1997A&A...328..617M

Although molecular oxygen O₂ is predicted to be an important constituent of the interstellar gas, it has not yet been observed in Galactic interstellar clouds. While ongoing O₂ observation projects in our Galaxy have to use stratospheric balloons or satellites, several teams have already tried to observe its isotopomer ¹⁶O¹⁸O with ground-based telescopes. In this paper, a chemical model of interstellar clouds including ¹⁸O and ¹³C isotopic reactions with a non-LTE calculation of ¹⁶O¹⁸O rotational population has been used to predict the intensity of the 3 ¹⁶O¹⁸O lines observable from the ground, namely the 234, 298 and 402 GHz lines. These predictions are compared to the upper limits obtained by Marechal et al. (1997A&A...318..252M) and to the tentative detection obtained by Pagani et al. (1993A&A...274L..13P) towards L134N. We show that only the 234 GHz line is observable with today's receivers and that the strong upper limits we have reported are not compatible with a standard steady-state model. Clumpiness, high (0.7) C/O ratio or other non-standard models are required to explain our observations. The consequence of the ¹⁸O isotopic fractionation is also discussed.