Room temperature deposition of a TiO2 thin film from aqueous peroxotitanate solution Yanfeng Gao, Yoshitake Masuda, Zifei Peng, Tetsu Yonezawa and Kunihito Koumoto* Department of Applied Chemistry, Graduate School of Engineering, Nagoya University, Furo-cho, Chikusa-ku, Nagoya 464-8603, Japan. E-mail:
[email protected] Received 5th September 2002, Accepted 13th December 2002 First published as an Advance Article on the web 14th January 2003 A transparent, high purity, amorphous titanium dioxide thin film composed of densely packed nanometersized grains has been successfully deposited on a glass substrate at room temperature from an aqueous peroxotitanate solution using a simple, inexpensive, reproducible, and environmentally friendly method. The as-deposited thin film was 117 nm thick and composed of closely packed particles of 10–15 nm in diameter, but they aggregated into large grains of 50–100 nm in diameter. The aqueous peroxotitanate solution was obtained by dissolving metatitanic acid (H2TiO3) in a mixture of concentrated H2O2 and NH3?H2O. An anatase TiO2 thin film was obtained by heating the as-deposited thin film at 500 uC for 1 h in air. A chemical composition of TiO1.4(O2)0.5(OH)0.2?1.34H2O is proposed for the as-deposited thin film, based on XPS, FT-IR and TG-DTA data. Band gap energies of 3.20 eV for indirect transition and 3.63 eV for direct transition were obtained for the anatase TiO2 film.
Introduction Titanium dioxide thin films have attracted both industrial and academic attention for such potential applications as the gate dielectric in metal oxide semiconductor field-effect transistors (MOSFETs),1 as surfaces for solar energy conversion,2 and as high efficiency photocatalysts3 because of their well-known properties, including high refractive index, high permittivity, and transmittance in the visible region. TiO2 films are usually synthesized by vapor-phase deposition techniques4 such as chemical vapor deposition or atomic layer epitaxy, which require expensive high vacuum equipment and consume a great deal of energy. Novel direct low temperature deposition processes for TiO2 thin films involve controlled hydrolysis of titanium species using an organic solvent5 or in an aqueous solution.6–8 Amorphous TiO2 thin films were successfully synthesized on the silanol regions of UV-irradiated octadecyltrichlorosilane (OTS) self-assembled monolayers (SAMs) under a nitrogen atmosphere by using a series of organic solutions of titanium compounds.5 The hydrolysis of TiCl4,6 TiF47 or (NH4)2TiF68 can yield either amorphous or crystallized TiO2 thin films, depending on the synthetic conditions employed. A complex peroxo precursor of titanium can also be employed for the deposition of TiO2 thin films by the electrochemical method,9a the SAM technique,9b sol–gel method,9c and the dip-coating method.9d The peroxide-based route for synthesizing titanium oxide powder is well established.10 The peroxotitanium solution is usually prepared by adding droplets of pure TiCl4 to an ice-cooled aqueous solution containing H2O2 with or without excess acid.9a,b,11 Even in the case of no excess acid, the pH of the solution is still below 1. The color of the resulting homogeneous solution is intense red, and it changes to yellow–green with increasing pH as a result of deprotonation by water molecules of the peroxotitanium complex.11 However, TiCl4 is highly moisture-sensitive, making it inconvenient and rather costly to use. Moreover, a small concentration of residual chlorine in TiO2 films may result in degradation of their electronic properties.12 In contrast, Tada et al.9d fabricated a TiO2 thin film from a chloride-free aqueous solution by a dip-coating method using the water-soluble peroxotitanate precursor13 ammonium 608
citratoperoxotitanate(IV), (NH4)8[Ti4(C6H4O7)4(O2)4]?8H2O, which was prepared by dissolving citric acid in a transparent aqueous titanium peroxo solution. Although no composition information about their film was reported, impurities such as hydrocarbons might exist in the as-deposited film, which means a heating process is needed to decompose them. However, to our knowledge, no attempt has been made to directly deposit TiO2 thin films from a simple peroxotitanium solution prepared under alkaline conditions. We have succeeded in preparing a peroxotitanium solution by dissolving metatitanic acid (H2TiO3) in a mixture of concentrated H2O2 and NH3?H2O. When the as-obtained homogeneous solution was left at room temperature, it became muddy and precipitation occurred. Judging from the components of the solution, the precipitate was probably TiO2 or TiO2-based hydrate. Therefore, it seemed that it ought to be possible to fabricate a TiO2 thin film at room temperature by controlling the supersaturation of this solution. In this paper, we report that a transparent, high purity, amorphous TiO2 thin film was successfully deposited on a glass substrate at room temperature. The phases present, the chemical composition, the morphology, and the degree of transparency of the resulting film were all characterized. The optical band gaps of the as-deposited film and of others annealed at different temperatures were measured.
