HAL Id: cea-01365822
https://hal-cea.archives-ouvertes.fr/cea-01365822
Submitted on 13 Sep 2016
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.
Determination of the cation site distribution of the
spinel in multiferroic CoFe2O4 / BaTiO3 layers by
X-ray photoelectron spectroscopy
T. Aghavnian, J.-B. Moussy, D. Stanescu, Rachid Belkhou, N Jedrecy, Hélène
Magnan, P. Ohresser, Arrio M.-A., Ph. Sainctavit, Antoine Barbier
To cite this version:
T. Aghavnian, J.-B. Moussy, D. Stanescu, Rachid Belkhou, N Jedrecy, et al.. Determination of the
cation site distribution of the spinel in multiferroic CoFe2O4 / BaTiO3 layers by X-ray photoelectron
spectroscopy . Journal of Electron Spectroscopy and Related Phenomena, Elsevier, 2015, 202,
pp.16-21. �10.1016/j.elspec.2015.02.006�. �cea-01365822�
ContentslistsavailableatScienceDirect
Journal
of
Electron
Spectroscopy
and
Related
Phenomena
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 / e l s p e c
Determination
of
the
cation
site
distribution
of
the
spinel
in
multiferroic
CoFe
2
O
4
/BaTiO
3
layers
by
X-ray
photoelectron
spectroscopy
T.
Aghavnian
a,b,
J.-B.
Moussy
a,
D.
Stanescu
a,
R.
Belkhou
b,
N.
Jedrecy
c,
H.
Magnan
a,
P.
Ohresser
b,
M.-A.
Arrio
d,
Ph.
Sainctavit
d,
A.
Barbier
a,∗aCEA/Saclay,DSM/IRAMIS/SPEC,CNRSUMR3680,F-91191Gif-sur-Yvette,France bSynchrotronSOLEIL,L’OrmedesMerisiersSaint-Aubin,F-91192Gif-sur-Yvette,France cSorbonneUniversités,UPMCUnivParis06,UMR7588,INSP,F-75005Paris,France dIMPMC,F-75015Paris,France
a
r
t
i
c
l
e
i
n
f
o
Articlehistory:
Received24October2014
Receivedinrevisedform3February2015 Accepted7February2015
Availableonline16February2015 Keywords:
Oxides Ferrites
Cationdistribution
X-rayphotoelectronspectroscopy(XPS) X-raymagneticcirculardichroism(XMCD)
a
b
s
t
r
a
c
t
ThepropertiesofCoFe2O4/BaTiO3artificialmultiferroicmultilayersstronglydependonthecrystalline
structure,thestoichiometryandthecationdistributionbetweenoctahedral(Oh)andtetrahedral(Td)sites (inversionfactor).Inthepresentstudy,wehaveinvestigatedepitaxialCoFe2O4layersgrownonBaTiO3,
withdifferentCo/Feratios.WedeterminedthecationdistributioninoursamplesbyX-raymagnetic circulardichroism(XMCD),awellacceptedmethodtodoso,andbyX-rayphotoelectronspectroscopy (XPS),usingafittingmethodbasedonphysicalconsiderations.WeobservedthatourXPSapproach convergedonresultsconsistentwithXMCDmeasurementsmadeonthesamesamples.Thus,withina carefuldecompositionbasedonindividualchemicalenvironmentsitisshownthatXPSisfullyableto determinetheactualinversionfactor.
©2015ElsevierB.V.Allrightsreserved.
1. Introduction
TheferroelectricperovskiteBaTiO3coupledwiththe
ferrimag-neticspinelCoFe2O4 isawellreferencedsystemforthestudyof
themagneticstatesuponmultiferroiccouplings[1,2].Suchastudy requirestofullyanalyzethemagneticpropertiesofthespinelof suchsystems,amongthemthespinelinversionfactor.
