Double Slit Experiment: Explained in Detail

[Honorable_Death]

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?

 

[Cosmo16]

I'm definately not qualified to really answer. And quite simply from what I can tell, you get the basic idea.

The only thing that I can't 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.

 

[ZapperZ]

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?

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.

Zz.

 

[Honorable_Death]

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?

 

[ZapperZ]

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?

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.

Zz.

 

[derz]

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.

 

[Honorable_Death]

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.
Zz.

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 shouldn't interfere with itself. right?

 

[ZapperZ]

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 shouldn't interfere with itself. right?

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.

Zz.

 

[Honorable_Death]

all I am 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?

 

[ZapperZ]

all I am 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?

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.

Zz.

 

[RandallB]

shoot a photon at 2 slits, at a random angle to make it not interfere with itself?

HD
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, let's 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.

 

[caribou]

all I am 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?

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 separate ones.

 

[McQueen]

Here is a site http://www.people.virginia.edu/~xy9z/qubit/qubit.doc" 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.

 

[RandallB]

Here is a site ... that claims that a new ...

New??
From the bad ref you gave us (Not a two slit anyway):

a half-silvered mirror. This mirror will reflect light 50% of the time, and let light pass through it 50% of the time. As might be expected, both light detectors pick up light being transmitted from the source. Now, if a single photon, an indivisibly small unit of light were emitted, what would happen? Quantum theory states that both detectors would see the photon, which actually travels both paths at the same time.

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!

 

[RandallB]

The particle on its own will always interfere, however, and will travel through no definite slit.

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.

 

[caribou]

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.

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.

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 separate realities because nothing happened physically and effects like these without physically causes are somewhat headache-inducing.

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* occurring but it can be supressed and even erased by the interference from the possibilities which aren't occurring. 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.

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.

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.

 

[Honorable_Death]

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.
Zz.

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.

 

[RandallB]

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.

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.

 

[Hooloovoo]

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.

 

[Hooloovoo]

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:

_...__...___...____.._____.______._____..____...___...__..._


|...|...|...|...|..|.|||.|..|...|...|...|...|

 

[-Job-]

Why not both a wave & a particle? I can imagine a photon as being a packet of energy which is being dragged around inside an electromagnetic wave. It might be an electromagnetic loop wherein the packet of energy is displaced by the surrounding environment, and the displacement of the particle advances the wave. In this manner the packet of energy can only go through one slit, but the wave in which it propagates, being a wave, will go through both slits and interfere with itself. Because the packet of energy is under the influence of the wave, its trajectory is dependent on the interference of the wave with itself. Hence a single photon can interfere with itself.
Maybe just needs a bit of imagination.
I'm confident that something like this is happening, it fits.

 

[ZapperZ]

Why not both a wave & a particle? I can imagine a photon as being a packet of energy which is being dragged around inside an electromagnetic wave. It might be an electromagnetic loop wherein the packet of energy is displaced by the surrounding environment, and the displacement of the particle advances the wave. In this manner the packet of energy can only go through one slit, but the wave in which it propagates, being a wave, will go through both slits and interfere with itself. Because the packet of energy is under the influence of the wave, its trajectory is dependent on the interference of the wave with itself. Hence a single photon can interfere with itself.
Maybe just needs a bit of imagination.
I'm confident that something like this is happening, it fits.

The problem with this picture comes when you understand what it actually takes to make a "wave packet" in real space. Remember from classical wave theory that a wave packet is a superposition of a number of waves. The sharper the edge of the wave packet, such as trying to produce a square wave, the more wave harmonics that you have to sum up. This means that if you do a Fourier transform of such a wave packet, you get not one, not two, but SEVERAL different wavelengths.

But yet, a monochromatic light source has a single, highly dominant frequency. It is very difficult, close to impossible, to make a distinct wave packet out of such a scenario. So this is an observation that you need to account for when using such an argument.

Zz.

 

[-Job-]

I'm sorry that i don't know what you're talking about. Maybe some day. :) I haven't gotten that far, but i'll take your word for it. I'm going to make a flash movie of a possible alternative explanation. Sounds like a fun project.

