Solar water splitting over Rh_0.5Cr_1.5O₃-loaded AgTaO₃ of a valence-band-controlled metal oxide photocatalyst

Abstract

Improvement of water splitting performance of AgTaO₃ (BG 3.4 eV) of a valence-band-controlled photocatalyst was examined. Survey of cocatalysts revealed that a Rh_0.5Cr_1.5O₃ cocatalyst was much more effective than Cr₂O₃, RuO₂, NiO and Pt for water splitting into H₂ and O₂ in a stoichiometric amount. The optimum loading amount of the Rh_0.5Cr_1.5O₃ cocatalyst was 0.2 wt%. The apparent quantum yield (AQY) at 340 nm of the optimized Rh_0.5Cr_1.5O₃(0.2 wt%)/AgTaO₃ photocatalyst reached to about 40%. Rh_0.5Cr_1.5O₃(0.2 wt%)/AgTaO₃ gave a solar to hydrogen conversion efficiency (STH) of 0.13% for photocatalytic water splitting under simulated sunlight irradiation. Bubbles of gasses evolved by the solar water splitting were visually observed under atmospheric pressure at room temperature.

---

Introduction

Photocatalytic solar water splitting is an important reaction because it has the potential for solving resources, energy and environmental issues. Hence, many researchers have aimed to develop water splitting systems of a one-step photoexcitation type and a two-step photoexcitation type (Z-scheme) with visible-light-driven powdered photocatalysts.^1–6 The simplicity of the one-step photoexcitation type will be an advantage in practical use. Domen and co-workers have reported that (oxy)nitride and oxysulfide photocatalysts such as GaN–ZnO,^7,8 ZnGeN₂–ZnO,^9,10 LaMg_1/3Ta_2/3O₂N,^11–13 TaON,¹⁴ CaTaO₂N,¹⁵ Ta₃N₅ (ref. 16) and Y₂Ti₂O₅S₂ (ref. 17) show activities for one-step photoexcitation type water splitting under visible light irradiation. We have also reported that Rh,Sb-codoped SrTiO₃ of a metal oxide photocatalyst shows the water splitting activity under visible light irradiation by loading of an IrO₂ cocatalyst.¹⁸ Although above visible-light-responsive photocatalysts split water under sunlight irradiation, their solar energy conversion efficiencies (solar to hydrogen conversion efficiency, STH) do not reach the level for practical use due to low apparent quantum yields (AQY). In terms of the high AQY, NiO-loaded La-doped NaTaO₃ (ref. 19) (BG 4.1 eV), Rh_0.5Cr_1.5O₃-loaded Zn,Ca-codoped Ga₂O₃ (ref. 20) (BG 4.6 eV) and MoO_y,RhCrO_x-coloaded Al-doped SrTiO₃ (ref. 21) (BG 3.2 eV) show the AQYs of 56% at 270 nm, 71% at 254 nm and 69% at 365 nm, for photocatalytic water splitting under UV irradiation, respectively. However, only Al-doped SrTiO₃ can respond to sunlight (λ > 300 nm) with STH of 0.4%.²² Thus, the development of sunlight-driven photocatalysts with one-step photoexcitation for water splitting is an important research topic.

We have reported that a AgTaO₃ photocatalyst (BG 3.4 eV) shows water splitting activity under UV irradiation by loading of a NiO cocatalyst.²³ The characteristics of AgTaO₃ is the valence band maximums formed by Ag 4d orbitals which are located above the bands formed by O 2p orbitals, and hence the band gap of AgTaO₃ is relatively narrow among the metal oxides containing Ta. AgTaO₃ is expected to respond to sunlight judging from the band gap of 3.4 eV, while the water splitting under sunlight irradiation has not been achieved yet. Moreover, the crystal structure of perovskite for AgTaO₃ is the same as that for SrTiO₃ and NaTaO₃ photocatalysts with high activities for water splitting. These backgrounds motivate us to investigate solar water splitting using the AgTaO₃ photocatalyst.

