α-Fe₂O₃ nanopillar arrays fabricated by electron beam evaporation for the photoassisted degradation of dyes with H₂O₂

Abstract

Vertically aligned α-Fe₂O₃ nanopillar arrays (NPAs) were fabricated by thermally oxidizing Fe NPAs on Si, quartz and F-doped SnO₂ (FTO) substrates prepared by glancing angle e-beam deposition (GLAD). The photocatalytic activity of these NPAs was evaluated by measuring the photodegradation of crystal violet (CV) and methyl orange (MO) in the presence of H₂O₂ under visible light irradiation. Moreover, the photoelectrochemical (PEC) performance was also studied. Typically the sample oxidized at 400 °C exhibits both the highest degradation efficiency and photocurrent density compared with those oxidized at other temperatures (e.g. 300 °C, 350 °C, 450 °C, 500 °C). This phenomenon might be attributed to a trade-off between two opposite effects. On the one hand, with the increase of the oxidation temperature, the improvement of NPAs' crystallinity will enhance the photocatalytic performance accordingly. On the other hand, increasing oxidation temperature may cause the reduction of oxygen vacancies on the NPAs' surface, which are regarded as the photoreaction active sites. This will thus degrade the photocatalytic performance.

---

Introduction

Since the first observation of photo electrochemical (PEC) splitting of water over TiO₂ electrodes by Fujishima and Honda in 1972,¹ semiconductor-based photocatalysis has attracted the increasing attention of researchers, as it provides an effective way to convert solar energy into chemical and electrical energy, and can be used in diverse range of fields, such as decomposing organic pollutants, hydrogen generation, etc.² However, as a wide band gap (∼3.0 eV) semiconductor, TiO₂ works only in the UV region which is only ∼4% of the incident solar radiation.³ In order to fully utilize the solar energy, efficient semiconductors that work in the visible light region as well as the UV region are highly demanded. However, developing brand-new high-performance catalysts that can efficiently extend photocatalytic activity into the visible spectral region still remains a big challenge. Actually, in addition to the effort in developing novel photocatalysts, traditional materials can also be modified by various strategies to achieve high catalytic activity under visible light irradiation.

Hematite (α-Fe₂O₃), an abundant environmentally friendly n-type semiconductor, has been widely used as catalysts,⁴ pigment,⁵ gas sensors.⁶ It has a band gap of ∼2.1 eV, and works in the visible light region with a high theoretical solar-to-hydrogen efficiency of ∼17% and good stability against photocorrosion.⁷ Therefore, α-Fe₂O₃ has attracted tremendous interests in recent years.^8–15 However, a drawback of α-Fe₂O₃ catalyst is the short hole diffusion length (2–4 nm) due to its poor minority charge carrier mobility (0.2 cm² V⁻¹ s⁻¹), which results in a high recombination rate of electrons and holes, a very short excited state lifetime (∼10 ps), and poor electrical conductivity.^16,17 To address or to overcome this issue, a lot of efforts have been paid to synthesize nanostructures of hematite, such as nanotubes, nanoparticles, nanocubes, nanowires, nanofibers, nanorods, and hierarchical structures.^8–15,18 Besides morphology control, other ways like doping, constructing heterojunctions, surface functionalization of nanostructures, etc., can also be used to optimize its photocatalytic performance.^9,19–22

Photocatalysts in previous reports are usually in the form of powders, which is hard to handle and collect. Here we report the synthesis of vertically aligned hematite nanopillar arrays on specific substrates to form self-standing structures, which are much easier to recycle. Furthermore, they do prevent the occurrence of secondary pollution which means more than recycles. The α-Fe₂O₃ arrays are fabricated by thermally oxidizing Fe nanopillars using a glancing angle deposition (GLAD) technique. The PEC and degradation performance under visible light radiation are evaluated. The influence of the oxidization temperature on the growth of hematite nanopillars and their catalytic performance are also investigated.

