Efficient quantum optical state engineering and applications

Over a century after the modern prediction of the existence of individual particles of light by Albert Einstein, a reliable source of this simple quantum state of one photon does not exist. While common light sources such as a light bulb, LED, or laser can produce a pulse of light with an average of one photon, there is (currently) no way of knowing the number of photons in that pulse without first absorbing (and thereby destroying) them. Spontaneous parametric down-conversion, a process in which one high-energy photon splits into two lower-energy photons, allows us to prepare a single-photon state by detecting one of the photons, which then heralds the existence of its twin. This process has been the workhorse of quantum optics, allowing demonstrations of a myriad of quantum processes and protocols, such as entanglement, cryptography, superdense coding, teleportation, and simple quantum computing demonstrations. All of these processes would benefit from better engineering of the underlying down-conversion process, but despite significant effort (both theoretical and experimental), optimization of this process is ongoing. The focus of this work is to optimize certain aspects of a down-conversion source, and then use this tool in novel experiments not otherwise feasible. Specifically, the goal is to optimize the heralding efficiency of the down-conversion photons, i.e., the probability that if one photon is detected, the other photon is also detected. This source is then applied to two experiments (a single-photon source, and a quantum cryptography implementation), and the detailed theory of an additional application (a source of Fock states and path-entangled states, called N00N states) is discussed, along with some other possible applications.