QUANTITATIVE ANALYSIS AND HEALTH RISK ASSESSMENT OF NOVEL BROMINATED FLAME RETARDANTS IN HOUSE DUST Taya Huang MSc Thesis Master's Degree Programme in Environmental Health Risk Assessment University of Eastern Finland, Faculty of Science and Forestry Department of Environmental and Biological Sciences 28 April, 2017
74
Embed
QUANTITATIVE ANALYSIS AND HEALTH RISK ASSESSMENT OF … · 2017-05-18 · TBP-DBPE 2,3-Dibromopropyl 2,4,6-tribromophenyl ether TCEP Tris(2-chloroethyl) phosphate ... 479/1996 for
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
QUANTITATIVE ANALYSIS AND HEALTH RISK
ASSESSMENT OF NOVEL BROMINATED FLAME
RETARDANTS IN HOUSE DUST
Taya Huang
MSc Thesis
Master's Degree Programme in Environmental Health Risk Assessment
University of Eastern Finland, Faculty of Science and Forestry
Department of Environmental and Biological Sciences
28 April, 2017
UNIVERSITY OF EASTERN FINLAND, Faculty of Science and Forestry
Department of Environmental Science
Master's Degree Programme in Environmental Health Risk Assessment
Taya Huang: Quantitative Analysis and Health Risk Assessment of Novel Brominated Flame
Retardants in House Dust
MSc thesis 66 pages, 8 appendixes ( 74 pages)
Supervisors: Panu Rantakokko, PhD; Matti Viluksela, PhD
Expanded Measurement Uncertainty with 95 % confidence interval
MU=2*utot
utot = Total MU
uAccu = Accuracy
uPres = Precision
Suitability of the Chromatographic System
This validation aimed to have chromatographic peaks that are symmetrical and without
interfering extra peaks in the channel of the compound to be measured. Also, there was to be
no serious interferences in the HRMS lock masses at the retention times of compounds to be
measured. Recovery rates of internal standards would preferably be in the range between 60%
and 120%. Lower recovery rates would be acceptable for highly volatile compounds. Ion
ratios of compounds to be measured should be within +/- 20% of the theoretical values or
values obtained from calibration standards.
26
2.2 RESULTS AND DISCUSSION
2.2.1 FRACTIONATION AND CLEAN-UP OPTIMIZATION
Results for Test 15T024
In this test, the standard 12C mixture of compounds in 500 μl of hexane were added to cleanup
columns. It was concluded that 100% Dcm was the most suitable solvent for the elution of
Fraction 1 from the Florisil column. With this solvent most of the BFRs could be separated
from the OPFRs. Ten% Acetone-Dcm was determined to be the most suitable solvent for the
elution of Fraction 2, where all of OPFR, could be eluted, except that 50% of TBOEP still
remained in the column. However, the use of 13C-labelled internal standard for TBOEP
corrects for this loss. BEH-TEBP was eluted in Fraction 2 together with OPFR, and would
remain so in subsequent tests.
Results for Test 15T031
In this test it was found that the detected concentration of each compound was quite consistent
across all 3 treatments. Table 7 below illustrates the results for this test.
Table 7. The detected concentration of each compound was quite consistent across all 3
treatments. BFR Extract (4 ml of
DCM) directly
to dual-column
Extract evaporated
to 1mL then to
dual-column
Extract evaporated to
0.5ML, 0.5mL hex added
then to dual-column
Certified/reference
values (range)
ng/g
Average Average 15T029-6-F1
ng/g ng/g ng/g
ab-DBE-DBCH 6 3 3
PBT 0 0 0
TBP-DBPE 0 0 0
EH-TBB 17 20 21 26-40 (2-6)
BTBPE 15 14 20 32-76 (4-14)
BEH-TEBP 1898* 6216* 0* 145-1300(16.7-
94)*
DBDPE 0 0 0 <20
BDE-28 24 24 25 46.9 (4.4)
BDE-47 243 249 279 497 (46)
BDE-100 74 78 95 145 (11)
BDE-99 486 502 577 892 (53)
BDE-154 40 41 49 83.5 (2.0)
BDE-153 59 60 68 119 (1)
BDE-183 no data no data no data 43.0 (3.5)
BDE-209 2306 2241 2336 2510 (190)
* It was found that BEH-TEBP had degraded in H2SO4-silica , therefore, these results for BEH-TEBP for Fraction 1 were not reliable.
27
The average concentration of compounds for 15T031 detected was approximately 50% of
certified/reference values. This error was due to calibration and was corrected in subsequent
tests. However, in the interest of determining the most simple and suitable fractionation
technique, the above technical issues were not pursued further.
In Test 15T029, activated Na2SO4 was found to be ineffective in removing impurities in
Fraction 2. Tests for the cleanup of Fraction 2 were not continued further.
2.2.2 EXTRACTION
Results for Test 15T029
The above-described clean-up facilitated the GC-HRMS analysis of Fractions 1-A and 1-B.
Based on the finding that the detected concentration of BFR and PBDE were consistent across
all three treatments in Test 15T031, direct pouring of 4ml Dcm extract was chosen for future
implementation. As Table 8 shows, the concentrations for PBDEs analyzed from SRM2585
by different solvents was generally consistent with certified values.
Table 8. The concentrations for PBDEs analyzed from SRM2585 by different extraction
solvents. Certified Value (SRM-
2585)
100%
Dichloromethane
25% Acetone-
Hexane
50%
Dichloromethane
-Ether
ng/g Average (ng/g) Average (ng/g) Average (ng/g)
BDE-28 46.9 (4.4) 45 46 45
BDE-47 497 (46) 531 519 524
BDE-100 145 (11) 175 170 173
BDE-99 892 (53) 1019 1009 1034
BDE-154 83.5 (2.0) 83 81 83
BDE-153 119 (1) 114 111 112
BDE-183 43.0 (3.5) 42 42 36
BDE-209 2510 (190) 2570 2871 2549
All three of the elution solvents tested, namely, 100% Dichloromethane, 25% Acetone-
Hexane, 50% Dichloromethane - Ether were shown to be equally efficient in extraction of
BFR. Table 9 shows that the amount of BFRs extracted are comparable across all three
elution solvents.
