[CP2K-user] [CP2K:21061] Re: CDFT geometry optimization convergence issue
uca...@ucl.ac.uk
ucapca1 at ucl.ac.uk
Sat Jan 18 03:07:14 UTC 2025
Hi Akanksha,
It looks like you have changed a number of the CDFT keywords from the
recommended values. Have you confirmed that your chosen settings work for
other systems?
Otherwise I suggest you follow the recommended keywords at
https://manual.cp2k.org/trunk/methods/dft/constrained.html
https://www.cp2k.org/_media/events:2018_user_meeting:cp2k-uk-2018-holmberg.pdf
https://pubs.acs.org/doi/full/10.1021/acs.jctc.2c00284
https://github.com/cp2k/cp2k/blob/master/tests/QS/regtest-cdft-hirshfeld-3/hirshfeld_qs.inc
The most important changes are probably:
OPTIMIZER BROYDEN -> NEWTON_LS
SHAPE_FUNCTION GAUSSIAN -> DENSITY
MAX_SCF 20 -> 60
I expect it would be best to remove SCCS and check that CDFT converges
properly without it. I don't have any experience with SCCS, but I assume
that it will make convergence more challenging.
Chris
On Friday, 17 January 2025 at 08:24:39 UTC Akanksha wrote:
> Dear CP2K community,
>
> I am trying to run *constrained DFT calculations* in CP2K 2024.2 to
> calculate the electron transfer parameters of the following reaction in an
> implicit DMF environment:
> *[trans-stilbene]- + 2bromophenol ------------> trans-stilbene +
> [2-bromophenol]-*
>
> 1) I have observed that *the SCF wavefunction optimization keeps on
> fluctuating*, even though I am restarting my calculation from an already
> optimized wavefunction under similar conditions (but without any charge
> constraint).
> 2) In other similar reactions, I have observed that even if the
> wavefunction reaches optimization somehow - *the charges do not reach the
> appropriate constraint target value. *The iterations are either too slow
> to reach convergence or start to give a drastically wrong number of valence
> electrons.
>
> Here is the CDFT section i am using:
> &CDFT
> TYPE_OF_CONSTRAINT HIRSHFELD
> ATOMIC_CHARGES TRUE
> COUNTER 0
> STRENGTH ${CDFT_STRENGTH}
> TARGET ${CDFT_TARGET}
> &ATOM_GROUP
> ATOMS ${CDFT_ATOMS}
> COEFF ${CDFT_COEFF}
> CONSTRAINT_TYPE CHARGE
> &END ATOM_GROUP
> &DUMMY_ATOMS
> ATOMS ${CDFT_DUMMY}
> &END DUMMY_ATOMS
> &OUTER_SCF ON
> TYPE CDFT_CONSTRAINT
> EXTRAPOLATION_ORDER 2
> MAX_SCF 50
> EPS_SCF 1.0E-3
> OPTIMIZER BROYDEN
> STEP_SIZE -2.0
> &CDFT_OPT ON
> BROYDEN_TYPE BT1_LS
> MAX_LS 5
> CONTINUE_LS
> FACTOR_LS 0.5
> JACOBIAN_STEP 1.0E-2
> JACOBIAN_FREQ 1 1
> JACOBIAN_TYPE FD1
> JACOBIAN_RESTART .FALSE.
> &END CDFT_OPT
> &END OUTER_SCF
> &HIRSHFELD_CONSTRAINT
> SHAPE_FUNCTION GAUSSIAN
> GAUSSIAN_SHAPE VDW
> &END HIRSHFELD_CONSTRAINT
> &PROGRAM_RUN_INFO ON
> &EACH
> QS_SCF 1
> &END EACH
> COMMON_ITERATION_LEVELS 2
> ADD_LAST NUMERIC
> FILENAME ./${NAME}
> &END PROGRAM_RUN_INFO
> &END CDFT
>
> Could you please help me figure out what is causing these issues? I have
> attached my input and output files for your reference.
>
> I really appreciate all your time and help. Thank you,
> Akanksha
>
> P.S. This question is a repost as I was not very clear in my previous post
> and didn't get a response. Sorry about that!
>
>
--
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 cp2k+unsubscribe at googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/cp2k/f1e71ab9-ef04-46ac-bbbd-61372e7bd383n%40googlegroups.com.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20250117/4e1ddf14/attachment.htm>
More information about the CP2K-user
mailing list