Preparation of polydopamine-functionalized graphene–Fe₃O₄ magnetic composites with high adsorption capacities

Abstract

In the present study, we report a simple solution mixing method to prepare polydopamine-functionalized graphene–Fe₃O₄ (DGF) nanocomposites with high adsorption capacities and an easy-separation ability. Water-soluble Fe₃O₄ particles are firmly deposited onto the surfaces of graphene oxide (GO) via electrostatic and hydrogen interactions. The interaction between the GO and Fe₃O₄ particles can prevent the graphene nanosheets from restacking and the Fe₃O₄ particles from agglomeration. The introduction of dopamine to functionalize GO not only reduces the GO but also endows abundant chemical groups. The existence of polydopamine affords more active sites for adsorption and further enhances the interaction of the GO and Fe₃O₄ particles to obtain adsorbent materials with stable structures. The adsorption capacity of DGF nanocomposites for methylene blue (MB) is 365.39 mg g⁻¹, which is much higher than that of graphene–Fe₃O₄ (GF) nanocomposite. Simultaneously, the DGF nanocomposites can be easily removed from polluted water after adsorption for MB by using a magnetic field, which is highly important for water conservation.

---

1. Introduction

Graphene is a new type of two-dimensional carbon nanostructure with excellent properties, such as extraordinary mechanical, electrical, and thermal properties. Its unique chemical structure and inspiring properties make it a promising material in many potential applications, including lithium ion batteries, supercapacitors, sensors, nanocomposites, and so on.^1–3 Also, its high specific surface area makes graphene an excellent choice for adsorbents of organic contaminants for wastewater treatment. However, how to achieve high adsorption capacities still has some technological challenges for graphene, due to the relatively low density of surface functional groups and the easy stacking of graphene sheets during reduction.⁴

A large amount of research has targeted on how to overcome these challenges. One attractive and efficient way to improve the dispersion of graphene is to use polydopamine (PDA) as the adhesive coating on two-dimensional nanofillers, which will endow graphene with new functionality. Dopamine (or 3,4-dihydroxyphenethylamine), which contains catechol and amine groups,^5,6 can self-polymerize under room temperature to produce PDA.⁷ PDA can strongly adhere to the surface of many organic and inorganic substance by forming covalent bonds and many other intermolecular interactions, such as hydrogen bonds. Yang et al. reported that dopamine makes each graphene sheet sandwich between two PDA, which prevents the agglomeration of graphene sheets and allows the specific surface area to be enhanced.⁸ Furthermore, there are many active groups on the surface of PDA, which endow it with secondary reaction abilities.^9–11 On the other hand, depositing nanoparticles on graphene oxide (GO) sheets can also prevent graphene nanosheets from restacking and thus endow new functionality to this 2D carbon nanomaterial. It is reported that Fe₃O₄,¹² TiO₂,¹³ and Co₃O₄ ^14,15 could be deposited on GO sheets. It is well known that Fe₃O₄ nanoparticles could be easily separated by an external magnetic field, owing to their unique superparamagnetism effect.^16,17 The hybrid of graphene and Fe₃O₄ could effectively combine their adsorption and magnet-separation abilities and represent a potential new class of adsorbents for wastewater treatment.^18–21 Zhang et al. fabricated highly crystalline Fe₃O₄–graphene composites by the one-step thermolysis of iron–organic complex and graphene oxide sheets under an oxygen-free condition, and which showed a high efficiency of removing methylene blue (MB) molecules.²⁰ Xie et al. prepared a superparamagnetic grapheme oxide–Fe₃O₄ hybrid composite by the deposition of amino-functionalized Fe₃O₄ particles onto grapheme oxide sheets, showing the adsorption capacities of 171.3 mg g⁻¹ for MB.²¹ Nevertheless, the combination of graphene and Fe₃O₄ by hydrothermal or other chemical processes is complicated compared to physical procedures,^20–23 which extremely limits the application of Fe₃O₄–grapheme composites in wastewater treatment.

