1184
Ind. Eng. Chem. Res. 2003, 42, 1184-1189
MATERIALS AND INTERFACES Photocatalytic Degradation of CH3Cl over a Nickel-Loaded Layered Perovskite Dong Won Hwang, Kyung Yong Cha, Jindo Kim, Hyun Gyu Kim, Sang Won Bae, and Jae Sung Lee* Department of Chemical Engineering and School of Environmental Science & Engineering, Pohang University of Science and Technology (POSTECH), San 31 Hyoja-dong, Pohang 790-784, Republic of Korea
A highly donor-doped (110) layered perovskite, La2Ti2O7, was found to be an efficient photocatalyst comparable to TiO2 P25 and better than K2La2Ti3O10, the latter being an undoped (100) layered perovskite, in oxidative degradation of CH3Cl in the gas phase under UV irradiation. The activity depended on the crystallinity and the surface area of perovskite oxide. The grinding methods of the mixed oxide precursors before sintering as well as the sintering temperature affected the crystallinity and the surface area of the formed perovskite. Loading of Ni to perovskite materials increased the photocatalytic activity in degradation of CH3Cl markedly by constructing a p-n junction between nickel oxide and perovskite material. The optimum nickel loading was dependent strongly on the sintering temperature and the surface area of the perovskite material, and a slight deviation from the optimal nickel loading caused a drastic decrease in the photocatalytic activity. 1. Introduction Much attention has been paid to the destruction of volatile organic compounds (VOCs), which leads to discomfort and health-related problems of human beings. Especially, VOCs containing chlorinated compounds are more strongly toxic. Many methods have been tried to remove these compounds such as adsorption, biochemical technology, UV oxidation, thermal oxidation, and catalytic combustion.1-4 However, photocatalysis using TiO2 has been considered as an alternative because it provides the advantages of operation at room temperature and a highly stable catalytic activity. When illuminated by UV light with a wavelength of less than 380 nm, TiO2 catalyzes the degradation of many kinds of organic compounds such as organophosphorus pesticides, 4-chlorophenol, acetaldehyde, and chloroform.5-8 The mechanism of photocatalytic decomposition is believed to be as follows: (1) the electrons and holes are generated when the photocatalyst is irradiated by light with a wavelength smaller than one corresponding to its band-gap energy; (ii) hydroxyl (•OH-), hydrosuperoxide (•HO2), and superoxide (•O2-) radicals as the main oxidative species are formed by a series of subsequent reactions involving electrons and holes; (iii) VOCs are oxidized by these oxidative species.9,10 Generation of the oxidative species by photogenerated electrons and holes is thought to be the rate-determining step.11,12 Various types of modified TiO2 were studied to improve the photocatalytic activity in VOC decomposition. Nanometer-sized TiO2, mesoporous-structured TiO2, * To whom correspondence should be addressed. Fax: 82562-279-5528. E-mail:
[email protected].
and pillared TiO2 in silica were reported to have improved activity relative to that of the unmodified one.13-15 Other semiconductors could also be alternatives to the most common TiO2 photocatalyst. V2O5, WO3, ZnO, ZrO2, and ZrTiO4 were reported to be efficient in the gas-phase decomposition of organic compounds.16-20 Although TiO2 or other bulk-type semiconductor photocatalysts showed good photocatalytic activities in VOC abatement, it is worthwhile to investigate new photocatalysts with different structures for higher activity. Recently, we have developed a series of new photocatalysts made of highly donor-doped (110) layered perovskites such as Sr2Nb2O7, La2Ti2O7, Ca2Nb2O7, and La4CaTi5O17, which were found to be much more efficient than bulk-type TiO2 and undoped (100) layered perovskite photocatalysts such as K4Nb6O17 in a photocatalytic water-splitting reaction.21,22 Layered perovskite means that there exists interspacing between perovskite slabs consisting of a TiO6 or NbO6 octahedron. This higher activity resulted from the highly donor-doped electronic structure, which leads to a narrower depletion region and a more facile separation