Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

How do you determine that a particle is/was entangled?

  1. Dec 11, 2008 #1
    If you encounter a single particle (photon, electron, etc.) in space, can you perform any measurement that will tell you whether that particle was entangled with another particle just prior to the measurement? More generally, can you determine if a particle is entangled if you know nothing about its entanglement partner(s) or "history"?
  2. jcsd
  3. Dec 11, 2008 #2

    Vanadium 50

    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor
    2017 Award

    You cannot look at a single particle and tell it "was entangled".
  4. Dec 11, 2008 #3


    User Avatar
    Science Advisor
    Gold Member

    That is true as far as I am aware.

    However, a *stream* of entangled particles do have an interesting property that allow it to identified as such: if the particles are passed through a double slit, they will not exhibit the traditional interference pattern (collectively).
  5. Dec 11, 2008 #4


    User Avatar
    Science Advisor

    True, but this will still not tell you, whether there is no interference pattern due to entanglement or whether the light is just extremely incoherent.
  6. Dec 11, 2008 #5
    good question .. great insight

    I believe that although it is not currently accepted that there is any viable manner in which we can scientifically test for entanglement - That such a method does in fact exist .

    A thought experiment or two reinforces this position . But more strongly , Scientific foundations themselves both Philosophically and experimentally defend such 'optimism' .

    If there exists a physical behavior that is repeatable and predictable , then there should be 'possible' (even when technologically unavailable) such knowledge .
    A person can be unaware of water inches under there feet and dry of thirst , but it is not the lack of water that the person dries of but the lack of working knowledge {technique} to access that water .

    It is my belief that all quantum are entangled . But that the open 'communication' channels for entanglement vary according to as-yet not well disclosed criteria .

    Once the communication association rule for quantum entanglement was accepted , it opened a whole new domain . We scientifically accept that this domain allows associative math (1/2 -> 1/2 spins commute etc) , but there is no deterministic principle nor basis theory why communicative operators must stop there .

    And here is a thought experiment .

    a, Disassociate the human concept of 'knowledge of self' from the physical .

    b, Allow this concept of awareness to be applied to all things .

    c, Attribute 'knowledge of self' to the particle in question .

    c, Can the Particle have knowledge of its entanglement ?
    Last edited: Dec 11, 2008
  7. Dec 11, 2008 #6


    User Avatar
    Science Advisor
    Gold Member

    Well, you can solve that 2 ways: use coherent light (which is the source for PDC entangled photons anyway) or send the particles through one at a time.
  8. Dec 11, 2008 #7
    No; this is still a widely repeated (were it true easy to confirm) but unsubstantiated claim often misunderstood.
    Even with a coherent light source if the two slits, and observation screen set up with it are in a “Near Field” configuration, it is geometrically impossible to give an interference pattern. If the distance from the single beam PDC source to the double slit is increased to a "Far Field" distance configuration the individual beam will certainly produce an interference pattern imbedded in the dispersion pattern.
    That is why DCE style experiments like Dopher (She describes the “Near Field” set up and requirement in her reported results) need a Near Field condition set up so only correlations in the proper conditions (which way unknown) can cause the interference pattern to appear in the Near Field set up.

    The only thing that has ever objectively shown that “entanglement” exists is demonstrating weird correlation results.
    Being able to pick a select group of photons from a near field double slit set up that classically could never produce an interference pattern, is an example of confirming the entanglement. By considering only those photons embedded in the plain dispersion pattern, that match or correlate with select photons (which way unknown) from an “entangled” other beam of light to ‘magically’ construct an interference pattern is an example of weird correlation results that confirms the two beams have an “entanglement” characteristic. That another group of photons can be selected (which way known) will return to the no pattern seen as when the beam is tested without correlation does not speak to if the single beam was from an entangled pair of beams; it speeks to "which way" issue.

    As Cthugha implied, No experiment ever has rigorously demonstrated the ability to determine if a single beam source of light came from half of a “entangled” source. Some form of correlation statistics testing (which excludes any chance of testing photons individually) is always required.

    P.S. Hey Dr.C, I finally caught up to your post count!
  9. Dec 11, 2008 #8


    User Avatar
    Science Advisor

    Well, ok, but when using coherent light (stimulated down conversion) you destroy the entanglement - or at least you will be out of the regime, where you can say, which two photons are entangled. There is a paper covering these topics:

    Control of Young’s fringes visibility by stimulated down-conversion (P. H. Souto Ribeiro, S. Pádua, J. C. Machado da Silva, and G. A. Barbosa) - Phys. Rev. A 51, 1631 - 1633 (1995)
    There are also some follow-up papers.

