Spin polarization and multiplicity optimization

[Dong-Hui Xu]

Dear Cp2k forum,

   I am currently performing an MTD calculation using PLUMED embedded in CP2K. The reactants are two radicals with a multiplicity of 3 or a singlet state of an open-shell configuration (with energies that are close), and the product is a closed-shell singlet state.   In my calculation, I employed UKS, RELAX_MULTIPLICITY, and DIAGONALIZATION. The issue encountered during the MTD simulation is that, upon product bond dissociation and transition back to the reactants, the electronic state of the reactants consistently adopts a closed-shell singlet configuration, which leads to an unreasonably high energy for the reactant in this state. Is there a method to enforce the reactants to adopt a triplet state or an open-shell singlet state, as opposed to the closed-shell singlet configuration?

Thank you very much,

Sincerely,

Donghui