Friday, January 16, 2009

New Quantum Camera

Information is physical because it takes energy to create it and transform it. Instead of the digital bits that you're used to thinking about in the terrestrial computing world, quantum information technologies (QIT) involve encoding information as quantum bits or qubits. The photon has turned out to be a very amenable quantum particle for encoding qubits. For anyone following my progress in the world of QIT, here's the latest. Our invited paper: "A Quantum Imager for Intensity Correlated Photons," which describes a new type of camera has now been published in the European open-access publication New Journal of Physics.

All cameras are quantum, in that all light is a quantum phenomenon resulting from the action of photons. This has been the correct description of light ever since Einstein's revolutionary 1905 paper (PDF) for explaining the photoelectric effect (amongst other things). Amazingly, even after 100 years of highest-precision experiments, some physicists have not accepted this particle model of light. Ironically, these same people happily talk about the electron as a quantum particle. Sigh!

Anyway, conventional cameras, and your eyes, only respond to differences in the intensity of light. In the quantum world, intensity corresponds to the average number of photons arriving at the image plane or your retina. What is less well known is that photons produced by most naturally occurring light sources are a little bit "sticky" in that there's a 1-in-a-million chance (at room temperature) that when one photon arrives at your retina, it will be accompanied by another photon. In other words, photons of pheather phlock together; an effect which neither your eye nor a conventional camera can image. Photon-pairing does not occur in all light sources, however. For example, laser light does not contain sticky photons because the artificial production method eliminates the effect.

Quantum stickiness is a result of the photon belonging to a class of quantum particles called bosons, which simply means they have whole-number values of intrinsic angular momentum (spin). In particular, the photon has a spin value of 1; in contrast to the electron which has spin 1/2 and is therefore not a boson. It's this spin-1 that gives light its polarization properties and makes polaroid sunglasses effective. The simultaneous arrival of photon-pairs constitutes a certain kind of statistical correlation in bosons and our new camera is able to "see" these intensity correlations.

One very important application of our new camera could be the validation of a weird, ultra-cold state of matter called a Bose-Einstein condensate (BEC). Although a BEC is unnatural, it can be produced in the lab. Irrespective of how you produce it, you need to be sure that what you're looking at is indeed a BEC. One way you can be sure, is to take a picture of it with our camera (not yet available at Best Buy). No image will appear because, like a laser, a true BEC does not produce sticky photons. Reminiscent of Michelson-Morley, this is another one of those potentially useful null results in physics.

Update: Our paper was slashdotted on Sunday January 25, 2009!

3 comments:

Vincent said...

Nice post. Here is a tangential question. Is it possible to create a light source that will efficiently produce many of these photon pairs.

Neil Gunther said...

See Table 1 in our paper. The implication is that you need 2 cameras; the
regular kind and ours. Currently, researchers are just using a regular
camera. In the context of BEC detection, you must have homogeneous matter
(e.g., atoms, polaritons) in order to facilitate potential condensate
production. The question becomes: What state is that homogeneous matter in?

In principle, a non-condensate blob (call it "!BEC") could act as either a
fully incoherent or a *semi-coherent* light source. A fully incoherent blob
would wash everything out, so no interference "image" would appear in
either camera. That would include your !BEC and not sticky. (Col 2 of Table
1) On the other hand, a semi-coherent !BEC blob will produce an
interference "image" in *both* cameras. (Col 4 of Table 1)

If, however, you attain the right conditions and some of the semi-coherent
blob now transitions to a genuine condensate, those areas of BEC within the
!BEC blob will act as fully *coherent* sources (by definition), at which
point that part of the "image" will disappear in our camera but remain in
the regular camera. (Col 3 of Table 1)

Neil Gunther said...

Because the relatively low production rate was a killer for some of our
experiments, I know we discussed this topic. My recollection is that it's
not so easy to increase the rate of super-Poisson photons. Without
getting into all the details, it's a Goldilocks problem. You don't want a
source that is completely coherent (like a laser) and you don't want a
source that is completely incoherent (like a light bulb). You want something
in between that's just right (like a Hg lamp), without reducing your overall intensity.

But never say never. A similar issue concerned the production rate of
entangled bi-photons. The first detection ca. 1948 involved the decay of
positronium. This, and later methods, have since been replaced by
spontaneous parametric down conversion using birefringent crystals,
http://en.wikipedia.org/wiki/Spontaneous_parametric_down_conversion .
Quantum dots might turn out to be useful, http://physics.technion.ac.il/~dg/PAPERS/PRL/prl01.pdf, although they're more usually associated with sub-Poisson sourcing.