Dry CH reforming on Co/Al of O catalysts reduced at differenttemperatures Catalysis Today

ContentslistsavailableatScienceDirect

Catalysis Today

jo u rn al h om ep a g e :w w w . e l s e v i e r . c o m / l o c a t e / c a t t o d

Dry reforming of CH ₄ on Co/Al ₂ O ₃ catalysts reduced at different temperatures

É. Horváth

, K. Baán

, E. Varga

, A. Oszkó

, Á. Vágó

, M. Tör ˝o

, A. Erd ˝ohelyi

aDepartmentofPhysicalChemistryandMaterialsScience,UniversityofSzeged,Aradivértanúktere1.,SzegedH-6720,Hungary

bExploration&BusinessDevelopment,R&D,MOLPlc,Hungary

a r t i c l e i n f o

Articlehistory:

Received30October2015

Receivedinrevisedform1April2016 Accepted11April2016

Availableonline30April2016

Keywords:

DryreformingofCH4

Co/Al2O3catalyst XPSofCo/Al2O3

Surfacecarbon

a b s t r a c t

Thereformingofmethanewithcarbondioxidehasbeeninvestigatedat773Kon10%Co/Al2O3reducedat differenttemperaturesupto1173K.ThecatalystswerecharacterizedbyBET,TPR,XRDandXPSmethods.

TPRandXPSresultsrevealedthatduringthepre-treatmentofthecatalyststheCoonlypartiallyreduced.

Thereductiondegreeandtheamountofthestructuredmetalliccobaltincreasedwiththereduction temperature.Theconversionofthereactantswasthehighestonthesamplereducedat973K,butthe turnoverfrequenciesofCOandH2formationincreasedasthereductiontemperatureincreased.

Theamountofsurfacecarbonsignificantlydependedonthepre-treatmenttemperatureofthecatalysts.

TheIRspectraregisteredatthebeginningofthereactionindicatethattiltedCOwasformed,butthe positionoftheabsorptionbandsdependsonthereductiontemperatureofthecatalysts.Similarly,the pre-treatmenttemperatureinfluencedthetypeofthesurfacecarbondeterminedbyXPS.Weassume thatthedifferentstructuresofthemetallicCoformedinpretreatmentresultedinthedifferentcatalytic behaviors.

©2016ElsevierB.V.Allrightsreserved.

1. Introduction

TheconversionofCH₄andCO₂–thetwomaincomponentsof biogasandthemajorgreenhousegases–intosynthesisgasisinthe focusofacademicandindustrialresearch.Oneofthepossibilities toachievethisgoalistoreactCH₄andCO₂witheachother.

AlloftheGroupVIIImetalsonavarietyofsupportshavebeen studiedasdryreformingcatalysts[1–3].Weshowedinourprevi- ousstudiesthatsupportednoblemetalsareactiveinthisreaction [4–7].SupportedCocatalystshavebeenalsoextensivelyinvesti- gatedinthedryreformingofmethaneduetotheirlowercostand availabilityinlargerscale.Theresultsobtainedonthesecatalysts aresummarizedindifferentreviews[1,2,8,9].

Ferreira-Aparicioetal.[10]comparedtheactivitiesofdifferent aluminasupportedcatalysts,andthoughtheCocontainingsample wasnotthebestonebutitshowedgoodstabilityinthewholetem- peraturerange.RuckensteinandWang[11]foundthatCo/Al2O3

exhibitedaverygoodactivityamongCocatalystssupportedondif- ferentoxides.ThestabilityofCo/Al₂O₃catalyststronglydepends ontheColoadingandthecalcinationtemperature[12].Overthe catalystwithhighCoconcentration (>12wt%)notable amounts of carbon were accumulated during reforming, and deactiva- tionwasobserved.It wasfoundthatnotonlythedepositionof

∗Correspondingauthor.

surfacecarbonbutalsotheoxidationofmetallicsitescouldresult inthedecreaseofactivity.Itwasstablewhenthebalancebetween theformationandoxidationofcarbon wasachieved.Thestruc- turalcharacteristicsoftheAl2O3 supportedCo(24wt%)catalyst werestronglyaffectedbythecalcinationtemperatureandsowas theirreducibility[13].The catalystslosttheiractivitiesin time independentlyofthecalcinationtemperature[11].Characteriza- tionof Co/␣-Al2O3 indicatedthat thesurfacespecies ofCo, the reformingactivityandtherateofcokeformationdependonthecal- cinationtemperatureandthereductiontime[14].Similarresults werefoundon␥-Al2O3supportedCocatalyst;theamountandthe reactivityofsurfacecarbonsignificantlydependedonthecalcina- tiontemperature[15].SanJosé-Alonsoetal.studiedthereaction onCo/Al2O3 withlow Cocontent[16].1%Co/Al2O3 deactivated duringthefirstminutesofthereactionat973KbecauseCoAl₂O₄ wasformed,whileon2.5%Co/Al₂O₃highstabilityandlowcarbon depositionwasobserved.Theeffectsofthepreparationmethods ontheefficiencyofCo/Al2O3werealsostudied[17,18].Thegood cokingresistivityofthecatalystswasattributedtothehighsur- faceconcentrationofOHspecies,tothesmallmetallicCoparticles andtothestronginteractionbetweenCoandthesupport.Itwas statedthatCoAl₂O₄playsanimportantroleinthepreventionof cokeformation[17].Co/Al₂O₃preparedbycontrolledadsorption methodshowedhigherCO2andCH4conversionthanthesample preparedbydryimpregnation;itwasattributedtothehigherdis- persionofCo.Onthissamplemorenon-deactivatingcokedeposit http://dx.doi.org/10.1016/j.cattod.2016.04.007

0920-5861/©2016ElsevierB.V.Allrightsreserved.

wasformed,whileontheothercatalystdeactivatingcokebuiltup whichdrasticallyhinderedthecatalyticactivity[18].

