Hi Lucas,

Normally GW is done on top of KS-LDA. You want to do GW on top of hybrid 
functionals whose validity needs to be carefully checked, because of double 
counting in the two approximations. I am not aware of the WEST code maybe their 
GW implementation already takes into account the exchange included in different 
hybrid functionals like HSE06. If not this might explain the discrepancies in 
your benchmark.

Cheers,

Sky

--------------------------------------------------------------------------------------------------
Jianqiang (Sky) ZHOU
European Theoretical Spectroscopy Facility
Institut des NanoSciences de Paris (INSP)
Sorbonne Université - Case 840 - 4 place Jussieu
Barre 2232, étage 2, pièce 11
75005 PARIS
http://etsf.polytechnique.fr/People/Sky<https://theory.polytechnique.fr/squirrelmail/src/compose.php?send_to=gaelle.bruant%40polytechnique.edu>
tel : +33 (0)1 69 33 44 85
________________________________
发件人: users <[email protected]> 代表 Lucas Nicolás Lodeiro 
Moraga <[email protected]>
发送时间: 2019年2月28日 15:44
收件人: [email protected]
主题: [QE-users] WEST CODE 3.0

Hello community.

I want to do a G0W0 calculation over HSE06 and PBE0 calculation, with WEST code 
(version 3.0), but I'm facing some difficulties and discrepancies.

After I have done a G0W0 over PBEsol and PBE calculation over the same 
structure (system) and the required time to complete the calculation it is 
around 24 hrs, but in hybrid case the time is absurdly greater. Is this normal?.
On the other hand, I have done a benchmark for an simple system, and i found 
discrepances in the quasi-particle energy (about 0.1 and 1.0 eV) between Gamma 
and GammaComplex (Kpoints = Gamma and Kpoints = Automatic 1 1 1 0 0 0 
respectively).

I read the Govoni's papers, and send to him some doubts, but he did not answer 
to me, for that I ask here. There are somebody with experience in WEST code?
I need someone with whom I can talk about these issues and give feedback.

Regards!
_______________________________________________
users mailing list
[email protected]
https://lists.quantum-espresso.org/mailman/listinfo/users

Reply via email to