Design-Led Solid Oxide Fuel Cell Manufacturing: Using 3D Imaging and Modeling to Optimize Electrode Performance

Abstract

Solid oxide fuel cells are among the most thermodynamically efficient energy conversion devices. They are critical for the future if we are to meet rising energy demands and increasing selectivity over our future energy mix. Traditionally, solid oxide fuel cell performance has been assessed using bulk parameters, and the microstructure-bulk properties relationship is yet to be well understood. Here, we use time-lapse 3D imaging to probe 3D electrode structure and move towards intelligently designing structures with specific performance attributes. Using 3D imaging allows for characterisation and quantification of parameters that are not well described in 2D, such as triple-phase boundary lengths, tortuosity and phase percolation. In this work, the parameters for an optimised porous nickel-scandia stabilized zirconia anode (Ni-ScSZ) were calculated using an electrochemical model to maximise triple-phase boundary length and density, gas permeation and electronic and ionic transport. Ni-ScSZ scaffolds were then produced using screen-printing and impregnation and characterized with FIB-SEM to obtain 3D microstructural parameters such as actual percolation of the phases, tortuosity and connectivity of the phases, and the length and density of triple-phase boundaries (Figure 1). The experimental microstructural parameters were then compared to the modelled optimised microstructure, allowing for an examination of the accuracy of the manufacturing parameters. Symmetrical cells were set up to compare the anode performance with the performance predicted by the electrochemical model. Early iterations of impregnated scaffolds show triple-phase boundary densities increased by approximately an order of magnitude and allowed for a critical assessment of the electrochemical model. The degraded electrode morphology was also characterized in 3D to link its changes during cycling to performance degradation. This approach enables us to move towards designing optimised electrodes with specific performance attributes. Figure 1