The universe, it turns out, is actively recycling its most violent debris. A landmark series of studies using the latest data from the international LIGO-Virgo-KAGRA Collaboration has revealed that a substantial fraction of merging black holes are not born directly from dying stars. They are second-generation black holes, forged from the remnants of earlier collisions, carrying the spin signature of their violent birth and hunting for new partners in the densest, most chaotic corners of the cosmos. MIT physicists Cailin Plunkett and Salvatore Vitale applied advanced statistical modeling to 155 merging pairs from the Gravitational-Wave Transient Catalog 5 and found that roughly 14 percent of all detected black hole mergers match this recycled, second-generation profile. The results were published in Physical Review Letters.
GWTC-5 Data | 400 Mergers and Two Hidden Populations
The breakthrough became possible because of scale. With the release of GWTC-5, the total number of recorded gravitational-wave events has grown to nearly 400 detected cosmic mergers. Rather than treating each signal as an isolated anomaly, the MIT team applied population-level statistical modeling across 155 merging pairs, searching for patterns invisible in any single event but detectable across the full dataset. The analysis isolated two starkly different sub-populations hiding in the data. The first is a stellar cohort of smaller, slowly spinning black holes that obey a strict mass ceiling, consistent with direct single-star collapse. The second is a recycled cohort of much heavier black holes that spin far faster and frequently appear at masses that stellar physics forbids them from reaching through any single-star death pathway. The statistical separation between these two populations is sufficiently clean that the team could characterize their properties independently and estimate what fraction of total detected mergers each accounts for.
The Pair-Instability Mass Gap | Why Impossible Black Holes Keep Appearing
This discovery provides an elegant solution to one of the most persistent paradoxes in gravitational-wave astronomy. Nuclear physics predicts a pair-instability mass gap between approximately 50 and 130 solar masses. Extraordinarily massive stars in this range generate cores so hot they produce electron-positron pairs, triggering a sudden pressure drop that ignites a runaway thermonuclear explosion called a pair-instability supernova. This explosion completely obliterates the star, leaving no remnant. No black hole of 50 to 130 solar masses should ever form from direct stellar collapse. Yet LIGO, Virgo, and KAGRA have repeatedly detected merger events whose component black holes sit squarely inside this forbidden mass range.
Hierarchical merging explains the paradox without requiring any revision of nuclear physics. The universe is not producing these mass-gap objects from single stars. It is building them incrementally by smashing together smaller, legitimately formed first-generation black holes that each sit below the 50 solar mass ceiling. Two 30-solar-mass objects colliding produce a roughly 57-solar-mass remnant, well inside the gap, without violating any constraint on stellar physics. Repeated collisions can then push the descendant further up the mass ladder, producing objects at any mass above the initial first-generation ceiling. The recycled cohort the MIT team identified in GWTC-5 is direct observational evidence that this incremental assembly line is operating across the cosmos.
Dense Stellar Environments | Where Hierarchical Mergers Happen
For a remnant black hole born from a collision to find a second partner and merge again, the environment must be extraordinarily crowded. In the vast emptiness of interstellar space, a newborn remnant would simply drift away from any other object, with essentially no probability of encountering a partner within any cosmologically relevant timescale. The 14 percent hierarchical merger rate identified in GWTC-5 therefore implies that a specific class of dense stellar environments is producing these events at significant rates. The two primary candidates are globular clusters and the accretion disks surrounding supermassive black holes at galactic centers. In globular clusters, stellar densities can be high enough that hundreds of black holes are packed within a few light-years, enabling dynamic capture, pairing, and repeated collision cycles. Galactic center accretion disks provide a similar density environment with the additional mechanism of disk torques that can herd black holes into co-planar orbits and accelerate their eventual merger.
As lead author Cailin Plunkett noted, "We are finding that, for some of these merging black holes, it's not their first rodeo." With LIGO, Virgo, and KAGRA continuing to increase operational sensitivity throughout 2026, researchers expect the growing catalog to reveal deeper levels of the hierarchy. Third- and fourth-generation black holes, objects that have survived multiple rounds of collisions over billions of years, may already be present in the data waiting for the statistical power to isolate them. For related OzoneNews coverage of gravitational-wave astronomy and related space physics, see our reporting on the CUNY lab recreation of black hole energy extraction, the Milky Way black hole wind discovery at Sagittarius A, the JWST galaxy-killing wind discovery in the early universe, and the Euclid telescope ancient quasar detections.