[QE-users] Au GBRV pseudopotential
gianluca.prandini at epfl.ch
Tue May 22 14:59:42 CEST 2018
1) the one you have is the correct pseudopotential (called GBRV-1.2 in the SSSP) that indeed you can find in the authors' webpage:
In the SSSP page we call GBRV-1.2 the full GBRV library for which the versions of the pseudos are <= 1.2.
(In GBRV-1.4 and GBRV-1.5 there are instead the updated versions of the pseudos for some elements)
2) yes, the GBRV pseudopotentials are ultrasoft so we perform the calculations with ecutrho=8 x ecutwfc.
3) in the SSSP we are interested at the convergence w.r.t. the ecutwfc of the cohesive energy rather than at the absolute value. As you rightly notice the value we obtain is different from literature results mainly because of the atom calculation (we use a small supercell size and smearing) but also to the bulk calculation (we use a coarse k-point grid).
Anyway under the section "About SSSP" at https://www.materialscloud.org/discover/sssp/ you can find the computational parameters effectively used to perform the calculations.
PhD student at Theory and Simulation of Materials (THEOS)
École Polytechnique Fédérale de Lausanne (EPFL)
E mail: gianluca.prandini at epfl.ch
Tel: +41 21 69 31040
From: users <users-bounces at lists.quantum-espresso.org> on behalf of Alexandra Davila <davila at theo-physik.uni-kiel.de>
Sent: Tuesday, May 22, 2018 8:12 AM
To: users at lists.quantum-espresso.org
Subject: [QE-users] Au GBRV pseudopotential
I was wondering about the Au GBRV-1.2 pseudopotential used for the
calculation reported in the website https://www.materialscloud.org (very
nice by the way!). I have only found (with which I have been working) Au
GBRV-1.0, I have taken a look also in the GBRV website, but not 1.2.
Maybe am I not looking in the right website?
Also I have noticed that for each pseudopotential it is written dual =
4 or 8, dual is ecutrho=8 x ecutwfc???
And about the cohesive energy for Au, there are some publications (e.g.
DOI: 10.1103/PhysRevB.87.214102) that have obtained for PBE 3.05 eV
(800meV difference to the reported value, but ), the source of this
difference is the atom calculation?? or another functional used??
Thanks for your time,
users mailing list
users at lists.quantum-espresso.org
More information about the users