One Stone, Three Birds: An Air and Interface Stable Argyrodite Solid Electrolyte with Multifunctional Nanoshells

Abstract Li6PS5Cl (LPSC) solid electrolytes, based on Argyrodite, have shown potential for developing high energy density and safe all‐solid‐state lithium metal batteries. However, challenges such as interfacial reactions, uneven Li deposition, and air instability remain unresolved. To address these issues, a simple and effective approach is proposed to design and prepare a solid electrolyte with unique structural features: Li6PS4Cl0.75‐OF0.25 (LPSC‐OF0.25) with protective LiF@Li2O nanoshells and F and O‐rich internal units. The LPSC‐OF0.25 electrolyte exhibits high ionic conductivity and the capability of “killing three birds with one stone” by improving the moist air tolerance, as well as the interface compatibility between the anode or cathode and the solid electrolyte. The improved performance is attributed to the peculiar morphology and the self‐generating and self‐healing interface coupling capability. When coupled with bare LiCoO2, the LPSC‐OF0.25 electrolyte enables stable operation under high cutoff voltage (≈4.65 V vs Li/Li+), thick cathodes (25 mg cm−2), and large current density (800 cycles at 2 mA cm−2). This rationally designed solid electrolyte offers promising prospects for solid‐state batteries with high energy and power density for future long‐range electric vehicles and aircrafts.


Electrolyte with Multifunctional Nanoshells
Junwu Sang, Kecheng Pan, Bin Tang, *             Table S1.Air stability of LPSC-OxFy and previously reported solid electrolytes.Table S2.Stability of LPSC-OF0.25 to lithium metal compared with the most advanced previously reported polymer, inorganic, and composite solid electrolytes.Table S3.Cycling stability of LPSC-OF0.25|LCOcells compared with the most advanced previously reported full cells with polymer, inorganic, and composite solid electrolytes matched with LFP, LCO, or NCM cathodes.inability to form a stable and uniform SEI with lithium metal to prevent S continuous reaction and diffusion.On the contrast, the interface of Li|LPSC-OF0.25 is much more complete and evener, and no significant bumpy and diffusion was found (Figure 3c).In addition, a significant amount of non-uniform lithium deposition was observed at the interface, and distinct columnar structures resembling dendrite precursors were identified (Figure 3e).

Figure
Figure S6.STEM-HADDF and EDS images of LPSC-O.

Figure S10 .
Figure S10.Contrast XRD patterns of pristine and post-annealed samples after exposure to air.

Figure
Figure S17.Cross-sectional SEM and EDS images for the Li|LPSC interface after cycles.

Figure
Figure S18.(a) Drawing of the in-situ Raman mold.(b) Image of laser focusing area on the Li|LPSC-OF0.25 interface.

Figure S19 .
Figure S19.Schematic diagram of a pressurized cell.

Figure
Figure S20.(a) Discharge capacity and coulombic efficiency, and (b-e) corresponding charge/discharge profiles at 0.5 mAh cm -2 within the voltage of 2.5-4 V (vs.Li-In).

Figure S25 .
Figure S25.EDS images of the as-prepared composite cathode powder after 50 cycles (with LPSC-OF0.25 as the solid electrolyte and ion conductive additive in composite cathodes).

Figure
Figure S26.STEM-HAADF and EDS images of the divested LCO after 50 cycles (with LPSC-OF0.25 as the solid electrolyte and ion conductive additive in composite cathodes).

Figure S27 .
Figure S27.Cry-TEM image of the divested LCO from composite cathodes after 50 cycles (with LPSC-OF0.25 as the solid electrolyte and ion conductive additive in composite cathodes).

Figure S1 .
Figure S1.XRD patterns of annealed samples.(a) XRD patterns of LPSC-OxFy.All obtained XRD data were normalized with the strongest peak at 29.7°, and the ratio of other peaks to the strongest peak represents the ratio of relative content to a certain extent, so the change of the relative content can be obtained by the change of peak intensity.Meanwhile the substrate was removed from the patterns caused by the halos at low angles (10−30°), which are mainly caused by the polyimide film used to prevent the air.The polyimide film leads to the interference of the strongest peak (~21°) of Li3PO4, so we chose the second-strongest peak (~34.5°) for comparison.(b) Partial enlarged drawing.(c) XRD patterns of LPSC-OF0.25 and standard XRD patterns of related impurities, Li2S (PDF No. 26-1188), Li2O (PDF No. 12-0254), Li3PO4, (PDF No. 48-0956) and LiF (PDF No. 45-1460).The LPSC-peak leads to the interference of the strongest peak (~45°) of LiF, so we chose the second-strongest peak (~38.7°) for analyses.

Figure
Figure S6.STEM-HADDF and EDS images of LPSC-O.No obvious shell structure was observed in LPSC-O, and O shows homogeneous distribution, indicating that O is successfully substituted for the S site.

Figure
Figure S7.STEM-HADDF and EDS images of LPSC-OF0.15.No obvious shell structure was observed in LPSC-OF0.15,and O and F show homogeneous distribution, indicating that O is successfully substituted for the S site, and F is replaced the Cl site.

