Designing the Ideal Catalyst Layer in PEMFCs

Abstract

Many design parameters affect the performance of PEMFCs, including catalyst density, porosity, nafion distribution, agglomerate size, etc. The ideal values of these parameters vary not only from layer to layer (for instance from catalyst layer to gas diffusion layer or from anode catalyst layer to cathode catalyst layer) but also from location to location inside the same layer, resulting in non-uniform 3-dimensional spatial distributions of the platinum, nafion, and pore and agglomerate sizes. Furthermore, the optimal values of these parameters also depends on the operating conditions, such as electrical current density, pressure, and temperature, which increases the complexity of the problem even more. To design the ideal catalyst layer, the designer must determine the optimum distributions of the above parameters within each layer. In this work, we focus on estimating the ideal structure of the catalyst layer in PEMFCs, in particular on determining the optimum spatial distributions of the platinum, porosity, nafion, and agglomerate size.To solve the above problem, we apply the adjoint method to compute the optimum parameters for the microstructure of the catalyst layer. We start by computing the sensitivity functions (which are related to the Gâteaux derivatives) of the cell voltage or current in the direction of the design parameter variation. The sensitivity functions are then applied to adjust the parameters of the entire catalyst layer in a point-by-point fashion. The process is repeated until convergence is obtained.Numerical results will be presented showing the non-optimized and optimized distributions at different values of the operating current density, temperature, and pressure for a 25 μm membrane PEMFC, as well as the polarization curves for non-optimized and optimized configurations.