Science of the Total Environment 454–455 (2013) 534–541

Contents lists available at SciVerse ScienceDirect

Science of the Total Environment

j ourna l homepage: www.e lsev ie r .com/ locate /sc i totenv

Seasonal variations in the atmospheric concentrations of polybrominated diphenylethers in Kuwait

Bondi Gevao ⁎, Abdul Nabi Ghadban, Massimiliano Porcelli, Lulwa Ali, Amal Rashdan, Majed Al-Bahloul,Khalid Matrouk, Jamal ZafarDepartment of Environmental Sciences, Environment and Urban Development Division, Kuwait Institute for Scientific Research, P. O. Box 24885, 13109 Safat, Kuwait

H I G H L I G H T S

• Fortnightly resolved PBDE concentrations are higher at urban than remote sites in Kuwait.• Urban–rural differences in PBDE concentrations are due to ongoing primary emissions.• Temperature driven air-surface exchange is not a dominant factor in measured levels.• Seasonal variations at urban location are masked by ongoing primary sources.

⁎ Corresponding author. Tel.: +965 24989178; fax: +E-mail address: bgevao@kisr.edu.kw (B. Gevao).

0048-9697/$ – see front matter © 2013 Elsevier B.V. Allhttp://dx.doi.org/10.1016/j.scitotenv.2013.02.073

a b s t r a c t

a r t i c l e i n f o

Article history:Received 31 December 2012Received in revised form 24 February 2013Accepted 24 February 2013Available online 9 April 2013

Keywords:PBDEsAtmospheric pollutionPOPsSeasonal variationsSemivolatile organic compoundsLong-range atmospheric transport

The study reports fortnightly atmospheric concentrations of PBDEs concomitantly measured at an urban anda remote location over a twelve-month period in Kuwait to examine seasonal variability and urban–ruralconcentration gradients. The annual mean (and range) of ∑PBDE concentrations was 32 (3–208) pg m−3

at the remote site and 57 (0.3–445) pg m−3 at the urban site. Although not statistically significant, the median(29 pg m−3) and mean (57 pg m−3) concentrations at the urban location were higher than those measured atthe remote location (18 and 29 pg m−3 respectively), consistent with the view that urban centers are an impor-tant net source of these compounds to the environment. Although Clausius–Clapeyron plots showed statisticallysignificant correlations (p b 0.05) with temperature for low molecular weight congeners (BDEs 28, 47, 100),correlations with the ΣPBDE concentrations were not significant at both urban and remote sites. The seasonalvariations in ΣPBDE concentrations were not markedly different at the urban location, but the median summerΣPBDE concentration at the remote location was significantly higher than winter median ΣPBDE concentrations.The absence in seasonality at the urban location may be due to ongoing primary emissions in urban areas.

© 2013 Elsevier B.V. All rights reserved.

1. Introduction

Polybrominated diphenyl ethers (PBDEs) are semivolatile organiccompounds (SOCs) with ubiquitous environmental distribution. Theyweremanufactured at three levels of bromination as technical mixtures:the penta-; octa-; and deca-formulations with each mixture having aunique set of congeners in differing proportions. The penta- mixture isprimarily used in polyurethane foamused in couches, chairs, automobileseats etc. The octa-mixture is used primarily in plastics used for electron-ic circuit boards or small appliances, while the deca-mixture is used intextile backings, wire insulation, electronic enclosures, carpet backings,and plastics (Allen et al., 2008; Batterman et al., 2009; Gearhart and

965 24989089.

rights reserved.

Posselt, 2006). They are highly hydrophobic, bioaccumulative (Burreauet al., 2004), and have the propensity to enter the gas phase at ambientconditions and undergo long range atmospheric transport (Gouin andHarner, 2003; Wania and Dugani, 2003). PBDEs are similar in molecularstructure to severalwell-knownpersistent organic pollutants (POPs) likepolychlorinated biphenyls (PCBs), polychlorinated dibenzo-p-dioxinsand furans (PCDD/Fs) andhave been shown tohave similar environmen-tal properties. As is the case with these legacy POPs, PBDEs have alsobeen measured at remote sites, such as the Arctic, where they hadnever been used (Gallego et al., 2007). These concerns led to the inclu-sion of the penta- and octa-technical mixtures in Annex A (elimination)of the convention (SC-4/14 and SC-4/18) at the meeting of the confer-ence of parties (COP) in May 2009 (UNEP, 2010). Due to the fact thatthese chemicals are trans-boundary pollutants and undergo long rangetransport (Beyer et al., 2000; Ockenden and Jones, 1999), environmentaldata are needed from all regions of the globe to better understand thekey processes that influence their global distribution. Research in recent

535B. Gevao et al. / Science of the Total Environment 454–455 (2013) 534–541

years on POPs have focused on understanding their ambient sources(Gevao et al., 2010; Jin et al., 2011;Wang et al., 2010, 2012), atmospherictransport (Meng et al., 2011; Pozo et al., 2009; vonWaldow et al., 2010;Wang et al., 2009), human exposure (Gonzalez-Mille et al., 2010; Harleyet al., 2010; Herbstman et al., 2010; Toms et al., 2009) and fate (Arnold etal., 2008; Lohmann et al., 2007; Shin et al., 2010; Zhang et al., 2009).

PBDEs and other SOCs are known to fluctuate on a diurnal, seasonaland annual basis, mostly driven by emissions from primary andsecondary sources, abiotic removal (e.g. photolysis, OH radical removal)from the atmosphere (Iacovidou et al., 2009; Mandalakis et al., 2003;Raff and Hites, 2006, 2007; Schenker et al., 2008; Soderstrom et al.,2004), advection of contaminated air from source regions (Chi et al.,2008), and local meteorology (e.g. wind speed, temperature, humidity).It iswidely reported that the concentrations of PBDEs are higher at urbanlocations compared to remote locations as a result of their ongoing pri-mary emissions from urban centers, especially from indoor environ-ments where concentrations are known to be higher due to the use ofthe compounds as flame retardants in consumer products.

Until recently, there was a dearth of information on persistentorganic pollutants (POPs) in the Middle East, and in Kuwait in particu-lar, especially for emerging POPs like PBDEs. PBDEs have recently beenreported in various environmental media in Kuwait including indoorair/dust (Gevao et al., 2006a, 2006b), ambient air (Gevao et al., 2006c,2010), sediments (Gevao et al., 2006d), biota (Gevao et al., 2011b), sew-age sludge (Gevao et al., 2008) and in soil (Gevao et al., 2011a) along anurban rural transect. In this study, we report on the seasonal changes inthe concentrations of PBDEs at an urban and remote site in Kuwait andassess the factors that mediate these fluctuations.

2. Materials and methods

All solvents used in this workwere of analytical grade and purchasedthrough VWR Scientific (USA). Silica (Baker, 100–200 mesh), Aluminaand Sodium Sulfate (Baker) were purchased through VWR Scientific(USA). The PBDE analytical standards (calibration, surrogate and inter-nal)were purchased fromCambridge Isotope Laboratories (CIL, Andover,MA, USA). The PUF plugs were certified as flame retardant free andpurchased from Tisch Environmental (OH, USA). The PUF plugs werecleaned for 48 h using dichloromethane in a giant Soxhlet, with thesolvent replaced after 24 h. The pre-extracted PUF plugs were dried ina clean desiccator under vacuum and stored in solvent rinsed amberglass jars linedwith solvent rinsed aluminum foil to avoid contaminationduring storage. Glass fiber filters (GFFs) used to collect the particulatephase were loosely wrapped with aluminum foil and baked in a mufflefurnace at 450 °C for 12 h to remove organic residues. They wereallowed to cool in a vacuum desiccator and stored in solvent-rinsed alu-minum foil until needed.

