Ultra-thin C₃N₄ nanosheets for rapid charge transfer in the core–shell heterojunction of α-sulfur@C₃N₄ for superior metal-free photocatalysis under visible light

Abstract

The core–shell heterojunctions of ultra-thin C₃N₄ nanosheet enwrapping spherical α-S composites (α-S@C₃N₄) were fabricated via a self-assembly process by electrostatic force to realize enhanced photocatalytic ability under visible light. The photocatalytic ability can be adjusted by tuning the amount of the ultra-thin C₃N₄ nanosheet. The α-S@C₃N₄ composite with 35% composition of C₃N₄ nanosheet has the highest photocatalytic ability. The degradation rate of Rhodamine B (RhB) with α-S@C₃N₄ (35% C₃N₄) is 6.72 times faster compared to α-S as photocatalyst. This increase could be attributed to the efficient photogenerated holes and electrons separation by the heterojunction, which has excellent charge transfer ability arising from the ultra-thin C₃N₄ nanosheet. The stability of the α-S is also largely improved by the heterojunction construction.

---

1. Introduction

Recently, more attention has been paid to semiconductor photocatalysis in the field of pollution elimination as a “green” technology that can completely mineralize most types of pollutants under the illumination of sunlight and moderate ambient conditions.^1–3 Development of an efficient visible-light-driven semiconductor material is an important need for practical application of this technology. With unique advantages of low-cost, abundance, and easy access, metal-free elemental photocatalysts are emerging and have great potential for application in industry.^4–6 Most recently, Liu et al. demonstrated that α-sulfur (α-S) can be a candidate for visible-light-driven photocatalysis because of its capability of degrading Rhodamine B (RhB) under visible light illumination.⁷ Since then, to enhance the photocatalytic ability of α-S, some studies have been carried out. Yu and co-workers⁸ constructed a heterojunction of graphene and g-C₃N₄ enwrapped α-S, which exhibited enhanced photocatalytic ability for bacterial inactivation. Peng et al.⁹ reported the fabrication of a sulfur/graphene composite, in which the incorporation of graphene enhanced its photocatalytic activity for the degradation of methyl orange due to the increased hydrophilicity and adsorption capacity. However, similar to the other photocatalysts, α-S still has low quantum efficiency attributed to the limited light absorption ability, the high recombination rate of photogenerated holes and electrons, and the sluggish surface reaction.

It is well known that photocatalytic activity is closely related to the size, morphology, structure, and surface area of the photocatalyst.¹⁰ Decrease in the size can enhance the photocatalytic activity due to increasing surface area exposed to the light and reactants. As a result, the light absorption and the surface reaction can be improved. Several surfactants have been used as the soft template in the fabrication of photocatalysts for the control of size and morphology.¹¹ One example is poly(vinyl pyrrolidone) (PVP), which has been proved to be an efficient template to either decrease the size or configure the shape of the sphere.¹²

