WO₃ nanorod photocatalysts decorated with few-layer g-C₃N₄ nanosheets: controllable synthesis and photocatalytic mechanism research

Abstract

Novel WO₃ nanorod (WO₃ NRs) photocatalysts decorated with few-layer g-C₃N₄ nanosheets (flg-C₃N₄) have been synthesized by a typical hydrothermal process. The structure, morphology, optical and electronic properties were researched by comprehensive characterization methods. The photocatalytic performance of flg-C₃N₄/WO₃ NRs nanocomposites was assessed by degrading Rhodamine B (RhB) dye under visible light irradiation. When the loading amount of the flg-C₃N₄ is 20 wt%, the flg-C₃N₄/WO₃ NRs nanocomposites exhibit the highest activity and the Chemical Oxygen Demand (COD) values decreased by about 64%. The improved light harvesting ability and higher separation efficiency of photoinduced electron–hole pairs by the decoration of flg-C₃N₄ lead to the boosting photocatalytic performance. In addition, through ESR analysis and a trapping experiment, the photocatalytic mechanism was also researched.

---

1. Introduction

Energy and environmental problems have prevented the sustainable development of social-economy; therefore, numerous researchers are continuously seeking advanced technologies to solve these issues.¹ Photocatalysis, one of the most promising technologies, can directly harness solar energy, for the degradation of environmental pollutants,² hydrogen evolution,³ and CO₂ photocatalytic reduction.⁴ To date, several photocatalytic nanomaterials, such as Fe₂O₃ and ZnO, have been researched and exhibited promising photocatalytic activity.^5,6 Although these materials are widely applied for photocatalytic applications, some of them are only active under ultraviolet irradiation,⁷ whereas solar power is only composed of 4% ultraviolet light.⁸ Therefore, there are more prospects for designing visible-light-driven catalysts to meet the need of industrial applications.

Considerable efforts have shown that WO₃, as one of traditional n-type semiconductors, is a visible light responsive photocatalyst with a relatively narrow band gap energy (2.4–2.8 eV) and suitable valence band (VB) potential, which is similar to that of TiO₂.⁹ Nevertheless, since the conduction band (CB) position of WO₃ is lower than the oxygen reduction process (O₂ + e⁻ → ˙O²⁻ (aq), −0.33 V vs. NHE),¹⁰ the transferring electrons in WO₃ CB cannot be utilized sufficiently, making it easy for the accumulative photogenerated electrons to recombine with holes, which results in the low photocatalytic performance of the WO₃ photocatalyst.^11,12 Therefore, several prevailing strategies have been applied to boost the photocatalytic performance of WO₃. Firstly, preparing materials with specific structures such as nanosheets,¹³ nanowires¹⁴ and nanorods.¹⁵ Secondly, introducing oxygen vacancies in WO₃ material.^16,17 Thirdly, coupling with different materials to build hybrid nanomaterials such as CaFe₂O₄/WO₃,¹⁸ Pd/WO₃,¹⁹ Cu/WO₃,²⁰ and WO₃/g-C₃N₄.²¹ Thus, designing WO₃ with a specific morphology or fabricating nanocomposites could significantly improve the photocatalytic activity of WO₃.

In the fields of photocatalysis, some nanomaterials with a graphene-like structure have been researched.^22,23 For example, SnS₂ nanosheets have been used as co-catalysts to improve the removal efficiency of pollutants.²⁴ Liu et al. rooted few-layer MoS₂ nanosheets into TiO₂ nanofibers for highly efficient photocatalytic hydrogen evolution.²⁵ Likewise, J. Choi et al. modified BN nanosheets on the surface of AgI to improve photocatalytic activity.²⁶ In 2009, Wang et al. showed the photocatalytic water splitting via graphitic-like carbon nitride, and such metal-free material has met extensive concern in the photocatalysis fields.^27,28 The bulk g-C₃N₄ nanomaterial possesses a layered structure and the π-conjugated s-triazine unit is formed by the sp² hybridization of the carbon and nitrogen, resulting in high structural stability of g-C₃N₄.^27,29 As a two-dimensional graphene-like nanostructure, the few-layer g-C₃N₄ nanosheets (flg-C₃N₄) show more superior physical and chemical properties such as the larger surface area, exposure of numerous active sites and optimized light harvesting, which may be favorable for promoting photocatalytic performance. Therefore, flg-C₃N₄ has been applied to enhance the photocatalytic performance of different semiconductors in recent research studies such as Ag₂CO₃/flg-C₃N₄,³⁰ Bi₂WO₆/flg-C₃N₄ (ref. 31) and Au/flg-C₃N₄.³² However, to the best of our knowledge, using flg-C₃N₄ to decorate WO₃ nanorods (WO₃ NRs) to form composite photocatalysts has never been reported.

