7 Evolution of the chemical abundances at the surface

Figure 12: Evolution as a function of $\log T_{\rm eff}$ of the abundance ratio N/C where N and C are the surface abundances of nitrogen and carbon respectively. The abundance ratios are normalized to their initial values. The tracks are for 9 $M_\odot$ for different values of the metallicity, Z, and rotation. The long-dashed line, at the bottom, corresponds to a non-rotating 9 $M_\odot$ stellar model at Z=10-5.

Figure 12 shows the evolution of the N/C ratios in models of rotating stars with 9 $M_\odot$ for the initial Z = 0.02, 0.004 and 10^-5. Values for other stellar masses may be found in Table 2.

At zero rotation, for any Z and any masses there is no enrichment during the MS phase (except at Z=0.02 for $M \geq 60~M_{\odot}$ due to very high mass loss). At 9 $M_\odot$ for an initial rotation of 300 km s^-1, we notice an increase of the N/C ratio already during the MS phase. In fact most of the increase in N/C is in general built during the MS phase, and this results from the steeper $\Omega$-gradients and greater compactness. The relative growths of the N/C ratio do not change very much from models with Z = 0.02 to models with Z = 0.004, however there is an increase by two orders of a magnitude for Z = 10^-5. Of course this large N/C enhancement is accompanied by a small enrichment in helium at the surface, typically of a few hundredths as shown by Table 2. Figure 12 illustrates the fact we find throughout this work, i.e. that the various effects of rotation on the internal structure, the surface composition and the yields are in general much higher at lower metallicities.

In Fig. 12, we notice for Z = 10^-5 an increase by a factor 4.5 of the N/C ratio for an increase of 100 km s^-1 of the initial rotation. As illustrated by Table 2, during the He-burning phase the rotation velocities become all the same whatever the initial rotational velocities. Thus, in the He-burning phase we may have very different surface chemical compositions for actually similar rotation velocities. This is likely true for all stellar masses where fast rotation is present, but the effect is in general larger for larger masses.

Curiously enough, at very low Z the fast rotating stars of intermediate masses which reach the TP-AGB phase (this occurs for $M \leq 7~ {M}_{\odot}$) get a higher Z during this phase due to their enrichment in CNO elements. As an example, a 7 $M_\odot$ has a $\rm X(CNO) = 3.1 \times 10^{-3}$ which is 430 times the initial CNO content. Thus very low Z stars may become higher Z stars near the end of their evolution. This might also affect the composition of planetary nebulae in low Z regions.