ARTICLE IN PRESS

1352-2310/$ - se

doi:10.1016/j.at

�Correspondfax: +1613 562

E-mail addr

Atmospheric Environment 42 (2008) 2948–2958

www.elsevier.com/locate/atmosenv

Seasonal trends in vegetation and atmospheric concentrations ofPAHs and PBDEs near a sanitary landfill

Annick D. St-Amanda, Paul M. Mayera, Jules M. Blaisb,�

aDepartment of Chemistry, University of Ottawa, 10 Marie-Curie, Ottawa, Ontario, Canada K1N 6N5bDepartment of Biology, University of Ottawa, 30 Marie-Curie, Ottawa, Ontario, Canada K1N 6N5

Received 1 July 2007; received in revised form 18 December 2007; accepted 20 December 2007

Abstract

Spruce needle and atmospheric (gaseous and particulate-bound) concentrations were surveyed near a sanitary landfill

from February 2004 to June 2005. The influence of several parameters such as temperature, relative humidity, wind speed

and direction, as well as increased domestic heating during the winter was assessed. In general, polybrominated diphenyl

ethers (PBDE) and polycyclic aromatic hydrocarbons (PAH) concentrations in spruce needles increased over time and

decreased following snowmelt with a minimum coinciding with bud burst of deciduous trees. Atmospheric concentrations

of PBDE and PAH, both particulate-bound and gaseous phase, were linked to daily weather events and thus showed more

variability than those in spruce needles. Highest PAH concentrations were encountered during the winter, likely reflecting

increased emission from heating homes. Pseudo Clausius-Clapeyron plots revealed higher PBDE gaseous concentrations

with increasing temperature, but showed no correlation between PAH gaseous concentrations and temperature as this

effect was obscured by seasonal emission patterns. Finally, air mass back trajectories and local wind directions revealed

that particulate-bound PBDEs, along with both gaseous and particulate-bound PAHs were from local sources, whereas

gaseous PBDEs were likely from distant sources.

r 2007 Elsevier Ltd. All rights reserved.

Keywords: PBDE; PAH; Sanitary landfill; Atmospheric; Vegetation; Temperature; Wind

1. Introduction

Both polybrominated diphenyl ethers (PBDEs)and polycyclic aromatic hydrocarbons (PAHs) areconsidered ubiquitous contaminants quantified in avariety of environmental media throughout theworld, including very remote pristine ecosystems(Hale et al., 2003; de Wit et al., 2006; Simcik and

e front matter r 2007 Elsevier Ltd. All rights reserved

mosenv.2007.12.050

ing author. Tel.: +1 613 562 5800x6650;

5486.

ess: jules.blais@uottawa.ca (J.M. Blais).

Offenberg, 2006). PBDEs are used as flame-retar-dants in a variety of consumables (Alaee et al.,2003) and increased usage is reflected in risingconcentrations in the environment (de Wit et al.,2006; Hale et al., 2006). It has also been demon-strated that PBDEs are toxic and persistent (de Wit,2002; de Wit et al., 2006; Hale et al., 2006); recentmodeling studies have shown that the more volatilecongeners have a moderate long-range atmospherictransport potential (Palm et al., 2002; Gouin andHarner, 2003). As they are simply mixed withmaterials during the manufacturing process and are

.

ARTICLE IN PRESSA.D. St-Amand et al. / Atmospheric Environment 42 (2008) 2948–2958 2949

not chemically bonded, PBDEs can leach in theenvironment from consumer products (Alaee et al.,2003; Osako et al., 2004). PAHs are also persistentorganic pollutants considered toxic and lipophilic,with a wide range of volatility; acenaphthylene ishighly volatile (supercooled liquid vapor pressure of4.14 Pa at 25 1C), whereas benzo(ghi)perylene has anegligible vapor pressure (supercooled liquid vaporpressure of 2.25� 10�5 Pa at 25 1C) (Mackay et al.,2006). Some are considered highly carcinogenic andtheir potential impact on human health may be dueto their association with particulates (de Kok et al.,2005). They can be released in the environment fromboth natural (forest fires and domestic woodburning) and anthropogenic sources (industrialactivities, incomplete combustion of petroleumand coal, and waste incineration) (Simcik andOffenberg, 2006).

Sanitary landfills accumulate vast quantities of awide variety of consumables, including general andhazardous wastes, electric and electronic equipmentand compost. Several studies have considered theenvironmental fate and atmospheric release ofvolatile organic compounds from landfills (Agrellet al., 2004; Osako et al., 2004; Wichmann et al.,2006). For semi-volatile organic compounds, likepolychlorinated biphenyls (PCBs), PAHs orPBDEs, most studies focus on leachate concentra-tions and the impact of incineration facilities (Agrellet al., 2004; Osako et al., 2004; Wichmann et al.,2006). Studies have shown that PBDEs, along witha wide array of brominated flame-retardants, arefound in waste electrical and electronic equipmentand that these can leach in the environment (Osakoet al., 2004). Furthermore, the presence of compost-ing facilities could also contribute to the overallrelease of semi-volatile organic compounds(SVOCs) from sanitary landfills (Deportes et al.,1995).

