ContentslistsavailableatScienceDirect
Applied
Catalysis
A:
General
j o ur na l h o me 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 a
An
improvement
of
photocatalytic
activity
of
TiO
2
Degussa
P25
powder
V.G.
Bessergenev
a,∗,
M.C.
Mateus
a,
A.M.
Botelho
do
Rego
b,
M.
Hantusch
c,
E.
Burkel
caUniversidadedoAlgarve,FCT,CampusdeGambelas,Faro8005-117,Portugal
bCQFMandIN,InstitutoSuperiorTécnico,UniversidadedeLisboa,Lisboa1049-001,Portugal cUniversityofRostock,InstituteofPhysics,Rostock,Germany
a
r
t
i
c
l
e
i
n
f
o
Articlehistory:
Received28January2015
Receivedinrevisedform29April2015 Accepted3May2015
Availableonline11May2015 Keywords: Nano-powder Titaniumdioxide Photocatalysis Watertreatment
a
b
s
t
r
a
c
t
ThephotocatalyticactivitiesofDegussaP25powdersannealedatvarioustemperaturesinvacuumand
airwerestudiedtogetherwithinvestigationsoftheircompositionsbyXPS,oftheircrystalstructuresby XRDandoftheirspecificsurfaceareasbyBET.ItisshownthatthephotocatalyticactivityofP25powders
wasremarkablyenhancedaftervacuumannealing;thekineticcoefficientcanberaisedby75%during
annealingat400◦C.ItisobviousthatthisenhancementisnotrelatedtotheadhesionabilityoftheP25 powders.
©2015ElsevierB.V.Allrightsreserved.
1. Introduction
Itisgenerallyacceptedthatthetheoreticalmodelof photocata-lysisonaTiO2surfaceconsistsofdifferentconsecutivesteps,where
eachoneisessentialfortheactivityandtheefficiencyofthe photo-catalyst[1].Theinitialstepofthephotocatalyticprocessconsistsin thegenerationofelectron–holepairsuponirradiationofthe mate-rialwithphotonsofenergiesthatareatleastequaltothoseofthe bandgapvalues.Inthesecondstep,theformedelectron–holepairs caneitherrecombineinthebulkordiffusetothesurfacewhere theycanparticipateinchemical reactions.In thisstep,the life-timeandthevelocityoftheelectron–holerecombinationiscrucial. Chargecarriertrapsareusedtopromotethetrappingofelectrons andholesatthesurfaceleadingtorecombinationsuppressingand tomoreefficientchargetransferprocesses.
ThenextstepisthecreationofH•*andOH•*radicalsasaresult oftheelectron–holeinteractionwithwaterand,finally,aprobable multiplestepreactionoforganiccompoundswithactiveradicals. Theelectrontransferprocessis moreefficientifthespeciesare pre-adsorbedonthesurface[2].However,sofar,theinfluences ofthedifferentstagesofthismodelonthefinalefficiencyofthe photocatalystswerenotwelldetermined.
∗ Correspondingauthor.Tel.:+351289800900x7614;fax:+351289800066. E-mailaddress:vbess@ualg.pt(V.G.Bessergenev).
ItshouldbeemphasizedthatthephotocatalyticactivityofTiO2
dependsonquiteanumber ofparametersincludingthecrystal structure,theratiobetweentheanataseandtherutilephases,the particlesize,thespecificsurfaceareaandthemeanporesize[3]. Inaddition,TiO2 isamaterialthatisveryadequatetobuildtiny
structuresinallsortsofsizesandshapes[4].Inparticular,TiO2
samplespreparedandtreatedbydifferentmethodsexhibitagreat varietyinphotocatalyticefficiencyeventhoughtheyhavethesame crystalform.Beyondthis,thekineticconstantofthephotocatalytic degradationdependsontheconditionsoftheexperiment,namely, whethertheheterogeneouscatalysisistakingplaceonthesolid/gas surfaceoronthesolid/liquidone.
Usingfirst-principlescalculations,thetotalenergiesofthe stoi-chiometricTiO2surfacesalongwiththeequilibriumshapes(the
Wulffconstruction)ofmacroscopicrutileTiO2crystalshavebeen
established[5].Itwasshownthatthe(110)surfacehasthe low-estsurface energy andthat (001)surface hasthehighest one. Therespectivecalculationsofthesurfaceenergeticsandtheshape reconstructionofmacroscopiccrystalsforanataseTiO2werealso
effectuated[6,7].Anatase(101)and(100)surfaceshavethelowest energiesof 0.44and 0.53J/m2,respectively, andthe(110)
sur-facehasthehighestenergyof1.09J/m2whichareallhigherthan
therutile (110)surfaceenergy of 0.31J/m2.Asa consequence,
someofthesesurfacescanbepresentinnano-crystalline mate-rials. However,itshouldbetaken intoaccount,for comparison withexperimentalcrystalshapes,thattheWulffconstructiongives theequilibriumcrystalshapeofmacroscopiccrystalandthatthe http://dx.doi.org/10.1016/j.apcata.2015.05.002
calculationsare strictlyvalidonlyat zerotemperature.Mostof themethodsfornanocrystalssynthesisdonotpresentequilibrium thermodynamicconditions.
ThebulkstructuresofreducedTiO2−xcrystalsarequitecomplex withvarioustypesofdefectssuchasoxygenvacancies,Ti3+and
Ti4+interstitialsandplanardefectssuchascrystallographicshear
planes(CSPs).Thedefectstructurevarieswithoxygendeficiency andthequestionwhichtypeofdefectisdominantinwhichregion ofoxygendeficiencyisstillasubjectofstudies[3].
Thereisavastspectroscopicandchemicalevidenceforthe pres-enceofpointdefectsinsamplesannealedinvacuumthatleadsto thepresenceofTi3+ions.Theirconcentrationistypicallyreported
asbeingofseveralpercent.Theultravioletphotoemission spec-troscopy(UPS)spectraofsingle-crystalsurfacescontainabandgap featureat∼1eVbelowtheconductingbandedgewhichhasbeen interpretedintermsofoccupiedTi3dstates[8].
Variousattemptsforimprovingthephotocatalyticpropertiesof TiO2byannealingthesamplesinairand/orinvacuumhavebeen
made.Theseattemptswerebasedondifferentideas.Someofthe obtainedresultsarediscussedbelow.Before,itshouldbe empha-sizedthatTiO2nanopowderspreparedbydifferentmethodscan
havenotonlydifferentmorphologyanddifferentsurfacechemical composition,butalsodifferentcrystalcompositions.Forthis rea-son,alltheexperimentalresultscanbedividedintotwoclasses, namely,into resultsobtainedwithpowderspreparedby differ-entmethods[9–13]andintoresultsobtainedwithP25powders [14–17].Itshouldbeemphasizedthatnostudiesofchangesin sur-facechemicalcompositionalongwithathermaltreatmentwere doneinalltheabovementionedexperimentalstudies.Thekind ofheterogeneouscatalysisusedforphotolysisexperimentisalso important.
