Quantum Entanglement: Einstein's "Spooky Action" Explained

[thenewmans]

Quantum entanglement’s been in the news lately. Every once in a while, a friend points out to me that quantum communication proves quantum entanglement works and faster than light communication is possible. One even pointed it out in a Michio Kaku book. But I thought quantum physics (the theory) stopped breaking special relativity ever since the EPR paradox. All these articles love to quote Einstein’s “Spooky action at a distance”.

 

[DopplerDog]

You could think of entanglement as "spooky action at a distance", but it can't be used to transmit information. Quantum entanglement is a real phenomenon, but it won't let you transmit a message faster than light.

 

[thenewmans]

You could think of entanglement as "spooky action at a distance", but it can't be used to transmit information. Quantum entanglement is a real phenomenon, but it won't let you transmit a message faster than light.

It can't be used to transmit information? I thought that was the idea.

 

[Alfi]

"spooky action at a distance"

What kind of distance are we talking here?
What distance have experiments made it to these days, in the physical separation of entangled things?

 

[Count Iblis]

It can't be used to transmit information? I thought that was the idea.

Yes, but not faster than light. You can use entanglement to transmit data securely in a way that is not possible classically.

 

[thenewmans]

Alfi, Good question. I don't know but they're talking about quantum communications with the ISS.

http://technology.newscientist.com/...station-to-be-quantum-communications-hub.html

 

[DopplerDog]

Yes, but not faster than light. You can use entanglement to transmit data securely in a way that is not possible classically.

That's right. Imagine an entangled pair of electrons, with opposite spins. Your friend takes one to point B, and you take one to point A. They are entangled, so you don't know if yours points up or down, they are both in a state of superposition. You take yours, it could point up or down, you won't know until you measure it. What's more, it will be in a state of superposition until you measure it - so it can't be thought of pointing in a specific direction UNTIL you measure it.

When you measure yours, say it shows as pointing up. Your wavefunction collapses into the up position. Then your friend's will obviously point down, his wavefunction automatically collapsing into the down position.

The point here is that the entanglement works across the distance between A and B. Whatever you measure your spin to be, your friend's spin will automatically become the opposite, even if your friend is in Alpha Centauri, and measures his within a few seconds of yours.

Note that although a signal could not possibly be transmitted between the two measurements faster than light, your friend's electron "knows" which way to point. Note also that you can't use this to send information faster than light, as this transmits no information to your friend (he just knows that you measured your spin in the opposite direction... but he knew that would have been the case anyway).

You could take a third electron with a known state, entangle it with one of the pair, and take a measurement. You can then send that measurement across to your friend through a classical signal (no faster than light), and your friend can securely use this, and his member of the pair, to make a copy of that third electron. This allows secure communications between you and your friend (you are transmitting the state of the third electron), so here information IS being sent - except it is done no faster than light (as it must). The benefit of this is that it is totally secure (third parties can't use the intercepted signal to reproduce the state of the third electron, as they need one of the entangled pair to "decode" it).

 

[thenewmans]

Now let me get this straight. You’re saying that our particles are in superpositions but we can’t look or they won’t be in superpositions any longer. Instead, we have to wait until we’re 4 lightyears apart and low and behold, they jump to opposite spins. Who’s to say they weren’t in that spin right from the start? Hmm, that sounds more like a second rate magic trick.

 

[DopplerDog]

Now let me get this straight. You’re saying that our particles are in superpositions but we can’t look or they won’t be in superpositions any longer. Instead, we have to wait until we’re 4 lightyears apart and low and behold, they jump to opposite spins. Who’s to say they weren’t in that spin right from the start? Hmm, that sounds more like a second rate magic trick.

That's a normal objection to make when first faced with QM. After all, take two "classical" electrons with opposite spins, mix them up, and take them far apart - and surprise, surprise, they'll also point in different directions. That, in itself, is not a quantum effect.

The quantum nature of the electron can be demonstrated, however, by performing different experiments, such as the double slit interference experiment on electrons. This experiment categorically shows that particles such as the electron do exist in a state of superposition.

The quantum "communications" thought experiment I mentioned takes for granted that the electron is in a state of superposition. It's not meant to be taken as evidence of this. There are other experiments that can do that.

