assessment of personal exposure to airborne re.pdf · pdf filetube material. in another...

Assessment ofPersonal Exposureto AirborneNanomaterialsA Guidance Document

Institute of Energy and Environmental Technology e.V. ( IUTA) , Germany

Federal Institute of Occupational Safety and Health (BAuA) , Germany

Institute of Occupational Medicine ( IOM) , Great Britain

University of Applied Sciences and Arts Northwestern Switzerland (FHNW), Switzerland

Catholic University of the Sacred Heart (UCSC) , Italy

Institute for the Research on Hazardous Substances( IGF) , Germany

Atomic Energy Commission (CEA) , France

University of Duisburg-Essen (UDE) , Germany

Assessment ofPersonal Exposureto AirborneNanomaterialsA Guidance Document

1. Introduction 4

2. Measurementandsamplingtechniquesincluding accuracy,comparabilityandfieldapplicability 2.1. Metricsissues 8 2.2. Personalmonitors 8 2.2.1. Partector 10 2.2.2. DiSCmini 10 2.2.3. nanoTracer 11 2.2.4. PUFPC100/C200 11 2.2.5. BlackcarbonmonitorMicroAethAE51 12 2.2.6. Accuracyandcomparabilityofthepersonalmonitors 12 2.3. Personalsamplers 13 2.3.1. Filterbasedsamplers 14 2.3.1.1. PersonalsamplingsystemPGP 14 2.3.1.2. NANOBADGE 14 2.3.1.3. NanoparticleRespiratoryDeposition(NRD)sampler 15 2.3.1.4. Prototypespersonalsamplers 16 2.3.2. Samplersforelectronmicroscopicanalysis 16 2.3.2.1. ESPnano 16 2.3.2.2. ThermalPrecipitatorSampler(TPS) 17 2.3.3. Specificcaseofcarbon-basedaerosols 17 2.3.4. Accuracyandcomparabilityofthepersonalsamplers 18 2.3.4.1. PGP-FAP 18 2.3.4.2. NANOBADGE 18 2.4. Periphery 20 2.4.1. Samplingtubes 20 2.4.2. Pre-separators 21 2.4.3. Personalpumps 21 2.5. Fieldapplicabilityofpersonalsamplersandmonitors 21 2.5.1. Practicalconsiderations 22 2.5.2. Effectsofwearingtheinstruments 22 2.5.3. Istheavailableinstrumentationapplicable 23 tofieldstudies?

3. Performanceofmeasurements 3.1. Descriptionandselectionoftheassessmenttask 25 3.1.1. Epidemiologicalstudies 25 3.1.2. Exposureassessmentwithinriskassessment/ 25 riskmanagementprocedures 3.2. Selectionofthemeasurementdevices 26 3.2.1. Epidemiologicalstudies 26 3.2.2. Exposureassessmentwithinriskassessment/ 26 riskmanagementprocedures

Content

3.3. Selectionoftheworkplacesoremissionsources 27 tobeinvestigated 3.3.1. Epidemiologicalstudies 27 3.3.2. Exposureassessmentwithinriskassessment/ 27 riskmanagementprocedures 3.4. Backgroundmanagement 28 3.4.1. Epidemiologicalstudiesandriskassessment/ 28 riskmanagementprocedures 23 3.5. Performanceofthemeasurements 28 3.5.1. Epidemiologicalstudies 28 3.5.2. Exposureassessmentwithinriskassessment/ 28 riskmanagementprocedures 3.6. Dataevaluation 29 3.7. Documentation 29 3.7.1. Epidemiologicalstudies 29 3.7.2. Exposureassessmentwithinriskassessment/ 29 riskmanagementprocedures 3.8. Qualityassurance 29

4. Datacollection,analysisandstorage 4.1. Datacollection 31 4.2. Dataanalysis 31 4.2.1. Preliminaryanalysis 31 4.2.2. Furtheranalysis 33 4.2.2.1. Bias 33 4.2.2.2. Precision 33 4.2.2.3. Distance 33 4.2.2.4. Auto-RegressiveIntegratedMovingAverage 34 (ARIMA)models

5. Exemplarydatafromfieldmeasurements 5.1. Experimental 36 5.2. Fieldstudyduringpreparationofpastes 36 5.3. FieldstudyduringproductionofTiO2nanoparticles 37 5.4. Fieldstudyinalaboratoryforthesynthesisofnanowires 38 5.5. Conclusionsfromfieldstudies 39

6. Lessonslearnedduringtheproject 6.1. Lesson1:Instrumentalissues 41 6.2. Lesson2:Issuesrelatedtoplanningandperformance 41 offieldmeasurements 6.3. Lesson3:Issuesrelatedtopersonalmonitoring 42 6.4. Lesson4:Issuesrelatedtopersonalsampling 42 6.5. Lesson5:Datacollectionandhandling 42 6.6. Lesson6:Samplingormonitoring? 43 Whatmetricshouldbedetermined?

7. Conclusions 45

Content

nanoIndEx – Assessment of Personal Exposure to Airborne Nanomaterials

4

Introduction

Theuseofmanufacturednanomaterials(MNMs)1hasincreasedataconstantpaceovertherecentyears.Theirapplicationsrangefromscratchresistantorself-cleaningsurfacecoatings,viaen-forcedpolymerstoenhancedcosmetics.Besidesthetremendousnewopportunitiesofferedbythesenovelmaterials,concernshavebeenraisedbecauseofpotentialadversehealtheffectsthatmayariseifMNMsaretakenupbythehumanbody[1].WhilehumanexposuretoMNMsmayinprincipleoccurduringanystageofthematerial’slifecycle,itismostlikelyinworkplaces,wherethesematerialsareproducedorhandledinlargequantitiesoroverlongperiodsoftime.Inhalationisconsideredasthemostcriticaluptakeroute,becausethesmallparticlesareabletopenetratedeepintothelunganddepositinthegasexchangeregion.Inhalationexposuretoairbornenanoma-terialsthereforeneedstobeassessedinviewofworkerprotection.

Exposuretoairborneparticlescangenerallybestbeassessedbymeasuringtheindividualexposu-reinthepersonalbreathingzone(PBZ)ofanindividual.ThePBZisdefinedasa30cmhemispherearoundmouthandnose[2].MeasurementsinthePBZrequireinstrumentsthataresmallandlight-weight.TheindividualexposurespecificallytoMNMshasnotbeenassessableinthepastduetothelackofsuitablepersonalsamplersand/ormonitors.Instead,moststudiesrelatedtoexposuretoMNMshavebeencarriedoutusingeitherbulkystaticmeasurementequipmentornotnanospe-cificpersonalsamplers.Inrecentyears,novelsamplersandmonitorshavebeenintroducedthatallowforanassessmentofthemorenanospecificpersonalexposuretoairborneMNMs.Intheter-minologyusedinnanoIndEx,samplersaredevicesthatcollectparticlesonasubstrate,e.g.afilterofflatsurface,forsubsequentanalysis,whereasmonitorsarereal-timeinstrumentsthatdeliverinformationontheairborneconcentrationswithhightimeresolution.Scientificallysoundinvesti-gationsontheaccuracy,comparabilityandfieldapplicabilityofthesenovelsamplersandmonitorshadbeenlacking.Thislackofknowledgewasthenucleusforstartingtheproject“Assessmentof

1Intheliterature,manufacturednanomaterialsarealsotermedengineerednanomaterials(ENMs)ornanoobjectsandtheiragglomeratesandaggregates(NOAA).Althoughtheirexactdefinitionmaybeslightlydifferent,thesetermsareusedsynonymouslyinthisdocument.

5

IndividualExposuretomanufacturednano-materialsbymeansofpersonalmonitorsandsamplers”(nanoIndEx).

PartnersinvolvedinthenanoIndExprojectare:

•FederalInstituteofOccupationalSafety andHealth(BAuA,Berlin,Germany),•FrenchAlternativeEnergiesandAtomic EnergyCommission(CEA,Grenoble, France),•UniversityofAppliedSciencesandArts NorthwesternSwitzerland(FHNW, Windisch,Switzerland),•InstituteofOccupationalMedicine (IOM,Edinburgh,UK),•InstituteofEnergyandEnvironmental Technologye.V.(IUTA,Duisburg,Germany),•InstituteforHazardousSubstance Research(IGF,Bochum,Germany),•CatholicUniversityoftheSacredHeart (UCSC,Rome,Italy).

Thethree-yearprojectstartedonJune1st,2013,andhasbeenfundedundertheframeofSIINN,theERA-NETforaSafeImplementationofInnovativeNanoscienceandNanotechnolo-gy.Theaimoftheprojectwastoscrutinisetheinstrumentationavailableforpersonalexposu-reassessmentconcerningtheirfieldreadinessandusabilityinordertousethisinformationtogeneratereliabledataonpersonalexposureinrealworkplacesandtoeventuallywidelydistri-butethefindingsamongtheinterestedpublic.ThisGuidanceDocumentyouareholdinginyourhandssummarisesthekeyfindingsoftheproject.

Initially,theliteraturewasthoroughlystudiedtoidentifysuitablepersonalmonitorsandsamplers.Thoseinstrumentsthatwereiden-tifiedassuitableandthathavebeenavailablefortheprojectunderwentintensivelaboratoryinvestigationsconcerningtheiraccuracyandcomparability.Theinvestigationscoveredabroadrangeofaerosolandparticleproperties,includingthefullrangeofmorphologiesfromsphericaloveragglomeratedtofibrouspar-ticles.Suchstudiesareofutmostimportanceintermsofqualityassuranceandtoeventually

judgewhetherpotentialdifferencesinconcen-trationsmeasuredinthePBZandintheback-groundorfarfieldaresignificant.Anoverviewoftheavailablepersonalsamplersandmoni-torsandtheiraccuracy,comparabilityandfieldapplicabilityispresentedinchapter2.StandardOperationProcedures(SOPs)havebeenpreparedfortheoperationofallpersonalsamplersandmonitorsandarefreelyavailableontheproject’swebsitewww.nanoindex.eu.

Exposuremeasurementsinthefieldrequireaclearstrategy.Theexactstrategycanvaryde-pendingonthelocalsettingsintheworkplaceandmayneedtobetailoredtothequestionstobetackled.Thechoiceofinstrumentsisaffectedbythemeasurementstrategy.If,forexample,taskbasedexposurewithshort-livedspikesintheconcentrationsaretobeasses-sed,theuseofpersonalmonitorswithhightimeresolutionisinevitable.Tothecontrary,forthedeterminationofshift-basedaverages,samplersmayalsobeused.Ifpersonalexpo-suretoacertainchemicalspeciesshallbeas-sessed,thenwiththecurrentlyavailabletech-nology,thiscanonlybeachievedbyparticlesamplingandsubsequentchemicalanalysisofthedeposit.Placementoftheinstrumentsformonitoringofthebackgroundorfarfieldconcentrationsisalsoanimportantcomponentofthemeasurementstrategy.Chapter3ofthisGuidanceDocumentpresentssuggestionsonhowtoconductfieldmeasurementsofperso-nalexposure.

Aftercompletionofameasurementcampaign,thecollecteddatahavetobeanalysedandstored.Besidesthemeasurementdata,con-textualinformationonthesurveyedworkers,theiractivities,theworkplacesetc.havetobegathered.Especiallyincaseofmonitorswithhightimeresolutionofe.g.onesecond,onemayeasilyloseoverviewofthehugedataset.nanoIndExhasdevelopeddatacollectionandanalysisprotocols,basedontheNanoExpo-sureandContextualInformationDatabase(NECID),thatsimplifiesthedatamanagementandanalysis.Chapter4providesrecommenda-tionsconcerningdatacollection,analysisandstorage.

1. Introduction


6

FieldstudieshavebeenconductedwithinnanoIndEx,tobringtheknowledgeontheins-trumentation,themeasurementstrategyandthedatacollectionandanalysisintopractice.Theinvestigatedworkplacesvariedfromlabo-ratories,wherenanoparticlesarebeingpro-ducedorcharacterised,viaapilotplantfortheproductionofengineerednanomaterialsinanintermediatescaletolargescaleindustrialproduction.Theaimofthefieldstudieswasnotonlytocollectdataonpersonalexposure,butalsotolearnmoreaboutthefieldreadi-nessofthesamplersandmonitors.Thefieldstudiesaresummarisedinchapter5.

(Notonly)InthesenseofThorsteinVeblen’squote,nanoIndExwasaseriousresearchpro-ject,becauseweexperiencednumeroussurpri-sesandnoveltiesduringthecourseofthepro-ject.Inonecase,oneoftheinstrumenttypesreactedcompletelydifferentlythanexpected,becauseofaninterferencewiththesamplingtubematerial.Inanothercase,theplacementoftheinstrumentsusedforbackgroundmo-nitoringinfieldmeasurementsturnedouttobemorecriticalthananticipated.Chapter6sharestheexpectedandunexpectedlessonswehavelearnedduringtheprojectwithyoutomakeroomfornewquestionsratherthanma-kingyougrowtheexactsamequestionsagain.

ThisGuidanceDocumentisintendedtopre-sentyouthestateoftheartinpersonalexpo-sureassessmentfornanomaterials.Whilethefocusoftheprojectwasonexposuretoma-nufacturednanomaterialsinworkplaces,most

findingsarealsodirectlyapplicabletotheassessmentofexposuretonon-engineerednanoscaleparticles,e.g.intheenviron-ment.Wehopethatyouwillfindthisbro-chureinterestinganduseful.Forfurtherin-formation,pleasealsorefertoourwebpagewww.nanoindex.eu.

„The outcome

of any serious research

can only be to make

two questions grow

where only one question

grew before“

T H O R S T E I N V E B L E N(1857–1929 )

7

Chapter 2

Measurement and sampling techniques including accuracy, comparability and field applicability


8

2.1. Metrics issues

CurrentoccupationalexposurelimitsforMNMsaresetasmassconcentrationlimits.Fornanofibres,fibreconcentrationlimitsmaybeimposedinanalogytoasbestos.Atpre-sent,nooccupationalexposurelimitsbasedonlungdepositedsurfacearea(LDSA)orparticlenumberlevelsareunderdiscussion.However,inthisregarditisimportanttonotethatevenifnumberconcentrationisoftendominatedbynanoscaleparticles,suchasMNMs,theirmassisusuallynegligiblecompa-redtothatofcoarseparticles.Consequently,othermetricsthanmassshouldbetakenintoaccountinordertomakeanadequateandcomprehensiveevaluationofexposurestoMNMsinworkplaces.Unfortunately,itisnotyetclearwhichkeyparticulateparameters(mass,surfacearea,numberorsizedistributi-on)couldbethemostrelevantmeasurementunitwithregardtoMNM-relatedoccupationalhealthissues.

Ofthecurrentlyavailablecommercialperso-nalinstruments,fewaimatderivingmassconcentrations.TheseincludeX-rayfluores-cence-basedmassdeterminationoffiltersamples,which,however,canonlybeappliedtonanoparticlesofspecificelementalcompo-sitionsignaturebeforeaparticlebackgroundfreeofthissignatureelement.Likewise,themassofgraphiticcarbon-basedMNMsmaybequantifiedbeforetheubiquitouscarbon-con-tainingbackgroundbyEC/OCanalysisoffiltersamples.Knowledgeofthebackgroundprofileandcompositionismandatoryfortheapplicationofsuchtechniques.Theblackcar-bonexposuremaybeassessedbyradiationabsorptionoffiltereddustandaerosols.Theparticlebackgroundmustthusalwaysbestu-diedbeforeorafteraworktaskassessmentand,ifpossible,eveninparallelbymonitoringthesupplyair.

Personalmonitoringinstrumentsusingelectri-calnanoparticledetectionprinciplesgenerallyapplyunipolardiffusionchargingtodetermineLDSAconcentrationsandinsomecasesalso

thenumberconcentration.Thenumbercon-centrationcanalsobedeterminedbyconden-sationparticlecounters(CPCs).However,asofnowonlyasinglepersonalCPCexists.

