An experimental and first-principles study of the effect of B / N doping in TiO2 thin films for visible light photo-catalysis

(1)

ContentslistsavailableatSciVerseScienceDirect

Journal

of

Photochemistry

and

Photobiology

A:

Chemistry

j o u r n al hom ep age : w w w. e l s e v i e r . c o m / l o c a t e / j p h o t o c h e m

An

experimental

and

ﬁrst-principles

study

of

the

effect

of

B/N

doping

in

TiO

2

thin

ﬁlms

for

visible

light

photo-catalysis

Md.

Nizam

Uddin

a,b,∗

_,

_Sayed

_Ul

_Alam

_Shibly

a

_,

_Rasim

_Ovali

c

_,

_Saiful

_Islam

a

_,

Md.

Motiur

Rahaman

Mazumder

a

_,

_Md.

_Saidul

_Islam

a

_,

_M.

_Jasim

_Uddin

d,e

_,

Oguz

Gulseren

c

_,

_Erman

_Bengu

b

a_Department_of_Chemistry,_Shahjalal_University_of_Science_and_Technology,_Sylhet-3114,_Bangladesh b_Department_of_Chemistry,_Bilkent_University,₀₆₈₀₀_Ankara,_Turkey

c_Department_of_Physics,_Bilkent_University,₀₆₈₀₀_Ankara,_Turkey

d_{High-Performance}_Materials_Institute,_Florida_State_University,_Tallahassee,_FL_32310,_USA e_Department_of_Chemical_Engineering_and_Polymer_Science,_Sylhet-3114,_Bangladesh

a

r

t

i

c

l

e

i

n

f

o

Articlehistory: Received30July2012 Receivedinrevisedform 15December2012 Accepted30December2012 Available online 19 January 2013 Keywords: TiO2 Photo-catalyst Thinﬁlm Doping Dipcoating

a

b

s

t

r

a

c

t

ThinﬁlmsofTiO2andboron–nitrogen(B/N)co-dopedTiO2onglasssubstrateshavebeenpreparedby

asimplesol–geldipcoatingroute.Titanium(IV)isopropoxide,boricacidandureahavebeenusedas titanium,boronandnitrogensources,respectively.TheﬁlmswerecharacterizedbyX-raydiffraction, X-rayphoto-electronspectroscopy,scanningelectronmicroscopy,RamanspectroscopyandUV–vis spec-troscopy.TheTiO2thinﬁlmswithco-dopingofdifferentB/Natomicratios(0.27–20.89)showedbetter

photo-catalyticdegradationabilityofmethylenebluecomparedtothatofbare-TiO2undervisiblelight.

TheTiO2ﬁlmdopedwiththehighestatomicconcentrationofNshowedrepeatedlythebest

photo-catalyticperformance.Thehighactivityofco-dopedTiO2thinﬁlmstowardorganicdegradationcanbe

relatedtothestrongerabsorptionobservedintheUV–visregion,redshiftinadsorptionedgesand sur-faceacidityinducedbyB/Ndoping.Furthermore,severalatomicmodelsforB/Ndopinghavebeenused toinvestigatetheeffectofdopingonelectronicstructureanddensityofstatesofTiO2throughab-initio

densityfunctionaltheorycalculations.Thecomputationalstudysuggestedasigniﬁcantnarrowingofthe bandgapduetotheformationofmidgapstatesandtheshiftofFermi-levelfortheinterstitialNmodel supportingtheexperimentalresults.

1. Introduction

Theutilizationofsolarirradiationtosupplyenergyortoinitiate chemicalreactionsisalreadyawellestablishedidea[1].Anatase phaseoftitaniumdioxide(TiO2),anon-toxicandbiocompatible wide-bandgap semiconductor, when irradiatedwitha suitable wavelengthlightisknowntofacilitatechemicalprocessesonits surfaceincludingdegradationreactions.Inaddition,TiO2isoneof themostimportantandwidelyinvestigatedphoto-catalyst materi-als[2].Itcanbeusedindecompositionofvariousenvironmentally hazardouscompounds(organic,inorganicandbiologicalmaterials) inbothgaseousandliquidphases[3].

ItiswellknownthatTiO2hasthreepolymorphs:brookite,rutile andanatase.Mixed-phasephoto-catalystswithrutileandanatase,

∗ Correspondingauthorat:DepartmentofChemistry,ShahjalalUniversityof Sci-enceandTechnology,Sylhet3114,Bangladesh.Tel.:+8801926372680;

fax:+88082171525.

E-mailaddresses:nizam3472@yahoo.com,uddinnizam@hotmail.com (Md.N.Uddin).

e.g.P25Degussa[4],phaseshavebeenreportedtoexhibitenhanced photo-activityrelativetosingle-phaseanatase.However, synthe-sisofrutile–anatasephasemixturerequiresahightemperature treatment; heatinganatase upto 600–700◦C is needed for the transformationofanatasetorutile[5].Furthermore,therearestill problemsintheuseofTiO2forpracticalandwide-spread photo-catalyticapplications:

(1)Recycling of nano-particulate TiO2 requires costly separa-tion/ﬁlteringprocesses.

(2)TiO2photo-catalysthasawidebandgap(3.2eV,foranatase), and can only be activated by UV radiation (<387nm) which constitutes only a small fraction (3–5%) of the solar spectrum. Thus, the use of visible light (400–750nm, ∼45% of thesolar spectrum) byanatase shouldbe enabled [6].

(3)TiO2 has a relatively low rate of electron transfer to oxy-genandahighrateofrecombinationwhichresultsinalow quantum yield rateand also a limited photo-oxidation rate [7].

(2)

remediationoftheissueslistedabove.Someoftheeffortsarelisted asaccordingly:

(1)Nano-crystallineTiO2thinﬁlmshaveattractedagreatdealof attention[8,9]owingtotheirﬂexibilityintreatingwastes,e.g. noseparationrequiredandreuseofphoto-catalyticproducts. (2)AnyredshiftintheopticalresponseofTiO2fromtheUVband

towardthevisiblespectrumwillhaveaprofoundeffectonthe photo-catalytic efficiency[10,11].Withthis view,doping of pureTiO2hasbeenundertakenbyanumberofresearchgroups [12–30]in thelast decade.Modificationby noble-transition metalshasbeenhistoricallyregardedasthefirststrategyandby nonmetalsasthesecondstrategy[12].However,these metal-doped TiO2 materials suffered from thermal instability and low-quantumefficiencybecauseofincreasedcarriertrapping afterdoping[13,14].

Incontrastaniondopinghasshowngreatpotentialin introduc-ingbathochromismandintensiveeffortshavebeenundertakento synthesizeanion-dopedtitaniatowardvisible-light-active photo-catalysts[15–23],usually byintroducing localizedstates inthe bandgap.Fittipaldietal.hascriticallydiscussedthelimitations andfuturechallenges intheuseofelectronparamagnetic reso-nancetechniquein theinvestigationofaniondopedTiO2 based photocatalyst[31].Recentemphasishasbeenplacedonco-doped systems,thatis,thoseinvolvingcombinationsofcationsandanions [24]ortwoanionstogetherwithintheoxidelattice,inwhicha dra-maticenhancementofphoto-catalyticbehaviorhasbeenreported [25–28].

