(324a) Polymer Design in the Era of Automated Simulation and Experimentation

Systematic advances in our ability to simulate polymeric materials and their constituents have raised the possibility of generating comprehensive data bases that are built from detailed atomistic models. Modern high throughput experimentation systems are also starting to produce vast amounts of data that could be used to populate data bases for specific classes of polymeric materials. In this presentation, I will describe our group's efforts to develop a multiscale simulation and machine learning approach that incorporates (i) featurization of the polymeric molecules of interest, (ii) multiscale simulation of their equilibrium and out-of-equilibrium properties, and (iii) incorporation of experimental and computer-generated data into algorithms for design of new materials. The capabilities of the proposed framework will be discussed in the context of three classes of materials, including polyolefins, polyacrylates, and photoactive polymers.