ZnO nanoparticle-functionalized WO₃ plates with enhanced photoelectrochemical properties

Abstract

In this work, ZnO NPs-functionalized WO₃ vertical plate-like arrays were first fabricated on FTO with a hydrothermal process for WO₃ vertical plate-like arrays and an electrodeposition process for the functionalization of ZnO. The ZnO nanoparticles are preferentially loaded on the active points of WO₃ in the shape of a sphere about 10 nm. The samples were characterized by various techniques, including X-ray diffraction (XRD), scanning electron microscopy (SEM) and transmission electron microscopy (TEM). Photoelectrochemical properties were investigated by photoelectrochemical measures, such as linear sweep voltammograms, electrochemical impedance spectroscopy (EIS), intensity-modulated photocurrent spectroscopy (IMPS) and incident photon to current conversion efficiency (IPCE). The results show that the photocurrent of WO₃ increases from 0.88 to 1.23 mA cm⁻² at 1.2 V (vs. Ag/AgCl) after functionalized with ZnO. Furthermore, the lifetime of the electron–hole has been prolonged from 6.44 to 8.56 ms, but there is no decrease in the electron transport time. In this case, the enhancement of the photoelectrochemical performance is attributed to effective transfer of photo-generated holes so as to retard the recombination of electrons and holes.

---

1. Introduction

Since the discovery of photoelectrochemical (PEC) water splitting using n-type TiO₂, hydrogen production by water dissociation has attracted great interest.¹ As a cheap and nontoxic material, TiO₂ and modified TiO₂ have been used as photocatalysts.^2–5 However, the wide bandgap energy which only allows absorption of UV light, restricts its efficiency in PEC devices. With the research progress of photoelectrochemical water splitting and photocatalysis, various kinds of semiconductors have also been invented and employed.^6–9 Among them, WO₃, which is widely used as a photocatalyst and gases sensor, is a suitable semiconductor because of some promising characteristics such as a suitable band gap (2.5–2.8 eV) that can utilize a portion of visible light and excellent stability to avoid photocorrosion.^10–14

In order to further improve the photoelectrochemical performance of WO₃ to fulfill the need of photolysis of water, two strategies have been adopted. One is to change its particle size or morphology, such as mesoporous, flower, nanorod and nanotube-based bundles.^15–18 For the WO₃ loaded on the FTO or ITO, the films have been prepared as nanowires, nanorods and nanoflakes.^11,19,20 The other approach to improve the photoelectrochemical properties is metal or non-metal element doping,^21–23 noble metal deposition,^24–26 graphene modifying^27,28 or semiconductor coupling.^29–32 Among them, semiconductor coupling is an effective method to promote the separation of the photo-generated electron–hole pairs because the valence and conduction band positions are different from each other. Therefore, the recombinations of electrons and holes are stymied, and the efficiency of water splitting is enhanced.

For the WO₃ based semiconductor material, numerous semiconductors (BiVO₄, NiWO₄, ZnWO₄ and C₃N₄) were used as a second semiconductor to form composites with WO₃.^29–32 All of these systems indicated that an appropriate amount of second semiconductor is best for the efficiency of photolysis of water. It is because that the excessive second semiconductor might play the role of obstacles that impede the contact of the WO₃ surface and the electrolyte solution. The second semiconductor loaded on the active sites of the base semiconductor can accelerate the separation of photo-generated electron–hole. So the selective loading is a great approach to improve the photoelectrochemical performance of WO₃.

In this paper, we selectively loaded ZnO on the active sites of WO₃ plates preferentially via the electrodeposition and the electrochromism reaction of WO₃. Moreover, we have studied the photoelectrochemical performances by liner sweep voltammogram, electrochemical impedance spectroscopy (EIS), intensity-modulated photocurrent spectroscopy (IMPS) and incident photon to current conversion efficiency (IPCE). It was found that the visible-light-driven photoelectrochemical properties of ZnO nanoparticles-functionalized WO₃ were enhanced as compared to the pristine WO₃ plates, and the possible mechanism for the enhanced photoelectrochemical performances was discussed on the basis of the obtained experimental results.

