Chemical deactivation by phosphorous under lean hydrothermal conditions over Cu/BEA NH3-SCR catalysts

(1)

ContentslistsavailableatScienceDirect

Applied

Catalysis

B:

Environmental

j ou rn a l h om epa g e : w w w . e l s e v i e r . c o m / l o c a t e / a p c a t b

Chemical

deactivation

by

phosphorous

under

lean

hydrothermal

conditions

over

Cu/BEA

NH

3

-SCR

catalysts

Stanislava

Andonova

a

_,

_Evgeny

_Vovk

b,c

_,

_Jonas

_Sjöblom

d

_,

Emrah

Ozensoy

b

_,

_Louise

_Olsson

a,∗

a_Competence_Centre_for_Catalysis,_Chemical_Engineering,_Chalmers_University,₄₁₂₉₆_Gothenburg,_Sweden b_Chemistry_Department,_Bilkent_University,₀₆₈₀₀_Bilkent,_Ankara,_Turkey

c_Boreskov_Institute_of_Catalysis,₆₃₀₀₉₀_Novosibirsk,_Russian_Federation d_Applied_Mechanics,_Chalmers_University,₄₁₂₉₆_Gothenburg,_Sweden

a

r

t

i

c

l

e

i

n

f

o

Articlehistory: Received5July2013 Accepted27August2013 Available online 1 September 2013 Keywords: NH3SCR NOxreduction Cu/BEAcatalysts Ppoisoning Deactivation.

a

b

s

t

r

a

c

t

ToobtainabetterunderstandingofthedeactivationofSCRcatalyststhatmaybeencountereddueto thepresenceofP-containingimpuritiesindieselexhausts,theeffectsinducedbyPoverCu/BEANH3

-SCRcatalystswerestudiedasfunctionsofthetemperatureofpoisoningandPconcentrationinthe feed.Cu/BEAcatalystswithdifferentCuloadings(4and1.3wt%Cu)wereexposedtoPbycontrolled evaporationofH3PO4inthepresenceof8%O2and5%H2Oat573and773K.Thereactionstudieswere

performedbyNH3-storage/TPD,NH3/NOoxidationandstandardNH3-SCR.Inaddition,acombinationof

severalcharacterisationtechniques(ICP–AES,BETsurfacearea,poresizedistribution,H2-TPRandXPS)

wasappliedtoprovideusefulinformationregardingthemechanismofPdeactivation.Porecondensation ofH3PO4incombinationwithporeblockingwasobserved.However,themeasuredoveralldeactivation

wasfoundtooccurmostlybychemicaldeactivationreducingthenumberoftheactiveCuspeciesand hencedeterioratingtheredoxpropertiesoftheCu/BEAcatalysts.TheprocessofPaccumulationonthe surfacepreferentiallyoccursonthe“overexchanged”Cuactivesiteswiththeformationofphosphate species.Thisislikelythereasonforthemoreseveredeactivationofthe4%Cu/BEAcomparedto1.3% Cu/BEA.Further,thehigherNOxreductionperformanceat773KoftheP-poisonedCu/BEAcatalystswas

foundtooriginatefromthelowerselectivitytowardsNH3oxidation,whichoccurspredominatelyonthe

“over-exchanged”sites.

1. Introduction

In thelast decade, the selectivecatalytic reduction (SCR) of nitrogenoxides(NOx)hasreceivedsigniﬁcantattentionduetoits

numerousapplicationstoreduceNOxemissionsintheexhaustof

stationarypowerplants,industryprocessesandrecentlyalsofrom automotivesources[1–3].ThecurrentstrategyofSCRbyusingNH3,

inparticularfordiesel-equippedvehicles[4–6]isnowadays con-sideredasoneofthemosteconomicalandeffectiveNOxabatement

catalytictechnology.

Vanadia-basedcatalysts(V2O5/WO3/TiO2)arethemost

com-monlyused and widely investigated for SCR [7]. However,the inadequatestabilityofthesecatalystsathightemperaturesand athighspacevelocities,incombinationwithtoxicity,shiftedthe focusoftheinvestigationstoanothergroupofmaterialsbasedon transition-metalion-exchangedzeoliteswhichofferanadvantage

∗ Correspondingauthor.Tel.:+4631-7724390;fax:+4631-7723035. E-mailaddress:louise.olsson@chalmers.se(L.Olsson).

ofimprovedNOxreductionperformanceandthermalstabilityina

widetemperaturerange.Hence,differentexperimentaland theo-reticalstudies[8–13]werefocussedontheeffectofthemetal(Fe, Cu,Cr,Ce,CoandRh)andthetypeofthezeolites[5,14,15](ZSM-5, MFI,FER,BEA,SSZ-13andSAPO-34)onthestabilityandtheoverall SCRperformanceoftheexhaustzeolite-basedcatalysts.Ingeneral, Fe-andCu-basedzeolitesareselectedasthemostactiveandstable SCRcatalystsforNOxreduction.Inparticular,itwasfound[16–22]

thatCu-ionexchangedzeolitesarecharacterisedbysuperiorlow temperatureNOxconversionand N2 selectivityin NH3-SCRand

directNOdecomposition.

Poisoningof thedieselexhaust catalystscausedby accumu-lationofimpuritiesintheformofsigniﬁcantamountofoil-and fuel-derived contaminants (P,Zn, Ca, K, Naand Mg) deposited onthesurface[23–30]isoneoftheproblemsthathavenotbeen totallysolvedwiththecurrentSCRtechnology.Inparticular,the effectsinducedbyphosphorous(P)areoneofthemajorproblems inpracticalapplicationsoftheSCRcatalystsduetotheir deacti-vation byP-containing impurities in biodiesel and lubricantoil additives[24].Ithasbeenshown[31,32]thatPcangreatlyimpair 0926-3373/$–seefrontmatter © 2013 Elsevier B.V. All rights reserved.

(2)

theeffectivenessoftheNOxemissioncontrolsystemsduetoits

cumulativeinﬂuenceover V2O5-WO3/TiO2 catalysts.Thus, even

verylowlevelsofPinthefuelmayleadtodeteriorationovertime, especiallywhenanengineconsumesasigniﬁcantamountof con-taminatedfuel.Itwasshown[33,34]thatthePcontaminationcan depositonthesurfaceoftheusedautomotivecatalysts,intheform ofdifferentphosphatespeciessuchasaglassy/amorphousphaseof Pb,ZnandCaphosphates.Inaddition,itwasfound[35–39]thatPis usuallyconcentratedintheforward-mostsectionofthemonolithic three-wayconverters(TWC).Adecreaseincatalyticactivityand changesincharacteristics,suchasalossofsurfacearea,inthefront sectionoftheTWC catalyticsystems havebeenassociated [36]

withextensivephosphorusdeposition.Inaddition,itwasreported

[40]thattheexposureofFe-zeoliteSCRcatalysttoPcanleadtoa lossofNOxconversionandanincreaseinNH3slip.Thiswaspartly

attributedtoalossofNH3storageabilityduetoPphysicalblockage.

Severalstudies [24,31,32,41–44] focussedonthe effect of P deactivation on both model V2O5-TiO2 and commercial V2O5

-WO3-TiO2catalysts.ThePpoisoningwasinvestigated[42,43]using

wetimpregnation of thecatalysts withan aqueoussolution of H3PO4.Thepoisoningstrengthwasfound[42]toberelativelylarge

atlowerreactiontemperaturesduetotheformationofdifferent deactivatingspeciesonthesurface.ThedecreaseinSCRactivity wasalsoobservedonV-basedcatalystswiththeadditionofH3PO4

asanextrusionbinder[45].InanattempttostudytheP poison-ingeffectbyusingdifferentapproachesforchemicaldeactivation

[24,31,32,46],itwasconcluded[42,43]that therealmechanism ofdeactivation cannotbereproduced bythe wetimpregnation methodoftenemployed.

Despitethenumerousinvestigations[24,31,32,41–44]carried outwithpoisonedV-basedcatalysts,theindividualeffectsofPas wellastheinﬂuenceofsimultaneouspoisoninginhydrothermal conditionsarestillnotsufﬁcientlyknown.Thereexistsonlyafew studiesintheliterature[40,47,48]addressingtheimpactof differ-entinorganicpoisons(P,Ca,Mg,Zn,K)ontheperformanceofFe zeolites-basedNH3-SCRcatalysts.Itwasfoundthattheexposure

ofFe-zeolitestoPleadstoastrongdeactivationofthecatalysts. However,adetailedknowledgeabouttheeffectsrelatedtoP deac-tivationofCuzeolitesforNH3-SCRcatalysisisstilllacking.

Inthelight oftheseﬁndings,theeffortsinthecurrentwork werefocussedonstudyingthe effectsinduced byP onCu/BEA NH3-SCRcatalystsbyclarifyingtheirmechanismofdeactivation

under well-deﬁned and more realistic conditions of poisoning. Themonolithsampleswereexposed toPat different tempera-turesbycontrolledevaporationofH3PO4 in thepresenceof5%

H2Oand 8%O2. Theoverall SCRoperationwastestedover the

freshCu/BEAcatalystswithdifferentCucontent(4and1.3wt%Cu) andcomparedtoP-poisonedCu/BEAsamples.Thereactionstudies duringNH3-storage/temperature-programmeddesorption(TPD),

NH3/NOoxidationandstandardNH3-SCRwereperformedinﬂow

reactorexperimentsintherangeof423–773K.Inaddition,a com-binationofseveralcharacterisationtechniqueswasapplied,such asinductivelycoupledplasmaatomicspectroscopy(ICP–AES), sur-faceareameasurements,pore sizedistribution,H2-temperature

programmedreduction(H2-TPR) andX-rayphotoelectron

spec-troscopy(XPS).Theanalysiswasdirectedtowardsclarifyingthe mechanismofPdeactivationofthecatalysts,byfocussingthe stud-iesonthenatureoftheformeddeactivatingspeciesonthesurface.

