ContentslistsavailableatScienceDirect
Applied
Surface
Science
j o u r n a l ho me 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 s u s c
TiO
2
–Al
2
O
3
binary
mixed
oxide
surfaces
for
photocatalytic
NO
x
abatement
Asli
Melike
Soylu
a,
Meryem
Polat
a,
Deniz
Altunoz
Erdogan
a,
Zafer
Say
a,
Cansu
Yıldırım
b,
Özgür
Birer
b,c,
Emrah
Ozensoy
a,∗aDepartmentofChemistry,BilkentUniversity,06800Ankara,Turkey
bKUYTAMSurfaceScienceandTechnologyCenter,Koc¸University,34450Istanbul,Turkey cDepartmentofChemistry,Koc¸University,34450Istanbul,Turkey
a
r
t
i
c
l
e
i
n
f
o
Articlehistory:Received15November2013
Receivedinrevisedform10February2014 Accepted12February2014
Availableonline22February2014 Keywords: TiO2 Al2O3 Photocatalysis NOxabatement DeNOx
a
b
s
t
r
a
c
t
TiO2–Al2O3binaryoxidesurfaceswereutilizedinordertodevelopanalternativephotocatalyticNOx
abatementapproach,whereTiO2siteswereusedforambientphotocatalyticoxidationofNOwithO2and
aluminasiteswereexploitedforNOxstorage.Chemical,crystallographicandelectronicstructureofthe
TiO2–Al2O3binaryoxidesurfaceswerecharacterized(viaBETsurfaceareameasurements,XRD,Raman
spectroscopyandDR-UV-VisSpectroscopy)asafunctionoftheTiO2loadinginthemixtureaswellasthe
calcinationtemperatureusedinthesynthesisprotocol.0.5Ti/Al-900photocatalystshowedremarkable photocatalyticNOxoxidationandstorageperformance,whichwasfoundtobemuchsuperiortothatof
aDegussaP25industrialbenchmarkphotocatalyst(i.e.160%higherNOxstorageand55%lowerNO2(g)
releasetotheatmosphere).OurresultsindicatethattheonsetofthephotocatalyticNOxabatement
activ-ityisconcomitanttotheswitchbetweenamorphoustoacrystallinephasewithanelectronicbandgap within3.05–3.10eV;wherethemostactivephotocatalystrevealedpredominantlyrutilephasetogether andanataseastheminorityphase.
©2014ElsevierB.V.Allrightsreserved.
1. Introduction
Indoorand outdoorair pollutants suchasNOx,SOx,volatile
organic compounds (VOCs) and particulate matter (PM) result insignificantly adverse effects onhumanhealth. Further nega-tiveimplicationsofair pollutioncanalsobeobservedonwater resources,agricultureandbiologicalhabitat[1–6].Amongthese airbornetoxicspecies,particularlynitrogenoxides(NOx)presenta
majorchallengeforairpurification.NOxspecies(i.e.mostlyNO(g),
NO2(g)andN2O(g))aregeneratedduringthefossilfuel
combus-tionprocessesviathehomogenousreactionofnitrogenandoxygen gasesathightemperatureswherethemajorcontributioncomes fromNO(g).NOxabatementcanbeperformedinaveryefficient
mannerusingthermalcatalytictechnologiessuchasselective cat-alyticreduction(SCR)[7–9]andNOxstorageandreduction(NSR)
(whichis alsocalled LeanNOx Traps, LNT)[10–12] atelevated
temperatures(i.e.T>300◦C).Inthesethermallyactivatedcatalytic DeNOx technologiesalthough SCRapproach requires utilization
∗ Correspondingauthor.Tel.:+903122902121;fax:+903122664068. E-mailaddress:ozensoy@fen.bilkent.edu.tr(E.Ozensoy).
URL:http://www.fen.bilkent.edu.tr/ozensoy(E.Ozensoy).
ofureaas anexternal reducing agent,NSR/LNTtechnologycan be used in the absenceof an additional reducing agent. How-ever,animportantchallengeinairpurificationistheabatement of gaseous NOx species under ambientconditions (i.e. atroom
temperatureandunderregularatmosphericconditions). Photocat-alyticsystemsofferpromisingopportunitiesinordertotacklethis importantenvironmentalchallenge,asthesesystemscanbe tail-oredtoefficientlyclean/purifyairunderambientconditionswith thehelpofultraviolet(UV)and/orvisible(VIS)light[13].Among thesesystemsTiO2-basedmaterialsarethemosteffective
photo-catalystsforair/waterpurificationapplications[14,15].Howeverit hasbeenreportedthatcompletephotocatalyticreductionoftoxic NOxspeciesintoharmlessN2occursonlywitharelativelylimited
performanceforthesesystems[13].
