Photoelectrochemical Oxidation of Methanol on Oxide Nanosheets

Department of Nano Science and Technology, Graduate School of Science and ... activity, which means that the degree of photooxidation of the alcohol ...
6 downloads 3 Views 233KB Size
J. Phys. Chem. B 2006, 110, 4645-4650

4645

Photoelectrochemical Oxidation of Methanol on Oxide Nanosheets Kazuyoshi Izawa,* Takashi Yamada, Ugur Unal, Shintaro Ida, Ozge Altuntasoglu, Michio Koinuma, and Yasumichi Matsumoto Department of Nano Science and Technology, Graduate School of Science and Technology, Kumamoto UniVersity, Kurokami 2-39-1, Kumamoto 860-8555, Japan ReceiVed: October 28, 2005; In Final Form: January 19, 2006

Photoelectrochemical oxidation of alcohol on various nanosheet electrodes such as Nb6O17, Ca2Nb3O10, Ti0.91O2, Ti4O9, and MnO2 system host layers were measured to evaluate the photocatalysis of water photolysis with alcohol as a sacrificial agent. The nanosheet electrodes were prepared by the layer-by-layer (LBL) method, using electrostatic principles. The highest photooxidation current density was observed in methanol solution for Nb6O17 and Ca2Nb3O10 nanosheets, while the density was lower for Ti0.91O2, Ti4O9, and MnO2 nanosheets in decreasing order. The rank in the photocurrent density was in agreement with that in the photocatalytic activity, which means that the degree of photooxidation of the alcohol determines the activity of the alcohol in the water photolysis process. The photocurrent was independent of the number of nanosheet layers on the electrode, indicating that only the mono-nanosheet layer attached directly on a substrate acts as a photoelectrocatalyst and that the interlayer space is not important. Consequently, higher photooxidation current on the Nb6O17 mono-nanosheet layer means that the charge separation of electron and hole under illumination is very large and that the hole-capturing process by CH3OH is very quick compared with the surface recombination on the Nb6O17 nanosheet. The adsorption of a transition metal cation on the nanosheet acted as the surface recombination center, because the photocurrent decreased after the adsorption. The photocatalytic mechanism has been discussed in detail in terms of various photoelectrochemical behaviors.

Introduction Layered oxides consisting of Ti, Nb, or Ta host layers have high catalytic activity in water photolysis especially in the presence of sacrificial reagents. In water photolysis with these semiconductive host layers, electrons and holes produced under the illumination of light with energy higher than the band gap of the host layer reduce and oxidize H2O to produce H2 and O2, respectively.1-9 The photocatalytic mechanism in layered oxides has been proposed that the intercalated H2O is preferentially oxidized by hole to produce O2 in the interlayer, while surface H2O is reduced by electron at the edge of the host layer.3 On the other hand, another mechanism, in which both reduction and oxidation reactions occur for the intercalated H2O, has also been proposed.4 CH3OH, a sacrificial reagent, strongly promotes the H2 evolution by its preferential photooxidation, as it acts as a hole captor. In this case, the alcohols with smaller molecular size such as CH3OH and C2H5OH promote the photocatalytic H2 evolution more strongly than those with larger molecular size.2,10-12 The size selectivity for the photooxidation of alcohol is explained by the intercalation mechanism. Alcohols with smaller molecular size is preferentially oxidized compared with those with larger molecular size because of easier intercalation of smaller alcohol molecules into the interlayer. Thus, the intercalation mechanism is one of the key points in the photocatalytic behavior of the layered oxide materials. On the other hand, the promotion of the photocatalytic evolution of H2 by alcohol photooxidation also has been observed at other photocatalysts such as TiO2.13,14 The mech-