2. Experimental section

2.1 Preparation of WO₃/ZnO

WO₃ plates were prepared from Na₂WO₄ by hydrothermal methods that had been well documented in our previous work.²⁰ A three electrode system was used for electrodeposition. The as prepared WO₃ film was used as a working electrode (1 × 1.5 cm²) while platinum sheet and Ag/AgCl (saturated KCl) were used as counter electrode and reference electrode, respectively. The electrodepositing process was carried out at the potential of −0.7 V (vs. Ag/AgCl) for 30 s from unstirred aqueous solution of 0.01 mol L⁻¹ Zn(NO₃)₂ and 0.1 mol L⁻¹ KCl. The electrolyte (bath) temperature was maintained at 55 °C after purged with oxygen for 10 min. Then the product was annealed at 350 °C for 30 min under air. So the sample was fabricated and labelled as WO₃/ZnO. For comparison, WO₃ plates film (labelled as WO₃) was also prepared without electrodeposition.

2.2 Characterization

For the structural studies, X-ray diffraction (XRD) patterns were recorded using X-ray diffractometer (XRD, D/Max2250) in the diffraction angle (2θ) between 10° and 70°. Scanning electron microscopy (SEM, JSM6700F) and transmission electron microscope (TEM, TECNAI G2 F20) were used to observe the morphologies and microstructures of the samples. UV-Vis diffused reflectance spectroscopy was carried out by an UV-Vis spectrometer (UV-Vis, Shimadzu 2450). XPS measurements were taken using a X-ray photoelectron spectrograph (XPS, ESCALAB 250Xi, ThermoFisher-VG Scientific). Photoelectrochemical test systems were composed of a standard three-electrode configuration. The WO₃ or WO₃/ZnO film with a Pt counter electrode and an Ag/AgCl reference electrode were immersed in a 0.2 mol L⁻¹ Na₂SO₄ solution. A 500 W Xe lamp served as the visible light source, and a 420 nm cutoff filter was placed on the path of the Xe lamp to remove UV irradiation.

3. Results and discussion

3.1 Preparation of characterization the films

As shown in Fig. 1, the reaction mechanism can be summarized as:

Zn(NO₃)₂ = Zn²⁺ + 2NO₃⁻ (1)

H₂O = H⁺ + OH⁻ (2)

Zn²⁺ + 2OH⁻ = Zn(OH)₂ (3)

Zn(OH)₂ = ZnO + H₂O (4)

WO₃ + xH⁺ + xe⁻ = H_xWO₃ (5)

Fig. 1 Synthesis schematic of the ZnO NPs-functionalized WO₃ vertical plate-like arrays.

Because the reaction eqn (5) ³³ preferentially occurred at the active sites of WO₃, near which the pH of the electrolyte shifts towards high value. Zn²⁺ in the interface region of semiconductor–electrolyte tended to form ZnO or Zn(OH)₂, which can be seen in the potential–pH diagram of T. Pauporté's work.³⁴ So ZnO or Zn(OH)₂ can be loaded on the active sites of WO₃. In order to improve the Zn(OH)₂ convert to ZnO and the contact between ZnO and WO₃, the samples were annealed, and ZnO NPs-functionalized WO₃ vertical plate-like arrays were fabricated on FTO.

The crystallographic properties of the as-prepared films were measured using XRD measurements. The corresponding results are shown in Fig. 2. It can be easily seen that the XRD data for the WO₃ and WO₃/ZnO film match well with the monocline WO₃ (PDF#72-0677) and tetragonal structure of SnO₂ (PDF#46-1088). There is no detection of ZnO phases in the patterns of XRD. It can be explained with two reasons. One is that the concentration of ZnO is too low to be detected by XRD. The other is that the zinc species in the sample are dispersing on the surface of WO₃ plates uniformly and sparsely.

Fig. 2 XRD patterns of the WO₃ and WO₃/ZnO films.

The morphology and microstructure of the WO₃ and WO₃/ZnO were investigated by SEM and TEM. Fig. 3 gives representative top-views and cross-views of the samples. In Fig. 3a and b, the low-magnification SEM images show that both samples present overall plate-like structure. The high-magnification SEM images in inset of Fig. 3a and b present almost the same morphology, while the thicknesses of plates are around 240-500 nm. And the cross sectional SEM images (Fig. 3a and b) show the same thickness of film about 1.63 μm. In other words, the nanostructures are well retained after electrodeposition and annealing. More details of morphological and structural features were studied using TEM and shown in Fig. 4. The EDS of WO₃/ZnO shown in Fig. S1† reveals the presence of W, O and Zn, and the diffraction patterns shown in Fig. S2† are indexed to the monocline WO₃. The image in Fig. 4a displays the stagger plate-like structure of the as prepared WO₃/ZnO. It can be observed that a lot of nanoparticles attach the surface of plates. From the high-magnification TEM image in Fig. 4b, the lattice fringe of nanoparticle is about 0.260 nm which corresponds to the interplanar spacing of (200) plane of ZnO.

