I'm trying to study a Yb-Ge-Sb compound and would like to include spin-orbit coupling (SOC). In Quantum ESPRESSO, this is done by including the following tags in the &SYSTEM name list.

lspinorb = .true.
noncolin = .true.
starting_magnetization = 0.0

But in the case of the pseudo-potentials included, should I use fully relativistic pseudo-potentials for all the three elements?

NOTE: As per the materials project repository, the compound is non-magnetic.

  • 1
    $\begingroup$ +1 for a great question! Welcome to the site and thank you for contributing your question to our new community! We hope to see much more of you in the future !!! $\endgroup$ Jan 19 '21 at 5:42

But in the case of the pseudo-potentials included, should I use fully relativistic pseudopotentials for all three elements?

Yes, you should use the fully relativistic pseudopotentials if you are considering spin-orbit coupling calculations.

When you should consider spin-orbit coupling?

If the investigated materials contain heavy atoms then you should consider spin-orbit coupling. This is irrelevant to its magnetic properties.

  • $\begingroup$ +1 For the quick answer for a new user, and I agree with what you're saying. I've added a second answer just to try to further address the "which elements" part in the title of the question. $\endgroup$ Jan 19 '21 at 5:41

Spin-orbit coupling is related to relativity which is increasingly important for heavier elements.

If you're aiming for ultra-high precision, as in this post on our site: How accurate are the most accurate calculations?, you will need to include every possible effect, including spin-orbit coupling, for all elements including hydrogen.

For most applications, one rule-of-thumb (which is still quite ad hoc) is that relativity becomes important for elements around Krypton (element #36) and onward. This would mean that Yb and Sb would most certainly need to be treated relativistically, and it would also be wise to treat Ge (element #32) relativistically if this is within your means.

What do I recommend?
Include spin-orbit coupling for all three of those elements because Yb and Sb are considered to be highly reltivistic, and you might as well include SO coupling for Ge too since it's also fairly heavy and you're already treating the other two elements with SO coupling.


The spin-orbit coupling is an interested phenomena very present in semiconductor physics (no heavy atoms here). It affects the behavior of electrons and is a very important property explored in the field of spintronics.

For example, in this work, the influence of spin-orbit coupling in the electronic band of $GaAs$ is studied.

Please, take a look at this seminal work of Prof. S. Das Sarma entitled Spintronics: Fundamentals and applications on the Review of Modern Physics journal (doi: 10.1103/RevModPhys.76.323, ArXiv version).


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.