On the role of metal atom doping in hematite for improved photoelectrochemical properties: a comparison study

Abstract

Doping with metals is an effective strategy to improve the charge transport and photoelectrochemical (PEC) properties of hematite photoelectrode. Herein, we report a comparison study of various metal atoms doped hematites to look into the effect of metal doping on the physical and chemical properties as well as the PEC performance of planar hematite thin film under the same synthetic and measurement condition. The efficient dopants, Ti, Cr, W, Pb, Sn, Zr and Si were selected and further investigated for their morphological and electronic properties. Nanorod arrays shaped hematite electrodes were also used to investigate the doping effect for the selected typical dopant elements. It was found that Ti, Zr, and Sn doping could greatly improve the photocurrent response, and Ce and Pb doping mildly increased the photocurrent. In contrast, Mo doping had a negative effect on the PEC property of hematite photoelectrodes. Based on the results of Mott–Schottky plots, valence band XPS spectra, electrochemical impedance spectroscopy and photocurrent, it can be found that the improvement of PEC performance of doped hematite is largely due to positively shifted flat band potential, which significantly influence the surface states and interface charge transport property. While for most dopants, the charge carrier densities were not changed and show less influence on the overall performance.

---

Introduction

Hematite is currently an intensely investigated metal oxide for photoelectrochemical (PEC) water splitting application because of its abundance, suitable band gap, stability and breakthrough of efficiency in recent years.^1–4 With a band gap of ∼2.2 eV, hematite has a theoretical solar to hydrogen conversion efficiency of 16.8%.⁵ However, there are some challenges of employing this material for PEC water splitting application: (a) conduction band potential is not high enough to drive water reduction without bias, (b) a relatively low absorption coefficient because of indirect bandgap and (c) poor majority carrier conductivity and short hole diffusion length (2–4 nm).^6–8 To overcome these drawbacks, morphology control and ion doping were mostly employed.^4,6,9–11 Nanoporous and nanowire/nanosheet structure with cocatalyst can be ideal structure for hematite photoanodes as the reduced distance for the hole diffusion and accelerated surface reaction.^6,12 For example, a nanoparticle-based α-Fe₂O₃ electrodes with a photocurrent of 4.0 mA cm⁻² at 1.53 V vs. RHE was reported by Gratzel¹³ and co-works. Recently, hematite nanosheet film modified with Ag nanoparticles (NPs) and Co–Pi cocatalyst reported by Wang gives a record photocurrent density of 4.68 mA cm⁻² at 1.23 V vs. RHE.¹⁴

To improve the conductivity and carrier mobility, hematite was doped with Si, Ti, Zn, Sn, Al, Mg, Nb, Zr, Mo, Cr, Pt, etc. in literatures.^9,10,15–24 It has been reported that a clear increase in the donor concentration was determined from doping with Ti⁴⁺, which is regarded accounting for the improved PEC performance.²⁴ Doping with Sn significantly improves the charge mobility of hematite nanostructures and thus show much higher photocurrent density than undoped samples.^25,26 Other dopants such as Zn, Si and Cr also exhibit a positive effect on photoelectrochemical performance of hematite photoelectrodes.^9,17,18 However, the reported doped hematites were typically synthesized with different approaches and investigated under various conditions. As a consequence, it is unclear to the different extents of these doping impacts upon the performance of hematites under various operation conditions. Thus, it's highly desired to compare the effects of various metal dopants on structural, electrical and photoelectrochemical properties in the same material structure and experimental condition of hematite and to study the diverse roles that dopants play in hematite. It is increasingly recognized that alongside new material or structure developments, improved knowledge of the change of optoelectronic, catalytic and carrier dynamic characteristics in doped materials will be important for their technological development.

In this paper we synthesized various metal atoms doped planar shaped hematites under the same synthetic procedure and compared their photocurrents to look into the effect of metal doping on hematite thin films. The efficient dopants were selected and further investigated for their morphological and electronic properties. The widely investigated hematite nanorod array was also used to look into the doping effects on its physical and chemical properties as well as the PEC performance. The hematite nanorod arrays were doped by surface deposition and thermal diffusion with selected dopants. The morphology and crystal structure of hematite photoelectrodes were studied by SEM, XRD, XPS and electrochemical measurement to investigated the effect of various dopants on PEC performance of hematite electrodes. Their flat band potentials, charge carrier densities and charge transport properties were compared to study the role of dopants that played on hematite.

