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

  1. Nov 1, 2005 #1
    From what I know of the double slit experiment you shine a light at two different slits and half of the photons go through one, and the other half go through the other (or a ratio similar to that), and i am pretty sure the second part of the experiment is to shoot one photon at the two slits and it comes out interfering with itself, can anyone go into further detail about this?
  2. jcsd
  3. Nov 1, 2005 #2
    I'm definitly not qualified to really answer. And quite simply from what I can tell, you get the basic idea.

    The only thing that I cant tell that you missed was the fact that if a photon is interfering with itself then it had to go through both slits. On photon had to have been in 2 places at the same time.

    You might want to read "The Universe in a Nutshell" and the 2 books by Brian Greene that I can't remember the title of but wre both really good.
  4. Nov 1, 2005 #3


    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor
    2018 Award

    The double slit experiment is actually a SINGLE-PHOTON INTERFERENCE phenomenon. [I capitalized that not to shout at you, but rather for emphasis]. The QM description of this phenomenon is that each photon has a superposition of paths since it can go through both slit with equal probability. So the superposition phenomena is the one responsible for producing the interference effect that you observe. Classically, this implies a picture where a photon interferes with itself. Two-photon interference is very rare (higher-order phenomena), and these do not produce the common interference pattern.

  5. Nov 1, 2005 #4
    But....what if you measure the photon before you shoot it at the two slits? is the singe-photon interference only if you havn't measure the photon so it therefore is missing position and spin according to GM?
  6. Nov 1, 2005 #5


    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor
    2018 Award

    What do you mean by "measure"?

    If what you do before doesn't tell you which slit the photon goes through, then it makes no difference. You still have the superposition. If what you do removes this and you have knowledge, or the ability to know which slit the photon passes through, then you have destroyed the superposition and you will not get the interference effect.

  7. Nov 1, 2005 #6
    Everybody interested about the strange things of QM should read Richard Feynmans QED: strange theory of light and matter. After you've read it you'll probably think that QM is even stranger than you thought.
  8. Nov 1, 2005 #7
    wait, i thought that a particle was never measured (any method of measurment) its position probability narrows, therefore if you measure it and the slits are still the same length apart....it shouldnt interfere with itself. right?
  9. Nov 1, 2005 #8


    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor
    2018 Award

    Remember that I ASKED what you mean by "measure". You have not offered any replies.

    You need to understand that the TYPE of measurement you make is crucial. There are non-commuting observables in which you do not destroy the superposition of one observable when you measure the other observable. That is how we could detect the superposition of the Schrodinger Cat-type experiments. You can't just say "oh, I made a measurement". That's ambiguous. I could make a measurement of the z-component of the angular momentum operator, yet the x and components remain in superposition of each other.

    Moral of the story: There are two very important aspect of QM that must be understood without fail - superposition principle, and the commutation relation of observables. In fact, that last part has often been called the First Quantization of QM. It is why we have the uncertainty principle.

  10. Nov 2, 2005 #9
    all im really asking, is there any way to shoot a photon at 2 slits, at a random angle (or whater random factor they use in the double slit experiment), is it possible to make it not interfere with itself?
  11. Nov 2, 2005 #10


    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor
    2018 Award

    Er... the 2-slit experiment shoot photons at "random angles"? Whoa! What experiment is this?

    There is still a disconnect with your understanding of what I have just said. If you REMOVE the superposition of paths that the electron can take, then you can destroy the interference pattern. If you do NOT remove this superpostion of paths, then you will get the interference pattern.

    You do not want to make it interfere? Then put a detector at one of the slit opening so that you have a "click" each time a photon passes by that slit, and no click when it passes by the other slit. You will then have 2 nice gaussian distribution centered behind each slit. Viola! No interference pattern.

  12. Nov 2, 2005 #11
    I don’t think your getting the whole picture here in short little bits:

    First, the double slit experiment is not a ‘single photon’ interference test. First done by Young at the start of the 1800’s it demonstrated that light actually was a wave as had been thought prior to Newton’s lights pieces (particles) ideas finally being accepted in the 1700’s. Then along came Plank in 1900 and we have Quantum’s of light! Clearly particles of light (that are not waves), cannot interfere with each other any more than bullets can (you have seen those diagrams right). So the issue is testing for it, early 20th century science was able to generate individual photons but even better individual electrons, and prove it.

