Quantum double slit experiment

[thgiepsluap]

Hi,

I've been reading about an experiment, where photons are fired through a double slit, and they act as if they go through both slits, by interfering with themselves, and creating an interference pattern on a screen behind.

Then if a measuring device is used to see which slit the photon goes through, it only goes through one, and doesn't create an interference pattern.

Some people suggest that the measuring device physically causes the photon to act this way.

Others suggest that it is the act of observing that does this, and that when a measuring device is left in place, but no-one is observing the results, then the photon still acts as if it goes through both slits.

please can someone help me out, as to what actually happens.

Also, where can i find a 1st hand account of this experiment?

many thanks.

Paul Speight

 

[lugita15]

Hi,

I've been reading about an experiment, where photons are fired through a double slit, and they act as if they go through both slits, by interfering with themselves, and creating an interference pattern on a screen behind.

Then if a measuring device is used to see which slit the photon goes through, it only goes through one, and doesn't create an interference pattern.

Some people suggest that the measuring device physically causes the photon to act this way.

Others suggest that it is the act of observing that does this, and that when a measuring device is left in place, but no-one is observing the results, then the photon still acts as if it goes through both slits.

please can someone help me out, as to what actually happens.

Also, where can i find a 1st hand account of this experiment?

many thanks.

Paul Speight

Contrary to what is often claimed in popular physics books, it is not the magical act of human "observation" that changes how the photon behaves. A measuring device is enough. As long as it is possible, *in principle*, to find out which slit the photon went through, the interference will go away. It doesn't matter that no observations actually took place, only that they could have taken place.

 

[Turtle492]

Contrary to what is often claimed in popular physics books, it is not the magical act of human "observation" that changes how the photon behaves. A measuring device is enough. As long as it is possible, *in principle*, to find out which slit the photon went through, the interference will go away. It doesn't matter that no observations actually took place, only that they could have taken place.

Surely then the invention of a suitable measuring device would make it in principle possible to find out which slit a photon went through? Just having one in the same room means that an observation *could* have taken place, what is the difference to the particle of whether you couldn't be bothered to take the actual measurement or whether you couldn't be bothered to set the equipment up? What if the measuring device is there but not switched on?

 

[lugita15]

Surely then the invention of a suitable measuring device would make it in principle possible to find out which slit a photon went through? Just having one in the same room means that an observation *could* have taken place, what is the difference to the particle of whether you couldn't be bothered to take the actual measurement or whether you couldn't be bothered to set the equipment up? What if the measuring device is there but not switched on?

I think you misunderstood me. Let me rephrase what I was trying to say. After the double slit experiment has been performed, if it is in any way possible to determine which slit the photon went through then there will be no interference. This is true regardless of whether someone actually tries to find out. Likewise, if, after the experiment has been performed, it is impossible to ever find out which slit the photon went through, then there will be interference.

To put it another way, what matters is whether the information exists, not whether it is known. Interference will be there as long as the information about the path of the photon does not exist. It would work the same way even if humans didn't exist.

This perspective aids in the understanding of many quantum mechanics concepts, such as Heisenberg's uncertainty principle: if a lot of information about the position a particle exists, not that much information about it's momentum exists.

 

[arkajad]

When you model this experiment on your PC, and it is easy to do it, you see the following.

First you have the source. It prepares quantum state of the particle. The you have the slits. You model the slits either as a pure Hamiltonian interaction or it may have some detection capacity. If it has some detection capacity, the interaction of the slits with the wave function is different - wave function will evolve differently with time. You can easily this and see the difference on the screen if you add, for instance, small detectors of positions or momenta around the slits. Then you have the screen. The screen has the detection capability and it reacts to the wave function when it reaches it. It also causes its collapse. Different parts of the screen reacta, after a somewhat random time, with the wave function, and the probabilities depend on the shape of the wave function and its change with time. This shape and its change with time depend on what was shaping the wave function before - that is on the Hamiltonian and non-Hamiltonian interactions before. One run of your simulation - you get one dot. Millions runs - millions dots - but some of the runs may give non-detection. Anyway after many runs you get the interference pattern. You can now play with the screen part or with the slit part, change its parameters and see how the pattern on the screen changes.