anism of this promotion can be explained by rapid capture of the produced hole at the photocatalyst surface. This phenomenon indicates that the reaction rates of electron and hole with chemical species at the photocatalyst surface are more important for the photocatalysis than energy positions of the CB and VB edges, since the presence of alcohol will not affect their energy positions. At the surface, holes and electrons are consumed by both charge transfer and surface recombination. Therefore, the rapid hole capture by alcohol will bring about the reduction in the surface recombination rate, leading to the promotion of H2 evolution by electron. However, above intercalation and surface mechanism in the photocatalysis of water in the presence of alcohol have not been cleared yet. Photoelectrochemical approach to the photolysis of H2O is very effective in making the mechanism clear, because both reduction and oxidation reactions can be analyzed separately by cathodic and anodic photocurrents, respectively,15-18 while both reactions take place in the photocatalytic reaction. Especially, analysis of the anodic photoelectrochemical oxidation of alcohol will be very important in making the promotion mechanism of the H2 photocatalytic evolution on a layered oxide photocatalysts clear. In this paper, the photoanodic behavior of layered oxide photocatalysts has been analyzed and the photocatalytic mechanism for H2O photolysis in the presence of alcohol will be discussed. The nanosheets are used as the starting material to prepare the electrode by the layer-by-layer (LBL) method19-26 in order to control the layered structure affecting the mechanism. Experimental Section

* Address correspondence to this author. E-mail: 049d8007@ gsst.stud.kumamoto-u.ac.jp. Phone: +81-96-342-3659. Fax: +81-96-3423679.

Materials. The starting materials K4Nb6O17, KCa2Nb3O10, K2Ti4O9, and K0.45MnO2 were prepared by a conventional solid-

10.1021/jp056210l CCC: $33.50 © 2006 American Chemical Society Published on Web 02/21/2006

4646 J. Phys. Chem. B, Vol. 110, No. 10, 2006 state method.27-30 K4Nb6O17 was prepared by heat treatment of the stoichiometric mixture of Nb2O5 and K2CO3 at 1200 °C for 15 min in a Pt crucible. KCa2Nb3O10 was prepared by the calcination of the stoichiometric mixture of Nb2O5, K2CO3, and CaCO3 at 1000 °C for 5 h and then at 1150 °C for 10 h. The mixture of K2CO3 and TiO2‚nH2O with a TiO2/K2CO3 molar ratio of 3 was dried at 90 °C for 10 h and then heated at 960 °C for 10 h to synthesize K2Ti4O9. K0.45MnO2 was prepared by heat treatment of the stoichiometric mixture of K2CO3 and Mn2O3 at 800 °C for 30 h. CsxTi(2-x/4)0x/4O4 (0: vacancy) was prepared by the complex polymerization method.15,21 Prior to the addition Ti(OCH(CH3)2)4 (10.796 mL) and Cs2CO3 (3.4983 g) were dissolved in a mixture of methanol (160 mL) and ethyleneglycol (60 mL). During the addition of titanium isopropoxide, the mixture was stirred vigorously with a magnetic stirrer and the temperature was raised gradually to 50 °C. The resultant solution was clear. Next, anhydrous citric acid (28.8 g) was added and the temperature was raised to 150 °C, yielding a resin-like mass, which transformed into ash and then into white powder upon raising the temperature to 300 °C. Heat treatment of the powder at 800 °C yielded the final CsxTi(2-x/4)0x/4O4 product. All the samples were analyzed by X-ray diffraction (XRD, using Cu KR radiation, RIGAKU, RINT-2500 VHF) patterns to confirm the structure. These powders were treated with 5 M HNO3 or 1 M HCl aqueous solution to protonate the interlayer by ion-exchange reaction between interlayer cations and protons in the solution. Protonated powders were exfoliated in tetrabutylammonium (TBA) solution for 72 h. The amount of TBA in the solution was 8-fold higher than that of powder in molar ratio. The subsequent centrifugation of the solution under 3000 rpm for 30 min yielded colloidal suspension of nanosheets. The host nanosheets obtained from K4Nb6O17, KCa2Nb3O10, CsxTi(2-x/4)0x/4O4, K2Ti4O9, and K0.45MnO2 were denoted as NbO, CaNbO, Ti0.91O2, Ti4O9, and MnO for convenience in the present paper. Preparation of the LBL Electrode. Deposition of the films by the LBL method was carried out by the same procedure as reported in other related studies.19-26 Au was used as a substrate in this study. It was cleaned by sonification in a 5 M HNO3 as well as in acetone, and washed with water in an ultrasonic bath. Substrates were primed in aqueous polyethyleneimine (PEI) (2.5 g/L) solution for 10 min to charge the surface of the substrate positively. Primed substrates were dipped into the colloidal exfoliation solution having negatively charged nanosheets for 10 min and then into 0.1 M aqueous lanthanum acetate solution for another 10 min to form the layered structure (number of layers (n) g2) by electrostatic deposition principle. Electrodes were rinsed with water to remove the excessively adsorbed species and dried under an N2 stream between the deposition steps. Repeating these layers produces n layers on the modified substrates. The deposition model of the prepared LBL film electrodes is illustrated in Figure 1, where n denotes the number of nanosheets and La3+ is intercalated in the film. La3+ only acts as a electrostatic binding cation between nanosheets to form the layered structure, but does not affect the photoelectrocatalysis, because La3+ adsorption had no effect on the photocurrent of the LBL electrode with n ) 1. As seen in the figure, it is considered that the nanosheets adsorbed onto PEI do not arrange uniformly, therefore they partially contact with the substrate because PEI is adsorbed nonuniformly onto the substrate. The effect of nonuniform adsorption of the nanosheets is described in a later section.

