- Email: [email protected]
Incorporation of CdS Nanoparticles Formed in Reverse Micelles into Mesoporous Silica
Journal of Colloid and Interface Science 235, 358–364 (2001) doi:10.1006/jcis.2000.7377, available online at http://www.idealibrary.com on
Incorporation of CdS Nanoparticles Formed in Reverse Micelles into Mesoporous Silica Takayuki Hirai,∗,1 Hironori Okubo,∗ and Isao Komasawa∗,† ∗ Department of Chemical Science and Engineering, Graduate School of Engineering Science, Osaka University, Toyonaka 560-8531, Japan; and †Research Center for Photoenergetics of Organic Materials, Osaka University, Toyonaka 560-8531, Japan Received July 10, 2000; accepted December 4, 2000
The incorporation of CdS nanoparticles, prepared in reverse micellar systems, into thiol-modified mesoporous silica, such as FM41 (functionalized MCM-41) and FM48 (functionalized MCM-48), has been investigated. The nanoparticles were immobilized in the mesopores via the incorporation of water droplets of the reverse micelles. A particle-sieving effect for FM41 having large (L-FM41, 3.8 nm) and medium (M-FM41, 3.6 nm) pore size was observed, in that the incorporation of the CdS nanoparticles was decreased with increasing particle size and with decreasing pore size of the FM41. Chemical vapor deposition treatment employed to narrow the mesopores of the CdS–FM41 enhanced the stability of CdS nanoparticles against heat treatment. The CdS–FM41 composites demonstrated photocatalytic activity for H2 generation from 2-propanol aqueous solution, the better photocatalytic activity being obtained with the larger pore size for CdS–L-FM41. °C 2001 Academic Press Key Words: mesoporous silica; CdS nanoparticles; size-selective incorporation; photocatalyst.
INTRODUCTION
There has been much recent interest in the utilization of the pores of mesoporous silica, such as MCM-41, as catalytic reaction fields. Metal oxide catalysts, such as Fe2 O3 (1) and TiO2 (2–4), are prepared mostly via the adsorption or impregnation of the metallic precursor, or by the chemical modification of the mesopores, followed by thermal oxidation. Xu and Langford have reported on the preparation of TiO2 -incorporated MCM-41 photocatalyst, formed by adding MCM-41 into TiO2 sol (5). However, the size selectivity of the mesopores for the nanoparticles was not described. In previous work (6), the size-selective incorporation of CdS nanoparticles into mesopores of mesoporous silica was investigated. CdS nanoparticles of ca. 3– 3.6 nm in diameter were prepared in reverse micelles, and incorporated into thiol-modified MCM-41 (FM41) of differing pore sizes, large (L-FM41), medium (M-FM41), and small (S-FM41). A particle-sieving effect for the FM41 was observed, in that the incorporation of the CdS nanoparticles was decreased both by 1 To whom correspondence should be addressed. E-mail: hirai@cheng. es.osaka-u.ac.jp. Fax: +81-6-6850-6273.
0021-9797/01 $35.00
C 2001 by Academic Press Copyright ° All rights of reproduction in any form reserved.
increasing the particle size and by decreasing the pore size for the FM41. In the present study, the incorporation of CdS nanoparticles into FM41 and FM48, thiol-modified MCM-48, and also into commercial mesoporous silica gel and on silica nanoparticles has been investigated in detail. The effect of chemical vapor deposition treatment on the stability of CdS nanoparticles in mesoporous silica, by narrowing the mesopores, was also studied. The resulting composite was utilized as photocatalyst for H2 generation from aqueous solution. EXPERIMENTAL
Chemicals. Cetyltrimethylammonium bromide (CTAB), 3-mercaptopropyltrimethoxysilane (TMPPS), sodium bis(2ethylhexyl)sulfosuccinate (Aerosol OT; AOT), and tetramethyl orthosilicate (TMOS) were supplied by Tokyo Chemical Industry, Ltd. (TCI). Fumed silica, used for M41S (MCM-41 and -48) preparation, was supplied by Sigma, and was also used as silica nanoparticles, abbreviated as SP7 hereafter (particle size, 7 nm; surface area, 390 m2 /g). Isooctane (2,2,4-trimethylpentane) was obtained from Ishizu Seiyaku Ltd. Tetramethylammonium hydroxide (TMAOH), sodium silicate, Cd(NO3 )2 · 4H2 O, Na2 S · 9H2 O, and all other chemicals were obtained from Wako Pure Chemical Industries, Ltd. All reagents were used without further purification. The preparation and filtration of the reverse micellar solutions were carried out as described in previous papers (7, 8). The mesoporous silica gel, abbreviated as SG6 hereafter, was supplied as spherical type CARiACT Q-6 (average pore size, 6 nm; surface area, 450 m2 /g; particle size, 28–150 µm) by Fuji Silysia Chemical Ltd. Preparation of the mesoporous silica materials. MCM-41 was prepared via a procedure similar to that reported by Cheng et al. (9). The mixture of CTAB, SiO2 (fumed silica), TMAOH, and H2 O (molar ratio, 0.27 : 1.0 : 1.19 : 40) was heated in an autoclave. The pore size of the MCM-41 was controlled by changing the reaction time (1–2 day) at 343–438 K and aging time in the mother liquor (0–7 days) at 423 K (10), as shown in Table 1. MCM-48 was prepared from the mixture of CTAB, SiO2 , Na2 O, ethanol, and H2 O (molar ratio, 1.0 : 1.4 : 0.35 : 5 : 140)
358
CdS NANOPARTICLES IN MESOPOROUS SILICA
TABLE 1 Preparation Conditions and Properties of Large (L-), Medium (M-), and Small (S-) Pore Size Thiol-Modified MCM-41 (FM41), FM-48, and FSG6 Specific Quantity Reaction Reaction Aging surface of –SH time temp time d100 area Pore size (mol/g (nm) of FM41) (day) (K) (day) (nm) (m2 /g) L-FM41 M-FM41 S-FM41 FM-48 FSG6
2 2 1 4
438 423 343 373
7 0 0
5.67 4.62 3.87
333 662 580 653 450
3.8 3.6 2.0 2.0 6.0
0.30 0.36 0.73 0.18
d100 values were determined by XRD analysis, specific surface area and pore sizes (mode diameter) were determined by N2 adsorption analysis and the Dollimore–Heal method, and the quantity of the thiol group was determined by iodometric titration.
