Investigation of absorber and heterojunction in the pure sulphide kesterite

(1)

HAL Id: hal-03032110

https://hal.archives-ouvertes.fr/hal-03032110

Submitted on 30 Nov 2020

HAL

is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire

HAL, est

destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

Charif Tamin, Denis Chaumont, Olivier Heintz, Remi Chassagnon, Aymeric Leray, Nicolas Geoffroy, Maxime Guerineau, Mohamed Adnane

To cite this version:

Charif Tamin, Denis Chaumont, Olivier Heintz, Remi Chassagnon, Aymeric Leray, et al.. Investigation

of absorber and heterojunction in the pure sulphide kesterite. Boletín de la Sociedad Española de

Cerámica y Vidrio, Elsevier, In press, �10.1016/j.bsecv.2020.05.004�. �hal-03032110�

(2)

Pleasecitethisarticleinpressas:C.Tamin,etal.,Investigationofabsorberandheterojunctioninthepuresulphidekesterite,Bol.Soc.Esp.

w w w . e l s e v i e r . e s / b s e c v

Original

Investigation of absorber and heterojunction in the pure sulphide kesterite

Charif Tamin

^a^,^b^,∗

, Denis Chaumont

^b

, Olivier Heintz

^b

, Remi Chassagnon

^b

, Aymeric Leray

^b

, Nicolas Geoffroy

^b

, Maxime Guerineau

^b

, Mohamed Adnane

^a

aLaboratoiredeMicroscopieElectroniqueetSciencesdesMatériaux(LMESM),DépartementdeTechnologiedesMatériaux,Facultéde physique,UniversitédesSciencesetdelaTechnologied’OranMohamedBoudiafUSTO-MB,ElM’naouar,BP1505,BirElDjir,31000 Oran,Algeria

bLaboratoireInterdisciplinaireCarnotdeBourgogne(ICB),UniversitédeBourgogneFranche-Comté,BP47870,21078Dijon,France

a r t i c l e i n f o

Articlehistory:

Received7January2020 Accepted25May2020 Availableonlinexxx

Keywords:

Kesterite CZTS Thinﬁlms Heterojunction Bandalignment

a bs t r a c t

Thispaperaimstostudythepropertiesoftheabsorberlayerandtheheterojunctionin kesteritesolarcells.TheCu₂ZnSnS₄(CZTS)thinfilmswerelayeredonaglasssubstrate fromacolloidalsolutionofmetalsaltsandthioureadissolvedinamixtureofwaterand ethanolanddepositedbyspincoatingtechnique.Thesampleswerethenheattreatedin afurnace,inthepresenceofsulphurpowderandunderanitrogengasflow.Theresults revealedtheformationofhomogeneouslayersofapurekesteritephaseofCZTScrystallites afterheattreatmentwithcorrectstoichiometryandoxidationstates.Theopticaltransmis- sionmeasurementsindicateanenergyband-gapof1.4eVandanabsorptioncoefficientof 10⁴cm⁻¹.TheseCZTSthinfilms,elaboratedbyspincoatingprocess,wereintegratedfor electronicpropertiesevaluationinaheterojunctioninthefollowingconfiguration:SnO2:F (FTO)andMolybdenumasbackcontact,CdSastamponandtheCZTSfilmasabsorberlayer.

ThebandalignmentattheCdS–CZTSheterojunctionindicatesacliff-likeconductionband offset(CBO)evenclosetobeaﬂatband.

Investigacióndelabsorbenteylaheterojunctionenelsulfuropuro kesterita

Palabrasclave:

Kesterita CZTS

Películasdelgadas Heterojunción Alineacióndebandas

r e su m e n

Estetrabajotienecomoobjetivoestudiarlaspropiedadesdelacapaabsorbenteylahetero- junciónenlascélulassolaresdekesterita.LasﬁnaspelículasdeCu2ZnSnS4(CZTS)fueron colocadasencapassobreunsustratodevidrioapartirdeunasolucióncoloidaldesales metálicasytioureadisueltasenunamezcladeaguayetanol,ydepositadasmediantela técnicaderevestimientoporcentrifugado.Acontinuaciónlasmuestrassetratarontérmi- camenteenunhorno,enpresenciadepolvodeazufreybajounﬂujodegasnitrógeno.Los

∗ Correspondingauthor.

E-mailaddresses:charif.tamin@univ-usto.dz,chariftamin@etu.u-bourgogne.fr(C.Tamin).

https://doi.org/10.1016/j.bsecv.2020.05.004

creativecommons.org/licenses/by-nc-nd/4.0/).

