J. Phys. Chem. C 2008, 112, 20007–20011
20007
Template-Free Formation of Meso-Structured Anatase TiO2 with Spherical Morphology† Poovathinthodiyil Raveendran,*,‡ Muthusamy Eswaramoorthy,*,§ Unnikrishnan Bindu,‡ Maya Chatterjee,| Yukiya Hakuta,| Hajime Kawanami,| and Fujio Mizukami| Department of Chemistry, UniVersity of Calicut, Kerala-673 635, India, Chemistry and Physics of Materials Unit and New Chemistry Unit, Jawaharlal Nehru Center for AdVanced Scientific Research, Bangalore-560 064, India, and Research Center for Compact Chemical Processes, and National Institute for AdVanced Industrial Science and Technology, 4-2-1, Nigatake, Miyagino-ku Sendai, 983-8551 Japan ReceiVed: July 7, 2008; ReVised Manuscript ReceiVed: October 18, 2008
Template-free synthetic approaches offer significant environmental and economical advantages in the synthesis of industrially important, high surface area, meso-structured metal oxide materials such as titania. Herein, we show that the simple, neutral, room-temperature, hydrolytic condensation of Ti(IV) n-butoxide dispersed in ethyl acetate followed by mild heating yields meso-structured, submicron-sized anatase spheres with high surface area and thermal stability (400 °C). 1. Introduction Titania, being an n-type wide band semiconductor, has several industrial applications,1-3 and the dye-sensitized, mesoporous anatase is currently being proposed as a potential replacement for silicon as the photoactive material in solar cells, offering tremendous opportunities for harvesting solar energy at low cost.4,5 High surface area and crystallinity (anatase) are two factors that are critical in enabling such technologies, and thus, inexpensive and environmentally benign methods6 for the synthesis of meso-structured anatase are important. Ever since the first synthesis of mesoporous titania was reported by Antonelli and Ying,7 several researchers have developed diverse techniques to improve its surface area and crystallinity.8-12 The majority of these approaches utilize templates. Although a few attempts13,14 toward the development of template-free routes have been reported, these methods did not provide mesostructured anatase. Self-assembly, often governed by site-specific intermolecular interactions, has been the focus of immense research in material science in the recent years for the controlled organization of nanoscale materials and the synthesis of meso-structured materials.15,16 For example, the self-assembly of several organic surfactants and block copolymers has been successfully used as templates and structure-directing agents for the synthesis of meso-structured metal oxides.1,7,8,17-21 The use of such auxiliary templates or structure-directing agents, some of them ionic liquids,22 necessitates techniques for their removal, such as calcination or solvent extraction under extreme conditions, raising environmental and economical concerns. Herein we investigate the template-free, room-temperature, hydrolytic condensation of inorganic-organic hybrid systems such as metal alkoxides in suitable solvents such as ethyl acetate as a “green” approach for the synthesis of meso-structured metal oxides. * Corresponding authors. P.R.: tel, (+91) 9446537724; fax, (+91) 494 2400269; e-mail,
[email protected]. M.E.: tel, (+91) 80 22082870; fax, (+91) 80 22082766; e-mail,
[email protected]. † This paper is dedicated to the fond memory of the late Prof. Yutaka Ikushima. ‡ University of Calicut. § Jawaharlal Nehru Center for Advanced Scientific Research. | Research Center for Compact Chemical Processes, and National Institute for Advanced Industrial Science and Technology.
