Nitrogen-doped holey graphene foams for high-performance lithium storage

Abstract

Although lithium ion batteries have gained commercial success in abundant portable devices, they still lack suitable electrode materials with high capacities and rate capabilities critical for broad applications. Herein, we demonstrate a facile and scalable approach toward nitrogen-doped holey graphene foams (NHGFs) via simultaneous etching and assembling of graphene oxide and subsequent annealing treatments. The resultant NHGFs possess some unique structural features such as hierarchically interconnected pores in the foams and a nitrogen-doped holey structure in the basal plane of the graphene walls, which facilitate the fast diffusion of both lithium and electrons through the NHGFs and meanwhile significantly improve electrochemical activity. As a result, the NHGFs exhibit excellent electrochemical performances for lithium storage, including a high reversible capacity (1194 mA h g⁻¹), good high-rate capability (420 mA h g⁻¹ at 5C) and long cycling stability with negligible capacity loss after 1000 continuous cycles.

---

Introduction

Graphene nanosheets have attracted great interest in the development of new anode materials for lithium ion batteries owing to their fascinating physical and chemical properties, such as high electrical conductivity, large surface area, good flexibility and high theoretical capacity (744 mA h g⁻¹).^1–3 However, due to the strong van der Waals interaction and large surface area, graphene nanosheets are prone to restacking and aggregating during the electrode fabrication process.^4–6 Moreover, the cross-plane diffusivity of lithium ions in graphene nanosheets is very low, leading to the predominant diffusion of lithium ions at the edges of the restacked graphene nanosheets,⁷ despite the high in-plane diffusivity of lithium ions (∼10⁻⁸ cm² s⁻¹).⁸ This severely hampers the electrochemical performance of graphene for lithium storage, especially at high current rates.^9,10

To circumvent the above obstacles of graphene nanosheets for lithium storage, several effective strategies including introducing defects, holes and heteroatoms (N, B, S)^11–13 into the basal plane of graphene, and assembling graphene into three-dimensional (3D) foams or networks have been recently developed.¹⁴ In the former, the presence of defects or holes in the graphene basal plane, also called holey graphene or graphene nanomesh,¹⁵ can be beneficial to the easy access of electrolytes and rapid diffusion of lithium ions.^8,9 Introducing heteroatoms into graphene nanosheets can greatly increase the adsorption energies and correspondingly decrease their energy barriers for lithium ion penetration, owing to the hybridization of the heteroatom lone pair electrons with the graphene aromatic system.⁵ Moreover, it has been demonstrated theoretically and experimentally that all the introductions of defects, holes and heteroatoms into graphene could improve their electrochemical active sites for lithium storage compared to pristine graphene nanosheets.^16,17 In the latter, different from the above chemical engineering in the basal plane of graphene, the rational arrangement of individual graphene nanosheets into 3D foams or networks, in which physical spaces (multi-sized pores) are created between the intact nanosheets, can also reduce significantly the diffusion length of lithium ion migration during the charge–discharge process and allow more graphene sheets to be available for lithium storage.^5,18,19 Although a separated strategy has been carefully investigated with the aid of various approaches such as chemical vapor deposition,^20,21 template-assisted lithography²² and hydrothermal methods²³ in recent years, facilely integrating several strategies to maximize their merits into one graphene-based material and realizing excellent electrochemical performances for lithium storage remain a great challenge.^24–26

Herein, we develop a rational approach to the facile and scalable fabrication of nitrogen-doped holey graphene foams (NHGFs) via a hydrothermal treatment of graphene oxide with hydrogen peroxide and subsequently annealing under ammonia gas. In the resultant NHGFs, the macroporous networks are constructed from abundant nitrogen-doped graphene nanosheets with in-plane holes, which can provide large amounts of cross-plane ion diffusion channels and meanwhile decrease the energy barriers for lithium ion penetration. Combining these with the 3D graphitic networks that maintain high electrical conductivity and structural integrity, a novel graphene-based electrode material successfully integrating the merits of several strategies is generated, which exhibits a high reversible capacity of 1194 mA h g⁻¹, good high-rate capability (420 mA h g⁻¹ at 5C) and long cycle stability for lithium storage.

