Facile synthesis of bismuth oxide/bismuth vanadate heterostructures for efficient photoelectrochemical cells

Abstract

Herein, we report a facile approach to synthesize Bi₂O₃/BiVO₄ heterostructures for photoelectrochemical (PEC) cells. Due to the fast separation of the electron–hole pairs as a result of the p–n junction, the Bi₂O₃/BiVO₄ heterostructures achieved a remarkable photocurrent of 2.58 mA cm⁻² at 1.2 V vs. Ag/AgCl, which is about 5 times that of the pristine BiVO₄.

---

Due to ever-growing environmental concerns and increasing energy demands, hydrogen is considered as a key energy of the future because of its clean, renewable, carbon-free, and high energy density properties.^1–6 Inspired by natural photosynthesis, artificial photoelectrochemical (PEC) water splitting is a promising pathway to economically produce hydrogen.^7–10 The efficiency of PEC water splitting is largely determined by the properties of the photoelectrode, and considerable efforts have been devoted to exploring the highly active photoelectrode materials. Semiconductor metal oxides such TiO₂,¹¹ ZnO,^12–14 Fe₂O₃,¹⁵ In₂O₃,¹⁶ and WO₃ (ref. 17) have been extensively studied as photoanodes for PEC water splitting and numerous successes have achieved. Among various metal oxides, binary metal oxide BiVO₄ hold great promise for its significant advantages of the photoactive phase i.e. the monoclinic scheelite phase with a band gap of ∼2.36 eV and high theoretical efficiency of 9.1%, which is capable of harvesting visible light.^9,18,19 Nevertheless, the efficiency of the BiVO₄ is still low as a result of its relatively poor light-harvesting ability and rapid recombination of photo-generated carriers.^20–22 Therefore, it is highly desirable and important to improve the photoactivity and of BiVO₄ by increasing the efficiency of absorption of the BiVO₄ and the separation of the electron–hole pairs.

To improve the PEC activity of BiVO₄ photoanodes, some strategies including element-doping, hydrogenation and composites have been proposed. Another effective strategy is to develop BiVO₄-based heterostructures. For example, various heterostructures such as WO₃/BiVO₄,²³ CaFe₂O₄/BiVO₄,²⁴ TiO₂/BiVO₄ (ref. 25) and BiVO₄/Bi₂S₃ (ref. 26) have been reported and enhanced PEC activity of these photoanodes have been achieved as a result of the increased carrier density, reduced electron–hole recombination and the narrowed band gap of the semiconductor. However, the present PEC performance of these heterostructures is still unsatisfactory.³ Therefore, the development of new BiVO₄-based heterostructure photoanodes with high PEC activity and excellent stability is very desirable.

In this work, we firstly reported the design and synthesis of the Bi₂O₃/BiVO₄ p–n heterostructures for efficient PEC cells. α-Bi₂O₃ is an intrinsic p-type semiconductor with good PEC water splitting and photocatalytic activity.²⁷ When the p-type Bi₂O₃ and n-type BiVO₄ are integrated together, a number of p–n junctions will be formed. Then, the holes in the valence band (VB) of the p-type Bi₂O₃ will combine with the electrons in the conduction band (CB) of the n-type BiVO₄ due to the p–n junctions. As a consequence, the photoexcited electron–hole pairs are effectively separated by this novel n–p junction structure, which is crucial for the enhancement of PEC activity. Our results show that the Bi₂O₃/BiVO₄ p–n heterostructures achieved a remarkable photocurrent density of 2.58 mA cm⁻² at 1.2 V vs. Ag/AgCl, which is about 5 folds that of pristine of BiVO₄. The Incident-photon-to-current-conversion efficiency (IPCE) of Bi₂O₃/BiVO₄ reaches 35.6% at 440 nm, which is much higher than that of the pristine BiVO₄ (about 10% at 440 nm). The Mott–Schottky analysis supports that the Bi₂O₃/BiVO₄ electrode possesses one order of magnitude improvement on donor density compared to the bare BiVO₄ electrode. This high photocatalytic activity makes Bi₂O₃/BiVO₄ a promising and an active photoanode.

