The roles of surface-doped metal ions (V, Mn, Fe, Cu, Ce, and W) in the interfacial behavior of TiO2 photocatalysts

ContentslistsavailableatScienceDirect

Applied

Catalysis

B:

Environmental

jo u r n al ho m e p ag 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

The

roles

of

surface-doped

metal

ions

(V,

Mn,

Fe,

Cu,

Ce,

and

W)

in

the

interfacial

behavior

of

TiO

2

photocatalysts

Sue-min

Chang

,

Wei-szu

Liu

InstituteofEnvironmentalEngineering,NationalChiaoTungUniversity,1001,UniversityRoad,Hsinchu30068,Taiwan

a

r

t

i

c

l

e

i

n

f

o

Articlehistory:

Received15December2013 Receivedinrevisedform8March2014 Accepted22March2014

Availableonline2April2014 Keywords:

Surfacedoping Photocatalyticactivity Electronicstructures Chargerecombination Interfacialchargetransfer

a

b

s

t

r

a

c

t

Sixtypesoftransitionmetalions,includingV,Mn,Fe,Cu,Ce,andW,aredopedintothesurfacelattice ofTiO2 powders,andtheirrolesinthechargetrapping,recombination,interfacialtransfer,and

pho-tocatalyticactivityaresystematicallystudied.Thesurface-dopedTiO2powdersexhibitphotocatalytic

activityintheorderofFe/TiO2>Cu/TiO2>V/TiO2>W/TiO2>Ce/TiO2>Mn/TiO2.WhiletheFe,CuandV

ionsimprovetheactivity,theW,Ce,andMnionscausedetrimentaleffects.Thedifferentinfluencesare associatedwiththeirenergylevels,coordinationnumbersandelectronegativity.Thesurface-dopedions trapchargecarriersandinteractwithadsorbatestoprovidealternativepathwaysforinterfacialcharge transfer.TheFeandCuionsinhibitdefect-mediatedannihilation,facilitatinginterfacialchargetransfer intermsofd–dtransitionsandthermallyinducedde-trapping.TheMnions,whichintroduceboth occu-piedandunoccupiedstatesinthemid-band-gapregion,incontrast,trapholesandelectronstoseverely consumechargecarriersviaintra-atomicrelaxation.TheCeandWions,whichhavehighcoordination numbersandelectronegativity,stronglybondtheO2−radicals,thuslimitingchargeutilizationaswellas

photocatalyticperformance.

©2014ElsevierB.V.Allrightsreserved.

1. Introduction

Photocatalysts, which absorb photons to promote chemical reactions,havebeenconsideredasimportantgreenmaterialsfor environmental and energy applications [1,2]. Titanium dioxide (TiO2)isthemostusedphotocatalystbecauseithashigh chemi-calstability,lowtoxicity,andsolarsensitivity.Inaddition,itshigh densityofstatesinbandsenablesefficientphoton-to-current con-version,and leadstoTiO2 catalysts becomingmoreactive than othersemiconductors[3].Photocatalyticactivitycorrelatestothe numberofchargecarrierswhichsuccessfullyescapefrom recom-binationand transfertotheadsorbates acrosstheinterface.To maintainahighlevelofeffectivechargecarriers,dopingtransition metalionsintotheTiO2latticetoinhibitrecombinationhasbeen extensivelystudied[4,5].

Conventionaldopingincorporatesionshomogeneouslyintothe bulklattice.Thedopedionsintroduceadditionalenergylevelsinto thebandstructure which areabletotrapelectronsor holesto

∗ Correspondingauthor.Tel.:+88635712121x55506;fax:+88635725958. E-mailaddress:chang@mail.nctu.edu.tw(S.-m.Chang).

1 _{Presentaddress:DepartmentofMaterialsScienceandEngineering,National}

TsingHuaUniversity,101,Section2,Kuang-FuRoad,Hsinchu,30013,Taiwan.

separate charge carriers from the bands, thus allowing more chargecarrierstosuccessfullydiffusetothesurface[5]. Depend-ingonthe typesand thechemical states, transitionmetalions providedifferentcontributionstothephotocatalyticactivity.Xu etal. [4]improvedthephotocatalytic activityofTiO2 nanopar-ticles by doping seven types of rare-earth ions individually into the TiO2 matrix, and the enhancement is in the order of Gd3+_>Nd3+_>La3+_>Pr3+_(Er3+_)>Ce3+_>Sm3+_{.Choietal.[5]reported} thatdopingTiO2 colloids withFe3+,Mo5+,Ru3+,Os3+,Re5+,V4+, andRh3+_{ionsimprovesphotoactivity,whereasCo}3+_andAl3+_ions reducetheactivity.Thedopantswithclosed-shellconfigurations areconsideredtohavelittleeffectontheactivity.Incontrast,good dopantsintroducetheirenergy levelsintheconduction/valence bands.Suchdopantsnotonlytrapchargecarriersinorderto sup-pressband-to-bandrecombination,butalsoreleasethetrapped carriers viathermaltransitionsoastomaintaina highlevelof effectivechargecarriers.Similarde-trappingofshallowlytrapped chargecarriers alsotakesplace inthedopantswiththeenergy levelsclosetothebandedges[7].

Althoughtheimprovedactivityofbulk-dopedTiO2particleshas beendemonstrated,manystudieshaveindicatedthatdetrimental effectsarethegeneralconsequenceofdopingbecausethetrapped charge carriers quickly recombine with the counter carriers, irrespectiveofshallowordeeptrapping[6,8,9].Thecontroversial http://dx.doi.org/10.1016/j.apcatb.2014.03.044

Fig.1.Thecross-sectionalviewofsurface-dopedTiO2particles.

effectsofdopingonthephotocatalyticperformanceareattributed tothedifferentdopingamountsandthedifferentsizesofdoped photocatalysts.Withtheappropriateamountsofdopants,small sacrificesinchargecarriersreducethecarrierdensityinthebands. The dopants thus inhibit band-to-band relaxation and allow a largefractionoftheremainingcarrierstosuccessfullydiffuseto thesurface. The optimal loadingfor the highest photocatalytic performance is particle-size dependent. According to reported data, a model developed by Bloh et al. indicates that optimal doping amounts decrease with increasing particle sizes [10]. Heavydopingisunfavorableforlargeparticlesbecauseitinduces moredefect-mediatedrecombinationduringthelongjourneyof thecarriers tothe surface[11–13]. In contrasttothe annihila-tioninsidethebulkparticles,thetrappedchargecarriersinthe quantum-sizedparticlesareabletoreachtheinterfacebecause theirwavefunctionspreadsovertheentirecatalystcluster[5].

