Journal of Colloid and Interface Science 231, 289–298 (2000) doi:10.1006/jcis.2000.

7147, available online at http://www.idealibrary.com on

The Effect of Alcohol Solvents on the Porosity and Phase Composition of Titania
Ki Chang Song1 and Sotiris E. Pratsinis2
Institute of Process Engineering, Swiss Federal Institute of Technology, ETH-Zentrum, CH-8092 Zurich, Switzerland Received March 22, 2000; accepted August 4, 2000

Bimodally porous titania powders were made by hydrolysis of titanium tetraisopropoxide (TTIP) dissolved in various alcohols (methanol, ethanol, isopropanol, and sec-butanol). The specific surface area (SSA) of the powders dried at 150◦ C ranged from 332 to 624 m2 /g as determined by nitrogen adsorption. At excess alcohol concentration, the SSA of the dried powders decreased in the order of sec-butanol, iso-propanol, ethanol, and methanol at a constant alcohol/TTIP molar ratio. The pore size distribution was bimodal with fine intraparticle pore diameters at 1–6 nm and larger interparticle pore diameters at 30–120 nm as determined by nitrogen adsorption isotherms. The average intraparticle pore diameter decreased with increasing alcohol concentration for methanol and ethanol, while it was rather constant at 3.3 nm, irrespective of alcohol concentration for iso-propanol and sec-butanol. The evolution of particle phase composition was determined by X-ray diffraction ranging from amorphous to crystalline anatase and rutile largely proportional to the calcination temperature and to a lesser extent on the type and concentration of alcohols. C 2000 Academic Press Key Words: titania powders; alcohol; bimodal pore size distribution; phase composition.

INTRODUCTION

Titania powders are used as cosmetics, pigments, photocatalysts, adsorbents, catalytic supports, and sensors (1–4). In these applications, the morphology, average particle size and particle size distribution, phase composition, and porosity of the titania powders are important factors to be controlled. Titania powders are produced by liquid and gas phase reactions. In the liquid phase process, aqueous-based methods using metal salts as a precursor material, such as the forced hydrolysis method (5) and the homogeneous precipitation method (6, 7), encounter several challenges; e.g., the concentration of reacting species should be low and counterions of the starting metal salts remain in the product and deteriorate the purity of powders (8). Sol–gel synthesis provides a way of making nanoparticles by a chemical reaction in solution starting with metal alkoxides as a precursor. An advantage of the sol–gel synthesis is that it is
1 2

On leave from Konyang University, South Korea. To whom correspondence should be addressed.

possible to obtain large amounts of powder with a high level of chemical purity (9). In this synthesis route hydrolysis and polycondensation reactions take place when TTIP reacts with water (10). These two reactions form nuclei (2 to 10 nm in diameter) which develop into primary particles (50 to 100 nm in diameter). These primary particles aggregate to form secondary particles (300 to 1000 nm in diameter) during aging (11). During solvent removal processes (filtration or centrifuging) from the solution, these secondary particles agglomerate to form wet gel powders. Through drying, solvents in the wet gel powders are evaporated to give dried powders (xerogel powders) with large shrinkage. In the above particle formation process, the TTIP is dissolved in alcohol to moderate the high reactivity of the TTIP with water during hydrolysis. When different type and concentration of alcohols are added during hydrolysis, it is expected that these may affect the hydrolysis rate of TTIP and, subsequently, the aggregation rate of the primary particles. As a result, this may affect primary and secondary particle sizes, morphology, phase composition, and porosity of the powders. St¨ ober et al. (12) synthesized monodisperse, spherical silica particles from combinations of different silicon alkoxides and alcohols. It was found that the hydrolysis rate became slower in the order of methanol, ethanol, n -propanol, and n butanol. They reported that final particle sizes obtained under comparable conditions were smallest in methanol and biggest in n -butanol. Harris and Byers (13) studied the growth of colloidal titania by titanium ethoxide hydrolysis in ethanol, n -butanol, and tert-butanol. They found that particle growth was controlled by the surface phenomena of particles and that the morphology of precipitates formed was affected by the solvent: monodisperse spherical particles were formed in the solvents of ethanol and n -butanol, but the precipitate formed in tert-butanol had no regular shape. Park et al. (14) investigated the formation of titania powders from thermal hydrolysis of TiCl4 and the morphology of the precipitates dispersed in various alcohols. They reported that the size of secondary particles decreases as the molecular weight of alcohols increases. Recently, Vorkapic and Matsoukas (15) synthesized titania powders from various titanium alkoxides and investigated the effect of alcohol on the size of the titania particles. They claimed that the smallest size of the particles was obtained when no alcohol was added to
0021-9797/00 $35.00
Copyright C 2000 by Academic Press All rights of reproduction in any form reserved.

