Weird occupation numbers

Message

jlrch · #1 Post by **jlrch** » Fri Mar 27, 2009 3:52 pm

Hello,

does anybody have ever found the following situation?

System: Cobalt cluster with five atoms and magnetic moment of 12 Bohr magnetons

Band occupation per spin near the Fermi level:

Spin up:
.
.
25 -5.4319 1.00000
26 -5.3884 1.00000
27 -4.6481 1.00000
28 -4.3434 1.00000
29 -4.1457 0.50000
30 -2.1596 0.00000
31 -1.8285 0.00000
.
.

Spin down:
.
.
13 -4.7772 1.00000
14 -4.4672 1.00000
15 -4.2492 1.00000
16 -4.0607 1.00000
17 -3.8720 0.50000
18 -3.7469 0.00000
19 -3.7432 0.00000
.
.

Can anybody explain why there is half occupation of the up and down levels
instead of just integer occupations as we should expect?

Thanks in advance for your comments.

- Jose Luis

#2 Post by **admin** » Tue Mar 31, 2009 3:44 pm

please show your input files, the result is very strange indeed (have you set NUPDOWN to enfoce such a strange magnetic moment?).

jlrch · #3 Post by **jlrch** » Fri Apr 03, 2009 5:15 pm

Here is the INCAR file:

SYSTEM = Co5 PAW-GGA

LWAVE = .TRUE.
LCHARG = .FALSE.

ISPIN = 2
VOSKOWN = 1
NUPDOWN = 12
MAGMOM = 5*2.400

IBRION = 2
NSW = 250
NELM = 250
NELMDL = 6
EDIFF = 0.0001
EDIFFG = -0.01

PREC = Med
ENMAX = 268
AMIX = 0.2
BMIX = 0.0001
AMIX_MAG = 0.8
BMIX_MAG = 0.0001

ISMEAR = 0
SIGMA = 0.06

NBANDS = 45
RWIGS = 1.302
ADDGRID = .TRUE.
LORBIT = 10

And the POSCAR file:

Co5
1.0
20.082 0.000 0.000
0.000 20.082 0.000
0.000 0.000 20.082
5
Selective dynamics
Cartesian
11.30347 10.23245 9.36778 T T T
9.42876 8.80005 8.54091 T T T
10.65373 11.28243 11.54157 T T T
8.93378 10.92409 9.76298 T T T
9.88648 8.96718 10.99297 T T T