By the way, that thought experiment was lifted straight out of Penrose's "Road To Reality", if anyone's interested.

 

[Degeneration]

How does causality play into this? I've heard it solves entanglement somehow

 

[DopplerDog]

How does causality play into this? I've heard it solves entanglement somehow

If you look at the comms thought experiment I mentioned, you'll see that both participants can take measurements of spin within seconds of each other and have their results consistent - in spite of the fact that they are separated by a distance that light could not cover in that time.

The interval between both measurements is space-like, meaning that some observers see A taking the measurement before B, and some observers see B taking the measurement before A. So whether A is before B or B before A is observer dependent. Hence, it cannot be said that A's measurement "causes" B's, nor that B's "causes" A's.

If A's electron were to somehow "tell" B's electron where to point via some sort of unknown classical signal, it would create a causality problem, because for some observers, B's measurement takes place before A's.

So causality requires that there cannot be an unknown classical signal sent between the two electrons to explain the phenomenon. Given that the electrons are not in a predetermined state but are in a superimposed state, this means entanglement is a quantum phenomenon that cannot be explained through classical means.

 

[peter0302]

Entanglement alone can't be used to transmit useful information for the simple reason that whatever IS exchanged between the two particles (if anything) is not apparent until the results of experiments on the respective particles are compared. And the only way to compare the results is to use slower-than-light communication. Before the comparison, the results still appear totally random, as though there was no entanglement.

Put another way - there's no evidence of entanglement until the particles are compared. We can't compare the particles FTL, so obviously we can't use entanglement to communicate FTL.

 

[moving_on]

A quick question:
In most places, the 'no communication' theorem is quoted as failing as a proof once 'time evolution' occurs. With relation to the usual scenario whereby one tries to send superluminal messages, what constitutes 'time evolution'?
How much 'time' must pass? Enough time for a sub-luminal signal to be sent?
What guarantees 'no communication' if 'some' time passes, but not enough for a sub-luminal message to be received?
I notice that in the experiments used to confirm the 'no communication' theorem that both entangled photons are sent through rows of polarisers set at random angles (even changing during their flight). Surely for any party to use more than one polariser constitutes 'time' passing? What was the point of that? I mean, o.k., no coincedences were observed (so I won't be going back in time to kill my grandfather before I was born...), but we haven't explained why not, as the theory being tested isn't designed to work in this scenario!

 

[Landru]

So the wave randomly colapses up or down and therefor it's useless as a binary communication scheme; so if technology found a way to intentionaly make the them colapse up or down then could it be used as communication? Is that necessarily impossible?

 

[LaserMind]

IMO - Think of two entangled particles as belonging to the same wavepacket (a combined packet), - even though they are spatially separated and it 'looks like' two wave packets.

When this wavepacket collapses - from whichever 'end' - it does so instantly across the whole packet - i.e. a message is not sent from one end of the packet to the other end. The two ends of the packet (with the particles) take up there complimentary states instantly as defined in the combined Hilbert Space tensor product wave equation.

A further aid to understanding is to think of a large wave packet containing one particle.
When it is captured by a detector, the whole packet collapses instantly. SO the outer part that has only say, 1% probability, does not hang around once the particle is captured. There is nothing there anyway - except a 1% probabilty (i.e. not a 'thing').

-IMO (this is 'wave function collapse' theory & not everyones choice, but seems to solve
practical issues).

 

[moving_on]

o.k. Can I run a specific example here?
Can't draw very well in this but I think I can do it without having to attach a proper diagram...
... a statistically relevant (?) number of entangled photons are sent through the paths A and B below: ( | -- and / represent polarising filters at 'vertical', 'horizontal' and 45 degrees)

Time x Time x+1 Time x+2

...|.......--..... D (detector) (path for photon stream A)

...../....... (no detector) (path for stream B)

What would one expect to detect at D?
Nothing at all ever? If so why does the filter at B make no difference?
I thought about the 'no communication theorem' but it seems to make no sense in this context, or am I missing something?
If the filter was at the relative position in the A path, one would expect to see
some detections some of the time, so why not with it at at B?

 

[hdsncts]

The point here is that the entanglement works across the distance between A and B. Whatever you measure your spin to be, your friend's spin will automatically become the opposite, even if your friend is in Alpha Centauri, and measures his within a few seconds of yours.

