Detailed microstructure analysis of as-deposited and etched porous ZnO films

(1)

To link to this article : DOI:10.1016/j.apsusc.2015.03.097

URL :

http://dx.doi.org/10.1016/j.apsusc.2015.03.097

This is an author-deposited version published in:

http://oatao.univ-toulouse.fr/

Eprints ID: 13981

To cite this version:

Shang, Congcong and Thimont, Yohann and Barnabé, Antoine and Presmanes,

Lionel and Pasquet, Isabelle and Tailhades, Philippe Detailed microstructure

analysis of as-deposited and etched porous ZnO films. (2015) Applied Surface

Science, vol. 344. pp. 242-248. ISSN 0169-4332

O

pen

A

rchive

T

oulouse

A

rchive

O

uverte (

OATAO

)

OATAO is an open access repository that collects the work of Toulouse researchers

and makes it freely available over the web where possible.

Any correspondence concerning this service should be sent to the repository

administrator:

staff-oatao@listes.diff.inp-toulouse.fr

(2)

Detailed

microstructure

analysis

of

as-deposited

and

etched

porous

ZnO

films

Congcong

Shang,

Yohann

Thimont,

Antoine

Barnabé

∗

_,

_Lionel

_Presmanes,

Isabelle

Pasquet,

Philippe

Tailhades

UniversitédeToulouseIIIPaulSabatier,CIRIMAT–UMRCNRS5085,InstitutCarnot,118routedeNarbonne,31062ToulouseCedex9,France

Keywords: ZnO Thinfilms Surfacemorphology Opticalproperties TCO Chemicaletching

a

b

s

t

r

a

c

t

ZnOnanostructuredmaterialsinthinfilmformsareofparticularinterestforphotovoltaicor photo-catalysisprocessesbuttheysufferfromalackofsimplemethodsforoptimizingtheirmicrostructure. WehavedemonstratedthatmicroporousZnOthinfilmswithoptimizedintergrainaccessibilitycan

beproducebyradiofrequencymagnetronsputteringprocessandchemicaletchingwith2.75mMHCl

solutionfordifferentduration.Theas-depositedZnOthinfilmswerefirstcharacterizedintermsof structure,grainsize,intergrainspace,opencavitydepthandtotalthicknessofthefilmbyXRD,AFM, SEM,profilometryandopticalmeasurements.Aspecificattentionwasdedicatedtothedetermination ofthesurfaceenhancementfactor(SEF)byusingbasicgeometricalconsiderationsandimages treat-ments.Inaddition,theporousfractionanditsdistributioninthethicknesshavebeenestimatedthanks totheopticalsimulationoftheexperimentalUV–Visible–IRspectrumsusingtheBruggemandielectric modelandcrosssectionSEMimagesanalysisrespectively.Thisstudyshowedthatthemicrostructure oftheas-depositedfilmsconsistsofadenselayercoveredbyaporousupperlayerdevelopingaSEFof 12–13m2_m−2_._This_two_layers_architecture_is_not_modified_by_the_etching_process._The_etching_process

onlyaffectstheupperporouslayerinwhichtheoverallporosityandtheinter-grainspaceincreasewith theetchingduration.Columndiameterandtotalfilmthicknessdecreaseatthesametimewhenthefilms aresoakedintheHClbath.Themicroporousstructureobtainedaftertheetchingprocesscould gener-ateagreatinterestfortheinterfaceselectronicexchangesforsolarcells,photocatalysisandgassensors applications.

1. Introduction

Thewidebandgapsemiconductorsplayanessentialrolein var-iouselectronicand opto-electronicapplicationsbeingproduced today[1,2].Inthis classofTransparentConductingOxide(TCO) materials,nanostructuredzincoxide(ZnO)receiveda consider-ableattentionoverpastfewyearsbecauseofitslowcost,itshigh transparencyassociatedwithahighmobilityofchargecarriersand becauseitdoesnotgiverisetoenvironmentalconcerns[3–6].For photovoltaicapplicationssuchasthinfilmsinSi-solarcells,ZnO canbeusedinTCOfrontcontact[7]butalsotoprovideadditional opticalfunctionslikelightscattering[8–10].NanostructuredZnOis alsoofparticularinterestfororganicanddyesensitizedsolarcells [11,12]orphotocatalysisprocess[13]becauseitcanbe nanostruc-turedintonanowiresandrods,whichmakesitidealforinfiltration

∗Correspondingauthor.Tel.:+33561557751;fax:+33561556163. E-mailaddress:barnabe@chimie.ups-tlse.fr(A.Barnabé).

oftheabsorptionlayer.ZnOnanostructuredmaterialsimportance ismanifestedinmanypapersthathavebeenissuedforits synthe-sisusingnumeroussoftchemistrypathways,chemicalandphysical depositiontechnologies[14–16].Amongthem,sputteringtechnic isparticularlyusedtodepositZnOthinfilmmaterialsbecauseof itsrobustness,large-areaandhighratesmanufacturingcapabilities [3,6,7,10,17].

