers were dominated by BDE 47, which
explained 50-60% of the total concentration, followed by 99 and 153. which both
contributed 13-20% of the total.
Thirteen PBDE congeners were quantified in serum in a group of 19 PC
technicians (PC techs), and the results were compared against hospital workers, and PC
clerks in Sweden (Jakobbson et al., 2002). Moreover, within these two comparison
groups, refined characterizations included women who had never breast fed (NBF),
hospital cleaners (HC), and computer clerks (PCC). Results were provided for BDEs 47,
5-42
-------
153, 154, 183, and 209. There were distinct differences among the groups. The median
value of BDE 47 was fairly similar among the groups; PC techs had a median of 1.3 ng/g
lwt, HCs had a median of 1.6 ng/g lwt, PCCs had a median of 1.5 ng/g lwt, and NBFs had
a median of 2.1 ng/g lwt. BDE 153 was the highest in the PC techs, and significantly
higher than the other three groups: 2.6 ng/g lwt for the PC techs, while it was 0.6, 0.8,
and 0.8 ng/g lwt for the other three groups. While these other three groups did not have
BDE 209, it was quanti fied at a median of 1.5 ng/g lwt for the PC techs. There were no
correlations with age and BDE levels, but a correlation was found between time on the
job and BDE 153 concentrations in the PC techs.
Eight Swedish employees at a recycling plant and 4 rubber mixers volunteered to
donate blood samples during their summer vacations, in 1998 and 2000, respectively
(Jakobsson et al., 2003). The first blood sample was drawn at start of vacation, 3-4 days
later, and then at additional time periods during the 4-5 week vacation. Tlie samples
were analyzed for BDEs 47. 153, 183. and 209. Assuming there was a constant baseline
of BDEs in these workers, which they obtained from serum measurements on non-
exposed individuals, the authors calculated half-lives. In addition to these 8 temporally
followed workers, the authors measured blood from 60 other subjects; in all, their study
had 107 observations from 68 subjects. They found distinct exposure levels and patterns
of BDE congeners as a function of which workers they measured. The rubber workers,
who were exposed only to decaBDE, had markedly elevated levels of BDE 209, which
averaged 27-35 ng/g lwt, with a high of 278 ng/g lwt. Unexposed individuals had levels
at ND-2.4 ng/g lwt. The electronic dismantlers were expected to be exposed to all BDE
congeners, and they were, with concentrations of BDEs 47, 153, 183, and 209 averaging
2.9, 4.5, 7.9, and 4.8 ng/g lwt for 47, 153, 183, and 209. While they found that BDE 209
is bioavailable based on finding high levels in an oecupationally exposed group (the
rubber workers), they also found it had the shortest half-life. There was, overall, an
inverse relationship between half-life and degree of bromination-lower brominated
congeners had longer half-lives: BDE 203 was 37 days, BDE 183 was 110 days, BDE
153 was 680 days, and BDE 154 was 270 days.
Qu et al. (2007) measured levels of 14 BDEs (including BDE 209) in electronic
waste dismantling workers, in residents living within 50 km of the dismantling area, and
a reference group with no exposure near an occupational setting in Quongdong, South
5-43
-------
China. The median concentrations of each congener were determined, and the sum of
these individual medians from the three groups were as follows; 126 ng/g hvt
(dismantling workers), 35.1 ng/g Iwt (residents living within 50 km), and 9.4 ng/g lwt
(reference group). BDE 209 was the highest among all groups, explaining between 50
and 70% of total concentration. Interestingly, BDE 207 was the second highest in the
electronic recycling group and the residents living within 50 km, explaining 8 and 15% of
the total, respectively. Exceedingly high concentrations were found in one 18-year-old
male electronic waste worker: BDEs 28, 183, 208, and 209 were found at 148.3, 60,2,
66.2, and 3,436 ng/g lwt, respectively.
Lee et al. (2007) measured 13 congeners (not including BDE 209) in 92 blood
samples, including 30 from incinerator workers, 51 from nearby residents, and 11 from
controls in 2001 and 2002. The average total concentration was 16.84 ng/g lwt, and there
was only a slight difference between the incinerator workers, who had the highest
concentrations at an average of 19.24 ng/g lwt, and the other two groups: residents at
15.22 ng/g lwt and controls at 17.74 ng/g lwt. The difference between the groups was not
significant, and there was no other correlations found, including to age, weight, dietary
habits, and others, with the exception of sex; males had 15% higher levels compared to
women. BDE 47 dominated the profile, explaining 33% of the profile, followed by
BDEs 153 (24%), BDE 183 (17%), BDE 99 (15%), and BDE 100 (7%).
5.6.4, Elevated Exposures at the High End of the General Adult Population
The analysis of the 2003/2004 NHANES blood concentration date described by
Sjodin et al. (2008), shown in Table 5.7 and used as a primary comparison to the
predicted concentration in the model, did not include statistics on total concentration, just
statistics on individual congeners. The congener-specific median concentrations are
shown in Table 5.7, and, when added together, they lead to a total concentration of about
36 ng/g lwt. The statistics provided for individual congeners show that the 90th percentile
concentration of all the key congeners in United States citizens are about 4-6 ti mes higher
than the 50Ul percentile and the 95lh is near or more than 10 times higher than the
medians of the congeners. Perhaps more noteworthy was the finding that the highest
total concentration found in an individual in NHANES 2003/2004 was 3680 ng/g lipid,
which included BDE 47 at 2350 ng/g lipid (Sjodin et al., 2008). This highest individual
5-44
-------
had concentrations about 100 times higher than the median in the population. Clearly,
there are individuals with very much higher concentrations than the central tendency
median selected for the point estimate exercise of this chapter.
In comparison, this is not the same trend as generally found for dioxin. Ferriby ct
al (2006) statistically evaluated the concentrations of polyehlorinated dibenzo-/?-dioxin
(PCDD) and dibenzofuran (PCDF; the combination abbreviated PCDD/F) concentrations
from NHANES 2001/2002. Determining the toxic equivalent (TEQ) concentrations of
the 17 toxic PCDD/F congeners for individuals in the survey, they provided population
statistics for PCDD/F TEQ concentrations. The median concentration in the population
was 14.4 pg/g lwt, the 95% was 45.2 ng/g lwt, and the maximum found was 139.2 ng/g
Iwt. In this case, the maximum found was 10 times and not 100 times the median. To
illustrate the difference in these NHANES results, Figure 5.3 shows the comparison of
NHANES percentiles for PCDD/F TEQ and BDE 47 findings at the 25%, 50%, 75%,
95%, compared to the maximum found in the NHANES surveys (dioxin data from
Ferriby et al, 2006, and PBDE data from Sjodin et at., 2008). While the percentiles of
dioxin concentrations within the general adult population, expressed as a fraction of the
maximum found, range from between about 0.10 at the 50th % and 0.30 at the 95th % for
PCDD/Fs, they range much more narrowly at <0.01 at the 50th % to 0.07 at the 95% for
BDE 47.
Other studies show a similar disparity between the median (or geometric mean, if
that is what the study authors provided) and high value of their body burden measurement
study. For example, Morcland et al. (2005) measured the blood of 93 individuals, 79 of
which were anglers. The highest congener found was BDE 47, at a geometric mean of
13.3 ng/g lwt, followed by BDE 99, at 3.2 ng/'g lwt. Although they did not report total
concentrations, BDE 47 was found at a high of 1,388 ng/g lwt, which was 100 times the
geometric mean for this congener, and the high BDE 99 concentration was 546 ng/g lwt,
about 20 times the geometric mean. Petreas et al (2003) measured BDE 47 in adipose
tissue from 32 women in the San Francisco area, and found a median of 16.5 ng/g lwt,
with a high of 510 ng/g lwt, 30 times the median.
Evidence suggests that these elevated exposures at the very high end of the
population could very well be due to dust exposures. Nearly every study that has been
conducted measuring dust concentrations from different locations very similarly finds a
5-45
-------
reasonably log normal range until the very last few samples which are substantially
higher than the rest of the population. Following is a bullet summary showing this trend
from several dust studies:
a) Stapleton et al (2008b) measured BDEs on hands using hand-wipes. Their results and
exposure estimates were described earlier in Section 5.3. The key finding of note for this
discussion is that they measured BDEs in 33 individuals, and that for 32 of them, the
measurement was under 500 ng hand, but the 33Kl individual had about 2000 ng on his
hand. The median from this population was 128 ng/hand.
b) Allen et al (2008) collected 108 bulk dust samples from about 20 homes in the Boston,
MA area, during two sampling events in 2006. Samples were characterized as having
come from the "living room", "bedroom", and "vacuum bag" (meaning location
unspecified). The geometric mean total concentrations found in the living room,
bedroom, and vacuum bag werel3,732; 6,255; and 4,269 ng/g dwt, respectively. The two
highest total concentrations were 544,000 and 269,000 ng/g dwt, found in vacuum bag
samples from a single individual's home. The first high measurement was dominated by
BDE 209, explaining 97% of the 544,000 ng/g dwt found; the dominant congener was not
identified for the second highest sample.
c) Harrad et al (2008) measured 78 samples in four cities in Canada, New Zealand, the
United Kingdom, and the United States. The median total concentration of 28 samples
from the United Kingdom was 2,900 ng/g dwt, but one sample was found at 520,000 ng/g
dwt, with essentially all of it (>99%) explained by BDE 209. In 20 United States
samples (from 2 cities in Texas), the median was 3,500 ng/g dwt, with a maximum of
17,000 ng/g dwt, dominated by BDEs 47, 99, 154, and 209.
d) Two studies of house dust in the United States showed similar trends of having a
relatively consistent finding for most of the homes, and then one or two having
substantially higher concentrations. In one, Stapleton et al (2005) analyzed dust samples
from 17 homes in the Washington, DC area. They found total concentrations (including
BDE 209) to be less than or near 7,000 ng/g dwt in 15 homes, but in two homes they
5-46
-------
found 14.990 and 30,100 ng/g dwt, respectively. These samples were dominated by
BDEs 47, 99, and 209 in comparable amounts. In the other, Sharp and Lunder (2004)
analyzed dust samples in 10 homes from 9 states. Concentrations found in 8 of the
homes were near or less than 6,000 ng/g dwt total, but in two homes, the total
concentrations were 16,000 and 41,000 ng/g dwt. These high samples were dominated
by BDEs 47, 99, and 209.
Perhaps a certain behavior within a household or office that puts one in close
contact with a product containing BDEs or in close contact with dust, heavily laden with
BDEs explains this tendency in populations to have very high exposures at the high end.
This is clearly an uncertainty that requires further investigation. Besides dust, the cause
for these higher exposures could also be a high consumption rate of a particular food
product that may be impacted by local conditions. For example, Sjodin et al (2000)
found that Swedish fisherman consuming large amounts of Baltic Sea fish had a median
median BDE 47 level five times higher than that of nonconsumers. However, only a
small statistically insignificant difference was found between urban anglers and non-
anglers in a study of 93 individuals (79 anglers, 14 non-anglers) in the United States
(Morland et al, 2004). Discussions below focus on the question of wliether dust-related
exposures are likely to dominate the median exposures, ft should also be noted that
findings of PBDE body burdens in other United States studies are not. very different from
the NHANES study. As described in Section 5,2., the means or medians from essentially
all United States studies are within a reasonably narrow range of PBDE concentrations:
30 to 100 ng/g Ivvt. The consistency in central tendency findings in these studies suggests
consistency in exposure of Americans to PBDE.
