[Pw_forum] NRAPP not giving phonon frequencies

Lorenzo Paulatto paulatto at sissa.it
Tue Dec 1 08:44:27 CET 2009


Quoting Bipul Rakshit <bipulrr at gmail.com>:
> The calculation is finished with the following lines
>
>   *   End of self-consistent calculation
>      Convergence has been achieved
>      Stopping because representation    4 is not done*
>
> But in LaMnO3.dyn2 or in the output file, no where it is giving me
> frequencies.
> So can anybody tell me how to get the frequencies of only first three modes,
> rather than calculating all 60 modes?
> What I have to modify in my input file?


Dear Bipul,
it is not possible to compute the frequency of a single mode; each  
mode gives you a line of the dynamical matrix; the matrix is then  
diagonalized and each eigenvalue correspond to a frequency. If you  
only compute the modes corresponding to the first atom you can still  
compute the frequency belonging to the first atom ***but*** under the  
approximation that all the other atoms have infinite masses; i.e. a  
mass much larger than the first one.

This clearly is **not** your case, as Lanthanium is 2.5 times heavier  
than Manganese and almost 9 times heavier than Oxygen. Nevertheless,  
I'll give you the procedure as reference, but don't be tempted to use  
it in your case, unless you really know what you're doing.

After computing the first 3 mode you have the first 3 lines of the  
dynamical matrix. You can take the 3x3 minor of the the matrix (first  
3 rows and first 3 columns) and diagonalize it by hand (using octave,  
matlab or whatever software you prefer). The eigenvalues are the  
frequencies you are looking for in the said approximation. They are in  
Rydberg atomic units, you can get cm-1 multiplying by 109737.

best regards.


-- 
Lorenzo Paulatto (MS)
SISSA & Democritos









----------------------------------------------------------------
   SISSA Webmail https://webmail.sissa.it/
   Powered by Horde http://www.horde.org/




More information about the Pw_forum mailing list