3014
Ind. Eng. Chem. Res. 1993,32, 3014-3019
Mechanism of Formation of Titanium Dioxide Ultrafme Particles in Reverse Micelles by Hydrolysis of Titanium Tetrabutoxide Takayuki Hirai,' Hiroshi Sato, and Isao Komasawa Department of Chemical Engineering, Faculty of Engineering Science, Osaka University, Toyonaka, Osaka 560, Japan
The mechanism of the formation of titanium dioxide ultrafine particles in reverse micelles by the hydrolysis of titanium tetrabutoxide (TTB)was studied, using sodium bis(2-ethylhexyl) sulfosuccinate (AOT)/isooctane as a reverse micellar solution. The effects of water, TTB, and surfactant concentrations were investigated. The particle formation process was followed by the change of UV-visible absorption spectra. T h e conditions for the formation of partic1.e~were studied, and the particle formation was found t o be controlled by the T T B concentration and the number of micelles. A reaction scheme for the particle formation, based on the number of micelles, was proposed, which successfully explained the data for the combined stages.
Introduction
Experimental Section
There has been much recent interest both in ultrafine particles and in procedures to prepare and stabilize uniform ultrafine particles. Various chemical methods for preparation have been proposed. Since the particles are prone to aggregate into larger particles, it is necessary to prevent uncontrolled aggregation in the course of preparation. Reverse micelles consist of nanometer-sized water cores which are dispersed in an apolar solvent. Various water-soluble molecules can be solubilized in the water cores, in which ultrafine particles can be prepared. Numerous papers deal with the preparation methods for ultrafine particles in reverse micelles and are summarized in a review (Osseo-Asare and Arriagada, 1990). I t is now possible to prepare composite particles such as CdSe/ZnS in the micelles (Kortan et al., 1990). Recently, several models have been proposed to explain the formation process of ultrafine particles in reverse micelles. For metal boride particles, a quantitative model based on the number of micelles and metal ions was presented to explain the number of formed metal boride particles (Ravet et al., 1987). For silicone dioxide particles, a qualitative model was proposed to explain the variation of particle size with water content (Osseo-Asare and Arriagada, 1990). For cadmium sulfide particles, rapid coagulation at the initial stage of particle growth was found to explain a UV-visible absorbance change at 280 nm (Towey et al., 1990). Kinetic studies are of great importance in understanding the mechanism, controlling the stagewise processes, and controlling the rate of particle formation. These studies can lead to further development for the preparation of particles having special composition and functions. In the present work, the mechanism of particle formation by hydrolysis of titanium tetrabutoxide (TTB) in AOT reverse micelles was studied. This system was employed because titanium and zirconium dioxide particles were successfullyprepared in a previous work (Hirai et al., 1992). The whole process from the hydrolysis of TTB to form hydrolyzed species to the formation of particles was followed by the measurement of the UV-visible spectra of the micellar solution. A kinetic scheme was then proposed, for the stagewise reactions including nucleation and nuclear growth in the micelles, based on the analysis of the spectra obtained with the micellar solution and with a homogeneous organic solution.
Sodium bis(2-ethylhexyl) sulfosuccinate (AOT), titanium tetrabutoxide (TTB), and l-butanol were supplied by Wako Pure Chemical Industries, LM., and used without further purification. Isooctane (2,2,Ctrimethylpentane) supplied by Ishizu Seiyaku, Ltd., was dehydrated by using molecular sieves 3A and was filtered using a 0.2-fim membrane filter prior to use. Reverse micellar solution was prepared by dissolving AOT in isooctane and by filtrating using the membrane filter, followed by addition of a required amount of prefiltered distilled water. The concentration of surfactant, [AOTI, was 0.05 M or 0.1 M (M = mol/L). The water content (water to surfactant molar ratio, W O= [H201/[AOTl) was determined by KarlFisher titration and was varied in the range of 9-30. The TTB solution was prepared by injecting a required amount of 1M TTB/1-butanol stock solution into the dehydrated isooctane. The hydrolysis of TTB was carried out in a beaker-type reactor (10 mL) at 25 "C. The reaction was initiated by injecting the TTB solution (0.1 mL) into the reverse micellar solution (5 mL) with mild stirring (300 min-1) generated by a magnetic stirrer. UV-visible absorption spectra were recorded on a Shimadzu UV-265FW spectrophotometer. About 3 mL of reaction mixture was poured into a quartz cell with a cap for the measurement of the change of absorption spectra for a period of up to 12 h. When the reaction time was greater than 1day, the reaction was carried out in a capped reactor to reduce vaporization and leakage of isooctane. The solution was transferred to the cell to measure the UV spectra when necessary. The size distribution of the reverse micelles and the ultrafine particles was measured via a dynamic light scattering spectrophotometer (DLS, Otsuka Electronics DLS-700Ar). The diameter of the particles was measured as follows. Ethanol (3 mL) was added to 5 mL of the reverse micellar solution containing particles. The solution was separated into two phases and the particles, suspended in the lower phase, were measured with DLS. The formation of particles in a homogeneous organic solution was also carried out for comparisonwith the results in micellar solutions. In this case, l-butanolwas employed as an organic solution instead of isooctane, since isooctane could dissolve only a very limited amount of water. The other operations were the same as those for the reverse micellar solutions.
