Use this URL to cite or link to this record in EThOS:
Title: Bose gases in double-well potentials
Author: Barker, Adam James
Awarding Body: University of Oxford
Current Institution: University of Oxford
Date of Award: 2020
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
This thesis presents experimental results which utilise multiple-radiofrequency-dressed potentials as a method of confinement for cold atoms, and details progress towards investigations of the out-of-equilibrium dynamics of two-dimensional systems. This work has demonstrated the matter-wave interference of Bose gases released from a double-well potential. We investigate the scaling of the matter-wave inteference fringe wavelength, verify the coherence of the experimental procedure, and demonstrate the controllability of the multiple-radiofrequency-dressed potentials to load a prescribed fraction of atoms into each potential well. Methods presented elsewhere were also used to produce two-dimensional Bose gases in the Berezinskii-Kosterlitz-Thouless regime. We present a discussion of phase correlations in two-dimensional Bose gases, including their time-dependent behaviour as the system equilibrates following a quench. Radiofrequency-dressed potentials are intrinsically state-selective, as we demonstrate through manipulations of a mixture of Rb-87 atoms in different hyperfine states within a double-well potential. This can be used to implement species-selective potentials and is a promising architecture for future experiments concerning impurity physics. We find this mixture to be long-lived in the radiofrequency-dressed potential, in stark contrast to recent experiments on this apparatus, in which mixtures of radiofrequency-dressed Rb-87 and Rb-85 were found to suffer from very fast inelastic loss. Furthermore, we describe the application of machine learning methods to optimise the experimental sequence, in particular the stages of laser cooling and evaporative cooling which are necessary to achieve a quantum gas. We compare several machine learning methods, including Artificial Neural Networks and Gaussian Process regression, and determine experimental parameters which have the strongest influence on the desired outcome. We focus our efforts initially on producing Bose-Einstein condensates with very large atom numbers, although we later highlight the versatility of this method by improving other metrics, such as sequence repetition rate or the final temperature of the gas.
Supervisor: Foot, Christopher Sponsor: Engineering and Physical Sciences Research Council
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available
Keywords: Atomic and Laser Physics