The existence of anomalies in temperature dependence of dielectricconstantinanatasethinfilmswasexperimentally estab-lishedin[18].Theseanomaliescanbeeasilycreatedbyannealingin vacuumandbesuppressedinlowpressureofoxygenand, there-fore,theyaresupposedtobeconnectedwithoxygenvacancies. Itmeansthatelectronicchargedistributionsintheregionaround vacanciesarenotsymmetricalsothatadditionaldipolemoments arepresent[18],whichcanfavourablyinfluencetheelectron–hole separation.Thismechanismwasnotconsideredneitherby the-orynorinexperiments.Therefore,theessentialnewfeatureinthe presentpaperis theaimtocreateoxygenvacanciesby anneal-inginvacuumandtocomparetheresultswithannealinginair. Since,thepresenceofvariousspeciesadsorbedontheTiO2surface
influencesthephotocatalyticactivityandchangesinthesurface chemicalcompositionduringtemperaturetreatmentXPS measure-mentswere performed.In addition, based onthe factthat the phasesofanataseandrutileofP25containseparated nanocrys-talsandthatadirectelectrontransportbetweenrutileandanatase throughthe commonboundarydoesnot exist, itis possibleto estimatechangesinthephotocatalyticactivityofanatasephase, separately.
2. Materialsandmethods
2.1. Samplepreparation
SampleswerepreparedusingtheDegussaP25powderwitha grainsizeofabout20nm.Oneseriesofthesampleswasannealed invacuum(∼2.5×10−5mbar)attemperaturesbetween20 and 750◦Cfor4h.Duringannealingthesampleswereplacedin tita-niumfoilcrucibleswithinastainlesssteelfurnacewithacopper cylinderfortemperaturehomogenization.Asheaters,twohalogen lamps(1kW)wereused.Anotherseriesofsampleswereprepared byannealinginairatthesametemperaturesfor4h.
Fig.1. TEMimageoftheDegussaP25powdersampleannealedat200◦C,during
4h.
InFig.1,atypicaltransmissionelectronmicroscopy(TEM)image ofthesampleannealedinvacuumat200◦Cisshown,evidencing thatthegrainsizesremainedafterannealing.
2.2. Phasecontrolmeasurements
Thecrystalstructureofthesampleswasstudiedusinga high-resolution PANalytical x’PertPRO XRD instrument with Cu K␣ radiation(=1.54184 ´˚A)in–2configuration.Theanatase/rutile ratioin thesampleswasdeterminedusingRietveld refinement procedureprovidedbyFullProfprogram[19,20]aswellas com-mercialPANalyticalx’PertPROsoftware.For typicalXRDpattern 2intervalwas10–100◦withthestep=0.008◦,theresultsof RietveldrefinementareshowninFig.2.Thetypicalparametersof therefinementwereRP≈5.4,RWP≈7.81,GOF–index≤1.7.
2.3. XPScompositioncontrol
TheTiO2sampleswereanalyzedbyX-rayphotoelectron
spec-troscopy (XPS) using XSAM800 (KRATOS) X-ray spectrometer operatedinthefixedanalysertransmission(FAT)mode.AnAlK␣ (1486.7eV)X-raysourcewasused.Theanalyserwasoperatedat 20eVpassenergybothfordetailedandsurveyspectra.Allthe bind-ingenergieswerereferencedtotheC1speakat285.0eV,assigned tocarbonsinglyboundtocarbonandhydrogen.Forquantification purposes,thefollowingsensitivityfactorswereused:Ti2p–1.8, O1s–0.66andC1s–0.25.Otherdetailsondataacquisitionand treatmentwereaspublishedelsewhere[21].Somestudies dedi-catedtothedeterminationoftherelationbetweenTi3+/Ti4+states
forthesamplesannealedinsituinultrahighvacuum(10−9mbar) weredoneusingXPSsetupwithaPHI5600cispectrometeratthe LeibnizInstitutfürWerkstoffkunde,Dresden.Amonochromated AlK␣(1486.7eV)X-raysourceisused.Theanalyseroperatedat 5.85eVpassenergyfordetailedspectrausinga0.05eVstepwidth. CalculationoftheTi3+/Ti4+concentrationisbasedonthearearatio
ofthesetwotitaniumspeciesintheTi2p3/2-peak.
2.4. Effectivesurfaceareameasurement
Todeterminethespecificsurfaceareaandtheporesize dis-tribution,aMicromeriticsASAP2020–PhysisorptionAnalyzerwas
Fig.2. RietveldanalysisoftheX-raypowderdiffractionfromDegussaP25sampleannealedinvacuumat400◦Cduring4h.ThepositionsoftheBraggreflectionsareindicated
byverticalbars(|).Thedifferencebetweentheexperimental(dots)andthecalculated(solidline)intensitiesfromtherefinedmodelisshownbytheplotinthelowerpart ofthediagram(WinPLOTR–programoftheFullProfsuit).
used.Beforethemeasurement,thesamplesweredegassedat50◦C for 10hat a pressure of 50mmHg. The weight of the sample wasdeterminedbeforeandafterthemeasurement.Anisothermal adsorptionofnitrogenwasdoneat77K.Thespecificsurfacearea iscalculatedbytheBETequationintherangeof0.05≤p/p0≤0.25.
The pore size distributions and pore areas are derived by the Barret–Joyner–Halenda(BJH)analysiswithHalseycorrectionsin therangeof0.42≤p/p0≤1oftheadsorptionbranchesforN2on
thepowders.
2.5. Photocatalyticactivitymeasurements
Thelayoutofthephotoreactorapparatuswasdescribedindetail in[22].Briefly,thevolumeofirradiatedsolutionwasof7.9mlwith anopticalthicknessof1mm.A200WXenonlampwasusedas anUV lightsourceand awaterfilterwithfusedsilicawindows wasusedinordertoavoidanexcessiveheatingofthesolution. ThephotocatalyticprocessinwaterrequiresthepresenceofO2as
electronacceptor;thusacontinuousflowofairwasinjectedby meansofanairpump.Flowcontrolwasperformedbystabilization ofthesameairpressureattheinletofthephotoreactor. Fenari-mol(Riedel,99.7%)5mg/lsolutionswerepreparedinbidistilled water.TheFenarimolsolutionwasleftincontactwiththereactor cellabout12hinthedarkbeforeirradiationinordertoachieve theadsorptionequilibriumofthepesticidebetweenthesolution andthecellsurfaces.Duringirradiation,100lsamplesweretaken every0.25h,andimmediatelyanalyzedonaHPLCsystem (Agi-lentTechnologies1220LCInfinitysystemwithUVdetector)under thefollowingexperimentalconditions:LichroCART125-4column: Lichrospher100RP-18,5m;eluent:acetonitrile(Merck Lichro-solv)65%, MilliQ water 35%; 1.0ml/min flow; UV detection at 220nm.Thetotalirradiationtimewasbetween1.5and3hforeach sample.