Experimental

Preparation of doped planar-shaped hematite photoelectrodes

The planar shaped hematites were deposited on FTO substrates by spincoating. The precursor for spincoating is prepared by mixing certain amount of Fe(NO₃)₃ aqueous solution (2 M), dopant solution (0.1 M), polyvinyl alcohol (PVA) solution (6 g/100 ml) and distilled water to obtain required concentration and molar ratio (1 mol% and 5 mol% doped with concentration of 1 M). Dopant solutions were obtained by dissolving 0.0005 mol dopant reagents in 5 ml 6.5 wt% HNO₃ to obtain a dopant ion concentration of 0.1 M. Those dopant reagents used are Li₂CO₃, Mg(NO₃)₂, Al(NO₃)₃·9H₂O, tetraethyl orthosilicate (TEOS), Ca(NO₃)₂·4H₂O, titanium butoxide, NH₃VO₃, Cr(NO₃)₂·9H₂O, Mn(NO₃)₂, Co(NO₃)₂·6H₂O, Ni(NO₃)₂·6H₂O, Cu(NO₃)₂·3H₂O, Zn(NO₃)₂·6H₂O, Sr(NO₃)₂, Y(NO₃)₃, Zr(NO₃)₄, (NH₄)₆Mo₇O₂₄·4H₂O, Cd(NO₃)₂, In(NO₃)₃, SnCl₄·5H₂O, Ba(NO₃)₂, PbO, Bi(NO₃)₃, La(NO₃)₃·nH₂O, Ce(NO₃)₂·6H₂O, NH₄WO₄, which were purchased from Sinopharm Chemical Reagent Co., Ltd. and used directly without further purification. The obtained precursor solution was then spincoated on the FTO glass at 4000 rpm for 15 s. This process was repeated for 2 circles using 0.05 ml precursor and the sample was heated on a heating plate for 10 min at 130 °C for each circle. The obtained thin film was then thermal annealed in air at 450 °C for 2 h in a muffle furnace.

Preparation of doped hematite nanorod photoelectrodes

The pristine hematite nanorod arrays were grown on FTO substrates using a facile hydrothermal method.^24,27 First, a Teflon-lined stainless steel autoclave was filled with 0.15 M FeCl₃·6H₂O and 1 M NaNO₃ solution at pH 1.5 (adjusted by HCl). Then a piece of FTO substrate was put into the autoclave with its conducting side facing to the wall. Subsequently the autoclave was sealed and heated in an oven at 100 °C for 24 h. After cooling down naturally, the obtained yellow β-FeOOH film was washed with deionized water and then sintered at 550 °C for 2 h and 800 °C for 90 s.

For preparation of samples doped with metal elements, the unannealed yellow β-FeOOH film was soaped in 0.2 M (NH₄)₆Mo₇O₂₄·4H₂O, 0.2 M Pb(NO₃)₂, 0.2 M SnCl₄·5H₂O, 0.2 M Ce(NO₃)₂·6H₂O or 0.2 M Zr(NO₃)₄·3H₂O aqueous solution with stirring for 12 h. For titanium doping, 0.2 M TiCl₄ aqueous solution was used. The dopant solutions were prepared by dissolving corresponding salts in distilled water and the pH was adjusted by dropping NaOH aqueous solution (1 M) or HNO₃ aqueous solution (1 M) until hydrolysis was just starting. Then the solution treated films were further sintered at 550 °C for 2 h and 800 °C for 90 s, the same as the annealing treatment of the pristine hematite films.

Characterization

The morphologies of the samples were examined by field-emission scanning electron microscopy (SEM, 7800F, JEOL, Japan). X-ray diffraction spectra were conducted on an X-ray diffractometer (XRD, X'pert PRO MPD, PANalytical, The Netherlands, λ ∼ 0.154056 nm). X-ray photoelectron spectrometer (XPS, AXIS Ultra DLD, Shimadzu/Kratos Analytical) was used to detect the chemical states of elements.

