Expanded 3D Electrode Architecture for Low Temperature Direct Liquid Fuel Cells

AbstractThe US Navy continues to pursue electrochemical power sources with high energy density for low rate, long endurance undersea applications. The direct electro-oxidation and electro-reduction of sodium borohydride and hydrogen peroxide are being investigated to meet these goals. In an effort to minimize polarization losses and increase power density, a novel carbon microfiber array (CMA) electrode is being investigated.The CMA is composed of 750 micron long, 10 micron diameter graphite fibers that protrude from a current collector like blades of grass. The CMA was developed for the direct reaction of peroxide in the Mg-H2O2 semi fuel cell. [1] There, the high surface area of the microfiber cathode reduces peroxide concentration polarization, resulting in increased power and energy density. For this work the CMA architecture was adapted into a novel membrane electrode assembly and evaluated in the direct BH4- / H2O2 fuel cell. The unique feature of this architecture vs. traditional membrane electrode assemblies (MEAs) is how all three components of the triple boundary interface are optimized: electrical connectivity, ionic connectivity and mass transport. The current iteration of this electrode architecture utilizes a carbon cloth that has been hot pressed into N115 membrane. This component is then placed over the CMFA electrode. The carbon microfibers of the CMFA protrude up into the carbon cloth matrix forming a 3-dimewnsional, interdigitated electrode architecture. The result of this modification is improved electrolyte flow through the CMFA and improved utilization of the surface area afforded by the carbon microfibers that was not observed in the non modified CMFA. Half cell polarization measurements were obtained simultaneously with the fuel cell polarization. Initial results using this modified CMFA electrode architecture show that the polarization losses observed for both the reduction of hydrogen peroxide and for the oxidation of borohydride were 5.2 times lower than for the non-modified CMAs electrode (0.014 ohms vs. 0.074 ohms). Comparing these results to those calculated from the literature [2, 3], where traditional membrane electrode assemblies were used for borohydride oxidation, 5 and 2.6 time improvements were obtained (0.07 ohms and 0.037 ohms were the effective resistive losses seen in the anode half cell polarizations).