28
Table 9. The amount of BFRs extracted are comparable across all three elution solvents.
BFR Comparison value
(range)
100%
Dichloromethane
25%
Acetone-
Hexane
50%
Dichlorometh
ane-Ether
Average (ng/g) Average
(ng/g)
Average (ng/g)
ab-DBE-
DBCH
5 3 3
PBT 0 0 0
TBP-DBPE 0 0 0
EH-TBB 26-40 (2-6) 38 38 39
BTBPE 32-76 (4-14) 50 33 33
BEH-TEBP 145-1300(16.7-94) n/a n/a n/a
DBDPE <20 0 1 1
100% dichloromethane was chosen to be the most suitable extraction solvent for subsequent
extractions due to ease of preparation and use. It was found that an additional H2SO4- silica
clean-up column was necessary for the BFR fraction in order for the extract to be sufficiently
clean for the GC-HRMS analysis. The main benefit of using 100% dichloromethane is that it
can be directly poured to dual Florisil - H2SO4- silica clean-up column after extraction
without prior evaporation that is needed for 25% Acetone-Hexane and 50% Dichloromethane-
Ether.
29
2.2.3 SUMMARY OF SAMPLE TREATMENT IN FLOW CHART
Fig. 2 Flow-chart of analysis method.
Figure 2 shows a flow-chart of the analysis method.
2.2.4 RESULTS OF VALIDATION
Summary of Results
LOQ & MU were acceptable for all BFRs. LOQ of BFRs ranged between 0.5 – 5.0 ng/g. For
OPFR, LOQ ranged between 6.9-613 ng/g. Precision for OPFR was good, meaning the
relative SD were consistent for each compound. Accuracy for OPFR was poor compared to
other labs’ analysis result for the certified material for indoor dust SRM 2585. Therefore,
method for OPFR analysis requires further testing that was not conducted within the scope of
this thesis. Tables 10, 11, 12 and 13 below present a summary of validation results, taken
from an internal validation report written by Panu Rantakokko, PhD, Senior Researcher of
THL on 11 September 2015.
30
Table 10. Summary of LODs, LOQs and MUs to be reported for BFRs.
Compound LOD (ng/g) LOQ (ng/g)
MU (%)
< 50 ng/g
MU (%)
> 50 ng/g
ab-DBE-DBCH 1.0 2.5 55 55
PBT 0.2 0.5 55 40
TBP-DBPE 0.2 0.5 65 45
EH-TBB 2.0 5.0 25 25
BTBPE 0.2 0.5 25 25
BEH-TEBP* 10 25 100 50
DBDPE 2.0 5.0 30 30
BDE-28 0.2 0.5 25 20
BDE-47 0.2 0.5 30 20
BDE-100 1.0 2.5 40 40
BDE-99 1.0 2.5 25 25
BDE-154 0.3 0.7 25 25
BDE-153 0.3 0.7 30 30
BDE-183 0.6 1.5 35 35
BDE-209 1.0 2.5 75 75
Table 11. LODs and LOQs based on blank (n=4) and MassLynx (OPFRs, n=21).
Compound
/Sample LOD (ng/g) LOQ (ng/g)
TIBP 20 52
TNBP 12 32
TCEP 25 67
TCIPP 230 613
TDCIPP 21 57
TPHP 2.6 6.9
TBOEP 6.5 17
EHDPP 34 91
TEHP* 40 106
TMPP 7.1 19
*TEHP has very poor sensitivity with the ions selected, but alternative ions would be low
mass and extremely noisy and non-specific.
31
Table 12. Accuracy, precision and MU from results of SRM 2585 a (BFRs, n=8).
COMPOUND
Average (range)
(ng/g)
RSD
(%)
Certified/Other
(ng/g)
Recovery
(%) MU (%) Source
ab-DBE-DBCH <LOQ
PBT 0.22 (0.12-0.26) 24
TBP-DBPE <LOQ
EH-TBB 33 (31-36) 5.0 26 127 56 Van den Eede 2012
BTBPE 54 (35-86) 40 b 39 140 b 112 b Van den Eede 2012
BEH-TEBP c 1212 (1141-1281) 5.6 574 211 c 223 c Van den Eede 2012
DBDPE <LOQ <7.1 Van den Eede 2012
BDE-28 49 (47-51) 3.5 46.9 104 11 Certified
BDE-47 487 (443-551) 8.2 497 98 17 Certified
BDE-100 164 (148-191) 9.5 145 113 32 Certified
BDE-99 954 (870-1114) 8.5 892 107 22 Certified
BDE-154 83 (78-96) 7.6 84 99 15 Certified
BDE-153 115 (109-133) 7.5 119 96 17 Certified
BDE-183 37 (33-46) 11 43 87 34 Certified
BDE-209 3138 (2368-4807) 26 d 2510 125 d 72 d Certified a SRM 2585 has certified concentrations for PBDEs only. For other BFRs many sources exist, but Van den
Eede represents generally a recognized high quality laboratory. b For BTBPE some results were outliers, especially in the series 2. Source of deviation needs to be tested in
future work. c Sahlström et al 2012 measured a concentration of 1300 ng/g for BEH-TEBP. d Range of results for BDE-209 is large. Possible laboratory contamination needs to be tackled in future
testing.
Table 13. Accuracy, precision and MU from results of SRM 2585 a (OPFRs, n=4).
