V.F.1 Effect of System Contaminants on PEMFC Performance ...FY 2014 Annual Progress Report V–129...

V–128DOE Hydrogen and Fuel Cells Program FY 2014 Annual Progress Report

Huyen N. Dinh (Primary Contact), Guido Bender, Clay Macomber, Heli Wang, Bryan PivovarNational Renewable Energy Laboratory (NREL)15013 Denver West ParkwayGolden, CO 80401-3305Phone: (303) 275-3605Email: Huyen.Dinh@nrel.gov

DOE Manager David Peterson Phone: (720) 356-1747Email: David.Peterson@ee.doe.gov

Subcontractors• PaulT.Yu,BalsuLakshmanan,GeneralMotorsLLC,HoneoyeFalls,NY

• JohnWeidner,JohnVanZee,UniversityofSouthCarolina, Columbia, SC

• RyanRichards,JasonChrist,ColoradoSchoolofMines,Golden, CO

• SteveHamrock,AndrewSteinbach(in-kindpartner), 3M, St. Paul, MN

• PaulBeattie(in-kindpartner),BallardPowerSystems,Vancouver, BC, Canada

• OlgaPolevaya(in-kindpartner),NuveraFuelCellsInc.,Billerica,MA

ProjectStartDate:July20,2009 Project End Date: Project continuation and direction determined annually by DOE

Overall Objectives Our overall objective is to decrease the cost associated

withsystemcomponentswithoutcompromisingfunction,fuelcellperformance,ordurability.Ourspecificprojectobjectives are to:

Identifyandquantifysystem-derivedcontaminants.•

Developexsituandinsitutestmethodstostudy•contaminantsderivedfromsystemcomponents.

Identifyseverityofsystemcontaminantsandimpactof•operatingconditions.

Identifycontaminationmechanisms.•

Developmodels/predictivecapability.•

Guidesystemdevelopersonfuturematerialselection.•

Disseminate knowledge gained to the community.•

Fiscal Year (FY) 2014 Objectives Identifyimpactofoperatingconditions.•

Developamechanisticmodelforcontamination.•

Disseminateprojectinformationtothefuelcell•community.

Developunderstandingofleachingconditions’impacton•contaminant concentration.

Technical BarriersThisprojectaddressesthefollowingtechnical

barriersfromtheFuelCellssection(3.4.4)oftheFuelCellTechnologiesOfficeMulti-YearResearch,Development,andDemonstration Plan:

(A) Durability

(B) Cost

Technical TargetsThisprojectfocusesonquantifyingtheimpactofsystem

contaminantsonfuelcellperformanceanddurability.Insightsgainedfromthesestudieswillincreaseperformanceand durability by limiting contamination-related losses and decreasingoverallfuelcellsystemcostsbyloweringbalance-of-plant(BOP)materialcosts.ProperselectionofBOPmaterialswillhelpmeetthefollowingDOE2020targets:

Cost:$30/kWfortransportation;$1,000–1,700/kWfor•stationary

Lifetime:5,000hoursfortransportation;60,000hours•forstationary

FY 2014 Accomplishments Developedtheleachingindexasaquickmaterial•screening method.

Identifiedimpactofvariousfuelcelloperating•conditions (contaminant concentration, relative humidity, celltemperature,currentdensity,andcatalystloading)onfuelcellperformanceandrecoveryforselectedstructuralmaterialextracts.Thisknowledgecanhelpidentifyfuturemitigationstrategiesforcontaminants.

Developedamodelforcontaminationmechanismbased•onexperimentswithmodelorganiccompounds.

ImprovedNRELwebsite(www.nrel.gov/hydrogen/•contaminants.html) and interactive material data tool (www.nrel.gov/hydrogen/system_contaminants_data/)byaddingmoredata(60systemcomponentmaterials

V.F.1 Effect of System Contaminants on PEMFC Performance and Durability

V–129FY 2014 Annual Progress Report DOE Hydrogen and Fuel Cells Program

V.F Fuel Cells / ImpuritiesDinh – National Renewable Energy Laboratory

total)andprojectinformationandimprovinguserexperience.

PresentedDOEwebinaron“AnOverviewofNREL’s•OnlineDataToolforFuelCellSystem-DerivedContaminants” [1].

