Anion Exchange Membranes with Tuned Ionic Conductivity

The potential of anion exchange membrane (AEM) fuel cells to provide inexpensive compact power from a wider variety of fuels than is possible with a proton exchange membrane (PEM) fuel cell, has continued to drive the research interest in this area.  Alkaline catalysis in fuel cells has been demonstrated with non-precious metal catalysts, and with a variety of fuels beyond H₂and methanol. Alkaline fuel cells (AFCs), based on aqueous solutions of KOH, have serious drawbacks associated with system complexity and carbonate formation. Anion exchange membrane (AEMs) fuel cells have a number of advantages over both PEM fuel cells and traditional AFCs; however, ionic conductivity in AEMs is still lower than PEMs and chemical stability of membrane attached cations in hydroxide is still not sufficient for practical applications.

Our goal is to synthesize an AEM with excellent anion transport properties that can be fabricated into a thin robust film suitable for fuel cell applications.  To do this with have started with polymer architectures based on homo- and co-polymers of vinylbenzyltrimethylammonium, PVBTMA, as these are readily synthesized as organized or random diblock polymers, which we can fully characterize, use for the study of anionic transport and build in-silico so that predictive modeling can be used to design next generation materials. 

Block copolymers that can self-assemble into well-defined hydrophobic and hydrophilic domains will better be able to achieve the stringent requirements for excellent AEM fuel cells. Membranes made from block copolymers will provide well-oriented and continuous conductive hydrophilic channels to enhance ion conductivity. Because of the presence of the hydrophobic domain in the membranes, the mechanical property of the membranes can also be enhanced. Therefore, high IEC can be achieved leading to higher conductivity. In contrast to random copolymers, it is difficult to achieve high IEC because of the associated swelling encountered at high states of hydration resulted in disintegration of the membranes. By using block copolymers containing polycation as AEMs, the relationship between structure and ionic conductivity of the membranes can be investigated as well. Several studies about structure-morphology-property relationships of block copolymers for PEM have shown that the morphology of the conductive membranes strongly influences their proton conductivity on the aspect of type and orientation of structure.

We studied water partitioning in PVBTMA diblock copolymers with alternating hydrophobic lamellae (for structural support) and hydrophilic lamellae (to conduct ions). In order to be conductive, the hydrophilic block must absorb water. The water is often thought of as uniformly permeating the hydrophilic lamella, but that need not be so. Since the hydrophilic block is effectively tethered to the hydrophilic-hydrophobic interface, the polymers must stretch on average to accommodate the water. For typical well-solvated grafted layers the optimal stretching profile leads to strongly non-uniform swelling with solvent. In a simple example the polymer concentration decreases quadratically with the distance z from the tethering surface and reaches zero at a specific height h set by the solvent quality and the elastic properties of the polymer coils. Likewise for weaker solvation, there will be more polymer near the hydrophilic-hydrophobic interface, and more water at the midplane. This nonuniformity takes a special form for marginal solvation, where two concentrations exist in equilibrium. Here the tethering constraint favors the polymer-rich phase to reside near the tethering surface, while the dilute phase occupies a “water channel" near the midplane. The water configuration has potential implications for the ion conductivity of the membrane. The polymer-rich region is packed with ions (good for conductivity), but the small amount of water there inhibits ion mobility (bad for conductivity).  In the water channel, this tradeoff is reversed.  The consequences of this for ionic conductivity will be discussed.