The ABINIT Users Mailing List ( CLOSED )

[Xavier Gonze <gonze@pcpm.ucl.ac.be>]

Dear Miguel,

Miguel Angel Salvado Sanchez wrote:
>         Hello!
>
>         Thank you for the answers.
>         My question was specifically for methods (codes) that solve the 
> Kohn-Sham
> equations
>         without using pseudopotentials or other frozen-core methods.
>         In some moment I have thinked that this could be equivalent in ABINIT
>         to use some odd kind of pseudopotential. For first row elements may be
>         not so unreasonable to use an all-electron method with plane waves as 
> basis
> functions.

You can use indeed all-electron potentials, with an extremely small smearing
at the origin. The easiest is to enter a Goedecker-type potential file,
with only the local part, see e.g.
~ABINIT/Psps_for_tests , the files 01h.bare and 02he.bare .
I give an example for C below (this might be called 06c.bare).
It is smoothed within 0.001 Bohr of the origin.
Note that for such a potential, you will need extremely high cut-off energy
to be converged.

Xavier

---------------------------------------------------------------------------

Bare (erfc) Goedecker-Teter-Hutter  Wed May  8 14:27:44 EDT 1996
6   6   960508                     zatom,zion,pspdat
2   1   0    0    2001    0.       pspcod,pspxc,lmax,lloc,mmax,r2well
0.0010000  0   0  0  0     rloc, c1, c2, c3, c4
0 0 0                              rs, h1s, h2s
0 0                                rp, h1p