G G G G G

IntroDuCtIon Costanddurabilityissuesofpolymerelectrolyte

membranefuelcell(PEMFC)systemshavebeenchallengingforthefuelcellindustry.Thecurrentstatusoffuelcellsystemcostsis$55/kW,muchlowerthan$124/kWin2006,butstillhigherthantheultimatetargetof$30/kW[2].Asfuelcellsystemsbecomemorecommerciallycompetitive,theimpactofcontaminantsderivedfromfuelcellsystemcomponentmaterialshasriseninimportance.Contaminantsderivedfromfuelcellsystemcomponentmaterials—structural materials, lubricants, greases, adhesives, sealants, andhoses—havebeenshowntoaffecttheperformanceanddurabilityoffuelcellsystems.LoweringthecostofPEMFCsystemcomponentsrequiresunderstandingofthematerialsusedinthesecomponentsandthecontaminantsthatarederivedfromthem.Unfortunately,therearemanypossiblecontaminationsourcesfromsystemcomponents[3-5].Currentlydeployed,high-cost,limited-productionsystemsuseexpensivematerialsforsystemcomponents.Inordertomakefuelcellsystemscommerciallycompetitive,thecostofBOPcomponentsneedstobeloweredwithoutsacrificingperformanceanddurability.FuelcelldurabilityrequirementslimittheperformancelossattributabletocontaminantstoatmostafewmVoverrequiredlifetimes(thousandsofhours),which means system contaminants must have a near-zero impact.

Ascatalystloadingsdecreaseandmembranesaremadethinner(botharecurrenttrendsinautomotivefuelcellresearchanddevelopment),operationoffuelcellsbecomesevenmoresusceptibletocontaminants.Inconsumerautomotivemarkets,low-costmaterialsareusuallyrequired,butlowercosttypicallyimplieshighercontaminationpotential.Theresultsofthisprojectwillprovidetheinformationnecessarytohelpthefuelcellindustrymakeinformeddecisionsregardingthecostofspecificmaterialsversusthepotentialcontaminantimpactonfuelcellperformanceanddurability.Theprojectresultswillalsoidentifytheimpactofdifferentoperatingconditionsandpossiblemitigationstrategiesforcontaminants.

APProACh Ourgoalistoprovideanincreasedunderstanding

offuelcellsystemcontaminantsandtohelpguidetheimplementationand,wherenecessary,developmentofsystem

materialstosupportfuelcellcommercialization.Whilemuchattentionhasbeenpaidtoairandfuelcontaminants,systemcontaminantshavereceivedlimitedpublicattentionandverylittleresearchhasbeenpubliclyreported[6-8].Ourapproachistoperformparametricstudiestocharacterizetheeffectsofsystemcontaminantsonfuelcellperformance,aswellastoidentifytheseverityofcontamination,identifycontaminationmechanisms,developamodel,anddisseminateinformationaboutmaterialcontaminationpotentialthatwouldbenefitthefuelcellindustryinmakingcost-benefitanalysesforsystemcomponents.TheBOPmaterialsselectedforthisstudyare commercially available commodity materials and are generallydevelopedforotherapplicationsforwhichcommonadditives/processingaidsmaynotbeaconcern,buttheymaypresentproblemsforfuelcells.Westudiedleachatesaswellasmodelcompoundsthatarecapableofreplicatingthedeleteriousimpactofsystem-basedcontaminants.

rESultS Oneofthisyear’saccomplishmentswasexpanding

theBOPmaterialdatabaseandprojectinformationaswellasimprovingtheuserexperienceontheNRELwebsite.Thescreeningresultsfor60commerciallyavailableBOPmaterials (structural, hoses, assembly aids such as seals, gaskets,andadhesives),usingmultiplescreeningmethodstoidentifyandquantifysystem-derivedcontaminants,arearchivedandmadepubliclyaccessibleontheNRELwebsite.TheNRELmaterialscreeningdatatoolwasdesignedtobeinteractive,easytouseandinformativetothefuelcellcommunity.Furthermore,aDOEwebinarwaspresentedbyDinhtogiveanoverviewofNREL’sonlinedatatoolandprovideatutorialonhowtousetheWeb-basedtooltoaccessprojectresults[1].