2.1. Sampling

Two sites, an urban and a rural site, were selected for collecting airusing modified high volume air samplers (Tisch Environmental, Inc).The urban site was located on the premises of the Kuwait Institute forScientific Research (29° 20.227 N; 47° 54.208 E) close to the meteorolog-ical station. The remote site (29° 58.809N; 47° 40.171 E)was on a farm inAbdali, close to the border with Iraq. This site is agricultural, comprisingapproximately 800 farms in a total area estimated at 20,000 ha used togrow vegetables (primarily tomatoes, cucumbers, eggplants, and sweetpeppers in green houses) and date palm treeswith a lowpopulation den-sity comprisingmainly of farmworkers and shepherds. The land is gener-ally flat with the soil consisting of 80–90% sand with very low organiccarbon content generally between 0.4 and 2.5% (Gevao et al., 2011a).The samplerswere stationed approximately 50 m from thenearest build-ing to prevent contaminated indoor air from influencing the results.Twenty-four-hour air samples were collected fortnightly betweenNovember 2008 and February 2010 to assess seasonal fluctuations in

ambient concentrations of POPs. The samplers were located about1.5 m above ground. Over each sampling period, approximately 900 ±50 m3of airwas pumped through aWhatmanGFF (8″ × 10″) to trappar-ticulate matter and the gas phase compounds trapped on two PUF plugs(85 mm in diameter and 70 mm in length) located downstream of theGFF. Exact air volumes were determined for each sample with the useof a calibrated Magnehelic gauge (Tisch Environmental, Inc) which mea-sured pressure at the start and end of each sampling period. The averageof the two readings was used to calculate the air sampled over that mea-surement period. Upon retrieval, the PUF and filter samples were storedin separate cleaned amber glass jars and kept at−15 °C until extractionto minimize losses by photolysis and/or volatilization. The GFF wasweighed using a micro-balance before and after deployment to deter-mine the total suspended particulate (TSP) concentration over that sam-pling period, which is nominally defined as the difference between thetwo measurements. Field blanks, collected fortnightly, consist of a PUFand filter assembled in the sampler, immediately removed and processedin an identical manner to actual samples.

2.2. Extraction and clean-up

PUF and filter samples were extracted separately in a Soxhlet ap-paratus using 1:1 v/v mixture of DCM:hexane. Prior to extraction,the samples were spiked with BDEs 35, 181 and 13C-209 to monitoranalytical recoveries. The sample extracts were reduced in volumeon a Turbovap® II concentration workstation (Hopkinton, MA, USA),solvent exchanged to hexane, and interfering compounds removedby column chromatography (i. d. 9 mm) using 2 g of silica gel(Merck, 60–230 mesh) and 1 g alumina (BDH, neutral alumina). Thesilica gel used throughout this work was pre-extracted with hexanefor 24 h, changing solvents every 12 h, dried under vacuum at 90bars of pressure and stored in solvent-rinsed glass jars lined withsolvent-rinsed aluminum foil. Prior to use, the silica gel was activatedat 130 °C for 3 h. The alumina and sodium sulfate were baked at450 °C and stored in solvent-rinsed glass jars until ready for use.The sorbents (silica gel then alumina) were slurry packed in hexaneand capped with approximately 0.2 g anhydrous Na2SO4 to preventthe column from contact with air. The sample extract was spiked ontop of the column and the target compounds were eluted with40 ml 1:1 mixture of hexane:DCM. Dodecane (50 μl) was added tothe eluent to ensure that the samples did not dry out during solventreduction under a gentle nitrogen stream. The samples were thentransferred to 100-μl glass inserts, and spiked with Mirex (10 μl of10 ng/μl) as an internal standard, used for volume correction and toadjust for variations in instrument response.

2.3. Instrumental analysis

The sample extracts were analyzed on a Shimadzu GC 2010(Shimadzu, Tokyo, Japan) gas chromatograph using splitless injectionon a 15 m DB 5 ms column (0.25 mm i.d., 0.25 μm film thickness) andhelium as carrier gas. The oven program was 80 °C for 1 min, rampedat 12 °C min−1 to 155 °C, 4 °C min−1 to 215 °C, and further rampedat 3 °C min−1 to 300 °C and held for 10 min. This gas chromatographwas coupled to a Shimadzu 2010 Mass Selective Detector operated inelectron capture negative chemical ionization (ECNCI) mode usingselected ion monitoring (SIM), with methane as reagent gas. Theions m/z 79 and 81 were monitored for PBDEs except for BDE 209and 13C-209, for which ions m/z 487 and 495 were monitored respec-tively, whereas the ions 402/404 were monitored for Mirex. Operatingconditions were as follows: injector temperature was set at 250 °C;ion source 230 °C; quadrupole 106 °C; transfer line 300 °C. Identifica-tion and quantification were carried out against 5 calibration standardsof known concentration. A peakwaspositively identified if itwaswithin±0.05 min of the retention time in the calibration standard and

536 B. Gevao et al. / Science of the Total Environment 454–455 (2013) 534–541

quantifiedonly if the S/N ≥ 3, and the ratio of the target ion to its qualifierion was within ±20% of the standard value.

2.4. QA/QC

Strict quality-assurance and quality-control measures were put inplace to monitor the analytical process. Laboratory and field blankscomprised of pre-extracted PUF plugs and baked filters were treatedas samples. There was no discernible difference between laboratoryand field blanks, indicating minimal contamination during transport,storage and analysis. The analytes present in the appropriate blankswere subtracted from those in the sample extracts. The method de-tection limits (MDLs) were calculated as the mean blank + 3 × SD.These were converted to concentrations by dividing by the averageair volume. The detection limits varied from 0.1 to 1.6 pg m−3 forPBDEs obtained from the results of the analysis of field blanks. Aver-age recoveries (%) for surrogates spiked in samples for all sampleswere between 70 (± 12 SD) for BDE 35, 84 (± 8 SD) for BDE 181,and 76 (± 15 SD) for 13C-209. External recoveries for the entire methodwere carried out by spiking six PUF plugs with a working standardcontaining all PBDE congeners. The extraction and work-up procedureswere similar to those for actual samples. The recoveries were found tobe 90 ± 12% for all congeners. Since there are no reference materialsfor air, the accuracy and precision of the analytical method was assessedby replicate analyses (n = 6) of a certified indoor dust referencematerial(SRM 2585). The results compared very well with the certified values ofall the congeners (Gevao et al., 2010).

