Observation of room-temperature ferroelectricity

Observation of room-temperature ferroelectricity in tetragonal strontium titanate thin films on SrTiO3 (001) substratesY. S. Kim, D. J. Kim, T. H. Kim, T. W. Noh, J. S. Choi et al.Citation: Appl. Phys. Lett. 91, 042908 (2007); doi: 10.1063/1.2764437View online: /10.1063/1.2764437View Table of Contents: /resource/1/APPLAB/v91/i4Published by the American Institute of Physics.Related ArticlesFeCr2O4 and CoCr2O4 spinels: Multiferroicity in the collinear magnetic state?Appl. Phys. Lett. 99, 172903 (2011)Effect of bottom electrodes on nanoscale switching characteristics and piezoelectric response in polycrystalline BiFeO3 thin filmsJ. Appl. Phys. 110, 084102 (2011)Nanoscale phase separation in quasi-uniaxial and biaxial strained multiferroic thin filmsAppl. Phys. Lett. 99, 132905 (2011)Experimental evidence of ferroelectric negative capacitance in nanoscale heterostructuresAppl. Phys. Lett. 99, 113501 (2011)Piezoresponse force microscopy observation of domain switching in Bi3.15Nd0.85Ti3O12 thin film prepared by pulsed laser depositionJ. Appl. Phys. 110, 054105 (2011)Additional information on Appl. Phys. Lett.Journal Homepage: /Journal Information: /about/about_the_journalTop downloads: /features/most_downloadedInformation for Authors: /authorsObservation of room-temperature ferroelectricity in tetragonal strontium titanate thin films on SrTiO 3…001…substratesY .S.Kim,D.J.Kim,T.H.Kim,and T.W.Noh a ͒ReCOE and FPRD,Department of Physics and Astronomy,Seoul National University,Seoul 151-747,KoreaJ.S.Choi and B.H.ParkDepartment of Physics,Konkuk University,Seoul 143-701,KoreaJ.-G.YoonDepartment of Physics,University of Suwon,Kyunggi-do 445-743,Korea͑Received 2May 2007;accepted 3July 2007;published online 26July 2007͒The authors investigated the ferroelectric properties of strontium titanate ͑STO ͒thin films deposited on SrTiO 3͑001͒substrates with SrRuO 3electrodes.The STO layer was grown coherently on the SrTiO 3substrate without in-plane lattice relaxation,but its out-of-plane lattice constant increased with a decrease in the oxygen pressure during ing piezoresponse force microscopy and P -V measurements,the authors showed that the tetragonal STO films possess room-temperature ferroelectricity.The authors discuss the possible origins of the observed ferroelectricity.©2007American Institute of Physics .͓DOI:10.1063/1.2764437͔SrTiO 3is a well-known incipient ferroelectric ͑FE ͒ma-terial.At low temperature,it has a very large dielectric con-stant but remains in a paraelectric state.1Its FE stability seems to be hampered by competing interactions,such as quantum fluctuation and antiferrodistortion.2,3Such a deli-cate balance can be easily upset by external parameters.For example,other studies have reported ferroelectricity in SrTiO 3due to doping with other cations,4substituting oxy-gen isotopes,5applying electric fields,6or using strain engi-neering techniques.7–10In this letter,we report the observation of room-temperature ferroelectricity in tetragonal strontium titanate ͑STO ͒thin films.We grew STO thin films coherently on SrTiO 3͑001͒substrates between SrRuO 3͑SRO ͒top and bot-tom electrodes.We discovered that our STO films,grown at a lower oxygen partial pressure ͑P O 2͒,have larger out-of-lattice constants and show ferroelectric behavior at room temperature.We discuss the roles of strain and defects on the observed ferroelectricity in the STO films.To fabricate SRO/STO/SRO heterostructures on SrTiO 3͑001͒substrate,we used pulsed laser deposition ͑PLD ͒with in situ reflection high-energy electron diffraction ͑RHEED ͒.11We used a KrF excimer laser with repetition rate of 1Hz and fluence of ϳ2J/cm 2.We deposited the bottom SRO,STO,and top SRO layers sequentially at 600°C with-out breaking the vacuum.Before deposition,we prepared TiO 2-terminated SrTiO 3substrates.10Then,we deposited an ϳ15-nm-thick SRO bottom electrode layer using a polycrys-talline SrRuO 3target at P O 2of 2ϫ10−2Torr.Next,we grew the STO layer using a SrTiO 3single crystal target at various P O 2,ranging from 