Delayed-choice double slit experiment

[Jamma]

Hi PF. I've finished reading Hawking's "The Grand Design", it was a great book. I'm not a taught physicist but I've always been interested in physics, and I study maths at university.

My question was relating to the section where Hawking describes the delayed-choice version of the double slit experiment, the only part of the book that I didn't really grasp. This is from the book:

"The fact that the past takes no definite form means that observations you make on a system in the present affect its past. That is underlined rather dramatically by a type of experiment thought up by John Wheeler, called a delayed-choice experiment. Schematically, a delayed-choice experiment is like the double-slit experiment we just described. in which you have the option of observing the path the particle takes, except in the delated-choice experiment you postpone your decision about whether or not to observe the path until just before the particle hits the detection screen.
Delayed-choice experiments result in data identical to those we get when we choose to observe (or not observe) the which-path information by watching the slits themselves. But in this case the path each particle takes- that is, its opast- is determined long after it poassed through the slits and presumably had to "decide" whether to travel through just one slit, which does not produce interference, or both slits, which does."

I am confused by his wording here.

How exactly do you postpone the choice of whether to measure the particle until it has just hit the screen? Surely if it has already been interfered by a photon at the slit, then whether a human decides to scrap the information or not is irrelevant. On the other hand, if he meant that you don't send a photon out to measure which slit it has gone through, then how exactly do you acquire this information if you do decide to measure it at the slit?

On wikipedia, it mentions removing the screen and using some sort of telescope to find the "which-path" information. Again, I don't understand how this version works. If the telescope is focused on the slit, then a photon is needed to intefere with the particle at the slit for the telescope to work doesn't it? Or is this some sort of theoretical telescope which doesn't need light to work, i.e. is this all just a thought experiment? And besides, we don't know that the interference pattern disappears now, because the screen was removed.

Sorry for my ignorance, I am sure that I am just interpreting this incorrectly. Thanks in advance for any help.

 

[DrChinese]

There are a variety of delayed choice experiments which lead to the same point: that the past does NOT appear to be etched in stone. It is not clear by what mechanism or ultimate rule this occurs. This is demonstrated to great effect in experiments with entangled particles. They can actually be entangled - and demonstrate violation of Bell Inequalities (which is an accepted standard for entanglement) - after they are detected!

Experimental Nonlocality Proof of Quantum Teleportation and Entanglement Swapping

"A seemingly paradoxical situation arises — as suggested by Peres [4] — when Alice’s Bellstate analysis is delayed long after Bob’s measurements. This seems paradoxical, because Alice’s measurement projects photons 0 and 3 into an entangled state after they have been measured. Nevertheless, quantum mechanics predicts the same correlations. Remarkably, Alice is even free to choose the kind of measurement she wants to perform on photons 1 and 2. Instead of a Bell-state measurement she could also measure the polarizations of these photons individually. Thus depending on Alice’s later measurement, Bob’s earlier results either indicate that photons 0 and 3 were entangled or photons 0 and 1 and photons 2 and 3. This means that the physical interpretation of his results depends on Alice’s later decision.

"Such a delayed-choice experiment was performed by including two 10 m optical fiber
delays for both outputs of the BSA. In this case photons 1 and 2 hit the detectors delayed by about 50 ns. As shown in Fig. 3, the observed fidelity of the entanglement of photon 0 and photon 3 matches the fidelity in the non-delayed case within experimental errors. Therefore, this result indicate that the time ordering of the detection events has no influence on the results and strengthens the argument of A. Peres [4]: this paradox does not arise if the correctness of quantum mechanics is firmly believed."

Specifically, in the experiment: The decision to entangle 2 photons can be made after they are already detected. (This technique does not allow one to signal from the future to the past however or anything similar - all results are random.)

 

[etamorphmagus]

The entanglement experiments bring a lot of confusion, for beginners such as ourselves. But I might add something to the wheeler delayed choice experiment.

So here is the problem and how to get it right (I hope):
In the experiment you put telescopes pointing at the slits—watching what path the particle will take—and after the particle goes through, you quickly add the screen, to detect the wave distribution, the interference pattern, thus bringing the so called delayed-choice-paradox. (or maybe you leave the screen, and then quickly pull it our, doesn't matter)
Thing is, the photon, electron, or any particle at all are not classical-wave-or-particle, but are quantum-mechanical, with quantum-mechanical behavior. In this behavior, you can only get some information about it by measuring, but the problem is that with measurement comes the uncertainty principle, and you can only get one classical result, either you see interference, or you see location. so it is not as if the particle changes its mind about what behavior it has, it's just that you measure a different property:

Telescopes: you measure particle behavior, what path it took
Screen: you measure wave behavior via particle detection, after many particles you see wave-distribution - interference pattern.​

This measurement is also called collapse of the wave-function to a single state, or decoherence from superposition to single state.


I hope I got it right.

 

[Jamma]

Argh, sorry, I still don't get it.

How do the telescopes work, with photons right? Am I right in saying that the delayed choice is in whether or not we decide to measure the photon (i.e. by removing the screen, or not)? Is it not the photon interfering the electron but infact, the measurement of the photon by the telescope which causes the change?

