Marigo (2003) showed that variable molecular opacities may decrease the efficiency of HBB - or even prevent it in some cases - especially in the more massive AGB stars.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

We call attention to the fact that the MM02 yields for helium are currently the only ones to ensure a good agreement between chemical evolution models for the Milky Way (hereafter MW) and the solar helium abundance - see CMM03. This is essentially due to mass loss in massive stars. In the massive range, the yields computed by MM02 for He, C, N and O would be essentially unchanged by explosive nucleosynthesis and can thus be considered robust calculations which take into account important physics (i.e. rotation and mass loss - see Hirschi et al. 2004 for a detailed description of these models for massive stars).

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... CMM03

For type Ia SNe we adopted the stellar yields of model W7 of Thielemann et al. (1993).

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... models

Here we adopted their standard models and took their tables where the mass loss parameter varies with metallicity - see CRM03 for details. vdHG97 also computed another set of models where less HBB was assumed. We also computed a chemical evolution model where the latter stellar prescriptions were adopted. In Fig. 2 this model would fall in between the two solid curves discussed here - it is not shown to make the figure less crowded.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... points

Romano & Matteucci (2003) have shown that by adopting the vdHG97 set of stellar yields with less HBB it is possible to reproduce the trend of ¹²C/¹³C, which decreases in time in the solar neighbourhood.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... C-rich

Notice that Chieffi & Limongi (2002) do not include rotation in their computations.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... metallicity

Some mechanism able to increase ¹⁴N at low metallicities is still needed in this case as models computed with WW95 stellar yields are not able to fit the new Spite et al. (2005) data for N/O, as shown by the thin solid curve in Fig. 2. WW95 also did not include rotation in their calculations.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... metallicities

Lower mass loss rates lead to longer stellar lifetimes. As a consequence the star undergoes more dredge up episodes thus increasing the amount of C brought to the surface and later ejected into the interstellar medium.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... data

Optical data suggest a flat N/O abundance gradient for the MW, whereas infrared data suggest a negative one (Simpson et al. 1995). Moreover, the recent data by Daflon & Cunha (2004) also show a gradient in N/O very similar to the one obtained by our model once the MM02 stellar yields are adopted (compare asterisks and thick line shown in the middle panel of Fig. 5).

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.