COMPOUND
Average
(range) (ng/g)
RSD
(%) b
Certified/
Other (ng/g)
Recovery
(%) b
MU
(%) Source
TIBP 18 101.5
TNBP 1196 1.6 190 629 1059 Van den Eede 2012
TCEP 3363 6.2 680 495 789 Van den Eede 2012
TCIPP 3780 9.8 860 440 679 Van den Eede 2012
TDCIPP 7744 14.0 3180 244 288 Van den Eede 2012
TPHP 4335 5.9 1160 374 548 Van den Eede 2012
TBOEP 57568 7.5 63000 91 23 Van den Eede 2012
EHDPP 3039 9.5 1300 234 268 Bergh 2012
TEHP 1091 9.1 370 295 390 Bergh 2012
TMPP 3435 10.0 1140 301 403 Van den Eede 2012
32
Suitability of Chromatographic System
For the BFRs, percentage recovery of internal standards were found to be in the range of 60%
- 120%. Ion ratios of compounds to be measured were within +/- 20% of the theoretical
values, or values from calibration standards.
For the OPFRs, peak sizes tended to be too small in the calibration standard for all
compounds, and for some compounds in the Internal Standard solution added to samples.
Table 14 below illustrates the amount of dust detected in Home Settled Dust samples (n=4),
Home Exhaust Air Filter Dust samples (n=4) and Home Air Condition Filter Dust samples
(n=4) during the validation process. Home settle dust samples and home exhaust air filter dust
samples are from the same home (Home 1), while Home air condition filter dust samples are
from another home (Home 2).
33
Table 14. Amount of dust detected in Home Settled Dust samples, Home Exhaust Air Filter
Dust samples, and Home Air Condition Filter Dust samples during validation. Vapor pressure
of each compound is included for comparison. Sample
type
Settled Dust (n=4,
Home 1)
Exhaust Air Filter Dust
(n=4, Home 1)
Air Condition Filter Dust
(n=4, Home 2)
Compound Average
(ng/g)
RSD
(%)
Average (ng/g) RSD
(%)
Average (ng/g) RSD
(%)
Vapour
Pressure
(Pa)(25C)
ab-DBE-
DBCH
0.56 29.2 0.41 22 0.31 5.8 2.97E-03
PBT 0.70 11 0.63 36a 1 19 6.00E-04
TBP-
DBPE
0.64 4.9 2.0 5.4 0.08 13 1.26E-05
EH-TBB 5.6 15 8.7 3.4 2.7 23 3.71E-07
BTBPE 3.5 6.1 2.7 9.2 8.5 135 3.88E-10
BEH-
TEBP
136 19 138 8.2 192 11 1.55E-11
DBDPE 158 17 212 5.7 327 140 n/a
BDE-28 0.95 5.1 1.2 6 0.27 8.1 n/a
BDE-47 23 4.3 24 2.7 4.1 4.1 n/a
BDE-100 4.6 3.9 4.5 10.7 0.94 7.7 n/a
BDE-99 41 2.9 38 2 5.3 11 n/a
BDE-154 2.4 5.6 2.7 1.9 0.49 31 n/a
BDE-153 5.6 7.2 8.0 3.9 1.2 60 n/a
BDE-183 1.3 14 2.2 37 2.9 98 n/a
BDE-209 752 2.2 785 3.6 184 36 n/a
34
2.2.5 CONCLUSIONS OF METHOD DEVELOPMENT AS OF 2016
It was concluded that 100% Dcm was the most suitable solvent for the elution of Fraction 1
from the Florisil column. With this solvent most of the BFRs could be separated from the
OPFRs. 10% Acetone-Dcm was determined to be the most suitable solvent for the elution of
Fraction 2, where all of OPFR, could be eluted, except that 50% of TBOEP still remained in
the column. BEH-TEBP was eluted in Fraction 2 together with OPFR. Direct pouring of 4ml
Dcm extract was chosen for future implementation. 100% dichloromethane was chosen to be
the most suitable extraction solvent for subsequent extractions due to ease of preparation and
use. Settled dust was considered to be a preferable matrix.
For the BFRs, percentage recovery of internal standards were found to be in the range of 60%
- 120%. Ion ratios of compounds to be measured were within +/- 20% of the theoretical
values, or values from calibration standards. For the OPFRs, chromatography peak sizes
tended to be too small in the calibration standard for all compounds, and for some compounds
in the Internal Standard solution added to samples.
For the validation, Limit of Quantification (LOQ) & Measurement Uncertainties were
acceptable for all BFRs. LOQ of BFRs ranged between 0.5 – 5.0 ng/g. For OPFR, LOQ
ranged between 6.9-613 ng/g. Precision for OPFR was good, but accuracy was poor compared
to other labs’ analysis result for the certified material for indoor dust SRM 2585. Therefore,
method for OPFR analysis requires further testing that was not conducted within the scope of
this thesis.
As compared to sampling of settled dust by collecting settled dust on surfaces with a vacuum
cleaner, indoor dust collected on the exhaust air filter could be a good time and space
integrated sample from the entire indoor space of a household, in the case where this
particular exhaust mechanism is installed. Therefore, exhaust filter dust can be representative
of one indoor compartment as all air that exit the house goes through the filter. However, the
concentration of more volatile BFRs such as ab-DBE-DBCH and PBT tend to be lower than
that on settled dust, as illustrated in Table 14. Therefore, settled dust would be a preferable
matrix than exhaust dust within the scope of this particular validation test described in this
thesis. Air conditioner filter dust showed large deviation, as illustrated by the high relative
standard deviation in Table 14. Therefore, air conditioner filter dust was not an ideal matrix.
35
Figure 3 shows a schematic of an example of a heat recovery unit of a house ventilation
system with Stale air from inside filter used for sample collection.
Figure 3. Schematic illustration of exhaust ventilation. This figure has been provided by Dr.
Panu Rantakokko of the National Institute for Health and Welfare (THL), Kuopio.
It was noticed that the spiked samples were more homogenous than non-spiked samples when
1g of sample was spiked at one time. The reason may be that the FRs in the samples have
been extracted and equally redistributed during the spiking process with dichloromethane,
resulting in a more homogeneous overall distribution of the FRs present in the sample.
However, it would be difficult to obtain such a large quantity of indoor dust sample from a
single site unless taken from exhaust air filter. Large dust mass available from the filter is one
significant benefit of using it for dust sampling.
