Design of a solar-driven TiO₂ nanofilm on Ti foil by self-structure modifications

Abstract

A novel solar-driven TiO₂ nanofilm (V_O–N–TiO₂ (A/R) nanofilm), which was obtained by self-structure modifications, was designed on a Ti substrate surface. The V_O–N–TiO₂ (A/R) nanofilm showed a three-dimensional pine branched-like structure and a disordered lattice structure, which consisted of V_O, N-doping and an anatase/rutile phase, was built in the V_O–N–TiO₂ (A/R) nanofilm. The experimental results show that the band gap of the V_O–N–TiO₂ (A/R) nanofilm can be narrowed to 0.72 eV and 2.4 eV and that it exhibited excellent optical absorption capacity in the 200–2500 nm region. Using the V_O–N–TiO₂ (A/R) nanofilm to degrade methyl orange, the removal efficiency could reach 96% within 150 min under sunlight, which shows its high photoelectrocatalytic activity and stability.

---

1 Introduction

Titanium dioxide (TiO₂) is one of the most promising multifunctional materials that can be used in extensive applications such as photocatalysis, air purification, photovoltaic cells, solar cells, and water splitting.^1–3 However, TiO₂ can only be excited under UV irradiation (<387.5 nm) because of its large band gap (anatase of 3.2 eV and rutile of 3.0 eV).⁴ The introduction of dopants and coupling with narrow band gap semiconductors have been carried out to extend the optical absorption into the visible-light region.^5–7 N-doping of TiO₂ is considered as the optimal method to improve its photocatalytic activity under visible light.^8,9 N-doped TiO₂ was first synthesized by sputtering a TiO₂ target in an N₂ (40%)/Ar gas mixture, and this can extend the optical absorption to 550 nm.¹⁰ Recently, a N-doped TiO₂ film was prepared using the anodization process, and this film also has a photoresponse under visible light.¹¹ In addition, crystal phase of TiO₂, morphology, and disorder structure consisting Ti³⁺ and oxygen vacancy (V_O) have been proved to be effective approaches in improving its photochemical activity.¹² The mixed-phase of TiO₂ displays an enhanced photocatalytic activity due to the rapid transfer of electrons from rutile to anatase on the TiO₂ interface.^13,14 V_O-doped TiO₂ shows considerable potential in extending the visible light absorption.^15,16 The band gap of TiO₂ can be narrowed to 1.85 eV by the synergistic effect of V_O and surface disorder.¹⁷ Its optical absorption can be further extended to 1200 nm as a result of introducing a disordered structure in the nanophase TiO₂ surface using the hydrogenation process.¹⁸ Recent research has been carried out to investigate the fundamental properties and photocatalytic applications of hydrogenated black TiO₂.^19,20

In this study, we design a novel solar-driven TiO₂ nanofilm (V_O–N–TiO₂ (A/R) nanofilm) by adopting a new electrolyte system in the anodic oxidation process. A three-dimensional pine-bough-like structure nanofilm that has a disordered lattice structure, consisting of V_O, N-doping and an anatase/rutile phase, was obtained. These variations in structure were defined as “self-structural modification”. The optical absorption of the V_O–N–TiO₂ (A/R) nanofilm by self-structural modification covered the ultraviolet, visible and near-infrared (NIR) region from 200 nm to 2500 nm and displayed improved photoelectrocatalytic activity under sunlight.

2 Experimental

2.1 Synthesis of V_O–N–TiO₂ (A/R) nanofilm

Titanium foil (thickness 0.2 mm, purity 99.6%, Baoji Titanium Nickel Gold Port Equipment Manufacturing Co., Ltd.) was clipped to the size of 3.3 cm × 4 cm, then ultrasonically degreased in acetone and ethyl alcohol, and rinsed in distilled water for 15 min. The Ti sheets were chemically polished in a mixed solution of HF (40%), HNO₃ (65%) and distilled water (volume ratio 1 : 4 : 5) for 20 s and then rinsed in distilled water. The Ti sheet was used as the anode, and Pt foil was used as the counter electrode. Anodization was performed in an ethylene glycol solution containing HNO₃ (1 wt%) and water (1 wt%) at 40 V for 10 min using a DC power supply at room temperature. The samples were cleaned and dried and then annealed at 773 K for 1 h in a muffle furnace. The annealed samples were designated as V_O–N–TiO₂ (A/R) nanofilm.