A cocatalyst is one of the most effective factors for improvement of the photocatalytic activity. Actually, the water splitting activity of AgTaO₃ is drastically improved by loading of a NiO cocatalyst which is widely used in water splitting over metal oxides.²³ Recently, Rh_0.5Cr_1.5O₃ has also arisen as a potential candidate cocatalyst for water splitting over metal oxides.^20,21 In the present study, we investigated the loading effect of various cocatalysts on water splitting over AgTaO₃. Solar water splitting was also demonstrated using the AgTaO₃ with the optimized cocatalyst.

Experimental

Preparation of photocatalysts

AgTaO₃ was synthesized by a solid-state reaction using Ag₂O (Kanto Chemical; 99.0%) and Ta₂O₅ (Rare Metallic; 99.99%) as starting materials. The starting materials were mixed in an alumina mortar in a ratio of Ag : Ta = 1 + x : 1 (x = 0, 0.03, 0.05, 0.07). The mixture was calcined in air at 1173–1373 K for 5–20 h in an alumina crucible. The calcined materials were washed with 1 mol L⁻¹ of an aqueous HNO₃ solution in ultrasonication for 1 h in order to remove Ag metals deposited on the surface of AgTaO₃.

A Pt cocatalyst was loaded in situ by a photodeposition method using an aqueous H₂PtCl₆ (Tanaka Kikinzoku; 37.7 wt% as Pt) solution. NiO,²³ RuO₂ (ref. 24) and Cr₂O₃ (ref. 25) of cocatalysts were loaded on AgTaO₃ photocatalysts by an impregnation method using Ni(NO₃)₂·6H₂O (Wako Pure Chemical; 98.0%), Ru₃(CO)₁₂ (Aldrich; 99%), Cr(NO₃)₃·9H₂O (Kanto Chemical; 98.0–103.0%). AgTaO₃ powder was suspended in aqueous solutions dissolving Ni(NO₃)₂ or Cr(NO₃)₃, and an acetone solution dissolving Ru₃(CO)₁₂ in porcelain crucibles. The AgTaO₃-suspended solutions were dried up on a water bath and subsequently calcined in an electric furnace with conditions of 543 K – 1 h for NiO, 673 K – 2 h for RuO₂ and 623 K – 1 h for Cr₂O₃. Rh_0.5Cr_1.5O₃ (ref. 25) cocatalyst was loaded by a simple impregnation method using Cr(NO₃)₃·9H₂O and Rh(NO₃)₃ (Kanto Chemical; >80.0% as anhydrous). AgTaO₃-suspended aqueous solution containing both Rh(NO₃)₃ and Cr(NO₃)₃ was dried up and subsequently calcined at 623 K for 1 h. It is reported that Rh_0.5Cr_1.5O₃ of a mixed oxide is naturally formed on a photocatalyst by this process.²⁶ The loading amount of Rh_0.5Cr_1.5O₃ was controlled by adjusting the amount of starting materials.

Characterization

The crystal structure of the synthesized AgTaO₃ was identified by powder X-ray diffraction (Rigaku; MiniFlex600). The elemental composition of the synthesized AgTaO₃ was measured by an X-ray fluorescence spectrometer (Rigaku; NEX DE). A Diffuse reflectance spectrum was measured using a UV-vis-NIR spectrometer with an integrating sphere (Jasco; UbeatV-570) and was converted from reflection to absorbance mode by the Kubelka–Munk method. Morphologies of photocatalyst particles and Rh_0.5Cr_1.5O₃-cocatalysts were observed using a scanning electron microscope (JEOL; JSM-7400F) and a transmission electron microscope (JEOL; JEM-2100F). A chemical state of Rh in Rh_0.5Cr_1.5O₃-loaded AgTaO₃ was analyzed using an X-ray photoelectron spectroscopy (JEOL; JPS-9010MC).

Photocatalytic reaction

Photocatalytic water splitting was carried out in a gas-closed-circulation system. Photocatalyst powder (0.3 g) was dispersed in distilled water (120 mL) in a top-irradiation cell with a Pyrex window. A 300 W Xe-arc lamp (PerkinElmer; Cermax-PE300BF) and a solar simulator (Asahi spectra; HAL-320) were employed as a light source. An inner-irradiation reaction cell made of quartz equipped with a 400 W high-pressure Hg lamp (SEN; HL400EH-5) was also used for photocatalytic water splitting in order to compare with the activity of NiO/NaTaO₃:La.¹⁹ Amounts of evolved H₂ and O₂ gases were determined with a gas chromatograph (Shimadzu; GC-8A, MS-5A, Ar carrier gas, TCD). Turnover number (TON) of the amount of reacted electrons/holes to the molar quantity of AgTaO₃ was estimated using the eqn (1).