(3)

resultadosrevelaronlaformacióndecapashomogéneasdeunafasedekesteritapurade loscristalesdeCZTSdespuésdeltratamientotérmicoconlaestequiometríaylosestados deoxidacióncorrectos.Lasmedicionesdelatransmisiónópticaindicanunabrechaenla bandadeenergíade1,4eVyuncoeﬁcientedeabsorciónde10⁴cm⁻¹.Estaspelículasdelgadas deCZTS,elaboradasporelprocesoderevestimientodeespín,seintegraronparalaevalu- acióndelaspropiedadeselectrónicasenunaheterojunciónenlasiguienteconﬁguración:

SnO₂:F(FTO)ymolibdenocomocontactoposterior,CdScomotampónylapelículaCZTS comocapaabsorbente.LaalineacióndelabandaenlaheterojunciónCdS-CZTSindicaun desplazamientodelabandadeconducción(CBO)similaraldeunacantilado,inclusocerca deserunabandaplana.

Introduction

Inrecentyears,photovoltaicenergyhasbecomeapromising sourcefor meeting the energythat society demands. Cur- rently,silicon-basedsolarcellsarethemostcommercialised productonthephotovoltaicmarket.Theexpensiveandpollut- ingmanufacturingtechniquesofcrystallinesiliconsolarcells ledscientiststoﬁndanewgenerationofsolarcellsthatare moreeconomicalandenvironmentallyfriendly.

Thin-filmsolarcellsareapromisingalternativethatcould achieve this goal by reducing the amount of materials.In recent years, three thin-film solar cells technologies were developed:amorphoussilicon(a-Si),cadmiumtelluride(CdTe) and CIGS(CuInxGa_1−xSe2).TheCIGS technologyachieved a higher efficiency (22.9%)than the cadmium telluride (21%) andamorphoussilicon(10.2%)[1].However,(CdTe)technology usestwotoxicelement(CdandTe)andtheCIGStechnology involvesthe use ofrareelements suchasindium andgal- lium.Inaddition,indiumisnowwidelyusedintouchscreen technology[2]anditspricecouldsignificantlyincreaseinthe futureandaffecttheproductionoftheCIGStechnology.

Toavoidtheuse oftoxicandrare elements,newmate- rial kesterite (Cu₂ZnSnS₄ or Cu₂ZnSnSe₄ referred as CZTS and CZTSe) using copper, zinc, tin, and sulphur or sele- niumelements,hasattractedconsiderableinterestinrecent years and became promising for the development of eco- friendlysolarcells.Thiscompoundisap-typesemiconductor andisthereforeusedasanabsorberlayerinsolarcells.Its near-optimumdirectbandgapenergyof1.5eVanditslarge absorptioncoefﬁcient,greaterthan10⁴cm⁻¹,makeitoneof the mostpromisingmaterialsforphotovoltaicapplications [3,4].Althoughthisinnovativetechnologyhasgrownrapidlyin recentyears,itsperformanceisweakcomparedtoCIGS.The bestefﬁciencyofkesteritesolarcellsis12.6%forCu2ZnSn(S, Se)4and10%forCu2ZnSnS4[1,5].

Thedevelopmentofkesteritesolarcellsiscurrentlylimited bythelargeopencircuitvoltage(Voc)deﬁcit[6,7].Thishigh Voc deﬁcit(Eg/q−Voc)isstronglydependentonseveralfac- tors,inparticularrelatedtothepurityoftheCZTSabsorbent materialandthequalityandnatureoftheinterfacesbetween thedifferentlayersandnon-ideal bandalignmentthatlim- itschargeseparationattheabsorber/tamponheterojunction [8–10]. Thus, the CZTS layer must be pure (without para- siticphases)andstableduringthedepositionandtheheat

treatment processes (no loss of tin and sulphur by evap- oration or diffusion in the other layers of the cell). As well as, the heterojunction needs more investigations to understandtherecombinationlimitsattheheterojunction- interfaceabsorber/buffer(CZTS/CdS),whichisoneofthemain factorsinadditiontotheabsorbentlayerproblemslimitingthe efﬁciency.

