Calculation of photo-emission spectra on the basis of a combination
of the KKR-method and the DMFT


Hubert Ebert

Dept. of Chemistry and Biochemistry / Physical Chemistry, University of Munich, Butenandtstr. 5-13, D-81377 München, Germany


H. Ebert (a), J. Minar (a), L.Chioncel (b), M.I. Katsnelson (b), and A.I. Lichtenstein (c),

(a) Phys. Chem., University of Munich, Germany
(b) University of Nijmegen, Netherlands
(a) Phys. Chem., University of Munich, Germany
(c) Institute for Theo. Physics, University of Hamburg, Germany


The Korringa-Kohn-Rostoker (KKR) or, equivalently, multiple scattering technique, is an extremely flexible band structure method because it gives direct access to the electronic Green's function. This feature is exploited for example when dealing with impurities, disordered alloys or surfaces. In addition this supplies a very reliable and sound basis when dealing with spectroscopic properties. In particular concerning photo-emission a straightforward implementation of the so-called one step model can be achieved. Corresponding calculations of photo-emission spectra have been done in general within the framework of local spin density functional theory (LSDA). Recently, a combination of the KKR-method with dynamical mean field theory (DMFT) has been presented by us, that allows an improved treatment of correlation effects compared with standard LSDA calculations. As it is shown, this development opens the way to corresponding photo-emission calculations. The most important technical aspects of the KKR+DMFT scheme will be presented together with applications to transition metal alloys. This is followed by a description of the KKR-based implementation of the one-step model of photo-emission. Results of corresponding calculations of various types of spectra (spin-resolved, angle-integrated and -resolved, respectively) will be presented and discussed. In addition it will be shown, that the investigation of spin-orbit induced effects as the Fano-effect and magnetic dichroism is possible by using a fully relativistic formulation of our scheme.