Fabrication of the heterojunction is another strategy for improving photocatalytic activity potentially attributed to the highly effective charge separation.¹³ When designing a heterojunction, the critical factor that should be considered is the charge transporting ability of the loading material. One good example is graphene, which is widely used for the construction of the heterojunction due to its extremely high electron mobility.¹⁴ Recently, it was reported that bulk g-C₃N₄ can degrade organic pollutants under visible light irradiation.^15–17 It is a layered material and can be exfoliated into an ultra-thin nanosheet with a few layers and even into a single atomic layer nanosheet.^18,19 The exfoliated C₃N₄ nanosheet has a big surface area with very thin layer thickness and excellent conductivity with less resistance between the layers, which make it an excellent material for constructing heterojunctions with other semiconductors. Furthermore, the interface area of the heterojunction is a key parameter that can affect the photocatalytic ability.²⁰ The large contact area can promote charge transfer across the interface, thus increasing the separation of electron–hole pairs and leading to the enhancement of photocatalytic ability. Ideally, the approach to enhance photocatalytic ability is to increase the interface area of the semiconductors at the heterojunction. Therefore, based on this hypothesis, enwrapping the spherical α-S with the selected ultra-thin C₃N₄ nanosheet to obtain large contact area would improve the catalytic efficiency. Another consideration when fabricating an efficient heterojunction is the matching of the band potential of the semiconductors (denoted as semiconductor 1 and semiconductor 2). When the conduction band (CB) and valence band (VB) potential of semiconductor 1 are lower than those of semiconductor 2, the potential gradient drives photogenerated electrons in the CB of semiconductor 2 to that of semiconductor 1 and photogenerated holes in the VB of semiconductor 1 to that of semiconductor 2. Therefore, the electrons and holes are further separated and stored in different semiconductors, and the recombination of them can be suppressed. It is reported that the VB of α-S is 0.25 eV higher than that of anatase TiO₂.⁷ The band gap of α-S is 2.79 eV. Therefore, the calculated CB and VB of α-S are −0.40 and 2.39 eV vs. NHE, respectively. The CB and VB level of g-C₃N₄ are −1.3 and 1.4 V vs. NHE, respectively.²¹ Considering the match of the CB and VB levels with those of α-S, ultra-thin C₃N₄ is suitable for fabricating a heterojunction with α-S.

In this study, the core–shell heterojunctions of an ultra-thin C₃N₄ nanosheet enwrapping a spherical α-S (α-S@C₃N₄) was fabricated, for the first time, for enhanced photocatalytic ability under visible light illumination. The ultra-thin C₃N₄ nanosheet was used as an efficient charge transporter in the heterojunction structure. Based on the rapid charge transfer, big contact surface area and matching VB and CB levels of the two semiconductors, enhanced photocatalytic ability was expected. RhB, a common pollutant in the industrial wastewater, was selected as a test substance to evaluate the photocatalytic performance of the as-prepared samples.

2. Experimental section

Sample preparation

Preparation of hollow spherical α-S. Spherical α-S was fabricated by a soft template process with PVP as the template. First, 1.5 g of Na₂S₂O₃·2H₂O was dissolved in 250 mL water at about 50 °C. This solution was marked as solution A. Moreover, 200 mg of PVP was dissolved in 25 mL of C₂H₅OH/CH₃COOH/H₂O mixture (volume ratio of 1 : 1 : 3) with vigorous stirring, and the resulting PVP solution was noted as solution B. In the following step, solution B was added to solution A followed by stirring for 5 minutes. The resulting mixture was heated to 70 °C followed by the addition of 12 mL of concentrated hydrochloric acid. The temperature was then maintained at 70 °C for 5 min with magnetic stirring. The products were collected by filtration and washed with distilled water until the pH reached 7. The collected samples were then dried at 60 °C for 2 h.

Preparation of ultra-thin C₃N₄ nanosheets. The ultra-thin C₃N₄ nanosheet was fabricated by a “bulk-nanosheet-ultra-thin nanosheet” route, and the detailed procedure is described elsewhere.¹⁹ Briefly, the bulk g-C₃N₄ was ground into powder in a mortar and transferred to a ceramic boat, followed by annealing at 550 °C for 200 min (heating rate of 5 °C min⁻¹). After annealing, a light yellow powder of C₃N₄ nanosheet was produced. Subsequently, 100 mg of C₃N₄ nanosheet powder was dispersed in 100 mL of methanol at room temperature and exfoliated by ultrasonication for 4 h. The final product (ultra-thin nanosheet) was centrifuged and dried in vacuum at 50 °C for 6 h.

