Single Photon Double Slit Experiment

[TJonline]

TL;DR Summary

We're told that single photons passing through a double slit produce an interference pattern, but the act of observing which slit the photon passes through causes the interference pattern to show a simple ballistic pattern instead. But observing which slit the photon passes through necessitates that the photon be influenced by the observation. Why does that fact not invalidate the ability to make definitive conclusions about the observer in relation to quantum events at a fundamental level?

We're told that single photons passing through a double slit produce an interference pattern, but the act of observing which slit the photon passes through causes the interference pattern to show a simple ballistic pattern instead. But observing which slit the photon passes through necessitates that the photon be influenced by the observation. Why does that fact not invalidate the ability to make definitive conclusions about the observer in relation to quantum events at a fundamental level? In other words, why is that not just considered an experimental limitation rather than being considered evidence of the weirdness of the quantum phenomena?

 

[PeterDonis]

We're told

Can you give a specific reference? There are a lot of treatments of the double slit experiment out there, and not all of them do a very good job of explaining the issues involved.

observing which slit the photon passes through necessitates that the photon be influenced by the observation

In the sense that doing this changes the results, certainly. Whether the cause of that change in the results is properly described as the photon being "influenced by the observation" is a different question; that depends on what you mean by "influence" and "observation".

why is that not just considered an experimental limitation rather than being considered evidence of the weirdness of the quantum phenomena?

If it were an experimental limitation, there would be some way of getting around it. There isn't.

Since the term "observation" carries connotations that are often unhelpful, let's take that out of the picture and discuss a scenario in which no humans or other conscious beings are involved; there is just a setup that allows some random choice mechanism (flipping a coin, running a random number generator, seeing whether a radioactive atom decays or not, etc.) to determine whether or not an automated device is put in place at each slit that will record a macroscopic record--for concreteness, let's say making a mark on a sheet of paper, including a time stamp--if and only if a photon passes through that slit. We also have automated devices that record when each photon is emitted from the source, and when and where each photon hits the detector screen.

If we then make a large number of runs of this experimental setup, we will have results falling into three categories:

(1) Record of photon emitted from source, record of photon passing through slit #1, record of photon hitting detector at some point.

(2) Record of photon emitted from source, record of photon passing through slit #2, record of photon hitting detector at some point.

(3) Record of photon emitted from source, no record at either slit (because the random choice mechanism chose not to put the recording devices at each slit in place), record of photon hitting detector at some point.

Results in category #1 will show an image of slit #1 on the detector screen, without any interference pattern; results in category #2 will show an image of slit #2 on the detector screen, without any interference pattern; results in category #3 will show an interference pattern on the detector screen.

The point is that there is no way to design an experiment that (a) produces records as described above, and (b) does not have results that fall into the three categories above, with the detector images as described for each category.

Does this help to resolve your question? If not, can you rephrase your question in terms of the setup described above, to make clear what unresolved issue you see?

 

[Dale]

In other words, why is that not just considered an experimental limitation rather than being considered evidence of the weirdness of the quantum phenomena?

Well, quantum weirdness isn’t a well defined thing, but I certainly find it weird that an electron interferes with itself. And I also find it weird that it is never measured to go through both slits.

 

[vanhees71]

Quantum theory becomes weird only in most popular-science books on the topic, because weirdness sells unfortunately better than science, or when philosophers write about it. There's nothing weird about it in the lab nor in most scientific papers which deal with physics, i.e., objective observable facts about nature.

 

[troglodyte]

Quantum theory becomes weird only in most popular-science books on the topic, because weirdness sells unfortunately better than science, or when philosophers write about it. There's nothing weird about it in the lab nor in most scientific papers which deal with physics, i.e., objective observable facts about nature.

compared to the classical physics there is definitely some strangeness in the quantum world .Without this comparison you are right.This are all empirical facts.So ,it is on you which view do prefer and how do you want to look on the quantum world.

