Enhanced Spin and Electronic Reconstructions at the Cuprate-Manganite Interface B. A. Gray,1,a) E. J. Moon,1 I. C. Tung,2,3 M. Kareev,1 Jian Liu,4 D. J. Meyers,1 M. J. Bedzyk,3 J. W. Freeland,2 and J. Chakhalian1 1)Department of Physics, University of Arkansas, Fayetteville, Arkansas 70701, USA 2)Advanced Photon Source, Argonne National Laboratory, Argonne, Illinois 60439, 3 USA 1 0 3)MaterialsScienceandEngineering,NorthwesternUniversity, Evanston, Illinois,60208, 2 n USA a J 4)Department of Physics, University of California, Berkeley, California 94720, 6 1 USA ] l e We report on a resonant soft X-ray spectroscopy study of the electronic and magnetic - r t structureofthecuprate-manganiteinterface. PolarizedX-rayspectroscopymeasurements s . t takenattheCuLedgerevealuptoafive-foldincreaseinthedichroicsignalascomparedto a m pastexperimentalandtheoreticalvalues. Furthermoreanincreaseinthedegreeofinterlayer - d charge transfer up to 0.25e (where e is charge of an electron) per copper ion is observed n o c leadingto aprofoundreconstructionin theorbitalschemefor theseinterfacialcopperions. [ 1 ItisinferredthattheseenhancementarerelatedtoanincreaseinTMI observedformanganite v layersgrownwithrapidlymodulatedflux. 6 3 7 3 . 1 0 3 1 : v i X r a a)Electronicmail: [email protected] 1 The confluence of multiple competing order parameters stemming from the rich spectrum of groundstatesaccessibleinartificialtransitionmetaloxidemultilayersfacilitatestheexplorationof unique quantum states and phenomena at the interface. A quintessential example is the junction between the high-T cuprate YBa Cu O (YBCO) and the colossal magnetoresistance manganite c 2 3 7 La Ca MnO (LCMO),forwhichtheexactnatureoftheelectronicandmagneticstructureat 2/3 1/3 3 and near the interface remains actively debated. On one hand, proximity effects (PE) are excepted to play a dominate role in determining properties,1 and indeed prior extensive measurements of YBCO/LCMO heterostructures have revealed suppressions in the critical temperature and free carrierresponseoperatingoveralengthscalefarexceedingtheanticipatedsuperconductive(SC) condensatepenetrationdepthξ .2,3 OnepossibleexplanationfortheobservedlongrangePEisa F triplet pairing component of the SC condensate due to the presence of magnetic inhomogeneity neartheinterface.4 Throughthe combined utilizationofpolarizedneutron reflectivityandX-raymagnetic circular dichroism (XMCD), a definitive picture of the microscopic magnetic profile in YBCO/LCMO multilayerswasobtainedwhichcoincidedwiththeaforementionedtheoreticalpredictionsincluding aninducednetmomentintheSClayerattheinterfacewithanantiferromagneticcouplingtothe ferromagnetic (FM)layer.5–7 However, subsequent X-raylinear dichroism (XLD)measurements established the role of orbital reconstruction in which a strong covalent Cu-O-Mn bond at the interfacepromoteschargetransferbetweenlayersandarearrangementoftheCud occupation.8 z2−r2 ThesignofthemagneticinteractionbetweenCuandMnissimplydeterminedbytheGoodenough- Kanamorirules,whilechargetransferacrosstheinterfaceoffersanalternativeexplanationtothe previouslyseenlongrangePE.3,9 Despitetheseadvancements,recenttheoreticalandexperimental results have reproduced the magnetic dichroic signal on Cu but failed to account for the orbital reconstruction.10,11 Furthermore,recentlyevidenceforatripletSCcomponentinYBCO/LCMO heterostructureshasbeenemerging.12 In this letter, we report on a resonant X-ray absorption spectroscopy (XAS) study of the electronicandmagneticprofileofthecuprate-manganiteinterfaceintherepresentative[YBCO(9 u.c.)