[Octopus-users] luminescence

[David Strubbe]

It sounds like you are thinking of starting the system in some excited
state, and seeing how long it takes until it emits a photon and goes back
down to the ground state.  But there are no photons in an electronic
structure framework, because doing that requires treating not merely quantum
mechanics of electrons but also quantum electrodynamics of the
electromagnetic field.  You will never see the system suddenly induce a
uniform (external!) electric field in the simulation box.  And even if you
could, the timescale is not a fixed wait time, but rather like a half life,
so you would need to do very many simulations to plot the different randomly
distributed wait times into an exponential decay.  How many timesteps you
can feasibly do in Octopus will not be the limiting factor here, but rather
that the underlying physics is not described in the way it would need to be
to do a direct calculation of fluorescence.

Instead, why not take advantage of time-reversal symmetry?  Fluorescence is
simply absorption backwards -- but could be from a different electronic or
atomic structure, which is what gives rise to differences between absorption
and fluorescence spectra.

David

On Thu, May 14, 2009 at 9:20 AM, david marlan <davma7 at gmail.com> wrote:

> Colleagues!
> Octopus allows to calculate absorption spectras as written in manual.
> Is it possible to
> calculate luminescence of the system?
>
> Ok, one difficulty is clear. Since the luminescence time scale can be
> from femtoseconds
> to nanoseconds and higher (phosphorescence  can be many times longer) it is
>  a
> computationally intensive  proccess (if I correctly understood time
> step of TDDFT
> method should be always sufficiently small to maintain stability,
> since that fact
> the number of iterations must be very large).
>
> But in principle is it possible by current octopus? (having
> computational resources   etc.)
>
> Thanks!
> _______________________________________________
> Octopus-users mailing list
> Octopus-users at tddft.org
> http://www.tddft.org/mailman/listinfo/octopus-users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.tddft.org/pipermail/octopus-users/attachments/20090514/90f40386/attachment.html