[Wien] (no subject)

[Peter Blaha]

> does anyone know a way to :
>
> * restrict the Bader charge integration to a particular l-value?  (the difficulty here is that the Bader volume of an atom often extends into the interstitial region)
> * plot the charge density as a function of r (distance to the nucleus, i.e., radial coordinate in spherical coordinate system) for a particular l-value?  For the total charge inside the MT this is easy, but is there a good way for, eg., s-charge or d-charge ...

First you need to restrict the density calculation inside the spheres to
summations over selected lm values only.

More difficult is the interstital.
In principles you can expand the plane waves into an lm-expansion of
bessel functions times Ylm(K) Ylm(r). With such an expansion it is in
principle possible to "expand" an interstitial charge density into an
expansion similar to that inside the spheres.

The obvious problems are: which part of the density should be described
from which atom (this would be solvable using the Bader-surfaces)

The further you go away from the atom, the higher you need to go in the LM
expansion. Again, for your question of s,p,d charges this is not a real
problem.

So all what is necessary is to expand PW into bessel functions for large
R-values. (Such an expansion at R=RMT is done eg. in lapw0 for the
potential, where the coulomb potential in the interstitial is expanded
into an LM series at RMT to give the boundary values necessary to solve
Poisons equation inside the sphere.

Regards

                                      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/
--------------------------------------------------------------------------