A noble-metal-free photocatalytic hydrogen production system based on cobalt(III) complex and eosin Y-sensitized TiO₂

Abstract

Although dye-sensitized TiO₂ for photocatalytic H₂ production have been investigated for more than 20 years, research on using low-cost transition metal complexes as catalysts for proton reduction on TiO₂ is still relatively rare. In this paper, three cobalt complexes (CoL1–CoL3) were prepared and used as hydrogen evolution catalysts, and a noble-metal-free H₂-evolving system was fabricated by using an organic dye, eosin Y (EY), as photosensitizer and triethylamine (TEA) as sacrificial agent. Under visible light irradiation, a turnover number of 90 (based on CoL2) was obtained for 2 wt% CoL2/TiO₂ within 6 h in aqueous solution under the optimal conditions of 4 μmol EY, 5% TEA (v/v) and pH 11. In addition, much higher hydrogen production efficiency and stability were achieved for the heterogeneous system, in comparison with the homogeneous one.

---

1 Introduction

Since Fujishima and Honda discovered that water could be decomposed into H₂ and O₂ on a TiO₂ photoelectrode,¹ more and more attention has been paid to photocatalytic hydrogen production from water splitting using semiconductor materials, and TiO₂ has been considered as one of the most promising photocatalyst due to its low cost, nontoxicity, and photostability.² However, pure TiO₂ requires UV irradiation, shows fast electron–hole recombination rate and a high overpotential for hydrogen generation.^3,4 It is well-known that the absorption wavelength range can be extended to visible-light region by dye sensitization,^5,6 and the recombination rate can be restrained by loading Pt or enzymes known as hydrogenases. The loaded catalyst not only acts as electron sink to suppress the recombination of the photoexcited electrons and holes, but also serves as active sites to lower the H₂ evolution overpotential.^7,8

Up to now, the most effective catalyst for proton reduction is Pt, whereas it apparently raises the cost of hydrogen production and limits its practical application. Therefore, numerous efforts have been made to replace Pt with low-cost alternatives, in which transition metal complexes have also been investigated.⁸ For example, Reisner et al. reported the attachment of a complex Db[NiFeSe]-H to Ru(II) complex-sensitized TiO₂. A turnover frequency of approximately 50 (mol H₂) s⁻¹ (mol total hydrogenase)⁻¹ was obtained.⁹ However, to the best of our knowledge, researches on the photocatalytic hydrogen production based on transition metal complexes almost all focused on its use as H₂-evolving molecular catalysts in homogeneous systems,^10–19 though the heterogeneous systems usually possess better stability and practicality than the homogeneous ones. For instance, organic solvents are usually indispensable for the homogeneous systems, while for the heterogeneous ones, most reactions can proceed in an aqueous solution. In addition, the catalyst recovery is often difficult for the homogeneous systems. In comparison, studies on the application of cobalt complexes in the heterogeneous systems are relatively rare.^20–25 For example, Ji et al. reported the photoelectrochemical hydrogen production of CodmgBF₂ (dmg = dimethylglyoxime anion), which was linked to a Ru(II) sensitizer anchored onto NiO film. The resultant photoelectrode exhibited superior stability in aqueous solutions.²⁴ In particular, for the exploiting of cobalt complexes in heterogeneous TiO₂ systems, the only example was a cobaloxime complex [Co^III(dmgH)₂(pyridyl-4-hydrophosphonate)Cl] (abbreviated as CoP), which was anchored on TiO₂ and acted as the H₂ evolution catalyst.^26–29 With phosphonate-functionalized Ru(bpy)₃²⁺ (RuP) as photosensitizer (PS), 53 TON (based on CoP) of H₂ was generated within 5 h under visible light irradiation.²⁶ However, the PS was still precious Ru(bpy)₃²⁺ derivative in the above systems.