2.2 Characterization

The morphology of the samples was observed using field-emission scanning electron microscopy (FE-SEM, S-4800, Hitachi). Crystalline structures were investigated by glancing angle X-ray diffraction (GLXRD) using Cu-Kα (0.15418 nm) radiation at 40 kV and 30 mA from 20° to 80° at 0.03° step and a glancing angle of 1°. Raman measurement was recorded in the range of 100–1000 cm⁻¹ using a Horiba Jobin Yvon LabRAM HR800 with 532 nm laser excitation, and the power of the sample was estimated to be 5 mW and the spectral resolution of the spectrometer was 1 cm⁻¹. The chemical state of the elements was evaluated using X-ray photoelectron spectroscopy (XPS, Perkin-Elmer Corporation, Eden Prairie, MN) with an Mg K anode (1253.6 eV photon energy, 15 kV, 300 W) at a takeoff angle of 45°. The binding energy values were calibrated using the contaminant carbon (C 1s = 284.4 eV) as a reference. Multiplex XPS spectra of Ti 2p, O 1s, and N 1s were recorded using the band-pass energy of 35.75 eV, which corresponds to an energy resolution of 1.2 eV. Atomic concentrations of these elements were obtained by comparing the peak areas of their spectra. The V_O identification was carried out using electron spin resonance spectroscopy (ESR, JES FA-200). The settings were as follows: 323.717 mT center field, 323.717 ± 25 mT sweep width, 9077.574 mHz microwave frequency, 100.00 kHz modulation amplitude, and 3.0 mW power. UV-vis-NIR absorption spectra were obtained using a TU-1901 spectrometer with the wavelength range of 200 nm–2500 nm. Photoluminescence (PL) spectra were obtained using a fluorescence spectrophotometer (FLsp 920) with excitation wavelengths of 280 and 354 nm.

2.3 Electrochemical performance

Photoelectrochemical properties were evaluated using an electrochemical workstation (CHI 830C Instruments, China) in the standard three-electrode configuration. A 0.5 M Na₂SO₄ solution was used as the electrolyte. The V_O–N–TiO₂ (A/R) nanofilm was used as the working electrode with an active area of 1 cm², and Pt foil and Ag|AgCl were used as the counter and reference electrodes, respectively. The I–t curve was used to evaluate the photocurrent density. Light irradiation intensity was recorded using an illuminometer (TES-1339).

2.4 Photoelectrocatalytic activity

The photoelectrocatalytic activity of the V_O–N–TiO₂ (A/R) nanofilm was evaluated by the decomposition of methyl orange (MO) under sunlight (MO concentration 20 mg L⁻¹, pH value 6.8 and volume 40 mL). A bias voltage of 4 V (the effect of different bias potentials on MO degradation see ESI Fig. S1†) was applied during the MO decomposition. The solution absorbance was measured using a UV-vis spectrophotometer (UV3000, China) at 464 nm. The decomposition rate was calculated using eqn (1),where C₀ is the initial absorbance of MO, and C_t is the absorbance after reaction time t.

3 Results and discussion

3.1 Morphologies and crystal structure

Irregular holes could be observed on the Ti substrate surface after chemical polishing (Fig. 1a). The flocculent structures were formed on the Ti substrate surface when the voltage of 40 V was applied to the two electrodes for 10 min (Fig. 1b). Interestingly, a three-dimensional pine-bough-like structure nanofilm was obtained after calcination at 773 K for 1 h in oxic atmosphere (Fig. 1c), and the pine-bough-like nanoclusters consisted of many nano-sized branches grown on the bole (Fig. 1d).

