Re: Shape of MEP curve

[manishk]

Hi All,

I was performing NEB, but not getting the smooth shape of the mep curve (attached). What could be the possible reason for this?

Here are the settings of my INCAR file

ADDGRID = True
ALGO = Fast
EDIFF = 1e-06
EDIFFG = -0.05
ENCUT = 520.0
IBRION = 1
ISIF = 2
ISMEAR = 0
ISPIN = 2
ISYM = 0
LASPH = True
LDAU = True
LDAUJ = 0 0 0 0
LDAUL = 0 0 2 0
LDAUPRINT = 1
LDAUTYPE = 2
LDAUU = 0 0 4.38 0
LMAXMIX = 4
LORBIT = 11
LREAL = False
LWAVE = False
MAGMOM = 93*0.6
NELM = 100
NELMIN = 3
NSW = 450
POTIM = 0.2
PREC = Accurate
SIGMA = 0.05
IMAGES = 7
SPRING = -5.0
ICHAIN = 0
LCLIMB = .TRUE.
LNEBCELL = .FALSE.
IOPT = 3

I am using vasp version 6.3.2

[graeme]

I would have to see your calculation, but my guess that that you do not have our vtstcode linked into vasp and so it is not extracting the forces at each image. If those forces are zero, the spline will have a derivative of zero at each image. You can grep for VTST in any OUTCAR file to check if the vtstcode is link into your binary.

Another possible reason is that there is net translation and/or rotation along your path. If the distance between images is not consistent with the force at each point, the spline will look weird. The solution to that is to freeze one atom in the same position for each image.

[manishk]

Thanks for the quick reply.

I have checked the OUTCAR of an image and found that the VTST is not linked to VASP.