By clicking “Accept All Cookies”, you agree to the storing of cookies on your device to enhance site navigation, analyze site usage, and assist in our marketing efforts. View our Privacy Policy for more information.
PreferencesDenyAccept

AI-accelerated design of next-generation battery cathode material

A collaboration between Polaron and WMG combines WMG’s expertise in lithium manganese iron phosphate (LMFP) technology and electrode manufacturing with Polaron’s machine learning-driven microstructural optimisation.

Polaron’s AI tool offers an impactful and rapid way of analysing and improving battery production and performance, reducing both development time and cost associated with the adoption of new chemistries.
Dr Gerard Bree
Assistant Professor, WMG
Manufacturing battery electrodes involves complex, interdependent processes where factors such as particle size, composition, and processing conditions directly impact performance. Traditional optimisation methods rely on iterative prototyping, which is costly and time-consuming. By applying AI-driven characterisation, reconstruction, and optimisation, Polaron enables data-driven microstructural optimisation of LMFP electrodes. This project is utilising Polaron’s AI tools to guide LMFP electrode manufacturing to improve cell performance, deepen understanding of LMFP processing, and demonstrate the value of AI-driven process optimisation for wider adoption across the battery industry and beyond.

The manufacture of lithium manganese iron phosphate (LMFP) battery electrodes involves multiple interdependent processes, including material preparation, mixing, coating, drying, and pressing. Each step affects the electrode’s microstructure, influencing cell performance and long-term stability. However, understanding these relationships using traditional experimental approaches requires extensive prototyping, making development slow and costly.

‍

As part of this project, WMG, University of Warwick, initially manufactured a range of different battery cells, each produced with variations in three key manufacturing parameters. The collected data included both high-resolution 2D imaging data of the electrodes and electrochemical cycling data from battery performance tests.

‍

Using the Polaron platform, firstly rapid batch-segmentation was performed using our AI-based algorithms to process the 2D electrode images, identifying key microstructural features. These segmented images were then reconstructed into 3D electrode models, providing insight into battery transport resistivity and diffusivity properties that can only be observed in three dimensions. By integrating these reconstructions with electrochemical data, Polaron’s platform determined the optimal manufacturing parameters for improving cell performance, which WMG is now implementing in production.

‍

By integrating machine learning into the manufacturing workflow, this project is improving efficiency, reducing production costs, and accelerating the commercialisation of LMFP technology. The findings will contribute to the wider adoption of AI in battery manufacturing, supporting the development of next-generation energy storage solutions.

‍