So, you have a computational model of the phenomenon. It describes pretty well what you observe.

Now, where is the particle? You do not need it in the simulation experiment. If you wish, you can add to that some Bohmian trajectory, but it does not play any role in this approach. All you need is the wave function and its interaction with the slits and with the screen. You get in this way exactly what you see in a real experiment.

You can model the same way tracks in a cloud chamber. Again, you do not need a particle. All you need is the interaction of the wave function with detectors - this cause repeated collapses of the wave function (which we do not see) and detectors reactions that accompany these collapses - which we can record and see. You can simulate this way particle tracks, including their timing. But there are no particles there! A mystery? Perhaps.

 

[physics pfan]

The screen has the detection capability and it reacts to the wave function ...It also causes its collapse.

Now, where is the particle? You do not need it ...All you need is the wave function and its interaction with the slits and with the screen. ... Again, you do not need a particle. All you need is the interaction of the wave function with detectors

Regretfully I find some problems in this post. First, "wave function collapse" is normally equivalent to "collapse of the state vector Ψ" and |ψ|2 gives you the probability of finding an electron in a specific location whereas a photon is subject to Maxwell's equations and the probability of a photon in a location is proportional to the square of the radiation energy density at that location. So I think this post has confused computational ψ waves with real-world radiation energy waves. Not the first time this has happened.

A second problem is the easy identification of photon with particle. There is scant evidence for the photon as particle aside from the lazy assumption that anything traversing space and terminating at a point must be a particle. As to why this "particle" should exhibit all sorts of wave behavior before its "particle" termination, many don't want to be bothered with that.

So the sensible answer as to why a photon can pass through both slits is that it is a wave and a wave, unlike a particle, can subdivide (and later rejoin and interfere).

The real question is why an electron can do the same thing. That is 1) tied up with the wave nature of the moving electron and 2) the subject of a different thread.

Regards,

PP

 

[FREEDOM2]

When you model this experiment on your PC, and it is easy to do it, you see the following.

First you have the source. It prepares quantum state of the particle. The you have the slits. You model the slits either as a pure Hamiltonian interaction or it may have some detection capacity. If it has some detection capacity, the interaction of the slits with the wave function is different - wave function will evolve differently with time. You can easily this and see the difference on the screen if you add, for instance, small detectors of positions or momenta around the slits. Then you have the screen. The screen has the detection capability and it reacts to the wave function when it reaches it. It also causes its collapse. Different parts of the screen reacta, after a somewhat random time, with the wave function, and the probabilities depend on the shape of the wave function and its change with time. This shape and its change with time depend on what was shaping the wave function before - that is on the Hamiltonian and non-Hamiltonian interactions before. One run of your simulation - you get one dot. Millions runs - millions dots - but some of the runs may give non-detection. Anyway after many runs you get the interference pattern. You can now play with the screen part or with the slit part, change its parameters and see how the pattern on the screen changes.

So, you have a computational model of the phenomenon. It describes pretty well what you observe.

Now, where is the particle? You do not need it in the simulation experiment. If you wish, you can add to that some Bohmian trajectory, but it does not play any role in this approach. All you need is the wave function and its interaction with the slits and with the screen. You get in this way exactly what you see in a real experiment.

You can model the same way tracks in a cloud chamber. Again, you do not need a particle. All you need is the interaction of the wave function with detectors - this cause repeated collapses of the wave function (which we do not see) and detectors reactions that accompany these collapses - which we can record and see. You can simulate this way particle tracks, including their timing. But there are no particles there! A mystery? Perhaps.

What is a Hamiltonian Interaction?