 

[DrChinese]

We're told that single photons passing through a double slit produce an interference pattern, but the act of observing which slit the photon passes through causes the interference pattern to show a simple ballistic pattern instead. But observing which slit the photon passes through necessitates that the photon be influenced by the observation. Why does that fact not invalidate the ability to make definitive conclusions about the observer in relation to quantum events at a fundamental level? In other words, why is that not just considered an experimental limitation rather than being considered evidence of the weirdness of the quantum phenomena?

It is not an experimental limitation. You can in fact make observations at BOTH slits simultaneously and get interference - or not! For a double slit setup using a stream of photons:

a. Place a polarizer oriented at 0 degrees over each slit, and then build up a pattern. That pattern will display the usual interference effects, and you are observing the polarization of every photon going through.

b. Now rotate one of the two polarizers by 90 degrees. There is no more, and no less, interaction between the apparatus for light emitted from the source and arriving at the screen. The total intensity does not change. You are still observing the polarization of each photon, correct? And yet: the interference pattern completely disappears.

The reason the interference disappears in configuration b. : There is the possibility (in principle) of determining which slit the photons go through. (It matters not whether that information is actually obtained.) By the usual explanation, interference is absent when which slit information is available. On the other hand, in configuration a.: If the photon goes ONLY through one slit, how does it know to act differently depending on the setting (polarizer orientation) of the other slit?

http://sciencedemonstrations.fas.ha...-demonstrations/files/single_photon_paper.pdf

 

[vanhees71]

Again one must emphasize that photons are never ever to be visualized as pointlike particles. You rather have to think in electromagnetic waves even if you have only one photon each time in the experiment.

Particularly if you have setup, where you get double-slit interference fringes the transverse extension of the wave packets describing the photons in the single-photon Fock states in the only correct QED picture, (transverse to the momentum of the photon) must be large enough to cover both slits. Otherwise there's no double-slit interference at all, because the two slits are not both "illuminated" in the incoming channel. For the same reason the detector must be far enough from the slits such that the outgoing partial waves from both slits overlap well, because otherwise there's no interference of the partial waves in the outgoing channel.

For the same reason the explanation in #6 how to gain which-way information (e.g., using quarter-wave plates in 90-degrees relative orientation in the slits) by storing the which-way information in the photons themselves in terms of polarization. This is only possible, because the photon is not simply describable as a billard-ball like miniature particle but a (quantized) wave field.

 

[PeterDonis]

You can in fact make observations at BOTH slits simultaneously and get interference - or not!

But, as you yourself point out, these observations only tell you which slit the photon went through in case b, not case a. So your case a. corresponds to my case (3) in post #2, and your case b. corresponds to my cases (1) and (2) in post #2.

However, your setup now has an additional case, namely:

(4) Record of photon emitted from source, and nothing else (because the photon was absorbed by the polarizers instead of being passed through).

 

[bhobba]

Again one must emphasize that photons are never ever to be visualized as pointlike particles.

Why picture it at all? This is beyond usual experience in which human intuition and mental pictures are developed. We describe what's going on mathematically, and can show as a limiting case the usual intuitive mental pictures emerge. Maybe some people's expectations are overly ambitious?

Thanks
Bill

 

[vanhees71]

The problem is that in discussing the double slit experiment people think about photons or particles as if they could be described as minaturized bullets, i.e., within the classical-point-particle concept. Then it's claimed that the observed interference phenomena are weird and thus that quantum theory is weird.

It's, however, important to note that the opposite is in fact the case! Quantum theory is resolving all weirdness into the most successful theory in physics ever discovered. There's no weirdness at all but a successful description of all phenomena known today (except "quantum gravity" though there are no phenomena related with it observed yet).