5.7. UNCERTAINTY AND VARIABILITY IN ESTIMATING INTAKE DOSE
AND CONVERTING THAT DOSE TO A BODY BURDEN
The analysis in this chapter has taken a "point estimate" approach in the intake
estimation, the pharmacokinetic modeling, and the comparison with measured body
burdens. In so doing, it has arrived at a fairly broad-reaching finding that the bulk of
exposures are indoor-dust related. The exercise revealed that between 7 and 8 ng/kg-day
of total PBDE exposure appears necessary in order to reproduce the median body burdens
5-47
-------
seen in the adult population. This finding is based on pharmacokinetic modeling—given
the model and parameters chosen for BDE congeners, the median body burden would not
be duplicated unless the intake dose were in the 7-8 ng/kg-day range. The point estimate
approach suggested that over 80% of this total intake came from dust ingestion,
inhalation, and dermal contact with dust. However, there are uncertainties throughout
this exercise, as well as variabilities in exposure. As discussed in the previous section,
there appears to be exposures at the very high end of the general population that are
substantially higher than the central tendency exposures that are characterized by a dose
in the range of 7 to 8 ng/kg-day, dominated by dust exposures. This construct of the
exercise does not guarantee that the finding of the importance of the dust pathways is
proven. The purpose of this section is to lay out all of these uncertainties and variabilities
so that this primary finding and other findings in this chapter can be understood in their
proper context.
a) Uncertainties With Estimates of Dust Intakes of PBDE:
/. soil/dust ingestion amount: The amount of 50 mg/day assumed for adults is a
classic uncertainty. Described as an "average" value for soil ingestion in the Exposure
Factors Handbook (EPA, 1997), the actual amount of ingestion of indoor dust + outdoor
soil could very easily be much lower. While Sjodin et al. (2004b) assumed a dust
ingestion rate of 100 mg/day, Stapleton et al. (2005), on the other hand, assumed an adult
housedust ingestion rate of 0.56 mg/day, over two orders of magnitude lower, explaining
their estimate of 3.3 ng/day of exposure to PBDEs via ingestion exposure to household
dust. As noted earlier in the chapter, this value of 0.56 mg/day housedust was listed in
EPA's Exposure Factors Handbook (EPA, 1997), which cited Hawley (1985) for using a
value of 0.56 mg/day to characterize adult exposure to housedust from normal activities
in the house (higher exposures of over 100 mg/day resulted from "work in the attic"). As
noted, Stapleton et al. (2005) estimated an exposure to total PBDEs of 3.3 ng/day based
on this assumption. It would seem that their subsequent studies, and those of others,
essentially prove that this is much too low an estimate for indoor dust ingestion, at least
in the context of estimating exposure to indoor dust. Speci fically, their recent study
(Stapleton et al., 2008b) directly measuring BDEs on hands using wipes, and then using
5-48
-------
an empirical model to estimate hand-to-mouth exposures, found a median exposure
estimate for adults of 154 ng/day.
2. soil/dust fraction: The procedure here to characterize ingestion of housedust
is to multiply a total amount ingested by a fraction that comes from the house. The
assumption that 0.90 for this fraction, based on data suggesting 90% of the time is spent
indoors, is conservative. Paustenbach et al. (1997) looked at data suggesting that 50% of
indoor dust originates from outdoor soil. This may be less important for the current
exercise because dust concentrations were taken from indoor dust measurements, so the
origin of the dust is not relevant when using measured dust concentrations directly. Still,
the assumption that 90% of the ingested amount is indoor dust is really not substantiated
in the literature. It is appropriate for "time spent indoors" and, for obvious reasons,
would be reasonable for inhalation exposures.
3. bioavailability ofPBDEs in dust: Bioavailability was considered in the
context of absorption fractions for dust ingestion and dust dermal contact. In the dermal
contact pathway, an absorption fraction of 0.03 (3% absorbed) was assumed for all
congeners (Table 5.5), and in the dust ingestion pathway, different absorption fractions
for each congener ranging from about 0.20 to 0.80, were assumed. There is a subtle
difference in the way in which results are presented in this study, in that for all pathways
except dermal contact, the intake dose is calculated just by the contact rate (food
ingestion rate, inhalation rate, etc) multiplied by the concentration of BDEs in the contact
media (in food, air, etc). For the dermal contact pathway, this intake dose has already
considered this absorption fraction of 0.03. For the other pathways, the absorption
fractions are used in the PK modeling of body burdens due to intake doses; the absorption
fractions reduce the intake dose to consider absorbed dose. In any case, all of the
absorption fractions are uncertain. The value of 0.03 for dermal contact was used in the
modeling of background exposures to dioxin based on literature showing that this tightly
sorbed contaminant would not desorb readily from soil contacting the skin and then
penetrate the skin surface to a great extent (EPA, 2003). However, this may not be true
for PBDEs, or at least true for the case where the vehicle is house dust rather than soil.
Only a few studies were found that have measured dermal absorption of PBDEs.
Although none of these used PBDEs sorbed to dust, soil or other solid matrices, they
provide some indicat ion of the dermal absorption poten tial of these chemicals. Hughes et
5-49
-------
al (2001) examined the in vitro dermal absorption of [!4C]decabromodiphenyl oxide
(DBDPO). Skin from the adult hairless female mouse was removed and mounted in
flow-through diffusion eells. The chemical was applied to the skin at three dose levels (6,
30 and 60 nmol) in a volatile vehicle (tetrahydrofuran). The 24-h cumulative percent of
the dose in the receptor fluid was 0.07-0.34%. The percent of the applied dose detected
in the skin after 24 hr ranged from 2 to 20%. Staskal et al (2005) conducted a mouse in
vivo study to measure dermal absorption of 2,2\4,4'-tetrabromodiphenyl ether (RDE 47).
A single dose was applied to the skin in an acetone solution. About 62% of the
administered dose was absorbed over a 5 day period. Roper et al (2006) studied the
dermal absorption of 2,2,,4,4"-tetrabromodiphenyl ether (BDE 47) using rat and human
skin in vitro. A single dose was applied to the skin in an acetone solution. The total
absorbed dose in human skin after 24 hr was 3.13% (1.93% in the receptor solution and
1.2% in the skin). The total absorbed dose in rat skin after 24 hr was 17.94% (14.81 % in
the receptor solution and 3.13% in the skin). Roper et al (2006) showed that rat skin was
more permeable than human skin to TBDE and this has also been observed for other
chemicals and other rodents in multiple studies (i.e. van Ravenzwaay et al., 2004). All
three studies applied the chemical in a solvent which evaporated rapidly leaving a residue
of pure chemical on the skin. This would tend to increase the absorption relative to a
similar dose which has been absorbed to dust. Thus, the 3% absorption based on human
skin in vitro testing by Roper et al (2006) is probably the most relevant data, but may be
high for dust exposures. USEPA (2004) recommends 3% absorption for TCDD in soil.
The similarity of TCDD to PBDEs in terms of lipophilicity and molecular size, adds
support to the assignment of 0.03 for PBDEs. The absorption fractions for dust
ingestion were derived from a study by Huwe et al (2008), where BDE congeners were
administered to male rats in com oil and household dust. It was found that absorption
amounts were similar for corn oil and dust, and the congener-specific results from that
study, showing absorption fractions ranging between 0.18 and 0.78, were directly used
for the dust ingestion pathway in this study. The absorption fractions for food ingestion
and inhalation were higher at between 0.78 and 0.94. It is known that organic
compounds sorbed to soil, or dust, are less bioavailable than when ingested in food or
inhaled. Paustenbach et al. (2006) examined the literature on the bioavailability of
2,3,7,8-TCDD. In their monte carlo simulations on contaminated soil exposures, they
5-50
-------
assumed that the oral absorption of this compound in contaminated soil followed a
lognormal distribution with a range of 0.5 to 63% and a mean value of 35%. ATSDR
(2004) reviewed the literature on bioavailability of PCBs in soil and concluded that a
range of 40-65% was appropriate. Paustenbach et al. (1997) did comment on the fact that
the bioavailability of contaminants on house dust, in general, is much greater than that in
outdoor soil, because house dust particles are finer than soil particles. In a backward
pharmacokinetic modeling exercise in which McDonald (2005) derived intake estimates
that would correspond to measured body burdens, he assumed absorption fractions of
0.78 to 0.94 for BDEs 47, 99, 100, 153, and 154. This was not pathway-specific and was
based on absorption experiments in rats, and although the carrier in the experiments was
not noted, it was unlikely to be soil. EPA (2003) assumed that the absorption fraction of
dioxin from a total intake dose was 0.80, but dioxin intakes are dominated by food and
not dust/soil. In general, the point is made that while an intake dose of PBDEs could be
dominated by dust ingestion and dermal contact with dust, there remains uncertainty as to
how much of the doses by these pathways get absorbed to eventually appear in blood.
c) Uncertainties and Variabilities With Other Pathways:
It is reasonable to conclude that there is less uncertainty in characterizing central
tendency for water ingestion, food ingestion, and inhalation, as compared to soil/dust
ingestion and soil/dust dermal contact. Food intake quantities as developed in the
Exposure Factors Handbook (EPA, 1997) were derived from the U.S.Department of
Agriculture Continuing Survey of Intakes by Individuals (CSFIT), which is a survey of
high quality used in many assessments, such as the U.S. Food and Drug Administration's
market basket surveys, which are used to determine intakes from measurements of
contaminants in sampled food. Water ingestion and air inhalation are also
comprehensively studied and the contact rates chosen are characterized as very
reasonable central tendency point estimates. Furthermore, most food surveys arrive at
comparable concentrations in food products. One exception described in Chapter 4 was
the finding by the United Kingdom's Food Standards Agency of high BDE 209 in all
food sampled (FSA, 2006). The quality of this data are unknown, and might be
questioned since it differs from essentially all other studies of PBDEs in food.
Interestingly, BDE 209 was found at high levels also in dry cat food (Dye et al, 2007).
5-51
-------
While BDE 209 was found at low concentrations, <0.01 ng/g wwt, and comprised only a
small percent of total concentration, <10%, in canned "wet" cat food products, it was
found at levels between 0.42 and 2.28 ng/g wwt in dry cat food and comprised 73 - 87%
of total concentrations in four types of dry cat food (chicken, salmon, poulty and fish, and
adult dry food). When combining food concentrations of PBDEs with food intakes, most
studies around the world, including the United States, have arrived at adult food intakes
in the 1-2 ng total BDE/kg-day. Inhalation and water ingestion have mostly arrived at
adult exposures less than 1 ng/kg-day. These findings can easily be seen in Table 5-4,
which lists intake estimates from all pathways from studies around the world. Only the
UK FSA study showing unusually high BDE 209 concentrations arrived at food intakes
greater than this 1-2 ng total BDE/kg-day range.
It should also be noted that there may be pathways not considered in this
assessment For example, while the assessment considered dermal contact with dust, it
did not consider dermal contact with PBDE-treated products. Dermal absorption of
PBDEs may occur via direct contact with treated materials such as clothing, carpeting,
upholstery, etc. Wester et al (1996) has shown that chemicals in fabric can transfer from
fabric into and through human skin. This was based on human skin in vitro experiments
with glyphosphate and malathion applied to cotton sheets. Absorption from diy cloth
was found to occur but it was less than the chemicals in aqueous solutions. When the
cloth was wetted with water to simulate sweating, absorption increased. For example,
absorption of malathion from aqueous ethanol solution was 8.77 ± 1.43%. This decreased
to about 0.60% for dry cotton sheets. However, absorption from cotton sheets increased
to 7.34 ± 0.61% when wetted with aqueous ethanol.
d) Uncertainties With Pharmacokinetic Modeling
The choice of the simple 1-compartment Ist order model is reasonable for PBDEs.
Like dioxins, for which the model lias been extensively and successfully used (EPA,
2003, and other citations not provided), PBDEs are lipophilic and persistent in body.
The application of the model at stead y state instead of in a temporally variable mode
could introduce uncertainties. For dioxins, it was found that application of the model at
steady state using dose estimates developed for current conditions would underestimate
average adult body burdens by about one-half (EPA, 2003). This is because of high
5-52
-------
dioxin intakes in the middle decades of the twentieth century (the 1960s until about 1980;
much higher than current intakes), such that body burdens of older adults are higher than
younger adults, driving up the current average adult population body burden. Similar age
trends were not found in BDE population studies. In fact, the younger age ranges were
found to have higher body burdens in some studies. Unlike dioxins, PBDEs were not in
the environment prior to their introduction into household products beginning in the
1970s and body burdens were first found to contain PBDEs in the 1980s. With the
voluntary withdrawal of pentaBDE and octaBDE formulations in 2004, there may be
declines seen in the key congeners of these formulations in the most recent surveys.
Generally, though, it may be reasonable to assume that exposures have at least remained
steady if not increased (because of perhaps rising levels in dust as more products using
the PBDEs were incorporated into modem living) between 1980 and the present. This
provides a reasonable justification for use of the simple PK model in a steady-state mode,
at least in comparison with dioxin, where steadily declining exposures leads to an
underestimate of population body burdens if using a steady state model
The uncertainties associated with the dose estimates were discussed in die
previous section. Absorption was also discussed in the previous section in the context of
absorption of PBDEs from ingested dust. There was limited choice in the literature from
which to make selections of elimination half-lives. Only one study, Gever et al. (2004)
was found to assign half-lives to BDEs 47, 99, 100, 153, and 154 of 2.9 to 11.0 yr. Only
one study was found, Thuresson et al. (2006), which provided estimates for BDE 183 of
0.26 yr (94 d) and for BDE 209 of 0.041 yr (15 d).
