Studies into the Yb-doping effects on photoelectrochemical properties of WO₃ photocatalysts

Abstract

Yb-doped WO₃ photocatalysts were prepared by co-sputtering Yb and WO₃ on FTO glass and then sintering at elevated temperatures in air. At sintering temperatures between 450 and 550 °C, the photocurrent densities of Yb-doped WO₃ photocatalysts were higher and more stable than those of pristine WO₃ photocatalysts. The Yb-doping effects on the photoelectrochemical properties were investigated using Raman, transient absorption and electrochemical impedance spectroscopies. Compared to pristine WO₃ photocatalysts, Yb-doped photocatalysts sintered at up to 550 °C consist of sub-stoichiometric WO_3−x due to the substitution of W⁶⁺ cations with Yb³⁺ dopants, displaying shorter electron–hole recombination times and higher levels of donor densities.

---

Introduction

Tungsten trioxide (WO₃) finds many applications as photocatalysts and gas sensors owing to its intrinsic properties of visible light-responsive band gap, free carrier diffusion length and good chemical stability in aqueous media. Over the years, much effort has been made to incorporate another metal into the semiconducting metal oxide to further enhance its photoelectrochemical properties. For example, the catalytic modification of a WO₃ surface via the electrodeposition of Pt particles could promote the reaction rate of methanol oxidation at the photoanode, leading to a significantly enhanced H₂ production efficiency, compared with when using unmodified WO₃.¹ It has also been reported that Pt-modified WO₃ films can be used as H₂ sensors which exhibit lower electrical resistances at higher Pt concentrations, favouring H₂ detection possibly due to the catalytic effects of Pt clusters on the dissociative adsorption of H₂ to form H-atoms. This catalytic dissociative effect releases electrons into the WO₃ matrix which in turn decreases the electrical resistance.² Pt nanoparticles can also be loaded onto nanoporous WO₃ nanotubes, which greatly increases photocatalytic activity because the large surface area of WO₃ nanotubes offers many more active sites in close proximity to the Pt nanoparticles to facilitate the transfer of photoelectrons to the Pt nanoparticles.³

To enhance the photoelectrochemical properties of WO₃, the incorporation of other transition metals such as Mo, Ti and Fe has also been investigated.^4–6 In Mo_xW_1−xO₃ hydrates, Mo⁶⁺ ions were preferred to W⁶⁺ ions as trapping sites for electrons.⁴ Therefore the substitution of W⁶⁺ with Mo⁶⁺ cations led to a red shift in the absorption edge due to the reduction of Mo⁶⁺ to Mo⁵⁺. Hence the photoelectrochemical response of WO₃ photoanodes to visible light was enhanced together with the red shift. Doping of Ti⁴⁺ into a WO₃ semiconductor also improved the photocatalytic performance under visible light due to the change in the band structure of WO₃ caused by impurity states between the conduction band and valence band resulting in a narrower band gap.⁵ The modification of a WO₃ surface with Fe³⁺ extended the absorption region and reduced the charge transfer resistance, leading to an increased photocurrent density.⁶ All of these dopings of transition metals had the common effect of narrowing the band gap of WO₃ and extending the absorption to the visible region of the solar spectrum.

A different mechanism that proceeds through harnessing the electron-rich f-orbitals of Yb dopants in WO₃ was reported for the first time in our earlier study.⁷ We prepared and investigated Yb-doped WO₃ photocatalysts of varying Yb concentrations for photoelectrochemical water splitting. For a Yb doping concentration as low as 0.34 at%, determined by X-ray photoelectron spectroscopy, an evident enhancement in the photocurrent density was observed, which was attributed to an increase in conductive carrier paths caused by oxygen vacancies and the 4f¹³ orbital configuration of the Yb³⁺ doped into the WO₃. However, a higher doping concentration of 1.6 at% increased the charge transfer resistance and reduced the photocurrent density compared to pristine WO₃, possibly due to a distortion of the WO₃ crystal structure induced by the larger ionic radius of Yb³⁺ relative to W⁶⁺. Notwithstanding the limiting factor of the size of Yb³⁺ dopant cations, the study showed that tapping into the electron-rich f-orbitals of the lanthanide proved to be an effective means to increase charge carrier generation for enhanced photocurrent densities in WO₃.