Preparation of core–shell heterojunctions of α-S@C₃N₄ composites. The spherical α-S was enwrapped by the ultra-thin C₃N₄ nanosheet via a self-assembly process using electrostatic force. The schematic procedure for preparing the core–shell heterojunctions of α-S@C₃N₄ is presented in Scheme 1. The ultra-thin C₃N₄ nanosheets were post functionalized in HCl (0.5 mol L⁻¹) for 1 h for protonation to get a well-dispersed aqueous solution (1 mg mL⁻¹) prior to use.²² The ultra-thin C₃N₄ nanosheet solution was then added dropwise to the negatively-charged spherical α-S suspension with vigorous stirring. After 15 min, the products were collected by filtration and washed with distilled water until the pH reached 7. The samples were obtained by drying at 60 °C for 2 h. The samples with 25%, 35% and 45% mass ratio of ultra-thin C₃N₄ nanosheets to α-S were denoted as S₁, S₂, and S₃, respectively.

Scheme 1 Schematic procedure for preparing the core–shell heterojunctions of α-S@C₃N₄.

Sample characterization

The thickness and the morphology of the C₃N₄ nanosheets were measured by atomic force microscopy (AFM Agilent Pico Plus) and transmission electron microscopy (TEM; JEM-2100(UHR) JEOL). X-ray diffraction (XRD) spectra were recorded by a Rigaku D/MAX-2400 with Cu Kα radiation, with an accelerating voltage of 40 kV and current of 30 mA. The scanning rate was 8° (2θ) min⁻¹, and the scanning range was 10°–80°. Light absorption intensities were measured using a UV-vis spectrophotometer (Shimadzu, UV-2450) with a wavelength range of 200–800 nm. The morphology of the heterojunction composites was obtained with a field emission scanning electron microscope (FE-SEM, Hitachi S-4800) equipped with an energy dispersive spectrometer (EDS) using a voltage of 5 kV, and the structure of the heterojunction composites was characterized with a high applied voltage of 15 kV and by EDS. Fluorescence (FL) spectra were recorded by an FL spectrometer (LS-55, PE). The specific surface area was determined by an adsorption instrument (Tristar 3000) and calculated using the linear portion of the Brunauer–Emmett–Teller (BET) model.

Measurement of photocatalytic activity

The photocatalytic activities of the photocatalyst products were monitored through the degradation of RhB under visible light irradiation. The photocatalytic reactions were conducted in a 100 mL cuboid quartz reactor. A 300 W Xe lamp was used as the visible light source. The light was passed through a filter to shield any wavelength below 400 nm. The used light intensity was 30 mW cm⁻². In all the experiments, 0.08 g of the photocatalyst was added to 80 mL of a 5 mg L⁻¹ aqueous RhB solution. During each photocatalytic experiment, 5 mL of the suspension was collected at predetermined time intervals. The suspension was centrifuged at 9500 rpm for 10 min, and the concentration of RhB was determined by measuring the absorbance at λ = 554 nm with a Shimadzu UV2000 spectrophotometer. The used photocatalyst was separated from the suspension, washed with deionized water and ethanol three times, and then dried at 60 °C for 6 h for the next experimental run.

3. Results and discussion

Analyses of C₃N₄ nanosheet characterization

The collective analyses indicated that the ultra-thin C₃N₄ nanosheets were obtained by exfoliation. The TEM image illustrates the highly transparent feature of the C₃N₄ nanosheets (Fig. 1(a)), indicating the ultra-thin thickness of the nanosheets. The thickness of the nanosheets is 0.8–1.2 nm, as determined by the analysis of the AFM image (Fig. 1(b)). Because the theoretical distance between each layer of the bulk g-C₃N₄ is 0.326 nm,^23,24 it is assumed that two or three layers of the ultra-thin C₃N₄ nanosheet was successfully fabricated. The XRD patterns were recorded and are shown in Fig. 1(c). In the XRD pattern of bulk g-C₃N₄, the (002) diffraction at around 27.5° relates to the characteristic interlayer stacking structure, while the (100) diffraction at 13.1° indicates the interplanar structural packing. In the pattern of the ultra-thin C₃N₄ nanosheet, the (002) diffraction related to the interlayer stacking is also found. However, the intensity is significantly less after exfoliation. This further demonstrated that the thickness of the C₃N₄ has been largely decreased.

Fig. 1 (a) TEM, (b) AFM image and (c) XRD pattern of the C₃N₄ nanosheet.