With these selections, it was found BDE 47 appears to be underpredicted. The
prediction was 10.0 ng/g lwt, compared to observations of 20.5 ng/g lwt in blood and
26.0 ng/g lwt in breast milk. The cause for this underprediction is not known, but it could
very easily be the assumed half-life in humans. At 3.0 years, it dissipated nearly twice as
fast as BDE 99, which was assigned a half-life of 5.4 years. Had it dissipated at a half-
life of 10 years, the prediction would jump to greater than 30 ng/g lwt, now more than
twice the prediction for BDE 99, and more in line with current measurements. In the
same vein, BDE 99 was overpredicted. It was predicted at 14.6 ng/g lwt, while is
measured as 5.0 ng/g lwt in blood and 5.4 ng/g lwt in breast milk. Perhaps the half-lives
of both congeners should be reversed. However, other evidence in the literature supports
5-53
-------
the assumption that BDE 47 is eliminated more rapidly than BDE 99. In an experiment
where BDEs 47, 99, 100, and 153 were all administered intravenously to mice, Staskal et
al (2006) found that tissue concentrations were highest for BDE 153, followed by BDEs
100, 99, and 47. Similar to the human data, this mice data suggests a more rapid
elimination of BDE 47 as compared to BDE 99 and also other key congeners. On the
other hand, the human body burden data show that BDE 47 concentrations are the highest
of all congeners and about four times higher than BDE 99 concentrations. The modeling
in this study (given uncertainties of course) suggests that the dose of the two congeners is
about equal. Logically, therefore, one would speculate that BDE 99 would be more
rapidly eliminated as compared to BDE 47, but that is not what the human toxicokinetic
data, or the rodent data cited, has found. This trend of higher BDE 47 concentrations in
humans may suggest debromination of BDE 99 (or higher brominated congeners) to form
BDE 47 in the body. In other words, the modeling may correctly determine body
burdens due to intakes, but the half-lives assigned might not reflect full elimination from
the body but rather higher brominated congeners are being transformed to lower
brominated congeners. However, this is much too speculative given current information
on human and animal metabolism. It might be possible that the one study measuring the
elimination of these congeners in an occupational cohort was not appropriate for the half-
lives of these two congeners in the background population. In any case, it would be
premature to use this framework to "calibrate" congener-specific half-lives in humans.
Still, it gives an indication of where key information gaps are.
With regard to the final total concentrations, they were predicted at 35.9 ng/g lwt
and measured at 36.3 ng/g lwt in blood and 44.1 ng/g lwt in breast milk. Other studies in
breast milk showed different concentrations, some lower and some higher. Most other
blood measurement studies were somewhat higher than this NHANES result, but most
studies found central tendency total concentrations to be well under 100 ng/g lwt, so the
prediction is about within a factor of 2 of most measurements, and in the case of
NEf ANES blood and the selected breast milk concentrations, predictions came very close
to measurements.
The results of this simple PK modeling exercise are consistent with those of
McDonald (2005), although he took the opposite approach: he started with body burdens
and used reverse PK modeling to derive the intake doses which would explain the body
5-54
-------
burdens. He used the same PK parameters (absorption fractions and congener-specific
half-lives) as were used in this exercise, although he did not separately consider dust
ingestion with a different absorption fraction. Compiling data on PBDEs in breast milk
(including some of the same data summarized above), he found a median concentration
of 48 ng/g Iwt, a mean of 90 ng/g lwt, and a 95% of 302 ng/g lwt These totals were the
sum of BDE congeners 47, 99, 100, 153, and 154. Modeling these congeners
individually, he found that the doses of total PBDEs that would explain these body
burdens were 8.5, 16.0, and 53.6 ng/kg-d for the median, mean, and 95%. In the exercise
above, dose was forward calculated from exposure media concentrations and contact
rates, and the adult dose was estimated at 547 ng/day total, or on a body weight basis
assuming a 70 kg adult, 7.8 ng/kg-day. This dose was estimated based on
"representative" central tendency media concentrations and average contact rates, so the
consistency of 7.8 ng/kg-day derived by a forward calculation for "average" conditions
with McDonald's 8.5 ng/kg-day derived by a retrocalculation from the median, body-
burden is very encouraging.
What is equally interesting is that McDonald (2005) cites European body burdens
which are more like 10 ng/g lwt, as well as estimates of dose made by Europeans based
on food alone, which are in the range of 1.0 ng/kg-day, and using his PK framework.
McDonald (2005) found quantitative consistency in those two quantities (i.e.. the dose he
reverse-calculated from body burdens matches the forward calculated dose by food
ingestion developed by European researchers). In essence, he is implying that the
primary pathway of BDE exposure to Europeans is through food consumption. It is
unclear why house dust and other indoor sources of PBDEs were not considered. These
pathways were considered by Harrad et al. (2006) in a study conducted in the United
Kingdom. They observe that dust concentrations are much lower in the United Kingdom
as compared to the United States, while food concentrations are comparable in the two
countries. They suggest that the higher body burdens found in the United States are
likely to indoor dust exposures. If not entirely due to indoor dust, there may be other
pathways of exposure Americans have that Europeans do not. In. any case, the prevalence
of lower brominated BDEs in dust in the United States and the subsequent impact on
United States body burdens is the key issue identified in this assessment, and, needless to
say, it requires further research.
5-55
-------
5.8. OVERALL FINDINGS OF EXPOSURE OF AMERICANS TO PBDES
Examination of literature data and exposure exercises in this chapter support these
general findings:
1. PBDEs bioaccumulate in lipids, and body burden measurements are expressed on a
ng/g lipid weight (lwt) basis. Nearly all body burden data are from blood and breast
milk. Body burdens of Americans are higher than body burdens of individuals in other
countries; most of the non-United States data are from Europe. Data suggests total PBDE
body burdens in the range of 30 to 100 ng/g lwt in. Americans, while it is Jess than 10
ng/g lwt for people in other countries. There is some suggestion that body burdens are
higher in women than men; breast milk data suggest lipid-based concentrations that are
higher than the lipid-based blood concentrations (and blood measurements are from both
men and women).
2. The predominant congener found in body burden studies is BDE 47, explaining about
50% of the total concentration. The second most found congeners are 99 and 153, both
explaining in the range of 10-20% of total concentrations. Most of the studies have not
measured BDE 209, but when measured, it was generally found in about half the samples
at low levels near 1-2 ng/g lwt
3. The limited data on adipose tissue are consistent with these findings for blood and
breast milk. Limited occupational data support the observation that individuals in
occupations which would lead to higher exposures to specific congeners have higher
concentrations of those congeners than the general population. Very limited data are
available on body burdens of children and infants, but in one study of a family including
2 parents and 2 young children— the children's body burdens were in the 200 - 400 ng/g
lwt range, while the parents had blood measurements near 100 ng/g lwt. An important
trend that warrants further investigation is that, even in background adult populations,
there are individuals experiencing very high exposures. This has been seen in studies of
PBDEs in blood as well as indoor dust measurement studies, suggesting that dust
exposures could explain these unusually high exposures.
5-56
-------
4. Intakes have been expressed as a straight-dose basis, ng total PBDE/day, or on a body-
weight basis, ng total PBDE/kg-day. Intake estimates in the literature have tended to
focus more on intake by food than by housedust, although this has changed in recent
years as researchers recognize the importance of indoor dust in the overall exposure
paradigm for PBDEs. Several of the researchers measuring PBDEs in dust also estimated
intakes by soil/dust ingestion using their measurement data, and others have more
directly measured potential hand-to-mouth intakes by sampling the hands of individuals
using alcohol wipes (see Stapleton et al., 2008b). Some estimates of exposure via
housedust were high, up to 400 ng/day, but other estimates were as low as 3 ng/day; the
latter assumed less than 1 nig/day dust ingestion while the former assumed a conservative
100 mg/day dust ingestion rate. Stapleton et al. (2008b) estimated a median adult hand-
to-mouth intake of 154 ng/day (> 2 ng/kg-day), based on hand wipe data in conjunction
with a model on exposure to contaminants on hands. Literature estimates of intakes from
food ingestion were in the range of 0.5 to 2.0 ng/kg-day.
5. Intake estimates derived in this study, based on exposure media concentrations
derived in Chapter 4 combined with average contact rates, arrived at total daily intakes in
the range of 450 to 600 ng/day for children and adults. These intakes were driven by
indoor exposures via soil/dust ingestion, dermal contact with dust, and inhalation of
indoor air; those three pathways accounted for about 83% of total intakes, with food and
water ingestion explaining the remaining 17%. Infant intakes via breast milk exceeded
1,400 ng/day total PBDEs.
6. Using a simple pharmacokinetic (PK) model parameterized with available literature
values, lipid-based concentrations (not specific to blood or milk) were predicted, starting
with the intake values as summarized above. On a total PBDE basis, the prediction was
low at 35.9 ng/g lwt, while it was observed at 36.3 ng/g Iwt in blood and 44.1 ng/g lwt in
milk as central tendency values (geometric mean in blood and median in breast milk)
values in studies selected as representative of the general populat ion. Predictions were
reasonably close to measurements for 7 of 9 congeners. The prediction of BDE 47 at
10.0 ng g lwt did not match the observed measurements of 20,5 ng/g lwt in blood and
5-57
-------
26.0 ng/'g lvvt in milk, and the prediction of 14.6 ng/g lwt of BDE 99 was judged
meaningfully higher than the observed concentrations of 5.0 ng/g lwt for blood and 5.6
ng/g lwt for breast milk. The causes for these discrepancies in the BDEs 47 and 99 are
not known, but it could very easily be the assumed half-lives in humans. At 3.0 years,
BDE 47 was eliminated nearly twice as fast as BDE 99, which was assigned a half-life of
5.4 years. Had BDE 47 been assigned an elimination half-life of 10 years, the prediction
would jump to over 30 ng/g lwt, now twice as high as BDE 99 prediction, and more in
line with measurements. Similarly, if the BDE 99 half-life was half as much as it was,
predictions of BDE 99 in lipid would better match the observed lipid concentrations.
7. While these predictions encouragingly match observations, uncertainties exist in the
exercise, starting from development of dose estimates based on limited environmental
measurements, to indoor contact rates with houscdiist, to the PK parameters of absorption
and elimination half-life. There is also variability in United States body burdens.
Contact rates for food/water ingestion and inhalation are fairly well established, and the
exposure media concentration summaries suggest similarities among different studies in
food and air concentrations. It was assumed that the remainder of the exposures came
from house dust through the pathways of ingestion and dermal contact. Circumstantial
evidence supporting this hypothesis was the high concentrations found in United States
house dust, and other researchers have also identified house dust as a key matrix of
exposure concern for these compounds. The overall weight-of-evidencc of this exercise
supports the finding that the bulk of United States exposures occur in the indoor
environment through contact with house dust. The exercise suggests these exposures
account for between 80 and 90% of total exposures, with the remainder due primarily to
food ingestion.
8. This PK model was used to model infant body burden impacts from consumption of
breast milk. Lipid concentrations were modeled to rise to 200 ng/g lwt through age 5, to
then drop gradually to below 100 ng/g lwt by age 19. However, this result was very
sensitive to assumed elimination half life for total BDEs. When an adult-like half-life of
6 years was used in this infant model, concentrations rose to nearly 200 ng/g lwt by age 1
to continue to rise to 325 ng/g lwt by age 5, to then drop to below 100 ng/g lwt by age 19.
5-58
-------
Only one study was available with which to compare these results. Studying a family of
four - two adults, a child, and a toddler, concentrations in the 18-month old toddler rose
to above 400 ng/g Iwt, and concentrations in the 5 year-old were above 200 ng/g Iwt. At
the same time, the adult concentrations in the family were near 100 ng/g Iwt. While this
study was insufficient to provide verification of either modeling assumption on half-li fe
in infants and children, this ancillary data supports the conclusion that infant and
childhood body burdens appear likely to be significantly higher than that of adults.
REFERENCES
Allen JG, MD McClean, HM Stapleton, TF Webster. 2008. Critical factors in assessing
exposure to PBDEs via house dust. Env, Int., article-in-press. Available May 15, 2008.
Athanasiadou M, SN Cuardra, G Marsh, A Bergman, K Jakobsson. 2008.
Polybrominated diphcnvl ethers (PBDEs) and bioaccumulative hydroxylated PBDE
metabolites in young humans from Managua, Nicaragua. Env. Health Persp 116; 400-
408.
