Delayed choice quantum eraser

[Spacezilla]

http://en.wikipedia.org/wiki/Delayed_choice_quantum_eraser" : The results from Kim, et al. have shown that, in fact, observing the second photon's path will determine the particle or wavelike behavior of the first photon at the detector, even if the second photon is not observed until after the first photon arrives at the detector. In other words, the delayed choice to observe or not observe the second photon will change the outcome of an event in the past.

Question: What would happen if this was delayed even further? We are talking nanoseconds here, but what if the particle was slowed or sent far away, before being reflected back? What I'm getting at is... What if the delay wasn't nanoseconds, but seconds? What if it was minutes? What if it was enough time to manually change whether the second photon was observed or not?

The results from Kim, et al. have shown that, in fact, observing the second photon's path will determine the particle or wavelike behavior of the first photon at the detector, even if the second photon is not observed until after the first photon arrives at the detector.

In other words, the photon behaves one way if we observe the second photon nanoseconds later and another way if we don't. So we start the experiment and we observe our detector. Depending on whether we see an interference pattern or not, we observe each secondary photon or we don't and that's the point.

I'm been thinking about this for a while and I can't figure out what would happen. I searched Google and found http://uplink.space.com/showflat.php?Board=sciastro&Number=141733 where someone else asks the same question, but unfortunately no one really seems to understand the experiment, so he gets no real answer.

Now, I turn on my photon generator.. and what do I see? Either a "particle" pattern or an interference pattern, one would presume.

Well, according to the results of the already-performed original "Delayed Choice Quantum Eraser" experiment, what I see depends on what the state of the 'which-path/no which-path' switch will be 10 minutes from now when the idler photon hits it?

What if I see an interference pattern, so then when I see that, I make sure the switch is on "which-path" no matter what... which will clearly violate the results of the previous expirement. I can deliberately violate the rules of what I see on the detector, by switching the switch to what WOULD give me the opposite result, using my own free will.

What is the deal here, and what would happen in this case?

This is exactly what I'm asking too. Now, depending on how you choose to interpret the delayed choice quantum eraser experiment, there's no getting around that we're either changing the past or predicting the future. So what happens if we delay it long enough to change whether we'll detect or erase the information about the second photon's after observing the results?

This is not just a theoretical question, this is something that should be possible to test for someone who has access to this kind of equipment, but unfortunately I don't. However, many people have in the past predicted the outcomes of quantum experiments correctly, so I hope someone here can predict the outcome of this experiment as well.

Thank you for your time. :)

 

[JesseM]

Referring to the first set of photons as "signal photons" and the second set as "idler photons", the way it works is that the total pattern of signal photons actually never shows interference--even if you measure all the idlers in such a way that the which-path information is erased, you will see interference if you do a "coincidence count" between signal photons and idlers which went to a certain detector, but if you add all the subsets, they add up to a non-interference pattern. So, it's impossible to find an interference pattern until after the idlers have already been measured, ruling out the possibility of any backwards-in-time hijinx. This was discussed on this thread a while ago. And here's my summary from this thread:

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://xxx.lanl.gov/PS_cache/quant-ph/pdf/9903/9903047.pdf

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.

 

[Demystifier]

See
http://arxiv.org/abs/quant-ph/0611034

 

[Demystifier]

For those who have not read the paper above, here is a plain-English summary:
The delayed-choice experiment looks paradoxical if you think that the photon passes either through both slits or through one slit only. The paradox removes if you assume that something (the wave) passes through both slits, while something different (the particle) passes through one slit only. But then you need to assume that the photon consists of two separate things, which contradicts the standard interpretation of QM. Fortunately, there is an interpretation - the Bohmian interpretation - that provides such a wave-and-particle picture consistently.

 

[Spacezilla]

Thank you for your comments, I greatly appreciate it.

 

[vanesch]

It might be interesting to do a search in this forum on delayed choice quantum eraser. Many threads have discussed this in the past here.

BTW, although Bohmian mechanics can give an explanation for the phenomenon, every self-consistent interpretation of quantum theory can do so.

 

[Demystifier]

BTW, although Bohmian mechanics can give an explanation for the phenomenon, every self-consistent interpretation of quantum theory can do so.