Analyses of α-S@C₃N₄ heterojunction characterization

XRD analyses. XRD patterns of the α-S, S₁, S₂ and S₃ are shown in Fig. 2. Clear characteristic peaks at 2θ of 23.1°, 25.9°, 26.7°, 27.6°, 29.0° and 31.4° are observed in the pattern of α-S. The crystal form of α-S is orthorhombic, which is indexed to the standard cards (PDF no. 77-0145). The patterns of S₁, S₂ and S₃ are similar to that of α-S, indicating that the α-S in the heterojunction composites is also orthorhombic in structure, and the fabrication of the heterojunctions cannot change the crystal form of α-S. No diffraction peaks of the ultra-thin C₃N₄ nanosheet are observed for the composite samples, which may be hidden by a spectral line of the α-S.

Fig. 2 XRD patterns of the α-S, S₁, S₂ and S₃.

Morphologies. The morphology of the α-S is spherical with a size of approximately 1 μm (Fig. 3(a)). From Fig. 3(b) and (c), it can be inferred that the particle is almost spherical in shape. With the increase in the amount of ultra-thin C₃N₄ nanosheets loaded, the size of the particle becomes larger, indicating that the C₃N₄ nanosheets are loaded on the α-S sphere and the thickness of the loaded layer of C₃N₄ nanosheets becomes larger. However, when 45% C₃N₄ nanosheet was added to the composite, the original morphology of the α-S disappears (Fig. 3(d)). Some of the C₃N₄ nanosheets cannot be enwrapped on the α-S particles and exist randomly in the composite. To further verify that some C₃N₄ nanosheets have been successfully enwrapped on the α-S sphere, the particle structure of S₁ was characterized by SEM with high detection voltage combined with elemental detection with EDS. It was observed from SEM results that the particles are constructed by two layers (Fig. 3(e)), and from the EDS analysis, it was inferred that the particles are composed of elemental C, N and S (Fig. 3(f)), demonstrating that the ultra-thin C₃N₄ nanosheets were enwrapped on the α-S sphere.

Fig. 3 SEM images of (a) α-S, (b) S₁, (c) S₂, (d) S₃ and (e) SEM image of S₁ with high detection voltage (15 eV) and (f) EDS spectra of the rectangular area outlined in image (e).

Optical absorption. The optical absorption of the as-prepared samples was investigated using a UV-vis spectrometer. From Fig. 4(a), α-S shows intense absorption from 200 to 450 nm, indicating its potential as a visible-light-induced photocatalyst. The band gap was determined by the following formula:

αhν = A(hν − E_g)^n/2 (1)

Fig. 4 (a) UV-vis absorption spectra and (b) calculation of the band gap by Kubelka–Munk function of the α-S, C₃N₄ nanosheets, S₁, S₂ and S₃.

The calculated band gap is 2.80 eV (Fig. 4(b)), similar to previously reported values. The ultra-thin C₃N₄ nanosheet can absorb light from 200 to about 410 nm. Its exact band gap is 3.03 eV, which is obviously smaller than that of bulk g-C₃N₄ (2.70 eV). The similar increase of the band gap has been reported previously and is presumably attributed to the decrease of conjugation length and the strong quantum confinement effect from the ultra-thin structure of the prepared C₃N₄ nanosheets.^18,25 The increased band gap of the ultra-thin C₃N₄ nanosheet is probably due to the negative shift of its CB and/or the positive shift of its VB. The VB and CB potentials of the ultra-thin C₃N₄ nanosheets were estimated based on the band gap difference (0.43 eV) of the ultra-thin C₃N₄ nanosheet and the bulk g-C₃N₄ from the VB and CB levels of the bulk g-C₃N₄. The VB potential is between 1.4 and 1.83 eV, and the CB potential is between −1.73 and −1.3 eV. The calculated band gap of S₁, S₂ and S₃ are 2.78, 2.72 and 2.76 eV, respectively (shown in Fig. 4(b)). Compared to α-S (2.8 eV), the decreases of the band gaps are very minimal and can be ignored.

