Organizers: Luigi Delle Site, Cecilia Clementi, Sara Jand

Abstract:

Molecular modeling based on electronic properties assures the inclusion into effective models of the intrinsic quantum nature of atoms and molecules. The resulting models assure a high computational efficiency in treating large systems for long time with the minimal lost of microscopic accuracy and eventually to go back to the electronic scale. The corresponding scaling up and down of physical accuracy allows to identify the interconnection between different scales and as a consequence to pin point the relevant degrees of freedom that drive properties of systems of interest.  In this minisymposium we will see techniques and applications that realize the idea above.

Speakers:

11:30h: Paolo Carloni (FZJ)

QM/MM MD simulations of proton-coupled transporters

We will discuss some recent applications of our massively parallel QM/MM code MiMiC, developed within a consortium of european university. We will focus on transporters which exploit proton gradients across the membrane to transport organic molecules. We will close with a perspective of using data science approaches to develop potentials and calculate accurate free energies using QM/MM MD data.

12:15h: Daniele Passerone (Empa, Swiss Federal Laboratories for Materials Science and Technology)

ML Potentials and Enhanced Sampling: A Powerful Strategy for Nanoscience Modelling

Being a computational scientist in an environment in which experiments are conducted in controlled conditions at the nanoscale is both a blessing and a formidable challenge. Modelling of surface chemistry, surface-supported assembly, creation of nanodevices, characterization of low-dimensional materials, single-atom manipulations, magnetism and electronic correlation effects require both a high level of theory and access to extended time and length scales in simulation. In the age of artificial intelligence, ab initio simulations, machine-learning-derived interaction schemes and enhanced sampling methods [1] work together to provide quantitative insight into the experiments next-door. I will focus on an application to enantioselective chemical reactions of a method based on the committor function [1] to sample and characterize that transition state ensemble in the scope of rare but important events.

[1] P. Kang, E. Trizio, M. Parrinello, “Computing the committor with the committor to study the transition state ensemble” Nature Computational Science 2024, 4, 451–460.