[TON] = [the molar quantity of reacted electrons]/[the molar quantity of AgTaO₃] = [(the amount of evolved H₂) × 2/mol]/[the molar quantity of AgTaO₃/mol] (1)

Apparent quantum yields (AQY) were measured using a 300 W Xe-arc lamp (Asahi Spectra; MAX-302) with band-pass filters (Asahi Spectra). The photon flux of the monochromatic light through the long-pass filters was measured using a silicon diode head (OPHIR; PD300-UV head and NOVA display). An AQY and a solar to hydrogen conversion efficiency (STH) were estimated using the following eqn (2) and (3).

[AQY%] = 100 × [the number of reacted electrons or holes]/[the number of incident photons] = 100 × [(the number of evolved H₂ molecules) × 2]/[the number of incident photons] (2)

[STH%] = 100 × ([ΔG⁰(H₂O)/kJ mol⁻¹] × [rate of H₂ evolution/μmol h⁻¹])/([irradiation time/s] × [solar energy (AM1.5G)/mW cm⁻²] × [irradiation area/cm²]) = (237 × [rate of H₂ evolution/μmol h⁻¹])/(3600 × 100 × 25) × 100 (3)

Results and discussion

XRD measurement confirmed that trigonal AgTaO₃ with a perovskite structure was successfully synthesized in a single phase as previously reported (Fig. S1†).²³ The peaks due to metallic Ag were not observed in XRD patterns of AgTaO₃ both before and after washing with an aqueous HNO₃ solution. However, the absorption due to the surface plasmonic resonance of Ag was observed in diffuse reflectance spectra of the AgTaO₃ before and after the washing (Fig. S2†). The absorption due to the surface plasmonic resonance after the washing was a little bit smaller than that before the washing. Therefore, metallic Ag on the surface would be removed by the washing but the amount of removed Ag was quite small. Actually, atomic ratios of Ag to Ta in the AgTaO₃ after the washing were calculated to be 1.00 and 0.99 from XPS and XRF measurements, respectively. These results indicate that the ratios of Ag to Ta on the surface and in the bulk were almost stoichiometric even if after the washing. The absorption edge of AgTaO₃ was around 380 nm in a diffuse reflectance spectrum (Fig. S3†), indicating 3.4 eV of the band gap. Scanning electron microscope observation revealed that the size of AgTaO₃ particles was from several hundreds nm to several μm (Fig. S4†).

Table 1 shows the activities for photocatalytic water splitting over AgTaO₃ loaded with various cocatalysts under UV irradiation. Non-loaded AgTaO₃ produced only a small amount of H₂ without O₂ evolution. In other words, water splitting did not proceed. In contrast, all AgTaO₃ loaded with various cocatalysts produced both H₂ and O₂. However, the amount of evolved H₂ was more than a stoichiometric amount when NiO, RuO₂ and Pt were loaded. When Cr₂O₃ and Rh_0.5Cr_1.5O₃ were loaded, the stoichiometric amounts of H₂ and O₂ evolved, indicating that ideal water splitting proceeded. In particular, the activity of AgTaO₃ was greatly improved by loading of the Rh_0.5Cr_1.5O₃ cocatalyst. This result is appropriate because Rh_0.5Cr_1.5O₃ is known as an effective cocatalyst for photocatalytic water splitting. It is reported that Cr₂O₃ does not work as a cocatalyst for photocatalytic water splitting but Ag coloaded with Cr works as a cocatalyst.²⁵ In the present case, only Cr₂O₃-loaded AgTaO₃ showed the water splitting activity. This is probably because a small amount of metallic Ag still remained on the surface of AgTaO₃ even after washing with an aqueous HNO₃ solution and a Ag–Cr cocatalyst was formed by loading of Cr₂O₃ as an effective cocatalyst. We optimized the loading amount of the Rh_0.5Cr_1.5O₃ cocatalyst and synthesis conditions of AgTaO₃. The water splitting activity of AgTaO₃ slightly depended on synthesis conditions and 1273 K – 15 h was the best condition for the activity (Table S1†). In addition, the activity was insensitive for addition of an excess amount of Ag. On the other hand, the activity much depended on the loading amount of Rh_0.5Cr_1.5O₃ cocatalyst and 0.2 wt% was optimum (Table 1). The optimized Rh_0.5Cr_1.5O₃(0.2 wt%)/AgTaO₃ stably and continuously produced H₂ and O₂ using a Xe lamp (Fig. 1).

