Abstract-Spectra for certain comets show the presence of crystalline silicate dust grains believed to have been incorporated during comet formation. While grain crystallization is widely assumed to result from the thermal annealing of precursor amorphous grains, the physical processes behind the silicate amorphous-to-crystalline transition are poorly understood. This makes it difficult to place constraints on the evolutionary histories of both grains and comets, and consequently, on the nebular conditions in which they formed. It has, therefore, become necessary to study this process in the laboratory using simulated grain materials. In this paper, we discuss recent results from laboratory investigations into a basic amorphous MgSiO 3 silicate annealed in the region of 1000 K. Our object is not to model the behavior of dust grains per se, but to study the underlying process of crystallization and separate the physics of the material from the astrophysics of dust grains. In our experiments, we bring together spectroscopic measurements made in the infrared with the high resolution structural probing capabilities of synchrotron X-ray powder diffraction. The combined use of these complementary techniques provides insights into the crystallization process that would not be easily obtained if each was used in isolation. In particular, we focus on the extent to which the identification of certain spectral features attributed to crystalline phases extends to the physical structure of the grain material itself. Specifically, we have identified several key features in the way amorphous MgSiO 3 behaves when annealed. Rather than crystallize directly to enstatite (MgSiO 3 ) structures, in crystallographic terms, amorphous MgSiO 3 can enter a mixed phase of crystalline forsterite (Mg 2 SiO 4 ) and SiO 2 -rich amorphous silicate where structural evolution appears to stall. Spectroscopically, the evolution of the 10 µm band does not appear to correlate directly with structural evolution, and therefore, may be a poor indicator of the degree of crystallinity. Indeed, certain features in this band may not be indicators of crystal type. However, the 20 µm band is found to be a good indicator of crystal structure. We suggest that forsterite forms from the ordering of pre-existing regions rich in SiO 4 and that this phase separation is aided by a dehydrogenation processes that results in the evolutionary stall. The implications of this work regarding future observations of comets are discussed.

Abstract. We present the results of combining in situ high resolution synchrotron X-ray powder diffraction and infrared spectroscopic measurements on an amorphous pyroxene powder sample annealed in the region of 1000 K. We find using both techniques that the crystalline structure formed during annealing is Mg 2 SiO 4 (forsterite), but that the presence of certain features in the 10 µm band normally attributed to crystalline enstatite (MgSiO 3 ) is contradicted by spectroscopy in the 20 µm region (along with certain other 10 µm band features) and X-ray diffraction. Both indicate crystalline forsterite as the only crystalline phase formed at this temperature. We discuss the possible mechanism of forsterite formation from amorphous pyroxene and identify the presence of proto-forsteritic structures in the amorphous starting material. We suggest the likely origin of the 10 µm band "crystalline enstatite" features as being due to short-range improvements in the amorphous MgSiO 3 network ordering which are not necessarily accompanied by the formation of crystalline enstatite structure. These results not only suggest that the formation of crystalline enstatite dust grains via annealing may be difficult to realise at this temperature, but also highlight the possibility, in the absence of additional corroborating evidence, of misidentifying the nature of the carrier of the 10 µm "enstatite" features when observed in the spectra of objects such as comets. We also discuss the evolution of fine structure in the region of 15 to 16 µm, which may serve as an observational indicator of grain processing in stellar sources.

No abstract