1
$\begingroup$

The final system to model is an interface made of 2 surfaces of 12 fcc-Pt atoms each, with a NH3 molecule in between. I'm using ultra-soft pseudopotentials. If I take the energy cutoff (ecutwfc) that gives a converged lattice parameter within 0.01 Angstrom, I find 30 Rydberg for Pt and 40 for NH3 (NH bond length converged within 0.01 Angstrom).

My question is, given a large number of atoms with lower cutoff and a minority number of atoms with larger cutoff, how should I pick the final cutoff? Should I always go for safer option (higher cutoff) even though it causes the majority of the system to be treated with a cutoff higher than needed, increasing computational time?

$\endgroup$
1
  • 2
    $\begingroup$ Take Ecut at higher side $\endgroup$ Feb 2, 2023 at 2:14

1 Answer 1

1
$\begingroup$

Yes, you should always choose the largest cutoff for the combined system; otherwise, your molecular system won't be treated correctly.

You can check out our paper on reaction pathways of a molecule on a Si surface here: https://pubs.acs.org/doi/abs/10.1021/acs.jpcc.6b04450

$\endgroup$
2
  • 1
    $\begingroup$ As it’s currently written, your answer is unclear. Please edit to add additional details that will help others understand how this addresses the question asked. You can find more information on how to write good answers in the help center. $\endgroup$
    – Community Bot
    Feb 2, 2023 at 15:40
  • $\begingroup$ Welcome to our new community and thank you for contributing your answer here! We hope to see much more of you in the future! I see you've had an account for more than a year but you're posting for the first time. Answers often get "flagged" for their length or content. In this case the issue was the "length". Your entire answer fits within a comment with more than 300 characters remaining. Is it possible for you to go into more detail and explain the answer (perhaps pointing to which part of the paper to examine, for example)? It might help to look at some other answers on the site. $\endgroup$ Feb 2, 2023 at 15:42

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .