Despite the gains that have been made in fuel cell technology over the last decade, major barriers to implementation of commercial fuel cells remain. One of the key shortcomings of contemporary fuel cells is the proton exchange membrane (PEM), which serves as the electrolyte for proton transfer and as the separator to prevent direct physical mixing of the hydrogen and the oxygen at the anode and cathode, respectively. The state-of-the-art polymer electrolytes are perfluorosulfonate ionomers. The next generation of fuel cells being designed for transportation require improvements in several aspects of the current membranes, including their high cost, insufficient durability, especially at higher temperatures, poor resistance to fuel crossover, unacceptable water transport rates and inadequate properties above 100°C. A higher temperature PEM that is less sensitive to water can improve the efficiency of a fuel cell by speeding up the reaction at the anode, and higher operating temperature also improves the CO tolerance and reduces the need for pressurization of the fuel cell. Developmental work on new membrane materials involves mostly the synthesis of new sulfonate ionomers based on aromatic hydrocarbon polymer backbones. These materials, however, are likely to have similar nanostructures as the perfluorosulfonate ionomers, i.e., water-swollen ionic channels that form due to aggregation of water and the sulfonic acid groups and similar deficiencies as the perfluorosulfonate membranes with regard to water management and fuel crossover. A relatively high concentration of water is required to develop a percolation pathway of water-swollen “ionic” channels, and water concentration and ion transport are coupled in water-swollen perfluorosulfonate membranes. As a consequence, the transport and mechanical properties of the membranes are coupled; that is, improving the conductivity by increasing ion-exchange capacity (IEC) or water concentration degrade the mechanical properties and vice versa.

Despite the considerable research on the synthesis and characterization of new membrane materials, surprisingly little work has been directed at controlling the morphology of PEMs. This is surprising, because it is the ionic microstructure that controls the conductivity and the transport of water and methanol through the membrane. In this paper, we will discuss the properties of PEMs produced from polymer blends and how the morphology of the blend affects transport properties. The blends were prepared from a conductive polymer (i.e., an ionomer) and either a non-conducting polymer or a poorly conducting polymer. With a two-phase blend, one can decouple the mechanical and transport properties and, in principle, optimize both independently. In our case, the non-conducting phase provides mechanical stability and durability to the PEM, while the ionomer provides the ionic pathway for conductivity. We will also discuss the electrical field alignment of the conductive phase in a polymer blend can significantly improve conductivity.

Blend membranes were made comprising sulfonated poly(ether ketone ketone), SPEKK, with polyetherimide (PEI), polyethersulfone (PES), another SPEKK, and sulfonated crosslinked polystyrene particles (SXLPS). PEKK and PEI are miscible. Below an SPEKK sulfonation level (IEC) of 0.8 meq/g, SPEKK and PEI remained miscible; for SPEKK with an IEC > 0.8, phase separation occurred during film casting from either NMP or DMAc solutions. Blends of SPEEK with an IEC = 2 meq/g (SPEKK2.0), with PEI as the minority component exhibited a morphology that consisted of ~0.5-0.6 µm PEI domains dispersed in the ionomer matrix. The addition of PEI to the SPEKK improved the mechanical integrity of the membranes, though it also lowered the conductivity. Even so, above a sulfonation level of 1.9 meq/g, as much as 30% PEI could be incorporated into the membrane with less than an order of magnitude reduction in conductivity. The addition of PEI also significantly lowered the water absorbed by the membrane. Unlike the SPEKK/PEI blends, solution cast blends of SPEKK and PES were homogeneous over the entire range of sulfonation and blend concentrations studied. Because of the improved intermolecular interactions between the two polymers in the SPEKK/PES blends, higher sulfonation levels, even above that where SPEKK is water soluble, could be used to prepare viable membranes. The homogenous SPEKK/PES blends provided significant advantages with regard to membrane swelling. For example, the addition of 15 wt% PES to SPEKK1.7 (IEC = 1.7 meq/g) decreased the water concentration at 90°C from 80 H2O/SO3H for the neat SPEKK to below 20 H2O/SO3H for the blend. More remarkable, water-soluble SPEKKs (IEC > 2.4 meq/g) were rendered insoluble by the addition of 50% PES. While the addition of PES to SPEKK lowers the proton conductivity of the blend, that can be compensated by increasing the IEC of the SPEKK (e.g., to 3.5 meq/g). Thus, the design of the SPEKK/PES membranes offers several degrees of freedom, including the SPEKK IEC and the blend composition that may make possible a membrane with high conductivity ( 10-2 – 10-1 S/cm), yet relatively low water swelling (< 20 H2O/SO3H) and good mechanical properties. Low IEC SPEKK (< 1.5 meq/g) produced fairly robust membranes, due to low swelling of those membranes by water. However, the proton conductivity was low. In the two blends described above, a non-ionic polymer was used to improve the mechanical properties of high IEC SPEKKs. Although that approach was successful at improving the mechanical integrity of the PEM, it also lowered the proton conductivity. In a blend of a high IEC SPEKK with a low IEC SPEKK, we expected addition of the lower IEC SPEKK to have a less detrimental effect on the conductivity and fuel cell performance (because it is still a conductive ionomer), and because of its lower water sorption, to also improve the mechanical properties of the membrane. In addition, the two component polymers are likely to be more compatible than the SPEKK with PEI or PES, since the sulfonic acid groups in both polymers can hydrogen bond intermolecularly. The SPEKK/SPEKK blends were generally two-phased, but they were more resistant to swelling than the neat SPEKKs. If two high IEC SPEKKs were used, e.g., SPEKK2.0 and SPEKK1.5, the blends were homogenous. The morphology of the solution cast membrane was controlled by varying the difference in IEC of the two SPEKKS, the composition of the ionomers and the casting solvent and temperature. The best compromise in transport and swelling resistance (and mechanical properties) was obtained when a two-phase, co-continuous microstructure was developed. Polarization curves indicated that at lower temperatures and high humidity (e.g., 80°C and 100% RH), the SPEKK/SPEKK membranes were superior to neat SPEKK membranes in performance and structural integrity. However, at high temperature (> 100°C) and low humidity operation (< 35% RH), the performance significantly suffered, due to the low water content of the membrane.

We have also used an electric field to orient the conducting phase in a two-phase polymer blend PEM. This significantly enhances the transport properties of the PEM. For example, alignment of the dispersed droplet SPEKK phase in an SPEKK/PEI membrane improved the conductivity by as much as two orders of magnitude. In this talk, we will discuss in more detail this approach to the development of PEMs.