The Bi₂O₃/BiVO₄ heterostructure nanospheres were prepared on FTO substrates via a two-step process, as schematically illustrated in Fig. 1. Metallic Bi nanobelts were firstly grown on FTO substrates by electrodeposition (Experimental section). As shown in Fig. 1b, the deposited film is black in colour and consisting of nanobelts with a 300 nm wide dendritical structure. XRD spectrum confirms that the deposited film is rhombohedral Bi (Fig. S4a and S5†). To obtain the Bi₂O₃/BiVO₄ heterostructures, the as-prepared Bi nanobelts were impregnated in the solution of 0.1 mol L⁻¹ NH₄VO₃ for 6 h and then annealed at 550 °C in air (Experimental section). Interestingly, the colour of the film has been changed from black to yellow (inset in Fig. 1c), and XRD studies clearly demonstrate the successful transformation from metallic Bi to Bi₂O₃/BiVO₄ (Fig. S4d†). As shown in Fig. 1c, the dendritical structure is basically retained after formed the Bi₂O₃/BiVO₄ heterostructures. Additionally, it should be noted that the Bi nanobelts have been transformed into some connected nanospheres with a dendritical structure. Furthermore, the composition of the Bi₂O₃/BiVO₄ heterostructures can be readily adjusting the immersion time in the NH₄VO₃ solution. For example, Bi₂O₃/BiVO₄ heterostructures was obtained when the immersion time was fixed to 6 h, as shown in Fig. S1.† When the immersion time increased to 12 h, pure BiVO₄ nanospheres were obtained (Fig. S2 and S4c†). In addition, pure Bi₂O₃ nanobelts could be formed by directly annealed the as-prepared Bi nanobelts in air (Fig. S3 and S4b†). All these results clearly show that our present method is an effective method to synthesize the heterostructures with controllable composition. More XRD details of Bi₂O₃, BiVO₄, and composite were shown in Fig. S4.†

Fig. 1 (a) Schematic diagrams for the growth process of Bi₂O₃/BiVO₄ heterostructures, SEM images of (b) Bi, (c) Bi₂O₃/BiVO₄ heterostructures.

In order to further investigate the chemical composition and defect state of the products, X-ray photoelectron spectroscopy (XPS) analysis was performed. The XPS survey spectra collected for the pristine Bi₂O₃, BiVO₄ and Bi₂O₃/BiVO₄ heterostructures were shown in Fig. 2a. Besides Sn, Si signals originating from the FTO substrate and C signals originating from the adventitious carbon, only Bi, V and O are detected on the sample surface. Fig. 2b shows the Bi 4f XPS spectra of the samples. The doublet broad peaks with higher binding energy of 159.4 eV and 164.6 eV are observed for all the samples, which are consistent with the characteristic Bi 4f_7/2 and Bi 4f_5/2 of Bi³⁺ peaks.²⁸ This reveals that the Bi has been successfully converted into Bi³⁺. Moreover, the V 2p spectra confirm the presence of V⁵⁺ in the Bi₂O₃/BiVO₄ and BiVO₄ samples (Fig. 2c). The binding energies of the synthetic peaks centred at 516.3 eV are consistent with the reported values for V⁵⁺.²⁹ Fig. 2d displays the Raman spectra of Bi₂O₃, Bi₂O₃/BiVO₄ and BiVO₄. The peaks located at 329 cm⁻¹ and 447 cm⁻¹ are the characteristic peaks for Bi₂O₃,³⁰ while the peaks at 826 and 128 cm⁻¹ are assigned to the V–O vibration of BiVO₄ structure units.³⁰ For the Bi₂O₃/BiVO₄ sample, the peak at 826 cm⁻¹ is assigned to BiVO₄ and the peak at 329 cm⁻¹ is assigned to Bi₂O₃, respectively. Therefore, the Bi₂O₃/BiVO₄ sample consists of Bi₂O₃ and BiVO₄, which also affirmed the successful formation of the heterostructure.

