Microstructure-property relationships in a gas diffusion layer (GDL) for Polymer Electrolyte Fuel Cells, Part I: effect of compression and anisotropy of dry GDL

Highlights: • Methods are developed to predict transport properties of dry GDL in PE Fuel Cells. • Diffusivity and Permeability are reliably predicted based on 3D characteristics. • Predictions based on 3D microstructure match well with numerical simulations. • Anisotropy is due to in- and through-plane variation of tortuosity and hydraulic rad. • The methods can be used to predict relative permeability and diffusivity in wet GDL. - Abstract: New quantitative relationships are established between effective properties (gas diffusivity, permeability and electrical conductivity) for a dry GDL (25 BA) from SGL Carbon with the corresponding microstructure characteristics from 3D analysis. These microstructure characteristics include phase volume fractions, geodesic tortuosity, constrictivity and hydraulic radius. The latter two parameters include information from two different size distribution curves for bulges (continuous PSD) and for bottlenecks (MIP-PSD). X-ray tomographic microscopy is performed for GDL at different compression levels and the micro-macro-relationships are then established for the in-plane and through-plane directions. The predicted properties based on these relationships are compared with numerical transport simulations, which give very similar results and which can be summarized as follows: Gas diffusivity is higher in the in-plane than in the through-plane direction. Its variation with compression is mainly related to changes of porosity and geodesic tortuosity. Permeability is dominated by variations in hydraulic radius. Through-plane permeability is slightly higher than in-plane. Anisotropy of electrical conductivity is controlled by tortuosity, which is higher for the through-plane direction. A table with new quantitative relationships is provided, which are considered to be more accurate and precise than older descriptions (e.g. Carman-Kozeny, Bruggeman), because they are based on detailed topological information from 3D analysis. Furthermore, when using these relationships as input for macro-homogenous modeling, this enables to simulate microstructure effects of real GDL (SGL 25 BA) more accurately. In future, the same methodology can be used to study micro-macro relationships in wet GDL and to predict relative liquid permeability and relative gas diffusivity.