My Community

I computed (G03/G09) orbital energies of molecules containing heavy atoms such as Ge. Now I'd like to compare the respective orbital energies with VASP calculations but the results seem to be rather different? I used the samePOTCAR-files as one would use for bulk calculations with VASP. What would be the best approach to calculate the orbital energies of a molecule?

Thanks a lot for your help!

Hi,

it's not that straightforward, because in VASP you calculate a 3D periodic image of you your molecule and hence there is no vacuum for a zero energy level. You would like to calc. the workfunction of the system to gain this value. Check the forum for an appropriate post.

Cheers,

alex

I would proceed as follows:
-- in 3D periodic systems, the E=0 level is not defined to within a constant offset Evacuum, this is correct; HOWEVER, you can use the tool vtotav.f (to process the LOCPOT file
written if LVTOT= .True. is set) in order to estimate Evacuum in the molecule calculation (i.e. a unit cell with a sufficiently large vacuum region). The energies of the Kohn-Sham levels should then be referred to that E, or to the workfunction (which is the energy of the Fermi level with respect to Evacuum). Please choose which one is more appropriate in your case.
-- please check the character of a certain Kohn-Sham state by having a look at the orbital (charge) character as written in the file PROCAR (set LORBIT = 11 or 12 in INCAR). As you should use the Gamma point only for the free molecule, there is no band dispersion of the states)


[ Edited Thu Aug 30 2012, 03:53PM ]