Distributions of Polycyclic Aromatic Hydrocarbons, AromaticKetones, Carboxylic Acids, and Trace Metals in Arctic Aerosols: Long-Range Atmospheric Transport, Photochemical Degradation/Production at Polar SunriseDharmendra Kumar Singh,† Kimitaka Kawamura,*,†,‡ Ayako Yanase,‡ and Leonard A. Barrie§,⊥

†Chubu Institute for Advanced Studies, Chubu University, Kasugai 487-8501, Japan‡Department of Chemistry, Tokyo Metropolitan University, Hachioji, Tokyo 192-0397, Japan§Bolin Centre Research, Stockholm University, Stockholm SE-106 91, Sweden

*S Supporting Information

ABSTRACT: The distributions, correlations, and source apportionment ofaromatic acids, aromatic ketones, polycyclic aromatic hydrocarbons (PAHs),and trace metals were studied in Canadian high Arctic aerosols. NineteenPAHs including minor sulfur-containing heterocyclic PAH (dibenzothio-phene) and major 6 carcinogenic PAHs were detected with a highproportion of fluoranthene followed by benzo[k]fluoranthene, pyrene, andchrysene. However, in the sunlit period of spring, their concentrationssignificantly declined likely due to photochemical decomposition. Duringthe polar sunrise from mid-March to mid-April, benzo[a]pyrene tobenzo[e]pyrene ratios significantly dropped, and the ratios diminishedfurther from late April to May onward. These results suggest that PAHstransported over the Arctic are subjected to strong photochemicaldegradation at polar sunrise. Although aromatic ketones decreased inspring, concentrations of some aromatic acids such as benzoic and phthalicacids increased during the course of polar sunrise, suggesting that aromatic hydrocarbons are oxidized to result in aromatic acids.However, PAHs do not act as the major source for low molecular weight (LMW) diacids such as oxalic acid that are largelyformed at polar sunrise in the arctic atmosphere because PAHs are 1 to 2 orders of magnitude less abundant than LMW diacids.Correlations of trace metals with organics, their sources, and the possible role of trace transition metals are explained.

1. INTRODUCTION

The Arctic, which is covered by the Eurasian and NorthAmerican continents, is known to receive polluted air massescontaining organic and inorganic contaminants from thenorthern midlatitudes by long-range atmospheric transport.1,2

Sea-to-air flux of marine organic materials are limited duringwinter due to the coverage of sea ice in the Arctic Ocean, andatmospheric transport of aerosols and their precursor gases areonly the sources of the arctic winter aerosols. In spring aroundmid-March when polar sunrise begins, photochemical reactionsmodify the atmospheric composition of arctic aerosols; forexample, formation of sulfate via gas-to-particle conversion isenhanced2,3 as well as secondary production of water-solubledicarboxylic acids.4,5 The previous polar sunrise experimentsshowed that organic pollutants transported to the Arctic frommidlatitudes of Eurasia, Asia, and North America are severelysubjected to photochemical oxidation in the arctic atmos-phere.2,6

Fu et al.1 reported that, during winter-spring season, suddenappearance of solar irradiance and long-range atmospherictransport are the key components regulating the chemical

composition of organic aerosols in the atmosphere of the Arctic.Furthermore, a distinct rise in ambient temperature from winterto spring substantially impacts the partitioning of semivolatileorganic compounds between gas and particle in the atmosphere.Aromatic hydrocarbons are one of the typical organic

pollutants emitted from fossil fuel combustion and biomassburning processes. They should be long-range transported in theatmosphere to the Arctic in winter and spring. Latitudinaldistributions of PAHs in the deep-sea sediments from theequatorial to northern North Pacific at 175°E transect showedthat concentrations of PAHs increase from the equatorial Pacificto the northern North Pacific nearby the Bering Sea,7 suggestingthat anthropogenic PAHs are long-range transported over thehigh altitudes in the Northern Hemisphere, although they aredeposited by wet/dry processes during the transport across theocean and are perched in the deep-sea ocean floor via

Received: March 30, 2017Revised: July 11, 2017Accepted: July 21, 2017Published: July 21, 2017

Article

pubs.acs.org/est

© 2017 American Chemical Society 8992 DOI: 10.1021/acs.est.7b01644Environ. Sci. Technol. 2017, 51, 8992−9004

sedimentation processes. Zhou et al.8 reported that interment ofPAHs by solid organics is a feasible mechanism to restrain theheterogeneous/multiphase reaction during long-range atmos-pheric transport for several days to weeks. Although theheterogeneous reactivity of surface-bound PAHs is exceedinglyfast in the atmosphere, the above-mentioned mechanism maysubstantially extend the lifetime of PAHs, allowing them toexperience long-range transport to distant locations.8

The PAHs that escape the wet/dry deposition in theatmosphere are long-range transported to the Arctic. Depositionof PAH from the atmosphere to snow and ice sheets haspersisted in the Canadian high Arctic over the last 20 years.9

However, there are only a few studies10,1 on PAHs that havebeen carried out in aerosols of the arctic atmosphere.In present study, we analyzed the samples of atmospheric

aerosols collected from Alert, Canada in the high Arctic todetermine PAHs employing a gas chromatography (GC) andGC/mass spectrometry (GC/MS). We also measured aromaticketones, aromatic carboxylic acids, and trace metals in the arcticaerosols. Here, we investigate seasonal and temporal variationsof these aromatic compounds with the variations in benzo[a]-pyrene to benzo[e]pyrene concentration ratios and carcinogenicPAHs from winter to early summer and discuss theirphotochemical behaviors. In addition to speciation anddistributions of detected organic species, we found uniquecorrelations of trace metals with organic species detected.

2. EXPERIMENTAL SECTION2.1. Aerosol Samples. Total suspended particles (TSP)

were collected at Alert (82.5°N; 62.3°W) alfresco adjacent tothe Special Study Trailer Laboratory (175 m a.s.l.) from 19February to 10 June 1991 employing a high volume air sampler(without a denuder). Quartz fiber filter (Pallflex 2500 QAT-UP,20 × 25 cm), which was precombusted at 450 °C for 3 h toeliminate potential organic contaminants. Samples werecollected on a weekly basis and field blank filters were procuredonce every 4 weeks. Average ambient surface air temperatureduring sample collection ranged from −34.9 °C in February to−1.9 °C in June. The sun fully rises to 24 h above the horizon byApril 1 and polar sunrise starts March 5. Before and aftersampling, filters were stored in a precleaned glass jar (150 mL)with a Teflon-lined screw cap. Samples were transported toTokyo in a cooler at temperature below 0 °C and stored indarkness at −20 °C until the analyses. Details of sampling siteand sampling method are provided elsewhere.11,4 Sincesampling phase (a week) is long, a positive artifact by theadsorption of gaseous organics on the quartz filter would beinsignificant and not alter their concentrations. We simulta-neously collected TSP samples using nitrocellulose filters for theanalyses of metals at the same site.2.2. Extraction and Derivatization. PAHs, aromatic acids

and ketones were extracted with 0.1 M KOH in methanol,containing ca. 5% distilled water from sample filters. Aprecleaned Whatmann GF/A filter was used for separation ofextracts by filtration. Under ultrasonication the filtrate wasfurther extracted with methanol and then dichloromethane.Rotary evaporator was employed for concentration of thecombined extracts under vacuum and divided into neutral andacidic fractions; the neutrals were extracted with n-hexanecontaining 10% dichloromethane (DCM) whereas carboxylicacids isolated by extraction with DCM after acidification with 6M HCl. Four fractions of (1) alkanes, (2) PAHs, (3) ketones/aldehydes, and (4) fatty alcohols were separated from neutral

fraction on a silica gel column chromatography. PAHs andaromatic ketones/aldehydes were determined using a CarloErba MEGA 5160 gas chromatograph (GC) equipped with anon-column injector, a fused silica HP-5 column and an FIDdetector and GC/mass spectrometer (Finnigan MAT ITS-40).12