2. Experimental

2.1. Catalystpreparation

Cu/BEAcatalystswithdifferentCucontent(1.3and4wt%Cu) werepreparedfollowing theprocedureof ionexchange ofBEA

zeolite(SiO2/Al2O3=38,ZeolystInternational)withNaNO3(Merck)

andthenwithCu(CH3COO)2(Merck).Intheﬁrststep,theNa-form

ofBEAsampleswaspreparedbyexchanging50gofthezeoliteina solutionofNaNO3byvaryingtheconcentrationofthesolution(21.6

and108mMNaNO3,respectively).Theexchangewascarriedoutby

agitatingtheslurryatroomtemperaturefor1h,withtheinitialpH adjustedto6.6andkeptconstantusingNH4OH.Thesolidwasthen

filteredandwasheduntiltheneutralpHofthefiltratewasreached. Theaboveprocessincludingtheion-exchange,filteringand wash-ingwasrepeatedtwotimes.Inthesecondstage,theNa-formof BEAsamplesdriedat353Kfor12h,wereusedforthenextstep ofCuionexchangewithCu(CH3COO)2byvaryingtheCu

precur-sorconcentration(2.2and11mMCu(CH3COO)2,respectively).The

exchangewascarriedoutbyagitatingtheslurryatroom temper-aturefor1h,withtheinitialpHadjustedto6.0andkeptconstant duringthestirring.Afterﬁltrationandwashingtheaboveprocess ofCuion-exchangewasrepeatedtwomoretimestogiveatotalof threeexchanges.Finally,theresultingpowderwasdriedat353K for12handcalcinedat723Kfor3h.

The calcined powder catalysts were used to coat cordierite monoliths. The monoliths were cut from a commercial honey-combcordieritestructure(length=20mm,diameter=22mmand celldensityof400cpsi)andheatedto773Kfor2h.Asolidphaseof 5wt%boehmite(DisperalD,Sasol,GmbH)dissolvedinaslurry mix-tureconsistingoftheliquidphase(distilledwater/ethanol=50/50) wasﬁrstusedfortheimpregnationofthecalcinedmonolithsin ordertoenhancetheattachmentof theion exchangedcatalyst. Thealumina-coatedmonolithswerecalcinedat773Kfor2h.Then, theprocedureconsistedofimmersingthemonolithsintoaslurry composedofaliquidphaseofequalamountsofdistilledwaterand ethanolandasolidphaseof5wt%boehmiteand95wt%catalyst. Thesolidintheslurrywas20%w/w.Theprocedureofthe immer-sion,blowingawaytheexcessslurry,drying(363Kfor2min)and heating(823Kfor2min)inairwasrepeatedseveraltimesuntil themonolithwascoatedwiththedesiredamountof washcoat (∼700mg).Finally,thewash-coated monolithswerecalcinedat 773Kfor2h.

2.2. PexposureofCu/BEAcatalystsinleanhydrothermal conditions

ThePpoisoningoftheCu/BEAmonolithsampleswasperformed byusingtheexperimentalset-upwhichhasbeendescribedindetail elsewhere[49,50].Themonolithcatalystwasinsertedinthemiddle oftheheatedzoneofahorizontalquartztubereactor,whichwas equippedwithaninsulatedheatingwirecontrolledbyEurotherm temperature-controller. The temperature was measuredwith a thermocouplepositionedabout10mminfrontofthemonolithand asecondoneplacedinthecentreofthemonolith.TosimulateP poisoninginleanhydrothermalconditions,anaqueoussolutionof H3PO4intheformofsteamwasfedintothequartzreactor(byusing

acontrolledevaporationandmixingviaaBronkhorstsystem)inthe presenceof5%H2O,8%O2andAr.Thetotalgasﬂowratewasheld

constantat3500mlmin−1,givingaspacevelocityof30,300h−1, basedonmonolithvolume.Theresultingmixturewasthenpassed overthemonolithsamples.Theprocedurewasdevelopedto com-paretheeffectsofpoisoningattwodifferenttemperaturesat573 and773K,whilethedurationoftheexposure(4.4h)waskept con-stant.Themonolithswereﬁrstexposedat573Kto50ppmPand thento100ppmPbyincreasingtheamountofH3PO4inthefeed.In

asimilarway,thePpoisoningatthehighertemperature(773K)was carriedoutbyusinganewmonolithsample.Todeterminetheeffect ofchangingthePconcentrationandthetemperatureof poison-ing,theactivitymeasurements(describedbelow)wereperformed overthefreshandP-poisonedcatalystsaftereachstepofP expo-sure.Topreventtheformationof(NH4)3PO4,aftertheprocedure

(3)

ofPpoisoning,anextensivecleaningofthereactorwascarriedout beforestartingtheactivitymeasurementsoverthepoisoned cata-lysts.Alllineswereheatedandmaintainedattemperaturesabove 423KtopreventwaterandH3PO4condensation.

2.3. Catalystcharacterisation

Theelementalanalysisof thefreshand P-poisonedcatalysts (crushedmonoliths)wasdeterminedbyanICP–AESafterLiBO2

-fusionandaciddigestionofthesamples.

The textural properties of the monolith samples previ-ously degassed at 523K for 3h were measured based on N2

adsorption–desorptionisothermsusingaMicromeriticsASAP2000 apparatus.BETsurfacearea(SBET)andtotalporevolume(Vpores)

werecalculatedusingtheBETand Barret–Joyner–Halenda(BJH) method, respectively. Pore size distributions were obtained by applyingtheBJHmodeltoN2desorptiondata.

Theexperimentalset-upusedfortheH2-TPRexperiments

com-prisesaverticalquartztubereactormountedinanelectricfurnace, partoftheassemblyoftheheat-fluxdifferentialscanning calorime-try(SetaramSensysDSC)instrument.Thegasflowintothereactor wascontrolledbyusingasystemofBronkhorstmassflow con-trollers.Priortoeachmeasurement,thecatalyst(approximately 0.1gofpowderofcrushedmonolithsamplesplaced onthe sin-teredbedofthequartztube)wasfirsttreatedwithamixtureof5% O2inArat773Kfor2h.Thetemperaturewasthendecreasedto

323Kunderthesamegasenvironment.AfterﬂushingwithonlyAr at323Kfor30min,aﬂowof1%H2/Ar(20mlmin−1)waspassed

throughthesampleat323Kfor20minandthenthetemperature inthepresenceof1%H2/Arwasraisedatarateof10Kmin−1upto

1073K.TheeffluentfromthereactorwasmonitoredusingaHiden HPR20quadrupolemassspectrometer(MS)equippedwitha capil-laryprobeconnecteddirectlytotheexitofthereactor.Theanalysis wasperformedbyrecordingtheMSsignalswithmasstocharge ratio(m/e)equalto2,18,20and32inpressureversustimemode. XPSdatawererecordedwithaThermofisherK-Alpha spectrom-eterusingnon-monochromaticAlK␣X-rayirradiation.Thepowder sampleswereaffixedonaCu-basedelectricallyconductingtape beforetheXPSanalysis.Ane-beamfloodgunwasusedforcharge compensationduringthespectralacquisition.Thebindingenergies (BE)ofallXPspectrawerecalibratedbyutilisingthereferenceC1s signallocatedat284.6eVandtheXPintensitieswerenormalised usingtheintensityoftheO1ssignalofeachXPspectrum.

2.4. Flowreactormeasurementswithmonolithcatalysts

Thereactionstudiesonthemonolithcatalystswereperformed ontheexperimentalset-updescribedabovefor thePpoisoning experiments.Approximately700mgcatalystwashcoatedonthe monolithwas usedin each experiment yielding a space veloc-ityof30,300h−1,basedonmonolithvolume.Thetotalgasflow washeldconstantat3500mlmin−1andcontrolledbyasystemof Bronkhorstmassflowcontrollers.Thewaterintheformofsteam wasintroducedintothereactor byusinga controlled evapora-tionandBronkhorstmixingsystem.Themonolithswerewrapped withquartzwooltoensurethatnogasslippedaroundthesample. Theoutletgascompositionfromthereactorflowwasmonitored andanalysedon-linewithrespecttoNO,NO2,N2O,NH3,andH2O

contentbyusing MKSMultiGas2030HSFTIRspectrometer.To maintaina constant catalytic behaviourover thecourseof the study,thecatalystsweredegreenedbyincreasingthetemperature to773KinAr;thenthesampleswerecleaned/conditionedwith 8%O2inArfor15minandthenthecatalystsweretreatedwitha

gasmixtureof400ppmNO+400ppmNH3+8%O2+5%H2Oand

balancingamountsofArfor30min.Priortoeachexperiment,the catalystswerepre-treatedat773KinArand8%O2for15min.

Table1

Flowreactormeasurementsperformedinapredeﬁnedsequenceofstepsoverthe freshandP-poisonedCu/BEAmonolithcatalysts.

Samples Reactionstudies Fresh 1.NH3storage/TPD 2.NH3-SCR 3.NOoxidation 4.NH3oxidation P–poisoned 1.NH3-SCRupto573K 2.NOoxidationupto573K 3.NH3oxidationupto773K 4.NH3-SCRupto773K 5.NH3storage/TPD 6.NOoxidationupto773K 7.NH3oxidationupto773K(rep.)

ThefollowingexperimentsoverthefreshandP-poisoned cata-lystswerecarriedout:

(a)NH3storagetestsandTPDinthepresenceofH2O–Thecatalysts wereinitiallyexposedto400ppmNH3inthepresenceof5% H2Ofor40minat423K.AfterﬂushingwithAr+5%H2Ofor 30min,thetemperaturewasraisedto773Kwitharampspeed of10Kmin−1.TheoutletNH3concentrationwasmonitoredas afunctionoftimeandthenconvertedtothecumulativeNH3 storedduringtheuptakeperiodasapercentageoftheNH3fed, byintegratingtheareaincludedbetweentheinletNH3andthe outletNH3concentrationcurve.

(b)Flowreactorstudies–Theactivitymeasurementswerecarried outat423,473,523,573,673and773K.Theresultsforeach temperature wereobtained afterthe systemhad reacheda steady-stateandthenthereactortemperaturewasincreasedto thenexttargettestreactiontemperature.Inthisway,the exper-imentswereconductedwithin423–773Kwhilethereaction mixturewascontinuouslyfedduringthewholetemperature range.TheexperimentsofNH3orNOoxidationunderlean con-ditionswereperformedwithaninletgasmixtureconsistingof 8%O2,400ppmNH3(or400ppmNO),5%H2Oandabalanceof Ar.ThereactionstudiesofSCRwithNH3wereperformedwith aninletgasmixtureconsistingof8%O2,400ppmNH3,400ppm NO,5%H2OandabalanceofAr.

Forcomparisonoftheresults,thereactionstudiespriortoand after P poisoning were performedin a predeﬁned sequence of experiments, presentedin Table 1.In the ﬁrststage, the activ-itymeasurementsofNH3storage/TPD,standardNH3-SCR,NOand

NH3oxidationwereconductedoverthefreshdegreenedcatalysts.

