Characterization of a magneto-optical trap of 87Rb atoms

This thesis presents the optimization and characterization of a magneto-optical trap (MOT) setup for rubidium-87 atoms. The apparatus is designed to provide a source of cold atoms, with temperatures around 100 $\mu$K, near the tip of a hollow-core photonic crystal fiber. The ultimate goal is to load the cold atomic cloud into the fiber to explore the unique properties of this hybrid system. Since the number of captured atoms depends on the interplay between the magnetic field and detuning, a detailed study was conducted to investigate the number of trapped atoms as a function of different detuning values and magnetic field gradients. By comparing two measurement methods, it was determined that the MOT is capable of trapping approximately $10^7$ rubidium atoms. The optimal magnetic field gradient and laser detuning frequency were found to be $B' = 9.32 \, \text{G/cm}$ and $\delta \approx -13$ MHz, respectively. Additionally, the number of captured atoms was found to be strongly dependent on the background pressure inside the MOT chamber. In this regard, it was observed that by closing the rubidium source valve, the pressure can be reduced by up to 80\% over a 24-hour period, and one possible way to measure it could be through the background signal with the magnetic field gradient turned off. Then we analyzed the trap lifetime $\tau$, which is also inversely proportional to the partial pressure of rubidium. Under optimal conditions, we obtained a trap lifetime of $0.31 \, \text{s}$. Furthermore, a temperature estimation of the atomic cloud was performed using the time-of-flight (TOF) method. The velocity distribution along the $x$-axis, which lie in the imaging plane, was captured with a CMOS camera and yielded temperature of $T_x = (252 \pm 20) \, \mu\text{K}$. These results provide essential insights into the optimization of the MOT parameters and lay the foundation for future studies involving the loading of cold atoms into hollow-core photonic crystal fibers.