On Thu, 7 Jul 2005, Oleksandr Voznyy wrote: | Hi, | So does SIESTA support spin-orbit interaction calculations?
No. | I saw in very old messages in the list that it doesn't. That ATOM generates | spin dependent potentials but SIESTA neglects them. You seem to mix up spin-orbit with spin-polarized calculation. Spin-polarized case does not necessarily includes spin-orbit, and a calculation with spin-orbit (in other codes) is not necessarily a spin-polarized one. Siesta can run magnetic calculation also using non-magnetic (relativistic or not) pseudopotentials. | Is 'SpinPolarized True' related to spin-orbit interactions???? No. It allows spin-up and spin-down components (of potential, charge density...) to be different. But there is no spin-orbit term which would mix them. Such mixing comes about when one uses non-collinear option, but even there the direction of magnetization stands in no relation to crystallographic axes. Andrei Postnikov