Izawa et al.

Figure 1. Model of the LBL film on the electrode (n shows the number of layers).

Photoelectrochemical Properties. All electrochemical experiments were carried out in a conventional three-electrode electrochemical cell with a Pt counter electrode and a saturated Ag/AgCl reference electrode. The working electrode potentials were referred to this reference electrode unless otherwise stated in this paper. A 500 W ultrahigh-pressure Hg lamp having the 60 mW/cm2 light intensity was used as a light source to measure the photoelectrochemical properties. Cyclic voltammograms (CV) were measured under a potential sweep rate of 20 mV/s. A 0.5 M K2SO4 solution (pH 6.2) was used as an electrolyte solution. N2 saturation was reached in the electrolytes in a quartz cell before the electrochemical measurement. Characterization. The crystal structure was analyzed from X-ray diffraction (XRD) patterns. The sizes of the prepared nanosheets were observed with an atomic force microscope (AFM, Molecular Imaging). The surface condition of LBL films was observed by scanning electron microscopy (SEM, JEOL, JSM-6060LV). The compositions of the deposited films were analyzed by an inductively coupled plasma (ICP) spectrophotometer (Seiko Instruments, SPS7800) after dissolving the film in acid solutions (HF or H2SO4). Results and Discussion Figure 2 shows typical AFM images of the exfoliated nanosheets of CaNbO (A), NbO (B), and Ti4O9(C) used in this work. The samples for AFM measurements were prepared by drying after dropwise application of the present nanosheet suspensions onto a mica substrate. For all the samples, the sizes were in the range from about 20 to 1000 nm, and the thickness was from 0.84 to 1.61 nm, depending on the kind of nanosheets. The nanosheets used in this work are summarized in Table 1. The thicknesses of all the nanosheets are close to the values reported previously,31-35 and are listed in Table 1. This indicates that almost all nanosheets exist as a mono-nanosheet in the suspension. The small difference in the thickness value will be based on the adsorbed species on the nanosheet surface such as H2O.33 Figure 3 shows a typical SEM micrograph of the surface of the LBL electrode covered with NbO nanosheets (n ) 1). Other electrodes also showed similar surface structure. The surface is smooth at the micrometer level, but somewhat rough at the nanometer level. This characteristic of the surface may be important to bring about the high photocurrent as stated in a later section. The nanosheets on the LBL electrode (n ) 1) were dissolved in HCl or HF solution, and then the solutions were

Oxidation of Methanol on Oxide Nanosheets

J. Phys. Chem. B, Vol. 110, No. 10, 2006 4647

Figure 4. Cyclic voltammograms of NbO electrodes (n ) 1) in the first sweep: (a) in 0.1 M CH3OH + 0.5 M K2SO4 and (b) in 0.5 M K2SO4.

