Double slit experiment with detectors not recording

[tofu_]

In the double slit experiment, a beam of electrons forms an interference pattern on a screen after passing through two slits. If you place a detector on one or both of the slits, the electrons do not form an interference pattern.

In this video:


Physicist Thomas Campbell makes the following claim (at 2:45) if you leave the detectors turned on, but you throw away the data from the detectors without looking at it, you get a wave interference pattern on the screen behind the slits.

Is this true?

 

[blandrew]

In the double slit experiment, a beam of electrons forms an interference pattern on a screen after passing through two slits. If you place a detector on one or both of the slits, the electrons do not form an interference pattern.

In this video:


Physicist Thomas Campbell makes the following claim (at 2:45) if you leave the detectors turned on, but you throw away the data from the detectors without looking at it, you get a wave interference pattern on the screen behind the slits.

Is this true?

Here is a link to an experiment that was done regarding a "Delayed Choice Quantum Erasure"

http://arxiv.org/PS_cache/quant-ph/pdf/9903/9903047v1.pdf

 

[Jivesh]

How is that even possible? I though that the fundamental idea of quantum mechanics is that the act of measurement collapses the wavefunction of the electron and thus makes it behave as a particle, irrespective of whether we acknowledge the presence of the detector being there or not.
The detector here is essentially measuring the position observable X, whose only eigenfunctions are Dirac Delta functions, which gives the wavefunction a highly localised behavior and thus make it exhibit particle nature. How does it matter if we are not looking at the data of the detector? If the detector is there, then the electron should behave as a particle.

 

[Cthugha]

Physicist Thomas Campbell makes the following claim (at 2:45) if you leave the detectors turned on, but you throw away the data from the detectors without looking at it, you get a wave interference pattern on the screen behind the slits.

Is this true?

No. Once the data is detected by an irreversible process, deleting the data does not induce any changes. It does not matter whether you throw away the data or look at it.

The only thing you can erase are reversible markers. For example you can make the paths in the double slit experiment distinguishable by using polarizers at each slit. Afterwards it is possible to change this polarization without destroying the photon, so you can shift the polarization of the beams originating from both slits such, that they are the same again. As no irreversible process happened, the interference pattern will reappear. In this case you could get which-way information and destroy the interference pattern if you measured the photon at the right position and time, but as you never measure, it persists.

This is very different from actually measuring and throwing away the data, which will never give a persisting interference pattern.

 

[tofu_]

No. Once the data is detected by an irreversible process, deleting the data does not induce any changes. It does not matter whether you throw away the data or look at it.

So basically this guy, who is calling himself a physicist, is just a crackpot and doesn't know what he's talking about!

 

[Cthugha]

Well, it is very hard to find information about that guy. The only info I found is as follows:

"Tom holds a Bachelor of Science in Physics and Math from Bethany College and a Master of Science in physics from Purdue University, as well as having done doctoral-level work at the University of Virginia. He is the physicist described as “TC” in Bob Monroe’s Far Journeys. Tom began researching altered states of consciousness with Bob in the early 1970s. He and Dennis Mennerich helped to design experiments and develop the technology for creating specific altered states. They were also the main subjects of Bob’s investigations at that time. For the past thirty years, Campbell has been focused on scientifically exploring the properties, boundaries, and abilities of consciousness. During that same time period, he excelled as a working scientist—a professional physicist dedicated to pushing back the frontiers of cutting-edge technology.

Using his mastery of the out-of-body experience as a springboard, he dedicated his research to discovering the outer boundaries, inner workings, and causal dynamics of the larger reality system. In February of 2003, Tom published the My Big TOE trilogy. The acronym “TOE” is a standard term in the physics community that stands for “Theory Of Everything” and has been the Holy Grail of that community for fifty years. My Big TOE represents the results and conclusions of Tom’s personal and scientific exploration of the nature of existence. This overarching model of reality, mind, and consciousness merges physics with metaphysics, explains the paranormal as well as the normal, places spirituality within a scientific context, and provides direction for those wishing to personally experience an expanded awareness of All That Is."

This sounds very much like crackpottery. But besides: if he had evidence that throwing away data alters the interference pattern, why wouldn't he publish it in a scientific journal?

 

[GeorgCantor]

There is another crackpot?:


Professor David P. Jackson
Dickinson College
Dept. of Physics and Astronomy
Carlisle, PA 17013
(717) 245-1073
jacksond@dickinson.edu


This project is funded in part by the National Science Foundation (CCLI grant DUE-0737230) and Dickinson College.


Here is the experimental setup and their conclusions:

"The mystery of the Quantum Eraser is that when we tag the photon so that we can determine which path the photon travels it seems to know that we are watching and only travels one way. However when we erase path information and we can no longer determine which path the photon travels the interference returns!

To say it simply: The photon seems to know when we are watching and behaves differently when we can say where it has traveled."


http://singlephoton.wikidot.com/quantum-eraser



But if this were true, it would be possible to send signals ftl. Makes you wonder how fairly those grants are distributed.

 

[my_wan]

So basically this guy, who is calling himself a physicist, is just a crackpot and doesn't know what he's talking about!

Yes, this guy is trying to support his theory of quantum universal consciousness theory of everything. He has to embellish on the meaning in order to claim physical consistency such a crackpot theory.

The problem with the proliferation of such crackpots is that at the edge of theoretical physics it's hard to explain to many people what is wrong with it, and even involves some things none of us yet understand. It too often allows crackpots to have their cake and eat it to.

 

[unusualname]

I don't see anthing wrong with Jackson's experiment above, even though the phrasing is too cute, but the other youtube one is very confused and without specific experimental details you can't understand what he's talking about regarding the detectors.

However, he's almost certainly wrong if he's claiming conscious observation of the data caused the interference to disappear.

For a proper peer-reviewed double-slit with erasure experiment see http://grad.physics.sunysb.edu/~amarch/

This has interesting details, such as the fact that the which-way information can seemingly be obtained after the photon is detected and still destroy the interference! (Though, if you're happy with a non-local interpretation of quantum mechanics that's not too surprising, as a FTL mechanism can explain how the photon still "knows" about the which-way marker)

 

[Cthugha]

Here is the experimental setup and their conclusions:

"The mystery of the Quantum Eraser is that when we tag the photon so that we can determine which path the photon travels it seems to know that we are watching and only travels one way. However when we erase path information and we can no longer determine which path the photon travels the interference returns!
[...]
But if this were true, it would be possible to send signals ftl. Makes you wonder how fairly those grants are distributed.

