Dick Stufkens Prijs 2018 awarded to chemical physicist Stefan Vuckovic

Holland Research School of Molecular Chemistry honours innovative PhD-research

16 October 2018

The Dick Stufkens Prize 2018 for the best PhD thesis of the Holland Research School of Molecular Chemistry (HRSMC) will be awarded to Dr Stefan Vuckovic. In his thesis, Vuckovic presents a new and innovative approach to the quantum mechanical calculation of atomic and molecular energies. It lays the foundation for a more accurate modelling of drugs, catalysts, solar cells and batteries.

The calculation of molecular properties with the help of quantum mechanics (the Schrödinger equation) has become an integral part of chemical research. In principle "everything" can be calculated: geometries, excitation energies, NMR, ESR, IR, transition barriers in reactions, etc. etc.. But the accuracy still needs to be improved. A system of many electrons which repel each other and are attracted by the nuclei (an N-particle system with N possibly quite large) is extraordinarily complicated. The correct incorporation of the instantaneous interaction of the electrons is denoted the electron correlation problem. Exact solutions are not available.
Among the approximate methods the so-called density functional theory (DFT) based ones are currently most popular. This theory simplifies the N-particle problem to a number of 1-particle problems.  DFT can boast many successes for large systems, ranging from medical drugs to solar cells, catalysts and materials research (e.g. for batteries). Now, Stefan Vuckovic has devised a new approach to achieve a next level of accuracy of DFT calculations. The jury of the Dick Stufkens Prize is impressed by the high level of Vuckovic's thesis and in particular by its innovative character.

New starting point yields new insight

Stefan Vuckovic focuses on the crucial electron correlation energy. In general, DFT calculations contain rather bold approximations of this correlation energy. Vuckovic now takes a new point of departure, i.e. the case of strong correlation. Here, the electron repulsion is taken so large that it dominates all other properties (also the kinetic energy). A necessary constraint is that the electron density is fixed. As it turns out, these extreme conditions lead to completely new and surprising insights, according to the jury: "When this work fulfils on its promise, the application of DFT in (computational) chemistry will make a big step forward, in particular regarding the accuracy of the results."

Elegant and promising approach

In his thesis, Vuckovic presents a thorough mathematical-physical analysis of the "strongly correlated electrons" (SCE) model. The jury was particularly pleased that this analysis leads him to one of the most interesting proposals of recent years for improving DFT methods. He employs the simple concept of an "exchange-correlation hole" that an electron creates around itself by pushing aside the other electrons. A proposal is made for a completely nonempirical calculation of this exchange-correlation hole. The first results with a simple approximation are promising. Much work will still be needed in order to make the method efficient and competitive, but the solid basis in the physics of electron correlation is a great advantage of this method over the many heavily parametrised and often not very transparent DFT functionals that have come to dominate applications of DFT in chemistry. 

Stefan Vuckovic conducted his research at the Theoretical Chemistry group of the Vrije Universiteit Amsterdam. His dissertation is entitled 'Fully non-local exchange-correlation functionals from the strong-coupling limit of density functional theory'. The defense took place on November 14, 2017, with Prof. Paola Gori Giorgi as promoter and Dr Michael W.J. Seidl as co-promotor. Vuckovic will shortly be working in California with an NWO-Rubicon grant, with the well-known DFT theoretician Prof. Kieron Burke.

Source: University of Amsterdam UvA