[QE-users] Applicability of DFT+U+J to Anisotropy Calculations

[Peter Stoeckl]

Dear all,

In recent calculations I have been attempting to calculate uniaxial
magnetocrystalline anisotropies in transition-metal magnetic compounds
using DFT+U method.

As I understand it, these calculations require (a) non-collinear spins
`noncolin = .true.` and (b) DFT+U with both U and J parameter
`lda_plus_u_kind = 1`, as the simpler  `lda_plus_u_kind = 0` is
incompatible with non-collinear spin calculations.
(This is occasionally inconvenient, as DFT+U+J itself is not supported for
structural relaxations.)

1. Are there any known issues (or theoretical concerns) with using DFT+U+J
in calculations of magnetic anisotropy in QE?

2. For reference, I calculate E_MCA = E_IP – E_OP, i.e. the difference in
(SCF) energy between configurations with in-plane and out-of-plane
magnetization respectively, rather than the perhaps easier force theorem
approach. While I don't expect this to change much, does this affect the
considerations from (1), if any?

Any information (or corrections) regarding the above points would be
greatly appreciated.

Peter Stoeckl
PhD Candidate, Department of Physics
University of Minnesota
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20220217/a11feaad/attachment.html>