DOS analysis for metalorganic material including metal

[luke419]

I'd like to obtain DOS (density of state) of metallorganic material including Zn.
The unit cell is 17.141 Ang * 12.112 Ang * 12.112 Ang with angle of 60.05 45.06 45.06.
It is supposed that it has experimental energy gap of 3.5 eV (2.7 eV when calculated using LDA) or less.

I use ENCUT 520 eV since zinc is included.
KPINTS mesh of 2X3X3 is used preliminarily.
The sytem is composed o 46 atoms.

What should I write keywords and their parameter in DOS part of INCAR file for this job?
That is, I'd like to know what is recommandable contents (standard method) in DOS related part in INCAR file for it.
For instance, what values should I assign for "LVTOT, LCHARG, LWAVE, NBANDS, NEDOS, SMEAR, SIMGA, LORBTT, RWIGS, EMIN, EMAX etc."?

With best regards,


Luke

[alex]

Hi Luke419,

you may try an INCAR like:

Electronic minimisation
GGA = PE
PREC = Accu
EDIFFG = -0.01
ENCUT = 400.
NELMIN = 4
LREAL = A

DOS related values
SIGMA = 0.1
ISMEAR = 0 ! Gaussian smearing for big cells

Do not write huge files
LCHARG = .FALSE.
LWAVE = .FALSE.

Performance options
NPAR = 1 ! adjust to number of boxes/processors
ALGO = V

in case you want to optimize your structure, just add
Ionic relaxation
NSW = 50 number of steps for IOM
IBRION = 2 ionic relax: 0-MD 1-quasi-New 2-CG
ISIF = 0 ionic relaxation + stress tensor calcn.

You may play with other variables you've mentioned in your posting, but normally there is no need to change except for very special needs. The above INCAR is my standard set for many oxygen containing compounds and mostly it works without changing too much of the standard settings in VASP.

I rather doubt you'll need 520 eV for Zn. Normally oxygen determines the size of the basis set (likee.g. 400eV) ...
Also I'd suggest you check your primitive cell. 45° for beta and gamma look a little strange. A Gamma point calculation is very likely to be sufficient for your MOF. I'd assume it is an insulator ...

cheers

alex

[luke419]

Thanks for your answer really.
Can I use ISMEAR 0 and SIGMA 0.1 for semiconcucting material?
It seems that ISMEAR -5 requires much computational costs but gives more accurate result.
Is it right?

[alex]

Hmm,

why not to use your suggestion?
But how yould you define 'accurate' in your ISMEAR=-5 case?
If you have a semiconducting system you may wish to check to increase the KPOINTS until convergence of whatever property you are interested in.

Cheers

alex