[DFTB-Plus-User] How to effect a non-scc dftb calculation, it the current SK parameter compatible ?

[Peter Yen]

Dear Mr. Balint
Thank you for your kind reply. You have mention that the repulsive energy
(RP) is fitted to bulk system of gold and since crystal gold will not have
charge transfer or nearly zero charge transfer, the RP fitted will be
similar to the non-scc case. However, in the case of gold clusters, there
are in fact non zero charge transfer among atoms. And i believe the Rp in
this case will be quite different to the bulk case since  charge transfer
now occurs in the former. In these two extreme cases ( bulk gold and gold
clusters Au_n, where n is less than 10 ), do you think two separate RPs is
necessary ? Because the Rp in Au-Au.skf is actually fitted to system with
nearly zero charge transfer, yet in small clusters case, the charge
transfer can't be neglected. I was speculating maybe  the Rps will be quite
different in these two cases. Would you provide any comments or advice ?
Many thanks.


                                                      With Best Regards
                                                      Peter




2017-03-24 10:39 GMT+08:00 Bálint Aradi <aradi at uni-bremen.de>:

> Dear Peter,
>
> > I would like to carry out a non-scc dftb calculation. But i have one
> > part not clear and need your advice. Say for example, the gold case.
> > The SK parameter for gold in auorg folder is Au-Au.skf. Do i just
> > simply set the SCC=no option  in the  script for my non-scc
> > calculation? One part to pay attention to is that the repulsive
> > energy in Au-Au.skf is fitted by assuming "band structure" and
> > "charge fluctuation energy" existed and both terms are subtracted
> > with respect to DFT energy during the fitting process of the
> > repulsive energy. If now i just turn off " charge fluctuation energy
> > " by setting "SCC=no", will the repulsive energy i use in Au-Au.skf
> > not compatible to my non-scc calculation ? Shouldn't it be more
> > reasonable to have a separate repulsive energy profile for the
> > non-scc case ? Does the admin have this non-scc repulsive energy
> > profile that i can request from ?  Thank for your suggestions!
>
> You are correct, if you set SCC=no, the repulsives in the Auorg set
> (which where made with SCC = Yes) will not be correct. But depending on
> the amount of charge transfer between the atoms in your system, you may
> still get qualitatively correct forces this way, so you could prerelax
> structures in non-scc mode and refine them with scc at the end. (This is
> mostly what non-scc is used for.)
>
> We usually do not bother to create a non-scc repulsive, as it is not
> really useful. Either your system has not much charge transfer (e.g. a
> slightly distorted Au supercell), then scc and non-scc repulsives would
> be very similar. Or your system has a substantial charge transfer (e.g.
> some defect in gold, or some molecule containing one Au atom), but then
> making a non-scc calculation would give a wrong Hamiltonian (one not
> considering this charge transfer), so fitting a repulsive with that
> unphysical non-scc Hamiltonain would yield a non-physical and therefore
> non-transferable repulsive.
>
> Best regards,
>
> Bálint
>
> --
> Dr. Bálint Aradi
> Bremen Center for Computational Materials Science, University of Bremen
> http://www.bccms.uni-bremen.de/cms/people/b-aradi/
>
>
>
> _______________________________________________
> DFTB-Plus-User mailing list
> DFTB-Plus-User at mailman.zfn.uni-bremen.de
> https://mailman.zfn.uni-bremen.de/cgi-bin/mailman/listinfo/dftb-plus-user
>



-- 
Postdoctoral fellow, Physics department,
National Central University, Taiwan(R.O.C)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.zfn.uni-bremen.de/pipermail/dftb-plus-user/attachments/20170324/73c8d5b7/attachment-0001.html>