Letter pubs.acs.

org/NanoLett

Ferroelectricity in Simple Binary ZrO2 and HfO2

Johannes Muller,*, Tim S. Boscke,, Uwe Schroder,, Stefan Mueller, Dennis Brau haus,, Ulrich Bottger, Lothar Frey, and Thomas Mikolajick,#

Fraunhofer Center for Nanoelectronic Technology, Dresden, Germany Qimonda GmbH, Dresden, Germany Namlab gGmbH, Dresden, Germany RWTH Aachen, Aachen, Germany Fraunhofer Institute of Integrated Systems and Device Technology, Erlangen, Germany # Chair of Nanoelectronic Materials, University of Technology Dresden, Dresden, Germany
S Supporting Information *

ABSTRACT: The transition metal oxides ZrO2 and HfO2 as well as their solid solution are widely researched and, like most binary oxides, are expected to exhibit centrosymmetric crystal structure and therewith linear dielectric characteristics. For this reason, those oxides, even though successfully introduced into microelectronics, were never considered to be more than simple dielectrics possessing limited functionality. Here we report the discovery of a eld-driven ferroelectric phase transition in pure, sub 10 nm ZrO2 thin lms and a composition- and temperature-dependent transition to a stable ferroelectric phase in the HfO2ZrO2 mixed oxide. These unusual ndings are attributed to a size-driven tetragonal to orthorhombic phase transition that in thin lms, similar to the anticipated tetragonal to monoclinic transition, is lowered to room temperature. A structural investigation revealed the orthorhombic phase to be of space group Pbc21, whose noncentrosymmetric nature is deemed responsible for the spontaneous polarization in this novel, nanoscale ferroelectrics. KEYWORDS: Ferroelectric, hafnium oxide, zirconium oxide, phase transition, thin lm he binary oxides HfO2 and ZrO2 have been extensively studied for more than a century. Especially their martensitic phase transition from tetragonal to monoclinic and its implications on the mechanical properties of the material system have always been and still are of great scientic and commercial interest. The mechanical strain released during the volume expanding transformation to the monoclinic phase (P21/c) can be directly utilized in the transformation toughening of ceramic steel.1 To avoid this defect generating mechanism when using the pure oxides, stabilization of the high temperature polymorphs of tetragonal (P42/nmc) or cubic 24 (Fm3m ) symmetry is usually pursued. Only in recent years, driven by microelectronic scaling and the industrys strive to nd a suitable high-k replacement for integrated gate and capacitor dielectrics, extensive research has been conducted on HfO2 and ZrO2 based thin lms.5,6 In the course of this material development it was found that for thin layers in the range of several nanometers the tetragonal to monoclinic transition temperatures, as estimated from ceramic bulk samples, are signicantly lowered.79 This size-induced phase transition is frequently observed in the context of free as well as conned ZrO2 nanoparticles and was attributed to the lower surface energy developed by the high temperature polymorphs of HfO2 and ZrO2.1012 However, as in the bulk phase diagram,13 the tetragonal to monoclinic transition
2012 American Chemical Society

temperature for HfO2 is still higher compared to the one of ZrO2. In sub 20 nm lms this usually leads to tetragonal, undoped ZrO2 and partially monoclinic HfO2 thin lms14 that still require, similar to stabilized bulk ceramics, small amounts of group III,15 IV,16 or rare earth element1720 dopants to reach a full stabilization of the high temperature polymorphs. Recently, however, we reported that, until a complete stabilization of those theoretically predicted21,22 and experimentally conrmed16,23 higher-k polymorphs in HfO2 thin lms is reached, those phase transitions are accompanied by the occurrence of ferroelectricity.24 Several dopants such as Si,24,25 Y,26 and Al27 as well as the admixture of 50 mol % ZrO228 were identied to provoke ferroelectricity in HfO2 that due to its centrosymmetric phase relation (P21/c P42/nmc Fm3m) was widely believed to be paraelectric. The occurrence of ferroelectricity in binary oxides is of high scientic interest and has so far theoretically been predicted only for alkaline oxides.29 The discovery of ferroelectricity in thin lms of the broadly commercialized binary oxides of hafnium and zirconium is unexpected and aects multiple elds of application. Especially in the context of ferroelectric memories, which since their
Received: May 30, 2012 Revised: July 12, 2012 Published: July 19, 2012
4318
dx.doi.org/10.1021/nl302049k | Nano Lett. 2012, 12, 43184323