289

and Halender (BJH) method with cylindrical pore size (20). 10. another exothermic peak at 390◦ C is observed.05–0. Although many research groups have explored the effect of alcoholic solvent on the morphology and particle size. E10. Micromeritics) in the relative pressure ( P / P0 ) range of 0. the alcohol added replaces the alcohol attached to titanium atom in TTIP) took place (17). 97%. respectively. In especial. Bahr) at a heating rate of 10◦ C/min in flowing air. the resulting precipitate consists of agglomerated masses of particles that have been fused together by further reaction. When TTIP mixed with alcohols. However. The filtered powder was dried at 150◦ C for 24 h. E5. it is expected that the E20 powder is amorphous hydrated titania up to 350◦ C before the strong exothermic peak appears. and E10 powders which Titanium tetraisopropoxide (TTIP. The molar ratio of H2 O to TTIP was 20. . As this peak is more noticeable in intensity for amorphous powders. while IA and IR are the X-ray integrated intensities of the (101) reflection of anatase and the (110) reflection of rutile. the alcoholysis reaction (i. iso-propanol. Hettich) at 9000 rpm for 10 min. This indicates that the E5. while the presence of alcohol resulted in larger particles. 5. Prior to reaction with water. The specific surface area of the powders was obtained by multipoint BET nitrogen adsorption (ASAP2010. the N0. which is associated with the oxidation of residual organic components after hydrolysis (21).e. The desorption isotherm was used to determine the pore size distribution using the Barrett. RESULTS AND DISCUSSION (1) Thermal Analysis Figure 1 shows the differential thermal analysis (DTA) of powders dried at 150◦ C and made at different ethanol concentrations. none of the above referenced studies has examined the effect of these solvents on the phase transformation and porosity of the titania powders. while methanol. The powders were analyzed by thermogravimetric differential thermal analysis (TG-DTA. ethanol.290 SONG AND PRATSINIS the reaction mixture. Look and Zukoski (16) studied shear induced agglomeration during the homogeneous nucleation and growth of particles formed through the hydrolysis and condensation of tetraethylortho-titanate in aqueous ethanol solutions. On the other hand. the precipitate was not produced in the other alcohols. Differential thermal analysis (DTA) curves of sol–gel made titania powders. and sec-butanol were used as solvents.. 1. Clear liquid was removed from the slurry by separation in a superspeed centrifuge (Rotina35. In this paper. EXPERIMENTAL where X is the weight fraction of rutile in the powders. a slurry containing white hydrated titania precipitate was formed (18). this peak is not found in the N0 and E5 powders and shows a weak intensity in the E10 powder. The weight fraction of rutile or anatase in the titania powders was calculated by (19) X = 1/[1 + 0. a strong exothermic peak is observed at 260◦ C. and E20 (Table 1). Ti(OC3 H7 )4 . The powders were made at ethanol/TTIP molar ratios of 0.8 IA / IR ]. An X-ray diffractometer (D5000. we investigate the effect of the type and concentration of alcoholic solvents added during hydrolysis on the texture and phase composition of product titania powders. and E20 powders contain more residual organic components than the N0 powder possibly from incomplete hydrolysis. followed by calcination at different temperatures ranging from 300 to 800◦ C for 1 h at a heating rate of 10◦ C/min. when TTIP was dissolved in methanol. TTIP was first dissolved in various concentrations of different alcoholic solvents. Siemens) was used for identification of crystalline species in the calcined powders with Ni-filtered Cu K α radiation as the X-ray source for diffraction angles 2θ between 20◦ and 60◦ . In contrast. for the N0 powder (Table 1) made in the absence of alcohol. a weak exothermic peak is observed at 260◦ C. a precipitate was produced since the alcoholysis reaction produced titanium isopropoxy methoxides insoluble in methanol. On adding water into the mixed solution. and dried at 150◦ C.25. Aldrich) was used as a precursor. Hydrolysis was carried out at room temperature by adding a constant amount of distilled water slowly to the mixed solution (TTIP and alcohol) for 30 min with a vigorous stirring and maintaining the stirring for another 30 min to attain the equilibrium of the reactions (hydrolysis and particle formation). E5. Joyner. corresponding to N0. For the E5. However. For the E20 powder. FIG. All powders were degassed at 150◦ C prior to measurement. and E20 powders (Table 1) made in the presence of ethanol. corresponding to the crystallization of amorphous hydrated titania to anatase TiO2 (21). E10. however. The molar ratio of alcohol/ TTIP was varied in the range of 0–40. E10. They found that there is a minimum shear rate below which uniform particles are precipitated but when the precipitating suspension is sheared above this value. STA501. and 20.

