Localization of waves is the class of phenomena which refer to interferences effects
in wave transport through disordered scattering media. It was first studied in the context
of solid-state physics when Anderson showed, in a seminal paper ,
that propagation could be stopped by destructive interference. This subtle mechanism, now
known as strong (or Anderson) localization, is predicted to occur for sufficiently strong
disorder, i.e. when k l < 1 (k is the wavenumber and l the elastic scattering mean
free path).
Since then, it was realized that interferenceeffects do have persistent effects at low disorder
(a mesoscopic regime known as weak localization) and still hamper
transport. In fact the physical
ingredients of localization, originally introduced for electron waves, apply to any linear
wave and in particular to light. This triggered active research in the
field of optics during the past two decades leading to the observation
of coherent backscattering and universal conductance
fluctuations to cite a few.
A challenge in this field is still the observation of strong localization of visible light. It
was recently reported for near-infrared light using semi-conductors powders
but the validity of the experiments was questioned. Atoms have
been considered as promising scattering media to achieve strong localization.
Indeed they are natural realizations of resonant point-dipole scatterers with
large cross-sections and they constitute
perfectly monodisperse samples. In addition, cooling techniques, which led to BEC observation
in dilute alkali gases, allow high atomic spatial densities of
the order of 1014 at/cm3. This should bring, at least, an ultra-cold
cloud of atoms close to the strong localization transition. However, the first step in this
direction weas only performed recently in the weak localization regime.
In these experiments coherent backscattering (CBS) of light by a cloud of laser-cooled Rubidium atoms was
observed. CBS is an interferential enhancement of the average reflected intensity off a disordered sample
in the backscattering direction. It is a two-wave interference originating from pairs of reciprocal
scattering paths inside the medium. Perfect interference contrast has been
predicted and observd in the helicity preserving polarization channel for classical scatterers with
spherical symmetry.
The detection of the coherent backscattering is now a well estabilshed techinqieu, which
we have adapted to our Rubidium and Strontium traps.
Surprisingly small
enhancement factors were reported and subsequently explained : light scattering by an atomic
dipole transition with any degeneracy dramatically reduces the interference contrast except
for the elementary 0-1 dipole transition. Thus, in order to secure the way to strong localization, it
seems necessary to use atoms with such a simple dipole transition.
Strontium atoms fulfill
this requirement :
The full contrast of the enhancement factor in the helicity preserving channel
for J = O -> J = 1 transition has interesting consequences for
wave localization experiments. It is known that, in the low saturation regime,
scattering of light with an atomic dipole does not destroy the phase information
of the wave. As it was predicted, it is the internal
atomic structure that is at the origin of the small enhancement factor in the similar
Rubidium experiment.
The simplified internal structure of the Strontium atoms have allowed for a
more quantitative
comparison with theory; using a
Monte-Carlo simulation;
which allows also to include the geometrical
effect of our gaussian shaped cloud of scatterers. The excellent agreement between the
experimental data and the Monte Carlo simulation will allow us to study in future more subtle
effects of coherent backscattering which might have a visible effect on the cone shape.
It will be interesting e.g. to study effect due to the velocity of the atoms, correlations between
atoms, magnetic field, laser intensity etc.
Concerning Anderson localization, where interferences
play a crucial role, a J = O -> J = 1 transition appears to be a good
choice. Is it now possible to increase the cloud density to reach the Anderson localization threshold?
Cooling Strontium with the intercombination line in a dipole trap seems to by an
every promising technique as shown by the group of H. Katori in Japan.
CBS with Sr