Herein, polydopamine-functionalized graphene–Fe₃O₄ nanocomposites (DGF) were successfully prepared by a facile solution mixing process. Positive water-soluble Fe₃O₄ are chosen to deposit onto negative GO sheets by electrostatic adsorption, as this is an easy operational process. The interaction between the GO and Fe₃O₄ particles could prevent graphene nanosheets from restacking and Fe₃O₄ particles from agglomeration. Furthermore, dopamine is used to functionalize GO–Fe₃O₄, which plays many functional roles as follows (1) reduces GO into graphene; (2) coats onto the surfaces of graphene–Fe₃O₄ and prevents the restacking of GO sheets; (3) enhances the interaction of GO and Fe₃O₄ particles to obtain stable structure; and (4) endows abundant chemical groups for adsorption. The DGF hybrids not only have excellent adsorptive properties, but also can be separated quickly by an external magnetic field instead of a time-consuming process, such as sedimentation, centrifugation and filtration.^18,19 The results showed that DGF hybrids could become one of the most promising candidates for water purification due to their high adsorption capacity and convenient magnetic separation.

2. Experimental section

2.1 Synthesis of polydopamine-graphene–Fe₃O₄ (DGF)

GO was prepared from natural graphite by the modified Hummer's method. Natural graphite (1 g) and NaNO₃ (1 g) were ground together, and then the mixtures were transferred to a 1000 mL flask. Concentrated H₂SO₄ (60 mL) was slowly dropped into the flask and stirred in an ice bath for 20 min. Further, KMnO₄ (6 g) was slowly added within 10 min. The mixture was further stirred for 5 days at room temperature. Deionized water (150 mL, 50–60 °C) was slowly added and then stirred for 2 h. Afterwards, 30% H₂O₂ (10 mL) was dropped into the solution to reduce the residues until the slurry turned golden yellow. After being washed with HCl solution and water, the brown solutions were diluted and ultrasonicated to a brown-yellow, homogeneous dispersion of GO. Water-soluble Fe₃O₄–OH₂⁺ nanoparticle solutions were synthesized according to the method reported previously.²⁴

The fabrication process of DGF hybrids is illustrated in Scheme 1. Dopamine (6.25 mg) was added into the mixture of GO (12.50 mg) aqueous dispersions and Fe₃O₄ (6.25 mg) nanoparticle solutions. Then, the pH was adjusted to about 7.0. The volume of the solution was controlled to 5 mL. When the mixture was heated at 60 °C for 24 h without any disturbance, the dark DGF solution was obtained. DGF could be separated by a magnetic field for further use. In contrast, the preparation of graphene–Fe₃O₄ (GF) nanocomposites was similar to DGF but without dopamine.

Scheme 1 Illustration for the synthesis of DGF hybrids.

2.2 Materials characterizations

The as-synthesized materials were characterized by X-ray diffraction (XRD), performed on a Rigaku D/max 2550VB/PC diffractometer at room temperature. The patterns were recorded over the angular range 3–80° (2θ), using Cu Kα radiation (λ = 0.154056 nm) with a working voltage and current of 40 kV and 100 mA, respectively. High-resolution transmission electron microscopy (HRTEM: JEOL JEM-2100), and field emission scanning electron microscopy (FE-SEM: HITACHI S-4800) with energy dispersive X-ray spectroscopy (EDS) were used to characterize the morphology microstructures. The Raman spectra were obtained using a spectrometer (Via + Reflex) using an excitation laser wavelength of 514 nm. The magnetic properties were measured at room temperature on a vibrating sample magnetometer (VSM, lakeshore7407) at room temperature with an applied magnetic field from −18 kOe to 18 kOe.

2.3 Adsorption and desorption experiments

Adsorption experiments were carried out in a glass beaker at 25 °C, and 25 mL of MB solution of a known initial concentration was stirred with 25 mg of magnetic DGF. After magnetic separation using the permanent magnet, the equilibrium concentrations of the dyes were measured with a UV-vis spectrophotometer at a wavelength of 664 nm. The adsorption capacity of MB on the adsorbents was evaluated according to the equation:
where q (mg g⁻¹) is the amount of MB adsorbed onto the adsorbents at equilibrium, C₀ (mg L⁻¹) and C_e (mg L⁻¹) are the initial and equilibrated MB concentrations, respectively, V (L) is the volume of solution added, and W (g) is the mass of the adsorbent.