In this study, concentrations of PBDEs andPAHs in environmental media were surveyed at asanitary landfill for 16 months (February 2004–June2005). Atmospheric (gaseous and particulate-bound) and spruce needle samples were collectedon a bi-weekly basis. Norway spruce (Picea abies)was considered for this study since these conifershave previously been shown to be efficient passiveair samplers for SVOCs (Schroter-Kermani et al.,2006; St-Amand et al., 2007). Temporal trends ofPBDEs and PAHs in these three media were studiedand influence of environmental and meteorologicalparameters along with seasonal effects were as-

sessed. These observations were then used toidentify emission sources and release patterns.

2. Experimental procedure

2.1. Sampling site and techniques

Samples were collected at the Trail Waste Facility(situated at 451130N, 751460W), a sanitary landfillfor the city of Ottawa, ON, Canada. This landfill isdivided into two major areas: general waste andcompost. Hazardous waste, tires and appliances arealso collected and eventually sent elsewhere. It islocated close to a major highway in a mostlyresidential agricultural area with little industrialactivity. During the study period, wind directionwas predominantly from the west at an averagewind speed of 3.62m s�1. Temperatures rangedfrom �30.7 1C (15 January 2004) to 32.4 1C (11July 2005) with lowest temperatures generallyencountered in January and highest temperaturesin July. Maximum precipitation for a 24 h periodwas 135.4mm on 9 September 2004, which was dueto a tropical storm remnant. Maximum snowfall fora 24 h period was recorded on 7 March 2005(15.4 cm) which contributed to a 40 cm maximumsnow cover observed between 8 and 10 March 2005.

Atmospheric and spruce needle samples werecollected bi-weekly from February 2004 to June2005. Full details on sampling techniques have beenpresented previously (St-Amand et al., 2007).Briefly, a high-volume air sampler was deployedfor approximately 18 h to collect particulates andgaseous species, which were retained with a glassmembrane filter and polyurethane foam (PUF),respectively. Filters were stored in solvent-cleanedaluminum foil and PUFs were kept in solvent-cleaned jars. Norway spruce needles were collectedand placed in solvent-cleaned aluminum foil. Allsamples were kept at �80 1C until extraction (lessthan 3 months).

2.2. Extraction procedure

Details about the extraction procedures foratmospheric and spruce needle samples are alsodescribed previously (St-Amand et al., 2007). Beforeextraction, the filters were dried and weighed todetermine total suspended particulate (TSP). Allsamples were spiked with two surrogate standards(BDE-77 and BDE-118) before extraction toassess recovery. Filters and PUFs were extracted

ARTICLE IN PRESSA.D. St-Amand et al. / Atmospheric Environment 42 (2008) 2948–29582950

separately with hexane using a Soxhlet apparatusfor approximately 18 h and extracts were concen-trated using a rotary evaporator. Dried spruceneedles were extracted using an accelerated solventextraction module (Dionex ASE-200) at 140 1Cusing 1:1 hexane:dichloromethane mixture. Extractswere concentrated using a TurboVap (Zymark) anda 10% aliquot was removed for lipid determination.These were then put through a small-deactivatedsilica column, which was eluted with hexane andconcentrated to approximately 200 mL using aTurboVap. Final extract volume was 500 mL forall samples.

2.3. Instrumental analysis

An Agilent HP 6890 gas chromatograph coupledwith a quadrupole mass selective detector HP5973N was used for all analyses. PBDEs wereanalyzed using chemical ionization mass spectro-metry (GC–CIMS). Injections of 2 mL were made inpulsed splitless mode at 280 1C. A fused-silicacapillary column (15m� 0.25mm) coated with0.25 mm chemically bonded HP-5MS phase (5%phenyl methyl siloxane) was used. Initial oventemperature was set at 120 1C, followed by an initialramp of 60 1Cmin�1 to 180 1C and a second ramp of25 1Cmin�1 to 300 1C which was maintained for5.20min for a total run time of 11min. Columncarrier gas helium was set at a constant flow of1.1mLmin�1 for lower PBDE congeners (tri tohepta) to ensure optimal resolution between chro-matographic peaks. This was increased to3mLmin�1 for BDE-209 analysis and the finaltemperature of 300 1C was kept for 10.20min for atotal run of 16min.

The mass spectrometry parameters were the samefor all PBDE congeners. The GC–MS interface andsource were kept at 300 and 250 1C, respectively, toensure maximum sensitivity and methane was usedas reagent gas. The following mass fragments weremonitored in SIM mode: m/z 79 and 81, forquantification and qualification of PBDEs, respec-tively, and m/z 404 for internal standard (octa-chloronaphthalene) monitoring. The followingPBDEs were separated and quantified: tri-BDE 17and 28, tetra-BDE 47, 66, 71 and 77 (surrogate),penta-BDE 85, 99, 100 and 118 (surrogate), hexa-BDE 138, 153, and 154, hepta-BDE 183 and 190and deca-BDE 209.

