A Novel Method to Indirectly Measure Electro-osmotic Drag and Back Diffusion From Total Water Flow Experiments in PEM Fuel Cells

The objective of the research is to quantify the electro-osmotic drag and back diffusion portions of the experimentally measured total water flow across a proton exchange membrane. Part of the deficiency of prior research calculated individual coefficients by forcing trendlines through the origin that implicitly assumes other drivers to be negligible. If one or more other drivers are present, their impact will be lumped in with the intended coefficient measurement. Also, using linear trendlines assumes that the indirectly measured coefficient does not change with current density, which is likely the case for liquids but not for vapor. For fuel cell membranes like Nafion, the electro-osmotic drag and back diffusion coefficients are dependent on the hydration state of the membrane. To overcome this dependency, a higher-order polynomial data fit is used for the total water flow. A methodology of combining the theoretical model with experimental data sets is proposed here to determine the component coefficients at each data point. Applying this method to prior data reveals that other fluid drivers assumed to be negligible were likely not so. A more accurate understanding of electro-osmotic drag and back diffusion in turn enables more accurate prediction of total water flow from each driver, leading to improving water management and reducing the risks of cathode catalyst layer flooding and membrane dry-out.