1 Introduction

Lithium-ion batteries (LIBs) are required to meet the huge demand for high-specific energy and power, long-term stability, environmental friendliness, safety, etc. [1,2,3,4]. In particular, LIBs should be developed to operate within the daily-life temperature range, i.e., from − 20 to 60 °C [5]. Even though tremendous efforts have been devoted to improving cell performance under wide temperature ranges, stable operation of LIBs at low and high temperatures is very challenging, which mainly originates from the vulnerability of the electrolyte [6, 7]. So far, the direct modification of the electrolyte system, such as the use of various solvents [8,9,10], additives [11,12,13,14,15], salt and its concentration [16,17,18], etc., are yet to be considered the most effective solution that can widen the usable temperature range of batteries. Hence, the rational design of the electrolyte system is critical to operating the cell under a wide temperature range.

The battery performance at a low temperature shows serious degradation mainly due to the poor Li-ion transport [19, 20]. The electrolyte typically suffers from a severe increase in viscosity, which leads to sluggish Li-ion transport. Meanwhile, low boiling points of carbonate-based solvents contribute to the decomposition of the electrolyte during high-temperature operation of LIBs [21]. The evolution of gas phases by decomposition of the electrolyte results in the swelling of the cell, being responsible for the increase in the resistance or cell failure. Hence, LIBs that can operate under a wide temperature range require a rationally designed electrolyte with facile Li-ion mobility, high thermal and electrochemical stability, high-quality solid electrolyte interphase (SEI) and cathode electrolyte interphase (CEI) layer forming capability, etc. The attempts to optimize the composition of the electrolyte as well as explore additives have been suggested by earlier studies. Vinylene carbonate (VC) and fluorinated ethylene carbonate (FEC) are the most successful additives that can aid in forming a high-quality solid-electrolyte-interphase (SEI) layer on the graphite surface [22]. Meanwhile, nitrile-based additives are well known to form the cathode-electrolyte-interphase (CEI) layer on the cathode surface as well as suppress the decomposition of the electrolyte at the elevated temperature [23].

Herein, we scrutinized the effect of the electrolyte solvent composition on the battery performance. We studied the battery behavior with the various electrolyte series blended with the most common carbonate-based solvents. Afterward, the effects of various additives, such as VC, FEC, and nitrile-based organic molecules, were explored to optimize the electrolyte system for stable battery operation under the wide temperature range. Therefore, we could provide the design rule for the electrolyte that enables the cells to operate under a wide temperature range, from − 20 to 60 °C, while minimizing the sacrifice of the battery performance.

2 Experiment Section

2.1 Electrolyte Preparation

Electrolytes were prepared by blending various carbonate-based solvents, ethylene carbonate (EC), ethyl methyl carbonate (EMC), dimethyl carbonate (DMC), and diethyl carbonate (DEC), with fixed salt and its concentration of 1 M LiPF6 as tabulated in Table 1. Base electrolyte series without any additives (Base#1, Base#2, Base#3, Base#4, and Base#5) were prepared with a combination of solvents.

Table 1 Compositions of the electrolyte. Salt and its concentration are fixed as LiPF6 and 1 M, respectively
Full size table

Based on the preliminary battery performance screening of full cells with the base electrolyte system, Base#1, Base#4, and Base#5 electrolytes were selected to be further investigated with various additives. FEC and VC were added to the Base#X (X = 1, 4, and 5) electrolytes, denoted by Base#XF and Base#XV, respectively (Table 1). Then, nitrile-based additives were studied as the additives for the high-temperature stability. In order to understand the effect of the chain length and the cyano-functional group in nitrile-based additives, butyronitrile (BN), succinonitrile (SN), dodecanenitrile with one cyano group (DO1N), and dodecanenitrile with two cyano groups (DO2N), were added to the only FEC added electrolyte, Base#XF (X = 1, 4, and 5) electrolyte, which are referred by Base#XFBN, Base#XFSN, Base#XFDO1N, and Base#XFDO2N, respectively (Table 1). Molecular structures of BN, SN, DO1N, and DO2N are shown in Figure S1.