Experimental methods

Fabrication of α-Fe₂O₃ nanopillar arrays

Vertically aligned Fe nanopillars were deposited by the e-beam GLAD technique onto three different substrates as below: (1) planar silicon substrates with (001) orientation for material characterization, (2) quartz substrates for degradation reaction, and (3) F-doped SnO₂ (FTO) substrates (20 Ω per square) for PEC test. All the substrates were ultrasonically cleaned in sequence in acetone, ethanol and deionized (DI) water baths for 10 min respectively. Prior to the deposition, the chamber was evacuated to a vacuum level above 1 × 10⁻⁸ torr. During deposition, the vapor flux incident angle was set to ∼86° off the surface normal to the substrates, rotating at a speed of 10 RPM. The deposition rate (∼0.75 nm s⁻¹) and the height of the nanopillars were monitored by a quartz crystal microbalance. After deposition, the samples were subjected to the oxidation process in a quartz tube furnace for 1 h at a ramp of 2 °C min⁻¹ under ambient conditions to temperatures of 300, 350, 400, 450 and 500 °C, respectively. For comparison with powder sample, we prepared α-Fe₂O₃ sol using Fe₂O₃ powders. Then we paint it onto the same SiO₂ substrate to test properties.

Characterization and measurement

The morphology and structure of the samples were examined by field-emission scanning electron microscope (FESEM, JEOL-7001F), high-resolution transmission electron microscope (HRTEM, JEOL-2011) and Raman spectroscopy (LABRAM HR800, excitation wavelength of 633 nm), respectively. The chemical structure of the samples was analyzed by X-ray photon electron spectrometer (XPS, Perkin Elmer PHI 5300), and the binding energy was calibrated with the reference to the C 1s peak centered at 284.6 eV. The optical properties of the samples were examined by a UV-vis spectrometer (Perkin Elmer Lambda 35) in a wavelength range of 200 to 900 nm at room temperature.

The steady sate current density (j–V) and electrochemical impedance spectroscopy (EIS) measurements were carried out by an electrochemistry workstation (CHI 660D, Chenhua instrument). The nanostructured films were used as the working electrode, an Ag/AgCl electrode (saturated KCl) and Pt sheet were used as the reference and counter electrodes, respectively. The working electrode was illuminated with a 300 W Xe lamp. An ultraviolet filter was placed between the light source and the quartz cell to cut off the UV light in wavelength <420 nm. Photocurrent densities were measured in the light on–off process with a pulse of 30 s under visible light illumination (200 mW cm⁻²) at 0.4 V bias vs. Ag/AgCl electrode.

The photocatalytic activity of the hematite nanopillar arrays was evaluated by the photodegradation of crystal violet (CV) and methyl orange (MO) photodegradation with the light source of 300 W Xe lamp at ambient temperature. The sample on quartz substrate (15 mm × 15 mm) was placed in a quartz cell containing 5 mL of CV (5 μM) and MO (5 μM). Prior to light irradiation, the photocatalyst was immersed into solution in the dark room for 30 min to reach an adsorption/desorption equilibrium, then 12 μL of 30 wt% H₂O₂ was added to the solution, and the Xe lamp was turned on for different time spans. After that, the UV-vis absorbance spectra of the dye aqueous solution were obtained to measure the dye concentration after photodegradation.

Results and discussion

Fig. 1(a), (c), (e) and (f) show typical top-view SEM images of the as-deposited film and those oxidized at 300, 400 and 500 °C respectively. Fig. 1(b) and (d) show typical side-view SEM images of the as-deposited film and the film oxidized at 300 °C, respectively. It can be seen that the as-deposited film consists of vertically aligned Fe nanopillars, with a diameter of ∼30 nm and a length of ∼200 nm. After thermal oxidation, as shown by Fig. 1(c) and (d), the diameter of the nanopillars was increased to ∼55 nm, and they touched with each other slightly. No obvious difference is observed among samples oxidized at 300, 400 and 500 °C.

Fig. 1 SEM images of the different samples: (a) Fe nanopillar film; (b) cross-section of Fe nanopillar film; (c) α-Fe₂O₃ films oxidized at 300 °C; (d) cross-section of sample oxidized at 300 °C; (e) α-Fe₂O₃ films oxidized at 400 °C; (f) α-Fe₂O₃ films oxidized at 500 °C.

Fig. 2 shows the XRD patterns of the as-prepared samples. The XRD pattern of the Fe sample is well matched with the standard spectrum of Fe (JCPDS 50-1275). When the oxidation temperature are 300 and 350 °C, the main phase is α-Fe₂O₃ (JCPDS 33-0664) with a little diffraction peak located at 30.1°, which can be ascribed to the (220) plane of γ-Fe₂O₃ (JCPDS 39-1346). This indicates that a few amount of γ-Fe₂O₃ co-existing in the sample. When the annealing temperature is higher than 350 °C, the phase will be totally α-Fe₂O₃.

