Quantum eraser thought experiment?

[Johan0001]

http://en.wikipedia.org/wiki/Delayed_choice_quantum_eraser

From the DCQE experiment above , assuming we could ,


1. Send the idler photons on a long round trip that too a year or so before going through the
double slit which is in the same lab as the detector D0.
2. Take note of each photon leaving the laser beam ( kind of tag it with a number 1.,2, 3 as they leave)
3. Send each photon at say 10 second intervals.

Then say I sent a few thousand of these photons and I "Detected" them on D0 as a white block having no pattern , only a superposition of all the states ( I.E. interference and no interference patterns).

But now while the idler photons are doing their round trip , I decide to Pre-select on detector D0, Only those photons that for my intensive purpose , completes a curve that looks like an interference pattern that I would want to pass through the Double slit.

Once I have these Pre selected photons , I decide to direct the unselected ones away from the double slit ( since they have no use for me now), by use of some fast switching mechanism.


For the preselected photons I decide to put polarisers onto the slits, to give me the which way information AND subsequently my NON interference pattern which I can detect behind the Double slit.

So now I have an interference pattern at D0 but a NON interference pattern behind the double slit on another Detector , with the SAME ENTANGLED PHOTONS!

What am I missing here?

 

[.Scott]

Once I have these Pre selected photons, I decide to direct the unselected ones away from the double slit ( since they have no use for me now), by use of some fast switching mechanism.

Here's the sketch from your wiki article.
Notice that the beam splitter, the vertical black bar at the top left, is the very first component that any photons pass through.

Unless you say otherwise, I assume you meant to say that you would direct the unselected photons away from prism PS, not from the double slit.

----- edit -----
You're going to have to restate. While the idlers are on their 1-year voyage, the signal photons have already been detected at D0, so there is nothing to deflect from the splitter. You must mean PS. But you say splitter so many times, I cannot tell which are misstatements and which are as intended.

I get the general gist that you are trying to impose an interference pattern at D0. And you could do that by blocking the splitter in sync with the motion of D0. You could also do that by moving D0 rapidly past the areas that you want to keep dark and slower past the area you want light.

 

[Cthugha]

So now I have an interference pattern at D0 but a NON interference pattern behind the double slit on another Detector , with the SAME ENTANGLED PHOTONS!

What am I missing here?

edit 2: forget what I wrote first.

Edit: Maybe I got your idea wrong...you might need to restate...why do you think there should be a problem with seeing an interference pattern in one arm, but not in the other if you send them through different double slits? Seeing an interference pattern at the signal side means that you assured that your selected ensemble has a reasonably well defined wave vector and you have no which-way information here. The idler photons will also share the well defined wave vector which is a necessary condition to see interference, but not a sufficient one. You will not see an interference pattern on the idler side if you have which-way info there. However, which way info at the idler side does not give you which way info about the signal side in that case, so I do not see a real problem.

 

[Johan0001]

Hi
Thanks for the replies guys.
After reading my post with your replies , I realize that I did not present the thought experiment quite right.
Using the diagram posted earlier.

Yes you are correct , in assuming that :
The Preselected Idler photons that I chose , using their corresponding Signal photon at detector D0. Must pass through the prism PS and the Photons that I did not select must be Directed away from the prism PS (after they have completed their long round trip, not neccesarrily 1 year).


Senario 1.

Now if I chose photons at Detector D0 that plotted out an interference curve , are we saying that:
All the photons passing through PS will continue to either D1,D2 because they represent a subset of an interference pattern at D0 that I earlier preselected.


Which implies that , I can predict what is going to happen with my preselected photons at the "Beam splitters"!


Senario 2

I could just as well have selected photons at DO that represent a particle like pattern i.e. Bell shape curve , which would "Imply" that the photons should only be detected at D3,D4 one year later!
Again predicting what would happen to the photon's at the beam splitters, one year later!


Is this correct , if not what will the results be of the experiment?

 

[Cthugha]

Hi
Now if I chose photons at Detector D0 that plotted out an interference curve , are we saying that:
All the photons passing through PS will continue to either D1,D2 because they represent a subset of an interference pattern at D0 that I earlier preselected.