OtherfavorableaspectsofZnOincludeitschemicalreactivityto manyacidsleadingtomanyopportunitiesforwetchemicaletching. ManyetchingmethodsusingHClhavebeenreportedtodevelopa roughsurfaceandtheninducealightscatteringeffect,whichcan significantlyimprovelighttrapping(andthereforecurrent photo-generation)[8,9,18–21].HCletchingagentwasalsousedtostudy thecorrosionbehaviorofZnOthinfilmsinacidsolutions[22].

Inthisstudy,wereportanovelapproachtothefabricationof sub-micrometricporousZnOthinfilms.First,ZnOthinfilmsare depositedbyRadioFrequency(RF)cathodicmagnetronsputtering. Depositionparametersareadjustedtomaximizetheinter-granular porosityintheas-depositedZnOfilms.Thenanostructureofthese

(3)

films is then controlled using wet chemical etching in weakly concentrated HCl media.However, unlike the usually obtained micrometricpitbyHCletchingforlightscatteringproperty[8,9], wepresentanewconsequenceoftheetchingeffectatthe nano-metric scale, forthemodification ofaccessibilityof thesurface togasesand liquids.A detailedcharacterization,includingXRD, SEM,AFM,opticalanalysis,opticalsimulation,ofboth structure andmicrostructureofZnOthinfilmsbeforeandafteretching,is presented.Furthermore,twocalculationsmethodsareevaluated fortheestimationofthesurfaceenhancementfactor(SEF). 2. Materialandmethods

2.1. DepositionofZnOthinfilms

Thinfilmsweredeposited inanAlcatelCIT,type A450, con-ventionalplanarsystem,usingacommercialceramicZnOtarget (Neyco,purity99.99%).Thestudiedsamplesweredepositedonto glasssubstrateswith1.3×2.6cm2indimensioncarriedbywater cooledsubstrateholder.TheRFpowerwassetas50Wwith mag-netron,andargonwasusedasdepositiongas.

Manyauthorsreportedthatthemicrostructureofthethinfilms preparedbysputteringcanbetunedbychangingdeposition con-ditions[23,24]. Thishasbeen confirmedby themicrostructure studyofspineloxidethinfilms(Zn0.87Fe2.13O4[25]andCoMnFeO4

[26,27]) deposited using the same RF sputtering apparatus by changingthedepositiongaspressure(PAr)andsubstrateto

tar-getdistance(DST).WiththeincreaseofPArandDST,particleswith

shortmeanfreepathareproduced.Suchdepositionconditionsgive risetoparticleswithlowenergyandlargeincidenceangle,i.e.with lowmobilityofad-atomsandhighshadowingeffectforthegrowth ofthinfilms.Therefore,thehighestPAr=2PaandDST=8cm

val-uesallowedby thedepositionsystemwereusedinanattempt toprepareZnOthin filmswithporousmicrostructureand large accessiblesurface.

Undertheseconditions,thebombardmentofthetargetisvery intense especially on theerosion ring due tothe concentrated plasmacausedbymagnetron.Thechemicalreductionofthe tar-getsurface tendstoincreaseprogressivelyeach time plasmais switchedon.Topreventthisphenomenonandtoavoidhighoxygen deficiencyandheterogeneityinthedepositedthinfilms,oxygen (1%ofO2inAr)wassystematicallyintroducedinthepre-sputtering

gasduring20minpriortoeachdeposition(underpureAr). Adepositionratecloseto8nm/minwasdeterminedfromsetof referencesamplesdepositedinthe10–90mindurationrange. 2.2. WetchemicaletchingofZnOthinfilms

Asimpleexperimentalset-upwasusedforalltheetchingtests. Inapolymercontainerequippedwithatightcap,50mlwater solu-tionofHCl(2.75mM)waspreparedfirst.Duringeachetching,the wholepieceofsamplewasimmersedintothesolution.Oncethe etchingwasfinished,thesamplewasrinsedimmediatelyby de-ionizedwater,andrinsedforseveraltimes,usinganultrasound bathofde-ionizedwater.Finally,thesamplewasdriedusing com-pressednitrogengasflow.Alltheseexperimentsweredoneatroom temperature.FivedifferentsamplesweresubmittedtoHClbathfor durationsrangingfrom10to60min.

2.3. Characterizationtechniques

Thethinfilmthicknesswasmeasuredbybothmechanical pro-filometry (tprofilo)using aDektak 3030ST apparatusand optical

transmittance (toptical)according tothe interferometric method

[28].ThecrystallinestructureofthinfilmswasstudiedusingBruker AXS D4 Endeavor diffractometer, equippedwith a 1D LynxEye

detector. Thediffraction patterns wererecorded in –2 mode from20to100◦ _with_a_step _size_of _0.016◦ _and _a_counting_time

of0.13sforeachstep.CopperanodeisusedasX-raysource,with thecorrespondingKaradiation(Ka1=1.5405 ˚A,Ka2=1.5443 ˚A).

NickelfilterwasusedtoeliminatetheKbrayandfluorescence.