Fig. 2 XRD spectra of the as-deposited Fe films and oxidized samples at different temperatures.

Fig. 3 shows the Raman spectra of the as-deposited film and films oxidized at temperatures ranging from 300 to 500 °C, measured at room-temperature. It can be seen that, all samples oxidized at temperatures above 300 °C demonstrate the characteristic peaks corresponding to α-Fe₂O₃, i.e. peaks at 226 and 493 cm⁻¹ of the A_1g mode, and peaks at 295, 411 and 610 cm⁻¹ of the E_g mode.^23,24 The peak centered at 1318 cm⁻¹ may be caused by the two-magnon scattering due to the interaction between two magnons created on antiparallel close spin sites. It is noticed that the films oxidized at low temperatures (e.g. 300 °C, 350 °C) exhibited broader and lower intensity Raman peaks, whilst the films oxidized at higher temperatures exhibited sharper and higher intense Raman peaks because of better crystallinity.²⁵

Fig. 3 Raman spectra of the as-deposited Fe films and oxidized samples at different temperatures.

Fig. 4 shows XPS spectra of the films oxidized at 300, 400 and 500 °C, where the binding energies were corrected by referencing the C 1s line to 284.6 eV. The peaks located at ∼710 and ∼724 eV are in good agreement with the Fe 2p^3/2 and Fe 2p^1/2 binding energy of Fe³⁺.²⁶ The little shakeup satellite structure between two main peaks is the fingerprint of the electronic structure of Fe³⁺.²⁷ Based on Raman and XPS analysis, it can be known that Fe nanopillars are oxidized into hematite nanopillars by the oxidization treatment at 300–500 °C.

Fig. 4 Fe 2p XPS spectra of samples oxidized at 300, 400 and 500 °C, respectively.

The morphology and structure of individual nanopillars after oxidization were also investigated by TEM analysis. Fig. 5(a)–(d) shows typical low-magnification and high-magnification HRTEM images of the nanopillars oxidized at 300 and 400 °C, respectively. Insets of Fig. 5(b) and (d) show their corresponding selected-area electron diffraction (SAED) patterns. Patterns are of 300 and 400 °C oxidation temperature. We check α-Fe₂O₃ PDF card with some indices of crystal face to confirm its phase, marked in figures. The SAED rings of the sample oxidized at 400 °C correspond to the (113), (110), (104), (012), and (101) planes of α-Fe₂O₃, respectively. It can be seen that the nanopillars are of 1D cone-shaped rod-like morphology, ∼50 nm in diameter and ∼200 nm in length. The HRTEM image of 400 °C shows that the lattice fringes are about 0.20 nm, which is in good agreement with the interplanar spacing of α-Fe₂O₃ (113) planes. Similarly, the 0.27 nm lattice fringes of 300 °C sample can be attributed to the α-Fe₂O₃ (104) plane. It is worthwhile to mention that diffraction rings of α-Fe₂O₃ become more distinguishable at higher oxidization temperatures (e.g. 400 °C), which is in good agreement with the Raman analysis.

Fig. 5 TEM images and HRTEM images: (a) Fe nanopillar oxidized at 300 °C; (b) HRTEM of white box in (a), where the inset is the corresponding SAED pattern; (c) Fe nanopillar oxidized at 400 °C; (d) HRTEM of white box in (c), where the inset is the corresponding SAED pattern.

PEC experiments were performed to evaluate the photoelectrochemical property of these oxidized samples in a three-electrode cell, using the oxidized samples as the working electrode with an exposed area of 1.5 cm² under visible light irradiation. Here 0.01 M KOH aqueous solution was used as the electrolyte. Fig. 6(a) shows the photocurrent density measured for these samples with light on and off, at a bias potential of +0.4 V, from which photo-generated electron–hole separation is clearly observed.²⁸ It can be seen from Fig. 6(a) that for samples oxidized at 300, 350 and 400 °C, an increase in the photocurrent was observed for samples. For samples oxidized at >400 °C, i.e., 450 and 500 °C, a monotonic decrease in the photocurrent density was observed. The largest photocurrent density (i.e. 1.4 mA cm⁻²) was obtained with the sample oxidized at 400 °C, suggesting a higher efficiency in the separation of electrons and holes and a better photocatalytic activity.²⁹ Fig. 6(b) shows the photocurrent density of the sample oxidized at 400 °C, as a function of time with the light on and off. It can be seen that the photocurrent density is quite stable and repeatable.