Experimental

Synthesis of nitrogen-doped holey graphene foams

In a typical experiment, graphene oxide was first synthesized from natural graphite flakes using a modified Hummers method,²⁷ and then diluted to a concentration of 2 mg mL⁻¹ with milli-Q water. 10 mL of the as-synthesized graphene oxide aqueous dispersion was then mixed with 1 mL of hydrogen peroxide (3 wt%) by sonication for 10 min. The resulting mixture was then sealed in a Teflon-lined autoclave and hydrothermally treated at 180 °C for 6 h, in which graphene oxide nanosheets were chemically etched to holey graphene oxide and simultaneously assembled into a 3D holey graphene oxide foam. Finally, the as-prepared foam was freeze- or critical point-dried to preserve the 3D network formed during the synthesis process. Finally, the 3D holey graphene oxide foams were annealed under a mixed atmosphere of ammonia gas (150 sccm) and nitrogen gas (450 sccm) at 1000 °C for 45 min, generating nitrogen-doped holey graphene foams, denoted as NHGFs. For comparison, slightly etched nitrogen-doped graphene foams (se-NGFs) and non-etched nitrogen-doped graphene foams (ne-NGFs) were also fabricated under similar synthesis conditions except for adding 1 mL of 0.3 wt% H₂O₂ and not adding H₂O₂, respectively. Meanwhile, NHGFs-800 and NHGFs-900 were fabricated by adjusting the annealing temperature to 800 °C and 900 °C, respectively. A pure holey graphene foam (denoted as HGF) was also fabricated via the direct annealing of graphene oxide foam in N₂ at 1000 °C.

Assembly of lithium ion cells

The working electrodes were prepared by spreading a slurry of the active materials (80 wt%), acetylene black (10 wt%) and the polyvinylidene fluoride (10 wt%) binder onto copper foil with N-methyl-2-pyrrolidone (NMP) as the mixture solvent. The as-prepared electrodes were then dried at 120 °C under vacuum for 12 h. To test the lithium ion battery assembly, metallic lithium foil was used as the counter-electrode, 1 M LiPF₆ in ethylene carbonate (EC)/dimethyl carbonate (DMC)/diethyl carbonate (DEC) (1 : 1 : 1 by volume) was used as the electrolyte, and a glassy fiber film as the separator. All of the coin-type (CR2032) cells were assembled in an argon-filled glove box (H₂O, O₂ level < 0.1 ppm).

Characterizations

The morphology and microstructure of the NHGFs were investigated using transmission electron microscopy (TEM: JEOL-2100F), field-emission scanning electron microscopy (FE-SEM: JEOL-7500) and X-ray diffraction (XRD, Rigaku D/max2500PC) using Cu Kα radiation over the range of 10–60° (two-theta) at room temperature (XRD patterns shown in Fig. S1†). X-ray photoelectron spectroscopy (XPS: Thermo Electron Corporation ESCALAB 250) measurements were obtained using an XPS spectrometer with monochromatized Al Kα radiation and a 200 eV pass energy with a 30 eV step, over a sample area of 500 μm × 500 μm. Nitrogen sorption isotherms, BET surface area, porous volume and pore size distribution were measured at 77 K with a Quadrasorb (Quantachrome, USA).

The charge–discharge measurements were carried out on a Land CT2001A battery test system over a voltage range of 0.01–3 V at different current rates (theoretical capacity of graphene: 1C = 744 mA h g⁻¹) at room temperature. Electrochemical impedance spectroscopy (EIS) measurements of the electrodes were carried out on an electrochemical workstation (Autolab). The impedance spectra were recorded by applying a sine wave with amplitude of 5.0 mV over the frequency range from 100 kHz to 0.01 Hz.

