Authors: Julia Bodensteiner et al.
First Author’s Institution: European Southern Observatory (ESO)
Status: Published on ESO’s The Messenger [open access]
Unraveling the mysteries behind the fates of the most massive stars is key to understanding the present state of the universe. This is because massive stars are origins of elements heavier than helium, as a result of thermonuclear interactions in their cores; as well as being sources of electromagnetic radiation, strong stellar winds, and supernovae, which help seed these elements throughout the cosmos. Since the most massive stars end their lives as black holes, understanding the distribution and characteristics of these objects is key to understanding the lifecycle of said stars.
The problem with searching for black holes, however, is that by their very nature, they are nearly impossible to detect on their own. We infer their existence through two main techniques. The first is when a black hole accretes material from a stellar binary companion – this gas and dust can form an accretion disk around the black hole, heating up and emitting x-ray radiation. The second is from the detection of gravitational waves that occur when a black hole merges with another compact object.
In our galaxy, we have detected around 100 or so black holes from X-ray binaries. However, since we expect most massive stars to end their lives as black holes, theories suggest we should see ~107 stellar-mass black holes in the Milky Way. As such, it is suspected that the vast majority of black holes are what we call quiescent – that is, they do not accrete enough to show up on x-ray observations, and thus can only be detected via gravitational effects on other nearby bodies.
To date there have been only a handful of reported candidate quiescent black holes. These have all been in binary systems, whose initial signature was detected through spectroscopy and radial velocity measurements. In short, if looking at the light spectrum of a star shows its spectral lines varying sinusoidally, as if orbiting a companion, but there are no appropriate lines that vary in opposite cadence, it could represent an unseen companion that does not emit light – such as a black hole.
The authors of today’s paper, however, caution that a black hole need not be the only explanation for this. There could instead be a companion star that is emitting light, but is not detected due to low-quality data or being relatively faint compared to the much brighter companion. Alternatively, it could be rotating so fast that its spectral lines are broad and shallow, and thus are much less distinguishable, among other theories.
In particular, the authors look in detail at two systems – LB-1 and HR 6819 – whose initial spectra prompted them to be reported as quiescent black holes orbited by a B-type star (luminous, blue, and usually more massive than the sun). However, subsequent analyses have proposed that they are instead binary star systems that consist of a B-type star whose atmosphere has been stripped, and another luminous star.
The spectra of LB-1 and HR 6819 both share similar features that are shown in Figure 1. In particular, in this wavelength region there are two bright, stationary, broad emission lines, and two dark, narrow, shallow absorption lines, which vary sinusoidally in time. The absorption lines are those of the B-type star, and vary on a scale of tens of days. The emission lines are instead characteristic of a classical Be star – a specific kind of B-type star that contains an emitting circumstellar gaseous disk.
The initial hypothesis was that since the Be emission lines appear stationary, the B-star and Be star did not orbit one another closely. Either the B-star must orbit with an invisible companion, and the Be emission corresponded to either an unrelated third star which appeared in the spectra due to chance superposition, or this was a triple star system and the Be-star orbited much further away. These ideas were backed up by calculations which showed that if the B-type star had its typical mass of ~5 solar masses, the Be star would need to have an unphysical mass to have such an effect on the radial velocity of the B-type star’s spectral lines.
The Stripped Star Solution
However, subsequent studies have since suggested that such a triple system would most likely be unstable. Furthermore, the Be star’s emission lines do in fact seem to show a very small, subtle variation in opposite cadence to the B-star’s absorption lines; thus the two stars could in fact be orbiting one another, removing the need for a third, invisible companion. If this is the case, then, how do we justify the non-physicality of the Be star’s estimated mass? The projected orbital velocity of the B-type stars, based on the movement of their spectral lines, is much larger than it should be in comparison to that of the Be stars, in both the LB-1 and HR 6819 systems.
We can resolve this by reinterpreting the physical nature of the B-type star. If we assume its mass to be on the order of ~0.5 solar masses, rather than the typical ~5-6, the radial velocities would make sense. Under this interpretation, the B-type star is not a standard main-sequence B-type star, but is instead in a “post-mass-transfer” phase – a star that has been fully ‘stripped’ by its binary companion, losing the majority of its mass, while its companion accreted all that matter and angular momentum, spinning up into a rapidly rotating Be star. As the outer hydrogen layers were stripped away from the B-type star, its now exposed Helium core would puff outwards and re-contract into a new equilibrium phase. In the early contraction phase, its luminosity and surface temperature can appear to overlap with those of a typical main-sequence B-type star, and thus it could be easily confused for one. If this is the case, then said B-type star would continue contracting over millions of years, eventually becoming a sub-dwarf OB star. The start of the contraction phase, however, is the brightest and most easily spectroscopically detectable phase of this evolution, and so it makes sense that these systems are detected in this phase.
The authors posit that high-resolution interferometry might be the definitive way to determine whether these systems contain quiescent black holes or stripped B-type stars. The binary scenario has the two companions orbiting at 1-2 milliarcsecond separations, with an orbital period on the order of tens of days. On the other hand, the triple scenario should have the Be star orbiting much further apart, and appearing stationary on month-long timescales. Initial observations of HR 6819 from the GRAVITY instrument at the VLT Interferometer seem to favor the binary hypothesis, and further observations in April-September of 2022 will allow for the derivation of stellar parameters such as an accurate mass of the stripped B-type star. Similarly, GRAVITY observations of LB-1 are planned for this year, and will hopefully shed light on the nature of that system as well.
With the possible elimination of these two candidates, however, the search for the missing quiescent black holes continues, and the authors hope that the lessons learned from these two systems pushes for an interdisciplinary approach to finding and characterizing these objects, in particular with ongoing and upcoming large scale surveys on the horizon.
Astrobite edited by Yoni Brande
Featured image credit: NASA/Wikimedia Commons