    For the other argument: Sending the particles through one at a time does not help much. Light from down conversion is still equivalent to light with extremely short coherence length. If the slit distance is shorter than the coherence length you will see interference. Otherwise you won't.

    edit: RandallB is of course right. If I remember correctly, Zeilinger also demonstrated somewhere that single photon interference and two photon interference are complementary.
    Last edited: Dec 11, 2008
  10. Dec 11, 2008 #9
    There is the research in slowing down photons . http://www.hno.harvard.edu/gazette/2001/01.24/01-stoplight.html" [Broken]

    I wonder what happens to an entangled 2nd when a entangled 1st slows down .
    (another entanglement variable to be disclosed ?)

    Seems to me that slowing photons down must also have a bit of a deterministic effect on the probability of fundamental uncertainty also :)
    Last edited by a moderator: May 3, 2017
  11. Dec 12, 2008 #10


    User Avatar
    Science Advisor
    Gold Member

    A couple of points... I could not read the reference as I don't have a subscription. But another paper by one of the authors (Padua) provides another take on double slit and entanglement. It is a bit complicated, as there are a variety of issues.

    A double-slit quantum eraser (2001/2008)

    First, they were able to get an interference pattern out of entangled photons. They even can do this receiving which path information - a seeming contradiction.

    Second, that result is ONLY possible by coincidence counting. I am sure the same thing is true of other experiments involving interference patterns of entangled photons. Now, why am I so sure?

    Because third, IF you could get an interference pattern WITHOUT coincidence counting, THEN you could send FTL messages! We know that can't make sense. You would do this by choosing, at Alice, to either block one slit of the double slit or not; at Bob, you would see the double slit interference pattern either appear or disappear.

    See Zeilinger, page 290, figure 2, there is no direct interference pattern for entangled photons:

    Experiment and the foundations of quantum physics (1999)

    But fourth, that result appears to be somewhat at odds with the Padua et al experiment, their Figures 3 and 4, quoted below:

    FIG. 3. Coincidence counts when QWP1 and QWP2 are placed in front of the double-slit. Interference has been destroyed.
    FIG. 4. Coincidence counts when QPW1, QWP2 and POL1 are in place. POL1 was set to , the angle of the fast axis of QWP1. Interference has been restored in the fringe pattern.

    QWP1/QWP2 are at Alice, POL1 is at Bob. The only difference in the setups is at Bob, and yet this results in some photons appearing in fringe positions at Alice - which if true could be used to send an FTL signal.

    The only way that makes sense to me is if there is a different pattern for entangled photons (one bar in middle instead of the expected two, with sufficient diffusion in all positions to account for some at the fringe positions) than for unentangled photons. QED.
  12. Dec 12, 2008 #11


    User Avatar
    Science Advisor

    They can? Assuming you mean a single photon interference pattern this would puzzle me as this would usually mean that you have an extremely short distance between the down conversion crystal and the double slit, which means, that entanglement does not have much meaning anymore as the allowed range of k-vectors, which actually make it through the slits, becomes rather narrow. After having a look at that paper I see, that they did not have which way information and an two-photon interference pattern. They had either one or the other. This is a pretty standard DCQE setup.

    Well, I do not see your point. Of course there is no interference pattern for entangled photons (unless you increase the distance between the down conversion crystal and the double slit to rather large lengths.). I never opposed that. However a single photon out of an entangled twin photon pair created by spontaneous parametric down conversion behaves like thermal light with extremely short coherence length. Light with short coherence length will also show no interference pattern if the slit distance is larger than the coherence length. Of course one can increase the coherence length by increasing the distance between the conversion crystal and the double slit, which is equal to picking a smaller portion of the light source. However this also means, that the range of allowed angles corresponding to different k-vectors narrows, so entanglement becomes meaningless in the extreme limit. See for example the Dopfer thesis (unfortunately still not available in English, I suppose, but I can translate some parts if needed), which even gives a criterion for which distances between double slit and down conversion crystal single photon interference and two photon interference are visible or http://arxiv.org/pdf/quant-ph/0112065. Of course single and two photon interference exclude each other. So it is true that entangled photons do not produce a single photon interference pattern, but this is still by no means enough to identify them as entangled.

