Enhancing high-throughput battery materials discovery and design with artificial intelligence
Year: 2026
Type: article
Abstract: Battery material discovery and design face sophisticated multi-objective, multi-scale optimization challenges in high-dimensional variable spaces, a core bottleneck limiting advanced energy storage device upgrading. Prompted by the Materials Genome Initiative (MGI), high-throughput computation (HTC) has become a powerful approach to address these challenges via high-throughput automated computational workflows and hierarchical screening technologies. With the continuous accumulation of battery materials data and advances in artificial intelligence (AI) development, HTC is well-positioned to integrate with AI for accelerating the discovery and design of battery materials. Herein, we thoroughly analyze the challenges and strategies for achieving this goal, from the fundamentals to the applications of HTC. The crucial hurdles in HTC, e.g., data waste, dimension disaster, multi- and cross-scale calculations, data quality control, and the experimental validation gap, are highlighted, and strategies to overcome those are provided, such as the integration of HTC and AI with experimentation in a closed-loop workflow under the framework of MGI. Looking ahead, the development of HTC for battery materials discovery and design presents compelling opportunities, particularly through synergistically integrating HTC with data mining and large language models. The presented insights and strategies are expected to stimulate work on the deep integration of HTC and AI, to guide the discovery and design of battery materials as well as other functional materials.
Source: eScience
Authors: Qian Zhao, Bing He, Yurong Ren, Maxim Avdeev, Yuxiao Lin, Siqi Shi
Language: English
Subfield: Materials Chemistry
Field: Materials Science
Domain: Physical Sciences
Open Access status: diamond
Awards: G2323951130, G4953414827, G5689145776
OpenAlex ID: W7204056623