In order to obtain a deeper and better understanding of Yb-doped WO₃ photocatalysts for practical applications, herein we have studied the effects of Yb-doping, at varying sintering temperatures, on the photoelectrochemical properties of the material. The results showed that Yb doping enlarged the thermal envelope of WO₃. Specifically, photocurrent densities of Yb-doped WO₃ were sustained up to 550 °C. However for pristine WO₃ photocatalysts, the photocurrent density continually decreased with an increasing sintering temperature over the same range. These improvements were evaluated in the light of structural properties, charge dynamics and donor densities of thermally treated photocatalysts investigated with Raman, transient absorption and electrochemical impedance spectroscopies. The Yb-doped WO₃ photocatalysts sintered at up to 550 °C were sub-stoichiometric due to the substitution of W⁶⁺ with Yb³⁺ dopants and showed a shorter electron–hole recombination time but higher donor densities compared to pristine WO₃ photocatalysts. Yb-doping led to more stable and higher photocurrent densities with lower charge transfer resistances for PEC water splitting with the Yb-doped WO₃ photocatalysts.

Experimental details

Deposition

WO₃ and Yb were co-sputtered onto ultrasonically cleaned FTO glass substrates inside a magnetron sputtering chamber when the base pressure reached ∼10⁻⁷ Torr. The sputtering power was set at 100 W and 5 W for WO₃ and Yb, respectively, in order to yield as-deposited film thicknesses of ∼450 nm. The sputtered samples were sintered in air in a box furnace in a temperature range between 400 and 600 °C for 1 hour. A set of pristine WO₃ samples (without Yb) was also prepared using the same process parameters.

Characterisations

Structural analyses were carried out with a WITec alpha 300R microscope system using a 532 nm Nd YAG laser of ∼1 mW power to obtain Raman spectra of the samples. The spectroscopy for each sample was averaged at 1 μm per pixel and 0.1 s per point from a 100 × 100 μm scan area at a 100× objective. Transient absorption spectroscopy was carried out using a Nd:YAG laser (1 mJ, 7 ns, 10 Hz, 355 nm excitation) as the pump source and a 450 W xenon arc lamp as the probe source. Each spectral curve is the average of 256 measurements while each kinetics decay curve is the average of four measurements at 16 laser shots per measurement. Secondary ion mass spectroscopy (SIMS) using ION SIMS IV to profile the distribution of Yb in WO₃ and optical absorbance measured with a Shimadzu UV-3101 PC scanning spectrophotometer were carried out to supplement the spectroscopic results.

Photoelectrochemical (PEC) measurements and detection of gases (H₂ and O₂)

PEC measurements were carried out in a custom-made two-compartment glass cell with a three-electrode configuration. A Pt wire as the counter electrode was placed in one compartment, separated from a Ag/AgCl reference electrode and the working electrode (sputter-deposited WO₃ on FTO substrate) in another compartment. 400 mL of Na₂SO₄ (0.05 M, pH ∼ 5.4) electrolyte solution, which had already been purged with Ar, was added to the glass cell. Current–voltage measurements under dark and illuminated conditions at a scan rate of 100 mV s⁻¹ between −0.1 and 1.6 V Ag/AgCl were controlled using a potentiostat (Autolab PGSTAT101) connected to Nova 1.8 software. For the illuminated measurements, a xenon light source (Hamamastu LC8) with a radiant spectrum that resembles the visible portion of the solar spectrum was directed at the working electrode through a quartz window in the PEC glass cell. The intensity of the light had already been adjusted to 100 mW cm⁻² based on the output reading of a calibrated Si solar cell. The illuminated area on the working electrode was measured to be ∼1.5 cm by 1.0 cm. Electrochemical impedance spectroscopy was also performed using the same test cell set-up to obtain Nyquist and Mott–Schottky plots to determine the charge transfer resistance and donor density respectively. For the Nyquist plots, the cell was illuminated and maintained at constant bias voltages superimposed with a sinusoidal voltage of 10 mV in a frequency range between 0.1 Hz and 10 kHz. For the Mott–Schottky plots, measurements were made in the dark at selected frequencies between bias voltages of −0.2 and 1.3 V Ag/AgCl.

