IAEA利用LGR水同位素分析仪测量高丰度富集D样品

来源：纷纭教育

ResearchArticle

Received:16October2015

Revised:10November2015

Accepted:12November2015

PublishedonlineinWileyOnlineLibrary

RapidCommun.MassSpectrom.2016,30,415–422(wileyonlinelibrary.com)DOI:10.1002/rcm.7459

Measurementofextremely2H-enrichedwatersamplesbylaserspectrometry:applicationtobatchelectrolyticconcentrationofenvironmentaltritiumsamples

L.I.Wassenaar*,B.Kumar,C.Douence,D.L.BelachewandP.K.Aggarwal

InternationalAtomicEnergyAgency,ViennaInternationalCenter,A-1400,Vienna,Austria

RATIONALE:Naturalwatersamplesartiﬁciallyorexperimentallyenrichedindeuterium(2H)atconcentrationsupto

10,000ppmarerequiredforvariousmedical,environmentalandhydrologicaltracerapplications,butaredifﬁculttomeasureusingconventionalstableisotoperatiomassspectrometry.

METHODS:2Herewedemonstratethatoff-axisintegratedcavityoutput(OA-ICOS)laserspectrometry,alongwithH-enrichedlaboratorycalibrationstandardsandappropriateanalysistemplates,allowsforlow-cost,fast,andaccuratedeterminationsofwatersampleshavingδ2HVSMOW-SLAPvaluesuptoatleast57,000‰(~9000ppm)ataprocessingrateof60samplesperday.

RESULTS:Asonepracticalapplication,extremely2H-enrichedsamplesweremeasuredbylaserspectrometryandcomparedtothetraditional3HSpike-Proxymethodinordertodeterminetritiumenrichmentfactorsinthebatchelectrolysisofenvironmentalwaters.Highly2H-enrichedsamplesweretakenfromdifferentsetsofelectrolyticallyconcentratedstandardsandlow-level3(<10TU)IAEAinter-comparisontritiumsamples,andallcasesreturnedaccurateandpreciseinitiallow-levelHresults.

CONCLUSIONS:Theabilitytoquicklyandaccuratelymeasureextremely2H-enrichedwatersbylaserspectrometrywillfacilitatetheuseofdeuteriumasatracerinnumerousenvironmentalandotherapplications.Forlow-leveltritiumoperations,thisnewanalyticalabilityfacilitateda10–20%increaseinsampleproductivitythroughtheeliminationofspikestandardsandgravimetrics,andprovidesimmediatefeedbackonelectrolyticenrichmentcellperformance.Copyright©2016JohnWiley&Sons,Ltd.

Naturalwaters(1H2HO)artiﬁciallyenrichedindeuterium(2H)intopureH2gas.[8–11]However,mostIRMSlaboratoriesarewellabovenaturalabundancemassfractionsareusedasreluctanttomeasurewatersextremelyenrichedin2HonIRMSpowerfultracersinmedical,environmental,2andhydrologicalinstruments(minorcollectornotoptimized),oronsampleapplications.AwidespreadapplicationofH-enrichedwaterpreparationapparatusroutinelyusedfornaturalabundanceisindoublylabelledwater(DLW)[1,2]usedforenergyexpenditurewaters(contamination).SomeIRMSsamplepreparationtestingofhumansandanimals.Otherapplicationsincludedevicessufferfromconsiderablebetween-samplecarryover,experimentsoftissueturnover,[3,4]usingdeuteriumasanespeciallywhenmeasuring2H-enrichedwatersamplesthatartiﬁcialtracerinﬁeld[5,6]andlaboratoryhydrogeologicaloraffectdozensofsubsequentsampleswithouttheapplicationdiffusionexperiments,andforbatchelectrolyticenrichmentofcarryovercorrectionmodels.[12]IRMSmaysufferfromlargeoftritiuminenvironmentalwatersamples.[7,8]Thehighδscaleexpansionatenriched2Hconcentrationswhenusingconcentrationsof2HcomparedtonaturalwatersrangefromH2gas.Moreover,moststableisotopelaboratoriesdonotslightlyabovenaturalabundancemassfractions(~150ppm)haveappropriatestandardswithδ2HVSMOW-SLAPvaluesoftoextremevaluespotentiallysurpassing10,000ppmdeuterium1000‰ormore.[11]Asaresult,moststableisotopelaboratories(e.g.δ2HVSMOW-SLAPupto60,000‰).

areHydrogenisotope(δ2H)assaysofliquidwatersamplesare2unwillingtoaccept,orcannotmeasure,extremelyH-enrichedwatersamplesforanyoftheaforementionedtraditionallycarriedoutusingdual-inletorcontinuous-ﬂowapplications.

isotope-ratiomassspectrometry(IRMS)usingeitherH2O(water)/H2(gas)equilibration,orZn/Cr/Chigh-temperature2In2001,theﬁrstlaser-basedmeasurementsofhighlyH-enrichedwatershadaδ2HVSMOW-SLAPlimitofaboutchemicalreactor(HTC)reductionmethods,viaconversion

15,000‰,withdemonstrablyimprovedreductioninbetween-samplememorycomparedtocontemporaryIRMSmethods,withsamplemeasurementtimesontheorderof*Correspondence40min.[13]Since2009,low-cost,commercialwaterisotopeEnergyAgency,to:ViennaL.I.Wassenaar,InternationalInternationalCenter,AtomicA-1400laserspectrometershaveovertakenVienna,18IRMSastheprimarymeanstomeasureδ2H(andδO)innaturalwaters.RequiringE-mail:l.wassenaar@iaea.org

Austria.

littlewater(<1000nL)andfewconsumables,andwith

415L.I.Wassenaaretal.

minimaltraining,laser-basedwaterisotopemeasurementsareatastageofwidespreadadoptionandaffordability.[14,15]Theﬁrsttestsof2H-enrichedDLWwaterbycommercialcavityringdown(CRDS)laserspectrometryshowedsuccesswithsampleshavingδ2HVSMOW-SLAPvaluesupto~750‰,butrequiredmorethan20sampleinjectionstoovercomesigniﬁcantbetween-samplememory;hence,only15samplesperdaycouldbemeasured.[16]However,withrecentdevelopmentsinlaserspectrometry,thepotentialformodernliquidwaterisotopelaserinstrumentationhasnotbeenadequatelyexploredforextremely2H-enrichedwatersamples,whichmaybeusefulfortheaforementionedapplications,orintracerorexperimentalstudies.