This is not really crackpottery, but - well - very simplifying language. He uses a very specialized meaning of tagging and watching which can lead to very misleading interpretations, but he gets the experimental results straight.

This has interesting details, such as the fact that the which-way information can seemingly be obtained after the photon is detected and still destroy the interference! (Though, if you're happy with a non-local interpretation of quantum mechanics that's not too surprising, as a FTL mechanism can explain how the photon still "knows" about the which-way marker)

This is not really true. The information is not really obtained after detection of the first photon because all information is available only in coincidence counting, when both photons are already detected. What they do is more like a clever kind of filtering process than changing the past.
I gave a (atmittedly very simplifying) rough account of the basic physics behind DCQE (however, based on Kim's paper) in a different thread a rather long while ago, in case you are interested:
https://www.physicsforums.com/showpost.php?p=2241460&postcount=8".

The Walborn paper just introduces a different mechanism of introducing, removing and recovering indistinguishability of both possible photon paths.

 

[unusualname]

This is not really true. The information is not really obtained after detection of the first photon because all information is available only in coincidence counting, when both photons are already detected. What they do is more like a clever kind of filtering process than changing the past.
I gave a (atmittedly very simplifying) rough account of the basic physics behind DCQE (however, based on Kim's paper) in a different thread a rather long while ago, in case you are interested:
https://www.physicsforums.com/showpost.php?p=2241460&postcount=8".

The Walborn paper just introduces a different mechanism of introducing, removing and recovering indistinguishability of both possible photon paths.

Thanks for the clarification, I assumed a ftl mechanism was required, but using this coincidence counting method does seem to complicate the analysis.

 

[Cthugha]

Well, if you go into details, you will indeed find that one must assume some nonlocal model to explain some of the experiments performed with entangled particles (depending on your preferred interpretation this can be immediate collapse of the wave function or several different proposed mechanisms - there are enough threads in these forums discussing the different interpretations). You just do not get a possibility for ftl information transfer.

 

[Jivesh]

But I don't understand that why does he say that "when we leave the detector on, but don't make any measurements, that is, the magnetic head is moving, but there is no magnetic tape to record the readings of the moving head, then we see an interference/diffraction patter". Has this been proved in any experiment? The part where measurement affects the state of the system is understood, but why does he narrows down the measurement process only to consciousness, while the same effects can also be produced whenever electromagnetic waves interact with matter of any form, whether living or not.

 

[Cthugha]

Has this been proved in any experiment? The part where measurement affects the state of the system is understood, but why does he narrows down the measurement process only to consciousness, while the same effects can also be produced whenever electromagnetic waves interact with matter of any form, whether living or not.

No, it has not been proved in any experiment and it is just a plain wrong claim. Whether there is an consious observer at the end of the detection chain or not does not make any difference.

 

[unusualname]

Well, if you go into details, you will indeed find that one must assume some nonlocal model to explain some of the experiments performed with entangled particles (depending on your preferred interpretation this can be immediate collapse of the wave function or several different proposed mechanisms - there are enough threads in these forums discussing the different interpretations). You just do not get a possibility for ftl information transfer.

Yes, at least no classical information transfer, due to the non-deterministic selection of the quantum state.

I do wonder if a more ideal experiment with single (pairs) of photons (or other particles) could be realized, perhaps with the "which-way" particle detected over a much larger distance (so we're talking about several seconds or even minutes) than the "interference" particle so that coincidence counting wouldn't be required and perhaps a full detection screen for the interference pattern could be employed.

ie make the ftl mechanism explicit, since the "which-way" particle won't even be detected for several seconds/minutes after the "interference" particle hits the detection screen. That way you would be sure that some ftl "wave-function" mechanism was operating.

 

[Cthugha]

I do wonder if a more ideal experiment with single (pairs) of photons (or other particles) could be realized, perhaps with the "which-way" particle detected over a much larger distance (so we're talking about several seconds or even minutes) than the "interference" particle so that coincidence counting wouldn't be required and perhaps a full detection screen for the interference pattern could be employed.

The interference seen is a genuine two-photon interference effect comparable to Hong-Ou-Mandel interference. Therefore coincidence counting is always required to see the interference pattern even if the longer distance taken by the which-way particle corresponds to minutes.
It is only the two-photon state which is coherent. The which-way arm signal is as incoherent as it gets and will therefore on its own never show any meaningful interference pattern (unless you break entanglement).

 

[unusualname]

The interference seen is a genuine two-photon interference effect comparable to Hong-Ou-Mandel interference. Therefore coincidence counting is always required to see the interference pattern even if the longer distance taken by the which-way particle corresponds to minutes.
It is only the two-photon state which is coherent. The which-way arm signal is as incoherent as it gets and will therefore on its own never show any meaningful interference pattern (unless you break entanglement).

I don't understand, are you referring to the double-slit experiment I posted? The interference is created by single photons going through a double slit, the entangled partners are used to obtain which-way information only.

 

[Cthugha]

This is the most common misconception about DCQE experiments. It is not true that the interference pattern in DCQE experiments is caused by single photon interference and the other photon is just used for which-way information. If that was true, you would never need coincidence counting.

In fact, the light reaching the double slit is spatially incoherent. This means that its wavevector is rather undefined and by just looking at this arm you NEVER see any interference. Low coherence corresponds to a large spread in wavevectors. If you just shone light with well defined wavevector at the double slit, you would see a clear interference pattern. If you just shone light with a different well defined wavevector at the double slit, you would see a different clear interference pattern. However, as the spread in wavevectors is large, you will see a superposition of all possible interference patterns, which corresponds to no interference pattern at all. By using a small detector at one end and a large detector at the other end, you can choose some well defined wavevector at the small detector side. As you have entangled photons, the coincidence counting now acts like a filter. The entangled counterparts detected at the other side which are correlated to the ones detected at the small detector will now also be a subset with a well defined wavevector and therefore this small subset will have a well defined interference pattern. However, it is by no means possible to choose such a subset without doing coincidence counting.

