Hexabromocyclododecanes and Tetrabromobisphenol-A in Indoor Air

Aug 7, 2008 - In cars, only the surface of the seats with which occupants would have direct contact (i.e., not including seat backs) was sampled for 2...
0 downloads 12 Views 111KB Size
Environ. Sci. Technol. 2008, 42, 6855–6861

Hexabromocyclododecanes and Tetrabromobisphenol-A in Indoor Air and Dust in Birmingham, UK: Implications for Human Exposure M O H A M E D A B O U - E L W A F A A B D A L L A H , * ,†,§ STUART HARRAD,† AND ADRIAN COVACI‡ Division of Environmental Health and Risk Management, School of Geography, Earth and Environmental Sciences, University of Birmingham, Birmingham, B15 2TT, United Kingdom, Toxicological Center, University of Antwerp, Universiteitsplein 12610 Wilrijk, Belgium, and Department of Analytical Chemistry, Faculty of Pharmacy, Assiut University, 71526 Assiut, Egypt

Received April 22, 2008. Revised manuscript received June 26, 2008. Accepted July 2, 2008.

Hexabromocyclododecanes (R-, β-, and γ-HBCDs) and tetrabromobisphenol-A (TBBP-A) were determined in indoor air from homes (n ) 33; median concentrations ΣHBCDs ) 180 pg m-3; TBBP-A ) 15 pg m-3), offices (n ) 25; 170; 11), public microenvironments (n ) 4; 900; 27) and outdoor air (n ) 5; 37; 1). HBCDs and TBBP-A were also determined in dust from homes (n ) 45; median concentrations ΣHBCDs ) 1300 ng g-1; TBBP-A ) 62 ng g-1), offices (n ) 28; 760; 36), cars (n ) 20; 13,000; 2), and public microenvironments (n ) 4; 2700; 230). While ΣHBCDs in car dust significantly exceeded (p < 0.05) those in homes and offices, TBBP-A in car dust was significantly lower (p < 0.05) than that in homes and offices. No significant differences were observed between ΣHBCDs and TBBP-A in air or dust from homes and offices. Compared to dietary and inhalation exposures, dust ingestion constitutes an important pathway of exposure to HBCDs and TBBP-A for the UK population. Specifically, using average dust ingestion rates and concentrations in dust, dust ingestion constitutes for adults 34% (TBBP-A) and 24% (HBCDs) of overall exposure, and for toddlers 90% (TBBP-A) and 63% (HBCDs). Inhalation appears a minor exposure pathway to both HBCDs and TBBPA. On average, dust is 33% R-, 11% β-, and 56% γ-HBCD, while air is 22% R-, 11% β-, and 65% γ-HBCD.

Introduction Brominated flame retardants (BFRs) constitute a diverse group of compounds used to prevent or minimize fire hazards. The most widely used BFRs are tetrabromobisphenol A (TBBP-A) with a global demand of 170,000 t in 2004, alongside decabromodiphenyl ether, hexabromocyclododecanes (HBCDs), and pentabromodiphenyl ether, for which the worldwide market demands in 2001 were respectively 56,100, 16,700, and 7,500 t (1). TBBP-A is used mainly as a reactive flame retardant covalently bonded to the polymer matrix in epoxy and * Corresponding author e-mail: [email protected]; tel: +44 121 414 5431; fax: +44 121 414 3078. † University of Birmingham. § Assiut University. ‡ University of Antwerp. 10.1021/es801110a CCC: $40.75