Rareearthoxideadditivesimprovedtheanti-cokeperformance ofthe20%Co/Al₂O₃catalyst[19]intheCO₂+CH₄reaction.Rhaddi- tivesenhancedthereducibilityofCoandtheactivityofthecatalyst [15].KdopingofCo/Al₂O₃alsoreducesthecarbondepositionbut thecatalyticactivityaswell;itwasexplainedbyitsabilitytocat- alyzethecarbongasificationandbypartialcoverageofcobaltactive sitesforCH₄decomposition[20].

Takanabeetal.[21,22]studiedtheinfluenceofthereduction temperatureonthecatalyticbehaviorofCo/TiO2inthedryreform- ing of methane.The great difference in the initialactivities of 10%Co/TiO₂reducedatdifferenttemperaturescorrespondstothe differentcrystalstructuresofTiO₂ afterthereduction[21].0.5%

Co/TiO2reducedbelow1123Kshowedrapidandcompletedeac- tivationatthebeginningofthereaction;thedeactivationwould becorrelatedtothecokingontheactivesites.Ontheotherhand, theCo/TiO2reducedatandabove1123Kshowedrelativelystable activity[22].

Thepresentpapergivesanaccountoftheeffectofthereduction temperatureonthestructureandcatalyticefficiencyofCo/Al2O3in theCO₂+CH₄reaction.Earlierwefoundthatwhenthereducibility ofCoonaluminaincreased,soincreasedtheactivityofthecatalysts [15].Thechangesofthereductiontemperaturealsoinfluencedthe ratioofoxidizedandreducedCoonthesurfaceofthesupport.With thepresentcontribution,wewouldliketoshowtheeffectofthe above-mentionedratioontheactivity,thestabilityofthecatalyst andonthecokeformation.

2. Experimental

2.1. Preparationofthecatalysts

Thesupport(Al2O3,DegussaP110C)wasimpregnatedwiththe aqueoussolutionofCo(NO₃)₂toyieldanominal10wt%metalcon- tent.The samplesweredriedat 383K and calcinedat 973K in air.

Beforethemeasurements,thecatalystswereoxidizedat673K inO₂flowfor30minandreducedatdifferenttemperaturesfrom 773Kto1173KinflowingH2for60mininsitu.Afterwardsthecat- alystwasflushedwithN₂atthereductiontemperaturefor15min, andcooleddowntothereactiontemperature.

2.2. Kineticmeasurements

Catalyticreactionswerecarriedoutinafixedbedcontinuous- flowreactor.TheratioofCH4/CO2inthereactinggasmixturewas 1:1.Theamountofcatalystusedwasusually0.15g.Theflowrate ofthereactantswas60ml/minatroomtemperatureandpressure.

Analysesof thegaseswereperformedwithgaschromatograph (Agilent6890N)using HP-PLOT Q column. TC and FI detectors detectedthegasessimultaneously.ToincreasethesensitivityofCO andCO₂detectionamethanizerwasappliedbeforethedetectors.

Infraredspectra wererecorded during thecatalytic reaction withanAgilentCary670typeFTIRspectrometerequippedwith diffusereflectanceattachment(Harrick)withBaF₂windowswith awavenumberaccuracyof±4cm⁻¹.Typically,32scanswerereg- istered.Thecatalystswerepre-treatedasmentionedabove,and thentheCH₄+CO₂gasmixturewasintroducedintothecellatthe reactiontemperatureandtheIRspectrawererecorded.Allspec- trawereratedtothespectraofthecatalystspre-treatedbeforethe measurements.Whenthesampleshadtobereducedabove873K (thecellisheatableonlyto873K),thecatalystsweretreatedwith H₂exsituatthegiventemperatureandcooleddowninN₂stream toroomtemperature.Beforethemeasurementsthecatalystwas reducedagaininsituat773K.SeparateTPRmeasurementsrevealed

thatthistreatmentissufficienttoremovetheadsorbedoxygenor oxidizedspecieswhileonlyonereductionstagewasdetectableat about600K.

2.3. Characterizationofthecatalysts

Theamountofsurfacemetalwasdeterminedafterpretreatment byH2adsorptionusinganimpulsemethod.

TemperatureprogrammedreductionandtheBETsurfacemea- surements of the catalysts were carried out by a BELCAT-A instrument.TheBETsurfacewasmeasuredusingN2adsorptionat thetemperatureofliquidnitrogen.BeforeTPRmeasurements,the catalystsweretreatedinoxygenat673Kfor30minandflushed withArfor15min.AfterwardsthesamplewascooledinflowingAr toroomtemperature.ThepureArflowwaschangedto10%H2con- tainingAr,andthereactorwasheatedlinearlyatarateof20K/min upto1373KandtheH₂consumptionwasdetected.