If not, I don't understand where the change occurs, because in both cases, the photon needed to interfere with the electron to give the possibility of it being measured later by the telescope.

I think I can see why this can't send messages back to the past though, the changes of the electron need to actual be measured in the past for them to give the information, and this would not have allowed the manipulation that we see in this experiment later on.

 

[etamorphmagus]

You are correct, the delayed choice is either to measure the "where they came from" or not, after they apparently ALREADY did the wave thingy, catching nature with her pants does. That is the "paradox" but of course, quantum-mechanically it is all consistent. It is only a paradox if you restrict yourself to classical bahavior (how can it be both wave and particle?).

There are no electrons in "[URL delayed choice experiment[/URL].

I don't really understand the rest of your entry, please rephrase.

 

[Jamma]

Ok, thanks, that is quite useful, I think that I understand.

Apologies for mentioning the electron, by that, I meant the particle in question, it's just the way it was phrased in one of the explanations that I have read.

I'm still not 100% on this though; whether or not you measure the particle with the telescope, the particle must have been interfered with by the photon at the slit for it to be later observed by the photon, must it not? Therefore, why does it matter whether or not we observe the photon with the telescope or not, the photon will have still interfered at the slit? Or is it that there would not be enough information to deduce which slit that the particle passed through if the photon doesn't make it to the telescope? It seems that the experiment is the same to the variant where we just measure at the slit but now the device used to measure it is further away, pointing in a different direction; how does the particle know that?

Basically, am I right in saying that it is the measurement of the photon which causes the difference, and not the interference of the photon with the particle? I find this difficult to believe, what criterion is the particle using to deduce that the photon is being measured if this is indeed the correct explanation (which therefore leads me to believe that this is the incorrect explanation).

 

[Jamma]

Oh, and to clear up the confusion here, by photon, I am always referring to something that is measuring the particle which is being fired through the slits, and by particle, I mean the particle in question.

I can see why this could cause confusion, because on the wiki page that you have provided, Wheeler fires a photon through the slits.

 

[Jamma]

I think that I see the confusion.

I was assuming that we needed to fire a photon at the particle for it to be measured by the telescope.

But on the wiki page, no interactions such as this occur, because we are observing the photon directly. But then how does this determine the slit that is being used? Is it not possible for the photon to have came from one slit and be measured by the wrong telescope? I mean, in the interference pattern, the pattern is caused by the fact that there is a probability that the particle can interfere with itself on one side of the screen by approaching it from the other slit, right?

 

[etamorphmagus]

Ok, thanks, that is quite useful, I think that I understand.

Apologies for mentioning the electron, by that, I meant the particle in question, it's just the way it was phrased in one of the explanations that I have read.

I'm still not 100% on this though; whether or not you measure the particle with the telescope, the particle must have been interfered with by the photon at the slit for it to be later observed by the photon, must it not? Therefore, why does it matter whether or not we observe the photon with the telescope or not, the photon will have still interfered at the slit? Or is it that there would not be enough information to deduce which slit that the particle passed through if the photon doesn't make it to the telescope? It seems that the experiment is the same to the variant where we just measure at the slit but now the device used to measure it is further away, pointing in a different direction; how does the particle know that?

Exactly, it is measuring the which path information, by measuring it disturbed the momentum of the photos (uncertainty principle), therefore its wavelength, and therefore all the photons that would create interference pattern together are no longer coherent and you see no interference.

The device doesn't need to know anything, the point of the experiment is to show that photons are quantum-mechanical and not only particles or only waves. You choose the classical aspect you will see, by measuring classical particle (which path info), or measuring classical wave (interference).

Basically, am I right in saying that it is the measurement of the photon which causes the difference, and not the interference of the photon with the particle? I find this difficult to believe, what criterion is the particle using to deduce that the photon is being measured if this is indeed the correct explanation (which therefore leads me to believe that this is the incorrect explanation).

Yeah (for the bold part).
The uncertainty principle is a statement that governs the way quantum-mechanical behavior exhibits itself to us in the form of classical behavior (wave\particle). You should read more about it http://hyperphysics.phy-astr.gsu.edu/hbase/uncer.html#c2".

think that I see the confusion.

I was assuming that we needed to fire a photon at the particle for it to be measured by the telescope.

But on the wiki page, no interactions such as this occur, because we are observing the photon directly. But then how does this determine the slit that is being used? Is it not possible for the photon to have came from one slit and be measured by the wrong telescope? I mean, in the interference pattern, the pattern is caused by the fact that there is a probability that the particle can interfere with itself on one side of the screen by approaching it from the other slit, righ

I also thought this was weird, that the photon make go to any of the two telescopes and not provide you with which-path info, but the Wheeler wiki page says:

Sufficiently far beyond the region of the plate, the beams from upper and lower slits cease to overlap and become well separated. There place photodetectors. Let each have an opening such that it records with essentially 100 percent probability a quantum of energy arriving in its own beam, and with essentially zero probability a quantum arriving in the other beam

So there's no probability of photon from slit A to reach telescope B. It doesn't really matter how it works, it works (Don't ask me for the math, I'm new at this too) and supplies which-path info.