Published on Web 08/07/2008

 2008 American Chemical Society

polycarbonate resins used in printed circuit boards and electronic equipment. It can also be used as an additive, for instance in high-impact polystyrene (HIPS) and acrylonitrilebutadiene-styrene resins. Additive usage accounts for 10% of TBBP-A total applications. However, even when used as a reactive flame retardant, excessive nonpolymerized TBBP-A is always present which can be emitted to the environment (2). Due to its low water solubility (63 µg/L) and low vapor pressure (6.24 × 10-6 Pa), TBBP-A is likely to be associated with suspended particulate matter following release (3). It has been detected in several environmental compartments including air (4, 5), dust (6, 7), sewage sludge, sediment (8, 9), and human blood (10). TBBP-A has been identified as an endocrine disrupter. It also displays a high potency to bind to human transthyretin and is immunotoxic (11). However, the potential TBBP-A toxicity is mitigated to some extent by its estimated human half-life of 2.2 days (3). HBCDs are used as additives to expanded and extruded polystyrene foams for thermal insulation of buildings, backcoating of fabrics, and to a lesser extent in HIPS (1). The commercial formulations consist mainly of R-, β-, and γ-diastereomers with γ- predominant. HBCD has low water solubility (49, 15, and 2 µg/L for R-, β-, and γ-HBCD, respectively), a fairly low vapor pressure (6.27 × 10-5 Pa), and is persistent. It can therefore bioaccumulate and undergo long-range transport (12). Oral exposure to HBCDs may induce hepatic cytochrome P450 enzymes in rats and alter the normal uptake of neurotransmitters in rat brain. It can disrupt the thyroid hormone system and induce cancer through a nonmutagenic mechanism in humans (12). HBCD has been detected previously in outdoor air (13), dust (14, 15), sewage sludge, and sediment (9). We reported recently for the first time on diastereomerspecific concentrations of HBCDs in household dust from Canada, the UK, and the United States (15). To the authors’ knowledge, there have been no reports of diastereomerspecific HBCD concentrations in indoor air, save for those reported recently by our group in three office microenvironments monitored for the purposes of calibrating PUF disk passive samplers (16). In addition, very little is known about concentrations of HBCDs in outdoor air (13, 17). Information on the presence of TBBP-A in indoor air and dust is scarce. There is only one study of TBBP-A in dust from offices in the UK (6) and two studies of TBBP-A in indoor air from offices in Sweden (4, 18). Although we have previously estimated human exposure to ΣHBCDs via ingestion of household dust (15), there issto datesno report on human exposure to TBBP-A or HBCDs via air inhalation and dust ingestion that considers exposure in both domestic and nondomestic microenvironments, such as cars, offices, and public microenvironments (PMEs). Our recent report on HBCDs in indoor dust (15) also highlighted the possibility that the predominance of R-HBCD observed in humans (12) may not solely be due to in vivo biotransformation of β- and γ-HBCD (19), but, at least partly attributable to the diastereomer pattern in dust. Specifically, many dust samples displayed a marked shift from the predominance of γ-HBCD in the commercial formulations toward the R-diastereomer (15). The mechanism of this predominant shift was recently studied and explained (20). In light of the above, the aims of this study are to (1) report concentrations of the three main HBCD diastereomers and TBBP-A in air and dust from cars, homes, offices, and PMEs; (2) evaluate whether the previously observed shift in HBCD diastereomer pattern from predominantly γ-HBCD in the commercial formulation toward R-HBCD observed in VOL. 42, NO. 18, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6855

dust samples (15) is reflected in air samples; and (3) estimate exposure of UK adults and toddlers to HBCDs and TBBP-A via both inhalation and dust ingestion.

