Nine Out of Ten Black Holes in the Milky Way Have No Companion Star — Which Is Why We Have Never Seen Them

Nearly every black hole astronomers have found in the Milky Way was found by accident of company. A black hole with a companion star gives itself away — it tears gas off its partner, heats it until it glows in X-rays, or tugs the star into an orbit that wobbles measurably. A black hole with nothing nearby does none of that. It emits no light and moves through the galaxy leaving no trace.

New work led by Tom Wagg at the Flatiron Institute in New York suggests that second category is not a curiosity but the overwhelming majority. Running a computational model of the Milky Way's history forward through billions of years, the team estimates that about 91% of stellar-mass black holes in the galaxy have no companion at all. Another 3% or so are not in the galaxy anymore, having been thrown clear by the recoil kick of the supernova that created them.

The model tracks the full life cycle: stars forming out of gas, evolving, exhausting their fuel and collapsing, with the resulting explosions delivering asymmetric kicks that shove the newborn black hole in some direction at high speed. Whether a black hole keeps its companion depends on how hard that kick lands relative to how tightly the pair is bound. Most binaries do not survive the event. The results were posted to the arXiv preprint server.

The implication reframes what the current catalog of black holes represents. The systems astronomers have spent decades cataloguing are not a random sample of the population; they are the specific subset that stayed bound to a partner, which means they are biased toward the gentlest kicks and the tightest orbits. Any inference about how massive stars die that rests on that catalog inherits the bias.

Finding the hidden nine-tenths requires a different technique. A lone black hole passing in front of a background star bends and briefly brightens that star's light — gravitational microlensing, an effect that needs no companion and no glowing gas, only a coincidental alignment. Those alignments are rare, which is why the method needs instruments that watch enormous numbers of stars at once.

Three are coming into position. NASA's Nancy Grace Roman Space Telescope is designed for exactly this kind of wide-field monitoring toward the crowded galactic bulge; the next Gaia data release will add years of precision astrometry that can catch the subtle astrometric shifts lensing produces; and large spectroscopic surveys can follow up on candidates. If the Flatiron model is right, those surveys should turn up isolated black holes in numbers that dwarf the known binaries — and if they do not, the physics of supernova kicks needs rethinking.