[CP2K-user] [CP2K:22362] Error in BS calculation: KS energy is an abnornal value
Thomas Kühne
tkuehne at gmail.com
Sun Aug 2 15:51:04 UTC 2026
his is not a normal SCF convergence failure. The calculation aborts before the first SCF iteration while CP2K is building the initial Kohn–Sham energy. Therefore changing MAX_SCF, Pulay/Broyden mixing, or the electronic temperature will not fix the underlying NaN.
Please provide the complete input and the output from the beginning up to the abort, especially:
UKS, CHARGE, and MULTIPLICITY
both Gd &KIND sections, including the complete &BS definitions
basis-set and pseudopotential names/files
DFT+U and XC settings
coordinates, cell, and Poisson periodicity
CP2K version
As a diagnostic, first remove the &BS sections, smearing, level shift, and explicit CHOLESKY INVERSE, and try a plain UKS calculation with the default Cholesky method. If that works, the problem is in the broken-symmetry occupations.
Note that NEL in &BS specifies changes relative to the atomic reference occupation, not the final occupation. Check the “Guess for atomic kind” section in the output. Also make sure the Gd pseudopotential includes the 4f electrons in the valence. The LnPP2/q11 potentials are 4f-in-core and cannot describe opposite local 4f spin configurations.
Finally, LEVEL_SHIFT 0.2 without a unit means 0.2 Hartree, not 0.2 eV.
Once the calculation gives a finite initial energy, SCF convergence can be improved separately, for example with a smaller mixing alpha such as 0.1–0.2.
> Am 30.07.2026 um 07:37 schrieb 'Parismita Borgohain' via cp2k <cp2k at googlegroups.com>:
>
> Hello, I am trying to do BS calculation in Gd2 system but i got this error all the time. i have tried &SCF
> SCF_GUESS ATOMIC
> EPS_SCF 1.0E-7
> MAX_SCF 1000
> CHOLESKY INVERSE
> ADDED_MOS 50
> &SMEAR ON
> METHOD FERMI_DIRAC
> ELECTRONIC_TEMPERATURE [K] 300
> &END SMEAR
> &DIAGONALIZATION
> ALGORITHM STANDARD
> &END DIAGONALIZATION
> &MIXING
> METHOD PULAY_MIXING
> PULAY_ALPHA 0.4
> NPULAY 8
> &END MIXING
> &END SCF
>
> and &SCF
> SCF_GUESS ATOMIC
> EPS_SCF 1.0E-5
> MAX_SCF 1000
> CHOLESKY INVERSE
> LEVEL_SHIFT 0.2
> ADDED_MOS 50
> &SMEAR ON
> METHOD FERMI_DIRAC
> ELECTRONIC_TEMPERATURE [K] 500
> &END SMEAR
> &DIAGONALIZATION
> ALGORITHM STANDARD
> &END DIAGONALIZATION
> &MIXING
> METHOD BROYDEN_MIXING
> ALPHA 0.4
> NBROYDEN 8
> &END MIXING
> &END SCF. But I got the same error. Can anyone please suggest me some scf convergence tips.
>
>
> SCF WAVEFUNCTION OPTIMIZATION
>
> Step Update method Time Convergence Total energy Change
> ------------------------------------------------------------------------------
>
> *******************************************************************************
> * ___ *
> * / \ *
> * [ABORT] *
> * \___/ KS energy is an abnormal value (NaN/Inf). *
> * | *
> * O/| *
> * /| | *
> * / \ qs_ks_methods.F:889 *
> *******************************************************************************
>
> Disclaimer:- This footer text is to convey that this email is sent by one of the users of IITH. So, do not mark it as SPAM.
>
> --
> 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 <mailto:cp2k+unsubscribe at googlegroups.com>.
> To view this discussion visit https://groups.google.com/d/msgid/cp2k/bbeb3025-1a57-416b-aa2f-e8dff83cd40en%40googlegroups.com <https://groups.google.com/d/msgid/cp2k/bbeb3025-1a57-416b-aa2f-e8dff83cd40en%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/E01EDE42-60A4-4946-8166-B3533F944D0A%40gmail.com.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20260802/a2193b03/attachment-0001.htm>
More information about the CP2K-user
mailing list