3. Results and discussion

The major congeners detected in the vast majority of samples ana-lyzed were: BDE 28 (2, 4, 4′-TriBDE), BDE 47 (2, 2′, 4, 4′-TetraBDE),BDE 99 (2, 2′, 4, 4′, 5-PentaBDE) and BDE 100 (2, 2′, 4, 4′,6-PentaBDE), BDE-153 (2, 2′, 4, 4′, 5, 5′-HexaBDE), BDE-154 (2, 2′, 4,4′, 5, 6′-HexaBDE), and BDE-183 (2, 2′, 3, 4, 4′, 5′, 6-HeptaBDE). BDEs17, 71, 66, 138 and 190 were sometimes present close to the detectionlimits of the method. Due to the erratic concentration of BDE 209,thishas also been removed from further analysis. ∑PBDEs refers to the

0

50

100

150

200

250

300

350

400

450

500

5/10

/200

812

/10/

2008

19/1

0/08

26/1

0/08

2/11

/200

810

/11/

2008

17/1

1/08

23/1

1/08

30/1

1/08

14/1

2/08

21/1

2/08

4/1/

2009

12/1

/200

918

/1/0

925

/01/

092/

2/20

098/

2/20

0922

/02/

098/

3/20

0922

/03/

09

Con

cent

rati

on (

pg m

-3)

Samplin

PBDEs-Rural Site

PBDEs-Urban Site

Temperature

Fig. 1. ΣPBDE concentrations (pg m−3) at the urban site in

sum of BDEs 28, 47, 99, 100, 154, 153 and 183 which are the dominantcongeners in penta technical mixtures (La Guardia et al., 2006).

The concentrations of ΣPBDEs at both urban and remote locationstogether with the temperature on the day of sampling are given inFig. 1 whereas Table 1 summarizes congener specific concentrationsin both gas and particle-phase fractions. The annual mean (and range)of ∑PBDE concentrations was 32 (3–208) pg m−3 at the remote siteand 57 (0.3–445) pg m−3 at the urban site. The concentrations mea-sured in this study are broadly consistent with those reported in levelsat semi-rural and some urban cities in other parts of the world. Forexample, concentrations of 2.8–37 pg m−3 have been reported forHazelrigg, a semi-rural site in the UK, 0.2–5 pg m−3 at Maze Hill,Ireland, 3.4–33 pg m−3 at Chilton, UK (Lee et al., 2004), 1–8 pg m−3

in Sweden (de Wit, 2002), 54–120 pg m−3 in Davis, California (Cahillet al., 2007); 40–70 pg m−3 in Chicago, USA (Chen et al., 2009); 41.5–256.8 pg m−3 in Guangzhou China (Chen et al., 2006) and 0.48–52 pg m−3 in Northern China (Wang et al., 2012) using high volumesamplers and concentrations of 1.3–6.7 pg m−3 along an urban ruraltransect in Birmingham, UK (Harrad and Hunter, 2004), 0.06–43 pg m−3 across Europe (Jaward et al., 2004), b0.1–4.4 pg m−3 forOttawa, Canada (Wilford et al., 2004) estimated from polyurethanefoam passive sampling.

The box plot in Fig. 2 summarizes the concentrations measuredthroughout the year and allows for easy visualization of the urbanrural differences in concentrations. The box contains the middle 50%of the data, whereas the top and bottom end of the box representthe 75th and 25th percentiles of the data set, respectively. The exten-sions (“whiskers”) at either end of the box indicate the 99 and 1 percen-tile and the outliers represent individual data that exceed 1.5 times theinterquartile range. The median concentrations are indicated by thehorizontal line whereas the mean concentrations are depicted by thered bars. In general, the size of the box and the length of the whiskersare an indicator of the variability in concentrations at a given site for agiven compound. A small box shows that the distribution is uniformover the entire sampling period and vice versa. The concentrations atthe remote site were fairly identical throughout the year judging fromthe spread of the data. It can be seen from Fig. 2 that half of the dataranged from 5 to 25 pg m−3 at the remote location and from 10 to

0

5

10

15

20

25

30

35

40

45

5/4/

2009

19/0

4/09

3/5/

2009

17/0

5/09

31/0

5/09

14/0

6/09

28/0

6/09

12/7

/200

926

/07/

099/

8/20

0923

/08/

096/

9/20

0928

/09/

0911

/10/

2009

25/1

0/09

6/12

/200

921

/12/

094/

1/20

1017

/01/

102/

2/20

10

Tem

pera

ture

(°°C

)

g Date

Kuwait City and a remote site close to the Iraqi border.

Table 1Summary of PBDE Congener concentrations [mean ± SD (range)] for fornightly samples collected between October 2008 and February 2010 at a remote and urban location inKuwait.

Rural location

Congener Gas phase concentration Particle phase concentration Total (gas + particle) % on filter (observed) % on filter (model prediction)

28 1.1 ± 1.4 (bd.l.−8.6) 0.3 ± 1.1 (bd. l.−6.9) 1.4 ± 1.6 (bd. l.−8.6) 16 ± 22 2.4 ± 2.347 11 ± 11 (0.5–46) 2.3 ± 3.4 (0.2–16) 13 ± 11 (1.4–47) 20 ± 21 22 ± 18100 2.9 ± 8.2 (0.1–51) 2.0 ± 9.6 (0.1–59) 4.8 ± 17 (0.3–110) 30 ± 21 54 ± 2899 8.7 ± 14 (0.4–67) 1.4 ± 2.1 (0.2–13) 10 ± 14 (0.8–68) 24 ± 21 62 ± 28154 0.7 ± 1.1 (bd. l.−5.9) 0.2 ± 0.2 (bd. l.−1.1) 0.9 ± 1.1 (bd. l.−6.3) 30 ± 26 82 ± 19153 0.8 ± 1.2 (bd.l.−6.3) 0.2 ± 0.2 (bd. l−1.3) 1.0 ± 1.3 (0.1–6.6) 26 ± 26 88 ± 14183 0.1 ± 0.3 (bd.l.−1.6) 0.1 ± 0.2 (bd. l−0.7) 0.3 ± 0.4 (bd. l.−2.0) 49 ± 37 96 ± 5.2Σ7PBDEs 26 ± 33 (1.4–143) 6.5 ± 12 (0.5–64) 32 ± 39 (2.9–208)Urban location

Congener Gas phase concentration Particle phase concentration Total (gas + particle) % on filter (observed)

28 2.3 ± 2.0 (bd. l.−11) 0.3 ± 0.4 (bd. l.−2.3) 2.4 ± 2.1 (bd. l.−11) 14 ± 1547 22 ± 28 (2.6–157) 2.1 ± 5.4 (0.1–35) 23 ± 28 (0.1–158) 13 ± 14100 4.6 ± 8.8 (0.3–40) 1.2 ± 4.8 (0.1–30) 5.6 ± 12 (0.1–67) 22 ± 1899 21 ± 40 (0.8–208) 2.1 ± 5.3 (0.1–34) 22 ± 39 (0.1–208) 22 ± 20154 1.5 ± 2.7 (bd. l.−15) 0.3 ± 0.5 (bd. l.−2.4) 1.7 ± 2.7 (bd. l.−15) 31 ± 28153 1.7 ± 3.5 (bd. l.−19) 0.3 ± 0.6 (bd. l.−2.7) 1.9 ± 3.4 (bd. l.−19) 33 ± 33183 0.1 ± 0.1 (bd. l.−0.5) 0.2 ± 0.3 (bd. l.−1.8) 0.3 ± 0.4 (bd. l.−1.8) 76 ± 31Σ7PBDEs 52 ± 82 (4.4–444) 6.6 ± 14 (0.3–83) 57 ± 83 (0.3–445)

Remote Annual Urban Annual

Con

cent

rati

on (

pg m

-3)

0

100

200

300

400

500

Fig. 2. Summary of ΣPBDE concentrations measured at the urban site in Kuwait Cityand a remote site close to the Iraqi border. The box contains the middle 50% of thedata, whereas the top and bottom end of the box represents the 75th and 25th percen-tiles of the data set, respectively. The extensions (“whiskers”) at either ends of the boxindicate the 99 and 1 percentile and the outliers represent individual data that exceed1.5 times the interquartile range.