General Motors (GM) screened and categorized 34 structuralplasticmaterialsintogroupsbasedontheirbasicpolymerresin(e.g.,polyamideorPA)andmanufacturers.Theyfoundthattheleachingindex(LI),whichisthesumofthesolutionconductivityandtotalorganiccarbon(TOC),isaquickwaytoscreenplasticmaterials.Theleachingindexisanindicatoroftheamountofcontaminants(organics,inorganics,andions)leachingfromthematerial.Figure1showsthathigherleachingindexgenerallyresultsinhighercell voltage loss and is correlated with lower material cost. TheimplicationisthatfuelcelldeveloperscandoaquickscreeningoftheBOPmaterialcandidatesbycarryingouttheleachingexperimentandmeasuretheTOCandsolutionconductivityoftheextractsolution.Thesemeasurementsarequickandeasytodo.Ifsomegoodmaterialcandidatesarefound,thenfurthertesting,suchaselectrochemistry,membrane conductivity, advanced analytical characterization,andinsituinfusionexperimentscanbecarriedouttobetterunderstandwhatcontaminantspeciesarepresentandhowtheyimpactfuelcellperformance.

Dinh – National Renewable Energy LaboratoryV.F Fuel Cells / Impurities

V–130DOE Hydrogen and Fuel Cells Program FY 2014 Annual Progress Report

From34structuralplasticmaterialsscreened,threewereselectedforinsituinfusionparametricstudiestounderstandtheeffectofthepolymerresin(PAandPPAorpolyphthalamide),additive(e.g.,percentofglassfiberaddedforplasticstructuralintegrity),anddifferentoperatingconditions (contaminant concentration, relative humidity (RH),celltemperature,currentdensity(CD),andcatalystloading)onfuelcellperformanceandrecovery.Theparametersstudiedreflect80%oftypicalfuelcelloperatingconditions.Figure2showsthatthePAmaterial(EMS-4),whichhasthehighestLI,resultedinhighervoltagelossthanPPAmaterials.Furthermore,thePPAmaterialthathasthelowerglassfiber(GF)content(30%GFforEMS-10vs.50%GFforEMS-7)resultedinalowerLIandlowerfuelcellperformanceloss.Theseresultsimplythatthepolymerresintypeandadditivesareimportantcontaminantsourceconsiderations.Inaddition,Figure2showsthattheinsitu

Figure 1. Higher leaching index (conductivity + total organic carbon) is generally correlated with higher fuel cell performance loss and lower material cost. BES = Bakelite epoxy-based material – Sumitomo; BPS = Bakelite phenolic-based material – Sumitomo; S = Solvay; C = Chevron Philips; B = BASF; D = DuPont; E = EMS; Information provided by GM.

Figure 2. In situ fuel cell voltage loss due to contaminants (dV1) increases linearly as a function of structural material leachate concentration due to contamination of the fuel cell cathode: (a) EMS-4 50% glass fiber PA, (b) EMS-7 50% glass fiber PPA, and (c) EMS-10 30% glass fiber PPA. The voltage loss after passive recovery (dV2) is also shown. The plots also show that polymer resin type and additives in plastic materials matter. The LI for the different materials is also shown for comparison. Standard operating conditions (SOC): 80°C, 32/32% inlet RH, 0.2 A/cm2, H2/air stoichiometry = 2/2; 150/150 kPa; Information provided by GM.

V–131FY 2014 Annual Progress Report DOE Hydrogen and Fuel Cells Program

V.F Fuel Cells / ImpuritiesDinh – National Renewable Energy Laboratory

fuelcellvoltagelossduetocontaminants(dV1)increaseslinearlyasafunctionofleachateconcentration(redline)andthecontaminationeffectcanbepartiallyreversedintheabsenceofcontaminants(blueline).Asimilartrendwasobservedforallthreestructuralmaterialsstudied.