In order to investigate the utilization of new kinds of cobalt complexes as H₂ evolution catalysts of heterogeneous TiO₂ system and construct noble-metal-free hydrogen evolution systems, herein, three cobalt(III) complexes (abbreviated as CoL1–CoL3, as depicted in Fig. 1) were prepared and loaded onto TiO₂ surface through different bridging groups. Their hydrogen production performances were studied by using an organic dye eosin Y (abbreviated as EY, Fig. 1) as the PS and triethylamine (TEA) as the sacrificial agent under visible light irradiation. In addition, the influences of experimental parameters, such as the loading amount of CoL2 on TiO₂, the concentration of EY and TEA, and the pH value of the reaction system on hydrogen production activity were investigated in detail. Furthermore, the hydrogen evolution performance of CoL1–CoL3 was compared and the deactivation reason was explored preliminarily.

Fig. 1 Structures of cobalt complexes CoL1, CoL2, CoL3 and EY.

2 Experimental

2.1 Materials

1,10-Phenanthroline-5,6-dione (L1),³⁰ [Co(phen)₂Cl₂]Cl³¹ and 5-aldehydosalicylic acid³² were prepared and recrystallized according to literature procedures. TiO₂ used in this work was Degussa P25 obtained from a commercial source. SiO₂ was the silica gel used for thin-layer chromatography. All other chemicals were of analytical grade purity, obtained form Sinopharm Chemical Reagent Co. and used as received without further purification.

2.2 Photocatalyst preparation

2.2.1 Synthesis of [Co(phen)₂L1](PF₆)₃ (abbreviated as CoL1). CoL1 was synthesized according to the method reported by Song et al.³³ [Co(phen)₂Cl₂]Cl (0.50 g, 0.95 mmol) and 1,10-phenanthroline-5,6-dione (0.24 g, 1.14 mmol) was refluxed for 4 h in 60 mL 1 : 1 (v/v) ethanol–water. After that, the reaction mixture was cooled to room temperature and a precipitation was obtained by adding saturated aqueous NH₄PF₆ solution. The precipitation was collected, washed with water and ether, and then recrystallized from acetonitrile–ether. After filtrated and dried in vacuo for 24 h, the complex was achieved (yield: 65% based on Co). Anal. calcd (%) for C₃₆H₂₂CoF₁₈N₆O₂P₃: C, 40.52; H, 2.06; N, 7.88. Found: C, 40.83; H, 2.18; N, 7.62.

2.2.2 Synthesis of [Co(phen)₂L2](PF₆)₃ (abbreviated as CoL2). For the synthesis of [Co(phen)₂L2](PF₆)₃ (abbreviated as CoL2), a mixture of CoL1 (0.4775 g, 0.5 mmol), 5-aldehydosalicylic acid (0.1163 g, 0.7 mmol) and ammonium acetate (0.8004 g, 10 mmol) in 5 mL glacial acetic acid was refluxed for 3 h under nitrogen. After cooled, diluted with 40 mL H₂O and neutralized with ammonia to pH 5, the obtained precipitation was filtered off and washed repeatedly with water and diethyl ether to get a brown solid (yield: 63% based on Co). Anal. calcd (%) for C₄₄H₂₈CoF₁₈N₈O₃P₃: C, 43.93; H, 2.33; N, 9.32. Found: C, 44.03; H, 2.38; N, 9.02. IR (KBr, cm⁻¹): 3406 (O–H), 1625 (C=O), 1581 (C=N), 643 (Co–N (L2)), 558 (Co–N (phen)); ¹H NMR (300 MHz, DMSO-d₆, δ ppm): 9.15 (6H, d); 8.56 (6H, s); 7.96 (7H, m); 7.66 (6H, d).

2.2.3 Synthesis of [Co(phen)₂L3](PF₆)₃ (abbreviated as CoL3). It was prepared by the same procedure with CoL2, expect that 5-aldehydosalicylic acid was replaced by 4-carboxybenzaldehyde (0.1051 g, 0.7 mmol) (yield: 65% based on Co). Anal. calcd (%) for C₄₄H₂₈CoF₁₈N₈O₂P₃: C, 44.52; H, 2.36; N, 9.44. Found: C, 44.80; H, 2.40; N, 9.22. IR (KBr, cm⁻¹): 3395 (O–H), 1592 (C=O), 1544 (C=N), 643 (Co–N (L3)), 557 (Co–N (phen)); ¹H NMR (300 MHz, DMSO-d₆, δ ppm): 9.18 (6H, d); 8.56 (6H, s); 7.97 (7H, m); 7.72 (2H, m); 7.67 (4H, d).

