As far as persistence is concerned, the measurement technique will determine the longevity of the entanglement. Generally photons are measured destructively, as in, for example, measurement by the photoelectric effect. Polarization measurements, however, do not have to be destructive, so one has a choice. One can set polarizer to pass a photon with spin +1, and block spin -1. So, a +1 particle will still be in a +1 state after measurement, and a spin -1 photon will be destroyed by absorption. It all depends on the details of the measurement process.
If a singlet photon state gives a +1 along any axis, then the other photon will have spin -1 along the same axis, as is required by conservation of angular momentum. If the axes are different, then a measurement of 1 photon implies a probability distribution for the other measurement. A +1 measurewment in the z direction says that the other particle will be +1 50% of the time, -1 otherwise, if measured alnong the x axis. Further, the same is true for Sz = -1. So, as is required by the fact that Sx and Sy do not commute, a measurement of Sz does not specify the measurement for, say, Sx whether or not the two photons are entangled.
Klaynos and I pretty much agree.
Regards,
Reilly Atkinson