Herein, we demonstrate a controllable hydrothermal method to prepare flg-C₃N₄/WO₃ NRs nanocomposites with excellent catalytic performance. The stability of the photocatalyst was evaluated by four consecutive catalytic runs. The photocatalytic mechanism was also studied in detail.

2. Experimental

2.1 Synthesis of WO₃ NRs

WO₃ NRs were prepared by a facile hydrothermal method.¹⁰ Firstly, 2.112 g Na₂WO₄·2H₂O and 1.87 g NaCl were mixed fully in 50 mL water. The solution pH was adjusted by 3.0 mol L⁻¹ HCl and stirred continuously for 4 h. Afterwards, the solution was placed into a Teflon-lined autoclave and heated at 180 °C for 24 h. Thereafter, the obtained products were washed and separated. The products were dried at 70 °C.

2.2 Synthesis of the flg-C₃N₄

Bulk g-C₃N₄ was synthesized by calcining melamine molecules at high temperature. Melamine was heated at 550 °C with 2 °C min⁻¹ for 4 h. Flg-C₃N₄ was synthesized by calcining as-prepared bulk g-C₃N₄ at 550 °C for 100 min with 2 °C min⁻¹.

2.3 Synthesis of the flg-C₃N₄/WO₃ NRs

0.05 g WO₃ NRs was placed in 15 mL of DI water, which includes a certain amount of as-prepared flg-C₃N₄, and then the above mixture was ultrasonicated for 20 min and further stirred for 30 min. Then, the mixed solution was added in a 25 mL Teflon-lined autoclave and heated at 150 °C for 4 h. Thereafter, the final WO₃ NRs with various flg-C₃N₄ contents were obtained and dried at 70 °C. A schematic is shown in Scheme 1.

Scheme 1 Schematic for the preparation of the WO₃ NRs, the flg-C₃N₄ and the flg-C₃N₄/WO₃ NRs nanocomposites.

3. Results and discussion

3.1 XRD analysis

The XRD spectra of WO₃ NRs, flg-C₃N₄ and flg-C₃N₄/WO₃ NRs are displayed in Fig. 1. The flg-C₃N₄/WO₃ NRs nanocomposites with different contents of flg-C₃N₄ exhibited a similar pattern compared with WO₃ NRs. The XRD patterns of the nanocomposites matched perfectly with the JCPDS card of WO₃ (no. 33-1387). The diffraction peaks appearing at 14.0°, 22.8°, 24.2°, 26.9°, 28.3°, 33.5°, 36.6°, 49.7° and 55.7° are indexed to the (100), (001), (110), (101), (200), (111), (201), (220), and (221) crystal planes of WO₃,²¹ respectively. After introducing flg-C₃N₄, the intensity of the diffraction peak for the (100) crystal planes almost disappeared and the peak for the (002) crystal planes dramatically decreased compared with the reports of previous researchers,^22,33 manifesting that the flg-C₃N₄ structure might be composed of few-layer s-triazine unit layers.³⁴ In addition, for the nanocomposites, the diffraction peak of flg-C₃N₄ for the (002) crystal planes might overlap with the peak of WO₃ NRs for (101), and the peak for (100) could not be discovered, which might be attributed to the low contents of flg-C₃N₄.

Fig. 1 XRD spectra of the as-prepared WO₃ NRs, flg-C₃N₄/WO₃ NRs and flg-C₃N₄.