For the desorption study, 25 mg of DGF adsorbent was added to 25 mL MB solution with a concentration of 50 mg L⁻¹, and the mixture was stirred at ambient temperature for 20 min. After the magnetic separation, the supernatant dye solution was discarded, and the adsorbent was separated. Then, 25 mL ethanol was added to the adsorbent and stirred for 20 min. The adsorbent was collected using a magnet and reused. The supernatant solutions were measured by UV-vis spectra, and the cycle of absorption–desorption processes were successively conducted 5 times.

3. Results and discussion

3.1 Morphology and composition

Fig. 1 shows the XRD patterns of GO, DGF, GF, Fe₃O₄. As shown in Fig. 1a, a sharp diffraction (002) peak at 2Θ = 10.26° is clearly observed, indicating that nature graphite is converted into GO after acute oxidation by the modified Hummer's method. The corresponding d-spacing of (002) peak is 0.86 nm, much higher than that of graphite (∼0.34 nm).¹⁹ This may be attributed to the formation of oxygen-containing functional groups on the graphite surface. Fig. 1d shows the typical XRD pattern of as-prepared Fe₃O₄. The diffraction peaks at 30.16°, 35.60°, 43.22°, 53.78°, 57.32° and 62.92° correspond to the (220), (311), (400), (422), (511) and (440) planes of Fe₃O₄ (JCPDS no. 65-3107). Compared to Fe₃O₄, DGF and GF show all the diffraction peaks of Fe₃O₄ without the peak of GO at 2Θ = 10.26°. The phenomenon of the disappearance of the (002) peak of GO is also reported by other groups. It may be attributed to the strong peaks of Fe₃O₄ which overwhelm the weak GO peaks. On the other hand, the presence of Fe₃O₄ and PDA reduces the aggregation of graphene sheets, which results in more monolayer graphene, resulting in the formation of a broad peak at 2Θ = 25°.^25,26

Fig. 1 XRD patterns of (a) GO, (b) DGF, (c) GF and (d) Fe₃O₄.

Raman spectroscopy is a very powerful and valuable technique to characterize carbon materials. The Raman spectrum of graphene is usually characterized by two main features, the G band at 1575 cm⁻¹, indicating the E_2g phonon of C sp² atoms in a 2-dimensional hexagonal lattice, while the D band at 1350 cm⁻¹ arises from a breathing mode of κ-point photons of A1g symmetry.^27,28 The intensity ratio of the D band to the G band (I_D/I_G) is usually used as a measure of the disorder. Fig. 2 shows the Raman spectra of GO, DGF, and GF. The I_D/I_G of GF (0.84) shows a similar value compared with that of GO (0.83), indicating the non-destruction of the GO structure by the electrostatic adhesion between GO and Fe₃O₄–OH₂⁺. Furthermore, DGF has a higher I_D/I_G value of 0.87. This may be attributed to the functional groups on the surface of dopamine, which increases the edge defect of DGF by forming adhesive layers on graphene.

Fig. 2 Raman spectra of (a) GO, (b) DGF and (c) GF composite.

The successful incorporation of polydopamine into GO was further confirmed by FTIR spectroscopy. Fig. 3 displays the FTIR spectrum of GO and DGF. The bands of GO at 1729 cm⁻¹ and 3414 cm⁻¹ indicates the existence of a large amount of carbonyl groups (C=O) and hydroxyl groups (–OH) on the surface of GO. Furthermore, the band of aromatic double bond (C=C) stretching could be visible at 1618 cm⁻¹. After combination with dopamine and Fe₃O₄, the wide peak at 3414 cm⁻¹ significantly decreased and the peak of C=O at 1729 cm⁻¹ almost disappeared in DGF. These are mainly attributed to the removal of oxygen functionalities from GO as a result of the reduction effect of dopamine. In addition, the peaks at 1557 cm⁻¹ and 1005 cm⁻¹ correspond to the N–H and C–N stretching vibrations, respectively. The FTIR data suggests the successful introduction of dopamine into the DGF hybrid.

