Light, which is composed of discrete quanta, or photons, is one of the most fundamental concepts in physics. Being an elementary entity, the behaviour of photons is governed by the rules of quantum mechanics. The ability to create, manipulate and measure quantum states of light is not only useful in foundational tests of quantum theory, but also in a wide range of quantum technologies – which aim to utilize non-classical properties of quantum systems to perform tasks not possible with classical resources. Only recently has it been possible to control the properties of number states of light, which have a fixed photon-number. Two-photon states are central to testing fundamental physical theories (such as locality and reality) and the implementation of quantum information technologies. The versatility of photon-pair states is en- abled by the potential entanglement properties it can posses. Thus controlling the correlations between photons is crucial to both pure and applied physics. To produce a single photon, a photon-pair state can be used. Detection of one photon indicates its twin’s existence. Many applications, such as optical quantum computation, require pure indistinguishable single photons. Heralding single pho- tons from a photon-pair will, in general, produce single photons in a mixed quantum state due to correlations within the pair. A common approach to creating photon-pairs is through the nonlinear sponta- neous four-wave mixing interaction in optical fibres. This thesis presents a theoreti- cal and experimental implementation of a scheme to tailor the spectral correlations within the pairs. Emphasis is placed on engineering the two-photon state such that they are completely uncorrelated. Spatial entanglement is naturally avoided due to the discrete nature of the optical fibre modes. Spectral correlations are eliminated by careful choice of dispersion characteristics and conditions. The purity of the photons generated by this scheme is demonstrated by means of two-photon inter- ference from independent sources. We measure a purity of (85.9 ± 1.6)% with no spectral filtering, exhibiting the usefulness of this source for quantum technologies and applications.

Due to its inherent stability and compactness, integrated optics can allow for experimental complexity not currently achievable with bulk optics. This opens up the possibility for large-scale quantum technological applications, such as quantum communication networks and quantum information processing. Quantum information processing relies on efficient sources of entangled photon pairs. Most demonstrations in integrated photonics so far have featured the on-chip manipulation of photon states using a free-space bulk-optic source of photons. This has the drawback of introducing loss due to the spatial mode mismatch between waveguide modes of the chip and modes of the produced photons. In this way, loss limits the number of photons that are simultaneously carried in the integrated optical device, and thus limits the number of qubits. One way to avoid this loss is to generate the photons in another waveguide device. This can be done through, for example, spontaneous four-wave mixing (SFWM). In this third-order nonlinear process, two pump photons spontaneously scatter off each other to create two photons of two new frequencies, satisfying momentum and energy conservation. This has been studied in birefringent optical fibers and photonic crystal fibers. In this work, we investigate the SFWM generation of photons in a waveguide coupler comprised of two touching tapered optical fibers, which we call a microcoupler. The two silica fibers are kept in contact and tapered to be 1 micron in diameter in the 10 cm long uniform interaction region. This device has three main advantages over a standard telecom 2x2 fiber coupler. 1) The small mode area enhances the photon generation rate; 2) The microcoupler supports four modes which is the minimum number required for two-photon entanglement. So in principle the device should be able to produce polarization-entangled photon pairs; 3) The strong waveguide-waveguide coupling and waveguide dispersion (due to the tapering) forces the photons to be far in wavelength from the background light around the pump. We present the 28 allowed phasematching processes for the microcoupler, as well as predict the frequencies of the generated photons. We report the first experimental observation of photon pairs produced via SFWM in a microcoupler. We also analyze the polarization state of the observed photons to figure out which phasematching processes are responsible for generating the photons. We expect to observe more photon pairs in future devices, with the ultimate goal being the generation of polarization-entangled photon pairs for integrated optics.