Cycling-induced damage of silicon-based lithium-ion batteries: Modeling and experimental validation

Currently, there are few analyses available to quantitatively uncover the effects of cycling-induced damage in silicon-based batteries on the stress evolution and capacity loss. We develop a comprehensive model to address this issue. The comparisons between numerical and experimental results validate the proposed model and illustrate the damage effects on the decrease of structural stiffness and the stress evolution. In contrast to the common perception that damage is unfavorable to batteries, we propose a concept of training of batteries which introduces sufficient damage to batteries to improve retention performance. The training process is implemented by electrochemical cycling under a large C-rate, and its merit is validated experimentally. We also discuss optimization of the training method.

Automated summary

Currently, there is a lack of analysis methods to quantitatively study the effects of cycling-induced damage on the stress evolution and capacity loss in silicon-based batteries. We propose a comprehensive model to address this issue, which is validated through comparisons with experimental results. Contrary to common belief, we suggest a concept of training batteries by introducing sufficient damage to improve retention performance, and its effectiveness is experimentally confirmed. Optimization of the training method is also discussed.