Structural and Vibrational Properties of Corundum-type In2O3 Nanocrystals under Compression

1 Structural and Vibrational Properties of Corundum-type In2O3 Nanocrystals under Compression J.A. Sans,1,* R. Vilaplana,2 D. Errandonea,3 V.P. Cuenca-Gotor,1 B. García-Domene,3 C. Popescu,4 F. J. Manjón,1 A. Singhal,5 S. N. Achary,5 D. Martinez-Garcia,3 J. PellicerPorres,3 P. Rodríguez-Hernández,6 and A. Muñoz 6 1 Instituto de Diseño para la Fabricación y Producción Automatizada, MALTA Consolider Team-Universitat Politècnica de València, 46022 València, Spain 2 Centro de Tecnologías Físicas: Acústica, Materiales y Astrofísica, MALTA Consolider Team, Universitat Politècnica de València, 46022 València, Spain 3 Departamento de Física Aplicada-ICMUV, MALTA Consolider Team, Universidad de Valencia, Edificio de Investigación, C/Dr. Moliner 50, Burjassot, 46100 Valencia, Spain 4 ALBA-CELLS, Cerdanyola, 08290 Barcelona, Spain 5 Chemistry Division, Bhabha Atomic Research Centre, Trombay, Mumbai 400085, India 6 Departamento de Física, Instituto de Materiales y Nanotecnología, MALTA Consolider Team, Universidad de La Laguna, 38205 Tenerife, Spain E-mail: [email protected] ABSTRACT This work reports the structural and vibrational properties of nanocrystals of corundum-type In2O3 (rh-In2O3) at high pressures by using angle-dispersive x-ray diffraction and Raman scattering measurements up to 30 GPa. The equation of state and the pressure dependence of the Raman-active modes of the corundum phase in nanocrystals are in good agreement with previous studies on bulk material and compare nicely with theoretical simulations on bulk rh-In2O3. Nanocrystalline rh-In2O3 showed stability under compression at least up to 20 GPa, unlike bulk rh-In2O3 which gradually transforms to the orthorhombic Pbca (Rh2O3-III-type) structure above 12-14 GPa. The different stability range found in nanocrystalline and bulk In2O3 is discussed. KEYWORDS: indium oxide, nanocrystals, corundum, X-ray diffraction, Raman spectroscopy, high-pressure, ab initio calculations 1. Introduction Indium oxide (In2O3) is a sesquioxide whose stable phase at room conditions is the cubic bixbyite-type structure1 (space group (SG) Ia-3, N. 206, Z=16). This phase has been widely studied for optoelectronic applications due to its special properties as a transparent conducting oxide (TCO).2,3 The widespread use of cubic In 2O3 (c-In2O3) in industrial processes covers the 2 fabrication of window layers in solar cells,3-5 light-emitting diodes,6,7 electrochromic windows,8 liquid-crystal displays,9,10 and gas sensors. 11,12 However, the scarcity and high cost to obtain indium metal has led to the study of new systems, like Sn-doped In2O3 (ITO) and In2−2xZnxSnxO3 (x ≤ 0.4, also named ZITO or IZTO), in order to cut the price of its production process.13−15 Rhombohedral corundum-type (SG R-3c, N. 167, Z= 6) In 2O3 is also known to be a metastable phase at room conditions that can be obtained after high pressure (HP)-high temperature (HT) treatment, but also at room pressure.16-40 Modern calculations of pure and doped rhombohedral In2O3 (rh-In2O3), suggest that this phase could be even better TCO than c-In2O3; thus opening a new way to search for improved TCOs.41-44 Given the above result, it is clear that the study of the HP phases (perhaps metastable at ambient conditions) of a technological important compound, like In 2O3, is crucial for the advancement of science and technology.45-47 In this respect, three additional orthorhombic phases of In2O3 have been theoretically predicted and indeed obtained by application of HP: Rh2O3-III type (SG Pbca, N. 61, Z= 8),36,37 Rh2O3-II type (SG Pbcn, N. 60, Z= 4),32, 48-51 and -Gd2S3-type (SG Pnma, N. 62, Z= 4).52 Among them only the Pbcn phase has been claimed to be metastable at ambient conditions,49-51 despite the Pbca phase is a phase reported at lower pressures than the Pbcn phase.36,37 Working with nanocrystals opens the door to enhanced properties of materials;53,54 and adds the possibility of obtaining metastable HP crystalline and amorphous phases at room conditions with different properties from the original material.47,55-57 It is well known that most nanocrystals undergo a pressure-induced