AI-enabled materials discovery for advanced ceramic electrochemical cells

Abstract

Ceramic electrochemical cells (CECs) are promising devices for clean and efficient energy conversion and storage due to their high energy efficiency, more extended system durability, and less expensive materials. However, the search for suitable materials with desired properties, including high ionic and electronic conductivity, thermal stability, and chemical compatibility, presents ongoing challenges that impede widespread adoption and further advancement in the field. Artificial intelligence (AI) has emerged as a versatile tool capable of enhancing and expediting the materials discovery cycle in CECs through data-driven modeling, simulation, and optimization techniques. Herein, we comprehensively review the state-of-the-art AI applications for materials design and optimization for CECs, covering various material aspects, database construction, data pre-processing, and AI methods. We also present some representative case studies of AI-predicted and synthesized materials for CECs and provide insightful highlights about their approaches. We emphasize the main implications and contributions of the AI approach for advancing the CEC technology, such as reducing the trial-and-error experiments, exploring the vast materials space, discovering novel and optimal materials, and enhancing the understanding of the materials-performance relationships. We also discuss the AI approach's main limitations and future directions for CECs, such as addressing the data and model challenges, improving and extending the AI models and methods, and integrating with other computational and experimental techniques. We conclude by suggesting some potential applications and collaborations for AI in materials design for CECs.

Keywords

• Ceramic electrochemical cells
• Artificial intelligence
• Materials design
• Materials optimization
• Materials performance
• Machine learning