    Note that using coherent light for pumping does not change that, neither does decreasing the photon level to photons arriving one at a time. A single photon of an entangled pair is still no different from incoherent light, which is not entangled by any means.
  13. Dec 13, 2008 #12


    User Avatar
    Science Advisor
    Gold Member

    1. They did get which-path PLUS an interference pattern together, but that was ONLY with coincidence counting. The actual pattern at Alice was a dispersed single "hill" on one side. But once you orient the filter to determine which-path at Bob and then combine the results, the interference emerges.

    2. I do not get your point about the distance from the PDC crystal to the double slit. You can funnel the PDC output into fiber and do anything you want with it, just as with unentangled photons. There should be no issue getting an interference pattern if there is one to get.

    Now in the reference you gave (1994), I am not sure I see the relationship to this discussion. They used a single Type I crystal, which does not produce a polarization entangled beam. You need 2 of those (aligned perpendicularlyy) to get entanglement, and they didn't claim they were polarization entangled. They were coherent, as photon pairs coming out of a Type I crystal are of know polarization (both orthogonal to the input).

    Nor do I follow your thinking about single particles. All particles - be they photons, electrons, or large molecules - exhibit double slit interference with suitable slit arrangements. (Obviously a bunch released together incoherently may exhibit destructive interference.) On the other hand, PDC output is generally very low, on the order of perhaps 100-1000 pairs per second. So they are essentially coming through one at a time, and the issue of one destructively interfering with its successor is not an issue. Thus there is no issue of the beam being coherent or not, but technically they are all parallel or orthogonal to the one that precedes it anyway and so again coherence is not a problem.

    It is definitely not like thermal light - i.e. polarized in a random direction. All PDC output is either vertically or horizontally aligned - of course actually a superposition of those if they are spin entangled.
  14. Dec 14, 2008 #13


    User Avatar
    Science Advisor

    Just for defining terms for other readers, who might not recognize the terminology I use:

    What I call two-photon interference pattern is the pattern received in coincidence counting. What I call single-photon interference is the usual interference pattern you get directly in a Young type double slit experiment.

    Well, they get the pattern by doing projective measurements, which erase the which-path information. That there is only a pattern in coincidence counting is of course true. Only the two-photon-state is coherent.

    Fiber do just shift the difference. The point of taking the distance is as follows: single photon interference needs coherent light. So you either take laser light or you use a double slit. However for the double slit it is necessary, that the coherence length of the incident light is longer than the slit seperation. Otherwise the fields at the slits do not necessarily have a fixed phase relationship and you will still not get an interference pattern. Higher coherence corresponds to a lower spread in the emission angles - or equivalently wave vectors k. It is well known, that you can decrease the spread in k by increasing the distance between source and detector. This is what is used in stellar interferometry and this is why the HBT effect worked in the first experiments of Hanbury Brown and Twiss.

    Now, in order to get a two-photon interference pattern, the size of your interference pattern depends on the range of k-vectors you have. The wider your range is, the larger the pattern will be and the higher your visibility will be. So by choosing a large distance between slit and source, you can get a single-photon interference pattern out of a single one of an entangled photon pair. If you increase the distance you widen the range of k-vectors, the single-photon interference pattern vanishes and the two-photon interference pattern emerges. The Dopfer thesis covers this in some depth.

    Yes...with suitable slit arrangements and this is even true for entangled photons. As I said before, you can also get a single-photon interference pattern out of a single photon out of an entangled photon pair, if you use suitable slit arrangements. My point is, that you can't distinguish between entangled or unentangled photons by just looking at the absence of an interference pattern in a certain geometry. The only double slit, where this does not matter is a perfectly centered double slit, where the slits themselves do not have any width and are delta-like. But in this geometry you will again see a single photon interference pattern even out of entangled photons.

    The main property of thermal light is in my opinion that it is incoherent, which means for me that the coherence time is on the order of 1 ps or below. This determines the double slit geometry, in which you are able to measure single-photon interference, which brings me back to the beginning of the discussion.

    Well, in the correct geometry they will. And in the same geometry, where a stream of entangled particles does not show interference, you will also be able to find a light source, which does not produce entangled photons, but is so incoherent that there will also be no interference pattern. You cannot by any means know for sure, that a stream of photons is entangled by just looking at one half of the entangled partners.
  15. Dec 20, 2008 #14
    Thank you all for your answers to my query, which was perhaps naive, but seems to have led to some interesting discussions. What I had in mind was a possible FTL model, where Bob could either measure or not measure his particle, leading, if Alice could determine whether her particle had been entangled prior to her measurement, to a binary code that Alice could read with no knowledge of the outcome of Bob's measurement. She would only need to know that Bob had made a measurement, and thus "collapsed the wave function." I see now that this model has been presented on the Physics Forum previously ("Is it possible to determine if a photon is entangled?", posted by ACG 5/18/08.