The approach presented herein is based on the microphase separation and organization of Ti(IV) n-butoxide, TiOB (monomers or polyalkoxides, as many alkoxides exist at room temperature), when dispersed in a medium like ethyl acetate, as a result of their different intermolecular association behavior resulting from site-specific polar and nonpolar interactions. Such microphase separation (or microheterogenity) is a wellestablished phenomenon even in simple solvent mixtures.23 2. Experimental Section TiOB (97%, Adrich) and ethyl acetate (EA, WAKO Chemicals) were used as purchased. In a typical procedure, 2 mL of TiOB was added to 6 mL of dehydrated EA with continuous stirring at 25 °C (rt). After stirring the mixture rigorously for about 5 min, 3 mL of deionized water was added with stirring to hydrolyze TiOB. A white slurry of TiO2 was formed almost instantaneously and it was kept overnight at rt for drying. These samples are referred to as the as-synthesized or room-temperature samples. For hydrothermal treatment, the samples were heated at desired temperatures in ovens having temperature regulators, with sufficient amounts of water, in closed stainless steel autoclaves. Blank samples were also prepared in the absence of ethyl acetate, by simply adding about 3 mL of deionized water to about 2 mL of TiOB under constant stirring. The white slurry formed was allowed to dry at room temperature. High-resolution transmission electron microscope (HR TEM) images of the samples were obtained using TECNAI G2 type field emission TEM on standard Cu grids covered by amorphous carbon films containing the sample. X-ray diffraction studies were carried out using Rigaku 2000 diffractometer having low angle accessory, using Cu KR radiation. Brunauer-Emmett-Teller (BET) adsorption-desorption experiments for the surface area measurements were carried out using a BELSORP 28SA apparatus. 3. Results and Discussion HR TEM images of the as-synthesized samples reveal that the sample consists of TiO2 spheres in the nano and submicron size range, typically between 50 and 700 nm (Figure 1). In some cases, these spheres appear to be associated (Figure 1A). Closer
10.1021/jp805963s CCC: $40.75 2008 American Chemical Society Published on Web 11/21/2008
20008 J. Phys. Chem. C, Vol. 112, No. 50, 2008
Figure 1. TEM images of the rt-synthesized (A, B, C, and D) and heat treated (E and F, 60 °C, 1 week) mesostructured titania nanospheres at various magnifications. Images G and H correspond to the bulk TiO2 (prepared in the absence of EA) in the rt-synthesized form and after heat treatment (60 °C, 1 week), respectively.
observations using high-resolution TEM (HR TEM) reveal that these nanospheres are micro/submesoporous (Figure 1B-D). These pores correspond to an interparticle, worm-hole type architecture with the pore sizes ranging from 1.2 to 1.6 nm. Small angle XRD (Figure 2A, blue curve) for this sample clearly shows the presence of mesoscale order in this system with a peak around 2θ ) 2° (d-spacing ) 4.8 nm). The broad nature of this peak and the absence of higher order reflections indicate that the mesoscale order is short ranged. The absence of any typical crystalline peaks in the higher angle XRD reveals that the sample is amorphous. To study the effects of heat treatment, about 2 mL of water was added to the sample, which was stirred well and left to dry at 60 °C. After about 48 h, the small angle XRD (Figure 2A, dark red) of the sample showed some degree of sharpening and a slight shift to higher d-values (5.3 nm), indicating an increase in the mesoscale order. In fact, since no templates are employed, we anticipated that
Raveendran et al. the meso-scale order at rt may collapse on heating. The observed increase in ordering may be due to a coupling between the drying-induced self-assembly24 and the Ostwald ripening within the localized, nanoparticulate TiO2 spheres. Corresponding higher angle XRD (Figure 2B, dark red) showed that the sample is in the anatase phase. Continued heating of the sample at 60 °C for 7 days resulted in a significant increase in intensity with increase in d-spacing and sharpening of the low angle XRD peak (Figure 2A, black curve), showing a further increase in the meso-scale order. This may be explained as due to the sharpening of an initially broader particle size distribution at higher temperature, with increase in particle sizes. The particle size from the Scherrer formula and the observed d-spacing agree rather well, supporting this notion. Also, the small angle peak shifted to relatively larger d-values (11.3 nm), indicating a notable reorganization within the spheres as a result of the crystallization. This observation is in contrast to the previous reports21 of a decrease in the d-value on heat