Table 1 Photocatalytic water splitting over AgTaO₃ loaded with various cocatalysts under UV irradiation^a

Cocatalyst (wt%)	Loading method	Activity/μmol h^−1
		H2	O2
a Photocatalyst: 0.3 g, reactant solution: distilled water (120 mL), cell: top-irradiation cell with a Pyrex window, light source: 300 W Xe-arc lamp (λ > 300 nm).
None	—	0.5	0
Rh0.5Cr1.5O3 (0.05)	Impregnation (623 K – 1 h)	56	28
Rh0.5Cr1.5O3 (0.1)	Impregnation (623 K – 1 h)	318	162
Rh0.5Cr1.5O3 (0.2)	Impregnation (623 K – 1 h)	400	192
Rh0.5Cr1.5O3 (0.3)	Impregnation (623 K – 1 h)	217	111
Cr2O3 (0.13)	Impregnation (623 K – 1 h)	1.8	0.9
RuO2 (0.2)	Impregnation (673 K – 2 h)	1.6	0.5
NiO (0.2)	Impregnation (543 K – 1 h)	18	4
Pt (0.2)	Photodeposition (in situ)	45	4

---

Fig. 1 Photocatalytic water splitting over Rh_0.5Cr_1.5O₃(0.2 wt%)/AgTaO₃. AgTaO₃ was synthesized by a solid state reaction at 1273 K for 15 h without excess Ag. Rh_0.5Cr_1.5O₃ was loaded by an impregnation method at 623 K for 1 h. Photocatalyst: 0.3 g, reactant solution: distilled water (120 mL), cell: top-irradiation cell with a Pyrex window, light source: 300 W Xe-arc lamp (λ > 300 nm).

Chemical and physical states of the loaded Rh_0.5Cr_1.5O₃ in Rh_0.5Cr_1.5O₃(0.2 wt%)/AgTaO₃ before and after water splitting were characterized using a transmission electron microscope (TEM) and an X-ray photoelectron spectroscopy (XPS) because the water splitting activity of AgTaO₃ was greatly improved by loading a Rh_0.5Cr_1.5O₃ cocatalyst. It was confirmed by TEM that Rh_0.5Cr_1.5O₃ particles with a diameter of about 10–20 nm were loaded on AgTaO₃ (Fig. 2), as observed for GaN–ZnO.²⁶ In XPS spectra, the binding energy of the Rh 3d_5/2 peak for Rh_0.5Cr_1.5O₃(0.2 wt%)/AgTaO₃ was 309.0 eV (Fig. S5†) which is close to not that for Rh₂O₃ (308.3 eV) but that for Rh_0.5Cr_1.5O₃ (308.9 eV).²⁶ The slightly positive shift in the binding energy was also observed for the Rh_0.5Cr_1.5O₃ loaded on GaN–ZnO (309.1 eV).²⁶ These results indicate that the loaded Rh_0.5Cr_1.5O₃ on AgTaO₃ in the present study was similar to that on GaN–ZnO. Additionally, the TEM image of Rh_0.5Cr_1.5O₃ cocatalysts did not change so much after photocatalytic water splitting. On the other hand, in the XPS spectra, the Rh 3d_5/2 peak for Rh_0.5Cr_1.5O₃(0.2 wt%)/AgTaO₃ after photocatalytic water splitting was broadened to lower binding energy than that before photocatalytic water splitting. This observation indicated that the photocatalyst was activated by partial reduction of Rh in loaded Rh_0.5Cr_1.5O₃ cocatalyst to the metallic form during the induction period in Fig. 1.

