[Pw_forum] Error in Ph.x for calculation Raman spectrum

[Ari P Seitsonen]

Dear LIANG Xiongyi,

   Some answers:

  - Yes, one needs fixed occupations, because the implementation in PH is 
for non-resonant Raman scattering; the formalism in metallic systems is 
much more involved, there are some recent developments in the Quantum 
ESPRESSO Community at least by Michele Lazzeri and Francesco Mauri

  - About LDA, I somehow thought that the restriction is lifted in recent 
versions (or, one just neglects the error message by commenting it 
out...). But I do not know whether for example hybrid functionals would 
work/give reasonable results, GGA yes

  - If the problem appears in the self-consistent calculation, the problem 
is not in the calculation of Raman spectrum. Without more details about 
your system (input, pseudo potentials) we can only guess for what you are 
calculating, it is difficult to help, sorry

  - First one should of course achieve self-consistency, in principle 
having an odd number of electrons is not a problem. Whether the phonon 
code works with spin-polarised systems, sorry, I do not know but I would 
guess so

     Greetings,

        apsi

-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-
   Ari Paavo Seitsonen / Ari.P.Seitsonen at iki.fi / http://www.iki.fi/~apsi/
     Ecole Normale Supérieure (ENS), Département de Chimie, Paris
     Mobile (F) : +33 789 37 24 25    (CH) : +41 79 71 90 935


On Wed, 8 Mar 2017, LEUNG Clarence wrote:

> 
> Dear QE users,
> 
> 
> As far as I know, when using ph.x to calculate raman spectrum. We should use fixed occupation and LDA in scf, right?
> 
> 
> However a error occur (charge is wrong, smearing is needed ) when I use fixed occupation in scf calculation because of my system is odd electrons.
> 
> 
> what should I do to calculate the raman spectrum of odd electron system?
> 
> 
> Thanks in advanced!
> 
> 
> Best regards,
> 
> LIANG Xiongyi
> 
> Department of Physics and Materials Science
> 
> City University of Hong Kong
> 
> 
>