Inthecurrentwork,ratherthanattemptingtoperformcomplete photocatalytic reduction of NOx,an alternative NOx abatement
strategy hasbeen demonstrated, which includes photocatalytic oxidationofNOxonaTiO2/Al2O3binaryoxidephotocatalyst
sur-faceanditsstorageinthesolidstateintheformofnitratesand nitrites.Thisalternativestrategywasinspiredbyourrecent stud-ies onNSR technology which is used for thethermal catalytic aftertreatmentofautomotiveNOxemissions[12,16–21].In
cou-pleof theseformer studies,we spectroscopicallydemonstrated
http://dx.doi.org/10.1016/j.apsusc.2014.02.065 0169-4332/©2014ElsevierB.V.Allrightsreserved.
that[12,16–21]ontheTiO2/Al2O3binaryoxidesurface,oxidized
NOx species suchasNO2(g)canreadilyundergoa thermal
dis-proportionationreactionforming adsorbednitrites and nitrates allowingeffectivesolidstateNOxstorage.HoweverNO(g)hasa
limitedadsorptionenergyonmanymetaloxidesurfacescompared tothatofNO2,hinderingthestorageofNOinthesolid(adsorbed)
state.Thus,forsolidstateNOxstorage,NOshouldbefirstoxidized
toNO2 andthensubsequentlystoredontheavailableadsorption
sitesofthecatalystsurfaceintheformofnitrites/nitrates.Although thiscan bedonereadilyatelevatedtemperaturesusinga plat-inumgroupmetal(PGM)promotedmetaloxidecatalystsuchas Pt/Al2O3,itcannotbeefficientlyachievedunderambient
condi-tions(i.e.atroomtemperature)duetokineticlimitations.However thislimitationcanbeovercomebydesigningacatalytic system includinga photocatalyticNO(g)oxidation componentwhich is coupledtoa NOx storage component. Along theselines, inthe
currentwork,weshowthatTiO2/Al2O3binaryoxidesurfacescan
beexploitedtoperformphotocatalyticNOxoxidationandstorage,
whereTiO2surfacedomainsprovideNOoxidationcapabilityunder
ambientconditions,convertingNO(g)+O2(g)intonitrites/nitrates
whilethehigh-surfaceareaAl2O3componentenablesboththe
dis-persionofthephotocatalyticTiO2domainsaswellasthecreation
ofadditionalstoragesitesforoxidizedNOx.Oncesaturatedwith
NOx,suchaphotocatalyticNOxoxidationandstoragecatalystcan
readilyberegeneratedbytreatmentwithwater,whichcandissolve theadsorbednitrites/nitratesandrestoretheNOxadsorptionsites
[22].
In order to demonstrate this alternative strategy, in the current study, a set of TiO2/Al2O3 binary oxide
photocata-lysts were synthesized and characterized. A sol–gel synthesis method was used to co-precipitate titania with alumina. The influences of the surface structure on the photocatalytic NO oxidation and storage was investigated by modifying the sur-face structure via calcination. Photocatalytic performances of thisnewfamilyofTiO2/Al2O3 binaryoxidephotocatalystswere
alsocomparedwitha commerciallyavailablephotocatalyst(i.e. DegussaP25)inordertodeterminetherelative performanceof theTiO2/Al2O3 system against a widelyused industrial
bench-mark.
2. Experimental
2.1. Samplepreparation
Titanium (IV) isopropoxide (TIP, 97%, Sigma–Aldrich) and aluminum-tri-sec-butoxide(ASB,97%,Sigma–Aldrich)wereused as the main ingredients in the preparation of the TiO2/Al2O3
binaryoxides via sol–gelmethod [16,18].Three series of sam-ples were prepared by varying therelative molar composition of the TiO2 component in the TiO2/Al2O3 binary oxide. These
samples are labeledas “xTi/Al-y”, where x represents theTiO2
to Al2O3 mole ratio (i.e. 0.25, 0.5 and 1.0) and y represents
the calcination temperature (150–1000◦C) of the sample. In thesynthesis,dependingonthecorrespondingTiO2–Al2O3mole
ratio, an appropriate amount of ASB was mixed with propan-2-ol(99.5%,Sigma–Aldrich)andacetylacetone(99.3%,Fluka)for 30min. Subsequently, TIP was added in a drop wise fashion to the mixture over the course of another 30min. All of the synthesis steps were carried out at room temperature under vigorous stirring. The co-precipitation of the obtained hydrox-ides was accomplished after the gradual addition of 0.5M HNO3(aq) to the solution which led to the formation of a
gel. The resulting yellow gel was aged under ambient condi-tions for 2 days and the dried sample was ground to form a fine powder. Next, synthesized TiO2/Al2O3 binary oxides were
calcinedinairfor2hatvarioustemperaturesrangingfrom150 to1000◦C.
2.2. Structuralcharacterizationmeasurements
Determinationofthecrystalstructureofthesynthesized mate-rialswerecarriedoutwithaRigakuMiniflexX-raydiffractometer (XRD)equippedwithCuK␣radiationoperatedat30kV,15mA, and 1.54 ˚A (wavelengthof copper X-raysource). The XRD pat-ternswererecordedinthe2rangeof10–60◦withastepwidth of0.02s−1.Ramanspectraofthesampleswerecollectedin the rangeof200–1500cm−1witharesolutionof4cm−1usingaHoriba JobinYvonLabRAMHR800spectrometerequippedwithaconfocal RamanBX41microscope.TheRamanspectrometerwasequipped witha Nd:YAGlaser (=532.1nm)where thelaser powerwas 20mW. Thespecific surfacearea(SSA) valuesof theTiO2
sam-plesweredeterminedbyconventionalBrunauer–Emmett–Teller (BET)N2 adsorptionmethodusinga MicromeriticsTristar 3000
surface areaand pore sizeanalyzer. Priorto theBET measure-ments,all ofthe sampleswereoutgassed in vacuumfor 2hat 150◦C.DiffuseReflectanceUV–vis(DR-UV–vis)spectrawere uti-lizedinordertoobtainelectronicbandgapvalues.Thesespectra wererecordedwithaShimadzuUV-3600UV-Vis-NIR spectropho-tometerusingtheISR-3100integratingsphereattachmentinthe specularreflection(8◦)mode.Bariumsulfate(BaSO4)wasusedas
thereferencematerialintheDR-UV–vismeasurements.Obtained DR-UV–visspectrawerefinallycorrectedusingtheKubelka-Munk transformation.
