[CP2K-user] [CP2K:21717] Discrepancy in using Range-Separated Hybrid Functional with D4 Correction

[Johann Pototschnig]

Hi

You could try comparing to the alternative implementation:
D4_REFERENCE_CODE F

Did you also check the other parameters?
FACTOR_S9_TERM
D4_CUTOFF
D4_CN_CUTOFF
R_CUTOFF

best wishes


On Wednesday, July 30, 2025 at 1:27:27 PM UTC+2 Jürg Hutter wrote:

> Hi
>
> I have to guess, but here I my two top options:
>
> 1) Poisson solver: use WAVELET, it is more accurate than MT
> 2) D4: check that CP2K and ORCA have the same D4 contributions
>
> regards
> JH
>
> ________________________________________
> From: cp... at googlegroups.com <cp... at googlegroups.com> on behalf of Lim 
> Jia Yang <limjiay... at gmail.com>
> Sent: Tuesday, July 29, 2025 2:53 PM
> To: cp2k
> Subject: [CP2K:21699] Discrepancy in using Range-Separated Hybrid 
> Functional with D4 Correction
>
> Dear CP2K community,
>
> CP2K version 2025.1, ORCA version 6.1.0
> I am doing benchmarking on vertical detachment energy. Both functionals 
> used are from Libxc for calculations in ORCA and CP2K.
>
> First method: LC-ωPBE+D3BJ(2016)
> Second method: LC-ωPBE08+D4
>
> The first method gives similar values with a ~0.0002 eV difference between 
> ORCA and CP2K. The second method shows differences of ~0.024 eV, with CP2K 
> being lower.
>
> Any advice is really appreciated. I hope I didn't make some mistakes in 
> the input.
>
> Attached are the CP2K input files for both methods. The output file for 
> the second method shows "POWELL| Error in trust region". The same error 
> appears when D4 correction is removed.
>
> ORCA input line for the first method:
> ! D3BJ aug-cc-pVTZ TightSCF DEFGRID3
> %method
> method dft
> functional hyb_gga_xc_lc_wpbe
> D3S6 1.00
> D3A1 0.563761
> D3S8 0.906564
> D3A2 3.593680
> end
> ORCA input line for the second method:
> ! D4 aug-cc-pVTZ TightSCF DEFGRID3
> %method
> method dft
> functional hyb_gga_xc_lc_wpbe08_whs
> D4S6 1.00
> D4A1 0.378
> D4S8 1.170
> D4A2 4.816
> end
>
> --
> 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 visit 
> https://groups.google.com/d/msgid/cp2k/6860d009-2149-4781-84ee-227085e3a60bn%40googlegroups.com
> <
> https://groups.google.com/d/msgid/cp2k/6860d009-2149-4781-84ee-227085e3a60bn%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 visit https://groups.google.com/d/msgid/cp2k/8c8e504d-b514-4d93-a8d5-89f9fac1eb1cn%40googlegroups.com.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20250730/c3dec99f/attachment.htm>