Fig. 3 SEM images of the surface morphology of (a) WO₃ and (b) WO₃/ZnO; cross-sectional micrographs of (c) WO₃ and (d) WO₃/ZnO; high-magnification SEM images in inset of (a) and (b).

Fig. 4 Transmission electron micrographs and High resolution TEM (HRTEM) image of WO₃/ZnO.

XPS was performed to know the surface chemical composition and the oxidation state of the as-prepared film. The XPS peaks indicate that the film contains W, O, Zn and C. The signal of C originates from the adventitious hydrocarbon in the XPS instrument itself. Fig. 5 shows the high-resolution XPS spectra of W 4f, O 1s, and Zn 2p, respectively. Two peaks at 35.56 and 37.71 eV in the spectrum were assigned to W 4f_7/2 and W 4f_5/2, respectively, which are consistent with the reported values.³³ The asymmetric O 1s peak was fitted by three peaks, centered at 530.2 eV (O–Zn), 530.5 eV (O–W) and 532.6 (surface hydroxyl).³⁵ The binding energies of Zn 2p_1/2 and Zn 2p_3/2 are respectively located at 1044.05 and 1021.05 eV, with a spin energy separation of 23 eV.³⁶ The above results indicate that ZnO nanoparticles are successfully decorated on WO₃ plates, and the molar content of loaded ZnO is determined by XPS to be 2.02 mol%.

Fig. 5 XPS of WO₃/ZnO (a) full spectrum; (b) W 4f; (c) O 1s and (d) Zn 2p.

Fig. S3† shows the UV-Vis absorption spectra of WO₃ and WO₃/ZnO. The UV-Vis spectrum of the WO₃ indicates that it absorbs light with a wavelength less than 465 nm, corresponding to 2.67 eV of band gap energy. There is a slight blue-shift in the absorption edge of WO₃/ZnO, which is 460 nm corresponding to 2.70 eV of band gap energy. At the wavelengths >465 nm, the absorbance of both samples are almost the same. So we believe that the light absorption ability has not been improved in this system.

3.2 Photoelectrochemical measurements of the films

Fig. 6a shows the transient photocurrent responses with several on–off cycles of intermittent light irradiation (>420 nm) at 1.0 V (vs. Ag/AgCl). As can be seen, the photocurrent value rapidly decreases to about zero as soon as the light turns off, and the photocurrent comes back to a constant value when the light is on again. These results are well reproducible. Within the first few seconds of illumination, a slight photocurrent decay was observed for each sample. It may have two reasons. One is that concentrated gradient exists at the interface of photoanode and electrolyte, which is because the photoelectrochemical properties were measured in the 0.2 mol L⁻¹ Na₂SO₄ solution without stirring. The other, which is more important, is that sudden irradiation leads to rapid initial generation of electron–hole pairs and a great photocurrent excitement at the early stage of the illumination. From the beginning of irradiation to the end of measure, the photocurrent of WO₃ decreases 33.3% (from 1.01 to 0.67 mA cm⁻²). The WO₃/ZnO exhibits higher photocurrent than that of the bare WO₃ under visible irradiation, and the photocurrent decreases 22.2% (from 1.22 to 0.95 mA cm⁻²). It means that the composite structure film shows better stability compared to the bare WO₃.

Fig. 6 (a) Photocurrent–time plots with chopping light at 1.0 V (vs. Ag/AgCl), (b) linear sweep voltammograms and (c) photoconversion efficiency of the as-prepared samples.