Twokindsofexperimentswereperformed.Itwasnoticedthat thesedimentationrateswerethesameandthattheywerevery slowforthesamplesannealedattemperatures200–600◦C. How-ever,forthesamplesannealedatevenhighertemperaturesthe sedimentationratesweredrasticallyhigher.Thefirstkindof exper-imentswasperformedwithpowdersamplesdispersedinwater
solutions (10mg/50ml,for thesampleswithTannealing≤600◦C).
Inordertobeabletocomparethereactionratesforallthe sam-ples,anothertypeofexperimentwasperformed.Hereby,samples withidenticalweightweredeposited onfusedsilica substrates byevaporationfromaTiO2aqueoussuspension(10mg/10ml,the
finalamountofpowderwasdeterminedbyweightingafterwater evaporation)andtestedunderidenticalexperimentalconditions.
3. Results
InTable1theresultsofXRDandBETexperimentsforthe sam-plesannealedinvacuumandinairfor4hareshown.
3.1. X-raydiffractionresults
ItcanbeseenfromTable1,thatsignificantmodificationsof crystalstructureandcorrespondingchangesinanatase/rutileratios takeplaceforthesamplesannealedattemperatureshigherthan 550◦C. For the samples prepared at temperatures higher than 600◦C,thereisalsosignificantdifferenceintheanatase/rutileratio forthesamplesannealedinvacuumandinair,thedataareshown inFig.3.
Forthesamplesannealedinvacuum,thephasetransformation fromtheanatasephasetotherutilephaseoccursfasterthanforthe samplesannealedinairatthesametemperatures.Itcanbe sup-posedthatsurfaceoxygendefects,appearingduringthesample annealinginvacuum,createmoredisorderedsurfacestates hav-inglowerpotentialbarriersforatomicdiffusion,thusallowingfor fasterphasetransformationsthaninthecaseofsampleannealing inair.
3.2. XPSsurfacecompositioncontrol
The quantitative results of the surface compositions of the samples,annealed inairandin vacuum,obtainedfromXPSare presentedinTable2.Ti2pspectrawerefittedwithasinglepeak assignedtoTi4+.However,giventhepronouncedtailofthepeak
atlowerbindingenergies(Ti2p3/2)–aGaussian–Lorentzian
Table1
RutileandanatasecontentinthesamplesannealedinvacuumandairfromXRDdatabyRietveldrefinementanalysisandcorrespondingsurfaceandcumulativeporeareas determinedbyBETexperiments.
Annealingtemperature(◦C) Rutile Anatase BETsurface
area(SA),m2/g
Cumulativepore area(CA),m2/g
RatioCA/SA
%,offraction Particlesize,nm %,offraction Particlesize,nm Vacuum 200 16.37 42.4 83.63 25.4 36.0 15.3 0.425 400 16.17 43.3 83.83 25.6 29.0 25.8 0.889 550 22.29 43.5 77.71 26.8 41.0 37.1 0.904 600 32.97 51.7 67.03 28.4 39.9 35.8 0.897 620 59.74 54.5 40.26 27.7 31.4 21.9 0.697 650 89.98 68.4 10.02 33.9 19.9 14.1 0.709 670 91.46 75.5 8.54 39.1 15.5 10.5 0.677 700 100 86.0 0 – 11.0 7.1 0.645 Air Notannealed 16.67 42.3 83.33 25.0 41.27 – 200 16.53 42.6 83.47 25.2 55.5 44.1 0.794 400 16.57 44.5 83.43 25.7 57.2 41.0 0.717 550 21.78 48.5 78.22 28.6 44.9 32.9 0.733 600 32.29 50.0 67.71 28.4 38.2 31.4 0.822 620 46.58 61.0 53.43 32.6 29.8 30.2 1.013 650 63.41 66.5 36.59 32.4 22.3 19.9 0.892 670 77.03 75.9 22.97 35.2 20.3 19.4 0.956 700 93.44 97.8 6.56 45.8 13.3 11.6 0.872 750 99.2 119.6 0.8 – – –
thespectra–wecannotdiscardtheexistenceofasmall contri-bution(around2%)correspondingtoTi3+species.Someimportant
relationsarealsoshowninFig.4(a–d).
AsitcanbeseenfromXPSdata,thecontentofthestructural oxygen(latticeO2−)is,withintheexperimentalerror,thesame foralltemperaturesofannealinggivingthecompositionformula TiO2.15± 0.2forthesamplesannealedinvacuumandTiO2.10± 0.11
forthesamplesannealedinairthatcorrespondstoanexperimental errorofapproximately10%,takingintoaccountthatthe informa-tionaboutthe O2− species is extractedfromthe fittedpeak. It
shouldbeemphasizedthat thesamplesinXPSanalysispresent differentanatase/rutileratios. Up to-day,theequilibrium phase diagramforoxygencontentinTiO2anataseandrutilephasesat
differenttemperaturesinvacuumisstillunknown.Itisnot obvi-ousthat,thechangesinoxygencontentinvacuumarethesame, inthesetwophases.Theobtainedresultsshowthatthosechanges aresmallerthan10%.
Essentialdecreasesincarbonandnon-structuraloxygen con-tents are observed when the annealing temperature increases from500to600◦Cforbothseriesofsamplesannealedinairandin
Table2
XPSpeaksintensitiesandrelativecontentsobtainedforthesamplesannealedinairandinvacuum.
