Effect of Ti(IV) loading on CH4 oxidation activity and SO2 tolerance of Pd catalysts supported on silica SBA-15 and HMS

(1)

ContentslistsavailableatScienceDirect

Applied

Catalysis

B:

Environmental

j ou rna l h o m e p a 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

Effect

of

Ti(IV)

loading

on

CH

4

oxidation

activity

and

SO

2

tolerance

of

Pd

catalysts

supported

on

silica

SBA-15

and

HMS

A.M.

Venezia

a,∗

_,

_G.

_Di

_Carlo

b

_,

_L.F.

_Liotta

a

_,

_G.

_Pantaleo

a

_,

_M.

_Kantcheva

c

a_Istituto_per_lo_Studio_dei_Materiali_{Nanostrutturati}_(ISMN-CNR),_via_Ugo_La_Malfa,_153,_Palermo_I-_90146,_Italy

b_Istituto_per_lo_Studio_dei_Materiali_{Nanostrutturati}_(ISMN-CNR),_via_Salaria_km_29300,₀₀₀₁₅_Monterotondo_Stazione,_Rome,_Italy c_Department_of_Chemistry,_Bilkent_University,₀₆₈₀₀_Bilkent,_Ankara,_Turkey

a

r

t

i

c

l

e

i

n

f

o

Articlehistory: Received4April2011

Receivedinrevisedform3June2011 Accepted11June2011

Available online 17 June 2011 Keywords: Pdcatalyst CH4oxidation TiO2-SBA-15 TiO2-HMS SO2effect

a

b

s

t

r

a

c

t

PuresilicaSBA-15andHMSandcorrespondingTi(IV)modiﬁedmesoporoussilica,with5and10wt% ofTiO2,werepreparedandusedassupportforpalladium(1wt%)catalysts.Thematerials,analysedby XPS,XRD,BET,NH3-TPDandFT-IRtechniques,weretestedinthetotaloxidationofmethane.The cat-alyticactivitywasmeasuredinleanconditionsatWHSV=60,000mlg−1h−1intheabsenceandpresence of10vol.ppmSO2.Moreover,theeffectofaprolongedreactionagingandsevereSO2poisoningonthe catalyticperformanceofthebestperformingcatalystwasinvestigated.TheadditionofTiO2improved thecatalyticperformanceoftheSBA-15supportedcatalystsbyincreasingthesulfurtoleranceandmost importantlybyfavoringtheregenerationofthecatalystinsubsequentSO2-freeruns.Anopposite behav-iorwasobservedwiththepalladiumsupportedonTi(IV)-modiﬁedHMSsupportwhichexhibitedlower activityandasubstantialworseningoftheSO2toleranceascomparedtopalladiumsupportedonpure HMS.Onthebasesofthestructuralandchemicalinvestigation,thedifferencesbetweenthetwoseries ofcatalystswereascribedtothedistinctstructuralandacidicpropertiesofthesupports.Inparticular, thegoodperformanceoftheTi(IV)dopedSBA-15supportedcatalystswasduetothecombinationof Ti(IV)structurallyincorporatedintothesilicalatticeandpresentassurfacedispersedTiO2particles.The negativeeffectoftheTi(IV)overtheHMSsupportedcatalystswasrelatedtothehighacidityinducedby themorehomogeneousincorporationofTi(IV)intothesilicastructure.

1. Introduction

Natural gas, mainly consisting of methane, is increasingly

replacinggasolineordieselasfuelfortransportationvehicles[1,2].

Thereasonforthischangeisthedeclineofthepetroleumreserves

and theloweremission of pollutantsassociated withthe

com-bustionofmethane.Sincethenaturalgasfuelledvehicles(NGV)

typicallyrunatlowtemperature(320–420◦C)theemissionofNOx

islowerandduetothehighH:CratioalsotheproducedCO2 is

lessascomparedtotheotherfueldrivenvehicles[2].Nevertheless

amajorconcernrelatedtotheuseofnaturalgasistheemission

oftheunburnedmethanewhichhasanevenstrongergreenhouse

effectascomparedtotheCO2[3].Sincemethaneisanalmostinert

molecule,itrequireshightemperatureforitscompleteoxidation.

Acrucialstepintheachievementofthetotalcombustionprocess

istheactivationoftheﬁrstC–Hbond[4,5].Eitherhomolyticbond

cleavagewiththeformationofradicals,orheterolyticC–Hbond

cleavageattheacid–basepairofsitesisgenerallyconsidered.In

anycase,inordertoincreasetheefﬁciencyofthemethane

com-∗ Correspondingauthor.Tel.:+390916809372;fax:+390916809399. E-mailaddresses:venezia@pa.ismn.cnr.it,anna@pa.ismn.cnr.it(A.M.Venezia).

bustionatlowtemperature(below600◦C)itisnecessarytousea

suitablecatalyst.Forthisspeciﬁcreaction,twoclassesofcatalysts

arecurrentlyinvestigated,thosebasedontransitionmetaloxides

assolidsolutionoxides[6,7],perovskites[8,9],hexaaluminate[10],

andthosebasedonnoblemetals[11–14].Amongthislatterclass,

Pd-basedcatalystsarethemostactiveforthemethanetotal

oxi-dationatlowtemperatures.Aspointedoutinseveralstudies,their

catalyticactivitydependsstronglyonthenatureofthesupport[12],

onthepalladiumprecursors[15,16]andonthesizeofthePdO

particles[17].Theirmajordrawbackisrepresentedbytheireasy

poisoningbysulfurderivedfromthegasandenginelubricating

oil[11].Accordingtotheliterature,whenasulfatingsupportlike

Al2O3isused,palladiumdeactivatesslowly,duetothe

preferen-tialinteractionofSOxwiththesupport,atvariancewithpalladium

overtheinertSiO2deactivatingquicklybecauseofthedirect

inter-actionbetweenpalladiumandSOx.Nevertheless,theuseofsilica,

intheabsenceofparticlesintering,allowsaneasierregenerationof

thesulfur-poisonedcatalystthroughthethermaldecompositionof

thepalladiumsulfateoccurringattemperatureabove600◦C[1,11].

Recentstudiesinourgrouphaveshownthatbothmethane

conver-sionactivityandsulfurtolerancecouldbeimprovedsubstantially

by using high surface areasilica with speciﬁc characteristic of

acidityand morphology. In particular,supportingpalladium on

(2)

mesoporous silica HMS yielded more efﬁcientand more sulfur

resistantcatalystsascomparedtoalowersurfaceareasilica

sup-portedcatalyst[18].Incorporationofapreciseamount(10wt%)

ofTiO2 duringthesol–gelpreparation ofamorphoussilica

pro-ducedanadditionalimprovementofthesulfurtoleranceduring

themethaneoxidation inthepresenceofSO2,stillallowingthe

completecatalystregenerationtypicalofasilicasupport[19,20].

Thepositive behaviorwasattributedtothecombination ofthe

highsurfaceareasupportandthescavengeractionofthesulfating

oxidelikeTiO2.Withinthisframe,thepresentstudyintendsto

fur-therimprovethecatalyticbehaviorofpalladiumfortheoxidation

ofmethanebyusingtitania-dopedsilicamesoporousmaterialsas

catalystsupports.Tothisaimtwoseriesofpalladiumcatalysts

sup-portedonTi(IV)-modiﬁedHMSandTi(IV)-modiﬁedSBA-15were

prepared.Inordertocomparewiththeprevioussol–gelprepared

titaniamodiﬁedsilica[19],Ti(IV)wasincorporatedusingtheone

stepsynthesisinwhichtitaniumprecursorwasaddedduringthe

preparationofthemesoporousmaterials.Toevaluatetheeffectof

thesupportstructureonthePdactivityandstability,analysesby

XPS,XRD,BET,NH3-TPDandFT-IRtechniqueswereperformed.

