The actinide serie shows a rich variety of electronic and structural behaviors. This serie can be roughly halved with the light actinides, firstly, that have little or no electronic correlations (delocalized f electrons), and on the other hand, the heavy actinides for which they are significant (localized f electrons). This transition from an itinerant to localized electrons has a significant influence on the structural properties of these metals.

We now have a code for calculating the electronic structure, ABINIT, in which we have developed methods that can account for these different behaviors: PBE0, LSDA + U, DMFT ... An original parallelization has also recently been implemented in the context of this code to effectively take advantage of the supercomputer capacity available at the Military Applications Division (one of the most powerful in Europe).

We propose, as part of this thesis, to study the influence of electron correlations on the reactivity of actinide surfaces. By means of ab initio calculations, the candidate will focus on the early stages of oxidation. At the same time a systematic will be considered to determine the influence of these correlations on the structural and electronic properties of actinide oxides. They play an important role in UO2 whereas they do not appear in the metal. The causes of this behavior remain an open issue and require a theoretical effort combined with the simulation.

Strong expertise in solid state physics is required at both theoretical and material level. A taste for programming would be desirable in the context of this study.