I agree. But why then this phenomenon is regarded as a big deal, even by experts?

 

[vanesch]

I agree. But why then this phenomenon is regarded as a big deal, even by experts?

I already pointed out a few times that many experimental papers on this issue, which are *experimentally* very rich, are very misleading/poorly written on the interpretation side.

I guess that both Bohmians and MWI-ers don't find it a big deal (although nevertheless quite fun). The "projectionists" are of course a bit puzzled because they don't know anymore when to project
(although, if you really understand von Neumann, you wouldn't be puzzled either).

 

[RProgrammer]

Simplification

I too have seen this a lot and have found a one-line version of the question:

What would you see if you looked at the detector in a DCQE before the choice was made?



Now as for the implementation, I see Spacezilla's point also that the implementation would seem at least as straightforward as the original experiment.
I would imagine you would use a Bose-Einstein condensate to considerably slow the signal photons. While you replaced the mirrors with a manual (manually controlled electronic) switch which merged the signal photon path from slit B into the path from slit A.

 

[RProgrammer]

Understanding Bohmian

Demystifier

So, under the bohmian interpretation of quantum mechanics, every particle is a particle and a wave always.
And that the wave influences the particle based on where the wave is most intense.

But does that mean manipulating the which-way information affects the wave aspect? Otherwise it would always show an interference pattern (over time).

 

[RProgrammer]

Which pattern?

Now take a modification of the experiment where many photons are sent to the interferometer and hit the detector so that a pattern could emerge (and those are the idlers?).

While the signal(?) photons are still traveling for such a time that the last idler has hit the detector before the first choice is made on a signal photon.

Then what might you see on the detector if measured prematurely?

 

[cesiumfrog]

RP, if you look at only the detector, you just see noise. Gaussian. Doesn't matter whether the "choice" has been made "yet" or "not". Read those other threads.

 

[RProgrammer]

Gaussian

So if you look at the detector prematurely you see noise?
Because if you always see noise this isn't much of an experiment.

 

[RProgrammer]

Signal vs Idler

Sorry, you had them correct JesseM,
The signal is measured by the detector, and the idler is put through the choice.

 

[cesiumfrog]

RP, the only way not to see noise is by looking at correlations.. which is obviously impossible until after you receive information about the result of the measurement on the other (idler) set of photons.

 

[RProgrammer]

Confusion

Ok, now I think I've been confused all this time.

Here's how I thought the measurement was taken (of the signal photons, D0 in the paper):

The signal photons either exhibited a particle or a interference tendency (although only to be seen over meny repetitions).
D0 was on a stepper motor moving left and right to capture the photons in a certain position.
D0 would be moved to the next position after a good number of photons were sent through from the pump through the whole setup.
Then the total number of hits from each position was mapped together with each other position to form a sort of 2D graph from which either an interference pattern or a particle pattern would be seen.

Please correct me if I'm wrong.

 

[Demystifier]

I already pointed out a few times that many experimental papers on this issue, which are *experimentally* very rich, are very misleading/poorly written on the interpretation side.

I guess that both Bohmians and MWI-ers don't find it a big deal (although nevertheless quite fun). The "projectionists" are of course a bit puzzled because they don't know anymore when to project
(although, if you really understand von Neumann, you wouldn't be puzzled either).

Good points!

 

[cesiumfrog]

Please correct me if I'm wrong.

..or you could just carefully read the original journal article yourself. Or those other threads you've been directed to here.

Think of it this way: if it worked the way you were assuming, it would be possible to tell the future or transmit information faster than light (and if there's one thing that physicists certainly would have noticed by now...).

 

[Abiologist]

Kim et al.

RP, the only way not to see noise is by looking at correlations.. which is obviously impossible until after you receive information about the result of the measurement on the other (idler) set of photons.

I thought that the signal photons were passing through a Young's slit type apparatus - in which case shouldn't they usually show an interference pattern overall ? - in other words why should the interference pattern only appear with the correlations ??

 

[cesiumfrog]

I thought that the signal photons were passing through a Young's slit type apparatus - in which case shouldn't they usually show an interference pattern overall ?

not if there's some way (in principle) to figure which slit each went through

 

[Abiologist]

Kim et al.

