Modification of TaON with ZrO₂ to improve photocatalytic hydrogen evolution activity under visible light: influence of preparation conditions on activity

Abstract

The effects of the preparation conditions of ZrO₂-modified TaON (referred to as ZrO₂/TaON) on the photocatalytic activity for H₂ evolution under visible light (λ > 420 nm) were investigated. ZrO₂/TaON catalysts were prepared by loading particulate Ta₂O₅ with ZrO₂ using different zirconium precursors, followed by nitridation at 1123 K for different durations (5–20 h) under NH₃ flow. Nitridation of ZrO₂/Ta₂O₅ for 15 h resulted in the production of ZrO₂/TaON, regardless of the zirconium precursors used, but the physicochemical properties varied. The highest activity was obtained for ZrO₂/TaON synthesized from Ta₂O₅ and ZrO(NO₃)₂·2H₂O (Zr/Ta = 0.1 by mole) after nitridation for 15 h. Physicochemical analysis suggested that a lower density of anionic defects in TaON, which was realized using highly-dispersed ZrO₂ nanoparticles (10–30 nm in size), contributed primarily to the enhanced activity.

---

1. Introduction

Hydrogen, which may one day serve as a primary energy carrier, could alleviate the threat of global climate change and help avoid the undesirable effects of fossil fuel use. Photocatalytic water splitting has been widely studied as a potential means to produce H₂ from renewable resources.¹ To better utilize the available solar energy, photocatalysts that can operate under visible light irradiation (λ > 400 nm) are highly desirable.

Among the potential candidates for visible-light-induced overall water splitting, tantalum oxynitride (TaON), which has a band gap of 2.5 eV, suitable band edge positions, and high stability under photochemical reaction conditions, has been extensively studied in recent years.^2–8 Although TaON has a relatively high activity for O₂ evolution, its H₂ evolution activity is moderate.^2a,b This is at least partially caused by its high defect density, and the H₂ evolution activity can be improved by applying an appropriate preparation method.⁸ We recently reported that the feasibility of the modification of TaON with ZrO₂ can enhance the activity for H₂ evolution.^8a In particular, ZrO₂/TaON loaded with nanoparticulate Pt functions efficiently as an H₂ evolution component for a two-step water splitting system in combination with Pt/WO₃,^8b RuO₂/TaON,^8c or Ir/TiO₂/Ta₃N₅^8d as an O₂ evolution photocatalyst in the presence of an IO₃⁻/I⁻ redox mediator, exhibiting a much higher performance than an analogous Pt/TaON system.

In general, it is important for preparing a highly efficient photocatalytic material to examine both the preparation conditions and the physicochemical factors that determine the photocatalytic activity of a given photocatalyst.^9–18 The precursors of Zn–Ga and Zn–Ge oxynitrides, which are active photocatalysts for overall water splitting under visible light, can have a significant effect on their photocatalytic activity.^13b,14b In these cases, the use of inappropriate precursors resulted in failure to achieve overall water splitting. For a ZrO₂/TaON photocatalyst, however, the detailed relationship between the structure and activity of the material has yet to be investigated. We therefore investigated several factors affecting the photocatalytic activity of ZrO₂/TaON for H₂ production in order to obtain guidelines for the preparation of a highly active photocatalyst. In this paper, the physicochemical properties of ZrO₂/TaON and the optimal preparation conditions will be discussed. Strategies to develop a highly active ZrO₂/TaON photocatalyst were determined on the basis of structural analysis.

2. Experimental

2.1 Materials and reagents

For the preparation of ZrO₂/TaON, Ta₂O₅ (99.9%; High Purity Chemicals), ZrO(NO₃)₂·2H₂O (99%; Kanto Chemicals), ZrOCl₂·8H₂O (98%; High Purity Chemicals), ZrCl₄ (95+%; Wako Pure Chemicals), Zr(O-i-C₃H₇)₄ (99.9%; High Purity Chemicals), ZrO₂ (99.9%; Kanto Chemicals), and methanol (Reagent grade; Kanto Chemicals) were used as-received without further purification. H₂PtCl₆·6H₂O (98.5%; Kanto Chemicals) was used as a precursor for Pt nanoparticles, which were employed as cocatalysts on ZrO₂/TaON for photocatalytic H₂ evolution.

