Last year, Ugo Di Carlo (former postdoc in our group, now an assistant professor at SISSA in Italy), wrote a paper arguing that the detached black hole-star binaries, like the first two identified in data provided by the Gaia mission, were likely formed in young star clusters in the Milky Way.
For context, this comes back to how you detect black holes in the first place. Traditionally, people have relied on either X-ray emission from the black holes (as they accrete gas from a companion star), or gravitational-wave emission from two black holes that merge. But staring in 2018, people were able to use the motion of a companion star in orbit of the black hole to infer the presence of the black hole. The first three black holes were detected this way (using the “radial velocity” method, the same that is used to find exoplanets) in the globular cluster NGC 3201 (see Giesers et al. 2018, 2019).
The next two black holes, however, were found by the proper motion of their companions (watching them orbit in 2D across the sky over time). They were found very close to Earth (relatively speaking) by a team lead by El Badry et al. (2023a, 2023b). But unlike the black holes in NGC 3201 (which we argued were formed dynamically), it’s not obvious that you can form Gaia BH1 and BH2 from isolated binary stars.
To explore whether the two Gaia black holes could have instead formed in open clusters, Ugo used a set of N-body models of disrupting clusters from his PhD thesis to simulate a full Milky Way of star clusters. The details of the paper can be found here (Di Carlo et al., 2024), but the punch line is that we found star clusters are nearly 50 times as efficient at producing detached black hole-star binaries as isolated binary evolution.
There are a lot of different dynamical and binary processes that can produce these systems. Take a look at a couple from Ugo’s paper here:
More recently, the Gaia collaboration themselves have announced the detected of a new Gaia black hole. Unlike the first two, Gaia BH3 has a mass of nearly 33 solar masses, fully consistent with the peak at ~30 solar masses for black holes coming from the LIGO/Virgo gravitational-wave data. It’s even consistent with being in the core of a disrupted globular cluster, suggesting that the origin of LIGO’s 30 solar mass black holes might finally be close to being settled!