SIMBAD references

Star-to-star abundance variations of C, N, O, Na and Al in globular-cluster red giants have been recently supplemented by the finding that [Mg/Fe] is depleted in stars with extremely large [Al/Fe] (Shetrone, 1996AJ....112.1517S). To find out which of the magnesium isotopes is responsible for the observed depletion of [Mg/Fe] Shetrone (1996AJ....112.2639S) also undertook an isotopic analysis of Mg and found that it is ²⁴Mg which is depleted in Al-rich giants. On the other hand, Norris & Da Costa (1995ApJ...447..680N) demonstrated that even in the massive globular cluster ω Cen which has intrinsic spreads in both [Fe/H] and the abundances of the s-process elements, [O/Fe] anticorrelates with [Na/Fe] and [Al/Fe] as in ``normal'' monometallic clusters. These new spectroscopic results allow us to test current models of stellar evolution and nucleosynthesis, as well as those of the formation and chemical enrichment of globular clusters. In an effort to explain self-consistently these observations we have considered two possibilities: (1) a deep mixing scenario which assumes that in red giants some kind of (extra)mixing transports products of nuclear reactions from the hydrogen burning shell (HBS) to the base of the convective envelope; and (2) a combination of primordial and deep mixing scenarios. It is shown that (1) cannot account for the anticorrelation of [Mg/Fe] vs. [Al/Fe] without additional ad hoc assumptions, among which we identify a strong but still undetected low energy resonance in the reaction ²⁴Mg(p,γ)²⁵Al, and episodical increases of the HBS temperature up to the value T≃74x10⁶K. In (2) intermediate mass AGB stars are assumed to produce the decreased ²⁴Mg and increased ²⁵Mg initial abundances in some globular-cluster low mass stars and Al is synthesized at the expense of ²⁵Mg in the HBS and transported to the surface of the red giant by extramixing. We discuss advantages and deficiencies of both scenarios and propose some observational tests.