[Pw_forum] how to determine the mass of mixing pseudopotential by virtual.x

[Stefano Baroni]

On Oct 24, 2007, at 4:36 AM, Yanyan wrote:

> Dear Stefano de Gironcoli
>   Recently,  i am testing the virtual-atom pseudopotential  
> generated by virtual.x.
> I want to ask if the testing results of the virtual pseudo made by  
> 0% pseudo1 + 100% pseudo2
> are in same agreement with the results calculated directly with  
> pseudo2.

They should, as far as I understand, but others may have something  
more to add.

> If both
> values are different, only the convergence trend is equal. Is the  
> virtual-pseudo OK?
> While i want to calculate phonon frequencies, how can I do to  
> obtain the mass
> of virtual-atom pseudopotential. Because of the mixing  
> pseudopotential as (1-x)
> pseudopotential1 + x pseudopotential2, whether the mass of virtual- 
> atom
> pseudopotential can be determined by (1-x) mass of potential1 + x  
> mass of potential2.

As already said in this forum a few days ago, it depends on the  
application. To be (a little bit) more specific: the weighted mass  
gives reasonable results for acoustic phonons, where atoms of  
different mass vibrate in phase. For the optic branches, instead,  
different atoms vibrate out of phase, so that taking the average mass  
is only meaningful when the optic branches of the pure compounds  
overlap considerably (which in turn implies that the masses are close  
enough). Beware, though, that the time-consuming part of any phonon  
calculation is the computation of the interatomic force constants,  
which do not depend on mass. Once these calculated, the calculation  
and diagonalization of the dynamical matrices is a relatively minor  
task. It may make sense therefore to calculate the interatomic force  
constants using the VCA, and then to calculate the dynamical matrices  
using real masses. This procedure has been followed successfully  
since the eve of DFPT. Examples of its application can be found,  
e.g., in Rev. Mod. Phys.  73, 515 (2001).

Hope this helps.
Stefano (Baroni)

> Thank you for your advice.
> Best regards
> Junjie Wang
>
> 雅虎邮箱，终生伙伴！
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum

---
Stefano Baroni - SISSA  &  DEMOCRITOS National Simulation Center -  
Trieste
[+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype)

Please, if possible, don't  send me MS Word or PowerPoint attachments
Why? See:  http://www.gnu.org/philosophy/no-word-attachments.html



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20071024/99d33808/attachment.html>