Valence band egde higher than Fermi level

[qavo08]

Dear VASP users,

I have been using GGA+U to calculate the band gap of CoO. First, I performed standard self-consistent run, then I used preconverged CHGCAR file and added the high symmetry points in KPOINTS file for band structure calculation.

For the FM and AFM2 cases, I always got the valence band edge higher than Fermi level about 0.2eV. For the AFM1 case, the valence band edge is below Fermi level. Could anyone explain why in the FM and AFM2 the valence band edge higher than Fermi level and how to fix it. Thank you very much.

Here is my INCAR file:

ENCUT = 550.000000
SIGMA = 0.010000
EDIFF = 1.00e-05
EDIFFG = -2.00e-02
PREC = Accurate
ALGO = Fast
ISPIN = 2
ISMEAR = 0
NELM = 500
ISTART = 1
VOSKOWN = 1
ICHARG = 11
LORBIT = 11
NSW = 0
IBRION = -1
NPAR = 4
LDAUTYPE = 2
LMAXMIX = 4
LDAUPRINT = 2
LCHARG = .TRUE.
LWAVE = .TRUE.
LDAU = .TRUE.
LREAL = Auto
LDAUL = 2 -1
LDAUU = 4.500 0.000
LDAUJ = 1.000 0.000
MAGMOM = 1*2.6000 4*-2.6000 3*2.6000 1*-2.6000 4*2.6000 4*-2.6000 4*2.6000 3*-2.6000 1*2.6000 4*-2.6000 3*2.6000 32*0.0000

[Neutrino]

What is your definition for the Fermi level?

[qavo08]

I think I figured out the problem. It's just a matter of writting Fermi level in OUTCAR. Thank admin for nice explanation and thank Neutrino for the concern.

http://cms.mpi.univie.ac.at/vasp-forum/ ... =3&t=12817