The ABINIT Users Mailing List ( CLOSED )

[<michel.cote@umontreal.ca>]

Dear Xavier,

Thanks for pointing out the relevant places where the matrix elements are 
calculated. I am very interested in adding this feature in Abinit. It seems 
that Andriy Nevidomskyy is also interested. We can discuss it further at the 
Abinit Workshop.

Michel

> -----Message d'origine-----
> De : Xavier Gonze [mailto:gonze@pcpm.ucl.ac.be]
> Envoyé : 21 octobre, 2002 02:16
> À : forum@abinit.org
> Objet : Re: [abinit-forum] electron-phonon coupling
> 
> 
> Dear Michel,
> 
> michel.cote@umontreal.ca wrote:
> > Dear Abinitioners,
> > 
> > Sorry for the previous email, I press the wrong button...
> > 
> > I was wondering if anyone has attented to calculate 
> electron-phonon coupling with Abinit.
> 
> In the linear-reponse part of ABINIT (routine cgwf3.f) it is done :
> 
> H(1)k,k+q | psi(0) k+q > is computed at line 377-378 of 
> cgwf3.f of v3.4.3
> 
> the matrix elements <psi(0) k | H(1)k,k+q | psi(0) k+q >
> are computed line 401-402, and stored in eig1_k .
> 
> The matrix elements eig1_k are output as arguments of cgwf3.f , 
> transmitted to vtowfk3, then to vtorho3, where they are stored,
> for all k points, in eigen1 . Then, eigen1 is tranmitted to scfcv3.f,
> where they are written in the first-order wavefunction files,
> line 499 of outwf.f (formeig=1 in the case of linear response -
> note that outwf is common to unperturbed and perturbed wavefunctions
> outputs, hence the change of notation).
> 
> So, they are available in different places.
> 
> Unfortunately, nobody has yet written a postprocessor to deduce
> things like the electrical conductivity, or the superconductivity
> critical temperature.
> 
> Good continuation,
> Xavier
> 
> 
> 
>