My Community

Dear all,
I am trying to remove semicore d electron from Ga.
According to what I read in the manual I at first perform a non selfconsistent calculation ICHARG=12 in order to get the initial band structure. then

NOW I fix the occupation (Ga with 13 electron in the valence)

ISTART=0
ISPIN=2
ISMEAR=-2
ENCUT=287.594
ENAUG=531.356
NUPDOWN=1
NELMDL=5
FERWE=5*1 1*1 3*0.333 3*0
FERDO=5*1 1*1 6*0
SIGMA=0.1
LDIAG=.FALSE.

I get the TOTEN from OUTCAR

Now I remove 1 electron d

ISTART=0
ISPIN=2
ISMEAR=-2
NELECT=12
ENCUT=287.594
ENAUG=531.356
NUPDOWN=2
FERWE=5*1 1*1 3*0.333 3*0
FERDO=4*1 1*1 7*0
SIGMA=0.1
LDIAG=.FALSE.

BUT I obtain the same IDENTICAL energy of s electron removal

ISTART=0
ISPIN=2
ISMEAR=-2
NELECT=12
ENCUT=287.594
ENAUG=531.356
NUPDOWN=2
FERWE=5*1 1*1 3*0.333 3*0
FERDO=5*1 7*0
SIGMA=0.1
LDIAG=.FALSE.

with ISMEAR=-2 the occupancy it is fixed, isn't it?
Please tell me where I fail!

Best
G

Perhaps a degeneracy issue? That would physically cause the orbitals to have the same energy...

FERDO=4*1 1*1 7*0 for d and FERDO=5*1 7*0 for s and FERDO=5*1 1*1 6*0 for fixed. Removing one electron for d or s adds up to 12 in either case, and for the fixed occupation adds up to 13, so that part looks correct.

I think it has something to do with degeneracy, but I could be wrong... ?!