STEPS TOWARD INTERSTELLAR SILICATE MINERALOGY .1. LABORATORY RESULTS OF A SILICATE GLASS OF MEAN COSMIC COMPOSITION

Although extrasolar silicates were detected more than 25 years ago, im portant questions concerning chemical composition, material properties , and grain structure are still without reliable answers. The most imp ortant of these questions are listed at the beginning of this paper be cause they play decisive roles as guide-posts for the silicate researc h program of the Jena laboratory astrophysics group. This paper commun icates the first results of this program aimed at a closer mineralogic al characterization of the interstellar/circumstellar silicates that h ave been observed in different types of objects. In this first approac h, pyroxene glass samples, the cation content of which reflects mean c osmic proportions of the four most abundant metals, have been prepared and analytically characterized. They are expected to be good candidat es for matching the silicate spectra of star-forming regions and young stellar objects (YSOs). For the pyroxene glass, optical constants fro m 250 nm to 500 mu m have been determined. Particles having sizes with in the Rayleigh limit show broad bands peaking at 9.5 and 18.8 mu m. F or the sake of comparison, a crystalline sample of the same compositio n was also measured. Its narrow bands are positioned at 9.4, 10.5, 11. 1, 13.7, 15.6, 18.1, 19.5, 26.5, 29.5, 37.5, and 49 mu m in agreement with expectations for a chemical composition corresponding to hypersth ene. In addition to the vibration bands weak crystal field bands at 1 and 2 mu m due to Fe2+ have also been detected for the pyroxene glass. If these bands were detectable in interstellar and circumstellar sour ces they would offer a unique possibility of discriminating the pyroxe ne-type from the olivine-type silicates. The FIR absorption coefficien t measured for the glass sample turned out to be proportional to lambd a(-2). The centroids of the 10 and 19 mu m bands of the pyroxene glass satisfactorily match those observed in the Orion Trapezium and massiv e YSOs in molecular clouds. While the mean 10 mu m band profile derive d from observations of six massive YSOs is excellently fitted by profi les calculated with the pyroxene glass data for Rayleigh grains, the O rion Trapezium profile cannot be satisfactorily represented by analogo usly calculated profiles. Broadening mechanisms that could give an exp lanation for the great width of the Orion profile are discussed. In co ntrast to former conclusions, the new pyroxene glasses do not satisfac torily fit silicate emission profiles observed with less massive YSOs (Herbig Ae/Be stars, T Tauri stars). This suggests that the properties of silicate grains present in circumstellar envelopes of massive YSOs could be considerably different from those of the less massive ones. Finally, the new data are compared with former results on pyroxene gla sses of different composition which were derived from transmission spe ctra. The comparison underscores the point that the kind of preparatio n sensitively influences the resulting optical data.