3.2 XPS analysis

The surface elements and chemical states of WO₃ NRs and flg-C₃N₄/WO₃ NRs nanocomposites were further investigated by the XPS method. Fig. 2A shows the XPS survey spectrum of WO₃ NRs and 20 wt% flg-C₃N₄/WO₃ NRs nanocomposites, revealing the presence of C, N, O, and W in the nanocomposites. Fig. 2B–E shows the high resolution spectra of W 4f, O 1s, C 1s, and N 1s. As depicted in Fig. 2B, the peaks were discovered at 35.6 eV and 37.7 eV, respectively, which are consistent with the typical binding energies of W 4f_7/2 and W 4f_5/2 of W⁶⁺.¹⁰ The O 1s peak (Fig. 2C) at 530.4 eV was attributed to the O²⁻ in the WO₃ NRs. The characteristic peaks of C 1s (Fig. 2D) were observed at binding energies 284.7 eV and 288.6 eV. The peak at 284.4 eV could be ascribed to carbon bonds and the peak observed at 288.6 eV was assigned to defect-containing sp²-bonded carbon (N–C=N).³⁵ The peak of N 1s (Fig. 2E) was observed at 399.0 eV. All these XPS results are further confirmed by the FT-IR analysis.

Fig. 2 (A) XPS survey spectra of WO₃ NRs, 20 wt% flg-C₃N₄/WO₃ NRs. High resolution spectra of (B) W 4f and (C) O 1s peak of the WO₃ NRs and 20 wt% flg-C₃N₄/WO₃ NRs, (D) C 1s and (E) N 1s peak of 20 wt% flg-C₃N₄/WO₃.

3.3 FT-IR analysis

FT-IR spectra are a robust tool to investigate chemical functional groups of the nanomaterials. Fig. 3 shows the FT-IR spectra of WO₃ NRs, flg-C₃N₄ and flg-C₃N₄/WO₃ NRs nanocomposites with different contents. The broad stretching vibrations observed at 3100–3500 cm⁻¹ were attributed to O–H and N–H bands, which was relative to surface absorbed water molecules and uncondensed amino groups.³⁶ Moreover, a series of absorption peaks observed between 1200 cm⁻¹ and 1650 cm⁻¹ were due to characteristic absorption peaks of the CN heterocycles.³⁷ The peak appearing at 810 cm⁻¹ was attributed to the stretching peaks of s-triazine units.³⁸ Furthermore, with increased flg-C₃N₄ contents in the nanocomposites, the absorption peak positions at 828 cm⁻¹ ascribed to the vibrations of O–W–O bands slightly shifted,³⁹ which demonstrated the existence of the strong interaction between WO₃ NRs and s-triazine units of flg-C₃N₄ as well. Therefore, the FT-IR results provided further evidence that WO₃ NRs were decorated with flg-C₃N₄ successfully and an interaction existed between WO₃ NRs and flg-C₃N₄.

Fig. 3 FT-IR of the WO₃ NRs and flg-C₃N₄/WO₃ NRs nanocomposites and flg-C₃N₄.

3.4 SEM and TEM analysis

The morphology and superficial microstructure of flg-C₃N₄/WO₃ NRs nanocomposites were studied by SEM and TEM. Fig. 4A and B show the SEM images of WO₃ NRs and flg-C₃N₄/WO₃ NRs (20 wt%). As revealed in Fig. 4A, it could be clearly observed that rod-like WO₃ was successfully synthesized. As revealed in Fig. 4B, WO₃ NRs were decorated by flg-C₃N₄ on the surface, which revealed that flg-C₃N₄ and WO₃ NRs were successfully coupled. In the EDS pattern, C, N, W and O could be observed (Fig. 4C), which further proved that flg-C₃N₄/WO₃ NRs nanocomposites were synthesized without other impurities. Fig. 4D–G shows the TEM images of WO₃ NRs and flg-C₃N₄/WO₃ NRs (20 wt%). As shown in Fig. 4D and E, WO₃ NRs displayed the inerratic rod-like morphology with the crystal width of around 20 nm. As shown in Fig. 4F and G, the WO₃ NRs are well dispersed on the surface of the flg-C₃N₄ and intimately contacted between WO₃ NRs and flg-C₃N₄, which is favorable to enhance the transfer of photogenerated electrons.³⁴ The thickness of flg-C₃N₄ was about 5.10 nm (Fig. S2†), and the representative thickness of the g-C₃N₄ monolayers as suggested by our previous report⁴⁰ was ∼0.64 nm. Therefore, the as-synthesized flg-C₃N₄ consisted of ∼8 monolayers. Fig. 4F and G also showed the wrinkled morphology and relatively large surface of flg-C₃N₄, which shows that flg-C₃N₄ could increase the specific surface area of WO₃ NRs and provide more active reaction sites.