Well, not ALL the photons will go to D1 or D2, but you will get some tendency. Have a look at the plots of the joint detection rates R01 to R04 in figure 3 of the wiki article. If you preselect photons on a specific pattern and discard all the other ones, the number of photons going to D1, D2, D3 and D4 will more or less be determined by the overlap of the pattern you selected and the coincidence count patterns. So if you pick a pattern that looks exactly like R01, you will get lots of detections at D1, few at D2 and moderate amounts at D3 and D4.

For the same reasons you will get more or less equally distributed detection rates when picking a bell shaped curve.

 

[Johan0001]

Thanks Cthugha that clears some of my confusion

Keeping this train of thought though , am I correct in saying if we perform the experiment (without the Beam splitter {BS_C}, the one that "erases" the which path information)

Will we ALWAYS get Bell shaped distributions on all the detectors D0 ...D4.

If this is so

Then in my thought experiment , what would we see on the detectors D0..D4 after I year has passed , if I only removed the Beamsplitter {BS_C} at that point when all the Signal photons have reached D0, but the Idler photons were still in the process of doing their long round trip?

Obviously I ask because when all the Signal photons have hit the detector D0 , I have no knowledge of which path they took. They should show an interference pattern at that instance , right?
But after I remove the BeamSplitter {BS_C} , I WILL KNOW which path they took, when they arrive later.

Is this not a contradiction , since we are testing the fact that knowing the which path , collapses the interference and produces a Bell shape?

 

[Cthugha]

Will we ALWAYS get Bell shaped distributions on all the detectors D0 ...D4.

You need to be a bit more precise. I cannot follow your scenario. Are you interested in what you see when looking at D0 alone or at the coincidence counts between D0 and another detector? This is a crucial difference. The sum of all detections at D0 will always just show a bell shaped curve, no matter what you do.

 

[Johan0001]

This is the crux of my question:


Then in my thought experiment , what would we see on the detector D0, after one year has passed.
If I only removed the Beamsplitter {BS_C} at that point when all the Signal photons reached D0, but the Idler photons were still in the process of doing their long round trip?

Obviously I ask because when all the Signal photons have hit the detector D0 , I have no knowledge of which path they took at that time.
They should show an interference pattern at that instance , right?

But after I remove the BeamSplitter {BS_C} , I WILL KNOW which path they took, when they arrive later.

Is this not a contradiction , since we are testing the fact that knowing the which path , collapses the interference and produces a Bell shape?
OR
otherwise put why did I see interference if I now Know the which path by removing the Erasure component during the experiment?

 

[Cthugha]

Then in my thought experiment , what would we see on the detector D0, after one year has passed.
If I only removed the Beamsplitter {BS_C} at that point when all the Signal photons reached D0, but the Idler photons were still in the process of doing their long round trip?

Obviously I ask because when all the Signal photons have hit the detector D0 , I have no knowledge of which path they took at that time.
They should show an interference pattern at that instance , right?

No. As I tried to emphasize before, you NEVER see an interference pattern at D0 alone. Having no which-way information is a necessary condition for an interference pattern to form, but not a sufficient one. Try putting a broad double slit in sunlight and you will not see an interference pattern.

 

[Johan0001]

"Having no which-way information is a necessary condition for an interference pattern to form, but not a sufficient one. Try putting a broad double slit in sunlight and you will not see an interference pattern. "

I've lost you here. Then what else is necessary , to see an interference pattern??

Assuming of course you have the right slit spacings relative to the wavelength of the incoming
light / photon . Which I assume you are not referring to.

http://en.wikipedia.org/wiki/Double-slit_experiment

 

[DrChinese]

"Having no which-way information is a necessary condition for an interference pattern to form, but not a sufficient one. Try putting a broad double slit in sunlight and you will not see an interference pattern. "

I've lost you here. Then what else is necessary , to see an interference pattern?

Coherence is one, and please note that as a general rule: Entangled photons do not generate an interference pattern in a double slit setup. This can be a bit confusing and I believe you already understood that.

 

[Johan0001]

Coherence is one
True if we treat the single photon as series of interfering waves after the slits , during its propagation from emission to absorption. The math works I concede that.

