SIMBAD references

The dust composition and dynamics of the circumstellar envelopes of oxygen-rich AGB stars with low mass-loss rates (5x10^–8-10^–5M_☉/yr) have been investigated. We have analyzed the ISO-SWS spectra of twenty-eight oxygen-rich AGB stars with optically thin shells, and modelled the observations with the radiative transfer code DUSTY using the optical constants from laboratory dust analogues. This has allowed us to determine the composition of the dust and the physical conditions at the inner edge of the shell. Moreover, by comparing with CO observations available in the literature, we have determined the gas-to-dust mass ratios and the mass-loss rates of these sources, and analyzed the wind-driving mechanism. The results show that the small amounts of dust present in these envelopes, characterized by visual optical depths in the 0.03-0.6 range, are enough to drive the wind by radiation pressure on the grains. In some sources there are indications of circumstellar dust that does not contribute to the wind-driving, and that may distributed in a disk or clumps. Other sources show signs of variable mass-loss rates. A grain mixture in the shell consisting of aluminium oxide, melilite, olivine, spinel and Mg_0.1Fe_0.9O fit the observed spectra well. From these species, only melilite is required to have a fractional abundance greater than 25% in all cases. Although spinel reproduces the 13µm feature, the absence of the 16.8µm peak in our SWS spectra casts doubts on this identification. The outcome of the modelling reveals that the olivine content in these CSEs increases with pressure and temperature at the inner edge. Moreover, the aluminium oxide percentage in the dust of the envelopes shows a positive correlation with the gas-to-dust mass ratio. These results, together with the derived dust compositions, are consistent with the thermodynamic dust condensation sequence scenario and its freezing-out due to kinetics. However, the temperatures at the inner edge of the shell are substantially lower than those predicted by theory.