Fig. 2 (a) Survey XPS spectra and high-resolution XPS spectra of Bi₂O₃, Bi₂O₃/BiVO₄ heterostructures and BiVO₄. XPS spectra of (b) Bi_4/f (c) V_2p. (d) Enlarged room temperature Raman-scattering spectrum in range 50–800 cm⁻¹ of Bi₂O₃, Bi₂O₃/BiVO₄ heterostructures and BiVO₄.

In order to better understand the microstructure of Bi₂O₃/BiVO₄ heterostructures, transmission electron microscopy (TEM) analysis was carried out. Fig. 3a is a typical TEM image of the as-prepared Bi₂O₃/BiVO₄ heterostructures, showing that a lot of nanoparticles of 5 nm in diameter are uniformly covered on the surface of the spheres. Fig. 3b shows the high-resolution TEM (HRTEM) image of the sample, suggesting the Bi₂O₃/BiVO₄ heterostructure nanospheres are well crystalline. The well-resolved lattice fringes of 0.32 nm that corresponding to the (120) plane of monoclinic Bi₂O₃ are well observed. The lattice fringes of 0.29 nm that corresponding to the (040) plane of monoclinic BiVO₄ are also observed. Therefore, the Bi₂O₃/BiVO₄ heterostructures are successfully prepared. As shown in Fig. 3c–f, distributions of Bi, V and O are clearly presented by the STEM EDS elemental maps, which indicate that the Bi, V and O are uniformly embedded in the nanospheres, and high combining degree of each other.

Fig. 3 (a and b) TEM bright field images of the Bi₂O₃/BiVO₄ heterostructures. (c) HAADF-STEM image of the Bi₂O₃/BiVO₄ heterostructures. (d–f) STEM-EDS elemental mappings of Bi, V and O, respectively.

To study the effect of the p–n heterojunction on the PEC activity of BiVO₄ photoanode, PEC measurements were performed on the pristine BiVO₄, Bi₂O₃/BiVO₄ and Bi₂O₃ samples in a three-electrode electrochemical cell with 0.1 mol L⁻¹ Na₂SO₄ solution as the electrolyte. Fig. 4a compares the current vs. potential (i–E) curves for the pristine BiVO₄, Bi₂O₃/BiVO₄ heterostructures and Bi₂O₃ photoelectrodes in the dark and under visible light (λ > 420 nm) irradiation. As expected, the Bi₂O₃/BiVO₄ electrode exhibited the best photocurrent density compared to the BiVO₄ and Bi₂O₃. The photocurrent density of Bi₂O₃/BiVO₄ achieved a current density of 2.58 mA cm⁻² at the potential of 1.2 V vs. Ag/AgCl, which is about 5 times higher than those of pristine BiVO₄ (0.47 mA cm⁻²). This present value is also higher than recent reported BiVO₄-based photoanodes, such as Co-Pi modified BiVO₄/ZnO (2.0 mA cm⁻² at 1.2 V vs. Ag/AgCl),³¹ WO₃/BiVO₄ (0.55 mA cm⁻² at 1.23 V vs. Ag/AgCl),³² BiVO₄–TiO₂ (0.53 mA cm⁻² at 1.2 V vs. Ag/AgCl),²⁵ macro–mesoporous Mo:BiVO₄ (2.0 mA cm⁻² at 1.0 V vs. Ag/AgCl).³³ Furthermore, the current vs. potential (i–E) curve for the pure p-type Bi₂O₃ is shown in Fig. S6.† Incident-photon-to-current-conversion efficiency (IPCE) measurements were further performed to investigate the PEC performances of the as-prepared photoelectrodes. Fig. 4b shows the IPCE spectra of the pristine Bi₂O₃, BiVO₄ and Bi₂O₃/BiVO₄ photoelectrodes measured at 0.6 V vs. Ag/AgCl as a function of incident light wavelength. Obviously, the Bi₂O₃/BiVO₄ photoanode showed substantially enhanced IPCE values compared to the pristine Bi₂O₃ and BiVO₄ at all measured wavelengths, which agree well with their i–E characteristics. The maximum IPCE value of the Bi₂O₃/BiVO₄ reaches 35% at 440 nm, which is much higher than that of pristine BiVO₄ (about 10% at 440 nm) and pure Bi₂O₃ (about 0.16% at 440 nm). This result conveniently supports our hypothesis that the PEC performance of BiVO₄ can be greatly improved by forming Bi₂O₃/BiVO₄ p–n type heterostructures.

