4
$\begingroup$

I am trying to optimize a copper supercell for the study of chalcones adsorption using Quantum ESPRESSO, however, the size of my supercell is large (>200 atoms) and QE can not handle it.

What is an alternative for it?

Improve this question
$\endgroup$

1 Answer 1

Reset to default
5
$\begingroup$

It should be possible to use Quantum Espresso to handle more than 200 atoms.

I guess you are studying a surface, in that case, please note that in PW-based codes the vacuum makes your calculation more expensive since it is considered as if it was matter. For this reason, you can try to lower a little the vacuum (while always ensuring convergence). Always on the 2D case, you can try to use a Coulomb potential along the non-periodic directions to ease the convergence.

Lastly, if you are optimizing the cell and you need to study absorption, you can try to consider the optimization of only 1/2 layers on the surface while keeping frozen the bottom layers.

To finally answer your question, if you don't want to use Quantum Espresso, there are alternative PW codes (such as Abinit, VASP) and localized basis sets that would ease the fact that you are dealing with vacuum (such as Siesta).

Improve this answer
$\endgroup$
5
  • $\begingroup$ Thanks for your response. Yes, I'm working with a surface and my supercell has over 200 atoms. I already tried using a little vacuum and only 1/2 optimization, however, my calculations did not continue. Maybe it's impossible to use a PW base for this case? $\endgroup$
    – Everson Gomes
    Jan 31 at 19:02
  • 1
    $\begingroup$ I wouldn't say that it's impossible, several studies report approximately this amount of atoms with PW codes. Of course, sometimes with low-dimentional materials, localized basis set could help you speed-up the calculations but theoretically you are free to use PW successfully. At this point, I would suggest just using more processors for your calculation could do the trick. Also, if it can help you, you can start by optimizing with lower parameters (cut-off, k-points...) to reach a better initial guess and then re-start with the converged parameters. $\endgroup$
    – Laura
    Jan 31 at 19:09
  • $\begingroup$ Thank you so much. $\endgroup$
    – Everson Gomes
    Jan 31 at 19:55
  • $\begingroup$ Unfortunately, I already used the minimum values for ecut, K points, mixingbete, and my calculations always stopped or can not achieve final coordinates. $\endgroup$
    – Everson Gomes
    Jan 31 at 20:05
  • $\begingroup$ If lots of comments are needed, don't hesitate to use the Quantum ESPRESSO chat room, rather than clicking the link SE gives you for making new chat rooms, since we already have a good room for QE discussions. $\endgroup$
    – Nike Dattani
    Feb 1 at 0:02

You must log in to answer this question.

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