Astron. Astrophys. 363, 605-616 (2000)

1. Introduction

Following the pioniering studies by Schwarzschild & Harm (1965, 1967) and Weigert (1966), it is now a quarter century since the first extensive modeling of Asymptotic Giant Branch (AGB) structures, in which, above the carbon oxygen core, hydrogen and helium alternatively burn in shells, with the He-burning phase being initiated by a thermonuclear runaway (Thermal Pulse - TP - phase, e.g. Iben 1981). The convective shells developped during the TPs, and the following "dredge up" of inner nuclearly processed material to the surface (Iben 1975) leads to the formation of Carbon and s-process enhanced stars. Already two decades ago (see Iben 1981) it was realized that in the LMC, where the AGB luminosities are more reliable than in the Galaxy, there were no Carbon stars more luminous than (Blanco et al. 1980), contrary to the theoretical expectations on the extension of the AGB up to . This old finding is confirmed by recent studies (Costa & Frogel 1996).

Later on, very luminous AGBs were discovered (Wood et al. 1983): they were oxygen rich (M-type) stars, and were few compared to the numbers expected if massive AGBs evolve at the theoretical nuclear rate of magyr. This is confirmed also by the scarciness of these stars in the MCs clusters (e.g. Frogel et al. 1980, Mould & Reid 1987).

It is generally accepted that the lack of a C-star stage above is due to nuclear processing at the bottom of the convective mantle of massive AGBs (Hot Bottom Burning, HBB), which cycles into nitrogen the carbon dredged up, if any (Iben & Renzini 1983, Wood et al. 1983). This interpretation has been widely confirmed by the discovery that almost all the luminous oxygen rich AGBs in the MCs are lithium rich, that is they show at the surface a lithium abundance (where ) (Smith & Lambert 1989, 1990; Plez et al. 1993; Smith et al. 1995; Abia et al. 1991). Production of lithium is possible via the so-called Cameron & Fowler (1971) mechanism, if the temperature at the bottom of the convective envelope is K and non instantaneous mixing is accounted for. Modelling of HBB started early with envelope models including non instantaneous mixing coupled to the nuclear evolution (Sackmann et al. 1974) and the lithium production is well reproduced in the recent full models (Sackmann & Boothroyd 1992; Mazzitelli et al. 1999, hereinafter MDV99; Blöcker et al. 2000). Although the details of lithium production depend on the input physics, mainly on the modelling of convection (MDV99), the luminosity range at which lithium rich AGBs should appear is not sensibly dependent on the details, and agrees well with the range observed in the MCs, namely . Let us recall that modelization of the lithium production in AGB is necessary to understand the galactic chemical evolution of lithium from the population II values (Spite & Spite 1993; Bonifacio & Molaro 1997) to the present solar system abundance of (see Romano et al. 1999 for a recent reevaluation of the problem).

The scarciness of luminous AGBs, on the other hand, must be attributed to the onset of strong mass loss which terminates the "visible" evolution and leads to a phase in which the stars are heavily obscured by a circumstellar envelope (CSE) and eventually evolve to the white dwarf stage. This occurs when matter of the outermost layers is found at a distance from the star where temperature and density allow for dust formation, and collisional coupling of the grains with the gas drives a very efficient stellar wind (Habing et al. 1994; Ivezic & Elitzur 1995). The stars will then traverse a phase during which they are surrounded by a thick CSE.

A major problem in building up realistic upper AGB models is then the modelization of mass loss, a necessary ingredient from several independent points of view. For the nucleosynthesis and galactic chemical evolution, the yields from massive AGBs (in particular the lithium yields, as we will see in the present calculations) are influenced by mass loss during the HBB phase, not only because different amounts of processed envelopes are shed to the interstellar medium at a given time (reflecting the stage of nucleosynthesis which has been reached), but also because the stellar structure, and thus and the nucleosynthesis itself, depends on the rate of mass loss: therefore selfconsistent models must be explored and they have not yet been developped.

On the other hand, while at first the searches for AGB stars in the Galaxy and the MCs had been limited to optically bright stars, the surveys in the infrared, starting from IRAS databases or from near IR observations, are making the cornerstones for the understanding of the phase during which, by heavy mass loss, the objects become enshrouded in dust, making them practically invisible at the optical wavelengths and accessible only in the infrared (e.g. Habing 1996). Based on these surveys (see, e.g., Zijlstra et al. 1996; van Loon et al. 1997), follow up observations have given information on the pulsation periods; mass loss rates () and expansion velocity of the envelope have been derived via the OH/IR associated masers (e.g. Wood et al. 1992); ISO spectroscopy and/or photometry has allowed to model the bolometric luminosity and (e.g. van Loon et al. 1999a). The obscured AGBs observations will be a powerful test for the stellar models expected to have extended CSE.  ¹

In MDV99 we presented results from detailed computations focused on lithium production in massive AGB stars, with the aim of studying the influence of convection modelling and other physical inputs on the surface lithium abundance. Although the models were run for population I composition stars, a first comparison with the LMC and SMC lithium rich AGBs was attempted (Fig. 16 in MDV99). However, a full and more detailed comparison with the MCs requires models computed with the appropriate metallicity. Further, in MDV99 we did not touch the problem of visibility and of a possible calibration of mass loss.

In this paper we present stellar models starting from the pre-main sequence and evolved to the AGB phase, with different prescriptions for , with the aim to describe the lithium rich AGBs observed in different luminosity bins in the MCs bright sources. A further constraint can be put based on the s-process enrichment observed in these same stars.

We compare the paths of evolutions having consistent with these observations with the versus pulsation period observations of obscured stars in the LMC (Wood et al. 1992) and with the versus derived from ISO spectrophotometry (van Loon et al 1999a).

We finally show how the models vary with the main physical inputs. In particular we find that, within the framework of our convection model, the lithium vs. luminosity trend is not influenced by the overshooting distance, but this latter is relevant to determine the range of masses involved in lithium production, which our computations show to be extablished within . The minimum mass achieving HBB leading to large lithium abundances is slightly dependent on the mass loss rate adopted, lowest models having more chances of achieving high temperatures at the base of the external envelope during their AGB evolution.

These new tracks represent a first numerical attempt to quantify the mass loss in the massive AGB evolution. The mass loss parametrization has a number of implications on the problems of population synthesis and galactic chemical evolution, whose modeling remains very qualitative if it is not based on such full computations. In particular, the calibration we obtain implies that the lithium production from the massive AGB stars is not relevant for the lithium galactic chemical evolution.

© European Southern Observatory (ESO) 2000

Online publication: December 11, 2000
helpdesk.link@springer.de