So, indeed, as you say, when learning QT you have to forget the classical concept (except the symmetry principles which successfully describe classical and quantum theory) and learn the "new" quantum concepts of how to describe the phenomena! The double-slit experiment is a great opportunity to do so. See the introductory chapter of the Feynman Lectures vol. III:

https://www.feynmanlectures.caltech.edu/III_01.html

 

[TJonline]

Ok, if I understand correctly (in oversimplified terms), a photon passes through both slits (and every possible path in some sense) because it acts as a wave rather than as a particle. And observing necessitates interfering with its path(s), simply because learning anything about a photon in transit necessitates that you interact with it somehow (i.e. It can't both travel freely AND tell you where it is at some point on the way without bouncing something else off of it or bouncing it off of something else). What got me wondering was actually a recent broadcast of the new Cosmos series in which the double slit experiment is being 'entertained'. Maybe the following is technically true, but is made to sound more like magic than good science, because magic sounds cooler.

"The mere act of observation changes reality"

"So the reason we didn't get the interference pattern earlier wasn't because we chopped up the light into single photons. It was because we were observing which slit the photons passed through. But how could a photon know if someone is watching? A photon doesn't have eyes. A photon doesn't have a brain. How could it know it was being watched? You might reasonably conclude that a single photon is such a tiny thing that it's very hard to see without using complex technology. This machinery does violence to the delicate photon. Changes it. But that doesn't explain why photons behave like particles when we're watching, but waves when we're not. If light is fundamentally a particle, then it should never create a wave pattern whether we're observing or not. How could individual photons know where to take their places so that as a group, they create the interference patterns of waves? This is a maddening conundrum at the heart of quantum physics."

"Until we make an observation, the photon exists in a state of uncertainty governed by laws of probability. And when we do observe it, it becomes something completely different."

 

[DrChinese]

Ok, if I understand correctly (in oversimplified terms), a photon passes through both slits (and every possible path in some sense) because it acts as a wave rather than as a particle. And observing necessitates interfering with its path(s), simply because learning anything about a photon in transit necessitates that you interact with it somehow (i.e. It can't both travel freely AND tell you where it is at some point on the way without bouncing something else off of it or bouncing it off of something else).

Please re-read my post #6. You can observe its polarization (in every case) and still either have interference or not. It is not just the act of observation. It is the specific information gained (from the specific setup). The only thing changing is the RELATIVE orientation of the polarizers. Nothing is otherwise added to any version.

So then the question becomes... how does it know the relative orientation is one way or the other? This is the part of things that is something amazing - and confusing. There are different interpretations that attempt to explain this in "common sense" terms. However, this is described correctly by standard application of QM. So technically, things stop there.

 

[TJonline]

Please re-read my post #6. You can observe its polarization (in every case) and still either have interference or not. It is not just the act of observation. It is the specific information gained (from the specific setup). The only thing changing is the RELATIVE orientation of the polarizers. Nothing is otherwise added to any version.

Is is true that the photon that would pass through both slits has identical polarization through both paths? If so, then (b) seems to say that one or the other slit always blocks one or the other path (or both), so you never see an interference pattern (or nothing). And (a) always keeps both paths open (or blocks both paths), so you do see an interference pattern (or nothing).

 

[bhobba]

Ok, if I understand correctly (in oversimplified terms), a photon passes through both slits (and every possible path in some sense) because it acts as a wave rather than as a particle.

Here is a better explanation (though it still has issues - as Vanhees may elaborate on) of the double slit experiment, that, correctly IMHO, avoids the wave-particle duality thing. More advanced textbooks such as my bible - Ballentine - QM - A Modern Development never even mentions this wave particle stuff because really it is only an aid for learning QM at the beginning:
https://arxiv.org/pdf/quant-ph/0703126.pdf

Once you start to understand the true basis is the mathematics, rather than pictures like wave-particle duality, you can see its best to just think in terms of the math. It's not really shut-up and calculate - it's more like - this is outside human everyday experience - trying to pigeonhole it in some way to that paradigm (ie everyday experience) just causes confusion. A much better way of thinking about the double slit is as demonstration of the indeterminacy principle and superposition principle. The indeterminacy principle means when the wave-function of the quantum particle encounters the slit the momentum behind the slit is unknown because we know its position ie have conducted a position measurement. When you have two slits we have conducted a measurement to determine if it is at one or the other slit and the momentum behind each slit is not determined. So to get the wave-function you must take the superposition of both. When you chug through the math you get the double slit interference pattern as explained in the paper.

