Gravitational-wave catalog reveals long-predicted ‘missing’ range in black hole masses

On some nights, the most massive stars in the universe end their lives in explosions so violent that nothing is left to collapse into a black hole. For decades, astronomers have predicted that this extreme kind of stellar death should carve a missing band into the black hole mass spectrum — a range where nature simply does not make them.

Now, an international team says it has found that missing band.

In a study published April 1 in Nature, researchers led by Hui Tong of Monash University in Australia report strong evidence for a long‑predicted “pair‑instability” gap in the masses of stellar‑origin black holes. By analyzing dozens of black hole mergers detected through gravitational waves, the team reconstructed how heavy black holes tend to be and found a striking shortage in a specific mass range.

“Stellar theory predicts a forbidden range of black‑hole masses between approximately 50 solar masses and 130 solar masses owing to pair‑instability supernovae,” Tong and colleagues write. “Here we report evidence of the pair‑instability gap in LIGO–Virgo–KAGRA’s fourth Gravitational‑Wave Transient Catalog.”

The work draws on GWTC‑4, the latest catalog of gravitational‑wave events compiled by the LIGO, Virgo and KAGRA observatories. These laser interferometers in the United States, Italy and Japan listen for tiny ripples in spacetime produced when pairs of black holes spiral together and collide.

Using data from GWTC‑4 and earlier catalogs, Tong’s team examined the masses of the two black holes in each binary system. In each merger, the heavier object is labeled the primary (m₁) and the lighter the secondary (m₂). When they combined information from the entire sample and corrected for selection effects — the fact that heavy systems are easier to detect — a pattern emerged.

The distribution of secondary masses showed a clear deficit between about 44 and 116 times the mass of the sun. Tong’s group estimates the lower edge of this gap at roughly 44 to 45 solar masses, with a quoted value of 44 plus 5 minus 4 solar masses at 90% credibility.

“The gap is not present in the distribution of primary masses m₁,” the authors write. “It appears unambiguously in the distribution of secondary masses m₂.”

That distinction — primaries versus secondaries — is crucial to how scientists interpret the result.

Stars too massive to leave black holes

The newly reported gap lines up with a theoretical concept known as the pair‑instability mass gap.

In normal core‑collapse supernovae, a massive star exhausts its nuclear fuel, its core collapses, and the outer layers are blasted away, often leaving a black hole or neutron star behind. But in stars born with far greater masses, conditions in the core can become so extreme that high‑energy photons — gamma rays — transform into pairs of electrons and their antimatter counterparts, positrons.

This process reduces the radiation pressure that helps support the core against gravity. The core contracts, heats up further, and can trigger runaway burning of oxygen. Depending on the star’s mass, one of two outcomes is expected. In slightly less extreme cases, the star experiences violent pulses that eject large amounts of mass — a pulsational pair‑instability supernova — so that the eventual remnant is a smaller black hole than otherwise expected. For higher masses, models predict a full pair‑instability supernova that completely disrupts the star, leaving no black hole at all.

As a result, many models of massive star evolution predict that black holes should not be born with masses in a broad range, often estimated between roughly 50 and 130 solar masses, with the exact edges depending on factors such as metallicity, rotation and uncertain nuclear reaction rates.

Astronomers have been searching for this gap in observational data for years. Early gravitational‑wave catalogs hinted at a cutoff around 45 to 60 solar masses, but those signs became murkier as more heavy mergers turned up, including objects apparently inside the forbidden zone.

A 2020 event known as GW190521 featured a primary black hole about 85 times the mass of the sun. When first announced, that object was widely described as an “impossible” or “forbidden” black hole because it appeared to fall within the pair‑instability gap. A 2023 event, GW231123, went further: both components weighed in at roughly 100 solar masses, squarely in the predicted no‑go zone for black holes formed from single stars.

Those mergers forced astronomers to consider alternative formation routes, such as “hierarchical mergers,” where black holes grow in dense star clusters by colliding with one another multiple times.

Tong’s analysis suggests that those exotic‑sounding routes may be exactly what is happening — and that, once the full population is considered, the basic pair‑instability prediction holds.

