Laser cooling of atoms is a powerful and widely used tool in atomic physics. Traditional laser cooling relies primarily on the mechanical effect of single-photon transitions between ground states and electronically excited states. The goal of this project was to extend these techniques to explore using multiple laser wavelengths and excited-to-excited state transitions to cool and trap atoms. The new methods may have practical applications for background-free detection of single trapped atoms and for cooling and trapping of atomic species (such as Hydrogen) with extremely inconvenient wavelengths.
This project explored the use of multiple laser wavelengths and multiple atomic transitions to laser cool and trap atoms. Traditional laser cooling relies largely on mechanical forces due to light scattering from a single color laser. Multiple wavelengths and transitions provide interesting extensions to usual "single-photon" cooling: access to substantially different effective photon momenta, access to different atomic line widths and saturation intensities, the possibility of coherence and EIT effects, and the possibility to easily separate atom fluorescence from laser excitation. This approach relies on mechanical forces arising from excited state to excited state transitions. For example, replacing the single laser excitation shown in a) by the three-laser excitation in b) would result in larger light scattering forces, and the fluorescence would be at a very different wavelength from the excitation lasers, which could be easily filtered away.
Our research project explored these ideas in magneto-optically trapped Cesium. Single photon cooling is normally performed on the 852 nm cycling transition, shown in red. Three photon cooling would operate through the three different transitions (shown as orange, blue and green in b), with wavelengths at 894 nm, 795 nm and 761 nm, respectively. (There are other possible wavelength choices, but these have convenient diode laser wavelengths.) A longer term goal (assuming early successes) would be to make a single atom surface MOT by coating the surface to reflect 894 nm, 794 nm, and 761 nm, but to transmit the fluorescing 852 nm. More speculatively, there may be interesting four-wave mixing applications, and it may be possible to demonstrate this scheme using Rydberg transitions, where the interactions between atoms are strong and long ranged.
We demonstrated "2-photon" and "3-photon" laser cooling and trapping in cesium. The experiment used traditional cooling beams at 852 nm in the x-y plane, but replaced the usual two beams along z with lasers at 795 nm. This laser only couples excited-to-excited atomic states, and any cooling and trapping in that direction involves at least two-photon transitions. We found that the cooling and trapping is quite efficient, and have even observed sub-doppler cooling in this new regime. Based on our results, we proposed an efficient cooling scheme for anti-hydrogen that uses excited-to-excited transitions to improve cooling efficiency.