Tannealing,◦C 25 200 400 550 600 620 650 670 700 750 BE,eV Assignment
Air Ti,2p3/2 7.49 6.89 8.09 8.29 12.52 13.23 13.67 13.48 13.64 14.07 458.6 TiO2 Ti,2p1/2 3.44 3.28 4.05 3.99 5.98 6.37 7.41 7.99 7.26 7.38 464.2 O1,1s 24.78 22.85 24.82 26.57 39.04 38.90 41.03 42.28 43.76 45.83 529.7 O2–,structural O2,1s 4.57 3.50 3.16 2.68 4.58 4.51 5.82 4.83 4.77 4.79 531.0 O,OH O3,1s 4.51 4.87 4.66 4.58 3.59 3.54 3.28 3.46 3.04 3.11 532.2 O C O4,1s 1.96 3.01 2.56 2.95 1.29 1.18 0 0 0 0 533.5 H2O,O–C C1,1s 41.24 44.97 44.38 42.14 27.98 26.60 24.26 21.49 24.80 21.69 285.0 C–C,C–H C2,1s 8.46 7.09 5.26 4.84 3.42 4.06 2.40 4.36 1.45 1.52 286.8 C–O C3,1s 3.55 3.54 3.01 3.96 1.59 1.60 1.92 2.11 1.27 1.61 289.0 O–C O,CO3 O/Ti 3.3 3.4 2.9 3.0 2.6 2.5 2.4 2.4 2.5 2.5 O2–/Ti 2.3 2.2 2.0 2.2 2.1 2.0 1.9 2.0 2.1 2.1 C/Ti 4.9 5.5 4.3 4.1 1.8 1.6 1.4 1.3 1.3 1.2 O(nO2−)/Cox 0.9 1.1 1.3 1.2 1.9 1.6 2.1 1.3 2.9 2.5 Cox/Ctotal 0.2 0.2 0.20 0.2 0.2 0.2 0.2 0.2 0.1 0.1 Vacuum Ti,2p3/2 8.14 7.80 8.16 11.1 14.93 13.50 12.97 12.87 459.1 TiO2 Ti,2p1/2 4.26 3.73 4.01 5.70 7.67 6.39 6.93 6.18 464.8 O1,1s 26.49 26.15 26.36 36.90 45.22 43.24 40.82 41.54 529.7 O2−,structural O2,1s 6.55 7.50 2.59 4.54 4.97 4.45 5.27 6.04 531.0 O,OH O3,1s 3.23 3.59 5.14 2.51 1.91 3.49 3.07 3.71 532.2 O C O4,1s 0 0 2.54 0 0.64 0.72 0.81 0.46 533.5 O–C O5,1s 0 0 0.51 0 0.34 0.40 0.11 0.22 535.4 H2O C1,1s 41.40 41.18 41.68 32.30 20.40 23.08 25.87 23.68 285.0 C–C,C–H C2,1s 6.84 7.20 5.52 4.98 2.47 3.08 2.57 3.38 286.4 C–O C3,1s 3.10 2.85 3.48 1.95 1.45 1.64 1.57 1.93 288.9 O–C O O/Ti 2.93 3.23 3.05 2.61 2.35 2.63 2.52 2.73 O2–/Ti 2.14 2.27 2.17 2.19 2.00 2.17 2.05 2.18 C/Ti 4.14 4.44 4.17 2.33 1.08 1.40 1.51 1.52 O(nO2−)/Cox 0.98 1.10 1.20 1.02 2.01 1.92 2.24 1.96 Cox/Ctotal 0.19 0.20 0.18 0.18 0.16 0.17 0.14 0.18
Fig.3. Anatase/rutileratioindependenceonannealingtemperatureinvacuumand inair.
vacuum.Alsotheoxidizedcarbongetsmorestronglyoxidized:(O nonO2−)/Cox changesfromanaveragevaluearound1till550◦C
toanaveragevaluearound2from600◦Con(Fig.4aandc).These speciesofoxygenandcarbonusuallycomefromO–,OH–,H
2O,O–C
absorbedonthesurfaceorareobtainedduringtheTiO2synthesis.
Itcanbeaccentuatedthat,afterbeingannealed,thesampleswere exposedtoairanddidnotreturntotheformercontentofthese kindsofspecies.CommonfeaturesinFig.4aandcarethatthevalue O2–/TiremainsconstantandthevalueO/Ti(O–totaloxygen
con-tent)decreasesfrom3to2.5startingfrom550◦Cunderannealing inairandinvacuum.Relativecontentsofvariousoxygen contain-ingspeciesobtainedfromoxygenpeakbyfitting,namelyO1(O2–,
structural),O2 (OH),O3(O=C),O4,O5 (O-C,H2O)are shownin
Fig.4bforthesamplesannealedinairandinFig.4dforthesamples annealedinvacuum.Itcanbeseenthattemperaturedependence oftheO1/Ototalratioisdifferentforthesamplesannealedinairand
invacuum.TheO1/Ototalratioalwaysincreaseduponannealingin
airwhichmeansthatbothphases,anataseandrutile,gainsome oxygenimprovingtheirstoichiometries,andthattheO1/Ototalratio
decreaseduponannealinginvacuumstartingfrom620◦C,where predominantlytherutilephaseexists(seeTable1).TheO2/Ototal
ratio(O-OandO-H)hasacomplexbehaviouruponannealinginair, but,takingintoaccounttheerrorofabout10%,itcanbeconsidered asapproximatelyconstantandinvacuumadecreaseofthisratiois observed.TheO3/Ototalratio(O=Cspecies)decreasesinbothcases
andthe(O4,O5)/Ototal(H2O,O-Cspecies)ratioalsodecreasesupon
annealinginairanditisapproximatelyzeroinvacuum.
3.3. SpecificBETsurfacearea
An adsorption measurement exhibits several structural fea-tures.Itispossibletoconcludeabouttheparticleshapefromthe
Fig.4. (a)XPSatomicratiosOtotal/Ti,O2–/Ti,Ctotal/Ti,(OnonO2−)/Cobtainedforthesamplesannealedinair.(b)XPSatomicratiosO1/Ototal(O1-O2−),O2/Ototal(O2-O,OH),
O3/Ototal(O3-O=C)andO4/Ototal(O4-H2O,O-C)obtainedforthesamplesannealedinair.(c)XPSatomicratiosOtotal/Ti,O2–/Ti,Ctotal/Ti,(OnonO2−)/Cobtainedforthesamples
annealedinvacuum.(d)XPSatomicratiosO1/Ototal(O1-O2−),O2/Ototal(O2-O,OH),O3/Ototal(O3-O=C)and(O4,O5)/Ototal(O4-H2O,O-C)obtainedforthesamplesannealed
Fig.5.AdsorptionbranchesofN2-adsorptionmeasurementsat77K. shapeoftheisothermandtheporousstructurefromthetypeofthe
hysteresis[23,24].Eachheattreatedtitaniapowderinthispaper showsatypeIVisotherm.Thecharacteristicsfortheadsorption branchofthisisothermareshowninFig.5andaresimilartothe isothermtypeII.