Experimental Preparation of precursor solution and deposition of thin films H2TiO3 can be dissolved only in the presence of H2O2 and NH3?H2O in appropriate quantities. Specifically, 3 g of H2TiO3 (80%, Mitsuwa) was added to an ice-cooled mixture consisting of 25 cm3 H2O2 (30% in H2O, Mitsubishi) and 5 cm3 ammonia (25% in H2O, Kishida). After stirring for 90 min, a homogeneous pale yellow–green solution was obtained, which can be stable for several weeks at temperatures of less than 5 uC. Fresh aqueous peroxotitanium solution was diluted with deionized water (w18 MV cm) to 5 mM Ti41 at pH 2.4 (adjusted with HNO3). The cleaned substrate (Corning 1737; barium borosilicate glass) was then floated on the surface of the dilute solution at room temperature (y24 uC) to deposit
J. Mater. Chem., 2003, 13, 608–613 This journal is # The Royal Society of Chemistry 2003
DOI: 10.1039/b208681f
a thin film. The solution became turbid after soaking for about 15 min. After soaking for 12 h, the substrate was carefully rinsed with deionized water before drying in air. Characterization techniques The phases present in the as-deposited thin film, its chemical composition, and its morphology were determined by X-ray diffraction (XRD; Rigaku RAD-C; Cu-Ka, 40 kV, 30 mA), X-ray photoelectron spectroscopy (XPS; VG Scientific ESCALAB 210; Mg-Ka, 15 kV, 18 mA), and atomic force microscopy (AFM; Seiko SPI3800N; scanning frequency 1–2 Hz), along with scanning electron microscopy (SEM; Hitachi model S-3000), respectively. The thickness of the film was measured by laser ellipsometry (Philips PZ2000), with an incidence angle of 70u and wavelength of 632.8 nm. Optical transmission spectra were recorded at room temperature using a V-570 spectrophotometer (JASCO, 100 nm min21) in the wavelength range 200–1000 nm. Fourier-transform infrared spectroscopy (JASCO FT/IR-610) was carried out at a resolution of 4 cm21, with KBr as a reference. Thermogravimetric analysis (TG), as well as differential thermal analysis (DTA) were conducted (Rigaku Thermo plus TG8120, 5 uC min21) to further characterize the collected precipitate after the thin films had been deposited.
Results and discussion Peroxotitanium groups and their hydrolysis It is well known that Ti41 in a H2O2-containing aqueous solution can form a series of colored peroxotitanium complexes depending on the pH of the solution.11 Various colors are closely related to the valency state of titanium and the structures of the complexes. Species with higher peroxo group ratios can be obtained only from strongly alkaline solutions.11b,c At pH w 10, a colorless solution was obtained, and either [Ti(O2)2(OH)2]22 or [Ti(O)(O2)(OH)2]22 was proposed to be the major species in this solution.11c For our as-prepared aqueous solution of the H2TiO3–NH3?H2O–H2O2 system, the pH was found to be greater than 10 and the color was yellow–green. Neither the pH nor the color of the solution is consistent with the experimental results reported in the literature,11c which makes it difficult to determine what was formed in our as-prepared high pH solution. At pH v 3, it is generally accepted that the ratio of peroxo groups to titanium in the species is 1/1, although the exact species present in the acid solutions remain controversial.11b Under our deposition conditions, where the pH was adjusted to be approximately 2.4, according to a quantitative relationship between hydrolysis and electronegativity (based on the socalled partial charge model),9a,b,14 hydrolyzed titanium cations would lead to hydrous oxide precipitation by condensing through both olation and oxolation, which may be written schematically as follows. Olation: M–OH 1 M–OH2 A M–OH–M 1 H2O Oxolation: M–OH 1 M–OH A M–O–M 1 H2O Phase composition XRD [see Fig. 1(a)] of the as-deposited thin film showed no significant peak, indicating that the as-deposited film was amorphous. A diffraction peak due to anatase (101) can be clearly observed after heating at 500 uC for 1 h in air, suggesting that the as-deposited amorphous phase crystallizes to anatase TiO2. Crystallization and phase transformation were also investigated using a powder X-ray diffractometer (PXRD; Rigaku
Fig. 1 (a) XRD profiles of the as-deposited thin film and of others after annealing at different temperatures. (b) XRD profiles of the precipitate after deposition of thin films.