2.2 Characterization and Electrochemistry

680 mAh pouch-type full cells were manufactured using LiNi0.4Co0.2Mn0.4O2 (NCM424) and natural graphite as the cathode and the anode, respectively. The cathode was prepared using NCM, conductive additive (Super P), and binder (polyvinylidene fluoride (PVDF)) by weight ratio (94: 3: 3). Meanwhile, the anode was composed of natural graphite (97.5 wt%) and binder (styrene butadien rubber (SBR, 1.5 wt%) and carboxymethyl cellulose (CMC, 1 wt%)). Polypropylene membrane was used as a separator. Various electrolytes listed in Table 1 were injected into the cells. Cell assembly procedures were carried out in the dry room. The cells were tested under various conditions ((dis)charging current, temperature, etc.). The viscosity of the electrolyte was measured using the rheometer (Brookfield DV-III Ultra). High-temperature instability was studied by measuring the thickness change of cells that were exposed to the elevated temperature.

3 Results and Discussion

3.1 Effect of the Solvent Composition of the Electrolyte on the Battery Performance

The effect of the base electrolyte composition on the battery performance was studied using 680 mAh full cells. The five different base electrolytes were injected into the full cells. Galvanostatic cycling tests proceeded for 500 cycles at 60 °C as shown in Fig. 1a. In the case of the electrolytes blended with DMC (Base#2 and Base#5), early cell failures after ~ 75th and ~ 200th cycle, respectively, were observed. In these cells, severe swelling of the cells occurred as well, which could trigger cell failure. However, full cells with the electrolytes blended with DEC (Base#3, Base#4, and Base#5) retarded cell failure. In particular, swelling did not occur in the cell with the Base#3 electrolyte. Approximately 50% capacity could be preserved in the cell with the Base#3 electrolyte revealed even after the 500th cycle without cell failure. However, it should be noted that IR drop was observed only in the discharge curve of the cell with the Base#3 electrolyte at the low-temperature (− 23 °C) test (Fig. 1b).

Fig. 1
figure 1

a Cycle-life performance at 60 °C and b discharge capacity retention of the full cells with various electrolytes at  − 23 °C compared with those at RT. c Charge and d discharge capacity retention of the full cells at various current densities compared with those at 0.1 C

Full size image

The disparate temperature behavior of the full cell performance could originate from the boiling point (B.P.) and viscosity of the solvent. In general, the stronger intermolecular force in the liquid is attributed to the higher boiling points and viscosities. At the elevated temperature, B.P. is the most important factor that determines the stability of the electrolyte. B.P.s of DMC and DEC are 90 °C and 126 °C, respectively [24, 25]. Hence, the cells with DEC showed relatively stable long-term operation at the elevated temperature. Meanwhile, the viscosity is critical to the battery response at a low temperature. The mobility of Li ions is impeded due to high viscosity at − 23 °C.[26] Despite very similar discharge curves among the cells, only the cell with the electrolytes blended by EC and DEC, i.e., a combination of two solvents with the highest B.P. and viscosity, revealed the IR drop. The viscosities and ion conductivities of the electrolyte solution were also measured, as depicted in Figure S2, which is consistent with our observation.

The swelling of the cell is directly related to the gas evolution resulting from the decomposition of the electrolyte, SEI, and CEI layer, etc., during the high-temperature exposure of the cell. To investigate the gas evolution of the cells at a high temperature, the cell thicknesses were measured after thermal treatment, as presented in Fig. 2. Cells were stored in the convection oven for 1 h at 23 °C. Then, the temperature was ramped to 90 °C for 1 h and maintained for 4 h. After high-temperature exposure of the cells, the temperature was cooled down to RT for 1 h. Cells with the electrolytes blended by DEC, Base#3 and #4, exhibited thickness increases between 12.5 and 40%, as shown in Fig. 2a. In contrast, the thickness of the cell with the electrolyte blended by DMC, Base#2, was increased by ~ 211%. Because the thickness change is commensurate with the gas evolution during cycling at the elevated temperature, our result strongly suggests that the high-temperature safety of the cell is directly attributed to the B.P. of the electrolyte solvent in the case of using the electrolyte without additives.