Theobjectiveofthispaperistwofold:(i)todemonstratethatcommercialoff-axisintegratedcavityoutputlaserspectrometry(OA-ICOS)canbeusedtorapidlyobtainaccurateandpreciseδ2HVSMOW-SLAPvaluesforbothnaturalabundanceandwatersamplesextremelyenrichedindeuteriumupto57,000‰,and(ii)subsequentlytodemonstratetheefﬁcacyofusingthisanalyticalcapabilityforlaser-based2H-enrichmentmethodstoimprovetheproductivityofenvironmentaltritiumlaboratoriesengagedinbatchmodeelectrolyticenrichmentof3H.

(e.g.gramsH2O)ofthespikesamplebeforeandafterelectrolysis.InitialtritiumconcentrationsofunknownsamplesprocessedthroughtheTEUaredeterminedbyrearrangement:

Ti¼Tf=ððVi=VfÞÁRÞ

(2)

Thetritiumrecoveryfactor(R),unfortunately,canonlybedeterminedonspikecells.HencespikerecoveriesandderivedenrichmentparametersareaveragedandappliedequallytoallTEUcellscontainingunknownsamples.QuantitativewaterrecoveriesandaccurateweighingarecriticalintheSpike-Proxymethod.Disadvantagesareasigniﬁcantreductioninsamplethroughputbecauseofthespikerequirement.

AnalternativetotheSpike-Proxymethodisthe2H-enrichmentmethod,[8]whichleveragesthefactthat2H(HDO)iscorrespondinglyconcentratedinaTEUelectrolysisprocess,albeittoalesserextentthantritium(HTO)duetodifferentnetisotopefractionationfactorsandvaporlosses.Nevertheless,tritium(ifpresent)anddeuteriumareverystronglycorrelatedduringelectrolyticenrichment.[8,9]Becausetheelectrolyticallyenriched2Hsamplecanbemeasuredasanindependentvariable,itprovidesameansfordeterminingthe3H-enrichmentfactorsforeachcell.Thisledtotheconceptofacellconstant(k)thatcorrelatesthe3H-and2H-enrichmentfactorstoeachother:[19]k¼lnðTf=TiÞ=lnðDf=DiÞ

(3)

HMETHODFORDETERMINING

ELECTROLYTICTRITIUMENRICHMENT

Tritiumisapopularradiotracerofshort-termhydrologicandgroundwaterresidencetimes,[17]butexceedinglylowconcentrationsinenvironmentalwatersnowadaysaretoolowfordirectdecaycounting.Thuswatersamplestypicallyrequirepre-concentrationof3Hbyelectrolyticenrichment.[7,18]Methodsforpre-concentratingtritiumusing250–1000mLwatersamplesemploysetsofmild-steelalkalineelectrolysiscells,orpolymerelectrolyticmembraneunits.[9,19,20]Alltritiumenrichmentunits(TEUs)havethreecommonalities:(i)pre-distillationofsamplestoremovedissolvedions,(ii)electrolytic3Henrichmentofthedistilledsamplesto8–60mLﬁnalvolume,and(iii)decaycountingbyliquidscintillation(LSC)orgasproportionalcounting(GPC)instruments.Dependingonthe3Hconcentration,startingandﬁnalsamplevolumes,electrolyticcell-type,andoperationalconditions,watersamplesmaybeenrichedin3Hbyfactorsof10–90timesormore[7]asneededforLSCorGPC.Notably,2Hiscorrespondinglyconcentratedduringelectrolysis.

AcriticalrequirementforanyTEUisaccuratedeterminationofthetritiumenrichmentfactorforeachcellinordertocorrecteachsamplebacktoitsoriginal3Hconcentrationintheenvironment(withappropriatedecaycorrection).MostlaboratoriesusetheSpike-Proxymethod,whereby10–20%ofsamplesprocessedthroughtheTEUarehigh-concentrationtritiumstandards(spikes).Thetritiumrecovery(R)isdeterminedfrompre-andpost-electrolysisgravimetricH2Orecoveries(Vﬁnal/Vinitial)andmeasurementofspike3Hconcentrationsbeforeandafterelectrolyticenrichment:

R¼ðTfÁVfÞ=ðTiÁViÞ

(1)

2whereDistheﬁnal(f)andinitial(i)sample2Hconcentrationinppm,andTisasabove.Rearrangementallowsdeterminationoftheinitialunknowntritiumconcentration(Ti)ofasamplebyknowingk(foreach,orbyaveragingidenticalcells),measuring3Hintheelectrolyticallyenrichedsample(Tf),andmeasuringtheinitialandenriched2Hconcentrations:

Ti¼Tf=ðDf=DiÞk

(4)

Akeyrequirementofthe2Hmethodfordeterminingtritiumenrichmentfactorsisaccuratedeterminationofthecellconstant(k)forallTEUcells,bytheabilitytomeasureextremely2H-enrichedliquidwatersamples.Thecellconstantisdeterminedempiricallybycoupled2Hand3Hspiketesting,alongwithcarefulgravimetricrecoveries.Thecellconstantcanbedeterminedforindividualcells,oraveragedifidenticalbehaviorcanbedemonstratedforeachTEUset.[19]Anaddedbeneﬁtofthe2Happroachiseliminationoflaboriousgravimetricweighingsteps(seeEqn.(4)).Whileeliminationofgravimetricweighingseemsattractive,bymaintainingitthedeuteriumrecoveryfactor(Rd)foreachcellcanalsobedetermined:

Rd¼ðDfÁVfÞ=ðDiÁViÞ

(5)

whereTisthetritiumconcentrationofthespike(netcountsperminute(CPM)afterbackgroundsubtraction)before(i)andafter(f)electrolysis,andVisthewatervolume

The2Hrecoveryfactorfacilitatesimmediatedetectionofelectrolyticcellperformancedegradation,bynothavingtowaitforweeksforspikeLSCcountingcompletion.Todate,the2Hmethodisemployedbyfewlaboratories,andinallcaseshighly2H-enrichedsamplesareusuallydilutedbyafactorof1000ormoretothenaturalabundancerange[7]andmeasuredbytraditionalIRMSatconsiderableextracost.Therequisitelargedilutionsandisotopemassbalancebudgetingcontributesigniﬁcanterrortotheoverallprocess.

