Double Slit Quantum Eraser, Really?

[i8uncertainty]

Hello all, I have been reading about quantum entanglement on this webpage. It builds up to the quantum eraser where two entangled photons are produced using a BBO crystal, that have opposite polarity, with one going through a double slit to a detector and the other going to another detector linked - to correlate detections. This setup produces an interference pattern. Then special crystals (that don't destroy entanglement) are placed in front of the slits to impart rotational polarisation either clockwise or anticlockwise depending on if the incident polarisation is horizontal or vertical. The polarisation is measured at both detectors - if its horizontal at one and clockwise at the other you can tell which slit it went through. This destroys the interference pattern.
The finale is the placement of a polarising plate on the photon on the none slit pathway so that the polarisation's horizontal and vertical components become a mix and the original polarisation cannot be determined.. the interference pattern reappears.

Well that's my following of it, the question I have though is isn't the last polarising plate simply acting as a filter so that counts are only made when the photons have a certain range of polarities that would produce a pattern i.e surely the plate blocks some photons with an exact opposite polarity of the plate and if these counts were added back to the pattern then it would just become a blob?

 

[San K]

Hello all, I have been reading about quantum entanglement on this webpage. It builds up to the quantum eraser where two entangled photons are produced using a BBO crystal, that have opposite polarity, with one going through a double slit to a detector and the other going to another detector linked - to correlate detections. This setup produces an interference pattern. Then special crystals (that don't destroy entanglement) are placed in front of the slits to impart rotational polarisation either clockwise or anticlockwise depending on if the incident polarisation is horizontal or vertical. The polarisation is measured at both detectors - if its horizontal at one and clockwise at the other you can tell which slit it went through. This destroys the interference pattern.
The finale is the placement of a polarising plate on the photon on the none slit pathway so that the polarisation's horizontal and vertical components become a mix and the original polarisation cannot be determined.. the interference pattern reappears.

Well that's my following of it, the question I have though is isn't the last polarising plate simply acting as a filter so that counts are only made when the photons have a certain range of polarities that would produce a pattern i.e surely the plate blocks some photons with an exact opposite polarity of the plate and if these counts were added back to the pattern then it would just become a blob?

welcome to the forum i8uncertainty!

you are correct...:)

this applies in the delayed choice quantum erasure experiment as well

 

[i8uncertainty]

If that's the case then this experiment doesn't show entanglement - just an elaborate (but rather nifty) way to selectively filter photons indirectly, and the delayed choice experiment after it seems a bit pointless. So I'm inclined to think I've missed something here?

 

[San K]

If that's the case then this experiment doesn't show entanglement - just an elaborate (but rather nifty) way to selectively filter photons indirectly, and the delayed choice experiment after it seems a bit pointless. So I'm inclined to think I've missed something here?

they show entanglement however what they don't show is:

the future effecting/changing the past...

 

[i8uncertainty]

Where do they show entanglement?
If we take the situation to its bare bones without detectors and even discarding the photons that take the non-slits route, then if we count all the photons going through the circularly polarising plates(Quarter Wave Plates) we get no pattern. If we take away all the photons that are predominantly polarised in the vertical or horizontal, leaving ones with components in both to go through the QWP's then we get a pattern. This to me seems like exactly the same situation, and hence why I fail see how a measurement on the entangled pair is supposedly collapsing the wave-packet.

 

[San K]

Where do they show entanglement?
If we take the situation to its bare bones without detectors and even discarding the photons that take the non-slits route, then if we count all the photons going through the circularly polarising plates(Quarter Wave Plates) we get no pattern. If we take away all the photons that are predominantly polarised in the vertical or horizontal, leaving ones with components in both to go through the QWP's then we get a pattern. This to me seems like exactly the same situation, and hence why I fail see how a measurement on the entangled pair is supposedly collapsing the wave-packet.

the short/quick answer is that ...the which-way or no-which-way is being performed only on photon A...nothing is being done to photon B...however the other photon B is behaving as if it's effected too...

There are other experiments, as well, that support Quantum entanglement.

are you, for example, aware of Bell's test/theorem?

http://en.wikipedia.org/wiki/Bell's_theorem[/PLAIN]

 

[harrylin]

Where do they show entanglement?
If we take the situation to its bare bones without detectors and even discarding the photons that take the non-slits route, then if we count all the photons going through the circularly polarising plates(Quarter Wave Plates) we get no pattern. If we take away all the photons that are predominantly polarised in the vertical or horizontal, leaving ones with components in both to go through the QWP's then we get a pattern. This to me seems like exactly the same situation, and hence why I fail see how a measurement on the entangled pair is supposedly collapsing the wave-packet.

San K already gave the short answer.

I joined this forum because of discussions about QM and entanglement (although nowadays I spend too much times on other sub forums). And although some of the literature claims that it is now a fact, this has far from been settled.

So, the long answer is Very Long. If you search this forum (use Advanced Search), you will find many discussions on Bell's Theorem.

 

[i8uncertainty]

Ok, thanks for the links and the responses, I'm starting to see the weirdness in the example of the non-liner quantum correlation of readings, still going to take a bit to get my head round though!

 

[Sylvia Else]

The interpretation of these erasure experiments bothers me. The ones I've looked at all have a similar characteristic, which is that the interference patterns that 'result' from the erasure are not directly detectable, but instead have to be found by dividing photons into two sets, which produce fringes, and anti-fringes respectively.

In the paper cited by the web page linked by the OP, it looks to me as if the results can be interpreted as indicating that the addition of the quarter wave plates has the effect of splitting the photons into two groups, each of which forms an interference pattern, but with a phase difference such that the patterns cancel out, leaving the appearance of non-interference.

The experimenter can then either position the POL1 polariser so as to obtain the information required to determine which path each s photon took, or can position it so as to determine which group the photon belonged to, but not both.