Thanks
Bill

 

[vanhees71]

Ok, if I understand correctly (in oversimplified terms), a photon passes through both slits (and every possible path in some sense) because it acts as a wave rather than as a particle. And observing necessitates interfering with its path(s), simply because learning anything about a photon in transit necessitates that you interact with it somehow (i.e. It can't both travel freely AND tell you where it is at some point on the way without bouncing something else off of it or bouncing it off of something else). What got me wondering was actually a recent broadcast of the new Cosmos series in which the double slit experiment is being 'entertained'. Maybe the following is technically true, but is made to sound more like magic than good science, because magic sounds cooler.

"The mere act of observation changes reality"

"So the reason we didn't get the interference pattern earlier wasn't because we chopped up the light into single photons. It was because we were observing which slit the photons passed through. But how could a photon know if someone is watching? A photon doesn't have eyes. A photon doesn't have a brain. How could it know it was being watched? You might reasonably conclude that a single photon is such a tiny thing that it's very hard to see without using complex technology. This machinery does violence to the delicate photon. Changes it. But that doesn't explain why photons behave like particles when we're watching, but waves when we're not. If light is fundamentally a particle, then it should never create a wave pattern whether we're observing or not. How could individual photons know where to take their places so that as a group, they create the interference patterns of waves? This is a maddening conundrum at the heart of quantum physics."

"Until we make an observation, the photon exists in a state of uncertainty governed by laws of probability. And when we do observe it, it becomes something completely different."

That's not I would express it. It's not the "act of observation" (whatever this should be) but the interaction between the photon with the setup of the experiment.

Let's carefully rethink the idea to gain (or not gain) "which-way information" using the photon's polarization. In the first setup we just send photons polarized in the $x$ direction through the double slit. Then there is no way to decide through which slit each photon went when hitting the screen, where it is absorbed leaving a mark so that the position of this absorption is measured (within the resolution of the measurement device, e.g., the pixel size of the CCD cam, if this is your screen). The probability of detection a photon at a given place is given by the usual double-slit formula for waves with interference fringes due to the superposition of partial waves going through the openings (leading to both single- and double-slit fringe patterns).

Now put quarter-wave plates in each slit, one oriented $45^{\circ}$ and the other $-45^{\circ}$. All you need to know is that an (idealized) quarter-wave plate oriented in $\pm 45^{\circ}$ is described as a unitary operator on the polarization state of the photon such that a photon polarized in $x$-direction becomes a photon with helicities $\pm 1$ (i.e., left- and right-circular polarized em. waves). Since these are mutually orthogonal states the photon after the slit carries the information through which slit it came, since its helicity is in 100% correspondence between the slit through which it came. You don't need to measure the polarization though. It's enough that now the photon is prepared such that its polarization implies with 100% certainty the information through which slit it came. That this correlation between polarization and which-way information is 100% is due to the fact that the polarization states are orthogonal to each other, but this orthogonality also implies that there is no interference term between the different ways the photons go through the double slit, and thus the double-slit interference pattern is completely wiped out and the interference pattern at the screen is just the incoherent superposition of the single-slit interference patterns of the slits. That's an example for the fact that the diffraction property of photons ("wave property") is excluded by the (principle) possibility to know with certainty through which slit they came ("particle property"). It's clear that "reality is changed" just due to the setup of the entire experiment, i.e., because it makes a difference whether the photons interact with the quarter-wave plates within the slits or if they don't. There's nothing weird about this since any measurement necessarily means that the measured object has to interact with the meaurement apparatus, and usually this interaction alters the state of the measured object (there are exceptions like socalled quantu nondemolation measurements, but that's another topic).

