The ABINIT Users Mailing List ( CLOSED )

[Nicola Marzari <marzari@MIT.EDU>]

This, by the way, raises an old question - atomic codes all have
spherical symmetry, but there are cases where fractional occupations
give raise to "cylindrical" atoms with lower (often significantly lower)
total energy,  (suppose you have Fe+, that has 3d6 4s1 -
what do you do e.g. with the minority d electron ? You could have one 
minority orbital filled 1, 2 orbitals filled 0.5, 3 orbitals filled
1/3, or 5 orbitals filled 0.2 . Or maybe even something in between.).

Does anyone have expertise, or a theoretical/fundamental comment
on this ?

I vaguely remember a comment by Walter Kohn saying that cylindrical
atoms should be more accurate (even when their total energy is
higher than the spherical atom), but I can't find the reference.
There is also a paper by Englisch and Englisch from 1983, that mentions 
that fractional occupancies are not v-representable.

Best,

                        nicola



Eric J. Walter wrote:

> Dear David, 
>
> I think the easiest and most straight forward thing to do is to run
> the atomic pseudopotential calculations within abinit.  Just use a
> very large unit cell.  Also, keep in mind that you may need to enforce
> a particular symmetry (to get a particular atomic configuration) and
> you may want to run the calculation spin-polarized.
>
> Hope this helps,
>
> Eric
---------------------------------------------------------------------
Prof Nicola Marzari   Department of Materials Science and Engineering
13-5066   MIT   77 Massachusetts Avenue   Cambridge MA 02139-4307 USA
tel 617.4522758  fax 617.2586534  marzari@mit.edu  http://nnn.mit.edu