2. Experimental

2.1. Supportandcatalystpreparation

ThemesostructuredHMSmaterialwassynthesizedaccording

toapublishedprocedure[21].Basically,HMSwasassembledfrom

4:1molarmixturesoftetraethylorthosilicate(TEOS)(Aldrich)as

theinorganicprecursoranddodecylamine(DDA)(Aldrich)asthe

structure-directingsurfactantin90:10(v/v)water/ethanol.About

49mmolofDDAweredissolvedin50mlofethanoland450mlof

H2O.Tothesurfactantsolution,heatedto60◦C,196mmolofTEOS

wereadded.Thegelmixturewaskeptina closedTeﬂonvessel

at60◦Cfor20h.Thereactionproductwasﬁltered,washedwith

distilledwateranddriedatroomtemperaturefor24h.The

surfac-tantwasremovedbycalcinationinairat500◦Cfor4h.Titanium

containingHMSmaterialsweresynthesizedaccordingtothe

pro-cedurereportedbyTanevetal.[22]usingTi(iso-OC3H7)4(Aldrich)

astitaniumprecursor,andfollowingthesamestepsasabove.The

titaniumprecursor wasaddedin appropriateamounttoyield a

5wt%and10wt%ofTiO2intheﬁnalmixedoxide.

SBA-15 was synthesized following a procedure described

by Zhao et al. [23]. Accordingly, 8.1g of Pluronic P123

(EO20PO70EO20,Aldrich)wasdissolvedin146.8gdistilledwater

and4.4gof conc.HCl (37%) and stirredover night at35◦C. To

thissolution16gofTEOS(Si(OC2H5)4,Aldrich98%)wasquickly

addedandstirredfor24hat35◦C.Thesuspensionwasannealed

at100◦Cfor24hinclosedpolypropylenebottle.Thesolidproduct

wasﬁltered,washedwithwaterandcalcinedat500◦Cfor5hin

air.Similarprocedurewasusedforthesynthesisofthetitanium

containingSBA-15materials.Themaindifferenceconsistedinthe

pre-hydrolysisofTEOSfor5hat35◦Cbeforeaddingappropriate

amountofTi(iso-OC3H7)4 dropbydropundervigorousstirring.

Thetitaniumprecursorwasaddedinappropriateamounttoyield

sampleswith5wt%and10wt%ofTiO2intheﬁnaloxide.

Attainment of the ordered mesoporous structures was

con-ﬁrmed by the SAXS patterns and by the typical type IV N2

adsorption–desorptionisotherms[21,24].

Palladiumwasdepositedbywetimpregnationusingan

aque-oussolutionofpalladiumnitrateintheappropriateamounttoyield

1wt%Pd loadedcatalysts.Thesampleswere calcinedat 500◦C

for4h.Thechemicalcompositionofthesampleswascheckedby

X-rayﬂuorescenceanalyses.ThesampleswerelabelledPd/HMS,

Pd/TixHMS,Pd/SBA-15andPd/TixSBA-15wherexrepresentedthe

TiO2weightpercentage(wt%).

2.2. Catalystcharacterization

X-raydiffractionpatternsweremeasuredwithaPhilips

verti-calgoniometerusingNi-ﬁlteredCuK␣radiation.Aproportional

counterand0.05◦stepsizesin2wereused.Theassignmentofthe

variouscrystallinephaseswasbasedontheJPDSpowder

diffrac-tionﬁlecards[25].Fromthelinebroadeningofthemainreﬂection

peaks,usingtheScherrerequation,particlesizesabovethe

detec-tionlimitsof3nmweredetermined[26].

SAXS measurements were performed with BRUKER AXS

NANOSTARwithstepsizesof0.02◦in2.

X-rayﬂuorescencewasperformedusingtheBrukerS2Ranger

spectrophotometer.

The microstructural properties of thematerials were

deter-minedfromN2adsorption–desorptionisothermsat−196◦Cusing

aSorptomatic1900(CarloErba)instrument.Beforethe

measure-ments,sampleswereheatedinvacuumat250◦Cfor2h.Speciﬁc

surfaceareasandporedistributionsofthematerialswereobtained

respectively using the Brunauer–Emmett–Teller (BET) and the

Barret–Joyner–Hatenda(BJH)calculationmethods[24].

TheX-rayphotoelectronspectroscopyanalyseswereperformed

withaVGMicrotechESCA3000Multilab,usingthe

unmonochro-matisedAlK␣source(1486.6eV)runat14kVand15mA.Forthe

individualpeakenergyregions,apassenergyof20eVwasused.

Samplesweremountedwithdouble-sidedadhesivetape.Binding

energieswerereferencedtotheC1sbindingenergyofadventitious

carbonsetat285.1eV.ThesoftwareprovidedbyVGwasusedfor

peakanalysesandforthecalculationoftheatomicconcentrations.

Theprecisiononthebindingenergyandontheatomicpercentage

valueswasrespectively±0.15eVand±10%.

The acidity of the oxide catalysts wasdetermined by

mea-surements oftemperature-programmeddesorptionof ammonia

(NH3-TPD).Thesampleamountof0.1gwasout-gassedina5vol.%

O2/He ﬂow at 500◦C for 1h. This was followed by

ammonia-saturation by ﬂowing 5% NH3/He stream (30ml/min) at room

temperaturefor1h.Afterpurgingwith100ml/minHeﬂowfor1h

at100◦Candthencoolingdowntoroomtemperature,thecatalyst

washeatedunderHe(30ml/min)inalinearrateof10◦C/minto

950◦Candtheammoniadesorptionwascontinuouslymonitored

bytheTCD.In ordertodeterminethetotalacidityof the

cata-lystfromitsNH3desorptionproﬁle,theareaunderthecurvewas

integrated.

The type of acidity (Lewis and Brønsted) was investigated

by FT-IR spectroscopy of adsorbed pyridine. The FT-IR spectra

wererecordedusingaBomemHartman&BraunMB-102model

FT-IRspectrometerwith a liquid-nitrogencooled MCTdetector

ata resolution of 4cm−1 (100scans).The self-supportingdiscs

(∼0.013g/cm2₎_were_activated_in_the_IR_cell_by_heating_for₁_h_in_a

vacuumat450◦Candinoxygen(100mbar,passedthroughatrap,

cooledinliquidnitrogen) atthesametemperature,followedby

evacuationfor1hat450◦C.Thepyridine(Sigma–Aldrich)

adsorp-tiontestwascarriedoutbytheadmissionof1.6mbarofthebase

intotheIRcell,leftincontactwiththesamplefor15min.Theexcess

pyridinewasthenevacuatedatroomtemperaturefor15min,

fol-lowedbyadesorptionofthestronglybondedbasefractioninthe

temperaturerange 25–250◦C.Thespectraoftheadsorbed

com-poundswereobtainedbysubtractingthespectraoftheactivated

samplesfromthespectrarecorded.Thesamplespectrawerealso

gas-phasecorrected.

2.3. Catalyticactivity

Methaneoxidation catalytictests wereperformedusinga U

shapedquartzreactorwithaninnerdiameterof12mm,electrically

heatedinafurnace.Thecatalystpowder(sievedfractionbetween

(3)

thermalgradients,anditwasplacedonaporousquartzdisk.The

reactiontemperaturewasmeasuredbyaK-typethermocouplein

contactwiththecatalyticbedlong12mm.Priortothecatalytic

testing,thesamplesweretreated“insitu”underﬂowingO2(5vol.%

inHe,50ml/min)at350◦Cfor0.5handunderHeduringcooling

at200◦C.Thestandardreagentgasmixture,consistingof0.3vol.%

ofCH4+2.4vol.%O2 inHe,wasledoverthecatalyst(50mg)ata

ﬂowrateof50ml/min(STP),equivalenttoaweighthourlyspace

velocity(WHSV)of60,000mlg−1h−1.Activitiesweremeasuredby

increasingthetemperaturefrom200◦Cto600◦C(bystepsof50◦C,

holdtime45min).Theinletandoutletgascompositionswere

anal-ysedbyonlinemassquadrupole(ThermostarTM_,_Balzers),_in_order

tofollowtheevolutionofallthespecies,CH4,CO,CO2,H2,H2O,

O2.Moreover,theconcentrationsofCO,CO2andCH4specieswere

checkedbyIRanalysers(ABBUras14,Uras26),calibratedinthe

range0–3000ppmforCO,0–10,000ppmforCO2and0–30,000ppm

forCH4.ThereactionproductsofmethaneoxidationwereCO2and

H2O.NoCOwasdetectedintheoverallrangeoftemperature.

Car-bonbalancewascloseto±5%inallthecatalytictests.Experiments

ofmethaneoxidationinthepresenceofSO2wereperformedby

co-feeding10vol.ppmofSO2.Betweenconsecutiverunsthesample

wascooleddowninHeatmosphere.