54 Sample no.1 596. (22).58 0. Hence. did not show such a strong peak have already had anatase TiO2 before 390◦ C as shown by XRD in Figs.19 0. and E20 powders calcined at three temperatures as a function of ethanol/TTIP molar ratios. The powders were made at ethanol/TTIP molar ratios of 0. P10.56 0. . the amount of residual alkyls preventing crystallization to anatase was small. This indicates that the E5. On the other hand. The weight reduction of sol–gel made titania powders by thermogravimetric (TG) analysis in flowing air. E5. and E20 powders have more residual organic components than the N0 powder due to the incomplete hydrolysis. E10. 3b–3d that the phase transformation from amorphous to anatase occurs at 300◦ C for the E5 and E10 powders and at 400◦ C for the E20 powder. 10. In our study. inhibiting the hydrolysis of TTIP. This is in agreement with the TG-DTA results of Figs. However. It is seen in Figs. 3a. and 20. E10.63 0. E5. (2) Phase Composition Figures 3a–3d represent the XRD patterns of the N0. the phase transformation temperature of amorphous to anatase increases. respectively. and sec-butanol.37 0.29 0. E10.5 592. E10.2 Pore volume (cm2 /g) 0.4 331. 5.4 559. while the N0 powder is anatase. When the initial ethanol concentration increases.36 0. and dried at 150◦ C. 1 and 2.0 489. respectively.54 0.64 0. regardless of the initial ethanol concentration at a constant calcination temperature. E5.4 523.6 581. However. and E20 powders. regardless of the initial ethanol/TTIP molar ratio. TTIP molecules in ethanol are more diluted.4 402. E5. the N0 powder shows anatase at 150◦ C and maintains it till 500◦ C in Fig. This is in a good agreement with the DTA data in Fig. 1. respectively. The E5.6 554. This confirms that the anatase to rutile transformation rate is not influenced by the ethanol concentration. It is seen that the phase transformations from amorphous to anatase and anatase to rutile occur at 300 and 600◦ C. 1 and 2.7 564. only the first zone is observed. when the TTIP is dissolved in ethanol. E10. which is confirmed by TG-DTA results in Figs.3 516. at various calcination temperatures.52 0. The rutile weight fraction shows almost a constant value. when the water and HCl concentration were large.31 0. and E20 (Table 1). and E20 powders. iso-propanol. Figure 4 shows the calculated rutile weight fraction in the N0. The second zone from 230 to 380◦ C corresponds to the oxidation of residual organic components and removal of chemically adsorbed water (21). which corresponds to the removal of physically adsorbed water and alcohol. and B10 powders (Table 1) calcined at various temperatures after being prepared with methanol.7 446. E10. E10. Figure 2 shows the thermogravimetric (TG) analysis of the N0. two main zones in weight loss are shown.SOL–GEL MADE TITANIA IN ALCOHOL SOLVENTS 291 TABLE 1 Effect of Type and Concentration of Alcohols on the Specific Surface Area and Pore Volumes of the powders dried at 150◦ C Alcohol/TTIP molar ratio 0 5 10 20 40 5 10 20 40 5 10 20 40 5 10 20 40 Specific surface area (m2 /g) 354.44 0. for the N0 powder. N0 M5 M10 M20 M40 E5 E10 E20 E40 P5 P10 P20 P40 B5 B10 B20 B40 Alcohol type No alcohol Methanol Methanol Methanol Methanol Ethanol Ethanol Ethanol Ethanol Iso-propanol Iso-propanol Iso-propanol Iso-propanol Sec-butanol Sec-butanol Sec-butanol Sec-butanol physically adsorbed water and alcohol. However. For the E5. The first zone from 50 to 230◦ C corresponds to the removal of FIG. when the amount of water and HCl addition during hydrolysis of TTIP was small. and E20 powders dried at 150◦ C are amorphous. Figures 5a–5c represent the XRD patterns of the M10. Thus.2 624. the dried powders made in ethanol are amorphous. so the molecular structure of the dried powder resembled that of anatase. corresponding to N0.25 0. 2. large amounts of unhydrolyzed alkyls remained in the dried powder. 3a–3d. the phase transformation from anatase to rutile begins to appear at 600◦ C for all powders. According to Terabe et al.26 0.2 540.4 432. These alkyls prevent crystallization to anatase so that the dried powders are amorphous. and E20 powders dried at 150◦ C.33 0.2 546. which is more stable than amorphous titania. dried powders made in ethanol contain more residual organic components than the N0 powder.41 0. This means that the initial ethanol concentration has no effect on the phase transformation temperature of anatase to rutile. while those made without ethanol are anatase.