Theamountandthereactivityofsurfacecarbonformedinthe catalyticreactionsweredeterminedbytemperature-programmed hydrogenation.Afterthecatalyticrun(2h)thereactorwasflushed withN2atthereactiontemperature,andthesamplewascooled downto373K.ThentheN₂flowwaschangedtoH₂andthesam- plewasheatedupto1173Kandtheforminghydrocarbonswere determinedbygaschromatograph.

ForXPSstudies,thepowdersampleswerepressedintotablets withca.1cmdiameterandafewtenthofmmthicknessandplaced intotheloadlockofthespectrometer.Sampletreatmentswere carried out in a high-pressure cell (catalysis chamber) directly attachedtotheanalysischamberand isolatedfromthatwitha gatevalve.Withthehelpofthesamplemanipulatorsitwaspos- sibletotransferthesamplesfromtheanalysischamberintothe highpressurecellinhighvacuum,withoutthereachofair.The sampleswerepre-treatedinthesamewayasdescribedabove.The high-pressurecellisheatablealsoonlyto873K;sowhenthesam- pleshadtobereducedabovethistemperaturetheyweretreated aswrittenabove.Afterthepre-treatment,thesampleswerecooled toroomtemperatureinflowingnitrogengasandthesamplewas takenbacktotheanalysischamber.Afterspectrumacquisition,the samplewasmovedbackintothecatalyticchamber andtreated withthereactinggasmixtureatthereactiontemperatureunderthe sameexperimentalconditionsasthecatalyticreaction.Theexper- imentwasinterruptedafterthe5th,30thand120thminutesofthe reactionandtheXPspectrawererecorded.

XPspectraweretakenwithaSPECSinstrumentequippedwitha PHOIBOS150MCD9hemisphericalelectronenergyanalyzeroper- atedintheFATmode.TheexcitationsourcewastheK␣radiationof analuminumanode(h=1486.6eV).TheX-raygunwasoperated at210Wpower(14kV,15mA).Thepassenergywassetto20eV, thestepsizewas50meV.Typically,fivescanswereaddedtogeta singlespectrum.ThebindingenergyoftheAl2plineinAl2O3was usedasenergyreference:itwastaken74.7eV.Fordataacquisition andevaluationbothmanufacturers’(SpecsLab2)andcommercial (CasaXPS)softwarewereused.

TheXRDstudywascarriedoutona RigakuMiniflexIIpow- derX-raydiffractometerequippedwithaCuK␣radiationsource (=0.15418nm)byapplyingascanningrateof4^◦/mininthe2 rangeof3–80^◦.

3. Resultsanddiscussion 3.1. Characterizationofthecatalysts

TheBETsurfacearea,theamountofH₂–adsorbedat373K– andthedispersionoftheCoontheCo/Al₂O₃reducedatdifferent temperaturesaresummarizedinTable1.

Table1

Somecharacteristicdataofthe10%Co/Al2O3catalystsreducedatdifferenttemperature.

Reductiontemperature BETsurface AdsorbedH2at373K Dispersion Averageparticlesize^a _(Co0^Co+Co⁰²⁺) b

K m²/g ␮mol/g % nm reduced usedinthereactionfor

10min30min120min

773 95 227 24 12.6 0.37 0.37 0.31 0.30

973 94 211 22 21.5 0.39 0.35 0.35 0.31

1173 96 129 14 26.1 0.57 0.57 0.57 0.41

aDeterminedfromXRDdatausingtheScherrerequation.

bDeterminedfromXPSresults.

0 20 40 60 80 100 120

0 2 4 6 8 10

CH4 conversion (%)

Reaction time (min) 1

2

3

A

0 20 40 60 80 100 120

0 2 4 6 8 10 12 14 16

3

CO2 conversion (%)

Reaction time (min) 1

2

B

Fig.1. TheconversionofCH4(A)andCO2(B)intheCH4+CO2reactionat773KonCo/Al2O3reducedat773(1),973K(2)and1173K(3).

The BET surfaces of thesamples were practically thesame, 94–96m²/g,irrespectiveofthereductiontemperature(Table1).

The Co dispersion calculated from the amount of adsorbed hydrogendecreasedasthereductiontemperatureincreased.Itwas thehighest(24%)onthecatalystreducedat773K,andthelow- est(14%)whentheCo/Al₂O₃wastreatedwithhydrogenat1173K (Table1).

TheXRDprofileoftheCo/Al2O3calcinedat973Kshowsdiffrac- tionpeaksat31.5^◦,37.1^◦,59.5^◦,and65.5^◦ [15]correspondingto CoAl₂O₄ (JCPDScardNo.82-2248)orCo₃O₄ (JCPDScardNo.78- 1970)[23–25].Bothhavespinelstructurewithalmostidentical diffractionpatterns[25].Onthereducedsamplesthediffractions characteristicforthespinelstructuresdisappearedandnewpeaks weredetectedat44.2^◦ and51.5^◦ correspondingtoCo(111) and Co(200)(JCPDScardNo.15-0806)facets.Theintensitiesofthese featuresincreasedasthereductiontemperatureincreased.

FromtheXRD datausingthe Scherrerequationthe average crystalsizeofthereducedcatalystswasdetermined(Table1).We foundthattheparticlesizeincreasedasthereductiontempera- tureincreased.Itwas12.6nmonthecatalystreducedat773Kand 26.1nmwhenthesamplewastreatedwithhydrogenat1173K.

