Cobalt nanoparticles encapsulated in N-doped graphene nanoshells as an efficient cathode electrocatalyst for a mechanical rechargeable zinc–air battery

Abstract

Air-cathodes with properties of efficiency, durability and low cost are essential for high performance metal–air batteries and fuel cells for practical applications. In this study, non-precious metal ORR electrocatalysts derived by the encapsulation of Co nanoparticles in N-doped graphene nanoshells were synthesized by a typical one-step pyrolysis process. Compared with commercial Pt/C catalysts, the prepared Co-30@N-G hybrid electrocatalyst showed a high ORR activity at the same level in an alkaline medium. Subsequently, the Co-30@N-G hybrid electrocatalyst has been used as a cathode of Zn–air batteries, which displays equivalent performance to the systems derived using a commercial Pt/C catalyst. The Co-30@N-G derived mechanical rechargeable Zn–air battery showed a persistent flat discharge curve with minimum voltage loss at a high discharge rate of 40 mA cm⁻². The robustness of the Co-30@N-G ORR catalyst can allow the batteries to work constantly by periodically replacing the Zn anode and electrolyte, presenting an efficient and economical cathode for Zn–air flow batteries or Zn–air fuel cells.

---

1. Introduction

Due to global warming and fossil energy shortage, it is imperative to develop a clean and sustainable energy. Currently, much effort have been made to fabricate efficient, safe, affordable and environmentally friendly energy conversion and storage devices.^1,2 To date, Li-ion batteries (LIBs) are one of the most popular commercial energy conversion and storage devices to power various electronic devices and electric vehicles owing to their high cycle capability and energy efficiency. However, safety aspect and unsatisfactory storage capacity of LIBs (100–200 W h kg⁻¹) restrict a long-term application.^3,4 Recently, metal (M)–air batteries (M = Zn, Al, and Fe) have drawn considerable attention as a possible alternative due to their remarkably high energy density, low cost, and environmentally friendliness.^5–7 Among them, Zn–air battery has been considered to be a promising device for practical application owing to its abundant reserves and high theoretical specific energy density of 2500 W h kg⁻¹, which is about 6 times higher than the best LiBs reported.^8–10 According to the technological concept or the reaction mechanism of the energy system, Zn–air battery has been classified as a primary battery, a secondary battery, and a fuel cell.^11–13 Compared to the primary battery, the secondary Zn–air battery suffers from some problems such as redistribution of Zn, formation of undesirable Zn-morphologies, and formation of nonconducting ZnO layer on the anode surface, which limit the use and lifetime of the Zn–air batteries.^14,15 The high specific energy density of Zn–air battery is ascribed to the utilisation of environmentally abundant oxygen (O₂) in air, which greatly reduce the cathode size and total weight of the batteries. However, the sluggish oxygen reduction reaction (ORR) activity at the cathode side decreases the overall efficiency of the Zn–air battery.^16,17 Furthermore, the high cost and less availability of the currently used Pt-based electrocatalysts hinder an extensive use in Zn–air battery.^18,19 To overcome these problems, one of the best solutions is to fabricate a mechanical rechargeable Zn–air battery with inexpensive, corrosion-resistant, and highly efficient ORR electrocatalysts, from which maximum output can be attained at a significantly reduced investment.^10,20,21

Recently, carbon-based materials, particularly N-doped carbon materials are proved to be potential alternatives to Pt as electrocatalysts of ORR.²² Study results revealed that the N atoms inserted in the graphitic lattice can promote exterior hydrophilicity and change exterior electronic structures of carbon materials, which can improve the adsorption of oxygen molecules and the subsequent electron transfer process.^23,24 The morphology and structure are also very important for the ORR catalytic activity of carbon materials. Due to the large specific surface area, excellent conductivity and stability, N-doped graphene shows comparable ORR electroactivity to the commercial Pt/C catalyst at same loading in an alkaline electrolyte.^25–27 However, N-doped graphene with low density, which form a much thicker catalyst layer, may limit its volumetric specific activity, thus resulting in deteriorating mass transport in the Zn–air battery.²⁵ For the sake of solving the problem and improving the ORR activity, very recently, transition metal nanoparticles with ORR activity (e.g., Co, Fe, and Ni) are encapsulated in carbon materials. This approach can prepare composite catalysts displaying remarkably upgraded ORR activity owing to the synergy of metal and carbonaceous material.^26,28–31 On the one hand, the surface carbon layer in the unique encapsulation structure can prevent acid corrosion, oxidation and agglomeration of metal nanoparticles in the process of electrocatalysis, resulting in an excellent durability in wide range of pH.³² On the other hand, metal nanoparticles not only increase the graphitization degree of surface carbon layer during carbonization but also transfer electron to the carbon layer.³³ A synergetic role of the doped N in graphene lattice and the encapsulated metal nanoparticles produces an enhanced intrinsic electrocatalytic activity on the hybrid ORR catalysts.