PAHs were analyzed using electron ionizationmass spectrometry (GC–EIMS). Injections of 1 mL

were made in pulsed splitless mode at 280 1C. GCseparation was realized using the same column asdescribed above (with column guard). Initial oventemperature was set at 150 1C and held for 2min,followed by a first ramp of 10 1Cmin�1 to 240 1Cand second ramp of 5 1Cmin�1 to 300 1C, which washeld for 5min. The GC–MS interface was kept at300 1C and the source at 150 1C. The followingfragments were monitored in SIM mode: m/z 152for acenaphthylene (AceN), m/z 166 for fluorene(Flu), m/z 178 for phenanthrene (Phe) and anthra-cene (Ant), m/z 202 for pyrene (Pyr), m/z 228 forbenz(a)anthracene and chrysene (BaA+Chr), m/z240 for chrysene-d12 (internal standard), m/z 252for benzo(b)fluoranthene and benzo(k)fluoranthene(BbF+BkF), and m/z 276 for indeno(123-cd)pyrene(IndP) and benzo(ghi)perylene (BghiP). Benz(a)-anthracene and chrysene as well as benzo(b)fluor-anthene and benzo(k)fluoranthene were combinedbecause of poor peak resolution. Ratios of thepeak areas of the analytes and the internal standardwere used for quantification. Calibration curves con-structed using 10 standards for PBDEs (0–5pgmL�1)and five standards for PAHs (0–0.1ngmL�1) gave0.99 or higher coefficients of determination.

2.4. Quality control

All non-graduated glassware was triple-washedwith laboratory grade detergent, triple-rinsed withdistilled water, rinsed with ACS grade acetone andhexane before placing in an oven at 200 1C forseveral hours. Graduated glassware was washed asdescribed, but rinsed with Omnisolv grade acetoneand hexane and was not oven-dried. Blanks (PUFs,filters, and ASE) were processed to check forlaboratory contamination and concentrations ofmost compounds (PBDEs and PAHs) were belowdetection limit; BDE-47 and -99 were found at verylow concentrations and thus the samples were notblank corrected. PUF breakthrough artifacts werealso assessed by cutting PUFs in half and compar-ing contribution of the top and bottom halves.Average total concentration breakthrough to thesecond PUF half was generally less than 10% andwas highest for acenaphthylene (average below30%). Sample recovery was evaluated for allsamples using two surrogate standards (BDE-77and -118). Average recoveries for spruce needle,filter and PUF samples were 70.975%, 86.377%,and 89.078%, respectively. If recovery fell outside

ARTICLE IN PRESSA.D. St-Amand et al. / Atmospheric Environment 42 (2008) 2948–2958 2951

these ranges, samples were recovery corrected tomean values.

Recovery was also evaluated on a compound basisby spiking blanks with a low- concentration stan-dard. For PBDEs these were found to be 93.874%for spruce needle samples (n ¼ 6), 87.275% for filtersamples (n ¼ 3) and 83.574% for PUF samples(n ¼ 9). For PAHs these generally ranged from 30%to 80% for spruce needle samples (n ¼ 5), from 60%to 110% for filter samples (n ¼ 5) and above 90% forPUF samples (n ¼ 5). PAH sample concentrationswere adjusted accordingly. Method detection limitswere determined using both the standard deviation ofseveral injections (nX7) of a low-concentrationstandard and the error of the regression. For PBDEsdetection limits ranged from 0.002 to 0.056 pgm�3

for atmospheric samples (assuming a sample volumeof 800m3), and 0.25 to 5.3 pg g�1 for spruce samples(assuming a sample mass of 8 g). For PAHs theseranged from 0.001 to 0.007ngm�3 for atmosphericsamples and from 0.13 to 0.67pg g�1 for spruceneedles.

3. Results and discussion

3.1. PBDE and PAH spruce needle concentrations

Spruce needles showed gradual accumulation ofboth PBDEs and PAHs starting at bud burst (as

Fig. 1. Spruce needles total concentration of PAHs and PBDEs (ng g�

buds).

shown in Fig. 1). For PBDEs, total concentra-tions generally ranged from 210 to 9940 pg g�1

dry weight with an average of 2370 pg g�1. Table1S presents a summary of spruce needle concentra-tions for selected PBDEs. Without BDE-209total PBDE concentrations were much lower ran-ging from 20 to 3750 pg g�1 with an average920 pg g�1 dry weight. It should also be noted thatBDE-71 was not detected in spruce needle samples.Predominant PBDE congeners quantified in spruceneedles were BDE-209, 99 and 47 which accountfor more than 85% of the total PBDE burden inspruce needles (Fig. 1S) and reflect current usageof Penta and DecaBDE technical formulations(Hale et al., 2003). PBDE vegetation congenerprofiles remained fairly constant throughout thestudy period.

Total PAH spruce needle concentrations weresignificantly higher than PBDEs (paired t-test,po0.05) ranging from 10.7 to 608 ng g�1 dry weightwith an average 82.8 ng g�1 and were within valuesreported in the literature (Schroter-Kermani et al.,2006). Table 1S also presents a summary of spruceneedle concentrations for selected PAHs. Contri-buting over 80% of the total PAH spruce needlesburden were benzo(b)fluoranthene+benzo(k) fluor-anthene, benzo(a)pyrene and indeno(123-cd)pyrene(Fig. 2S). PAH congener profiles were also fairlyconstant throughout the study period.

