Page 1 of 1

Problems in Fe clean surface optimization

Posted: Tue Mar 08, 2016 4:43 pm
by luisazofra
Hi All,

I'm trying to run some calculations on Fe metal. However, it seems I've some problems; I assume with the MAGMOM-tag.

Attending to the VASP website, the MAGMOM-tag should be specifically indicated as follows: http://cms.mpi.univie.ac.at/vasp/vasp/MAGMOM_tag.html

i.e. NIONS*1.0 for ISPIN = 2 / OR / 3*NIONS*1.0 for non-collinear magnetic systems

In my case, it's a 80 Fe atoms cell.

The INCAR I've created for this (ionic relaxation, magnetic moment specification, dispersion corrections via DFT-D3 and DFT+U):

System = Fe

# IONIC RELAX
ENCUT = 450
PREC = Medium
EDIFF = 1E-4
EDIFFG = -0.02
NSW = 500
ISIF = 2
ISMEAR = 0
SIGMA = 0.02
IBRION = 2
POTIM = 0.2
LCHARG = .FALSE.
LWAVE = .FALSE.

# Fe SPIN-POL. AND SETTING MAGMON
ICHARG = 2
ISPIN = 2
MAGMON = 80*1.0
#LNONCOLLINEAR=.TRUE.
#LMAXMIX = 4

# DFTD3
IVDW = 11
VDW_RADIUS = 50.2
VDW_CNRADIUS = 20.0
VDW_S6 = 14.0
VDW_S8 = 16.0
VDW_SR = 1.0

# DFT+U
LDAU=.TRUE.
LDAUTYPE = 2
LDAUL = 2
LDAUU = 3
LDAUJ = 0
LDAUPRINT = 1
LASPH=.TRUE.

The OUTCAR stops at:

The magnetic configuration has the point symmetry C_6v.
The point group associated with its full space group is D_6h.

Any help for this? Thanks a lot in advance.

Additional information: PAW Fe_GW 31Mar2010 pseudo-potential in the POTCAR file, 3x3x1 k-points in basis to the Monkhorst Pack scheme, and this POSCAR:

