How to calculate DOS for Molecule??

[mharis]

Hi All,

I have a problem in calculating DOS for CO molecule. Here is the detail.
First I optimized the CO bond length in a big cell size of 15x15x15. I used only gamma point in this calculation. As results, I got the CHGCAR and CONTCAR files.

I used these two files to calculate DOS, I increased the number of k-points to 7x7x7 and used the following INCAR file:
SYSTEM = CO
NWRITE = 2
ISTART = 0
ICHARG = 11
INIWAV = 1

NELM = 100
NELMIN = 4
EDIFF = 1E-06
IBRION = -1
NSW = 0

ISMEAR = -5
LREAL = .FALSE.
LORBIT = 2
RWIGS = 0.863 0.820
NPAR = 1

I got the DOSCAR file, but it doesn't seem to be correct. I expected the DOS to have 2 peaks (2pi* and 5s), but it has only 1 peak.

Is there something wrong in how I calculated DOS for molecule??
Please help me.
Thank you.

[alex]

Hi there,

you may consider a couple of things:
- molecules do not need a DOS description, they come with discrete energy levels and do not show energy dispersion, so a single k-point is enough (and shows the discrete energy eigenvalue spectra)
- VASP is a code which comes with pseudopotential, so some expected states might be within the PPs core

Hth

Alex