Description | Mechanistic Hypothesis Tests, Reaction Coordinates, and Rate Laws: The Rare Events Approach to Multiscale Modeling Abstract: It is widely appreciated that rare events methods can overcome time scale separations in computational studies of chemical reactions, nucleation phenomena, protein folding, electron transfer, etc. Their ability to also overcome length scale separations is less widely appreciated. Rare events typically involve much shorter length scales than those associated with temperature, composition, density, or stress gradients. As such, the rate laws that naturally emerge from rare events analyses provide an 'off-diagonal' and essentially exact route from the atomistic modeling scale to the continuum scale. In such multiscale modeling applications, rare events techniques that yield theories, activation parameters, and rate laws are more useful than techniques that directly predict numerical rates. Three classic examples are harmonic transition state theory, classical nucleation theory, and Marcus theory, each of which frequently appears within continuum scale models and which has also inspired countless computational methods. These classic theories are unrivaled in their ability to predict (and interpret) kinetic trends as a function of temperature, supersaturation, electrochemical potential, and many other properties. Their capabilities derive from a shared feature at their foundations: each was built around a scalar reaction coordinate with a clear physical meaning and widespread applicability across many reactants. By comparison, frameworks that directly use committors, eigenfunctions of the transition matrix, or diffusion maps invest a massive simulation effort to generate one rate at one condition. The same can be argued for methods that choose collective variables on a case-by-case system specific basis. Taking the three classic theories as a guide, I ask how we might discover generalizable reaction coordinates and useful theories for those activated processes that as yet remain poorly understood. I will demonstrate our recent progress toward this goal for problems related to ionic crystal growth and for SN2 reactions in aqueous solution. Bio: Molecular Engineering and Sciences Seminar Series This weekly seminar brings together students, faculty and invited guests from various disciplines across campus to explore current trends in molecular engineering and nanotechnology. It is a forum for active interdisciplinary discussions. These talks are open to the public and attract a diverse audience of students and faculty. |
---|