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

Double slit experiment - QM or wave?

  1. Aug 21, 2008 #1
    I propose a setup to distinguish between the quantum interpretation and the wave interpretation of the double slit experiment.

    Briefly, they are:

    The photon gun fires a photon that travels through either slit A or slit B, then hits a detector cell registering a spot in the detector pattern. All photons follow the same procedure, and with the help of quantum mechanics, the result is a nice interference pattern.

    The photon gun fires a photon that travels through both slit A and slit B, then hits a detector cell with either destructive or constructive interference, depending on the location of the detector cell. If the superposition is "too destructive", no spot will register in the detector pattern. All photons interfere destructively or constructively and the result is a nice interference pattern.

    So how do we distinguish between these interpretations? We do it by counting detections.

    If the photon gun fires 1000 photons per minute, quantum mechanics predict that, during a minute, a 1000 detections will be registered (ideally). The wave theory on the other hand, predict that during a minute, only photons that are not interfering destructively will register at the detector.

    So the setup is simple. First fire the photon gun without the double slit to determine the photon firing rate, then add the double slit and count the detections.

    Since I adhere to the wave interpretation, I predict that there will be a significant difference between firing rate and detection rate. However, if this was the case, you could still argue that the non-detections "got lost" somewhere at the slits. But even if this was the case, you could still distinguish the theories from each other. Because, if the quantum interpretation is correct, the "lost photons" are lost, just random mishaps at the slits. On the other hand, in the wave interpretation all "lost photons" are destructive interference. This means that, if we alter the detector so that the destructive areas (dark bands) become constructive instead of destructive, there should be a significant increase in detections which cannot be explained by quantum mechanics.

    An illustration of the setup:

    Any thoughts?
  2. jcsd
  3. Aug 21, 2008 #2


    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor

    Hate to break this to you, but there's a more elegant experimental demonstration that has been done for undergraduate physics labs based on the Mach-Zehnder interferometer. See T.L. Dimitrova and A. Weis, Am. J. Phys. v.76, p.137 (2008).

    One of the component of the experiment that was done by the students was this:

    You'll have a tough time convincing anyone that your "wave interpretation" is valid with such observation.

  4. Aug 21, 2008 #3
    Oook? .. The quote you supplied completely proves my point. I don't get it.

    The "revival of the clicks" cannot be interpreted as anything other than elimination of destructive interference.

    I'm guessing this is more of a semantics issue now then. What would you say is the difference between a wave and a photon that "interferes with itself"?
  5. Aug 21, 2008 #4


    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor

    The difference happened at the beam splitter. You need to read the entire article, and also the J.J. Thorn et al paper.

  6. Aug 21, 2008 #5
    I don't have access to the article via AJP. The Thorn paper seems to be more of a bell test setup with beam splitting and coincidence counts, i.e. no interference. So how is it relevant?

    Concerning duality, different experiments show different things, and one does not take precedence over another (even if the wave interpretation is more easily observed). My original point was that the double slit is one of those experiments that cannot be explained satisfactory with a quanta of light. But you obviously believe otherwise, so I ask you. Why is the double slit not explained with a wave model of light? Which interpretation would Occam prefer?
  7. Aug 21, 2008 #6


    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor

    The point being that "interference" by itself isn't sufficient, because both scenario results in the same observation (unless you want to consider higher order interference). What is important is the "which-way" observation that can't be resolved via the wave scenario.

    But it can't. If you look at ALL the observation, your application of Occam will require the minimal sent of description to describe ALL, not just one, not just two, observations. I haven't even yet invoked other phenomena such as the antibunching observation, multiphoton photoemission, etc. that current has zero description via wave scenario. Do you think your Occam principle will allow you to switch gears of describing light one way for one set of experiments, and switching to another to describe another set? I can stick to just one for all of them.

  8. Aug 21, 2008 #7


    User Avatar
    Science Advisor

    The whole idea is completely flawed. Interference is a phenomenon arising due to superposition of possible paths/probability amplitudes. There is no such thing as photons just interfering destructively and vanishing completely. For every zone of destructive interference, there is also a zone of constructive interference, so the total photon number does not change anyway. Otherwise this would be a harsh violation of energy conservation.
  9. Aug 21, 2008 #8
    Which scenarios? and what about the "which-way" observation can't be resolved?