2.2 . Personal monitorsWithinnanoIndEx,threetypesofpersonalmonitorswerethoroughlycharacterised:(1)theMiniatureDiffusionSizeClassifierDiSCmini(Testo,Titisee-Neustadt,Germany,identicalwithminiDiSC)[3],(2)theAera-sensenanoTracer(oxility,Eindhoven,theNetherlands)[4]and(3)thepartector(na-neos,Windisch,Switzerland)[5].Allthreeinstrumentsarebasedondiffusionchargingofthenanoparticles,followedbythedetec-tionofcurrentsonthefemto-Amplevel.AllinstrumentscanmeasuretheLDSAcon-centration,theDiSCminiandthenanoTracercanadditionallyalsomeasureparticlenumberconcentrationandaverageparticlediameter.Inallthreeinstruments,aerosolsenterthein-strumentandarechargedinaunipolardiffusi-oncharger,wheretheyacquireachargewhichisnearlyproportionaltotheparticlediameter(q~dx,withxapproximately1.1),and,bycoin-cidence,alsonearlyproportionaltotheLDSA.Theiontrapremovesexcessionsremainingafterthechargingprocess.TheproportionalitytoLDSAisnotexact,butitcanbethoughtofasagoodapproximationtoLDSAatleastinthesizerangeof20–350nm.Figure1showstherelationofchargetoLDSAovertherangeof5–10,000nm:

Thegraphshowsclearlythatthechargeacquiredisareasonableapproximation(±25%)fortheLDSAinthesizerangeof20–350nm.If±30%deviationistolerable,thesizerangecanbeextendedto400nm.Largerdeviationsoccuroutsideofthissizerange:formicronsizedparticles,theLDSAinferredfromchargingisabouttentimestoolow,forverysmallparticlesaround10nm,theLDSAinferredisabouttwiceashighasinreality.Itshouldbenotedthatthecon-tributionofsub-20nmparticlestothetotalLDSAconcentrationistypicallylow,whereasthedeviationforlargeparticlescanbequite

9

significant.FurtheruncertaintiestotheLDSAdeterminationapply,suchasdifferencesduetoparticlematerialandmorphology,anddif-ferentbreathingpattersofindividuals.AscanbeseenfromFigure1,thereisalsoanun-certaintyregardingthecalibrationoftheinstruments;instrumentscouldbecalibratedatsay50or100nm–nostandardoncali-bratingLDSAhasbeenestablishedasyet,andthereforeinstrumentsofdifferentmanu-facturersmayeasilydisagreesystematicallyby10–20%.

Inadditiontothesethreemonitors,aPerso-nalUltrafineParticleCounter(PUFPC100,Enmont,Cincinnati,USA)wasbrieflytestedtowardstheendoftheproject.ThePUFPC100isapersonalwaterbasedcondensationparticlecounterthatmeasurestheparticle

2. Measurement and sampling techniques

INST RUMENT MIN IDIS CDIS CMINI

N A NOTR ACER PA RTEC TOR PUFPC10 0

PUFPC 2 0 0

MICROA E TH A E 51

SIZE(H x W x D)(cm x cm x cm)

18 x 9 x 4.5 16.5 x 9.5 x 3 13.4 x 7.8 x 2.9 19 x 11 x 7 13 x 10 x 7 11.7 x 6.6 x 3.8

W EIGH T (g) 670 750 400 1,000 750 280

PA RT ICLE S IZER A NGE (nm) 10–300

Fastmode

20–120

Advanced mode

10–30010–10,000 ≥ 4.5 –

CONCENTR AT ION R A NGE 103–106 #/cm3 0–106 #/cm3 0–2*104 μm2/cm3 0–2*105 #/cm3 0–1 mg BC/m3

ME T RIC NC/dp/LDSA NC NC/d

p/LDSA LDSA NC Black Carbon con-centration

ACCUR ACY ± 30% ± 1,500 cm-3 ± 20% ± 10% ±1 μg BC/m3

S A MPLEFLOW (lpm) 1 0.3–0.4 0.5 0.3 0.05/0.1/0.15/0.2

T IMERE S OLU T ION (s) 1 3 16 1 1 1/10/30/60/300

BAT TERYL IFE T IME ( h ) 6–8 7 15 3.3– 6 6–24

T A B L E 1 : Specifications of personal monitors.

F I G U R E 1 : Charge acquired by the particlesvs calculated LDSA for spherical particles.


10

numberconcentration.ThenewerversionC200ismainlyidenticalwiththeC100,butsmallerandquieter.Table1providesanoverviewofthespecificationsoftheavailablepersonalmonitors.

2 .2 .1. PartectorThepartectoristhesimplestandsmallestoftheavailablepersonalmonitors.ItsschemeandaphotographareshowninFigure2.

Theunipolarchargerispulsedon-offsothatcloudsofchargedparticlesaregeneratedperiodically.[5]ThesechargecloudspassthroughaFaradaycageconnectedtoanelectrometer,which“sees”thechargecloudsandreactstothembyalwayskeepingtheentirecageelectricallyneutral,i.e.thechargeontheFaradaycageisalwaystheoppositeofthatinsidethecage.Bymeasuringthechargeflowingtothecage,thechargeontheaerosolcanbeinferred.Thesignaloftheelectrome-terhasasinusoidalshape,anditsamplitudeisameasureforthetotalchargeontheparticles,andiscalibratedforLDSAconcentration.ThisAC-typemeasurementhasthebigadvantagethatelec-trometerzerooffsetdriftsareautomaticallycompensatedfor,andthustemperature/humidityvari-ationshardlyaffectthedevice,anditsstart-uptimeisveryshort(16s)comparedtotheothers.ThetechnicalspecificationsofthepartectoraresummarisedinTable1.AnenhancedversionofthepartectorisadditionallyequippedwithanelectrostaticprecipitatorthatcancollectparticlesontoaTEMgridforsubsequentanalysis.Theinstrument,basedonthemeasuredconcentration,recquiresanadequatesamplingtime.

2.2 .2 . DiSCminiTheDiSCminihastwoelectrometerstagesthatcanbeusedtoinfermoreinformationabouttheparticles.[3]Particlesarechargedcontinuouslyanddetectedfirstina“diffusionstage”consistingofastackofstainlesssteelgrids,wherepreferentiallysmallerparticlesaredepositedbydiffusion.Thelargerparticleshaveahigherprobabilityofpassingthroughthisstage,andendupinafilterstage,whereallparticlesarecollected.TheschematicoftheinstrumentisshowninFigure3.Bymeasuringtheratioofthetwoelectrometerstages,theaverageparticlesizeisinferred,andtheparticlenumberconcentrationiscalculatedfromthetotalcurrentandtheparticlesizeinfor-mation.

TheDiSCminimeasuresbothcurrentsinparallel,andthusdeterminestheLDSAconcentration,particlenumberconcentration,andaverageparticlesize.DiSCminiistheonlyinstrumentthatusesapre-separator(impactor)thatremovesallincomingparticlesgreaterthan700nm.TheparticlesizerangeforaccurateLDSAconcentrationmeasurementsislimitedto20–400nm(seeabove).Fornum-berconcentrationmeasurements,thereisinprinciplenolowersizelimit.Onlythechargingefficiencydecreaseswithdecreasingparticlesizesuchthataveryhighconcentrationmaybeneededinordertoproducesufficientcurrent.nanoIndExexperimentsshowedthatDiSCminicanstillmeasurethe

F I G U R E 2 : Scheme (left) and photograph (right) of the partector. [6]

11

numberconcentrationof10nmparticleswithreasonableaccuracy.TechnicalspecificationsaregiveninTable1.

2.2 .3 . nanoTracer

TheAerasensenanoTracerusesaswitchedelectrostaticprecipitationzonetoachieveessentiallythesamemeasurementcapabilitiesastheDiSCminiwithasingleelectrometerdetectionstage.[4]Theprecipitatorpreferentiallyremovessmallparticlesfromthegasstream,i.e.whenitisturnedon,theelectrometermeasuresmostlylargeparticles;whentheprecipitatorisoff,theelectrometermeasu-resallparticles.AsinthecaseofDiSCmini,thetotalcurrent,measuredwhentheprecipitatorisoff,isproportionaltotheLDSAconcentrationandtheaverageparticlediameterandparticlenumbercon-centrationaredeterminedfromtheratioofthetwocurrents.TechnicalspecificationsofthenanoTra-cercanbefoundinTable1.

2.2 .4 . PUFP C100/C200ThePersonalUltrafineParticleCounter(PUFPmodelC100,Enmont,Cincinnati,USA)[9]isawaterbasedcondensationparticlecounter.Theincomingaerosolisexposedtoanatmosphere,supersa-turatedwithwatervapour.Thevapourcondensesontotheparticlesurfacesandmakesthemgrowtoopticallydetectablesizes.Thewaterreservoirlastsfor6hcontinuousoperationbeforeitneedstoberefilled.TheC100isequippedwithaGPSreceiverthattracksthemovementsofitsuserandallowsforlinkingtheexposuretothelocation.However,thisfeatureisintendedforoutdooruseandusuallydoesnotworkfor(indoor)workplacemeasurements.

F I G U R E 3 : Scheme (left) and photograph (right) of the DiSCmini. [7]

F I G U R E 4 : Scheme (left) and photograph (right) of the nanoTracer. [8]

F I G U R E 5 : Photograph of the Personal Ultrafine ParticleCounter; left: PUFP C100, right: PUFP C200. [10]



12

AnewerversionofthePUFP,themodelC200,whichissmaller,lighterandquieterthantheC100butwithotherwiseidenticalspecificationshasjustbeenintroduced.TechnicalspecificationsofboththeC100andC200aregiveninTable1.

2.2 .5. Black carbon monitor MicroAeth AE51

Theblackcarbonmonitor(BCM,seeFigure6)deviceMicroAeth(modelAE51,Aethlabs,SanFran-cisco,CA,USA)isaportable,self-containedandbattery-poweredaerosolmonitorthatusesairflowfiltrationandiscapableofmeasuringBlackCarbon(BC)withuptooneminutetimeresolution.Theinstrumentusesreal-timeabsorptionmeasurementsofawhite,PTFE-coatedborosilicateglassfib-refilter.Infraredabsorptionat880nmisinterpretedasareal-timesignatureforthemassofblackcarbonparticlesonthefilter.Quantificationcanbeachievedbyusingtheabsorptioncoefficientforblackcarbonof16.6m2/gfromtheliterature.[11]Theobtainedresultcorrespondstoanequivalentblackcarbonconcentration.ByusinganinletcyclonewithPM2.5at50ml/minflowrateandPM1.6at100ml/min,thedevicecanbeusedasananoparticlemonitor.

Anyparticleswithabsorptionat880nmdeviatingfromthatofblackcarbonwillcauseincorrectmasspredictions.Ithasbeenreportedthattheconcentrationofcarbonnanofibresand-tubescanbedeterminedwiththeMicroAethdevice.[12]However,theauthorreportedtheresponseoftheBCMtodropwithincreasingnanotubefilterloadalreadyatabout1/10ofthemanufacturer’srecommendedfilterload.Inaddition,nanotube-specificcalibrationwasreportedtobenecessary.

2 .2 .6 . Accuracy and comparability of the personal monitorsInthenanoIndExproject,alargecomparisonstudywasperformedinthelaboratorytocharacte-risetheaccuracy(comparedtoreferenceinstruments)andcomparability(deviationsbetweenNdevicesofthesametype)ofthepersonalmonitorsDiSCmini,partectorandnanoTracerfor17testaerosolswithparticlediametersrangingfrom10–700nm.ThestudyhasbeenconductedinthenanoTestCenteratIGF.

Asreference,ascanningmobilityparticlesizer(SMPS)wasused,whichrecordstheentireparticlesizedistribution,fromwhichallparametersmeasuredbythepersonalmonitorscanbecalculated.Ingeneral,agoodaccuracy[14]andgoodcomparabilitywasfoundforLDSAforalldevicesinvesti-gated.[15]Theaveragedeviationfromthereferenceinstrumentwasabout±10%inallcases,andthevariabilityaround±20%(withafewoutliers).Ageneraldependenceonparticlemorphologyorconcentrationcouldnotbefound.Onlyforparticleswithdiametersbelow20andabove250nm,largerdeviationswerefoundasexpectedanddescribedintheintroductionofthischapter.

Fortheparticlediameters,theaveragedeviationwasapproximately-20%fortheDiSCmini,and+5%forthenanoTracer,thenumberconcentrationwasoverestimatedby30%onaveragebytheDiSCminiand10%bythenanoTracer.Thevariabilitybetweeninstrumentsinthenumberconcentra-tionwasabouttwiceashigh(±20%)thaninLDSA–thisisnotsurprising,sincetheLDSAmeasu-rementisadirectmeasurementofacurrent,whereastheparticlediameterandnumberconcentra-tionareinferredbyassumingparametersoftheparticlesizedistributionwhicharenotnecessarilycorrect.Nevertheless,theperformanceofthepersonalmonitorsisclearlysatisfactory,asevenexpensivestationarynanoparticledetectorsareusuallyspecifiedtoanaccuracyof±10%atbest.

AloanunitofthePUFPC100hasonlyshortlybeenavailabletowardstheendoftheproject.IthasundergoneasmallerstudytocompareresultsobtainedwiththeC100withresultsfromstationaryreferenceCPCs.Awaterbased(TSImodel3787)andabutanolbased(TSImodel3776)CPCwere

13

better.ThewaterbasedreferenceCPCsho-wedinprinciplethesamebehaviouragainsthydrophobicDEHSparticles.Forworkplaceorambientmeasurements,itisexpectedthatthisfindingdoesnotlimittheusabilityofthePUFPC100,sinceitisveryunlikelythatsuchhighlypurehydrophobicparticlesordropletsareencountered.

2.3. Personal samplersIncontrasttodirect-readingpersonalmo-nitors,personalsamplersaredevicesthatcollectparticlesforsubsequentanalysis.Typicalsubstratesusedinpersonalsamplersarefiltersfortheanalysisofthemassconcen-trationsand/orchemicalcompositionoftheparticles,andflatsurfaces(e.g.Siwafer)orTEMgridsforelectronmicroscopicanalysisoftheparticlesizeandmorphologyor–ifcou-pledwithenergydispersiveX-rayfluorescencespectroscopy(EDXorEDS)–thechemicalcomposition.

usedasreferences.MeasurementswereconductedwithhygroscopicNaClandhydro-phobicDEHSparticlesofdifferentsizesandconcentrations.TheresultsshowthattheC100typicallyagreedwithin±10%withthereferenceCPCs.However,theinstrumentwasalmostblindforpurehydrophobicDEHSparticles.WhenthedispersedDEHScontainedonlyminorimpurities,theagreementwiththebutanolbasedreferenceCPCwasagainmuch


F I G U R E 6 : MicroAeth AE51 Black carbon monitor. [13]

(1) Depends on model (E, A or FAP)

T A B L E 2 : Technical specifications of commercial personal samplers.

INST RUMENT PGP N A NOBA DGE NRD T EMPA RTEC TOR

E SPN A NO10 0

TP S

2013 2015

SIZE(H x W x D)(cm x cm x cm)

(1) 16.5 x 9.5x 3

16.5 x 9.5x 3 – 14.2 x 7.8

x 2.915.24 x 10.16

x 7.62 15 x 6 x 3.5

W EIGH T (g) (1) 150 255 – 430 907 320

PA RT ICLESIZER A NGE (nm)

(1) 10–4,000 < 300 10–10,00020 nm-

supermicron range

20–600

S A MPLEFLOW (lpm) 2 0.6 1 2.5 0.45 0.1 0.001–0.01

S UBST R AT E

gold-coated track-etch membrane

filter

polycarbonatetrack-etched

membrane filterquartz filter

nylonmesh

screensTEM grid

TEM grid metallic/silicon

substrate

nickel TEM grid

BAT TERYL IFE T IME ( h )

depends on the pump > 8

dependson the pump

15 6–24 8


14

Measurementstrategiesthatusesamplinginstrumentsandsubjectsampledaerosolensemblestomicroscopicparticle(orfibre)analysiscanprovidevaluableinformationonthisparticlespectrumwithrespecttonumber,sizeandmorphology.However,missingagglo-merateandparticledensitiestogetherwithadeterminationofonlythe2D-projectedparticleareacurrentlylimitthereliabilityofsuchindi-rectparticlemassestimation.

Thesamplingofworkplaceatmospheresforsuchanindividualparticle-basedanalyticalapproachrequiresatleastanapproximateknowledgeoftheparticle(number)concen-tration.Thereasonisthatindividualaerosolparticleanalysisdoesnottoleratetoohighfiltercoverage,whichwouldleadtoattachingoroverlappingparticles.Inordertokeepthefiltercoverageatanacceptablelevel,approp-riatesamplingdurationsmustbechosenbasedonprevalentparticleconcentrations.Suchanapproachisapplied,e.g.,bythepartectorTEMdevice.ItintegratestheLDSAconcentrationduringTEMgridsamplingtostopelectrostaticdepositionatTEMgridcoverageoptimisedforindividualnanoparticleanalysis.Ifnotintegra-tedintheinstrument,theuseofanadditionalmonitormayberequiredtoestimatetheopti-malsamplingtime.Anoverviewofthetechni-calspecificationsofthepersonalsamplersisgiveninTable2.

2.3.1. Filter based samplers2.3.1.1. Personal sampling system PGP

ThepersonalsamplingsystemPGP(German:personengetragenesProbenahmesystem)is

apersonalfilterholdersystemforcollectingparticlesofdifferentfractions.ThePGP-EAisequippedwithawelldefinedporouspoly-urethanefoamasapre-selectorfortheE-(re-spirable)andA-(alveolar)fractions.Foroccu-pationalworkplacesamplingoffibres,thePGPversionPGP-FAPcanbeused.Itisgenerallyoperatedat2l/minairflowwiththefiltersurfa-cebeingorienteddownwards.Thefacevelocityofthefilteriskeptatalowlevelbyawideinletnozzleof30mmdiameter,seeFigure7.Techni-calspecificationscanbefoundinTable2.

