General
The experiments involve producing two photons from an incident photon. A crystal converts in a very inefficient way one out of say 10^9 photons. But this is okay: a laser puts out 10^15 photons per second, so we get plenty. The process is called spontaneous parametric down-conversion. The photons are entangled in time, energy, momentum, spatial mode, and often polarization. That means that the parameters that define the properties of the photons are correlated. For example, by conservation of energy the sum of the photon’s energy is equal to the pump’s photon energy. Momentum is also conserved inside the crystal.
Once the photons are produced they go in general in separate but correlated directions. They do something, such as going through an interferometer or other optical elements, and reach their final destination: detectors. Not just any detector, single photon detectors, which detect individual photons with efficiency. Normal detectors (photodiodes or digital cameras) can be efficient, but for many photons, having a threshold for minimum detection of a few photons. Photomultipliers also do not work well because they are not efficient enough in the infrared. We need other accessories, such as filters to restrict the range of wavelength light going into the detectors, optical fibers to channel the light to the detectors, which are not on the same table as the optical elements. The setup takes a space of less than 2 meters. We use optical breadboards of dimensions 2-feet by 5-feet (or the metric equivalent) to place all the optical components. Then finally the signals from the detectors have to eventually reach the computer. The computer needs to get information about the photon pairs, so we need a “coincidence” circuit. This solves many problems. Room lights and detector dark counts are eliminated by this circuit, so life is good!
Essential components of the experiments are:
Procedures
Here we include some information on setting up and alignment. It contains our accumulated experience after many years of experience and Immersion workshops since 2011.
- Instructions for setting up the labs and equipment information. This is an old file that has been updated. More recent information in individual files may be presented more concisely. It still has a lot of useful information. File
- Step-by-step procedures to set up labs. These are procedures that we use in our yearly Alpha immersions. Enclosed are the latest ones from 2019, which give instructions to setup and align for 4-5 experiments. File
- Troubleshooting! The first time is tough because alignments are subtle and often we ignore an important component (It took us one year to get down-conversions! The culprit was a missing lens…). I often get queries about the initial setup. I am happy to help (email to egalvez@colgate.edu), but this file has some suggestions to get out of the dead end. File
Equipment Cost
The experiments are not cheap. It is not a single big ticket item, but there are many parts. Here are some considerations:
- The pump laser. This has to be short wavelength. Thanks to blu-ray players, we have plenty of gallium-nitride diode lasers that put out plenty light at a wavelength of 405 nm. Their price goes from $20 50-mW laser “pointers” (not to be used in talks!), up to about $300 for more reliability and current stability, and up even further to about $4000 for temperature stability.
- The detectors must be single-photon avalanche-photodiode detectors. They are almost exclusively made by one company (Perkin-Elmer/Excelitas) for about $4000. The good news is that Alpha can get ones for educational prices for about $1700 each.
- This file has an overview of price considerations when starting out: File
- This file has actual prices and order information. Prices are current as of early 2019. File
Calculations and other Information
Parametric down-conversion is a nonlinear process that exploits the birefringence of crystals. We use beta barium borate crystals for this purpose. There are commercial vendors that already have what we want (Newlight photonics), but if you want to order to other from vendors (there are many), you need to specify more. Here we present some calculations done in Mathematica and Mathcad (down-conversion, compensation for entanglement).