Thus,manyresearchersstartedtoinvestigateanionicnonmetal dopantssuchasC[32–35],N[34,36–43],S[44,45],andB[46,47]for extendingthephoto-catalyticactivityintothevisible-lightregion because,therelatedimpuritystatesforsuchdopantsappearnear thevalencebandedgebutdonotactaschargecarriers[34]. Gom-bacetal.havestudiedBandNco-dopedpowdered-TiO2,where asigniﬁcantimprovementinactivityarisesmainlyfromthered shift in the absorption edge, and also B is reported to inhibit growththerebyresultinginhighsurfaceareapowders[48]. Espe-cially,surface constructionof titaniaduring N-dopinghasbeen demonstratedbothexperimentallyandtheoretically[37,49]. How-everN-dopingoftitaniabythermaltreatmentunderanammonia atmoshphereusuallyleadstoverylimitedvisiblelightactivitybut greatlyimpairstheUVactivity[11].Divergentresultshavebeen reportedforless-studiedsysteminvolvingB-dopedTiO2[17,46,47]. Itisthereforehighlyimportanttodesignandconstructeffective photocatalystsurfacestructureswithsomesortsofco-dopingwith properratioofdifferentanionsaimingsynergyeffectsthatenhance theseparationandtransferofthecarrierstodevelopefﬁcient vis-iblelightphotocatalysts.However,there isstillgreat debateon thelocationof thedopants,thesynergyeffects on photoactivi-tiesandtheimportanceoftherelativeratiosonphotoactivities [17,50].

In this work, we focused on thesynthesis of B/N co-doped anatasethinfilmsonglasssubstratesbyasimplesol–geldip coat-ingroute.Itwasouraimtoexplorepossiblesynergisticadvantages arisingfromthesimultaneouspresenceofBandNasdopantsin theTiO2structure.Asmentionedearliertherearealreadyreports onthephoto-catalyticbehaviorofboronandnitrogendopedTiO2 nano-powders[51–54]andmesoporousforms[50].Tothebestof ourknowledge,thereisnoreportonthephoto-catalyticactivity ofboronandnitrogendopedanatasethin films.Detailed struc-tural,compositionalandopticalcharacterizationoftheco-doped titaniathinfilmshasbeendone.Then,thesefilmswereemployed asphoto-catalystsundervisiblelightirradiationinthe degrada-tionofaqueousmethyleneblue(MB),adyeofteninvestigatedas

tivestudyofthecatalyticefﬁciencywasevaluatedasafunctionof thedopantnatureandatomicratioofthedopant.

A varietyof internal charge transfersmay takeplace inthe dopedsystems[57].Co-dopingmayalsobebeneficialtoreducethe numberofintrinsicdefectswhicharesupposedtobe detrimen-talinphotocatalyticprocessessincetheyareconsideredtofavor electron–holerecombination[57].Inordertoexplainthe improve-mentinthephoto-catalyticactivityobservedfortheco-dopedfilms weperformedfirst-principlesplane-wavecalculations[58]based onthedensityfunctionaltheory(DFT)[59,60]onvariousatomic modelsdepictingB/Ndopinginanatasestructure.Theresultsof thesetheoreticalworkswereusedtobetterexplainthechanges inducedinthebondingbehaviorandthebandgapofanataseby doping.

2. Experimental 2.1. Materials

Titanium(IV)isopropoxide(TIP)(≥97.0%,SigmaAldrich)and triethylamine(TA)(≥99%,SigmaAldrich)wereusedasTisource andstabilizer,respectively.Microscopicsodalimeslideglasswas usedassubstrate.MBwaspurchasedfromMerck.Solutionsof1M HCland1MNaOHwereusedtoadjustthepHofthesolution.All reagentswereanalyticalgradeandusedwithoutfurther puriﬁca-tion.

2.2. PreparationofTiO2ﬁlm

TIP (1.5ml) was added to anhydrous (Anh.) ethanol (10ml) under vigorous stirring conditions and then TA (0.35ml) was addedas a stabilizerof thesolutionand stirredat 200rpm for 2–3min under N2 environment (solution-A). A secondsolution wasprepared separatelyby mixing hydrochloricacid(0.92ml), water(0.15ml)andAnh.ethanol(10ml)usingamagneticstirrer at200rpm(solution-B).Thetwosolutionswerethenmixeddrop wiseandstirredvigorouslyfor60minunderN2environment.The formedTiO2solwastransparent,quitestableandhighlysensitive totheamountofTAandwater.Thenthesolwasagedfor24hand servedforﬁlmpreparation.Thetransparentsolwasstableforthree weeks.

TiO2 thinfilmswerepreparedbyadip-coatingmethod.Prior tothecoating process,soda-lime–silicaglasssubstrates (micro-scopeslides)withdimensionof 10mm×60mm×1.5mmwere grinded by a commercial bench grinder (model: ST-150), then cleanedinpotassiumdichromateanddichloromethanesolution. Finally those abrasive substrateswere rinsed withalcohol and deionizedwaterandthendriedat100◦Cinamicrooven.TheTiO2 gelfilmwasobtainedbydippingthesubstrateintheprecursor solutionbathandpulledupwardswithaspeedabout4cm/min. Thesubstratescoatedwithgelwerepretreatedinroom temper-atureandthenannealedfor20minusingamicroovenat200◦C and finally thefilm wasvapor treated in the vapor of boiling waterfor30storemovethelooselybondedparticles.The coat-ingprocesswasrepeatedfivetimesforthickfilmpreparationand finallyannealedat500◦Cfor2husingamufflefurnace(JSMF-30T, Korea).

B/NdopedTiO2 wassynthesizedbyasimilarmethodwhere appropriateamountsofboricacid(H3BO3)(solution-C)andurea (solution-D)weredissolvedseparatelyinAnh.ethanol(5ml)and rapidly added to the mixture of solution-A and solution-B. A schematicﬂow chartof thepreparationof dopedTiO2 ﬁlmsby sol–geldipcoatingisshowninFig.1.

(3)

Fig.1.SchematicﬂowchartofthepreparationofdopedTiO2thinﬁlmonglasssubstratebysol–geldip-coatingprocess.