Nano Letters introduction strongly rely on the challenging integration of perovskite-based materials, these fully CMOS compatible and highly scalable HfO2-based ferroelectrics have the potential to oer a novel and much simpler approach.30,31 Here we report the discovery of a composition- and temperature-dependent ferroelectric phase transition covering the full mixing range of HfO2ZrO2 thin lms. A stable ferroelectric phase is observed for the solid solution at room temperature, whereas a eld driven transition is observed in pure ZrO2. A structural investigation by grazing incidence X-ray diraction (GI-XRD) suggests a composition-dependent tetragonal to orthorhombic to monoclinic phase change in HfO2ZrO2 thin lms when starting from Zr-rich compositions. As further revealed by in situ high temperature GI-XRD, connement of the thin lms by a TiN electrode assists in the suppression of the tetragonal to monoclinic transition during cooling, resulting in an alternative tetragonal to orthorhombic pathway. This noncentrosymmetric, orthorhombic, and thereby potentially ferroelectric Pbc21 phase of the mixed oxide was found to further stabilize with decreasing temperature. This led to a stable ferroelectric phase at 80 K in Zr-rich samples that exhibited only eld-driven ferroelectric transition at room temperature. Those nonlinear and temperature-dependent dielectric characteristics found in pure ZrO2 and in its solid solution with HfO2 conrm the existence of a tetragonal to orthorhombic phase transition and further underline the assumption that the occurrence of ferroelectricity in HfO2 and ZrO2 based thin lms is an intrinsic property of those material systems and cannot be explained by doping related defect dipoles or ionic mobility. We utilized atomic layer deposition (ALD) to prepare a set of 9 nm HfO2ZrO2 thin lms on TiN, spanning a wide composition range of the solid solution starting and ending with the pure oxides (Figure 1A). All lms crystallized during deposition of a TiN top electrode at 500 C. The polycrystalline nature was conrmed by high-resolution transmission electron microscopy (HR-TEM; Figure 1B). Polarization voltage (PV) as well as small signal capacitancevoltage (CV) characteristics of TiN-based metalinsulatormetal (MIM) capacitors of those lms are depicted in Figure 2A. Just as expected from the centrosymmetry of its monoclinic structure, the pure HfO2 lm shows a linear relation between the displacement current and the applied electric eld as well as a fairly constant capacitance in this eld range. However, as ZrO2 content increases, characteristic ferroelectric PV and CV hysteresis evolve until for a nearly equal mixture of ZrO2 and HfO2 a remanent polarization of 17 C/cm2 is reached. The coercive eld of this sample was approximately 1 MV/cm. Further increasing the ZrO2-content in the solid solution leads to a thinning of the hysteresis loop at zero bias, phenomenologically best described as superimposed antiferroelectric-like characteristics. This thinning continues until for pure ZrO2 the remanent polarization has completely vanished and only a distinct double-loop hysteresis remains. When taking a direct look at the current owing on and o the capacitor during a triangular voltage excitation, as depicted in Figure 2B, the polarization switching can clearly be separated from leakage current contributions that often lead to a confusion of simple paraelectric materials with ferroelectrics.32 A summary of the composition-dependent phase transition witnessed in Figure 2A and B is given in Figure 2C. The remanent polarization, the dielectric constant, and the monoclinic phase fraction estimated from the peak area relation
4319