3. Resultsanddiscussion

3.1. Characterization

3.1.1. N2adsorption–desorptionanalyses

TheN2adsorptionanddesorptionisothermsofthepuresilica

supportsandofthetitaniadopedonesareshowninFig.1.The

Fig.1. Nitrogenadsorption–desorptionisothermsofpureandTi(IV)dopedHMS andSBA-15supports.

Table1

TexturalandstructuraldataofpureandTi(IV)-modiﬁedSBA-15andHMSsupports.

Sample Sa_(m2_/g) _d pb(nm) Vpc(cm3/g) a0d dwe(nm) HMS 824 3.2 0.99 4.6 1.4 Ti5HMS 845 2.5 1.33 4.5 2.0 Ti10HMS 728 2.6 1.00 4.3 1.7 SBA-15 838 7.7 0.99 11.3 3.6 Ti5SBA-15 912 7.1 1.35 11.3 4.2 Ti10SBA 841 7.1 0.96 11.0 3.9

a_The_BET_surface_area_values_were_calculated_in_the_range_{0.05–0.2p/p}0_. b_The_mean_pore_diameter_determined_using_the_BJH_model_from_N

2

adsorp-tion/desorptionisotherms.

c _The_pore_volume_determined_considering_the_range_p/p0_from_0.1_until_to_0.98. d_Unit_cell_parameter_calculated_as_a

0=2d100√3withd100beingtheplanedistance

computedaccordingtotheBragg’slaw(=2dsin).a0correspondstothedistance

betweenthepores.

e_The_wall_thickness_calculated_as_d w=a0−dp.

isotherms areoftype IV,withhysteresisloopscharacteristicof

mesoporouscompounds.AccordingtotheIUPACclassiﬁcation,the

hysteresisloopsoftheSBA-15andHMScanbeclassiﬁedasH1and

H3types,respectively[24].Asalreadyreportedinliterature,the

hysteresisloopsoftheSBA-15andTixSBA-15supportsarelarger

andtypicallyofthepresenceofmesopores[27].Incontrast,the

hysteresisloopsofHMSandTixHMSareratherﬂatandextended

overa largerange ofrelativepressures,indicating thepresence

ofbothframeworkmesoporosityandinterparticlemacroporosity

[27].ThetexturalpropertiesaresummarisedinTable1.TheBET

surfaceareasoftheHMSandSBA-15supportsaresimilarandquite

large.Bothsupportsareslightlyaffectedbythepresenceoftitania.

Asa functionofthetitanialoadingthesurfaceareasdo not

fol-lowanycleartrend,onthecontrary,theporediametersdecrease.

Asexpected,theaverageporediametersoftheSBA-15seriesare

morethandoublethoseoftheHMSseries.

3.1.2. SAXS

The SAXSpatterns of thetwo series of supports are shown

in Fig. 2. The correspondingunit cellparameters (a0), and the

wallthickness(dw)areprovidedinTable1.Theunitcell

param-eterswerecalculatedasa0=2d100√3,assumingahexagonalunit

cellforbothseriesofsupports.Thewallthicknesswasestimated

bysubtractingfroma0 theporediametersobtainedfromtheN2

physisorptionmeasurements.TheSAXScurvesoftheHMS

sam-plesarecharacterizedbyonemainbroadpeak,ataround2≈2◦

correspondingtoaunitcellparametersof≈4.4nmattributedto

abi-dimensionalhexagonalstructure[21].Theinsertionof

tita-niuminHMSresultedinaprogressivelossofthislong-rangeorder

andin aslightdecreaseoftheunitcellparameter.Onthe

con-trarySBA-15exhibitsthreemainreﬂectionpeaksindexedas(100),

(110)and(200)typicalofamoreorderedmaterial,withaunitcell

parameterof≈11.3nm.Theinsertionoftitaniadoesnotchange

sig-niﬁcantlytheoriginalhexagonalstructurebut,inaccordwiththe

decreasedpeakintensity,slightlycausesadeteriorationofthelong

rangeorder.AsshowninTable1,aslightdecreaseoftheunitcell

parametersisobservedforthetitaniacontainingsupportsand

cor-respondinglyanincreaseoftheframeworkwallthickness,toalarge

extentintheHMSseries.Thiseffectwouldconﬁrmacertaindegree

oftitaniaincorporationinthestructureofthesilica[28,29].The

Table2

AcidpropertiesofthesupportsasdeterminedbyTPD-NH3intherange50–500◦C.

Sample T(◦C) VNH3(mlg−1)

HMS 114 1.4

Ti10HMS 157 7.8

SBA-15 118 2.0

(4)

Fig.2. SAXSCurvesofpureandTi(IV)dopedHMSandSBA-15supports.

subsequentimpregnationofpalladiumleavesunchangedtheSAXS

patterns(notshowninhereforbrevity)ofbothseriesofsupports.

3.1.3. TPD-NH3

Thechange of acidityupon insertion of titania was

investi-gatedbytemperatureprogrammeddesorptionofammonia.The

techniqueprovidedinformationonthetotalacidityofthesolids,

includingbothBrønstedandLewisacidityandaccountingforthe

differentstrengthandnumberofacidsites.TheTPDpatternsofthe

pureandofthe10wt%TiO2dopedmesoporoussilicasareshown

inFig.3.TheresultsaresummarisedinTable2,intermsof

tem-peratureofNH3desorptionandvolumesofdesorbedNH3.Forboth

typesofmesoporousmaterialsonlytheﬁrstpeakwasconsidered.

Indeedtheotherfeature,consistingofabroadpeakobservedonly

inthe SBA-15samples,wasdue to theevolution of water and

organicmaterials,asdetectedbythequadrupolemeasurements,

andthereforenotincludedintheacidityevaluation.Accordingto

thepatternsandtothevalueslistedinTable2,itisevidentthatthe

insertionofTi(IV)inHMScausedasubstantialincreaseofacidity

intermsofboth,strengthandnumberofacidsites.Onthecontrary

theadditionoftitaniuminSBA-15producedonlyalimitedacidity

changes.

3.1.4. XRDanalyses

Inordertoidentifythecrystallinephasespresentinthesamples

theX-raydiffractionanalysesof thesupportedpalladium

cata-lystswereperformed.InFig.4thecorrespondingdiffractograms

areshownalongwiththediffractionlinesofthereferencephases,

anataseTiO2andPdO.TheXRDpatternsshowabroaddiffraction

bandbetween20◦and30◦2attributedtotheamorphouspartof

thesubstrate.CharacteristicPdOpeaksarepresentinthepatterns

Fig.3. TPD-NH3patternsintherange50–500◦Cofthemesoporoussupports.

ofbothseriesofsamples.Theanatasephaseisvisibleinthe

Ti(IV)-dopedSBA-15samples,butnotintheTi(IV)–dopedHMSseries.

Theanatasephasewasalsoobservedinasimilarlyprepared

Ti-SBA-15materialcontaining2.5at%Ti.DRSUV–visspectraofthis

samplepresentedanabsorptionbandaround334nmduetoTi4+

inoctahedralcoordinationtypicaloftheanatase[30].Generally

speaking,preparationof ion-dopedSBA-15bydirect

incorpora-tionduringthesynthesisofthemesoporousoxideisratherdifﬁcult

duetothestrongacidicmedia[31].Indeed,understronglyacidic

conditions(pH<1)free titaniumspeciessuchasTi4+_exist _only

incationicformandthusarenotabletoentertheframeworkof

SBA-15[32].Thegraftingprocedureusingseveraltypesoftitanium

precursorsallowsinsertingonlyabout6at%oftitaniumrelativeto

siliconintoSBA-15matrix.Highertitanialoadings,8–10at%asin

thepresentcase(usedSi/Tiratiosof25and12),wouldproduce

anatasecrystallites ofTiO2 [33].Inthepresentwork,thedirect

synthesisprocedureadoptedforthepureSBA-15was

intention-allyused,aimingtoinvestigatetheeffectofTiO2eitherformedas

aggregatesandalsoasstructuralmodiﬁer.ByapplyingtheScherrer

equationtothelinewidthofthemainreﬂectionpeaks,the

diam-etersofthePdOandTiO2 particlesinthedifferentsampleswere

estimated.ThecorrespondingcalculatedvaluesarelistedinTable3.