Fig. 1 FE-SEM images of the V_O–N–TiO₂ (A/R) nanofilm. (a) Ti substrate polished by chemical polishing, (b) as-anodized sample, (c) V_O–N–TiO₂ (A/R) nanofilm calcinated at 773 K for 1 h, and (d) enlarged scale of the red region in (c).

The as-anodized sample presented an amorphous structure, whereas the V_O–N–TiO₂ (A/R) nanofilm showed a mixed crystal structure of rutile and anatase (Fig. 2a). The mixed crystal parameters were calculated using Scherrer's equation²¹ (Table 1) and were different from their theoretical value. This is due to the disordered lattice structure, which was caused by the reconstruction of the anatase/rutile interfaces.²² Moreover, the following peaks were observed in the Raman spectrum (Fig. 2b). E_g (a) ∼ 146 cm⁻¹, E_g (a) ∼ 197 cm⁻¹, B_1g (a) ∼ 399 cm⁻¹, A_1g (a) + B_1g (a) ∼ 517 cm⁻¹, and E_g (a) ∼ 639 cm⁻¹ which were assigned to anatase.^18,23,24 E_g (r) ∼ 442 cm⁻¹ and A_1g (r) ∼ 619 cm⁻¹ were assigned to rutile.^23,25 Clearly, the two complementary methods, XRD and Raman spectroscopy, revealed that the as-synthesized TiO₂ film was composed of an anatase and rutile phase.

Fig. 2 (a) XRD patterns for the as-anodized sample and V_O–N–TiO₂ (A/R) nanofilm and (b) Raman scattering pattern of the V_O–N–TiO₂ (A/R) nanofilm.

Table 1 Effect of annealing on the XRD parameters

Annealing	2-Theta (°)	Phase ID	Area	FWHM	Crystallite size (nm)	Lattice parameters		Cell volume (Å^3)
						a = b	c
Unannealed	—	Amorphous	—	—	—	—	—	—
Annealed	25.45	Anatase	566	0.401	16.01	3.785	9.514	142.86
	27.39	Rutile	680	0.495	16.24	4.953	2.959	72.59

---

3.2 Chemical state of the elements

The V_O–N–TiO₂ (A/R) nanofilm contained Ti, O, N, and C (Fig. 3a). The amount of N is ∼3.79 at%, which is higher than that mentioned in previous reports.^11,26,27 The C element was introduced from residual organic carbons.²⁸ The peaks at 458.47 and 464.12 eV were attributed to Ti⁴⁺ from Ti 2p^3/2 and 2p^1/2, respectively (Fig. 3b).²⁹ However, the binding energy of Ti 2p^3/2 shifted ∼0.3 eV towards a lower binding energy, which shows the presence of V_O and a few N species.^30,31 The peak at 399.6 eV was attributed to the interstitial nitrogen dopant^32,33 (Fig. 3c). The peak at 407.2 eV originated from the nitrate species in the anodization process.^34,35 The peak at 529.8 eV was attributed to O–Ti–O in the TiO₂ crystal lattice (Fig. 3d), whereas the peak at 531.6 eV was assigned to the hydroxyl species.^16,36

Fig. 3 (a) Full XPS spectra of the V_O–N–TiO₂ (A/R) nanofilm, (b) multiplex high-resolution scan over Ti 2p, (c) N 1s, (d) O 1s, and (e) low temperature ESR spectra of the V_O–N–TiO₂ (A/R) nanofilm.

The signal at g = 2.003 in the ESR spectrum (Fig. 3e) confirmed the presence of V_O in the V_O–N–TiO₂ (A/R) nanofilm,³⁷ and the concentration of V_O was approximately 1.0 × 10¹⁵. N doping always leads to thermal instability in TiO₂ and thus increases V_O formation.^38,39

3.3 Formation mechanism of the V_O–N–TiO₂ (A/R) nanofilm

Based on the abovementioned experimental results and analysis, the formation mechanism of the V_O–N–TiO₂ (A/R) nanofilm, which contains three stages, was proposed.