2 .3 .1.2 . NANOBADGE

TheNANOBADGE(NANOINSPECT,Alcengroup,Paris,FranceandFrenchAlternativeEnergiesandAtomicEnergyCommissionCEA,Grenoble,France)isalightweight,battery-operatedandportabledevice,whichcancollectairbornepar-ticlesdirectlyinthebreathingzoneofaworker.Thesamplerisconnectedtoacassette,whosefilterisanalysedofflinebyX-rayfluorescen-cespectroscopy(XRF)providingacumulativemass-basedquantificationofthechemicalele-mentspresentonthefilters.ThemeasurementoftheengineerednanoparticleconcentrationbytheirconstitutiveelementusingXRFrepre-sentsaverypowerfulstrategy,becauseitisawaytogetridoftheexistinghighandfluctua-tingbackgroundlevelofnaturalandanthropo-genicnanoparticles.Moreover,itisanon-des-tructiveanalyticaltechnique,meaningthatthesamesamplecanbecharacterisedfurtherwithothertechniquessuchasscanningelectronmicroscopy(SEM).Theinstrumentisprovidedwithfilterunits(singleuse)inindividualzip

F I G U R E 7 : The PGP-FAP sampler.

F I G U R E 8 : The NANOBADGE sampler (2015 version).

15

bagsandpersonalIDbadge(personaluse,oneforeachpersonoperatingthesampler).Thefilterunitisasealedcassettecontainingapolycarbonatetrack-etchedmembranetocol-lectparticlesandisequippedwithaRFIDchiptostoredata(samplingtime,date,flowrate,errors,workerID,sampleID,...).Track-etchedmembranesallowparticlecollectionforsubse-quentanalysisbyXRF(elementalcompositionandconcentration)andSEM-EDX(particlesize,morphologyandchemicalidentification).Itcanbeequippedwithtwodifferentpre-separatorstoremovecoarseparticles(impactorsforre-spirablefraction,i.e.withd50=4μmorPM2.5)thatwerenotevaluatedinthisstudy.

Aftersampling,thecassettesareextractedfromtheNANOBADGEandsentdirectlyforanalysisandsubsequentdatarestitution(e.g.elementalmassconcentrationinthebreathingzoneaveragedoverthetotalsamplingtime).

Table2providesthetechnicalspecificationsoftheNANOBADGE.The2013versionoftheNANOBADGEsamplerwasevaluatedintheprojectnanoIndExandisreferredhereinas‘NANOBADGE’.Asingleon-offswitchmakestheNANOBADGEdevicerobustandverysimp-letouse.ThesamplingtimeandthesampledvolumeareautomaticallyloggedintotheRFIDtaglocatedinthesamplingunit.

Thedeviceisequippedwithred/greenlightsandanalarmsoundtowarntheuseraboutanymalfunction(e.g.inletclogged,dischar-gedbattery,etc.).Theencounterederrorsarelogged.Thedevicehasbeenrecognisedtobecomfortable,securelyfastenableanddoesnotrestrictthemobilityoftheuser.However,inquietworkplacesthedeviceisperceivedbysomeusersasnoisyandproducingannoyingvibrations2.

2 .3 .1.3 . Nanoparticle RespiratoryDeposition (NRD) sampler

Thepersonalnanoparticlerespiratorydeposi-tion(NRD,ZefonInternational,Ocala,FL,USA)[16]samplerwasdevelopedtobeusedasafull-shiftpersonalsamplerthatselectivelycollectsnanoparticlesinaworkplaceatmo-sphere.Todothis,firstlyanewcollectioncriterion,namelythenanoparticulatematter(NPM),wasdevisedinordertogetthetargetcollectionefficiencyofthesampler.TheNPMisthefractionofairborneparticlesthatwoulddepositinthehumanrespiratorytractbyBrowniandiffusion.BasedonthiscriteriontheNRDsamplerwouldcollectallparticlessmallerthan300nm,theminimumdepositionforsub-micrometreparticles,thatwhenin-haledcandepositanywhereintherespiratorytract(seeFigure9).

2Thisinformationreferstothe2013versionoftheinstrument.Thenewestversion(2015)oftheNANOBADGEisquieter.

F I G U R E 9 : Schematic of the NRD (left) ; NPM sampling criterion, ICRP total respiratory deposition and effective deposition on the dif fusion stage of the NRD sampler (right) [16].


1 cm


16

Thesamplerconsistsofarespirablealuminiumcycloneusedtoeliminateparticleslargerthan4μm,followedbyanimpactionplatewhereparticleslargerthan300nmarecollectedandadiffusionstagecontainingeighthydrophilicnylonmeshscreenswith11μmporesizeand6%porositythatcollectparticleswithaneffi-ciencythatmatchestheNPMcriterion.

Theparticlescollectedonthenylonfibersofthemeshscreenscanbecharacterisedeitherbychemicalanalysisorbyscanningelectronmicroscopytodeterminethesize,numberandchemicalcompositionofthecollectedpar-ticles.TheNRDsamplerhasnotbeenavailabletothenanoIndExprojectandisthereforenotfurthercoveredinthisGuidanceDocument.

2 .3 .1.4 . Prototypes personal samplers

Avarietyofdevelopmentsofpersonalnano-particlesamplerscanbefoundinthescientificliterature.TwoofthemhavebeenavailableinnanoIndExandarehenceexemplarilypresen-tedhere,namelythePM0.1personalsampler(PNS)[17]andthepersonalnanoparticlesampler(PENS)[18].

ThePM0.1personalsamplerconsistsofacom-merciallyavailabletwo-stagepre-cutimpactorusedtoremoveparticlesinthemicronsizerange(PM1.4-TSP),followedbyapre-cutinerti-alfilterthatuseswebbedstainlesssteel(SUS-316L)fiberstoremovefineparticles(PM0.5–PM1.4)andalayeredmeshinertialfilterusedforthePM0.1separation.ThelayeredmeshinertialfilterconsistsofcommerciallyavailablemeshcopperTEMgridssandwichedbetweencopperspacersandhastheadvantagethattheseprovideauniformstructureoffibersalignedperpendiculartotheflowdirection,maximisingtheinertialeffectonparticleswithlesspressuredropandnolossinseparationperformance.ByimmersingtheTEMgridsinanappropriatesolution,thecollectedparticlescanbeextractedforchemicalanalysis.

ThePersonalNanoparticleSampler(PENS)enablesthecollectionofbothrespirableparti-culatemass(RPM)andnanoparticlessimulta-

neouslyataflowrateof2L/min.Itconsistsofarespirablecyclone,usedtoremoveparticleslargerthan4μminaerodynamicdiameter,amicro-orificeimpactorwithacut-offdiameterof100nmandafiltercassettecontaininga37mmTeflonfilter.Themicro-orificeimpactorconsistsofafixedmicro-orificeplatewith137nozzlesof55μminnerdiameterandasiliconeoil-coatedTeflonfiltersubstraterotatingat1rpmtoachieveauniformparticledepositionandavoidsolidparticlebouncing.Particlesrangingfrom4μmdownto100nmarecollec-tedontheimpactionplateofthemicro-orificeimpactor,whilenanoparticlesarecollectedonthefilterofthefinalstage,althoughataratherhighpressuredrop14kPa.

2.3.2 . Samplers for electron microscopic analysis2.3.2 .1. ESPnano

Thecommercialhandheldelectrostaticpreci-pitator(ESP)isavailablefromESPnano(model100,ESPnano,Spokane,WA,USA).[19]ThissamplerissmallandbatteryoperatedandcollectsairborneparticlesontoTEMgrids.AschematicoftheESPisshowninFigure10.Thesamplerisintendedtobeusedmainlyinworkplaceexposureassessmenttotakesam-plesinlocations,whereareleaseofparticlesissuspected.TheTEManalysiscanthenpro-videproofforthepresenceorabsenceofcertainsubstances.Thesampleris,however,handheldandnotpersonal.

TheESPnanomodel100usesaunipolarcoro-nachargertogenerateionsnearatipelectro-

F I G U R E 1 0 : Schematic ( left) and photograph (right)of ESPnano 100. [20]

17

de.Whenapositivehighvoltageisappliedtothetip,acoronaisformedthationisestheair.ThetipelectrodefacesthesamplingelectrodewiththeTEMgrid.Consequently,thegenera-tedionsfollowtheelectricfieldlinesintotheperpendicularaerosolflow,wheretheycollidewiththeparticlestochargethem.Thechar-gedparticlesaredepositedontotheTEMgridwithinthesameelectricfieldthatisusedtogeneratetheionsandchargetheparticles.Asthedeviceisintendedtobeusedunderfieldconditionsaremovable“key”systemwasdesi-gnedthatwouldinsureafastandeasyrepla-cementofthesamplemediabetweendifferentsamplings.Thesamplemediacanbepre-loa-dedinthelabontothekeyandaftersamplecollectionthekeycanbekeptinairtighthol-dersuntilthesampleanalysisareperformed.

2 .3 .2 .2 . Thermal PrecipitatorSampler (TPS)

TheThermalPrecipitatorSampler(TPS,RJLeeGroup,Monroeville,PA,USA)usesthether-mophoreticforcetocollectnanoparticlesontostandardTEMgrids,forsubsequentanalysisofparticlesize,concentrationandchemicalcomposition.Itisthusnotamasssampler.Thesamplercollectsairborneparticlesbyapplyingarelativelylargetemperaturegra-dienttoanarrowflowchannel.Becauseofthetemperaturegradient,gasmoleculesonthe

hottersideoftheparticlehavegreaterkineticenergythanthoseonthecolderside,transfer-ringmorenetmomentumpercollisiontotheparticlethandomoleculesonthecolderside,causingathermophoreticparticlemotion.Theparticleswillmoveinthedirectionofdecrea-singtemperatureandwilleventuallydepositontothecoldersideoftheflowchannelthatincludestheTEMgrid.

TheTPSsamplesaerosolataflowratebet-ween1and10mL/minandutilisesaremo-vablesamplecartridgethatholdsahole-freecarbonfilmsupportedbya200meshnickelTEMgridontowhichparticlesaredeposited.ThecartridgecanbeslidintotheTPSbodyforsamplingimmediatelybelowthehotplatewhilemaintainingthermalcontactwiththecoldplatetoestablishthethermophoresiszone(seeFigure11).Becausenickelisferro-magnetic,thegridisheldinplacebyasmallmagnetlocatedbetweenthecoldplateandthegriditself.

Atransferfunctionwasdevelopedthatrelatesthenumber,sizeandcompositionofthecol-lectedparticlestotheonesofthetestaerosolinordertoreconstructtheparticlenumbersizedistribution.[21]TheTPShasnotbeenavailableduringthenanoIndExprojectandisthereforenotfurthercoveredinthisGuidanceDocument.

2.3.3 . Specific caseof carbon-based aerosolsTheservicetoprovideaquantificationofcar-bon-basedaerosolswasnotfullyoperationalin2013–2015fortheNANOBADGEsampler.Therefore,duringtheprojecttheNANOBADGEcassettewasadapted(prototyping)toallowsamplingcarbon-basedaerosolsonquartzfilterforsubsequentanalysisinathermal-opticalanalyser[22](LabOC-ECAerosolAnalyserfromSunsetLaboratory).Sootpar-ticlesweregeneratedbysparkgeneratorandbydieselenginesandweresuccessfullycol-lectedbythesampleronfreshlyfiredquartzfilters.Themassofelementalcarbonde-positedonthefiltershasbeendetermined

F I G U R E 1 1 : The thermal precipitation sampler (TPS): the overall device including the removable sample cartrid-ge (top); bottom: view of the TPS region containing the hot plate (a) , TEM grid holder (b) and cold plate (c). [21]



18

bythermal-opticalanalysis.Thelowmassofelementalcarbononthefilter,combinedwithcontaminationbyorganiccompoundswhenmountingthefilters,madeitdifficulttodrawreliableconclusionsontheresultsobtained.Nevertheless,aproof-of-concepthasbeenobtainedandthepreliminaryresultssuggestedthatwithsometechnicalimprovements(e.g.newwaystomountthefiltersandtosamplearepresentativepieceoffilter,etc)theNANO-BADGEsamplercouldprovidequantitativeanalysisofelementalcarbon(“black”car-bon).In2016,newsamplingcassetteswereproposedfortheNANOBADGEsamplerbuttheywerenotavailableduringthenanoIndExprojecttobeevaluatedandfurthercoveredinthisGuidanceDocument.

2.3.4 . Accuracy and comparability of the personal samplers2.3.4 .1. PGP-FAP

ThefiltrationefficiencyofthemembranesusedinthePGP-FAP(seeTable2)wasestima-tedtobe99.8%byconnectingaNanometerAerosolSampler(NAS,TSImodel3089)[23]downstreamofthefilterandevaluatingthembySEM.Collectedparticleswerecharacteri-sed,classifiedandcountedwithrespecttotheirmorphology.Duetothehighcollectingfilterareaof707mm2,inprincipleverylownanofibreconcentrationscanbedetected.However,thisrequiresacquiringasufficientnumberofSEMimages.TheGermanconcen-trationthresholdforclearancemeasurementoflessthan1000fibres/m3canforinstancebetestedbyevaluatingafilterareathathascollectedtheparticlesof2litresaerosol.Afteracollectiontimeof8hat2l/minflow,afilterareaof0.74mm2needstobeanalysedtotestforfibreclearance.ThenumberofSEMimagestomapthisareadependsonthecharacteristicstructuresizeoftheparticleorfibrestobecounted.Asaruleofthumbforthedetectionoffibres,thepixelresolutionofanSEMimageshouldcorrespondtothediameterofthefibrestobecounted.Ifverythinnanofibresneedtobedetectedandcountedhowever,therequiredpixelsize,i.e.,highmagnificationcanleadtoanenormousamountofSEMimages.

2 .3 .4 .2 . NANOBADGE

TheNANOBADGEfiltersareanalysedbyX-rayfluorescencespectroscopy(XRF)providingacumulativemass-basedquantificationofthechemicalelementspresentonthefilters.Thus,thesamplerprovidesthemassconcentrationinthebreathingzoneaveragedoverthetotalsamplingtime.ThequantificationbymassoftheelementsdepositedonthefiltersrequiresthattheXRFspectrometeriscalibrated,whichwasdoneusingapreviouslyreportedmetho-dology.[24]Inshort,setsoffiltersofincrea-singparticleloadingaregeneratedbysamplingcontrolledaerosols(e.g.ZnO,TiO2,etc.).ThefiltersarethenanalysedbyXRF,followedbydissolutionoftheparticlesforelementalquan-tificationbyICP-MS.Theplotofthenormali-sedXRFintensityversusthemassdeterminedbyICP-MSyieldsthecalibrationcurvesforthedifferentelementsstudied.TheseareusedtoconverttheX-rayfluorescenceintensitytomass.Thelimitsofdetection(LOD)forthefol-lowingelementshavebeendeterminedfortheNANOINSPECTXRFasshowninTable3.

Tofurtherillustratethevalidityofthesamplerforpersonalexposureassessments,thelevelofdetection(LOD)forZnOandTiO2havebeendeterminedusinganotherinstrument,theRi-gakuNANOHUNTERXRFthatwasusedatCEAfortheprojectnanoIndEx.[25]TheEuropeanAgencyforSafetyandHealthatWorkdistin-guisheslong-termandacuteexposure,theformerbeingarepeatedexposureaveragedoverworkingshiftsof8handthelatterapeakexposureaveragedover15min.[26]Thus,theLODshavebeenconvertedtoaerosolmassconcentrationsforafullshiftbasedonthe

LOD (ng/filter)

Al 100.2

Si 20.1

Ti 2.6

Ca 6.8

Zn 1.5

T A B L E 3 : Limits ofdetection of the NANO- BADGE using XRF analysis (NANO INSPECTXRF, optimised Z offset,0.1° angle, 200 secacquisition).

19

latestrecommendedexposurelevels(REL)oftheNationalInstituteforOccupationalSaf-etyandHealth(NIOSH).[27]TheminimumsamplingtimerequiredtodetectanexposureatorabovetheRELhasbeencalculated.AsshowninTable4,thelimitsofdetectionaremuchlowerthantheRELforthetwooxidesconsideredinthisstudy.Thedetectionofpeakexposureisalsopossible,sinceafewsecondsofsamplingatorabovetheRELaresufficienttoexceedtheLOD.SincetheLODareseveralordersofmagnitudesmallerthanthecurrentREL,theNANOBADGEsamplercanalreadyaccommodatetougherregulation,shouldtheexposurelevelsbeloweredinthefuture.