of charge carriers. If this superior property to those of TiO2 and (100) undoped layered perovskite oxides could be applied to photocatalytic degradation of VOC, a series of novel catalysts with much better activity might be obtained. This paper describes photocatalytic degradation of methyl chloride (CH3Cl) as a model compound of chlorinated VOC over a highly donor-doped (110) layered perovskite, La2Ti2O7. Gas-phase photocatalytic activity of the (110) layered perovskite oxide under UV light will be compared with that of TiO2 P25 (Degussa) and a (100) layered perovskite oxide, K2La2Ti3O10.23 The two perovskites are differentiated by the direction of their
10.1021/ie020457c CCC: $25.00 © 2003 American Chemical Society Published on Web 02/11/2003
Ind. Eng. Chem. Res., Vol. 42, No. 6, 2003 1185
layers relative to the structure of perovskite. These layered perovskites, to our knowledge, have never been used for the treatment of VOCs. The effects of catalyst preparation variables are also investigated on their catalytic performance. 2. Experimental Section 2.1. Catalyst Preparation. Although the highest activity in the water-splitting reaction over (110) layered perovskite materials was obtained over Sr2Nb2O7 with the quantum efficiency of 23%, La2Ti2O7 was selected in this study of CH3Cl degradation for its diversity and simplicity of sample fabrication. La2Ti2O7 photocatalyst was synthesized by a solid-state reaction. A mixture of La2O3 (99.9%, Aldrich) and TiO2 (99.9%, Aldrich) in a stoichiometric ratio was ground. At this stage, two methods of grinding (“dry” grinding and “wet” grinding in an ethanol solvent) were used. The ground powder was pressed with a pressure of 3000 kg/cm2 in the form of a pellet. Pellets were sintered at 1173-1473 K for 10 h in an electric furnace under an air atmosphere. Nickel metal was loaded on the La2Ti2O7 synthesized above by the impregnation method; the perovskite material was added in an aqueous solution containing a required amount (0.01-5.0 wt % of La2Ti2O7 powder) of a nickel nitrate and then was dried in an oven at 373 K and calcined at 573 K in air for 1 h. The nickel-loaded catalysts were then pretreated in a closed gas-circulation system; the sample was reduced by H2 (22 µmol/s) at 773 K for 2 h and then oxidized by air (22 µmol/s) at 473 K for 1 h. 2.2. Catalyst Characterization. The crystal structure of the sintered powder was determined by X-ray diffraction (XRD; Mac Science Co., M18XHF), and the band-gap energy was measured by UV-vis diffuse reflectance spectroscopy (Shimadzu, UV 525). The Brunauer-Emmett-Teller (BET) surface area was determined by N2 adsorption in a constant-volume adsorption apparatus (Micrometrics, ASAP 2012), and the morphology was examined by scanning electron microscopy (SEM; Hitachi, S-2460N). Transmission electron microscopy (Philips, CM 200) was used to observe the dispersion of the metal on the oxide powders. 2.3. Film Preparation and Photocatalytic Degradation of CH3Cl. The materials obtained from the above procedures were ground to fine powders in agate. The powders were coated on a glass plate (2 cm × 2 cm) with double-faced tape. The loading of catalysts on the glass plate was approximately 10 mg/4 cm2. A photocatalytic reaction was carried out over this film at room temperature and at atmospheric pressure in a closed gas-circulation system, shown in Figure 1. The photocatalytic reactor containing the glass plate was connected to the reaction system and sealed with silicone grease. The reactant gas of CH3Cl was injected into the gas-circulation system with a gas-tight syringe. The initial concentration of CH3Cl was 200 ppm balanced with air, which was controlled according to gas chromatograph (GC) analyses (HP 5890 II). The irradiation source was a high-pressure Hg lamp (Oriel Inc., 500 W), and IR radiation was removed by a water filter in front of the UV lamp to eliminate the thermal excitation of catalyst. Upon reaching a steady state, UV irradiation was initiated and continued for 3 h. The change in the CH3Cl concentration was followed by analysis with a GC equipped with an flame ionization
Figure 1. Reaction system for gas-phase photocatalytic CH3Cl degradation.
