A flexible dual solid-stateelectrolyte supercapacitor with suppressed self-discharge and enhanced stability

Abstract

Redox supercapacitors (SCs) have recently been developed as a high-efficiency strategy for energy storage due to their high charge storage capacity and easy accessibility. Developing a low-cost, highly efficient and universal method to control the self-discharge (SDC) process of redox supercapacitors (SCs) would amplify their applications in energy storage. This report on a novel dual solid-state electrolyte supercapacitor (DSES) has demonstrated that the addition of polyvinyl alcohol (PVA) into a CuCl₂/H₂SO₄ active electrolyte could significantly suppress the SDC process, which takes 10 times longer for 50% energy retention, benefiting from moderate kinetics and ion crossover rate. The role of the dual electrolyte system is also illustrated to be effective for improving the cycling stability. Hence, our strategy, which is effective in enhancing the energy storage, mitigating the SDC process and improving the cycling stability, could provide a versatile method for designing low-cost and stable redox SCs for energy storage.

---

Introduction

Electrochemical double-layer capacitors (EDLCs), known as supercapacitors (SCs), accumulate ions of opposite charge in a double layer on the surface of high-surface-area carbon, and are able to deliver higher power (10³ to 10⁵ W kg⁻¹) and longer cycle life than batteries.^1–3 Despite that, the charge being confined to the surface results in a medium energy density (4–5 W h kg⁻¹), which is much lower than that of batteries (20 to 150 W h kg⁻¹) and is popularly perceived as the main impediment to the growth of the SC market.^4–6 The energy gap between batteries and SCs usually can be minimized via introducing faradaic reactions into the purely capacitive process. Hence, various strategies such as hybridizing with heteroatoms,^7–11 organic molecules,^12,13 transition metal oxides/hydroxides^14–18 or conducting polymers¹⁹ have been implemented on porous carbon. Apart from the capacitance improvement, most of them still suffer from decreased conductivity, inferior cycling stability and complicated synthetic routes.

The recently developed ‘redox SCs’, with redox-active electrolytes substituting for inert ones, are another promising alternative. The approach involves highly soluble redox couples and less-expensive carbon, and thus is highly efficient and cost-effective compared with solid redox active electrodes. Until now, diverse redox additives including hydroquinone, copper salts, halides, vanadium complexes, methylene blue and ferricyanide^20–25 have been reported to significantly augment the capacitance of SCs. Nevertheless, one of the biggest challenges that prevent their wide applications remains the fast self-discharge (SDC) process, which can occur via migration of mobile redox species, diffusion overvoltage losses or intermixing of redox couples.^20,21,23,26 Accounting for high SDC rates of little practical use, some strategies such as Nafion cation-exchange membranes,^21,24 active copper salts which can convert into insoluble Cu species,²¹ or an electrostatic adsorption mechanism²⁷ have been proposed to control the SDC process. But some negative effects such as high cost, inferior cycling stability or slender progress still seriously prohibit their wide application. Boettcher et al. have recently achieved great progress by facilitating electrostatic adsorption of charged redox couples on an electrode with methyl viologen/bromide as the electrolyte, thereby significantly slowing down the SDC rate.²³ Besides the improvement, its application is still limited by the specific type of electrolytes that could be used and little work has been reported on successful mitigation of the SDC process of redox SCs.