phase transition at larger pressures than in the bulk and that many nanocrystals undergo a pressure-induced amorphization;55 however, many questions remain open regarding the behaviour of nanocrystals at HP in comparison with bulk materials. In 3 particular, the different thermodynamic stability of crystalline phases in nanocrystals as compared to bulk material,55,56 and whether the compressibility of nanocrystals is different to that of bulk material or if the different compressibility comes from a different degree of hydrostaticity affecting micro and nanocrystal powders. Even if several works have been devoted to the study of In2O3 nanocrystals at HP, most of them have focused on nanocrystals of cubic bixbyite-type In2O3,58-60 and only a small part have been devoted to study nanocrystals of rhombohedral corundum-type In2O3.49,50,59 In fact, a different stability range of the rhomobohedral phase under compression has been observed depending on the sample size. Gurlo et al., who studied nanocrystals of rh-In2O3, suggested that this compound is stable at room temperature between room pressure and 30 GPa;49 however, two recent studies reported that bulk rh-In2O3 undergoes a transformation above 12-14 GPa in good agreement with theoretical simulations.36,37 The apparent lack of phase transition in nanocrystals up to 30 GPa suggests the existence of large kinetic barriers to explain their extended structural stability. In this work we report a HP study of the structural and vibrational properties of nanocrystals of corundum-type In2O3 by means of angle-dispersive powder x-ray diffraction (XRD) and Raman scattering (RS) measurements up to 30 GPa. Both the equation of state and the pressure dependence of the Raman-active modes of the corundum phase in nanocrystals are reported, showing a nice agreement with previous studies on rh-In2O3 nanocrystals.49 These results have been compared with experimental and theoretical structural and lattice dynamics results on bulk rh-In2O3 offering an explanation to the previously apparent contradicting results observed between bulk and nanocrystalline rh-In2O3. 4 2. Methods 2.1. Experimental details Rh-In2O3 nanoparticles were synthesized by thermal dehydration of InOOH nanoparticles previously prepared from a solvothermal reaction between indium nitrate and tetramethylammonium hydroxide.61 XRD measurements confirm that nanocrystalline powders are pure and have corundum-type structure with an hexagonal unit cell with lattice parameters a=b= 5.485 Å and c= 14.530 Å, which are in good agreement with reference JCPDS card 220336 (a=b= 5.487 Å and c= 14.510 Å). On the other hand, transmission electron microscopy (TEM) confirmed both the crystalline structure and purity and determined that nanocrystals have a rice-grain type morphology with an average length (width) of 20 (13) nm and a uniform distribution between 10 (6) and 30 (20) nm.61 Room-temperature experiments in rh-In2O3 nanocrystals were conducted up to 30 GPa in a membrane-type diamond anvil cell. In HP-XRD and HP-RS experiments, powder sample was loaded inside a 150 m diameter hole drilled in an inconel gasket together with a mixture of methanol-ethanol-water (MEW) in a 16:3:1 proportion as a quasi-hydrostatic pressuretransmitting medium (PTM) as in a previous work,37 in order to compare results for bulk and nanocrystalline rh-In2O3 under the same conditions. Additionally, a second XRD experiment with Ar as PTM was done in order to prove if nanopowders could be affected by non-hydrostatic effects using MEW and to compare with previous results described in the literature.37 For pressure calibration in angle-dispersive powder XRD and RS measurements we introduced inside the pressure cavity Cu powder62 and ruby,63 respectively. In the XRD experiments we collected also two patterns, one in the region where the Cu peaks were strong and the other where there is no Cu signal. The first pattern was used to determine pressure and the second one to analyse the crystal structure of In2O3. 