As is known, the practicability and electrocatalytic performance of catalysts are largely depended on the synthetic method. Recently, pyrolysis of solid phase precursors has been widely used to synthesize carbon nanomaterials due to several advantages such as high yield, convenience of N doping, low cost by using naturally available compounds as precursors and simple process.^25,26,28,29 In this study, we report using cobalt nitrate (Co(NO₃)₂·6H₂O), β-cyclodextrin (β-CD), and urea as solid-phase precursors to prepare cobalt nanoparticles encapsulated in N-doped graphene nanoshell by pyrolyzing at high temperature and acid-leaching. The carbon morphology and structure show strong dependence on the composition of cobalt precursor. With the mass of cobalt nitrate increased from 0 to 0.1 g, the carbon morphology and structure changed from graphene-like nanoshell to thread-like carbon, whereas the amount of encapsulated cobalt nanoparticles increased and content of doped N decreased. The optimized electrocatalyst exhibited equivalent ORR activity in an alkaline medium to the commercial Pt/C and superior performance and cycling durability for zinc–air battery.

2. Experimental

2.1 Synthesis of electrocatalysts

All reagents were of analytical grade and used without further treatment. The catalysts were synthesized by an uncomplicated one-step pyrolysis process. Typically, 0.1 g β-cyclodextrin, 3.0 g urea and different amount of cobalt nitrate were dispersed in 20 mL deionized water. The mixture was constantly stirred and then dried at 80 °C. The acquired product was heated from 25 °C to 800 °C at a rate of 5 °C min⁻¹ under Ar with a flow rate of 50 cm³ min⁻¹ and maintained at 800 °C for 1 h in a tube furnace. Furthermore, in order to remove uncovered cobalt species, the samples were immersed in 0.5 M HClO₄ solution, maintained at 80 °C for 3 h and then washed completely with deionized water. Finally, the samples were dried at 60 °C in vacuum. The obtained loose black powder was labeled as Co-x@N-G, in which x represents the mass of cobalt nitrate (x = 0, 10, 30 and 100 mg, respectively).

2.2 Physical characterizations

The XRD data for the samples were recorded on a Rigaku/Max-3A X-ray diffractometer with CuKα radiation. FE-SEM images were obtained using a Supra 55 Sapphire apparatus equipped with an energy dispersive X-ray analysis system. TEM and HRTEM images were acquired on a JEOL-2010 microscope with an accelerating voltage of 200 kV. Raman spectra were obtained on a DXR Raman Microscope with 633 nm laser. Specific surface area of catalysts was measured by Brunauer–Emmette–Teller (BET) isotherms with a SSA-4200 analyzer using N₂ adsorption–desorption analysis at 77 K. Moreover, on the base of the BJH model, the pore size distribution data stem from the desorption branch of the isotherm.