Measurementofextremely2H-enrichedwatersbylaserspectrometry

EXPERIMENTAL

resultsofelectrolyticallyenrichedsamplesandspikesas2measuredbylaserspectrometry.EachlaboratorystandardH-enrichedcalibrationstandards

wasmeasuredseveralhundredtimesagainstIAEA605andInordertomeasureextremely2H-enrichedwatersamples,IAEA606.Notably,IHLW-68waspredictedtofallappropriateenrichedcalibrationstandardsareneededthatconsiderablyoutsidethecalibrationrangeoftheIAEA605spantheδrangeofthesamples.[11,12]Untilrecently,noandIAEA606referencematerials.

primaryreferencewatershighlyenrichedindeuteriumwereavailable.In2015,theInternationalAtomicEnergyAgencyTritiumenrichmentunittestwatersamples

(IAEA,Vienna,Austria)announcednew2H-(and18O-)Allthe3Hwatersamplesfor2H-enrichmenttestingwereenrichedprimaryHDOstandards.[21]Forthepurposesofspikestandardstritiumenrichment,twoofthesestandards(IAEA605,3thatweremeasuredforvolumetricmassbalancesandHand2Hconcentrations,using250mLandIAEA606)wereusefulforthedevelopmentof2H-enriched500mLpre-andpost-electrolyticTEUsystems.Asecondsetlaboratorycalibrationstandards.Theδ2HVSMOW-SLAPvaluesoftestsamplesconsistedof500mLlow-level3HsamplesoftheIAEAprimaryreferencewatersaresummarizedinfromthe2012IAEATRICinter-comparison.[23]TheTRICTable1.Owingtothesmallamountofreferencematerialssamplesallowedustodeterminehowwelllaser-based2Hprovided(20mL),weprepared20Lofthree2H-enrichedmethodsresultedintritiumenrichmentfactorsthatreturnedlaboratorystandardwatersforuseindailynormalization;accurateandpreciseoriginallow-level3Hresults.

theselaboratorystandardswerecalibratedusingassignedBatchelectrolyticenrichmentswereconductedusingmildvaluesforIAEA605andIAEA606.

steel24-cellTEUsystems(250mL,500mL)attheIAEAThreenewlaboratorystandardswerepreparedgravimet-IsotopeHydrologyLaboratory(IHL)whichhavebeeninrically[22]usingdistilledtapwater(δ2HVSMOW-SLAP=–77‰),useforover20years.ThealkalineTEUsystemanditsandbyadding32–200gof99.9993%deuteriumoxideoperationarefullydescribedintheIAEAtritiumstandard(Sigma-Aldrich,St.Louis,MO,USA)inordertospantheoperatingprocedure(SOP)availablefromtheIsotopeδ2Hrangefor250–2000mLelectrolyticallyenrichedsamplesHydrologyLaboratory.[24]Brieﬂy,theIHLusestheSpike-obtainedintypicaltritiumoperations(estimatedtobeintheProxymethodwhere3ofthe24(13%)samplesineach6000–60,000‰rangeforδ2HVSMOW-SLAP).ThelaboratoryTEUprocessingbatcharespikes.Thesespiketripletsarestandardswerehomogenizedfor1weekpriortoisotopicadvancedbyonepositionineachnewelectrolyticrun.Thus,analysisandstoredat0.5barargongaspressureinsteelspikesarecycledthroughtheTEUevery8runs,andareusedsiphon-dispensingcontainers.Allδ2HVSMOW-SLAPcalibrationstodetermineenrichmentparametersforalltheremainingwereconductedbyoff-axisintegratedcavityoutputlasercells.Forthistest,0.5mLpre-andenrichedspikesamplesspectrometry,asdescribedbelow.

weresampledopportunisticallyfor2HfromroutineIHLInordertoobtainassignedδ2HVSMOW-SLAPvaluesforthesampleprocessingoperations.Onediscrete500mL,laboratorystandards,weconductedtwo-pointdata24-sample,TEUanalysissetcontainedtriplicatesamplesofnormalizationusingIAEA605andIAEA606ascalibrationlow-levelTRICtestwatersamples.Deuteriumsamplingstandards,withthelaboratory2standardsmeasuredasconsistedoftakinga0.5mLsub-sampleofthepre-electrolysisunknowns.Aftercalibration,δHVSMOW-SLAPvalueswerespike(naturalabundance,afterpre-distillation)anda0.5mLassignedtoallthree2H-enrichedlaboratorystandards(IHLsub-samplefromthe10–12mLﬁnalelectrolyticallyenrichedW-62,IHLW-63andIHLW-68)andusedtonormalizethe

post-distilledsample(extremely2H-enriched).Allthe

Table1.2HVSMOW‐SLAPvaluesofprimaryisotopicallyenrichedreferencewaters,gravimetricestimatesandlaserspectrometricassayofthreehighly2H‐enrichedIHLlaboratorywaterstandardsusedinroutinetritiumprocessingIAEAPrimaryStandardsδ2HVSMOW‐SLAP(‰)

2H(ppm)

VSMOW20142.8SLAP2IAEA6055,997.9–427.51088.8.2IAEA606

15,993.62639.9EnrichedLaboratoryStandardsδ22(‰H,VSMOWgravimetric)‐SLAPδ2(‰HVSMOW,measured)‐SLAPNH(ppm)

IHLW‐629249–909173.7±34901582.1IHLIHLWW‐‐6368

17,91958,211––18,28257,057

18,0.2±656,842.8±10

320116

2960.629.1

aGravimetricestimateswerebasedon20L(kg)ofdistilledtapwater(δ2HVSMOW‐SLAP=–77‰)andadding32.00,.00and200.00gof99.9993deuteriumoxide,assuming1%weighingerror.Gravimetricestimatesweremadebymixingandisotopelaboratory.