Then,thesamplesweresubjectedtoPpoisoningwith50/100ppm Pat573K,followedbyactivitymeasurementsbetweeneachstep ofexposurewiththedifferentconcentrationofH3PO4.Ina

sim-ilarway,theexperimentsofPpoisoningathighertemperatures (773K)wereperformedbyusinganotherfreshmonolithsample. DuetothepossibilityofformationoflooselyboundPspecieswhich caneasilyberemovedbyheatingthesampleathightemperatures, theP-poisonedsampleswereﬁrsttestedinstandardNH3-SCRand

NOoxidationupto573K(Table1).Then,theexperimentswere repeatedatvarioustemperatureswithin423–773K.

ToevaluatetheoverallSCRperformanceofthecatalysts,the outletconcentrationcurveswereusedtodeterminetheactivityper Cusite,expressedasaratiooftheamount(kmol)ofNOxreducedor

NH3convertedspeciespermolofCusitespersecond.Theresults

foreachtemperaturewereobtainedafterthesystemhadreached asteady-state.TheoutletNH3andNOxconcentrationsduringNH3

oxidationandNH3-SCRweremonitoredasafunctionoftimeand

thenconvertedtoNOxandNH3reacted,accordingtoEq.(1):

(4)

(NOin₎_and_(NHin

3)aretheNOandNH3concentrationsatthereactor

inlet;(NOout_x )and(NHout₃ )aretotalNOxandNH3concentrationsat

thereactoroutlet,respectively.

TheamountofreactedNOx(orNH3)perCusite,deﬁnedasthe

numberofNOx(NH3)molecules(kmol)convertedpermoleofCu

persecond(overthefreshandP-poisonedcatalysts)wascalculated byusingEq.(2),as:

kmolNO_x(NH3)reactedpermolCu=

NOreacted_x (NHreacted₃ )(ppm×s)×TF(ml×min−1)×10−6(ppm−1)

22414(ml×kmol−1)×60(s×min−1)×mwashcoat(g)×Cu(mol×g−1cat)

×100 (2)

where, TF (mlmin−1)is the gas ﬂow rate (3500mlmin−1) and mwashcoat(g)isthemassofthewashcoatonthemonolith.

To estimatethe degree of deactivation of thecatalysts, the reductionofNOxconversioninthereactionofNH3-SCRduetoP

poisoningwasestimatedbyusingEq.(3): ReductionofNO_xafterpoisoning

=

NOreacted_x (kmol)

Fresh−

NOreacted_x (kmol)

Aged

NOreacted_x (kmol)

Fresh

×100 (3)

NOreacted_x

Freshand

NO_xreacted

Agedaretheamounts(kmol)oftotal NOxconvertedpermolofCusitesoverthefreshandP-poisoned

catalystspersecond.

Ina similarway,thedecreasein theNH3 conversionduring

NH3-SCRwascalculatedaccordingtoEq.(3),byusingtheamounts

(kmol)ofNH3convertedoverthefreshandP-poisonedcatalysts.

3. Resultsanddiscussion

3.1. Chemicalcompositionandtexturalcharacteristicsofthe catalysts

3.1.1. ICP,BETsurfaceareaandporesizedistribution

TheICP–AESanalysisoftheCu/BEAcatalystswascarriedoutto quantifytheamountofCu,AlandSionallofthepowdersamples beforetheirwashcoating,whilePanalysiswasonlyperformedfor P-poisonedmonolithsamples.TheresultsarelistedinTable2.The composition-dependentchangesofthetexturalproperties(SBET,

Vpore)ofthefreshandP-poisonedCu/BEAcatalystswithdifferent

CuloadingsarealsosummarisedinTable2.Inaddition,BJHpore sizedistributionsfordifferentmonolithsarealsopresentedinFig.1. InthecaseofBEAion-exchangedsampleswithdifferentCu con-tent,theICP–AESanalysisindicatedthatthesynthesiswhichwas controlledbychangingtheconcentrationofCuintheion-exchange solutionhasresultedinCu/BEAcatalystswithCucontentsof4.0 and1.3wt%Cu.Inourpreviousstudy[10],elementalanalysisof thepowdercatalystsbeforetheirwashcoatingshowedthattheCu ionexchangelevelin4Cu/BEAsampleiscloseto88%.This sam-plecanberegardedasanover-exchangedsystemcomparedtothe 1.3Cu/BEAsamplewithalowerCucontentandalowerCuexchange levelof∼30%.Thisisalsoinagreementwithotherstudies[12,51]

intheliterature,wheretheCu/BEAcatalystshavebeenconsidered as“over-exchanged”whentheexchangedlevelbasedontheCu/Al ratiois morethan50%.It isworthmentioningthat,thecurrent ICP–AESanalysisshowedthatSi/Alratiotypicallyremainssimilar fortheanalysedsamples,whiletraceamountsofP(i.e. compara-bletotheinstrumentaldetectionlimit)werealsoobservedonthe freshsamples.

Conversely,PcontenttotheP-poisoned4Cu/BEAsampleswas noticeablyhigherthanthefreshcatalysts,indicatingthatthe expo-sureofthemonolithsampleswithH3PO4 inlean hydrothermal

conditionshasresultedtotheaccumulationofPinthesamples.The Pcontentdetectedforboth4Cu/BEAsamplesexposedtoH3PO4at

573K(P1)and773K(P2)was11.3and11.9%,respectively.Onthe

otherhand,onthe1.3Cu/BEAsamplewiththelowerCucontent, whichwaspoisonedunderidenticalconditionsasforthe4Cu/BEA catalyst(i.e.573K(P1),Paccumulationwasobservedtobeonly 1.4%.

Thetexturalcharacteristicsofthefresh,non-poisonedCu/BEA catalystssynthesised withdifferentCucontentshowedthatthe ion-exchanging of thezeolite withthehigher concentration of

Cu(i.e.4.0wt%Cu)resultedinaslightdecreaseofthetotal sur-faceareaandtheporevolumecomparedtothesamplewiththe lowerCucontent(i.e.1.3wt%Cu).Suchbehaviourisexpectedand indicatesthattheincorporationofCuionsoccursthrough substitu-tionofexistingNa+_cations_at_the_ion-exchange_sites,_and_proceeds

withoutsignificantocclusionofporenetwork.Ontheotherhand, aconsiderabledifferenceinthetexturalcharacteristicsofthe P-poisonedCu/BEAcatalystswasobserved(Table2).Thepoisoningby Pwasfoundtohaveasignificanteffectonthespecificsurfacearea andporevolumeofthecatalystswhichwerediminishedcompared tothefreshmonolithsamples.Theseresultsarealsoconsistent withthedatapresentedin Fig.1,whereitcanbeseenthatthe depositionofH3PO4producedasignificantchangeintheporesize

distributionforthemonolithsexposedtoP.Itisvisiblethatthe fresh,non-poisonedmonolithsampleshaveahigherporevolume andtheporesizedistributionplotcontainstwopeaks,around3.6 and5.1nm.Thesmallerporesareattributedtointercrystalline dis-tancewithintheaggregates whereasthebiggerporesarelikely tooriginatefromtheinter-aggregatedistance.Itwasfoundthat theporeswithlargerdiameterofthebimodalmesoporous struc-turewerepartiallyﬁlledafterPpoisoningofthe4Cu/BEAcatalysts, indicatingtheoccurrenceofphysicaldeactivationmostlikelydue toporeblockingandcondensation.Accordingtotheseresults,it wassuggestedthatthedepositedPmayactasimpuritiesblocking thepores.Therefore,furtherinformationregardingthepossibility forchemicaldeactivationofthecatalystswasobtainedbyH2-TPR

ofthesamples. 3.1.2. H2-TPR

TheredoxbehaviouroftheCu/BEAcatalystsafterPpoisoning wasstudiedbyTPRanalysisperformedbyrecordingtheH2

con-sumedasfunctionofthetemperatureintherangeof323–1073K.

Fig.2presentstheH2-TPRproﬁlesforfresh1.3Cu/BEAand4Cu/BEA

samplesaswellassimilarmeasurementsfor thesamesamples poisonedwithPat573and773K.

ThereductionsignalobservedintheTPRproﬁleofthe4Cu/BEA samplerevealsthreemajorfeatures at454,520and 578K.The H2-TPRofCu/BEAcatalystswithdifferentCuloadingshasbeen

thoroughlydiscussedinvariousformerstudies[12,52–54].Based onthesereports,theﬁrstprominentsignalat454KintheTPRof the4Cu/BEAsamplewasattributedtothereductionofCu2+_ions_in

Cu O Custructures,whichcanbeformedathighCuionexchange levels.It wasreported[12,52–54]thatthesedimericCuspecies observedforlargeCu-loadingscontainbridgingoxygenatomsthat canreactwithH2atcomparablylowtemperaturesthanisolated

Cu-sites.ThisisalsoingoodagreementwiththeH2-TPRcurveof

1.3Cu/BEAcatalystwiththelowerCucontent,wherethe tempera-turemaximashifttohighertemperatures(atabout671and839K) withdecreasingtheCucontent.Signiﬁcantlylowertemperature maximaareobservedforthe4Cu/BEAcatalystascomparedtothe 1.3Cu/BEAsystem,whichclearlyshowsthatthereducibilityofthe over-exchanged4Cu/BEAsampleissubstantiallyhigherthanthat ofthe1.3Cu/BEAcatalyst.Thiscanbeassociatedwiththesmaller populationofisolatedCu-sitesinthecaseof4Cu/BEAsamplewhich

(5)

Table2

Elementalcomposition,speciﬁcsurfacearea(SBET)andtotalporevolume(Vpores)ofthefreshandP-poisonedCu/BEAmonolithcatalysts.

Monolithsample TemperatureofPpoisoning(K) Elementalanalysis(%) SBET(m2g−1) Vp(cm3g−1)

Cuc _P _Alc _Sic 4Cu/BEAa _– ₄ _1.1 _1.9 ₄₂ ₁₄₆ _0.105 4Cu/BEA–P1b ₅₇₃ ₄ _11.3 _1.9 ₄₂ ₁₁₆ _0.0892 4Cu/BEA–P2b ₇₇₃ ₄ _11.9 _1.9 ₄₂ ₁₀₉ _0.0806 1.3Cu/BEAa _– _1.3 _0.92 _1.9 _41.3 ₁₅₄ _0.107 1.3Cu/BEA–P1 573 1.3 1.4 1.9 41.3 141 0.090

a_Fresh,_non_poisoned_sample.

b_P1/P2–P_poisoned_samples_after_exposure_of_the_monoliths_with_P_at₅₇₃_and₇₇₃_K,_{respectively.}

c _Cu,_Al_and_Si_content_in_the_samples_(wt%)_was_determined_by_ICP_analysis_of_Cu/BEA_powder_catalysts_without_the_monolith_and_the_binder.

