About this Abstract |
Meeting |
MS&T22: Materials Science & Technology
|
Symposium
|
2022 Undergraduate Student Poster Contest
|
Presentation Title |
Molecular Dynamics Simulations with Machine Learning Potential for Amorphous Li7La3Zr2O12 |
Author(s) |
Ziyao Luo |
On-Site Speaker (Planned) |
Ziyao Luo |
Abstract Scope |
Li7La3Zr2O12 (LLZO) is considered a promising candidate for solid-state electrolytes (SSE) due to various advantageous properties. However, polycrystalline LLZO suffers from Li dendrite penetrating through the grain boundaries. Amorphous LLZO (a-LLZO) can solve the problem of dendritic penetration, but the relatively low Li-ion conductivity (~10-6 S/cm) limits its use as bulk SSE. Simulations of doped a-LLZO are conducted using molecular dynamics with a generic machine learning potential (trained by ab initio data), and the Li-diffusivity is analyzed with materials analysis libraries such as Pymatgen. The results indicate that the generic MLP is potentially capable of yielding accurate results, and doping can indeed enhance the Li+ conductivity in a-LLZO. This study sheds light on the structural mechanisms affecting Li+ mobility in amorphous Li-ion conductors and provides insight into the design of a-LLZO with high Li-ion conductivity. |