Toenablethetrappedchargecarrierstoeffectivelyparticipate inthesurfacereactions,Vionsweredopedintothesurface lat-ticeofmicrometer-sizedTiO2 particlesinapreviousstudy[14]. Inaddition,theinfluencesofsurfacedopingandbulkdopingon photocatalyticactivitywereclarified.ItwasfoundthattheVions withinthesurfacelatticesignificantlyenhancetheactivityasthe concentrationincreases.Incontrast,unlesstraceamountsofVions arepresent,bulkdopingreducestheactivity.Thesurface-doped ionscreateaninternalelectricfield todrivethecharge carriers thataredriftingfromthebulk tothesurfaceandsuppress sur-facerecombinationviatrapping,thuspromotingchargeutilization. Similarimprovements where surface dopingcontributes tothe activityarealsodemonstratedintheP5+_-andZr4+_-dopedTiO

2,the Zn2+_-dopedSnO

2andtheBa-dopedCd0.8Zn0.2Spowders[15–18]. Unlikethedopantswithinthebulklattice,thesurface-dopedions areinvolvedintheadsorptionandinterfacialchargetransferin additiontotrapping.Therolesofdifferentsurface-dopedionsinthe interfacialbehaviorsandtheircontributionstothephotocatalytic activityarestillunclear.

Inthisstudy,weseparatelyincorporatesixdifferenttypesof transitionmetalions,includingV,Mn,Fe,Cu,Ce,andW,intothe TiO2surfacelattice,andthephotocatalyticactivityofthe surface-dopedTiO2powdersaresystematicallyexamined.Surfacedopingis carriedoutbycoatingas-preparedTiO2particleswithathindoped filmfollowedbycalcination.Fig.1showsthecross-sectionalview ofthesurface-dopedTiO2samples.Toexplorethedifferent contrib-utionstotheactivitybythesurface-dopedions,wedeterminethe chargetrapping,recombination,andinterfacialchargetransferat thesurfaces.Therolesofthedopantsintheinterfacialbehaviorsare elucidatedwithrespecttotheirelectronicstructures,coordination numbersandelecotronegativity.

2. Experimental

2.1. SurfacedopingofTiO2photocatalystswithtransitionmetal ions

Bothpureandsurface-dopedTiO2sampleswerepreparedusing sol–gelprocesses, in which titanium isopropoxide (TTIP, Acros, 98%)compoundswereusedastheprecursorfortheTiO2powders. TheTTIPliquidwasdilutedwithisopropanol(IPA,J.Backer,99.5%) toreacha molarratioof TTIP/IPA=1/30.Theprecursor solution washeatedat150◦Cunderambientconditionsfor3htoinduce hydrolysisandcondensation.Thesolidwasthencalcinedat500◦C for3htoobtainpureTiO2powders.Surface-dopedTiO2powders werepreparedbycoatingtheas-driedTiO2powders(2.0g)witha thindopedTiO2film.TheTiO2powdersweredispersedina coat-ingsolutionthatcontainsTTIPandtransitionmetalions(M)inIPA atamolarratioofM/TTIP/IPA=0.2/1.0/1200.Sixtypesof transi-tionmetalions,VO(OC3H7)3(VTIP,Aldrich,99%),Mn(NO3)2·4H2O (Merck),Fe(NO3)3·9H2O(Riedel-deHaën),CuCl2·2H2O(SHOWA), WCl6 (Aldrich),andCe(NO3)3·6H2O(SHOWA),wereusedasthe dopants. To completely coat the TiO2 particles with the coat-ingsolutions,thesuspensionsweretreatedusingsonicationfor 10min.Then,thepowderswereseparatedfromthecoating solu-tionthroughcentrifugationat15,000rpmfor3min.Afterremoving thesupernatant,theremainingsolutionwasfurtherdrainedusing suction.Simultaneously,theadsorbedprecursorsbegantoreact withthewatervaporintheambientairtoformadopedthinfilm ontheTiO2surface.Priortocalcinationat500◦Cfor3h,thesamples weredriedat100◦Ctostabilizethestructureofthefilms.Doping concentrationswereadjustedbychangingtheM/TTIPratiosinthe coatingsolution.

2.2. Characterizations

Themorphologyofthesurface-dopedTiO2powderswas char-acterized usinga transmission electron microscope(TEM, JOEL JEM-3000F) operated at an acceleration voltageof 300kV. Ele-mentalline-scanningandmappinganalysisweremeasuredusing an energy-dispersivespectrometer (EDS)equippedin the TEM. The chemical states of thedopants and thechemical composi-tionswereidentifiedusinganX-rayphotoelectronspectrometer (XPS,PHI1600)operatedwithanAlK␣radiation(1486.6eV).The photoelectrons werecollectedin the analyzerat a passenergy of 23.5eVand ata collectioninterval of 0.1eV.Chemicalshifts resultingfromthechargingeffectswerecalibratedbyfixingthe C1speakofthesurfacecarbonaceouscontaminantsat284.8eV. The crystalline structures of the TiO2 powders were examined usinganX-raypowderdiffractometer(XRPD,MACScience,MXP18) withaCuK␣radiation(=1.5406 ´˚A),anacceleratingvoltageof 30kV, and an emission current of 20mA. The diffraction pat-ternswererecordedatthe10–80◦2positionswithasampling widthof 0.03◦ and a scanningspeedof 10◦/min.The electronic structuresof thesampleswereanalyzedusinganUV–vis spec-trometer (HITACHI 3010) in diffused reflectance mode with a scanning range from 900to 200nm. Aluminum oxide (Al2O3), whichwasconsideredtodelivertotalreflection,wasusedasthe referenceforallthemeasurements.Thediffusedreflectancewas thenconvertedintoabsorptionsaccordingtotheKubelka–Munk formula[19].Nitrogenadsorption/desorptionisothermswere mea-suredusingagassorptionanalyzer(Micromeritics,Tristar3000)at 77K.TheBrunauer–Emmett–Teller(BET)modelwasusedto esti-mate thesurfaceareas ofthesamplesbased ontheabsorption data.

2.3. Chargetrappingandinterfacialtransfer

Electronparamagneticresonance(EPR)spectrawererecorded usingaBrukerEMXspectrometeroperatedatanX-bandfrequency. A500WXelamp(UshioInc.),withitsmajoroutputwavelength setto365nm,servedasthelightsourcefortheactivationofthe photocatalystsandwaspositionedatafixeddistancefromasample cavity.Priortotheanalysis,thepowdersinthesampletubeswere evacuatedat120◦Cfor2htoremoveadsorbedwaterfromthe sur-face.Thespectraofthetrappedchargesinthephotocatalystswere recordedat77Kunderavacuumbothinthedarknessandwith UVillumination.Toexaminetheinterfacialchargetransferfrom thesurfacetotheadsorbates,thesametubesweresubsequently filledwithO2gas,andthespectraofthesamplesintheO2 atmo-spherewereacquiredusingthesameprocesses.Theinstrumental conditionsweresetatacenterfieldof3050Gandasweepwidth of6000.0G.Themicrowavefrequencywas9.65GHzwithapower of15.0mW.