2.3. Filmcharacterization

Thesurfacemorphologyand thethicknessofthefilmswere investigated by scanning electron microscopy (Carl-Zeiss EVO 40 usingLaB6 filament).TheRamanspectrafromtheannealed sampleswererecordedwithaHoriva(Jobin–YvonMicroRaman) spectrometer.XRDpatternswererecordedbyRigaku(MiniFlesx) usingCuK␣radiation(=1.5406 ˚A).TheScherrer equationwas appliedtotheanatase(101)diffractionpeaktocalculatethe aver-agecrystallinesizes.XPSinaSPECSPHOIBOS100hemispherical electrostaticenergyanalyzerwasusedforthecheckingofatomic percentage of the dopant and others species of thefilms. The absorptionspectraofthebandwidthofthedopedandbare (un-doped)-TiO2filmsrangingfrom300to800nmwereinvestigated byUV–visspectroscopy(ThermoScientific:Evolution160). 2.4. Photo-catalyticstudies

Photo-chemicaldegradationwascarriedout inanopen visi-blelightchamberasshowninFig.2.Anouterwaterpumpwas usedtocirculateconstanttemperaturewaterthroughthesystem continuouslytokeepthetemperatureconstantduringthe degra-dationstudy.Thatchamberconsistedoftwomagneticstirrers,two coolingfansanda200Wtungstenlamp.Degradationwascarried outundervisiblelight.50ml1×10−5MMBsolutionwastaken andtwocatalystﬁlmswereusedfordegradationstudy.The sur-faceareaofeachﬁlmwas6cm2_._Change_in_the_{concentration}_of MBsolutionduringphoto-catalyticdegradationatdifferenttime intervalswasmonitoredbyUV–visspectroscopy(Simadju1800). Theabsorptionspectrawererecordedandrateofdecolorizationof

MBwasobservedintermsofchangeintheintensityatmaxofthe dye.Thedecolorizationefficiency(%)hasbeencalculatedas: effi-ciency(%)=((A0−At)/A0)×100,whereA0isthelightabsorbanceof MBbeforethetreatmentandAtisthatofaftertreatmentattimet. Beforetakingthesamplesundervisiblelightirradiation,the solu-tionofMBwastreatedwithcoatedfilmsfor30mininthedarkto getadsorptionanddesorptionequilibrium.

2.5. Computationaldetails

Wehaveperformedtheﬁrstprinciplesplane-wavecalculations [58] basedonDFT [59,60]using theprojector-augmented-wave (PAW)potentials[61,62]implementedinViennaab-initio simu-lationpackage(VASP)program[61–65].Theexchange-correlation potential was expressed in terms of the generalized gradient approximation(GGA)(Perdew–Wang91type)[66].Aplane-wave cut-offenergyof500eVwasusedinallcalculationstoachievethe desiredaccuracy.Thevariable-cellstructuraloptimizationwas per-formedfordopedmodelsinsuchawaythattheexternalpressure waslessthan1.0Kbarandmaximumforcemagnituderemained oneachatomwassetatmostto0.05eV/Å.Weused2×2×2cell forwhichthepureanatasecellcontains48ionsintotal.7× 7×7 Monkhorst–Pack(MP)[67]meshwasusedfork-pointsamplingin theBrillouinzone.ThepartialoccupancyaroundtheFermilevel wastreatedbyGaussiansmearingwithasmearingparameterof 0.05eV.Spinpolarizationwasincludedandthetotalenergywas minimizedupto10−5eVaccuracyforallcalculations.Duringthe optimizations no symmetry constraintswere imposed in order toavoidentrapmentinlocalminimumconﬁgurationsfordoped models.

(4)

Fig.2. Experimentalsetupforphoto-catalyticreactions.

3. Resultsanddiscussion 3.1. Morphologyandstructure

Inordertoinvestigatethemorphologyoftheobtained sam-ples,acomparisonbetweentheSEMimagesofthe500◦Cfor2h annealedsamplesweremade;bare-TiO2 anddifferentco-doped TiO2 filmsisillustratedinFig.3(a)–(d).Topandbottomimages ofthosefiguresshowedlowandhighmagnificationoftheimages, respectively.Fig.3(e)showsatypicalcrosssectionviewtakenfrom sample‘b’.AsestimatedfromtheFig.3(e),thethicknessofthe filmswasapproximately4␮m.Significantcrackingwasobserved onthesurfaceofallfilms.Thefilmsmorphologyarecomparableto thoseofFe3+_and_W6+_doped_TiO

2ﬁlmspreparedonceramicand Tisubstrate,respectivelybyYaoetal.[68].

Fig.4showstheRamanspectraofthestandardTiO2(anatase) andthesampleswithdifferentB/Nratios.Comparisonwiththe RamanspectrumofpureanataseTiO2(curved),whichshowsband at143,196,396,515,and638cm−1,demonstratesthatthesurface layerismainlyconstitutedbyanatase.Conversely,Raman spec-trumofthesampleswithN andBdopingshowastrongsharp bandat140.7cm−1 (Eg),threemid-intensity bandsat393(B1g), 513.5(A1gandB1g)and634.8cm−1(Eg),andaveryweakbandat ∼200cm−1(Eg).ThosebandshavebeenascribedforanataseTiO2 [69–71].TheintensitiesofvariousTiO2featureshavebeenclearly decreasedwithB/NdopingwithrespecttothatofbareTiO2[69]. SincechangeintheparticlesizeswithrespecttoB/Ncontentisnot sosigniﬁcantforoursamples(havebeenexplainedinXRDpart) comparetothoseshowninRef.[69],thereisnotveryclearpicture ofintensitychangesofdifferentanatasefeatureswithindifferent

Fig.3. SEMimagesoftheB/Nco-dopedandbare-TiO2ﬁlmsobtainedonglasssubstratesandannealedat500◦Cfor2hwhereatomicratioofeachdopingisgivenatthe

(5)

0 100 200 300 400 500 600 700 800 900 B/N: 0.27 (a) B/N: 3.83 (b) B/N: 20.89 (c) Bare TiO2 (d) E g R a m a n In te n s it y (a .u .) 634.8 E g 513.5 A 1g B 1g 393 B 1g Raman shift (cm-1) 140.7 E g

Fig.4.RamanspectraofdifferentTiO2thinﬁlms.Themodesofsymmetryofthe

anataseareindicated.

B/Ndopingfilms.ItisusefultounderlinethattheRamanpeaksof theB/NdopedTiO2arebroaderthanthoseofthepureanatase.In particular,particleshavingsmalldimensionandlowcrystallinity arecharacterizedbyRamanpeakbroadening.Theshifts(5cm−1) intheRamanbandscomparetothereferences[69]mightbedue totheNdopinganddifferenceinparticlesizephononconfinement andoxygendeficiency.

XRDpatternsofdopedand bare-TiO2 thin ﬁlmsannealedat 500◦C for 2h are shown in Fig. 5. It is identiﬁed that all the diffraction peaks can beindexed tothe anatase phase of TiO2 (101),(004),(200),(105),(211),(204),(116)and(215)planes [JCPDS:00-002-0406] and noother phase can be detected.No B-and N-derived peaksdue tootheroxides andnitrides have alsobeendetectedinallthepatterns,indicatingthatBandNas dopantin TiO2 exhibitnotendencytosegregateand/or precip-itate indifferent phases duringthe syntheticprocess [72].The Band N in thematrix are assumed to beeither interstitial or systematicallysubstituteTiorOwithoutchangingthehostTiO2 matrix.ItcanbefoundthattheXRDpeakpositionsofdoped sam-plesareingoodagreementwiththosereferenceanatasephaseof TiO2[48,JCPDS:00-002-0406].

