(379f) Modeling of Chemically-Specific Separations in Lyotropic Liquid Crystal Self-Assembled Nanoporous Membranes

The nanopores that form in cross-linked lyotropic liquid crystal (LLC) membranes are chemically complex and uniform in size which makes them promising for highly selective separations. We have developed an atomistic molecular model of an LLC membrane that is maximally consistent with experimental X-ray diffraction for the purpose of studying small molecule transport within it.

We have characterized transport of water, sodium ions and small polar solutes within the pores of our model. We find that the transport rate of a species is dependent not only on molecular size, but on chemical functionality as well, leading to anomalous diffusive behavior. Electrostatic interactions between the membrane and solutes provide an interesting diversity of mechanistic behaviors. In general, all solutes perform intermittent hops between lengthy periods of entrapment, and a range of different trapping mechanisms are responsible for this behavior. The particular frequency of each mechanism depends to large extent on the identity of the chemical species, meaning that such membranes have significant capability for separations.

We use our detailed mechanistic understanding to construct a range of stochastic models rooted in anomalous diffusion theory, which we can use to project out the long term behavior of the solutes and ultimately predict selectivity. Using the insights developed in this study, we can begin to think about how to redesign existing LLC membranes in order to perform solute-specific separations.