Fig. 3 FTIR spectra of (a) GO and (b) DGF.

To further identify the chemical bonding information in the composites, the GO and polydopamine-graphene (DG) samples were examined by XPS. As shown in Fig. 4a, the C 1s peak of GO is deconvoluted into four components with binding energies at 284.5, 286.7, 288.2, and 288.9 eV, corresponding to C–C/C=C, C–O, C=O and O–C=O species, respectively. The C 1s spectra of DG are similar to GO, but the intensities of the peaks for the oxygen functional groups decreased significantly, while the peak of C–C/C=C became dominant. This suggests that most GO was reduced to graphene by dopamine. The O 1s spectra of DG can be divided into the three components of C–O (531.3 eV), O–C=O (532.3 eV), and C=O (533.1 eV), which correspond to the spectra of C 1s (Fig. 4c). Furthermore, a new peak of N 1s emerges in DG, which proves the existence of PDA on the graphene sheets. This peak could be deconvoluted into three components, whose binding energies at 398.9 eV, 399.9 eV, and 401.7 eV correspond to N–R, R–NH–R, and R–NH₂ groups, respectively (Fig. 4d). The presence of secondary amine in DG suggests that dopamine was spontaneously polymerized by PDA during its oxidation by GO. The conclusion in DG could indirectly reflect the polymerization of dopamine on the surface of graphene and Fe₃O₄ and the reduction of GO by dopamine.

Fig. 4 XPS spectra for (a) C 1s peak of GO and (b) C 1s, (c) O 1s, and (d) N 1s peaks of DG.

The morphology and structures of the as-prepared GO, Fe₃O₄, DGF and GF are fully detected by TEM and HRTEM observations. The representative TEM image of the GO sample is depicted in Fig. 5a. The obtained GO samples are transparent and have a crumpled and rippled structure. This can attributed to the deformation upon the exfoliation process. Fig. 5b shows the TEM image of pure Fe₃O₄ nanoparticles, which possess a narrow particle size distribution ranging from 16–22 nm. Moreover, with a combination of GO sheets and Fe₃O₄ nanoparticles, as shown in Fig. 5c and d, the magnetite particles are firmly anchored onto the surfaces of GO substrates, because of the electrostatic interaction between negative GO and positive Fe₃O₄. The existence of graphene-nanosheets substrate could effectively prevent the nanoparticles from agglomeration and endow a good dispersion of these magnetite nanoparticles. The morphology of DGF composites has no apparent difference when compared with that of GF. However, it should be emphasized that with the presence of dopamine, more magnetite particles are anchored onto the graphene (TEM image in Fig. 5d), owing to the bridge-like linking of polydopamine with the abundant groups (i.e., hydroxyl and amino-groups). Furthermore, the corresponding HRTEM image in Fig. 5d reveals that the obtained Fe₃O₄ nanoparticles have a good crystallinity and lattice spacings of 0.29 nm, which is in good agreement with the (200) planes of Fe₃O₄.²¹ Therefore, it is believed that such a perfect DGF structure would have a high adsorption performance, as a result of the good dispersion of nanoparticles and nanosheets, and abundant groups from polydopamine and the excellent magnetic response ability.

Fig. 5 TEM images of (a) GO, (b) Fe₃O₄, (c) GF hybrid, and (d) DGF hybrid and inset is HR-TEM image of Fe₃O₄.

