[Wien] CXZ and P

[Peter Blaha]

> At beginning, I constructed supercell in P lattice
> structure. But when run 'x sgroup', the structure is
> changed to 'CXZ', and I followed the suggestion on the
> websit of wien2k to change gamma near 90 degree. But
> if gamma equals 91 or 89 degree, it is useless ('x
> symmetry' has error "FORTRAN STOP-3 missing" ). If
> chosed gamma=90.001, 'x symmetry' run and no any
> error. Then I changed gamma in case.struct and
> case.struct_st back to the real degree. My question is
> if it is right if I put gamma like that??
> ( I attached P and CXZ structure and their symmetry
> output files)

I cannot verify your problem. symmetry runs without problems with all the
struct files you sent.

> When I used the above process constructing (113) and
> (115) slab structure (CXZ), there is no any error or
> abnormal during 'init_lapw' and self-iteration.
> Unfortunately, (117) slab structure (CXZ) calculation
> has something wrong during self-iteration. '7
> EIGENVALUES BELOW THE ENERGY -8.00000'. We checked all
> parameters and couldn't find where the error came
> from.  Maybe the structure file of CXZ is

Divergence ?   --> smaller mixing in case.inm
grep :DIS case.scf

Also: You are aware of the possible monoclinic CXZ bug ?
Whenever you run a monoclinic CXZ lattice, you MUST do a comparison with
the corresponding P cell!!

                                      P.Blaha
--------------------------------------------------------------------------
Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-1-58801-15671             FAX: +43-1-58801-15698
Email: blaha at theochem.tuwien.ac.at    WWW: http://info.tuwien.ac.at/theochem/
--------------------------------------------------------------------------