 

[eaglelake]

Hi PF. I've finished reading Hawking's "The Grand Design", it was a great book. I'm not a taught physicist but I've always been interested in physics, and I study maths at university.

My question was relating to the section where Hawking describes the delayed-choice version of the double slit experiment, the only part of the book that I didn't really grasp. This is from the book:

"The fact that the past takes no definite form means that observations you make on a system in the present affect its past. That is underlined rather dramatically by a type of experiment thought up by John Wheeler, called a delayed-choice experiment. Schematically, a delayed-choice experiment is like the double-slit experiment we just described. in which you have the option of observing the path the particle takes, except in the delated-choice experiment you postpone your decision about whether or not to observe the path until just before the particle hits the detection screen.
Delayed-choice experiments result in data identical to those we get when we choose to observe (or not observe) the which-path information by watching the slits themselves. But in this case the path each particle takes- that is, its opast- is determined long after it poassed through the slits and presumably had to "decide" whether to travel through just one slit, which does not produce interference, or both slits, which does."

I am confused by his wording here.

How exactly do you postpone the choice of whether to measure the particle until it has just hit the screen? Surely if it has already been interfered by a photon at the slit, then whether a human decides to scrap the information or not is irrelevant. On the other hand, if he meant that you don't send a photon out to measure which slit it has gone through, then how exactly do you acquire this information if you do decide to measure it at the slit?

On wikipedia, it mentions removing the screen and using some sort of telescope to find the "which-path" information. Again, I don't understand how this version works. If the telescope is focused on the slit, then a photon is needed to intefere with the particle at the slit for the telescope to work doesn't it? Or is this some sort of theoretical telescope which doesn't need light to work, i.e. is this all just a thought experiment? And besides, we don't know that the interference pattern disappears now, because the screen was removed.

Sorry for my ignorance, I am sure that I am just interpreting this incorrectly. Thanks in advance for any help.

We must first understand how quantum experiments differ from classical physics. The ideas given below are from Bohr and are embodied in his Principle of Complementarity.
1. Every experiment must have an experimental result that gives it closure. There is no experiment until there is an experimental result. You may remember Wheeler's famous statement, "No elementary phenomenon is a phenomenon until it is a registered phenomenon." When the particle triggers the detector the experiment is over and done.
2. (Non-Separability Principle) The entire experimental configuration, including the particle source, the particle, the preparation apparatus, the measurement device, and the measurement result must be treated as a single entity. It is indivisible and it cannot be broken into its separate parts. Nor is it a sequence of physical events. The results obtained depend on the whole experimental arrangement.
3. (Delayed Choice) If we change the apparatus at any time during the experiment, the results obtained correspond to the experimental configuration in place at the moment the experiment is closed, i.e. the instant when the particle is detected. We can wait until the last possible moment to make changes. It doesn't matter. Only the final experimental configuration matters. Delayed choice experiments confirm such behavior. This is actually a corollary of the first two statements.
We are forced to give up some of the most cherished tenets of classical physics. The particle is no longer an independent object that travels from the source through the apparatus and into the measuring device. It does not have a trajectory. Nor does it have any pre-existing physical quantities prior to measurement. We don't know anything until we have a result. If we insist that the particle still behaves classically, then paradoxes and contradictions follow.
Particles cannot decide anything. We have no idea what the particle is doing before it is detected in the delayed choice experiment. A quantum experiment has no past. It only has a measured result corresponding to the experimental configuration in place at the instant the particle is detected.
You delay making a choice by using classical physics. We tacitly do a time-of-flight calculation and then make our final choice at a time after the particle should have passed through the slit(s) but before it would hit the screen. Of course, this requires a trajectory, so the whole exercise is questionable. Do we really know where the particle is when we make our choice?? I doubt it.
You state, "Surely if it has already been interfered by a photon at the slit---------", which is speculation that cannot be verified experimentally. We have no idea what is going on before we obtain the measured result. All we know is that if we choose the which-way experiment, there is no interference. If we choose the set up for which there is no way to determine which slit the particle went through, then there is interference. Many of us refuse to accept such a meager description. But, apparently, that’s all there is!
With regard to using a telescope to to get which-way information------- If the telescope only 'sees' slit A, then if you replace the telescope with a detector, then only particles from slit A are detected, and there is no interference.
Sorry to be so harsh, but you are asking for a classical explanation of the double-slit experiment. There isn't any!
Best wishes.

 

[Jamma]

I understand what you are saying, and I wasn't looking for a classical explanation, in a sense.

3. (Delayed Choice) If we change the apparatus at any time during the experiment, the results obtained correspond to the experimental configuration in place at the moment the experiment is closed, i.e. the instant when the particle is detected. We can wait until the last possible moment to make changes. It doesn't matter. Only the final experimental configuration matters. Delayed choice experiments confirm such behavior. This is actually a corollary of the first two statements.