4

8

12

16

0,00 0,04 0,08 0,12 0,16

4

8

12

16

0,00 0,04 0,08 0,12 0,16 4Cu/BEA - P2 4Cu/BEA - P1

BJH

Desorption

dV/dD

Pore

Volume

(c

m³

g

-1

) x

10

0 Pore

Diameter (

nm)

3.6

5.

1 A

4Cu/BEA 1.3Cu/BEA - P1 1.3Cu/BEA

B

3.

6

5.

1

Fig.1.BJHporesizedistributionofthefreshandP-poisonedat573K(P1)and773K(P2)Cu/BEAmonolithssampleswithdifferentCucontent(4and1.3wt%Cu): 4Cu/BEA–Fresh/P1/P2(A)and1.3Cu/BEA–Fresh/P1(B).

requiresalessfacile[12,52–54],two-stepreductionmechanism

[53]inwhichisolatedCu2+_ions_are_initially_reduced_to_the_Cu+

intermediateandthentometallicCuspecies.

400 600 800 1000 0 200 400 600 800 1000 1200 1.3Cu/BEA - P1 1.3Cu/BEA 4Cu/BEA - P2 4Cu/BEA - P1 4Cu/BEA 67 1

83

9

52

0 H

2

consumption (ppm)

Temperature (K

)

45

4

57

8

Fig.2.H2-TPRofthefreshandP-poisonedat573K(P1)and773K(P2)Cu/BEA

monolithssampleswithdifferentCucontent(4and1.3wt%Cu).

TheTPRsignalsat520andat578Kforthe4Cu/BEAsamplecan alsobeinterpretedviatwo differentexplanationsbasedonthe formerstudies[12,52–55]intheliterature.ItwasshowninRef.

[12,52–54]thattheseH2 consumptionpeakscouldberelatedto

thetwo-stepreductionofCu2+_to_Cu+_(i.e.₅₂₀_K_signal)_followed_by

Cu+_to_Cu0_(i.e.₅₇₈_K_signal)._{Alternatively,}_by_referring_to_another

experimentalstudy[55],itcanalsobearguedthatthesetwo differ-entTPRpeaksareindicativeofthereductionoftwodifferenttypes ofCu2+_sites_into_Cu+_species.

Fig. 2 clearly shows that the P-poisoning results in signifi-cantchanges in theTPRprofiles. The4Cu/BEA catalystsafterP poisoning are characterised withsignificantly higher reduction temperatures compared to that of the fresh 4Cu/BEA. More-over, this effectis more discerniblefor thecatalyst exposed to Pat573K.Inaddition,theTPRsignalintensitiesforP-poisoned samplesdecreaseddramaticallyincomparisonwiththefresh cat-alysts.The fraction of thereduced Cu sites (inmolg−1cat)and thetotalintegralH2 consumptionsignalsobtainedfromtheH2

-TPRresultsarepresented inTable3.Theseresultsshowedthat the total H2 consumption (6.16×10−4molg−1cat)of the fresh

4Cu/BEAcatalystcloselymatchestheCuloadinginthesame cata-lyst(6.30×10−4molg−1cat)).Suchbehavioursuggeststhatalmost 100%oftheexistingCu2+_sites_were_reduced_during_the_TPR

experi-ments.Ontheotherhand,TPRdataforthepoisonedsamplesreveal that only76–78%oftheCusitesexisting onthefreshcatalysts remainedavailableforreductionafterPpoisoning.Inotherwords, thefractionofthereducedCuspeciesoverP-poisoned4Cu/BEA cat-alystsisabout20%lowerincomparisontothatofthefreshsample, indicatingtheattenuationofthenumberofavailableCusitesfor reduction.Inasimilarway,thetotalamountofH2consumption

(6)

Table3

CalculatedparametersviaH2-TPRandsurfacecompositionsoftheanalysedfreshandP-poisonedCu/BEAmonolithsamplesviaXPSanalyses.

Samples H2-TPR XPS

TotalH2consumed (molg−1cat)x10−4

TotalCureduced(%)a _{Cu(II)/(Cu(I)}₊_Cu(0))b _Cu(II)%c _Cu/Sid _P/Sid

4Cu/BEA 6.16 97.8 1.6 62 0.04 –

4Cu/BEA–P1 4.83 76.7 3.3 77 0.05 0.03

4Cu/BEA–P2 4.95 78.7 3.3 77 0.06 0.03

1.3Cu/BEA 1.97 96.4 1.0 50 0.01 –

1.3Cu/BEA–P1 1.73 84.3 0.3 25 0.01 –

a_Percent_of_total_Cu_reduced_was_calculated_based_on_the_total_amount_of_Cu_(mol_g−1_cat)_in_the_samples_and_the_total_integral_H₂_consumption_during_H₂_-TPR. b _Relative_abundance_of_Cu(II)_species_with_respect_to_the_abundance_of_all_Cu(I)_and_Cu(0)_species

c _Percent_abundance_of_Cu(II)_species_with_respect_to_the_total_abundance_of_Cu(II),_Cu(I)_and_Cu(0)_species

d _Relative_surface_atomic_ratios_obtained_from_the_{corresponding}_integrated_XPS_signals_and_atomic_sensitivity_factors_(ASF)

coincideswiththeCucontent(2.05×10−4molg−1cat)inthis sam-ple.Althoughtheeffectismuchmoresuppressedincomparisonto thesamplewiththehigherCuloading,thetotalH2consumption

ofthe1.3Cu/BEA–P1sampledecreasedby∼10%afterPpoisoning. In the light of these observations, it can be suggested that thepoisoningbyPfollowsbothphysicalandchemical deactiva-tionpathways.It is evidentthat theexposureof themonoliths withPchemicallydeactivatestheCu/BEAcatalystsbydecreasing thenumber oftheactive Cuspecies and hencehinderingtheir redoxcapabilities. Furthermore,Ppoisoning hasa considerable effectonthemetal zeolite interactionbyproducing Cu species whicharestronglybondedtotheframeworkoxygenresultingina highertemperatureofreduction.Moreover,itisalsolikelythat P-poisoningmayalsoleadtotheformationofCu-phosphatespecies. Thisisparticularlylikely asthePsourceusedin thepoisoning experimentswasH3PO4,whichcanreadilygeneratephosphates

uponitsdepositiononthecatalystsurface.Therefore,further infor-mationregardingthenatureoftheP-containingspeciesgenerated after thepoisoning process was obtainedvia XPS experiments whichwillbediscussedinSection3.1.3.

3.1.3. XPS

TheCu2p3/2 andP2pXPspectraofthefreshand P-poisoned

Cu/BEAsamplesarepresentedinFig.3.Thissetofdatacorresponds totrituratedpowdersampleswhichincludeamixtureofwashcoat togetherwiththemonolith.ItisworthmentioningthatXPS analy-seswerealsoperformedoverthesamesetofcatalystsamplesusing differentsamplingtechniques(e.g.byscrappingthewashcoatfrom themonolithwallsorbydirectlyanalysingtheinteriorwallsofthe monolithbybreakingthemonolithchannels),whichrevealed simi-larresultsascomparedtothetrituratedsamplesdiscussedbelow.It isknown[56]thattheshake-upsatellitepositionedatc.a.943eVin theCu2p3/2spectraisanindicationofthepresenceofCu(II)species.

TwodiscerniblefeaturesofthemainCu2p3/2signalseeninFig.3A

at934.7and933.6eVcanbeattributedtoCu(II)andCu(I)/Cu(0) species,respectively.TheCu(I)statecanbequalitatively differenti-atedfromCu(0)signalfromthecorrespondingLMMAugersignal,

[56]howeverduetolowCucontentoftheanalysedsamples, acqui-sitionofareliableLMMAugersignalwasnotfeasibleinthecurrent XPSmeasurements.Thus,thelatterCu2p3/2signalat933.6eVis

tentativelyassignedtoCu(I)and/orCu(0) species.Althoughthe current XPSresultsdo not providea direct evidence for ruling outtheexistenceofCu(0)species,presumablyexistenceofsuch ahighlyreducedCustateseemsratherunlikely.Comparisonofthe spectracorrespondingtothefreshandP-poisoned4Cu/BEA cata-lystspresentedinFig.3AindicatesthattheCu2p3/2peaksforboth

4Cu/BEA–P1(P2)samplesafterPpoisoningshowsigniﬁcant asym-metrywithrespecttothatofthefreshcatalyticsystem.Itisvisible thattheCu2p3/2spectraofthe4Cu/BEA–P1(P2)catalystsconsistof

stronglypronouncedshoulderonthehigherbinderenergysideat

934.7eVofthemainpeakat933.6eVwhiletheCu2p3/2peakofthe

fresh4Cu/BEAsamplelooksmoresymmetric.Thus,itis presum-ablethatthepresenceofPbringsaboutavisiblevariationinthe populationsofCuspecieswithdifferentoxidation states. There-fore,therelativeamountofCu(II)versusCu(I)/Cu(0)specieswas estimatedwiththehelpoftheshake-upsatellitesignal,sinceCu(I) andCu(0)speciesdonotrevealthisparticularsatellitesignal.Itis possibletocalculateCu(II)/(Cu(0)+Cu(I))ratiobycomparing inte-gratedareasoftheshake-upsatelliteandthemainCu2p3/2signals.

ThismethodisdescribedcomprehensivelyintheworkofBiesinger etal.[56].Asdescribedinthisformerreport[56],forthe calcula-tionofCu(II)/(Cu(0)+Cu(I))ratio,oneneedstoknowtheratioof theintegratedpeakareasofthemainCu2psignallocatedatc.a. 933–935eVandthesatellitefeatureatc.a.943eVforapureCuO referencematerial.Inourcalculations,weusedavalueof1.9for thispurpose,whichhasbeenreportedbyBiesingeretal.[56].The resultsobtainedviathisanalysisarepresentedinTable3.

XPSdatacorrespondingtobothfreshandP-poisoned4Cu/BEA catalystsindicatetheincreasedfractionofCu(II)intheP-poisoned sampleswithrespecttothefreshsystem.It wasfoundthatthe fractionoftheCu(II)afterPpoisoningofthe4Cu/BEAcatalystis about15%higherincomparisontothatoftheP-freesample.

