New ether-functionalized pyrazolium ionic liquid electrolytes based on the bis(fluorosulfonyl)imide anion for lithium-ion batteries

Abstract

Four new ionic liquids (ILs) composed of ether-functionalized pyrazolium cations and the bis(fluorosulfonyl)imide (FSI) anion are prepared and characterized in this report. The physicochemical properties of these ILs, such as melting point, thermal stability, viscosity, conductivity and electrochemical stability, are systematically investigated. These four FSI-based ILs are all in the liquid state and their viscosities are lower than 40 mPa s at room temperature. The charge–discharge performances of Li/LiFePO₄ cells containing these IL electrolytes with 0.8 mol L⁻¹ LiFSI are also examined. Three electrolytes exhibit nice cycling stability at 0.1C, and the PZ2o2-2-FSI electrolyte shows excellent rate performance.

---

1. Introduction

As power sources for electric vehicles (EVs) and hybrid electric vehicles (HEVs) lithium-ion batteries attract considerable attention continuously. However, their implementation for large-scale applications have been restricted due to safety concerns which mainly caused by the flammable organic solution-based electrolytes.^1–4 In order to enhance the safety of lithium-ion batteries, a large number of investigations have focused on ionic liquids (ILs) as safe electrolytes because of their preferable properties: negligible volatility, nonflammability, and high thermal and electrochemical stability.^5–9 However, the inherently higher viscosity of IL electrolytes still hinders the diffusion of lithium ion and results in the sluggishness of electrochemical processes.¹⁰ Therefore, low viscous ILs are always pursued by researchers for improving the electrochemical performances of lithium-ion batteries.

The physicochemical properties of ILs composed of organic cations and different anions can be altered diversely by structural variation. Fluorinated sulfonyl anions are often used to obtain low-viscosity ILs due to high charge delocalization and weak coordinating nature,^10–14 and their representatives are bis(trifluoromethanesulfonyl)imide (TFSI) and bis(fluorosulfonyl)imide (FSI) anions. Compared to TFSI anion, FSI anion is more beneficial to viscosity and conductivity, according to the research results about imidazolium,¹⁵ pyrrolidinium¹⁶ and phosphonium¹⁷ ILs. And the outstanding advantage of FSI anion in viscosity and conductivity can be attributed to smaller size and weaker electrostatic and induction interactions between cations and anions.¹⁸ Like TFSI anion, FSI anion has been proved to support the lithium deposition and stripping on the surface of lithium metal, ascribed to the formation of favorable solid electrolyte interphase (SEI) film. And good cycling performance of LiCoO₂ or LiFePO₄ cathode has been illustrated in several FSI-based IL electrolytes, including 1-ethyl-3-methyl imidazolium FSI (EMI-FSI),^16,19 N-methyl-N-propylpiperidinium FSI (PP13-FSI),¹⁶ N-methyl-N-propylpyrrolidinium FSI (Py13-FSI),¹⁹ N-ethyl-N-butylpyrrolidinium FSI (Py24-FSI)²⁰ and tributylmethylphosphonium FSI (P1444-FSI).²¹

Furthermore, functionalization of cation is a useful means to obtain superior ILs and then provides more choices for electrochemical application.^22,23 Recently, ether-functionalized oniums (i.e. imidazolium,^24–26 tetraalkylammonium,^11,27 pyrrolidinium,²⁸ piperidinium^28,29) with various anions have gotten extensive attention. Ether group with electron-donating effect can reduce melting points and viscosities of ILs but not cause the apparent decline of thermal and electrochemical stability.^{24,27,28,30,31} Many ether-functionalized ILs based on TFSI anion have been utilized as electrolytes in lithium-ion batteries. A typical example is N,N-diethyl-N-methyl-N-(2-methoxyethyl)ammonium TFSI (DEME-TFSI). Seki et al. have reported the excellent capacity retention of Li/LiCoO₂ cells with DEME-TFSI electrolyte, and deeply investigated the interfacial property of LiCoO₂ cathode and lithium metal anode.^32,33 Nevertheless, ether-functionalized ILs based on FSI anion have not been mentioned as an electrolyte in lithium-ion batteries.

