[Wien] Engel-Vosko's GGA

[Peter Blaha]

> This question is not particularly related to WIEN but to DFT in general,
> in special to the Engel-Vosko's GGA.
>
> I know that this GGA is a good aproximation for the exchange-correlation
> potential, giving reasonable gaps for semiconductors, but it produces bad
> total energies, ie. Energy vs Volume curves are wrong.
>
> I am wondering how good are the relative energies among different magnetic
> states using this GGA.
>
> For instance, could I obtain good Heisenberg exchange parameters, J, using
> it?.

I would very much doubt that any quantity derived from the XC-energy is
good.
The only exception might be, where standard GGA underestimates magnetism
and the EV-GGA gives stronger magnetism.

                                      P.Blaha
--------------------------------------------------------------------------
Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-1-58801-15671             FAX: +43-1-58801-15698
Email: blaha at theochem.tuwien.ac.at    WWW: http://info.tuwien.ac.at/theochem/
--------------------------------------------------------------------------