    Both my setup, the quote you supplied, and Cthugha's post tells you why interference is enough, and both interpretations do not result in equal observations. QM predicts no change in clicks / number of detections when blocking one path or changing the resulting interference pattern. Wave theory does.

    I'm not sure you can. How do you fit longer wavelength radiation in your description? radio transmission, microwave ovens... and refraction, diffraction? .. Even if you could, would you consider your description to be the final say in the matter? I can certainly appreciate the benefits and practicality of having a single description for all phenomena, but that doesn't mean I'm ruling out all other possibilities.

    So please, I'm interested... Just looking at the double slit experiment, what about it makes you consider it to be better explained with quantum mechanics than with a wave model? Because I've always wondered why physicists use it as an example of how special QM is, when the actual explanation involves quanta of light "interfering with itself". What does this mean, and how does it differ from a wave model?

    Read the quote in ZapperZ first post. Your claim is not true locally in an experiment like this, the photon detection does change. At the position where there is destructive interference, the light doesn't interact with the matter (eg. the detector), so it continues it's propagation and eventually interferes differently and interacts with something else. Energy is conserved, but not confined to the experiment.
  10. Aug 21, 2008 #9
    You have grossly miss read something somewhere. Half the clicks with no interferance pattern just a dispersion patttern just like wave theroy does is excactly what QM says.

    Just google for double slit basics or reread whatever your looking at.
  11. Aug 21, 2008 #10


    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor

    In a beam splitter with wave picture, half of the way goes through while the other half gets reflected. In a photon picture, a photon has a probability of 50% to go through and 50% to get reflected. This means that if you put your detector over both branch of the interferometer, you'll only get one detection, not both, in the photon picture. This is what the J.J. Thorn et al. expt. detected.

    I don't see how Cthugha's post contradicts mine. I described a generic interference experiment and that you can described it with both pictures.

    Actually, yes you can since YOU were the one who wanted to invoke Occam. Refraction and diffraction requires interaction with the material, and in condensed matter physics, optical transport are often described using the photon picture. See for example, the microscopic explanation on the interaction of photons with the phonon structure of a material that can result in the explanation for why things are opaque, transparent, and why light moves "slower" in a dispersive material.

    Shoot one photon at a time, i.e. light with extremely low intensity. Look at it over time. Now do the same with electrons, protons, etc. Tell me if what you observe is different. If yes, then light is different than the "physical particles" such as electron, protons, etc. and you may have a leg to stand on. If they look the same and have the same characteristics, then you have a lot of explaining to do.

  12. Aug 22, 2008 #11


    User Avatar
    Science Advisor

    Of course the photon detection rate is changed locally. This is what interference is about. So the total count rate of the detector will only change, if the detector is small compared to the total dimensions of your interference pattern. But if you use a small detector and place it inside a spot of destructive interference, qm also tells you, there will be a significant difference between the counting rate with and without double slit in this spot. So I do not see any difference.

    The only possibility to have no change in the counting rates would be to use a deterministic photon gun, which fires single photons in completely defined directions at completely defined moments. This would not produce an interference pattern. However this case corresponds to the case of completely incoherent light, where the distance of the slits is far longer than the coherence length of your wave. So any wave model will not predict any changes in the counting rates either.
  13. Aug 22, 2008 #12
    Cthugha beat me to it. I'm afraid it looks like your conceptual understanding of the two-slit experiment is wrong. The experiment has been performed many times before, and the QM description wins.
    Suppose your "wave" interpretation
    were correct. Then a single photon, travelling through the slits, would produce dots at the location of every constructive interference fringe. This simply isn't what you see when you do the experiment.
    Also, it's technically wrong to say that in the quantum description
    The WHOLE POINT of QM is that this is NOT what we say happens. What we say happens in QM is that there is a wavefunction, which goes through both slits, and interferes with itself constructively and destructively at the interferences fringes along the screen. The square of the amplitude of this wavefuction at a point describes the probability that an electron will be detected at that point if we choose to make a measurement there. So the interference pattern is slowly manifested on the screen as increasing numbers of photons are detected in the places earmarked as 'probable' by the wavefunction. When a photon is detected in one place, its wavefunction is said to "collapse", and there's no longer any chance of finding it anywhere else.
    Last edited: Aug 22, 2008
  14. Aug 22, 2008 #13
    I think you're right about the double slit experiment. Feynmann once said that all the essential mysteries of QM are included in the double slit experiment, but he was talking about
    a thought experiment using electrons, not light. He meant that the essential mystery is that particles (like electrons) show interference. He never said anything about the double slit experiment with regard to the wave vs particle theories of light.