2.3. Photocatalyticactivitymeasurements
The custom-designed photocatalytic flow reactor system (Scheme1)wasusedtomeasure thephotocatalyticNOx
oxida-tionandstorageperformancesofTiO2/Al2O3binaryoxidesunder
UVA excitation. The photocatalytic flow reactor system mainly consistedofa gasmanifoldsystem,a samplecompartmentand a chemiluminiscenceNOx analyzer(Horiba APNA-370).The gas
manifoldsystemwasconnectedtogascylinderscontainingN2(g)
(99.998%,LindeGmbH),O2(g)(99.998%,LindeGmbH)and100ppm
NO diluted in N2 (Linde GmbH). Mass flow controllers (MFCs,
MKS 1479A)wereused tocontrolthe volumetric flowrates of gasesandacapacitancepressuregauge(MKSBaratron)wasused tomeasure total pressure of theflowing gaswhich wasset to 1atm. The following flow rates were used to prepare the gas mixture,0.750SLM(standardlitersperminute)forN2(g),0.250
SLM for O2(g),and 0.010SLM for NO(g) witha total gas flow
rate of 1.010 SLM. Prior to mixing, N2(g) and O2(g) were also
bubbledthroughahumidifier.Therelativehumidityofthetotal gas mixture was 70% RH which was measured with a Hanna HI 9565 humidity analyzerat the sample position in the pho-tocatalyticflow reactor.Thisgasmixturerepresentsasynthetic polluted air sample. Before the performance tests, synthesized powdersampleswereplacedona2mm×40mm×40mm poly-methyl methacrylate (PMMA) sample holder and subsequently irradiated with UVA (350nm) light bulbs (F8W/T5/BL350, Syl-vania/Germany) under ambientconditions for 18houtside the flow reactor in order to remove the surface contaminations and toactivatethephotocatalysts. For each measurement, typ-ically a 950mg activatedphotocatalyst samplewasplaced into the flow reactor. The photocatalytic flow reactor was illumi-nated with8W UVA lamps (F8W/T5/BL350, Sylvania/Germany) whoseemissionwavelengthwas350nm.ConcentrationsofNO(g), NO2(g)andtotalNOx(g)speciesinthephotocatalyticreactorwere
quantitativelymeasuredonlinewiththechemiluminiscenceNOx
Scheme1. Descriptionofthecustom-designedphotocatalyticflowreactorsystem.
Gasphasephotocatalyticactivitymeasurementsarereportedin termsofpercentphotonicefficiencies(%)asdescribedinEqs.(1) and(2).
%= nNOx
nphoton×
100 (1)
wherenNOx correspondstoeitherthedecreaseinthetotal
num-berofmolesofallgaseousNOxspeciesorthenumberofmoles
ofNO2(g)generatedina 60min(i.e.3600s)photocatalytic
per-formancetest.Ontheotherhand,nphotoncorrespondstothetotal
numberofincidentUVAphotonsimpingingonthecatalystsurface in3600s,whichcanbecalculatedthroughEq.(2)as:
nphoton=ISt
Nhc (2)
whereIrepresentsthephotonpowerdensity oftheUVAlamp, experimentally measuredat the sample positionin the photo-catalytic reactor (typically, 7.5Wm−2), is the representative emissionwavelengthoftheUVAlamp(i.e.350nm),Sisthe sur-faceareaofthephotocatalystsampleholderinthereactorthatis exposedtotheUVAirradiation(i.e.40mm×40mm=1600mm2);
tisthedurationoftheperformancetest(i.e.3600s),Nisthe Avo-gadro’snumber,hisPlanck’sconstantandcisthespeedoflight.
3. Resultsanddiscussion
3.1. Specificsurfaceareameasurements
Thermal evolution and the structural variations of the TiO2/Al2O3binaryoxidesampleswithvaryingmolarcompositions
wereinvestigatedaftercalcinationstepsatdifferenttemperatures (Fig.1).Fig.1revealsthatTiO2/Al2O3 samplespossesseda
rela-tivelyhighsurfaceareaafterpreparationandcalcinationatlow temperatures(e.g.≥420m2/g).ThesehighSSAvalueswere
pre-servedtoalargeextentupto600◦C.Thisobservationisinverygood accordancewiththecurrentXRDandRamanresults(Figs.2and3) suggestingapredominantlyamorphousstructureforallTiO2/Al2O3