Fig. 6b shows the liner sweep voltammogram measurements under visible light irradiation for the WO₃ and WO₃/ZnO electrodes at a scan rate of 20 mV s⁻¹. The photocurrent values of both electrodes increase with the upgrade of applied potential of the working electrode. The photocurrent densities of the WO₃ and WO₃/ZnO at 1.2 V (vs. Ag/AgCl) are about 0.88 and 1.23 mA cm⁻², respectively. The photoelectrochemical properties of photoanodes fabricated with WO₃ and WO₃/ZnO plates are also evaluated by the light energy to chemical energy efficiency, which is calculated as an equation.^37,38

ε (%) = j_p{(E⁰_r − |E_a|)}/I_o × 100

In this equation, j_p, E⁰_r, E_a and I_o are the photocurrent density, the standard reversible potential, applied potential of the working electrode and power density of the incident light, respectively. As shown in Fig. 6c, the maximum photoconversion efficiency of WO₃/ZnO (0.475%) is greater than that of WO₃ (0.339%).

Following liner sweep voltammogram measurements, EIS measurements were used to investigate the properties of electron transfer resistance across the semiconductor–electrolyte interfaces under visible irradiation. It is well known that EIS Nyquist plots can be used to characterize the charge transfer resistance and the separation efficiency of the photo-generated electrons and holes. Fig. 7a shows the Nyquist plots of samples at an AC frequency varying from 10 kHz to 0.1 Hz under the bias of 0.7 V (vs. Ag/AgCl). Each of the electrodes presents a semicircle, which can be fitted to an equivalent circuit including a series resistance (R_s), a charge transfer resistance (R_ct) and a constant phase element (CPE), and the equivalent circuit is exhibited in the inset of Fig. 7a. Bode-phase plots of the EIS spectra for the samples are shown in Fig. 7b. The maximum oscillation frequency (f_max) of the impedance semicircle of WO₃/ZnO is less than that of WO₃. Thus, the corresponding lifetime of photoelectrons in WO₃/ZnO increases according to the formula (τ = 1/2πf_max).³⁹ And the values of τ are 6.44 and 8.56 ms for WO₃ and WO₃/ZnO, respectively. It can be ascribed to efficient separation of the light-generated electrons and holes in the WO₃/ZnO. In other words, the electrons transfer from the conduction band of WO₃ to the FTO, while the holes are transported to the valence band of the ZnO. So the recombination of electron–hole pairs are inhibited, and lifetime of photoelectrons are prolonged.

Fig. 7 EIS plots of the WO₃ and WO₃/ZnO: (a) Nyquist plots and (b) Bode plots.

Mott–Schottky is used to determine the flat band potential of each semiconductor material. As shown in the Fig. S4,† the flat band potentials for WO₃ and WO₃/ZnO are −0.168 and −0.198 V (vs. Ag/AgCl), respectively, which are calculated from the x-intercept of the linear portion of the Mott–Schottky plots. It means that the shift in the Fermi level to negative potential is about 0.03 V after modified with ZnO.⁴⁰

In order to identify the main factor of the enhancement of photoelectrochemical properties, intensity-modulated photocurrent spectroscopy (IMPS) was carried out to investigate electron transport. The lamp-house was fitted with a blue light emitting diode (LED) (390 nm) driven by a PP210 (Zahner) frequency response analyser. Fig. 8 shows the complex plane spectra of the IMPS response. The electron transport time (τ_n) can be determined from the frequency at the imaginary maximum (f_im), given by the formula (τ_n = 1/2πf_im).²⁹ The electron transport time calculated for WO₃ and WO₃/ZnO is 2.35 and 2.97 ms, respectively. The slightly longer electron transport time might be caused by the electron transfer from the conduction band of ZnO to FTO via the conduction band of WO₃. However, the content of ZnO is too little to generate much electrons. Considering the light source for other photoelectrochemical measurements is filtered by a 420 nm cutoff filter, the photo-generated electrons produced by ZnO are negligible. The above results confirm that the photo-generated holes of WO₃ transferred to the valence band of ZnO improve the separation of electrons and holes so as to enhance the photoelectrochemical performance of the electrode.

Fig. 8 Intensity-modulated photocurrent spectroscopies of WO₃ and WO₃/ZnO.

The photocurrent responses of samples as a function of wavelength of incident light were measured at a bias of 1.2 V. The calculated incident photon to current conversion efficiency (IPCE) is shown in Fig. 9. The IPCE of WO₃ drops to zero at wavelengths longer than 470 nm, which is consistent with the results of UV-Vis spectra. In the photoresponse region, WO₃/ZnO shows a better IPCE value than WO₃. Both photoanodes display a maximum IPCE value at the wavelength of around 360 nm. In the case of the WO₃/ZnO film, the maximum IPCE value is 47.24%, whereas WO₃ is 28.88%. Because of the similar photoresponse region of WO₃ and WO₃/ZnO, the enhancement of IPCE indicates that the effective inhibition in the recombination of electron–hole pairs by transferring the photo-generated holes is a great choice for improving the photoelectrochemical properties.