via heating at 373 K for 4 days (11). The functionalization of the mesoporous silica M41S with surface thiol groups was carried out, as reported by Feng et al. (12) for the synthesis of 75% functionalized monolayers on mesoporous support samples. In this, 5 g of M41S was refluxed in 200 mL of toluene with 25 mL of TMPPS. The resulting thiol-modified M41S (functionalized M41S; FM41S) powders, approximately 17 µm in size, were denoted hereafter as L-FM41 (functionalized MCM-41), M-FM41, and S-FM41, representing large, medium, and small pore sizes, respectively, and FM48 (functionalized MCM-48). The d100 value and pore size for the FM41S samples were determined by XRD (Philips, PW-3050) and N2 adsorption (Bel Japan Inc., Belsorp28SP) analyses using the Dollimore–Heal method (13), respectively, with results that are detailed in Table 1. The thiol-modification of commercial silica, SG6 and SP7, was carried out in a similar way (14), and the resulting functionalized silica denoted as FSG6 and FSP7, respectively. Preparation of CdS nanoparticles in reverse micellar solution and the incorporation into functionalized silica. The CdS nanoparticles were prepared in a reverse micellar system consisting of 0.1 mol/L AOT, water, and isooctane. The molar ratio of water to AOT (=[H2 O]/[AOT]; water content) is denoted as Wo . An AOT–isooctane reverse micellar solution of the required Wo value and containing Cd(NO3 )2 was added rapidly to another micellar solution containing Na2 S and of the same Wo , and was stirred vigorously with a magnetic stirrer at 298 K in a glass vessel. The parameter y is defined to express the feed reactant compositions in the reverse micellar solutions, as y = [S2− ]/[Cd2+ ]. Three minutes following the formation of the CdS nanoparticles, 0.01–0.4 g of FM41S, FSG6, or FSP7 was added to 20 mL of the reverse micellar solution, stirred for 12 h, and separated by centrifugation. The percentage incorporation of the CdS nanoparticles was determined by subtracting the absorbance of CdS in the supernatant from that in the feed micellar solution. Analysis. The water content of the reverse micellar solution (Wo ) was determined by a Karl-Fischer moisture meter (Kyoto
359
Electronics MKS-1). The absorption spectra for the semiconductor nanoparticles in the micellar solutions were recorded on a diode-array UV–visible spectrophotometer (Hewlett-Packard 8452A). The diffuse reflectance spectra for the CdS nanoparticles incorporated in silica, CdS–FM41S, CdS–FSG6, and CdS–FSP7, were recorded, following dispersion of the CdS– silica composites in water, by a UV–vis spectrophotometer (Japan Spectroscopy V-550) equipped with an integrating sphere attachment (ISV-469). The diameter of the CdS nanoparticles was estimated on the basis of their absorption thresholds, as shown in the previous paper (7, 8), according to the Brus equation (15). TEM measurements was carried out by JEOL JM2010. The quantity of CdS incorporated into the functionalized silica was determined by using an inductively coupled argon plasma emission spectrometer (ICP-AES, Nippon JarrellAsh ICAP-575 Mark II), following dissolution of CdS into 1 mol/L HCl. Photoirradiation experiment. About 0.1 g of CdS-containing FM41 or FSG6 was dispersed in 20 mL of a 15 vol% 2-propanol aqueous solution. In this, 2-propanol was employed as a sacrificial electron donor for the photogenerated positive hole in the CdS nanoparticles. The solution in a test tube was purged with argon for 1 h, sealed with a septum, and photoirradiated with a 2-kW xenon lamp (Ushio UXL-2003D-O). Irradiation light of wavelength (λ) < 300 nm and light in the IR range were cut off by the Pyrex glass of the tube and by the water filter, respectively. The quantity of H2 formed in the gas phase of the tube was measured by gas chromatography (Shimadzu GC-14B), as described previously (7). RESULTS AND DISCUSSION
Incorporation of CdS Nanoparticles into Thiol-Modified Mesoporous Silica As shown in the previous paper (6), the size-selective incorporation of CdS nanoparticles into L- and M-FM41 has been achieved, whereas the nanoparticles are not incorporated into MCM-41 without thiol modification. Thus, the incorporation needs thiol groups and proceeds via chemical bonding between CdS and –SH. A typical TEM image for CdS-containing MFM41 (CdS–M-FM41) is shown in Fig. 1. The CdS nanoparticles are found in the mesopores of FM41, as shown by the arrows. Some CdS particles found on the FM41 surface are slightly larger than those in the mesopores. The diffuse reflectance spectra for the surface-modified silicas, measured following incorporation of the CdS nanoparticles, separation, and redispersion in water, demonstrate the characteristic absorption of size-quantized CdS, as shown in Fig. 2. From the shift in the onset wavelength from the bulk value (500 nm), the particle size of the CdS thus immobilized was thus estimated as 3.3, 3.2, 3.1, 3.1, 3.4, and 3.2 nm, respectively, for L-FM41, M-FM41, S-FM41, FM48, FSG6, and FSP7. As shown in Fig. 3, the percentage incorporation is affected remarkably by the Wo value of the reverse micellar solution,
360
HIRAI, OKUBO, AND KOMASAWA
FIG. 1. TEM image for CdS-containing M-FM41. CdS nanoparticles were prepared in reverse micellar solution (Wo = 2, [Cd2+ ] = 0.1 mmol/L, [S2− ] = 0.05 mmol/L, y = 0.5) and incorporated into 0.1 g of M-FM41.
where it is shown that the percentage incorporation decreases progressively with increasing Wo for L-FM41, while for MFM41 incorporation hardly occurs for Wo > 3. This is attributable to a particle-sieving effect for the FM41, since smaller CdS particles are formed in the reverse micellar system at lower Wo (7, 16). The estimated diameters for the CdS particles, after
FIG. 2. Diffuse reflectance spectra for CdS nanoparticles incorporated in (a) L-FM41, (b) M-FM41, (c) S-FM41, (d) FM48, (e) FSG6, and (f) FSP7, measured following dispersion in water. The CdS nanoparticles were prepared in reverse micellar solutions (Wo = 2, [Cd2+ ] = 1 mmol/L, [S2− ] = 0.5 mmol/L, y = 0.5) and incorporated into 0.04 g of functionalized silica.
FIG. 3. Percentage incorporation of CdS nanoparticles prepared in reverse micellar solutions ([Cd2+ ] = 0.1 mmol/L, [S2− ] = 0.05 mmol/L, y = 0.5) at various Wo values for (a) 0.04 g of L-FM41, M-FM41, and FSG6, and (b) 0.01 g of S-FM41, FM48, and FSP7. The percentage incorporation of the CdS nanoparticles was determined by subtracting the absorbance of CdS in the supernatant from that in the feed micellar solution.
