基于pmf的污染源解析-

Available online at www.sciencedirect.com

Journal of Environmental Sciences

www.jesc.ac.cn

Sizedistribution,characteristicsandsourcesofheavymetalsinhazeepisodinBeijing

JingchunDuan1,JihuaTan2,∗,JimingHao3,FaheChai1

1.StateKeyLaboratoryofEnvironmentalCriteriaandRiskAssessment,ChineseResearchAcademyofEnvironmentalSciences,Beijing100012,China.E-mail:duanjc@craes.org.cn

2.UniversityofChineseAcademyofSciences,Beijing100049,China3.SchoolofEnvironment,TsinghuaUniversity,Beijing100084,China

articleinfo

Articlehistory:

Specialissue:ProgressandprospectsofatmosphericenvironmentalscienceinChina

Keywords:

sourceapportionment;positivematrixfactorization;PM2.5;watersolubleion;healtheffect

DOI:10.1016/S1001-0742(13)60397-6

abstract

Sizesegragatedsampleswerecollectedduringhighpollutedwinterhazedaysin2006inBeijing,China.Twentynineelementsand9watersolubleionsweredetermined.HeavymetalsofZn,Pb,Mn,Cu,As,Cr,Ni,VandCdweredeeplystudiedconsideringtheirtoxiceffectonhumanbeing.Amongtheseheavymetals,thelevelsofMn,AsandCdexceededthereferencevaluesofNationalAmbientAirQualityStandard(GB3095-2012)andguidelinesofWorldHealthOrganization.Byestimation,highpercentageofatmosphericheavymetalsinPM2.5indicatesitisaneffectivewaytocontrolatmosphericheavymetalsbyPM2.5controlling.Pb,Cd,andZnshowmostlyinaccumulationmode,V,MnandCuexistmostlyinbothcoarseandaccumulationmodes,andNiandCrexistinallofthethreemodes.Consideringthehealtheffect,thebreakthroughratesofatmosphericheavymetalsintopulmonaryalveoliare:Pb(62.1%)>As(58.1%)>Cd(57.9%)>Zn(57.7%)>Cu(55.8%)>Ni(53.5%)>Cr(52.2%)>Mn(49.2%)>V(43.5%).PositivematrixfactorizationmethodwasappliedforsourceapportionmentofstudiedheavymetalscombinedwithsomemarkerelementsandionssuchasK,As,SO42−etc.,andfourfactors(dust,vehicle,agedandtransportation,unknown)areidentifiedandthesizedistributioncontributionofthemtoatmosphericheavymetalsarediscussed.

Introduction

Urbanairpollutionisaworldwideproblemcausingavarietyofriskstohumanhealth.FastgrowingcitiesindevelopingcountriessuchasChinaarestrongestaffectedbyurbanairpollutionshowinghighestburdenofdis-eases(Cohenetal.,2004;Chenetal.,2011).TheWorldHealthOrganization(WHO)estimatedthatannuallyabout300,000peopleprematurelydieduetourbanairpollutioninChina.

Atmosphericheavymetalscanimposealongtermbur-denonbiogeochemicalcyclingintheecosystem(Kellyetal.,1996;Nriagu,1988;NriaguandPacyna,1988).Among

∗Corresponding

author.E-mail:tanjh@ucas.ac.cn

them,As,Cr,Ni,Pb,Zn,Cu,VandCdarecarcinogenic,AsandCdarepotentiallymutagenic,PbandHgarefetaltoxic(Cheng,2003;Heetal.,2001).Inrecentyears,theheavymetalpollutionaccidentshavebeenreportedfrequentlyinChina(Zhou,2011).InFebruary2011,theStateCouncilofficiallyapprovedthe“12thFive-YearPlan”forcomprehensivepreventionandcontrolofheavymetalspollution.AnewNationalAmbientAirQualityStandard(NAAQS)(GB3095-2012)waspromulgatedinChinain2012,anditwillbeformallyimplementedsinceJanuary1,2016.InadditiontotighteningthePblimit,forthefirsttime,NAAQS(GB3095-2012)includedinthelimitsofCd,Hg,As,Cr(VI)inappendixasareferenceforlocalgovernmentstosetuplocalambientairqualitystandardsinduetime.Atmosphericemissionisoneofimportantwaysofheavymetalpollution;however,insufficientattention

190

JournalofEnvironmentalSciences2014,26(1)189–196

hasbeenpaidtoitinChina.

Knowledgeofthesizedistributionofparticulatematter(PM)isnotonlyvitalinunderstandingitseffectsonhumanhealth,itssourcesandtransformationprocessesduringatmospherictransport,butalsoforestimationthedrydepositionandlightextinctionoftheaerosol(Trijonis,1983;Pakkanenetal.,2001;Salmaetal.,2005;Tangetal.,2006;Duanetal.,2005,2007,2012;Tanetal.,2009a;Taoetal.,2012a,2012b).However,fewresearchonsizedistributionofinorganiccomponentsinaerosolhavebeencarriedoutinChina,especiallyforheavymetalsinBeijing(Ningetal.,1996),particularlyforaerosoldowntoultrafinefraction.

