Reduction of graphene oxide gel with carbon nanotubes, sulfur cathode material preparation and electrochemical performance

Abstract

Reduction of graphene oxide with different reduction degrees was carried out using a hydrothermal reduction method at 130–190 °C. X-ray diffraction, scanning electron microscopy, Fourier transform infrared spectrometry, Raman spectroscopy, and electrochemical testing methods were conducted to investigate the effect of reduction of the graphene oxide gel on the Li–S battery electrochemical performance. The results indicated that the reduced graphene oxide gel was a 3D network graphite structure. The sulfur-based composite material was uniformly embedded, and sulfur polymer loss was suppressed. The reduction degree of graphene oxide (GO) gel increased with increasing water fluid temperatures. In contrast, when the water fluid temperature was 190 °C, the best rate performance of the electrode was observed.

---

1. Introduction

Chemical power sources as energy storage devices have properties between super capacitor and capacitors. Lithium sulfur batteries (Li–S), as a high-energy chemical power supply system with high capacity (1675 mA h g⁻¹), high specific energy (2600 W h kg⁻¹), low cost, no pollution, and many other significant advantages,^1–3 and possess broad application space in the field of mobile electronic devices and other fields. In addition, lithium battery itself is abundant, low cost, and environmentally friendly; therefore, this battery has attracted considerable attention. However, numerous limiting factors have hindered its extensive practical application. The first limiting factor is the electronic insulation of elemental sulfur; its room temperature conductivity is only about 5 × 10³⁰ S cm⁻¹,^4–6 which causes low utilization of the active material. Second, the high solubility of the sulfur polymer ion S_n²⁻ (3 ≦ n ≦ 6) in an electrode reaction results in loss of active substances. Moreover, S_n²⁻ shuttling back and forth causes the shuttle between the anode and cathode.^7–9 Third, the end product of electrode reaction is insulative lithium sulfide (Li₂S, Li₂S₂), which is deposited on the surface of cathode metal lithium and causes loss of active substances. The apparent volume effect of elemental sulfur in the process of charging and discharging also leads to a decline in battery performance.

To attempt to solve the aforementioned issues encountered by Li–S batteries, researchers have conducted excellent studies, such as mixing different types of carbon materials (carbon nanotubes^10–14 and graphene^15,16) and elemental sulfur, to improve the materials conductive performance and active material utilization, but many disadvantages still exist. In the electrode reaction process, carbon nanotubes cannot satisfactorily inhibit the sulfur polymer ion dissolution in the electrolyte, eventually leading to cell cycle performance degradation. To prevent this excessive dissolution, covering a layer of conductive material outside the carbon/sulfur compound material is the best approach. This approach not only improves the conductive performance of sulfur electrode, but also prevents excessive S_n²⁻ diffusion and dissolution. Given the unique structure and excellent physical and chemical properties of 2D,^17,18 graphene is often used to improve the performance of carbon/sulfur composites.

In this study, graphene oxide (GO) was prepared using Hummers' method, and then 1 mg ml⁻¹ GO aqueous solution was prepared. After GO was ultrasonically dispersed, RGO hydrogel with different reduction degrees was prepared by hydrothermal method at different temperatures, in which the carbon nanotube/sulfur (CNT/S) was dispersed as the base material. Subsequently, the preparation of RGO/CNT/S composites was completed after freeze-drying. The structure, morphology, and electrochemical properties were characterized by X-ray diffraction (XRD), scanning electron microscopy (SEM), and Fourier transform infrared spectrometry (FTIR). A LAND test instrument and electrochemical workstation were also employed. The results showed that RGO/CNT/S materials prepared by a hydrothermal reduction method and after freeze-drying greatly improved the lithium sulfur battery performance.

2. Experimental

2.1 Preparation of GO samples

GO solution was produced with natural flake graphite (200 mesh, carbon content: 99.9%; Qingdao Hua Tai Lubrication Technology Company, Shandong, China) as the raw material through the improved Hummers method.¹⁹ The GO solution was then added with 0.01 mol l⁻¹ HCl (AR, Nanjing Chemical Reagent Co., Ltd., Jiangsu, China). It was washed to neutral with deionized water (10 MΩ cm), and GO was obtained after filtering and drying.