At the first stage, nitric ions are forced to accumulate at the titania/solution interface under an electric field and attack the oxide layer on the surface of the Ti substrate to form a soluble complex such as [TiO_2−x(NO₃)_x]^m−n (m > n, 0 < x < 2), which leads to the local dissolution/thinning of the oxide layer and the penetration of the oxide layer or irregular holes into the Ti substrate (Fig. S2†).⁴⁰ Similarly, the accumulated nitric anions form a soluble complex such as [Ti(NO₃)_n]^m−n (m = 3, 4; n > 4) in concurrence with the formation of titania, which results in pore generation and growth along the electric field direction. Nevertheless, the complex anions [Ti(NO₃)_n]^m−n are thermodynamically unstable and further decompose through electrochemical reactions into the thermodynamically stable TiO₂ and interstitial nitrogen dopant.⁴⁰ Moreover, the NO species derived from the decomposition of nitric acid could be chemisorbed on the surface or incorporated into the lattice of the resultant TiO₂ (ref. 41) as indicated by XPS. The formation of flocculent structure films on the Ti substrate is dependent on the anodic oxidation rate of titanium and the anodic dissolution rate of the resultant titania in the electrolyte. Under the current conditions, the dynamic equilibrium of the anodization-dissolution rate led to the uniform growth of flocculent structures.

At the second stage, the flocculent structure film is further dehydrated, crystallized and expanded during calcination. The anatase-to-rutile transformation was conducted at the wide temperature range of 673–1373 K.⁴² The rutile phase evolved mainly from the inner regions of the agglomerated TiO₂ particles, and the anatase phase emerged in the surface region during the phase transformation because of the asymmetric heat dissipation at the interface.^43,44 The Vo defects increase with the increase in the annealing temperature, and N doping can increase V_O formation.^42,45 In addition, the interface between the Ti substrate and oxide layer may react as Ti + TiO₂ → TiO_x (x = 0–2), which may form more V_O.⁴⁶ Such V_O can ulteriorly promote the anatase-to-rutile transformation, because it involves the overall shrinkage of the oxygen structure and a cooperative movement of ions.^47,48 Due to the different thermal expansion coefficients of anatase (10.2 × 10⁻⁶ K⁻¹) and rutile (7.14 × 10⁻⁶ K⁻¹),⁴⁹ the outer anatase layer grew faster than the inner rutile core during calcinations. The flocculent structures ultimately evolved into three-dimensional pine-bough-like structures after calcination for 1 h.

At the last stage, fast cooling is key process for the formation of a disordered structure and V_O defects. The metastable defective phase on the surface such as V_O can subsequently become nearly stoichiometric when the sample is calcined at high temperatures and is quickly exposed to oxygen-rich air.¹⁷ This means that the V_O defects on the surface immediately decreases and reaches a steady state in oxygen-rich conditions,⁵⁰ but the V_O defects still maintains a high level in the internal lattice on the lower surface.

3.4 Optical absorption

Ti foil almost has no optical absorption (Fig. 4a). However, the optical absorption of the V_O–N–TiO₂ (A/R) nanofilm covered the ultraviolet, visible and near-infrared (NIR) region from 200 nm to 2500 nm. One of the absorption edges was extended to 475 nm. The extension of light absorption from UV to visible light is attributed to the localized states near the valence band (VB), which is induced by interstitial nitrogen atoms.^8,11 Another huge absorption peak appeared at ∼830 nm and extended to 2500 nm, which was due to the disordered lattice structure that was introduced by the anatase/rutile interface and V_O.^14,17,18 The multi-peaks in the absorbance spectra indicated that the band gap of the V_O–N–TiO₂ (A/R) nanofilm was drastically narrowed by intra-band transitions.¹ The band gap was evaluated by the plots of (Ahv)^1/2 vs. excitation energy (hv).²⁹ Fig. 4b shows that the band gaps were narrowed to 0.72 eV and 2.4 eV.

