SpringerLink
Forum Springer Astron. Astrophys.
Forum Whats New Search Orders


Astron. Astrophys. 322, 924-932 (1997)

Previous Section Next Section Title Page Table of Contents

1. Introduction

Asymptotic Giant Branch (AGB) stars are characterized by intense mass loss processes that play a central role in the star's subsequent evolution and lead to the formation of a cold circumstellar envelope (CSE) of gas and dust. The envelope chemical composition is determined by the amount of carbon enrichment in the stellar atmosphere, due to mixing processes (the so called third dredge-up), induced by thermal pulses near the end of the AGB phase. C-rich envelopes (having [C]/[O] [FORMULA] 1 by number) mainly exist around carbon stars, while O-rich CSEs ([C]/[O] [FORMULA] 1) are associated to M giants. The transition from O-rich to C-rich envelopes is controlled by the extent of the third dredge-up, and it can occur only for AGB stars in a narrow mass range: indeed, stars with a core mass lower than 0.6 [FORMULA]   (Straniero et al. 1995) are unable to undergo the third dredge-up, while H burning at the base of the convective envelope (Hot Bottom Burning) is expected to deplete carbon in the envelope of stars more massive than [FORMULA] 5 [FORMULA]   (Wood et al. 1983; Boothroyd et al. 1993; Frost & Lattanzio 1995), preventing them from becoming C-rich.

Despite the low dust-to-gas mass ratio (in the range [FORMULA] 0.001-0.01), the optical properties of AGB CSEs are mainly determined by the dust grains: silicates in O-rich envelopes (Pégourié & Papoular 1985) and a mixture of Hydrogenated Amorphous Carbon (HAC, Jones et al. 1990; Duley 1993) with inclusions of SiC (Skinner & Whitmore 1988) and possibly Policyclic Aromatic Hydrocarbons (PAH, Puget & Léger, 1989; Cherchneff & Barker 1992) in C-rich CSEs. Both oxidic and carbonaceous star dust is characterized by vibrational bands positioned in the mid-IR window (at 9.7 and 18 µm for silicates, at 11.3 µm for SiC and at 3.3, 6.2, 7.7, 8.6 and 11.3 µm for PAH); furthermore, most of the continuous thermal radiation from the dusty optically opaque envelope is emitted in the same wavelenght range.

For these reasons a mid-IR search was carried out with the imaging camera TIRCAM (Busso et al. 1996, hereafter Paper I) for a sample of 16 sources, including both AGB (O-rich and C-rich) and post-AGB stars. In Paper I, photometry and colors were derived using 10% bandwidth filters at 8.8, 9.8, 11.7 and 12.5 µm, and compared with the Low Resolution Spectra (LRS) from IRAS (1986), in order to establish suitable photometric criteria for discriminating between O-rich and C-rich sources and for estimating mass loss rates. The observations formed the basis for discussing the evolutionary status of the sources and their mass loss history. In Sect. 2 we discuss the general features of their spectra, from the point of view of the optical properties that can be inferred for dust.

Detailed numerical simulations of the CSE thermal structure is the subsequent step for a better determination of the physical parameters. One has to compute the source spectrum and the radial brightness distribution for a wide range of input parameters; comparison with observations will then allow us to estimate the optical depth of the envelopes, to derive mass loss rates, and to extract information on chemical abundances which are relevant to study the stellar nucleosynthesis (see e.g. Busso et al. 1995). With these aims in mind, we developed a numerical code for solving the problem of non-grey radiative transfer through a spherically symmetric dust shell around an evolved AGB star. We present our model in Sect. 3, with emphasis on the input model parameters and on the dust opacities used. In Sect. 4 the model results are shown for the sources observed in Paper I whose LRS is known, and a new estimate for the dust mass loss rates and the dust-to-gas mass ratio is obtained.

In Sect. 5 a more general discussion of the model spectra is carried on. We point out how available dirty silicates dust opacities alone seem not to be completely adequate for a detailed simulation of the mid-IR features of circumstellar dust, as previously suggested by different authors (Onaka et al. 1989a, b; Little-Marenin & Little 1990, hereafter LML90; Simpson 1991), and we test possible explanations. Our conclusions are summarized in Sect. 6.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1997

Online publication: June 5, 1998

helpdesk.link@springer.de