AtomicForceMicroscope(AFM)analysesofthethin filmswere performedusing a nanoscopeIIIDimension3000 VeecoDigital Instrument(tappingmode)withsupersharptipsforthestudyof surfacemorphology.AJEOLJEM6700FfieldemissiongunScanning ElectronMicroscope(SEM)wasalsoengagedinthesurface mor-phologyanalysis,aswellasthemorphologyofthecross-section. Opticalpropertiesofthethinfilms,notablythetransmittanceand reflectance,were analyzedby aBentham PVE300photovoltaic characterizationsystem,withintegratingsphereoperationmode. 2.4. ImageanalysistreatmentforSEFcalculation

The Surface Enhanced Factor (SEF) is thesurface on which moleculescanbeadsorbedforafilmperunitarea.SEFissimply definedastheratiooftherealtotheprojectedgeometricsurface.

InthemodeldevelopedforCoMnFeO4thinfilmsby

Oudrhiri-Hassanietal.[27],themicrostructureofathinsputter-deposited film is simply described by cylindrical grainshape with hemi-sphericaldomes.Nearthesubstrate,theinterfaceisconsideredas denselayeraccordingtothemodelofCziganyetal.[29].A sim-pleschemeofthismodelisrepresentedinFig.1a.Inthismodel, SEFrepresentsthesumofthenormalizedareaofthebottomofthe intersticesbetweenthecolumns(SCavity),thelateralsurfaceareaof

thecolumns(SWall)andthesurfaceareaofthedomes(SDome).SEF

iscalculatedbygeometricalconsiderationusingthethicknessof theporouslayer(tPLentitledL3intheoriginalpaper)andthemean

diameterofthecolumns(d)estimatedfromSEMmicrographsin cross-sectionandplanarviews,accordingtoEq.(1):

SEF₌ (4tPL+d)+2d √

3

2d√3 (1)

Inthiswork,theGwyddionsoftware[30]wasusedtotreatand analyzeSEMsurfaceviewoftheZnOthinfilminordertomore finelyestimatetheSEFthroughabetterestimationoftheareaof thebottomoftheintersticesbetweenthecolumnsandthelateral surfaceareaofthecolumns.Afteranormalizationofthegrayscale, amathematicalsegmentationtreatmentwasappliedandfollowed byaborderkilloperationtoobtainanequivalentthresholdmask oftheimage(Fig.1b).Thresholdhasbeenfixedat64%ofthe maxi-mumofthenormalizedgreyscale.Thisvaluewasarbitrarilysetin ordertorevealamaximumofinter-granularareas(areaofthe bot-tomoftheintersticesbetweenthecolumns)withoutaddingwrong grainboundariesinsidethegrains.Aftertheinversionofthemask, markedpatternshavebeenusedforthelocalization,quotationand thesizecharacterizationoftheopencavities(astheperimeterand theprojectedsurfacearea)(Fig.1c).Thetotalperimeter(Pt)has

beendeducedfromthesumoftheperimeterofeachmarkedopen cavities.Herewesupposedthattheunmarkedzonesrepresentedin blackinFig.1cconsistofgrainsonly.Thegrainshavealsobeen mod-elledaspseudo-cylinderswithhemisphericaldomesatthesurface ofthefilm.Asforthepreviousmodel,SEFisthesumofSCavity,SWall

andSDome.Theareaofthebottomoftheintersticesbetweenthe

columns,i.e.thesurfaceofthewholeprojectedopencavitiesis determineddirectlybythethresholdoperation(SCavity).Thelateral

surfaceareaofthecolumns(SWall),i.e.theinternalsurfacewallof

theopencavitiescanbecalculatedbymultiplyingthetotal perime-teroftheopencavitiesPtbythethicknessoftheporouslayertPL.

Therelationbetweenthedomesurface(SDome)andthegrain

pro-jectedsurface(SProjected Total−SCavity)ismathematicallyequalto2:

thedomesurfaceisequalto2pr2_,_i.e._half_of_the_surface_area_of_a

(4)

Fig.1.(a)SimplifiedmicrostructureoftheasdepositedZnOthinfilmusedfortheestimationofSurfaceEnhancementFactor(SEF)whereasdistheaveragecolumndiameter andtRL,tPL,tDLthethicknessesoftheRoughnessLayer,PorousLayerandDenseLayerrespectively.Thetotalthicknessofthefilmist=tRL+tPL+tDL.(b)calculatedmask(in

red)superposedtotheoriginalimageSEMmicrographobtainedafterthethresholdoperationand(c)extractedmaskusedforthecalculationofthetotalperimeterofopen cavities(Pt).

theareaofadiskofthesameradiusr.Asaresult,fromtPLestimated

fromtheSEMmicrograph,andPtandSCavityobtainedbytheimage

treatment,SEFcanbecalculatedaccordingtoEq.(2):

SEF= SCavity+tPL·Pt_S+2(SProjectedTotal−SCavity)

ProjectedTotal

(2) 3. Resultsanddiscussion

3.1. As-depositedZnOthinfilms

XRDmeasurementindicatesthattheas-depositedZnOfilmsare polycrystallinewiththehexagonalwurtzitestructure.Ahighc-axis orientationperpendiculartothesubstrateisobservedincontrast tothetargetusedfordeposition.AnexampleoftheXRDpatternof thethinfilmisshowninFig.2.Thelatticeparametersa=3.29(1) ˚A andc=5.183(7) ˚Aforthesethinfilmswerededucedfromthe(002) and(103)peaksrespectivelylocatedat34.586◦_and_62.875◦_in_2.