The acidic fraction containing various types of carboxylicacids was derivatized with 14% BF3 in methanol tocorresponding methyl esters. The esters were isolated with n-hexane and then further separated into three fractions using asilica gel column chromatography; monocarboxylic acids,dicarboxylic acids and ketoacids, and hydroxy fatty acids.Carboxylic acid methyl esters were determined using a CarloErba MEGA 5160 GC as above.12 The desired compounds wereidentified by comparing the GC retention times with those ofauthentic standards. The compound identification was con-firmed by the examination of mass spectra obtained by a massspectrometer (Finnigan-MAT ITS-40). All the chemicalanalyses were completed by 1995.

2.3. Chemical Analysis of Trace Elements. Vanadium-(V), aluminum (Al) and manganese (Mn) were determinedusing instrumental neutron activation analysis (INAA) whereasother trace metals (Zn, Mg, Pb, Fe, Ni, Cu, and Ca) wereassessed by inductively coupled plasma emission (ICP)spectroscopy. Analysis with INAA was executed at theUniversity of Toronto Slowpoke Reactor using short irradiationof 1/8 filter in plastic vials followed by counting of the samplesin separate nonirradiated vials. Calibration is checked by analysisof National Institute of Standards and Technology (NIST) flyash standards. ICP analysis was conducted on the residue of 1/8of a filter. Filters were ashed at 475 °C and mixture of ultrapurehydrochloric and nitric acid were used for extraction. Finalextracts were prepared in 1 mL of concentrated HNO3 and 30mL distilled deionized water. All the metal analyses werecompleted by 1992.

2.4. Quality Control and Quality Assurance. Field blankfilters (n = 4) were analyzed for the above-mentioned organiccompound classes. However, any major peaks were detected forthe target compounds on GC chromatograms and GC/MStraces. The data shown here are rectified for the field blanks. Thedetection limits were typically <0.005 pg m−3. Recoveries forauthentic standards using the analytical procedure presented>80%. The data were not corrected for the recoveries. Analyticalerrors of major species by duplicate analyses of samples were<10%. On the basis of composite solutions of metals or onstandard reference materials such as NBS1648 urban particulatematter, NBS1633a trace elements in coal fly ash, and NRCCMESS-1 marine sediment reference material were performedthrough the ashing digestion and analysis procedure on aroutine basis for several quality control samples. Loss ofelements attributable to volatilization during ashing of thesample was less than 10%. Extraction efficiency of metals fromthe ash residue in the acid digestion step was better than 98%.With varimax rotation, principle component analysis (PCA)

was also performed for clarification of complex data that can beabridged by decreasing a set of variables, called factors orcomponents. Reducing the dimensionality for a set of variables,PCA can be applied as an appropriate statistical procedure.According to Galarneau,13 PCA separates observed ambientconcentrations corresponding to groups of covarying elementsand evaluating those groups to doubted source profiles.

Environmental Science & Technology Article

DOI: 10.1021/acs.est.7b01644Environ. Sci. Technol. 2017, 51, 8992−9004

8993

3. RESULTS AND DISCUSSION

Thirty aromatic species (19 PAHs, 6 carboxylic acids, and 5ketones) were identified in the arctic aerosol samples. Amongthem, aromatic acids were the most abundant compound class,followed by aromatic ketones and PAHs. The molecularformulas, ring numbers, and abbreviations of detected PAHsare given in Table 1.3.1. Seasonal and Temporal Variations in Concen-

trations and Molecular Distributions of PAHs, AromaticAcids, and Ketones with Total Carbon (TC). Temporalvariations in mass concentrations of PAHs and aromatic acidswith total carbon (TC) detected in the Alert aerosols are shownin Figure 1. The trend of TC mass concentrations showed a riseand fall in different sampling weeks, with the utmostconcentration (639 ng m−3) during the first week of samplingin February (before polar sunrise) and the lowest concentration(91 ng m−3) in the last week in June (well after polar sunrise).The mean concentration of TC was found to be 360 ± 176 ngm−3. We observed that all PAHs were not present in all thesampling weeks, higher concentrations of PAHs were obtainedin late February (before polar sunrise), and concentrationsgenerally decreased during the polar sunrise (Figure 1a). Theseresults may be caused by photochemical degradation of airbornePAHs into their derivatives with additional functional groups.1

The another reason is an increase in the concentration of ozone(O3), which could augment the ozonolysis in the atmosphere,leading to a decrease in particulate PAH concentrations.14 Incontrast, concentrations of aromatic acids were correspondinglyhigher even after polar sunrise with dominance of γ-(2,4-dimethylphenyl)butanoic acid (Figure 1b).Average mass concentrations of each PAH measured in total

suspended particulate matter before and after polar sunrise areranked in the following order, ANTH < FLUO < DBTHP <D[ah]A < BINAP < PERY < B[b]FLUO < CORO < B[a]A <B[a]P < B[ghi]FLA < B[ghi]P < B[e]P < INDP < PHEN <CHR < PYR < B[k]F < FLA (see Figure 2 and Table 1 forabbreviations). However, naphthalens, xylenes, and toluene

were not detected in aerosol phase. They should be abundantlypresent in gaseous phase.The lowest concentration was obtained for anthanthrene

(0.012 pg m−3), and the highest for fluoranthene (24.0 pg m−3).The observation that fluoranthene is most abundant PAH isconsistent with our previous study1 and with other studiescarried out during winter dense foggy episode in the centralIndo-Gangetic Plain of India.15,16 The highest average massconcentration of fluoranthene in particulate matter implies thatfossil fuel combustion and other combusted organic matter arethe principal sources, which are associated with higheranthropogenic activity.17 Predominance of fluoranthene (21%)followed by benzo[k]fluoranthene (20%), pyrene (12%) andchrysene (11%) is characterized in the molecular distributions ofPAHs in Alert aerosols. We found that PAHs, for instance, FLA,CHR, PYR, B[k]F, B[a]A, and INDP prevailed in the arcticaerosols (Figure 2). Dibenzothiophene (DBTHP) was alsoobserved in the Alert aerosols, however, its average massconcentration (0.087 pg m−3) was lower than all detected PAHs,except for ANTH and FLUO. DBTHP has been proposed as amarker compound released from the Canadian oil sand activityregion.18 Kelly et al.18 reported that distribution of PAHs in anoil slick produced on the melted snow surface was dominated bydibenzothiophene, phenanthrene, anthracene, fluoranthene,pyrene, benzanthracenes and chrysene.The mean concentrations of aromatic acids are ranked in the

following order; 2,6-naphalene dicarboxylic acid < salicylic acid< benzoic acid <2-carboxybenzaldehyde < phthalic acid < γ-(2,4-dimethylphenyl)butanoic acid. The lowest concentration ofaromatic acids for 2,6-naphalene dicarboxylic acid was 24.6 pgm−3, whereas the highest was for γ-(2,4-dimethylphenyl)-butanoic acid at 478 pg m−3. The reason for γ-(2,4-dimethylphenyl)butanoic acid to be much more abundantthan other organic species detected is not clear at this moment.It could be due to the presence of some specific dominant localsource (unlikely), long-range transport or photochemicalreactions (Figure 2). 9-Fluorenone, 9,10-anthracenedione, 4H-cyclopenta[def]phenanthren-4-one and benz[de]anthracen-7-

