Spectroscopy measurements often contain a mixture of pure component signals of unknown shape and concentration. An important measurement challenge is how to deconvolute these pure component signals given a limited set of experiments and how to use this data to improve machine learning models. This project focuses on using signal processing techniques, machine learning, and AI explainability to improve polymer spectroscopy. This work is done with NIST's Polymer Analytics project.
When polymer materials are stretched, they bend, deform, and ultimately break. How do we understand this process and control it to yield better materials? This project explores how glassy structure, crystallinity, and entanglements affect polymer mechanics. Employing computational, theoretical, and data-science-driven techniques, we develop next-generation plastics and compatibilizers. This work is in conjunction with NIST's Polymer Analytics project.
While branched polymers have many technological applications, the structural characterization of these polymers poses experimental challenges. Though many theories assume a single, well-defined structure, synthesizing these macromolecules produces a wide distribution of architectures. Using in silico methods can solve these issues by allowing precise structural control and rapid fabrication of new materials. We generate relationships between polymer architecture and dilute solution properties, such as the intrinsic viscosity, radius of gyration, and hydrodynamic radius. Working closely with experiments, we plan to use these relationships to improve U.S. industrial competitiveness through better plastic characterization. This work is part of NIST's Macromolecular Architectures project.
A selection of non-NIST publications is provided below. A complete list is available on my Google Scholar. * co-first authorship.
NIST Sigma Xi Early-Career Poster Presentation for outstanding poster, 2023
MML Postdoctoral Fellow Accolade, 2022