Many thanks for your reply. However, if we choose not to observe which slit the photons went through - then I suppose the photons are free to produce interference patterns. In which case it is surprising to me that the interference patterns produced by two sets of photons cancel to give a Gaussian. Could you tell me in simple language where the phase shift comes from - and why this is not present in the original Young's slit experiment ??

 

[cesiumfrog]

One way to interpret it is this: in the original Young's slit, once the signal photon's position is measured, there is no way even in principle to determine which slit the photon came from. But in the DCQE, when the signal photon's position is measured, it is still possible in principle to determine (by making a particular measurement of the idler) which slit the photon went through.

 

[Abiologist]

Kim et al

Many thanks again. It appears from further reading that there are many interpretations of quantum mechanics which physicists have difficulty in deciding which one is the more appropriate.
In Kim et al 1999 the two wave components from an idler photon are well separated and presumably pass through D1 and D2, impinge on the beam splitter BS thereby erasing the 'which slit' information - allowing the interference pattern to be seen from the signal photon.
Does this mean that it might be possible to send one of these wave components a very long way and back again - and then to the beam splitter BS - (processing one photon at a time) - if the same results are obtained this would provide experimental evidence that the wave components of a photon are effectively timeless and probably infinitely extended ?

 

[Cane_Toad]

... would provide experimental evidence that the wave components of a photon are effectively timeless and probably infinitely extended ?

I thought this was already implied because it is a wave "function". Yes/no?

 

[Abiologist]

Kim et al.

I thought this was already implied because it is a wave "function". Yes/no?

Implication from a theoretical equation does not negate the positive benefits of getting empirical evidence to support the relevance of the equation - does it ?

 

[JesseM]

In Kim et al 1999 the two wave components from an idler photon are well separated and presumably pass through D1 and D2, impinge on the beam splitter BS thereby erasing the 'which slit' information - allowing the interference pattern to be seen from the signal photon.
Does this mean that it might be possible to send one of these wave components a very long way and back again - and then to the beam splitter BS - (processing one photon at a time) - if the same results are obtained this would provide experimental evidence that the wave components of a photon are effectively timeless and probably infinitely extended ?

I don't understand what you're asking here. You can in theory make the path the idler travels before reaching the beam splitter as large as you want (although this would probably increase the number of idlers which simply miss the beam-splitter and the detectors altogether), but why would this show that "the wave components of a photon are effectively timeless and probably infinitely extended?" You could send an ordinary electromagnetic wave in classical electromagnetism a large distance and still measure its effects at the spot you send it to, would this show that electromagnetic waves are timeless and infinitely extended? If not, what's the difference?

 

[Abiologist]

You could send an ordinary electromagnetic wave in classical electromagnetism a large distance and still measure its effects at the spot you send it to, would this show that electromagnetic waves are timeless and infinitely extended? If not, what's the difference?

Yes you're quite right. It seems that the results have a lot to do with the nature of the beam splitter. If two wave components hit the beam splitter (whether or not at the same time) then the photons segregate in order that their entangled counterparts give (or rather gave) interference patterns. It is incredibly strange that they are able to do this.

 

[Abiologist]

would this show that electromagnetic waves are timeless and infinitely extended? If not, what's the difference?

It seems credible to me that in fact the future paths of the signal and idler photons (or rather their wave components) are already mapped in some way - and that is why the system reacts as a whole to a change to the environment.

 

[johnson15]

Photon can go back in time, and change the source.

 

[johnson15]

...or the 'witch way' idea is wrong.

choice...

 

[insightforge]

Now take a modification of the experiment where many photons are sent to the interferometer and hit the detector so that a pattern could emerge (and those are the idlers?).

While the signal(?) photons are still traveling for such a time that the last idler has hit the detector before the first choice is made on a signal photon.

Then what might you see on the detector if measured prematurely?

Given the point above, I am still confused about why you can only interpret the results at the detector as a Gaussian pattern? With a sufficient delay you would be able to generate a set of points at the detector that should either fit an interference pattern or not fit. I see a lot of comments about the necessity of coincidence counting with the detector results. If coincidence counting is the game breaker about FTL communication couldn’t you overcome the issue by setting up a standard method for how you interact with the signal photons?