    A test on just one and only one particle doesn’t mean much. Big deal it hits a spot on the detector screen, one or two slits there is no difference. Now for a Quantum Double Slit Test! By sending what we can prove are one at a time individual particles (even by ‘testing’ them) towards the double slit and then counting and mapping there locations we see the interference pattern slowly build just as we find in patterns of water waves (made up of lots of particles) going through two openings. That what we call a paradox, it’s just not possible; how do we explain it? What else more experiments, using the case of electrons, which amazingly are showing individual particle interference just as reliably as light, lets MOVE the test that confirmed we were sending just one photon at a time.

    This new test is the important one that you are referring to, not just testing to see that an individual particle is going towards both slits. Rather carefully moving the test up close to the slits and check only for those approaching slit one. Without a chance of accidentally detecting an electron headed to slit two. (It’s easier to test electrons than photons) Now we can carefully count and track what they do. Do these all stay on one side of the pattern or spread out evening into every part of the interference pattern?

    The results of your suggested test – neither, the pattern now looks like a single slit pattern with the electrons detected from slit 1 spread evenly with those from slit 2. The Paradox continues, why did the pattern stop, how do you explain it! You cannot, at least not until after 1926, then you can argue each particle has “superposition” allowing it to go though multiple slits. If you “collapse” the superposition at a point that it cannot reform and reach second slit, there is no reason to expect an interference pattern. The ability to have “superposition”, that's the whole point of QM. No “random angle” of test will change that.

    Now does this still fly in the face of any common sense understanding of reality? Of course, that’s why they say you must abandon common sense, at least while understanding QM. In takes QM to understand and resolve this paradox. If you can explain it better without superposition, than you need to write a book about it.
  13. Nov 2, 2005 #12
    The interference can be greatly reduced and essentially eliminated by the particle encountering other particles near the slits like a gas or a measuring device or many other things. The interference spreads into the other particles and is lost. We are then left with two possible paths which don't interfere anymore and we find one of them occurs. The other possibility may also occur but it would need to do so in a different reality to ours! The particle on its own will always interfere, however, and will travel through no definite slit. It needs other particles to make the possibilities seperate ones. :smile:
    Last edited: Nov 2, 2005
  14. Nov 2, 2005 #13
    Here is a site http://www.people.virginia.edu/~xy9z/qubit/qubit.doc" [Broken]that claims that a new experiment using single photons , offers definitive proof of single photon interference. See the detials of the experiment and the accompanying diagrams at the bottom of the page.
    Last edited by a moderator: May 2, 2017
  15. Nov 2, 2005 #14
    From the bad ref you gave us (Not a two slit anyway):
    Wrong QM does not say that a half-silvered mirror will divide one photon into two photons that can be detected by each detector. And experiments agree with QM that only one will see the photon, but you cannot predict which one.
    Best take this one off your review list, they have this wrong. At least until they update with what we can hope was a poor review from the Prof. is says it was last submitted to in 12-2003!
  16. Nov 2, 2005 #15
    I don’t think I get what your driving at with your other comments. If your suggesting something like MWI etc. those are better considered with none double slit experiments, like the foolish “dead cat” idea or maybe entanglement would be better.

    However this comment seems interesting. “The particle on its own”, going through the 2nd slit and not being affected by the test going on only in front of slit one. Should its superposition be affected at all??

    If we check the locations detected NOT in synchronization with the detections by the slit one detector, wouldn’t they be the particles going though the second slit. And if we found a pattern produced, wouldn’t that mean we would have a pattern while knowing “Which way” the particles came through! Violating QM rules!

