Fig. 1 Characterization of the solvation structure and the schematic of the SEI formation. a Schematic of CTAB as a LiNO₃ promoter as well as drawing anions into EDL with optical images of prepared electrolytes w and w/o CTAB. b Schematic of the proposed SEI formation process w and w/o the LiNO₃ addition. c High-resolution Raman spectra of the BE and the CNE-30 electrolyte in the 1010-1105 cm⁻¹ range. d The MD simulations of Li⁺-O and Li⁺-N radial distribution functions g(r) in the BE and the CNE-30 electrolyte. e CV curves of Li||Cu cells scanned at 0.1 mV s⁻¹ in the voltage range of 0-2.5 V vs. Li⁺/Li

Fig. 2 Exploration of EDL evolution by MD simulations and in situ EC-SERS during the first SEI formation. Intensities of FSI⁻/EC ratios in a the BE electrolyte, b the CNE-30 electrolyte and c NO₃⁻/EC ratio, d CTA⁺ in the CNE-30 electrolyte as a function of distance (Å) from graphene electrode under different applied voltages (V). Schematic of EDL model during the first SEI formation in e the BE and f the CNE-30 electrolyte. g Calculated anion/EC ratios from 1.7 to 0.2 V according to Raman peaking fitting. h Schematic of in situ EC-SERS cell configuration: Cu mesh coated with Au nanoparticles as the working electrode, lithium chip as both the counter electrode and the reference electrode

Fig. 3 In-depth analysis of chemical structure of SEI by XPS and ToF-SIMS. XPS spectra of O 1s, N 1s, and F 1s for SEI formed in a the BE and c the CNE-30 electrolyte and the corresponding fitted curves. The detections were conducted on electrochemically deposited Li on Cu foil after five cycles, and the Li surface was sputtered by Ar⁺ for 5, 10, 15, and 20 min, respectively. b Elemental ratios of F/C and N/C in the BE and the CNE-30 electrolyte after the different sputtering times. d Li₂O content (%) in the BE and the CNE-30 electrolyte. The distribution of N and F formed in the CNE-30 electrolyte in e cross section and f 3D reconstructions via ToF-SIMS with a detection area of 100 μm × 100 μm. g Elemental ratios of N/C and F/C in the BE and the CNE-30 electrolyte via ToF-SIMS

Fig. 4 Average CE measurement in asymmetric cells and cycling performance of symmetric cells with the BE and the CNE-30 electrolyte. a Galvanostatic discharge curves in Li||Cu half cells under 0.5 mA cm⁻²/1 mAh cm⁻² during the 1st cycle and the inset indicates the nucleation overpotential (${\eta }_{1}$) and mass-transfer overpotential (${\eta }_{2}$). Average CE measurement after b 10 cycles and c 50 cycles. d Rate performance of the Li||Li cells under 0.5/1/1.5/2/2.5/3 mA cm⁻² and a fixed capacity of 1 mAh cm⁻². e Comparison of long-term cycling of the Li||Li cells under 0.5 mA cm⁻²/1 mAh cm⁻² with insets showing magnified overpotential at 300 h/600 h/1000 h, respectively. Performance comparison between our work and recent advances in f ACEs and g symmetric cells

Fig. 5 The kinetics of Li⁺ transfer at Li/electrolyte interface. a CV curves of the BE and the CNE-30 electrolyte under 0.1 mV s⁻¹ in the voltage range of −0.2 to 2.5 V vs. Li⁺/Li. b Tafel plots obtained by the Li||Li cells of the BE and the CNE-30 electrolyte with inset showing the exchange current densities (i₀). c Charge transfer activation energies E_a calculated using Arrhenius equation. d In situ optical images of Li deposition in the BE and the CNE-30 electrolyte (scale bars, 100 μm). e SEM images of Li deposition morphology

Fig. 6 Full cell performance in the BE and the CNE-30 electrolyte. a Long-term cycling performance of the Li||LiFePO₄ full cell with the BE and the CNE-30 in an electrolyte of 1.68 μL mg⁻¹ with a N/P ratio of 12 and cathode loading of 10.5 mg cm⁻² under 0.5 °C. b Rate performance of the Li||LiFePO₄ full cell under 0.1/0.2/0.5/1/2/3 C with the BE and the CNE-30 electrolyte. Capacity-voltage curves during the 20th/50th/100th cycle with c the BE and d the CNE-30 electrolyte. e Long-term cycling performance of the Li||LiCoO₂ full cells in an electrolyte of 1.72 μL mg⁻¹ with an N/P ratio of 10 and cathode loading of about 10.3 mg cm.⁻² under 0.5 C (charging)/1.0 C (discharging)