2.2 Synthesis of ZrO₂/TaON and loading Pt nanoparticles

Ta₂O₅ powder and a Zr precursor were mixed (Zr/Ta = 0.1 by mole) in a methanol solvent, dried, and subsequently heated in air at 1073 K for 2 h to yield a ZrO₂/Ta₂O₅ composite. ZrO₂/TaON materials were obtained through nitridation of the as-prepared ZrO₂/Ta₂O₅ powder at 1123 K under a flow of NH₃ (20 mL min⁻¹) for different durations (5–20 h). Prior to reaction, ZrO₂/TaON was modified with 0.5 wt% of Pt as a water reduction cocatalyst. The Pt cocatalyst was loaded on ZrO₂/TaON by impregnation from an H₂PtCl₆ aqueous solution followed by a hydrogen reduction treatment at 473 K for 1 h.

2.3 Characterization of ZrO₂/TaON

The prepared samples were studied by powder X-ray diffraction (XRD; RINT-UltimaIII + D/Tex Ultra UltimaIII detector, Rigaku; Cu Kα), transmission electron microscopy (HRTEM; JEM-2010F, JEOL), X-ray photoelectron spectroscopy (XPS; JPS-9000, JEOL), UV-visible diffuse reflectance spectroscopy (DRS; V-570, Jasco), and photoluminescence spectroscopy (PL; FP-6600, Jasco). The binding energies determined by XPS were corrected by reference to the C1s peak (284.6 eV) for each sample. The PL spectra were measured at liquid nitrogen temperature under excitation at 420 nm from a xenon lamp. The Brunauer–Emmett–Teller (BET) surface area was measured using a BELSORP-mini instrument (BET Japan) at liquid nitrogen temperature (77 K). The experimental error in the surface area measurement was ∼1 m² g⁻¹.

2.4 Photocatalytic reactions

Photocatalytic reactions were performed in a Pyrex side-irradiation reaction vessel connected to a glass closed gas circulation system at room temperature. 0.1 g of catalyst was dispersed in an aqueous methanol solution (10 vol%, 250 mL). The reaction vessel was evacuated several times to completely remove the air from the reaction system, followed by the introduction of argon (ca. 5 kPa) into the system. The reaction was initiated by irradiation with a 300 W xenon lamp fitted with a cutoff filter (λ > 420 nm). The evolved gases were analyzed by gas chromatography. The temperature of the reactant solution was typically 298–303 K.

The apparent quantum yield (AQY) for H₂ evolution was measured using the same experimental setup, but with a top-irradiation vessel, a band pass filter (λ = 420.5 nm), and a neutral density filter, and was estimated as:

AQY (%) = (2 × R/I) × 100 (1)

where R and I represent the rates of H₂ evolution and incident photons, respectively. The total number of incident photons (5.1 mW) was measured using a calibrated silicon photodiode.

3. Results and discussion

3.1 Effect of zirconium precursors in the starting material on the physicochemical and photocatalytic properties

ZrO₂/TaON catalysts were first prepared using various zirconium precursors under the same nitridation conditions (15 h). XRD patterns of the synthesized materials are shown in Fig. 1A. All of the prepared samples exhibited single-phase diffraction patterns identical to that of TaON, indicating a monoclinic baddeleyite structure similar to that of ZrO₂. There were no peaks attributable to Ta₂O₅, Ta₃N₅, or zirconium (oxy)nitride from any of the prepared samples. Although no noticeable change in the XRD patterns was observed among the prepared samples, there was a minor difference at 2θ = 28–29.5°, which corresponds to the main peak region of baddeleyite TaON and ZrO₂, as shown in Fig. 1B. A small peak assignable to monoclinic ZrO₂ was evident in samples prepared from Zr(O-i-C₃H₇)₄ and ZrO₂ precursors, while no ZrO₂ peaks were observed in samples derived from other precursors. This indicates that larger (or more aggregated) ZrO₂ particles were present on the surface of TaON prepared from Zr(O-i-C₃H₇)₄ and ZrO₂. Furthermore, in samples prepared from ZrO(NO₃)₂·2H₂O, Zr(O-i-C₃H₇)₄ and ZrO₂ precursors, the positions of the TaON diffraction peaks were slightly shifted to lower 2θ angles. This peak shift probably resulted from the incorporation of a small amount of ZrO₂ species into the TaON lattice (i.e., the formation of a solid solution between the two), considering the relatively large a and b lattice constants of ZrO₂ (a = 5.3129 Å, b = 5.2125 Å, c = 5.1471 Å)¹⁹ compared to those of TaON (a = 4.9494 Å, b = 5.0166 Å, c = 5.1642 Å),^3c with the lattice mismatch being calculated to be ca. 7%. It has been also reported that ZrO₂ and TaON can form a solid solution (Zr_xTa_1−xO_1+xN_1−x) with a baddeleyite structure over the compositional range of 0 < x < 0.28.²⁰

Fig. 1 (A) X-Ray diffraction patterns of ZrO₂/TaON catalysts synthesized from different Zr precursors with a common nitridation time of 15 h. (B) An enlarged view of panel (A).