2.2.7 DISCUSSION
It was recommended that elution of BEH-TEBP need to be tested further to reduce BEH-
TEBP degradation in H2SO4-silica clean-up column. There was an aim to get BEH-TEBP to
36
Fraction 1. However, it has since proven to be unsuccessful. OPFR accuracy needs to be
improved, but at the present, difficulties with impurities remain. There may be a need to
increase OPFR concentrations in Internal Standard and Calibration Standard Solutions to
match the levels found in actual samples. Also, it may be good to have separate calibration
solutions for BFR and OPFR.
3. HUMAN HEALTH RISK ASSESSMENT
According to the definition put forward by the United States Environmental Protection
Agency (USEPA 2017), human health risk assessment is a process of estimating the
probability of human health effects, especially adverse health effects for a given exposure to a
substance, which can be a chemical present in an environmental medium. There are four basic
steps to human health risk assessment: hazard identification, dose-response assessment,
exposure assessment and risk characterization. Hazard identification attempts to determine
whether a substance may cause harm to humans by putting together available information on
toxicokinetics and possible adverse effects of a chemical on human health. Dose-response
assessment attempts to determine a numerical relationship between exposure to the substance
and the effects. The dose-response relationship links the probability and severity of adverse
health effects to the level of exposure to the substance. Exposure assessment is a process of
determining the magnitude of exposure, frequency of exposure and duration of exposure to
the substance. Exposure assessment also takes into consideration the population exposed to
the substance. Risk characterization summarises the information gathered for the first three
steps of the risk assessment process, and from these information, conclusions may be drawn
regarding the extent of risk resulting from exposure to a substance.
Figure 4 below shows a schematic representation of the four steps to human health risk
assessment. This figure was taken from the USEPA website (USEPA 2017).
37
Figure 4. The four-step process to human health risk assessment.
Hazard characterization in this section will be based on currently available information from
literature. The endpoint Derived No-Effect Level (DNEL) will be employed. Risk
Characterisation Ratio will be used to perform hazard characterization.
The DNEL is a part of human health hazard assessment stipulated in Annex I of REACH,
which comprises of four steps: evaluation of non-human information, evaluation of human
information, classification and labelling, and finally, the derivation of DNEL (Munn 2007).
DNEL is the level of human exposure that should not be exceeded. In hazard characterization,
the estimated exposure of a population is compared with the corresponding DNEL. The risk
of adverse health effect is considered to be adequately controlled if the exposure level does
not exceed the corresponding DNEL, as stipulated in REACH Annex I, Section 6.4 (ECHA
2016a).
The Risk Characterisation Ratio (RCR) is a way to quantify risk in a given exposure scenario.
Based on the guidance from ECHA (2016a), the RCR is calculated as follows:
RCR = Exposure/DNEL
If exposure is less than DNEL, i.e. RCR<1, the risk is adequately controlled. If exposure is
larger than DNEL, i.e. RCR>1, the risk is not adequately controlled.
38
This section will focus on the six novel and emerging BFR listed in Table 1, namely, ab-
DBE-DBCH, PBT, TBP-DBPE, EH-TBB, BTBPE and BEH-TEBP. The OECD eChem
portal, HSDB database and the ECHA CHEM database were consulted for available relevant
information on the above substances.
There was insufficient toxicological information available for ab-DBE-DBCH, PBT, TBP-
DBPE to perform a hazard identification. There is some information available for BTBPE
from the Scientific Opinion on Emerging and Novel Brominated Flame Retardants (BFRs) in
Food, issued by the European Food Safety Authority in 2012. An evaluation of human-health
related toxicity has been included in the United States Environmental Protection Agency
Alternatives Assessment update (2015) for EH-TBB and BEH-TEBP. However, the substance
EH-TBB is not found on the OECD eChem portal. Chemical registration dossier (ECHA
2016b) is available for BEH-TEBP through the European Chemical Agency’s ECHA CHEM
database. In this dossier, a Derived no-effect level (DNEL) for oral exposure has been
estimated for BEH-TEBP.
This section summarises relevant toxicological information available for the substances
BTBPE, EH-TBB and BEH-TEBP. An exposure estimation has been performed based on the
amount of BFR detected from a recent unpublished study by THL. Based on this exposure
estimation, Risk Characterisation Ratio based on the Guidance on Information Requirements
and Chemical Safety Assessment (ECHA 2016a) has been calculated for BEH-TEBP
according to the DNEL for oral exposure.
The substance Bis(2-ethyl-1hexyl)tetrabromophthalate (BEH-TEBP) with CAS Number
26040-51-7 was selected for hazard characterization based on the availability of a DNEL
value for oral exposure for the general population.
The DNEL value for BEH-TEBP has been derived for long-term oral exposure according to a
repeated dose oral toxicity study which will be described in Section 3.3.2.
A commercial mixture FM-550 containing EH-TBB and BEH-TEBP has been used in several
toxicological experiments mentioned in this section. The exact composition of FM-550 is
39
proprietary. However, it has been found by Stapleton et al. (2008) that FM-550 contains
approximately 50% of isopropylated triaryl phosphate and triphenylphosphate. The other 50%
consisted of brominated compounds EH-TBB and BEH-TEBP in approximately 4:1 ratio by
mass.
Another commercial mixture, FMBZ-54 has also been used in several experiments mentioned
in this section. FMBZ-54 comprises of EH-TBB:BEH-TEBP in an approximately 4:1 ratio
(Bearr et al. 2012).
3.1 BTBPE
The European Food Safety Authority has published a Scientific Opinion on Emerging and
Novel Brominated Flame Retardants (BFRs) in Food, updated on 6 December 2013. It was
concluded that 1,2-bis(2,4,6-tribromophenoxy)ethane (BTBPE), with the CAS Number
37853-59-1, has a possibility to raise concerns for bioaccumulation, based on available
experimental and environmental behavior data. BTBPE was considered to be of high
persistency in the environment. In the European Union, BTBPE has been classified as low
production volume (LPV) chemical, which implies that the import or production volume is
more than 10 tonnes but less than 1000 tonnes per year. BTBPE is in pre-registration under
REACH. However, no registration dossier has been made publicly available. There is very
limited information on toxicity in humans. This section will focus on relevant information on
possible toxicity in humans, therefore, ecotoxicological information will not be discussed.