Pyrazolium cation possesses a C–N heteroaromatic ring which is analogical to imidazolium cation, but studies of pyrazolium ILs are quite scarce compared with imidazolium ILs.^34–39 The previous researches involving pyrazolium ILs still focused on the TFSI-based ILs, some of which have been applied as electrolytes in lithium-ion batteries.^40,41 In this work, we combined ether-functionalized pyrazolium cation and FSI anion to prepare four new ILs for the first time, whose structure were shown in Fig. 1. After the investigation of physicochemical properties for these ILs, their low-melting point, low-viscosity and high-conductivity characteristics were confirmed. The charge–discharge test of Li/LiFePO₄ cells containing these IL electrolytes with 0.8 mol L⁻¹ LiFSI was also performed. PZ2o1-2-FSI, PZ2o2-1-FSI and PZ2o2-2-FSI electrolytes exhibited nice cycling stability at 0.1C, and PZ2o2-2-FSI electrolyte showed excellent rate performance.

Fig. 1 Structure of these FSI-based ether-functionalized pyrazolium ILs.

2. Experimental

2.1 Reagents and materials

Potassium bis(fluorosulfonyl)imide (KFSI) was purchased from Kanto chemical Co., Ltd. Lithium bis(fluorosulfonyl)imide (LiFSI) was kindly provided by Morita Chemical Industries Co., Ltd. 1-(2-Ethoxymethyl) pyrazole and 1-(2-ethoxyethyl) pyrazole were prepared according to the published method.⁴² The other reagents were purchased from Alfa Aesar.

2.2 Synthesis of FSI-based ILs

Pyrazolium iodides were prepared via alkylation of ether-functionalized pyrazoles with iodomethane or iodoethane, as the previous work.⁴³ The next step was anion exchange between pyrazolium iodides and KFSI to obtain the final FSI-based products following the literatures.³⁰ Detailed synthesis procedures and NMR data were recorded in the ESI.† The pyrazolium iodides also were used to prepare PZ2o1-1-TFSI and PZ2o2-2-TFSI via anion exchange with LiTFSI.

2.3 Measurement

The structures of products were identified by ¹H NMR and ¹³C NMR spectra (Bruker, Advance III HD 400). The NMR results revealed that the chemical structures of all synthesized ILs are expected and there was no evident impurity in these samples. The corresponding peaks for the position of each proton had been marked in Fig. S1† as an example. The water contents were all below 50 ppm, determined by a moisture titrator (Metrohm 73KF Karl Fischer coulometer). The melting points of ILs were detected by using a differential scanning calorimeter (DSC, TA Instrument Q2000). Packing by a small aluminum crucible under dry atmosphere, each sample (5 mg approximately) was cooled down to −60 °C and held for 10 minutes to insure its complete crystallization (if possible). And then it was heated and cooled from −60 °C to 60 °C at a scan rate of 10 °C min⁻¹. The above steps were repeated twice and then the thermal data of the second heating–cooling scan was analyzed. Thermal stability was measured by thermal gravimetric analysis (TGA, TA Instrument Q5000). Dropping into a platinum pan the sample was heated up to 600 °C under nitrogen atmosphere (heating rate was 10 °C min⁻¹).

The density was tested by weighing 1.00 mL IL in an argon-filled glove box at 25 °C. A Brookfield viscometer (DV-III) and a conductivity meter (DDS-11A) were used to measure the viscosity and conductivity of ILs respectively. The values of viscosity and conductivity were recorded at 5 °C intervals in the temperature range from 25 to 80 °C by using a Brookfield temperature controller (TC-502). The electrochemical stabilities were analyzed by means of linear sweep voltammetry (LSV) in the glove box. Glassy carbon electrode (3 mm diameter) was used as the working electrode and lithium metal was used as both reference electrode and counter electrode. The positive and negative scans of pure IL were carried out separately. Before each scan, the glassy carbon electrode was polished via nano-alumina powder and rinsed by deionized water.

The battery performances of FSI-based (or TFSI-based) IL electrolytes with 0.8 mol L⁻¹ of LiFSI (or LiTFSI) were tested by 2032 coin cells. The cathode was composed of LiFePO₄, acetylene black and PVdF (weight ratio 8 : 1 : 1). All the electrode materials was coated on aluminum foil (battery use) and dried under vacuum at 110 °C. The loading of active material was about 2.0 mg cm⁻² without pressing. Lithium foil was used as anode. The Li/LiFePO₄ coin cells were fabricated in the glove box using the PE separator (SK, 20 μm). The galvanostatic charge–discharge cycling tests were examined by a LAND cell test instrument in the voltage range of 2.0–4.0 V. And the rate current was calculated by using the nominal capacity of 170 mA h g⁻¹. In addition, LSV test and electrochemical impedance spectrum (EIS) were carried out by an electrochemical workstation (CHI600D).