and E20 powders as a function of calcination temperature. irrespective of alcohol type. and (d) 20. in this study. The XRD patterns of sol–gel made titania powders calcined at various temperatures. the alcoholysis reaction. (c) 10. it is interesting to observe that the alcoholysis reaction has no significant effect on the phase transformations from amorphous to anatase and anatase to rutile. All powders show a decrease in the SSA with increasing calcination temperature from the phase transformation and crystallite growth. might take place. while the powders prepared with ethanol exhibit a . 3. (b) 5. Figure 6 shows the specific surface area (SSA) of the N0. However. E10. The powders were made at ethanol/TTIP molar ratios of (a) 0. resulting in different hydrolysis rate of TTIP.292 SONG AND PRATSINIS FIG. E5. When TTIP mixed with various types of alcohols. It is interesting to observe that the N0 powder shows a moderate decrease without a rapid decrease in SSA during calcination. where the alcohol added replaces the alcohol attached to titanium atom in TTIP. respectively.

leading to more dense packing of the primary particles within the secondary particle. irrespective of the alcohol concentration. However. the dilution effect of excess alcohol seems to have . the isotherm shows high adsorption. respectively. and methanol for high alcohol/ TTIP molar ratios of 20 and 40. iso-propanol. iso-propanol. (3) Texture Property Table 1 summarizes the SSA and pore volume of the powders dried at 150◦ C and made with different type and concentration of alcohols. In this study. we stirred the solution during particle formation vigorously. is attributed to the way in which the aggregates of primary and secondary particles are packed. The pore volume of the dried powders decreases in the order of sec-butanol. Figures 8a and 8b show the pore size distribution of the powders dried at 150◦ C and made with different concentrations of ethanol and sec-butanol. all powders made with sec-butanol show type IV isotherm with two hysteresis loops (Fig. In the sol–gel synthesis of titania. The SSA and pore volume show a maximum at molar ratio of alcohol/TTIP = 20 and then decrease with increasing iso-propanol and sec-butanol concentration. indicating that the powder contains micropores (type I). These primary particles then organize into secondary particles (300 to 1000 nm in diameter) by aggregation (11). rapid decrease in the SSA between 150 and 300◦ C for E5 and E10 powders and between 300 and 400◦ C for E20 powder. This rapid decrease in the SSA is attributed to the phase transformation from amorphous to anatase phase. The shape of the hysteresis loops is of type H3 associated with aggregates of plate-like particles giving rise to slit-like pores (20. E10. It is seen that the average intraparticle pore diameter decreases with increasing ethanol concentration (Fig. while that shows almost a constant value (3. irrespective of alcohol concentration for sec-butanol (Fig. the SSA decreases in the order of secbutanol. indicating the presence of mesopores (type IV).3 nm). 23). 7b). the isotherm of the E40 powder with excess ethanol concentration is a combination of types I and IV (20. when TTIP reacts with water. The dried powders made with methanol show smaller SSA and pore volume than those from the other alcohols at a given alcohol/TTIP molar ratio. ethanol. sec-butanol. the curve exhibits a hysteresis loop. 4. 