My Community

Hi Dear
I have a question about magnetic moment. I'm calculating DMS with 38 atoms, where 1 atom is Mn. In first step I did relaxation of my structure, and now I want to calculate magnetic moment, but I don't know what I should write to my INCAR file. I've read manual of Vasp, but there is information only about MAGMOM. Should I write only this?:
ICHARG =1
ISPIN = 2 and
MAGMOM = 18*0 19*0 1*5
or something else?
I want also analyse magnetic order of my system with 2 atoms Mn. What I should write in INCAR file in case of FM or AFM ordering?
Please help.
Thank you
Margo

I personally thinks it very helpful when I analyze the magnetic order to include the LORBIT tag in the INCAR:

http://cms.mpi.univie.ac.at/vasp/vasp/node129.html

In this case you get the local magnetic moment for each atom written to the OUTCAR, which can be good in order to get a feeling of the magnetic structure. The magnetization is of course ambiguous and will depend on the projection radius, but still usually provides a guideline.

If you would like to investigate the FM and AFM states it's really recommended that you set the MAGMOM term so that you have the corresponding initial magnetic struture, otherwise the solution can sometimes relax to another local equilibrium, e.g. the nonmagnetic solution. The experimental (or known calculated) moments, for each respective state, multiplied with a factor of 1.3 or 1.5 is typically a good choice, but it may vary depending on your system.

Cheers
/Dan Fors



<span class='smallblacktext'>[ Edited Fri Sep 19 2008, 08:09AM ]</span>

Thank you for answer.
I have another question.
In first step I relaxed my structure and in the second step I calculate magnetic moment.
When I'm calculating magnetic moment should I still relax my structure or no?
This is my INCAR file:

SYSTEM = InMnAs
ISTART = 1
LCHARG = .TRUE.
LWAVE = .TRUE.
ALGO = Fast
ISPIN = 2
ICHARG = 1
LREAL = Auto
PREC = High
MAGMOM = 18*0 19*0 1*6
LORBIT = 2

ISIF = 2
NSW = 0
IBRION = -1
EDIFFG = -0.01

ISMEAR = 0
SIGMA = 0.05

LPLANE = .TRUE.
NPAR = 1
NSIM = 4

Is this file correct or no?
Margo

the structure might slightly change for spin-polarized structures.
therefore set NSW > 0 and do some more relaxation steps (IBRION>0)
if you choose LORBIT=2, you have to give RWIGS explicitely