binaryoxidesamplesbelow600◦C.Athighertemperatures,a dras-ticand a monotonic decreasein theSSA values wereobserved in line withthe enhanced crystallinity and structural ordering ofthesamplesatelevatedtemperatureswhicharealsoevident in the current XRD and Ramanmeasurements (Figs. 2 and 3). Itisworthmentioningthatuponcalcinationat900◦C, SSA val-uesfor0.25Ti/Al-900,0.5Ti/Al-900,1.0Ti/Al-900samplesdecreased to108,64and25m2/g,respectively.Theseparticularvaluesare
ratherclosetotheSSAofthecommercialDegussaP25catalyst(i.e. 55m2/g)whichisusedasthebenchmarkphotocatalystinthe
cur-rentstudy.Athighercalcinationtemperaturessuchas1000◦C,SSA valuesforalloftheTiO2/Al2O3 binaryoxidesamplesdrastically
decreasetoc.a.9–17m2/gwhichisinperfectagreementwiththe
increasedcrystallinityandtheformationofthelowsurfacearea phasessuchasrutileand␣-Al2O3(corundum)observedintheXRD
andRamanexperiments(Figs.2and3). 3.2. XRDandRamanspectroscopyexperiments
Fig.2presentsXRDprofilesobtainedfortheTiO2/Al2O3samples
withdifferentmolarcompositionsthatwerecalcinedatvarious temperatureswithin150–1000◦C.Itisapparentthatforall sam-ples,calcinationattemperatureslessthanorequalto600◦Cyields amorphous structures. Calcination at 800◦C resultsin the first discernibleindicationsofcrystallinity,where␥-Al2O3(JCPDS
29-0063)phasestartstobevisiblefor0.25Ti/Aland0.5Ti/Alsamples. Forthe1.0Ti/Alsample,inadditiontothe␥-Al2O3phase,
forma-tionofanatase(JCPDS21-1272)andrutile(JCPDS04-0551)phases ofTiO2alsobecomesvisible.ItisclearthatwithincreasingTiO2to
Al2O3moleratiointhephotocatalystcomposition,crystallinityof
300 400 500 600 700 800 900 1000 1100 0 100 200 300 400 500 600 a er A ec afr u S cifi c e p S (m 2 /g ) Temperature (oC) 0.25 Ti/Al 0.5 Ti/Al 1.0 Ti/Al 470487 285 256 108 17 424 393 131 64 9 390 86 25 9
Fig.1. SpecificsurfaceareavaluesfortheTiO2/Al2O3binaryoxidesampleswith
differentmolarcompositionsthatwerecalcinedatvarioustemperatureswithin 150–1000◦Cinair.
Fig.2. XRDpatternsfortheTiO2/Al2O3binaryoxidesampleswithdifferentmolarcompositionsthatwerecalcinedatvarioustemperatureswithin150–1000◦Cinair.
theobservedphasesincreases.Thisisinlinewiththefactthatpure (bulk)TiO2hasmuchlowerphasetransitiontemperaturesbetween
amorphous,anataseandrutilephasesthantheTiO2 domainson
theTiO2/Al2O3 surface[16,18].Thus atlowTiO2 toAl2O3 mole
ratios,thereexistsastronginteractionbetweentheTiO2
minor-itydomainsandtheAl2O3majoritydomains,whichisdecreasing
thesurfacemobilityoftheTiO2domainsandhinderingthe
nuclea-tionandgrowthofanataseandrutilephasesatlowtemperatures. HoweverathigherTiO2toAl2O3moleratios,interactionbetween
theTiO2andAl2O3domainsweakenstoacertainextentasTiO2
convergestoa morebulk-like configuration,pushingthephase transitiontemperaturestolower(bulk-like)values.
Uponcalcinationat900◦C,although␥-Al2O3seemstobethe
onlydiscerniblecrystallinephaseonthe0.25Ti/Alsurface(where TiO2isstillinamorphousstate),anataseandrutilephasesbecome
clearlyvisibleonthe0.5 Ti/Aland1.0 Ti/Alsurfaceswherethe crystallinityofthelatterissignificantlygreater.Thisisinperfect agreementwiththeSSAvalues presentedinFig.1,suggestinga muchlowerSSAfor the1.0 Ti/Al-900samplecompared to0.25 Ti/Al-900and0.5Ti/Al-900samples.Itisalsoworthmentioning thatalthough␣-Al2O3(JCPDS10-0173)phaseisnotsignificantly
visibleat900◦CforlowerTiO2toAl2O3moleratios;thisphaseis
noticeablydiscernibleforthe1.0Ti/Al-900sample.Furthermore, ␣-Al2O3(corundum)phasestartstoappearduringtheanataseto
rutilephasetransition.Asdiscussedinoneofourformerreports
[16],this canbeexplained bytheformationof asolid solution betweenanataseand alumina.In this solid solution,when the anatasephase isconvertedintorutileatelevatedtemperatures, aphasesegregationoccurswhichtriggersaphasetransitioninthe aluminacomponentfrom␥to␣-phase.Finally,aftercalcinationat 1000◦C,allsamplesseemtobehighlyordered,wherecorundum andrutilearetheonlyvisiblecrystallinephases,inverygood har-monywiththedrasticSSAdecreasesobservedforthesesamplesin
Fig.1.