A custom-built photocatalytic reactor comprising a photoelectrochemical cell and a gas chromatograph was used to simultaneously measure photocurrents and quantify gaseous products from water splitting. The cell was an 820 cm³ glass vessel configured with a two-electrode set-up consisting of a Pt counter electrode and the sputter-deposited WO₃ working electrode. During the photoelectrochemical measurements, the WO₃ photocatalyst inside the cell was illuminated at a light intensity of 50 mW cm⁻² using a 150 W xenon arc lamp fitted with an AM1.5 filter. A constant 1.5 V bias voltage was applied through a potentiostat to the counter electrode during the measurements. ∼220 mL of the same electrolyte as was used in the previous section was used in this test. The gases from the photoelectrochemical cell were delivered by a high purity Ar carrier gas to a gas chromatograph (GC-2014 Shimadzu) for analysis.

Results and discussion

Structural analyses

At a sintering temperature of 400 °C, both the Yb-doped and pristine WO₃ samples displayed broad Raman spectra with a weak band region around a wavenumber of 200 cm⁻¹ and a peak at 270 cm⁻¹, as shown in Fig. 1. This band region evolved into peaks at 130, 190 and 117 cm⁻¹ and new peaks appeared at 706 and 809 cm⁻¹ when the sintering temperature was increased to 450 °C and beyond. All of the peaks which increased in intensity and sharpness with increasing sintering temperature, clearly identified both the Yb-doped and pristine photocatalysts as monoclinic WO₃ with peaks at the 200 cm⁻¹ band region ascribed to lattice vibration, 270 cm⁻¹ to bending and 706–809 cm⁻¹ to stretching vibrations of W–O bonds in monoclinic WO₃.^8,9 At sintering temperatures of 500 °C and above, Raman peaks of the Yb-doped WO₃ appeared weaker than those of its pristine counterpart with clear differences observed at 550 and 600 °C in all of the peak regions. In addition, Yb-doping also caused a minor shift of ∼2 cm⁻¹ to lower bond energies as shown in the inset graph which indicates a less ordered crystal lattice in the Yb-doped WO₃ compared to the pristine WO₃.

Fig. 1 Raman spectra of Yb-doped (dashed lines) and pristine (solid lines) WO₃ sintered from 400 to 600 °C. Peak indexing is for the first appearance of the peak. The inset spectrum at 550 °C shows a blue shift in the Raman peak at 809 cm⁻¹ of doped WO₃ by ∼2 cm⁻¹.

Distribution of Yb

Elemental depth profiling of Yb-doped WO₃ samples with SIMS (see Fig. S1a†) showed that Yb distribution followed those of W and O, indicating that Yb was uniformly incorporated into the WO₃ lattice. The average composition of the Yb-doped samples was WYb_0.025O_3−x from elemental analyses using an Oxford INCA X-ray energy dispersive spectrometer (EDS). The Raman and SIMS results in combination showed that it is very likely that the Yb³⁺ dopants substituted W⁶⁺ in the crystal lattice to form sub-stoichiometric WO_3−x which would be less crystalline than pristine WO₃.¹⁰ This is in agreement with the colour change from pale green in the pristine WO₃ to deep blue in the Yb-doped WO₃ as displayed by the samples shown in Fig. S1b.†

