The rare carbonaceous chondrite which fell around the township of Murchison in northern Victoria, Australia, in 1969 has probably been the subject of more scientific publications than any other meteorite. The discovery of grains formed in a presolar environment and the abundance of organic molecules within the Murchison meteorite have facilitated studies of the formation of the Solar System, the development of stars and the origins of life.

Recent advances in the knowledge of the evolutionary status of asymptotic giant branch (AGB) stars and of the nucleosynthesis processes occurring in them are discussed, and used to interpret abundance determinations for s-process elements, lithium and CNO isotopes in several types of AGB stars. We focus our attention mainly on carbon-rich AGB stars. By combining these different constraints we conclude that most carbon stars in the solar neighborhood are of low mass (M≤3 M⊙), their abundances being a consequence of the operation of thermal pulses and the third dredge-up. However, the observed abundances in carbon stars of the R and J types cannot be explained by this standard scenario. These stars may not be on the AGB, but possibly in the core-He burning phases; their envelopes may have been polluted with nuclear ashes of the core-He flash, followed by CNO re-processing enhancing 13C. Observational evidence suggesting the operation of non-standard mixing mechanisms during the AGB phase is also discussed.

We discuss the production sites of the nuclei from C to Al in solar metallicity stars in the range 13–35 M⊙. We will show how, contrary to current beliefs, the advanced burning phases and the passage of the blast wave play a pivotal role in determining the final yields of quite a few ‘light’ nuclei. We will also show how the relative contributions of the hydrostatic and explosive burning depend on the initial mass of the star: the smaller the mass the larger the importance of the explosive burning.

A numerical model for ion implantation into spherical grains in free space has been developed. It can be applied to single grains or collections of grains with known grain-size distributions. Ion-scattering effects were taken into account using results of computer simulations. Possible isotope and element fractionation of the implanted species was investigated using this model. The astrophysical significance of the model lies in the possible identification of energetically different components (such as noble gases) implanted into presolar grains (such as diamond and SiC) and in establishing implantation energies of the components.

We derive an updated, analytic formula for the gravitational radius, rg, of an irradiated, thin disk. For such a disk, the region outside the gravitational radius will produce a thermal wind, while the region interior to the gravitational radius will be stable. We find that rg ≈ 1.4 (M*/M⊙)/(T0/104K) AU. The value of 1.4 AU is approximately one fifth of currently used values.

The analysis uses the adiabatic approximation. We argue that the same formula applies for the non-adiabatic, isothermal case.

The abundance of the neutron-rich magnesium isotopes observed in metal-poor stars is explained quantitatively with a chemical evolution model of the local Galaxy that considers — for the first time — the metallicity-dependent contribution from intermediate mass stars. Previous models that simulate the variation of Mg isotopic ratios with metallicity in the solar neighbourhood have attributed the production of 25Mg and 26Mg exclusively to hydrostatic burning in massive stars. These models match the data well for [Fe/H] > –1.0 but severely underestimate 25,26Mg/24Mg at lower metallicities. Earlier studies have noted that this discrepancy may indicate a significant role played by intermediate mass stars. Only recently have detailed calculations of intermediate mass stellar yields of 25Mg and 26Mg become available with which to test this hypothesis. In an extension of previous work, we present a model that successfully matches the Mg isotopic abundances in nearby Galactic disk stars through the incorporation of nucleosynthesis predictions of Mg isotopic production in asymptotic giant branch stars.

We investigate the effect of duplicity on stellar yields of carbon, nitrogen, and oxygen. Populations of single and binary stars are modelled and the yields calculated for the whole population. The effects of explosive nucleosynthesis in novae and supernovae are included but by artificially removing these effects from our populations we determine the influence of a binary companion on asymptotic giant branch yields of the CNO elements.

Infrared observations of nova light curves reveal that classical novae form grains in the expanding shells, ejected into the interstellar medium as a consequence of a violent outburst. Such grains contain nucleo-synthetic fingerprints of the nova explosion. In this paper, we analyse different isotopic signatures expected to be present in nova grains on the basis of detailed hydrodynamic calculations of CO and ONe novae and compare them with recent determinations of presolar nova grains from the Acfer 094 and Murchison meteorites.

We discuss possible stellar origins of short-lived radioactive nuclei with meanlife τ ≤ 100 Myr, which were shown to be alive in the Early Solar System (ESS). We first review current ideas on the production of nuclides having 10 ≤ τ ≤ 100 Myr, which presumably derive from the continuous interplay of galactic astration, nucleosynthesis from massive supernovae and free decay in the interstellar medium. The abundance of the shorter lived 53Mn might be explained by this same scenario. Then we consider the nuclei 107Pd, 26Al, 41Ca and 60Fe, whose early solar system abundances are too high to have originated in this way. Present evidence favours a stellar origin, particularly for 107Pd, 26Al and 60Fe, rather than an in situ production by energetic solar particles. The idea of an encounter (rather close in time and space) between the forming Sun and a dying star is therefore discussed: this star may or may not have also triggered the solar formation. Recent nucleosynthesis calculations for the yields of the relevant short-lived isotopes and of their stable reference nuclei are discussed. Massive stars evolving to type II supernovae (either leaving a neutron star or a black hole as a remnant) seem incapable of explaining the four most critical ESS radioactivities in their observed abundance ratios. An asymptotic giant branch (AGB) star seems to be a viable source, especially if of relatively low initial mass (M ≤ 3 M⊙) and with low neutron exposure: this model can provide a solution for 26Al, 41Ca and 107Pd, with important contributions to 60Fe, which are inside the present uncertainty range of the 60Fe early solar system abundance. Such a model requires that 26Al is produced substantially on the AGB by cool bottom processing. The remaining inventory of short-lived species in the solar nebula would then be attributed to the continuous galactic processing, with the exception of 10Be, which must reflect production by later proton bombardment at a low level during early solar history.