Figure 2. CH3Cl degradation as a function of the irradiation time. Reaction conditions: catalyst, 10 mg; CH3Cl, 200 ppm; 25 °C; volume, 40 cm3. Catalyst: La2Ti2O7 was sintered at 1373 K for 10 h, and 0.30 wt % of nickel was loaded and then pretreated by reduction at 773 K for 2 h followed by oxidation at 473 K for 1 h.
detector (He carrier) and an AT-1 column. The product distribution was analyzed by a GC equipped with a quadrupole mass spectrometer (HP 5972). 3. Results 3.1. Reaction Profile and Product Distribution. Figure 2 shows the reaction profile of CH3Cl degradaion over nickel-loaded La2Ti2O7. The La2Ti2O7 sample sintered at 1373 K for 10 h was loaded with 0.30 wt % of nickel and then pretreated by reduction at 773 K for 2 h followed by oxidation at 473 K for 1 h. The band gap of La2Ti2O7 was estimated to be 3.2 eV from its absorption edge of 360 nm of the UV diffuse reflectance spectrum. Before photocatalytic reaction, a blank test was performed without catalyst and irradiation. The concentration of CH3Cl did not change with time under UV irradiation without catalyst or under dark conditions in the presence of the catalyst. When nickel-loaded La2Ti2O7 was used as a photocatalyst under UV light, however, a significant decrease in the CH3Cl concentration from 200 to 54 ppm was observed for an irradiation time of 3 h. Products were analyzed by GC-MS, and the main products were CO2, CH4, H2O, and HCl, but quantitative analysis for these products was impossible because the reactant, CH3Cl, had a very low initial
1186
Ind. Eng. Chem. Res., Vol. 42, No. 6, 2003
Table 1. Physical Properties and Photocatalytic Activity in Oxidative Degradation of CH3Cl photocatalyst
BET surface area (m2/g)
CH3Cl conversiona (%)
TiO2 P25 K2La2Ti3O10b La2Ti2O7c
50 2.0 2.0
77 55 73
a After reaction of 3 h at room temperature and atmospheric pressure with 200 ppm of CH3Cl in air. b K2La2Ti3O10: sintered at 1273 K for 10 h and nickel loading of 1.0 wt %. c La2Ti2O7: sintered at 1373 K for 10 h and nickel loading of 0.3 wt %.
concentration (about 200 ppm). This product distribution was nearly consistent with the result of Solymosi et al. that CO2, H2O, and HCl were produced mainly from the catalytic decomposition of CH3Cl over Cr2O3doped SnO2 in the presence of O2. CH4 might come mainly from the incomplete oxidation.23 This result was also supported by many other works where CO2 and H2O were reported to be the main products in the photocatalytic oxidation of many VOCs such as acetaldehyde, acetic acid, 1-butene, and ethylene over a TiO2 photocatalyst.7,25-27 3.2. Photocatalytic Activity of La2Ti2O7. Table 1 shows a comparison of the BET surface area and photocatalytic activity of La2Ti2O7 with those of TiO2 P25, which has been accepted as the standard photocatalyst for most photocatalytic reactions. The photocatalytic activity of K2La2Ti3O10 was also shown in this table to compare a (100) layered perovskite photocatalyst with a (110) layered perovskite photocatalyst because this material also showed good photocatalytic activity in the water-splitting reaction.23 For La2Ti2O7, the highest activity in CH3Cl degradation was obtained when La2Ti2O7 was sintered at 1373 K for 10 h and 0.30 wt % of nickel was loaded, while for K2La2Ti3O10, the highest activity was obtained when K2La2Ti3O10 was sintered at 1273 K for 10 h and 1.0 wt % of nickel was loaded. We tested also Ni-loaded TiO2 as well as pure TiO2 for CH3Cl degradation. However, Ni loading on TiO2 did not result in the increase of activity of TiO2. The conversion of CH3Cl during the reaction time of 3 h was on the order of TiO2 P25 (77%) > La2Ti2O7 (73%) > K2La2Ti3O10 (55%) for the same amount (10 mg) of each catalyst. Thus, it could be concluded that layered perovskite photocatalyst, La2Ti2O7, has an excellent activity of CH3Cl degradation that is comparable to the that of a conventional TiO2 photocatalyst and better than that of the similar layered perovskite, K2La2Ti3O10, on a total weight basis. 3.3. Effect of the Grinding Methods of the Precursor Oxide Mixture. The photocatalytic activity was highly dependent on the grinding method for a La2O3TiO2 mixture, as shown in Figure 3. The catalyst labeled “wet grinding” was prepared from the precursor mixture ground in the presence of an ethanol solvent, and the “dry grinding” catalyst was made with grinding without the solvent. Both catalysts prepared from dry-ground and wet-ground precursors were sintered at 1273 K for 10 h. Optimum nickel loading was about 0.5 wt % for both oxides, which was very sensitive to the sintering temperature of an oxide mixture, as will be discussed later. The wet-ground catalyst was much more active (65% conversion) than the dry-ground one (23%). The properties of La2Ti2O7 prepared from differently ground precursors were investigated by both SEM and XRD. The particle sizes of La2Ti2O7 obtained from wet grinding were 1-2 µm, but those of La2Ti2O7 obtained from
Figure 3. Dependence of the photocatalytic activity on the grinding method (wet grinding vs dry grinding). Reaction conditions: catalyst, 10 mg; CH3Cl, 200 ppm; 25 °C; reactor volume, 40 cm3. Catalyst: La2Ti2O7 was sintered at 1273 K for 10 h, and nickel was loaded and then pretreated by reduction at 773 K for 2 h followed by oxidation at 473 K for 1 h.