2.3 Electrochemical measurements

Electrocatalytic activity was measured in a traditional three-electrode electrochemical system by a CHI-660E electrochemical station (Shanghai, Chenhua). A glassy carbon (GC) electrode (3 mm in diameter), saturated calomel electrode (SCE) and Pt foil served as the working electrode, reference electrode and counter electrode, respectively, were immersed in 0.1 M KOH aqueous solution. The potentials in this work were converted to the reversible hydrogen electrode (RHE) with the equation:³⁴

E_RHE = E_SCE + 0.059pH + 0.24 (1)

The catalyst ink was produced as follows: 1.5 mg prepared catalyst was dispersed in 735 μL of water, 205 μL of isopropanol and 60 μL of Nafion (5 wt%, Sigma-Aldrich) by ultrasound (>30 min) to form a homogeneous slurry. Then, the prepared ink (5 μL) was dropped on the mirror-polished glassy electrode and dried in vacuum at room temperature. The commercial Pt/C catalyst (20 wt%, Johnson Matthey) was also examined in the same process.

2.4 Air electrode preparation and zinc–air battery tests

A carbon paper (Toray TGP-H-060) was chosen as the conductive skeleton. The mixture, constituting carbon black (VulcanXC-72R) and polyvinylidene fluoride (PVDF) with a mass ratio of 8 : 1, was painted onto the skeleton to form an air diffusion layer. The mass loading was about 2 mg cm⁻². Then, 10 mg of catalysts and 200 μL of Nafion solution (5 wt%) were ultrasonically suspended in 600 μL of water and 200 μL of isopropanol, then the ink (200 μL) was coated on the air diffusion layer and dried in vacuum at room temperature. The loading of the catalysts was 1.14 mg cm⁻² unless otherwise stated.

A homemade zinc–air battery device was used for performance and cycling durability test of the mechanical rechargeable zinc–air battery. Zinc plate and 6 M KOH solution were used as the anode and electrolyte, respectively. Discharge polarization curves and long-time galvanostatic discharge curves were measured by a CHI-660E electrochemical station. The battery test was operated with purging pure oxygen gas.

3. Results and discussion

3.1 Structural characterization

In our experiment, the ORR catalysts were synthesized by a one-step pyrolysis method. During this pyrolysis process under Ar atmosphere, β-CD and urea provided C and N source, respectively. The graphitic-C₃N₄ from β-CD and the nitrogen atoms from urea can embed into the graphitic lattice, which form N-doped carbon materials.²⁵ The crystal structure and morphology of the obtained Co-x@N-G nanocomposites were characterized by XRD, SEM and TEM in detail. Fig. 1 shows the XRD curves of the Co-x@N-G samples. The distinct peaks at around 44.2°, 51.6°, and 75.8° are consistent with JCPDS card 15-0806, indicating the existence of metallic cobalt in the Co-x@N-G nanocomposites. The intensity of the peaks accord with the amount of cobalt nitrate used. The more cobalt nitrate, the stronger the intensity of the peaks. Furthermore, the broad diffraction peak at ∼26° can be ascribed to the G (002) plane of N-G.³⁵

Fig. 1 XRD patterns of Co-0@N-G, Co-10@N-G, Co-30@N-G and Co-100@N-G treated at 800 °C for 1 h in Ar atmosphere.

SEM image of Co-30@N-G shows that cobalt nanoparticles are uniformly dispersed in the graphene-like carbon materials (Fig. 2a), and the corresponding elemental mapping images also confirm the homogeneous distribution of Co and N elements in carbon matrix (Fig. 2b–d). The TEM image in Fig. 2e reveals that the cobalt nanoparticles are encapsulated in graphene nanoshell and HRTEM image in Fig. 2f exhibits some completely wrapped metallic cobalt nanoparticles showing (111) fringes (0.206 nm). No lattice fringes ascribe to CoO or Co₃O₄ have been found, which is in accord with the result of XRD. The specific layer distance (0.34 nm) also can be clearly seen in Fig. 2f, indicating high graphitized carbon shell. These nonporous graphene nanoshells, tightly wrapping around cobalt nanoparticles, can prevent Co nanoparticles from aggregation and oxidation. In comparison, the morphologies of Co-x@N-G materials pyrolyzed with different amount of cobalt nitrate also observed and shown in Fig. 3. Co-0@N-G is prepared without cobalt nitrate, showing a graphene-like structure (Fig. 3a), which is very similar to that of previous report.²⁵ When cobalt nitrate are introduced to the reaction system, the hybrid materials which consist of Co nanoparticles and carbon materials can be obtained. The amount of the cobalt nanoparticles in the Co-x@N-G nanocomposites is relevant to that of cobalt nitrate and the morphologies of carbon also changed.

