Multiscale Simulation of Proton Transport in the Catalyst Layer with Consideration of Ionomer Thickness Distribution

摘要

To spread polymer electrolyte fuel cells (PEFCs) widely, which have many advantages, such as low environment load and high energy conversion efficiency, design and optimization of cathode catalyst layers (CLs) are necessary. CLs have inhomogeneous microstructures where Pt catalysts covered with ionomer thin films are supported by carbon particles. Electrons and protons transport to Pt surface through carbon particles and ionomer thin films while oxygen molecules diffuse in void and permeate through ionomer thin film to reach the Pt surface. Therefore, analyzing and understanding of mass transport of proton and oxygen in CLs are important for improving fuel cell performance. However, since an ionomer thin film on the Pt surface has a thickness of about 4 to 10 nm, detail understanding of the relationship between the transport mechanism within the nanostructures of ionomers and the fuel cell performance remains an issue. Therefore, we have focused on the analysis of transport phenomena using mesoscale numerical simulations by considering the effects of nanostructures on mass transport obtained from molecular dynamics (MD) simulations. Kobayashi et al. have studied the thickness dependence of ionomer thin films on proton transport using MD simulations and introduced these effects into the mesoscale simulations. The authors found that the self-diffusion coefficients of proton (D_H~+) show a peak at the ionomer thickness of 7 nm, in which the value of D_H~+ is almost two times larger than that in the bulk membrane, leading to the high proton conductivity that is 1.6 times larger than the bulk membrane. These results suggest that the influence of ionomer thickness at the nanometer scale on the performance of PEFCs is significant. However, in the mesoscale simulations, the resulting cell performance (i.e., I-V curve) show similar results between the systems with and without consideration of the thickness effects because the thickness distributions obtained from experiments was not sufficiently reproduced using the hydrophilic adhesion model for ionomer distribution in the mesoscale simulation. In this study, we develop a new method to improve the non-uniform ionomer distribution that reproduces the experimental thickness distribution. Furthermore, we take into account the oxygen transport results obtained from MD simulation in the mesoscale simulations to understand the effects of ionomer nanostructures on the oxygen transport at larger scale in CLs and the fuel cell performance.