Results and discussion

As illustrated in Fig. 1, the overall synthesis procedure of the nitrogen-doped holey graphene foams involves two steps: the first is the hydrothermal treatment of graphene oxide with hydrogen peroxide at 180 °C, in which graphene oxide nanosheets were chemically etched to holey graphene oxide and simultaneously assembled into a 3D foam; the second is the annealing of the as-prepared 3D holey graphene oxide foams at different temperatures under ammonia gas. The morphology and structure of the resulting NHGFs were elucidated using SEM, TEM, and HRTEM measurements (Fig. 1 and 2). Apparently, a 3D porous configuration with interconnected pores ranging from sub-micrometers to several micrometers is clearly visible (Fig. 1), which is similar to those reported for graphene and graphene oxide hydrogels.^28,29 A typical TEM image (Fig. 2a) discloses that the nanosheet walls of these pores are highly transparent to electron beams, demonstrating their ultrathin nature. Close inspection further reveals the holey structure with a uniform pore size of ca. 2 nm in the nanosheets (Fig. 2b). This clearly demonstrates that some oxidants like hydrogen peroxide could effectively make holes in the basal plane of a graphene nanosheet during the hydrothermal process as we expected. The elemental mapping images further reveal that carbon, nitrogen and oxygen atoms are homogeneously dispersed in the walls of the NHGFs (Fig. 2c–e).

Fig. 1 Schematic illustration of the fabrication of the nitrogen-doped holey graphene foams (NHGFs). It mainly involves two steps: (1) hydrothermal treatment of graphene oxide with hydrogen peroxide at 180 °C, in which graphene oxide nanosheets were chemically etched to holey graphene oxide and simultaneously assembled into a 3D foam; (2) annealing of the holey graphene oxide foam at high temperatures under ammonia gas. The typical SEM image of the NHGFs (bottom left) reveals a 3D porous configuration with interconnected pores ranging from sub-micrometers to several micrometers.

Fig. 2 Morphological and structural analysis of the nitrogen-doped holey graphene foams. (a) Typical TEM and (b) HRTEM images of the NHGFs, revealing the holey structure with a uniform pore size of ca. 2 nm in the nanosheets. Elemental mapping images of (c) nitrogen, (d) carbon and (e) oxygen, indicating the homogeneous dispersion of N, C and O in the holey graphene.

The porous nature of the NHGFs was further confirmed by nitrogen physisorption measurements. Their adsorption–desorption isotherms exhibit a typical IV hysteresis loop (Fig. 3a), which is characteristic of mesopores with different pore sizes. Barrett–Joyner–Halenda (BJH) calculations disclose that the pore size distribution is in the range of 5–15 nm (Fig. 3b), except for the open macropores identified from the SEM images (Fig. 2a). A high BET surface area of up to 623 m² g⁻¹ is observed for the NHGFs, which is much higher than those for the slightly etched nitrogen-doped graphene foam (507 m² g⁻¹) and pure holey graphene foam (131 m² g⁻¹). This clearly demonstrates that the BET surface areas of the NHGFs are strongly dependent on the concentrations of hydrogen peroxide during our fabrication process. Because the BET surface area mainly originated from the mesopores and micropores, we further conducted a methylene blue (MB) adsorption investigation to evaluate the macroporous surface area of the NHGFs²⁹. According to this investigation (Fig. S2†),³⁰ the surface area of the NHGFs is 863 m² g⁻¹, which is also much higher than those of the slightly etched nitrogen-doped graphene foam (495 m² g⁻¹) and pure holey graphene foam (206 m² g⁻¹), which is consistent with the above BET analysis. In combination with the analysis based on the SEM and TEM images, it is reasonable to believe that the high surface area of the 3D NHGFs is attributed to both the holy graphene nanosheets and their constructed porous structures. Thus, this unique hierarchically porous structure should greatly facilitate the access of electrolytes and the fast diffusion of lithium ions during lithium storage.

Fig. 3 (a) Nitrogen adsorption/desorption isotherms of the NHGFs, se-NGFs, ne-NGFs and NGFs, confirming the porous structure with different pore sizes, (b) BJH calculation discloses that the pore sizes in the 3D nitrogen-doped holey graphene architectures are mainly in the range of 5–20 nm.

