3. Experimental results
Several spectra of icy samples have been taken at two different incidence angles (45 and 0 degrees) and in the two different polarizations (P and S). In some cases ("weak" transitions) absorption band profiles coincide when the angle or the polarization is changed. However some instances exist ("strong" transitions) in which the spectra strongly depend on the incidence angle and on the polarization. For example this is the case for the stretching and bending modes of solid 12CO2, for the stretching mode of solid CO and the C-O strething mode of OCS and in general depends on the optical constants (). Here we consider as an example the case for CO2 and in particular a mixture CH3OH:CO2=1:1 and pure CO2.
3.1. The CH3OH:CO2 mixture
Figs. 2 to 5 show the absorption spectrum of a CH3OH:CO2=1:1 mixture in four different ranges. The mixture has been deposited at 12.5 K and then warmed up to 105 K. In Fig. 2 the 3800-2700 cm-1 range is shown where the bands at about 3250 cm-1 and 2950 and 2830 cm-1 are due to the O-H stretching mode and C-H stretching modes in methanol respectively (see e.g., Sandford & Allamandola 1993); the bands at about 3600 cm-1 and 3700 cm-1 are due to combination modes of CO2 (see e.g., Sandford & Allamandola 1990). Fig. 3 shows the 2400-2300 cm-1 spectral range. The band at about 2340 cm-1 is due to the asymmetric stretching mode of 12CO2. Fig. 4 shows the 2300-2250 cm-1 region. The absorption band at about 2280 cm-1 is due to the asymmetric stretching mode in 13CO2. Finally, Fig. 5 shows the 700-600 cm-1 region in which the absorption due to the bending mode of 12CO2 occurs. Here we show the spectra after warm-up to 105 K however several spectra have been taken at 12.5 K and during warm-up and the results here discussed are valid at all temperatures. In all figures the top panels show two laboratory spectra taken at oblique incidence (45 degrees) and at normal incidence (0 degrees) respectively. It is evident that these two laboratory spectra are almost equal in Figs. 2 and 4 and are quite different in Figs. 3 and 5. In the middle panels the case of oblique incidence is considered and spectra have been taken in the two different polarizations (P and S). The absorption band profiles shown in Figs. 2 and 4 do not depend on the polarization of the IR beam while the profiles shown in Figs. 3 and 5 strongly depend on the polarization. In the same panels it is shown the unpolarized spectrum computed from the spectra taken in P and S polarization using the equation
which takes into account the different background spectra in the two polarizations. is the unpolarized transmission spectrum, and are the spectra of the sample in P and S polarization respectively and and are the spectra of the background (which include the substrate) in P and S polarization respectively. As shown in the bottom panels the computed unpolarized spectrum at oblique incidence perfectly matches the laboratory unpolarized spectrum at the same incidence angle. Furthermore it is evident from all these figures that the spectrum at oblique incidence taken in S polarization is equivalent to the unpolarized spectrum at normal incidence. On the other hand at normal incidence the polarized spectra are always exactly equal to the unpolarized ones; in fact, due to the symmetry, in this case all the possible orientations of the electric vector are equivalent.
3.2. Pure CO2
Baratta & Palumbo (1998) have already shown that the laboratory spectra of pure CO2 at 45 degrees taken in P and S polarizations differ to each other. Furthermore they have measured the optical constants (n and k) of pure CO2 and have shown that the computed spectrum at 45 degrees perfectly compares with the laboratory spectrum taken at the same incidence angle. Using the optical constants measured by Baratta & Palumbo (1998) we have computed the absorption spectra of pure CO2 at different incidence angles. Fig. 6 shows the calculated spectra for thin films in the region of the main absorption bands: the 12CO2 asymmetric stretching and bending modes and the 13CO2 asymmetric stretching mode. It is evident that for 12CO2 the band profiles strongly depend on the incidence angle. In fact at any angle different from zero new bands appear at 2375 cm-1 and 675 cm-1 respectively whose intensity depends on the incidence angle. Baratta & Palumbo (1998) have shown that these bands are real. However they are not due to absorption (k) but to the increased reflectivity corresponding to the region across the absorption band where . In the bottom panel of Fig. 6 it is shown the profile of the 13CO2 stretching mode. In this case the profile does not depend on the incidence angle but only the intensity changes due to the different path of the IR beam in the icy sample.
Using Mie theory we have calculated the extinction cross section for two different grain models: spheres and core+mantles spheres. In this latter case we have assumed that the core is made of a typical glass (dielectric) with and . In the calculation we have used r0=0.12 µm for the spheres and r0=0.12 µm and r1=0.1512 µm for the coated spheres (core+mantles) in such a way that the volume of the core is the same as the volume of the mantle. Results are reported in Fig. 7. We verified that in both cases the scattering cross section is negligible (particles can be considered in the Rayleigh limit). In Fig. 7 the spectra calculated using the Mie theory are compared with the synthetic spectra computed for thin films assuming normal incidence (0 degrees) and oblique incidence (60 degrees), already shown in Fig. 6. It is evident that these four spectra exactly coincide for the asymmetric stretching mode of 13CO2 while they are quite different for the 12CO2 bands. In particular, the core+mantle grain model shows two features at about 675 cm-1 in the CO2 bending mode region and at about 2380 cm-1 in the CO2 stretching mode region which are similar to the features which appears in the spectra of the thin film at 45 and 60 degrees. However these features differ in position to each other. In the case of spherical grains no additional feature appears and the main peak shifts to higher wavenumbers. A discussion on the details of the origin of the features observed in small particle absorption cross section can be found in Ehrenfreund et al. (1997).
© European Southern Observatory (ESO) 2000
Online publication: June 5, 2000