    I’ve never seen results documenting a test in this detail, but my guess is we would should find no pattern. And the explanation that the testing going on in front of slit one is affecting or collapsing; the Guide wave, superposition, or whatever is trying to find its way though it to signal the untouched particle going though slit 2.
  17. Nov 2, 2005 #16
    I've decided I've got my detector watching both slits in the case when there is a measurement being made because an experiment with a detector watching only one slit leads to some awkward questions about interaction-free measurement. :biggrin:

    By interaction-free measurement, I mean that if you don't detect the particle at the slit with the detector, then you obviously know the particle is going through the other slit without any physical interaction telling you. It's a *lack* of physical interaction telling you! You then have, say, wave function collapse or universes splitting into seperate realities because nothing happened physically and effects like these without physically causes are somewhat headache-inducing. :bugeye:

    I think possibly the way out is perhaps what Gell-Mann and Hartle are meaning when they talk of "mutually exclusive" (only *one* possibility may occur) and "exhaustive" (one possibility *must* occur) interfering histories of the universe. I suspect they mean one possibility really *is* occuring but it can be supressed and even erased by the interference from the possibilities which aren't occuring. I know that's the case for the ideal von Neumann experiment so maybe that's the answer. I've got to admit I'm a little confused here. It's something I hope to get round to answering some time in the distant future. :uhh:

    Right now, I'd like to measure both slits in the case when a measurement is made in the two slit experiment and so not worry about anything else. :wink:

    Anyhow, in what I meant in my previous post, knowledge which allows you to assign a definite slit to the particle is always associated with no interference pattern on the detection screen. :smile:
    Last edited: Nov 2, 2005
  18. Nov 3, 2005 #17
    doesnt it have to shoot at a random angle? because if you knew the angle then you would know what slit it would go through.
  19. Nov 3, 2005 #18
    Like a flashlight covers a larger area down the line, same for the uncertainty of where the individual photon will be spreading out over a larger area down the line, that area must cover both slits. If you test close enough to the slits so that only one slit can be reached from the test point, is when you've changed from confirming one photon is being sent towards both slits, to checking for Which Way information.
    Once, you test for WW the pattern stops.
  20. Nov 3, 2005 #19
    The angle isn't important. Here's what's going on:

    When you shine light through a single slit, with a piece of photographic film on the other side, the various photons (acting as waves) bend around the obstacle as waves do. This is called "diffraction" and the resulting image will be a "diffraction pattern" such as

    _ __ ___ ____ _____ ______ _____ ____ ___ __ _

    (varying depending on how wide the slit is, among other things).

    When you shine light through a pair of parallel slits, the various photons do their bending thing. But there are now two sets of diffraction patterns coming out the other side. These waves interfere with each other, as waves do. The resulting pattern is therefore called an "interference pattern" and looks like this

    | | | | | | ||| | | | | |

    (varying depending on how far apart the slits are, among other things).

    Now, take your single slit, and beam single photons through it one at a time. As you would expect, they bend around the obstacle as waves tend to do, and the individual photons eventually build up a diffraction pattern like the one before.

    _ __ ___ ____ _____ ______ _____ ____ ___ __ _

    And when you take your double slit, and beam single photons through it one at a time, you would expect a doubled diffraction pattern to build up, because there's no other photon waves to interfere with on the other side of the slits.

    __ ___ ____ _____ __________________ _____ ____ ___ __

    But what actually happens when you send individual photons through the double-slit, is an interference pattern builds up. It's as if each individual photon's wave went through both slits at the same time, and then the two wave paths interfered with each other on the other side. In shorthand, you'd say it's as if the photon interfered with itself.

    | | | | | | ||| | | | | |

    But that's not accurate. The photon isn't "interfering with itself" because it isn't REALLY a wave. A photon BEHAVES like a wave, and also behaves like a partice, depending on how you look at it. But it IS neither a wave nor a particle. It is its own thing.

    When you get your head around the concept that photons are not waves or particles, but are something else that merely behaves like waves and particles, it becomes much easier to start thinking about how they really behave.

    And that makes it easier to comprehend what's really going on here (and also to get a handle on many other amazing attributes of photons, such as the fact that they do not experience time).

    Hope this helps.
  21. Nov 3, 2005 #20
    Dang, the forum parses multiple spaces into single spaces, screwing up the diagrams I put there.

    Ignore the dots, and they should look like this:


Share this great discussion with others via Reddit, Google+, Twitter, or Facebook