2006A&A...457..995V

The chemical content of the ejecta of Asymptotic Giant Branch stars (AGBs) is important to understand whether intermediate mass stars played a role in the ``self-enrichment scenario'', to explain the chemical anomalies observed among Globular Clusters stars. One difficulty of this model is that it is not fully consistent with the observed oxygen-sodium anticorrelation. We look for the combinations of the input-physics parameters by which the yields of massive AGBs are sodium-rich and oxygen-poor. Many evolutions from the early evolutionary phases until the mass of the envelope drops below ∼0.5M_☉ are calculated for a 5M_☉Z=0.001 model for various assumptions concerning the extra mixing from the bottom of the envelope and the relevant cross sections involving sodium. A modest amount of extra mixing from the bottom of the surface convective zone leads to ejecta that are moderately sodium rich and oxygen depleted. A different but appealing modelization allows us to reproduce the observations of stars with a lower oxygen content: extra mixing is not included, but the cross sections of the reaction ²³Na(p, α)²⁰Ne must be a factor of ∼4 lower than the recommended values. If the initial neon content of the star is not solar-scaled but enhanced as the other α elements by a factor ∼2, the afore mentioned cross section must be lessened by only a factor ∼2.