Crystal Isomers of ScFeO3 - Crystal Growth & Design (ACS Publications)


Crystal Isomers of ScFeO3 - Crystal Growth & Design (ACS Publications)pubs.acs.org/doi/abs/10.1021/acs.cgd.6b00770Ca...

0 downloads 185 Views 6MB Size

Article pubs.acs.org/crystal

Crystal Isomers of ScFeO3 Yosuke Hamasaki,† Takao Shimizu,‡ Shintaro Yasui,*,† Tomoyasu Taniyama,† Osami Sakata,‡,§,∥ and Mitsuru Itoh*,† †

Laboratory for Materials and Structures, Institute for Innovative Research, ‡Center for Elemental Research, and ∥Department of Materials Science and Engineering, School of Materials and Chemical Technology, Tokyo Institute of Technology, 4259, Nagatsuta, Midori-ku, Yokohama, 226-8503 Japan § Synchrotron X-ray Station at SPring-8, National Institute for Materials Science, 1-1-1 Kouto, Sayocho, Sayogun, Hyogo 679-5148 Japan S Supporting Information *

ABSTRACT: In inorganic compounds, “crystal isomers”, which can exist in metastable phases, are obtained by various solution-processing techniques, highpressure syntheses, as well as physical and chemical thin film fabrication techniques. The metastable phase depends on the processing, allowing the hierarchy of the Gibbs free energy to be controlled in a phase at a given temperature. In this study, we successfully stabilize five metastable phases, four phases of ScFeO3 and one Sc0.48Fe1.52O3, prepared from one ScFeO3 target by the pulsed laser deposition technique. The crystal structures are identified by X-ray diffraction and high-angle annular dark field-scanning transmitted electron microscopy measurements. The relationship between the crystal structure of the film and the substrate is κ-Al2O3-type Sc0.48Fe1.52O3 on SrTiO3(111), spinel-type ScFeO3 on MgO(001), corundum-type ScFeO3 on Fe2O3/Al2O3(0001) and NdCaAlO4(001), YMnO3-type ScFeO3 on Al2O3(0001), and bixbyite-type ScFeO3 on YSZ(001). Four of these structures (all except the bixbyite structure) have not been reported by other processing techniques. These results suggest that the thin film growth technique is a strong tool for exploring novel functional materials and the metastable phases of oxide isomers.



INTRODUCTION Structural isomerism is defined as compounds with the same chemical formula but different crystal structures, symmetries, and atomic arrangements. Covalently bonded organic compounds have many structural and stereoisomers because the four valence electrons of carbon in organic compounds allow branching in the carbon chain and structurally different positions of atoms. On the other hand, nondirectional and ionically bonded inorganic compounds tend to have relatively higher symmetrical structures with a smaller number of constitutional atoms in the unit cell compared to organic compounds. Most simple metal oxides, except for compounds with three or four coordination numbers such as borates and silicates, have stronger ionic characters and fewer crystal isomers. Complex oxides, which are composed of multiple cations, generally have uneven chemical bonds with anions to yield a uniform coordination number of cations. Crystal isomers (≈ crystal polymorphs) in complex oxides are usually available by controlling the temperature, crystal size, and cell volume. Temperature controls the successive phase transition of BaTiO3 from cubic at high temperature to a rhombohedral structure through tetragonal and orthorhombic structures.1 Furthermore, BaTiO3 can be transformed from a perovskite-type structure to a hexagonal-type structure with a drastic rearrangement of atoms above 1460 °C. Meanwhile, © 2016 American Chemical Society

MgSiO3 transforms from an orthoenstatite structure through a clinoenstatite, β-Mg2SiO4 + stishovite, γ-Mg2SiO4 + stishovite, ilmenite, and perovskite to eventually realize a postperovskite structure by applying pressure.2,3 This leads to a reduction in the cell volume and a change in the coordination number. A relative increase in the surface energy by reducing the size sometimes inverts the hierarchy of the appearance of relevant metastable phases. Simple oxides of Al2O34 and Fe2O35 have many metastable phases of γ, η, δ, θ, θ′, θ″, κ, ... and β, γ, ε, perovskite, postperovskite, Rh2O3(II), ..., respectively. However, freezing the desired metastable phases is general challenging. Actually, only a few studies have reported the formation of more than two metastable phases in the film form with the same composition. Thin film fabrication techniques have contributed to the development of novel functional electronic materials and devices. Epitaxial films deposited on single crystal substrates suffering from interfacial strain due to a mismatch in the lattice constants between the film and the substrate allow the physical properties, such as electrical conductivity, ferroelectricity,6,7 and magnetism,8 to be tuned using so-called strain engineering. In Received: May 22, 2016 Revised: July 20, 2016 Published: July 29, 2016 5214

DOI: 10.1021/acs.cgd.6b00770 Cryst. Growth Des. 2016, 16, 5214−5222

Crystal Growth & Design

Article

Figure 1. Phase relationship plotted as a function of the ionic radius for (a) A3+FeO3 and (b) A3+2O3.

bixbyite-type and corundum-type structures coexist for the composition of ScFeO3.16 Bréard et al. obtained a single phase of the bixbyite-type ScFeO3 using a conventional solid-state reaction technique.17 The perovskite-type ScFeO3 is only stable under 15 GPa and above 800 °C. As it decompresses, it transforms into the LiNbO3-type structure.18 Li et al. also reported that polar corundum ScFeO3 can be synthesized under 6 GPa and 1500 °C.19 A YMnO3-type REFeO3 has been reported as a metastable phase for several rare earth ions, through various synthesizing routes of spray ICP technique for RE = Eu and Yb,20 partial melting for RE = In,21 containerless melt crystallization process for RE = Y, Sm, Eu, Gd, Dy, Er, Yb, and Lu,22,23 and crystallization from amorphous phase for RE = Y.24 For Al3+ and Ga3+ with small ionic radii, the κ-Al2O3-type structure of GaFeO325 can appear in the equilibrium phase diagram, and AlFeO326 in the high-temperature phase above 1300 °C. Figure 1a summarizes the reported phases with respect to the ionic radii of six-coordinated A3+ ions, while Figure 1b plots the crystal structures for A3+2O3. It is known that α-Al2O3 and α-Fe2O3 with corundum-type structures are in stable phases at ambient conditions, while γ-Al2O3 and γ-Fe2O3 with spinel-type structures are metastable phases. κ-Al2O3 and ε-Fe2O3 described as possessing the GaFeO3-type structure are also metastable. Sakurai reported that as the particle size increases, the crystal structure changes in the sequence γ-Fe2O3 → ε-Fe2O3 → β-Fe2O3 → α-Fe2O3.27 A Rh2O3(II)-type structure appears in Fe2O3 and Al2O3 under pressures above 50 GPa.28,29 Sc2O3 and In2O3 with relatively larger ionic radii have a bixbyite-type structure as a stable phase. This structure has been reported for Fe2O3 as β-Fe2O3. Corundum-type In2O3 can be synthesized under high pressure above 12.8 GPa.30 As seen in Figure 1a,b, Sc2O3 and ScFeO3 are located at the cross point of various kinds of crystal structures, indicating that the combination of Sc3+ and Fe3+ may be suitable to test the formation of metastable structures (i.e., crystal isomers) due to the criticality of ScFeO3. We anticipated that the six crystal isomers (corundum, bixbyite, spinel, κ-Al2O3-type, YMnO3-type, and LiNbO3-type structures) (Figure 2) would appear if the deposition conditions of ScFeO3 epitaxial films on various kinds of single crystal substrates control the growth energy.

