[Wien] AFM LaMNO3 Symmetry

[Stefaan Cottenier]

> Paramgnetic symmetry of LaMnO3 is Pnma, (ie, number 62 Int. Tab. Cryst.).
> When AFM (A-phase) sets in at low temperature, the two glide planes (n and
> a) are lost, because they are connected with time-reversal.
> If we call Mn1 and Mn2 the ions with spin up and Mn3 and Mn4 the two
> ions with spin down, the lost symmetries are those connecting Mn1 with Mn4
> (the glide plane n) and Mn1 with Mn2 (the glide plane a), even if these
> last two have the same spin !
> On the contrary, Mn1 and Mn3 are still connected by the mirror plane m,
> even if they have opposite spin (because the spin lies in the glide plane
> and thus changes sign).
> What symmetry group we should specify in the structure file for this
> AFM-phase ???

Instead of answering your question, let me suggest this alternative approach
:

In your original case.struct, split the positions of the Mn-atoms according
to their spin. If e.g. a position with MULT=4 has 2 atoms with spin up and
two with spin dn, make two positions with MULT=2 out of it. Give these two
different spin types a different label. Run now init_lapw, and accept all
structure changes proposed by nn and sgroup. In the end, you should be left
with the correct structure, and case.outputsgroup will even tell you all
symmetry information (=> symmetry information is not necessarily input for
wien, sgroup is often smart enough to recognize the correct space group if
you provide it with enough information about which atoms are magnetically
different). Do not forget to put the correct up and dn occupations for Mn in
the final case.inst file before proceeding with lstart.

Stefaan