Effects of Template and Precursor Chemistry on Structure and

Effects of aging conditions on the structural properties of mesoporous SiO2/TiO2 composite materials with crystallized framework. Hua Li , Wei-hua She...
0 downloads 0 Views 356KB Size
Download PDF
Langmuir 2004, 20, 9095-9102

9095

Effects of Template and Precursor Chemistry on Structure and Properties of Mesoporous TiO2 Thin Films X. Shari Li, Glen E. Fryxell,* Jerome C. Birnbaum, and Chongmin Wang Pacific Northwest National Laboratory, P.O. Box 999, Richland, Washington 99352 Received June 3, 2004. In Final Form: July 7, 2004 Mesoporous TiO2 thin films were synthesized by sol-gel processing using an aqueous-based, inexpensive, and environmentally friendly precursor and cationic surfactants as templates under mild reaction conditions. The films were prepared by spin-coating on glass substrates followed by calcination to remove the surfactant. N2 sorption, X-ray diffraction, and transmission electron microscopy were used to characterize the porosity, pore size, and pore structure before and after calcination. Films were found to have wormlike pore structures after calcination and surface areas on the order of 200 m2/g. These results show that the mesostructure and porosity of the thin films can be controlled by the surfactant template chemistry such as surfactant/Ti ratio, pH, and rate of solvent evaporation.

Introduction There has been a great interest in TiO2 films for a variety of applications, for example, gas sensors,1-3 photocatalysis,4 and photoelectrodes.5-8 TiO2 thin films are commonly prepared using a sol-gel process.3,4,6,9 Sol-gel methodology is one of the most convenient technologies for preparation of oxide thin films due to its low cost, ease of execution, and low processing temperatures. For many applications, large surface area mesoporous TiO2 films are desired. Mesoporous silica films with pore sizes in the range of 1-10 nm have been synthesized by the sol-gel process using surfactant templates,11-13 in which the pores are formed in a spin-coated14-17 or dipcoated film18,19 after removal of the pore former. This molecularly templated synthetic strategy allows rational control of the porosity, pore size, pore shape, film texture, * To whom correspondence should be addressed. (1) Yamazoe, N.; Miua, N. Chem. Sens. Technol. 1992, 4, 19-42. (2) Go¨pel, W. Sens. Actuators 1996, 56, 83-102. (3) Atashbar, M. Z. IEEE-NANO 2001, 544-549. (4) Katoh, K.; Tsuzuki, A.; Torii, Y.; Taoda, H. J. Mater. Sci. 1995, 30, 837-841. (5) Regan, B. O.; Gra¨tzel, M. Nature 1991, 353, 737-739. (6) Ruhman, M. M.; Tanaka, H.; Soga, T.; Jimbo, T.; Umeno, M. IEEE 2000, 806-809. (7) Bach, U.; Lupo, D.; Comte, P.; Moser, J. E.; Weisso¨rtel, F.; Salbeck, J.; Spereitzer, H.; Gra¨tzel, M. Nature 1998, 395, 583-585. (8) Gra¨tzel, M. Nature 2001, 414, 338-344. (9) Toko, T.; Yuasa, A.; Kamiya, K.; Saka, S. J. Electrochem. Soc. 1991, 138, 2279-2285. (10) Yusuf, M. M.; Imai, H.; Hirashima, H. J. Non-Cryst. Solids 2001, 285, 90-95. (11) Yang, H.; Kuperman, A.; Coombs, N.; Mamiche-Afara, S.; Ozin, G. A. Nature 1996, 379, 703-705. (12) Yang, H.; Coombs, N.; Sokolov, I.; Ozin, G. A. Nature 1996, 381, 589-592. (13) Aksay, A.; Trau, M.; Manne, S.; Honma, I.; Yao, N.; Zhou, L.; Fenter, P.; Eisenberger, P. M.; Gruner, S. M. Science 1996, 273, 892898. (14) Liu, J.; Bontha, J. R.; Kim, A. Y.; Baskaran, S. MRS Symp. Proc. 1996, 431, 245-249. (15) Ogawa, M. Chem. Commun. 1996, 1149-1150. (16) Bruinsma, P. J.; Hess, N. J.; Bontha, J. R.; Liu, J.; Baskaran, S. MRS Symp. Proc. 1997, 445, 105-110. (17) Bruinsma, P. J.; Bontha, J. R.; Liu, J.; Baskaran, S. Mesoporoussilica films, fibers, and powders by evaporation. U.S. Patent No. 5,922,299, July 13, 1999. (18) Lu, Y.; Ganguli, R.; Drewien, C. A.; Anderson, M. T.; Brinker, C. J.; Gong, W.; Guo, Y.; Soyez, H.; Dunn, B.; Huang, M. H.; Zink, J. I. Nature 1997, 389, 364-368. (19) Zhao, D.; Yang, P.; Melosh, N.; Feng, J.; Chmelka, B. F.; Stucky, G. D. Adv. Mater. 1998, 10 (16), 1380-1385.