3. Results and discussion

3.1 Physicochemical properties of ILs

Table 1 displayed the properties of four FSI-based ether-functionalized pyrazolium ILs, including melting points, thermal decomposition temperatures, viscosities and conductivities.

Table 1 Properties of these FSI-based ether-functionalized pyrazolium ILs

ILs	Mw^a/g mol^−1	Tm^b/°C	d^c/g cm^−3	η^d/mPa s	σ^e/mS cm^−1	Td^f/°C
a Molecular weight.b Melting point noted from the onset.c Density at 25 °C.d Viscosity at 25 °C.e Conductivity at 25 °C.f Decomposition temperature of 5% weight loss.
PZ2o1-1-FSI	321.31	<−60	1.32	39.8	5.41	262.5
PZ2o1-2-FSI	335.34	<−60	1.28	32.2	6.21	264.5
PZ2o2-1-FSI	335.34	<−60	1.28	39.4	5.09	275.3
PZ2o2-2-FSI	349.37	<−60	1.22	31.7	5.8	265.2

---

It was universally accepted that introducing an ether group into cation could be beneficial to reduce the melting point attributed to the asymmetry of cation, high flexibility and electron-donating effect of ether group.^11,30,31,44 For TFSI-based pyrazolium ILs, it had been proven that ether-functionalization contributed to reduce the melting point.⁴³ Here, all of these FSI-based ether-functionalized pyrazolium ILs did not exhibit any phase transition behavior from −60 to 60 °C. DSC curves of two ILs were illustrated in Fig. S9† as examples. Their melting points (T_m) were defined as “<−60 °C” according to the previous papers,^30,45,46 which were much lower than those of ether-functionalized phosphonium⁷ or tetraalkylammonium¹¹ based on FSI anion. Thus these four new functionalized pyrazolium ILs could be classified as low-melting point ILs.

The TGA curves were illustrated in Fig. 2 which indicated that all ILs had one-stage decomposition behavior and their decomposition temperatures (T_dec) were around 265 °C. Compared to imidazolium (∼225 °C),¹⁵ pyrrolidinium (∼230 °C)⁴⁷ and sulfonium (∼255 °C)¹⁰ ILs based on FSI anion, these pyrazolium ILs had better thermal stability. Beyond the type of cations, anion also could affect T_dec obviously. T_dec of these FSI-based pyrazolium ILs were relatively lower than their TFSI counterparts.^44,48 Similar results were also observed in phosphonium,⁷ tetraalkylammonium¹¹ and pyrrolidinium.^28,47 This was mainly caused by the pyrolysis inclination of FSO₂-group.⁴⁹ Even so, these FSI-based pyrazolium ILs still possessed enough thermal stability to be safe electrolytes when compared to the conventional electrolytes.⁵⁰

Fig. 2 TGA curves of these ILs.

Low viscosity was a crucial property for ILs applied in electrochemical devices because it was propitious to mass transmission significantly.³⁰ Usually, ion size, ion–ion interactions (van der Waals force, electrostatic interaction, etc.) and ion complexes were main factors for the viscosity of ILs. And the decrease of viscosity could also be realized by the substitution of TFSI anion with smaller FSI anion owning to weaker electrostatic and induction interactions between cations and anions.¹⁸ Like imidazolium,¹⁵ pyrrolidinium¹⁶ and phosphonium⁷ ILs, the fluidities of the FSI-based pyrazolium ILs in this work were obviously superior to those of the TFSI analogues. For instance, the viscosity of PZ2o1-1-FSI was only 39.8 mPa s but that of PZ2o1-1-TFSI was 52.1 mPa s at 25 °C.⁴⁴ Considering that their viscosities were all lower than 40 mPa s at ambient temperature, these four FSI-based pyrazolium ILs could belong to a new sort of low-viscosity ILs.