1 dry weight) from February 2004 to June 2005 (2003 and 2004

ARTICLE IN PRESSA.D. St-Amand et al. / Atmospheric Environment 42 (2008) 2948–29582952

PBDE spruce needle concentrations generallyincreased during the warmer summer monthsstabilizing in November. This trend is however;largely influenced by BDE-209 one of the predomi-nant congeners present in spruce needles, andlighter PBDEs show a more gradual increase.PAH spruce needle concentrations increased slowlystarting at bud burst until November, at whichpoint there was a sudden increase in concentration.This is most probably due to increased fuelconsumption for domestic heating at the arrival ofcolder temperatures, causing higher PAH atmo-spheric concentrations (see below). Although asimilar increase can also be noted for PBDEs, thisis limited to a few data points, and appliesexclusively to BDE-209, which, as mentionedpreviously, is the largest contributor to total PBDEvegetation concentration. With the arrival of spring,a decrease in total PBDE and PAH concentrationsis seen following snowmelt (February and March)with a minimum at spring bud burst. Similar trendshave also been observed for DDT (and to someextent HCHs) in pine needles (Kylin and Sjodin,2003). This was suggested to be due to thedegradation and abrasion of the wax matrix duringthe winter-to-spring transition. In fact, the winterseason brings forth stressful conditions for conifers,such as low water apportionment, windy conditions,and sun radiation. During winter cold temperatureprogress and the protective layer on conifer needlesthin out. This layer can thin rapidly with the arrivalof spring, especially if this transition is dramatic,such as a few days of elevated temperatures(Larcher, 1996; Fitter and Hay, 2003). With thearrival of spring, it has also been suggestedthat PBDEs could be re-released from the meltingsnowpack followed by subsequent depositiononto available surfaces (Gouin and Harner, 2003).The new leaves of the deciduous trees could offernew deposition surfaces resulting in competitiveuptake.

3.2. PBDE and PAH atmospheric concentrations

As seen in Fig. 2, gaseous and particulate-boundtotal PBDE concentrations were generally below 2and 20 pgm�3, and ranged from 0.36 to 6.9 pgm�3

and 0.72 to 145 pgm�3, respectively, and arecomparable to PBDE atmospheric concentrationsreported in the region (Butt et al., 2004; Wilfordet al., 2004). Table 2S presents a summary of theatmospheric concentrations for selected com-

pounds. Volatile congeners were predominantlyfound in the gaseous fraction, with BDE-28, -47and -99 accounting, on average, for more than 80%of the total concentration (Fig. 1S). PBDEsassociated to particulates generally had a higherlog KOA; BDE-183 and -209 were found exclusivelyassociated to particulates. BDE-209 contributed onaverage 55% to the total PBDE particulate burden,with BDE-47 and -99 also present in significantamounts (Fig. 1S). The omnipresence of BDE-47,-99 and -209 in all media studied (spruce needles,gaseous and particulate-bound) reflects the currentusage pattern of PentaBDE and DecaBDE technicalPBDE formulations in North America (Hale et al.,2003).

Atmospheric PAH concentrations were an orderof magnitude higher than those of PBDEs. Asshown in Fig. 2, total gaseous PAHs ranged from244 to 3680 pgm�3 (generally below 2000 pgm�3)while particulate-bound concentrations ranged from28 to 1870 pgm�3 (generally below 500 pg/m3) andfell within values reported in the literature forCanada (Sanderson et al., 2004). The three-ringPAHs (acenaphthylene, fluorene, phenanthrene,and anthracene) constituted more than 90% ofthe total gaseous concentrations. Almost all PAHswere found in appreciable amounts in particulates,except for acenaphthylene, fluorene and anthracene(Fig. 2S).

PBDE vegetation uptake occurs through bothparticulate-bound and gaseous depositions. Therelative importance of these processes is affectedby their logKOA and particulate-gas partitioningbehavior (St-Amand et al., 2007).

Particulate-gas partitioning was assessed bycalculating KP for each sampling sessions asdescribed by

KP ¼ðCP=TSPÞ

CG, (1)

where CP and CG are the particulate-bound andgaseous concentrations, respectively (pgm�3) andTSP is the total suspended particulates (mgm�3).Measured TSP values are presented in Fig. 2.

As shown in Fig. 3, particulate-gas partitioning ofboth PBDEs and PAHs were strongly correlatedto logKOA (po0.001), as the proportion ofSVOCs associated to particulates increased withdecreasing volatility. On average, penta-BDE andhigher congeners (logKOA above 11) were mostlyassociated to particulates (above 70%) whichcorroborates previous results (Shoeib et al., 2004).

ARTICLE IN PRESS

Fig. 2. Atmospheric total PBDE and PAH concentrations (pgm�3) and total suspended particulates (mgm�3) for each sampling session.

A.D. St-Amand et al. / Atmospheric Environment 42 (2008) 2948–2958 2953

PAHs containing at least four rings with a logKOA

above 10 (benzo(a)anthracene+chrysene, benzo(b)-fluoranthene+benzo(k)fluoranthene, benzo(a)pyrene,indeno(123-cd)pyrene, and benzo(ghi)perylene) werealso found predominantly associated to particulates

(average also above 70%). This corroboratesother studies, except for benzo(a)anthracene+chrysene, which has been found to split more evenlybetween the gas phase and particulates (Su et al.,2006).