and thickness and can result in good mechanical properties in the film.20 This methodology has been extended to the templated syntheses of various transition metal oxides. For example, porous TiO2 films have been prepared using surfactant templating of a Ti(C4H9O)4 as a precursor.10 The micromorphologies and pore sizes of TiO2 films can be controlled by changing the type of the surfactant species. Much of the interest behind making nanostructured titania phases stems from titania’s activity as a semiconductor and photocatalyst.4,7,8 Therefore, it is important to be able to prepare mesoporous titania phases that have been chemically modified. For example, Antonelli and coworkers have doped mesoporous titania with alkali metals,21 potassium fulleride wires,22 reduced Ti species,23 and cobaltocene,24 to create low valent reactive interfaces which undergo useful chemical processes, such as the fixation of dinitrogen to form ammonia.23 Therefore, for maximum utility in this area, the chemistry involved in the synthesis of mesoporous titania thin films needs to be compatible with the incorporation of catalytically active species, photosensitizers, or nanoparticle adjuncts. In this paper, we present a convenient method to synthesize mesoporous TiO2 thin films by spin-coating using surfactants as templates and using an aqueousbased, inexpensive, and environmentally friendly precursor, using mild reaction conditions, allowing this method to be tailored to include an assortment of modification chemistries (e.g., nanoparticle inclusions). Deposition solutions are composed of surfactant templates and titanium lactate in aqueous alcohol at modest pH. The mesostructure and porosity of the thin films can be controlled by the surfactant template chemistry, which is affected by parameters such as surfactant/Ti ratio, pH, and rate of solvent evaporation. (20) Baskaran, S.; Liu, J.; Domansky, K.; Kohler, N.; Li, X.; Coyle, C.; Fryxell, G. E.; Thevuthasan, S.; Williford, R. E. Proceedings of the SEMATECH Ultra Low K Workshop; SEMATECH: Austin, TX, 1999; pp 55-80. (21) Vettraino, M.; Trudeau, M.; Antonelli, D. M. Inorg. Chem. 2001, 40, 2088-2095. (22) Ye, B.; Trudeau, M.; Antonelli, D. Chem. Mater. 2001, 13, 27302741. (23) Vettraino, M.; Trudeau, M.; Lo, A. Y. H.; Schurko, R. W.; Antonelli, D. J. Am. Chem. Soc. 2002, 124, 9567-9573. (24) Murray, S.; Trudeau, M.; Antonelli, D. M. Inorg. Chem. 2000, 39, 5901-5908.

10.1021/la0486279 CCC: $27.50 © 2004 American Chemical Society Published on Web 09/14/2004

9096

Langmuir, Vol. 20, No. 21, 2004

Li et al.

Figure 1. XRD patterns of TiO2 films prepared by slow and rapid removal of solvent: (A) slow removal of solvent (evaporation, ca. 10 min); (B) rapid removal of solvent during spin-coating.