From the wiki definition:

Does delayed choice violate causality?[edit]

Experiments that involve entanglement exhibit phenomena that may make some people doubt their ordinary ideas about causal sequence. In the delayed choice quantum eraser, an interference pattern will form on D0 even if which-path data pertinent to photons that form it are only erased later in time than the signal photons hit that primary detector. Not only that feature of the experiment is puzzling; D0 can, in principle at least, be on one side of the universe, and the other four detectors can be "on the other side of the universe" to each other.[20]:197f

However, the interference pattern can only be seen retroactively once the idler photons have been detected and the experimenter has had information about them available, with the interference pattern being seen when the experimenter looks at particular subsets of signal photons that were matched with idlers that went to particular detectors.[20]:197




Why does one need to define interference after the IDLER photons are absorbed at a much later time.

Are we saying that there was no interference until we measured the idler photon?
Does it follow then as well, that we have changed our conscious experience of what we experienced previously, by our measurement of Idler photons in our conscious future.

I find this a paradox to say the least , there must be hidden variables or the theory is flawed.

 

[Cthugha]

Coherence is one
True if we treat the single photon as series of interfering waves after the slits , during its propagation from emission to absorption. The math works I concede that.

This is fundamental. I am not sure you get it right. Incoherent light just does not create an interference pattern, regardless of whether you have single photons, sunlight or a laser beam unless you filter it to become more coherent.

Why does one need to define interference after the IDLER photons are absorbed at a much later time.

Are we saying that there was no interference until we measured the idler photon?
Does it follow then as well, that we have changed our conscious experience of what we experienced previously, by our measurement of Idler photons in our conscious future.

No, none of this is happening. There are no interference patterns appearing or disappearing. The results at D0 NEVER EVER show interference. Not beforehand. Not after the idler photons are detected. Once you have detected the idler photons at, say, D1 you are able to identify the corresponding signal photons. If you just keep those and throw away all the other detection events, this filtered subset shows an interference pattern. You do not actively choose whether there is an interference pattern or not retroactively. You choose whether you will be able to access the information telling you which detection events to keep and which ones to throw away in order to see an interference pattern.

 

[DrChinese]

I find this a paradox to say the least , there must be hidden variables or the theory is flawed.

You have skipped a number of things to arrive at this conclusion. Hidden variables are not a deduction from the evidence. And the theory isn't flawed if it makes useful predictions.

Please note that few scientists actually believe there are hidden variables. If there are hidden variables, they are non-local.

While paradoxical, this experiment is just one of many strange ones involving entanglement. For example, it is possible to entangle a pair of particles AFTER they have been measured. In fact it is possible to entangle particles that have never existed in each other's light cones. It is possible to entangle particles that never existed at the same time.

http://arxiv.org/abs/quant-ph/0201134
http://arxiv.org/abs/1209.4191

 

[Johan0001]

No, none of this is happening. There are no interference patterns appearing or disappearing. The results at D0 NEVER EVER show interference. Not beforehand. Not after the idler photons are detected. Once you have detected the idler photons at, say, D1 you are able to identify the corresponding signal photons. If you just keep those and throw away all the other detection events, this filtered subset shows an interference pattern. You do not actively choose whether there is an interference pattern or not retroactively. You choose whether you will be able to access the information telling you which detection events to keep and which ones to throw away in order to see an interference pattern.

This makes perfect sense and I agree totally.
The only puzzle is why the idler photon , picks the right detector every time , in our future , and when we correlate back to the Signal photon IN THE PAST , the "pattern" corresponds to which way info being available or not .

I can only think that a photon in free space has no time associated with it , the emission and absorption , in the photons frame of reference occurs simultaneously. There is no in between.

As Einstein said "Ïf we could travel on a photon time would stop for us."

So the IDLER photon information at D1 is available to the SIGNAL photon at the same time even though for us they are separated in space and time , they are not for the photons..

 

[Johan0001]

In fact it is possible to entangle particles that have never existed in each other's light cones. It is possible to entangle particles that never existed at the same time.


Time in the future present or past does not exist in the photons frame of reference . So I guess they are possible. The mechanism of entanglement may be time invariant. Since we are bounded by time , we cannot escape it.

"its like asking a fish in water , what is water " "änd entanglement only happens in air"

Great posts guys , thanks for your insights.