ARTICLE IN PRESS

Fig. 3. Particulate-gas partitioning of PBDEs and PAHs studied.

Individual compounds are indicated. Error bars: standard error

of the mean (n ¼ 36 PBDEs and n ¼ 35 PAHs).

Table 1

Calculated p-values for the correlation of total PBDE and PAH gas

parameters

Meteorological parametersPBDE

Gaseous

Temperature (K) 0.000*(+)

Relative humidity (%) 0.953

Air masses backward trajectories (categorical) 0.001*

Air masses backward trajectories (numerical) 0.048*

Wind direction (categorical) 0.353

Wind direction (numerical) 0.283

Wind speed (km/h) 0.423

Presence of an asterisk indicates significant correlation (po0.05). Sign

Table 2

Summary of regression slopes, correlation coefficients and p-values from

selected PBDEs

Compound Slope R2 p-values DH

BDE-17 �6.44 0.52 po0.5 53.

BDE-28 �5.50 0.59 po0.001 45.

BDE-47 �4.83 0.72 po0.001 40.

BDE-66 �4.76 0.43 po0.001 39.

BDE-85 �2.97 0.19 po0.1 24.

BDE-99 �2.83 0.42 po0.001 23.

BDE-100 �3.33 0.45 po0.001 27.

BDE-138 �5.27 0.28 – 43.

BDE-153 �2.92 0.18 po0.01 24.

BDE-154 �2.24 0.12 po0.05 18.

BDE-183 �2.86 0.18 po0.1 23.

a(Tittlemier et al., 2002).b(Harner and Shoeib, 2002).

A.D. St-Amand et al. / Atmospheric Environment 42 (2008) 2948–29582954

3.3. Seasonal variations in atmospheric PBDE

concentrations

Gaseous concentrations of PBDEs, althoughrelatively low, were generally higher in the summerand were significantly correlated with temperature(po0.001) (Table 1). Pseudo Clausius–Clapeyronplots were constructed and results are presented inTable 2, as well as Fig. 3S, which illustrates theresults for BDE-28, -47, -99 and -153. For mostPBDEs, the natural logarithm of partial pressurecorrelates significantly (po0.05) with the reciprocaltemperature. Exceptions are BDE-17, -85 and -183and this may be due to their very low concentrationsin the gas phase, which very often were belowdetection limit. Clausius-Clapeyron plots can alsobe used to calculate air-surface exchange enthalpies,if the system considered is at equilibrium. These

eous and particulate-bound concentrations with meteorological

PAH

Particulate-bound Gaseous Particulate-bound

0.064 0.184 0.001*(�)

0.905 0.110 0.048*(�)

0.045* 0.771 0.956

0.010* 0.471 0.587

0.222 0.005* 0.311

0.004* 0.002* 0.113

0.652 0.189 0.457

of correlation is also indicated.

pseudo Clausius-Clapeyron plots used to calculate enthalpies for

calc (kJ/mol) DHVAPa (kJ/mol) DHOA

b (kJ/mol)

5712.4 – 72.8

876.6 79.7 74.5

174.2 94.6 97.0

579.6 97.8 107.0

7711.8 – 102.0

574.8 108.0 91.1

775.3 102.0 105.0

8725.0 118.5 –

3711.4 110.0 98.2

6711.2 113.0 94.4

8713.0 118.0 89.5

ARTICLE IN PRESSA.D. St-Amand et al. / Atmospheric Environment 42 (2008) 2948–2958 2955

were calculated using their respective pseudoClausius-Clapeyron plots regression slopes and weresmaller than DHOA and DHVAP reported in theliterature (Harner and Shoeib, 2002; Tittlemieret al., 2002). However, it has been suggested byHoff et al. (1998) that Clausius–Clapeyron plotsand corresponding slopes can also be used to assessthe input of SVOCs from long-range transport andlocal volatilization. Slopes similar to DHOA orDHVAP slopes would indicate that local sourcesare important, whereas shallower slopes wouldindicate that long-range transport is the dominantcontribution process to the gaseous concentration(Hoff et al., 1998). In this study, the Clausius–Clapeyron plots drawn for PBDEs present slopes(and pseudo enthalpies) that are shallower thanthose predicted by the known vaporization enthal-pies. Furthermore, slopes for lighter PBDEs arehighly linear (high R2, low p-values) thus demon-strating that one process is dominant. Theseobservations indicate that long-range transport isthe dominant process contributing to gaseousconcentrations.

Particulate-bound PBDE concentrations weregenerally highest in the fall and spring seasons andwere not significantly correlated to temperature(Table 1). The lower concentrations encountered inwinter could be due to the atmospheric scavengingby snowfall, and could be re-released during thesnowmelt, thus causing an increased concentrationin air during the spring snowmelt (Gouin andHarner, 2003).

