[CP2K-user] [CP2K:13550] IC-QMMM with single charge in front of a metal plane: open boundary corrections

[Dorothea Golze]

Dear Katharina,

I just ran a quick test for the EXT_POTENTIAL keyword. System: small
organic molecule on top of gold (charge-neutral periodic system using
standard periodic solver).  The keyword still seems to work as intended:
when preparing a plane-averaged plot in z-direction, the potential is not
zero in the region of the metal, but the value defined by EXT_POTENTIAL. I
can send you my example input files; feel free to contact me directly (
doroth... at aalto.fi).

Best regards,
Dorothea

Am Fr., 26. Juni 2020 um 17:00 Uhr schrieb Dorothea Golze <
doroth... at googlemail.com>:

> Dear Katharina,
>
> in short, it is very likely no bug; I think the basic problem is that you
> try to run a periodic DFT calculation for a charged system.
> The constant potential condition is a hard constraint for the average. I
> tested with the standard periodic solver (for periodic systems) and the MT
> solver (for a cluster-type of approach).
> I don't have it in the paper, but in my thesis, I have an example, where I
> placed an ion above a slab (Figure 3.7 page 31,
> https://www.zora.uzh.ch/id/eprint/116638/)
> For the MT solver, I have the suspicion you are not applying this solver
> correctly, which can lead to technical artifacts. You need lots of vacuum
> around your system since you also specified a periodic cell, this seems not
> to be the case. I suggest to test the MT solver for a single molecule (DFT
> only) first to get a feeling how much vacuum you need, i.e., to converge
> out the cell size and how this solver behaves.
>
> Best regards,
> Dorothea
>
>
> Am Fr., 26. Juni 2020 um 16:32 Uhr schrieb Katharina Doblhoff-Dier <
> k.doblh... at lic.leidenuniv.nl>:
>
>> Dear Dorothea,
>> To break a complex problem down: Should I consider it a bug, that I get
>> non-constant potentials in the plot of the hartree energies, when using MT
>> or implicit BCs and that potential in the metal does not match
>> EXT_POTENTIAL? (see plot below - also posted in the original question) Or
>> is there a good reason for this behavior?
>>
>> red:      IMPLICIT poissn solver with Neumann BC at z=0 and z=60 and
>>>> homogeneous Dirichlet BC at z=30
>>>> green:  MT (Martyna-Tuckerman) poisson solver (shifted by 0.4V)
>>>> blue:    normal Ewald summation (shifted by 0.4V)
>>>>
>>>> [image: tmp3.png]
>>>>
>>>>
>>>>
>>>> Thank you for your help!
>> Best regards,
>> Katharina
>>
>> --
>> You received this message because you are subscribed to the Google Groups
>> "cp2k" group.
>> To unsubscribe from this group and stop receiving emails from it, send an
>> email to cp... at googlegroups.com.
>> To view this discussion on the web visit
>> https://groups.google.com/d/msgid/cp2k/168440d5-5b57-4075-900e-876a99f8e437o%40googlegroups.com
>> <https://groups.google.com/d/msgid/cp2k/168440d5-5b57-4075-900e-876a99f8e437o%40googlegroups.com?utm_medium=email&utm_source=footer>
>> .
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20200626/f49b6736/attachment.htm>