537B. Gevao et al. / Science of the Total Environment 454–455 (2013) 534–541

60 pg m−3 at the urban location. The size of the box in Fig. 2 at the urbanlocation is larger indicating wider variations in concentrations measuredat this site. Although not statistically significant (p = 0.054), themedian(29 pg m−3) and mean (57 pg m−3) concentrations at the urban loca-tion were higher than those measured at the remote location (18 and29 pg m−3 respectively), consistent with the view that urban centersare an important net source of these compounds to the environment.Higher concentrations of most industrial POPs have been reported forurban locations relative to those at remote locations (Cetin andOdabasi, 2007; Genualdi et al., 2010; Gouin et al., 2007; Harner et al.,2006; Harrad and Hunter, 2006; Jafari et al., 2008; Motelay-Massei etal., 2005; Pozo et al., 2006). Strandberg et al. (Strandberg et al., 2001),for example, reported PBDE concentrations which were 4 to 7 timeshigher in Chicago air compared to measurements at three rural sitesaround the Great Lakes in 1999, whereas Harner et al. (Harner et al.,2006) reported a two-fold difference in atmospheric PBDE concentrationbetween Toronto and its surroundings, similar to that reported byHarradet al. in Birmingham, UK in 2006 (Harrad and Hunter, 2006). We also re-cently published the concentrations of PBDEs in surface soil samples col-lected along a 140 km transect across Kuwait to assess the role of urbancenters as sources of persistent organic pollutants to the surrounding en-vironment (Gevao et al., 2011a). The concentrations of PBDEs in KuwaitCity were about 250 times higher (P b 0.01) than those collected fromsites outside the city supporting the hypothesis that urban centers aresources of PBDEs and other POPs to the surrounding environment. Thedifference in concentration between the urban site and remote locationin this study and in the studies cited above are strongly influenced bythe proximity of the sampling site to potential primary sources. The con-centrations measured at most remote locations are more a reflection ofregional background atmospheric concentrations of these compoundspossibly from long-range transport from other regions. At urban loca-tions, however, it is likely that the sources of PBDEsmeasured in the sam-ples were from both primary emissions and secondary sources. Thesecan include the outgassing of contaminated air from homes and offices(Gevao et al., 2006a, 2006b), volatilization from soils (Gevao et al.,2011a) and sewage sludge drying operations (Gevao et al., 2008), volatil-ization from landfill sites around the city, and incineration of PBDE-containing waste in addition to possible long-range atmospheric trans-port from other regions.

The box plots in Fig. 2 show that although therewere isolated eventswith very high penta BDE signatures at both sites, these spikes in con-centrations could not be linked to any specific events. The influence of

air mass origin on concentrations was examined in an attempt to ex-plain these episodic spikes in concentrations. This was done by runningair mass back trajectory using the online HYSPLIT trajectory model(Draxler and Rolph, 2003). The analysis did not suggest that concentra-tions were elevated when the air mass originated from any specific re-gion. Episodic spikes in the atmospheric concentration of POPs, someassociated with known events, have been reported in the literature.For example, an increase in the atmospheric concentration in PBDEsand PCBs was observed in the early spring of 2000 in Ontario, Canada(Gouin et al., 2005). This “spring pulse” is hypothesized to be associatedwith the sudden release with snowmelt of pollutants accumulated in

538 B. Gevao et al. / Science of the Total Environment 454–455 (2013) 534–541

snow during winter months. In another study in the UK, elevated con-centrations of PBDEs and PAHs were measured during the national“bonfire night” festivities whenmany large fires are lit across the entirecountry (Farrar et al., 2004). The PAH concentration increased from ca20–30 ng m−3 to a maximum of ~180 ng m−3; whereas the PBDEconcentrations increased by a factor of 25 reaching concentrations of95 pg m−3. Alaee et al. (2003), for example, reported surprisinglyhigh concentrations of 700 and 2000 pg m−3 in summer at extremelyremote locations of Tagish and Alert which were thought to be relatedto the burning of refuse in the vicinity (Alaee et al., 2003). The extremelyhigh concentrationsmeasured in this study at the urban site (e.g. 17/11/2008, 21/12/2008, 5/4/2009 and 17/5/2009) whichwas not observed atthe remote sitemay be due to a localized event like refuse combustion ofproducts containing PBDEs or some other local event causing the epi-sodic release into the atmosphere. There are other studies where spikesin the atmospheric concentrations of pollutants were unexplained. Yaoet al. (Yao et al., 2006) (cited in (Moreau-Guigon et al., 2007)), for exam-ple, measured elevated concentrations of lindane of approximately3000 pg m−3 in air samples collected from downtown Toronto and600 pg m−3 further north on the 27th June, 2005 following a lindaneban in Canada.We also recently observed a spike in PBDE concentrationsin air in Kuwait during an extreme dust storm event that occurred inMay 2007where peak PBDE concentrationswere 30 times higher duringthe dust storm compared to post dust storm concentrations (Gevao etal., 2010). In the present study, however, we could not identify any spe-cific events to explain the localized spikes in concentrations at the urbansampling location particularly on17/11/2008, 21/12/2008, 5/4/2009 and17/5/2009.

3.1. Seasonal variations in concentrations

The mean seasonal concentrations are plotted as box plots for boththe remote and urban sampling locations in Fig. 3 to determine whetherthe concentrations of PBDEswere different for different seasons. The datawas divided into two seasonal groups based on temperature. Summer isdefined as the period fromApril toNovember duringwhich temperatures

Remote Winter Urban Winter Remote Summer Urban Summer

Con

cent

rati

on (

pg m

-3)

0

100

200

300

400

500

Fig. 3. Seasonally averaged concentrations of ΣPBDE at the urban site in Kuwait Cityand a remote site close to the Iraqi border. The box contains the middle 50% of thedata, whereas the top and bottom end of the box represents the 75th and 25th percen-tiles of the data set, respectively. The extensions (“whiskers”) at either ends of the boxindicate the 99 and 1 percentile and the outliers represent individual data that exceed1.5 times the interquartile range.

ranged between 20 and 45 °C and winter was defined as the period be-tween November and April when temperatures ranged between 6 and20 °C. There were no seasonal differences in concentrations at theurban location (p = 0.133, Mann–Whitney test). At the remote location,however, the median summer concentration (26 pg m−3) was signifi-cantly higher (p = 0.041,Mann–Whitney test) than themedian concen-tration measured in the winter months (14 pg m−3).