Separation ability of photogenerated charge carriers. PL spectra analysis is used to reveal the separation efficiency of the photogenerated electrons and holes in semiconductors. Fig. 5 shows the PL spectra of S₁, S₂ and S₃ together with that of α-S for comparison. Considering that the PL emission results from free charge carrier recombination, a lower peak indicates lower rate of their recombination. Compared to α-S, all three heterojunction composites showed significant quenching of the PL, which suggests that the separation of photogenerated electrons and holes can be improved by the construction of heterojunctions between the α-S and ultra-thin C₃N₄ nanosheets. The gradient of the VB and CB levels of α-S and of C₃N₄ drives the holes generated in the α-S to the C₃N₄ nanosheet, and the electrons generated in the C₃N₄ nanosheet to the α-S. Due to the excellent charge transfer ability of the ultra-thin C₃N₄ nanosheet, the separated holes in the ultra-thin C₃N₄ nanosheet are rapidly transferred to the reaction site, which then participate in the surface reaction. This rapid transportation further promotes the separation of photogenerated carrier pairs. Thus, the recombination of holes and electrons can be largely suppressed.

Fig. 5 PL spectra of α-S, S₁, S₂ and S₃.

The PL intensities of S₁, S₂ and S₃ are different. The intensity of S₂ is similar to that of S₃, which is lower than that of S₁. From the perspective of the separation ability of photogenerated carriers, there is an optimal thickness of the coating layer in the heterojunction.²⁶ Compared to S₁, the C₃N₄ nanosheet layer of S₂ is thicker but still within the optimal thickness. Thus, the separation ability of the S₁ for photogenerated holes and electrons of S₁ is lower than that of S₂. Abundant C₃N₄ nanosheets loaded on S₁ increase the thickness of the C₃N₄ nanosheet layer to thicker than the optimal thickness. Thus, the separation efficiency of the electrons and holes is not enhanced.

Photocatalytic activity and stability. Fig. 6(a) shows the photocatalytic activities of the α-S@C₃N₄ composites together with α-S for comparison. RhB concentration gradually decreases in the presence of α-S under visible light, and only 31.1% of RhB is removed after 40 min. However, under the same reaction conditions, the removal rates of RhB with S₁, S₂ and S₃ are 78.1%, 93.4% and 84.1%, respectively. To quantitatively investigate the reaction kinetics of the RhB degradation, the experimental data were fitted by applying a first-order model, as expressed by eqn (2). This equation is well established for photocatalytic experiments when the pollutant is in the millimolar concentration range.

ln(C₀/C_t) = kt (2)

where C₀ and C_t are the dye concentrations at 0 and t of time, respectively, and k is the apparent first-order rate constant. In Fig. 6(b), the calculated k values of S₁, S₂ and S₃ are 0.0382, 0.0635 and 0.0453 min⁻¹, respectively, which are higher than that of α-S (0.00945 min⁻¹), demonstrating that the heterojunction structure formed by enwrapping ultra-thin C₃N₄ nanosheet can enhance the photocatalytic ability of α-S. The k value of S₂ is 6.72 times higher than that of α-S. This enhancement can be attributed to the separation efficiency of the photogenerated holes and electrons. The C₃N₄ nanosheet itself is also a visible-light-driven photocatalyst. To confirm the heterojunction effect, the same amounts of ultra-thin C₃N₄ nanosheet (28 mg) and α-S (52 mg) to those of S₂ were used as the photocatalyst. Under the same conditions, the RhB removal rate was largely lower than that for S₂, confirming that the enhanced photocatalytic ability of S₂ is attributed to the heterojunction effect. The photocatalytic ability of the ultra-thin C₃N₄ nanosheet was also determined (Fig. S1†). The RhB removal rate is also lower than that of S₂, which further confirms the heterojunction effect. The specific surface areas of α-S and α-S@C₃N₄ heterojunctions were also determined (see the ESI†). The exact causal relationship of the specific surface area and the photocatalytic ability will be discussed later.

