Graphene nanodots decorated ultrathin P doped ZnO nanosheets as highly efficient photocatalysts

Abstract

A novel kind of highly efficient photocatalyst composed of ultrathin P doped ZnO (ZnO:P) nanosheets decorated with graphene nanodots (GNDs) was fabricated. Taking the unique advantages of the ultrathin ZnO nanosheet structure, and the excellent properties of the GNDs, a 1.6 wt% ZnO:P/GNDs nanosheet composite exhibited an optimum visible-light photocatalytic oxidation activity of 0.442 min⁻¹ for RhB degradation and an excellent photostability over five cycles. This can be attributed to the synergistic effect between the ZnO:P nanosheets and GNDs, leading to a significantly enhanced charge efficiency and visible light absorption.

---

Introduction

In the past few decades, semiconductor based photocatalysis has attracted widespread attention in environmental pollutant treatment by the utilization of the visible light region in the solar spectrum.^1,2 However, it is strictly restrained by the low absorption efficiency and fast charge recombination of semiconductor materials.^3,4 Thus the development of semiconductor materials with a suitable band gap and high specific surface area is of intensive interest for potential photocatalytic applications. Compared with widely used TiO₂ etc.,^5,6 ZnO with a band gap of 3.27 eV has been regarded as another promising candidate for the degradation of organic pollutants, due to its high electronic mobility, high exciton binding energy, abundant morphologies, low cost, and non-toxicity.⁷ Various strategies, including nanostructuring, ion doping, and coupling with other semiconductors, have been proposed to suppress its photogenerated charge recombination and extend the light absorption range.^3,7 Among them, the fabrication of one or two-dimensional (1 or 2D) ZnO based nanostructure composites has emerged as one of the most promising strategies due to the specific properties of the nanostructures, including large surface-to-volume ratios and direct electrical pathways for the rapid transport of photogenerated carriers.^8–10 Compared with ZnO–metal systems, ZnO–carbon systems (such as carbon nanotubes, fullerene and graphene) allow faster charge separation across the interface,⁷ in which the carbon materials act as excellent electron-acceptors/transports to facilitate the migration of electrons and hinder charge recombination, consequently the photocatalytic performance is enhanced significantly.¹¹ Therefore, it is regarded as an effective approach to enhance the photocatalytic performance.

Graphene, as the star of the carbon material family, has been demonstrated to be a promising component for composite fabrication in photocatalytic applications.¹² Enhanced photocatalytic performances of graphene/graphene oxide combined with ZnO,³ CdS,^10,13 MoS₂,¹⁴ and TiO₂ (ref. 5 and 15) were realized, among which the ZnO–graphene composites showed excellent charge transfer performances and good adsorption capabilities for organic pollutants.^7,16,17 The enhancement of the photocatalytic activity of the ZnO–graphene composites is attributed to a strengthened interfacial contact and optimized interfacial composition,¹ where graphene serves as an electron acceptor to inhibit the recombination of charge carriers. The combination of ZnO nanostructures with graphene would be an ideal system to accelerate the charge transfer from the photocatalyst to the interface coming into contact with the organic pollutants. Particularly, zero-dimensional graphene quantum dots (GQDs) exhibit unique properties such as good surface grafting, excellent solubility, low chemical toxicity, and cost-effective preparation, making them fascinating materials for ZnO based composite photocatalysts.¹⁸ However, the contact area between ZnO nanostructures is usually limited and their performances cannot be maximized.

In this study, 2D nanostructured ultrathin P-doped ZnO nanosheet/GND composites were fabricated by using conventional chemical vapor transport and condensation growth of ZnO:P nanosheets,^19,20 followed by the uniform surface decoration of graphene nanodots (GNDs). The composites were characterized by X-ray diffraction (XRD), scanning electron microscopy (SEM), transmission electron microscopy (TEM), atomic force microscopy (AFM), Raman spectroscopy, UV-VIS absorption spectroscopy, and photoluminescence (PL) spectroscopy. In our previous work, it was observed that during the growth of the ZnO nanostructures, P atoms as dopants incorporated into ZnO to partially occupy the O sites. However, due to the different lengths of Zn–P and Zn–O bonds, P atoms introduced lattice distortion when occupying the substitutional O sites, and strain relaxation altered the growth direction of ZnO from the [0001] to [101̄0] direction.²⁰ Consequently, a two-dimensional ultrathin nanosheet structure for ZnO:P was formed. Without P doping only a nanowire array could be synthesized for undoped ZnO. Therefore, P dopants are essential for the synthesis of two-dimensional ultrathin ZnO nanosheets. This work demonstrates that P doping promotes the growth of ultrathin ZnO:P nanosheets to offer a large interfacial area for light adsorption and photocatalytic reactions.²¹ With the increased amount of GNDs, the photocatalytic performance of the composite increases first and reaches an optimum with 1.6 wt% GNDs, and subsequently decreases. The collective advantages of the large surface and interfacial areas not only suppress charge recombination, but also enhance the visible light adsorption of the composite photocatalyst. This can be attributed to the formation of intimate contacts with large interfacial areas between the ultrathin ZnO:P nanosheets and GNDs, thus facilitating the fast charge transfer.