3.4. Seasonal variations in atmospheric PAH

concentrations

Seasonal trends for gaseous and particulate-bound PAH concentrations are not as straightfor-ward since meteorological effects are combined tousage patterns. Prior investigations have used massconcentration ratios (diagnostic ratios) to identifypossible atmospheric PAH sources (Schauer et al.,2003; Blasco et al., 2006). For example, the ratio ofconcentrations of anthracene/(anthracene+phenan-threne) can be used to assess possible combustion(incomplete combustion of organic matter) versuspetroleum sources (geochemical alterations of or-ganic matter). If this ratio is higher than 0.1 thencombustion sources dominate, whereas a ratiolower than 0.1 indicates prevalence of petroleumsources (Blasco et al., 2006). For particulate-boundPAH concentrations this ratio was always higher

than 0.1 suggesting combustion sources. In fact,particulate-bound PAH concentrations were gener-ally higher in the colder winter months as domesticheating increases (Fig. 2), and correlated signifi-cantly and negatively with temperature (po0.001)(Table 1). Furthermore, five and six rings PAHshave been linked to heating emissions (Schaueret al., 2003) which constituted, on average, 2

3of

particulate-bound PAH burden.PAH gaseous concentrations were found to be

more variable and were not significantly correlatedwith temperature (Table 1). For gaseous PAHconcentrations, the above diagnostic ratio waslower than 0.1 from February to mid-July 2004,then higher than 0.1 until December, when itdecreased below 0.1 until the end of April 2005,indicating both pyrogenic and petroleum origins.Pseudo Clausius–Clapeyron plots were also con-structed, even though gaseous acenaphthylene wasthe only PAH to correlate significantly with 1000/T,though positively (R2

¼ 0.31, po0.001). This maybe due to an increase in PAH emissions encounteredat colder temperatures. Other gaseous PAHs did notcorrelate significantly with temperature. Althoughhigher gaseous concentrations are usually expectedat higher temperatures, it is possible that theincrease emission during the winter season concealsthis effect. Pyrene is used in Fig. 3S to represent atypical PAH.

3.5. Other meteorological effects and emissions

sources

To distinguish between long-range and localemissions and meteorologically related effects, andexplain some of the variability encountered inatmospheric concentrations for both PBDEs andPAHs, back trajectories using HYSPLIT (Draxlerand Rolph, 2003) and wind roses were constructedfor each sampling sessions. Air mass back trajec-tories were computed for 96 h starting at the end ofeach sampling sessions. Their overall trajectorieswere then categorized according to 16 cardinalpoints: N, ENE, NE, NEE, E, ESE, SE, SSE, S,SSW, SW, WSW, W, WNW, NW, and NNW. Windroses were compiled using hourly weather datacollected at the Ottawa International Airport for theduration of the sampling period. Predominant winddirections were also categorized according to 16cardinal points. A thorough statistical analysisusing SYSTAT was performed to study the impactof air masses and wind direction. Both were not

ARTICLE IN PRESSA.D. St-Amand et al. / Atmospheric Environment 42 (2008) 2948–29582956

only defined categorically as above, but alsonumerically (degrees). These were also consideredwith other meteorological parameters such astemperature, relative humidity and wind speed toassess their impact (Table 1).

Gaseous PBDE concentrations were found to belinked to long-range air masses. In general, airmasses deriving from northerly directions resultedin lower PBDE gaseous concentrations, whereasthose deriving from the west and south had highPBDE gaseous concentration. These were found tocorrelate significantly with air mass directions bothcategorically (po0.001) and numerically (po0.05).These observations demonstrate that gaseousPBDEs found at the sanitary landfill could be fromlong-range transport, thus corroborating the ob-servations made using pseudo Clausius–Clapeyronplots. PBDE gaseous concentrations did not corre-late significantly with local wind direction, relativehumidity, and wind speed.

Particulate-bound PBDE concentrations corre-lated significantly with local wind direction (numer-ical po0.005) and to a lesser extent with air massback trajectories. Previous studies indicate thatSVOCs in the gas phase should travel farther andpossibly undergo long-range atmospheric transportto a larger extent than compounds associated toparticulates which are more prone to deposit ontoavailable surfaces (Palm et al., 2002; Shoeib andHarner, 2002). Winds from the north, west and eastwere linked to highest particulate-bound SVOCsconcentrations demonstrating the possible impact oflocal sources such as the city center and thecomposting, electronic refuse and general wasteareas at the sanitary landfill which are situatednorth, west and east, respectively, from the sam-pling station. Temperature, relative humidity andwind speed did not correlate significantly with theparticulate-bound PBDE concentrations; however,particulate-gas partitioning of most PBDEs wasshown to be temperature dependent (most com-pounds po0.001) (Fig. 4S found in Appendix). Astemperature decreases, exchange is thus shiftedtowards particulates reducing the overall mobilityand long-range transport potential.

Gaseous PAH concentrations were found tocorrelate significantly with local wind direction(po0.005), possibly indicating a local source. Ingeneral, winds blowing from the north and eastshowed highest gaseous PAH concentrations corre-sponding with the general direction of the highwayand suburbs of the city. No significant correlations

were found with temperature, relative humidity, airmass back trajectories or wind speed. Particulate-bound PAH concentrations did not correlatesignificantly with local wind direction, air massback trajectories or wind speed (p40.05). Particu-late-bound PAH concentrations correlated nega-tively with temperature (po0.001) which may bedue to an increased association with particulatesencountered at lower temperatures (as seen with thevan’t Hoff plots in Fig. 4S). A significant correlationwas also observed between PAH concentration andrelative humidity (po0.05).

