Method for In-Operando Contamination of Lithium Ion Batteries for Prediction of Impurity-Induced Non-Obvious Cell Damage

The safety of lithium-ion batteries within electrified vehicles plays an important role. Hazards can arise from contaminated resulting non-obvious damages or insufficient production processes. A systematic examination requires experimental methods to provoke a defined contamination. Two prerequisites were required: First, the extent and type contamination should be determinable exclude randomness. Second, specimens work properly before contamination, enabling realistic behavior. In this study, two developed allow for first time controlled reproducible application water oxygen into 11 single-layer full cells (Li4Ti5O12/LiCoO2) used as during electrical cycling. Electrochemical impedance spectroscopy was continuously monitor fit parameters equivalent circuit model (ECM). For time, these calibrate machine-learning algorithm which able predict state. decision tree calibrated with ECM eight (training data) validated by predicting state three remaining (test data). prediction quality proved usability classification algorithms contaminations battery damage after manufacturing use. It integral part management systems that increases vehicle safety.