2016MNRAS.456L..40V

We derive the carbon monoxide (CO) luminosity function (LF) for different rotational transitions [i.e. (1-0), (3-2), (5-4)] starting from the Herschel LF by Gruppioni et al. and using appropriate LCO-LIR conversions for different galaxy classes. Our predicted LFs fit the data so far available at z ~ 0 and 2. We compare our results with those obtained by semi-analytical models (SAMs): while we find a good agreement over the whole range of luminosities at z ~ 0, at z ~ 1 and z ~ 2, the tension between our LFs and SAMs in the faint and bright ends increases. We finally discuss the contribution of luminous active galactic nucleus (LX > 10⁴⁴ergs- 1) to the bright end of the CO LF concluding that they are too rare to reproduce the actual CO LF at z ~ 2.