Fig. 9 Incident photo to current conversion efficiency of WO₃ and WO₃/ZnO.

A simple mechanism of the conduction and valence band positions for a WO₃/ZnO composite structure is shown in Fig. 10. As can be seen, ZnO has neglectful activity under visible light, and much less photoactive than WO₃. However, when formed into a composite with WO₃, electrons excited into the conduction band of ZnO can transfer to the current collector via the conduction band of the WO₃. More importantly, the valence band edge of the ZnO is at a higher energy than the valence band edge of the WO₃. The holes leaving in the valence band of WO₃ can move into the valence of ZnO, which facilitate the electron/hole separation. Therefore, it can be concluded that the WO₃/ZnO has a better photoelectrochemical performance than WO₃ by promoting electron/hole separation.

Fig. 10 Schematic of conduction and valence band positions for transference of electrons and holes.

4. Conclusion

In summary, ZnO NPs-functionalized WO₃ vertical plate-like arrays film was first synthesized by a combination of hydrothermal and electrodeposition methods, and characterized by various techniques. Experimental results indicate that ZnO NPs-functionalized WO₃ has a higher photocurrent under visible light irradiation. And the photoconversion efficiency is greater than that of bare WO₃. On the basis of EIS, IMPS and IPCE, the high photoelectrochemical performance can be attributed to the effective transfer of the photo-generated holes retarding the recombination of electrons and holes. The combination of electrodeposition and the electrochromism reaction of WO₃ provide an approach to synthesize WO₃ based 1D or 2D composite materials film.

Acknowledgements

This study was supported by the National Nature Science Foundation of China (no. 21171175), the Hunan Provincial Natural Science Foundation of China (no. 13JJ6003), the Open-End Fund for the Valuable and Precision Instruments of Central South University.