Letter

Figure 1. (A) ZrO2 and HfO2 content in the HfO2ZrO2 solid solution measured by XPS plotted against the ALD pulsing ratio of the utilized alkylamide precursors, TEMAH and TEMAZ. The nearly similar growth per cycle of both precursors enables a linear and almost direct stoichiometry control by the ALD cycle ratio. (B) HR-TEM micrographs of the metalinsulatormetal capacitor used for electrical testing in this work. The polycrystalline nature of the HfO2ZrO2 thin lms as well as of the TiN metal electrodes is clearly visible.

of the coinciding tetragonal 011t and orthorhombic 111o against the monoclinic 111m reection are plotted versus the ZrO2 content. The underlying grazing-incident X-ray diraction (GIXRD) measurements are given in the Supporting Information. It is clearly observed that with the surface energy driven destabilization of the monoclinic phase toward pure ZrO2 the dielectric constant steeply increases and the remanent polarization is maximized at the phase boundary of this transition. However, when directly comparing the monoclinic phase fraction to the evolution of permittivity, one notices that the monoclinic phase has almost completely vanished above 50 mol % ZrO2 admixture while permittivity still increases toward ZrO2. This gives a rst indication that the tetragonal phase such as present in the pure ZrO2 is not immediately reached. As will be elaborated further, the phase transition is possibly bridged by an additional phase of intermediate permittivity. A eld-driven ferroelectric transition in pure ZrO2 and an adjacent ferroelectric transition with HfO2 admixture is truly surprising and has so far neither for bulk ceramics nor for thin lms of ZrO2 or HfO2ZrO2 been reported. As already mentioned, the temperature-driven and well-understood P21/c P42/nmc Fm3m transition does not allow for a stable spontaneous polarization. However, detailed GI-XRD measurements depicted in Figure 3A reveal that the assumption of a simple tetragonal to monoclinic phase transition does not prove satisfactory for the observations made in this work. As expected the pure HfO2 shows a predominantly monoclinic P21/c crystal structure, with the 111m and 111m reections being the most prominent feature in the diractogram. Likewise the structure
dx.doi.org/10.1021/nl302049k | Nano Lett. 2012, 12, 43184323

Nano Letters

Letter

Figure 2. (A) PV hysteresis at 1 kHz and small signal CV hysteresis at 10 kHz (50 mV level) of 9 nm thin HfO2ZrO2 based metalinsulatormetal capacitors at room temperature. An evolution from paraelectric HfO2 to ferroelectric HfO2ZrO2 to an antiferroelectric-like behavior in ZrO2 can be observed in PV as well as in CV characteristics. (B) Current response to a triangular voltage excitation reveals polarization switching to be clearly separable from leakage current contributions at high elds. (C) Remanent polarization, dielectric constant, and monoclinic phase fraction in the HfO2ZrO2 solid solution with respect to the mixing ratio of the oxides. With increasing ZrO2 content the dielectric constant increases due to a reduction in monoclinic phase fraction, whereas the remanent polarization is maximized in the transition region.

Figure 3. (A) GI-XRD diractograms of 9 nm ZrO2, Hf0.5Zr0.5O2, and HfO2 thin lms at an incident angle of 0.55. Starting from ZrO2 a composition-dependent tetragonal to orthorhombic to monoclinic transition is observed. Strong reections are labeled in the graph. Reference powder patterns for HfO2 and ZrO2 were calculated from the literature (P42/nmc, Mal ek et al.;33 Pbc21, Kisi et al.;37 P21/c, Ruh et al.38). (B) In situ GI-XRD measurements allow the direct observation of a suppressed tetragonal to monoclinic transformation during cooling in samples conned by a TiN electrode.

of the pure ZrO2 can be identied as the tetragonal P42/nmc phase frequently reported in literature.3336 However, it has to
4320

be noted that especially in thin lms separating the tetragonal phase with a c/a ratio close to 1 from the cubic phase proves
dx.doi.org/10.1021/nl302049k | Nano Lett. 2012, 12, 43184323