/LCMO(26u.c.)]×3superlattice(SL).CircularlypolarizedX-raysareusedtoacquireelement sensitiveinformationaboutthemagneticstructureoftheinterface,whilelinearpolarizationsprobe the electronicstructure. Temperature dependent dctransport measurements explorethe electronic andmagneticqualitiesofthefilms. Weconcludethatbothspinandelectronicreconstructionsare presentandmarkedlyenhanced. Furthermore,theseenhancementsarelinkedtothestructuraland 2 chemicalpropertiesofthemanganite-cupratelayersobtainedbyrapidlymodulatedflux. SamplefabricationwascarriedoutinarecentlydevelopedPulsedLaserEpitaxy(PLE)facility featuringinfraredlasersubstrateheatingandin-situgrowthmonitoringviahigh-pressureReflection HighEnergyElectronDiffraction(HPRHEED).Duringdeposition,stoichiometrictargetsofYBCO andLCMOwereablatedbyaKrFlaser(λ =248nm). Thesubstratetemperaturewasheldat750 ◦C, and a partial oxygen pressure of 250 mTorr was maintained inside the chamber throughout the deposition. Immediatelyafterdeposition,thesubstratetemperaturewasloweredto530◦C,andthe samplewasannealedin500Torrofultra-pureoxygenfor1hour. Amutuallycompatiblegrowthregime(i.e. temperatureandpressure)for YBCO andLCMO was achieved through interval deposition in which material is deposited by high frequency pulses followedbyanextendeddwelltimebetweenthedepositionofeachunitcell(u.c.).13 Undamped RHEED specular intensity oscillations (not shown) were observed for both YBCO and LCMO layersoveralargenumberofcyclesallowingforalayer-by-layergrowthwithu.c. control. Theleft inset to Fig. 1 shows the post growth RHEED image for the SL along the (0 0 1) direction. The presenceofunbrokencrystaltruncationrodsuptothesecondordercombinedwithwelldefined specular(00)andoffspecularBraggreflectionstestifiestothequalityofthe2-Dgrowth. To determine the structural properties of the SL, we performed X-ray scattering at beamline 5-BM-D ofthe AdvancedPhoton Source(APS) atArgonne National Laboratory. Figure1 shows thespecularX-raydiffractionalongthe(00L)crystaltruncationrodasafunctionofthemagnitude of the out-of-plane momentum transfer vector, Q . As seen in Fig. 1, the SL shows all expected z Braggreflectionsforc-axisorientedlayersandKiessigfringes testifying tothequalityofthefilm andsharpnessoftheinterfaces. BasedonthepositionoftheYBCO(005)reflection,theaverage c-axislatticeconstantfortheYBCOlayersintheSLis11.70A˚,whichagreeswiththereported bulkvalueforoptimalstoichiometry.14 In order to establish the electronic qualities of the samples, we investigated the dc transport properties of single layer YBCO and LCMO films and the SL. Figure 2 shows resistivity versus temperatureforsinglelayersofYBCO(leftaxis)andLCMO(rightaxis). AsseenfortheYBCO film, thesuperconducting transitionT takes place at 93K attestingto theproper optimallydoped c stoichiometry. InthecaseoftheLCMOsample,themetal-to-insulatortransitiontakesplaceatT MI = 212 K (inflection point). Note, the previously reported values of T for LCMO films grown MI undersimilarconditions(i.e. substrate,thicknessetc.) butwithouttheuseofintervaldepositionare