Fig. 2 shows typical TEM images of ZrO₂/TaON samples prepared from ZrO(NO₃)₂·2H₂O and Zr(O-i-C₃H₇)₄, which, according to the XRD results, contained less and more aggregated ZrO₂ particles, respectively. As shown in Fig. 2A, the ZrO(NO₃)₂·2H₂O sample consisted of larger base TaON particles and loaded ZrO₂ nanoparticles with a primary particle size of ca. 10 nm, although some of the ZrO₂ nanoparticles aggregated to form large secondary particles (∼30 nm). TEM-EDS analysis indicated that the loaded nanoparticles were crystalline, as suggested by the lattice fringes. On the other hand, significant aggregation of ZrO₂ particles was observed on the TaON component in a sample derived from Zr(O-i-C₃H₇)₄, as shown in Fig. 2B. These results are consistent with the results of XRD measurements (Fig. 1).

Fig. 2 TEM images of ZrO₂/TaON catalysts synthesized from (A) ZrO(NO₃)₂·2H₂O and (B) Zr(O-i-C₃H₇)₄ with a common nitridation time of 15 h.

Further analysis was conducted using XPS to identify the changes in the valence states of zirconium and tantalum. The binding energies of Zr3d and Ta4f peaks of the prepared samples were almost the same as those of ZrO₂ and TaON references (Fig. S1, ESI†). These results further support the idea that the zirconium and tantalum in the prepared samples were close to ZrO₂ and TaON, respectively. Although a slight incorporation of ZrO₂ species into TaON has been observed in samples prepared from ZrO(NO₃)₂·2H₂O, Zr(O-i-C₃H₇)₄ and ZrO₂ precursors (Fig. 1), it did not strongly affect the valence states of the zirconium and tantalum.

UV-visible DRS for the same set of samples are shown in Fig. 3. A steep absorption occurred at 520 nm in each sample, which corresponds to electron transitions from hybridized N2p and O2p orbitals to empty Ta5d orbitals,²¹ and resembles the spectral feature of TaON. Consistent with our previous study,⁸ however, ZrO₂/TaON samples display a lower absorption band in the longer wavelength region, which is derived from the defect sites, as compared to unmodified TaON. As we discussed previously,^8a during nitridation of tantalum oxide precursors, the reduction of Ta⁵⁺ cations in TaON and/or the oxide precursor generates reduced tantalum species (e.g., Ta³⁺ and Ta⁴⁺), resulting in the formation of defect sites due to unbalance of cation–anion stoichiometry. On the other hand, when ZrO₂ is loaded on the surface of Ta₂O₅, tantalum cations at the interface between Ta₂O₅ (and/or TaON) and the loaded ZrO₂ would interact with ZrO₂ and thereby become more cationic. As a result, the formation of reduced tantalum species on the TaON surface is considered to be suppressed by prior loading with nanoparticulate ZrO₂. The results of DRS analyses thus indicate that upon modification with ZrO₂, the original band gap structure of TaON remains largely unchanged, but the defect density is reduced. The above-mentioned idea that the reduction of tantalum species at the TaON surface during nitridation can be suppressed by modification with ZrO₂ seems to contradict the results of XPS analysis, which showed that the Ta4f peak of the composite sample appeared at the same position as that of TaON (Fig. S1, ESI†). Presumably, the change in the state of the surface tantalum species between the two samples was too small to be detected by XPS measurements.

Fig. 3 (A) UV-visible diffuse reflectance spectra of ZrO₂/TaON catalysts synthesized from different Zr precursors with a common nitridation time of 15 h. (B) An enlarged view of panel (A).