To identify the chemical composition of the 3D NHGFs, XPS measurements and elemental analysis were carried out. As shown in Fig. 4, C1s, N1s and O1s peaks are observed clearly in the survey curves. In the case of the NHGFs, an obvious N1s signal is evident, and the corresponding N content was measured to be 3.4 at%, close to that analyzed using an Organic Elements Analyzer (3.6 at%). This value is higher than those of the slightly etched nitrogen-doped graphene foam (3.0 at%) and non-etched nitrogen-doped foam (1.3 at%), as well as being comparable to those reported for N-doped graphene foams or networks (2–4 at%).²⁹ The high nitrogen content in the NHGFs should arise from more accessible heteroatom doping at the holes and defects of graphene compared with that of pristine graphene. The complex N1s spectrum can be fitted to three different peaks (Fig. 4c), located at the binding energies of 398.2 eV, 399.7 eV, and 401.3 eV, corresponding to the pyridinic, pyrrolic and graphitic nitrogens, respectively. The corresponding O1s spectrum (Fig. 4d) shows O–C=O, C=O and C–O peaks at 531.9 eV, 533.0 eV and 534 eV, respectively.

Fig. 4 (a) XPS survey spectra of the NHGFs, se-NGFs, ne-NGFs and NGFs. High-resolution XPS spectra of (a) survey, (b) C1s, (c) N1s and (d) O1s. These N1s peaks are fitted to three energy components centered at around 398.2 eV, 399.7 eV, and 401.3 eV, corresponding to pyridinic-N (N1), pyrrolic-N (N2), and graphitic-N (N3), respectively.

The electrochemical performances of the 3D NHGFs were systematically evaluated using galvanostatic discharge (lithium insertion)–charge (lithium extraction) measurements at various rates (1C = 744 mA h g⁻¹) over a voltage range from 0.01 to 3.0 V. It is striking that a very high capacity of 2847 mA h g⁻¹ is achieved during the initial discharge process at a current rate of 0.1C (Fig. 5a), corresponding to a lithium insertion coefficient of 7.7 in Li_7.7C₆. This value is much higher than the theoretical capacity of graphene (744 mA h g⁻¹) with a maximum coefficient of 2 (Li₂C₆),³⁰ suggesting the existence of extra lithium storage sites in our 3D NHGFs. The detailed discharge–charge curve analysis (Fig. 5a and b) reveals that two domains including an approximate plateau and a slope mainly contribute to the overall capacity of the 3D NHGFs during the discharge process. Obviously, the plateau at ∼0.9 V is similar to that reported for carbonaceous materials,³¹ characteristic of the formation of a solid electrolyte interface (SEI) layer on the surface of the NHGF electrode. The typical cyclic voltammograms of the NHGFs, se-NGFs, ne-NGFs and NGFs are compared in Fig. S3 (see ESI†). As shown in Fig. S3a,† in the case of the NHGFs, there is a main reduction peak centred at 0.8 V in the first cathodic scanning process, corresponding to the discharge flat plateau at around 0.85 V in the first discharge curve, which disappears in the second cycle. This should be ascribed to the formation of solid electrolyte interphase (SEI) layers on the surface of the NHGFs. The other cathodic peak ranged from 0.5 V to 0.05 V, which is clearly different from that of graphite with a narrow peak at 0.05 V. It is reasonable to believe that the slope below 0.5 V in the discharge curves should be attributed to lithium adsorption onto both sides of graphene owing to the large surface area, which is similar to the reported graphene nanosheets. The slope also indicates that the severe re-stacking and aggregation of graphene nanosheets were effectively avoided during our fabrication process since the re-stacked graphene sheets commonly have a long plateau at ∼0.15 V for lithium storage. The first charge curve further shows that the reversible capacity of the NHGFs is mainly in the voltage range of above 0.2 V.³² Moreover, this phenomenon with high reversible capacity at above 0.2 V is similar to those reported for holey graphene or graphene nanomesh and other nanostructured carbons with high capacities, in which additional micro and mesoporous lithium storage mechanisms exist.³³ It is worth noting that the foams’ reversible capacity exceeds the theoretical capacity of graphene, 744 mA h g⁻¹, which could be ascribed to the synergistic effect of the porous lithium storage mechanism and nitrogen-doping, which improved the BET surface area of the foams. Thereby, such a porous lithium storage mechanism combined with the common surface storage of graphene gives rise to a very high capacity of 1194 mA h g⁻¹ for the NHGFs during the initial charge process.

Fig. 5 Electrochemical performances of the 3D nitrogen-doped holey graphene foams. (a) Representative discharge–charge curves of the NHGFs at a current rate of 0.1C, (b) representative discharge–charge curves of the ne-NGFs at 0.1C. (c) Cycling performances of the NHGFs, se-NGFs, ne-NGFs and NGFs at 0.1C. (d) Rate capacities of the NHGFs, se-NGFs, ne-NGFs and NGFs at different current rates from 0.25C to 5C. (e) Cycling performance of the NHGFs at 5C for 1000 cycles.