2.2.4 Preparation of photocatalyst CoL/TiO₂. 2 mL DMF solution of a certain amount of CoL was added dropwise to a stirred dispersion of 1.0 g TiO₂ in 150 mL water. Then the mixture was stirred at room temperature for 24 h in dark. After centrifuged, the obtained solid was washed thoroughly with water and ethanol before dried at 60 °C for 12 h.

The amount of CoL loaded onto TiO₂ was determined by the UV-vis absorption spectra through measuring the absorbance difference of the clear supernatants before and after adding TiO₂. The results showed that almost both CoL2 and CoL3 were loaded onto TiO₂ when less than 6 wt% CoL was used because no absorption spectra of CoL2 and CoL3 were detected for the supernatants after the loading reaction. While the loading amount of CoL1 was only 50% due to the weak interaction between it and TiO₂, in comparison to that of CoL2 or CoL3.

2.2.5 Preparation of photocatalyst CoL2/SiO₂. 2 mL DMF solution of 20 mg CoL2 was added dropwise to a stirred dispersion of 1.0 g SiO₂ in 150 mL water. Then the mixture was stirred at room temperature for 24 h in dark. After centrifuged, the obtained solid was washed thoroughly with water and ethanol before dried at 60 °C for 12 h. The amount of CoL2 loaded onto SiO₂ was also determined by the UV-vis spectra and the result showed that only about 20% of CoL2 was adsorbed onto SiO₂.

2.3 Characterization

The ¹H NMR spectra were carried out on a Bruker 300 MHz NMR spectrometer by using DMSO-d₆ as the solvent and TMS as an internal standard. The IR spectra were recorded in the region of 4000–400 cm⁻¹ on a Nicolet IR-470 infrared spectrometer with pressed KBr pellets. The diffuse reflectance spectra (DRS) were recorded on a Shimadzu UV-2550 UV-vis spectrophotometer using BaSO₄ as reference, and transformed to the absorption spectra according to Kubelka–Munk relationship. The UV-vis absorption spectra were recorded using a Hitachi U2000 spectrophotometer. The cyclic voltammetries of the three CoL complexes were determined on a CHI-630B electrochemical workstation by using a glassy carbon as the working electrode, a Pt flag as the counter electrode, and a saturated calomel as the reference electrode (SCE) with a scan rate of 100 mV s⁻¹. All three electrodes were immersed in a 1 : 1 CH₃CN–H₂O (v/v) solution containing 0.05 mol L⁻¹ n-Bu₄NPF₆ as the electrolyte.

2.4 Photocatalytic hydrogen production test

The photocatalytic hydrogen production experiments were performed in a 230 mL quartz reactor with a flat window. 50 mg CoL/TiO₂ and a certain amount of EY were added to a 100 mL aqueous TEA solution. The pH was adjusted to a certain value by hydrochloric acid. The mixture was ultrasonically dispersed for 5 min, and then bubbled with nitrogen for 15 min before irradiated by a 300 W Xe lamp equipped with 420 nm cut-off filter. During the reaction, a continuous magnetic stir was applied and the temperature was kept to 30 °C through a circulating water system. The H₂ evolved was detected by Gas Chromatograph (GC, Agilent 4890D) equipped with a thermal conductivity detector (TCD) and a porapak Q stainless column. Nitrogen served as the carrier gas at a flow rate of 20 mL min⁻¹.

3 Results and discussion

3.1 Co(III) complexes preparation

In order to avoid the coordination of carboxylate group on ligands L2 and L3 with Co³⁺, complex CoL2 or CoL3 was prepared by reacting CoL1 with 5-aldehydosalicylic acid or 4-carboxybenzaldehyde, respectively. The IR and ¹H NMR results confirmed the successful achievement of CoL2 and CoL3.

