4
$\begingroup$

I have been using QuantumEspresso (QE) for quite some time, but I have been thinking about trying SIESTA for some of my structures that are in a slab geometry with vacuum on either side. QE will use plane waves to fill the vacuum region and as it gets bigger then the calculation takes longer and longer.

How does SIESTA treat the vacuum region? How big of a vacuum region can I use in SIESTA?

$\endgroup$

1 Answer 1

2
$\begingroup$

Vacuum in siesta is very cheap, generally it is not something you should worry about. E.g. for graphene you can easily use 100 ang or 300 ang in the vacuum region.

Of course siesta will use more memory, but that depends on your available memory for the hardware you are using.

$\endgroup$
3
  • $\begingroup$ Will SIESTA use more memory than QE? $\endgroup$
    – Camps
    Commented Jul 5 at 17:02
  • 1
    $\begingroup$ Generally no, but as always, it depends ;) if you do the same system in ge and siesta, then siesta will use less memory. $\endgroup$
    – nickpapior
    Commented Jul 5 at 17:17
  • $\begingroup$ thank you for this clarification! if you don't mind I have another quick question: QE has a flag called assume_isolated='2D' which is a different (slightly cheaper) method for dipole correction. does siesta have the same thing? i haven't found it but maybe it called differently $\endgroup$
    – lucian
    Commented 5 hours ago

You must log in to answer this question.

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