Astrongincreaseofadsorbedgasmoleculesisobserveduptoa relativepressureofp/p0=0.05whereamonomolecularcoverageis
reached.Thehigherthequantityofadsorptionatthispointis,the biggerthespecificsurfaceareais.Typicaladsorptionbranchesare showninFig.5.
Athigherrelativepressures,theadsorptionbranchespresent asmallincreaseuptop/p0=0.7.Theslopeofthecurvedepends
onthemesoporousstructure.Inmesoporesitispossibletobuild upamultimolecularcoatingwhileinmicroporesitisnot possi-ble.Thecurveforthepowdertreatedat600◦Cinvacuumshows
a bigger slope than the one treated at 200◦C in vacuum.The conclusion is that powders annealed at higher temperatures includemoremesoporeswhichisillustratedinFig.7.
Atrelativepressuresnear1,structuralinformationforthe exter-nalsurfaceareaandmacroporesareincluded,sinceonlyforthese structuresafurtheradsorptionispossible.Thesamplesannealed at lower temperatures are showing a higher increase, so it is expectedthattheratioofporeareaandspecificsurfaceareaislower [23,24].
Thedifferences betweentheisothermtypesIIand IV arein thedesorptionbranchand areillustratedinFig.6. Ahysteresis isdetectablebecauseofcapillarycondensationinthepores.The shape of thehysteresisdeterminesthefavourableshape of the poresandtheendofthehysteresis,wheretheadsorptionbranch and the desorption branch are equal, determines the smallest
Fig.7.Poresizedistributionofselectedsamples.
diameteroftheopenpores.Becausethispointisatratherhigh relativepressuresforeachsample,itcanbeconcludedthatonly themacroporesareopenporeswhilethemesoporesareclosed.
ThecombinationofatypeIVisothermandahysteresistypeH3 foreverysampleleadstothefactthatthepowdersaggregateto plate-likeparticleswithslitshapedpores[23,24].Thecalculated specificsurfaceareasandporeareasarelistedinTable1.
TheTiO2 powderstreated in air showa higherspecific
sur-faceareaincomparison withthesamplesannealedinvacuum, except,forthesamples,treatedat600and620◦C.Forthese sam-ples,specificsurfaceareasarenearlyequal.Forsamplestreated inair, theratiobetweenthepore areaandthespecific surface areais increasingwithraisingthetemperaturefrom400upto 700◦C,whereasthisratioisdecreasingforvacuumtreated sam-ples.Between600and620◦Cthementionedratiohasthelargest changeforbothatmospheres.Thistemperaturerangecorresponds toastrongphasetransitionfromanatasetorutile[25],asitcanbe seeninTable1andFig.3.
Forvacuumannealedsamples,aconsiderableincreaseinthe specificsurfacearea,observedat550and600◦C,canbeprobably explainedbyvacuumcleaningofthesurfacebecauseofevaporation ofspeciescontainingcarbon(seeFig.4c).Thiseffectofincreasein thespecificsurfaceareafor vacuumannealedsampleswasalso observedin[16].
Theporesizedistribution,Fig.7,showsamaximumofpores between1.7and4nmforeverysampleexceptforthe200◦Ctreated invacuum.Forthissample,noporesaredetectedinthissizerange. Theairtreatedpowdersupto600◦Ccontainahigheramountof detectablemicropores,between1.7and2.0nmdiameter,aswell asmesopores.
Thepowdertreatedat400◦CexhibitsasmallBETspecific sur-face area under vacuum conditions and the largest one under normalairconditions,becauseofthemesoporousarea.The evapo-ratingorburningoforganicrestsorabsorbedwaterinthepowder producescavities[23].However,thisrequiresoxygeninthe atmo-sphereorhightemperatures,whichexplainsthedifferencesinthe poresizedistributionofthe200and400◦Csamplesforboth atmo-spheres.Intheoxygenrichatmosphere,organicrestsareburned outinthetemperaturerangeupto300◦C,whereasan evapora-tionofthesereststakesplaceinvacuuminatemperaturerange
between200and400◦C.Therefore,theairtreatedsamplesforma higherspecificsurfaceareaduetoahigherporeareacomparedto thesamplestreatedinvacuumatlowertemperatures.
Asaconsequenceofthegraingrowth,poresareclosedandthe specificsurfaceareaaswellastheporeareaisdecreasing.Athigher temperatures,anewformationofporesisdue tocrystallization fromamorphousmaterialtoanataseandre-crystallization,from anatasetorutile[26].Thehigheramountofrutileinthevacuum treatedpowderscombinedwiththesmallergrainsizeimpliesthat poreformationattemperaturesaround600◦Cismorefavourable forthesesamplesthanfortheairtreatedones.Thisleadstothe highamountofporesintherangefrom4to10nmforthe600◦C samplesandthechangeinthementionedratiobetweenthepore areaandthespecificBETsurfaceareabetween600and620◦C. 3.4. Photocatalyticactivity
Table3showstheresultsofthestudyofthekineticsofFenarimol photocatalyticdegradation.
Therearethreedifferentdegradationprocessesthatoccurinthe photocatalyticreactorsimultaneouslywithverydifferentkinetic constants.Thefirstprocessisthedegradationontheanatasephase surfaceofTiO2powder.Thisprocesshasthehighestkinetic
con-stant.Thesecondprocessisthedegradationontherutilephase surfacewiththekineticconstantmuchlowerthanthatofanatase. Samplesannealedinvacuumattemperaturesequaltoorhigher than 700◦C have 100% of rutile phase and show such kinetic constants(seeTable3).ThethirdprocessistheFenarimoldirect degradationbyUVlightinthereactorwithoutTiO2powderwhich
hasthelowestkineticconstant.ThekineticconstantofFenarimol degradationcanbecalculatedfromthefirstorderkineticeq.(1): ln
I t I0 =−k×t (1)whereItandI0arethecurrentandinitialHPLCchromatographic
integratedpeakintensitiesat220nm,kisthekineticconstantand tistimeinhours.Allthreeprocessesofdegradationare indepen-dentoneachotherand,accordingwith[17],thereisnosynergetic effectofco-presenceofanataseandrutileonphotocatalytic activ-ityofP25,consequently,kineticconstantcanberepresentedasa
Table3
Photocatalyticactivityofthesamplesannealedinvacuumandinairwithtwodifferentfunctionmodesofthereactor:powderdepositedontheglassanddispersedinthe solution.