WealreadyknowthatbulkCoFe2O4hasseveralkeyproperties
suchasstrongmagneto-crystallineanisotropy,high magnetostric-tionandsaturationmagnetization,aswellashighcoercivefields atroomtemperature[3,4].BulkCoFe2O4crystallizesinthespinel
structurewithalatticeparameterof0.838nm.Theminimalcubic unitcell(i.e.gatheringthe4rhombohedralunitcellsimposedby thefcclattice)contains8formulaunitsandthespinelspacegroup isFd ¯3m.TheCoandFecationscanoccupyeithertetrahedralTd oroctahedralOhsitesoftheoxygensublattice.Theirdistribution followstheformula [FeyCo1−y]Td[Fe2−yCoy]OhO4,withthespinel
inversionparameterybeing0fora normalspineland1 foran inverseone. Latticeenergycalculations [5]showthatCo2+ions
∗ Correspondingauthor.
E-mailaddress:abarbier@cea.fr(A.Barbier).
favorOhsitesinCoFe2O4,thusaninversespinelstructureis
pre-dicted.Inpracticeyisusuallysmallerthan1,especiallyinthinfilms
[6].
Tothebestsofourknowledge,apreciseanalysisofthecation distributioninsidethespinelofBaTiO3–CoFe2O4epitaxialthin
lay-ersmultiferroicshasnotbeenlookedintobymostofthemain references on the subject of magneto-electric coupling [2,7,8]. However,thedistributionofthosecationswithinthespinel struc-turestronglyaffectstheelectricandmagneticpropertiesofthe samples.Forinstance,theresultingmagneticmomentwillbe dif-ferentdependingonthecationdistribution(anillustrationofthe conceptisdetailedforMnFe2O4ferriteinRef.[9]).
Itisthereforeofhighimportancetodeterminethisionsite dis-tributioninordertodeducetheglobalmagneticmoment,aswellas foranymagneticcharacterization,andinparticularinthecontext of magnetoelectriccoupling. X-raymagnetic circular dichroism (XMCD)isconsideredthemostprecisetechniquetodeterminethe inversionparameter;howeveritrequiresaccesstoasynchrotron beamline.Alternatively,authorsoftenchoosetoestimatethecation distributionusingX-rayphotoelectronspectroscopy(XPS),asitisa commonapparatusforspectroscopyinlaboratories,andsinceitis obviousthatthespectrahavetocontainallnecessaryinformation likeintheX-rayabsorptionspectroscopy(XAS)case.Nevertheless,
http://dx.doi.org/10.1016/j.elspec.2015.02.006
frompreviousreportsnoclearandreliablestrategyfortheXPS fit-tingprocedurewasestablished.Itisstillunclearhowtocorrectly exploitXPSspectratoextractthatdata,especiallyconsideringthat agivenfittingmethodtendstohaveanon-negligibleinfluenceon theresult,andmethodsseemtovaryfrompapertopaper[10,11], sometimesleadingtocontroversies[12,13].Thelackofdirect com-parisonwithXMCDdoesnotallowtoconcludeonthepertinence oftheproposedapproaches.
OuraimistodefineareliablestrategyforXPSdataprocessing basedonphysicalconsiderationsandconsistentwithXMCD mea-surements,inordertoextracttheadequatecationdistributionin CoFe2O4thinfilms.Forthatpurpose,weusewell-defined
epitax-ialthinfilmsofCoFe2O4(001)grownonBaTiO3underlayers.The
choiceofanalyzingspinelsampleswithdifferentCo/Feratioswas adoptedtoinsurethestabilityofthemethodforaseriesof sam-plesforwhichthecationorganizationislikelytovary.Weexpect thatthepresentmethodcanbeextendedtoalargevarietyofother spinels.
Maghemiteistheoxidizedformofmagnetitewhereallthe diva-lentironhavebeenreplacedbytrivalentones.IndeedthreeFe(II) ionscanbereplacedbytwoFe(III)ionsandavacancy. Magnemi-tizationofspinelisobservedwhenpartofFe(II)ionsarereplaced byFe(III).Byconvention,wediscussthedegreeofspinelinversion independentlyofthedegreeofmaghemitization,whichbecomes anadditionalparameterwhenCo/Fediffersfrom0.5.Inthesame manner,wedefinethecationicstoichiometryastheCo/Feratio regardlessofthedivalent/trivalentdistribution.