2.4. Photocatalyticactivity

Thephotocatalyticactivitiesofthepureandthesurface-doped TiO2powderswereexaminedbasedonthedegradationof bisphe-nolA(BPA,20mg/L)inanaqueoussolutionundertheirradiation of305nmUVlight.Thecatalystsweresuspendedultrasonicallyin theBPAsolutionsinfused-silicatubes.Priortoirradiation,the sus-pensionwaspurgedusingO2gasindarknessaccompaniedwith stirringfor30mintoequilibratetheadsorptionanddesorptionof BPAandtosaturatethesolutionwithO2.Thedegradationofthe BPAmoleculeswasmonitoredbysamplingthesuspensionat differ-entintervalsofirradiationtimeandanalyzingtheconcentrationof themoleculesintheremainingsolutionusingahighperformance liquid chromatograph (HPLC, Waters Alliance 2695) equipped witha C18column(5␮m,4.6mm×250mm)and aphotodiode array detector (PDA, Waters 2996). The mobile phase was a methanol–watermixture(80/20,V/V)ataflowrateof1.0mL/min.

3. Resultsanddiscussion

3.1. Bulkcharacteristicsofthesurface-dopedTiO2photocatalysts Fig.2showsTEMimage,EDSelementalline-scanningand map-pingresultsfortheFe-dopedTiO2particles.Thesizesandshapes ofthesol–gel-derivedTiO2powderswereirregular,andmostwere ata nanometerscale.High concentrations ofFeelementswere measuredintheedgesoftheparticlesatathicknessof10–20nm. Thisresultrevealsthesurface-dopedfeatureofthesample.Other dopedpowdersexhibitedsimilarmorphologies,buthadthinner dopedlayers (<10nm).The bulk properties,includingthe crys-tallinestructures,theelectronicstructuresandthesurfaceareas,of thesampleswereexaminedusingXRD,UV–visspectroscopy,and N2 adsorption–desorption measurement, respectively (see Sup-plementarymaterials,TableS1).PureTiO2powdersexhibitedan anatasephasewithacrystallitesizeof16.5nm.Thebandgapof theTiO2particleswas3.28eV,whichisclosetothatofbulkanatase TiO2crystals[20].Surfacedopingwithtransitionmetalions, includ-ingV,Mn,Fe,Cu,Ce,andW,hadlittleeffectonthebulkcrystalline structures,andsimilarsizedanataseTiO2crystals(14.8–16.8nm) werealsomeasuredinthesurface-dopedpowders.An insignifi-cantshiftinthe(101)diffractionpeak(2=25.29–25.32)indicates thatthedopedionsmainlyremainwithinthesurfacelattice,and thermally-inducedinwardmigrationisunfavorable(see Supple-mentarymaterials,Fig.S1). Thesedoped ionsslightly inhibited particlecoalescenceandincreasedthespecificsurfaceareaofthe TiO2powdersfrom48to52–60m2/g.ThedopedTiO2powdershad

Fig.2. (a)TEMimage,(b)EDSline-scanningand(c)mappingresultsforthe Fe-dopedTiO2particles.GreenandreddotsinthemappingresultstandforFeandTi

elements,respectively.(Forinterpretationofthereferencestocolorinthisfigure legend,thereaderisreferredtothewebversionofthisarticle.)

Table1

TheM/Tiratios,chemicalstates,andthepeakfeaturesofthephotoelectronlinesofdopedtransition-metalions.

Dopant XPS-surface XPS-subsurface EPR

BEa_{(eV) DS}b_{(eV) FWHM}c

(eV)

Chemical states

M/Tiratios BEa_{(eV) DS}b_{(eV) FWHM}c

(eV) Chemical states Chemical states V 516.7 7.6 1.5 V4+ _{0.05 516.6 7.6 1.5 V}4+ _V4+_(V3+₎d Mn 641.6 11.5 4.0 Mn3+ _{0.11 641.5 11.5 4.5 Mn}3+ _Mn4+ (Mn3+₎ Fe 710.8 13.1 5.0 Fe3+ _{0.48 709.6 13.6 4.0 Fe}2+ _Fe3+_(Fe2+₎ Cu 933.0 19.9 2.0 Cu+ _{0.11 932.7 19.9 2.5 Cu}+ _Cu2+_(Cu+₎ Ce 886.1 882.0 18.5 18.1 4.5 4.0 Ce0₍₅₅₎e Ce4+₍₄₅₎ 0.09 885.7 881.6 18.5 18.3 4.5 4.0 Ce0₍₆₆₎ Ce4+₍₃₄₎ Ce0_,Ce4+ W 36.3 2.2 1.8 W6+ _{0.12 35.9 2.2 1.7 W}6+ _W6+ a_{BE—bindingenergy.}

b_{DS:energydifferencebetweenthedoubletsplittingpeaks.} c _{FWHM—fullwidthathalfmaximum.}

d _{ThespeciesintheparenthesesaredeterminedbasedontheincreasedintensityofoxidizedspeciesintheEPRspectraafterexposuretoO}

2molecules. e_{Valuesintheparenthesesaretheatomicpercentageofthespecies.}

anintrinsicbandgapof3.21–3.31eV.Inaddition,thesurface-doped V,Mn,andFeionsintroducedextrinsicbandgapsof2.73,2.54,and 2.15eV,respectively,inthedopeddomain.

3.2. Surfacecompositionsandchemicalstatesofdopants

Thesurfacecompositionsandthechemicalstatesofthedopants werecharacterizedusingbothXPSandEPRmethods.TheXPS spec-traofthedopants inthetop-mostandsub-surfacelayerswere acquiredbothbeforeandaftersoftArsputtering.Toensurethat thesoft sputtering didnot induce reduction ofthe dopedions in thesamples,Fe2O3 powderswereused asthe standard and thephotoelectronlinesoftheFeionsweremeasuredbothbefore and after thesputtering. A minute chemical shift betweenthe photoelectronpeaksconfirmsthatthisactionhaslittleinfluence onchangingthechemicalstates.Table1summarizesM/Tiratios (Mdenotesthetransitionmetalions),thechemicalstatesofthe dopants,anddetailedfeaturesoftheircorresponding photoelec-tronlinesintheXPSspectra.ThemajorityofthemeasuredM/Ti ratiosofthedopedpowders(0.05–0.12)werelessthanthe nom-inalvalue(0.20),exceptfortheFe-dopedTiO2sample,whichhad anFe/Tiratioof0.48.ThelowM/Tiratiosareastheresultofthe reducedthicknessesofthedopedshells(<10nm)comparedtothe samplingdepthofXPS.Infact,thesurfaceFe/Tiratiowas0.22prior tocalcination.Theincreaseintheratioaftercalcinationindicates thatcalcinationforcestheFeionstomovetowardthesurface.The segregationoftheFeionsisduetotheirlowsolubilityintheTiO2 matrixandtheHüttigtemperaturesofironoxides(defined empir-icallyas0.3Tm,whereTmisthemeltingpoint.Fe2O3:470◦C,FeO: 413◦C)thatarelowerthanthecalcinationtemperature(500◦C) [21,22].AlthoughMn-,V-,andCu-basedoxidesalsohavelow Hüt-tigtemperatures(Mn2O3:266◦C,V2O5:207◦C,CuO:316◦C,Cu2O: 289◦C),increasedM/TiratiosarenotmeasurableusingXPSbecause oftheultra-thincoatingofthedopedshells[23].

