Controlling Spin of an Entangled Particle, Causality

[piezas]

Everything I've read has said that causality in the universe isn't violated because entangled particles don't transmit information (because measured values are random compliments). All of this spoke about measurement.

If entangled atoms are created (such as those 2006 on), can properties such as spin be directly controlled on one of the atoms (and therefore measured on the other)? I thought spin could be influenced by fields.

Thanks,

 

[xepma]

This process only happens locally. Suppose you start out with the (unnormalized) entangled spin-state:

$$|++\rangle + |--\rangle$$

So the if the first particle is measured to have spin up the second will also have spin up. Using a magnetic field we flip the spin of the first particle. This is not a measurement, so we flip the spin without even knowing "what it is". The resulting state is:

$$|-+\rangle + |+-\rangle$$

So we still have entanglement, but the correlation between the two particles changed. This time, when we measure the first particle to have spin up, the second one will have spin down (and vice versa). The second particle is not influenced - only the correlation between the two particles has changed.

 

[alexepascual]

This process only happens locally. Suppose you start out with the (unnormalized) entangled spin-state:

$$|++\rangle + |--\rangle$$

So the if the first particle is measured to have spin up the second will also have spin up. Using a magnetic field we flip the spin of the first particle. This is not a measurement, so we flip the spin without even knowing "what it is". The resulting state is:

$$|-+\rangle + |+-\rangle$$

So we still have entanglement, but the correlation between the two particles changed. This time, when we measure the first particle to have spin up, the second one will have spin down (and vice versa). The second particle is not influenced - only the correlation between the two particles has changed.

I guess if you flip the state of the first particle to the oposite of what it was without knowing what it was and wothout knowing what you get, then, as there is no measurement, entanglement is preserved (this is equivalent to what you said).
But if you forced the spin of the first particle to be spin up only, then the entanglement would be destroyed. Of course you can't do this with just a magnetic field, and this does imply measurement.
The result is again that you can't use this for communiation.
But there are some experiments that have been proposed which could be interpreted as allowing for communication between far away locations using entanglement. I think the reason why these schemes are being rejected is that they seem to violate causality. So they present a paradox. Maybe most physicists are a little wary of introducing new paradoxes. So they prefer to reject these ideas using as the main argument that they violate causality, and they try to predict results for those experiments that won't violate causality.
One of the first experiments in this respect that showed some curious features was probably a delayed-choice thought experiment presente by Wheeler. In this experiment, a photon form a distant galaxy may go on either side (or both as a wave) of some large mass that produces gravitational lensing before reaching Earth. At the detection point we may decide to measure which-way information by foucusing telescopes on each direction of arrival, or to take the measurement in suh a way that which-way information is lost. Depending on our choice, the photon went on both sides of the gravitational lens or on one side. Wheeler notes that this would imply changing something that is very far into the past and that both paths are separated by a huge distance. Of course this experiment would have some characteristics that would make it almost impossible to perform. But the main characteristic that makes it not extremely interesting is that we are talking about photons, which are 100% quantum things.
The interesting thing would be to run into a paradox like this where what you are "changing" is some macroscopic event, such as the result of measurement.
Birgit Dopfer came up with such an experiment. This experiment consisted of two entangled photons that are produced by type-I spontaneous parametric down conversion (SPDC).
These photons have their momenta entangled such that their sum always adds up to the same number. Each photon goes along one arm of the apparatus. In the shorter arm, the photon goes through a double slit and then hits a detector. On the longer arm there is a lens and a detector that can be positioned in such a way that it can be determined through which slit the first photon went through or otherwise it could be positioned out of focus so that this information is lost. The photons are sent one at a time, but after a few photons have been sent (which can be done in a relatively short time), an interference pattern emerges or not on each side of the setup depending on the position of the detector in the longer arm. Birgit Dopfer did the experiment and it worked as described. But there are some loopholes that could give rise to that behavior and which might prevent it to work at long distances. If it did work, information could be transferred by having an interference pattern represent a "1" and no interference represent a "0". The paradox that this experiment presents is that tyou could send messages faster than light and even into the past.
This experiment was analized by Zeilinger and also by Cramer and their conclusions were very different.
Ray Jensen came up with a thought experiment for faster-than-light communication based on the Dopfer experiment that uses polarization-entangled photons. As far as I know he was not able to publish his paper.
Cramer has proposed several modifications of the Dopfer experiment and he was working on a new version of the experiment but there are no news about his progress (as far as I know).
I think this subject is still controversial. Although entanglement has been known for quite a while, perhaps we still don't understand it in depth.