2. Experimental
ThreesamplesofCoFe2O4/BaTiO3/SrTiO3(001)wereprepared
by Atomic-Oxygen assisted Molecular Beam Epitaxy (AO-MBE)
[14–16].
The growth was performed in a dedicated AO-MBE setup workinginthe10−10mbarpressurerangeequippedwitha radio-frequency plasmasource,anAugerelectron spectroscopy(AES) analyzer,areflectionhighenergyelectrondiffraction(RHEED)gun andconnectedinultrahighvacuumconditionstoaseparate ultra-highvacuumchamberwithanX-rayphotoelectronspectroscopy apparatus.Duringgrowth,thesamplesrotatecontinuouslyaround thesurfacenormaltoobtainhomogenousfilms.
Themainfeatures ofthesamplesatdifferentstagesof their growtharedetailedinTable1.Theymainlydifferbyhaving dif-ferentCo/Feratios.TwooftheSrTiO3(001)substrateswere1%Nb
doped(forsamplesnumber1and2),thelastwasundoped(sample 3).Allthreesubstrateswereannealedinairat900◦Cfor3hafew weeksbeforedepositioninordertorestorethenominal stoichiom-etryandgetridofthepolishinginduceddefects.Wefirstperformed theepitaxialgrowthofBaTiO3(001)onSrTiO3(001)inthe
con-ditionsdetailedin[14].BaandTiweredepositedfromindividual Knudsencells.Thegrowthprocesswasmonitoredinrealtimeby RHEED,whichisatechniquethat permitstocontrolthecrystal quality.ThestoichiometryofthesampleswascheckedbyAESand XPSafterthegrowth.
Table1
MaincharacteristicsoftheCoFe2O4/BaTiO3/SrTiO3samples.
Sample 1 2 3
BaTiO3thickness(nm) 15 13 11
CoFe2O4thickness(nm) 6 6 6
XPSCo/Fe 0.28 0.48 0.63
SpinelinversionparameteryXMCD 0.90 0.80 0.90
SpinelinversionparameteryXPS 0.808 0.806 0.824
%Fe3+
OhbyXMCD 55% 60% 55%
%Fe3+
OhbyXPS 59.6% 59.7% 58.8%
AftertheBaTiO3 (001)crystallinegrowththesampleswere
extractedfrom theAO-MBEchamber andmounted ona differ-entsampleholderoptimizedforCoFe2O4 growth,thenlater
re-introducedinthesameepitaxychamber.Thecarbonobservedby AESontheBaTiO3,duetoair-exposure,wasremovedby45min
highbrillianceoxygenplasmaexposure.Thentheferritegrowth wasperformedasinRef.[15],andatatemperatureof450◦C.Then again,realtimeRHEEDandpost-growthinsituAESandXPS tech-niqueswereused.
TheXPSapparatususestwinanodeswithaXR3X-raysourceand thehemisphericalanalyserAlpha110fromsocietyThermoElectron Corporation.XPSspectrawereobtainedbyusingAl-K˛(1486.6eV) andMg-K˛(1253.6eV)X-raysources.Theprecautionofusingtwo sourcesistoinsurethattheAugerorphotoemissionpeaks can-not beconfused.A widescan of thesampleswasfirst realized (0–1200eV)todetectallthespeciesinthesample.Thenwefocus onFe2p,Co2pandO1s.Thosespecificenergyintervalswerethen scannedmultipletimes–typically10scansforCoandFe,5forO–in ordertomaximizethesignaltonoiseratio.Theenergyresolutionof theXPSspectraistypicallyaround0.1eV.Analysiswasperformed usingthesoftware“AvantageDataSpectrumProcessing”TM.
