An experimental and computational study of solid oxide fuel cell electrodes

Composite nano-LSCF-GDC cathodes were sythesized using an infiltration method achieving polarization resistances as low as 0.2 Ocm 2 at 600°C. Instead of firing both the LSCF and GDC phase in one step, as is done conventionally, the processing of each phase was separate, allowing for individual control to optimize the microstructure of each phase. It was found that firing the GDC scaffold at 1100°C followed by infiltrating LSCF and firing at 800°C yielded optimal performance. The resulting microstructure was a sub-micron scale GDC matrix covered with a nano-scaled LSCF particle network. It was also found that up to 12 vol% LSCF (the highest loading tested), the performance of the cathodes continually improved. A simple model relating the performance of the cathodes to the surface area of the LSCF, electrocatalyst phase showed good agreement with experimental results without the use of fitting parameters at 600°C.;Based on the knowledge that the surface area of the LSCF directly impacts the performance, coarsening was then assumed to be the primary degradation mechanism of the cathode. An accelerated testing method was developed, where the cathodes were aged at high temperatures (650--850°C), and impedance measurements were taken at 600°C. This data was fit to a power-law model for coarsening of a 3D particle on a 2D surface. The fit was very good indicating that coarsening was in fact responsible for the decrease in performance. SEM images in the same location after various aging treatments also showed a loss of surface area in the LSCF phase. An activation energy range was found from fitting the power-law model, and predicted degradation over 40,000h was between 39--74%. Finally, a dual infiltration of both LSCF and GDC into a GDC matrix showed decreased degradation, likely because the GDC particles were acting to pin the LSCF and slow the coarsening.;Additionally, a phase field model was developed to describe the growth and coarsening of electrocatalyst particles. Both bulk and surface diffusion were accounted for in this model, and a method for relating the mobility at the surface to the surface diffusion coefficient was developed. Since the diffusion coefficients of this novel system remain unknown, a parametric study of diffusion coefficients was done revealing how various ratios of bulk and surface diffusion in the substrate and particle affect the microstructural evolution of the anode.