ATSDR. 2004. Public Health Assessment. Saipan Capacitors (a/k/a Tanapag Village
(Saipan)) Tanapag Village, Saipan, Commonwealth of the Northern Marianas Island
EPA Facility ID: MPD982524506. August 31, 2004. Oral bioavailability of
polychlorinated biphenyl (PCB) residues in soil Available at,
http://wwnv.atsdr.cdc.gov/hac/pha/saipan083104-CM/saipan083104-CM-p7.html.
Accessed 7/7/06.
Bakker Ml. R De Winter-Sorkina, A De MuL PE Boon, G Van Donkersgoed, JD Van
Klaveren, BA Bauman, WC Hijman, SPJ Van Leeuwen, J De Boer, MJ Zeilmaker. 2008.
Dietary intake and risk evaluation of polybrominated diphenyl ethers in The Netherlands.
Mol N'utr Food Res. 52: 204-216.
Bocio, L, JM Llobet, J Domingo, J Corbella, A Teixido, and C Casas. 2003.
Polybrominated Diphenyl Ethers (PBDEs) in Foodstuffs; Human Exposure through the
Diet. J. Agric. Food Chcm. 2003(51): 3191-3195.
Bradman A, L Fenster, A Sjodin, RS. Jones, DG Patterson, B Eskenazi, 2007,
Polybrominated diphenyl ether levels in the blood of pregnant women
living in an agricultural community in California. Env. Health Persp. 115: 71-74.
Darnerud PO, GS Eriksen, T Johannesson, PB Larscn, M Viluksela. 2001.
Polybrominated diphenyl ethers: occurrence, dietary exposure, and toxicology. Env
Health Persp 109:49-68.
Dye JA, M Venier, L Zhu, CR Ward, RA Kites, LS Birnbaum. 2007. Elevated PBDE
levels in pet cats: sentinels for humans? Env. Sci and Tech 41: 6350 - 6356.
5-59
-------
EPA, 1997, Exposure Factors Handbook, United States Environmental Protection
Agency, Office of Research and Development, Washington, DC. EPA/600/P-95/002B.
EPA. 2000. Estimated per capita fish consumption in the United States. Report
prepared by the United States Environmental Protection Agency, Office of Water.
EPA. 2003. Exposure and Human Health Reassessment of 2,3,7,8-Tetrachlorodibenzo-p-
Dioxin (TCDD) and Related Compounds. United Stales Environmental Protection
Agency, Office of Research and Development, National Center for Environmental
Assessment, NAS Review Draft. December, 2003. EPA/600/P-00/001 C(a-f). Available
at, http://www.epa.gov/ncea/dioxin.htm.
EPA. 2004. Risk assessment guidance for Superfund. Vol. I. Human health evaluation
manual (part E, supplemental guidance for dermal risk assessment). Office of Superfund
Remediation and Technology Innovation, Washington, DC, EPA/540/R/99/005.
Available online at http://www.epa.gov/superfund/programs/risk/ragse/index.httn
Fabrellas, B, A Martinez, B Ramos, ML Ruiz, 1 Navaroo, A Torre. 2005. Results of an
European survey based on PBDEs analysis in household dust. Organohalogen
Compounds 67: 452- 454
Faldt E, SN Cuadra, M Athanasiadou, A Bergman, K Jakobbson. 2005. Polybrominatcd
diphenyl ethers (PBDEs) in serum from teenagers working in a waste disposal site, and in
women with high consumption offish in Nicaragua. Organohalogen Compounds 67:
502-504.
Fangstrom B, L Hovander, A Bignert, I Athanassiadis, L Linderhom, P Grandjean, P
Wethe, A Bergman, 2005. Concentrations of polybrominatcd diphenyl ethers,
polychlorinated biphenyls, and polychlorobiphenylois in serum from pregnant Faraoese
women and their children 7 years later. Env Sci &Tech 39: 9457-9463.
Fangstrom B, I Athanassiadis, T Odsjo, K Noren, A Bergman . 2008. Temporal trends
ofbrominated flame retardants in milk from Stockholm mothers, 1980-2004. Mol Nutr
Food Res. 2008 Feb;52(2): 187-93.
Fischer D, K. Hooper, M Athanasiadou, I Athanassiadis, A Bergman. 2006. Children
show highest levels of polybrominated diphenyl ethers (PBDEs) in a California family of
four - a case study. Env, Flealth Persp 114: 1581-1585.
Focant J F. A. Sjodin, WE Turner, DG Patterson. 2004. Measurement of selected
polybrominated diphenyl ethers, poybrominated and polychlorinated biphenyls, and
organochlorine pesticides in human serum and milk using comprehensive two-
dimensional gas chromatography isotope dilution time-of-flight mass spectrometry.
Analytical Chemistry 76:6313-6320.
5-60
-------
FSA, 2006. Brominated Chemicals: UK Dietary Estimates. Published by Food
Standards Agency, UK, Information Sheet 10/06. published at,
http://wwvv.food.gov.uk/science/surveillance/fsisbranch2006/fsisi006. Accessed Sep 28,
2006.
Fukata H. M Omori-Inoue, H Osada, E Todaka, N Omori, Y Kawashiro, S Takahashi, H
Takigami, S Sakai. C Mori. 2005. Current status of maternal and fetal exposure to
brominated flame rctardants, PCBs and dioxins in Japan. Organohalogen Compounds
67:1617-1619.
Furst P. 2006. Dioxins, polychlorinated biphenyls and other organohalogen compounds
in human milk. Levels, correlations, trends and exposure through breastfeeding. Mol.
Nutr. Food Res. 50: 922-933.
Gevao B, L Ali, A Al-Omair, M Helaleh. 2005. Non-dietary human exposure to
polybrominated diphenyl ethers in Kuwait. Organohalogen Compounds, Volume 67:
1526-1.529.
Geycr HJ, KVV Schramm, PO Darnerud. et al. 2004. Terminal elimination half-lives of
the brominated flame retardants TBBPA, IIBCD. and lower brominated PBDEs in
humans. Organohalogen Compounds 66: 3820-3825.
Gomara R, L Herrero, JJ Ramos, JR Mateo, MA Fernandez, JF Garcia, MJ Gonzalez.
2007. Distribution of polybrominated diphenyl ethers in human umbilical cord serum,
paternal serum, maternal serum, placentas, and breast milk from Madrid population,
Spain. Env. Sci and Tech 41: 6961-6968.
Gomara B, L Herrero, MJ Gonzalez. 2006. Survey of polybrominated diphenyl ether
levels in Spanish commercial foodstuffs. Env. Sci. and Tech. 40: 7541-7547.
Guvenuis DM, A Aronsson, G Ekman-Ordeberg, A Bergman, K. Noren. 2003. Human
prenatal and postnatal exposure to polybrominated diphenyl ethers, polychlorinated
biphenyls, polychlorobiphenylols, and pentachlorophenol. Env. Health Persp. Ill: 1235-
1241.
Harrad S, C Ibarra, M Diamond, L Melymuk, M Robson, J Douwes, L Roosens, AC
Dirtu, A Covaci. 2008. Polybrominated diphenyl ethers in domestic indoor dust from
Canada, New Zealand, United Kingdom and United States. Env. Int 34: 232-238,
Harrad S, S Hazrati, C Ibarra. 2006. Concentrations of polychlorinated biphenyls in
indoor air and polybrominated diphenyl ethers in indoor air and dust in Birmingham,
United Kingdom: implications for human exposure. Env. Sci and Tech. 40: 4633-4638.
Harrad S, L Porter. 2007. Concentrations of polybrominated diphenyl ethers in blood
serum from New Zealand. Chemosphere 66:2019-2023.
5-61
-------
Harrad S, R Wijcsckera, S Hunger, C Halliweil, R Baker, 2004. Preliminary assessment
of U.K. human dietary and inhalation exposure to polybrominatcd diphenyl ethers. Env,
Sci. and Tech. 38: 2345-2350.
Hawley, JK. 1985. Assessment of health risk from exposure to eontaminated soil. Risk
Anal. 5:289-302.
Hazrati S, S Harrad 2005. Implications of passive sampling derived concentrations of
airborne PCBs and PBDEs in urban indoor microenviromnents. Organohalogen
Compounds 66: 1033-1036.
Hughes MF, BC Edwards, CT Mitchell, B Bhooshan. 2001. In vitro dermal absorption of
flame rctardant chemicals. Food and Chem. Tox. 39:1263-1270,
Huwe JK, H Hakk, DJ Smith, JT Diliberto, V Richardson, HM Stapleton, LS Bimbaum.
2008, Comparative absorption and bioaccumulation of polybrominated diphenyl ethers
following ingestion via dust and oil in male rats. Env. Sci Tech. 42: 2694-2700.
lnoue K, K Harada, K Takenaka, S Uehara, M Kono, T Shimizu, T tadasuga, K
Senthilkumar, F Yamashita, A Koizumi. 2006. Levels and concentration ratios of
polychlorinated biphenyls and polybrominated diphenyl ethers in serum and breast milk
in Japanese mothers. Env. Health Persp 114: 1179-1185.
Jakobsson K, M Athanasiadou, A Christiansson, A Bergman, L Hagmar. 2005.
Polybrominated diphenyl ethers (PBDEs) in serum from Swedish men 1988 - 2002. A
longitudinal study. Organohalogen Compounds 67: 533-536.
Jakobsson K, KThuresson, P Hoglund, A Sjodin, L Hagmar, A Bergman. 2003. A
summary of exposures to polybrominated diphenyl ethers (PBDEs) in Swedish workers,
and determination of half-lives of PBDEs. Organohalogen Compounds 60-65, Dioxin
2003, Boston, MA.
Jakobsson K, K Thuresson, L Rylander, A Sjodin, L Hagmar, A Bergmen. 2002.
Exposure to polybrominated diphenyl ethers and tctrabromobisphenol A among computer
technicians. Chemosphere 46: 709-716.
Johnson-Restrepo B, K Kanna, DP Rapaport, BD Rodan. 2005. Polybrominated
diphenyl ethers and polychlorinated biphenyls in human adipose tissue from New York.
Env Sci and Tech. 39: 5177-5182.
Jones-Otazo, HA, JP Clarke, ML Diamond, JA Archbold, G Ferguson, T Harrier, GM
Richardson, JJ Ryan, B Wilford. 2005 Is house dust the missing exposure pathway for
PBDEs? An analysis of the urban fate and human exposure to PBDEs. Env. Sci. and
Tech .39: 5121-5130.
5-62
-------
Kalantzi O, RF Brown, C Erdmann, M Caleffi, R Goth-Goldstein, M Petrcas. 2005.
Polybrominatcd diphcnyl ethers (PBDEs) in human breast adipose tissue samples from
Brazil. Organohalogen Compounds 67: 479-481.
Karlsson M, A Julander, B van Bavel, L Hardell. 2007. Levels of brominated flame
retardants in blood in relation to levels in household and dust. Env. Int. 33: 62-69.
Knutsen H, C Bergsten, C Thomsen, A Slctta, G Becher, J Alexander, HM Meltzer.
2005. Preliminary assessment of PBDE exposure from food in Norway. Organohalogen
Compounds 67: 1624 - 1627.
Kotz A, R Malish, K Kype, M Oehme. 2005. PBDE, PBDD/F and mixed chlorinated-
brominated PXDD/F in pooled human milk samples from different countries.
Organohalogen Compounds 67: 1540-1544
Krcuzer PE, GA Canady, C Baur, W Kessler, O Papke, H Greim, JG Filser. 1997.
2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) and congeners in infants. A toxicokinetics
model of human lifetime body burden by TCDD with special emphasis on its uptake and
nutrition. Archives of Toxicology 71:383-400.
Lee SJ, MG Ikonomou, II Park, SY Back, YS Chang. 2007. Polybrominated diphenyl
ethers in blood from Korean incinerator workers and general population. Chemosphere
67: 489-497.
Lorber M, and L Phillips. 2002. Infant exposure to dioxin-like compounds in breast
milk. Env Health Persp 110: A325-A332.
Lunder S, R Sharp. 2004. Mothers' Milk Record Levels of Toxic Fire Retardants Found
in American mothers' breast milk. Environmental Working Group, http://www.ewg.org
Luksemburg W, R Wenning, M Maier, A Patterson, S Braithwaite. 2004.