Carrier lifetime

The effect of Yb-doping on carrier transport was probed using transient absorption (TA) spectroscopic measurements. Fig. 2a shows the TA spectra obtained in the spectral range between 300 and 650 nm for the WO₃ photocatalysts and FTO substrate, recorded after a 100 ns delay upon excitation by a 355 nm laser. Strong TA signals with maxima (in the ∼352 to 357 nm region) were registered for both Yb-doped and pristine WO₃ with Yb-doped WO₃ exhibiting a lower signal level and a blue shift relative to pristine WO₃. TA spectra with maxima in the UV-visible region have been ascribed to absorption by photoholes in metal oxides such as Fe₂O₃, WO₃, TiO₂ and BiVO₄.^11–14 A lower TA signal for Yb-doped WO₃ would imply that fewer photoholes were probed in the spectral range possibly due to faster recombination of electrons and holes in the less ordered crystal lattice. The blue shift observation corroborates with the optical absorption spectra in Fig. S2† which show that Yb doping did not produce the desired red shifts in the absorption edges of WO₃ for all of the sintering temperatures, unlike for the reported doping effects of Mo, Ti and Fe transition metals into a WO₃ semiconductor where the band gap was narrowed and the absorption region was extended to the visible range of the solar spectrum (i.e. red shift).^4–6 Both the TA and optical absorption spectra in this present study suggested that the presence of Yb oxides with wide band gaps in the Yb-doped WO₃ probably contributed to the non-red shifts in the absorption edges. Fig. 2b shows the decay kinetics of TA probed at a 400 nm wavelength, which revealed a rapid decrease in the TA signals for the first 10 ms followed by a gradual decrease for up to 0.1 s for both the Yb-doped and pristine photocatalysts, which resembles the occurrence of bulk recombination of electrons and holes in semiconductor photocatalysts in similar excitation conditions.^11,13,15 However the reported recombination times t_1/2 (time to reach half of the initial concentration of charge carriers) of these photocatalysts were usually in the order of microseconds. In the present study with no bias voltage applied, the t_1/2 values for both the Yb-doped and pristine WO₃ samples were higher than the literature values^11,13,15 with the Yb-doped WO₃ having a shorter t_1/2 of ∼0.46 ms compared to 0.87 ms for the pristine WO₃. t_1/2 of the Yb-doped WO₃ shows that the mobility of free carriers in the doped photocatalyst was not severely impaired by the doping with Yb even though Raman spectroscopic measurements indicated that the crystal lattice was less ordered in the Yb-doped WO₃ compared to the pristine WO₃. The long recombination time in WO₃ is linked to its free carrier diffusion length which can be up to two orders of magnitude higher than that of either TiO₂ or Fe₂O₃. A long carrier lifetime is desired in photocatalytic applications for increased free electron–hole pairs to enhance photocurrents although it does not necessarily lead to increased photocurrents.

Fig. 2 (a) Transient absorption spectra of Yb-doped and pristine WO₃ sintered at 550 °C. Bare FTO substrate was also measured as a reference. The spectra were measured at a 100 ns delay after excitation by a 355 nm laser. (b) Transient kinetic decays of Yb-doped and pristine WO₃ probed at 400 nm.

Photoelectrochemical (PEC) studies

The photocurrent densities and electrochemical impedances from the PEC testing are plotted in Fig. 3. Between the sintering temperatures of 450 and 550 °C, the photocurrent densities of Yb-doped WO₃ photocatalysts were both fairly constant and higher than for pristine WO₃. However, the photocurrents of Yb-doped WO₃ photocatalysts sintered at 400 and 600 °C were low. The charge transfer resistance values corresponded to the photocurrent densities albeit in an opposite trend. Fig. 3c summarises the contrasting effects of the sintering temperature from 450 to 550 °C (at 1.2 V Ag/AgCl) for Yb-doped WO₃ which maintained a constant level of 1.2 mA cm⁻² (corresponding to 525 to 626 Ω) and for pristine WO₃ which displayed a decreasing level from 1.0 to 0.4 mA cm⁻² (corresponding to 710 to 1300 Ω).

