Item

Abstract

Electrode degradation currently limits the wide-spread commercial adoption of highly efficient solid oxide cell based energy storage systems. While the phenomenon has been reported and studied excessively, the structure of the active site of the relevant oxygen evolution reaction located at the triple phase boundary in electrolysis mode is elusive. This prevents the detailed analysis the actual degradation mechanism, which is crucial for the systematic development of mitigation and prevention strategies.
In this work, the fundamental structure of the triple phase boundary and its underlying complexion at the solid/solid interface are uncovered by physics based simulations. The simulations uncover a new deactivation pathway by cation segregation trough the complexion. The findings are supported by experimental data, whose analysis is aided by a newly developed, robust workflow to detect minute concentrational changes of electron microscope images. To enable the access to the active sites’ electronic structure, three neural network based generative models for inverse material design are assessed and adequate metrics for model training and comparison developed.