Figure3summarizesthemaineffectofdifferentoperatingconditions(concentration,RH,CD,andcatalystloading)onfuelcellperformancelossduetocontamination(dV1)andrecovery(dV2)intheabsenceofcontaminants(alsoknownaspassiverecovery).Asexpected,fuelcellswithlowPtloadingaremoresensitivetoBOPplasticleached contaminants and result in higher cell voltage loss, regardlessofthematerialstudied.Figure3ashowsthatthevoltage loss increases with increasing current density while RHappearstohaveaminimumeffectonvoltageloss.RHisacomplicatedfactorsinceitcontrolsthemolefractionofbothwaterandcontaminantintothefuelcell.AsRHincreases,morewatervaporentersthefuelcellandcanhelpflushoutthecontaminants.However,morewatervaporalsomeansmorecontaminantsarebroughtintothefuelcellandresultsinhighervoltageloss.ThesetwophenomenamaycountereachotherandleadtoinsensitivityofRHtofuelcellperformanceloss.Figure3bshowsthatthesefourparametershavesimilareffectonthevoltagelossafterpassiverecovery(dV2),butthemagnitudeofthevoltagelossislowercomparedtodV1.Thesevoltagelosseswereobtainedduringinfusionatrelativelylowcurrentdensity(0.2A/cm2). Analysisofthepolarizationcurvesbeforecontamination(beginningoflife)andafterpassiverecoveryshowedthatthetrendonfuelcellvoltagelossduetotheseoperatingparametersissimilaratlowandhighcurrentdensities(e.g.,1.2A/cm2).

StatisticalanalysisoftheparametricresultsshowedthatCDand/ordosageare/isthemostsignificantfactor(s)affectingcellperformance,followedbyleachate

concentration,interactionofRHandPtloading,Ptloading,andinteractionofRHandconcentration.Itisimportanttonotethatinteractionbetweendifferentoperatingconditionsshouldbeconsideredwithrespecttocontaminationeffect.Forexample,trendstowardlowercatalystloadingsmaymeanthatfuelcellsneedtooperateathigherRHsincethesetwoparametersinteractwithoneanother.

Fromtheparametricstudy,wehaveidentifiedseveralmitigation strategies to minimize the leachate concentration (leachingindex).Thesestrategiesincludeminimizingthecontacttimeandcontactratiooftheplasticmaterialswithwaterinthefuelcell,minimizingexposureofplasticmaterialtohightemperature,increasing the RH or increasing the RHandpotentialcycling(exsiturecovery),choosingcleanBOPmaterials(usuallymoreexpensive,e.g.,resintypeandadditive),andworkingwithmaterialsupplierstominimizecontaminants(i.e.,removingadditivesthatarenotapplicabletofuelcellsandusinglessoralternativeadditivesthatdonotleachoutcontaminants).Thesestrategiescanminimizefuelcellperformancelossduetosystem-derivedcontaminants.

ConCluSIonS AnD FuturE DIrECtIonSWeimprovedtheNRELprojectwebsiteandinteractive•datatoolbyexpandingthematerialdatabase,enhancinguserexperience,archivingtheresults,andmakingthempubliclyaccessible.

Wedevelopedtheleachingindexasagood,quick•screeningmethodforpotentialsystemcomponents.Thisdata is also included on the NREL website.

Wefoundthatcost,polymerresintypeandadditives•needtobeconsideredwhenselectingBOPplasticmaterialsforfuelcellsystemsbecausethechoicecanhavedifferentdegreesofcontaminationimpact.

Figure 3. Summary of the effects of different operating conditions on fuel cell performance loss (dV1) and passive recovery (dV2). SOC were used.

65.048.532.0

60

50

40

30

20

0.200.130.06

0.40.20.1

60

50

40

30

20

1.000.550.10

Me

an

volt

ag

elo

ss[m

V]

Main Effects Plot for dV1Fitted Means

Main Effects on Contamination (dV1)

RH [%] CD: [A/cm2]

Pt loading [mg/cm2] Conc. [dilution factor]

65.048.532.0

30

25

20

15

100.200.130.06

0.40.20.1

30

25

20

15

101.000.550.10

Me

an

volt

ag

elo

ss[m

V]

Main Effects Plot for dV2Fitted Means

Main Effects on Recovery (dV2)

RH [%] CD: [A/cm2]

Conc. [dilution factor]

Pt loading [mg/cm2]

(a) (b)

Dinh – National Renewable Energy LaboratoryV.F Fuel Cells / Impurities

V–132DOE Hydrogen and Fuel Cells Program FY 2014 Annual Progress Report

Wefoundthatcontaminationimpactdependsonfuel•celloperatingconditions(CD,concentration,Ptloading,RH interaction with Pt loading and concentration, temperature)andthatinteractionsbetweendifferentoperatingconditionsneedtobeconsidered.

Wefoundthatoperatingconditions(e.g.,time,•temperature)thatcausemoreliquid/plasticcontactneedtobeconsideredindevelopingafuelcellsystembecausethey can lead to higher contaminant concentration (higherleachingindex).

