[QE-users] HSE XC - mismatch between energy- and force-minimizing bond length for H2 and O2

[Tobias Kloeffel]

Dear Robert Wexler,

there are some additional force terms when using Vanderbilt Ultrasoft 
pseudopotentials in HFX calculations. These are not implemented, and 
using a more recent QE version will actually stop, because these force 
terms are missing.

Regards,

Tobias Klöffel

On 5/13/19 5:33 PM, Robert Wexler wrote:
> Dear QE users,
>
> I obtained some perplexing results using the HSE XC functional that I 
> would like to share with you.
>
> I calculated the ground state potential energy surfaces of H2 and O2 
> as a function of bond length strain (i.e. % different from some 
> arbitrary starting bond length). The results are attached in this 
> email as "h2-compare-force-energy.pdf" and 
> "o2-compare-force-energy.pdf", respectively. The right panel for H2 
> and O2 shows the total energy as a function of bond length strain. As 
> you can see, a slight contraction of the original bond length is 
> preferred. If you look at the left panel, however, the total force is 
> minimized at positive strains of greater magnitude. The minimum force 
> and total energy should occur at the same bond length strain, which 
> could present some problems for geometry optimization. I have attached 
> inputs and outputs for each calculation in .tar.gz files 
> ("h2-qe-in-out.tar.gz" and "o2-qe-in-out.tar.gz").
>
> I took this one step further and tried to relax a water molecule using 
> HSE and found that the relaxation does not converge. This can be seen 
> in "h2o-hse-relax.pdf". As the BFGS proceeds, it cannot converge 
> because one configuration decreases the total energy (steps 1, 4-12, 
> 15-16) and another decreases the total force (steps 2-3, 13-14).
>
> Here is a quick summary of my computational resource and QE settings:
> - Cori, NERSC, KNL
> - QE 6.3
> - GBRV PSPs
> - SCF calculation
> - 35/350 Ry cutoffs
> - LSDA for O2 (I get the correct total/absolute spin magnetization of 
> the O2 molecule)
> - HSE, 1x1x1 q-grid
> - Assume isolated, "mt"
> - Gamma point
> - More details can be found in the inputs and outputs I've supplied
>
> Thank you in advance for any help you can provide.
>
> Best,
> Rob
> ---------------------------------------------
> Robert B. Wexler, PhD
> University of Pennsylvania, 2019
> rwexler at sas.upenn.edu <mailto:rwexler at sas.upenn.edu>
> (215) 801-8741
> ---------------------------------------------
>
>
> _______________________________________________
> Quantum Espresso is supported by MaX (www.max-centre.eu/quantum-espresso)
> users mailing list users at lists.quantum-espresso.org
> https://lists.quantum-espresso.org/mailman/listinfo/users


-- 
M.Sc. Tobias Klöffel
=======================================================
Interdisciplinary Center for Molecular Materials (ICMM)
and Computer-Chemistry-Center (CCC)
Department Chemie und Pharmazie
Friedrich-Alexander-Universität Erlangen-Nürnberg
Nägelsbachstr. 25
D-91052 Erlangen, Germany

Room: 2.305
Phone: +49 (0) 9131 / 85 - 20423
Fax: +49 (0) 9131 / 85 - 26565

=======================================================

E-mail: tobias.kloeffel at fau.de

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20190513/6fe483ae/attachment.html>