TiO₂/Bi₂S₃ core–shell nanowire arrays for photoelectrochemical hydrogen generation

Abstract

A facile and effective procedure for the synthesis of quasi-vertically aligned TiO₂/Bi₂S₃ core–shell nanowire arrays (NWAs) is reported. The nontoxic and earth abundant Bi₂S₃ was assembled onto the hydrothermal pre-grown TiO₂ NWAs via a successive ionic layer adsorption and reaction (SILAR) method. The morphologies, microstructures, and optical properties of the pristine TiO₂ and composite TiO₂/Bi₂S₃ with different SILAR circles were characterized in detail. For photoelectrochemical (PEC) measurements, the TiO₂/Bi₂S₃ core–shell NWAs exhibited not only an enhanced photocurrent density (2.8 times higher than that of pristine TiO₂), but also a negatively shifted onset potential from 0.067 to −0.072 V vs. RHE, as compared to the TiO₂. This better PEC performance results from the broadened light absorption and the improved charge carrier separation efficiency. Our results provide a green photoelectrode for PEC hydrogen generation.

---

Introduction

Since the seminal work of Fujishima and Honda,¹ titanium dioxide (TiO₂) has been widely investigated as a photoanode material for photoelectrochemical (PEC) hydrogen production owing to its excellent chemical stability, low cost, nontoxicity and environment-friendly features.^2–7 Among various nanostructured TiO₂, the one dimensional TiO₂ nanowire arrays (NWAs) on transparent FTO substrates have attracted considerable interest due to their facile fabrication, good oriented charge-transport properties, and large specific surfaces for light trapping.^8–14 Despite these advantages, the photoconversion efficiencies of TiO₂ NWA photoanodes are extremely depressed due to the large energy band gap (3.0 eV for rutile TiO₂).^15,16 This large energy band gap restricts their solar energy utilization only in the ultraviolet region. Therefore, considerable efforts have been made to enhance their visible light harvesting ability.^17–23 For example, Wang et al. reported that the transition-metal doped TiO₂ NWAs exhibited dramatically improved visible light absorptions and photocatalytic properties in comparison to the pristine TiO₂ samples.²⁴ Moreover, using morphology controlling strategies to design novel architectures is also explored to enhance the light utilization ability of TiO₂, such as the branched TiO₂ nanorod array electrode reported by Cho et al.²⁵

Sensitizing TiO₂ NWAs with short band gap semiconductors is another approach to extend the working spectrum into the visible range. In this case, the type-II heterostructure NWAs is integrated, and the light absorption can be extended by the narrower band gap sensitizers. In addition, the charge separation efficiency can also be largely improved due to the build-in space electric field. The conventional II–VI semiconductors such as CdS and CdSe, have exhibited considerable PEC performance improvement as sensitizers for TiO₂ NWAs. However, their fatal toxicity is a worrying issue when used for practical applications. Recently, Bi₂S₃ has been emerging to be an effective, inexpensive and environmentally friendly light absorber material for wide band gap semiconductors.²⁶ For example, Gao et al. reported the Bi₂S₃-coated BiVO₄ discoids by in situ transform the surface BiVO₄ into Bi₂S₃.²⁷ The heterostructured BiVO₄/Bi₂S₃ exhibited superior photocurrent response and photocatalytic activity for the reduction of Cr^VI under visible light illumination. Besides, Lin et al. demonstrated that the Bi₂S₃ quantum dots sensitized SnO₂ porous film photoelectrode achieved a high photocurrent density of 0.9 mA cm⁻² under 0.109 sun illumination.²⁸ Also, there has been several reports on the hybridization of TiO₂ with Bi₂S₃ for PEC application.^29–31 For example, Zumeta-Dubé et al. reported the Bi₂S₃ quantum dots (QDs) sensitized TiO₂-P25 with PEC conversion efficiency of 0.84%.³²