    People also invoke the image being built up one click at a time, or one silver halide crystal at a time as evidence of the particle nature of light. This argument of course has nothing to do with double slits or triple slits or any kind of interference. It seems convincing at first glance but is actually quite hard to nail down. The problem is that it's hard to know when you have exactly one photon at a time: and both the wave theory and the particle theory allow for a diminishing probability of coincidence counts as you make the source weaker and weaker.
  15. Aug 22, 2008 #14
    The evidence that light has some particulate nature is nothing to do with the two-slit experiment. It's to do with the photoelectric effect.
  16. Aug 23, 2008 #15
    I'm not disagreeing with you, just would like to add that, however, the two slits experiment, in my hopinion, has to do with the fact that a corpuscle of light has or not a definite individuality from the source to the detectors. What I mean is this:

    If a corpuscle with such individuality exists, it must go through one or the other or both (splitting itself) the slits. It's well known that if it goes through one or the other, you don't get an interference pattern; it's probably less known that if it splits in two and goes through both...nothing changes because you have two particles passing through the slits and the same you don't have an interference pattern (ah yes, you could say it interferes with itself, but this is what a wave does, not a corpuscle).

    The only possibility, in my opinion, is that the photon has not a definite individuality between the source and the detectors, the field constantly creates and destroyes it, so it's the field that passes through both slits (exactly how you wrote), interfering with itself and then creating the single photon which hits the detectors.

    Just another crazy idea, of course. :smile:
    Last edited: Aug 23, 2008
  17. Aug 23, 2008 #16
    Hey all. I appreciate you joining the discussion, but I really would like to narrow this thread down. Wave-particle duality is a huge subject that can be debated forever, so I would appreciate if we could focus on the original statement. Which was...

    • QM and wave theory predicts different outcomes in the double slit experiment. Wave theory predicts a significant decrease in detection rate compared to firing rate, QM does not.

    Is there any data on this subject?

    What I'm talking about, is using a detector that will cover the entire area behind the slits, so that all photons passing the slits can be detected. In that case, QM predicts that... when a photon is fired, it's probabillity wave will pass the slits, interfere, and collapse at the detector determining where the photon will strike. This means that all (ideally) photons will strike the detector. This is NOT what wave theory predicts. In wave theory, the dark fringes in the pattern is a result of the fact that this area was hit by a photon (wave) that was interfering destructively, and therefore didn't set off the detector. This means that wave theory predicts a significant decrease in detection rate compared to firing rate. Unlike QM.

    Off Topic:

    And why would a quanta bend it's path when interacting with the material?

    You know as well as me that these experiments are on a completely different scale when shooting particles. Based on what I've been able to find on the web the setup is very different, but results are kind of similar, yes. They both display fringes anyway. I have some questions though. Have a look at the video on this site: http://www.hqrd.hitachi.co.jp/em/doubleslit.cfm
    Here's a quick analysis of the brightness of the fringes: http://physics.soderholms.com/electron_fringes.jpg
    Which fringe would you say is the center one? A typical double slit pattern has an intensity distribution looking something like this: http://img.sparknotes.com/figures/C/c33e2bffc162212e1d9aa769ad3ae54f/envelope.gif
    It looks to me like, either there's something shifty going on with the intensities, or the center fringe is a minimum. So I'm not convinced it's actually an interference pattern we're looking at. In the setup (http://physics.soderholms.com/electron_biprism.jpg) there's a few (to say the least) things in the way of the detector. And the paper (American Journal of Physics 57 117-120) doesn't say anything about what it looks like without the biprism, or what happens if you change the distance to the detector, or block one of the paths. It's basically an experiment set up and performed only to confirm quantum mechanics. They might as well just put the conclusion in there.