Ontheotherhand,thedataforthesampleswith1.3wt%Cu load-ingdemonstratesufﬁcientlylowerCu(II)%contentincomparison tothesampleswith4wt%Culoading,particularlyforP-poisoned samples.ItisimportanttonotethatCu(II)speciesareproneto reductionuponexposuretoX-rays.ReductionofCu(II)speciesdue toX-rayirradiationhasbeenreportedin formerstudies[57,58]

associated withCu-based catalytic systems.In thecurrent XPS measurements,wedidnotparticularlyfocusontheX-rayinduced reductionoftheanalysedsamples.However,itisworthmentioning thatwhentheCu2pXPspectraobtainedafter15minofX-ray expo-surewerecomparedwiththeXPspectraobtainedfromidentical setofsamplesafterlonger(i.e.2h)X-rayexposure;noapparent differencesweredetectedbetweenthesetwosetofdata.Onthe otherhand,thisobservationdoesnotexcludethepossibilityof X-rayinducedreductionofCu(II)siteswhichcouldhavetakenplace intheveryﬁrst15minoftheXPSanalysis.ThedatafromtheH2-TPR

showsthat98and96%ofthecopperinthe4Cu/BEAand1.3Cu/BEA wereintheformofCu(II)(calculationbasedonCu(II)toCu(0)inthe TPR),respectively.TheXPSdatashowsignificantlylessCu(II)and thereasoncouldbereductionofthecopperbythebeaminthefirst minutesoftheexperiments,asseenbyWilkenetal.[57].Ifthisis thecase,XPSgivesimportantinformationaboutthereducibilityof thecopperbetweenthedifferentsamples.TheXPSrevealthatthe copperspeciesin4Cu/BEAaremoredifficulttoreduceafter phos-phorousexposure,whichisinlinewiththehighertemperaturefor reductionintheTPR.Thisisnotseenforthelowloadingsample, butontheotherhandtheloadingisverylow,makingthisanalysis moredifficult.

(7)

Fig.3. Cu2p(A)andP2p(B)XPspectraofthefreshandP-poisonedat573K(P1)and773K(P2)Cu/BEAmonolithssampleswithdifferentCucontent(4and1.3wt%Cu).

WehavealsoanalysedtheP2psignalintheXPspectra(Fig.3B) for the samples given in Table 3. Phosphorous could only be detectedfortheP-poisoned 4Cu/BEAsampleswitha higherCu loading,whileP2psignalwasbelowtheinstrumentaldetection limit for the catalyst samples with the lower Cu content (i.e. 1.3Cu/BEA).ThisobservationsuggeststhatCusitesmaybe func-tioningasPOxanchoringsitesintheCu/BEAcatalyticsystem.The

P2pXPsignalsgiveninFig.3Brevealabroadfeaturelocatedat 134.4eV,whichcanbeassociatedwithphosphate,metaphosphate

[59,60],and/ordihydrogenphosphate[61]functionalities,asallof thesespeciesrevealrelativelysimilarP2pbindingenergyvalues fordifferentmetalcations.

BasedontheXPSresultscombinedwiththeH2-TPR,itis

appar-entthatPOxspecies(mostlikelyintheformofphosphates)strongly

interactwithactiveCusitesonthecatalysts,leavingasmaller num-berofaccessibleCusitesavailableforreductionviaH2-TPR.

3.2. Flowreactormeasurements

Furtherinformationregarding theeffectsinduced by chemi-caldeactivationofCu/BEAcatalystswithPwasobtainedviaﬂow reactormeasurementsonthewash-coatedmonolithsamples. 3.2.1. NH3-storageandTPD

In Fig. 4, time-dependent NH3 uptake (at 423K) and TPD

(423–773K)measurementsin thepresence of5% H2Oare

pre-sented.

ItisvisibleinFig.4AthatthepresenceofPinthe4Cu/BEA sam-plehasasigniﬁcanteffectonthebreakthroughproﬁleofammonia andonthecorrespondingNH3uptakebehaviourofthesystem

poi-sonedwith50ppmPandthenwith100ppmPat573K.TheNH3

signalduringtheexposureperiodismoresigniﬁcantlydelayedfor thefresh4Cu/BEAincomparisontothe4Cu/BEA–P1sample.On theotherhand,theNH3adsorptionexperimentsperformedover

the4Cu/BEA–P2catalysts(Fig.4B)showedNH3uptakebehaviour

similartothatofthefresh4Cu/BEAwithverysmalldeviationsin theNH3concentrationproﬁles.Suchasituationisalsovalidforthe

1.3Cu/BEAcatalystwithlowerCucontentpresentedinFig.4C. TheanalysisoftheTPDdatainFig.4forthepoisonedsamples showssimilartrends,withoutanysigniﬁcanttemperatureshiftin thedesorptionmaxima.However,theNH3 signalsintheTPDof

theP-poisoned4Cu/BEAcatalystsreveallowerdesorption inten-sities(Fig.4A).Thiseffectismuchmorestronglypronouncedin theTPDproﬁlesofthe4Cu/BEAcatalystsPpoisonedat573Kwith

50and100ppmP.BasedonthedatapresentedFig.4D,itcanbe seenthattheextentofPpoisoningincreasesmonotonicallywith theincreasingamountofH3PO4inthefeed.

TheeffectofthePpoisoningat773KontheNH3storage

abil-ity,particularlyinthecaseofthe4Cu/BEA–P2(Fig.4B)islimited incomparisontothatforthe4Cu/BEA–P1sample.Thisbehaviour can beexplained byconsideringthe proposedmechanism ofP depositioninthestudy[31]investigatingdeactivationofV-based commercialSCRcatalystsbyH3PO4.Itwasshowninthisformer

studythatH3PO4moleculesstartcondensationreactionsforming

polyphosphoricacidswhichcanbedepositedonthesurfaceat tem-peratureslowerthan773K.Oncedepositedonthecatalystouter surface,thesespecieswerefound[31]tohavehighmobilityand abilitytopenetrateandevenbetrappedintothewallsbycapillary forces.

Inthelightoftheseﬁndings,itcanbesuggestedthatPpoisoning ofthe4Cu/BEAcatalystsat773Khindersthechemical deactiva-tionofthesamplesduetoinefﬁcientcondensation/polymerisation reactionsofthedepositedH3PO4atelevatedtemperatures,

result-ing in loosely bound POx species. It is worth mentioning that

XPSandICP–AESresultsforthe4Cu/BEA–P1/P2samplesgivenin

Tables2and3revealsimilarPsurfaceatomicratiosforthesetwo samples.Thisobservationpointstothefactthatchemicalnature ofthepoisoningPOxspeciescouldhaveamorecentralrolethan

thesolesurfacecoveragesofsuchfunctionalities.Alongtheselines, althoughthechemicalstructuresofthepoisoningPOxspeciesare

likelytobedifferentonthe4Cu/BEA–P1and4Cu/BEA–P2samples, suchstructuraldifferencesseemtobeelusivetocaptureviaXPSas thesetwosamplesyieldverysimilarP2pXPspectra(Fig.3B).

TheNH3uptakebehaviourofthefreshandP-poisonedat573K

1.3Cu/BEAcatalystispresentedinFig.4C.Despitethelower tem-peratureofPexposureofthemonolithsampleat573K,theresults showedanNH3storagebehaviour,whichissimilartothatofthe

freshcatalyst.TheamountofstoredNH3on1.3Cu/BEAgradually

decreaseswithincreasingthePconcentration.However,the poi-soningprocessismuchlesspronouncedcomparedtothatforthe “overexchanged”4Cu/BEAsamplewithhigherCucontent.Thus,it canbearguedthattheprocessofPaccumulationonthesurface hasoccurredpreferentiallyonthesocalled“overexchanged”Cu activesites(whichareabundantonthe4Cu/BEAsample).Thisis inagreementwiththecurrentICP–AESdatawhichshowedthat thePcontentisonly1.4%P,althoughtheconditionsofpoisoning wereidenticalforthe4Cu/BEA–P1catalyst.Inaddition,theH2-TPR

(8)

Fig.4.EvolutionofNH3concentrationasfunctionofthetimeduringNH3uptake(at423K)andTPD(423–773K)inthepresenceof5%H2OoverthefreshandP-poisonedat

573K(P1)and773K(P2)Cu/BEAcatalystsafterexposureofthesampleswith50and100ppmP:4Cu/BEA(AandB)and1.3Cu/BEA(C).Theinset(D)presentstheestimated amountsofNH3storedonthesurface(mmol)pergramcatalystsasfunctionofthetemperatureofPpoisoning.

P-poisoned1.3Cu/BEAcatalystcomparedtothechangesobserved forthe4Cu/BEAsamplewithhigherCucontentafterpoisoning. 3.2.2. NH3andNOoxidation

Fig.5showstheevolutionofNH3concentrationasafunctionof

thetimeduringNH3oxidation(423–773K)overthefreshand

P-poisonedCu/BEAcatalysts.ItcanbeseeninFig.5thatbothfreshand P-poisonedCu/BEAcatalystsexhibittypicalproﬁlesconsistentwith similarNH3oxidationstudiesreportedintheliterature[5,10,62].

Fig.5.EvolutionofNH3concentrationasfunctionofthetimeduringNH3oxidation

(423–773K)overthefreshandP-poisonedat573K(P1)and773K(P2)Cu/BEA catalystsafterexposureofthesampleswith50and100ppmP:4Cu/BEA(AandB) and1.3Cu/BEA(C).Thereactionstudieswereperformedinthepresenceof400ppm NH3,8%O2and5%H2O.

Accordingly, upon NH3 admission at 423K, NH3 breakthrough

appeared,withasteadyincreaseintheexitNH3concentrationover

time,graduallyconvergingtotheinletNH3concentrationlevelof

400ppm.TheCu/BEAcatalystsexposedtoPexhibitanNH3storage

at423Ksimilartothedatadiscussedintheprevioussection, indi-catingadecreasedNH3adsorptionabilitycomparedwiththefresh

Cu/BEAsamples.

Increasingthetemperatureupto573Kdoesnotresultinany signiﬁcantdifferences intheNH3 oxidationbehaviourinanyof

theanalysedsamples.Duringthetransitionsfromalow temper-aturetoa highertemperature,ammoniadesorptionpeakswere alsoobserved,howeverthesepeaksarenotfullyshowninFig.5,in ordertopresenttheoxidationbehaviourinamorevisiblemanner. InFig.5,thelaterstagesoftheoxidationattemperatures≥573K showsigniﬁcantdissimilarities.NH3oxidationoccursoverthe

non-poisoned4Cu/BEAsample(Fig.5A)intherangeof573–773Kwitha maximumconversionat773K,whereNH3iscompletelyoxidised.