Polybrominated diphenyl ethers (PBDE) and polychlorinated dibenzo-p-dioxins
(PCDD/F) and biphenyls (PCB) in fish, beef, and fowl purchased in food markets in
Northern California U.S.A. Organohalogen Compounds 66: 3932 - 3937.
Mandalakis M, EG Stephanou, Y Horii, K Kannan. 2008. Emerging contaminants in car
interiors: evaluating the impact of airborne PBDEs and PBDD/Fs. Env. Sci and Tech.
Advance on-line, published on-line 3/26/08.
Mazdi An, NG Dodder, MP Abernathy, RA Hites, RM Bigsby. 2003. Polybrominated
diphenyl ethers in maternal and fetal blood samples. Env Health Persp. 111:1249-1252.
McDonald TA. 2005. Polybrominated diphenyl ether levels among United States
residents: daily intake and risk of harm to the developing brain and reproductive organs.
Int Env. Ass. & Mgmt 1 (4):343-354.
5-63
-------
Meng, X-Z, EY Zcng, LPYu, Y Guo, BX Mai. 2007. Assessment of human exposure to
polybrominated diphenyl ethers in China via fish consumption and inhalation. Env, Sci
and Tech 41: 4882-4887.
Mori and, KE3, PJ Landrigan, A Sjodin, AK Gobeille, RS Jones, EE MeGahee, LL
Needham, DG Patterson. 2005. Body Burdens of Polybrominated Diphenyl Ethers
among Urgan Anglers. Em Health Persp 113:1689-1692.
Naert C\ M Piette, N Bruneel, CV Pcteghem. 2006. Occurrence of polychlorinated
biphenyls and polybrominated diphenyl ethers in Belgian human adipose tissue samples.
Arch Environ. Contain. Toxicol 50: 290-296.
NEW. 2004. Northwest Environment Watch. Flame Retardants in the bodies of Pacific
Northwest Residents. Report released Sep 29, 2004. available at,
http://wvvw.northvvestwatch.org.
Ohta S, D Ishizuka, H Nishimura, T Nakao, O Aozasa, Y Shimidzu, F Ochiai, T Kida, M
Nishi, H Miyata. 2002. Comparison of polybrominated diphenyl ethers in fish,
vegetables, and meats and levels in human milk of nursing women in Japan.
Chemosphere 46:689-696
Pausteobach, DJ, BL Finley, TF Long. 1997. The critical role of house dust in
understanding the hazards posed by contaminated soils. Tnt J. Tox 16: 339-362.
Paustenbaeh DJ, K Fehling, P Scott, M Harris, and B. Kerger, 2006. Identifying soil
cleanup criteria for dioxins in urban residential soils: how have 20 years of research and
risk assessment experience affected the analysis? J. Tox. and Env. Health, Part B, 9: 87-
145.
Petreas M, J She, FR Brown, J Winkler, G Windham, E Rogers, G Zhao, R Bhatia, MJ
Charles. 2003. High body burdens of 2,2',4,4'-tetrabromodiphenyl ether (BDE-47) in
California women. Env Health Persp. Ill: 1175-1179.
Qu W, X Bi, G Sheng, S Lu, J Fu, J Yuan, L Li. 2007. Exposure to polybrominated
diphenyl ethers among workers at an electronic waste dismantling region in Guangdong,
China. 2007. Env Internal 33; 1029 - 1034.
Roper CS. AG Simpson, S Madden, TL Serex, JA Biesemeier. 2006. Absorption of
(C 4)-tetrabromodipheny! ether (TeBDE) through human and rat skin in vitro. Drug and
Chem. Tox. 29:289-301.
Ryan JJ, J van Oostdam. 2004. Polybrominated diphenyl ethers (PBDEs) in maternal
and cord blood plasma of several Northern Canadian populations. Organohalogen
Compounds 66: 2579-2585.
5-64
-------
Ryan J J, BC Wainman, A Schecter, J Moiscy, I Kosarae, WF Sun. 2006. Trends of the
brominated flame rcatardants, PBDEs and IIBCD, in human milk samples from North
America. Organohalogen Compounds 68: 778-781.
Schecter A, TR Harris, O Papke, KC Tung, A Musumba. 2006a. Polybrominatcd
diphenyl ether (PBDE) levels in the blood of pure vegetarians (vegans). Tox and Env.
Chem 88: 107-112.
Schecter A, O Papke, TR Harris, J Olson, KC Tung, A Musumba, L Birnbaum, 2006b.
Polybrominatcd Diphenyl Ether (PBDE) levels in an expanded market basket survey of
U.S. food and estimated PBDE dietary intake by age and sex. Env Health Persp
114:1515-1520.
Schecter A, S Johnson-Welch, KC Tung, TR Harris, O Papke, R Rosen. 2007.
Polybrominated diphenyl ether (PBDE) levels in livers of U.S. human fetuses and
newborns. J Tox and Env Health Part A, 70( 1): 1-6
Schecter A, O Papke, KC Tung, J Joseph. TR Hanis, J Dahlgren. 2005. Polybrominated
Diphenyl Ether Flame Retardants in the U.S. Population: Current Levels, Temporal
trends, and Comparison with Dioxins, Dibenzofurans, and Polychlorinated Biphenyls. J.
Occup. Env. Medicine 47:199-211.
Schecter A, M Pavuk, O Papke, JJ Ryan, L Birnbaum, R Rosen. 2003. Polybrominated
diphenyl ethers (PBDEs) in U.S. mother's milk. Env Health Persp 111:1723-1729.
Sharp, R, S. Lunder. 2004. In the Dust. Toxic Fire Retardants in American Homes.
Published by Environmental Working Group, report posted at
http://www.ewg.org/reports/inthedust/index.php.
She J, A Holden, M Shart, CW Deny. K Hooper. 2005. PBDE Congener Pattern as an
indicator of human exposure pathway in North America. Organohalogen Compounds 67:
525-528.
Schuhmachcr M, H Kiviranta, T Vartiainen, JL Domingo. 2007. Concentrations of
polychlorinated biphenyls (PCBs) and polybrominated diphenyl ethers (PBDEs) in milk
of women from Catalonia, Spain, Chemosphere 67: S295-S300,
Sjodin A, L Hagmar, E Klasson-Wehler, J Bjork A Bergman. 2000. Influence of
consumption of fatty Baltic Sea fish on plasma levels of halogenated environmental
contaminants in Latvian and Swedish men. Env. Health Perspeet 108: 1035-1041.
Sjodin A, DG Patterson, A Bergman. 2001. Brominated flame retardants in serum from
U.S. Blood Donors. Env. Sci and Tech. 35: 3830-3833
Sjodin A, RS Jones, JF Focant, C Lapeza, RY Want, EE McGahee, Y Ahang, WE
Turner, B Slazyk, LL Needham, DG Patterson. 2004a. Retrospective time-trend study of
5-65
-------
polybrominated diphenyl ether and polybrominated and polychlorinatcd biphenyl levels
in human serum from the United States. Env Health Pcrsp 112: 654-658.
Sjodin A, O Papke, E McGahee, R Jones, JF Foeant, T Pless-Mulloli, LM Toms, R
Wang, Y Zhang, L Needham, T Heimrmann, D Patterson. 2004b. Concentration of
polybrominated diphenyl ethers (PBDEs) in house hold dust from various countries -
inhalation a potential route of human exposure. Organohalogen Compounds 66: 3770-
3775
Sjodin A, JS LaKind, DG Patterson, LL Needham, R Wang, IM Paul, J Stokes, CM
Berlin. 2005. Current concentrations and changes in concentrations of PBDEs,
persistent pesticides, and PCBs in human milk. Organohalogen Compounds 67:1745-
1748.
Sjodin, A, L Wong, R Jones, A Park, Y Ahang, C Hodge, E DePietro, C McClure, W
Turner, L Needham. D Patterson. 2008. Scrum concentrations of polybrominated
diphenyl ethers (PBDEs) and polybrominated biphenyl (PBB) in the United States
Population: 2003-2004. Env. Sci. Tcclmol. 42: 1377-1384.
Smith AII. 1987. Infant exposure assessment for breast milk dioxins and furans derived
from waste incineration emissions. Risk Analysis 7(3):347-353.
Stapleton li\l. NG Dodder. JH Offenbcrt, MM Schantz, and SA Wise. 2005
Polybrominated diphenyl ethers in house dust and clothes dryer lint. Env. Sci. and Tech.
39: 925-931.
Stapleton HM, A Sjodin, RS Jones, S Niehuser, Y Zhang, DG Patterson. 2008a. Serum
level of polybrominated diphenyl ethers (PBDEs) in foam recyclers and carpet installers
working in the United States. Env. Sci Technol 42: 3453-3458.
Stapleton HM, SM Kelly, iG Allen, MD McClean, TF Webster. 2008b. Measurement of
polybrominated diphenyl ethers on hand wipes: estimating exposure from hand-to-mouth
contact. Env. Sci Tech 42: 3329-3334.
Staskal DF, J J Diliberto, MJ DeVito, and LS Birnbaum. 2005. Toxicokinetics of BDE
47 in female mice: effect of dose, route of exposure, and time. Tox, Sci. 83: 213-223.
Staskal DF, H Hakk, D Bauer. JJ Diliberto, LS Birnbaum. 2006. Toxicokinetics of
polybrominated diphenyl ether congeners 47, 99, 100, and 153 in mice. Tox. Sci. 94:28-
37.
Thomas GO, M Wilkinson, S Hodson, KC Jones. 2006, Organohalogen chemicals in
human blood from the United Kingdon. Env. Poll. 141: 30-41.
Thomsen C, V Liane, M Froshaug, G Becher. 2005a. Levels of brominated flame
retardants in human samples from Norway through three decades. Organohalogen
Compounds 67: 658-661.
5-66
-------
Thomson C, M Froshaug, SL Broadwcll, G Bcchcr, M Eggesbo. 2005b. Levels of
brominated flame retardants in milk from the Norweigan human milk study: HUMIS.
Organohalogen Compounds 67: 509-512.
Thuresson K, P Hoglund, L Hagmar, A Sjodin, A Bergman, and K. Jakobsson. 2006.
Apparent half-lives of hcpta- to decabrominated diphenyl ethers in human serum as
determined in oecupationally exposed workers. Env Health Persp 114: 176-181.
Toms LL, FA Harden, RK Symons, D Burniston, P Furst, JF Muller. 2007.
Polybrominated diphenyl ethers (PBDEs) in human milk from Australia. Chemosphere
68: 797-803.
USDA. 1995. Food and nutrient intakes by individuals in the United States, 1 day, 1989
- 1991. United States Department of Agriculture, Agricultural Research Service,
Washington, DC. NFS Report No. 91-2.
van Ravenzwaay B, E Leibold. 2004. A comparison between in vitro rat and human and
in vivo rat skin absorption studies. Hum. Exp. Toxicol. 23, 421-430.
Voorspocls S, A C'ovaci, H Neels, P. Schepens. 2007. Dietary PBDE intake: A market-
basket study in Belgium. Env. hiter. 33: 93-97.
Webster T, V Vieira, A Schccter. 2005. Estimating human exposure to PBDE-47 via air,
food, and dust using Monte Carlo methods. Organohalogen Compounds 67: 505-508.
Wester RC, D Quan, HI Maibach. 1996 In vitro percutaneous absorption of model
compounds glyphosate and malathion from cotton fabric into and through human skin.
Food and Chem Tox 34: 731 -735.
Wilford BF1, T ITarner, J Zhu, M Shoeib, KC Jones. 2004. Passive sampling survey of
polybrominated diphenyl ether flame retardants in indoor and outdoor air in Ottowa,
Canada: Implications for sources and exposure. Env Sci and Tech 38:5312-5318
Wolff MS, SL Tcitelbaum, PJ Lioyet al. 2005. Exposures among pregnant women near
the World Trade Center site on 9/11. Env Health Perspect 113: 739-748.
Wu N, T Hermann, O Paepke, J Tickner, R Hale, E Harvey, M La Guardia, MD
McCLean, T Webstor. 2007. Human exposure to PBDEs: associations of PBDE body
burdens with food consumption and house dust concentrations. Env. Sci. and Tech. 41:
1584-1589.