Fig. 2 TEM images of Rh_0.5Cr_1.5O₃(0.2 wt%)/AgTaO₃ (a) before and (b) after photocatalytic water splitting under UV irradiation using a 300 W Xe-arc lamp. AgTaO₃ was synthesized by a solid state reaction at 1273 K for 15 h without excess Ag. Rh_0.5Cr_1.5O₃ was loaded by an impregnation method at 623 K for 1 h.

An action spectrum of water splitting on the optimized Rh_0.5Cr_1.5O₃(0.2 wt%)/AgTaO₃ was measured, as shown in Fig. 3. The maximum AQY was 38% at 340 nm. As far as we know, the AQY is the highest among one-step photoexcitation type water splitting using metal oxide photocatalysts with valence bands consisting of metal cation orbitals. In order to compare the activity of Rh_0.5Cr_1.5O₃(0.2 wt%)/AgTaO₃ with that of highly active NiO/NaTaO₃:La, photocatalytic water splitting over Rh_0.5Cr_1.5O₃(0.2 wt%)/AgTaO₃ was carried out in an inner-irradiation reaction cell made of quartz equipped with a 400 W high pressure Hg-lamp using 1 g of photocatalyst powder suspended in 340 mL of distilled water, as shown in Fig. 4. Evolution rates of H₂ and O₂ were 20 mmol h⁻¹ and 10 mmol h⁻¹, respectively. The activity of Rh_0.5Cr_1.5O₃(0.2 wt%)/AgTaO₃ was similar to that of NiO/NaTaO₃:La.¹⁹ This reaction catalytically proceeded because TON was 47. Thus, Rh_0.5Cr_1.5O₃(0.2 wt%)/AgTaO₃ interestingly showed the high activity for water splitting even without doping unlike La-doped NaTaO₃, Zn,Ca-codoped Ga₂O₃, and Al-doped SrTiO₃, suggesting that AgTaO₃ itself has good potential for water splitting. One possible explanation about the potential of AgTaO₃ will be the distortion of its crystal structure. The Ta–O–Ta bond angle of AgTaO₃ (164 degree) is very similar to that of NaTaO₃ (163 degree).²³ The similarity in the lattice distortion between AgTaO₃ and NaTaO₃ causes a similar property of a conduction band of AgTaO₃ to a highly active NaTaO₃. Therefore, mobility of photogenerated electrons and a reduction potential of water reduction of AgTaO₃ should be similar to those of NaTaO₃. Additionally, the valence band of NaTaO₃ is formed by only O 2p orbitals, whereas that of AgTaO₃ is formed by hybridized orbitals between Ag 4d and O 2p.²³ Therefore, photogenerated holes in AgTaO₃ could migrate more easily than those in NaTaO₃. These positive factors in AgTaO₃ gave high photocatalytic activity.

Fig. 3 An action spectrum of water splitting over Rh_0.5Cr_1.5O₃(0.2 wt%)/AgTaO₃ and a diffuse reflectance spectrum of AgTaO₃ prepared by a solid state reaction at 1273 K for 15 h without excess Ag. Rh_0.5Cr_1.5O₃ was loaded by an impregnation method at 623 K for 1 h. Photocatalyst: 0.3 g, reactant solution: distilled water (120 mL), cell: top-irradiation cell with a Pyrex window, light source: 300 W Xe-arc lamp with band-pass filters.

Fig. 4 Photocatalytic water splitting over Rh_0.5Cr_1.5O₃(0.2 wt%)/AgTaO₃ under UV irradiation. Photocatalyst: 1 g, reactant solution: distilled water (340 mL), cell: inner-irradiation cell made of quartz, light source: 400 W high pressure Hg-lamp. AgTaO₃ was synthesized by a solid state reaction at 1273 K for 15 h without excess Ag. Rh_0.5Cr_1.5O₃ was loaded by an impregnation method at 623 K for 1 h.

