3.1. Structure of the discs
Yet before first images of Vega-type systems were obtained it had become clear that the observed excess emission in the IR stems from circumstellar dust grains rather than from other sources like free-free emission, background IS cirrus, cool companions, etc ([I]; see also ). Already first coronograph images of ¡3 Pic showed a clear disc-like circumstellar structure. Explanations alternative to the dust disc — equatorial mass loss or bipolar outflows from a pre- or post-main-sequence object — were quickly ruled out (see [31,103] for details). A set of subsequent images of ¡3 Pic resulted in a relatively good understanding of the global spatial distribution of dust in the disc. Not going into details (which can be found in other reviews centred on this intriguing object — e.g. [31,103-105]), we outline here the main ideas.
The disc seen nearly edge-on extends outward to more than 1000 AU [65,107]. The material is confined within the half-opening angle ~ 7°  and the mid-plane number density of dust particles falls off approximately as r~3 (see  and references therein). At about 100 AU, however, the radial density distribution changes its profile [63-66,68].
Within this zone the number density increases much slower, so that the normal optical depth remains constant or even drops slightly. Two wings of the disc are substantially asymmetric. By analyzing their own and other observations, Kalas and Kewitt  identified in the disc of (3 Pic five types of asymmetry. These comprise a general brightness/radial extension asymmetry of two wings, their width asymmetry, and warp-like disc asymmetries, some of which are especially pronounced in the inner parts of the disc [64,109],
The asymmetries are also seen in the discs of some other stars (e.g. [23,58,80]) attracting considerable interest as a presumed manifestation of planets orbiting these stars. Whereas the warps and local asymmetries are, most probably, really the signatures of the alleged planets [64,109,110], the global brightness asymmetries may also be due to other mechanisms. The difference by about 20% observed in the brightness and polarization of the two wings of (3 Pic is well explained by slightly different size distributions . Such a difference could, under plausible conditions, be put down to interactions of the circumstellar particles with interstellar dust grains [53,106,112]. On the other hand, the difference in the sizes may be attributed to the action of a planet, too. Lecavelier des Etangs et al.  suggests that, if the planetary orbit is slightly eccentric, it would cause alignment of periastrons of comets/planetesimals at larger distances from the star and therefore a non-axisymmetric dust production and distribution in the disc. Krivova et al.  point out another possible mechanism: as small dust particles are blown away by stellar radiation pressure, their enhanced collisional production in the regions where the disc parent bodies are trapped in planetary resonances would lead to an overabundance of hyperbolic grains in the adjacent outer parts of the disc.
The so-called inner depletion zones are another widely present feature of Vega-type systems, also attributed to the presumed planets. Apart from (3 Pic , the depletion zones are found in all cases when the disc around a MS or transitive-aged star is resolved — see Table 1. Many other MS stars with circumstellar discs have low near- and mid-IR fluxes, yet detectable emission at 60 and 100 ^m. It is very difficult to explain such a distribution of energy unless the regions close to the stellar photospheres are devoid of dust particles. The gaps apparently occur more frequently in stars older than 10 Myr, indicating a clearing with age . The inner parts of younger systems are not yet accessible to observations. But the spectral energy distribution suggests that the inner gaps are not yet formed and the inner boundary of the dust discs/shells is determined by dust grain sublimation (e.g. [88,116-118]). For a few T Tau stars, however, the spectral energy distribution shows an evidence of the gaps [119,120].
The density profile is not established in great detail for stars other than ¡3 Pic . It is clear, however, that the radial density slope for MS objects is typically much steeper than for young stars with discs/shells (cf. values of a in Table 1). In some cases, as for HD 141569 or HR 4796A [71,74,78,80], the discs have an abruptly bounded ring structure with intensity falling off fast both inwards and outwards.
In comparison with the dust distribution and structure of the discs, the properties of dust particles are much more difficult to find out. Nevertheless, some constraints can be placed, especially when a combination of different observational data, like in the case of
Chemical composition. IR and optical data considered in combination allow an estimation of the mean albedo of dust particles. For (3 Pic it appears to be rather high — 0.5 ± 0.2 (see Artymowicz  for a discussion) thereby ruling out such materials as carbonaceous species, pure metals and iron-rich minerals. A similar value (about 0.4) is derived for HD 141569 . Comparison of spectrophotometric data of (3 Pic with laboratory reflectance spectra together with the albedo restrictions led Artymowicz et al.  to the conclusion that the particles are likely to be composed of magnesiumdominated (i.e. iron-poor) olivines and pyroxenes or astronomical ices (H2O, CO2, NH3, CH4) with a small admixture of carbonaceous compounds. Spectroscopic observations of other Vega-type and younger systems imply essentially akin composition — silicates, typically magnesium-rich, small carbon particles (PAHs), iron oxide and crystalline water ice (e.g. [24,43-45,47,121,122]).
Spectrophotometry in the 10 ¡im spectral region has shown that silicates are indeed important constituents of dust in the disc of (3 Pic [124-126], The silicate feature is quite similar to those in some Solar System comets with characteristic features (e.g. secondary peak at 11.3 pm) of crystalline silicates and differs significantly from those in very young stars. A similar profile of the feature is observed in some other cases, including intermediate-stage (post-Herbig Ae/Be) objects [44,43,47,127,128]. The presence of crystalline silicates in the discs is corroborated by emission features in the 20 — 70 /¿m region [42-45]. This suggests that the dust was reprocessed and supports the idea that the cometary bodies resupply the grains to the discs. For yet younger, PMS, objects, however, the spectral structure due to cristalline material is typically not seen, implying probably amorphous materials [121,122,129,130].
