The Effect of Electric Double Layers on Ionic Conductivity in the Agglomerates of PEM Fuel Cell Anodes

Abstract

We present a theoretical analysis of ionic transport inside the catalyst particle agglomerates that form the electrodes of proton exchange membrane (PEM) fuel cells. The electrodes continue to be the subject of intense research and development because they are still the largest cost and source of performance degradation in PEM fuel cells. The advancement of electrodes requires proper understanding of the electrode structure and the relevant transport processes. However, the details of the electrode microstructure and the micro-scale and nano-scale transport mechanisms are still not well understood. A common hypothesis, supported by recent coarse-grained molecular dynamics simulations, is that the primary pores (the pores inside the agglomerates) are void space and not filled with Nafion electrolyte. Instead, it has been postulated that the primary pores are saturated with liquid water during operation. Here, we report on the effect of the electric double layers (EDLs), which form at the interface between the water and the carbon catalyst supports, on the ionic transport within the agglomerates. The multi-scale model addresses phenomena at two length scales: (1) the nano-scale EDL thickness and (2) the microscale agglomerate radius. We model the EDL using the Gouy-Chapman-Stern model, which provides a pore average conductivity for the spherical conduction-reaction model of the agglomerate. We use a spherical agglomerate model to calculate an effectiveness factor for the electrochemical reactions. Here we present the application of the model to the anode, where the low activation overpotential allows linearizations and convenient analytical solutions. A key finding of this work is the important role the EDLs have in establishing the effectiveness of the platinum catalyst utilization. In addition, we resolve the dependence of the agglomerate effectiveness factor on the activation overpotential and agglomerate radius. We observe a significant nonmonotonic dependence of the catalyst effectiveness factor on the overpotential and dramatic improvement in effectiveness of catalyst utilization with smaller agglomerates.