Table 1. Concentrations of Polycyclic Aromatic Hydrocarbons (PAHs) and Their Molecular Formula with Abbreviations andNumber of Rings

concentration (pgm−3)

PAHs abbreviation molecular formula ring number range (average ± STD)

fluorene FLUO C13H10 3 0−0.2 0.014 ± 0.045dibenzothiophene DBTHP C12H8S 3 0−0.8 0.087 ± 0.22phenanthrene PHEN C14H10 3 0−51 7.4 ± 13.7fluoranthene FLA C16H10 4 0−189 24.0 ± 52.7pyrene PYR C16H10 4 0−97.4 13.3 ± 29.1benzo[b]fluorene B[b]FLUO C17H12 4 0−7.4 0.94 ± 2.241,1′-binaphthalene BINAP C20H14 4 0−3.6 0.68 ± 1.18benzo[a]anthracene B[a]A C18H12 4 0−40.4 4.76 ± 11.7chrysene CHR C18H12 4 0−20.1 2.76 ± 6.56benzo[k]fluoranthene B[k]F C20H12 5 0−93.0 12.6 ± 27.2benzo[e]pyrene B[e]P C20H12 5 0−160 21.9 ± 46.6benzo[a]pyrene B[a]P C20H12 5 0−47.3 5.84 ± 13.4perylene PERY C20H12 5 0−24.6 3.17 ± 7.35benzo[ghi]fluoranthene B[ghi]FLA C18H10 5 0−14.7 0.92 ± 3.67dibenz[a,h]anthracene D[ah]A C22H14 5 0−44.0 5.6 ± 13.0benzo[ghi]perylene B[ghi]P C22H12 6 0−5.3 0.52 ± 1.48anthanthrene ANTH C22H12 6 0−37.9 5.3 ± 11.6indeno[1,2,3-cd]pyrene INDP C22H12 6 0−0.2 0.012 ± 0.049coronene CORO C24H12 7 0−17.0 2.01 ± 5.16

Environmental Science & Technology Article

DOI: 10.1021/acs.est.7b01644Environ. Sci. Technol. 2017, 51, 8992−9004

8994

one are known as oxidation products of PAHs. Diphenylme-

thanone is also called benzophenone. Their concentrations were

found to decrease according to the following order; 9-

fluorenone > benz(de)anthracen-7-one >9, 10-anthracenedione

>4H-cyclopenta(def)phenanthren-4-one > diphenylmethanone.

The highest concentration of 9-fluorenone was 39.3 pg m−3

(Figure 2). Oxygenated PAHs are produced in the atmospheric

reactions of PAHs with O3 and hydroxy radicals.19

Figure 1. Weekly variations in mass concentrations of total carbon (TC) with (a) PAHs and (b) aromatic acids detected in the Alert aerosols.

Environmental Science & Technology Article

DOI: 10.1021/acs.est.7b01644Environ. Sci. Technol. 2017, 51, 8992−9004

8995

Chen et al.20 reported the characteristics of secondary organicaerosol (SOA) production from naphthalene, 1-methylnaph-thalene and 2-methylnaphthalene under the conditions of highand low NOx and the absence of NOx in a chamber study andobserved that yields of SOA from naphthalene and methyl-naphthalenes are more with increasing yields of 1-methylnaph-thalene > naphthalene >2-methylnaphthalene. They concludedthat OH radicals, NOx levels, initial PAH/NO ratios, NO2/NOratios influence the system reactivity, and all affected the SOAformation from the PAH precursors. Similarly, Chan et al.21

stated that SOA was semivolatile under high-NOx andeffectively nonvolatile under low-NOx conditions, owing tothe greater fraction of ring-retaining outcomes produced underlow-NOx conditions. They reported that PAHs are estimated toyield SOA 3−5 times higher than light aromatic compoundsover photooxidation and PAHs can comprise up to 54% of thetotal SOA from the oxidation of diesel emissions, playing apotentially large source of urban SOA.The increased concentrations of aromatic acids from mid-

April to late April and from mid-May to late May potentiallysupport an augmented photochemical production in the Arcticand surroundings. The winter maximum of PAHs suggests thatthe Arctic receives air masses with polluted aerosols and theirprecursors produced from the midlatitudes through long-rangeatmospheric transport.1 According to Halsall et al.,10 PAHconcentrations during October to April were highest in highArctic Alert due to the predominance of haze. The multiple

peaks in aromatic acids are shown in May onward as a result ofplausible photochemical degradation of PAHs and other sourcesincluding local pollution. Aerosol elimination rates areminimum in winter because of the absence of solar radiationunder stagnant meteorological conditions with surface basedinversion.22

Figure 3 displays box-plots of monthly variations in totalconcentrations of PAHs and aromatic acids. Average concen-trations of ∑19-PAHs and ∑6-aromatic acids are 112 and 727pg m−3, respectively. A seasonal fluctuation was noticed in theconcentrations of PAHs. Before polar sunrise (late February-early March), box plots display higher concentrations of PAHsthan the warmer months (April-June). The median concen-tration of ∑PAHs during the cold episode was much higherthan the rest of the months. Enhanced intrusion of air massesoriginated from the Eurasian and North American continentswas due to meteorological conditions.10 Furthermore, theexistence of intense temperature inversions in the boundarylayer, predominantly due to the presence of sea ice in the ArcticOcean, inhibits the deposition and dispersal of pollutants.23 As aresult of these processes, atmospheric PAHs peak in winter.10

The seasonal pattern of variations in total mass concentration ofaromatic acids was opposite to that of PAHs (Figure 3);concentrations of ∑6-aromatic acids increased from lateFebruary to May and then decreased in June.

3.2. Relative Abundances of PAHs in Terms of Numberof Rings. Figure 4 shows the relative abundances of 3-, 4-, 5-, 6-,

Figure 2. Average mass concentrations of individual PAHs, aromatic acids and ketones detected in the Alert aerosols.

Environmental Science & Technology Article

DOI: 10.1021/acs.est.7b01644Environ. Sci. Technol. 2017, 51, 8992−9004

8996

and 7-aromatic ring PAHs in the Alert aerosols during lateFebruary to early June (before and after polar sunrise). Thepercentage contributions are 7% (3-rings), 48% (4-rings), 33%(5-rings), 10% (6-rings) and 2% (7-rings). Thus, the dominantcontributors to total PAHs in the Alert aerosols were 4-, 5- and6-ring PAHs, which is in agreement with previous research

carried out in the Canadian Arctic, Alert.1,10 High molecularweight PAHs are the main contributor in the particulate phase.24

Heavier PAHs are predominantly associated with the particulatephase at ambient temperatures normally in the Arctic; thepreponderance of PAHs (70−90%) are adsorbed on suspendedparticles whereas lighter PAHs (2−3 benzene rings) are

Figure 3. Temporal variations in the total mass concentrations of PAHs and aromatic acids.

Figure 4. Mass concentrations and relative contribution (%) of PAHs based on the number of aromatic rings.

