Recent developments in multiconfigurational density functional theory and applications in actinide chemistry

e-mail:efr@ifk.sdu.dk

Lundi 19 mars 2007, 11^h00
Density-Functional Theory (DFT) is widely used nowadays in quantum chemistry, in particular because it is computationally cheap and thus can be applied to large scale systems. But, even if standard approximate functionals enable a rather accurate calculation of the dynamic correlation (Coulomb hole), the static correlation effects (degeneracy or near-degeneracy) cannot be treated adequately in DFT. Nevertheless those can be described in Wave-Function Theory (WFT), for example with a Multi-Configurational Self-Consistent Field (MCSCF) model, but then an important part of the dynamic correlation has to be neglected. It can be recovered by perturbation-theory based methods like CASPT2 or NEVPT2, but their computational complexity prevents large scale calculations. It is therefore of high interest to develop hybrid models which combine the best of both WFT and DFT. This can be achieved for example by splitting the regular two-electron Coulomb interaction into long-range and short-range parts as proposed by Andreas Savin. In this so-called "Short-range DFT" (SRDFT) approach, the long-range interaction is treated in WFT and the short-range interaction in DFT.

The first part of the talk deals with the development of perturbation theory based SRDFT methods. The derivation of the Möller-Plesset-SRDFT energy and wave function is presented. We show in particular that the self-consistent generalized Bloch equation is equivalent at any order n of perturbation (n>1) to a linear self-consistent equation fulfilled by the contribution of order n to the density matrix. A Multi-Reference extension of the MPn-SRDFT such as a NEVPTn-SRDFT method is also proposed. The MP2-SRDFT method (implemented in a development version of the DALTON program package) is then used as a tool for investigating the optimal long/short-range separation.

The second part of the talk deals with the application of SRDFT methods in actinide chemistry. The optimal long/short-range separation for actinides is in particular discussed considering actinide compounds such as ThO₂, PaO₂+, UN₂, UCO, UO₂++, NpO₂+++ and PuO₂++++.