It is difficult to study the possible trends inconcentrations in environmental media for PAHsby only looking at meteorological effects, withoutconsidering usage patterns and potential emissionsources. Not only there is a possible influence fromtemperature, wind direction may also play animportant role in the concentration of SVOCs asthey transport contaminants from the sources to thesampling site. These parameters may influence theparticulate-bound and gaseous concentrations withgreater impact than the spruce needle concentra-tions as SVOCs are accumulated through time inspruce needles and thus a sudden short-lived risein atmospheric concentrations may not be reflectedin the spruce needle.

4. Conclusions

This study was designed not only to survey PBDEand PAH concentrations in the environment at asanitary landfill, but also to consider the influence oftemperature, wind direction and emissions on theseconcentrations. Seasonal effects were noted in allmedia, which were due to weather related phenom-ena and/or usage patterns (higher concentration inair during the colder months, due to increaseddomestic heating and fuel consumption). PseudoClausius–Clapeyron plots showed a high correlationbetween gaseous concentration and temperature forPBDEs; corresponding slopes were shallower thanthose predicted from known DHOA or DHVAP valuesthus indicating that long-range transport is thedominant process contributing to gaseous concen-trations. Pseudo Clausius–Clapeyron plots failed toshow a temperature effect for PAHs because theywere overshadowed by increased emission to airduring the colder months. Finally, using statisticalanalysis with air mass back trajectories, wind rosesand meteorological parameters, it was possible todemonstrate that particulate-bound PBDEs, as well

ARTICLE IN PRESSA.D. St-Amand et al. / Atmospheric Environment 42 (2008) 2948–2958 2957

as particulate-bound and gaseous PAHs were fromlocal sources (sanitary landfill, highway, and sub-urbs), whereas gaseous PBDEs were mainly fromlong-range transport.

Acknowledgments

A.D.S. acknowledges the following: the NaturalSciences and Engineering Research Council ofCanada, the Ontario Graduate Scholarships inScience and Technology and the University ofOttawa for the financial support; and H. Shaw forthe help with sample preparation. P.M.M. andJ.M.B. thank the Natural Sciences Research Coun-cil of Canada for continuing support. The authorsthank the city of Ottawa for access to sampling siteand the NOAA Air Resources Laboratory (ARL)for the provision of the HYSPLIT transport anddispersion model used in this publication.

Appendix A. Supplementary materials

The online version of this article contains addi-tional supplementary data. Please visit doi:10.1016/j.atmosenv.2007.12.050

References

Agrell, C., ter Shure, A.F.H., Sveder, J., Bokenstrand, A.,

Larsson, P., Zegers, B.N., 2004. Polybrominated diphenyl

ethers (PBDEs) at a solid waste incineration plant I: atmo-

spheric concentrations. Atmospheric Environment 38,

5139–5148.

Alaee, M., Arias, P., Sjodin, A., Bergman, A., 2003. An overview

of commercially used brominated flame retardants, their

applications, their use patterns in different countries/regions

and possible modes of release. Environment International 29

(6), 683–689.

Blasco, M., Domeno, C., Nerin, C., 2006. Use of lichens as

pollution biomonitors in remote areas: comparison of PAHs

extracted from lichens and atmospheric particles sampled in

and around the Somport Tunnel (Pyrenees). Environmental

Science and Technology 40, 6384–6391.

Butt, C.M., Diamond, M.L., Truong, J., Ikonomou, M.G., Ter

Schure, A.F.H., 2004. Spatial distribution of polybrominated

diphenyl ethers in Southern Ontario as measured in indoor

and outdoor window organic films. Environmental Science

and Technology 38 (3), 724–731.

de Kok, T.M., Hogervorst, J.G., Briede, J.J., van Herwijnen,

M.H., Maas, L.M., Moonen, E.J., Driece, H.A., Kleinjans,

J.C., 2005. Genotoxicity and physicochemical characteristics

of traffic-related ambient particulate matter. Environmental

and Molecular Mutagenesis 46, 71–81.

de Wit, C.A., 2002. An overview of brominated flame retardants

in the environment. Chemosphere 46 (5), 583–624.

de Wit, C.A., Alaee, M., Muir, D.C.G., 2006. Levels and trends

of brominated flame retardants in the Arctic. Chemosphere

64, 209–233.

Deportes, I., Benoit-Guyod, J.-L., Zmirou, D., 1995. Hazard to

man and the environment posed by the use of urban waste

compost: a review. The Science of the Total Environment 172,

197–222.

Draxler, R. R. and Rolph, G. D. (2003). ‘‘HYSPLIT (HYbrid

Single-Particle Lagrangian Integrated Trajectory).’’ from

/http://www.arl.noaa.gov/ready/hysplit4.htmlS. NOAA Air

Rescources Laboratory, Silver Spring, MD.

Fitter, A.H., Hay, R.K.M., 2003. Environmental Physiology of

Plants. Academic Press, San Diego, CA, USA.