treatment as a result of wall condensations, suggesting significant differences from the previously reported cases. In the present case, the pores are of interparticle nature and are lesser defined as compared to the structured pores resulting from the templatebased strategies. This also provides an explanation for an observed d-value of 11.9 nm for the calcined material with an associated particle size of around 10 nm. However, there is no notable change in the full width at half-maximum (fwhm) of the higher angle anatase peaks (Figure 2B, black curve, particle size from Scherrer formula is ∼4 nm) for the sample heated for 1 week in comparison with that of the 48 h sample. The microstructures of the TEM images of the heat-treated samples are quite distinct from the as-synthesized samples. The spherical TiO2 assemblies in the heat-treated samples (Figure 1E,F) appear to be constituted by a stone-wall type packing (where the pores are the interparticle voids) of anatase nanocrystals. Just like the nanosphere association observed, there appears to be connectivity between the individual anatase nanocrystals. This and the random misalignment of the crystallographic axes along which these crystals are grown may provide high thermal stability for these mesoscale assemblies, by preventing particle aggregation. Since one key aspect for the utilization of these materials is their thermal stability, we followed the heating effects on these samples up to 400 °C. Small angle XRD (Figure 2A, red curve) shows that while the heat treatment at 400 °C increases the mesoscale order (as evidenced from the significant sharpening of the peak), there is no pronounced shift in the d-spacing (11.9 nm). Also, the higher angle XRD (Figure 2B, red curve) peaks reveal an increase in the crystallite sizes (∼10 nm, using the Scherrer formula). The TEM images also reveal that the spherical assemblies of the TiO2 nanoparticles, of size around 5-10 nm, are intact even after heating at 400 °C. We also studied the effects of hydrothermal treatment at different temperatures on these systems. Typically, about 4 mL of deionized water was added to about 1 g each of the assynthesized samples taken in three 20-mL, Teflon-lined autoclaves, and the sealed autoclaves were heated at 60, 80 and 100 °C, respectively, for about 24 h. Small angle XRD patterns of these samples suggest a dependence of the d-spacing with the temperature. The absence of any structural collapse on hydrothermal treatment also suggests that there is an inherent degree of order in the directed self-assembly of the TiO2 nanoparticles. High resolution TEM images and the electron diffraction patterns of these samples are shown in parts A and B of Figure
Template-Free Formation of Titania Spheres
J. Phys. Chem. C, Vol. 112, No. 50, 2008 20009
Figure 2. Small angle (A) and higher angle (B) XRD of the samples: (blue) as synthesized, (dark red) after heating at 60 °C for 48 h, (black) after heating at 60 °C for 1 week, and (red) after heat treatment at 400 °C for 5 h.
Figure 3. Crystallinity of the hydrothermally treated nanospheres. (A) HRTEM image of the hydrothermally treated (100 °C) samples showing the lattice fringes corresponding to the anatase phase. (B) The corresponding electron diffraction pattern. (C) Small angle XRD of the samples corresponding to different treatment temperatures: (black) rt; (red) 60 °C; (dark red) 80 °C; (blue) 100 °C.
3, respectively, supporting a stone-wall-type packing of highly crystalline anatase nanoparticles. For the removal of the organic residues (the n-butanol formed as the hydrolysis product and the residual EA) and water from the sample, we employed extrusion using supercritical CO2, which is widely regarded as an environmentally benign and efficient medium due to its nontoxicity, nonflammability, abundance, low cost, and high diffusivity.25 Also, supercritical fluid (SCF) extrusion is an inherently “dry” process and there is no need for any postextrusion drying. The BET surface area for the rt sample after CO2 extrusion is 554 m2/g. The N2 adsorption-desorption isotherm (Figure 4A) falls on the borderline between type I and type IV with no hysterisis. The Dollimore-Heal (DH) pore size distribution shows that the pores are mostly below 1.7 nm and the Horvath-Kawazoe (HK) analysis for the micropore region shows a discrete pore size distribution in the range from 0.5 to 1.4 nm (Figure 1A, inset). These are in agreement with the results obtained from the HRTEM images. The SCF extrusion did not affect the spherical morphology or the d-spacings as evidenced from the TEM and small-angle XRD profiles. It is observed that even conventional heating under vacuum at 100 °C can be used for
