[Wien] Fermi contact Interaction

[pieper]

On Mon, 8 Mar 2010 19:49:31 -0600, Vladimir Michaelis
<vladimirkm at gmail.com> wrote:
> Hello,
> 
>    I have a paramagnetic unit cell with a Re with an electron spin of 1
>    and three other non-paramagnetic ions.  I have built a supercell and
>    have run these calculations a few times using various number of
kpoints,
>    magnetic calculations with antiferromagnetic calculations, etc.
However,
>    I keep getting a fermi contact of zero, however I know that this is
not
>    real.  Are there any known limitations in obtaining the fermi contact
>    terms using this method for Re or Ru? I have also tried Ru and it
>    predicts a rather large value which was slightly counter intuitive
from
>    the experimental value recorded.    I am trying to compare these to
>    experimentally determined data that were acquired using NMR, is this
the
>    best approach?
> 
> PS: Peter I thank you for your previous response however still having a
> few problems...
> 
> 	In ionic compounds, the Fermi contact term can sometimes be only a
> 	"small" part of the total Hyperfine field. The orbital (and dipolar)
> 	terms can become large, even for 3d materials, in particular when the
> 	electronic configuration contains eg. a single 3d electron in one
> 	spin chanel (eg. high-spin Fe2+). I realize this and have looked into
the
> 	dipolar and orbital terms however I don't seem to see an affect

Maybe a question from the outside is allowed: Did you call lapwdm with  "
3 3 " (instead of " 0 0 ") in the last line of case.indm once after your
SCF? This should give you the orbital contribution to HFF (grep for :XOP in
case.scfdmup)

> 	
> 	However, it is also well known that the contact term is often
> 	underestimated by about 10-20% for eg. Fe. - OK
> 
> Also be sure, you do the correct magnetic structure (often they are
> antiferromagnetic): I have tried this and gets me closure but still not
> what I was expecting