December 1 - 6, 2024
Boston, Massachusetts
Symposium Supporters
2024 MRS Fall Meeting & Exhibit
MT04.05.16

Revolutionizing Battery Development with AI

When and Where

Dec 3, 2024
8:00pm - 10:00pm
Hynes, Level 1, Hall A

Presenter(s)

Co-Author(s)

Jie Liu1,2,Xiao Shen3

The University of Hong Kong1,Hong Kong Quantum AI Lab2,The Australian National University3

Abstract

Jie Liu1,2,Xiao Shen3

The University of Hong Kong1,Hong Kong Quantum AI Lab2,The Australian National University3
The global lithium-ion battery market achieved a shipment volume of 1,203 GWh in 2023, with a compound annual growth rate (CAGR) of 27.4% and a price learning rate (LR) of 27.3% over the past 20 years. However, rapid battery industry growth and substantial resource consumption have made increasing energy density crucial for sustainable development. Despite significant research progress, energy densities of mainstream LFP and NCM batteries remain around 160-250 Wh/kg, with an energy density learning rate (EDLR) of 5-7%, indicating a 5-7% improvement in energy density each time production doubles. This is in stark contrast to the 17.9% EDLR for best practice batteries, which achieve energy densities exceeding 700 Wh/kg. This substantial gap necessitates a shift from traditional R&D paradigms to an AI -driven approach for battery development, with their full potential impact yet to be completely realized.<br/>In this study, we provide a comprehensive framework for AI driving battery design by integrating quantum chemistry, domain knowledge, databases, knowledge graphs, expert systems, large language models, and automated experiments, which could help bridge the EDLR gap between academic research and practical application. We assess its potential contributions to the future development of lithium batteries, with a focus on material efficiency and cost savings. A scenario analysis under different EDLR conditions for mainstream batteries reveals that even slight improvements in EDLR can lead to substantial material savings. Specifically, if the EDLR of mainstream batteries matches the 17.9% rate of best practice batteries by 2030, the industry could save approximately 16.1 Mt of battery materials and 151 billion USD in material costs.<br/>The findings underscore the significant advantages, both economically and in terms of resource conservation, of accelerating lithium battery R&D. There is a compelling case for seizing the opportunity to integrate AI and Quantum Chemistry in lithium battery R&D. This integration not only promises enhanced energy densities but also marks a significant stride towards more sustainable and cost-effective battery production, potentially yielding returns far greater than the current investments in these cutting-edge technologies.

Symposium Organizers

Kjell Jorner, ETH Zurich
Jian Lin, University of Missouri-Columbia
Daniel Tabor, Texas A&M University
Dmitry Zubarev, IBM

Session Chairs

Kjell Jorner
Jian Lin

In this Session