Figure 2. Typical AFM images of the prepared nanosheets of CaNbO (A), NbO (B), and Ti4O9 (C) used in the current analysis.

TABLE 1: Various Nanosheets Used in the Current Work

NbO CaNbO Ti0.91O2 Ti4O9 MnO

lateral size (nm)

thickness (nm)

thickness of host layera (nm)

20-200 20-200 50-1000 20-600 50-800

1.03 1.61 0.84 1.18 1.09

1.0 1.47 0.73 1.08 0.52

a The thickness is calculated from our XRD results or the data in refs 31-35.

analyzed by ICP measurement to evaluate the state of the film in terms of the number of the nanosheets on the electrode surface. Assuming that nanosheets cover the whole surface of a substrate, the theoretical number of nanosheets deposited on

Figure 3. SEM micrographs of the LBL films (NbO) with n ) 1.

a substrate was estimated from the number of corresponding metal cations, i.e., Ti, Nb, etc., dissolved in the solution. According to the ICP data, the number has such a value that nanosheets are piled up to form 1 to 10 layers on some parts for all the electrodes with n ) 1. Considering the roughness factor of the substrate surface, which is usually small for a smooth metal plate, the above result suggests that the electrode surfaces are not always covered with a layer formed of mononanosheets even for the LBL electrode with n ) 1, but are sometimes covered with an aggregated form of the nanosheets. The estimated number of individual nanosheets dissolved in the solution increased linearly with the number of LBL layers (n), indicating that films with n layers can be formed in the layerby-layer style despite the existence of agglomerates on some parts of the electrode. This was confirmed by an almost linear increase in the band gap absorption intensity of films with increasing n. Consequently, the present LBL film on the electrode exists in the state illustrated in Figure 1. Figure 4 shows typical voltammograms of NbO electrodes (n ) 1) in the first sweep cycle. CaNbO electrode showed a similar voltammogram. Higher photocurrent was observed in the solution with CH3OH than that in the solution without CH3OH, indicating that photooxidation of CH3OH occurred in preference to that of H2O. The photocurrent increased from the start potential (0 V) to about 0.5 V, and then saturated at 0.9 V for the NbO electrode as shown in Figure 4. A similar increase in the photocurrent was also observed for the reverse scan from the start potential to the negative potential. These results suggest that the increase in the photocurrent will be based on the photoinduced activation of the surface but not the potential-induced activation. This phenomenon may contribute to the photoinduced conversion of the surface to hydrophilicity.36 After the photoinduced activation in the first cycle, the voltammogram was scarcely changed. Other electrodes also showed similar behavior. The second scan voltammograms for all of the electrodes in 0.1 M CH3OH are shown in Figure 5. The photocurrents for various

4648 J. Phys. Chem. B, Vol. 110, No. 10, 2006

Izawa et al.

Figure 7. The dependence of the photocurrent on the number of NbO nanosheets, n. The photocurrent was measured at 0 V vs Ag/AgCl in 0.1 M CH3OH + 0.5 M K2SO4.

Figure 5. Cyclic voltammograms of various LBL electrodes (n ) 1) in 0.1 M CH3OH + 0.5 M K2SO4: (a) NbO, (b) CaNbO, (c) Ti4O9, (d) Ti0.91O2, and (e) MnO.

Figure 6. Photocurrent intensities for LBL electrodes (n ) 1) at 0 V in various alcohol solutions. Electrolyte: 0.1 alcohol + 0.5 M K2SO4. NbO film (dashed bar), CaNbO film (white bar), Ti0.91O2 film (dotted bar), and Ti4O9 film (black bar).