Water splitting proceeded over the Rh_0.5Cr_1.5O₃(0.2 wt%)/AgTaO₃ photocatalyst even under simulated sunlight irradiation, as shown in Fig. 5. H₂ and O₂ were steadily evolved with the rates of 486 mL m⁻² h⁻¹ and 247 mL m⁻² h⁻¹, respectively. The STH was estimated to be 0.13%. Additionally, we were able to watch evolved bubbles when Rh_0.5Cr_1.5O₃(0.2 wt%)/AgTaO₃ on the bottom of the reaction cell was irradiated with simulated sunlight under 1 atm at room temperature. The STH of Rh_0.5Cr_1.5O₃(0.2 wt%)/AgTaO₃ is lower than those of TiO₂/CoOOH/RhCrO_x/SrTiO₃:Al (STH = 0.4%)²² and SrTiO₃:Rh,La-Au-BiVO₄:Mo photocatalyst sheet (STH = 1.1%),²⁷ whereas it is higher than those of Z-schematic water splitting using SrTiO₃:Rh of a H₂-evolving photocatalyst, BiVO₄ of an O₂-evolving photocatalyst, and ionic electron mediators such as Fe^3+/2+ and [Co(bpy)₃]^3+/2+.^28,29 Thus, we successfully achieved highly efficient one-step photoexcitation type solar water splitting using Rh_0.5Cr_1.5O₃(0.2 wt%)/AgTaO₃ of a valence-band-controlled metal oxide photocatalyst.

Fig. 5 Photocatalytic solar water splitting over Rh_0.5Cr_1.5O₃(0.2 wt%)/AgTaO₃. AgTaO₃ was synthesized by a solid state reaction at 1273 K for 15 h without excess Ag. Rh_0.5Cr_1.5O₃ was loaded by an impregnation method at 623 K for 1 h. Rh_0.5Cr_1.5O₃(0.2 wt%)/AgTaO₃ was activated by a 300 W Xe-arc lamp in distilled water for 20 h before this reaction (Fig. 1) in order to eliminate the induction period early. Photocatalyst: 0.3 g, reactant solution: distilled water (120 mL), cell: top-irradiation cell with a Pyrex window, light source: a solar simulator with an AM1.5-filter (100 mW cm⁻²), irradiation area: 25 cm².

Conclusions

Rh_0.5Cr_1.5O₃(0.2 wt%)-loaded AgTaO₃ has arisen as a photocatalyst for solar water splitting in a suspension system. The AQY of Rh_0.5Cr_1.5O₃(0.2 wt%)/AgTaO₃ was about 40% at 340 nm. The activity of Rh_0.5Cr_1.5O₃(0.2 wt%)/AgTaO₃ was similar to that of NiO/NaTaO₃:La under the same experimental condition, using 1 g of photocatalyst powder suspended in 340 mL of distilled water, in an inner-irradiation cell made of quartz equipped with a 400 W high pressure Hg-lamp. AgTaO₃ seems to have good potential for water splitting because Rh_0.5Cr_1.5O₃(0.2 wt%)/AgTaO₃ showed the high AQY even without doping of elements. Rh_0.5Cr_1.5O₃(0.2 wt%)/AgTaO₃ also showed the photocatalytic activity for water splitting under simulated sunlight irradiation with the STH of 0.13%. The AQY and STH of Rh_0.5Cr_1.5O₃(0.2 wt%)/AgTaO₃ were the highest for one-step photoexcitation type water splitting using a valence-band-controlled metal oxide photocatalyst. Thus, it was clarified that a metal oxide photocatalyst with a valence band consisting of metal cation orbitals rather than O 2p could be utilized as a photocatalyst for highly efficient water splitting. The knowledge is expected to be applied to visible-light-driven photocatalysts.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by JSPS KAKENHI Grant Numbers 17H06433 in Scientific Research on Innovative Areas “Innovations for Light-Energy Conversion (I⁴LEC)” and 17H0127. The authors gratefully thank Dr T. Ichihashi for assistance with TEM measurements.

