4712
Langmuir 2007, 23, 4712-4714
Thick Transparent Rutile TiO2 Films Crystallized in Solution Y.-F. Gao,*,† M. Nagai,† W. S. Seo,‡ and K. Koumoto§ Musashi Institute of Technology, AdVanced Research Laboratories, Tokyo 158-0082, Japan, Korea Institute of Ceramic Engineering and Technology, Gumcheon-Gu, Seoul, Korea, and Graduate School of Engineering, Nagoya UniVersity, Nagoya 464-8603, Japan ReceiVed December 6, 2006. In Final Form: February 13, 2007 We report a novel process for the preparation of dense, transparent TiO2 films of 2.5 µm thickness on a F-doped SnO2-covered glass substrate. The starting solution contained peroxotitanate complex ions, which are relatively stable under the experimental conditions, permitting the deposition of highly textured rutile nanocrystalline films. The nanocrystals exhibit specific orientations along the (101) and (002) crystalline planes. Kinetic studies suggest that the precipitation started from the formation of amorphous solids, followed by crystallization through a dissolutionrecrystallization process. Although a minor phase of anatase was detected only for powders collected from solutions after film preparation, not for films, the transformation from amorphous to anatase was believed to occur before further transformation of anatase to rutile. The present method enables film synthesis on a surface with a large area, and therefore could be integrated into the processing of electroluminescent devices.
Low-dimensional TiO2 nanostructures with controllable crystalline phases show unique optical, electronic, and bioactive properties.1,2 Due to their high surface-to-volume ratios, this kind of material is especially attractive for application to photocatalysts,1 supports for heterogeneous catalysts,3 electrodes for dye-sensitized solar cells,4 and candidate materials for H2 storage.5 Soft solution approaches to synthesize these well-defined TiO2 nanostructures mainly comprise surfactant-directed methods,2,6 hydrothermal treatment of anatase powders or their precursors,3a,7 and templating8 using porous aluminum,8a-e polymer fibers,8f organic gelators,8g or inorganic rods such as ZnO.8h Preparation of crack-free, highly transparent in the visible range, thick (micrometer-order), crystalline TiO2 films over large areas is of importance to a range of optoelectronic applications. TiO2 nanorod films have been recently synthesized by hydro†
Advanced Research Laboratories. Korea Institute of Ceramic Engineering and Technology. § Nagoya University. ‡
(1) (a) Manna, L.; Scher, E. C.; Li, L. S.; Alivisatos, A. P. J. Am. Chem. Soc. 2002, 124, 7136. (b) Ma, R.; Sasaki, T.; Bando, Y. J. Am. Chem. Soc. 2004, 126, 10382. (c) Patzke, G. R.; Krumeich, F.; Nesper, R. Angew. Chem., Int. Ed. 2002, 41, 2446. (2) Zhong, Z.; Ang, T.-P.; Luo, J.-Z.; Gan, H.-C.; Gedanken, A. Chem. Mater. 2005, 17, 6814. (3) (a) Wen, B.; Liu, C.; Liu, Y. Inorg. Chem. 2005, 44, 6503. (b) Guo, Y. G.; Wan, L.-J.; Bai, C.-L. J. Phys. Chem. 2003, 107, 5441. (c) Cozzoli, P. D.; Curri, M. L.; Agostiano, A. Chem. Commun. 2005, 3186. (4) Chou, T. P.; Fryxell, G. E.; Li, X.; Cao, G. Proc. SPIEsInt. Soc. Opt. Eng. 2004, 5510, 129. (5) Lim, S. H.; Luo, J.; Zhong, Z. Y.; Ji, W.; Lin, J. Inorg. Chem. 2005, 44, 12, 4124. (6) (a) Chemseddine, A.; Moritz, T. Eur. J. Inorg. Chem. 1999, 235. (b) Kanic, K.; Sugimoto, T. Chem. Commun. 2004, 1584. (c) Sugimoto, T.; Okada, K.; Itoh, H. J. Colloid Interface Sci. 1997, 193. (7) (a) Kasuga, T.; Hiramatsu, M.; Hoson, A.; Sekino, T.; Niihara, K. Langmuir 1998, 14, 3160. (b) Kasuga, T.; Hiramatsu, M.; Hoson, A.; Sekino, T.; Niihara, K. AdV. Mater. 1999, 11 (15), 1307. (c) Zhu, Y.; Li, H.; Koltypin, Y.; Hacoben, Y. R.; Gedanken, A. Chem. Commun. 2001, 2616. (d) Du, G. H.; Chen, R. C.; Che, R. C.; Yuan, Z. Y.; Peng, L. M. Appl. Phys. Lett. 2001, 79, 32. (e) Tian, Z. R.; Voigt, J. A.; Liu, J., Mckenzie, B.; Xu, H.-F. J. Am. Chem. Soc. 2003, 125, 12384. (8) (a) Imai, H.; Takei, Y.; Simizu, K.; Matsuda, M.; Hirashima, H. J. Mater. Chem. 1999, 9, 2917. (b) Miao, Z.; Xu, D.; Ouyang, J.; Guo, G.; Zhao, X.; Tang, Y. Nano Lett. 2002, 2, 717. (c) Chu, S. Z.; Wada, K.; Inoue, S.; Todoroki, S. Electrochim. Acta 2003, 48, 3147. (d) Lei, Y.; Zhang, L. D. J. Mater. Res. 2001, 16, 1138. (e) Shyue, J.-J.; Padture, N. P. Mater. Lett. 2007, 61, 182. (f) Caruso, R. A.; Schattka, J. H.; Greiner, A. AdV. Mater. 2001, 13, 1577. (g) Kobayashi, S.; Hanabusa, K.; Hamasaki, N.; Kimura, M.; Shirai, H. Chem. Mater. 2000, 12, 1523. (h) Lee, J.-H.; Leu, I.-C.; Hsu, M.-C.; Chung, Y.-W.; Hon, M.-H. J. Phys. Chem. B 2005, 109, 13056.