12 h of stirring, are 3.0 (Wo = 2), 3.3 (Wo = 3), 3.6 (Wo = 4), and 3.7 (Wo = 5) nm, respectively, and the pore size of M-FM41 is 3.6 nm. The smaller CdS particles are thus incorporated more easily into the mesopores of the FM41. The incorporation is also more obvious for the case of L-FM41 having the larger pores (3.8 nm in diameter). For the case of FSG6, having pores of ca. 6 nm in diameter, a weak particle-sieving effect is observed at Wo < 4, but the effect is not obvious at Wo = 5. This is because the pore size is sufficiently large to incorporate CdS nanoparticles smaller than 3.73 nm in diameter, and the pore size distribution may be wider than that of FM41. In addition, relatively smaller values for percentage incorporation for FSG6 may indicate that the number of effective incorporation site, effected by –SH groups, in FSG6 is less than that in FM41, and this is consistent with the result that the number of –SH groups in FSG6 is less than that in FM41 (Table 1). In these functionalized silicas with greater mesopores, L-FM41, M-FM41, and FSG6, the immobilized CdS particles are likely to grow slightly compared to the particles in reverse micelles during the incorporation, as
CdS NANOPARTICLES IN MESOPOROUS SILICA
361
shown in Fig. 2, but are still smaller than the mesopore size and thus retain their quantum size effect. For the cases of S-FM41, FM48, and FSP7, the particlesieving effect is hardly observed, as shown in Fig. 3b. In the previous study (6), 0.04 g of S-FM41 was used, and more than 90% of the CdS particles, irrespective of the Wo value, were immobilized, although the size of the CdS particles prepared at Wo = 2–5 is greater than the pore size (2.0 nm). The percentage incorporation is much smaller in the present case, since the quantity of S-FM41 is decreased to 0.01 g. The N2 adsorption measurement showed that S-FM41 had macropores 20–40 nm in diameter. Thus, the CdS nanoparticles are probably immobilized in the macropores and also possibly on the surface, rather than in the mesopores, and the relatively rapid and sparse immobilization may suppress the particle growth. For FM48 (pore size, 2 nm) and FSP7, a similar tendency is observed, with the percentage incorporation being increased at increasing Wo . The flexibility and instability of reverse micelles are likely to be increased by increasing the Wo value, and thus the CdS nanoparticles are immobilized more easily. L- and M-FM41 are therefore suitable for nano-CdS incorporation into the mesopores of a silica matrix. Figure 4 shows the effect of the feed reactant composition in the reverse micellar solution, y = [S2− ]/[Cd2+ ], on the incorporation of CdS into M-FM41. The percentage incorporation for the CdS nanoparticles into M-FM41 is decreased by a decreasing value of y. This finding is inconsistent with the results for the surface modification of CdS with thiophenol in reverse micelles (8), where the redispersion ratio for the surface-modified CdS nanoparticles into pyridine is much lower at y > 0.9, because the thiophenol molecules need surface binding sites, at which excess Cd2+ ions are located. In the present case for FM41, fewer thiol groups, and thus a lesser excess of Cd2+ ions, are needed for the immobilization of CdS. A too great excess of Cd2+ ions may react predominantly with the surface –SH groups, and thus
reduce the immobilization of the CdS nanoparticles into FM41. Figure 5 shows the quantity of CdS incorporated into M-FM41. For y = 1, the quantity of incorporation increases with increasing CdS concentration in the reverse micellar solution and occurs probably owing to the increase in the contact of the CdS nanoparticles with FM41. In contrast, for y = 0.5, the quantity of CdS incorporated decreases with increasing CdS concentration, since the quantity of excess Cd2+ is increased under this condition. The time course variation in the percentage incorporation, following the addition of M-FM41 into the reverse micellar solution, is shown in Fig. 6. Approximately 500 min of agitation is needed to attain a constant value. Where M-FM41 is precontacted with a CdS-free reverse micellar solution (Wo = 2), prior to the CdS incorporation, the percentage incorporation is greatly decreased. This occurs because the FM41 mesopores are likely to be filled with water when contacted with the CdS-free reverse
FIG. 4. Effect of the feed reactant composition in the reverse micellar solution, y = [S2− ]/[Cd2+ ], on the percentage incorporation of CdS nanoparticles prepared in reverse micellar solution (Wo = 2, [CdS] = 0.05 mmol/L) into 0.03 g of M-FM41.
FIG. 6. Time course variation for the percentage incorporation of CdS nanoparticles prepared in reverse micellar solution (Wo = 2, [CdS] = 0.15 mmol/L, y = 1) into 0.1 g of M-FM41. The effect of precontact of MFM41 with CdS-free reverse micellar solution is shown by the open circles.
FIG. 5. Quantity of CdS incorporated into M-FM41 (0.03 g) as a function of CdS concentration in reverse micellar solution (Wo = 2, y = 0.5 or 1).
362
HIRAI, OKUBO, AND KOMASAWA
micelles, and are thus unable to incorporate the CdS nanoparticles. A mechanism for the incorporation of CdS into FM41 may therefore be suggested in which the water droplets in the micellar solution are adsorbed into the mesopores, and the CdS nanoparticles in the droplets are immobilized by the –SH groups of the FM41. The increase in the percentage incorporation after a 200-min mixing time in the case of the precontact treatment (open circles) may also suggest that an exchange of a CdS-free water droplet in the mesopore for a CdS-containing water droplet in the reverse micelles can occur during agitation. This proposed mechanism is supported by the results of repeated incorporation of CdS nanoparticles into FM41. In this, the FM41 powders were contacted with a CdS-containing reverse micellar solution, followed by washing with hexane and acetone and drying in vacuo, and then contacted again with a fresh reverse micellar solution containing CdS nanoparticles. As shown in Fig. 7, the cumulative quantity of CdS incorporated (open simbols) is increased by the repeated treatment, although the successive increase in the quantity for each treatment gradually decreases (closed symbols). This occurs because the incorporation sites (mesopores) filled by CdS-free water droplets are dehydrated by the washing and the drying of the FM41. The quantity of incorporated CdS reaches a saturation value of ca. 75 µmol of CdS/g of M-FM41, irrespective of the quantity of M-FM41 used. The maximum value is increased to ca. 0.1 mol of CdS/g of M-FM41 when the Wo value of reverse micellar solution is reduced to 1, thus indicating that the smaller particles, prepared at smaller Wo , can be incorporated more easily into the interior of the mesopores. CVD Treatment of the CdS–FM41 To Improve the Stability against Heat Treatment Heat treatment or calcination of the particles is sometimes effective to improve the catalytic or photocatalytic properties. Although MCM-41 itself is stable against heat treatment, CdS–
FIG. 8. Diffuse reflectance spectra for CdS nanoparticles prepared in reverse micellar solution (Wo = 2, [CdS] = 0.15 mmol/L, y = 1) when incorporated into (a) 0.15 g of M-FM41 and (b) 0.05 g of L-FM41. Also shown is the effect of CVD treatment for CdS–FM41 on the stability against heat treatment at 523 K for 30 min.
FM41 is not stable when heated at 523 K, as shown in Fig.8, showing the red shift of the onset wavelength. During heat treatment, the surface –SH groups of the FM41, immobilizing CdS nanoparticles, are probably decomposed and allow the aggregation of the CdS nanoparticles. One possible way to prevent this undesirable aggregation of the CdS nanoparticles is to narrow the pore size of the FM41, thus enclosing the nanoparticles within the mesopores. The CVD method employed in this study was that previously employed for zeolites (17, 18). TMOS and water were supplied alternately five times with N2 carrier gas at 423 K. N2 adsorption analysis showed that, following the CVD treatment, the resulting M-FM41 had mesopores of ca. 2.26 nm in diameter. Thus, by employing CVD treatment, as shown in Fig. 8, the aggregation of CdS nanoparticles, following heat treatment at 523 K, is effectively suppressed for both CdS–M-FM41 and CdS–L-FM41. Photocatalytic Properties of CdS–FM41
FIG. 7. Individual (closed symbols) and cumulative (open symbols) quantities of CdS incorporated by repeated use of M-FM41. CdS nanoparticles were prepared in reverse micellar solution (Wo = 2, [CdS] = 0.15 mmol/L, y = 1) and incorporated into 0.03, 0.06, or 0.09 g of M-FM41.
