next previous
Up: A survey of SiO


7 Conclusions

We have detected a large number of new thermal SiO sources in the regions of massive star formation. Most detections are made towards strong H2O masers. Further investigation showed that the detection rate is the highest towards objects associated with IRAS sources with large FIR luminosities. This is probably related to the earlier results that the FIR luminosity correlates with the maximum H2O luminosity (Wouterloot et al. 1995), and that the mechanical luminosity of outflow associated with H2O masers correlates with the maser luminosity (Felli et al. 1992). Although more subtle mechanisms have been suggested, the fact that powerful shocks are required for both destruction of silicate grains and the most commonly accepted pumping mechanism of H2O masers, makes these relations understandable. Since FIR emission comes from circumstellar dust heated by the radiation from the central star, its correlation with the SiO detection rate suggests that the occurrence of shocks depends on the stellar luminosity.

A FIR luminosity limited sample of IRAS sources revealed also a spatial dependence in the SiO detection rate. For sources with $\log L_{\rm FIR}/L_\odot \gt 4$ the detection rate in the galactic longitude range $300^\circ \leq l < 60^\circ$ is roughly twice as high as in the other two equal intervals of longitude. About 90% of the sources in the mentioned longitude range are located between the galactocentric radii 3 and 7 kpc.

The full widths above two sigma of the SiO lines vary from 2 to 103 km s-1 and they always have a single peak. The latter characteristic implies that bow-shocks do not provide a generally applicable model for the SiO emission regions. A better agreement with the observations is obtained by the model of turbulent wakes behind bow-shocks by Raga & Cabrit (1993), where the FWZI and the line asymmetry depends on the viewing angle. However, the number of narrow and symmetric line profiles suggests that the SiO formation also takes place in relatively quiescent gas, perhaps due to grain mantle evaporation and subsequent chemical processing aided by ionizing radiation from shocks. The higher detection rate in the inner 7 kpc of the Galaxy may be partly due to the latter processes. It is possible that in this region, where the molecular gas surface density is large and the star formation rate is high (e.g. Blitz 1997; Bronfman 1992), the dense cores are warmer and have higher column or volume densities, and thereby provide appropriate conditions for the production and excitation of SiO.

The data provide a basis for further investigation of shocks associated with massive young stars. In a subsequent paper we shall study the velocities of H2O, OH and CH3OH masers with respect to the observed SiO line profiles, utilizing the existing interferometric maps of the spatial distribution of masers and UC HII regions.

The excitation temperatures derived from the SiO data towards three sources are subthermal and do not show variation in velocity ranges of about 20 kms-1. This agrees with the results of Acord et al. (1997) towards the massive outflow source G5.89-0.39 and indicates that the determination of kinetic temperature of shocked gas is not possible with the aid of two transitions of SiO only.


next previous
Up: A survey of SiO

Copyright The European Southern Observatory (ESO)