3.2 UV-vis diffuse reflectance spectra of CoL2/TiO₂

The UV-vis diffuse reflectance spectra (Fig. 2) of TiO₂, 2 wt% CoL2/TiO₂ and 4 wt% CoL2/TiO₂ show that in comparison to pure TiO₂, CoL2/TiO₂ exhibits a broad absorption band in the region of 400–800 nm, and it increases gradually with the increasing of CoL2 from 2 wt% to 4 wt%, indicating the successful attachment of CoL2 to TiO₂ surface.

Fig. 2 UV-vis diffuse reflectance spectra of TiO₂ and CoL2/TiO₂.

3.3 Effect of different parameters on hydrogen production performance

The photocatalytic hydrogen production activities of CoL/TiO₂ catalysts were tested in a closed system. As a representative, firstly, various experimental factors, such as the amounts of CoL2, EY and TEA, and the pH value of the reaction system were studied to investigate their influences on hydrogen evolution activities.

3.3.1 Effect of CoL2 loading amount. The hydrogen production activities of TiO₂, CoL2 and CoL2/TiO₂ are given in Fig. 3. It indicates that when EY was used as sensitizer, almost no or only trace of H₂ was detected for either TiO₂ or CoL2, whereas attaching CoL2 on TiO₂ surface leads to a significant improvement on the hydrogen generation activity. For 2 wt% CoL2/TiO₂, the TON (vs. CoL2) reached to 90 within 6 h, indicating that the TiO₂ and CoL2 are both indispensable for hydrogen generation. However, for 4 wt% CoL2/TiO₂, the H₂-evolving activity decreased and the TON was only 58.

Fig. 3 Time courses of the photocatalytic hydrogen production of CoL2, TiO₂ and CoL2/TiO₂ (EY 4 μmol, catalyst 50 mg, TEA 5 mL and pH = 11; for the homogeneous system, the amount of CoL2 is 1 mg).

It is probably due to the following causes: (i) excessive loading of CoL2 decreases (or shields) the active sites of TiO₂ surface, leading to a decrease in the adsorption amount of EY on TiO₂;²⁸ (ii) more loading of CoL2 will result in more reverse reaction of hydrogen evolution; (iii) excessive loading of CoL2 will lead to more direct electron transfers between EY and CoL2, as the electron transfer route (2) shown in Fig. 4, which results in a lower hydrogen production efficiency. Correspondingly, the much higher H₂-evolving performance of CoL2/TiO₂ can be attributed to the efficient electron transfer from the excited EY* to the CB of TiO₂, and then to the H₂ evolution catalyst CoL, as the route (1) depicted in Fig. 4.^27,28,34,35

Fig. 4 Schematic representation of the possible electron transfers from EY to CoL in the heterogeneous EY-sensitized CoL/TiO₂ system (EY and CoL molecules are simplified as circles).

3.3.2 Effect of the pH value. On the other hand, the pH value of the reaction systems also affects its hydrogen production activity. As shown in Fig. 5, for 2 wt% CoL2/TiO₂, with the increase of pH from 9 to 11, the H₂ evolution activity was improved and achieved a maximum at pH 11. However, the TON drastically decreased to about 7 at pH 12. The low activity at lower pH value is probably due to the fact that the rate of Co^I species formation is apparently slowed, and it is difficult for TEA to be oxidized at the lower pH to provide electrons, whereas the activity decrease at higher pH value is possibly owing to the fact that the driving force for H₂ production is greatly reduced, and the protonation of the Co catalyst, an essential step postulated in forming H₂, may become very unfavorable.^17,18,36

Fig. 5 The amounts of hydrogen evolved under different pH values for 2 wt% CoL2/TiO₂ (EY 4 μmol, catalyst 50 mg, TEA 5 mL).

