[Wannier] Constrain to the minimization

[xiaochuan Ge]

Dear all,

For example I have a conjugated carbon chain ( CH2-(CH)n-CH2) which I
expect to show two types of wannier functions: Sigma-like orbitals sitting
on carbon-hydrogen and carbon-carbon bonds, and Pi-like orbitals that sits
on two carbons. However, it occurs that wannier90 always mixes the Pi-like
orbital with one of the sigma-like orbitals in order lower the
localization. I understand this may produce more localized orbitals, but
the shape of them have less chemical meaning.

What I would like to do is to separate the occupied KS orbitals into two
manifolds: pi-like orbitals ( they look like the linear combination of
Carbon-pz orbital), and the rest. Then when we I rotate the KS orbitals in
order to produce Wannier functions I do not mix the orbitals between these
two manifolds.

I hope I have explained the question clear enough. Has this kind of thing
or similar already been considered in wannier90? Would anyone suggest me
how to achieve this? Thank you very much.

===================
Dr. Xiaochuan Ge (Giovanni)
Center for Functional Nanomaterials
Brookhaven national laboratory
===================
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/wannier/attachments/20141124/46b66f57/attachment.html>