A facile approach to prepare biomimetic composite separators toward safety-enhanced lithium secondary batteries

Abstract

A mussel-inspired polydopamine (PDA) coating turns radio-frequency (RF) Al₂O₃ sputtering – which thus far has not been appropriate for the surface treatment of porous polyolefin-based separators – into a damage-free, reliable, and cost-efficient process. Due to the thermally resistive PDA layers, polyethylene (PE) separators can sustain high-power Al₂O₃ sputtering conditions over 75 W, which significantly reduces processing time. Furthermore, compared to the as-prepared separators, PDA/Al₂O₃-coated PE separators also reveal improved thermal stability and cycle performance for lithium secondary batteries. PDA/Al₂O₃-coated PE separators retained their original size when exposed to temperatures of 145 °C over 30 min, while the bare PE separators shrank to 9% of their original size. At a temperature of 25 °C, the unit cell (LiMn₂O₄/separator/Li metal) employing the PDA/Al₂O₃-coated PE separators maintained 94.8% (103.4 mA h g⁻¹) of the initial discharge capacity after 500 cycles at C/2 rate and 51.7% (56.7 mA h g⁻¹) at 25 C rate, while the corresponding values for the bare PE separators were 89% (98.6 mA h g⁻¹) at C/2 rate and 24.5% (27.2 mA h g⁻¹) at 25 C rate.

---

Introduction

Due to their high specific-energy density and long cycle life, lithium ion batteries (LIBs) are vital power sources predominantly used in consumer electronics.^1–5 With the ever-increasing demand for high energy density LIBs in emerging fields such as electric vehicles (EVs) and energy storage systems (ESSs), it has become imperative to find a way to increase the energy density of LIBs at low cost.^6,7 Especially for large-scale applications of LIBs, the issue of safety takes a pivotal position because consumer safety is a major concern for battery manufacturers. Fast charging accompanying high current and/or over-charging can cause lithium dendrite growth on the anode surface owing to the Li-insertion potential of conventional carbonaceous anodes, which is almost 0 V vs. Li/Li⁺.^8–11 Out of the various battery constituents including anodes, cathodes, and electrolytes, the separators are the key to determining the safety of the whole LIB system.^5–7,12 As separators physically block a direct contact between the cathode and the anode, if the heat generated from large-scale batteries equipped with high-power energy devices ruptures the dimensional stability of the separators, a catastrophic thermal failure of the LIB would occur, accompanied by explosive flames and venting. To overcome these drawbacks, a large number of approaches to apply a surface coating on commercial porous polyolefin-based (in particular, polyethylene (PE) and polypropylene (PP)) separators have been reported.^13–16 Heat-resistant coatings, including ceramic particles and/or high-melting polymers, can effectively improve the thermal stability of commercial polyolefin-based separators. On the other hand, the coating layers inherently alter the pore structure of the separators, and thus influence the ionic pathway within the separators, which in turn affects the cell performance of the LIBs. In general, coating layers on separators are likely to block and/or hinder the existing pore structure of the separators.

To the best of our knowledge, we were the first to report the feasibility of using radio-frequency (RF) magnetron sputtering to introduce binder-free inorganic coating layers (Al₂O₃ in the study) onto commercial PE separators.¹⁷ An optimal amount of hydrophilic Al₂O₃ coating layers creates desirable ionic pathways, resulting in enhanced rate capabilities of LIBs without blocking the pores, which are beneficial for holding larger amounts of liquid electrolyte. However, the sputtering time of longer than 10 minutes should be shortened to a few minutes for commercial application. Hence, in order to maintain the appropriate amount of sputtered coating materials, the RF power has to be increased, which might cause thermal damage to PE separators. A facile approach to circumvent this problem is to pre-coat the bare separator with polydopamine (PDA) for enhancing thermal stability, followed by high-power RF sputtering.