Theyareveryclosetothetheoreticalvaluesofa=3.2498 ˚Aand c=5.2066 ˚Arespectively(JCPDScardNo.36-1451).

Fig.2. X-Raydiffractionpatternsof(a)ZnOpowderand(b)600nm-thickZnO as-depositedthinfilm.

Themicrostructureoftheas-depositedthinfilmswasstudied bybothAFMandSEM(insurfaceandcross-sectionviews).A rel-ativelyflatsurfacewasobtainedforthesethinfilmswithaRMS roughnessof13nmmeasuredbyAFM(Fig.3a).FromtheSEM sur-faceview(Fig.3b),anaverageparticlesizeofapproximately110nm was measured and narrow inter-grain spaces were evidenced. Thecross-sectional SEM image(Fig.3c)showed columnar-type microstructurewithinvertedconical-shaped.Thesegrainsgrow perpendiculartotheglasssubstrateandareconfinedinabi-layer configurationaspreviouslydescribedbyC.Besleagaetal.[31].From thesubstratetothesurface,aCompactLayer(CL)withsmallgrain sizefirstappearswithanaveragethicknessof200nmfollowedby a400nmthickPorousLayer(PL)withbetterdefinedlargergrains. Theupperporouslayerisduetotheshadoweffectoftheinitially developedcolumns.Thisbi-layerconfigurationcanusuallybeseen forsuchrelativelythicksampleduetotheevolutionofwavy inter-faceaspreviouslyreportedbyCziganyetal.[29]andsimulatedby mathematicalmodels.Forlowerthicknesses,thewavysurfaceof thethinfilmthroughoutthegrowingdirectionhasnosignificant evolution:theshadoweffectdoesnotoccurandthefilmformedis compact.

TheSEF,whichissimplydefinedastheratiooftherealtothe projectedgeometricsurface,isaparameterthatcanbeusedfor porousthinfilmevaluation.FortheSEFdetermination,BET mea-surementonthinfilmsisthenveryefficientbutrequiresalarge quantityofdepositswhichmakesittimeconsumingandnotso easy to use. An easierapproach is given by the determination ofSEFusingasimplegeometricalmodelinwhichcharacteristic parametersaresuppliedbymicrostructuralobservations.Indeed wedemonstrated previouslythat SEFcanbewellestimatedby usingthethicknessoftheporouslayerandthemeangrainsize [27].Thusinourcurrentwork,thismethodbasedona mathemat-icalmodelhasbeenusedtoestimateSEFvalue(Eq.(1)).Fromthe SEMobservationsofthet=600nmthickZnOfilm(Fig.3bandc Fig.3),thedenselayerthicknesshasbeenapproximatedto1/3of thetotalthickness(tDL=200nm)andthecolumnaverage

diame-terestimatedtod=110nm.Withathicknessoftheroughnesslayer estimatedtod/2(tRL=55nm),thethicknessoftheporouslayerwas

(5)

Fig.3. (a)AFMand(b)SEMsurfaceviewsand(c)SEMtiltedcrosssectionviewofas-deposited600nmthickZnOfilm.

deduced(tPL=t−tRL−tDL=345nm).Finallythewhole

microstruc-tureofthefilmcanbemodeledandtheSEFcanthenbeestimated tobe13.3m2_m−2_using_this_model.

UsingtheSEMsurface view(Fig.3b),atotal perimeterPt of

30.2mmandasurfaceofthewholeprojectedopencavitiesS_Cavity of0.126mm2havebeendeterminedbyimagetreatmentrealized onatotalprojectedsurfaceSProjectedTotalof1mm2.Withidentical

thicknessoftheporouslayer(tPL=345nm),theSEFisthen

esti-matedtobe12.3m2_m−2_by_application_of_Eq.₍₂₎_._This_is_in_good

accordancewiththeresultobtainedpreviouslyandthenvalidates thesimplemicrostructuralmodelrepresentedinFig.1a.

Finally themicrostructureof600nmthickas-deposited ZnO thin films can be summarized as follow. A dense under layer accountingforabout1/3ofthetotalthickness,i.e.about200nm, andaporousupperlayer(∼350nmthick)surroundedbya rough-ness layer corresponding to the surface waviness of the film (∼50nmthick)whichbothoccupiestheremainingabout2/3ofthe totalthickness.Fromthisgeometricalconsiderationandbyusing asimplegeometricalmodel,thesurfaceenhancementfactor(SEF) hasbeenestimatedtobeworthcloseto12–13m2_m−2_._Even_though

thedepositionconditionwasadjustedtoproduceahighlyporous microstructure(highargonpressureandlargesubstrate-to-target distance),theas-depositedZnOthinfilmdevelopmuchless acces-siblesurfacecomparedtospinelCoMnFeO4thinfilms[27]prepared

undersimilardepositionconditions.Thisresultismainlyexplained bythelargercolumndiametersandhigherdepositionrateswhich areobtainedinthezincoxidelayer.Itcanbeduetothesimple chemicalcompositionofZnOthatallowsaneasiercrystallization

thanmorecomplexoxidelikethecobalt-manganesespinelferrite composedofthreecationsandtwocationicsublattices.