Fig. 2 (a) SEM image; (b–d) elemental mappings of C, Co and N; (e) TEM image; (f) HRTEM image of Co-30@N-G.

Fig. 3 SEM and TEM images of (a and b) Co-0@N-G; (c and d) Co-10@N-G; and (e and f) Co-100@N-G.

The structure and defects of synthesized Co-30@N-G and Co-0@N-G materials were further detected by Raman spectra. In Fig. 4, the two distinct peaks at 1354 and 1580 cm⁻¹ and a broad peak around 2800 cm⁻¹ can be assigned to the D, G and 2D band of carbon materials, respectively.^25,26,36 It is well-know that the D band is relevant to the vibration of sp³ hybridized C atoms, representing the partially disordered structures on the graphitic plane or structural defects in carbon-based materials, while the G band is usually associated to the E_2g vibration of sp² hybridized C atoms, representing the graphitized degree. The intensity ratio of D band to G band (I_D/I_G) calculated from the peak intensity can be used to calculate the graphitized degree of the carbon-based materials.³⁵ In this work, the I_D/I_G value of Co-30@N-G (1.06) is slightly lower than that of Co-0@N-G (1.13), indicating an enhanced graphitization degree due to the existence of cobalt nanoparticles during carbonization.

Fig. 4 Raman spectra of Co-0@N-G and Co-30@N-G.

The Brunauer–Emmett–Teller (BET) analysis of N₂ adsorption–desorption isotherms (Fig. 5) revealed that both the plots of Co-30@N-G and Co-0@N-G were close to type IV patterns according to IUPAC classification with an obvious hysteresis loop in the high pressure region, indicative of the existence of mesopores.³⁷ Notably, the BET surface area and total pore volume of Co-30@N-G are 144.8 m² g⁻¹ and 0.74 cm³ g⁻¹, respectively, which is superior to that of Co-0@N-G (127.8 m² g⁻¹ and 0.69 cm³ g⁻¹, respectively) despite lower carbon content in Co-30@N-G.

Fig. 5 N₂ adsorption–desorption isotherms and the corresponding pore size distribution of Co-0@N-G and Co-30@N-G.

3.2 Electrochemical research

Using a three electrode device in 0.1 M KOH electrolyte, the properties of electrocatalysts were analyzed on CHI 660E. The loading of electrocatalysts on the glassy carbon electrode was same for all. CV plots of diverse catalysts are displayed in Fig. 6. In general, the confined area arising from the CV plot is proportional to the activity of specific surface area of catalyst based on carbon. In this case, the activities of Co-x@N-G electrocatalyst series decreased in the order of Co-30@N-G > Co-10@N-G > Co-0@N-G > Co-100@N-G. It is clear that Co-30@N-G has the largest active surface area of all the Co-x@N-G electrocatalysts, which is beneficial for Co-30@N-G to be applied as a cathode electrocatalyst for zinc–air batteries. In order to evaluate the ORR catalytic activity, series of Co-x@N-G electrocatalysts were evaluated in O₂-saturated and N₂-saturated electrolyte, respectively, under the static conditions. As the CV curves shown in Fig. 7a and Table 1, Co-30@N-G presents the highest catalytic activity among all the Co-x@N-G electrocatalysts. The onset potential, cathodic peak potential and cathodic peak current were 0.928 V (vs. RHE), 0.796 V (vs. RHE) and 1.22 mA cm⁻², respectively, which are equal and even better than the highly efficient Pt/C catalyst (Table 1). The long-term stability is also one of the most important requirements for excellent electrocatalyst.³⁸ The stability of Co-30@N-G catalyst for the ORR was evaluated using chronoamperometry at 0.8 V (vs. RHE) in an O₂-saturated 0.1 M KOH electrolyte and was compared with that of the commercial Pt/C (20 wt%). As shown in Fig. 7b, the current density of Co-30@N-G only decreases by 1.5% after 20 000 s continuous operation. In contrast, the ORR current of commercial Pt/C decreases by 9.1% after 20 000 s.

