5. Summary and conclusions
We have presented the results of a search for SiO (v=0, J=3-2) and HCN J=1-0 line emission from a sample of southern S stars, using the SEST telescope. The seven stars observed all show evidence of circumstellar envelopes created by mass loss, as indicated by relatively strong CO lines and excess emission in the IRAS bands. We use a statistical equilibrium/radiative transfer code for spherical shells to determine, or at least constrain, the molecular abundances of SiO and HCN for the stars detected in this survey as well as those detected by BL. The observations and analysis lead to the following conclusions.
(1) The SiO (v=0, J=3-2) line is detected in all 7 stars. Excitation models which assume a constant SiO abundance, X(SiO), provide a lower limit to X(SiO) in the range (0.4-5) 10-6. A more realistic model, where the SiO abundance falls off exponentially with distance from the star, with an e-folding radius of 3 1015 cm, gives photospheric SiO abundances in the range (0.4-8) 10-5. The star Gru is an exception, with an order of magnitude lower SiO abundance than the other stars in our sample.
(2) The detection and inferred abundances of SiO are consistent with formation of the molecule under TE conditions near the stellar photosphere, if C/O 0.97 is typical of these S stars, and the gas density and temperature are cm-3 and T 2300 K, according to the chemical models presented in BL. However, the formation of SiO may be controlled by the passage of pulsationally-driven shocks through the stellar atmosphere, if the models of Willacy & Cherchneff (1998) for carbon star atmospheres are also applicable to S stars.
(3) The HCN J=1-0 line is detected in 2 of the 7 southern S stars observed. This detection rate is comparable to that by BL for northern S stars. All S stars in the present work and in BL, which are detected in HCN emission, show the 9.7 micron silicate feature in emission in IRAS LRS spectra (Chen & Kwok 1993).
(4) If HCN is produced near the stellar photosphere, our models require abundances which are much higher than predicted by TE chemistry, unless the gas temperature is 1300 K and the density n(H2) 1012 cm-3. Such conditions would also be likely to result in grain formation. HCN production could then be enhanced if condensation of silicate (or oxide) grains resulted in an increase in the gas-phase C/O ratio to be 1, as proposed by Sharp (1988). Such a scenario would be consistent with the presence of the 9.7 micron silicate feature in emission in these stars. Alternatively, HCN formation may occur through photochemical reactions in the outer envelope. If so, a measurement of the size of the HCN distribution is necessary to estimate the HCN abundance. Interferometric imaging of HCN emission in these stars is needed to discriminate between the possible chemical origins.
© European Southern Observatory (ESO) 1998
Online publication: October 22, 1998