Fig.5. XRDpatternsofthethinﬁlmsofco-dopedTiO2:(a)B/N;0.27,(b)B/N;3.83,

(c)B/N;20.89and(d)bare-TiO2. 600 550 500 450 400 350 300 250 200 150 100 50 0 N 1s Si 2s Si 2p Ar Ti 3 p Ti 3 s O 1 s Ti 2 s Ti 2 p B 1s C 1s B/N: 20.89 (c) Bare-TiO₂ (d) Intensity (a.u)

Binding Energy (eV)

Fig.6. TypicalXPSsurveyscansofbare-TiO2andB/N:20.89dopedTiO2ﬁlms.

Thecrystalsizesofallthesamplesareestimatedusingthe Scher-rerequation:

B(2)= K Lcos

whereListhefullwidthathalfmaxima(FWHM)ofthe diffrac-tionpeakofanatase,K=0.89istheshapefactor,isthediffraction angleandistheX-raywavelengthcorrespondingtotheCuK␣ irradiation.Consideringpeakof(101)planesinaccount,the aver-agecrystallinesizesofthesample‘a’,‘b’,‘c’and‘d’are13.43,11.67, 12.99and13.43nm,respectively.Itsuggestedthatthechangein particlesizewithrespecttoB/Ncontentisnotsigniﬁcant.In addi-tion,thereisnochangeinthedspacingvalue;3.5 ˚A,whichimplies that B/N modiﬁcationin co-doping samplesdo not change the averageunitcelldimension.Foroursamplestheaveragelattice parametersareasfollows:a=3.78 ˚A,c=9.44 ˚Aandunitcell vol-ume=134.52 ˚A3_.

3.2. XPSstudy

XPSstudiesusingmono-chromatedAlK␣X-raysourcewere performedtochecktheatomicpercentageofdifferentdoping ele-mentswithintheannealedfilmsandbondingenvironmentamong Ti,Oanddopant.Atomicpercentagesofdifferentelementswere calculatedfromtheXPSnarrowscanpeakintensityconsidering therelativesensitivityfactorofeachelement.Theatomicratiosof borontonitrogen(B/N)fordifferentsamplesareasfollows;0.27(a), 3.83(b)and20.89(c).Fig.6showsthetypicalXPSspectrumofthe B/N:20.89co-dopedTiO2film.Thespectrumfrombare-TiO2 sam-pleisalsoshowninthatfigureforcomparison.XPSpeaksshowed thatthedopedsamplescontainedTi,O,C,SielementswithBand Nasdopant.Thepresenceofcarboncouldbeascribedtothe resid-ualcarbonfromtheprecursor’ssolution.Fig.7(i)showstheB1s XPSspectrumforthesample‘c’whichcontainshighestatomic per-centageofB.Itshowssinglepeakcenteredatbindingenergy(BE) of192.1eV.ReferringtothestandardBEofB1sinB2O3(193.1eV, B Obond)[73]andTiB2(187.5eV,B Tibond)[74],itisspeculated thatboronatommightprobablybeincorporatedintoTiO2matrix andformTi O BorTi B OorO Ti Bbond[52,75].Hencepeak at192.1eVhavebeenassignedtoTi O BorTi B OorO Ti B bond.Fig.7(ii)showstheN1sXPSspectrumforthesample‘a’which containshighestatomicpercentageofN.Consideringsignificant asymmetry,N1speakhasbeendeconvolutedwithintwopeaksat

(6)

196 195 194 193 192 191 190 189 192.1 B/N : 20.89 (c)

(i)

B1s Intensity (a.u.)

Binding Energy (eV)

405 404 403 402 401 400 399 398 397 396 395 B/N: 0.27 (a)

(ii )

400.0 398.5 N1s Intensity (a.u.)

Binding Energy (eV)

Fig.7.XPSnarrowscanofB1sandN1sofdoped-TiO2thinﬁlmswithB/N:20.89

and0.27,respectively.

398.5and400.0eV.Thepeakat398.5eVmaybebecauseofanionic NincorporatedinTiO2inO Ti Nlinkages[22,76],whichmight beresponsibleforvisiblelightphoto-catalysis.Thehigher bind-ingenergypeakat400.0eVareattributedtooxidizednitrogenin theformofTi O NorTi N Olinkages[22,76].Theabsenceofa peakatornear396eVfortheN1scorelevelimpliedthattheTiN phaseorchemisorbednitrogenisnotformedinthenanomaterials [77].

3.3. UV–visabsorptionspectra

Fig.8comparestheUV–visabsorptionspectraofthinfilmsof bare-andB/Nco-dopedTiO2 withdifferentatomicratios.After B/Nco-dopingtotheTiO2,theabsorptionsofcatalystsincreased significantlyintherangeofwavelengthsfrom400to800nm.This clearred-shiftinUV–visspectrarevealsthevisiblelight absorp-tionwithB/Nco-dopedTiO2[52,78–81].Thered-shiftwithinthe absorptionedgesfollowsanincreasingorderas(a)>(b)>(c)(d). ItimpliesthatasatomicpercentageofNishigherinB/Nco-doping case,theredshiftintheabsorptionedgeishigherwhichisingood agreementwiththeresultinreference[52].Morevertheredshift observedintheco-dopingcasesaresignificantlyhighercompare tothatofbare-TiO2case.

300 350 400 450 500 550 600 650 700 750 800 A b sor b anc e ( a .u .) (nm) B/N:0.27 (a) B/N:3.83 (b) B/N:20.89 (c) Bare TiO 2 (d)

Fig.8. UV–visabsorptionspectraofco-dopedandbare-TiO2thinﬁlms.

3.4. Evaluationofphoto-catalyticactivity

Thephoto-catalyticactivityofbare-andB/Nco-dopedTiO2has beenexaminedbyphoto-catalyticdegradationofMB.Fig.9shows thepercentageofdegradationefficiencyforthesamefilmsin(i) cyclenumber1(1◦ cycle),(ii)cyclenumber2(2◦cycle)and(iii) cyclenumber3(3◦cycle).Dataclearlyshowsthatthedegradation efficiencyincreasescontinuouslyduringtheprocessofdegradation forallcycles.Howevertherateofdegradationisclearlydiffered frombare-toco-dopedTiO2films.ThedopedTiO2filmswithB/N atomicratios:0.27,3.83and20.89showedhigherdegradation per-formance,upto70%,whereasbare-TiO2filmsshowedthatof30%. Thephoto-catalyticperformancewithintheco-dopedfilmswere increasedwithNcontentsanddecreasedwithBcontents.However allco-dopedfilmsshowedsignificantlybetteractivityundervisible rangelightcomparetobare-TiO2films.For1◦cycle,degradation performanceforco-dopedfilmsandbare-TiO2filmsare51–63% and28%,respectively.