I think that this needs to be more thorough. I mean, we could change the apparatus to be a brick wall instead of slits after "the particle had passed the slits" (I appreciate that such statements are sketchy as, since you mentioned, we don't know where the particle is). The results wouldn't be based on this final configuration. Unless you mean that the changes to the experiment are logged, and this is part of the configuration?

My confusion wasn't really that I am stuck in a classical mindset, it's just as to how the telescopes measure which slit the particle passes through when we choose to measure it. The first point of confusion was that I was thinking that they work like the sensors of the variant where we measure at the slit; by sending out a photon. In fact, we measure the particle directly by pointing the telescope at the slit and allowing the particle to travel into the telescope. What I don't understand is, how does this determine that the particle definitely came from the claimed slit; can't the particle have come from the opposite slit? Or is it that the chances are so low that we may as well assume with near 100% certainty which it came from, and that this destroys the interference.

Also, how do we know that the interference pattern is no longer apparent, since we removed the screen, and we now only know a which-path information (I'm guessing that we don't know that the pattern dissapears, just that, as remarked before, we now know the "which-path" i.e. particle like behaviour, but no longer see the interference pattern i.e. the wave-like behaviour).

 

[eaglelake]

I understand what you are saying, and I wasn't looking for a classical explanation, in a sense.



I think that this needs to be more thorough. I mean, we could change the apparatus to be a brick wall instead of slits after "the particle had passed the slits" (I appreciate that such statements are sketchy as, since you mentioned, we don't know where the particle is). The results wouldn't be based on this final configuration. Unless you mean that the changes to the experiment are logged, and this is part of the configuration?

My confusion wasn't really that I am stuck in a classical mindset, it's just as to how the telescopes measure which slit the particle passes through when we choose to measure it. The first point of confusion was that I was thinking that they work like the sensors of the variant where we measure at the slit; by sending out a photon. In fact, we measure the particle directly by pointing the telescope at the slit and allowing the particle to travel into the telescope. What I don't understand is, how does this determine that the particle definitely came from the claimed slit; can't the particle have come from the opposite slit? Or is it that the chances are so low that we may as well assume with near 100% certainty which it came from, and that this destroys the interference.

Also, how do we know that the interference pattern is no longer apparent, since we removed the screen, and we now only know a which-path information (I'm guessing that we don't know that the pattern dissapears, just that, as remarked before, we now know the "which-path" i.e. particle like behaviour, but no longer see the interference pattern i.e. the wave-like behaviour).

If we put a brick wall in place of the slits then there would be no result. No result means the experiment is never finished?? There is no final configuration without a result. Or, does the brick wall become the detector when the particle hits it, so that we do have a result?? This is a very perceptive question and I wish I had an answer for you. Such an experiment could be done with a shutter across the slits, but I don't know if anyone has actually done it. Bohr's three statements tell us how quantum experiments "behave" in a non-classical way. As far as I know, there is no deeper "understanding".
Lets try to simplify the experiment by eliminating the telescope and the human observer. In particular, it is not necessary to send out a photon to interact with the particle. We can obtain which-way information without a human observer ever knowing which slit the particle actually went through. For example, place polarizers across each slit so that particles going through slit A are x-polarized and particles going through slit B are y-polarized. Now its polarization determines which-way a particle went, even though no one has measured the particle's polarization and no one actually knows which slit the particle did go through. With both slits open and no telescopes looking at the slits, we wait till the last possible moment to insert the polarizers. Now, the which-way information is contained in the polarization, and there is no interference.
The general rule is this: If there is no way, in principle, to determine which slit the particle went through, then there is interference. If there is a way to determine which slit the particle went through, then there is no interference.
If we use a detector instead of a screen then we must measure the angular distribution of many particles by moving the detector and counting particles at all possible angles. If you have which path information then there is no interference. You are correct that detecting a single particle doesn't give the interference pattern, if any.
Best wishes.

 

[etamorphmagus]

. We can obtain which-way information without a human observer ever knowing which slit the particle actually went through. For example, place polarizers across each slit so that particles going through slit A are x-polarized and particles going through slit B are y-polarized. Now its polarization determines which-way a particle went, even though no one has measured the particle's polarization and no one actually knows which slit the particle did go through. With both slits open and no telescopes looking at the slits, we wait till the last possible moment to insert the polarizers. Now, the which-way information is contained in the polarization, and there is no interference.

Wait, if you place the polarizers on the slits after the quantum passed the slits, the intereference is destroyed? Doesn't make sense, you have no information on that quantum after it passed the slits, so adding polarizers changes nothing.

I don't know much about polarization, but it does disturb the photons, so say you have a beam of photons going through the slits with different polarizers, the beams would lose coherence + you could in principle get the which-path information and the interference would be destroyed.
How does it disturb the photons? momentum? phase?

 

[JJRittenhouse]

I don't know much about polarization, but it does disturb the photons

As far as I know, polarization is more like a slit which insures that only photons with wavelengths parallel to the polarization can get through. This is more like a gate than disturbance. It limits what kind of photons gets through just as a slit in the double slit experiment would. If your polarizations and different over each slit, you only need to measure the polarization at the result of the experiment to determine which slit the photon went through.