typically<170K.11 InaconventionalgrowthtoachieveanelevatedT asinourfilms,significantly MI 3 increasedfabricationtemperatureswouldberequiredwhich isadetrimentalconstraintoncuprate growth.15 Thestabilizationofmaterialphasesoutsideofregionsofthermodynamicstabilityisa hallmarkofintervaldeposition,andtheelectronicandmagneticpropertiesofLCMOareknownto dependsensitivelyonstoichiometryandthepresenceofdefects.16,17 Together,thesehelpexplain theoriginoftheenhancementsseeninthespectroscopydatabelow. Inaddition,thetemperature dependent resistance for the SL is plotted on the bottom left axis of Fig. 2. A suppression in the criticaltemperatureT =56.7Kisobservedandismostlikelytheresultofholedepletionthrough c interlayerchargetransferascorroboratedbytheXASdatabelow. Next we explored the electronic and magnetic structure with XLD and XMCD resonant soft X-rayspectroscopies. Theexperimentswerecarriedoutatthe4-ID-CbeamlineoftheAdvanced Photon Source in Argonne National Laboratory. In the XLD studies at the Cu L edge, we inves- tigated the orbital occupation by measuring the difference in absorption for polarizations in the ab-planeandalongthec-axis,whiletheXMCDexperimentsattheCuandMnLedgesobtained information regarding element specific magnetic moments by measuring the differences in the absorptionofrightandleftcircularpolarizations. Spectrawererecordedatanincidentangleof15 degreesoutofplanesimultaneouslyinbothfluorescenceyield(FY)andtotalelectronyield(TEY) acquisitionmodes. Thelinearpolarization-dependentCuL absorptionspectraoftheSLispresentedinFig. 3(a). 3 FirstweconsiderthebulksensitiveFYdatasetwhichshowsthestrongXLDexpectedfortheCu2+ (3d9)state ofcuprates.18,19 Note,the in-planesignalstemsfrom formallydivalentCuwithin the CuO sheets. Severalimportantfeaturescanbeidentifiedinthespectrumofwhichtheexcitonic 2 linenear931 eVisthemost prominent andcorrespondstotransitions fromtheCu2pcorelevels into the unoccupied bands of mainly Cu d orbital character (2p63d9 → 2p53d10). A clear x2−y2 shoulderisobservedonthehighenergysideofthewhiteline. Theshoulderisascribedtotransitions coupled with Cu ligand hole states (2p63d9L → 2p53d10L) and is the well-known signature of theZhang-Ricestate.20 AnotherimportantfeatureoftheFYspectrumisconnectedtothec-axis polarization,wherethemaximumoftheabsorptionpeakisshiftedby0.3eVtohigherenergy;the largernumberofligandholestatesinthechainstransferspectralweightfromtheexcitoniclineto thehighenergyshoulder. Sincethed orbitaloftheCuO planesisoccupied,theoutofplane z2−r2 2 polarizationmainlyprobesunoccupiedstatesfrommonovalentCuinthechainsofd orbital y2−z2 character. In the TEY data, distinct transformations in polarization dependence and lineshape manifest 4 in the spectra indicating charge transfer across the interface leading to a reconstruction in the orbital scheme for interfacial Cu ions. As indicated in Fig. 3(a), the position of the Cu white lineisshiftedby0.5eVtolowerenergyintheTEYspectrum,whichexceedsthevalue(0.4eV) reportedin theprior study. This substantialchemical shift isthe markof interfacialchargetransfer. Through acomparison to Cu1+ and Cu2+ reference materials,an approximate calculationyields a chargetransferamplitudeof0.25epercopperion.21 Furthermore,theshoulderathighenergyisnot