Understanding Ion Transport and Interfacial Stability in Fluorine Containing Lithium Argyrodite Electrolytes for Solid-State Lithium-Sulfur Batteries

Rechargeable all solid-state lithium-sulfur batteries (ASLSBs) hold tremendous promise for use in electric vehicles due to their high theoretical capacity (~1675 Ah/kg; ~5-6 times higher than state-of-the-art Li-ion batteries), high promised energy density (400 Wh/kg) and cycle life, and low costs owing to natural abundance of sulfur. Lithium argyrodites (e.g., Li₇PS₆, and its halogen-doped derivatives) have emerged as a lucrative class of solid-state electrolytes (SSEs) for ASLSBs; owing to their high Li-ion conductivity (~10^-3 S/cm), good elastic stiffness (~30 GPa), and low flammability. Despite their promise, ASLSBs (even using argyrodite SSEs) remain far from commercialization due to paucity of fundamental understanding of physical factors underlying key electrochemical phenomenon including ion-conduction, charge transport, structural evolution, and interfacial reactions (e.g., dendrite growth, electrolyte decomposition etc.). Specifically, such knowledge gap has thwarted development of high-performance SSEs that provide rare combination of superionic Li conduction and interfacial stability. Here, we integrate <i>ab initio</i> molecular dynamics (AIMD) simulations, density functional theory (DFT) calculations, liquid-phase synthesis and characterization methods to advance the understanding of lithium-ion conduction and interfacial reactions in fluorine (F) containing lithium argyrodites. The choice of F-containing argyrodites was motivated by recent reports indicating that F (or LiF) could suppress formation of deleterious dendrites at the Li-anode. Our AIMD simulations showed that lithium argyrodite doped with F alone, Li₆PS₅F possesses much lower Li-ion conductivity (~0.32 mS/cm) as compared to state-of-the-art Li₆PS₅Cl electrolytes (~0.9 mS/cm). Interestingly, we found that simultaneous doping of argyrodites with two dissimilar halogens (e.g., F and Cl) results in much higher Li⁺ ion conductivity as compared to the counterparts doped with one halogen. For instance, we found that the Li⁺ ion conductivity of Li₆PS₅F_0.5Cl_0.5 electrolyte (2 mS/cm) is ~2.2 times higher than that of Li₆PS₅Cl (0.9 mS/cm) and ~6 times that of Li₆PS₅F (0.32 mS/cm); consistent with our experimental measurements. Careful analysis of our AIMD trajectories, and DFT calculations indicate that this enhanced Li-ion diffusion can be attributed to the unique defect structure stabilized by F ions, and halogen-mediated hopping of Li ions across the SSE. More importantly, we found that such mixed halogen doping enhances stability of SSE against Li-anode, as evidenced by our galvanostatic Li stripping/plating experiments that show flat voltage profile and low polarization voltage for Li₆PS₅F_0.5Cl_0.5even at current density of 0.2 mA/cm²; in contrast, the cells with Li₆PS₅F and Li₆PS₅Cl electrolytes die at 0.05 mA/cm²and 0.1 mA/cm²respectively. AIMD simulations show that this enhanced interfacial stability in Li₆PS₅F_0.5Cl_0.5SSE is enabled by formation of strong Li-F bonds at the anode/electrolyte interface, which reduce the dissociation of PS₄^3- tetrahedra. These results will be discussed in the context of developing novel solid-state electrolytes for high-performance all solid-state lithium-sulfur batteries.