Actually, come to think of it, if you used polarization, then you would have results which would not reach your screen if they were blocked by the polarization. So by limiting the kinds of photons that get through the experiment, that would change the results as only photons matching your two polarization choices could be measured.

 

[JJRittenhouse]

If you placed glass over the slits in the experiment, so that you'd have photons interacting with the electrons in the glass, would that be the same as observing the path of the photon, even if no data was collected?

 

[etamorphmagus]

If you placed glass over the slits in the experiment, so that you'd have photons interacting with the electrons in the glass, would that be the same as observing the path of the photon, even if no data was collected?

It's not about data it's about physical interaction that is strong enough to decohere the photon from the rest of the photons that would "follow" it after.
Each photon's momentum would be disturbed in a random way, and momentum equals wavelength, so overall you would have lots of different wavelengths (lose coherence) that won't create the interference pattern but a classical average line.


So about putting 2 polarizers with different angles, I understand, not mathematically, that the quantum-mechanical photon would not interfere with itself the way normal slits would cause it too, since it is "forcing" it to choose orientation and in essence decohering it, collapsing it, into a single state, so you'll get a classical curve. But why should it happen if you "delay" your placing of the polarizers after the photon passes. I don't think it will.

 

[Jamma]

If we put a brick wall in place of the slits then there would be no result. No result means the experiment is never finished?? There is no final configuration without a result. Or, does the brick wall become the detector when the particle hits it, so that we do have a result?? This is a very perceptive question and I wish I had an answer for you. Such an experiment could be done with a shutter across the slits, but I don't know if anyone has actually done it. Bohr's three statements tell us how quantum experiments "behave" in a non-classical way. As far as I know, there is no deeper "understanding".
Lets try to simplify the experiment by eliminating the telescope and the human observer. In particular, it is not necessary to send out a photon to interact with the particle. We can obtain which-way information without a human observer ever knowing which slit the particle actually went through. For example, place polarizers across each slit so that particles going through slit A are x-polarized and particles going through slit B are y-polarized. Now its polarization determines which-way a particle went, even though no one has measured the particle's polarization and no one actually knows which slit the particle did go through. With both slits open and no telescopes looking at the slits, we wait till the last possible moment to insert the polarizers. Now, the which-way information is contained in the polarization, and there is no interference.
The general rule is this: If there is no way, in principle, to determine which slit the particle went through, then there is interference. If there is a way to determine which slit the particle went through, then there is no interference.
If we use a detector instead of a screen then we must measure the angular distribution of many particles by moving the detector and counting particles at all possible angles. If you have which path information then there is no interference. You are correct that detecting a single particle doesn't give the interference pattern, if any.
Best wishes.

My point with the brick wall was indeed that we would see no result (or, a very different result, i.e. no electron observed at the screen). I was just trying to suggest that your explanation should be more thorough; you suggested that all that matters is the final configuration, but clearly it also depends on the changes you make to the apparatus (i.e. if we changed to the brick wall, but after the particle had passed by, we see that it is more than just the final configuration that is important; I think that I just misinterpreted what you were saying).

I indeed see why it is odd that we should see interference when it is polarised because it should technically only be allowed through one slit, yet it can pass through both and interfere with itself if we have not yet obtained information about the polarisation of the particle. The question is, can we polarise it without being able to gain information about what polarisation it has to begin with?

 

[eaglelake]

Wait, if you place the polarizers on the slits after the quantum passed the slits, the intereference is destroyed? Doesn't make sense, you have no information on that quantum after it passed the slits, so adding polarizers changes nothing.

I don't know much about polarization, but it does disturb the photons, so say you have a beam of photons going through the slits with different polarizers, the beams would lose coherence + you could in principle get the which-path information and the interference would be destroyed.
How does it disturb the photons? momentum? phase?

Yes, there is no interference if the polarizers are in place at the instant the particle is detected. Even if we wait until the particle has already passed through the slits, there is still no interference. You can wait until the last possible moment and the human observer is not required to know which way the photon went. You say this doesn't make sense, but what you mean is it doesn't make classical sense. Remember that quantum events are not classical. Adding polarizers does change the quantum experiment along with the wave function and the experimental results. Such delayed choice experiments with polarizers have been done. Photons from slit A are now different from the photons coming through slit B. You are correct - they are no longer coherent, which is necessary for interference. There is a classical explanation for this. In classical wave optics, it is known that only parallel components of the electric field contribute to the interference. x-polarizers only pass the x-component of the electric field, y-polarizers only pass the y-component of the electric field. Now the electric fields from the two slit are not parallel and, hence, there is no interference. I do not know how the photon is disturbed, if at all, by passing through the polarizer.