N=numbermassofbalancemeasurementsequations.[22]bmadebylaserspectroscopy.Boldfaceismean±SEM,withassignedδvaluesusedintheIHLRapidCommun.MassSpectrom.2016,30,415–422Copyright©2016JohnWiley&Sons,Ltd.wileyonlinelibrary.com/journal/rcm

417L.I.Wassenaaretal.

distilledpre-andpost-electrolysissamplesfor2Hanalysiswerestoredintightlysealed1.5mLTeﬂonseptumcappedlaserspectrometryvialsuntilanalysis.Laseranalyseswereconductedwithin1dayto1weekaftercollection.Tritiummeasurements

Unenrichedandenrichedspikesamplescorrespondingtothesame2Hsub-samplesweremeasuredbyliquidscintillationcounting(LSC)onaQuantulus1220™(PerkinElmer,Akron,OH,USA).Thetritium-countingprocedureisfullydescribedintheIAEASOP.[24]Brieﬂy,8mLofpre-andpost-enrichedspikesweremixedwith12mLofUltimaGolduLLT™(PerkinElmer)scintillationcocktailandallowedtostabilizeinthedark.TheLSCtemplateconsistedof24unknownenrichedsamplesincluding3enrichedspikes,aswellas2unenrichedspikes,2calibrationstandards,and4backgroundsamples.Thetotalcountingtimewas500min/sampleconductedover10countingcycles(approx.12days).OptimaloperationoftheQuantulusLSChasbeendescribedpreviously.[25]Allspectralregionsofinterestwereoptimizedforlow-level3Hbetacountinginwatertoobtainthehighestﬁgureofmerit(FOM).NetCPMvaluesforpre-andpost-enrichedspikeswereobtainedbysubtractionofthemeanbackgroundCPM(~0.5to0.8CPM)onaperrunbasis.Forthisstudy,weextractednetCPMdataofpre-andpost-spikedataobtainedfrommultipleQuantulusLSCrunscoveringtheperiodfromOctober2014toJuly2015,andwheredeuteriummeasurementswereconducted.AlldataprocessingwasdoneusinganExcel-basedTritiumInformationManagementSystem(TRIMS)developedattheIHL.ErrorpropagationwasdeterminedforalltheprocessingstepsasoutlinedintheIAEASOP,andthiswasincludedintheﬁnaluncertaintyreportingofthe3Hresults.Laserspectrometryfor2H-enrichedsamples

Severaloff-axisintegratedcavityoutputwaterisotopelaserinstrumentsfromLosGatosResearch(MountainView,CAUSA)weretestedfordirectlymeasuringhighly2H-enrichedwaters.Wetested1st-3rdgenerationLosGatosResearchlaserinstrumentsandfoundtheydidnotcorrectlyreportδ2Hvaluesabove~12,000–15,000‰,despiteapparentlygoodisotopologueopticalspectrometry,dueinternalsoftwarelimitationsintheseolderinstruments(DougBaer,LosGatosResearch,personalcommunication).InthisstudywethereforeusedthenewestLosGatosResearchliquidwaterisotopeanalyzermodel912-0032(4thgeneration).

Forlaser-basedHDOmeasurements,allthepre-andenrichedpost-electrolysissamplesandlaboratorycalibrationstandardswereseparatedintothreegroups.Theﬁrstcomprisedpre-enrichmentsampleshavingnaturalabundance2Hlevels.Theseweremeasuredbyroutinelaserspectrometryusingnaturalabundancewaterstandards,controls,anddata-processingproceduresusingLIMSforLasers2015,asfullydescribedelsewhere.[15,26]Abatchof24TEUpre-enrichedwatersamplesandstandardsemployedananalysisprotocolof8injections,ignoringtheﬁrst4,atarateof11min/sample.

Forextremely2H-enrichedwatersamplesbeyondtheVSMOW-SLAPcalibrationscale,thesameprocedurewasusedbutwithsomemodiﬁcations.First,watersampleswereseparatedbyTEUsizeandtheexpecteddegreeof2Henrichment.Forexample,250mLcelland500mLcell

deuteriumsamplesweregroupedseparately.Forthe250mLTEUsystem,wepredictedδvaluesaround12,000‰andthereforeusedatwo-pointcalibrationusinglaboratorystandardsIHLW-62(lowstandard)andIHLW-63(highstandard);inthecaseofthe500mLTEUor1000mLandhigherexperimentalenrichmentssamples,weusedIHLW-62(lowstandard)andIHLW-68(highstandard),andIHLW-63wasusedasacontrol(Table1).

Theonlyotherminormodiﬁcationforlaserspectrometrycomparedwiththenaturalabundanceprocedureswasthenumberofinjectionspervial.Owingtothehighlyenriched2Hlevels,weempiricallydeterminedthat12injections,ignoringtheﬁrst4,wassufﬁcienttoobtainaccurateresultsandminimizebetween-samplememorycorrectionsasquantiﬁedbyLIMSforLasers2015.Thisresultedinananalysisrateof16min/sampleforeachenrichedsample.FortheroutineTEUsetsof24samples,thistranslatedtoonly504injectionspersetonthelaserinstrument,wellwithinhigh-performanceseptaspeciﬁcations(SupelcoThermogreenLB-2,6mmdisks;Supelco,Bellefonte,PA,USA).Includinglaboratorystandardsandcontrols,theprocessingratewas60unknownsamplesperday.Ifthelaserinstrumentusedforhighly2H-enrichedsampleswassubsequentlyneededfornaturalabundancewatersamples(pre-enrichedsamples,allothersamples),approximately50sequentialinjections(approximately1h)oftapwaterwassufﬁcienttoﬂushawaytheresidualmemoryofthe2H-enrichedwatersamples.Inshort,completesetsofpre-andpost-electrolysiswatersamplescouldbeanalyzedbylaserspectrometryinlessthan48h,therebyprovidingnearimmediatefeedbackonthebatchTEUelectrolyticenrichmentperformance.

