order to detect heavy elements and to determine the corresponding chemical abundances in hot white dwarfs. More precisely, the ions considered in our work were Zr IV-VII, Cu IV-VII , In IV-VII, Ag IV-VII and Cs IV-VII. All these new data were then introduced into the synthetic spectra highlighting large overabundances of such elements in some hot white dwarfs, interpreted as being due to radiative levitation.

Titre:Theoretical Study of the Atomic Structure and Radiative Parameters of Lowly and Moderately-Ionized Heavy Elements: Applications to Cosmochronology and to the Spectral Analysis of Compact Astrophysical Objects.

Promoteurs: Monsieur Pascal Quinet et Monsieur Patrick Palmeri

Résumé de la dissertation:The main goal of our work was to provide atomic parameters which are urgently needed to address some of the most important current astrophysical problems such as those related to cosmochronology, stellar nucleosynthesis and the study of compact objects.To do so, we essentially used two theoretical methods to model the atomic structures and to compute the radiative data in heavy atomic systems of astrophysical interest, namely the pseudo-relativistic Hartree-Fock (HFR) approach including core-polarization effects (HFR+CPOL), and the fully relativistic multiconfiguration Dirac-Fock method (MCDF).A first section of our research was focused on theoretical atomic structure calculations in singly ionized uranium and thorium. The main goal was to produce new reliable atomic data in order to help astrophysicists detect those two very important but hardly detectable elements in stellar spectra. Transition probabilities and oscillator strengths for U II and Th II are important in astrophysics. Indeed, these two ions are used as cosmochronometers in order to determine the age of stars. For this particular use in cosmochronology, the knowledge of U II and Th II abundances in stars is of high importance. Indeed, in order to date a star, the following equation can be used:where R is the uranium/thorium abundance ratio, P is the production rate, τ is the half-life and T is the age of the star. Therefore, in order to precisely determine R, a better knowledge of Th II and U II spectra is necessary. The accuracy of this dating technique is still hampered by the lack of available radiative parameters in these two elements. In this context, we managed to establish a reliable list of U II spectral lines that could be used in cosmochronology, corresponding to 38 strong electric dipole transitions in the wavelength visible region. To do so, a pseudo-relativistic HFR+CPOL model was employed to compute the oscillator strengths, the latter being found in satisfactory overall agreement with the most accurate