thermal treatment at 200 °C of TiCl3 aqueous solutions in the presence of NaCl.9a A highly crystalline rectangular parallelepiped rutile TiO2 on the submicrometer scale (150-250 nm in width and 3-4 µm in length) was prepared on a glass substrate. On a functionalized substrate, transparent, oriented nanocrystalline anatase films of 100-200 nm in thickness were also formed using TiCl3 as a starting material.9b However, either thickness or transparency alone is not enough for application in electroluminescent devices. We report here a highly oriented, transparent rutile TiO2 film synthesized at a normal pressure and a low temperature (95 °C) on a SnO2:F substrate. The process was initiated with preparation of a peroxotitanium complex solution using H2TiO3 (Mitsuwa Chem.), H2O2 (Kishida, 30% in water), and ammonia (NH3, Kishida, 28% in water) as raw materials. The detailed procedure for preparation of stock solutions has been reported previously.10 The deposition solution contained 4 mM Ti4+ in a total volume of 100 mL solution. The solution pH ()1) was regulated by adding an appropriate amount of either acids (HCl, HNO3, or H2SO4) or bases (NaOH), and all solutions were reagent grade purchased from the Wako Chemical Corporation. The SnO2:F substrate (Asahi Glass, 12 Ω/0, 1 × 2 cm2) was first cleaned with acetone, ethanol, and distilled water. The cleaned substrates were hung vertically in the deposition solution with the back side against the beaker wall. Deposition was conducted at 95 °C for 12-144 h. After deposition, the film was rinsed with distilled water and dried at 50 °C. The morphology of the film was characterized with a scanning electron microscope (FE-SEM, JSM-6700F, JEOL). A flat homogeneous surface with small cracks was observed at low magnification (Figure 1a). The appearance of small cracks can be attributed to the specific growth mode. The spaces between neighboring nanowire-like deposits greatly decreased the stress occurring during drying. At high magnification, the film surface exhibits a particulate characteristic with particle sizes around (9) (a) Hosono, E.; Fujihara, S.; Kakiuchi, K.; Imai, H. J. Am. Chem. Soc. 2004, 126, 7790. (b) Wang, D.; Liu, J.; Huo, D.; Nie, Z.; Lu, W.; Williford, R. E.; Jiang, Y. J. Am. Chem. Soc. 2006, 128, 13670. (10) (a) Gao, Y.-F.; Masuda, Y.; Peng, Z.-F.; Yonezawa, T.; Koumoto, K. J. Mater. Chem. 2003, 13, 608. (b) Gao, Y.-F.; Masuda, Y.; Koumoto, K. Chem. Mater. 2004, 16, 1062. (c) Gao, Y.; Masuda, Y.; Koumoto, K. Langmuir 2004, 20, 3188. (d) Gao, Y. F.; Nagai, M.; Seo, W.; Koumoto, K. J. Am. Ceram. Soc. 2007, 90, 831.
10.1021/la0635404 CCC: $37.00 © 2007 American Chemical Society Published on Web 03/21/2007
Letters
Langmuir, Vol. 23, No. 9, 2007 4713
Figure 3. UV-vis spectrum and optical band gap (inset) of the as-deposited rutile TiO2 film.
Figure 1. SEM micrographs (a-c) and XRD patterns (d) of TiO2 thin film prepared on a glass substrate precoated with a transparent conducting oxide (FTO) layer. (a) Low magnification. (b) High magnification. (c) Cross-sectional. R refers to the rutile phase. The red XRD pattern is for the FTO substrate.