In this paper, we report the heterostructured TiO₂/Bi₂S₃ core–shell NWAs synthesized by successive ionic layer adsorption and reaction of Bi₂S₃ onto the pre-grown TiO₂ NWAs for solar-light-driven PEC hydrogen generation. This hybrid TiO₂/Bi₂S₃ outstands itself from the early reported TiO₂/Bi₂S₃ composites by its one dimensional morphology. Comparing to the conventional II–VI semiconductors (such as CdS and CdSe) sensitized TiO₂, this composite TiO₂/Bi₂S₃ is environment-friendly and low-cost. The as-obtained photoanode exhibits strong absorption in the visible spectrum up to 800 nm. With light illumination, the optimized photoanode yields a photocurrent of ∼0.97 mA cm⁻² at 0.2671 V vs. RHE, which is 2.8 times of that of pure TiO₂ samples. Moreover, the TiO₂/Bi₂S₃ composite photoanodes induce a cathodic shift of the photocurrent onset potential. The excellent PEC properties of our photoanodes show that the hybridization of Bi₂S₃ on the TiO₂ NWAs has potential applications in QDs sensitized solar cells, PEC hydrogen generation and photocatalysis.

Experimental details

Preparation of TiO₂ NWAs by hydrothermal method

TiO₂ NWAs were grown onto fluorine-doped tin oxide (FTO) substrates using a hydrothermal method reported previously.³³ Briefly, 0.45 ml titanium n-butoxide and 0.6 g NaCl were added into the 30 ml diluted hydrochloric acid (mass fraction 18.25–19%) and stirred until the mixture became clear. After that, the above precursor was transferred into a 100 ml Teflon-lined stainless steel autoclave in which the FTO substrates were placed with an angle against the in-wall with the conductive side facing down. The hydrothermal process was conducted in an electric oven at 150 °C for 8 h. After cooling, the resultant samples were removed and rinsed with water to get rid of any residual reactants and dried in air at 80 °C.

Preparation of Bi₂S₃-sensitized TiO₂ NWAs by SILAR

For Bi₂S₃ sensitization, the as-synthesized TiO₂ NWAs were successively immersed into two different solutions for 1 min each, first in 0.05 M Bi(NO₃)₃ solution and then in 0.1 M Na₂S in aqueous solution. The Bi(NO₃)₃ solution was ultrasonic dissolved with the pH adjusted to 1–2 by concentrated nitric acid. Following each immersion, the electrodes were rinsed with deionized water and dried with a nitrogen stream to remove excess precursors before the next dipping. The two-step dipping procedure was called one SILAR cycle. Sample that went through n SILAR cycles is referred to as TiO₂/Bi₂S₃(n). Finally, the as-prepared TiO₂/Bi₂S₃ hybrid electrodes and the bare TiO₂ electrode were annealed at 250 °C in Ar atmosphere for 1 hour.

Structural and optical characterizations

The morphology and microstructure of the samples were characterized using the field emission scanning electron microscope (NOVA NANOSEM 450, FEI, USA) equipped with an energy-dispersive X-ray spectroscopy (EDS) and transmission electron microscopy (TEM, JEM2010-HR, 200 kV). Raman spectra of the as-obtained samples were recorded on a Renishaw InVia system with a laser operating at λ = 532 nm as the excitation source. The crystal structure was characterized by a Rigaku Dmax 2500 X-ray diffractometer with Cu K_α radiation. The light absorption was recorded by UV-vis spectrophotometer (UV-2500, Shimadzu, Japan).