3.2. HCl-etchedZnOthinfilms

TheorientedmicrostructureoftheasdepositedZnOthinfilms allowsgeneratingattractiveSEFvaluesbutwithlimitedinter-grain spacefor gassensorand dyesensitized solarcells applications. Chemical etching with acid solution can generate anisotropic effects[21,32,33]andthenincreasestheinter-grainspace with-outnotablemicrostructural modification.Thisanisotropiceffect canbemodulated bytheuseof variousetchingmethod,agent andconditions.Forinstance,adominantverticaletchingprocess hasbeenobtainedonrf-sputteredZnOfilmetched electrochemi-callyusing10wt%KOHsolution[34,35].Inthiswork,post-chemical etchingwasthenperformedinasecondstepinordertoenlargethe inter-grainspaceandincreasetheaccessibilityofthethinfilm sur-facearea.Aserieof600nmthickZnOthinfilmswereetchedin 2.75mMHClwatersolutionfordifferentduration.Thethinfilms characteristicswerethenevaluatedaccordingtoetchingduration. The optical characteristics (total transmittance TT, total reflectance TR and deduced total absorbance TA) of the as-depositedandetchedZnOthinfilmsforvariousetchingtimesare plottedinFig.4.

Theas-depositedandetchedthinfilmthicknesseswere mea-sured by both mechanical profilometry (tprofilo) and optical

transmittance(toptical).ResultsarereportedintheTable1within

(6)

Fig.4.OpticalspectraofZnOthinfilmsasafunctionoftheetchingdurationt.(a) TotalTransmittanceTT.InsetshowstheexperimentalandsimulatedTTspectraof as-depositedfilm.(b)TotalReflectanceTR.(c)TotalabsorbanceTA.Insetshowsthe Taucplotforestimationofdirectallowedopticalgap.

(400<<800nm)extractedfromTTmeasurementsrepresented inFig.4a.Hazeparameterdefinedastheratioofdiffusedtototal transmittedlightisalsoreportedassoonasdiffusedtransmitted lightismeasurable.Accuraciesestimatedbyreproducibilityand

Table1

CharacteristicvaluesfromdifferentanalysesofHCletchedsamplesfordifferent etchingduration.

Etchingduration(min) 0 10 20 30 45 60

tprofilo(nm) 610 558 491 428 365 –

toptical(nm) 620 547 506 420 401 –

TT400–800(%) 83 84 85 85 88 88

Haze(%) – – – 2.7 2.4 2.4

repeatabilitytestsarelessthan3%forallthereportedvaluesinthe Table1.

Acleardecreaseofthicknesses(tprofiloandtoptical)wasobserved

withtheincreaseofetchingdurationforthefirst45minofetching. Forlongeretchingduration(60min),theestimationofthicknesses frombothprofilometryandopticalmeasurementsbecamevery dif-ficultandleadstolimitedaccuracydue tothelessdefinedstep andinterferencesontransmittancespectrarespectively.An aver-ageverticaletchingrateof5.6nmmin−1 _is_deduced_for_etching

durationsmallerthan45min.

Theaveragetransmittanceinvisiblerangeandthetotal trans-mittanceinabsorptionrangebothincreasedwiththeincreaseof etchingduration,whichisinaccordancewiththedecreaseofthin filmthickness.Theabsorptionedgeisunaffectedbytheetching treatment(Eg=3.24eV)asshownbytheTaucplot[36]intheinsert oftheFig.4c.For30minofetchingdurationandmore,the diff-usedtransmittedlightstartstobemeasurableand fewpercent (2.4–2.7%)couldbecalculatedforthehazefactor(Table1)dueto theincreaseofsurfaceroughnessandlightscattering.

TTandTRspectraoftheas-depositedfilmwerealsosimulated withtheSCOUTsoftware[37]usingaDrudemodelcoupledwith aninterbandKimoscillator[38].Duringsimulation,thetotal thick-nessofthethinfilmswasdividedintotwolayers(atopporous layerstackedonadenselayer)accordingtotheprevious micro-structuralcharacterizations(Fig.3c)andmodelization(Fig.1a).It canbealsonoticedthattheestimationofthevolumefractionofthe ZnOtoplayerwasobtainedbytheBruggemandielectricmodel[39]. Byrefinementofdifferentparameters,includingthetopporous layervolumefraction(PL%Vol₎_and_thickness_(t

PL)andtheunder

(7)

denselayersthickness(tDL),thesimulatedspectrawerefittedto

the experimentaldata. The optimalfit of the experimental as-depositedtotaltransmittancespectrum(insertFig.4a)isobtained withPL%Vol₌_93%,_t