Fig. 2
figure 2

The thickness change of the full cell by exposure to the high temperature (90 °C). Comparison of thickness changes among the full cells with a base, b FEC and VC added, c Base#1-based nitrile-additives added, d Base#4-based nitrile-additives added, e Base#5-based nitrile-additives added, and f FEC and BN added electrolytes

Full size image

In order to study the kinetics of the electrolytes, charge and discharge capacity retention were evaluated by varying current densities from 0.2 to 5 C at 23 °C as depicted in Fig. 1c and d, respectively. The most thermally stable cell, the cell with the Base#3 electrolyte, revealed the most sluggish response from both the charge and discharge process. The ~ 68% and ~ 72% capacities (vs. those of 0.1 C) could be delivered during the charge and discharge process at 5 C, respectively, in the cell with the Base#3 electrolyte. The cells with the electrolyte blended by DMC showed a faster response and 75% capacity retention (vs. those of 0.1 C) in both the charge and discharge process at 5 C, compared with other cells. Namely, the trend of the cells with various electrolytes in the rate capability is consistent with the ionic conductivities of the electrolyte system (Figure S2) but is completely opposite to that in the thermal stability test (Fig. 1a).

3.2 Effect of SEI Layer Forming Additives on the Electrolyte Stability

The compositions of the electrolyte solution were selected by the preliminary battery test at the elevated temperature. The electrolytes with DMC (Base#2 and Base#4) were excluded. The further optimization of the electrolyte is allowed by additives, such as fluorinated ethylene carbonate (FEC) and vinylene carbonate (VC), which are well known to participate in the formation of the high-quality SEI layer on the graphite surface. FEC and VC are decomposed before that of EC during the initial charging process, forming an SEI layer on the graphite anode. Thereby, FEC and VC are oftentimes used as additives to improve not only capacity retention but also rate capability by suppressing the direct reaction of the anode with the electrolyte followed by degradation of graphite structure [27]. FEC and VC suffer from cathodic decomposition before EC, considering their low LUMO values [28]. The decomposed FEC and VC lead to the formation of the high-quality SEI layer, and thus preventing the decomposition reaction of EC and/or degradation of the graphite structure. The potentials for the cathodic decomposition could be estimated through the differential capacity (dQ/dV) vs. voltage plots, which confirms 2.7 V, 3 V, and 3.1 V for FEC, VC, and EC, respectively (Figure S3).

All cells with the FEC- and VC-added electrolytes showed excellent capacity retention during 500 cycles at RT (Fig. 3a). Full cells with the Base#1 electrolyte system with FEC and VC (Base#1F and Base#1 V, respectively) showed the better capacity retention corresponding to ~ 97% of the initial capacity. The slightly lower capacity retention, 94–97%, could be achieved from cells with the Base#4 and Base#5 electrolytes with FEC and VC. To evaluate low-temperature battery performance, cells that were fully charged to 4.2 V at 23 °C were discharged at − 23 °C. Both charge and discharge current densities were 0.2 C. Since FEC is reported to form a thinner SEI layer than VC, FEC could lead to faster kinetics of cells at low temperatures compared with VC [29]. The discharge capacities of the cells with the VC-added electrolyte were similar to those with the electrolytes without SEI layer-forming additives. However, the higher discharge capacity and less IR drop could be observed in the cell with FEC added electrolyte than others (Fig. 3b). As aforementioned, battery performance at a low temperature is highly dependent on the conductivity (or viscosity).

Fig. 3
figure 3

a Discharge capacity retention of the cells with the FEC and VC added Base#1, Base#4, and Base#5 electrolytes for 500 cycles at RT. b Discharge capacity retention of cells with FEC and VC added electrolytes at − 23 °C compared with those at RT

Full size image

In addition, the high-quality SEI layer that originates from FEC and VC additives reduces the resistance of the cell, resulting in higher rate capability [30, 31]. All cells showed enhanced capacity retention when FEC and VC were added to the electrolytes at the accelerated current rates (Fig. 4). The cells with the Base#1 and Base#5 electrolyte series showed more improvement in the rate capability. In the case of the cells with the Base#1 series electrolyte system (Base#1, Base#1F, and Base#1 V), the charge capacity retentions of the cells with FEC and VC electrolyte additives at 10 C (vs. those at 0.1 C) were 47% and 44% higher, respectively, compared with those without electrolyte additive. FEC and VC additives also led to a 12% higher discharge capacity retention in the cell at 10 C. In the case of the cells with the Base#4 series electrolyte system (Base#4, Base#4F, and Base#4 V), the charge capacity retention at 10 C was ~ 36% higher in the cells using the electrolytes with additives (Base#4F and Base#4 V) than the cell using the electrolyte without additives (Base#4).