Experimental Section The synthesis of these mesoporous anatase TiO2 films is straightforward. All reagents are aqueous-based, inexpensive, environmentally friendly, and readily available on a large scale. The surfactant cetyl trimethylammonium chloride (CTAC) (from TCI America) was dissolved in deionized (DI) water and 200 proof ethyl alcohol (Gold Shield Chemical Co.); then (NH4)2Ti(OH)2(C3H5O3)2 (Tyzor LA- from Dupont, 2.23 M in Ti) was slowly added to the surfactant solution. The molar ratio of Ti/ CTAC/water/ethanol is 1:0.1-0.7:15-30:20. The solution thus prepared was clear. The mixture was shaken for 4-12 h and spin-coated on 1 in.2 glass at 1000 rpm. The spin-coated films were then dried at 150 °C for 2-5 min on a hot plate. Finally, films were calcined on a hot plate at 375 °C for 0.5-5 min in air to remove surfactants. Attempts to remove the surfactants were also performed using acid extraction,25 UV/ozone,26 and air plasma, all of which were unsuccessful. To study the effects of surfactant concentration on the pore structures and properties of films, the Ti/CTAC molar ratio was varied from 0.1 to 0.7. In addition, concentrated HNO3 (25) Mercier L.; Pinnavaia, T. J. Environ. Sci. Technol. 1998, 32, 2749-2754. (26) Clark, T., Jr.; Ruiz, J. D.; Fan, H.; Brinker, C. J.; Swanson, B. I.; Parikh, A. N. Chem. Mater. 2000, 12, 3879-3884.

was used to adjust the solution pH from 8.2 to 1 in order to study the effect of solution pH (without the addition of acid, the pH of the precursor solution is 8.2). The pore structure and pore size of films before and after calcination were studied using X-ray diffraction (XRD) and transmission electron microscopy (TEM). The crystalline phase of TiO2 was also identified by XRD. The X-ray diffraction data of films were obtained using a Philips X’PENT-MPD diffractometer with Cu Ka radiation. Highresolution TEM analysis was carried out on a JEOL JEM 2010F microscope with a specified point-to-point resolution of 0.194 nm. TEM samples were prepared by scraping the film off of the substrate. Brunauer-Emmett-Teller (BET) surface area and pore size analyses were determined using nitrogen adsorption/ desorption collected with a Quantachrome Autosorb 6-B gas sorption system on degassed samples. The samples for BET analyses were obtained by evaporating the solvent in a hood at room temperature followed by heating the dried samples in air at 5 °C/min to 375 °C and then holding them at 375 °C for 2 h. Carbon-13 NMR spectra were collected at ambient temperature on a Varian VXR-300S spectrometer operating at 75.429 MHz. Titrimetric analyses were carried out using a ORION pH meter (model 720A) with a Corning G-P Combo W/RJ electrode. The Tyzor solution concentration was adjusted to 0.480 M and was titrated against a 0.315 M HNO3 solution.

