Tuesday, January 31, 2012

1201.6263 (Markus Meinert et al.)

Proving the LDA+U description of Co2FeSi wrong: an x-ray magnetic linear
dichroism study
   [PDF]

Markus Meinert, Jan-Michael Schmalhorst, Manuel Glas, Günter Reiss, Elke Arenholz
There is experimental evidence both for and against a half-metallic
ground-state of the Heusler compound Co2FeSi. Density functional theory based
calculations suggest a non half-metallic ground state. It has been argued, that
strong on-site Coulomb interaction has to be taken into account via the LDA+U
method, which predicts a half-metallic ground-state for a range of values for
U. X-ray magnetic linear dichroism (XMLD) can be used as a tool to assess the
correctness of the LDA+U approach: the calculated spectra within the LDA+U or
GGA+U schemes are clearly different from those within the LDA or GGA. Due to
its ability to separate different orbital symmetries, XMLD allows us to
distinguish unambigously between different models of the electronic structure
of Co2FeSi. In this article we discuss experimental XMLD spectra and compare
them with detailed first principles calculations. Our findings give clear
evidence for the incorrectness of the LDA+U or GGA+U band structures, whereas
constrained calculations with the GGA and a fixed spin moment of 6\mu_B give
the best overall agreement between experiment and theory.
View original: http://arxiv.org/abs/1201.6263

No comments:

Post a Comment