Rint2100; Cu-Ka, 40 kV, 30 mA, scan rate 1u min21). For this purpose, the collected precipitate was heated at different temperatures for 1 h in air (10u min21). Powders heated at 400 uC had a light yellow color, which changed to white after annealing at 600 uC. White is characteristic of crystallized TiO2, and the amorphous–crystalline transformation was assumed by eye to be completed at 600 uC. The PXRD results showed that the as-obtained powder was amorphous. Crystalline anatase was observed after annealing at temperatures as low as 300 uC and the anatase phase remained until 800 uC, which is a much higher temperature than was observed for sol–gel-derived TiO2.15 This indicates that it may be easy to prepare a thin film of well-crystallized pure anatase using the technique described here. Obvious signs of transformation to the rutile phase were found at about 850 uC, when some small diffraction peaks due to rutile were observed. After annealing at 900 uC, about 39% rutile was formed, as evaluated by the relative intensity of the XRD peaks due to anatase (101) and rutile (110). The full width at half-maximum of the anatase (101) diffraction peak decreased with increasing temperature, suggesting that the average particle size is increased by annealing. XPS quantitative analysis For the as-deposited thin film, XPS analysis (see Fig. 2) shows Ti2p3/2 and Ti2p1/2 located at 458.75 and 464.3 eV, respectively. The Ti2p3/2 binding energy exceeds that of Ti metal (454.0 eV), TiO (455.0 eV), and Ti2O3 (456.7 eV), but is similar to that of TiO2 (458.4–458.7 eV),5 which suggests that Ti is in the 14 oxidation state and directly bonded to oxygen. The Ti2p spectrum showed almost no obvious change after Ar1 bombardment for 30–150 min, which probably implies that the thin film is homogeneous in depth. The O1s spectrum (see inset of Fig. 2) of the as-deposited film consists of a main J. Mater. Chem., 2003, 13, 608–613
609
Fig. 2 XPS spectra of the as-deposited thin film; inset is a close-up of the O1s range for the as-deposited thin film before (a) and after (b) Ar1-bombardment for 30 min.
peak at about 530.25 eV and an obvious shoulder located at about 532.54 eV. The former can be attributed to TiO25 and the latter confirms the presence of water.9b The shoulder peak became weak after 30 min Ar1 bombardment, indicating the detected water may be adsorbed on the surface of the film. Only a very small peak for N1s was detected on the surface of the as-deposited film and it disappeared after 30 min Ar1 bombardment, suggesting that the detected nitrogen existed only in the surface region, probably because of contamination. No Ba3d or Si2p signals (due to the substrate) were detected, even after 150 min Ar1 bombardment, suggesting that the film is dense and completely covers the glass substrate. Quantitative XPS analysis showed a Ti : O : C : N molar ratio of 1 : 2.6 : 0.49 : 0.117 at the surface of the thin film, which changed slightly to 1 : 2.47 : 0.24 : 0.08 after 150 min Ar1bombardment. Both carbon and nitrogen were detected due to contamination, but the Ti : O atomic ratio is smaller than the stoichiometric ratio of 1 : 2 for TiO2. Thermal behavior The total weight loss (see Fig. 3) in the temperature region up to 1000 uC investigated by TG analysis is about 29% of the initial sample weight and mostly occurs below 400 uC. There is no appreciable weight loss beyond 400 uC. The weight loss consists of two distinct steps: about 21% at 250 uC and 29% at 350 uC. In the DTA curve, two obvious thermal effects are detected; an endothermic peak at 100 uC and a broad exothermic peak centered at 780 uC. The former is accompanied by obvious weight loss, while no weight change corresponding to the latter exothermic peak occurs. Therefore, these are likely attributable to the release of adsorbed water and the phase transformation of anatase to rutile, respectively. Meanwhile, we also examined the weak thermic effects appearing at 170–340 uC. The exothermic peak at around 250 uC confirms the crystallization of the amorphous phase to anatase and, hence, crystallization should have indeed occurred after heating at 400 uC for our thin films, although no clear
Fig. 3 TG-DTA curves for the precipitate collected after deposition of thin films.
