[CP2K-user] [CP2K:19368] Isolate dispersion contribution in vdW-DF2 calculations

[Michael Fischer]

Dear Prof. Hutter,

Thank you very much for the clarification.
I presume that it cannot be expected that a simple "switching off" of the 
dispersion part (commenting out the entire vdW_POTENTIAL block and 
re-running a single-point calculation with the same combination of exchange 
and correlation functionals) would not give a physically meaningful 
estimation of the dispersion contribution (as the difference between "full" 
rev-vdW-DF2 energy and the "exchange-corellation only" energy), because 
some part of the local contribution is implicitly included in the 
exchange-correlation part. Or could that be a plausible option?

Thanks again, best regards,
Michael

Jürg Hutter schrieb am Montag, 16. Oktober 2023 um 11:08:42 UTC+2:

> Hi
>
> the printed dispersion energy is the calculated energy of the non-local
> vdW energy functional (only the non-local part!). Together with the XC
> energy you get the full functional energy. No other information is 
> available.
>
> regards
> JH
>
> ________________________________________
> From: cp... at googlegroups.com <cp... at googlegroups.com> on behalf of 
> Michael Fischer <doktor... at gmail.com>
> Sent: Friday, October 13, 2023 10:16 AM
> To: cp2k
> Subject: [CP2K:19362] Isolate dispersion contribution in vdW-DF2 
> calculations
>
> Dear all,
>
> I am studying host-guest systems (molecules in zeolites) with the 
> rev-vdW-DF2 functional (X_B86_R, PW92 correlation, LMKLL dispersion) and I 
> am looking at the possibility to calculate the dispersion contribution to 
> the total interaction energy.
>
> In the past, I have used Grimme-type dispersion correction, e.g., PBE-D3. 
> Here, it was fairly trivial to calculate the dispersion contribution as:
> E_disp = E_disp(host+guest) - E_disp(host) - E_disp(guest)
> ...where the E_disp terms on the right-hand side correspond to the energy 
> values given in the line "Dispersion energy" in the CP2K output.
>
> With rev-vdW-DF2, "Dispersion energy" is also printed. However, what is 
> notable in the first place is that the dispersion energy is always 
> positive. A subtraction as in the above equation gives results that look 
> plausible at first, but a comparison to the total interaction energies 
> shows that the E_disp contribution is sometimes more negative than E_tot, 
> something which seems implausible.
>
> Is there some way to isolate the dispersion contribution in calculations 
> using non-local vdW-DF approximations?
>
> Any suggestion will be highly appreciated!
> Thank you and best regards,
> Michael
>
> --
> 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... at googlegroups.com<mailto:cp2k+uns... at googlegroups.com
> >.
> To view this discussion on the web visit 
> https://groups.google.com/d/msgid/cp2k/8432682b-63da-4365-ac25-e53733f7e262n%40googlegroups.com
> <
> https://groups.google.com/d/msgid/cp2k/8432682b-63da-4365-ac25-e53733f7e262n%40googlegroups.com?utm_medium=email&utm_source=footer
> >.
>

-- 
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 on the web visit https://groups.google.com/d/msgid/cp2k/97f8688a-52b4-47f2-98cb-6733be1a2a9cn%40googlegroups.com.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20231016/62f11026/attachment.htm>