[QE-users] negative phonon frequency when calculate lambda

Lorenzo Paulatto paulatz at gmail.com
Mon Oct 29 09:52:26 CET 2018


Matdyn just interpolates the 2-body force constants to get the dynamical 
matrix at an arbitrary q point and diagonalizes it. It does not care if 
the frequencies are positive or negative, it also does not compute any 
electron-phonon coupling.

kind regards

On 23/10/2018 23:21, Chao Lei wrote:
> Hello, Lorenzo,
> 
> Thanks for your reply, I am using matdyn.x and it already got the lambda 
> vs broadening without NaN or mathematical errors, the negative frequency 
> makes sense for me. So in matdyn.x the code does not do any check on the 
> frequencies too ? or it just neglected the negative frequency when doing 
> the sum ?
> 
> Best wishes
> 
> Chao Lei
> Pos-doc
> The University of Texas at Austin
> 
> On Tue, 16 Oct 2018 at 02:43, Lorenzo Paulatto <paulatz at gmail.com 
> <mailto:paulatz at gmail.com>> wrote:
> 
>     Hello Chao,
>     do you mean, if you can compute lambda, or you have computed it
>     already and you are asking if it makes sense?
>     THe big problem with negative frequencies with this kind of system
>     is that you cannot really impose conservation of energy, but as in
>     elph conservation of energy is maybe not crucial. Anyway, the
>     original lambda.x code does not do any check on the frequencies, so
>     you probably get NaN or mathematical errors if you try to use it.
> 
>     Lambda.f90 is a very simple code, I think you may try to edit it and
>     see if different recipes can give something reasonable (i.e. skip a
>     mode when w2 is negative, use a complex omega=sqrt(w2) but take
>     delta( |e-omega|)...), sum the contribution with a weight
>     exp(-IMAG(omega)/something),....
> 
>     Of course, playing with the code can help thinking, but does not
>     replace it.
> 
>     cheers
> 
>     On Mon, 15 Oct 2018 at 18:22, Chao Lei <leichao.forum at gmail.com
>     <mailto:leichao.forum at gmail.com>> wrote:
> 
>         Dear users,
> 
>         I am calculating electron-phonon using ph.x, which got negative
>         frequency. This is OK for me as it is expected, however, I have
>         a question that how does ph.x treat the negative frequency when
>         use them to calculate the electron-phonon coupling constant
>         lambda ? Thanks for your reply
>           if you know the answer.
> 
>         Best
> 
>         Chao Lei
>         Pos-doc
>         The University of Texas at Austin
>         _______________________________________________
>         users mailing list
>         users at lists.quantum-espresso.org
>         <mailto:users at lists.quantum-espresso.org>
>         https://lists.quantum-espresso.org/mailman/listinfo/users
> 
> 
> 
>     -- 
>     Lorenzo Paulatto - Paris
>     _______________________________________________
>     users mailing list
>     users at lists.quantum-espresso.org
>     <mailto:users at lists.quantum-espresso.org>
>     https://lists.quantum-espresso.org/mailman/listinfo/users
> 
> 
> _______________________________________________
> users mailing list
> users at lists.quantum-espresso.org
> https://lists.quantum-espresso.org/mailman/listinfo/users
> 

-- 
Lorenzo Paulatto - Paris


More information about the users mailing list