Alllaser-baseddeuteriumresults(δ2HVSMOW-SLAP)wereprocessedandreportedinthe‰notationrelativetotheVSMOWstandard,andprocessedusingLIMSforLasers2015.Theδvaluesdataweretransformedintomassfractionsdeuterium(ppm)using:

ÀÁ

Dppm¼δ2HVSMOW-SLAPþ1000=ðδ2HVSMOW-SLAPþ1000

(6)þ1000=0:00015576ÞÁ1000000

RESULTSANDDISCUSSION

2H-enrichedlaboratorystandards

Gravimetricallydeterminedandlaser-measuredδ2HVSMOW-SLAPvaluesforthreeIHLlaboratorystandardsmeasured

relativetoIAEA605andIAEA606aresummarizedinTable1.InthecaseofIHLW-62andIHLW-63,thelaserspectrometry-measuredδ2Hresultsfellwithintheδrangepredictedbygravimetricdeterminations.ForsampleIHLW-68,themeasuredδ2Hvaluewasonlyslightlylowerthanpredictedbygravimetriccalculations,probablybecausethesamplewasfaroutsidetheIAEA605andIAEA606referencecalibrationrange,andduetopossibleeffectsofveryslightisotopicδ-scalecompressiononthelaserinstrument.Asafurthercheck,wemeasuredIHLW-63asanunknownsampleusingassignedvaluesofIHLW-62andIHLW-68ascalibrationstandards.Inthiscase,IHLW-63returnedaδ2HVSMOW-SLAPvalueof+18,232±33‰(n=38),closetothatreportedinTable1,andwithintherangepredictedby

Measurementofextremely2H-enrichedwatersbylaserspectrometrygravimetricestimates.Withoutprimarystandardscoveringeachtripletwasthemeanoftwomeasurements,summarizedthishigherδrange,theIHLW-68valueremainstentative,inTable2.Theindividualcellconstantswereremarkablybutforsucha2highenrichmentwasremarkablyclosetosimilarwithinandacrossbothmild-steelTEUsystems.ForthepredictedδHvalue.Forourpurposes,weacceptedthe250mLTEUsystem,themeancellconstantkwasthelaser-measuredδ2HvaluesinTable1ascorrect,and1.0542±0.0044,withanoverallrangefortheindividualcellstheseδvaluesandtheiruncertaintieswereassignedforallbetween1.0460and1.0620.Forthe500mLTEUsystemthesubsequentroutine3Helectrolysisanddata-processingmeancellconstantwas1.0592±0.0101,withtheoverallrangeoperations.Sampleδvalueconversionsintomassfractionsforindividualcellsrangingfrom1.0509to1.0670.AStudent’s(ppm)arelistedinTable1.FortheprimaryRMsandt-testrevealedthattherewasnosigniﬁcantdifferencelaboratorystandards,themassfractionsofdeuteriumbetweenthemeancellconstantsofthe250mLand500mLrangedfrom.2ppm(VSLAP2)to2639.9ppm(IAEA606),TEUsystems(p=0.058).ThemeancellconstantforthewhereastheenrichedIHLlabstandardsrangedfromcombined250and500mLTEUsystemswastherefore1582.1ppm(IHLW-62)to29.1ppm(IHLW-68).These1.0555±0.0048.Forallsubsequentcalculations,andduetoﬁndingsshowthatcommerciallaserspectrometrycanbetheabsenceoflongtrackrecordsforindividualcellconstants,usedtoquicklyandaccuratelymeasurewatersamplesweusedthemeancellconstantanditsuncertaintyasextremelyenrichedin2H,andhavingδ2Hvaluesuptoatapplicabletothe250mLand500mLIHLTEUsystems.

least57,000‰,andpossiblyevenhigher.

TodemonstratehowwellusingacellconstantkrecoveredTheenrichedlaboratorystandardIHLW-68waspreparedintheinitialmeasuredCPMforbothIHLelectrolysissystems,anticipationofelectrolyticenrichmentoflarge1–2Lsamples,Eqn.(4)wasusedtodeterminetheinitialsample3HCPMwhichwillbeincreasinglyrequiredasenvironmental3Hlevelsvalues.InTable2,acomparisonof3Hinit(e.g.un-enricheddeclinetobackgroundlevels.Indeed,therecentTRICtestspike)withtheelectrolyticallyenrichedspiketriplicatesrevealedthatonlylaboratoriesthatenrichedwatersamplesCPM,andapplicationoftheuniformcellconstant,revealed50×orhigherwereabletoachieveaccurateandpreciseexcellentperformanceindeterminingtheinitial3Hcontentresultsforlow-level(<5TU)tritiumtestsamples.[23](lastcolumn),despitethefactTEUsampleenrichmentsCurrently,fewlaboratoriesenrichsamplesgreaterthanwereconductedovera10-monthperiodwithtwodifferent500mL(e.g.δ2HVSMOW-SLAPof~25,000‰)duetoextracostTEUsystemsandusingthreedifferentLSCcounters.Notably,orthelack2ofasuitableTEUapparatus.Theabilitytodirectlynogravimetricdatawasrequiredwiththe2HmethodmeasureHatextremeconcentrationsbylaserspectrometrytodetermineprovidesanewwaytoquicklydeterminecellperformance3thetritiumenrichmentfactorsortheinitialsampleHcontent.

and3H-enrichmentfactorsbeforehand,withouthavingtoPoorlyperformingcell#4showedamarkedlydifferentwaitweeksforLSCspike-countingresults.kconstant(1.1219)fromthegoodcells.Importantly,applicationofthemeasuredcellconstantyieldedanResultsoftritiumspikestandards

inaccurate3Hrecovery(10.91CPMvsactual8.82CPM),indicatingthatthe2H-derivedcellconstantscannotbeusedTheresultsof3Hand2Hassaysonspikesprocessedonblindlytoremedybadlyperformingcells.Poorlyperforming250mLand500mLTEUsystemsbetweenOctober32014andcellsshouldberemovedfromtheTEUandreconditioned.July2015aretabulatedinTable2.The2HandHrecoveriesMonitoringdeuteriumrecoveryfactorsisthereforeanforeachelectrolyticcellweredeterminedusingEqns.(1)importantcontrolinevaluatingTEUcellperformance.Ourand(5).Theaveragedeuteriumrecoveryforthe250mLdataTEUwas74.4±1.6%,withanoverallrangeforindividual2suggestthat3HresultsobtainedfromcellswithpoorHrecoveries(e.g.<70%onIHLTEUsystems)shouldnotcellsofbetween71and77%.Theaveragedeuteriumbeaccepted.These2Hcontrolcriteriawillneedtoberecoveryonthe500mLTEUwas77.0±1.0%,witharangedevelopedforeachuniquelaboratoryTEUsystem.Ideally,forindividualcellsofbetween75and79%.Notably,cell#4overtimeitmaybepossibletoobtainrobustindividualmean(excludedfromsummaries)wasknowntobeperformingcellkconstantcriteriabymaintainingregularoroccasionalbadly,andwasintheprocessofbeingreconditionedafterspiketesting.Whiletheuseofindividualcellconstantsseemschemicaltreatment.ThisTEUcell2stillindicatedpoor2Hintuitivelysuperior,itremainstobeseenwhetherindividualrecoveryofonly54.8%.Thus,Hcanalsobeusedasancellconstants,versusindicatorformonitoringcell-enrichmentperformanceprior3averagedcellconstants,providesdemonstrablybetterHresultsafterallsourcesofuncertaintytostartingLSCcounting.Inourcase,athresholdof70%arefactoredin.