One of the main problems in the useof TiO2-based photo-catalystsisthedeactivation;mostlyattributedtoblockageofthe activesitesonthesurface,lossofcoatingmaterialfromthesurface duetoerosionoretc[82,83].Toevaluatethepossibilityofcatalyst recoveryand re-use,successivephoto-catalytic MBdegradation wasperformedforallofthedopedandbare-films.Aftereachcycle, filmswereannealedfor1hat500◦Candthenre-used.Fig.9(ii) showsthatsuccessiveutilizationfor2◦cycleinducedevenbetter performanceforallthefilms(a),(b),(c)and(d).For2◦and3◦cycles theefficiencyincreasedto62–69%forco-dopedfilmswhereasthat increasedto32–34%forbare-films.Theincreasingofefficiencyfor 2◦and3◦cyclesmightbe(i)duetothesurfacechemistrychanging ateachtimeheattreatmentaftereachdegradationstudyand/or(ii) thenitrogenconcentrationwithinthestructuremightbeincreased fortheMBsolutionadsorptiononthefilm.Itisstillingreatdebate ofthesynergyeffectsonphotoactivitiesandtheimportanceofthe relativeratiosonphotoactivities[17,50].Furtherworkisongoing tounderstandthisbehavior.

3.5. Theoreticalresults

WehavealsoperformedDFTcalculationsforbulkanatase,single dopedB-toO-andN-toO-andco-dopedBN-orNB-toTiO-cases whereatitaniumatomwasreplacedwithBorNandanoxygen atomwasreplacedbyNorB,respectively.Wehaveinvestigated varioussingledopedcases(e.g.BandNsubstitutiononTisiteand

(7)

0 40 80 120 160 200 240 0 10 20 30 40 50 60 70

(i)

B/N : 0.27 (a) B/N : 3.83 (b) B/N : 20.89 (c) Bare TiO2 (d) % of ef feciency

Irradiation time (min)

0 40 80 120 160 200 240 0 10 20 30 40 50 60 70

(ii)

B/N : 0.27 B/N : 3.83 (a) (b) B/N : 20.89 (c) Bare TiO2 (d) % of ef ficiency

0 40 80 120 160 200 240 0 10 20 30 40 50 60 70

_(iii)

B/N : 0.27 (a) B/N : 3.83 (b) B/N : 20.89 (c) Bare-TiO2 (d) % of Ef ficiency

Fig.9.Percentageofdegradationefﬁciencyofﬁlms(i)incyclenumber1(1◦_cycle),_(ii)_in_cycle_number₂₍₂◦_cycle)_and_(iii)_in_cycle_number₃₍₃◦_cycle).

onOsite,andinterstitialofBandN)andco-dopedcases(e.g. sub-stitutionofBandNondifferentsites,andfordifferentseparations betweenBandN).Wearepresentingthemostrelevantcasesfor theexperimentalsituations.Thelatticeparametersofbulkanatase werefoundtobe3.81 ˚Aand9.72 ˚A(aandc),andc/aratiowas2.55 whichareclosetoourexperimentalvalues(a=3.78 ˚A,c=9.44 ˚A andc/a=2.50).Contrarytotheearlierreportclaiminglarge vol-umeexpansion,whichhassigniﬁcanteffectontheelectronicband structures,upondoping[84],thevolumechangeislessthan1.5% forallofthedopedstructuresinourcalculations.

Fig.10displaysthestructuralmodelsandthetotal(DOS)and partialdensityofstates(PDOS)ofpureTiO2,singledopedB-or N-toO-site andinterstitialN.Thebandgapofanataseis2.04eV, lowerthan experimentalvalue; 3.2eV [85] butconsistent with theprevioustheoreticalstudies[39,86]underestimationofband gapisawell-knowndeﬁciencyofDFTcalculations.Acorrection canbemadeusinghybridfunctional, self-interactioncorrection or‘DFT+U’methods,butinthisstudyourprimarypurposeisto understandthechangesinducedbythedopantsontheelectronic structuresofbulkanatase,nottheexactvalueofthebandgap.PDOS analysisshowsthatthetopmostvalencebandsareformedmainly fromO2pstateswhiletheconductionbandsminimumareformed fromTi3dstates.ForthecaseofBsubstitutingOatom(Fig.10(b)), thehybridizationofB2p,Ti3dandO2porbitalsgivesrisetogap statesappearingat0.27eVbelowtheconductionbandforB-doped O-model.Inthisstructure,Batommakesastrongbondingwithone oftheneighborOatomswithbondlengthas1.36 ˚A.B Tibond dis-tanceis2.11 ˚AwhichisevenlargerthanTi Obonddistanceofpure anatase(1.95 ˚A).Thesystematicinvestigationofnon-metaldoping

ofspeciesofB,C,NandFforsubstitutionalandinterstitialcases hasbeencarriedbyValentinandPacchioni[57].Thegapstatesare expectedforsubstitutionalnon-metaldopingandthepositionsof thesestatesaremainlydependonthenucleareffectivechargeof thedopant.Forlighterelements,the2pstatesinthegapappear higherinthegapthatisclosetotheconductionbandminimum (CBM).OurresultsforBsubstitutiontoOcaseareingood agree-mentwiththisobservation.However,thestates inthegapalso containTi3dstates.ThetypicalpositionofTi3+_states_is_just_below theCBM.Thissituationcanalsobeaddressedtolow electroneg-ativityofBatomcomparedtoOandlesschargetransferfromTi toB.We canconcludethatsubstitutionof BtoOsite produces statesinthegap,whichprovidestheabsorptionofvisiblelight,as wellasTi3+_ions_which_act_like_{electron–hole}_{recombination} cen-tersandthereforediminishthephotocatalyticactivity.Moreover, ahighfrequencyelectronparamagneticresonance(HF-EPR) mea-surementsonBdopedanatase[51]observedsomespindensity onB,TiandOatomswhichmightbeexplainedastheformation ofsubstitutionofBtoOsite.ThevalenceelectronsfromBmight betransferredtolatticeTiions(Ti4+₎_producing_Ti3+_ions_and_it_is well-knownthatthepresenceofTi3+_ions_produce_midgap_states [85].

ForthecaseofNsubstitutingOatom(Fig.10(c)),N2pdefect statesarelocatedatthetopofthevalancebandmaximum.The positionofthesestatesmightberelatedwiththeeffectivenuclear chargeofNatom[57],thatistheN2pstatesappearlowcompared tothatofB2pintheforbiddenregionduetohigheffectivenuclear chargeofNatomcomparedtoBatom.TheN2pstatesarelocalized atthedefectsiteandthepositionofN2pstates.Alongwiththis,

(8)

Fig.10.Relaxedstructuresandpartialdensityofstates(PDOS)of(a)pureanatase,(b)B-dopedO-,(c)N-dopedO-and(d)Ninterstitialmodels.Energiesareshiftedsuch thatFermilevelsarematchedwithzeroofenergy.Thebondlengthsbetweendopedatomsandﬁrstneighborsareindicated.Solidlines,grayshaded,greenshadedandblue shadedareasrepresentthetotalDOS,PDOSofTi3d,O2pandN2pstates,respectively.(Forinterpretationofthereferencestocolorinthisﬁgurelegend,thereaderisreferred tothewebversionofthearticle.)

theEPRmeasurementsalsoindicatethelocalizationofunpaired electrononNdopant[37].Thevisiblelightabsorptionmightbe enhancedbythesubstitutionofNatomtoOsiteduelocalizedstates inthegapregion.Duetothesedefectstates,bandgapisreduced to1.66eVand1.87eVfordownandupspinstates,respectively. Nitrogenatomhasﬁveelectronsinthevalanceshellthereforethe structureisparamagneticandanacceptorstateappearsjustabove theFermilevel.