TheTPRprofileofthecatalystconsistsofabroadpeakbetween 580and980Kwiththreedifferentmaximaat767,823,and885K, andahightemperaturereductionstage(Tmax=1103K)wasalso observed[15].Thepeak atlowtemperaturecorresponds tothe reductionoflargecrystallineCo₃O₄particles.Thehightempera- turereductionstagecanbeattributedtothereductionofCoOand CoAl2O4.Someauthorssuggestthatthepeakatlowtemperatureis inconnectionwiththereductionoflargecrystallineCo₃O₄particles toCo⁰viaCoOformation[23],othersproposethatthislowtemper- aturepeakcanbeascribedtothereductionofCo3O4toCoO[26].

Thereisanagreementthatthehightemperaturepeakcorresponds

tothereductionofCo³⁺andCo²⁺species[23,26],butEwbanketal.

[18]suggestedthatCoAl₂O₄reducedatthishightemperature.

TheCo2pXPspectraofthe10%Co/Al₂O₃ sampleshowsthe characteristicdoublet(Co2p_3/2at781.4eV)withshakeupsatellite peaksbothintheasreceivedandtheoxidizedstate,corresponding toCo²⁺.Thisvalueagreedwellwiththeearlierfindings[27]usually citedintheliteratureforcobalt-oxides.

Afterreductionastrongshoulder,morepreciselyanearlydis- tinctpeakappearedatthelowbindingenergysideoftheCo2p_3/2 component at 778.0eV, which can be attributed to themetal- liccobaltstate[28,29].Theintensityofthispeakincreasedwith increasingthereductiontemperature,butde-convolutingtheCo 2p_3/2 envelopeafterthetreatmentofthesampleat1173K,the peakandthesatellitecharacteristicofCo²⁺arestillunequivocally detectable.TheratioofCo⁰/(Co⁰+Co²⁺)calculatedfromtheXPS peakareashowsthattheratioofmetalliccobaltishardlymore than50%evenafterthehightemperaturereduction(Table1).

3.2. Reactionofmethanewithcarbondioxide

ThereactionofCO₂andCH₄wasstudiedunderisothermcondi- tionsat773K.Theactivitiesofthesamples,theconversionofCH₄ andCO2 increasedintimeinallcasesduring2hofthereaction (Fig.1).

OnlyCO, H₂ and a smallamountof water weredetected.It shouldbementionedthatwhenthecatalystswereusedforlonger time(atleast100h)theactivityofthecatalystsslowlydecreased.It wasfoundthattheCO₂conversionwashigherthanthatofmethane inallcasesandtheCO/H₂ratiowashigherthan1(Table2).These observationsindicatedthatthereaction

CH4+CO22CO+2H2

Table2

SomecharacteristicdatafortheCH4+CO2reactionat773Kon10%Co/Al2O3reducedtdifferenttemperature.

Reduction temperature

Formationrate(␮mol/gs) Conversion(%) CO/H2 Turnoverfrequency(1/s) Surface

carbonaceous deposit␮mol/g

CO H2 CO2 CH4 H2 CO

5min 115min 5min 115min 5min 115min 5min 115min 5min 115min 5min 115min 5min 115min

773K 9 12.7 2 3 4.3 6.1 2.2 3.1 4.4 4.2 0.005 0.01 0.025 0.036 884.8

973K 24.7 28.9 8.6 11.9 14.3 16.6 7 8.9 2.9 2.4 0.023 0.032 0.066 0.077 7984.7

1173K 17.1 22.2 5.2 8.7 9.8 12.5 4.7 6.7 3.3 2.5 0.022 0.037 0.072 0.093 6285.3

wasfollowedbyseveralsecondaryprocesses,includingthehydro- genationofCO,CO2andthewatergasshiftreaction.Itisknown thatintheequilibriumat773KtheCO₂conversionishigherthan thatofmethanebyabout25%andtheCO/H₂ratioisabout1.6[9], butinourcasesthedifferenceintheconversionswasnear100%

andtheCO/H₂ ratioswerehigher.Comparingtheconversionof CH₄andCO₂obtainedondifferentcatalysts,wecanconcludethat theactivitywasthehighestonthesamplereducedat973Kand thelowestonthecatalysthydrogenatedat773K.Theorderofthe hydrogenandCOformationratewasthesameaswasobservedin theCH₄andCO₂(Table2).

Adifferentorderresultedwhenwecalculatedtheturnoverfre- quenciesforCOandH₂formation.Thehighestvaluewasfoundin bothcasesonthesamplereducedat1173K(Fig.2andTable2)after twohoursofthereaction.

Howcouldweexplainthedifferencesintheactivityofthecat- alystsreducedatdifferenttemperatures?Earlierweassumed[5]

inthecaseofsupportedRhcatalyststhatthefirststepofthereac- tionisthemethanedecompositiontohydrogenandperhapstoCH_x fragments.ThesespeciescouldpromotetheCO₂dissociationand theadsorbedoxygenfacilitatedthedissociationofCH4.Recentlya theoreticalcalculationshowedthatCO2couldbeactivatedonthe CosurfaceandCO⁻₂couldbeformed,butthisinteractionisstructure sensitive[30].Takingintoaccountthatwithincreasingthereduc- tiontemperatureincreasedtheareaofaspeciallyorientedsurface ofCo(fcc)particles,wecouldsupposethatthevaryingstructureof Co⁰resultedinthedifferentactivitiesofthesamples.

TheCO/H2ratiochangedinawiderange.Itwas4.2–4.4onthe samplereducedat773Kand2.4–2.5ontheothersamples(Table2).