 

[piezas]

Thank you both for the help. I am always amazed at the helpfulness of the scientific community in sharing information. I'm recently getting back into the science world. And the fundamental quantum nature of space and matter is more interesting to me than ever now.

Do either of you have a reference as to the fundamental difference between what mechanically constitutes measurement (that doesn't destroy entanglement) and what constitutes control (forcing a spin up or down). I was under the impression before that both decohere the behaviour. I didn't know control had another aspect that destroyed entanglement.

It's funny. I was thinking of something like that position/momentum entangled photon case. I'll look into the experiment. Thanks. Entanglement is so mysterious and I imagine it will take a while to generate a unified model. Even if information can never be transferred faster than light, the underlying mechanism of instantaneous affect between non localized space is amazing.

I was reading a paper regarding quantum dot cellular automata in which cells were created with single Silicon atoms one nanometer in size. The automatic signal propogation in the system is brilliant. It almost seems as if there were a way to entangle atoms to drive the system and receive output, you could develop QCA chips that would require no power and require no connection. Too good to be true. Remote, infinitely vast computational power.

 

[piezas]

BTW, I've been looking for the Birgit Dopfer paper. It seems to have been removed. And googling around results in only news articles reporting the results. Has the thesis been discredited in some way?

 

[alexepascual]

BTW, I've been looking for the Birgit Dopfer paper. It seems to have been removed. And googling around results in only news articles reporting the results. Has the thesis been discredited in some way?

No. I don't think it has been discredited. She did the experiment and even if her interpretation of it were questioned, I don't think the experiment itself would.
Her paper was maintained at the University of Viena website by Zeilinger. But now for some reason it has been removed. I have to warn you that the paper is in German. I don't speak German but I was also interested in looking at it so I asked Birgit for a copy. (This was the day before yesterday). So let's wait and see. If she emails it to me I can send you a copy.
For the moment you can start by reading Anton Zeilinger's 1999 paper "Experiment and the Foundations of Quantum Physics" (I think it is in Zeilinger's university webpage).
You may also want to look at the website: http://www.paulfriedlander.com/text/timetravel/experiment
A link to Ray Jensen's paper is at:
https://www.physicsforums.com/showthread.php?t=61181

 

[piezas]

Thanks, again. I'll check them out. I've been look at various light traps to see how long it takes a signal to degrade. And if the reflection interaction destroys the entanglement. Seems that it could be useful in boosting the delay above error.

 

[alexepascual]

Do either of you have a reference as to the fundamental difference between what mechanically constitutes measurement (that doesn't destroy entanglement) and what constitutes control (forcing a spin up or down). I was under the impression before that both decohere the behaviour. I didn't know control had another aspect that destroyed entanglement.