Nano Letters rather challenging, usually resulting in an ambiguous interpretability of the published results.23 In this work the absence of diractions from the TiN electrode (see sample preparation in Supporting Information) and the high signal-to-noise ratio allow for the identication of the tetragonal phase by the 012t reection at 42 originating from the slightly distorted oxygen lattice in the tetragonal modication. Unexpected diractions on the other hand appear for the equal mixture of ZrO2 and HfO2 that was found to exhibit the highest remanent polarization at room temperature. Several reections are indicated that cannot be explained by a simple phase mixture of P21/c and P42/nmc, suggesting the presence of a third phase stable during the transition from m-HfO2 to tZrO2. Similar to in the doped HfO2 systems that were found to exhibit ferroelectricity, this phase best matches the diraction pattern of an orthorhombic phase with Pbc21 symmetry.24,28,27 The existence of this noncentrosymmetric phase in ZrO2, unfortunately published without data on its dielectric properties, was rst and solely reported by Kisi et al.37 in 1989. In a neutron diraction study they observed the Pbc21 phase in small particles of Mg stabilized ZrO2 conned in a cubic matrix. A martensitic tetragonal to orthorhombic phase transition induced by a displacement of the Zr and O1 ions against the O2 ion was suggested. This tetragonal to orthorhombic transformation eventually circumvents the volume expanding tetragonal to monoclinic transformation not favorable in a highly constrained environment, such as in a cubic matrix or in our case of a thin lm additionally conned by a metal electrode. The supporting nature of the mechanical connement provided by the TiN electrode in this process can be demonstrated by in situ GI-XRD measurements recorded during cooling. As shown in Figure 3B, a thin lm crystallized into the orthorhombic phase is stable toward any further heat treatment above 800 C as long as the TiN electrode is still in place. However, if this mechanical encapsulation is removed, a transformation to the monoclinic phase can be observed, strongly increasing when temperature drops below 200 C. These ndings are in accordance to the assumption of a reversible, temperature-dependent tetragonal to orthorhombic transition. In the constrained system the orthorhombic phase transforms to the tetragonal phase during heat up and back into the orthorhombic phase during cooling. If the encapsulation is removed, the orthorhombic phase is still stable at room temperature, as depicted in Figure 2A, but after renewed heating into the tetragonal phase part of the system undergoes the well-established tetragonal to monoclinic transformation. As already mentioned, compared to this well-established transition very little research has been conducted on the less common polymorphs of HfO2 and ZrO2. Nevertheless, for pure HfO2 and ZrO2 the stability of the Pbc21 phase was calculated from ab initio by Lowther et al.,39 indicating that the relative stability of this phase with respect to the monoclinic phase lies within a narrow range of only 10 meV/atom. The calculated and experimental lattice parameters as well as our results estimated from the HfO2ZrO2 diractograms in Figure 3A are summarized in Table 1. Considering the dierent preparation methods the estimated lattice parameters of this work are in good agreement with these previous results. In this context it is interesting to add that already in 1985 Suyama et al.40 published the synthesis of an HfO2ZrO2 solid solution, which exhibited pure orthorhombic symmetry. Additionally orthorhombic phases have been observed in
4321

Letter

Table 1. Estimated Lattice Parameters of the Investigated HfO2 and ZrO2 Polymorphs Compared to Literature Data
GI-XRD, this work HfO2 space group a b c
a

ab initio simulationa HfO2 Pbc21 5.3 5.11 5.1 ZrO2 Pbc21 5.26 5.07 5.08

neutron diractionb ZrO2 Pbc21 5.26 5.07 5.08

ZrO2 P42/nmc 3.59 3.59 5.17

Hf0.5Zr0.5O2 Pbc21 5.24 5.01 5.05

P21/c 5.14 5.07 5.29

Ab initio simulation results obtained by Lowther et al.37 bA neutron diraction study conducted by Kisi et al.39

pure41 as well as Y-doped42 HfO2 thin lms grown by ALD. A complete summary of earlier sightings of orthorhombic phases in ZrO2 is further given by Heuer et al.43 However, in all cases the nonlinear dielectric properties and in most cases the exact space groups of those orthorhombic phases were not further elaborated. The temperature dependence of the ferroelectric phase stability was further investigated down to 80 K. Due to a reduced leakage current at 80 K, a fully saturated antiferroelectric-like PV hysteresis loop can be recorded for pure ZrO2 and is shown in the Supporting Information, Figure 1. To rule out a distortion of the hysteresis loop due to leakage contributions, the comparison of this low temperature hysteresis loop to hysteresis loops at elevated temperatures was done at a smaller excitation signal (Figure 4). For ZrO2 it

