Dear all,

Hope you have a good day.

As far as I know, the imaginary frequency means that one of harmonic squared frequencies contains imaginary number, and if it's not the transition calculation, especially if it is optimization task, this imaginary number means that the scf orbital gradient does not converged tightly.

The problem is the imaginary frequency keeps obtained during the optimization calculation, though I set the customized parameterization of the SCF convergence as 1.0e-10 in the DMol3 calculation.

Is there anything wrong about my knowledge?

Or is there anything I miss-set in my calculation?

Thank you.

Best regards,

Rakhoon Choi