A tale of two black hole families

By separating primary and secondary masses and incorporating spin information, the team finds evidence for two distinct subpopulations of merging black holes.

Systems with both components below the gap tend to have relatively low “effective spin,” a parameter that combines the spins of the two black holes along the axis of their orbit. That is broadly consistent with expectations for black holes born from the collapse of single, massive stars, which some models suggest may not spin rapidly at merger.

In contrast, systems whose primaries lie in or above the gap have significantly higher effective spins. That pattern, Tong and others argue, fits with black holes that are themselves the remnants of previous mergers. When two black holes combine, the resulting object is predicted to spin rapidly. If that remnant later merges again as part of a new binary, it can show up as a high‑spin primary in the mass gap.

“The new data reveal two populations: a low‑spin group with no black holes above the gap and a high‑spin group that extends across the full mass range and occupies the gap,” astrophysicist Fabio Antonini and colleagues wrote in a separate 2025 preprint that also used GWTC‑4 data to infer a gap with a lower edge near 45 solar masses.

What Tong’s Nature paper adds, experts say, is a cleaner view of where the deficit lies — in the masses of secondary black holes — and a more direct link between the observed gap and the physics of how massive stars burn and explode.

Because the gap shows up most clearly among secondaries, the data suggest that nature rarely produces a lighter companion black hole in the gap through direct stellar collapse. Instead, the black holes inside the gap appear mostly as primaries, consistent with heavier, merger‑built remnants.

Gravitational waves as a nuclear physics tool

The location of the pair‑instability gap does more than confirm a piece of stellar astrophysics. It also encodes information about the nuclear reactions that power the last stages of a massive star’s life.

One of the most important of those reactions is the fusion of carbon‑12 and helium‑4 to form oxygen‑16, written ^12C(α,γ)^16O. The rate at which this reaction proceeds, especially at energies around 300 kilo‑electronvolts, helps determine how large a star’s carbon‑oxygen core becomes and where the boundary between pulsational and full pair‑instability lies.

Measuring this reaction rate in terrestrial laboratories is extremely challenging. Experiments must probe very low cross‑sections at energies where background noise is high. As a result, the ^12C(α,γ)^16O rate remains one of the largest uncertainties in stellar evolution calculations.

Using their mass‑gap measurement as a constraint on stellar models, Tong’s team derives an estimate for the reaction’s so‑called S‑factor at 300 keV of about 260 keV‑barn, with sizable error bars. While that will not settle the nuclear physics debate on its own, it provides an independent, astrophysical piece of evidence that can be compared with laboratory measurements.

A maturing field, and unresolved questions

The claim of a pair‑instability gap from gravitational‑wave data is not entirely new. Antonini and other groups have published preprints over the past year arguing for “strong evidence” of a gap starting near 45 solar masses, using similar GWTC‑4 data. At least one reanalysis has questioned how robust the evidence is to different modeling assumptions.

What makes the new Nature paper notable is the combination of a high‑profile, peer‑reviewed venue; a focus on secondary masses and spin that sharpens the physical interpretation; and public release of the analysis code, which allows others in the community to test and build on the result.

Even so, some uncertainties remain. The exact boundaries and depth of the gap depend on how scientists model the underlying black hole population and the sensitivity of the detectors. Alternative formation channels, such as primordial black holes born in the early universe, have also been proposed as possible contributors to objects inside the gap, though most researchers see them as speculative at this stage.

As LIGO, Virgo and KAGRA continue their fourth observing run and plan future campaigns, the catalog of black hole mergers is expected to grow into the thousands. Next‑generation observatories such as the planned Einstein Telescope in Europe and the Cosmic Explorer projects in the United States aim to increase sensitivity by an order of magnitude, potentially allowing astronomers to track how the mass gap and black hole populations evolve over cosmic time.

For now, Tong’s study suggests that an absence can be as informative as any individual detection. In the masses of black holes stitched together from ripples in spacetime, scientists are beginning to see the outline of stars so massive that they skip making black holes altogether — leaving behind, in the data, a conspicuous emptiness where black holes should have been.