particular, strain engineering has been successfully demonstrated in perovskite-related compounds;9 many kinds of perovskite-type substrates are commercially available. These have lattice constants that range from a = 3.68 Å for NdCaAlO4 to ap = 4.01 Å for NdScO3, where ap is the lattice parameter of the pseudocubic cell. Strain engineering has also achieved great success in stabilizing metastable phases and investigating their physical properties. Examples include the metal−insulator transition in perovskite-type RENiO310 (RE = rare earth atom) as well as ferroelectricity and ferromagnetism in perovskite-type BiMnO3,11 which is stable only under high pressure. This means that the epitaxial growth technique could modify the hierarchy of metastable phases through the interaction between the film and the substrate. The type of substrate and the magnitude of the lattice mismatch between the film and the substrate from a chemical and physical point of view, respectively, may be dominant factors in stabilizing metastable crystal structures. In this study, we attempt to establish “crystal isomers” via the thin film fabrication technique. Perovskite-structured ABO3 materials have been intensively studied due to their diverse physical properties. In the inorganic chemistry field, the combination of cation A with a relatively large ionic radius and cation B with a smaller ionic radius tends to form a perovskite structure. In ABO3 oxide compounds, a tolerance factor, t = (rA + rO)/√2(rB + rO), where rA, rB, and rO are the ionic radii of the A and B cations and the O anion (oxygen), respectively,12,13 is a scale to grasp the yielded structures; t = 1 cubic perovskite, t > 1 tetragonal or hexagonal perovskite, and t < 1 is an orthorhombic or rhombohedral perovskite. In the case of an even smaller t value, which consists of relatively similar ionic radii of rA and rB in A3+B3+O3, sesquioxide structures, including corundum, bixbyite, etc. appear. On the other hand, the ilmenite- and LiNbO3-type structures form in A2+B4+O3 and A1+B5+O3.14 This fact suggests that the difference in ionic radii plays an important role in forming an ordered structure of cations in A3+B3+O3, while the difference in the cation valences strongly affects the crystal structure. Hence, it is estimated that the structural change from the ABO3 perovskite structure to the (A,B)2O3 sesquioxide structure will pass through a “cross point” with a decrease in t. Here, we focus on iron-based oxides because simple iron oxides have four structures,4 and many iron-based perovskite have been reported.15 When we employ A3+ cations with large ionic radii (such as rare earth metal ions), the perovskite-type structure is stable, except for the relatively small Sc3+. According to the equilibrium phase diagram of the Sc2O3−Fe2O3 system,



EXPERIMENTAL METHODS

ScFeO3 thin films were prepared at an optimized growth temperature on various single crystals using a pulsed laser deposition technique, 5215

DOI: 10.1021/acs.cgd.6b00770 Cryst. Growth Des. 2016, 16, 5214−5222

Crystal Growth & Design

Article

SrTiO3(111) substrate shows 00l (l = 2n) and 20l (l = m) diffractions, which correspond to the κ-Al2O3-type structure (Figure 3a). The in-plane relationship between the

Figure 2. Schematic illustration of the crystal structures: (a) Bixbyitetype (Ia3̅), (b) corundum-type (R3̅c), (c) spinel-type (Fd3̅m), (d) κAl2O3-type (Pna21), (e) YMnO3-type (P63cm), and (f) LiNbO3-type (R3c).

Figure 3. (a) Wide-range XRD-RSM of the Sc0.48Fe1.52O3 film on SrTiO3(111) along SrTiO3(111) [112]̅ . Red and white indexes indicate the SrTiO3 substrate and the ScFeO3 film, respectively. (b) HAADF-STEM image of the Sc0.48Fe1.52O3 film on SrTiO3(111). View direction is SrTiO3(111) [112̅]. Circles and broken lines denote simulated cationic positions and unit cells, respectively. (c) Schematic atomic arrangement of the κ-Al2O3-type structure viewed along the [100] direction. Black lines indicate a unit cell.

which utilized the fourth-harmonic wave of a Nd: YAG laser (λ = 266 nm) with a repetition rate of 5 Hz and an optimized energy density. The target was prepared by mixing bixbyite-type Sc2O3 (Rare Metallic, 99.9%) and α-Fe2O3 (Rare Metallic, 99.99%) powders, and then sintering at 1400 °C for 14 h in air. Powder X-ray diffraction (XRD) analysis for the target was a mixture of bixbyite- and corundum-type solid solutions, which was consistent with the phase diagram.16 The crystal structure of the ScFeO3 films was investigated by highresolution X-ray diffraction (HRXRD) using a Rigaku Smart-Lab diffractometer with CuKα1 monochromatic radiation (λ = 1.54059 Å) by Ge(220) two-bounce crystals. Wide-range XRD reciprocal space mappings (RSMs) were obtained by a 2D X-ray detector (PILATUS). The chemical compositions of the ScFeO3 films were measured by wavelength dispersion X-ray spectroscopy (WDX, PANalitical PW2404). Standard samples of Sc2O3 and Fe2O3 films were used. The atomic arrangement of the ScFeO3 films was observed by highangle annular dark field-scanning transmitted electron microscopy (HAADF-STEM) using JEOL ARM200F. Transmission electron microscopy (TEM) specimens were prepared using a focused ion beam (FIB) technique. Synchrotron XRD (SXRD) was performed at SPring-8 BL15XU using a 12.4 keV energy wave.