Raman spectra of the synthesized TiO2/Al2O3 binary oxide
sampleswithdifferentmolarcompositionsthatwerecalcinedat varioustemperatureswithin150–1000◦CaregiveninFig.3.These Ramanspectralfeaturescanbereadilyexplainedinthelightofthe XRDresultsgiveninFig.2,aswellastheformerRaman spectro-scopicstudiesintheliterature[16,18,23,24].Itisknownthatthe RamanspectrumofanatasephaseshowssixRamanfeatures(1A1g,
2B1g,and3Eg)at144(Eg),197(Eg),399(B1g),516(A1g+B1g),639
(Eg)and796cm−1 (Eg)[23].Ontheotherhand,therutilephase
canbecharacterizedbyaRamanspectrumwithfourmajorRaman activefeatures(A1g+B1g+B2g+Eg)at143(B1g),447(Eg),612(A1g),
826cm−1(B2g)andalsoatwo-phononscatteringbandat236cm−1
[24].InverygoodagreementwiththeXRDresultsgiveninFig.2,up to600◦C,allsamplesrevealanamorphousstructurewithnosharp Ramanfeatures.Itisworthmentioningthatsample1.0Ti/Al-600 revealsverybroadandconvolutedRamansignalscorresponding tosmallandpoorlycrystallineanataseandrutiledomainswhich seemtobeelusivetodetectinXRD(Fig.2c).Atcalcination temper-atureshigherthan600◦C,anatasephaseappearsasthedominant phasetogetherwithaminorcontributionfromrutile.With increas-ingtemperature,anatasetorutileratiointhesamplesdecreases whereat900◦Crutilebecomesthepredominantphasedetectedin theRamanspectra.Forthe0.5Ti/Al-900sample,anatasephaseis stillvisibleintheRamanspectra(Fig.3b),althoughrutileis defi-nitelythemajorityphase.InperfectharmonywiththeXRDresults (Fig.2),RamanspectrainFig.3alsosuggestthatincreasingTiO2
toAl2O3moleratioenhancesthecrystallinityofthephasesonthe
TiO2/Al2O3binaryoxidesurfaceswhichisevidentbythesharper
andstrongerRamanscatteringfeatures. 3.3. Photocatalyticperformanceexperiments
Fig.4shows atypicalconcentrationversustime plotthat is obtainedduringaphotocatalyticperformancetest.InFig.4,the totalNOxconcentration(i.e.sumoftheconcentrationsofallofthe
NOxspeciesexistinginthereactor,i.e.bluecurve)aswellas
sep-arateNO(g)(blackcurve)andNO2(g)(redcurve)concentrations
inthephotocatalyticreactormeasuredbythechemiluminiscence NOxanalyzerarepresented.Duringtheinitialc.a.20minofthe
analysis,asyntheticpollutedairgasmixturecomprisedofN2(g),
O2(g),H2O(g) aswell as 1ppm NO(g) is fed to the
photocata-lystsurfaceunderdarkconditionswheretheUVAlampisoffand anybackgroundexposuretosunlightisprevented. Underthese conditions(i.e.inthefirst15min), aminortransientfallinthe
0 20 40 60 80 0.0 0.2 0.4 0.6 0.8 1.0
0.5 Ti/Al-900
Concentration (ppm) Time(min) Light-on Light-off Thermal Adsorpon NOx(g) NO(g) NO2 (g)Fig.4.Concentrationversustimeplotforthephotocatalyticperformancetestofthe 0.5Ti/Al-900sample.Blue,blackandredcurvescorrespondtotheconcentrationsof totalNOx(g),NO(g)andNO2(g),respectively(seetextfordetails).(Forinterpretation
ofthereferencestocolorinthisfigurelegend,thereaderisreferredtotheweb versionofthisarticle.)
totalNOx(g)andNO(g)concentrationswasobserved.Thiscanbe
attributedtothedilutionofthegasinthereactorpipelineandthe thermaladsorptionofNOxspeciesonthegaslines,reactorwallsas
wellasadsorptiononthephotocatalystsurface.Sincethereactor iskeptincompletedarknessundertheseconditions,no photocat-alyticactivityisobservedduringthisinitialstageevidentbythe presenceofaminoramountofNO2(g)productionduetothermal
catalytic disproportionation processesoccurring onthecatalyst surface.Followingthisinitialtransientperiod,reactorwallsandthe photocatalystsurfacearesaturatedwithNOx,afterwhichNOx(g)
andNO(g)tracesquicklyreturntotheoriginalinletconcentration valueofc.a.1ppm,signifyingtheendofthermalcatalyticactivity. Afterthispreliminarytransientperiod,UVAexcitationsource isturnedonandthephotocatalyticreactionisstarted.UponUVA illumination,adrasticandapermanentfallintheNO(g)andtotal NOx(g)concentrationsconcomitanttoaquickandsignificantjump
in theNO2(g)level, were observed.This behavior suggeststhe
conversionofNO(g)intoNO2(g)viaphotocatalyticoxidation.
Fur-thermore,producedNO2(g)canadsorbonthephotocatalystsurface
intheformofchemisorbedNO2,nitritesandnitrates[16,18]and
storedin thesolidstate,resultingin afurtherfallintheNO(g) and total NOx signals. It is worth mentioning that, fall in the
NO(g)concentrationmightalsohavesomecontributionfromthe directphotocatalyticdecompositionandphoto-reductionofNO(g) formingN2(g)and/orN2O(g)[25].However,sincethedirect
photo-catalyticreductionisknowntobearelativelyinefficientpathway, thisreactionchannelmaybeexpectedtobeaminor photochem-icalroute.Consequently,thetotalNOxconcentration(blue)curve
(whichismostlycomprisedofthesumofNO(g)andNO2(g)signals)
inFig.4remains mostlybelow1ppmduringtheUVA-activated regime,illustratingthecontinuousphotocatalyticactivityandNOx
storageinthesolidstate.
Photochemical NO oxidation and storage performance tests wereperformedforallofthesynthesizedsamplesandthe sum-maryoftheseperformancetestswerepresentedintermsofpercent photonicefficienciesinFig.5,alongwiththecorrespondingdatafor theDegussaP25industrialbenchmark.Inthehistogramgivenin
Fig.5,blueandredbarsrepresentthepercentphotonic efficien-ciesfortotalNOx(g)decreaseandNO2(g)production,respectively.
Thesevalueswereobtainedbyintegratingthecorrespondingareas undertheconcentrationversustimecurvesforthedatasimilarto theonesgiveninFig.4.