Notes and references

1. F. E. Osterloh, Chem. Mater., 2008, 20, 35–54.
2. Y. Inoue, Energy Environ. Sci., 2009, 2, 364–386.
3. A. Kudo and Y. Miseki, Chem. Soc. Rev., 2009, 38, 253–278.
4. R. Abe, J. Photochem. Photobiol., C, 2010, 11, 179–209.
5. T. Hisatomi, J. Kubota and K. Domen, Chem. Soc. Rev., 2014, 43, 7520–7535.
6. K. Maeda and K. Domen, Bull. Chem. Soc. Jpn., 2016, 89, 627–648.
7. K. Maeda, T. Takata, M. Hara, N. Saito, Y. Inoue, H. Kobayashi and K. Domen, J. Am. Chem. Soc., 2005, 127, 8286–8287.
8. K. Maeda, K. Teramura, D. Lu, T. Takata, N. Saito, Y. Inoue and K. Domen, Nature, 2006, 440, 295.
9. Y. Lee, H. Terashima, Y. Shimodaira, K. Teramura, M. Hara, H. Kobayashi, K. Domen and M. Yashima, J. Phys. Chem. C, 2007, 111, 1042–1048.
10. K. Takanabe, T. Uzawa, X. Wang, K. Maeda, M. Katayama, J. Kubota, A. Kudo and K. Domen, Dalton Trans., 2009, 10055–10062.
11. C. Pan, T. Takata, M. Nakabayashi, T. Matsumoto, N. Shibata, Y. Ikuhara and K. Domen, Angew. Chem., Int. Ed., 2015, 54, 2955–2959.
12. C. Pan, T. Takata and K. Domen, Chem.–Eur. J., 2016, 22, 1854–1862.
13. C. Pan, T. Takata, K. Kumamoto, T. Minegishi, M. Nakabayashi, T. Matsumoto, N. Shibata, Y. Ikuhara and K. Domen, J. Mater. Chem. A, 2016, 4, 4544–4552.
14. K. Maeda, D. Lu and K. Domen, Chem.–Eur. J., 2013, 19, 4986–4991.
15. J. Xu, C. Pan, T. Takata and K. Domen, Chem. Commun., 2015, 51, 7191–7194.
16. Z. Wang, Y. Inoue, T. Hisatomi, R. Ishikawa, Q. Wang, T. Takata, S. Chen, N. Shibata, Y. Ikuhara and K. Domen, Nat. Catal., 2018, 1, 756–763.
17. Q. Wang, M. Nakabayashi, T. Hisatomi, S. Sun, S. Akiyama, Z. Wang, Z. Pan, X. Xiao, T. Watanabe, T. Yamada, N. Shibata, T. Takata and K. Domen, Nat. Mater., 2019, 18, 827–832.
18. R. Asai, H. Nemoto, Q. Jia, K. Saito, A. Iwase and A. Kudo, Chem. Commun., 2014, 50, 2543–2546.
19. H. Kato, K. Asakura and A. Kudo, J. Am. Chem. Soc., 2003, 125, 3082–3089.
20. Y. Sakata, T. Hayashi, R. Yasunaga, N. Yanaga and H. Imamura, Chem. Commun., 2015, 51, 12935–12938.
21. H. T. Chiang, H. Lyu, T. Hisatomi, Y. Goto, T. Takata, M. Katayama, T. Minegishi and K. Domen, ACS Catal., 2018, 8, 2782–2788.
22. H. Lyu, T. Hisatomi, Y. Goto, M. Yoshida, T. Higashi, M. Katayama, T. Takata, T. Minegishi, H. Nishiyama, T. Yamada, Y. Sakata, K. Asakura and K. Domen, Chem. Sci., 2019, 10, 3196–3201.
23. H. Kato, H. Kobayashi and A. Kudo, J. Phys. Chem. B, 2002, 106, 12441–12447.
24. Y. Inoue, Energy Environ. Sci., 2009, 2, 364–386.
25. K. Maeda, K. Teramura, N. Saito, Y. Inoue and K. Domen, J. Catal., 2006, 243, 303–308.
26. K. Maeda, K. Teramura, D. Lu, T. Takata, N. Saito, Y. Inoue and K. Domen, J. Phys. Chem. B, 2006, 110, 13753–13758.
27. Q. Wang, T. Hisatomi, Q. Jia, H. Tokudome, M. Zhong, C. Wang, Z. Pan, T. Takata, M. Nakabayashi, N. Shibata, Y. Li, I. D. Sharp, A. Kudo, T. Yamada and K. Domen, Nat. Mater., 2016, 15, 611–615.
28. Y. Sasaki, H. Kato and A. Kudo, J. Am. Chem. Soc., 2013, 135, 5441–5449.
29. H. Kato, Y. Sasaki, N. Shirakura and A. Kudo, J. Mater. Chem. A, 2013, 1, 12327–12333.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c9sc05909a

---