Managing silicon burn-out via onboard material diagnostics for durable high-energy density batteries

Silicon-graphite (Si/Gr) anodes increase battery energy density, but rapid Si "burn-out" limits lifetime. Yet, battery management systems lack effective tools to diagnose Si aging beyond capacity loss and to manage its degradation during operation. Here, we develop a diagnostic framework that jointly estimates material-specific health and degradation-induced Si open-circuit potential deformation from constant-current charge data. The framework tracks the transition state-of-charge (SoC) below which Si dominates, revealing pathway-dependent Si utilization shifts: from 48% fresh to 73% with lithium loss but 33% with active-Si loss. Transition SoC error remains below 3% up to an effective 0.3C, and error bounds across sampling and charge windows support onboard use. Guided by the finding that elevated temperature improves Si capacity retention in our cells, we propose adaptive thermal management that warms the cell during Si-dominant operation and cools it otherwise, projecting an approximately doubled cycle life without limiting capacity access.