From this perspective, the s photons are completely unaffected, and it's just the experimenter's choice as to what he wants to know. It then becomes completely unsurprising that delaying the choice makes no difference - one wouldn't expect it to.

Now there may be other experiments that I'm not aware of, or which haven't been done yet, which render such a perpective manifestly untenable, but for the moment I'm not convinced these experiements give as much insight into entanglement as is advertised.

Sylvia.

 

[Demystifier]

I'm not convinced these experiements give as much insight into entanglement as is advertised.

I agree. Or in other words (see https://www.physicsforums.com/blog.php?bt=6217#comment6217 )
"If delayed choice quantum eraser shocked you more than the rest of quantum mechanics, you haven't understood the rest of quantum mechanics yet."

 

[DrChinese]

Now there may be other experiments that I'm not aware of, or which haven't been done yet, which render such a perpective manifestly untenable, but for the moment I'm not convinced these experiements give as much insight into entanglement as is advertised.

Sylvia.

I don't know about the "advertised" part. Each person has their own threshold. The essential part was laid down with the EPR 1935 paper, although they guessed wrong about entanglement (thinking that there must be local hidden variables). The context of the experiment cannot allow more information to be derived about a state than the Heisenberg Uncertainty Principle allows.

As to other experiments: you can delay the decision to entangle 2 particles until AFTER they are detected. This is done by the method called "entanglement swapping". Alice makes a decision to entangle (or not) a pair of photons sent to Bob. Those photons have never previously interacted, but will nonetheless violate a Bell inequality if they become entangled by Alice - and that can be done by Alice either before or after Bob sees them. Entangled pairs of photons will follow the cos^2(theta) prediction of QM, while un-entangled pairs will not.

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

Note: You still must perform coincidence counting to see the QM predicted results as the pairs of photons sent to Bob can be entangled either in a state in which they are the same (++ or --) or different (-+ or +-). That occurs randomly, removing any possibility of FTL signaling.

 

[Kyxha]

In the case of the quantum eraser experiment, how do the idler photons "know" if they are falling on 2 separate detectors that can distinguish the slit that their partner photons went through, or falling on one big detector that cannot distinguish the slit, thus producing interference.
It seems that in both cases information IS encoded in the idler photons, yet two patterns are observed. What am I missing?

 

[i8uncertainty]

Hello Sylvia, I came to the same conclusion - that the example doesn't seem to be showing entanglement at all, I think that webpage was meant more as a general introduction to spark the mind of a newcomer to the subject but didn't quite hit the quanta on the head regarding entanglement. However it lead me to delve a little deeper and the explanation on the wiki-page link that was posted gets into the subject a bit more.
The experiment in that explanation has a coincidence counter set up again so that the polarity of two entangled photons can be measured. Each of the photon detectors can measure the polarity at a changeable angle, so you could have one vertical and the other at 45 degrees so to speak. The setup then generates entangled photons but with opposite polarity (vertical/anti-vertical), each going to one of the detectors and when the detectors are set to measure the same polarity then the results of the incident photons will show opposite polarity all of the time, which is correct and classed as correlated.
If you set up the detectors with one measuring the horizontal and the other vertical then you will also get an expected correlation of 50% i.e half the times the detectors show both positive (one of them isn't correct). If one photon is nearly vertical with a vertical detector then at the other the photon will be anti-vertical against a horizontal detector so it could go either way hence the 50%. This may be my over simplification of the event but I think the principle is generally right. Anyway it is when the detectors are set with an angle between these that entanglement shows its head as when you total up the correlation percentage, it differs from an expected linear change in correlation per change in angle, to a curve so that when measuring at 45 degrees you actually get a greater incidence of correlated results than would be expected.
Have to say going through it now has given me some more queries, might have to start another thread!

 

[Sylvia Else]

i8uncertainty, EPR style experiments with angles of 30 and 60 degrees certainly show the non-locaility result, with 30 degrees producing a 1 in 4 difference, and 60 degrees producing a 3 in 4 difference.

But do an EPR style experiment with 45 degrees and 90 degrees, and you get a 2 in 4 difference and a 4 in 4 difference, which could be explained by a hidden local variable model.

So the point is that although entanglement can produce strange outcomes, that doesn't mean that all experiements involving entanglement do.

The paper cited by Dr Chinese

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

is certainly worth a read. My initial thought was that the result could be explained locally by the split into four subsets, but it can't, or at least, if it can, I don't see it.

 

[Sylvia Else]

As to other experiments: you can delay the decision to entangle 2 particles until AFTER they are detected. This is done by the method called "entanglement swapping". Alice makes a decision to entangle (or not) a pair of photons sent to Bob. Those photons have never previously interacted, but will nonetheless violate a Bell inequality if they become entangled by Alice - and that can be done by Alice either before or after Bob sees them. Entangled pairs of photons will follow the cos^2(theta) prediction of QM, while un-entangled pairs will not.

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

Just some meandering thoughts from me: One would expect Bob's measurements on the pairs of photons to produce results that are random, and uncorrelated, regardless of what angles Bob chooses for his measurements. It's apparent that if Alice later decides to entangle her photons, the resulting Bell states indicate different correlations required of Bob's measurements on the subsets. But it is always possible for the subsets in combination to correspond to Bob's original random results. That is, if Alice causes the entanglement after Bob has performed the measurements, there is no need for any retroactive changing of Bob's results. However, the Bell state that Alice gets for each photon pair does appear to depend on the measurements that Bob made and the results that he got.

On the other hand, if Alice causes the entanglement before Bob performs the measurement then Bob must get results that, while still being random and uncorrelated overall, are consistent with the Bell states obtained by Alice.

It's all very strange.

Sylvia.