(277j) Integration of Multi-Scale Multi-Phenomena Simulations of Direct Methanol Fuel Cell Via Lattice Boltzmann Methods

Authors: 
Chung, P. S., Carnegie Mellon University
Chen, H., Carnegie Mellon University
Jhon, M. S., Carnegie Mellon University


The chemical reaction rate on the solid surface is significantly influenced by the transport of species near the surface. One of the compelling examples is the reaction occurring at or near the liquid/solid interface on the surface of the anode catalyst of a proton exchange membrane fuel cell (PEMFC) and direct methanol fuel cell (DMFC), which has been a favorite and promising candidate for the new energy sources due to several advantages: low environmental impact, high electrical conversion efficiency (up to 55%) independent of size, production of heat usable for co-generation cycles, and flexibility of the fuel used. DMFC are very attractive compared to the other fuel cells primarily because of their usability at ambient temperatures [1]. During the operation of a DMFC, the methanol (CH3OH)-water solution comes in contact with the catalyst at the anode side undergoing electro-catalytic oxidation owing to the bifunctional catalytic activity of the catalyst produces gaseous CO2 and protons. CO2 is released back from the anode catalyst to the fluids at the anode in the form of bubbles through the porous media. The diffusion of gaseous products, the gas nucleation on the solid surface, the migration of gas molecules into nanosized bubbles, and the evolution of the bubbles show complex physical phenomena, which have a drastic influence on the reaction rate. Here, by using lattice Boltzmann method (LBM), we studied the nanoscale transport phenomena in the sub-systems of the anode (i.e. nanotube catalyst, porous media, and gas diffusion layer), which will provide a foundation for the successful design of the nanostructure of the solid surface and the optimal operation of the reactive system. Experimentally, if the concentration of CO2 gas molecules near the solid surface increases in the absence of a pre-existing nucleation site, the supersaturated CO2 gas may initiate nucleation on the surface. When the gas bubbles grow, they migrate within the nanotube array. If there are empty zones, the bubbles will have a tendency to fill these zones and move out of the array to the porous media due to the net surface tension force resulting from different interfacial curvatures. In order to establish the numerical model for anode catalyst, we have focused on the mechanism of CO2 bubble nucleation, their agglomeration, detachment from the catalyst surface, and transport under high deformation due to the confinement in the nanotube catalyst via LBM. By using discretized Boltzmann equation, this method is one of the promising numerical schemes for simulating multi-scale fluid flows and complex physics with complex geometry. We have developed novel applications and fundamentals in rule-based mathematics and physics via LBM to model nanoscale transport processes as well as integrated modeling for the design of nano/information technology systems [2]. Especially, LBM can be implemented to different length and time scale of the system by using different governing equations. In order to track the vapor-liquid interface in the nanotube catalyst, we introduced mesoscale transport equation. After CO2 bubbles move out of the catalyst, CO2 gases diffusion through the porous media to the convective fuel channel. We have developed novel methodology for the complex boundary conditions by simulating fluid flow between the head-disk-interface on the hard disk drive system. Based on the technique for the complex boundary condition and the observed transport phenomena in the anode catalyst, we are further simulating CO2 bubble transport to the gas diffusion layer through the porous media. In order to simulate different transport phenomena from the nanotube catalyst, we applied different governing equations for each system, i.e. Darcy-type porous equation for the porous media, and standard continuum multi-phase flow for the convective fuel channel. Since the complex moving geometry of liquid/vapor phase, we are currently developing our boundary conditions.

[References] 1. C. K. Witham, W. Chun, T. I. Valdez, and S. R. Narayanan, Electrochemical and Solid-State Letters 3(11), 497 (2000). 2. S. S. Ghai, W. T. Kim, R. A. Escobar, C. H. Amon, and M. S. Jhon, J. Appl. Phys. 97, 10P703 (2005)