5 Angle-dispersive powder XRD measurements at different pressures were performed in the MSPD beamline64 at ALBA synchrotron facility. This beamline is equipped with Kirkpatrick-Baez mirrors to focus the monochromatic beam and a SX165 CCD detector with a diameter of 165 mm. We used a wavelength of 0.4246 Å and the sample-detector distance during the experiment was set to 240 mm. The 2-D diffraction images were integrated with FIT2D software.65 Lattice parameters of XRD patterns were obtained with Rietveld refinements performed using POWDERCELL66 and GSAS67 program packages. On the other hand, RS measurements at room temperature and different pressures were excited using the 632.8 nm HeNe laser (with a power below10 mW) and collected in backscattering geometry with a Horiba Jobin Yvon LabRam HR UV spectrometer equipped with a thermoelectrically cooled multichannel CCD detector (resolution below 2 cm-1). 2.2.Theoretical Calculations Total-energy ab initio calculations were performed for bulk rh-In2O3 within the density functional theory (DFT)68 using the Vienna Ab initio Simulation Package (VASP)69 as described in Ref. 37 for bulk rh-In2O3. In particular, we have used the projector-augmented wave (PAW)70 scheme implemented in this package and the generalized gradient approximation (GGA) was used for the description of the exchange-correlation energy with the PBEsol prescription.71 We have also performed lattice dynamics calculations of the phonon modes in bulk rhIn2O3 at the center of the BZ ( point) using the direct force constant approach (or supercell method)72 already described in Ref. 37. Our theoretical results provide both the frequencies of the normal modes and their polarization vectors and enable us to assign the Raman-active modes observed in nanocrystalline rh-In2O3 as previously done for bulk rh-In2O3.37 6 3. Results 3.1. XRD measurements Figure 1a shows a selection of the experimental angle-dispersive powder XRD patterns of rh-In2O3 at room temperature at different pressures up to 21 GPa obtained in the first experiment (using MEW mixture as PTM). Figure 1b shows similar results up to 29.8 GPa obtained in the second experiment (using Ar as PTM). All the XRD patterns can be indexed with the corundumtype structure. Under compression some of the peaks shift faster to higher angles than others due to the anisotropic compressibility of In2O3 which will be discussed in detail later. This phenomenon is illustrated in Figs. 1a and 1b by the (104) and (110) Bragg peaks which gradually merge under compression because the interplanar distance associated to the (104) peak decreases faster under compression than that corresponding to the (110) reflection. In the second experiment, in addition to the Bragg peaks of the sample, Bragg peaks of Ar are also detected. These peaks can be easily identified because, due to the large compressibility of Ar, they shift faster to higher angles under compression than the peaks corresponding to In2O3. The Ar peaks were used to confirm the pressure determined from Cu using the equation of state of Ar.73 The pressure determined from Ar agrees within 0.1 GPa with the pressure determined from Cu. The results of both XRD experiments indicate that the phase transition to the Pbca-type phase observed in bulk rh-In2O3 above 12-14 GPa,36,37 is not observed in nanocrystals even at 29.8 GPa. We would like to comment here that beyond 20 GPa all the peaks assigned to corundum-type In2O3 can be also explained with the Pbca-type structure. However, this structure has a peak at low angles, the (002) peak expected to be located around 3.2º, which is not present in the corundum-type structure. In our experiments, we did not observe this peak, which supports the stability of the corundum-type structure up to 7 29.8 GPa. This result is in agreement with the larger stability of the corundum phase measured by Gurlo et al. in nanocrystalline rh-In2O3 samples with particle size between 50 and 100 nm.49 XRD peaks in our first experiment show slight broadening with increasing pressure above 10 GPa which could be due to a slight loss of quasi-hydrostatic conditions of the PTM. The same broadening is observed in the second experiment but around 20 GPa. This is a consequence of the better quasihydrostatic conditions generated by Ar than by MEW above 10 GPa.74 In this regard, it is expected that small non-hydrostatic stresses do not