Table 1 lists the rates of visible-light-driven H₂ evolution achieved using samples modified by Pt nanoparticles as H₂ evolution sites. The specific surface areas are also shown for comparison. H₂ evolution was observed in all cases. The H₂ evolution rate was dependent on the zirconium precursors employed, but was relatively robust with respect to the specific surface area. The fastest rate was obtained with a sample prepared using ZrO(NO₃)₂·2H₂O. Furthermore, some of the samples had a lower rate of H₂ evolution than unmodified TaON. These results demonstrate that an appropriate choice of zirconium precursor is important for enhancing the water reduction activity of TaON. No H₂ was produced under dark conditions or without a photocatalyst in the presence of light.

Table 1 Rates of H₂ evolution from ZrO₂/TaON catalysts synthesized from different Zr precursors with a common nitridation time of 15 h^a

Entry	Precursor	Rate of H2 evolution/μmol h^−1	Specific surface area/m^2 g^−1
a Reaction conditions: catalyst, 0.1 g (0.5 wt% Pt-loaded); aqueous methanol solution, 250 mL (10 vol%); light source, xenon lamp (300 W) fitted with a cold mirror (CM-1); reaction vessel, Pyrex side-irradiation type; irradiation wavelength, 420 < λ < 800 nm.
1	ZrO(NO3)2·2H2O	44.5 ± 3.5	4.1
2	ZrOCl2·8H2O	38.2	5.1
3	ZrCl4	30.3	4.5
4	Zr(O-i-C3H7)4	31.8	4.7
5	ZrO2	25.3	3.6
6	None	36.1	4.4

---

3.2 Effect of nitridation time on physicochemical and photocatalytic properties

The effect of changing the NH₃ exposure time was examined using ZrO₂/Ta₂O₅ samples prepared from ZrO(NO₃)₂·2H₂O, the most suitable precursor. Fig. 4A shows XRD patterns of the corresponding nitrided products of ZrO₂/TaON samples with nitridation times of 5, 10, 15, and 20 h. All of the samples exhibited single-phase diffraction patterns assignable to TaON, with the exception of the sample with a 5 h nitridation time, for which additional Ta₂O₅ peaks were observable. With nitridation times longer than 10 h, a single-phase diffraction pattern assigned to TaON was formed. With a further increase in nitridation time, the TaON peaks became sharper, indicating improved crystallinity. This indicates that sufficient exposure time in an NH₃ atmosphere is needed for the complete formation of TaON from Ta₂O₅. Furthermore, there was no evidence of impurity phases in the XRD patterns.

Fig. 4 (A) X-Ray diffraction patterns and (B) UV-visible diffuse reflectance spectra of ZrO₂/TaON catalysts synthesized from the ZrO(NO₃)₂·2H₂O precursor with different nitridation times.

Fig. 4B shows DRS for the same set of samples, along with TaON (prepared with 15 h of nitridation) data for comparison. The steep absorption band around 520 nm due to the band gap transition of TaON was similar for all samples. The weaker absorption band observed at longer wavelengths in the spectra decreased with increasing nitridation time, indicating that an increase in exposure time to NH₃ led to the production of fewer defects. It is noteworthy that the 5 h and 10 h samples exhibited a higher intensity than unmodified TaON prepared with 15 h of nitridation. This suggests the presence of a higher density of defects in those two samples, compared to that of the TaON. These results are in good agreement with the XRD results.

Table 2 lists the H₂ evolution rates from an aqueous 10 vol% methanol solution containing Pt/ZrO₂/TaON catalysts nitrided for different durations under visible light. The H₂ evolution rate was remarkably enhanced with an increase in nitridation time, reaching a maximum at 15 h, beyond which it began to decrease slightly. As a result, the ZrO₂/TaON sample with a 15 h nitridation time exhibited the highest activity, with a H₂ production rate of 44.5 ± 3.5 μmol h⁻¹ and a corresponding AQY of ca. 1.7% at 420 nm. Neither N₂ nor O₂ were evolved during the reaction. Although the optimal ZrO₂/TaON functioned as a good H₂ evolution photocatalyst, the H₂ evolution activities of samples prepared with less than 10 h of nitridation were lower than that of unmodified TaON prepared for 15 h. These activity tests confirm the dependence of H₂ evolution activity on the nitridation time. Under UV irradiation (λ > 300 nm), 1068 μmol of H₂ was produced over the optimized catalyst after 18 h of irradiation. This amount far exceeds the amount of catalyst used (ca. 450 μmol), confirming the catalytic cycle of this reaction. Furthermore, no noticeable change was observed in XRD patterns before and after reaction, indicating stability of this material (Fig. S2, ESI†).