0888-588519312632-3014$O4.0O/0
0 1993 American Chemical Society
Ind. Eng. Chem. Res., Vol. 32, No. 12, 1993 3016
Wave Length ( n m )
Wave Length ( n r n )
Wave Length ( n m )
Figure 1. Absorption spectra taken from 5 min to 3 days after the initiation of hydrolysis of TTB.
Results and Discussion Change i n Absorption Spectrum of Solution during Hydrolysis of TTB. The formation of metal oxide particles by hydrolysis of metal alkoxides involves a hydrolysis stage, a nucleation stage, and a particle growth stage. The first stage is the hydrolysis of the alkoxides and the formation of hydrolyzed species. In the case of TTB, the overall hydrolysis reaction is described as
-
Ti(OC,Hg), + 4H,O Ti(OH), + 4C4HgOH (1) Titanium dioxide is, then, formed by the condensation of the hydrolyzed species.
-
Ti(OH), TiO, + 2H,O (2) In the nucleation stage, the nuclei are formed from the hydrolyzed molecules. Since small particles have high solubility (La Mer, 1952),a certain number of hydrolyzed molecules in the oversaturated solution are required in order to form stable nuclei which grow to larger particles. In the growth stage, larger particles are formed by particle growth, aggregation, and Ostwald ripening. Particle growth, however, is the principal process and this occurs by addition of hydrolyzed molecules. On the other hand, aggregation and Ostwald ripening are secondary processes and do not require hydrolyzed molecules. Aggregation is the result of Brownian motion and the coagulation of particles. The latter is via growth of larger particles accompanied by the dissolution of the smaller particles and differences in solubility. Figure 1 shows the UV-visible absorption spectra obtained during the hydrolysis of 3 X 106 M TTB for various systems. In the 3 M HzO/l-butanol (homogeneous) system, the absorption increased slightly and monotonously and no further spectral change was observed after 12h, as shown in Figure la. In the Wo= 30 reverse micellar system, where the overallwater concentration was the same as that in the homogeneous system, different behavior patterns were observed, as shown in Figure lb. In this, the absorption increased until 6 h and then decreased at a wavelength less than 300 nm. On the other hand, in the WO= 9 reverse micellar system, the spectral changes were similar to those in the homogeneous system, as shown in Figure IC. However,when water was added to this system, following 6 h, 1day, or 3 days, in order to increase W Oto 30, the absorption was found to decrease as in the case of WO= 30 within 1 day, following the addition of water. These results indicate that the particle formation process can be separated into two stages. The first stage proceeds in all systems within 12 h, but the second stage occurs only in the WO= 30 reverse micellar system. Because of the large excess of water (more than 104-fold), the hydrolysis of TTB is likely to proceed in all cases. The
subsequent nucleation requires a certain number of hydrolyzed molecules and proceeds only under specific conditions. The difference in the spectral changes can, therefore, be attributed to the nucleation and particle formation. Only the hydrolysis of TTB occurred in the WO= 9 reverse micellar system and in the homogeneous solution system. In these systems the stable spectra observed can be ascribed to hydrolyzed species. On the other hand, both the hydrolysis and the particle formation occurred in the WO= 30 reverse micellar system, and the spectrum observed after 3 days can be ascribed to titanium dioxide particles. Kinetics of Particle Formation. The kinetics of the overall hydrolysis reaction of alkoxides in the particle formation process was studied by the measurement of the induction period, i.e., the time between the mixing of the reagents and the first observation of particles. The kinetics of the hydrolysis of titanium alkoxides in water/alcohol homogeneous solution has been reported (Harris and Byers, 1988). In their study, titanium tetraethoxide was diluted with l-butanol to obtain titanium butoxide. Then the alkoxide solution was added to water/alcohol solution and particles were formed. The induction period obtained in l-butanol was much longer than that in ethanol and was 120 min when the