Itmeansthat oneofthesecondary reactionstookplace, which consumesH2and/orproducesCO,suchasreversewatergasshift reaction

CO2+H2CO+H2O, orthemethanationofCO₂ CO2+4H2CH4+2H2O

Onthecatalystspre-treatedat973and1173KtheCO/H₂ratio decreasedintime.ItispossiblethattheH2formationrateincreased morerapidlythanthatofCO,butmoreprobably,therateofasec- ondaryprocesschangedandresultedinthedecreaseoftheCO/H₂ ratio.

After2hofthereaction,theamountandthereactivityofsur- facecarbon formedduringthe dryreforming of methanewere determinedbyhydrogenTPR.Thereactivityofthecarbondeposit wasnearly thesame in all cases; the TPR peak maxima were about820Kindependentlyofthereductiontemperature(Fig.3).

Onthecontrary,theamountofsurfacecarbondeterminedwith temperature-programmedhydrogenationwasnearly8timesless onthesamplereducedat773Kthanontheothercatalysts(Table2).

Ifwerelatedtheamountofsurface carbonformedinthereac- tiontotheconverted CH4+CO2 wefoundthatlessthan3.5%of thetransformedgasaccumulatedonthesurfaceintheworstcase.

3.3. Hydrogenationofcarbondioxide

TounderstandthedifferencesintheCO/H₂ratiomeasuredon thesamplesreducedatdifferenttemperatures,thehydrogenation ofCO₂wasalsostudiedonthesamplesreducedat773and973K.

Atlowtemperature,at548K,theconversionofCO₂wasbelow3%

andtheCH4 selectivitywasabout90%inbothcases.Whenthe reactionwascarriedoutat773K,theCO₂ conversionincreased withincreasingthereductiontemperature(63and85%),butthe selectivityofCH4formationwas99%inbothcases.Thesefindings donotsupporttheideathatthedifferencesintheCO/H₂ratioare theresultsofsecondaryreactions.WhentheCO₂+H₂reactionwas carriedoutonacatalystwhichwasusedearlierintheCH4+CO2

reaction,theamountofCOformedintheCO2+H2 reactionwas threetimeshigheronthesamplereducedat773Kthanintheother case.Fromtheseresultswemayassumethatthesurfaceofthe catalystschangedduringtheCO2+CH4reactionanditresultedin thedifferentCOselectivityinthereversewatergasshiftreaction andsointhedifferentCO/H₂ratiosinthedryreforming.

3.4. Characterizationofthecatalystsafterthecatalyticreaction TheinfraredspectrawereregisteredintheDRIFTcellduringthe catalyticreactionofCH₄+CO₂at773Kon10%Co/Al₂O₃reducedat 773K.Theyshowedthatbeyondthecarbonatebandsatthebegin- ningofthereactionaweakabsorptionwaspresentat1984cm⁻¹ attributedtoadsorbedCO.Featuresat1592and1353cm⁻¹could beassignedtotheasymmetricandsymmetricstretchingvibrations ofO C Ogroupsofformatespecies.Theintensitiesoftheformate bandsslightlyincreasedinthefirst10minofthereactionandthen remainedconstant.Betweenthesepeaks,somenotwell-resolved absorbancescouldbeattributedtodifferentcarbonatesandC H groups(Fig.4).TheCObandwasobservedonlyinthefirstminutes ofthereaction.

Similarspectrawererecordedwhenthecatalystswerereduced athighertemperature,onlytheCObandshiftedtohigherwave number(2015cm⁻¹).Khassinetal.[29]foundalsothesamediffer- encesinthepositionofCOabsorptionbandsonCo/Al2O3reducedat 753and923K;thosefeatureswereattributedtotiltedCObonded toCo⁰ [29].Itwassupposedthattheappearanceofthistypeof adsorbedCOcouldbeanindicationofthesecondaryinteraction of COwitha neighboring cobalt atom,which shouldcausethe lateraltiltoftheCO.Theoccurrenceofthisinteractionresultsin thesignificantlengtheningofC Obond[31].Itmeansthatinour casesthedistancebetweentheneighboringcobaltatomsisdif- ferentwhenthecatalystswerereducedatdifferenttemperatures.

AnotherexplanationfortheappearanceofthetiltedCOistheelec- tronicchangesinthemetalatomsonthesurface[32,33].

Earlierwefoundthattheformatespecieslocatedonthealumina support[34].Cobaltformatedecomposesat470–490Kproducing Co3O4[35,36]sointhepresentcaseswecouldalsoacceptthatthe formategroupsarelocatedonthesupport,butinspiteofthelower decompositiontemperaturewecannotruleoutthattheybondedto cobaltoxides.TheCObandwasdetectableonlyinthefirstminutes

0 20 40 60 80 100 120 0,000

0,005 0,010 0,015 0,020 0,025 0,030 0,035 0,040

3 2

Turnocer frequency (1/sec)

Reaction time (min) 1

A

0 20 40 60 80 100 120

0,00 0,01 0,02 0,03 0,04 0,05 0,06 0,07 0,08 0,09 0,10

3

1 2

Turnover frequency (1/sec)

Reaction time (min)

B

Fig.2. TurnoverrateofH2(A)andCO(B)formationintheCH4+CO2reactionat773KonCo/Al2O3reducedat773(1),973K(2)and1173K(3).