PL=458nmandtDL=130nm.Theseparameters

areingoodagreementwiththepreviousSEMcharacterization. Toestimatetheevolutionoftheporousfractionwiththeetching duration,theevolutionofthemicrostructureoftheZnOfilmsafter 10,30,45and60minetchingdurationwastheninvestigatedby SEManalysis.TheresultsarepresentedinFig.5.Thecorresponding imageoftheas-depositedsamplewaspreviouslyshowninFig.3. Uponetchingduration,theSEMmicrographsincross-sectionviews (Fig.3candFig.5)showedareductionofthethicknessofthetop porouslayer(tPL)associatedwithadecreaseofitsvolumefraction,

i.e.andincreaseofitsporosity.However,theunderdenselayer thickness(tDL)variationhasnocleartrendatdifferentetching

dura-tion.Accordingtothesurfaceviews(Fig.3bandFig.5)andwiththe increaseofetchingduration,thegrainsizedecreasessignificantly whereas theinter-grainspaceincreases significantly.Therefore, consideringthethicknessdiminutionshowedpreviouslywiththe increaseofetchingduration,HClactedbothonthetopofthegrains andonthesidesofthecolumnsoftheporousupperlayerwhichare exposedtotheetchingsolutionthroughtheinter-columnarpores previouslyhighlighted.Fromthecross-sectionalviews(Fig.5),it appearsthatthebi-layermicrostructureobservedinas-deposited thinfilm(Fig.3b)hasbeenpreserved.After10minofetching,the grainsofporouslayershowedlesscompactmicrostructure com-paredtotheas-depositedthinfilms.Theinter-grainspaceslightly increasedfortheentireporouslayer.Whentheetchingduration increasedto30min,thegrainsizeappearstobemuchsmallerthan thatofas-deposited and10minetchedthin films.However,for 60minetching,theporouslayerwasheavilyattacked.Inthe mean-time,thedenselayerappearstobethesameforthethreesamples. With10minofetching,theetchingeffectalreadyreachedthe inter-faceofthetwolayers,howeverwith50minmore,thedenselayer remainsroughlythesame.Thedifferentmicrostructuresobserved inupperandunderlayersleadthentoetchingrateswhichdiffers stronglyforthecolumnlocatedintheporouslayerandthedense layerplacedinthedeeperpartoftheZnOfilmandthusless acces-sible.TheschematicmicrostructureofaZnOthinfilmcanbeseen intheFig.6aatinitialandfinalstageofthepartialetchingprocess. Theaverageinter-grainspacewasestimatedusingthesame SEMimagestreatmentusedfortheas-depositedanalysis.These valuesrepresenttheaveragesizeofthemarkedzones(Fig.6b). Theinter-grainspace(dIGS)distributionsatdifferentetching

dura-tionwerealsoinvestigatedusingthesurfacetreatmentanalysis whichallowsdeterminingthedistributionofthemaximumradius oftheinscribedcircleinthemarkedzoneofthemaskandhasbeen reportedintheinsertofFig.6b.Fromthisdistribution,the per-centageofdIGS expressedincumulativepast(%)versusthesize

wasusedtodeterminethemedianvalue(d50%=dIGS)andtheerror

bars(d45%andd55%).Fortheas-depositedthinfilms,thedIGS

dis-tributionisquitenarrowwithamaximumoccurrencenear30nm. Withtheincrease ofetchingduration to60min,thedIGS

distri-butionbecame broaderand themaximumis slightly shiftedto highervalues (dIGS≈60nmfor60min).Theseresultsconfirmed againtheincreaseinaverageinter-grainspace,andfurthermore, thebroadeningofdistributionsuggeststhenon-homogenous etch-ingindepththroughoutthewholethinfilmlayer.

Duringtheetching,boththetopporouslayersincludingthe roughnesslayer(tPL+tRL)andthebottomdenselayer(tDL)decrease

(Fig.6c). However,accordingtotheslightestaccessibilityofthe denseunderlayer,theverticaletchingrate(from thesurfaceof thefilmtothefilm-substrateinterface)oftheporouslayerismore important.TheevolutionoftheSEFcalculatedbySEMimage anal-ysisasafunctionoftheetchingdurationisreportedintheFig.6d. TheSurfaceEnhancementFactor decreasesfrom12.3m2_m−2_to

Fig.6.(a)SchematicrepresentationofthemicrostructureofaZnOthinfilmas depositedandafterpartialetchinginHClsolution.(b)Averageintergrainspace (dIGS)anditsdistributiondeterminedfromthemaximumradiusoftheinscribed

circleinthemarkedzones(insert),(c)thicknessofthelayers(totalthicknesstand denselayerthicknesstDL)and(d)theSEFfordifferentetchingdurations.Thedotted

curvesareplacedtoguideviewer.

7.2m2 −_m2_with_the_etching_duration_due_to_both_the_decrease_in

thicknessofthetopporouslayerandthedecreaseofthemean col-umndiameter.Finely,theincreaseofinter-grainspacepromotes theaccessibilityofmoleculeswithinthefilm,whichcouldreveal ahighpotentialforexampleforapplicationssuchasgassensing (improvementofsensibilityandresponsetime)orGräetzelcells (absorbingdyeinsidetheporosity).