alkoxide and water concentrations were 0.016 and 0.38 M, respectively. This result indicates that the overall hydrolysis reaction of titanium tetrabutoxide is slower than that of titanium tetraethoxide. Furthermore, the induction period varied inversely with alkoxide concentration in l-butanol, thus indicating that the kinetics of the overall hydrolysis of TTB was first order with respect to TTB. The particle formation was followed by taking the absorption spectra of the reaction solution. Figure 2 shows the absorbance change at 280 nm during the hydrolysis of TTB in the reverse micellar solutions. In the WO= 30 system, the absorbance increased to a maximum value and then decreased. In the WO= 9 system where particle formation was not observed, the absorbance increased rapidly together with that of the WO= 30 system at the initial part of the reaction and then increased slightly. This may be due to the vaporization of isooctane and the increase of concentration of AOT in solution. This vaporization was confirmed by a reduction in the volume of solution. The increase and decrease of absorbance are related to the slow hydrolysis of TTB and particle formation, respectively. To study the rate of the hydrolysis reaction in the reverse micelles, the first part of the kinetic transient up to 120 min in the absorbance was analyzed in terms of first-order kinetics. The half period of the hydrolysis of 3 X 10-5M TTB shown in Figure 2 was about 1 h. The latter part of the absorbance change was also analyzed in terms of first-order kinetics, and for the system
3016 Ind. Eng. Chem. Res., Vol. 32, No. 12, 1993
-
-I I
0.14
$
0
o
-
Wo = 30 (observed) Wo = 9 (observed) Wo = 30 (Calculated) 0.1M AOT/isooctane [ TTB ] = 3x1 0-5M
200 400 600 Time ( min )
i 800
Figure 2. Absorbance changes at 280 nm during hydrolysis of TTB. Comparison of observed data with calculated data.
of [TTBI = 3 X lo4 M and W O= 30, the half period was about 6.5 h. This latter part may correspond to the consumption of hydrolyzed molecules for particle formation. These values of half period will be used for the simulation in the latter part of this work. Factors of Particle Formation. In homogeneous solution, both the alkoxide concentration and the water concentration are known to affect particle formation and the size distribution of the particles (Ogihara et al., 1989). In the reverse micellar system, the overall water concentration is controlled by the water content (IHzOl/[AOTl) and by the surfactant concentration ([AOTI). The effects of water content, TTB concentration, and surfactant concentration on particle formation were investigated. Figure 3a-1 shows the absorption spectra obtained after the hydrolysis of 3 X 10-5 M TTB for 3 days at various water contents. In the W O= 30 system, the titanium dioxide particles were formed as mentioned above. When the water content was less than 30, however, the spectral features were similar to that of the hydrolyzed species, showing that no particles were formed in these systems. The increase in the water content thus facilitates the particle formation. The results obtained with 9 X 10-5 M TTB are shown in Figure 3a-3. The absorption in the W O= 30 system was increased by three times compared to that obtained with 3 X 106 M TTB. This indicates that the absorption of formed particles is proportional to the initial TTB concentration. In this TTB concentration, the particles were formed except in the case of the system of [AOTI = 0.1 M and W O= 9. In the case of W O= 15 and [TTBI = 6 X lo6 M, the spectrum lay between that of hydrolyzed species and that of particles, as shown in Figure 3a-2, indicating that the hydrolyzed species was not completely converted to the particles. The increase of TTB concentration also facilitates the particle formation. Furthermore, the decrease of surfactant concentration also facilitates particle formation. In the case of [TTBI = 9 X lo6 M and W O= 9, the particle formation occurred in the [AOTI = 0.05 M system, but not in the [AOTI = 0.1 M system, as shown in Figure 3a-3. Both the increase of water content and the decrease of surfactant concentration are found to facilitate the particle formation. The former increases the overall water concentration of the reverse micellar solution, while the latter causes a decrease. The overall water concentration does not seem to be a controlling factor for particle formation. This problem is reconciled from a viewpoint taken from the number of micelles. The number of micelles, Nm,can be calculated by the average diameter (d,) determined by using DLS and its standard deviation (a) as previously