Materials and Methods Sampling Strategy. Air samples were collected between February and December 2007 in a total of 62 microenvironments within the West Midlands conurbation. Samples were taken under normal room use conditions to reflect actual human exposure. We selected the following microenvironment categories for study: homes (living rooms, n ) 33), offices (n ) 25), and PMEs (3 pubs and 1 restaurant). Outdoor air sampling (n ) 5) was performed in December 2007 at the Elms Road Observatory Site (EROS) in Birmingham, UK. Dust samples were collected between September 2006 and June 2007 from 97 microenvironments within the West Midlands conurbation and in Basingstoke, Hampshire. All microenvironments comprised a convenience sample of acquaintances of the authors. Samples were taken under normal room use conditions to reflect actual human exposure. The following microenvironment categories were selected for study: homes (living rooms, n ) 45 (concentrations of HBCDs in 31 of these samples have been reported previously in our study of international variations in contamination of household dust (15), but are included here to better facilitate study of differences in contamination between microenvironment categories)), offices (n ) 28), PMEs (3 pubs and 1 restaurant), and cars (n ) 20, further information in Table SI-1 in the Supporting Information). Sampling Methods. Indoor Dust Sampling. Dust samples were collected using a Nilfisk Sprint Plus 1600 W vacuum cleaner. Sampling was conducted according to a clearly defined standard protocol by one of the research team. In offices and homes, 1 m2 of carpet was vacuumed for 2 min and in case of bare floors 4 m2 was vacuumed for 4 min. In cars, only the surface of the seats with which occupants would have direct contact (i.e., not including seat backs) was sampled for 2 min. Samples were collected using nylon sample socks (25 µm pore size) that were mounted in the furniture attachment tube of the vacuum cleaner. After sampling, socks were closed with a twist tie, sealed in a plastic bag and stored at -20 °C. Before and after sampling, the furniture attachment was cleaned thoroughly using an isopropanol-impregnated disposable wipe. Indoor Air Sampling. Passive air samplers (i.e., polyurethane foam (PUF) disks) were used to sample HBCDs in indoor air. They were employed to provide a time-integrated sample over a 30 day sampling period. PUF disk samplers have been used successfully in studies monitoring PBDEs in indoor air (21, 22), and have been validated recently for monitoring HBCDs in indoor air (16). Further specific details of passive air sampling protocols can be found as Supporting Information. Active Air Sampling of Both Indoor and Outdoor Air. During method development, it became evident that, unlike HBCDs, airborne TBBP-A is present mainly in the particulate phase. This is in agreement with the findings of Sjodin et al. (4). Hence, PUF disk samplers which sample mainly the vapor phase are unsuitable for monitoring TBBP-A. Furthermore, the PUF disk passive sampling rates derived for monitoring HBCDs in indoor air (16) cannot be extrapolated to sampling outdoor air. For these reasons, low volume active air samples were taken for the purposes of monitoring concentrations of HBCDs in outdoor air and TBBP-A in both indoor and outdoor air. Further specific information on active air sampling protocols can be found as Supporting Information. Analytical Protocols. Sample Preparation and Extraction. Dust samples were passed through a 500 µm mesh size sieve, weighed accurately, and extracted using pressurized liquid 6856

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 18, 2008

extraction (Dionex Europe, UK, ASE 300). Dust samples (typically between 100 and 300 mg) were loaded into precleaned 66 mL cells containing 1.5 g of Florisil and Hydromatrix (Varian Inc., UK) to fill the void volume of the cells, spiked with 25 ng of each of 13C-labeled R-, β-, γ-HBCD and TBBP-A as internal (surrogate) standards (i.e., standards used for determination of analyte concentrations) and extracted with hexane/dichloromethane (1:9, v/v) at 90 °C and 1500 psi. The heating time was 5 min, static time 4 min, purge time 90 s, flush volume 50%, with three static cycles. PUF disks and filters were spiked separately with 25 ng of each of 13C-labeled R-, β-, γ-HBCD and TBBP-A as internal standards and Soxhlet extracted with hexane/CH2Cl2 (1:9, v/v) for 8 h. Cleanup. The crude extracts were concentrated to 0.5 mL using a Zymark Turbovap II then cleaned up by loading onto SPE cartridges filled with 8 g of precleaned acidified silica (44% concentrated sulfuric acid, w/w). The analytes were eluted with 25 mL of hexane/dichloromethane (1:1, v/v). The eluate was evaporated to dryness under a gentle stream of N2, then reconstituted in 200 µL of d18-γ-HBCD (25 pg µL-1 in methanol) as recovery determination (or syringe) standard, used to determine the recoveries of internal standards for QA/QC purposes. Analysis. Separation of R-, β-, and γ-HBCDs and TBBP-A was achieved using a dual pump Shimadzu LC-20AB Prominence liquid chromatograph equipped with a SIL-20A autosampler, a DGU-20A3 vacuum degasser, and a Varian Pursuit XRS3 C18 reversed phase analytical column (150 mm × 2 mm i.d., 3 µm particle size). A mobile phase program based upon (a) 1:1 methanol/water and (b) methanol at a flow rate of 150 µL min-1 was applied for elution of the target compounds; starting at 50% (b) then increased linearly to 100% (b) over 4 min, held for 7 min followed by a linear decrease to 60% (b) over 4 min, held for 1 min, and finishing with 100% (a) for 10 min. TBBP-A and the three HBCD diastereomers were baseline separated with retention times of 9.0, 10.6, 11.2, and 11.7 min for TBBP-A, R-, β-, and γ-HBCD, respectively. Mass spectrometric analysis was performed using a Sciex API 2000 triple quadrupole mass spectrometer operated in electrospray negative ionization mode. MS/MS detection operated in the MRM mode was used for quantitative determination based on m/z 640.6f79, m/z 652.4f79, and m/z 657.7f79 for the native, 13C-labeled, and d18-labeled HBCD diastereomers, respectively, and m/z 540.8f79, m/z 552.8f79 for the native and 13C-labeled TBBP-A, respectively. Specific instrumental calibration parameters are given in Table SI-2 in the Supporting Information. Quality Assurance/Quality Control. Good recoveries of both 13C-labeled internal standards and d18-R-HBCD employed as SES for air samples were obtained for all the studied compounds (Table SI-3 in the Supporting Information). No TBBP-A or HBCDs were detected in method blanks (n ) 10) for air samples consisting of a precleaned PUF disk (for passive samplers) and a filter and PUF plug (for active samplers) treated as a sample. Method quantification limits (MQLs) for TBBP-A and individual HBCD diastereomers were therefore governed by the S:N ratio. Based on a sampled air volume of 30 m3, this equates to 1.8 pg m-3 for TBBP-A and 3.3 pg m-3 for HBCDs. Similarly, TBBP-A or HBCDs were below MQLs in method blanks (n ) 25) for dust analysis. Detectable, but very low HBCD concentrations (typically 0.1-0.5 ng ΣHBCDs g-1) were obtained for field blanks (n ) 12). These consisted of sodium sulfate (0.2 g) “sampled” using the vacuum cleaner according to the standard protocol and treated as a sample. Concentrations in samples in each batch of 10 were thus corrected for the contamination detected in the associated field blank. MQLs for individual HBCD diastereomers were governed by