We present new results on the evolution and nucleosynthesis in rotating AGB stars. We analyse the role of the gradient of mean molecular weight in the mixing process and show that neglecting this component induces a potentially strong third dredge-up. We also quantify the impact of rotation on the structure and conclude that the effects of rotation (1) mainly concern the inner, fast rotating regions of the stars and (2) are relatively weak as long as rotational mixing does not induce a deep third dredge-up. We also focus our investigations on the s-process nucleosynthesis and show that rotational mixing tends to inhibit the production of s-elements. This results from the contamination of the 13C-rich layers responsible for the neutron production by the poisonous 14N. Our calculations also indicate that the distribution of s-process elements depends sensitively on the magnitude of the diffusion coefficient. These results suggest that rotational mixing is not the main mechanism responsible for the production of s-elements in AGB stars, but that it can influence, and in particular reduce, the final enrichment in s-elements.

Presolar graphite contains a 22Ne-rich component called Ne-E(L). Noble gas studies on graphite aggregates and single grains have shown that although a dominant source of the 22Ne is 22Na, 22Ne in the He-shell of asymptotic giant branch stars have also contributed to the Ne-E(L). In addition to novae that have been considered to be a possible source of 22Na, supernovae are a likely source as well. Krypton isotopic ratios of the separates indicate that part of graphite formed in low-mass (≤3 M⊙) asymptotic giant branch stars of low metallicity (Z ≤ 0.006).

The oxygen isotopic compositions of densely packed submicron oxide grains in two grain separates of different grain size from the CM2 carbonaceous chondrite Murray were determined by multi-detection raster imaging on the NanoSIMS ion microprobe. This led to the identification of 81 presolar spinel and 3 presolar corundum grains among ∼51 700 grains in the CF residue (mean diameter 0.15 µm) and 171 presolar spinel and 29 presolar corundum grains among ˜21 500 grains in the CG residue (mean diameter 0.45 µm). Previous NanoSIMS analysis of individual grains from the same residues has led to the discovery of 15 presolar spinel and 3 presolar corundum grains among 628 CF grains, and 9 presolar spinels among 753 CG grains. The oxygen isotopic compositions of the presolar oxides found by raster imaging are comparable to those of the presolar oxides measured individually. While the abundance of presolar spinel among the (larger) grains of the CG residue is the same for both techniques, the detection efficiency for presolar spinel by imaging among the (smaller) grains in CF is lower due to the small size of these grains. Nonetheless, it is possible to identify presolar grains of this size range. Though single grain measurements are effective for determining the precise isotopic compositions and abundances of presolar grains, raster ion imaging is the method of choice in searches for rare presolar grain types such as presolar silicates.

We have extended our published set of low-mass AGB stellar modelsto lower metallicities. Different mass-loss rates have been explored. We provide interpolation formulae for the luminosity, effective temperature, core mass, mass of dredge up material and maximum temperature in the convective zone generated by thermal pulses. Finally, we discuss the resultant modification of these quantities when we use an appropriate treatment of the inward propagation of the convective instability, as caused by the steeprise in radiative opacity when the convective envelope penetratesthe H-depleted region.

Asymptotic giant branch stars are expected to produce 22Ne through the combined H and He burning that operates during their thermally pulsing evolution. However, observationally there is a fairly tight correlation between the O and Ne abundances as measured in planetary nebulae in various populations. In this paper we use recent detailed stellar evolutionary calculations for compositions appropriate to the Galaxy and the Large Magellanic Cloud, in an attempt to determine if the models are consistent with the observed abundances. We show that there is only a narrow range in stellar mass, about 2 to 4 M⊙ (lower for lower [Fe/H]) where 22Ne is produced in sufficient quantities to affect the total observed elemental neon abundance, which is mostly 20Ne. The models appear to be consistent with the observations, but a more thorough analysis is required.

The primary present-day observables upon which theories of galaxy evolution are based are a system’s morphology, dynamics, colour, and chemistry. Individually, each provides an important constraint to any given model; in concert, the four represent a fundamental (intractable) boundary condition for chemodynamical simulations. We review the current state-of-the-art semi-analytical and chemodynamical models for the Milky Way, emphasising the strengths and weaknesses of both approaches.