Researchers at the Pritzker School of Molecular Engineering at the University of Chicago have developed a new AI model called ElectrolyteGPT. This technology is capable of not just selecting individual chemical components for batteries, but creating ready-to-use electrolyte formulas. Electrolytes are key elements of modern batteries, and their composition directly affects capacity, charging speed, and safety. This is reported by Ixbt.com .
Developing new electrolytes is one of the most complex tasks in materials science, as the number of potential molecules exceeds 10⁷⁰. Unlike existing solutions, the ElectrolyteGPT model offers ready-made recipes that include component concentrations and mixing ratios. This process is performed several times faster than traditional laboratory research.
The team, led by Professor Chibueze Amanchukwu, tested the neural network's recommendations in a laboratory setting. The results showed that some compositions suggested by the AI can compete in their properties with solutions in the most advanced lithium-metal batteries. This proves that AI can work at the level of experienced materials design experts.
During the system training process, scientists formed a special database. In the initial stage, the model relied on pharmaceutical data to suggest medical molecules. However, after being retrained specifically on electrolyte-related compounds, ElectrolyteGPT began creating chemically stable and useful compositions for batteries.
Currently, all AI suggestions are undergoing laboratory verification. Nevertheless, ElectrolyteGPT is analyzing new material variants much faster than humans. This technology is expected to significantly accelerate the creation of more efficient batteries for electric vehicles, gadgets, and energy storage systems in the future.
Read “Zamin” on Telegram!