TABLE 1. Summary of HBCDs and TBBP-A Concentrations (pg m-3) in Air from the Studied Microenvironments r-HBCD

location

statistical parameter

homes, n ) 33, n ) 5 for TBBP-A

average standard deviation median minimum maximum

59 77 37 14 430

offices, n ) 25, n ) 5 for TBBP-A

average standard deviation median minimum maximum

public microenvironments, n ) 4

average standard deviation median minimum maximum

outdoor air, n ) 5

average standard deviation median minimum maximum

a

From ref 23.

b

γ-HBCD

Σ HBCDs

22 9 22 5 54

170 140 120 39 710

250 240 180 67 1300

16 5 15 9 22

52 61 24 4 250

43 20 36 18 87

24 6 23 14 34

120 68 110 43 370

180 90 170 70 460

16 12 11 4 33

170 280 71 10 1400

250 110 210 180 400

28 12 24 19 46

550 140 570 360 690

900 60 900 820 960

26 7 27 17 32

110 72 140 29 160

33 2 33 31 35

37 2 37 34 40

3.0 0.5 2.9 2.3 3.7

β-HBCD

1.1 0.1 1.0 0.9 1.2

TBBP-A

0.8 0.1 0.7 0.7 0.9

Σtrihexa-BDEs

a

21b

From ref 21.

the field blanks and were typically 0.1 ng g-1, while for TBBP-A the MQL was 0.05 ng g-1 based on S/N ratio of 10:1. In the absence of an appropriate standard reference material for TBBP-A and HBCDs, the accuracy and precision of the analytical method for HBCDs was assessed via replicate analysis (n ) 10) of SRM 2585. The results obtained compared favorably with the indicative values reported elsewhere (ref 23 and Table SI-4a). For TBBP-A, a standard addition or “matrix spike” method to SRM 2585 at 3 concentration levels (n ) 5 at each level) was used to assess the accuracy and precision of the method and good results were obtained (Table SI-4b). In samples where particularly elevated TBBP-A or ΣHBCDs concentrations were found (i.e., such that the internal standard “spiking” levels were inappropriately low) (n ) 25), a second aliquot of the dust sample in question was analyzed using a smaller quantity of dust and a higher amount of internal standards. The results of the second analysis are reported in this study. Statistical Analysis. Statistical analysis of the data was conducted using Excel (Microsoft Office 2003) and SPSS version 13.0. In all instances, where concentrations were below the MQL, concentrations were assumed to equal half the MQL. The distribution of each data set was evaluated using both the Kolmogorov-Smirnov test and visual inspection. The results revealed concentrations in all data sets to be log-normally distributed. Hence, further ANOVA and t tests were performed on log-transformed concentrations.