Tannealing,◦C Samplemass,mg k,h−1 kA/g,(h×g)−1 Effectivearea,m2/g kAS/m2,(h×m2)−1
Powderannealedinvacuum
20 8.7 3.02 421 41.27 10.2 200 9.0 3.45 425 36.0 11.81 400 9.1 5.28 618 29.0 21.31 550 7.9 2.89 424 41.0 10.34 600 8.5 2.31 354 39.9 8.87 620 7.9 2.63 725 31.4 23.09 650 8.2 0.77 498 19.9 25.03 670 7.3 0.67 519 15.5 33.48 700 9.5 0.39 – 11.0 – Directdegradation 0 0.26 – – –
Powderannealedinair
20 8.7 3.02 421 41.27 10.2 200 8.0 2.57 343 55.5 6.18 400 8.4 2.58 327 57.2 5.72 550 7.6 2.74 414 44.9 9.22 600 8.5 2.43 370 38.2 9.69 620 7.9 2.57 505 29.8 16.95 650 9.2 2.14 536 22.3 24.04 670 8.0 0.87 288 20.3 14.19 700 7.4 0.55 394 13.3 29.62
Powderannealedinvacuum.Suspension
20 10.0 0.891 69 41.27 1.67
200 10.0 0.838 62 36.0 1.72
400 10.0 1.039 86.1 29.0 2.97
600 10.0 0.725 52.1 39.9 1.31
700 10.0 0.350 – 13.3 –
sumoftheconstantsk=kA+kR+kD,wherekA,kRaretheconstants
foranataseandrutilephasesandkDistheconstantofdirect
degra-dation.FromTable1thefractionsoftheanataseandrutilephases canbedeterminedforeachsampleannealedinvacuumorinair. BycalculatingspecifickineticconstantfortherutilephasekR/m,
wheremisthemassofthe100%rutilesample,itispossibleto cal-culatethespecifickineticconstantfortheanatasephase,according toeq.(2):
kA=k−
(mR×kR+kD)
mA
(2)
The specific kinetic constants normalized for the mass in dependenceonannealingtemperatureareshowninFig.8.
Asit can beseen from Table3 thechanges in specific area forthesamplesannealedinvacuumandin airattemperatures T≤400◦C arequitedifferent.Thesesamples,however, havethe sameanatase/rutileratios(seeFig.3orTable1)andconsistmostly oftheanatasephase(83%).Thus,thechangesinthespecificareacan beattributedforanatasephaseandthespecifickineticconstants, normalizedinrespectofthearea,canbecalculatedfromEq.(3). In supportof thisconclusionitcanbementioned thatthe spe-cificareasforthesamplesannealedinvacuumandinairat700◦C
Fig.8. Specifickineticcoefficient(kA)ofanatasephasecalculatedperunitmass.Thegraphofatypicalkineticcurve,obtainedforthesampleannealedinvacuumat650◦C
Fig.9. Specifickineticcoefficient(kAS)ofanatasephasecalculatedperunitof
spe-cificarea.
arealmostthesame(thesampleannealedinairhassmallfraction (6.56%)ofanatase).
kAS= kA
Specificarea (3)
TheresultsarepresentedinTable3andinFig.9.Theyshowthat thesamplesannealedat670◦C,bothinvacuumandinair,have valuesofkAS ofaboutthreetimeshigherthanthenotannealed
samples.
4. Discussion
Thereareaseriesofstudiesdedicatedtotheimprovementofthe photocatalyticpropertiesofnon-P25samplespreparedbydifferent methods[9–13].Generally,itwasshownthatbythemethodofheat treatmentofthesekindsofsamples,inairorinvacuum,sometimes, thephotocatalyticactivitycouldbeimproved,however,nostudy succeededtorisethephotocatalyticactivitysignificantlyhigher thanthatoftheP25powders.
Someworks oncommercial Degussa P25 TiO2 powders can
bementioned,separately,because,asminimum,theinitial mate-rials have more or less the same characteristics in different studies.
ItwasfoundbyPorteretal.[14]thatannealingtheP25 sam-plesinairattemperaturesfrom600to1000◦Cfor3hsignificantly affectsboth the microstructureand thephotoactivity. Overthe rangeofcalcinationtemperatures,thesamplescalcinedat650◦C revealedthehighestphotoactivities,whichcanbeascribedtoan improvementinthecrystallinityduetothecalcination[14].An increaseinthephotocatalyticactivityforthesamplecalcinedat 650◦CincomparisonwiththenottreatedP25wasfound∼11.5% (dataobtainedfromTableIVin[14]).
Yuetal.[15]studiedP25powderannealedinairandinvacuum at200,300,400and500◦Cfor3h.Amaximuminthephotocatalytic activitywasobservedforthesampleannealedat400◦Cinair.The increaseinthephotocatalyticactivityincomparisontothevacuum treatedsampleatthesametemperaturewasfoundtobeabout32% (dataobtainedfromFig.2in[15]).Thisincreasewasattributedto thecatalystsurfacewithmoreadsorbedoxygenthatwould gener-atemoresuperoxideanionradicalsunderUVirradiation,resulting inahigherphotoactivity[15].
Theinfluenceofthethermaltreatmentinaironthestructure andthephotocatalyticactivityofTiO2P25(P25nottreated,and
annealedat200,300,400,500,600and700◦C)wasstudiedby Fer-nandesMachadoandSantana[16].Thegreatestactivityforphenol degradationwasobtainedwithuntreated(original)TiO2P25[16].
Recently,acrystallinecompositionofDegussaP25powderswas carefullystudiedalong withthephotocatalyticactivityof artifi-ciallyprepared powdersobtainedbymixing anatase,rutile and amorphous phases previously separated from P25 by selective dissolution[17].Itwasfoundanexistenceofinhomogeneityof as-suppliedP25inthesamepackagepresumablybecauseP25is preparedbygas-flamesynthesis;anatase,rutileandamorphous phasesmaybeproduceddependingonflameconditions[17].It wasalsofoundthatthephotocatalyticactivityoftheamorphous phase isnegligible in comparison withtheonesofthe anatase andtherutilephases.It wasalsorevealedthatthecomponents ofP25,theanatase,therutileandtheamorphousphasesbehave independently withoutany interactions (absence of synergism) [17].
From this short description of theoretical and experimental resultsit canbeconcludedthat:first,there isnoevidencethat nanoparticles,obtainedbydifferentprocedureshavethesame rela-tionbetweensurfaceswithdifferentorientations(likein Wulff construction);second,anexistenceofadditionalelectroniclevels createdbyoxygenvacanciesandsituatedatabout1eVbelowthe conductingbandlevelisprovedonlyfortherutile(110)surface; third,oxygenvacanciescanbealsocreatedontheanatasephase surfacesbuttheenergiesforthecreationofthesevacanciesare higherthanthatonesfortherutilesurfaces.Experimentally,itwas proventhattherutilesurfacehasalowerphotocatalyticactivity thanthatoneoftheanatasephase,exceptinthecaseswhensmall metalclustersarepresentontherutilesurface[17].Theprocesses ofthecrystalstructureformation(confirmedbyXRD),thechanges inthemeso-andmicroporesdistributions(confirmedbyBETand BJHanalysis)andtheformationofoxygenvacanciesandofTi3+and
Ti4+interstitials(confirmedbyESR)canaffectthephotocatalytic
activitiesandallthesefactorscontributesimultaneously.
