How to calculate the charge transfer in VASP?

Message

Anonymous · #1 Post by **Anonymous** » Fri Sep 17, 2004 1:07 pm

For the metal/oxide interface structure, how can I determine the charge transfer from the oxide to metal or the reverse?

#2 Post by **admin** » Wed Oct 27, 2004 8:10 am

PLEASE NOTE: unless for methods based on localized atomic wavefunctions
as basis sets, the 'atomic/ionic' charge and hence the charge transfer is
arbitrary to some extent, because it depends on the 'atomic volume' over which the charge density is integrated.

in VASP, the volume is defined by RWIGS (the Wigner-Seitz radius).
hence, you could either
1) use PAW potentials (RWIGS defined) or
2) set RWIGS to a reasonable value in INCAR
-- set LORBIT to an appropriate value (have a look into the online-manual)
and take the differences of the charges within these spheres before/after
adsorption as a measure for the charge transfer.

graeme · #3 Post by **graeme** » Sun Jul 03, 2005 6:34 pm

There is also a small program to calculate a Bader partitioning of the vasp charge density at http://theory.cm.utexas.edu/bader . This is a somewhat more natural partitioning of atomic charges because it does not require specifying atomic radii, and it is based only upon the charge density, CHGCAR file. In the Bader analysis, atomic regions are separated by surfaces of charge density minima between atoms. Atomic charges are calculated as the integrated charge density within these Bader regions.

Jiaye Li · #4 Post by **Jiaye Li** » Wed Mar 28, 2007 9:13 am

Recently I want to calculate the charge transfer of elements, i.e. how many electrons left of a specific atom after optimization of a slab