The photoirradiation of CdS nanoparticles sometimes causes undesirable aggregation of the nanoparticles, as for the case of CdS–polyurea composite (19, 20). In the present case of CdS–FM41, the relative ratio in the quantity of FM41 to CdS nanoparticles is found to be an important factor in stabilizing the nanoparticles. As shown in Fig. 9a, with 0.05 g of M-FM41 used for 20 mL of Wo = 2 reverse micellar solution containing
363
CdS NANOPARTICLES IN MESOPOROUS SILICA
The photocatalytic generation of H2 on CdS–FM41 and CdS– FSG6 in 15 vol% 2-propanol aqueous solution was demonstrated, under conditions similar to those used for the previous study with CdS– and ZnS–polyurea composites (19, 20), except that hexametaphosphate was not used in the present case. The quantities of H2 formed as a function of photoirradiation time are shown in Fig. 10. For this experiment, CdS–L-FM41, CdS–M-FM41, and CdS–FSG6 had the same absorption onset wavelength (ca. 420 nm) and therefore the same CdS size, and were stable against photoirradiation. A continuous generation of H2 is observed, occurring probably via the reduction of water with 2-propanol acting as the sacrificial agent. A much greater photocatalytic property is obtained for the greater pore sizes of L-FM41 and FSG6, which probably results because the adsorption of water and the desorption of H2 occur more easily in the greater mesopores. The pore size dependency is also consistent with previous reports concerning the photocatalytic CO2 reduction on TiO2 –MCM-41 and MCM-48 (4), and the photoionization of porphyrins in MCM-41 (21).
CONCLUSION FIG. 9. Diffuse reflectance spectra for CdS nanoparticles prepared in reverse micellar solution (Wo = 2, [CdS] = 0.15 mmol/L, y = 1) when incorporated into (a) 0.05 and (b) 0.15 g of M-FM41. Also shown is the effect of the quantity of M-FM41 on the stability against photoirradiation for 24 or 48 h.
0.15 mmol/L CdS, the incorporated CdS nanoparticles are not stable against photoirradiation by the Xe lamp and a red shift of the onset wavelength shows that the nanoparticles are aggregated within the FM41. As shown in Fig. 9b, where the quantity of MFM41 is increased to 0.15 g, the nanoparticle stability is greatly enhanced, probably owing to the relative sparse incorporation of the CdS nanoparticles.
The incorporation of CdS nanoparticles, prepared in reverse micellar systems, into thiol-modified mesoporous silica has been investigated, with the following results. 1. The incorporation of the CdS nanoparticles is decreased both by an increasing particle size and by a decreasing pore size of the FM41 (functionalized MCM-41). Thus, a particlesieving effect is observed, especially for the L-FM41 and MFM41 cases, having a relatively large and medium pore size, respectively. A rather weak particle-sieving effect is obtained for the thiol-modified commercial mesoporous silica gel, FSG6. 2. CVD treatment for CdS–FM41 employed to narrow the mesopore size is effective in enhancing the stability of the CdS nanoparticles against heat treatment at 523 K. 3. The stability of CdS nanoparticles in FM41 against photoirradiation is enhanced by the sparse incorporation of CdS, obtained by using a greater quantity of FM41 in the incorporation procedure. The resulting CdS–FM41 composites demonstrate photocatalytic activity for H2 generation from 2-propanol aqueous solution. Higher photocatalytic activity is obtained for the larger pore sizes with CdS–L-FM41 and CdS–FSG6.
ACKNOWLEDGMENTS
FIG. 10. Quantities of H2 formed from 15 vol% 2-propanol aqueous solution by photoirradiation for dispersed CdS–L-FM41, CdS–M-FM41, and CdS–FSG6. The CdS nanoparticles were prepared in reverse micellar solution (Wo = 2, [CdS] = 0.15 mmol/L, y = 1) and incorporated into L-FM41 (0.05 g), M-FM41 (0.15 g), and FSG6 (0.4 g).
The authors acknowledge Professor Mituji Hirata and Professor Seiji Takeda (Osaka University) for their help in the TEM measurement, and Professor Hajime Tamon (Kyoto University) for his help in the measurement of the pore size of FM41S. The authors are also grateful to the Division of Chemical Engineering, Department of Chemical Science and Engineering, Osaka University, for the scientific support for the Gas-Hydrate Analyzing System (GHAS) and Lend-Lease Laboratory System, and to the financial support by Grants-in-Aid for Scientific Research (Nos.10450286 and 12450311) from the Ministry of Education, Science, Sports and Culture, Japan.
364
HIRAI, OKUBO, AND KOMASAWA
REFERENCES 1. Abe, T., Tachibana, Y., Uematsu, T., and Iwamoto, M., J. Chem. Soc., Chem. Commun. 1617 (1995). 2. Maschmeyer, T., Rey, F., Sankar, G., and Thomas, J. M., Nature 378, 159 (1995). 3. Aronson, B. J., Blanford, C. F., and Stein, A., Chem. Mater. 9, 2842 (1997). 4. Zhang, S. G., Fujii, Y., Yamashita, H., Koyano, K., Tatsumi, T., and Anpo, M., Chem. Lett. 659 (1997). 5. Xu, Y., and Langford, C. H., J. Phys. Chem. B 101, 3115 (1997). 6. Hirai, T., Okubo, H., and Komasawa, I., J. Phys. Chem. B 103, 4228 (1999). 7. Hirai, T., Shiojiri, S., and Komasawa, I., J. Chem. Eng. Jpn. 27, 590 (1994). 8. Shiojiri, S., Hirai, T., and Komasawa, I., J. Chem. Eng. Jpn. 30, 86 (1997). 9. Cheng, C.-F., Zhou, W., Park, D. H., Klinowski, J., Hargreaves, M., and Gladden, L. F., J. Chem. Soc., Faraday Trans. 93, 359 (1997).
10. Khushalani, D., Kuperman, A., Ozin, G. A., Tanaka, K., Graces, J., Olken, M. M., and Coombs, N., Adv. Mater. 7, 842 (1995). 11. Kim, J. M., Kim, S. K., and Ryoo, R., Chem. Commun. 259 (1998). 12. Feng, X., Fryxell, G. E., Wang, L.-Q., Kim, A. Y., Liu, J., and Kemner, K. M., Science 276, 923 (1997). 13. Dollimore, D., and Heal, G. R., J. Appl. Chem. 14, 109 (1964). 14. Vieira, E. F. S., Simoni, J., and Airoldi, C., J. Mater. Chem. 7, 2249 (1997). 15. Brus, L. E., J. Chem. Phys. 80, 4403 (1984). 16. Pileni, M. P., Langmuir 13, 3266 (1997). 17. Niwa, M., and Murakami, Y., J. Phys. Chem. Solids 50, 487 (1989). 18. Yamazaki, K., Niwa, M., and Murakami, Y., Kagaku Kogaku Ronbunshu 16, 564 (1990). 19. Shiojiri, S., Hirai, T., and Komasawa, I., Chem. Commun. 1439 (1998). 20. Hirai, T., Watanabe, T., and Komasawa, I., J. Phys. Chem. B 103, 10120 (1999). 21. Sung-Suh, H. M., Luan, Z., and Kevan, L., J. Phys. Chem. B 101, 10455 (1997).