    What are you saying, that light can't be wave because we can shine a dot on the wall? That light must be corpuscular because it travels in straight lines? That's a bit 18th century of you.

    Anyway. Too much off topic now.
  18. Aug 23, 2008 #17


    User Avatar
    Science Advisor
    Gold Member

    ZapperZ is correct and you are closing your eyes to an experiment that answers directly the question you are asking: whether photons are discrete or not. The Thorn experiment (which is simply an undergrad version of experiments previously performed by others) asks: are photons simply waves of probability? The answer is a big NO. There are definite predictive differences - as you imagine - between various possibilities. The one you suggest is clearly ruled out, by 377 standard deviations if I recall. The QM predictions are confirmed in the process. *Read* the paper. I might suggest you quit insisting that the experiment is not relevant when it obviously is. How can anyone take your suggestion seriously when you ignore facts you don't like?
  19. Aug 23, 2008 #18
    Sorry but I don't understand it. Can you elaborate? Also, where exactly is the difference between firing rate and detection rate?
    If 10 photons are sent in a time interval of one second, ten photons will be detected in a time interval of one second, wave theory or QM.

    Also, when you say that:
    1. Your description of the "Quantum interpretation" is wrong, as others have said, because "a photon that travels through either slit A or slit B" is exactly a photon of which you have the "which way" information, and that means no interference pattern. Example: you detect the photon near one of the slits putting there a little detector; clearly this is not the two slits experiment we are talking about.

    2. About your wave interpretation. As I've written in my previos post, if you are certain that the particle splits in two and passes through both slits, nothing changes, it's as if you had two photons passing simultaneously through both slits; it could seem strange, but you have the "which way" information the same, and so no interference pattern. Think to the fact that to say "the particles splits in two and both passes through the slits" means that you have that specific information, and to have it you should put two little detectors near every slit and detect the two particles (the two "half photon") near the slits; so, no interference pattern. Unless you invoke De Broglie Bohm interpretation, the only possibility is that no particles passes through one or the other or both slits. (In my opinion).
  20. Aug 23, 2008 #19
    I'm not saying that light can't be a wave because we can shine a dot on the wall, and I'm certainly not saying that it's corpuscular because it travels in straight lines.
    What I'm saying is that if light were a classical wave, then an arbitrarily small "amount" of light would show evidence of an interference pattern when diffracted through one or more slits. It's perfectly possible to "shine a dot on the wall" with a laser pointer (say), but if you take a laser pointer (emitting innumerable photons) and shine it through a diffraction grating you see an interference pattern, clear as day. If light were a pure, classical wave, then it wouldn't be possible to shine any amount of light through a two-slit setup in such a way that you didn't see interference fringes immediately.

    Lightarrow: Schroedinger originally suggested that there were no particles, only localised wave packets. Lorentz (I think!) demonstrated that such "matter waves" would spread out over time due to dispersion, and lose their localisation, which of course conflicts with the existence extremely well localised particles in the world today. This is the reason why people don't think of the wavefunction as describing a particle as being physically spread out. I suppose I can't honestly say, however, that your idea is any crazier than the idea that a particle pops into existence when we decide to look for it :rolleyes:
  21. Aug 23, 2008 #20
    That's not the question he was asking. Hydromatic is wrong on certain points of physics but he was quite clear on this point: he asks why people choose to use the double-slit experiment to support the particle theory over the wave theory. I wouldn't normally butt in on one of ZapperZ's discussions because of his tendency to use his power as a moderator to silence his opponents, but since you found it necessary to weigh in on his behalf, I thought it was only fair for me to point out that ZapperZ has not answered this question. In fact, the extensive quote he provides in his first answer shows that the wave theory rather than the particle theory is the obvious explanation for the double slit experiment. Hydromatic rightly pointed this out in his response.
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?