3.2 Adsorption properties

The adsorption properties of the as-prepared samples toward MB dye from aqueous solution are investigated in detail. Fig. 6 shows the effect of contact time on the amount of dye adsorbed on GF and DGF at an initial concentration of 50 mg L⁻¹ of MB. It can be seen that the adsorption process is very fast and could attain equilibrium within 5 min. The extremely short equilibrium time suggests the high adsorption efficiency of graphene composites. The adsorption ability of DGF and GF is above 49 mg g⁻¹, the corresponding removal efficiency is close to 100%. Such an excellent adsorption capability and fast adsorption ability may be attributed to the π–π and electrostatic interactions between the aromatic rings of MB molecules and graphene.^18,19 Besides, the Fe₃O₄ nanoparticles could disperse onto graphene uniformly and prevent graphene sheets from agglomeration. This enhanced contact area between graphene and MB is favorable for the adsorption of MB.

Fig. 6 The effect of contact time on the adsorption of 50 mg L⁻¹ MB onto (a) GF and (b) DGF.

Fig. 7 shows the adsorption isotherm for MB with different initial concentrations on GF and DGF. The adsorption abilities of DGF and GF increase with the increasing initial concentration of MB and finally reached saturation. The saturated adsorption capabilities for GF and DGF are almost 293.37 mg g⁻¹ and 365.39 mg g⁻¹. It can be clearly seen that the adsorption capability of DGF is much better than that of GF. GF shows a higher residual MB concentration than DGF when the initial MB concentration is above 100 mg L⁻¹. The high adsorption ability of DGF could be attributed to there being more active sites on graphene surfaces, such as catechol groups, resulting from the PDA polymerized on the graphene nanosheets. DGF hybrids have a large amount of negatively charged oxygen functional groups, which are favorable for the adsorption of cationic MB.

Fig. 7 Isotherms of MB adsorption on the (a) GF and (b) DGF composites (temperature: 25 °C; contact time: 20 min).

The adsorption isotherm is the most important parameter to describe the process of an adsorption system. Fig. 7 shows the adsorption systems of MB dye on the GF and DGF composite at different initial MB concentrations. The MB adsorption on different samples can be simulated by Langmuir and Freundlich isotherm models,¹⁶ which could be expressed as follows:

where q_e (mg g⁻¹) is the amount of MB adsorbed by the absorbents at equilibrium; C_e (mg L⁻¹) is the equilibrium concentration of MB; b is the constant of Langmuir; q_max (mg g⁻¹) is the maximum adsorption capacity of MB calculated by the Langmuir isotherm model; and k and n are the Freundlich constants related to the adsorption ability.

Table 1 summarizes the constants of Langmuir and Freundlich isotherm models and the calculated coefficients. It can be found that the maximum adsorption capacity of DGF is 371.75 mg g⁻¹, which is much higher than that of GF (289.27 mg g⁻¹). The regression coefficient R² of DGF calculated from the Langmuir model is much higher than that from the Freundlich model, indicating that the Langmuir isotherm model matches better with the experimental data. These results demonstrate that the adsorption of MB is a monomolecule layer adsorption process, and the adsorption takes place at the surface of DGF. However, the Freundlich isotherm model is more aligned with GF, in which R² is higher than that of Langmuir model.

Table 1 Adsorption kinetic parameters for MB adsorption on the DGF and GF composite

Adsorbent	Langmuir			Freundlich
	qmax (mg g^−1)	b (L mg^−1)	R^2	k	n	R^2
GF	289.27	0.20	0.92	92.68	3.85	0.95
DGF	371.75	0.81	0.95	162.40	5.26	0.81

