Rationalizing the interphase stability of Li|doped-Li₇La₃Zr₂O₁₂via automated reaction screening and machine learning

Abstract

Lithium metal batteries are a promising candidate for future high-energy-density energy storage.

Modelling Bulk Electrolytes and Electrolyte Interfaces with Atomistic Machine Learning

“…Liu et al . established an automated workflow using DFT calculations to determine all possible reactions when Li 7 La 3 Zr 2 O 12 (LLZOM, M=dopant) comes into contact with differing amounts of Lithium and trained a KRR model with 100 data to map 15 structural features (such as the coordination number) to the reaction energy for Li|LLZOM interfaces [147] . With an RMSE of 0.04 eV for the reaction energy, their KRR model predicted 18 unexplored LLZOM systems, which were then validated by DFT calculations.…”

Recent progress on discovery and properties prediction of energy materials: Simple machine learning meets complex quantum chemistry

“…With the help of MGI, ML can be used to mine the data of new reaction processes, new properties, or new structures. For example, high-throughput automated reaction screening and ML approaches have been introduced to mimic the interface reactions of solid electrolyte of lithium [144]. It has been well known that the solid electrolyte interface formed on Li metal surface will protect cells from Li-dendrites.…”

High-throughput computational screening of oxide double perovskites for optoelectronic and photocatalysis applications

“…Moreover, this method can be used in other fields, such as semiconductor light-emitting diodes and transparent conducting materials [31]. In addition, novel materials design via machine learning has drawn much interest in many fields, such as photovoltaic materials [32], lithium-ion battery materials [33,34], and glass material system [35,36]. Moreover, this approach is being optimized and updated continually [37,38].…”