(Invited) Structural Characterization of Polymer Electrolyte Fuel Cell Electrodes Using Nano-Scale Synchrotron X-Ray Computed Tomography

Abstract

With improvements in the intrinsic oxygen reduction reaction activity of both Pt-based and platinum group metal-free (PGM-free) catalysts for polymer electrolyte fuel cells, in recent years more attention is being focused on utilization of the active sites in these catalysts under conditions relevant to the automotive application. Taking full advantage of the performance offered by these high activity catalysts requires comprehensive knowledge of the electrode microstructure and the spatial distribution of the catalyst, support, ionomer, and pores. Nano-scale resolution X-ray computed tomography (nano-CT) has been used as a powerful tool for non-destructive quantification of three-dimensional pore morphology and estimation of ionomer distribution in both Pt-based and PGM-free cathode catalyst layers (CCL) [1, 2]. On the other hand, due to insufficient resolution (20 nm voxel size) this technique cannot resolve the detailed structures of support, ionomer, and catalyst. We developed the “hybrid computational method” [2] combining statistical information from transmission electron microscopy (TEM), ultra-small angle X-ray scattering (USAXS), and BET with nano-CT to reconstruct the features below the nano-CT resolution. In this study, an example will be given of application of this method to a Pt alloy-based electrode. Various shape, size and composition agglomerates are reconstructed to 1 nm voxel resolution using the hybrid methodology. It is shown that ionomer film thickness is spatially non-uniform inside the agglomerates and strongly correlates with ionomer to carbon weight ratio (I/C) while it is independent of the agglomerate size. Compared to experimental observations, the application of capillary condensation theory to the reconstructed agglomerate structure is shown to accurately represent the relative humidity (RH) dependence of the electrochemically-active surface area (ECA). Furthermore, direct numerical simulations are performed to quantify localized transport losses. Effectiveness factor correlations are derived for use in macroscale models. The effects of RH, agglomerate structural properties, and pores in the carbon support on the local oxygen transport resistance are examined. It is demonstrated that low catalyst utilization is responsible for the experimentally-observed high local transport resistance under dry conditions and that agglomerate shape and size affect the local oxygen transport resistance only if the primary pores are poorly accessible.