I have been trying to construct the spin-density plot of an α-alumina supercell. Being a diamagnetic material, I would not expect to see any spins in this structure. However, when running a spin-polarized calculation, all O atoms seem to have a spin.
This calculation was done on a 2x2 supercell using Quantum ESPRESSO (spin-polarized with 0.1 initial magnetic moments on all atoms). I repeated the same calculation (SCF run followed by extraction of Spin Densities) using CP2K and GPAW and got similar results. I also tried with the unit cell of α-alumina, which had spins on O atoms.
Could someone help understand what could be happening here? Any input would be much appreciated.