Properties of Mesoporous TiO2 Thin Films

Langmuir, Vol. 20, No. 21, 2004 9097

Results and Discussion Rate of Solvent Removal. The solvent evaporation rate has been observed to have an impact on the morphology of mesoporous silica thin films. Gibaud and co-workers27 have observed that mesostructured dipcoated thin films undergo an order-disorder transition that is related to the rate of solvent removal. This transition is thought to involve intermediate hybrid mesophases that are related to the concentration gradients within the reaction me´lange. The concentration of nonvolatile components enhances micelle formation and selfassembly of the template and of the inorganic components. For example, rapid evaporation (less than 20 s) of a silicic acid/cetyltrimethylammonium bromide (CTAB) sol resulted in a 3D hexagonal mesostructure in the dip-coated thin film, while slow evaporation (ca. 20 min) of the same reaction mixture led to a disordered wormlike structure.27 The term “evaporation controlled self-assembly” has been coined to describe this phenomenon.27 The chemical dynamics of this process has been monitored in situ using small-angle X-ray scattering and interferometry measurements.28,29 In our studies of spin-cast titania thin films, the rate at which the solvent was removed was also found to have a notable impact on film morphology. Figure 1 shows the XRD patterns of TiO2 films prepared by slow and rapid removal of solvent during the spin-casting of the film. Slow removal of solvent was accomplished by dropping the deposition solution onto a glass slide and placing the sample in a hood until the solvent was evaporated (about 10 min). Rapid removal of solvent sample was prepared by spin-coating solution on a glass slide at 1000 rpm. The films then were dried at 150 °C for 5 min on a hot plate before XRD measurement. The small-angle XRD pattern for the spin-coated film shows four well-resolved peaks at 2θ of 2.75°, 5.5°, 8.25°, and 11°, that are indexable as (100), (200), (300), and (400) of a lamellar phase, with a 32.1 Å d spacing. An additional peak in the 2θ range of 2°-2.5° corresponds to the wormlike structure with a d spacing of about 35-40 Å. For the film prepared by slow removal of solvent, this wormlike structure peak became much more intense and the peaks corresponding to the lamellar phase became much weaker. TEM images of the samples scraped from these films are shown in Figure 2. These TEM results confirm that a lamellar phase is formed from the spincoated Tyzor precursor and wormlike structure is a major phase from a similar sample prepared by slow removal of solvent. Rapid removal of solvent had a significant impact on film structure. Bruinsma et al.17 reported that mesoporous silica film prepared by spin-coating using CTAC as a template has a strong primary reflection in the XRD pattern, a qualitative indicator of structural order. When the surfactant was mixed in the aqueous solution, the surfactant molecules were homogeneously dispersed. During the spin-coating process, the solvent evaporated quickly, causing the surfactant to form aggregated micelles (“evaporation induced self-assembly” 30-32), and the mi(27) Gibaud, A.; Grosso, D.; Smarsly, B.; Baptiste, A.; Bardeau, J. F.; Babonneau, F.; Doshi, D. A.; Chen, Z.; Brinker, C. J.; Sanchez, C. J. Phys. Chem. B 2003, 107, 6114-6118. (28) Grosso, D.; Balkenende, A. R.; Albouy, P. A.; Lavergne, M.; Babonneau, F. J. Mater. Chem. 2000, 10, 2085-2089. (29) Grosso, D.; Babonneau, F.; Albouy, P. A.; Amenitsch, H.; Balkenende, A. R.; Brunet-Bruneau, A.; Rivory, J. Chem. Mater. 2002, 14, 931-939. (30) Brinker, C. J.; Lu, Y.; Sellinger, A.; Fan, H. Adv. Mater. 1999, 11, 579-585.

Figure 2. TEM images of TiO2 films prepared by slow and rapid removal of solvent: (A) slow removal of solvent (evaporation, ca. 10 min); (B) rapid removal of solvent during spincoating.

celles directed the formation of ordered arrays in the film. Subsequent aging and heating served to condense the preceramic species into a continuous network around the micelles, and an ordered ceramic phase can ultimately be formed. Rapid removal of solvent results in a kinetic structure that is dictated by the self-assembly of the surfactant. The silica sample obtained through slow evaporation of the solvent only had a broad peak of very low intensity in the XRD, indicative of a low degree of order (in accord with previous observations for dip-coated silica thin films27). In this case, the lamellar morphology of the micelles is slowly overwhelmed by the slower condensation and growth of the titania phase, ultimately resulting in wormlike pores and disordered wall structure. In this case, the origin of the structure appears to be influenced strongly (31) Lu, Y.; Fan, H.; Stump, A.; Ward, T. L.; Rieker, T.; Brinker, C. J. Nature 1999, 398, 223-226. (32) Bore, M. T.; Rathod, S. B.; Ward, T. L.; Datye, A. K. Langmuir 2003, 19, 256-264.

