New ways to bridge the gap between microscopic simulations and macroscopic chemistry

Date: 18 October 2018

Venue: Campus Middelheim, A.143 - Middelheimlaan 1 - 2020 Antwerpen (route: UAntwerpen, Campus Middelheim)

Time: 2:30 PM

Organization / co-organization: Department of Chemistry

PhD candidate: Kristof Bal

Principal investigator: Erik Neyts

Short description: PhD defence Kristof Bal, Faculty of Science - Department of Chemistry


Over the last decades, computational chemistry has become a mature and widely practiced discipline of the chemical sciences. Computational methods run the gamut from molecular dynamics (MD) simulations with approximate classical potentials to high-level quantum mechanical electronic structure calculations. At their best, computational models provide a fundamental peek into atomic-level properties of matter and greatly improve our understanding of processes all across chemistry. It is sometimes even possible to predict the outcome of an experiment, besides being able to give a rationalization of it afterward. However, atomic-level models are naturally limited to microscopic time and length scales. A key difficulty of computational investigations is hence their inability to match the size and complexity of most experiments. This poses an important challenge to computational chemists: how to achieve a useful interplay between model and experiment?

In this thesis, new ways to bridge theory and reality are investigated. In particular, two case studies are presented. In the first part of this thesis, a new general method to extend the short time scale of atomistic simulations is developed—the model is brought closer to the experiment. The second part of this thesis concerns the construction of a new model for the interpretation of a specific experiment—the model simplifies the experiment. These two approaches represent two sides of the same coin. On one hand, the limited scale of the model can be perceived as a weakness, and therefore approaches must be developed that allow to extrapolate simulated microscopic molecular dynamics to experimentally measurable macroscopic observables. On the other hand, the former idea can also be turned on its head: in this view, the model can be a “pure” representation of a phenomenon, whilst macroscopic experiments are only the sum of many different microscopic components, none of which can be studied on its own—in that sense, the model is the bottom-up inverse of a top-down experimental approach. This philosophy is applied to the interaction between a plasma and catalyst: a model is constructed to specifically investigate the plasma-induced changes of the catalyst’s electronic structure, i.e., how a plasma can change the chemical properties of a catalyst.

Both cases, however, clearly show that model and experiment can and should co-exist: precisely because of their vastly differing scales, they offer different but complementary views of the same problem.