PEC measurements

The PEC properties of samples were investigated in a conventional three-electrode cell with a 1.5 cm² Pt plate as counter electrode, Ag/AgCl electrode as reference electrode, and prepared films as the working electrode. The area contacted with electrolyte of hematite films was about 0.785 cm². The electrolyte used for hematite nanorod arrays was 0.5 M Na₂SO₄ aqueous solution. An electrochemical analyzer (CHI760D, Shanghai Chenhua) was used for all PEC measurements. The scan rate of the applied potential was 0.02 V s⁻¹. The simulated light was obtained using a xenon light source equipped with an AM 1.5 G filter, and the light power density was set to be 100 mW cm⁻². Electrochemical impedance spectra were recorded at 0.8 V vs. Ag/AgCl with a frequency range of 1 Hz to 100 kHz. Mott–Schottky analysis was carried out at potential range from −0.3 to 1.4 V vs. Ag/AgCl in the dark with a frequency of 977 Hz.

Results and discussion

Doped planar hematites were synthesized by spincoating/annealing method and investigated to rapid screen and identify the efficient dopants for hematite. Current density versus applied voltage measurements (C–V) were performed for different metal ion doped planar hematites with doping level of 1% and 5% in 0.5 M NaOH. The photocurrent and dark current values at potential of 0 V (vs. Ag/AgCl) were derived from the current–voltage (C–V) curves and plotted in Fig. 1.

Fig. 1 Photocurrent comparison of bare hematite photoelectrode and 1% and 5% metal element doped hematite photoelectrodes. The inset shows how the photocurrents were derived from the current–voltage (C–V) curves.

It can be found that the pristine hematite photoelectrode showed a relatively low photocurrent, which is about 8.6 μA cm⁻² at 0 V vs. Ag/AgCl. When comparing to the doped hematites, it is revealed that most elements doping show positive effect on photocurrents and among these dopants tested Ti gives the highest enhancement in photocurrent of hematite (about 47 μA cm⁻²), followed by Cr, W, Pb, Sn, Zr, etc. In particular, 5% doping level gives more enhancement than 1% doping level for the top 5 samples with highest photocurrents. While for other samples, those with 1% doping level give higher or similar photocurrent than those with 5% doping level. This result indicated that more effective dopants possess a higher optimal doping level. The decrease of photocurrent with higher doping level could result from exceeding dopants acting as recombination center. For those effective dopants, however high doping concentration does not act as recombination but make positive contribution to the photocurrent.

To look into the detailed influence from the different elements doping, those doped samples with top 7 performances were selected for further investigation. Fig. 2 presents the SEM images of flat-shaped hematite films. Apparently, the undoped hematite film consists of leaf shape particles with 30–50 nm in width and 120–150 nm in length. For hematite films doped with Cr, Ti, W, Sn, Zr or Pb, no obvious morphology change could be observed. The thickness of flat-shaped hematite film is determined to be around 100 nm.

Fig. 2 SEM images of (a) undoped (b) 5% Cr doped (c) 5% Ti doped (d) 5% W doped (e) 5% Sn doped (f) 5% Zr (g) 5% Pb doped flat-shaped hematite photoelectrodes.

The chemical state of doped hematites identified by XPS spectra can be used to understand the chemical valence and bonding situation of elements. The original and fitted XPS spectra of the O 1s collected for pristine and doped hematite films are shown in Fig. 3. In O 1s XPS spectra, the pristine film shows a main peak at 529.7 eV, which is a characteristic peak of lattice oxygen in Fe₂O₃. The small peak at 531.2 eV is due to absorbed hydroxyl group on the surface of hematite. After doping treatment, the main peaks of all doped films shift to higher binding energy, similar as the previous reports.²⁸ The foreign dopant atoms would exert influence on the Fe–O lattice and increased the binding energy. Fig. S1† shows the XPS spectra of each dopant and their valence states and molar ration to Fe were calculated and shown in Table S1.† Since the valence of Cr and Pb (+3 and +2, respectively) were not higher than Fe, the shift of O 1s peak to higher binding energy is relatively smaller. The peak around 532.8 eV for Ti, Pb and W doped samples can be ascribed to oxygen of SiO₂, this could be from the exposed glass substrate. In Fig. S2,† two typical Fe 2p_1/2 and Fe 2p_3/2 peaks appear in all the samples and the absence of Fe²⁺ satellite peak at around 716 eV indicates that the main state of Fe is Fe³⁺.²⁹

Fig. 3 O 1s XPS spectra of bare and doped planar-shaped hematite.