Fig. 3 Photoelectrochemical testing of Yb-doped and pristine WO₃ sintered at different temperatures. Photocurrent density versus voltage (left) and electrochemical impedance (right, at a constant bias voltage of 1.2 V and illuminated) for (a) Yb-doped WO₃ and (b) pristine WO₃. (c) Effects of sintering temperature on photocurrent densities and series resistances.

To further investigate the effects of Yb-doping on photocurrents, electrochemical impedance spectroscopy (EIS) measurements for each photocatalyst were carried out in the dark at 1000 Hz between −0.2 and 1.4 V to obtain capacitance values which were fitted to the Mott–Schottky (MS) equation as shown in Fig. 4. Except for samples sintered at 400 and 600 °C, the slope of each MS plot, which is equal to , was significantly lower for Yb-doped WO₃ compared to pristine WO₃, in other words doping resulted in increased donor densities (note: e is the electron charge, ε is the dielectric constant of WO₃ (50), ε_o is vacuum permittivity, and N_D is donor density).¹⁶ As Yb was co-sputtered at the same sputtering power for all of the doped samples, it is expected that N_D will be similar for all of the doped samples regardless of the sintering temperature. As shown in Fig. 4b, the N_D values ranged from 2.92 × 10²⁰ to 6.33 × 10²⁰ cm⁻³ for Yb-doped WO₃ versus a lower level of 6.08 × 10¹⁹ to 1.98 × 10²⁰ cm⁻³ for pristine WO₃ for sintering temperatures between 450 and 550 °C. The N_D values for both doped and pristine WO₃ were however low at the extreme temperatures (400 and 600 °C) when compared to the respective N_D values for the 450 to 550 °C range.

Fig. 4 Electrochemical impedance spectroscopy in dark conditions for Yb-doped and pristine WO₃ sintered at different temperatures. (a) Mott–Schottky plots to derive donor densities. The insets include plots for samples sintered at 600 °C. (b) Effects of sintering temperatures on donor densities.

Thus far, it has been shown that Yb-doped WO₃ photocatalysts demonstrated photocurrent enhancement with reduced charge transfer resistances. Raman spectroscopy and SIMS results (Fig. 1 and S1†) showed that Yb³⁺ dopants substituted W⁶⁺ in the crystal lattice which would have resulted in oxygen vacancies to form sub-stoichiometric WO_3−x. Oxygen vacancies created in PO₄ salts have been reported to widen the valence band and thus narrow the band gap to yield enhanced photoactivity.^17,18 However in the present study, a blue shift (hence there is no narrowing of the band gap) was observed in the transient absorption spectra instead (Fig. 2) which more probably reflects the presence of Yb oxides with wide band gaps. Besides, the carrier life time of Yb-doped WO₃ was shorter than that of pristine WO₃. The enhanced photocatalytic activity of Yb-doped WO₃ is attributed to increased donor densities as shown in the Mott–Schottky plots in Fig. 4. The electrons available from the 4f and 6s orbitals of Yb and the shallow electron donors from the oxygen vacancies¹⁹ arising from the substitution of W by Yb contributed to a higher electron density for the doped WO₃. Even at increased sintering temperatures where an amorphous layer formed at the WO₃/electrolyte interface would thicken,²⁰ the higher electron density would enable more free electrons to tunnel through to result in reduced charge transfer resistances and hence increased photocurrent densities. The higher electron density effect was sustained to sintering temperature of up to 550 °C as indicated by the stabilities of the photocurrent densities and the electrochemical impedances of the Yb-doped WO₃, as shown in Fig. 3c. In contrast, a lower electron density in the pristine WO₃ would be impeded by the amorphous layer formed at the WO₃/electrolyte interface and led to decreasing photocurrents and increasing electrochemical impedances over the same sintering temperature range. However at the extreme sintering temperatures of 400 and 600 °C, the photocurrent densities of both the Yb-doped and pristine WO₃ were lower as shown in Fig. 3c, which corresponds to the decreased electron density levels shown in Fig. 4. At 400 °C, both the doped and pristine WO₃ were not fully crystallised, as indicated by the broad Raman spectra for 400 °​C in Fig. 1, which implies the presence of grain boundaries to trap free electrons. On the other hand at 600 °C, although the crystallinities of both the doped and pristine WO₃ increased, as shown by the sharp peaks in the Raman spectra for 600 °​C in Fig. 1, the amorphous layer formed at the photocatalyst/electrolyte interface by oxidation²⁰ could have altered the effective dielectric constants of the photocatalysts and lowered the capacitance values and the free donor carrier densities. While the present study shows that doping a small amount of electron-rich Yb can increase donor densities and lead to higher PEC photocurrent densities, increasing the doping level of Yb in WO₃ continuously may nullify the benefits of increasing donor density due to the distortion of the WO₃ crystal structure which impedes charge transport and increases charge transfer resistance, as reported in our earlier study.⁷