2ϫ10−4to 2ϫ10−2Torr.We monitored the layer-by-layer growth mode carefully and obtained an ϳ100-nm-thick STO layer by counting RHEED intensity oscillations.After depositing an ϳ15-nm-thick SRO top electrode,we lowered the temperature to room temperature while keeping P O 2constant.We found that all of the layers,especially the STO layer,were grown coherently on SrTiO 3͑001͒substrate.To in-vestigate the crystal structure and lattice constants of the STO thin films,we measured x-ray reciprocal space maps ͑X-RSMs ͒around the asymmetric ͑1¯03͒Bragg reflections.For the X-RSM measurements,we used a Bruker AXS D8advanced x-ray diffractometer with a Vactec-1detector.We plotted X-RSM in reciprocal lattice unit ͑r.l.u.͒,which is calculated from measured angular scans ͑1r.l.u.=2␲/3.905Å−1͒.Figures 1͑a ͒and 1͑b ͒show the X-RSM ofa ͒Electronic mail:twnoh@phya.snu.ac.krFIG.1.͑Color online ͒X-ray reciprocal space mapping around the asym-metric ͑1¯03͒Bragg reflection of SrRuO 3/STO/SrRuO 3heterostructure grown with oxygen partial pressure ͑a ͒͑P O 2͒=10mTorr and ͑b ͒P O 2=0.2mTorr.͑c ͒The P O 2dependences of in-plane and out-of-plane lat-tice constants of STO thin films are shown as solid red squares and solid red triangles,respectively.The tetragonality of STO thin films as a function of the P O 2is shown as solid blue circles.APPLIED PHYSICS LETTERS 91,042908͑2007͒0003-6951/2007/91͑4͒/042908/3/$23.00©2007American Institute of Physics91,042908-1the SRO/STO/SRO heterostructures grown with P O2=10and 0.2mTorr,respectively.These maps show clear͑1¯03͒Bragg peaks of the SrTiO3substrate,SRO,and STO layers in the H0L scattering plane.All of these peaks are on the same H-value line,indicating that the SRO and STO layers could be grown coherently on the SrTiO3substrates without lattice relaxation.The crystal structure of the STO layer has a systematicdependence on P O2.Note that the difference in the L valuesof the STO layer peaks in Figs.1͑a͒and1͑b͒from that of the SrTiO3substrate indicates that our STO layers have tetrago-nal symmetry,not cubic symmetry.The X-RSM also shows that the L value becomes smaller for thefilms grown at asmaller P O2.The solid red squares and triangles in Fig.1͑c͒show the measured values of the in-plane͑a-axis͒and the out-of-plane͑c-axis͒lattice constants,respectively.The a-axis lattice constant remains the same,independent of P O2 due to the clamping effect of the substrate,while the c-axislattice constant increases systematically as P O2decreases.The increase in the c-axis lattice constant is related to the expansion of the unit cell volume,which could be attributed to the formation of defects inside the STO layer.We define the tetragonality as͑c/a͒-1,where a and c are the a-and c-axis lattice constants,respectively.In displacive ferroelectrics,such as BaTiO3,ferroelectricity comes from the displacement of transition metal ions;therefore,large tetragonality is a prerequisite for ferroelectricity.10,12As shown in Fig.1͑c͒,the tetragonality of ourfilms͑the solidblue circles͒increases as P O2decreases.The tetragonalitybecomes even larger than that of FE BaTiO3,shown by thedashed line,especially for our STOfilms deposited at P O2ofless than10−3Torr.To check whether our STOfilms could be FE,we per-formed piezoresponse force microscopy͑PFM͒tests.These included room-temperature PFM tests using a Park Systems XE-100atomic force microscope͑AFM͒and temperature-dependent piezoresponse studies using another AFM,a Seiko Instruments SPA-300HV.Figure2͑a͒shows a schematic dia-gram of our PFM study setup.First,we scanned a3ϫ3␮m2square region with a dc bias of+10.5V.Then,we scanned a center1ϫ1␮m2square region with a reversed dcbias of−10.5V.Finally,we obtained PFM images by apply-ing an ac bias to the tip.Figures2͑b͒and2͑c͒show the room-temperature phase and amplitude images of the piezo-electric domains on a STO layer that was grown at P O2of0.2mTorr.Thesefigures indicate that the read/write and re-versible piezoelectric domains can be formed.Figure2͑d͒shows the temperature dependence of the maximum piezo-electric coefficient,2d33max.Thisfigure