Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 357, 1045-1050 (2000)

Previous Section Next Section Title Page Table of Contents

4. Discussion and conclusions

We have studied the profile of infrared absorption bands of icy films containing different molecules of astrophysical interest as a function of the incidence angle and of the polarization of the IR beam of the spectrometer. Here we have presented the results relative to the profile of CH3OH and CO2 absorption bands in an icy mixture CH3OH:CO2=1:1 and of pure CO2. We have shown that in some cases the band profiles do not depend on the incidence angle or on the polarization (e.g., absorption bands due to CH3OH in the CH3OH:CO2 mixture, asymmetric stretching mode of 13CO2). However some instances exist in which the absorption profiles strongly depend on the incidence angle and the polarization. For example this is the case of the stretching and bending mode of CO2. Here we have discussed the results relative to pure CO2 and CO2 in mixture with CH3OH. Baratta & Palumbo (1998) present the results relative to pure CO2 and CO; Palumbo & Baratta (2000) discuss the case of CO2 in mixture with H2O and CH3OH. We have compared the synthetic laboratory spectra of pure CO2 at different incidence angle with Mie-calculated absorption cross sections. We have shown that shape and peak position of the CO2 fundamental bands are strongly influenced by "surface modes" while for the CO2 isotope stretching mode the laboratory measured absorption spectra give a very accurate representation of the particle extinction spectrum.

It is well known that the interaction of light with small particles can be different from that with bulk films, resulting in differences in the profile of absorption features between laboratory and interstellar spectra. In order to compare the experimental data with astronomical observations the extinction cross sections for small particles must be derived. For this purpose accurate sets of optical constants are needed. The shape and peak position of the bands strongly depends on the adopted model (size and grain shape). In fact small particles may give rise to multiple extinction peaks where their number and position depend on the grain shape and on the [FORMULA] value across the band (see e.g., Ehrenfreund et al. 1997). In general, the band profile will depend in a complex way on the material (n and k values) and on the geometry (size and shape for grains; incidence angle, polarization and thickness for thin films). Anyway the degree of dependence of the band profile on the geometry is related to n and k values. In particular, if the absorption feature is "weak" (small value of k and consequently, according to the K-K relation, small variation of n across the band) all these differences are faded away and all the absorption profiles will be similar independently of the grain shape and size or the incidence angle, polarization and thickness. Under these circumstances a direct comparison between transmission spectra of laboratory ice mixtures and astronomical observations is allowed. On the other hand, if the band profile observed in a transmission spectrum of a thin film depends on the geometry (i.e., incidence angle or equivalently on P and S polarization at oblique incidence), then the laboratory spectrum differs from the extinction cross section for small particles and in addition different grain shapes give rise to different band profiles. If this is the case a direct comparison is not possible even because one should wonder which laboratory spectrum (at which incidence angle) has to be compared with astronomical observations. Unfortunately the answer is "none".

Hence from the results here shown it is possible to conclude that when the band profiles in thin films depend on the incidence angle or on the polarization then laboratory spectra are not representative of particle extinction cross section and surface modes are not negligible.

Furthermore our results show that when the optical constants of a particular icy mixtures are not known and hence small particles cross-section calculations cannot be performed, it is possible to know whether surface modes would affect the band profile experimentally, both taking the spectra in the laboratory at oblique incidence at different polarizations and taking the spectra at different incidence angle. However even if subpeaks appear in these conditions their position depend on particles shape (Ehrenfreund et al. 1997) and in turn these laboratory spectra cannot be directly compared to astronomical spectra. Quoting Bohren & Huffman (1983):"All of this illustrates a general rule, which we can state but not prove: if there is an interesting effect in a thin film, there will be a corresponding effect in small particles".

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 2000

Online publication: June 5, 2000