A large percentage of stars, both in star-forming clusters and in isolation, are surrounded by dusty, disk-like structures. Infrared observations over the past few decades have yielded substantial information on these systems.

If a dusty star is younger than a few million years, its disk is likely to consist of a rotating cloud of hydrogen in which primordial dust particles are suspended. Such structures are known as protoplanetary disks (see Evolution of Planetary Systems).

Around an older star, the circumstellar disk will be free of gas, and its primordial dust will have disappeared – either through dispersal by stellar winds, or through absorption by the star, or by accretion into planetoids and planets. The dust that we observe must therefore be the result of collisions between remnant planetesimals (e.g., asteroids and comets) orbiting in close proximity to one other. Such orbiting fields of rock, ice, and dust are known as debris disks.

Debris disks have been detected around stars of most spectral classes, from hot A stars to cool M dwarfs. The debris around several nearby stars has even been directly imaged. Robust surveys of the Solar neighborhood demonstrate that 15-20% of F, G, and K stars older than 1 billion years are surrounded by dusty debris disks (Trilling et al. 2008). The percentage is still higher for younger stars, as well as for all stars of spectral class A.

Such disks are thought to resemble one or both of the Solar System’s major debris fields – the Asteroid Belt and the Kuiper Belt. The former is a ring of mostly rocky debris, scattered between 2 and 4 AU. The latter is a much wider belt beyond the orbit of Neptune (30-55 AU) that contains hundreds of thousands of icy objects, ranging in size from a few meters to more than a thousand kilometers in diameter.

However, the Solar System’s dusty debris is much more tenuous than the structures observed around nearby stars – so much so that existing technologies would be unable to detect our own Kuiper Belt if it were only 10 parsecs distant. Beichman and colleagues (2006b) note that the typical extrasolar debris belt is about 100 times more luminous at infrared wavelengths than is the Solar System’s complement of debris. The observed debris fields must therefore be substantially more massive than our own, or they must have experienced a major collisional cascade in the recent past.

Millions of craters preserved on the Solar System’s airless planets and moons bear witness to an epoch of asteroid impacts that occurred during the first billion years of our system’s history. This era concluded with the so-called Late Heavy Bombardment (Gomes et al. 2005), after which our system’s planets and debris belts assumed their mature configuration. Because the epoch of bombardment is widely thought to be responsible for the relative depletion of our Asteroid and Kuiper Belts, an equally violent interlude may have created the dusty disks now observed around other Sun-like stars (Wyatt et al. 2007). Although the Solar System experienced its final bombardment at an early age (about 700 million years), similarly transient events might occur in other systems at later times.

Debris disks have become a key focus in the observational and theoretical investigation of exoplanetary systems. David Trilling and colleagues call them “signposts of planetary system formation” (Trilling et al. 2007), arguing as follows:

“The presence of a debris disk indicates that planetary system formation progressed at least to the planetesimal stage in a given system. . . . This may argue that planetary system formation is quite robust – able to occur in many different conditions. While none of the debris disks we observed are very similar to our own solar system, there can be no question that the process of planetary system formation is quite common.” (Trilling et al. 2008)

Although current technologies have confirmed the existence of planets in only a fraction of debris disk systems, most are likely to contain, at minimum, low-mass rocky or icy objects comparable to the Galilean moons of Jupiter. Many may harbor objects as big as Mars or Mercury, and some might even sustain terrestrial, Super Earth, or ice giant planets.

To date, observations support a correlation between spectral type and planetary system architecture (Laughlin et al. 2004; Johnson et al. 2007a, 2007b, 2008; Kenyon & Kennedy 2008). Just as the likelihood of planetary companions increases with spectral type, from M dwarfs to A-type stars, so does the likelihood of detectable debris. Among the best-studied examples of extrasolar debris disks are AU Microscopii, Epsilon Eridani, HD 69830, Tau Ceti, Beta Pictoris, Vega, and Fomalhaut, representing spectral classes M, K, G, and A. With the exception of HD 69830 and Tau Ceti, these systems are younger than 1 billion years.

debris disks around Sun-like stars

Recent observations using the Spitzer Space Telescope and other advanced instruments have dramatically enlarged the sample of nearby Sun-like known to harbor debris (Greaves et al. 2004, 2005; Beichman et al. 2006; Bryden et al. 2006; Moro-Martin et al. 2007, Hillenbrand et al. 2008, Trilling et al. 2008). Within 65 parsecs (212 light years), at least 53 mature stars of spectral classes F, G, and K have detectable debris. A table of physical and orbital information on these systems is available at Sun-like Stars with Debris.

This sample includes 11 systems that are also known to host gas giant or ice giant planets. All of them happen to be located within 45 parsecs of the Sun. In this subsample are four stars that host at least two planets each, in addition to substantial debris fields. HD 128311 and HD 82943 harbor two gas giants each, and HD 38529 hosts an inner gas giant and an outer object that overlaps the mass range of brown dwarfs (Moro-Martin et al. 2007). HD 69830 harbors three planets, all of them probably ice giants in the mass range of Uranus and Neptune. Around these four stars we have begun to glimpse the large-scale architecture of an entire planetary system, a perspective that until very recently was limited to the single example of our Solar System.