 

[weirdoguy]

Why picture it at all?

I'm going to steal that question from you if you don't mind. It's pure gold

 

[TJonline]

I'm going to steal that question from you if you don't mind. It's pure gold

Answer (to the quesiton): To help simple minded dufuses like me to try to get an intuitive understanding of complex phenomena that require brain stretching. Like: A water drop in a pool creates a wave that in turn (if big enough) may cause another particle to pop back out as an intro to particle wave duality. Oh yea! I see that! Fascinating! Wha? It's a simplistic model? I'll look deeper.

 

[DrChinese]

Is is true that the photon that would pass through both slits...

This is my point. The particle cannot be said to independently go through one slit and one slit only. And that statement actually applies to either of the setups.

Assume the source emits its beam 45 degrees offset from the polarizers at the slits. It should be clear that since random chance determines whether the light will pass each polarizer, some of the light paths could have a chance go through BOTH filters. Even when the filters are crossed; as in those cases there are different peaks and valleys than when the filters are parallel. (Admittedly it looks "as if" the light went through one or the other and not both, in the crossed case.)

So the conclusion is: you must look at the ENTIRE setup, apply the quantum description, and get a good answer. You can't pick and choose elements to focus on without looking at the whole picture. Otherwise, you end up concluding that the light knows what it needs to do "here" based on something that occurs "there". Yes, in all cases the photons can be said in some sense to go through both filters. But the resulting patterns are different when the relative orientation of the filters change, as described by QM.

 

[bhobba]

Answer (to the quesiton): To help simple minded dufuses like me to try to get an intuitive understanding of complex phenomena that require brain stretching.

Then stick wth the basic textbook/popularisation wave-particle idea. But 'simple-minded' is not always compatible with a deep analysis. You pull that idea apart by deeply looking at it then it falls to pieces so to speak. The only one that doesn't do that is if you just stick to the math, which is what advanced texts do - you have now moved beyond the basic level. It is unfortunate physics is sometimes like that. One of it's greatest teachers - Feynman - didn't like it was like that - but never could figure out how to completely avoid it, and had to accept it. He too said the double slit contains the essential mystery - and it does - but its resolution in terms of the wave-particle idea is useful only at the beginner level.

Thanks
Bill

 

[bhobba]

This is my point. The particle cannot be said to independently go through one slit and one slit only.

True. But just to put it another way let's say you have one slit. You can say behind the slit the particle has a definite position - the position of the slit. If you measure the position behind the slit you will always get an exact answer or you can look at the slit as measuring the position - either view is fine. Now what position did it have before the measurement? QM says - like Newton once quipped - I make no hypothesis. People get that from Newton quite easily - it wasn't until Einstein further progress was made. Thats what I mean by just trusting the math - it's not hard really. Have two slits and if you measure the position of the particle behind the slits you will find it behind one slit or the other. It's the same view as the single slit. Now there is another principle of QM - called the principle of superposition (plus a very reasonable symmetry argument) that allows one to figure out the wave function behind the slits - and chugging through the math you get an interference type pattern when you measure the position again in the screen behind the slits. Whats going on between the slits and the screen - QM makes no hypothesis - like Newton made no hypothesis. You can calculate the probability of what position you would get when you measure it - but that is all. Once you accept what the math says - like accepting what Newton said about his math - you do not run into issues. But try to form a picture - then as Feynman says - you go down a hole nobody has been able to escape from - just like before Einstein you go down a hole nobody was able to escape from if you wondered why gravity acted like it did. Actually any theory is like that - it has some things you just accept as the way it is. Ultimate knowledge is beyond the scope of science - but that is getting into the philosophy of science, which for various reasons is not discussed on this forum.

Thanks
Bill