Quantum Entanglement, Dark Counts, Coincidence Detection

[mindwalker_i]

20. That's a log of really good information!

Thank you for your reply, this is exactly the kind of stuff I need to think about.

What Dr. Cramer is trying to do and I'm trying to follow is to eventually get to the point of not using a coincidence circuit. The idea is that normally, all photons going into the Mach-Zehnder interferometer generate nice interference patterns - some light spots, some dark spots. However, when the momentum-superposition of the entangled photons is destroyed by making a "measurement" on the beam not goign through the interferometer, the entangled photons no longer make an interference pattern. That means some of them will hit the parts of the pattern that were dark. Cramer says that the dark count is too much for him to be able to tell when the entangled photons are hitting those spots. Hence, I'm really worrying about dark count.

From what you said above, filter and such are big culprits in destroying photons prematurely. I imagine that by absorbing one photon of a pair, that's going to have an amplified effect of coincidence detection. I was kind of thinking that I would filter out the "degenerate" photons with a block of opaque material with holes drilled to just get those parts of the cones - with holes as big as the diameter of the pump beam. My thought process is that the pump goes into the BBOs (yes, I'll do a two-crystal type 1 SPDC), and all across the area of the cross-section of the pump beam, downconverted photons will be created and will exit with angles relative to the spot where they were downconverted. Actually, that beam should be slightly larger than the pump since downconversion can happen anywhere in the thickness of the crystals. However, once downconverted, they should all be going in the same direction - for all degenerate photons. Having another block-and-hole further away, or a tube, would filter out just those photons (like 99% would just be the degenerate photons), I would think.

As for fiber-coupling the detectors, I was thinking I wouldn't get that option but am rethinking it on your advice. The critical concern is injection into the fiber: the bare detectors have 180um, which is a good size relative to the beam to pick out just parts of the interference pattern. However, I hope to tune the interferometer so that there's just one interference fringe - on one port of the interferometer, it will be light in the middle, the other will be dark in the middle. If I could do that and inject most of the area from the dark-center output, then when the entangled photons are kicked out of superposition, I'll capture more of them if I can inject most of the beam into the fiber. Lenses would help, but do I also need to collimate it to get injection?

In order to measure wavelength of the laser, my first thought is to use a double-slit to make an interference pattern and measure the distance between fringes. Then wavelength can be back-calculated. I'm sure there's a better way, and probably more accurate. Cramer is temperature-controlling his laser to get it to exactly 405nm. That might be a possibility, or I could adjust the angle of the crystals as you say. I think maintaining a constant temperature on the diode would be critical.

I see experimental setups using a half-wave plate. What is this for? It adjusts the phase of the light, so is it necessary to have a certain phase when it hits the BBO?

Again, thank you very much for the information! It gives me a lot of ideas on how to filter, or not filter, and things that might help detection. The idea of being able to use a visible beam for alignment is really good - possibly I could use a visible laser and put a mirror in place to reflect that beam into the interferometer, or turn it to reflect the downconvered beam into the interferometer.

You rock!