TheXMCDmeasurementswereperformedonDEIMOS (Dichro-ismExperimentalInstallationforMagneto-OpticalSpectroscopy) beamlineatsynchrotronSOLEIL(Saint-Aubin,France).DEIMOSis abeam-linededicatedtothestudyofthemagneticandelectronic propertiesusingpolarizedlight.ThemainadvantageofXMCDover othertechniquestoinvestigatemagneticbehaviorisitschemical and orbitalselectivity.Theenergyrange is 350–2500eV,which allowsthestudyofL2,3edgesofFeandCoandtheKedgeofO.The
energyresolutionisE/Efrom6000to10,000.Theoverall prop-ertiesofthebeamline[17]giveusveryclearspectradespitethe smallthicknessofthethinfilmofspinelof6nm.Measurements wereperformedat4K.Thecryo-magnetdelivers±7Talongthe beamdirectionand±2Tperpendiculartothebeam.AllourXMCD measurementswecarriedoutat4K,withanincidenceangleof30◦ withrespecttothesurfaceandamagneticfieldof5Talongthe beamdirection.
3. Results
TheRHEEDpatterns wererecordedduringthedepositionof both BaTiO3 and CoFe2O4.Anexample offinal RHEED patterns
is reportedin Fig.1. It indicatesthat thesamplesare epitaxial and single crystalline.Also, we can observefrom the compari-sonofsample3(non-stoichiometric,Co/Fe=0.63)withareference stoichiometricsample(Co/Fe=0.50)thattheRHEEDpatternsare similar.Moderatestoichiometrychangesdonotseemtoaffectthe observationsmadeusingtheRHEEDtechnique.X-rayreflectivity measurementsperformedat8keVgaveusanaveragedeposition rateof0.06nm/minforBaTiO3andof0.13nm/minforCoFe2O4.The
deducedthicknessesofdepositionaregiveninTable1.
Becauseouraimistodeterminethecationdistributionbetween octahedralandtetrahedralsitesinthespinelusingXPS,itis nec-essarytohaveareliableandwellacceptedbenchmark.Thusthis distributionwasfirstdeterminedusingXMCDmeasurementson DEIMOSbeamline.ThemeasurementswereperformedattheL3
edgeofFearound716.7eV.Theexperimentalplotsarepresented inFig.2, left,forthedifferentsamples.The similarformof the XMCDplotsindicatesthattheaverageelectronicstructureforFe issimilarforthethreesamples.Inaddition,onenoticesthatthe samplesareinversespinel.Thoseplotswerefittedusingalinear combinationofthetheoretical[16,18]spectraofFe3+in
octahe-dralandtetrahedralsitesandofFe2+inoctahedralsites(Fig.2,
middle),asthesethreecomponentsaresufficientsinceFe2+isnot
Fig.1.RHEEDpatternsafterthedepositionofCoFe2O4onBaTiO3–stoichiometricsample(top),sample3(bottom),[100]direction(left),[110]direction(right). absorptionspectratakenfrom[19,16]originatefromacrystalfield
atomicmultipletapproach,takingintoaccounttheoccupied3d electronconfigurations,thecrystalfield,thespin–orbitcoupling andelectron–electroninteractionswithinFe,andthehybridization of3delectronstoothervalenceelectrons.
Theleastsquaresmethodwasappliedusingtwoparameters, theFe3+proportiononthetotalofFe2+andFe3+cations(closeto
1asFe2+appearedneglectful),andtheFe3+Ohproportiononthe
totalofsitesFe3+OhandFe3+Td.Thismethodwasusedtoconclude
onthebestfittingvalues(Fig.2,right).
ThepercentageofFe3+inoctahedralsitesisrespectivelyof55%,
60%and55%forsamplesnumber1,2and3.Wethusconcludethat theCoFe2O4thinlayersgrownonBaTiO3crystallizemainlyinthe
inversespinelstructure.
Fig.2.ExperimentalXMCDspectraattheL2,3edgeofFeinthesamples(a);theoreticalXMCDplotsofFe3+Oh,Fe3+TdandFe2+Oh(b);fromRef.[19],fittedXMCDspectraof
Fig.3.XPSspectrumofCoinCoFe2O4forsample3(a).XPSspectrumofFeinCoFe2O4forsample3(b).
TheXPSspectrathatwemeasuredforCoandFearepresentedin
Fig.3.Themaincomponentsarethepeakscorrespondingto elec-troniclevels2p3/2and2p1/2andtheassociatedshake-upsatellites
forbothmetals.Theshake-uppeaksareattributedtothetransition ofanelectronfroma3dorbitaltotheempty4sorbitalduringthe ejectionofthecore2pphotoelectron[20].