 

[etamorphmagus]

Yes, there is no interference if the polarizers are in place at the instant the particle is detected. Even if we wait until the particle has already passed through the slits, there is still no interference. You can wait until the last possible moment and the human observer is not required to know which way the photon went. You say this doesn't make sense, but what you mean is it doesn't make classical sense. Remember that quantum events are not classical. Adding polarizers does change the quantum experiment along with the wave function and the experimental results. Such delayed choice experiments with polarizers have been done. Photons from slit A are now different from the photons coming through slit B. You are correct - they are no longer coherent, which is necessary for interference. There is a classical explanation for this. In classical wave optics, it is known that only parallel components of the electric field contribute to the interference. x-polarizers only pass the x-component of the electric field, y-polarizers only pass the y-component of the electric field. Now the electric fields from the two slit are not parallel and, hence, there is no interference. I do not know how the photon is disturbed, if at all, by passing through the polarizer.

So if the photon source is 1 meter from the slits, and the detector is a light-hour away — if you wait 50 minutes since release of photon, and add polarizers, you kill the interference by messing with the photon's wave function? Even quantum mechanically this kind of doesn't make sense, because you DON'T have the inforamtion on the photon - you KNOW you put the polarizers after it passes. Unless it is as if you snag on the photon's "tail".
If this is really true then this is a type of spooky-action-at-a-distance and classical-retrocausality isn't it?

 

[zonde]

Yes, there is no interference if the polarizers are in place at the instant the particle is detected. Even if we wait until the particle has already passed through the slits, there is still no interference. You can wait until the last possible moment and the human observer is not required to know which way the photon went. You say this doesn't make sense, but what you mean is it doesn't make classical sense. Remember that quantum events are not classical. Adding polarizers does change the quantum experiment along with the wave function and the experimental results. Such delayed choice experiments with polarizers have been done. Photons from slit A are now different from the photons coming through slit B. You are correct - they are no longer coherent, which is necessary for interference. There is a classical explanation for this. In classical wave optics, it is known that only parallel components of the electric field contribute to the interference. x-polarizers only pass the x-component of the electric field, y-polarizers only pass the y-component of the electric field. Now the electric fields from the two slit are not parallel and, hence, there is no interference. I do not know how the photon is disturbed, if at all, by passing through the polarizer.

No, this is nonsense. If polarizer is added after photon have passed that place it will not affect result.
If you have some particular experiment in mind then tell what it is.

Apart from other things this makes FTL communication possible.

 

[zonde]

I don't know much about polarization, but it does disturb the photons, so say you have a beam of photons going through the slits with different polarizers, the beams would lose coherence + you could in principle get the which-path information and the interference would be destroyed.
How does it disturb the photons? momentum? phase?

In order for interference to be observable two beams have to have correlated phase. When you add polarizers you destroy this phase correlation between two beams.

 

[DrChinese]

So if the photon source is 1 meter from the slits, and the detector is a light-hour away — if you wait 50 minutes since release of photon, and add polarizers, you kill the interference by messing with the photon's wave function? Even quantum mechanically this kind of doesn't make sense, because you DON'T have the inforamtion on the photon - you KNOW you put the polarizers after it passes. Unless it is as if you snag on the photon's "tail".
If this is really true then this is a type of spooky-action-at-a-distance and classical-retrocausality isn't it?

As zonde says, this is not how it works. These delayed choice double slit experiments are notoriously complicated to discuss. So don't be surprised if you are talking about one thing and another person is talking about something else.

The entire context must be considered. Clearly, the relevant context for the slits is around the time that the photon is passing by. Before or after doesn't really make a difference. For the screen, the relevant context is around the time of detection. And in between, etc.

Polarizers can be placed over the slits. As the relative angle of those polarizers is varied, the interference will appear or disappear. Again, this is considering the entire context as there may be other factors to consider depending on the setup. So generally, you see no interference when the polarizers are crossed and you do see interference when they are parallel. That is because in one setup, you get which slit information and the other you don't.

 

[Jamma]

As zonde says, this is not how it works. These delayed choice double slit experiments are notoriously complicated to discuss. So don't be surprised if you are talking about one thing and another person is talking about something else.

The entire context must be considered. Clearly, the relevant context for the slits is around the time that the photon is passing by. Before or after doesn't really make a difference. For the screen, the relevant context is around the time of detection. And in between, etc.

Polarizers can be placed over the slits. As the relative angle of those polarizers is varied, the interference will appear or disappear. Again, this is considering the entire context as there may be other factors to consider depending on the setup. So generally, you see no interference when the polarizers are crossed and you do see interference when they are parallel. That is because in one setup, you get which slit information and the other you don't.

Yes, this makes sense, I thought something seemed a bit fishy.

Because I am new to this, can someone just tell me if my way of thinking of things is correct.

How I imagine everything is the particle as being a probability distribution, a sort of cloud in three dimensional space, but also equipped with a phase (actually, thinking about it, the phase is needed to work out this probability distribution I think). To work the probabilities out of where the particle is, we compute all different paths to the point and take into account the phase; therefore, we can get interference, as in the double slit experiment.