TiO2particlesizeof8nmwereobtainedfortheSBA-15supported

catalysts,whereasPdOparticlesizesof6–7nmwereobtainedfor

allthesamples,includedalsotheHMSsupportedones.

3.1.5. XPSanalyses

Informationonthe chemical stateand the surface chemical

compositionofthepalladiumsamples,beforethecatalytictests,

(5)

Fig.4.XRDpatternsofSBA-15andHMSsupportedpalladiumcatalystswithand withoutTi(IV).ThepatternsofpureanataseTiO2andPdOoxidesareshownas

references.

Table3

XRDderivedPdOandTiO2particlediametersinthefreshsamples.

Sample dPdO(nm) dTiO2(nm)

Pd/SBA-15 6 – Pd/Ti5SBA-15 6 8 Pd/Ti10SBA-15 7 8 Pd/HMS 7 – Pd/Ti5HMS 6 n.d.a Pd/Ti10HMS 6 n.d.a a_n.d.₌_not_detectable.

2p3/2andtheO1sbindingenergiesarecompiledalongwiththe

XPS-derivedatomicratiosTi/SiandPd/(Ti+Si).Thepalladium3d

photoelectronspectraarecharacterizedbythePd3d5/2binding

energyof337.2eV_±0.2eV.ThevalueistypicalofPd4+_as_in_PdO

2.

Fig.5. Ti2pphotoelectronspectraofPdcatalystssupportedon10wt%TiO2doped

mesoporoussilicas.

Thisoxidehasbeenalreadyreportedforsimilarsamples[19,34].

Itsformationmaybeattainedbyoxygenincorporation intothe

PdOcrystallatticeduringcalcination[34].TheTi2pspectraofthe

10wt%Ti(IV)dopedcatalystsareshowninFig.5.Theyare

char-acterizedbythetwospin-orbitcomponents,Ti2p3/2andTi2p1/2,

5.7eVapart.Thethirdcomponentat≈462.5eV±0.5eVincluded

intheﬁttedspectraisattributabletoacontributionfromtheO1s

lineexcitedbytheAlK␤lineofthenon-monochromatized

radi-ation(h=69.7eV)[31].ThebindingenergyoftheTi2p3/2 in

thepureTiO2oxidesis458.9eVtypicalofaTi4+species[30,35].In

accordwithapreviousstudy[19],largervaluesofTi2p3/2binding

energyareobservedforthetitaniaincorporatedsilicas.The

chem-icalshifts,withrespecttothepureTiO2value,rangefrom1.6eVin

theHMSseriesto0.5eVfortheSBA-15series.Twocontributions

mayberesponsibleforsuchsigniﬁcantshift,a“ﬁnalstate”andan

“initialstateeffect”.Astotheformereffect,itgenerallyaccountsfor

theTi2p3/2chemicalshiftwhenmovingfromPd/TixSitoPd/TiO2

[36]. It originatesfrom theextraatomic relaxationenergy [37]

whichisrelatedtothepolarizabilityoftheoxidecarrier.Theless

mobiletheelectronsare,asinSiO2,thesmalleristherelaxation

energyandthereforethelargeristhemeasuredbindingenergyof

thephotoelectrons.Thiseffect,relatedtotheelectricalpropertyof

thematerial,wouldproduceanalogousshiftinthetwotypesof

mesoporoussilicasupportedsamples,thereforeitcouldaccount

forpartofthetotalshift.Thesecondeffect,socalled“initialstate”

effect,discussedalreadyforaseriesofTiO2-graftedSiO2[35,38]and

responsibleforalargeshiftbetweenthetwoseriesofsamples,may

Table4

XPSbindingenergiesandXPS-derivedatomicratiosofthefreshcatalysts.

Sample Pd3d5/2(eV) Ti2p3/2(eV) O1sa(eV) Ti/Sib Pd/(Ti+Si)

Pd/HMS 337.4 533.6 0.02 Pd/Ti5HMS 337.3 460.5 533.4(97%) 0.02 (0.04) 0.02 530.6(3%) Pd/Ti10HMS 337.4 460.3 533.4(94%) 0.04 (0.08) 0.02 530.9(6%) Pd/SBA 337.2 533.4 0.02 Pd/Ti5SBA-15 337.0 459.3 533.5(90%) 0.06(0.04) 0.02 530.6(10%) Pd/Ti10SBA-15 337.0 459.4 533.5(85%) 0.08 (0.08) 0.02 530.8(15%)

a_The_values_in_parentheses_represent_the_atomic_percentages_of_the_oxygen_chemical_components. b_The_values_in_parentheses_represent_the_nominal_atomic_ratios.

(6)

Fig.6. O1sphotoelectronspectraofPdcatalystssupportedon10wt%TiO2doped

mesoporoussilicas.

reﬂecttheoccurrenceintheTi(IV)-dopedsamplesofanintimate

associationofTiO2andSiO2,producingTi–O–SibondswhereTi4+

occupiestetrahedralcoordinationsitessimilartoSiinSiO2 [39].

Sincetitaniumismoreelectropositivethansilicon,theincreaseof

thepositivechargeonthetitaniumcenterofaTi–O–Sibondas

com-paredtoTi–O–Ti,woulddetermineanincreaseoftheTi2pbinding

energy.ThelargershiftobservedintheHMSderivedsamplesisin

accordwiththeXRDresults,suggestingabettertitaniadispersion

inthesesamplesascomparedtotheSBA-15samples.TheO1score

levelspectraofthemixedoxidesupportedcatalystsareshownin

Fig.6andthecorrespondingbindingenergiesarelistedinTable4.

TheO1sspectrawereﬁttedwithtwocomponents,ahighbinding

energyoneat≈533.5eV±0.1eVassignedtoSiO2andalow

bind-ingenergyfeatureat530.7eV±0.2eVclosertothevaluereported

forTiO2[39].Itisworthnoticingthattherelativeintensityofthe

lowenergypeakwithrespecttothehighenergypeakis

signiﬁ-cantlyhigherinthePdTi10SBA-15ascomparedtothePdTi10HMS.

Thisdifferencereﬂectslargersurfaceamountofoxygen

interact-ingwithtitaniumandthereforelargeramountofsegregatedTiO2

atthesurfaceofthePdTi10SBA-15sample.AsshowninTable4,the

experimentalTi/Siatomicratioincreaseswithincreasingloadings

ofTiO2,asexpectedfromthenominalratiosgiveninparentheses.

Itisworthnoticingthelargervaluesofthisratioobtainedforthe

titaniadopedSBA-15samplesascomparedtothecorresponding

titaniadopedHMSsamples.Inaccordwiththeoxygenresults,this

againindicatesapreferentialsurfacesegregationoftitaniuminthe

SBA-15supportedcatalysts.ThePd/(Ti+Si)atomicratioslistedin

Table4arethesameforallsamplesand,aspredictablebecause

oftheadoptedimpregnationprocedure,theyarelargerthanthe

theoreticalvalueof∼0.006.

3.1.6. FT-IR

InFig.7theFT-IRspectraintheOHstretchingregionofthe

activatedHMSandSBA-15samplesarecomparedwiththespectra

of Pd-free and Pd-loaded 10% titania-doped mesoporous

com-posites.ThespectraofHMS andSBA-15exhibitasharpbandat

3744cm−1,assigned toisolated Si–OHspecies[40]. Thedoping

ofHMSwith10wt%TiO2 resultsina shiftoftheOHstretching

vibrationtowardlowerwavenumbersandappearanceofadistinct

shoulderatapproximately3690cm−1.Thedepositionofpalladium

causesdisappearanceoftheabsorptionsat3734and3690cm−1

and regeneration of thesignal at 3744cm−1. This leads tothe

conclusionthattheformertwobandscanbeascribedtotitanol

groupsthatserveasanchoringsitesofthepalladiumspecies.The

Fig.7.FT-IRspectraofthesamplesintheOHstretchingregion.