Iftheobtainedspectracanbesimilarfordifferentlaboratories, theexploitationmethodtendstovaryfrompapertopaperin liter-ature.Zhouetal.[10]fitXPSspectraonbulkandpowderCoFe2O4
anddifferentiatesTdandOhsites,butdonotusethesame back-groundasusforironnorfitshake-upsatellites.Wangetal.[11]fit alsoCoandFewithtwopeaks,andaddthefittingoftheshake-up satellite,butonlyconsiderthe2p3/2ionlevelsandnotthe2p1/2
aswell.Thefollowingparagraphsdetailourapproachtowardsthe exploitationofthoseXPSspectra.
Firstofall,beyondthequestionofsiteoccupation,theXPS spec-traallowusalsotoestimatethestoichiometryofthesamples.This wasdoneusingaShirley-typebackgroundbyfittingtheCo2p3/2
andFe2p3/2peaks(octahedralsitesandtetrahedralsitesand
shake-up)anddeducingtheirarea.Letsnoticethatthisdecompositionhas morevaliditythantryingtohighlightaseparationbetween differ-entoxidationstatesofironasXMCDmeasurementshaveshown thatFe2+islessthan1%ofFe3+.Thisareaisthencorrectedusing
thecorrespondingScofieldfactors[21].Wechosetoestimatethe Co/FeratiobycalculatingtheonebetweentheScofield-corrected intensityareaofCo2p3/2andFe2p3/2.Thechoiceof2p3/2over2p1/2
wasbecauseofthehigherintensityofsignalgivingusmore pre-cision.Thisratioisfoundtoberespectivelyof0.28,0.48and0.63 (Table1)forsamples1,2and3.
Experimentally,itiseasiertodiscriminatethedifferent contrib-utionstothepeaksfortheCospectrathanfortheFespectra.This observationwasmostlyconfirmedduringthefittingprocesswith the“AvantageDataSpectrumProcessing”TMsoftware,whichwas
morereproducibleforCo.ThemainCo2p3/2peakwasdecomposed
intotwodifferentpeaks(Co2p3/2OhandCo2p3/2Td(Fig.4))
how-everthemainFe2p3/2 peakwastoobroadanddidnotallowan
obviousdiscriminationofindividualcomponents(Fe2p3/2Ohand
Fe2p3/2Td)bythisfittechnique.Asaconsequence,wechosethe
XPSspectraforCoasareferenceforestimatingthespinelinversion parameter,ratherthanFe.ThisdecompositioninOhandTdpeaks iscommononFe[22],andweproceededbyanalogytodothesame onCo.
WesearchamethodologyallowingtofittheCo2pXPS spec-truminareproduciblewaybasedonphysicalpropertiesinterms ofelectronicstructureinordertoquantitativelyassesstheratios ofOhandTdsites.Thismethodisbaseduponthefollowing con-straints.Firstofall,weuseaShirleybackgroundthatcoversallthe energyrangefromCo2p3/2toCo2p1/2.Asaguardrule,forallfits
ofallelements,weusedaShirleybackground.Thenextstepisto
imposeanidenticalspinelinversionparameterfortheCo2p3/2and
Co2p1/2states,whichimpliesthattheratiobetweentheareasof
theXPSpeaksofCoinTdandOhsites,Co-Td/Co-Oh,mustbethe sameforbothCo2p3/2 andCo2p1/2.Thisconstraintisrelaxedat
theendofthefitandthefitisacceptedifthereisnomorethan 2%shiftfroma centralvalue.Thethirdconstraint derivesfrom therigidbandapproximation[23]linkedtothecontinuityinthe perceivedenvironmentforthedifferentsites.Itimpliestokeep,for thedecompositionofCo2p3/2andCo2p1/2inpeaksassociatedtoOh
andTdsites,theenergygapbetweenbothsitesECo−Oh–ECo−Td con-stantforeachelectronicconfiguration.Thisconstraintisrelaxedat theendofthefitandthefitisacceptedifthereisnomorethan0.1eV shiftfromacentralvalue.Theshake-upisabroadpeakconsisting ofamixofdifferentpeaks,andisconsideredaswellrepresented byauniquepeak.Wefixnoconstraintsonitsenergypositioning forthesamereasons.