Here, we demonstrate a novel dual solid-state electrolyte supercapacitor (DSES) to increase capacitance and meanwhile mitigate the SDC process, as well as improve the cycling stability. The DSES device with CuCl₂/polyvinyl alcohol (PVA)/H₂SO₄ and PVA/H₂SO₄ as the electrolytes at the positive and negative electrodes, respectively, is designed for the assembly of porous carbon cloth (CC)-based SCs. CuCl₂ in acidic media has been chosen as the active couple because of its fast electron-transfer kinetics during charge/discharge. The combination of double-layer capacitance with pseudocapacitance provided by the redox active electrolytes leads to a tremendous increase in the overall capacitance of the DSES device up to 4148 mF cm⁻² at 4 mA cm⁻², which is five times higher than that in the absence of CuCl₂. Although the intermixing of redox species can be suppressed by conversion to insoluble Cu,²¹ the SDC rate in the CuCl₂/H₂SO₄ aqueous electrolyte is still rapid. Also, it was found that the fast SDC process complied with the diffusion-control mechanism. In consideration of the slower kinetics and lower ionic crossover between polymer electrolyte/electrode interfaces, PVA polymer was further added into the aqueous system. It was demonstrated to be effective in controlling the SDC effect, which took 10 times longer for 50% energy retention (18 465 s) than the time taken in the aqueous electrolyte (1720 s). Moreover, the pore blockage of the negative electrode in a single solid-state electrolyte supercapacitor (SSES) caused by insoluble species could be mitigated with dual polymer electrolytes, and a better cycling stability of 92% capacitance retention after 3000 cycles, compared with 35% for SSES, is achieved. This new system might shed light on developing low-cost and stable redox energy storage devices.

Results and discussion

Ultramicroporous carbon cloth (TCC), prepared by a simple high-temperature calcination of commercial carbon cloth, was chosen as the electrode material as it possesses high flexibility, good conductivity and high specific surface area.²⁸ The SEM image of TCC is shown in Fig. S1.† The TCC electrode exhibits a quasi-rectangular shape in its cyclic voltammetry (CV) curve with broad pseudocapacitive peaks at ∼0.3 V in 1 M H₂SO₄ resulting from additional oxygen functional groups (Fig. S2†).²² As shown in Fig. 1A, the addition of CuCl₂ to the H₂SO₄ supporting electrolyte leads to a significant increase in the area of the CV curve at the same scan rate, indicating a higher capacitance. Here a redox peak at around 0.45 V is observed in 0.1 M CuCl₂ acidic solution, which can be attributed to the redox reaction involving Cu²⁺. A further increase in the concentration of CuCl₂ leads to higher currents and broader CVs, demonstrating an enhanced redox pseudocapacitance induced by electroactive Cu²⁺. The GCD curves in Fig. S3† and 1B–D also demonstrate an increase of capacitance with the increase of CuCl₂ concentration. The maximum capacitance for the TCC electrode calculated from galvanostatic charge and discharge (GCD) curves is 22 430 mF cm⁻² in 0.3 M CuCl₂/H₂SO₄ (Fig. 1D), which is about 20 times larger than the capacitance obtained in the absence of CuCl₂. The redox reactions were also pH dependent and the peak current decreased or even disappeared in CV curves at higher pH (Fig. S4†), indicating that the significantly increased capacitance is attributed to pseudocapacitance from redox reactions of Cu²⁺.

Fig. 1 (A) CVs of TCC electrodes in 0, 0.1, 0.2 and 0.3 M CuCl₂/1 M H₂SO₄ solution at 5 mV s⁻¹. GCD curves of TCC electrodes in (B) 0.1 M, (C) 0.2 M, (D) 0.3 M CuCl₂/1 M H₂SO₄ at various current densities.

Even though Cu²⁺ was chosen as the redox species to inhibit the migration of active species, the SDC behavior of the redox aqueous system was also unsatisfactory. As shown in Fig. 2A, the voltage of the CuCl₂/H₂SO₄ aqueous device quickly goes down to 0.5 V within 1720 s and further down to 0.2 V within 4100 s. Fitting of the SDC curve shows a linear relationship between V and t^1/2 with a high correlation coefficient R² (higher than 0.99) (Fig. 2B), demonstrating a diffusion control model.²⁹ At the beginning of the SDC process, a small deviation of the fitting curve was observed, which can be attributed to the more equilibrated state of ions at the electrode/electrolyte interface.³⁰ Thus, we can conclude that the SDC process in the CuCl₂/H₂SO₄ system is probably controlled by ion diffusion/transport.

Fig. 2 (A) The comparison of SDC profiles plotted as the self-discharge time of the aqueous SC with the CuCl₂/H₂SO₄ electrolyte and SSES device with 0.5 M CuCl₂/PVA/H₂SO₄ electrolyte. (B) The corresponding SDC profiles of the aqueous SC plotted as t^1/2. (C) The device structure of the SSES device with CuCl₂/PVA/H₂SO₄ as the redox active electrolyte. (D) CV curves of the SSES device at 0 M, 0.1 M, and 0.5 M CuCl₂ at 5 mV s⁻¹.

