average potential

Message

zsoltrak · #1 Post by **zsoltrak** » Tue Aug 19, 2008 9:21 pm

Dear VASP users,

I am doing defect calculations using supercell models. In oreder to calculate the energy of the VBM in the presence of the defects, I want to align the average potentials of the defect containing supercell (away from the defect) and the pure supercell. Can somebody, please, tell me how to calculate these average potentials? Can I get them from the LOCPOT file?

Thanks,
Zsolt

Gu Chenjie · #2 Post by **Gu Chenjie** » Mon Feb 21, 2011 7:58 am

hi, could you share the experience on this problem with me, thanks a lot.
archygu at gmail.com

tlchan · #3 Post by **tlchan** » Mon Feb 21, 2011 5:14 pm

One accepted method is to pick a region (typically an atomic sphere, muffin-tin radius, or a region containing a few atoms depending on your specific calculation) for the defect-free supercell and average the local potential within the region. Then you pick a corresponding region for the defect-containing supercell but far away from the defect. The two average potentials can be aligned. The local potential is written in LOCPOT. Another method is to align using core levels instead of the local potential.

superyoyo · #4 Post by **superyoyo** » Tue Feb 22, 2011 8:02 am

with PAW potentials, you can simply use the "average (electrostatic) potential at core" of the atom which is far from defects and align with that of atom in defect-free supercell. Please search keyword "norm" in OUTCAR file.

Cheers!