Stars of spectral type 'Be' are often found with neutron stars or other evolved analogues, but a black-hole companion has never been spotted before. Optical emission from a black hole's surroundings has given it away. See Letter p.378
Stellar-mass black holes, which are formed from the gravitational collapse of massive stars, are unusually scarce in our Universe. It is not yet clear whether there are fewer of them than expected or whether they are just hard to find, but either way they are deserving of their exotic reputation. Therefore, the discovery by Casares et al.1, reported on page 378 of this issue, of a stellar-mass black hole orbiting around a star dubbed MWC 656 is like finding a needle in a haystack. This black hole does not emit X-ray radiation — as black holes are expected to do — so it could be the first sign of a large population of 'quiescent' black holes.
MWC 656 is itself interesting because it is surrounded by a dense outflow from the star's equator caused by a combination of its fast rotation (the projected rotational velocity2 is about 300 kilometres per second) and pulsations that can eject material from the equator3. The resulting circumstellar disk produces spectral emission lines from hydrogen and other elements, meriting its classification as a 'B-emission' or 'Be' star. Fast rotation is a requirement in the formation of Be stars, and they probably acquire their high angular momentum during mass transfer from a massive companion star that eventually explodes as a supernova. In fact, many Be stars are found with highly evolved companions, usually neutron stars that are remnants of the post-supernova massive stars4. A few Be stars have companion stars that have been stripped down to just their helium cores5, but such a hot object is ruled out in the case of MWC 656 because there is no significant ultraviolet-light contribution coming from anywhere other than the Be star. MWC 656 is the first Be star to have a black-hole companion detected (Fig. 1).
Casares et al. have identified emission lines from helium plasma that is trapped in an accretion disk around the black hole, as well as emission from the disk around the Be star, that provide a robust measurement of the mass ratio between the star and the black hole. Such emission lines are notoriously difficult to measure: they are broad and often asymmetric, complicating the usual procedures used to measure the line centres, but the authors have taken care to reduce any underlying systematic errors. Taken together with the mass of the Be star, which is about 10–16 solar masses, the measured ratio implies that the black hole has a mass of between 3.8 and 6.9 solar masses.
In studies of stellar evolution, conventional wisdom tells us that stellar-mass black holes form during the collapse of the cores of very massive stars — those with masses more than 25 times that of the Sun6 — once the stars exhaust their fuel, and that the collapse is possibly accompanied by a supernova. The supernovae that massive stars (8–25 solar masses) undergo are expected to produce neutron star remnants instead. These massive stars tend to form within close groups of stars, so binary star systems are the norm, and triple and quadruple systems are not unusual. The catastrophe of a supernova in a binary has dramatic consequences: if more than half of the total-system mass is lost, or if 'kick' velocity from the explosion propels the newly formed supernova remnant with enough momentum, the remnant and companion star could fly off in opposite directions7. But if the companion star does remain gravitationally bound to the remnant, an X-ray binary is formed: the black hole or neutron star remnant interacts with the remaining star to produce X-ray emission.
Theorists predict that stellar-mass black holes are abundant. If this is so, we should find them all over the Milky Way. Many of them ought to be bound in X-ray binaries, whereas others should be freely floating through space. There are probably tens of millions of massive stars in the Milky Way that could potentially collapse into black holes, but there are only about 50 stellar-mass black holes known with good confidence8. X-ray studies of young star-forming regions such as the Carina Nebula, which might contain at least a few recent supernovae products, have not found any black holes9. Even large sky surveys are coming up with little as they search for the subtle brightness variations of stars whose light bends around a foreground black hole that passes in front of the star (an effect known as microlensing)10.
MWC 656 presents a rare opportunity to study mass transfer, angular momentum and accretion-disk physics around a quiescent black hole. Casares et al. find a hint of a hotspot on the black hole's accretion disk that suggests that mass is pulled away from the Be star's disk, crashing into the accretion disk when the stars make their closest approach during their orbit around one another. The absence of X-ray emission from this system is evidence that material is not channelled into the black hole; rather, it must be retained in a holding pattern within the accretion disk. Gas in the outer regions of the Be star's disk will have high angular momentum, which will be transferred to the accretion disk during the mass transfer. Without an efficient mechanism to remove this angular momentum, accretion will be suppressed and the black hole will remain quiet. If there exists a larger population of Be star–black-hole binaries, such quiescence is probably the rule, not the exception. Casares et al. have shown us a way to find them.