610
J. Mater. Chem., 2003, 13, 608–613
diffraction peak appears in the film XRD profile [see Fig. 1(a)]. After annealing at 500 uC, the diffraction peak is still weak, which may be because our films are composed of very small particles. The gel–crystalline conversion temperature has been reported to be 360–410 uC for TiO2 thin films prepared by the sol–gel method.15 However, direct observation of the exothermic effect in the DTA curve has been difficult, probably because the crystallization temperature is usually close to the temperatures for decomposition of residual hydroxyl groups and removal of chemisorbed water, which demonstrate endothermic effects. For our precipitate collected after deposition of the thin films, a small exothermic effect was observed, but it was still weak, which might be because the exothermic effect was partly counteracted by the endothermic effect of peroxide decomposition. We must consider the other small DTA peaks. The peroxo (O–O) group cannot be decomposed completely by annealing at 350 uC,16 so we consider the weight loss at 240–400 uC (see Fig. 3) may result from the decomposition of the peroxo group, which enables us to accurately evaluate the oxygen content contained in the as-deposited thin film. FT-IR characterization In the FT-IR spectrum (see Fig. 4) of the collected precipitate, the broad peak appearing at 3100–3600 cm21 is assigned to the fundamental stretching vibration of hydroxyl groups (free or bonded),17a which is further confirmed by the weak band at about 1620 cm21.18 This absorption band is caused by the bending vibration of coordinated H2O as well as Ti–OH. The bending vibrational mode of water may appear as shoulders on the spectrum, such as that at 3240 cm21.17a The peaks located at 500 and 430 cm21 are likely due to the vibration of the Ti–O bonds in the TiO2 lattice.17 The peak centered at 900 cm21 may be assigned to the characteristic O–O stretching vibration in H2O2,16a thus, the shoulder observed at 690 cm21 may be due to the vibration of the Ti–O–O bond. Although the peak detected at 1409 cm21 could not be assigned, the FT-IR spectrum definately confirms the presence of Ti–O bonds, peroxo groups, and OH groups in the precipitate. Based on the FT-IR analysis, a possible chemical formula for the as-deposited thin film is TiO2 2 x 2 0.5y(O2)x(OH)y?zH2O. Chemical composition of the as-deposited TiO2 thin film Attributing the weight loss observed by TGA of 8% at 250– 400 uC to the decomposition of peroxo groups, x and y were calculated to be about 0.5 and 0.2, respectively. The amount of adsorbed water was evaluated from the weight loss of 21% before 250 uC. According to XPS analysis, the intensity of the O1s peak assigned to the adsorbed water decreased after Ar1 bombardment, suggesting that the adsorbed water exists on the surface of the film. Thus, we inferred that the as-deposited thin film had a composition of TiO1.4(O2)0.5(OH)0.2?1.34H2O, which is different from those proposed for the thin films deposited from TiCl4–H2O2 aqueous solutions by the electrochemical
Fig. 4 FT-IR spectrum of the precipitate collected after deposition of thin films.
method9a or SAM technique,9b where the hydroxide group does not directly bond to titanium. Morphology and roughness Fig. 5 shows AFM images of both the as-deposited thin film and those after annealing (5 uC min21) at various temperatures for 1 h. The as-deposited thin film [see Fig. 5(a) and (f)] is composed of closely packed nanoparticles of about 10–15 nm in diameter. The statistical root mean square (RMS) roughness for the measured area (about 100 6 100 nm2) is 2–3 nm, depending on the investigated area, corresponding to 2–3% of the film thickness, which was measured by ellipsometry to be 117 nm. Such topography and roughness are typical for the whole specimen (1 6 1 cm2, measured areas are 100 6 100 or 300 6 300 nm2) and were easily reproduced. For the annealed films, the RMS (sample areas 1 6 1 cm2, measured area 300 6 300 nm2) changes a little with temperature, but the largest value is still less than 7 nm. However, cracks are also observed. Cross-sectional profile analysis of the roughness (see Fig. 5) shows the largest depth of cracks is 26 nm [see Fig. 5(d)], which is much smaller than the thickness of the film (108 nm), suggesting that the cracks are superficial and do not extend throughout the film. Although closely packed particles are observed in the AFM images, aggregation must have occurred under our deposition conditions. Fig. 6 shows SEM images of the as-deposited thin film and that after annealing at 700 uC. The as-deposited thin film exhibits a smooth surface. After annealing, grains of the