the removal of these residues, which could be of importance in the commercialization of this approach. The anatase sample after heat treatment for 48 h shows a type IV N2 adsorption-desorption isotherm with a surface area of 388 m2/g and a DH pore size of 2 nm, while continued heat treatment for 1 week causes a decrease in the surface area (260 m2/g) and an increase in the DH pore size (4 nm). The samples subjected to heat treatment at 400 °C also present a surface area of 230 m2/g but with an increased DH pore size (8.4 nm). A HR TEM image of the calcined sample is provided in Figure S1 (Supporting Information). These observations may be explained on the basis of the continued crystal growth process and structural reorganizations within the nanospheres. The hydrothermally treated samples also show surface areas around 250 m2/g and DH pore sizes in the range of 5-6 nm. The mechanistic details regarding the formation of the solution phase nanostructures of TiOB in EA, their size control, and the thermodynamic control are of importance. Compositions with higher amounts of TiOB (above 50% TiOB by volume, in the TiOB-EA mixture) resulted in porous, but nonstructured titania, while lower concentrations (below 10% TiOB by volume, in the TiOB-EA mixture) show small aggregates with
20010 J. Phys. Chem. C, Vol. 112, No. 50, 2008
Figure 4. N2 adsorption-desorption isotherms for the rt- synthesized (A) and the heat-treated (60 °C, 48 h, B) samples. The pore size distributions using the Horvath-Kawazoe and Dollimore-Heal methods are shown in the insets.
lesser morphological preferences. In order to examine the role of EA in the self-assembly process, we also carried out blank experiments and observed that the mesoscale order exists even in the absence of EA, although the spherical morphology is absent. The TEM images for the bulk (synthesized in the absence of EA) sample in the as-synthesized form and after heat treatment (60 °C, 1 week) are given in Figure 1, parts G and H, respectively. The BET surface area and pore size corre-
Raveendran et al. sponding to Figure 1H is 246 m2/g and ∼4 nm, respectively. These data are in general consistent with those observed for the samples with spherical morphology. This indicates that the observed meso-scale order arises from the self-assembly of TiOB molecules themselves and the role of EA is only in dispersing the bulk associations into thermodynamically stable spherical assemblies. A tentative mechanism is presented in Figure 5. It is plausible that the charge separations in the TiOB molecule (with partial positive and negative charges on the Ti and the O atoms, respectively, and the essentially nonpolar alkyl chains) may guide the self-assembly of these systems. Although there is no long-range meso-scale order in the systems (as evidenced from the absence of higher order reflections in the small angle XRD), these results are significant, not only from an industrial standpoint but also from the understanding of nanoscale organizations. We observed similar, though less ordered, associations in the case of niobium oxide formed from niobium ethoxide, suggesting the possibilities of extending this method to other oxide systems. In fact, previous works have already used niobium ethoxide as a precursor for the synthesis of niobium oxide, in the template-assisted strategy.26 The spherical morphology of the aggregates formed is also of importance. This may be explained on the basis of the energetics of cluster speciation between the highly charge polarized TiOB and the less polarized EA. In the dispersion concentrations studied, there will be a competition between selfinteractions vs cross-interactions among EA and TiOB systems (as well as their partially hydrolyzed derivatives), and it is natural that the more interacting (in terms of electrostatics) TiOB molecules will maximize the self-interaction (and minimizes the surface area), thereby causing the observed spherical morphology. The situation is best described by microheterogenity. While such effects will be less prominent on the morphology of very small clusters, it is bound to be significant in the case of larger clusters. The solvent molecules, thus, may play a significant role in guiding the self-assembly of TiO2 into spherical clusters, although in a passive way. The possibility of the direct participation of EA in the formation of the solution phase nanostructures also cannot be ruled out completely. 4. Conclusions and Outlook In summary, we have demonstrated a template-free method for the preparation of mesoporous, anatase titania having
Figure 5. Mechanism: The hierarchical organizations of TiOB on hydrolysis forms mesostructured TiO2 with amorphous walls, which on heating transforms into anatase phase. EA helps in the dispersion of such clusters into spherical assemblies.
