[QE-users] Pseudopotentials for magnetized computations

[Johnson, Miles R.]

Hi all,

I'm doing various computations for NiPS3, which is antiferromagnetic so the bulk of my computations are spin polarized. I was reading through the paper on GBRV pseudopotentials by David Vanderbilt's group, and they "recommend significant testing before using any pseudopotential in detailed magnetic calculations" while noting that their pseudopotentials did not perform super well for NiO at least partially due to complications with computing the magnetic properties of Ni. They also say the pseudopotentials were made without spin-orbit interactions, which I think may be important in my case (one paper suggests it as a cause for magnetic anisotropy).

Does anyone have suggestions for how to choose and test pseudopotentials when trying to do detailed magnetic computations? I would think a fully relativistic potential is best, but at first glance looking at the qe webpage (https://www.quantum-espresso.org/other-resources/), Don Hamann's code seems to be the only choice for those. Could other pseudopotentials potentially be better for my spin-polarized computations? How do I test their viability?

Any advice is welcome.

Thanks!
-Miles

---
Miles Johnson
Applied Physics PhD candidate
California Institute of Technology
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20221101/ebdb09ab/attachment.html>