fordeuteriumrecoveryappearedtobeareasonablelowerboundtorevealcelldegradation,uponwhichcellreconditioningmaybewarranted,andsamplesfromdegradedcellsshouldEvaluatingthe2Hmethodforaccuracyandprecisionnotbeused.Similarly,themean3Hrecoveryonthe250mLTheﬁnaltestconsistedofapplyingthe2H-3H-determinedTEUsystemwas86.3±2.2%,withanoverallrangeformeancellconstantktoaroutinesetof24samplesprocessedindividualcellsof81–%.Themean3Hrecoveryforbyenrichmentonthe500mLTEUsystem(includingpoorlythe500mLTEUsystemwas92.4±2.1%,withanoverallperformingcell#4),andthencomparingtheﬁnaltritiumrangeforindividualcellsoffrom90to97%.Poorlyperforming(TU±SD)outcomeswiththoseobtainedusingthecell#4hadalow33Hrecoveryofonly81.0%.

conventionalSpike-Proxymethod.ThistestsetincludedThecell2H-Hconstant(k)valuesforthe250mLandtriplicatesoftheIAEAtritiuminter-comparisonsamples,500mLTEUs2weredeterminedusingEqn.(3)andthewhichconsistedofﬁvelow-levelsamples<8TUhavingmeasuredδHand3Hvaluesfromeachtriplicatedspikewell-established3Hvalues,[23]threespikestandards(knownandcell.TheinitialCPMoftheunenrichedspikesamplefor

TU),twounknownsamples(oneDIwaterplacedincell#4)

419L.I.Wassenaaretal.

Table2.Resultsof250mLand500mLTEUspikesamplesfordeterminationofindividualandaverage2H-3Hcellconstants(k)Electrolysis

ID250mL

140514051405140614061406140714071407140814081408140914091409141014101410141114111411500mL

241924192419242024202420242124212421242224222422242424242424242524252425CellConstant

MeanSD

aCounterIDQ3Q3Q3Q3Q3Q3Q2Q2Q2Q1Q1Q1Q1Q1Q1Q1Q1Q1Q1Q1Q1Q2Q2Q2Q1Q1Q1Q1Q1Q1Q3Q3Q3Q3Q3Q3Q2Q2Q2

Cell#1011121617181920212223241234567192021222324123456101112131415

VoVfgrams

2Hinit2Hﬁnal2H

3Hinit3Hﬁnal3HCell

ppm143.5143.4143.5143.6143.7143.7143.5143.6143.6143.7143.7143.7143.6143.6143.6143.7143.7143.6143.5143.6143.7143.6143.4143.4143.5143.5143.4143.4143.5143.6143.6143.6143.5143.4143.5143.7143.5143.4143.4

2365.32398.62363.82303.92323.32300.92178.92208.22329.72204.32247.92102.32172.52207.42076.921.12176.22186.22299.02177.32219.73754.33385.83251.03373.23676.43530.03578.43347.63639.13548.03526.03530.63454.73486.63634.13331.13292.43533.0

Recovery%

75.475.072.775.373.373.971.275.371.673.175.373.876.476.474.675.877.074.275.874.172.478.476.175.677.276.676.176.576.478.154.877.776.477.377.477.178.179.175.5

9.138.918.739.219.109.019.24

cpm173.48177.31174.46165.53167.76167.28150.80153.73162.83167.28168.82158.12160.33163.61154.80158.21158.25160.29168.18161.22163.20271.96242.11233.05270.34294.93282.27282.85265.97288.30322.03261.80261.41252.06252.266.23229.95226.48244.42

Recovery%constantInitial

87.087.184.487.185.386.781.186.282.486.688.286.688.9.387.788.4.386.786.185.282.4.290.1.792.792.091.190.190.592.381.093.992.191.991.692.296.497.393.4

1.05081.05331.05311.05271.05461.05721.04751.04951.05001.06201.05771.05971.05601.05721.06071.05681.05451.05731.04601.05161.04901.05601.05351.05481.05791.05661.05631.05091.05381.05161.12191.05931.05831.05471.05281.05561.06711.06611.06

9.009.069.068.848.8.958.548.588.599.379.269.319.119.149.239.048.9.058.999.149.078.688.618.9.659.619.609.4.579.5010.918.928.908.778.728.808.328.298.31

249.4249.2249.2249.5249.3249.1249.3249.3249.1249.6249.4249.8249.6249.6250.0249.4249.5249.5249.4249.8249.98.4498.98.98.5499.0498.7498.5498.9498.98.7498.4498.0498.7499.0498.9499.5499.3499.5

11.411.211.011.711.311.511.712.211.011.912.012.612.612.412.912.512.712.211.812.211.715.016.116.616.414.915.415.316.315.411.115.815.516.015.915.216.817.215.3

8.669.5.638.828.808.02

250mL500mLAll1.05421.06081.05550.00440.00500.0048

Themeancellconstantkwasusedtodetermine3Hinitialintheﬁnalcolumn.Deﬁnitionsandequationsforrecoveryandthecellconstantaregiveninthetext.Q=QuantulusID.bResultsaresortedbyincreasingdateofanalysis(October2014toJuly2015)asspikeswererotatedthrougheachTEUsystem.cNoteunderperformingCell#4inthe500mLsystemisevidentbypoor2Hand3Hrecoveries.

andoneenrichedbackgroundsample.Theuseoftheseinter-comparisonsamplesallowedindependentveriﬁcationoftheaccuracyofperformanceofthe2HandSpike-ProxymethodstoobtainenrichmentfactorsandproduceaccurateﬁnalTUresults.