Sc0.48Fe1.52O3 film and the substrate was revealed by the phiscan recorded around Sc0.48Fe1.52O3 {209} and SrTiO3 {110}. The peaks of Sc0.48Fe1.52O3 {209} show a 6-fold symmetry rotated 60° from each other, and the peaks of SrTiO3 {110} have a 3-fold symmetry (Figure S1). These results indicate that the Sc0.48Fe1.52O3 film has three in-plane domains rotated 120° with an in-plane relationship of Sc 0.48 Fe 1.52 O 3 {100}/ SrTiO3{112̅}. The domain structure of the Sc0.48Fe1.52O 3 film on SrTiO3(111) is the same as the result of our previous report of an epitaxial AlFeO3 film on a SrTiO3(111) substrate.33 The lattice constants of Sc0.48Fe1.52O3 are a = 5.145 Å, b = 8.865 Å, and c = 9.637 Å, which are larger than those for ε-Fe2O3 (a = 5.1019 Å, b = 8.7807 Å, and c = 9.4661 Å).35 This is consistent with the fact that the ionic radius of Sc is larger than that of Fe and includes the incorporation of Sc2O3 in the ε-Fe2O3 lattice. The atomic arrangement of κ-Al2O3-type ScFeO3 film was investigated by HAADF-STEM (Figure 3b,c). The HAADFSTEM image along SrTiO3(111) [112̅] clearly indicates that two- or four-atom groups form each row and stack alternately along the c-axis direction. The two-atom groups correspond to paired tetrahedral or octahedral cations, whereas the four-atom groups are cations in four-corner shared octahedral sites. At the film and substrate interface, an interfacial bottom layer about 2 nm thick is observed. This layer shows a similar κ-Al2O3-type atomic pattern. However, there are mirror planes between two cation layers. These mirror planes could not be observed in the κ-Al2O3-type structure because the nonpolar structure excludes such a symmetric element. From these results, we conclude that the crystal structure of the Sc0.48Fe1.52O3 film on the SrTiO3(111) substrate has a κAl2O3-type structure. Spinel-Type Structure on MgO(001). The general formula of spinel-type compounds is AB2O4, where A and B are tetrahedrally and octahedrally coordinated cations to oxygen, respectively. Compounds with A2O3 also have spineltype structure (e.g., γ-Fe2O3 and γ-Al2O3) and vacancies



RESULTS κ-Al2O3-Type Structure on SrTiO3(111). The κ-Al2O3type structure with SG Pna21 has four cation sites: three octahedral sites and one tetrahedral site. Close-packed oxygen layers are stacked along the c-axis with a sequence of ABACABAC···, and the cations are distributed over the octahedral + tetrahedral cation sites and two octahedral cation sites in each interoxygen layer.31 The stable phase of ABO3 with a κ-Al2O3-type structure has been reported only for GaFeO3. Compounds with the κ-Al2O3-type structure such as GaFeO3, AlFeO3, ε-Fe2O3, and κ-Al2O3 have been reported in the form of a film deposited on an SrTiO3(111) substrate.32−34 The similarity of the atomic arrangements might lead to the stabilization of the κ-Al2O3-type structure on SrTiO3(111). In our previous study, an ε-Fe2O3 film on SrTiO3(111) was obtained only under limited conditions with an ablation energy smaller than 0.32 J cm−2. Subsequently, we attempted to stabilize the κ-Al2O3-type ScFeO3 on a SrTiO3(111) substrate based on the summary in Figure 1. WDX for the obtained film reveals that the ratio of Fe/Sc is 76:24. Compared to the other structures described in the later sections, the chemical composition of the κ-Al2O3-type film markedly deviates from the nominal composition. The Sc0.48Fe1.52O3 film on the 5216

DOI: 10.1021/acs.cgd.6b00770 Cryst. Growth Des. 2016, 16, 5214−5222

Crystal Growth & Design

Article

occupying the octahedral cation sites [i.e., AB5/3□1/3O4 (□; vacancy)]. From the viewpoint of anion packing, a spinel-type structure has stacking similar to that of a NaCl-type structure. Specifically, the spinel-type structure has the doubled-lattice constants of the NaCl-type structure. To stabilize the spineltype structure, we selected NaCl-type MgO(001) and spineltype MgAl2O4(001) substrates. ScFeO3 films were deposited on those substrates under the same conditions as the bixbyite-type ScFeO3 film. The wide-range XRD-RSM for the ScFeO3 film on MgO was performed along the MgO(001) [100] direction. Figure 4a

The lattice mismatch was estimated using the lattice constants of γ-Fe2O3 (a = 8.351 Å). The lattice mismatches between γ-Fe2O3 and the substrate are 3.89% for MgAl2O4 and −1.19% for MgO. Here since the ionic radius of Sc3+ is larger than that of Fe3+, the lattice constant of the defect-spinel-type ScFeO3 can also be larger than that of γ-Fe2O3. Thus, the lattice mismatch between the defect-spinel-type ScFeO3 and the substrate may be larger than that between γ-Fe2O3 and the substrates. The larger lattice mismatch suggests that the defectspinel-type ScFeO3 is not stabilized on the MgAl2O4 substrate. We noted that the difference in the crystal structures between the film and the substrate is not crucial to stabilize the metastable phase. However, the lattice mismatch and the structural similarity of the atomic arrangement and stacking are key points in the metastable phase stabilization. Corundum-Type Structure on Fe 2 O 3 (0001)/ Al2O3(0001). The corundum structure belongs to SG R3̅c. Closed-packed oxygen layers are stacked along the c-axis in the sequence ABAB···. A metastable phase with a corundum structure of Ga2O3 and In2O3 has been stabilized on Al2O3(0001).36,37 As explained in the previous section, the crystal structure of the ScFeO3 film on the Al2O3(0001) substrate is not the corundum-type but the YMnO3-type. The in-plane lattice mismatch between corundum-type Fe2O3 (a = 5.036 Å and c = 13.749 Å) and Al2O3 (a = 4.77 Å and c = 13.04 Å) is 5.58%. Therefore, the lattice mismatch between the corundum-type ScFeO3 and the Al2O3 substrate exceeds 5.58%. This large mismatch may be responsible for the stabilization of the YMnO3-type structure rather than the corundum-type ScFeO3 on the Al2O3(0001) substrate. To reduce the lattice mismatch, we chose Fe2O3 for the buffer layer with the corundum structure. Consequently, a ScFeO3 film was grown on an epitaxially deposited Fe2O3 buffer layer over the Al2O3(0001) substrate. Fe2O3 and ScFeO3 films were deposited under the same condition as the bixbyite ScFeO3 film described in the previous section. The diffraction spots of ScFeO3 film revealed by the widerange XRD-RSM show the exact same trends as those of Fe2O3 buffered film and Al2O3 substrate (Figure 5a), indicating that the corundum-type ScFeO3 film is epitaxially grown on the

Figure 4. (a) Wide-range XRD-RSM of ScFeO3 film on MgO(001) along MgO(001) [100]. Red and white indexes show the MgO substrate and the ScFeO3 film, respectively. (b) HAADF-STEM image of the ScFeO3 film on MgO(001). View direction is MgO(001) [110]. Circles and broken lines indicate the simulated cationic positions and the unit cell of spinel-type structure, respectively. (c) Schematic atomic arrangement of the spinel-type structure viewed along the [110] direction. Black lines mark the unit cell. Red and blue spheres represent cations in the 4-fold symmetry and 6-fold symmetry sites, respectively.

