Single photons - how do we know ?

[Hydr0matic]

If we were to perform a double-slit experiment with single photons, how exactly would we know that the detections registered actually are single photons ?

If there was no duality, and light was just a wave, then there would have to be a limit on how low intensity we could measure, right ? So how do we know there isn't low intensity waves hitting the detector without triggering it ?


Follow-up question:

Performing this single photon experiment one assumes that there actually are single photons. The result of this is crazy quantum weirdness. Why is this experiment held as "displaying the non-intuitive nature of QM", and not as "verifying the wave-nature of light" ? ... to me that seems strange.

 

[quartodeciman]

If you prescribe an experiment with single photons, then you are assuming light consists of quanta. What you need to prescribe instead is an experiment with a very weak light source.

The double-slit arrangement is not designed especially to distinguish wave theory from quantum theory, but you might succeed by using a photographic screen with fine, dense photosensitive crystals in the emulsion. The screen should be set at sufficient distance from the slit rig to be photographically slow. Take a series of time exposures, beginning with small fractions of a second and incrementing the time by small amounts. Develop and fix these photographs. If wave theory prevails, the sequence would probably show the positions of interference maxima early, then later filling in the intermediate spots uniformly, until the full interference pattern emerges, as from disappearance of a fog. I think what actually happens is that the spots for the shortest time exposures appear early in a completely sporadic and random pattern, until the interference pattern emerges statistically later in the series. This favors a quantum theory of light.

Charge Coupled Devices (CCDs), with a matrix of small pixel-like detectors, would work even better, but the design of CCDs assumes the quantum nature of light, so that might be a problem as a clean discriminating test between representations of light.

Photoemission of electrons (photoelectric effect) is a sharper discriminator between wave and quantum theories, but you didn't ask about that.

 

[Hydr0matic]

Originally posted by quartodeciman
I think what actually happens is that the spots for the shortest time exposures appear early in a completely sporadic and random pattern, until the interference pattern emerges statistically later in the series. This favors a quantum theory of light.

Are you telling me that one could actually discriminate a single atom/molecule in this crystal that was hit by a photon ? I know this is what happens with CCD's, but with photographic plates ? u sure ?

 

[quartodeciman]

No, I am not telling you that one could actually discriminate a single atom/molecule in this crystal that was hit by a photon. They are grains that become active, each of which consists of lots of atoms. We can't discuss how light acts on separate atoms or ions without invoking a theory. The important thing about a photo series experiment is the observed pattern of grain activations, whether uniform or random, through the series of photographs. Each theory carries a different expectation. It helps to have a dense emulsion of very fine grained crystals, but activation of single atoms would be asking too much. This is also why I demur from a CCD experiment. It is a theoretical matter how it works, so it isn't really available to us for the purpose of distinguishing light theories.

A scattering experiment, like Compton's scattering of xrays off hydrogen atomic bound electrons, would give one something to look for (scattering angle and energy dispersion of emergent xrays), but it still assumes some scattering mechanism at the atomic level. It is a good encouragement for use of quantum theory, once the wave explanation of light has been discarded.

 

[chroot]

Use an array of photodetectors, and a weak light source. Each click of the photodetectors is a single quantum.

- Warrens

 

[Hydr0matic]

quartodeciman, thnx for two great responses *thumbs up*

So how do you get a light source appropriate for this kind of experiment ? Cause every light source I can think of now seems to strong.. so there must be some kind of filtering process or something, perhaps ?

How is this accomplished without disturbing the uniformity of the beam ?

You don't happen to know the name of someone who has performed this experiment ? ... something I could search for ..

 

[quartodeciman]

I've heard of light sources that emit one photon at a time. Just call it "a very, very weak source".

I've seen plates of photographs series in support of quantum theory- no slits.

Maybe someone else knows about actual performance of an experiment like this. Most experimental setups simply add support to an already assumed quantum theory, rather than establishing it as a preference.

 

[McQueen]

quote:
----------------------------
If we were to perform a double-slit experiment with single photons, how exactly would we know that the detections registered actually are single photons ?
--------------------------------
I think that the point here is that quantum physics is mainly a statistics based theory , so even if there is an approximation of single photons the experiment should still yield verifiable results.