Fig. 4 (a) i–E curves recorded with a scan rate of 25 mV s⁻¹ under a full-arc xenon lamp irradiation, (b) IPCE spectra collected at the incident wavelength range from 300 to 700 nm at 0.6 V vs. Ag/AgCl of the Bi₂O₃, BiVO₄, and Bi₂O₃/BiVO₄ heterostructures, (c) chronoamperometry (i–t) of Bi₂O₃/BiVO₄ heterostructures.

The PEC stability of a photoelectrode is another crucial point for a PEC cell to produce hydrogen. The stability of the Bi₂O₃/BiVO₄ photoanode was also evaluated, and the photocurrent–time (i–t) curves of Bi₂O₃/BiVO₄ photoanode collected at 1.2 V vs. Ag/AgCl are shown in Fig. 4c. The photocurrent of Bi₂O₃/BiVO₄ is initially about 2.58 mA cm⁻² and the photocurrent only decreases to 2.28 mA cm⁻² within the initial 20 minutes and then photocurrent decreases slowly to 1.91 mA cm⁻² within 120 min of illumination. This demonstrates that the Bi₂O₃/BiVO₄ heterostructures are very stable during the long time irradiation at extreme voltage, which is due to the enhanced separation of electron–hole pairs in the p–n heterostructures.

It is known that both the light absorption and separation efficiency of photoexcited electron–hole pairs have important influences on the PEC property of the photoelectrode. Diffuse reflectance UV-visible spectra of the Bi₂O₃, BiVO₄ and Bi₂O₃/BiVO₄ heterostructures were collected to understand the influence of p–n heterojunction on the light harvesting capability. As shown in Fig. 5a, the band edges of the Bi₂O₃ and BiVO₄ samples were at about 400 nm and 512 nm, respectively, which are consistent with the recent reports.^34,35 There is only one absorption band edge as the Bi₂O₃ and BiVO₄, distribute so uniformly in the heterostructures, which is consistent with the recent reports, such as BiOI/TiO₂ heterostructures,³⁶ BiVO₄/TiO₂ heterostructures.³⁷ Additionally, the band edge of Bi₂O₃/BiVO₄ heterostructures sample (460 nm) lied between the Bi₂O₃ (400 nm) and BiVO₄ (512 nm). As an indirect semiconductor, the band gap of Bi₂O₃, BiVO₄ and Bi₂O₃/BiVO₄ heterostructures could be determined with the formula:

αhν = A(hν − E_g)^n/2,

where α, h, ν, A, E_g, and n are the absorption coefficient, Planck's constant, the incident light frequency, a constant, the band-gap energy, and an integer, respectively. Among them, n depends on the characteristics of the optical transition in a semiconductor, i.e., direct transition (n = 1) or indirect transition (n = 4). For BiVO₄ and Bi₂O₃, both of them pertain to direct transition and the values of n are 1.³⁸ The band-gap energy (E_g value) of BiVO₄ can be thus estimated from a plot of (αhν)² versus the photon energy (hν). From the Fig. S7,† the estimated band gap of Bi₂O₃ sample is about 2.32 eV, which is much smaller than those of untreated BiVO₄ (2.36 eV) and Bi₂O₃ (2.85 eV) samples. The conduction band edge of a semiconductor at the point of zero charge can be calculated by the empirical equation:³⁹