influence the structural stability of relatively uncompressible compounds like sesquioxides.75 In this work, we have extracted the evolution of the structural parameters of nanocrystalline rh-In2O3 under compression by Rietveld refinement. In Figure 1, one can see the experimental diffraction pattern (black symbols), the Rietveld refinement performed (black line) and the residuals (red lines) showing a good fit to the experimental data. Figure 2 represents the experimental pressure dependence of the hexagonal unit-cell volume, lattice parameters a and c and the c/a ratio in nanocrystalline rh-In2O3. The experimental data of the unit-cell volume up to 29.8 GPa can be fitted with a third-order Birch-Murnaghan (BM) EOS,76,77 which yields V0= 376.7(4) A3, B0= 169(6) GPa and B0’= 6.2(9). These values are in good agreement with theoretical data of bulk material fitted up to 30 GPa (V0= 383.3 A3, B0= 169 GPa and B0’= 3.3) and with previous experimental (B0= 176-180 GPa) and theoretical estimations in bulk material.37 Note that the value of the bulk modulus for rh-In2O3 nanocrystals is slightly smaller than that in bulk material (see Table I). In relation to this, it must be noted that a decrease of the bulk modulus in nanocrystals with respect to bulk material has been observed in a number of oxides, like TiO2,78,79 ZrO2,80,81 SnO2,82 MgO,83 ZnO,84 and Al2O3.85 8 A deviation of the compressibility of nanocrystals above certain pressure has been found in other works. In particular, nanocrystals of anatase TiO2 pressurized with a 4:1 methanol-ethanol mixture were found to undergo deviations of the compressibility of the a axis and unit-cell volume above 10-12 GPa which were attributed to the appearance of an intermediate local distortion (~ 2 Å) with TiO2-II structure.86 On the other hand, a significantly lower bulk modulus (148±5 GPa) compared to theoretical calculations (169.4 GPa) has been recently observed in c-In2O3 nanocrystals with a high defect density under compression with a 4:1 methanol-ethanol mixture as PTM.60 In that work, authors claimed that the different bulk modulus of two nanocrystalline samples with similar sizes and morphologies were due to their different defect densities. Another possibility is that the change in compressibility observed above 8 GPa in In2O3 nanocrystals could be due to non-hydrostatic conditions as it was reasoned for nanocrystalline TiO2 anatase.78 Note that nanocrystalline powder is more prone to non-hydrostatic conditions than bulk powder because the surface area of nanocrystalline grains tend to be more surrounded by other nanocrystalline grains than by the molecules of the PTM. On the other hand, the surface area of microcrystalline grains tend to be more surrounded by the molecules of the PTM than by other grains. In any case, it was demonstrated for nanocrystalline TiO2 anatase that the EOS obtained prior to observation of non-hydrostatic conditions is in good agreement with the EOS obtained under truly hydrostatic conditions along the whole range of the experiments.78 Since measurements in nanocrystalline rhIn2O3 using MEW or Ar show similar results (see Fig. 2), we are confident of the bulk modulus obtained for nanocrystals with an average grain size around 20 nm up to 30 GPa. It must be stressed that the bulk modulus we have obtained for rh-In2O3 nanocrystals in the present study is much smaller than that reported (305 GPa) in a previous work.58 In fact, Ref. 58 also reports an overestimation of the bulk modulus in rh-In2O3 as was also confirmed in our previous work.37 The 9 main reason for the overestimation of the bulk modulus in Ref. 58 is the lack of structural data for the corundum-type phase at low pressures and the inherent uncertainty caused by the extrapolation of the bulk modulus from data above 20 GPa. On the other hand, both a and c lattice parameters show a rather similar pressure dependence in both bulk material and nanocrystals, also observed in the c/a ratio (inset of figure 2b). The lattice parameters of rh-In2O3 nanocrystals show an almost linear variation with pressure. Both a and c parameters have been fitted to a modified Murnaghan’s equation with free parameters a0, B0 and B0’. 