My Community

Hi all,

I'm running several slab optimizations (with vasp.5.3.2) modifying a bond distance of an adsorbed molecule and inheriting the WAVECAR file of the previous calculation. In these calculations and I have found that in many cases the OUTCAR file ends up with the message:

"reached required accuracy - stopping structural energy minimisation"

but only after doing 1 ionic step (when the default is NELMIN= 2) and without reaching the EDIFFG criteria. For example:

1 F= -.29207540E+03 E0= -.29207282E+03 d E =-.292075E+03 mag= 0.0000

I have checked the OSZICAR files in these cases and some of them have achieved the EDIFF criteria and some do not.
I have never found this problem in previous calculations. My INCAR file looks like:

ISTART = 2
ISPIN = 2
NUPDOWN= 0
GGA = PE

# ==> Electronic Relaxation <== #
ENCUT = 500
EDIFF = 1E-7
EDIFFG = 1E-4
LREAL = Auto
NELM = 250

LMAXMIX= 6
INIMIX = 0
AMIX = 0.2
BMIX = 0.0001
AMIX_MAG= 0.8
BMIX_MAG= 0.0001

# ==> Ionic Relaxation <== #
NSW = 400
IBRION = 1
POTIM = 0.05

ISMEAR = 0
SIGMA = 0.05

# ==> Other Parameters <== #
LASPH = .TRUE.
LDAU = .TRUE.
LDAUL = 3 -1 -1
LDAUU = 5.5 0.0 0.0
LDAUJ = 1.0 0.0 0.0
NPAR = 4

LDIPOL = .TRUE.
IDIPOL = 3

I have tried to run an optimization of one of these problematic calculations without reading a WAVECAR file and I get exactly the same result. I have also tried to change the EDIFFG criteria to a force criteria (EDIFFG= -0.02) but without result. It's very weird because it seems to happen randomly. Does anyone know what can be going on?

Thanks in advance!

When your slab is relaxed and there is only weak interaction between the slab and the molecule the whole system fulfill the stopping criteria just at the beginning of the calculation. Try starting from the non-relaxed slab.