I don't understand this. How can his spin "become" something?

I mean, think of this: a friend of yours has two playing cards, a queen and a king. The fact that one is a king and the other is a queen is a reality. The type of card is an intrinsic property. So now, the cards are turned over and you take one at random, go halfway across the country and look at your card. If you see that it is a king, then you know that your friend still has the queen. But his card didn't "become" a queen, you have simply been able to logically conclude that his is a queen because yours is a king, and there was one king and one queen to begin with.

Isn't the spin of an electron also an intrinsic property? It just seems to me that when you measure one spin, you can rationally conclude that the other "entangled" electron is the opposite spin.

Do I have this completely wrong? This is hard for me to comprehend.

 

[Landru]

I don't understand this. How can his spin "become" something?

I mean, think of this: a friend of yours has two playing cards, a queen and a king. The fact that one is a king and the other is a queen is a reality. The type of card is an intrinsic property. So now, the cards are turned over and you take one at random, go halfway across the country and look at your card. If you see that it is a king, then you know that your friend still has the queen. But his card didn't "become" a queen, you have simply been able to logically conclude that his is a queen because yours is a king, and there was one king and one queen to begin with.

Isn't the spin of an electron also an intrinsic property? It just seems to me that when you measure one spin, you can rationally conclude that the other "entangled" electron is the opposite spin.

Do I have this completely wrong? This is hard for me to comprehend.

I think they aren't sure whether the hypothetical cards leave with determined outcomes or if the outcome is determined at the time of measuring.

 

[LaserMind]

In path A the filters are orthogonal so there is zero probability
of transmission, whereas in B there is a chance of transmission
for both vertical and horizontal states.

 

[KJK]

I don't understand this. How can his spin "become" something?

I mean, think of this: a friend of yours has two playing cards, a queen and a king. The fact that one is a king and the other is a queen is a reality. The type of card is an intrinsic property. So now, the cards are turned over and you take one at random, go halfway across the country and look at your card. If you see that it is a king, then you know that your friend still has the queen. But his card didn't "become" a queen, you have simply been able to logically conclude that his is a queen because yours is a king, and there was one king and one queen to begin with.

Isn't the spin of an electron also an intrinsic property? It just seems to me that when you measure one spin, you can rationally conclude that the other "entangled" electron is the opposite spin.

Do I have this completely wrong? This is hard for me to comprehend.

Well, as I have understood, in QM theory this the example would go like this:

The cards are in superposition (a wave containing both realities), so you actually have both a king/queen in the hand until you look (50% probability for both). When you look at your card you randomly choose (wavefunction collapse/decoherence) a timeline (reality) where you had either a King/Queen, and then you intrinsicly know what the other guy had.

Also, as I have understood more recent theory of decoherence implies that the choosing of timeline isn't actually wavefunction collapse caused by the observer but due to decoherence you get entagled to the card and thus continue living in that reality, but both realities continue to exist in different universes... (multiverse?)

Is this right?

Anyway, I'm no physisict, just interested in the subject, and still have some reservations against the experiments that have proven entaglement. Need to read more about those...

Very interesting discussion in this forum :)

 

[pallidin]

Why would the cards be in superposition? They have no quantum entanglement to provide for that, in my opinion.

 

[hdsncts]

Why would the cards be in superposition? They have no quantum entanglement to provide for that, in my opinion.

Right... I used a macroscopic set of objects as an example to try and explain how I don't understand. I think he was just going along with my metaphor that the cards are basically an electron singlet.

 

[DrChinese]

Isn't the spin of an electron also an intrinsic property?

That it has a spin, yes. What is its value with respect to a particular axis: this is questionable.

 

[KJK]

Right... I used a macroscopic set of objects as an example to try and explain how I don't understand. I think he was just going along with my metaphor that the cards are basically an electron singlet.

Yes, just a metaphor, the macroscopic cards are of course not in superposition. :)

 

[Landru]

What is the post-hypothesis scientific support of quantum probabilty suggesting that it is a reality and that outcomes are not determined by undiscovered variables? This is another one of those assertions I see everywhere on the interweb for which I can't find scientific support.