shows that only h0l (h = 2n and l = 2m) reflections appear. The 202 reflection indicates that the lattice constant of the ScFeO3 film is almost double that of the MgO substrate. Consequently, the crystal structure of the ScFeO3 film is the spinel-type. In the wide-range XRD-RSM along MgO(001) [110], hhl (h and l = 2n, 2m or 2n+1, 2m+1) corresponding to the super lattice peaks is observed (Figure S2). The lattice constant of the spinel-type ScFeO3 film calculated from out-of-plane patterns is 8.64 Å. It is difficult to estimate the in-plane lattice constant from the wide-range XRD-RSM because the two lattice constants between spinel-type ScFeO3 and MgO are almost the same. Thus, to estimate the in-plane lattice constant, we also measured the HRXRD-RSM around the MgO 204 reflection (Figure S3). This result implies that the in-plane lattice constant of the ScFeO3 film is the same as that of the MgO substrate (a = 4.21 Å), indicating that the ScFeO3 film has a tetragonal lattice. In the HAADF-STEM measurement viewed along the MgO(001) [110] direction (Figure 4b,c), the atomic arrangement of the ScFeO3 film on the MgO substrate clearly matches that of a spinel-type structure coherently, where red and blue spheres are cations in the 4-fold symmetry site and the 6-fold symmetry site, respectively. In addition, the WDX result suggests that the ratio of Fe and Sc is 57:43. These results show that the defect-spinel-type ScFeO3 film is epitaxially grown on the MgO substrate. We also deposited a ScFeO3 film on the MgAl2O4(001) substrate (a = 8.083 Å), but the spinel-type structure did not appear.

Figure 5. (a) Wide-range XRD-RSM of a ScFeO3 film on Fe2O3/ Al2O3 (0001) along Fe2O3/Al2O3(0001) [101̅0]. Red and white indexes indicate the Al2O3 substrate and the ScFeO3 film, respectively. (b) HAADF-STEM image of the ScFeO3 film on Fe2O3/Al2O3 (0001). View direction is Al2O3(0001) [101̅0]. Circles and broken lines indicate the simulated cationic positions and the unit cell of the corundum-type structure, respectively. (c) Schematic atomic arrangement of the corundum-type structure viewed along the [101̅0] direction. Black lines denote the unit cell. 5217

DOI: 10.1021/acs.cgd.6b00770 Cryst. Growth Des. 2016, 16, 5214−5222

Crystal Growth & Design

Article

Fe2O3/Al2O3 substrate. The lattice constants of the corundumtype ScFeO3 are a = 5.151 Å and c = 13.991 Å. The lattice mismatches between the corundum-type ScFeO3 and the Al2O3 substrate or Fe2O3 are 7.99% and 2.28%, respectively. The small mismatch of the latter relationship may stabilize the corundum-type ScFeO3. It is generally known that there are two structures relevant to the corundum structure: the ilmenitetype and the LiNbO3-type. These structures are distinguished by the ordering manner of the cations. In the case of corundum-type structures, the stacking sequence of two cations, A and B, along the c-axis is (A, B)-(A, B)-□-(A, B)(A, B), where A, B, (A, B) and □ correspond to the A-site, Bsite, disordered A and B, and a vacancy, respectively. The stacking sequences along the c-axis of the ilmenite-type and LiNbO3-type structures are A-B-□-B-A and A-B-□-A-B, respectively. In the XRD patterns, the ilmenite-type structure should show 0003 and 0009 reflections, whereas these diffractions are forbidden in the corundum-type and LiNbO3-type structures. However, it is difficult to distinguish between the corundum and LiNbO3-type structures by XRD measurements. Thus, we measured the arrangement of cations in the film observed by HAADF-STEM (Figure 5b,c). In the HAADF-STEM image with the zone axis of Al2O3(0001) [101̅0], bright spots correspond to the two cation sites, which have equal intensities. In the case of the LiNbO3-type structure, one should be a bright spot and the other should be dark.18 We also performed STEM-EDX analysis for the two spots (Figure S4), which revealed that the Fe and Sc ratio is 57:43 (Table S1) and the Fe and Sc atoms are disordered at the two cation sites. Therefore, we concluded that the crystal structure of ScFeO3 film deposited on the Fe2 O3/Al2O 3(0001) substrate is the corundum-type. Corundum-Type Structure on NdCaAlO4(001). In terms of the rotation (Φ) of the BO6 octahedra along the perovskite [111] direction, the LiNbO3-type structure can be described as a distorted perovskite with a tilting system of a−a−a−.38 Perovskite with Φ = 0° and 0 < Φ < 15° are regarded as the cubic and rhombohedral structures, respectively. In the case of Φ > 20°, the crystal structure is the LiNbO3-type.39 It should be noted that LiNbO3(012) corresponds to pseudocubic perovskite (001). LiNbO3 and LiTaO3 were epitaxially grown on perovskite-type substrates with the relationship of LiNbO3(012)//SrTiO3(001) and LiTaO3(012)// SrTiO3(001),40 respectively. To stabilize the LiNbO3-type structure, we deposited ScFeO3 film on SrTiO3(001), LaSrAlO4(001), and NdCaAlO4(001) substrates. LaSrAlO4(001) and NdCaAlO4(001) substrates have a K2NiF4-type structure with lattice constants of a = 3.756 Å and a = 3.685 Å, respectively. ScFeO3 films were deposited under the same condition as the bixbyite ScFeO3 film. Among the three perovskite-related substrates, only the NdCaAlO4 substrate yields a single-phased ScFeO3 film. In wide-range XRD-RSM along the NdCaAlO4(001) [100], the ScFeO3 film shows diffraction patterns like the perovskite structure. In Figure 6a, the white circles indicate the peaks of the ScFeO3 film. Around the 103 diffraction of NdCaAlO4, the two split peaks of ScFeO3 film occur along the out-of-plane direction. This split is generally observed in the rhombohedral perovskite film because this film on NdCaAlO4(001) has four in-plane domains tilted along the directions. There are two expected domain patterns: pseudocubic perovskite (001) parallel to either the in-plane or the out-of-plane of the

Figure 6. (a) Wide-range XRD-RSM of ScFeO 3 film on NdCaAlO4(001) along NdCaAlO4(001) [100]. Red indexes and white circles indicate the Al2O3 substrate and the ScFeO3 film, respectively. (b) HAADF-STEM image of the ScFeO3 film on NdCaAlO4(001). View direction is NdCaAlO4(001) [100]. Circles and broken lines indicate simulated cationic positions and the pseudocubic unit cells, respectively. (c) Atomic arrangements of Al2O3 from the viewpoint of the pseudocubic cell. Black lines denote the pseudocubic unit cell.

substrate (Figure 7). The ScFeO3 film on the NdCaAlO4 substrate corresponds to the former.