E_VB = X − E^e + 0.5E_g,

where E_VB is the valence band-edge potential, X is the electronegativity of the semiconductor, expressed as the geometric mean of the absolute electronegativity of the constituent atoms, E^e is the energy of free electrons on the hydrogen scale (about 4.5 eV), E_g is the band-gap energy of the semiconductor, and E_CB can be determined by E_CB = E_VB − E_g. The X values for BiVO₄ and Bi₂O₃ are ca. 6.04 and 5.95 eV. The band-gap energies of BiVO₄ and Bi₂O₃ adopt 2.36 and 2.85 eV, respectively, which are consistent with the recent reports.^34,35 Given the equation above, the CB and the bottom of the VB of BiVO₄ are calculated to be 0.36 and 2.72 eV, the CB and VB of Bi₂O₃ are calculated to be 0.03 and 2.88 eV, respectively. According to the above results, Scheme 1 was proposed to illustrate a possible charge-separation process. The p-type Bi₂O₃ is with Fermi energy level close to the valence band while the n-type BiVO₄ is with Fermi energy level close to the conduction band (XPS VBM was shown in Fig. S8†). When the two semiconductors are in contact, the CB potential of Bi₂O₃ is more negative than that of BiVO₄. As shown in Scheme 1, electron–hole pairs will generate on Bi₂O₃ and BiVO₄ under the irradiation of visible-light. The excited holes on the VB of Bi₂O₃ transfer to the CB of BiVO₄ and combine together. Then, the electrons produced on the CB of Bi₂O₃ will transfer to Pt electrode to reduce H⁺ to form H₂, while the excited holes on the VB of BiVO₄ will oxide H₂O to O₂. Furthermore, the migration of photogenerated electrons and holes could be promoted by the internal electric field. Therefore, the formation of Bi₂O₃/BiVO₄ p–n heterojunction on the surface could effectively separate the photoexcited electron–hole pairs and could greatly reduce the recombination of the photogenerated charge carries.

Fig. 5 (a) UV-vis diffuses absorption spectra and (b) Mott–Schottky curves of BiVO₄ and Bi₂O₃/BiVO₄ heterostructures.

Scheme 1 Schematic diagram of the band energy of Bi₂O₃ and BiVO₄ before contact and the formation of a p–n heterojunction and the proposed charge transfer and separation process of Bi₂O₃/BiVO₄ p–n heterostructure under visible-light irradiation.

To elucidate the influence of heterostructure on photoelectrical properties of BiVO₄, the electrochemical impedance measurements were carried out. Fig. 5b shows the Mott–Schottky plots of the BiVO₄ electrodes at a frequency of 1 kHz in the dark, which were generated based on capacitances that were derived from the electrochemical impedance. Both the BiVO₄ and Bi₂O₃/BiVO₄ electrodes show positive slopes, as expected for n-type semiconductors. Notably, Bi₂O₃/BiVO₄ shows substantially smaller slope of Mott–Schottky plot compared to bare BiVO₄, suggesting significantly increased donor densities based on the following equation:

N_d = (2/e₀εε₀)[d(1/C²)/dV]⁻¹,

where N_d is the donor density, e₀ the electron charge, ε the dielectric constant of BiVO₄ (ε = 86),²⁰ ε₀ the permittivity of vacuum, and V the applied bias at the electrode. The carrier densities of the BiVO₄ and Bi₂O₃/BiVO₄ electrodes are calculated to be 1.85 × 10¹⁶ and 3.80 × 10¹⁷ cm⁻³, respectively. Notably, Bi₂O₃/BiVO₄ electrode possesses one order of magnitude improvement on donor density compared to the bare BiVO₄ electrode. The drastically increasing donor density of Bi₂O₃/BiVO₄ clearly indicated the enhancement of the conductivity. On the other hand, the Mott–Schottky curve of p-type Bi₂O₃ is shown in Fig. S9.† All the results clearly verified the assumption that the heterostructures influence the PEC performance deeply.