400 500 600 700 800 900 1000 1100 1200

0 1000 2000 3000 4000 5000 6000 7000 8000 9000

W ( nm ol /g s)

Tempe rature (K) 1 2 3

Fig.3.TemperatureprogrammedhydrogenationofsurfacecarbonaceousdepositformedintheCH4+CO2reactionon10%Co/Al2O3catalystsreducedat773K(1),973K(2) and1173K(3).

ofthereactionandtheintensityofitwasveryweak.Bandsinthe C Hregion(2800–2900cm⁻¹)werealsoobserved,butthesefea- turesarenotcharacteristic,whilethespectrawereregisteredinthe presenceofmethaneinCH₄+CO₂flow.WhentheDRIFTcellwas flushedwithHeafterthecatalyticreactionallbandsdisappeared immediatelyfromthespectra.Itissurprisingtodetectadsorbed COandformategroupswellabovetheirdesorptiontemperature.It meansthattheformationrateofthesespeciesishigherthantheir desorption,decompositionorfurtherreactionrates.

TheCo2pXPspectraof10%Co/Al2O3reducedatdifferenttem- peraturesareshown beforeandafterthereactionin Fig.5.The samplesweretreatedinthecatalysischamber(HPC)oftheinstru- mentunderthesameexperimentalconditionsasinthecatalytic tests.Theexperimentwasinterruptedafterthe10th,30th, and 120thminutesofthereactionandXPspectraweretaken.

TheCo2pspectraofthe10%Co/Al2O3sampleshowsthechar- acteristicdoublet(Co2p_3/2at781.4eV)withsatellitepeaksinthe oxidizedstateas wasmentioned above,correspondingtoCo²⁺. Theseresultsagreedwellwiththeearlierfindings[27].

Afterreduction andduring thereactiona strongshoulder, a nearlydistinctpeakappearedatthelowbindingenergysideofthe Co2p_3/2componentat778.0eV.ItsCo2p_1/2counterpartcouldalso bedetectedonthespectra.Therefore,thisfeaturecanbeattributed tothemetalliccobaltstate.Thebindingenergyofthisfeatureisin agreementwiththeliteraturedata[28,29].Theintensityofthis peakincreasedwithincreasingthereductiontemperature,butthe surfaceCoconcentrationsignificantlydecreasedduringthereduc- tion(Fig.5).DeconvolutingtheCo2p_3/2envelopeobtainedafter thereductionthepeakandthesatellitecharacteristicforCo²⁺is unequivocallydetectableevenwhenthecatalystwasreducedat 1173K(Fig.5).Table1showsthattheratioofmetalliccobalton thesurfacetothewholecobaltcontentdeterminedbyXPSwas hardlymorethan50%,andthesevaluesdecreasedfurtherduring thereaction; itseemsthatthecatalystsslightlyoxidizedduring thereaction.TherearesomeobservationsthatthesupportedCo catalystsarenotoxidized[37]butreduced[38]underthecatalytic test,probablyduetothereducingnatureoftheevolvedproducts (H2 and CO).It hastobenotedthat inboth casesthereaction

2000 1800 1600 1400 0,0

0,1 0,2 0,3 0,4 0,5

1984

1533

1503 10

9

8.

7

6

5

4 3

2

Absorbance (a.u.)

Wavenumber (cm^-1) 1

1592

Fig.4. Infraredspectrarecordedon10%Co/Al2O3reducedat773Kinthefirst(1), 2nd(2),3rd(3),5th(4),10th(5),15th(6),30th(7),60th(8),90th(9)120th(10) minutesoftheCO2+CH4reactionat773K.

temperaturewashigher(1073K)thanthereductiontemperature.

Inourcasesthecatalytictestswerecarriedoutonlyat773K,but probablythereisasensitiveequilibriumbetweenthereductiveand oxidativeprocesses.Ourresultsdemonstratedthatat773Ktherate oftheoxidativereactionssuchastheCO2[39]dissociationorthe

formatedecomposition[35,36]onCosurfaceishigherthanthatof reduction.

Ithastobenotedthattheactivityofthecatalystsincreased intimewhilsttheratioofmetalliccobaltdecreased,althoughthe activityoftheunreducedsamplewasrelativelylowinthereac- tion.Inspiteofthistheoxidationofmetalhasbeenreportedasa deactivationmechanismofcobalt-basedcatalysts[12,21].

TheC1speakontheXPspectrawaslocatedinthereducedsam- plesat285.0eVinallcases,originatedfromadventitiouscarbon.

Theintensityofthispeakdecreasedasthereductiontemperature increased(Fig.6).

Therewerenochangesinthepeakpositioninthisregionwhen thesamplereducedat773Kwasusedinthecatalyticreactionbut theintensityofitincreasedsignificantlyespeciallyinthesecond hourofthereaction(Fig.6A).

VerydifferentC1sspectraweredetectedwhenthecatalysts werereducedat973or1173K.Alreadyafter10minofthereac- tionnewpeaksweredetectedat282.8eVandlateranotherone wasobservedat281.5eV(Fig.6B)onthesamplereducedat973K;

thissignalbecamemoreandmoredominant.Similarspectrawere detectedonthesamplereducedat1173K.

Ewbanketal.[18]alsofoundaC1sXPSpeakat281.7eVafterthe dryreformingofmethaneonCo/Al₂O₃,whichwasattributedtothe presenceofcarbidiccarbon.Inanotherwork,thebindingenergy ofC1sinCoC2isreportedat283.2eV[40,41].Bulkcobaltcarbide decomposedtoformgraphiticcarbonabove700K[41].Fengetal.