Figure 4. SEM images of La2Ti2O7 prepared from precursor mixtures ground with and without an ethanol solvent: (A) wet grinding; (B) dry grinding. Both catalysts were sintered at 1273 K for 10 h.
dry grinding were 1-5 µm, as shown in Figure 4. XRD patterns in Figure 5 show that the crystallinity of La2Ti2O7 was higher for the catalyst prepared from wetground precursors as indicated by the narrower width of XRD peaks. Thus, it appears that a more homogeneous particle size distribution of the oxide mixture is obtained with wet grinding of La2O3 and TiO2, and this could result in a more crystalline La2Ti2O7 and eventually the higher photocatalytic activity. 3.4. Dependence of Optimum Nickel Loading on the Sintering Temperature. Bare La2Ti2O7 was not efficient in CH3Cl decomposition under UV irradiation,
Ind. Eng. Chem. Res., Vol. 42, No. 6, 2003 1187
Figure 5. XRD patterns of La2Ti2O7 prepared from precursor mixtures ground with and without an ethanol solvent: (A) wet grinding; (B) dry grinding. Both catalysts were sintered at 1273 K for 10 h.
Figure 7. Relationship between the BET area and optimum Ni loading on La2Ti2O7 for the highest photocatalytic activity of CH3Cl degradation. Table 2. Dependence of the Photocatalytic Activity on Crystallinity and the Surface Area sintering temp (K)
relative crystallinity (%)a
BET surface area (m2/g)
CH3Cl conversion (%)
1173 1273 1373 1473
90 100 100 100
4.0 2.9 2.0 1.5
20 65 73 16
a
Figure 6. Dependence of the photocatalytic activity on the sintering temperature and nickel loading. Reaction conditions: catalyst, 10 mg; CH3Cl, 200 ppm; 25 °C; reactor volume, 40 cm3. Catalyst: nickel-loaded La2Ti2O7 (wet grinding) was pretreated by reduction at 773 K for 2 h followed by oxidation at 473 K for 1 h.
as was also observed in the water-splitting reaction. Thus, Ni metal was indispensable for the activation of perovskite material in photocatalytic degradation of CH3Cl under UV irradiation, as described in many other reports.22,28 When nickel particles are deposited on the external surface of La2Ti2O7, a p-n junction between nickel and a layered perovskite material could be formed, where more efficient charge separation occurs. The photocatalytic activity increased with the amount of external dopant (nickel loading) up to some point, but the further increase of nickel loading resulted in a lowering of the photocatalytic activity. Both nickel oxide, a p-type semiconductor, and the perovskite material, an n-type semiconductor, should absorb the sufficient photons needed for its band-gap excitation so that the p-n junction can be operated properly. When this ratio was not optimized, the photocatalytic activity of the catalyst was reduced. In addition, it was highly dependent on the sintering temperature of the perovskite material, as shown in Figure 6. The optimum loading varied from 0.75 to 0.25% for sintering temperature from 1173 to 1473 K. Figure 7 shows the dependence of the optimum nickel loading on the BET surface area of each catalyst. The BET surface area decreased linearly as the sintering temperature increased in this region (1173-1473 K), and from the slope of this plot, it could be calculated that La2Ti2O7 needed about 0.2 wt % of nickel loading/1 m2 of La2Ti2O7 surface area for maxi-
From XRD intensities of the (212) plane of La2Ti2O7.