PEC measurements

The PEC properties of the photoelectrodes were studied on an electrochemical workstation (CH Instruments, model CHI660E). The three-electrode system was used with the TiO₂ or TiO₂/Bi₂S₃ NWAs on FTO substrates as the working electrodes, a Pt foil as counter electrode and Ag/AgCl as reference electrode. A mixture aqueous solution of 0.24 M Na₂S and 0.35 M Na₂SO₃ (pH = 11.5) was used as the electrolyte and sacrificial reagent for all PEC measurements. A 150 W Xe lamp was used as the light source. Before test, the incident light intensity was calibrated to 100 mW cm⁻² with a digital power meter. The photocurrent densities vs. measured potential (i–v curves) measurements were carried out at a scan rate of 25 mV s⁻¹. The i–t curves were measured with a chopped light illumination at −0.6 V bias vs. Ag/AgCl. Electrochemical impedance spectroscopy (EIS) was measured under light at the open-circuit potential, with frequency range from 0.01 Hz to 100 kHz and the modulation amplitude of 5 mV. Mott–Schottky plots were measured in the dark at an AC frequency of 1.0 kHz. For all PEC measurements, the active area of the working electrodes was strictly kept within 1 cm². Hereafter, the electrode potential versus the Ag/AgCl is converted to the reversible hydrogen electrode (RHE) potential according to the Nernst equation³⁴

E_RHE = E_Ag/AgCl + 0.059pH + E^θ_Ag/AgCl

where E_RHE is the converted potential vs. RHE, E^θ_Ag/AgCl = 0.1976 V at 25 °C, and E_Ag/AgCl is the experimental potential measured against the Ag/AgCl reference electrode.

Results and discussion

Fig. 1a shows the typical top view SEM image of the bare TiO₂ NWAs. As seen, the TiO₂ NWAs are uniformly grown on the entire substrate with diameter around 75 nm, and the rod exhibits a regular tetragonal growth. For composite photoanodes, we synthesized three TiO₂/Bi₂S₃ core–shell NWAs samples with 10, 20 and 30 SILAR circles, and referred to them as TiO₂/Bi₂S₃(10), TiO₂/Bi₂S₃(20) and TiO₂/Bi₂S₃(30), respectively. After the deposition of Bi₂S₃ layer, the surfaces of the TiO₂ NW become rough. Fig. 1b shows the low magnification SEM image of the TiO₂/Bi₂S₃ NWAs with 20 SILAR deposition circles. Compare to Fig. 1a, Bi₂S₃ modified TiO₂ shows similar NWA feature but the image is slightly bright, suggest that the Bi₂S₃ sensitizers make the surfaces more conductive. Fig. 1c shows the morphology of the TiO₂/Bi₂S₃(20) sample under higher magnification. As can be seen from Fig. 1c, Bi₂S₃ is uniformly deposited onto the TiO₂ NW surfaces and the diameter TiO₂/Bi₂S₃(20) is increased to approximately 100 nm. Hence, the Bi₂S₃ deposition thickness is approximately 13 nm. These observations indicate the successful synthesis of core–shell TiO₂/Bi₂S₃ NWA heterostructures. The morphologies of the composite TiO₂/Bi₂S₃ NWAs with different SILAR circles are also exhibited in Fig. S1(a)–(c) (see ESI†). With increasing of SILAR circles, a larger diameter and rougher surface of TiO₂/Bi₂S₃ composite NWAs are displayed. In addition, some overdosed Bi₂S₃ nanoparticles are blocked at the interspaces of NWAs when the SILAR circles increased up to 30. This overloading of Bi₂S₃ is detrimental for PEC performance of TiO₂/Bi₂S₃ photoelectrode which will be discussed later. The TEM images of the Fig. S1(d)–(f)† give a more straight-forward observation of the thickness of the Bi₂S₃ layer with different SILAR circles, from which we can also reckon the particle-initiated growth behavior of the Bi₂S₃ layer. When the SILAR circle is 10, only a few separated Bi₂S₃ particles are decorated on the TiO₂ NW, the average diameter of the Bi₂S₃ particles is 20 nm. Increasing the SILAR circle up to 20, more Bi₂S₃ particles are loaded on the TiO₂ NW (dark dot in Fig. S1e†), and the particles tend to interlink to each other resulting in an overall covering of the TiO₂ NW surface. The 30 SILAR circles make a ∼25 nm thickness shell on the TiO₂ NW (inset in Fig. S1f†), and the particle-stacked Bi₂S₃ shell can be directly seen. Fig. 1d displays the EDS spectrum of the TiO₂/Bi₂S₃ NWAs, the atomic ratio of Bi : S is found to be ∼2 : 3, showing that the samples are in even stoichiometric. The other peaks of Ti, Si, Sn, Ca, and O are attributed to the chemical components of TiO₂ NWAs and the FTO glass substrate. Fig. 1e shows the TEM image of TiO₂/Bi₂S₃(20) core–shell NWs, it can be observed that the 1D TiO₂ NW is sheathed by the Bi₂S₃ nanoparticles throughout the length. Fig. 1f is the HRTEM image taken from the red square in Fig. 1e. The observed interplanar spacing of 0.397 nm corresponds to the (220) plane of orthorhombic Bi₂S₃, which is agree with the result observed by XRD below.