    But is it crystal clear that entanglement does not somehow permit FTL communication? My thinking is as follows:The speed of light limit on signaling times results from a "classical" physics theory, i.e., relativity. But the apparent instantaneity (>10,000xc according to that recent experiment) of "communication" between entangled particles arises from QM theory. At this point, the two theories are of course incompatible, since no unified theory exists. But when experiments tested Bell's inequality, and thus whether (classical) local reality or QM entanglement is correct, QM won out. This implies to me that our current QM theory may ultimately prove a far more basic description of the universe than does our current relativity theory. If so, perhaps the final unified theory would in fact permit FTL signaling utilizing some aspect of entanglement.

    I would appreciate your comments on this admittedly intuitive argument.
  16. Dec 20, 2008 #15


    User Avatar
    Science Advisor

    bruce2g gave you an example in this thread awhile back, showing a graph in post #20 where interference was seen in the coincidence count but where they also showed the total pattern of signal photons without coincidence counting, and no interference was seen. In that post he also mentioned that the paper derived the predicted probability distribution for the total pattern of signal photons, and that it was simply a constant function, obviously implying no interference. So both on a theoretical and experimental level, this paper shows that photons entangled in a certain way will not show interference when you look at the total pattern going through a double slit.

    I also mentioned some other papers which also seem to show experimental demonstrations of this in post #24 of that thread. You never responded to either bruce2g's post or mine.
  17. Dec 21, 2008 #16


    User Avatar
    Science Advisor

    Just one comment: one has to be careful about the statements one derives from these experiments. While it is true that an ensemble of signal or idler photons will not show interference in most common geometries, which allow to maintain entanglement, this is by no means a proof, that you have entangled photons present contrary to the claims at the beginning of this topic. You can say, that the light present must be very incoherent, but nothing else (see for example the Dopfer thesis, page 44-47).
  18. Dec 21, 2008 #17


    User Avatar
    Science Advisor

    Yes, I didn't mean to say that having a lack of interference pattern is itself de facto proof of entanglement; I was just addressing RandallB, who doesn't believe that even for a source of coherent photons, if they are entangled their total pattern will fail to show interference when you send them through a double slit.
  19. Dec 21, 2008 #18
    1) If we think about an entangled pair, say separated
    by a large distance. If one state in observed for
    one in the pair then we know at that exact time what
    the other state would be if it were possible to observe
    it at the exact same time. That's roughly entanglement.
    But nothing physically happens to the partner, its just
    that its wave function passed through a known point at a
    known time (we could theoretically measure it - and then
    argue for a long time about the results in this forum
    ) the particle simply
    continues as if nothing happened. There's nothing we could detect
    to tell us that this happened. But the entanglement has gone
    now as any subsequent observation would be uncorrelated.

    2) The Quantum Eraser uses an entangled partner as marker.
    Isn't it true that entangled particles have a common wave
    function? Is anyone else worried by the claimed results of
    this experiment?
  20. Dec 21, 2008 #19
    Bare with me, as i'm only in year 12...

    But does this "entaglement" mean that something is travelling at faster than the speed of light, communicating what happens to it's partner?

    Feel free to totally bash me :)
  21. Dec 21, 2008 #20


    User Avatar
    Science Advisor
    Gold Member

    That could be true, but that is not the definition of entanglement. Entangled means they share a common wave state. We do not actually know what happens at any lower or more detail level.

    What is true is that it acts "as if" they were in instantaneous contact. But there are some other explanations for this behavior that do not involve FTL propagation. One example is the Many Worlds Interpretation (MWI).
  22. Dec 21, 2008 #21
    Oh, I think I understand that it's by no means a definition, I just meant that as a consequence...

    Once again forgive my lack of understanding, but are these two electrons connected via cosmic "strings"?

  23. Dec 28, 2008 #22


    User Avatar
    Science Advisor

    I don't think that the "coherent photons"-part is strictly true, unless you just mean the seed for the down-conversion crystal. Having coherence and entanglement present simultaneously is very complicated. Stuff like maximally entangled coherent fields are usually used in quantum telportation protocols, where one wants to transfer entanglement from a pair of atoms to some light beams.