Pd-freeand Pd-loadedTi10SBA-15 sample spectradisplay very

strongabsorptionintheOHstretchingregionandtheintensities

ofthecorrespondingbandsareoff-scale.Thisindicatesthatthe

OHgroupsinducedbythemodiﬁcationofSBA-15withtitaniaare

Fig.8. (a)FT-IRspectrainthe1700–1400cm−1 _region_of_the_Ti10HMS_sample

recordedafteradsorptionofpyat25◦_C_and_subsequent_desorption_at_various

tem-peratures.(L,Lewisacidity;B,Brønstedacidity;andH,hydrogenbondedpy)(b) FT-IRspectrainthe1700–1400cm−1regionofTi10SBA-15samplerecordedafter adsorptionofpyat25◦_C_and_subsequent_desorption_at_various_{temperatures.}_(L,

(7)

Fig.9.(a)FT-IRspectrainthe1700–1400cm−1_region_of_the_Pd/Ti10HMS_sample

recordedafteradsorptionofpyat25◦Candsubsequentdesorptionatvarious tem-peratures.(L,Lewisacidity;B,Brønstedacidity;andH,hydrogenbondedpy)(b) FT-IRspectrainthe1700–1400cm−1_region_of_the_{Pd/Ti10SBA-15}_sample_recorded

afteradsorptionofpyat25◦_C_and_subsequent_desorption_at_various_{temperatures.}

(L,Lewisacidity;B,Brønstedacidity;andH,hydrogenbondedpy.)

associatedwiththeTiO2 phase whichis inagreementwiththe

XRDandXPSresultsshowinganatasecrystallitesatthesurfaceof

SBA-15.Theshoulderat∼3715cm−1,visibleinthespectraofboth

Ti-containingsupportsisassignedtoterminalTi–OHgroups[41].

Thebroadbandat∼3560cm−1inthespectraofTi10SBA-15and

Pd/Ti10SBA-15samplesisattributedtoH-bondedsilanol/titanol

groups[40].ThisabsorptionisabsentinthespectraofTi10HMS

andPd/Ti10HMSsamples.Thediscrepancyisprobablyrelatedto

somestructuraldifferencesbetweenthetwomesoporoussilicas.

In ordertodiscriminate theBrønstedand theLewis acidity,

theadsorptionofpyridine(py)hasbeenstudiedovertheTi-free

andTi-containingsupports,andforthePdcatalysts.Thespectra

obtaineduponroom-temperatureadsorptionofthebasefollowed

byevacuationatvarioustemperaturesareshowninFigs.8and9for

theTi-dopedsupportsandforthePdcatalystsrespectively.Allthe

spectraincludingthoseoftheTi-freesupports(notshowninhere)

exhibitedbandsat1598and1446cm−1duetonon-acidicsilanols

andcorrespondingtopymoleculesweaklyperturbedbyH-bonding

withthesurfaceOH-groups[29,42–44].Inaddition,aweak

absorp-tionat1626cm−1frompyridiniumiononacidicBrønstedsiteswas

observedalsoinbothTi-freesupports[28,42–44].However,the

protonatedbasewasweaklyboundand disappearedduringthe

evacuationat100◦C.ThedopingoftheHMSandSBA-15supports

bytitaniacaused signiﬁcantchangesin thespectraofadsorbed

py.AsshowninFig.8,thebandsat1603and1462cm−1present

onlyintheTi-dopedsamplesareattributedtothe␯8aand␯19b

ringvibrations,respectively,ofpymoleculescoordinatedtostrong

Lewisacidsites,whilethebandat1584cm−1revealsthepresence

ofweakLewisacidsites[28,42–44].Byevacuationatdifferent

tem-peratures,theintensitiesofthebandsofpycoordinatedtoLewis

acidsites decreasesconsiderably. Smallamount of coordinated

pyremainsafterdegassingat250◦Candpressureof∼10−3_mbar.

Thisindicatesthatthesamplesarecharacterizedbyamoderate

Lewisacidity.ThespectraoftheTi10HMStakenbetween50and

250◦C(Fig.8a)containasplitbroadbandwithmaximaat1648

and1628cm−1whichisattributedtothe␯8aand␯8bvibrations

of protonatedpy indicatingthe presenceof Brønstedacidsites

[28,42–44].Itshouldbenotedthatthebandataround1540cm−1

thatcorrespondstothe␯19bvibrationofpyadsorbedonBrønsted

acidsitesisnotobserved.Thereasonforthiscouldbethelow

con-centrationofsurfacehydroxylswithacidicprotons.However,the

absorptionsat1648and1628cm−1areclearlyvisibleinallofthe

spectratakeninthe50–200◦Ctemperaturerangewhichconﬁrm

thepresenceofcertainamountofweakBrønsted acidityin the

Ti10HMSsample.Pyridiniumionvibrationsat1646and1628cm−1

arenotobservedintheTi10SBA-15sample(Fig.8b)indicatingthe

absenceofBrønstedacidity.Thereforeforthismaterial,theband

at1476cm−1 (whichcanbeassignedtopycoordinatedtoboth

BrønstedandLewisacidsites)isrelatedonlytoLewisacidsites.

Severalmodelshavebeenproposedtoexplaintheappearanceof

acidsitesinamixedoxide.Themostacceptedoneisthemodel

byTanabeassumingthattheextrachargegeneratedbyintimately

mixingtwooxidesproducesacidicsitesofBrønstedorLewis

char-acterdependingonthesignofthischarge[45,46].Negativecharge,

balanced by association withprotons, generatesBrønsted sites

whereaspositivechargegeneratesLewissites.Applyingthismodel

tothecaseofthemixedtitaniaandsilicaoxideintheexcessofsilica,

followingtherulesofthemodel,thechargeexcessatthetitanium

cationwouldbenegative,thereforeproducingBrønstedacidsites.

Theeffectivenessofthemodelimpliesarelationshipbetweenthe

molecularhomogeneityandtheacidsitedensity.Then,the

pres-enceofBrønstedacidsitesontheHMSsamplesmayreﬂectabetter

Si/Tistructuralhomogeneity.Theresultwouldbeinaccordwith

theabsenceofanatasephaseintheXRDpatternsandalsowiththe

largerTi2pchemicalshiftandthelowerintensityofthelowenergy

O1scomponentobservedintheXPspectraoftheHMSsamples.

ThespectraofpyridineadsorbedonthePdpromotedsamples

aregiveninFig.9.ThecomparisonwiththespectraofFig.8

indi-catesthattheincorporationofpalladiumintothesupportslowers

theamountofLewisacidsites,especially inthePd/Ti10SBA-15

sample. Moreover,theabsorptionbandin the1675–1620cm−1

rangeofthespectrumofthePd/Ti10HMScatalyst(Fig.9a),assigned

topyridinecoordinatedtoBrønstedacidsites,becomesbroaderand

strongerthanthatofthesupport.Thissuggeststhatthepromotion

ofTi10HMSwithPdgeneratesadditionalBrønstedacidityandas

aresult,thetotalsurfaceacidityofthePd-containingTi10HMSis

higherthanthatofthePd-freesample.

In order to get preliminarymechanistic information on the

oxidation of methane and look for the formation of

interme-diate species, in situ FT-IR experiments were performed. The

spectra of the palladium catalysts, during 15min of exposure

to a gas mixture containing 15mbar CH4 and 85mbar O2 at

the various temperatures, were collected. The spectra for the

Pd/Ti10HMS andPd/Ti10SBA-15are shown inFig.10.The

pos-itiveabsorptioninthe3690–2700cm−1 rangeindicatesthatthe

amountofH-bondedhydroxylshasincreasedasaresultofwater

formation. The bands at 1635–1627 and 1534–1545cm−1 are

attributedtoadsorbedcarbonatespeciesmostlikelycoordinated

toTisites[15,43]whicharetheintermediatesforproducingCO2.

Inspiteoftheclosesimilaritybetweenthespectraofthetwo

sam-ples,itisworthnoticingthelargerintensityofthepeaksrelated

(8)

Fig.10.FT-IRspectrarecordedduringtheexposureofthePd/10Ti-HMSandPd/10Ti-SBA-15catalyststoamixtureof15mbarCH4+85mbarO2atvarioustemperatures.

spectrumofPd/Ti10HMSsampleascomparedtothesamepeaks

inthespectrum of Pd/Ti10SBA-15sample.No otheradsorption

speciesrelatedtotheactivatedmethaneweredetected.The

pres-enceofadsorbedformatespecies(HCOO−),withtypicalbandsat

3000–2700cm−1(C–Handcombinationsoffundamental

frequen-cies)and at1600–1550cm−1 (as C–O),as intermediatesofthe

reaction,couldnotbeexcluded[15,47].Thegasphasespectra(not

shownhere) indicatedformation ofCO2 withpeak observedat

2350cm−1.