ThehighlysensitiveXRFtechniqueyieldstheelementalcompositionofthecollectedpar-

ticleswithsensitivityintheorderofafewtensofnanogramsperfilter.Cconsequently,itcouldbeusedeitheroverafullshift(e.g.8h)orduringshortoperations(e.g.15min)todetectacuteexposureevents.Themaindrawbackobservedisthatthesensitivityofthisanalyticaltechniqueisdecreasingdramaticallyforlightelements(Z<13)andthereforecarbon-basedparticlescannotbeanalysedwiththistechnique.

Severalmeasurementcampaignswereorgani-sedduringthecourseoftheprojectatIUTA,IGFandCEAonmonodisperse,polydisperse,compact,andagglomeratedparticles.Tho-semeasurementsallowedustoevaluatetheNANOBADGEsamplerinvariousconditionswithdifferentaerosols(sizedistribution,morphology,chemicalcomposition,etc.)andagainstdifferentgranulometers,countersandmonitors.

TheexampleshowninFigure12illustratestheperformanceoftheNANOBADGEcom-paredtoascanningmobilityparticlesizer(SMPS)bycarryingoutsimultaneousmeasu-rementsontestaerosolsofZnO.Theeffecti-vedensityandshapeoftheparticlespresentinthetestaerosolsweredeterminedexpe-rimentallyusingatandemDMA-ELPIsetup[28]tocomparenumber-baseddataobtainedwiththeSMPSwithmass-baseddataob-tainedwiththeNANOBADGE.

Thesamplerhasbeenevaluatedandvali-dateduptoasizeof200nmusingseveralaerosolsofZnOandTiO2particles.TheagreementbetweentheSMPSandtheNANOBADGEsamplerwaswithin±25%onalltestaerosolsforwhichtheeffectiveden-sitywasdetermined(seeFigure12).

Thisstudyhighlightsthefactthatthedensityoftheparticlesinaerosolsisofgreatimport-ancetocompareelectrical-mobility-basedresultstomass-basedmeasurements.Whenaerosolsaremonodispersewithperfectlysphericalnon-agglomeratedparticles,re-sultsfromSMPSandCPCmightbeeasilyconvertedtomass.However,incaseofmorecomplexaerosols(i.e.polydisperseoragglomerated),theeffectivedensityoftheagglomerateshastobepreciselyknowninordertoreducethedeviationbetweenmonitorsandsamplers.Therefore,metricconversionhasnotbeenperformedondatageneratedduringfieldmeasurements.Qua-litativeevaluationofevents,fromdirectreadinginstrumentsandcumulativemass-ba-sedquantificationofthechemicalelementspresentontheNANOBADGEfilters,couldbeofhighvaluefortheoccupationalexposureassessment.

REL for ultrafine dust fromNIOSH (μg/m3)

LOD(ng/filter)

LOD (μg/m3) for 8 h of sampling

Minimum samplingtime at the REL

ZnO 5000 30 ±20% 0.1 ±25% < 1 min

TiO2 300 12 ±25% 0.04 ±30% < 1 min

T A B L E 4 : Comparison between the recommended exposure levels (REL) published by the NIOSH and the limits of detection (LOD) of the NANOBADGE sampler for shift and acute exposure (Rigaku NANOHU TER XRF, optimised Z offset, 0.75° angle, 200 sec acquisition).



20

coronacharger[31]thatmayaffectthechar-gingefficiency[32].

Duringtheproject,wefoundthatbesidestheadverseeffectonthecoronaelectrode,siloxa-nesinthegasphasearepreferentiallyionisedincoronachargers,liketheonesusedinDiSC-mini,PartectororNanoTracer.Siloxanesareverylargeandheavymoleculesandthustheionpropertiesinthechargerchangedrastical-ly,resultinginadecreasedparticlechargingefficiency.WefoundthatespeciallytheDiSC-minisignificantlyunderreportstheparticleconcentrations,whensamplingthrough(new)conductivesiliconetubes[33].Thediscrepan-cycanreachafactoroftwoormoreincaseoftheDiSCmini/miniDiSC,butaremuchlowerforthepartector(seeFigure13).

Duringcomparisonbetweendifferentsam-plingtubematerials,thelowestdiscrepancywasexpectedlyfoundwithstainlesssteeltubes.However,thesearenotflexibleandcanhencenotbeusedinpersonalmeasu-rements.Tygon®tubeswerefoundtobethebestcompromisebetweenlowmeasurementartefactsandgoodpracticability.Thisisingoodagreementwithworkdoneinthe1980’s[36]whenconductivetubingwasnotyetavailable.Ifsamplingtubesareneededinpersonalexposuremeasurements,theuseofTygon®tubesisthereforerecommended.Incaseofpersonalsamplers,thesampling

2.4 . Periphery2.4 .1. Sampling tubes

Apersonalsamplerormonitorcanonlybemounteddirectlyinthebreathingzone,ifitissufficientlysmallandlightweight.Other-wise,theinstrumentcanbefixedonabelttosamplefromthebreathingzoneviaflexibletubing.Thesetubescanintroduceartefactsthatbiasthemeasurement.Particlelossesareunavoidableduringaerosoltransport.Theselossesaremainlydrivenbyparticlediffusion,sedimentation,andinertialand/orelectrost-aticdeposition.Sedimentationandinertiallossesaretypicallynegligibleincaseofnano-particlesduetotheirlowmass,butdiffusionlossesincreasewithdecreasingparticlesizeandwithincreasingresidencetimeinsidethetubeofagivendiameter.Therefore,samplingtubesshouldbekeptasshortaspossible.Inordertoavoidelectrostaticparticlelosses,thetubesusedshouldgeneratenoelectrost-aticchargesorelectricfields.Thisisbestachievedbytheuseofelectricallyconductivetubing.Forthispurpose,carbonimpregnatedsiliconetubesareverycommonlyconsideredastheoptimalchoiceinaerosolmeasure-ments.However,itwasfoundpreviously,thatsiloxanesmightdegasparticularlyfromnewsiliconetubes[29].Thesesiloxanescanbeadsorbedbyparticlesandaltertheirchemicalcomposition[30].Furthermore,theyformasiliconoxidedepositontheelectrodesofa

F I G U R E 1 3 : Deviation of LDSA concentrations measured with miniDiSC and partector caused by conductive silicone tubing [34] and Tygon® tubing [35].

F I G U R E 1 2 : Mass of ZnO calculated from the SMPS data and mass of ZnO measured by XRF analysis of the NANOBADGE filters (calculated from the mass of Zn). An effective density of 2.2 g/cm³ was used.

21

headisusuallymountedatornearthecollarboneoftheindividual,whilethepumpthatgeneratesthesamplingflowraterestsonthebelt.Flexibletubesareusedtoconnectthepumpandthesamplinghead.Aslongasthesamplingsubstrate(e.g.filter)residesinsidethesamplinghead,thechoiceoftubemate-rialdoesnotplayarole.If,however,thepar-ticlestobesampledaretransportedthroughatube,cautionshouldbetakentoassurelowlossesinsideandlowdegassingfromthema-terial.Degassedmoleculesmaybeadsorbedontheparticles’surfacesandchangetheirchemicalcomposition.

Henceitisrecommended,nottouseanysili-conetubestotransporttheaerosolinmeasu-rementsinvolvingunipolardiffusionchargersorsamplersusedforsubsequentchemicalspeciation.TheuseofTygon®tubingcurrentlyseemstobethebestcompromise.

2.4 .2 . Pre-separatorsInertialpre-separatorssuchasimpactorsorcyclonesarecommonlyusedinaerosolmeasurementsinordertolimittheparticlesizerange.Suchpre-separatorsremoveallparticlesthatarelargerthantheso-calledcut-offsizeofthepreseparator.Theyareusedtoadjusttheparticlesizerangeoftheaerosoleithertomatchacertainsamplingconvention,e.g.particles<4μm(d50)incaseoftherespirablefractionortolimitthepar-ticlesizestothemeasurementrangeoftheinstruments.Industyworkplaceatmosphe-res,suchpre-separatorsarealsohelpfulinprotectingtheinstruments.Limitationoftheaerosoltothemeasuringrangeoftheinst-rumentisofparticularimportanceforsomeofthepersonalmonitors.Themonitorsaredesignedtomeasurethenumberand/orlungdepositedsurfacearea(LDSA)concentrationoftheairborneparticles.Asdescribedabo-ve,measurementswithreasonableaccuracyareonlypossibleforalimitedsizerange.Apre-separatorwith400nmcut-offwouldforexampleberequiredforaccurateLDSAmea-surements,whichasofnowisnotcommerci-allyavailable.

2.4 .3 . Personal pumpsMostpersonal(filter)samplersrequiretheuseofapersonal,battery-operatedpumpinordertodrawtherequiredflowratethroughthesampler.Inprinciple,apumpshouldbechosentoprovidethenecessaryflowrate,takingintoconsiderationthepressuredropofthesampler.Novelnanospecificsamplerswithcut-offsizesinthenanometrerangeexhibitahigher-pressuredropthanclassice.g.respirablecyclones,thusnecessitatingstrongerpumps.Anotherimportantissuetotakeintoconsiderationwhenchoosingapumpfornanospecificsamplers,isthatthehigher-pressuredropincreasesthebatteryconsumption,i.e.thebatterylifetimeissignificantlyreduced.DuringthenanoIndExproject,itwasfoundthatpumpbatteries,designedfor>8hcontinuousoperationwithconventionalsamplerswerealreadydischar-gedafteronlyaround4hwhenusedtosamp-lewith(prototypes)ofsamplerswith100nmcut-size.

2.5. Field applicabilityof personal samplers and monitorsInstrumentsintendedtoassessindividualexposureofworkerstoMNMsinthefieldneedtosatisfyanumberofspecificrequire-ments.Basicaspectsaddressportabilityandincludebatteryoperationtime,robust-nessandwearabilityduringregularworkaswellasaspectsofmechanical,electri-calandexplosionsafety.Furthermore,theinstrumentshavetoproducereliablere-sultsnotonlyunderlaboratoryconditions,butalsowhencarriedbyapersoninthefield.Theusabilityofinstrumentsfordailyassessmentdependspredominantlyonitsweightandbulkiness.Sinceaminimumof8hbatterylifeisnecessarytocollectdataoverafullworkshift,batterycapacitycanbeasignificantweightandvolumefactorofportableinstruments.Therefore,measure-mentprincipleswithhigh-energyconsump-tionmaynotbeappropriateforpersonalapplication.



22

2.5.1. Practical considerationsTodeterminepersonalinhalationexposure,theinletoftheinstrumentmustbefixedreliablyintheworker’sbreathingzone.Highweightandvolumeimpedesdirectpositioningoftheinstrumentinthebreathingzonewithoutrestrictingtheworker’smobility.Ifatubeextensionoftheinletmustbeusedtoallowwearingabulkyorheavydeviceatthebeltorinabackpack,carefulselectionofappropriatetubematerialsbecomesnecessarytogetherwithestimationofassociatedwalllossesinsideofthetube.Asdescribedinsection2.4.1,degassingofsiloxanesfromsiliconetubingcansignificantlyaffectthemeasurementaccuracyofsomediffusioncharginginstrumentsandmayalsoaltertheparticles’chemicalcomposition.Inverydustyenvironments,impactorsorcyclonesusedforlimitingtheparticlesizerangeand/orforprotectionoftheinstrumentmayclogandre-quireperiodiccleaning.Inaddition,thecalibrationneedstobecheckedperiodicallybycomparingtheresultsfromseveralinstruments.

2.5.2 . Effects of wearing the instrumentsThequalityofpersonalexposuremeasurementsmayalsobeaffectedbythemodeofwearingtheinstruments.Potentialinfluencesmaystemfromavarietyoffactors,includingpersonalactivities,vibration,localairflowsorrelativevelocitiesbetweenthepersoncarryingtheinstrumentandtheaerosoltobemeasured.Sucheffectshavebeenobservedinnumericalstudies[37]andduringmeasurementsofexposuretomicronparticles[38].InnanoIndEx,weconductedmeasurementsofpersonalexposuretowellcontrolledNaClnanoparticleconcentrations.Duringthemeasurements,anindividualwasinsidea23m³chamber,whichwashomogenouslyfilledwiththetestaerosol.TheindividualcarriedtwopartectorinthePBZ,oneleftandoneright,whilecarryingoutcertainactivi-ties.Athirdpartectormeasuredwhilerestingonatableinsidethechamber.Nosignificantdiffe-rencesbetweentheresultsobtainedfromtheleftandrightsideofthebreathingzonewasobser-ved.Thepersonalresultsalsoagreedwithin±10%withthoseobtainedbytheinstrumentrestingonthetable.However,itwasnotedthatthedatashowedhigherfluctuationsduringactivitieslikewalkingthanduringquietsitting,buttheaveragewasunaffected(seeFigure14).Thiseffectisnotunexpected,aswalkingmaydisturbthelocalflowandparticledistribution,leadingtoshortliveddifferences.Intheexperiments,thiseffectwaspurelyrandomandthereforeonlyaffectedsingledatapoints,butnotthemeanconcentrations.Thesameexperimentswererepeatedafterswapping

F I G U R E 1 4 : Boxplots of LDSA concentration ratios measured left and right in the PBZ and in comparison with reference concentrations measured on a table

23

theinstrumentsleftandright(datanotshownhere).Inthatcase,theratiosreversed,i.e.the(small)differencesstemfromtheinstru-mentsthemselves,butnotfromthepositio-ningoftheinstrumentsinthebreathingzone.Itisexpectedthatthesefindingsalsoapplytoothermonitorsandsamplers,atleastaslongasthesamplingflowratesaresimilar.

2.5.3 . Is the availableinstrumentation applicableto field studies?Thecommerciallyavailablepersonalinstru-mentsstudiedintheprojectnanoIndExinclu-dedpartector,partectorTEM,NANOBADGE,MicroAethAE51,FAP-PGP,DiSCmini,ESPnanoandPUFPC100.Theyallsatisfythebasicrequirementsofwearabilitythatwerediscus-sedaboveandshowedgoodfieldapplicability.DiSCmini,ESPnanoandPUFPC100requireadditionaltubeextensionstomeasureintheworker’sbreathingzone.Selectionofappropri-atetubesrequirescareandneedstobedo-cumented(see2.4.1).ForprototypesamplerslikePNSandPENSthathavecut-offdiametersinthenanometrerangeandhencehaveahigh-pressuredrop,thepumpoperationtimedidnotalwayscoverafullshift’sdurationof8h.Itwasshownthatwearingoftheinstrumentsdoesnotaffectthemeasurementaccuracy,butitmaycausealargerscatterofthedataincaseofmeasurementswithhightimeresolution.

Thecurrentlycommerciallyavailablerangeofinstrumentsdoesnotallowpersonalsur-veillanceofcompliancetomassbasedMNMexposurelimitsinaneasyway.Onlyinexpo-suresituationswhereMNMofclearchemicalormorphologicalsignaturearebeinghandled,filterbasedsamplingwithsubsequentanalysiscanprovideestimatesofMNM-specificmassorfibrenumberconcentrations.ExamplesareXRF-basedquantificationofMNMmadefromlow-backgroundelementsandSEM-basedindi-vidualfibrecounting.

Toconclude,theavailableinstrumentsaretechnicallymatureandreliablyuseableinfieldmeasurementsofpersonalexposuretoairbor-

nenanomaterials.Eachoftheinstrumentshasitsadvantagesanddisadvantages.Datacanbemeasuredbothwithhightimeresolutionorastimeweighted(e.g.shiftbased)averages.Avarietyofexposuremetricscanbedetermined.Theirapplicabilitytolegislativestandardsisyetunclearasthereisnoofficialguidanceonwhichmetricshouldbeusedtoexpressexpo-suretoairbornenanomaterialsmeasuredbothwithhightimeresolutionorastimeweighted(e.g.shiftbased)averages.Avarietyofexpo-suremetricscanbedetermined.Theirapplica-bilitytolegislativestandardsisyetunclearasthereisnoofficialguidanceonwhichmetricshouldbeusedtoexpressexposuretoairbor-nenanomaterials.


Chapter 3

Performance of measurements

25

duringashift,itmayadditionallybenecessarytodeterminetheaverageshorttermexposureconcentrationduringtheseepisodesofpotenti-allyhighexposure.Asdosecalculationwillalsobeofconcerninthosecases,acommontimebaseforshort-termmeasurementsshouldalsobedefined.Forthispurpose,atimebaseof15minisrecommended,eveniftheepisodesaresignificantlyshorter.

3.1.2 . Exposure assessmentwithin risk assessment/riskmanagement proceduresThisdocumentdealswithexposureassessmentwithinriskassessment/riskmanagementpro-cedures.Incaseswhereoccupationalexposurelimits(OELs)forNOAAareexisting,exposureassessmentstrategyetc.willhavetobeper-formedaccordingtonationalstandardsandpossiblyEN689.Specificdirectionsforthesecasescanbefoundinthisstandard.