Conclusions

In summary, the Bi₂O₃/BiVO₄ heterostructures were successfully prepared by a simple electrodeposition method and followed calcination. Raman, TEM and XPS analyses confirm the formation of the p–n heterostructures, which can facilitate the transportation and separation of the photo-generated electron–hole pairs. The Bi₂O₃/BiVO₄ heterostructures exhibited significantly enhanced PEC activity under visible light irradiation. The photocurrent density of Bi₂O₃/BiVO₄ heterostructure photoanode achieved a high photocurrent of 2.58 mA cm⁻² at 1.2 V vs. Ag/AgCl, which is about 5 times than that of pristine BiVO₄. These finding indicates the Bi₂O₃/BiVO₄ heterostructures are very promising candidates for PEC cells.

Acknowledgements

Y.M.Z. acknowledges the financial support of this work received by the Natural Science Foundation of China (no. 21276104). K.H.Y. acknowledges the Jinan University Scientific Research Innovation Cultivation Project of Excellent Postgraduate Candidates Exempt from Admission Exam (no. 33220131114).

Notes and references

1. J. Gan, X. Lu and Y. Tong, Nanoscale, 2014, 6, 7142–7164.
2. X. Lu, S. Xie, H. Yang, Y. Tong and H. Ji, Chem. Soc. Rev., 2014, 43, 7581–7593.
3. H. G. Kim, P. H. Borse, W. Y. Choi and J. S. Lee, Angew. Chem., Int. Ed., 2005, 44, 4585–4589.
4. J. Shi, Y. Hara, C. Sun, M. A. Anderson and X. Wang, Nano Lett., 2011, 11, 3413–3419.
5. X. Lu, G. Wang, S. Xie, J. Shi, W. Li, Y. Tong and Y. Li, Chem. Commun., 2012, 48, 7717–7719.
6. S. Yang, L. Fan and S. Yang, J. Phys. Chem. B, 2004, 108, 4394–4404.
7. S. Xie, T. Zhai, W. Li, M. Yu, C. Liang, J. Gan, X. Lu and Y. Tong, Green Chem., 2013, 15, 2434–2440.
8. G. Wang, Y. Ling, H. Wang, X. Lu and Y. Li, J. Photochem. Photobiol., C, 2014, 19, 35–51.
9. S. Xie, T. Zhai, Y. Zhu, W. Li, R. Qiu, Y. Tong and X. Lu, Int. J. Hydrogen Energy, 2014, 39, 4820–4827.
10. C.-H. Zeng, S. Xie, M. Yu, Y. Yang, X. Lu and Y. Tong, J. Power Sources, 2014, 247, 545–550.
11. J. Luo, L. Ma, T. He, C. F. Ng, S. Wang, H. Sun and H. J. Fan, J. Phys. Chem. C, 2012, 116, 11956–11963.
12. K.-H. Ye, J.-Y. Wang, N. Li, Z.-Q. Liu, S.-H. Guo, Y.-P. Guo and Y.-Z. Su, Inorg. Chem. Commun., 2014, 45, 116–119.
13. H. Li, Y. Fu, H. Liu, M. Zhu, Z. Peng, J. Yang, J. Li, X. Huang, Y. Jiang and Q. Liu, Inorg. Chem. Commun., 2013, 30, 182–186.
14. C. Yao, B. Wei, H. Ma, L. Meng, X. Zhang and Q. Gong, J. Power Sources, 2013, 237, 295–299.
15. A. Ikram, S. Sahai, S. Rai, S. Dass, R. Shrivastav and V. R. Satsangi, J. Power Sources, 2014, 267, 664–672.
