A Particle-Based Model for Effective Properties in Infiltrated Solid Oxide Fuel Cell Electrodes

Abstract

A modeling framework for the numerical reconstruction of the microstructure of infiltrated electrodes is presented in this study. A particle-based sedimentation algorithm is used to generate the backbone, while a novel packing algorithm is used to randomly infiltrate nanoparticles on the surface of backbone particles. The effective properties, such as the connected triple-phase boundary length, the effective conductivity, the effective diffusivity, are evaluated on the reconstructed electrodes by using geometric analysis, finite volume and random-walk methods, and reported in dimensionless form to provide generality to the results. A parametric study on the effect of the main model and operating parameters is performed. Simulations show that the critical loading (i.e., the percolation threshold) increases as the backbone porosity decreases and the nanoparticle diameter increases. Large triple-phase boundary length, specific surface area and good effective conductivity can be reached by infiltration, without detrimental effects on the effective transport properties in gas phase. Simulations reveal a significant sensitivity to the size and contact angle of infiltrated particles, suggesting that the preparation process of infiltrated electrodes should be properly tailored in order to obtain the optimized structures predicted by the model.