Furthermore, the structure of pyrazolium cation was influential in the viscosity as well. It was found that the ILs containing ethyl group at the N-2 position had minor viscosity when ether group at N-1 position was identical. And ether group at N-1 position (2-methoxyethyl or 2-ethoxyethyl group) did not affect apparently the viscosity. So the viscosity of four FSI-based pyrazolium ILs changed in the order as follows: PZ2o2-2-FSI (31.7 mPa s) ≈ PZ2o1-2-FSI (32.2 mPa s) < PZ2o2-1-FSI (39.4 mPa s) ≈ PZ2o1-1-FSI (39.8 mPa s). This phenomenon also appeared in the TFSI-based pyrazolium^44,48 or imidazolium²⁶ ILs.

Fig. 3(a) illustrated the variation of viscosity with temperature over the range from 25 to 80 °C fitted by Vogel–Tammann–Fulcher (VTF) model (eqn (1)).

herein, the physical meanings of η₀ (mPa s), B (K) and T₀ (K) parameters were high-temperature viscosity limiting value, pseudo-activation energy and ideal glass transition temperature respectively.⁵¹ The values of three parameters and the corresponding fitting coefficient R² were presented in Table 2. The values of R² were all close to 1, which indicated that the temperature dependence of viscosity was shown a good agreement with VTF model. Compared with the corresponding TFSI-based ILs, these FSI-based pyrazolium ILs had smaller B values and bigger T₀ values.^44,48

Fig. 3 VTF plots of (a) viscosity and (b) conductivity.

Table 2 Parameters of VTF equation for viscosity^a

ILs	η0 (mPa s)	B (K)	T0 (K)	R^2
a The percentage standard errors for η0, B and T0 have been included, and R^2 is the VTF fitting parameter.
PZ2o1-1-FSI	0.25 ± 5%	595 ± 2%	180 ± 1%	0.99999
PZ2o1-2-FSI	0.14 ± 5%	740 ± 2%	162 ± 1%	0.99999
PZ2o2-1-FSI	0.14 ± 7%	745 ± 3%	165 ± 1%	0.99999
PZ2o2-2-FSI	0.16 ± 9%	693 ± 4%	166 ± 2%	0.99997

---

Smaller size of FSI anion was contributed to increase the conductivity of ILs. The FSI-based imidazolium,¹⁵ pyrrolidinium¹⁶ and phosphonium⁷ ILs owned higher conductivities than the corresponding TFSI-based ILs. Here, these FSI-based pyrazolium ILs also had higher conductivities than their TFSI counterparts. For example, the conductivity of PZ2o1-1-TFSI was 3.26 mS cm⁻¹,⁴⁴ and the conductivity of PZ2o1-1-FSI was 5.41 mS cm⁻¹ at 25 °C. Moreover, the relationship between conductivity and the structure of pyrazolium cation was complicated. The ILs containing smaller 2-methoxyethyl group at the N-1 position had higher conductivities when alkyl group at N-2 position was identical. It seemed that the smaller size of pyrazolium cation was helpful to improve the conductivity. But the ILs with longer ethyl group at the N-2 position had higher conductivities when ether group at N-1 position was identical. For instance, the conductivity of PZ2o2-2-FSI (5.8 mS cm⁻¹) was higher than that of PZ2o2-1-FSI (5.09 mS cm⁻¹). And the similar rule could be found in TFSI-based imidazolium ILs.²⁶ It was inferred that ion complexes might influence the conductivity besides cation size.⁵²

Temperature dependence of conductivity from 25 to 80 °C was described in Fig. 3(b). A VTF model (eqn (2)) was also used to represent the relationship of conductivity values and temperature where σ₀ (mS cm⁻¹), B (K) and T₀ (K) are three parameters which have similar physical meaning to those in eqn (1).

Table 3 summarized the values of three parameters and the corresponding fitting coefficient R². The temperature dependence of conductivity was also fitted well by VTF model. Compared to the corresponding TFSI-based ILs, FSI-based pyrazolium ILs had smaller B values and bigger T₀ values.^44,48

Table 3 Parameters of VTF equation for conductivity^a

ILs	σ0 (mS cm^−1)	B (K)	T0 (K)	R^2
a The percentage standard errors for σ0, B and T0 have been included, and R^2 is the VTF fitting parameter.
PZ2o1-1-FSI	138 ± 2%	298 ± 2%	206 ± 1%	0.99999
PZ2o1-2-FSI	138 ± 5%	295 ± 4%	203 ± 1%	0.99994
PZ2o2-1-FSI	169 ± 2%	350 ± 1%	198 ± 0.4%	0.99999
PZ2o2-2-FSI	164 ± 4%	350 ± 3%	193 ± 1%	0.99997