described (Kuboi et al., 1990). The results for the present systems are shown in Table I. The number of micelles decreases with increasing water content and decreasing surfactant concentration. The average number of the hydrolyzed molecules per micelle, thus, increases with decreasing number of micelles. This facilitates the formation of the nuclei of the particles, since a certain number of hydrolyzed molecules are required to form a stable nucleus. Thus the number of micelles controls the particle formation in the reverse micelles. Figure 4 shows the effect of TTB concentration on particle diameter measured with DLS. The diameters are slightly affected by ?r'B concentration. Sincethe ultrafiie particles have rather weak scattering characteristics, the DLS measurement was not very reproducible. The data represent the average of four to six separate runs, and the arrows indicate the degree of scatter of the data. Quantitative Model for Formation of Particles. Reverse micelles can exchangetheir contents in water cores via collision and redispersion processes. The exchange rate constant is reported to be 106-108 M-I s-l for the AOT/ isooctane system (Fletcher et al., 1987; Lang et al., 1988). Thus, the rearrangement of micelles occurs on amillisecond time scale in the present system, since the concentrations of micelles are in the range of 106-1W M (Table I). The reactions in the particle formation process are much slower than the exchange of the contents. Thus the reactants are likely to be distributed among the micelles according to an equilibrium distribution. Thus, the reaction rate is independent of the exchange rate of micelles (Oldfield, 19911, and is controlled by the number of such micelles. This analysis has been proved experimentally by the oxidation of iodide by persulfate (Muiioz et al., 19911, in which the reaction rate was observed to depend on the reactant concentration per water volume and on the water content. Water-soluble molecules are consideredto be distributed among the reverse micelles according to a Poisson distribution (Atik and Thomas, 1981). The probability to have i molecules in a micelle is given by eq 3, where h is the average number of molecules in a micelle. p i = A' exp(-h)/i! (3) The rate of hydrolysis is expressed by the following relation:
-dNm/dt = kHNm (4) where k~ and N m are the pseudo-first-order rate constant and the number of TTB molecules, respectively. The second step of the particle formation is the nucleation. The nucleation rate is considered to be proportional to the number of micelles containing a sufficient number of hydrolyzed molecules for the nucleation (Nmp). Thus, dNJdt = kNN,,n (5) where kN and N p are the rate constant for nucleation and the number of formed nuclei, respectively. Nm,n is calculated from eq 6, assuming that the nucleation requires n molecules of hydrolyzed TTB, n-1
where Nh is the number of hydrolyzed molecules. The distribution of hydrolyzed molecules @ i ) can be calculated from eq 3, using NJNm for A. The particle growth can occur in micelles containing a nucleus or a particle, in addition to hydrolyzed molecules.
1
0.2-
-
'
.\
-- -
1
'
1
~
1
Wo = 30, [AOTJ = 0.1M Wo E 15, [ A 0 1 = 0.1M Wo.9, [AOT]=O.lM Wo = 9, [AOT] = 0.05M
-
I
/
I
'
/
'
I
- Wo = 30, [AOT] = 0 . l M
0.4-
-
\
-
\'\
Wo = 15, [AOT] = 0.1M Wo = 9, [AOT] = 0.1 M W o = 9 . [AOT]=O.O5M
a-2)
-
-
[ T B ] = ~ x ~ O -- ~ M
- Wo = 30, [AOT] = 0.1M l
0.6-
'
__ \
\\a-3)
-
l
-
-
[TBI = 9 x 1 0 - 5 ~
\
Wo = 9. [AOT] = 0.05M
\\
260
Wave Length (nrn)
@
Wo = 30, [AOTJ = 0.1 M Wo = 9, 15, [[AOT] A 0 7 = 0.1M
Wo = 15, [AOT] = 0 . l M W o = 9 , [AOT]=O.lM Wo = 9. [AOTJ 0.05M
Wave Length (nrn)
l
\
- Wo = 30, [AOTI = 0.1M
Wo = 30, [AOT] = 0.1M
'
\
0.4-
\
W0'=15,[AOTJ=0.1M E 9, [AOTJ = 0.1M Wo W, = 9, [AOT] = 0.05M
l
Wo = 15, [AOTJ = 0 . l M Wo E 9. [AOTJ = 0.1 M Wo.9, [AOT] =O.O5M..