Environmental Science & Technology Article

DOI: 10.1021/acs.est.7b01644Environ. Sci. Technol. 2017, 51, 8992−9004

8997

predominantly found in gas phase.24 Emission of PAHs to theatmosphere from heavy duty diesel engines are predominantly4-ringed structures whereas gasoline engines emit highermolecular weight PAHs with more ring structures.25 Woodburning and coal combustion sources are also possible in thearctic air mass source regions.16,1,10

The reduced height of the atmospheric surface-based mixinglayer, decreased atmospheric reactivity of PAH compounds,greater emissions (biomass, wood, and coal burning), weakersolar radiation flux, and increased atmospheric stability are likelyfactors contributing to a peak of PAHs in the winter season inthe Arctic.26,27 As a result of residential heating in winter,atmospheric emissions were found to increase marked by higher

levels of 4-ring PAHs,28 which supports the measured highestconcentration (48%) of 4-ring PAHs in the present study.Lowered atmospheric mixing height and strong stability in thelower arctic atmosphere together with a fall in ambienttemperature further indicate to an entrapment of pollutantsnear the ground surface.29

3.3. Speciation of Carcinogenic PAHs and Variations inthe Ratio of Benzo[a]Pyrene to Benzo[e]Pyrene. TheUnited States Environmental Protection Agency30 and Interna-tional Agency for Research on Cancer (IARC)31 classify 7-PAHssuch as B[a]A, B[a]P, B[b]F, B[k]F, CHR, D[ah]A, and INDPto be possible carcinogens for humans. Seven PAHs arerecognized as Group B2 carcinogens.32 In this study, six of the

Figure 5. (a) Variations in mass concentrations and relative contribution of carcinogenic PAHs and (b) temporal variations in the ratio of B[a]P toB[e]P detected in the Alert aerosols.

Environmental Science & Technology Article

DOI: 10.1021/acs.est.7b01644Environ. Sci. Technol. 2017, 51, 8992−9004

8998

seven carcinogenic PAHs (B[a]A, B[a]P, B[k]F, CHR, D[ah]A,and INDP) were detected. Figure 5a displays the seasonal massconcentrations of these six carcinogenic PAHs. The meanconcentration of carcinogenic PAHs ranked as follows: B[k]F >CHR > INDP > B[a]P > B[a]A > D[ah]A. The averageconcentration of these six carcinogenic PAHs in Alert aerosols is46.8 pgm−3, accounting for ∼41.7% of total concentration of 19PAHs measured. Most PAHs generate their products, forexample, nitro- and oxygenated-PAHs upon reaction withradicals and other chemicals (e.g., SO2, NOx, and O3) in theatmosphere, which are even more toxic.25 It was observed that ingeneral comparison with SO2, NO2 has a feebler correlation withparticle bound PAHs, because of their reactive nature andcomplex mechanism of source and sink.14 It was reported thatSO2 and NO2 someway apportion a common emission sourcesuch as vehicular emission.14

High solar radiation and ambient temperature enhance theformation of O3, consequently, the O3 level is noticed to becomehighest throughout summer.14 The O3 concentration increaseswith a rise in ambient temperature and solar radiation duringclear days.33 It was found that PAHs can react with O3 and NO2

through ozonolysis and nitration, respectively, forming productsthat are more reactive than the parent compounds.34

Conversely, it has been confirmed that ozonolysis can takeplace in a laboratory condition similar to the ambientatmosphere, subsequently forming various oxy-PAHs.34 Never-

theless, as a result of the severe complicated atmosphericconditions, findings on the real atmosphere have hitherto to beascertained. Particle-borne B[a]P with extra O3 display pseudo-first-order kinetics in terms of selective loss of B[a]P over B[e]P,and reactions with a liquid organic coating ensue by themechanism of Langmuir−Hinshelwood.8 Kwamena et al.35

investigated the surface-bound PAHs kinetics with O3 andsuggested that the O3 partitioning constant is a signifier of theO3-aerosol surface contact, being independent of the amounts ofPAHs adsorbed.Figure 5b displays progressive variations in the concentration

ratios of B[a]P to B[e]P in the Alert aerosols. B[a]P and B[e]Phave similar physical properties, since they are a pair of isomers.The ratio of B[a]P to B[e]P was used to understand the freshand aged inputs of PAHs from their different sources into theCanadian high arctic region. B[a]P relative to B[e]P shows thereactiveness or stability of PAHs. The ratios peaked on Marchfirst and April 19th weeks of the sampling period. During thepolar sunrise from mid-March to mid-April, the B[a]P to B[e]Pratios drop and further diminish from late April to June. Theseresults indicate that PAHs carried over the Arctic are subjectedto strong photochemical degradation at polar sunrise and after.Average mass concentration of B[e]P is higher than that ofB[a]P, which is in agreement with another study10 accomplishedin the Alert region. The B[a]P to B[e]P ratios ranged from 0.0to 0.9 with an average of 0.3 during the period of February to

Figure 6. Concentrations and molecular distributions of PAHs and aromatic acids before (Feb. 19−25, left) and after polar sunrise (Apr. 1−8, right).

Environmental Science & Technology Article

DOI: 10.1021/acs.est.7b01644Environ. Sci. Technol. 2017, 51, 8992−9004

8999

June. In the week of April 19, the B[a]P to B[e]P ratio isrelatively high (0.91), which may be associated with a localemission from the military base’s incineration although we donot have such records.It was reported that a higher concentration ratio of reactive/

stable PAHs is prominent in winter months, which is a sign ofnew inputs of PAHs into the Arctic. This specifies that onlypartial breakdown of B[a]P relative to B[e]P occurs in the arcticatmosphere in winter, with the accumulation of PAHs duringpossible long-range transport to the Arctic.10 PAHs, containing apredominantly toxic species B[a]P, are present in pristine areas,e.g., the Arctic and Antarctic regions, which are long-rangetransported from distant combustion sources as reported in fieldmeasurements and modeling studies.36 During PAH transport, aloss of PAHs takes place via both heterogeneous and gas-phasephoto-oxidation reactions. To date laboratory35 and modelingstudies37 propose that heterogeneous reactions may be themajor atmospheric loss process of PAHs.3.4. Effects of Polar Sunrise on PAHs and Aromatic

Acids at Alert. Concentrations and molecular distributions ofPAHs and aromatic acids before (February 19−25) and afterpolar sunrise (April 1−8) are displayed in Figure 6.Concentrations of all PAHs were found to significantly decreaseafter polar sunrise. Total mass concentration of the 19 PAHsmeasured was 850 pg m−3 before polar sunrise and, after polarsunrise, it became ∼27 times lower at 31.7 pg m−3. Thedominant PAHs, FLA, and B[k]F, became ∼36 and ∼18 timesless abundant after polar sunrise than those before polar sunrise.Similarly, concentrations of B[a]P and B[e]P are ∼53 and ∼43times lower than those before sunrise, respectively. The B[a]Pto B[e]P ratio decreased greatly during polar sunrise, stronglyconsistent with a process of photochemical degradation of PAHsin sunlight during polar sunrise in the Arctic. The prevailingPAHs are fluoranthene, benzo[k]fluoranthene, pyrene, phenan-threne, and chrysene (Figure 6).Before polar sunrise aromatic acids were detected in high

