Quantify the impact of fast-charging cycling on crack evolution in electrode material through machine learning algorithms

The development of nickel-rich layered oxide (NRLO) cathode materials for power batteries, aiming to solve the range anxiety problem of electric vehicles. However, NRLO materials face the challenge of structural degradation during fast-charging cycling process. Therefore, it is necessary to comprehensively understand the crack evolution mechanism and the relationship between the capacity degradation and the severity of crack in NRLO materials. In this paper, the quantitative statistical analysis of the cracks in LiNi0.8Co0.1Mn0.1O2 (NCM811) particles is conducted by processing high-resolution Scanning Electron Microscope (SEM) image with machine learning algorithms. The evolution mechanism of NCM811 particle cracks is identified under fast-charging cycles. Through extracting the features of crack, the quantitative relationship of capacity fading and microcrack is explored. The understanding of the continuous crack evolution mechanism for NRLO materials can provide an internal perspective for optimizing the structural design of cathode materials. Machine learning-based image recognition methods are crucial for precisely analyzing the mechanisms of aging performance degradation in batteries. These meticulously quantified crack characteristics play a crucial role in advancing the development and refining the validation of lithium-ion battery life degradation models.