Hence,obtainingapurephaseofCZTSrequiredoptimi- sation. The mechanism of formation of secondary phases anddecompositionofCZTSphaseduringheattreatmentis stillunderstudy[11].Severalauthorshavereportedthepres- ence of secondary phases in the kesterite structure, such as Cu_2−xS, SnS, SnS2, Cu2SnS3, whichare p-type orn-type semiconductorsorinsulators,thatcanformsecondarydiodes insideoftheCZTS[11–13].AreviewbyKumaetal.indicates that, under equilibriumconditions, acopper-rich layercan suppress thesecondaryphasesand assistintheformation ofthe purephase ofCZTS[12]. Kermadiet al.proved that copper-richCZTScouldremovesomesecondaryphasesbut wasunabletoremovethecoppersulphidephase[14].Ash- faqetal.recentlydemonstratedthattheuseofIn2O3:Sn(ITO) substratepromotesthegrowthofasingle-phase,namelyCZTS [15].However,mostofthepreviousstudiesdidnottakeinto accountthediffusionmechanismswhenusingITOtoobtain asingleandpureCZTSphase:crystallographicallythephase iskesteritebutchemically,indiumcanpartiallysubstitutetin andformaCu₂ZnSn1−xInxS4alloy(CZITS),whichmaychanges thepropertiesoftheCZTSandpromotetheappearanceofthe In2O3phase[16].

For theheterojunctionengineering,anunoptimisedhet- erointerfacebetweentheabsorberandthebufferlayercould result in a major limitation on device performance. How- ever,theheterojunctionengineeringhasachievedrelatively littleattentioninthekesteritecommunity.Recently,Yanetal.

(2018) obtained the efﬁciency record of kesterite sulphide solarcellsbyheterojunctionheattreatment[5].Thus,further focused experimental investigations at the absorber-buffer hetero-interfaceswillbeimpactfulinthekesteritedevelop- ment.

Inthispaper,wepresentthedepositionprocessandthe optical, structural,and chemicalcharacterisationsofsingle phase kesterite on a glass substrate by spin coating and sulphurisation process. These CZTS thin ﬁlms were used in astack oflayerstoproduce all solution heterojunction.

The electronic properties of the CdS/CZTS heterojunction

(4)

were collected experimentally by XPS including core-level, valance band positions and UV–visible bandgap measure- ments.

Experimental

CZTSthinﬁlmsweredepositedonsubstratesbyspincoating andheattreatedinafurnaceundercontrolledatmosphere.

Thecompleteheterojunctionwaselaboratedusingchemical bathdepositionandspincoatingtechniques.Variouscharac- terisationtechniqueswerecarriedouttostudytheproperties ofCZTSthinﬁlmsandheterojunctions.

CZTSthinsﬁlmspreparation

The colloidal solution was prepared by dissolving copper chloridedihydrate,zinc chloride,tinchloridedihydrateand thiourea(H₂NCSNH₂)withrespectiveconcentrationsof0.4M, 0.2M,0.2Mand1Minamixofwater(75%)andethanol(25%).

Somedropsofethanolaminewereaddedtothesolutionas stabiliser,wherethe pHofthesolution wasaround4. The solutionwasstirredfor6hatroomtemperaturetoobtaina clearyellow-colouredcolloidalsolution.

CZTSthinﬁlmswerelayeredonglasssubstratesfromthe colloidalwater-ethanolsolutionbyspincoatingprocessingat 4000rpmfor30s.Sampleswerethendrieddirectlyonahot plateheatedat250^◦Cfor10minunderairambient.Thispro- cesswasrepeated10times(foraﬁnalthicknessof1.2␮m).

Samplesweresubsequentlyheattreatedinafurnaceat520^◦C during30mininthepresenceofsulphur(0.5g)undernitrogen gasﬂow.

SamplesCTB01andCTB01Srespectivelycorrespondtothe CZTSthinﬁlmsbeforeandafterheattreatment.

Heterojunctionfabrication

Theheterojunctionprototypewasbuiltasfollows:

- Theﬁrst layerwas Molybdenum(Mo).ThisMolayerwas depositedusingane-beamevaporator(PlassysMEB400)on FTO(ﬂuorine-dopedtinoxide,SnO₂:F)glasssubstratespro- vided bySOLEMCompany.Thethicknessofthe Molayer was100nm;

- ThesecondlayerwastheCZTS(5layers)thinﬁlmdeposited using aspin coating techniqueasdescribed above (CZTS thinsﬁlmspreparationsection);

- ThethirdlayerwastheCdSbufferlayerdepositedbyChem- icalBathDepositionCBD(70^◦C,15min).Thechemicalbath compositionwas: 20mLofdeionisedwater,5mLof0.1M CdCl2,20mLof0.1Mthioureaand17.5mLof6.5MNH4OH.

ThepHofthechemicalbathduringthereactionwasabout 10.

ThisstackoflayerswillbereferredtoasHJ19-01.Finally, arapidannealing wasperformedinafurnaceat270^◦C for 10minundernitrogengasﬂowwithoutsulphur.Theas-built heterojunctionwillbelabelledasHJ19-01S.