abundances, implying that they are formed by photochemicalprocesses in the midlatitudes and long-range transported to theArctic.1 Their concentrations decreased after polar sunrise. Incontrast, benzoic acid, salicylic acid, phthalic acid, and γ-(2,4-dimethylphenyl)butanoic acid did not show a significant declineafter polar sunrise. However, similar to aromatics, 2-carbox-ybenzaldehyde and 2,6-naphthalenedicarboxylic acid were foundto be degraded with concentrations ∼72 and ∼18 times lowerafter polar sunrise than before polar sunrise. This is probablydue to the oxidation of aldehyde group and naphthalenestructure in increasing sunlight.The concentrations of aromatic acids increased in mid-May to

late May. They are very likely formed by in situ photochemicaloxidation of organic precursors such as naphthalenes, toluene,and xylenes.38 Oxidation of PAHs can produce secondarycompounds such as fluorenone and phenanthroquinone duringcombustion and photo-oxidation processes. They can also beformed by the oxidation of phenanthrene or benzofluorenones,which are oxidation products of benzofluorene.39 Li et al.40

investigated the influence of methyl group to formation of SOAin the photooxidation of aromatic hydrocarbons under low NOxcondition. They concluded that oxidation products of methylgroup carbon of aromatic compounds have a lower rate ofpartitioning to the particle-phase than the products derived fromthe ring opening of aromatic hydrocarbons.3.5. Source Apportionment of Organic Aerosols.Many

investigators in the recent past have employed diagnostic ratios

of 3- to 7-ring PAHs categorized on the origin of varioussources, some of which have been summarized in Table S1.Pyrolytic products from coal and wood burning and diesel/gasoline engine exhausts, and aged aerosols of combustionemission origin are the major sources of PAHs and relatedcompounds observed in this study based on diagnostic ratios(Table S1). The average ratio of B[e]P/(B[e]P + B[a]P) is 0.65,suggesting that combustion-derived aerosols are more aged41

because the concentration ratio of B[e]P/(B[e]P + B[a]P) forfreshly emitted PAHs is equal to 0.50.42 The most abundantPAH’s diagnostic ratio, FLA/(FLA + PYR), is >0.5, indicatingthe contributions from coal, grass, and wood burning (TableS1).Principle component analysis provides the replacement of a

large set of intercorrelated original variables with reducednumber of independent variables or principal components.43 Anemission source can be recognized by these components orfactors. Eigen values >1 were considered for retention ofprincipal components. Principal components with greater than5% of total variance of data set were utilized as components.Loadings affected most PAHs (Table S2) and aromatic acids(Table S3) in each component and a value greater than 0.5 wasselected. SPSS (version 24) was used to enhance the variance ofthe squared elements in the column, a factor matrix wasgenerated (Tables S2 and S3). Three principal components orfactors for PAHs and two for aromatic acids were set by thescree tests and their component plots in rotated space aredisplayed in Figure S1a,b.Table S2 shows the outcomes of factor analysis on the

concentrations of total PAHs. Three factors explain 98.4%variability in the PAHs data. Factor 1 account for 62.4% of thetotal variance, which is loaded with B[b]FLUO, B[a]A, B[ghi]P,PYR, CORO, B[a]P, IndP, CHR, B[k]F, B[ghi]FLA, B[e]P,FLA, D[ah]A, and PHEN; these are indicators of coal andorganic matter combustion, traffic emission, coal, grass, andwood burning and their long-range transport.1,44 Factor 2explain 30.6% of the total variance, which is loaded with FLA,D[ah]A, PHEN, DBTH, FLUO, PERY, and ANTH. ThesePAHs are derived from traffic emission, coal, and organic mattercombustion via long-range transport.1,44 Factor 3 accounts for5.3% of the total variance and is loaded with only BINAP, whichis possibly transported long distances to the Arctic from low andmid latitudes.Table S3 shows the outcomes of factor analysis of

concentrations of total aromatic acids. Two factors accountfor 82.5% variability in the acid data. Factor 1 accounts for 59.3%of the total variance, which is loaded with 2,6-naphalenedi-carboxylic acid, 2-carboxybenzaldehyde, salicylic acid, andphthalic acid; these compounds are derived from the oxidationof various organic precursors containing aromatic structures.Phthalic acid was found to be a proxy to understand the organicaerosol formation via secondary oxidation.1 Factor 2 accountsfor 23.2% of the total variance and is loaded with γ-(2,4-dimethylphenyl)butanoic acid and benzoic acid, which are theproducts derived from motor exhausts and photochemicaldegradation of organic precursors. Benzoic acid has beenregarded as a primary pollutant released from exhausts ofmotors and a photochemical degradation of toluene, other alkylbenzenes, and naphthalenes emitted by automobiles.38,45

Oxygenated PAHs are produced via photo-oxidation of PAHswith oxidants (ozone, OH radicals, and nitrogen oxides) presentin the atmosphere.19 Moreover, these compounds have beenoriginated in brake lining wear particles, road dust, and

Environmental Science & Technology Article

DOI: 10.1021/acs.est.7b01644Environ. Sci. Technol. 2017, 51, 8992−9004

9000

emissions of particulate exhaust from heavy duty dieseltrucks.46,47

3.6. Speciation and Variation in Mass Concentrationsof Trace Elements and Their Correlations with DetectedOrganic Species. Trace elements were categorized in twogroups: major and minor elements. Major elements arecomprised of those metals whose mass concentrations are >3ng m−3 such as Al, Mg, Fe, and Ca while minor elements such asV, Zn, Pb, Ni, Cu, and Mn are considered as those withconcentrations <3 ng m−3. Figure 7 shows the massconcentrations with relative contribution (%) of major andminor trace elements detected in the Alert aerosols. We foundthat among the major trace elements, Al is most abundant(29%) followed by Mg (28%), Ca (27%), and Fe (16%),whereas Zn is the most abundant (43%) among the minorelements followed by Pb (20%), Cu (17%), Mn (12%), V (5%),and Ni (3%).Four major (Al, Mg, Fe, and Ca) and six minor (V, Zn, Pb, Ni,

Cu, and Mn) metal elements were selected for ensuring thecorrelations and their plausible role with detected organicspecies. Table S4 shows correlations of Mg and Pb with somePAHs (pg m−3) before (n = 5) and after polar sunrise (n = 11)during sampling period. We found that Mg and Pb are highlycorrelated (R2 = 0.67 to 0.94; p < 0.05) with FLA, PYR, CHR,and PHEN before polar sunrise whereas after polar sunrise thecorrelations are weak (R2 = 0.04 to 0.78; p > 0.05). TheDurbin−Watson statistic for all significantly correlated species isfound to be <2, which determines the well autocorrelationbetween species. The contrary correlations before and afterpolar sunrise suggest that photodegradation of PAHs that takesplace during polar sunrise and after. Reduction of metal massconcentration from “before” to “after polar sunrise” could be

explained by another reason for contrary correlations. Thehigher correlations before polar sunrise also suggest samesources of Mg, Pb, FLA, PYR, CHR, and PHEN by either long-range transport from midlatitude or local sources.Table S4 shows the correlations of 5 aromatic acids (γ-(2,4-

dimethylphenyl)butanoic acid, phthalic acid, 2-carboxybenzal-dehyde, salicylic acid, and 2,6-naphthalene dicarboxylic acid)with 7 trace metals (Mg, V, Mn, Al, Zn, Pb, and Ca) before andafter polar sunrise during sampling period. We observed thatthese 5 aromatic acids were highly correlated (R2 = 0.60 to 0.99;p < 0.05) with 7 trace metals before polar sunrise, whereas thecorrelations were declined (down to R2 = 0.0017 to 0.67; p >0.05) after polar sunrise, signifying photo oxidation/degradationof aromatic acids, signifying that those species were transportedfrom similar sources to the Arctic. We observed that γ-(2,4-dimethylphenyl)butanoic acid was negatively correlated with V(R2 = 0.88; p < 0.05), Mn (R2 = 0.95; p < 0.01), and Al (R2 =0.60; p = 0.06), indicating that those species were transportedfrom dissimilar source regions to the Arctic before polar sunrise(Table S4). Siois and Barrie48 reported that Al, Mn, and V arefound in soil component of aerosols in spring likely due to long-range transport of dust from Gobi desert.Positive correlations (R2 = 0.43 to 0.90; p = 0.39 to 0.05) of V,

