The influence of residual gas expulsion on the evolution of the Galactic globular cluster system and the origin of the Population II halo

We present new results on the evolution of the mass function of the globular cluster system of the Milky Way, taking the effect of residual gas expulsion into account. We assume that gas embedded star clusters start with a power-law mass function with slope beta = 2, similar to what is observed for the Galactic open clusters and young, massive star clusters in interacting galaxies. The dissolution of the clusters is then studied under the combined influence of residual gas expulsion driven by energy feedback from massive stars, stellar mass loss, two-body relaxation and an external tidal field. The influence of residual gas expulsion is studied by applying results from a large grid of N-body simulations computed by Baumgardt & Kroupa. In our model, star clusters with masses less than 10(5) M-circle dot lose their residual gas on time-scales much shorter than their crossing time and residual gas expulsion is the main dissolution mechanism for star clusters, destroying about 95 per cent of all clusters within a few tens of Myr. We find that in this case the final mass function of globular clusters is established mainly by the gas expulsion and therefore nearly independent of the strength of the external tidal field, and that a power-law mass function for the gas embedded star clusters is turned into a present-day lognormal one, verifying the theory proposed by Kroupa & Boily. Our model provides a natural explanation for the observed (near-)universality of the peak of the globular cluster mass function within a galaxy and among different galaxies. Our simulations also show that globular clusters must have started a factor of a few more concentrated than as we see them today. Another consequence of residual gas expulsion and the associated strong infant mortality of star clusters is that the Galactic halo stars come from dissolved star clusters. Since field halo stars would come mainly from low-mass, short-lived clusters, our model would provide an explanation for the observed abundance variations of light elements among globular cluster stars and the absence of such variations among the halo field stars. Furthermore, our modelling suggests a natural tendency of >10(7) M-circle dot gas clouds to retain their residual gas despite multiple supernova events, possibly explaining the complex stellar populations observed in the most massive globular clusters.