Experimental

Synthesis of ZnO:P/GNDs composites

Ultrathin P-doped ZnO nanosheets were fabricated by using a conventional chemical vapor transport and condensation method.²⁰ Single crystal Si coated with a thin Au layer (1.5 nm) was used as the substrate. ZnO (Alfa Aesar 99.99%) and graphite (200 mesh, Alfa Aesar 99.9%) powders mixed with a mass ratio of 50 to 50 were used as the precursors. Phosphor pentoxide (P₂O₅, Alfa Aesar 99.99%) nanopowder with a mass ratio of 2.5% was added into the precursors. High purity argon (Ar, 99.999%) and oxygen (O₂, 99.999%) with a volume ratio of 7 to 4 were the carrier and/or reaction gases at a constant total flow rate (110 sccm). The growth of the ZnO:P nanosheets was carried out at 1000 °C with a growth time of 8 minutes, and then the furnace was cooled down to room temperature naturally.

For the preparation of the GNDs, graphene fabricated by using the Hummers’ method was firstly heated in a drying oven at 400 °C for 30 s. 80 mL of concentrated HNO₃ mixed with 150 mL of concentrated H₂SO₄ was stirred for 10 minutes. Then, graphene with a mass weight of 4 g was slowly added into the mixed acid, and stirred for 8 h. Subsequently, 85 mL of NaCl solution with a concentration of 0.78 g mL⁻¹ was trickled into the mixed acid with a speed of 1 mL min⁻¹. After that, ammonia was trickled into the mixed solution to obtain a neutral pH value, and dialysis bags were used to remove salt ions and achieve electric neutrality.

The P-doped ZnO nanosheets (50 mg) were dispersed in distilled water (100 mL), and GND solution (5 mg mL⁻¹) was added into the ZnO:P nanosheet suspension. The mixed solution was ultrasonically dispersed (30 minutes), then centrifuged and dried (60 °C, 24 hours) to obtain the composites. To investigate the effect of the GND amount on the photocatalytic performance, 0.25 mL, 0.5 mL, 0.75 mL, 1 mL, and 1.25 mL of the GND solution were separately added into the ZnO:P nanosheet suspension, and the corresponding weight ratios of GNDs in the resultant samples were 0.4 wt%, 1.0 wt%, 1.6 wt%, 2.1 wt%, and 2.7 wt% determined by thermogravimetric analysis (TGA) as is shown in Fig. S1.†

Characterization of ZnO:P/GNDs composites

The thickness of the ZnO:P nanosheets was characterized by using atomic force microscopy (AFM, Cypher, Oxford Inc.) and the surface morphology was examined by using field-emission scanning electron microscopy (FE-SEM, Quanta FEG, FEI). Room temperature photoluminescence (PL) spectra of the GNDs were collected by using a LabRAM HR Evolution spectrometer (Horiba, France) with an excitation wavelength of 325 nm. The decoration of the GNDs on the surface of the ZnO:P nanosheets was characterized by using high resolution transmission electron microscopy (HRTEM, Tencai G20, FEI). X-ray photoelectron spectroscopy (XPS) characterization was performed to investigate the composition and chemical bonding with an upgraded PHI 5000C ESCA System with Mg K_α irradiation (hν = 1253.6 eV). The absorption spectrum was recorded by using a UV/VIS absorption spectrophotometer (UV-VIS-2550). The photocurrent response was measured by using a semiconductor characterization system (Keithley 4200), and a 300 W Xenon lamp was employed as the light irradiation source.