indicates that 2d33max has nearly temperature-independent values in the range of17–25pm/V.In the configuration of Fig.2͑a͒,surface charges trapped on the free dielectric surface might result in spurious signals.13To remove any systematic errors,we performed another PFM test using our SRO/STO/SRO heterostructures, which are shown schematically in Fig.3͑a͒.In this configu-ration,the STO layer was covered with a conducting SRO top electrode,so we could avoid the surface charge trapping effects and apply a uniform electricfield to the STO layer.In addition,we applied the dc and ac biases using a separate electrical connection and measured only the piezoresponse using the PFM tip.We found that this experimental setup reduced the noise significantly and provided fairly good d33-V hysteresis loops.Figure3͑b͒shows a room-temperature d33-V hysteresis loop for the STO layer,depos-ited at P O2of0.2mTorr.The observed2d33max value is ϳ27pm/V.The observations of piezoelectric domains and hysteresis loops suggested that the occurrence of ferroelec-tricity at room temperature in our tetragonal STO thinfilms is genuine.Further experimental evidence for ferroelectricity came from electrical measurements.To measure the P-V hysteresis loops,we used a Sawyer-Tower circuit,which consisted of a Yokogawa FG300function generator and a DL7100digital oscilloscope,and applied triangular waves of200kHz to the SRO/STO/SRO capacitors.Figures4͑a͒and4͑b͒show the P-V loops,measured on50ϫ50␮m2square capacitors,for thefilms grown at P O2of1.0and0.2mTorr,respectively. These P-V loops show typical FE responses.Figure4͑c͒illustrates the temperature-dependent remnant polarization ͑P r*͒values of our STO thinfilms.As the temperature de-creases,the value of2P r*increases.At10K,the2P r*value of the STOfilm grown at P O2of 1.0mTorr becomes ϳ6␮C/cm2.Note that this2P r*value is larger thanthe FIG.2.͑Color online͒͑a͒Schematic diagram of the PFM setup.͑b͒Phase and͑c͒amplitude of the piezoelectric domain after application of+10.5and −10.5V bias to the3ϫ3and1ϫ1␮m2areas,respectively.The PFM image was obtained at room temperature for the STOfilm grown at P O2=0.2mTorr.͑d͒Temperature dependence of the maximum piezoelectric coefficient,2d33max of the STOfilm͓2d33max=d33͑+͒+d33͑−͔͒.FIG.3.͑Color online͒͑a͒Schematic diagram of another PFM setup.In this setup,uniform dc and ac biases can be applied on the SRO top electrode with a separate electrical circuit.The PFM tip only reads mechanical piezo-electric response.͑b͒Room-temperature d33-V hysteresis loop for the STO films grown at P O2=0.2mTorr.reported 2P r *values of Ca x Sr 1−x TiO 3͑ϳ0.55␮C/cm 2at 4.3K ͒͑Ref.4͒and isotope-exchanged SrTiO 3͑ϳ0.6␮C/cm 2at 18K ͒.5The piezoelectric coefficient can be expressed as d 33=2Q ␧33P S ,where Q is the electrostric-tion coefficient and ␧33is the dielectric constant along the caxis.Note that the temperature dependence of 2P r*in Fig.4͑c ͒is similar to that of 2d 33max in Fig.2͑d ͒.At present,we are unable to determine the origins of the room-temperature ferroelectricity in our STO films.The strain coming from the coherent growth,shown in Fig.1,is one possible candidate.To check this possibility,we esti-mated the misfit strains of STO films grown at P O 2of 1.0and 0.2mTorr to be ϳ−3.7and −4.3ϫ10−3,respectively.Ac-cording to a thermodynamic calculation for stoichiometric SrTiO 3,7the misfit strains of our STO films are not large enough to raise their Curie temperatures to room tempera-ture.Moreover,we fabricated an ϳ30-nm-thick STO film in addition to the 100-nm-thick films used in our studies.X-ray studies showed that the thinner film has nearly the same a -and c -axes lattice constants.If the room-temperature ferroelectricity comes from the strain effects only,the thinnerfilm should have the same 2P r*values.However,as shown in Fig.4͑d ͒,its 2P r *values are smaller than those of the 100-nm-thick films by a factor of 2–3.Therefore,the strain effect alone could not be the main origin of the ferroelectricity.Other candidates for the room-temperature ferroelectric-ity are defects