BeijingisthecapitalofChinawithapopulationofover19.61million.SeriousairpollutionhasbeenreportedinBeijing,andgreateffortshavebeenmadetoimprovetheairquality,especiallysince1998.Fromthenon,theairpollutioninBeijinghaschangedfromsimpletypeofcoalburningtomixedtypeoftrafficexhaustandcoalburningcontributedbyimprovementsofindustrialandenergystructure,usageofcleanenergyincludingnaturegasandcleancoal,reductioninpollutionsources,andimplementationofadvancedenvironmentalstandards(Zhangetal.,2011).Inthisstudy,sourceapportionmentofheavymetalswasfirstemployedinBeijingbypositivematrixfactorization(PMF)baseonsizesegregateddata,andbothsourcesandtheirsizedistributionwereobtained.Thedataisimportantforstudiesonbothhealtheffectandpolicy-making.AsFig.1shows,thesamplingperiodwasintheyearof2006withhighpollutionofPM10inBeijing,sotheresultofthisstudywillalsohelptoevaluatetheeffectofenvironmentalpoliciessincethen.

1Materialsandmethods

1.1Studyareaandsampling

A13-stagelowpressureimpactor(Dekati,Tampere,Fin-land)wasusedtoprovideresolutionofthesizedistribution

400350)3300TSP

m/gμ250∆(01M200∆P,PS150TPM10

100500

199419951996199719981999200020012002200320042005200620072008200920102011

Fig.1TrendsofTSPandPM10inBeijing,China,during1994–2011.Datawerecollectedfromhttp://www.bjepb.gov.cn.

ofaerosolpopulationsfromcoarseparticlestoultrafineparticles.PTFEfiltersubstrates(Waters,USA)wereusedonallstages,andataflowrateof10L/minthe50%cutoffdiameters(D50,μm)were9.92,6.68,4.00,2.39,1.60,0.948,0.613,0.382,0.263,0.157,0.095,0.056and0.028.SamplingsiteofTsinghuaisanurbansiteontherooftopofabuilding(5mabovetheground)withinTsinghuaUni-versity,whichislocated20kminthenorthwestdirectionfromthecenterofBeijingand2kmfrom4thRingRoad.Someboilersforheatingandcookingmayexistinthevicinity.Moredetailinformationcanbefoundinpreviouspublication(Duanetal.,2008).

FoursetsofsamplesBJ1,BJ2,BJ3andBJ4withdurationrangesfrom48–72hrwerecollectedduring4th–7th,10th–13th,17th–20thand25th–27th,Dec.2006,respectively.Totally,48sampleswerecollected.AsFig.1shows,thePM10levelswereveryhighin2006,andthesamplescollectedinthisstudycanrepresentthesituationbeforeBeijingOlympicGamesin2008forwhichthegovernmenthadtakengreateffortstocontroltheairpollutionandthePM10levelshadgraduallydecreasedfrom161to114μg/m3sincethen.1.2Chemicalanalysis

Theaerosol-loadedfilterswerecutintohalfs,andsamplesofonehalfwereplacedinTeflontubeswitha4-mLmixtureofconcentratedhydrochloricacidandnitricacidatvolumeratioof3:1,andthenmicrowavedfor58mintoensurethecompletedigestionofparticlescollectedonTeflonfilters.Themicrowaveprocessconsistedoffourstepsundertheoperationpowerof1200W.Thefirststepistohavetemperaturerampingto120°Cfromroomtemperaturein5minandholdingfor5min;thesecondisrampingto155°Cin8minandholdingfor8min;thethirdisrampingto180°Cin5minandholdingfor5min;andthelaststepisrampingto195°Candholdingfor12min.Then,thedigestedsolutionwasdilutedto10mLusingultrapurewater(specificresistance󰀃8.3MΩ·cm)toperformthemetalanalysisbyinductivelycoupledplasma-massspectrometry(ICP-MS)(Thermo,Xserial)andICP-AES(Thermo,IRISIntrepidIIXSP).Thecalibrationwasmadeusingmulti-element(metal)standards(certifiedreferencematerials;NationalAnalysisCenterforIronandSteel,China)ina3%(V/V)HNO3solution.Sixblankfiltersweretreatedandanalyzedinthesamewayasfortheactualsamples.Inaddition,theapproximatedetectionlimits(threetimesthestandarddeviationofblanksanalyzed)wereSi(216μg/L),Al(33μμgg//L),L),NaCa(55(221μgμ/L),g/L),MgSc(40(0.15μg/L),μg/SL),(92Tiμg(5.14/L),Kμg(102/L),V(0.083μg/L),Cr(0.93μg/L),Mn(1.14μg/L),Fe(40μg/L),Co(0.71μg/L),Ni(0.41μg/L),Cu(0.74μg/L),Zn(2.85μg/L),Ga(0.16μg/L),As(0.17μg/L),Se(0.22μg/L),Sr(0.77μg/L),Zr(0.57μg/L),Mo(0.09μg/L),Ag(0.11μg/L),Cd(0.027μg/L),Sb(0.14μg/L),Ba(1.43

JournalofEnvironmentalSciences2014,26(1)189–196

191

μg/L),Tl(0.10μg/L)andPb(1.11μg/L).