An absence of detectable dust has been noted around five other nearby exoplanetary host stars: 55 Cancri, Upsilon Andromedae, 51 Pegasi, Tau Bootis, and Pi Mensae (Beichman et al. 2006a).

debris disks around M dwarfs

Despite dedicated searches, few M dwarfs are known to harbor debris disks, and those that do are younger than about 200 million years (Gautier 2007, Shankland et al. 2008). This contrasts strongly with the situation in Sun-like stars, as noted above (Trilling et al. 2008).

AU Microscopii is the only red star so far identified whose debris disk has actually been imaged. Located only 9.9 parsecs (32 light years) away, AU Mic is evidently a sibling to Beta Pictoris (Kalas et al. 2004). Both stars were probably born in the same star-forming nebula between 10 and 20 million years ago. With a spectral type of M1, a mass of 0.5 MSOL, and a luminosity of 0.1 Solar, AU Mic is relatively large and bright for its spectral class, but still much dimmer and cooler than Sun-like stars. As a result, the outer regions of its disk still seem to preserve primordial dust grains. We see the disk in an edge-on orientation, with an inner radius of about 12 AU (Krist et al. 2005) and a central clearing that may be the result of perturbations by at least one newly-accreted planet. The overall radius of the disk is more than 120 AU and may extend to 210 AU (Kalas et al. 2004), a much greater area than the disks observed around larger and more mature stars such as Tau Ceti and our own Sun.

Two other, more distant M dwarfs have also been shown to harbor debris disks. Located about 21 parsecs away, GJ 842.2 has an estimated age of 200 million years and a large disk of cold debris extending to a radius of 300 AU (Lestrade et al. 2006). This is the most massive debris field yet detected around an M dwarf. Located a little farther away, at about 27 parsecs, GJ 182 is probably 20 to 50 million years old (with an upper age limit of 150 million years), and thus about the same age as our Sun during the era of terrestrial planet formation by giant impacts (see Evolution of Planetary Systems). The star's infrared excess attests to the presence of warm debris at small radii, compatible with ongoing collisional cascades, with cooler dust extending outward to 120 AU (Liu et al. 2006).

Intensive observations have revealed no debris around another well-known M dwarf: GJ 876, the first red dwarf star ever shown to harbor a planetary system. Despite a pair of gas giants and one Super Earth, whose collective presence hints at a primordial disk that was at least as extensive as the one theorized for GJ 842.2, this star exhibits no infrared excess (Shankland et al. 2008). Its non-detection may be due to the age of GJ 876, which is at least 1 billion years and possibly much older (Marcy et al. 1998, Rivera et al. 2005), because the debris in a more evolved system might have had time to be ground into dust and dispersed by stellar winds. Or it might result from the gravitational effects of the two gas giants, which could have swept any debris out of the system through long-term perturbations (Shankland et al. 2008). Alternatively, GJ 876 - and perhaps other older M dwarfs - may harbor debris of unknown composition or properties that make it more difficult to observe than the debris around Solar-type stars (Shankland et al. 2008).

debris disks around A-type stars

The dustiest population of all is spectral type A, in which at least 50% of young stars, and 25% of stars nearing the end of their main sequence lifetimes, harbor debris (Su et al. 2006, Trilling et al. 2008). This frequency can be attributed to two factors. First, given their short lifetimes, A stars are younger on average than Sun-like stars or M dwarfs, and thus more likely to preserve their disks. Second, A stars are brighter than average stars of the more numerous spectral classes, providing the intense luminosity that enables their disks to be detected.

* Recently published sources vary considerably in their estimates of Vega's mass, ranging from 2.3 to 2.9 Msol.
Sources: Su et al. 2005 ("The Vega debris disk"); Wyatt et al. 2007b ("Steady-state evolution of debris disks around A stars"); Aufdenberg et al. 2006 ("First results from the CHARA Array VII: Long-baseline interferometric measurements of Vega"). See references.

By far the best-studied debris disks are those of three A stars located within 20 parsecs: Vega, Fomalhaut, and Beta Pictoris. These are often called “Vega-type stars” because Vega's dusty envelope was the first to be discovered, through infrared observations by the IRAS satellite in 1983 (Aumann et al. 1984). Considerably less attention has been paid to two other nearby A stars, Denebola and Iota Centauri, despite their comparable debris disks. Although they are all quite young in astronomical terms, these stars have already completed a substantial portion of their time on the main sequence. As a class, A stars have lifetimes in the range of 850 million years (Su et al. 2006).

In analyzing these three nearby systems we must take into account their age, since all of them are less than half a billion years old. At that stage in the Solar System's evolution, it still preserved extensive debris fields, with objects in frequent collision as a result of secular interactions with the system's four giant planets. Many A stars will experience similarly violent histories, but they are likely to leave the main sequence and expand into giant stars by the time their planetary systems reach an equilibrium.

As the principal representatives of the class, the Vega trio is discussed in more detail in a related section. In each of these three systems the debris accompanying the star has acquired structure, either in the form of a central clearing, or in a series of rings, or through a combination of the two. In each case the most likely explanation for the disk structure is the accretion of planets or planetoids out of the rocky and icy particles that comprise the debris. Since all three stars long ago lost their primordial complement of dust, the fine particles that currently allow their disks to be detected must be the result of collisional processes comparable to the Late Heavy Bombardment.

Last update June 2008