forum@abinit.org

Subject: The ABINIT Users Mailing List ( CLOSED )

List archive

Re: [abinit-forum] The frequency-dependent dielectric tensor

From: Xavier Gonze <xavier.gonze@uclouvain.be>
To: forum@abinit.org
Subject: Re: [abinit-forum] The frequency-dependent dielectric tensor
Date: Fri, 23 Jan 2009 02:06:48 +0100
Domainkey-signature: a=rsa-sha1; c=nofws; d=uclouvain.be; h=message-id:from: to:in-reply-to:content-type:content-transfer-encoding: mime-version:subject:date:references; q=dns; s=selucl; b=nf4MJxn lR0REf8cgz8lyAlWaSfSzttC+oyrhIE/k960a47gMyjjNBjHL8EsD3Z0j4sql6+l OsjzyyzeRwwDa4lY+WcjO0mJEP+d1hB5qbdvluEvznXh72gAPhGqcQVoxMgjCg8v jN1TWQLnj4/6c/TJFl8fNzwZFfwl0YsunC94=

Dear Andrey,

On 21 Jan 2009, at 02:56, Andrey Chibisov wrote:

Hello!
How to calculate the frequency-dependent dielectric tensor in the Response-function calculation?

You mean, with the rfelfd input variable ?
If so, the answer is that it is not implemented. rfelfd only drives response with respect to static calculations.

On the other hand, the computation of the independent-particle frequency-dependent dielectric tensor (well optical response),
is possible using the "optics" utility, or the "conducti" utility, both distributed in the ABINIT package.
At the level of the more elaborate RPA, such a data is "hidden" in the GW
part of ABINIT, but there, you need to extract from the code, and this is not easy. If you want RPA or more
elasophisticated approximations (i.e. including excitons), you have to continue an ABINIT calculation using the DP or EXC
software applications.
See the links at http://www.abinit.org/community/?text=links

Wish you the best,
Xavier

[abinit-forum] The frequency-dependent dielectric tensor, Andrey Chibisov, 01/21/2009
- Re: [abinit-forum] The frequency-dependent dielectric tensor, Xavier Gonze, 01/23/2009
  - Re: [abinit-forum] The frequency-dependent dielectric tensor, Andrey Chibisov, 01/23/2009