Theelectrolysis,2HandLSCcountingtestresultsareshowninTable3,andgivecomparativeoutcomesusingthelaser2H-enrichmentandSpike-Proxymethods.WiththetraditionalSpike-Proxymethod,threemeasuredspikesshowninTable3wereusedtodeterminethemeantritium

Measurementofextremely2H-enrichedwatersbylaserspectrometry

Tablesamples,3.Resultsofa500mL,24-sample,TEUrunwithspikes,TRICandunknownsamples,2comparingthetritiumenrichmentfactors,2HrecoveryandtheﬁnalTUplusuncertainties,andanenrichedcomparingbackgroundtheH-enrichmentandSpike-Proxymethods

Cell#SamplenameLabNo.Sampling

dateEnrichment

TUUncertaintyRecoveryEnrichment

TUUncertainty

Spikemethod

2H-Enrichmentmethod

Spike24227/30/201429.7537.865.3876.529.8535.343.3423SpikeSpike

239724217/30/20147/30/201427.829.5539.16550.905.165.6176.478.127.830.3540.42536.513.573.4945AlmendrasTRICJuly24547/15/20112.106TRICT252011T242455244912/1/201239.812/1/201228.328.47.560.12.12.4.360.120.1977.534.076.928.728.67.450.134.330.100.1878TRIC2.8TRICT23245112/1/201229.0TRICT22T212509251412/1/201212/1/201227.930.21.100.1076.52.880.450.120.1176.929.177.928.131.01.100.090.440.120.111011TRICTRICT20T252510251512/1/201212/1/201230.829.60.007.380.070.2675.976.630.829.80.007.340.070.251213TRIC29.414TRICT242511TRICT23T22251912/1/2012252012/1/201212/1/201228.84.600.1529.827.62.881.280.1077.20.1178.079.029.54.5428.52.820.151.240.100.111516TRICTRICT20T202521251612/1/201212/1/201231.629.3-0.120.30-0.090.0974.976.531.229.4-0.120.31-0.090.091718TRIC29.219TRICT212512TRICT21T22251712/1/2012251312/1/201212/1/201230.30.650.1128.930.40.661.180.1375.50.0976.479.730.40.650.1131.90.651.120.130.092021TRICTRICT23T24251835012/1/201212/1/201229.028.82.724.570.130.1477.275.729.328.62.694.610.120.142223TRIC28.024

SampleT25250EnrichedTK-11

Deadwater

250250

10/12/201412/1/20121/26/2009

30.27.790.2928.329.8

1.530.42

0.0677.30.16

76.776.1

30.57.710.2829.8

1.520.42

0.060.16

Table4.Comparisonofknown-valueIAEATRICinter-recoveryfactorswereabove75%,indicatingthattheTEUcomparisoncellswereperformingaswellasexpected.Therewasalsolaser2H-enrichmentsamples,andenrichedSpike-Proxyandprocessedmethods

usingtheexcellentagreementamongthe3H-enrichmentfactorsdeterminedbythe2H-enrichmentandSpike-Proxymethods,2withnosigniﬁcantdifferenceinthemeantritiumenrichmentEnrichmentH-factorsobtainedusingthesemethods(Student3’st-test,sampleTRICKnownvaluemethodSpikemethodN

p=0.388).AsummaryoftheinternationalHinter-comparisontestsampleresultsMean

SDMeanSD2isgiveninTable4.FortheTRICtestsamples,theH-enrichmentandSpike-ProxyTRICT257.517.580.217.500.193methodsbothyieldedaccurateandpreciseoutcomesTRIC4.49comparedwiththeirwell-establishedvalues.Inshort,theTRICT24T234.372.744.512.830.130.092.800.150.1033TRICtestsampleresultsconﬁrmedthatthelaser-based2H-TRIC1.15enrichmentmethodprovidedafasterandreliablewaytoTRICT221.120.090.580.08obtainaccurate3H-enrichmentfactorsforalltypesofTRICT21T200.431.1900.590.06

0.1230.22

0.06

0.120.22

electrolyticenrichmentcellsystems.

CONCLUSIONS

enrichmentparameter,whichwasapplieduniformlytoallthesamplecells,versusthe2HmethodwhereweusedtheTheuseofoff-axisintegratedcavityoutputlaserspectrometryaveragecellconstantasdeterminedabovefortheTEUtoaccuratelyanddirectlymeasureextremely2H-enrichedsystem.ThereportedﬁnalTUvaluesandtheirpropagatedwatersupto57,000‰(~9000ppmdeuterium)uncertaintieswereobtainedbydataprocessingwiththe2atarateof60samples/daywillfacilitatetheuseofHasatracerinIHLTRIMSsoftware.Encouragingly,itwasobviousfromenvironmentalandhydrologicalapplications.TheincorporationtheresultsinTable3thattheSpike-Proxyandlaser2H-ofthelaser-based,2H-enrichmentmethodintolow-levelenrichmentmethodsbothreturnednearlyidenticalandtritiumoperationscanfacilitatea10–20%increaseinsampleaccurateresultsforthelaboratoryspike(542.1CPM)withinthroughputbyeliminationofreplicatedspikestandards,andtheexpectedmeasurementuncertainty(10%).Forallcellslaboriousgravimetrics,andbyprovidingimmediatefeedbackinthe500mLTEUsystem(except#4),thedeuterium

onthequalityofelectrolyticcellperformance.Additional

421L.I.Wassenaaretal.

beneﬁtsoflaserspectrometryincludelowcost,easeofuse,andtheabilitytoquicklyswitchbacktonaturalabundancewatersamplesforconductingconventionalstableisotopeanalysesofenvironmentalwaters.

[12]S.Guidotti,H.G.Jansen,A.T.Aerts-Bijma,

B.M.Verstappen-Dumoulin,G.vanDijk,H.A.Meijer.Doublylabelledwateranalysis:preparation,memorycorrection,calibrationandqualityassuranceforδ2Handδ18Omeasurementsoverfourordersofmagnitudes.RapidCommun.MassSpectrom.2013,27,1055.

[13]R.vanTrigt,E.R.Kerstel,G.H.Visser,H.A.Meijer.Stable

isotoperatiomeasurementsonhighlyenrichedwatersamplesbymeansoflaserspectrometry.Anal.Chem.2001,73,2445.