Figure 7. Schematic illustration (a), (b) the two expected domain patterns and (c), (d) the simulation figures of their diffraction patterns.

As described in the section on the corundum-type structure on Fe2O3/Al2O3(0001), it is difficult to distinguish between LiNbO3 and corundum-type structures by XRD measurements. Therefore, the atomic arrangement of ScFeO3 film was observed by HAADF-STEM. HAADF-STEM images with a zone axis of NdCaAlO4(001) [100] show the perovskite-like pattern, and all atoms have the same intensity (Figure 6b,c). We also carried out STEM-EDX analysis for the four spots of ScFeO3 film (Figure S5). The ratio of Fe and Sc at the four spots are almost 50:50 (Table S2), indicating Sc and Fe are disordered at all cation sites. Moreover, the atomic arrangements of film are consistent with a corundum-type structure. These observations indicate that the corundum-type ScFeO3 is epitaxially grown on the NdCaAlO4 substrate. YMnO3-Type Structure on Al2O3(0001). In previous reports, YMnO3-type structures such as REFeO3 and REMnO3 were stabilized on YSZ(111) and Al2O3(0001) substrates.41−43 Thus, we deposited ScFeO 3 films on YSZ(111) and Al2O3(0001) substrates. The growth conditions of the ScFeO3 film were the same as those of the bixbyite-type ScFeO3 film. The ScFeO3 film on the YSZ(111) substrates has 5218

DOI: 10.1021/acs.cgd.6b00770 Cryst. Growth Des. 2016, 16, 5214−5222

Crystal Growth & Design

Article

We then carried out synchrotron XRD around 10−14 and 20−28 diffractions (Figures 9 and 10). The synchrotron XRD

mixed phases of bixbyite and YMnO3-type phases, whereas the ScFeO3 film on the Al2O3(0001) substrate has a single phase of the YMnO 3 -type structure (Figure 8a). The chemical

Figure 9. SXRD 2θ−ω pattern of the YMnO3-type ScFeO3 film around the YMnO3-type ScFeO3 101̅4 reflection. Figure 8. (a) Wide-range XRD-RSM of ScFeO3 film on Al2O3(0001) along Al2O3(0001) [101̅0]. Red and white indexes indicate Al2O3 substrate and ScFeO3 film, respectively. (b) HAADF-STEM image of ScFeO3 film on Al2O3(0001). View direction is Al2O3(0001) [101̅0]. Simulated cationic positions and the unit cell of YMnO3-type structure are indicated by spheres and broken lines, respectively. (c) Schematic atomic arrangement of the YMnO3-type structure viewed along the [101̅0] direction. Black lines indicate the unit cell. Red and green spheres represent Sc and Fe, respectively.

composition of the ScFeO3 film on the Al2O3(0001) substrates by WDX is Fe/Sc = 52:48. In the wide-range XRD-RSM along the Al2O3(0001) [112̅0] direction (Figure 8a), the 000l (l = 2m) and hh2hl (h = n and l = m) diffractions of the YMnO3type structure are observed. To investigate the in-plane relationship between the film and the substrate, phi-scan measurements were carried out around ScFeO3 {3031̅ 2} and Al2O3 {2024̅ } reflections (Figure S6). The YMnO3-type ScFeO3 film has a 6-fold symmetry along the in-plane epitaxial relationship, suggesting that YMnO3-type ScFeO3 [101̅0]//Al2O3 [101̅0]. The lattice constants of the YMnO3-type ScFeO3 film estimated from XRD results are a = 5.72 Å and c = 11.71 Å, respectively. According to a previous report by Masuno et al. that considered synthesis of the YMnO3-type Lu1−xScxFeO3 (0 ≤ x ≤ 0.8) by containerless processing,44 both the a and c lattice constants decrease as the Sc content increases. Note that the lattice constant c changes significantly compared to a in the YMnO3-type structures such as REFeO3 and REMnO3. The ratio of the change in the lattice constant from LuFeO3 to ScFeO3 is −4.5% for the a-axis and −0.34% for the c-axis. REFeO3 (RE = Tb - Lu, Y) has a polar structure with SG P63cm,4,45 whereas InFeO3 has a nonpolar structure with SG P63/mmc13 (YAlO3-type structure) (i.e., nonpolar hexagonal REFeO3 is stabilized for a small ionic radius of rare earth ions). Because of the tilting of the FeO5 bipyramids and the displacement of RE ions, the unit cell volume of the polar phase is tripled, √3a × √3b, compared to that of the nonpolar phase.46 To distinguish polar and nonpolar structures, we carried out a 2θ−θ scan along the ScFeO3 [303̅12] direction (P63cm notation). The 1014̅ and 2028̅ reflections can be observed in the P63cm phase (Figure S7), but not in the P63/mmc phase (Figure S8). The origins of the two peaks at 35.60° and 75.38°, which are estimated from the 30−312 diffraction, are unclear.

Figure 10. SXRD 2θ−ω pattern of the YMnO3-type ScFeO3 film around the YMnO3-type ScFeO3 202̅8 reflection.

patterns clearly show these two peaks, indicating that the ScFeO3 film has a polar structure. Recently, the antiferroelectric phase with SG P3c̅ as a ground state was reported for InMnO3.47 Since antiferroelectric and ferroelectric phases exhibit the same extinction rule, it is difficult to distinguish these two phases by XRD. Therefore, we performed HAADFSTEM along the zone axis of Al2O3(0001) [1120̅ ] (Figure 8b,c). Since the intensity is proportional to Z2, where Z corresponds to an atomic number, bright spots correspond to Fe ions and dark spots to Sc ions, indicating that Fe and Sc ions occupy the inside of the bipyramid and the RE site, respectively. The Sc ions shift upward and downward periodically; displacement patterns are up−up−down and down−down− up. In the case of the polar structure with P63cm, the atomic displacement pattern of the RE ions is up−up−down or down−down−up, whereas in the case of the nonpolar structure with P63/mmc and P3̅c, they are center−center−center and up−center−down, respectively.48,49 In terms of the RE ion arrangements, it is confirmed that ScFeO3 film on Al2O3(0001) is a polar structure with P63cm. Bixbyite-Type Structure on YSZ(001). The bixbyite-type structure with space group (SG) Ia3̅, which is also known as the C-type rare-earth sesquioxide structure, is one kind of deficient fluorite-type structure.50 Since one-fourth of its anion sites are empty, the bixbyite-type structure is a fluorite-type superstructure with a doubled-lattice parameter. Considering the similarity of the crystal structure, the ScFeO3 film was deposited on a YSZ(001) substrate. During the deposition, the substrate temperature was kept at 800 °C under P(O2) = 100 mTorr, and the laser fluence was set to 2.4 J cm−2. 5219

DOI: 10.1021/acs.cgd.6b00770 Cryst. Growth Des. 2016, 16, 5214−5222

Crystal Growth & Design

Article

structure.51 The chemical composition of the ScFeO3 film determined by WDX is Fe/Sc = 60:40. These results clearly demonstrate that a bixbyite-type ScFeO3 film can be grown epitaxially on the YSZ(001) substrate.

