ChE 2025 Fall Seminar Series

Details

Speaker: Kevin Dorfman, PhD
Distinguished McKnight University Professor Head, Chemical Engineering and Materials Science, University of Minnesota

Biosketch:

Kevin Dorfman is presently a Distinguished McKnight University Professor and Head of the Department of Chemical Engineering and Materials Science at the University of Minnesota. His research interests lie primarily in polymer physics, where he is especially well known for his work on applications of self-consistent field theory to block copolymer self-assembly and integrated experimental and computational research into DNA confinement in nanochannels. His research has been recognized by numerous awards, most notably the Colburn Award of the AIChE and a Packard Fellowship, and he is an Associate Editor for the ACS journal Macromolecules. Prior to joining the University of Minnesota as an Assistant Professor in 2006, Prof. Dorfman was a postdoctoral fellow at Institut Curie (Paris, France) working with Jean-Louis Viovy from 2002-2005. He received his MS (2001) and PhD (2002) in chemical engineering from MIT, under the supervision of Howard Brenner, and received his BS in chemical engineering from Penn State in 1999.

Talk Title: Network phases in block polymers

Abstract:
This presentation will explore our recent computational research self-assembled network phases in block copolymers, which can serve as templates for creating metallic metamaterials. In the first part of the presentation, I will describe routes to produce single-gyroid phases, chiral network with three-fold connectors. While single gyroid is metastable in simple AB diblock copolymers, blends of linear diblock polymers and neat melts of nonlinear copolymer architectures expose stability windows for alternating gyroid and single gyroid, respectively, in experimentally accessible systems. The second part of the presentation will discuss a new principle, known as boundary frustration, that guides termination plane selection between two non-preferential surfaces. Finally, I will describe a new approach known as “generative SCFT” that leverages generative adversarial networks to learn from self-consistent field theory (SCFT) solution trajectories to propose new initial guesses for subsequent SCFT calculations. This approach not only identified all known block polymer network phases, but also uncovered a vast library of candidate network phases.