Generallythequestiontobeansweredinthedescribedtypesofexposureassessment(riskassessment/riskmanagement)willbe:“Arethecurrentmeasurestakeninriskmanagementsufficienttominimiseworker’sriskfornegativehealtheffectsbyexposureagainstNOAA?”Forthispurpose,afirstconsiderationmustbegiventothehazardresultingfromaspecificNOAA,i.e.itstoxicity.Inordertogiveavalidestimationofriskthecurrentstateofdevelop-mentofpersonalmonitorsdoesnotallowfortheiruseinexposureassessmentofhighlytoxicNOAA.Apossibleexampleofthisclassofsubstancesmightbesomespecificcarbonnanotubes(CNTs),i.e.longandrigidones[40].

Personcarriedinstrumentscanbeusedfordirectandon-linenumberconcentrationorlungdepositedsurfaceareaconcentrationmea-surements(“monitoring”,seesection2.2).InadditiondifferentpersoncarriedinstrumentscanbeusedforsamplingtheNOAAinquesti-onandtosubsequentlyanalysethesesampleswithappropriate(wet-)chemicalorinstrumentalmethodslikeXRF,ICP-MSorelectronmicros-copy(“sampling”,seesection2.3).Thelatter,i.e.thequalitativeidentificationofNOAAin

3.1. Description andselection of theassessment task3.1.1. Epidemiological studiesEpidemiologicalstudieswithrespecttona-no-objectsandtheiragglomeratesandaggre-gates(NOAA),trytoestablishdoseresponserelationshipsforNOAAandselectedmedicalendpoints.WhereasthelatteraspectsarenotwithinthescopeofthisGuidanceDocument,theaccuratemeasurementanddocumentati-onofNOAA-doseswithintheselectedgroupofworkersandtheirworklife(oraselectedperiodwithinthatworklife)needtofulfilcertainpre-requisites(forgeneralaspectsofexposureassessmentwithinepi-studiesseee.g.ref.[39]withtheexampleofcrystallinesilica).Asageneralprinciple,allparametersfortheintendedmeasurements(“protocol”)needtobediscussedwiththeepidemiologicalscientistsastheywillinalmostallcasesbeacompromisebetweentheepidemiologicallywantedandtheanalyticallypossible.Atfirst,theexactairbornecomponent(s)tobeinves-tigatedneed(s)tobedefined.Thisincludesthechoiceofthemetrictobeusedfortheairborneparticles.Asnormallyadoseisthetargetofexposureassessmentwithinepi-stu-dies,additionallythetimeperiods(periodsofactualsampling)overwhichexposurehastobequantifiedneedtobedefinedaswell.FormanyepidemiologicalstudiesasocalledJobExposureMatrix(JEM)willbethefinaloutco-meofexposureassessment.AJEMassignsaverageexposurelevelstojobtitlesbyca-lendarperiod.Basedonathoroughstructuralevaluationoftheworkplacesinquestionjobtitles(i.e.jobsforwhichhomogenousexposurelevelscanbereasonablyassumed)needtobedefined.Forthesejobtitles,typicallyaverageshiftexposureneedstobedetermined.Nor-mallytheaveragingtimeis8h,butsignificantdeviationsarepossible.Thesumofshiftdosesovertheworkperiodinquestionallowsforcalculationoftherespectivedose.Incaseswheretheexposurepatternisveryinhomoge-neous,forexamplebecauseNOAA-exposureonlyhappensduringveryshorttimeperiods

3. Performance of measurements


26

Themostimportanttaskwhenchoosingatieredapproachistodefinedecisioncriteria,whichal-lowforoneoftheabovementionedcasestobeidentified.Examplesforthesearegivenin[41].

3.2 . Selection of the measurement devices3.2 .1. Epidemiological studiesDependingontheoutcomeofthedescriptionandselectionoftheassessment,task(seeabove),suitablemeasurementequipmentneedstobeselected.Availabilityofproceduresandequipmentwillalmostcertainlyhavealreadybeenamajortopicintheabovementioneddiscussionandselectionprocess.However,devicesandmeasurement/samplingproceduresshouldatthispointbeselectedfromtheavail-able(chapters2.2and2.3)andsuitable(chap-ter2.4)instrumentationwithaclearviewoftheabove-discussedparameters.Thismustalsoincludepracticalconsiderationsliketheneedforuseofexplosionproofequipmentorpossiblesamplingtimerestrictionsinbat-tery-poweredequipmentwithrespectoftheneedofcoveringshiftexposure.

Personalsamplingshouldindoubtbeselectedinsteadofstaticsampling,althoughbothcanhaveanaddedvalue.Shiftexposureshouldpreferablybemeasuredbydirectcoverageofthewholeshift.However,itmayalsobecalcu-latedfromtimeweightedaveragingofdistinctlydifferentperiodsofexposureperiodswithinashift,ifmorepracticale.g.:selectivity,sensitivi-tyandfurtherqualityparametersoftheselec-tedequipment/proceduresneedtobecoveredaswell.

3.2 .2 . Exposure assessmentwithin risk assessment/riskmanagement proceduresWhichpersonalmonitororsamplerischosendependsonthespecificworkplacesandNOAAinquestion.Thediffusionchargertypeinstru-mentsarerobustandwidelyapplicable;how-ever,theirlimitationsatparticlediametersabove400nmshouldbetakenintoaccount.[14]Iftheworkplaceexposurewillbecharacterised

airbornestate,ismandatory,ifdirectreadinginstrumentscannotsufficientlydifferentiatebetweenNOAA-concentrationandtherespec-tivebackground,whichfrequentlywillbethecase.Inordertoanswerthecentralquestionofsufficiencyofriskmanagementmeasu-resandtooptimisetheuseoftheavailableresources,quiteoftentheuseofaso-calledtieredapproach[41]isadvisable.Thismeansthatinafirststeptherelevanceofexposureassessmentischeckedbyevaluatingpossib-leexposureusingavailabledocumentsandpre-knowledgeoftheworkplacesinquestion(tier1).Ifrelevancecannotbedenied,thesecondtierofexposureassessmentwillbenecessary,whichinvolvessimple,fastandsufficientlyreliablemeasurements.Personalmonitorsareespeciallyusefulinthiscontext(tier2or“basicassessment”),althoughtheyarenotexplicitlyincludedintheapproach.TheuseofpersonalsamplersmayincreasethereliabilityoftheirresultsbyhelpingtoclearlyidentifythepresenceoftheNOAAinquestion.

Resultsofthesetier2measurementscanbe:

a.Exposureassessmentwasinconclusive: Insufficientinformationonthenature/ quantityoftheriskisavailable/was obtained.Inthatcase,aso-called“ex- pertassessment”ortier3assessment willbenecessarywithmuchmoreelabo- rateequipmentandeffort.Personcarri- ed instrumentsMAYbepartoftier3 measurements,butwillnotbesufficient inmostcases.b.Exposureassessmentwasconclusive– riskmanagementmeasurementsare notsufficient:Inthatcase,additionalmea sureshavetobeimplementedandsub sequentlyexposureassessmenthasto berepeated.c.Exposureassessmentwasconclu- sive–riskmanagementmeasurements aresufficient:Inthatcase,theoutcome oftheexposureassessmenthastobe documentedandthenormalrepetition cycleofriskassessmentprocedures enteredintheparticularcompany.

27


bysignificantagglomeration/aggregationoftheprimarynanoparticles,largerparticlesizesneedtobetakenintoaccount.Thiswillactuallyregularlybethecase.Opticalinstru-ments(photometers,opticalparticlecoun-ters)mayadditionallybeusedthentocovertheparticlesizerangeof300nmtoafewμm(i.e.therespirabledust).

Theselectionofpersonalsamplerswillmainlyberuledbytheintendedsubsequentanaly-sis.SochemicalanalysislikeXRForICP-MSwillrequireaminimumofparticlemasstobesampled(LOD)withafilter.XRFandSEManalysesrequiretheparticlestobedeposi-tedonaflatsurfaceofe.g.atrack-etchedmembranefilter.Alternatively,transmissionelectronmicroscopy(TEM)andsubsequentelementaryanalysiswillrequirespecificsamplingmedia(see2.3).

Selectionofthepropermonitorsandsamp-lersiscrucialforthesuccessofexposureassessmentandneedstobedonewithrespectofconditionsintheworkplace(likeworkpattern,suspectedconcentrationrange,natureofbackground,NOAAinquestionetc.)andshouldbewelldocumented(seebelow).

3.3. Selection of the work-places or emission sources to be investigated3.3.1. Epidemiological studiesFromthejobtitleselectionprocessdescri-bedinsection3.1.1,atthispointthegene-raltypesofworkplacestobecoveredarealreadyknown.Forepidemiologicalstudies,emissionmeasurementsarenormallynotsuited.

Forallworkplacetypes(jobtitles)tobecovered,atthispointadetaileddiscussionoftheexposuredeterminants,i.e.alltheparametersinfluencingtheheightofexposu-reduringagivenshift,needtobediscussedanddocumented(seeforexampleTRGS402[42]).Theresultsofexposureassessmentforeachjobtitleneedtoberepresentativefor

thatjobtitleandtherespectivecalendartimeperiodoftheJEM(seeabove).

3.3.2 . Exposure assessmentwithin risk assessment/riskmanagement proceduresOncetheequipment(monitors,samplers,samplingmediaetc.)hasbeenchosenithastobedecidedwhetheremissionorexposureconcentrationswillbebettersuitedtoanswerthecentralriskmanagementquestion.

Generally,therespectiveconcentrationswillhavetobedeterminedwithandwithoutthespecificcontrolmeasures(technical,organisa-tional,personalinthatorderofrelevance)inplace.Themostimportantobstacletounam-biguousresultsofthesemeasurementswillbelargeandfluctuatingbackgroundconcen-trationsinthesamemetricaschosenforthemonitoringexposureassessment.Therefore,recordingcontinuously(oratleastquasi-con-tinuously)oftheconcentrationsandcompari-sonwithadetailed“diaryofevents”(log)overthecourseofmeasurementsisaveryusefulapproachinordertoidentifynon-workrelatedepisodes/eventsinfluencingtheexposure/emissionconcentrationandthebackgroundalike.[43]Additionally,qualitativeidentificationoftheNOAAinquestionbysuitableanalysesofthesamples(seeabove)willbenecessaryinmostcases.Inmanycases,groupingoftheinvestigatedworkplacesintoso-calledsourcedomains[44]willalsobehelpful.

Timebaseofthemeasurementswilllargelyberuledbythetimecharacteristicsofthetaskinquestion.Therefore,shortactivities(e.g.emptyingofacontainerofnanomaterialsorcleaningofasmallproductionsite)shouldbeaccompaniedbymonitoringduringtheseactivities.Thereisacertainconflictbetweenmonitoringandsamplingifveryshortactivitiesaretobeinvestigated,aslimitsofdetectionofcertainanalyticalmethodsmayrequirelongersamplingtimesthanmonitoringperiods.Infact,sometimessampledmassmaysimplynotbeadequateforthesetypesofanalyses.Inthelattercases,repletionoftasksandsampling


28

•Spatialcompensationofthebackgroundby measurementclosetothe workplaceinquestion(duringactivities)and awayfromit(“Nearfield”,“Farfield”)•Temporalcompensationbymeasuring withandwithoutthespecificactivitiesof theworkplace•Acombinationofthelattertwo

Inaddition,specialconsiderationshouldbegiventothe“outdoor”(i.e.outsidetherespec-tivebuilding)background,whichmaymostlybeinfluencedbycombustionengineexhaust.Thefinalagreeduponmethodofbackgroundtreatmentneedstobepartoftheprotocolfortheperformanceofexposureassessmentintherespectiveepi-study.Forriskassessment/riskmanagement,thenon-workplacerelatedbackgroundofultrafineparticlesmustbepro-perlyaddressedandtreatedaswell.Thesameconsiderationsasmentionedaboveapplyaswell.

Specificguidanceonbackgroundtreatmentisgivenin[45].

3.5. Performance of the measurements3.5.1. Epidemiological studiesTheresultsofdiscussionsanddecisionsasofparagraphs3.1.1to3.4.1arecondensedintoafinal“protocolfortheassessmentofexposu-re”whichisbasicallythestandardoperationprocedureforthatmeasurementcampaign.Allmeasurementsshallbeperformedaccor-dingtothatprotocolanddocumentedrespec-tively.

3.5.2 . Exposure assessmentwithin risk assessment/riskmanagement proceduresTheactualmeasurementsshallbeperformedaccordingtoapre-determinedworkplan(“protocol”),preferablyfollowinganexistingstandardoperationprocedure.Thisprotocolhastodescribeindetail,howthetieredappro-ach(ifany)forexposureassessmentisworkingintherespectivecase.Specialconsideration

ontoidenticalsamplingmediamaybeawayoutofthisproblem.

NOTE:Forcomparisonofthemeasuredex-posureconcentrationswithanexistingOELshiftmeasurements/samplingandadditionallyshort-termmeasurement/sampling(15minrecommended)mayberequiredadditionally.

3.4 . Backgroundmanagement3.4 .1. Epidemiological studies and risk assessment/risk management proceduresAsbackgroundtreatmentanditsdescriptionismoreorlessidenticalforepidemiologicalandriskrelatedstudieswedonotdescribethemseparately.

Dependingonthenatureofthecomponent(NOAA)tobedetermined,theprocedureforbackgroundtreatmentneedstobediscussedbeforehand.

Theepidemiologicalstudyinquestionmaynotbeinterestedinthediscriminationofurbanbackgroundparticlesfromtheonesresultingfromworkplaceactivities,dependingonthemedicalendpointtobedetermined.Ifthatisthecase(e.g.becauseunspecificresponseoftheairwaystoultrafineparticleswithinaspe-cifiedsizerangeistheselectedendpoint),nofurtherseparatetreatmentofthebackgroundmaybenecessary.

Inmanycases,however,thestudyinquestionwillbeinterestedinspecificNOAAworkplaceexposureandinthosecasestheomnipresent,non-workplacerelatedbackgroundofultrafi-neparticlesmustbeproperlyaddressedandtreated.Howwellthisisdonewilldeterminethequalityofthestudyinamajorway.Thefollowingpossibilitiesexist:

•SpecificmeasurementofONLYthe NOAAinquestionwithdirectdiscrimi- nationofthebackground(e.g.chemi- calormorphologicalspeciation)

29

hastobegiventothedecisioncriteriaasof3.1.2above.Actualmeasurements(applicationofmonitorsandsamplers)shallfollowexistingstandardoperationprocedures.[46]

3.6. Data evaluationDatatreatmentandevaluationisdescribedinchapter4indetail.

3.7. Documentation3.7.1. Epidemiological studiesTheactualdocumentationisoneofthecoreelementsoftherespectiveepidemiologicalstudyandisnotfurtherdiscussedhere.Ofcourse,itshallcontainALLrelevantaspectsdiscussedinthischapter.

3.7.2 . Exposure assessmentwithin risk assessment/riskmanagement proceduresThedocumentationoftheresultshastobedescribedintheprotocolasof3.5.2above.Ithastoinclude

•allrelevantdatafortheactualper- formanceofassessment(“who,where, when”),•decisionstakeninplanningoftheactual measurementcampaign(seeabove)in- cludingthedecisioncriteriawithinthe scopeofatieredapproach(ifany),•theresultsofallmeasurementsduring monitoring(preferablyalsoprimarydata ofmonitors),•theresultsofaveragingofthemonitors’ saveddataduringthesamplingperiods,•thesamplingdetailsofmonitorsand samplers,•andtheresultsofsampling

Thelatterincludesdataforallpre-selectedme-trics,e.g.respirablemassconcentration,massconcentrationofaspecificNOAA,qualitativeanalysesoffiltersamplesandidentifiedNOAAetc.Aformalevaluationofthemeasurements/assessmentswithrespecttoandtakingintoaccountthepre-selecteddecisioncriteriaofa

tieredapproach,ifany,ismandatory.Thiswillresultinoneofthepossibleformaloutcomesdescribedin3.2.2.

3.8 . Quality assuranceAllinstrumentsusedinfieldstudiesshouldworkasreliablyandasreasonablypossible.Amanufacturercalibrationofeachinstrumentpriortoeachfieldstudyiscertainlynotpos-sibleduetotimeandfinancialrestrictions,butthecalibrationoftheinstrumentsshouldfrequentlybecheckedbycomparingthemwitheachother.Thesimplestcheckthatshouldbedoneveryfrequentlyistomeasure(andadjust,ifnecessary)theflowrateoftheinstruments.Anotherrathersimplecheckoftheinstru-ments’calibrationistoletseveralinstrumentsrunsidebysideoveracertaintime.Resultsshouldbecomparedandeachdatasetshouldbecheckedforanyobviousanomalies.Thissimpletestshouldbeperformedpriortoeachmeasurementcampaign,ideallyuponarrivalattheworkplace,becauseofpotentialdamagesduringthetransportationoftheinstrumentstothesite.