---

Fig. 8a shows the magnetic hysteresis loops of water-soluble Fe₃O₄ nanoparticles, GF and DGF. The magnetization loops are critical S-like curves. The saturation magnetization (M_s) of Fe₃O₄, GF and DGF is 69.6, 20.6, and 18.0 emu g⁻¹, respectively. All the samples show superparamagnetic behavior. DGF has a lower value of M_s than that of GF and pure Fe₃O₄, which could be attributed to the relatively low amounts of Fe₃O₄ on graphene (the estimated Fe₃O₄ amount in DGF is 25%). The DGF composites can be quickly removed by an external magnetic field instead of time-consuming processes such as sedimentation, centrifugation and filtration, meaning that they can also be recycled many times in this simple and easy way.^16,18,21 As known, the cyclic utilization of the adsorbent is critical for industrial applications. Fig. 8b shows the adsorption capacity of MB on the DGF composite in 5 cycles of adsorption–desorption processes. After the adsorption of MB, the adsorbent could be quickly separated by a magnet. Then the hybrid could be desorbed by the ethanol and reused for the next absorption process. After five cycles, the absorption ability of DGF decreases to 36.79 mg g⁻¹, and the corresponding removal efficiency is 74%. On the other hand, even after 5 adsorption–desorption cycles, DGF composites still have a better external magnetic response and can be removed by a magnet in just 1 min. This suggests that a close integration between graphene and Fe₃O₄, owing to the electrostatic interaction and hydrogen-bond interaction.

Fig. 8 (a) Magnetic hysteresis loops of Fe₃O₄, GF and DGF. The inset picture shows aqueous solutions of MB before and after adsorption; (b) adsorption capacity of MB on the DGF composite in 5 successive cycles of adsorption–desorption (initial MB concentration: 50 mg L⁻¹; temperature: 25 °C; contact time: 20 min).

4. Conclusions

In summary, polydopamine-functionalized graphene–Fe₃O₄ (DGF) nanocomposites were successfully prepared by a facile solution method. The hybrids were used to absorb the MB dyes, and the adsorption mechanisms were investigated in detail. The Fe₃O₄ particles were adsorbed onto the graphene surfaces by electrostatic and hydrogen interactions. DGF and GF could absorb MB dyes in 5 min and could be separated by an external magnetic field in 1 minute, instead of a time-consuming process such as centrifugation and filtration. The saturated adsorption capabilities for DGF and GF are 365.39 mg g⁻¹ and 293.37 mg g⁻¹, respectively. The enhancement of adsorption capability could be ascribed to the functionalization of the polydopamine, which not only endows abundant chemical groups for adsorption, but also could form adhesive coatings on the graphene to prevent the aggregation of graphene sheets and thus it protects the stable structure of the hybrids. The magnetic capability of nanocomposites could also endow it with recycling properties. The removal efficiency of DGF could reach 74% even after 5 successive cycles of adsorption–desorption processes. DGF could be used to remove the contaminants from water, which is very important for water conservation. This study demonstrates an easy way to fabricate reusable graphene based nanocomposites with high adsorption properties. We believe that DGF will become a promising candidate for the water purification.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (51173043, 21136006, 21236003, 21322607), the Special Projects for Nanotechnology of Shanghai (11nm0500200, 12nm0502700), the Basic Research Program of Shanghai (13JC1408100, 13NM1400801), Program for New Century Excellent Talents in University (NCET-11-0641), the Fundamental Research Funds for the Central Universities.

