Two phase Monte Carlo modeling of gas and water transport in PEM fuel cells

Abstract

Low temperature polymer electrolyte membrane fuel cells are known as clean energy converters, which produce electrical energy as the result of the cell electrochemical reaction between hydrogen and oxygen, where water and heat are the only byproducts. Although an adequate hydration is required for the ionic conductivity of the polymer membrane of the cell, the excess water in liquid phase can obstruct the paths of the reactant gases, which diffuse from the gas flow channels through the gas diffusion layer (GDL) to the reaction sites within the catalyst layer (CL). This problem, which can result in a decreased cell performance, is more severe at high current densities where more water is produced. Therefore, a comprehensive understanding about the effect of all the possible structural and operating parameters on the quantity and distribution of liquid water within the cell is essential for a proper water management. The focus of this study is on water distribution within the GDL and CL structures, taking into account the surface wetting properties, and thermodynamic and operating boundary conditions. This dissertation includes two major parts: 1) steady-state water distribution (with no time axis), and 2) the evolution of water distribution with time. In the first part, a grand canonical Monte Carlo model (GCMC), which had been previously developed by Dr. Katrin Seidenberger to study the steady-state water distribution in GDL structures, is further developed for simulations in CLs, as well as their interface with the adjacent microporous layers. The GCMC simulation results represent a qualitative 3D water distribution in the structure, together with 2D and 1D pore filling degree analysis. However, in order to extract the information on the pore sizes, a complementary pore analysis method is developed. This method is used to investigate the pore size distribution and the relation between the pore filling degree and pore size in the structure. The suggestion of this study for increasing the performance of the PEMFCs is to decrease the gradient of the wetting properties at the structure interfaces. This can facilitate the water transport from the CL to the flow channels by reducing the surface energy barrier to water transport paths. In the second part of this work a new multi-timescale kinetic Monte Carlo (KMC) model is developed to study the evolution of water distribution within the GDL structures with time. In this model the effect of different current densities has been studied, and since water production and movement take place on - by orders of magnitude - two different timescales, a multi-timescaling strategy is pursued to include both processes in the simulations. The KMC simulation results for all current densities show an increase in the amount of water near the CL, and as opposed to that a decreasing behavior in pore filling degrees in the regions near the flow channel with time. However, increasing the current density results in an increased overall water quantity in the GDL structure. The main results of this PhD thesis are published in three papers.