[42]alsodetectedanunusuallylow bindingenergyfor C1s,at 281.4eV,intheconversionofethanoltoacetonitrileonalumina supportedNiorCocatalysts.Theysupposedthatthecarbonhad combinedwithCoorNiinthecatalyticruns.Takingintoaccount theseobservations,thelowbindingenergyC1sspeciescouldbe assignedasCoC_xcarbidelikeform,ratherthanasastructuredcar- bonlayer.Thedifferencesinthestructureofthecarbonformed duringthereactiononthecatalystsreducedatdifferenttemper- aturescouldbealsoexplainedbythedifferentsurfacestructures ofmetallicCoproducedinthereductionat773Kandat973Kor highertemperatures.

Comparingtheamountofsurfacecarbon,determinedbyTPR (Fig. 3 and Table 2), and the C 1speak intensitieson theXPS curvesobtainedonthesamplesreducedatdifferenttemperatures,

815 810 805 800 795 790 785 780 775 770

781.4 778.0

2000 cps

(a) (b) (c) (d) (e)

815 810 805 800 795 790 785 780 775 770

781.4 778.0

(a) (b) (c) (d) (e)

2000 cps

815 810 805 800 795 790 785 780 775 770

781.4 778.0

2000 cps

(e) (d) (c) (b)

(a)

Bind ing energy (eV)

A B C

Fig.5.XPspectraofCo2precordedaftertheCO2+CH4reactionat773KonCo/Al2O3reducedat773K(A),973K(B),1173K(C);oxidized(a),reducedsample(b),after5th (c),30th(d)and120th(e)minutesofthereaction.

292 290 288 286 284 282 280 278 292 290 288 286 284 282 280 278

281.5 282.8 285.0

(a) (b) (c) (d)

250 cps

Bind ing energy (eV) (e)

A B

(e) (d) (c)

(a) (b)

500 cps

285.0

Fig.6. XPspectraofC1srecordedaftertheCO2+CH4reactionat773KonCo/Al2O3reducedat773K(A),973K(B),oxidized(a),reducedsample(b),after5th(c),30th(d) and120th(e)minutesofthereaction.

itseemsthatthereisacontradiction.Onthesamplereducedat 773K,arelativelyintensivepeakwasdetectedafterthereactionon theXPSspectraat285eV.ItisassignedtosurfacecarbonfromC C orC Cgroups,butonlyasmallamountofcarbonwasdetectedby TPR.Onthesamplesreducedathighertemperatures,theamountof surfacecarbondeterminedbyTPRwasmorethanoneorderofmag- nitudehigher(Fig.3,Table2)buttheXPSpeakintensities,although atlowerenergies,weremuchless.Thesefeatureswereassignedto carbidelikestructures.Howcouldthisdiscrepancybeexplained?

Wehavetosupposethatthecarbononthesamplereducedat 773KisdirectlyonthesurfaceandsotheXPSsignalisnotshielded bythecobaltorbythesupport.Ontheothersamples,themain partofcarbondiffusesduringthereactionintothebulkandsonot detectablebyXPS,butsegregatestothesurfaceabove770Kduring theTPRandreactswithhydrogen.

WemayassumethattheratioofthemetallictooxidizedCo evolved in the pretreatment resulted in the different catalytic behaviors.

4. Conclusion

XRDresultsrevealedthat cobaltspinelstructures formedin thethermaltreatmentofCo/Al2O3.TPRandXPSresultsrevealed thatduringthepre-treatmentofthecatalyststheCoonlypartially reduced.Thereductiondegreeandtheamountofthestructured metalliccobaltincreasedwiththereductiontemperature,butthe ratioofmetallic cobaltonthesurfacetothewholecobalt con- tentdeterminedbyXPSwashardlymorethan50%evenafterthe reductionat1173K.

Theconversionof thereactantwasthehighest onthesam- plereducedat973K,buttheturnoverfrequenciesofCOandH₂ formationincreasedasthereductiontemperatureincreased.

Theamountof surfacecarbonsignificantly dependedonthe pre-treatmenttemperatureofthecatalysts.Bymeansof insitu DRIFTspectroscopyinsomecasesadsorbedCOandformatespecies weredetectedduringthecatalyticrunfarabovetheirdesorption temperature;theformationrateofthemishigherthanthedes- orption,decompositionorfurtherreactionrates.Fromthespectra, wemaysupposethatatiltedCOwasformed,butthepositionof theabsorptionbandsdependsonthereductiontemperatureofthe catalysts.

Similarly,thepre-treatmenttemperatureinfluencedthetypeof thesurfacecarbondeterminedbyXPS.Onthecatalystsreducedat lowtemperature,theamountoftheadventitiouscarbonincreased, butontheothersamples,thecarbidictypecarbonwasdetected.We maysupposethatthedifferentstructuresofthecatalystsformedat differentreductiontemperaturesresultedinthedifferentcatalytic behavior.

References

[1]M.C.J.Bradford,M.A.Vannice,Catal.Rev.41(1999)1–42.

[2]M.-S.Fang,A.Z.Abdullah,S.Bhatia,Chem.Cat.Chem.1(2009)192–208.

[3]D.Pakhare,J.Spivey,Chem.Soc.Rev.43(2014)7813–7837.

[4]F.Solymosi,Gy.Kutsán,A.Erd"ohelyi,Catal.Lett.11(1991)149–156.