Experimental

Materials

The cathode was prepared by mixing LiMn₂O₄ (LMO, Kyushu Ceramics, Japan), polyvinylidene fluoride (PVDF, KF-1300, Kureha, Japan), and conductive carbon (Super-P, Timcal, Switzerland). The anode was Li metal foil (400 μm, Honjo Metal, Japan). A mixture of ethylene carbonate/ethyl methyl carbonate (EC/EMC = 3/7 by volume) containing 1.15 M LiPF₆ was purchased from PANAX ETEC (Korea) and used without further purification. N-Methyl-2-pyrrolidone (NMP), Trizma® base (99.9%), Trizma® hydrochloride (99%), 2-(3,4-dihydroxyphenyl)ethylamine hydrochloride (dopamine hydrochloride, 98%), and methanol (CH₃OH) were purchased form Aldrich and used without further purification. Deionized (DI) water (Milli-Q system, Millipore Co., USA, 18.2 MΩ cm) was used. Microporous polyethylene (PE, ND420, Asahi Kasei E-materials, Japan) separators were used (porosity: 41%, thickness: 20 μm). The Al₂O₃ target was purchased from Taewon Scientific Co., Ltd (iTASCO, Korea) with 99.99% purity.

Polydopamine (PDA) surface coating

Separators were surface-coated via a simple polydopamine coating method.⁵ Dopamine solution (2 mg mL⁻¹) was prepared using a mixture of Tris buffer solution (pH 8.5, 10 mM) and methanol (CH₃OH/Tris buffer = 1/1 in wt%) as a co-solvent.

Preparation of PDA/Al₂O₃-coated PE separators

Radio-frequency (RF) magnetron sputtering was used to deposit Al₂O₃ on both the bare PE and the PDA-coated PE separators. The target-to-substrate distance was 150 mm and the diameter of the target was 50 mm. The working vacuum pressure of the stainless steel chamber was 7 × 10⁻³ Torr of argon (99.999%); cooling water was circulated around the target and throughout the chamber to prevent over-heating during deposition. Prior to Al₂O₃ sputtering, the targets were pre-sputtered for 60 min at an RF power of 50 W to remove residual surface contaminants.

Characterization of separators

The surface morphology of various types of separators including bare PE, PDA-coated PE, and PDA/Al₂O₃-coated PE was characterized by field-emission scanning electron microscopy (FE-SEM; JSM-6390, JEOL, Japan). The thermal stability of the composite separator was observed by placing it in an oven and heating at 145 °C for 30 min. The thermal shrinkage ratio was calculated using eqn (1), where A_i and A_f represent the area of the sample before and after the high-temperature storage, respectively:

Thermal shrinkage ratio (%) = (A_i − A_f)/A_i × 100 (1)

The electrolyte uptake amount was determined using eqn (2), where W_i and W_f indicate the separator weight before and after electrolyte absorption, respectively:

Uptake amount (wt%) = (W_f − W_i)/W_i × 100 (2)

The air permeability represented by the Gurley number was examined with a Gurley densometer (4110N, Thwing-Albert, USA) by measuring the time taken for air to pass through a determined volume under a given pressure. The Gurley number was determined according to procedure JIS-P8117 (Japanese Industrial Standards) by measuring the time (s) for 100 cm³ of air to pass through the membrane under a constant air pressure (6.52 psi).

Electrode preparation

A slurry mixture containing 90 wt% LiMn₂O₄, 5 wt% conductive carbon (Super-P), and 5 wt% PVDF in NMP was employed. The slurry was cast on aluminum foil (15 μm, Sam-A Aluminum, Korea) using a doctor blade. The cast slurry was dried under air at 130 °C for 1 h, and the electrodes were roll-pressed with a gap-control-type roll-pressing machine (CLP-2025, CIS, Korea). The cathode (density: 1.67 g cm⁻³; loading amount: 7.35 mg cm⁻²; thickness: 44 μm) was punched into a disc shape (radius: 12 mm) and dried at 60 °C for 12 h under vacuum before assembly.