9098

Langmuir, Vol. 20, No. 21, 2004

Li et al.

Figure 3. XRD as a function of CTAC/Ti ratio.

by the condensation of the preceramic to form anatase and secondarily by the surfactant. The fact that this reorganization is only observed with the slow removal of solvent suggests what a careful balancing act is involved in terms of dictating structure. Effect of Surfactant Concentration. Figure 3 shows the XRD patterns of TiO2 films prepared with different surfactant concentrations. Samples were prepared by spincoating at 1000 rpm followed by drying at 150 °C for 5 min on a hot plate before XRD measurement. With the CTAC/Ti molar ratio equal to or higher than 0.49, the small-angle XRD pattern for spin-coated film shows four well-resolved peaks at 2θ of 2.75°, 5.5°, 8.25°, and 11°, that are indexable as (100), (200), (300), and (400) of a lamellar phase with a d spacing of 32.1 Å. An additional peak in the 2θ range of 2°-2.5° corresponds to the wormlike structure. When the CTAC/Ti molar ratio was less than 0.29, a disordered structure was formed. Based on the typical surfactant-oil-water phase diagram,33 spherical and cylindrical micelles are preferentially formed at low surfactant concentrations. At increasing surfactant concentrations, hexagonal and cubic micelle phases are formed preferentially, and lamellar phases are formed at high surfactant concentrations. For the SiO2-CTAB system in an acidic medium, the hexagonal phase was observed with CTAB/Si ) 0.12 (CTAB ) 0.15 wt %) and cubic phase was formed with CTAB/Si ) 0.13 (CTAB )

0.16 wt %).34 Ogawa produced lamellar phase SiO2 by using a CTAB/Si mole ratio of 0.4.35 The BET surface area of TiO2 powder prepared with different surfactant/Ti molar ratios is summarized in Figure 4. As can be seen from the data, as the surfactant concentration increases, BET surface area first increases and appears to reach a maximum at a CTAC/Ti ratio of approximately 0.5. Above this value, it levels off and drops slightly, possibly due to phase segregation of the surfactant and TiO2 at higher surfactant concentrations.

(33) Baskaran, S.; Liu, J.; Li, X.; Fryxell, G. E.; Kohler, N.; Coyle, C.; Birnbaum, J. C.; Dunham, G. Sol-Gel Commercialization and Applications; Ceramic Transactions, Vol. 123; American Ceramic Society: Westerville, OH, 2001; pp 39-47.

(34) Huo, Q.; Margolese, D. I.; Ciesla, U.; Feng, P.; Gier, T. E.; Siwger, P.; Leon, R.; Petroff, P. M.; Schuthand, F.; Stucky, G. D. Nature 1994, 368, 317-320. (35) Ogawa, M. J. Am. Chem. Soc. 1994, 116, 7941-7942.

Figure 4. Surface area as a function of CTAC/Ti ratio.

Properties of Mesoporous TiO2 Thin Films

Langmuir, Vol. 20, No. 21, 2004 9099

Figure 5. XRD as a function of pH.

Effect of Solution pH. The XRD patterns of TiO2 films prepared with a surfactant/Ti molar ratio of 0.49 while the solution pH was adjusted to 0.9, 3.9, 5.4, 6.1, 6.8, and 8.2 are shown in Figure 5. These samples were prepared by spin-coating at 1000 rpm, followed by drying at 150 °C for 5 min on a hot plate before XRD data were collected. Within the pH range studied, the small-angle XRD pattern for the spin-coated film shows four well-resolved peaks at 2θ of 2.75°, 5.5°, 8.25°, and 11°, that are attributed to the (100), (200), (300), and (400) lines of a lamellar phase. An additional peak in the 2θ range of 2°-2.5° corresponds to the wormlike structure. As the pH decreases, the d spacing of the lamellar phase decreases nonlinearly. Figure 6 shows the plot of the d spacing of the (100) lamellar peak as a function of solution pH, which was found to undergo a step increase above a pH of about 6. The titania-lactate solutions were studied by carbon13 NMR at pH 6.8, 6.2, and 5.4 (the solutions were identical to the spin-coating solutions, except that surfactant was left out). The spectra (not shown) revealed a 1.16 ((0.05) to 1 ratio of free lactate ion to titania-lactate complex for all three solutions. This ratio indicates the average titania-lactate species present has 0.92 mol of lactate complexed for each mole of titanium. The fact that this ratio does not change over this pH range might be expected

