Faster than light communication through statistical analysis of entangled photons

[bascule]

I believe (or perhaps I should say I'm under the mistaken impression) that we can use statistical analysis of the collision points of entangled photons to send information faster than the speed of light.

 

To begin with, we need two light beams, produced by passing our source beam through a beam splitter, which is 50% reflective in one direction and 50% transparent in the other. These beams travel along for awhile and then are optically directed to interfere with each other, producing the classical interference pattern of the double slit experiment.

 

If we are to introduce a "which path detector" on one of the beams coming out of the beam splitter, trying to figure out which path a particular photon took, the interference pattern is destroyed. When the "which path detector" is off, we see the interference pattern, and when it's on, the interference pattern is destroyed and we see only the non-interfering beams.

 

We can place the "which path detector" immediately next to the optical devices causing the two beams to cross. It doesn't matter if the source and the destination + "which path detector" are a light year apart... as long as the "which path detector" is on the interference pattern will be destroyed.

 

So, let's complicate the experiment, by placing down converters which produce entangled photons on each side of the beam splitter, immediately next to where the photons are originally emitted.

 

Now rather than two beams coming out of the beam splitter, being directed at a single endpoint, we have four beams, two coming out of the beam splitter, and then subsequently passing through the two down converters, resulting in a total of 4 beams. The two down converters send parallel beams of entangled photons in opposite directions, and equidistant from the source are devices that direct the beams at each other, resulting in two copies of the interference pattern of the double slit experiment, generated by entangled photons.

 

Provided there is nothing causing waveform collapse besides the "screens" the two beams are being directed onto...

 

So my question is: what happens to the signal on either side if we place a "which path detector" on one of the beams on one of the endpoints? If we take a peek at just one of these four beams to see if a photon has decided to travel that direction, wouldn't the interference pattern be destroyed on both sides simultaneously, because the photons are entangled?

Edited September 27, 2009 by bascule

[uncool]

It will take time for the pattern to be destroyed - the interference pattern is a result of the light propagating. If you place your detector right next to the splitter, only after the appropriate interval will the pattern be destroyed. In the meantime, the person would be unable to detect the difference.

=Uncool-

[ydoaPs]

I believe (or perhaps I should say I'm under the mistaken impression) that we can use statistical analysis of the collision points of entangled photons to send information faster than the speed of light.

 

To begin with, we need two light beams, produced by passing our source beam through a beam splitter, which is 50% reflective in one direction and 50% transparent in the other. These beams travel along for awhile and then are optically directed to interfere with each other, producing the classical interference pattern of the double slit experiment.

 

If we are to introduce a "which path detector" on one of the beams coming out of the beam splitter, trying to figure out which path a particular photon took, the interference pattern is destroyed. When the "which path detector" is off, we see the interference pattern, and when it's on, the interference pattern is destroyed and we see only the non-interfering beams.

 

We can place the "which path detector" immediately next to the optical devices causing the two beams to cross. It doesn't matter if the source and the destination + "which path detector" are a light year apart... as long as the "which path detector" is on the interference pattern will be destroyed.

 

So, let's complicate the experiment, by placing down converters which produce entangled photons on each side of the beam splitter, immediately next to where the photons are originally emitted.

 

Now rather than two beams coming out of the beam splitter, being directed at a single endpoint, we have four beams, two coming out of the beam splitter, and then subsequently passing through the two down converters, resulting in a total of 4 beams. The two down converters send parallel beams of entangled photons in opposite directions, and equidistant from the source are devices that direct the beams at each other, resulting in two copies of the interference pattern of the double slit experiment, generated by entangled photons.

 

Provided there is nothing causing waveform collapse besides the "screens" the two beams are being directed onto...

 

So my question is: what happens to the signal on either side if we place a "which path detector" on one of the beams on one of the endpoints? If we take a peek at just one of these four beams to see if a photon has decided to travel that direction, wouldn't the interference pattern be destroyed on both sides simultaneously, because the photons are entangled?

Haven't we been through this already?

[bascule]

It will take time for the pattern to be destroyed - the interference pattern is a result of the light propagating. If you place your detector right next to the splitter, only after the appropriate interval will the pattern be destroyed. In the meantime, the person would be unable to detect the difference.

=Uncool-

 

The "which path detector" is placed next to one of the endpoints, not the beam splitter, the idea being that flipping it on will cause "spooky action" at a distance and collapse the photons on the other side, destroying the interference pattern immediately before the beams are caused to converge onto the screen.

 

Haven't we been through this already?

 

Yes, I suppose I could've revived that thread

 

I still don't get what the problem is.

---

Merged post follows:

Consecutive posts merged

And how about this:

 

Forget the big elaborate nonsensical description above.

 

The device just needs to duplicate the double slit experiment, using pairs of entangled photons. Something like this:

 

|<--------------------- (source) --------------------->|

 

Collapsing the interference pattern on one side should collapse the interference pattern on the other side, no?

