DIY Quantum Eraser: Exploring the Interference of Orthogonal Light Waves

[JAguilon]

Click here for the publication.

Having performed this experiment, I have gotten clean results. Essentially, a double slit is made by putting an electron beam in the way of a wire with orthogonal polarizers on either side. This destroys the expected interference pattern since the polarized filters "measure" the path of the photons. However if one places a 45 degree polarizer that allows the orthogonal light waves to both pass through, the interference pattern restores. According to the article, this is a "quantum eraser" since the wave nature was destroyed with the perpendicular polarizers and restored afterwards with the 45 degree filter.

This being said, I also understand that the classical Fresnel-Arago laws (link) state that orthogonal waves do not interfere. Wikipedia also mentions that when particle detectors are at the slits, the wave function should collapse. But it also states that this experiment has never been published. Here we have an experiment that places a "detector" at the slits, and as far as Scientific American says, it has collapsed and even restored the wave function. Now, I can only think of 2 conclusions to this:

1) The Fresnel-Arago Laws were a precursor to quantum mechanics and there is no interference because the information has been leaked into the outside environment
2) This is purely a classical experiment and can be explained as such

Is this experiment just a demonstration of classical optics or is there actually a quantum nature to this? I also wonder if Fresnel and Arago had an explanation to the nature of orthogonal light waves, or if the quantum mechanical wave collapse due to observation is the only reason. Does anybody have information on this?

Much gratitude for your thoughts! This is for a science fair project for my high school, so I would greatly appreciate it since I no longer know whether I should present it as a classical twist to the double slit experiment or a true quantum mechanical phenomenon.

 

[DrDu]

As far as I remember, they even state in the article that it is a completely classical phenomenon.

 

[mfb]

It is a purely classical experiment if the intensity is large. If you use single photon sources, a classical treatment breaks down (as there are no photons in a classical wave description), but the experiment still gives the same result.

 

[Jilang]

Classically, if light were polarised in a certain direction, how would any light get through a filter at 45 degrees to that direction?

 

[mfb]

Classically, if light were polarised in a certain direction, how would any light get through a filter at 45 degrees to that direction?

There is a field component in the direction of the filter, and this component* can get through.
It's like moving diagonally on a street, you are moving forwards (or backwards) as your motion has a velocity component in this direction.*actually the parallel one for the magnetic field and the orthogonal one for the electric field, if the filter consists of a set of parallel, conducting wires.

 

[Jilang]

I can't quite see that. If I were moving diagonally across the street I wouldn't be able to cross at a zebra crossing even if I had a component in that direction.

 

[mfb]

If you replace the stripes by walls and approach the zebra diagonally, you will collide with the walls and keep some velocity component to cross the streets.

 

[Jilang]

I guess the analogy is what I am not comfortable with. Say I am moving across the as street with my arms out in the 45 degree direction and then go though a narrow aperture across the street, then it would take my arms off.

 

[DrChinese]

It is a purely classical experiment if the intensity is large. If you use single photon sources, a classical treatment breaks down (as there are no photons in a classical wave description), but the experiment still gives the same result.

...

There is a field component in the direction of the filter, and this component* can get through.

I would think that if we applied the classical approach to this setup (orthogonal polarizers) there would be a percentage chance that a component could go though both filters. After all, classically there should be a separate probability of transmission at each polarizer. But the actual likelihood is zero.

In other words, there would be some interference effect even if it were a reduced effect, on the average. But the experiment shows none. So in my mind, this isn't explainable classically. You said as much, I guess, by your comment about intensity.

 

[mfb]

I guess the analogy is what I am not comfortable with. Say I am moving across the as street with my arms out in the 45 degree direction and then go though a narrow aperture across the street, then it would take my arms off.

Okay, this is certainly beyond the applicability of my comparison now.
The classic situation is simple, you can use the Maxwell equation and see that 50% of the intensity goes through.

I would think that if we applied the classical approach to this setup (orthogonal polarizers) there would be a percentage chance that a component could go though both filters. After all, classically there should be a separate probability of transmission at each polarizer. But the actual likelihood is zero.

Do you mean without eraser behind the slits? Then the wave passes through both slits, does not interfere in a relevant way afterwards (you can treat both orthogonal components separately - classical electrodynamics is linear) and you get the same result as with quantum mechanics.

