Photon Antibunching: Clarifying Inverse Not Necessarily True

[Niles]

Hi

My teacher said that sub-Poissonian statistics (g⁽²⁾(0)<1) implies photon antibunching (g⁽²⁾(t)>g⁽²⁾(0)), but that the inverse is not necessarily true, since g⁽²⁾(t) = 1 for t very large. I am not quite sure I can see why the inverse is not the case.

Can anyone clarify this?


Niles.

 

[Cthugha]

You can find a demonstration of such a case for example in "Transition from Antibunching to Bunching in Cavity QED" by M. Hennrich et al., Phys. Rev. Lett. 94, 053604 (2005). If you do not have access to PRL, you can also find the paper on the pages of the Max Planck institute here: http://www.mpq.mpg.de/cms/mpq/en/departments/quanten/homepage_cms/publications/papers/library/PRL94p053604_Hennrich.pdf"

In that case antibunching without sub-Poissonian statistics occurs because they study the emission of a number of single atoms falling through a cavity and the dependence on the mean atom number. Although each atom is in principle a source of non-classical light and should emit light showing sub-Poissonian statistics, the actual number of photons falling through the cavity is following a Poisson distribution which then alsomanifests in the emission statistics, yielding antibunching, but no sub-Poissonian statistics because the number of emitters is fluctuating.

 

[Niles]

Thanks for the paper -- I will have to read it. My teacher said that a two-mode state being bunched and having sub-Poissonian photon statistics has been constructed. I cannot quite see how this is possible since we know that sub-Poissonian => antibunching, then how does it make sense to have bunched light which is sub-Poissonian?

Thanks for your time.