󰇡1 󰆓 󰇢󰇧   󰆓󰇨 , Eq. (1) The results of these fits are shown in Table II. The axial compressibilities of the aand c-axis defined as P x x x  1  (where x = a, c) can be obtained as 0 3 1 B x  with the B0 values obtained with the Murnaghan EOS fit for the lattice parameters. The experimental compressibility of the a (c) axis is 1.4 (3.0) ·10-3 GPa-1 This result indicates that the c axis is more than two times more compressible than the a axis as it occurs in bulk rh-In2O3.37 Consequently, we can conclude that the major contribution to the compressibility of nanocrystalline rh-In2O3 comes from the larger compressibility of the c axis than that of the a axis. These behaviours are in agreement with experiments and theoretical calculations already published for bulk rh-In2O337 and theoretical calculations (Table II). Similarly, the c/a ratio in bulk and nanocrystalline rh-In2O3 shows the same tendency in good agreement with our theoretical calculations; however, this behaviour is different to that reported in corundum-type Al2O3, V2O3, Cr2O3, and Fe2O3,87 as already discussed in our previous work regarding bulk rh-In2O3.37 16 1b Figure 1. Selected X-ray diffraction patterns of nanocrystalline rh-In2O3 at several pressures using MEW (a) and Ar (b) as PTM. Diffraction patterns have been shifted vertically for clarity. Different diffraction peaks are assigned in the graph to rh-In2O3 (R-3c), Ar and Cu phases. Refinements and residuals of the fit to these phases are also shown. 17 Figure 2. Pressure dependence of the unit-cell parameters (a) and volume (b) in nanocrystalline rh-In 2 O 3 for our first experiment (close symbols) and the second experiment (open symbols). The inset shows the pressure dependence of the c/a ratio in nanocrystalline rh-In 2 O 3 . Red dashed lines correspond to theoretical data of bulk rh-In 2 O 3 . 18 Figure 3. Selected RS spectra of nanocrystalline rh-In2O3 at different pressures on upstroke up to 29.3 GPa. In the RS spectrum at 0.2 GPa, first-order Raman modes of rh-In2O3 are indicated by arrows, while second-order Raman modes are denoted with asterisks. Above 20 GPa, arrows in the RS spectra indicate the new peaks assigned to the Pbca phase. The top RS spectrum at 2.3 GPa shows the reversibility of the sample to the corundum phase on downstroke. 19 Figure 4. Pressure dependence of the Raman-active mode frequencies in nanocrystalline rh-In2O3 (violet circles) and Pbca-type In2O3 (orange squares). Full (empty) symbols correspond to the upstroke (downstroke) process. Solid lines correspond to ab initio simulated Raman-active modes of bulk rh-In2O3. The symmetry of the Raman-active modes of rh-In2O3 is indicated. 20 TABLES Table I. Comparison of theoretical and experimental zero pressure unit-cell volume per formula unit, bulk modulus, and its pressure derivative for bulk and nanocrystalline rh-In2O3. Phase V 0 (Å3) B0 (GPa) B’0 Method nano R-3c 62.78(1) 169(6) 6.2(9) Exp. (XRD)a R-3c 61.49 305.5 2.9 Exp. (XRD) b bulk R-3c 62.86(4) 180(7) 3.2(1.2) Exp. (XRD)c R-3c 63.88 167 3.3 Theo. (GGAPBEsol)d a This work b Ref. 58 c Ref. 37 d Data up to 30 GPa. 21 Table II. Lattice parameters and their bulk modulus and pressure derivative of bulk modulus at room pressure in nanocrystalline rh-In2O3 (top) as compared to experimental and theoretical parameters in bulk Exp. rh-In2O3 (bottom) a0 (Å) B0a (GPa) Ba’ c0 (Å) B0c (GPa) Bc’ Method 5.4831(8) 233(16) 11.8 14.461(8) 109(3) 0.48 Our Exp. (XRD) 5.479(2) 220(30) 11.8 14.511(5) 121(7) 0.48 Exp. (XRD) 5.520(1) 216(3) 11.8(4) 14.552(1) 106.0(9) 0.48(11) Theo. (GGAPBEsol) 22 Table III. Room-pressure experimental frequencies, pressure coefficients and Grüneisen parameters of the Raman active modes of bulk and nanocrystalline rh-In2O3. Bulk rh-In2O3 Nano rh-In2O3 Abinitiocalculations ExperimentalbExperimental Mode (Sym)  cm-1  a cm-1    cm-1  c A1 g 156(1) 0.65 (3) -0.012 (2) 0.66 162.3 (2) 0.85 (4) -0.015 (1) 0.93 163.2(2) 0.590(14) 0.61 E g 169(1) 0.73 (3) -0.010 (2) 0.68 177.1 (2) 1.01 (6) -0.022 (3) 1.01 181.3(5) 0.54(4) 0.50 E g 214(1) 1.54 (3) -0.023 (2) 1.14 218.1 (3) 1.85 (8) -0.039 (5) 1.50 223.9(5) 1.30(3) 0.97 E g 265(1) 1.11 (6) -0.005 (5) 0.66 270.7 (4) 1.33 (13) -0.011 (7) 0.87 270(1) 1.49(11) 0.93 E g 373(2) 4.91 (5) -0.040 (4) 2.08 383.8 (6) 4.69 (16) -0.022 (8) 2.16 388.3(9) 4.61(6) 2.0 A1 g 486(2) 4.46 (6) -0.033 (4) 1.45 501.9 (9) 3.92 (25) 0.005 (14) 1.38 502.5(6) 4.75(5) 1.6 E g 569(2) 6.14 (5) -0.059 (4) 1.70 590.3 (6) 5.95 (18) -0.058 (12) 1.78 599(1) 5.42(11) 1.53 a Theoretical B0= 158 GPa used for calculating the Grüneisen parameter. b From ref. 37 (experimental B0= 177 GPa used for calculating the Grüneisen parameter). c Experimental B0= 169 GPa used for calculating the Grüneisen parameter. P           GPa cm 1 2 2 P           2 1 GPa cm P            GPa cm 1 2 2 P           2 1 GPa cm P            GPa cm 1 23 Author Contributions The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Acknowledgments This work is supported by the Spanish MICINN projects MAT2016-75586-C4-1/2/4-P and MAT2015-71070-REDC. A.M. and P.R-H acknowledge computing time provided by Red Española de Supercomputación (RES) and MALTA-Cluster. J.A.S. acknowledges financial support through the Ramon y Cajal fellowship. References 1 Marezio M 1966 Refinement of the Crystal Structure of In2O3 at two Wavelengths Acta Crystallogr. 20 723-728. 24 2 Mizoguchi H and Woodward P M 2004 Electronic Structure Studies of Main Group Oxides Possessing Edge-Sharing Octahedra: Implications for the Design of Transparent Conducting Oxides Chemistry of Materials 16 5233-5248. 3 Breeze A J et al 2001 Charge transport in TiO2/MEH-PPV polymer photovoltaics Physical Review B 64 125205. 4 Tahar R B H et al 1998 Tin doped indium oxide thin films: Electrical properties J. Appl. Phys. 83 2631-2645. 5 Kaspar T C et al 2011 ZnO/Sn:In2O3 and ZnO/ CdTe Band Offsets for Extremely Thin Absorber Photovoltaics Appl. Phys. Lett. 99 263504. 6 Tang C W and Vanslyke S A 1987 Organic electroluminescent diodes Appl. Phys. Lett. 51 913 7 Tang L-M et al 2010 Donor-donor binding in In2O3: Engineering shallow donor levels J. Appl. Phys. 107 083704 8 Steele B C H and Golden S J 1991 Variable transmission electrochromic windows utilizing tindoped indium oxide counterelectrodes Appl. Phys. Lett. 59 2357 9 Lee B H et al 1997 Effect of process parameters on the characteristics of indium tin oxide thin film for flat panel display application Thin Solid Films 302 25-30 10 Hsu S F et al 2005 Highly efficient top-emitting white organic electroluminescent devices Appl. Phys. Lett. 86 253508 11 Himmerlich M et al 2012 Surface properties of stoichiometric and defect-rich indium oxide films grown by MOCVD J. Appl. Phys. 111 093704 12 X. Lai et al 2010 Ordered Arrays of Bead-Chain-like In2O3 Nanorods and Their Enhanced Sensing Performance for Formaldehyde Chemistry of Materials 22 3033-3042 25 13 Rauf I A 1993 Low Resistivity and High Mobility Tin-Doped Indium Oxide Films Mater. Lett. 18 123−127 14 Palmer G B et al 1997 Conductivity and Transparency of ZnO/SnO2-Cosubstituted In2O3 Chem. Mater. 9 3121−3126. 15 Hoel C A et al 2010 High-Pressure Synthesis and Local Structure of Corundum-Type In2−2xZnxSnxO3 (x ≤ 0.7) J. Am. Chem. Soc. 132 16479−16487. 16 Shannon R D 1966 New High Pressure Phases Having The Corundum Structure Solid State Commun. 4 629-630 17 Nordlund Christensen A et al 1967 Hydrothermal Investigation of the Systems In2O3·H2O·Na2O and In2O3·D2O·Na2O. The Crystal Structure of Rhombohedral In2O3 and of In(OH)3 Act. Chem. Scan. 21 1046-1056 18 Prewitt C T et al 1969 The C Rare Earth Oxide-Corundum Transition and Crystal Chemistry of Oxides Having the Corundum Structure Inorg. Chem. 8 1985-1993 19 Atou T et al 1990 Shock-Induced Phase Transition of M203 (M = SC, Y, Sm, Gd, and In)- Type Compounds J. Solid State Chem. 89 378-384 20 Gurlo A et al 2003 Polycrystalline Well-Shaped Blocks of Indium Oxide Obtained by the Sol-Gel Method and their Gas-Sensing Properties Chemistry of Materials 15 4377-4383 21 Gurlo A et al 2003 Corundum-type indium (III) oxide: formation under ambient conditions in Fe2O3–In2O3 system Inorg. Chem. Comm. 6 569-572 22 Yu D B et al 2003 Metastable Hexagonal In2O3 Nanofibers Templated from InOOH Nanofibers under Ambient Pressure Adv. Funct. Mater. 13 497-501 23 Seo W S et al 2003 Preparation and Optical Properties of Highly Crystalline, Colloidal, and Size-Controlled Indium Oxide Nanoparticles Adv. Mater. 15 795-797