electrodes (n ) 1) in 0.1 M CH3OH at 0 V in Figure 5 are shown in Figure 6 together with the dependence on the type of alcohol to be oxidized. The photocurrent onset potentials under illumination were around -0.71, -0.65, -0.60, and -0.49 V for NbO, CaNbO, Ti4O9, and Ti0.91O2 electrodes, respectively. We have also measured the rest photopotential, which can be defined as the initial potential of the electrode surface under irradiation with a Hg lamp right before starting the cyclic sweep. The rest photopotentials were around -0.77, -0.69, -0.58, and -0.58 V for NbO, CaNbO, Ti4O9, and Ti0.91O2 electrodes, respectively. When the NbO and CaNbO electrodes were held at the photocurrent onset potentials, H2 evolution was observed because the equilibrium hydrogen potential was -0.57 V in the solution (pH 6.2). Higher photocatalytic activity of the NbO system for H2 evolution in the presence of CH3OH than that of the TiO system is based on more negative photocurrent onset potentials for the former system than for the latter system. This is related to a higher energy level of the conduction band of the NbO system with an n-type semiconducting property.16 The photocurrent due to the oxidation of CH3OH was the highest for NbO system electrodes, and was scarcely observed for the MnO electrode. These tendencies are in agreement with the photocatalytic activities for H2 evolution in the presence of CH3OH.7,12,37-41 Higher photocurrent is related to higher

photooxidation rate of CH3OH by hole in preference to the surface recombination rate between electron and hole at the surface. Consequently, the highest photocatalytic activity of the NbO system is based on the highest photocurrents of CH3OH oxidation as well as the most negative photocurrent onset potentials of the NbO system electrodes as stated already. From the dependence of the photocurrent on the type of alcohol as shown in Figure 6, it was found that the highest photocurrent was observed for CH3OH and the photocurrent decreased with the increase in molecular weight of the alcohol. The tendency of the increase in the photocurrent with the molecular weight is similar to that of the photocatalytic H2 evolution in the presence of alcohol. It has been suggested that alcohol molecules with smaller size intercalate much more easily than those with larger size because the intercalated alcohol can be photooxidized.12 However, this photocatalytic mechanism of the intercalation selectivity does not completely explain the present results. The interlayer structure did not affect the photocurrent, since the layer number, n, scarcely affected the photocurrent. That is, the LBL electrode even with n ) 1 showed the highest possible photocurrent as stated below. Consequently, photoelectrocatalytic activity at the surface of the nanosheet used in the present study is the highest for CH3OH and is lower for the alcohol with heavier molecular weight. HCOOH and HCHO were also easily photooxidized in the same photoelectrochemical test. These results may suggest that hydrophilic property of organic compounds is important for them to be easily oxidized, if the nanosheets are assumed to have the same property on their surfaces.36 Easier adsorption of organic compound with higher hydrophilic property will occur on the nanosheet surface with the same property. In fact, the photocurrent decreased with the reduction in the hydrophilic property of the nanosheet surface, which was obtained by heat treatment at 300-400 °C before the photoelectrochemical test. The dependence of the photocurrent of CH3OH at 0 V on the apparent number of nanosheet layers, n, was measured for all of the LBL electrodes. The typical result for the NbO electrode is shown in Figure 7. The photocurrent was almost independent of the number of layers, where the samples with n higher than 2 will have the layered structure. This result indicates that the adsorption of CH3OH on the nanosheet surface is more important than the intercalation for the photooxidation of CH3OH. This mechanism is somewhat different from the mechanism of H2O photooxidation at the layered oxide photocatalysts, where the intercalated H2O is photooxidized in preference to free H2O in the solution. Moreover, the independence of photocurrent on the number of layers indicates that only nanosheets in direct contact with the substrate surface contribute to the photocurrent. Even in the case of the contact of mononanosheets with the substrate through PEI molecules and uniform coverage of the whole surface, the electron transfer

Oxidation of Methanol on Oxide Nanosheets

J. Phys. Chem. B, Vol. 110, No. 10, 2006 4649

Figure 8. Change in the photocurrent (at 0 V vs Ag/AgCl) as a function of the concentration of CH3OH at the NbO electrode (n ) 1). Figure 10. A model of the photocatalytic mechanism in the presence of CH3OH.