7a are of type IV with two hysteresis loops indicating bimodal pore size distributions. the reactions result in the formation of uniform spherical particles. at high relative pressures between 0. The intraparticle pores are associated with the voids between the primary particles within the secondary particles. Thus. 8b). ethanol. According to BDDT classification (20. This resulted in the formation of agglomerates and not uniform particles. and methanol. for iso-propanol and sec-butanol.SOL–GEL MADE TITANIA IN ALCOHOL SOLVENTS 293 FIG. respectively. it goes through the hydrolysis and polycondensation reactions to give nuclei (2 to 10 nm in diameter) which develop primary particles (50 to 100 nm in diameter). The SSA ranges from 332 to 624 m2 /g and decreases along with pore volume with increasing methanol and ethanol concentration. and E20 powders in Fig. The calculated rutile weight fractions of the powders made at differential ethanol/TTIP molar ratios and calcined at various temperatures for 1 h. Figures 7a and 7b show the adsorption isotherms of the powders dried at 150◦ C and made with various alcohol concentrations of ethanol and sec-butanol. 23). and E20 powders. On the other hand. associated with narrow slit-like pores.0. The shape of the hysteresis loops is also of type H3. while the interparticle pores are due to the voids between the secondary particles within the agglomerates (24). The shape of the hysteresis loop is also of type H3. above a critical shear rate the final precipitate is composed of agglomerates consisting of fused particles. However. 23) with two distinct regions: at low relative pressures. This results in a decrease in the average diameter of intraparticle pores and the SSA with increasing alcohol concentration. E10. Further aggregation of these secondary particles gives agglomerates with several tens of micrometers in diameter (11). and methanol for all alcohol/TTIP molar ratios. the isotherms corresponding to the E5.8 and 1. However. The bimodal pore size distributions of the titania powders in our study are closely related with the above particle formation mechanism. 8a). All powders show bimodal pore size distribution consisting of smaller intraparticle pores and larger interparticle pores. The pore size distribution of the powders made in methanol shows similar behavior with ethanol. When methanol and ethanol are used as solvents. iso-propanol. the observed difference in our experiment. in terms of the SSA and porosity of the dried powders. However. higher alcohol concentration separates the primary particles formed and hinders the progress of aggregation due to the dilution effect of excess alcohol. Our work showed results similar to those of Look and Zukoski. while that made in iso-propanol exhibits similar behavior with sec-butanol. For low alcohol/TTIP molar ratios of 5 and 10 the SSA decreases in the order of ethanol. Unlike the isotherms of the E5. Look and Zukoski (16) claimed that in the absence of shear.

3 nm). The powders were made with (a) methanol. During drying. The molar ratio of alcohol/TTIP was 10. The XRD patterns of sol–gel made titania powders calcined at various temperatures. 5. (b) iso-propanol. while powder made in methanol has smaller average of intraparticle pore diameter than ethanol-made powder. Figures 9a and 9b represent the pore size distributions of the powders dried at 150◦ C and made with different types of alcohol for alcohol/TTIP molar ratios of 10 and 20. respectively. For both ratios. and (c) sec-butanol. This can be explained by the difference of capillary force during drying of wet gel powders. It is interesting to observe that powders made in iso-propanol and sec-butanol show the same average intraparticle pore diameter (3. large shrinkage . the average intraparticle pore diameters of powders made in the presence of iso-propanol and sec-butanol are larger than those made in the presence of methanol and ethanol. little effect on the average diameter of the intraparticle pores and the SSA.294 SONG AND PRATSINIS FIG.