Electrochemical measurements

The ionic conductivity of the separators was evaluated by sandwiching a liquid electrolyte-soaked separator between two stainless steel electrodes. To evaluate the effect of the various types of separators on the cell performance, CR2032-type unit half-cells (LiMn₂O₄/Li metal) were assembled in a glove box filled with argon. The unit cells were aged for 12 h and cycled between 3.0 and 4.5 V vs. Li/Li⁺ at C/10 rate (0.088 mA cm⁻²) using a constant-current (CC) mode for both charging and discharging processes at room temperature. They were then stabilized in three subsequent cycles between 3.0 and 4.5 V vs. Li/Li⁺ at C/5 rate (0.176 mA cm⁻²) in constant-current/constant-voltage (CC/CV) mode for charging and CC mode during discharging processes at room temperature using a charge/discharge cycle tester (PNE Solutions, Korea).

The pulse-power of the unit cells was investigated using a hybrid pulse-power characterization (HPPC) test sequence.^18–20 The test consisted of a 10 s discharge pulse (5 C, 4.3 mA cm⁻²), a 40 s rest, and a 10 s charge pulse (3.75 C, 3.225 mA cm⁻²) performed at every 10% state-of-charge (SOC) after precycling and high temperature storage. For the latter, the unit cells were stored at 60 °C for 3 days, followed by cooling down to 25 °C for 10 h.

To evaluate the rate capability, the unit half-cells (LiMn₂O₄/Li metal) were discharged with a varying discharge current density from 1 C to 30 C (1 C, 5 C, 10 C, 15 C, 20 C, 25 C, and 30 C), while maintaining the same charging current density of C/2 (0.44 mA cm⁻²). The unit cells were subsequently cycled at a C-rate of C/2 (0.44 mA cm⁻²) for 500 cycles to evaluate the cycle performance. For the evaluation of rate capability and cycle performance, CC/CV mode and CC mode were used during the charging process and discharging process, respectively, between 3.0 and 4.5 V vs. Li/Li⁺ at room temperature.

Results and discussion

As shown in Fig. 1(a–c), the bare PE separators are thermally damaged during high-power RF sputtering at a power over 25 W due to the heat accumulated during the process, and the extent of the damage increases with increasing sputtering power. This implies that the RF sputtering process only allows a 25 W operating power to be used for commercial PE separators. In this study, we introduced a sophisticated mussel-inspired polydopamine (PDA) surface coating upon PE separators, followed by RF sputtering for Al₂O₃ coating.^5,17,21 PDA coatings can readily adhere to various types of surfaces such as organic, inorganic, and even metallic, and render the surface hydrophilic.^5,22,23 At least under our experimental conditions for PDA and RF sputtering coating processes, due to the nano-scale coating layer thickness, the original pore structures of the bare PE separators were not significantly altered, as shown in Fig. 2.^5,17

Fig. 1 Digital camera images of (a–c) bare PE separators and (d) polydopamine-coated PE separators exposed to various RF sputtering powers and times (red-dotted square indicates the original size of the separator i.e. 3 cm × 3 cm).

Fig. 2 Scanning electron microscopy (SEM) images of (a and b) bare PE separators, (c and d) PDA-coated PE separators, and (e and f) PDA/Al₂O₃-coated PE separators.

The physical properties of the separators, including thickness, Gurley number, liquid uptake amount, and ionic conductivity, were investigated and the results are listed in Table 1. As discussed above, the nano-scale coating layer thickness did not influence the total thickness of the separators. On the other hand, the Gurley number, a good indicator of the permeability of a membrane, increased from 271 to 295 and 299 s 100 mL⁻¹ after PDA coating and Al₂O₃ sputtering treatment, respectively. As the pores were slightly blocked due to the surface coating, more time was inherently required for air to pass through at the same pressure.