Fig. 4 (A) SEM image of WO₃ NRs; (B) SEM image of WO₃ NRs, 20 wt% flg-C₃N₄/WO₃ NRs; (C) EDS of the flg-C₃N₄/WO₃ NRs; (D and E) TEM images of WO₃ NRs; (F and G) TEM images of 20 wt% flg-C₃N₄/WO₃ NRs.

3.5 DRS analysis

The UV-Vis diffuse reflectance spectra (DRS) of WO₃ NRs and flg-C₃N₄/WO₃ NRs nanocomposites are displayed in Fig. 5A. By comparing with the WO₃ NRs, the optical absorption of flg-C₃N₄/WO₃ NRs in the range of 450–600 nm was apparently enhanced and the absorption edge showed a slight red-shift. The results indicated that the photocatalysts could increase the absorption of visible light, which would be beneficial for their photocatalytic performance. The energy gap of WO₃ NRs and flg-C₃N₄ was estimated via the classical Tauc approach, (αhν)² = A(hν − E_g)ⁿ, as revealed in Fig. 5B, the energy gap (E_g) of WO₃ NRs was estimated to be 2.70 eV and E_g of flg-C₃N₄ was 2.80 eV.

Fig. 5 (A) The UV-Vis absorption spectra of the as prepared WO₃ NRs and the flg-C₃N₄/WO₃ NRs nanocomposites with different contents of flg-C₃N₄ (B) estimated band gap curves for the as prepared WO₃ NRs and the flg-C₃N₄.

3.6 Nitrogen adsorption analysis

The Brunauer–Emmett–Teller (BET) surface areas of WO₃ NRs and 20 wt% flg-C₃N₄/WO₃ NRs have also been analyzed. From Fig. 6, the specific surface area of the 20 wt% flg-C₃N₄/WO₃ NRs was 40.6 m² g⁻¹, which was about 1.7 times as high as that of the WO₃ NRs (24.3 m² g⁻¹), showing that the decoration of the flg-C₃N₄ could further increase the specific surface areas of the WO₃ NRs. The higher surface area values implied that there are more adsorption and active sites for the photocatalytic reaction.

Fig. 6 Nitrogen absorption–desorption isotherm of WO₃ NRs and the 20 wt% flg-C₃N₄/WO₃ NRs nanocomposites.

3.7 Electrochemistry analysis

It can be noted that the separation efficiency of electrons and holes plays a critical role for photocatalysis. The separation efficiency of photoinduced-charges was assessed by photocurrent and electrochemical impedance spectroscopy (EIS). From Fig. 7A, in view of photocurrent response, the introduction of flg-C₃N₄ enhanced the photocurrent density remarkably, indicating boosting separation efficiencies of electrons and holes and longer lifetimes of photogenerated charge carriers.⁴¹ In addition, as shown in Fig. 7B, the 20% flg-C₃N₄/WO₃ NRs nanocomposites exhibited a smaller diameter of the Nyquist circle than WO₃ NRs, suggesting that 20 wt% flg-C₃N₄/WO₃ NRs nanocomposites had a lower charge transfer resistance and more effective electron–hole pair separation efficiency.⁴² Thus, these results further supported that decorating with flg-C₃N₄ facilitated the charge separation.

Fig. 7 (A) Transient photocurrent response for the as prepared WO₃ NRs and the flg-C₃N₄/WO₃ NRs nanocomposites with different contents of flg-C₃N₄; (B) electrochemical impedance spectroscopy (EIS) Nyquist plots of the WO₃ NRs and the 20 wt% flg-C₃N₄/WO₃ NRs nanocomposites.

3.8 Photocatalytic performance

As observed in Fig. 8A, the photoactivities of WO₃ NRs and flg-C₃N₄/WO₃ NRs nanocomposites were investigated by degrading RhB. The results indicated that RhB was not degraded obviously without catalysts, indicating that RhB self-degradation could be ignored. The photocatalytic activity of WO₃ NRs was improved gradually with the increasing proportion of flg-C₃N₄. When the contents of flg-C₃N₄ reached 20%, the highest photocatalytic performance was obtained, and 96% of RhB was degraded after 120 min irradiation. Nevertheless, further increasing the contents of flg-C₃N₄ in the nanocomposites up to 30% caused decreased degradation efficiency. This result could be attributed to that excessive flg-C₃N₄ might cover part of the active site of WO₃ NRs. Based on the abovementioned results, there might be some synergistic interaction between WO₃ NRs and flg-C₃N₄, which plays a vital role in enhancing the photocatalytic performance. From Fig. 8B and C, the intensity of the characteristic absorption peak dropped as the time increased. In addition, the maximum absorption peak was left-shifted, which was related to the step-by-step de-ethylation process.⁴³ The COD values for the degradation of RhB (Table 1) show that the COD values decreased by about 64% by the 20 wt% flg-C₃N₄/WO₃ NRs. In addition, the corresponding total organic carbon (TOC) values decreased by about 46.2% (Fig. S3†). The rate constant of the 20 wt% flg-C₃N₄/WO₃ NRs nanocomposites was about 7.5 times as high as that of the WO₃ NRs (the kinetics were fitted by a first-order model as shown in Table S1†).