The spin coating deposition is a facile method to prepare metal ion doped hematite with precise doping level and provide a rapid screen of doping elements for efficient dopants. However, the synthesized planar hematite thin films showed very low activities. The most reported hematite photoanodes with high photocurrents are Fe₂O₃ nanowire arrays. To check the role of different metal element doping in this hematite structure, we further selected typical dopants to compare their effects on Fe₂O₃ nanowire arrays. The elements of Ti, Sn, Zr, Pb, Mo and Ce with improved or decreased photocurrent in planar hematite were selected.

The nanorod arrays were prepared using a hydrothermal method in previous report.²⁷ The doping procedure was conducted by soaking in dopant solution followed with heat treatment. In this work, we adopted a short-time annealing process (800 °C for 90 s) to induce surface dopants diffusion into hematite and prevent Sn of the FTO substrates from diffusing into hematite nanorods, which has been demonstrated in our previous report.²⁴ Fig. 4 shows SEM images of annealed hematite nanorods doped with different dopants. It was found that the undoped nanorods were 80 nm in diameter and about 500 nm in length perpendicular to the FTO glass substrate, similar with previous reports.²⁶ For the doped samples, there is no significant change in morphology of hematite nanorods after doping with Mo, Pb, Ce, Zr and Ti elements. While the Sn doping treatment make the surface of nanorods capped with small particles which can be due to the formation of SnO₂ particles. As the Sn atom is relatively large compared to other dopants, it is more difficult to incorporate at interstitial sites or to diffuse into the hematite host.

Fig. 4 Low- and high-magnification SEM images of pristine and various metal element doped hematite nanorod arrays and cross-section SEM image of pristine hematite.

X-ray diffraction (XRD) measurements were conducted to examine the phase purity and crystallinity of each sample. As shown in Fig. 5, the peaks of all films can be indexed to α-Fe₂O₃ with a rhombohedral structure (JCPDS card no. 24-0072). In the patterns of all films apart from FTO, the dominant hematite peak is (110) as a result of the (110) preferential orientation, which also consists with previous reports.²⁴ There are no peak shifts or other phases exist in all the spectra. This result suggests that the doping has little effect on the crystal structure of hematite films and no detectable structure phase formed, including the amorphous SnO₂ particles which are visible in SEM images.

Fig. 5 XRD patterns collected for bare Fe₂O₃ and element doped Fe₂O₃ nanorod photoelectrodes.

Fig. 6 presents the PEC performance of pristine Fe₂O₃ and Fe₂O₃ nanorods doped with different elements. It can be observed that the hematite nanorod films show much more positive onset potentials than planar hematite films. The pristine Fe₂O₃ nanorods exhibit a photocurrent density of 37 μA cm⁻² at 1 V vs. Ag/AgCl. After being doped with foreign elements, significant improvements of photocurrents were observed for most doped nanorods. In consistency with result of planar samples, the Ti doped hematite nanorod exhibits the highest photocurrent. Similarly, Zr, Sn, and Pb doped hematite nanorods also show greatly improved photocurrent response and the Mo doped sample produce a decreased photocurrent. While the Ce doped hematite showed an increased photocurrent, which is different from that of planar one.

Fig. 6 Photocurrent responses of bare and doped hematite nanorods recorded under chopped light irradiation. The electrolyte is 0.5 M Na₂SO₄ aqueous solution.

To investigate the effect of dopants on the electrochemical properties, the Mott–Schottky plots of pristine and doped hematite nanorod were tested in 0.5 M Na₂SO₄ aqueous solution. The flat band potential and carrier densities of films can be estimated from the Mott–Schottky eqn (1):^26,30,31

where C_sc is the space charge capacitance, e is the electron charge, ε is the dielectric constant of hematite (ε = 80),^26,30,32 ε₀ is the permittivity of free space, N_D is the carrier density, E is the applied bias, E_fb is the flat band potential, k is the Boltzmann constant, and T is the temperature in K. The results were shown in Fig. S3.† According to eqn (1), the carrier density in photoelectrodes is inversely proportional to the slopes of linear parts of Mott–Schottky plots and the flat-band potentials of each sample can be obtained from the intercept on the x-axis in the linear region of the curves. Fig. 7 shows the comparison of photocurrents, flat-band potentials and carrier densities of different dopants for hematite photoelectrodes. It was shown that the flat-band potential shifted positively for samples doped with Ti, Zr and Sn, those also exhibit significantly enhanced photocurrents, while for samples doped with Pb, Ce and Mo, negatively shifted or unchanged flat-band potentials were observed and relatively lower photocurrents were obtained. It can be expected that sample with a larger positively shift of flat-band potential gives a better catalytic activity regarding the oxygen evolution reaction.³³ Furthermore, it is also found that the carrier densities calculated from slopes of Mott–Schottky plots of the doped samples are within the same order of magnitude (10¹⁷ cm⁻³) and are not necessary higher than that of pristine sample. So the dopants may not acting as an electrical dopant to increase the carrier density, which is consistent with some previous reports.³⁴ However the dopants in hematite may relate to other factors such as flat-band potential or surface catalytic effects for water oxidation.³⁵

