Heey Vipul Kumar Ambasta

In fact, relaxing the top layers (maybe more than just one, say two) and the adsorbed molecules, while keeping the bulk fixed, is a popular approach to balance accuracy and computational efficiency. This "partial" but suitable relaxation of just the top few layers and adsorbed species before the phonon calculation may prevent negative frequencies -try and see. And yes, the nat_todo parameter is the one to limit phonon perturbations to specific atoms like the adsorbed molecule and top layers. You're on the right track!

Best,

Gabriel Vinicius

Hey Gabriel Vinicius

Thanks for your kind reply. Your message helped me clear some concerns, thanks for that.

I will try out your appraoch and get back to you.

--

Best Regards,

Vipul Kumar Ambasta

While using nat_todo to perturb only the adsorbed molecules and top-layer of the crystal is an efficient strategy, ensure that the non-relaxed lattice is in a stable configuration to avoid the introduction of unwanted negative frequencies in your phonon dispersion.