Fig. 4
figure 4

Charge capacity retention of the full cells with FEC and VC added a Base#1, b Base#4, and c Base#5 electrolyte system at various current densities compared with those at 0.1 C. d Comparison of charge capacity retention of the cells with the FEC and VC added electrolyte system at various currents rates

Full size image

Discharge capacity retention of the cells at 10 C was increased by 6% and 4% if FEC and VC were added into the Base#4 electrolyte, respectively (Fig. 5). FEC and VC additives in the Base#5 electrolyte system also showed the increase in the capacity retention at 10 C. Charge capacity retention of the cells with electrolyte additives at 10 C was increased by 48% compared with that without electrolyte additive. Discharge capacity retentions at 10 C were also increased by 7% and 4% in cells using the Base#5 electrolyte with FEC and VC, respectively. Based on the rate capability test, the FEC additive contributed to a slightly faster response than the VC additive, regardless of the composition of the electrolyte. The resistance of the SEI layer originated by FEC seems to be lower than that by VC. Meanwhile, cells with Base#1F or Base#4F electrolyte systems showed ~ 9% and ~ 10% increase in charging and discharging rate capabilities, respectively, compared with the cell with the Base#4F electrolyte. Such difference can be highly dependent on the ionic conductivity of the solvent in the electrolyte, as presented in Figure S4.

Fig. 5
figure 5

Discharge capacity retention of the full cells with FEC and VC added a Base#1, b Base#4, and c Base#5 electrolyte system at various current densities compared with those at 0.1 C. d Comparison of charge capacity retention of the cells with FEC and VC added electrolyte system at various currents rates

Full size image

In order to study the effect of the SEI-layer forming additives on the high-temperature cycling, cells were cycled at the 1 C rate. Galvanostatic cycling of cells was carried out at 60°C. All cells with electrolyte additives exhibited earlier cell failure than counterparts without electrolyte additives (Fig. 6). The earlier cell failure accounted for the accelerated gas evolution by FEC and VC. The SEI layers originating from FEC and VC can contain many inorganic compounds, readily decomposed into the gas phases during high-temperature operation. The swelling experiment was also performed in the same manner above. After high-temperature exposure, thickness changes of the cells were monitored (Fig. 2b). Indeed, FEC and VC led to the larger thickness increase by the more gas evolution [32, 33]. Overall, a greater amount of gas evolution occurs in the cell with the Base#5 electrolyte series than those with the Base#1 and Base#4 electrolyte systems, possibly owing to the lower B.P. of the electrolyte solvent.

Fig. 6
figure 6

Cycle-life performance at 60 °C of the full cells with FEC and VC added electrolytes. All cells showed earlier cell failure compared with the cells with electrolytes without additives

Full size image

3.3 Study about Nitrile-Based Additives for High-Temperature Stability

The high-temperature gas generation by the use of additives like FEC and VC should be overcome for the stable operation of Li-ion batteries. In order to enhance high-temperature stability in the electrolyte solution, nitrile-based additives were further added to the selected electrolyte system. Only FEC-added electrolyte systems (Base#1F, Base#4F, and Base#5F) have been selected. Nitrile-based additives are relatively stable in the electrolyte and are considered excellent electrolyte additives since they only protect the cathode surface without damage due to the SEI layer at the graphite surface [34, 35]. Cyano-functional groups (–CN) of nitrile-based additives form a complex with the transition metal of the cathode, mitigating the side reaction that can occur by the reaction between the electrolyte and the cathode [36].

Galvanostatic cyclings of full cells with nitrile-based additives were performed at the 1 C rate. It should be noted that ~ 10% decrease in capacity retention after 250 cycles was observed in the cells with nitrile-based electrolyte additives compared with those only with FEC additives (Fig. 7). Nitrile-based additives deteriorated the capacity retention due to the formation of the cathode electrolyte interphase (CEI) film on the cathode surface. The conduction of Li+ is sluggish at the CEI layer, which gradually degrades the capacity retention. Interestingly, capacity retention is higher as the chain length and cyano functional group (–CN) of the nitrile-based additives become smaller and less, respectively, as shown in Fig. 7a, b, and c.[37,38,39] Among various nitrile-based additives, butyronitrile additives resulted in higher capacity retention (Fig. 7d).

Fig. 7
figure 7

Discharge capacity retention of the cells with the FEC and nitrile-based additives added a Base#1, b Base#4, and c Base#5 electrolyte system for 250 cycles at RT. d Comparison of discharge capacity retention of the cells with FEC and BN added electrolyte system

Full size image

Low-temperature discharge capacities of full cells were also investigated. Cells using the electrolyte with only FEC additive delivered the highest capacities (Fig. 8). All cells exhibited lower capacities if nitrile-based additives were added to the electrolyte with FEC. This can be understood by the lower viscosity of the electrolyte by nitrile additives. The greater the viscosity of the electrolyte, the smaller the Li-ion conductivity. The linear chain structure of the nitrile-based additives renders the electrolyte viscous, as shown in Fig. 8a, b, and c. Therefore, the butyronitrile (BN) additive with the shorter chain length additives resulted in higher capacity retention in all cells (Fig. 8d).

Fig. 8
figure 8

Discharge capacity retention of the cells with the FEC and nitrile-based additives added a Base#1, b Base#4, and c Base#5 electrolyte system at − 23 °C compared with those at RT. d Comparison of discharge capacity retention of the cells with FEC and BN added electrolyte system at − 23 °C

Full size image

Furthermore, rate charge and discharge capabilities were also studied from 0.2 to 10 C at 23 °C, as presented in Figs. 9 and 10, respectively. When nitrile-based additives were used, the rate capability was degraded due to the impedance of the CEI layer. The cells with shorter chain lengths and the smaller cyano-functional group (–CN) of the nitrile-based additives exhibited higher capacities during both the charging and discharging processes. Therefore, degradation was small or negligible in the case of using BN as the additive, which is also confirmed by the ionic conductivity measurement (Figure S5).

Fig. 9
figure 9

Charge capacity retention of the full cells with FEC and nitrile-based additives added a Base#1, b Base#4, and c Base#5 electrolyte system at various current densities compared with those at 0.1 C. d Comparison of charge capacity retention of the cells with FEC and BN added electrolyte system at various currents rates

Full size image
Fig. 10
figure 10

Discharge capacity retention of the full cells with FEC and nitrile-based additives added a Base#1, b Base#4, and c Base#5 electrolyte system at various current densities compared with those at 0.1 C. d Comparison of discharge capacity retention of the cells with FEC and BN added electrolyte system at various currents rates

Full size image

The swelling tests of the 4.2 V charged cells with nitrile-based additives at 90 °C are presented in Fig. 2. When the nitrile-based additives with the longer chain length were added to the electrolyte, less swelling occured in the cells with Base#1 and Base#4 electrolyte series at the high temperature (Fig. 2c and d). In contrast, similar swelling occurred in the cells with Base#5 electrolyte system regardless of nitrile-based additives due to the presence of DMC in the electrolyte solvent (Fig. 2e). BN additives resulted in 7% less swelling compared with the FEC-added Base#5 electrolyte. In the meantime, significantly less swelling by BN additives in Base#1 and Base#4 electrolytes, 38% and 16%, was achieved compared with FEC-added Base#1 and Base#4 electrolytes, respectively (Fig. 2f). Considering that the swelling test is one of the standards to assess the high-temperature safety of batteries, the nitrile-based additives are necessary to ensure high-temperature safety of Li-ion batteries.

Next, the safety of cells at a high temperature was further investigated. All cells did not show explosion or leakage when charged to 4.2 V and stored at 160 °C for 1 h (Figures S6, S7, and S8). Lastly, in order to understand the thermal behavior of the cathode, the DSC test was carried out. The 4.2 V charged full cells at 0.2 C were disassembled in a dry room, and cathodes were carefully harvested. The exothermic reaction of the cathode with the nitrile-based additives revealed a higher onset temperature by ~ 10–~ 15 °C, which confirms the high-temperature stability of the cathode due to the CEI layer on the surface (Figure S9).

4 Conclusion

In our work, we investigated various additives for the electrolyte of Li-ion batteries. In order to form the high-quality SEI layer, FEC was added to the electrolyte, which improved the charge and discharge capacity retention of the cells at 10 C by 47% and 12%, respectively, compared with the cells without FEC. In addition, the cells with FEC additives exhibited only 6% capacity loss after 500 cycles at the 1 C rate. Despite significant enhancement of the battery performance, FEC led to the severe degradation of high-temperature stability. Though nitrile-based additives slightly deteriorate the battery performance, they ameliorated the high-temperature instability of cells. We unveiled that the shorter chain length and the smaller cyano-functional group in the molecular structures of nitrile-based additives resulted in the higher thermal stability of cells. We unveiled that the composition and additive of the electrolyte could tune the characteristics of Li-ion batteries. Therefore, we believe that our work can provide deep insight into the design of the electrolyte of Li-ion batteries for the desired application.