observedintheTEYdataconfirmingthedepletionofholesfromtheYBCOregionoftheinterface. WhiletheevolutionintheamplitudeofthehighenergyshoulderisacommonfeatureofYBCO XASdopingprofiles,theshiftinthewhitelineisunexpected.18 Previously,thishasbeenattributed tothereconstructionofthed orbitalthroughtheformationofacovalentbondwithMnthrough z2−r2 apical oxygen across the interface.8,22 Confirming this picture, the large linear dichroism of the bulk is no longer present, and the absorption along the c-axis now even surpasses that of the the in-planefortheTEYspectrum. Finally, we turn our attention to magnetism on Cu and Mn. Figure 3(b) shows the XMCD spectra at both the Mn and Cu L-edges acquired in TEY mode. As anticipated for a FM system, thestrongmagneticdichroismreachesamaximumvalueof37.4%attheMnL edge. OntheCu 3 L-edge,however,onlyamoderatemagneticsignalispresent,whichindicatesanuncompensated magneticmomentonCuneartheinterface. FromthesignoftheXMCDonbothMnandCuedges, an antiparallel alignment of Cu and Mn moments coupled across the interface can be deduced. The magnitude of the dichroism (6.9 %) a the Cu L -edge is up to a factor five larger than the 3 previouslyreported values providing additional evidence for the high quality ofthe interface.7 The measurements at the Cu L-edge were repeated at 100 mT without loss in the magnitude of the magneticdichroism. In summary, our resonant soft X-ray spectroscopy study of the cuprate-manganite interface hasconfirmedthepresenceofbothspinandorbitalreconstructions. Weobservedincreasesinthe dichroicsignalandthechemicalshiftattheCuLedgesignalinganenhancementintheinduced momentonCuandthedegreeofinterlayerchargetransfer,respectively. Furthermore,anincreasein T hasbeenobservedformanganitelayersgrownwithintervaldeposition,whichisfundamentally MI relatedtoitsmagneticandelectronicproperties. J.C.wassupportedbytheDOD-AROunderGrantNo. 0402-17291andtheNSFunderGrant No. DMR-0747808. Work at the APS is supported by the U.S. DOE under grant no. DEAC02- 06CH11357. 5 REFERENCES 1A.I.Buzdin,Rev.Mod.Phys.77,935(2005);F.S.Bergeret,A.F.Volkov,andK.B.Efetov,Rev. Mod.Phys.77,1321(2005). 2Z. Sefrioui, D. Arias, V. Pena, J. E. Villegas, M. Varela, P. Prieto, C. Leon, J. L. Martinez, and J. Santamaria, Phys. Rev. B 67, 214511 (2003); V. Pena, Z. Sefrioui, D. Arias, C. Leon, J. Santamaria,M.Varela,S.J.Pennycook,andJ.L.Martinez,Phys.Rev.B69,224502(2004). 3T.Holden,H.-U. Habermeier, G.Cristiani,A.Boris, A.Pimenov,J.Humlic¨ek,O.I.Lebedev, G. VanTendeloo,B.Keimer,andC.Bernhard,Phys.Rev.B69,064505(2004). 4F.S.Bergeret,A.F.Volkov,andK.B.Efetov,Phys.Rev.Lett.86,4096(2001). 5F.S.Bergeret,A.F.Volkov,andK.B.Efetov,Phys.Rev.B69,174504(2004). 6J. Stahn, J. Chakhalian, Ch. Niedermayer, J. Hoppler, T. Gutberlet, J. Voigt, F. Treubel, H-U. Habermeier,G.Cristiani,B.Keimer,andC.Bernhard,Phys.Rev.B71,140509(2005). 7J.Chakhalian,J.W.Freeland,G.Srajer,J.Strempfer,G.Khaliullin,J.C.Cezar,T.Charlton,R. Dalgliesh,C.Bernhard, G.Cristiani,H.-U.Habermeier,andB. Keimer, NaturePhysics2,244 (2006). 8J. Chakhalian, J.W. Freeland, H.-U. Habermeier, G. Cristiani, G.Khaliullin, M. vanVeenendaal, andB.Keimer,Science318,1114(2007). 9J.W.Freeland,J.Chakhalian,H.-U.Habermeier,G.Cristiani,andB.Keimer,Appl.Phys.Lett. 90,242502(2007). 10X.Yang,A.Yaresko,V.Antonov,andO.Andersen,arXiv:0911.4349(unpublished). 