In considering the trade-off between suppression of the SDC behavior and the reaction kinetics, PVA polymer and a higher concentration of Cu²⁺ were added in the original aqueous solution to further control the shuttling of electrolytes between the two electrodes, as polymer electrolytes usually have moderate kinetics and ion mobility relative to that in aqueous solution. The structure of the SSES device with CuCl₂/PVA/H₂SO₄ as the redox electrolyte is displayed in Fig. 2C. CV and GCD tests were then performed to evaluate the electrochemical performance of the device (Fig. 2D and S5†). Prominent redox waves centered at 0.75 V are exhibited in the CV curves, and their intensities increase with the addition of CuCl₂, corresponding to the enhanced capacitance from GCD results. To study the ability of energy retention, a potential decay effect was then monitored at open circuit. As depicted in Fig. 2A, compared with the original system, the SSES with PVA displays a much slower SDC rate. The voltage declining to half of its original value needs 18 465 s, which is 10 times longer than that of CuCl₂/H₂SO₄ aqueous solution. To explore the SDC mechanism in the device, the SDC curve was analyzed with the well-accepted potential driving model and the diffusion control model.³¹ However, fitting of the corresponding SDC curve shows that the behavior does not comply with either single control model (Fig. 3A and B), but the SDC process can be controlled with the dual-mechanism model by separating the SDC process into two parts, as depicted in Fig. 3C. The beginning of the SDC process obeys the potential driving model and 50% voltage drop is governed by the diffusion controlled SDC process. The effects of gel electrolytes, electrolyte/electrode interface and the ionic transportation were then reflected on Nyquist plots (Fig. 3D and Table S1†) based on the electrochemical impedance spectroscopy (EIS) test. The Nyquist plots display both a semi-circle in the high-frequency region and a straight line in the low-frequency region, corresponding to an electron-transfer-limited process and a diffusion-limited process, respectively. The higher intersection on the real axis of the Nyquist plot for SSES (3.32 vs. 2.11 Ω, Table S1†), known as series resistance R_s, represents an increased internal resistance associated with the electrolyte resistance, the intrinsic resistance of electrode materials, and the contact resistance at the interface of the electrode. Meanwhile, the larger semi-circle in the high frequency region (charge transfer resistance, R_ct, Table S1†) for SSES, involving the process occurring at the electrode/electrolyte interface, suggests relatively poor ionic conductivity at the electrode–electrolyte interface with the addition of the polymer electrolyte. These variations all relate to the slightly slow kinetics in polymer electrolytes, compared with aqueous electrolytes. In addition, the Warburg region (W_o) between the semicircle and the long tail correlates with the ion diffusion/transport to electrolyte/electrode interfaces and the relatively high Warburg resistance in SSES (2.51 Ω vs. 0.56 Ω) indicates a slower ion diffusion to the electrode surface. Therefore, we can conclude that the slow kinetics and low ion transport rate between the electrolyte and electrode/electrolyte interface can be the main factors for the supression of the SDC process in SSES.

Fig. 3 SDC curves of SSES with 0.5 M CuCl₂/H₂SO₄/PVA as the solid-state electrolyte and its fitting curves based on the (A) diffusion-controlled model and (B) potential driving model. (C) Corresponding SDC curve of the SSES device and its simulation by the dual-mechanism control model. (D) Nyquist plots and the corresponding equivalent circuits of the redox aqueous SC and SSES devices.