 

[Hydr0matic]

ok, thnx..

Just one more question about the double slit..

When performing this experiment with single photons, has anyone measured the difference in detection rate with and without the slit ? Cause this should be, as you say, a "sharp discriminator" between the theories.

QM predicts no difference in detection rate because all photons go through the slit and hit the detector. Wave theory predicts fewer detections with the slit than without, because some light interferes destructively and doesn't trigger the detector.

Strangely, searching google for such a test didn't turn up anything. Am I missing something ?

 

[thumper]

Quantum mechanics does not describe the motion of photons, that is left to maxwells equations, which do an adequatly good job. Show me the Schroedinger equation for massless particles please.

The profoundity of the double slit experiment comes when you replace the beam of photons with a beam of massive particles, in the first instance electrons.
However when you consider the interaction of an electromagnetic field with matter which is Lorentz invariant, quantum field theory is required.

 

[ZapperZ]

Originally posted by Hydr0matic
quartodeciman, thnx for two great responses *thumbs up*

So how do you get a light source appropriate for this kind of experiment ? Cause every light source I can think of now seems to strong.. so there must be some kind of filtering process or something, perhaps ?

How is this accomplished without disturbing the uniformity of the beam ?

You don't happen to know the name of someone who has performed this experiment ? ... something I could search for ..

Try

http://www.optica.tn.tudelft.nl/Education/photons.htm

Neutral density filters work in most cases when one wants to cut down on the intensity of a light source (assuming the original intensity isn't so high as to burn the filters and the freq is within the working range of the filters). However, we should keep in mind that "single photon sources" usually are not source in which a single photon is emitted and then stop. They are usually sources in which a low number of photons are emitted per second and thus, on average, the photons are separated over a large distance. Sources such as these are typically used in EPR-type experiments.

The other point I want to mention is that single-photon interference (which is what the 2-slit experiment is about) is different than 2-photon and multi-photon interference.[1] Each individual photon passing through the 2-slit assembly acts as if it passed through both slits and interfered with itself, even though it is registering only one "dot" or blob at the "screen" or detector. You only see this effect after many of them pass through the slits and register at the detector.

Zz.

[1] L. Mendel, Rev. Mod. Phys., v.71, p274 (1999).

 

[quartodeciman]

Alas, my photographic experiment proposal isn't so pure an answer to hydr0matic's question.

I assumed that each photon activates one development center on an emulsion. That turns out to be untrue. It takes multiple photons hits to establish a single development center (stable neutral silver atoms within a sea of silver ions). Sorry to mislead anyone!

 

[Hydr0matic]

don' worry about it

.. but what about this:

QM predicts no difference in detection rate because all photons go through the slit and hit the detector. Wave theory predicts fewer detections with the slit than without, because some light interferes destructively and doesn't trigger the detector.

?

 

[MathNerd]

The double-slit experiment was not used to discover the particle nature of light. It was the wave nature of matter that was discovered with the double-slit experiment. Max Plank hypothesized the particle (quantum) nature of light in his study of the ultraviolet catastrophe in the radiation of black bodies. Albert Einstein later used this hypothesis in an explanation of the photoelectric effect and Compton in an explanation of electron-electromagnetic radiation scattering. Both were confirmed by experiment. So the quantum nature of light was experimentally proven via the Compton effect and the photoelectric effect.

 

[MathNerd]

Originally posted by thumper
Quantum mechanics does not describe the motion of photons, that is left to maxwells equations, which do an adequatly good job. Show me the Schroedinger equation for massless particles please.

The profoundity of the double slit experiment comes when you replace the beam of photons with a beam of massive particles, in the first instance electrons.
However when you consider the interaction of an electromagnetic field with matter which is Lorentz invariant, quantum field theory is required.

Maxwell’s equations are classical equations so they don’t even deal with photons and thus they don’t describe a photon’s behavior. Photons being quantum objects are described by quantum mechanics via quantum electrodynamics.

 

[Hydr0matic]

MathNerd, we are discussing the details of the double slit experiment with single photons, not photons in general. Why would you assume we don't already know the stuff you posted ?

