Collisional interaction between metastable neon atoms

In this thesis, the study of cold gases of neon atoms in different metastable states is described. It contains measurements of the collisional parameters for both the 3s[3/2]2 and the 3s'[1/2]0 metastable state and the dependence of the inelastic loss on external fields. Furthermore, the investigation of frequency dependent laser-induced collisions, and the possibility to excite photoassociation resonances is presented. Based on previous measurements we have selected 22Ne for evaporative cooling. Although, we can experimentally achieve an increase in phase-space density with evaporative cooling, the relatively high inelastic collision parameters prevent the realization of a Bose-Einstein condensate in neon. For the measurements described here, neon atoms have been confined in a magneto-optical trap, in a magnetostatic trap, or in an optical dipole trap, respectively. By laser cooling inside the magnetic trap, atomic samples with more than 95 percent occupation of the magnetic substate m=+2 could be prepared. They have a typical temperature of 0.5 mK, central densities up to 1e11 cm^-3, and a central phase-space density of up to 2.2e-7. After loading the optical dipole trap from the magnetic trap, 2.5e6 atoms with typical temperatures of 0.1 mK, and central densities up to 5e10 cm^-3 were trapped. By evaporative cooling of the atoms in the magnetic trap we could increase the phase-space density by a factor of 200 to 5e-5. Also simulations of optimized evaporation for our experimental parameters show clearly that we are limited to a phase-space density on the order of 1e-5. From these simulations it became clear that a 5-fold increase in the "good-to-bad" ratio for evaporative cooling suffices to reach the quantum degenerate regime. Investigating the frequency dependence of laser-induced collisions did not reveal an experimental signature for the excitation of photoassociation resonances. For the 3D3 line a frequency dependence of laser enhanced Penning ionization was observed, which is interesting in itself. The absence of the collisional enhancement effect by laser light for the transition to the 3D2 line is intriguing and for an explanation calculations are required. Measurement of the two-body loss coefficient as function of the magnetic field showed a field dependence of the inelastic loss. These losses increase towards both small and large offset fields. In the magnetic trap, we are limited to offset fields < 50 G. In this range of fields, the two-body losses are too large to achieve a Bose-Einstein condensate of magnetically trapped metastable neon. The implementation of an optical dipole trap allowed us to trap the 3P0 metastable state. From the trap loss measurements we determined the two-body loss coefficient of the 3P0 metastable state for both bosonic isotopes 20Ne and 22Ne. For 20Ne we obtained s=6(+5,-4)e-10 cm^3/s and for 22Ne s=11(+7,-6)e-10 cm^3/s. These large two-body losses make it extremely unlikely to reach degeneracy with this metastable state. Nevertheless it is important that the 3P0 metastable state can be trapped to investigate other interesting physical effects. For example, it is essential to apply the STIRAP technique for trapped atoms and to realize the proposed coherent control of collisions. There is also a large interest in a precise determination of the lifetime of this metastable state, which is of importance for the verification of QED. We can also trap neon atoms in their energetically lowest magnetic substate 3P2(m=-2) with the perspective of reducing inelastic collisions in the energetically lowest state.