Characterisations

TransmissionElectronMicroscopy(TEM)wasperformedona JEOLJEM-2100FmicroscopetocharacterisetheCZTSthinfilm microstructure.Forthepreparation,thinfilmswereremoved fromthesubstratebyscratchingthemwithascalpel.After- wards,thesamplesurfaceiswipedwithacarbon-coatedgold grid,resultinginpiecesofthefilmsstucktothecarbonfilm.

Thelocalmorphologyandthecrystallographywerestudied usingconventional,highresolutionmicroscopy(HRTEM)and SelectedAreaElectronDiffraction(SAED)modes.

Thesurface morphology(Plain view)and the multilayer stack(cross-section)oftheheterojunctionwereexaminedby ScanningElectronMicroscopy(SEM).Theobservationswere performedonaHitachiSU8230SEMequippedwithanEnergy DispersiveSpectrometer(ThermoScientiﬁcNSSSDD)allow- ingchemicalanalysesoftheﬁlms.

The atomic states, chemical compositions and valence band data were determined by X-ray Photoelectron Spec- troscopy (XPS)and performedusing aPHIVersaprobe5000 apparatus with monochromated Al K␣1 X-rays (energy of 1486.6eV,powerof50WandX-rayspotdiameterof200␮m).

Experiments were realised after sputtering of the CTB01, CTB01SandHJ19-01Ssamplesinordertoremovethemajor partofthethinlayerofoxidefromatmosphericcontamina- tion.Sputteringwasdonewithargonionsof500eVfor5min forthethinﬁlms(incidenceangleof45^◦).Fortheheterojunc- tion,the XPSdatawere collected aftereach 60s ofetching time. Inthis condition, the sputteringrate ofstandard sil- icondioxide isaround2nm/min.Adventitiouscarbonfrom atmosphericpollutionisusedasinternalstandardforenergy calibration(1s levelat284.8eV). Duringmeasurements, the residual pressure ofthe analysis chamberwas maintained below10⁻⁷Pa.SpectrawereprocessedwiththeCasasoftware packageandtheionisationcross-sectionsfromLandaumodel wereusedinordertoquantifythesemi-empiricalrelativesen- sitivityfactors.

The crystallographic structures of samples was investi- gatedbyX-RayDiffraction(XRD)andperformedwithaBruker D8DISCOVER(CuK␣radiationsource=1.5406 ˚A).

ThesizeofnanocrystalswasestimatedfromtheScherrer formula[17]:

Dhkl=0.89/Bhklcos

whereistheX-raywavelengthused,B_hkl thefullwidthat halfmaximumandtheBraggangleofthestudiedpeak.

TheRaman spectroscopy(Raman)wasperformedwitha custom-built epi-confocal microscope equippedwitha40× objective lens(0.6 NA,Nikon).Thesamplewas illuminated withalaserexcitationwavelengthof784nmandanintensity of0.6mWatthefocus.Raman spectrawere acquiredwith aspectrometer (equippedwitha gratingof1200lines/mm) associatedwithacooledCCDcamera(1024×256pixels).An expositiontimeof10swasusedandthecalibrationwasper- formedusinga(111)siliconwaferat520cm⁻¹.

The Ultraviolet–visible Spectroscopy (UV–vis) was per- formedwithaThermoSpectronicHeliosGammaspectropho- tometer,inthewavelengthrange190–1100nmwitha0.5nm

(5)

Fig.1–XRDdiffractogramsoftheCZTSthinﬁlm,(a)raw sampleCTB01,(b)heattreatedsampleCTB01S.

steptorecordthetransmittanceofthinﬁlmsandtocalculate theabsorptioncoefﬁcient,˛,fromthefollowingequation:

␣= 1 eln

₁

T

whereeisthe ﬁlmthickness andT thetransmittance[18].

NotedthattheﬁlmthicknesswasmeasuredusingaDektak 6MProﬁlometer(Veeco).

Theopticalbandgap valuesweredetermined usingthe Taucdiagram[16]where(˛h)²isplottedasafunctionofthe energyh.Thevalueofthegapisthatattheintersectionofthe extrapolationofthelinearpartofthecurvewiththeabscissa axis.

Resultsanddiscussion

Crystallographicandchemicalcharacterisation

TheXRD diffractogramoftheCZTSlayerbeforeheattreat- ment(CTB01)isshowedinFig.1a.Thethreemajorpeaksat 28.43^◦,47.27^◦and56.15^◦arerespectivelyassignedtothe(112), (220)and(312)crystallographicplanesofthekesteritetetrag- onalstructureofCZTS(PDFcard #04-015-0223).Inaddition totheCZTSmajorphase,theSnO2cassiteritephaseisalso detected,seepeaks(110),(101)and(211)atrespectiveangles 26.49^◦,33.77^◦ and51.72^◦(PDFcard#04-003-0649).ThisSnO₂ phaseemergesduringthedryingprocessinambientairon thehotplate.