Fig. 1 SEM images of bare TiO₂ (a) and TiO₂/Bi₂S₃(20) (b) NWAs, respectively. (c) Magnified SEM image of TiO₂/Bi₂S₃(20) NWAs. (d) EDS of TiO₂/Bi₂S₃(20) NWAs. TEM (e) and HRTEM (f) images of TiO₂/Bi₂S₃(20) NWAs, respectively.

The XRD patterns of the as-obtained samples were collected to character their crystallographic phases (Fig. 2a). The pattern of the pristine TiO₂ NWAs (black line in Fig. 2a) reveal that the TiO₂ have a rutile structure (JPCDS 21-1276). Apart from the peaks of TiO₂ and FTO substrate (JPCDS 77-0448), the additional diffraction peaks (red line in Fig. 2a) can be readily indexed to the orthorhombic phase of Bi₂S₃ (JCPDS 17-0320) for TiO₂/Bi₂S₃ composite electrode. The main peaks at 2θ of 23.7, 24.9, 25.2, 28.6, and 31.76 are corresponding to the (101), (130), (310), (211) and (221) planes of orthorhombic Bi₂S₃, respectively. Fig. 2b shows the Raman spectra of the TiO₂ and TiO₂/Bi₂S₃ NWAs photoanodes. The pristine TiO₂ exhibits two sharp peaks at 447 and 608 cm⁻¹, which can be attributed to the E_g and A_1g mode in rutile TiO₂ as reported by Portq et al.³⁵ The broad peak at 239 cm⁻¹ is of a complex nature, which is relevant to the second-order scattering and disorder effects. The coating of Bi₂S₃ layer weakens characteristic peaks of the TiO₂ and three new peaks at 237, 261 and 187 cm⁻¹ are emerged. The added peaks can be attributed to the A_g, B_g1 and A_g mode of the Bi₂S₃,^36,37 which indicates the crystalline character of Bi₂S₃ layer in the TiO₂/Bi₂S₃ composite NWAs.

Fig. 2 (a) XRD patterns and (b) Raman scattering spectra of the pristine TiO₂ and composite TiO₂/Bi₂S₃ NWAs on FTO substrates.

The optical absorption properties of the TiO₂ and TiO₂/Bi₂S₃ were characterized by absorption spectra, as shown in Fig. 3. We can see that the absorption wavelength of TiO₂ NWAs is less than 410 nm. In the case of the composite TiO₂/Bi₂S₃ NWAs, its absorption extends to cover the whole visible light range and even beyond 800 nm. In addition, UV-vis absorption spectra show that as the SILAR deposition circle increases the light absorption ability also increases. The relationship between the incident photon energy and the absorption coefficient is given by the following equation: (αhν)² = A(hν − E_g), where α, h, ν and E_g are the optical absorption coefficient, Planck constant, the photon frequency and band gap of semiconductor, respectively. A is a constant. The band-gap values of TiO₂ and TiO₂/Bi₂S₃(20) are derived to be 3.02 and 1.6 eV, respectively. This enhancement of light trapping effect for TiO₂/Bi₂S₃(20) is beneficial for the PEC hydrogen generation application.

