The infrared spectrum of the Galactic center and the composition of interstellar dust.

We have obtained 5-8 micrometers spectra of the Galactic center from the Kuiper Airborne Observatory at resolving powers of approximately 50, approximately 150, and approximately 300. These spectra show absorption features at 5.5, 5.8, 6.1, and 6.8 micrometers. Together with previously observed features in the 3 micrometers region, these features are compared with laboratory spectra of candidate materials. The 3.0 and 6.1 micrometers features are due to the OH stretching and bending variations of H2O and are well fitted by water of hydration in silicates (e.g., talc). The 3.0 micrometer band is equally well fitted by ice mixtures containing 30% H2O, but such mixtures do not provide a good fit to the observed 6.1 micrometer band. The 3.4 and 6.8 micrometers features are identified with the CH stretching and deformation modes in CH2 and CH3 groups in saturated aliphatic hydrocarbons. The 6.1 micrometer band shows a short wavelength shoulder centered on 5.8 micrometer, attributed to carbonyl (C double bond O) groups in this interstellar hydrocarbon dust component. Finally, the narrow 5.5 micrometer feature is also attributed to carbonyl groups, but in the form of metal carbonyls [e.g., Fe(CO)4]. We have derived column densities and abundances along the line of sight toward the Galactic center for the various identified dust components. This analysis shows that hydrocarbon grains contain only 0.08 of the elemental abundance of C and contribute only a relatively minor fraction (0.1) of the total dust volume. Most of the interstellar dust volume is made up of silicates (approximately 0.6). Small graphite grains, responsible for the 2200 angstroms bump, account for 0.07 of the total dust volume. The remaining one-quarter of the interstellar dust volume consists of a material(s) without strong IR absorption features. Likely candidates include large graphite grains, diamonds, or amorphous carbon grains, which all have weak or no IR active modes. Finally, various models for the origin of the hydrocarbon dust component of the interstellar dust are discussed. All of them face some problems in explaining the observations, in particular, the absence of the spectroscopic signature of hydrocrbon grains in sources associated with molecular clouds.