Photon Guns: Is It Possible to Shoot One Photon at a Time?

[student85]

Hi.
I've been wondering if it is possible to have a "gun" that shoots one photon at a time. Does it exist already? If it doesn't, what are the technological limitations that prevent its production? Can you actually build/get one easily or is it so high-tech that only advanced labs have them?
Thanks.

 

[Nabeshin]

It looks like they're pretty difficult to construct, and they have only recently succeeded in creating one. I did a quick google search of "single photon gun" which turned up a ton of articles, so you can look through those.

 

[mgb_phys]

A "gun" that shoots one photon at a time is failry easy.
Take a small light bulb, run it at too low a voltage and put some smoked glass or neutral density filters in front - it's easy to get down to the point where it is emmitting only 1 photon/s or less.

To emit one photon at a time on demand is a little more tricky - you can probably do it with a quantum dot LED

 

[Nabeshin]

A "gun" that shoots one photon at a time is failry easy.
Take a small light bulb, run it at too low a voltage and put some smoked glass or neutral density filters in front - it's easy to get down to the point where it is emmitting only 1 photon/s or less.

Well, it's easy to get down to the point where it's emitting only one photon ON AVERAGE. But you'll still get periods of 2+ photons or zero photons, which apparently is where the complications arise.

 

[f95toli]

A "gun" that shoots one photon at a time is failry easy.
Take a small light bulb, run it at too low a voltage and put some smoked glass or neutral density filters in front - it's easy to get down to the point where it is emmitting only 1 photon/s or less.

Well, that is sort of true. The problem is -as Nabeshin has already pointed out- that light from a light bulb is thermal and has a complettely different "signature" (second order correlation function) than true single photon sources regardless of the intensity.
The "on-demand" problem is a somewhat separate issue and is not direcly related to whether or not a source is a true single photon source of not.

That said, reasonably good single-photon sources do exist and are used routinely in many experiments in e.g. quantum optics.

 

[ZapperZ]

There are loads of papers/reviews on single-photon sources here:

http://www.iop.org/EJ/abstract/1367-2630/6/1/E04

I believe all of them are on the New Journal of Physics, which means that you could get these papers for free.

Zz.

 

[student85]

Thanks for your input.
A follow-up question...
Supposing we have this single-photon gun, what would happen if we do the famous Young (double slit) experiment using this gun as the source? Imagine we point the gun straight into the middle of the two holes and we shoot one photon. What would we see if we placed a screen to the other side? Would we see a single dot? Would we see nothing because the photon never really made it through one of the holes (since we are pointing it to the space in between the 2 holes)?

 

[ZapperZ]

Thanks for your input.
A follow-up question...
Supposing we have this single-photon gun, what would happen if we do the famous Young (double slit) experiment using this gun as the source? Imagine we point the gun straight into the middle of the two holes and we shoot one photon. What would we see if we placed a screen to the other side? Would we see a single dot? Would we see nothing because the photon never really made it through one of the holes (since we are pointing it to the space in between the 2 holes)?

This is no longer "exotic". In fact, it is part of many college physics experiments.

http://www.physics.brown.edu/physics/demopages/Demo/modern/demo/7a5520.htm

Zz.

 

[Cthugha]

This is no longer "exotic". In fact, it is part of many college physics experiments.

http://www.physics.brown.edu/physics/demopages/Demo/modern/demo/7a5520.htm

Zz.

This experiment is however routinely done in the one-photon-on-average approach, not in the quantum optical sense of what single photons mean. However the basic problem is the same in both cases. If the coherence time of your light exceeds the difference in travel time from the source to the slits, those two paths are indistinguishable and you will see an interference pattern. Otherwise you won't.

Using a "on-demand" photon gun you will face the problem that the emission process will be equivalent to a measurement of the emission time and will therefore strongly reduce the coherence time of your emission, so you will only see an interference pattern, if your slits are really close to each other.