Theory and computation, in synergy with experiment, are playing an increasingly important role in the design and characterization of new materials. In this talk, I will describe the efforts we are making in my group to develop new computational methodologies that address specific challenges in free energy exploration and generation. In particular, I will describe our recent development of enhanced free energy based methodologies for predicting structure and polymorphism in molecular crystals [1-3] and for determining conformational equilibria of oligopeptides[4-7]. The strategies we are pursuing include heterogeneous multiscale modeling techniques, which allow “landmark” locations (minima and saddles) on a high-dimensional free energy surface to be mapped out, and temperature-accelerated methods, which allow relative free energies of the landmarks to be generated efficiently and reliably. I will then discuss a new scheme for using neural networks to represent multidimensional free energy surfaces and the use of the aforementioned enhanced sampling methods to generate the data needed to train the networks.
[1] T. -Q. Yu and M. E. Tuckerman Phys. Rev. Lett. 107, 015701 (2011).
[2] T. –Q. Yu, E. Vanden-Eijnden, P. –Y. Chen, A. Samanta, and M. E. Tuckerman J. Chem. Phys. 140, 214109 (2014).
[3] E. Schneider, L. Vogt, and M. E. Tuckerman, Acta Cryst. B (in press).
[4] J. B. Abrams and M. E. Tuckerman J. Phys. Chem. B 112, 15742 (2008).
[5] M. Chen, M. A. Cuendet, and M. E. Tuckerman J. Chem. Phys. 137, 024102 (2012).
[6] A. T. Tzanov, M. A. Cuendet, and M. E. Tuckerman J. Phys. Chem. B 118, 6529 (2014).
[7] M. Chen, T. –Q. Yu, and M. E. Tuckerman Proc. Natl. Acad. Sci. U.S.A. 112, 3235 (2015)
Copyright. All Rights Reserved.