The ABINIT Users Mailing List ( CLOSED )

[Brad Foreman <phbaf@ust.hk>]

Hi Orlin,

There are two conventions in ABINIT that determine the absolute value of 
the local potential.  First, the average value of the Hartree potential is 
set to zero by definition [1,2,3], because in a neutral system the net 
contribution to the total energy from the Coulomb interaction terms (for 
the electron-ion system) at wave vector G = 0 is identically zero.

Second, the G = 0 part of the local pseudopotential is also set to zero 
[1].  Although the contribution from this term to the total energy is not 
zero, it is merely a constant that depends only on the type of ions present 
in the system and on the average valence electron density, not on the ionic 
positions or the functional form of the electron density [4,5].  This term 
can therefore be set to zero in the Kohn-Sham potential and added on 
afterwards in the total energy.

If you would like to restore the G = 0 part of the local pseudopotential to 
its actual nonzero value, you can find this information in the internal 
variable ecore, which is evaluated in the subroutines pspini.f and pspcor.f 
and printed in the output file.  To obtain the mean value of the local 
pseudopotential, just divide ecore by the total number of valence electrons 
and by the volume of the unit cell.  This number can then be added to the 
Kohn-Sham potential and to any calculated band structures, producing 
results in which the absolute potential is defined only by the convention 
of setting the mean Hartree potential to zero.

Best regards,
Brad


[1] J. Ihm, A. Zunger, and M. L. Cohen, J. Phys. C 12, 4409 (1979); 13, 
3095 (1980).
[2] M. T. Yin and M. L. Cohen, Phys. Rev. B 26, 3259 (1982).
[3] M. T. Yin and M. L. Cohen, Phys. Rev. B 26, 5668 (1982).
[4] M. C. Payne et al., Rev. Mod. Phys. 64, 1045 (1992), Eq. (2.15).
[5] R. M. Martin, Electronic Structure: Basic Theory and Practical Methods 
(Cambridge Univ., 2004), Eqs. (12.22) and (13.1).


At 03:35 PM 7/10/2006 +0200, you wrote:
> Hi,
>
> I have done macroave calculation of a metal surface, but I see that the 
> potential does not go to zero far from it. Is it normal? I send a graph of it.
>
> Best regards,
> Orlin
>
> --
> Orlin Blajiev
> Department of Metallurgy, Electrochemistry and Materials Science
> Faculty of Applied Science Vrije Universiteit Brussel
> Pleinlaan 2, B-1050 Brussels
> Belgium
> http://www.vub.ac.be/META/
>
> tel.: 32-(0)2-6293538
> fax : 32-(0)2-6293200