mum photocatalytic activity. Thus, the optimum nickel loading should decrease as the sintering temperature increases because the surface area of La2Ti2O7 decreases as the sintering temperature increases. At an optimum nickel loading, the photocatalytic activity (the conversion of CH3Cl) with sintering temperature was in the order 1373 K (73%) > 1273 K (65%) > 1173 K (20%) > 1473 K (16%), as shown in Table 2. Although the BET surface area of La2Ti2O7 sintered at 1173 K was the highest (4.0 m2/g) among four catalysts, the crystalline La2Ti2O7 phase was not formed completely yet at 1173 K, and thus this catalyst showed the lowest activity. When the catalysts sintered at 1273 and 1373 K are compared, it is thought that crystallinity of perovskite is more important than the surface area because the photocatalytic activity of the catalyst sintered at 1373 K with a better crystallinity but a lower surface area is higher than that sintered at 1273 K with a worse crystallinity but a higher surface area. The photocatalytic activity of La2Ti2O7 sintered at 1473 K decreased despite the formation of a maximal crystalline phase because now the effect of the surface area took over. Table 2 summarizes the relationship between the photocatalytic activity and relative crystallinity. The crystallinity was calculated from XRD intensities of the (212) plane (diffraction angle of about 30°) of La2Ti2O7 with the highest intensity. The intensity of this peak increased with sintering temperature but remained unaltered for La2Ti2O7 catalysts synthesized at temperatures higher than 1273 K. Another important observation made in Figure 6 is that nickel-loaded La2Ti2O7 is active only at a very narrow range of nickel loading on La2Ti2O7 in CH3Cl oxidation. Thus, the photocatalytic activity of CH3Cl oxidation decreases drastically even by 0.1 wt % variation from the optimum nickel loading, which is a very
1188
Ind. Eng. Chem. Res., Vol. 42, No. 6, 2003
peculiar result compared with that of the water-splitting reaction over nickel-loaded perovskite oxide. For CH3Cl oxidation, the range showing the photocatalytic activity was only 0.2-0.4 wt % for the catalyst sintered at 1373 K, and it was totally inactive outside this range. For the water-splitting reaction, however, it was almost equally active in a wide range of 0.1-5.0 wt %. Thus, it is very important to control the nickel loading precisely in CH3Cl oxidation using La2Ti2O7 as the photocatalyst in CH3Cl degradation. 4. Discussions 4.1. Highly Donor-Doped Layered Perovskite Oxides. As a member of recently found photocatalysts, (110) layered perovskite oxides, La2Ti2O7 has a comparable activity relative to TiO2 P25, a well-known photocatalyst in the field of photocatalytic VOC degradation, as shown in Table 1. Its hypothetical intrinsic activity is greater than that of TiO2 P25 when the rates based on surface cation sites are compared. In addition, compared with K2La2Ti3O10, a member of (100) layered perovskite, La2Ti2O7 is found to be a more efficient photocatalyst in CH3Cl oxidation in a comparison of their mass activities as well as activities per surface sites. In any case, this work is the first demonstration, to our best knowledge, of photocatalytic VOC degradation activities of layered perovskite materials, which are known for their high water-splitting activities.21,22 If we compare the intrinsic activity per surface cationic sites, the order of the photocatalytic activity in CH3Cl degradation is consistent with that in the water-splitting reaction. That is, (110) layered perovskite catalyst is much more active than the previously known (100) layered materials or bulk-type oxide for the photocatalytic oxidation of CH3Cl into H2O and CO2 under UV irradiation. Probably, its structure might be important for the photocatalytic activity of CH3Cl oxidation, as it was in the photocatalytic water splitting. As described in an earlier paper, La2Ti2O7 is a member of highly donor-doped perovskite oxides. Replacement of Ca2+ or Sr2+ in common perovskite materials of CaTiO3 or SrTiO3 by a trivalent cation such as La3+ results in excess electrons and slabs of a distorted perovskite structure of four unit cells thick in order to accommodate excess oxygen. In the photocatalytic oxidation of CH3Cl, excited electron-hole pairs are generated when the catalyst is illuminated with light having energy equal to or greater than the band gap of the photocatalyst. A possible route to improve the photocatalytic activity in this process is to suppress the recombination of generated electrons and holes. In the highly donor-doped (110) perovskite, the charge separation of electron and hole would be more facile than other semiconductor materials such as TiO2 and K2La2Ti3O10 because of a narrower depletion layer and thus more drastic band bending.22 4.2. Crystallinity and Surface Area. An efficient photocatalyst requires a high surface area and a high crystallinity because many reaction steps of photocatalysis take place on the surface and a defect in the structure could provide a site for electron-hole recombination. Unfortunately, a synthesis condition favoring one deteriorates the other. The sintering temperature is a good example. At too low sintering temperatures, the formation of a crystalline phase is difficult, whereas at too high temperatures, the BET surface area becomes too small, although a highly crystalline phase is formed.
