Machine Learning-Enabled Self-Consistent Field Theory for Soft Materials

University Of California-Santa Barbara, Santa Barbara CA

Investigators

Abstract

Computer simulations are a powerful tool for understanding, predicting, and discovering soft material formulations such as polymer systems. Polymers are composed of long molecular chains and are ubiquitous in both synthetic (e.g., nylon, polyethylene, polyester, Teflon) and natural (e.g., DNA, proteins, cellulose, nucleic acids) settings. In this project, computational tools will be developed that combine machine learning and scientific computing for the exploration and prediction of polymer systems, which will also help to accelerate the discovery of new materials. More broadly, this project will provide a framework for similar computationally costly problems that could be dramatically expedited by using machine learning. Last but not least, the project will serve as an anchor for the interdisciplinary training of both undergraduate and graduate students in an emerging field of much demand. Parameter space exploration for a soft material is an instance of the forward problem: given a set of parameters, find the corresponding stable morphology. But the inverse design problem, which consists of obtaining the formulation parameters that stabilize a given target morphology, is also of great technological importance as it facilitates the design of new functional materials with highly-tuned and desired properties. The numerical solution of both forward and inverse design problems requires the repeated evaluation of the computationally costly functions. The research team will develop efficient computational methods to enable polymer self-consistent field theory with machine learning to accelerate the solution of both forward and inverse design problems aimed at facilitating the discovery of new structures and the design of polymers and polymer systems. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.