The rate capabilities and cycling performances of the NHGFs were investigated via galvanostatic discharge–charge measurements at various current rates from 0.1C to 5C (Fig. 5c and d). For comparison, the slightly etched and non-etched nitrogen-doped graphene foams as well as the pure holey graphene foam (the rate performances of NHGFs-800 and NHGFs-900 are shown in Fig. S6†) were also tested under the same electrochemical conditions. Remarkably, in the case of the 3D NHGFs, the reversible capacities are retained at 1150 and 900 mA h g⁻¹ after 50 cycles at 0.1C and 0.25C, respectively, and the coulombic efficiencies (calculated from the discharge and charge capacities) approach almost 100% (Fig. S4 in ESI†). In contrast, the capacities of the slightly etched and non-etched nitrogen-doped graphene foams rapidly decrease to 610 and 320 mA h g⁻¹ even at the lowest current rate of 0.1C, although these values are higher than that of the pure holey graphene foam (450 mA h g⁻¹). Such a prominent difference strongly validates the efficiency of our protocol to improve the electrochemical performance of graphene by incorporating the strategies of doping nitrogen and creating holes or defects. More importantly, the 3D NHGFs exhibit an excellent cycling performance at high current rates, indicating the ultrafast diffusion of lithium ions in the electrode material owing to the short diffusion path length and the stable structure. Apparently, such high-rate performances of the 3D NHGF nanosheets are superior to most of those reported for previous graphene-based materials including graphene nanosheets, holey graphene, graphene nanomesh, nitrogen-doped hollow graphene spheres and 3D graphene foams obtained under similar testing conditions.^34,35

In addition, from Fig. 5d, it can be seen that the capacity has a tendency to slightly decrease at a high current rate of 5C, whereas it increases steadily at the same rate during long charge–discharge processes (Fig. 5e). On close inspection of the cycling performance in Fig. 5e, it is visible that the capacity fluctuates and slightly increases during the long testing process. This should be related to two influencing factors: (1) testing room temperature, and (2) the effective availability of active materials at high current rates. During our testing processes, we found that the capacity of a cell is sensitive to the testing room temperature during the day and night; a higher temperature commonly leads to a higher capacity. More importantly, at high current rates, the charge–discharge time is very short, which significantly affects the effective availability of the active materials (less than 100%). Thus, with increasing the number of cycles, the holey structure and nitrogen-doping should be more favorable for the fast diffusion of both lithium ions and electrons, leading to a higher effective availability of the active materials during charge–discharge processes.

To gain further insight into the reason why the 3D NHGFs possess an excellent electrochemical performance, we carried out AC impedance spectra measurements after 50 cycles at the rate of 0.1C. As presented in Fig. 6a (and Fig. S7†), the Nyquist plot of the NHGFs consists of a broad semicircle at high frequencies for the charge-transfer kinetics-controlled region and a straight linear line at low frequencies for the mass transfer-controlled Warburg region (Warburg impedance coefficient, σ_w, calculated in Table S1†).^8,36 The diffusion coefficient of lithium ions in the 3D NHGFs is estimated to be ∼1.4 × 10⁻⁸ cm² s⁻¹ from the impedance spectra (for details, see ESI†). This value is much higher than those of the slightly etched and non-etched nitrogen-doped graphene foams (1.3 × 10⁻⁸ to 5.2 × 10⁻⁹ cm² s⁻¹) and even comparable to that reported for vertically aligned graphene sheets (∼1.5 × 10⁻⁸ cm² s⁻¹),³⁷ demonstrating that the holey structure of graphene and the hierarchically porous structure of the NHGFs can significantly improve the diffusivity of lithium ions in the electrode. This can be further confirmed by the low electrolyte resistance (R_e) of 4.0 ohms (Table S2†) calculated on the basis of a modified Randles equivalent circuit shown in Fig. S5.† Meanwhile, the charge-transfer resistance (R_ct) of the NHGFs is only 26.7 ohms, prominently lower than those of the slightly etched and non-etched nitrogen-doped graphene foams (32.0–45.8 ohms), holey graphene foam (39.4 ohms) and those reported for graphene-based materials (40–70 ohms). These values (R_e, R_f) are slightly decreased when accompanied with a degree of chemical etching by hydrogen peroxide; the charge-transfer resistance (R_ct) is notably downturned, corresponding to a resistance of 26.7 ohms which is lower than the other slightly or non-etched foams, indicating the charger transfer and mass transfer of lithium ions was much more conductive. Such a low charge-transfer resistance indicates the high electrochemical activity of the NHGFs for lithium storage since the exchange current density is inversely proportional to the charge-transfer resistance. Associated with the unique structures of the NHGFs as discussed above, it is reasonable to believe that the pronounced electrochemical performances of the NHGFs can be attributed to their favorable structures for the fast diffusion of both lithium ions and electrons during the charge–discharge processes, successfully satisfying the kinetics requirements of high-power lithium ion batteries.