 

[vanesch]

See the thread on "entanglement" https://www.physicsforums.com/showthread.php?t=246314
or just any thread where one talks about "EPR" or "Bell theorem" or the like.

 

[vanesch]

IMO - Think of two entangled particles as belonging to the same wavepacket (a combined packet), - even though they are spatially separated and it 'looks like' two wave packets.

When this wavepacket collapses - from whichever 'end' - it does so instantly across the whole packet - i.e. a message is not sent from one end of the packet to the other end. The two ends of the packet (with the particles) take up there complimentary states instantly as defined in the combined Hilbert Space tensor product wave equation.

A further aid to understanding is to think of a large wave packet containing one particle.
When it is captured by a detector, the whole packet collapses instantly. SO the outer part that has only say, 1% probability, does not hang around once the particle is captured. There is nothing there anyway - except a 1% probabilty (i.e. not a 'thing').

-IMO (this is 'wave function collapse' theory & not everyones choice, but seems to solve
practical issues).

What you describe is indeed "what happens on paper" when you do the practical quantum calculation of an entanglement experiment a la Aspect. And the whole discussion is about how it happens "in nature" and not "on paper".

 

[wawenspop]

To quote Alain Aspect "In some sense, both photons keep in contact through space and time".
That 'contact' he is referring to, relates to the fact that the entangled photons belong to the same 'wave packet' even with a large separation between the two different parts of the packet.

So it could be like, say, two wave packets connected by a 'thin string', -of any length -and the combination behaves as if the two particles were 'on top' of each other, exactly as they were 'at birth'. In other words, the spatial separation is not recognised by our two entangled particles and so the long string is superfluous.

Don't forget that the number of dimensions in the Hilbert product space of the two particles rises from 3 to 6. In multiple entanglements the number of dimensions becomes too large to think about.

 

[moving_on]

Lasermind,
is the following scenario not 'possible':
a pair of entangled photons emerge and begin their journey down their routes.
In terms of 'time' the first filter either of them visits is the vertical filter in the first
path (by the photon in the first path). There is a probability that the photon in this path will pass the filter. Imagine it does.
Hence it (and by proxy) it's entangled twin will attain a 'vertical' polarisation probability
function (i.e. will pass a vertical filter if it is met, and have a 50% chance of passing
a filter at 45 degrees).
In terms of 'time' the next filter that either of them visits is, indeed the 45 degree
filter in the second path (by the photon in the second path).
There is, therefore a 50% chance that this photon will pass this filter.
Imagine it duly does. It AND it's entangled twin will attain a 45 degree polarisation
probability function (i.e. will pass a 45 degree filter if it is met, and have a 50%
chance of passing EITHER a vertical or horizontal filter.
In terms of 'time' the third filter that is met by either photon is the horizontal filter
in the first path (by the photon in the first path).
There is a 50% chance that this photon will pass this filter and may be detected.

I agree that the simple solution seems at first likely,
but once I got this into my mind I couldn't resolve it.
Have I made an error here, and if so, where?

 

[DrChinese]

Lasermind,
is the following scenario not 'possible':
a pair of entangled photons emerge and begin their journey down their routes.
In terms of 'time' the first filter either of them visits is the vertical filter in the first
path (by the photon in the first path). There is a probability that the photon in this path will pass the filter. Imagine it does.
Hence it (and by proxy) it's entangled twin will attain a 'vertical' polarisation probability
function (i.e. will pass a vertical filter if it is met, and have a 50% chance of passing
a filter at 45 degrees).
In terms of 'time' the next filter that either of them visits is, indeed the 45 degree
filter in the second path (by the photon in the second path).
There is, therefore a 50% chance that this photon will pass this filter.
Imagine it duly does. It AND it's entangled twin will attain a 45 degree polarisation
probability function (i.e. will pass a 45 degree filter if it is met, and have a 50%
chance of passing EITHER a vertical or horizontal filter.
In terms of 'time' the third filter that is met by either photon is the horizontal filter
in the first path (by the photon in the first path).
There is a 50% chance that this photon will pass this filter and may be detected.

I agree that the simple solution seems at first likely,
but once I got this into my mind I couldn't resolve it.
Have I made an error here, and if so, where?

Of course, the particles do NOT remain entangled through filter after filter. Once either is observed, their entanglement ends (as to the that observable).