Figure 9. Effect of the adsorbed transition metal cations on the NbO electrode (n ) 1) to the photocurrent (at 0 V vs Ag/AgCl). The electrolyte was 0.1 M CH3OH + 0.5 M K2SO4.

from the film surface to the Au substrate may take place without difficulty. Probably, the mono-nanosheets will not cover the surface uniformly and most of the nanosheets partly contact with the Au surface while some of them contact through the PEI molecules, which have a thickness of about 8 to 12 Å42,43 on the Au substrate. It is considered that the electron transfer from the nanosheet surface to the Au substrate was easier through the contact point than through the PEI molecules of which resistance can be negligible. The nonuniformity of the nanosheets is observed as stated already in Figure 1. Figure 8 shows the change in the photocurrent as a function of the concentration of CH3OH at the NbO electrode. The photocurrent increased with the concentration and saturated at about 0.1 M. The number of holes produced in the NbO nanosheet is constant under a constant intensity of UV light. The concentration of the hole used for the oxidation of CH3OH at the surface is not enough for the oxidation of CH3OH in a concentration range higher than about 0.1 M with the present light source. Figure 9 shows the effect of the adsorbed transition metal cation on the photocurrent of CH3OH. The adsorption was made by the immersion of the LBL electrodes into the solutions containing transition metal cations with positive charge (Fe3+, Pt(NH3)22+, Ni(NH3)62+, Co(NH3)63+) for 10 min. The adsorbed transition metals caused a reduction in the intensity of the photocurrent. This phenomenon is in agreement with the nonphotoresponse of Ti layered oxide intercalated with transition metal cations, where transition metal cations act as a recombination center of holes and electrons.15 The present adsorbed transition metal cations will also act as a surface recombination center, leading to the low photocurrent, although the effect of Pt(NH3)22+ is not so large. A model of the photocatalytic mechanism in presence of CH3OH is illustrated in Figure 10. The figure shows the chargetransfer mechanisms and possible electrochemical reactions

involving produced electrons and holes in the semiconductor nanosheets under illumination. The electron-hole separation occurs serially on the surface of nanosheets under illumination and is stronger in the Nb and Ti nanosheet films, which produce higher photocurrents as seen in Figure 5. The photocurrent remains constant in the 0-1 V range and it is not related to the potential range but the oxidation of methanol species adsorbed on the surface of the film. Almost all of the produced electrons and holes will immediately reach the nanosheet surface or edge because their thickness is very thin where bulk recombination can be neglected. CH3OH will easily adsorb on the nanosheet surface that is covered with OH leading to the hydrophilic property. The adsorbed CH3OH will be easily oxidized by the hole reaching the surface, while the surface electron will bring about the H2O reduction to produce H2. The presence of the adsorbed CH3OH will relatively decrease the rate of the surface recombination between the surface hole and electron pairs because the concentration of holes on the surface decreases as a result of the rapid transfer to CH3OH due to its higher tendency to capture holes. Consequently, the presence of CH3OH promotes H2 evolution under illumination because of a decrease in the surface recombination rate of the electron and hole. On the other hand, the adsorbed 3d transition metal cations promote the surface recombination rate, leading to a decrease in the photocurrent of CH3OH oxidation as stated already. Conclusion Photoelectrochemical oxidation of alcohol on various nanosheet electrodes (TiO, NbO, and MnO system host layers) was measured to evaluate the photocatalysis of water photolysis with alcohol as the sacrificial reagent. The nanosheet electrodes were prepared by the LBL method, using the electrostatic principle. The highest photooxidation current was observed in methanol solution for the NbO systems, and it was followed by TiO and MnO systems in decreasing order of intensity. The order of photocurrent intensity from the highest to lowest was in agreement with that of the photocatalytic activity meaning that the degree of photooxidation of alcohol determines the activity for the water photolysis when alcohol is used. Moreover, the photocurrent onset potentials reflecting the CB edge energy position were more negative for the nanosheet with higher activity. The photocurrent was independent of the number of layers on the electrode indicating that only a mono-nanosheet layer attached directly to the electrode acts as the photoelectrocatalyst and that the adsorbed species rather than the intercalated species are important for the photooxidation. The photooxidation rate decreased with the number of carbons in