To analyze its crystal structure, wide-range XRD-RSM was used to measure the ScFeO3 film on YSZ(001). In Figure 11a,



DISCUSSION Table 1 summarizes the deposited films on various substrates, including the main phase, orientation of the crystal structure in Table 1. Summary of the Crystal Structures of ScFeO3 Films Deposited on Various Substratesa

Figure 11. (a) Wide-range XRD-RSM of ScFeO3 film on YSZ(001) along YSZ(001) [100]. Red and white indexes indicate the YSZ substrate and the ScFeO3 film, respectively. (b) HAADF-STEM image of ScFeO3 film on YSZ(001). View direction is YSZ(001) [100]. Circles and broken lines denote the simulated cationic positions and the unit cell, respectively. (c) Schematic atomic arrangement of the bixbyite-type structure viewed along the [100] direction. Black lines mark the unit cell.

the vertical and horizontal axes correspond to the directions of YSZ(001) [001] and YSZ(001) [110], respectively. The out-ofplane XRD patterns show four peaks from the film and two peaks from the substrates (Figure 12). Moreover, ScFeO3 film

a

Rectangle, triangle, and square indicate two-, three-, and four-fold inplane symmetry, respectively.

the films and substrates, in-plane symmetry of the films and substrates, and secondary phase. Among these results, the κAl2O3-type phase is only obtained for the film deposited on the SrTiO3(111) substrate. Although the target composition is Fe/ Sc = 50:50, the chemical composition of the κ-Al2O3 film is Fe/ Sc = 76:24. Trial and error to realize the single-phase κ-Al2O3-type film revealed that the optimum conditions occur when an energy less than the target ablation (less than 0.32 J cm−2) is used for film deposition. Usually, such a change in the ablation energy alters the composition of the ablated species. Therefore, the smaller ablation energy may have tuned the film composition to Fe/Sc = 76:24 when forming the κ-Al2O3-type structure, which may have a smaller formation energy than the same structure with a film composition of Fe/Sc = 50:50 on SrTiO3(111). The assumption is supported by the fact that the deposition of the film using Sc-rich Sc1.5Fe0.5O3 target gives bixbyite-type ScFeO3 thin film without a trace of a κ-Al2O3-type ScFeO3 film. A necessary condition to achieve spinel- and corundum-types of ScFeO3 is the small lattice mismatch between ScFeO3 and the substrate. This idea is based on the fact that the corundum and spinel phases appear in mismatches of 2.28% between corundum-type ScFeO3 and α-Fe2O3 on Al2O3(0001) and 0.83% between spinel-type ScFeO3 and MgO(001), respectively, while the bixbyite-type phase appears on YSZ(111) (6.60% mismatch). The appearance of corundum-type ScFeO3 on NdCaAlO4(001) is attributed to the lattice mismatch of 2.31%, which is calculated using the pseudcubic lattice constant of 3.77 Å for the corundum-type ScFeO3 and NdCaAlO4(100). The appearance of LiNbO3-type ScFeO3 has been reported in the decompression of perovskite-type ScFeO3 stabilized at a static pressure of 15 GPa, which is a reasonable result from the

Figure 12. HRXRD 2θ−θ pattern of the ScFeO3 film on YSZ(001). * indicates peaks of the YSZ substrate.

112 and 332 reflections in wide-range XRD-RSM also appear (Figure 11a). The diffraction patterns of the film are similar to that of YSZ substrate, giving 00l (l = 2n) and hhl (h = 2n and l = 2m) reflections, respectively. These results indicate that a ScFeO3 film on the YSZ(001) has a doubled-lattice parameter of YSZ with a fluorite-type structure. Therefore, the crystal structure of ScFeO3 film on YSZ(001) is attributed to bixbyite-type structure. The out-ofplane XRD patterns of the film give a lattice constant of a = 9.60 Å. The HAADF-STEM image for the ScFeO3 film on YSZ(001) viewed along the YSZ(001) [100] agrees with the atomic arrangement of a bixbyite-type structure (Figure 11b,c). Moreover, a 90°-rotated domain structure is observed under the same view condition (Figure S9). This structure has been reported in an epitaxial In2O3 film deposited on a YSZ(001) substrate, which was attributed to the loss of the 2-fold rotation operations along the [110] directions and the 4-fold rotation operation along the [100] directions from the fluorite 5220

DOI: 10.1021/acs.cgd.6b00770 Cryst. Growth Des. 2016, 16, 5214−5222

Crystal Growth & Design



structural viewpoint that Sc and Fe are ordered in both structures.18 The cation-ordered perovskite-type ScFeO3 cannot be stabilized by the thin film technique. Changing from the cation-disordered corundum-type ScFeO3 structure to the cation-ordered LiNbO3 or the perovskite structure may be difficult by the thin film technique utilizing two-dimensional interfacial strain. From these points, we expect that LiNbO3type ScFeO3 may be grown by applying three-dimensional isostatic strain to the film or by applying a chemical pressure using ions with similar valences but different ionic radii (e.g., a combination of lanthanide family and iron or applying chemical pressure using ions with +2/+4 or +1/+5 valencies and different radii). In the case of substrates with a large mismatch for spinel- and corundum-type ScFeO3, either bixbyite- or YMnO3-type phase or both are grown. These results indicate that formation energies of bixbyite- and YMnO3-type phases are smaller than the other phases in the strained thin film state. Considering that YMnO3-type ScFeO3 has yet to be obtained in bulk form, the hierarchy of the energies of the relevant phases in the strained film form differs from that in a bulk state at ambient or high pressure. Finally, the in-plane symmetries of the deposited films and the substrates were considered. κ-Al2O3-type Sc0.48Fe1.52O3 on SrTiO3(111), YMnO3-type ScFeO3 on SrTiO3(001), and MgAl2O4(001) show different in-plane symmetries in the deposited films and the substrates. The formation of three inplane domains rotated 120° from each other is already explained in the previous section for κ-Al 2 O 3 -type Sc0.48Fe1.52O3 on SrTiO3(111). The formation of hexagonal YMnO3-type ScFeO3 films on cubic SrTiO3(001) and MgAl2O4(001) substrates may be ascribed to the effect of the interfacial layer such as the amorphous phase, which has been reported in the YMnO3 and YbFeO3 films deposited on YSZ(111) and Si(001) substrates, respectively.52,53

