Low barriers from CI-NEB calculations

[Marcella Iannuzzi]

 Hi Natalie,

apparently the TS is already at the second bead of the NEB, very close to 
the initial state (the RMSD is the smallest 0.307756). 
This is unusual. Are you sure that the initial state is optimised? What 
happens if you allow the optimisation of the end points during the NEB?

Regards
Marcella



On Monday, July 18, 2016 at 4:22:09 PM UTC+2, Natalie Austin wrote:
>
> Hello,
>
> I've been running CI-NEB calculations on CO2 dissociation on 55-atom NPs. 
> I was able to get a converged NEB but the resulting barrier for 
> dissociation was very small ~0.05 eV. I am a bit unsure about this result 
> considering the TS energies I've come across for CO2 dissociation on 
> transition metals are ~0.2 - 1 eV. My initial and final states were 
> optimized using SCF 1E-8 and forces 4E-5. When the NEB completed I found 
> one single imaginary frequency, which appears to correspond to CO - O 
> dissociation.  I was wondering if anyone has come across the issue where 
> only small barriers were detected but not expected?
>
> I've included the CI-NEB section in below. I've also attached my whole NEB 
> input file and output and the resulting energy file and input file and 
> molden file for the transition state. 
> &MOTION
>   &BAND
>    NPROC_REP 20
>    BAND_TYPE CI-NEB
>    NUMBER_OF_REPLICA 8
>    K_SPRING 0.01
>     &OPTIMIZE_BAND
>       OPT_TYPE DIIS
>       OPTIMIZE_END_POINTS .FALSE.
>       &DIIS
>        MAX_STEPS 1000
>        N_DIIS 7
>       &END
>     &END
>     &REPLICA
>      COORD_FILE_NAME 1.xyz
>     &END REPLICA
>     &REPLICA
>     COORD_FILE_NAME 2.xyz
>     &END REPLICA
>     &REPLICA
>      COORD_FILE_NAME 3.xyz
>     &END REPLICA
>     &REPLICA
>     COORD_FILE_NAME 4.xyz
>     &END REPLICA
>     &REPLICA
>      COORD_FILE_NAME 5.xyz
>     &END REPLICA
>     &REPLICA
>     COORD_FILE_NAME 6.xyz
>     &END REPLICA
>     &REPLICA
>      COORD_FILE_NAME 7.xyz
>     &END REPLICA
>     &REPLICA
>     COORD_FILE_NAME 8.xyz
>     &END REPLICA
>    ALIGN_FRAMES .FALSE.
>    ROTATE_FRAMES .FALSE.
>   &CONVERGENCE_CONTROL
>    MAX_DR    0.002
>    MAX_FORCE 0.002  
>    RMS_DR    0.001
>    RMS_FORCE 0.001
>   &END CONVERGENCE_CONTROL
>
> Thanks, 
>
> Natalie
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20160718/eb75f544/attachment.htm>