Fe
1.00000
11.46400 0.00000 0.00000
5.73200 9.92800 0.00000
0.00000 0.00000 26.46400
Fe
80
Direct
0.000000000000000 0.000000000000000 0.000000000000000
0.250000000000000 0.000000000000000 0.000000000000000
0.500000000000000 0.000000000000000 0.000000000000000
0.750000000000000 0.000000000000000 0.000000000000000
0.000000000000000 0.250000000000000 0.000000000000000
0.000000000000000 0.500000000000000 0.000000000000000
0.000000000000000 0.750000000000000 0.000000000000000
0.250000000000000 0.250000000000000 0.000000000000000
0.250000000000000 0.500000000000000 0.000000000000000
0.250000000000000 0.750000000000000 0.000000000000000
0.500000000000000 0.250000000000000 0.000000000000000
0.500000000000000 0.500000000000000 0.000000000000000
0.500000000000000 0.750000000000000 0.000000000000000
0.750000000000000 0.250000000000000 0.000000000000000
0.750000000000000 0.500000000000000 0.000000000000000
0.750000000000000 0.750000000000000 0.000000000000000
0.000000000000000 0.000000000000000 0.108298065296252
0.250000000000000 0.000000000000000 0.108298065296252
0.500000000000000 0.000000000000000 0.108298065296252
0.750000000000000 0.000000000000000 0.108298065296252
0.000000000000000 0.250000000000000 0.108298065296252
0.000000000000000 0.500000000000000 0.108298065296252
0.000000000000000 0.750000000000000 0.108298065296252
0.250000000000000 0.250000000000000 0.108298065296252
0.250000000000000 0.500000000000000 0.108298065296252
0.250000000000000 0.750000000000000 0.108298065296252
0.500000000000000 0.250000000000000 0.108298065296252
0.500000000000000 0.500000000000000 0.108298065296252
0.500000000000000 0.750000000000000 0.108298065296252
0.750000000000000 0.250000000000000 0.108298065296252
0.750000000000000 0.500000000000000 0.108298065296252
0.750000000000000 0.750000000000000 0.108298065296252
0.000000000000000 0.000000000000000 0.216596130592503
0.250000000000000 0.000000000000000 0.216596130592503
0.500000000000000 0.000000000000000 0.216596130592503
0.750000000000000 0.000000000000000 0.216596130592503
0.000000000000000 0.250000000000000 0.216596130592503
0.000000000000000 0.500000000000000 0.216596130592503
0.000000000000000 0.750000000000000 0.216596130592503
0.250000000000000 0.250000000000000 0.216596130592503
0.250000000000000 0.500000000000000 0.216596130592503
0.250000000000000 0.750000000000000 0.216596130592503
0.500000000000000 0.250000000000000 0.216596130592503
0.500000000000000 0.500000000000000 0.216596130592503
0.500000000000000 0.750000000000000 0.216596130592503
0.750000000000000 0.250000000000000 0.216596130592503
0.750000000000000 0.500000000000000 0.216596130592503
0.750000000000000 0.750000000000000 0.216596130592503
0.000000000000000 0.000000000000000 0.324894195888755
0.250000000000000 0.000000000000000 0.324894195888755
0.500000000000000 0.000000000000000 0.324894195888755
0.750000000000000 0.000000000000000 0.324894195888755
0.000000000000000 0.250000000000000 0.324894195888755
0.000000000000000 0.500000000000000 0.324894195888755
0.000000000000000 0.750000000000000 0.324894195888755
0.250000000000000 0.250000000000000 0.324894195888755
0.250000000000000 0.500000000000000 0.324894195888755
0.250000000000000 0.750000000000000 0.324894195888755
0.500000000000000 0.250000000000000 0.324894195888755
0.500000000000000 0.500000000000000 0.324894195888755
0.500000000000000 0.750000000000000 0.324894195888755
0.750000000000000 0.250000000000000 0.324894195888755
0.750000000000000 0.500000000000000 0.324894195888755
0.750000000000000 0.750000000000000 0.324894195888755
0.000000000000000 0.000000000000000 0.433192261185006
0.250000000000000 0.000000000000000 0.433192261185006
0.500000000000000 0.000000000000000 0.433192261185006
0.750000000000000 0.000000000000000 0.433192261185006
0.000000000000000 0.250000000000000 0.433192261185006
0.000000000000000 0.500000000000000 0.433192261185006
0.000000000000000 0.750000000000000 0.433192261185006
0.250000000000000 0.250000000000000 0.433192261185006
0.250000000000000 0.500000000000000 0.433192261185006
0.250000000000000 0.750000000000000 0.433192261185006
0.500000000000000 0.250000000000000 0.433192261185006
0.500000000000000 0.500000000000000 0.433192261185006
0.500000000000000 0.750000000000000 0.433192261185006
0.750000000000000 0.250000000000000 0.433192261185006
0.750000000000000 0.500000000000000 0.433192261185006
0.750000000000000 0.750000000000000 0.433192261185006

Re: Problems in Fe clean surface optimization

Posted: Wed Mar 09, 2016 9:48 am
by alex
Hi luisazofra,

the symmetry of atomic and magnetic structure are not the same.

Easy way out: ISYM = 0 (putting the code to P1 symmetry and much more computer time)
Sophisticated way: alter the structure a bit to get a C6v (get rid of the inversion centre)

Cheers,

alex

Re: Problems in Fe clean surface optimization

Posted: Wed Mar 09, 2016 12:41 pm
by luisazofra
Thanks a lot Alex. Could be another solution establish a symmetry D_6h for the magnetic moment as third option and keep the cell as it is?

In positive case, how to define the MAGMOM-tag in D_6h space group?

Cheers,

Re: Problems in Fe clean surface optimization

Posted: Thu Mar 10, 2016 6:28 pm
by alex
Visualize your structure and think a little about it. I'm sure it'll jump right at you! ;-)

Cheers,

alex