Mass-driven natural aging performance prediction model of thermal battery considering the multistage corrosion mechanism of anode

The natural aging of thermal battery manifests as a continuous decrease in discharge performance, attributable to the corrosion-induced surface evolution of lithium metal anode. Due to the insufficient understanding of anode corrosion mechanism, evaluating the effects of natural aging persists as a critical barrier. This study identifies the multistage corrosion mechanism of Li₁₃Si₄ anode in natural aging, facilitated by self-built humidity-temperature collaborative platform. On this basis, a multistage corrosion mass model is proposed for corrosion state prediction in different environments. Post-mortem analysis reveals distinct corrosion stages: initial surface corrosion dominated by lithium-hydroxide formation, followed by the combination of oxygen with the active metal through gas channels, which eventually led to the corrosion accumulation. These morphological and compositional transitions degrade the discharge performance. Furthermore, according to Nernst equation and resistance model, the effects of corrosion products on discharge performance is analyzed, and a mass-driven performance degradation prediction model is established. Compared with 6-year natural aging data, the mean error of performance prediction is 2.48 % which verifies the model has high accuracy. This work provides basic insights into the natural aging mechanism of metal anodes and establishes a method for predicting the performance degradation of power-sources considering corrosion state and discharge mechanism.