On the 10th, Lee Byung-joo, a senior researcher at the Korea Institute of Science and Technology (KIST), noted that a research team discovered the mechanism by which lithium ions move in solid electrolytes, a key material for all-solid-state batteries, and proposed a new design strategy utilizing artificial intelligence (AI). This research was published on December 27 of last year in the international journal Advanced Energy Materials.
In batteries, electrolytes serve as pathways for ion movement. Traditionally, liquid electrolytes have been primarily used. However, solid electrolytes, which pose a lower fire risk and are structurally stable, are garnering attention as key materials for next-generation batteries, specifically all-solid-state batteries.
The challenge is that solid electrolytes have lower ionic conductivity, which indicates how well ions move compared to liquid electrolytes, and they exhibit significant friction at the interface with electrodes, as well as unexpected chemical reactions that degrade performance. As a result, there have been difficulties in commercializing all-solid-state batteries.
To address these issues, the research team collaborated with POSCO Holdings' Applied AI Research Team to develop computing simulations utilizing AI. While traditional methods had slow computational speeds that made accurate predictions difficult, the research team applied machine learning technology that allows AI to learn on its own, enabling more precise predictions of how atoms within the battery move. This advancement allowed the research team to analyze the movements of all-solid-state battery materials containing over 3,000 atoms at nanosecond (ns, one billionth of a second) intervals.
Using this technology, the research team analyzed the movement of lithium ions and discovered that sulfur ions hinder the movement of lithium ions within the battery. The team confirmed that optimizing the quantity and arrangement of sulfur ions could enable lithium ions to move faster, with speeds increased by up to 100 times. To validate this, the results from comparing and analyzing existing experiments showed a high correlation, proving its reliability.
This research is expected not only to contribute to the commercialization of all-solid-state batteries but also to serve as a foundation for the development of next-generation battery technologies in various fields, such as electric vehicles, energy storage systems (ESS), and wearable devices. Additionally, the AI-based simulation techniques developed by the research team can be applied to the exploration and optimization of ionic conductive materials, showing significant potential for expansion into industries such as semiconductors, hydrogen fuel cells, and catalytic materials.
Senior researcher Lee Byung-joo said, "By establishing a methodology for developing high-performance battery materials incorporating AI technology, we aim to provide important guidelines for subsequent research and development." He added that he expects this to contribute to the realization and commercialization of high-performance all-solid-state batteries.
References
Advanced Energy Materials (2024), DOI: https://doi.org/10.1002/aenm.202402396