Moreelaborateroundrobintestsusingwelldefinedtestaerosolsase.g.carriedoutbyKaminskietal.[47]shouldalsobecarriedoutperiodicallytoelaborateonthelimitsoftheinstruments’comparabilityandassurethatinstrumentsarealsocomparablewiththeonesfromotherinstitutions.


Chapter 4

Data collection,analysisand storage

31

wasonlyavailabletopartnersinthePEROSHgroupandtwoconsortiummemberswhohadbeengrantedpermissiontouseit.Notallpart-nersinnanoIndExhadaccesstoNECIDandsotofacilitatethedatacollectionforstorageinNECIDexceltemplatesweredeveloped,basedonNECID3.Anaccompanyingdatacollectionprotocolwasalsodevelopedtoassisttheex-posurescientistincollectingtherelevantdataandensuringthatthesamplers/monitorswereidentifiable.AsNECIDaimstocollectdataonallpossibleaspectsofthefieldstudythetemplatesarequitelargeandcouldpotentiallytakesometimetofilloutforaspecificstudy.Thereforeasimplertemplatewasdevelopedtocollectthecontextualdataessentialforthedataanalysissothattheanalysiscouldbeundertakenwithoutdelay.

Itisimportanttoconsiderusingastandarddatacollectiontemplateinfieldstudies,par-ticularlywhenconductedbymultipleorgani-sationsasthiswillensurethatthesamedataarecollectedforallfieldmeasurements.WhiletheNECIDstructureisnotyetaccessibletoallorganisations,theideaofharmonisingdatacollectiongoingforwardwillresultinthepotentialtopooldatafromfieldstudiesforfuturework,suchasinvestigatingexposureforepidemiologicalstudies.

4.2 . Data analysisThefocusofthedataanalysisinnanoIndExwastocomparethemeasurementsobtainedfrommonitors.Inparticular,tocomparethemeasurementsobtainedfromthepersonalmo-nitorstoeachother,andtoreferencemonitors(stationaryequipment).

4.2 .1. Preliminary analysisAswithalldataanalysis,thefirststepistovisualisethedatatostarttoformanopinionofwhatthestatisticalanalysiswilltellyou.Time-seriesplots(Figure15a)allowforavisualcomparisonoftheentiresetofdata.Whenthemainpurposeistocomparemeasurementsof

4.1. Data collectionIncarryingoutfieldstudies,itisimportanttoconsiderhowtherelevantdatashouldbecol-lected.Inordertoallowforappropriateana-lysisandinterpretationofthedataobtainedfromthemeasurementequipment,contextualinformationshouldalsobecollected.Thiscon-textualinformationshouldincludeanumberofvitalpiecesofinformation:

•Adescriptionofthetask(s)thatarebeing carriedoutduringthemeasurementperiod•Thepositionofthesamplersandmonitors, onpeopleandintheroom.Includingidenti- ficationofmovementofanyoftheinstru- ments(i.e.movingfromnearfieldtofar field)•Informationonanyeventsthatoccur- redduringthemeasurementperiod · Relevanttotheequipmentsuchas periodofcleaning · Relevanttothetasksuchas emptyingofbagofparticles · Relevanttoactivityintheroom(or outside)suchasopeningwindows•Informationonthenanomaterials beingusedintheprocess · Typeofmaterial,size,density, morphology•Informationontheroom · Roomdimensions · Ventilation

AdatabasestructurehasbeendevelopedbythePEROSHgroupwiththeaimofenablingcollectionofmeasurementdatainaharmo-nisedformat.ThisdatabaseiscalledNECID(NanoExposureandContextualInformationDatabase)[48].TheNECIDdatabasehasbeendesignedtostoreallavailablecontextu-alinformationforameasurementseriesandthereforeincludestablesontheworkplaceitself,thematerials,taskscarriedout,othersourcesofemission,availabilityofgeneralandlocalcontrols,theworkers,theirexperienceanduseofPPE.AtthetimeofplanningthenanoIndExfieldstudies,theNECIDdatabase

3ThesedocumentsareavailablefordownloadonthenanoIndExwebpagewww.nanoindex.eu

4. Data collection, analysis and storage


32

differentinstruments,themainthingstolookforarewhetherthemonitorsbehavethesame,dotheymeasureexactlythesame,dotheymoveinparallel(presenceofanoffset),doesonemonitorbehavedifferentlyathigher/lowerconcentrations,andisthisbehaviourconsistentovertheentiremeasurementperiod.Thesameprocesscouldbeusedtocomparemeasurementstakenindifferentpositions.Timeseriesplotsarealsoimportantforidentifyingpeaksandtroughsandallowforanini-tialimpressiontomade,whencomparedwiththecontextualinformationregardingthetimingsoftasksandbackgroundevents,onwhethertheyareassociatedwithanythingthathasbeenobservedandrecordedduringthemeasurementperiod.

Boxplots(Figure15b)allowforanimpressionoftheoveralldistributionofconcentrationsmeasu-red,whetheronemonitorhasmorevariationthananotherandwhetherthemedianandrangeisthesame.Scatterplots(Figure15c)ofthemeasu-rementsobtainedfromtwomonitorsshouldbeexaminedtodeterminewhetherthetwomonitorsmeasureexactlythesame(allpointsliealongthelineofequality),thereisaconsistentoffset(thepointslieaboveorbelowthelineofequalitybutareparalleltoit),orwhetherthereisevidenceofsomeotherrelationshipbetweentheresultsfromthetwomonitors(astraightlineorcurvethatdoesnotfollowthelineofequality).Afinalplotthatisusefulwhencomparingtwomeasure-mentsistheBland-Altmanplot(Figure15d).Thisplotshowsthedifferencebetweenmeasurementsagainsttheaverageofthetwo.Fromthisplotyoucanseewhattheactualdifferenceisbetweenthetwomonitorsandmakesomeevaluationofwhetherthisdifferenceisimportantornot(i.e.adifferenceof5willbeimportantintermsofmeasuringtemperature,butnotintermsofmea-suringnumberconcentration).Lookingattheplotasawhole,youcanevaluatewhetherthereisanypatterninthepoints,doesthedifferenceincreasewithincreasingconcentration(upwardsslopingline),decreasewithincreasingconcentration(downwardslopingline)oristheresomeotherrelationshipevident.

Theseplotsandyourevaluationofthemwillpro-videsomeimpressionoftherelationshipbetweentwomonitors.

F I G U R E 1 5 : Examples for graphical representation of the measurement data, (a) time series, (b) box plot, (c) scatter plot, (d) Bland-Altman plot.

a

b

c

d

33

Thiscanbecalculatedfortheentiremeasure-mentperiodandsummarised.Themoststra-ightforwardwaytoevaluatethebiasmaybetopresentitasaboxplot(Figure16).Therangeofbiascanthenbeexaminedandanevalua-tionmadeaboutwhetheritfallswithinanac-ceptablerange(±30%isusuallythoughofofasacceptableforequipmentmeasuringpartic-leconcentrations).Theboxplotshowsthatthebiasiswithintherangeof±30%.Twooftheinstrumentsarenegativelybiased,indicatingthatthemeasureconsistentlylowerthanthereference,whiletheothersarepositivelybia-sed,indicatingthattheymeasureconsistentlyhigher.ThebiasislowerfortheminiDiSCandDiSCminithanthepartector.

4.2 .2 .2 . Precision

Theprecisionisameasureofthevariability.Thiscanbecalculatedby:

Plottingthebiasandprecisionagainstthetimeand‘true’concentrationallowustoseeifeitherofthesefactorshasanimpactonthebiasandprecision.Doesthebiasincreasewithincreasingconcentration?Doestheprecisiondecreasewithincreasingtime?Furthertothis,regressionmodelscanbeusedtoinvestigatewhethertheconcentrationand/ortime(Figure17)haveanimpactontheratio,orthediffe-rence.

Theboxplotoftheprecision(Figure18)agreeswiththeimpressionsgainedfromthebiasplots,thereismorevariabilityinthepartector,inthiscase,andsolowerprecision.

4.2 .2 .3 . Distance

Distancebetweentwotimeseriescanbeusedasanothermeasureofhowsimilartheyare.SimplermeasuressuchasEuclideandistance,Manhattandistanceanddynamictimewarpingareoftenusedbuttheseignorethetemporalaspectandserialcorrelationsofthedata.

4.2 .2 . Further analysisThenextstageisthentoconsidersomenume-ricalevaluationofwhetherthetwosetsofmea-surementsdifferandhow.Generally,thiswouldinvolveinvestigatingpaireddifferences,butinthecaseoftime-seriesdatathereisanaddedcomplicationasthedatawithinatime-seriesisnotindependent(i.e.thevalueobtainedattimeTidependsonthevalueattheprevioustimepointTi-l).WithinnanoIndExwehaveconside-redanumberofmethodstoevaluatethediffe-rencebetweentwotimeseries,andattemptingtotakeaccountofthedependenceofthedata.Initially,however,weconsideredsomemorestandardmeasuresforthecomparisons.

4.2 .2 .1. Bias

Biasisanimportantmeasureforhowcloseaninstrumentistothetruevalue.Oftenthe‘true’valueisnotknown;inthiscase,youcaneithertakethereferencemonitorasthe‘true’valueortheaveragevalueofalloftheinstrumentsbeingcompared.Oftenthelatteristakenasthemedianratherthanthemeantoavoidthevaluebeinginfluencedbyanyoutliers.

Essentiallythebiasistheratioofthedifferen-cebetweenthevalueobtainedfromtheinstru-ment,I,andthetruevalue,T.

Thebiascanthenbemultipliedby100andrepresentedasapercentage;biasof0.2=20%bias.

F I G U R E 1 6 : Boxplot of bias of dif ferent instruments compared to NSAM.

4. Data collection, analysis and storage


34

considerationsofnon-stationarityofthetimeseriesresultinARIMAmodels.ARIMAmodels,canbefittedusingavarietyofsta-tisticalsoftware,inRtheforecastpackageisonepackagethatallowsforthefittingofARIMAmodels.ThesecanbefittedusingthearimafunctionandexperimentingwiththechoicesforAR,IandMA.Theoptimalvaluesofthesefunctionsarechosenbyexa-miningplotsofauto-correlationandpartialauto-correlationoftheresidualsaswellascomparingbetweenmodelsusingtheAIC.WithinthispackagethereisalsoanoptiontofitthebestfitARIMAmodeltothedata,usingauto.arima.Theresultingmodelal-lowsfortheaverageconcentrationoverthetimeperiodtobecalculated,whiletakingaccountofthecorrelationsbetweenpointsinthetimeseries.Usingthefunctionsavai-lableinR,andotherpackages,itispossibletocomparedifferentperiodswithinasingletimeseries(i.e.comparingbetweenbackg-roundandtaskswherethemeasurementsaretakensequentially).

4.2 .2 .4 . Auto-Regressive IntegratedMoving Average (ARIMA) models

ThefinalmethodutilisedinnanoIndExisARIMAmodels.Thesemodelsaredesignedtospecificallyanalysetimeseriesdata.Themodelissplitintothreemaincomponents:Au-to-Regressive(AR),whichtakescurrentvaluestobealinearcombinationofpreviousvalues,pluswhitenoise;MovingAverage(MA),whichconsiderlinearcombinationsofthewhitenoiseinputs.ThesecanbecombinedinanARMAmodel,whichwhenextendedtoalsoinclude

Eachofthemethodsusesaslightlydifferentequationtocalculatethedistance,butthedistanceisessentiallycalculatedasthedifferencebetweenthetwotimeseriesateachtimepointandsum-medoveralltimepoints.Thegreaterthisvaluethefurtherapartthetwotimeseriesareoverthetimeperiodsothiscanbeusedtogiveameasureofhowsimilarthetwotimeseriesare.

Thereareotherdistancemeasures,whichaccountforthetemporalaspectandarethereforemoreappropriatetocomparetimeseries.Thefirstordertemporalcorrelationcoefficientisoneofthese,givingaresultbetween-1and1,where1indicatesthatbothtimeseriesbehaveinaverysimilarwayinboththeirdirectionandrate,-1indicatesthattherateofchangeissimilarbutthattheymoveinoppositedirectionsand0indicatesthatthereisnosimilarityinthebehaviourbetweenthetwotimeseries.TherearepackagesavailableinR,includingTSclust,whichcanbeusedtodetermi-nethevariousdistancemeasuresavailable.

F I G U R E 1 8 : Boxplot of precision of dif ferent instruments.

F I G U R E 1 7 : Precision and Bias by concentration and time.

35

Chapter 5

Exemplary datafrom fieldmeasurements


36

5.1. ExperimentalFieldmeasurementsconductedinthescopeoftheprojectnanoIndExusedstationaryandper-sonalinstrumentsandhaveunderlinedtheimportanceofpersonalinstrumentationforindividualexposuremeasurement.Personalmonitorsusedincludedthepartector,partectorTEM,DiSCmini,nanoTracerandinasinglecasethePUFPC100.PGPandNANOBADGEwereusedascommercialpersonalsamplersinadditiontoafewprototypes.SamplesforelectronmicroscopicanalysesweretakenwithPartectorTEMandESPnano.Stationarymeasurementequipmentcomprisedthestateoftheartaerosolinstrumentation,includingscanningmobilityparticlesizers(SMPS),fastmobilityparticlesizers(FMPS),condensationparticlecounters(CPC),opticalparticlecounters(OPC),ae-rodynamicparticlesizers(APS),andelectrostaticprecipitatorsamplersforcollectingparticlesforsubsequentanalyses.Thesuiteofinstrumentsforameasurementcampaignwaschosenbasedonthematerialsandactivitiesintheworkplace.

Duringtheperformedmeasurements,theresultsofpersonalinstrumentswerecomparedtothoseofstationaryinstruments,whichwereplacedinthenearandfarfieldoftheemissionsourceorinthebackground.Movementsofworkerscarryingpersonalinstrumentationthroughtheroomwereprotocoledtotestforcorrelationsbetweenpersonalandstationaryinstrumentsatselectedsitua-tionsandlocations.

5.2 . Field study during preparation of pastesFieldmeasurementswereperformedinanindustrialpilotplantduringpreparationofpolymer-ba-sedconductiveimpregnationpastesfromnanostructuredexfoliatedgraphites.Theworkerwaswea-ringfullpersonalprotectionandmovingfreelybetweenthemixerstationandthenearbypowderstore.Theworkwasinterruptedfromtimetotimeforremovalofpowderspillsbyvacuumcleaning.Nolocalexhaustventilationwasinstalledatthemixer.AcomparisonofpersonalandstationarymonitorresponsesinthenearfieldatthemixerandthefarfieldinthepilotplanthallisshowninFigure19.Whilethefarfieldparticlenumberconcentrationwashardlyincreasedduringthemixingtask,theindividualdosefrequentlyexceededpeakconcentrationsof100000cm-3andwas5timeshigherthaninthefarfieldonaverage.Thenearfieldmonitorwasplaceddirectlynexttothemixer.However,sincetheroomventilationwasunintentionallydirectedfromthenearfieldmonitorinlettowardstheemissionsource,thenearfielddevicedetectedonlyfewpeaksandthreetimesloweraverageconcentrations.

F I G U R E 1 9 : Comparison of worker, near and far f ield particle number concentrations during preparation of a MNM-containing paste involving dry powder handling and mixing. The numbers given in angle brackets in the legend are mean values obtained by averaging the data over the time span indicated by the grey frame.

37

5. Exemplary data from field measurements

ThisshowsthatonlypersonalinstrumentationinthebreathingzoneiscapableofmeasuringMNMexposure,whereasstationaryinstrumentscanonlymeasureemission.Thisway,personalmonitorswereshowntobecapableofidenti-fyingunknownemissionsourcesandpeakemis-sionsrelatedtospecificworktasks.Differenttostationarydevicestheyarenotaffectedbyunk-nownroomventilationconditions.Roomairflownotdirectedfromemissionsourcetoastatio-naryinstrumentdilutepeakemissionsintolar-gerroomvolumesbeforebeingdetectedbythestationarymonitor.Thismakesanidentificationofcriticalworktasksthatcausestrongpeakemissionscomplicatedorimpossible.Fromanoccupationalhygienepointofviewitisdesirab-letoreducetheemissionofcriticalworktasks,totestandoptimisetheeffectivenessoflocalventilationmeasuresandtoimplementmoreadequatepreventiveandprotectivemeasures.

