Binary and multiple star systems result from the fragmentation of dense material in young molecular clouds. Observations reveal that this can occur on small scales, supporting a previous model of star formation. See Letter p.483
When we look at the night sky with the naked eye, most stars look like single points of light. Taking a closer look using binoculars, one can quickly find two or more light sources in close proximity — a binary or multiple star system. As the sensitivity and resolution of telescopes have improved, astronomers have been able to determine the multiplicity of different types of star. For instance, about half of all Sun-like stars have at least one companion1. On page 483, Tobin et al.2 study a system of three young stars in close proximity to one another and provide evidence for a mechanism that explains how such systems are formed.
The process of star formation is extremely short relative to the lifetime of a star. Young stellar objects called protostars are created from the fragmentation of gas and dust in molecular clouds, and start burning hydrogen in their cores a few million years later3. Multiplicity is common at these early stages — an outcome of the star-formation process4. Astronomers therefore search for the earliest examples of binary or multiple star systems to better understand the mechanisms that give rise to Sun-like stars.
Studies of newborn star systems require highly sensitive astronomical instruments, because protostars are cool, low-mass (and therefore faint) objects. The instruments also need to have high resolution because the protostars in these systems are in close proximity to one another. Interferometers that detect light of wavelengths between several tenths and a few millimetres are now able to study systems whose protostars are separated by as little as several astronomical units (1 AU is Earth's distance from the Sun) at the distance of our nearest protostellar neighbours.
How exactly do binary and multiple star systems form? Theoretical studies predict two main mechanisms that can be distinguished, most generally, by the distances over which the fragmentation of gas and dust occurs5,6. More specifically, the first mechanism can be described as large-scale fragmentation that is caused by the turbulence of gas in the molecular cloud. By contrast, the second mechanism is characterized by small-scale fragmentation, resulting from gravitational instabilities developing in dense material that has formed a small disk in the molecular cloud.
Large-scale fragmentation was first directly observed7 in the Perseus star-forming region. A concentration of gas and dust in this region, known as Barnard 5, was shown to be a young quadruple protostar system that is on track to form a binary system whose members will be widely separated (more than 1,000 AU apart). At least two members of the original arrangement are likely eventually to become unbound (or be ejected) from the system. However, observationally confirming the small-scale mechanism required much higher resolution than was used in previous studies7. To achieve this, Tobin and colleagues used a powerful interferometer, the Atacama Large Millimeter/submillimeter Array (ALMA) in northern Chile.
The authors observed the triple protostar system L1448 IRS3B (Fig. 1), which, like Barnard 5, is located in the Perseus region. Near the centre of this system are two protostars that are separated by 61 AU. Following a spiral arm outward, the third member resides at a distance of 183 AU from the central-most protostar. This complicated structure was revealed thanks to the ten times higher sensitivity and two times higher resolution provided by ALMA compared with previous observations2. The disk that surrounds these objects acts as an intermediate reservoir through which material from the disk's outer envelope can move inward and eventually transfer mass to the forming stars.
Tobin and collaborators' results represent the first direct observational evidence that small-scale disk fragmentation can lead to the formation of binary and multiple star systems. The authors observe two of the expected outcomes of gravitational instability of a protostellar disk: namely, a hierarchical configuration of protostars and a spiral-shaped disk.
The strength of this study is the attention to detail in Tobin and colleagues' analysis. For instance, the authors show that the disk of the protostar system is gravitationally unstable by calculating the 'Toomre Q parameter' (ref. 8). Specifically, they determine the size at which the disk would fragment because of a gravitational instability, and conclude that it would be unstable at radii of between 150 and 320 AU. This range agrees with the observed separation between the two protostars at the centre of the system and the third member in the spiral arm. Moreover, the authors find that this fragmentation probably occurred recently (within the past few thousand years), which is consistent with the young age of the protostar system.
It is important to emphasize that both of the fragmentation mechanisms are plausible and expected, rather than being mutually exclusive. Determining the frequency at which each mechanism occurs will require follow-up studies. Fragmenting disks like the one observed by Tobin and colleagues are probably not rare — rather, they are waiting to be studied in more detail using the powerful (sub-)millimetre-wavelength telescopes that are now available.Footnote 1
Raghavan, D. et al. Astrophys. J. Suppl. 190, 1–42 (2010).
Tobin, J. J. et al. Nature 538, 483–486 (2016).
Evans, N. J. II et al. Astrophys. J. Suppl. 181, 321–350 (2009).
Mathieu, R. D. Annu. Rev. Astron. Astrophys. 32, 465–530 (1994).
Duchêne, G. & Kraus, A. Annu. Rev. Astron. Astrophys. 51, 269–310 (2013).
Reipurth, B. et al. in Protostars and Planets Vol. VI (eds Beuther, H. et al.) 267–290 (Univ. Arizona Press, 2014).
Pineda, J. E. Nature 518, 213–215 (2015).
Toomre, A. Astrophys. J. 139, 1217–1238 (1964).