 

[ZapperZ]

I'm sorry that i don't know what you're talking about. Maybe some day. :)

http://hyperphysics.phy-astr.gsu.edu/hbase/waves/wpack.html

Look at the mathematics. You have to either do a discrete sum, or an integral, to obtain a wave packet. And look at the last figure at the bottom. To make a wave packet, you have to combine a number of waves, each with different wavelengths.

So if you are saying that photons are wavepackets, then a monochromatic light source (a souce with only one predominant wavelength such as a laser) cannot exist.

Zz.

 

[-Job-]

I actually saw that link before, I'm going to study it. I hope you're not misunderstanding me though, because i used the words wave and packet, i wasn't talking directly about wave packets. I was just trying to provide an alternative explanation to the phenomena of the two slit experiment with the following model:
Suppose we have a tank of water with some little ball floating around in it. In the middle of the tank we have an obstruction with two slits. If we make a wave, this will displace the little ball. The ball itself goes either through one or the other slit. The wave that is carrying the ball however, goes through both slits and would show the common interference ripples. When the little ball crosses one of the slits it enters the zone of interference, after which its path is likely to change. If you had only one slit, this wouldn't happen, because there wouldn't be a zone of interference so the ball would just keep on it's path.
My wave packet thing was only to try to explain how it may be that a photon travels inside a wave, i thought up that idea in a different scenario and I'm happy to see that it works in here.

 

[ZapperZ]

I actually saw that link before, I'm going to study it. I hope you're not misunderstanding me though, because i used the words wave and packet, i wasn't talking directly about wave packets. I just trying to provide an alternative explanation to the phenomena of the two slit experiment with the following model:
Suppose we have a tank of water with some little ball floating around in it. In the middle of the tank we have an obstruction with two slits. If we make a wave, this will displace the little ball. The ball itself goes either through one or the other slit. The wave that is carrying the ball however, goes through both slits and would show the common interference ripples. When the little ball crosses one of the slits it enters the zone of interference, after which it's path is likely to change.

There is still the problem. What you DETECT are the photons. We do not detect your "wave envelope". In your water scenario, what is making the interference is the water wave, not the ball. Yet, what is detected in a light interference is the photon. There is no external wave.

The same can be said about electrons, protons, neutrons, buckyballs, etc. What you detect that is interfering are the object themselves. We have not detected this additional "water wave" equivalent. So what you are doing here is introducing a non-detected entity to explain an observation that has been sufficiently described by QM.

Zz.

 

[-Job-]

Couldn't the wave be an alternating field, magnetic-to electric? I'm using the wave to explain the distribution of the photons in the detector. In a way the wave is detected by the behavior of the photon. And i wouldn't think to question the Qm explanation if it didn't offer such a bizarre view, I'm not confortable in accepting it before i can consider at least a couple of others. Of course i don't have the means to make observations on this phenomenon, but i'd like to know if the model i suggested is feasible or not, even if it does not fit in QM, superficially it seems decent. What i'd really like to know is how light propagates in space.

 

[ZapperZ]

Couldn't the wave be an alternating field, magnetic-to electric? I'm using the wave to explain the distribution of the photons in the detector. In a way the wave is detected by the behavior of the photon. And i wouldn't think to question the Qm explanation if it didn't offer such a bizarre view, I'm not confortable in accepting it before i can consider at least a couple of others. Of course i don't have the means to make observations on this phenomenon, but i'd like to know if the model i suggested is feasible or not, even if it does not fit in QM, superficially it seems decent. What i'd really like to know is how light propagates in space.

If the wave is the EM field, then what is left for the photon? You have taken out the definition of the photon and given it to this "wave", and gotten back the classical EM wave that we know already. What is then left to your "ball"? How would you be able to explain the single-photon detection scenario? What is being "detected" if a photon doesn't contain the EM field?

Zz.