Fig. 3 Normalized UV-vis absorption spectra of TiO₂ and TiO₂/Bi₂S₃ core–shell NWAs on FTO substrates.

To evaluate the Bi₂S₃ sensitization effect on PEC performance of TiO₂, Fig. 4a shows the i–v curves of the bare TiO₂ and the TiO₂/Bi₂S₃ core–shell NWAs with 10, 20 and 30 SILAR circles. By analyzing the experimental data in Fig. 4a, three conclusions can be reached. First, for all of the TiO₂/Bi₂S₃ NWAs photoanodes, the saturate photocurrent densities are much higher than that of the pristine TiO₂. The TiO₂/Bi₂S₃(20) sample achieved a value of 0.98 mA cm⁻² at 0.2671 V vs. RHE (−0.6 V vs. Ag/AgCl), which is higher than that of the bare TiO₂ (0.35 mA cm⁻²) and other composite photoanodes (0.8 mA cm⁻² for TiO₂/Bi₂S₃(10) and 0.95 mA cm⁻² for TiO₂/Bi₂S₃(30)). It confirms that the coating of Bi₂S₃ is a simple but effective method for enhancing the PEC performance of TiO₂. Second, as compared to TiO₂, the onset potential of TiO₂/Bi₂S₃ is much negatively shifted and leads to the photocurrent saturation at a more negative potential. Typically, the photocurrent increases from the onset potential around −0.072 V vs. RHE and approaches a plateau at 0.05 V vs. RHE for TiO₂/Bi₂S₃(20) sample, whereas the photocurrent plateau in the bare TiO₂ photoanode is 0.3 V vs. RHE. Achieving the saturated photocurrent level at a more negative potential is significantly important because it reduces the applied external bias, and thus increases the overall efficiency of the PEC cells.²² Third, the TiO₂/Bi₂S₃(20) overmatches the one with 10 or 30 SILAR circles deposition. As the Bi₂S₃ deposition increasing, the light absorption of TiO₂/Bi₂S₃ is also increased (see Fig. 3). However, too much Bi₂S₃ deposition blocks the porous structure of the electrodes (see SEM images of Fig. S1 in ESI†) and lengthens the electron diffusion path from Bi₂S₃ to TiO₂. In this case, a more positive external bias is needed to separate the photo-induced charge efficiently for TiO₂/Bi₂S₃(30), which can explain the smaller photocurrent between 0 and 0.3 V vs. RHE but similar photocurrent when the potential exceeds 0.35 V vs. RHE in comparison with TiO₂/Bi₂S₃(20). Fig. 4b shows the i–t curves of the TiO₂ and the TiO₂/Bi₂S₃ core–shell NWAs photoelectrodes at 0.2671 bias vs. RHE under chopped light irradiation. Upon illumination, the photocurrent of TiO₂/Bi₂S₃ flashily rise to a high level, and decay to a relatively steady-state after a few second, displaying a series of spikes in the photoresponse. This behavior might be attributed to the charge carrier accumulation at the electrode–electrolyte interface due to the slow oxygen evolution kinetics or oxidization of trap states on the surface and in the bulk by charge carriers.³⁸ And when the light is cut off, the photocurrent densities drop to nearly zero instantaneously. Such fast rise and fall of the photocurrents indicates that carrier transport in the heterojunction material proceeds very quickly.

Fig. 4 (a) i–v curves under AM 1.5G in the dark and at 100 mW cm⁻² illumination for pristine TiO₂ NWAs and TiO₂/Bi₂S₃ core–shell NWAs with different circles SILAR. (b) Amperometric i–t curves of the TiO₂/Bi₂S₃(10, 20, 30) core–shell NWAs at 0.2761 V versus RHE with light on/off cycles.