Thefitisperformedusingsymmetricpeaksinordertolimit thenumberoffittingparameters,evenifwecannotexcludesome asymmetry.Howeverthismethodissuitabletogiveustherelative areaofoctahedralandtetrahedralpeakswhichmakesaphysical sense.
Andindeedalltheseconstraintsputtogethergiveusaunique andperfectlyreproduciblefitfortheCo2pspectra(Table2). What-evertheinitialpositionbeforetheprocess,thefitalwaysconverges onthesamevalues.ThisfullyjustifiestotakeCoasareference. However,ifweusethesamefittingprocedureforfittingtheFe2p lines,weobservedthattheresultsarestronglydependentofthe initialparameter.Itismaybeduetolargerpeaksandpresenceof differentcomponents(Fe2+andFe3+).Moreover,inthesameenergy
range,thereisalsotheKLLAugerofoxygen(Figs.3and5)whichis mixedwiththeshake-upofFe2p1/2.
Table2
BindingenergypositionofthefittingpeaksinXPSforsample3.
Sample3 Co Fe 2p3/2Oh 786.95eV 717.50eV 2p3/2Td 789.35eV 720.15eV 2p3/2shake-up 793.35eV 725.65eV 2p1/2Oh 802.35eV 731.65eV 2p1/2Td 804.6eV 733.05eV 2p1/2shake-up 809.5eV 740.6eV
Fig.5. XPSspectrumofFeinCoFe2O4–decompositioninindividualcontributions ofthebestfitforsample3.
Becauseaninversespinelhas100%ofitsCoinOhsites(y=1), andanormalspinel100%inTdsites(y=0),wededucethespinel inversionparameter(y)tobeequaltotheoccupationpercentage ofthecobaltoctahedralsites.FromthefitofCo2pspectra,this percentagewasfoundtobe,forsamplesnumber1,2and3,of80.8%, 80.6%and82.4%respectively.Thoseresultsarewithina2%error marginlinkedtotheshiftafterreleasingconstraints.
WetriedtofittheFe2pXPSspectrainthesamemanner,but thereproducibilitywasaproblem.Indeed,weobtainedgoodfits correspondingtosignificantlydifferentinversionratios.Asa con-sequence,thosespectracouldnotbeusedasthereferenceforour XPSmethod.However,asweexaminethesamesample,itis pos-sibletoimposethesupplementaryconstraintofhavingthesame spinelinversionparameterasdeterminedfromtheCo2pspectrum. Usingthisaddedconstraint,wecanfindagoodfitthatcorresponds totheexpectedresults,andwhichismoreoverunique(Fig.5and
Table2).
ConcerningtheFe2p fitting,wefoundthatitisimportantto determinethebackgroundoveralargerangeincludingthehuge peakat≈748eVcorrespondingtotheKLLAugerpeakofoxygen, whichcontributesmassivelytothebackground.Itisherea bet-tersolutionthanswitchingtoanAlK˛sourcetoeliminatethe AugercontributionofO,astheAugercontributionofCowouldthen overlapinenergywithFe2p.
4. Discussion
The difference of binding energy between octahedral and tetrahedralsitesforsamplesnumber1,2and3arerespectively of2.75eV,2.23eVand2.63eVforFe2p(averagevaluesofFe2p3/2
andFe2p1/2),withina0.1eVmargin,andof2.40eV,2.35eVand
2.38eV for Co2p. Thesevalues are closeto theone found for CoFe2O4 nanoparticles[10](3eVforFe2pand2.10eVforCo2p),
wheretheamountofnormalspinelstructureishigherthaninour samples.Comparingourresultswithotherexperimentsonsimilar samplescangivediscrepancies[24]sincesomedonottakeinto accounttheCo2p1/2peakorhavenotfittedtheshake-upsatellite.
Itsoseemsthattakingintoaccountallparameters–background,
electroniclevels2p3/2 and2p1/2,shake-upsatellites– iskey in
findingaunifiedmethodofXPSanalysis.