 

[-Job-]

The ball could be the energy of the electromagnetic wave. A photon has finite energy after all, how is energy distributed throughout an electromagnetic wave? It may be concentrated in a packet. This energy packet may be what reacts with matter, causing electrons to jump to other orbits.

 

[ZapperZ]

The ball could be the energy of the electromagnetic wave. A photon has finite energy after all, how is energy distributed throughout an electromagnetic wave? It may be concentrated in a packet. This energy packet may be what reacts with matter, causing electrons to jump to other orbits.

What "energy"? You do know that all the "energy" that is in a photon is in the from of the electromagnetic energy? What other forms of energy do you wish a photon to have? It is how we DEFINED a photon. But you have taken that out and gave it to an external "wave". So what is a photon now in your book?

Zz.

 

[-Job-]

Doesn't an electromagnetic wave propagate in 3 Dimensions (i.e. it's not a line but a surface)? If so how can a photon, whose energy is accounted primarily by the electromagnetic field, not interact with more than one particle at the same time in the same manner?

 

[ZapperZ]

Doesn't an electromagnetic wave propagate in 3 Dimensions (i.e. it's not a line but a surface)? If so how can a photon, whose energy is accounted primarily by the electromagnetic field, not interact with more than one particle at the same time in the same manner?

You do know, of course, that under conventional physics, EM wave = photons. They are not separate entity.

Now look at your question. I could also ask "How come EM waves don't interact with each other, since they are "...accounted primarily by the electromagnetic field"?"

And if all EM wave "propagates in 3D", then there's no point in solving any boundary conditions in the Poisson's equation for various geometry, is there?

I am trying VERY hard not to invoke any quantum field theory here. However, you need to be reasonable in your question because at some point, YOU have to do the work yourself in finding out the very basic E&M theory AND basic quantum theory.

Zz.

 

[RandallB]

"diffraction pattern" "interference pattern"

For a better understanding of the single slit diffraction pattern vs. the multi-slit interference patterns; see the link:
http://www.physics.northwestern.edu/ugrad/vpl/optics/diffraction.html"
Note that the interference pattern seen forms inside the shape of the single slit diffraction pattern.

I believe the old code here has a small error dealing with slit spacing, but it doesn’t hurt the look of the display. Actual slit spacing is half the number you put in.

 

[-Job-]

I know traditionally an EM wave = photon, i just wanted to use the wave as the propagation for a particle which would be what actually interacts with matter (ignoring any EM interaction), and calling that particle the photon instead. But i don't think the "water wave" can be explained using electromagnetism anyway, so whatever. Maybe it's space that's rippling.

 

[ZapperZ]

I know traditionally an EM wave = photon, i just wanted to use the wave as the propagation for a particle which would be what actually interacts with matter (ignoring any EM interaction), and calling that particle the photon instead. But i don't think the "water wave" can be explained using electromagnetism anyway, so whatever. Maybe it's space that's rippling.

Then maybe you can show me how a rippling space can somehow be blocked by a silly 2-slit set up. This implies that a slit isn't part of space and CAN block or hinder space. Last time I check with Dr. Einstein, space (and spacetime) can only be affected by gravity and/or extremely, extremely large density of energy, and not by some silly slit that can be made out of cardboard.

Zz.

 

[-Job-]

Ha, that's a good point, lucky for me i wasn't really serious about it.

 

[-Job-]

To be more serious, was the environment where this experiment was performed controlled with respect to electromagnetic emissions? I understand it must be hard to completely eliminate it as there should always be some background radiation but isn't it possible that this residual radiation affects the behavior of the photon? For example, with one slit, this background radiation flows from one side to the other without any interference, so the electromagnetic wave corresponding to the photon is free to take its path. With two slits the background radiation might generate some interference pattern which might affect the photon. Just considering some possibilities.
One question i have about the experiment is, does the photon always make it to the other side of the slits, or does it sometimes collide with the wall and not reach the detector? If it's the second scenario then that's just odd.

 

[huffypuffy]

all I am 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?

Yea, look at look at which gate each photon goes through and it won't interfere with itself