Gouin, T., Harner, T., 2003. Modelling the environmental fate of

the polybrominated diphenyl ethers. Environment Interna-

tional 29 (6), 717–724.

Hale, R.C., Alaee, M., Manchester-Neesvig, J.B., Stapleton,

H.M., Ikonomou, M.G., 2003. Polybrominated diphenyl

ether flame retardants in the North American environment.

Environment International 29 (6), 771–779.

Hale, R.C., La Guardia, M.J., Harvey, E., Gaylor, M.O.,

Mainor, T.M., 2006. Brominated flame retardant concentra-

tions and trends in abiotic media. Chemosphere 64, 181–186.

Harner, T., Shoeib, M., 2002. Measurements of octanol-air

partition coefficients (K-OA) for polybrominated diphenyl

ethers (PBDEs): predicting partitioning in the environ-

ment. Journal of Chemical and Engineering Data 47 (2),

228–232.

Hoff, R.M., Brice, K.A., Halsall, C.J., 1998. Nonlinearity in the

slopes of Clausius–Clapeyron plots for SVOCs. Environmen-

tal Science and Technology 32, 1793–1798.

Kylin, H., Sjodin, A., 2003. Accumulation of airborne hexa-

chlorocyclohexanes and DDT in pine needles. Environmental

Science & Technology 37, 2350–2355.

Larcher, W., 1996. Physiological Plant Ecology. Springer, New

York, NY, USA.

Mackay, D., Shiu, W.Y., Ma, K.-C., Lee, S.C., 2006. Handbook

of Physical–Chemical Properties and Environmental Fate of

Organic Chemicals. CRC Press/Taylor & Francis, Boca

Raton, FL.

Osako, M., Kim, Y.-J., Sakai, S.-i., 2004. Leaching of bromi-

nated flame retardants in leachate from landfills in Japan.

Chemosphere 57, 1571–1579.

Palm, A., Cousins, I.T., Mackay, D., Tysklind, M., Metcalfe, C.,

Alaee, M., 2002. Assessing the environmental fate of

chemicals of emerging concern: a case study of the

polybrominated diphenyl ethers. Environmental Pollution

117 (2), 195–213.

Sanderson, E.G., Raqbi, A., Vyskocil, A., Farant, J.-P., 2004.

Comparison of particulate polycyclic aromatic hydrocarbon

profiles in different regions of Canada. Atmospheric Envir-

onment 38, 3417–3429.

Schauer, C., Niessner, R., Poschl, U., 2003. Polycyclic aromatic

hydrocarbons in urban air particulate air matter: decadal and

seasonal trends, chemical degradation, and sampling artifacts.

Environmental Science and Technology 37, 2861–2868.

Schroter-Kermani, C., Kreft, D., Schilling, B., Herrchen, M.,

Wagner, G., 2006. Polycyclic aromatic hydrocarbons in pine

and spruce shoots-temporal trends and spatial distribution.

Journal of Environmental Monitoring 8, 806–811.

Shoeib, M., Harner, T., 2002. Using measured octanol-air

partition coefficients to explain environmental partitioning

ARTICLE IN PRESSA.D. St-Amand et al. / Atmospheric Environment 42 (2008) 2948–29582958

of organochlorine pesticides. Environmental Toxicology and

Chemistry 21 (5), 984–990.

Shoeib, M., Harner, T., Ikonomou, M., Kannan, K., 2004. Indoor

and outdoor air concentrations and phase partitioning of

perfluoroalkyl sulfonamides and polybrominated diphenyl ethers.

Environmental Science and Technology 38 (5), 1313–1320.

Simcik, M.F., Offenberg, J.H., 2006. Polycyclic aromatic hydro-

carbons in the Great Lakes. In: Hites, R.A. (Ed.), Persistent

Organic Pollutants in the Great Lakes. The Handbook

of Environmental Chemistry 5N. Springer, Berlin, Germany,

pp. 307–353.

St-Amand, A.D., Mayer, P.M., Blais, J.M., 2007. Modeling atmo-

spheric vegetation uptake of PBDEs using field measurements.

Environmental Science and Technology 41 (12), 4234–4239.

Su, Y., Lei, Y.D., Wania, F., Shoeib, M., Harner, T., 2006.

Regressing gas/particle partitioning data for polycyclic

aromatic hydrocarbons. Environmental Science and Technol-

ogy 40, 3558–3564.

Tittlemier, S.A., Halldorson, T., Stern, G.A., Tomy, G.T., 2002.

Vapor pressures, aqueous solubilities, and Henry’s law

constants of some brominated flame retardants. Environ-

mental Toxicology and Chemistry 21 (9), 1804–1810.

Wichmann, W., Kolb, M., Schmidt, C., Alawi, M., Bahadir, M.,

2006. Assessment of the environmental impact of landfill sites

with open combustion located in arid regions by combined

chemical and ecotoxicological studies. Chemosphere 65,

1778–1783.

Wilford, B.H., Harner, T., Zhu, J.P., Shoeib, M., Jones, K.C.,

2004. Passive sampling survey of polybrominated diphenyl

ether flame retardants in indoor and outdoor air in Ottawa,

Canada: implications for sources and exposure. Environ-

mental Science and Technology 38 (20), 5312–5318.