Table 1 Properties of separators

System	Thickness (μm)	Gurley number (sec 100 mL^−1)	Liquid uptake (%)	Ionic conductivity (mS cm^−1)
Bare PE	20	271	58	0.719
PDA-coated PE	20	295	112	0.759
PDA/Al2O3-coated PE	20	299	124	0.758

---

Both PDA and Al₂O₃ coating layers are known to be hydrophilic due to the presence of oxygen moieties,^5,17 and thus both surface-coated separators show affinity for polar organic electrolytes as shown in Fig. 3. The released liquid electrolyte droplet (EC/EMC = 3/7 containing 1 M LiPF₆) immediately spread out over the surface-coated separators, while the hydrophobic bare PE separators maintained the liquid electrolyte droplet in its as-released form over a long period. Consequently, both surface-coated separators turned semi-transparent as liquid filled the pore structures, while the bare PE separators remained opaque.

Fig. 3 Digital camera images taken after a droplet of liquid electrolyte (EC/EMC = 3/7 containing 1 M LiPF₆) was released upon (a) bare PE separators, (b) PDA-coated PE separators, and (c) PDA/Al₂O₃-coated PE separators.

In general, improved wetting ability by surface treatment can increase the ionic conductivity of bare PE separators, because ionic conductivity mainly depends on the amount of available Li⁺ ions associated with the liquid electrolyte, under the assumption that the pore structures of the membrane are uniform.^5,21 As listed in Table 1, the liquid electrolyte holding capacities of the PDA-coated PE separators and the PDA/Al₂O₃-coated PE separators are 193% and 214% more than that of the bare PE separators, respectively. As a result, both surface-coated PE separators revealed enhanced ionic conductivity compared to the bare PE separators.

For safety reasons, dimensional stability of the separators at high temperatures should be the overriding criterion for separator selection. Under abnormal conditions or harsh operating conditions such as high-power operation, LIBs in general generate a large amount of heat, and the separator should not be ruptured and/or deformed under such conditions to keep the whole LIB system safe. As shown in Fig. 4, various types of separators were exposed to a high temperature of 145 °C for 30 min, and the dimensional changes were monitored in each case. Bare PE separators shrank drastically to only 9% of their original size during the test. Due to the thermally stable melanin-like properties of PDA, the PDA-coated separators showed improved thermal stability: they maintained 67% of the original size.^7,24 Surprisingly, the PDA/Al₂O₃-coated separators maintained their original dimensions after the same treatment. Considering the fact that 40 min of RF Al₂O₃ sputtering (at 25 W) was required to achieve a similar level of dimensional stability for the Al₂O₃-coated PE separators in our previous study, our new approach, requiring only 2 min of RF Al₂O₃ sputtering (at 75 W), is a more efficient and cost-effective process for mass production.¹⁷

Fig. 4 Digital camera images of (a) bare PE separators, (b) PDA-coated PE separators, and (c) PDA/Al₂O₃-coated PE separators after exposure to a temperature of 145 °C for 30 min (red-dotted square indicates the original size of the separator i.e. 3 cm × 3 cm).

To investigate the effect of surface coating of PE separators on electrochemical properties such as rate capability and discharge capacity retention ability, 2032 coin-type unit cells consisting of LiMn₂O₄/separator/Li metal were prepared.

As shown in Fig. 5(a), each unit cell, employing various types of separators including bare PE, PDA-coated PE, and PDA/Al₂O₃-coated PE, showed almost identical performance during the first cycle operated in constant current (CC) mode for both charging and discharging processes at C/10 rate between 3.0–4.5 V vs. Li/Li⁺: for the bare PE separators charge capacity = 108.9 mA h g⁻¹ and discharge capacity = 107.6 mA h g⁻¹; for the PDA-coated PE separators charge capacity = 111.0 mA h g⁻¹ and discharge capacity = 108.0 mA h g⁻¹; and for the PDA/Al₂O₃-coated PE separators charge capacity = 110.1 mA h g⁻¹ and discharge capacity = 107.8 mA h g⁻¹.