5-67
-------
Table 5-1. Blood concentrations ofPBDE congeners in the United States
Congener
Concentration,
ng/g hvt
Comment
Citation
TriBDE
17
0.3
N-l: Pool of 15 individuals from
Philadelphia. Memphis, Miami
Focantet al. 2004
0 1 ft 03, 0.05
3 2003 results: I pooled serum (n-100) from
1 X, 1 pooled whole blood (tr-100) from TX;
mean of 39 individuals 29 MS, 10 NY
Schecter et al,, 2005
0.013
N=8: mean from vegans (ND = 0)
Sehccter et al., 2006a
0.003
N=11; mean from blood/milk study (ND=0)
Schecter et al,, 2006c
28
1.2
N=2040; NHANES 2003/4; geometric mean
Sjodin et al.. 2008
0.8
N=l: Pool of 15 individuals from
Philadelphia, Memphis. Miami
Focantet al. 2004
1.3, 1.9, 1,1
3 2003 results: 1 pooled serum (n 100) from
I X. ] pooled whole blood (n=100) from TX;
mean of 39 individuals - 29 MS. 10 NY
Schecter et al., 2005
0.65
N = 100; median serum level in 100 mothers
who were pregnant and near WTC on 9/11
Wolffet alet al„ 2005
0.92
N=8: mean from vegans (ND = 0)
Schecter et al., 2006a
0.58
N=11; mean from blood/milk study (NT>=0)
Schecter et al., 2006c
TetrnBDE
47
20.5
N=2040; NHANES 2003/4; geometric mean
Sjodin et al., 2008
0.6 (<0.4-23.8)
Median, range; 6 of 12 blood donor samples
collected in 111 in 1988 quantified
Sjodin et al., 2001
13.2; .7-1388.6
Geometric mean; range of 93 urban angler
samples in NY and NJ; 93% detected
Morlandet alet al,
2005
28.1
N=I; Pool of 15 individuals from
Philadelphia, Memphis, Miami
Focantet al. 2004
34 (29-98)
Median from 2000/2 (range) in trend study
from pooled blood from around U.S.
Sjodin et al.. 2004
28; 9.1-310
Median; range from maternal/cord blood
(n=T2): 2001; cord blood was identical
Mazdai et al., 2003
50.6 (<10-511)
Mean, range from Laotian reproductive age
women in San Francisco area, 97-99
Petreas et al. 2003
32.5, 44.2. 25.0
3 2003 results: 1 pooled serum (n^lOO) from
TX, 1 pooled whole blood (ti"T00) from TX:
mean of 39 individuals 29 MS, 10 NY
Schecter et al., 2005
0.6 (<0.4-23.8)
Median, range (n=12) from individual donors
collected in 1988 in Illinois
Sjodin et al., 2001
23 60: 94 -
137;136 - 245
Range from 2 parents, 1 5 year-old, and 1 18-
month old from case study in CA in 2004
Fischer et al., 2006
9.7
N 100; median serum level in 100 mothers
who were pregnant and near WTC on 9/11
Wolffet alet al„ 2005
22.6
N=8; mean from vegans (ND - 0)
Schecter et al., 2006a
13.1
N= 11: mean from blood/milk study (ND=0)
Schecter et al.. 2006c
11; 205
N=24: median and max from a cohort of
pregnant Latina women in CA
Bradman et al., 2006
66
0.3
N= 1; Pool of 15 individuals from
Philadelphia, Memphis. Miami
Focantet al. 2004
5-68
-------
0.3, NA. 0.4
3 2003 results: 1 pooled serum (iv-lOO) from
! X. 1 pooled whole blood (ir=l00) from TX:
mean of 39 individuals - 29 MS. 10 NY
Schecter et aL 2005
ND
N---8; mean from vegans (ND = 0)
Scheeter et al., 2006a
0.1
N=11; mean from blood/milk study (ND=4))
Schecter et al., 2006c
77
ND, NA. 0.01
3 2003 results: 1 pooled serum ui iOt>> from
TX. 1 pooled whole blood (n = 100) from TX;
mean of 39 individuals 29 MS. 10 NY
Schecter et al.. 2005
PcntaBDE
85
1.0; 0.2-109.1
Geometric mean; range of 92 urban angler
samples in NY and NJ; 27% delected
Morlandct alet al.,
2005
0.8
N= 1; Pool of 15 individuals from
Philadelphia, Memphis, Miami
Focantet al. 2004
0.7(0.5-1.4)
Median from 2000/2 (range) in trend study
from pooled blood from around U.S.
Sjodin et al., 2004
NA, 1.1, 1.2
3 2003 results: 1 pooled serum (n=l00) from
TX, 1 pooled whole blood (n=100) from TX;
mean of 39 individuals 29 MS. 10 NY
Schecter et al., 2005
0.4
N=8: mean from vegans (ND = 0)
Schecter et al.. 2006a
0.3
N=11; mean from blood/milk study (ND=0)
Schecter ct al 2006c
0.3; 5.0
N=24; median and maximum from a cohort of
pregnant Latina women in CA
Bradman et al., -iuuo
99
5,0
N-2040; NHANES 2003/4; geometric mean
Sjodin et al.. 2008
0.3; <0.2-3.7
Median, range; 8 of 12 U.S. blood donor
samples collected in III in 1988 quantified;
Sjodin et al., 2001
3.2; 0.3-545.5
Geometric mean; range of 93 urban angler
samples in NY and NJ; 66% detected
Morlandct alet al..
2005
9.2
N-1; Pool of 15 individuals from
Philadelphia. Memphis, Miami
Focantet al. 2004
11 (6.8-26)
Median from 2000/2 (range) in trend study
from pooled blood from around U.S.
Sjodin et al,, 2004
5.7; 2.4-68
Median; range from maternal/cord blood
-------
5.7.5.2.4.7
3 2003 results: 1 pooled serum (n-100) from
TX, 1 pooled whole blood (n-11) from Tx;
mean of 39 individuals - 28 MS. 10 NY
Schecter et al.. 2005
8-22; 30-39; 57-
87
Range from 2 parents, 1 5 year-old. and 1 18-
month old from case study in C'A in 2004
Fischer et al.. 2006
1.8
N-100; median serum level in 100 mothers
who were pregnant and near WTC on 9/11
Wolff et al.. 2005
5.7
N=8; mean from vegans > (M
Schecter et al., 2006a
3.7
1; mean from blood/milk study (ND=0)
Schecter et al., 2006c
1.8; 44.0
N=24; median and maximum from a cohort of
pregnant Latina women in CA
Bradman et al.. 2006
HexaBDE
138
NA, 0.3, 0.2
3 2003 results: 1 pooled serum (n=100) from
TX. 1 pooled whole blood (n=100) from TX;
mean of 39 individuals 29 MS. 10 NY
Schecter et al., 2005
0.07
N=8; mean from vegans (ND = 0)
Schecter et al., 2006a
0.04
N=11; mean from blood'milk study (ND=0)
Schecter et al., 2006c
153
5.7
N=2040; NHANES 2003/4; geometric mean
Sjodin et al.. 2008
0.3; 0.1-2.0
Median, range; 12 of 12 U.S. blood donor
samples collected in HI in 1988 quantified
Sjodin et al., 2001
3.2; 0.4-165.2
Geometric mean; range of 93 urban angler
samplers in NY and NJ; 96% detected
Morlandet alet al.,
2005
6.2
Pool of 15 individuals from Philadelphia,
Memphis, Miami
Focantet al. 2004
7.3 (1.8-17)
Median from 2000/2 (range) in trend study
from pooled blood from around U.S.
Sjodin et al.. 2004
2.9; 1 83
Median; range from maternal/cord blood
in 1 2>: 2001; cord blood was identical
Mazdai et al., 2003
12.3. 11.7. 5.7
3 2003 results: 1 pooled serum in ](»>) from
TX, 1 pooled whole blood (n-100) from TX;
mean of 39 individuals 29 MS, 10 NY
Schecter et a!.. 2005
19 42; 49 65;
75 141
Range from 2 parents, 1 5 year-old, and 1 18-
month old from case study in CA in 2004
Fischer et al., 2006
1.8
N 100; median serum level in 100 mothers
w ho were pregnant and near WTC on 9/11
Woltfet alet al., 2005
14.6
N-8; mean from vegans (ND = 0)
Schecter et al., 2006a
8.9
N=11; mean from blood'milk study (ND=0)
Schecter et al.. 2006c
1.5; 35
N--24; median and maximum from a cohort of
24 pregnant Latina women in C'A
Bradman et al., 2006
154
0.6; 0.09-24.7
Geometric mean; range of 89 urban angler
samples in NY and NJ; 25% detected
Morland et al., 2005
1.2
N 1; pool of 15 individuals trom
Philadelphia. Memphis, Miami
Focant et al. 2004
0.95 (0.5-1.8)
Median from 2000/2 (range) in trend study
from pooled blood from around LJ.S.
Sjodin et al.. 2004
0.3 (ND-6.1)
Median, range from maternal/cord blood
(n=l 2); 2001; cord blood was identical
Mazdai et al., 2003
0.8.0.8, 1.0
3 2003 results: 1 pooled serum (nH00) from
TX. 1 pooled whole blood (n=U)0) from TX;
mean of 39 individuals 29 MS, 10 NY'
Schecter et al., 2005
0.6-6; 3-7; 4-17
Range from 2 parents, 1 5 year-old, and 1 18
month-old from case studv in CA in 2004
Fischer et al.. 2006
0.6
N"8; mean from vegans (ND--0)
Schecter et al., 2006a
0.4
N=11; mean from blood/milk study (ND--0)
Schecter et al.. 2006c
0.3; 4.2
N=24; median and maximum from a cohort of | Bradman et al., 2006
5-70
-------
| 24 pregnant Latina women in C"A
HeptaBDE
183
0.2:0.1-1.3
Median, range; 12 of 12 U.S. blood donor
samples collected in HI in 1988 quantified
Sjodin ct al, 2001
0.5; 0.1-2.0
Geometric mean; range of 93 urban angler
samplers in NY and NJ; 29% detected
Morlandet alet al.,
2005
\|v 0-2.7
Median; range from maternal/cord blood
{n= 12); 2001; cord blood was identical
Mazdai et al., 2003
0.3,0 4 0 4
3 2003 results; 1 pooled serum (n-100) from
TX. 1 pooled whole blood (n=100) fromTX;
mean of 39 individuals 29 MS. 10 NY
Schecter et al.. 2005
0.04
N=8; mean from vegans (ND = 0)
Schccter et al., 2006a
0.05
N=11; mean from blood/milk study (ND=0)
Schecter et al., 2006c
NonaBDE
203
SDfU.I <0.1-
0.2
Median, range; 5 of 12 U.S. blood donor
samples collected in 111 in 1988 quantified
Sjodin et al., 2001
DccaBDE
209
ND(LO); <1.0-
33.6
Median, range; 5 of 12 U.S. blood donor
samples collected in 111 in 1988 quantified
Sjodin ct al., 2001
NA, 1.4,2.7
3 2003 results: 1 pooled serum (n=100) from
TX, 1 pooled whole blood (n=100) from TX;
mean of 39 individuals 29 MS. 10 NY
Schecter et al., 2005
2 - 23; 9 143;
19 233
Range from 2 parents. 1 5 year-old, and 1 18-
moiith old from case study in CA in 2004
Fischer et al., 2006
ND
N=8; mean from vegans (ND = 0)
Schecter et al., 2006a
3.7
N-=l 1; mean from blood/milk study i \D n i
Schecter et al.. 2006c
5-71
-------
Table 5.2. Breast milk concentrations of PBDE congeners in the United States
Congener
Concentration,
ng/g hvt
Comment
Citation
MonoBDE
3
0.15
Mean from 20 primaparae women from
around U.S..
[.under & Sharp,
2004
BiBDE
7
0.01
Mean from 20 primaparae women from
around U.S..
Lunder & Sharp.
2004
8 h- 11
0.02
Mean from 20 primaparae women from
around U.S..
Lunder & Sharp,
2004
12
0.01
Mean from 20 primaparae women from
around U.S..
Lunder & Sharp,
2004
13
0.01
Mean from 20 primaparae women from
around U.S..
Lunder & Sharp,
2004
15
1.46
Mean from 20 primaparae women from
around U.S..
Lunder & Sharp,
2004
1.5
WHO inter-lab calibration study with results
from 17 of 24 countries; U.S. listed here
Kotz et al., 2005
TriBDE
17
0.1
Mean from 20 primaparae women from
around U.S..
Lunder & Sharp.
2004
0.02 (0.01)
Mean (median), 47 women in Texas 2002
Schecter et al. 2003
0.02
N= 11; mean from blood/milk study (ND=0)
Scheeter et al.. 2006c
DL (0.01)
N=46; median from 46 women in the Boston,
MA area
Wu et al.. 2007
25
0.02
Mean from 20 primaparae women from
around U.S..