2.2 Preparation of reduced graphene oxide (RGO) hydrogel

About 1 mg ml⁻¹ GO aqueous solution was prepared using a certain amount of the abovementioned GO. This solution was uniformly dispersed after 7 h of through ultrasonication. Subsequently, the GO aqueous solution with uniform dispersion was obtained. A certain volume of GO water solution was placed into a 50 ml Teflon polytetrafluoroethylene reaction kettle and heated up to 130 °C (5 °C min⁻¹ heating rate), which was maintained for 14 h. When the temperature dropped to room temperature, different reduction degrees of GO hydrogel were produced.

2.3 Preparation of 3D RGO aerogel

Three types of reduction degrees of GO hydrogel were placed in a refrigerator set at −20 °C to precool for 24 h, and then were subjected to freeze-drying in a vacuum freeze dryer at −50 °C for 48 h to produce different reduction degrees of 3D RGO aerogels. The samples were marked as follows: RGO-130, RGO-160, and RGO-190. The main preparation process of the samples is shown in Fig. 1.

Fig. 1 3D reduction oxidation process of graphene aerogel preparation.

2.4 Preparation of CNT/S active material

CNT and elemental sulfur (S₈) with a 6 : 4 mass ratio were placed in an agate mortar and completely ground for 1 h. They were added into a 50 ml PTFE reaction kettle after stirring well and then were transferred into a glove box. The box was allowed to stand for 0.5 h to exclude residual air and prevent oxidation of S at high temperature. The reaction was then removed from the glove box and placed in a drying oven at 160 °C for 14 h so that S can be fully fused and spread into the CNTs voids. After cooling to room temperature, CNT/S active material was produced and marked as CNT/S.

2.5 Preparation of RGO/CNT/S composites

A certain amount of the CNT/S active material prepared above was dispersed into a certain volume of RGO hydrogel, and then was treated ultrasonically at 70 °C for 7 h to increase dispersion onto the RGO hydrogel. The samples were placed in a refrigerator at −20 °C for 24 h to precool and then were freeze-dried for 36 h to produce RGO/CNT/S composites, which were marked as RGO-130/CNT/S, RGO-160/CNT/S, and RGO-190/CNT/S.

2.6 Material characterization

The samples microstructures were observed through field-emission SEM (HITACHIS-4800). Phase and structure analyses were conducted on a MiniFlex600-type XRD (MiniFlex600) with the following test parameters: scan range, 7–90° and 5.0° min⁻¹ scan rate. FTIR analysis was performed with a Thermo Nicolet NEXUS 670-type FTIR and Raman spectrometer (United States Thermoelectric Company), with a 350–4000 cm⁻¹ scanning range. Samples were produced by the KBr pellet method.

2.7 Battery preparation and electrochemical performance testing

Battery preparation. The active material, super P Li (conductive agent), and PVDF (binder) were uniformly mixed in a 7 : 2 : 1 mass ratio. The mixture was scattered in NMP and then ground for 1 h to produce a well-proportioned serous, which was painted on the aluminum foil current collector. The collector was placed in a vacuum drying oven at 60 °C for 12 h, and then tailored in a positive plate (with a diameter of 14 mm) and placed in a dry glove box filled with argon (contents of O₂ and H₂O were no more than 0.1 ppm). The composite electrode was the positive electrode and the lithium tablet was the negative electrode; microporous polyethylene was the diaphragm material for the Li–S battery, and the electrolyte was 1 mol l⁻¹ Li TFSI/DME + DOL (volume ratio: 1 : 1). All these materials comprise a CR-2025 button cell.

