2.3. Interstellar Dust: Early Modelling Efforts
1930s: Metallic grains
-- C. Schalén; J.L. Greenstein
But what caused the interstellar reddening? Since hyperbolic meteors were thought to exist, first attempts were made to tie the interstellar dust to the meteors. Small metallic particles were among the materials initially proposed to be responsible for the interstellar reddening, based on an analogy with small meteors or micrometeorites supposedly fragmented into finer dust (Schalén 1936; Greenstein 1938). (6) Reasonably good fits to the -1 extinction law were obtained in terms of small metallic grains with sizes of the order of 0.01 µm. (7) It became evident later that meteors or micrometeorites are not of interstellar origin.
In 1948 Whitford published measurements of star colors versus spectral types over a wavelength range from about 3500Å (UV) to the near IR. The relation was not the expected straight line, but showed curvature at the near UV and IR regions. Things were beginning to make some physical sense from the point of view of small particle scattering.
1940s: Dirty ice grains
-- J.H. Oort & H.C. van de Hulst
Based on the correlation between gas concentration and extinction, Lindblad (1935) argued that it seemed reasonable to grow particles in space since, as hypothesized by Sir Arthur Eddington, gaseous atoms and ions which hit a solid particle in space would freeze down upon it. Lindblad (1935) further put forward the hypothesis that interstellar dust could have formed by condensation (or more properly, accretion) of interstellar gas.
In the 1940's van de Hulst (1949) broke with tradition and published the results of making particles out of atoms that were known to exist in space: H, O, C, and N. He assumed these atoms combined on the surface to form frozen saturated molecules. The gas condensation scenario was further investigated by Oort & van de Hulst (1946) and led to what later became known as the "dirty ice" model. (8)
The dirty ice model of dust by van de Hulst was a logical followup of the then existing information about the interstellar medium and contained the major idea of surface chemistry leading to the ices H2O, CH4, NH3. But it was not until the advent of IR astronomical techniques made it possible to observe silicate particles emitting at their characteristic 10 µm wavelength in the atmospheres of cool stars that we had the cores on which the matter could form. Interestingly, their presence was predicted on theoretical grounds by Kamijo (1963). As van de Hulst said, he chose to ignore the nucleation problem and just go ahead (where no one had gone before) with the assumption that "something" would provide the seeds for the mantle to grow on. By 1945 we had many of the theoretical basics to understand the sources of interstellar dust "ices" but it was not until about 1970 that the silicates were established. Without having a realistic dust model, van de Hulst developed the scattering tools to provide a good idea of dust properties.
1960s: Graphite grains
-- F. Hoyle & N.C. Wickramasinghe
A "challenge" to the dirty ice model came out just after the discovery of interstellar polarization (Hall 1949; Hiltner 1949) since it seemed that the dirty ice model could not explain the rather high degree of polarization relative to extinction (see van de Hulst 1957; this was later shown not to be true by Greenberg et al. 1963a, b).
This led to the re-consideration of metallic grains and the consideration of graphite condensed in the atmospheres of carbon stars as a dust component (Cayrel & Schatzman 1954; Hoyle & Wickramasinghe 1962) because of their enormous potential for polarizing stellar radiation (as a result of its anisotropic optical properties of graphite). The graphite proposal seemed to be further supported by the detection of the 2175 Å hump extinction (Stecher & Donn 1965), although we know nowadays that the graphite model is not fully successful in explaining the 2175 Å hump and the ultimate identification is still not made (see Section 3.1.2).
Kamijo (1963) first proposed that SiO2, condensed in the atmospheres of cool stars and blown out into the interstellar space, could provide condensation cores for the formation of "dirty ices". It was later shown by Gilman (1969) that grains around oxygen-rich cool giants are mainly silicates such as Al2SiO3 and Mg2SiO4. Silicates were first detected in emission in M stars (Woolf & Ney 1969; Knacke et al. 1969a), in the Trapezium region of the Orion Nebula (Stein & Gillett 1969), and in comet Bennett 1969i (Maas, Ney, & Woolf 1970); in absorption toward the Galactic Center (Hackwell, Gehrz, & Woolf 1970), and toward the Becklin-Neugebauer object and Kleinmann-Low Nebula (Gillett & Forrest 1973). Silicates are now known to be ubiquitous, seen in interstellar clouds, circumstellar disks around young stellar objects (YSOs), main-sequence stars and evolved stars, in HII regions, and in interplanetary and cometary dust (see Li & Draine 2001a for a review).
6 Greenstein (1938) concluded that a dust size distribution of dn/da ~ a-3.6 "seems to provide the best agreement of theory and observation (interstellar extinction)." Interesting enough, this power law distribution was very close to the dn/da ~ a-3.5 distribution derived about 40 years later for the silicate/graphite dust model (Mathis, Rumpl, & Nordsieck 1977; Draine & Lee 1984). Back.
7 Another possible influencing factor of proposing the metallic dust model might have been the fact that it was easier to compute the scattering by metallic particles using the Mie theory because to get a -1 law required smaller particles than if they were dielectric and computations for large particles were too tedious (van de Hulst 1986). Back.
8 A. Li noted: the term "dirty ice" was invented by J.M. Greenberg as recalled by H.C. van de Hulst (1997). Back.