Towards a Membrane Electrode Assembly for a Thermally Regenerative Fuel Cell

Loading...
Thumbnail Image
Date
2013-04-15
Authors
Skerritt, Mark
Keyword
Nafion , PEMFC , mPBI , Polymer , Electrochemistry , Fuel Cell , Polybenzimidazole
Abstract
The thermally regenerative fuel cell (TRFC) concept that is analyzed is a polymer electrolyte membrane fuel cell (PEMFC), powered by the electro-oxidation of H2 and the electro-reduction of propiophenone. The main products of this fuel cell should be 1-phenyl-1-propanol and electricity. The 1-phenyl-1-propanol should then be converted back to propiophenone, while hydrogen is regenerated by using waste heat and a metal catalyst (Pd/SiO2). The first objective was to find a compatible polymer that would work as either an ionomer/binding agent and as a membrane in the membrane electrode assembly (MEA) of the TRFC. This was achieved by checking the compatibility of each polymer with 1-phenyl-1-propanol and propiophenone (the alcohol-ketone pair). Catalyst coated gas diffusion layers or catalyst coated membranes were made to test the stability of the polymers in the catalyst bed when exposed to the alcohol-ketone pair. If the polymer was compatible with the alcohol-ketone pair, MEAs were constructed using this polymer. The second objective was to test these MEAs inside a H2/propiophenone fuel cell that would prove the concept of our envisioned TRFC. It was found that the only polymer that was stable in the alcohol-ketone pair was mPBI (m-phenylene polybenzimidazole). The mPBI had to be doped with H3PO4 to enable H+ conductivity. Unfortunately, some H3PO4 leached out of the H3PO4-doped mPBI when in the presence of the alcohol-ketone pair. MEAs that were created using H3PO4-doped mPBI were found to work for H2/air and H2/propiophenone fed PEMFCs. The best performance achieved with the H2/propiophenone powered fuel cell was 6.23 μW/cm2. Unfortunately, the presence of the 1-phenyl-1-propanol product could not be proved by EIS or CV on the fuel cell, or by GC-FID of the cathode effluent. Other unknown products were seen in the GC-FID spectrum of the cathode effluent. Therefore, it is possible that the propiophenone did reduce at the cathode but it produced an unknown product. In conclusion, the viability of the proposed TRFC system was not verified. H3PO4 leaching from the MEA makes it impossible to use H3PO4-doped mPBI as the electrolyte in the final version of the MEA in the TRFC system.
External DOI