Electrochemical performance testing. The charge–discharge cycle performance tests of batteries were performed using the LAND test instrument at a 27 °C constant temperature with the voltage window set at 1.5–2.8 V. The charge–discharge performance was evaluated with a 1.5–2.8 V voltage window (temperature condition: 28 °C), and the charge–discharge specific capacity calculation was based on the quality of the active substance sulfur. Cyclic voltammetry (CV) and alternating current (AC) impedance of the battery were tested on a CHI750E electrochemical workstation with a 1.0–3.0 V scanning window, a 0.0001 V s⁻¹ scanning speed, a 0.01 Hz to 100 kHz impedance test frequency window and a 5 mV AC signal amplitude window.

3. Results and discussion

3.1 Morphology characterization and phase analysis

The XRD patterns of GO and RGO-X (X = 130, 160, 190) are shown in Fig. 2. A strong diffraction peak exists near 2T = 13° with the GO map, which is a GO typical characteristic diffraction peak.²⁰ Compared with the GO and RGO-X, the high RGO-X diffraction peaks of the spectral lines appear at a slightly offset angle, and the positions of the three peaks after shifting are around 2T = 25°. The intensity of the diffraction peak of RGO-130 near 2T = 25° was weakened and gradually enhanced with increasing reduction temperature, whereas the half peak width and interplanar spacing (d) gradually decreased. The corresponding interplanar spacing (d) with GO and RGO-X are listed in Table 1. The data showed that with increasing hydrothermal reduction temperature, peeling occurs between the layers of RGO, leading to the interlayer spacing (d) being decreased and become closer to the ordered graphite interlayer spacing (d = 0.335 nm), so that the overall structure becomes more similar to ordered graphite.

Fig. 2 GO and RGO-X (X = 130, 160, 190) XRD spectra.

Table 1 GO and RGO-X (X = 130, 160, 190) interplanar spacing

	2T (°)	d (nm)
GO	11.7	0.715
RGO-130	24.7	0.347
RGO-160	25.0	0.343
RGO-190	25.2	0.341

---

FTIR spectra of GO and RGO-X (X = 130, 160, 190) are shown in Fig. 3. GO contains abundant hydroxyls, carboxyls, carbonyls, epoxy groups, and other functional groups at the 1051, 1227, 1632, 1731, 3127, and 3422 cm⁻¹ positions, which correspond to the C–OH stretching vibration peak, C–O–C stretching vibration peak, β_OH, C=O stretching vibration bending vibration peak, V′_OH (related to the formation of hydrogen bond), and V_OH stretching vibration peak, respectively. Interaction between these functional groups overcame the π–π van der Waals force, which led to satisfactory dispersion in an aqueous solution.^21–23 Compared with GO, the intensities of C–OH, C–O–C, C=O, V_OH stretching vibration peaks, and V′_OH stretching vibration peak after hydrothermal reaction, which are related to the formation of a hydrogen bond in the 3D RGO, were significantly decreased. Higher reduction temperature results in more evidently decreased stretching vibration peak of each functional group. This phenomenon rendered the carbon atoms structure to be similar to the initial graphite²⁴ and entirely enhanced the electrical conductivity.

Fig. 3 FTIR spectra of GO and RGO-X (X = 130, 160, 190).

GO and RGO-X (X = 130, 160, 190) Raman spectra are shown in Fig. 4, and it can be seen that all the samples produced Raman characteristic peaks at 1353 cm⁻¹ and 1593 cm⁻¹, respectively, corresponding to D and G bands. The D and G band intensity ratio (I_D/I_G) determined the degree of GO that was reduced; the greater the I_D/I_G value, the lower the degree of order of RGO, and the degree of order may be caused by sample defects.²⁵ The Raman spectra of all samples were fitted with the fit Peak software. After taking the integral, I_D/I_G values before and after hydrothermal reduction were obtained. At the same time, the conductivity was tested by the four probe method before and after reduction; the obtained electrode conductivity (κ) values, I_D/I_G values, and κ values are listed in Table 2. The data in the table show that with the increase of hydrothermal temperature, the value of I_D/I_G and κ increased. The increased I_D/I_G value was caused by an increased quantity of sample defects. The increased κ value was due to the functionalized conjugate being weakened in RGO, which is in agreement with the FTIR results in Fig. 3.