 

[arkajad]

What is a Hamiltonian Interaction?

Hamiltonian interaction is an interaction described by http://en.wikipedia.org/wiki/Schr%C3%B6dinger_equation" from the textbooks

with a self-adjoint Hamiltonian $$\hat{H}=\hat{H}^\dag$$.

It implies a reversible, completely deterministic evolution, continuous in time - no collapses, no events, no catastrophes, nothing spectacular really ever "happens".

 

[ZapperZ]

Regretfully I find some problems in this post. First, "wave function collapse" is normally equivalent to "collapse of the state vector Ψ" and |ψ|2 gives you the probability of finding an electron in a specific location whereas a photon is subject to Maxwell's equations and the probability of a photon in a location is proportional to the square of the radiation energy density at that location. So I think this post has confused computational ψ waves with real-world radiation energy waves. Not the first time this has happened.

A second problem is the easy identification of photon with particle. There is scant evidence for the photon as particle aside from the lazy assumption that anything traversing space and terminating at a point must be a particle. As to why this "particle" should exhibit all sorts of wave behavior before its "particle" termination, many don't want to be bothered with that.

So the sensible answer as to why a photon can pass through both slits is that it is a wave and a wave, unlike a particle, can subdivide (and later rejoin and interfere).

The real question is why an electron can do the same thing. That is 1) tied up with the wave nature of the moving electron and 2) the subject of a different thread.

Regards,

PP

This makes very little sense due to the presence of many which-way experiments. If a wave "sub-divides", then a detection will occur on BOTH slits, not one or the other, when detectors are placed at both paths.

If you are proposing some new physics that hasn't been published before, this is the wrong forum to do it (re-read the https://www.physicsforums.com/showthread.php?t=414380" that you have agreed to).

Zz.

 

[Phrak]

I think you misunderstood me. Let me rephrase what I was trying to say. After the double slit experiment has been performed, if it is in any way possible to determine which slit the photon went through then there will be no interference. This is true regardless of whether someone actually tries to find out. Likewise, if, after the experiment has been performed, it is impossible to ever find out which slit the photon went through, then there will be interference.

To put it another way, what matters is whether the information exists, not whether it is known. Interference will be there as long as the information about the path of the photon does not exist. It would work the same way even if humans didn't exist.

Very well. I have an instrument, that measures the change in momenta of both slits imparted by each photon passing through slit 1 and slit 2. I stick this information away somewhere. Whether I access this information or not has no effect on the photographic plate I've put behind the slits.

 

[rahuljayanthb]

the double slit experiment highlights the wave-particle duality.
at the time of creation and annihilation,the particle behaves as a particle.and in between as a wave.so the after creation,the particle travels as a wave and since there are two possibles paths(two slits) the wave (and wave function) splits in two and each wave passes through each slit.when a detector is placed,it will have a probability of 1/2 of finding it and when detected it is effectively annihilated thus behaves as a particle.

 

[ZapperZ]

the double slit experiment highlights the wave-particle duality.
at the time of creation and annihilation,the particle behaves as a particle.and in between as a wave.so the after creation,the particle travels as a wave and since there are two possibles paths(two slits) the wave (and wave function) splits in two and each wave passes through each slit.when a detector is placed,it will have a probability of 1/2 of finding it and when detected it is effectively annihilated thus behaves as a particle.

This is misleading, and perpetuating a myth that one needs two different descriptions to fully describe such a phenomenon. One does not. A single, consistent description of light for such a setup exists within the QM. There is no "duality" within QM. Read our FAQ in the General Physics forum on this topic, and then read the Marcella paper. There is no need to switch gears to go from wave to particle, or vice versa.

http://arxiv.org/ftp/quant-ph/papers/0703/0703126.pdf

Zz.

 

[arkajad]

http://arxiv.org/ftp/quant-ph/papers/0703/0703126.pdf[/url]

Zz.

