[Pw_forum] QE-GIPAW capabilities

[Davide Ceresoli]

Dear Kris,
     the orbital shift for metals is in 5.1 and in the SVN. You can try
it on your systems, and it would be great if you report how slow is
the convergence w.r.t. k-points, and if you find any difference
between the smearing methods.

To appreciate the problem, I've attached the convergence of chemical
shift, and orbital susceptibility of Al-fcc as a function of 1/n_k.
The Fermi-Dirac smearing appears to converge faster (e.g. the f-sum
rule is very close to -3), but the susceptibility (chi_bare_pGv) is
negative or vanishing. The bare and para contribution to the chemical
shift of 27Al are very dependent of smearing. Comparing to PARATEC,
I've found that MV and GAUSS smearing are correct for the orbital shift,
but the susceptibility is wrong by orders of magnitude.

Beware, even if there is only one atom, these calculations run for
several hours!

Good luck and let us know! In the meanwhile I'll work on the Knight
shift.

Best,
     Davide



On 06/20/2014 11:49 PM, Kris Harris wrote:
> I'm interested in some 1H shifts for hydroxides on a metal surface, so the
> orbital terms are hopefully(?) not so influenced by the metallic character,
> while the Knight shift is probably important.  Though I'm not sure if the Knight
> shift calculation runs into the same difficulties with sensitivity to smearing
> method.
>
> There are some other nuclei inside the metal itself that would also likely tell
> me something about the structure.  Getting the orbital shift right seems an open
> question still from the old mailing list comments, can I fool around with
> calculation results versus smearing method in the current (5.1) version or is
> that only with developer versions?
>
> thanks,
> Kris
>
>
> On Fri, Jun 20, 2014 at 11:22 AM, Davide Ceresoli <davide.ceresoli at istm.cnr.it
> <mailto:davide.ceresoli at istm.cnr.it>> wrote:
>
>     Dear Kris,
>           up to now, only the orbital shift is calculated, but I've found
>     very different results depending on the smearing method.
>
>     Regarding the Knight shift, I've been quite busy and I could not
>     overcome the "activation" barrier.
>
>     To the other developers: how about a three days hackathon?
>
>     Best,
>           Davide
>
>
>
>     On 06/19/2014 05:56 PM, Kris Harris wrote:
>      > Hi,
>      >
>      > I don't think this has been asked since the last revision of the code, so
>     I'll
>      > ask for an update: any chance v. 5.1 or the SVN tree can be used for Knight
>      > shifts in metallic systems?
>      >
>      > I've got some interesting experimental shifts in a metallic system, and I'd
>      > really like to use QE-GIPAW to reproduce/understand them and perhaps select
>      > between possible structural models.
>      >
>      > thanks,
>      > Kris
>     _______________________________________________
>     Pw_forum mailing list
>     Pw_forum at pwscf.org <mailto:Pw_forum at pwscf.org>
>     http://pwscf.org/mailman/listinfo/pw_forum
>
>

-- 
+--------------------------------------------------------------+
   Davide Ceresoli
   CNR Institute of Molecular Science and Technology (CNR-ISTM)
   c/o University of Milan, via Golgi 19, 20133 Milan, Italy
   Email: davide.ceresoli at istm.cnr.it
   Phone: +39-02-50314276, +39-347-1001570 (mobile)
   Skype: dceresoli
   Website: http://sites.google.com/site/dceresoli/
+--------------------------------------------------------------+
-------------- next part --------------
A non-text attachment was scrubbed...
Name: NMR-convergence.pdf
Type: application/pdf
Size: 5792 bytes
Desc: not available
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20140623/8a53c264/attachment.pdf>