Figure 4. PV hysteresis of ZrO2 revealing a decreasing critical eld for polarization switching when temperature is lowered from 230 to 80 K. For the same temperature range a complete transition from an antiferroelectric-like hysteresis loop to a ferroelectric hysteresis can be observed in Zr-rich samples of the solid solution.

becomes apparent that with a decreasing temperature similar to with increasing HfO2 content the stability of the ferroelectric phase increases. In accordance with observations made on the temperature dependence of the antiferroelectric hysteresis loops in PLZTS, the critical eld for the back switching of the spontaneous polarization decreases when with decreasing temperature a ferroelectric phase is approached.44 For a Zr-rich thin lm, containing only 22 mol % HfO2, at room temperature already being comprised of a eld-driven and a stable ferroelectric phase, an almost complete transition to a purely ferroelectric phase with decreasing temperature can be observed. As assumed for the ferroelectric transition in Sidoped HfO2, the higher symmetry tetragonal structure as compared to the lower symmetry orthorhombic structure is
dx.doi.org/10.1021/nl302049k | Nano Lett. 2012, 12, 43184323

Nano Letters stabilized with increasing temperature.25 Especially for the ZrO2 system this observation is in good agreement with the reports on the appearance of an orthorhombic phase in ZrO2 at cryogenic temperatures and conrms the temperature dependence of the related tetragonal to orthorhombic transition suggested by those authors.37,45 Further measurements, especially to higher temperatures, were conducted to determine the Curie temperature of the system. The results are given in the Supporting Information. Due to high leakage contributions at elevated temperatures as well as the technical setup, measurements were limited to 400 K. Within this temperature range no paraelectric behavior in PV measurements and therewith no characteristic permittivity signatures indicating the Curie temperature were observed. Nevertheless, the characterization of the Curie temperature as well as its frequency dispersion would be of great value for a deeper understanding on the nature of the underlying phase transitions. Ferroelectricity originating from a rarely witnessed, noncentrosymmetric modication of HfO2 and ZrO2, namely, the orthorhombic Pbc21 phase, seems reasonable and has proven to be highly reproducible independent of the system investigated. Especially the ferroelectric transition in the ZrO2HfO2 solid solution described in this work underlies this assumption, since due to the equally sized and completely mixable cations no eect on the oxygen vacancy density is expected and therewith on ionic conductivity. Even though we believe that true, structure-related ferroelectricity is also responsible for the nonlinear dielectric characteristics in the trivalent and tetravalent doped HfO2 systems we reported earlier, the contribution of ionic conduction in these undersized and oversized substituted systems might not be negligible. Recent experiments using electrochemical strain microscopy revealed that the local O-vacancy distribution, density, and mobility can be precisely controlled by an electrical eld yielding a surprising electromechanical responds in otherwise nonpiezoelectric systems like, for example, yttria-stabilized ZrO2.46 The authors do not want to rule out the signicant involvement of Ovacancies to the observed phenomena, especially since the stability of those defects are believed to play an important role in the stabilization of the individual phases in HfO2 and ZrO2.47 Nevertheless, given the structural data and the compositional eects observed, we believe that the ever-present O-vacancies might be actively involved but are unlikely the root cause for the formation of the polarization in those systems. On the other hand, a true antiferroelectric nature of pure ZrO2 and its origin in the tetragonal phase remains questionable. Nevertheless, our earlier work indicates that the appearance of this eld-driven transition is closely linked to the appearance of the tetragonal phase. This antiferroelectric-like behavior was observed in Si-24 and Al-doped-27 HfO2 thin lms as well. Both systems are stabilized into the tetragonal P42nmc structure with sucient dopant content. The same is true for the pure ZrO2 layers investigated in this work. Only ferroelectric, Y-doped HfO226 did not exhibit such a elddriven transition and as proven by experiments48 as well as from a theoretical point of view47 favors the cubic structure over the tetragonal at high doping levels. In the context of the tetragonal to orthorhombic phase transition described earlier, this leads to the assumption that this transition is not only temperature-dependent but can further be induced by an electric eld alone. However, without further experiments as already suggested earlier, the possibility of a relaxor ferro4322