b-3)
[ T B ~= 9~10'~M
300
340
Wave Length (nrn)
Figure 3. Absorption spectra after hydrolysis of TTB for 3 days in reverse micellar systems. (a) Observed; (b) cdculated. Table I. Reverse Micellar Systems Used for the Preparation of Titanium Dioxide Particles [AOTl(M) W O d,(nm) u N,,JlOlg Cm/lOb(M) 0.267 1.42 2.36 0.1 30 19.30 11.97 0.188 7.20 0.1 15 10.72 0.206 9.56 15.88 0.1 9 8.97 4.75 7.90 0.206 0.05 9 8.98
1
0';M AOT/isooctane W,, = 30
10
ddays
observed
- calculated
'I /--0 10-5
10-
5
[TTBI (MI Figure 4. Effect of ?TB concbntration on particle diameter and conversion of TTB molecules to particles. Comparison of observed diameter with calculated diameter.
By assuming that the rate of consumption of the hydrolyzed molecules by the particle growth is proportional to the number of such pregnant micelles and is independent of the particle size, the mass balance on the hydrolyzed molecule can be expressed by the following equation: -dNh/dt = n dNJdt
+ kGNhNdNm+ dNmB/dt
(7)
where 120 is the rate constant taking into account the probability of the particle growth. The number of micelles containing both particle and hydrolyzed molecules is represented by Ndv,/Nm. Each term of eq 7 represenb
the consumption rate of hydrolyzed molecules for nucleation, the consumption rate of hydrolyzed molecules for particle growth, and the rate for the formation of hydrolyzed molecules by hydrolysis. Assuming that no ripening and aggregation occur, that is, the number of particles is identical to the number of nuclei formed, one can simulate the variation of the number of TTB molecules, hydrolyzed molecules, and particles with time by solving the above equations. Application of the SimulationModel to the Present Particle Formation. The proposed model contains four parameters to be determined such as the pseudo-firstorder hydrolysis rate constant (kH),the pseudo-first-order growth reaction rate constant (kN),the pseudo-first-order nucleation rate constant ( k G ) , and the number of hydrolyzed molecules required to form a stable nucleus (n).The value of kH can be determined by the value of the half period of the first part of the absorbance change at 280 nm. This was determined to be 1.9 X 1V s-l, since the half period was 1h. The latter part of the absorbance change after 500 min represents the consumption of hydrolyzed molecules in the nucleation and particle growth stages. In this part, the hydrolysis of TTB is almost finished, and the consumption of hydrolyzed molecules by nucleation is much less than that by particle growth. The first and third terms of eq 7 are, thus, negligible after 500 min, and eq 7 is now simplified as -dNhfdt = (kGNdNm)Nh
(8) Assuming that N p is constant, the observed pseudofirst-order rate constant (k&p/Nm) is calculated to be 2.96 X 10-5 s-1 using the value of the half period, 6.5 h. For the present case, the number of micelles, Nm, was 1.42 X 1019as shown in Table I. The diameter of particles formed was about 3 nm as shown in Figure 4; thus, the number of particles was calculated to be 4.0 X 10'6. The value of kG is now estimated as 1.05 X 10-4 s-l.
3018 Ind. Eng. Chem. Res., Vol. 32, No. 12, 1993 I " " I " "
0.2
0 1M AOTiisooctane 6h
h
I v
a
0
C
fi 0
0.1 -
ii
W, = 30 [ TTB ] = 3 ~ 1 0 - ~ M
(0
P
d
0-
W a v e Length ( n m ) Figure 5. Calculated absorption spectra during hydrolysis of TTB.