Fig. 4 (a) UV-vis-NIR absorption spectra of the V_O–N–TiO₂ (A/R) nanofilm and Ti foil, and (b) the variation of (Ahv)^1/2 vs. excitation energy (hv) for the V_O–N–TiO₂ (A/R) nanofilm.

To further examine the electronic structure and defect levels of the V_O–N–TiO₂ (A/R) film, we analyzed the valence band (VB) edge spectra, which were obtained using XPS and PL emission spectroscopy. As shown in Fig. 5a, the typical value of the VB maximum (VMB) for TiO₂ NTs was located at ∼1.2 eV below the Fermi energy.¹⁸ However, for the V_O–N–TiO₂ (A/R) film, the VMB was located at ∼0.4 eV, and the band tail continuously extended to −0.28 eV. That is, a blue shift about 1.48 eV toward the vacuum level was observed, which indicated the presence of extra states above the VB.^18,51

Fig. 5 (a) Valence band spectra from XPS and (b) a schematic of the band alignment model based on the XPS VB spectra and UV-vis-NIR spectra. The work function of TiO₂ is 4.1 eV, which corresponds to the energy distance from the Fermi level (E_f) to the vacuum level (E_vac).

The VB edge at ∼0.4 eV was introduced by the N 2p band level located at 0.8 eV above the VB.^31,52,53 As observed in the PL emission spectra (Fig. S3(a)†), the peaks at 515 nm (2.4 eV) and 561 nm (2.2 eV) correspond to the band energy from the N 2p to conduction band (CB) for anatase and rutile.^54,55 Surface disorder, which is induced by hydrogenation, can induce a substantial shift of 2.18 eV above the VB position,¹⁸ and that caused by carbon doping can cause a shift of 1.6 eV toward the vacuum level.⁵¹ Core–shell black TiO₂ with the fast cooling step also induces a substantial shift of 1.5 eV above the VBM.¹⁷ In our case, a number of states, which were caused by the presence of interstitial N atoms and interface disorder of anatase/rutile, were possible for a maximum energy of 1.48 eV above the VBM of V_O–N–TiO₂ (A/R).

Furthermore, the high concentration of V_O can create shallow levels at 0.7–1.0 eV below the CBM.^17,56,57 Serried peaks from 506 nm (2.45 eV) to 590 nm (2.1 eV) were observed in the PL emission spectra (Fig. S3(b)†), which were assigned to the recombination of defective states from the V_O to VB for anatase and rutile.⁵⁸ In addition, the peaks at 437 nm (2.8 eV) and 456.7 nm (2.7 eV) can be ascribed to self-trapped excitons located at the TiO₆ octahedra.²⁹ Based on the abovementioned analysis, a schematic of the band alignment model from the VB to CB was constructed and is shown in Fig. 5b, and a schematic of the electronic transitions in different energy levels is summarized in Fig. 6.

Fig. 6 Schematic of the electronic transitions in different energy levels in the V_O–N–TiO₂ (A/R) nanofilm.

The CB edge of TiO₂ (A) is about 0.2 eV higher than that of TiO₂ (R).⁵⁹ All of the excitation energies from (1) to (14) existed in the composite system. The excited wavelength and energy gap for each excited channel are listed in Table 2. Only channel (1) needs to be excited by UV light irradiation (<388 nm). Channels (2) to (6) can be excited by visible light (<721 nm), and (7) to (14) can even be excited by near-infrared light (<1824 nm).