Size distribution. Another important characteristic of dust grains in the discs is their size distribution. The results obtained in the case of (3 Pic by different authors differ somewhat and at the first glance seem even somewhat controversial. First of all, the grey colours of the disc [67,70] are indicative of big particles — bigger than a few micron. Even larger sizes are suggested by far-IR and millimetre data [21,131]. To account for both IR and optical observations together, the major contribution to the cross section should be given by grains about 1 — 20 pm in size , though an additional population of hotter and smaller grains is required by some models (e.g. [98,132]). Micron-sized grains are also responsible for the observed silicate emission feature [125,126]. Though grain collisions should really supply the disc with small grains, particles smaller than ~ 2 pm should be quickly swept out of the disc by stellar radiation pressure .
A detailed consideration of polarimetric data available for ¡3 Pic  shows, that there is, in fact, no contradiction in these data. The observed spatial distribution and the wavelength dependence of the polarization together with the disc colours can only be explained if particles in a wide size range are present in the disc with grains smaller than ~ 3 — 10 pm in size being somewhat depleted, though still of importance. This general shape of the size distribution is confirmed by a model of the dynamical evolution of circumstellar dust grains — it is a result of the joint action of the particle collisions and stellar radiation pressure [53,114], A detailed dynamical modelling  implies that the overall distribution has basically three different slopes. The slope is steeper, with the power-law exponent close to 3.5 typical of collision-governed systems, for grains smaller than the blowout limit and for big grains (tens of microns and bigger). For intermediate-sized, just above the blowout limit, particles prone not only to mutual collisions, but also to the destructive impacts of small grains streaming outwards, there is a gentle dip in the distribution. A similar shape of the distribution has long been known for the Zodiacal Cloud , although the distribution is set by different mechanisms.
As principally the same mechanisms of dust production and removal act in discs of other MS stars, the general shape of the size distribution is expected to be similar. Indeed, though the presence of large grains (^5 — 20 fim) is inferred by observational data for many stars (see e.g. [21,31,57,58,135]), evidence of smaller hotter grains also exists (e.g. [71,72,80,99,102,123,127,136,137]). In younger and dustier systems, collisions should be very intensive, so that the depletion of smaller grains could be much less pronounced. On the other hand, the interaction of the dust particles with gas, which is not depleted in young systems, can dampen down the relative velocities of the grains, decreasing the probability of catastrophic collisions.
We are still in the dark about the structure of circumstellar dust grains. If, as is commonly believed, the particles in the discs are really of cometary origin (Sect. 3.3), then the aggregate structure proposed by Li and Greenberg  is a good probability. Fluffy aggregates in this model are made up of primitive interstellar dust particles consisting of a silicate core and an organic refractory mantle. An additional ice mantle could cover the grains in the outer region of the disc. Such a model provides a good fit to the 10 ^m silicate feature and the continuum emission from near-IR to millimetre observed in the case of ¡3 Pic . An extremely high porosity of the aggregates (consisting of more than 95% of vacuum) suggested by the authors seems, however, to be in conflict with the observed high level of the polarization (; see also ).
It was quickly realized that dust particles in the discs around MS stars can not be left over after the star formation. Particles, the presence of which is indicated by different observational data, should be destroyed or removed from the discs by stellar radiation pressure, grain-grain collisions and sublimation of grains (e.g. ). Other possible, though less crucial, sinks of dust include sweeping up by larger bodies, destruction due to impacts of interstellar particles, sputtering due to the stellar wind. This implies that the particles must be continuously replenished. By analogy with the Solar System, Weissman  suggested that the dust material could be supplied by large bodies, like comets or planetesimals. Further observational evidence was found in support of this idea called Falling and Orbiting Evaporating Bodies (FEBs, OEBs) scenario [113,140-142].
1. Spectroscopic observation of /3 Pic in the visible and UV revealed strong variations of the circumstellar absorption lines Call, Mg II, A1 III, CIV and others (see Lagrange et al. , Lagrange-Henri et al. , Beust et al.  and later works of this group for more details). The lines are almost always red-shifted and the variations are sporadic, which was interpreted as the result of vaporization of large bodies falling towards the star. The phenomenon was later observed for many other stars, including PMS ones (e.g. [145-149]). Moreover, spectroscopic monitoring by Grady  revealed red-shifted absorption lines in 17 of 18 programme MS and PMS stars.
2. Identification of the stable CO and C I circumstellar features in f3 Pic implies that the disc is continuously replenished with these gaseous species . A permanent or quasipermanent source of gas is attested by stable circumstellar lines of many other elements . The source should have standard abundances in refractory elements and the FEBs scenario provides a natural explanation of this.
3. A similarity of the profile of the 10/im silicate emission feature in /? Pic and some Solar Sytem comets discussed above indicates that the dust in the disc may be a product of the comet disruptions.
4. The hypothesis of planetesimals as the replenishing source of dust in circumstellar discs is implicitely supported by the existence of trans-Neptunian objects believed to be a primordial population left over after the formation of the planetary system [14,15].
5. The so-called UX Ori phenomenon — irregular brightness dimmings accompanied by an increase of the polarization and a blueing of the object observed for many PMS stars — is believed to be another manifestation of large comet-like bodies or planatesimals moving around the stars (e.g. [153,154]).
It should be mentioned that comet-like bodies or planetesimals produce mainly large dust particles [155,156] which then supply smaller grains to the disc through their colli-sional fragmentation (see ).
Was this article helpful?