The most common processes that are known to control the atmo-spheric concentrations of semivolatile compounds include temperaturemediated air-surface exchange (Lee et al., 1998; Lohmann et al., 2003;Wallace and Hites, 1996), abiotic removal (e.g. photolysis, OH radicalremoval) from the atmosphere (Iacovidou et al., 2009; Mandalakis etal., 2003; Raff and Hites, 2006, 2007; Schenker et al., 2008; Soderstromet al., 2004), advection of contaminated air from source regions (Chiet al., 2008), and local meteorology (e.g. wind speed, temperature,humidity). The observed fluctuations in concentrations of pollutantsat a given location therefore provide important clues about their be-havior in that particular environment. It is thought for example thattemperature-mediated air-surface exchange is more important in de-termining the seasonal fluctuations of POPs in urban areas with reser-voirs from past usage (Lee et al., 1998; Lohmann et al., 2003; Wallaceand Hites, 1996) whereas abiotic removal processes and atmospherictransport may be more significant at remote background areas whereemissions are negligible (Mandalakis et al., 2003; Raff and Hites, 2007).

The temperature dependence of the gas-phase ambient PBDE con-centrations was investigated using Clausius–Clapeyron plots (Eq. (1)).

lnP ¼ −ΔHSA

R

� �1T

� �þ constant ð1Þ

Where P is the partial pressure of the PBDE congener (atm), ΔHSA isthe enthalpy of surface-air exchange (kJ mol−1), and T is the ambienttemperature (K), and R is the ideal gas constant (0.008314 kJ mol−1).The gas phase concentration of individual congeners were convertedintopartial pressures and the natural logarithmsof these values regressedagainst the reciprocalmean temperature for each sampling event (ln P vs.1/T). The regression results are summarized in Table 2 for both samplinglocations. The temperature dependence of the partial pressures of conge-ners 28, 47 and 100 were statistically significant at the remote samplingsite whereas only congener 47 showed any statistically significant corre-lation with temperature at the urban location. Although statisticallysignificant correlations existed for come congeners the slopes were shal-low ranging from−2732 to−3659 suggesting that long-range transportwas influenced by long-range transport. When the atmospheric concen-trations are controlling gas phase concentrations, the slopes tend to besteeper. For example, Cetin and Odabasi (Cetin and Odabasi, 2008)reported slopes which were an order of magnitude higher than those

Table 2Temperature dependence of the partial pressure for selected PBDE Congeners at the remoteand urban sites in Kuwait.

Congener Slope Intercept R2 Significance ΔHSA (kj mol−1)

Urban site28 −1116 −21.4 0.03 P > 0.05 9.347 −2732 −14.2 0.11 p b 0.05* 22.7100 −2750 −16.3 0.06 P > 0.05 22.999 −2419 −16 0.04 P > 0.05 20.1154 −626 −25 0.002 P > 0.05 5.2

Remote site28 −2948 −16 0.273 p b 0.002** 24.547 −2795 −14.6 0.104 P b 0.05* 23.2100 −3659 −13.8 0.12 P b 0.05* 30.499 −2912 −15 0.08 P > 0.05 24.2154 −2304 −19.6 0.06 P > 0.05 19.2153 −2368 −19.3 0.06 P > 0.05 19.7

** Very significant correlation; *significant correlation.

539B. Gevao et al. / Science of the Total Environment 454–455 (2013) 534–541

reported here. Temperature explained 11–27% of the atmospheric vari-ability of these congeners but explained only 0.2–8% for the congenerswhich were not significantly correlated with temperature. In this study,the ΣPBDE concentrations measured at both urban and remote siteswere not correlated with ambient temperature (p = 0.36). This lack ofcorrelation suggests that factors other than temperature are controllingthe concentrations in this study. Although temperature is not the over-arching factor controlling the concentrations in Kuwait, a plot of theratio “R” (Jiang et al., 2010; Wang et al., 2012; Yum et al., 2008), definedas the sum of the more volatile congeners (BDE 47 + BDE 99 + BDE100) to the sum of the higher molecular weight, less volatile congeners(BDE 153 + BDE 154) were positively correlated with temperature atboth sites (spearman correlation r = 0.33, p =0.037 at the rural site;r = 0.4, p = 0.013 at the urban location). The plot of the Ratio “R” forall sampling events plotted in Fig. 4 shows higher values in summerand lower values in winter suggesting that at higher temperatures theproportion of lowermolecularweight,more volatile congeners arehigherin the atmosphere compared to higher molecular congeners in line withpredictions based on the physico-chemical properties (vapor pressureand Log KOA).

3.1.1. Gas-particle partitioningThe partitioning of semivolatile organic compounds between the gas

and particle phases is usually described by the gas/particle partitioncoefficient, Kp (m3 μg−1)(Pankow, 1994).

KP ¼ F= A:TSPð Þ ð2Þ

Where A and F are the concentrations (pg m−3) in the gas and par-ticle phases respectively, and the TSP (μgm−3) is the total suspendedparticulate matter in air. The fraction on particles, ø, was calculated as:

∅ ¼ KP � TSPKP � TSP þ 1

ð3Þ

0

10

20

30

40

50

60

70

5/10

/200

812

/10/

2008

19/1

0/08

26/1

0/08

2/11

/200

810

/11/

2008

17/1

1/08

23/1

1/08

30/1

1/08

14/1

2/08

21/1

2/08

4/1/

2009

12/1

/200

918

/1/0

925

/01/

092/

2/20

098/

2/20

0922

/02/

098/

3/20

0922

/03/

09

R =

[47

+99+

100]

/[15

4+15

3]

Sampling

R-Rural

R-Urban

Temp

Fig. 4. Seasonal variations in the ratio “R” ([BDEs 47 + 99 + 100]/[BDEs 153 + 154]) a

The percentages of various congeners associated with particleswere calculated on the basis of Eqs. (2) and (3) for all sampling eventsand summarized in Table 1. Also given in Table 1 are the predictedfractions based on the Junge–Pankow adsorption model (PLo-basedmodel). In the classical PLo-based model, the fraction of a semivolatilecompound adsorbed onto particles is related to its subcooled liquid vaporpressure (PLo) and the surface area per unit volume of the particle (θ):

∅ ¼ cθ= PoL þ cθ

� � ð4Þ

Where c is a constant (17.2 Pa cm) relating to the heat of condensa-tion of the chemical and the surface area properties, and the surfacearea parameter (θ) assumed to be 1.1 × 10–5 cm2/cm3 (Bidleman,1988). A good agreement was obtained between the predicted andmeasured ø values for BDE 47 and to a lesser extent, BDE 100 at bothsites. However, the Junge–Pankow adsorption model underestimatedthe sorption of BDE-28 while overestimating the concentrations of allthe other congeners. The Junge–Pankow adsorption model have alsobeen reported to overestimate ∅ for PBDEs (Chen et al., 2006) andother semivolatile organic compounds including PCBs, PCDD/Fs andPCNs (Helm and Bidleman, 2005; Lee and Jones, 1999). The distributionof congeners between the gas and particulate phases observed in thisstudy have also been reported in other studies. For example, Strandberget al. (Strandberg et al., 2001) reported that 20% of BDE 47 was foundin the particle phase, 35 to 45% for BDE-99 and 100, and 70% for BDE153 and 154 in Great Lakes air. The deviations from the predicted behav-iors show that PBDEswere far fromequilibriumbetween the gas and par-ticle phases. The lower than expected particle fraction in this study maybe related to high total suspended particulate matter in air relative tothe concentration of chemicals. High concentrations of TSP are oftenassociated with dust storm episodes which are important weather phe-nomena in arid countries around the world (Al-Awadhi, 2005; Chi etal., 2008; Garrison et al., 2006).