Photocatalytic degradation

The ZnO:P/GNDs composites were utilized to degrade RhB under simulated sunlight irradiation by using a 300 W Xenon lamp. The ZnO:P/GNDs composites (20 mg) were placed in RhB solution (40 mL, 10 mg L⁻¹), and prior to the measurement of photocatalytic performance the solution was magnetically stirred in the dark (15 minutes) to achieve an absorption–desorption equilibrium. The photodegradation experiments were carried out by using a 300 W Xenon lamp to simulate solar irradiation. The solution was sampled at intervals of 2 minutes, and the characteristic absorption band of RhB at a wavelength of 554 nm was measured by using a UV-VIS absorption spectrophotometer.

Detection of hydroxyl radicals (˙OH)

To examine the photocatalytic performance of the ZnO:P/GNDs composites, ˙OH produced during light irradiation was detected by a PL technique using terephthalic acid (TA) as the probe molecule. TA can react with ˙OH to form highly fluorescent 2-hydroxyterephthalic acid (TAOH), which shows a PL peak at 425 nm, and its intensity is proportional to the amount of ˙OH. Experimental procedures were similar to the measurement of photodegradation, except that the RhB solution was replaced with fresh prepared TA (5 × 10⁻⁴ M) mixed with NaOH (2 × 10⁻³ M) solution. The ZnO:P/GNDs composites (7 mg) were dispersed in TA solution (20 mL), and after simulated sunlight irradiation the reaction solution was used to measure the PL intensity at a wavelength of 425 nm.

Results and discussions

Fig. 1a shows the SEM image of the as-grown ZnO:P sample with a nanosheet morphology of several-micrometers width. By using AFM scanning of the cross-sectional profile, the mean thickness of the ZnO:P nanosheets is identified to be about 20 nm, as is shown in Fig. S2.† Compared with their width of several micrometers, the ultrathin feature of the ZnO:P nanosheets is realized. The PL emission of the GNDs dissolved in water (5 mg mL⁻¹) is shown in Fig. S3.† A strong emission peak at 526 nm is observed for the GNDs with an excitation wavelength at 325 nm. The surface morphology of the ZnO:P/GNDs composites was also investigated by the utilization of TEM. As is shown in Fig. 1b, it can be seen that the GNDs are homogenously dispersed on the surface of the ZnO:P nanosheets. Furthermore, the formation of the ZnO:P/GNDs composites is confirmed by Raman spectra. In Fig. 1c, three characteristic peaks (380 cm⁻¹, 438 cm⁻¹ and 579 cm⁻¹) are observed for both the ZnO:P/GNDs composites and ZnO:P nanosheets, which are correspondingly attributed to the A₁ (transverse optical), E^t₁ (transverse optical), and E^l₁ (longitudinal optical) modes of ZnO.^22,23 Two peaks (1350 cm⁻¹ and 1590 cm⁻¹) observed for the ZnO:P/GNDs composites are assigned to the D and G bands of the GNDs, indicating that the GNDs are successfully coated on the surface of ZnO. The chemical bonding properties of the ZnO:P/GNDs composites are characterized by XPS, and the C 1s spectra are shown in Fig. 1d. The main peak at 284.6 eV is attributed to the C–C bond with an sp² orbital. The peaks at 285.9 eV and 288.4 eV are assigned to the C–O and C=O bonds, respectively. Thus, it is further identified that the GNDs were well decorated on the surface of the ultrathin ZnO:P nanosheets. Fig. S4† shows the XPS spectra of the Zn and P elements in the ZnO:P nanosheets. Two peaks of Zn 2p_1/2 and 2p_3/2 respectively located at 1022.2 eV and 1045.2 eV are observed, indicating that Zn is in the chemical state of Zn²⁺. A P related peak at 187.4 eV (P 2s) is observed, confirming that P dopants are essentially incorporated into ZnO, and it can be reasonably inferred that P is negatively charged (P³⁻) and bonded with Zn.²¹

Fig. 1 (a) The SEM image of the ZnO:P nanosheets. (b) The TEM image of the ZnO:P/GNDs composites, inset: high magnification image. (c) Raman spectra of the ZnO:P/GNDs composites and ZnO:P nanosheets. (d) The C 1s spectrum of the ZnO:P/GNDs composites.