Ultrasonicmethodwasusedtoextractinorganicionsfromanotherhalfofsamplesandnormallyover98%ofsulfate,nitrate,andammoniumcanbeextracted.Thefilterwassubmergedinavialwith10mLultrapurewater,sealedandsubjectedtoultrasoundfor20minforextrac-tion.Theextractwasthenanalyzedbyionchromatography(Dionex1400)todeterminetheconcentrationsofinorgan-icions.Theconcentrationsofthewater-solubleinorganicionsinthefieldblankswere0.22,0.11,notdetected(nd),0.21,nd,0.32,nd,0.35and0.21μg/m3forF−,Cl−,NO3−,SO42−,NH4+,Na+,K+,Ca2+andMg2+,respectively.Therecoveryofeachionwasintherangeof80%–120%.Therelativestandarddeviationofeachionwaslessthan6%forthereproducibilitytest.Blankvaluesweresubtractedfromsampledeterminations.Thedetailsweregivenelsewhere(Tanetal.,2009b).1.3PMFanalysis

ThePMFmodelwasdevelopedbyPaateroandTapper(1994)andPaatero(2004).EPAPMFisoneofthereceptormodelsdevelopedbytheUSEnvironmentalProtectionAgency(EPA)’sOfficeofResearchandDevelopment.ThealgorithmsusedinEPAPMFmodeltocomputeprofilesandcontributionshavebeenpeerreviewedbyleadingscientistsintheairqualitymanagementcommunityandhavebeencertifiedtobescientificallyrobust.Inthisstudy,massconcentration,17elementsand8ionswereincludedinthemodelEPAPMF3.0.Theinputdataincludingcon-centrationsofsizesegregatedspeciesandequation-baseduncertainties.Theequation-baseduncertaintyincludingdetectionlimits(listedinSection1.2)anderrorfractions(5%).Iftheconcentrationislessthanorequaltothemethoddetectionlimit(MDL)provided,theuncertainty(Unc)iscalculatedusingthefollowingEq.(1)(Polissaretal.,1998):

Unc=

5

×MDL6

(1)

IftheconcentrationisgreaterthantheMDLprovided,thecalculationis

󰀉

Unc=

(Errorfraction×Concentration)2+MDL2

(2)

2Resultsanddiscussion

2.1LevelsofatmosphericPMandheavymetalsSomeepidemiologicalevidencessuggestmortalityinur-banareasmaybelinkedtoPM10(Lippmann,1998).AsshowninTable1,theaveragedPM10andPM2.5were192.2and119.0μg/m3respectivelyduringthestudyperiodinwinterinBeijing,muchhigherthantheNAAQSofChinaforPM10(100and70μg/m3annuallyforGB3095-1996andGB3095-2012,respectively)andPM2.5(35μg/m3annuallyGB3095-2012).ThePM10andPM2.5alsoexceedWHOguidelinesmorethan8.61and10.90times.ItindicatesthePMpollutionsinwinterinBeijingismuchserious.

Pbisaheavymetalnervetoxic.AccordingtoTable1,theaverageconcentrationofatmosphericPbis281.6ng/m3inBeijingcomparabletoreportedaverageconcen-trationof261.0ng/m3inChina,whichfallbelowtheannuallimitsofcurrentNAAQSofChina(GB3095-1996)of1000ng/m3,appendixofNAAQS(GB3095-2012,500ng/m3)andtheWHOguidelineof500ng/m3.ThiscanbemainlyattributedtothenationwideprohibitionofleadedgasolineinChinasinceJuly1,2000.Previousstudy(Li

Table1ComparisonofPM10,PM2.5andatmosphericheavymetalswithNAAQSandWHOguidelinesNAAQSGB3095-1996

NAAQSGB3095-20127035505006

WHO(2000)2010100050010006.6150250.255

Thisstudy192.2119.0281.610.747.1163.722.741.65.8501.972.2

RatiotoWHO9.6111.900.560.017.141.090.911.16

261.017.951.0198.829.085.713.2424.5117.0

AveragedlevelsinChina∗PM10(μg/m3)PM2.5(μg/m3)Hg(ng/m3)Pb(ng/m3)V(ng/m3)As(ng/m3)Mn(ng/m3)Ni(ng/m3)Cr(VI)(ng/m3)Cd(ng/m3)Zn(ng/m3)Cu(ng/m3)

100

1000

0.0255

*DuanandTan,2013192

JournalofEnvironmentalSciences2014,26(1)189–196

etal.,2000)hadshownthat,sincethephaseoutoftheleadedgasoline,theaverageannualconcentrationofPbdecreasedby74%comparedwithbefore.However,exceptforvehicleemissions,coalcombustionisalsoamajorsourceforanthropogenicPb(Duanetal.,2012).