3.1.1 Hazard Identification for BTBPE
Toxicokinetic information
In a study of rats given 0.05-5% 14C-BTBPE in the diet for one day by Nomeir et al (1993),
there was a very limited amount of radioactivity eliminated in urine, of less than 1% of dose
ingested, but a high percentage of faecal excretion, of 80-100% of the dose ingested. In most
of the tissues, there were undetectable levels of radiolabeled compounds. This suggested that
BTBPE gastrointestinal absorption is poor in rats. However, in rats given a diet with 500
mg/kg bw/day 14C-BTBPE for a duration of 10 days, it was found that the adipose tissue,
kidney, skin, the thymus contained the highest concentration. Less than 0.01 % of the dose
was found in the majority of the tissues.
40
In another study by Hakk et al. (2004), rats were given a single dose of 2 mg/kg bw 14C-
BTBPE by gavage. 100% of the dose was recovered in the faeces. The same research group
also demonstrated that elimination of radioactivity by bile was less than 1%, which suggested
that the faecal elimination was primarily from unabsorbed BTBPE. Due to this low level of
absorption, tissue level of BTBPE was low. After 72 h of the single dose given, more than
0.1% of the dose was found only in the gastrointestinal tract and carcass.
3.1.2 Toxicity and Dose-Response Information on BTBPE
LD50 for BTBPE was estimated to be >10g/kg bw for rats and dogs. No compound-related
effects were observed in rats after being fed up to 10% BTBPE in the diet, at an estimated
concentration of 35 mg/kg bw/day, for 14 days. In an inhalation study, rats inhaled BTBPE at
5 or 20 mg/liter in the atmosphere for 21 days. No gross pathological changes were observed.
However, it was observed in the lungs unspecified histopathological lesions (Matthews 1984,
cited by Nomeir et al. 1993).
The reproductive and developmental toxicity of BTBPE was studied by Egloff et al. (2011). It
has been found that there are no hatching effects in chicken. In the WHO/IPCS evaluation
(2005) for genotoxicity and carcinogenicity, BTBPE was found to be not mutagenic in Ames
test and S. cerevisiae. No information was available for BTBPE on human health endpoints.
3.2 EH-TBB
In a report published in August 2015, The USEPA has evaluated a number of novel and
emerging brominated flame retardants that have been used as alternatives for phased-out
PBDEs. 2-Ethylhexyl-2,3,4,5-tetrabromobenzoate (EH-TBB) with CAS Number 183658-27-7
was one of the brominated flame retardants evaluated. EH-TBB is not found on the OECD
eChem portal. EH-TBB is not registered under REACH, and no information on production
volume in the EU is available (EFSA 2013).
USEPA has stipulated a hazard criteria used to interpret available data and assign a hazard
level. These hazard criteria, named the “Design for the Environment Alternatives Assessment
41
Criteria for Hazard Evaluation” by USEPA were finalized in 2011. When insufficient
information is available, hazard designation would be assigned conservatively by weight of
evidence (USEPA 2015). The criteria used by USEPA to assign hazard designations is
included as a table in Appendix 2.
3.2.1 Hazard Identification for EH-TBB
Toxicokinetic Information
Experimental data with FM-550 showed that it was possible for EH-TBB to be absorbed after
oral exposure from gestation and through lactation. EH-TBB was found in the tissues of
exposed dams and pups after exposure to the FM-550 (Patisaul et al. 2013). In a study by
Patusaul et al. (2013), pregnant rats were given 0, 0.1 or 1 mg/kg bw/day FM-550 in the diet
through gestation day 8 until post-natal day 21. FM550 components, including EH-TBB and
BEH-TEBP were detected in adipose, liver, and muscle tissues of dams at post-natal day 21 at
768 ng/g w.w. for high dose, and 29.6 ng/g w.w for low dose, and less than 7 ng/g w.w. in
controls. EH-TBB was also detected in pooled post-natal day 21 pup adipose tissue.
The primary metabolite of EH-TBB was found to be tetrabromobenzoic acid (CAS number
27581-13-1) by in vitro metabolism experiments with human liver microsomes, rat liver
microsomes, rat cytosol, rat intestinal microsomes, and rat serum following exposure to EH-
TBB. Phase two metabolites of tetrabromobenzoic acid was not found (Roberts et al. 2012).
Tetrabromobenzoic acid was also detected in liver tissue of dams on post-natal day 21 in the
experiment by Patusaul et al. (2013).
Hazard Identification
EH-TBB was evaluated by USEPA, according to the hazard criteria mentioned above and
included in Appendix 2, to have low acute toxicity and low genotoxicity. However, it was
evaluated to have a moderate carcinogenicity, moderate reproductive toxicity, developmental
toxicity, neurotoxicity and repeated dose toxicity. EH-TBB was considered to have a high
persistence and high tendency for bioaccumulation in the environment.
42
3.2.2 Toxicity and Dose-Response Information for EH-TBB
For Acute Mammalian Toxicity, the Acute Oral Lethality was estimated to be LD50 >5000
mg/kg bw based on several studies. EH-TBB was estimated to have uncertain potential for
carcinogenicity, based on professional judgement and analogy with closely related chemicals.
EH-TBB is considered to have a low genotoxicity by USEPA. Gene Mutation test in vitro and
Chromosomal Aberration in vitro tests yielded negative results (Chemtura 2006).