Theoverallprocesscanbedescribedinlinewiththeprevious liter-ature[5,10,17,62]andnegligibleamountsofNOxandN2Oisformed

(datanotshown),resultinginthatammoniaismostlyoxidisedto N2,accordingtoEq.(4),asfollows:

4NH3+3O2→2N2+6H2O (4)

Ontheotherhand,theresultsinFig.5AandCclearlyshowthat theNH3oxidationat673and773Kissubstantiallyloweroverthe

fresh1.3Cu/BEAcatalyst.Thisresultisingoodagreementwithour previousstudy[10],whereitwasfoundthattheNH3 oxidation

rateperCusiteissigniﬁcantlyhigherfortheover-exchangedCu samples.

Incontrasttothefreshcatalysts,theresultsinFig.5AandC (cor-respondingtothe4Cu/BEAand1.3Cu/BEAsamples)obtainedafter Ppoisoningat573KshowedaclearpoisoningoftheNH3oxidation.

Inasimilarway,theNH3oxidationoverthe4Cu/BEAcatalysts

poi-sonedat773Kwith50and100ppmPshowedaparalleltrendof progressiveNH3oxidationdeterioration(Fig.5B)withincreasingP

(9)

0 100 200 300 400 Total NO x (N O+ NO 2 ) conce. (ppm)

A

NO in NH 3 conce. (ppm) 423 K 473 K 523 K 573 K 673 K 773 K 0 100 200 300 400 NH3 in

B

423 K 473 K 523 K 573 K 673 K 773 K 0 100 200 300 400 0 10 20 30 40 NO 2 con ce. (ppm) Time (min)

C

0 100 200 300 400 0 10 20 30 40 Cu/BEA Cu/BEA-P1-50 ppm Cu/BEA-P1-100 ppm N2 O conce. (ppm) Time (min)

D

Fig.6.EvolutionoftotalNOx(A),NH3(B),NO2(C)andN2O(D)oftheoutletgascompositionduringstandardNH3-SCRinthetemperaturerangeof423–773Koverthefresh

andP-poisonedat573K4Cu/BEAcatalystafterexposureofthesamplewith50and100ppmP.Thereactionstudieswereperformedinthepresenceof400ppmNO(NH3),

8%O2and5%H2O.

comparisontothatobservedforthe4Cu/BEA–P1catalysts.Hence, theresultsclearlyshowthatbothparametersofpoisoning (temper-atureandPconcentration)areinﬂuentialinthestorageprocessof NH3anditsoxidation.Thelowertemperatureofpoisoningat573K

andthepresenceofPinhigherconcentrations(100ppmP)result inthelargestdecreaseintheNH3oxidationoverCu/BEAcatalysts.

ThiswasexplainedbyapartialeliminationoftheNH3adsorption

sitesduetoPdepositionontheCuactivesites.

Inaddition,therepeatedexperimentsofNH3oxidationshowed

thattheloweredNH3 oxidationconversionover thePpoisoned

catalystsisbecomingevenfurtherreducedincomparisontothat observedaftertheﬁrstNH3oxidationexperiment.Thiscanbevery

clearlyseenespeciallyforthe4Cu/BEA–P1sample(Fig.5A)when thetemperatureofthereactionisincreasedto673K.Thisresult canbeexplainedbyconsideringthatthehightemperature treat-mentofthePpoisonedcatalystsintheexperiments(standardSCR, NOoxidation)beforerepeatingthesecondtestofNH3 oxidation

(seeTable1)hasresultedinthemigrationofcondensed H3PO4

fromtheporesofthecatalyststothesurfaceleadingtoadditional PdepositionandfurthereliminationofactiveCusites.

ThestudiescarriedoutovertheP-poisonedCu/BEAcatalysts regardingtheirperformanceinthereactionofNOoxidation(data notpresented)showedthattheNOconversiontoNO2isslightly

lowerthanthatobservedforthefreshsamples.Similartothe stud-iesofNH3oxidation,thisbecomesmoreobviouswithanincrease

ofthetemperatureto673and773KwhereNOoxidationprocessis moresuppressedforthepoisonedcatalystswhilethefreshCu/BEA catalystswerestillabletokeephigherNOoxidationactivity.

3.2.3. NOxreductionperformanceinthereactionofstandard

NH3-SCR

Fig.6presentstheconcentrationversustimecurvesforNOx(A)

andNH3(B)alongwiththeNO2(C)andN2O(D)intheoutletstream

whichwereusedtodeterminetheactivityofthefresh4Cu/BEA catalystinthereactionofstandardNH3-SCRinthetemperature

rangeof423–773KanditsdeactivationcausedbyPpoisoning with50and 100ppmPat573K.Inaddition,Fig.7displaysthe percentdecreaseintheNOx(Fig.7A)andNH3(Fig.7B)conversion

afterPpoisoningasafunctionofthetemperatureofthereaction. ThesevalueswerecalculatedbyusingtheamountofNOxreduced

andNH3convertedpermolofCusitespersecondoverthefresh

andP-poisonedcatalysts,asdescribedinSection2.Thecalculations werecarriedoutfortwodifferent4Cu/BEAcatalystsexposedto phosphorous(50and100ppmP)at573Kand773K.

ConcerningtheresultsgiveninFig.6,thefresh4Cu/BEA cata-lystexhibitstypicalNOxandNH3proﬁles(blackcurves)consistent

withtheresultsreportedinourpreviousstudy[10]focussedon theeffectofCuloadingontheSCRoperationoftheCu/BEA cata-lysts.Accordingly,uponNOandNH3admissiontotheoxygenrich

atmosphereat423K,theNOxbecomesimmediatelydetectableand

reachesasteadystatelevelofabout260ppm(Fig.6A).Ontheother hand,theexitNH3concentration(Fig.6B)steadilyincreaseswith

time, approachinga concentrationlevel ofabout260ppm after approximately40minwhere thesaturationofthesample with NH3isalmostcompletelyachieved.Afurtherincreaseinthe

reac-tiontemperatureintherangeof473–573Kincreasestheactivityof the4Cu/BEAcatalystforNOxreductionwithamaximumNOx

con-versionat573K.Theanalysisregardingtheproductsexitingthe reactoralsoshowedthattheoverallreactionhasresultedmainly inproductionofN2 (estimatedbasedonmeasuredNO,NO2and

N2O)accompaniedwiththeformationofsmallquantitiesofN2O.

TheconcentrationproﬁlesdemonstratingthechangesintheNOx

andNH3conversionwithincreasingthetemperatureupto573K

showedthatthereductionofNOoccursbyconsuming approxi-matelyequimolecularamountsofNH3 andNO,accordingtoEq.

(5),asfollows:

4NH3+4NO+O2→4N2+6H2O (5)

Theanalysisofthedataforthefresh4Cu/BEAcatalystathigher temperaturesrevealedthattheNOxconversionstartedtodecrease

withincreasingthetemperatureto673and773KwhiletheNH3

conversion shows continuousincrease where a maximum NH3

conversion(100%)isachieved(seeFig.6AandB).Thisbehaviour was previouslyexplained in theliterature [5,18,63–66], by the increasedNH3oxidationathighertemperatures.Furthermore,in

ourpreviousstudy[10],itwasshownthattheoxidationrateofNH3

perCusiteoccursfasterovertheCu/BEAcatalystwiththehigher Culoading(4wt%Cu)withrespecttothelowerCucontent(1.3wt% Cu).

(10)

400 500 600 700 0 5 10 15 20 25 30 35 400 500 600 700 800 0 5 10 15 20 25 30 35

50 ppm P at

573 K

10 0 ppm P at

573 K

50 ppm P at 773 K

10 0 ppm P at

773 K

Temperature (K)

Reduction of N

O

x

conversion

after

P

poi

soning

(%)

Temperature (K)

A

Reduction

of

NH

3

conversion

after

P

poi

soning (%)

B

Fig.7.ReductionofNOx(A)andNH3conversion(B)asfunctionofthetemperature(423–673K)ofthereactionofNH3-SCRover4Cu/BEAcatalystafterexposureofthe

samplewithPat573and773K.

Onthebasisoftheseresults,itisapparentthattheP-poisoning ofthe4Cu/BEAcatalystleadstoavisiblechangeinthe concen-trationproﬁlesoftotalNOx,NH3,NO2andN2O(Fig.6).TheNOx

removalperformancein comparisontothecorrespondingfresh samplewasdecreased,duetopoisoning.Thiscanbeclearlyseen inthewholetemperaturerangeof473–673Kforboth 4Cu/BEA samplesexposedto50and100ppmPat573K.Ourcalculations showedamaximumdeactivation(∼35%)ofthe4Cu/BEAcatalysts exposedtoPat573K(Fig.7AandB)whenthereactionofNH3-SCR

wasperformedatthelowesttemperature(423K).Thisdrasticloss ofactivityatthistemperatureislikelyrelatedtothechangesinthe redox-propertiesofthe4Cu/BEA–P1catalyst,asdiscussedearlier, whichcouldbeduetoblockingofactivesites.

Previousstudiesintheliterature[52,67,68]showedthatthe abilitytoundergoredoxCu2+_↔_Cu+_cycle_is_important_for_the_SCR

activityandthattheredox-activeCusitesareinvolvedinthe kineti-callyrelevantstepoftheSCRreaction.Thus,itcanbeexpectedthat theformationofphosphatespeciesonthesurfaceprobably pro-hibitstheCuactivesitestoparticipateintheCu2+_↔_Cu+_redox_cycle

duringtheSCRreaction.Further,theanalysispresentedinFig.7

showedthat thedeactivationofthe4Cu/BEAcatalystdecreases withincreasingthe SCRreactiontemperature upto 673K. The decreaseintheNOxandNH3 conversionafterPpoisoningofthe

4Cu/BEAcatalyst(poisonedat573K)isabout15%at473K,and thedeactivationslowlydecreasestoabout10%at573K. Consid-eringthattheoverallprocessofreductionofCuspeciesoverthe 4Cu/BEA–P1catalystisshiftedtowardshighertemperatureregion, itcanbesuggestedthatthispartiallyrecoversthelossof activ-itycausedbyPwithincreasingthereactiontemperature.Another possibleexplanationcouldbedesorption/evaporationofthe con-densedH3PO4acidcausingphysicalblockageoftheporestructure

atincreasingthereactiontemperatures.

