Detecting electrochemical changes in a nickel-zinc battery by operando x-ray computed-tomography analysis during charge-discharge tests

This study developed a method to detect and analyze deterioration in sections of a battery during charge-discharge tests in real time. This method is based on a time-series analysis using x-ray computed tomography, three-dimensional reconstruction of the battery volume, and unsupervised machine learning. The developed method detects not only electrochemical changes in a battery through the conventional voltage-capacity diagram but also physical changes such as the deterioration of the parts of a battery that cannot be found via human inspection directly from the sliced images of the three-dimensional reconstructed volumes. Furthermore, the characteristics of these changes inside a battery can be captured through precise analysis using persistent homology, a mathematical machinery, at degrees 0 and 1. This demonstrates that our method can capture both continuous and discrete structural changes (e.g., a continuous deformation of active materials and compounds that are precipitated randomly in the electrolytes) within a battery. As a by-product, the start of the venting system implemented near the anode of a battery can be detected using the method from a specific cycle during the charge-discharge tests.