Incorporation of CdS Nanoparticles Formed in Reverse Micelles into Mesoporous Silica Takayuki Hirai,∗,1 Hironori Okubo,∗ and Isao Komasawa∗,† ∗ Department of Chemical Science and Engineering, Graduate School of Engineering Science, Osaka University, Toyonaka 560-8531, Japan; and †Research Center for Photoenergetics of Organic Materials, Osaka University, Toyonaka 560-8531, Japan Received July 10, 2000; accepted December 4, 2000
The incorporation of CdS nanoparticles, prepared in reverse micellar systems, into thiol-modified mesoporous silica, such as FM41 (functionalized MCM-41) and FM48 (functionalized MCM-48), has been investigated. The nanoparticles were immobilized in the mesopores via the incorporation of water droplets of the reverse micelles. A particle-sieving effect for FM41 having large (L-FM41, 3.8 nm) and medium (M-FM41, 3.6 nm) pore size was observed, in that the incorporation of the CdS nanoparticles was decreased with increasing particle size and with decreasing pore size of the FM41. Chemical vapor deposition treatment employed to narrow the mesopores of the CdS–FM41 enhanced the stability of CdS nanoparticles against heat treatment. The CdS–FM41 composites demonstrated photocatalytic activity for H2 generation from 2-propanol aqueous solution, the better photocatalytic activity being obtained with the larger pore size for CdS–L-FM41. °C 2001 Academic Press Key Words: mesoporous silica; CdS nanoparticles; size-selective incorporation; photocatalyst.
INTRODUCTION
There has been much recent interest in the utilization of the pores of mesoporous silica, such as MCM-41, as catalytic reaction fields. Metal oxide catalysts, such as Fe2 O3 (1) and TiO2 (2–4), are prepared mostly via the adsorption or impregnation of the metallic precursor, or by the chemical modification of the mesopores, followed by thermal oxidation. Xu and Langford have reported on the preparation of TiO2 -incorporated MCM-41 photocatalyst, formed by adding MCM-41 into TiO2 sol (5). However, the size selectivity of the mesopores for the nanoparticles was not described. In previous work (6), the size-selective incorporation of CdS nanoparticles into mesopores of mesoporous silica was investigated. CdS nanoparticles of ca. 3– 3.6 nm in diameter were prepared in reverse micelles, and incorporated into thiol-modified MCM-41 (FM41) of differing pore sizes, large (L-FM41), medium (M-FM41), and small (S-FM41). A particle-sieving effect for the FM41 was observed, in that the incorporation of the CdS nanoparticles was decreased both by 1 To whom correspondence should be addressed. E-mail: hirai@cheng. es.osaka-u.ac.jp. Fax: +81-6-6850-6273.
0021-9797/01 $35.00
C 2001 by Academic Press Copyright ° All rights of reproduction in any form reserved.
increasing the particle size and by decreasing the pore size for the FM41. In the present study, the incorporation of CdS nanoparticles into FM41 and FM48, thiol-modified MCM-48, and also into commercial mesoporous silica gel and on silica nanoparticles has been investigated in detail. The effect of chemical vapor deposition treatment on the stability of CdS nanoparticles in mesoporous silica, by narrowing the mesopores, was also studied. The resulting composite was utilized as photocatalyst for H2 generation from aqueous solution. EXPERIMENTAL
Chemicals. Cetyltrimethylammonium bromide (CTAB), 3-mercaptopropyltrimethoxysilane (TMPPS), sodium bis(2ethylhexyl)sulfosuccinate (Aerosol OT; AOT), and tetramethyl orthosilicate (TMOS) were supplied by Tokyo Chemical Industry, Ltd. (TCI). Fumed silica, used for M41S (MCM-41 and -48) preparation, was supplied by Sigma, and was also used as silica nanoparticles, abbreviated as SP7 hereafter (particle size, 7 nm; surface area, 390 m2 /g). Isooctane (2,2,4-trimethylpentane) was obtained from Ishizu Seiyaku Ltd. Tetramethylammonium hydroxide (TMAOH), sodium silicate, Cd(NO3 )2 · 4H2 O, Na2 S · 9H2 O, and all other chemicals were obtained from Wako Pure Chemical Industries, Ltd. All reagents were used without further purification. The preparation and filtration of the reverse micellar solutions were carried out as described in previous papers (7, 8). The mesoporous silica gel, abbreviated as SG6 hereafter, was supplied as spherical type CARiACT Q-6 (average pore size, 6 nm; surface area, 450 m2 /g; particle size, 28–150 µm) by Fuji Silysia Chemical Ltd. Preparation of the mesoporous silica materials. MCM-41 was prepared via a procedure similar to that reported by Cheng et al. (9). The mixture of CTAB, SiO2 (fumed silica), TMAOH, and H2 O (molar ratio, 0.27 : 1.0 : 1.19 : 40) was heated in an autoclave. The pore size of the MCM-41 was controlled by changing the reaction time (1–2 day) at 343–438 K and aging time in the mother liquor (0–7 days) at 423 K (10), as shown in Table 1. MCM-48 was prepared from the mixture of CTAB, SiO2 , Na2 O, ethanol, and H2 O (molar ratio, 1.0 : 1.4 : 0.35 : 5 : 140)
358
CdS NANOPARTICLES IN MESOPOROUS SILICA
TABLE 1 Preparation Conditions and Properties of Large (L-), Medium (M-), and Small (S-) Pore Size Thiol-Modified MCM-41 (FM41), FM-48, and FSG6 Specific Quantity Reaction Reaction Aging surface of –SH time temp time d100 area Pore size (mol/g (nm) of FM41) (day) (K) (day) (nm) (m2 /g) L-FM41 M-FM41 S-FM41 FM-48 FSG6
2 2 1 4
438 423 343 373
7 0 0
5.67 4.62 3.87
333 662 580 653 450
3.8 3.6 2.0 2.0 6.0
0.30 0.36 0.73 0.18
d100 values were determined by XRD analysis, specific surface area and pore sizes (mode diameter) were determined by N2 adsorption analysis and the Dollimore–Heal method, and the quantity of the thiol group was determined by iodometric titration.
