Characterizing low frequency Li-ion resistance variation with aging using Laplace transformation

Li-ion cells change significantly in their performance over time due to multiple aging mechanisms each sensitive to usage parameters such as temperature, charge rate and usable state-of-charge (SoC) range. This can cause cells even of identical design to age differently based on lifetime usage conditions. It is therefore important to employ methods able to quantify the changes in resistance components due to aging and give insight into the underlying causes. This work explains an automated approach to extract the diffusion resistance from voltage relaxation curves. This combines Laplace inversion, finite length Warburg elements and particle swarm parameter optimization to fit a frequency domain diffusion model to time domain data. This approach was demonstrated using design-of-experiments aging tests performed as part of the work. Nickel-Manganese-Cobalt/Gr cells were cycled for 9 months at 45 °C using an automotive duty cycle. Relaxation points were taken at 10% SoC intervals in monthly reference tests. The charging protocol and SoC range was varied across cells to evaluate the relative impact of these parameters. Analysis showed a common trend of all cells with an average resistance decrease in the first month, and a noticeable increase in resistance towards month 9. This was shown through Scanning Electron Microscopy to arise from particle cracking in the cathode. Diffusion resistances showed localized trends due to the movement of electrode phase changes relative to cell state-of-charge. Diffusion time constants elongate consistently over life. This shows control systems must adapt for complex resistance parameter changes and significant dynamic effects during lifetime.