11R.Werner,C.Raisch,A.Ruosi,B.A.Davidson,P.Nagel,M.Merz,S.Schuppler,M.Glaser,J. Fujii,T.Chasse,R.Kleiner,andD.Koelle,Phys.Rev.B82224509(2010). 12Y.Kalcheim,T.Kirzhner,G.Koren,andO.Millo,Phys.Rev.B83064510(2011);K.Dybko,K. Werner-Malento,P.Aleshkevych,M.Wojcik,M.Sawicki,andP.Przyslupski,Phys.Rev.B80 144504(2009);A.F.VolkovandK.B.Efetov,Phys.Rev.Lett.102077002(2009). 13G.Koster,G.J.H.M.Rijnders,D.H.A.Blank,andH.Ragalla,Appl.PhysLett.74,3729(1999); M.Kareev,S.Prosandeev,B.Gray,J. Liu,P.Ryan,A.Kareev,E.J.Moon,andJ.Chakhalian,J. Appl.Phys.109,114303(2011). 14R.Liang,D.A.Bonn,andW.N.Hardy,Phys.Rev.B73,180505(2006). 15O.I.Lebedev,G.vanTendeloo,andS.Amelinckx,Phys.Rev.B58,8065(1998). 16P.Schiffer,A.P.Ramirez,W.Bao,andS.-W.Cheong,Phys.Rev.Lett.75,3336(1995). 6 17J.Aarts,S.Freisem,R.Hendrikx,andH.W.Zandbergen,Appl.Phys.Lett.72,2975(1998). 18N. Nu¨cker, E. Pellegrin, P. Schweiss, J. Fink, S. L. Molodtsov, C. T. Simmons, G. Kaindl, W. Frentrup,A.Erb,andG.Mu¨ller-Vogt,Phys.Rev.B51,8529(1995). 19C.T.Chen, F.Sette,Y.Ma,M.S.Hybertsen,E. B.Stechel,W.M.C.Foulkes,M.Schulter,S-W. Cheong, A. S. Cooper, L. W. Rupp, Jr., B. Batlogg, Y. L. Soo, Z. H. Ming, A. Krol, and Y. H. Kao,Phys.Rev.Lett.66,104(1991). 20F.C.ZhangandT.M.Rice,Phys.Rev.B37,3759(1998). 21F.M.F.deGroot,J.ElectronSpectrosc.Relat.Phenom.67,529(1194). 22JianLiu,B.J.Kirby,B.Gray,M.Kareev,H.-U.Habermeier,G.Cristiani,J.W.Freeland,andJ. ChakhalianPhys.Rev.B84,092506(2011). 7 101 O (002) .u.) 10-1 (0 1)(0 0)(0 1-‐) STYB C OM (O 0(0060)2) a C ( -3 L 7) y 10 0 vit YBCO (005) O (0 flecti 10-5 YBCO (004) YBC O (008) e C R B Y -7 10 -9 10 2.0 2.5 3.0 3.5 4.0 Q (Å-1) Z FIG.1. (Coloronline)X-rayreflectivitydata(dots)anderrorbar(solidline)fora[YBCO(9u.c.)/LCMO (26u.c.)]×10SLSrTiO . TheinsetontheleftshowstheaftergrowthRHEEDimageforthesameSLona 3 ) 7 SrTiO300substrate. Whitetriangular markersindicatetheweakhalf-order signalwhichisattributedtothe ( O PbnmcCrystalsymmetryofthecappingLCMOlayer. Theright-handinsetisahighangleannulardarkfield B Y scanningtransmissionelectronmicroscopyimageoftheYBCO/LCMOinterfacialregion. Labelsindicate differentlayers. 8 ) 50 R m 160 150 u.c. LCMO e c s 50 u.c. YBCO 40 i Ω s µ 120 [YBCO (9 u.c.)/LCMO (26 u.c.)]x3 it v y ( 30 yti vit 80 20 1( ti 0 s 3 esi 40 10 Ωµ R c 0 0 m ) ) 60 Ω ( e 40 c n a 20 t s i s 0 e R 0 50 100 150 200 250 300 Temperature (K) FIG.2. (Coloronline)Temperature-dependentdctransportfor50u.c. YBCOand150u.c. LCMOsingle layersandthe[YBCO(9u.c.)/LCMO(26u.c.)]×3SLonSrTiO substrates. Measurementswereperformed 3 intheconventionalvanderPawconfiguration. 9 (a) Cu L Edge T = 15 K 6 3 Z.R. Pol. in ab-plane )) u.u.5 Pol. along c-axis .. aa (( 4 nn FY - Bulk oo ii ptpt3 rr oo ss2 bb AA 1 TEY - Interface 0 924 926 928 930 932 934 936 EEnneerrggyy ((eeVV)) (b) 10 x 2 ) % 0 ( D -10 Mn L Edge C M -20 Cu L Edge X -30 T = 15 K H = 5 T -40 630 640 650 660930 940 950 960 Energy (eV) FIG.3. (Coloronline)SoftX-rayspectroscopiesofthe[9u.c. YBCO/26u.c. LCMO]×3SLonSrTiO . 3 (a) XLD measurements at the Cu L -edge. The top and bottom sets of spectra correspond to FY and 3 TEY detection modes, respectively. Due to TEY’s shallow probing depth, only Cu at the first interface predominantlycontributestotheCuLintensityinthismode. (b)XMCDmeasurementsattheMn(left)and Cu(right)L-edgesinTEY.Note,toruleoutartifactstheXMCDdatasetswerecheckedatbothmagnetic fieldorientations. 10