Fig. 6 (a) Current versus time measurements of α-Fe₂O₃ at different oxidation temperature during visible light illuminations under 0.4 V versus Ag/AgCl electrode bias. (b) Current versus longer time measurements of α-Fe₂O₃ at 400 °C; (c) j–V characteristics of all samples; (d) Nyquist plots under visible light at open circuit voltage of all oxidized sample.

Fig. 6(c) shows the photocurrent density of these samples measured at various bias potentials ranging from −0.2 V to 0.4 V. The open circuit potential (V_oc) relates closely to the difference in the apparent Fermi level between α-Fe₂O₃ and the reference electrode,³⁰ is dependent on the oxidation temperature. From Fig. 6(c), it can be seen that the sample oxidized at 400 °C has a minimum of V_oc of ∼−0.049 V, while the sample oxidized at 500 °C has a maximum value of ∼0.183 V. It is well known that a more negative V_oc means better electron and hole separation in semiconductors.^31–33 This can be responsible partially for the largest and smallest photocurrent density observed for the two samples.

EIS measurements were conducted for all samples at frequencies ranging from 0.1 Hz to 100 kHz in 0.01 M KOH electrolyte at open circuit voltage conditions under visible light irradiation. An equivalent circuit was used to fit the Nyquist plots, as shown in Fig. 6(d), where the R_s, C₁, R₁, R_ct, and C₂ are the series resistance, the capacitance of α-Fe₂O₃ and the electrolyte, the resistance of the holes trapped at surface states, the resistance of holes transferred to the electrolyte through surface states and the capacitance of surface states, respectively. All above values are obtained and listed in Table 1. The specific values of C₁ and C₂ for different oxidation temperatures are basically in the same order of magnitude with little change, respectively. One sees that all samples exhibited a similar R_s with a value around 0.5 Ω. While R_ct is normally larger than R₁ two magnitude which indicates the former plays a more important role in circuit. Though values of R_ct on 400, 450 and 500 °C are close, R₁ increases from 349.5 to 3092 Ω gradually, which also leads to reduction on photocurrent.^34–36 Based on above analysis it is concluded that the oxidation temperature of 400 °C owns the maximum photocurrent, which is also confirmed by the PEC results.

Table 1 The calculated values of R_s, R₁ and R_ct through fitting of the experimental impedance spectra based upon the proposed equivalent circuit in Fig. 6(d)

Oxidation temperature (°C)	300	350	400	450	500
Rs (Ω)	0.52	0.54	0.51	0.51	0.52
C1 (F)	1.2 × 10^−5	3.1 × 10^−6	1.9 × 10^−6	1.3 × 10^−6	1.1 × 10^−6
R1 (Ω)	31.3	139.7	349.5	431.1	3092
C2 (F)	8.9 × 10^−5	4.6 × 10^−5	9.3 × 10^−5	8.4 × 10^−5	6.7 × 10^−5
Rct (Ω)	106 620	102 700	57 749	65 029	58 155

---

The photoactivities of the obtained α-Fe₂O₃ NPAs are investigated by the degradation of CV with H₂O₂ under visiblelight irradiation, and the degradation processes were tested by recording the characteristic absorbance peak at 584 nm. Fig. 7(a) shows the time course of the decrease in the dye concentration. In absence of α-Fe₂O₃, the CV is degraded by H₂O₂ 79.4% until 60 min. While the enhancement is observed after the introduction of α-Fe₂O₃ film; for example, the degradation efficiency of sample oxidized at 400 °C reaches to 90.9% due to the constitution of a Fenton system.³⁷ After data processing, all reaction follows first-order kinetics as shown in Fig. 7(b). The rate constant of 400 °C oxidation sample presented by slope is 1.34 times higher than that in absence of α-Fe₂O₃.

Fig. 7 (a) Time course of the decrease in the concentration; (b) ln(C₀/C) for the degradation of CV with visible irradiation under different conditions; (c) (αE_photon)² versus E_photon plots of corresponding films; (d) dependence of photocurrent density, degradation efficiency and bandgap energy on different oxidation temperatures.