Piezas, I can't perhaps give you the formal definition, but I can tell you the way I see it.
After either measurement of forcing (my terminology), you know the result, therefore there was "collapse" or "decoherence" (whichever you prefer) and the process is irreversible.
The only difference would be that in measurement you are selecting one of the eigenstates at random. The only influence of measurement would be collapsing the state to one of those eigenstates, and if a particle you are measuring is entangled, then your result is going to be correlated with the remote result.
If you force your particle to a particular eigenstate, then you are disturbing the original correlation between the entangled particles. You can foce your particle to be spin-down and that does not mean that the other partile is going to flip up in response to this.
The way that I see it is that the correlations were formed when the particles were interating, and they will keep that correlation as long as you don't disturb them.
You can see the entanglement as a bunch of sticks held together in the middle by a rubber band and somewhat spread appart at the ends. the end of each stick on one side represents one of the possible results of measuremment on one particle, and the other ends represent the possible result of measurement on the other particle. Each stick represents the correlation between the measurements represented by it's ends. (It would be easier to draw a picture). Of course this picture would not be very adequate if we choose different basis (as for example different polarizer orientation) between both particles.
If you make a measurement, you are just picking the end of one stick at random. Now, if you force it it would be more like picking the end of the stick and moving it, but here the stick example does not work very good either. You would have to think about the stiks as very fragile so that when you force one end you break it in the middle. The correlation is destroyed. If this was not the case, we would have had already experiments designed to send information using the "forcing" on one side. The only place where I have seen such experiments suggested has been in this forum, but I have not seen such a thing suggested in the literature, so I would assume that those who were proposing it had a wrong idea about entanglement.

 

[alexepascual]

Thanks, again. I'll check them out. I've been look at various light traps to see how long it takes a signal to degrade. And if the reflection interaction destroys the entanglement. Seems that it could be useful in boosting the delay above error.

Reflection does not destroy entanglement. Measurement of the eigenstate that is entangled does not appear to destroy it either, as you can still expect the other side to yield the correlated value.

 

[piezas]

BTW, looking at the extensions of the Dopfer experiment I couldn't see why observing the photon on 'the short arm' wouldn't break down the entanglement of the photon on the long arm before it is observed. It seems that if the lens position is shifted to a length longer than that of the short arm to the screen, the difference is explainable by the entanglement breaking on the short arm. I may be reading it wrong.

 

[DrChinese]

BTW, looking at the extensions of the Dopfer experiment I couldn't see why observing the photon on 'the short arm' wouldn't break down the entanglement of the photon on the long arm before it is observed. It seems that if the lens position is shifted to a length longer than that of the short arm to the screen, the difference is explainable by the entanglement breaking on the short arm. I may be reading it wrong.

"Breaking the entanglement" is a relative action because it happens separately on each particle (at least as far as can be objectively determined). Clearly you can lengthen or shorten the paths, and that makes no difference to the outcomes.

 

[alexepascual]

BTW, looking at the extensions of the Dopfer experiment I couldn't see why observing the photon on 'the short arm' wouldn't break down the entanglement of the photon on the long arm before it is observed. It seems that if the lens position is shifted to a length longer than that of the short arm to the screen, the difference is explainable by the entanglement breaking on the short arm. I may be reading it wrong.

