Low Mass Binaries

Brown dwarfs also exist as companions of late-type main sequence stars. Gl 229B, the archetypical T dwarf, lies at a separation of — 50 AU from an M0.5 primary.

Since that discovery, the nearest stars have been subjected to intense scrutiny (Simons et al, 1997; Oppenheimer et al, 2001), but only two have yielded brown dwarf companions: Gl 570D is an extremely cool T8 dwarf, lying 1525 AU from Gl 570ABC, a K5/M1/M2 triple (Burgasser et al, 2000); and e Indi, a K5 dwarf, is accompanied at a separation of 1459 AU by a T1/T5 binary, e Indi Bab (Scholz et al, 2003; McCaughrean et al, 2006). Other binaries within 10-20 parsecs are known, including Gl 802B (Sect. 5.3.3); G196-3B, a resolved L-dwarf companion of a young (— 107 year-old) M dwarf (Rebolo et al); and LP 216-75/2M0951+35, a Pleiades-age M4.5 dwarf with a - 0.02M© wide (A -450 AU) L6 companion (Reid & Walkowicz, 2006). The implication, however, is that luminous (L/T) brown dwarf companions of late-type dwarfs (particularly M dwarfs) are rare.

The qualifier "luminous" in the last sentence of the previous paragraph highlights one of the key observational issues: brown dwarfs cool and fade with age (Figs. 5.1 and 5.2), making it difficult to confidently assess their presence among local field stars, which have typical ages of several Gyrs. Radial velocity measurements can set some limits on brown dwarfs at small separations. Approximately 100 field M dwarfs (mainly early types, M0-M3) are included in the radial velocity planet-search programs described by Irwin (this volume). So far, four systems (Gl 436, Gl 581, Gl 849 and Gl 876) are known to harbour planetary companions. Only Gl 849b has a minimum mass close to that of Jupiter (M2 sin(i) = 0.82MJup, Butler et al, 2006); the remaining planets have M2 sin i < 0.1MJup. While many nearby M dwarfs still lack adequate observations, no brown dwarf companions have been discovered to date.

An alternative means of circumventing the visibility issue is to survey a much younger population. The nearest star forming regions, however, are several hundred parsecs distant, setting strong constraints on our ability to detect faint companions. Recently, a number of young (— 107 years) stellar associations have been identified, with sparser membership but lying at distances of 40-60 parsecs from the Sun (Zuckerman & Song, 2004). Principal among those is the TW Hydrae Association (TWA), which has at least 28 stellar/brown dwarf systems as members. Nine are binaries, three are triples and one is a quadruple system, giving an overall multiplicity of 43%. The companions include two brown dwarfs (TWA 5B and 2M1207-39B), both resolved systems.

The 2M1207-39AB system deserves particular comment. The primary is itself a brown dwarf, whose spectral type of M8 (Gizis, 2002) implies a mass of 25-45MJup for an age of —107 years. The companion, discovered by Chauvin et al (2004), is — 8 magnitudes fainter with an inferred mass of 4-6MJup, corresponding to a mass ratio of q — 0.2. At a separation of > 40 AU, the companion lies well beyond the plausible extent of 2M1207-39A's protoplanetary disk. This indicates that both components probably formed like stars, from the gravitational collapse of molecular gas, rather than like planets, from the accretion of rocks and gas in a circumstellar protoplanetary disk. 2M1207-39B is therefore the lowest mass brown dwarf yet identified.

2M1207-39AB also stands apart from most other very low-mass binaries. Follow-up observations of 2MASS, DENIS and SDSS sources have provided extensive cat alogues of nearby L and T dwarfs (Cruz et al, 2006; Burgasser et al, 2006). Those datasets have been surveyed for binary companions, although current observations are largely limited to high resolution imaging with either the Hubble Space Telescope (e.g. Reid et al, 2006; Burgasser et al, 2006) or ground-based AO systems (e.g. Close et al, 2003). With angular resolutions better than 0.1 arcseconds, those observations can resolve binaries with separations of a few AU, corresponding to periods of 10-50 years for massive brown dwarfs. More than 90 L dwarfs, including 72 within 20 parsecs, and ~ 25 T dwarfs have been observed to date. Spectroscopic observations are sparser (e.g. Reid et al, 2002), and are complicated by the broad absorption features of L dwarfs that limit the accuracy of the measured velocities.

The results from recent investigations have been summarised by Burgasser et al (2007). In brief, there are three main conclusions: first, the multiplicity fraction for ultracool dwarfs is probably less than 20%, with ~ 12% in resolved systems and no more than ~ 6% in spectroscopic binaries. Second, there is a clear preference for equal-mass systems. Fig. 5.10 (from Reid et al, 2006) illustrates this point, where we compare the point-spread function distribution for HST near-infrared images

Separation (arcsec)

Fig. 5.10. The HST NICMOS J-band (F110W) point-spread function, plotted in relative magnitudes, matched against the peak brightness of known companions of ultracool dwarfs. The dotted lines mark the effective detection limits. It is clear that there is a substantial area of discovery space that is accessible, but unoccupied.

Separation (arcsec)

Fig. 5.10. The HST NICMOS J-band (F110W) point-spread function, plotted in relative magnitudes, matched against the peak brightness of known companions of ultracool dwarfs. The dotted lines mark the effective detection limits. It is clear that there is a substantial area of discovery space that is accessible, but unoccupied.

against the peak brightness of known L dwarf binaries. All of the systems have nearly equal luminosities, which, for brown dwarfs, implies nearly equal masses, yet there is a vast expanse of discovery space for lower luminosity companions. Finally, the component separation distribution for brown dwarf binaries peaks at 3-10 AU, with few binaries having separations exceeding 15 AU. All of the wide binaries, including 2M1207-39AB, GG TauBab and DENIS 0551-4434AB, are younger than 108 years, raising the possibility that an evolutionary effect, perhaps dynamical stripping, might play a role in modifying the population.

Was this article helpful?

0 0

Post a comment