3.3.3 Effect of the EY and TEA concentrations. As illustrated in Fig. 6a, increasing the EY amount from 2 to 4 μmol results in an obvious improvement on hydrogen evolution efficiency due to the enhancement of light harvesting ability, whereas when the EY amount was further increased to 6 μmol, the H₂-evolving efficiency decreased because that some dye molecules absorbing input light can not participate in the electron transfer to the CB of TiO₂, which leads to loss of part of the incident light. On the other hand, the higher concentration of EY, the closer distance between adjacent dye molecules, results in more serious concentration quenching.^34,37 It results in less efficient utilization of dye molecules and a lower hydrogen evolution rate. However, from Fig. 6a, it can also be seen that when 6 μmol EY was used, the initial hydrogen evolution rate (from 0–2.5 h) was higher than that of 4 μmol EY, which indicates that except for the heterogeneous reaction, the homogeneous one also occurs namely the direct electron transfer from EY to CoL happens. So the activity decrease after 3 h can be attributed to the deactivation of homogeneous system, as discussed later. Similarly, when the TEA amount was increased from 2.5 to 5.0 mL, the H₂ produced increased, while it dramatically decreased at 7.5 mL and only 20 TON was produced within 6 h, as illustrated in Fig. 6b. Control experiments showed that no hydrogen was detected in the absence of either the sensitizer EY or TEA.

Fig. 6 (a) Time courses of hydrogen generation with various EY concentrations for 2 wt% CoL2/TiO₂ (catalyst 50 mg, TEA 5 mL and pH = 11). (b) The amounts of H₂ evolved with different amounts of TEA (EY 4 μmol, catalyst 50 mg, pH = 11).

3.3.4 Effect of CoL species and TiO₂ on hydrogen production. As the H₂-evolving catalyst, the species of CoL will determine its H₂ evolution performance. So the hydrogen production activities of CoL1–CoL3 were investigated under the same conditions of 50 mg 2 wt% CoL/TiO₂, 4 μmol EY, 5% (v/v) TEA and pH 11. As shown in Fig. 7, the results indicated that CoL2 was the most effective and the TON (vs. CoL2) was up to 90 after 6 h, which was higher than that of the RuP-sensitized CoP–TiO₂ system (53 vs. CoP).²⁶ While when CoL1 or CoL3 was utilized, the amount of H₂ produced decreased significantly and the TON was only 43 or 9, respectively. This can be explained by the redox potential of CoL and the interaction between it and TiO₂.²⁶ The cyclic voltammetry results show that the redox potential of Co(II)/Co(I) occurs at −0.45 V, −0.61 V, and −0.64 V (vs. NHE) for CoL1, CoL2 and CoL3, respectively, and there is no significant difference in the redox potential of CoL2 and CoL3 possibly due to their similar structures. Then the electron transfer form CB of TiO₂ (−0.70 V vs. NHE) to CoL catalysts is thermodynamically favorable²⁷ and in principle CoL1 with more positive potential should be more active than the other two CoL catalysts. However, the results indicated that the hydrogen-evolving activity of CoL1 was lower than that of CoL2. It is probably because that except for the redox potential, the interaction between CoL and TiO₂ also plays an important role in its H₂-evolving activity, namely the anchoring group on ligand L also affects its activity. Therefore, CoL2 with carboxylate and hydroxyl groups will result in stronger interaction with TiO₂ and exhibits higher hydrogen evolution activity than CoL1 possessing diketone ligand.

Fig. 7 Time courses of hydrogen production over different 2 wt% CoL/TiO₂ (EY 4 μmol, catalyst 50 mg, TEA 5 mL and pH = 11).

As we all know, TiO₂ serves not only as a catalyst carrier, but also a critical electronic component for the photocatalytic process. It harvests electrons from the excited EY molecules, facilitates the formation of a long-lived charge-separated state and transfers electrons to CoL.²⁶ In order to confirm this point, an electronically inert support SiO₂ was used to replace TiO₂ and the result showed that only 3.7 μmol H₂ was produced within 3 h under the same conditions.