 

[unusualname]

This is the most common misconception about DCQE experiments. It is not true that the interference pattern in DCQE experiments is caused by single photon interference and the other photon is just used for which-way information. If that was true, you would never need coincidence counting.

In fact, the light reaching the double slit is spatially incoherent. This means that its wavevector is rather undefined and by just looking at this arm you NEVER see any interference. Low coherence corresponds to a large spread in wavevectors. If you just shone light with well defined wavevector at the double slit, you would see a clear interference pattern. If you just shone light with a different well defined wavevector at the double slit, you would see a different clear interference pattern. However, as the spread in wavevectors is large, you will see a superposition of all possible interference patterns, which corresponds to no interference pattern at all. By using a small detector at one end and a large detector at the other end, you can choose some well defined wavevector at the small detector side. As you have entangled photons, the coincidence counting now acts like a filter. The entangled counterparts detected at the other side which are correlated to the ones detected at the small detector will now also be a subset with a well defined wavevector and therefore this small subset will have a well defined interference pattern. However, it is by no means possible to choose such a subset without doing coincidence counting.

In the experiment I'm referring to the coincidence counts are used just to ensure both entangled photons have passed through the apparatus.

http://grad.physics.sunysb.edu/~amarch/

This experiment uses single photon pairs via SPDC.

I think you're thinking of another experiment.

But correct me if I'm mistaken.

 

[Cthugha]

In the experiment I'm referring to the coincidence counts are used just to ensure both entangled photons have passed through the apparatus.

http://grad.physics.sunysb.edu/~amarch/

Sorry, but: no, they are not. As that page says: "It is not accurate to consider these photons as separate entities, but rather as one. They can travel very far away from each other, but they will not loose their correlation." Even if you completely removed all stray light and had no other background photon detections, you would not see an interference pattern without doing coincidence counting.

This experiment uses single photon pairs via SPDC.

This is right, but i would change the emphasis to:
This experiment uses single photon pairs via SPDC.

There are indistinguishable two-photon probability amplitudes leading to the same results. It is not necessary that these two photons ever meet for two-photon interference to occur.

I think you're thinking of another experiment.

But correct me if I'm mistaken.

No, I am considering the same experiment. Doing coincidence counting in such experiments is NEVER a means of just reducing the background noise due to stray light and other photons which are not part of an entangled pair.

In fact, single and two-photon interference are even complementary as has been shown in "Demonstration of the complementarity of one- and two-photon interference" by Abouraddi et al. (Phys. Rev. A 63, 063803 (2001)), also available on Arxiv.

 

[imiyakawa]

http://www.bottomlayer.com/bottom/reality/chap2.html

Is this guy also selling crackpottery? It's another claim of a double slit experiment being done wherein the data is thrown out prior to backwall observation. (interference)

Can someone link an experiment where the no-data-before-backwall-observation experiment is actually done? Why are all these guys lying to the QM-noob's faces?

 

[blandrew]

http://www.bottomlayer.com/bottom/reality/chap2.html

Is this guy also selling crackpottery? It's another claim of a double slit experiment being done wherein the data is thrown out prior to backwall observation. (interference)

Can someone link an experiment where the no-data-before-backwall-observation experiment is actually done? Why are all these guys lying to the QM-noob's faces?

Is this the experiment your looking for?

http://arxiv.org/PS_cache/quant-ph/pdf/9903/9903047v1.pdf

This is their summery, open for debate-

In conclusion, we have realized a quantum eraser experiment
of the type proposed in ref. [3]. The experimental
results demonstrate the possibility of observing both
particle-like and wave-like behavior of a light quantum
via quantum mechanical entanglement. The which-path
or both-path information of a quantum can be erased or
marked by its entangled twin even after the registration
of the quantum.

 

[peteratcam]

http://www.bottomlayer.com/bottom/reality/chap2.html

Is this guy also selling crackpottery? It's another claim of a double slit experiment being done wherein the data is thrown out prior to backwall observation. (interference)

Yes, a crack pot.

Why are all these guys lying to the QM-noob's faces?

Too much crack and pot.

 

[Cthugha]

http://www.bottomlayer.com/bottom/reality/chap2.html

Is this guy also selling crackpottery? It's another claim of a double slit experiment being done wherein the data is thrown out prior to backwall observation. (interference)

The peer-reviewed references he cites (and also the one blandrew just cited) are all valid, but this guy's interpretation of what these mean is a clear case of crackpottery.

 

[unusualname]

Sorry, but: no, they are not. As that page says: "It is not accurate to consider these photons as separate entities, but rather as one. They can travel very far away from each other, but they will not loose their correlation."

Yeah, so?

Even if you completely removed all stray light and had no other background photon detections, you would not see an interference pattern without doing coincidence counting.

If there was no stray light and 100% of the photons passed through the apparatus to the detectors you wouldn't need coincidence counting (duh!)

In practice some of the photons won't pass through the slits/polarisers no matter how perfect your apparatus, so you need coincidence counters in any case.

This double-slit eraser stuff can be confusing, but it's not as subtle as you're suggesting here.

It has a perfectly easy explanation in terms of non-local wave-functions without worrying about what happens at the which-way detector (which could be moved off towards infinity without changing what's happening at the interference detector)

No, I am considering the same experiment. Doing coincidence counting in such experiments is NEVER a means of just reducing the background noise due to stray light and other photons which are not part of an entangled pair.

In fact, single and two-photon interference are even complementary as has been shown in "Demonstration of the complementarity of one- and two-photon interference" by Abouraddi et al. (Phys. Rev. A 63, 063803 (2001)), also available on Arxiv.

Well you've not understood the experiment or its explanation, I don't even understand your point, coincidence counts means both photons were detected, that's all, why would you want to record a non-coincident blip?

 

[my_wan]

Why are all these guys lying to the QM-noob's faces?

It sells more books than real science does.

Many actually believe this. Some simply prefer to believe it. Others got it from from some sloppy wording somewhere, and never actually bothered to investigate their assumptions. Or what they did investigate was the lack of a classical explanation, rather than what observation physically entailed, and assumed that was equivalent. Evidence by the lack of evidence they weren't looking for.

