Why the optimized structure is different with those of literature ??

[ilmarekw]

Hi,

My question is on the structure optimization including unit cell and atoms.

I'm optimizing Na2CO3 crystal structure with the same potential function given in literature (PW91-GGA). I did simulation several time sequentially to relax the structures. However, the final converged structure does not same with that in literature.

My incar file is below:

general:
PREC = Accurate
IBRION = 2
ISIF = 3
NSW = 20
NELMIN = 5
EDIFF = 1.0e-08
EDIFFG = -1.0e-08
IALGO = 38
ISMEAR = 0; SIGMA = 0.1
LREAL = .FALSE.
ADDGRID = .TRUE.
LWAVE = .FALSE.
LCHARG = .FLASE.

-------------

I have not an idea how to solve this..

Any reply would be gratefully thankful.

[admin]

please
1) do the relaxation in steps of 3-5 ionic steps (NSW) and delete the WAVECAR files between the runs (in particular if you started with a bad guess of the the cell )
2) increase ENCUT to at least 1.3*max(ENMAX) of POTCAR
3) you may need to use the hard PPs if the CO3 bond-lenghts are small
4) check if EACH of the ionic steps is converged electronically
5) EDIFFG = -1e-08 is too strict, 1e-02 should be sufficient
6) note that there may be a general discrepancy with experiment (underbinding) if you use PW91 PPs.