Wehaveidentifiedseveralmitigationstrategiesto•minimizetheleachingindexandhenceminimizetheperformanceloss.

Wewilldeterminethefuelcellperformanceimpactof•lower leachate concentrations.

Wewilldevelopanalyticalmethodstomeasuresoluble•leachatesinsolutionandvolatilesinheadspace.

Wewillperformmechanisticstudiesonorganicand•ionicmodelcompoundsderivedfromstructuralplasticstounderstandtheeffectofindividualandmixturesofcompoundsonfuelcellperformance.

WewilldisseminateprojectinformationviatheNREL•website,publications,reports,andpresentations.

FY 2014 PublICAtIonS/PrESEntAtIonS 1.Wang,H.;Macomber,C.S.;Christ,J.;Bender,G.;Pivovar,B.;Dinh,H.N.“EvaluatingtheInfluenceofPEMFCSystemContaminantsonthePerformanceofPtCatalystviaCyclicVoltammetry.”Electrocatalysis(5),2014;pp.62-67.

2.Yu,P.T.;Bonn,E.A.;Lakshmanan,B.“ImpactofStructuralPlasticsasBalanceofPlantComponentsonPolymerElectrolyteMembraneFuelCellPerformance.”ECS Transactions (58:1),2013;pp.665-680.

3.Dinh,H.N.“EffectofSystemContaminantsonPEMFCPerformanceandDurability.”SectionV.B.1.2013 DOE Hydrogen and Fuel Cells Program Annual Progress Report;pp.V129-V134.

4. Dinh, H.N.“EffectofSystemContaminantsonPEMFCPerformanceandDurability.”DOEFuelCellTechnologiesOfficeAnnualMeritReview,Washington,DC,June2014.

5.Dinh,H.N.“AnOverviewofNREL’sOnlineDataToolforFuelCellSystem-DerivedContaminants.”DOEFuelCellTechnologiesOfficeWebinar,May27,2014.

6.Christ,J.M.“AdsorptionCharacteristicsofPolymerElectrolyteMembraneChemicalDegradationProductsandtheirImpactonOxygenReductionReactionActivityforPlatinumCatalysts.”ColoradoSchoolofMinesChemistryDepartmentalSeminar,Golden,CO,April2014.

7.Christ,J.M.;Neyerlin,K.C.;Richards,R.;Dinh,H.N.PresentationattheElectrochemistryGordonResearchConference,Ventura,CA,January2014.

8.Dinh,H.N.;Yu,P.T.;Weidner,J.“EffectofSystemContaminantsonPEMFCPerformanceandDurability.”PresentationtotheDOEFuelCellTechTeam,Southfield,MI,Jan.15,2014.

rEFErEnCES 1.Dinh,H.N.“AnOverviewofNREL’sOnlineDataToolforFuelCellSystem-DerivedContaminants.”DOEFuelCellTechnologiesOfficeWebinar,May27,2014.http://energy.gov/sites/prod/files/2014/06/f16/webinarslides_fuel_cell_contaminants_052714.pdf.

2.http://www.hydrogen.energy.gov/pdfs/13012_fuel_cell_system_cost_2013.pdf.

3. D.Gerard.“MaterialsandMaterialsChallengesforAdvancedAutomotiveTechnologiesBeingDevelopedtoReduceGlobalPetroleumConsumption.” 2007MaterialsScienceandTechnology(Invited),Detroit,MI,September16–20,2007.

4.D.A.Masten,A.B.Bosco.Handbook of Fuel Cells (eds.: W.Vielstich,A.Lamm,H.A.Gasteiger),Wiley(2003):Vol.4,Chapter 53,p.714.

5. Q. Guo,R.Pollard,J.Ruby,T.Bendon,P.Graney,J.Elter.ECS Trans.,Vol.5(1),p.187(2007).

6.G.A.James,et.al.“PreventionofMembraneContaminationinElectrochemicalFuelCells.”U.S.PatentApplicationUS2005089746A.

7.R.Moses.“MaterialsEffectsonFuelCellPerformance.”NationalResearchCouncilCanadaInstituteforFuelCellInnovation,InternationalFuelCellTestingWorkshop,September20–21,2006.

8.K.O’Leary,B.Lakshmanan,M.Budinski.“MethodologiesforEvaluatingAutomotivePEMFuelCellSystemContaminants.”2009Canada-USAPEMNetworkResearchWorkshop,February16,2009.