# Double slit experiment and entanglement [duplicate]

Just wondering, what would happen in this experiment.

In the experiment you would first have two entangled particles.

Then you fire one of the particles, lets say "Particle A", at a double slit towards a detector.

While in transit to the detector, what if the other entangled particle, lets call it "Particle B" was observed / had it's wave function collapsed?

Would "Particle A" still generate a wave-like interference pattern or would the wave function for both be collapsed?

In theory you cannot send classical data by entanglement, so this experiment must somehow fail, but I can't quite figure out why. If this experiment were to succeed, then you could read and send data about wave function states over entangled particles.

• See this paper for more details. – Trimok Jan 13 '14 at 20:25
• I asked a very similar question here and eventually posted an answer to my own question. – user1247 Jan 14 '14 at 1:35
• I see, got it. Yes that answers my question completely, thanks! – Aniseedwolf Jan 14 '14 at 15:53
• It always pays, when you're doing this sort of thought experiment, to write down an explicit example --- specify the state spaces, specify the initial joint state of the particles, specify the measurements you're going to make, and then work out exactly what happens. Usually a single example, fully worked out, will give you all the intuition you need to see what happens in general. – WillO Jun 21 '16 at 17:18

This experiment has been done, first by Birgit Dopfer in 1998, then later by Dr. John Cramer of the University Of WA. In Dopfer's experiment, there was a "coincidence detector" which is basically an AND gate to filter out only the entangled pairs. By moving the detector in the beam of photons not going to the double-slit, the information about the photon's momentum could be preserved or erased, affecting whether the interference pattern showed up or not.

However, this coincidence detector is a classical information channel. It prevents using entanglement to send information, unfortunately. Cramer tried several different methods to eliminate the coincidence detector and what he found was that there is an "anti-interference" pattern that drowns out the signal. I wonder what the properties of this anti-signal are and whether there's some way to get around it. The answer is almost certainly "no," but it's an interesting thing to study.