    In an usual down conversion setup coherence and entanglement exclude each other pretty much. As a rule of thumb coherence is a measure of indistinguishability. You can't distinguish photons inside one coherence volume. The whole idea of entanglement is now a bit spoiled if you have a few photons emitted simultaneously, those are pair correlated, but you don't know, which of the photons in one leg of the experiment corresponds to one certain photon in the other leg of the experiment. In this case the two photon interference should be drastically reduced.

    The missing single photon interference pattern is a hint at some other phenomenon: nonclassical light emission. Getting the photon number noise below the shot noise limit at low photon numbers also results in strongly reduced coherence time - the ideal case of a completely deterministic single photon source would never show an interference pattern. In most real cases the coherence time is on the order of what one would expect for thermal light, mostly due to some mixing with the vacuum state.

    Of Course I agree with your conclusion. Photons, which are entangled in a meaningful way will not show single photon interference in a double slit experiment.
  24. Dec 29, 2008 #23


    User Avatar
    Science Advisor
    Gold Member

    No one really knows the specifics of the underlying mechanism. However, you don't need to postulate strings to complete the theory itself... it is essentially complete as is.

    String theory is a work in progress which is attempting to connect gravity and the other fundamental forces. It is not directly related to entanglement.
  25. Jan 14, 2009 #24
    Guess I missed updating this thread over the holidays.

    Your first link here shows that the Zeilinger claims in your second link are wrong.

    Zeilinger incorrectly uses the results from “Near Field” experiments.

    Reread the experimental results in your first link with attention to:

    - The first two lines of sect V.
    - Also, the comments between formulas 6 & 7 to confirm that this actually preformed experiment was conducted in a “Far Field” set up. (A requirement to produce the interference pattern with or without the beam on the double slit originating from a entanglement source.)
    - The results in Figure 2
    They contradict the Zeilinger claim as we have discussed before in prior threads.

    Since no filters are in place anywhere in the Fig 2 test; 100% of the S photons are used.
    Even if the correlation counter turned off the same 100% of the S photons would be detected in the same place just as shown in Fig 2.
    Remember Dp I positioned to collect all photons unless blocked by a polar filter; no lens or slits scatter any of them away from Dp. Ds must be moved due to the slit scattering.

    Thus: A far field beam, even if one of a pair of “entangled” beams, will give a interference pattern when applied to a double slit. QED.

    The trouble your are having in understanding the results of this experiment is mostly related to understanding the QWF filters. Remember 100% of the photons are still going through a QWF. Thus the only time 100% of the photons going through the slits are not being detected is when a polar filter is used on Dp.

    Although they didn’t show an example it helps to realize that if they set the polar filter to θ + π/4 with the QWF’s in no pattern would be seen.
    They likely assumed that as understood as normal and only reported the “interesting results. I’m sure they saw that during their set ups, just as they likely saw the single beam interference pattern with coincident counting as normal during the set up.

    Wish they would have documented their set up observations in more complete detail.
    It such a widely held misconception with plenty of False proofs sited to support Zeilnger that usually pointlessly apply “Near Field” results like Dopher. (Some just have a hard time understanding Near vs. Far field issues on this one, even though Dopher covered it in her paper.

    Although the results here are obvious, without including the detail from their set up observations I’m sure some can go on for pages of denials to spin these results to their personal opinion.
    But they cannot produce an equivalent experiment with the set up observations recorded that does not show interferance.
  26. Jan 14, 2009 #25


    User Avatar
    Science Advisor
    Gold Member

    Zeilinger is not wrong. I think the issue is the lingo you are using (or Zeilinger and I are using, as the case may be).

    There are 2 different cases, which in the referenced experiment are distinguished by coincidence counting:

    a) where there is NO possibility of determining which slit: and in this case there IS an interference pattern. To get this case, which only occurs sometimes in this experiment, you must "erase" the possibility of determining which slit for the entangled partner.

    b) where there IS a possibility of determining which slit; and in this case there is NO interference pattern. This cases occurs naturally much of the time even when you are trying to "erase" the possibility of determining which slit for the entangled partner.

    When you add the a) and b) cases (i.e. ignore coincidence counts), you still get a pattern which looks ALMOST like b) above. If that were NOT true, then there would be a potential for an FTL signaling mechanism.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook