Asymmetric fluorinated sulfonamide electrolyte solvent for high-voltage and wide-temperature Li-metal batteries

Abstract

Stable cycling of Ni-rich positive electrodes under high voltages (≥ 4.6 V) is critically challenging due to severe interfacial parasitic reactions and structural degradation, especially under all-climate conditions (i.e., ≤ −20 °C and ≥ 55 °C). Herein, we rationally design and synthesize an asymmetric fluorinated sulfonamide solvent, N-ethyl-N-methyltrifluoromethane-sulfonamide, where the fluorination of sulfone backbone greatly promotes the oxidation stability, while the introduction of asymmetric amine group bestows the solvent with an low freezing point (below −150 °C). This obtained solvent, with 1 M lithium bis(fluorosulfonyl)imide, delivers high oxidation stability (above 5.0 V) and good Li reversibility (up to 99.8% in Li | |Cu cells). Coupled with its ability to form highly robust LiF-rich solid-electrolyte/cathode-electrolyte interphases, this single-salt single-solvent electrolyte enables stable cycling of Li | |Ni_0.8Co_0.1Mn_0.1 cells under an high voltage of 4.8 V across an wide temperature of −60 to 100 °C. Remarkably, a prototype 5 Ah-level Li | | Ni_0.8Co_0.1Mn_0.1 pouch cell at an cut-off voltage of 4.7 V and 0.1 C demonstrates a high specific energy (based on the mass of all components) of 514 Wh kg⁻¹ over 20 cycles.

Stable cycling of Ni-rich positive electrodes at high-voltage (≥ 4.6 V) remains challenging. Here, authors propose a single-salt single-solvent electrolyte based on an asymmetric fluorinated sulfonamide solvent, which sustains stable cycling of 4.8 V LillNCM811 cells from −60 to 100 °C.

Introduction

Due to the specific energy limit of approximately 250 Wh kg⁻¹, the state-of-the-art lithium-ion batteries (LIBs) fail to satisfy the pressing demand for next-generation high-energy-density battery technologies¹. Lithium metal batteries (LMBs) have been considered one of the most promising alternative high-energy-density energy storage technologies (exceeding 500 Wh kg⁻¹). These batteries typically couples with high-energy-density positive electrode (e.g., Ni-rich layered oxide positive electrodes (LiNi_xCo_yMn_1-x-yO₂), abbreviated as NCMs) and Li-metal^2–4. Especially, elevating the upper cut-off voltage for NCMs can enhance both the discharge capacity (C + ΔC) and working voltage (V + ΔV), which significantly promotes the overall specific energy (E_high = (C + ΔC) × (V + ΔV) in Supplementary Fig. 1). Despite considerable progress has been made, achieving stable operation of LMBs under high voltages (≥ 4.6 V) and all-climate conditions (especially ≤ −20 °C and ≥ 55 °C) remains a critical challenge.

From the perspective of battery configuration, two primary issues must be addressed during the design of high-energy-density LMBs (Supplementary Fig. 2). On the negative electrode side, a stable Li plating/stripping process with high Coulombic efficiency (CE) has been recognized as a critical requirement for achieving high reversibility of Li-metal. The formation of a surface layer on Li-metal, referred to as the solid−electrolyte interphase (SEI), is ascribed to the higher Fermi energy level of Li-metal vs. the lowest unoccupied molecular orbital (LUMO) level of solvents^4,5. This SEI must simultaneously meet the following critical prerequisites^6–10, including (i) high Li⁺ conductivity with electronically insulative to facilitate rapid ion transport while blocking electron conduction; (ii) good mechanical properties to suppress the growth of Li dendrites; (iii) good interfacial compatibility to accommodate the volume expansion of Li-metal; and (iv) favorable electrochemical stability to prevent irreversible reactions with electrolytes. On the positive electrode side, NCMs with agglomerated polycrystalline structures have emerged as the predominant choice for high-energy-density batteries¹¹. Concerning the failure mechanism of NCMs, the transition metal (TM) dissolution represents a significant factor that warrants attention. This dissolution further leads to interfacial structural degradation, thereby exposing fresh surfaces to react with the electrolytes, resulting in the formation of a thick and/or uneven cathode–electrolyte interphase (CEI)^12,13. Moreover, the irreversible phase transition from a layered phase to inactive rock-salt phase may exacerbate the capacity fading, especially at high cut-off voltages exceeding 4.3 V^14,15. Given the aforementioned challenges associated with both the negative electrode and positive electrode sides, the electrode–electrolyte interphase plays a vital role to implement high-energy-density LMBs, which needs to re-scrutinize the current electrolyte chemistry^16,17.

Carbonate-based electrolytes with regular concentration (~1 M) have been successfully applied in LIBs, while they are not suitable for LMBs due to their high reactivity with Li-metal¹⁸. Ether-based electrolytes have been progressively employed in LMBs over the last decade, primarily due to their reliable reductive stability¹⁹. While the inferior oxidation stability (below 4.2 V) restricts their application in high-energy-density LMBs²⁰. It is urgent to develop new electrolyte systems to fulfill the increasing demand of high-energy-density LMBs. Recently, various electrolyte design strategies have been proposed to enhance the compatibility with high-voltage LMBs. For instance, steric hindrance was introduced to the ether solvent to weaken their solvation strength, which not only promotes the Li⁺–anion interactions to generate anion-derived SEIs but also increases the oxidation stability^21,22. Moreover, the introduction of fluorinated groups with strong electron-withdrawing property to the ether solvents^23–28 was shown to significantly enhance their oxidative stability and further enhance Li compatibility. These explorations aim to expand battery longevity through the formation of SEIs/CEIs containing LiF. Alternative to the well-explored carbonate and ether solvents, some sulfur-contained solvents including sulfone-²⁹, sulfoxide-³⁰, and sulfonamide-based solvents^31,32 exhibit high oxidation stability and enable to form robust sulfur-containing interphases. Yet, so far, the reported charging cut-off voltages in these electrolytes have been mostly limited to ≤ 4.6 V. Besides, the electrolyte oxidation and Li compatibility issues will be exacerbated when operating the LMBs across a wide temperature range (especially ≤ −20 °C and ≥ 55 °C).

Herein, we demonstrate a single-salt single-solvent electrolyte that enables the stable operation of LiNi_0.8Mn_0.1Co_0.1O₂ (NCM811) LMBs under high voltage (4.8 V) across a wide temperature range (−60 to 100 °C). As shown in Fig. 1a, sulfone features good oxidation stability was selected as backbone, the introduction of an electron-withdrawing –CF₃ moiety is expected to further promote the oxidation stability, whereas the amination aims to elevate the Li-metal compatibility. Next, by incorporating asymmetry, we designed and synthesized an asymmetric fluorinated sulfonamide, namely N-ethyl-N-methyltrifluoromethane-sulfonamide (FEMS), with the aims to increase the disorder intermolecular arrangement, leading to low freezing point and low Li⁺ desolvation barrier. When paired with 1 M lithium bis(fluorosulfonyl)imide (LiFSI), the resulting single-salt single-solvent LiFSI-FEMS electrolyte not only possesses good oxidation stability (above 5.0 V) but also enables highly reversible Li plating/stripping with a high Li | |Cu CE of 99.8% at 30 °C. Furthermore, it also promises the stable operation of Li | |NCM811 cells under an high voltage of 4.8 V and wide temperature ranging from −60 to 100 °C. Remarkably, a prototype 5 Ah-level Li | |NCM811 pouch cell is also examined at a cut-off voltage of 4.7 V, which shows a high specific energy of 514 Wh kg⁻¹ over 20 cycles.

Fig. 1. Design rationale and solvation structure of the asymmetric electrolyte.

a Design rationale of asymmetric fluorinated sulfonamide solvent. RDFs calculated from MD simulations of LiFSI-FDMS and LiFSI-FEMS electrolytes at b 30 °C and c −40 °C. Distributions of primary solvation sheath compositions of LiFSI-FDMS and LiFSI-FEMS electrolytes at d 30 °C and e −40 °C.

Results

Solvent asymmetry-induced low freezing point

As confirmed by ¹H, ¹³C, and ⁹F NMR spectra, two asymmetric fluorinated sulfonamides, including FEMS and N-propyl-N-methyltrifluoromethane-sulfonamide (FDPS), were successfully synthesized by the reaction of trifluoromethanesulfonyl chloride with N-ethly-N-methylamine (Supplementary Fig. 3) and N-propyl-N-methylamine (Supplementary Fig. 4), respectively. The FDPS solvent, characterized by its highly asymmetric structure with methyl/propylamine, possesses significant steric hindrance, resulting in a limited solubility for lithium bis(fluorosulfonyl)imide (LiFSI) with only up to 0.6 M (Supplementary Fig. 5). This low solubility inevitably reduces the electrolyte’s ionic conductivity (Supplementary Fig. 6). To balance the conflict between solvent asymmetry and ionic conductivity, we focus on the investigation of LiFSI in FEMS electrolyte (denoted as LiFSI-FEMS), which features a methyl/ethylamine asymmetric structure.

To validate the design principle of asymmetry-induced low freezing point, differential scanning calorimetry (DSC) was applied to assess the freezing point of various solvents and their corresponding electrolytes. As shown in Supplementary Fig. 7a, the electrolyte of 1 M LiPF₆ in EC/DEC (denoted as LiPF₆-EC/DEC) exhibits two distinct endothermic peaks at −24.4 °C and 5.2 °C, respectively, due to the high melting point of EC. Meanwhile, the electrolyte of 1 M LiFSI in symmetric FDMS solvent (denoted as LiFSI-FDMS) shows an obvious endothermic peak around −55.5 °C, indicating a low melting point of the electrolyte. In sharp contrast, the pure FEMS solvent, with an asymmetric methyl/ethylamine structure, shows no endothermic peak even at temperature as low as −150 °C, indicating it remains liquid state at this temperature (Supplementary Fig. 7b). Similarly, no obvious endothermic peak can be observed for the LiFSI-FEMS electrolyte within the testing temperature range (down to −150 °C). More intuitively (Supplementary Fig. 7c), LiPF₆-EC/DEC and LiFSI-FDMS rapidly freeze after being stored at −20 °C and −60 °C, respectively, while the LiFSI-FEMS maintains invisible liquidity even at −100 °C. Furthermore, the LiFSI-FEMS still demonstrates low viscosity (below 2.5 mPa s) even at −60 °C (Supplementary Fig. 7d), holding great promise for low temperature application. Furthermore, the thermogravimetry analysis (TGA) demonstrates that the LiFSI-FEMS shows the lowest mass loss of 4.2% at 100 °C (Supplementary Fig. 8), indicating its good stability at high-temperature, which is favorable for wide temperature operation.

Solvation structure and physicochemical properties of the electrolytes

⁷Li NMR measurement was performed to investigate the Li⁺ coordination with solvents in the different electrolytes at the same concentration of 0.6 M (Supplementary Fig. 9), where higher chemical shift represents less shielding effect of Li nuclei due to the weak coordination of the surrounding solvents. The LiFSI-FDMS exhibits the lowest chemical shift of −1.22 ppm, indicating the strongest interaction between Li⁺ and FDMS. By gradually incorporating asymmetry to the solvents, the ⁷Li chemical shift was gradually increased to −1.21 ppm for LiFSI-FEMS and further to −1.15 ppm for LiFSI-FDPS electrolyte. These results indicate that the interaction of Li⁺–solvent decreases as the degree of solvent asymmetry increases, leading to a weakly solvated structure that is expected to facilitate the desolvation process of Li⁺ solvates.

To investigate the solvation structures of various electrolytes, molecular dynamics (MD) simulations were performed and the radial distribution functions (RDFs) of both the anion and solvent with respect to Li⁺ was calculated for each electrolyte. As shown in Supplementary Fig. 10, for LiPF₆-EC/DEC electrolyte, the average coordination environment of Li⁺ within 3 Å is found to be Li⁺(O-EC)_3.97(O-DEC)_0.04 at 30 °C, demonstrating the strong coordination of EC with Li⁺ in the first solvation sheath. This EC-dominated solvation structure not only hinders the desolvation at the interphase, but also tends to form an organic-rich SEI layer with high interfacial impedance and poor stability. In contrast, both LiFSI-FDMS and LiFSI-FEMS exhibit anion-rich solvation structure with the average coordination environment calculated to be Li⁺(O-FDMS)_1.64(O-FSI⁻)_2.75 and Li⁺(O-FEMS)_1.54(O-FSI⁻)_3.13, respectively (Fig. 1b and Supplementary Fig. 11a, b and Supplementary Data 1–3). Specifically, the LiFSI-FEMS electrolyte exhibits a much higher FSI⁻/solvent ratio (i.e., 2.03) than that of LiFSI-FDMS (i.e., 1.68). According to the MD statistics (Fig. 1d), 59% first solvation sheath of LiFSI-FEMS electrolyte is dominated by three or more FSI⁻ anions, which is much higher than that of LiFSI-FDMS electrolyte (i.e., 42%), leading to more anion-derived inorganic SEI for LiFSI-FEMS.

When the temperature was reduced to −40 °C, the difference of Li⁺ coordination environment become more significant for these two electrolytes. As shown in Fig. 1c and Supplementary Fig. 11c, d, and Supplementary Data 4–5, the LiFSI-FEMS still possesses an anion-rich solvation structure of Li⁺(O-FEMS)_2.14(O-FSI⁻)_2.19, whereas the first solvation sheath of LiFSI-FDMS is changed to solvent-rich solvation structure of Li⁺(O-FDMS)_2.59(O-FSI⁻)_1.68. Furthermore, the MD statistic results reveal that the LiFSI-FEMS electrolyte still maintains a high proportion of 49% solvates with three or more FSI⁻ anions, while only 16% is remained for LiFSI-FDMS electrolyte (Fig. 1e). These results indicate that the asymmetric structure can effectively mitigate the Li⁺–solvent interaction, especially under low temperatures, favoring the desolvation process and formation of anion-derived SEI.

The binding energy between Li⁺ and solvent in each solvation sheath plays a critical role since that it directly influences the desolvation process near the electrode surface. When the solvated Li⁺ moves towards the negative electrode, the anion would depart first easily due to the electrostatic repulsion, rendering the desolvation energy between Li⁺ and the anion negligible³³. The desolvation energies for the remaining Li⁺(solvent)_x complexes were calculated using density functional theory (DFT) based on their distributions and respective binding energies (Supplementary Fig. 12 and Supplementary Data 6–17), yielding a desolvation energy of 205.6 kJ mol⁻¹ and 148.2 kJ mol⁻¹ for LiFSI-FDMS and LiFSI-FEMS electrolytes, respectively at 30 °C. At low temperature (i.e., −40 °C), the desolvation process becomes the domination step for Li deposition, which is much lower for LiFSI-FEMS (i.e., 186.5 kJ mol⁻¹) than that LiFSI-FDMS (i.e., 295.8 kJ mol⁻¹), highlighting the great advantage of asymmetric FEMS in enabling low temperature operation.

As shown in Supplementary Fig. 13, although the ionic conductivities of LiPF₆-EC/DEC are an order of magnitude higher than that of LiFSI-FEMS from 0 °C to 60 °C, the ionic conductivity of LiPF₆-EC/DEC decreases sharply at −20 °C due to its solidification (Supplementary Fig. 7b, c). Notably, LiFSI-FEMS still exhibits a relatively high ionic conductivity of 0.07 mS cm⁻¹ at −40 °C and 0.023 mS cm⁻¹ at −60 °C, enabling the low temperature operation. Furthermore, as shown in Supplementary Fig. 14, the LiFSI-FEMS electrolyte delivers a significantly higher Li⁺ transference number (0.72) compared to that of LiPF₆-EC/DEC (0.35), which can facilitate the uniform Li⁺ deposition and effectively suppress the growth of Li dendrites by mitigating concentration polarization³⁴. Moreover, the electrochemical impedance of the Li | |Li cells, including the resistances of solvated Li⁺ transport in the bulk electrolyte, the desolvation of solvated Li⁺, and Li⁺ migration through the SEI, were carefully investigated at various temperatures. As shown in Supplementary Fig. 15, compared with LiPF₆-EC/DEC at 20 °C, although the LiFSI-FEMS electrolyte exhibits higher electrolyte resistance (R_el) due to its lower ionic conductivity, the total resistance can be fully compensated by its much lower desolvation and SEI resistances (R_des+R_SEI). When the temperature was decreased to 0 °C and further to −20 °C, the advantage of the LiFSI-FEMS electrolyte becomes more significant with much lower total resistance, demonstrating the unique advantages of this asymmetric electrolyte (i.e., low freezing point, low desolvation and SEI resistances) for enabling low temperature LMBs.

Next, the safety performance of various electrolytes was assessed through combustion tests. As shown in Supplementary Fig. 16, the LiPF₆-EC/DEC electrolyte immediately catches fire on ignition and continue burns fiercely even after the fire torch was removed, while LiFSI-FEMS electrolyte shows no signs of combustion even after several ignitions, demonstrating its intrinsic non-flammability. Besides, LiPF₆-EC/DEC electrolyte exhibits poor wettability with a contact angle of 51.4°, whereas LiFSI-FEMS electrolyte displays favorable electrolyte wettability with a smaller contact angle of 32.0° (Supplementary Fig. 17).

Highly reversible Li plating/stripping across wide temperatures

The compatibility between electrolytes and Li-metal was first evaluated in Li | |Cu asymmetric cells. As shown in Fig. 2a, Li | |Cu cell using LiPF₆-EC/DEC electrolyte exhibits rapid CE decay due to the detrimental interfacial reactions between Li-metal and carbonate solvents, leading to cell failure within 50 cycles. Conversely, Li | |Cu cell using LiFSI-FEMS electrolyte displays good compatibility with a high average CE of 99.2% over 550 cycles. We further employed the Aurbach method to evaluate the Li plating/stripping CEs. As shown in Fig. 2b, the LiFSI-FEMS electrolyte achieves an high CE of 99.8% at 30 °C, far exceeding that of LiPF₆-EC/DEC electrolyte (84.6%). In the initial plating curve shown in Supplementary Fig. 18, compared to the large nucleation overpotential using LiPF₆-EC/DEC electrolyte (145.7 mV), LiFSI-FEMS electrolyte enables a much lower nucleation overpotential (34.6 mV), implying a lower nucleation barrier of Li³⁵.

Fig. 2. Li-metal reversibility at various temperatures.

a Long-term cycling test of Li | |Cu cells with LiPF₆-EC/DEC and LiFSI-FEMS electrolytes at a current density of 1.0 mA cm⁻² for 1.0 mAh cm⁻² at 30 °C. b Modified Aurbach measurement of Li | |Cu CEs in LiPF₆-EC/DEC and LiFSI-FEMS electrolytes at 30 °C. c Modified Aurbach measurement of Li | |Cu CEs in LiFSI-FEMS electrolyte at 55 °C at a current density of 0.5 mA cm⁻² for 1.0 mAh cm⁻², −40 °C at a current density of 0.5 mA cm⁻² for 1.0 mAh cm⁻², and −60 °C at a current density of 0.1 mA cm⁻² for 0.2 mAh cm⁻². d Long-term cycling test of Li | |Li symmetric cells using LiPF₆-EC/DEC and LiFSI-FEMS electrolytes at a current density of 0.5 mA cm⁻² for 1.0 mAh cm⁻² at 30 °C. Insert: selected polarization voltage profile from 2940 to 3000 h using LiFSI-FEMS electrolyte.

The ion-transfer kinetic was first evaluated using a Li | |Cu cell with a potential window of −0.2 to 0.6 V. The current of Li plating/stripping using LiFSI-FEMS electrolyte is significantly higher than that using LiPF₆-EC/DEC electrolyte, indicating rapid and reversible plating/stripping kinetics (Supplementary Fig. 19). The exchange currents derived from Tafel curves of Li | |Li symmetric cells were then calculated to reflect the charge-transfer kinetics (Supplementary Fig. 20). The exchange current for the cell using LiFSI-FEMS electrolyte at 30 °C (0.71 mA cm⁻²) is higher than that using LiPF₆-EC/DEC electrolyte (0.15 mA cm⁻²). Even at −20 °C, LiFSI-FEMS electrolyte still shows a high exchange current during the Li deposition, confirming its rapid kinetics at low temperatures.

Elevating the temperature (i.e., ≥ 55 °C) would deteriorate the Li-metal/electrolyte interfacial parasitic reactions, while lowering the temperature (i.e., ≤ −20 °C) would dramatically compromise the Li deposition kinetics, therefore, achieving reversible Li plating/stripping across wide temperatures remains a significant challenge. To justify the advantage of this asymmetric electrolyte, Li | |Cu cell was then subjected to the test at wide temperatures. As shown in Fig. 2c and Supplementary Figs. 21–22a, using LiFSI-FEMS electrolyte, Li | |Cu cells achieve average CEs of 99.6%, 99.2%, 98.7%, and 95.5% at 55 °C, −20 °C, −40 °C, and −60 °C, respectively, attributed to the facile desolvation and conductive SEI. The good compatibility towards Li-metal was further evaluated in Li | |Li symmetric cells (Fig. 2d). Using LiPF₆-EC/DEC electrolyte, Li | |Li symmetric cell exhibits significant fluctuations in polarization voltage and experiences a quick short circuit within 500 h, due to the unstable SEI that results in continuously parasitic reaction and Li dendrite growth. In contrast, the cell using LiFSI-FEMS electrolyte maintains highly stable cycling over 3000 h with a low polarization voltage of ~12 mV (Fig. 2d and Supplementary Fig. 22b, c). Furthermore, the LiFSI-FEMS electrolyte still promises good cycling stability with low polarization voltage at a higher current density of 1.0 mA cm⁻² (Supplementary Fig. 23a–c). Even at −60 °C, Li | |Li cell using FEMS electrolyte still operate stably over 600 h (Supplementary Fig. 23d).

Li deposition morphology and interfacial chemistry

The Li deposition morphology at various temperatures was then characterized using scanning electron microscopy (SEM) in Fig. 3a, b. For the cell using LiPF₆-EC/DEC, highly loose and whisker-like Li deposition morphology was observed on Cu even at a mild temperature of 30 °C, exacerbating the parasitic reaction with electrolyte that leads to rapid cell failure (Fig. 2a). In contrast, the cell using LiFSI-FEMS shows a flat and compact granular morphology, indicating homogenous Li deposition. Elevating the temperature to 55 °C further deteriorates the parasitic reaction for that using LiPF₆-EC/DEC, whereas the LiFSI-FEMS still maintains highly dense and smooth surface that approaches ideal planar Li deposition. When the temperature was reduced to −20 °C, Li plating/stripping is dominated by charge-transfer kinetic (i.e., desolvation), abundant needle-like dendrites with rough surface was observed for the cell using LiPF₆-EC/DEC due to the difficulty in desolvation. Extraordinarily, the cell using LiFSI-FEMS electrolyte still yields uniform Li deposition morphology with a shiny silver appearance, primarily owing to the facile Li⁺ desolvation in LiFSI-FEMS electrolyte. Even at −40 °C and −60 °C, uniform spherical Li deposits without dendrites was achieved for that using LiFSI-FEMS electrolyte (Supplementary Fig. 24), which is crucial for operating LMBs at low temperatures.

Fig. 3. Li deposition morphology and interfacial chemistry.

SEM images and corresponding optical photographs of Li deposited on Cu foils at a current density of 0.5 mA cm⁻² for 1.5 mAh cm⁻² with a LiPF₆-EC/DEC and b LiFSI-FEMS electrolytes from 55 to −20 °C. c C 1 s in-depth XPS spectra of cycled Li-metal after 100 cycles in LiPF₆-EC/DEC and LiFSI-FEMS electrolytes. d S 2p and N 1 s in-depth XPS spectra of the cycled Li-metal after 100 cycles in LiFSI-FEMS electrolyte. e Atomic ratio of the SEI layer after 100 cycles in LiPF₆-EC/DEC and LiFSI-FEMS electrolytes before and after 60 s of sputtering. Schematic illustration of the SEI formed in f LiPF₆-EC/DEC and g LiFSI-FEMS electrolytes.

In addition to the facile desolvation, the components of SEI on Li-metal also plays a crucial role to enable reversible Li plating/stripping, which was carefully investigated by X-ray photoelectron spectroscopy (XPS) with an Ar⁺ sputtering technique. With LiPF₆-EC/DEC electrolyte, the Li-metal surface is enriched with ROCO₂Li, Li₂CO₃, LiF, and Li_xPO_yF_z species (Fig. 3c and Supplementary Fig. 25), attributing to the excessive decomposition of organic solvents (i.e., EC and DEC) and LiPF₆, which is not only unstable but also too resistive for repeated Li plating/stripping³⁶. In contrast, using LiFSI-FEMS electrolyte, the Li-metal surface displays distinct sulfur species (i.e., SO_x, CSO₂/NSO₂, and Li₂S_x) and nitrogen species (i.e., S–N/C–N, LiN_xO_y, and Li₃N) (Fig. 3d, e, and Supplementary Fig. 25), which has been recognized to be stable with low resistance^37,38. As the etching time increases, the amounts of low-valent S-based species (i.e., S_x²⁻ and S²⁻) and LiF steadily increase, which is beneficial for reducing the Li⁺ diffusion energy barrier and facilitate homogeneous Li deposition³⁹. Moreover, as shown in the Li 1 s in-depth XPS spectra, the Li-metal cycled in LiFSI-FEMS shows earlier detection of Li-metal peak, indicating that the SEI formed in LiFSI-FEMS is much thinner than that in LiPF₆-EC/DEC electrolyte (Supplementary Fig. 25c).

Based on above discussion, the detailed chemical components of SEI using both LiPF₆-EC/DEC and LiFSI-FEMS electrolytes are schematically illustrated in Fig. 3f, g. For LiPF₆-EC/DEC electrolyte, organic components (i.e., ROCO₂Li) are distributed throughout the thick SEI, leading to inhomogeneous Li deposition and dendrite formation. In contrast, the thin SEI formed with LiFSI-FEMS electrolyte is enriched with LiF, sulfur-containing and nitrogen-containing inorganic species, which not only regulates Li⁺ flux for homogenous deposition but also suppresses the parasitic reaction with electrolyte, enabling dendrite-free Li deposition with high CE across a wide temperature range.

High anodic stability towards high-voltage positive electrode

The stability of solvent towards positive electrode during the delithiation process is critical for building high-energy LMBs. Therefore, the H-transfer reaction energy of DEC, EC, and FEMS on the surface of NCM811 at a 1/3 charge state was calculated based on DFT. The calculated reaction energies (ΔE_reaction) of DEC, EC, and FEMS are −1.43 eV, −1.22 eV, and 0.98 eV, respectively (Fig. 4a and Supplementary Data 18–20), which indicates that carbonate solvents readily react with oxygen on the NCM811 surface, while FEMS exhibits the highest H-transfer reaction energy, demonstrating its good oxidation stability on high voltage positive electrode. Furthermore, linear sweep voltammetry (LSV) with Pt as the working electrode was performed to assess the intrinsic oxidation stability of electrolyte in a three-electrode cell. As shown in Fig. 4b, the LiPF₆-EC/DEC electrolyte shows an onset oxidation of about 4.3 V, whereas the LiFSI-FEMS electrolyte is stable up to ~5.0 V without any apparent oxidation current. This validates our design rationale that sulfone backbone with electron-withdrawing –CF₃ moiety can significantly enhance the oxidation stability of the electrolyte. Meanwhile, the anodic corrosion of Al current collector is a significant drawback associated with imide salt (e.g., LiFSI)⁴⁰, which was examined by a chronoamperometry (CA) test of Li | |Al cell. As shown in Fig. 4c, the SEM images reveal no obvious Al corrosion for both electrolytes after holding at 4.8 V for 10 h. While high leakage current was observed for LiPF₆-EC/DEC electrolyte, which was ascribed to the oxidation decomposition of the carbonate solvents. In contrast, LiFSI-FEMS electrolyte exhibits negligible leakage current below 1.0 μA cm⁻², indicating significant inhibition of Al corrosion and electrolyte oxidation, which can be attributed to the weaker solvation ability of FEMS that allows the accumulation of passivation layer on Al surface^26,41.

Fig. 4. Electrochemical performance of high-voltage Li | |NCM811 cell at 30 °C.

a The H-transfer reaction energies of DEC, EC, and FEMS molecules on the delithiated NCM811 (003) surface. Different colored spheres represent atoms: gray (C), yellow (S), light blue (F), tan (O), dark blue (N), light pink (H), blue (Ni), light gray (Mn), and light red (Co). b LSV curves of electrolytes in a three-electrode cell with Pt as the working electrode and Li as the counter and reference electrodes at a scan rate of 1 mV s⁻¹. c CA profiles of Li | |Al cells at 4.8 V for 10 h. Inserts: SEM images of the Al foils after CA test for 10 h. d Rate performance and e cycling performance of Li | |NCM811 cells at a high cut-off voltage of 4.8 V and 30 °C using LiPF₆-EC/DEC and LiFSI-FEMS electrolytes. 1 C = 220 mA g⁻¹ for all the NCM811 cells tested in this study.

Having confirmed its good Li-metal compatibility, rapid electrochemical kinetics, and high anodic stability, the LiFSI-FEMS electrolyte was subjected to the test in high-voltage LMBs with Ni-rich NCM811 positive electrode. As shown in Fig. 4d, under a high cut-off voltage of 4.8 V, Li | |NCM811 cell using LiFSI-FEMS demonstrates good rate performance, which still delivers a high capacity of 170.3 mAh g⁻¹ at a high rate of 5.0 C, owing to its accelerated desolvation kinetics and lower interfacial resistance compared to that using LiPF₆-EC/DEC electrolyte. From the detailed galvanostatic charge−discharge (GCD) profiles at high cut-off voltage of 4.8 V (Supplementary Fig. 26a, b), Li | |NCM811 cell with LiPF₆-EC/DEC electrolyte shows an initial capacity of 234.4 mAh g⁻¹ with an initial CE of 84.7%, while the cell using LiFSI-FEMS electrolyte achieves an initial capacity of 253.7 mAh g⁻¹ and a high initial CE of 91.6%. These results suggest that during the initial cycle, LiFSI-FEMS electrolyte is able to form favorable electrolyte–electrode interphases, which effectively prevent the parasitic reaction and irreversible capacity loss. Remarkably, the cell with LiFSI-FEMS electrolyte demonstrates a high capacity retention of 95% with low voltage polarization after 100 cycles (Fig. 4e and Supplementary Fig. 26c, d). It is particularly noteworthy that such high cut-off voltage (i.e., 4.8 V) is competitive to most reported “single-solvent single-salt” electrolytes (Supplementary Table 1). If cycling under a slightly lower cut-off voltage of 4.7 V, the LiFSI-FEMS electrolyte is able to sustain the stable cycling of Li | |NCM811 cell for more than 200 cycles without obvious capacity decay (Supplementary Fig. 27).

Galvanostatic intermittent titration technique (GITT) was employed to quantitatively evaluate the diffusion kinetics of Li⁺ on positive electrode before and after cycle⁴². Before cycle, both LiFSI-FEMS and LiPF₆-EC/DEC electrolytes exhibit similar polarization overpotential and Li⁺ diffusion coefficients at various state-of-charges (SoCs) in Supplementary Fig. 28a–d. After 100 cycles, the overpotentials at various SoCs using LiFSI-FEMS electrolyte are much smaller than that using LiPF₆-EC/DEC electrolyte (Supplementary Fig. 28e, f), implying that the CEI derived from LiFSI-FEMS electrolyte possesses a much lower resistance. More importantly, Li⁺ diffusion coefficients using LiFSI-FEMS electrolyte remain almost same as before cycling, of which the values were about one order of magnitude higher than those using LiPF₆-EC/DEC electrolyte. These results strongly confirm that the LiFSI-FEMS electrolyte effectively regulates the formation of thin and robust CEI to facilitate the Li⁺ diffusion with low resistance.

Interfacial chemistry and structural evolution of NCM811 at high-voltage

To gain an in-depth analysis of the LiFSI-FEMS electrolyte in protecting high-voltage positive electrodes, XPS was carried out to investigate the NCM811 positive electrodes after 100 cycles at 4.8 V. As shown in Fig. 5a, b, and Supplementary Fig. 29, for the positive electrode cycled in LiPF₆-EC/DEC electrolyte, the surface contains a large amount of unstable ROCO₂Li and Li_xPO_yF_z species, indicating excessive depletion and decomposition of carbonate solvents and LiPF₆ at high-voltage. Those species remain quite high even after 60 s of Ar⁺ sputtering time. In contrast, using LiFSI-FEMS electrolyte, the surface shows less organic species but more inorganic species (i.e., LiF, nitride and sulfur compounds). Notably, as sputtering time increases, the content of LiF increases significantly with relative atomic ratio of F rises from 37 to 66% (Fig. 5b, c), endowing the CEI with high thermal stability and low resistance⁴³.

Fig. 5. Interfacial and structural characterizations of the cycled NCM811 positive electrodes after 100 cycles.

a C 1 s and b F 1 s in-depth XPS spectra of the NCM811 positive electrodes cycled in LiPF₆-EC/DEC and LiFSI-FEMS electrolytes. c Atomic ratio of the CEI layer in LiPF₆-EC/DEC and LiFSI-FEMS electrolytes before and after 60 s of sputtering. d XRD patterns of the pristine and cycled NCM811 positive electrodes in LiPF₆-EC/DEC and LiFSI-FEMS electrolytes. e TMs concentrations after cycling in different electrolytes. SEM and HR-TEM images of NCM811 particles after 100 cycles at 4.8 V in f LiPF₆-EC/DEC and g LiFSI-FEMS electrolytes. Schematic illustration of the CEI formed on the surface of NCM811 particles in h LiPF₆-EC/DEC and i LiFSI-FEMS electrolytes.

Moreover, the structural reversibility of NCM811 positive electrodes during repeated cycling was investigated by X-ray diffraction (XRD). As shown in Fig. 5d, the (003) peak for the NCM811 positive electrode cycled in LiPF₆-EC/DEC electrolyte shifts to lower angle and loses intensity, indicating a clear structural degradation during cycling. In contrast, the positive electrode cycled using LiFSI-FEMS electrolyte remains nearly unchanged, implying that the structural integrity was preserved during cycling. Moreover, the curves of differential capacity vs. voltage (dQ/dV) derived from CV test were plotted to assess the stability of the internal lattice structure during cycling (Supplementary Fig. 30). After 100 cycles, the phase transition process of H₂ → H₃ with LiPF₆-EC/DEC electrolyte disappears during both charging and discharging process, indicating an obviously internal structural change. In contrast, the dQ/dV curves for the cell with LiFSI-FEMS electrolyte are almost overlapped upon cycling, indicating that the LiFSI-FEMS electrolyte could effectively prevent the NCM811 positive electrode from distortion at high-voltage.

The transition metals (TMs) dissolution after cycling was also examined using inductively coupled plasma optical emission spectrometry (ICP-OES). The dissolved TMs (i.e., Ni, Co, and Mn) contents for the cell using LiPF₆-EC/DEC electrolyte were obviously higher than those using LiFSI-FEMS electrolyte (Fig. 5e), confirming that CEI derived from LiFSI-FEMS electrolyte effectively inhibits the dissolution of TMs. Furthermore, the cycled Li-metal in LiPF₆-EC/DEC electrolyte exhibits an obvious Ni signal from energy-dispersive X-ray spectroscopy (EDS) in Supplementary Fig. 31, indicating the crossover of TM ions that get reduced on Li-metal, which can catalyze the parasitic reactions and destruct the SEIs. In contrast, with LiFSI-FEMS electrolyte, much less TM ions crossover was found on the surface of cycled Li-metal, which provides additional evidence for suppressing TMs dissolution.

Meanwhile, SEM and high-resolution transmission electron microscopy (HR-TEM) were conducted to assess the structural evolution of NCM811 positive electrodes after cycling. The NCM811 particle cycled using LiPF₆-EC/DEC electrolyte is severely collapsed (Fig. 5f), suggesting severe parasitic reaction between positive electrode and electrolyte. Besides, the surface of cycled NCM811 is coated with a thick CEI of about 9 nm and the lattice spacing of the (101) planes increases from 0.246 to 0.256 nm, indicating serious irreversible structural damage. In sharp contrast, using LiFSI-FEMS electrolyte, the NCM811 positive electrode remains an intact and spherical microstructure (Fig. 5g), intuitively showing high structural integrity of NCM811 positive electrode. Furthermore, the cycled NCM811 particle exhibits a dense and homogeneous CEI of about 2 nm without obvious change in the lattice spacing of the (101) planes, confirming that the LiFSI-FEMS electrolyte can effectively suppress side reactions and preserve positive electrode integrity at high-voltage.

Based on the above discussions, in order to realize high voltage operation of Ni-rich layered oxide positive electrodes, the electrolyte should not only possess intrinsically high oxidation stability, but also being able to generate highly robust CEI on positive electrodes. As schematically illustrated in Fig. 5h, the LiPF₆-EC/DEC electrolyte itself suffers from insufficient oxidation stability and the derived CEI primarily comprises organic component, resulting in over decomposition of the electrolyte and continuous parasitic reactions with charged positive electrodes, which leads to serious structural disruption and TMs dissolution of the positive electrodes. In contrast, in addition to the high oxidation stability, the CEI generated by LiFSI-FEMS electrolyte is mainly consisting of LiF and other inorganic components (e.g., Li₃N and S-containing species) that is thin and robust, which ensures rapid Li⁺ migration and preserves the integrity of the NCM811 positive electrode under high cut-off voltages (Fig. 5i).

Stable operation of high-voltage LMBs from − 60 to 100 °C

To assess the feasibility of the electrolytes under wide temperature conditions, Li | |NCM811 cells were tested over a temperature range from −60 to 100 °C. As shown in Fig. 6a and Supplementary Fig. 32, Li | |NCM811 cell with LiPF₆-EC/DEC electrolyte fails to operate at below −20 °C due to the solidification of the electrolyte (Fig. 2b). In stark contrast, using LiFSI-FEMS electrolyte, Li | |NCM811 cell still delivers high capacities of 176.2, 144.5, and 103.9 mAh g⁻¹ at −20, −40, and −60 °C, respectively. Upon returning to 25 °C from −60 °C, Li | |NCM811 cell delivers similar capacity (238.0 mAh g⁻¹) as in the initial three cycles (239.0 mAh g⁻¹). Furthermore, when the temperature was raised to 70 °C and further to 100 °C, Li | |NCM811 cell still exhibits steady capacities of 241.3 and 236.4 mAh g⁻¹, respectively, demonstrating the feasibility of LiFSI-FEMS electrolyte to enable high voltage LMBs across a wide temperature range.

Fig. 6. Electrochemical performance of high-voltage Li | |NCM811 coin cells at various temperatures.

a Cycling performance of Li | |NCM811 cells at 0.1 C under different temperatures (−60 °C, −40 °C, −20 °C, 30 °C, 70 °C, 100 °C), except for −60 °C at 0.05 C. The cut-off voltage of Li | |NCM811 cell was set at 4.8 V from –60 to 30 °C, while at 4.7 V for 70 and 100 °C. Cycling performance of Li | |NCM811 cells at b 55 °C (2 C), c 70 °C (2 C), d −20 °C (0.2 C), e −40 °C (0.1 C), and f −60 °C (0.05 C). The charging/discharging was performed at the same C-rate at each temperature.

High temperature would seriously aggravate the positive electrode/electrolyte parasitic reactions, especially under high cut-off voltages. The advantage of the LiFSI-FEMS electrolyte was first demonstrated by testing the Li | |NCM811 cell under a cut-off voltage of 4.7 V at 55 °C. As shown in Fig. 6b, the Li | |NCM811 cell with LiFSI-FEMS electrolyte delivers good cycling performance with a high capacity retention of 88% and high average CE of 99.4% after 100 cycles. Besides, the charge-discharge curves reveal that the voltage polarization remains stable and low upon cycling, indicating good interfacial stability (Supplementary Fig. 33). In contrast, the cell using LiPF₆-EC/DEC electrolyte experiences rapid capacity decay with large voltage polarization due to continuous electrolyte decomposition and severe side reaction accelerated by high temperature and high cut-off voltage. Furthermore, the NCM811 positive electrode particles maintain high integrity after cycling in LiFSI-FEMS at 55 °C, whereas severe pulverization was found for that in LiPF₆-EC/DEC (Supplementary Fig. 34). We then further raise the temperature to 70 °C, as shown in Fig. 6c, the cell is still able to cycle stably with a high retention of 80% after 100 cycles, which is attributed to the highly robust and thermally stable inorganic-rich CEI/SEI generated by LiFSI-FEMS.

Low temperature would greatly compromise the electrochemical kinetics. Due to the solidification of the electrolyte, poor oxidation stability, and uneven Li plating/stripping, the Li | |NCM811 cell using LiPF₆-EC/DEC electrolyte experiences quick capacity decay with less than 40% capacity retained after 50 cycles at −20 °C (Fig. 6d and Supplementary Fig. 35). For that using LiFSI-FEMS electrolyte, the cell demonstrates good cycling stability with a capacity retention of 83% and low polarization voltage after 200 cycles. Additionally, the Li | |NCM811 using LiFSI-FEMS electrolyte also shows good rate performance at −20 °C (Supplementary Fig. 36). To further justify the advantage of the LiFSI-FEMS electrolyte, the cell using LiFSI-FEMS was then subjected to the test at −40 °C with a cut-off voltage of 4.8 V. As shown in Fig. 6e and Supplementary Fig. 37, the Li | |NCM811 achieves a high initial capacity of 143.4 mAh g⁻¹ at 0.1 C, with negligible decay and highly stable polarization voltage after 300 cycles. Additionally, the cell also demonstrates good rate performance with a capacity of 88.2 mAh g⁻¹ released at 0.5 C (Supplementary Fig. 38). Remarkably, even when the operating temperature was further reduced to −60 °C, the Li | |NCM811 cell still displays a high initial capacity of 104.9 mAh g⁻¹ without obvious capacity decay after 100 cycles (Fig. 6f). To mimic the realistic assembly of LMBs for practical application, 50 μm thin Li foil and high mass loading NCM811 positive electrode (24 mg cm⁻²) were adopted to assemble the Li | |NCM811 coin cells, leading to a practically low negative to positive electrode capacity (N/P) ratio of 2.15. As presented in Supplementary Figs. 39–40, the Li | |NCM811 cells using LiFSI-FEMS electrolyte can also operate stably at 30 °C and −40 °C.

Currently, it remains a critical challenge for cycling the high-voltage (≥4.5 V) Li | |NCM cell at low temperatures of ≤ −40 °C and high temperatures of ≥60 °C simultaneously (Supplementary Table 2). As schematically summarized in Fig. 7a, the stable operation of Li | |NCM811 cell using LiFSI-FEMS under high voltage and wide temperature range can be attributed to the following merits: (1) the introduction of –CF₃ moiety to the high voltage tolerant sulfone backbone endows the electrolyte with high intrinsic oxidation stability; (2) the asymmetric structure of the solvent significantly increases the disordered intermolecular arrangement, which bestows the FEMS solvent low freezing point; (3) the fluorination and steric hindrance dual effects significantly weaken the solvation ability of the FEMS solvent, promoting the Li⁺–anion pairing to constitute anion-rich solvation sheath, which not only bestows low desolvation energy but also leads to the formation of LiF-rich inorganic CEI/SEI that is highly robust and thin.

Fig. 7. Electrolyte merits and evaluation of 4.7 V Li | |NCM811 pouch cell.

a Schematic summary of the electrolyte merits towards high voltage and wide temperature LMBs. Different colored spheres represent atoms: gray (C), blue (S), red (O), green (F), pink (N), white (H), and light blue (Li). b Cycling performance and c GCD curves of the Li | |NCM811 pouch cell with the LiFSI-FEMS electrolyte at a cut-off voltage of 4.7 V and 30 °C.

4.7 V Li-metal pouch cell

The practicability of the LiFSI-FEMS electrolyte was further evaluated in a pouch-type cell. To achieve a high specific energy exceeding 500 Wh kg⁻¹, all cell parameters were strictly controlled in harsh conditions, including high mass loading positive electrode (25 mg cm⁻²), low negative/positive ratio (N/P = 1.9), and lean electrolyte (1.83 g Ah⁻¹). The specific energy of the Li | |NCM811 pouch cell was calculated based on the total mass of all components, of which the detailed parameters are shown in Supplementary Table 3 and Supplementary Fig. 41. As shown in Fig. 7b, the pouch cell delivers a discharge capacity of 5.46 Ah at 4.7 V, corresponding to an high specific energy of 514.0 Wh kg⁻¹. Furthermore, the pouch cell can stably charging/discharging for more than 20 cycles at such a high cut-off voltage (i.e., 4.7 V), retaining over 95.6% capacity with low voltage polarization (Fig. 7c), which represents a competitive cut-off voltage ever reported for Li-metal pouch cells based on Ni-rich layered oxide positive electrodes. In addition, adopting a slightly more electrolyte (E/C = 2.20 g Ah⁻¹), the operation of Li-metal pouch cell could be extended to 40 cycles (Supplementary Fig. 42).

Discussion

In conclusion, a single-salt single-solvent electrolyte with an asymmetric fluorinated sulfonamide solvent, FEMS, has been designed to enable high voltage LMBs up to 4.8 V and across a wide temperature range from −60 to 100 °C. The systematic theoretical and experimental studies have demonstrated that (1) the introduction of –CF₃ moiety to the high voltage tolerant sulfone backbone endows the electrolyte with high intrinsic oxidation stability (~5.0 V); (2) the asymmetric methyl/ethylamine groups of FEMS could increase the disorder intermolecular arrangement, bestowing the electrolyte with low freezing point (below −150 °C), and (3) the FEMS with weak solvation ability endows the electrolyte with anion-rich solvation structure, which not only possesses low desolvation energy barrier but also generates LiF-rich, S/N-containing inorganic CEI/SEI, significantly promoting the electrochemical kinetics and Li reversibility (up to a high CE of 99.8%) from −60 to 55 °C. Consequently, this single-salt single-solvent LiFSI-FEMS electrolyte promises the Li | |NCM811 coin cells to operate steady at a high cut-off voltage of 4.8 V with a capacity retention of 95% after 100 cycles at 30 °C, and no obvious capacity decay after 300 cycles at −40 °C and 100 cycles at −60 °C, respectively. Remarkably, the practicality of this single-salt single-solvent FEMS electrolyte has also been verified by a 5.46 Ah Li | |NCM811 pouch cell at a cut-off voltage of 4.7 V, which achieves a high specific energy of 514 Wh kg⁻¹ over 20 cycles. This study provides valuable insights of solvent asymmetry on widening operation temperature range, representing a significant advancement in the electrolyte design to maximize the specific energy and operation temperature range of practical LMBs.

Methods

Materials

Triflyl chloride (98%), N-ethly-N-methylamine (98%), N-propyl-N-methylamine, anhydrous triethylamine (TEA, 98%), and dichloromethane (DCM, 99.5%) were purchased from Energy Chemical. Lithium bis(fluorosulfonyl)imide (LiFSI, >99.9%), 1.0 M LiPF₆ in EC/DEC electrolyte (1:1 by volume), and LiNi_0.8Mn_0.1Co_0.1O₂ (NCM811) powders were friendly provided by Guangzhou Tinci Materials Technology Co. Ltd. PP separator (Celgard 2500, 25 μm thick, 55% porosity, and 0.064 μm average pore size) was purchased from Celgard, LLC. Li foil (99.95%, 500 or 50 μm thick), Cu foil, Al foil, and CR 2032-type coin-cell cases were purchased from Guangdong Canrd New Energy Technology Co. Ltd.

Synthesis of N-ethyl-N-methyltrifluoromethane-sulfonamide (FEMS)

Firstly, 50 ml of anhydrous DCM, 7.8 ml of anhydrous TEA, and 3.37 g of N-ethly-N-methylamine were sequentially added to a 250 ml two-neck round-bottom flask under anhydrous and oxygen-free conditions. The mixture was cooled to 0 °C in an ice-water bath. In the pressure-equalizing dropping funnel, 10 ml of DCM and 8 g of triflyl chloride were mixed and dropwise added to the round-bottom flask under stirring. After stirring for 0.5 h at 0 °C, the mixture was allowed to warm to 25 °C and react for 8 h under stirring. Then, the reaction mixture was washed twice with 100 ml of 1 mol L⁻¹ hydrochloric acid to gather the organic phase, which was further washed twice by 80 ml of saturated brine. The organic phase was dried over anhydrous sodium sulfate, filtered, and condensed using a rotary evaporator. The pure FEMS product was obtained after vacuum distillation (70 °C, 40.0 mbar). Then, the FEMS solvent was dried over molecular sieve for 72 h before use.

¹H-NMR (600 MHz, CDCl₃, δ/ppm): 3.38–3.54 (ws, 2H), δ 3.05–3.06 (s, 3H); δ 1.27–1.31 (m, 3H). ¹³C-NMR (150 MHz, CDCl₃, δ/ppm): 119.1–121.3 (CF₃); δ 45.9–46.0 (N − CH₃); δ 34.4–34.5 (N − CH₂ − ); δ 13.3–13.4 ( − CH₃). ¹⁹F-NMR (450 MHz, CDCl₃, δ/ppm): −75.6~ −75.5 (CF₃).

Synthesis of N-propyl-N-methyltrifluoromethane-sulfonamide (FDPS)

The synthesis procedure for FDPS was the same as the FEMS, except for replacing N-ethyl-N-methylamine with N-propyl-N-methylamine. The FDPS was collected at 80 °C during the vacuum distillation.

¹H-NMR (600 MHz, CDCl₃, δ/ppm): 3.40–3.25 (ws, 2H), δ 3.03–3.06 (d, 3H); δ 1.73–1.65 (m, 2H); δ 0.96–0.99 (d, 3H). ¹³C-NMR (150 MHz, CDCl₃, δ/ppm): 117.0–123.5 (CF₃); δ 52.3–52.5 (N − CH₃); δ 34.9–35.0 (N − CH₂ − ); δ 20.8–21.0 (C − CH₂ − C);δ 10.6–10.7 ( − CH₃). ¹⁹F-NMR (450 MHz, CDCl₃, δ/ppm): −75.1~ −75.0 (CF₃).

Preparation of electrolytes

The single LiFSI salt single sulfonamide solvent electrolyte mentioned in this study was prepared by dissolving 1 mmol of LiFSI in 1 ml solvent of FDMS and FEMS and stirred for 12 h until fully dissolved. For LiFSI-FDPS electrolyte, due to the limited solubility of the FDPS, 0.6 mmol of LiFSI was adopted. All the electrolytes were prepared in an Ar-filled glovebox at 25 °C with water and oxygen levels below 0.01 ppm.

Preparation of electrodes

The NCM811 positive electrode was prepared by mixing NCM811 power, acetylene black (Li400, Denka Black), and poly(vinylidene fluoride) (PVDF, HSV900, Kynar) in N-methyl-2-pyrrolidone (NMP, 99.9%) with a weight ratio of 8: 1: 1. The slurry was coated onto 12 μm Al foil using a doctor blade and dried over 6 h at 60 °C. Subsequently, it was further vacuum dried at 120 °C for 24 h. The low and high mass loadings of active material are about 3.2 and 24 mg cm⁻², respectively.

Material characterizations

Differential scanning calorimetry (DSC) measurements of solvents and electrolytes were carried out from 30 to −150 °C in a sealed aluminum pan at a rate of 5 °C min-using a DSC 214 polyma (NETZSCH). The density and viscosity of electrolyte was measured by a viscometer (SVM3001, Anton Paar) with the temperature range from−60 to 55 °C. The contact angle was on a PP separator (Celgard 2500) using a contact angle measurement system (JC2000D3P, Shanghai Zhongchen). The chemical compositions on the electrodes surface were analyzed using X-ray photoelectron spectroscopy (XPS, PerkinElmer PHI 1600 ESCA). The cycled electrodes were rinsed by solvents three times and transferred from the glovebox to the XPS chamber without exposure to air. The Ar⁺ sputtering for the XPS depth-profiling was conducted by a beam energy of 500 eV over an area of 2.5 mm × 2.5 mm. Scanning electron microscopy (SEM, MAIA3, TESCAN Brno, s.r.o.), transmission electron microscopy (TEM, JEM-2100 HR), and high-resolution transmission electron microscopy (HR-TEM, FEI Talos F200X) were conducted to observe the morphology of the electrodes. Energy dispersive X-ray spectroscopy (EDS, Oxford) was used to characterize the transition metals (TMs) of the Li | |NCM811 cells with a beam energy of 5 keV. The X-ray diffraction (XRD) measurement was performed with a diffractometer (D8 Advance, Bruker) with Cu Kα radiation with a scan rate of 5° min⁻¹ and a step size of 0.02. NMR spectra were measured using a 600 MHz spectrometer (Avance NED, Bruker). The chemical shifts in parts per million (ppm) were referenced to a tetramethylsilane standard. The contents of the dissolved TM ions were measured using an inductively coupled plasma optical emission spectrometry (ICP-OES, SPECTRO ARCOS MV). For the test of dissolved TM ions, cycled Li-metal and separator was collected from the disassembled cells in the Ar-filled glovebox and immersed in DME solvent for three days. Then the solution was heat at 80 °C to remove the solvent, and the residues was dissolved in 1 mL HNO₃ and diluted to 10 mL with deionized water.

Electrochemical measurements

All coin cells were assembled using a 2032-type case with 400 or 50 μm Li foil (14 mm), 60 μL electrolyte, Celgard 2500 (19 mm) as the separator, one spring, and one spacer in an Ar filled glovebox (25 °C, O₂ < 0.01 ppm, H₂O < 0.01 ppm). Charge-discharge tests for Li | |NCM811, Li | |Cu, and Li | |Li cells were performed at various current densities using a LAND CT2001A instrument (Wuhan Jinnuo Electronic Co. Ltd.) under different temperatures. For the cycling test of Li | |Cu cell, after three activation cycles (plating 1 h with a current density of 0.05 mA cm⁻² and then stripping to 1.0 V), Li (8 mm) was plated onto the 9 μm Cu foil (14 mm) with a capacity of 1 mAh cm⁻² and followed by stripping to 1.0 V. For the Aurbach method tests at 30, 55, and −40 °C, Li was initially plated on the Cu electrode with a capacity of 5 mAh cm⁻² and fully stripped. Then, Li was re-plated on Cu electrode with a capacity of 5 mAh cm⁻² as a reservoir, and further stripped/plated for 10 cycles at a current density of 0.5 mA cm⁻² with a capacity of 1 mAh cm⁻². Finally, the reserved Li was completely stripped to 1.0 V. When tested at −60 °C, the current density and corresponding capacity decreased to 0.1 mA cm⁻² and 0.2 mAh cm⁻², respectively. CV test of Li | |NCM811 cells was conducted in a two-electrode coin-cell assembly with NCM811 (12 mm) as the positive electrode and 400 μm Li metal (14 mm) as the negative electrode with a scan rate of 0.1 mV s⁻¹ and a voltage range of 2.8–4.8 V. LSV, CA, CV, and EIS measurements were carried out using a Potentiostat (VMP3, Bio-Logic). The LSV tests were performed in a three-electrode cell from OCV to 7.0 V at a scan rate of 1 mV s⁻¹, with Pt as the working electrode and Li-metal as both the reference and counter electrodes. The CA test was conducted by holding Li | |Al cell at 4.8 V for 10 h. The EIS spectrum was recorded over a frequency range from 10 mHz to 1 MHz (except 1 mHz at −20 °C). All cells were tested with at least three replicates, and representative data showing an intermediate performance level were presented in this manuscript.

GITT was applied to cycled coin cells within a voltage range of 2.8–4.8 V at 0.1 C, with a charging/discharging duration of 20 min and a resting period of 60 min. Li-ion diffusion coefficient (D_Li+) utilizing GITT methodology, which adhered to Fick’s second law and formula:

$${\mathrm{D}}_s=\frac{4}{\mathrm{\pi }\mathrm{\tau }}{\left(\frac{m_{\mathrm{e}}{V}_{\mathrm{e}}}{M_{\mathrm{e}}S}\right)}^2{\left(\frac{\mathrm{\Delta }{E}_{\mathrm{s}}}{\mathrm{\Delta }{E}_{\mathrm{t}}}\right)}^2$$ 1

Where m_e, M_e, and V_e represent the mass, molar mass, and molar volume of the active material, S is the contact area between electrode and electrolyte. ∆E_s denotes the change in pulse voltage after relaxation, ∆E_t represents the potential difference that occurs during the current pulse.

The ionic conductivity was measured by a two-electrode cell, in which two pieces of Pt of the same size were placed in parallel with a certain distance. The cell constant k was calibrated using KCl standard solution (147.4 μS cm⁻¹, 25 °C). The ionic conductivity value was obtained via EIS using the following equation:

$$\mathrm{\delta }=\frac{\mathrm{k}}{\mathrm{R}}$$ 2

where R is the resistance, and k is the cell constant, respectively.

The Li⁺ transference number (t_Li⁺) was measured by applying a small DC polarization potential (i.e., 10 mV) to Li | |Li cell for 20 min to reach a steady state. The t_Li⁺ was calculated according to the following equation:

$${\mathrm{t}}_{{\mathrm{Li}}^{+}}=\frac{{\mathrm{I}}_{\mathrm{S}}\left(\mathrm{\Delta }\mathrm{V}-{\mathrm{I}}_0{\mathrm{R}}_0\right)}{{\mathrm{I}}_0\left(\mathrm{\Delta }\mathrm{V}-{\mathrm{I}}_{\mathrm{S}}{\mathrm{R}}_{\mathrm{S}}\right)}$$ 3

where ∆V is the applied potential, R₀ and R_S are the initial state and steady state resistance, respectively. I₀ and I_S are the initial state and steady state current, respectively.

The Tafel curve was obtained from CV at a scan rate of 0.5 mV s⁻¹ using a Li | |Li symmetrical cell. The exchange current density was calculated using the Tafel equation

$$\mathrm{\eta }=\mathrm{a}+\mathrm{blog}\left(\mathrm{I}\right)$$ 4

when η equals to 0, a and b are the constants determined by data fitting.

Preparation and test of 5 Ah-level pouch cells

5 Ah-level Li | |NCM811 pouch cells were assembled at dry room with a dew point of −60 °C. The detailed parameters of 5 Ah-level Li | |NCM811 pouch cells were provided in Supplementary Table 3. First, both positive and negative electrodes were stacked layer-by-layer using a PP separator. Second, nickel and aluminum tabs were welded to the copper and aluminum currents, respectively. Third, it was sealed by aluminum plastic film without short-circuit. Finally, the pouch cell was injected with a certain amount of electrolyte and sealed for resetting 12 h. The cycling performance of the pouch cells was tested in a constant-temperature test chamber of 30 ± 0.5 °C with a cut-off voltage of 4.7 V after two initial activation cycles at 0.1 C with a cut-off voltage of 4.5 V. The cycling performance of the pouch cells was tested under a fixing device to provide 200 kPa external pressure. Because each pouch cell has different weight, the representing one was selected to be presented in this manuscript.

Theoretical calculations

MD simulations were performed using the LAMMPS package. Molecules and ions were described by using the optimized potentials for a liquid simulations all-atom (OPLS-AA) force field. The simulation box was composed of 39 LiFSI molecules and 264 FEMS or 299 FDMS, or 39 LiPF₆ in 289 EC and 159 DEC molecules with lattice constants of 41 Å × 41 Å × 41 Å, which is corresponding to 1 M LiFSI in FEMS, 1 M LiFSI in FDMS, and 1 M LiPF₆ in EC/DEC electrolytes, respectively. The electrolyte systems were equilibrated for 2000 ps (time step of 1 fs) followed by a 1000 ps (time step of 1 fs) simulation run in isothermal-isobaric ensemble (NPT) and a 1000 ps (time step of 1 fs) simulation run in isothermal-isovolumetric ensemble (NVT). The temperature was controlled to 303 K and 233 K by V-rescale thermostat and pressure was controlled to 1 bar by Berendsen barostat. VMD was used to sample the snapshots and analyze the RDF of different electrolytes from the simulation trajectory.

The binding energy of the Li⁺(solvent)_n complexes was optimized and calculated by DFT, which was performed using the Gaussian 16 package at the B3LYP/6-311 + + (d, p) level with DFT-D3 dispersion correction. The binding energy (E_B) between Li⁺ and (solvent)_n was evaluated as following:

$$E_B=E_{{Li}^{+}-\left(solvent\right)n}-\left(E_{{Li}^{+}}+n*E_{solvent}\right)$$ 5

Where E_Li⁺_{−(Solvent)n} are the total energy of the Li⁺(solvent)_n complexes, E_Li⁺ and E_solvent are the energy of the individual Li⁺ ion and solvent molecule, respectively. The initial structures for the geometry optimization were taken from the snapshot of the MD simulation.

Author contributions

Q.Z. conceived the idea and directed the project. L.C. and Jingwen L. proposed the concept and designed the experiments. Jingwen L., Z.L., Z.P., and W.Y. carried out material synthesis, electrochemical measurements, and most characterizations. X.G. and D.L. helped with the pouch cell assembly and evaluation. Jiawei L. conducted the NMR measurement and analysis. P.X. and L.M. carried out the computational simulations. L.C. and Jingwen L. wrote the draft paper, Q. Z. and Y. C. edited and polished the paper. All authors contributed to the interpretation of the results.

Peer review

Peer review information

Nature Communications thanks Ya-Jun Cheng, Weibing Xing and the other anonymous reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Funding

This work was supported by the National Natural Science Foundation of China (Grant Nos. 92372123 to Q.Z., 22208118 to L.C.), Natural Science Foundation of Guangdong Province (Grant Nos. 2024A1515012236 to L.C., 2023B1515130004 to Q.Z.), and Science and Technology Program of Guangzhou (Grant No. 2024A04J4109 to L.C.).

Data availability

All data supporting the findings of this study are included in the Article and its Supplementary Information. Source data are provided with this paper. Source Data file and DFT calculations data have been deposited in Figshare under accession code DOI link https://doi.org/10.6084/m9.figshare.32141917. Any other data that support the findings of this study are also available from the corresponding authors on request.

Competing interests

A patent related to the work has been submitted (application number CN117410568B) by South China Normal University. The inventors are Qifeng Zheng, Luyi Chen, Jingwen Liu, and Zehang Peng. The patent refers to the synthesis of solvent and the preparation of electrolyte in this paper but provides more analogous solvent than this work. The other authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at https://doi.org/10.1038/s41467-026-76496-2.