order of tens nanometers can be seen, which is larger than those observed by AFM, suggesting the occurrence of aggregation. Even the grain sizes we observed by AFM for our as-deposited thin film, 10–15 nm in diameter, or those annealed at different temperatures are still larger than the mean particle size we estimated using the Scherrer equation (see following section). Hence, aggregation must have occurred. Many cracks can be also observed, indicating that further efforts need to be made to prepare a crack-free TiO2 thin film. Optical transmission spectra and band gaps The optical transmission spectra shown in Fig. 7(a) demonstrate that the as-deposited film and those annealed at different teperatures are highly transparent (transmittance ¢ 82%) in the visible region. The high transmittances indicate fairly smooth surfaces and relatively good film homogeneity, which is consistent with the AFM observations. For the high energy absorption region, the relation between transmittance (T) and absorption coefficient (a) can be expressed as a ~ 2ln(T)/d, where d is the film thickness.19 The relation between absorption coefficient and incident photon energy (hv) can be written as (ahv) ~ A1(hv 2 Eg1)1/2 and (ahv) ~ A2(hv 2 Eg2)2 for allowed direct and indirect transitions, respectively, where A1 and A2 are constants, and Eg1 and Eg2 are the direct and indirect band gaps, respectively.20 The energy band structure of rutile has been extensively studied. Both experimental and theoretical calculations suggest that rutile TiO2 has a direct forbidden gap (3.03 eV) which is
Fig. 5 AFM images of the as-deposited thin film (a), along with those after annealing at different temperatures: (b) 400; (c) 500; (d) 600; (e) 700 uC. (f) Enlarged three-dimensional image of the as-deposited film.
Fig. 6 SEM photographs of the as-deposited thin film (a) and that after annealing at 700 uC (b).
J. Mater. Chem., 2003, 13, 608–613
611
annealing temperature was not observed for our 117 nm TiO2 thin film heated at above 500 uC. We also tried to identify the changes in band gap energies with the film morphology. As shown in Fig. 5, the as-deposited thin film was composed of closely packed nanometer-sized grains of 10–15 nm in size, which is remarkable because nanostructured titania exhibits important properties and is now the subject of intensive investigation. Annealing (5 uC min21 in air) resulted in an increase in grain size. Note that the grain size remained nanometer-scale even after annealing at 700 uC. Band gap energies are affected by particle size because of quantum-size effects.21,24,25 For our crystallized thin film, the particle size, evaluated by the Scherrer equation [Dhkl ~ 0.89l/(bhklcosh)], is less than 2 nm, even after annealing at 700 uC, but no obvious increase in band gap energy was observed.
Conclusions A new environmentally friendly aqueous solution system for fabricating TiO2 thin films on glass substrates at room temperature has been successfully developed. Compared with the conventional approaches for growth of TiO2 thin films, the present method is simple, safe, cost-effective, and reproducible. The as-deposited film is a transparent amorphous thin film composed of densely packed nanometer-sized grains, and the chemical composition TiO1.4(O2)0.5(OH)0.2?1.34H2O is proposed based on XPS, FT-IR and TG-DTA data. Anatase TiO2 thin films were obtained by heating the as-deposited thin films at 500 uC for 1 h in air. After crystallization, band gap energies of 3.20 eV for indirect transition and 3.63 eV for direct transition were obtained.
Fig. 7 Optical transmittance spectra (a), indirect transition band gap energies (b), and direct transition band gap energies (c) of the asdeposited TiO2 thin film and those after annealing. The insets in (b) and (c) are derived from the data in the corresponding figures.