Therefore, there exists an optimum sintering temperature that gives maximum photocatalytic activity. This is nicely demonstrated in Table 2. More strictly speaking, the crystallinity seems more important than the surface area in the CH3Cl degradation because the activity of the catalyst changed greatly with sintering temperature, although their surface areas varied only slightly. However, the low activity for the catalyst sintered at the higher temperature indicates that the surface area is also an important factor for a good photocatalyst. The crystallinity of the perovskite oxide is determined by the grinding method as well as the sintering temperature. In the sintering process of perovskite oxide, the mixed state and particle size distribution of metal oxide precursor seems important for the crystalline phase formation. The metal oxide mixture ground in an ethanol solvent (wet grinding) showed the higher photocatalytic activity by forming the more homogeneous distribution of particle size and eventually the better crystalline phase than the oxide mixture ground without a solvent. 4.3. Nickel Loading. Nickel metal loading is also the key step for the efficient photocatalyst for oxidation of CH3Cl, as it was for the water-splitting reaction. It has been reported that the nickel metal oxide plays the role of the one axis in the construction of a p-n junction along with perovskite oxide by being located on the external surface of (110) layered perovskite oxide.22,23 It has been presumed that there exists an optimum loading of nickel irrespective of the sintering temperature of the metal oxide precursor. However, it was found that the optimum loading varied significantly with the sintering temperature. As the sintering temperature increases, the optimum loading of nickel decreases almost linearly. This linear decrease of optimum nickel loading with the increase of the sintering temperature resulted from the decrease in the surface of the perovskite oxide host, as shown by the nice correlation in Figure 7 between optimum nickel loading and the BET surface area of the catalysts. It appears that a balance of both exposed perovskite oxide and NiO is required for the maximum photocatalytic activity. In addition, a slight deviation from the optimum loading caused a drastic decrease of the photocatalytic activity, unlike in the water-splitting reaction. 5. Conclusions Nickel-loaded La2Ti2O7 was found to be an efficient photocatalyst comparable to TiO2 P25 (bulk-type) and better than K2La2Ti3O10 (undoped layered perovskite type) in photocatalytic degradation of CH3Cl. This high activity may result from their layered structure and the high concentration of donor level compared with undoped photocatalysts. Crystallinity and the surface area of perovskite oxide are very important factors for an efficient photocatalyst. A slightly reduced crystallinity caused a drastic decrease in the photocatalytic activity, and photocatalysts with the same crystallinity showed a linear relationship between the surface area and activity. These crystallinities and surface areas of perovskite oxides were determined by the sintering temperature as well as the grinding method of the precursor oxide mixture. Catalysts prepared from the wet-ground oxide mixture showed higher activities than those from the dry-ground mixture. The optimum nickel loading was determined by the surface area of perov-
