Expediting Ionic Conductivity Prediction of Solid-State Battery Electrodes Using Machine Learning

Solid-state batteries can offer enhanced safety and potentially higher energy density compared to traditional lithium-ion batteries. However, their power density remains a challenge due to limited ionic conductivity in composite electrodes caused by non-ideal microstructures. Laborious experimental...

Full description

Saved in:
Bibliographic Details
Published inIEEE journal on multiscale and multiphysics computational techniques Vol. 9; pp. 375 - 382
Main Authors Le, Mai, Yao, Alan, Zhang, Amie, Le, Hieu, Chen, Zhaoyang, Wu, Xuqing, Zhao, Lihong, Chen, Jiefu
Format Journal Article
LanguageEnglish
Published IEEE 2024
Subjects
Batteries
Conductivity
Electrodes
finite difference method
Finite difference methods
Frequency division multiplexing
Image analysis
Machine learning
Microstructure
Noise reduction
numerical modeling
Numerical models
Predictive models
Solid state drives
solid-state batteries
Online AccessGet full text

Cover

More Information
Summary:Solid-state batteries can offer enhanced safety and potentially higher energy density compared to traditional lithium-ion batteries. However, their power density remains a challenge due to limited ionic conductivity in composite electrodes caused by non-ideal microstructures. Laborious experimental processes and time-consuming data analysis algorithms are obstacles to establishing structure-performance correlations and optimizing electrode microstructure. In this paper, we present a machine learning approach to predict the effective conductivity of a composite electrode based on scanning electron microscopy images, using binary images composed of conductive and non-conductive regions and an ionic conductivity value of the conductive region. We show that our proposed method is two orders of magnitude more efficient than conventional numerical schemes such as the finite difference method.
ISSN:2379-8815
2379-8815
DOI:10.1109/JMMCT.2024.3475988