---

The LSV curves of four FSI-based pyrazolium ILs were illustrated in Fig. 4. As seen in Table 4, for PZ2o1-1-FSI and PZ2o1-2-FSI, the cathodic limiting potentials were around +1.1 V versus Li/Li⁺ and the anodic limiting potentials were around +5.0 V versus Li/Li⁺. Their electrochemical windows were about 3.9 V. For PZ2o2-1-FSI and PZ2o2-2-FSI, the cathodic and anodic limiting potentials shifted slightly toward positive direction, and the electrochemical windows were about 4.0 V (from +1.2 V to +5.2 V versus Li/Li⁺). It was obvious that the structure of ether group at N-1 position could affect the electrochemical stability of these FSI-based pyrazolium ILs. Furthermore, the electrochemical windows of the corresponding TFSI-based pyrazolium ILs were about 4.4 V (from +1.0 V to +5.4 V versus Li/Li⁺).^44,48 So it meant that substituting FSI anion for TFSI anion would reduce marginally the anodic and cathodic stabilities of pyrazolium ILs. And this regulation was also observed in imidazolium and piperidinium ILs.¹⁶

Fig. 4 LSV curves of these ILs at room temperature. Working electrode: glassy carbon; counter electrode: lithium metal; reference electrode: lithium metal; scan rate: 10 mV s⁻¹.

Table 4 Cathodic and anodic limiting potentials and electrochemical windows values at 25 °C^a

ILs	Cathodic limiting potential (V vs. Li/Li^+)	Anodic limiting potential (V vs. Li/Li^+)	Electrochemical window (V)
a Working electrode: glassy carbon; counter electrode: lithium metal; reference electrode: lithium metal; scan rate: 10 mV s^−1.
PZ2o1-1-FSI	+1.1	+5.0	3.9
PZ2o1-2-FSI	+1.1	+5.0	3.9
PZ2o2-1-FSI	+1.2	+5.2	4.0
PZ2o2-2-FSI	+1.2	+5.2	4.0

---

3.2 Electrochemical performance of Li/LiFePO₄ cells

The charge–discharge tests of Li/LiFePO₄ cells were carried out to assess the cycling and rate performances of these FSI-based pyrazolium IL electrolytes containing 0.8 mol L⁻¹ LiFSI. For comparison, the performances of PZ2o1-1-TFSI and PZ2o2-2-TFSI electrolytes with 0.8 mol L⁻¹ LiTFSI were also examined under the same test condition.

Fig. 5(a) showed the charge–discharge curves of Li/LiFePO₄ cell using PZ2o2-2-FSI electrolyte at 0.1C as a sample, and the discharge capacity and coulombic efficiency during cycling of Li/LiFePO₄ cells using these four IL electrolytes at 0.1C were illustrated in Fig. 5(b). The initial discharge capacities of PZ2o1-1-FSI, PZ2o1-2-FSI, PZ2o2-1-FSI and PZ2o2-2-FSI electrolytes were 158.1, 159.8, 161.8 and 162.5 mA h g⁻¹ respectively. Although their initial capacities did not have apparent difference, these FSI-based IL electrolytes exhibited various capacity retentions of 50 cycles. PZ2o1-2-FSI and PZ2o2-2-FSI electrolytes had the better capacity retentions (99.2% and 99.3%), and the capacity retention of PZ2o2-1-FSI electrolyte was 95.6%. But the discharge capacity of PZ2o1-1-FSI electrolyte only remained 87.9% after 50 cycles. The initial coulombic efficiencies of these four IL electrolytes were in the range from 95% to 97%, and their efficiencies could stabilize over 99% after several cycles. The high and stable efficiency indicated that the capacity decay of PZ2o1-1-FSI electrolyte was not caused by the electrochemical decomposition of electrolyte during the charge–discharge processes. As shown in Fig. S10,† the interfacial resistance of PZ2o1-1-FSI electrolyte increased with the cycling number, while that of PZ2o2-2-FSI electrolyte was extremely stable. Therefore, the inferior cycling performance of PZ2o1-1-FSI electrolyte could be attributed to the gradual increase of interfacial resistance. Furthermore, the discharge capacity and coulombic efficiency of PZ2o1-1-TFSI and PZ2o2-2-TFSI electrolytes at 0.1C were also displayed in Fig. S11(a).† After 50 cycles, the capacity retentions of PZ2o1-1-TFSI and PZ2o2-2-TFSI electrolytes were 99.1% and 99.3%, respectively, and the efficiencies were both above 99% after several cycles. According to the EIS plots illustrated in Fig. S11(b) and (c),† the interfacial resistances of two TFSI-based IL electrolytes varied slightly. The above results manifested that the substitution of TFSI anion with FSI anion could not ensure stable electrode/electrolyte interface and good cycling performance.