Figure 6. The d spacing of the as-spun lamellar phase as a function of pH.

considering the pKa of lactic acid is 3.08;25 thus even at pH 5.4 the salt-to-acid ratio is over 200 to 1. A second NMR experiment was conducted on the actual sols used, to investigate if adding surfactant and alcohol would have any effect on the salt-to-complex ratio at the same three pHs. No significant difference in speciation was detected. Finally a spin-coat solution was adjusted to a pH of 1.2, well below the pKa of lactic acid, and analyzed by carbon-13 NMR. At this pH, the lactate salt to lactic acid ratio is 1.3 × 10-2 (98.7% lactic acid). Not surprisingly,

9100

Langmuir, Vol. 20, No. 21, 2004

Li et al.

Figure 7. Titanium lactate titration data.

Figure 8. Low-angle XRD patterns of TiO2 films after calcination showing collapse of the lamellar structure after calcination at 375 °C.

the spectrum revealed that virtually no titania-lactate complex exists at this pH. It is clear from the NMR data that the effect pH has on the d spacing of the resultant films is not due to a difference in lactate ion/titania-lactate complex ratio, in the pH range of 5.4-6.8. The amount of lactic acid present in solution in this pH range is very small and has no significant effect on complex formation. To further investigate the change in d spacing in the pH range 5.4-6.8, the titania-lactate species was titrated with dilute nitric acid solution. An inflection point was observed (see Figure 7), and the pKb of the titania-lactate complex was determined to be 7.2 ( 0.2. A second calculated curve is shown for comparative purposes of a lactate ion solution titrated with nitric acid at the same concentrations. Not only is it apparent that the titanialactate species present in solution exhibits acid/base character, but the data also suggest that well-defined speciation is present (i.e., oligomeric species with a very small range of molecular weights or perhaps a specific molecular weight) since a reasonably well-defined end point was observed in the titration. Previous work has suggested that this titanium lactate species is oligomeric in this pH range,37 and therefore the formula could be represented as [Ti(Lac)0.92(µ-O)x((OH)y]z, where titanium

Figure 9. TEM images of the mesoporous TiO2 films after calcination.

ions complexed with lactate ions are linked together with bridging oxo ligands and the remaining open sites on the titanium ions are occupied by hydroxyl groups. The inflection point observed for the titanium lactate species is very near the pH range being investigated, and clearly a significant difference in unprotonated to protonated titania-lactate species ratio will be present. For example, at pH 6.8 the ratio is calculated to be 2.5:1 (protonated to unprotonated), while at pH 6.2 this ratio is 10:1, and at pH 5.4 it is 63:1. Presumably the site of protonation is not a lactate ligand and is more likely to be the more basic bridging oxo ligand. Protonation of an oxo bridging ligand would form a hydroxy (36) Handbook of Chemistry and Physics, 72nd ed.; Lide, D. R., Ed.; CRC Press: Boca Raton, FL, 1991-1992; pp 8-40. (37) Graff, G.; Bunker, B. Unpublished work.

