Accelerated Discovery of Novel Garnet-Type Solid-State Electrolyte Candidates via Machine Learning

Abstract

All-solid-state batteries (ASSBs) have attracted considerable attention because of their higher energy density and stability than conventional lithium-ion batteries (LIBs). For the development of promising ASSBs, solid-state electrolytes (SSEs) are essential to achieve structural integrity. Thus, in this study, a machine-learning-based surrogate model was developed to search for ideal garnet-type SSE candidates. The well-known Li 7 La 3 Zr 2 O 12 structure was used as a base material, and 73 chemical elements were substituted on La and Zr sites, leading to 5329 potential structures. First, the elasticity database and machine learning descriptors were adopted from previous studies. Subsequently, the machine-learning-based surrogate model was applied to predict the elastic properties of potential SSE materials, followed by first-principles calculations for validation. Furthermore, the active learning process demonstrated that it can effectively decrease prediction uncertainty. Finally, the ionic conductivity of the mechanically superior materials was predicted to suggest optimal SSE candidates. Then, ab initio molecular dynamics simulations are followed for confirmation of diffusion behavior for materials classified as superionic; 10 new tetragonal-phase garnet SSEs are verified with superior mechanical and ionic conductivity properties. We believe that the current model and the constructed database will become a cornerstone for the development of next-generation SSE materials.

Discovery of Superionic Solid-State Electrolyte for Li-Ion Batteries via Machine Learning

“…We recognized that a method of obtaining ionic conductivity at room temperature by extrapolating a result at high temperature could lead to a significant error for some cases. However, since such a method has been widely employed for this type of research, – we strongly believe that current approach is still reliable at least for qualitatively comparing the ionic conductivity values of considered materials.…”

h-MBenes (M/B = 1:1) as Promising Electrocatalysts for Nitrogen Reduction Reaction: A Theoretical Study

“…50,54 Therefore, h-MABs are more advantageous in the study of catalytic performance than o-MABs. Moreover, theoretical 55 and experimental 56,57 studies have demonstrated the promising exfoliation feasibility in the M/B = 1:1 family (h-M 2 AB 2 and h-(M′ 2/3 M″ 1/3 ) 2 AB 2 ). In 2020, Rosen and co-workers successfully synthesized twodimensional Mo 4/3 B 2−x and reported their excellent HER catalytic activity.…”

Machine-Learning-Driven High-Throughput Screening for High-Energy Density and Stable NASICON Cathodes

“…On the other hand, in combination with DFT and ML, it is possible to investigate the extensive exploration space through employing high-throughput computation for the database construction and surrogate model for rapid inference, thereby accelerating the discovery of promising materials. Thus, in the research field of rechargeable batteries, various studies combining ML and DFT are actively progressing, achieving remarkable outcomes: (1) solid-state electrolyte screening, − (2) cathode development, , and (3) anode design are some representative examples.…”