For the estimation of the nucleation rate constant, kN, in eq 5, the value of n expressing the number of hydrolyzed molecules to form a nuclei is first needed, since the value of Nm,,, is dependent on the value of n. The calculated diameter of formed particles is very sensitive to the ratio of k G and kN. The value of kN is determined in each case for integer n greater than 1using the observed diameter shown in Figure 4. The variation of particle diameter with TTB concentration was calculated for each n value by simply changing the number of initial TTB molecules to produce the best correlation with the observed data. The value of n was thus determined to be 5. The nucleation rate constant, kN, on the basis of n = 5 is now estimated as 1.9 X 10" s-l. The nucleation process is generally taken to occur in the size range 1-10 nm. In the reverse micelles, however, it must occur at much smaller sizes as suggested by Towey et al. (1990). Actually, Ravet et al. obtained the value of n = 2 for metal boride particles, from the analysis of variation of the number of formed particles with the number of metal ions (1987). The absorption spectrum of the reaction solution can now be calculated using eqs 3-7 and the parameters ( k ~ , kN,kc,and n). The calculated spectra are shown in Figure 5. These spectra show good agreement with the observed data shown in Figure lb. The calculated time course of absorbance at 280 nm is shown in Figure 2. The data obtained at the maximum absorbance were not used in determination of parameters. However, the calculated time course of absorbance shows good agreement with the observed data. The solid line in Figure 4 shows the calculated diameter of particles and the conversionof TTB molecules to the particles after 3 days as a function of the TTB concentration. The calculated diameters show good agreement with the observed data. This is to be expected, since the value of n was estimated based on the observed data of particle diameters. In the range of TTB concentration less than 2 X 10-6 M, the calculation suggests that the conversion to particles is not complete in 3 days of hydrolysis in the reverse micelles, as shown in Figure 4. The particle could grow more in this low concentration range. However, the experiment was not plausible due to the inevitable vaporization and leakage of isooctane and a large absorption due to surfactant compared to that due to particles. The spectra of reverse micellar solutions after hydrolysis of TTB for 3 days were also calculated, and typical results are shown in Figure 3b. These spectra show very good agreement with the corresponding observed data shown in Figure 3a. The presented reaction scheme gives a satisfactory description of the slow particle formation process by hydrolysis of titanium tetrabutoxide in reverse micelles from the hydrolysis stage to particle growth stage.
Conclusion The mechanism of formation of titanium dioxide ultrafine particles in AOT/isooctane reverse micelles by hydrolysis of titanium tetrabutoxide (TTB) was studied. The particle formation process was followed by the UVvisible absorption spectra. The following results were obtained: 1. Different spectrum changes were observed between an organic homogeneous solution containing 3 X 1od M TTB and 3 M water and a reverse micellar solution containing identical amounts of TTB and water. The difference was caused by the formation of particles. The particles were formed in reverse micelles, but not in homogeneous organic solution under the relevant conditions. 2. The effects of water content, TTB concentration, and surfactant concentration on the particle formation were investigated. The increase of TTB concentration and water content and the decrease of surfactant concentration were found to facilitate particle formation. Particle formation was controlled by TTB concentration and the number of micelles. 3. A reaction scheme for the particle formation was proposed. The number of micelles is very important, since a certain number of hydrolyzed molecules are required to form a stable nucleus. The rate of consumption of the hydrolyzed molecules by particle growth is assumed to be proportional to the number of micelles having a nucleus or a particle and to be independent of particle size. The scheme can explain successfully the observed data for the combined stages of particle formation. Acknowledgment The authors gratefully acknowledge financial support from a Grant-in-Aid for Scientific Research of the Ministry of Education, Science and Culture, Japan (No. 04453124, 1992).