Asitfollowsfromtheobtainedresultsofthepresentwork,there arefewimportantmechanismsthathavestronginfluencesonthe photocatalyticactivities.
First,itcanbeemphasizedthat,forthesamplesannealedat T≤550◦C in vacuum both configurations of the photocatalytic experiments—TiO2powderdispersedinsolutionanddepositedon
fusedsilicasubstratesgavethesameresults,namelyanincreaseof thephotocatalyticactivityforthesamplesannealedatT=400◦Cin comparisonwiththesamplesannealedinair.Atthistemperature, thereareneithersignificantchangesintheanatase/rutileratio(see Table1)norin thesurfacecomposition(seeFig.4a–d).The dif-ferencebetweenthesurfaceareasforthesamplesannealedinair andinvacuumonlyincreasesthedifferenceinkASvalues.Oxygen
vacancies(orTi3+sites)haveastronginfluenceonthedielectric
constant[18]andontheelectricalresistivity[27,28]andcanbe easilycreatedatthesetemperaturesbyannealinginvacuumor byhigh-energyballmilling[29].RecentlyTi3+surfacedefects,its
properties,generationand photocatalyticapplicationhavebeen reviewed[30].
In this workP25 Degussa powderswereannealed insitu in ultrahighvacuumandrespectiveXPSspectraofTiwereobtained afterannealing. Theregularincrease in Ti3+/Ti4+ratio hasbeen
observedwhentemperatureofannealingwasincreased.The val-uesofTi3+/Ti4+ratioschangesfromapproximately1.3%atroom
temperature(thesamplewasonlyplacedintothechamberwith ultrahighvacuumwithoutheating)to1.38%at200◦Candto2.23% at 350◦C (the samples were heated at respective temperature underultrahighvacuumduring2h).Despitethereducing atmo-sphereofaround10−9mbartheamountofTi3+sitesisverysmall.
However,thepronouncedtailtheTi2p3/2componentinthelow
stronglysuggeststheexistenceoftheTi3+speciesevenifitsrelative
%isjustoftheorderof2%.
Second,whenthetemperatureofannealinginvacuumincreases from550to600◦C,therearestronglossesinthenon-stoichiometric oxygen and surface carbon species, as it is shown in Fig. 4a and c. Adsorbed oxygenon TiO2 surface scavenges the
photo-inducedelectronsandinhibitstheformationofTi3+sites[1].Hence,
whensurfaceadsorbedoxygenandcarboncontainingspeciesare removed,theprocessoftheelectron–holerecombinationbecomes moreeffective.This,inturn,resultsindecreasingthephotocatalytic activityasitisshowninFig.8.Asitcanbealsonoted,theeffective areaat550–600◦C(seeTable1)isalmostequalforthesamples annealedinvacuumandinairandtheinfluenceofchangesinthe effectiveareacanbeexcludedatthesetemperatures.
Third, further increase in the annealing temperatures (Tann>600◦C) results in fast anatase to rutile phase
transfor-mations asit is shown in Fig.3. Theoverall kinetic constant k decreasesstronglyatthesetemperaturesbecausetherutilephase haslowerphotocatalyticactivity.Nevertheless,atatemperature of620◦CthereisacriticalincreaseinkAaswellasinkAS.Unlike
the 400◦C samples, this increase occurs in vacuum and in air treatedpowder.Therearefewprocessestakingplaceatthesame timeatT>500◦C,namely:phasetransformation,improvementof crystallinity(crystalgrowth)andoxygenvacanciesformation.By theabovedescribedanalysisitispossibletoseparatetheeffect of losses in the adsorption ability (decrease in specific surface area) from the electronic processes (hole action) and to show thattheprocesseswhich dependontheelectron–holecatalytic activityincreasetheirefficiencies(kASgrowinguponannealing).
In addition,thespecifickinetic constant kA doesnot changeat
650<T<700◦C in vacuumand 670<T<700◦C in air becauseof two concurrent processes, namely, oxygenvacancies formation and losses in specific area (see Table 1 and Fig. 8). Contrarily, kAS strongly increasesin thistemperature intervalshowingthe
efficiency of the induced electron trapping mechanism by Ti3+
sites(Fig.9).
Fourth,asitcanbeseenfromFig.7andTable1,theformation of poresis more effective during annealing in air than in vac-uum.Thesizedistributionoftheporesisapproximatelythesame inboth cases.However,thephotocatalyticactivity,represented throughtheoverallkineticconstant kand thespecificconstant kA, of vacuum treated samples is higher than that of samples
treatedinair(fortemperaturesupto600◦C).Itmeansthatthe poreswithsizesintheintervalof1.4–4nmdonothave signifi-cantinfluenceonthephotocatalyticactivitiesoftheTiO2powder
samples.
Thereis somecontradiction in theabove mentioned litera-tureabouttheexistence[10]ornon-existence[17]ofsynergism betweentheanataseandrutilephases.Additionally,Hurumetal. [31]observedthetimedependenceintheESRsignalscomingfrom therutileandanataseinaqueousP25slurryaftertheUV illumina-tionwasswitchedoff.Inthistimedependence,theESRsignalfrom therutilephasewasdecreasedand thesignalfromtheanatase phasewasincreasedatthesametime.Itwasinterpretedinterms oftheexistenceofanataseelectrontrappingsitesthatcanonlybe populatedbyelectrontransferfromtherutiletotheanatasephase. Thus,accordingtotheauthors,therutilephaseactsasanantenna toimprovethephotoactivity[31].However,inmostofthecases, theanataseandrutileparticlesrepresentseparatenanocrystals,so thatthedirectelectrontransferacrossrutile/anataseborderwas notconfirmedyetandthereisnoextensionofthephotoactivity intovisiblewavelengths.In[10],aparticularcasewasconsidered, namely,relativelybigrutileparticles(smallspecificsurfacearea) arephotocatalyticallyactiveandcoveredbysmallanatase parti-cles(bigspecificsurfacearea).Therefore,anenhancementinthe photocatalyticactivitycanarisefromanadsorptionabilityofthe
anataseparticles.Inthiswork,itwasconfirmedthatthepurerutile phasehasamuchloweractivitythantheanatasephaseina Fenari-moldecomposition.Sucheffectwasearlierobservedforthinfilms forthesampleswithapproximatelythesamesurfacearea[32]. Thisgivesanopportunitytoestimatethephotocatalyticactivityof anatasephaseseparately.