FortheinterstitialNmodel(Fig.10(d)),therewillbeastrong N Obondingwithabondlengthof1.34 ˚A.Sincethe electronega-tivityofOishigher,NtendstodonateelectronstoOandN O Ti conﬁgurationisformedatthedefectregion.Theseexcesselectrons behaveinthesamemannerofB-toO-model(Fig.10(b)) result-ingtheformationofmidgapstates[37].However,inthiscase,in

additiontotheun-occupiedstateformedattheedgeofconduction band,extramidgapstatesappearbelowtheFermilevel.

Inordertounderstandtheeffectofco-dopinginanatasewe sub-stitutedBandNatomsforvariouspositionsofTiandOsites.Fig.11 displaystherelaxedstructuresandPDOSforBN-andNB-doping toTiO-positionsaswellasthebondlengthstonearestatomsof dopedspeciesisindicated.ForBN-dopedtoTiO-model,Batom transfersitselectronstoOatomsandreachesclosed-shell conﬁgu-rationwhichpreventsBNbonding.TheinteractionofNatomwith itssurroundingatomsproduces midgapbandswhichbehaveas acceptorstatesandalocalizednitrogenstateappearsatthetop ofthevalenceband.Thus,N Ti Otypeofhybridizationproduces midgapstates.PDOSanalysisindicatesthatthereisnoboron con-tributiontoDOSaroundFermilevelbutbandgapofthiscaseis

Fig.11.Co-dopingofanatase:(a)BatomreplacedwithTiandNatomreplacedwithOand(b)NatomreplacedwithTi,andBatomreplacedwithO.Solidlines,grayshaded, greenshadedandblueshadedareasrepresentthetotalDOS,PDOSofTi3d,O2pandN2pstates,respectively.(Forinterpretationofthereferencestocolorinthisﬁgurelegend, thereaderisreferredtothewebversionofthearticle.)

(9)

-4.7 -4.0 -3.0 -2.0 -1.0 0.0 0 1 2 3 4 5 6 7 Ti-Rich O-Rich N doped O B doped O N Interstitial BN doped TiO NB doped TiO Fo rma tio n En erg y (e V) O

Fig.12.Formationenergyasafunctionofchemicalpotentialofoxygenforthe dopedstructures.

foundtobe1.76eVforthespinupstatesand1.2eVforspindown states.Comparedtosingle-dopedmodels,thisisarelativelylarger narrowinginthebandgap.Thus,co-dopingofanatasewithB/Ncan beexpectedtoimprovephoto-degradationwithrespectto single-dopedcasesviabandgapnarrowing.

Incontrastwiththepreviousco-dopingcase,fortheNB-doped TiO-model,thereisabondingbetweenBandNatomswith bond-lengthof1.617 ˚A.SimilartotheB-toO-case,theFermilevelis alsopinnedforthismodeltothebottomoftheconductionband causingaconductorlikebandstructure,whichobviouslydoesnot contributeinthephoto-catalyticactivityofTiO2.

In summary, band gap narrowing leads to the red shift of the absorbance spectra to the visible light region. Meanwhile, the appearance of midgap states could be responsible for the enhancementofthephoto-catalyticactivityeitherbyimproving theabsorbance oravoiding the recombinationof electron–hole pairs[87].

XPS results of the doped films suggested the formation of N Ti OconfigurationupontheexaminationoftheN1sandTi2p spectra(Section3.2).AsshowninFig.10(c)and(d),N Ti O con-figurationcanbeformedinbothN-toO-andNinterstitialcases. Therefore,in ordertounderstandtheenergetically morestable structures,we havecompared theformation energies ofdoped modelsusingthefollowingformalism:

Eform=E

TiO2+B N

−E(TiO2)−kB−lN+mO+nTi whereE(TiO2+B/N)and E(TiO2)arethetotalenergies ofdoped structureand pureanatase,respectively and’sarethe chemi-calpotentialsofcorrespondingspecies.ChemicalpotentialsofN andBarecalculatedfromgaseousphaseofN2and␣-boronphase, respectively.Oissetasthechemicalpotentialofthegasstateof O2forthecaseofOrichenvironment.Ontheotherhand,Tiisset frommetallicTiphaseforthecaseofOpoor(Tirich)environment. FurthermoreweimposedtheconstraintasTi+2O=TiO2 in

ordertoensuretheequilibriumofbulkanatasephase.Integersm andnarethenumbersofOandTivacancies,respectivelywhilek andlarethenumbersofdopedBandNatoms(ifany),respectively. Fig.12displaystheformationenergyofthedifferentdopedmodels asafunctionofthechemicalpotentialofoxygenwithrespecttothe valueofgasphase(O).Variationofoxygenchemicalpotential indicatestherelativeabundanceofoxygeninthesynthesismedium withintwolimitingcases;O2-richandO2-poorenvironments.Two limitsintheﬁgurerepresenttheenvironments;O=0eVforthe

O2-richandO=−4.7eVfortheO2-poororTi-rich.Asdepictedin Fig.12,N-dopedO-caseisenergeticallythemostfavorabledoping modelintheOpoorenvironment,whileBN-dopedTiO-isthestable oneintheOrichenvironment.Thereisacrossoverbetweenthese twomodelsatO=−2.3eV.HoweverNinterstitialcasemight bestabilizedinOrichenvironmentwithinthesingledoped mod-els.ThismightbethereasonfortheformationofN O Tior/and N Ti Obondinginthecaseofourexperiment.

4. Conclusions

Anexperimentalandtheoreticalstudyhasbeenperformedto realizethesynergisticeffectofnonmetaldopinginTiO2for photo-catalytic degradation.Co-doped TiO2 thinfilmswithboronand nitrogenhavebeensuccessfullysynthesizedbysimplesol–geldip coatingmethod.TheB/Nco-dopedTiO2filmsdemonstratedupto 40%higherphoto-catalyticactivitiesthanbare-TiO2 filmsunder visiblelightirradiation.Theabsorptionedgesforthedopedfilms werefoundtobeshiftedtowardthevisibleregion,whilethe over-allabsorptionremarkablyincreasedfordopedfilms.Thefilmwith theB/Natomicratioof0.27displayedthehighestdegradationrate amongalldopedfilms.Thedopedfilmsretainedtheirsuperior cat-alyticactivityforextended periods.Computationalstudieswere conducted onseveralatomic models describing various doping schemes.TheresultsshowedthatdopingwithBand/orNinduced (a)bandgapnarrowing (redshiftof theabsorbance spectrato thevisiblelightregion)and(b)formationofmidgapstates espe-ciallyincaseofNinterstitialmodel.Theseresultsalsosupported theobservedsynergisticeffectsofB/Ndopingforhigher photo-degradationactivity.Thesecomputationalfindingssupportedour experimentaldatabyindicatingthepossibleroutesthat canbe responsiblefortheimprovementofthephoto-catalyticactivityin TiO2 duetoBand Ndoping.Itis revealedthatB/NdopedTiO2 filmscouldbeapotentialcandidateforscalingupforindustrial applications.