Figure 2. HRTEM graphs of a rutile TiO2 film along [211] direction. (A) Low-magnification cross-sectional image. (B) High-magnification image.
10 nm in diameter. In some parts of the surface, the particles were observed to be arranged into a rod shape (Figure 1b). In a cross-sectional view, vertically arranged bundles are present lengthwise against the substrate surface (Figure 1c). This characteristic is also obvious in a TEM micrograph (Figure 2). The thickness is about 2.5 µm. Moreover, unlike what appeared on the surface, the deposits seen from cross sections are smooth. This finding may imply that the rods probably grew by orientationattachment of small particles, preferably at the tips of rods (Figure 2b). All these observations suggest that the growth might be along the direction perpendicular to the substrate. A randomly distributed layer of spherical particles was also present in the area between the oriented bundle and the substrate, implying that the initial precipitates may act as a seed layer to induce the
subsequently oriented growth. In addition, the film adheres well to the substrate as checked by the Scotch tape method. The film demonstrates a mean transparency of about 70-80% in the visible range and an optical band gap of 3.15 eV, slightly larger than the values reported for rutile TiO2 thin films (3.05 eV11) (Figure 3). The X-ray diffraction (XRD, Rigaku 2500, 40 kV, 300 mA) pattern in Figure 1d suggests that the as-deposited film shows enhanced (101) and (002) diffraction peaks of rutile TiO2, in accordance with the SEM observations and Raman spectra (Supporting Information Figure S1). Growth with these specific orientations was also confirmed for both powders collected after treatment and films using plastic or polymer substrates (without FTO layers), suggesting that it is not dependent on the crystallography of substrates. Further observation with a transmission electron microscope (TEM, 400 kV, JEM4010) confirmed the column-shaped growth along the direction perpendicular to the substrate surface (Figure 2A). Dissolution of the preformed amorphous solids supports limited ions in solution, permitting oriented growth of highly transparent rutile TiO2 under a relatively low-level degree of solution supersaturation. High-resolution TEM micrographs clearly show the lattice fringes of the (200) and (210) planes of rutile TiO2 observed from the [211] direction (Figure 2B) and of the (110) and (002) planes observed from the [110] direction (Supporting Information Figure S2). Image simulation of the structure of an ideal rutile crystal agrees well with that observed by HRTEM along the [211] direction (Supporting Information Figure S3).12 The single-crystalline nature of rutile nanocrystals is confirmed by the selected area electron diffraction (SAED) pattern, which is sharp and consists of a group of isolated spots (Figures 2 and S2). For the powders collected in solution after deposition, we further carried out time-dependent studies on their crystallinity, color, and morphology. The deposition starts from the formation of a porous amorphous TiO2 solid with a light yellowish color in the initial stage (time 19 h), implying changes in chemical composition and crystalline states. Actually, the yellowish precipitates were confirmed to be amorphous by XRD, and the hydroxyl group residue was also suggested by X-ray photoelectron spectra. For the white solid, the crystallized rutile was present. For some precipatates collected after treating for different times, a minor phase of anatase was also detected by XRD. These results suggest that it undergoes a transformation from amorphous to anatase (11) (a) Pascual, J.; Camassel, J.; Mathieu, M. Phys. ReV. B 1978, 18, 5606. (b) Daude, N.; Gout, C.; Jouanin, C. Phys. ReV. B 1977, 15, 3229. (12) Simulation by the multislice method was conducted on a high-resolution TEM (model JEM-4010; JEOL Ltd., Tokyo, Japan) with a point resolution of 0.15 nm. The accelerating voltage and the defocus were 400 kV and 40 nm, respectively. The specimen thickness was 10 nm.