Thestates ofthetransition metalions,V4+_,Mn3+_,Fe3+_,Cu+_, Ce0_/Ce4+_{, and W}6+_{, were characterized in the top-most lattice.} ExceptfortheW6+_andFe3+_{ions,whichmaintainthesamevalence} numbersasthoseoftheirprecursors,theV4+_,Cu+_,andCe0_species arethereducedforms,and theMn3+_andCe4+_ionsarethe oxi-dizedformsoftheirprecursors.Afterremovalofseveralatomic layersusingsoft sputtering,wedeterminedthechemical states ofthedopantslocatedwithinthesub-surfacelattice.Whilethe otherdopedionsexhibitedthesamechemicalstatesasthosein theouterlattice,reducedFe2+_{speciesandanincreasedfraction} ofCe0 _{atomsweremeasured.TheFe}3+_andV5+_{ionscouldhave} beenreducedbytheisopropanolintheprecursorsolutionsbecause thereductionpotentialoftheFe3+_/Fe2+_{(0.711V)andtheV}5+_/V4+ (E0_{=1.00V)couples is higher than that of isopropanol/acetone}

5000 4500 4000 3500 3000 2500 2000

Intensity

(A.

U.

)

Magnetic field

(G

)

Fig.3. EPRspectraforboththepureandtheM-dopedTiO2powders(M=V,Mn,Fe,

Cu,Ce,andWionswithinthesurfacelattice).Allspectraarerecordedindarkness inthevacuum.

(E0_{=0.28V) [14,24]. Moreover, thermally-induced} dehydroxyla-tionanddeoxygenationalsoforcethereductionofthedopedions [24,25].TheexposureofsurfaceFeionstoO2 moleculesathigh temperaturesdominates theFe3+_{stateinthetop-mostlayer.In} contrast to reduction, Mn2+_-to-Mn3+ _{conversion occurs via the} thermal decompositionof a MnO2 moiety atabout 500◦C [26]. TheincompatiblestructureoftheMnOmoietyintheTiO2matrix couldbethereasonfortheMnO-to-MnO2transformationatlow temperatures.ThedisproportionationoftheCe3+_{ionsleadstothe} coexistenceoftheoxidizedCe4+_andreducedCe0_species.

Fig.3showstheEPRspectraforthepureandthesurface-doped TiO2 powders recorded in a vacuum.EPR is sensitive to para-magneticspeciesandprovidesadditionalinformationaboutthe chemicalstatesofthedopants.ThepureTiO2powdersexhibited aweaksignalatg=2.003indarkness,whichdenotesthetrapped holesatthesurfaceO−species[15].AllthedopedTiO2 samples showedintensivepeaks,otherthantheO−signal,intheirEPR spec-tra,indicatingtheparamagneticfeaturesofthedopedions.Vions (S=1/2,I=7/2)inthedopedTiO2powderscontributedtoanoctet in theERPspectrum astheresultofnucleus hyperfine interac-tions.ThesimulatedspinHamiltonianparametersfortheVions areg_⊥=1.962,A_⊥=52G,g//=1.938,andA//=165G,indicatingthe V4+_{ions[14].Thehyperfineinteractionsof}55_{Mnelements(I=5/2)}

4000 2000 -13000 0 13000 4000 2000 -40000 0 40000 4000 2000 -30000 0 30000

W-Ti

O

Ce-TiO

Cu-TiO

Fe-TiO

Mn-TiO

V-TiO

Δ

I

Magnetic field

(G

)

Fig.4. EPRdifferencespectrafortheM-dopedTiO2powders.I=IO2−Ivac,whereIO2andIvacrepresenttheintensityofthesignalsrecordedintheO2atmosphereandin

thevacuum,respectively.

showedtwosetsofsextetintherangeof2835–3976G[27].The simulatedHamiltonianparametersfortheMnionsaregx=2.008, gy=1.994,gz=1.992,andAiso=86G,indicatingtheMn4+ionswith ad3 _{electronicconfigurationandS=3/2.Mn}3+_{ions,whichwere} examinedasthemajordopedspeciesinXPS,weresilentinEPR. TheFe-dopedTiO2sampleshowedtworesonancepeaksatg=4.3 and2.0,indicatingtheFe3+_{ionsatthesubstitutionalsiteinthe} anataseTiO2lattice[28].Cuions(I=3/2)showedthreesetsof4-fold hyperfine-structurelines,indicatingCu2+_ionswithad9_electronic configuration.Thesepeakscanberesolvedusingtwogroups of Hamiltonianparameters.Thefirstisg_⊥=2.07,A_⊥=30G,g//=2.33, A//=70G,andtheotherisg⊥=2.07,A⊥=30G,g//=2.24,A//=70G, whichindicatetheCu2+_{ionsatthesubstitutionalsitesandtheCu}2+

ionsintheCuOcluster,respectively[29].CeandWionsweresilent intheEPRspectra,revealingtheirdiamagneticproperties.These resultsareinagreementwiththechemicalspecies(W6+_,Ce4+_,and Ce0_{)determinedfromXPS.WefurtheracquiredtheEPRspectrafor} thedopedsamplesintheO2atmospheretoexaminetheinteraction ofthesampleswithO2molecules.Fig.4showstheEPRdifference spectraobtainedbysubtractingthesignalsrecordedinthe vac-uumfromthoserecordedintheO2atmosphere.Inthepresenceof O2molecules,theintensityoftheV4+,Mn4+,Fe3+,andCu2+species increasedinthedopedTiO2powders.Theseresultsrevealthe exist-enceofreducedspeciesinthelattice,includingV3+_,Mn3+_,Fe2+_,and Cu+_{ions.Inaddition,thedopedionsareabletotransferelectrons} totheadsorbedO2moleculesacrosstheinterface.

