A study of the effect of rotational mixing on massive stars evolution: surface abundances of Galactic O7-8 giant stars

Massive star evolution remains only partly constrained. In particular, the exact role of rotation has been questioned by puzzling properties of OB stars in the Magellanic Clouds. Our goal is to study the relation between surface chemical composition and rotational velocity, and to test predictions of evolutionary models including rotation. We have performed a spectroscopic analysis of a sample of fifteen Galactic O7-8 giant stars. This sample is homogeneous in terms of mass, metallicity and evolutionary state. It is made of stars with a wide range of projected rotational velocities. We show that the sample stars are located on the second half of the main sequence, in a relatively narrow mass range (25-40 Msun). Almost all stars with projected rotational velocities above 100 km/s have N/C ratios about ten times the initial value. Below 100 km/s a wide range of N/C values is observed. The relation between N/C and surface gravity is well reproduced by various sets of models. Some evolutionary models including rotation are also able to consistently explain slowly rotating, highly enriched stars. This is due to differential rotation which efficiently transports nucleosynthesis products and allows the surface to rotate slower than the core. In addition, angular momentum removal by winds amplifies surface braking on the main sequence. Comparison of the surface composition of O7-8 giant stars with a sample of B stars with initial masses about four times smaller reveal that chemical enrichment scales with initial mass, as expected from theory. Although evolutionary models that include rotation face difficulties in explaining the chemical properties of O- and B-type stars at low metallicity, some of them can consistently account for the properties of main-sequence Galactic O stars in the mass range 25-40 Msun