Fig. 6 Comparison of CV curves of Co-0@N-G, Co-10@N-G, Co-30@N-G, and Co-100@N-G in N₂-purged 0.1 M KOH at 10 mV s⁻¹.

Fig. 7 (a) CV curves for five different electrocatalysts in N₂-purged and O₂-saturated 0.1 M KOH at 10 mV s⁻¹; (b) current–time (i–t) chronoamperometric response for the ORR on Co-30@N-G and commercial Pt/C in O₂-saturated 0.1 M KOH at 0.8 V (vs. RHE).

Table 1 Catalytic activity data of ORR for various electrocatalysts

Electrocatalyst	Onset potential (V vs. RHE)	Cathodic peak potential (V vs. RHE)	Cathodic peak current (mA cm^−2)
Pt/C	0.956	0.782	1.09
Co-0@N-G	0.827	0.706	0.74
Co-10@N-G	0.868	0.767	0.99
Co-30@N-G	0.928	0.796	1.22
Co-100@N-G	0.887	0.794	0.68

---

Although Co-30@N-G presents a superior ORR activity under static CV study, it is not sufficient to conclude that the catalyst can ensure an enhanced performance in a practical Zn–air battery because the concentration of dissolved oxygen will be different in the electrolyte used for the static CV investigation (0.1 M KOH) and the electrolyte used in the battery testing (6 M KOH). Along with this, the overall performance will also be decided by the formation of an effective triple-phase boundary. Hence, it is necessary to analyze the catalyst under similar electrolyte conditions, which are favorable for real Zn–air battery. The prepared battery was composed of Co-30@N-G ORR electrocatalyst coated on the carbon fiber paper as the cathode and Zn foil as the anode in 6 M KOH (Fig. 8a). Moreover, the performance of the system derived from Co-30@N-G electrocatalyst was also been compared with a similar system made from the Pt/C-derived air electrode. The battery reaction can be described as follows:

Anode: Zn + 4OH⁻ → Zn(OH)₄²⁻ + 2e⁻ (2)

Zn(OH)₄²⁻ → ZnO + H₂O + 2OH⁻ (3)

Cathode: O₂ + 4e⁻ + 2H₂O → 4OH⁻ (4)

Overall reaction: 2Zn + O₂ → 2ZnO (5)

Fig. 8 Zn–air batteries performance of the catalysts: (a) Schematic representation of the mechanical rechargeable Zn–air battery; (b) polarization plots of the Zn–air batteries made with Co-30@N-G and Pt/C as the air catalysts; (c) typical discharge curves of Zn–air batteries with Co-30@N-G as the cathode catalyst under continuous discharge until complete consumption of Zn at two different current densities. Specific capacity was normalized to the mass of consumed Zn.

The open-circuit voltages for the systems made from Co-30@N-G and Pt/C are 1.42 V and 1.45 V, respectively. The polarization curve of the cell presents a current density up to 131 mA cm⁻² (at 1.0 V) arising from Co-30@N-G and the peak power density achieved 227 mW cm⁻² at the voltage of 0.65 V (Fig. 8b). With the same loading, the battery made up of Pt/C catalyst presents slightly lower electrocatalytic performance (125 mA cm⁻² for the current density and 219 mW cm⁻² for peak power density). The difference in the ORR activity of the cells derived from these two air electrodes could be directly correlated to the previously observed difference in the electrochemical properties in the three-electrode system performed in 0.1 M KOH electrolyte. A primary Zn–air battery composed of Co-30@N-G catalyst is very robust. No obvious voltage drops were observed at a constant current density of 10 mA cm⁻² or 40 mA cm⁻² until the Zn foil is completely consumed (Fig. 8c). The calculated specific capacity of Co-30@N-G derived battery normalized to the mass of consumed Zn was 671 mA h g_Zn⁻¹, which was also better than that of the Pt/C derived battery (∼645 mA h g_Zn⁻¹) (all discharged at 10 mA cm⁻²).