Fig. 5 (a) Charge–discharge curves of Li/LiFePO₄ cell using PZ2o2-2-FSI electrolyte at 0.1C, and (b) discharge capacity and coulombic efficiency during cycling of Li/LiFePO₄ cells.

Fig. 6(a) displayed the variation of discharge capacity with rate for Li/LiFePO₄ cells at room temperature, and the discharge capacity was normalized according to the stable value at 0.1C. Among these four FSI-based IL electrolytes, PZ2o2-FSI electrolyte possessed the best rate property, whose discharge curves at different current rates were shown in Fig. 6(b). It can be seen that there was only a small amount of capacity loss when the cell discharged quickly. When the discharge rate increased to 2.0C, the capacity still remained 91.4% retention of that at 0.1C. Even at the current rate of 3.0C, the discharge capacity was 139.4 mA h g⁻¹, which was 88.4% of the capacity at 0.1C. Though the rate properties of PZ2o1-2-FSI and PZ2o2-1-FSI electrolytes were not as well as that of PZ2o2-2-FSI electrolyte, their capacities at 3.0C still remained over 80%. By contrast, the discharge capacity of PZ2o1-1-FSI electrolyte declined rapidly as the current increased, and its capacity remained only 58.6% when the rate rose to 3.0C. EIS was conducted to clarify the interfacial characteristics of Li/LiFePO₄ cells. Fig. 6(c) depicted the EIS plots of Li/LiFePO₄ cells after 5 cycles at 0.1C. The diameter of semicircle was assigned to the interfacial resistance of SEI film. PZ2o2-2-FSI electrolyte had the lower interfacial resistance, and it suggested that the beneficial interfacial characteristics could result in the superior rate property. Moreover, the rate properties of these FSI-based pyrazolium IL electrolytes were better than those of the corresponding TFSI-based IL electrolytes. For example, the capacity of PZ2o1-1-TFSI and PZ2o2-2-TFSI electrolyte at 3.0C remained 54.6% and 65.0% respectively (as shown in Fig. S12†), which were lower than the results of the corresponding FSI-based IL electrolytes. This improvement of rate property was ascribed to the advantage of FSI anion in viscosity.

Fig. 6 (a) Variation of normalized discharge capacity with rate for Li/LiFePO₄ cells, (b) the corresponding discharge curves of Li/LiFePO₄ cell using PZ2o2-2-FSI electrolyte at room temperature, and (c) EIS plots of Li/LiFePO₄ cells after 5 cycles at 0.1C at room temperature. Charge rate was 0.1C and discharge rates were 0.1, 0.2, 0.5, 1.0, 1.5, 2.0, 2.5 and 3.0C.

4. Conclusion

Four new ILs based on ether-functionalized pyrazolium cations and FSI anion were synthesized and characterized. The melting points were lower than −60 °C and the thermal decomposition temperature were higher than 260 °C. Introducing ether group and FSI anion, these ILs owned lower viscosity and higher conductivity. Among of them, PZ2o2-2-FSI possessed the lowest viscosity (31.7 mPa s) while PZ2o1-2-FSI had the highest conductivity (6.21 mS cm⁻¹). The electrochemical windows of four ILs were about 4.0 V. Three IL electrolytes with 0.8 mol L⁻¹ of LiFSI exhibited good cycling performance in the application of Li/LiFePO₄ cells. Thereinto, PZ2o2-2-FSI electrolyte had the best rate property. Its discharge capacity still remained 139.4 mA h g⁻¹ when the discharge rate increased to 3.0C.

Acknowledgements

The authors would like to thank Research Center of Analysis and Measurement of Shanghai Jiao Tong University for their kind help in NMR tests. This work was financially supported by National Natural Science Foundation of China (Grants No. 21373136).

