Which relativistic approximations are necessary and when for modeling the X-ray spectroscopy of actinides?

The need for relativistic Hamiltonians in the calculation of the electronic structure of heavy elements is already well established. For calculating X-ray spectroscopies, such as absorption (XAS) and emission (XES), these effects are even more pronounced due to the participation of core electrons. A variety of approximations have been made to treat relativistic effects more cheaply than the full 4-component Dirac equation, including effective core potentials, all-electron scalar relativistic, and all-electron spin-orbit calculations. These methods can then be combined with a variety of approaches, including density functional theory (DFT), coupled-cluster theory, or configuration interaction to include electron correlation and model the relevant excited states. In this talk I will be presenting calculations using several different levels of relativistic theory on uranium and other actinide complexes to better understand the role of these approximations on the expected accuracy of computed X-ray spectra. An assessment of the differences between computed and experimental spectra shows that while sometime spin-orbit calculations may be required for even qualitative agreement, for other cases there is little quantitative difference with a simpler scalar relativistic approach. This is due not only to the specific complex of interest, but also the selection rules governing the transitions for the X-ray edge of interest.