Vcanbeenemittedintotheatmospherebythreeways(Al-Momani,2003;Linetal.,2005):naturalrockweath-ering;combustionoffossilfuelssuchascoalandoil;miningandsmeltingofvanadiumcontainingmineralssuchasvanadium-titaniummagnetite.TheconcentrationofatmosphericVis10.7ng/m3inBeijingwhichismuchlowerthanthereportedaverageconcentrationof17.9ng/m3inChina,andmuchlowerthanthelimitofWHO1000ng/m3aswell.ThereisnolimitofVinbothcurrentandnewNAAQS(GB3095-1996,GB3095-2012).

Asisoneofthemostdangerouscarcinogenicelementstohuman.AlthoughAsisnotaheavymetal,consideringitshealtheffect,itisdiscussedasaheavymetal(oid)inthecontext.Studieshaveshownthatcoalcombustionisoneoftheimportantanthropogenicsource(NriaguandPacyna,1988).TheconcentrationofatmosphericAsis47.1ng/m3inthisstudywhichissimilartothereportedaverageconcentrationof51.0ng/m3inChina,andmuchhigherthanthelimitofthenewNAAQS(GB3095-2012)inChina(6ng/m3)andthelimitofWHO(6.6ng/m3).ExcessMnwilldoharmtothenervous,immuneandreproductivesystemsofhumanbeings.AtmosphericMncomesmainlynotonlyfromanthropogenicsourcessuchascoalcombustion,metalsmeltingandgasolineantiknockadditive,butalsofromnaturalsourcesofsoilerosion.AtmosphericMnwasenrichedinbothcoarseandfineparticles(Duanetal.,2012).Theconcentrationofatmo-sphericMnis163.7ng/m3inthisstudy,whichisclosetoWHO’slimitof150ng/m3.ThereisnolimitofMninbothcurrentandnewNAAQS(GB3095-1996,GB3095-2012).NiisdefinedasthefirstclassofcarcinogensbyCancerResearchCenter.PetroleumandcoalcombustionarethemajorsourcesofNi,andsomeindustryprocessessuchassmeltingofnickeloreandnickel-containingmetalores(especiallyironorsteel)canalsoemitNi.Theconcentra-tionofatmosphericNiis22.7ng/m3inthisstudy,neartoWHO’slimitof25ng/m3.ThereisnolimitofNiinbothcurrentandnewNAAQS.

MostatmosphericCrexistsintwoformsofinorganicCr(III)andCr(VI).Cr(VI)candoharmtothekidneysandheart,andhaveacarcinogeniceffect.BothnewNAAQS(GB3095-2012)ofChinaandlimitofWHOonlylistouttheCr(VI)concentrationlimit,however,generallythedatareportedintheliteratureisoftotalCrandfewdataonCr(VI).ThetotalconcentrationofatmosphericCris41.6ng/m3inthisstudy.ThestudiesontheformsofatmosphericCrarelimited,andatmosphericCr(VI)needstobefurtherstudied.

Afterbreathingintothebody,mostCdwillgetintotheliverandkidneys.Theconcentrationofatmospheric

Cdis5.8ng/m3inthisstudy,similartothenewNAAQS(GB3095-2012)inChinaandtheWHOlimitof5ng/m3.CuandZncanbecometoxicwhenpresentinginexcessiveamount.TheconcentrationofatmosphericCuandZnare72.2ng/m3and501.9ng/m3respectivelyinthisstudy.AsTable1shows,comparedwithconcentrations,theatmosphericheavymetalsinBeijinginwinterrankas:Zn>Pb>Mn>Cu>As>Cr>Ni>V>Cd,andalltheseheavymetalsareneartheaveragelevelsinChinawhichreviewedbyDuanandTan(2013).TheconcentrationratiosofatmosphericheavymetalstoWHOguidelinescanreachashighas7.14(0.01–7.14),however,comparedwiththeratiosofPM10andPM2.5(9.61and11.90),theyarestillmuchlower.ItindicatedthatitismoreimportanttocontrolatmosphericheavymetalsbycontrolPMfirst.2.2SizedistributionofatmosphericheavymetalsGenerally,typicalaerosolspectrahavethreepeaks(modes).Coarsemodeparticles(>2μm)tendtobeme-chanicallygeneratedsuchassoildust,seasprayaerosol,androaddustinurbanareas.Nucleation(<0.1μm)andaccumulationmode(0.1–2μm)particlestendtobeproducedeitherdirectlyfromcombustionsources,orbygastoparticleconversioninvolvingreactionproductsofsulphates,nitrates,ammoniumandorganics.Figure2showsthesizedistributionofthestudiedatmosphericheavymetalsinBeijing.