Mg, Pb, Zn, Fe, Ca, and Ni with total carbon and n-docosanewere found (Figure S2a,b), proposing that those species werealso transported from similar source regions to the Arctic beforepolar sunrise. n-Alkanes such as n-docosane are typicallyascribed to the emission of fossil fuel.49 Coal combustion isalso contributed as a dominant source for n-docosane.50

PHEN, CHR, B[a]A, B[k]F, B[b]FLUO, FLA, B[e]P, andB[a]P also showed strong correlations with Mg, Ni, Zn, Pb, andCa (R2 = 0.44 to 0.89; p = 0.37 to 0.05) before polar sunrise

Figure 7. Concentrations with (%) relative contribution of major and minor trace metals detected in the Alert aerosols.

Environmental Science & Technology Article

DOI: 10.1021/acs.est.7b01644Environ. Sci. Technol. 2017, 51, 8992−9004

9001

(Figure S3). Similarly, aromatic acids (benzoic acid, salicylicacid, 2-carboxybenzaldehyde, phthalic acid, 2,6-naphthalenedi-carboxylic acid, and γ-(2,4-dimethylphenyl)butanoic acid) gavegood correlations (R2 = 0.41 to 0.64; p = 0.38 to 0.05) with tracemetals (mainly Fe, Cu, Ni, Zn, Ca, Mg, and Pb) as shown inFigure S4. Figure S4 depicts fair correlations of benzoic acidwith Fe (R2 = 0.58; p < 0.05) and γ-(2,4-dimethylphenyl)-butanoic acid with Cu (R2 = 0.64; p < 0.05). On the basis ofthese correlations, we can propose here a plausible role of Feand Cu as important reagents in Fenton chemistry with theseorganic acids in dark reaction (before polar sunrise).All 5 aromatic ketones (diphenylmethanone, 4H-cyclopenta-

[def]phenanthren-4-one, 9,10-anthracenedione, benz[de]-anthracen-7-one, and 9-fluorenone) are also highly correlated(R2 = 0.41 to 0.97; p = 0.36 to 0.01) with Mg, Ni, Zn, Pb, Zn, Ca,and Mn. We found strong correlations of 4H-cyclopenta[def]-phenanthren-4-one with 9,10-anthracenedione (R2 = 0.99) andbenz[de]anthracen-7-one (R2 = 0.92; p < 0.05). 9,10-Anthracenedione had also a strong correlation (R2 = 0.94; p <0.05) with benz[de]anthracen-7-one as shown in Figure S5. Theabove correlations signify that these species are also originatedfrom similar sources to the Arctic before polar sunrise.Vanadium (V) and aromatic acids are useful tracers of fossil

fuel combustion, while V is in part released from natural source,for example, wind-blown dust.51 The detected trace transitionmetals may be interesting to better understand the role oftransition metals such as Fe, Cu, and Mn in the interaction withorganic compounds (acid, ketones, and their derivatives),because a recent aerosol study52 from the central Indo-GangeticPlain reported strong correlations of water-soluble organiccarbon with transition metals and stated the role of Fentonreagent (Fe, Cu/H2O2) in the formation of secondary organicaerosols. Table S4 also displays the correlations of 6carcinogenic PAHs with Ni, a carcinogenic transition metal(cancer slope factor = 0.84)41 before and after polar sunrise inthe Alert aerosols. We found stronger correlations (R2 = 0.86 to0.92; p < 0.05) after than before polar sunrise, suggesting thatthese (6 PAHs and Ni) carcinogenic species were evolved fromthe same sources. It was reported that Ni primarily originatesfrom total vehicular emissions as a result of burning oflubricating oil.53 Cu, Zn, Ni, and Fe primarily originate fromanthropogenic sources, for example, industries, petroleum, coalcombustion, and fine soil dust resuspension.54

In conclusion, higher concentrations of PAHs and aromaticacids in the Alert aerosols before polar sunrise provide clues tobetter understand the role of photochemical processes in themid latitudes and long distance transport to the Arctic. Lowerconcentrations of PAHs and aromatic acids after polar sunrisesuggest a photochemical degradation of combustion-derivedPAHs and secondary production of organic aerosols in thepresence of oxidants in the Arctic atmosphere. 2,6-Naphthale-nedicarboxylic acid declined from late April onward like otherPAHs, thus we can conclude that dicarboxylic acids (e.g., oxalicacid), and their derivatives are not photochemical products ofPAHs at arctic polar sunrise. The possible sources of PAHs arecoal and organic matter combustion, traffic emission, coal/grass/wood burning, and long-range transport, whereas severalaromatic acids were the photochemical oxidation products ofaromatic compounds derived from various sources includingmotor exhausts. Predominance of fluoranthene (21%), γ-(2,4-dimethylphenyl)butanoic acid (66%), and 9-fluorenone (39%)demonstrated the highest contributor to their respective class ofcompounds (PAHs, aromatic acids, and ketones). Relative

contribution of 4-ring PAHs in total PAHs showed the highestvalue (48%) followed by 5-, 6-, 3-, and 7-ring PAHs. Lowestdegree of sulfur containing heterocyclic PAH (dibenzothio-phene) and high abundances of 6-carcinogenic PAHs wereidentified in the Alert aerosols. A comparison of correlations ofaerosol organics and metals before and after polar sunriseprovides independent evidence that in situ production ordestruction of organics is occurring for many compounds duringpolar sunrise and after. This process weakens correlationsbetween organics and metals during the sunlit period.

■ ASSOCIATED CONTENT*S Supporting InformationThe Supporting Information is available free of charge on theACS Publications website at DOI: 10.1021/acs.est.7b01644.

Table S1. Diagnostic ratios and possible sources of PAHs.Table S2. Principal component analysis. Table S3.Principal component analysis. Table S4. Correlations oftrace metals with PAHs and aromatic acids detectedbefore and after polar sunrise in the Alert aerosols. FigureS1. Component plots in rotated space for (a) PAHs and(b) aromatic acids. Figures S2−S5. Correlations of tracemetals. (PDF)

■ AUTHOR INFORMATIONCorresponding Author*Phone: +81-568-51-9330; fax: +81-568-51-4736; e-mail:kkawamura@isc.chubu.ac.jp (K.K.).ORCIDDharmendra Kumar Singh: 0000-0003-1428-2849Present Address⊥L.A.B.: McGill University, Department of Atmospheric andOceanic Sciences, 801 Sherbrooke St. W., Montreal H3A 2K6,Quebec, Canada.NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTSThis study was partly endorsed by the Japanese Ministry ofEducation, Culture, Sports, Science, and Technology (MEXT)and the Japan Society for the Promotion of Science (JSPS) bygrant-in-aid Nos. 17340166, 19204055, and 24221001.