Fig. 8 (A) Photodegradation of RhB by the as-prepared WO₃ NRs and the flg-C₃N₄/WO₃ NRs nanocomposites with different contents of flg-C₃N₄; (B) time-dependent UV-Vis absorption spectra of the as-prepared WO₃ NRs; (C) time-dependent UV-Vis absorption spectra of the 20 wt% flg-C₃N₄/WO₃ NRs.

Table 1 The changes of COD values in the degradation processes for RhB under visible light irradiation

Time (min)	0	60	120
COD values by WO3 NRs (mg mL^−1)	213.6	169.7	151.5
COD values by 20 wt% flg-C3N4/WO3 NRs (mg mL^−1)	150.0	75.7	54.5

---

3.9 Stability evaluation

To test the practical reusability of the flg-C₃N₄/WO₃ NRs nanocomposites, the circulating tests were investigated under the same conditions. As revealed in Fig. 9A, the photocatalytic activity of 20 wt% flg-C₃N₄/WO₃ NRs nanocomposites exhibited no obvious distinction after four recycles. Notably, from Fig. 9B, no distinction was found between the XRD results before and after four consecutive tests, suggesting that the nanocomposites were stable during the degradation process of the pollutants.

Fig. 9 (A) Cycling runs of 20 wt% flg-C₃N₄/WO₃ NRs nanocomposites; (B) XRD patterns of 20 wt% flg-C₃N₄/WO₃ NRs nanocomposites before and after the photocatalytic reactions.

3.10 Proposed photocatalytic mechanism

According to previous reports, the CB and VB positions of WO₃ NRs are +0.74 V and +3.44 V, respectively.⁴⁴ The CB and VB positions of flg-C₃N₄ are −0.97 eV and 1.83 eV, respectively.²² For the general electron–hole separation process, the electrons in the CB of flg-C₃N₄ will transfer to the WO₃ CB, and holes in the VB of WO₃ NRs will transfer to the flg-C₃N₄ VB. These accumulated electrons lying in WO₃ CB cannot be utilized sufficiently via the oxygen reduction process (O₂ + e⁻ → ˙O₂⁻ (aq), −0.33 V vs. NHE). Likewise, the holes lying in the flg-C₃N₄ VB cannot have enough potential to effectively oxidize OH⁻ (OH⁻ + h⁺ → ˙OH (aq), −2.4 V vs. NHE). However, according to ESR analysis, the characteristic peaks of the DMPO- ˙O₂⁻ and ˙OH radicals could be observed (Fig. 10A and B), which showed that the ˙O₂⁻ and ˙OH radical species were produced under a visible light irradiation process. Therefore, the transfer process of photoinduced charge carriers is not in accordance with the traditional model in this system. Namely, the separation of charge carriers may conform to the Z-scheme.⁴⁴ As shown in Fig. 11, the photoexcited electrons in the WO₃ CB and photoexcited holes in the flg-C₃N₄ VB combine rapidly. Moreover, not only the molecular oxygen is reduced into ˙O₂⁻ by the electrons in the flg-C₃N₄ CB, which has more negative potential, but also the water molecule is oxidized into ˙OH radicals by the holes in the WO₃ VB, which has more positive potential. Therefore, a typical Z-scheme is favorable for the flg-C₃N₄/WO₃ NRs photocatalyst. In addition, the effect of photogenerated holes was investigated by a trapping experiment (as shown in Fig. 10C). After triethanolamine was added, the photocatalytic activity significantly decreased, which showed that the holes might play an important role.

