CI-NEB: nearly there but not quite converged

[Jörg Saßmannshausen]

Dear all,

I am battling a problem with  some CI-NEB calculations I am doing. 
They are nearly converged but not quite:

 RMS DISPLACEMENT =   0.00010     [ 0.00010]                               (YES)
 MAX DISPLACEMENT =   0.00051     [ 0.00020]                               ( NO)
 RMS FORCE        =   0.00019     [ 0.00050]                               (YES)
 MAX FORCE        =   0.00102     [ 0.00100]                               ( NO)

If I do a frequency calculation on this geometry for the transition state I usually find the imaginary frequency of the 
transition state but also one or two more which corespond to say a rotation of a Me group or a rotation of a Cp ring of 
my metallocene. 

Now, I could say: I ignore it but I really would like to get that done 'properly'.

I am already using quite tight parameters, for example:

EPS_SCF 1.0E-8
EPS_DEFAULT 1.0E-16

I am doing a PBE-D3/DZVP level of theory calculations. 

Also, if I am selecting such a step for a restart, it seems to jump off too far:

 RMS DISPLACEMENT =   0.00130     [ 0.00010]                               ( NO)
 MAX DISPLACEMENT =   0.00893     [ 0.00020]                               ( NO)
 RMS FORCE        =   0.00253     [ 0.00050]                               ( NO)
 MAX FORCE        =   0.01721     [ 0.00100]                               ( NO)

These are the parameters I am using:

    BAND_TYPE CI-NEB
    K_SPRING 0.02
    ROTATE_FRAMES T 
    &CI_NEB
       NSTEPS_IT  5
    &END

and 

    &OPTIMIZE_BAND
      OPTIMIZE_END_POINTS T
      OPT_TYPE DIIS
      &DIIS
       MAX_STEPS 600
       N_DIIS 7
       NO_LS
       STEPSIZE 0.5
       MAX_STEPSIZE 1.0
      &END

I am happy to post the input,the force_eval and subsys files if required.

Does anybody have some good advice here?

All the best from London

Jörg

-- 
*************************************************************
Dr. Jörg Saßmannshausen, MRSC
University College London
Department of Chemistry
20 Gordon Street
London
WC1H 0AJ 

email: j.sassma... at ucl.ac.uk
web: http://sassy.formativ.net

Please avoid sending me Word or PowerPoint attachments.
See http://www.gnu.org/philosophy/no-word-attachments.html
-------------- next part --------------
Q29udGVudC1UeXBlOiBhcHBsaWNhdGlvbi9wZ3Atc2lnbmF0dXJlOyBuYW1lPXNpZ25hdHVyZS5h
c2MgDQpDb250ZW50LURlc2NyaXB0aW9uOiBUaGlzIGlzIGEgZGlnaXRhbGx5IHNpZ25lZCBtZXNz
YWdlIHBhcnQuDQoNCi0tLS0tQkVHSU4gUEdQIFNJR05BVFVSRS0tLS0tDQpWZXJzaW9uOiBHbnVQ
RyB2MS40LjEyIChHTlUvTGludXgpDQoNCmlGNEVBQkVJQUFZRkFsa0dYZmdBQ2drUVRDbi9MeTlT
L3djR3JBRC9SZXFjZzN4RDdBWVhCWWl6aDNVM1ZBYzcNCmdiWlIyTmhqUm5uV0FselpRUVVBLzB5
RmViUDRMUExaeTd1ZjBFaDZOWUsyL3E1NS9NUmJhU002a091WHpaS0cNCj1Xd1p1DQotLS0tLUVO
RCBQR1AgU0lHTkFUVVJFLS0tLS0NCg==