Fig. 6 (a) C_t/C₀ and (b) ln(C₀/C_t) of the RhB concentration versus reaction time in photocatalytic degradation with α-S, S₁, S₂, S₃ and α-S and ultra-thin C₃N₄ nanosheet under visible light illumination.

Possible mechanisms for the enhanced photocatalytic ability of α-S@C₃N₄ under visible light illumination are proposed (Fig. 7). Once illuminated by the visible light, the α-S and ultra-thin C₃N₄ nanosheet are excited. Because the CB (between −1.73 and −1.3 eV vs. NHE) of the ultra-thin C₃N₄ nanosheet is more negative than that of α-S, the photogenerated electrons in the CB of the C₃N₄ nanosheet are ejected to the CB of α-S. At the same time, the holes generated by α-S are transported to the VB of the C₃N₄ nanosheet due to the positive potential of α-S (between 1.4 and 1.83 eV vs. NHE). The matching values of the CB and VB potential between the two semiconductors keeps the electrons and holes separated. Therefore, the recombination of electron–hole pairs is restrained and more electrons and holes can transfer to the surface of the composite and join in the surface reaction. Thus, the photocatalytic activity can be enhanced. The oxidizing species in the photocatalytic degradation of RhB with α-S and α-S@C₃N₄ are also probed (Fig. S2†). Compared with α-S, the heterojunction construction changes the photocatalytic degradation pathway of RhB. ˙OH radicals no longer serve as oxidizing species with α-S@C₃N₄ as catalyst, and only holes and ˙O₂⁻ radicals contribute to the degradation of RhB. The data and detailed discussion can be read in the ESI.†

Fig. 7 Proposed mechanism for the enhanced photocatalytic ability of α-S@C₃N₄.

Stability is one of the key parameters for evaluating the performance of a photocatalyst. Fig. 8 shows the cycling performance of α-S and S₂ for degrading RhB in a photocatalytic process under visible light illumination. After recycling and reusing in each cycle, the removal rate of α-S dramatically decreases from 31.1% for the 1st cycle to 3.0% for the 5th cycle, indicating that α-S is aging in the photocatalytic process. In contrast, the removal rate of S₂ does not decrease obviously, inferring that coating C₃N₄ nanosheets on α-S can largely enhance the stability of α-S. Based on the results listed above, the α-S@C₃N₄ heterojunction composite is proven to be a good candidate for an efficient, recyclable and stable photocatalyst.

Fig. 8 Cycling runs in the photocatalytic degradation of RhB with α-S and S₂ under visible light illumination.

4. Conclusions

The core–shell heterojunctions of α-S@C₃N₄ were successfully fabricated via a self-assembly process by electrostatic force. The enhancement of the photocatalytic ability of the α-S@C₃N₄ composite could be attributed to the efficient separation of photogenerated holes and electrons by the heterojunction, which has excellent charge transfer ability arising from the ultra-thin C₃N₄ nanosheet. The stability of the α-S is largely enhanced by the heterojunction construction. The excellent photocatalytic ability and stability of α-S@C₃N₄ composite demonstrates its great potential for usage in environmental remediation technology as a highly efficient photocatalyst.

Acknowledgements

This work was supported by the National Science Fund China (project no. 21107007) and Cultivation Program for Excellent Talents of Science and Technology Department of Liaoning Province (no. 2014026009).

