[Wien] Symmetry error in building type P cell

[Peter Blaha]

You have done almost everythink correct:

supercell does not work with R lattice.

So you created a "supercell" yourself using a primitive cell with three 
times as many atoms according to the Cryst.tables.

a) Please specify lattice type H  (not P)  "hexagonal" lattices get a 
special treatment within WIEN and you need to tell the program this by 
selecting H

b) Of course, according to symmetry the 3 atoms like 
                           Mn (0,0,0), (2/3,1/3,1/3), (1/3,2/3,2/3)
are "equivalent" (and sgroup therefore wants to reduce the cell to the 
smaller R cell). However, as is mentioned in the faqs,... one CANNOT run a 
supercell WIOTHOUT "breaking the symmetry". 
When you neglect this, dstart will try to find symmetry operations which 
transfer these Mn atoms into each other, and of course there is NO such 
symmetry operation present.

c) So you must break symmetry. How to do this depends a little on what you 
actually want to do with such a supercell, but the easiest is to eg. label 
one of the Mn atoms as "Mn1". This tells all WIEN-programs (in particular 
nn and sgroup) that these Mn1 atoms are "different" then Mn atoms (eg. 
because they should get AF ordered or be replaced by impurities, 
vacancies,.....). The setup will no longer make these (and others will 
follow automatically) atoms equivalent, but they will be independend atoms 
resulting in reduced symmetry.
Another type of symmetry break would be to move one of the atoms slightly 
away from the ideal position.

As I said, it depends what you actually want to do. 

> Since the supercell program won't run on R lattice, I'm setting up my R
> lattice from a P type structure.
> My material: #166(R-3m),
> Nonequivalent atoms: 3
> Atom positions: O (0,0,0.26), Mn (0,0,0), Li (0,0,0.5) in rhombohedral
> coordinates.
> This structure works well for initialization and scf cycles.
> 
> What I did:
> lattice type: P
> angles: 90,90,120
> nonequivalent atoms: 3
> atom positions: O (0,0,0.26), (2/3,1/3,0.26+1/3), (1/3,2/3,0.26+2/3)
>                              (0,0,0.74), (1/3,2/3,0.74-1/3),
> (2/3,1/3,0.74-2/3)
>                           Mn (0,0,0), (2/3,1/3,1/3), (1/3,2/3,2/3)
>                        Li (0,0,0.5), (1/3,2/3, 0.5-1/3), (2/3,1/3,0.5+1/3)
> These are all atoms in the P cell with distance shorter than one lattice
> vector.
> Then I started initialization.
> The x sgroup found the right space group #166, but I still used my
> manually generated .struct file for all the following steps. However,
> problems occured when running x dstart. The dstart.error said:
>  'ROTDEF' - no symmetry operation found.
>  'ROTDEF' - for jatom, index 1 2
>  'ROTDEF' - atomposition of jatom   0.0000000   0.0000000   0.2600000
>  'ROTDEF' - atomposition of index   0.6666667   0.3333333   0.5933333
> 
> I carefully checked the atom positions, all were correct and positive.
> Seems Wien2k tends to think oxygen atoms are not all equivalent. But they
> are equivalent. Actually, I tried to separate oxygen into O1 and O2 ( the
> 1st line and the 2nd line). It didn't hel. I have no idea what's
> happening.
> 
> Please give your advice.
> 
> Many thanks
> 
> Shu Miao
> 
> 
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
> 


                                      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/
--------------------------------------------------------------------------