In other words, there would be some interference effect even if it were a reduced effect, on the average. But the experiment shows none. So in my mind, this isn't explainable classically. You said as much, I guess, by your comment about intensity.

In the quantum eraser setup, there is interference.

 

[JAguilon]

Thank you everybody for the input. I think I have a bearing of what all of this means now.

When I had orthogonal filters on each slit, the result looked very much like the single slit experiment. The light extended outwards horizontally, but no interference was present. Does this mean both waves are spreading out--but not interfering with each other--on the same spot?

 

[mfb]

Thank you everybody for the input. I think I have a bearing of what all of this means now.

When I had orthogonal filters on each slit, the result looked very much like the single slit experiment. The light extended outwards horizontally, but no interference was present. Does this mean both waves are spreading out--but not interfering with each other--on the same spot?

Right.

 

[JAguilon]

Right.

Now I am confused as to how this could ever be quantum, even if you fired one photon at a time. I thought that the wave function collapse would mean that the photon starts acting out as a particle, meaning there should be two spots of light. In this experiment, you just get two waves that overlap each other when they are polarized orthogonally. Firing single photons at a time should yield the same result. Am I missing something?

 

[DrChinese]

Do you mean without eraser behind the slits? Then the wave passes through both slits, does not interfere in a relevant way afterwards (you can treat both orthogonal components separately - classical electrodynamics is linear) and you get the same result as with quantum mechanics.

In the quantum eraser setup, there is interference.

Well it's a bit difficult for me to simultaneously think in terms of classical waves and individual particles.

But my point is that a classical wave polarized at 45 degrees has an independent probability of traversing a left polarizer at 0 degrees as well as a right polarizer at 90 degrees. That's because the setting at one place (left) cannot affect the passage at the other (right). This is regardless of whether there is interference or not. Of course, this is tantamount to saying that you can have intensity of less than one wave.

Of course in a quantum world, none of that ever happens. So I see experiments such as this as providing confirmation of quantum behavior. But that's just me.

 

[mfb]

Now I am confused as to how this could ever be quantum, even if you fired one photon at a time. I thought that the wave function collapse would mean that the photon starts acting out as a particle, meaning there should be two spots of light. In this experiment, you just get two waves that overlap each other when they are polarized orthogonally. Firing single photons at a time should yield the same result. Am I missing something?

You cannot get two spots with a single photon. Each photon will get detected at some position and nowhere else.

Firing single photons at a time should yield the same result

Sure, but that is not classical mechanics any more, as the concept of photons does not exist there.

But my point is that a classical wave polarized at 45 degrees has an independent probability of traversing a left polarizer at 0 degrees as well as a right polarizer at 90 degrees. That's because the setting at one place (left) cannot affect the passage at the other (right). This is regardless of whether there is interference or not. Of course, this is tantamount to saying that you can have intensity of less than one wave.

A classical wave does not have probabilities. A classical wave has intensities, and those will split every time the wave hits a polarizer (unless 100% are reflected, or 100% are transmitted)

 

[JAguilon]

You cannot get two spots with a single photon. Each photon will get detected at some position and nowhere else.

But overtime, if single photons were fired repeatedly, having the orthogonal filters should result in two spots. Perhaps I should show an image of my results to help clear up my question. Attached is what I got when orthogonal filters were used. From my understanding, doing this experiment with quantum photons is supposed to yield the same result as firing a stream of photons. So why am I getting a wave pattern (two waves that don't interfere) rather than two sections of light? I accept that my results could be explained by a classical wave, but if we did this firing single photons, would I still be getting a wave like pattern on the other side?

 

[DParlevliet]

This has a nice relation to my topic Quantum erase explained by waves and is much easier. It don't have the Scientific American, so I hope I have the right setup in mind. It can be easy explained classical. But first some remarks:
- The polarizers before the slits are no detectors. They change the wave in a way it could be detected later on through which slit the photon went. This destroys the interference, even if this measurement is not actually done. The 45º polarizer erases this which-slit information, so interference is visible again.
- So photons are not absorbed in the polarizers (otherwise they would not reach the detector).
- The strangeness of quantum is that you explain everything with waves but when you finally measure it: there is a particle! But with a particle you could not explain the interference (perhaps Fresnel-Arago is also Quantum strangeness, I don't know).
- For the classical explanation, always visualise this in diagrams:

Choose (grey) polarizing axes x and y parallel to the (purple) polarizers at the slits. A single photon has a wave (don't ask what that is, just calculate with it). Suppose its polarization is green. You can decompose that (is that the right English word?) on the x and y axis. One polarizer transfers only the x polarization, the other only the z polarization. If those add up at the detector they are orthogonal and you know, those don't interfere.
Then place the 45º polarizer. Again you can decompose the x and y to the polarizer axis, and only polarization components parallel to this axis (green) are transfered. Those waves are now polarized in the same direction, so when added at the detector there is again interference! That is all.