the dielectric constant of the alcohol decreases (25). Figures 10a and 10b show the pore size distributions of the N0 and E20 powders (Table 1). respectively. FIG. where γ is the surface tension. r is the pore radius. leading to a looser packing of secondary particles in the former solvents. The surface tensions of the alcohol considered in this study have similar values (25). and θ is the contact angle. in the wet gel powders takes place by solvent evaporation in the pores. . the most important factor which affects the capillary force becomes the pore radius in the wet gel powders. 6. The capillary force ( P ) can be expressed by (17) P = 2γ cos θ/ r. the aggregation rate may be expected to be faster in isopropanol and sec-butanol than in methanol and ethanol. and n -propanol. E10. 7. leading to the dried powders with looser packing. Thus. the pores FIG. and E20 powders (Table 1) as a function of calcination temperatures. For both N0 and E20 powders. the intraparticle pores of the N0 powder have not disappeared completely and show a small peak. leading to an enhanced rate of aggregation of primary particles in the solution (14). At 600◦ C. Thus. E5. This looser packing of secondary particles has large pores and thus the capillary force during drying becomes smaller. This results in decreased stability. This is probably the reason why the average intraparticle pore diameters of the dried powders made with iso-propanol and sec-butanol are bigger than those made with methanol and ethanol. Similar result was reported by Vorkapic and Matsoukas (15) who made titania powders at various solvents. ethanol. as a function of calcination temperatures.SOL–GEL MADE TITANIA IN ALCOHOL SOLVENTS 295 show bimodal distributions consisting of the intraparticle and interparticle pores at 500◦ C. The specific surface areas of the N0. As the molecular weight of alcohol is increased. while those of the E20 powder collapse completely. This shrinkage determines the agglomeration state of the resultant dried powders and is strongly affected by the capillary force which arises from the evaporation of the solvents in the wet gel powders. Nitrogen adsorption isotherms of the powders dried at 150◦ C and prepared with various concentrations of (a) ethanol and (b) sec-butanol. They found that the degree of aggregation increased in the following order: methanol.

. the FIG. That means that the intraparticle pores of FIG. 8. When the calcination temperature increases from 600 to 800◦ C. The molar ratios of alcohol/TTIP were fixed to (a) 10 and (b) 20.296 SONG AND PRATSINIS interparticle pores have completely disappeared for both powders. suggesting pore growth. compared to that in the E20 powder. It is seen in both powders that the pores exhibit monomodal distribution of the interparticle pores due to the complete collapse of the intraparticle pores at 800◦ C. Pore size distributions of the powders dried at 150◦ C and prepared with various types of alcohol. the average interparticle pore diameter increases. At 1000◦ C. It is interesting to note that the intraparticle pores in the N0 powder have not disappeared completely at 600◦ C. Pore size distributions of the powders dried at 150◦ C and prepared with various concentrations of (a) ethanol and (b) sec-butanol. 9.

The effect of alcohol type on particle aggregation rate is explained by dilution and colloidal stabilization. For a given alcohol. . 9a and 9b. The average intraparticle pore diameter decreased with increasing alcohol concentration for methanol and ethanol. (26) observed the microstructures of titania membranes calcined at various temperatures by high-resolution scanning electron micrograph. while that showed a constant value (3.SOL–GEL MADE TITANIA IN ALCOHOL SOLVENTS 297 could be attributed to the phase transformation of anatase to rutile. independent of the type and concentration of alcohols. the dilution effect has little effect on the average diameter of the intraparticle pores in Fig. the dielectric constant of the alcohol decreases. This increases the average diameter of intraparticle pores with increasing molecular weight of alcohol as shown in Figs. The intraparticle pores disappeared at 600◦ C. whereas the interparticle pores collapsed completely at 1000◦ C. as its concentration increases. Kumar et al. They found that the powders before anatase to rutile phase transformation contain small anatase crystallites. the phase transformation temperature of anatase to rutile was at 600◦ C. The porosity of the powders dried at 150◦ C could be controlled by the alcohol type and concentration. Pore size distributions of the (a) N0 and (b) E20 powders (Table 1) calcined at various temperatures. Another interesting result from these figures is that the main cause of the disappearance of most intraparticle pores at 600◦ C The type and concentration of alcohol added during hydrolysis of TTIP play an important role on the phase composition and texture of sol–gel made titania powders. for iso-propanol and sec-butanol. SUMMARY FIG. 10. leading to the disappearance of the voids between anatase crystallites which results in the collapse of intraparticle pores in the bimodal pore size distributions. It is thought that smaller anatase crystallites grow into bigger rutile crystallites through the phase transformation. 8a. the dried powders were amorphous as residual organic components prevented the crystallization to anatase. This results in a decrease in the average diameter of intraparticle pores with increasing alcohol concentration in Fig. All dried powders showed bimodal pore size distributions consisting of fine intraparticle pores and larger interparticle pores. irrespective of alcohol concentration for iso-propanol and sec-butanol. Alcohol retards the hydrolysis reaction of the TTIP and results in residual organic components in the powders by incomplete hydrolysis. as confirmed in Figs. but the powders after transformation have smaller anatase crystallites and bigger densified rutile regions. when ethanol was added. leading to an enhanced rate of aggregation of primary particles within the secondary particle. the powder made with ethanol collapse more easily than those of the powder made without ethanol. at a given alcohol/TTIP molar ratio as the molecular weight of alcohol increases. leading to more dense packing of the primary particles within the secondary particle. However. On the other hand. On the other hand. However. the powders dried at 150◦ C were anatase.3 nm). 8b. the number concentration of primary particles is decreased and the aggregation rate is reduced. 3a and 3d. When ethanol was not added during hydrolysis of TTIP.