via heating at 373 K for 4 days (11). The functionalization of the mesoporous silica M41S with surface thiol groups was carried out, as reported by Feng et al. (12) for the synthesis of 75% functionalized monolayers on mesoporous support samples. In this, 5 g of M41S was refluxed in 200 mL of toluene with 25 mL of TMPPS. The resulting thiol-modified M41S (functionalized M41S; FM41S) powders, approximately 17 µm in size, were denoted hereafter as L-FM41 (functionalized MCM-41), M-FM41, and S-FM41, representing large, medium, and small pore sizes, respectively, and FM48 (functionalized MCM-48). The d100 value and pore size for the FM41S samples were determined by XRD (Philips, PW-3050) and N2 adsorption (Bel Japan Inc., Belsorp28SP) analyses using the Dollimore–Heal method (13), respectively, with results that are detailed in Table 1. The thiol-modification of commercial silica, SG6 and SP7, was carried out in a similar way (14), and the resulting functionalized silica denoted as FSG6 and FSP7, respectively. Preparation of CdS nanoparticles in reverse micellar solution and the incorporation into functionalized silica. The CdS nanoparticles were prepared in a reverse micellar system consisting of 0.1 mol/L AOT, water, and isooctane. The molar ratio of water to AOT (=[H2 O]/[AOT]; water content) is denoted as Wo . An AOT–isooctane reverse micellar solution of the required Wo value and containing Cd(NO3 )2 was added rapidly to another micellar solution containing Na2 S and of the same Wo , and was stirred vigorously with a magnetic stirrer at 298 K in a glass vessel. The parameter y is defined to express the feed reactant compositions in the reverse micellar solutions, as y = [S2− ]/[Cd2+ ]. Three minutes following the formation of the CdS nanoparticles, 0.01–0.4 g of FM41S, FSG6, or FSP7 was added to 20 mL of the reverse micellar solution, stirred for 12 h, and separated by centrifugation. The percentage incorporation of the CdS nanoparticles was determined by subtracting the absorbance of CdS in the supernatant from that in the feed micellar solution. Analysis. The water content of the reverse micellar solution (Wo ) was determined by a Karl-Fischer moisture meter (Kyoto
359
Electronics MKS-1). The absorption spectra for the semiconductor nanoparticles in the micellar solutions were recorded on a diode-array UV–visible spectrophotometer (Hewlett-Packard 8452A). The diffuse reflectance spectra for the CdS nanoparticles incorporated in silica, CdS–FM41S, CdS–FSG6, and CdS–FSP7, were recorded, following dispersion of the CdS– silica composites in water, by a UV–vis spectrophotometer (Japan Spectroscopy V-550) equipped with an integrating sphere attachment (ISV-469). The diameter of the CdS nanoparticles was estimated on the basis of their absorption thresholds, as shown in the previous paper (7, 8), according to the Brus equation (15). TEM measurements was carried out by JEOL JM2010. The quantity of CdS incorporated into the functionalized silica was determined by using an inductively coupled argon plasma emission spectrometer (ICP-AES, Nippon JarrellAsh ICAP-575 Mark II), following dissolution of CdS into 1 mol/L HCl. Photoirradiation experiment. About 0.1 g of CdS-containing FM41 or FSG6 was dispersed in 20 mL of a 15 vol% 2-propanol aqueous solution. In this, 2-propanol was employed as a sacrificial electron donor for the photogenerated positive hole in the CdS nanoparticles. The solution in a test tube was purged with argon for 1 h, sealed with a septum, and photoirradiated with a 2-kW xenon lamp (Ushio UXL-2003D-O). Irradiation light of wavelength (λ) < 300 nm and light in the IR range were cut off by the Pyrex glass of the tube and by the water filter, respectively. The quantity of H2 formed in the gas phase of the tube was measured by gas chromatography (Shimadzu GC-14B), as described previously (7). RESULTS AND DISCUSSION
Incorporation of CdS Nanoparticles into Thiol-Modified Mesoporous Silica As shown in the previous paper (6), the size-selective incorporation of CdS nanoparticles into L- and M-FM41 has been achieved, whereas the nanoparticles are not incorporated into MCM-41 without thiol modification. Thus, the incorporation needs thiol groups and proceeds via chemical bonding between CdS and –SH. A typical TEM image for CdS-containing MFM41 (CdS–M-FM41) is shown in Fig. 1. The CdS nanoparticles are found in the mesopores of FM41, as shown by the arrows. Some CdS particles found on the FM41 surface are slightly larger than those in the mesopores. The diffuse reflectance spectra for the surface-modified silicas, measured following incorporation of the CdS nanoparticles, separation, and redispersion in water, demonstrate the characteristic absorption of size-quantized CdS, as shown in Fig. 2. From the shift in the onset wavelength from the bulk value (500 nm), the particle size of the CdS thus immobilized was thus estimated as 3.3, 3.2, 3.1, 3.1, 3.4, and 3.2 nm, respectively, for L-FM41, M-FM41, S-FM41, FM48, FSG6, and FSP7. As shown in Fig. 3, the percentage incorporation is affected remarkably by the Wo value of the reverse micellar solution,
360
HIRAI, OKUBO, AND KOMASAWA
FIG. 1. TEM image for CdS-containing M-FM41. CdS nanoparticles were prepared in reverse micellar solution (Wo = 2, [Cd2+ ] = 0.1 mmol/L, [S2− ] = 0.05 mmol/L, y = 0.5) and incorporated into 0.1 g of M-FM41.
where it is shown that the percentage incorporation decreases progressively with increasing Wo for L-FM41, while for MFM41 incorporation hardly occurs for Wo > 3. This is attributable to a particle-sieving effect for the FM41, since smaller CdS particles are formed in the reverse micellar system at lower Wo (7, 16). The estimated diameters for the CdS particles, after
FIG. 2. Diffuse reflectance spectra for CdS nanoparticles incorporated in (a) L-FM41, (b) M-FM41, (c) S-FM41, (d) FM48, (e) FSG6, and (f) FSP7, measured following dispersion in water. The CdS nanoparticles were prepared in reverse micellar solutions (Wo = 2, [Cd2+ ] = 1 mmol/L, [S2− ] = 0.5 mmol/L, y = 0.5) and incorporated into 0.04 g of functionalized silica.
FIG. 3. Percentage incorporation of CdS nanoparticles prepared in reverse micellar solutions ([Cd2+ ] = 0.1 mmol/L, [S2− ] = 0.05 mmol/L, y = 0.5) at various Wo values for (a) 0.04 g of L-FM41, M-FM41, and FSG6, and (b) 0.01 g of S-FM41, FM48, and FSP7. The percentage incorporation of the CdS nanoparticles was determined by subtracting the absorbance of CdS in the supernatant from that in the feed micellar solution.