Thefactthatthedifferenceinbindingenergybetween octahe-dralsiteandtetrahedralsiteinXPSisbiggerthatthedifferenceof absorptionedgesofthesamesitesisnotreallysurprising.Indeed, thescreeningofthecoreholeisreflectedinthespectraina com-pletelydifferentmannerforthetwoprocesses.
TheinversionparameteryobtainedusingtheXPSmethod,of 80.8%,80.6%and82.4%forsamplesnumber1,2and3,canbelinked totheestimatedpercentageofFe3+inoctahedralandtetrahedral
sites.Ifweconsiderthatinaninversespinel,Fe3+isinsimilar
pro-portioninbothOhandTdsites,andthatinanormalspinelallFe3+
areinOh sites,wefindthat%Fe(Oh)=(y×0.5+(1−y)×1)×100 whichis59.6%,59.7%and58.8%.Eachofthoseresultsislessthan 5%differentfromresultsfromXMCD–55%,60%and55%for sam-plesnumber1,2and3–whichisintheerrormarginoftheXPS technique.
Wecanalsoobservethatthespinelinversionparameteryseems fairlyindependentoftheCo/Feratio,soofthestoichiometryofthe spinel.Thisresultisquitesurprising[25].Indeed,for stoichiomet-rieswithlowCo2+cationscontent,wewouldhaveexpectediron
cationstoreplacesomeoftheOhsites,modifyingthusthe inver-sionparameter.Oramaghemitizationofthecrystal,whichwould meancationicvacanciesbutwouldimplyalsoasmallchangeofthe inversionparameter.
TheresultsindicatethatthespinelofourBaTiO3–CoFe2O4
mul-tiferroicthinlayerscrystallizesinamostlyinverseconfiguration andthusofhighmagneticmoment.
IfXMCDremainsarecognizedmethodwhenlookingfora pre-cisedeterminationofthespinelinversionparametery,weshow thatXPSspectracoupledtophysicalconstraintsfortheanalysis, cangiveacomplementaryestimation.BothXMCDandXPS meth-odsyieldverysimilarresults.ThisconvincesusthatXPScanbean excellentsynchrotronfreeapproach,whena5%erroronthesought informationremainsanacceptableprecision.
5. Conclusion
We have grown CoFe2O4/BaTiO3/SrTiO3 samples by
oxygen-assisted molecular beamepitaxy, fully controlledin crystalline structureandstoichiometry.ThespinelfilmofCoFe2O4was
stud-iedbothinXMCDandXPS.Toperformthatstudy,wefocusedon theFespectrainXMCDandtheCospectrainXPS.Tofitcorrectly theFespectrainXPS,itappearednecessarytoperformthescanon alargebandofenergyaroundFetoobtaintheactualbackground. Usingphysicalconsiderationstoimposeconstraintstothefitof theXPSspectra,wemanagedtoobtainquantitativevaluesofthe distributionoftheCoandFecationsbetweenoctahedraland tetra-hedralsitesofthespinel.Thosevalueswerefullyconsistentwith theXMCDresultsobtainedonthesamesamples,therefore con-firmingourapproach.Wealsohighlightedtheinversenatureof thespinel.Thismethodcanbeofmajorutilitytoobtainroutinely thosevaluesinlaboratoriesequippedwithanXPSapparatus,rather thanhavingtoperformthedeterminationonasynchrotron beam-line.Itisalsolikelyapplicabletoanyotherspinelferritessuchas CrFe2O4,MnFe2O4,NiFe2O4,CuFe2O4.
Acknowledgements
WeacknowledgeSOLEILforprovisionofsynchrotronradiation facilities and are gratefulto theSOLEIL-DEIMOSbeamline staff memberswhohelpedachievethoseexperiments.Thisworkwas partlysupportedbytheCNano-MAEBAgrant,forwhichwearealso grateful.
References
[1]R.Chopdekar,V.Malik,A.F.Rodríguez,L.LeGuyader,Y.Takamura,A.Scholl, D.Stender,C.Schneider,C.Bernhard,F.Nolting,etal.,Phys.Rev.B86(2012) 014408.
