Solid oxide fuel cells (SOFCs) have been recognized as one of the most potent next-generation energy conversion systems, owing to their higher efficiency, better fuel flexibility, and smaller amount of pollution, unlike their counterpart fuel cells. SOFCs can be analyzed as a highly-complex multi-component materials system consisting of porous electrodes (cathodes and anodes), densified electrolytes, interconnects, and sealing materials, which are sequentially layered and characteristic of heterogeneous and multi-phase materials. The resultant power generation is achieved through the electrochemical reactions involving electronic carriers, ionic carries, and gaseous transports at anodes and cathodes, influencing their electrical, mechanical, and thermal properties. In particular, both cathodes and anodes are typically constructed based on three components: mixed ionic/electronic electrodes, electrolytes, and a porous section. Those three constituents should be optimized at both electrodes with the aim of maximizing the electrochemical reaction sites, i.e., triple-phase boundaries in association with the gaseous transport involved in oxygen reduction reactions at cathodes and hydrogen oxidation reactions at anodes, which leads to high cell performance and guaranteed durability. The electrochemical performances are crucially dependent on their complex microstructures, demanding for highly informative analyses on the electron microscopy images under investigation.This work reports quantitative microstructural analysis methodology incorporating machine learning-assisted microstructural analysis strategy. Quantitative microstructural interpretations were carried out without human involvement through an integrated combination of deep learning and focused ion beam-scanning electron microscopy (FIB-SEM) analytics on Ni/Y2O3-stabilized ZrO2 (Ni/YSZ) cermets. The Ni/YSZ/pore composites were analyzed for the automated extraction of microstructural parameters with the aim to preventing the subjective analysis problems and unavoidable artifacts frequently encountered in lengthy image processing tasks and eliminate biased evaluations. Considering the high volume of image data and future expectations for electron microscopy usage, FIB-SEM was efficiently combined with semantic segmentation targeted at detailed phase recognition on three constituents, i.e, Ni, YSZ, and pores. Traditional image processing analysis tools are in the synergistic manner combined with phase separation predictions by semantic segmentation algorithms, allowing for a quantitative evaluation of microstructural parameters. The combined strategy enables one to significantly enhance poor image quality originating from artifacts in electron microscopy, including charging effects, curtain effects, out-of-focus problems, and unclear phase boundaries encountered in searching for high-efficiency solid oxide fuel cells (SOFCs).
Read full abstract