4000 2000 -20000 -10000 0 10000 20000

Δ

I

TiO

V-TiO

Fe-

TiO

Cu-TiO

Mn-TiO

Ce-TiO

(a)

Magnetic field

(G)

W-TiO

Δ

I

Magnetic fie

ld (G)

W-TiO

Ce-TiO

Cu-TiO

Fe-TiO

Mn-TiO

V-TiO

TiO

(b)

Fig.6. EPRdifferencespectraforthepureandthesurface-dopedTiO2powders(a)inthevacuumand(b)intheO2atmosphere.I=Iirradiation−Idarkness,whereIdarknessand

Iirradiationrepresenttheintensitiesacquiredbeforeandafterirradiation,respectively.

BasedonthespeciesdeterminedfromXPSandEPR character-izations,weillustratetheelectronicstructuresofthedopedTiO2 domainsinFig.5.Theconductionbandandthevalencebandof TiO2matrixareprimarilycomprisedofTi(3d)andO(2p)orbitals, respectively[30].Whenadditionalionsareintroducedintothe lat-tice,interactionbetweentheouter-shellorbitalsoftheseionsand theenergystatesinthebandscreatesimpuritylevelsandchanges thebandstructures.Theinfluenceofthedopantsontheelectronic structuresisassociatedwiththeiratomicnumbers,ionicradiiand oxidationstates[30].Ingeneral,theconductionbandminimum

(CBM)moves downward toa lowerenergy whena severe dis-tortionoftheoctahedralgeometryiscausedbysmalldopedions. Theions,whichhavehighoxidationstates,decreasetheenergyof boththevalencebandmaximum(VBM)andtheCBMbecausemore oxygenionsareneededforchargecompensation.Thehigh occu-pancyofd-orbitalslowerstheenergyofthedstates,thusmoving theenergylevelsawayfromtheconductionbandandcontributing totheVBM.IntheV-dopedTiO2domain,thede-localizeddxz/dyz statesoftheV3+_andV4+_{ionsextendtheCBMtoalowerenergy} [30].TheoccupiedV3+_andV4+_(3d

0.43and1.00eV,respectively,belowtheCBM[31].DopedMn3+ ionshaveoccupied3dxyand3dxz/3dyzstatespositionedat1.0eV abovetheVBMandanunoccupied3d_x2−y2statesat0.9eVbelow theCBM[32,33].RelativetotheMn3+_{ions,theenergystatesof} Mn4+_{ionsshifttolowerenergyvaluesinthebandgap[33].The} highsplittingenergyofFe3+_{ionsintroduceslocalizedFe(3d}

x2−y2) statestotheVBMtoincreasetheband-edgeenergy.De-localized Ti4+_/Fe3+_(3d

z2)states,whichareoccupiedwithelectrons,are pos-itionedat0.5–0.8eVabovetheVBM,whilethelocalizedFe3+_(3d

xy) andthedelocalizedTi4+_/Fe3+_(3d

xz/3dyz)states,whichare unoc-cupied,arelocatedat0.7eVbelowtheCBM[30].Fe2+_ionshave alower oxidationnumber anda largerionicradii (0.78 ˚A)than Fe3+ _{ions(0.65 ˚A).Accordingly,theVBMof theFe}2+_-dopedTiO

2 domainsand theenergy statesof theFe2+ _{ionsmovetohigher} energyvalues.Cu+ _{ionshavelow potentiald statesbecausethe} d-orbitalsarecompletelyfilled.ThelocalizedCu+_(3d

x2−y2)andthe de-localizedTi4+_/Cu+ _(3d

z2)statesdominatetheVBMofthe Cu-dopedTiO2clustersandincreasetheband-edgetoahigherenergy [30,34].DopedCu2+_{ionshaveenergystatesclosetothoseoftheCu}+ ionsinthebandstructure[34,35].Ce4+_{ionsintroduceunoccupied} 4fstateslocatedat1.5eVbelowtheCBM[36,37].Twounoccupied statesfromtheW6+_(5d

xz)and(5dx2−y2)orbitalslieat0.8and2.1eV belowtheCBM,respectively[37,38].Thesemodelsimulationsare inaccordancewiththeelectronicstructuresmeasuredinthisstudy. TheextrinsicbandgapsofthepowdersdopedwithV (2.73eV), Mn(2.54eV),andFe(2.15eV)ionscorrespondtotheenergyfor theelectrontransitionfromtheVBMtothede-localizedVdxz/dyz states,theoccupiedMn3dxyand3dxz/3dyzstatestotheCBM,and theoccupiedFedstatestotheunoccupieddstates,respectively [12].

3.3. Chargetrapping,recombinationandinterfacialtransfer Thedopedionswithinthesurfacelatticeareinvolvedincharge trapping,recombinationandinterfacialtransfer.Toascertainthese behaviors,werecordedtheEPRspectrafortheTiO2powders irra-diated with UV light and semi-quantified the photo-generated speciesbothintheabsenceandinthepresenceofO2molecules. Fig. 6 shows the EPR difference spectra for the pure and the dopedTiO2 powdersboth in the vacuumand in theO2 atmo-sphere.Thespectral differencesareobtainedbysubtractingthe spectrarecordedinthedarknessfromthoseafterirradiation,and thesignalsin thespectraindicatethephoto-generated species. Thequantitiesofthesamespeciescanbedirectlyrelatedtothe correspondingintensitiesbecauseallthesampleshaveanequal massandasimilarsurfacearea.ThepureTiO2 powdersshowed anintensivepeakatg=2.003inthevacuum,indicatinga substan-tialnumberofphoto-generatedholestrappedatthesurfaceO− sites.However,therewerefewerthephoto-generatedholesinthe O2atmosphere.Coronadoetal.[39]reportedthatO2 molecules areabletoparticipateinthesurfacechargetrappingbygenerating eitherozonideions(Ti4+ _O

3−)orsuperoxidespecies(Ti4+ O2−) whentheyreactwiththesurfacetrappedholesorelectrons, respec-tively.Inthisstudy,thedisappearanceofthephoto-generatedholes andtheabsenceofTi4+ _O

3−andTi4+ O2−speciesindicatethatthe adsorbedO2moleculesmediatechargerecombination.Infact, sur-faceO−speciesdisappearedwhenrecordedindarknessafterthe pureTiO2powderswereexposedtoO2gas.Thisfindingsupports theunderstandingthattheadsorbedO2 moleculesinteractwith thetrappedholes.