Nomenclature C, = molar concentration of micelles, M (mol/L) d, = average diameter of reverse micelles, nm d, = average diameter of particles, nm k H = pseudo-first-orderrate constant of hydrolysis of TTB, 5-1
k~ = pseudo-first-orderrate constant of particle growth, s-1 kN = pseudo-first-order rate constant of nucleation, s-1 n = number of required hydrolyzed moleculesto form astable nucleus Nh = number of hydrolyzed molecules in 1 L of reaction solution N , = number of micelles in 1L of reverse micellar solution N,,,, = number of micellescontaining more than n hydrolyzed molecules in 1 L of reverse micellar solution N , = number of formed particles in 1 L of reaction solution N ~ = number B of TTB molecules in 1L of reaction solution pi = probability to have i molecules in a micelle W , = [HzO]/[AOT] = water to surfactant molar ratio (water content) u = standard deviation of diameter distribution X = average number of molecules in a micelle [ ] = molar concentration of species in the brackets, M (mol/ L)
Literature Cited Atik, S. S.; Thomas, J. K. Transport of Photoproduced Ions in Water in Oil Microemulsions: Movement of Ions from One Water Pool to Another. J. Am. Chem. SOC.1981, 103 (12), 3543-3550.
Ind. Eng. Chem. Res., Vol. 32,No. 12, 1993 3019 Fletcher, P. D. I.; Howe, A. M.; Robinson, B. H. The Kinetics of Solubilisate Exchange between Water Droplets of a Water-in-oil Microemulsion. J. Chem. SOC.,Faraday Trans. 1 1987, 83 (4), 985-1006. Harris, M. T.;Byers, C. H. Effect of Solvent on the Homogeneous Precipitation of Titania by Titanium Ethoxide Hydrolysis.J.NonCryst. Solid 1988, 103,49-64. Hirai, T.;Imamura, E.; Matsumoto, T.; Kuboi, R.; Komasawa, I. Preparation of Metal Oxide Ultrafine Particles by Hydrolysis of Metal Alkoxide in Reverse Micelles (in Japanese). Kagaku Kogaku Ronbunshu 1992,18 (3),296-302. Kortan, A. R.; Hull, R.; Opila, R. L.; Bawendi, M. G.; Steigerwald, M. L.; Carroll, P. J.; Brus, L. E. Nucleation and Growth of CdSe on ZnS Quantum Crystallite Seeds, and Vice Versa, in Inverse Micelle Media. J. Am. Chem. SOC.1990, 112 (4),1327-1332. Kuboi, R.; Mori, Y.; Komasawa, I. Reverse Micelle Size Distribution and Mechanism of Protein Solubilization into Reverse Micelles (in Japanese). Kagaku Kogaku Ronbunshu 1990,16 (4),763-771. La Mer, V. K. Nucleation in Phase Transitions. Znd. Eng. Chem. 1952,44 (6),1270-1277. Lang, J.; Jada, A.; Malliaris, A. Structure and Dynamics of Waterin-Oil Droplets Stabilized by Sodium Bis(2-ethylhexyl) Sulfosuccinate. J. Phys. Chem. 1988,92 (7), 1946-1953. Muiloz, E.;G6mez-Herrera,C.; Graciani,M. M.; Moyh, M. L.; Shchez, F. Kinetics of the Oxidation of Iodide by Persulphate in AOT-
Oil-Water Microemulsions. J. Chem. SOC.,Faraday Trans. 1991, 87 (l), 129-132.
Ogihara, T.; Ikeda, M.; Kato, M.; Mizutani, N. Continuous Proceseing of Monodispersed Titania Powders. J. Am. Ceram. SOC. 1989, 72 (9),1598-1601. Oldfield, C. Exchange Concept in Water-in-oil Microemulsions: Consequences for ‘Slow’ Chemical Reactions. J. Chem. SOC., Faraday Trans. 1991,87 (16),2607-2612. Osseo-hare, K.; Arriagada, F. J. Synthesis of Nanoeize Particles in Reverse Microemulsions. Ceram. Trans. 1990,12, 3-16. Ravet, I.; Nagy, J. B.; Derouane, E. G. On the Mechanism of Formation of Colloidal Monodisperse Metal Boride Particles from Reversed Micelles Composed of CTAB-1-Hexanol-Water. Stud. Surf. Sci. Catal. 1987, 31, 506516. Towey, T. F.; Khan-Lodhi, A.; Robinson, B. H. Kinetics and Mechanism of Formation of Quantum-sized Cadmium Sulphide Particles in Water-Aerosol-OT-OilMicroemulsions. J. Chem. SOC., Faraday Tram. 1990,86 (22),3757-3762.
Received for review March 22, 1993 Accepted August 23, 1993. Abstract published in Advance ACS Abstracts, October 15, 1993. @
