Studying the performance of a direct methanol fuel cell (DMFC) is complicated by the complex interactions of kinetic and transport processes. As a result, changes in one aspect of the cell have consequences in other aspects, which are difficult to elucidate from full-cell polarization (i.e., voltage versus current) behavior. This study outlines a strategy to use current and voltage relationships from anode half-cells, cathode half-cells, and hydrogen pump coupled with methanol crossover data and a mathematical model. In this way, all the kinetic and transport processes have been quantified, and the cell voltage was deconstructed (i.e., individual voltage losses were quantified). This data analysis accounts for all of the voltage losses observed during the operation of the full cell. As expected, the anode and cathode overpotentials accounted for most of the losses (i.e., 92% on average). Also, the cathode flow rate has been shown to affect the methanol crossover by diffusion. Cells operated at constant stoichiometry or where the cathode flow rate is small can show a parabolic shape in the methanol crossover because the electroosmotic drag dominates over diffusion as the primary transport mechanism for methanol through the membrane. Decrease in the methanol crossover was observed for cells with high compression and thicker cathode electrodes. The one-dimensional model, developed previously (García et al., 2004, “Mathematical Model of a Direct Methanol Fuel Cell,” J. Fuel Cell Sci. Technol., 1(1), pp. 43–48), was improved by: (1) including methanol transport from the anode flow channel to the backing layer using a mass transfer resistance and (2) accounting for the unreacted methanol transport through the cathode. The model was able to reasonably predict the anode, cathode, full-cell polarization, and methanol crossover data for methanol concentrations between 0.05 M and 2 M at all operating currents.
Read full abstract