References

1. M. Armand, F. Endres, D. R. MacFarlane, H. Ohno and B. Scrosati, Nat. Mater., 2009, 8, 621–629.
2. J. Kalhoff, G. G. Eshetu, D. Bresser and S. Passerini, ChemSusChem, 2015, 8, 2154–2175.
3. A. Lewandowski and A. Świderska-Mocek, J. Power Sources, 2009, 194, 601–609.
4. V. Etacheri, R. Marom, R. Elazari, G. Salitra and D. Aurbach, Energy Environ. Sci., 2011, 4, 3243.
5. D. R. MacFarlane, N. Tachikawa, M. Forsyth, J. M. Pringle, P. C. Howlett, G. D. Elliott, J. H. Davis, M. Watanabe, P. Simon and C. A. Angell, Energy Environ. Sci., 2014, 7, 232–250.
6. S. Fang, L. Yang, J. Wang, H. Zhang, K. Tachibana and K. Kamijima, J. Power Sources, 2009, 191, 619–622.
7. K. Tsunashima, A. Kawabata, M. Matsumiya, S. Kodama, R. Enomoto, M. Sugiya and Y. Kunugi, Electrochem. Commun., 2011, 13, 178–181.
8. A. Swiderska-Mocek, Electrochim. Acta, 2014, 132, 504–511.
9. M. Galiński, A. Lewandowski and I. Stępniak, Electrochim. Acta, 2006, 51, 5567–5580.
10. H.-B. Han, J. Nie, K. Liu, W.-K. Li, W.-F. Feng, M. Armand, H. Matsumoto and Z.-B. Zhou, Electrochim. Acta, 2010, 55, 1221–1226.
11. H.-B. Han, K. Liu, S.-W. Feng, S.-S. Zhou, W.-F. Feng, J. Nie, H. Li, X.-J. Huang, H. Matsumoto, M. Armand and Z.-B. Zhou, Electrochim. Acta, 2010, 55, 7134–7144.
12. J. Reiter, S. Jeremias, E. Paillard, M. Winter and S. Passerini, Phys. Chem. Chem. Phys., 2013, 15, 2565–2571.
13. G. A. Giffin, N. Laszczynski, S. Jeong, S. Jeremias and S. Passerini, J. Phys. Chem. C, 2013, 117, 24206–24212.
14. J. Reiter, E. Paillard, L. Grande, M. Winter and S. Passerini, Electrochim. Acta, 2013, 91, 101–107.
15. M. Ishikawa, T. Sugimoto, M. Kikuta, E. Ishiko and M. Kono, J. Power Sources, 2006, 162, 658–662.
16. H. Matsumoto, H. Sakaebe, K. Tatsumi, M. Kikuta, E. Ishiko and M. Kono, J. Power Sources, 2006, 160, 1308–1313.
17. K. Tsunashima, Y. Sakai and M. Matsumiya, Electrochem. Commun., 2014, 39, 30–33.
18. S. Tsuzuki, K. Hayamizu and S. Seki, J. Phys. Chem. B, 2010, 114, 16329–16336.
19. A. Guerfi, S. Duchesne, Y. Kobayashi, A. Vijh and K. Zaghib, J. Power Sources, 2008, 175, 866–873.
20. J. S. Moreno, Y. Deguchi, S. Panero, B. Scrosati, H. Ohno, E. Simonetti and G. B. Appetecchi, Electrochim. Acta, 2016, 191, 624–630.
21. L. Jin, P. C. Howlett, J. M. Pringle, J. Janikowski, M. Armand, D. R. MacFarlane and M. Forsyth, Energy Environ. Sci., 2014, 7, 3352–3361.
22. J. H. Davis Jr, Chem. Lett., 2004, 33, 1072–1077.
23. Z. Fei, T. J. Geldbach, D. Zhao and P. J. Dyson, Chem.–Eur. J., 2006, 12, 2122–2130.
24. Z. B. Zhou, H. Matsumoto and K. Tatsumi, Chem.–Eur. J., 2004, 10, 6581–6591.
25. W. A. Henderson, V. G. Young Jr, D. M. Fox, H. C. De Long and P. C. Trulove, Chem. Commun., 2006, 3708–3710,  10.1039/b606381k.
26. J. Zhang, S. Fang, L. Qu, Y. Jin, L. Yang and S.-i. Hirano, Ind. Eng. Chem. Res., 2014, 53, 16633–16643.