Interestingly, the battery can be recharged again and again by simply refueling the anode (i.e., Zn) and the electrolyte periodically, and the produced waste (zincate species) can be assembled and recovered in recycling plants. In this way, the cathode electrode made with Co-30@N-G can work robustly for a long time at a high discharge current density (40 mA cm⁻²) with minimal voltage loss (Fig. 9). Thus, Co-30@N-G as the air electrode in terms of activity, stability and durability can be used in future metal–air batteries and electric vehicles (EVs).

Fig. 9 Recharging of the Zn–air battery using Co-30@N-G catalyst as the cathode material and by reloading the Zn anode and electrolyte periodically. The recharging process is indicated by the arrows. Conditions: anode: Zn-foil, cathode: Co-30@N-G catalyst, electrolyte: 6 M KOH, discharge rate: 40 mA cm⁻².

4. Conclusions

In summary, non-precious metal ORR electrocatalysts derived by encapsulating Co nanoparticles in N-doped graphene nanoshells have been synthesized by a controllable one-step method of pyrolysis of inexpensive and easily available raw materials. The morphology and ORR activity of the hybrid electrocatalysts can be easily modulated by adjusting the mass of cobalt nitrate added in the reaction system. The Co-30@N-G electrocatalyst presented a significant ORR activity that was close to the up-to-date Pt/C electrocatalyst in an alkaline system. Moreover, the Zn–air batteries using Co-30@N-G as the cathode eletrocatalyst delivered a large current density of 131 mA cm⁻² with a high peak power density of 227 mW cm⁻². At a discharge rate of 10 mA cm⁻², Co-30@N-G derived Zn–air battery showed a superior specific capacity of 671 mA h g_Zn⁻¹ compared to Pt/C derived setup (645 mA h g_Zn⁻¹). The robustness of the Co-30@N-G eletrocatalyst can allow the battery to work continuously by replacing the Zn anode and electrolyte periodically, presenting an ideal air catalyst for Zn–air batteries applied in EVs.

Acknowledgements

The authors are grateful to the National High Technology Research and Development Program of China (No. 2009AA03Z319) and the Fundamental Research Funds for the Central Universities of China (No. DUT12LK04).