To confirm that the photocurrents of the Yb-doped and pristine WO₃ photocatalysts sputter-deposited in this study were due to photoelectrochemical reaction of water, the water splitting products (H₂ and O₂) were collected and analysed using the in-house photocatalytic reactor coupled with a gas chromatograph. H₂ gas was detected at hourly intervals as shown in the gas evolution plot in Fig. 5; however O₂ gas was not so conclusively identified during the first half of the reaction (∼3 hours), which was probably due to the lower amount being produced and its higher solubility in water. Electrolysis in the dark with the same photocatalyst was also carried out to verify that no gaseous products were collected without light irradiation. 1.46 μmol of H₂ had been collected at the end of six hours of the photocatalytic reaction using Yb-doped WO₃. Given that the surface area of the photocatalyst was 1.5 × 1.0 cm², the H₂ yield was ∼2000 μmol h⁻¹ m⁻² which is considerably lower than those reported for WO₃ nanostructures.^21–23 Yet high photo-stability was demonstrated with only an ∼10% drop in the photocurrent at the end of 6 hours of photocatalysis, as shown in Fig. S3.†

Fig. 5 Gaseous H₂ evolution from Yb-doped and pristine WO₃ during photoelectrochemical water splitting. Electrolysis in the dark was also carried out to verify that no gaseous products were collected without light irradiation.

Conclusions

We prepared Yb-doped WO₃ photocatalysts via the co-sputtering of Yb and WO₃, followed by sintering at elevated temperatures in air from 400 to 600 °C. The Yb-doping effects on photoelectrochemical (PEC) properties were investigated using Raman, transient absorption and electrochemical impedance spectroscopies. Raman spectroscopic investigations showed that the intensities of lattice vibration and the vibration of W–O bonds in monoclinic WO₃ were reduced in Yb-doped WO₃ relative to pristine WO₃, which is due to the substitution of W⁶⁺ by Yb³⁺ dopants to form sub-stoichiometric WO_3−x. Transient absorption (TA) spectroscopy showed lower TA signals for Yb-doped WO₃ relative to the pristine WO₃. The recombination time of electrons and holes in Yb-doped WO₃ was shorter than for pristine WO₃. The electrochemical impedance spectra revealed higher levels of donor densities for Yb-doped WO₃. The Yb doping effect led to more stable and higher photocurrent densities with lower charge transfer resistances in PEC water splitting with Yb-doped WO₃ photocatalysts sintered between 450 and 550 °C. In contrast, photocurrent densities of the pristine WO₃ decreased steadily with increased charge transfer resistances.

Acknowledgements

This work was supported by the Institute of Materials Research and Engineering (IMRE), Agency for Science, Technology and Research (A*STAR) (Grant code: IMRE/12-1C0101).