It is worth mentioning that for an ideal catalyst with an utmostphotocatalyticDeNOxperformance,bluebars(i.e.NOx(g)
storage/conversion) should be maximized; while red bars are simultaneouslyminimized(i.e.minimumslipoftoxicNO2(g)into
theatmosphere).WhenthebehavioroftheDegussaP25 indus-trialbenchmarkphotocatalystgiveninFig.5isinvestigated,itis immediatelyseenthatthisindustrialphotocatalysthasaveryhigh NO(g)photo-oxidation capabilitygeneratinga largequantityof NO2(g),whilethesamecatalyst hasa verylimited NOxstorage
capability(bluebar).ConsideringthefactthatNO2(g)isamuch
moretoxicpollutantthanNO(g),althoughDegussaP25industrial benchmarksystemisveryactiveinphoto-oxidation,thismaterial doesnotqualifytobeaveryefficientphotocatalyticDeNOx
sys-temforNOxabatement.Anotherbenchmarksampleusedinthe
controlexperimentswas␥-Al2O3.Fig.5unambiguouslyindicates
that,␥-Al2O3hasneithersignificantphotocatalyticNOxstoragenor
photocatalyticNO2(g)productioncapabilities.
Ontheotherhand,whenthephotocatalyticperformancedata fortheTiO2/Al2O3 binaryoxidesamplesare examined,onecan
immediatelynotetheremarkableimprovementinthe photocat-alyticDeNOxperformancecomparedtotheDegussaP25industrial
benchmark.InFig.5,performanceresultsfortheTiO2/Al2O3binary
oxidesamplesareassembledinthreegroupsbasedonTiO2toAl2O3
moleratio(i.e.0.25,0.5and1.0)inthephotocatalyststructure.It isvisiblethatforthe0.25Ti/Alsamplescalcinedatvarious tem-peratures,catalystscalcinedbelow900◦CrevealverylowDeNOx
performance,wheretheperformancereachesanoptimumvalue between900and950◦Candstartstofallat1000◦C.
Asimilarperformancetrendisobservedfor0.5Ti/Alcatalysts calcinedatvarious temperatures(Fig.5).For thisfamily of cat-alysts,althoughnosignificantactivityisobservedatcalcination temperatureslessthan900◦C,photocatalyticDeNOxperformance
presentsaveryradicalenhancementat900◦C, revealingvalues thataremuchbetterthananyofthephotocatalystsinthe0.25Ti/Al family.Itisworthmentioningthatafurtherincreaseinthe calci-nationtemperatureto1000◦CresultsinthephotocatalyticDeNOx
performanceofthe0.5Ti/Alsystem.
Fig.5indicatesthatforthe1.0Ti/Alphotocatalystfamily,no significantphotocatalyticactivityisdetectedupto800◦C,whileat thiscalcinationtemperaturearemarkableincrease inthe activ-ity is observed, though this catalyst is not as effective as the 05 Ti/Al-900 catalyst in total NOx abatement, due to the
significantNO2(g)generationoftheformer.ItcanbeseeninFig.5
that for calcinationtemperaturesabove 800◦C, NOx abatement
startstofall,evidentbytheincreasedNO2(g)slipintothe
atmo-sphereaswellasdecreasingNOxstorageinthesolidstate.Thus,
ageneralanalysisoftheperformanceresultspresentedinFig.5
revealsthat,0.5Ti/Al-900binaryoxidecatalystshowsthehighest NOxabatementperformanceamongalloftheanalyzed
photocata-lysts,whereitperforms160%higherNOxstorageand55%lower
NO2(g)releasetotheatmospherecompared totheDegussaP25
industrialbenchmark.
PhotocatalyticperformanceoftheTiO2/Al2O3binaryoxide
sam-plescanbereadilyinterpretedinthelightofcurrentstructural characterization experiments (Figs. 1–3) which reveal valuable insightregardingthespecificsurfaceareasaswellasthe crystal-lographicphasesthatarepresentontheTi/Alsamples.Firstly,itis apparentinFig.5thatforthebestperformingphotocatalystfamily (i.e.0.5Ti/Al),onsetofactivityisobservedinaverydrasticmanner asthecalcinationtemperatureisincreasedfrom800◦Cto900◦C. BET,XRDandRamanmeasurementsgiveninFigs.1–3suggests thatthis thermalwindowdirectlyoverlapswiththe crystalliza-tionoftheamorphousTiO2toformamixtureofanataseandrutile
phaseswherethelatteristhedominantphase.Inotherwords,itis apparentthatinordertoachievethebestphotocatalyticNOx
abate-mentperformance,auniquecrystallographicmixtureofanatase andrutilephaseshastobeobtained.
Secondly,Fig.5alsosuggeststhatforTi/Alfamilieswithdifferent TiO2loadings,ultimateperformanceisobservedforthe
interme-diateloadingandtheperformancewasseentodecreaseforvery loworveryhighTiO2loadings.Thiscanbeexplainedbythefact
thatatlowTiO2loadings,itislikelythatTiO2loadingisnothigh
enoughtobedispersedonalloftheAl2O3 surface.Thusnot all
oftheNOxadsorption/storage(i.e.Al2O3)sitescanbeutilizeddue
tolimitedphoto-oxidationcapabilityoftheinadequatenumberof TiO2oxidationsitesonthesurface.Ontheotherhand,atveryhigh
TiO2loadings,TiO2coversmostoftheAl2O3surfaceandupon
cal-cinationabove800◦C,SSAofthecatalystsamplefallsdrastically togetherwiththeformationofcrystallineanataseandrutile mix-ture;limitingtheavailablenumberofNOxstoragesitesthatare
availableafterphoto-oxidation.