Letter

electric, resulting in equally behaved antiferroelectric-like hysteresis loops, cannot be ruled out.49,50 In conclusion, a composition- and temperature-dependent ferroelectric phase transition was observed in thin lms of the HfO2ZrO2 solid solution. Hf-rich samples exhibited a predominantly monoclinic structure, whereas Zr-rich samples were crystallized into a tetragonal phase. Ferroelectricity appeared in the mixed oxide and was attributed to a tetragonal to orthorhombic transformation during cooling resulting in a noncentrosymmetric, orthorhombic phase of the space group Pbc21. A eld-driven and temperature dependent ferroelectric transition was observed in pure ZrO2 and in the Zr-rich part of the phase diagram suggesting a high reversibility of this tetragonal to orthorhombic transformation. These ndings shed new light on ZrO2, a dielectric material extensively used in DRAM storage nodes and embedded capacitors, expected to exhibit linear dielectric functionality.

S Supporting Information *

ASSOCIATED CONTENT

Detailed description of characterization methods and sample preparation, GI-XRD measurements of the full HfO2ZrO2 composition range, and supporting low and high temperature PV-hysteresis and CV measurements of the HfO2ZrO2 system. This material is available free of charge via the Internet at http://pubs.acs.org.

AUTHOR INFORMATION

Corresponding Author Present Addresses

*E-mail: johannes.mueller@ieee.org. Bosch Solar Energy AG, Erfurt, Germany Aixtron, Herzogenrath, Germany

Notes

The authors declare no competing nancial interest. T.B. and U.S. were with Qimonda Dresden during the initial stage of the work.

ACKNOWLEDGMENTS We would like to thank Marcus Mildner for preparing the HRTEM micrographs. The work for this Letter was supported within the scope of technology development by the EFRE fund of the European Community and by funding of the Free State of Saxony (Project MERLIN). The authors are responsible for the content of the paper. REFERENCES
(1) Garvie, R. C.; Hannink, R. H. J.; Pascoe, R. T. Nature 1975, 258, 703. (2) Weiss, L. Z. Anorg. Chem. 1910, 65 (1), 178227. (3) Curtis, C. E. J. Am. Ceram. Soc. 1947, 30 (6), 180196. (4) Curtis, C. E.; Doney, L. M.; Johnson, J. R. J. Am. Ceram. Soc. 1954, 37 (10), 458465. (5) Choi, J. H.; Mao, Y.; Chang, J. P. Mater. Sci. Eng. R 2011, 72 (6), 97136. (6) Kittl, J. A.; Opsomer, K.; Popovici, M.; Menou, N.; Kaczer, B.; Wang, X. P.; Adelmann, C.; Pawlak, M. A.; Tomida, K.; Rothschild, A.; Govoreanu, B.; Degraeve, R.; Schaekers, M.; Zahid, M.; Delabie, A.; Meersschaut, J.; Polspoel, W.; Clima, S.; Pourtois, G.; Knaepen, W.; Detavernier, C.; Afanasev, V. V.; Blomberg, T.; Pierreux, D.; Swerts, J.; Fischer, P.; Maes, J. W.; Manger, D.; Vandervorst, W.; Conard, T.; Franquet, A.; Favia, P.; Bender, H.; Brijs, B.; van Elshocht, S.; Jurczak, M.; van Houdt, J.; Wouters, D. J. Microelectron. Eng. 2009, 86 (79), 17891795.
dx.doi.org/10.1021/nl302049k | Nano Lett. 2012, 12, 43184323