0

5

10

15

20

25

30

35

40

45

5/4/

2009

19/0

4/09

3/5/

2009

17/0

5/09

31/0

5/09

14/0

6/09

28/0

6/09

12/7

/200

926

/07/

099/

8/20

0923

/08/

096/

9/20

0928

/09/

0911

/10/

2009

25/1

0/09

6/12

/200

921

/12/

094/

1/20

1017

/01/

102/

2/20

10

Tem

pera

ture

(°°C

)

Date

t sampling locations in Kuwait City and a remote location close to the Iraqi border.

540 B. Gevao et al. / Science of the Total Environment 454–455 (2013) 534–541

3.2. Technical mixtures

The major congeners detected in most samples analyzed were:BDE-47, -99, and -100 which together constitute ca 85 and 90% ofthe Σ5PBDEs at the remote and urban sites, respectively. Thesethree congeners are the dominant congeners in technical pentamixtures (Sjodin et al., 1998) suggesting that the main source ofPBDEs in the air during this period had a penta signature. The annu-al average percent contribution of the major congeners in decreas-ing order of importance at the remote site is: BDE-47 (42.2%), -99(31.6%), -100 (15%) with the others contributing less than 10%.Similarly, at the urban site, the same congeners also dominatedthe congener mixture with the distribution in decreasing order ofimportance being BDE-47 (40%), -99 (39%), -100 (9.8%). This simi-larity in the profiles to that in the technical penta- formulations in-dicates that the transfer of congeners from source-sink operateswith similar efficiencies across the penta-PBDE congeners andthat minimal weathering would have occurred during atmospherictransport. The distribution is similar to that reported in northernChina (Wang et al., 2012) and at a Mediterranean coastal site inFrance where BDEs 47 and 99 were reported to contribute 45 and39% respectively. This similarity between the profile in air and in thetechnical mixture suggests that volatilization from penta-treated prod-ucts is significant, assuming that the congener composition reported bySjodin et al. (1998) is representative of penta-BDE technical formula-tions (Sjödin et al., 1998).

4. Conclusions

The concentrations of PBDEs in Kuwait were generally higher at theurban site compared to the levels at the remote site (although the annu-ally averaged concentration differences were not statistically significant),consistent with other studies in the literature and with the view thaturban centers are a net source of industrial chemicals to the surroundingenvironment. The most likely source of PBDEs at the remote location isthought to be long-range transport from other regions, whereas at theurban locations other sources such as outgassing of contaminated airfrom homes and offices, emissions from vehicles and incineration ofPBDE containing waste will contribute to the measured concentrationsin addition to long-range atmospheric transport from other regions.There was no statistically significant seasonal difference in the medianconcentration at the urban location, but at the remote location,the median summer concentration was significantly higher thanwinter median concentrations. The absence in seasonality at theurban location may be due to ongoing primary emissions in urbanareas. The ΣPBDE concentrations measured at both urban and re-mote sites were not significantly correlated with ambient tempera-ture as is often reported for semivolatile organic compounds. Thislack of correlation suggests that factors other than temperaturewere controlling the concentrations in this study. Although tem-perature was not the overarching factor controlling the concentra-tions at the sampling sites, a plot of the ratio “R”, defined as theratio of the more volatile congeners (BDE 47 + BDE 99 + BDE100) to the sum of the less volatile compounds (BDE 153 + BDE154), was positively correlated with temperature at both sitessuggesting that temperature-dependent air-surface exchange ofPBDEs from soil and other reservoirs contributed to the observedconcentrations.

Acknowledgment

We are grateful to the Managements of the Kuwait Institute ofScientific Research, the Kuwait Foundation for the Advancement ofScience, and the Kuwait Environmental Public Authority for fundingthis research.

References

Alaee M, Muir D, Cannon C, Helm P, Harner T, Bidleman T. New persistent chemicals inArctic air and water. In: Bidleman T, Macdonald R, Stow J, editors. Canadian ArcticContaminants Assessment Report II. Northern Contaminants Program. Ottawa, On-tario, Canada: Ministry of Indian Affairs and Northern Development; 2003.p. 116–24.

Al-Awadhi J. Dust fallout characteristics in Kuwait: a case study. Kuwait J Sci Eng2005;32:135–52.

Allen JG, McClean MD, Stapleton HM, Webster TF. Critical factors in assessing exposureto PBDEs via house dust. Environ Int 2008;34:1085–91.

Arnold RG, Teske S, Tomanek M, Engstrom J, Leung C, Zhang J, et al. Fate ofpolybrominated diphenyl ethers during wastewater treatment/polishing andsludge stabilization/disposal. Ann N Y Acad Sci 2008;1140:394–411.

Batterman SA, Chernyak S, Jia CR, Godwin C, Charles S. Concentrations and emmisionsof polybrominated diphenyl ethers from U.S. houses and garages. Environ SciTechnol 2009;43:2693–700.

Beyer A, Mackay D, Matthies M, Wanai F, Webster E. Assessing long-range transportpotential of persistent organic pollutants. Environ Sci Technol 2000;34:699–703.

Bidleman TF. Atmospheric processes. Environ Sci Technol 1988;22:361–7.Burreau S, Zebuhr Y, Broman D, Ishaq R. Biomagnification of polychlorinated biphenyls

(PCBs) and polybrominated diphenyl ethers (PBDEs) studied in pike (Esox lucius),perch (Perca fluviatilis) and roach (Rutilus rutilus) from the Baltic Sea. Chemosphere2004;55:1043–52.

Cahill TM, Groskova D, Charles MJ, Sanborn JR, Denison MS, Baker L. Atmospheric con-centrations of polybrominated diphenyl ethers at near-source sites. Environ SciTechnol 2007;41:6370–7.

Cetin B, Odabasi M. Particle-phase dry deposition and air–soil gas-exchange ofpolybrominated diphenyl ethers (PBDEs) in Izmir, Turkey. Environ Sci Technol2007;41:4986–92.

Cetin B, OdabasiM. Atmospheric concentrations and phase partitioning of polybrominateddiphenyl ethers (PBDEs) in Izmir, Turkey. Chemosphere 2008;71:1067–78.

Chen LG, Mai BX, Bi XH, Chen SJ, Wang XM, Ran Y, et al. Concentration levels, compo-sitional profiles, and gas-particle partitioning of polybrominated diphenyl ethersin the atmosphere of an urban city in South China. Environ Sci Technol 2006;40:1190–6.

Chen D, Bi X, Zhao J, Chen L, Tan J, Mai B, et al. Pollution characterization and diurnalvariation of PBDEs in the atmosphere of an E-waste dismantling region. EnvironPollut 2009;157:1051–7.

Chi KH, Hsu SC, Wang SH, Chang MB. Increases in ambient PCDD/F and PCB concentra-tions in Northern Taiwan during an Asian dust storm episode. Sci Total Environ2008;401:100–8.

deWit CA. An overview of brominated flame retardants in the environment. Chemosphere2002;46:583–624.

Draxler RR, Rolph GD. HYSPLIT (HYbrid Single-Particle Lagrangian Integrated Trajectory)Model access via NOAA ARL READY Website. (http://www.arl.noaa.gov/ready/hysplit4.html). Silver Spring, MD: NOAA Air Resources Laboratory; 2003.