At some levels it's no different from the psychic industry. People actually lose their homes as a result of being taken in by the Barnum effect. It's far bigger business than people realize.

 

[imiyakawa]

Amit Goswami subscribes to this kind of interpretation. How does he still get work? "Amit Goswami is a theoretical nuclear physicist and member of The University of Oregon Institute for Theoretical Physics since 1968, teaching physics for 32..."

 

[Cthugha]

If there was no stray light and 100% of the photons passed through the apparatus to the detectors you wouldn't need coincidence counting (duh!)

Sorry, but this is plain wrong. You need spatially coherent light to form an interference pattern. The light in one arm of the entangled beams is not coherent enough for that. Even without stray light and every photon detected at the double slit side, you would NOT see an interference pattern at the double slit side alone. If you could, you would be able to send ftl information using a quantum eraser.

If you do not believe me, read Zeilinger's famous review article:
http://www.hep.yorku.ca/menary/courses/phys2040/misc/foundations.pdf"

He also states that the pattern formed behind the double slit by one arm of the entangled beams alone will never show any interference at the beginning of section III.

If you have access to a BBO crystal, just try it out. If you increase the distance between the BBO and the double slit, you will see an interference pattern without the need to perform coincidence counting (well, if you do not have too much stray light of course) and you will not see an interference pattern in the coincidence counting data. If you decrease the distance, you will notice that the direct interference pattern vanishes and the coincidence counting pattern appears. This has also been calculated and experimentally tested in the PHD thesis of Birgit Dopfer, a former PHD student of Zeilinger. Unfortunately the thesis is in German only and has vanished from the web. If anybody still has a copy, please tell me.

In practice some of the photons won't pass through the slits/polarisers no matter how perfect your apparatus, so you need coincidence counters in any case.

Yes, but this is not the reason why you need coincidence counting.

Well you've not understood the experiment or its explanation, I don't even understand your point, coincidence counts means both photons were detected, that's all, why would you want to record a non-coincident blip?

Now you confuse me. It was your point to record a non-coincident blip in an earlier post:

The interference is created by single photons going through a double slit, the entangled partners are used to obtain which-way information only.

Single photon interference patterns mean that you do not need coincidence counting. Two-photon interferences like in the Hong-Ou-Mandel effect always need coincidence counting. Saying the interference is created by single photons means that coincidence counting is unnecessary or used only to reduce noise. This is NOT the case here.

edit: it took me a while to find some other experimental data, but you can also have a look at post 20 in this thread:
https://www.physicsforums.com/showthread.php?t=186341&page=2"

and the paper attached there for clear experimental data that there is no interference pattern without coincidence counting.

 

[unusualname]

(I lost a longer reply on my netbook, here's most of it again)

Sorry, but this is plain wrong. You need spatially coherent light to form an interference pattern. The light in one arm of the entangled beams is not coherent enough for that. Even without stray light and every photon detected at the double slit side, you would NOT see an interference pattern at the double slit side alone. If you could, you would be able to send ftl information using a quantum eraser.

If you do not believe me, read Zeilinger's famous review article:
http://www.hep.yorku.ca/menary/courses/phys2040/misc/foundations.pdf"

He also states that the pattern formed behind the double slit by one arm of the entangled beams alone will never show any interference at the beginning of section III.

No it's not wrong. Zeilinger says (at the beginning of section III) that there is no interference if the entangled partners have path information, this is not in dispute. If the path information is erased then the interference appears. The only way to be sure the path information is erased is to have a coincidence counter tell you what happened to the entangled partner. If there is no coincindence count then we do not know what happened to the entangled partner (it might have got zapped by a cosmic omg particle for all we know) so we may or may not get coherent superposition states for the other particle through the double-slit, and so the non-coincidence blips will mess the interference pattern up (cause noise if you like)
And I don't know why you keep going on about spatially coherent light for single photons, are you suggesting a similar experiment can't be done using neutrons or buckyballs?

Here's a question for you:

1. Suppose I move the detector for the which-way particles to alpha-centuri, then run the experiment with everything else set up exacly as in a previous run which gave us an interference pattern (when matched with coincindence blips).

2. Now, if I stick my data for the interference detector blips on a floppy disk does it contain an interference distribution amongst its data or do I have to wait several years for the photons to reach alpha-centuri before the data on the floppy disk is finalised? (With the data from alpha centuri I assume I can restrospectively do a coincidence match assuming the photons traveled undisturbed at uniform speed)

(I didn't go through your mathematical analysis you linked to earlier in your first post, you may well have correctly analysed the phases or similar at the detectors but I don't see how it's relevant, the explanation for this experiment is that the wave-function is non-local)

 

[Cthugha]

No it's not wrong. Zeilinger says (at the beginning of section III) that there is no interference if the entangled partners have path information, this is not in dispute.

Right.

If the path information is erased then the interference appears. The only way to be sure the path information is erased is to have a coincidence counter tell you what happened to the entangled partner. If there is no coincindence count then we do not know what happened to the entangled partner (it might have got zapped by a cosmic omg particle for all we know) so we may or may not get coherent superposition states for the other particle through the double-slit, and so the non-coincidence blips will mess the interference pattern up (cause noise if you like)

This is the point I (and Zeilinger also) dispute. You assume the only reason for doing coincidence counting is assuring that which-way info gets dumped. I disagree. Let me introduce a simple experiment which points out the difference:

Imagine an ideal experiment. No influence of stray light, ideal photodetectors and such stuff. You again have a double-slit side and the entangled partner on the other which-way side. Now you directly dump the entangled which-way side (in a manner that does not give any which-way information) and are not interested in the exact detections. The assurance that which-way information is lost is enough. If the only sense of coincidence counting was assuring that stray light and noise is reduced and to assure that which-way information is indeed destroyed, this would mean that you should now see an interference pattern in all detections at the detections at the double slit side alone. I say this is not the case.

Now modify this experiment to the delayed choice variant again. The which-way side travels to Alpha-Centauri and the double-slit side has already been recorded. While the which-way particles are moving on, you can choose whether to take which-way information or not and therefore whether you get an interference pattern in coincidence counting or not. Again you detect all which-way particles without any losses. Assuming that ONLY nonlocality of the wavefunction causes the interference pattern to disappear and reappear would mean that in this case your choice at the which-way side alters the detections already made on the double slit side, allowing to send ftl-info. I say this is not the case and coincidence counting is needed to get additional information from the which-way side (k-space information) to pick a suitable subset of detected photons at the double slit side which will show interference.