References

1. A. Fujishima and K. Honda, Nature, 1972, 238, 37–38.
2. W. Zhou, W. Li, J.-Q. Wang, Y. Qu, Y. Yang, Y. Xie, K. Zhang, L. Wang, H. Fu and D. Zhao, J. Am. Chem. Soc., 2014, 136, 9280–9283.
3. 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.
4. S. Hoang, S. P. Berglund, N. T. Hahn, A. J. Bard and C. B. Mullins, J. Am. Chem. Soc., 2012, 134, 3659–3662.
5. K. Li, B. Chai, T. Peng, J. Mao and L. Zan, ACS Catal., 2013, 3, 170–177.
6. X. Guan and L. Guo, ACS Catal., 2014, 4, 3020–3026.
7. K. Chang, Z. Mei, T. Wang, Q. Kang, S. Ouyang and J. Ye, ACS Nano, 2014, 8, 7078–7087.
8. L. Zhang, D. Jing, L. Guo and X. Yao, ACS Sustainable Chem. Eng., 2014, 2, 1446–1452.
9. P. Zhou, J. Yu and M. Jaroniec, Adv. Mater., 2014, 26, 4920–4935.
10. P. M. Rao, L. Cai, C. Liu, I. S. Cho, C. H. Lee, J. M. Weisse, P. Yang and X. Zheng, Nano Lett., 2014, 14, 1099–1105.
11. J. Y. Zheng, G. Song, C. W. Kim and Y. S. Kang, Nanoscale, 2013, 5, 5279–5282.
12. L. Yin, D. Chen, M. Hu, H. Shi, D. Yang, B. Fan, G. Shao, R. Zhang and G. Shao, J. Mater. Chem. A, 2014, 2, 18867–18874.
13. L. Yin, D. Chen, M. Feng, L. Ge, D. Yang, Z. Song, B. Fan, R. Zhang and G. Shao, RSC Adv., 2015, 5, 328–337.
14. D. Chen, L. Yin, L. Ge, B. Fan, R. Zhang, J. Sun and G. Shao, Sens. Actuators, B, 2013, 185, 445–455.
15. O. Yayapao, T. Thongtem, A. Phuruangrat and S. Thongtem, J. Alloys Compd., 2011, 509, 2294–2299.
16. R. Solarska, B. D. Alexander, A. Braun, R. Jurczakowski, G. Fortunato, M. Stiefel, T. Graule and J. Augustynski, Electrochim. Acta, 2010, 55, 7780–7787.
17. V. B. Patil, P. V. Adhyapak, S. S. Suryavanshi and I. S. Mulla, J. Alloys Compd., 2014, 590, 283–288.
18. P. M. Rao, I. S. Cho and X. Zheng, Proc. Combust. Inst., 2013, 34, 2187–2195.
19. J. Su, X. Feng, J. D. Sloppy, L. Guo and C. A. Grimes, Nano Lett., 2010, 11, 203–208.
20. J. Yang, W. Li, J. Li, D. Sun and Q. Chen, J. Mater. Chem., 2012, 22, 17744–17752.
21. Y. Sun, C. J. Murphy, K. R. Reyes-Gil, E. A. Reyes-Garcia, J. M. Thornton, N. A. Morris and D. Raftery, Int. J. Hydrogen Energy, 2009, 34, 8476–8484.
22. W. Mu, X. Xie, X. Li, R. Zhang, Q. Yu, K. Lv, H. Wei and Y. Jian, RSC Adv., 2014, 4, 36064–36070.
23. M.-T. Chang, L.-J. Chou, Y.-L. Chueh, Y.-C. Lee, C.-H. Hsieh, C.-D. Chen, Y.-W. Lan and L.-J. Chen, Small, 2007, 3, 658–664.
24. S. Sun, W. Wang, S. Zeng, M. Shang and L. Zhang, J. Hazard. Mater., 2010, 178, 427–433.
25. A. Tanaka, K. Hashimoto and H. Kominami, J. Am. Chem. Soc., 2013, 136, 586–589.
26. D. Chen, T. Li, Q. Chen, J. Gao, B. Fan, J. Li, X. Li, R. Zhang, J. Sun and L. Gao, Nanoscale, 2012, 4, 5431–5439.
27. X. Chang, L. Dong, Y. Yin and S. Sun, RSC Adv., 2013, 3, 15005–15013.
28. Y. Liu, W. Li, J. Li, Y. Yang and Q. Chen, RSC Adv., 2014, 4, 3219–3225.
29. J. Su, L. Guo, N. Bao and C. A. Grimes, Nano Lett., 2011, 11, 1928–1933.
30. K. C. Leonard, K. M. Nam, H. C. Lee, S. H. Kang, H. S. Park and A. J. Bard, J. Phys. Chem. C, 2013, 117, 15901–15910.
31. J. Zhu, W. Li, J. Li, Y. Li, H. Hu and Y. Yang, Electrochim. Acta, 2013, 112, 191–198.
32. Y. Zang, L. Li, Y. Zuo, H. Lin, G. Li and X. Guan, RSC Adv., 2013, 3, 13646–13650.
33. S. Cong, Y. Tian, Q. Li, Z. Zhao and F. Geng, Adv. Mater., 2014, 26, 4260–4267.
34. A. Goux, T. Pauporté, J. Chivot and D. Lincot, Electrochim. Acta, 2005, 50, 2239–2248.
35. P. T. Hsieh, Y. C. Chen, K. S. Kao and C. M. Wang, Appl. Phys. A, 2008, 90, 317–321.
36. M. Shao, F. Ning, M. Wei, D. G. Evans and X. Duan, Adv. Funct. Mater., 2013, 24, 580–586.
37. P.-T. Hsiao, L.-C. Chen, T.-L. Li and H. Teng, J. Mater. Chem., 2011, 21, 19402–19409.
38. A. Devadoss, P. Sudhagar, C. Ravidhas, R. Hishinuma, C. Terashima, K. Nakata, T. Kondo, I. Shitanda, M. Yuasa and A. Fujishima, Phys. Chem. Chem. Phys., 2014, 16, 21237–21242.
39. K. Chen, X. Feng, R. Hu, Y. Li, K. Xie, Y. Li and H. Gu, J. Alloys Compd., 2013, 554, 72–79.
40. J. Zhang, J. H. Bang, C. Tang and P. V. Kamat, ACS Nano, 2009, 4, 387–395.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra03918e

‡ Contributed equally to this work.

---