Ind. Eng. Chem. Res., Vol. 42, No. 6, 2003 1189
skite oxide, and a slight deviation from optimum loading caused a drastic decrease of the photocatalytic activity. Acknowledgment This work was supported by the Brain Korea 21 project and from Research Center for Energy Conversion and Storage. Literature Cited (1) Elliott, J.; Watkins, J. Controlling volatile organic compound emission from industrial wastewater. Noyes Data 1990. (2) Morreti, E. C.; Mukhopadhyay, N. Chem. Eng. Prog. 1993, July, 20. (3) Josephson, J. Environ. Sci. Technol. 1984, 18, 222A. (4) Kosuko, M.; Nunez, C. N. J. Air Waste Manage. Assoc. 1990, 40, 254. (5) Zhao, M.; Chen, S.; Tao, Y. J. Chem. Technol. Biotechnol. 1995, 64, 339. (6) Hu¨gu¨l, M.; Boz, I.; Apak, R. J. Hazard. Mater. 1999, 64, 313. (7) Obuchi, E.; Sakamoto, T.; Nakano, K.; Shiraishi, F. Chem. Eng. Sci. 1999, 54, 1525. (8) Martin, C. A.; Baltanas, M. A.; Cassano, A. E. Catal. Today 1996, 27, 221. (9) Kormann, C.; Bahnemann, D. W.; Hoffmann, M. R. Environ. Sci. Technol. 1994, 22, 798. (10) Hoffmann, A. J.; Carraway, E. R.; Hoffmann, M. R. Environ. Sci. Technol. 1994, 28, 776. (11) Wang, C. M.; Heller, A.; Gerischer, H. J. Am. Chem. Soc. 1992, 114, 5230. (12) Gerischer, H.; Heller, A. J. Phys. Chem. 1991, 95, 5261. (13) Ding, X. Z.; Qi, Z. Z.; He, Y. Z. J. Mater. Sci. Lett. 1995, 14, 21.
(14) Dai, Q.; Zhang, Z. L.; He, N. Y.; Li, P.; Yuan, C. W. Mater. Sci. Eng., C 1999, 8-9, 417. (15) Jung, K. Y.; Park, S. B. Appl. Catal. B 2000, 25, 249. (16) Tanaka, T.; Nishimura, Y.; Kawasaki, S.; Ooe, M.; Funabiki, T.; Yoshida, S. J. Catal. 1989, 118, 327. (17) Sclafani, A.; Palmisano, L.; Marci, G.; Venezia, A. M. Sol. Energy Mater. Sol. Cells 1998, 51, 203. (18) Gouvea, C. A. K.; Wypych, F.; Moraes, S. G.; Duran, N.; Peralta-Zamora, P. Chemosphere 2000, 40, 427. (19) Navio, J. A.; Colon, G.; Hermann, J. M. J. Photochem. Photobiol. A: Chem. 1997, 108, 179. (20) Navio, J. M.; Marchena, F. J.; Macias, M.; Sanchez-Soto, P. J.; Pichat, P. J. Mater. Sci. 1992, 27, 2463. (21) Kim, H. G.; Hwang, D. W.; Kim, J.; Kim, Y. G.; Lee, J. S. Chem. Commun. 1999, 1077. (22) Hwang, D. W.; Kim, H. G.; Kim, J.; Cha, K. Y.; Kim, Y. G.; Lee, J. S. J. Catal. 2000, 193, 40. (23) Ikeda, S.; Hara, M.; Kondo, J. N.; Domen, K.; Takahashi, H.; Okubo, T.; Kakihana, M. J. Mater. Res. 1998, 13, 852. (24) Solymosi, F.; Rasko, J.; Papp, E.; Oszko, A.; Bansagi, T. Appl. Catal. A 1995, 131, 55. (25) Muggli, D. S.; Falconer, J. L. J. Catal. 1999, 187, 230. (26) Cao, L.; Huang, A.; Spiess, F.-J.; Suib, S. L. J. Catal. 1999, 188, 48. (27) Park, D. R.; Zhang, J.; Ikeue, K.; Yamashita, H.; Anpo, M. J. Catal. 1999, 185, 114. (28) Domen, K.; Kudo, A.; Onishi, T. J. Catal. 1986, 102, 92. (29) Nozik, A. J. Appl. Phys. Lett. 1977, 30, 567.
Received for review June 20, 2002 Revised manuscript received December 2, 2002 Accepted December 7, 2002 IE020457C