Keep it relevant, please ...

 

[quartodeciman]

QM predicts no difference in detection rate because all photons go through the slit and hit the detector. Wave theory predicts fewer detections with the slit than without, because some light interferes destructively and doesn't trigger the detector.

I suspect you meant to say "slits" rather than "slit", since you bring up interference, which requires multiple slits. But if there is destructive interference then there is also constructive interference, augmenting the intensity received at certain places along the detectors and thus offsetting any lowering of total detection rate due to destructive interference.

 

[Phabi]

Originally posted by Hydr0matic

You don't happen to know the name of someone who has performed this experiment ? ... something I could search for ..

I am looking too for references to some real experiment of this kind (I found only thought experiments, which assume the real experiment works and build upon it).

In particular I am interested into the experiments where electrons are sent, one by one, through the two slits and, only when the "observer" observes them passing through one of the two slits, they behaves as particles.

How is it possible to make this observation ? Doesn't it strides with the uncertainty principle ?
Excuse me if the question maybe stupid, but I am new to quantum physics and I am trying to understand some of it...

(I have hundreds of other questions, but before I need to assimilate this concept of "observation" which seems so unnatural)

 

[FZ+]

Doesn't it strides with the uncertainty principle ?

Hmm? This affirms the uncertainty principle - the measurement of the electrons/photons passing the slit collapses the probability wavefunction of the system.

 

[Phabi]

Originally posted by FZ+
Hmm? This affirms the uncertainty principle - the measurement of the electrons/photons passing the slit collapses the probability wavefunction of the system.

I mean: how can you tell exactly which slit the electron is passing through without interfering with it ?

I actually don't buy the idea of a wave that becomes a particle in the very moment we look at it, I am more on the idea that a wave is a particle (whether we look at it or not). We think that waves and particles are two different (and incompatible) things because, in our macroscopic world, we have two concepts, waves and particle, which refer to tangible objects, but we don't have any idea of what things are like at subatomic scale and try to adapt our world conceptions to those objects which behave like particles (which we think of as billiard balls) or waves (which we think of as water waves on a lake) but which instead are quite a different thing.

Our everyday experience is our main obstacle in understanding the world.
For example experiencing the world from the "inside" gives us the misconception that space and time (or mass and energy) are two different things just because our perceptions of these things are different (our perception of time is different from our perception of space).
But everything would be simpler if space and time were the same thing (with time simply a fourth dimension of space without a special difference from the other three or more). It's just the way we perceive time (one instant after the other) which make us think it has a different meaning.
But here I'm going off-topic, sorry for that.

 

[ZapperZ]

Originally posted by Phabi
I am looking too for references to some real experiment of this kind (I found only thought experiments, which assume the real experiment works and build upon it).

In particular I am interested into the experiments where electrons are sent, one by one, through the two slits and, only when the "observer" observes them passing through one of the two slits, they behaves as particles.

How is it possible to make this observation ? Doesn't it strides with the uncertainty principle ?
Excuse me if the question maybe stupid, but I am new to quantum physics and I am trying to understand some of it...

(I have hundreds of other questions, but before I need to assimilate this concept of "observation" which seems so unnatural)

There are a couple of similar experimental results, but done on photons. You may want to check these out:

1. U. Eichman et al., PRL v.70, p.2359 (1993).
2. P. Grangier et al., PRL v.54, p.418 (1985).

Zz.

 

[FZ+]

I mean: how can you tell exactly which slit the electron is passing through without interfering with it ?

That's what I meant. In the case of the changed double slit experiment, the results suggest exactly that - it is impossible to measure the electron/photon without interfering with it. The changing of the interference pattern is an example of the uncertainty principle.

I actually don't buy the idea of a wave that becomes a particle in the very moment we look at it, I am more on the idea that a wave is a particle (whether we look at it or not).

The current thinking is that such entities are neither classical waves, nor classical particles. They simply do not respect our classical models based on macroscopic events in general.

 

[Loren Booda]

The question might be stated: does a measured photon represent the probabilities of all virtual paths leading to its given experimental trial? Or, does such an experiment require an infinite time between sucessive photons to insure their absolute particle nature?