And I don't know why you keep going on about spatially coherent light for single photons, are you suggesting a similar experiment can't be done using neutrons or buckyballs?

No, that is not my point. Entangling neutrons or buckyballs should be pretty complicated. The main point of my argument is that the spatial coherence of the subset chosen by coincidence counting is significantly higher than the spatial coherence of the whole set of sdetected photons in one arm without doing coincidence counting and that it is this increase which causes the interference pattern to appear. This is, however, much easier to see in the Kim et al. quantum eraser paper.

Here's a question for you:

1. Suppose I move the detector for the which-way particles to alpha-centuri, then run the experiment with everything else set up exacly as in a previous run which gave us an interference pattern (when matched with coincindence blips).

Ok.

2. Now, if I stick my data for the interference detector blips on a floppy disk does it contain an interference distribution amongst its data or do I have to wait several years for the photons to reach alpha-centuri before the data on the floppy disk is finalised? (With the data from alpha centuri I assume I can restrospectively do a coincidence match assuming the photons traveled undisturbed at uniform speed)

You will of course have to wait until you also have the data from the which-way side.

(I didn't go through your mathematical analysis you linked to earlier in your first post, you may well have correctly analysed the phases or similar at the detectors but I don't see how it's relevant, the explanation for this experiment is that the wave-function is non-local)

See my above example for an experiment where it makes a difference. Nonlocality ALONE is not enough.

 

[unusualname]

This is the point I (and Zeilinger also) dispute. You assume the only reason for doing coincidence counting is assuring that which-way info gets dumped. I disagree. Let me introduce a simple experiment which points out the difference:

Imagine an ideal experiment. No influence of stray light, ideal photodetectors and such stuff. You again have a double-slit side and the entangled partner on the other which-way side. Now you directly dump the entangled which-way side (in a manner that does not give any which-way information) and are not interested in the exact detections. The assurance that which-way information is lost is enough. If the only sense of coincidence counting was assuring that stray light and noise is reduced and to assure that which-way information is indeed destroyed, this would mean that you should now see an interference pattern in all detections at the detections at the double slit side alone. I say this is not the case.

How do you "directly dump" the entangled which-way side? When the entangled pair leaves the source they have path information (which is equivalent to which-way information) due to momentum conservation, you must erase this information before either photon can show interference through a double-slit. If you "just dump" the which-way side by letting it hit a screen or ignoring it and letting it shoot out into empty space you will not get an interference pattern with the other photon.

You must carefully ensure path/which-way information is erased by passing the photon through some erasure mechanism, and you must check it passed through this mechanism by using a coincidence counter to match it to a partner.

You can't just talk about "dumping" the which-way side. What you really mean is if we can be sure 100% of the which way side is going through the erasure mechanism imagine we dump them afterwards (ie don't bother detecting them afterwards).

But since this 100% ideal situation can't be realized we have to employ a coincidence counter.

I'll skip the rest of the suggested experimental details since this makes it invalid.

No, that is not my point. Entangling neutrons or buckyballs should be pretty complicated. The main point of my argument is that the spatial coherence of the subset chosen by coincidence counting is significantly higher than the spatial coherence of the whole set of sdetected photons in one arm without doing coincidence counting and that it is this increase which causes the interference pattern to appear. This is, however, much easier to see in the Kim et al. quantum eraser paper.

I don't understand this, it sounds like you're suggesting that if there is no coincidence count some weird type of photon pair was emitted from the source with properties not consistent with the photons creating the interference pattern (different wavelength/frequencies, wrong directions??)

You will of course have to wait until you also have the data from the which-way side.

Of course you wouldn't, are you crazy! Several years later when you get the data back from alpha centuri you can do a retrospective coincidence match based on timings of detection at alpha-centuri and adding the appropriately large offset (assuming the photons traveled undisturbed and at uniform speed to alpha-centuri)

The floppy disk data isn't going to magically change when the photons hit the detectors at alpha centuri.!

The point is, that the photons traveling to alpha centuri have had their which-way information erased, so it doesn't really matter what happens to them after this, but (to repeat myself for the umpteenth time) we'd like to know which photons these had as entangled partners to extract our interference pattern from the floppy disk data (which will contain a lot of detections for photons whose entangled partner didn't make it through the erasure mechanism)Once the which-way information is erased, the photon hitting the double-slit can split into two wavelike coherent paths and interfere with itself.

The point of suggesting moving the which-way detector to alpha-centuri was to lead on to something much more non-trivial in these experiments which hasn't so far been addressed (afaik)

What if you move the erasure mechanism on the which-way path to alpha centuri?

Copenhaginists will say you still get an interference pattern (yes, with coincidence matching, sigh), but I wonder how big the delay in these delayed choice experiments can b?. Increasing the erasure delay even beyond a few seconds is not acheivable with current technology.

But this would get to the crux of just how non-local is the wave-function? (it breaks einstein locality for sure), and would (in theory) allow a refutaion of the copenhagen interpretation if favor of some non-local pilot-wave type mechanism (where the pilot wave has some bounded type of locality based on communication in higher dimensions or via some other exotic topology or via undetected tachyons etc etc)

This last part sounds speculative, of course, but the reason why "no one understands quantum mechanics" is because no one knows how the wave function enables non-local physics, and we can only speculate.The moral of this debate is that they really should teach pilot-wave interpretations in mainstream physics courses, even if the mechanism turns out to be nonsense in reality, it's hardly worse than the CI and it does enable easy analysis of delayed choice quantum experiments.

 

[Cthugha]

How do you "directly dump" the entangled which-way side?

You just have to project it into a momentum eigenstate which cannot carry any position information. See figure 3 in the Zeilinger paper which I linked earlier.

But since this 100% ideal situation can't be realized we have to employ a coincidence counter.

I'll skip the rest of the suggested experimental details since this makes it invalid.

Sorry, but this is wrong. You can in principle realize a setup where all of the which-way photons are perfectly dumped at the beginning with which-way information destroyed. This does not invalidate anything.

