Environ. Sci. Technol. 2008, 42, 4222–4228
Linking PBDEs in House Dust to Consumer Products using X-ray Fluorescence J O S E P H G . A L L E N , * ,‡,† MICHAEL D. MCCLEAN,‡ HEATHER M. STAPLETON,§ AND THOMAS F. WEBSTER‡ Environmental Health & Engineering, Inc., 117 Fourth Avenue, Needham, Massachusetts 02494, Boston University School of Public Health, Department of Environmental Health, 715 Albany Street, Boston, Massachusetts 02118, and Duke University, Nicholas School of the Environment & Earth Sciences, Durham, North Carolina 27708
Received November 27, 2007. Revised manuscript received March 4, 2008. Accepted March 18, 2008.
The indoor environment is an important source of exposure to polybrominated diphenyl ethers (PBDEs), a class of fire retardants used in many household products. Previous attempts to link PBDE concentrations in house dust to consumer products have been hampered by the inability to determine the presence of PBDEs in otherwise similar products. We used a portable X-ray fluorescence (XRF) analyzer to nondestructively quantify bromine concentrations in consumer goods. In the validation phase, XRF-measured bromine was highly correlated with GC/MS-measured bromine for furniture foam and plastic from electronics (n ) 29, r ) 0.93, p < 0.0001). In the field study phase, the XRF-measured bromine in room furniture was associated with pentaBDE concentrations in room dust in the bedroom (r ) 0.68, p ) 0.001) and main living area (r ) 0.51, p ) 0.02). We also found an association between XRFmeasured bromine levels in electronics and decaBDE levels in dust, largely driven by the high levels in televisions (r ) 0.64, p ) 0.003 for bedrooms). For the main living area, predicting decaBDE in dust improved when we included an interaction effect between the bromine content of televisions and the number of persons in the house (p < 0.005), a potential surrogate for television usage.
Introduction Polybrominated diphenyl ethers (PBDEs) are commonly used as flame retardants in household consumer products such as electronics, furniture containing polyurethane foam (PUF), certain fabrics, and other products (1, 2). There are three commercial products (penta, octa, deca) and though penta and octa are no longer produced, large quantities are still present in consumer products. PBDE concentrations in the indoor environment exceed outdoor concentrations (3, 4). Animal studies have shown that PBDEs are neurotoxic and endocrine disruptors (5–8) and one of the few human studies conducted to date found an association between PBDE body * Corresponding author phone: 800-825-5343; fax: 781-247-4305; e-mail:
[email protected] † Environmental Health & Engineering, Inc. ‡ Boston University School of Public Health. § Duke University. 4222
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 11, 2008
burdens and cryptorchidism (9). To minimize indoor exposure to these compounds, identification of sources and their relationship to exposure pathways must be understood. In principle, it should be possible to link PBDE concentrations in air and dust samples collected from indoor spaces to the consumer products in those spaces. However, the attempts of several groups to do sosrelying primarily on counts of electronics and PUF-containing furnitureshave resulted in only modest success (10–16). We hypothesized that counts may be insufficient if there are large differences in PBDE concentrations between otherwise similar objects; the resulting misclassification error would tend to make associations harder to detect (17). While a better method of characterizing potential PBDE sources is needed, it is generally not feasible to “biopsy” consumer goods from homes and analyze them for PBDEs using conventional methods. We have therefore employed an innovative approach for obtaining a surrogate PBDE measure: testing products for bromine using X-ray fluorescence (XRF). Portable XRF analyzers have been used for years to nondestructively test for the presence of lead in homes, analyze museum artifacts for identification and conservation, test for lead in contaminated soils and air filters, and assist in crime-scene investigations including forensic odontology and analysis of gunshot residues (18–23). XRF technology has expanded in recent years, due in part to two European Union Directives: Restriction of Hazardous Substances and Waste Electrical and Electronic Equipment (RoHS/WEEE) (24, 25). XRF technology illuminates a sample with high-energy photons generated by a low-power X-ray source. Upon hitting an atom, photons dislodge electrons in inner orbitals. The vacancy is filled by outer orbital electrons which then release a fluorescent X-ray pattern which is unique by element. By measuring the scattered X-rays the XRF can estimate density and calculate a concentration. Accordingly, the primary objectives of our study were to (1) determine whether bromine concentrations determined using the XRF analyzer provide a valid method of estimating the PBDE content of consumer products, (2) characterize the bromine content of different types of consumer products, and (3) evaluate the relationship between the bromine content of consumer products and PBDEs in house dust.