Cycling stability has been considered as another major issue for redox SCs. Fig. 4A shows the cycling performance of the SSES and a big deterioration of the electrochemical performance is observed. As Cu²⁺ ions can be reduced to insoluble metallic and/or oxidized copper deposited on the negative electrode, the loss of energy can be attributed to the reduced redox-active ions in electrolytes in SSES or the pore blockage effect on porous carbon. To explore the main reason for the reduced charge storage capacity, a two-electrode liquid electrolyte system was further chosen as redox electrolytes could be easily updated in this system during the cycling test. As shown in Fig. S6,† the area of the CV diminishes significantly after 400 cycles and it cannot be restored even when the two electrodes are then transferred to a new 0.2 M CuCl₂/H₂SO₄ electrolyte solution which is similar to the original, indicating that the lowered concentration of the redox-active species cannot be the main reason for the decline of capacitance. Further clear analysis of the positive and negative electrodes with scanning electron microscopy (SEM) images (Fig. 4B and C and S7†) and the comparison with that of TCC before the cycling test (Fig. S1†) reveal that a layer of deposit covers the surface of the negative electrode and plugs the pores of TCC, but the positive electrode surface shows no significant change. The reason can be ascribed to the occurrence of the reduction process of Cu²⁺ on the surface of the negative electrode during the cathodic polarization. The X-ray diffraction (XRD) pattern of the surface layer on the negative electrode after the CV cycling performance is shown in Fig. 4D. It exhibits three distinct peaks at 43.4°, 50.5° and 74.3°, corresponding to the (111), (200) and (208) planes of Cu, respectively. The peaks at 28.5°, 47.5°, 56.3°, 69.7°, and 76.6° which are characteristic of CuCl and Cu₄Cl(OH)₃ also exist. As a result of the chemical instability of Cu⁺, a small amount of Cu₂O is present. The associated reaction processes are described in the ESI.† Hence, we can conclude that the poor cycling performance of SSES is primarily caused by the deposition of metallic and nonmetallic copper precipitates on the surface of the negative electrode. The ultramicropores of TCC were blocked by a thick layer of deposits and therefore, this result in a lower ion accessible surface area and poor reaction kinetics, as well as decreased electrical double-layer capacitance and redox capacitance. Also, the surface layer of the deposits contains a complex mix of metallic and oxidized copper, which increases the contact resistance of the electrode and meanwhile impedes the charge storage performance.³²

Fig. 4 (A) Cycle performance of SSES at a current density of 30 mA cm⁻². SEM images for (B) positive and (C) negative electrodes tested after cycling for 400 CV cycles in 0.2 M CuCl₂/1 M H₂SO₄ solution. (D) XRD analysis of the negative electrode after cycling tests in CuCl₂/H₂SO₄ solution.

To improve the cycling performance and remit the pore blockage effect, a new DSES was designed with two kinds of solid-state electrolytes. In the DSES device, as displayed in Fig. 5A, the electrolyte in the negative electrode was replaced with PVA/H₂SO₄, while PVA/H₂SO₄/CuCl₂ was retained as the active electrolyte in the positive electrode. Based on the structure of DSES, the working mechanism is similar to that of an asymmetric SC, where the positive and negative electrodes provide pseudocapacitive and double-layer capacitive behaviors, respectively. As shown in Fig. 5B, excellent cycling performance with 92% capacitance retention after 3000 continuous cycles was achieved with DSES, which is significantly better than the 35% retention for SSES. This is due to the suppression of pore blockage at the surface of the negative electrode with double solid-state electrolytes. Also, Nyquist plots display similar trends after 1000 cycles, suggesting the good electrochemical stability of DSES during the cycling (Fig. S8†). The electrochemical data were collected when the device was at equilibrium. In addition to the improved cycling performance, similar to SSES, DSES shows enhanced capacitive behavior. Fig. 5C shows the CV curves for DSES at different concentrations of the CuCl₂/PVA/H₂SO₄ electrolyte and prominent redox peaks of Cu²⁺ ions can be observed. A maximum capacitance of 4148 mF cm⁻² at a current of 4 mA cm⁻² is achieved and also maintains 1680 mF cm⁻² at a fast charge/discharge rate of 20 mA cm⁻² (Fig. S9†). The corresponding volumetric capacitances are shown in Fig. S10,† and the maximum volumetric capacitance is calculated to be about 46 F cm⁻³. A Ragone plot is obtained based on the GCD tests and the energy density reaches 576 μW h cm⁻2, which is 5 times larger than that of SSES with PVA/H₂SO₄ as the electrolyte (Fig. 5D and S11†). Additionally, the maximum power density reaches 20 mW cm⁻², which is slightly lower than that of the aqueous device with 0.3 M CuCl₂/1 M H₂SO₄ as the electrolyte (Fig. S12†). The reason can be ascribed to the relatively slow ion diffusion and the depressed electrochemical kinetics with the addition of PVA.^33,34 Also, the DSES device shows a slow SDC rate, as shown in Fig. S13.† Moreover, in order to evaluate the potential of DSES for flexible energy storage, the mechanical flexibility performance was analyzed. The digital photo of DSES in the inset figure shows excellent flexibility. The corresponding CV curves also remain stable under different bending angles (Fig. S14†). These results all prove the effectiveness of this new system in high-performance and stable flexible redox SCs.

