Desorption of CO and O \mathsf interstellar ice analogs

Aims. Solid O sub(2) has been proposed as a possible reservoir for molecular oxygen in dense clouds through freeze-out processes. The aim of this work is to characterize quantitatively the physical processes that are involved in the desorption kinetics of CO-O sub(2) ices by interpreting laboratory temperature programmed desorption (TPD) data. This information is used to simulate the behavior of CO-O sub(2) ices under astrophysical conditions. Methods. The TPD spectra have been recorded under ultra high vacuum conditions for pure, layered and mixed morphologies for different thicknesses, temperatures and mixing ratios. An empirical kinetic model is used to interpret the results and to provide input parameters for astrophysical models. Results. Binding energies are determined for different ice morphologies. Independent of the ice morphology, the desorption of O sub(2) is found to follow 0{\rm th}-order kinetics. Binding energies and temperature-dependent sticking probabilities for CO-CO, O sub(2)-O sub(2) and CO-O sub(2) are determined. O sub(2) is slightly less volatile than CO, with a binding energy of 912\pm15 versus 858\pm15 K for pure ices. In mixed and layered ices, CO does not co-desorb with O sub(2) but its binding energy is slightly increased compared to pure ice whereas that of O sub(2) is slightly decreased. Lower limits to the sticking probabilities of CO and O sub(2) are 0.9 and 0.85, respectively, at temperatures below 20 K. The balance between accretion and desorption is studied for O sub(2) and CO in astrophysically relevant scenarios. Only minor differences are found between the two species, i.e., both desorb between 16 and 18 K in typical environments around young stars. Thus, clouds with significant abundances of gaseous CO are unlikely to have large amounts of solid O sub(2).