Fig. 4 Raman spectra of GO and RGO-X (X = 130, 160, 190).

Table 2 I_D/I_G values and the conductivity for the electrode

	GO	RGO-130	RGO-160	RGO-190
ID/IG	1.27	1.46	1.57	1.63
κ (S cm^−1)	4.976481 × 10^−3	2.48824 × 10^−2	2.9858886 × 10^−2	3.4835367 × 10^−2

---

SEM images of GO and different reduction degrees of RGO-X (X = 130, 160, 190) gel are shown in Fig. 5. As can be seen from Fig. 5a, deep rolling folds on the GO surface and interlayer reunion phenomenon were observed, but the overall appearance had no fixed structure. Fig. 5b shows that at a 130 °C reduction temperature, the GO lamella structure was faintly visible, and the interlayer reunion phenomenon evidently decreased, but the overall appearance had no fixed structure.²⁶ At a 160 °C reduction temperature, the GO lamellar structure was more apparent, and the interlayer reunion phenomenon largely disappeared; a 3D structure was also present, as shown in Fig. 5c. In Fig. 5d, at a 190 °C reduction temperature, the GO lamella structure was more apparent, and the samples maintained a 3D network structure.

Fig. 5 SEM images of GO and RGO-X (X = 130, 160, 190) (a) GO, (b) RGO-130, (c) RGO-160 (10k), (d) RGO-190 (10k), (e) RGO-160 (20k), (f) RGO-190 (20k).

3.2 Electrochemical properties

Fig. 6a and b show the first and third CV graphs of the RGO-X/CNT/S composite electrode, respectively, and Fig. 6c presents the first and third CV curve graph of the CNT/S composite electrode. All the electrodes had two reduction peaks and an oxidation peak, which is typical electrochemical behavior for lithium sulfur batteries. After the first loop, the two reduction peaks for the CNT/S electrode and RGO-X/CNT/S electrode were both around 2.0 V (S₄²⁻ → S²⁻) and 2.25 V (S₈ → S₄²⁻); the positions of the reduction peaks that correspond to RGO-130/CNT/S and RGO-190/CNT/S were slightly higher than the RGO-160/CNT/S electrode. This result is attributed to the S₈ viscosity, which is lowest at the reduction temperature of 160 °C; thus, the cladding structure cannot be formed well with the 3D RGO gel. After three loops, the reduction peaks of RGO-X/CNT/S composite electrode overlapped, which indicates that the number of oxygen-containing functional groups decreased under the high reduction temperature, which caused the increase of the overall conductivity and the rate of electron transport in the electrode reaction, thereby increasing the S₈ active material utilization. In addition, after the first cycle, the SEI films of the three composite electrodes were basically formed, which also rendered the electrode structure stable. In Fig. 6c, the CNT/S electrode without RGO gel could effectively inhibit the “shuttle” effect,^7,11,12 resulting in loss of many active materials. This phenomenon led to an unchanged reduction peak position of the electrode after three loops. The abovementioned CV analysis shows that the RGO-190/CNT/S composite electrode had better electrochemical behavior at 190 °C reduction temperature.

Fig. 6 Cyclic voltammetry (CV) curves of different samples (a) RGO-X/CNT/S (1st), (b) RGO-X/CNT/S (3rd), and (c) CNT/S.

The first charge and discharge performance of different electrodes at a 200 mA g⁻¹ current density are shown in Fig. 7. Each electrode has two discharge platforms; this finding is in agreement with the CV curves in Fig. 6, which is a typical discharge platform curve for the Li–S battery. In Fig. 7b, the first specific discharge capacity of the CNT/S electrode was 907.1 mA h g⁻¹ at a 200 mA g⁻¹ current density, whereas in Fig. 7a, the first specific discharge capacities of RGO-X (130, 160, 190)/CNT/S increased to 1081.7, 1181, and 1297.4 mA h g⁻¹, respectively, and the first Coulomb efficiencies were 83.8%, 91.7%, and 91.2%. The batteries showed a better charging–discharging retention rate and repeated utilization of the active materials. On the one hand, this result is due to the RGO gel conductivity. The conductivity enhanced with increasing reduction temperature, speeding up the electronic conductivity of the composite electrode, thus more active materials became involved in the electrode reaction. On the other hand, the 3D network structure of RGO gel can easily fix active materials, so the “shuttle” effect was suppressed to a certain degree and a loss of active materials in the electrode reaction was reduced.