Lunder & Sharp.
2004
28
7.4
Mean of 3 pooled sample analyses from CO,
C'A, and NC
Focant et al. 2004
8.6
WHO inter-lab calibration study with results
from 17 of 24 countries; U.S. listed here
Kotz et al., 2005
2.4(1.2)
Mean (median). 47 women in Texas 2002
Schecter et al.. 2003
t.l
N=11; mean from blood/milk study (ND=0)
Schecter et al.. 2006c
1.80. 0.47
N=38; mean, median from 23 towns in MA
Johnson-Restrepo et
al. 2007
0.93
\ 4(<: median from 46 women in the Boston,
MA area
Wu et al., 2007
28 + 33
6.2
Mean from 20 primaparae women from
around U.S..
Lunder & Sharp,
2004
28/32
3.8
Mean of 40 samples from primeparae women
from Pacific Northwest
NEW. 2004
32
0.2
Mean of 40 samples from primeparae women
from Pacific Northwest
NEW, 2004
35
0.02
Mean from 20 primaparae women from
around U.S..
Lunder & Sharp,
2004
37
0.06
Mean from 20 primaparae women from
around U.S..
Lunder & Sharp,
2004
TetraBDE
47
193.0
Mean of 3 pooled sample analyses from CO,
C'A, and NC
Focant et al., 2004
5-72
-------
84.9
Mean from 20 primaparae women from
around U.S..
Lunder & Sharp,
2004
50; 26
Mean; median of 40 samples from priineparae
women from Pacific Northwest
NEW. 2004
233.9
WHO inter-lab calibration study with results
from 17 of 24 countries; U.S. listed here
Kotz et al„ 2005
35
Partial data from longitudinal study in
Pennsylvania; 3 women
Sjoditi et al., 2005
40.8; 18.4
Mean, median, 47 women in Texas 2002
Schecter et al., 2003
17.1
N=11; mean from blood/milk study (ND=0)
Schecter et al. 2006c
40.7, 7.7
N=38: mean, median from 23 towns in MA
Johnson-Restrepo et
al. 2007
13.9
Nv=46; median from 46 women in the Boston.
MA area
Wu et al.. 2007
49
0.6
Mean from 20 primaparae women from
around U.S..
Lunder & Sharp,
2004
66
1.1
Mean from 20 primaparae women from
around U.S..
Lunder & Sharp,
2004
0.3
Mean of 40 samples from priineparae women
from Pacific Northwest
NEW. 2004
0.65 (0.14)
Mean (median), 47 women in Texas 2002
Schecter et al.. 2003
0.07
N= 1 1; mean from blood/milk study (ND=0)
Schecter et al., 2006c
1.07. <0.84
N =38; mean, median from 23 towns in MA
Johnson-Restrepo et
al. 2007
0.12
N =46; median from 46 women in the Boston,
MA area
Wu et al., 2007
71
0.05
Mean from 20 primaparae women from
around U.S..
Lunder & Sharp,
2004
0.5. 0.2
Mean, median of 40 samples from priineparae
women from Pacific Northwest
NEW, 2004
75
0.07
Mean from 20 primaparae women from
around U.S..
Lunder & Sharp,
2004
77
0.01 (NA)
Mean (median), 47 women in Texas 2002
Schecter et al. (2003)
0.02, <0.84
N=38; mean, median from 23 towns in MA
Johnson-Restrepo et
al. 2007
PentaBDE
85
5.5
Mean of 3 pooled sample analyses from CO,
CA. and NC
Focant et al., 2004
2.3
Mean from 20 primaparae women from
around U.S..
Lunder & Sharp.
2004
1.1,0.6
Mean, median of 40 samples from priineparae
women from Pacific Northwest
NEW, 2004
1.15, 0.41
Mean, median, 47 women in Texas 2002
Schecter et al (2003)
0.4
N=11; mean from blood/milk study (M) ot
Schecter et al., 2006c
1.02, <0.46
N=38; mean, median from 23 towns in MA
Johnson-Restrepo et
al. 2007
0.26
N= 46; median from 46 women in the Boston,
M A area
Wu et al., 2007
5-73
-------
99
55.0
Mean of 3 pooled sample analyses from CO,
CA, and NC
Focant et al., 2004
20.9
Mean from 20 primaparae women from
around U.S..
Lunder & Sharp,
2004
10. 5.4
Mean, median of 40 samples from primeparae
women from Pacific Northwest
NEW. 2004
58.7
WHO inter-lab calibration study with results
from 17 of 24 countries; U.S. listed here
Kotz et al.. 2005
8
Partial data from longitudinal study in
Pennsylvania: 3 women
Sjodin et a!., 2005
14.0,5.7
Mean, median, 47 women in Texas 2002
Scheeter et al. (2003)
3.5
*N=11; mean from blood/milk study (ND=0)
Schecter et al,. 2006c
11.8. 1.5
N-=38: mean, median from 23 towns in MA
Johnson-Restrepo et
al. 2007
2.42
N=46; median from 46 women in the Boston.
MA area
Wu et al., 2007
100
34.4
Mean of 3 pooled sample analyses from CO,
CA, and NC
Focant et al., 2004
18,4
Mean from 20 primaparae women from
around U.S..
Lunder & Sharp.
2004
12.5.2
Mean, median of 40 samples from primeparae
women from Pacific Northwest
NEW, 2004
39.2
WHO inter-lab calibration study with results
from 17 of 24 countries; U.S. listed here
Kotz et al., 2005
8.2, 2.9
Mean, median, 47 women in Texas 2002
Schecter et al„ 2003
4.4
N=11; mean from blood/milk study (ND-0)
Schecter et al., 2006c
6.9, <0.46
N=38: mean, median from 23 towns in MA
Johnson-Restrepo et
al. 2007
2.40
N= 46; median from 46 women in the Boston.
M A area
Wu et al.. 2007
118
0.2, <0.46
N=38; mean, median from 23 towns in MA
Johnson-Restrepo et
al. 2007
119
0.06
Mean from 20 primaparae women from
around U.S..
Lunder & Sharp,
2004
126
0.04
Mean from 20 primaparae women from
around U.S..
Lunder & Sharp,
2004
HexaBDE
138
0.60. 0.09
Mean (median), 47 women in Texas 2002
Scheeter et al. (2003)
0.04
N=11; mean from blood/milk study (ND=0)
Schecter et al.. 2006c
0.16. <0.03
N=38: mean, median from 23 towns in MA
Johnson-Restrepo et
al. 2007
0.03
N=46; median from 46 women in the Boston,
MA area
Wu et al., 2007
138-1-166
0.26
Mean from 20 primaparae women from
around U.S..
Lunder & Sharp,
2004
140
0.21
Mean from 20 primaparae women from
around U.S.,
Lunder & Sharp,
2004
153
16.4
Mean of 3 pooled sample analyses from CO,
CA. and NC
Focant et al. 2004
5-74
-------
19.8
Mean from 20 pnmaparac women from
around U.S..
Lunder & Sharp,
2004
16. 4.8
Mean, median, of 40 samples from
primeparae women from Pacific Northwest
NEW. 2004
17.6
WHO inter-lab calibration study with results
from 17 of 24 countries; U.S. listed here
Kotz et al,, 2005
5.3, 2.0
Mean, median. 47 women in Texas 2002
Sehecter et al., 2003
7.0
N=-l 1; mean from blood/milk study ( ND; 0)
Scheeter et al., 2006c
5.2, 1.1
N»38; mean, median from 23 towns in MA
Johnson-Restrepo et
al. 2007
3.05
N 46; median from 46 women in the Boston,
MA area
Wu et al., 2007
154
3.6
Mean of 3 pooled sample analyses from CO,
CA, and NC
Foeant et al., 2004
1.5
Mean from 20 primaparae women from
around U.S..
Lunder & Shaip,
2004
0.8, 0.4
Mean, median, of 40 samples front
primeparae women from Pacific Northwest
NEW, 2004
3.0
WHO inter-lab calibration study with results
from 17 of 24 countries; U.S. listed here
Kotz et al. 2005
0.76, 0.22
Mean, median, 47 women in Texas 2002
Sehecter et al., 2003
0.30
N' -11; mean from blood/milk study (N7> 0)
Sehecter et al., 2006c
0.63. 0.05
N=38; mean, median from 23 towns in MA
Johnson-Restrepo et
al. 2007
0.17
N 46: median from 46 women in the Boston.
MA area
Wu et al., 2007
155
0.3
Mean from 20 primaparae women from
around U.S..
Lunder & Sharp,
2004
HeptaBDE
183
0.15
Mean from 20 primaparae women from
around U.S..
Lunder & Sharp,
2004
0.3, 0.2
Mean, median, of 40 samples from
primeparae women from Pacific Northwest
NEW, 2004
0.13,0.07
Mean, median. 47 women in Texas 2002
Sehecter et al., 2003
0.09
N=11; mean from blood/milk study (ND=0)
Sehecter et al., 2006c
6.2, 4.1
N=38; mean, median from 23 towns in MA
Johnson-Restrepo et
al. 2007
0.065
N=46; median from 46 women in the Boston,
MA area
Wu et al., 2007
190
0.01
Mean from 20 primaparae women from
around U.S..
Lunder & Sharp.
2004
NonaBDE
203
< 7 (ND)
N=38; mean, median from 23 towns in MA
Johnson-Restrepo et
al. 2007
206
0.01
Mean from 20 primaparae women from
around U.S..
Lunder & Sharp,
2004
207
0.06
Mean from 20 primaparae women from
around U.S..
Lunder & Sharp,
2004
208
0.01
Mean from 20 primaparae women from
around U.S..
Lunder & Sharp,
2004
5-75
-------
DecaBDE
209
0.24
Mean from 20 primaparae women from
around U.S..
Lunder & Sharp.
2004
0.8. 0.4
Mean, median, of 40 samples from
primeparae women from Pacific Northwest
NEW, 2004
0.92, ND
Mean, median, 47 women in Texas 2002
Schecter et ai., 2003
0.42
N 11; mean from blood/milk study (ND 0)
Seheeter et al, 2006c
<204 (ND)
N=38: mean, median from 23 towns in MA
Johnson-Restrepo et
al. 2007
DL (0.25)
N 46; median from 46 women in the Boston.
MA area
Wu et al.. 2007
5-76
-------
Table 5.3. Representative body burden levels of PBDEs in Americans
Description
Blood
ng/g lwt
Mother's milk
ng/g lwt
BDE 17
NA
NA
BDE 28
1.2
1.7
BDE 47
20.5
26.0
BDE 66
NA
0.1
BDE 85
NA
0.6
BDE 99
5.0
5,4
BDE 100
3.9
5.2
BDE 138
NA
NA
BDE 153
5.7
4.8
BDE 154
NA
0.4
BDE 183
NA
0.3
BDE 197
NA
NA
BDE 206
NA
NA
BDE 209
NA
0.4
TOTAL
36.3
44.9
Note: blood data from Sjodin et al., 2008; mother's milk data from NEW, 2004.
NA = not available.
5-77
-------
Table 5,4. Estimates of general population intake, or exposure, dose of total PBDEs
provided in the literature.
Pathway; units
Exposure dose; target
individual
Reference; comment
United States anc
Canada
Dust ingestion,
ng/d
120 - 6000 for children;
3.3 for adults
Stapleton et al. (2005); children's
estimate pertinent to ages 1-4, assuming
0.02 - 0.2 g/day ingestion rate; adult
estimate assumes 0.00056 g/day.
Dust ingestion,
ng/d
Child median - 1380
Adult median - 154
Stapleton et al. (2008); measurements
on hands on 33 individuals; empirical
model to estimate dust ingestion
Dust ingestion,
ng/d
400 for adults
Sjodin et al. (2004); based on median of
4200 ng/g found in Atlanta vacuum dust
samples, and 100 mg day adult ingestion
Dust ingestion,
ng/d
147; 228 for adults
580; 910 for toddlers
Harrad et al (2008); dust cone from TX;
median & average dust cone; 50, 200
mg d dust ing for adults, toddlers
Diet, Dust,
Inhalation, ng/d
155 for adults; 264 for
toddlers (4 mo - 2 yrs);
227 for fish eater; 2190
for occupational
Jones-Otazo et al. (2005); 66% of adult
exposure due to dust; 90% of toddler
due to dust; air/dust modeling + food
concentrations; Toronto Canada
Diet, ng/kg-d
0.9-1.5 for males/females
above 12 yr; 2.6 for 2-5
yr; 306 for infant
Schecter et al. (2006); based on market
basket survey and age-based intake
rates; infant was for breast feeding
Diet, ng/kg-d
0.3 - 0.8 for adults
Huwe et al. (2005); lower estimate
based on "lean meats" and higher on
meats of high fat content.