I am reading there:

It can be scattered with anyone of the continuum of momentum eigenvalues $$p_y = p \sin \theta,$$ where $$-\pi/2\leq\theta\leq\pi/2.$$

Measurement of a well-defined scattering angle $$\theta$$ constitutes a measurement of the observable $$\hat{p}_y$$ and, therefore, the basis vectors in Hilbert space are the momentum eigenvectors $$|p_y>$$ .

My question: why measuring of the angle is considered to be the same as measuring the component of the momentum?

 

[ZapperZ]

I am reading there:



My question: why measuring of the angle is considered to be the same as measuring the component of the momentum?

I'm not sure I understand the problem that you've having with this.

You know the total momentum of the photon. Since it is moving at an angle from the normal, knowing the angle will give you the transverse component of the momentum. Rather straight-forward to me.

Zz.

 

[arkajad]

First, we do not know its total momentum after the scattering. Unless we assume something about the scattering that has not been explicitly stated. But should have been.

But then we have another problem. The angle theta refers to a coordinate system in momentum space, not in position space. But real measurement are done with the screen in x-y space. Again, this should have been explicitly stated but was not in the paper. Any real screen, put at a certain distance from the slits will not show the calculated distribution. Theta measured with respect to what?

The paper therefore assumes the screen is to be put infinitely far from the slits and should be infinitely long. Which should have been stated but was not.

With these two amendments the paper would be better, I think.

 

[ZapperZ]

First, we do not know its total momentum after the scattering. Unless we assume something about the scattering that has not been explicitly stated. But should have been.

Unless you want to make the assertion that the frequency changes after it passes through the slit, then since the speed remains the same, the momentum hasn't changed.

But then we have another problem. The angle theta refers to a coordinate system in momentum space, not in position space. But real measurement are done with the screen in x-y space. Again, this should have been explicitly stated but was not in the paper. Any real screen, put at a certain distance from the slits will not show the calculated distribution. Theta measured with respect to what?

The paper therefore assumes the screen is to be put infinitely far from the slits and should be infinitely long. Which should have been stated but was not.

But even if this is an angle in momentum space, I'm not sure what the problem is in obtaining an angle in real space. Condensed matter physicists do this all the time, i.e. going from reciprocal space to real space. There is no difficulty in obtaining such an angle. In fact, this is how we obtain the band structure of occupied states using ARPES, where the angle of spread corresponds to momentum.

Zz.

 

[arkajad]

Unless you want to make the assertion that the frequency changes after it passes through the slit, then since the speed remains the same, the momentum hasn't changed.

There is no term "frequency" in the paper. There are terms "particle" and "momentum". The author should state it explicitly that he restricts himself to an ideal elastic scattering.

But even if this is an angle in momentum space, I'm not sure what the problem is in obtaining an angle in real space.

The problem is that the author should have stressed that his calculation are not valid if the screen is too close to the slits. They are exactly valid only if the screen is infinitely far and infinitely long.

Perhaps the author himself did not realize this fact - otherwise, I am sure, he would have pointed it out. The paper is so precise in all other respect, why not to be precise and pedagogical (not to mislead readers) also in these two respects? The clarity could only gain from these two additions - don't you think so?

Or am I in error? Would it mess up such an otherwise nice paper?

 

[unusualname]

That paper uses position eigenstates so isn't valid for relativistic particles like photons.

Did he compare his results with a real experimental setup and real interference patterns? It seems to me he's just derived a rather trivial generic wave interference result, which applies qualitatively (and approximately) to any slit experiment.

 

[arkajad]

Moreover, with, say, electrons, I am not sure how anything can be derived using this method if we add a deflecting electric field somewhere between the source and the screen - assuming they are at a finite distance.

 

[physics pfan]

This makes very little sense due to the presence of many which-way experiments. If a wave "sub-divides", then a detection will occur on BOTH slits, not one or the other, when detectors are placed at both paths.