Fig. 10 ESR spectra of (A) DMPO-˙O₂⁻ and (B) DMPO-˙OH adducts in 20 wt% flg-C₃N₄/WO₃ NRs nanocomposites aqueous dispersion systems before and after visible light irradiation; (C) comparison of the photocatalytic activities of 20 wt% flg-C₃N₄/WO₃ NRs nanocomposites for the degradation of RhB with or without adding triethanolamine.

Fig. 11 Schematic of the proposed mechanism for the degradation of pollutants.

Thus, based on the abovementioned results, the photoexcited holes, ˙O₂⁻ and ˙OH radicals act as the major reactive species during the photodegradation process. The transport process of photogenerated electrons and holes is consistent with the Z-scheme in this system.

4. Conclusions

The novel flg-C₃N₄/WO₃ NRs nanocomposites were prepared via a facile hydrothermal process. The 20 wt% flg-C₃N₄/WO₃ NRs nanocomposites showed the highest degradation efficiency for RhB. The improved photocatalytic activity might be attributed to more active sites, the formation of a heterojunction and the high separation of electrons and holes. In addition, the flg-C₃N₄/WO₃ NRs nanocomposites were typical Z-scheme photocatalysts, and a possible mechanism for the improved photocatalytic performance was also proposed. Furthermore, from the perspectives of the decreased COD values and the stability of nanocomposites, this work is also meaningful in designing and fabricating the flg-C₃N₄-based nanocomposites.

Acknowledgements

The authors genuinely appreciate the financial support for this study by the National Nature Science Foundation of China (21476097) and the Six talent peaks project in Jiangsu Province (2014-JNHB-014).