Reproductive Toxicity
EH-TBB is considered to have a moderate reproductive toxicity by USEPA. In a 2-generation
oral gavage reproductive toxicity study in rats, no reproductive effects were identified at
doses up to 165 mg/kg bw/day. This study was done with the commercial mixture FMBZ-54,
containing EH-TBB and BEH-TEBP, with the larger constituent being EH-TBB. 165 mg/kg
bw/day was the highest dose tested with the mixture FMBZ-54 and was considered as the
NOAEL. No adverse effects were observed in reproductive performance and fertility. This
NOAEL falls within the range of Moderate hazard criteria set up by USEPA (MPI Research
2008a, USEPA 2015).
Developmental Toxicity
EH-TBB is considered to have a moderate developmental toxicity by USEPA. In a 2-
generation oral gavage reproductive toxicity study in rats, given 15, 50, or 165 mg/kg-day of
FMBZ-54, it was found that pups at birth had lower body weights. Body weights were also
lower throughout lactation, in both first and second generation offspring. In the first-
generation female, premating body weight was lower. At lactation day 21, spleen weights
were decreased in first generation male pups and both male and female pups in second
generation. A NOAEL of 50 mg/kg bw/day was identified from this study based on the effects
on body weight. This NOAEL falls within the range of Moderate hazard criteria. However, it
was not very clear which component or components of the commercial mixture had caused
the observed developmental effects. LOAEL was estimated to be 165 mg/kg bw/day based on
this study. (MPI Research 2008a, USEPA 2015).
In an unpublished prenatal study by MPI Research (2008b), rats were exposed to 0, 50, 100,
300 mg/kg bw/day FMBZ-54 mixture on gestation days 6-19. There was increased incidence
of dams with sparse hair in the abdomen, lower gestation body weight, and lower food
43
consumption during gestation at doses higher than or equal to 100 mg/kg bw/day. At 100
mg/kg bw/day, lower fetal weight was observed. Incidence of fused cervical vertebral neural
arches increased in fetuses at the highest dose. At highest dose, increased incidence of fetal
ossification variations, including additional ossification centres to the cervical vertebral neural
arches, incomplete ossified skull bones (jugal, parietal, and squamosal), and unossified
sternebrae were also observed. For maternal toxicity, a NOAEL of 50 mg/kg bw/day and a
LOAEL of 100 mg/kg bw/day was estimated based on the above-described effects. For
developmental toxicity, NOAEL of 50 mg/kg bw/day and LOAEL of 100 mg/kg bw/day were
estimated based on decreased fetal weight (2008b, USEPA 2015).
Neurotoxicity
EH-TBB is conservatively designated to have a moderate neurotoxicity by USEPA. There is
very limited experimental data available and no data available on neurotoxicity screening. In a
28-day sub-chronic oral toxicity study in rats treated with FM-550 in doses 0, 160, 400, 1000
mg/kg bw/day, no neurotoxic effects were reported. The NOAEL in this study was reported to
be 1000 mg/kg bw/day, which was the highest dose tested with FM-550 (Chemtura 2006).
Repeated Dose Effects
EH-TBB is considered to have a moderate repeated dose effects, designated by USEPA based
on the two developmental and prenatal study already described above (MPI Research 2008a,
2008b).
A 28-day sub-chronic oral toxicity study was performed in rats, treated with 0, 160, 400, 1000
mg/kg bw/day. Unspecified kidney effects were reported at 1000 mg/kg bw/day. No systemic
effects were observed at 160 mg/kg bw/day, and therefore a NOEL was estimated based on
this observation. Based on kidney effects, a LOAEL of 1000 mg/kg bw/day was estimated.
NOAEL was estimated at 400 mg/kg/day (Chemtura 2006).
3.3 BEH-TEBP
In a report published in August 2015, The USEPA has evaluated a number of novel and
emerging brominated flame retardants that has been used as alternatives for phased-out
PBDEs. Bis(2-ethyl-1hexyl)tetrabromophthalate (BEH-TEBP) with CAS Number 26040-51-7
was one of the brominated flame retardants evaluated.
44
In addition, BEH-TEBP is a pre-registration substance under REACH (ECHA 2016b). It has
been classified as low production volume (LPV) chemical, which implies that the import or
production volume is more than 10 tonnes but less than 1000 tonnes per year.
3.3.1 Hazard Identification for BEH-TEBP
Toxicokinetic information
Experimental data with a commercial mixture FM-550 showed that it was possible for BEH-
TEBP to be absorbed after oral exposure. BEH-TEBP was found in the tissues of exposed
dams after exposure to the commercial mixture, however, not in the pups even though
exposure has been from gestation to lactation (Patisaul et al. 2013).
In in vitro tests, mono(2-ethylhexyl)tetrabromophthalate (CAS number 61776-60-1) was
found to the primary metabolite. In rat or human subcellular fractions, no metabolites of
BEH-TEBP was found. This metabolite was formed by purified porcine carboxylesterase at a
rate of 1.08 mol/min mg/protein. No phase two metabolite of the primary metabolite was
found (Roberts et al. 2012). BEH-TEBP has not been evaluated in humans (USEPA 2015).
BEH-TEBP has been metabolized in vitro in hepatic subcellular fractions of fathead minnow,
common carp, snapping turtle and wild-type mice (Bearr et al. 2012). There was no data
available on toxicokinetic properties of the pure BEH-TEBP compound after oral, inhalation
or dermal exposure.
Hazard Identification
This FR was evaluated by USEPA to have low acute toxicity. However, it was evaluated to
have a moderate carcinogenicity, genotoxicity, reproductive toxicity, developmental toxicity,
neurotoxicity and repeated dose toxicity.
It was stated by the REACH registration applicant that conclusion cannot be drawn for
bioaccumulation potential in mammals based on the results of study.