Fig. 7 Comparison of photocurrents, flat-band potentials and carrier densities of planar shaped hematite.

In particular, the carrier density of Ce doped sample was increased but its flat band potential shifted negatively, resulting a reduced photocurrent. While for Sn doped hematite, the positively shifted flat-band potential induced an enhanced photocurrent even though its carrier density was decreased. It can be seen that the PEC performance of hematite photoelectrode is greatly related to flat band potential and carrier density which involves surface states and charge transport property, respectively. In general, a positively shifted flat-band potential is required and a decreased carrier density need to be avoided for an efficient PEC performance.

Valence band XPS spectra were recorded to determine the valence-band edge.³⁶ The valence band XPS spectra for different element doped hematite nanorods were plotted and shown in Fig. S4† and the valence-band edge values were listed in Table S3.† It was found that the ranking of valence-band position with different element doped samples is similar to the results of flat band potential. It is easy to understand that the flat-band potential is close to the conduction band (CB) position for n-type semiconductor, and the difference between the valence band (VB) position and CB is band gap energy.³⁷ Same trend for photocurrent and valence band edge of samples with different elements doping was observed. More specifically, samples with a more positive binding energy of valence band edge show a higher photocurrent. This result again verified that the shift of band edge position upon doping is an important factor that determined the photocurrent.

The improvement of photocurrent by the positively band edge position shift could be a result of improved surface reaction or surface charge transfer. To get more information of this process, electrochemical impedance spectroscopy (EIS) of both undoped and doped films were recorded. As shown in Fig. 8, the representative Nyquist plots for all hematite films can be fitted to a 2-RC-circuit model which consists of the resistance of FTO, R_s, the charge transport resistance in the bulk, R_CT1, and the space-charge capacitance, CPE₁, of the bulk hematite, as well as the Helmholtz capacitance at the hematite/electrolyte interface, CPE₂, and the charge transfer resistance at the hematite/electrolyte interface, R_CT2.¹⁸ The calculated equivalent circuit parameters by fitting according to the equivalent circuits are shown in Table 1. As can be observed, all doped hematite samples show reduced charge transport resistance in the bulk in comparison with the bare hematite. According to the Mott–Schottky results, the carrier density didn't increased much after doping, but the electrical conductivities of hematites upon doping improved, this could be explained by increased polaron hopping probability (the conducting mechanism of hematite) with lattice strain (shorter intraplane Fe–Fe distance) originated from the incorporated dopants.³⁸ Further, the lowest bulk charge transport resistance was observed with Ti–Fe₂O₃ photoelectrode, followed by Zr–Fe₂O₃ and Sn–Fe₂O₃, in agreement with the photocurrent response. It is interesting to note that Mo–Fe₂O₃ photoelectrode shows lower bulk charge transport resistance but much higher resistance at the hematite/electrolyte interface, which can be responsible for the low photocurrent density.

Fig. 8 Electrochemical impedance spectra (EIS) for bare and doped hematite nanorod photoelectrodes measured at 1.0 V vs. Ag/AgCl under AM 1.5 G irradiation. The inset is the equivalent circuit model used for fitting the Nyquist plots.