I don't understand this, it sounds like you're suggesting that if there is no coincidence count some weird type of photon pair was emitted from the source with properties not consistent with the photons creating the interference pattern (different wavelength/frequencies, wrong directions??)

Ok, just one question. Take incoherent light, maybe some white light source, put it directly in front of a double slit. Will you see an interference pattern? No you won't. The light is too incoherent. This is why Young had to use a pinhole in front of his famous first double-slit experiment. Do you accept this basic physics?

Of course you wouldn't, are you crazy! Several years later when you get the data back from alpha centuri you can do a retrospective coincidence match based on timings of detection at alpha-centuri and adding the appropriately large offset (assuming the photons traveled undisturbed and at uniform speed to alpha-centuri)

And for this you do not have to wait until you have the data from Alpha-Centauri? I am puzzled what you mean by this.

The floppy disk data isn't going to magically change when the photons hit the detectors at alpha centuri.!

Exactly! But if it was only nonlocality defining whether you see an interference pattern or not, the recorded data at the floppy would have to change. It obviously does not. THIS is my argument.

The point is, that the photons traveling to alpha centuri have had their which-way information erased, so it doesn't really matter what happens to them after this, but (to repeat myself for the umpteenth time) we'd like to know which photons these had as entangled partners to extract our interference pattern from the floppy disk data (which will contain a lot of detections for photons whose entangled partner didn't make it through the erasure mechanism)

You do not seem to get the point. What happens if the which-way data is not erased? You can still erase it 5 seconds before the photons reach Alpha Centauri. If you do this erasure 5 seconds before, you will be able to get the interference pattern. If you do not, you will not. In your picture the detections would have to change. In my description this is not necessary.

Once the which-way information is erased, the photon hitting the double-slit can split into two wavelike coherent paths and interfere with itself.

They can do so anyway. However without coincidence counting you get a lot of different superposed interference patterns which add up to no interference pattern at all. Only by "filtering" one of these subsets by picking coincidence counts on the other side with some well defined property (usually wave vector), you can pick one of the superposed interference patterns. Have a look at the Kim and Scully quantum eraser paper. There you have two possible paths in which position information is erased. The interference patterns you get in coincidence counting from these two possible paths are exactly out of phase and sum up to no pattern at all. How do you explain this experimental fact in your picture?

The point of suggesting moving the which-way detector to alpha-centuri was to lead on to something much more non-trivial in these experiments which hasn't so far been addressed (afaik)

What if you move the erasure mechanism on the which-way path to alpha centuri?

Copenhaginists will say you still get an interference pattern (yes, with coincidence matching, sigh), but I wonder how big the delay in these delayed choice experiments can b?. Increasing the erasure delay even beyond a few seconds is not acheivable with current technology.

But this would get to the crux of just how non-local is the wave-function? (it breaks einstein locality for sure), and would (in theory) allow a refutaion of the copenhagen interpretation if favor of some non-local pilot-wave type mechanism (where the pilot wave has some bounded type of locality based on communication in higher dimensions or via some other exotic topology or via undetected tachyons etc etc)

This last part sounds speculative, of course, but the reason why "no one understands quantum mechanics" is because no one knows how the wave function enables non-local physics, and we can only speculate.

Now you are getting into crackpottery. These points have been addressed time and again. My comments and the references I quoted are exactly about this point and the references clearly show that Einstein locality/causality is NOT violated (it is nonlocal, but does not allow information transfer). See also Brian Greene's Fabric of the Cosmos, any paper by Zeilinger, Weihs or Scully for clear accounts why Einstein locality is not violated.

 

[unusualname]

For the record let's just say that we have contrary explanations of the coincidence counter's role.

 

[Cthugha]

I agree.

I will think about an experiment that is capable of directly testing the different models.

 

[lcdisplay]

I have read this entire thread and I'm leaning towards unusualname. The DCQE which is a variant of the QE shows a much deeper interpretation of time that I won't get into.

Just sticking with the standard QE though it is not evident from the sources I have read that the coicidence counter is required in the experiments except for filtering and backup. These entangled photons come out like once every 100 pulses so they are rare enough that without the coincidence counter noise would dominate your results.

In the equations of quantum mechanics I don't know anybody (yet) who has shown that you must employ a coincidence counter.

It is true that you need one for rare single photon experiments in less than ideal light isolation conditions but if you had an ideal lab the need is not obvious.

Indeed it seems like in 1998 Weihs et al showed this. "G. Weihs, T. Jennewein, C. Simon, H. Weinfurter, A. Zeilinger, Phys. Rev. Lett. 81 5039 (1998).

I'm intrigued why though, Cthuga, you are so adamant that it is necessary. I'm not ridiculing you mind you as you sound quite well versed in QM. I am skeptical if anything like you as I have not performed the experiment myself but can only have faith that the data is correct. This is relatively cutting edge and if you have a different set of sources that point to a coincidence counter being needed I would love to read them.

 

[lcdisplay]

As far a FTL communication is concerned, this is an area of intrigue for anybody with a pulse and as a Quantum Physics student with a pulse I am very intrigued.

I challenge the notion that a photon pair in a singlet state for example cannot transfer information FTL. Indeed it does transfer non usable information faster than light. That is not in question at this time (or is it).
So we develop the no signaling, no communication theorems which say no matter what you do you cannot use the wavefunction collapse to transmit information. And this today seems is correct but whether or not you can use the existence of an interference pattern or lack thereof as a way to transmit a 1 or 0 via a QE is a completely different phenomenon.
In this concept the no communication/no signaling theorem does not apply, at least not to me. And so I challenge the notion that FTL comm is not possible. As iron sharpens iron I would love to be challenged by you gentelemen into a discussion on this subject to see if sense can be made. I am fully open to the possibility that I am wrong as I make no special claim that my understanding is fool proof.

 

[lcdisplay]

If we take a Zeigler type of QE setup and somehow produce 100 nanoseconds worth of back to back to back photons so that there is thousands of them back to back in that 100 ns and send them off to NewYork and the entangled "pulse" to San Francisco (we produce them somewhere in the middle) we should be able to dicipher the inteference patterns (or lack of) quickly without the use of movable single detectors. Assuming little noise of course. This should be able to do the job but I could be wrong. What do you guys think?