Fig. 5a shows the Mott–Schottky (M–S) plots, in which the flat band potential at the electrode–electrolyte interface can be estimated by the M–S equation³⁹

1/C² = (2/εε₀eN_dA²)[(V − V_FB) − kT/e₀]

where C is the specific capacity, ε is the dielectric constant of Bi₂S₃ or TiO₂, ε₀ is the electric permittivity of vacuum, N_d is the carrier density, A is the area, V is the applied potential, V_FB is the flat band potential, k is the Boltzmann constant, T is the absolute temperature, and e₀ is the electron charge. The V_FB values of the TiO₂/Bi₂S₃ and TiO₂ NWAs are determined to be 0.07 V and 0.26 V vs. RHE by extrapolating the X-intercepts of the linear region in the M–S plot, respectively. Such a negative shift of the flat band for the TiO₂/Bi₂S₃ electrode suggests a higher carrier concentration and lower charge recombination in the TiO₂/Bi₂S₃ electrode in comparison with the pristine TiO₂. In addition, the positive slope indicates that the TiO₂/Bi₂S₃ NWA is an n-type semiconductor. The TiO₂/Bi₂S₃ shows substantially smaller M–S plot slope compared to the TiO₂, suggesting an increase of donor density in the TiO₂/Bi₂S₃.⁹ Using the resulted V_FB and band gap values of the TiO₂/Bi₂S₃ and TiO₂, the energy band level aliments can be estimated as displayed in Fig. 5b.⁴⁰ The charge transfer and hydrogen production mechanism in the TiO₂/Bi₂S₃ PEC cell is proposed. When the TiO₂/Bi₂S₃ heterojunction is irradiated by solar light, electrons are excited to the conduction band (CB) of Bi₂S₃ and TiO₂. Subsequently, electrons migrate from Bi₂S₃ to TiO₂ and finally to the counter electrode to produce H₂. At the same time, the holes are transported to the valence band (VB) of Bi₂S₃ and then scavenged by the sacrificial reagents at the electrode–electrolyte interface. Thus, the photogenerated charge separation efficiency is improved because of the formation of type-II heterojunction interface in the TiO₂/Bi₂S₃ electrode, which is also responsible for its high photocurrent density.

Fig. 5 (a) Mott–Schottky plots of pristine TiO₂ and TiO₂/Bi₂S₃(20) core–shell NWAs. (b) Schematic of energy band level aliment and carrier transferring mechanism in TiO₂/Bi₂S₃ core–shell NWAs. (c) Nyquist plots and (d) Bode plots of the impedance spectra measured at open circuit condition under 1 sun illumination.

To further examine the intrinsic electronic properties of the TiO₂/Bi₂S₃ NWAs, the EIS were collected and shown in Fig. 5c. Normally, the smaller the radius in Nyquist plot, the lower the charge transfer impedance at the electrode–electrolyte interface for corresponding electrode.⁴¹ The TiO₂/Bi₂S₃ composite electrodes represent much smaller radius than the pristine TiO₂ electrode, implying its more effective charge carrier separations compared to the pristine TiO₂. The inset in the Nyquist plots is the equivalent circuit model of the PEC cell, where the R_s is the series resistance of the electrochemical device, R and C represent the resistance and capacitance of space charge region, respectively, including charge transfer across the FTO/TiO₂–Bi₂S₃ interface and the TiO₂–Bi₂S₃/electrolyte interface, R_Pt and C_Pt are the resistance and capacitance of the Pt cathode, respectively. By fitting the experimental data (dot points) into solid line using the equivalent circuit model, a more quantitative approach to compare the interface charge transfer impedance of TiO₂ and TiO₂/Bi₂S₃ is reached. As a result, the determined value of R is 3270 and 735 Ω for pristine TiO₂ and composite TiO₂/Bi₂S₃ photoelectrodes, respectively. Correspondingly, the TiO₂/Bi₂S₃ heterojunction shows a lower characteristic peak frequency in the Bode phase plot in Fig. 5d, implying the longer electron lifetime than that of the pristine TiO₂ electrode.⁴² The enhanced charge separation and prolonged electron lifetime in the TiO₂/Bi₂S₃ composite is also responsible for the improved PEC performance for TiO₂.

