Multi-scale Analysis and Phase Segmentation of FIB and X-ray Tomographic Data of Electrolyzer Electrodes Using Machine Learning Algorithms

Abstract

An efficient conversion of electrical energy into hydrogen could be one of the most important key issues for future green energy supply. In case of electrolyzer cells, this can be achieved by increasing the active surface of Ni-based electrodes. This is a crucial topic for the research of alkaline electrolyzer cells in terms of improving the cell performance. Typical coating processes are scalable and effective techniques that are well suited for this purpose. Variations of sintering process parameters (i.e. duration or temperature ramps) have a distinct impact on the homogeneity of the coating layer and its thickness as well as on possible media transport paths during cell operation. In this study, we analyzed the impact of the coating layers on the structure and morphology of the electrodes using laboratory X-ray computer tomography. We introduce a workflow that is based on machine learning algorithms which enable to distinguish between substrate and coating layers. More detailed information was obtained using FIB/SEM (focused ion beam/scanning electron microscopy). This multi-scale approach enhances the understanding of the sintering processes and serve as a basis for CFD simulations.