Eachofthedopantsalsoeliminatedthephoto-generatedholes todifferentextentswhenirradiatedinthevacuum,andtheholes wereevenunabletobedetectedintheMn-dopedsamples.The dopedions,which introduceenergylevelsbetweenthe conduc-tionand valencebands,areabletomediatechargeannihilation

Fig.7.ThephotocatalyticactivityofthepureandthedopedTiO2powders.

through trapping. Because there are noobvious signals of the dopantspresentinthedifferencespectra,mediatedannihilationis rapidinallthedopedsamples.Basedonthepeak-to-peakheights, wesemi-quantifiedtheremainingtrappedholesinthedopedTiO2 powders.ItwasfoundthatFeionsmaintainedthehighest num-bersoftrappedholes,followedinorderbyW-,V-,Ce-,Cu-,and Mn-ions.Thed–dtransitionoftheFeionsinhibitschargetrapping andsoreducesthemediatedrecombination.ComparedtotheV ionswhichintroduceasingleoccupiedstatebelowtheCBM,the twounoccupiedstatesoftheWionsmediatecharge recombina-tionmoreslowlybecauseelectronsrequiremultipleprocessesto recombinewithholes.TheCe-dopedpowderscontaineda substan-tialnumberofmetallicCeatoms.Thesemetallicatomsmediate chargerecombinationbyreceivingelectronsfromtheTiO2 con-ductionband,andthenreturningtheelectronstorefilltheholes inthevalenceband.Similarmediatedrecombinationisalsofound attheinterfacebetweenPtnanoparticlesandTiO2 crystals[40]. Mn3+_andMn4+_{ions,whichintroducebothoccupiedand} unoccu-piedlevelsinthemid-band-gapregion,trapelectronsandholesat almostthesametime.Intra-atomicrelaxationsoonannihilatesthe trappedchargecarriers.TheoccupiedstatesoftheCu+_andCu2+ ionsarepositionedinthevalenceband.Theremarkablereduction inthenumberoftrappedholesandtheslightlyincreasedintensity oftheCu2+_{peakindicatethattheCuionsareabletoefficientlytrap} holes.

Throughlatticetrappingandinterfacialtransfer,bothdopants andoxygenmoleculesareabletomediaterecombination.To deter-minewhichprocessismoreefficient,wecomparethequantities ofthetrappedholesinthedopedTiO2 powdersinthevacuum (Fig.6a,dopedpowders)withthoseinthepureTiO2powdersin theO2 atmosphere (Fig.6b,TiO2).The higherintensitiesofthe trappedholesintheFe-,W-,andV-dopedsamplesrevealthat inter-facialchargetransferfromtheconductionbandtotheadsorbedO2 moleculesismoreefficientthanthelatticetrappingbythedopants, whereas,incontrast,theCe,Cu,andMnionsinducedanopposite effect.

IntheO2atmosphere,adsorbedO2moleculesmediated recom-bination,andfurthereliminatedthephoto-generatedholesinthe dopedTiO2powders.However,areductioninthequantityofthe holes of only 6%was foundin theW-doped powders.Because charge transfer betweentheTiO2 surfaceand the adsorbedO2 moleculeshasbeendemonstratedtobefasterthanlatticetrapping bytheWions,thissmallreductionsuggeststhattheWionsinhibit theinteractionoftheO2−anionswiththeholes.Similarbehavior wasfoundontheCe-dopedpowders,forwhichO2moleculeseven increasedthenumberofphoto-generatedholes.

Fig.8.Therolesofthesurface-dopedionsinchargetrapping,recombination,interfacialtransfer,andinteractionwithsurfacespecies.

3.4. Photocatalyticactivity

The photocatalytic activities of the pure and surface-doped TiO2 powders were examined based on the degradation of BPA (20mg/L) under UV irradiation. The degradation followed pseudo-first order kinetics, and the rate constants (k) were determined from the slope of the linear dependence between ln(C/C0) and the reaction time, where C0 and C are the BPA concentrations at the beginning and after a certain reaction time, respectively. Fig. 7 shows the photocatalytic activity of boththepureanddopedpowders.Thephotocatalystsexhibited activityintheorderofFe-dopedTiO2(k=3.59×10−2min−1)> Cu-doped TiO2 (k=2.35×10−2min−1)>TiO2 (k=1.94×10−2min−1) ∼ V-doped TiO2 (k=1.78×10−2min−1)>

W-doped TiO2 (k=1.14×10−2min−1)>Ce-doped TiO2 (k=9.33×10−3min−1)>Mn-doped TiO2 (k=6.40×10−3min−1). WefurtheroptimizedtheFeloadingandfoundthatthehighest activity (k=3.88_×10−2min−1)occurred ata surface Fe/Tiratio of0.67 (seeSupplementarymaterials,Fig.S2). Thisactivitywas two-timeshigherthanthatofthepureTiO2powders.

Photocatalysisisinvolvedwithchargegeneration,charge dif-fusionfromthebulktothesurface,andinterfacialchargetransfer totheadsorbates.Chargerecombinationreducesthenumberof availablechargecarriersforthesurfacereactions,andisthekey tolowphotocatalyticactivity.Chargerecombinationoccurseither throughband-to-bandrelaxationorthroughdefect-mediated anni-hilation.Incorporationofionsintothesurfacelatticecreatesan internalelectricfieldtotriggerthechargecarrierstodiffusefrom

thebulktothesurface.In addition,surface-dopedionstrapthe chargecarriersandinhibitband-to-bandrecombination.Unlikethe dopantspresentinthephotocatalysts,thedopantswithinthe sur-facelatticeareabletointeractwithadsorbates.Interfacialcharge transferviathesurface-dopedionstotheadsorbatescompeteswith mediatedannihilationtogovernthephotocatalyticactivity,and theprobabilityoftheseprocessesoccurringisdeterminedbythe energystates,thecoordinationnumbersandtheelectronegativity ofthesurface-dopedions.

Fig.8schematicallyillustratestherolesoftheV,Mn,Fe,Cu,Ce, andWionsintheinterfacialbehaviorsofthesurface-dopedTiO2 photocatalysts.Surface-dopedFeandCuionsimprovethe photo-catalyticactivitybecausetheytrapthechargecarriersandinteract withtheadsorbatestoprovidealternativepathwaysfor interfa-cialchargetransfer.Inaddition,thed–dtransitionsoftheFeions provideadditionalchargecarrierstocompensateforthecharge recombination.TheoptimalsurfaceFe/Tiratioforthehighest pho-tocatalyticactivityis0.67.Althoughthisvalueishigherthanthe solubilityofFeions(1–10at.%)intheTiO2matrix,Fe2O3clusters aretoominutetobedetectedinthedopeddomainexploredinthis study[22,41].Abovethisvalue,tunnelingofthetrappedcarriers promotesannihilationandconsequentlyreducethephotocatalytic activity[5].Cu+_{ions,whichintroduceoccupiedandunoccupied} statesinthevalenceandconductionbands,respectively,efficiently trapthechargecarriers.Moreover,thetrappedchargecarrierscan leavethetrappingsitesandescapefromrecombinationthrough thermallyinducedtransitioninthebands.Theappearanceofsmall Cu2+_{signalsin theEPRspectrumafterirradiationsupports this} deduction(seeFig.6).Theelectronicstructureenablesthe surface-dopedCuionstostabilizeahighnumberofchargecarriersatthe surface,thusfacilitatinginterfacialchargetransfer.Surface-doped Vionshavebeendemonstratedtoimprovethephotocatalytic activ-ityofthedopedTiO2powderswhenthesurfaceV/Tiratioishigher than0.14[14].However,theVionsexaminedinthisstudy con-tributelittletotheactivityasaconsequenceofthelowdopinglevel (i.e.theV/Tiratioisonly0.05).Fromtheelectronicstructurepoint ofview,V3+/4+_{ionshaveoccupiedstatesclosetheCBM.The} high-energystatesofferathermodynamicallyfavorable pathwayand allowtheelectronsthathaveahighpotentialtoundergointerfacial transferthroughtheVions,andinhibitsurfacerecombination.