Fig. 7 (a) RGO-X/CNT/S and (b) CNT/S cycle discharge/charge voltage profiles at a 200 mA g⁻¹ current density.

Fig. 8 shows the circulating performance of different composite electrodes at a 400 mA g⁻¹ current density. The first specific discharge capacity of the CNT/S electrode was 984 mA h g⁻¹ and became 679.1 mA h g⁻¹ after 50 loops. The capacity attenuation was extremely severe, the retention rate was only 69%, and the Coulomb efficiency remained at around 80%. The discharge curves of RGO-X (130, 160, 190)/CNT/S electrodes were smooth, and their specific capacity was larger than the CNT/S electrodes. The capacity retention ratios of the composited RGO were 81.4%, 82.9%, and 84.6%, and the Coulomb efficiencies were maintained at more than 90%. RGO gel greatly improved the electrode cycling performance, which was mainly due to the RGO gel with a strong physical adsorption function to fix the active substances effectively and prevent excessive dissolution of polysulfide lithium in the electrolyte. The RGO gel was slightly different at different reduction temperatures. The data showed that higher reduction temperature resulted in more superior RGO/CNT/S electrode performance. This superior performance is due to the conductivity of the composite materials, which was further enhanced with increasing the reduction temperature, causing the kinetic rate to accelerate the electrochemical reaction. The RGO gel showed an evident layer structure with the high reduction temperature, which made it easier for active materials to disperse evenly on the RGO gel surface and then improve the active material utilization.

Fig. 8 Cycle performance and Coulomb efficiency of different composites at a 400 mA g⁻¹ current density.

Fig. 9 presents a comparison of the electrodes rate capacities. All the electrodes were tested with the following current densities: 100 mA g⁻¹ → 200 mA g⁻¹ → 400 mA g⁻¹ → 800 mA g⁻¹ → 100 mA g⁻¹. When the current density was 100 mA g⁻¹, the discharge capacity of CNT/S was 980 mA h g⁻¹. After 50 laps, the reversible specific capacity was 384 mA h g⁻¹, and the capacity retention rate was only 39.2%. The CNT/S electrode exhibited a poor reversibility under high current density conditions. The reversible specific capacity of RGO-X (X = 130, 160, 190)/CNT/S electrode was 689, 846, and 934 mA h g⁻¹, respectively, after 50 cycles. The capacity retention rates were 60%, 66.8%, and 68%, respectively, which showed good reversibility. This finding is due to the 3D network structure of RGO gel, which could effectively fix active substances. This structure can also be used as an elastic buffer and weaken the influence of the volume effect on the electrode structure. The data also showed that the ratio properties of the electrode rose with increasing reduction temperature; the best rate performance of the electrode was observed at 190 °C reduction temperature. This was due to the increase in reduction temperature resulting in an increased RGO gel layer of a specific surface area;²⁴ this phenomenon makes it easy for the active substance to be infiltrated by the electrolyte so that more lithium sulfide is reduced.

Fig. 9 Rate performance of CNT/S and RGO-X/CNT/S cathode.