Farrar NJ, Smith KE, Lee RG, Thomas GO, Sweetman AJ, Jones KC. Atmospheric emissionsof polybrominated diphenyl ethers and other persistent organic pollutants during amajor anthropogenic combustion event. Environ Sci Technol 2004;38:1681–5.

Gallego E, Grimalt JO, BartronsM, Lopez JF, Camarero L, Catalan J, et al. Altitudinal gradientsof PBDEs and PCBs in fish from European high mountain lakes. Environ Sci Technol2007;41:2196–202.

Garrison VH, ForemanWT, Genualdi S, Griffin DW, Kellogg CA, Majewski MS, et al. Saharadust— a carrier of persistent organic pollutants, metals andmicrobes to the Caribbean.Rev Biol Trop 2006;54(Suppl. 3):9-21.

Gearhart J, Posselt H. Toxic at any speed. Chemicals in cars and the need for safe alterna-tives, vol. 28. The Ecology Center: Ann Arbor, Michigan; 2006. Available at: http://www.ecocenter.org/dust/ToxicAtAnySpeed.pdf. Accessed September 15, 2012.

Genualdi S, Lee SC, Shoeib M, Gawor A, Ahrens L, Harner T. Global pilot study of legacyand emerging persistent organic pollutants using sorbent-impregnated polyure-thane foam disk passive air samplers. Environ Sci Technol 2010;44:5534–9.

Gevao B, Al-Bahloul M, Al-Ghadban AN, Al-Omair A, Ali L, Zafar J, et al. House dust as asource of human exposure to polybrominated diphenyl ethers in Kuwait. Chemosphere2006a;64:603–8.

Gevao B, Al-Bahloul M, Al-Ghadban AN, Ali L, Al-Omair A, Helaleh M, et al.Polybrominated diphenyl ethers in indoor air Kuwait: implications for human ex-posure. Atmos Environ 2006b;40:1419–26.

Gevao B, Al-Omair A, SweetmanA, Al-BahloulM, Al-Ali L, HelalehM, et al. Passive-samplerderived air concentrations for polybrominated diphenyl ethers and polycyclic aro-matic hydrocarbons in Kuwait. Environ Toxicol Chem 2006c;25:1496–502.

Gevao B, Beg MU, Al-Ghadban AN, Al-Omair A, Helaleh M, Zafar J. Spatial distribution ofpolybrominated diphenyl ethers in coastal marine sediments receiving industrialand municipal effluents in Kuwait. Chemosphere 2006d;62:1078–86.

Gevao B, Muzaini S, Helaleh M. Occurrence and concentrations of polybrominateddiphenyl ethers in sewage sludge from three wastewater treatment plants in Kuwait.Chemosphere 2008;71:242–7.

Gevao B, Jaward FM, MacLeod M, Jones KC. Diurnal fluctuations in polybrominateddiphenyl ether concentrations during and after a severe dust storm episode in KuwaitCity, Kuwait. Environ Sci Technol 2010;44:8114–20.

Gevao B, Ghadban AN, Uddin S, Jaward FM, Bahloul M, Zafar J. Polybrominated diphenylethers (PBDEs) in soils along a rural–urban–rural transect: Sources, concentrationgradients, and profiles. Environ Pollut 2011a;159:3666–72.

Gevao B, Jaward FM, Al-Bahloul M, Ud din S, Beg MU, Zafar J. Polybrominated diphenylethers in three commercially important fish from theNorthwestern Arabian Gulf: occur-rence, concentration, and profiles. Arch Environ Contam Toxicol 2011b;60:636–42.

541B. Gevao et al. / Science of the Total Environment 454–455 (2013) 534–541

Gonzalez-Mille DJ, Ilizaliturri-Hernandez CA, Espinosa-Reyes G, Costilla-Salazar R,Diaz-Barriga F, Ize-Lema I, et al. Exposure to persistent organic pollutants (POPs)and DNA damage as an indicator of environmental stress in fish of different feedinghabits of Coatzacoalcos, Veracruz, Mexico. Ecotoxicology 2010;19:1238–48.

Gouin T, Harner T. Modelling the environmental fate of the brominated diphenylethers. Environ Int 2003;29:717–24.

Gouin T, Harner T, Daly GL, Wania F, Mackay D, Jones KC. Variability of concentrationsof polybrominated diphenyl ethers and polychlorinated biphenyls in air: implica-tions for monitoring, modelling and control. Atmos Environ 2005;39:151–66.

Gouin T, Jantunen L, Harner T, Blanchard P, Bidleman T. Spatial and temporal trends ofchiral organochlorine signatures in Great Lakes air using passive air samplers. EnvironSci Technol 2007;41:3877–83.

Harley KG, Marks AR, Chevrier J, Bradman A, Sjodin A, Eskenazi B. PBDE concentrationsin women's serum and fecundability. Environ Health Perspect 2010;118:699–704.

Harner T, Shoeib M, Diamond M, Ikonomou M, Stern G. Passsive sampler derived airconcentrations of PBDE along an urban–rural transect: spatial and temporal trends.Chemosphere 2006;64:262–7.

Harrad S, Hunter S. Spatial variation in atmospheric levels of PBDEs in passive air sampleson an urban–rural transect. Organohalogen Compd 2004;66:3786–92.

Harrad S, Hunter S. Concentrations of polybrominated diphenyl ethers in air and soil ona rural–urban transect across a major UK conurbation. Environ Sci Technol 2006;40:4548–53.

Helm P, Bidleman TF. Gas-particle partitioning of polychlorinated naphthalenes andnon- and mono-ortho-substituted polychlorinated biphenyls in arctic air. Sci TotalEnviron 2005;342:161–73.

Herbstman JB, Sjodin A, Kurzon M, Lederman SA, Jones RS, Rauh V, et al. Prenatal expo-sure to PBDEs and neurodevelopment. Environ Health Perspect 2010;118:712–9.

Iacovidou E, Mandalakis M, Stephanou EG. Occurrence and diurnal variation ofpolychlorinated biphenyls and polybrominated diphenyl ethers in the backgroundatmosphere of Eastern Mediterranean. Chemosphere 2009;77:1161–7.

Jafari A, Moeckel C, Jones KC. Spatial biomonitoring of persistent organic pollutants inIran: a study using locally produced butter. J Environ Monit 2008;10:861–6.

Jaward FM, Farrar NJ, Harner T, Sweetman AJ, Jones KC. Passive air sampling of PCBs,PBDEs, and organochlorine pesticides across Europe. Environ Sci Technol 2004;38:34–41.

Jiang Y, Wang X, Zhu K, Wu MS, Sheng G, Fu J. Occurrence, compositional profiles andpossible sources of polybrominated diphenyl ethers in urban soils of Shanghai,China. Chemosphere 2010;80:131–6.

Jin J, Wang Y, Liu W, Yang C, Hu J, Cui J. Polybrominated diphenyl ethers in atmosphereand soil of a production area in China: levels and partitioning. J Environ Sci (China)2011;23:427–33.

La Guardia L, Hale RC, Harvey E. Detailed polybrominated diphenyl ether (PBDE) con-gener composition of the widely used penta-, octa-, and deca-PBDE technicalflame-retardant mixtures. Environ Sci Technol 2006;40:6247–54.

