Dear Editor,

please find enclosed a manuscript entitled "Narrowband source of correlated
photon pairs via four-wave mixing in a cold atomic ensemble", where we present
an experimental setup and measurements on it.

While interfaces between different quantum systems, like photons and atoms, is
a key ingredient of many proposals for numerous quantum information protocols,
and photon pair sources have been used for a long time to demonstrate such
protocols, there are only very few of them implemented experimentally that
actually allow an interaction with atoms due to a bandwidth mismatch. So far,
parametric conversion in nonlinear optical crystals has been used, where the
optical bandwidth was filtered or engineered to match atomic transitions. Such
techniques are notoriously lossy, and therefore not very bright.

In the work we present here, we use a four-wave mixing process in a
moderately cold atomic ensemble to generate photon pairs that are not only
spectrally narrow, but also matched in their wavelength to an atomic
transition in Rubidium, one of the workhorses for many implementations of
quantum information processes. Compared to similar experiments reported earlier
with atoms in warm or hot gas cells, we are able to observe a photon band
width on the order of 10-20 MHz that is comparable with the atomic
transition. The photon heralding efficiency of 15-20% that is on par with many
pair sources based on three wave mixing in nonlinear optical crystals, and its
brightness is sufficiently high that a number of experiments can be carried
out. In comparison with earlier work in this direction, we are able to
increase the violation of a Cauchy-Schwartz inequality by almost three orders
of magnitude. Therefore, we believe this technique will have a significant
impact on several research directions in quantum information processing. 

As a nice side outcome, the narrow bandwidth does allow an almost
textbook-like demonstration of the thermal character of the emitted radiation
in a Hanburry-Brown--Twiss experiment, which is not easy to observe with other
'thermal' light sources.

Therefore, we believe that Physical Review Letters would be a suitable
platform for this work, and hope for your favorable consideration.

With Best Regards on behalf of all authors,

Christian Kurtsiefer