New tool predicts lithium battery failure, could help make safer EVs

Researchers at the University of California, San Diego have developed a new, straightforward method to accurately measure lithium metal battery performance using scanning electron microscopy (SEM) combined with an algorithm. This innovation addresses a critical challenge in lithium metal batteries: uneven lithium deposition on electrodes, which leads to the formation of dendrites—spiky lithium structures that can pierce battery separators, causing short circuits and battery failure. Previously, assessments of lithium deposit uniformity were subjective and inconsistent across labs, hindering progress in battery research. The team created an algorithm that analyzes SEM images by converting them into black-and-white pixels representing lithium deposits and calculates an Index of Dispersion (ID) to quantify lithium uniformity. A lower ID indicates more uniform lithium deposition, while a higher ID signals clustering and potential battery degradation. Validated on over 2,000 computer-generated images and real battery tests, the ID score correlated with battery health, with fluctuations in the score serving as early warnings of impending failure. This accessible method, leveraging standard