Spin optimization

[tak]

Dear VASP forum

I am new VASP user (moving from Gaussian to VASP), and learning it by myself. I went through most of Hands-on materials and started doing my own calculations. I am dealing with Fe cluster. I want to optimize spin while fixing cell parameter, and vice versa. Below is the INCAR and POSCAR files I created. Could anyone evaluate if I am doing wight? This type of questions may not be appropriate in this forum, but here is the only place I could post my questions.

Thank you very much.

Tak

INCAR file for spin optimization
SYSTEM = Fe bcc
ISTART = 0
ISMEAR = 2
SIGMA = 0.2
ENCUT = 300
ISPIN = 2
IBRION = 2
ISIF = 0
NSW = 15
EDIFF = 1.0E-05

POSCAR file
bcc Fe
2.866
1.0 0.0 0.0
0.0 1.0 0.0
0.0 0.0 1.0
2
Cartesian
0.0 0.0 0.0
0.5 0.5 0.5

INCAR file for cell parameter optimization while fixing spin
SYSTEM = Fe bcc
ISTART = 0
ISMEAR = 2
SIGMA = 0.2
ENMAX = 300
ISPIN = 2
IBRION = 2
ISIF = 6
NUPDOWN = 4
NSW = 15
EDIFF = 1.0E-05

[forsdan]

1. For the first case there is actually no reason to do any ionic relaxation since the forces for your bcc structure will be zero in this specific case.

2. It's an good idea to set the MAGMOM tag as well. I typically set it to NIONS*3 for bulk Fe.

3. The cutoff energy is bit lower than what I usually typically use for bulk Fe. Have you performed convergence tests wrt to ENCUT, k-point sampling and smearing width? In addition it should say ENCUT in the second case as well and not ENMAX.

4. ENAUG is also a good thing to explicitly define in the INCAR, not necessary but preferable.

5. Finally, one last note about the potential. Do you include the semi-core states for Fe? If you're just performing bulk calculations it will not effect the results to a great extent but for defects and solution energies it can have an impact. Maybe surface energies as well.

Otherwise I think it looks fine.

Best regards,
/Dan
<span class='smallblacktext'>[ Edited Thu Feb 28 2008, 12:47AM ]</span>