Fig. 6 (a) Nyquist plots of the NHGF, se-NGF, ne-NGF and NGF electrodes obtained by applying a sine wave with amplitude of 5.0 mV over the frequency range from 100 kHz to 0.01 Hz. (b) Relationship between Z_Re and ω_−1/2 at low frequencies, the slope of each line corresponds to the Warburg impedance coefficient, σ_w.

Conclusions

In summary, we have demonstrated an efficient fabrication of nitrogen-doped holey graphene foams via the hydrothermal treatment of graphene oxide with hydrogen peroxide and subsequent annealing under ammonia gas. The unique structural features of the NHGFs including the hierarchically interconnected pores and nitrogen-doped holey structure in the basal plane of the graphene walls not only allow for the fast diffusion of lithium ions and electrons through the NHGFs, but also render significantly improved electrochemical activities for lithium storage. As a consequence, the NHGFs exhibit superior electrochemical performances in terms of the high reversible capacity, good high-rate capability, and long cyclability when they are used as the anode material for lithium storage. We believe that such a facile and scalable synthesis protocol can be further extended to the fabrication of other heteroatom (B, S, BN) doped holey graphene foams, with wide applications in catalysis, sensors, supercapacitors, batteries, and fuel cells.

Acknowledgements

This work was financially supported by the Scientific Research Foundation for the Returned Overseas Chinese Scholars (No. 37-7290-01), State Education Ministry, the “Recruitment Program of Global Experts”, and the Fundamental Research Funds for the Central Universities (No. 30-4233-01 and 30-4041-02).