Table 1 The fitting results of resistance and capacitance values for Nyquist plots

Sample	RS (Ω cm^−2)	CPE1 (F cm^−2)	RCT1 (Ω cm^−2)	CPE2 (F cm^−2)	RCT2 (Ω cm^−2)
Bare Fe2O3	78.35	2.73 × 10^−5	1768.0	4.20 × 10^−6	879.7
Mo–Fe2O3	64.83	1.74 × 10^−7	231.1	1.09 × 10^−5	16 340.0
Pb–Fe2O3	71.99	1.08 × 10^−7	361.0	7.57 × 10^−5	664.3
Ce–Fe2O3	68.94	1.13 × 10^−5	505.6	1.26 × 10^−4	729.0
Sn–Fe2O3	72.45	1.49 × 10^−5	146.7	2.63 × 10^−4	266.9
Zr–Fe2O3	66.14	1.30 × 10^−5	122.9	2.87 × 10^−4	259.8
Ti–Fe2O3	73.68	3.46 × 10^−5	89.53	2.66 × 10^−4	363.1

---

Conclusions

In summary, we have reported synthesis of planar and nanorod shaped hematite electrodes doped with various metal atoms and investigation of the effect of metal doping on the physical, chemical and PEC properties. The performances of doped planar hematite thin films were compared under the same synthetic and measurement condition. The Ti, Cr, W, Pb, Sn and Zr were found to be the most efficient dopants to improve the PEC performance. Samples doped with these elements were further investigated for their morphological and electronic properties. For widely investigated nanorod arrays shaped hematite were also doped with selected dopant elements. The result showed that Ti, Zr, and Sn doping greatly improve the photocurrent, followed with Ce and Pb doping which also increased the photocurrent. The Mo doping however had a negative effect on the PEC property of hematite photoelectrodes. Mott–Schottky plots indicated that the charge carrier densities were not changed for most dopants. Valence band XPS spectra samples with a more positive binding energy of valence band edge show a higher photocurrent. Electrochemical impedance spectroscopy showed that the doped hematite samples show reduced charge transport resistance. These results demonstrated that the improvement of PEC performance of doped hematite is largely due to their positively shifted flat band potential, which influenced the surface charge transport property. This result provides us new explanation on the effect of doping on hematite photoelectrodes.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (No. 51202186, 51236007, 21606175).