inside the STO films.In our STO films,we expect the formation of defects during deposition under lower P O 2.It has been reported that oxygen vacancies could increase the unit cell volume,consistent with our observations.14However,our films showed unit cell volume expansion up to 1.3%,much greater than that of SrTiO 3−␦,which is typically less than 0.3%.14In addition,other studies have reported that oxygen vacancies can induce electron doping,which makes SrTiO 3−␦metallic.14,15Our STO filmgrown at P O 2of 0.2mTorr had a resistivity of ϳ106⍀cm,which is much greater than that of metallic SrTiO 3−␦.An-other kind of defect comprises SrO vacancies.This possibil-ity is supported theoretically by formation energy calcula-tions;the energy of a SrO vacancy in SrTiO 3is only 1.53eV,which is smaller than those of other defects.16Re-cent PLD studies of STO thin films by Ohnishi et al.support the formation of SrO vacancies.17We also have optical spec-troscopy data and first-principles calculation results that are consistent with this view.18If the SrO vacancies can be formed with similar probability to oxygen vacancies,the doping effect will be minimal due to the neutral charge of the SrO vacancies leaving the STO films as insulators.Under the strain developed in our SRO/STO/SRO geometry,the SrO vacancies might be aligned and could provide defect dipoles in a given direction.19Further investigations on the forma-tion of SrO vacancies and their roles in ferroelectricity are warranted.In summary,we grew epitaxial STO thin films on SrTiO 3͑001͒substrates with SrRuO 3top and bottom elec-trodes under low oxygen pressures.The resulting tetragonal STO thin films showed ferroelectric responses even at room temperature.This interesting experimental finding is very easy to implement when making ferroelectric STO films and could have significant practical implications.The authors thank Y .J.Chang,J.Y .Jo,S.S.A.Seo,and S.H.Chang for their valuable discussions.This work was supported financially by the Creative Research Initiatives ͑Functionally Integrated Oxide Heterostructure ͒of the Korea Science and Engineering Foundation ͑KOSEF ͒.1K.A.Muller and H.Burkard,Phys.Rev.B 19,3593͑1979͒.2H.Unoki and T.Sakudo,J.Phys.Soc.Jpn.23,546͑1967͒.3W.Zhong and D.Vanderbilt,Phys.Rev.Lett.74,2587͑1995͒.4T.Mitsui and W.B.Westphal,Phys.Rev.124,1354͑1961͒.5M.Itoh,R.Wang,Y .Inaguma,T.Yamaguchi,Y .J.Shan,and T.Nakamura,Phys.Rev.Lett.82,3540͑1999͒.6P.A.Fleury and J.M.Worlock,Phys.Rev.174,613͑1968͒.7N.A.Pertsev,A.K.Tagantsev,and N.Setter,Phys.Rev.B 61,R825͑2000͒.8K.J.Choi,M.Biegalski,Y .L.Li,A.Sharan,J.Schubert,R.Uecker,P.Reiche,Y .B.Chen,X.Q.Pan,V .Gopalan,L.Q.Chen,D.G.Schlom,and C.B.Eom,Science 306,1005͑2004͒.9J.H.Haeni,P.Irvin,W.Chang,R.Uecker,P.Reiche,Y .L.Li,S.Choudhury,W.Tian,M.E.Hawley,B.Craigo,A.K.Tagantsev,X.Q.Pan,S.K.Streiffer,L.Q.Chen,S.W.Kirchoefer,J.Levy,and D.G.Schlom,Nature ͑London ͒430,758͑2004͒.10Y .S.Kim,D.H.Kim,J.D.Kim,Y .J.Chang,T.W.Noh,J.H.Kong,K.Char,Y .D.Park,S.D.Bu,J.G.Yoon,and J.S.Chung,Appl.Phys.Lett.86,102907͑2005͒.11D.H.Kim,J.S.Yang,K.W.Lee,S.D.Bu,T.W.Noh,S.J.Oh,Y .W.Kim,J.S.Chung,H.Tanaka,H.Y .Lee,and T.Kawai,Appl.Phys.Lett.81,2421͑2002͒.12R.E.Cohen,Nature ͑London ͒358,136͑1992͒.13J.Y .Son,S.H.Bang,and J.H.Cho,Appl.Phys.Lett.82,3505͑2003͒.14W.Gong,H.Yun,Y .B.Ning,J.E.Greedan,W.R.Datars,and C.V .Stager,J.Solid State Chem.90,320͑1991͒.15N.Shanthi and D.D.Sarma,Phys.Rev.B 57,2153͑1998͒.16M.J.Akhtar,Z.-U.-N.Akhtar,and R.A.Jackson,J.Am.Ceram.Soc.78,421͑1995͒.17T.Ohnishi,M.Lippmaa,T.Yamamoto,S.Meguro,and H.Koinuma,Appl.Phys.Lett.87,241919͑2005͒.18Y .S.Kim,J.Y .Kim,S.J.Moon,Y .J.Chang,W.S.Choi,J.Yu,and T.W.Noh ͑unpublished ͒.19C.H.Park andD.J.Chadi,Phys.Rev.B 57,R13961͑1998͒.FIG.4.͑Color online ͒Temperature dependence of P -V hysteresis loops in STO films grown at P O 2=1mTorr ͑a ͒and P O 2=0.2mTorr ͑b ͒.͑c ͒Tempera-ture dependence of the remnant polarization 2P r *for STO films grown at P O 2=0.2and 1mTorr.The films are 100nm thick.͑d ͒Temperaturedependence of 2P r *for 100-and 30-nm-thick STO films,grown at P O 2=0.2mTorr ͓2P r *=P r ͑+͒+P r ͑−͔͒.。