AGB stars

All stars with initial masses <8M. end their lives on the Asymptotic Giant Branch. An AGB star consists of a degenerate C/O core surrounded by a very extended convective atmosphere from which mass is lost via a dense and dusty outflow at rates of 10^-8 to 10^-4M./yr and expansion velocities of 5-30 km/s. The mass-loss mechanism in AGB stars is poorly understood. It is believed to be related to the slow pulsations and the formation of dust, which is subsequently pushed out by radiation pressure. Improving our understanding of the physical mechanisms that drive this process is important because mass loss dominates AGB evolution and also because AGB stars play an important role in the chemical evolution of galaxies by returning gas and dust to the ISM.

Questions that may be addressed are: (i) Where does the dust form in the extended atmosphere? If it forms too far away from the photosphere the mass loss rate resulting from radiation pressure on the dust grains is insufficient to explain the observations. (ii) What is the role of the pulsations in the mass loss process? Is the star distorted due to the large convective motions in the envelope? (iii) What molecules are depleted in the dust formation region? (iv) How does dust formation depend on the phase of the pulsation and on the chemical composition of the star?

Several theoretical studies have highlighted the potential of mid-infrared interferometry for studying AGB envelopes and addressing some of these issues (Lorentz-Martins et al., 1995; Winters et al., 1995; Ivezic & Elitzur, 1996). The mid-infrared region is ideal for studying dust formation near AGB stars and the accompanying depletion of atoms and molecules. A start has been made by Danchi et al. (1994) using the Infrared Spatial Interferometer, which has two 1.65-metre apertures (see Table 1). Their measurements, made at 10 microns using baselines up to 13m, allowed a detailed study of the inner radii for 13 of the brightest late-type stars. The VLTI, with its 8-m apertures and 10-20microns capability, is uniquely suited to extending this work to fainter and more distant objects and with higher spectral resolution. For example, the location and properties of the silicate dust can be studied by measuring the change in size of the object as a function of wavelength through the silicate features at 9.7 and 18 microns (Lorentz-Martins et al. 1995). The layers above the photosphere in which dust forms may extend to about 10 stellar radii, which is several tens of milliarcsec at distances of 500-1000 pc and easily accessible to VLTI. Direct imaging of the stellar disk will also be possible, so limb darkening and distortions from sphericity can be measured. If an AGB star is imaged throughout a pulsation cycle and if simultaneous radial velocity data are taken, the distance can be measured.

The mid-infrared also is the obvious wavelength region for studying post-AGB stars. Many post-AGB candidates were discovered in the IRAS point source catalogue to show warm dust (500 K) and turn out to be binaries. The most famous example of such an object is the Red Rectangle (see Van Winckel et al, 1995). It appears that mass loss on the AGB can be affected by the presence of an unseen companion, with mass being stored in a circum-binary disk. It is currently unclear whether these disks are stable and how they affect the further evolution of the object and the formation of a planetary nebula. The disks should be a few to several tens of AU in size, which means they can be resolved by VLTI at a distance of 500pc.