Raman Cooling to High Phase Space Density

Abstract

Experiments on quantum degenerate gasses have become widespread since the first experimental observations of Bose-Einstein condensates (BECs) a few decades ago. Traditionally such experiments have relied heavily on evaporation to produce their quantum degenerate Bose or Fermi gases. Although evaporation has proven to be effective for many atomic species, it leads to the loss of many trapped atoms and is generally slow. The work presented in this thesis aims to improve upon the performance of evaporation by using Raman cooling. First it is shown that Raman cooling alone can produce (impure) BECs, cooling clouds all the way to condensation without evaporation. This is the first demonstration of direct laser cooling to a true three-dimensional BEC. Next it is shown that when evaporation and Raman cooling are both used in the same sequence, pure BECs can be produced in as little as 575 ms. This is the fastest BEC production time known to the author and it is achieved with a much simpler apparatus than other sub-second BEC experiments.

Raman sideband cooling in a 3D optical lattice may provide a way to prepare BECs even faster and potentially with very few collisions between atoms. Some preliminary work along these lines is also presented. Results from Raman sideband cooling in 1D and 2D optical lattices are also presented. Raman sideband cooling in a 1D lattice is shown to produce clouds with phase space densities of about 0.1, which is significantly larger than that achieved in previous work, but still shy of quantum degeneracy. Raman sideband cooling in a 2D lattice is shown to lead to nonthermalized clouds when the atom number and trap frequencies are sufficiently large. These unusual clouds are significantly hotter along the loosely-confined direction of the trap than in the tightly-confined direction, indicating the lack of thermalization in these effectively one-dimensional systems.

Along a separate line of research, the design of a Rydberg cavity quantum electrodynamics (cQED) experiment is discussed. The apparatus is designed to house two optical cavities and two imaging systems. One optical cavity is asymmetric in mirror transmission, and a method is demonstrated for measuring the offset between the trap and probe light standing waves very precisely using only frequency measurements. The imaging systems have moderately large numerical aperture (NA) and are designed to be flexible, cost effective, relatively simple, cause only small aberrations, and require little in-vacuum alignment. The system will be capable of creating atom arrays in the cavities. These arrays could be used to implement quantum logic gates locally using Rydberg interactions between nearby atoms, and the cavity modes could be used as a bus to meditate gates between distant atoms.