Degradation Mode Analysis for Lithium-Ion Cells with Silicon-Dominant Anodes Using Reference Electrodes

The degradation mode analysis widely investigates lithium-ion cell aging based on voltage reconstructions. In this work, we adapt and validate this degradation mode analysis method for silicon-dominant anodes with feature-poor potentials. Our motivation is to investigate why reducing the capacity utilization improves lifetime significantly. Firstly, different initial silicon anode and NCA cathode potential parametrizations are investigated. Using T-cells with lithium metal reference electrodes, the fitted anode and cathode potentials are validated by the measured ones. Secondly, different algorithm parameters for the voltage reconstruction are investigated for reliable results. The resulting root mean square error (RMSE) between the measured and the reconstructed voltages is ≈3.1 mV at the beginning-of-life and stays below 15 mV over aging until ≈55% state-of-health (SoH). The main aging mechanism is the loss of lithium inventory at ≈49%, with a linear correlation to the SoH. The loss of cathode active material is minor, at 1.0% to 3.2%, dependent on the voltage window. The apparent loss of anode active material is initially negative due to the gain of amorphous silicon. The degradation modes lead to a significant change in the electrode balancing over aging, reproducing the measured increasing anode and cathode potentials. Finally, coin cell and previous post-mortem results validate the T-cell results.