Liquid crystals (LCs) are chemically responsive materials that can be used to fabricate inexpensive, wearable sensors. These sensors are comprised of a micrometer-thick LC film supported on a solid, chemically-functionalized surface. When exposed to a gaseous ozone-chlorine mixture, the ozone and chlorine diffuse through the LC film and bind to the chemically-functionalized surface. This binding disrupts the molecular interactions between the surface and the film, causing the LCs to reorient and refract the light traveling through them. Different concentrations of ozone and chlorine result in different optical responses; however, it is not easy to analyze the LC optical response to predict the concentrations of ozone and chlorine. Our group decided to use computational techniques to overcome this obstacle. Specifically, we were given 46 LC optical responses with 16 concentrations combinations (ozone: 1.5, 5, 100, and 650 ppm; chlorine: 0, 1, 2, and 5 ppm). The data contained three vectors corresponding to the temporal evolution of red, green, and blue intensities. We hypothesized that the time-series signals contained some characteristic components capable of distinguishing different mixture concentrations. The analysis methods included statistical machine learning models (e.g., principal component analysis, support vector machines) and different types of neural networks (e.g., convolutional neural networks, recurrent neural networks). It was found that these methods could accurately predict high ozone concentrations, but performance decreased at low concentrations of both compounds.

Analyzing the optical response of liquid crystals to an ozone-chlorine gas mixture

ACS Spring Meeting 2021

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.