5.3. Field study duringproduction of TiO2

nanoparticlesAnotherfieldstudywasconductedinapilotplantfortheproductionofirondopedTiO2nanoparticles.Afastmobilityparticlesizer(FMPS)wasusedtomonitorthefarfield.Thelungdepositedsurfaceareaconcentrationinthefarfieldwascalculatedfromthemeasuredsizedistributionsbyassumingparticlestobespherical.Severalworkersinthefacilitywereequippedwithpersonalmonitorsandperso-nalsamplers.Figure20showsanexampleforatimeseriesplotoftheLDSAconcentrationmeasuredinthefarfieldwiththeFMPSandthepersonalexposureofaworkerintermsofLDSAconcentration,measuredwithapar-tector.Thetimeseriescanbesplitintothreephases.Initially,theworkerconductedregularworkinsidethefacility,wherenolocalemis-sionsourcewasdetected.Later,duringthelunchbreak,thepartectorwasplacedneartheFMPStorechargeandforasidebysidecomparisonoftheinstruments.Finallyafterthelunchbreak,theworkercontinuedwithhisactivities,butastronglocalemissionofNaClparticleswasdeliberatelyinitiatedinsidethe

productionfacility.Figure20clearlyshowsthatduringperiodswithnolocalemission,theconcentrationsmeasuredinthefarfieldandonthepersonagreedratherwell.Thesameistrueduringthesidebysidecomparison,whe-reaswiththestronglocalsource,thepersonalexposureconcentrationwasmuchhigherthanthefarfieldconcentration.

AscatterplotofthesamedataasinFigure20isgiveninFigure21.Theplotalsorepresentstherathergoodagreementofthedatamea-suredsidebysideandwithnolocalsource.Thenegativespikesinthescatterplotduringthemeasurementswithoutlocalsource(blue)stemfromshortperiods,wheretheworkeren-teredaroomwithfilteredairsupply.ThesearealsoclearlyvisibleinFigure20at11:27,12:25and12:30.

F I G U R E 2 0 : Time series plot of the lung deposited surface area concentration measured in the far f ield with FMPS and personal exposure measured with a partector.

F I G U R E 2 1 : Scatter plot of the LDSA concentration measured in the far f ield with an FMPS and personal ex-posure concentration measured with a partector; dif ferent colours refer to the three phases, specified in Figure 20.


38

ofthehighlyagglomeratednanoparticlesofTiO2emittedduringthecleaningstepatthe“source”.

ThecalculatedmassconcentrationofTiO2measuredwiththetwoNANOBADGErespecti-velyforleftandrightchestwas1.31and1.57μg/m³averagedoverthewholeshift.SEMimagesconfirmthepresenceofnano-structuredparticlesofTiO2onthefiltersasshowninFigure24.ThosevaluesremainsmuchlowerthantheRELoftheNIOSH[27](300μg/m3).Ironoriginatingcertainlyfrommachiningandweldingactivitiesinthesur-roundingareahasalsobeendetectedandconfirmedbySEMimages(Figure23).

5.4 . Field study in alaboratory for thesynthesis of nanowiresAfieldstudywasconductedinananotech-nologyresearchfacility,whereamongothersnanowiresareproduced.Thelaboratoryhasafilteredairsupply.Backgroundparticlecon-centrationsintheroomwerethereforequitelow.Thisstudywasconductedaccordingtotier2ofatieredapproach,i.e.solelywithpor-tableandpersonalmeasurementequipment.TheinstrumentationusedincludedDiSCmini,

Figure22showsanexampleofXRFanalysisonthefourmainelementsmeasuredontheNANOBADGEfilters(Ti,Fe,CaandCl).Twosamplerswerewornbyaworker(positionedonleftandrightchestinthebreathingzone)overthewholeshiftwhiletwootherswererespec-tivelylocatedatthesource(nearfield)andinthemainhallforbackgroundmeasurement(farfield).

Duringthatday,theTiO2pyrolysisreactorwasemptiedandcleaned.IthasbeenshownbasedonXRFanalysisthatthequantityoftitanumonthe“source”filterwas7timeshigheronthatdaythanonthedaybeforewhenthereactorwaskeptclosed.Figure25showsSEMimages

F I G U R E 2 3 : SEM images from the NANOBADGE filers showing Fe-basedspherical particles.

F I G U R E 2 4 : SEM images fromthe NANOBADGE filers showingnanostructured TiO2 particles.

F I G U R E 2 2 : XRF analysis of the NANOBADGE filters from one day of measurement.

F I G U R E 2 5 : SEM images from the NANOBADGE filers (source) showing highly agglomerated TiO2 nanoparticles.

39

partector,partectorTEM,PUFPC100,ESPnanoaswellasprototypesofpersonalsamplers.Duringthismeasurementcampaign,severalmonitorswereplacedindifferentfarfieldlocationstokeeptrackofpotentialspatialcon-centrationdifferences.AtleastoneofthelocalstaffandoneofthenanoIndExinvestigatorscarriedapersonalexposuremonitor.

Figure26presentsthetimeseriesofthelungdepositedsurfaceareaconcentrationduringasingletask,i.e.preparationofasubstrateforthesynthesisofnanowires.Thisincludedthetransferofawaferchunkintoaglovebox,inwhichIndiumwasmeltedonahotplatetotreatthewafer.Thepersonalexposureoftwopeopleinsidethelaboratorywasmonito-redwithoneminiDiSC/DiSCminieach.Onepartectorwaslocatedinsidethegloveboxandanotherpartectorontopoftheglovebox

suchthatitwouldaspireparticlepotentiallyreleasedthroughtheopeningsforthegloves.AnotherminiDiSCwascollocatedtothepar-tector.Thegloveboxwasmaintainedatunderpressure,areleasefromthegloveboxundernormaloperationwouldthusbequiteunlikely.OnepartectorandonepartectorTEMwereadditionallyplacedseveralmetersawayfromthegloveboxtomonitorthefarfieldconcen-tration.Figure26showsthatallinstrumentsmeasuringoutsidethegloveboxshowedverylowandconstantconcentrations,whereasthepartectorinsidethegloveboxmeasuredsignificantlyhigherconcentrationsduringtheperiodfrom11:05to11:13,whenthehotplatewasswitchedon.ThesamedataasshowninFigure26arepresentedasboxplotsinFigu-re27.Thegraphsclearlyshowthatonlytheconcentrationinsidethegloveboxincreasedtohighlevels,whereasalltheotherconcentra-tions,includingthepersonalexposureconcen-trationsremainedquitelowandcomparablewitheachother.Onlyonepartectorinthefarfieldconsistentlymeasuredhigherconcentra-tionsthantheothers,which,however,donotseemtobecorrelatedwithanyactivitiesinthelaboratory.Thereasonfortheslightlyhigherconcentrationmeasuredwiththisinstrumentremainsunclear.

Nevertheless,thetier2assessmentprovedthatthesafetymeasuresinplace,i.e.e.opera-tinginaglovebox,areeffectiveandefficientlypreventtheoperatorsfromexposuretonano-materials.

5.5. Conclusions fromfield studiesTheuseofpersonalsamplersandmonitorstoevaluateindividualoccupationalexposureofworkerstoMNMscansignificantlyimpro-vetheprocessofriskassessmentandriskmanagement.Infact,personalmonitoringandsamplinghasprovencapableofprovidingrelevantandreliabledataregardingtheindivi-dualexposureofworkers.Insomecasesthepersonalequipmenthasproventobesuperioroverstaticdevices,incaseswithhighspatialvariabilityoftheworkplaceaerosol.

F I G U R E 2 6 : Time series of lung deposited surface areaconcentration measurement in a tier 2 assessment of the exposure in a research facility during the preparation of asubstrate for the synthesis of nanowires.

F I G U R E 2 7 : Box plots of the data measured in a tier 2assessment of the exposure during preparation for the synthesis of nanowires.

5. Exemplary data from field measurements

Chapter 6

Lessons learnedduringthe project

41

ThecomparabilityofnumberconcentrationsmeasuredwiththePUFPC100withthosemeasuredwithstationaryCPCswastypically±10%orbetter.Theinstrumentalsocoversabroadersizerangethanthediffusionchargers,i.e.from4.5nmuptoseveralmicrometers.However,whentheparticlesarehighlyhydro-phobic(pureDEHS),thePUFPC100reportedbackdrasticallytoolowconcentrations.Tothecontrary,theagreementwasmuchbetter,whentheDEHScontainedonlyminorimpu-rities.Alsomeasurementswithhydrophobicsoot-likecarbonparticlesdeliveredverygoodandaccurateresults.Itcanthereforebeex-pectedthatthePUFPC100isabletomeasurethenumberconcentrationinalmostanyrealworkplacesetting,wherehighlypurehydro-phobicsubstancesareratherunlikely.

Anotherissuewenoticedrelatedtotheinstrumentsisthattheinternalclockswereratherinaccurate.Propersynchronisationoftheinstrumentclocksis,however,inevitable.Withtheavailableinstruments,atleastdailysynchronisationisrequired.Manufacturersareencouragedtousebetterclocksthatsynchro-nisethemselves,basedonaradiosignal,orautomaticallysynchronisewithacom-puterclockassoonasthedeviceisconnec-ted.

6.2 . Lesson 2:Issues related to planning and performance of field measurementsWehadtolearnthehardwaythatpositioningofthestaticinstrumentsfornearfield,farfieldand/orbackgroundmeasurementsneedstotakeintoconsiderationsmanyfactors.Anyairflowsandespeciallytheir3dimensionaldirectionneedtobedetermined,asotherwise,forexamplethenearfieldmonitorsmaynotbeaffectedbytheactivity,whereasinextre-mecases,eventhebackgroundorfarfieldcanbebiased.

Anotherimportanttopicthatneedstobedis-cussedandconcludedbeforehandiswhether

AsThorsteinVeblenpointedout,seriousre-searchwillalwaysgeneratenewopenques-tions,andagoodresearchprojectwillteachtheresearchersseverallessons.DuringthenanoIndExproject,manynewquestionscameup.Someofthemwererathereasytoanswer,whileittookusquitesomeefforttotackleothersandtolearnourlessons.Andofcoursesomequestionsstillremainopentomakeourfutureinteresting.Inthischapter,wewouldliketosharewithyoutheroadswehavetra-velledduringthenanoIndExproject,sothatyoucantakeashortcutwithoutduplicatingourdetours.

6.1. Lesson 1:Instrumental issuesBycomparingthesheersizeofapersonalmonitorwiththesizeofconventionalaerosolmeasurementequipment,itbecomesobviousthatsomethinghastobecompromised.Whilethepartectorisjustalittlelargerthanthesizeofacigarettebox,anFMPSweighs32kgandhashalfthesizeofadishwasher.Inordertobesosmall,mostpersonalmonitorsusetheindirectmeasurementprincipleofchar-gingtheincomingparticlesandmeasureacurrent,inducedbytheso-chargedparticles.Theinterpretationofthiscurrentisbasedonseveralassumptionsandonlyholdsinacertainsizerange.Weprovidedexperimen-talproofforthepriorassumptionthattheseinstrumentsareonlyableofdeterminingthenumberandLDSAconcentrationforparticlesbetween20nmand400nm.TheexpectedaccuracyandcomparabilityofLDSAconcent-rationmeasurementsisintherangeof±30%.Fornumberconcentrationmeasurementswithdiffusionchargersitislower.Whiletheaccuracyandcomparabilityofthepersonalmonitorsarehencecertainlybelowthoseofconventionalaerosolmeasurementequipment,theinstrumentsstillprovidereasonablygoodandsufficientlyaccuratedataonthepersonalexposure.

Theonlyavailablepersonal(waterbased)condensationparticlecounterPUFPC100wasonlyshortlyavailablewithinnanoIndEx.

6. Lessons learned during the project


42

WeconductedathoroughstudyontheeffectofdifferenttypesoftubingonthemeasurementandconcludedthatcurrentlyTygon®seemstobethemostrecommendabletubematerial.

6.4 . Lesson 4:Issues related topersonal samplingOnecriticalpointinpersonalparticlesamp-lingforsubsequentanalysesisthatthesamp-lerstypicallyoperateatratherlowflowrates,amongotherstokeeppumprequirementslow.Ontheotherside,theanalyticaltechniquesusedtoevaluatethesamplesneedacertainminimumamountofmaterialtobeanalysed(LOD).Especiallyincaseofsamplesforelec-tronmicroscopicanalyses,itcanbequitedifficulttodeterminethecorrectsamplingtime.Thechallengeistocollectenoughma-terialtoprovideproperstatistics,butatthesametimenottooverloadthesample.Par-tectorTEMsuggestsasamplingtimeforop-timalcoverageofthesubstrate.However,inseveralcases,wefoundthedurationofatasktobemonitoredtobemuchtooshortforthesuggestedtime,especiallyiftheparticlecon-centrationsarelow.

Besidescommercialpersonalsamplers,wealsousedafewprototypesamplersinnanoIndEx.Twoofthemuseanimpactionstagewithacutoffsizeatonly100nm.Suchanimpactorge-neratesahighpressuredrop,whichrequiresastrongpumpandreducesthebatterylifetimeofthepump.Wefoundthatwiththetestedsamp-lers,8houroperationofthepersonalpumpswasimpossibleasthebatteriesweretypicallyemptyafter4–6hoursalready.

6.5. Lesson 5:Data collection andhandlingCollectionandhandlingofpersonalexposuremeasurementswithhighlytimeresolvedmo-nitorscanbequitechallenging.Monitorswith1stimeresolutionproduce3600datapointsperhouror28,800datapointsper8hshift.Ifseveralmonitorsareused,likeintheexamples

shortandtaskbasedexposures/dosesaretobedeterminedorifshiftbasedaveragesarerequired.Ifthelatterisneeded,thehighlytimeresolveddatasetofpersonalmonitorscanbedrasticallyreducedtobehandledmoreeasily(seebelow).

6.3. Lesson 3:Issues related topersonal monitoringAnimportantquestioninpersonalexposuremeasurementsisalwayswhereinthebre-athingzonetofixthesamplinginlet.Theaspirationefficiencycanbeaffectedbymanyparameters,includingtheactivityandwhetherthepersonisleft-orrighthanded.Inadedica-tedlaboratorystudy,weexposedtwoindividu-alstoNaClaerosol,whiletheywerecarryingoutcertainactivities.Bothcarriedtwoidenti-calpersonalmonitorsbothsamplingfromthebreathingzone,onefromneartheleftandtheotherfromneartherightcollarbone.Nosigni-ficantdifferenceswerefound(seeFigure14)formeasurementscarriedoutwithpartectorwithoutsamplingtubes.Wethereforeconclu-dethattheplacementofthesamplinginlet(leftvs.right)isnotcritical.

TheotherindividualinthechamberwasequippedwithminiDiSCinstruments,whichsampledthrough75cmlongconductivesili-conetubes.Thesetubesaretypicallyconside-redastheoptimumfortransportingaerosolsthroughflexibletubes,becausetheyminimiseparticlelosses.BothpersoncarriedminiDiSCdrasticallyunderreportedtheairborneparticleconcentrations.Theresultsweretheproba-blybiggestsurpriseweexperiencedduringtheprojectandcanhaveamajorimpactonpersonalexposuremonitoringwithdiffusionchargers.Whenweinvestigatedthereasonsforthediscrepancyfurther,wenoticedthatdegassingofsiloxanesfromthesiliconetubingaffectedtheionpropertiesinthechargeroftheminiDiSC,resultingintoolowcurrentstobemeasured.Thecurrentsarethenmisinter-pretedaslowparticleconcentrations.ThiseffectwasparticularlypronouncedincaseofDiSCmini,butstillnoticeablewithpartector.

43

berandLDSAconcentrationcanonlybedeter-minedwithmonitors(exception:samplingandelectronmicroscopicanalysis,butthisisverytimeconsuming),whereasthemassconcent-rationcanonlybedeterminedwithreasonableaccuracybyusingfiltersamples.Ifexposuretoaspecificchemicalentityshallbedeter-mined,thiscancurrentlyonlybeachievedbysamplingandsubsequentchemicalanalysis.

Whilemonitorsprovidemore(timeresolved)information,thisalsomeansthatmoreinfor-mationneedstobeanalysed,whereastheanalysisofsamplersdirectlyprovidesasinglevaluewhichcanbemoreeasilyusedinworkermedicalfilesorfutureepidemiologicalstudies.

Workplacesmeasurementsgenerallypresentparticlespectraofunknowncomposition.IndividualMNMexposureassessmentatworkplacesthusoftenrequiresacombinationofmonitoringandsamplinginstruments.Anyexposurecausesaparticletype-andsize-de-pendentdose.Forabetterdistinction,notonlyparticlenumberormassconcentrationsaretobedeterminedbuttheMNMdosemustbedifferentiatedbyidentityandorigin.Thisrequiresmorphologicalandchemicalcompo-sitionanalysisofsampledparticlesbymicros-copyandX-ray(XRF,EDX)orRamanspectros-copy.Thiswayinformationcanbegeneratedthatisnecessarytodistinguishmanufacturednanoparticlefromthoseoriginatingfromna-turalorbackgroundsourceslikecombustionandroadtraffic.

inchapter5,onemaynotonlyeasilylooseoverviewofthedata.Thespreadsheetfilesgetquitelarge,especiallywhenplottingthedatawith1stimeresolution,whichmanytimescausedsoftwareorcomputercrash.Itishenceveryrecommendabletoprepareclearspreadsheetsandtoreducetheamountofdatathroughaveragingwherepossible.Ifdatareductionisnotpossible,thenumberofdiagramsperfileshouldbelimited.