4. Conclusion

Highlyc-orientedhexagonalwurtziteZnOthinfilmshavebeen preparedbyradio-frequencycathodicmagnetronsputteringunder an argonpressure of 2Pa and a target-to-substrate distanceof 8cm. The as-deposited films are made of a denseunder layer accountingfor1/3ofthefilm,surroundedbyaporous microstruc-turefortheremainingpart.Asurfaceenhancementfactor(SEF) closeto12–13m2_m−2 _has_been_determined_by_two _calculation

methods.ThinfilmswereetchedbylowconcentrationHCl solu-tionthenanalyzedbyprofilometry,BET,SEM,AFMandUV–Vis–IR spectrophotometer. By applying SEM imagetreatment method, mathematicalmodelandopticalsimulations,thethicknessesofthe

(8)

denseandporouspartsofthefilm,andtheinter-grainspacewere studiedindetail.Theetchedfilmsshowasimilarmicrostructure tothatoftheasdepositedthinfilm,withaporouslayerstacked ondenselayer.Thetopporouslayerthicknessdecreaseswiththe etchingdurationinoppositionwiththethicknessofthedenselayer remainingconstant. Globally,a significantdecrease ofthetotal thinfilmthicknessandtheenlargementoftheinter-grainspace throughouttheentireporouslayerwereobservedwiththeincrease inHCletchingduration.Accordingly,HCletchingperformedboth along(topofthecolumns)andperpendicular(sideofthecolumns) toc-axiswithaloweretchingrateatgrainboundarieswhich gen-eratedanincrease oftheporosity of theporousupper layerin nanometerscale.

Finally, even with partial loss of thickness, the HCl etching canstillbeveryinterestingforincreasingtheaccessibilityforgas andmolecules withinZnO thin films.Themicrostructural char-acteristicsof theetchedZnOthin filmsare veryinteresting for theelectronicexchange as,for instance,in solarcells, gas sen-sorsandphotocatalystsapplications.Moreover,thewetetching atnanometerscaleofafilmdepositedunderoptimized sputter-ingconditionswithawell-controlledporousmicrostructurecan beappliedtoanyotherthinfilmcomposition.

References

[1]P.Barquinha,R.Martins,L.Pereira,E.Fortunato,TransparentOxideElectronics: FromMaterialstoDevices,Wiley,2012,978-0-470-68373-6.

[2]D.S.Ginley,C.Bright,Transparentandconductingoxides,MRSBull.25(8) (2000)15–21.

[3]T.Minami,T.Yamamoto,T.Miyata,Highlytransparentandconductiverare earth-dopedZnOthinfilmspreparedbymagnetronsputtering,ThinSolidFilms 366(2000)63–68.

[4]Ü.Özgür,D.Hofstetter,H.Morkoc¸,ZnOdevicesandapplications:areviewof currentstatusandfutureprospects,Proc.IEEE98(2010)1255–1268.

[5]Ü.Özgür,Y.I.Alivov,C.Liu,A.Teke,M.A.Reshchikov,S.Do˘gan,V.Avrutin,S.J. Cho,H.Morkoc¸,AcomprehensivereviewofZnOmaterialsanddevices,J.Appl. Phys.98(2005),041301-1–103.

[6]K.Ellmer,A.Klein,B.Rech,TransparentConductiveZincOxide:Basicsand ApplicationsinThinFilmSolarCells,Springer,2008,ISBN978-3-540-73612-7.

[7]E.Fortunato,D.Ginley,H.Hosono,D.Paine,Transparentconductingoxidesfor photovoltaics,Mater.Res.Soc.Bull.32(2007)242–247.

[8]K.C.Lai,F.J.Tsai,J.H.Wang,C.H.Yeh,M.P.Houng,Wet-etchtexturingofZnO:Ga backlayeronsuperstrate-typemicrocrystallinesiliconsolarcells,Sol.Energy Mater.Sol.Cells95(2011)1583–1586.

[9]J.Hüpkes,H.Zhu,J.I.Owen,G.Jost,E.Bunte,Instabilitiesinreactivesputtering ofZnO:Alandreliabletexture-etchingsolutionforlighttrappinginsiliconthin filmsolarcells,ThinSolidFilms520(2012)1913–1917.

[10]M.Berginski,J.Huepkes,W.Reetz,B.Rech,M.Wuttig,Recentdevelopmenton surface-texturedZnO:Alfilmspreparedbysputteringforthin-filmsolarcell application,ThinSolidFilms516(2008)5836–5841.

[11]L.Bahadur,S.Kushwaha,HighlyefficientnanocrystallineZnOthinfilms pre-paredbyanovelmethodandtheirapplicationindye-sensitizedsolarcells, Appl.Phys.A109(2012)655–663.

[12]G.Jimenez-Cadena,E.Comini,M.Ferroni,A.Vomiero,G.Sberveglieri,Synthesis ofdifferentZnOnanostructuresbymodifiedPVDprocessandpotentialusefor dye-sensitizedsolarcells,Mater.Chem.Phys.124(2010)694–698.

[13]Y. Sivalingam, E.Martinelli, A. Catini, G. Magna, G. Pomarico, F. Basoli, R. Paolesse, C. Di Natale, Gas-sensitive photoconductivity of porphyrin-functionalizedZnOnanorods,J.Phys.Chem.C116(2012)9151–9157.

[14]Z.Fan,J.G.Lu,Zincoxidenanostructures:synthesisandproperties,J.Nanosci. Nanotechnol.5(10)(2005)1561–1573.

[15]L.Vayssieres,Anaqueoussolutionapproachtoadvancedmetaloxidearrayson substrates,Appl.Phys.A89(2007)1–8.

[16]C.Shang,A.Barnabé,Structuralstudyandphasetransitioninvestigationina simplesynthesisofporousarchitected-ZnOnanopowder,Mater.Charact.86 (2013)206–211.