Nano Letters
(7) Ushakov, S. V.; Navrotsky, A.; Yang, Y.; Stemmer, S.; Kukli, K.; Ritala, M.; Leskela, M. A.; Fejes, P.; Demkov, A.; Wang, C.; Nguyen, B. Y.; Triyoso, D.; Tobin, P. Phys. Status Solidi B 2004, 241 (10), 2268 2278. (8) Boscke, T. S.; Hung, P. Y.; Kirsch, P. D.; Quevedo-Lopez, M. A.; Ramirez-Bon, R. Appl. Phys. Lett. 2009, 95 (5), 0529043. (9) Muller, J.; Boscke, T. S.; Schroder, U.; Reinicke, M.; Oberbeck, L.; Zhou, D.; Weinreich, W.; Kucher, P.; Lemberger, M.; Frey, L. Microelectron. Eng. 2009, 86 (79), 18181821. (10) Shen, P.; Lee, W. H. Nano Lett. 2001, 1 (12), 707711. (11) Shukla, S.; Seal, S.; Vij, R.; Bandyopadhyay, S.; Rahman, Z. Nano Lett. 2002, 2 (9), 989993. (12) Navrotsky, A. J. Mater. Chem. 2005, 15, 18831890. (13) Ruh, R.; Garrett, H. J.; Domagala, R. F.; Tallan, N. M. J. Am. Ceram. Soc. 1968, 51 (1), 2328. (14) Kim, H.; McIntyre, P. C.; Saraswat, K. C. J. Mater. Res. 2004, 19 (2), 643650. (15) Park, P. K.; Kang, S.-W. Appl. Phys. Lett. 2006, 89 (19), 192905. (16) Tomida, K.; Kita, K.; Toriumi, A. Appl. Phys. Lett. 2006, 89 (14), 142902. (17) Kita, K.; Kyuno, K.; Toriumi, A. Appl. Phys. Lett. 2005, 86 (10), 102906. (18) Ushakov, S. V.; Brown, C. E.; Navrotsky, A. J. Mater. Res. 2004, 19 (3), 693696. (19) Losovyj, Y. B.; Ketsman, I.; Sokolov, A.; Belashchenko, K. D.; Dowben, P. A.; Tang, J.; Wang, Z. Appl. Phys. Lett. 2007, 91 (13), 132908. (20) Wiemer, C.; Lamagna, L.; Baldovino, S.; Perego, M.; SchammChardon, S.; Coulon, P. E.; Salicio, O.; Congedo, G.; Spiga, S.; Fanciulli, M. Appl. Phys. Lett. 2010, 96 (18), 182901182903. (21) Zhao, X.; Vanderbilt, D. Phys. Rev. B 2002, 65 (23), 233106. (22) Zhao, X.; Vanderbilt, D. Phys. Rev. B 2002, 65 (7), 75105. (23) Kim, S. K.; Hwang, C. S. Electrochem. Solid-State Lett. 2008, 11 (3), G9G11. (24) Boscke, T. S.; Muller, J.; Brau haus, D.; Schroder, U.; Bottger, U. Appl. Phys. Lett. 2011, 99 (10), 102903. (25) Boscke, T. S.; Teichert, S.; Brau haus, D.; Muller, J.; Schroder, U.; Bottger, U.; Mikolajick, T. Appl. Phys. Lett. 2011, 99 (11), 112904. (26) Muller, J.; Schroder, U.; Boscke, T. S.; Muller, I.; Bottger, U.; Wilde, L.; Sundqvist, J.; Lemberger, M.; Kucher, P.; Mikolajick, T.; Frey, L. J. Appl. Phys. 2011, 110 (11), 114113. (27) Muller, S.; Muller, J.; Singh, A.; Riedel, S.; Sundqvist, J.; Schroder, U.; Mikolajick, T. Adv. Funct. Mater. 2012, 22 (11), 2412 2417. (28) Muller, J.; Boscke, T. S.; Brau haus, D.; Schroder, U.; Bottger, U.; Sundqvist, J.; Kucher, P.; Mikolajick, T.; Frey, L. Appl. Phys. Lett. 2011, 99 (11), 112901. (29) Bousquet, E.; Spaldin, N. A.; Ghosez, P. Phys. Rev. Lett. 2010, 104 (3), 37601. (30) Boscke, T. S.; Muller, J.; Brau haus, D.; Schroder, U.; Bottger, U. Tech. Dig. IEDM 2011, 547550. (31) Muller, J.; Boscke, T. S.; Schro der, U.; Hoffmann, R.; Mikolajick, T.; Frey, L. IEEE Electron Device Lett. 2012, 33 (2), 185. (32) Scott, J. F. J. Phys.: Condens. Matter. 2008, 20 (2), 21001. (33) Mal ek, J.; Benes, L.; Mitsuhashi, T. Powder Diffraction 1997, 12 (2), 9698. (34) Heuer, A. H.; Ruhle, M.; Marshall., D. B. J. Am. Ceram. Soc. 1990, 73, 10841093. (35) Bouvier, P.; Djurado, E.; Lucazeau, G.; Le Bihan, T. Phys. Rev. B 2000, 62 (13), 8731. (36) Kisi, E. H.; Howard, C. J. Key Eng. Mater. 1998, 153 - 154, 1 36. (37) Kisi, E. H.; Howard, C. J.; Hill, R. J. J. Am. Ceram. Soc. 1989, 72 (9), 17571760. (38) Ruh, R.; Corfield, P. W. R. J. Am. Ceram. Soc. 1970, 53 (3), 126129. (39) Lowther, J. E.; Dewhurst, J. K.; Leger, J. M.; Haines, J. Phys. Rev. B 1999, 60 (21), 14485.
4323