Lee RGM, Jones KC. Gas-particle partitioning of atmospheric PCDD/Fs: measurementsand observations on modelling. Environ Sci Technol 1999;33:3596–604.

Lee RGM, Hung H, Mackay D, Jones KC. Measurement and modelling of the diurnalcycling of atmospheric PCBs and PAHs. Environ Sci Technol 1998;32:2172–9.

Lee RGM, Thomas GO, Jones KC. PBDEs in the atmosphere of three locations in WesternEurope. Environ Sci Technol 2004;38:699–706.

Lohmann R, Brunciak PA, Dachs J, Gigliotti CL, Nelson E, van-Ry D, et al. Processes con-trolling diurnal variations of PCDD/Fs in the New Jersey coastal atmosphere. AtmosEnviron 2003;37:959–69.

Lohmann R, Breivik K, Dachs J, Muir DC. Global fate of POPs: current and future researchdirections. Environ Pollut 2007;150:150–65.

Mandalakis M, Berresheim H, Stephanou EG. Direct evidence for destruction ofpolychlorobiphenyls by OH radicals in the subtropical troposphere. Environ SciTechnol 2003;37:542–7.

Meng XZ, Pan ZY, Wu JJ, Qiu YL, Chen L, Li GM. Occurrence of polybrominated diphenylethers in soil from the central Loess Plateau, China: role of regional range atmo-spheric transport. Chemosphere 2011;83:1391–7.

Moreau-Guigon E, Motelay-Massei A, Harner T, Pozo K, Diamond M, Chevreuil M, et al.Vertical and temporal distribution of persistent organic pollutants in Toronto. 1.Organochlorine pesticides. Environ Sci Technol 2007;41:2172–7.

Motelay-Massei A, Harner T, Shoeib M, Diamond M, Stern G, Rosenberg B. Using pas-sive air samplers to assess urban–rural trends for persistent organic pollutants

and polycyclic aromatic hydrocarbons. 2. Seasonal trends for PAHs, PCBs, and or-ganochlorine pesticides. Environ Sci Technol 2005;39:5763–73.

Ockenden WA, Jones KC. Global fractionation. Prog Environ Sci 1999;1:119–51.Pankow JF. An absorption model of gas/particle partitioning of organic compounds in

the atmosphere. Atmos Environ 1994;28:185–8.Pozo K, Harner T, Wania F, Muir DC, Jones KC, Barrie LA. Toward a global network for

persistent organic pollutants in air: results from the GAPS study. Environ SciTechnol 2006;40:4867–73.

Pozo K, Harner T, Lee SC, Wania F, Muir DC, Jones KC. Seasonally resolved concentra-tions of persistent organic pollutants in the global atmosphere from the first yearof the GAPS study. Environ Sci Technol 2009;43:796–803.

Raff JD, Hites RA. Gas-phase reactions of brominated diphenyl ethers with OH radicals. JPhys Chem A 2006;110:10783–92.

Raff JD, Hites RA. Deposition versus photochemical removal of PBDEs from Lake Superiorair. Environ Sci Technol 2007;41:6725–31.

Schenker U, Soltermann F, Scheringer M, Hungerbuhler K. Modeling the environmentalfate of polybrominated diphenyl ethers (PBDEs): the importance of photolysis forthe formation of lighter PBDEs. Environ Sci Technol 2008;42:9244–9.

Shin M, Duncan B, Seto P, Falletta P, Lee DY. Dynamics of selected pre-existingpolybrominated diphenylethers (PBDEs) in municipal wastewater sludge underanaerobic conditions. Chemosphere 2010;78:1220–4.

Sjodin A, Jakobsson E, Kierkegaard A, Marsh G, Sellstrom U. Gas chromatographic iden-tification and quantification of polybrominated diphenyl ethers in a commercialproduct, Bromkal 70-5DE. J Chromatogr A 1998;822:83–9.

Sjödin A, Jakobsson E, Kierkegaard A, Marsh G, Sellstrom U. Gas chromatographic iden-tification and quantification of polybrominated diphenyl ethers in a commercialproduct, Bromkal 70-5DE. J Chromatogr A 1998;822:83–9.

Soderstrom G, Sellstrom U, de Wit CA, Tysklind M. Photolytic debromination ofdecabromodiphenyl ethyl (BDE 209). Environ Sci Technol 2004;38:127–32.

Strandberg B, Dodder NG, Basu I, Hites RA. Concentrations and spatial variations ofpolybrominated diphenyl ethers and other organohalogen compounds in GreatLakes air. Environ Sci Technol 2001;35:1078–83.

Toms LM, Sjodin A, Harden F, Hobson P, Jones R, Edenfield E, et al. Serumpolybrominated diphenyl ether (PBDE) levels are higher in children (2–5 yearsof age) than in infants and adults. Environ Health Perspect 2009;117:1461–5.

UNEP. The nine new POPs under the Stockholm Convention; 2010.von Waldow H, Macleod M, Scheringer M, Hungerbuhler K. Quantifying remoteness

from emission sources of persistent organic pollutants on a global scale. EnvironSci Technol 2010;44:2791–6.

Wallace JC, Hites RA. Diurnal variations in atmospheric concentrations of polychlorinatedbiphenyls and endosulfan: Implications for sampling protocols. Environ Sci Technol1996;30:444–6.

Wang P, Zhang Q, Wang Y, Wang T, Li X, Li Y, et al. Altitude dependence ofpolychlorinated biphenyls (PCBs) and polybrominated diphenyl ethers (PBDEs)in surface soil from Tibetan Plateau, China. Chemosphere 2009;76:1498–504.

WangMS, Chen SJ, HuangKL, Lai YC, Chang-ChienGP, Tsai JH, et al. Determination of levelsof persistent organic pollutants (PCDD/Fs, PBDD/Fs, PBDEs, PCBs, and PBBs) in atmo-sphere near a municipal solid waste incinerator. Chemosphere 2010;80:1220–6.

Wang C, Li W, Chen J, Wang H, Li T, Shen G, et al. Summer atmospheric polybrominateddiphenyl ethers in urban and rural areas of northern China. Environ Pollut 2012;171:234–40.

Wania F, Dugani CB. Assessing the long-range transport potential of polybrominateddiphenyl ethers: a comparison of four multimedia models. Environ Toxicol Chem2003;22:1252–61.

Wilford BH, Harner T, Zhu J, Shoeib M, Jones KC. A passive sampling survey ofpolybrominated diphenyl ether flame retardants in indoor and outdoor air in Ottawa,Canada: implications for sources and exposure. Environ Sci Technol 2004;38:5312–8.

Yao Y, Harner T, Brice K, Su K, Blanchard P,Ma J, et al. Episodic transport of g-HCH (lindane)to the Great Lakes basin after the official Canadian Ban. In SETAC, 27th Annual Meting(Montreal, Canada); 2006.

Yum SH, Addink R, McCabe JM, Ostaszewski A, Mackenzie-Taylor D, Taylor AB, et al.Polybrominated diphenyl ethers and polybrominated biphenyls in sediment andfloodplain soils of the Saginaw river watershed, Michigan, USA. Arch EnvironContam Toxicol 2008;55:1-10.

Zhang X, Diamond ML, Ibarra C, Harrad S. Multimedia modeling of polybrominateddiphenyl ether emissions and fate indoors. Environ Sci Technol 2009;43:2845–50.