UT theoretical chemistry code forum

Dear all,

i have no experience for Bader charge. Recently I used VASP to calculate the Bader charge for metal/metal interface.

1. i found charge transfer about 1 e between metals: one looses about 0.5e, and another gains about 0.5 e.

i think it's too big value...?

i checked charges in OUTCAR,but does not match with that from Bader calculation.

why do they give so much different results?

2. and next, what is difference between CHGCAR and CHGCAR_sum?

i know that the CHGCAR_sum is sum of core and valence.

the CHGCAR also includes all s, p, and d electrons including the core and valence.

Thanks a lot in advance.

Azsa

The partial charges calculated in vasp are integrated around spheres of size RWIGS. Depending upon this radius, the result could be quite different from the Bader charges. One thing that the Bader charges have going for them is that they partition the total charge uniquely among the atoms so that no charge is lost or double counted. This is a nice property for determining charge transfer between atoms.

The total CHGCAR_sum includes both valence and the core charges from the PAW potentials; CHGCAR includes just the valence. The former reproduces the charge density maxima at the atomic cores, which is important for the analysis. The later has the charge which you want to partition, over which the integration of charge is done.