12 h of stirring, are 3.0 (Wo = 2), 3.3 (Wo = 3), 3.6 (Wo = 4), and 3.7 (Wo = 5) nm, respectively, and the pore size of M-FM41 is 3.6 nm. The smaller CdS particles are thus incorporated more easily into the mesopores of the FM41. The incorporation is also more obvious for the case of L-FM41 having the larger pores (3.8 nm in diameter). For the case of FSG6, having pores of ca. 6 nm in diameter, a weak particle-sieving effect is observed at Wo < 4, but the effect is not obvious at Wo = 5. This is because the pore size is sufficiently large to incorporate CdS nanoparticles smaller than 3.73 nm in diameter, and the pore size distribution may be wider than that of FM41. In addition, relatively smaller values for percentage incorporation for FSG6 may indicate that the number of effective incorporation site, effected by –SH groups, in FSG6 is less than that in FM41, and this is consistent with the result that the number of –SH groups in FSG6 is less than that in FM41 (Table 1). In these functionalized silicas with greater mesopores, L-FM41, M-FM41, and FSG6, the immobilized CdS particles are likely to grow slightly compared to the particles in reverse micelles during the incorporation, as
CdS NANOPARTICLES IN MESOPOROUS SILICA
361
shown in Fig. 2, but are still smaller than the mesopore size and thus retain their quantum size effect. For the cases of S-FM41, FM48, and FSP7, the particlesieving effect is hardly observed, as shown in Fig. 3b. In the previous study (6), 0.04 g of S-FM41 was used, and more than 90% of the CdS particles, irrespective of the Wo value, were immobilized, although the size of the CdS particles prepared at Wo = 2–5 is greater than the pore size (2.0 nm). The percentage incorporation is much smaller in the present case, since the quantity of S-FM41 is decreased to 0.01 g. The N2 adsorption measurement showed that S-FM41 had macropores 20–40 nm in diameter. Thus, the CdS nanoparticles are probably immobilized in the macropores and also possibly on the surface, rather than in the mesopores, and the relatively rapid and sparse immobilization may suppress the particle growth. For FM48 (pore size, 2 nm) and FSP7, a similar tendency is observed, with the percentage incorporation being increased at increasing Wo . The flexibility and instability of reverse micelles are likely to be increased by increasing the Wo value, and thus the CdS nanoparticles are immobilized more easily. L- and M-FM41 are therefore suitable for nano-CdS incorporation into the mesopores of a silica matrix. Figure 4 shows the effect of the feed reactant composition in the reverse micellar solution, y = [S2− ]/[Cd2+ ], on the incorporation of CdS into M-FM41. The percentage incorporation for the CdS nanoparticles into M-FM41 is decreased by a decreasing value of y. This finding is inconsistent with the results for the surface modification of CdS with thiophenol in reverse micelles (8), where the redispersion ratio for the surface-modified CdS nanoparticles into pyridine is much lower at y > 0.9, because the thiophenol molecules need surface binding sites, at which excess Cd2+ ions are located. In the present case for FM41, fewer thiol groups, and thus a lesser excess of Cd2+ ions, are needed for the immobilization of CdS. A too great excess of Cd2+ ions may react predominantly with the surface –SH groups, and thus
reduce the immobilization of the CdS nanoparticles into FM41. Figure 5 shows the quantity of CdS incorporated into M-FM41. For y = 1, the quantity of incorporation increases with increasing CdS concentration in the reverse micellar solution and occurs probably owing to the increase in the contact of the CdS nanoparticles with FM41. In contrast, for y = 0.5, the quantity of CdS incorporated decreases with increasing CdS concentration, since the quantity of excess Cd2+ is increased under this condition. The time course variation in the percentage incorporation, following the addition of M-FM41 into the reverse micellar solution, is shown in Fig. 6. Approximately 500 min of agitation is needed to attain a constant value. Where M-FM41 is precontacted with a CdS-free reverse micellar solution (Wo = 2), prior to the CdS incorporation, the percentage incorporation is greatly decreased. This occurs because the FM41 mesopores are likely to be filled with water when contacted with the CdS-free reverse
FIG. 4. Effect of the feed reactant composition in the reverse micellar solution, y = [S2− ]/[Cd2+ ], on the percentage incorporation of CdS nanoparticles prepared in reverse micellar solution (Wo = 2, [CdS] = 0.05 mmol/L) into 0.03 g of M-FM41.
FIG. 6. Time course variation for the percentage incorporation of CdS nanoparticles prepared in reverse micellar solution (Wo = 2, [CdS] = 0.15 mmol/L, y = 1) into 0.1 g of M-FM41. The effect of precontact of MFM41 with CdS-free reverse micellar solution is shown by the open circles.
FIG. 5. Quantity of CdS incorporated into M-FM41 (0.03 g) as a function of CdS concentration in reverse micellar solution (Wo = 2, y = 0.5 or 1).
362
HIRAI, OKUBO, AND KOMASAWA
micelles, and are thus unable to incorporate the CdS nanoparticles. A mechanism for the incorporation of CdS into FM41 may therefore be suggested in which the water droplets in the micellar solution are adsorbed into the mesopores, and the CdS nanoparticles in the droplets are immobilized by the –SH groups of the FM41. The increase in the percentage incorporation after a 200-min mixing time in the case of the precontact treatment (open circles) may also suggest that an exchange of a CdS-free water droplet in the mesopore for a CdS-containing water droplet in the reverse micelles can occur during agitation. This proposed mechanism is supported by the results of repeated incorporation of CdS nanoparticles into FM41. In this, the FM41 powders were contacted with a CdS-containing reverse micellar solution, followed by washing with hexane and acetone and drying in vacuo, and then contacted again with a fresh reverse micellar solution containing CdS nanoparticles. As shown in Fig. 7, the cumulative quantity of CdS incorporated (open simbols) is increased by the repeated treatment, although the successive increase in the quantity for each treatment gradually decreases (closed symbols). This occurs because the incorporation sites (mesopores) filled by CdS-free water droplets are dehydrated by the washing and the drying of the FM41. The quantity of incorporated CdS reaches a saturation value of ca. 75 µmol of CdS/g of M-FM41, irrespective of the quantity of M-FM41 used. The maximum value is increased to ca. 0.1 mol of CdS/g of M-FM41 when the Wo value of reverse micellar solution is reduced to 1, thus indicating that the smaller particles, prepared at smaller Wo , can be incorporated more easily into the interior of the mesopores. CVD Treatment of the CdS–FM41 To Improve the Stability against Heat Treatment Heat treatment or calcination of the particles is sometimes effective to improve the catalytic or photocatalytic properties. Although MCM-41 itself is stable against heat treatment, CdS–
FIG. 8. Diffuse reflectance spectra for CdS nanoparticles prepared in reverse micellar solution (Wo = 2, [CdS] = 0.15 mmol/L, y = 1) when incorporated into (a) 0.15 g of M-FM41 and (b) 0.05 g of L-FM41. Also shown is the effect of CVD treatment for CdS–FM41 on the stability against heat treatment at 523 K for 30 min.
FM41 is not stable when heated at 523 K, as shown in Fig.8, showing the red shift of the onset wavelength. During heat treatment, the surface –SH groups of the FM41, immobilizing CdS nanoparticles, are probably decomposed and allow the aggregation of the CdS nanoparticles. One possible way to prevent this undesirable aggregation of the CdS nanoparticles is to narrow the pore size of the FM41, thus enclosing the nanoparticles within the mesopores. The CVD method employed in this study was that previously employed for zeolites (17, 18). TMOS and water were supplied alternately five times with N2 carrier gas at 423 K. N2 adsorption analysis showed that, following the CVD treatment, the resulting M-FM41 had mesopores of ca. 2.26 nm in diameter. Thus, by employing CVD treatment, as shown in Fig. 8, the aggregation of CdS nanoparticles, following heat treatment at 523 K, is effectively suppressed for both CdS–M-FM41 and CdS–L-FM41. Photocatalytic Properties of CdS–FM41
FIG. 7. Individual (closed symbols) and cumulative (open symbols) quantities of CdS incorporated by repeated use of M-FM41. CdS nanoparticles were prepared in reverse micellar solution (Wo = 2, [CdS] = 0.15 mmol/L, y = 1) and incorporated into 0.03, 0.06, or 0.09 g of M-FM41.
