strange in DOS result for CO free molecule

[pnita]

Hi all,

I'm trying to model CO free molecule and do the DOS calculation. However, I found a strange result in my DOS peaks. There are peaks representing d-orbitals in my DOS analysis of CO (which shouldn't occur)...even the intensity is just about 0.8 states/eV and very small compared to the s- and p- orbital peaks.

Do you have any suggestion on this strange result?

Below is the INCAR I used in the calculation.

general:
SYSTEM = CO
ISTART = 1
ICHARG = 11
PREC = high
LREAL = Auto

Electronic Relaxation:
ISMEAR = 1
SIGMA = 0.1
ENCUT = 400
GGA = 91

DOS related values
RWIGS = 0.863 0.820 !Wigner-Seitz radii
LORBIT=1
NPAR=1
IDIPOL=3
LDIPOL=.TRUE.
LVTOT=.TRUE.

Thank you in advance for your help.

[admin]

1) free molecules have a discrete line spectrum, not a dos. therefore use the gamma point only, Gaussian Integration and narrow smearing
(SIGMA=0.001)
2) please note that the partial DOS is obtained by projection of l-like charge onto the spheres (please check whether your spheres overlap!), vasp does not use local basis functions.