Materials and Methods The study was conducted in two phases: a laboratory-based validation of the XRF method and a field investigation of PBDEs using the XRF in residential settings. Validation Phase. The validation phase included an assessment of foam collected from chairs and plastic samples obtained from the housings of televisions and computer products. A sample of at least one square inch was removed from each product for testing. A portable XRF analyzer and test stand (Innov-X Systems) were used to obtain multiple ten-second bromine measurements of each sample (typically ten readings per sample). The samples were then analyzed via gas chromatography/mass spectrometry (GC/MS) by laboratory staff who were blinded to the results from the XRF tests. The foam samples were weighed and then extracted using accelerated solvent extraction (ASE) with dichloromethane. Prior to extraction, the ASE cells containing the foam and sodium sulfate (as a filler) were spiked with two labeled recovery standards: 13C-labeled 2,2′,3,4,5-pentachlorodiphenyl ether (13CDE-86) and 13C-labeled 2,2′,3,3′,4,4′,5,5′,6,6′10.1021/es702964a CCC: $40.75
2008 American Chemical Society
Published on Web 04/30/2008
decabromodiphenyl ether (13C BDE 209), both supplied by Cambridge Isotope Laboratories (Andover, MA). Conditions for the ASE extraction were as follows: 5 min heating to 100 °C, pressurized to 2000 psi and flushed with 60% volume in a cycle that repeated three times. Recovered extracts were reduced in volume to 1.0 mL using rapid nitrogen evaporation systems (Turvo Vap) and filtered through a 25 mm syringe filter equipped with a 0.45 µm PTFE membrane (VWR, Atlanta GA). The plastic samples from electronics casings were extracted by completely dissolving the polymer. A small crosssectional piece of plastic was cut using a razor blade and its mass was recorded (approximately 0.1 g). The plastic piece was transferred to a 250 mL volumetric flask. Following addition of toluene (250 mL), the flask was wrapped in aluminum foil (to prevent any degradation of the BDE 209) and the contents were stirred for 24 h using a Teflon magnetic stirrer. Afterward a 1 mL aliquot was transferred to a small 4 mL vial spiked with 50 ng of 13C BDE 209. The extracts from both the plastic and foam samples were treated in a similar manner to prepare for GC/MS analysis. To remove coextracted polymers the extracts were injected into an HPLC unit equipped with a size exclusion chromatography column (divinylbenzene-polystyrene column, 10 µm particle size, 100 Å pore size, 2.5 cm i.d. × 60 cm, PL-Gel, Polymer Laboratories, Inc., Amherst, MA). Samples were eluted through the column using a mobile phase of 100% dichloromethane at a flow rate of 5.0 mL/min. The collected eluant was concentrated in volume and transferred into hexane. Samples were further cleaned using silica plus solid phase extraction cartridges (Waters, Milford, CT) and eluted with 20 mL of hexane. All extracts were then reduced in volume using the automated evaporation system, transferred to a GC auto sampler vial, spiked with 50 ng of 4′-fluoro2,3′,4,6-hexabromodiphenyl ether (F-BDE 69, Chiron Inc., Trondheim, Norway) and analyzed for PBDEs, tetrabromobisphenol A (TBBPA), and bistribromophenoxyethane (BTBPE) using gas chromatography electron impact negative chemical ionization (GC/ECNI-MS) as previously described (26). Standards for TBBPA and BTBPE were purchased from Wellington Laboratories (Guelph, ON). Based on the composition of detected compounds, we computed the concentration of “GC/MS-measured bromine” in the samples. Field Investigation Phase. The field investigation included collection of dust samples from 19 residences in the Boston area (Massachusetts) from October to November 2006. All protocols were reviewed and approved by the Boston University Institutional Review Board, and informed consent was obtained prior to participation. Two dust samples were collected from each homesone from the bedroom and one from the main living areasusing a Eureka Mighty-Mite canister vacuum cleaner and a crevice tool fitted with a cellulose extraction thimble (Whatman International) as previously described (26, 27). The main living area was defined as the room where participants spent the majority of their waking hours (not including the kitchen). The bedroom and main living area were selected as the two rooms that (a) would likely have the largest number of PBDE sources and (b) where participants would likely spend the majority of their time. Dust samples were sieved (
