Current performance targets for anion exchange membrane (AEM) fuel cells call for greater than 95% alkaline stability for 5000 h at temperatures of up to 120 °C. Using this target temperature of 120 °C, we provide an incisive 1 H nuclear magnetic resonance-based alkaline degradation method to identify the degradation products of n-alkyl spacer tetraalkylammonium cations in various AEM polymers and small molecule analogues. The operative alkaline degradation mechanisms and rates on benzyltrimethylammonium-, n-alkyl interstitial spacer-, and n-alkyl terminal chain-cations are compared in several architectures. Our findings indicate that benzyltrimethylammonium and n-alkyl terminal pendant cations are significantly more labile than an nalkyl interstitial spacer cation. Additionally, we found that the alkaline stability of an n-alkyl interstitial spacer cation is enhanced when it is combined with an n-alkyl terminal pendant. At 120 °C, an inverse trend was observed in the overall stability of AEM poly(styrene) and AEM poly(phenylene oxide) samples compared to what has been shown at 80 °C. Follow-up small molecule studies suggest that at 120 °C, a 1,4-elimination degradation mechanism may be activated on styrenic AEM polymers capable of forming hyperconjugated resonance hybrids.

It has been demonstrated that a micropatterned surface can decrease the resistance of anion exchange membranes (AEMs) and can induce desirable flow properties in devices, such as mixing. Previously, a model that related the resistance of flat and patterned membranes with the same equivalent thickness was proposed, which used the patterned area and thickness ratio of the features to describe the membrane resistance. Here, we explored the validity of the parallel resistance model for a variety of membrane surface designs and area ratios. We demonstrated that the model can predict the resistance of a wide range of patterned AEMs. We showed that the resistance is independent of the spatial ordering of the design by examining random patterns, which is relevant for applications that require, for example, increased turbulent liquid flow in multilayered devices. Some experimental values of resistance obtained for patterned membranes presented deviations from the model. Scanning electron microscopy (SEM) images of the patterned membranes revealed resolution variations and pattern replication errors due to the stereolithographic process. A geometric correction of the target ratios improved the fit of the modelled data to the experimental values, showing that light bleeding during curing was a source of error. Two additional experimental factors were not accounted for in the model: a distinct interface between the bottom and top layer, and overcuring of the bottom layer during successive steps. These sources of error were investigated by examining the resistance of single and double layered membranes, and single layer membranes with different curing times. The differences obtained in the resistances for control samples demonstrated that both the interface and overcuring influence the resistance of the membrane. The results obtained in this study enlighten the discussion relating membrane surface morphology and transport properties, as well as the optimization of 3D printed membranes using a stereolithography process.

Cell membranes control mass, energy, and information flow to and from the cell. In the cell membrane a lipid bilayer serves as the barrier layer, with highly efficient molecular machines, membrane proteins, serving as the transport elements. In this way, highly specialized transport properties are achieved by these composite materials by segregating the matrix function from the transport function using different components. For example, cell membranes containing aquaporin proteins can transport ∼4 billion water molecules per second per aquaporin while rejecting all other molecules including salts, a feat unmatched by any synthetic system, while the impermeable lipid bilayer provides the barrier and matrix properties. True separation of functions between the matrix and the transport elements has been difficult to achieve in conventional solute separation synthetic membranes. In this study, we created membranes with distinct matrix and transport elements through designed coassembly of solvent-stable artificial (peptide-appended pillar[5]arene, PAP5) or natural (gramicidin A) model channels with block copolymers into lamellar multilayered membranes. Self-assembly of a lamellar structure from cross-linkable triblock copolymers was used as a scalable replacement for lipid bilayers, offering better stability and mechanical properties. By coassembly of channel molecules with block copolymers, we were able to synthesize nanofiltration membranes with sharp selectivity profiles as well as uncharged ion exchange membranes exhibiting ion selectivity. The developed method can be used for incorporation of different artificial and biological ion and water channels into synthetic polymer membranes. The strategy reported here could promote the construction of a range of channel-based membranes and sensors with desired properties, such as ion separations, stimuli responsiveness, and high sensitivity.

Redox-responsive anion exchange membranes were developed using photoinitiated free-radical polymerization and reversible oxidation and reduction of viologen. The membranes were formulated using poly­(ethylene glycol diacrylate) and diurethane dimethacrylate oligomers, dipentaerythritol penta-/hexa-acrylate cross-linker, photoinitiators, and 4-vinylbenzyl chloride as precursors for functionalization. In the membrane, 4,4′-bipyridine reacted with the 4-vinylbenzyl chloride residues, and subsequently, unreacted amines were methylated with iodomethane to obtain viologen as both the ion carrier and redox-responsive group. Upon oxidation, viologen supports two cations, where the reduced form only contains one cation. Thus, the redox responsiveness changed the membrane ionicity by a factor of 2. The area-specific resistance of the membranes in the oxidized, +2, state was lower than in the reduced, +1, state. The resistance increased between 40.6 ± 0.1 and 111.6 ± 0.1%, depending on membrane thickness, with the most significant increment being a resistance change from 4.88 × 10–4 Ω m2 in the oxidized state to 1.03 × 10–3 Ω m2 in the reduced state. Membrane permselectivity in the reduced, +1, state was between 15.9 ± 0.1 and 26.5 ± 0.01% lower than in the oxidized, +2, state, with no change in water uptake, spanning an average of 0.87 ± 0.02 in the oxidized state to an average of 0.7 ± 0.01 in the reduced state. Upon reduction, membrane ion-exchange capacity decreases, increasing ionic resistance and decreasing membrane permselectivity due to a reduction in fixed charge concentration without a measurable change in water uptake. This trend is not generally observed for ion-exchange membranes and explains that the changes in transport properties result from changes in ionicity, not water uptake or domain size. The reversibility and stability of the stimuli responsiveness were confirmed by the absence of transport property changes after redox cycling.