Article

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.cgd.6b00770. XRD patterns, HAADF-STEM images, schematic illustration of the difference of diffraction patterns, and results of STEM-EDX analysis (PDF)



AUTHOR INFORMATION

Corresponding Authors

*(M.I.) E-mail: [email protected]. *(S.Y.) E-mail: [email protected]. Author Contributions

All authors wrote the manuscript. H.Y., T.S., S.Y., T.T., and M.I. conducted this study. H.Y. prepared and characterized thin films. T.S., S.Y., and O.S. measured SXRD. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This work was supported by the Japan Agency of the Promotion of Science through Grants-in-Aid for Scientific Research (A) (M.I.: Nos. 15H02292 and 26620190), a Grantin-Aid for young scientists (B) (S.Y.: No. 40616687), and the MEXT Elements Strategy Initiative to Form Core Research Center. The synchrotron radiation experiments were performed at BL15XU and approved by the NIMS synchrotron at SPring8 under Proposal Nos. 2013A4711, 2015A4905, and 2016A4909 for S.Y. This work was partly supported by NIMS microstructural characterization platform as a program "Nanotechnology Platform" Project No. A-15-NM-0006 and A16-NM-0010 for S.Y.





REFERENCES

(1) Kay, H. F.; Vousden, P. London, Edinburgh, Dublin Philos. Mag. J. Sci. 1949, 40, 1019. (2) Fei, Y.; Bertka, C.; Mantle, M. Petrol. F. Obs. High Press. Exp. Geochem. Soc. Publ. 1999, 189. (3) Murakami, M. Science 2004, 304, 855. (4) Levin, I.; Brandon, D. J. Am. Ceram. Soc. 1998, 81, 1995. (5) MacHala, L.; Tuček, J.; Zbořil, R. Chem. Mater. 2011, 23, 3255. (6) Haeni, J. H.; Irvin, P.; Chang, W.; Uecker, R.; Reiche, P.; Li, Y. L.; Choudhury, S.; Tian, W.; Hawley, M. E.; Craigo, B.; Tagantsev, K.; Pan, X. Q.; Streiffer, S. K.; Chen, L. Q.; Kirchoefer, S. W.; Levy, J.; Schlom, D. G. Nature 2004, 430, 758. (7) Choi, K. J.; Biegalski, M.; Li, Y. L.; Sharan, A.; Schubert, J.; Uecker, R.; Reiche, P.; Chen, Y. B.; Pan, X. Q.; Gopalan, V.; Chen, L. Q.; Schlom, D. G.; Eom, C. B. Science 2004, 306, 1005. (8) Lee, M. K.; Nath, T. K.; Eom, C. B.; Smoak, M. C.; Tsui, F. Appl. Phys. Lett. 2000, 77, 3547. (9) Schlom, D. G.; Chen, L. Q.; Pan, X.; Schmehl, A.; Zurbuchen, M. A. J. Am. Ceram. Soc. 2008, 91, 2429. (10) Ikeda, A.; Manabe, T.; Naito, M. Phys. C 2014, 505, 24. (11) Moreira Dos Santos, A. F.; Cheetham, A. K.; Tian, W.; Pan, X.; Jia, Y.; Murphy, N. J.; Lettieri, J.; Schlom, D. G. Appl. Phys. Lett. 2004, 84, 91. (12) Ramadass, N. Mater. Sci. Eng. 1978, 36, 231. (13) Giaquinta, D. M.; zur Loye, H. C. Chem. Mater. 1994, 6, 365. (14) Inaguma, Y.; Yoshida, M.; Tsuchiya, T.; Aimi, A.; Tanaka, K.; Katsumata, T.; Mori, D. J. Phys. Conf. Ser. 2010, 215, 012131. (15) Schneider, S. J.; Roth, R. S.; Waring, J. L. J. Res. Natl. Bur. Stand., Sect. A 1961, 65A, 345. (16) Cassedanne, J.; Forestier, H. C. R. Hebd. Seances Acad. Sci. 1960, 250, 2898.

CONCLUSION In conclusion, using the ScFeO3 target, we stabilized five crystal structures in the film form by selecting the appropriate substrates or using a buffer layer: κ-Al2O3-, spinel-, corundum-, YMnO3-, and bixbyite-type structures. XRD and HAADFSTEM were used to identify the phases. The lattice mismatch likely plays an important role in the phase stabilization of spinel- and corundum-type ScFeO3 films. Additionally, the chemical atomic arrangement between the film and the substrate plays a vital role in all structures. The bottom layer is only formed in the κ-Al2O3-type structure. Four of structures were obtained for the first time: κ-Al2O3-, spinel-, corundum-, and YMnO3-type ScFeO3. From the results of the present study, we can infer that targeted structures can be obtained by carefully selecting the substrate, crystal structure, and preparation conditions of the film (e.g., the oxygen partial pressure, temperature, and laser fluence), even if a structure has not been achieved by conventional methods. Moreover, this study suggests that the epitaxial thin film fabrication technique may be a powerful tool to explore new functional materials. The selection of the appropriate combination of materials, substrates, and preparation conditions may overcome the small potential barrier to the adjacent metastable phase or bypass the hierarchy of Gibbs energy to reach other metastable phases. 5221