However, when we try to observe this "cloud" we destroy parts of it as being observed, or possibly reveal the location of the particle (but only to a limited accuracy; the uncertainty principle). When this happens, we sort of reset things. So for example, we can perform the double slit experiment and see interference. But when we observe the particle past the slit, we pin down it's location, and then we don't see the interference, because the point that it hits the screen is now essentially coming from one source i.e. the point where we found the particle (of course, where we observe the particle after the slit has it's own probability distribution which should be an interference pattern, but observing it early will smudge everything in a sense by the time it hits the screen).

For example; if we observe the particle just before the screen, do we still see some interference? I would think so since the probability distributions for finding the particle just before the screen should have an interference pattern (just as if we moved the screen closer) and from here, the particles have almost hit the screen anyway, so are likely to hit it in a pretty straight line from their trajectory.I'm sure that I've abused lots of terms and laws here, but can someone experienced with this please tell me what I have right/wrong? (if any of it!).

 

[DrChinese]

Yes, this makes sense, I thought something seemed a bit fishy.

Because I am new to this, can someone just tell me if my way of thinking of things is correct.

How I imagine everything is the particle as being a probability distribution, a sort of cloud in three dimensional space, but also equipped with a phase (actually, thinking about it, the phase is needed to work out this probability distribution I think). To work the probabilities out of where the particle is, we compute all different paths to the point and take into account the phase; therefore, we can get interference, as in the double slit experiment.

However, when we try to observe this "cloud" we destroy parts of it as being observed, or possibly reveal the location of the particle (but only to a limited accuracy; the uncertainty principle). When this happens, we sort of reset things. So for example, we can perform the double slit experiment and see interference. But when we observe the particle past the slit, we pin down it's location, and then we don't see the interference, because the point that it hits the screen is now essentially coming from one source i.e. the point where we found the particle (of course, where we observe the particle after the slit has it's own probability distribution which should be an interference pattern, but observing it early will smudge everything in a sense by the time it hits the screen).

For example; if we observe the particle just before the screen, do we still see some interference? I would think so since the probability distributions for finding the particle just before the screen should have an interference pattern (just as if we moved the screen closer) and from here, the particles have almost hit the screen anyway, so are likely to hit it in a pretty straight line from their trajectory.


I'm sure that I've abused lots of terms and laws here, but can someone experienced with this please tell me what I have right/wrong? (if any of it!).

Probably as good a way as any to think of it.

One thing that is unclear to anyone: is collapse a physical process? It may or may not be. The process of erasing a measurement implies it is not (since collapse is supposed to be an irreversible process). But all of this implies an understanding that involves a degree of speculation. So my point is that your image will be fuzzy in this particular but so is everyone's.

 

[etamorphmagus]

In order for interference to be observable two beams have to have correlated phase. When you add polarizers you destroy this phase correlation between two beams.

What is phase correlation? Do you mean phase difference on the sources? That must remain constant for the sources to be coherent.

As zonde says, this is not how it works. These delayed choice double slit experiments are notoriously complicated to discuss. So don't be surprised if you are talking about one thing and another person is talking about something else.

The entire context must be considered. Clearly, the relevant context for the slits is around the time that the photon is passing by. Before or after doesn't really make a difference. For the screen, the relevant context is around the time of detection. And in between, etc.

Polarizers can be placed over the slits. As the relative angle of those polarizers is varied, the interference will appear or disappear. Again, this is considering the entire context as there may be other factors to consider depending on the setup. So generally, you see no interference when the polarizers are crossed and you do see interference when they are parallel. That is because in one setup, you get which slit information and the other you don't.

Alright, I accept the complexity. But what happens if you place polarizers just before detection? Does it greatly depend on the configuration? I think the answer is a plain straight yes or no.

 

[DrChinese]

Alright, I accept the complexity. But what happens if you place polarizers just before detection? Does it greatly depend on the configuration? I think the answer is a plain straight yes or no.

As a general rule, placing polarizers in front of the detection screen does not yield additional information about which slit. So it does not change the interference pattern, only the intensity of the pattern.

This may seem counterintuitive if there are polarizers in front of the slits.

 

[etamorphmagus]

As a general rule, placing polarizers in front of the detection screen does not yield additional information about which slit. So it does not change the interference pattern, only the intensity of the pattern.

This may seem counterintuitive if there are polarizers in front of the slits.

Why unintuitive? polarizers filter light, don't they?

 

[DrChinese]

Why unintuitive? polarizers filter light, don't they?

Yes. Polarizers by the slits may supply additional information about which slit information. But one in front of the detection screen does not.

 

[etamorphmagus]

Yes. Polarizers by the slits may supply additional information about which slit information. But one in front of the detection screen does not.

Ohhhh you meant right close to the detector. I was thinking about placing them at the slits after the light already passed.

So like Jamma said:

For example; if we observe the particle just before the screen, do we still see some interference? I would think so since the probability distributions for finding the particle just before the screen should have an interference pattern (just as if we moved the screen closer) and from here, the particles have almost hit the screen anyway, so are likely to hit it in a pretty straight line from their trajectory.

At that point it doesn't really disturb the interference, some photons don't have to go in there anyways + you don't get which-path information.

 

[eaglelake]

No, this is nonsense. If polarizer is added after photon have passed that place it will not affect result.
If you have some particular experiment in mind then tell what it is.

