Next Contents Previous

2.4. Contemporary Interstellar Dust Models

Since 1970s: Modern models

-- J.M. Greenberg; J.S. Mathis; B.T. Draine; and their co-workers

The first attempt to find the 3.1 µm feature of H2O was unsuccessful (Danielson, Woolf, & Gaustad 1965; Knacke, Cudaback, Gaustad 1969b). This was, at first, a total surprise to those who had accepted the dirty ice model. However, this gave the incentive to perform the early experiments on the UV photoprocessing of low temperature mixtures of volatile molecules simulating the "original" dirty ice grains (Greenberg et al. 1972; Greenberg 1973) to understand how and why the predicted H2O was not clearly present. From such experiments was predicted a new component of interstellar dust in the form of complex organic molecules, as mantles on the silicates. This idea was further developed in the framework of the cyclic evolutionary silicate core-organic refractory mantle dust model (Greenberg 1982a; Greenberg & Li 1999a). Similar core-mantle models have also been proposed by others (Désert, Boulanger, & Puget 1990; Duley, Jones, & Williams 1989; Jones, Duley, & Williams 1990).

According to the cyclic evolutionary model, ices evolve chemically and physically in interstellar space, so do the organics. Where and how the interstellar dust is formed appears to involve a complex evolutionary picture. The rates of production of refractory components such as silicates in stars do not seem to be able to provide more than about 10% of what is observed in space because they are competing with destruction which is about 10 times faster by, generally, supernova shocks (Draine & Salpeter 1979a, b; Jones et al. 1994). At present the only way to account for the observed extinction amount is to resupply the dust by processes which occur in the interstellar medium itself. The organic mantles on the silicate particles must be created at a rate sufficient to balance their destruction. Furthermore, they provide a shield against destruction of the silicates. Without them the silicates would indeed be underabundant unless most of the grain mass was condensed in the ISM, as suggested by Draine (1990).

What is currently known about the organic dust component is based very largely on results of laboratory experiments which attempt to simulate interstellar processes. The organic refractories which are derived from the photoprocessing of ices contain a mixture of aliphatic and aromatic carbonaceous molecules (Greenberg et al. 2000). The laboratory analog suggests the presence of abundant prebiotic organic molecules in interstellar dust (Briggs et al. 1992).

The silicate core-organic mantle model is recently revisited by Li & Greenberg (1997) in terms of a trimodal size distribution consisting of (1) large core-mantle grains which account for the interstellar polarization, the visual/near-IR extinction, and the far-IR emission; (2) small carbonaceous grains of graphitic nature to produce the 2175 Å extinction hump; (3) polycyclic aromatic hydrocarbons (PAHs) to account for the far-UV extinction as well as the observed near- and mid-IR emission features at 3.3, 6.2, 7.7, 8.6, and 11.3 µm. This model is able to reproduce both the interstellar extinction and linear and circular polarization.

An alternative model was proposed by Mathis, Rumpl, & Nordsieck (1977) and thoroughly extended by Draine & Lee (1984). It consists of two separate dust components - bare silicate and graphite particles. Modifications to this model was later made by Sorrell (1990), Siebenmorgen & Krügel (1992), and Rowan-Robinson (1992) by adding new dust components (amorphous carbon, PAHs) and changing dust sizes.

Very recently, Draine and his co-workers (Li & Draine 2001b, 2002a; Weingartner & Draine 2001a) have extended the silicate/graphite grain model to explicitly include a PAH component as the small-size end of the carbonaceous grain population. The silicate/graphite-PAHs model provides an excellent quantitative agreement with the observations of IR emission as well as extinction from the diffuse ISM of the Milky Way Galaxy and the Small Magellanic Cloud.

Mathis & Whiffen (1989) have proposed that interstellar grains are composite collections of small silicates, vacuum ( approx 80% in volume), and carbon of various kinds (amorphous carbon, hydrogenated amorphous carbon, organic refractories). However, the composite grains may be too cold and produce too flat a far-IR emissivity to explain the observational data (Draine 1994). (9) This is also true for the fractal grain model (Wright 1987).

In view of the recent thoughts that the reference abundance of the ISM (the abundances of heavy elements in both solid and gas phases) is subsolar (Snow & Witt 1995, 1996), Mathis (1996, 1998) updated the composite grain model envisioned as consisting of three components: (1) small silicate grains to produce the far-UV (lambda-1 > 6 µm-1) extinction rise; (2) small graphitic grains to produce the 2175 Å extinction hump; (3) composite aggregates of small silicates, carbon, and vacuum ( approx 45% in volume) to account for the visual/near-IR extinction. The new composite model is able to reproduce the interstellar extinction curve and the 10 µm silicate absorption feature. But it produces too much far-IR emission in comparison with the observational data (Dwek 1997).

9 Let Cabs(lambda) propto lambda-beta be the far-IR absorption cross section; Td be the characteristic dust temperature; jlambda propto Cabs(lambda) × 4pi Blambda(Td) propto lambda-(4+beta) be the dust far-IR emissivity (where Blambda[Td] is the Planck function at wavelength lambda and temperature Td). While the observed emission spectrum between 100 µm and 3000 µm (Wright et al. 1991; Reach et al. 1995) is well represented by dust with beta = 1.7, Td approx 19.5 K or beta = 2.0, Td approx 18.5 K (Draine 1999), fluffy composite grains have beta approx 1.60 (Mathis & Whiffen 1989). A beta = 2.0 emissivity law is naturally expected for solid compact dust: the far-IR absorption formula for spherical submicron-sized grains is Cabs(lambda) / V = 18pi / lambda × {epsilonim / [(epsilonre + 2)2 + epsilonim2]} where V is the grain volume; epsilonre and epsilonim are respectively the real and imaginary part of the dielectric function. For dielectrics epsilonim propto lambda-1 and epsilonre propto const (epsilonim << epsilonre) while for metals epsilonim propto lambda, epsilonre propto const (epsilonim >> epsilonre) so that one gets the same asymptotic relation for both dielectrics and metals Cabs(lambda) propto lambda-2. Back.

Next Contents Previous