27. Z. B. Zhou, H. Matsumoto and K. Tatsumi, Chem.–Eur. J., 2005, 11, 752–766.
28. Z. B. Zhou, H. Matsumoto and K. Tatsumi, Chem.–Eur. J., 2006, 12, 2196–2212.
29. H. Matsumoto, H. Sakaebe and K. Tatsumi, J. Power Sources, 2005, 146, 45–50.
30. K. Tsunashima and M. Sugiya, Electrochem. Commun., 2007, 9, 2353–2358.
31. S. Fang, Y. Tang, X. Tai, L. Yang, K. Tachibana and K. Kamijima, J. Power Sources, 2011, 196, 1433–1441.
32. S. Seki, Y. Kobayashi, H. Miyashiro, Y. Ohno, Y. Mita, A. Usami, N. Terada and M. Watanabe, Electrochem. Solid-State Lett., 2005, 8, A577–A578.
33. S. Seki, Y. Ohno, H. Miyashiro, Y. Kobayashi, A. Usami, Y. Mita, N. Terada, K. Hayamizu, S. Tsuzuki and M. Watanabe, J. Electrochem. Soc., 2008, 155, A421–A427.
34. Y. Abu-Lebdeh, P. J. Alarco and M. Armand, Angew. Chem., Int. Ed., 2003, 42, 4499–4501.
35. P. J. Alarco, Y. Abu-Lebdeh and M. Armand, Solid State Ionics, 2004, 175, 717–720.
36. N. Ishimaru, W. Kubo, T. Kitamura, S. Yanagida, Y. Tsukahara, M. M. Maitani and Y. Wada, Mater. Sci. Eng., B, 2011, 176, 996–1001.
37. N. A. Negm, M. M. Said and S. M. I. Morsy, J. Surfactants Deterg., 2010, 13, 521–528.
38. Y. Han, H. V. Huynh and G. K. Tan, Organometallics, 2007, 26, 6581–6585.
39. C. Chiappe, A. Sanzone, D. Mendola, F. Castiglione, A. Famulari, G. Raos and A. Mele, J. Phys. Chem. B, 2013, 117, 668–676.
40. Y. Abu-Lebdeh, A. Abouimrane, P. J. Alarco and M. Armand, J. Power Sources, 2006, 154, 255–261.
41. S. Seki, T. Kobayashi, N. Serizawa, Y. Kobayashi, K. Takei, H. Miyashiro, K. Hayamizu, S. Tsuzuki, T. Mitsugi, Y. Umebayashi and M. Watanabe, J. Power Sources, 2010, 195, 6207–6211.
42. S. Shen, S. Fang, L. Qu, D. Luo, L. Yang and S.-i. Hirano, RSC Adv., 2015, 5, 93888–93899.
43. M. Chai, Y. Jin, S. Fang, L. Yang, S.-i. Hirano and K. Tachibana, Electrochim. Acta, 2012, 66, 67–74.
44. M. Chai, Y. Jin, S. Fang, L. Yang, S. I. Hirano and K. Tachibana, J. Power Sources, 2012, 216, 323–329.
45. K. Tsunashima, S. Kodama, M. Sugiya and Y. Kunugi, Electrochim. Acta, 2010, 56, 762–766.
46. M. Kärnä, M. Lahtinen and J. Valkonen, J. Mol. Struct., 2009, 922, 64–76.
47. Q. Zhou, W. A. Henderson, G. B. Appetecchi, M. Montanino and S. Passerini, J. Phys. Chem. B, 2008, 112, 13577–13580.
48. J. Zhang, S. Fang, L. Qu, Y. Jin, L. Yang and S. I. Hirano, J. Appl. Electrochem., 2015, 45, 235–244.
49. H. B. Han, Y. X. Zhou, K. Liu, J. Nie, X. J. Huang, M. Armand and Z. B. Zhou, Chem. Lett., 2010, 39, 472–474.
50. N. Wongittharom, T.-C. Lee, I. M. Hung, S.-W. Lee, Y.-C. Wang and J.-K. Chang, J. Mater. Chem. A, 2014, 2, 3613.
51. R. Sescousse, K. A. Le, M. E. Ries and T. Budtova, J. Phys. Chem. B, 2010, 114, 7222–7228.
52. R. Ludwig and D. Paschek, ChemPhysChem, 2009, 10, 516–519.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra16348c

---