[17]M.Lalanne,J.Soon,A.Barnabé,L.Presmanes,I.Pasquet,P.Tailhades, Prepara-tionandcharacterizationofthedefect–conductivityrelationshipofGa-doped ZnOthinfilmsdepositedbynonreactiveradio-frequency–magnetron sputter-ing,J.Mater.Res.25(2010)2407–2414.

[18]S.H.Lee,B.H.Seo,J.H.Seo,Wetetchingofgalliumindiumzincoxide(GIZO) semiconductorforthinfilmtransistorapplication,J.KoreanPhys.Soc.53(5) (2008)2603–2606.

[19]H.Zhu,J.Hüpkes,E.Bunte,J.Owen,S.M.Huang,Noveletchingmethodonhigh rateZnO:Althinfilmsreactivelysputteredfromdualtubemetallictargetsfor silicon-basedsolarcells,Sol.EnergyMater.Sol.Cells95(2011)964–968.

[20]X.Liao,X.Zhang,Zincoxidenanostructuresandtheircore-shellluminescence properties,J.Phys.Chem.C111(2007)9081–9085.

[21]X.G.Han,Y.Q.Jiang,S.F.Xie,Q.Kuang,X.Zhou,D.P.Cai,Z.X.Xie,L.S.Zheng, ControlofthesurfaceofZnOnanostructuresbyselectivewet-chemicaletching, J.Phys.Chem.C114(2010)10114–10118.

[22]B.H.Seo,S.H.Lee,J.H.Seo,J.H.Jeon,H.Choe,Studyonthewetetchbehavior ofazinc-oxidesemiconductorinacidsolutions,J.KoreanPhys.Soc.53(2008) 402–405.

[23]J.A.Thornton,Themicrostructureofsputter-depositedcoatings,J.Vac.Sci. Technol.A4(1986)3059–3065.

[24]A.Anders,Astructurezonediagramincludingplasma-baseddepositionand ionetching,ThinSolidFilms518(2010)4087–4090.

[25]S. Capdeville, P. Alphonse, C. Bonningue, L. Presmanes, Ph. Tailhades, MicrostructureandelectricalpropertiesofsputterdepositedZn0.87Fe2.13O4

thinlayers,J.Appl.Phys.96(2004)6142–6146.

[26]I.Sandu,L.Presmanes,P.Alphonse,P.Tailhades,Nanostructuredcobalt man-ganeseferritethinfilmsforgassensorapplication,ThinsolidFilms495(2006) 130–133.

[27]F.Oudrhiri-Hassani,L.Presmanes,A.Barnabé,P.Tailhades,Microstructure, porosityandroughnessofRFsputteredoxidethinfilms:Characterizationand modelization,Appl.Surf.Sci.254(2008)5796–5802.

[28]R.Swanepoel,Determinationofthethicknessandopticalconstantsof amor-phoussilicon,J.Phys.E16(1983)1214–1222.

[29]Z.Czigany,G.Radnoczi,Columnargrowthstructureandevolutionofwavy interface morphologyinamorphousandpolycrystallinemultilayeredthin films,ThinSolidFilms347(1999)133–145.

[30]D.NecasandP.Klapetek,http://www.gwyddion.net/Gwyddionsoftware. [31]C.Besleaga,G.E.Stan,A.C.Galca,L.Ion,S.Antohe,DoublelayerstructureofZnO

thinfilmsdepositedbyRF-magnetronsputteringonglasssubstrate,Appl.Surf. Sci.258(2012)8819–8824.

[32]J. Elias, R. Tena-Zaera, G.Y. Wang, C. Lévy-Clément, Conversion of ZnO nanowiresintonanotubeswithtailoreddimensions,Chem.Mater.20(2008) 6633–6637.

[33]H.Q.Wang,G.H.Li,L.C.Jia,G.Z.Wang,L.Li,Generalinsituchemicaletching synthesisofZnOnanotipsarray,Appl.Phys.Lett.93(2008),153110-1–3.

[34]C.C.Ching,P.K.Ooi,S.S.Ng,M.A.Ahmad,Z.Hassan,H.AbuHassan,M.J.Abdullah, FabricationofporousZnOviaelectrochemicaletchingusing10wt%potassium hydroxidesolution,Mater.Sci.Semicond.Process.16(2013)70–76.

[35]C.C.Ching,S.Lee,P.K.Ooi,S.S.Ng,Z.Hassan,H.AbuHassan,M.Abdullah,Optical andstructuralpropertiesofporouszincoxidefabricatedviaelectrochemical etchingmethod,Mater.Sci.Eng.B178(2013)956–959.

[36]J.Tauc,R.Grigorovici,A.Vancu,Opticalpropertiesandelectronicstructureof amorphousgermanium.,Phys.StatusSolidi15(1966)627.

[37]W.Theiss,http://www.wtheiss.com/CODE-Athinfilmanalysisanddesign software.

[38]C.C.Kim,J.W.Garland,H.Abad,P.M.Raccah,Modelingtheoptical dielec-tricfunctionofsemiconductor:extensionofthecriticalpointparabolicband approximation,Phys.Rev.BCondens.Matter45(1992)11749–11767.

[39]A.A.G.Bruggeman,BerechnungverschiednerphysikalischerKonstantenvon heterogenenSystemen,AnnalenderPhysik24(1935)636–679.