 |
Center for Nonlinear Studies |
Recent developments in statistical inference and machine learning have paved the way for systematizing and expediting the process of colloidal material (i) design and (ii) characterization. With respect to the former, we demonstrate the utility of inverse statistical mechanics for the design of microphase-structured materials comprised of clusters and pores. The systematic nature of our approach enables the design of interactions that prompt the high fidelity equilibrium self-assembly of the target structures. Regarding characterization, we explore the application of machine learning for the automated detection of phase transitions. Leveraging Principal Components Analysis, order metrics that characterize equilibrium freezing, demixing, condensation, and a non-equilibrium absorbing phase transition in a model of active matter are discovered.
Host: Jeffery Leiding