Astron. Astrophys. 331, 742-748 (1998)

4. Discussion

4.1. PAH energy budget

Some 20 to 30% of the total IR-emission of a typical cirrus cloud is emitted by transiently heated very small particles in the 12 and 25 µm IRAS bands (see e.g. Boulanger et al. 1985, Laureijs et al. 1989, Bernard et al. 1994, Dwek et al. 1997). We have estimated the total infrared emission 6 - 1000 µm for the three positions in G 300.2-16.8 . Besides the ISOPHOT photometry we have utilised the 25, 60, and 100 µm IRAS data and, for extrapolation beyond 100 µm, the spectral energy distribution as derived by Dwek et al. (1997) from the COBE/DIRBE and FIRAS data. The resulting values are 1.5 (O1), 1.3 (O2), and 2.5 10 W cm^-2 sr^-1 (O3). Using the data in column (7) and (8) of Table 4 we find that the UIR bands contain a fraction of 5 - 6% and 2% of the total IR-emission at positions O1, O2 and O3, respectively. Very small particles emit a fraction of 20 - 40% of the total at O1, O2 and 10% at O3, the exact percentage depending on the maximum wavelength (16.5 or 30 µm) adopted for the very small particle emission. We conlude that the carriers of the UIR bands are an important agent in cirrus clouds processing stellar UV radiation energy into IR-emission.

4.2. Band ratios and PAH properties

A basic result of our observations is that the relative intensities of the UIR bands in G 300.2-16.8 are comparable to those in the high-ISRF objects (see Table 4). There is a longstanding problem with the explanation of the UIR band intensities in terms of the PAH hypothesis. For neutral, hydrogenated PAHs the bands around 11 - 13 µm which are due to C - H out-of-plane bending vibrations are stronger than the 6.2 and 7.7 µm C - C stretching bands. In highly UV irradiated objects like RNs, HII regions and PNs the contrary situation is observed. This now also seems to be indicated in the diffuse cloud G 300.2-16.8 located in the weak ISRF. A possible explanation for the small 11.3/7.7 µm ratio in these environments is that the PAHs are strongly dehydrogenated (see Léger et al. 1989). Since dehydrogenation is expected to be much weaker for a cirrus cloud, a substantially larger 11.3/7.7 µm ratio would be expected for G 300.2-16.8 which is not the case. We conclude that our observations do not support dehydrogenation as a solution to the band ratio problem.

A probable solution to this problem has emerged from recent laboratory and theoretical results for cross sections of singly ionized PAHs (PAH ) (for a review see Allamandola et al. 1995). The 6.2, 7.7 and 8.6 µm band cross sections for PAH s are typically a factor of 10 larger than for neutral PAHs. The observed 11.3/7.7 µm ratio (Table 4) could thus be understood if a substantial fraction of the PAHs are ionized not only in the high-ISRF objects but also in G 300.2-16.8. The ionisation degree results from an equilibrium between ionizing events, depending on the intensity of the ISRF, and recombination events, depending on the local electron density (Omont 1986). In regions with high UV radiation density and low electron density, such as RNs, the fraction of ionized PAHs approaches unity. In the local-ISRF diffuse medium the situation is less clear. Verstraete et al. (1990) have estimated that for the two compact PAHs, coronene and pyrene, the ionization degree ([PAH ]/[PAH]) is 6%. For the same PAH molecules Bakes & Tielens (1994) and Salama et al. (1996) have calculated similarly low [PAH ]/[PAH] ratios in moderately dense ( 200 cm^-3) diffuse clouds. Joblin et al. (1996) have found that the 8.6/11.3 µm ratio in the reflection nebula NGC 1333 decreases by a factor of 2 between the position of the star and outer nebula in accordance with the calculated change of ionization degree. A similar result has been obtained in NGC 7023 by Cesarsky et al. (1996) for the band ratios of the 6.2 - 8.6 µm and 11.3 - 12.7 µm groups. As a consequence, we would expect the 11.3/7.7 µm ratio to be higher in the low-ionization environment of G 300.2-16.8 than in the reflection nebulae with 100% PAH ionization. Although the value for G 300.2-16.8 is at the upper end of the wide range obtained for RN, PN and HII, it is not clear whether this can already be considered as a significant effect, given the variation and uncertainties of values now obtained for the cirrus cloud.

The continuum emission in the 10 and 16 µm bands may be explained by two alternative models: (1) non-equilibrium emission by large clusters of PAH molecules or very small carbonaceous grains (see e.g. Schutte et al. 1993, Moutou et al. 1996); (2) non-equilibrium emission by very small silicate grains (Draine & Anderson 1985, see also discussion in Désert et al. 1986). Any equilibrium emission by large grains at m is excluded due to the weak ISRF and thus low temperature ( K) of such grains in the cirrus. We notice that the 10 µm continuum flux is higher by a factor of 2 at O1 and O2 than at O3. The 16 µm continuum has its maximum value at O2 (IRAS 25 µm peak) while the emission levels at O1 (IRAS 12 µm peak) and O3 (IRAS 100 µm peak) are equal. This suggests that the 16 µm emission, if due to PAHs, is at least partially caused by a different population of the size distribution (larger species).

© European Southern Observatory (ESO) 1998

Online publication: February 16, 1998
helpdesk.link@springer.de