[Wien] exciton trapping

[Fabiana Da Pieve]

Hello wien2k users

I am maybe going to ask a trivial or a totally non-sense question, but I am
trying to do what I can with what I have.
I know that at the moment BSE calculations with wien2k (and other codes)
are very computationally demanding, and in any case the BSE part is still
not released.

In any case, even if it would be released, I think I could not run it on my
system because it is very heavy.
However, I would like to find a way (even a very simple one) to see if a
certain optical excitation would lead to a localized exciton (which
localizes because of local distortion for example) or if such optical
excitation would not lead to any distortion. I would like to study this
because sometimes excitonic transitions could change bond lengths and
change (slightly, I think) the covalency of a certain bond, and in such
distortion the exciton could be trapped, which is want I want to see.

So, I was thinking that MAYBE I could have "a first idea" of the eventual
distortion induce by my optical ideal exciton already at the DFT level in
the following way:

1) I remove an electron from one of the external levels of the anion in
which I believe the hole is created in the file......(ok, very bad thing,
optical excitations cannot be described as localized on a single site at
the beginning)
2) at this point, either I smear the charge uniformly in space  (or maybe I
could explicitly put an electron more in the orbital of the cation which I
know is the acceptor orbital ??)
3) instead of running a normal scf calculation, I run a structural
optimization calculation (keeping the lattice parameters fixed, i.e. moving
only atoms)
4) at the end I analyze the changes in the bond lengths near the atom

Is it correct (at least as a very zero approximation to the problem) ?
totally wrong ?
Any suggestions is welcome !

Thanks !
Fabiana
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20120301/ff10368f/attachment.htm>