almost degenerate with an indirect allowed transition (3.05 eV).21 In contrast, there are few experimental results or calculations available on the band structure of anatase.20a For this reason, many studies assume that the optical properties of anatase are similar to those of rutile and evaluate the indirect band gap of anatase thin films,20a,22 but there are still several papers on the evaluation assuming a direct band gap.17d,22b,e,23 Here, we evaluated the optical band gap of our TiO2 thin film assuming both direct and indirect transitions. The (ahv)2 or (ahv)1/2 versus hv curves for the as-deposited thin film and those annealed at different temperatures are shown in Fig. 6(b) and (c). Band gap energies were evaluated by extrapolating the linear parts of the curves to the energy axis. For the as-deposited thin film, both the indirect one at 3.63 eV [see Fig. 7(b)] and the allowed direct optical absorption band at about 4.20 eV [see Fig. 7(c)] were found. The band gap energy decreases on high temperature annealing and becomes saturated at 3.20 eV [see Fig. 7(b)] for indirect transition and at 3.63 eV [see Fig. 7(c)] for direct transition, which are comparable with the reported values for anatase TiO2 thin films (indirect transition 3.20,20a 3.23,22a 3.26 eV;22d direct transition 3.67,16a 3.40,17d 3.80,22a 3.65 eV22e), but larger than the values reported for rutile thin films (3.05 eV21). We can infer that the band gap energies are partly dependent on the crystallinity of the thin films, as the band gap energies for the well-crystallized thin films are comparable with those of crystallized bulk materials, whereas for the amorphous or poorly crystallized ones, the band gap energies are higher than those of the corresponding bulk materials. This finding is supported by other studies.11c,16b,19 However, the shift of the absorption edge to higher wavelength with increasing 612
J. Mater. Chem., 2003, 13, 608–613
Acknowledgement One of us (Y.-F. G.) is grateful to NGK Insulators Ltd. of Japan for supporting a scholarship.
References 1 P. S. Peercy, Nature, 2000, 406, 1023. 2 (a) B. O’Regan and M. Gra¨tzel, Nature, 1991, 353, 737; (b) U. Bach, D. Lupo, P. Comte, J. E. Moser, E. Weissortel, J. Salbeck, H. Spreitzer and M. Gra¨tzel, Nature, 1998, 395, 583. 3 (a) S. K. Zheng, T. M. Wang, W. C. Hao and R. Shen, Vacuum, 2002, 65, 155; (b) A. L. Linsebigler, L. Guangquan and J. T. Yates Jr., Chem. Rev., 1995, 95, 735. 4 (a) J. W. Klaus, O. Sneh and S. M. Goerge, Science, 1997, 278, 1934 and references therein; (b) M. L. Hitchman and S. E. Alexandrov, Interface, 2001, 10, 40. 5 Y. Masuda, Y. Jinbo, T. Yonezawa and K. Koumoto, Chem. Mater., 2002, 14, 1236. 6 (a) Y. Zheng, E. Shi, Z. Chen, W. Li and X. Hu, J. Mater. Chem., 2001, 11, 1547; (b) K. J. Kim, K. D. Benkstein, J. V. D. Lagemaat and A. J. Frank, Chem. Mater., 2002, 14, 1042. 7 K. Shimizu, H. Imai, H. Hirashima and K. Tsukuma, Thin Solid Films, 1999, 351, 220. 8 (a) K. Koumoto, K. Seo, T. Sugiyama, W. S. Seo and W. J. Dressick, Chem. Mater., 1999, 11, 2305; (b) S. Deki, Y. Aoi, O. Hiroi and A. Kajinami, Chem. Lett., 1996, 6, 433. 9 (a) I. Zhitomirsky, L. Gal-Or, A. Kohn and H. W. Hennicke, J. Mater. Sci., 1995, 30, 5307; (b) T. P. Niesen, B. Joachim and A. Fritz, Chem. Mater., 2001, 13, 1552; (c) R. S. Sonawane, S. G. Hegde and M. K. Dongare, Mater. Chem. Phys., 2002, 9451, 1; (d) M. Tada, Y. Yamashita, V. Petrykin, M. Osada, K. Yoshida and M. Kakihana, Chem. Mater., 2002, 14, 2845. 10 (a) V. Kumar, J. Am. Ceram. Soc., 1999, 82, 2580; (b) S. Gijp, L. Winnubst and H. Verweij, J. Mater. Chem., 1998, 8, 1251; (c) G. R. Fox, J. H. Adair and R. E. Newnham, J. Mater. Sci., 1990, 25, 3634; (d) J. A. Navio, F. J. Marchena, M. Macias, P. J. Sanchez-soto and P. Pichat, J. Mater. Sci., 1992, 27, 2643; (e) E. R. Camargo and M. Kakihana, Chem. Mater., 2001, 13, 1181. 11 (a) F. P. Rotzinger and M. Gra¨tzel, Inorg. Chem., 1987, 26, 3704;