 

[unusualname]

You can't know which photons contributed to the interference without measuring both the entangled photons.

What the photons do at the qm level is random as far as classical physics is concerned so there is no way of transmitting classical information.

An interesting question left unanswered by these experiments is how non-local the qm wave-function is, since the delays involved are rather small, and we can't do it via New York and San Francisco yet (or via alpha-centuri :) ) , Bell type results have been confirmed over a few kilometers in Switzerland, but that's about it.

If something like the Holographic Principle exists, then we may just be viewing a projection from another dimensional surface, which would help explain the non-locality. We simply don't know yet.

 

[lcdisplay]

You can't know which photons contributed to the interference without measuring both the entangled photons.

Please explain

 

[JesseM]

Please explain

There have been a few threads on the delayed choice quantum eraser (DCQE), see here for example. Here's what I said before about why you can't see an interference pattern without doing a coincidence count:

Even in the case of the normal delayed choice quantum eraser setup where the which-path information is erased, the total pattern of photons on the screen does not show any interference, it's only when you look at the subset of signal photons matched with idler photons that ended up in a particular detector that you see an interference pattern. For reference, look at the diagram of the setup in fig. 1 of this paper:

http://arxiv.org/abs/quant-ph/9903047

In this figure, pairs of entangled photons are emitted by one of two atoms at different positions, A and B. The signal photons move to the right on the diagram, and are detected at D0--you can think of the two atoms as corresponding to the two slits in the double-slit experiment, while D0 corresponds to the screen. Meanwhile, the idler photons move to the left on the diagram. If the idler is detected at D3, then you know that it came from atom A, and thus that the signal photon came from there also; so when you look at the subset of trials where the idler was detected at D3, you will not see any interference in the distribution of positions where the signal photon was detected at D0, just as you see no interference on the screen in the double-slit experiment when you measure which slit the particle went through. Likewise, if the idler is detected at D4, then you know both it and the signal photon came from atom B, and you won't see any interference in the signal photon's distribution. But if the idler is detected at either D1 or D2, then this is equally consistent with a path where it came from atom A and was reflected by the beam-splitter BSA or a path where it came from atom B and was reflected from beam-splitter BSB, thus you have no information about which atom the signal photon came from and will get interference in the signal photon's distribution, just like in the double-slit experiment when you don't measure which slit the particle came through. Note that if you removed the beam-splitters BSA and BSB you could guarantee that the idler would be detected at D3 or D4 and thus that the path of the signal photon would be known; likewise, if you replaced the beam-splitters BSA and BSB with mirrors, then you could guarantee that the idler would be detected at D1 or D2 and thus that the path of the signal photon would be unknown. By making the distances large enough you could even choose whether to make sure the idlers go to D3&D4 or to go to D1&D2 after you have already observed the position that the signal photon was detected, so in this sense you have the choice whether or not to retroactively "erase" your opportunity to know which atom the signal photon came from, after the signal photon's position has already been detected.

This confused me for a while since it seemed like this would imply your later choice determines whether or not you observe interference in the signal photons earlier, until I got into a discussion about it online and someone showed me the "trick". In the same paper, look at the graphs in Fig. 3 and Fig. 4, Fig. 3 showing the interference pattern in the signal photons in the subset of cases where the idler was detected at D1, and Fig. 4 showing the interference pattern in the signal photons in the subset of cases where the idler was detected at D2 (the two cases where the idler's 'which-path' information is lost). They do both show interference, but if you line the graphs up you see that the peaks of one interference pattern line up with the troughs of the other--so the "trick" here is that if you add the two patterns together, you get a non-interference pattern just like if the idlers had ended up at D3 or D4. This means that even if you did replace the beam-splitters BSA and BSB with mirrors, guaranteeing that the idlers would always be detected at D1 or D2 and that their which-path information would always be erased, you still wouldn't see any interference in the total pattern of the signal photons; only after the idlers have been detected at D1 or D2, and you look at the subset of signal photons whose corresponding idlers were detected at one or the other, do you see any kind of interference.

 

[unusualname]

^^JesseM is referrring to a more complicated experimental setup than the one in the experiment I have been referring to (Walborn et al):

http://grad.physics.sunysb.edu/~amarch/

In this experiment you need to measure both entangled photons to ensure the which-path interference was erased, otherwise you are dealing with a set of photons with unpredictable properties, either because the entangled partner never made it through the eraser mechanism or you're detecting photons from laboratory noise.

The Walborn experiment makes this much more clear than other quantum eraser experiments, where phase matching etc might come into play.

 

[lcdisplay]

Jesse the quantum eraser experiment you have referenced is one of many possible ways to erase the which way information. In that experiment a bad design for FTL communication was chosen (good for simple path erasing). Look at Dopfer et al. The Heisenberg detector does not have this weakness.

Unusual, about the coincidence detector. Even if coincidence detection was necessary that alone would not prohibit FTL communication and that was proven by Weihs et al in the aforementioned refrence of my prev post. They just synchronized quantum clocks.

What about the "subset" of detected pairs. Again Wineland/NIST addressed that loophole and at most it would correspond to an SNR question. Additionally, spontaneous Parametric Down Converted photon pairs, among other entangled photon creation schemes, can be made to filter the entangled pairs from the non entangled pairs at the point of creation so this should not be a problem.

We know stray light is just an engineering challenge and finally the efficiency of pair creation is also just another engineering challenge which is starting to get solved.

So far I don't see a problem. What am I not seeing?

In my previous post I mentioned using a Zeigler type of QE but I meant Walborn

 

[unusualname]

I don't understand what your question is. Just becasue the interference seems to require ftl or non-local physics why do you think that implies ftl communication?

QM is random as far as classical physics is concerned, at least we have no known way of deterministically selecting a quantum state, so you can't transfer classical information ftl. There may be quantum processes operting ftl but that doesn't break einstein causality, in fact for all we know, the quantum processes themselves may be fundamentally random or acausal.

 

[JesseM]

Jesse the quantum eraser experiment you have referenced is one of many possible ways to erase the which way information. In that experiment a bad design for FTL communication was chosen (good for simple path erasing). Look at Dopfer et al. The Heisenberg detector does not have this weakness.

