Reports: AC5

45666-AC5 Proton Transfer in Phosphoric Acid Based Fuel Cell Membranes

Joel R. Fried, University of Cincinnati

The ultimate goal of this project is to investigate the mechanism of proton transfer (PT) in a proton-exchange membrane for fuel cell use by application of a hybrid simulation method (ONIOM) using Atom-centered Density Matrix Propagation (ADMP) ab initio molecular dynamics and molecular mechanics. The specific focus of this work is phosphoric acid (PA)-doped polybenzimidazole (PBI). During this year, the ZPE-corrected proton affinities of PA, water, and several molecular species proposed to be formed during the PT events were calculated using B3LYP/6-31++G(d,p). The proton affinity of water at this level of theory was found to be 684 kJ/mol in good comparison with the experimental value of 691 kJ/mol. For PA, the value was 822 kJ/mol and 2404 kJ/mol for PO43- as an example of one anionic species that could form during proton transfer. In addition, detailed potential energy surface scans have been used to investigate the energy barriers associated with different PT pathways for PA and PA/water systems. Additional ADMP calculations were made using this DFT functional and basis set at 298 K. Detailed molecular dynamics (MD) studies of PA have also been finished including determination of the diffusion coefficient of PA over a range of temperatures using the COMPASS force field. As shown by the graft at the top of this text, the self-diffusion coefficient and the slope of the Arrhenius plot (i.e., the apparent activation energy for diffusion, 23.8 kJ/mol from our MD results) show good agreement with published PFG-NMR results (apparent activation energy of 29.8 kcal/mol). Radial distribution function (RDF) plots calculated from the MD results show evidence for hydrogen bonding in PA. Results of these background studies, have been written for publication and a second paper is being finished based upon our MD studies of polybenzimidazoles. In this coming final year, dynamic trajectories of a system consisting of several small PBI chains and PA molecules (with and without water molecules) will be analyzed to determine PT pathways using the hybrid ONION method employing ADMP (B3LYP/6-31++G(d,p)) for ab initio MD and the DREIDING force field for molecular mechanics. DREIDING will be validated against our COMPASS results for PA and PBI by comparison of RDF plots.