45
3.3.2 Toxicity and Dose-Response Information for BEH-TEBP
For acute mammalian toxicity, the acute oral lethality was estimated to have an LD50 of larger
than or equal to 2000 mg/kg in rats, based on two studies (Bradford et al. 1996, Chemtura
2006). BEH-TEBP was estimated to have uncertain potential for carcinogenicity, based on
professional judgement and analogy with closely related chemicals. BEH-TEBP is considered
to have a moderate genotoxicity. In a chromosomal aberration test with human lymphocytes,
there was a weak positive result for the test material RC9927; CASRN 26040-51-7 (Purity of
BEH-TEBP > 95%) (ACC 2004). Two other in vitro chromosomal aberration assays were
performed using a component of a commercial mixture FM-550 containing BEH-TEBP,
which yielded negative results (Chemtura 2006). In an in vivo mouse micronucleus assay, it
was found that BEH-TEBP did not cause gene mutation in bacteria or chromosomal
aberration (ACC 2004).
A study submitted by a REACH registration applicant estimated an LD50 of >5000 mg/kg
bw. 5 female and 5 male rats were administered a single oral dose of 5000 mg/kg bw of BEH-
TEBP by gavage and observed for mortality and clinical signs for 14 days. No death occurred
to any animal. Body weight gain was normal and on day 15, there was no relevant necropsy
finding (ECHA 2016b).
Reproductive Toxicity
BEH-TEBP is considered to have a moderate reproductive toxicity by USEPA. In a 2-
generation oral gavage reproductive toxicity study in rats treated with 15, 50, or 165 mg/kg-
day FMBZ-54, no reproductive effects were identified at doses up to 165 mg/kg bw/day. 165
mg/kg bw/day was the highest dose tested and was considered as the NOAEL. No adverse
effects on reproductive performance or fertility in rats were observed. This NOAEL falls
within the range of Moderate hazard criteria (MPI Research 2008a, USEPA 2015).
In a 28-day repeated dose dietary toxicity study in rats given 0, 200, 2,000, and 20,000 ppm in
diet (approx. 0, 21.1, 211, 2,110 mg/kg bw/day) of test material RC9927; CASRN 26040-51-
7 (Purity of BEH-TEBP > 95%), no adverse changes in testes or ovary weights was observed.
Gross necropsy and histopathology were performed on a full complement of male and female
reproductive organs and tissues and no adverse effects were observed. However, other
46
reproductive indicators have not been examined. A NOAEL of 2000 ppm (approx. 223.4
mg/kg bw/day) for dietary toxicity was established. 2100 mg/kg bw/day was the highest dose
tested.
Developmental Toxicity
BEH-TEBP is considered to have a moderate developmental toxicity by USEPA. In a 2-
generation oral gavage reproductive toxicity study in rats, given 15, 50, or 165 mg/kg-day of
FMBZ-54, it was found that pups at birth had lower body weights. Body weights were also
lower throughout lactation, in both first and second generation offspring. In the first-
generation female, premating body weight was lower. At lactation day 21, spleen weights
were decreased in first generation male pups and both male and female pups in second
generation. A NOAEL of 50 mg/kg bw/day was identified from this study based on the effects
on body weight. This NOAEL falls within the range of Moderate hazard criteria. However, it
was not very clear which component or components of the commercial mixture had caused
the observed developmental effects. LOAEL was estimated to be 165 mg/kg bw/day based on
this study. (MPI Research 2008a, USEPA 2015).
In an unpublished prenatal study by MPI Research (2008b), rats were exposed to 0, 50, 100,
300 mg/kg bw/day FMBZ-54 mixture on gestation days 6-19. There was increased incidence
of dams with sparse hair in the abdomen, lower gestation body weight, and lower food
consumption during gestation at doses higher than or equal to 100 mg/kg bw/day. At 100
mg/kg bw/day, lower fetal weight was observed. Incidence of fused cervical vertebral neural
arches increased in fetuses at the highest dose. At highest dose, increased incidence of fetal
ossification variations, including additional ossification centres to the cervical vertebral neural
arches, incomplete ossified skull bones (jugal, parietal, and squamosal), and unossified
sternebrae were also observed. For maternal toxicity, a NOAEL of 50 mg/kg bw/day and a
LOAEL of 100 mg/kg bw/day was estimated based on the above-described effects. For
developmental toxicity, NOAEL of 50 mg/kg bw/day and LOAEL of 100 mg/kg bw/day were
estimated based on decreased fetal weight (2008b, USEPA 2015).
In a study submitted by REACH registration applicant, pregnant rats were administered BEH-
TEBP by oral gavage at dose levels of 250, 500 or 1000 mg/kg bw/day during gestation. No
effect was found on body weight development and dietary intake. A NOEL for maternal
47
toxicity was established at 1000 mg/kg bw/day at the highest dose tested. No relevant adverse
effects were observed in the offspring. The NOEL for developmental toxicity was therefore
1000 mg/kg bw/day at the highest dose tested (ECHA 2016b).
Neurotoxicity
BEH-TEBP is conservatively designated to have a moderate neurotoxicity by USEPA. There
is very limited experimental data available and no data available on neurotoxicity screening.
In a 28-day sub-chronic oral toxicity study in rats treated with FM-550 in doses 0, 160, 400,
1000 mg/kg bw/day, no neurotoxic effects were reported. The NOAEL in this study was
reported to be 1000 mg/kg bw/day, which was the highest dose tested with FM-550
(Chemtura 2006).
Repeated Dose Effects
BEH-TEBP is considered to have a moderate repeated dose effects. In a 28-day dietary
toxicity study, a small decrease of body weight, as well as decreased phosphorus and calcium
levels in female rats was observed. A LOAEL of 2110 mg/kg bw/day was estimated. A
NOAEL was identified as 211mg/kg bw/day. A moderate hazard was designated
conservatively (ACC 2004).
In a 2-generation oral reproductive toxicity study in rats, a NOAEL of 50 mg/kg bw/day was
estimated based on reduced body weight or body weight gain during premating period in
parental F0 and F1 female rats dosed with 165 mg/kg bw/day of the same commercial
mixture. LOAEL was determined to be 165 mg/kg bw/day (Chemtura 2006).