16. H. Li, C. Chen, X. Huang, Y. Leng, M. Hou, X. Xiao, J. Bao, J. You, W. Zhang and Y. Wang, J. Power Sources, 2014, 247, 915–919.
17. T. Wang, J. Tang, S. Wu, X. Fan and J. He, J. Power Sources, 2014, 248, 510–516.
18. M.-L. Guan, D.-K. Ma, S.-W. Hu, Y.-J. Chen and S.-M. Huang, Inorg. Chem., 2011, 50, 800–805.
19. Y. Liang, T. Tsubota, L. P. A. Mooij and R. van de Krol, J. Phys. Chem. C, 2011, 115, 17594–17598.
20. D. K. Zhong, S. Choi and D. R. Gamelin, J. Am. Chem. Soc., 2011, 133, 18370–18377.
21. F. F. Abdi and R. van de Krol, J. Phys. Chem. C, 2012, 116, 9398–9404.
22. S. P. Berglund, A. J. E. Rettie, S. Hoang and C. B. Mullins, Phys. Chem. Chem. Phys., 2012, 14, 7065–7075.
23. Y. Pihosh, I. Turkevych, K. Mawatari, T. Asai, T. Hisatomi, J. Uemura, M. Tosa, K. Shimamura, J. Kubota and K. Domen, Small, 2014, 10, 3692–3699.
24. E. Kim, H. Kang, G. Magesh, J. Kim, J. Jang and J. Lee, ACS Appl. Mater. Interfaces, 2014, 6, 17762–17769.
25. S. Ho-Kimura, S. J. A. Moniz, A. D. Handoko and J. Tang, J. Mater. Chem. A, 2014, 2, 3948–3953.
26. S. Moniz, J. Zhu and J. Tang, Adv. Energy Mater., 2014, 4, 1301590.
27. L. Li and B. Yan, J. Alloys Compd., 2009, 476, 624–628.
28. S. K. Pilli, T. E. Furtak, L. D. Brown, T. G. Deutsch, J. A. Turner and A. M. Herring, Energy Environ. Sci., 2011, 4, 5028–5034.
29. L. Zhang, D. R. Chen and X. L. Jiao, J. Phys. Chem. B, 2006, 110, 2668–2673.
30. Y. Jianqiang and A. Kudo, Adv. Funct. Mater., 2006, 16, 2163–2169.
31. S. J. A. Moniz, J. Zhu and J. Tang, Adv. Energy Mater., 2014, 4, 1301590.
32. Y. Pihosh, I. Turkevych, K. Mawatari, T. Asai, T. Hisatomi, J. Uemura, M. Tosa, K. Shimamura, J. Kubota, K. Domen and T. Kitamori, Small, 2014, 10, 3692–3699.
33. M. Zhou, J. Bao, Y. Xu, J. Zhang, J. Xie, M. Guan, C. Wang, L. Wen, Y. Lei and Y. Xie, ACS Nano, 2014, 8, 7088–7098.
34. L. S. Zhang, W. Z. Wang, J. O. Yang, Z. G. Chen, W. Q. Zhang, L. Zhou and S. W. Liu, Appl. Catal., A, 2006, 308, 105–110.
35. L. Zhou, W. Wang, L. Zhang, H. Xu and W. Zhu, J. Phys. Chem. C, 2007, 111, 13659–13664.
36. X. Zhang, L. Zhang, T. Xie and D. Wang, J. Phys. Chem. C, 2009, 113, 7371–7378.
37. Y. Hu, D. Li, Y. Zheng, W. Chen, Y. He, Y. Shao, X. Fu and G. Xiao, Appl. Catal., B, 2011, 104, 30–36.
38. L. Zhou, W. Z. Wang, S. W. Liu, L. S. Zhang, H. L. Xu and W. Zhu, J. Mol. Catal. A: Chem., 2006, 252, 120–124.
39. D. Hou, X. Hu, P. Hu, W. Zhang, M. Zhang and Y. Huang, Nanoscale, 2013, 5, 9764–9772.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra03500g

---