Notes and references

1. S. Cao, J. Low, J. Yu and M. Jaroniec, Adv. Mater., 2015, 27, 2150–2176.
2. M. Guan, C. Xiao, J. Zhang, S. Fan, R. An, Q. Cheng, J. Xie, M. Zhou, B. Ye and Y. Xie, J. Am. Chem. Soc., 2013, 135, 10411–10417.
3. J. Tian, R. Ning, Q. Liu, A. Asiri, A. Al-Youbi and X. Sun, ACS Appl. Mater. Interfaces, 2014, 6, 1011–1017.
4. J. Mao, T. Y. Peng, X. H. Zhang, K. Li, L. Q. Ye and L. Zan, Catal. Sci. Technol., 2013, 3, 1253–1260.
5. S. Han, L. Hu, Z. Liang, S. Wageh, A. Al-Ghamdi, Y. Chen and X. Fang, Adv. Funct. Mater., 2014, 24, 5719–5727.
6. R. Pawar and C. Lee, Appl. Catal., B, 2014, 144, 57–65.
7. D. Chen, T. Li, Q. Chen, J. Gao, B. Fan, J. Li, X. Li and R. Zhang, Nanoscale, 2012, 4, 5431–5439.
8. J. Xu and X. Cao, Chem. Eng. J., 2015, 260, 642–648.
9. M. Miyauchi, M. Shibuya, Z. Zhao and Z. Liu, J. Phys. Chem. C, 2009, 113, 10642.
10. B. Weng, J. Wu, N. Zhang and Y. Xu, Langmuir, 2014, 30, 5574–5584.
11. P. Wang, Y. Bai, P. Luo and J. Liu, Catal. Commun., 2013, 38, 82–85.
12. Y. Xie, G. Liu, L. Yin and H. Cheng, J. Mater. Chem., 2012, 22, 6746–6751.
13. Z. Sun, T. Liao, Y. Dou, S. Hwang, M. Park, L. Jiang, J. Kim and S. Dou, Nat. Commun., 2014, 5, 3813.
14. J. Liu, O. Margeat, W. Dachraoui, X. Liu, M. Fahlman and J. Ackermann, Adv. Funct. Mater., 2014, 24, 6029–6037.
15. F. Zheng, H. Lu, M. Guo, M. Zhang and Q. Zhen, J. Mater. Chem. C, 2015, 3, 7612–7620.
16. M. Park, J. Seo, H. Song and K. Nam, J. Phys. Chem. C, 2016, 120, 9192–9199.
17. G. Liu, J. Han, X. Zhou, L. Huang, F. Zhang, X. Wang, C. Ding, X. Zheng, H. Han and C. Li, J. Catal., 2013, 307, 148–152.
18. Z. Liu, Z. Zhao and M. Miyauchi, J. Phys. Chem. C, 2009, 113, 17132–17137.
19. Y. Liu, Y. Ohko, R. Zhang, Y. Yang and Z. Zhang, J. Hazard. Mater., 2010, 184, 386–391.
20. Y. Nosaka, S. Takahashi, H. Sakamoto and A. Nosaka, J. Phys. Chem. C, 2011, 115, 21283–21290.
21. J. Zhao, Z. Jia, X. Shen, H. Zhou and L. Ma, Ceram. Int., 2015, 41, 5600–5606.
22. Y. Ide, F. Liu, J. Zhang, N. Kawamoto, K. Komaguchi, Y. Bando and D. Golberg, J. Mater. Chem. A, 2014, 2, 4150–4156.
23. S. Yang, Y. Gong, J. Zhang, L. Zhan, L. Ma, Z. Fang, R. Vajtai, X. Wang and P. Ajayan, Adv. Mater., 2013, 25, 2452–2456.
24. Z. Zhang, J. Huang, M. Zhang, Q. Yuan and B. Dong, Appl. Catal., B, 2015, 163, 298–305.
25. C. Liu, L. Wang, Y. Tang, S. Luo, Y. Liu, S. Zhang, Y. Zeng and Y. Xu, Appl. Catal., B, 2015, 164, 1–9.
26. J. Choi, D. Reddy and T. Kim, Ceram. Int., 2015, 41, 13793–13803.
27. X. Wang, K. Maeda, A. Thomas, K. Takanabe, G. Xin, J. Carlsson, K. Domen and M. Antonietti, Nat. Mater., 2009, 8, 76–80.
28. F. Dong, Z. Zhao, T. Xiong, Z. Ni, W. Zhang, Y. Sun and W. Ho, ACS Appl. Mater. Interfaces, 2013, 5, 11392–11401.
29. M. Groenewolt and M. Antonietti, Adv. Mater., 2005, 17, 1789–1792.
30. Y. Li, L. Fang, R. Jin, Y. Yang, X. Fang, Y. Xing and S. Song, Nanoscale, 2015, 7, 758–764.
31. W. Chen, T. Liu, T. Huang, X. Liu, J. Zhu, G. Duan and X. Yang, Appl. Surf. Sci., 2015, 355, 379–387.
32. N. Cheng, J. Tian, Q. Liu, C. Ge, A. Qusti, A. Asiri, A. Al-Youbi and X. Sun, ACS Appl. Mater. Interfaces, 2013, 5, 6815–6819.
33. H. Xu, J. Yan, Y. G. Xu, Y. Song, H. Li, J. Xia, C. Huang and H. Wan, Appl. Catal., B, 2013, 129, 182–193.
34. H. Xu, J. Yan, X. She, L. Xu, J. Xia, Y. Xu, Y. Song, L. Huang and H. Li, Nanoscale, 2014, 6, 1406–1415.
35. Y. Lv, Y. Zhu and Y. Zhu, J. Phys. Chem. C, 2013, 117, 18520–18528.
36. X. Bai, L. Wang, R. Zong and Y. Zhu, J. Phys. Chem. C, 2013, 117, 9952–9961.
37. J. Holst and E. Gillan, J. Am. Chem. Soc., 2008, 130, 7373–7379.
38. X. She, L. Liu, H. Ji, Z. Mo, Y. Li, L. Huang, D. Du, H. Xu and H. Li, Appl. Catal., B, 2016, 22, 144–153.
39. H. Xu, L. Liu, Y. Song, L. Huang, Y. Li, Z. Chen, Q. Zhang and H. Li, J. Alloys Compd., 2016, 660, 48–54.
40. X. She, J. Wu, J. Zhong, H. Xu, Y. Yang, R. Vajtai, J. Lou, Y. Liu, D. Du, H. Li and P. Ajayan, Nano Energy, 2016, 27, 138–146.
41. S. Chen, Y. Hu, S. Meng and X. Fu, Appl. Catal., B, 2014, 150, 564–573.
42. M. Zhang, X. J. Bai, D. Liu, J. Wang and Y. Zhu, Appl. Catal., B, 2015, 164, 77–81.
43. J. Di, J. Xia, S. Yin, H. Xu, L. Xu, Y. Xu, M. He and H. Li, J. Mater. Chem. A, 2014, 2, 5340–5351.
44. P. Zhou, J. Yu and M. Jaroniec, Adv. Mater., 2014, 26, 4920–4935.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra12861k

---