Allthestudiedheavymetalsexistinallofthenucleationmode,accumulationmodeandcoarsemodeparticles.Amongthem,Pb,Cd,AsandZnshowmostlyinaccumu-lationmode;V,MnandCuexistmostlyinbothcoarseandaccumulationmodes;andNiandCrexistinallofthethreemode.Consideringthehealtheffectandvisibilitydegra-dation,anewNAAQS(GB3095-2012)waspromulgatedinChinain2012,andPM2.5limitswereaddedintoitforthefirsttime.AsFig.2shows,byestimation,theratiosofatmosphericheavymetalsinPM2.5(PM2.39inthisstudy)tothoseinPM10(PM9.92inthisstudy)rankedasPb(88.5%)>>CdCr(71.7%)(81.8%)>>NiZn(67.9%)(81.5%)>>MnAs(63.3%)(77.1%)>>VCu(46.8%).(75.9%)ItindicatesthatitisaneffectivewaytocontrolatmosphericheavymetalsbyPM2.5controlling.

Consideringhealtheffect,sincesmalleraerosolcaneasilygetintothehumanrespiratorysystem,theheavymetalsassociatedwithsmalleraerosolwilldomoreharmtohumanhealth.USEPA(1982)presentedthedeposi-tionratesofparticlesintherespiratorytractformouthbreathingasafunctionofparticlesize.Accordingtoit,thebreakthroughofatmosphericheavymetalsintopulmonaryalveoliareshowninFig.2.Bycalculation,thebreak-throughratesofthestudiedheavymetalsintopulmonaryalveoliare:Pb(62.1%)>As(58.1%)>Cd(57.9%)>Zn(57.7%)>Cu(55.8%)>Ni(53.5%)>Cr(52.2%)>Mn(49.2%)>V(43.5%).

JournalofEnvironmentalSciences2014,26(1)189–196

1.2dC/(Ctotal*dlogDp)1.0 0.80.60.40.2345672345672345670.010

34562345623456 Pb Breakthrough PbdC/(Ctotal*dlogDp) As Breakthrough As193

1.21.00.80.60.40.20.0345670.1

2345671

Dp (μm)

23456710

1.2dC/(Ctotal*dlogDp)1.00.80.60.40.20.0 ZnBreakthrough Zn0.1

1

Dp (μm)

0.1

1

Dp (μm)

10

1.4dC/(Ctotal*dlogDp)1.21.00.80.60.40.20.0 Cd Breakthrough Cd dC/(Ctotal*dlogDp)1.00.80.60.40.2 V Breakthrough VdC/(Ctotal*dlogDp)1.41.21.00.80.60.40.20.0 Cr Breakthrough Cr345670.1

2345671

Dp (μm)

2345670.010

34560.1

234561

Dp (μm)

2345610

1.00.8dC/(Ctotal*dlogDp)0.60.40.20.0345670.1

2345671

Dp (μm)

23456710

1.0dC/(Ctotal*dlogDp)0.80.60.40.20.0

dC/(Ctotal*dlogDp) Cu Breakthrough Cu1.00.80.60.40.20.0 Mn Breakthrough Mn Ni Breakthrough Ni345670.1

2345671

Dp (μm)

23456710

34560.1

234561

Dp (μm)

2345610

345670.1

2345671

Dp (μm)

23456710

Fig.2SizedistributionofatmosphericheavymetalsandtheirbreakthroughintopulmonaryalveoliinBeijing.Theerrorbarsrepresentonestandarddeviationforthefoursetsofsamples.

2.3SourceapportionmentofatmosphericheavymetalsFoursetsof48sizesegregatedsampleswereinputtoEPAPMF3.0.Inordertosimplifythemodelcalculation,exceptforthenineheavymetals,onlyelementsandionsasgoodsourcemarkersandwithlowuncertaintieswereincludedinthemodeling:Al,Ca,Fe,K,Mg,V,Cr,Mn,Ni,Cu,Zn,As,

−2−

Se,Cd,Sb,Ba,Pb,Na+,K+,Mg2+,Ca2+,NH+4,Cl,SO4,NO−3andmassconcentration.Amongthem,AlandCaareindicatorsofcrustrelateddust(Duanetal.,2012);Kisamarkerofbiomassburning(Duanetal.,2004);AsandSegenerallymostcomefromcoalburning(Duanetal.,2012);

2−−

NH+4,SO4andNO3areindicatorofsecondaryaerosolandlongrangetransportation.Tanetal.(2009)reportedtheseionsaresignificantlyhigherinhazedaysthannormaldays.

FourfactorsareidentifiedbyEPAPMF3.0,andfactorprofiles(%ofspeciestotal)areillustratedasFig.3.Factor1(dust)hashighcontributionofAl(91.2%),Ca(95.8%),Fe(81.4%),Mg(89.3%)andBa(84.1%)indicatingcrustrelateddust(Heetal.,2001).Factor1isthemainsourceof

V(58.2%)andMn(35.1%),andcontributeslessthan20%tootherheavymetals.