For example, let’s assume you are not constrained by the length of the delay that you can add to the system (you use a Bose Einstein condensate to slow down the signal photon significantly or bounce it of an object far away in space). Given your large time margin in your ability to decide whether or not to measure the spin on the signal photons you now have significant time to add some complexity to how you send and interpret the message. Setup the system so that you do everything (measure the signal photons & analyze the idle photon detector patterns) in discrete time intervals. This interval would most likely be the amount of time required to collect enough data at the detector to determine, with a high probability, if the patter was either an interference pattern or not (assuming that you also knew whether or not you were measuring the signal photon for the same interval). Now, with these discrete timeframes of measurement at the detector, you would also use the same discrete time intervals to either measure the spin on the signal photons or not measure the spin. Based on this system, it seems that you could take the data that you collect at the detector over these discrete intervals and apply two coincidence set, one in which you assume there was a measurement of the signal photon for the discrete period and one where you assume there was no measurements. If in fact there was no measurement (ie you did not decide to measure the signal photon in the future during the same discrete interval) then the assumed coincidence count for the measurement model (ie you only have the detector data, but you apply a coincidence count as if you had the signal data with a measurement of the spin taking place) for the same discrete interval should produce an interference pattern at the detector. Finding an interference pattern under the assumed coincidence count for the discrete interval assuming there was a measurement would not fit the data for a series of photons with a collapsed sping passing through a perpendicular double set of polarized filter in front of the double slit. There for it seems that you would know that you did not measure the signal photon for that discrete interval in the future. Given your ability to discern you interaction with the signal photon over discrete intervals in the future could you not then setup a binary, morse code communication system to pass information back to the past?

Disclaimer – I have no formal education in quantum physics, and this particular experiment has always interested me. I am sure there is some specific reason why the proposed method does not work for FTL communication. I was thinking maybe it has something to do with a requirement of knowing if the signal photon does or does not pass through the measurement filter. Another way to create the same proposed system would be to measure the number of photons hitting the detector for the discrete intervals. If there was a measurement of the signal photons there should be half as many photons making it through for the same period. It seems to work that way too. Please let me know why I am wrong as I am sure that what I am suggesting is a bunch of junk.

 

[msumm]

insightforge: from what others are saying, it sounds like you will not see an interference pattern when you look at the cases in which you didn't determine the which-path info. you only see an interference pattern when you look at a subset of those cases -- in terms of the diagram mentioned earlier in this thread, you only see the interference pattern when you look at the cases when a photon was registered at detector D1 (or only D2). someone could correct me on this, I'm no expert, I'm just reading the thread because i had similar questions in the past.

in that diagram you could think of 4 different cases:
1.a.) idler photon detected at D3 and therefore we know the signal photon passed through "slit a"
1.b.) idler photon detected at D4 and therefore we know the signal photon passed through "slit b"
2.a.) idler photon detected at D1 and therefore we do NOT know which-way info
2.b.) idler photon detected at D2 and therefore we do NOT know which-way info

I think what they are saying is that if you look at cases 2.a and 2.b together, the signal photons do NOT show an interference pattern, instead that actually show the same pattern as is seen in 1.a or 1.b. If you look at the pattern produced in cases 2.a only, then you do indeed see an interference pattern, but you cannot know that you are in case 2.a until after you detect the idler at D1 (and then it's too late to change the measurement, ...). similarly for 2.b, and evidentially the interference pattern for 2.b is slightly different so that when you add it to the interference pattern from the 2.a cases you get the same pattern seen in the 1.a./1.b. cases.

 

[YLW]

Hi, I'm new and I've just been reading about the DCQE. I have very little physics knowledge so please bear with me if I ask stupid questions. It was explained many posts (and quite some time) ago in this thread by JesseM that the experiment does not show that the 'choice' made later in time does influences the earlier positioning of the signal photons.

But even if we have to look at the subsets of correlations between D1/D2 and related signal photons to see interference, doesn't that suggest that there is some kind of influence made on the earlier detection of the signal photons by a later "choice"? I mean, how would the signal photons which hit D0 arrange themselves in a way that would show interference when taggedwith their twin photons which hit D1 when at the time the signal photons hit D0, we do not even know that their twin photons would hit D1?