In a repeated-dose oral toxicity study submitted by REACH registration applicant, three
groups of ten male and ten female rats received BEH-TEBP (Tradename FR-45B) by diet at
200, 2000 or 20000 ppm concentrations (= approx. 21.97, 223.4 or 2331 mg/kg bw/day) for
four weeks. A similar control group received no treatment. A positive control group with five
males and five female rats received di-2-ethyl hexyl phthalate (DEHP) by diet at
concentration 15000 ppm for four weeks. Dietary administration of BEH-TEBP to rats at the
highest dose of 20000 ppm produced only minor changes, namely, there was a slightly lower
48
overall bodyweight gain for female rats. Marginally low alanine amino-transferase activities
were also observed in females receiving the highest dose. Marginally low phosphorus
concentrations were seen in all females and males receiving the highest dose. No evidence of
toxicity was observed at 200 or 2000 ppm. BEH-TEBP did not cause any toxicity in the liver
or testes observed in positive controls given DEHP, namely peroxisome proliferation. The
NOAEL was conservatively estimated at 2000 ppm, i.e. 223.4 mg/kg bw/day for the above-
described observed effect. This study and NOAEL has been used by the REACH registrant
and ECHA to estimate the DNEL for long-term oral exposure.
49
4 EXPOSURE ASSESSMENT
4.1 Available Relevant Exposure Information
Ali et al. (2011) estimated exposure to FRs via ingestion of indoor dust by assuming 100%
absorption of intake, as it was also done in a previous study by Jones-Otazo et al. 2005. The
amount of dust ingestion was assumed to be of an average 50 mg per day for toddlers and 20
mg per day for adults. For high dust ingestion, 200 mg per day was assumed for toddlers and
50 mg per day for adults. “Low”, “Typical” and “High” exposure scenarios for each
microenvironment were estimated using 5th percentile, median, and 95th percentile
concentrations in the dust samples. Overall exposure to FR by dust ingestion was calculated
according to the estimated relative time spent in each microenvironment for toddlers and
adults.
In the Ali study, in a typical exposure scenario, exposure to BTBPE was 0.01 ng/kg bw/day
for mean dust ingestion and 0.05 ng/kg bw/day for high dust ingestion for toddlers at the
median. For non-working adult, the typical exposure was 0.02 ng/kg bw/day for mean dust
ingestion and 0.06 ng/kg bw/day for high dust ingestion at the 95th percentile (below LOQ at
median). For working adult, the typical exposure was 0.04 ng/kg bw/day for mean dust
ingestion and 0.08 ng/kg bw/day for high dust ingestion at the 95th percentile (below LOQ at
median).
Exposure to EH-TBB was 0.02 ng/kg bw/day for mean dust ingestion and 0.08 ng/kg bw/day
for high dust ingestion for toddlers at the median. For non-working adult, the typical exposure
was 0.02 ng/kg bw/day for mean dust ingestion and 0.05 ng/kg bw/day for high dust ingestion
at the 95th percentile (below LOQ at median). For working adult, the typical exposure was
also 0.02 ng/kg bw/day for mean dust ingestion and 0.05 ng/kg bw/day for high dust ingestion
at the 95th percentile (below LOQ at median).
Exposure to BEH-TEBP was 0.10 ng/kg bw/day for mean dust ingestion and 0.40 ng/kg
bw/day for high dust ingestion for toddlers at the median. For non-working adult, the typical
exposure was 0.13 ng/kg bw/day for mean dust ingestion and 0.32 ng/kg bw/day for high dust
ingestion at the 95th percentile (below LOQ at median). For working adult, the typical
50
exposure was 0.01 ng/kg bw/day for mean dust ingestion and 0.02 ng/kg bw/day for high dust
ingestion at the median.
4.2 Brominated Flame Retardants Measured in Children’s Room in Kuopio, Finland
The Institute for Health and Welfare investigated the levels of novel and emerging flame
retardants in the indoor dust of 40 children’s room in the city of Kuopio, Finland. This study
has not been published. This section presents a part of the result from this study. The levels of
* (ng/g) = (µg/kg) [a] Ali, N. et al. 2011. ‘Novel’’ brominated flame retardants in Belgian and UK indoor dust: Implications for human exposure. Chemosphere 83 (2011) 1360-1365.
[b] Cristale,J.; Lacorte,S. 2013. Development and validation of a multiresidue method for the analysis of polybrominated diphenyl ethers, new brominated and organophosphorus flame retardants in sediment, sludge and
dust. J. Chromatogr. A. 1305 (2013) 267-275.
[c] Sahlström et al. 2012. Clean-up method for determination of established and emerging brominated flame retardants in dust. Anal. Bioana. Chem. 404 (2912) 459.
[d] Stapleton H.M. et al. 2008. Alternate and new brominated flame retardants detected in U.S. house dust. Enviro. Sci. Technol. 42 (2008) 6910. [e] Van den Eede, N. et al. 2011. Analytical developments and preliminary assessment of human exposure to organophosphate flame retardants from indoor dust. Environment International 37(2) 454-461.
[f] Van den Eede, N. et al. 2012. Multi-residue method for the determination of brominated and organophosphate flame retardants in indoor dust. Talanta 89 (2012) 292-300.
[g] Hoffman, K. et al. 2015. Monitoring Indoor Exposure to Organophosphate Flame Retardants: Hand Wipes and House Dust. Env. Health. Pespec. 123(2) 160-165.
[h] Bergh, C. et al. 2012. Organophosphate and phthalate esters in standard reference material 2585 organic contaminants in house dust. Anal. Bioanal. Chem. 402 (2012) 51-59. [i] Fan, X. et al. 2016. Non-PBDE halogenated flame retardants in Canadian indoor house dust: sampling, analysis, and occurrence. Environ Sci Pollut Res (In press).
[j] Ionas A.C.; Covaci A. 2013. Simplifying multi-residue analysis of flame retardants in indoor dust. Int J Environ Anal Chem 93:1074–1083.
[k] Ali N. et al. 2012. Assessment of human exposure to indoor organic contaminants via dust ingestion in Pakistan Indoor Air 22(3) 200-211.
69
APPENDIX 2: Criteria Used by USEPA to Assign Hazard Designation