(29i) Simulation-Driven Design of Chemoresponsive Liquid Crystal Sensors By Tuning Analyte Partitioning

Chemoresponsive liquid crystal (LC) sensors are promising platforms for the detection of vapor-phase analytes. These sensors consist of a LC thin film deposited on a reactive substrate. Upon exposure to an analyte, changes to the anchoring of the LC film to the substrate triggers a change to LC order which is observable optically. Experiments have determined that the activation time of the sensor depends on the transport of analytes within the LC films. Understanding and predicting these transport properties can guide the design of LC sensors with improved selectivity.

In this work, we use atomistic molecular dynamics simulations to quantify the partitioning and diffusion of nine small-molecule analytes, including four common atmospheric pollutants, in model systems representative of LC sensors. We first parameterize all-atom models for 4-cyano-4′-pentylbiphenyl (5CB), a mesogen typically used for LC sensors, and all analytes. We validate these models by reproducing experimentally determined 5CB structural parameters, 5CB diffusivity, and analyte Henry's law constants in 5CB. Using the all-atom models, we calculate analyte solvation free energies and diffusivities in bulk 5CB. These simulation-derived quantities are then used to parameterize a mass-transport model to predict sensor activation times. These results demonstrate that differences in analyte–LC interactions can translate into distinct activation times to distinguish activation by different analytes. Finally, we quantify the effect of LC composition by calculating analyte solvation free energies in TL205, a proprietary LC mixture. These calculations indicate that varying the LC composition can modulate activation times to further improve sensor selectivity. These results thus provide a computational framework for guiding LC sensor design by using molecular simulations to predict analyte transport as a function of LC composition.