3.4 The deactivation reason and the stability of CoL2

From the above results, it can be seen that after about 6 h, the hydrogen evolution of 2 wt% CoL2/TiO₂ ceased. To better understand the deactivation reason and the stability of CoL2, a homogeneous system was also investigated, in which CoL2 was used as molecular catalyst, EY as photosensitizer and TEA as the sacrificial electron donor. As shown in Fig. 8, its H₂-evolving activity is obviously lower than that of heterogeneous system and the hydrogen evolution ceases only after 3 h. However, readdition of EY (4 μmol) to the system leads to a recovered and only a little lower activity. In comparison, for the heterogeneous system, after the cessation of hydrogen evolution, readdition of EY only results in a slight recovery (Fig. 8b). And it is worth to note that the recovered hydrogen production activity for the heterogeneous system is almost the same as that of the homogeneous one. All this suggests that CoL2 is relatively stable, and the deterioration of hydrogen production activity can be ascribed to the desorption of CoL2 from TiO₂ surface and photolysis of the EY sensitizer.^18,38

Fig. 8 Two runs of hydrogen production for (a) the homogeneous system containing CoL2 (1 mg) and (b) the heterogeneous system containing 2 wt% CoL2/TiO₂ (50 mg) with 5 mL TEA, 4 μmol EY and pH = 11.

This deduction was further verified by the change of absorption spectra of EY before and after photocatalytic reaction. From Fig. 9, it can be seen that the absorption spectrum of the resulting clear solution is different from that of the original EY solution. The strong characteristic absorption peak at 520 nm for EY shifts to about 490 nm and it is very weak after irradiated by visible light for 6 h, which indicates that the sensitizer EY may be almost completely degraded during the reaction.

Fig. 9 UV-vis absorption spectra of the clear solutions containing EY only and after photocatalytic hydrogen evolution reaction for EY-sensitized 2 wt% CoL2/TiO₂ system.

4 Conclusions

In summary, as an inexpensive alternative to Pt, three cobalt complexes have been synthesized and acted as H₂ evolution catalysts, which were loaded onto TiO₂ by different bridging groups. By using an organic dye EY as photosensitizer and TEA as sacrificial electron donor, noble-metal free heterogeneous hydrogen production systems based on cobalt complexes have been firstly constructed and 90 turnovers of H₂ (based on CoL2) was produced within 6 h under visible light irradiation. This system is facile to assemble and can work in aqueous solution (free of organic solvents). The results also indicate that compared with the homogeneous system, the heterogeneous one displays much higher hydrogen-evolving activity and better stability. It will pave a way for the fabrication of noble-metal-free heterogeneous photocatalytic hydrogen production systems based on transition metal complexes, though the sensitizer and the composite catalyst CoL/TiO₂ were unstable in the present system. More researches on improving the activity and stability of cobalt complexes and exploiting more stable sensitizer are undergoing in our lab.

Acknowledgements

This research was financially supported by the National Science Foundation of China (no. 21171147), National Key Basic Research Program of China (973 Program) (no. 2009CB220005) and the China Postdoctoral Science Foundation (no. 20080441028).