Acknowledgments

WeacknowledgethesupportsfromtheDepartmentof Chem-istryofShahjalalUniversityofScienceandTechnology,Bangladesh andTÜB˙ITAK, TheScientiﬁcandTechnologicalResearchCouncil ofTurkey(Grantno:TBAG110T394).Computingresourcesused inthisworkwereprovidedbytheNationalCenterforHigh Per-formance Computing of Turkey (UYBHM) under grant number 10362008.OGacknowledgesthesupportofTheTurkishAcademy ofSciences,TÜBA.PartialfundingforM.N.Uddinisprovidedbythe TurkishUndersecretariatforDefenseIndustries.

References

[1]O.Carp,C.L.Huisman,A.Reller,A.Reller,ProgressinSolidStateChemistry32 (2004)33–177.

[2]B.Tryba,A.W.Morawski,M.Inagaki,AppliedCatalysisB:Environmental46 (2003)203–208.

[3]T.Watanabe,K.Hashimoto,A.Fujishima,in:D.F.Ollis,H.A.Ekabi(Eds.), Pho-tocatalyticPuriﬁcationandTreatmentofWaterandAir,Elsevier,Amsterdam, 1993.

[4]X.Qin,L.Jing,G.Tian,Y.Qu,Y.Feng,JournalofHazardousMaterials172(2009) 1168–1174.

[5]D.A.H.Hanaor,C.C.Sorrell,JournalofMaterialsScience46(2011)855–874. [6]M.Mrowetz,W.Balcerski,A.J.Colussi,M.R.Hoffmann,JournalofPhysical

ChemistryB108(45)(2004)17269–17273.

[7]X.Z.Li,F.B.Li,C.L.Yang,W.K.Ge,JournalofPhotochemistryandPhotobiology A:Chemistry141(2001)209–217.

[8]C.H.Kwon,J.H.Kim,I.S.Jung,H.Shin,K.H.Yoon,CeramicsInternational29 (2003)851–856.

[9]J.Yu,X.Zhao,Q.Zhao,JournalofMaterialsScienceLetters19(19)(2000) 1015–1017.

[10] R.Asahi,T.Morikawa,T.Ohwaki,K.Aoki,Y.Taga,Science293(5528)(2001) 269–271.

(10)

(2003)5483–5486.

[12]X.Chen,S.S.Mao,ChemicalReviews107(2007)2891–2959.

[13]W. Choi, A. Termin, M.R. Hoffmann, Angewandte Chemie 106 (1994) 1148–1149.

[14]C. Wang,D.W.Bahnemann,J.K. Dohrmann,Chemical Communications16 (2000)1539–1540.

[15] R.Asahi,T.Morikawa,T.Ohwaki,K.Aoki,Y.Taga,Science293(2001)269–271. [16]S.U.M.Khan,M.Al-Shahry,W.B.Ingler,Science297(5590)(2002)2243–2245. [17] S.In,A.Orlov,R.Berg,F.Garcia,S.P.Jimenez,M.S.Tikhov,D.S.Wright,R.M. Lambert,JournaloftheAmericanChemicalSociety129(2007)13790–13791. [18] K.Yang,Y.Dai,B.Huang,S.J.Han,TheJournalofPhysicalChemistryB110(2006)

24011–24014.

[19] J.C.Yu,W.Ho,J.Yu,H.Yip,P.K.Wong,J.Zhao,EnvironmentalScienceand Technology39(4)(2005)1175–1179.

[20] M.Sathish,B.Viswanathan,R.P.Viswanath,C.S.Gopinath,Chemistryof Mate-rials17(2005)6349–6353.

[21] L.Lin,W.Lin,J.L.Xie,Y.X.Zhu,B.Y.Zhao,Y.C.Xie,AppliedCatalysisB75(2007) 52–58.

[22] K. Yang, Y. Dai, B. Huang, Journal of Physical Chemistry C 111 (2007) 12086–12090.

[23](a)M.Maitri,K.S.Thushara,B.Saha,P.Chakraborty,C.M.Janet,R.P.Viswanath, C.M.Nair,K.V.G.K.Murty,C.S.Gopinath,ChemistryofMaterials21(2009) 2973–2979;

(b) M. Maitri, K. Sivaranjani,D.S. Bhange, B. Saha, P. Chakraborty, A.K. Viswanath,C.S.Gopinath,ChemistryofMaterials22(2010)565–578; (c)M.Maitri,C.S.Gopinath,ChemistryofMaterials21(2009)351–359. [24]H.Liu,L.Gao,JournaloftheAmericanCeramicSociety87(2004)1582–1584. [25]J.Yu,M.Zhou,B.Cheng,X.Zhao,JournalofMolecularCatalysisA:Chemical246

(2006)176–184.

[26]J.Xu,J.Li,W.Dai,Y.Cao,H.Li,K.Fan,AppliedCatalysisB79(2008)72–80. [27] M.Satish,B.Viswanathan,R.P.Viswanath,C.S.Gopinath,JournalofNanoscience

andNanotechnology9(2009)423–432.

[28](a)P.Xu,T.Xu,J.Lu,S.Gao,N.S.Hosmane,B.Huang,Y.Dai,Y.Wang,Energy& EnvironmentalScience3(2010)1128–1134;

(b)K.Yang,Y.Dai,B.Huang,JournalofPhysicalChemistryC111(2007) 18985–18994.

[29]E.Mete,D.Uner,O.Gulseren,S.Ellialtioglu,PhysicalReviewB79(2009)125418 (1–15).

[30] E.Mete,O.Gulseren,S.Ellialtioglu,PhysicalReviewB80(2009)035422(1–9). [31]M.Fittipaldi,D.Gatteschi,P.Fornasiero,CatalysisToday,inpress

[32]Y.Nakano,T.Morikawa,T.Ohwaki,Y.Taga,AppliedPhysicsLetters87(2005) 052111–052114.

[33] C.K.Xu,S.U.M.Khan,Solid-StateLetters10(2007)B56–B59.

[34] H. Wang, J.P. Lewis, Journal of Physics: Condensed Matter 18 (2006) 421.

[35]Y.F.Lee,K.H.Chang,C.C.Hu,K.M.Lin,JournalofMaterialsChemistry20(2010) 5682–5688.

[36] O.Diwald,T.L.Thompson,T.Zubkov,E.G.Goralski,S.D.Walck,J.T.Yates,Journal ofPhysicalChemistryB108(2004)6004–6008.

[37]C.D.Valentin,G.Pacchioni,A.Selloni,S.Livraghi,E.Giamello,JournalofPhysical ChemistryB109(2005)11414–11419.