Fig. 5 (A) Device structure of DSES with CuCl₂/PVA/H₂SO₄ and PVA/H₂SO₄ dual solid-state electrolytes at the positive and negative electrodes, respectively. (B) Comparison of the cycling performance of SSES (black) and DSES (red) at a current density of 30 mA cm⁻² (C) CV curves of DSES with various concentrations of CuCl₂ at 5 mV s⁻¹. (D) Ragone plot for DSES in 1.5 M CuCl₂/PVA/H₂SO₄, with SSES in the absence of CuCl₂ as a comparison.

Conclusions

In summary, we have demonstrated a novel DSES device, with CuCl₂/PVA/H₂SO₄ and PVA/H₂SO₄ as the electrolytes at positive and negative electrodes, respectively, to enhance the capacitance and mitigate the SDC process, as well as improve the cycling stability. The device could deliver high capacitance up to 4148 mF cm⁻² benefiting from the redox reaction of Cu²⁺ ions. The SDC process of the redox SC was effectively suppressed with the addition of PVA in the redox electrolyte as a result of moderate kinetics and ion crossover rate relative to the original electrolyte. In addition, the DSES device was demonstrated to be an effective strategy to improve the cycling stability of SSES by remitting pore blockage behavior on the negative electrode. These results might promise the application of our universal strategy in low-cost and stable redox energy storage devices.

Experimental

Synthesis of ultramicroporous carbon cloth (TCC)

TCC was synthesized according to the previous report.²⁸ Briefly, commercial CC (Shanghai Hesen Electric Co. Ltd, China, HCCP330) was calcined at 1000 °C for 1 h with a heating rate of 5 °C min. The product was directly used as the electrode material without the need for any other current collector and conductive additive.

Morphology and electrochemical characterizations

SEM images were obtained with a SU-70 microscope. X-ray diffraction (XRD) data were collected with a D/tex-Ultima TV wide angle X-ray diffractometer equipped with Cu Kα radiation (λ = 1.54 Å). The cyclic voltammetry (CV) tests, galvanostatic charge/discharge (GCD) data, SDC processes and electrochemical impedance spectroscopy (EIS) measurements were recorded on a Gamry Reference 600 electrochemical workstation and a LAND CT2001A. All the electrochemical tests were conducted at room temperature.

Electrochemical tests in aqueous redox electrolytes

A three-electrode electrochemical device was first used to test the electrochemical properties of TCC in CuCl₂/H₂SO₄ redox electrolytes, consisting of a working electrode, and a platinum sheet and an SCE electrode as the counter and reference electrodes, respectively. The two-electrode cell in the aqueous redox electrolyte consisted of two working electrodes separated by filter paper and packaged in CuCl₂/H₂SO₄ electrolytes. The electrochemical data were recorded with the device at equilibrium.

Fabrication of a flexible single solid-state electrolyte supercapacitor (SSES)

A flexible SSES was assembled in a symmetrical two-electrode configuration with the addition of PVA into the H₂SO₄/CuCl₂ electrolyte. Specifically, 2 g H₂SO₄ and corresponding amount of CuCl₂ were added into 20 mL water solution. Then, 2 g polyvinyl alcohol (PVA) powder (Aldrich) was added into the above solution and stirred at 85 °C until the PVA dissolved completely. After that, two TCC electrodes and a fiber filter paper were immersed into the electrolyte for 1 h and assembled face to face with a working area of 1 cm². The assembled device was finally placed under ambient conditions to allow the electrolyte to solidify. The thickness of the SSES is measured to be about 0.9 mm.

