Researchers at Skoltech and the AIRI Institute have teamed up with machine learning techniques to fast-track the discovery of materials for solid-state lithium-ion batteries. By replacing flammable liquid electrolytes with robust solid materials such as ceramics, these batteries could soon offer safer, longer-range electric vehicles.
In a recent study published in npj Computational Materials, neural networks swiftly identified promising substances for both solid electrolytes and protective coatings. While automakers have yet to widely adopt solid-state batteries, the potential to boost vehicle range by up to 50% and improve fire safety makes them a compelling alternative. Still, finding materials that meet all technical requirements remains challenging.
Ph.D. student Artem Dembitskiy explained that graph neural networks can pinpoint new battery materials with high ionic mobility much faster than traditional quantum chemistry methods. This acceleration in discovery could pave the way for more efficient development of battery components, including various protective coatings.
Assistant Professor Dmitry Aksyonov elaborated on the vital role of protective coatings. He noted that the metallic lithium at the anode is a strong reducing agent, causing most current electrolytes to degrade when in contact. Similarly, the oxidising nature of the cathode can compromise electrolyte structure, potentially leading to performance issues or short circuits. Hence, stable protective coatings are essential to ensure overall battery reliability.
The machine learning models used in this research also predict ionic conductivity—a key property in both electrolytes and coatings—thereby reducing the time required for such complex calculations. The study identified promising coating materials like Li₃AlF₆ and Li₂ZnCl₄, highlighting the potential of AI to streamline the discovery process and deliver safer, more efficient batteries.
If you’ve ever been frustrated by slow progress in research and development, this approach offers a glimpse into faster, smarter ways of solving complex challenges in battery technology.
