The ABINIT Users Mailing List ( CLOSED )

[Masayoshi Mikami <mmikami@rc.m-kagaku.co.jp>]

Dear Ariesto Pamungkas:

The mixture of LDA-PPs and GGA-PPs should be avoided in principle,
from the viewpoint of the following paper :
"Pseudopotential study of binding properties of solids within  
generalized gradient approximations: The role of core-valence exchange  
correlation"
" M. Fuchs, M. Bockstedte*, E. Pehlke§, and M. Scheffler":
Phys. Rev. B 57, 2134–2145 (1998)
http://prola.aps.org/abstract/PRB/v57/i4/p2134_1? 
qid=fca3aa34eb6a56c1&qseq=2&show=10

Once the xc potentials are consistent, the different type of PPs (TM,  
Hamann, Rappe et al, ...) may be mixed.

Are there some updates in this regard ?

Best wishes,
Masayoshi

On 2004/03/18, at 18:23, Ariesto Pamungkas wrote:

> Hello all abinit user
>
> I found that one type of pseudopotential or XC can
> give good result for one element but for other
> element , other type of pseudopotential give better
> result. Is it oke using different pseudopotential or
> XC to calculate a compound?  ,for example using
> Ge(Haman/GGA), Sb(TM/GGA), Te(Haman/GGA) and
> Sn(Haman/LDA) potential to calculate GE-Sb-Sn-Te
> compound?.