Thisbehaviour was also observed for the4Cu/BEA catalyst, P poisoned at 773K (the data regarding the SCR performance of the sample are not presented) although the deactivation (Fig.7)wasfoundtoberatherlimitedincomparisontothatforthe 4Cu/BEA–P1.Inparticular,thiscanbeclearlyseen(Fig.7)whenthe SCRmeasurementswereconductedat423Kforthe4Cu/BEA–P1 catalyst.The 4Cu/BEA–P1catalyst hasa maximumdeactivation of∼35%,whilethe4Cu/BEA–P2samplelostonlyabout16%ofits activity.Asitwasdiscussedabove,thehighertemperatureofP poisoningofthe4Cu/BEA catalystat773Kcompared tothat at 573Klimits(tosomeextent)theeffectofchemicaldeactivation ofthesamplesandPdepositionontheactiveCusites.Inanother work,itwasshown[31]thatoncePispresentinthegasphase, reactionswithO2andH2OmaythenformH3PO4,whichmaythen

start condensationreactionsand lead totheformation of ultra ﬁneparticles.Inparticular,condensationofthesespecieshasbeen estimated[31]tohappenattemperatureslowerthan773K.

Anotherimportantaspectregardingthecatalyticbehaviourof thestudiedfreshandP-poisoned4Cu/BEAsamplesisthedifference betweentheNOxconversionswithincreasingthetemperatureto

773K.ItcanbeseeninFig.6A,thatNOxreductionprocessbyNH3

occurredonthe4Cu/BEA–P1catalystswithahigherNOxconversion

andalowerconcentrationofgaseousNOxspeciesexitingthe

reac-torcomparedtothefresh4Cu/BEAsample.Inaddition,unreacted NH3,whichisalsocommonlyreferredtoasNH3slip,wasdetected

at673and773KfortheP-poisonedcatalysts(about15ppm).These resultswereingoodagreementwiththedatareportedinSection

3.2.2andwerefoundtooriginatemostlyfromthelowerselectivity towardsNH3oxidation(Fig.5A).

Inthelightoftheﬁndings,itcanbearguedthatthePpoisoning followsbothphysicalandchemicaldeactivationandPchemically deactivatesCu/BEAcatalystsbychangingtheirredoxproperties. Furthermore,PdepositionoccursmainlyontheactiveCuspecies responsibleforthecatalyticreductionofNOxbyNH3.Itis

possi-blethattheporecondensationofH3PO4incombinationwithpore

blockingistheprevailingmechanisminthebeginningofthe pro-cessofPdeposition.However,oncedeposited,Pspeciescanalso migrateonthesurfaceandpartiallycovertheactiveCusites.In addition,itcanbesuggestedthat theaccumulatedPactsasan effectivepoisoninducingchemicaldeactivationbyreducingthe numberoftheactivesitesthanasimpurityblockingtheporesof thecatalysts.

FurthertheactivitymeasurementsinthereactionofNH3-SCR

were alsoconducted over thefresh and P-poisoned 1.3Cu/BEA catalystswithsigniﬁcantlylowerCucontent(1.3wt%Cu).These experimentsarepresentedinFig.8.Asdescribedearlier,the mono-lithsampleswereexposedtoPwith50and100ppmPat573Kby changingtheconcentrationofH3PO4inthefeed.Fromtheresults

giveninFig.8AandB,showingtheevolutionoftotalNOxandNH3

concentrationproﬁlesinthetemperaturerangeof423–773K,it canbeseenthatPpoisoningdidnotresultinanysigniﬁcant deac-tivationofthesampleevenafterexposureofthemonolithwith 100ppmPat573K.Theonlynoteworthyindicationregardingthe effectofPcanbeseenwhenthetemperatureoftheSCRreaction wasincreasedto773KatwhichahigherNOxreductionactivity

thanthefreshcatalyticsystemwasobserved.

Basedonthedatadiscussedsofar,itcanbearguedthatP accu-mulationonthesurfacewithchemicaldeactivationoccurs prefer-entiallyonthesocalled“overexchanged”Cuactivesiteswhichare abundantinthe4Cu/BEAsamplewiththehigherCuloading.This

(11)

0 100 200 300 400 Total NO x (N O+N O2 ) conce. (ppm)

A

NO in NH 3 conce. (ppm) 423 K 473 K 523 K 573 K 673 K 773 K 0 100 200 300 400 NH3 in

B

423 K 473 K 523 K 573 K 673 K 773 K 0 100 200 300 400 0 2 4 6 8 10 NO 2 conce. (ppm ) Time (min)

C

0 100 200 300 400 0 2 4 6 8 10 Cu/BEA Cu/BEA-P1-50 ppm Cu/BEA-P1-100 ppm N2 O conce. (ppm ) Time (min)

D

Fig.8.EvolutionoftotalNOx(A),NH3(B),NO2(C)andN2O(D)oftheoutletgascompositionduringstandardNH3-SCRinthetemperaturerangeof423–773Koverthefresh

andP-poisonedat573K1.3Cu/BEAcatalystafterexposureofthesamplewith50and100ppmP.Thereactionstudieswereperformedinthepresenceof400ppmNO(NH3),

8%O2and5%H2O.

argumentwasconﬁrmedbycalculatingtheratiooftheamountof NOx(inkmol)reducedorNH3 convertedpermolofCusitesper

second.Thecalculationswereperformedforvarioustemperatures

andtheresultsareplottedinFig.9.Itshouldbenotedthatthisis notarate,sincetheconversionishighandtheplugﬂowbehaviour mustbeconsideredforratecalculations.ThedatainFig.9givesa

0,0

0,5

1,0

1,5

2,0

2,5

0,0

0,5

1,0

1,5

2,0

2,5

NO

x

reduced

(k

mol/mol

Cu*s

)

A

4Cu/BEA

Cu/BEA Cu/BEA - P1, 50 ppm P Cu/BEA - P1, 100 ppm P

NH

3

converted

(kmo

l/

m

o

l Cu*s

)

B

400

50

0

60

0

70

0

80

0

1

2

3

4

5

6

7

400

50

0

60

0

70

0

800

0

1

2

3

4

5

6

7 NO

x

redu

ce

d (

km

ol

/mol Cu*s)

Temperature (K)

C

1.3Cu/BE

A

NH

3

converte

d

(kmo

l/m

ol Cu*s

)

D

Fig.9. EstimatedamountsofNOxreducedandNH3converted(kmol)permolCuactivesitespersecondonthesurfaceduringstandardNH3-SCRinthetemperaturerange

(12)

measureofhowtheconversionpersiteischangedindifferent con-ditions.However,theamountsofreducedNOxorconvertedNH3

speciespermolofCusitesaremoresigniﬁcantlydecreasedinthe caseofthe4Cu/BEAsamplewiththehigherCucontentafterP poi-soningcomparedto1.3Cu/BEA.Theexpectedincreaseintheratioof thereducedNOxpermolofCusitesat773KfortheP-poisoned

cat-alystswithrespecttothecorrespondingfreshsamples,asindicated inFig.9,canbeattributedtothehigheramountsofNH3available

forthe SCRdue tothelowered selectivitytowardsNH3

oxida-tion.Thereasonforthisisthatthe“over-exchanged”Cusitesthat aremainlyresponsibleforammoniaoxidation,aremoreseverely poisoned.

4. Conclusions

TheeffectsinducedbyPoverCu/BEANH3-SCRcatalystswith

differentCuloadings(4and1.3wt%Cu)werestudiedasa func-tionofthetemperatureofpoisoningandPconcentrationinthe feed.TosimulatePpoisoninginleanhydrothermalconditions,the monolithcatalystswereexposedtodifferentconcentrationsofPby controlledevaporationofH3PO4,inthepresenceof8%O2and5%

H2O.TheprocedurewasdevelopedtocomparetheeffectsofP

poi-soning(50and100ppmP)attwodifferenttemperatures:573and 773K.The reactionstudiesinvolvingNH3-storage/TPD, NH3/NO

oxidationandstandardNH3-SCRwereperformedinﬂowreactor

experimentsintherangeof423–773K.Inaddition,acombination ofdifferentcharacterisationtechniques(ICP–AES,BETsurfacearea measurements,poresizedistribution,H2-TPRandXPS)wasapplied

toprovideusefulinformationregardingthemechanismofP deacti-vationofthecatalysts.Basedonthesestudies,themainconclusions aresummarised,asfollows:

(a)ThepoisoningoftheCu/BEAcatalystsbyPfollowsbothphysical and chemical deactivation. It wasfoundthat the pore con-densationofH3PO4 incombinationwithporeblockingisthe

mechanismoftheprocessofPdeposition,indicatingthe occur-renceofphysicaldeactivation.However,themeasuredoverall deactivationwasrelatedmostlytooccurduetochemical deac-tivationbyreducingthenumberoftheactiveCusitesandhence theredoxpropertiesofCu/BEAcatalysts.

(b)ItwasfoundthattheprocessofPaccumulationonthesurface occurspreferentiallyonthesocalled“overexchanged”Cuactive siteswiththeformationofphosphatespecies.Thehigherextent ofdeactivationofthe4Cu/BEAcatalystthanthatforthesample withlowerCucontent(1.3wt%Cu)wasexplainedbythe pres-enceof“overexchanged”Cuactivesiteswhichareabundanton the4Cu/BEAsample.

(c)ThePpoisoningwasfoundtohaveamoresevereeffectwhen conducted at 573K compared to that at 773K. The results clearly showed that theP poisoning at lowertemperatures (573K)hasamoresigniﬁcantnegativeeffectontheNH3uptake

behaviour ofthe Cu/BEA catalystsdue toa partial elimina-tionoftheNH3 adsorptionsitesonthesurface. Inaddition,

the NH3 oxidation was loweredand alsoa decrease in the

NOxremovalperformanceincomparisontothe

correspond-ingfreshsampleoverthetemperaturerangeof473–673Kwas observed.

(d)Amaximumdeactivation(ofabout35%)ofthe4Cu/BEAcatalyst exposedtoPat573Kwasfoundtooccurwhenthereactionof NH3-SCRwasperformedatthelowesttemperature(423K).On

theotherhand,at673Knosigniﬁcantdeactivationwasfound and evenat thehighertemperature (773K)theNOx

reduc-tionperformanceoftheP-poisonedCu/BEAwasincreased.The reasonforthiswasfoundtooriginatemostlyfromthelower selectivitytowardsNH3oxidation.

Acknowledgment

ThisworkhasbeenperformedattheCompetenceCentrefor CatalysisincollaborationwithCombustionEngineResearchCentre andBilkentUniversityinTurkey.Wewouldliketoacknowledgethe Swedishfoundationforstrategicresearch(F06-0006)andChalmers InitiativeTransport,forfunding.