References

1. A. S. Arico, P. Bruce, B. Scrosati, J.-M. Tarascon and W. V. Schalkwijk, Nat. Mater., 2005, 4, 366–377.
2. M. Armand and J. M. Tarascon, Nature, 2008, 451, 652–657.
3. B. Scrosati and J. Garche, J. Power Sources, 2010, 195, 2419–2430.
4. M. S. Whittingham, Chem. Rev., 2014, 114, 11414–11443.
5. M. He, X. Yejian, Z. Xufeng and L. Zhaoping, Prog. Chem., 2015, 7, 935–944.
6. Z. Chen, A. Yu, D. Higgins, H. Li, H. Wang and Z. Chen, Nano Lett., 2012, 12, 1946–1952.
7. Z. L. Wang, D. Xu, J. J. Xu and X. B. Zhang, Chem. Soc. Rev., 2014, 43, 7746–7786.
8. Y. Li and H. Dai, Chem. Soc. Rev., 2014, 43, 5257–5275.
9. N. Xu, J. Qiao, X. Zhang, C. Ma, S. Jian, Y. Liu and P. Pei, Appl. Energy, 2016, 175, 495–504.
10. Y. Li, M. Gong, Y. Liang, J. Feng, J.-E. Kim, H. Wang, G. Hong, B. Zhang and H. Dai, Nat. Commun., 2013, 4, 1805.
11. V. M. Dhavale and S. Kurungot, ACS Catal., 2015, 5, 1445–1452.
12. P. Ganesan, P. Ramakrishnan, M. Prabu and S. Shanmugam, Electrochim. Acta, 2015, 183, 63–69.
13. Z. Ma, P. Pei, K. Wang, X. Wang, H. Xu, Y. Liu and G. peng, J. Power Sources, 2015, 274, 56–64.
14. D. U. Lee, J.-Y. Choi, K. Feng, H. W. Park and Z. Chen, Adv. Energy Mater., 2014, 4, 1301389.
15. P. Pei, K. Wang and Z. Ma, Appl. Energy, 2014, 128, 315–324.
16. M. Shao, Q. Chang, J. P. Dodelet and R. Chenitz, Chem. Rev., 2016, 116, 3594–3657.
17. X. Ge, A. Sumboja, D. Wuu, T. An, B. Li, F. W. T. Goh, T. S. A. Hor, Y. Zong and Z. Liu, ACS Catal., 2015, 5, 4643–4667.
18. M. Gatalo, P. Jovanovič, G. Polymeros, J.-P. Grote, A. Pavlišič, F. Ruiz- Zepeda, V. S. Šelih, M. Šala, S. Hočevar, M. Bele, K. J. J. Mayrhofer, N. Hodnik and M. Gaberšček, ACS Catal., 2016, 6, 1630–1634.
19. G. V. Fortunato, F. de Lima and G. Maia, J. Power Sources, 2016, 302, 247–258.
20. S. K. Singh, V. M. Dhavale and S. Kurungot, ACS Appl. Mater. Interfaces, 2015, 7, 21138–21149.
21. Z. Chen, A. Yu, R. Ahmed, H. Wang, H. Li and Z. Chen, Electrochim. Acta, 2012, 69, 295–300.
22. D.-W. Wang and D. Su, Energy Environ. Sci., 2014, 7, 576.
23. N. Daems, X. Sheng, I. F. J. Vankelecom and P. P. Pescarmona, J. Mater. Chem. A, 2014, 2, 4085–4110.
24. L. Hadidi, E. Davari, M. Iqbal, T. K. Purkait, D. G. Ivey and J. G. C. Veinot, Nanoscale, 2015, 7, 20547–20556.
25. F. Pan, J. Jin, X. Fu, Q. Liu and J. Zhang, ACS Appl. Mater. Interfaces, 2013, 5, 11108–11114.
26. G. Zhang, W. Lu, F. Cao, Z. Xiao and X. Zheng, J. Power Sources, 2016, 302, 114–125.
27. G. Nam, J. Park, S. T. Kim, D.-b. Shin, N. Park, Y. Kim, J.-S. Lee and J. Cho, Nano Lett., 2014, 14, 1870–1876.
28. Y. Su, Y. Zhu, H. Jiang, J. Shen, X. Yang, W. Zou, J. Chen and C. Li, Nanoscale, 2014, 6, 15080–15089.
29. J. Wang, H. Wu, D. Gao, S. Miao, G. Wang and X. Bao, Nano Energy, 2015, 13, 387–396.
30. T. Kim, Y. Ohata, J. Kim, C. K. Rhee, J. Miyawaki and S.-H. Yoon, Carbon, 2014, 80, 698–707.
31. B. Li, Y. Chen, X. Ge, J. Chai, X. Zhang, T. S. A. Hor, G. Du, Z. Liu, H. Zhang and Y. Zong, Nanoscale, 2016, 8, 5067–5075.
32. J. Wang, D. Gao, G. Wang, S. Miao, H. Wu, J. Li and X. Bao, J. Mater. Chem. A, 2014, 2, 20067–20074.
33. L. Jiang, J. Yan, L. Hao, R. Xue, G. Sun and B. Yi, Carbon, 2013, 56, 146–154.
34. Y. Liang, Y. Li, H. Wang, J. Zhou, J. Wang, T. Regier and H. Dai, Nat. Mater., 2011, 10, 780–786.
35. S. M. Unni, S. Devulapally, N. Karjule and S. Kurungot, J. Mater. Chem., 2012, 22, 23506.
36. J. Liang, Y. Jiao, M. Jaroniec and S. Z. Qiao, Angew. Chem., Int. Ed., 2012, 51, 11496–11500.
37. C. H. Huang, Q. Zhang, T. C. Chou, C. M. Chen, D. S. Su and R. A. Doong, ChemSusChem, 2012, 5, 563–571.
38. W. Yan, W. Bian, C. Jin, J.-H. Tian and R. Yang, Electrochim. Acta, 2015, 177, 65–72.

---

Footnote

† Author contributions: Kai-Yuan Zhou and Guang-Yi Chen contributed equally to this work.

---