■ REFERENCES(1) Fu, P.; Kawamura, K.; Barrie, L. A. Photochemical and othersources of organic compounds in the Canadian high Arctic aerosolpollution during winter-spring. Environ. Sci. Technol. 2009, 43 (2),286−292.(2) Barrie, L. A. Arctic air pollution: an overview of currentknowledge. Atmos. Environ. 1986, 20, 643−663.(3) Barrie, L.A.; Platt, U. Arctic tropospheric chemistry: an overview.Tellus, Ser. B 1997, 49, 450−454.(4) Kawamura, K.; Kasukabe, H.; Barrie, L. A. Source and reactionpathways of dicarboxylic acids, ketoacids and dicarbonyls in arcticaerosols: One year of observations. Atmos. Environ. 1996, 30 (10−11),1709−1722.(5) Kawamura, K.; Kasukabe, H.; Barrie, L. A. Secondary formation ofwater-soluble organic acids and α-dicarbonyls and their contributionsto total carbon and water-soluble organic carbon: Photochemical agingof organic aerosols in the Arctic spring. J. Geophys. Res. 2010, 115,D21306.(6) Quinn, P. K.; Bates, T. S.; Baum, E.; Doubleday, N.; Fiore, A. M.;Flanner, M.; Fridlind, A.; Garrett, T. J.; Koch, D.; Menon, S.; et al.

Environmental Science & Technology Article

DOI: 10.1021/acs.est.7b01644Environ. Sci. Technol. 2017, 51, 8992−9004

9002

Short-lived pollutants in the Arctic: their climate impact and possiblemitigation strategies. Atmos. Chem. Phys. 2008, 8, 1723−1735.(7) Ohkouchi, N.; Kawamura, K.; Kawahata, H. Distributions of three-to seven-ring polynuclear aromatic hydrocarbons on the deep sea floorin the central Pacific. Environ. Sci. Technol. 1999, 33, 3086−3090.(8) Zhou, S.; Lee, A. K. Y.; McWhinney, R. D.; Abbatt, J. P. D. Burialeffects of organic coatings on the heterogeneous reactivity of particle-borne benzo[a]pyrene (BaP) toward ozone. J. Phys. Chem. A 2012, 116,7050−7056.(9) Peters, A. J.; Gregor, D. J.; Teixeira, C. F.; Jones, N. P.; Spencer, C.The Recent Depositional Trend of Polycyclic Aromatic Hydrocarbonsand Elemental Carbon to the Agassiz Ice Cap, Ellesmere Island,Canada. Sci. Total Environ. 1995, 160-161, 167−179.(10) Halsall, C. J.; Barrie, L. A.; Fellin, P.; Muir, D. C. G.; Billeck, B.N.; Lockhart, L.; Rovinsky, Fy.; Kononov; Ey; Pastukhov, B. Spatial andtemporal variation of polycyclic aromatic hydrocarbons in the Arcticatmosphere. Environ. Sci. Technol. 1997, 31 (12), 3593−3599.(11) Barrie, L. A.; Li, S. M.; Toom, D. M.; Landsberger, S.; Sturges, W.Measurements of aerosol and gaseous halogens, nitrates and sulphuroxides by denuder and filter systems during Polar Sunrise Experiment.J. Geophys. Res. 1992, 99, 25453−25468.(12) Kawamura, K. Land derived lipid class compounds in the deep-sea sediments and marine aerosols from North Pacific. InBiogeochemical Processes of Ocean Flux in the Western Pacific; Sakai, H.,Nozaki, Y., Eds.; Terra Scientific Publishing Co.: Tokyo, 1995, pp 31−51.(13) Galarneau, E. Source specificity and atmospheric processing ofairborne PAHs: Implications for source apportionment. Atmos. Environ.2008, 42, 8139−8149.(14) Tham, Y. W. F.; Takeda, K.; Sakugawa, H. Exploring thecorrelation of particulate PAHs, sulfur dioxide, nitrogen dioxide andozone, a preliminary study. Water, Air, Soil Pollut. 2008, 194, 5−12.(15) Singh, D. K.; Sharma, S.; Habib, G.; Gupta, T. Speciation ofatmospheric polycyclic aromatic hydrocarbons (PAHs) present duringfog time collected submicron particles. Environ. Sci. Pollut. Res. 2015, 22(16), 12458−12468.(16) Singh, D. K.; Gupta, T. Effect through inhalation on humanhealth of PM1 bound polycyclic aromatic hydrocarbons collected fromfoggy days in northern part of India. J. Hazard. Mater. 2016, 306, 257−268.(17) Mouwerik, M. V.; Stevens, L.; Seese, M. D.; Basham, W.Environmental contaminants encyclopaedia fluoranthene entry na-tional park service, water resource divisions, water operations branch,1201 Oakridge Drive, Suite 250, Fort Collins, Colorado 80525, 1997.(18) Kelly, E. N.; Short, J. W.; Schindler, D. W.; Hodson, P. V.; Ma,M.; Kwan, A. K.; Fortin, B. L.Oil sands development contributespolycyclic aromatic compounds to the Athabasca River and itstributaries. Proc. Natl. Acad. Sci. U. S. A. 2009, 106, 22346−22351.(19) Finlayson-Pitts, B.; Pitts, J. N. Tropospheric air pollution: ozone,airborne toxics and polycyclic aromatic hydrocarbons and particles.Science 1997, 276, 1045−1051.(20) Chen, C.-L.; Kacarab, M.; Tang, P.; Cocker, D. R., III SOAformation from naphthalene, 1-methylnaphthalene, and 2-methylnaph-thalene photooxidation. Atmos. Environ. 2016, 131, 424−433.(21) Chan, A. W. H.; et al. Secondary organic aerosol formation fromphotooxidation of naphthalene and alkylnaphthalenes: implications foroxidation of intermediate volatility organic compounds (IVOCs).Atmos. Chem. Phys. 2009, 9, 3049−3060.(22) Shaw, G. E. The Arctic haze phenomenon. Bull. Am. Meteorol.Soc. 1995, 76 (12), 2403−2413.(23) Barrie, L. A. Polar Snow. In Chemical Exchange Between theAtmosphere and; Wolff, E. W., Bales, R. C., Eds.; NATO ASI Series I:Global Environmental Change 1996, 43, 93−129.10.1007/978-3-642-61171-1_5(24) Bi, X.; Sheng, G.; Peng, P.; Chen, Y.; Zhang, Z.; Fu, J.Distribution of particulate- and vapor-phase n-alkanes and polycyclicaromatic hydrocarbons in urban atmosphere of Guangzhou, China.Atmos. Environ. 2003, 37 (2), 289−298.

