Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 351, 1066-1074 (1999)

Previous Section Next Section Title Page Table of Contents

5. Astrophysical implications

5.1. Interstellar candidate: CH3OH

Of all the molecules used to constrain the possible physical process occuring in ices containing CO2, only a few are astrophysically relevant, and the best candidate is certainly methanol. This molecule has indeed been detected with high abundances in RAFGL7009S and W 33A through methanol very specific combination modes (Dartois et al. 1999). In space, the ice mantles are generally water dominated and we have performed experiments with carbon dioxide, methanol and water. In these experiments, various proportions of these three molecules mixed in the same ice matrix have been used. The bending mode of CO2 can only split if the water content in the mixture does not exceed too much the methanol content because as shown above, hydrogen bonding inhibits the complex formation and thus the splitting of the mode. We thus show not only that methanol must be abundant to satisfy the stchiometric properties of the complex formation but also that the phase in which this complex forms is physically segregated from the bulk of the water ice (Ehrenfreund et al. 1998). This effect might be a line of sight effect (sampling different grains at various temperature spatially uncorrelated) or could be due to the mantles forming an onion-like structure, both hypotheses being indistinguishable whithout spatial information.

5.2. Temperature diagnostics

The comparison between experiments and interstellar CO2 profiles is instructive. The differences observed from source to source reveal the thermal evolution of the ice. By their own nature, the protostellar envelopes display a temperature evolution both in time (due to the embedded star evolution itself) and radius (the flux constancy in radiative transfer imply the temperature must decrease with radius if there is no emitting source apart from the central star). The resultant spectra we observe towards different lines of sight are then affected by both. We know from the laboratory experiments that to a spectrum corresponds an evolutionary step, and we can obtain a more complete overview of the degree of evolution by comparing different ice modes as now the complete spectra are available with ISO. Only with such a complete scan of the possible observable transitions can we infer the degree of evolution of the sources. This experimental knowledge is one additional clue to infer the age of the condensations, and we can say that given sources are more evolved than others. However, saying that it is due to a more massive star, heating more efficiently the dust, or that it is due to a longer exposition of dust to a moderate radiation requires information from other observations or modes.

Let us focus on RAFGL7009S, NGC7538 IRS9 and S140 sources. S140 is a source located in the L1024 dark cloud, in a massive star forming region. In S140, ices are at a high temperature, certainly at a very late stage of their evolution, as solid CO is absent from the spectrum and the water bending mode located around 6 µm is beginning to show some cristallisation. Ices in RAFGL7009S are at an earlier temperature stage (d'Hendecourt et al. 1996). Comparison of their respective spectra with laboratory data shows that a good agreement is obtained for an ice temperature of about 110-120 K for S140 (Fig. 6), higher than the one needed to reproduce RAFGL7009S spectrum (about 80-90 K). Note that in space this temperature is likely to be slightly lower as the time scale for the evolution of the ice is much longer than in the laboratory. NGC7538 IRS9 (Fig. 6, lower panel) represents a case similar to S140. However, the complete spectrum of this source reveals the presence of more volatile ices than in S140, like CO (Whittet et al. 1996) which means that ices reside in a less evolved environment.

The NGC7538 IRS9 spectrum presented in Fig. 6, together with the S140 one, is one of the two characteristic CO2 bending mode profiles first showed in de Graauw et al. (1996). It is dominated by a two peaks substructure. The other "class" of line profiles is represented by RAFGL7009S and GL2136 (three peaks), this last source being also reported in de Graauw et al. (1996). These spectra, together with our interpretation, clearly show that we probe the sources at various evolution stages. However, if we compare the solid phase to the CO2 gas component, the evolutionary scheme is more difficult to establish. In NGC7538 IRS9, the carbon dioxide gas to solid ratio is around 0.01 whereas for GL2136 it is 0.02 (van Dishoeck et al. 1996). The water gas to solid ratio is less than 0.04 for NGC7538 IRS9, about 0.4 in GL2136 and the gas component is much hotter in this last source. Taking only this into account, GL2136 appears hotter than NGC7538 IRS9, in which many molecules are still in the ice mantles (van Dishoeck & Helmich 1996). The solid CO2 bending mode region reveals something less straightforward as the substructures observed in GL2136 are associated, in the laboratory, with a less thermally evolved complex between CO2 and CH3OH than what we derive for NGC7538 IRS9. Even if we now understand the physical parameters which allow the interpretation of the spectra, it shows that we lack in such cases additional spatial information to know precisely to what extend the region where the gas is observed is related to the solid phase location in the clouds.

In addition to the CO2 bending mode line profile, a high resolution monitoring of the CO2 second isotope stretching mode can provide useful clues about the evolutionary stage of interstellar ices as previously discussed. Both modes behaviours and line shapes are good indicators of the degree of evolution of the sources. It can be understood as a sort of "third dimension" in the ice study. Indeed, if we can relate the profiles to given temperatures or physical state in the laboratory, we can infer the temperature at which the ices have been raised.

5.3. Uniqueness of the candidate

We chose to show the ethanol-CO2 mixture spectrum in Fig. 6 as well as the methanol-CO2 one in comparison to S140 to demonstrate that the effects of molecular interactions lead to infrared spectra whose interpretation may not be unique. To identify the astrophysical candidate, one must spectroscopically look at the full range of the infrared spectra provided by ISO to observe potential other molecular modes, or look at specific transitions from observations in ground based atmospheric windows, like we did for RAFGL7009S. Indeed, in our experiments the ethanol mixture provides a very good fit to the spectrum of the source but methanol is definitively the right interstellar candidate .

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1999

Online publication: November 16, 1999