I suspect it probably does have the same weakness, i.e. there are multiple places the "idler" photon might be detected, and if you look at the subset of "signal" photons whose idler was detected in one specific position you could see an interference pattern, but if you add together all the subsets corresponding to different possible positions for the idler (weighted by the probability of the idler landing at each position), the sum is a non-interference pattern, so the total pattern of signal photons doesn't show interference. I don't know how to do the actual quantum optics calculations to show that this is true for the Dopfer experiment, so this is just a speculation, but I think it's a plausible one. See this thread and this one for more detailed speculations on just how things might work out in a few variations on the Dopfer experiment.

 

[lcdisplay]

Unusual, it's true the walborn experiment does seem to be similar to the zeilinger one looking at it more closely. I think I understand what you and Jesse are referring to as the subset of total photons now. The subset of entangled photons "selected" or measured by the polarizer means their entangled partners were polarized parallel to a primary axis of the QWP, so the QWPs act as retarders not "rotators" if you will, and thus we see the interference pattern. BUT only using a coincidence counter. Without the coincidence counter you will record a superposition of patterns which is the blob.

I will have to analyze the Dopfer experiment Jesse to see how the subset selection is done.

 

[Cthugha]

Sorry for the very late answer. I was attending a conference and did not have the time to post here.

Just sticking with the standard QE though it is not evident from the sources I have read that the coicidence counter is required in the experiments except for filtering and backup. These entangled photons come out like once every 100 pulses so they are rare enough that without the coincidence counter noise would dominate your results.

In the equations of quantum mechanics I don't know anybody (yet) who has shown that you must employ a coincidence counter.

It is true that you need one for rare single photon experiments in less than ideal light isolation conditions but if you had an ideal lab the need is not obvious.

In fact this is not true. Single photon interference patterns for downconverted light have been demonstrated without coincidence counting. The background noise is not necessarily that intense when using the right equipment. See for example
"Demonstration of the complementarity of one- and two-photon interference" by Abouraddy et al. (Phys. Rev. A 63, 063803 (2001) , also available on arxiv: http://arxiv.org/abs/quant-ph/0112065" )

Indeed it seems like in 1998 Weihs et al showed this. "G. Weihs, T. Jennewein, C. Simon, H. Weinfurter, A. Zeilinger, Phys. Rev. Lett. 81 5039 (1998).

That paper aims at the nonlocality issues. I am not sure what you think it means for this discussion.

I'm intrigued why though, Cthuga, you are so adamant that it is necessary. I'm not ridiculing you mind you as you sound quite well versed in QM. I am skeptical if anything like you as I have not performed the experiment myself but can only have faith that the data is correct. This is relatively cutting edge and if you have a different set of sources that point to a coincidence counter being needed I would love to read them.

My point is also contained in the paper I cited above using an even simpler experimental geometry. In fact it is the main point of that paper. It clearly shows that single photon interference patterns (without coincidence counting) and two-photon interference patterns (with coincidence counting) are complementary like position and momentum. If one thinks about that for a moment that issue is trivial. Single photon interference relies on spatial coherence of the single photon state. Therefore you can only see it under far-field conditions. That is, you have a narrow distribution of the wavevectors contributing to the interference pattern.
Two-photon interference relies on the coherence of the two-photon state. However, for entanglement to be meaningful you need near-field conditions resulting in a broad distribution of the wavevectors in your beam. As you see, the necessary conditions for single- and two-photon interference exclude each other and it is not possible to see a two-photon interference pattern without coincidence counting by any means.

That point has also been shown in Birgit Dopfers PhD thesis, but unfortunately it is written in German and seems to have vanished from the web. However, there are also several other papers on the complementarity between single- and two-photon interference which can be found quite easily.

By the way: testing the different explanations of DCQE presented in this thread is quite easy. Usually position information is easily dumped by imaging this part of the entangled photons into the Fourier plane and placing a detector there. This is enough to erase position information. In my explanation you now need a small detector on this part to filter out a narrow range of wavevectors in the Fourier plane. If destruction of position information alone was enough, you could also use a large bucket counter spanning the whole Fourier plane. Walborn for example uses a small detector at this position. I wonder whether someone actually compared these situations.

 

[lcdisplay]

Two-photon interference relies on the coherence of the two-photon state. However, for entanglement to be meaningful you need near-field conditions resulting in a broad distribution of the wavevectors in your beam. As you see, the necessary conditions for single- and two-photon interference exclude each other and it is not possible to see a two-photon interference pattern without coincidence counting by any means.

Cthugha thanks for your response, it's never late. I will need to read your reference. Also I'm not quite sure what you mean by single and two photon interference?

I'm pretty sure I understand now why the Walborn experiment must have coincidence counting and I'm relatively sure that same reasoning extends to the Dopfer Heisenberg detector setup where he spatially filters a subset of photons. I just haven't mathematically proven it to myself yet.

I'm actually writing a paper right now on why the Quantum Eraser experiments as they are today will not allow for FTL communication. So I'm looking at all the references.
Thanks

 

[Cthugha]

Also I'm not quite sure what you mean by single and two photon interference?

Single photon interference is what you see in a common double slit experiment. In this situation - as Dirac calls it - "every photon interferes only with itself". This is true even at low photon count rates when photons are arriving one at a time.

Two-photon interference results from two indistinguishable photons. The best known example is the Hong-Ou-Mandel effect (http://en.wikipedia.org/wiki/Hong–Ou–Mandel_effect" ), where two indistinguishable photons arriving at a beam splitter always take the same exit port. Here you superpose probability amplitudes of these two-photon processes which lead to the same result and can get constructive or destructive interference. Therefore, some of the processes possible for two distinguishable photons do not occur for two indistinguishable photons.

A more detailed discussion of what two-photon interference is can be found in the paper linked in the wikipedia article: http://physics.nist.gov/Divisions/Div844/publications/migdall/psm96_twophoton_interference.pdf" .

 

[plexel]

That point has also been shown in Birgit Dopfers PhD thesis, but unfortunately it is written in German and seems to have vanished from the web.

B. Dopfer, Ph.D. Thesis, U. Innsbrück 1998 http://web.archive.org/web/20070714025355/www.quantum.univie.ac.at/publications/thesis/bddiss.pdf