4650 J. Phys. Chem. B, Vol. 110, No. 10, 2006 the alcohol molecule, which suggests that the hydrophilic surface state contributes to the alcohol adsorption. Consequently, the highest photooxidation current on the NbO mono-nanosheet means that charge separation to electron and hole under illumination is very large and that the hole capture rate by CH3OH is very quick compared with the surface recombination on the NbO nanosheet. Adsorbed transition metal cations on the nanosheet acted as surface recombination centers, because the photocurrent decreased after the adsorption. Acknowledgment. This work was supported by the Core Research for Evolutional Science and Technology (CREST) program of the Japan Science and Technology Co. (JST) and a Grant-in-Aid for Scientific Research (no. 440, Panoscopic Assembling and High Ordered Functions for Rare Earth Materials, and no. 15350123) from the Ministry of Education, Culture, Sports, Science, and Technology. References and Notes (1) Sekine, T.; Yoshimura, J.; Tanaka, A.; Domen, K.; Maruya, K.; Onishi, T. Bull. Chem. Soc. Jpn. 1990, 63, 2107-2109. (2) Kudo, A.; Kondo, T. J. Mater. Chem. 1997, 7, 777-780. (3) Takata, T.; Shinohara, K.; Tanaka, A.; Hara, M.; Kondo, J. N.; Domen, K. J. Photochem. Photobiol. A 1997, 106, 45-49. (4) Sayama, K.; Yase, K.; Arakawa, H.; Asakura, K.; Tanaka, A.; Domen, K.; Onishi, T. J. Photochem. Photobiol. A 1998, 114, 125-135. (5) Kudo, A.; Hijii, S. Chem. Lett. 1999, 1103-1104. (6) Machida, M.; Ma, W. X.; Taniguchi, H.; Yabunaka, J.; Kijima, T. J. Mol. Catal. A: Chem. 2000, 155, 131-142. (7) Ebina, Y.; Sasaki, T.; Harada, M.; Watanabe, M. Chem. Mater. 2002, 14, 4390-4395. (8) Choy, J.-H.; Lee, H.-C.; Jung, H.; Kim, H.; Boo, H. Chem. Mater. 2002, 14, 2486-2491. (9) Kudo, M.; Tsuzuki, S.; Katsumata, K.; Yasumori, A.; Sugahara, Y. Chem. Phys. Lett. 2004, 393, 12-16. (10) Domen, K.; Ebina, Y.; Ikeda, S.; Tanaka, A.; Kondo, J. N.; Maruya, K. Catal. Today 1996, 167-174. (11) Yin, S.; Wu, J.; Aki, M.; Sato, T. Int. J. Inorg. Mater. 2000, 2, 325-331. (12) Ohtani, B.; Ikeda, S.; Nakayama, H.; Nishimoto, S. Phys. Chem. Chem. Phys. 2000, 2, 5308-5313. (13) Galin´ska, A.; Walendziewski, J. Energy Fuels 2005, 19, 11431147. (14) Kawai, T.; Sakata, T. J. Chem. Soc., Chem. Commun. 1980, 15, 694-695. (15) Matsumoto, Y.; Funatsu, A.; Matsuo, D.; Unal, U.; Ozawa, K. J. Phys. Chem. B 2001, 105, 10893-10899. (16) Koinuma, M.; Seki, H.; Matsumoto, Y. J. Electroanal. Chem. 2002, 531, 81-85.