The electrochemical impedance spectra (EIS) of different electrodes before and after the cycle are shown in Fig. 10. Cycle conditions were to charge and discharge ten times with a 100 mA g⁻¹ current density. The EIS curve was composed of three parts. The curve and real axis Z′ intercepts were the intrinsic impedance of batteries (active material resistance, electrolyte resistance, and current collector interface resistance). The marker represents R₁, and the half arc diameter of the high frequency region corresponds to the charge transfer resistance (R₂). The low frequency region for a horizontal axis and a 45 degree diagonal (Warburg impedance, marked as Z_W) correspond to the Li⁺ diffusion impedance in the electrolyte.^27,28 ZView software was used to fit the electrochemical impedance spectra of different electrodes before and after cycling. R₁, R₂, and Z_W values of different electrodes were obtained and are listed in Table 3. Table 3 shows that after ten cycles, the R₁, R₂, and Z_W resistance of all the electrodes were lower than before, and this is consistent with the impedance spectra. This finding is attributed to the ten cycles of the electrodes, which resulted in active substance infiltration by electrolytes and reduction of Li⁺ diffusion resistance in electrolytes. The data also showed that the impedance value of the RGO-X/CNT/S electrode was less than that of the CNT/S electrode. Moreover, the resistance decreased gradually with increasing reduction temperature. The electrodes showed the best electrochemical performance at a 190 °C reduction temperature. This performance is due to higher reduction temperature, which results in better conductive properties of RGO gel and promotes the rapid transmission of electrons in composite materials. In addition, the 3D mesh structure is convenient for rapid electrolyte diffusion and reduced the electrolyte ions resistance in the penetration process.

Fig. 10 Impedance plots before and after the cycle for CNT/S and RGO-X/CNT/S cathode.

Table 3 Electrode kinetic parameters

Sample	Stage	R1 (Ω)	R2 (Ω)	ZW (Ω)
CNT/S	Cycle before	22.96	107.4	74.4
	After 10 cycles	26.21	88.2	62.23
RGO-130/CNT/S	Cycle before	8.631	75.99	42.68
	After 10 cycles	8.282	75.12	34.22
RGO-160/CNT/S	Cycle before	4.748	72.57	44.35
	After 10 cycles	3.616	50.72	30.79
RGO-190/CNT/S	Cycle before	3.669	45.39	20.83
	After 10 cycles	3.414	22.78	19.72

---

4. Conclusion

(1) Hydrothermal reduction combined with a freeze-drying method could successfully prepare a 3D RGO aerogel structure. When the temperature increased to 160 °C, the degree of order was enhanced and the significant lamellar structure of the RGO gel showed a 3D network structure.

(2) When RGO/CNT/S was used as a Li–S battery cathode material, the battery exhibited good electrochemical performance compared with the CNT/S electrode. Higher reduction temperature resulted in a greater degree of GO reduction, better 3D RGO gel performance, and more suitable elastic buffer to adapt to changes in electrode volume.

Conflict of interests

The authors declare that there is no conflict of interests regarding the publication of this study.

Acknowledgements

The authors are grateful to the National Natural Science Foundation of China (Grant No. 11364011), the Guangxi Natural Science Foundation (Grant No. 2015GXNSFAA139004) and the Innovation Project of Guangxi Graduate Education (Grant No. YCSZ2015164).