Itisinevitabletotakemanynotesduringfieldmeasurements.Wealsofounditusefultohavemorethanasinglepersontakingnotes.Althoughmeanwhilewetaketheutmostcaretorecordeveryevenminorincidentduringthemeasurements,weeventuallytypicallystillfindthatsomethingismissinginourrecords.Apossibilityforeventlogginginthemonitorswouldthereforebeverywelcome.Currently,onlytheNanoTraceroffersthepossibilitytoflagthedatasetatuserdefinabletimespots.

Besidesthemeasurementdata,alotofcon-textualinformationconcerningtheworker,workplace,materials,etc.isneeded.TheamountofdatathatcanbeputintotheNECIDdatabaseseemsinfiniteandinitiallywewereabitstumped.Inordertoprovideabetteroverviewofwhatisreallyneeded,nanoIndExdevelopeddatacollectionanddatainputpro-tocolsthatmakelifealoteasiernow.

6.6. Lesson 6:Sampling or monitoring? What metric should bedetermined?Therearenoclearandsimpleanswerstothesequestions.Ofcourse,ifexposureduringshorttasksshallbedetermined,onlymonitorswithhightimeresolutioncanbeused.Samp-lersmightbeusedfor15minsamplingaslongastheanalyticaltechniquesusedtocharacte-riseandquantifythecollectedmaterialhaveasufficientlylowlimitofdetection.Ifthegoalistoproduceshiftaverages,theuseofsamplersisalsofeasible.Anothermainquestionisthemetrictobedetermined.Currently,thenum-

6. Lessons learned during the project

Chapter 7

Conclusions

45

WhenwestartednanoIndExinJune2013,notmuchwasknownaboutthepossibilities,thenovelpersonalmonitorsandsamplersof-ferandhowtheycanbeutilisedinexposureassessment.Intheprojectwelookedintothemostpressingquestions,liketheaccuracyandcomparabilityofthesamplersandmonitors,theirfieldapplicability,howfieldmeasure-mentscanbeconductedandwhatkindofdataneedstobecollected.Thosepressingques-tionsarenowanswered.Thepersonalinstru-mentsstudiedintheprojectnanoIndExwillcontributetoprogressinthefieldofnanotoxi-cologysincetheyarecapableofcharacterisingpersonalexposurelevelsintermsofdifferentdose-metrics.Infact,oneofthemainissuesinthecontextofMNM-relatedriskassessmentishazardcharacterisation.ItisprimarilybasedoninvitroandinvivotoxicologicalstudiestoinvestigateMNM-specifictoxicityandtoclarifyrelationshipsbetweenthephysicalandchemi-calpropertiesofMNMsandtheirinductionoftoxicbiologicalresponses.Unfortunately,manynanotoxicologicalstudieshaveusedexcessive,unrealisticallyhighdosesofMNMsanditis

thereforedebatablewhattheirfindingsmeanforthelowerreal-worldexposuresofhumans.Moreover,itisnotclearhowtoestablishrealisticexposuredosetestingintoxicologi-calstudies,asavailabledataonoccupationalexposurelevelsarestillsparse.Futuretoxi-cologicalstudiesshouldfocusonpotentiallyadverseeffectsoflow-levelandrealisticMNMsexposure,especiallythroughtheuseofexposuredosessimilartothoseidentifiedinenvironmentalsampling.Theuseofper-sonalinstrumentsliketheonesevaluatedinnanoIndExwillfacilitatethedeterminationofrealistic(low)exposureanddoselevelsexpressedindifferentdose-metrics.Theywillthusprovideextremelyusefulinformationtonanotoxicologistsandhelptoimplementwell-designedinvitroandinvivostudiesbasedonrealisticexposuredoses.

nanoIndExhasthuslaidthefoundationforfutureassessmentofpersonalexposuretoairbornenanomaterials,facilitatingepide-miologicalandmoremeaningfultoxicologicalstudies.

7. Conclusions


46

[1] G. Oberdörster, “Toxicology of ultrafine particles: in vivo studies” Philosophical Transactions of the Royal Society A, vol. 358, p. 2719–2740, 2000.[2] EN, EN 1540:2012-03: Workplace Exposure - Terminology, Beuth Verlag, 2012.[3] M. Fierz, C. Houle, P. Steigmeier and H. Burtscher, “Design, Calibration, and Field Performance of a Miniature Diffusion Size Classifier” Aerosol Science and Technology, vol. 45, p. 1–10, 2011.[4] J. Marra, M. Voetz and H. Kiesling, “Monitor for detecting and assessing exposure to airborne nanoparticles” Journal of Nanoparticle Research, vol. 12 , p. 21–37, 2010.[5] M. Fierz, D. Meier, P. Steigmeier and H. Burtscher, “Aerosol measurement by induced currents” Aerosol Science and Technology, vol. 48, no. 4, p. 350–357, 2014.[6] “naneos.ch” [Online]. Available: http://www.naneos.ch/partector.html. [Accessed 16 04 2016].[7] “testo-partikel.de” [Online]. Available: http://testo- partikel.de/files/DiSCmini_EN_compressed.pdf. [Accessed 16 04 2016].[8] “aerasense.com” [Online]. Available: http://www.aerasense.com/. [Accessed 16 04 2016].[9] P. Ryan, S. Son, C. Wolfe, J. Lockey, C. Brokamp and G. LeMasters, “A field application of a personal sensor for ultrafine particle exposure in children” Science of the Total Environment, vol. 508, p. 366–373, 2015.[10] “enmont.com” [Online]. Available: http://enmont.com/xe/products1. [Accessed 16 04 2016].[11] L. Soto-García, M. Andreae, T. Andreae, P. Artaxo, W. Maenhaut, T. Kirchstetter and T. Novakov, „Evaluation of the carbon content of aerosols from the burning of biomass in the Brazilian Amazon using thermal, optical and thermal-optical analysis methods“ Atmospheric Chemistry and Physics, Bd. 11, Nr. 9, p. 4425–4444, 2011.[12] I. Ogura, „Guide to measuring airborne carbon nanotubes in workplaces“ AIST-RISS and TASC, p. 1–46, 2013.[13] “aethlabs.com” [Online]. Available: https://aethlabs.com/. [Accessed 16 04 2016].[14] A. Todea, S. Beckmann, H. Kaminski and C. Asbach, “Accuracy of electrical aerosol sensors measuring lung deposited surface area concentrations” Journal of Aerosol Science, vol. 89, p. 96– 109, 2015.[15] A. Todea, S. Beckmann, H. Kaminski, C. Monz, D. Dahmann, V. Neumann, B. Simonow, N. Dziurowitz, A. Meyer-Plath, M. Fierz, H. Dozol, S. Clavaguera, G. Lidén, I. Tuinman, J. Pelzer, D. Bard, P. Thali, S. Bau, A. van der Vleuten, H. Vroomen, A. van der Vleuten and C. Asbach, „Inter- comparison of personal monitors for nanoparticles exposure at work places and in the environment“ Environmental Science & Technology (in preparation), 2016.[16] L. Cena, T. Anthony and T. Peters, “A personal nanoparticle respiratory deposition (NRD) sampler” Environmental Science & Techno- logy, vol. 45, p. 6483–6490, 2011.[17] T. Thongyen, M. Hata, A. Toriba, T. Ikeda, H. Koyama, Y. Otani and M. Furuuchi, “Development of a PM0.1 personal sampler for evaluation of personal exposure to aerosol nanoparticles” Aerosol and Air Quality Research, vol. 15, p. 180–187, 2015.[18] C.-J. Tsai, C.-N. Liu, S.-M. Hung, S.-C. Chen, S.-N. Uang, Y.-S. Cheng and Y. Zhou, “Novel active personal nanoparticle sampler for the exposure assessment of nanoparticles in workplaces” Environmental Science & Technology, vol. 46, p. 4546–4552, 2012.[19] A. Miller, G. Frey, G. King and C. Sunderman, “A handheld electrostatic precipitator for sampling airborne particles and nanoparticles” Aerosol Science and Technology, vol. 44, no. 6, p. 417–427, 2010.[20] “espnano.com,” [Online]. Available: http://www.espnano.com/. [Accessed 16 04 2016].[21] D. Leith, D. Miller-Lionberg, G. Casuccio, T. Lersch, H. Lentz, A. Marchese and J. Volckens, “Development of a transfer function for a personal thermophoretic nanoparticle sampler” Aerosol Science and Technology, vol. 48, p. 81–89, 2014.[22] B. Faure, P. Babbar, A. Guiot, S. Artous, P. Tiquet, S. Clavaguera, S. Derrough and J.-F. Damlencourt, “Identification of carbon nanotubes by thermal optical analysis” in NanoSafe 2014 Conference, 2014.[23] J. Dixkens and H. Fissan, “Development of an Electrostatic Precipitator for Off-Line Particle Analysis” Aerosol Science & Technology, vol. 30, p. 438–453, 1999.[24] S. Motellier, K. Lhaute, A. Guiot, L. Golanski, C. Geoffroy and F. Tardif, “Direct quantification of airborne nanoparticles composition by TXRF after collection on filters” Journal of Physics: Conference Series, p. 012009, 2011.[25] B. Faure, H. Dozol, C. Brouard, A. Guiot and S. Clavaguera, “Evaluation of a personal sampler for mass-based measurement of airborne nanoparticles” Journal of Aerosol Science, p. submitted, 2016.[26] European Chemical Agency, „Guidance on information requirements and chemical safety assessment Chapter R.8: Characterisation of dose [concentration]-response for human health“ ECHA-2010-G-19-EN, p. 186, 2012.[27] NIOSH, Current Intelligence Bulletin 63: Occupational Exposure to Titanium Dioxide, no. 120, p. 120, 2011.[28] W. Kelly and P. McMurry, “Measurement of particle density by inertial classification of differential mobility analyzer–generated monodisperse aerosols” Aerosol Science and Technology, 17 (3), p. 199–212, 1992.[29] M. Timko, Z. Yu, J. Kroll, J. Jayne, D. Worsnop, R. Miake-Lye, B. Onasch, D. Liscinsky, T. Kirchstetter, H. Destaillats, A. Holder, J. Smith and K. Wilson, „Sampling artifacts from conductive silicone tubing“ Aerosol Science & Technology, Nr. 43, p. 855–865, 2009.[30] Y. Yu, M. Alexander, V. Perraud, E. Brins, S. Johnson, M. Ezell and B. Finlayson-Pitts, „Contamination from electrically conductive silicone tubing during aerosol chemical analysis“ Atmospheric Environment, Nr. 43, p. 2836–2839, 2009.[31] K. Nashimoto, „Morphology and structure of silicon-oxides grown on wire electrodes by positive corona discharges“ Journal of the Electrochemical Society, Nr. 136, p. 2320–2327, 1989.[32] M. Fierz, “miniDiSC application note #6: Silicon oxide deposits on the corona wire,” [Online]. Available: http://fierz.ch/minidisc/pdf/ miniDiSCApplicationNote6.pdf. [Accessed 08 10 2015].[33] C. Asbach, H. Kaminski, Y. Lamboy, U. Schneiderwind, M. Fierz and A. Todea, „Silicone sampling tubes can cause drastic artifacts in measurements with aerosol instrumentation based on unipolar diffusion charging“ Aerosol Science & Technology, 2016 (submitted).[34] [Online]. Available: http://liaomaoqiang.sell.everychina.com/p-92487145-silicone-hose-mh-st- sih-0005.html. [Accessed 16 04 2016].[35] [Online]. Available: https://www.malvernstore.com/en-no/consumables/tubing-and- connectors. [Accessed 14 04 2016].[36] D. Pui, F. Romay-Novas and B. Liu, „Experimental study of particle deposition in bends of circular cross sections“ Aerosol Science & Technology, Nr. 7, p. 301–315, 1987.[37] M. Salmandazeh, G. Zahedi, G. Ahmadi, D. Marr and M. Glauser, “Computational modeling of effects of thermal plume adjacent to the body on the indoor air flow and particle transport” Journal of Aerosol Science, vol. 53, p. 29–39, 2012.[38] P. Nilsson, S. Marini, A. Wierzbicka, M. Karedal, E. Blomgren, J. Nielsen, G. Buonanno and A. Gudmundsson, “Characterization of Hairdresser Exposure to Airborne Particles during Hair Bleaching” Annals of Occupational Hygiene, vol. 60, p. 90–100, 2016.[39] D. Dahmann, D. Taeger, K. M. S. Büchte, P. Morfeld, T. Brüning and B. Pesch, „Assessment of exposure in epidemiological studies: the example of silica dust“ Journal of Exposure Science and Environmental Epidemiology, Bd. 1, Nr. 1, 2007.[40] NIOSH, Current Intelligence Bulletin 65: Occupational Exposure to Carbon Nanotubes and Nanofibers, pp. 1–156, 2013.[41] C. Asbach, T. Kuhlbusch, B. Stahlmecke, H. Kaminski, H. Kiesling, M. Voetz, D. Dahmann, U. Götz, N. Dziurowitz and S. Plitzko, “Measurement and monitoring strategy for assessing workplace exposure to airborne nanomaterials” in Safety of nanomaterials along their lifecycle: Release, Exposure and Human Hazard, W. Wohlleben, T. Kuhlbusch, C. Lehr and S. J., Eds., Taylor & Francis, 2014, p. 233–245.[42] [Online]. Available: http://www.baua.de/en/Topics-from-A-to-Z/Hazardous- Substances/TRGS/TRGS-402.html. [Accessed 14 04 2016].[43] M. Methner, L. Hodson and C. Geraci, “Nanoparticle Emission Assessment Technique (NEAT) for the identification and Measurement of Potential Inhalation Exposure to Engineered nanomaterials – Part A” Journal of Occupational and Environmental Hygiene, vol. 7, p. 127–132, 2009.[44] D. Brouwer, Annals of Occupational Hygiene, Nr. 56, p. 506–514, Control banding approaches for nanomaterials. 2012.[45] T. Kuhlbusch, C. Asbach, H. Fissan, D. Göhler and M. Stintz, „Nanoparticle exposure at nanotechnology workplaces: A review“ Particle and Fibre Toxicology, Bd. 8, Nr. 22, 2011.[46] [Online]. Available: nanoindex.eu. [Accessed 15 04 2016].[47] H. Kaminski, S. Rath, U. Götz, M. Sprenger, D. Wels, J. Polloczek, V. Bachmann, N. Dziurowitz, H. Kiesling, A. Schwiegelshohn, C. Monz, D. Dahmann, T. Kuhlbusch and C. Asbach, “Comparability of Mobility Particle Sizers and Diffusion Chargers” Journal of Aerosol Science, vol. 57, p. 156–178, 2013.[48] “www.perosh.eu” [Online]. Available: http://www.perosh.eu/research-projects/perosh- projects/necid/. [Accessed 18 04 2016].

47

Imprint

ThisbrochureisafinalproductoftheprojectnanoIndEx.ThisprojectissupportedbytheBritishTechnologyStrategyBoard(TSB),theFrenchNationalFundingAgencyforResearch(ANR),theGermanFederalMinistryofEducationandResearch(BMBF)andtheSwissFederalOfficeofPublicHealth(FOPH),undertheframeofSIINN,theERA-NETforaSafeImplementationofInnovativeNanoscienceandnanotechnology.Theresponsibilityforthecontentsofthispublicationlieswiththeauthors.

Lead author and project monitoring:

ChristofAsbach(ProjectCoordinator), InstituteofEnergyandEnvironmentalTechnologye.V.(IUTA),Germany

With contributions from:

AsmusMeyer-Plath, FederalInstituteforOccupationalSafetyandHealth(BAuA),GermanySimonClavaguera, FrenchAlternativeEnergiesandAtomicEnergyCommission(CEA),FranceMartinFierz, UniversityofAppliedSciencesandArtsNorthwesternSwitzerland(FHNW),SwitzerlandDirkDahmann, InstitutefortheResearchonHazardousSubstances(IGF),GermanyLauraMacCalmanandCarlaAlexander, InstituteofOccupationalMedicine(IOM),GreatBritainAnaMariaTodea, InstituteofEnergyandEnvironmentalTechnologye.V.(IUTA),GermanyIvoIavicoli, CatholicUniversityoftheSacredHeart(UCSC),Italy

Project support: MaikeMüller,UniversityofDuisburg-Essen

Design: Dipl.-Des.KatrinSchmuck

Project Coordinator: ChristofAsbach InstituteofEnergyandEnvironmentalTechnologye.V.(IUTA),Germany Phone:+492065418-409 Email:[email protected] www.nanoindex.eu

All rights reserved includingphotomechanicalreproductionandthereprintingofextracts.First published: May2016

Copyright: ©2016bytheauthors

Imprint

assessment of personal exposure to airborne re.pdf · pdf filetube material. in another...

Documents