Dear MS Users,

i am trying to calculate harmonic frequencies of an adsorbed molecule on a copper oxide cluster. A geometry optimization of the whole structure worked fine.

The problem i'm facing is, that the SCF cycle fails to converge as soon as one of the atoms is displaced for the frequency calculation; although the first (optimal structure) cycle converges fine.

I'm using Pulay mixing with 10 steps and up to 0.1Ha smearing for the electrons. The adsorbate is in the "Hessian Atoms" set.

Find an example outmol attached, it shows convergence in the first cycle, then fails after one atom was displaced.

Thanks for any ideas!

Martin