 

The main problem I saw raised was that the photons would spontaneously collapse (due to interactions with various particles whizzing about in the "vacuum" of space, I guess?)

[swansont]

 

Forget the big elaborate nonsensical description above.

 

The device just needs to duplicate the double slit experiment, using pairs of entangled photons. Something like this:

 

|<--------------------- (source) --------------------->|

 

Collapsing the interference pattern on one side should collapse the interference pattern on the other side, no?

 

The main problem I saw raised was that the photons would spontaneously collapse (due to interactions with various particles whizzing about in the "vacuum" of space, I guess?)

 

I'm not seeing how knowing which path on one side gives you which path information on the other side.

[Klaynos]

I think the issue is bascule thinking that "which path" is or can be an entangled quantity, I don't think it can be.

[bascule]

I'm not seeing how knowing which path on one side gives you which path information on the other side.

 

If you attempted to detect which path/slit a particular photon was traveling through (on one side only), wouldn't you see a pattern more akin to this on both sides:

 

 

than something like:

 

Edited September 28, 2009 by bascule

[Klaynos]

If you attempted to detect which path/slit a particular photon was traveling through (on one side only), wouldn't you see a pattern more akin to this on both sides:

 

I don't think you will see it on both sides, which slit is not entangled, and I can't think of a way where it could be.

[bascule]

I don't think you will see it on both sides, which slit is not entangled, and I can't think of a way where it could be.

 

My proposed experiment would have a total of four slits.

 

To make it even simpler: how about a source that emits a single photon at a time, passed through a device (a down converter?) that produces pairs of entangled photons which leave the device traveling in opposite directions at a quasi-random angle, ala:

 

|<--------------------- (source) --------------------->|

 

the | things on either end actually consist of plates with two slits. Two plates, two slits, a total of four slits. Behind each of the plates is a material which provides us feedback as to where the photons are actually landing (by changing color or whatever) whenever they manage to pass through the two slits.

 

We allow the device to run for some time.

 

Would we not see this sort of distribution of photon "landing sites" over time occurring on both "sides" of the experiment:

 

http://gravityandlevity.files.wordpress.com/2009/05/double-slit_experiment_results-tanamura.jpg

[swansont]

My proposed experiment would have a total of four slits.

 

To make it even simpler: how about a source that emits a single photon at a time, passed through a device (a down converter?) that produces pairs of entangled photons which leave the device traveling in opposite directions at a quasi-random angle, ala:

 

|<--------------------- (source) --------------------->|

 

the | things on either end actually consist of plates with two slits. Two plates, two slits, a total of four slits. Behind each of the plates is a material which provides us feedback as to where the photons are actually landing (by changing color or whatever) whenever they manage to pass through the two slits.

 

We allow the device to run for some time.

 

Would we not see this sort of distribution of photon "landing sites" over time occurring on both "sides" of the experiment:

 

http://gravityandlevity.files.wordpress.com/2009/05/double-slit_experiment_results-tanamura.jpg

 

Yes. That's fine so far. But the polarization entanglement and the "which path" information are separate.

[bascule]

Yes. That's fine so far. But the polarization entanglement

 

Polarization entanglement?

 

...and the "which path" information are separate.

 

...well the "which path" part is the one I somewhat understand.

 

So, next question: let's say we place a "which path" detector on one of the four slits.

 

Would the interference patterns on both "screens" be destroyed, leaving us with two bars on either screen?

[swansont]

Polarization entanglement?

 

That's the property that is entangled in downconversion. The polarization of the photon.

 

 

...well the "which path" part is the one I somewhat understand.

 

So, next question: let's say we place a "which path" detector on one of the four slits.

 

Would the interference patterns on both "screens" be destroyed, leaving us with two bars on either screen?

 

 

No. There is no property of the photon that is entangled with the path.

[bascule]

That's the property that is entangled in downconversion. The polarization of the photon.

 

Okay, seems like my scenario is starting to unravel now... what other properties can be entangled?

 

No. There is no property of the photon that is entangled with the path.

 

So the result would be the interference pattern would be destroyed on the screen with the "which path detector" but would continue to appear on the other screen?

[swansont]

Okay, seems like my scenario is starting to unravel now... what other properties can be entangled?

 

Polarization of photons and spin of particles are two that have been used. Any relationship of (discrete) states, though, could conceivably be entangled. If your two photons could be made to be different energy, I think the wavelength could be entangled, e.g. a red and blue photon were emitted. They would be in a superposition of the red and blue state, and the photon color would be entangled.

 

So the result would be the interference pattern would be destroyed on the screen with the "which path detector" but would continue to appear on the other screen?

 

Right.

[bascule]

Okay, I think I finally have a clear understanding of why this doesn't work now. Thanks.