If you are proposing some new physics that hasn't been published before, this is the wrong forum to do it (re-read the https://www.physicsforums.com/showthread.php?t=414380" that you have agreed to).

Zz.

Kindly explain what you mean by “which-way experiments.” This is a learned, but non-specialist forum; if you are referring to the Mach–Zehnder interferometer experiments then say so. If not, then explain.

I assume that when you wrote “slits” you actually meant “paths.” If you place detectors immediately behind both slits there is no interference and nothing of interest to measure.

I suggest you re-read section 2.2 of “The Quantum Challenge” by Greenstein and Zajonc where they describe two experiments by Grainger, Roger and Aspect. The first experiment had a beam splitter and photon detectors on each of the two paths. When a single photon went through the beam splitter it was detected on one path or the other, never on both. This, however, does NOT prove that the other path was not taken; what it does prove is that the other path, if taken, was not selected for photon termination/detection. This can be seen by the second experiment of Grainger, Roger and Aspect. Here they sent the single photon through a Mach–Zehnder interferometer and found conclusive proof of a single photon interfering with itself (i.e., traveling both paths). So yes a wave (photon) can sub-divide (follow all paths) and still select only one path for termination/detection.

Pp

 

[Cthugha]

I suggest you re-read section 2.2 of “The Quantum Challenge” by Greenstein and Zajonc where they describe two experiments by Grainger, Roger and Aspect. The first experiment had a beam splitter and photon detectors on each of the two paths. When a single photon went through the beam splitter it was detected on one path or the other, never on both. This, however, does NOT prove that the other path was not taken; what it does prove is that the other path, if taken, was not selected for photon termination/detection. This can be seen by the second experiment of Grainger, Roger and Aspect. Here they sent the single photon through a Mach–Zehnder interferometer and found conclusive proof of a single photon interfering with itself (i.e., traveling both paths). So yes a wave (photon) can sub-divide (follow all paths) and still select only one path for termination/detection.

This is not true. Reread the common literature on the subject. Common antibunching experiments show that single photons are always detected at only one exit port after a beam splitter. If a photon had the ability to subdivide, you could place two detectors behind a beam splitter, fire single photons at it and would expect some coincidence detections due to the photon splitting in two. You never see those effects, so a photon cannot subdivide the way you suggest.

Please note that the signature of a single photon interfering with itself does not mean that it subdivides and travels along both paths. The concept of a well-defined photon path is already ill-defined in this context. It is the probability amplitudes for the several possible events which interfere, not the actual particles themselves. See for example Roy Glauber's Nobel lecture "one hundred years of light quanta" (frrely available in the net) for details.

 

[gtwace]

Contrary to what is often claimed in popular physics books, it is not the magical act of human "observation" that changes how the photon behaves. A measuring device is enough. As long as it is possible, *in principle*, to find out which slit the photon went through, the interference will go away. It doesn't matter that no observations actually took place, only that they could have taken place.

Part right and Part wrong i guess, due to the delay quantum eraser experiment. The detector will remove the interference but only if the data is kept. If say the detector recording the data is destroyed before the data is read and the image screen is not viewed (viewing the image screen with detector on will breakdown the interference pattern), then an interference pattern appears when you look at the image screen after that. Therefore, "knowledge" changes how the photon behaves.

To answer whether human observation changes how the photon behave, we have to unfortunately sacrifice a human. The experiment will go like this, we do the double slit experiment with the detector collecting the data but not looking at the image screen. A person then reads the data (now only he knows, which proton passes through which slit) but he does not look at the image screen. Then he kills himself without leaving any clue as to which slits the photon passes through. A second person then comes into the room and look at the image screen. If the image screen is no interference pattern, then human observation does affect how the photon behave, if there is a interference pattern, then human does not matter, it is just a matter of "knowledge".