According to previous research, α-Fe₂O₃ can be photodecomposed because of the photoreduction of Fe³⁺ to Fe²⁺ under light illumination. Once Fe²⁺ is generated, either it reacts with H₂O₂ generating hydroxy radicals or it detaches from the oxide surface, leaving a vacancy on the semiconductor surface. Once effective collision of dye molecule and the holes happens, CV and/or CV^± ions form. After the reaction between ions and hydroxyl radicals produced by the decomposition of H₂O₂, CV is degraded completely.³⁸ The very good photostability of the Fe₂O₃ film strongly suggests that the reaction between surface Fe²⁺ and H₂O₂ is kinetically faster than the Fe²⁺ dissolution reaction in the Fe₂O₃ catalyst surface, thus preventing occurrence of catalyst photocorrosion. We also supplemented experiment choosing methyl orange (MO) as model dye solution (5 μM) in Fig. S1,† which cannot be degraded under visible lights. According to figure, α-Fe₂O₃ samples with same quantity of H₂O₂ show the similar regular in photocatalysis, in which sample oxidized at 400 °C also shows best degradation performance. And we compare the photocatalytic performance of Fe₂O₃ nanopillar arrays with Fe₂O₃ powders shown in same picture. The sectional view of powder Fe₂O₃ sample is shown in Fig. S2.†

The band gap energy (E_g) of semiconductor following the equation:

αE_photon = K(E_photon − E_g)^1/2,

where K, α, E_photon and E_g are a constant, the absorption coefficient, the discrete photon energy and the band gap energy, are estimated by calculating the intercept of an extrapolated linear fit to the experimental date of a plot of (αE_photon)² vs. E_photon yielded E_g for a direct transition³⁹ as shown in Fig. 7(c). The band gaps of samples from 300 °C to 500 °C are figured out to be 2.27, 2.26, 2.15, 2.16 and 2.18 eV, respectively. It is suggested again that the sample oxidized at 400 °C with narrowest E_g could make more use of light energy as a result of best photocatalytic properties. The change of photocurrent density (under +0.4 V bias voltage), degradation efficiency and E_g values over oxidation temperature are summarized in Fig. 7(d) and it is obvious to find 400 °C is the optimal treatment temperature in this system. With the increase of oxidation temperature, grains grow according to the XRD analysis which leads to narrower band gap.⁴⁰ However, from XPS analysis in Table 2, with the increase of annealed temperature, oxygen defects on oxide surface becomes less. And this just make semiconductors' band gap become broader.⁴¹ Two opposite factor cause appearance of 400 °C minimum E_g value.

Table 2 Area ratio of oxygen vacancy/oxygen lattice in Fig. 8

Oxidation temperature (°C)	300	400	500
Area ratio	1.631	1.109	0.650

---

According to analysis mentioned above, the strong intensity of the Raman peak and small full width at half maximum (FWHM) form progressively with improving oxidation treatment temperature, which means better crystallinity and higher crystallinity can truly enhance the PEC and photodegradation performance of α-Fe₂O₃ nanostructures.^31,42,43 As for why there is an inflexion on property, we look into more information of XPS about oxygen element. The deconvolution peaks display in Fig. 8, O 1s spectrum were observed from 529.0 to 536.0 eV. The low binding energy (BE) component observed at ∼529 eV corresponds to the main lattice oxygen (O_lat) forming oxide with iron. The bands around ∼532 eV can be assigned to adsorbed oxygen (O_ads) while the peak between them (∼521 eV) indicates the ionization of oxygen species that could allow compensation for some deficiencies (O_def) connected in part to the variations in the concentration of oxygen vacancies in iron oxides.^3,44 Based on Table 2, it is obvious to find the area ratio of oxygen vacancy/oxygen lattice decreases step by step from 1.63 to 0.65 when raising the treatment temperature.

Fig. 8 XPS spectra of O 1s of α-Fe₂O₃ oxidized at (a) 300 °C, (b) 400 °C, (c) 500 °C. Red line and black dash lines are the fitted curve and original data, respectively.

Combining former research analysis, oxygen vacancies can trap the photogenerated electrons thus reducing the electron–hole recombination and enhancing photocurrent density and photocatalytic character.⁴⁵ When higher oxidation temperature was carried out, the photocatalytic activity of the α-Fe₂O₃ thin films gradually increases due to the improvement of crystallization. However it also reduces oxygen vacancies existing on the surface of oxide, which eventually cuts down active sites and weakens the catalytic performance.