IncontrasttoFe,Cu,andVions,surface-dopedW,Ce,andMn ionscausedetrimentaleffectsonthephotocatalyticactivity.The Mnand Ce elementsreduce theactivitybecause theyserve as recombinationcenters.Mn3+/4+ _{ions,which have bothoccupied} and unoccupiedstates withinthebandgap,trapelectrons and holesonthesamesitesandrapidlyannihilatethetrapped carri-ersthroughintra-atomicrelaxation.Althoughtheions,whichare abletotrapbothelectronsandholes,areconsideredtoeffectively inhibitrecombinationviaseparatingchargecarriersontheir dif-ferentsites[42,43],wedemonstrated,inthisstudy,thatsuchions aremorecapableofpromotingannihilation.UnlikeFeions,d–d transitionintheMnionsisinactiveintheirradiationspectrum. Chargerecombinationoccurstoofasttoallowinterfacialtransfer, thusconsumingasubstantialnumberofeffectivechargecarriers. MetallicCe elementsare responsiblefor thesevere recombina-tionandthelowactivityoftheCe-dopedTiO2powders.Schottky barrierat thehetero-junction hasbeenextensively reportedto effectivelyseparatethecharge carriersand enhancethe photo-catalyticactivity[44].However,theabsenceofelectronreceptors withintheTiO2matrixforcestheelectronsthataccumulateinthe metallicCeelementstorefilltheholesinthevalencebandofthe TiO2 matrix.Inaddition,theEPRresultshowsthattheadsorbed O2moleculesexperiencedifficultyinmediatingcharge recombi-nationontheCe-dopedTiO2powders.Thisphenomenonreveals thatthesurfaceCe4+_{ionsstronglybindthechemisorbedO}

2− radi-calstoinhibittheinteractionoftheradicalswiththetrappedholes.

SimilarEPRfeaturesandlowphotocatalyticactivityarealsofound intheW-doped-TiO2-basedsystem.Thesefindingsconnectthelow photocatalyticactivitytothestrongbindingofO2−radicalsonthe surface.WeattributethisadverseeffecttothehighLewisacidityof thesedopedions.TheW6+_andCe4+_{ionsare6-and8-coordinated,} respectively,intheWO3 andCeO2 moiety,whereastheyare 4-coordinatedintheTiO2matrix[45,46].Unsaturatedcoordination allowstheseionstobondwiththeO2−radicalswhentheytransfer thetrappedelectronstotheadsorbedO2molecules,orwhentheO2 moleculesreceiveelectronsfromtheconductionband.Lewis acid-ityhasbeenconsideredtoimprovephotocatalyticactivitybecause itpromoteschemisorption ofO2 moleculestofacilitate interfa-cialelectrontransfer[47–49].However,thehighelectronegativity oftheW6+_{ions(Pauling’sscale:2.36)andthehighoxygen} defi-ciencyoftheCe4+_ionsleadtheO

2−speciestobetightlybound, consequentlyimpedingsubsequentinterfacialchargetransferand reducingtheactivity.

4. Conclusions

Therolesof V,Mn, Fe,Cu,Ce,and Wions intheinterfacial behaviorofsurface-dopedTiO2particlesweresystematically stud-iedintermsofchargetrapping,recombination,interfacialcharge transferandphotocatalyticactivity.ThedopedFe,CuandVions improvethephotocatalyticactivity,whereastheMn,Ce,andW ionscauseadverseresults.Surface-dopedionscaneithermediate chargerecombinationortransfertrappedcarrierstoadsorbates. Theelectronicstructures,thecoordinationnumbersandthe elec-tronegativityofthedopedionsdeterminetheprobabilityofthese two processes occurring and the photocatalytic activity of the surface-dopedphotocatalysts.Suitabledopantshaved–d transi-tionsintheirradiationspectrum,energylevelsclosetoorwithin theconductionandvalencebands,ormultipleunoccupiedstatesin thebandgap.Theseelectronicstructuresretarddefect-mediated recombination and provide an alternative pathway that facili-tatesinterfacialcharge transfer viathedopedions. Incontrast, unsuitabledopantspromoterecombinationorconfinethecharge carriersatthesurface,therebyinhibitingchargeutilization.Such dopantsintroducebothoccupiedandunoccupiedstatesinthe mid-band-gapregion,whichtrapelectronsandholestorapidlyinduce annihilationviaintra-atomicrelaxation.Inaddition,theionswith highcoordinationnumbersorelectronegativitystronglybondwith O2− radicals,thuslimiting surfacereactivity. Theresultsin this studynotonlyextendtheknowledgerelatedtothebehaviorsof surface-dopedions atthe interface,butalso provideimportant information to improve materialsthat exhibit high photon-to-currentconversionefficiency.

Acknowledgment

WethanktheNationalScienceCouncil,Taiwan,R.O.C.(Grant no.NSC101-2628-E-009-022-MY3)forfinanciallysupportingthis study.

AppendixA. Supplementarydata

Supplementary data associated with this article can be found,intheonlineversion,athttp://dx.doi.org/10.1016/j.apcatb. 2014.03.044.

References

[1]X.L.Hu,G.S.Li,J.C.Yu,Langmuir26(2010)3031–3039.

[2]M.Anpo,PureAppl.Chem.72(2000)1265–1270.

[3]M.Quintana,T.Edvinsson,A.Hagfeldt,G.Boschloo,J.Phys.Chem.C111(2007) 1035–1041.

[4]A.W.Xu,Y.Gao,H.Q.Liu,J.Catal.207(2002)151–157.

[5]W.Y. Choi, A. Termin, M.R. Hoffmann, J. Phys. Chem.—US 98 (1994) 13669–13679.