Results and Discussion Concentrations of HBCDs and TBBP-A in Air from Different Categories of Indoor Microenvironments. Table 1 summarizes the concentrations of R-, β-, and γ-HBCD and ΣHBCDs in air samples from homes, offices, and PMEs. These indoor concentrations are an order of magnitude higher than those detected in outdoor air in this study (average ) 37 pg ΣHBCDs m-3), and U.S. outdoor air (maximum ) 11 pg ΣHBCDs m-3) (13). Although there were an insufficient number of samples from PMEs to permit meaningful statistical interpretation, concentrations in PMEs sampled are substantially above those detected in homes and offices. No statistically significant difference (p > 0.05) was observed

between concentrations of ΣHBCDs in homes and offices. It is interesting that concentrations of ΣHBCDs reported here in air from each of the microenvironment categories studied are typically several times greater than the concentrations of Σtrihexa-BDEs (i.e., those congeners predominant in the Penta-BDE formulation) reported in an earlier study of the same microenvironment categories in Birmingham, UK (22). The concentrations of TBBP-A in air samples from homes, offices, and PMEs are summarized in Table 1. The data demonstrate concentrations in indoor air exceed by an order of magnitude those present in outdoor airsconsistent with the presence of indoor emission sources. However, the relatively small sample numbers preclude statistical analysis of differences in contamination between microenvironment categories. Unlike HBCDs, there are no substantial differences apparent between concentrations in PMEs and those in homes and offices. The concentrations of TBBP-A in this study agree with previous reports of concentrations in offices and lecture halls from Sweden (4), but are lower than those reported from offices and houses from Japan (24). Based on their relative production volumes alone, one would anticipate concentrations to fall in the order TBBP-A > HBCDs.PentaBDE. While differences in vapor pressures will influence the relative atmospheric abundance of these BFRs, it is noteworthy that concentrations of TBBP-A in both indoor and outdoor air are much lower than those of HBCDs and are similar to those of the trihexa-BDEs that constitute the PentaBDE formulation (22, 25). This may be attributable to the widespread use of TBBP-A as a reactive flame retardant which makes its release from treated goods less facile than for an additive flame retardant like HBCD. Concentrations of HBCDs and TBBP-A in Dust from Different Categories of Indoor Microenvironments. Table 2 summarizes the concentrations of HBCD diastereomers in dust samples from cars, homes, offices, and PMEs. Of particular interest is the very high concentration of 140,000 ng ΣHBCDs g-1 (29% R-, 18% β-, and 53% γ-HBCD) detected in one UK house dust sample, which is the highest HBCD level reported to date. We are currently unable to provide an explanation for this high concentration based on a survey of potential BFR-treated items present in the room sampled. Statistical analysis reveals that concentrations of ΣHBCDs in VOL. 42, NO. 18, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6857

TABLE 2. Summary of HBCDs and TBBP-A Concentrations (ng g-1) in Dust from the Studied Microenvironments location (reference) homes, n ) 45, for TBBP-A n ) 35, nondetects ) 1

r-HBCD

β-HBCD

γ-HBCD

Σ HBCDs

3200

1000

4200

8300

11000 380 22 66000

3900 93 9 26000

13000 670 70 75000

26000 1300 140 140000

610

210

760

standard deviation median minimum maximum

780 220 15 2900

300 84 11 1300

average

3200

standard deviation median minimum maximum

statistical parameter average

cars, n ) 20, for TBBP-A nondetects ) 10

public microenvironments, n)4

a

From ref 25.

b

9

BDE-209a

77

260000

71 62