References

1. A. Fujishima and K. Honda, Nature, 1972, 238, 37–38.
2. A. Kudo and Y. Miseki, Chem. Soc. Rev., 2009, 38, 253–278.
3. X. B. Chen, S. H. Shen, L. J. Guo and S. S. Mao, Chem. Rev., 2010, 110, 6503–6570.
4. S. P. Xu and D. D. Sun, Int. J. Hydrogen Energy, 2009, 34, 6096–6104.
5. M. Grätzel, Inorg. Chem., 2005, 44, 6841–6851.
6. E. Bae and W. Choi, J. Phys. Chem. B, 2006, 110, 14792–14799.
7. L. S. Yin, Y. P. Yuan, S. W. Cao, Z. Y. Zhang and C. Xue, RSC Adv., 2014, 4, 6127–6132.
8. J. H. Yang, D. E. Wang, H. X. Han and C. Li, Acc. Chem. Res., 2013, 46, 1900–1909.
9. E. Reisner, D. J. Powell, C. Cavazza, J. C. Fontecilla-Camps and F. A. Armstrong, J. Am. Chem. Soc., 2009, 131, 18457–18466.
10. C. V. Krishnan and N. Sutin, J. Am. Chem. Soc., 1981, 103, 2141–2142.
11. J. Hawecker, J. M. Lehn and R. Ziessel, Nouv. J. Chim., 1983, 7, 271–277.
12. X. Hu, B. S. Brunschwig and J. C. Peters, J. Am. Chem. Soc., 2007, 129, 8988–8998.
13. X. Hu, B. M. Cossairt, B. S. Brunschwig, N. S. Lewis and J. C. Peters, Chem. Commun., 2005, 4723–4725.
14. M. Razavet, V. Artero and M. Fontecave, Inorg. Chem., 2005, 44, 4786–4795.
15. V. C. A. Baffert, V. Artero and M. Fontecave, Inorg. Chem., 2007, 46, 1817–1824.
16. A. Fihri, V. Artero, M. Razavet, C. Baffert, W. Leibl and M. Fontecave, Angew. Chem., Int. Ed., 2008, 47, 564–567.
17. P. Du, K. Knowles and R. Eisenberg, J. Am. Chem. Soc., 2008, 130, 12576–12577.
18. T. Lazarides, T. McCormick, P. Du, G. Luo, B. Lindley and R. Eisenberg, J. Am. Chem. Soc., 2009, 13, 9192–9194.
19. W. R. McNamara, Z. Han, P. J. Alperin, W. W. Brennessel, P. L. Holland and R. Eisenberg, J. Am. Chem. Soc., 2011, 133, 15368–15371.
20. J. Huang, K. L. Mulfort, P. Du and L. X. Chen, J. Am. Chem. Soc., 2012, 134, 16472–16475.
21. S. W. Cao, X. F. Liu, Y. P. Yuan, Z. Y. Zhang, J. Fang, S. C. J. Loo, J. Barber, T. C. Sumb and C. Xue, Phys. Chem. Chem. Phys., 2013, 15, 18363–18366.
22. X. W. Song, H. M. Wen, C. B. Ma, H. H. Cui, H. Chen and C. N. Chen, RSC Adv., 2014, 4, 18853–18861.
23. F. Y. Wen and C. Li, Acc. Chem. Res., 2013, 46, 2355–2364.
24. Z. Q. Ji, M. F. He, Z. J. Huang, U. Ozkan and Y. Y. Wu, J. Am. Chem. Soc., 2013, 135, 11696–11699.
25. F. Y. Wen, J. H. Yang, X. Zong, B. J. Ma, D. E. Wang and Can Li, J. Catal., 2011, 281, 318–324.
26. F. Lakadamyaliy and E. Reisner, Chem. Commun., 2011, 47, 1695–1697.
27. F. Lakadamyali, A. Reynal, M. Kato, J. R. Durrant and E. Reisner, Chem.–Eur. J., 2012, 18, 15464–15475.
28. F. Lakadamyali, M. Kato and E. Reisner, Faraday Discuss., 2012, 155, 191–205.
29. F. Lakadamyali, M. Kato, N. M. Muresan and E. Reisner, Angew. Chem., Int. Ed., 2012, 51, 9381–9384.
30. M. Yamada, Y. Tanaka, Y. Yoshimoto, S. Kuroda and I. Shimao, Bull. Chem. Soc. Jpn., 1992, 65, 1006–1011.
31. G. Grassini-Strazza and S. Isola, J. Chromatogr. A, 1978, 154, 122–126.
32. J. C. Duff and E. J. Bills, J. Chem. Soc., 1932, 1987–1988.
33. Y. F. Song and P. Yang, Polyhedron, 2001, 20, 501–506.
34. J. Moser and M. Grätzel, J. Am. Chem. Soc., 1984, 106, 6557–6564.
35. Y. X. Li, C. F. Xie, S. Q. Peng, G. X. Lu and S. B. Li, J. Mol. Catal. A: Chem., 2008, 282, 117–123.
36. P. Zhang, M. Wang, J. F. Dong, X. Q. Li, F. Wang, L. Z. Wu and L. C. Sun, J. Phys. Chem. C, 2010, 114, 15868–15874.
37. S. X. Min and G. X. Lu, Int. J. Hydrogen Energy, 2013, 38, 2106–2116.
38. Z. Y. Wang, H. Rao, M. F. Deng, Y. T. Fan and H. W. Hou, Phys. Chem. Chem. Phys., 2013, 15, 16665–16671.

---