Defect chemistry and ion transport in low-dimensional-networked Li-rich anti-perovskites as solid electrolytes for solid-state batteries

Abstract

Solid-state batteries are attracting significant attention due to their plethora of potential advantages, including energy density gains and safety enhancements. The heart of this technology can be found in its...

Computational Design of Antiperovskite Solid Electrolytes

“…For example, in the recent study of Gu et al that used graph neural networks to predict the synthesizability of perovskites, eight antiperovskite systems, namely, Li 3 CHg, Li 3 BeIr, Li 3 CBe, Li 3 ClSr, Li 3 SPb, Li 3 FK, Li 3 BeTi, and Li 3 PMn, simultaneously displayed good synthesizability scores but low thermodynamic stabilities, especially when compared to those computed for Li 3 OBr and Li 3 OCl. A similar scenario was also observed in our recent work, where the zero-dimensional antiperovskite Li 6 OBr 4 displayed an exceptionally high Li-ion diffusion coefficient of 6.08 × 10 –9 cm 2 s –1 at 300 K but was unstable at higher temperatures and has yet to be experimentally realized. Such examples illustrate the importance of establishing a multitargeted search for novel solid electrolytes that focuses concomitantly on multiple properties (e.g., ionic conductivity, thermodynamic stability, electrochemical stability, and synthesizability) to avoid incomplete analyses that can lead to premature promising status declarations.…”

“…For example, we studied the doping of Li 3 OCl with F – and divalent cations (Mg 2+ , Ca 2+ , Sr 2+ , and Ba 2+ ) via defect and MD simulations and found that while F-doped Li 3 OCl displayed high dopant-vacancy binding energies and low conductivity, its Mg-doped equivalent showed high ionic conductivity and low migration barriers with low dopant-vacancy binding energies. F doping was also predicted to cause high dopant-vacancy binding energies in Li-rich low-dimensional-networked antiperovskites . As a result, a significant reduction in Li-ion diffusion and an increase in activation energy were observed in F-doped Li x OA x –2 (A = Cl, Br; x = 3–6).…”

“…In this context, solid-state batteries (SSBs) are currently capturing significant interest as a next-generation technology endowed with a plethora of potential transformative advantages compared to traditional Li-ion batteries, including improved energy density and safety. ,,− In contrast to commercial rechargeable batteries, which are powered by liquid electrolytes, SSBs utilize solid electrolytes to provide fast ionic conduction between the battery electrodes. As a result, the heart of this technology and the key to its future success lies in the development of solid electrolytes with lower costs, suitable mechanical properties, and high ionic conductivity, stability, synthesizability, scalability, and electrode compatibility.…”

“…Computational modeling plays an important role in tackling these critical challenges, either by predicting novel high-performance solid electrolyte compositions or by providing a deeper understanding of the barriers preventing SSBs from being applied at larger scales. ,,, In the context of antiperovskites, atomistic simulations based on density functional theory (DFT), ab initio molecular dynamics (AIMD), and force-field-based molecular dynamics (MD) have been fundamental in understanding and tailoring a wide array of important phenomena and properties, including ionic and electronic conductivity, diffusion mechanisms, stability, defects and doping, and interfacial resistance and compatibility. Furthermore, machine learning (ML) and high-throughput approaches have proven to be powerful for identifying and screening promising antiperovskite compositions. , Computational modeling has therefore become a vital tool in complementing and assisting the experimental synthesis and characterization of antiperovskites, as well providing key insights that cannot be achieved experimentally.…”

“…This work also opens the interesting question of whether similar results would be expected for space groups with distortions (e.g., octahedral tilting) and for low-dimensionalnetworked antiperovskites. 6,43 One of the many challenges involved in predicting and confirming a material as a promising solid electrolyte is that its properties can very often compromise or contradict each another. For example, in the recent study of Gu et al 25 that used graph neural networks to predict the synthesizability of perovskites, eight antiperovskite systems, namely, Li 3 CHg, Li 3 BeIr, Li 3 CBe, Li 3 ClSr, Li 3 SPb, Li 3 FK, Li 3 BeTi, and Li 3 PMn, simultaneously displayed good synthesizability scores but low thermodynamic stabilities, especially when compared to those computed for Li 3 OBr and Li 3 OCl.…”