It is appropriate to single out infrared spectroscopy because of its great 

 potential for remote sensing of dust material. Infrared spectroscopy is a power- 

 ful tool for the study of the composition of interstellar dust. Broad absorption 

 features appear superimposed on the infrared continuum of many interstellar 

 sources. Because of their spectral width, these features are generally attributed 

 to molecular vibrational transitions in solid materials. 



A molecular group (such as the methyl group, CH 3 ) absorbs at a few charac- 

 teristic vibrational frequencies. The peak frequency and absorption strength of 

 such a group does not vary much between different molecules within a given 

 class of molecules (such as saturated hydrocarbons). This makes the identifica- 

 tion of molecular groups (and classes of molecules) from observed infrared 

 spectra relatively easy. It does, however, somewhat hamper the precise identifi- 

 cation of the specific molecule involved in the absorption or emission process. 

 Precise identification can be rendered even more difficult by the presence of a 

 collection of molecules and the possible overlap of their absorption bands. The 

 detection of rather subtle spectral variations and thus relatively high-spectral- 

 resolution studies (X/AX > 10 3 ) are needed to identify specific molecules. Fortu- 

 nately, identification can be assisted by studying the spectra of a large sample of 

 sources and correlating the spectral variations with the physical conditions in 

 these sources. 



The possibilities and pitfalls of infrared spectroscopy of interstellar grains will 

 be illustrated with two examples: interstellar grain mantles and large molecules. 



Ground-based observations around 3 ^m have confirmed the presence of 

 water ice in grain mantles inside molecular clouds through the detection of the 

 3.08-jum OH-stretching vibration. Evidence for the C-H bond stretch has also 

 been found in the 3.2- to 3.6-/jm band. The detection of other features, in 

 particular the carbon-bearing molecules, is hampered by atmospheric absorption 

 in the 5- to 8-^m region and the presence of the strong ice and silicate bands, 

 which dominate the 3- and 10-jum regions, respectively. Airborne or spaceborne 

 observations of the 5- to 8-/um region of the spectrum are therefore extremely 

 important to determine the composition of interstellar grain mantles. 



Figure 3-6 shows the 5- to 8-jum spectrum of the bright, protostellar object 

 W33A obtained with the KAO. Deep absorption features at 6.0 and 6.8 £tm 

 are apparent. The 6.0-jum band can be identified with the bending mode in water 

 ices, reinforcing the identification of the 3.08-jum band with the OH-stretching 

 mode in water ice. A good agreement is obtained with the spectrum of pure 

 water or mixtures of water and other molecules as long as the concentration of 

 water is greater than 50%. The 6.8-/jm band is at the correct wavelength to be 

 the C-H deformation mode in saturated hydrocarbons. A comparison of the 

 observed spectrum with that of methanol (CH 3 OH) shows reasonable agreement 

 (fig. 3-6). Higher-resolution observations are needed to confirm this identifica- 

 tion by detecting the subtle variations in the 6.8-/jm band, as present in the 

 methanol spectrum. The position, and shape of the 6.8-jum band show that 



67 