Fig. 5 (a) Voltage profiles of unit cells employing bare PE separators, PDA-coated PE separators, and PDA/Al₂O₃-coated PE separators for the first cycle (CC mode at C/10 between 3.0–4.5 V vs. Li/Li⁺). (b) Nyquist plots for the unit cells after the first cycle, and (c) the same Nyquist plots with a different scale (the unit cells were fully discharged to 3.0 V vs. Li/Li⁺ prior to EIS measurements).

After the first cycle, followed by full discharge to 3.0 V vs. Li/Li⁺, electrochemical impedance spectroscopy (EIS) measurements were carried out. In general, the EIS spectra of unit cells are composed of two depressed semicircles and a steep sloping line in the low-frequency regions.^5,25,26 A small semicircle in the high-frequency region (left-hand side) is related to the resistance associated with the solid electrolyte interphase (SEI) layer (R_SEI), and a large semicircle in the medium-frequency region (right-hand side) corresponds to the resistance of the charge-transfer process (R_ct), accompanied by migration of the Li⁺ ions at the electrode/electrolyte interface.^5,27 The x-axis intercept corresponds to the bulk resistance (R_b) including the resistances of the electrolyte and electrode.²⁷ As shown in Fig. 5(b), the sum of the resistances (R_b + R_SEI + R_ct) has a higher value for the bare PE than for the two surface-coated cases. As shown in Fig. 5(c), R_b of the bare PE is much larger compared to the others. Keeping in mind that both R_b and R_ct are closely related to Li⁺ ion migration, which in turn is closely associated with the wetting ability of separators (increasing the degree of electrolyte retention), it is easy to understand why the PDA-coated PE and PDA/Al₂O₃-coated PE separators have smaller resistance values compared to those of the bare PE separators.

We also investigated the pulse-power capability of the unit cells using an HPPC test sequence.^18–20 The discharge pulse-power capabilities of unit cells (LiMn₂O₄/Li metal) employing different types of separators are summarized in Fig. 6. After precycling, PDA-coated PE and PDA/Al₂O₃-coated PE separators reveal almost identical discharge pulse-power capabilities, which are both higher values compared to bare PE, over the whole range of SOC from 10 to 90%. After high temperature storage (60 °C, 3 days), the pulse-power of each unit cell was lower compared to that after precycling. On the other hand, PDA/Al₂O₃-coated PE separators revealed the best discharge power capability. The HPPC results were consistent with the EIS results shown and discussed above (Fig. 5(b)).

Fig. 6 Pulse-power capability plot for unit cells after (a) precycling and (b) high temperature storage (60 °C, 3 days).

To investigate the effect of the increased wetting ability of the PDA-coated PE and PDA/Al₂O₃-coated PE separators, we measured the rate capability and discharge retention ability of the corresponding unit cells.

For the rate capability test, the discharge current was varied from 1 C to 30 C (1, 5, 10, 15, 20, 25, and 30 C), keeping the charging current constant (C/2, 0.44 mA cm⁻²). As shown in Fig. 7(a), the PDA-coated PE and PDA/Al₂O₃-coated PE separators showed improved rate capability compared to the bare PE separators. At 25 C rate (at the 30^th cycle), the rate capabilities of the PDA-coated PE and PDA/Al₂O₃-coated PE separators were enhanced by 214% (58.2 mA h g⁻¹) and 208% (56.7 mA h g⁻¹), respectively, compared to the bare PE separators (27.2 mA h g⁻¹). The improvement can be explained by the improved wetting ability of the coated separators and the EIS results discussed in Table 1 and Fig. 5, respectively. The improved wetting ability of the coating layers helps the separators retain greater amounts of liquid electrolyte within their micro-pore structures, which lowers the bulk resistance R_b, thus ensuring smooth migration of Li⁺ ions between the separators and electrodes due to small interfacial resistance (R_SEI + R_ct). When the discharge current was set back to 1 C after 35 cycles, the discharge capacity of the unit cells recovered to the level observed prior to the rate capability test. This indicates that irreversible material loss during the test was insignificant. From the results, we can infer that the reduced discharge capacity of the unit cells during the rate capability test was due to kinetic reasons, and not the active material loss caused by electrochemical surface reactions.

