Simulations of Activity and Degradation in Solid Oxide Fuel Cells as a function of Electrode Microstructures

Several characterization techniques to quantify electrode microstructure in threedimensional (3D) space, such as focused-ion-beam scanning-electron-microscopy (FIBSEM) and X-ray nanotomography, have recently been improved to reconstruct very large volumes of electrodes. This progress in micro-scale 3D characterization has provided us a remarkable opportunity to connect electrode microstructure to performance measures, such as electrochemical activity and degradation. Conventional models such as effective medium theory (EMT) have been widely used to quantify the relationship between average microstructure and average performance in electrodes of solid oxide fuel cells (SOFCs). When microstructural distributions are narrow, as thought to occur in research-grade SOFCs, the average value may be sufficient to describe performance. When microstructural distributions are broad, however, as observed in commercial-grade SOFCs, or when performance is driven away from the mean, via processes such as degradation, models using average parameters are insufficient. 

To improve electrode microstructures and deal with complexity of microstructures, full-scale 3D microstructurally resolved simulations are required. The software “Electrochemical Reactions in Microstructural Networks” (ERMINE) developed by our group is an application based on the Multiphysics Object Oriented Simulation Environment (MOOSE) that allows us to investigate local electrochemistry of electrode microstructures. Thus, this work aims to quantify local distributions of electrochemical parameters throughout the microstructure in heterogeneous electrodes to discover the relationship between microstructure and electrode activity and degradation. 

In the first part of thesis, the effect of different types of infiltrates, electronic conductors or ionic conductors, on performance in experimental cathode microstructures is explored using ERMINE simulations. While infiltration is a well-known method to improve SOFC activity, it is challenging if not impracticable to extract the precise origin of improvement from experiments. Simulations thus provide insight that is otherwise unobtainable. The improved performance of ionic conductors as infiltrates in these cathode microstructures is demonstrated and its origin clarified. In the next part of the thesis, systematic trends for local overpotential-driven degradation are investigated using ERMINE simulations. One commonly observed degradation mechanism is chromium poisoning in the cathodes of SOFCs, whereby electrochemical deposition of Cr species at active sites decreases their performance towards oxygen reduction. Simulations provide better understanding of the performance degradation by active site poisoning based on local electrochemical values. Using large amounts of simulation data makes it possible to correlate systematic trends in SOFC cathodes to given operating conditions. 

In the last part of the thesis, the impact of microstructural heterogeneity on electrode activity is investigated, again via ERMINE simulations. Prior work with ERMINE has focused on experimental cathode microstructures and synthetic microstructures aimed at reproducing those experimental microstructures. Improving our understanding of microstructure-property-performance requires running simulations on different datasets with varied microstructural distributions. Towards this end, the last part of this thesis focuses on experimental anode microstructures with distinctive differences from the cathodes and which have also been operated to yield degraded microstructures.