Centauri may require special yields and evolutionary conditions to be explained, but the need of improved yields is a much more general problem. Examples of important elements for which improved yields are badly needed are the He, C, N and O stable isotopes.
In the late nineties it has been shown (e.g. [Galli et al. (1997), Tosi (2000)] and references therein) that to let the predictions of Galactic chemical evolution models reproduce the low 3He abundances measured in Galactic HII regions a mechanism is needed, able to drastically reduce the 3He production normally predicted for low-mass stars. Such a mechanism was suggested by e.g. [Charbonnel (1995)] and [Wasserburg, Boothroyd, & Sackmann (1995)] to be the consequence of extra-mixing at work in RGB stars, possibly as a consequence of rotation. To reconcile the low HII regions abundances with the high 3He measured in a few Planetary Nebulae (PNe), the extra-mixing should affect about 90% of low mass stars. Since this extra-mixing implies not only a significant 3He depletion, but also a larger conversion of 12C into 13C in such a large fraction of stars, it is important to check whether the corresponding yields are consistent with the carbon ratios observed in PNe.
In Fig. 2 the carbon isotopic ratios measured by [Palla et al. (2000)] in several PNe are compared with the predictions from various theoretical yields. It is apparent that the standard nucleosynthesis predictions of the left-hand panels overproduce the carbon ratio, while the deep-mixing predictions by [Boothroyd & Sackman (1999)] in the right-hand panel nicely fit the data. This nice fit is however misleading, because Boothroyd & Sackman's computations reach only the RGB tip and not the final evolutionary phases. The carbon abundances can be significantly affected by the remaining evolution. Unfortunately, no yields taking deep-mixing into account have been calculated beyond the RGB and none of the yields computed up to the final phases include the deep-mixing effect. In other words, a direct check of the existence of the extra-mixing and of the solution to the 3He problem, with all its cosmological consequences, is currently unfeasible !
Figure 2. From [Palla et al. (2000)]: carbon isotopic ratios measured in PNe (dots with error bars in both panels) compared with stellar nucleosynthesis predictions. The curves in the left-hand panel show the ratio predicted without deep-mixing just before the PN ejection by [Marigo (2001)], long-dashed, [van den Hoek & Groenewegen (1997)], dotted, and [Forestini & Charbonnel (1997)], dashed. The dotted curves in the right-hand panel show the ratio predicted at the end of the RGB phase by [Boothroyd & Sackman (1999)], with (the curves falling down to low ratios) and without deep-mixing (those staying at high ratios).