Conclusions

In summary, we have successfully synthesized the TiO₂/Bi₂S₃ core–shell NWAs as photoelectrode for PEC hydrogen generation. The coating of Bi₂S₃ layer achieves a remarkable enhancement of photocurrent density in the TiO₂ NWAs photoanode. Moreover, the onset potential of the composite photoanode is also largely shifted to a negative direction. The broadened light absorption and improved charge carrier separation efficiency in the type-II TiO₂/Bi₂S₃ core–shell NWAs are responsible for the PEC performance enhancement. The present work provides a new photoelectrode with non-toxicity, low cost and easy fabrication, which may have potential application in PEC hydrogen generation and photocatalysis.

Acknowledgements

We thank the financial support of the National Natural Science Foundation of China (Grant no. 51202208, 51172191, and 11274264), The Project supported by Scientific Research Fund of Hunan Provincial Education Department (Grant no. 12B129), National Basic Research Program of China (no. 2012CB921303) and the Program for Changjiang Scholars and Innovative Research Team in University (IRT13093).

References

1. A. Fujishima and K. Honda, Nature, 1972, 238, 37–38.
2. S. U. M. Khan, Science, 2002, 297, 2243–2245.
3. I. Mora-Seró, T. L. Villarreal, J. Bisquert, Á. Pitarch, R. Gómez and P. Salvador, J. Phys. Chem. B, 2005, 109, 3371–3380.
4. P. R. Mishra, P. K. Shukla and O. N. Srivastava, Int. J. Hydrogen Energy, 2007, 32, 1680–1685.
5. J. H. Pan, X. Z. Wang, Q. Huang, C. Shen, Z. Y. Koh, Q. Wang, A. Engel and D. W. Bahnemann, Adv. Funct. Mater., 2014, 24, 95–104.
6. H. Cui, W. Zhao, C. Yang, H. Yin, T. Lin, Y. Shan, Y. Xie, H. Gu and F. Huang, J. Mater. Chem. A, 2014, 2, 8612.
7. X. Zhang and Z. Chen, RSC Adv., 2015, 5, 9482–9488.
8. Q. Dong, H. Yu, Z. Jiao, G. Lu and Y. Bi, RSC Adv., 2014, 4, 59114–59117.
9. Q. Kang, J. Cao, Y. Zhang, L. Liu, H. Xu and J. Ye, J. Mater. Chem. A, 2013, 1, 5766.
10. W. Wu, B. Lei, H. Rao, Y. Xu, Y. Wang, C. Su and D. Kuang, Sci. Rep., 2013, 3, 1352,  DOI:10.1038/srep01352.
11. P. Roy, D. Kim, K. Lee, E. Spiecker and P. Schmuki, Nanoscale, 2010, 2, 45.
12. B. Liu and E. S. Aydil, J. Am. Chem. Soc., 2009, 131, 3985–3990.
13. A. Wolcott, W. A. Smith, T. R. Kuykendall, Y. Zhao and J. Z. Zhang, Small, 2009, 5, 104–111.
14. K. Hara, K. Sayama, H. Arakawa, Y. Ohga, A. Shinpo and S. Suga, Chem. Commun., 2001, 569–570.
15. N. Jain, Y. Zhu, D. Maurya, R. Varghese, S. Priya and M. K. Hudait, J. Appl. Phys., 2014, 115, 24303.
16. D. O. Scanlon, C. W. Dunnill, J. Buckeridge, S. A. Shevlin, A. J. Logsdail, S. M. Woodley, C. R. A. Catlow, M. J. Powell, R. G. Palgrave, I. P. Parkin, G. W. Watson, T. W. Keal, P. Sherwood, A. Walsh and A. A. Sokol, Nat. Mater., 2013, 12, 798–801.