Apart from other things this makes FTL communication possible.

It would be nonsense if this was a classical experiment, but it is not.
Most delayed choice experiments are done with a Mark-Zehnder interferometer and beam splitters replacing the slits. A short readable article that addresses your concerns is Hillmer and Kwiat, Scientific American, April 16, 2007. They cite Jacque, et al, Science 315, 966 (2000).
"Quantum Theory and Measurement" by Wheeler and Zureck has comments by Bohr and Wheeler about delayed choice. Although Wheeler first used the term "delayed choice", Bohr knew about this before any one.
An actual delayed choice experiment with slits and polarizers is Walborn, et al, Phys Rev A 033818 (2002). A Google search will yield many other articles on actual experiments.
There is no faster than light communication here. The particles, i.e. the things moving through space-time, always travel at speeds less than the speed of light.
The point is this: Bohr said we could wait until the very last minute to make a change. He was right! Modern delayed choice experiments confirm this.
I repeat for emphasis: If we change the apparatus at any time during the experiment, the results obtained correspond to the experimental configuration in place at the moment the experiment is closed, i.e. the instant when the particle is detected. We can wait until the last possible moment to make changes. It doesn't matter. Only the final experimental configuration matters.
Best wishes

 

[etamorphmagus]

And of course, in the entangled particles experiments - the delayed quantum erasers, you could wait well after one was "detected", according to wiki. Is this so, in the "delayed choice quantum eraser"? I don't understand it fully, but it's something like that right?

 

[Jamma]

I repeat for emphasis: If we change the apparatus at any time during the experiment, the results obtained correspond to the experimental configuration in place at the moment the experiment is closed, i.e. the instant when the particle is detected. We can wait until the last possible moment to make changes. It doesn't matter. Only the final experimental configuration matters.
Best wishes

I'm sorry, I'm either misinterpretting you, or this just can't be right. So if you fired a particle through a slit 1m from the source, and the detector was 10 light years away, you could technically just block off the slit just before the particle "hits" the detector screen, and it will appear back at the wall because in that configuration, particles wouldn't be able to get through? This is obviously ridiculous.

Also, just because the particles travel at less than the SoL, that doesn't mean that you can't come up with some way of doing FTL communication. For example; you could fire the particles through at a constant rate and agree to measure (or not measure) a particle at each time. The sender could put inplace a polariser, or brick wall even, according to what you are saying, and this would effect the particle which was just about to hit the screen. This is clearly FTL communication if we do things on a large enough scale.

 

[DrChinese]

An actual delayed choice experiment with slits and polarizers is Walborn, et al, Phys Rev A 033818 (2002).

Here is the link:

http://arxiv.org/abs/quant-ph/0106078

"We report a quantum eraser experiment which actually uses a Young double-slit to create interference. The experiment can be considered an optical analogy of an experiment proposed by Scully, Englert and Walther. One photon of an entangled pair is incident on a Young double-slit of appropriate dimensions to create an interference pattern in a distant detection region. Quarter-wave plates, oriented so that their fast axes are orthogonal, are placed in front of each slit to serve as which-path markers. The quarter-wave plates mark the polarization of the interfering photon and thus destroy the interference pattern. To recover interference, we measure the polarization of the other entangled photon. In addition, we perform the experiment under delayed erasure circumstances. "


On the other hand, this is not the same thing as saying what happens at the slits matters much AFTER the photon has passed by. Because that is not part of the context, it doesn't matter.

 

[DrChinese]

I'm sorry, I'm either misinterpretting you, or this just can't be right. So if you fired a particle through a slit 1m from the source, and the detector was 10 light years away, you could technically just block off the slit just before the particle "hits" the detector screen, and it will appear back at the wall because in that configuration, particles wouldn't be able to get through?

I think there are different versions of the time element being discussed. There is not a single point in time in which context is relevant. There are many different points in space time to consider. After a photon has passed the double slit, what happens at the slit doesn't change the outcome. But the words before and after are somewhat misleading, as a context 10 light years away CAN be relevant to something "here and now". It depends.

 

[etamorphmagus]

I think there are different versions of the time element being discussed. There is not a single point in time in which context is relevant. There are many different points in space time to consider.

Which points?

After a photon has passed the double slit, what happens at the slit doesn't change the outcome. But the words before and after are somewhat misleading, as a context 10 light years away CAN be relevant to something "here and now". It depends.

To what?

This?

"We report a quantum eraser experiment which actually uses a Young double-slit to create interference. The experiment can be considered an optical analogy of an experiment proposed by Scully, Englert and Walther. One photon of an entangled pair is incident on a Young double-slit of appropriate dimensions to create an interference pattern in a distant detection region. Quarter-wave plates, oriented so that their fast axes are orthogonal, are placed in front of each slit to serve as which-path markers. The quarter-wave plates mark the polarization of the interfering photon and thus destroy the interference pattern. To recover interference, we measure the polarization of the other entangled photon. In addition, we perform the experiment under delayed erasure circumstances. "