Table 2 Excited wavelength and energy gap for each excited channel

Channels	1	2	3	4	5	6	7	8	9	10	11	12	13	14
Eg (eV)	3.2	3.0	2.4	2.2	2.2	1.72	1.52	1.48	1.4	1.0	0.8	0.8	0.72	0.68
λ (nm)	388	413	517	564	564	721	816	838	886	1240	1550	1550	1722	1824

---

3.5 Electrochemical performance

The photocurrent density of the V_O–N–TiO₂ (A/R) nanofilm was approximately 20 μA cm⁻² under solar irradiation, whereas that of the Ti foil was almost zero (Fig. 7). The illumination intensity of sunlight and temperature are shown in Table 3.

Fig. 7 Photocurrent density of the V_O–N–TiO₂ (A/R) nanofilm and Ti foil measured at zero bias voltage versus Ag|AgCl in 0.5 M Na₂SO₄ aqueous solution.

Table 3 Illumination intensity of sunlight and temperature in the photocatalytic process

Degradation time (min)	30	60	90	120	150
Illumination intensity (lx)	49 045	54 890	55 360	55 430	53 830
Temperature (K)	283	283	284	284	283

---

The removal efficiency of MO reached 96% for V_O–N–TiO₂ (A/R), whereas the control sample was only 66% within 150 min under sunlight (Fig. 8a). In addition, neither the photocatalysis process (PC, 0 V bias) nor the electrocatalysis process (EC, with 4 V bias in the dark) could effectively decompose MO. Therefore the photoelectrocatalysis (PEC) process exhibited a notable synergistic effect. The photoelectrocatalytic activity did not exhibit any reduction after 10 cycles (Fig. 8b), indicating the exceptional stability of the V_O–N–TiO₂ (A/R) nanofilm in the PEC process.

Fig. 8 (a) Solar-driven photoelectrocatalytic activity of the V_O–N–TiO₂ (A/R) nanofilm and pure TiO₂ (control sample); (b) cycling tests of the V_O–N–TiO₂ (A/R) nanofilm.

4 Conclusions

The V_O–N–TiO₂ (A/R) nanofilm exhibits significant potential in high-efficiency utilization of solar energy. The high photoelectrocatalytic activity of the V_O–N–TiO₂ (A/R) nanofilm under sunlight was attributed to three main facts. (1) The distortion of the anatase/rutile interface with V_O and N atom doping caused a disordered lattice structure, narrowed the band gap of TiO₂, promoted the optical absorption from 200 nm to 2500 nm, and interfacial electron transfer. (2) The unique three-dimensional pine-bough-like structures increased the surface area of the nanofilm, providing more adsorptive and active sites, meanwhile promoting the light adsorption of scattering and diffuse reflection. (3) The photo-generated electron–hole pairs can be separated further by applying a bias voltage.

Conflicts of interest

The authors declare no competing financial interest.

Acknowledgements

This work was supported by Program for Changjiang Scholars and Innovative Research Team in University (IRT13083) from the Ministry of Education of China.