Fabrication of a flexible dual solid-state electrolyte supercapacitor (DSES)

To suppress the deposition of Cu on the negative electrode which leads to plugging of the pores of carbon and inferior cycling stability, a DSES was designed and assembled. The PVA/H₂SO₄/CuCl₂ gel electrolyte used for the positive electrode was prepared in the same way as that of SSES, while the PVA/H₂SO₄ gel electrolyte was used for the negative electrode. Finally, a flexible DSES was assembled with a working area of 1 cm² in the same way as SSES and the thickness of the device was about 0.9 mm. As suggested by Gogotsi and Simon et al., compared with the gravimetric performance, the electrochemical performance based on the total area/volume of the assembled device is more reliable,^4,35–38 our study mainly provides the areal and volumetric performance of the whole device.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

Financial support from the National Key R&D Program of China (2016YFA0202900), the Key Program Supported by the Natural Science Foundation of Zhejiang Province, China (LZ18B060002), the National Natural Science Foundation of China (21622308, 91534114) and the Fundamental Research Funds for the Central Universities (2017XZZX002-16) is greatly appreciated.

Notes and references

1. Z. Wu and X. B. Zhang, Acta Phys.-Chim.Sin., 2017, 33, 305–313.
2. P. Simon and Y. Gogotsi, Nat. Mater., 2008, 7, 845.
3. L. Xie, H. Y. Wang, C. H. Chen, S. J. Mao, Y. Q. Chen, H. R. Li and Y. Wang, Research, 2018 DOI:10.1155/2018/5807980.
4. P. Simon and Y. Gogotsi, Acc. Chem. Res., 2013, 46, 1094–1103.
5. P. S. J. R. Miller, science, 2008, 321, 651–652.
6. M. F. El-Kady, V. Strong, S. Dubin and R. B. Kaner, Science, 2012, 335, 1326–1330.
7. T. Lin, I.-W. Chen, F. Liu, C. Yang, H. Bi, F. Xu and F. Huang, Science, 2015, 350, 1508–1513.
8. Y. Zhou, X. Xu, B. Shan, Y. Wen, T. Jiang, J. Lu, S. Zhang, D. P. Wilkinson, J. Zhang and Y. Huang, Energy Storage Materials, 2015, 1, 103–111.
9. Z. Wu and X. B. Zhang, Sci. China Mater., 2016, 59, 547–557.
10. J. Zhao, H. W. Lai, Z. Y. Lyu, Y. F. Jiang, K. Xie, X. Z. Wang, Q. Wu, L. J. Yang, Z. Jin, Y. W. Ma, J. Liu and Z. Hu, Adv. Mater., 2015, 27, 3541–3545.
11. G. Zhu, T. Chen, Y. Hu, L. Ma, R. Chen, H. Lv, Y. Wang, J. Liang, X. Li, C. Yan, H. Zhu, H. Liu, Z. Tie, Z. Jin and J. Liu, Nano Energy, 2017, 33, 229–237.
12. S. Gan, L. Zhong, L. Gao, D. Han and L. Niu, J. Am. Chem. Soc., 2016, 138, 1490–1493.
13. M. Boota, C. Chen, M. Becuwe, L. Miao and Y. Gogotsi, Energy Environ. Sci., 2016, 9, 2586–2594.
14. H. Wang, R. Fan, J. Miao, J. Chen, S. Mao, J. Deng and Y. Wang, J. Mater. Chem. A, 2018, 6, 6780–6784.