(25) Mohanraj, R.; Solaraj, G.; Dhanakumar, S. Fine particulate phasePAHs in ambient atmosphere of Chennai metropolitan city, India.Environ. Sci. Pollut. Res. 2011, 18, 764−771.(26) European Commission. Ambient Air Pollution by PAH. PositionPaper European Commission.http://ec.europa.eu/environment/air/pdf/pp_pah.pdf, 2001.(27) Hong, H.; Yin, H.; Wang, X.; Ye, C. Seasonal variation of PM10-bound PAHs in the atmosphere of Xiamen, China. Atmos. Res. 2007, 85(3−4), 429−441.(28) Sklorz, M.; Schnelle-Kreis, J.; Liu, Y.; Zimmermann, R.; Orasche,J. Daytime resolved analysis of polycyclic aromatic hydrocarbonsinurban aerosol samplesimpact of sources and meteorologicalconditions. Chemosphere 2007, 67, 934−943.(29) Ravindra, K.; Sokhi, R.; Van Grieken, R. Atmospheric polycyclicaromatic hydrocarbons: Source attribution, emission factors andregulation. Atmos. Environ. 2008, 42 (13), 2895−2921.(30) US EPA (United States Environmental Protection Agency).Reference dose description and use in health risk assessments.Background Document 1A. Integrated risk information system 1994.(31) IARC (International Agency for Research on Cancer). Report ofthe advisory group to review the amended preamble to the IARCmonographs. 2006, (IARC 4 Int. Rep. No. 06/001).(32) US EPA, Provisional Guidance for Quantitative Risk Assessmentof PAH. US Environmental Protection Agency, EPA/600/R-93/089,1993.(33) Pudasainee, D.; Sapkota, B.; Shrestha, M. L.; Kaga, A.; Kondo, A.;Inoue, Y. Ground level ozone concentrations and its association withNOx and meteorological parameters in Kathmandu valley, Nepal.Atmos. Environ. 2006, 40, 8081−8087.(34) Pitts, J. N., Jr.; Paur, H.-R.; Zielinska, B.; Arey, J.; Winer, A. M.;Ramdahl, T.; Mejia, V.; et al. Factor influencing the reactivity ofpolycyclic aromatic hydrocarbons adsorbed on filters and ambientPOM with ozone. Chemosphere 1986, 15, 675−685.(35) Kwamena, N.-O. A.; Staikova, M. G.; Donaldson, D. J.; George, I.J.; Abbatt, J. P. D. Role of the aerosol substrate in the heterogeneousozonation reactions of surface-bound PAHs. J. Phys. Chem. A 2007, 111,11050−11058.(36) Sofowote, U. M.; Hung, H.; Rastogi, A. K.; Westgate, J. N.;Deluca, P. F.; Su, Y.; McCarry, B. E. Assessing the long-range transportof PAH to a sub-Arctic site using positive matrix factorization andpotential source contribution function. Atmos. Environ. 2011, 45, 967−976.(37) Kwamena, N.-O. A.; Clarke, J. P.; Kahan, T. F.; Diamond, M. L.;Donaldson, D. Assessing the Importance of Heterogeneous Reactionsof Polycyclic Aromatic Hydrocarbons in the Urban Atmosphere Usingthe Multimedia Urban Model. Atmos. Environ. 2007, 41, 37−50.(38) Fine, P. M.; Chakrabarti, B.; Krudysz, M.; Schauer, J. J.; Sioutas,C. Diurnal variations of individual organic compound constituents ofUltrafine and accumulation mode particulate matter in the Los Angelesbasin. Environ. Sci. Technol. 2004, 38 (5), 1296−1304.(39) Allen, J. O.; Durant, J. L.; Dookeran, N. M.; Taghizadeh, K.;Lafleur, A. L.; Smith, K. A.; Sarofim, A. F. Measurement of oxygenatedpolycyclic aromatic hydrocarbons associated with a size-segregatedurban aerosol. Environ. Sci. Technol. 1997, 31, 2064−2070.(40) Li, L.; Qi, Li; Cocker, D. R., III Contribution of methyl group tosecondary organic aerosol formation from aromatic hydrocarbonphotooxidation. Atmos. Environ. 2017, 151, 133−139.(41) Lammel, G.; Dvorska, A.; Klanova, J.; Kohoutek, J.; Kukucka, P.;Prokes, R.; Sehili, A. M. Long-range atmospheric transport of polycyclicaromatic hydrocarbons is worldwide problem-results from measure-ments at remote sites and modelling. Acta Chim. Slov. 2015, 62 (3),729−735.(42) Grimmer, G.; Jacob, J.; Naujack, K. W. Profile of the polycyclicaromatic compounds from crude oils-inventory by GC GC-MS. PAH inenvironmental materials, part 3. Fresenius. Fresenius' Z. Anal. Chem.1983, 314, 29−36.(43) Thurston, G. D.; Spengler, J. D. A quantitative assessment ofsource contributions to inhalable particulate matter pollution inmetropolitan Boston. Atmos. Environ. 1985, 19, 9−25.

Environmental Science & Technology Article

DOI: 10.1021/acs.est.7b01644Environ. Sci. Technol. 2017, 51, 8992−9004

9003

(44) Meharg, A.; Wright, J.; Dyke, H.; Osborn, D. Polycyclic aromatichydrocarbon (PAH) dispersion and deposition to vegetation and soilfollowing a large scale chemical fire. Environ. Pollut. 1998, 99, 29−36.(45) Suh, I.; Zhang, R.; Molina, L. T.; Molina, M. J. Oxidationmechanism of aromatic peroxy and bicyclic radicals from OH-toluenereactions. J. Am. Chem. Soc. 2003, 125, 12655−12665.(46) Rogge, W. F.; Hil demann, L. M.; Mazurek, M. A.; Cass, G. R.;Simoneit, B. R. T. Simoneit B.R.T. Sources of fine organic aerosol: 2. nocatalyst and catalyst equipped d automobiles and heavy-duty dieseltrucks. Environ. Sci. Technol. 1993, 27, 636−651.(47) Enya, T.; Suzuki, H.; Watanabe, T.; Hirayama, T.; Hisamatsu, Y.3-Nitrobenzanthrone, a powerful bacterial mutagen and suspectedhuman carcinogen found in diesel exhaust and airborne particulates.Environ. Sci. Technol. 1997, 31, 2772−2776.(48) Sirois, A.; Barrie, L. A. Arctic lower tropospheric aerosol trendsand composition at Alert, Canada: 1980−1995. J. Geophys. Res. 1999,104, 11599−11618.(49) Schauer, J. J.; Kleeman, M. J.; Cass, G. R.; Simoneit, B. R. T.Measurement of emissions from air pollution sources. 5. C1−C32organic compounds from gasoline-powered motor vehicles. Environ.Sci. Technol. 2002, 36, 1169−1180.(50) Zhang, Y.; Schauer, J. J.; Zhang, Y.; Zeng, L.; Wei, Y.; Liu, Y.;Shao, M. Characteristics of particulate carbon emissions from real-world Chinese coal combustion. Environ. Sci. Technol. 2008, 42, 5068−5073.(51) Pacyna, J. M. Chemical tracers of the origins of arctic airpollution. In Pollution of the Arctic Atmosphere; Sturges, W. T., Ed.;Elsevier Science Publishers: New York, 1991.(52) Singh, D. K.; Gupta, T. Role of transition metals with watersoluble organic carbon in the formation of secondary organic aerosoland metallo-organics in PM1 sampled during post monsoon and pre-winter time. J. Aerosol Sci. 2016, 94, 56−69.(53) Singh, D. K.; Gupta, T. Source apportionment and riskassessment of PM1 bound trace metals collected during foggy andnon-foggy episodes at a representative site in the Indo-Gangetic plain.Sci. Total Environ. 2016, 550, 80−94.(54) Singh, D. K.; Lakshay; Gupta, T. Field performance evaluationduring fog-dominated wintertime of a newly developed denuder-equipped PM1 sampler. Environ. Sci. Pollut. Res. 2014, 21 (6), 4551−4564.

Environmental Science & Technology Article

DOI: 10.1021/acs.est.7b01644Environ. Sci. Technol. 2017, 51, 8992−9004

9004