Letter

(40) Suyama, R.; Takubo, H.; Kume, S. J. Am. Ceram. Soc. 1985, 68 (9), C-237C-237. (41) Aarik, J.; Aidla, A.; Kiisler, A.-A.; Uustare, T.; Sammelselg, V. Thin Solid Films 1999, 340 (12), 110116. (42) Dubourdieu, C.; Rauwel, E.; Roussel, H.; Ducroquet, F.; Hollander, B.; Rossell, M.; van Tendeloo, G.; Lhostis, S.; Rushworth, S. J. Vac. Sci. Technol. A 2009, 27 (3), 503514. (43) Heuer, A. H.; Lanteri, V.; Farmer, S. C.; Chaim, R.; Lee, R. R.; Kibbel, B. W.; Dickerson, R. M. J. Mater. Sci. 1989, 24 (1), 124132. (44) Shebanov, L.; Kusnetsov, M.; Sternberg, A. J. Appl. Phys. 1994, 76 (7), 43014304. (45) Marshall, D. B.; Jarnes, M. R.; Porter, J. R. J. Am. Ceram. Soc. 1989, 72 (2), 218227. (46) Kumar, A.; Ciucci, F.; Morozovska, A. N.; Kalinin, S. V.; Jesse, S. Nat. Chem. 2011, 3 (9), 707713. (47) Lee, C.-K.; Cho, E.; Lee, H.-S.; Hwang, C. S.; Han, S. Phys. Rev. B 2008, 78 (1), 12102. (48) Rauwel, E.; Dubourdieu, C.; Hollander, B.; Rochat, N.; Ducroquet, F.; Rossell, M. D.; van Tendeloo, G.; Pelissier, B. Appl. Phys. Lett. 2006, 89 (1), 12902. (49) Chu, F.; Reaney, I. M.; Setter, N. J. Appl. Phys. 1995, 77 (4), 16711676. (50) Chu, F.; Setter, N.; Tagantsev, A. K. J. Appl. Phys. 1993, 74 (8), 51295134.

dx.doi.org/10.1021/nl302049k | Nano Lett. 2012, 12, 43184323