Izawa et al. (17) Unal, U.; Matsumoto, Y.; Tanaka, N.; Kimura, Y.; Tamoto, N. J. Phys. Chem. B 2003, 107, 12680-12689. (18) Matsumoto, Y.; Unal, U.; Tanaka, N.; Kudo, A.; Kato, H. J. Solid State Chem. 2004, 177, 4205-4212. (19) Kashak, D. M.; Mallouk, T. E. J. Am. Chem. Soc. 1996, 118, 4222. (20) Sasaki, T.; Ebina, Y.; Tanaka, T.; Harada, M.; Watanabe, M.; Decher, G. Chem. Mater. 2002, 14, 3524-3530. (21) Matsumoto, Y.; Unal, U.; Kimura, K.; Ohashi, S.; Izawa, K. J. Phys. Chem. B 2005, 109, 12748-12754. (22) Fang, M.; Kim, C. H.; Saupe, G. B.; Kim, H. N.; Waraksa, C. C.; Miwa, T.; Fujishima, A.; Mallouk, T. E. Chem. Mater. 1999, 11, 15261532. (23) Schaak, R. E.; Mallouk, T. E. Chem. Mater. 2000, 12, 2513-2516. (24) Wang, Z. S.; Ebina, Y.; Takada, K.; Watanabe, M.; Sasaki, T. Langmuir 2003, 19, 9534-9537. (25) Wang, Z. S.; Sasaki, T.; Muramatsu, M.; Ebina, Y.; Tanaka, T.; Wang, L. Z.; Watanabe, M. Chem. Mater. 2003, 15, 807-812. (26) Ma, R.; Sasaki, T.; Bando, Y. J. Am. Chem. Soc. 2004, 126, 1038210388. (27) Kudo, A.; Tanaka, A.; Domen, K.; Maruya, K.; Aika, K.; Onishi, T. J. Catal. 1988, 111, 67-76. (28) He, M.; Feng, X.; Lu, X.; Ji, X.; Liu, C.; Bao, N.; Xie, J. J. Mater. Sci. 2004, 39, 3745-3750. (29) Bao, N.; Feng, X.; Lu, X.; Yang, Z. J. Mater. Sci. 2002, 37, 30353043. (30) Wang, L.; Omomo, Y.; Sakai, N.; Fukuda, K.; Nakai, I.; Ebina, Y.; Takada, K.; Watanabe, M.; Sasaki, T. Chem. Mater. 2003, 15, 28732878. (31) Unal, U.; Matsumoto, Y.; Tamoto, N.; Koimuma, M.; Machida, M.; Izawa, K. J. Solid State Chem. 2005, 178, 3794-3801. (32) Fukuoka, H.; Isami, T.; Yamanaka, S. J. Solid State Chem. 2000, 151, 40-45. (33) Sasaki, T.; Ebina, Y.; Kitami, Y.; Watanabe, M. J. Phys. Chem. B 2001, 105, 6116-6121. (34) Verbaere, A.; Tournoux, M. Bull. Chim. Soc. Fr. 1973, 4, 12371241. (35) Omomo, Y.; Sasaki, T.; Wang, L.; Watanabe, M. J. Am. Chem. Soc. 2003, 125, 3568-3575. (36) Sasaki, T.; Ebina, Y.; Fukuda, K.; Tanaka, T.; Harada, M.; Watanabe, M. Chem. Mater. 2002, 14, 3524-3530. (37) Shibata, M.; Kudo, A.; Tanaka, A.; Domen, K.; Maruya, K.; Onishi, T. Chem. Lett. 1987, 1017-1018. (38) Domen, K.; Yoshimura, J.; Sekine, T.; Tanaka, A.; Onishi, T. Catal. Lett. 1990, 4, 339-343. (39) Domen, K.; Kudo, A.; Shibata, M.; Tanaka, A.; Maruya, K.; Onishi, T. J. Chem. Soc., Chem. Commun. 1986, 23, 1706-1707. (40) Sayama, K.; Tanaka, A.; Domen, K.; Maruya, K.; Onishi, T. J. Phys. Chem. 1991, 95, 1345-1348. (41) Abe, R. J. Mater. Res. 1998, 13, 861-865. (42) Sasaki, T.; Ebina, Y.; Tanaka, T.; Harada, M.; Watanabe, M. Chem. Mater. 2001, 13, 4661-4667. (43) Chaufer, B.; Baudry, M. R.; Bouguen, A.; Labbe´, J. P.; Que´merais, A. Langmuir 2000, 16, 1852-1860.