Table 2 Rates of H₂ evolution from ZrO₂/TaON catalysts synthesized from the ZrO(NO₃)₂·2H₂O precursor with different nitridation times^a

Entry	Nitridation time/h	Rate of H2 evolution/μmol h^−1	Specific surface area/m^2 g^−1
a Reaction conditions: catalyst, 0.1 g (0.5 wt% Pt-loaded); aqueous methanol solution, 250 mL (10 vol%); light source, xenon lamp (300 W) fitted with a cold mirror (CM-1); reaction vessel, Pyrex side-irradiation type; irradiation wavelength, 420 < λ < 800 nm.
1	5	2.6	5.4
2	10	26.4	5.6
3	15	44.5 ± 3.5	4.1
4	20	41.0	5.2
5	15 (TaON)	36.1	4.4

---

As reported previously, the optimized ZrO₂/TaON loaded with nanoparticulate Pt also achieved stoichiometric water splitting into H₂ and O₂ under visible light, with an AQY of 6.3% under irradiation by 420 nm monochromatic light under optimal reaction conditions, when combined with Pt-loaded WO₃ as an O₂ evolution photocatalyst and an iodate/iodide redox couple. This was 6 times greater than the yield achieved using a TaON analog.^8b

3.3 Strategy to improve the water reduction activity of ZrO₂/TaON photocatalyst

As shown in Table 1, the zirconium precursor was an important factor contributing to the enhancement of the photocatalytic activity of ZrO₂/TaON for H₂ evolution, and ZrO(NO₃)₂·2H₂O was the most suitable among the precursors examined. The ZrO(NO₃)₂·2H₂O-derived sample contained relatively well-dispersed ZrO₂ nanoparticles of 10–30 nm in size that were loaded on the TaON component, while samples prepared from either Zr(O-i-C₃H₇)₄ or ZrO₂ did not (Fig. 2). Therefore, the loaded ZrO₂ nanoparticles derived from ZrO(NO₃)₂·2H₂O make contact with the TaON surface more effectively than those from Zr(O-i-C₃H₇)₄ and ZrO₂. As a result, the generation of defects in the ZrO(NO₃)₂·2H₂O-derived sample is effectively suppressed during nitridation, as can be seen in the UV-visible DRS results (Fig. 3). The activities of samples prepared using such inappropriate precursors were lower than that of unmodified TaON (Table 1), even though they exhibited a lower background level of the UV-visible spectra than unmodified TaON. We have revealed that one of the positive effects of ZrO₂-modification of TaON is the reduction of defect site generation during nitridation, while one negative effect is a hindrance of active reaction sites on the surface of TaON by excessively loaded ZrO₂.^8a This negative effect can become dominant in samples prepared using inappropriate precursors. While the size of ZrO₂ modifiers in ZrOCl₂- and ZrCl₄-derived samples was similar to that in the most active ZrO(NO₃)₂·2H₂O-derived sample (Fig. S3, ESI†), however, they were less active than the ZrO(NO₃)₂·2H₂O-derived sample. XPS analyses revealed that these samples contained chlorine species, which might have a negative impact on photocatalytic activity. We also performed TEM observations for samples prepared from ZrO(NO₃)₂·2H₂O and ZrO₂, which were, respectively, most active and inactive (Table 1). The observations showed that the size of Pt nanoparticles in these samples was smaller than 2 nm in size (Fig. S4, ESI†), indicating that the activity-difference is not related to the Pt particle size.

Changing the nitridation time from 5 to 20 h while using the same precursor (ZrO(NO₃)₂·2H₂O), a volcano-type relationship between nitridation time and activity was obtained, as shown in Table 2. XRD results indicated that structural conversion from Ta₂O₅ to TaON occurred at 5 h of nitridation, and durations longer than 10 h resulted in the disappearance of the Ta₂O₅ phase, leaving only TaON peaks that became slightly sharper with further increases in nitridation time (Fig. 4A). The 5 h sample also had a stronger background level at longer wavelengths than the other samples (Fig. 4B), indicating that the density of defects was largest in that sample. Judging from the XRD pattern, it would be natural to consider that the 5 h sample was insufficiently nitrided and less crystallized. In such a situation, there would be a considerable number of anion defects in the sample, as suggested by the UV-visible DRS results. With increasing nitridation time, the background level in the DRS decreased, suggesting the suppression of defect site generation. This is reasonable, because structural imperfections can be healed by NH₃ annealing.^{13,14a,15–17} Therefore, the origin of the background level was probably anion defects in the TaON. Furthermore, the increase in activity with nitridation time was associated with a reduction of defect sites acting as recombination centres between photogenerated electrons and holes.