and Burggraaf. G. 16. S. B. 4.. 461 (1992).. 26. A. Dominguez. E. 18. 29. 2. and Siemieniewska. C. REFERENCES 1. H. J. F. Yamaguchi. Miyazaki.. M. 21. 207 (1992). S. Ceram. Mezey. 62 (1968).-L.. Ceram.. 301 (1984). J.. and Sing. Chem. J. 1617 (1994). L. 153. and Kim. “Sol-Gel Science. Sing. H. S. Matijevic.... Vemury. T. Sci. S. R. in press (2000). 1574 (1986). Y. D. E. 12... C. W. Eng. A.” Academic Press. Soc... A. K. H. J. and Burggraaf.. R.. San Diego. FL. Song. and Scherer. Soc... J. J. 81. 1087 (1986).. Kato. We acknowledge the assistance and contribution of Dr. Chem. M. R. K. Pierotti. and Porosity. J. D. “Adsorption. Takada. in press (2000). Lett. K. Inoue. L. P.. Pratsinis. Spurr. 5... Kominami. D. Chem. Res. Am. J. Tseng. S. W. Bull.. Haul. Ceram.. A. M. Boca Raton. H. Sci.. Fink. Chem. 1990. and Lee. C. Canales. E. I. Colloid Interface Sci. New York. and Iguchi. 49. Moscou. and Pratsinis. Surfae Area. J. J. W. J. S. D. Soc. 19. 6. Catal. Am. R. Look. 23. K. Ceram. Non-Cryst. Lett. and Akinc. 47 (1988). K. 49 (1988). 50. 24.. A. Brinker. C.” CRC Press. H. “Handbook of Chemistry and Physics. E. London. 14. and Pratsinis. Mater.. 65. Imai. Rouquerol. J. Vorkapic. K. E. Pure Appl. 1193 (1978).. and Bowen. Am. S. Mater. J.. Mater. Colloid Interface Sci. A. Kumar. A. Mater. J.. Y. 80. 9. Everett. Akhtar. H. 4939 (1994). 57. Harris. Colloid Interface Sci.. 15... Ceram. A. 3. Yamagiwa.. Am. and Bohn. West. Mater. 75. Mater. G. and Matsoukas. 1966. K..... 22.. V. 603 (1985). Soc. W.. J... Winnubst. Eng.. 414 (1985). Koebrugge. W. . R. Sorddet. H. Terabe. 3. and Pratsinis. Sci. J. S. 1989. J. 29. E.. and Byers. Jinsoo Kim in the preparation of the final manuscript.298 ACKNOWLEDGMENTS SONG AND PRATSINIS 10.. C. M. 3408 (1992).. E. J. Anal. Montoya. Kera.. Gregg. Chem. J...... M. Viveros. S. H. J. 13. 197 (1996). Park. T.. 21. Yoldas.. Y. St¨ ober. S.” Academic Press. Pure Appl. K.. and Zukoski. 743 (1997). H. K. 63. 2815 (1998). L. Mater. 26. Bowen. D. T. A.. A.” Wiley. K. J. and Inui. H. E. Barringer. 7. S. and Myers. P. J. Langmuir 1. J..... T. W. 11. 1982. Song. Y.. and Mastrangelo.... 1095 (1993). Chem. 20. T. Y. Sci. Sci.. C.. and Schifter. “Vapor Deposition. 17. M. I. 15. Kim. K. Soc. 760 (1957). Solids 103. Am. Keizer.. 25.. Sacks.. G.. 3. 8. 15. Fotou. K. 122. N. C.. T.. 1141 (1993). K. E. This research was sponsored by the SNF (Swiss National Science Foundation) and KOSEF (Korea Science and Engineering Foundation).

Sign up to vote on this title
UsefulNot useful