Using the SOGGA11-X hybrid functional

[Ari Paavo Seitsonen]

Hello,

  I try to use the SOGGA11-X hybrid functional, in order to reproduce some earlier results; but I seem to get results that differ from the earlier ones. Could some one please check whether the following section is fine?

    &XC
      &XC_FUNCTIONAL
        &LIBXC
          FUNCTIONAL  HYB_GGA_X_SOGGA11_X
        &END LIBXC
        &LIBXC
          FUNCTIONAL  GGA_C_SOGGA11_X
        &END LIBXC
      &END XC_FUNCTIONAL
      &HF
        FRACTION  0.4015
        &SCREENING
          EPS_SCHWARZ          1.0E-8
        &END SCREENING
        &MEMORY
          MAX_MEMORY  4000
        &END MEMORY
      &END HF
    &END XC

  Thank you in advance!

    Greetings from Sunny Paris,

       apsi

[Frederick Stein]

Dear Apsi,

Remove the line "FRACTION 0.4015". This value is set by cp2k.

Best regards

Frederick

[Ari Paavo Seitsonen]

Dear Frederick,

  Thank you for the note - are you fully sure about this? After commenting out the lnie with "FRACTION" the bond lengths have become unphysically short. Also in the file "tests/QS/regtest-libxc/H2O-hybrid-b3lyp_libxc.inp" the "FRACTION" is present. Well, I continue digging out for how this is supposed to work... :)

    Greetings,

       apsi

--
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/290c46eb-3a4e-430b-be70-f347c3e91ddf%40googlegroups.com.

--

-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-
  Ari Paavo Seitsonen / Ari.P.S...@iki.fi / http://www.iki.fi/~apsi/
    Ecole Normale Supérieure (ENS), Département de Chimie, Paris
    Mobile (F) : +33 789 37 24 25    (CH) : +41 79 71 90 935

[Frederick Stein]

Dear apsi,

Sorry, I mixed it up with something different. I had a look at the corresponding reference. They do not use the full functional but a fraction of 0.5985. You have to set that manually, too. When CP2K calls Libxc, Libxc probably calculates the functional with a fraction of 1.0 as if you call a plain GGA without any scaling. For functionals that are explicitly available combinations of functionals like B3LYP, the needed mixture is already included by Libxc.

Sorry for the confusion.

Best regards.

Frederick

Am Samstag, 10. August 2019 15:33:21 UTC+2 schrieb Ari Paavo Seitsonen:

Dear Frederick,

  Thank you for the note - are you fully sure about this? After commenting out the lnie with "FRACTION" the bond lengths have become unphysically short. Also in the file "tests/QS/regtest-libxc/H2O-hybrid-b3lyp_libxc.inp" the "FRACTION" is present. Well, I continue digging out for how this is supposed to work... :)

    Greetings,

       apsi

Le jeu. 8 août 2019 à 13:25, 'Frederick Stein' via cp2k <cp...@googlegroups.com> a écrit :

Dear Apsi,

Remove the line "FRACTION 0.4015". This value is set by cp2k.

Best regards

Frederick

--
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 cp...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/290c46eb-3a4e-430b-be70-f347c3e91ddf%40googlegroups.com.

--

-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-
  Ari Paavo Seitsonen / Ari.P....@iki.fi / http://www.iki.fi/~apsi/