Conclusions

In conclusion, we have introduced the glancing angle e-beam deposition technique with oxidation treatment of vertically aligned α-Fe₂O₃ nanopillars. The PEC and photodegradation properties of samples obtained at different oxidation temperatures from 300 °C to 500 °C are investigated. There are two factors (crystallinity and oxygen vacancies) affect the performances of photocatalyt. With higher oxidation temperature, the improvement of crystallinity will enhance the photocatalytic property while the reduction of oxygen vacancies on the surface of material regarded as photoreaction active sites weakens the performance. The results show that Fe nanopillar arrays oxidized at 400 °C in air demonstrates both the highest photocurrent density (1.4 mA cm⁻² at 0.4 V vs. Ag/AgCl electrode) and the degradation efficiency (90.97% in 1 h) on CV in presence of H₂O₂.

Acknowledgements

The authors are grateful to the financial support by the National Natural Science Foundation of China (grant No. 51372135 and 51572148).

Notes and references

1. A. Fujishima and K. Honda, Nature, 1972, 238, 37–38.
2. K. Sivula, F. le Formal and M. Gratzel, ChemSusChem, 2011, 4, 432–449.
3. S. M. Park, T. Ikegami and K. Ebihara, Thin Solid Films, 2006, 513, 90–94.
4. X. Wang, J. Wang, Z. Cui, S. Wang and M. Cao, RSC Adv., 2014, 4, 34387–34394.
5. S. A. Jadhav, R. Bongiovanni, D. L. Marchisio, D. Fontana and C. Egger, Pigm. Resin Technol., 2014, 43, 219–227.
6. Y. V. Kaneti, Q. M. D. Zakaria, Z. Zhang, C. Chen, J. Yue, M. Liu, X. Jiang and A. Yu, J. Mater. Chem. A, 2014, 2, 13283–13292.
7. T. W. Hamann, Dalton Trans., 2012, 41, 7830–7834.
8. Z. J. Zhang, B. Q. Wei and P. M. Ajayan, Appl. Phys. Lett., 2001, 79, 4207–4209.
9. R. H. Wang, C. H. Xu, J. Sun and L. Gao, Sci. Rep., 2014, 4, 7171.
10. K. Nagato, M. Furubayashi, T. Hamaguchi and M. Nakao, Microelectron. Eng., 2011, 88, 2697–2699.
11. Y. L. Cheng, B. L. Zou, C. J. Wang, Y. J. Liu, X. Z. Fan, L. Zhu, Y. Wang, H. M. Ma and X. Q. Cao, CrystEngComm, 2011, 13, 2863–2870.
12. K. Woo, H. J. Lee, J. P. Ahn and Y. S. Park, Adv. Mater., 2003, 15, 1761–1764.
13. Z. Liu, B. L. Lv, D. Wu, Y. H. Sun and Y. Xu, Particuology, 2013, 11, 327–333.
14. Q. Wei, Z. C. Li, Z. J. Zhang and Q. Zhou, Mater. Trans., 2009, 50, 1351–1354.
15. W. H. Peng, C. L. Zhu, S. M. Zhu, F. Yao, Y. Li and D. Zhang, J. Mater. Sci., 2013, 48, 4336–4344.
16. H. Dotan, K. Sivula, M. Gratzel, A. Rothschild and S. C. Warren, Energy Environ. Sci., 2011, 4, 958–964.
17. B. M. Klahr and T. W. Hamann, J. Phys. Chem. C, 2011, 115, 8393–8399.
18. Z. Yang, J. Xu, C. Wu, H. Jing, P. Li and H. Yin, Appl. Catal., B, 2014, 156–157, 249–256.
19. C. Q. Zhu, Y. R. Li, Q. Su, B. A. Lu, J. Q. Pan, J. W. Zhang, E. Q. Xie and W. Lan, J. Alloys Compd., 2013, 575, 333–338.