17. Q. Liu, H. Lu, Z. Shi, F. Wu, J. Guo, K. Deng and L. Li, ACS Appl. Mater. Interfaces, 2014, 6, 17200–17207.
18. M. Khan, P. Jiang, J. Li and W. Cao, J. Appl. Phys., 2014, 115, 153103.
19. C. Dette, M. A. Pérez-Osorio, C. S. Kley, P. Punke, C. E. Patrick, P. Jacobson, F. Giustino, S. J. Jung and K. Kern, Nano Lett., 2014, 14, 6533–6538.
20. Y. Li, W. Wu, P. Dai, L. Zhang, Z. Sun, G. Li, M. Wu, X. Chen and C. Chen, RSC Adv., 2014, 4, 23831.
21. S. Kurian, H. Seo and H. Jeon, J. Phys. Chem. C, 2013, 117, 16811–16819.
22. G. Wang, H. Wang, Y. Ling, Y. Tang, X. Yang, R. C. Fitzmorris, C. Wang, J. Z. Zhang and Y. Li, Nano Lett., 2011, 11, 3026–3033.
23. B. Liu, H. M. Chen, C. Liu, S. C. Andrews, C. Hahn and P. Yang, J. Am. Chem. Soc., 2013, 135, 9995–9998.
24. C. Wang, Z. Chen, H. Jin, C. Cao, J. Li and Z. Mi, J. Mater. Chem. A, 2014, 2, 17820–17827.
25. I. S. Cho, Z. Chen, A. J. Forman, D. R. Kim, P. M. Rao, T. F. Jaramillo and X. Zheng, Nano Lett., 2011, 11, 4978–4984.
26. T. Wu, X. Zhou, H. Zhang and X. Zhong, Nano Res., 2010, 3, 379–386.
27. X. Gao, H. B. Wu, L. Zheng, Y. Zhong, Y. Hu and X. W. D. Lou, Angew. Chem., Int. Ed., 2014, 53, 5917–5921.
28. Y. C. Lin and M. W. Lee, J. Electrochem. Soc., 2013, 161, H1–H5.
29. J. Kim and M. Kang, Int. J. Hydrogen Energy, 2012, 37, 8249–8256.
30. R. Brahimi, Y. Bessekhouad, A. Bouguelia and M. Trari, Catal. Today, 2007, 122, 62–65.
31. L. M. Peter, K. G. U. Wijayantha, D. J. Riley and J. P. Waggett, J. Phys. Chem. B, 2003, 107, 8378–8381.
32. I. Zumeta-Dubé, V. Ruiz-Ruiz, D. Díaz, S. Rodil-Posadas and A. Zeinert, J. Phys. Chem. C, 2014, 118, 11495–11504.
33. G. Ai, R. Mo, H. Xu, Q. Chen, S. Yang, H. Li and J. Zhong, J. Appl. Phys., 2014, 116, 174306.
34. M. Gratzel, Nature, 2001, 414, 338–344.
35. S. P. S. Portq, P. A. Fleury and T. C. Damen, Phys. Rev., 1967, 154, 522.
36. T. Lutz, A. MacLachlan, A. Sudlow, J. Nelson, M. S. Hill, K. C. Molloy and S. A. Haque, Phys. Chem. Chem. Phys., 2012, 14, 16192.
37. Y. Zhao, K. T. E. Chua, C. K. Gan, J. Zhang, B. Peng, Z. Peng and X. Qihua, Phys. Rev. B: Condens. Matter Mater. Phys., 2011, 84, 205330.
38. 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.
39. T. Majumder, J. J. L. Hmar, K. Debnath, N. Gogurla, J. N. Roy, S. K. Ray and S. P. Mondal, J. Appl. Phys., 2014, 116, 34311.
40. A. B. Ellis, S. W. Kaiser, J. M. Bolts and M. S. Wrighton, J. Am. Chem. Soc., 1977, 99, 2839–2848.
41. Y. Hou, F. Zuo, A. Dagg and P. Feng, Angew. Chem., 2013, 125, 1286–1290.
42. Q. Liu, H. Lu, Z. Shi, F. Wu, J. Guo, K. Deng and L. Li, ACS Appl. Mater. Interfaces, 2014, 6, 17200–17207.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c4ra15820b

---