References

1. D. V. Esposito, J. G. Chen, R. W. Birkmire, Y. Chang and N. Gaillard, Int. J. Hydrogen Energy, 2011, 36, 9632–9644.
2. C. Zhang, A. Boudiba, C. Navio, C. Bittencourt, M. G. Olivier, R. Snyders and M. Debliquy, Int. J. Hydrogen Energy, 2011, 36, 1107–1114.
3. Z. Zhao and M. Miyauchi, Angew. Chem., Int. Ed., 2008, 47, 7051–7055.
4. L. Zhou, J. Zhu, M. Yu, X. Huang, Z. Li, Y. Wang and C. Yu, J. Phys. Chem. C, 2010, 114, 20947–20954.
5. C. Feng, S. Wang and B. Geng, Nanoscale, 2011, 3, 3695–3699.
6. S. Han, J. Li, X. Chen, Y. Huang, C. Liu, Y. Yang and W. Li, Int. J. Hydrogen Energy, 2012, 36, 4724–4731.
7. S. L. Liew, Z. Zhang, T. W. Glenn Goh, G. S. Subramanian, H. L. Debbie Seng, T. S. Andy Hor, H.-K. Luo and D. Z. Chi, Int. J. Hydrogen Energy, 2014, 39, 4291–4298.
8. M. F. Daniel, B. Desbat and J. C. Lassegues, J. Solid State Chem., 1987, 67, 235–247.
9. C. V. Ramana, S. Utsunomiya, R. C. Ewing, C. M. Julien and U. Becker, J. Phys. Chem. B, 2006, 110, 10430–11035.
10. G. Wang, Y. Ling, H. Wang, X. Yang, C. Wang, J. Z. Zhang and Y. Li, Energy Environ. Sci., 2012, 5, 6180–6187.
11. S. R. Pendlebury, M. Barroso, A. J. Cowan, K. Sivula, J. Tang, M. Grätzel, D. Klug and J. R. Durrant, Chem. Commun., 2011, 47, 716–718.
12. F. M. Pesci, A. J. Cowan, B. D. Alexander, J. R. Durrant and D. R. Klug, J. Phys. Chem. Lett., 2011, 2, 1900–1903.
13. T. Yoshihara, R. Katoh, A. Furube, Y. Tamaki, M. Murai, K. Hara, S. Murata, H. Arakawa and M. Tachiya, J. Phys. Chem. B, 2004, 108, 3817–3823.
14. Y. Ma, S. R. Pendlebury, A. Reynal, F. Le Formal and J. R. Durrant, Chem. Sci., 2014, 5, 2964–2973.
15. J. Tang, J. R. Durrant and D. R. Klug, J. Phys. Chem. B, 2004, 108, 3817–3823.
16. M. A. Butler, J. Appl. Phys., 1977, 48, 1914–1920.
17. Y. Lv, Y. Zhu and Y. Zhu, J. Phys. Chem. C, 2013, 117(36), 18520–18528.
18. L. J. Chen, Y. J. Zhao, J. Y. Luo and Y. Y. Xia, Phys. Lett. A, 2011, 375, 934–938.
19. S. K. Deb, Phys. Rev. B: Condens. Matter Mater. Phys., 1977, 16, 1020–1024.
20. F. Lin, C. P. Li, G. Chen, R. C. Tenent, C. A. Wolden, D. T. Gillaspie, A. C. Dillon, R. M. Richards and C. Engtrakul, Nanotechnology, 2012, 23, 255601.
21. C. W. Lai and S. Sreekantan, Int. J. Hydrogen Energy, 2013, 38, 2156–2166.
22. J. Lin, P. Hu, Y. Zhang, M. Fan, Z. He, C. K. Ngaw, J. S. C. Loo, D. Liao and T. T. Y. Tan, RSC Adv., 2013, 3, 9330–9336.
23. Q. Chen, J. Li, B. Zhou, M. Long, H. Chen, Y. Liu, M. Long, W. Cai and W. Shangguan, Electrochem. Commun., 2012, 20, 153–156.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra26254b

---