The photocatalytic activity of the ZnO:P/GNDs composites was evaluated for the degradation of Rhodamine B (RhB) under simulated sunlight irradiation by using a 300 W xenon lamp. For comparison, the pristine ZnO:P nanosheet sample was also investigated. Fig. 2a and b show the adsorption spectra of RhB with the ZnO:P/GNDs composites and the pristine ZnO:P nanosheet samples, respectively. The RhB adsorption peak decreases evidently with increasing irradiation time by using the ZnO:P/GNDs composites. After exposure to simulated sunlight irradiation for only 2 minutes, 64% of the RhB dye is decomposed and after 6 minutes 91% is decomposed. In contrast, RhB does not degrade dramatically with increasing degradation time when using the pristine ZnO:P nanosheets, and 30% RhB remains after 10 minutes of irradiation. The relative concentration of RhB with respect to time for different amounts of GND loading is shown in Fig. 2c. Irradiation by simulated sunlight in the absence of GNDs with the ZnO:P nanosheets causes a slight decrease in RhB concentration. The photodegradation of RhB is significantly enhanced by using the ZnO:P/GNDs composites, and the photodegradation rate is strongly dependent on the amount of GNDs. The photocatalytic activity improves with an increasing amount of GNDs from 0.4 to 1.6 wt%, whereas further increasing the amount of GNDs leads to a decrease. A maximum photodegradation of 99.3% is obtained for the ZnO:P/GNDs composites with 1.6 wt% GNDs. The photocatalytic degradation process is fit to a pseudo-first-order reaction, expressed as ln(C₀/C), where C and C₀ are the absorbance of RhB at irradiation time t and before irradiation (t = 0), respectively, and the reaction rate constant (k) is calculated according to the UV-VIS spectra. As is shown in Table S1,† the k value for the ZnO:P nanosheets is 0.131 min⁻¹, and a small amount of GNDs can remarkably increase the degradation rate. A maximum k value of 0.442 min⁻¹ is achieved for the ZnO:P/GNDs-1.6 wt% sample, which is about 3 times higher than that of the ZnO:P nanosheets. It is reasonable that the photodegradation rate varies with an increasing GND amount, since a relatively large amount of GNDs would block the light, reducing the effective catalyst area of the ZnO:P nanosheets for the absorption of incident photons. The combination of the ZnO:P/GNDs composites is robust when performing repeat runs for the photocatalytic degradation of RhB, as is shown in Fig. 2d. After running five cycles, 93.3% of RhB can be degraded by the recycled ZnO:P/GNDs composite sample in 10 minutes, suggesting good cycling stability of the ZnO:P/GNDs composites.

Fig. 2 (a) and (b) Time-dependent absorption spectra of RhB solution after degradation with ZnO:P/GNDs-1.6 wt% and the ZnO:P nanosheets, respectively. (c) Relative variation in absorbance of RhB with different amounts of GNDs in the ZnO:P/GNDs composites.

The absorbance spectra of the ZnO:P/GNDs composites recorded by using a UV-VIS spectrophotometer are shown in Fig. 3a. The UV-VIS absorbance substantially improves with increasing GND amounts, indicating that the light harvesting capacity is enhanced by the decoration with the GNDs. The ZnO:P nanosheets show a significant absorption edge at a wavelength of 380 nm, which is associated with the intrinsic bandgap absorption of ZnO (3.27 eV), therefore only the light in the UV region can be adsorbed. Due to the introduction of the GNDs, the absorbance of the ZnO:P/GNDs composites in the visible region is significantly enhanced. With an increasing GND amount, the ZnO:P/GNDs composites exhibit a continuous increase of absorbance in the visible light region. The increased absorbance in the visible light region will be beneficial for the visible light harvesting capacities of the ZnO:P/GNDs composites.

Fig. 3 (a) Absorbance spectra of the ZnO:P/GNDs composites with different amounts of GNDs. (b) Transient photocurrent response.

The transient photocurrent responses without an applied bias potential are presented in Fig. 3b. All of the samples show a positive photocurrent response and a characteristic of an n-type semiconductor. The ZnO:P nanosheets show a quite weak response, while the ZnO:P/GNDs composites display a clear photocurrent boost as soon as the light is turned on, indicating that a number of electrons are generated. The generated photocurrent is gradually improved when increasing the amount of GNDs up to 1.6 wt%, and subsequently decreases with a further increase of the amount of GNDs. Impressively, the photocurrent density of the ZnO:P/GNDs-1.6 wt% sample is about 190 times higher than that of the ZnO:P nanosheets. When the light is switched off, the photocurrent decays to the baseline with a slow response because of the spatially separated electrons released from the GNDs. The increase in photocurrent density for the ZnO:P/GNDs composites can be attributed to the synergistic effect of fast charge transfer between ZnO and the GNDs, which provides effective charge transport channels to facilitate the separation of the photogenerated electrons and holes.