Figure
Figure S8.STEM-HADDF and EDS images of LPSC-OF0.35.Apparently, increasing the proportion of F resulted in a thicker nanoshell and even micron-scale aggregation.

Figure
Figure S9.EIS of selected frequencies for measurements in SS|SE|SS cells at 25℃: (a) high frequency and (b) low frequency.(c) Ionic conductivity of SEs.The steep linear spike at low frequencies indicates that the as-synthesized LPSC-OxFy argyrodites are ionic conductors.The incomplete semicircles indicate a small grain boundary resistance, a typical feature of sulfide-based SEs that is favorable for battery assembly.However, it is hard to distinguish the grain boundary and bulk contribution based on these measured impedance spectra.Because the grainboundary/bulk resistance cannot be clearly detected, it is hard to fit the Nyquist plots.The total ionic conductivity is thus calculated from the local minimal resistance at the intersection between the impedance spectrum and the x-axis.

Figure S12 .
Figure S12.Voltage (solid lines) and current (dotted lines) curves with time for LPSC, LPSC-O and LPSC-OF0.25.The observation from the figure clearly indicates that the symmetrical cell with the LPSC-OF0.25 electrolyte exhibits a lower overpotential than those with LPSC andLPSC-O electrolytes.This seems contradictory, considering that the ionic conductivity of LPSC-OF0.25 is slightly lower than those of LPSC and LPSC-O.However, the overpotential is determined by the overall impedance of the cell, which consists of both the bulk resistance of the electrolyte (directly linked to the ionic conductivity) and the interfacial resistance.In general, the interfacial resistance significantly outweighs the bulk resistance of the electrolyte, making it the primary factor in determining the value of overpotential.The symmetrical cell with the LPSC-OF0.25 electrolyte demonstrates a lower interfacial resistance (FigureS15), and consequently, a reduced overpotential.

Figure
Figure S16.SEM: (b) LPSC, (c) LPSC-O, and (d) LPSC-OF0.25 surface and the corresponding EDS of Cl element after cycles.The EDS mapping of the surface further to verify the above result; the decomposition products (chloride) in the LPSC-OF0.25 does not show any significant clustering, different from large aggregates in LPSC and some small clusters in LPSC-O.

Figure S17 .
Figure S17.Cross-sectional SEM and EDS images for the Li|LPSC interface (a-d) and Li anode (e) after cycles.Through the cross-section diagram in Li|LPSC, the chloride distribution at the interface is irregular, and the sulfide has obvious diffusion.This is most likely caused by LPSC

Figure
Figure S18.(a) Drawing of the in situ Raman mold.(b) Image of laser focusing area on the Li|LPSC-OF0.25 interface.

Figure S19 .
Figure S19.Schematic diagram of a pressurized cell.

Figure
Figure S20.(a) Discharge capacity and coulombic efficiency, and (b-e) corresponding charge/discharge profiles at 0.5 mAh cm -2 within the voltage of 2.5-4 V (vs.Li-In).Charge/discharge profiles for Li-In|SE|LCO cells with (b) LPSC, (c) LPSC-O, and (d) LPSC-OF0.25 as the solid electrolyte and ion conductive additive in composite cathodes.Charge/discharge profiles for Li-In|SE|LCO cells after 100 cycles.

Figure S23 .
Figure S23.Rate performance and Coulombic efficiency for Li-In|LPSC-OF0.25|LCOcells within the voltage of 2.5-4 V (vs.Li-In).In order to accurately obtain more realistic rate performance, we used the current density (mA cm -2 ) divided by the areal capacity (mAh cm -2 )to obtain the C-rate (h -1 ).

Figure S24 .
Figure S24.Charge/discharge profiles for Li-In|LPSC-OF0.25|LCOcells within different voltage ranges (2.5-4 V and 2.5-4.3V vs. Li-In) under 0.5 mA cm -2 .When considering the typical LCO behavior during discharge, the change of discharge curves to steep slope is around 3.8 V vs. Li/Li + while the value in Figure S22 is about 3.45 V vs. Li-In, so we can translate to an actual charging cutoff potential of about 4.65 V vs. Li/Li + .

Figure
Figure S26.STEM-HAADF and EDS images of the divested LCO after 50 cycles (with LPSC-OF0.25 as the solid electrolyte and ion conductive additive in composite cathodes).

Figure S27 .
Figure S27.Cry-TEM image of the divested LCO from composite cathodes after 50 cycles (with LPSC-OF0.25 as the solid electrolyte and ion conductive additive in composite cathodes).

Table S1 .
Air stability of LPSC-OxFy and previously reported solid electrolytes.

Table S2 .
Stability of LPSC-OF0.25 to lithium metal compared with the most advanced previously reported polymer, inorganic, and composite solid electrolytes.

Table S3 .
Cycling stability of LPSC-OF0.25|LCOcells compared with the most advanced previously reported full cells with polymer, inorganic, and composite solid electrolytes matched with LFP, LCO, or NCM cathodes.