It is possible to use photoluminescence spectroscopy to evaluate the fate of photogenerated charge carriers in a semiconductor photocatalyst.²² Our previous study revealed that ZrO₂/TaON exhibits an enhanced photoluminescence band centred at ca. 690 nm, compared to that of TaON.^8b We attributed this result to the reduced density of nonradiative recombination sites in ZrO₂/TaON. In a similar manner, the photoluminescence spectra of samples nitrided for different durations but with the same precursor (ZrO(NO₃)₂·2H₂O) were acquired. As shown in Fig. S5 (ESI†), all samples exhibited a luminescence band at ca. 690 nm upon photoexcitation at 420 nm. However, the luminescence intensity differed, suggesting different densities of defect sites in these materials. In samples prepared using different zirconium precursors, we observed similar luminescence but different intensities, as shown in Fig. S6 (ESI†). Fig. 5 shows the relationship between the luminescence intensity and activity of ZrO₂/TaON. It is clear that there is a linear relationship between the two factors, regardless of precursors used and nitridation time.

Fig. 5 Relationship between the rate of H₂ evolution and photoluminescence intensity of ZrO₂/TaON catalysts synthesized (A) at different nitridation times with the same ZrO(NO₃)₂·2H₂O precursor and (B) from different Zr precursors with a common nitridation time of 15 h. Photoluminescence intensities were normalized with respect to the most intense spectrum in Fig. S5 and S6 (ESI†).

These results indicate that suppressing nonradiative recombination sites in the TaON component is an effective strategy to improve the water reduction activity of ZrO₂/TaON, and this can be accomplished by refining the preparation conditions. This is clearly different from the general trend in the conventional metal-oxide photocatalysts for H₂ evolution, where the activity is dependent primarily on the crystallinity and particle size of the material, as determined by the preparation conditions.^9,11,18 In the case of non-oxide photocatalysts like ZrO₂/TaON, reducing the density of defects such as anionic vacancies appears to be important, and differentiates the non-oxide systems from the metal-oxide systems, resulting in their unique photocatalytic properties.

4. Conclusions

Using various zirconium precursors, ZrO₂/TaON photocatalysts were prepared under different conditions, and the factors affecting photocatalytic H₂ evolution were investigated. Using a suitable precursor of the ZrO₂ modifier was found to be essential for the suppression of anionic vacancies during nitridation. This was realized by forming a better dispersion of the ZrO₂ modifier on the TaON surface without leaving impurities such as chlorine species. Physicochemical analysis revealed that the activity was dependent on how effectively nonradiative recombination centres were suppressed, but was largely insensitive to the specific surface area of the catalysts.

Acknowledgements

This work was supported by a PRESTO/JST program (Chemical conversion of light energy), the Research and Development in a New Interdisciplinary Field Based on Nanotechnology and Materials Science program of the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan, and The KAITEKI Institute, Inc. is also extended to the Global Center of Excellence (GCOE) Program for Chemistry Innovation through Cooperation of Science and Engineering and Nippon Sheet Glass Foundation for Materials Science and Engineering.

