Davide Sangalli, Andrea Marini, Alberto Debernardi
We propose a first-principles scheme for the description of the magneto-optical kerr effect. Using as reference materials iron, cobalt and nickel we show that, in contrast to the common wisdom, pseudo-potential based calculations give accurate predictions. Moreover our scheme includes excitonic and local field effects which are crucial for the description of magnetic excitons in semiconductors. We show that the Kerr parameter must be constructed from the current-current response function instead of the commonly used density-density response function. This implies that the state-of-the-art density based approaches cannot be applied and macroscopic quantities need to be reformulated in a current based formalism.
View original:
http://arxiv.org/abs/1205.1994
No comments:
Post a Comment