The chemistry and physics of carbon clusters have long been recognized as being extremely important in such diverse applied fields as heterogeneous catalysis and soot formation during combustion. More recently new technological areas have developed which focus on fullerene chemistry (Kroto et al. 1985; Weltner & Van Zee 1989; Krätschmer et al. 1990; Fye & Jarrold 1997). Since carbon clusters have been identified in interstellar and circumstellar media by radiowave or infrared spectroscopy, this topic has also attracted considerable interest in astrophysics (see Williams 1994 for a review). Among these carbon clusters many of them are found to be small in size, possessing only three or four atoms (Omont 1993). However medium-sized species also exist appearing in the form of an archetypical cumulenic or acetylenic carbon chain, , with one or two foreign atoms such as or attached to the extremities (Cherchneff et al. 1993; Cherchneff 1995). In contrast to these nanoscopic structures, the presence of microscopic entities such as carbonaceous grains has long been attested, in circumstellar envelopes surrounding cool giant stars, by IR absorption and emission (Merrill 1977). Dust grains have been thus invoked in driven mass loss in these stars during late stages of their evolution (Kwok 1975; Tielens 1983). Following a classical scenario, grains acquire momentum from stellar radiation field and collisions with gas eventually eject mass at a higher rate in space (Knapp 1986). However, in order that this mechanism be really efficient, grains must form in the vicinity of the star, namely in a region located above the photosphere and over a distance not exceeding a few stellar radii (Sedlmayr 1990). Thus one of the major problems for the astrochemists relates to understanding better the chemical pathways towards microscopic structures (like carbonaceous grains composed of thousands of atoms) starting from very small radicalar clusters (such as , , or ) [A very similar problem can be seen in oxygen-rich stars but with formation of silicate grains (Tielens 1990)]. A second important point is that the structure of the interstellar and circumstellar medium-sized clusters is still largely unknown. Many candidates have been postulated such as polycyclic aromatic hydrocarbons (hereafter referred as PAHs) (Tielens 1993), fullerenes (Krätschmer 1993), but also carbynes, i.e. long carbon chains and monocyclic rings (Thaddeus 1994). PAHs are thought to be quite abundant and ubiquitous in the outer region of circumstellar envelopes and in interstellar medium (Léger & Puget 1984). Allamandola et al. (1989) concluded that interstellar PAHs should be singled ionized and estimated their carbon atom numbers to be in the range 20 to 200. A similar conclusion that interstellar PAHs are ionized and partially dehydrogenated due to the UV interstellar fields is made by Allain et al. (1996a,b). Various authors have shown that PAHs can be produced directly in situ via a complex circumstellar or interstellar chemistry (Herbst 1991; Tielens 1993) or can also originate from carbon star atmospheres (Cherchneff et al. 1992; Cadwell et al. 1994; Cherchneff 1995). In any case, the physical conditions prevailing in the region where these clusters are formed have to play an important role and, possibly, other more exotic species such as carbynes and fullerenes may exist as well. We can see that spectra of true (fully hydrogenated) PAHs in both neutral and cationic forms have been clearly identified by the experimentalists (Léger & Puget 1984; Salama & Allamandola 1993). Besides, only a few experimental studies have been devoted to carbyne spectra (Freivogel et al. 1995). Accordingly, the assignment of the so-called diffuse interstellar bands (DIBs) observed in circumstellar and interstellar media to either aromatic-like species (PAHs ionized or not, partially or strongly dehydrogenated), carbon chains or monocycles still remains an open debate (Kerr et al. 1996). The matter is especially complex given that the different forms of carbon enumerated above can, potentially, coexist in various proportions in space. In this paper, we are only concerned with aggregation and growth of carbon clusters in a carbon-rich stellar atmosphere. We are not restricted to a special class of clusters such as PAHs or carbynes but rather our aim is to simultaneously consider the three possible types of hybridization for carbons, i.e. sp (alkynes), sp2 (alkenes and aromatic structures) and sp3 (alkanes, branched or not). In addition to these standard species, dehydrogenated structures with dangling bonds (radicalar clusters) have been considered. Estimates of sp2/sp and sp3/sp ratios are given, together with dehydrogenation rates for the various compounds. The limiting action on cluster growth due to both photodetachment of small radicalar entities or saturation of dangling bonds by hydrogen is also analysed.
© European Southern Observatory (ESO) 2000
Online publication: July 7, 2000