The photoirradiation of CdS nanoparticles sometimes causes undesirable aggregation of the nanoparticles, as for the case of CdS–polyurea composite (19, 20). In the present case of CdS–FM41, the relative ratio in the quantity of FM41 to CdS nanoparticles is found to be an important factor in stabilizing the nanoparticles. As shown in Fig. 9a, with 0.05 g of M-FM41 used for 20 mL of Wo = 2 reverse micellar solution containing
363
CdS NANOPARTICLES IN MESOPOROUS SILICA
The photocatalytic generation of H2 on CdS–FM41 and CdS– FSG6 in 15 vol% 2-propanol aqueous solution was demonstrated, under conditions similar to those used for the previous study with CdS– and ZnS–polyurea composites (19, 20), except that hexametaphosphate was not used in the present case. The quantities of H2 formed as a function of photoirradiation time are shown in Fig. 10. For this experiment, CdS–L-FM41, CdS–M-FM41, and CdS–FSG6 had the same absorption onset wavelength (ca. 420 nm) and therefore the same CdS size, and were stable against photoirradiation. A continuous generation of H2 is observed, occurring probably via the reduction of water with 2-propanol acting as the sacrificial agent. A much greater photocatalytic property is obtained for the greater pore sizes of L-FM41 and FSG6, which probably results because the adsorption of water and the desorption of H2 occur more easily in the greater mesopores. The pore size dependency is also consistent with previous reports concerning the photocatalytic CO2 reduction on TiO2 –MCM-41 and MCM-48 (4), and the photoionization of porphyrins in MCM-41 (21).
CONCLUSION FIG. 9. Diffuse reflectance spectra for CdS nanoparticles prepared in reverse micellar solution (Wo = 2, [CdS] = 0.15 mmol/L, y = 1) when incorporated into (a) 0.05 and (b) 0.15 g of M-FM41. Also shown is the effect of the quantity of M-FM41 on the stability against photoirradiation for 24 or 48 h.
0.15 mmol/L CdS, the incorporated CdS nanoparticles are not stable against photoirradiation by the Xe lamp and a red shift of the onset wavelength shows that the nanoparticles are aggregated within the FM41. As shown in Fig. 9b, where the quantity of MFM41 is increased to 0.15 g, the nanoparticle stability is greatly enhanced, probably owing to the relative sparse incorporation of the CdS nanoparticles.
The incorporation of CdS nanoparticles, prepared in reverse micellar systems, into thiol-modified mesoporous silica has been investigated, with the following results. 1. The incorporation of the CdS nanoparticles is decreased both by an increasing particle size and by a decreasing pore size of the FM41 (functionalized MCM-41). Thus, a particlesieving effect is observed, especially for the L-FM41 and MFM41 cases, having a relatively large and medium pore size, respectively. A rather weak particle-sieving effect is obtained for the thiol-modified commercial mesoporous silica gel, FSG6. 2. CVD treatment for CdS–FM41 employed to narrow the mesopore size is effective in enhancing the stability of the CdS nanoparticles against heat treatment at 523 K. 3. The stability of CdS nanoparticles in FM41 against photoirradiation is enhanced by the sparse incorporation of CdS, obtained by using a greater quantity of FM41 in the incorporation procedure. The resulting CdS–FM41 composites demonstrate photocatalytic activity for H2 generation from 2-propanol aqueous solution. Higher photocatalytic activity is obtained for the larger pore sizes with CdS–L-FM41 and CdS–FSG6.
ACKNOWLEDGMENTS
FIG. 10. Quantities of H2 formed from 15 vol% 2-propanol aqueous solution by photoirradiation for dispersed CdS–L-FM41, CdS–M-FM41, and CdS–FSG6. The CdS nanoparticles were prepared in reverse micellar solution (Wo = 2, [CdS] = 0.15 mmol/L, y = 1) and incorporated into L-FM41 (0.05 g), M-FM41 (0.15 g), and FSG6 (0.4 g).
The authors acknowledge Professor Mituji Hirata and Professor Seiji Takeda (Osaka University) for their help in the TEM measurement, and Professor Hajime Tamon (Kyoto University) for his help in the measurement of the pore size of FM41S. The authors are also grateful to the Division of Chemical Engineering, Department of Chemical Science and Engineering, Osaka University, for the scientific support for the Gas-Hydrate Analyzing System (GHAS) and Lend-Lease Laboratory System, and to the financial support by Grants-in-Aid for Scientific Research (Nos.10450286 and 12450311) from the Ministry of Education, Science, Sports and Culture, Japan.
364
HIRAI, OKUBO, AND KOMASAWA
REFERENCES 1. Abe, T., Tachibana, Y., Uematsu, T., and Iwamoto, M., J. Chem. Soc., Chem. Commun. 1617 (1995). 2. Maschmeyer, T., Rey, F., Sankar, G., and Thomas, J. M., Nature 378, 159 (1995). 3. Aronson, B. J., Blanford, C. F., and Stein, A., Chem. Mater. 9, 2842 (1997). 4. Zhang, S. G., Fujii, Y., Yamashita, H., Koyano, K., Tatsumi, T., and Anpo, M., Chem. Lett. 659 (1997). 5. Xu, Y., and Langford, C. H., J. Phys. Chem. B 101, 3115 (1997). 6. Hirai, T., Okubo, H., and Komasawa, I., J. Phys. Chem. B 103, 4228 (1999). 7. Hirai, T., Shiojiri, S., and Komasawa, I., J. Chem. Eng. Jpn. 27, 590 (1994). 8. Shiojiri, S., Hirai, T., and Komasawa, I., J. Chem. Eng. Jpn. 30, 86 (1997). 9. Cheng, C.-F., Zhou, W., Park, D. H., Klinowski, J., Hargreaves, M., and Gladden, L. F., J. Chem. Soc., Faraday Trans. 93, 359 (1997).
10. Khushalani, D., Kuperman, A., Ozin, G. A., Tanaka, K., Graces, J., Olken, M. M., and Coombs, N., Adv. Mater. 7, 842 (1995). 11. Kim, J. M., Kim, S. K., and Ryoo, R., Chem. Commun. 259 (1998). 12. Feng, X., Fryxell, G. E., Wang, L.-Q., Kim, A. Y., Liu, J., and Kemner, K. M., Science 276, 923 (1997). 13. Dollimore, D., and Heal, G. R., J. Appl. Chem. 14, 109 (1964). 14. Vieira, E. F. S., Simoni, J., and Airoldi, C., J. Mater. Chem. 7, 2249 (1997). 15. Brus, L. E., J. Chem. Phys. 80, 4403 (1984). 16. Pileni, M. P., Langmuir 13, 3266 (1997). 17. Niwa, M., and Murakami, Y., J. Phys. Chem. Solids 50, 487 (1989). 18. Yamazaki, K., Niwa, M., and Murakami, Y., Kagaku Kogaku Ronbunshu 16, 564 (1990). 19. Shiojiri, S., Hirai, T., and Komasawa, I., Chem. Commun. 1439 (1998). 20. Hirai, T., Watanabe, T., and Komasawa, I., J. Phys. Chem. B 103, 10120 (1999). 21. Sung-Suh, H. M., Luan, Z., and Kevan, L., J. Phys. Chem. B 101, 10455 (1997).