15. H. Wang, C. Xu, Y. Chen and Y. Wang, Energy Storage Materials, 2017, 8, 127–133.
16. X. H. Cao, B. Zheng, W. H. Shi, J. Yang, Z. X. Fan, Z. M. Luo, X. H. Rui, B. Chen, Q. Y. Yan and H. Zhang, Adv. Mater., 2015, 27, 4695–4701.
17. J. Zhou, Y. Huang, X. Cao, B. Ouyang, W. Sun, C. Tan, Y. Zhang, Q. Ma, S. Liang, Q. Yan and H. Zhang, Nanoscale, 2015, 7, 7035–7039.
18. J. Chen, H. Wang, J. Deng, C. Xu and Y. Wang, J. Mater. Chem. A, 2018, 6, 8986–8991.
19. Q. Meng, K. Cai, Y. Chen and L. Chen, Nano Energy, 2017, 36, 268–285.
20. R. Silvia, B. Clara, G. Marcos, M. Rosa and S. Ricardo, Angew. Chem., Int. Ed., 2011, 123, 1737–1739.
21. L. B. Chen, H. Bai, Z. F. Huang and L. Li, Energy Environ. Sci., 2014, 7, 1750–1759.
22. L.-Q. Mai, A. Minhas-Khan, X. Tian, K. M. Hercule, Y.-L. Zhao, X. Lin and X. Xu, Nat. Commun., 2013, 4, 2923.
23. S. E. Chun, B. Evanko, X. F. Wang, D. Vonlanthen, X. L. Ji, G. D. Stucky and S. W. Boettcher, Nat. Commun., 2015, 6, 7818.
24. F. Elzbieta, F. Krzysztof, M. Mikolaj and L. Grzegorz, ChemSusChem, 2012, 5, 1181–1185.
25. C. Xu, H. Wang, J. Deng and Y. Wang, Sustainable Energy Fuels, 2018, 2, 357–360.
26. S. T. Senthilkumar, R. K. Selvan and J. S. Melo, J. Mater.Chem. A, 2013, 1, 12386–12394.
27. B. Wang, J. A. Macia-Agullo, D. G. Prendiville, X. Zheng, D. Liu, Y. Zhang, S. W. Boettcher, X. Ji and G. D. Stucky, J. Electrochem. Soc., 2014, 161, A1090–A1093.
28. H. Wang, J. Deng, C. Xu, Y. Chen, F. Xu, J. Wang and Y. Wang, Energy Storage Materials, 2017, 7, 216–221.
29. B. W. Ricketts and C. Ton-That, J. Power Sources, 2000, 89, 64–69.
30. Q. Zhang, J. Rong, D. Ma and B. Wei, Energy Environ. Sci., 2011, 4, 2152–2159.
31. Q. Zhang, C. Cai, J. Qin and B. Wei, Nano Energy, 2014, 4, 14–22.
32. J.-J. Velasco-Vélez, K. Skorupska, E. Frei, Y.-C. Huang, C.-L. Dong, B.-J. Su, C.-J. Hsu, H.-Y. Chou, J.-M. Chen, P. Strasser, R. Schlögl, A. Knop-Gericke and C.-H. Chuang, J. Phys. Chem. B, 2018, 122, 780–787.
33. Y. Huang, M. Zhong, Y. Huang, M. Zhu, Z. Pei, Z. Wang, Q. Xue, X. Xie and C. Zhi, Nat. Commun., 2015, 6, 10310.
34. Y. Huang, M. Zhong, F. Shi, X. Liu, Z. Tang, Y. Wang, Y. Huang, H. Hou, X. Xie and C. Zhi, Angew. Chem., Int. Ed., 2017, 56, 9141–9145.
35. Y. Gogotsi and P. Simon, Science, 2011, 334, 917–918.
36. P. Simon and Y. Gogotsi, Nat. Mater., 2008, 7, 845–854.
37. X. Lu, M. Yu, G. Wang, Y. Tong and Y. Li, Energy Environ. Sci., 2014, 7, 2160–2181.
38. Y. Z. Zhang, Y. Wang, T. Cheng, W. Y. Lai, H. Pang and W. Huang, Chem. Soc. Rev., 2015, 44, 5181–5199.

---

Footnote

† Electronic supplementary information (ESI) available: Calculations, characterizations, electrochemical data and references. See DOI: 10.1039/c8se00364e

---