restricted open-shell Kohn-Sham (ROKS) Setup

[Michael LaCount]

Hello I am trying to set up a fairly simple test run for ROKS. The simplest test I could think of for what I want to try to do is to set up 2 jobs, both will have a Na and Cl atom placed 1nm apart and with periodic boundary conditions turned off (and 1nm from nearest box wall). The 2 jobs differ in that I want the first to converge to neutral atoms, and the second to converge to ions. My understanding with ROKS is that this should be doable and fairly easy.

My jobs are running into an error. CPASSERT failed qs_ks_utils.F:263, so there is something wrong with my setup. The CPASSERT in question is:

      CPASSERT(Nelectron == k_alpha - k_beta)

The CP2K code block I am using is:

  &DFT
    CHARGE 0
    MULTIPLICITY 1
    UKS
    ROKS
    &LOW_SPIN_ROKS
      ENERGY_SCALING 1 -1 1 -1
      SPIN_CONFIGURATION 1 1
      SPIN_CONFIGURATION 1 2
      SPIN_CONFIGURATION 2 2
      SPIN_CONFIGURATION 2 1
    &END LOW_SPIN_ROKS

...

 &END DFT

My interpretation of how this would work is that it would find 2 unpaired electron levels: an 's' for Na and a 'p' for Cl. This would then bias the two orbitals to a certain spin. I don't think I am understanding the syntax of these commands properly and would like some guidance on how to properly cover the 2 scenarios I mentioned (neutral Na/Cl and ionic Na/Cl in a non-periodic box).

Thanks

Michael

[Krack, Matthias]

Hi Michael

 

The error occurs, because MULTIPLICITY must be set to 3 in that case also for the calculation of a restricted open-shell singlet with OT. That defines the number of open shell orbitals involved in the calculation. OT includes only the occupied orbitals and the calculated “base” term is a triplet. Have a look at the CP2K test input tests/QS/regtest-lsroks/O2.inp. Note that the terms defined by SPIN_CONFIGURATION are additional terms which are added to the triplet (base) term which is calculated in any case.

 

HTH

 

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 visit https://groups.google.com/d/msgid/cp2k/209eb892-0358-40ef-bf69-2e370dc8cba0n%40googlegroups.com.