Thesizeofthenanocrystals,derivedfromtheScherrerfor- mula,variesfrom5.5nmto8.8nmforeachpeak.

The XRD pattern of the annealed CZTS layer, CTB01S (Fig. 1b) shows narrow and intense diffraction peaks cor- responding to the pure kesterite CZTS structure without

secondary phase. The cassiterite peaks have disappeared duringannealing.Tindioxidecanturnintosolidtinmono- sulphide[19,20]duringannealingundersulphuratmosphere andtinmonosulphideisavolatilespecieabove500^◦C[21].

Thesizeofthenano-crystallitesvariedbetween18.7nm and 23.1nm depending on the chosen peak. The peak intensities highlight an isotropic growth of the crystallites. Logically, the crystallites size increases under heat treatment.

Theheattreatmentundersulphurandnitrogengasﬂow (oxygen free) prevents the formation of oxidised phases, increasesthecrystallisation,andhelpstheformationofthe CZTSkesteritephase.Insomecase,asitseemstobethecase here,italsoavoidsthedecompositionofthisstructureinto binary and ternarycompounds suchas Cu_2−xS, SnS2, ZnS, Cu2SnS3whichhasbeenobservedbyseveralauthorsinthe literature[11–13].

Inordertoconﬁrmtheabsenceofsecondarystructures(as ZnSandCu₂SnS₃,forwhichthethreemainsX-raydiffraction peaksare identicaltoCZTSones),theSAEDtechnique,the FastFourierTransformationofHRTEMimageandtheRaman spectroscopywerecomplementaryperformed.

Fig. 2(a, b) shows oneelectron diffraction patterns and theFastFourierTransformationofaHigh-Resolutionimage (HRTEM) of sample CTB01. The calculation of d spacing (Table1)fromtheSAEDconﬁrmtheresultsofXRD.Onlypure Cu2ZnSnS4 kesterite crystallites were found inthe studied samples.Fig.2cshowswheretheSAEDandthechemicalanal- ysis(EDS)wereperformed.Thechemicalcompositioninthis area(Fig.2d)showsthepresenceoffourelementsoftheCZTS compound:copper,zinc,tin,andsulphur.

Fig. 3a shows the Raman spectrum of the CTB01 sample. Thestrongest peak at335cm⁻¹ corresponds to the A mode ofthe kesterite structureofCZTS whichis theoreti- callyreportedbyGureletal.[22].Theweakpeakat363cm⁻¹ isassignedtotheE(LO)modeofkesteritestructureofCZTS [23]. Forthe calcinedCTB01Ssample(Fig.3b),thepeaksat 287cm⁻¹ and337cm⁻¹ areassignedtothetwoAmodesof thekesteritestructureaswellasthepeaksat366cm⁻¹ and 373cm⁻¹ whichcorresponds tothe E(LO) and B(LO)modes [24–26].Thenewintensepeakat287cm⁻¹,mayduetoabetter crystallinityofCTB01ScomparedtoCTB01.Asnootherpeaks are observed neitherthe ZnS (at 348cm⁻¹) nor theCu2−xS (at 476cm⁻¹), SnS2at(315cm⁻¹)and Cu2SnS3 (at352cm⁻¹) phasesarepresent[27–32].

ThedifferentoxidationstatesweremeasuredbyXPSbefore and afteretching for the CTB01 and CTB01S samples (see Fig.4).

XPSspectrainFig.4ashowthebindingenergyofCu2p_3/2 andCu2p_1/2corelevels,respectivelyataround932and952eV, fortherawandannealedCZTSsamples.Themeasuredbind- ingenergyfortheCu2p3/2 levelandtheabsenceofsatellite athighbindingenergyforbothpeakscorroboratethe Cu(I) oxidationstateofcopperinthewholeCZTSlayers[33–35].

Thebindingenergyvalueofabout1021.8eVforZn2p3/2 corelevelconﬁrmsthepresenceofZn(II)forbothCZTSlayers, beforeandaftercalcination(seeFig.4b)[35].

TheSn3d_5/2levelforCTB01andCTB01Ssamplesareshown inFig.4c.Energyofthislevelslightlyabove486eVconﬁrmsthe presenceofSn⁴⁺speciesinthetwosamples[35,36].