Factor2(vehicle)hashighcontributionofMn(39.5%),Cu(35.6%),Zn(50.5%),As(40.7%),Cd(55.7%)andPb(26.9%)indicatingvehicleemission,andFactor2contributeslessthan20%tootherheavymetals.ItisconcludedthatvehiclesemissionsmaybethemajorsourceofPb,Cu,ZnandCdurbancontamination(Johanssonetal.,2009).

Factor3(agedtransportation)contributeshightoK(59.7%),As(48.1%),Se(86.0%),Pb(62.8%)andmost

2−

ofthesecondaryionsofNH+4(93.0%),SO4(69.0%)andNO−3(73.2%).Itindicatesitisamixtureofagedlongrangetransportationsource.Themixturemayincludebiomassburning,coalburningandotherindustriessuchasironandsteelindustry.BeijingwassurroundedbyTianjinCityandHebeiProvince.Tianjinisoneofthebiggestindustrialcitywithapopulationof9.3million.HebeiProvincehasthehighestironandsteelproductivityinChina,andthecoalconsumptionwasmuchhigh.BiomassburningwasoftenreportedinHebeiProvince.Ruralbiomassburning

194

100%JournalofEnvironmentalSciences2014,26(1)189–196

Dust50%

0100%

Vehicle50%

Relative contribution0100%

Aged and transportation50%

0100%Unknown50%

0

AlCaFeKMgVCrMnNiCuZnAsSeCdSbBaPbNa+K+Mg2+Ca2+NH4+Cl-SO42-NO3-PMFig.3Factorprofiles(%ofspeciestotal)obtainedfromEPAPMF3.0model.

hasalsobeenidentifiedasanimportantcontributortofinePMconcentrationsinBeijing(Duanetal.,2004).Factor3alsocontributestootherheavymetalsofV(26.6%),Cr(5.3%),Cd(37.8%)andMn(25.4%).

Factor4(unknown)contributeshightoCr(65.7%),Ni(56.3%),Na(50.1%)andCl(33.55%).ThesourceofFactor4isstillneedtobefurtherstudied.Factor4contributesmuchlesstootherheavymetals(<10%).Astomassconcentrationcontributions,itcanberankedasFactor3(agedandtransportation:46.7%)>Factor1(dust:40.7%)>Factor2(vehicle:8.8%)>Factor4(unknown:3.8%).Thisresultissimilartothosefoundbyotherresearchersthatabout34%ofPM2.5onaveragecanbeattributedtosourcesoutsideBeijing,andduringsustainedwindflowfromthesouth,HebeiProvincecancontribute50%–70%ofPM2.5concentrationsinBeijing(Streetsetal.,2007).

Althoughthetotalcontributionofatmosphericheavymetalsasdiscussedafore,thecontributionstronglyrelyonthesizedistributionasFig.4shows.Generally,Factor1contributesmoretocoarsemodeandlittleinnucle-ationandaccumulationmodeconcentrations;Factor2contributesmoretoaccumulationmodeandalittlecoarsemodeconcentrations;Factor3contributesmosttoaccu-mulationmodeconcentrations;andFactor4contributestoallofthethreemode.

3Conclusions

AstudyonsizedistributionsofatmosphericheavymetalswascarriedoutduringhazeinwinterinBeijing,Chi-na.Thelevels,sizedistributionsandsourcesofheavymetals(Zn,Pb,Mn,Cu,As,Cr,Ni,VandCd)werediscussed.Amongthem,atmosphericMn,Ni,As,CdandPbexceedthereferencevaluesofNAAQS(GB3095-2012)andguidelinesofWHO.TheconcentrationratiosofatmosphericheavymetalstoWHOguidelinescanreachashighas7.14,however,comparedwiththeratiosofPM10(9.61)andPM2.5(11.90),theyarestillmuchlow.ItindicatedthatitismoreimportanttocontrolatmosphericheavymetalsbycontrolPMfirst.Pb,Cd,andZnshowmostlyinaccumulationmode,V,MnandCuexistmostlyinbothcoarseandaccumulationmodes,andNiandCrexistinallofthethreemode.Byestimation,highpercent-ageofatmosphericheavymetalsinPM2.5indicatesitisaneffectivewaytocontrolatmosphericheavymetalsbyPM2.5controlling.Consideringthehealtheffect,thebreak-throughratesofatmosphericheavymetalsintopulmonaryalveoliare:Pb(62.1%)>As(58.1%)>Cd(57.9%)>Zn(57.7%)>Cu(55.8%)>Ni(53.5%)>Cr(52.2%)>Mn(49.2%)>V(43.5%).PMFmethodwasappliedfor

JournalofEnvironmentalSciences2014,26(1)189–196

195

Unknown

As Dust

100Mass dM/dlogDp (ng/m3)806040200ZnMass dM/dlogDp (ng/m3)5040302010 VehiclePbAged and transportation

Mass dM/dlogDp (ng/m3)34567890.1

234567891

Dp (μm)

2345678910

86420

34567890.1

234567891

Dp (μm)