Thirdly, Fig.5 indicates that onset of photocatalytic activity is observed in a rather sharp manner at 950, 900 and 800◦C for the0.25Ti/Al, 0.5Ti/Al and 1.0Ti/Al samples,respectively. In
Fig.5.PhotocatalyticDeNOxperformanceresultsfortheTiO2/Al2O3binaryoxidesampleswithdifferentmolarcompositionsthatwerecalcinedatvarioustemperatures
Fig.6. ElectronicbandgapvaluesderivedfromDR-UV–visspectroscopicresultsfortheTiO2/Al2O3binaryoxidesampleswithdifferentmolarcompositionsthatwere
calcinedatvarioustemperatureswithin150–1000◦Cinair.
otherwords,astherelativeTiO2loadingintheTiO2/Al2O3binary
oxidesamplesincreases,onsettemperatureforthephotocatalytic activityshiftstolowertemperatures.Thiscanalsobeexplained bytheonsettemperatureforthecrystallizationofTi/Alsamples (and hence the formation of photo-active TiO2 sites) observed
inXRDandRamanmeasurements(Figs.2and3)whichsuggest thatincreasingTiO2 loadingincreasesthetemperaturerequired
to switch form an amorphous TiO2 structure to a crystalline
structure.
3.4. DR-UV–vismeasurementsandelectronicbandgap
Inordertoinvestigatetherelationshipbetweentheelectronic structure and the photocatalytic NOx abatement performance,
electronicband gap values were calculated from the currently performed(notshown)DR-UV–visspectroscopicmeasurements. ThesebandgapvaluesarepresentedinFig.6.Inverygood agree-mentwiththediscussiongivenabove,electronicbandgapvalues fortherelativelyinactiveamorphousTi/Alsampleswhichare cal-cinedatlowertemperatures,revealacharacteristicallyhighvalue within3.4–3.6eV.Ontheotherhand,withtheonsetofthe pho-tocatalyticactivity,a very sharpfallin theelectronicbandgap valueswereobserved,where thebandgapdecreasestoa typi-calvalueof3.05–3.10eV,in linewiththeformationof ordered anataseandrutilephases.Typicalbandgapvaluesforbulkanatase andrutilephasesarec.a.3.2and3.0eV,respectively[26].Thus,for theactivephotocatalystsamples,thebandgapvalueisinbetween thatofanataseandrutile,beingclosertothelatter,inaccordance withthefactthatin themostactivephotocatalyst,rutileexists asthepredominantphasetogetherwithanataseastheminority phase.
Itisalsoworthnotingthatalthoughonsetofthephotocatalytic activityasafunctionofcalcinationtemperaturecanbefollowed withtheelectronicbandgapvalues,electronicbandgapcannot beusedasasoleindicatorfortheestimationofthephotocatalytic activitytrends.Thisisduetothefactthatoncethe photocatalyt-icallyactivestructureisobtainedleadingtoadrasticdecreasein theelectronicbandgap,bandgapvaluesceasetochangeathigher calcinationtemperaturesalthoughphotocatalyticactivitystartsto decline.
4. Conclusions
TiO2–Al2O3 binary oxide surfaces were utilized in order to
develop analternative photocatalyticNOx abatement approach,
where TiO2 sites were used for ambient photocatalytic
oxida-tion of NO with O2 and alumina sites wereexploited for NOx
storage.Chemical,crystallographicandelectronicstructureofthe TiO2–Al2O3 binaryoxidesurfaceswerecharacterizedasa
func-tionoftheTiO2loadinginthemixtureaswellasthecalcination
temperatureusedinthesynthesisprotocol.0.5Ti/Al-900 photocat-alystshowedremarkablephotocatalyticNOxoxidationandstorage
performancewhichwasfoundtobemuchsuperiortothatofa DegussaP25industrialbenchmarkphotocatalyst(i.e.160%higher NOx storageand 55%lower NO2(g)release totheatmosphere).
OurresultsindicatethattheonsetofthephotocatalyticforNOx
abatementactivityisconcomitanttotheswitchbetween amor-phoustoacrystallinephase withanelectronicbandgapwithin 3.05–3.10eVwherethemostactivephotocatalystrevealed pre-dominantly rutile phase together withanatase as theminority phase.
Acknowledgments
AuthorsacknowledgeZaferSayforperformingBET measure-ments. E.O. also acknowledges financial support from Turkish AcademyofSciencesthroughthe“TUBA-GEBIPOutstandingYoung Scientist Prize” and from Fevzi Akkaya Science Fund (FABED) throughEserTümenScientificAchievementAwardaswellasthe Scientific and Technical Research Council of Turkey (TUBITAK) (ProjectCode:109M713).
References
[1]O.Carp,C.L.Huisman,A.Reller,Photoinducedreactivityoftitaniumdioxide, Prog.SolidStateChem.32(2004)33–177.
[2]A.K.Gupta,K.Karar,S.Ayoob,K.John,Spatio-temporalcharacteristicsof gaseousandparticulatepollutantsinanurbanregionofKolkata,India,Atmos. Res.87(2008)103–115.
[3]S.Ishii,J.N.B.Bell,F.M.Marshall,Phytotoxicriskassessmentofambientair pollutiononagriculturalcropsinSelangorState,Malaysia,Environ.Pollut.150 (2007)267–279.