[6]K.Nagaveni,M.S.Hegde,G.Madras,J.Phys.Chem.B108(2004)20204–20212.

[7]J.X.Li,J.H.Xu,W.L.Dai,H.X.Li,K.N.Fan,Appl.Catal.,B:Environ.85(2009) 162–170.

[8]M.Schiavello,V.Augugliaro,L.Palmisano,K.Sclafani,A.M.Venezia,Langmuir 11(1995)3278.

[9]D.Dvoranova,V.Brezova,M.Mazur,M.A.Malati,Appl.Catal.,B:Environ.37 (2002)91–105.

[10]J.Z.Bloh,R.Dillert,D.W.Bahnemann,J.Phys.Chem.C116(2012)25558–25562.

[11]Z.B. Zhang,C.C.Wang,R. Zakaria, J.Y.Ying,J. Phys.Chem.B102 (1998) 10871–10878.

[12]J.F.Zhu,F.Chen,J.L.Zhang,H.J.Chen,M.Anpo,J.Photochem.Photobiol.,A: Chem.180(2006)196–204.

[13]J.F.Zhu,Z.G.Deng,F.Chen,J.L.Zhang,H.J.Chen,M.Anpo,J.Z.Huang,L.Z.Zhang, Appl.Catal.,B:Environ.62(2006)329–335.

[14]S.M.Chang,W.S.Liu,Appl.Catal.,B:Environ.101(2011)333–342.

[15]S.M.Chang,P.H.Lo,C.T.Chang,Appl.Catal.,B:Environ.91(2009)619–627.

[16]S.M.Chang,C.Y.Hou,P.H.Lo,C.T.Chang,Appl.Catal.,B:Environ.90(2009) 233–241.

[17]L.P.Li,J.J.Liu,Y.G.Su,G.S.Li,X.B.Chen,X.Q.Qiu,T.J.Yan,Nanotechnology20 (2009).

[18]K.Zhang,Z.H.Zhou,L.J.Guo,Int.J.HydrogenEnergy36(2011)9469–9478.

[19]S.Lacombe,H.Cardy,N.Soggiu,S.Blanc,J.L.Habib-Jiwan,J.P.Soumillion, Micro-porousMesoporousMater.46(2001)311–325.

[20]H.J.Zhai,L.S.Wang,J.Am.Chem.Soc.129(2007)3022–3026.

[21]J.A.Moulijn,A.E.vanDiepen,F.Kapteijn,Appl.Catal.,A:Gen.212(2001) 3–16.

[22]F.Gracia,J.P.Holgado,A.Caballero,A.R.Gonzalez-Elipe,J.Phys.Chem.B108 (2004)17466–17476.

[23]Q.L.Song,W.Liu,C.D.Bohn,R.N.Harper,E.Sivaniah,S.A.Scott,J.S.Dennis, EnergyEnviron.Sci.6(2013)288–298.

[24]S.M.Chang,R.A.Doong,J.Phys.Chem.B108(2004)18098–18103.

[25]S.M.Chang,R.A.Doong,Chem.Mater.17(2005)4837–4844.

[26]J.H.Li,J.J.Chen,R.Ke,C.K.Luo,J.M.Hao,Catal.Commun.8(2007)1896–1900.

[27]V.Singh,V.Natarajan,J.J.Zhu,Opt.Mater.30(2007)468–472.

[28]C.Fabrega,T.Andreu,A.Cabot,J.R.Morante,J.Photochem.Photobiol.,A:Chem. 211(2010)170–175.

[29]G.H.Li,N.M.Dimitrijevic,L.Chen,T.Rajh,K.A.Gray,J.Phys.Chem.C112(2008) 19040–19044.

[30]A.M.Shough,D.J.Doren,B.Ogunnaike,Chem.Mater.21(2009)1232–1241.

[31]D.Y.Lee,W.J.Lee,J.S.Song,J.H.Koh,Y.S.Kim,Comput.Mater.Sci.30(2004) 383–388.

[32]A.L.J.Pereira,L.Gracia,A.Beltran,P.N.Lisboa,J.H.D.daSilva,J.Andres,J.Phys. Chem.C116(2012)8753–8762.

[33]H.W.Peng,J.B.Li,S.S.Li,J.B.Xia,J.Phys.Condens.Matter20(2008).

[34]M.L.Guo,J.L.Du,PhysicaB:Condens.Matter407(2012)1003–1007.

[35]N.Nakajima,H.Kato,T.Okazaki,Y.Sakisaka,Surf.Sci.561(2004)93–100.

[36]W.G.Chen,P.F.Yuan,S.Zhang,Q.Sun,E.J.Liang,Y.Jia,PhysicaB:Condens. Matter407(2012)1038–1043.

[37]H.Kamisaka,T.Suenaga,H.Nakamura,K.Yamashita,J.Phys.Chem.C114(2010) 12777–12783.

[38]R.Long,N.J.English,Appl.Phys.Lett.98(2011).

[39]J.M.Coronado,A.J.Maira,J.C.Conesa,K.L.Yeung,V.Augugliaro,J.Soria, Lang-muir17(2001)5368–5374.

[40]W.N.Wang,W.J.An,B.Ramalingam,S.Mukherjee,D.M.Niedzwiedzki,S. Gan-gopadhyay,P.Biswas,J.Am.Chem.Soc.134(2012)11276–11281.

[41]Q.P.Wu,Q.Zheng,R.vandeKrol,J.Phys.Chem.C116(2012)7219–7226.

[42]J.F.Zhu,W.Zheng,H.E.Bin,J.L.Zhang,M.Anpo,J.Mol.Catal.A:Chem.216 (2004)35–43.

[43]T.Z.Tong,J.L.Zhang,B.Z.Tian,F.Chen,D.N.He,J.Hazard.Mater.155(2008) 572–579.

[44]H.Chen,S.Chen,X.Quan,H.T.Yu,H.M.Zhao,Y.B.Zhang,J.Phys.Chem.C112 (2008)9285–9290.

[45]S.Yamazoe,Y.Masutani,T.Shishido,T.Tanaka,AIP.Conf.Proc.882(2007) 696–698.

[46]G.Dutta,U.V.Waghmare,T.Baidya,M.S.Hegde,K.R.Priolkar,P.R.Sarode,Chem. Mater.18(2006)3249–3256.

[47]T.Z.Tong,J.L.Zhang,B.Z.Tian,F.Chen,D.N.He,M.Anpo,J.ColloidInterfaceSci. 315(2007)382–388.

[48]Y.Cong,B.Z.Tian,J.L.Zhang,Appl.Catal.,B:Environ.101(2011)376–381.

[49]Y.F.Ma,J.L.Zhang,B.Z.Tian,F.Chen,L.Z.Wang,J.Hazard.Mater.182(2010) 386–393.