DOI: 10.1021/acs.cgd.6b00770 Cryst. Growth Des. 2016, 16, 5214−5222

Crystal Growth & Design

Article

(17) Bréard, Y.; Fjellvag, H.; Hauback, B. Solid State Commun. 2011, 151, 223. (18) Kawamoto, T.; Fujita, K.; Yamada, I.; Matoba, T.; Kim, S. J.; Gao, P.; Pan, X.; Findlay, S. D.; Tassel, C.; Kageyama, H.; Studer, A. J.; Hester, J.; Irifune, T.; Akamatsu, H.; Tanaka, K. J. Am. Chem. Soc. 2014, 136, 15291. (19) Li, M.; Adem, U.; McMitchell, S. R. C.; Xu, Z.; Thomas, C. I.; Warren, J. E.; Giap, D. V.; Niu, H.; Wan, X.; Palgrave, R. G.; Schiffmann, F.; Cora, F.; Slater, B.; Burnett, T. L.; Cain, M. G.; Abakumov, A. M.; van Tendeloo, G.; Thomas, M. F.; Rosseinsky, M. J.; Claridge, J. B. J. Am. Chem. Soc. 2012, 134, 3737. (20) Mizoguchi, Y.; Onodera, H.; Yamauchi, H.; Kagawa, M.; Syono, Y.; Hirai, T. Mater. Sci. Eng., A 1996, 217−218, 164. (21) Giaquinta, D. M.; Davis, W. M.; zur Loye, H. C. Acta Crystallogr., Sect. C: Cryst. Struct. Commun. 1994, 50, 5. (22) Nagashio, K.; Kuribayashi, K. J. Am. Ceram. Soc. 2002, 85, 2550. (23) Kuribayashi, K.; Kumar, M. S. V. J. Phys. Conf. Ser. 2011, 327, 012019. (24) Yamaguchi, O.; Takemura, H.; Yamashita, M.; Hayashida, A. J. Electrochem. Soc. 1991, 138, 1492. (25) Wood, E. A. Acta Crystallogr. 1960, 13, 682. (26) Dayal, R. R.; Gard, J. a.; Glasser, F. P. Acta Crystallogr. 1965, 18, 574. (27) Sakurai, S.; Namai, A.; Hashimoto, K.; Ohkoshi, S. J. Am. Chem. Soc. 2009, 131, 18299. (28) Rozenberg, G. K.; Dubrovinsky, L. S.; Pasternak, M. P.; Naaman, O.; Le Bihan, T.; Ahuja, R. Phys. Rev. B: Condens. Matter Mater. Phys. 2002, 65, 064112. (29) Tsuchiya, J.; Tsuchiya, T.; Wentzcovitch, R. M. Phys. Rev. B: Condens. Matter Mater. Phys. 2005, 72, 020103. (30) Liu, D.; Lei, W. W.; Zou, B.; Yu, S. D.; Hao, J.; Wang, K.; Liu, B. B.; Cui, Q. L.; Zou, G. T. J. Appl. Phys. 2008, 104, 083506. (31) Bouree, F.; Baudour, J. L.; Elbadraoui, E.; Musso, J.; Laurent, C.; Rousset, A. Acta Crystallogr., Sect. B: Struct. Sci. 1996, 52, 217. (32) Sun, Z. H.; Zhou, Y. L.; Dai, S. Y.; Cao, L. Z.; Chen, Z. H. Appl. Phys. A: Mater. Sci. Process. 2008, 91, 97. (33) Hamasaki, Y.; Shimizu, T.; Taniguchi, H.; Taniyama, T.; Yasui, S.; Itoh, M. Appl. Phys. Lett. 2014, 104, 082906. (34) Gich, M.; Gazquez, J.; Roig, A.; Crespi, A.; Fontcuberta, J.; Idrobo, J. C.; Pennycook, S. J.; Varela, M.; Skumryev, V.; Varela, M. Appl. Phys. Lett. 2010, 96, 112508. (35) Sakurai, S.; Jin, J.; Hashimoto, K.; Ohkoshi, S. J. Phys. Soc. Jpn. 2005, 74, 1946. (36) Oshima, T.; Okuno, T.; Fujita, S. Jpn. J. Appl. Phys. 2007, 46, 7217. (37) Suzuki, N.; Kaneko, K.; Fujita, S. J. Cryst. Growth 2013, 364, 30. (38) Megaw, H. D. Acta Crystallogr., Sect. A: Cryst. Phys., Diffr., Theor. Gen. Crystallogr. 1968, 24, 583. (39) Megaw, H. D.; Darlington, C. N. W. Acta Crystallogr., Sect. A: Cryst. Phys., Diffr., Theor. Gen. Crystallogr. 1975, 31, 161. (40) Tagliente, M. A.; De Caro, L.; Sacchetti, A.; Tapfer, L.; Balestrino, G.; Medaglia, P. G.; Tebano, A.; Tucciarone, A. J. Cryst. Growth 2000, 216, 335. (41) Bossak, A. a.; Graboy, I. E.; Gorbenko, O. Y.; Kaul, A. R.; Kartavtseva, M. S.; Svetchnikov, V. L.; Zandbergen, H. W. Chem. Mater. 2004, 16, 1751. (42) Lee, J. H.; Murugavel, P.; Ryu, H.; Lee, D.; Jo, J. Y.; Kim, J. W.; Kim, H. J.; Kim, K. H.; Jo, Y.; Jung, M.-H.; Oh, Y. H.; Kim, Y.-W.; Yoon, J. G.; Chung, J.-S.; Noh, T. W. Adv. Mater. 2006, 18, 3125. (43) Lee, D.; Lee, J. H.; Murugavel, P.; Jang, S. Y.; Noh, T. W.; Jo, Y.; Jung, M. H.; Ko, Y. D.; Chung, J. S. Appl. Phys. Lett. 2007, 90, 182504. (44) Masuno, A.; Ishimoto, A.; Moriyoshi, C.; Kawaji, H.; Kuroiwa, Y.; Inoue, H. Inorg. Chem. 2015, 54, 9432. (45) Ahn, S. J.; Lee, J. H.; Jeong, Y. K.; Na, E. H.; Koo, Y. M.; Jang, H. M. Mater. Chem. Phys. 2013, 138, 929. (46) Wang, W.; Zhao, J.; Wang, W.; Gai, Z.; Balke, N.; Chi, M.; Lee, H. N.; Tian, W.; Zhu, L.; Cheng, X.; Keavney, D. J.; Yi, J.; Ward, T. Z.; Snijders, P. C.; Christen, H. M.; Wu, W.; Shen, J.; Xu, X. Phys. Rev. Lett. 2013, 110, 237601.

(47) Kumagai, Y.; Belik, A. a.; Lilienblum, M.; Leo, N.; Fiebig, M.; Spaldin, N. a. Phys. Rev. B: Condens. Matter Mater. Phys. 2012, 85, 174422. (48) Roddatis, V. V.; Akbashev, A. R.; Lopatin, S.; Kaul, A. R. Appl. Phys. Lett. 2013, 103, 112907. (49) Huang, F. T.; Wang, X.; Oh, Y. S.; Kurushima, K.; Mori, S.; Horibe, Y.; Cheong, S. W. Phys. Rev. B: Condens. Matter Mater. Phys. 2013, 87, 184109. (50) Adachi, G.; Imanaka, N. Chem. Rev. 1998, 98, 1479. (51) Hu, Y. L.; Rind, E.; Speck, J. S. J. Appl. Crystallogr. 2014, 47, 443. (52) Yoo, D. C.; Lee, J. Y.; Kim, I. S.; Kim, Y. T. J. Cryst. Growth 2002, 234, 454. (53) Iida, H.; Koizumi, T.; Uesu, Y.; Kohn, K.; Ikeda, N.; Mori, S.; Haumont, R.; Janokin, P.-E.; Kiat, J.; Fukunaga, M.; Noda, Y. J. Phys. Soc. Jpn. 2012, 81, 024719.

5222

DOI: 10.1021/acs.cgd.6b00770 Cryst. Growth Des. 2016, 16, 5214−5222