Template-Free Formation of Titania Spheres spherical morphology, in the nanometer/submicron size range, by simple room-temperature, hydrolytic condensation of TiOB dispersed in ethyl acetate. The mesoporous anatase titania thus synthesized exhibits high surface area (388 m2/g), suggesting potential applications for the method in the synthesis of nanocrystalline, mesoporous anatase TiO2. While the formation of the mesoporous organization may be attributed to the selfassembly of TiOB molecules through site-specific intermolecular interactions, the role of EA may be in dispersing the bulk TiOB associations into thermodynamically stable spherical assemblies, retaining their meso-scale order. Acknowledgment. The authors thank JSPS and the CREST program of JST for partial financial support. Supporting Information Available: HR TEM image of the calcined sample (Figure S1). This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) Sojer-illia, G. J. D. A.; Sanchez, C.; Lebeau, B.; Patarin, J. Chem. ReV. 2002, 102, 4093. (2) Resmi, M. R.; Whitesell, J. K.; Fox, M. A. Res. Chem. Intermed. 2002, 28, 711. (3) Varghese, O. K.; Gong, D.; Paulose, M.; Ong, K. G.; Dickey, E. C.; Grimes, C. A. AdV. Mater. 2003, 15, 624. (4) Bach, U.; et al. Nature 1998, 395, 583. (5) Hagfeldt, A.; Gra¨tzel, M. Acc. Chem. Res. 2000, 33, 269.
J. Phys. Chem. C, Vol. 112, No. 50, 2008 20011 (6) Anastas, P. T.; Warner, J. C. Green Chemistry: Theory and Practice; Oxford University Press Inc.: New York, 1998. (7) Antonelli, D. M.; Ying, J. Y. Angew. Chem. Intl. Ed. 1995, 34, 2014. (8) Yang, P.; Zhao, D.; Margolese, D. I.; Chmelka, B. F.; Stucky, G. D. Nature 1998, 396, 152. (9) Yoshitake, H.; Sugihara, T.; Tatsumi, T. Chem. Mater. 2002, 14, 1023. (10) Luo, H.; Wang, C.; Yan, Y. Chem. Mater. 2003, 15, 3841. (11) Haseloh, S.; Choi, S. Y.; Mamak, M.; Coombs, N.; Petrov, S.; Chopra, N.; Ozin, G. A. Chem. Commun. 2004, n/a, 1460. (12) Crepaldi, E. L.; Sojer-illia, G. J. D. A.; Grosso, D.; Cagnol, F.; Ribot, F.; Sanchez, C. J. Am. Chem. Soc. 2003, 125, 9770. (13) Collins, A.; Carriazo, D.; Davis, S. A.; Mann, S. Chem. Commun. 2004, n/a, 568. (14) Liu, C.; Fu, L.; Economy, J. J. Mater. Chem. 2004, 14, 1187. (15) Whitesides, G. M.; Grzybowski, B. Science 2002, 295, 2418. (16) Li, M.; Schnablegger, H.; Mann, S. Nature 1999, 402, 393. (17) Kresge, C. T.; Leonowicz, M. E.; Roth, W. J.; Vartuli, J. C.; Beck, J. S. Nature 1992, 359, 710. (18) Huo, Q. S. Nature 1994, 368, 317. (19) Sun, T.; Ying, J. Y. Angew. Chem., Int. Ed. 1998, 37, 664. (20) Che, S.; Liu, Z.; Ohsuna, T.; Sakamoto, K.; Terasaki, O.; Tatsumi, T. Nature 2004, 429, 281. (21) Wong, M. S.; Jeng, E. S.; Ying, J. Y. Nano Lett. 2001, 1, 637. (22) Zhou, Y.; Antonietti, M. J. Am. Chem. Soc. 2003, 125, 14960. (23) Dawson, E. D.; Wallen, S. L. J. Am. Chem. Soc. 2002, 124, 14210. (24) Rabani, E.; Reichman, D. R.; Geissler, P. L.; Brus, L. E. Nature 2003, 426, 271. (25) Eckert, C. A.; Knutson, B. L.; Debenedetti, P. G. Nature 1996, 373, 313. (26) Antonelli, D. M.; Ying, J. Y. Angew. Chem., Int. Ed. 2003, 35, 426.
JP805963S