Fig. 7 (a) A comparison of the discharge capacities of unit cells employing different types of separators (bare PE, PDA-coated PE, and PDA/Al₂O₃-coated PE separators) at different discharging current densities from 1 C to 30 C, keeping the charging current density at C/2. (b) Cycle performance of unit cells operated at C/2 for both charging and discharging processes.

To investigate the cycle retention ability of the surface-coated separators, unit cells (LiMn₂O₄/separator/Li metal) employing different types of separators were cycled at C/2 rate for both the charging and discharging processes in constant current/constant voltage (CC/CV) mode between 3.0 and 4.5 V vs. Li/Li⁺. As shown in Fig. 7(b), both the surface-coated separators achieved highly improved cycle retention ability compared to the bare PE separators. Bare PE separators retained 89% (98.6 mA h g⁻¹) of the initial discharge capacity, while PDA-coated and PDA/Al₂O₃-coated PE separators achieved 94.6% (104.0 mA h g⁻¹) and 94.8% (103.4 mA h g⁻¹) retention, respectively. Recently, we reported that the homogeneous Li⁺ ion flux derived from the improved wetting ability of separators can highly improve the cycle life of Li metal.^7,28 Here, it is inferred that the improved wetting ability of surface-coated PE separators facilitates the homogeneous Li⁺ ion flux over the electrode surfaces, resulting in enhanced cycle life.

Fig. 8 shows the potential profiles of selected cycles during the rate capability test illustrated in Fig. 7(a). Clearly, the polarization became larger with increasing current density. On the other hand, the bare PE revealed a larger polarization compared to the others. These results were consistent with the larger EIS (Fig. 5(b)) results and the smaller HPPC pulse-power (Fig. 6) results.

Fig. 8 Potential profiles of selected cycles (every 5^th cycle: 5^th, 10^th, 15^th, 20^th, 25^th, and 30^th) of unit cells employing (a) bare PE, (b) PDA-coated PE, and (c) PDA/Al₂O₃-coated PE separators during the rate capability experiments illustrated in Fig. 7(a).

Conclusions

With the help of mussel-inspired PDA coating, followed by RF Al₂O₃ sputtering, highly functional PE separators were developed. They demonstrate superior thermal stability at high temperature (145 °C) and improved cell performance including rate capability and cycle retention ability. PDA coating enables PE separators to sustain high-power Al₂O₃ RF sputtering with remarkably short processing times, and enhances the process efficiency. This means PDA/Al₂O₃-coated PE separators are a promising separator material for large-scale LIB systems, ensuring better safety standards and better performance.

Acknowledgements

We acknowledge financial support from the Ministry of Education, Science and Technology (MEST) and National Research Foundation (NRF) of Korea through the Human Resource Training Project for Regional Innovation (2014066977) and IT R&D program of MOTIE/KEIT (10046314).