References

1. L. F. Yin, J. F. Niu, Z. Y. She and J. Chen, Environ. Sci. Technol., 2010, 44, 5581–5586.
2. X. D. Zhang, W. M. Liao, W. Wu, D. J. Zheng, Y. S. Zhou, B. L. Xue, W. Liu, Z. Q. Lin and Y. L. Deng, J. Mater. Chem. A, 2014, 2, 11035–11039.
3. M. D. Ye, M. Y. Wang, D. J. Zheng, N. Zhang, C. J. Lin and Z. Q. Lin, Nanoscale, 2014, 6, 3576–3584.
4. Z. H. Kang, C. H. A. Tsang, N. B. Wong, Z. D. Zhang and S. T. Lee, J. Am. Chem. Soc., 2007, 129, 12090–12091.
5. Y. D. Chiou and Y. J. Hsu, Appl. Catal., B, 2011, 105, 211–219.
6. F. Wang, W. K. H. Ng, J. C. Yu, H. J. Zhu, C. H. Li, L. Zhang, Z. F. Liu and Q. Li, Appl. Catal., B, 2012, 111, 409–414.
7. G. Liu, P. Niu, L. C. Yin and H. M. Cheng, J. Am. Chem. Soc., 2012, 134, 9070–9073.
8. W. J. Wang, J. C. Yu, D. H. Xia, P. K. Wong and Y. C. Li, Environ. Sci. Technol., 2013, 47, 8724–8732.
9. W. C. Peng and X. Y. Li, Nano Res., 2013, 6, 286–292.
10. M. R. Hoffmann, S. T. Martin, W. Choi and D. W. Bahemann, Chem. Rev., 1995, 95, 69–96.
11. Y. D. Liu, J. Goebla and Y. D. Yin, Chem. Soc. Rev., 2013, 42, 2610–2653.
12. X. J. Dai, Y. S. Luo, W. D. Zhang and S. Y. Fu, Dalton Trans., 2010, 39, 3426–3432.
13. Q. Xiang, J. Yu and M. Jaroniec, Chem. Soc. Rev., 2012, 41, 782–796.
14. X. F. Zhang, X. Quan, S. Chen and H. T. Yu, Appl. Catal., B, 2001, 105, 237–242.
15. X. C. Wang, K. Maeda, A. Thomas, K. Takanabe, G. Xin, J. M. Carlsson, K. Domen and M. Antonietti, Nat. Mater., 2009, 8, 76–80.
16. K. Maeda, X. C. Wang, Y. Nishihara, D. L. Lu, M. Antoniett and K. Domen, J. Phys. Chem. C, 2009, 113, 4940–4947.
17. S. C. Yan, Z. S. Li and Z. G. Zou, Langmuir, 2009, 25, 10397–10401.
18. J. Xu, L. W. Zhang, R. Shi and Y. F. Zhu, J. Mater. Chem. A, 2013, 1, 14766–14772.
19. H. X. Zhao, H. T. Yu, X. Quan, S. Chen, H. M. Zhao and H. Wang, RSC Adv., 2014, 4, 624–628.
20. S. Chu, X. M. Zheng, F. Kong, G. H. Wu, L. L. Luo, Y. Guo, H. L. Liu, Y. Wang, H. X. Yu and Z. G. Zou, Mater. Chem. Phys., 2011, 129, 1184–1188.
21. S. Chu, Y. Wang, Y. Guo, J. Y. Feng, C. C. Wang, W. J. Luo, X. X. Fan and Z. G. Zou, ACS Catal., 2013, 3, 912–919.
22. Y. J. Zhang, A. Thomas, M. Antonietti and X. C. Wang, J. Am. Chem. Soc., 2009, 131, 50–51.
23. K. Schwinghammer, B. Tuffy, M. B. Mesch, E. Wirnhie, C. Martineau, F. Taulelle, W. Schnick, J. Senker and B. V. Lotsch, Angew. Chem., Int. Ed., 2013, 52, 2435–2439.
24. J. Zhang, M. Zhang, R. Q. Sun and X. Wang, Angew. Chem., Int. Ed., 2012, 51, 10145–10149.
25. X. D. Zhang, X. Xie, H. Wang, J. J. Zhang, B. C. Pan and Y. Xie, J. Am. Chem. Soc., 2013, 135, 18–21.
26. H. Wang, X. Quan, H. T. Yu and S. Chen, Carbon, 2008, 46, 1126–1132.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c4ra14623a

---