[DFTB-Plus-User] Hubbard Derivs

Bálint Aradi aradi at uni-bremen.de
Fri Dec 23 12:34:41 CET 2016

Dear Nagaprasad,

> <http://www.dftb.org/fileadmin/DFTB/public/misc/slakoformat.pdf>).
> From that, the Hubbard values are in the skf file, line 2, 5th, 6th,
> 7th value for d, p and s angular momentum. Let us take an example
> from autotest, scc/C2H6_3rdfull. The HubbardDerivs are taken as H =
> -0.16 and C = -0.16, and they are using  mio-0-1 parametrization for
> that calculation. When we see the values in scc/C2H6_3rdfull-damp,
> the HubbardDerivs are different even using the same mio-0-1
> parametrization . My questions are 1) I didn't find those
> HubbardDerivs (H = -0.16  and C = -0.16 and with dampXH  C = -0.1492
> and H = -0.1857) in the skf files, how can we choose them? 2) Why the
> HubbardDerivs are different with dampXH? 3) How we choose
> DampXHExponent?

First of all, please never take any tests from the autotest suite as
template for your calculations. The autotest suite is there to test all
various features, sometimes even with completely arbitrary unphyiscal
and inconsisten input values. It is only meant for developers and
sysadmins, so that they can check that they DFTB+ build works as expected.

As for the HubbardU derivatives. Currently, the 3ob set is the only one,
which was created with DFTB3 (third order DFTB + damping of the X-H
interactions). So, you should only use those features with that set
only. The Hubbard U values and the damping exponents for that set are
given in the accompanying README file. For all other sets, use
conventional SCC (2nd order) DFTB without any X-H damping.

  Best regards,


Dr. Bálint Aradi
Bremen Center for Computational Materials Science, University of Bremen

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 836 bytes
Desc: OpenPGP digital signature
URL: <http://mailman.zfn.uni-bremen.de/pipermail/dftb-plus-user/attachments/20161223/19b03d06/attachment.sig>

More information about the DFTB-Plus-User mailing list