PhD Conferral Sara Giarrusso

03/02/2020

13:45

Aula, Vrije Universiteit Amsterdam

Approximations and exact properties of density functionals from the strong-interaction limit of DFT

Sara Giarrusso

Prof.dr. P. Gori Giorgi, copromotor dr. K.J.H. Giesbertz

Amsterdam Institute of Molecular and Life Sciences

Sciences

PhD conferral

In the last decades, Density Functional Theory, in its Kohn-Sham formulation (KS-DFT), has emerged as an essential method for predicting and understanding a wide range of chemical and physical properties, due to its excellent accuracy compared to its computational cost.
In practice, it accounts for over 90% of all quantum mechanical simulations.

The use of KS-DFT relies, however, on models for the so-called exchange-correlation (XC) functional, which is the term that collects all effects which we are not able to treat explicitly. Despite its numerous successes, there is still a lot of cases (dissociation processes, reactions involving transition metals, non-covalent interactions) where its accuracy and, therefore, its applicability have to be improved.
The overall goal of my research is to extend the applicability of KS-DFT by developing new models anchored on solid theoretical ground, from the standpoint of a mathematically exact limit of the theory, known as its “strong-interaction limit”.

In particular, the work presented in my Ph.D. thesis
• highlights the broad potential of models, containing the strong-interaction limit ingredient, for future applications on challenging compounds such as those containing noble metal clusters and complexes;
• proposes a novel approach towards an efficient resummation of the perturbation series expansion around the Hartree-Fock reference state which possesses many promising qualities at the theoretical level;
• reports for the first time several exact properties useful to guide density functionals development, such as an analytical relation between the two most commonly used definitions of the XC functional, a qualitative feature of the effective potential in the case of a stretched hetero-nuclear bond and a new sum rule (among others).

In summary, this work contributes to show that explicit functionals of the density which are distinctive of the strong-interaction limit can be used to build novel methodologies, in contrast with a famous classification of the possible methods in the field - known as the Jacob’s ladder towards the “heaven of chemical accuracy” – according to which implicit and computationally expensive functionals are required for improved quality.
Sara Giarrusso