References

1. K. Sivula, F. Le Formal and M. Graetzel, ChemSusChem, 2011, 4, 432–449.
2. Y. Yang, M. Forster, Y. Ling, G. Wang, T. Zhai, Y. Tong, A. J. Cowan and Y. Li, Angew. Chem., Int. Ed., 2016, 55, 3403–3407.
3. P. Zhang, T. Wang, X. Chang, L. Zhang and J. Gong, Angew. Chem., 2016, 55, 5851–5855.
4. P. Peerakiatkhajohn, J. H. Yun, H. Chen, M. Lyu, T. Butburee and L. Wang, Adv. Mater., 2016, 28, 6405–6410.
5. A. Murphy, P. Barnes, L. Randeniya, I. Plumb, I. Grey, M. Horne and J. Glasscock, Int. J. Hydrogen Energy, 2006, 31, 1999–2017.
6. J. Wang, J. Su and L. Guo, Chem.–Asian J., 2016, 11, 2328–2334.
7. J. Wang, N. Zhang, J. Su and L. Guo, RSC Adv., 2016, 6, 41060–41066.
8. F. Malara, F. Fabbri, M. Marelli and A. Naldoni, ACS Catal., 2016, 6, 3619–3628.
9. I. Cesar, K. Sivula, A. Kay, R. Zboril and M. Graetzel, J. Phys. Chem. C, 2009, 113, 772–782.
10. C. Toussaint, H. L. L. Tran, P. Colson, J. Dewalque, B. Vertruyen, B. Gilbert, N. D. Nguyen, R. Cloots and C. Henrist, J. Phys. Chem. C, 2015, 119, 1642–1650.
11. L. Wang, X. Zhou, N. T. Nguyen and P. Schmuki, ChemSusChem, 2015, 8, 618–622.
12. J. Y. Kim, G. Magesh, D. H. Youn, J. W. Jang, J. Kubota, K. Domen and J. S. Lee, Sci. Rep., 2013, 3, 2681.
13. S. C. Warren, K. Voitchovsky, H. Dotan, C. M. Leroy, M. Cornuz, F. Stellacci, C. Hebert, A. Rothschild and M. Gratzel, Nat. Mater., 2013, 12, 842–849.
14. P. Peerakiatkhajohn, J. H. Yun, H. J. Chen, M. Q. Lyu, T. Butburee and L. Z. Wang, Adv. Mater., 2016, 28, 6405–6410.
15. N. T. Hahn and C. B. Mullins, Chem. Mater., 2010, 22, 6474–6482.
16. J. Deng, J. Zhong, A. Pu, D. Zhang, M. Li, X. Sun and S.-T. Lee, J. Appl. Phys., 2012, 112, 084312.
17. S. Shen, J. Jiang, P. Guo, C. X. Kronawitter, S. S. Mao and L. Guo, Nano Energy, 2012, 1, 732–741.
18. N. Mirbagheri, D. Wang, C. Peng, J. Wang, Q. Huang, C. Fan and E. E. Ferapontova, ACS Catal., 2014, 4, 2006–2015.
19. H. Pan, X. Meng, D. Liu, S. Li and G. Qin, Phys. Chem. Chem. Phys., 2015, 17, 22179–22186.
20. D. Wang, H. Chen, G. Chang, X. Lin, Y. Zhang, A. Aldalbahi, C. Peng, J. Wang and C. Fan, ACS Appl. Mater. Interfaces, 2015, 7, 14072–14078.
21. Y. C. Zhang, S. Q. Jiang, W. J. Song, P. Zhou, H. W. Ji, W. H. Ma, W. C. Hao, C. C. Chen and J. C. Zhao, Energy Environ. Sci., 2015, 8, 1231–1236.
22. W. D. Chemelewski, O. Mabayoje, D. Tang, A. J. Rettie and C. B. Mullins, Phys. Chem. Chem. Phys., 2016, 18, 1644–1648.
23. J. Huang, G. Hu, Y. Ding, M. Pang and B. Ma, J. Catal., 2016, 340, 261–269.
24. J. Wang, B. Feng, J. Su and L. Guo, ACS Appl. Mater. Interfaces, 2016, 8, 23143–23150.
25. D. D. Qin, Y. L. Li, T. Wang, Y. Li, X. Q. Lu, J. Gu, Y. X. Zhao, Y. M. Song and C. L. Tao, J. Mater. Chem. A, 2015, 3, 6751–6755.
26. Y. Ling, G. Wang, D. A. Wheeler, J. Z. Zhang and Y. Li, Nano Lett., 2011, 11, 2119–2125.
27. L. Vayssieres, N. Beermann, S. E. Lindquist and A. Hagfeldt, Chem. Mater., 2001, 13, 233–235.
28. Q. Li, J. Bian, N. Zhang and D. H. L. Ng, Electrochim. Acta, 2015, 155, 383–390.
29. Y. Ling, G. Wang, J. Reddy, C. Wang, J. Z. Zhang and Y. Li, Angew. Chem., Int. Ed., 2012, 51, 4074–4079.
30. A. G. Tamirat, W. N. Su, A. A. Dubale, H. M. Chen and B. J. Hwang, J. Mater. Chem. A, 2015, 3, 5949–5961.
31. Y. B. Chen, Z. X. Qin, T. Chen, J. Z. Su, X. Y. Feng and M. C. Liu, RSC Adv., 2016, 6, 58409–58416.
32. G. Wang, Y. Ling, D. A. Wheeler, K. E. George, K. Horsley, C. Heske, J. Z. Zhang and Y. Li, Nano Lett., 2011, 11, 3503–3509.
33. F. Boudoire, R. Toth, J. Heier, A. Braun and E. C. Constable, Energy Environ. Sci., 2014, 7, 2680–2688.
34. Q. Li, J. C. Bian, N. Zhang and D. H. L. Ng, Electrochim. Acta, 2015, 155, 383–390.
35. O. Zandi, B. M. Klahr and T. W. Hamann, Energy Environ. Sci., 2013, 6, 634–642.
36. E. A. Kraut, R. W. Grant, J. R. Waldrop and S. P. Kowalczyk, Phys. Rev. Lett., 1980, 44, 1620–1623.
37. Q. L. Xu, J. G. Yu, J. Zhang, J. F. Zhang and G. Liu, Chem. Commun., 2015, 51, 7950–7953.
38. M. N. Huda, A. Walsh, Y. F. Yan, S. H. Wei and M. M. Al-Jassim, J. Appl. Phys., 2010, 107, 123712.

---

Footnote

† Electronic supplementary information (ESI) available: XPS spectra, Mott–Schottky plots, valence band spectra and tables for summary of electronic parameters. See DOI: 10.1039/c6ra22895j

---