Now back to Quantum and wave-particle duality, which you much use carefully:
1 It states that a photon has both wave and particle properties, but not at the same time.
2 "Properties": it does not say that the wave or particle physically exist (nobody knows yet), it is just that practice showed that you can calculate (to a certain limit) with their classical rules.
3 "Not at the same time": when you calculate with a wave, you are no allowed to think about a particle, and the other way around.
4 You cannot calculate a photon with only a wave, nor only a photon. It is parts of both.
5 During a measurement the wave-particle is determined by what you measure. So in the above setup: polarizers are wave equipment, so you will see (only) waves. But the detector is a particle detector, so you will only see only particles.
6 Quantum mechanics has written the above simple rules in complicated math, which can explain much more and can be used in much more complicated situations.

 

[mfb]

But overtime, if single photons were fired repeatedly, having the orthogonal filters should result in two spots. Perhaps I should show an image of my results to help clear up my question. Attached is what I got when orthogonal filters were used. From my understanding, doing this experiment with quantum photons is supposed to yield the same result as firing a stream of photons. So why am I getting a wave pattern (two waves that don't interfere) rather than two sections of light? I accept that my results could be explained by a classical wave, but if we did this firing single photons, would I still be getting a wave like pattern on the other side?

The slits are very close together compared to the size of the spot on the wall, I guess. This means you do get two spots, but they overlap so significantly that you don't see the difference (use a polarizer in front of the screen to verify this!).

doing this experiment with quantum photons is supposed to yield the same result as firing a stream of photons.

Photons give the same result as photons? Isn't this obvious? There are no special "quantum photons".

So why am I getting a wave pattern (two waves that don't interfere) rather than two sections of light?

Which wave pattern?

 

[JAguilon]

The slits are very close together compared to the size of the spot on the wall, I guess. This means you do get two spots, but they overlap so significantly that you don't see the difference (use a polarizer in front of the screen to verify this!).

Photons give the same result as photons? Isn't this obvious? There are no special "quantum photons".

Which wave pattern?

1. I used a polarizer and the light still appears to be expanding out horizontally like a wave. This is where I am confused. If the wave function collapsed, I always thought it would be like this:

Instead, the pattern spreads out very wide.
https://www.physicsforums.com/attachment.php?attachmentid=65426&d=1389231380

2. That is what I expected, but that doesn't seem to match up with my experiment. When I label the photons with x and y polarizations, I was expecting two spots (or at the very least an incoherent blur), but instead I got a band of light that expands very far out horizontally but not vertically (see my last post for an image).3. The fact that the photons expanded out horizontally seems to indicate that they are still behaving like a wave; they just don't interfere, possibly because of F-A laws.

Or, is the fact that no interference occurs enough to be considered "wave-collapse?"

 

[DParlevliet]

Probably you did not read my post on the first page? In nr. 3 you came to the right conclusion.

 

[mfb]

1. I used a polarizer and the light still appears to be expanding out horizontally like a wave. This is where I am confused. If the wave function collapsed, I always thought it would be like this:
http://4.bp.blogspot.com/_EZWNQhIJDqs/SjTdrIYVS8I/AAAAAAAABu8/un-zqCLxfCI/s400/DSE+2.jpg
Instead, the pattern spreads out very wide.
https://www.physicsforums.com/attachment.php?attachmentid=65426&d=1389231380

You still get single-slit interference, which basically looks like a single spot unless your experiment is very precise.

2. That is what I expected, but that doesn't seem to match up with my experiment. When I label the photons with x and y polarizations, I was expecting two spots (or at the very least an incoherent blur), but instead I got a band of light that expands very far out horizontally but not vertically (see my last post for an image).

Those are single-slit effects. Block one slit, and you should get the same result.

Or, is the fact that no interference occurs enough to be considered "wave-collapse?"

A collapse would be an irreversible process. You can (but do not have to) consider the detection at the screen as such a process, but nothing before that.