Does that suggest that the signal photon that hits D0 already knows that its twin will hit D1 and if it does, wouldn't there at least be a suggestion that there is some backwards-in-time information transfer? Does a violation of causality mean that one would need to tell from looking at signal photons where entangled twin is going to land or is it enough for there to be a suggestion that an event in the present is influenced by an event in the future?

Thanks so much. So sorry if I wasted everybody's time.

 

[hankaaron]

When a particle is vibrating, it is more wave than particle.

However, when the particle back coils from the energy of a

photon, or in the case of a photon, it loses energy after

interaction with an electron, it temporarily ceases to vibrate

and instead becomes a point particle- that is, it is more

particle than wave.

So in the double slit experiment, the undisturbed particle is

mostly a vibrating wave, and hence interference is seen. If the

particle is observered (i.e., measured) then it ceases vibration

and becomes a point particle.


In the Quantum Erasure Experiment, if one of the entangled

particle is tagged, then so is the other. This is because there

is no complimentary value to tagging. If one of the entangled

pair is tagged, then so is the other. So when the first

entangled particle of the pair is measured, the other one

becomes a point particle at the same time.

But in the Delayed Choice Quantum Erasure experiment, why does

the signal detection sensorshow interference? It is because of

discontinues. That is, the particle begans to vibrate again

when some other observable is unknown. It stops vibrating when

a complimentatry observable is known. So in other words, the

photon stops vibrating when tagged.

But what happens when the idler photon is further away? The

signal detector should have already been either a interference

pattern or non interference pattern long before the idler photon

completes its path (into a detector with path info or one

without path info).

To understand what is happening, it is necessary to continue to

view the two split photons as only still one. In one reality,

the photon went straight to the signal detector screen. In the

other reality, it went back to the idler detectors.

Moreso, the idler photon went to another splitter, and that

photon went to both an idler detector that discerns path, and it

went to an idler detector without path info. This is why the

signal detector can accurately reflect both interference and non

interference results. The idler detectors branch off into

different realities (or universes), but the signal detector is

"connected" to all three.

For L&R path, the photon vibrates. For Left or for Right path,

it stops vibrating. This isconsistent with the prevailing

theory behind Quantum Computing.

Signal detector (present, simutaneous with idler dectors)
/ | \
Left L&R Right (present, simutaneous with signal detector.

Causality is still observed).

The which-path information is irrelevant since all paths are

exercised.

But what stops the photon from vibrating when the path is known?
That is the one question I can not find an answer for. It seems

that as soon as we understand one aspect of QM, other questions

arise!

 

[JesseM]

Hi, I'm new and I've just been reading about the DCQE. I have very little physics knowledge so please bear with me if I ask stupid questions. It was explained many posts (and quite some time) ago in this thread by JesseM that the experiment does not show that the 'choice' made later in time does influences the earlier positioning of the signal photons.

But even if we have to look at the subsets of correlations between D1/D2 and related signal photons to see interference, doesn't that suggest that there is some kind of influence made on the earlier detection of the signal photons by a later "choice"? I mean, how would the signal photons which hit D0 arrange themselves in a way that would show interference when taggedwith their twin photons which hit D1 when at the time the signal photons hit D0, we do not even know that their twin photons would hit D1?

Why assume that the signal photons are influenced by the later event of the idler being received at D1 or D2, as opposed to the other way around? Can't you equally well imagine that depending on which D0 position a given signal photon is detected, that influences the probability that the idler will later end up at D1 or D2? If the signal photon is detected at a position that is closer to a peak of what will become the D0/D1 interference pattern, and closer to a valley of what will become the D0/D2 interference pattern, then that might influence the idler to make it more probable it will end up at D1 and less probable it will end up at D2. I think this is exactly what you'd find in the Copenhagen interpretation where the detection of the signal photon "collapses" the wavefunction for the signal/idler pair, altering the probabilities that the idler will later be detected at different detectors. Analyzing what's going on in the Bohmian interpretation or the many-worlds interpretation would probably be a bit more complicated (see this paper for a Bohmian analysis), but there shouldn't be a need for any backwards causation at any rate.