[4]R.Chen,B.Zhou,H.Kan,B.Zhao,Associationsofparticulateairpollution anddailymortalityin16Chinesecities:animprovedeffectestimateafter accountingfortheindoorexposuretoparticlesofoutdoororigin,Environ. Pollut.182(2013)278–282.
[5]AirQualityGuidelinesforEurope,2nded.,WorldHealthOrganizationRegional OfficeforEurope(WHORegionalPublications,EuropeanSeries,No.91), Copen-hagen,2000.
[6]Airqualityandhealth,WorldHealthOrganization,FactSheetNo:313,Updated September2011.
[7]R.M.Heck,Catalyticabatementofnitrogenoxides—stationaryapplications, Catal.Today53(1999)519–523.
[8]Y.Traa,B.Burger,J.Weitkamp,Zeolite-basedmaterialsfortheselective cat-alyticreductionofNOxwithhydrocarbons,MicroporousMesoporousMater.
30(1998)3–41.
[9]G.Busca,L.Lietti,G.Ramis,F.Berti,Chemicalandmechanisticaspectsofthe selectivecatalyticreductionofNOxbyammoniaoveroxidecatalysts:areview,
Appl.Catal.B:Environ.18(1998)1–36.
[10]W.S.Epling,L.E.Campbell,A.Yezerets,N.W.Currier,J.E.ParksII,Overviewofthe fundamentalreactionsanddegradationmechanismsofNOxstorage/reduction,
Catal.Rev.Sci.Eng.46(2004)163–245.
[11]S.Roy,A.Baiker,NOxstorage-reductioncatalysis:frommechanismand
mate-rialspropertiesto storage-reductionperformance,Chem.Rev. 109(2009) 4054–4091.
[12]E. Ozensoy, J. Szanyi, C.H.F. Peden, Model NOx storagesystems: storage
capacityandthermalagingofBaO/-Al2O3/NiAl(100),J.Catal.243(2006)
149–157.
[13]J.Lasek,Y.H.Yu,J.C.S.Wu,RemovalofNOxbyphotocatalyticprocesses,J.
Pho-tochem.Photobiol.C:Photochem.Rev.14(2013)29–52.
[14]V.Loddo,G.Marci,C.Martin,L.Palmisano,V.Rives,A.Sclafani,Preparation and characterisation of TiO2 (anatase) supported on TiO2 (rutile)
cata-lystsemployedfor4-nitrophenolphotodegradationinaqueousmediumand comparisonwithTiO2(anatase)supportedonAl2O3,Appl.Catal.B:Environ.20
(1999)29–45.
[15]A.Mitsionis,T.Vaimakis,C.Trapalis,N.Todorova,D.Bahnemann,R.Dillert, Hydroxyapatite/titaniumdioxidenanocompositeforcontrolledphotocatalytic NOxoxidation,Appl.Catal.B:Environ.106(2011)398–404.
[16]S.M.Andonova,G.S.Senturk,E.Kayhan,E.Ozensoy,NatureoftheTi–Ba inter-actionsontheBaO/TiO2/Al2O3NOxstoragesystem,J.Phys.Chem.C113(2009)
11014–11026.
[17]E.Kayhan,S.M.Andonova,G.S.Senturk,C.C.Chusuei,E.Ozensoy,Fepromoted NOxstoragematerials:structuralpropertiesandNOxuptake,J.Phys.Chem.C
114(2010)357–369.
[18]S.M.Andonova,G.S.Senturk,E.Ozensoy,Fine-tuningthedispersionandthe mobilityofBaOdomainsonNOxstoragematerialsviaTiO2anchoringsites,J.
Phys.Chem.C114(2010)17003–17016.
[19]E.Emmez,E.I.Vovk,V.I.Bukhtiyarov, E.Ozensoy,Directevidenceforthe instabilityanddeactivationofmixed-oxidesystems:influenceofsurface seg-regationandsubsurfacediffusion,J.Phys.Chem.C115(2011)22438–22443. [20]E.I.Vovk,E.Emmez,M.Erbudak,V.I.Bukhtiyarov,E.Ozensoy,Roleofthe
exposedPtactivesitesandBaO2formationinNOx storagereduction
sys-tems:amodelcatalyststudyonBaOx/Pt(111),J.Phys.Chem.C115(2011)
24256–24266.
[21]G.S.Senturk,E.I.Vovk,V.I.Zaikovskii,Z.Say,A.M.Soylu,V.I.Bukhtiyarov,E. Ozensoy,SOxuptakeandreleasepropertiesofTiO2/Al2O3andBaO/TiO2/Al2O3
mixedoxide,systemsasNOxstoragematerials,Catal.Today184(2012)54–71.
[22]Soylu,Polat,Ozensoymanuscript,inpreparation.
[23]T.Ohsaka,F.Izumi,Y.Fujiki,RamanspectrumofanataseTiO2,J.Raman
Spec-trosc.7(1978)321–324.
[24]H.L.Ma,J.Y.Yang,Y.Dai,Y.B.Zhang,B.Lu,G.H.Ma,Ramanstudyofphase transformationofTiO2rutilesinglecrystalirradiatedbyinfraredfemtosecond
laser,Appl.Surf.Sci.253(2007)7497–7500.
[25]O.Carp,C.L.Huisman,A.Reller,Photoinducedreactivityoftitaniumdioxide, Prog.SolidStateChem.32(2004)33–177.
[26]D.A.H.Hanaor,C.C.Sorrell,Reviewoftheanatasetorutilephasetransformation, J.Mater.Sci.46(2011)855–874.