Notes and references

1. T. Kuilla, S. Bhadra, D. H. Yao, N. H. Kim, S. Bose and J. H. Lee, Prog. Polym. Sci., 2010, 35, 1350.
2. X. Huang, Z. Y. Yin, S. X. Wu, X. Y. Qi, Q. Y. He, Q. C. Zhang, Q. Y. Yan, F. Boey and H. Zhang, Small, 2011, 7, 1876.
3. O. C. Compton and S. T. Nguyen, Small, 2010, 6, 711.
4. S. Stankovich, D. A. Dikin, R. D. Piner, K. A. Kohlhaas, A. Kleinhammes, Y. Y. Jia, Y. Wu, S. T. Nguyen and R. S. Ruoff, Carbon, 2007, 45, 1558.
5. H. Lee, S. M. Dellatore, W. M. Miller and P. B. Messersmith, Science, 2007, 318, 426.
6. I. Kaminska, M. R. Das, Y. Coffinier, J. Niedziolka-Jonsson, J. Sobczak, P. Woisel, J. Lyskawa, M. Opallo, R. Boukherroub and S. Szunerits, ACS Appl. Mater. Interfaces, 2012, 4, 1016.
7. L. Q. Xu, W. J. Yang, K. Neoh, E. Kang and G. D. Fu, Macromolecules, 2010, 43, 8336.
8. L. P. Yang, J. H. Kong, W. A. Yee, W. S. Liu, S. L. Phua, C. L. Toh, S. Huang and X. H. Lu, Nanoscale, 2012, 4, 4968.
9. C. Cheng, S. Li, S. Q. Nie, W. F. Zhao, H. Yang, S. D. Sun and C. S. Zhao, Biomacromolecules, 2012, 13, 4236.
10. X. L. Hu, R. R. Qi, J. Zhu, J. Q. Lu, Y. Luo, J. Y. Jin and P. K. Jiang, J. Appl. Polym. Sci., 2014, 39754.
11. J. Ryu, S. H. Ku, M. Lee and C. B. Park, Soft Matter, 2011, 7, 7201.
12. H. K. He and C. Gao, ACS Appl. Mater. Interfaces, 2010, 2, 3201.
13. T. W. Lu, R. B. Zhang, C. Y. Hu, F. Chen, S. W. Duo and Q. H. Hu, Phys. Chem. Chem. Phys., 2013, 15, 12963.
14. Z. S. Wu, W. C. Ren, L. Wen, L. B. Gao, J. P. Zhao, Z. P. Chen, G. M. Zhou, F. Li and H. M. Cheng, ACS Nano, 2010, 4, 3187.
15. Y. Y. Liang, Y. G. Li, H. L. Wang, J. G. Zhou, J. Wang, T. Regier and H. J. Dai, Nat. Mater., 2011, 10, 780.
16. W. Fan, W. Gao, C. Zhang, W. W. Tjiu, J. S. Pan and T. X. Liu, J. Mater. Chem., 2012, 22, 25108.
17. M. C. Liu, C. L. Chen, J. Hu, X. L. Wu and X. K. Wang, J. Phys. Chem. C, 2011, 115, 25234.
18. S. Bai, X. P. Shen, X. Zhong, Y. Liu, G. X. Zhu, X. Xu and K. M. Chen, Carbon, 2012, 50, 2337.
19. J. Su, M. H. Cao, L. Ren and C. W. Hu, J. Phys. Chem. C, 2011, 115, 14469.
20. F. J. Zhang, J. Liu, K. Zhang, W. Zhao, W. K. Jang and W. C. Oh, Korean J. Chem. Eng., 2012, 29, 989.
21. G. Q. Xie, P. X. Xi, H. Y. Liu, F. J. Chen, L. Huang, Y. J. Shi, F. P. Hou, Z. Z. Zeng, C. W. Shao and J. Wang, J. Mater. Chem., 2012, 22, 1033.
22. Y. H. Xue, H. Chen, D. S. Yu, S. Y. Wang, M. Yardeni, Q. B. Dai, M. M. Guo, Y. Liu, F. Lu, J. Qu and L. M. Dai, Chem. Commun., 2011, 47, 11689.
23. J. J. Liang, Y. F. Xu, D. Sui, L. Zhang, Y. Huang, Y. F. Ma, F. F. Li and Y. S. Chen, J. Phys. Chem. C, 2010, 114, 17465.
24. X. Yang, C. L. Chen, J. X. Li, G. X. Zhao, X. M. Ren and X. K. Wang, RSC Adv., 2012, 2, 8821.
25. J. H. Deng, X. R. Zhang, G. M. Zeng, J. L. Gong, Q. Y. Niu and J. Liang, Chem. Eng. J., 2013, 226, 189.
26. J. H. Byeon, J. H. Park, K. Y. Yoon and J. Hwang, Nanoscale, 2009, 1, 339.
27. L. M. Malard, M. A. Pimenta, G. Dresselhaus and M. S. Dresselhaus, Phys. Rep., 2009, 473, 51.
28. H. C. Gao, Y. M. Sun, J. J. Zhou, R. Xu and H. W. Duan, ACS Appl. Mater. Interfaces, 2013, 5, 425.

---