12 13 14 15
16 17
(b) J. Mu¨hlebach, K. Mu¨ller and G. Schwarzenbach, Inorg. Chem., 1970, 9, 2381; (c) M. Mori, N. Shibata, E. Kyono and S. Ito, Bull. Chem. Soc. Jpn., 1956, 29, 904. Y. Narendar and G. L. Messing, Chem. Mater., 1997, 9, 580. M. Kakihana, M. Tada, M. Shiro, V. Petrykin, M. Osada and Y. Nakamura, Inorg. Chem., 2001, 40, 891. J.-P. Jolivet, M. Henry, J. Livage and E. Bescher, Metal Oxide Chemistry and Synthesis, John Wiley & Sons, Chichester, 1994, p. 53. (a) Q. Xu and M. A. Anderson, J. Am. Ceram. Soc., 1994, 77, 1939; (b) A. M. Tonejc, A. Turkovic´, M. Gotic´, S. Music, M. Vukovic, R. Trojko and A. Tonejc, Mater. Lett., 1997, 31, 127; (c) Q. Xu and M. A. Anderson, J. Mater. Res., 1991, 6, 1073; (d) A. G. Gaynor, R. J. Gonzales, R. M. Davis and R. Zallen, J. Mater. Res., 1997, 12, 1755; (e) K. I. Gnanasekar, V. Subramanian, J. Robinson, J. C. Jiang, F. E. Posey and B. Rambabu, J. Mater. Res., 2002, 17, 1507; (f) C. L. Fan, D. Ciardullo, J. Paladino and W. Huebner, J. Mater. Res., 2002, 17, 1520. (a) T. Ivanova, A. Harizanova and M. Surtchev, Mater. Lett., 2002, 55, 327; (b) H. Wang, M. Zhang, S. Yang, L. Zhao and L. Ding, Sol. Energy Mater. Sol. Cells, 1996, 43, 345. (a) R. Zhang and L. Gao, Key Eng. Mater., 2002, 224–226, 573; (b) T. Yoko, K. Kamiya and K. Tanaka, J. Mater. Sci., 1990, 25, 3922; (c) J. Zhang, I. Boyd, B. J. O’Sullivan, P. K. Hurley, P. V. Kelly and J.-P. Senateur, J. Non-Cryst. Solids, 2002, 303, 134;
18 19 20
21 22
23 24 25
(d) L. Castan˜eda, J. C. Alonso, A. Ortiz, E. Andrade, J. M. Saniger and J. G. Ban˜uelos, Mater. Chem. Phys., 2003, 77, 938. B. Klingenberg and M. A. Vannice, Chem. Mater., 1996, 8, 2755. L. J. Meng and M. P. Dos Santos, Thin Solid Films, 1993, 226, 22. (a) H. Tang, K. Prasad, R. Sanjines, P. E. Schmid and F. Levy, J. Appl. Phys., 1994, 75, 2042; (b) D. Bao, X. Yao, N. Wakiya, K. Shinozaki and N. Muzutani, Appl. Phys. Lett., 2001, 79, 3767; (c) J. Callaway, Quantum Theory of the Solid State, Academic, New York, 1974, p. 516. (a) J. Pascual, J. Camassel and M. Mathieu, Phys. Rev. B, 1978, 18, 5606; (b) N. Daude, C. Gout and C. Jouanin, Phys. Rev. B, 1977, 15, 3229. (a) D. G. Syarif, A. Miyashita, T. Yamaki, T. Sumita, Y. Choi and H. Itoh, Appl. Surf. Sci., 2002, 193, 287; (b) Z. Wang, U. Helmersson and P.-O. Ka¨ll, Thin Solid Films, 2002, 405, 50; (c) H. Takikawa, T. Matsui, T. Sakakibara, A. Bendavid and P. J. Martin, Thin Solid Films, 1999, 348, 145; (d) G. K. Boschloo, A. Goossens and J. Schoonman, J. Electrochem. Soc., 1997, 144, 1311; (e) A. Aoki and G. Nogami, J. Electrochem. Soc., 1996, 143, L191. T. Yoko, A. Yuasa, K. Kamiya and S. Sakka, J. Electrochem. Soc., 1991, 138, 2279. (a) S. Zheng, T. Wang, G. Xing and C. Wang, Vacuum, 2001, 62, 361; (b) Q. Zhou, Q. Zhang, L. Zhang and X. Yao, Mater. Lett., 2002, 54, 21. A. Linsebigler, G. Lu and J. J. Yates, Chem. Rev., 1995, 95, 735.
J. Mater. Chem., 2003, 13, 608–613
613
