Photon detection in the EPR experiment

[kurt101]

In the photon version of the EPR experiment, how is the final polarization state of the photon detected?

I have read a number of high level descriptions of the EPR experiment, but I am having trouble with understanding the detection part.

Here is my understanding, please correct me where I am wrong.

I understand the type of polarizer used in the EPR experiment is a linear beam splitting polarizer as defined here https://en.wikipedia.org/wiki/Polarizer

If I understand correctly, there are some linear beam splitting polarizers that split the photon into 2 directions and there are some that do not. Just to be clear, I am not referring to the polarization state vector, but the direction the photon is traveling as it leaves the polarizer.

So I can imagine that for a beam splitting polarizer that does alter the direction of the photon, the detector would be a photon detector placed in the location of where the photon is expected to arrive for a given polarizer orientation.

Or I can imagine that for linear beam splitting polarizer that does NOT alter the direction of the photon, the detector would need to include another polarizer (absorption or beam splitting type) that has a fixed orientation relative to the EPR experiment in order to split the two states prior to being detected by a photon detector.

Are either of these ways of detecting the final polarization of the photon in the EPR experiment valid? Are there other ways?
Thanks

 

[vanhees71]

You should first read about classical crystal optics. In its full glory it's marvelously explained in the good old book by Sommerfeld:

A. Sommerfeld, Lectures on Theoretical Physics, vol. IV (optics)

Then you can think of this as an effective theory describing also the interaction of single photons with the optical elements like your polarizer.

 

[zonde]

If I understand correctly, there are some linear beam splitting polarizers that split the photon into 2 directions and there are some that do not. Just to be clear, I am not referring to the polarization state vector, but the direction the photon is traveling as it leaves the polarizer.

All polarizing beam splitters are splitting photon beam in two direction. Polarizing beam splitter with one output does not make sense. Polarizer has one beam at the output, but it absorbs part of the beam so that the beam is altered anyways.

So I can imagine that for a beam splitting polarizer that does alter the direction of the photon, the detector would be a photon detector placed in the location of where the photon is expected to arrive for a given polarizer orientation.

Yes

 

[vanhees71]

For an ideal polarizing beam splitter no light is absorbed, but split into two partial beams of different polarization:

https://en.wikipedia.org/wiki/Polarizer#Beam-splitting_polarizers