[14]G.Lis,L.I.Wassenaar,M.J.Hendry.High-precisionlaser

spectroscopyD/Hand18O/16Omeasurementsofmicroliternaturalwatersamples.Anal.Chem.2008,80,287.

[15]L.I.Wassenaar,T.B.Coplen,P.K.Aggarwal.Approaches

forachievinglong-termaccuracyandprecisionofδ18Oandδ2Hforwatersanalyzedusinglaserabsorptionspectrometers.Environ.Sci.Technol.2014,48,1123.

[16]T.Thorsen,T.Shriver,N.Racine,B.A.Richman,D.A.Schoeller.

Doublylabeledwateranalysisusingcavityring-downspectroscopy.RapidCommun.MassSpectrom.2011,25,3.[17]I.D.Clark,P.Fritz.EnvironmentalIsotopesinHydrogeology.

LewisPublishers,NewYork,1997.

[18]C.M.Hoffman,G.L.Stewart.Quantitativedetermination

oftritiuminnaturalwaters,USGSWater-SupplyPaper1696-D,1966.

[19]T.Muranaka,N.Shima.Electrolyticenrichmentoftritiumin

waterusingSPEﬁlm.inElectrolysis,(Eds:J.Kleperis,V.Linkov).InTech,2012,pp.141–62.

[20]C.Taylor,inProceedingsofInternationalConferenceon

Low-RadioactivityMeasurementsandApplications,CSSR.SlovenskéPedagogickéNakladatestvo,1977,p.131.

[21]V.Faghihi,B.Verstappen-Dumoulin,H.Jansen,G.Dijk,

A.Aerts-Bijma,E.Kerstel,M.Gröning,H.Meijer.Anewhigh-qualitysetofsingly(2H)anddoubly(2Hand18O)stableisotopelabeledreferencewatersforbiomedicalandotherisotope-labeledresearch.RapidCommun.MassSpectrom.2015,29,311.

[22]V.Faghihi,H.A.Meijer,M.Groning.Athoroughly

validatedspreadsheetforcalculatingisotopicabundances(2H,17O,18O)formixturesofwaterswithdifferentisotopiccompositions.RapidCommun.MassSpectrom.2015,29,1351.[23]D.Hillegonds,L.Wassenaar,P.Klaus,P.Aggarwal.

Synthesisreport:Intercomparisontestforthedeterminationoflow-leveltritiumactivitiesinnaturalwatersforagedatingpurposes(TRIC2012).IAEAInternalDocument2014,19.

[24]IAEA.InternationalAtomicEnergyAgencyIsotope

HydrologyLaboratory,2014.

[25]L.Pujol,J.A.Sanchez-Cabeza.Optimisationofliquid

scintillationcountingconditionsforrapidtritiumdeterminationinaqueoussamples.J.Radioanal.Nucl.Chem.1999,242,391.

[26]T.B.Coplen,L.I.Wassenaar.LIMSforLasers2015for

achievinglong-termaccuracyandprecisionofδ2H,δ17O,andδ18Oofwatersusinglaserabsorptionspectrometry.RapidCommun.MassSpectrom.2015,29,2122.

Acknowledgements

ThisresearchwasfundedbytheInternationalAtomicEnergyAgency.WethankD.BrummerandC.Sambandamforassistancewithsampleprocessing.Wethankthereviewersfortheirconstructivecomments.

REFERENCES

[1]D.A.Schoeller.Measurementofenergy-expenditurein

free-livinghumansbyusingdoublylabeledwater.J.Nutr.1988,118,1278.

[2]K.A.Nagy,I.A.Girard,T.K.Brown.Energeticsof

free-rangingmammals,reptiles,andbirds.Annu.Rev.Nutr.1999,19,247.

[3]D.X.Soto,K.A.Hobson,L.I.Wassenaar.Theinﬂuenceof

metaboliceffectsonstablehydrogenisotopesintissuesofaquaticorganisms.Isot.Environ.HealthStud.2013,49,305.[4]Z.D.Sharp,V.Atudorei,H.O.Panarello,J.Fernandez,

C.Douthitt.Hydrogenisotopesystematicsofhair:archeologicalandforensicapplications.J.Archaeol.Sci.2003,30,1709.

[5]G.vanderKamp,D.R.vanStempvoort,L.I.Wassenaar.The

radialdiffusionmethod.1.Usingintactcorestodetermineisotopiccomposition,chemistry,andeffectiveporositiesforgroundwaterinaquitards.WaterResourcesRes.1996,32,1815.[6]M.J.Hendry,S.L.Barbour,K.Novakowski,L.I.Wassenaar.

PaleohydrogeologyoftheCretaceoussedimentsoftheWillistonBasinusingstableisotopesofwater.WaterResourcesRes2013,49,4580.

[7]U.Morgenstern,C.B.Taylor.Ultralow-leveltritium

measurementusingelectrolyticenrichmentandLSC.Isot.Environ.HealthStud.2009,45,96.

[8]C.B.Taylor.Therelationshipbetweenelectrolyticdeuterium

andtritiumseparationfactors,andattainmentofimprovedaccuracyinradiometriclow-leveltritiummeasurement.Appl.Radiat.Isot.1994,45,683.

[9]H.Östlund,E.Werner,inTritiuminthePhysicalandBiological

Sciences,vol.1.ProceedingsofaSymposium,1962.

[10]J.Horita,C.Kendall.Stableisotopeanalysisofwaterand

aqueoussolutionsbyconventionaldual-inletmassspectrometry.inHandbookofStableIsotopeAnalyticalTechniques,(Ed:P.A.deGroot).Elsevier,2004,pp.1–37.[11]T.Donnelly,S.Waldron,A.Tait,J.Dougans,S.Bearhop.

Hydrogenisotopeanalysisofnaturalabundanceanddeuterium-enrichedwatersbyreductionoverchromiumon-linetoadynamicdualinletisotope-ratiomassspectrometer.RapidCommun.MassSpectrom.2001,15,1297.

422wileyonlinelibrary.com/journal/rcm

RapidCommun.MassSpectrom.2016,30,415–422

因篇幅问题不能全部显示，请点此查看更多更全内容

查看全文