References

1. X. Chen, L. Liu, Y. Y. Peter and S. S. Mao, Science, 2011, 331, 746–750.
2. 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.
3. W. Guo, F. Zhang, C. Lin and Z. L. Wang, Adv. Mater., 2012, 24, 4761–4764.
4. L. Cui, Y. Wang, M. Niu, G. Chen and Y. Cheng, J. Solid State Chem., 2009, 182, 2785–2790.
5. J. Xue, Q. Shen, W. Liang, X. Liu and F. Yang, Electrochim. Acta, 2013, 97, 10–16.
6. H. Xie, W. Que, Z. He, P. Zhong, Y. Liao and G. Wang, J. Alloys Compd., 2013, 550, 314–319.
7. C. W. Lai and S. Sreekantan, Int. J. Hydrogen Energy, 2013, 38, 2156–2166.
8. J. Wang, D. N. Tafen, J. P. Lewis, Z. Hong, A. Manivannan, M. Zhi, M. Li and N. Wu, J. Am. Chem. Soc., 2009, 131, 12290–12297.
9. R. Nakamura, T. Tanaka and Y. Nakato, J. Phys. Chem. B, 2004, 108, 10617–10620.
10. R. Asahi, T. Morikawa, T. Ohwaki, K. Aoki and Y. Taga, science, 2001, 293, 269–271.
11. K. Shankar, K. C. Tep, G. K. Mor and C. A. Grimes, J. Phys. D: Appl. Phys., 2006, 39, 2361.
12. L. Liu and X. Chen, Chem. Rev., 2014, 114, 9890–9918.
13. D. C. Hurum, A. G. Agrios, K. A. Gray, T. Rajh and M. C. Thurnauer, J. Phys. Chem. B, 2003, 107, 4545–4549.
14. X. Zhang, Y. Lin, D. He, J. Zhang, Z. Fan and T. Xie, Chem. Phys. Lett., 2011, 504, 71–75.
15. V. N. Kuznetsov and N. Serpone, J. Phys. Chem. C, 2009, 113, 15110–15123.
16. J. Zhuang, S. Weng, W. Dai, P. Liu and Q. Liu, J. Phys. Chem. C, 2012, 116, 25354–25361.
17. A. Naldoni, M. Allieta, S. Santangelo, M. Marelli, F. Fabbri, S. Cappelli, C. L. Bianchi, R. Psaro and V. Dal Santo, J. Am. Chem. Soc., 2012, 134, 7600–7603.
18. X. Chen, L. Liu, P. Y. Yu and S. S. Mao, Science, 2011, 331, 746–750.
19. B. Chen, J. A. Beach, D. Maurya, R. B. Moore and S. Priya, RSC Adv., 2014, 4, 29443–29449.
20. N. Liu, C. Schneider, D. Freitag, M. Hartmann, U. Venkatesan, J. Müller, E. Spiecker and P. Schmuki, Nano Lett., 2014, 14, 3309–3313.
21. Q. Wang, X. Yang, X. Wang, M. Huang and J. Hou, Electrochim. Acta, 2012, 62, 158–162.
22. N. A. Deskins, S. Kerisit, K. M. Rosso and M. Dupuis, J. Phys. Chem. C, 2007, 111, 9290–9298.
23. T. Xia, N. Li, Y. Zhang, M. B. Kruger, J. Murowchick, A. Selloni and X. Chen, ACS Appl. Mater. Interfaces, 2013, 5, 9883–9890.
24. X. Chen and S. S. Mao, Chem. Rev., 2007, 107, 2891–2959.
25. T. Lan, X. Tang and B. Fultz, Phys. Rev. B: Condens. Matter Mater. Phys., 2012, 85, 094305.
26. D. Kim, S. Fujimoto, P. Schmuki and H. Tsuchiya, Electrochem. Commun., 2008, 10, 910–913.
27. S. Hoang, S. Guo, N. T. Hahn, A. J. Bard and C. B. Mullins, Nano Lett., 2011, 12, 26–32.
28. Y. C. Zhang, M. Yang, G. Zhang and D. D. Dionysiou, Appl. Catal., B, 2013, 142–143, 249–258.
29. B. Santara, P. K. Giri, K. Imakita and M. Fujii, J. Phys. Chem. C, 2013, 117, 23402–23411.
30. V. Etacheri, M. K. Seery, S. J. Hinder and S. C. Pillai, Adv. Funct. Mater., 2011, 21, 3744–3752.
