[Wannier] Preparation of the projections block

Jonathan Yates jonathan.yates at materials.ox.ac.uk
Mon Aug 2 23:27:26 CEST 2010

```On 2 Aug 2010, at 21:53, Jian-Xin Zhu wrote:

> Dear Jonathan and Respectful Wannier Users,
>
> When I prepare the projections block in the case.win file, a question arises ---
>
> If I have a 2 types of Vanadium atoms in the unit cell for ab-inition calculations and each type has 4 equivalent V atoms,
> how many lines should I put in the
> Begin Projections
> ...
> End Projections
> block?
>
> From the Wannier90: User Guide, I read on page 34 "C, Al etc. applies to all atoms of that type".
> So I understand I should put in 2 lines since there are two types of Vanadium. As such, I prepare the
> projections block as follows:
>
> Begin Projections
> V: l=0;l=1;l=2
> V: l=0;l=1;l=2
> End Projections

Jianxin,

That's not quite correct.

Wannier90 doesn't know anything about the symmetry equivalence of the atoms.

So if you put

Begin Projections
V: l=0;l=1;l=2
End Projections

you will get s,p,d projections on each Vanadium atom in your system (ie all 8 in your example).
If this is what you want, this is sufficient.

You might have a situation where you do want different projections for the inequivalent V (maybe if you had an antiferromagnetic system). In which case you could do something like

Begin atoms_cart
V1 0.00 0.00 0.00
V1 0.00 0.00 0.50
V2 0.25 0.25 0.25
V2 0.25 0.25 0.75
end atoms_cart

Begin Projections
V1: l=0;l=1;l=2
End Projections

which puts projections on only half of the V.

Yours
Jonathan

(ps. apologies for the repeated email - I forgot to send it to the list the first time)

--
Department of Materials, University of Oxford, Parks Road, Oxford, OX1 3PH, UK
tel: +44 (0)1865 612797                http://users.ox.ac.uk/~oums0549/
```