 

[Cthugha]

Part right and Part wrong i guess, due to the delay quantum eraser experiment. The detector will remove the interference but only if the data is kept. If say the detector recording the data is destroyed before the data is read and the image screen is not viewed (viewing the image screen with detector on will breakdown the interference pattern), then an interference pattern appears when you look at the image screen after that. Therefore, "knowledge" changes how the photon behaves.

Sorry, but this is completely wrong. This is a wrong point of view mediated in some popular media layman explanations of DCQE. In Reality in DCQE you will never see an interference pattern if the data is not kept. The interference pattern is only available in the coincidence counts from both detectors. Just in the single counts at one detector, there will never be an interference pattern (unless you destroy entanglment of course). You just rob yourself of the opportunity to see the coincidence count dataset containing an interference pattern. Knowledge does not change how the photon behaves.

 

[azaharak]

From my understanding of the quantum eraser experiment...


you need to erase the recorded information before your photons are even detected at the screen. The experiment is designed with idler photons which tell information about the true path taken. If they are scrambled before seeing the screen results the interference pattern emerges.

If the screen is hit by the photons, it doesn't matter if you haven't looked at the information yet. You only see the preservation of interference if you destroy the data before the photons hit the screen.

AZ

 

[Taze]

Correct me if I'm wrong,

But wouldn't the screen that the photons are fired at classify as a measuring device? And if not, couldn't they make the object with the slits out of the same material as the screen to gather further information?

 

[azaharak]

Yes

the "screen" is a measuring device, if you want to see the interference pattern you must have a record of it.

But measuring interference on the screen doesn't tell you anything about which slit the photon traveled through, it only tells you that the photons path is a superposition of the two possible paths.

My question is...


Classical diffraction can be viewed as incomming light scattering with the slit medium instead of bending or a hyguens method.

could this be extended to the quantum picture? would you say you have a superposition of scattering with many atoms?

 

[taylaron]

Interesting thread indeed.

My question regarding the double slit experiment is whither or not the experiment has been tried using a wide variety of different materials in construction of the slits.
I've been unable to find any resources describing this potential variable.

Regards,
-Taylaron

 

[ZapperZ]

Interesting thread indeed.

My question regarding the double slit experiment is whither or not the experiment has been tried using a wide variety of different materials in construction of the slits.
I've been unable to find any resources describing this potential variable.

Regards,
-Taylaron

Why would this matter? After all, in SQUID experiments, what is the nature of the material of the "slit" there? Yet, you get the same result!

Zz.

 

[taylaron]

Why would this matter? After all, in SQUID experiments, what is the nature of the material of the "slit" there? Yet, you get the same result!

Zz.

To my understanding, the diffraction of electrons in a double silt experiment was completely unexpected by physicists. This experiment resulted in the formation of the particle-wave duality concept which was incorporated into quantum mechanics.

Although I do not have a legitimate reason why, I would be curious to explore the use of a wide range of grating materials as its effects on the diffraction pattern may be as shocking as the original experiment itself. I do not know how or why the interference would act differently, it is simply a matter of curiosity for me.

Back to my original question; (regardless of why…) has the double slit experiment been made using many different slit materials?

 

[azaharak]

Quoted from https://www.physicsforums.com/showthread.php?t=418772



"Yes that's true. That's actually true of any kind of scattering, diffraction, refraction or any -actiony thing you come up with in classical electromagnetics. There is a principle called the equivalence principle in electromagnetics that allows you to redefine a scattering problem as a homogeneous medium with your original incident field going through unperturbed and a secondary field (the scattered field) being produced by sources in the medium. The sum of the two gives you the total field that you would observe in the actual problem. The idea is that when the incident field impinges on a scatterer it excites conduction and displacement currents on and throughout the scatterer. These currents produce their own electromagnetic waves that are called the scattered field. The superposition of the scattered and incident waves gives you the diffracted, reflected and refracted waves that you would normally observe. And these are not ficticious sources either, the currents that we are talking about are the same currents that are exciting on say a receiving antenna."