Hydroxyl radicals (˙OH) are deemed to be the dominant oxidant species during the photocatalytic process. The concentration of ˙OH strongly influences the photocatalytic degradation efficiency, which can be determined by PL spectroscopy. Fig. 4a–c present the PL spectra of TAOH solution under simulated sunlight irradiation by using the ZnO:P/GNDs composites, pristine ZnO:P nanosheets, and GNDs, respectively. In the case of the ZnO:P/GNDs composites, the PL intensity at about 425 nm increases rapidly with irradiation time, indicating the massive formation of ˙OH under simulated sunlight irradiation. The small peak at about 360 nm is attributed to the capillary glass tube from the solution sampling. A gradual increase of PL intensity is observed with irradiation time for the ZnO:P nanosheets, demonstrating that ˙OH can be moderately generated by the ZnO:P nanosheets. No obvious PL signal is observed in the GND sample, indicating that almost no ˙OH is produced by the GNDs. Fig. 4d shows the PL spectra of TAOH solution generated by the different photocatalysts under simulated sunlight irradiation for 10 minutes. Evidently, the amount of ˙OH produced by the ZnO:P/GNDs composites is much larger than that produced by the ZnO:P nanosheets or GNDs. Therefore, it can be confirmed that the ZnO:P/GNDs composites exhibit a higher photodegradation activity.

Fig. 4 The PL spectra of TAOH solution generated by (a) ZnO:P/GNDs-1.6 wt% composites, (b) ZnO:P nanosheets, and (c) GNDs. (d) The PL spectra of TAOH solution generated by the different photocatalysts under simulated sunlight irradiation for 10 minutes.

Fig. 5 illustrates the photodegradation mechanism of the ZnO:P/GNDs composites. The proposed reaction processes involved are as follows:²⁴

ZnO + hν → ZnO(h⁺) + ZnO(e⁻)

ZnO(e⁻) → GNDs(e⁻) (electron migration)

GNDs(e⁻) + O₂ → ˙O₂⁻ + GNDs

h⁺ + H₂O → ˙OH + H⁺

˙O₂⁻ + H⁺ → HO₂⁻

HO₂⁻ + ˙O₂⁻ + H⁺ → H₂O₂ + O₂

H₂O₂ + O₂⁻ → ˙OH + OH⁻ + O₂

H₂O₂ + e⁻ → ˙OH + OH⁻

Dye + ˙OH → CO₂ + H₂O

Fig. 5 A tentative mechanism of the photodegradation process of ZnO:P/GNDs composites.

When irradiated by photons with an energy higher than the bandgap of the ZnO:P nanosheets, the photogenerated electrons and holes reacted with O₂ and H₂O on the surface of the ZnO:P nanosheets, respectively. The reactions generate highly reactive ˙OH to degrade the organic molecules. Ultrathin ZnO:P nanosheets with a high surface-to-volume ratio provide a large interfacial area for the absorption of incident photons, and offer intimate contact with the GNDs. Meanwhile, when the RhB molecules come into contact with the GNDs, they are stuck to the surface of the composites through strong π–π interactions with the GNDs, which effectively increases the degradation rate.²⁵ Importantly, the conduction band position of ZnO is higher than that of the GNDs, resulting in the migration of electrons from ZnO to the GNDs.²⁶ Due to the heterojunction at the interface between ZnO and the GNDs as well as the excellent charge-carrier mobility, the photogenerated electrons can be smoothly transferred to the GND sites, which effectively separates the electrons and the holes. Consequently, more photogenerated electrons and holes are involved in the photocatalytic processes, and the photocatalytic activity of the ZnO:P/GNDs nanosheet composite is significantly enhanced. Therefore, GNDs not only act as electron acceptors but also as an electron-transport pathway, resulting in the inhibition of charge recombination, thus an enhancement of the photodegradation rate is achieved. With increasing GNDs, the photocatalytic activity of the ZnO:P/GNDs composites increases first and then gradually decreases. This phenomenon is associated with the shielding effect of the GNDs.^5,27 The increased amount of the GNDs suppresses the recombination rate of the photogenerated carriers, and thus improves the photocatalytic activity.

Conclusions

A highly effective photocatalyst of ZnO:P/GNDs composites is achieved and possesses a photo-oxidation rate of 0.44 min⁻¹ for RhB degradation under simulated sunlight irradiation, which is about 3 times larger than that of pristine ZnO:P nanosheets. The exceptional photocatalytic activity of the ZnO:P/GNDs composites can be attributed to the role of the GNDs acting as photoabsorbers and charge separation channels, and the synergistic effects between the ZnO:P nanosheets and GNDs for the enhancement of light adsorption efficiency, suppression of charge recombination, and improvement of interfacial charge transfer. This study will provide important inspirations for the design of novel high efficient photocatalysts.