3.2. CatalyticactivitywithSO2-freeandSO2-containingreactant

mixtures

Inordertocomparethecatalyticperformanceofthecatalysts,

T50values(correspondingtotemperaturesof50%CH4conversion)

ofﬁrstandsecondrunsperformedsequentiallyusingdifferent

con-ditionsarelistedinTable5alongwiththeArrheniusparameters.

Firstcyclesonthefreshsampleswereperformedwithpure

reac-tantmixtureandwithSO2containingmixture.Then,inorderto

checkforthermal instabilityand for possibleresidual

contami-nanteffects,secondrunswereperformedonsamplesagedupon

exposuretothepurereactantmixturefor16hat600◦C.Likewise,

aimingtorecoverthecatalystactivityaftertheﬁrstruninthe

pres-enceofthepoisoningSO2,asecondcyclewasperformedusingSO2

–freereagents.

ByinspectionofTable5,judging fromthevalues oftheT50,

thefreshPd/HMSappearsslightlymoreactiveascomparedtothe

Pd/SBA-15.Quiteinterestingthelongexposuretothereactantsat

600◦C,withsomeexceptions,determinesanoveralldecreaseof

theT50valueswhichcouldbeattributedtotheremovalofsome

residualorganicprecursors.Asexpectedandinaccordwith

pre-viousresultsobtainedwithpalladiumsupportedonamorphous

titaniamodiﬁedsilica[19],thepresenceof10ppmofSO2 inthe

reactantmixturedeactivatesthecatalystsbycausingasubstantial

increaseoftheT50withrespecttothecycleswithoutSO2.Within

thesedataitisworthnotingthedifferentbehaviorofthecatalysts

supportedonsilicaHMSandsilicaSBA-15.Theeffectofthesupport

modiﬁcationbytitaniaappearstobestronglyrelatedtothetype

ofthesupport.InthecaseofSBA-15,anincreaseoftitanium

con-centrationimprovestheactivity(lowerT50).Onthecontrary,inthe

caseoftheHMScatalyststheadditionoftitaniuminthesupportis

detrimentalsinceitproducesanincreaseoftheT50.Onepossible

explanationforthesurprisingbehavioroftheHMScatalystcould

bethestrongsupportaciditywhichhindersthepreferential

reac-tionoftheSOxwiththesupport,beneﬁcialforthepalladiumactive

sites.ConsideringthesecondcycleaftertheSO2exposure,the

activ-ityimprovesinallthesamples,asalreadyreportedforpalladium

catalystssupportedonhighsurfaceareasilica[18,19].Theeffect

oftitaniaonthecatalyticperformanceisshowninFigs.11and12,

whereselectedplotsofmethaneconversionasafunctionof

tem-peraturearegivenrespectivelyfortheHMSandtheSBA-15series

ofsupportedPdcatalysts.Thecurvesrefertoaﬁrstcyclewiththe

SO2inthereactantmixtureandtothesubsequentcyclewiththe

SO2freereactantmixture.

Inordertogetinformationontheactivesitevariation,dueto

theoxideTiO2and/ortotheSO2poisoning,theapparentactivation

energiesweredeterminedforthedifferentcatalyticruns.

Neglect-ingtheeffectofsmallamountsofwaterformedintheprocess,the

integralequationofaﬁrstorderreactionwithrespecttomethane

andpseudo-zeroorderwithrespecttooxygenwasusedforthe

calculationof thereaction rateconstants k [18,48].Then, from

Fig.11.MethaneconversionasafunctionoftemperatureoverHMSandTi(IV) promotedHMSPdcatalystsfordifferentcyclesindifferentconditions.

(9)

Table5

ValuesofT50(◦C)andArrheniusparametersa,activationenergy(Eact)andpreexponentialfactorA(s−1)ofthecatalysts,forthedifferentreactioncycles.

Sample 1stcycle 2ndcycleafter16hat600◦C 1stcyclewithSO2 2ndcycle

T50(◦C) Eact(kJ/mol) lnA T50(◦C) Eact(kJ/mol) lnA T50(◦C) Eact(kJ/mol) lnA T50(◦C) Eact(kJ/mol) lnA

Pd/SBA-15 341 77 17 346 71 16 405 125 25 395 85 18

Pd/Ti5SBA-15 302 78 19 308 73 17 383 90 19 325 77 18

Pd/Ti10SBA-15 304 76 20 282 67 21 334 79 18 292 75 18

PdHMS 322 79 18 289 76 18 382 120 24 359 81 18

Pd/Ti5HMS 344 97 21 334 69 16 407 85 18 375 72 16

PdTi10HMSb _n.a. _n.a. ₄₂₂ ₈₁ ₁₆ ₃₈₃ ₇₃ ₁₄

a_E

actandAarecalculatedfromtheArrheniusplotinthetemperaturerange250–350◦C. b_n.a._indicates_that_the_tests_for_this_sample_were_not_performed.

theArrheniusplotofln(k)versus1/T,inthetemperaturerangeof

250–350◦Ccorrespondingtoconversionsrangingbetween10%and

80%,theactivationenergiesEactandthepre-exponentialfactorsA

oftheArrheniusequation,k=Aexp(−Eact/RT),werecalculatedand

listedalsoinTable5.Withsomeexceptions,thevaluesofthe

activa-tionenergiesarebetween70and85kJ/mol,whichareclosetothose

reportedpreviouslyforsimilarcatalystsforthesametemperature

range[19].Acarefulinspectionofthetablegivesinterestinghints

aboutthestructuralchangesuponthechemicalreactions.Although

mostofthevariationsarewithintheexperimentalerror(±10%),

somesystematicchangesareenvisagedwhichcanberelatedto

structuralmodiﬁcationduetothereaction.Afterathermal

treat-mentat600◦Cfor16h,alowervalueoftheactivationenergyis

observed,correspondinginsomecasestoanenhancementofthe

activity(lowerT50).Asreportedbefore,thepresenceofSO2inthe

reactantmixturecausesanincreaseoftheactivationenergy,quite

remarkableinthesampleswithoutTiO2inthesupport[18].Then,

forallthesamples,thesubsequentcycleaftertheSO2exposureis

characterizedbyloweractivationenergyandbyapartialrecovery

oftheactivity.Theactivityiscompletelyregainedinthecaseofthe

TipromotedSBAsupportedcatalysts,butonlypartiallyrecovered

inthepureSBA-15,HMSandTi-dopedHMSsupportedcatalysts.

Theincreaseoftheactivationenergyduringtheruninthe

pres-enceofSO2conﬁrmedtheearlierideathatsulfurdeactivationwas

relatedtotheformationofacompositespeciesofloweractivity,

likePdO–SO3,quitelabileandeasilydecomposingattemperature

≥450◦_C_[18,19]_.

3.3. Stability,deactivationandstructuralchangesof

Pd/Ti10SBA-15

Themostinterestingsample,thePdTi10SBA-15exhibitingthe

lowestT50duringthecyclewithSO2andduringthesecondcycles,

Fig.12.MethaneconversionasafunctionoftemperatureoverSBA-15andTi(IV) promotedSBA-15Pdcatalystsfordifferentcyclesindifferentconditions.

Fig.13.CH4conversionasafunctionoftemperatureoverPd/Ti10SBA-15for

differ-entcyclesindifferentconditions.

wasselectedforaseriesofconsecutivecyclesinordertotestits

stability, deactivationand activityrecovery.Particularly, aftera

ﬁrstcyclewiththeSO2 containingreagentsand asecondcycle

withSO2-freereagents,thesampleunderwentanovernight

treat-mentinﬂowing10vol.ppmofSO2inHeat350◦Cataﬂowrateof

50ml/min.Afterthistreatmentthreemorecycleswiththepure

reagentswere performed.Alltherelated conversioncurvesare

given in Fig.13including alsothetwo curves (I cycleSO2 and

IIcycle)givenalreadyinFig.2.Thecorrespondingvalues ofT50

and Arrheniusparametersare listedinTable6. Forthesake of

claritysomeofthevaluesareduplicatedfromTable5.A

remark-ableincreaseoftheT50isobservedafterthenighttreatmentwith

SO2.Atthesametime,anincreaseoftheactivationenergyupto

114kJ/molisobtained.Inthefollowing4thand5thcycles,the

activ-ityiscompletelyrecovered.Itisremarkablethattheﬁnalvalues

oftheactivationenergies(95–94kJ/mol),althoughdecreasedwith

respecttothevalueobtainedaftertheseveredeactivation,donot

reachthelowestvalueof67kJ/molobtainedforthesecondcycle.

