Postselected versus nonpostselected quantum teleportation using parametric down-conversion

Abstract

We study the experimental realisation of quantum teleportation as performed by Bouwmeester {\em et al}. [Nature {\bf 390}, 575 (1997)] and the adjustments to it suggested by Braunstein and Kimble [Nature {\bf 394}, 841 (1998)]. These suggestions include the employment of a detector cascade and a relative slow-down of one of the two down-converters. We show that coincidences between photon-pairs from parametric down-conversion automatically probe the non-Poissonian structure of these sources. Furthermore, we find that detector cascading is of limited use, and that modifying the relative strengths of the down-conversion efficiencies will increase the time of the experiment to the order of weeks. Our analysis therefore points to the benefits of single-photon detectors in non-post-selected type experiments, a technology currently requiring roughly $6^{\circ}$K operating conditions.