20. Q. Y. Hao, S. A. Liu, X. M. Yin, Z. F. Du, M. Zhang, L. M. Li, Y. G. Wang, T. H. Wang and Q. H. Li, CrystEngComm, 2011, 13, 806–812.
21. A. J. R. Hensley, Y. C. Hong, R. Q. Zhang, H. Zhang, J. M. Sun, Y. Wang and J. S. McEwen, ACS Catal., 2014, 4, 3381–3392.
22. Q. Wei, Z. J. Zhang, Z. C. Li, Q. Zhou and Y. Zhu, J. Phys. D: Appl. Phys., 2008, 41, 202002.
23. A. A. Tahir, K. G. U. Wijayantha, S. Saremi-Yarahmadi, M. Mazhar and V. Mckee, Chem. Mater., 2009, 21, 3763–3772.
24. D. Bersani, P. P. Lottici and A. Montenero, J. Raman Spectrosc., 1999, 30, 355–360.
25. S. H. Huang and J. Liu, Chin. Phys. B, 2014, 23(5), 058105.
26. J. S. Xu and Y. J. Zhu, CrystEngComm, 2012, 14, 2702–2710.
27. T. Fujii, F. M. F. de Groot, G. A. Sawatzky, F. C. Voogt, T. Hibma and K. Okada, Phys. Rev. B: Condens. Matter Mater. Phys., 1999, 59, 3195–3202.
28. Q. Kang, J. Y. Cao, Y. J. Zhang, L. Q. Liu, H. Xu and J. H. Ye, J. Mater. Chem. A, 2013, 1, 5766–5774.
29. Y. Liu, C. S. Xie, J. Li, T. Zou and D. W. Zeng, Appl. Catal., A, 2012, 433, 81–87.
30. D. R. Baker and P. V. Kamat, Adv. Funct. Mater., 2009, 19, 805–811.
31. Z. Xie, X. X. Liu, W. P. Wang, X. J. Wang, C. Liu, Q. Xie, Z. C. Li and Z. J. Zhang, Nano Energy, 2015, 11, 400–408.
32. M. Jakob, H. Levanon and P. V. Kamat, Nano Lett., 2003, 3, 353–358.
33. A. Wood, M. Giersig and P. Mulvaney, J. Phys. Chem. B, 2001, 105, 8810–8815.
34. J. C. Chou, S. A. Lin, C. Y. Lee and J. Y. Gan, J. Mater. Chem. A, 2013, 1, 5908–5914.
35. Y. Hou, F. Zuo, A. Dagg and P. Y. Feng, Nano Lett., 2012, 12, 6464–6473.
36. K. Y. Yoon, J. S. Lee, K. Kim, C. H. Bak, S. I. Kim, J. B. Kim and J. H. Jang, ACS Appl. Mater. Interfaces, 2014, 6, 22634–22639.
37. B. Xu, B. B. Huang, H. F. Cheng, Z. Y. Wang, X. Y. Qin, X. Y. Zhang and Y. Dai, Chem. Commun., 2012, 48, 6529–6531.
38. H. Xie, Y. Z. Li, S. F. Jin, J. J. Han and X. J. Zhao, J. Phys. Chem. C, 2010, 114, 9706–9712.
39. L. Xu, J. X. Xia, L. G. Wang, J. Qian, H. M. Li, K. Wang, K. Y. Sun and M. Q. He, Chem. - Eur. J., 2014, 20, 2244–2253.
40. J. G. Yu, H. G. Yu, B. Cheng, X. J. Zhao, J. C. Yu and W. K. Ho, J. Phys. Chem. B, 2003, 107, 13871–13879.
41. P. Zhan, Z. Xie, Z. C. Li, W. P. Wang, Z. J. Zhang, Z. X. Li, G. D. Cheng, P. Zhang, B. Y. Wang and X. Z. Cao, Appl. Phys. Lett., 2013, 102, 071914.
42. R. Rajendran, Z. Yaakob, M. Pudukudy, M. S. A. Rahaman and K. Sopian, J. Alloys Compd., 2014, 608, 207–212.
43. P. M. Kouotou, Z. Y. Tian, H. Vieker, A. Beyer, A. Golzhauser and K. Kohse-Hoinghaus, J. Mater. Chem. A, 2013, 1, 10495–10504.
44. J. C. Dupin, D. Gonbeau, P. Vinatier and A. Levasseur, Phys. Chem. Chem. Phys., 2000, 2, 1319–1324.
45. R. Lv, T. Wang, F. L. Su, P. Zhang, C. J. Li and J. L. Gong, Nano Energy, 2014, 7, 143–150.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra17894k

---