2345678910

63456789234567892345670.11

Dp (μm)Cr10

1.2Mass dM/dlogDp (ng/m3)1.00.80.60.40.2CdMass dM/dlogDp (ng/m3)54321VMass dM/dlogDp (ng/m3)2345678910

43210

0.034567890.112Mass dM/dlogDp (ng/m3)10864200.034567890.1

Cu234567891

Dp (μm)2345678910

0.03456789234567890.11

Dp (μm)25Mass dM/dlogDp (ng/m3)20151050Mn3456789234567892345670.11

Dp (μm)Ni10

2.5Mass dM/dlogDp (ng/m3)2.01.51.00.50.0234567891

Dp (μm)

2345678910

34567892345678923456789100.11

Dp (μm)

34567890.1

234567891

Dp (μm)

23456710

Fig.4Sizedistributionofatmosphericheavymetalsreconstructedbysourcecontributions.

sourceapportionmentofstudiedheavymetalscombined

−

withsomemarkerelementsandionssuchasK,As,SO24etc.Fourfactorsareidentified:Factor1(dust)isthemainsourceofV(58.2%)andMn(35.1%),andcontributelessthan20%tootherheavymetals.Factor2(vehicle)hashighcontributionofMn(39.5%),Cu(35.6%),Zn(50.5%),As(40.7%),Cd(55.7%)andPb(26.9%),howevercontributeslessthan20%tootherheavymetals.Factor3(agedandtransportation)contributeshightoAs(48.1%),Pb(62.8%),andalsocontributetoV(26.6%),Cr(5.3%),Cd(37.8%)andMn(25.4%).Factor4(unknown)contributeshighonlytoCr(65.7%),Ni(56.3%),andcontributesmuchlesstootherheavymetals(<10%).

Acknowledgments

ThisworkwassupportedbytheNationalNaturalSci-enceFoundationofChina(No.41105111,41275134),theNationalDepartmentPublicBenefitResearchFoundation(MEP)(No.201109005),andtheResearchFoundofCRAES(No.2012ysky09).references

Al-MomaniIF,2003.Traceelementsinatmosphericprecipitation

atNorthernJordanmeasuredbyICP-MS:Acidityandpossiblesources.AtmosphericEnvironment,37(32):4507–4515.

196

JournalofEnvironmentalSciences2014,26(1)189–196

ChenBH,KanHD,ChenRJ,JiangSH,HongCJ,2011.Airpollution

andhealthstudiesinChina-Policyimplications.JournaloftheAirandWasteManagementAssociation,61(11):1292–1299.

ChengSP,2003.HeavymetalpollutioninChina:Origin,patternand

control.EnvironmentalScienceandPollutionResearch,10(3):192–198.

CohenAJ,AndersonHR,OstroB,PandeyKD,KrzyzanowskiM,

K¨unzliNetal.,2010.Urbanairpollution.In:ComparativeQuan-tificationofHealthRisks:GlobalandRegionalBurdenofDiseaseAttributiontoSelectedMajorRiskFactors(EzzatiM,LopezAD,RodgersAetal.,eds.).Geneva:WorldHealthOrganization;2004.(accessed3September2010).1353–1434.

DuanFK,LiuXD,YuT,CachierH,2004.Identificationandestimateof

biomassburningcontributiontotheurbanaerosolorganiccarbonconcentrationsinBeijing.AtmosphericEnvironment,38(9):1275–1282.

DuanJC,BiXH,TanJH,ShengGY,FuJM,2005.Thedifferencesof

thesizedistributionofpolycyclicaromatichydrocarbons(PAHs)betweenurbanandruralsitesofGuangzhou,China.AtmosphericResearch,78(3-4):190–203.

DuanJC,BiXH,TanJH,ShengGY,FuJM,2007.Seasonalvariation

onsizedistributionandconcentrationofPAHsinGuangzhouCity,China.Chemosphere,67(3):614–622.

DuanJC,TanJH,WangSL,HaoJM,ChailFH,2012.Sizedistributions

andsourcesofelementsinparticulatematteratcurbside,urbanandruralsitesinBeijing.JournalofEnvironmentalSciences,24(1):87–94.

DuanJC,TanJH,YangL,WuS,HaoJM,2008.Concentration,sources

andozoneformationpotentialofvolatileorganiccompounds(VOCs)duringozoneepisodeinBeijing.AtmosphericResearch,88(1):25–35.

DuanJC,TanJH,2013.AtmosphericheavymetalsandarsenicinChina:

situation,sourcesandcontrolpolicies.AtmosphericEnvironment,74:93–101.

HeKB,YangFM,MaYL,ZhangQ,YaoXH,ChanCKetal.,

2001.ThecharacteristicsofPM2.5inBeijing,China.AtmosphericEnvironment,35(29):4959–4970.

JohanssonC,NormanM,BurmanL,2009.Roadtrafficemissionfactors

forheavymetals.AtmosphericEnvironment,43(31):4681–4688.KellyJ,ThorntonI,SimpsonPR,1996.Urbangeochemistry:Astudyof

theinfluenceofanthropogenicactivityontheheavymetalcontentofsoilsintraditionallyindustrialandnonindustrialareasofBritain.AppliedGeochemistry,11(1-2):363–370.

