Chemical evolution of the Milky Way: the origin of phosphorus

Context. Recently, for the first time the abundance of P has been measured in disk stars. This provides the opportunity of comparing the observed abundances with predictions from theoretical models. Aims. We aim at predicting the chemical evolution of P in the Milky Way and compare our results with the observed P abundances in disk stars in order to put constraints on the P nucleosynthesis. Methods. To do that we adopt the two-infall model of galactic chemical evolution, which is a good model for the Milky Way, and compute the evolution of the abundances of P and Fe. We adopt stellar yields for these elements from different sources. The element P should have been formed mainly in Type II supernovae. Finally, Fe is mainly produced by Type Ia supernovae. Results. Our results confirm that to reproduce the observed trend of [P/Fe] vs. [Fe/H] in disk stars, P is formed mainly in massive stars. However, none of the available yields for P can reproduce the solar abundance of this element. In other words, to reproduce the data one should assume that massive stars produce more P than predicted by a factor of ~ 3. Conclusions. We conclude that all the available yields of P from massive stars are largely underestimated and that nucleosynthesis calculations should be revised. We also predict the [P/Fe] expected in halo stars.