[Wien] Mg-inst-file

[Stefaan Cottenier]

>    I want to calculate antiferromagnetic coupling in two Mg-doped GaN. I 
> don't know how to design the inst-file of Mg, since Mg has two 3s 
> electrons. 

Move 1 electron in 3p. What matters is to have a starting configuration 
with a moment. If the moment survives the scf-cycle, then you'll see 
from case.scf which electrons are responsible for it. If Mg does not 
'want' to have a moment in this matrix, you'll loose your moment during 
scf (but if you start without initial moment, you'll never pick it up 
during scf).

Stefaan