[38] C.D.Valentin,G.Pacchioni,A.Selloni,PhysicalReviewB70(2004)085116(1–4). [39]C.D. Valentin,G. Pacchioni, A.Selloni, Chemistry of Materials17 (2005)

6656–6665.

[40]Y.Liu,X.Chen,J.Li,C.Burda,Chemosphere61(2005)11–18.

[41]Z.Lin,A.Orlov,R.Lambert,M.C.Payne,JournalofPhysicalChemistryB109 (2005)20948–20952.

[42]S.Livraghi,M.C.Paganini,E.Giamello,A.Selloni,C.D.Valentin,G.Pacchioni, JournaloftheAmericanChemicalSociety128(2006)15666–15671. [43]J.Y.Lee,J.Park,J.H.Cho,AppliedPhysicsLetters87(2005)011904(1–3). [44]W.K.Ho,J.C.Yu,S.C.Lee,JournalofSolidStateChemistry179(2006)1171–1176. [45]K.M.Reddy,B.Baruwati,M.Jayalakshmi,M.M.Rao,S.V.Manorama,Journalof

SolidStateChemistry178(2005)3352–3358.

[46]W.Zhao,W.H.Ma,C.C.Chen,J.C.Zhao,Z.G.Shuai,JournaloftheAmerican ChemicalSociety126(2004)4782–4783.

Research45(2006)4110–4116.

[48]V.Gombac,L.D.Rogatis,A.Gasparotto,G.Vicario,T.Montini,D.Barreca,G. Balducci,P.Fornasiero,E.Tondello,M.Graziani,ChemicalPhysics339(2007) 111–123.

[49]M.Batzill,E.H.Morales,U.Diebold,PhysicalReviewLetters96(2006)026103 (1–4).

[50] G.Liu,Y.Zhao,C.Sun,F.Li,G.Q.Lu,H.M.Cheng,AngewandteChemie Interna-tionalEdition47(2008)4516–4520.

[51] A.M.Czoska,S.Livraghi,M.C.Paganini,E.Giamello,C.DiValentin,G.Pacchioni, PhysicalChemistryChemicalPhysics13(2011)136–143.

[52] M.Xing,Y.Wu,J.Zhang,F.Chen,Nanoscale2(2010)1233–1239.

[53]X.Zhou,F.Peng,H.Wang,H.Yu,J.Yang,JournalofSolidStateChemistry184 (2011)134–140.

[54]Y.G.Yang,Z.L.Ping,L.Shen,Y.J.Hua,ChineseJournalofStructuralChemistry 27(11)(2008)1353–1359.

[55]C.Baiocchi,M.C.Brussino,E.Pramauro,A.B.Prevot,L.Palmisano,G.Marci, JournalofMassSpectrometry214(2002)247–256.

[56]I.M.Arabatzis,T.Stergiopoulos,M.C.Bernard,D.Labou,S.G.Neophytides,P. Falaras,AppliedCatalysisB:Environmental42(2003)187–201.

[57] C.D.Valentin,G.Pacchioni,CatalysisToday,inpress

[58]M.C.Payne,M.P.Teter,D.C.Allen,T.A.Arias,J.D.Joannopoulos,Reviewsof ModernPhysics64(1992)1045–1097.

[59]W.Kohn,L.J.Sham,PhysicalReview140(1965)A1133–A1138. [60]P.Hohenberg,W.Kohn,PhysicalReviewA136(1964)B864–B871. [61]P.E.Blöchl,PhysicalReviewB50(1994)17953–17979.

[62]G.Kresse,D.Joubert,PhysicalReviewB59(1999)1758–1775. [63]G.Kresse,J.Hafner,PhysicalReviewB48(1993)13115–13118. [64]G.Kresse,J.Furthmüller,ComputationMaterialsScience6(1996)15–50. [65]G.Kresse,J.Furthmüller,PhysicalReviewB54(1996)11169–11186. [66]J.P.Perdew,J.A.Chevary,S.H.Vosko,K.A.Jackson,M.R.Pederson,D.J.Singh,C.

Fiolhais,PhysicalReviewB:CondensedMatter46(1992)6671–6678. [67]H.J.Monkhorst,J.D.Pack,PhysicalReviewB13(1976)5188–5192. [68]M.Yao,J.Chen,C.Zhao,Y.Chen,ThinSolidFilms517(2009)5994–5999. [69]K. Sivaranjani, C.S. Gopinath, Journal of Materials Chemistry 21 (2011)

2639–2647.

[70]D.Bersani,P.P.Lottici,X.Z.Ding,AppliedPhysicsLetters72(1998)73–75. [71]M.J.Uddin,F.Cesano,F.Bonino,S.Bordiga,G.Spoto,D.Scarano,A.Zecchina,

JournalofPhotochemistryandPhotobiologyA:Chemistry189(2007)286–294. [72] Y.H.Peng,G.F.Huang,W.Q.Huang,AdvancedPowderTechnology23(2012)

8–12.

[73]J.A. Schreifels,P.C. Maybury, W.E.Swartz, Journalof Catalysis 65(1980) 195–206.

[74] G.Mavel,J.Escard,P.Costa,J.Castaing,SurfaceScience35(1973)109–116. [75] J.Xu,Y.Ao,M.Chen,D.Fu,JournalofAlloysandCompounds484(2009)

73–79.

[76]T.C.Jagadale,S.P.Takale,R.S.Sonawane,H.M.Joshi,S.I.Patil,B.B.Kale,S.B.Ogale, JournalofPhysicalChemistryC112(2008)14595–14602.

[77] C.D.Valentin,E.Finazzi,G.Pacchioni,A.Selloni,S.Livraghi,M.C.Paganini,E. Giamello,ChemicalPhysics339(2007)44–56.

[78]J.Yang,H.Bai,X.Tan,J.Lian,AppliedSurfaceScience253(2006)1988–1994. [79] J.Yang,H.Bai,Q.Jiang,J.Lian,ThinSolidFilms516(2008)1736–1742. [80] C.H.Wei,X.H.Tang,J.R.Liang,S.Y.Tan,JournalofEnvironmentalSciences19

(2007)90–96.

[81]L.Mi,P.Xu,P.N.Wang,AppliedSurfaceScience255(2008)2574–2580. [82]J.Arana,E.T.Rendon,J.M.D.Rodriguez,J.A.H.Melian,O.G.Diaz,J.P.Pena,Applied

CatalysisB:Environmental30(2001)1–10.

[83]L.Cao,Z.Gao,S.L.Suib,T.N.Obee,S.O.Hay,J.D.Freihaut,JournalofCatalysis 196(2000)253–261.

[84]K.Yang,Y.Dai,B.Huang,PhysicalReviewB76(2007)195201(1–6). [85]J. Pascual, J. Camassel, H. Mathieu, Physical Review Letters 39 (1977)

1490–1493.

[86]H.Weng,J.Dong,T.Fukumura,M.Kawasaki,Y.Kawazoe,PhysicalReviewB73 (2006)121201(R)(1–4).

[87]E.Finazzi,C.D.Valentin,G.Pacchioni,JournalofPhysicalChemistryC113(2009) 220–228.