4714 Langmuir, Vol. 23, No. 9, 2007
before ending with the final product rutile. In the peroxotitanium complex solution from an alternative preparation method, an anatase-containing TiO2 film was also successfully grown on a Si substrate covered with self-assembled monolayers by the continuous-flow technique. The anatase TiO2 was found to be a result of transformation from amorphous solids and to appear after a relatively long processing time.13 When treating the peroxotitanium solution in the presence of a small amount of ammonium vanadate, an anatase phase was also prepared at about 60 °C.14 The transformation between polymorphs of TiO2 involves fundamental issues of kinetics and thermodynamics and has been intensely investigated both experimentally and theoretically.15-17 Because of the difference in surface energy, different polymorphs are more or less stable. Transformation among these polymorphs needs fundamentally to reduce the surface energy through coarsening by coalescence of neighboring particles. This process commonly occurs during heating. In the soft chemical processes, metastable amorphous or anatase TiO2 on the nanometer scale is usually preferably formed.18 The transformation to the thermodynamically stable phase rutile requires reaching to a critical size that is closely dependent on the chemical species and reaction conditions.15-17 All the above analyses enable us to partially understand the formation mechanism of these nanocrystals. For our solution system, despite the fact that the rutile is a thermodynamically stable phase, the gradual transformation from amorphous to rutile through anatase suggests that crystallization to rutile in this case maybe undergo a kinetically controlled dissolution-recrystallization process. As described above, although the formation of solids at current pH is significantly delayed, the amorphous precipitates were still produced after retention at 95 °C for a long period of time. Amorphous TiO2-based solids prefer to form in the peroxotitanium complex solution. These solids vary in chemical composition under different preparation conditions. In some cases, even residues of peroxo groups were also detectable. Gradual release of these impurities occurs smoothly under this processing followed by the crystallization into anatase, which is a spontaneous reaction. Further increases in the particle size of anatase have reduced surface energy and promoted the transformation of anatase to rutile. Experimental results on the transformation of these polymorphors of TiO2 with heating suggest that the temperature for the transformation from amorphous to anatase is much lower than that from anatase to rutile.10a However, only limited anatase minor phase (from powders collected in solution) were detected in our experiments. This result reveals that the transformation from anatase to rutile seems much faster than that from amorphous to anatase. (13) Fuchs, T. M.; Hoffmann, R. C.; Niesen, T. P.; Tew, H.; Bill, J.; Aldinger, F. J. Mater. Chem. 2002, 12, 1597. (14) Hoffmann, R. C.; Jeurgens, L. P. H.; Wildhack, S.; Bill, J.; Aldinger, F. Chem. Mater. 2006, 18, 4465. (15) Gribb, A. A.; Banfield, J. F. Am. Mineral. 1997, 82, 717-728. (16) Zhang, H.; Banfield, J. F. Am. Mineral. 1999, 84, 528-535. (17) Zhang, H.; Banfield, J. F. J. Mater. Res. 2000, 15, 437-448. (18) Gao, Y. F.; Koumoto, K. Cryst. Growth Des. 2005, 5, 1983.
Letters
The present method is capable of producing dense, transparent rutile films under relatively mild conditions compared to the hydrothermal method developed for the synthesis of rutile nanorods, in which the addition of NaCl into a TiCl3 (not TiCl4) aqueous solution is critical for the development of 1D growth.9a Furthermore, the film is thick on the micrometer order while retaining sufficient transparency. The formation of similar nanostructures using HCl, H2SO4, or HNO3 to regulate the solution pH has been confirmed. In addition, the existence of NH4+ in the solution is not necessary, because even when the base was changed from ammonia to NaOH, similar nanostructures were obtained. However, for the electroluminescence application, further decreasing the small cracks to improve transparency is still needed. From a liquid-phase deposition processing, Sukenik et al. have succeeded in obtaining crack-free TiO2 thin films in thickness up to 0.5 µm by using drying procedures that combine temperature and humidity control.19 Using a solvent with a low surface tension to rinse the film obviously reduced the appearance of cracks or inhibited their development due to changes in capillary stresses.20 This reported understanding in controlling the occurrence of cracks should be adequate for the films from the current solution system. In summary, we report a novel process for the preparation of dense, transparent TiO2 films of 2.5 µm thickness on a F-doped SnO2-covered glass substrate. The nanocrystals exhibit specific orientations along the (101) and (002) crystalline planes. Kinetic studies suggest that the precipitation started from the formation of amorphous solids, followed by crystallization through a dissolution-recrystallization process. Although a minor phase of anatase was detected only for powders collected from solutions after film preparation, not for films, the transformation from amorphous to anatase was believed to occur before further transformation of anatase to rutile. The present film can be synthesized over a large surface and therefore could be integrated into electroluminescent devices. This work gives an example that one can control many important properties such as crystalline phase, thickness, transparency, and crack occurrence in a single treatment solution under optimized conditions, which is usually considered to be a challenging task for a solution deposition process. Acknowledgment. This work was supported in part by the Murata Science Foundation and by a grant-in-aid for scientific research from the Ministry of Education, Culture, Sports, Science, and Technology of Japan to promote multidisciplinary research projects. Supporting Information Available: Characterization of Raman spectra, HR-TEM, and simulation (Figures S1-S3). This material is available free of charge via the Internet at http://pubs.acs.org. LA0635404 (19) Razgon, A.; Sukenik, C. N. J. Mater. Res. 2005, 20, 2544. (20) Goh, G. K. L.; Donthu, S. K.; Pallathadka, P. K. Chem. Mater. 2004, 16, 2857.