LiWJ,LiuCX,WangL,2000.Non-leadgasolineandtrendoflead

pollutioninair.UrbanEnvironment&UrbanEcology,20:61–62.LinCC,ChenSJ,HuangKL,HwangWI,Chang-ChienGP,Lin

WY,2005.Characteristicsofmetalsinnano/ultrafine/fine/coarseparticlescollectedbesideaheavilytraffickedroad.EnvironmentalScienceandTechnology,39(21):8113–8122.

LippmannM,1998.The1997USEPAstandardsforparticulatematter

andozone.In:IssuesinEnvironmentalScienceandTechnology,10(HesterRE,HarrisonRM,eds.).RoyalSocietyofChemistry,Cambridge.75–99.

NingDT,ZhongLX,ChungYS,1996.Aerosolsizedistribution

andelementalcompositioninurbanareasofnorthernChina.AtmosphericEnvironment,30(13):2355–2362.NriaguJO,1988.Asilentepidemicofenvironmentalmetalpoisoning.

EnvironmentalPollution,50(1-2):139–161.

NriaguJO,PacynaJM,1988.Quantitativeassessmentofworldwide

contaminationofair,waterandsoilsbytrace-metals.Nature,333(6169):134–139.

PaateroP,2004.User’sguideforpositivematrixfactorizationprograms

PMF2andPMF3,Part1:tutorial.

PaateroP,TapperU,1994.Positivematrixfactorization:anon-negative

factormodelwithoptimalutilizationoferrorestimatesofdatavalues.Environmetrics,5(2):111–126.

PakkanenTA,KerminenVM,KorhonenCH,HillamoRE,Aarnio

P,KoskentaloTetal.,2001.UseofatmosphericelementalsizedistributionsinestimatingaerosolsourcesintheHelsinkiarea.AtmosphericEnvironment,35(32):5537–5551.

PolissarAV,HopkePK,PaateroP,MalmWC,SislerJF,1998.

AtmosphericaerosoloverAlaska-2.Elementalcompositionandsources.JournalofGeophysicalResearch-Atmospheres,103(D15):19045–19057.

SalmaI,OcskayR,RaesN,MaenhautW,2005.Finestructureof

masssizedistributionsinanurbanenvironment.AtmosphericEnvironment,39(29):5363–5374.

StreetsDG,FuJS,JangCJ,HaoJM,HeKB,TangXYetal.,2007.

Airqualityduringthe2008BeijingOlympicGames.AtmosphericEnvironment,41(3):480–492.

TanJH,DuanJC,ChenDH,WangXH,GuoSJ,BiXHetal.,

2009a.ChemicalcharacteristicsofhazeduringsummerandwinterinGuangzhou.AtmosphericResearch,94(2):238–245.

TanJH,DuanJC,HeKB,MaYL,DuanFK,ChenYet

al.,2009b.ChemicalcharacteristicsofPM2.5duringatypicalhazeepisodeinGuangzhou.JournalofEnvironmentalSciences,21(6):774–781.

TangXL,BiXH,ShengGY,TanJH,FuJM,2006.Seasonalvariation

oftheparticlesizedistributionofn-alkanesandpolycyclicaro-matichydrocarbons(PAHs)inurbanaerosolofGuangzhou,China.EnvironmentalMonitoringandAssessment,117(1-3):193–213.TaoJ,CaoJJ,ZhangRJ,ZhuLH,ZhangT,ShiSetal.,2012a.

Reconstructedlightextinctioncoefficientsusingchemicalcompo-sitionsofPM2.5inwinterinUrbanGuangzhou,China.AdvancesinAtmosphericSciences,29(2):359–368.

TaoJ,ShenZX,ZhuCS,YueJH,CaoJJ,LiuSXetal.,2012b.

Seasonalvariationsandchemicalcharacteristicsofsubmicrometerparticle(PM1)atGuangzhou,China.AtmosphericResearch,118:222–231.

TrijonisJ,1983.Developmentandapplicationofmethodsforestimating

inhalableandfineparticleconcentrationsfromroutineHi-Voldata.AtmosphericEnvironment(1967),17(5):999–1008.

USEPA(EnvironmentalProtectionAgency),1982.Airqualitycriterion

forparticulatematterandsulfuroxides.TechnicalReportEPA-600/8-82-029.

WHO,2000.WorldHealthOrganization.GuidelinesforAirQuality

Geneva.

ZhangJ,OuyangZY,MiaoH,WangXK,2011.Ambientairquality

trendsanddrivingfactoranalysisinBeijing,1983–2007.JournalofEnvironmentalSciences,23(12):2019–2028.

ZhouSX,2011.[EB/OL].(2010-1-25).http://www.mep.gov.cn/

zhxx/hjyw/201001/t20100129185134.htm.