I think that observing one of two entangled photons does not break the entanglement. It just causes collapse of the wavefunction for that photon. The only way you may ever verify that there is such a thing as entanglement between two photons is by making measurements on each and after comparing the results finding that there is a correlation. You can visualize this easily with an EPR-type setup that involves two spin-1/2 particles. If you always find that when you measure the second particle, it's spin is the opposite than that of the first, then there is entanglement. You can measure the spins with a Stern-Gerlach apparatus on each end. The Stern-Gerlach apparatus does not force the spin to be up or down, it just picks one of the two possible values from the superposition.
I think you are confusing destruction of entanglement with collapse of the wave function.
In the combined Hilbert space of the two particles, you get a collapse to one of the two possible spin combinations. But these combination still complies with the correlation determined by the entanglement. Now, going back to Dopfer's experiment. Measurement at D2 is done in such a way that you don't know by which slit the photon passed. So we could say that (ignoring whatever happens to the other photon) you have a superposition of states, each corresponding to the photon going through each slit. If, on the other hand you set D2 in such a way that you can detect which slit the photon went through, then you are
collapsing the superposition. If that is the case, then the other photon will also be collapsed to having gone through one of the two slits but not both. But this is not the way the Dopfer's experiment is constructed, in that experiment, you don't detect which-way information on D2.
The function of the lens in arm 1, is a little tricky. Paul Friedlander explains quite well how that works. So if you position detector D1 in the correct place, you can find out which slit the first photon went through. So this apears as a contradiction, and it is here where different people explain it differently.
Cramer says that if arm 1 is very long, then you can effectively choose after the fact, which slit the first photon went through. So this would be "retrocausality". According to him, if you position D1 in such a way that you don't detect which-way information, then you'll see an interference pattern in D2. (Of course interference patterns appear only after a few photons have been detected). I personaly doubt it that retrocausality could be the cause for this behavior.
Zeilinger on the other hand, says that because it is possible to eventually determine which-way information using the photon on arm 1, then you'll never see an interference pattern in D2. I don't find this argument very convincing either.
Those who say that you would never see an interference pattern in D2 explain the fact that you do in fact see it in Dopfer's experiment, by saying that the coincidence counter is what makes the pattern appear. I haven't see any convincing evidence that this is the case.
I do think that in Dopfer's experiment the difference in length between the arms is not enough and that could be a loophole, as signals traveling at the speed of light could travell from one detector to the other before each measurement is finished. (it takes a finite time for each detector to "click")

 

[piezas]

I don't know if it's completely clear. Has there been experimental verification? I've only found people looking to reproduce the experiment, the most notable saying that it probably wouldn't work (Cramer). It seems like an easy enough experiment to reproduce.

 

[DrChinese]

I don't know if it's completely clear. Has there been experimental verification?

http://arxiv.org/abs/quant-ph/0106078 A Double slit quantum eraser is pretty close.

 

[alexepascual]

http://arxiv.org/abs/quant-ph/0106078 A Double slit quantum eraser is pretty close.

I think the Walborn experiment is too different. It uses type-II SPDC and deals with filtering of the polarization states at each of the slits. This complicates things a lot.
Also, I have noticed that there are a lot of people saying that you need the coincidence counter in order to see interference. And then they point to the Walborn experiment. In the Walborn experiment it is clear that you need coincidence counting, but I don't think the same argument applies to Dopfer.
I would not doubt the validity of Birgit's results. But I would think about which factors may invalidate an interpretation that switching on one side having an effect on the other side represents signaling faster than light or retrocausality.
I think that, regardless of possible results of improved versions of this experiment, as a thought experiment it is great to think about some underlying mechanism that may be the cause of the reported behavior.

One simple question would be: You move D1 out of the way and let the photons travel toward a far-away galaxy. (Assume little chance of being absobed by interstellar gas in the near future). Would you see interference at D2 in this case?
According to Zeilinger you would not because you could in principle use the photons in space to find which-way information about the ones that hit D2.
What do you think about this argument?

 

[piezas]

Something to ponder. I think I'd have something more confident to say on that once I understand more regarding the equations that describe entanglement and create etanglement. It hasn't clicked yet (but that seems to be common in QM).

One thing has been popping lately that seems very interesting and perhaps adds information. Has anyone yet to entangle particles that are not local to being with? It always seems that the particles need to interact in some fashion and that they can't be forced to entangle when already distant. I don't know if that's a technological hurdle or a fundamental one (again, still trying to understand the equations).

 

[DrChinese]

One thing has been popping lately that seems very interesting and perhaps adds information. Has anyone yet to entangle particles that are not local to being with? It always seems that the particles need to interact in some fashion and that they can't be forced to entangle when already distant. I don't know if that's a technological hurdle or a fundamental one (again, still trying to understand the equations).

This has been done too.

http://arxiv.org/abs/quant-ph/0609135

 

[piezas]

dear lord, the universe is weirder than i imagined. it will be an amazing thing when the underlying mechanisms for this are unwound.