Properties of Mesoporous TiO2 Thin Films

Langmuir, Vol. 20, No. 21, 2004 9101

Figure 10. XRD patterns of TiO2 after calcination (with anatase lines shown for reference).

species and break the oxo linkage. The smaller d spacing (25.8 Å) suggests that upon protonation either a smaller titanium lactate species is formed by fragmentation of the original oligomeric species or a thinner oligomeric species is formed due to a conformational change, such as a globular structure transforming to a more linear elongated structure upon the cleavage of an oxo bridge. Without further analytical data, it is difficult to more rigorously establish the identity of the exact species present in this pH range. However, hydrated CTAC forms a lamellar structure with a d spacing of 25.8 Å in the absence of any ceramic or preceramic components.38 This would indicate that the thickness of the titanium lactate species present at pH 5.4 is very similar to that of a hydrated chloride ion (∼3 Å)39 in this CTAC lamellar phase. Furthermore, this observation indicates that the thickness of the titanium lactate oligomer present from pH 6.8 (d spacing, 32.1 Å) deposition is approximately 9-10 Å across. Clearly, at these pHs the titanium preceramic precursor is larger and/or perhaps more globular in conformation. The ordered lamellar phase structurally collapsed and the TiO2 phases became disordered during the calcination at 375 °C, as revealed by the XRD spectrum (see Figure 8). This structural collapse is also apparent from the TEM image of the calcined sample, showing the mesoporous ceramic structure to be wormlike (see Figure 9). Even through this is a disordered structure, the calcined film is still porous, with a porosity of about 45% and a BET surface area of 220 m2/g. XRD analysis of the calcined thin film revealed that the TiO2 phase was anatase (see Figure 10). The scanning electron microscopy (SEM) photomicrographs of the crosssectioned TiO2 thin film indicated that the film is (38) Shin, Y.; Wang, L.-Q.; Fryxell, G. E.; Exarhos, G. J. Hygroscopic growth of self-assembled layered surfactant molecules at the interface between air and organic salts. J. Colloid Interface Sci., submitted. (39) Harris, D. C. Quantitative Chemical Analysis, 3rd ed.; W. H. Freeman: New York, 1991; p 107.

Figure 11. SEM images of a cross section of the TiO2 thin film (the light layer is the substrate, and the darker layer is the 250 nm mesoporous anatase film).

continuous and crack-free but is decorated with modest surface striations (see Figure 11). The film thickness is about 0.25 µm (the thickness can be controlled by changing the precursor solution viscosity and spin-coating rate). Taken together, these results suggest that the titanium oligomeric precursor exists as in a globular conformation at pHs above about 6.5, leading to a larger d spacing. Below a pH of 6.5, the titanium oligomeric precursor appears to exist in an open, linear conformation, leading to a more compact d spacing. The NMR evidence does not reveal any major speciation change in these titanium lactate solutions. Calcination at 375 °C leads to loss of order and partial collapse of the mesostructure, but the film maintains a significant level of porosity and high surface area.

9102

Langmuir, Vol. 20, No. 21, 2004

Conclusions A novel method for preparing surfactant-templated mesoporous anatase TiO2 thin films using very mild reaction conditions and an aqueous-based, inexpensive precursor in conjunction with cationic surfactants has been demonstrated. These high surface area ceramic thin films were prepared by spin-coating the precursor solution onto glass substrates, followed by calcination to remove the surfactant. The as-spun preceramic films were found to have a lamellar structure, whose d spacing was dependent on the pH of the deposition solution. Experimental observations suggest that a conformational change of the preceramic oligomer from globlular to linear is responsible for this pH-dependent shift. After calcination, the films were found to have wormlike pore structures. The mesostructure and porosity of these anatase thin films can be

Li et al.

controlled by the surfactant template chemistry, which is dictated by parameters such as the surfactant/Ti ratio, the pH of the deposition solution, and the rate of solvent evaporation. Further work developing templated mesostructured titanium-based thin films is currently underway and will be reported in due course. Acknowledgment. Pacific Northwest National Laboratory is operated for the U.S. Department of Energy (DOE) under Contract DE-AC0676RLO1830 by Battelle Memorial Institute. This work was supported by the Laboratory Directed Research and Development Program. Helpful discussions from Gordon Graff and Yongsoon Shin are gratefully acknowledged. LA0486279