ON-BOARD ESTIMATION OF SILICON BURN-OUT IN LITHIUM-ION BATTERIES WITH SILICON-GRAPHITE COMPOSITE ANODES

Silicon-graphite (Si/Gr) composite anodes are highly attractive for modern batteries because the addition of silicon significantly boosts energy density. However, these anodes suffer from a complex aging process often referred to as 'silicon burn-out,' driven by mechanisms such as particle cracking and the continuous formation of the solid electrolyte interphase (SEI). Existing diagnostic tools, like incremental capacity analysis or standard electrode state-of-health tracking, struggle to distinguish between the degradation of the silicon phase and the graphite phase because their thermodynamic responses overlap significantly.

To overcome these limitations, the authors introduce a new, battery management system (BMS)-compatible diagnostic framework. This approach combines two distinct signals: the unique thermodynamic fingerprint of the composite electrode and the kinetic effects induced specifically by silicon during charge and discharge cycles. By mapping these signals, the algorithm can separately monitor how the material properties of the silicon and graphite components evolve over the lifetime of the battery.

Because this tool is designed for on-board use in electric vehicles, the authors focused heavily on practical constraints. The framework is designed to function even with the messy, non-ideal data generated by real-world usage—such as variable charging rates, sporadic partial state-of-charge windows, and inconsistent sampling rates. Furthermore, by providing explicit error bounds for the estimated metrics, the framework allows for a more reliable interpretation of battery health in diverse, uncontrolled driving scenarios, helping engineers effectively track the specific contributions of each material to overall performance loss.