References

1. A. K. Geim and K. S. Novoselov, Nat. Mater., 2007, 6, 183–191.
2. D. Chen, L. Tang and J. Li, Chem. Soc. Rev., 2010, 39, 3157–3180.
3. M. M. Thackeray, C. Wolverton and E. D. Isaacs, Energy Environ. Sci., 2012, 5, 7854–7863.
4. Z. Fan, J. Yan, L. Zhi, Q. Zhang, T. Wei, J. Feng, M. Zhang, W. Qian and F. Wei, Adv. Mater., 2010, 22, 3723–3728.
5. Y. Fang, Y. Lv, R. Che, H. Wu, X. Zhang, D. Gu, G. Zheng and D. Zhao, J. Am. Chem. Soc., 2013, 135, 1524–1530.
6. D. Li, M. B. Muller, S. Gilje, R. B. Kaner and G. G. Wallace, Nat. Nanotechnol., 2008, 3, 101–105.
7. Z. Fan, J. Yan, T. Wei, L. Zhi, G. Ning, T. Li and F. Wei, Adv. Funct. Mater., 2011, 21, 2366–2375.
8. Z. Jiang, B. Pei and A. Manthiram, J. Mater. Chem. A, 2013, 1, 7775–7781.
9. W. Yuan, Y. Zhou, Y. Li, C. Li, H. Peng, J. Zhang, Z. Liu, L. Dai and G. Shi, Sci. Rep., 2013, 3, 2248.
10. Y. Zhu, S. Murali, W. Cai, X. Li, J. W. Suk, J. R. Potts and R. S. Ruoff, Adv. Mater., 2010, 22, 3906–3924.
11. D. Yu, L. Wei, W. Jiang, H. Wang, B. Sun, Q. Zhang, K. Goh, R. Si and Y. Chen, Nanoscale, 2013, 5, 3457–3464.
12. X. Zhao, C. M. Hayner, M. C. Kung and H. H. Kung, ACS Nano, 2011, 5, 8739–8749.
13. J. Liang, Y. Jiao, M. Jaroniec and S. Z. Qiao, Angew. Chem., Int. Ed., 2012, 51, 11496–11500.
14. Z. Wu, G. Zhou, L. Yin, W. Ren, F. Li and H. Cheng, Nano Energy, 2012, 1, 107–131.
15. J. Bai, X. Zhong, S. Jiang, Y. Huang and X. Duan, Nat. Nanotechnol., 2010, 5, 190–194.
16. X. Zhao, C. M. Hayner, M. C. Kung and H. H. Kung, Adv. Energy Mater., 2011, 1, 1079–1084.
17. Y. Gong, S. Yang, Z. Liu, L. Ma, R. Vajtai and P. M. Ajayan, Adv. Mater., 2013, 25, 3979–3984.
18. M. D. Levi and D. Aurbach, J. Phys. Chem. B, 1997, 101, 4641–4647.
19. Z. L. Wang, D. Xu, H. G. Wang, Z. Wu and X. B. Zhang, ACS Nano, 2013, 7, 2422–2430.
20. S. C. O’Hern, C. A. Stewart, M. S. H. Boutilier, J. C. Idrobo, S. Bhaviripudi, S. K. Das, J. Kong, T. Laoui, M. Atieh and R. Karnik, ACS Nano, 2012, 6, 10130–10138.
21. Z. J. Fan, J. Yan, T. Wei, G. Q. Ning, L. J. Zhi, J. C. Liu, D. X. Cao, G. L. Wang and F. Wei, ACS Nano, 2011, 5, 2787–2794.
22. S. Yang, L. Zhan, X. Xu, Y. Wang, L. Ling and X. Feng, Adv. Mater., 2013, 25, 2130–2134.
23. Y. Sun, Q. Wu and G. Shi, Energy Environ. Sci., 2011, 4, 1113–1132.
24. S. Yang, X. Feng and K. Müllen, Adv. Mater., 2011, 23, 3575–3579.
25. B. Kang and G. Ceder, Nature, 2009, 458, 190–193.
26. Y. T. Liu, X. D. Zhu, Z. Q. Duan and X. M. Xie, Chem. Commun., 2013, 49, 10305–10307.
27. R. Offeman and W. Hummers, J. Am. Chem. Soc., 1958, 80, 1339.
28. Y. Xu, K. Sheng, C. Li and G. Shi, ACS Nano, 2010, 4, 4324–4330.
29. A. L. M. Reddy, A. Srivastava, S. R. Gowda, H. Gullapalli, M. Dubey and P. M. Ajayan, ACS Nano, 2010, 4, 6337–6342.
30. L. Pan, X. D. Zhu, X. M. Xie and Y. T. Liu, J. Mater. Chem. A, 2015, 3, 2726–2733.
31. Z. Q. Duan, Y. C. Sun, Y. T. Liu, X. M. Xie and X. D. Zhu, RSC Adv., 2014, 4, 41543–41550.
32. L. Pan, X. D. Zhu, X. M. Xie and Y. T. Liu, Adv. Funct. Mater., 2015, 25, 3341–3350.
33. S. L. Kuo, W. R. Liu, C. P. Kuo, N. L. Wu and H. C. Wu, J. Power Sources, 2013, 244, 552–556.
34. Y. Zhao, C. Hu, L. Song, L. Wang, G. Shi, L. Dai and L. Qu, Energy Environ. Sci., 2014, 7, 1913–1918.
35. Y. Zhao, C. Hu, Y. Hu, H. Cheng, G. Shi and L. Qu, Angew. Chem., 2012, 124, 11533–11537.
36. C. Uthaisar and V. Barone, Nano Lett., 2010, 10, 2838–2842.
37. L. Pan, K. X. Wang, X. D. Zhu, X. M. Xie and Y. T. Liu, J. Mater. Chem. A, 2015, 3, 6477–6483.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra19313c

---