(588e) A Coarse-Grained Model of Polyphenylene Oxide – Trimethylamine Membrane for Alkaline Fuel Cells

Developing alkaline anion-exchange membrane fuel cells (AAEMFCs) remains challenging in part from difficulty synthesizing anion-exchange membranes with high OH^- ion conductivity, acceptable mechanical stabilities, and low chemical deterioration in high pH media for fuel cell operation. Molecular simulations provide a versatile tool to study the anion conductivity and stability of AEM materials in an effort to provide a fundamental understanding of the AEM design. However, available molecular models for membranes do not meet the quest of large spatial and temporal scales required to model the multiscale structure and transport processes in the polymer electrolyte membranes. Here, we present the development of coarse-grained models for hydrated polyphenylene oxide - trimethylamine (PPO-TMA) membranes. To our knowledge, this is the first coarse-grained model that includes water, ions, hydrophobic, and intramolecular interactions, all explicitly parameterized to reproduce multiple properties of interest for AEM, including ionic solvation and water-driven hydrophobic association. The uncertainty quantification (UQ) method is used in an iterative approach to parameterize the high-dimensional parameter space of the force field to reproduce multiple properties of the atomistic reference systems. Because of the reduced degrees of freedom in a coarse grained model, the structural, energetic, and dynamic properties cannot be simultaneously reproduced with arbitrary precision; however, the interdependence of these properties are considered collectively. The coarse-grained membrane model gives a reasonable description of the mobility of water and ions, and consequently, the solvation and the electro-osmotic drag, which are of utmost relevance for the operation of fuel cell membranes. We use the coarse-grained membrane model to explore how varying the alkyl amine pendant groups attached to the PPO backbone affects the morphology and dimensions of segregation in the membrane, and compare the simulation results with experiments. We anticipate that the large spatial and temporal simulations made possible by the coarse-grained model will advance the quest for anion-exchange membranes with improved transport and mechanical properties.