about transition state and frequence

[fegg7502]

Dear all.
i have a problem when i calculate the frequence and transition state. There are always several f/i when i calculate the frequence with a relaxed
structure, while there is no f/i when i calculate the frequence with the NEB transition state result.

the incars are as listed here:
SYSTEM =
ENCUT = 400
ISTART = 0 ; ICHARG=2

ISMEAR = 2 ; SIGMA = 0.2
ISIF = 2

IBRION = 5
POTIM = 0.024
NFREE = 2
NSW =400
PREC = HIGH
EDIFF = 0.000001




SYSTEM =

ISTART = 0 ; ICHARG=2

ISMEAR = 2 ; SIGMA = 0.2

ISPIN = 2

IMAGES = 4

SPRING = -5

IALGO = 48

ENCUT = 400

dynamic:

IBRION = 1

NSW = 400

NFREE = 2

POTIM = 0.2

[admin]

I suppose that you use INCAR1for both calculations of the vibrational spectrum, and INCAR2 for the NEB run only, do you?
please check:
-- does the TS really correspond to the geometry with the maximum total energy along the reaction path?
-- are all displacement steps (for both, ground and transition state) converged electronically up to 10-6 or did some of the steps simply finish after 60 electronic scf steps without being actually converged
-- usually, POTIM should even be chosen smaller in INCAR1 (0.01 to 0.015) to make sure that you are in the harmonic limit

[fegg7502]

the TS really correspond to the geometry with the maximum total energy along the reaction path

all displacement steps (for both, ground and transition state) converged electronically up to the defult value

[camilo]

Isn't there some concern that basis set incompleteness makes a small POTIM a bad idea?