5. Conclusions
AnnealingoftheDegussaP25powdersinvacuumandinairat differenttemperatureshavebeenusedtostudytheinfluenceof oxygenvacanciesonthephotocatalyticactivityofthesamples.It wasshownthatthephotocatalyticactivityofthesamplescanbe significantlyincreasedbyannealinginvacuum.Thewhole tem-peratureintervalofannealingbetween25and750◦Cshouldbe dividedintotwozones:onezoneupto500◦Cand anotherone higherthan500◦C.ForT<500◦Ctheincreaseinthephotocatalytic activitiesforthesamplesannealedinvacuum(maximumat400◦C) isindependentonchangesinthespecificsurfacearea.Itcanbe explainedbytheabilityofoxygenvacancies(orTi3+sites)toact
ascentresthatpreventtheelectron–holerecombinationand to increasetheefficiencyofthephotocatalyst.Itwasalsoshownthat oxygenandcarboncontainingspeciesadsorbedonthesurfaceor createdin theDegussatechnologicalprocessofsynthesis ofthe powderplayanimportantroleinphotocatalysis.Suchaspecies scavengephoto-excitedelectronsandalsopreventthe recombina-tion.WhenT>500◦C,theprocessesofphasetransformation,the crystalgrowthandthedecreaseinthespecificsurfaceareaacts simultaneously.However,itwaspossibletoshowthatthespecific kineticconstantoftheanatasephase(kAS)increaseswithhigher
annealingtemperatures.Thismeansthattheelectron–hole mech-anismofthedecompositionoforganics(Fenarimol) isstillvery effective.
Acknowledgements
TheauthorsaregratefultoFundac¸ãoparaaCiênciaeTecnologia (FCT)Portugalforthefinancialsupportgivenintheframeworkof thedevelopmentofscientificcentresCIQAandCCMARatAlgarve UniversityandtoFCTprojectPest-OE/CTM/LA0024/2013.
TheauthorsaregratefultoProf.MartinKnupferoftheLeibnitz InstitutfürWerkstoffkunde,Dresden.
References
[1]A.I.Lisebigler,G.Lu,J.T.YatesJr.,Chem.Rev.95(1995)735–758. [2]R.W.Matthews,J.Catal.113(1988)549–555.
[3]U.Diebold,Surf.Sci.Rep.48(2003)53–229. [4]A.N.Banerjee,Nanotech.Sci.Appl.4(2011)35–65.
[5]M.Ramamoorthy,D.Vanderbilt,R.D.King-Smith,Phys.Rev.B49(1994)16721. [6]M.Lazzeri,A.Vittadini,A.Selloni,Phys.Rev.B63(2001)155409.
[7]M.Lazzeri,A.Vittadini,A.Selloni,Phys.Rev.B65(2002)199901.
[8]M.V.Ganduglia-Pirovano,A.Hofmann, J. Sauer,Surf. Sci. Rep.62(2007) 219–270.
[9]R.R.Bacsa,J.Kiwi,Appl.Catal.B16(1998)19–29.
[10]T.Ohno,K.Tokieda,S.Higashida,M.Matsumura,Appl.Catal.A244(2003) 383–391.
[11]S.Bakardjieva,J. ˇSubrt,V. ˇStengl,M.J.Dianez,M.J.Sayagues,Appl.Catal.B58 (2005)193–202.
[12]K.Suriye,P.Praserthdam,B.Jongsomjit,Appl.Surf.Sci.53(2007)3849–3855. [13]T.Xia,Y.Zhang,J.Murowchick,X.Chen,Catal.Today225(2014)2–9. [14]J.F.Porter,Yu-G.Li,C.K.Chan,J.Mater.Sci.34(1999)1523–1531. [15]J.C.Yu,J.Lin,D.Lo,S.K.Lam,Langmuir16(2000)7304–7308.
[16]N.R.C.FernandesMachado,V.S.Santana,Catal.Today107–108(2005)595–601. [17]B.Ohtani,O.O.Prieto-Mahaney,D.Li,R.Abe,J.Photochem.Photobiol.A216
(2010)179–182.
[18]V.Bessergenev,Mater.Res.Bull.44(2009)1722–1728.
[19]J.Rodriguez-Carvajal,AnintroductiontotheprogramFullProf2000, Laborato-rieLéonBrillouin(CEA–CNRS)CEA/Saclay,France.
[20]L.B.McCusker,R.B.VonDreele,D.E.Cox,D.Louer,P.Scardi,J.Appl.Crystallogr. 32(1999)36–50.
[21]A.Garbout,S.Bouattor,A.M.BotelhodoRego,A.M.Ferraria,A.W.Kolsi,J.Cryst. Growth304(2007)374–382.
[22]V.G.Bessergenev,R.J.F.Pereira,M.C.Mateus,I.V.Khmelinskii,E.Burkel,R.C. Nicula,Int.J.Photoenergy5(2003)99–105.
[23]G. Leofanti,M.Padovan,G.Tozzola,B. Venturelli,Catal.Today41(1998) 207–219.
[24]KennethSing,Colloid.Surf.A187(2001)3–9.
[25]C.Charles,A.SorrellDorian,H.Hanaor,J.Mater.Sci.46(2011)855–874. [26]K.Meyer,P.Lorenz,B.Böhl-Kuhn,P.Klobes,Cryst.Res.Technol.29(1994)
903–930.
[27]B.Huber,H.Gnaser,C.Ziegler,Surf.Sci.566–568(2004)419.
[28]R.Amade,P.Heitjans,S.Indris,M.Finger,A.Haeger,D.Hess,Phys.Chem.Chem. Phys.8(2006)777–782.
[29]R.Amade,P.Heitjans,S.Indris,M.Finger,A.Haeger,D.Hesse,J.Photochem. Photobiol.A207(2009)231–235.
[30]L.-B.Xiong,J.-L.Lin,B.Yang,Y.Yu,J.Nanomater.2012(2012)1–13. [31]D.C.Hurum,A.G.Agrios,K.A.Gray,J.Phys.Chem.B107(2003)4545–4549. [32]V.G.Bessergenev,R.J.F.Pereira,M.C.Mateus,I.V.Khmelinskii,D.A.Vasconcelos,
R.Nicula,E.Burkel,A.M.BotelhodoRego,A.I.Saprykin,ThinSolidFilms503 (2006)29–30.