We experimentally and theoretically study Bragg diffraction as a tool for large-momentum transfer beam splitters in atom interferometry. A theoretical framework is developed to quantify the diffraction phase systematic caused by Bragg diffraction and experiments are performed to confirm these predictions using a Ramsey-Bord\\'e atom interferometer. We then develop methods to systematically cancel and reduce the diffraction phase systematic by carefully selecting Bragg diffraction parameters and utilizing Bloch oscillations. These techniques are then applied to an ongoing precision measurement of $h/m_\\text{Cs}$ for cesium, with the end goal of measuring the fine structure constant
alpha. We demonstrate a high contrast simultaneous conjugate Ramsey-Bord\\'e interferometer using 5th order Bragg diffraction and 25 common mode Bloch oscillations which achieves
2.5times106 radians of phase. We also demonstrate an interferometer with a statistical uncertainty of
deltaalpha/alpha=0.25 ppb after 25 hours of integration time that has diffraction phase systematic error of around 1 ppb. Other sources of systematic uncertainty are also thoroughly explored and determined to better than 0.1 ppb. The techniques and theories developed in this thesis will hopefully help enable future precision measurements based on Bragg diffraction.
The transition efficiency of atomic Bragg diffraction as mirrors and beam splitters in Bragg atom interferometers plays an essential role in impacting the fringe contrast and measurement sensitivity. This can be attributed to the properties of atomic sources, Bragg pulse shapes, the pulse duration, and the relative position deviation of the atoms and Bragg pulses. Here, we investigate the effect of the atomic source's diffusion and velocity width on the efficiency of Bragg diffraction of the moving cold atomic cloud. The transfer efficiency of Bragg mirrors and beam splitters are numerically simulated and experimentally measured, which are well consistent in comparison. We quantify these effects of atomic diffusion and velocity width and precisely compute how Bragg pulses' efficiencies vary as functions of these parameters. Our results and methodology allow us to optimize the Bragg pulses at different atomic sources and will help in the design of large momentum transfer mirrors and beam splitters in atom interferometry experiments.