[Pw_forum] Phonon frequency: nat_todo option

[stefano de gironcoli]

Dear Vikan,
I dont have direct experience on this problem. here a few things i would 
look into:

from your NEB calculation around the TS you should be able to see what 
the unstable mode is expected to look like and how many atoms are involved.
If you compute interatomic force constants (IFC) only around a subset of 
the atoms there is the problem of how you impose the acoustic sum rule.
One way would be to complete your DynMat using the IFC of the bulk 
material (that you could compute separately), and impose ASR on this model.
To stay on the safe side you could set the mass of the atoms that you 
did not compute to a very large number so they do not move.

stefano


On 21/10/2016 17:38, Vikan Manmathan wrote:
> Dear Developers and Experts,
>
> I am interested in studying diffusion of metal clusters on oxide 
> support. As a first step, I took a simplest case (adatom hopping) and 
> obtained the transition state using CI-NEB method.  I want to find the 
> rate constant for the above, I need to know the normal mode 
> vibrational frequencies at initial and transition states.
>
> So I calculated the phonon frequencies using frozen atomic coordinate 
> (nat_todo option only on metal adatom) since the system size is very 
> large.  I expected the transition state to have one less real 
> frequencies (3N-1) relative to initial state (3N). However, I am 
> getting 3N real frequencies for the transition state (correct me if I 
> am wrong).
>
>  I have also checked it by including more atoms in the linear response 
> (up to 33 atoms) but still I got 3N real frequencies for the TS.
>
> I have attached the input files and the output file generated by dynmat.x
>
> Any suggestion will be highly appreciated.
>
> Thanks in advance.
>
> regards,
>
> *NANDHA KUMAR V.*
> Research scholar,
> IISER, PUNE.
>
>
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://pwscf.org/mailman/listinfo/pw_forum


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20161022/fbf01064/attachment.html>