Thesigniﬁcantdifference,wellabovetheexperimentalerror,could

reﬂectsomestructuralchangesoftheactivesiteand,aslongasthe

Table6

ValuesofT50 (◦C)andArrheniusparametersa,activationenergy(Eact)and

pre-exponentialfactorA(s−1₎_of_the_{Pd/Ti10SBA-15}_catalyst_for_the_different_reaction

cycles.

Reactioncycle T50(◦C)Eact(kJ/mol)lnA

1stcycle 304 76 20

2ndcycleafter16hexposuretothereactionmixtureat600◦282C 67 21

1stCycleSO2 334 79 18

2ndcycleSO2-free 292 75 18

3rdcycleafterovernightSO2treat. 375 114 23

4thcycle 287 95 22

5thcycle 286 94 22

a_E

act andAarecalculatedfromtheArrheniusplotinthetemperaturerange 250–350◦C.

(10)

Table7

XRDderivedPdOandTiO2particlediametersandXPPd3d5/2bindingenergyandTi/SiandPd/(Ti+Si)atomicratiosinspentcatalysts.

Sample dPdO(nm) dTiO2(nm) Pd3d5/2(eV) Ti/Si Pd/(Ti+Si)

PdTi10SBA-15after5thcycleincludingSO2nighttreat. 18 7 337.0 0.08 0.005

PdTi10SBA-15after4daysreactionat600◦_C ₁₆ ₈ _337.3 _0.07 _0.005

ﬁnalactivityisconcerned,iswellcompensatedbyanincreaseof

thepre-exponentialfactor.Onthesamesamplebutona

differ-entbatchastabilitytestwasperformedbyrunningacatalytictest

afterexposingthesampletothepurereagentmixturefor4days

at600◦C.Asrevealedbytheconversioncurve(notgiveninhere),

matchingexactlythecurveofthesecondcycle,(Table6secondrow

data)nochangeinthecatalyticperformanceoccurred.

Inordertocheckforstructuralmodiﬁcation,thetwoaged

sam-pleswereanalysedbySAXS,XRDandXPStechniques.According

totheSAXSpatterns(notshownhere) theSBA-15mesoporous

structureoftheagedcatalystswasmaintained.TheXRDpatterns

ofthefreshsampleandthecorrespondingpatternsofthe

sam-pleafter5cyclesincludingtheSO2overnighttreatment(referto

Table6)andthesampleafterthelongrunof4daysat600◦Care

comparedinFig.14.ThepatternscontaintheTiO2anatasepeaks

andthePdOrelatedpeaks.ThecalculatedTiO2andPdOparticles

sizearegiveninTable7.ItisworthnoticingthatwhereastheTiO2

particlesmaintainthesamesize,thePdOparticlesizedrastically

increasesafterthelongrunwiththepurereagentmixturesand

alsoafterﬁvecyclesincludingtheovernightexposuretotheSO2at

350◦C.TheincreaseofthePdOparticlesizeuponsimilarsequence

ofreactioncycleswasalsoobservedrecentlywithapalladium

cat-alystsupportedonamorphousTi10SiO2[20].Itshouldbepointed

outthattheXRDtechniqueidentiﬁesthelargerparticles,whereas

smallparticlesizes,belowthelimitof3nm,arenotdetected.Then,

inordertoultimatelyconﬁrmthedecreaseofpalladium

disper-sion,XPSanalysesofthesamples,afterseveralcyclesincluding

theovernightSO2exposureandafterthestabilitytests,were

per-formed.TheXPSresultsofthespentsamplesarelistedinTable7.

Ascomparedtothecorrespondingvaluesofthefreshsample,as

giveninTable4,nosigniﬁcantchangesinthepalladiumbinding

energyand intheTi/Siatomic ratioareobserved.Howeverthe

smallervaluesofthePd/(Ti+Si)atomicratiosascomparedtothe

freshsampleareinaccordwiththeenlargementofthePdO

par-ticlesizerevealedbytheXRD.Itisworthnoticingthatinspiteof

thisseriouspalladiumsinteringthecatalyticactivityofthe

cata-lystwasmaintained.Arecentstudyonpalladiumsupportedon

silicaSBA-15,modiﬁedbyceriaandbyzirconia,reportedaloss

ofactivity, upon a severedeactivation test, for catalysts

show-Fig.14.XRDpatternsofPdTi10SBA-15catalystasfresh,after5thcycleincluding overnightSO2treatmentat350◦Candafterfourdaysofcatalyticreactionat600◦C

(longrun).

inggoodPdOparticlestability[13].Theconclusionofthework

wasthatthelossofcatalystactivityduringreactionwasnot

nec-essarilyrelated tothesinteringof theactivesitesbutratherto

thevariationoftheoxygenstoragecapacityandoxygenmobility

strictlyrelatedtothesupportproperties.It isacceptedthat

Pd-catalysedcombustionofmethaneoccursthrougharedoxorMars

vanKrevelenmechanism[17,49].Accordingly,duringthisreaction

PdOislocallyreducedtoPdbymethane,producingH2OandCO2,

andthenPdisreoxidisedbyoxygen.Althoughtheactivespecies

isconsideredPdO,itiswidelyrecognizedthatmetallicPdplays

theimportantrole in decomposingandactivatingthemethane

molecule[3,15,50].Therefore,theefﬁciencyoftheallprocessis

relatedtotheredoxpropertiesofthecatalystandtotheoxygen

mobility[17].Inthepresentcase,themaintenanceofthe

activ-ity,inspiteofthesigniﬁcantpalladiumoxidesintering,couldbe

attributedtothegoodinteractionbetweensurfacetitaniaand

sil-ica.Such interaction,asreportedin previouspapers,favorsthe

formationof Ti–O–Silinkageswithanincrease oftheoxidizing

potentialoftheTi(IV)cations[38,51].Theincreaseofthispotential

wouldenhancetheoxygenmobilityattheinterfacebetweenthe

Ti–O–Siunitsandthepalladiumparticles[38,49].Onthisaccount,

thestabilityofthePd/Ti10SBA-15catalyst,observedafteralong

lasting reactionat 600◦C and afterseveralcyclesincludingthe

overnightSO2 exposure,islikelyrelatedtotheoxygenmobility

whichallowsthereoxidationofthePdtoPdO.Inthepresenceof

asupportmetal—interaction,theincreaseofthePdOparticlesize

afterthelongreactionisnotdetrimental.Indeed,aninverseparticle

sizestructuresensitivityofthepalladiumcatalysedmethane

com-bustionwasclaimed,relatedtothehighstabilityofthesmallPdO

particlesinteractingwiththesupportandnotbeingeasilyreduced

[13,50].

4. Conclusion

Theeffectoftheadditionoftitaniatodifferentmesoporoussilica

oxidesonthemethaneoxidationactivityofsupportedpalladium

wasfoundtodependonthetypeofmesostructuredmaterial.The

additionof10wt%ofTiO2tothesilicaSBA-15,improvedthe

cat-alyticperformanceofthesupportedpalladiumcatalysts,intermsof

betteractivityandbettersulfurtoleranceascomparedtopalladium

overpureSBA-15support.Thisresultconﬁrmedthepositiveeffect

playedbytitania,actingasSO2scavenger,onPdcatalystssupported

onamorphoussilica[19].Oppositebehaviorwasobservedwiththe

HMSsamples.InthiscaseadditionofTiO2producedpoorer

cata-lystsintermsofloweractivityandlowerSO2tolerance.According

tothestructuralandmorphologicalanalyses,thetwoseriesof

cat-alystshadsimilarpalladiumdispersionandsimilarsurfaceareabut

differentTiO2distribution.InthecaseoftheHMS,Ti(IV)ionswere

homogeneouslydistributedintothesilicaframework,whereasa

mixtureofframeworkTi(IV)andsurfacesegregatedTiO2anatase

particleswaspresentintheSBA-15.Duetothemoreuniform

dis-tributionofTi(IV)intothesilicamatrix,asubstantialincreaseof

thetotalacidity,inparticularoftheBrønstedtype,developedin

theTixHMSsupportascomparedtotheSBA-15.Moreover,through

theinsituFT-IR,moreadsorbedwaterandcarbonatespecieswere

observedonthePd/Ti10HMSascomparedtothePd/Ti10SBA-15.