Whether or not it should or shouldn't depend on the medium isn't my question, (it shouldn't I'm thinking).

If you picture one photon as scattering with the medium itself, wouldn't you conclude it scatters with many atoms, (a super position of scattering).

I don't know... too much going on.

 

[arkajad]

Isn't each diffraction pattern a multi slit experiment? There are all kinds of diffracting grids. Or do you mean one or two photons (at most three?) at a time?

 

[ZapperZ]

To my understanding, the diffraction of electrons in a double silt experiment was completely unexpected by physicists. This experiment resulted in the formation of the particle-wave duality concept which was incorporated into quantum mechanics.

Come again?

"Wave-particle" duality is a classical concept, because in classical physics, the wave-like and particle-like phenomena are two separate description. In QM, there is no such thing as wave-particle duality. Read the FAQ in the General Physics forum.

Although I do not have a legitimate reason why, I would be curious to explore the use of a wide range of grating materials as its effects on the diffraction pattern may be as shocking as the original experiment itself. I do not know how or why the interference would act differently, it is simply a matter of curiosity for me.

Again, I asked you, how would the nature of the material matter in the SQUID experiment. The superfluid does not care about the nature of the material due to quantum protectorate aspect. All it cares about is that there are two possible path. That's it.

Back to my original question; (regardless of why…) has the double slit experiment been made using many different slit materials?

Yes. All the double slit experiments conducted all over the world throughout history were not done using only one identical material. This would be utterly silly. Furthermore, x-ray diffraction experiments is one powerful diagnostic tool that is used to study materials. You can bet that a variety of different materials have been used in XRD studies.

This is a very puzzling query.

Zz.

 

[CaPhysics]

I suggest you re-read section 2.2 of “The Quantum Challenge” by Greenstein and Zajonc where they describe two experiments by Grainger, Roger and Aspect. The first experiment had a beam splitter and photon detectors on each of the two paths. When a single photon went through the beam splitter it was detected on one path or the other, never on both. This, however, does NOT prove that the other path was not taken; what it does prove is that the other path, if taken, was not selected for photon termination/detection. This can be seen by the second experiment of Grainger, Roger and Aspect. Here they sent the single photon through a Mach–Zehnder interferometer and found conclusive proof of a single photon interfering with itself (i.e., traveling both paths). So yes a wave (photon) can sub-divide (follow all paths) and still select only one path for termination/detection.
---------------------------------------------------------------------------
This is not true. Reread the common literature on the subject. Common antibunching experiments show that single photons are always detected at only one exit port after a beam splitter. If a photon had the ability to subdivide, you could place two detectors behind a beam splitter, fire single photons at it and would expect some coincidence detections due to the photon splitting in two. You never see those effects, so a photon cannot subdivide the way you suggest.

Please note that the signature of a single photon interfering with itself does not mean that it subdivides and travels along both paths. The concept of a well-defined photon path is already ill-defined in this context. It is the probability amplitudes for the several possible events which interfere, not the actual particles themselves. See for example Roy Glauber's Nobel lecture "one hundred years of light quanta" (frrely available in the net) for details.

Maybe I am missing something here but you state that [in paraphrase] "we have zero evidence that photons split in two (subdivide)." But the post you are refuting says "When a single photon went through the beam splitter it was detected on one path or the other, never on both." So what exactly are you refuting?
You also mention "probability amplitudes" interfering with each other to produce the photon wave effects. But "probability amplitudes" usually refers to the Schrodinger wave equation and you cannot write such an equation for a photon since it has zero rest mass http://en.wikipedia.org/wiki/Photon. If photon behavior was as easily solved as you indicate people would stop writing books about it (and they haven't).

Regards,

CaPhysics

 

[Unix60959]

 

[taylaron]

Is there any inaccurate information portrayed in this video?