4.2. Protoplanetary disks
Contrary perhaps to the expectation that protoplanetary disks would be deeply embedded within the clouds from which they form, and they would therefore be inaccessible to optical observations, HST revealed many dozens of protoplanetary disks ("proplyds"; e.g., Bally, O'Dell and McCaughrean 2000; O'Dell 2001; O'Dell et al. 1993; O'Dell and Wong 1996, following the initial correct identification by Churchwell et al. 1987 and Meaburn 1988). Many of these disks are seen silhouetted against the background nebular light (when they are shielded from photoionization), with some possessing ionized skins and tails (e.g., Bally, O'Dell and McCaughrean 2000; Henney and O'Dell 1999, Fig. 12).
Figure 12. Protoplanetary disks (Proplyds) in the Orion Nebula (M42), HST/WFPC2. Credit: NASA, C. R. O'Dell (Vanderbilt University), and M. McCaughrean (Max-Planck-Institute for Astronomy). http://hubblesite.org/newscenter/archive/1995/45/
The ubiquity of the protoplanetary dust disks (they are seen in 55%-97% of stars; Hillebrand et al. 1998, Lada et al. 2000) demonstrates that at least the raw materials for planet formation are in place around many young stars. Indeed, in a few cases, like the dust ring and disk in HR 4796A and the nearly edge-on disk surrounding Beta Pictoris, the detailed HST images reveal gaps and warping (respectively) that could represent the effects of orbiting planets (Schneider et al. 1999, Kalas et al. 2000).
Another aspect of the protoplanetary disks that is significant for planet formation is the discovery of evaporating disks in the Orion Nebula. As was noted in Section IIIA, some of the Orion proplyds were shown to be evaporating (due to photo-ablation by UV radiation from young, nearby stars) at rates of ~ 10-7 to 10-6 M yr-1 (e.g., Henney and O'Dell 1999). Given that the masses of these disks are typically of order 10-2 M (if normal interstellar grains are assumed, so that the observed dust emission can be scaled to the total mass), this implies lifetimes for these disks of 105 years or less. There exists, however, some evidence that the grain sizes in Orion's disks may, in fact, be relatively large - perhaps of the order of millimeters (Throop 2000). The latter conclusion is based on the fact that the outer portions of the disks appear to be gray (they do not redden background light), and on the failure to detect the disks at radio wavelengths in spite of the implied large extinction in the infrared (hiding the central star in some cases). The observations are thus consistent with grain sizes in excess of the radio wavelength used, of 1.3 mm. When we think about the potential implications of these two findings (about disk lifetimes and grain sizes), we realize that they may have interesting consequences for the demographics of planets in Orion. The relatively short disk lifetimes but relative large grain sizes may mean that while rocky (terrestrial) planets can form in these strongly irradiated environments, giant planets (that require the accretion of hydrogen and helium from the protoplanetary disk) cannot (unless their formation process is extremely fast; Boss 2000, Mayer et al. 2002). It is nevertheless clear from the many observations of "hot Jupiters" (giant planets with orbital radii 0.05 AU) that less extreme environments do exist, in which giant planets not only form, but also have sufficient time to gravitationally interact with their parent disk and migrate inward, to produce the distribution in orbital separations we observe today (see, for example, Lin, Bodenheimer, and Richardson 1996, Armitage et al. 2002).
While disks around young stars produce jets and form planets, similar structures around old stars help perhaps to shape incredible "sculptures" around dying stars.