[2]H.Zheng,J.Wang,S.Lofland,Z.Ma,L.Mohaddes-Ardabili,T.Zhao,L. Salamanca-Riba,S.Shinde,S.Ogale,F.Bai,etal.,Science303(2004)661.
[3]J.Ryu,A.V.Carazo,K.Uchino,H.-E.Kim,Jpn.J.Appl.Phys.40(2001)4948.
[4]E.Manova,B.Kunev,D.Paneva,I.Mitov,L.Petrov,C.Estournes,C.D’Orléan, J.-L.Rehspringer,M.Kurmoo,Chem.Mater.16(2004)5689.
[5]H.S.C.O’Neill,A.Navrotsky,Am.Mineral.68(1983)181.
[6]S.Matzen,FilmsultramincesépitaxiésdeMnFe2O4,CoFe2O4etNiFe2O4pour lefiltragedespinàtempératureambiante,UniversitéPierreetMarieCurie -ParisVI,2011(Ph.D.thesis).
[7]C.A.Vaz,J.Hoffman,C.H.Ahn,R.Ramesh,Adv.Mater.22(2010)2900.
[8]R.Chopdekar,Y.Suzuki,Appl.Phys.Lett.89(2006)182506.
[9]S.Matzen,J.-B.Moussy,R.Mattana,K.Bouzehouane,C.Deranlot,F.Petroff,J. Cezar,M.-A.Arrio,P.Sainctavit,C.Gatel,etal.,Phys.Rev.B83(2011)184402.
[10]Z.Zhou,Y.Zhang,Z.Wang,W.Wei,W.Tang,J.Shi,R.Xiong,Appl.Surf.Sci.254 (2008)6972.
[11]W.P.Wang,H.Yang,T.Xian,J.L.Jiang,Mater.Trans.53(2012)1586.
[12]E.Paparazzo,J.ElectronSpectrosc.Relat.Phenom.154(2006)38.
[13]T.Yamashita,P.Hayes,J.ElectronSpectrosc.Relat.Phenom.154(2006)41.
[14]A.Barbier,C.Mocuta,D.Stanescu,P.Jegou,N.Jedrecy,H.Magnan,J.Appl.Phys. 112(2012)114116.
[15]A.Ramos,J.-B.Moussy,M.-J.Guittet,M.Gautier-Soyer,C.Gatel,P. Bayle-Guillemaud,B.Warot-Fonrose,E.Snoeck,Phys.Rev.B75(2007)224421.
[16]S.Matzen,J.-B.Moussy,R.Mattana,F.Petroff,C.Gatel,B.Warot-Fonrose,J. Cezar,A.Barbier,M.-A.Arrio,P.Sainctavit,Appl.Phys.Lett.99(2011)052514.
[17]P.Ohresser,E.Otero,F.Choueikani,K.Chen,S.Stanescu,F.Deschamps,T. Moreno,F.Polack,B.Lagarde,J.-P.Daguerre,etal.,Rev.Sci.Instrum.85(2014) 013106.
[18]C.Carvallo,P.Sainctavit,M.-A.Arrio,N.Menguy,Y.Wang,G.Ona-Nguema,S. Brice-Profeta,Am.Mineral.93(2008)880.
[19]P.Kuiper,B.Searle,L.-C.Duda,R.Wolf,P.VanderZaag,J.ElectronSpectrosc. Relat.Phenom.86(1997)107.
[20]L.Yin,I.Adler,T.Tsang,L.Matienzo,S.Grim,Chem.Phys.Lett.24(1974)81.
[21]J.Scofield,J.ElectronSpectrosc.Relat.Phenom.8(1976)129.
[22]T.Fujii,F.DeGroot,G.Sawatzky,F.Voogt,T.Hibma,K.Okada,Phys.Rev.B59 (1999)3195.
[23]D.Spanjaard,C.Guillot,M.-C.Desjonqueres,G.Tréglia,J.Lecante,Surf.Sci.Rep. 5(1985)1.
[24]W.Wang,H.Yang,T.Xian,J.Jiang,Mater.Trans.53(2012)1586.