Diet, ng/kg-d
<1 for children/adults for
fish; 0.04 - 20 tor
children'adults for
beef/chicken
Luksembourg et al. (2005); used market
basket survey data on fish, beef, chicken
from Northern CA.
Total dose, (i.e.,
all pathways),
ng/kg-d
8.5, 16.0, and 53.6 for
women
McDonald (2005); mean, median and
95% estimate for total dose based on
backward PK modeling and body
burden data for women.
Europe and Asia
Inhalation and
diet, ng/d
20; 9 (inhalation) and
107; 90 (food) for adults
Harrad et al. (2004); UK, mean and
median provided for inhalation and food
Dust ingestion
and inhalation,
ng/d
0.9 - 22 (dust ing) and 2
(inh) for adults; 12-43
(dust ing) and 0.4 (inh)
for toddlers
Harrad et al (2006); UK; average
results based on dust in 8 homes and air
in 32 indoor locations; mean
concentrations & range of contact rates
Dust ingestion,
ng/d
143; 2205 for adults
572; 9020 for toddlers
Harrad et al. (2008); UK; median &
average dust cone; 50, 200 mg/d ing for
5-78
-------
adults, toddlers; Canadian results also
Inhalation
0.2, 0.0005 - 2.9
Mandalakis et al. (2008); median and
range for inhalation in cars only.
Diet, ng/d
97.3 for adults
Bocio et al. (2005); Spain
Diet, ng/d
72 and 63 for adults
Schuhmacher et al. (2007): industrial
and urban areas of Spain
Dust ingestion &
inhalation,
ng/kg-d
2.0 and 0.2 for children,
0.2 and 0.4 for adults
Gavao et al. (2005); Kuwait
Diet, ng/d
62.5, 48.6, and 149.0 for
adults
Knutsen et al. (2005); Norway
Diet and breast
milk, ng/d
51 for adults; 110 for
infants
Darnerud et al. (2001); Sweden; infant
dose assumed 4.2 ng/g lwt in breast milk
Diet, ng/kg-d
0.79, adult
Bakker et al. (2006); Netherlands; food
consumption survey combined with
composite measurements of 47, 99, 100,
153
Diet, ng/kg-d
5.9, of which BDE 209 is
4.5, adult
FSA (2006);'UK, based on total diet
survey food samples + consumption
data
Diet, ng/d
35, adult
Voorspoclset al. (2006); Belgium, based
on market basket samples +
consumption data, includes BDE 209
Fish, ng/d
Nursing, ng/d
Inhalation, ng/d
1.7 - 12.9, child to adult
48.2, infant
2.7 - 9.2, child to adult
Menget alet al. (2007); China; median
values for fish intake/inhalation for
different ages, 0-1 for nursing infant
5-79
-------
Table 5.5. Exposure pathways and factors for the PBDE intake dose estimate
Exposure Factors;
Units
Comment: description
Adult
Ages
1-5
Ages
6-11
Ages
12-19
body weight, kg
Used for converting ng'day to ng/kg body
weight/day
70
15
30
58
soil Ingestion, mg/d
Central tendency values
50
100
50
50
soil dermal contact,
mg/d
Surface area that contacts the skin (5700
ciiV7d for adults) * amount soil adhering
to skin (0.07 mg/cm") * fraction absorbed
thru skin (0.03); area corresponds to head,
hands, forearms, lower legs
12
2.2
3.2
11
inhalation, nrVd
Unpublished estimates from recent
studies at 16,1 nr/day lor adults
13.3
7.5
12
14
fraction indoor
Children > 12 years and adults assume 21
hr/d; 19 hr/d for children
0.875
0.792
0.792
0.875
water ingestion. I d
Estimates from EPA (1997) still
considered current
1.4
0.69
0.79
0.97
milk ingestion, g/d
Data from USD A (1995)
175
348
357
308
dairy ingestion, g/d
Data from USD A (1995)
55
103
88
77
egg ingestion, g/d
Data from EPA (1997), in units of g/kg
bw/d, assumes bw above
16.8
11.25
12.3
13.9
beef ingestion, g/d
Data from EPA (1997), in units of g/kg
bw/d, assumes bw above
49.7
21
33
48.1
pork ingestion, g/d
Data from EPA (3997), in units of g/kg
bw/d, assumes bw above
15.4
7.2
10.5
15.7
poultry ingestion, g/d
Data from EPA (1997), in units of g/kg
bw/d, assumes bw above
35
16.5
26.1
33,6
other meat, g/d
Data from EPA (1997), in units of g/kg
bw/d, assumes bw above
24.5
16.5
20.7
24.4
freshwater/marine fin
fish, g/d
Data developed in EPA (2003). based on
EPA (2000)
11.6
3
3.8
4.5
freshwater/marine
shellfish, g/d
Data developed in EPA (2003), based on
EPA (2000)
3.8
1
1.2
1,5
Note:
All exposure factors and approaches developed in EPA's Dioxin Reassessment ( EPA,
2003), which relied on EPA's Exposure Factors Handbook (EPA, 1997) and U.S.DA
(1995).
5-80
-------
Tabic 5.6. Congener-specific and total adult intake estimates of PBDEs (exposures in units of ng/day)
Exposure Pathways
17
28
47
66
85
99
100
138
153
154
183
197
206
209
TOTAL
Fraction
soil ingestion
0.00
0.00
81.26
0.92
4.38
102.92
39.86
7.92
10.67
6.86
2.86
0.08
0.01
104.83
362.55
0.66
soil dermal contact
0.00
0.00
19.50
0.22
1.05
24.70
9.57
1.90
2.56
1.65
0.69
0.02
O.00
25.16
87.01
0.16
Inhalation
0.10
0.32
2.15
0.05
0.04
1.00
0.21
0.00
0.06
0.09
0.00
0.00
0.00
1.45
5.47
0.01
water ingestion
0.01
0.00
0.06
0.00
0.00
0.04
0.01
0.00
0.01
0.00
0.01
0.00
0.00
0.06
0.20
0.00
Milk ingestion
0.01
0.04
5.25
0.05
0.14
5.25
0.88
0.01
0.70
0.35
0.35
0.00
0.00
7.00
20.02
0.04
dairy ingestion
0.00
0.01
1.65
0.02
0.04
1.65
0.28
0.00
0.22
0.11
0.11
0.00
0.00
2.20
6.29
0.01
egg ingestion
0.00
0.00
0.34
0.00
0.03
0.67
0.10
0.00
0.07
0.05
0.00
0.00
0.00
0.17
1.44
0.00
beef ingestion
0.02
0.99
2.49
0.01
0.03
1.99
0.30
0.00
0.30
0.20
0.05
0.00
0.00
0.15
6.53
0.01
pork ingestion
0.00
0.00
1.23
0.00
0.08
1.85
0.23
0.02
0.31
0.15
0.14
0.00
0.00
0.31
4.31
0.01
poultry ingestion
0.00
0.01
2.10
0.01
0.00
4.20
1.05
0.07
0.70
0.04
0.07
0.00
0.00
4.20
12.45
0.02
other meats
0.01
0.16
1.55
0.00
0.05
2 29
0.42
0.03
0.38
0.12
0.10
0.00
0.00
1.17
6.26
0.01
fresh/marine finftsh
0.12
0.35
6.96
0.23
0.05
1.97
1.51
0.01
0.23
0.58
0.02
0.00
0.00
1.04
13.07
0.02
fresh/marine shellfi sh
0.00
0.00
13.68
0.00
0.00
4.56
3.42
0.00
0.00
0.00
0.00
0.00
0.00
0.00
21.66
0.04
TOTAL
0.27
1.89
138.21
1.52
5.88
153.09
57.82
9.96
16.20
10.19
4.39
0.10
0.01
147.74
547.27
Fraction
0.00
0.00
0.25
0.00
0.01
0.28
0.11
0.02
0.03
0.02
0.01
0.00
0.00
0.27
1.00
5-81
-------
Table 5.7. Pharmacokinetic parameters and predicted concentrations of BDEs compared with measurements in blood and milk
Exposure Doses, Parameters,
Results
28
4? "
99
100
138
153
154
183
209
TOTAL
1. Doses
soil ingestion, ng/d
0.00
81.25
102.92
39.86
7.92
10.67
6.86
2.86
104.83
357.17
soil dermal contact, ng/d
0.00
19.50
24.70
9.57
1.90
2.56
1.65
0.69
25.16
85.73
Inhalation, ng/d
0.32
2.15
1.00
0.21
0.00
0.06
0.09
0.00
1.45
5.28
Food & water ingestion, ng/d
1,57
35.30
24.47
8,18
0.14
2.91
1.6
0.85
16.3
92.24
If. Parameters
Dust absorption fraction
0.33
0.69
0.44
0.78
0,67
0.73
0.19
0.48
0.04
...
Other absorption fraction
0.90
0.94
0.78
0.93
0.90
0.90
0.86
0.90
0.90
—
Elimination half-life, yrs
6.0
3.0
5.4
2.9
6.0
11.7
5.8
0.26
0.04
...
III. Results
Predicted concentration, ng/g lwt
0.3
8,6
15.6
3.3
l.l
3.7
1.1
0.02
0.1
33.8
Observed blood, ng/g lwt
1.2
20.5
5.0
3.9
NA
5.7
NA
NA
NA
36.3
Observed milk, ng/g lwt
1.7
26.0
5.4
5.2
NA
4.8
0.4
0.2
0.4
44.1
5-82
-------
Table 5.8. Pharmacokinetic parameters for modeling the body burden impacts to infants
via breast feeding, and then to children from food and household exposures
Time after
birth
PBDE half-
life, yr
Body lipid
fraction
Body weight,
kg
Total PBDE
dose, ng/day
0
0.40
0.14
3.3
1411
1 mon
0.50
0.16
4.3
1411
2 mon
0.60
0.18
4.6
1411
3 mon
0.70
0.20
6
1411
4 mon
0.75
0.22
6.7
1411
5 mon
0.80
0.23
7.4
1411
6 mon
0.85
0.25
7.9
1411
7 mon
0.90
0.25
8.4
1411
8 mon
0.95
0.24
8.8
1411
9 mon
1.00
0.24
9.2
1411
10 mon
1.05
0.23
9.4
141 1
11 mon
1.10
0.23
9.8
1411
12 mon
1.15
0.23
11.3
1411
1 yr, 3 mon
1.30
0.22
11.7
757
1 yr, 6 mon
1.50
0.21
12.5
757
1 yr, 9 mon
1.70
0.20
12.9
757
- yr
2.00
0.20
13.3
757
3 yr
2.50
0.18
15.6
757
4 yr
3.00
0.16
17.6
757
5 vr
3.50
0.15
19.7
757
6-11 yr*
4.00 - 6.00
0.15 - 0.13
24-41
450
P 19 yr*
6.00
0.13 0.15
41 - 64
550
*Note: the time increments of calculation for pk modeling was one year after age 5; body
weight and lipid fractions incrementally decreased/increased during that time within the
ranges noted. Doses were constant at the values noted.
5-83
-------
Figure 5.1. Approach for characterizing exposure to polybrominated diphenyl ethers in
this report.
Compare
Compile data on
body burdens
Compile data on
exposure media
concentrations
Develop representative
congener profiles in
the media
Develop representative
body burden congener
profiles
Combine media concentrations
with exposure factors to
estimate intake dose
Convert intake dose to
Body burden using simple
1st order 1 -compartment
PK model
5-84
-------
Figure 5.2. Modeled infant and childhood body burdens of PBDEs.
350
300 -1
250
200
150 -j
100
50
0
Half-life = 6 yrs
Half-life = 6 weeks at birth
J j 1 1 1 1 ! 1 1 1 1 1 1 1 1 1 1 1 1 1
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
Age, years
5-85
-------
Figure 5 .3, The fraction of the maximum concentrations of PCDD/F TEQ and BDE 47
concentrations found at various percentiles within NHANKS surveys of these two
contaminants in adults.
PCDD/F TEQ
25% 50% 75% 95% 100%
NHANES Percentiles Within Adult Population
Note: NHANHS data on PBDEs from Sjodin et al. (2008), and data on PCDD/F from
Ferriby et al. (2006)
5-86
------- |