31. G. Liu, H. G. Yang, X. Wang, L. Cheng, J. Pan, G. Q. Lu and H.-M. Cheng, J. Am. Chem. Soc., 2009, 131, 12868–12869.
32. P. Roy, S. Berger and P. Schmuki, Angew. Chem., Int. Ed., 2011, 50, 2904–2939.
33. Z. Wang, W. Cai, X. Hong, X. Zhao, F. Xu and C. Cai, Appl. Catal., B, 2005, 57, 223–231.
34. X. Chen and C. Burda, J. Am. Chem. Soc., 2008, 130, 5018–5019.
35. S.-K. Joung, T. Amemiya, M. Murabayashi and K. Itoh, Appl. Catal., A, 2006, 312, 20–26.
36. R. Li, H. Kobayashi, J. Guo and J. Fan, Chem. Commun., 2011, 47, 8584–8586.
37. J. Dong, J. Han, Y. Liu, A. Nakajima, S. Matsushita, S. Wei and W. Gao, ACS Appl. Mater. Interfaces, 2014, 6, 1385–1388.
38. M. Batzill, E. H. Morales and U. Diebold, Phys. Rev. Lett., 2006, 96, 026103.
39. C. Di Valentin, G. Pacchioni, A. Selloni, S. Livraghi and E. Giamello, J. Phys. Chem. B, 2005, 109, 11414–11419.
40. S. Chu, S. Inoue, K. Wada, S. Hishita and K. Kurashima, J. Electrochem. Soc., 2005, 152, B116–B124.
41. K. Hadjiivanov and H. Knözinger, Phys. Chem. Chem. Phys., 2000, 2, 2803–2806.
42. D. A. Hanaor and C. C. Sorrell, J. Mater. Sci., 2011, 46, 855–874.
43. J. Zhang, M. Li, Z. Feng, J. Chen and C. Li, J. Phys. Chem. B, 2006, 110, 927–935.
44. W. Junwei, M. Ashish Kumar, Z. Qing and H. Liping, J. Phys. D: Appl. Phys., 2013, 46, 255303.
45. V. Sivaram, E. J. Crossland, T. Leijtens, N. K. Noel, J. Alexander-Webber, P. Docampo and H. J. Snaith, J. Phys. Chem. C, 2014, 118, 1821–1827.
46. D. Fang, Z. Luo, K. Huang and D. C. Lagoudas, Appl. Surf. Sci., 2011, 257, 6451–6461.
47. C. Rath, P. Mohanty, A. Pandey and N. Mishra, J. Phys. D: Appl. Phys., 2009, 42, 205101.
48. R. D. Shannon and J. A. Pask, J. Am. Ceram. Soc., 1965, 48, 391–398.
49. S. Li, J. Chen, F. Zheng, Y. Li and F. Huang, Nanoscale, 2013, 5, 12150–12155.
50. S. Wendt, P. T. Sprunger, E. Lira, G. K. H. Madsen, Z. Li, J. Ø. Hansen, J. Matthiesen, A. Blekinge-Rasmussen, E. Lægsgaard, B. Hammer and F. Besenbacher, Science, 2008, 320, 1755–1759.
51. S. Kurian, H. Seo and H. Jeon, J. Phys. Chem. C, 2013, 117, 16811–16819.
52. Y. Wang, C. Feng, M. Zhang, J. Yang and Z. Zhang, Appl. Catal., B, 2010, 100, 84–90.
53. M. Harb, P. Sautet and P. Raybaud, J. Phys. Chem. C, 2011, 115, 19394–19404.
54. W. Zhang, M. Zhang, Z. Yin and Q. Chen, Appl. Phys. B: Lasers Opt., 2000, 70, 261–265.
55. J. L. Sumerel, W. Yang, D. Kisailus, J. C. Weaver, J. H. Choi and D. E. Morse, Chem. Mater., 2003, 15, 4804–4809.
56. F. Zuo, L. Wang, T. Wu, Z. Zhang, D. Borchardt and P. Feng, J. Am. Chem. Soc., 2010, 132, 11856–11857.
57. E. Finazzi, C. Di Valentin, G. Pacchioni and A. Selloni, J. Chem. Phys., 2008, 129, 154113.
58. Q. Xiang, K. Lv and J. Yu, Appl. Catal., B, 2010, 96, 557–564.
59. T. Kawahara, Y. Konishi, H. Tada, N. Tohge, J. Nishii and S. Ito, Angew. Chem., 2002, 114, 2935–2937.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra05367f

‡ These authors contributed equally to this work (co-first authors).

---