Acknowledgements

This work was supported by the NSFC (51402193, 51572173 and 11402149), STCSM (14ZR1428000, 16JC1402200, 15520720300), Shanghai Talent Development Funding and Hujiang Foundation of China (B14006).

Notes and references

1. N. T. Khoa, S. W. Kim, D.-H. Yoo, S. Cho, E. J. Kim and S. H. Hahn, ACS Appl. Mater. Interfaces, 2015, 7, 3524–3531.
2. W. Gao, M. Wang, C. Ran and L. Li, Chem. Commun., 2015, 51, 1709–1712.
3. C. Han, Z. Chen, N. Zhang, J. C. Colmenares and Y.-J. Xu, Adv. Funct. Mater., 2015, 25, 221–229.
4. X. Chen, S. Shen, L. Guo and S. S. Mao, Chem. Rev., 2010, 110, 6503–6570.
5. X. Gou, Y. Cheng, B. Liu, B. Yang and X. Yan, Eur. J. Inorg. Chem., 2015, 2222–2228.
6. C. Lu, R. Chen, X. Wu, M. Fan, Y. Liu, Z. Le, S. Jiang and S. Song, Appl. Surf. Sci., 2016, 360, 1016–1022.
7. N. T. Khoa, S. W. Kim, D. V. Thuan, H. N. Tien, S. H. Hur, E. J. Kim and S. H. Hahn, RSC Adv., 2015, 5, 63964–63969.
8. Y. Bu, Z. Chen and W. Li, RSC Adv., 2013, 3, 24118–24125.
9. P. Worajittiphon, K. Pingmuang, B. Inceesungvorn, N. Wetchakun and S. Phanichphant, Ceram. Int., 2015, 41, 1885–1889.
10. Y. Wang, F. Wang and J. He, Nanoscale, 2013, 5, 11291–11297.
11. D. Chen, D. Wang, Q. Ge, G. Ping, M. Fan, L. Qin, L. Bai, C. Lv and K. Shu, Thin Solid Films, 2015, 574, 1–9.
12. S. Song, B. Cheng, N. Wu, A. Meng, S. Cao and J. Yu, Appl. Catal., B, 2016, 181, 71–78.
13. C. Zhai, M. Zhu, F. Pang, D. Bin, C. Lu, M. C. Goh, P. Yang and Y. Du, ACS Appl. Mater. Interfaces, 2016, 8, 5972–5980.
14. D. Lang, T. Shen and Q. Xiang, ChemCatChem, 2015, 7, 943–951.
15. Y. Sang, H. Liu and A. Umar, ChemCatChem, 2015, 7, 559–573.
16. O. Akhavan, ACS Nano, 2010, 4, 4174–4180.
17. F. S. Omar, H. Nay Ming, S. M. Hafiz and L. H. Ngee, Int. J. Photoenergy, 2014, 2014, 1–8.
18. Y. Liu and D. Y. Kim, Chem. Commun., 2015, 51, 4176–4179.
19. Y. Liu, S. Xie, H. Li and X. Wang, ChemCatChem, 2014, 6, 2522–2526.
20. Y. Zhu, H. Yang, F. Sun and X. Wang, Nanoscale Res. Lett., 2016, 11, 1–8.
21. S. Xie, Y. Liu, Z. Chen, X. Chen and X. Wang, RSC Adv., 2013, 3, 26080–26085.
22. K. A. Alim, V. A. Fonoberov and A. A. Balandin, Appl. Phys. Lett., 2005, 86, 053103.
23. G. Han, S. Wang, X. Chen and Y. Hong, J. Univ. Shanghai Sci. Technol., 2008, 30, 573–577.
24. R. C. Pawar and C. S. Lee, Appl. Catal., B, 2014, 144, 57–65.
25. Y. Wen, H. Ding and Y. Shan, Nanoscale, 2011, 3, 4411–4417.
26. D. I. Son, B. W. Kwon, D. H. Park, W.-S. Seo, Y. Yi, B. Angadi, C.-L. Lee and W. K. Choi, Nat. Nanotechnol., 2012, 7, 465–471.
27. Q. Li, B. Guo, J. Yu, J. Ran, B. Zhang, H. Yan and J. R. Gong, J. Am. Chem. Soc., 2011, 133, 10878–10884.

---

Footnote

† Electronic supplementary information (ESI) available: Experimental section, additional figures and tables. See DOI: 10.1039/c6ra11446f

---