The ABINIT Users Mailing List ( CLOSED )

[matthieu verstraete <matthieu.jean.verstraete@gmail.com>]

The compensating background (abinit charge variable), or some other
compensating charge (not coded), _must_ be added in order to avoid
divergence of the Ewald sums. There is always an arbitrary constant in
the potential, so I don't think this is too much of a problem. If you
have a surface, you compare to the vacuum potential, otherwise to the
Fermi level or some deep core state... For the total energy, you will
be comparing to the undefected solid and to some energy for the
dopant. I'll have a look at the 2008 paper, but basically abinit uses
very primitive jellium compensation charge.

Matthieu

On Thu, Sep 3, 2009 at 5:11 PM, <rolando.saniz@ua.ac.be> wrote:
> Dear Matthieu,
>
> Thanks for your reply. It has clarified a lot to us.
>
> Regarding the total energy in the case of charged impurities, however, our
> question was not really refering to supercell size corrections. We are not 
> sure
> of what is done in Abinit in the following sense. It appears that total 
> energy
> calculations in most codes are done using the total energy expression of 
> Ihm,
> Zunger, and Cohen [J. Phys. C 12, 4409 (1979)]. But this expression is valid
> only for charge-neutral systems [see, e.g., Lany and Zunger, PRB 78, 235104
> (2008)]. If the same total energy expression is used for charged supercells,
> then the total energy will be defined only up to a constant. So the question
> would be, is a compensating background added in Abinit for charge 
> neutrality,
> or is the total energy expression used adapted to charged supercell cases,
> i.e., is the total energy free of an arbitrary reference?
>
> Thanks for your time!
>
> Best regards,
> Rolando
>
> **********
> Rolando Saniz
> Departement Fysica
> Universiteit Antwerpen
>



-- 
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Dr. Matthieu Verstraete

Universite de Liège
Institut de Physique, Bat. B5
Allée du 6 aout, 17
B- 4000 Sart Tilman, Liège
Belgium

Phone : +32 4 366 37 50
Fax   : +32 4 366 36 29

Mail : matthieu.jean.verstraete@gmail.com