[5]A.Erd"ohelyi,J.Cserényi,F.Solymosi,J.Catal.141(1993)287–299.

[6]A.Erd"ohelyi,K.Fodor,T.Szailer,Appl.Catal.B:Environ.53(2004)153–160.

[7]I.Sarusi,K.Fodor,K.Baán,A.Oszkó,G.Pótári,A.Erd"ohelyi,Catal.Today171 (2011)132–139.

[8]A.W.Budiman,S.-H.Song,T.-S.Chang,C.-H.Shin,M.-J.Choi,Catal.Surv.Asia 16(2012)183–197.

[9]C.Papadopoulou,H.Matralis,X.Verykios,CatalysisforAlternativeEnergy Generation,in:L.Guczi,A.Erd"ohelyi(Eds.),Springer,2012,pp.57–127.

[10]P.Ferreira-Aparicio,A.Guerrero-Ruiz,I.Rodrıguez-Ramos,Appl.Catal.A:Gen.

170(1998)177–187.

[11]E.Ruckenstein,H.Y.Wang,Appl.Catal.A:Gen.204(2000)257–263.

[12]E.Ruckenstein,H.Y.Wang,J.Catal.205(2002)289–293.

[13]H.Y.Wang,E.Ruckenstein,Catal.Lett.75(2001)13–18.

[14]Z.Hou,T.Yashima,React.Kinet.Catal.Lett.81(2004)153–159.

[15]Zs.Ferencz,K.Baán,A.Oszkó,Z.Kónya,T.Kecskés,A.Erd"ohelyi,Catal.Today 228(2014)123–130.

[16]D.SanJosé-Alonso,M.J.Illán-Gomez,M.C.Román-Martinez,Int.J.Hydrogen Energy38(2013)2230–2239.

[17]L.Ji,S.Tang,H.C.Zeng,J.Lin,K.L.Tan,Appl.Catal.A:Gen.207(2001)247–255.

[18]J.L.Ewbank,L.Kovarik,C.C.Kenvin,C.Sievers,GreenChem.16(2014) 885–896.

[19]S.Zeng,L.Zhang,X.Zhang,Y.Wang,H.Pan,H.Su,Int.HydrogenEnergy37 (2012)9994–10001.

[20]D.SanJosé-Alonso,M.J.Illián-Gomez,M.C.Román-Martinez,Catal.Today176 (2011)187–190.

[21]K.Takanabe,K.Nagaoka,K.Nariai,K.Aika,J.Catal.230(2005)75–85.

[22]K.Nagaoka,K.Takanabe,K.Aika,Appl.Catal.A:Gen.255(2003)13–21.

[23]B.Jongsomjit,J.Panpranot,J.G.GoodwinJr.,J.Catal.204(2001)98–109.

[24]L.P.R.Profeti,E.A.Ticianelli,E.M.Asaf,J.PowerSources175(2008)482–489.

[25]Y.Zhang,H.Xiong,K.Liew,J.Lin,J.Mol.Catal.A:Chem.237(2005)172–181.

[26]G.Jacobs,T.K.Das,P.M.Patterson,J.L.Li,L.Sanchez,B.H.Davis,Appl.Catal.A 247(2003)335–343.

[27]R.L.Chin,D.M.Hercules,J.Phys.Chem.86(1982)360–367.

[28]B.A.Sexton,A.E.Hughes,T.W.Turney,J.Catal.97(1986)390–406.

[29]A.A.Khassin,T.M.Yurieva,V.V.Kaichev,V.I.Bukhtiyarov,A.A.Budneva,E.A.

Paukshtis,V.N.Parmon,J.Mol.Catal.A:Chem.175(2001)189–204.

[30]V.A.delaPe ˜naO’shea,S.Gonzalez,F.Illas,J.L.G.Fierro,Chem.Phys.Lett.454 (2008)262–268.

[31]A.B.Anderson,D.Q.Dowd,J.Phys.Chem.91(1987)869–873.

[32]S.P.Mehandru,A.B.Anderson,Surf.Sci.201(1988)345–360.

[33]R.L.Toomes,D.A.King,Surf.Sci.349(1996)1–18.

[34]F.Solymosi,A.Erd"ohelyi,M.Kocsis,J.Catal.65(1980)428–436.

[35]X.Wang,L.Lu,P.Wu,Q.Song,X.Wang,Termochim.Acta165(1990)139–145.

[36]E.Ingier-Stocka,A.Grabowska,J.Therm.Anal.54(1998)115–123.

[37]J.Estephane,S.Aouad,S.Hany,B.ElKhoury,C.Gennequin,H.ElZakhem,J.El Nakat,A.Aboukais,E.AbiAad,Int.J.HydrogenEnergy40(2015)9201–9208.

[38]R.Bouarab,O.Akdim,A.Auroux,O.Cherifi,C.Mirodatos,Appl.Catal.A:Gen.

264(2004)161–168.

[39]J.Lahtinen,T.Anraku,G.A.Somorjai,Catal.Lett.25(1994)241–255.

[40]J.Xiong,Y.Ding,T.Wang,L.Yan,W.Chen,H.Zhu,Y.Lu,Catal.Lett.102(2005) 265–269.

[41]J.Nakamura,I.Toyoshima,K.Tanaka,Surf.Sci.201(1988)185–194.

[42]C.Feng,Y.Zhang,Y.Zhang,Y.Wen,J.ZhaoCatal.Lett.141(2011)168–177.