Examination of loading the isotopes $^{85}$Rb and $^{87}$Rb simultaneously into a shallow far-off-resonance trap (FORT) has revealed an unexpected decrease in maximum atom number loaded as compared to loading either isotope alone. The simultaneous lo
ading of the FORT will be affected by additional homonuclear and heteronuclear light-assisted collisional losses. However, these losses are measured and found to be insufficient to explain the observed drop in total number of atoms loaded into the FORT. We find that our observations are consistent with a decrease in loading rate caused by inter-isotope disruptions of the efficient laser cooling required to load atoms into the optical trap.
We have studied the effects of loading $^{87}$Rb into a far off resonant trap (FORT) in the presence of an ultracold cloud of $^{85}$Rb. The presence of the $^{85}$Rb resulted in a marked decrease of the $^{87}$Rb load rate. This decrease is consiste
nt with a decrease in the laser cooling efficiency needed for effective loading. While many dynamics which disrupt loading efficency arise when cooling in a dense cloud of atoms (reabsorption, adverse optical pumping, etc.), the large detuning between the transitions of $^{85}$Rb and $^{87}$Rb should isolate the isotopes from these effects. For our optical molasses conditions we calculate that our cooling efficiencies require induced ground-state coherences. We present data and estimates which are consistent with heteronuclear long-ranged induced dipole-dipole collisions disrupting these ground state coherences, leading to a loss of optical trap loading efficiency.
We have examined loading of 85Rb atoms into a shallow Far-Off-Resonance Trap (FORT) from an optical molasses and compared it to loading from a Magneto-Optical Trap (MOT). We found that substantially more atoms could be loaded into the FORT via an opt
ical molasses as compared to loading from the MOT alone. To determine why this was the case, we measured the rate of atoms loaded into the FORT and the losses from the FORT during the loading process. For both MOT and molasses loading, we examined atom load rate and losses over a range of detunings as well as hyperfine pump powers. We found that the losses induced during MOT loading were essentially the same as the losses induced during molasses loading at the same MOT/molasses detuning. In contrast, load rate of the molasses was higher than that of a MOT at a given detuning. This caused the optical molasses to be able to load more atoms than the MOT. Optimization of FORT loading form an optical molasses improved the number of atoms we could trap by a factor of two over that of optimal loading from a MOT.
We have studied hetero- and homonuclear excited state/ground state collisions by loading both $^{85}$Rb and $^{87}$Rb into a far off resonant trap (FORT). Because of the relatively weak confinement of the FORT, we expect the hyperfine structure of th
e different isotopes to play a crucial role in the collision rates. This dependence on hyperfine structure allows us to measure collisions associated with long range interatomic potentials of different structure: such as long and short ranged; or such as purely attractive, purely repulsive, or mixed attractive and repulsive. We observe significantly different loss rates for different excited state potentials. Additionally, we observe that some collisional channels loss rates are saturated at our operating intensities (~15 mW/cm$^{2}$). These losses are important limitations in loading dual isotope optical traps.