Notes and references

1. K. Xu, Chem. Rev., 2004, 104, 4303–4418.
2. M. Armand and J.-M. Tarascon, Nature, 2008, 451, 652–657.
3. C. Liu, F. Li, L. P. Ma and H. M. Cheng, Adv. Mater., 2010, 22, E28–E62.
4. N. S. Choi, Z. Chen, S. A. Freunberger, X. Ji, Y. K. Sun, K. Amine, G. Yushin, L. F. Nazar, J. Cho and P. G. Bruce, Angew. Chem., Int. Ed., 2012, 51, 9994–10024.
5. M.-H. Ryou, Y. M. Lee, J. K. Park and J. W. Choi, Adv. Mater., 2011, 23, 3066–3070.
6. J. J. Woo, Z. Zhang and K. Amine, Adv. Energy Mater., 2014, 4 DOI:10.1002/aenm.201301208.
7. M.-H. Ryou, D. J. Lee, J. N. Lee, Y. M. Lee, J. K. Park and J. W. Choi, Adv. Energy Mater., 2012, 2, 645–650.
8. Y. Wang, H. Liu, K. Wang, H. Eiji, Y. Wang and H. Zhou, J. Mater. Chem., 2009, 19, 6789–6795.
9. T.-F. Yi, J. Shu, Y.-R. Zhu, X.-D. Zhu, C.-B. Yue, A.-N. Zhou and R.-S. Zhu, Electrochim. Acta, 2009, 54, 7464–7470.
10. S. S. Zhang, J. Power Sources, 2006, 161, 1385–1391.
11. C. Lin, X. Fan, Y. Xin, F. Cheng, M. O. Lai, H. Zhou and L. Lu, J. Mater. Chem. A, 2014, 2, 9982–9993.
12. M. Baginska, B. J. Blaiszik, R. J. Merriman, N. R. Sottos, J. S. Moore and S. R. White, Adv. Energy Mater., 2012, 2, 583–590.
13. J. Shi, T. Shen, H. Hu, Y. Xia and Z. Liu, J. Power Sources, 2014, 271, 134–142.
14. Y. Ko, H. Yoo and J. Kim, RSC Adv., 2014, 4, 19229–19233.
15. H.-S. Jeong and S.-Y. Lee, J. Power Sources, 2011, 196, 6716–6722.
16. J. Song, M.-H. Ryou, B. Son, J.-N. Lee, D. J. Lee, Y. M. Lee, J. W. Choi and J.-K. Park, Electrochim. Acta, 2012, 85, 524–530.
17. T. Lee, W.-K. Kim, Y. Lee, M.-H. Ryou and Y. M. Lee, Macromol. Res., 2014, 22, 1190–1195.
18. I. Bloom, S. A. Jones, V. S. Battaglia, G. L. Henriksen, J. P. Christophersen, R. B. Wright, C. D. Ho, J. R. Belt and C. G. Motloch, J. Power Sources, 2003, 124, 538–550.
19. H.-W. Lee and Y.-M. Lee, J. Korean Electrochem. Soc., 2012, 15, 115–123.
20. PNGV Battery Test Manual, Revision 1, Idaho National Engineering Laboratory, Department of Energy, DOE/ID-10597, May 1998.
21. S. M. Kang, M.-H. Ryou, J. W. Choi and H. Lee, Chem. Mater., 2012, 24, 3481–3485.
22. J. H. Waite, Nat. Mater., 2008, 7, 8–9.
23. H. Lee, S. M. Dellatore, W. M. Miller and P. B. Messersmith, Science, 2007, 318, 426–430.
24. B. Simonovic, V. Vucelic, A. Hadzi-Pavlovic, K. Stepien, T. Wilczok and D. Vucelic, J. Therm. Anal., 1990, 36, 2475–2482.
25. J. Song, H. Lee, Y. Wang and C. Wan, J. Power Sources, 2002, 111, 255–267.
26. J. Li, C. F. Yuan, Z. H. Guo, Z. A. Zhang, Y. Q. Lai and J. Liu, Electrochim. Acta, 2012, 59, 69–74.
27. T.-H. Cho, M. Tanaka, H. Onishi, Y. Kondo, T. Nakamura, H. Yamazaki, S. Tanase and T. Sakai, J. Power Sources, 2008, 181, 155–160.
28. M.-H. Ryou, Y. M. Lee, Y. Lee, M. Winter and P. Bieker, Adv. Funct. Mater., 2015, 25, 825.

---

Footnote

† These authors contributed equally to this work.

---