[DFTB-Plus-User] Dear dftb+ developer, is a multipole expansion scheme available for the charge fluctuation term ?

Peter Yen peter308 at gmail.com
Fri Oct 21 10:26:46 CEST 2016

Dear Mr. Bálint:
Thank you for your reply. In my opinion, I think there are two factors
which might make the partial charges different for DFT and dftb+:

1. The population analysis method.

I have read some articles addressing that Mulliken charge can't capture the
anisotropic feature of the charge. In some special systems, such as gold,
it is manifested with 5d-6s hybridization and an anisotropic feature in
charge may play an important rule in gold clusters, especially for
predicting lowest energy structures. In order to capture this anisotropic
feature, a different population method is necessary, such as Lowdin or

2. higher pole terms are neglected in charge fluctuation term.

    Currently, dipole, quadruple and higher order terms are not included
and those direct and cross interaction terms are therefore not accounted in
the charge fluctuation term, which might affect the total energy as well as
the final partial charge.

It seems to me that since dftb+ always neglected some terms in DFT, the
partial charge can't never be the same with DFT. As a matter of fact, some
papers proposed to use a so called CM3 method to remedy such differences in
partial charge predicted by DFT and dftb+. I would like to know that
whether the two factors i listed above will affect the global structures
predicted by dftb+ ? Thank you sinceriously.

                                                            With Best

2016-10-18 15:13 GMT+08:00 Bálint Aradi <aradi at uni-bremen.de>:

> Dear Peter,
> > Dear DFTB+ users and developers Recently, we have encountered some
> > issues in calculating partial charge in small gold clusters. The
> > charge we obtained seem to deviate systematically with respect to
> > those calculated by DFT. I am wondering if a numerical scheme
> > proposed in this paper ( Phys. Status. solidi B, No.2 ,259-269
> > (2012)) is available, perhaps in the newest version of dftb+, or not
> > ? I would like to learn whether adding higher order terms in addition
> > to a mono-pole charge term can remedy the systematic deviance of
> > partial charge predicted by dftb+ and DFT. Any suggestions or advice
> > are much appreciated.Thanks.
> Yes, we are working on the multipoles and hope to come up with a
> reliably working within a few months. (First results are promising, but
> a few peaces, like forces are still missing.)
> As for the partial charges, you should be careful on comparing DFT and
> DFTB charges. Depending on the charge partitioning scheme you use, you
> would have different values. In DFTB+ (and in DFTB in general) we use
> Mulliken-charges, which are, however, strongly basis dependent, so
> comparing to DFT results could be tricky. One could try to plot total
> charges on a grid and make Bader charge analysis for both DFT and DFTB,
> that would be probably more comparable, as at least you would use the
> same (basis-independent) charge partition scheme in both methods.
>   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

Research Assistant,Physics
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.zfn.uni-bremen.de/pipermail/dftb-plus-user/attachments/20161021/49def447/attachment.html>

More information about the DFTB-Plus-User mailing list