understanding BSSE print out

[Victor Volkov]

Dear members:

good evening/morning.

it is not clear to me how to interpret CP2K print out for BSSE.

If I take a small guest (like benzene) next to a carbon layer using GWP/PBE/DZVP

Of course, since interactions are so week, DFTD3 is a must - otherwise a benzene may "fly away".

Then, using BSSE  for S1(guest) and S2 (carbon layer)

CP2K prints

S1 + S2ghost: Total energy -455.12112693238430 Hartree
S1ghost + S2: Total energy -455.12112693238430 Hartree
S1           + S2: Total energy -455.12112693238441 Hartree

The differences are none or tiny.

However, the final print is:

CP-corrected Total energy:        0.015729
1-body contribution:                 -44.96634268204556
1-body contribution:                -410.13905550190344
2-body contribution:                 455.121127
BSSE-free interaction energy:      455.121127

First, it is not clear what is "CP".

Next, if this is the value of the BSSE artifact (due to the wave-function truncation),

then its huge 41.2964895 kJ/mole.

This would be almost twice larger than the binding energy of benzene-like guests when at carbonaceous surfaces.

This would mean that quality evaluations of binding energies for such systems using CP2K would be very difficult.

Any comments to hear would be nice.

Finally, it is not clear why 2-body contribution and BSSE-free interaction are printed as positive numbers.

I hope the questions I ask be interesting for users and helpful for the on-line manual to update.

After all, I wish to note that it was by a rare chance I was able to track the old

notice on the importance of declaring

    &KIND X_ghost
     BASIS_SET DZVP-MOLOPT-SR-GTH
     GHOST
    &END KIND

Even though ...

if I skip such definition, the numbers would come "a bit better"

:-)

CP would be about 21 kJ/mole, what is about comparable to the binding energy

of a benzene-like guest to a carbonaceous material.

 

Thank you for your time and attention.

I wish everybody be well and happy.

Victor

 

[Krack Matthias (PSI)]

Hi Victor

 

Without the full input, it is difficult to make any suggestion. Do you have a &BSSE section defining the FRAGMENTS?

 

Best

 

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/44686fb2-400e-443b-956d-e5abefa70cebn%40googlegroups.com.

[Victor Volkov]

Dear Matthias:

thank you for the kind reply.

I did not wish to be "too specific" about the system in public because the results are under

a peer review.

Here I attach one of the exemplary cases: pyridazine at graphene. 

Thank you.

With best regards,

Victor

To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/ZRAP278MB08278F88114F9F43DE58BEDAF42B9%40ZRAP278MB0827.CHEP278.PROD.OUTLOOK.COM.

[Krack Matthias (PSI)]

Hi Victor

 

As you wrote already, you need to add for each atomic kind the corresponding &KIND section for the GHOST kind like

 

&KIND C_ghost

BASIS_SET DZVP-MOLOPT-GTH

GHOST

&END KIND

 

to get reasonable results.

Usually, one is interested in the last number, the “BSSE-free interaction energy” which is here the difference of the total energies from run (5) and runs (3) and (4), i.e. (5) – [(3) + (4)]. A positive value implies a repulsive interaction (no binding) after removing the BSSE via the counterpoise (CP) correction.

 

HTH

To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/CAK%2B-SSgY_70DJ4kwXhs8Df3OUMjURoxm11SssC1P1b-Nj9y%2BRQ%40mail.gmail.com.

[Victor Volkov]

Thank you dear Mattheas.

Best,

Victor

*****************

Hi Victor...

you need to add for each atomic kind the corresponding &KIND section for the GHOST kind like

 

&KIND C_ghost

BASIS_SET DZVP-MOLOPT-GTH

GHOST

&END KIND

 

to get reasonable results.

Usually, one is interested in the last number, the “BSSE-free interaction energy” which is here the difference of the total energies from run (5) and runs (3) and (4), i.e. (5) – [(3) + (4)]. A positive value implies a repulsive interaction (no binding) after removing the BSSE via the counterpoise (CP) correction.

On Wed, Oct 19, 2022 at 12:12 PM Krack Matthias (PSI) <matthia...@psi.ch> wrote:

To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/ZRAP278MB08278F88114F9F43DE58BEDAF42B9%40ZRAP278MB0827.CHEP278.PROD.OUTLOOK.COM.

[Victor Volkov]

Good morning Matthias:

here, is the result of BSSE application for the case when CP2K suggests a relatively optimal adsorption

of a heterocycle molecule at interface.

After including BSSE, it is not too clear how to operate with the computed numbers:

 -BSSE RESULTS                                -

-CP-corrected Total energy:     -455.120393                                                                                   -
-1-body contribution:      -44.966343                  -
-1-body contribution:     -410.139056                  -
-2-body contribution:       -0.014995                  -

-BSSE-free interaction energy:       -0.014995

because if the BSSE free interaction is -0.014995 Hartree = -39.3693725 kJ/mol.

This is a huge number,

which is significantly larger that any typical binding

for single cycle aromatic molecules at carbonaceous materials.

Curiously,

the energies of binding by CP2K without BSSE agree with experimental results relatively well.

Would the result (after BSSE inclusion) suggest that CP2K theory

is not yet fit to address the physics:

that a reasonable agreement without BSSE should be considered as a blind coincidence,

and that CP2K would need further improvements?

Thank you for any comment.

In any way,

schönes wochenende.

:-)

With best,

Victor

To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/ZRAP278MB0827751044A2BF40D31279B4F42B9%40ZRAP278MB0827.CHEP278.PROD.OUTLOOK.COM.