SCCS Parameters: No change with Alpha and Beta

Logan Ward

unread,

Feb 20, 2023, 12:02:59 PM2/20/23

to cp2k

I am running solvation energy computations for isolated molecules, and finding that my choices of Alpha and Beta do not affect the output energy.

While the solvation energies match well to PCM without tuning these parameters, the lack of any effect worries me.

The full input file is shown below, but there are a few points where I'm concerned:

- Do I need to change the defaults for RHO_MAX or RHO_MIN?

- Does SCCS work fully with non-periodic boundary conditions?

Thanks,

Logan

!!! Generated by ASE !!!
&FORCE_EVAL
METHOD Quickstep
&DFT
BASIS_SET_FILE_NAME GTH_BASIS_SETS
POTENTIAL_FILE_NAME POTENTIAL
&XC
&XC_FUNCTIONAL BLYP
&END XC_FUNCTIONAL
&END XC
&POISSON
PERIODIC NONE
PSOLVER MT
&END POISSON
&SCF
MAX_SCF 10
&OUTER_SCF
MAX_SCF 9
&END OUTER_SCF
&OT T
PRECONDITIONER FULL_ALL
&END OT
&END SCF
&SCCS
ALPHA -29.45
BETA 0.1
GAMMA 29.45
RELATIVE_PERMITTIVITY 37.5
DERIVATIVE_METHOD CD3
METHOD ANDREUSSI
&END SCCS
&MGRID
CUTOFF [eV] 8.163415807310173477e+03
&END MGRID
&LS_SCF
MAX_SCF 10
&END LS_SCF
&END DFT
&SUBSYS
&TOPOLOGY
&CENTER_COORDINATES
&END CENTER_COORDINATES
&END TOPOLOGY
&COORD
C 1.082093052380952258e+01 1.133453319047618990e+01 7.429094119047618960e+00
O 1.076758152380952360e+01 9.918479190476190865e+00 7.390901119047618373e+00
C 9.485502523809524078e+00 9.399504190476189791e+00 7.319581119047618323e+00
C 9.585050523809524492e+00 7.892009190476191094e+00 7.098288119047618139e+00
C 8.281867523809523846e+00 7.173732190476190596e+00 7.445223119047618354e+00
C 7.939781523809523733e+00 6.899449190476190985e+00 8.891974119047617364e+00
C 7.170046523809523897e+00 7.943473190476190382e+00 8.130744119047619733e+00
C 7.417144523809524159e+00 9.408576190476191314e+00 8.432348119047619051e+00
O 8.801966523809523935e+00 9.693925190476191389e+00 8.533224119047618572e+00
H 1.187744352380952506e+01 1.161050519047619112e+01 7.433736119047618551e+00
H 1.034048252380952349e+01 1.177429119047619110e+01 6.541877119047618372e+00
H 1.033506752380952420e+01 1.173072019047619108e+01 8.326848119047618013e+00
H 8.918108523809523902e+00 9.877577190476191760e+00 6.493754119047618900e+00
H 9.853268523809523671e+00 7.711232190476190240e+00 6.052890119047618533e+00
H 1.041464952380952269e+01 7.522907190476191275e+00 7.710534119047618873e+00
H 7.973253523809523458e+00 6.415422190476190778e+00 6.733249119047618692e+00
H 7.443506523809523934e+00 5.967659190476190645e+00 9.139735119047617928e+00
H 8.646955523809523925e+00 7.252218190476190429e+00 9.636019119047618986e+00
H 6.148317523809524232e+00 7.708863190476190397e+00 7.846544119047618615e+00
H 6.959795523809523488e+00 1.003365719047619109e+01 7.645480119047618928e+00
H 6.975102523809523447e+00 9.700901190476191260e+00 9.390809119047618836e+00
&END COORD
&CELL
PERIODIC NONE
A 1.772912600000000083e+01 0.000000000000000000e+00 0.000000000000000000e+00
B 0.000000000000000000e+00 1.780663200000000046e+01 0.000000000000000000e+00
C 0.000000000000000000e+00 0.000000000000000000e+00 1.558312899999999956e+01
&END CELL
&KIND O
BASIS_SET DZVP-GTH
POTENTIAL GTH-BLYP
&END KIND
&KIND H
BASIS_SET DZVP-GTH
POTENTIAL GTH-BLYP
&END KIND
&KIND C
BASIS_SET DZVP-GTH
POTENTIAL GTH-BLYP
&END KIND
&END SUBSYS
&END FORCE_EVAL
&GLOBAL
PROJECT cp2k
PRINT_LEVEL LOW
&END GLOBAL

Krack Matthias

unread,

Feb 21, 2023, 4:38:57 AM2/21/23

to cp...@googlegroups.com

Hi Logan

The SCCS parameters alpha and beta affect only the (non-electrostatic) repulsion and the dispersion energy terms, respectively. The electrostatic solvation free energy is not affected by these parameters. Check the CP2K manual and the corresponding reference for details.

The default values for RHO_MIN and RHO_MAX should give reasonable results for neutral systems. If the results for the solvation energies are fine, then there is no need to change these parameters.

Yes, SCCS works fine for non-PBC. For charged systems it is even mandatory to use non-PBC (0-periodicity), e.g. by employing MT.

I also suggest to use the latest CP2K version 2023.1 with the revised SCCS printout.

HTH

Matthias

--
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+uns...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/34c38485-67ec-4e04-9496-ab8586b14786n%40googlegroups.com.

Logan Ward

unread,

Feb 21, 2023, 7:51:32 AM2/21/23

to cp...@googlegroups.com

This does help a lot. Thanks, Matthias!

You received this message because you are subscribed to a topic in the Google Groups "cp2k" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/cp2k/7oYTqSIyIqI/unsubscribe.
To unsubscribe from this group and all its topics, send an email to cp2k+uns...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/ZRAP278MB0827A4EC015840FB5F8AA438F4A59%40ZRAP278MB0827.CHEP278.PROD.OUTLOOK.COM.

Reply all

Reply to author

Forward