|
Phonons in crystals with strong phonon-magnetic coupling. If you have a crystal with some antiferromagnetic structure, and
coupling of magnetic moments to atomic distortions is not negligeble,
then phonon dispertion curves can be entirely wrong, if antiferromagnetic
structure is not considered. Suppose that the crystallographic structure
belongs to high-symmetry space group Fd-3m, and remain such if the
magnetic structure is ferromagnetic. In antiferromagnetic configuration
up and down atoms, may not equivalently interact with the neighbors,
It could be so bad that there appear faith soft modes.
Therefore, to establishe proper space group one should name the atoms with
spin up and down as two spicies, hence in Phonon
the Names should be different.
The relevant low-symmetry space group can be found using the FindSym program
supplemented to Phonon. One may find then that
one of the Fd-3m can be I4_1/amd space subgroup.
The Phonon calculations should be performed in I4_1/amd. As a consequence
the list of non-equivalent atoms displacements used to derive the
Hellmann-Feyman forces increases. If you are lucky, then this approach may
remove the undesired soft modes.
|