References

[1]S.Brandenberger,O.Kröcher,A.Tissler,R.Althoff,Catal.Rev.Sci.Eng.50(2008) 492–531.

[2]H.Bosch,F.Janssen,Catal.Today2(1988)369–379. [3]P.Forzatti,Catal.Today62(2000)51–65.

[4]E.Tronconi,I.Nova,C.Ciardelli,D.Chatterjee,M.Weibel,J.Catal.245(2007) 1–10.

[5]H.Sjövall,L.Olsson,E.Fridell,R.J.Blint,Appl.Catal.B:Environ.64(2006) 180–188.

[6]R.M.Heck,Catal.Today53(1999)519–523.

[7]C.Ciardelli,I.Nova,E.Tronconi,D.Chatterjee,B.Bandl-Konrad,M.Weibel,B. Krutzsch,Appl.Catal.B:Environ.70(2007)80–90.

[8]H.Sjövall,R.J.Blint,L.Olsson,Appl.Catal.B:Environ.92(2009)138–153. [9]S.Shwan,J.Jansson,J.Korsgren,L.Olsson,M.Skoglundh,Catal.Today197

(2012)24–37.

[10]O.Mihai,C.R.Widyastuti,S.,Andonova,K.,Kamasamudram,J.,Li,S.,Joshi,N.W., Currier,A.,Yezeretz,L.Olsson,submitted.

[11]R.Nedyalkova,S.Shwan,M.Skoglundh,L.Olsson,Appl.Catal.B:Environ. 138–139(2013)373–380.

[12]J.H.Kwak,D.Tran,S.D.Burton,J.Szanyi,J.H.Lee,C.H.F.Peden,J.Catal.287(2012) 203–209.

[13]M.Colombo,I.Nova,E.Tronconi,Appl.Catal.B:Environ.111–112(2012) 433–444.

[14]M.Wallin,C.-J.Karlsson,M.Skoglundh,A.Palmqvist,J. Catal.218(2010) 354–364.

[15]J.H.Kwak,R.G.Tonkyn,D.H.Kim,J.Szanyi,ChH.F.Peden,J.Catal.275(2010) 187–190.

[16]B.Modén,J.M.Donohue,W.E.Cormier,H.X.Li,Stud.Surf.Sci.Catal.174(2008) 1219–1222.

[17]M.Colombo,I.Nova,E.Tronconi,Catal.Today151(2010)223–230. [18]M.Colombo,I.Nova,E.Tronconi,Catal.Today197(2012)243–255. [19]M.Iwamoto,H.Yahiro,K.Tanda,N.Mizuno,Y.Mine,S.Kagawa,J.Phys.Chem.

95(1991)3727–3730.

[20]M.H. Groothaert,J.A.vanBokhoven,A.A.Battiston,B.M.Weckhuysen,R.A. Schoonheydt,J.Am.Chem.Soc.125(2003)7629–7640.

[21]B.Modén,P.DaCosta,B.Fonfé,D.K.Lee,E.Iglesia,J.Catal.209(2002)75–86. [22]M.Y.Kustova,S.B.Rasmussen,A.L.Kustov,C.H.Christensen,Appl.Catal.B:

Environ.67(2006)60–67.

[23]D.Nicosia,I.Czekaj,O.Kröcher,Appl.Catal.B:Environ.77(2008)228–236. [24]O.Kröcher,M.Elsener,Appl.Catal.B:Environ.77(2008)215–227.

[25]V.Kröger,T.Kanerva,U.Lassi,K.Rahkamaa-Tolonen,M.Vippola,R.L.Keiski, Top.Catal.45(2007)153–157.

[26]V.Kröger,M.Hietikko,U.Lassi,J.Ahola,K.Kallinen,R.Laitinen,R.L.Keiski,Top. Catal.30/31(2004)469–473.

[27]V.Kröger,M.Hietikko,D.Angove,D.French,U.Lassi,A.Suopanki,R.Laitinen, R.L.Keiski,Top.Catal.42/43(2007)409–413.

[28]V.Kröger,T.Kanerva,U.Lassi,K.Rahkamaa-Tolonen,T.Lepistö,R.L.Keiski,Top. Catal.42/43(2007)433–436.

[29]T.Kanerva,V.Kröger,K.Rahkamaa-Tolonen,M.Vippola,T.Lepistö,R.L.Keiski, Top.Catal.45(2007)137–142.

[30]A.Williams,J.,Burton,R.L.,McCormick,T.,Toops,A.A.,Wereszczak,E.E.,Fo, M.J.,Lance,G.,Cavataio,D.,Dobson,J.,Warner,K.,Nguyen,D.W.Brookshear, SAEInternational2013-01-0513.

[31]F.Castellino,S.B.Rasmussen,A.D.Jensen,J.E.Johnsson,R.Fehrmann,Appl.Catal. B:Environ.83(2008)110–122.

[32]M.Klimczak,P.Kern,T.Heinzelmann,M.Lucas,P.Claus,Appl.Catal.B:Environ. 95(2010)39–47.

[33]D.R.Liu,J.-S.Park,Appl.Catal.B:Environ.2(1993)49–70.

[34]M.J.Rokosz,A.E.Chen,C.K.Lowe-Ma,A.V.Kucherov,D.Benson,M.C.Paputa Peck,R.W.McCabe,Appl.Catal.B:Environ.33(2001)205–215.

[35]S.A.Culley,T.F.,McDonnell,D.J.,Ball,C.W.,Kirby,S.W.Hawes(1996)SAE Tech-nicalPaperSeries961898:13-21.

[36]D.D.Beck,J.W.Sommers,C.L.DiMaggio,Appl.Catal.B:Environ.11(1997) 257–272.

[37]D.E.Angove,N.W.Cant,Catal.Today63(2000)371–378.

[38]C.C.Webb,G.J.J.Bartley,B.B.Bykowski,G.Fransworth,M.Riley,JSAE20030269 (2003)1–12.

[39]G.C.Joy,F.S.,Molinaro,E.H.HomeierEH.(1985)SAETechnicalPaperSeries 852099:53-64.

[40]R.G.Silver,M.O.Stefanick,B.I.Todd,Catal.Today136(2008)28–33. [41]D.Nicosia,M.Elsener,O.Kröcher,P.Jansohn,Top.Catal.42-43(2007)333–336. [42]J.P. Chen, M.A.Buzanowski, R.T. Yang, J. AirWaste Manage. 40 (1990)

1403–1409.

(13)

[44]J.Beck,R.Muller,J.Brandenstein,B.Matschenko,J.Matschke,S.Unterberger, K.R.G.Hein,Fuel84(2005)1911–1919.

[45]J.Blanco,P.Avila,C.Barthelemy,A.Bahamonde,J.A.Odriozola,J.F.Garciadela Banda,H.Heinemann,Appl.Catal.B:Environ.55(1989)151–164.

[46]V.Kröger,U.Lassi,K.Kynkäänniemi,A.Suopanki,R.L.Keiski,Chem.Eng.J.120 (2006)113–118.

[47]R.Silver,M.Stefanick,B.Todd,Catal.Today136(2008)28–33.

[48]P.Kern,M.Klimczak,T.Heinzelmann,M.Lucas,P.Claus,Appl.Catal.B:Environ. 95(2010)48–56.

[49]S.Andonova,V.Marchionni,M.Borelli,R.Nedyalkova,L.Lietti,L.Olsson,Appl. Catal.B:Environ.132–133(2013)266–281.

[50]S.Andonova,V.Marchionni,L.Lietti,L.Olsson,Top.Catal.56(2013)68–74. [51]G.Centi,S.Perathoner,Appl.Catal.A:Gen.132(1995)179–259.

[52]B.Moden,J.M.Donohue,W.E.Cormier,H.-X.Li,Zeolitesandrelatedmaterials: trends,targetsandchallenges,in:A.Gédéon,F.Massiani,F.Babonneau(Eds.), Proceedingsof4thInternationalFEZAConference,Elsevier,2008.

[53]P.DaCosta,B.Modén,G.D.Meitzner,D.K.Lee,E.Iglesia,Phys.Chem.Chem. Phys.4(2002)4590–4601.

[54]R.Bulánek,B.Wichterlová,Z.Sobalík,J.Tich ´y,Appl.Catal.B:Environ.31(2001) 13–25.

[55]J.Li,N.Wilken,K.Kamasamudram,N.W.Currier,L.Olsson,A.Yezerets,Top. Catal.56(2013)201–204.

[56]M.C.Biesinger,L.W.M.Lau,A.R.Gerson,R.S.C.Smart,Appl.Surf.Sci.257(2010) 887–898.

[57]N.Wilken,R.Nedylkova,K.Kamasamudram,J.Li,N.W.Currier,R.Vedaiyan,A. Yezerets,L.Olsson,Top.Catal.56(2013)317–322.

[58]W.Grünert,N.W.Hayes,R.W.Joyner,E.S.Shpiro,M.R.H.Siddiqui,G.N.Baeva,J. Phys.Chem.98(1994)10832–10846.

[59]J.F.Moulder,W.F.Stickle,P.E.Sobol,K.D.Bomben,in:J.Chastain(Ed.),Handbook ofX-RayPhotoelectronSpectroscopy,Perkin-Elmer,EdenPrairie,Minnesota, 1992.

[60]Y.Barbaux,M.Dekiouk,D.LeMaguer,L.Gengembre,D.Huchette,J.Grimblot, Appl.Catal.A:Gen.90(1992)51–60.

[61]P.-H.Lo,W.-T.Tsai,J.-T.Lee,M.-P.Hung,Surf.Coat.Technol.67(1994)27–34. [62]O.Kröcher,M.Devadas,M.Elsener,A.Wokaun,N.Söger,M.Pfeifer,Y.Demel,

L.Mussmann,Appl.Catal.B:Environ.66(2006)208.

[63]T.Komatsu,M.Nunokawa,I.S.Moon,T.Takahara,S.Namba,T.Yashima,J.Catal. 148(1994)427–437.

[64]H.Sjöval,L.Olsson,R.J.Blint,J.Phys.Chem.C113(2009)1393–1405. [65]A.Grossale,I.Nova,E.Tronconi,Catal.Today136(2008)18–27.

[66]N.Wilken,K.Wijayanti,K.Kamasamudram,N.W.Currier,R.Vedaiyan,A. Yez-erets,L.Olsson,Appl.Catal.B:Environ.111–112(2012)58–66.

[67]S.Kieger,G.Delahay,B.Coq,B.Neveu,J.Catal.183(1999)267–280. [68]G.Delahay,B.Coq,S.Kieger,B.Neveu,Catal.Today54(1999)431–438.