References

1. N. Jayaprakash, J. Shen, S. S. Moganty, A. Corona and L. A. Archer, Angew. Chem., Int. Ed., 2011, 123, 6026–6030.
2. C. Wang, J. J. Chen, Y. N. Shi, M. S. Zheng and Q. F. Dong, Electrochim. Acta, 2010, 55, 7010–7015.
3. Q. Wang, J. Pan, M. Li, Y. Luo, H. Wu, L. Zhong and G. H. Li, J. Mater. Sci. Technol., 2015, 31, 630–633.
4. A. K. Geim and K. S. Novoselov, Nat. Mater., 2007, 6, 183–191.
5. N. D. Mermint, Phys. Rev., 1968, 176, 250–254.
6. X. Ling and J. Zhang, Acta Phys.–Chim. Sin., 2012, 28, 2355–2362.
7. X. L. Ji, K. T. Lee and L. F. Nazar, Nat. Mater., 2009, 8, 500–506.
8. L. F. Xiao, Y. L. Cao, J. Xiao, B. Schwenzer, M. H. Engelhard, L. V. saraf, Z. Nie, G. J. Exarhos and J. Liu, Adv. Mater., 2012, 24, 1176–1181.
9. X. L. Ji and L. F. Nazar, J. Mater. Chem., 2010, 20, 9821–9826.
10. S. C. Han, M. S. Song, H. Lee, H. S. Kim, H. J. Ahn and J. Y. Lee, J. Electrochem. Soc., 2003, 150, A889–A893.
11. W. Zheng, Y. W. Liu, X. G. Hu and C. F. Zhang, Electrochim. Acta, 2006, 51, 1330–1335.
12. L. X. Yuan, H. P. Yuan, X. P. Qiu, L. Q. Chen and W. T. Zhu, J. Power Sources, 2009, 189, 1141–1446.
13. J. J. Chen, X. Jia, Q. J. She, C. Wang, Q. Zhang, M. S. Zheng and Q. F. Dong, Electrochim. Acta, 2010, 55, 8062–8066.
14. M. W. Xu, C. J. Cheng, Q. Q. Sun, S. J. Bao, Y. B. Niu, H. He, Y. T. Li and J. Song, RSC Adv., 2015, 5, 56686.
15. G. Y. Xu, B. Ding, P. Nie, H. J. Luo and X. G. Zhang, Acta Phys.–Chim. Sin., 2013, 29, 546–552.
16. S. Aziz, J. Q. Zhao, C. Cain and Y. Wang, J. Mater. Sci. Technol., 2014, 30, 427–433.
17. L. P. Xue, M. B. Zheng, C. F. Shen, H. L. Lu, N. W. Li, L. J. Pan and J. M. Cao, Chin. J. Inorg. Chem., 2010, 26, 1375–1381.
18. M. Ciszewski, A. Mianowski and G. Nawrat, J. Mater. Sci.: Mater. Electron., 2013, 24, 3382–3386.
19. W. S. Hummers and R. E. Offeman, J. Am. Chem. Soc., 1958, 80, 1339.
20. D. R. Dreyer, S. Park, C. W. Bielawski and R. S. Ruoff, Chem. Soc. Rev., 2010, 39, 228–240.
21. H. K. Jeong, Y. P. Lee, R. J. W. E. Lahaye, M. H. Park, K. H. An, I. J. Kim, C. W. Yang, C. Y. Park, R. S. Ruoff and Y. H. Lee, J. Am. Chem. Soc., 2008, 130, 1362–1366.
22. W. Cai, R. D. Piner, F. J. Stadermann, S. Park, M. A. Shaibat, Y. Ishii, D. Yang, A. Velamakanni, S. J. An, M. Stoller, J. An, D. Chen and R. S. Ruoff, Science, 2008, 321, 1815–1817.
23. A. Barinov, O. B. Malcioglu, S. Fabris, T. Sun, L. Gregoratti, M. Dalmiglio and M. Kiskinova, J. Phys. Chem. C, 2009, 113, 9009–9013.
24. S. Stankovich, D. A. Dikin, R. D. Piner, K. A. Kohlhaas, A. Kleinhammes, Y. Y. Jia, Y. Wu, S. T. Nguyenb and R. S. Ruoff, Carbon, 2007, 45, 1558–1565.
25. C. Thomsen and S. Reich, Phys. Rev. Lett., 2000, 85, 5214.
26. H. Bi, K. Yin, X. Xie, Y. Zhou, N. Wan, F. Xu, F. Banhart, L. Sun and R. S. Ruoff, Adv. Mater., 2012, 24, 5124–5129.
27. L. X. Yuan, X. P. Qiu, L. Q. Chen and W. T. Zhu, J. Power Sources, 2009, 189, 127–132.
28. K. L. Zhu, J. H. Tian, L. P. Liu, Q. W. Tang, X. M. Yu, Y. M. Zhu and Z. Q. Shan, RSC Adv., 2013, 3, 13149–13155.

---

Footnote

† Hongzhe Wang contributed with Jianrong Xiao, and they are both the first author.

---