Onthebasesoftheseevidences,theformationofastrongeracidity

andthepresenceofBrønstedsiteswereresponsibleforthe

(11)

HMSsupportedPdcatalysts.Indeed,thestrongerwateradsorption

overtheacidicsupportwouldinhibitthecatalyticmethane

oxida-tionfavoringformationofthelessactivePd(OH)2andalsoblocking

theactive sites through surface diffusion.The large amountof

adsorbedcarbonatespecies,detectedinthePd/Ti10HMScatalyst

bytheinsituFTIRanalysesat350◦C,couldhavealsocontributed

to the deactivation of the catalysts. Furthermore, the creation

ofthestrongeracidity,byhinderingthepreferentialreactionof

theSOx withthesupport,nulliﬁed thescavenger action ofthe

titania.

Acknowledgements

This research hasbeen performed in theframework of the

D36/003/06COSTprogramandaNATOgrantESP.CLG.No.984160.

References

[1]P.Gelin,M.Primet,Appl.Catal.B39(2002)1–37.

[2]F.Arosio,S.Colussi,G.Groppi,A.Trovarelli,Catal.Today117(2006)569–576. [3]T.V.Choudhary,S.Banerjee,V.R.Choudary,Appl.Catal.A234(2002)1–23. [4] R.Burch,M.J.Hayes,J.Mol.Catal.A100(1995)13–33.

[5]Y.-x.Pan,C.-j.Liu,P.Shi,Appl.Surf.Sci.254(2008)5587–5593.

[6]L.F.Liotta,G.DiCarlo,G.Pantaleo,A.M.Venezia,G.Deganello,Appl.Catal.B66 (2006)217–227.

[7] N.Bahlawane,Appl.Catal.B67(2006)168–176. [8]Z.Gao,R.Wang,Appl.Catal.B98(2010)147–153.

[9]E.Tzimpilis,N.Moschoudis,M.Stoukides,P.Bekiaroglou,Appl.Catal.B87 (2009)9–17.

[10] Z.Jiang,Z. Hao,J. Su,T.Xiao,P.P.Edwards, Chem.Commun. 22(2009) 3225–3227.

[11]J.K.Lambert,M.S.Kazi,R.Farrauto,J.Appl.Catal.14(1997)211–223. [12]H.Yoshida,T.Nakajima,Y.Yazawa,T.Hattori,Appl.Catal.B71(2007)70–79. [13] F.Yin,S.Ji,P.Wu,F.Zhao,C.Li,J.Catal.257(2008)108–116.

[14] K.Persson,K.Jansson,S.G.Järas,J.Catal.245(2007)401–414.

[15]M.Schmal,M.M.V.M.Souza,V.V.Alegre,M.A.PereiradaSilva,D.V.Cesar,C.A.C. Perez,Catal.Today118(2006)392–401.

[16] N.M.Kinnumen,M.Suvanto,M.A.Moreno,A.Savimaki,K.Kallinen,T.-J.J. Kin-nunen,T.A.Pakkanen,Appl.Catal.A370(2009)78–87.

[17]K.Fujimoto,F.H.Ribeiro,M.A.Borja,E.Iglesia,J.Catal.179(1998)431–442. [18]A.M.Venezia,R.Murania,G.Pantaleo,G.Deganello,J.Catal.251(2007)94–102. [19] A.M.Venezia,G.DiCarlo,G.Pantaleo,L.F.Liotta,G.Melaet,N.Kruse,Appl.Catal.

B88(2009)430–437.

[20]G.DiCarlo,G.Melaet,N.Kruse,L.F.Liotta,G.Pantaleo,A.M.Venezia,Chem. Commun.46(2010)6317–6319.

[21]N.Marin-Astorga,G.Pecchi,T.J.Pinnavaia,G.Alvez-Manoli,P.Reyes,J.Mol. Catal.A247(2006)145–152.

[22]P.T.Tanev,M.Chibwe,T.J.Pinnavaia,Nature368(1994)321–323.

[23]D.Zhao,J.Feng,Q.Huo,N.Melosh,G.H.Fredrickson,B.F.Chmelka,G.D.Stucky, Science279(1998)548–552.

[24]S.J.Gregg,K.S.Sing,Adsorption,SurfaceAreaandPorosity,seconded.,Academic Press,SanDiego,1982.

[25] JCPDSPowderDiffractionFileInt.CentreforDiffractionData,Swarthmore, 1989,FileNo.42-1467.

[26] H.P.Klug,X-rayDiffractionProcedureforPolycrystallineandAmorphous Mate-rials,Wiley,NewYork,1954.

[27] V.LaParola,B.Dragoi,A.Ungureanu,E.Dumitriu,A.M.Venezia,Appl.Catal.A 386(2010)43–50.

[28] A.Tuel,Micropor.Mesopor.Mater.27(1999)151–169.

[29]W.Zhang,M.Fröla,J.Wang,P.T.Tanev,J.Wong,T.J.Pinnavaia,J.Am.Chem.Soc. 118(1996)9164–9171.

[30]T.A.Zepeda,Appl.Catal.A347(2008)148–161.

[31] N.N.Trukhan,V.N.Romanikov,A.N.Shmakov,M.P.Vanina,E.A.Paukshtis,V.I. Bukhtiyarov,V.V.Kriventsov,I.Yu.Danilov,O.A.Kholdeeva,Micropor.Mesopor. Mater.59(2003)73–84.

[32] W.Li,S.J.Huang,S.B.Liu,M.O.Coppens,Langmuir21(2005)2078–2085. [33]Z.Luan,E.M.Maes,P.A.W.VanderHeide,D.Zhao,R.S.Czernuszewics,L.Kevan,

Chem.Mater.11(1999)3680–3686. [34]Y.Bi,G.Lu,Appl.Catal.B41(2003)279–286.

[35]E.Santacesaria,M.Cozzolino,M.DiSerio,A.M.Venezia,R.Tesser,Appl.Catal. A270(2004)177–192.

[36]G.Lassaletta,A.Fernandez,J.P.Espinòs,A.R.Gonzalez-Elipe,J.Phys.Chem.99 (1995)1484–1490.

[37]C.D.Wagner,FaradayDiscuss.Chem.Soc.60(1975)291–300.

[38]A.M.Venezia,F.L.Liotta,G.Pantaleo,A.Beck,a.Horvath,O.Geszti,A.Kocsonya, L.Guczi,Appl.Catal.A310(2006)114–121.

[39] R.Castillo,B.Koch,P.Ruiz,B.Delmon,J.Catal.161(1996)524–529. [40]N.N.Trukhan,A.A.Panchenko,E.Roduner,M.S.Mel’gunov,O.A.Kholdeeva,J.

Mrowiec-Biało ´n,A.B.Jarz ˛ebski,Langmuir21(2005)10545–10554.

[41]M.I.Zaki,M.A.Hasan,F.A.Al-Sagheer,L.Pasupulety,ColloidsSurf.A190(2001) 261–274.

[42]M.Hunger,U.Schenk,M.Breuninger,R.Gläser,J.Weitkamp,Micropor. Meso-por.Mater.27(1999)261–271.

[43]G.Busca,Catal.Today41(1998)191–206.

[44] R.S.Araujo,D.A.S.Maia, D.C.S.Azevedo, C.L.CavalcanteJr.,E. Rodrıguez-Castellon,A.Jimenez-Lopez,Appl.Surf.Sci.255(2009)6205–6209. [45]K.Tanabe,T.Sumiyoshi,K.Shibita,T.Kiyoura,J.Kitagawa,Bull.Chem.Soc.Jpn.

47(5)(1974)1064.

[46] J.B.Miller,E.I.Ko,Catal.Today35(1997)269–292. [47] G.Busca,V.Lorenzelli,Mater.Chem.7(1982)89–126.

[48]L.J.Hoyos,H.Praliad,M.Primet,Appl.Catal.A98(1993)125–138.

[49]C.A.Muller,M.Maciejewski,R.A. Koeppel,A.Baiker,J. Catal.166 (1997) 36–43.

[50] S.C.Su,J.N.Carstens,A.T.Bell,J.Catal.176(1998)125–135. [51]X.Gao,I.E.Wachs,Catal.Today51(1999)233–254.