Notes and references

1. (a) J. S. Lee, Catal. Surv. Asia, 2005, 9, 217; (b) K. Maeda and K. Domen, J. Phys. Chem. Lett., 2010, 1, 2655; (c) A. Kudo and Y. Miseki, Chem. Soc. Rev., 2009, 38, 253; (d) W. J. Youngblood, S.-H. A. Lee, K. Maeda and T. E. Mallouk, Acc. Chem. Res., 2009, 42, 1966.
2. (a) G. Hitoki, T. Takata, J. N. Kondo, M. Hara, H. Kobayashi and K. Domen, Chem. Commun., 2002, 1698; (b) M. Hara, G. Hitoki, T. Takata, J. N. Kondo, H. Kobayashi and K. Domen, Catal. Today, 2003, 78, 555; (c) M. Hara, J. Nunoshige, T. Takata, J. N. Kondo and K. Domen, Chem. Commun., 2003, 3000; (d) M. Hara, T. Takata, J. N. Kondo and K. Domen, Catal. Today, 2004, 90, 313.
3. (a) W. A. Chun, A. Ishikawa, H. Fujisawa, T. Takata, J. N. Kondo, M. Hara, M. Kawai, Y. Matsumoto and K. Domen, J. Phys. Chem. B, 2003, 107, 1798; (b) M. Hara, E. Chiba, A. Ishikawa, T. Takata, J. N. Kondo and K. Domen, J. Phys. Chem. B, 2003, 107, 13441; (c) M. Yashima, Y. Lee and K. Domen, Chem. Mater., 2007, 19, 588.
4. S. Ito, R. Thampi, P. Comte, P. Liska and M. Grätzel, Chem. Commun., 2005, 268.
5. (a) R. Abe, T. Takata, H. Sugihara and K. Domen, Chem. Commun., 2005, 3829; (b) R. Abe, M. Higashi and K. Domen, ChemSusChem, 2011, 4, 228.
6. R. Nakamura, T. Tanaka and Y. Nakato, J. Phys. Chem. B, 2005, 109, 8920.
7. (a) R. Abe, T. Takata, H. Sugihara and K. Domen, Chem. Lett., 2005, 34, 1162; (b) R. Abe, M. Higashi and K. Domen, J. Am. Chem. Soc., 2010, 132, 11828.
8. (a) K. Maeda, H. Terashima, K. Kase, M. Higashi, M. Tabata and K. Domen, Bull. Chem. Soc. Jpn., 2008, 81, 927; (b) K. Maeda, M. Higashi, D. Lu, R. Abe and K. Domen, J. Am. Chem. Soc., 2010, 132, 5858; (c) K. Maeda, R. Abe and K. Domen, J. Phys. Chem. C, 2011, 115, 3057; (d) M. Tabata, K. Maeda, M. Higashi, D. Lu, T. Takata, R. Abe and K. Domen, Langmuir, 2010, 26, 9161.
9. A. Kudo, A. Tanaka, K. Domen and T. Onishi, J. Catal., 1988, 111, 296.
10. S. Ikeda, M. Hara, J. N. Kondo, K. Domen, H. Takahashi, T. Okubo and M. Kakihana, Chem. Mater., 1998, 10, 72.
11. H. Kominami, S. Murakami, J. Kato, Y. Kera and B. Ohtani, J. Phys. Chem. B, 2002, 106, 10501.
12. J. Sato, H. Kobayashi, K. Ikarashi, N. Saito, H. Nishiyama and Y. Inoue, J. Phys. Chem. B, 2004, 108, 4369.
13. (a) K. Maeda, K. Teramura, T. Takata, M. Hara, N. Saito, K. Toda, Y. Inoue, H. Kobayashi and K. Domen, J. Phys. Chem. B, 2005, 109, 20504; (b) K. Maeda and K. Domen, Chem. Mater., 2010, 22, 612.
14. (a) Y. Lee, K. Teramura, M. Hara and K. Domen, Chem. Mater., 2007, 19, 2120; (b) F. Tessier, P. Maillard, Y. Lee, C. Bleugat and K. Domen, J. Phys. Chem. C, 2009, 113, 8526.
15. K. Maeda, N. Saito, Y. Inoue and K. Domen, Chem. Mater., 2007, 19, 4092.
16. T. Hisatomi, K. Teramura, J. Kubota and K. Domen, Bull. Chem. Soc. Jpn., 2008, 81, 1647.
17. K. Maeda, B. Lee, D. Lu and K. Domen, Chem. Mater., 2009, 21, 2286.
18. K. Maeda, M. Eguchi, W. J. Youngblood and T. E. Mallouk, Chem. Mater., 2009, 21, 3611.
19. F. J. Torres, J. M. Amigó and J. Alarcó, J. Solid State Chem., 2003, 173, 40.
20. J. Grins, P.-O. Käll and G. Svensson, J. Mater. Chem., 1994, 4, 1293.
21. C. M. Fang, E. Orhan, G. A. Wijs, H. T. Hintzen, R. A. Groot, R. Marchand, J. Y. Saillard and G. With, J. Mater. Chem., 2001, 11, 1248.
22. J. Y. Shi, J. Chen, Z. C. Feng, T. Chen, Y. X. Lian, X. L. Wang and C. Li, J. Phys. Chem. C, 2007, 111, 693.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c2cy00499b

---