AN INVISIBLE COMPANION AT THE NEUTRON-STAR LIMIT
How heavy can a neutron star get? The answer depends on how matter behaves at extreme densities — what physicists call the equation of state of dense matter. Only a handful of neutron stars heavier than two Suns are known, and each one tightens the constraints. Several studies put the ceiling for a neutron star that spins slowly or not at all between 2.15 and 2.3 solar masses. Above it lies a “mass gap” between neutron stars and black holes.
A new discovery sits right on that line.
A giant that sways every 1,218 days
Xinlin Zhao, Song Wang, Jifeng Liu and colleagues, at the National Astronomical Observatories of the Chinese Academy of Sciences, Chongqing University and institutes in the United States, Italy and Australia, searched sky surveys for stars pulled around by something unseen. They cross-matched Gaia’s catalogue of binary orbits with two spectroscopic surveys, LAMOST and APOGEE, and kept the stars whose speed changed by more than 30 km/s.
One stood out: a red giant called J0008+0128, about 1,465 parsecs away. It is 5.39 times wider than the Sun, 4,929 K at its surface, and about 16 times as luminous.
The team gathered 16 infrared spectra from APOGEE, taken from December 2013 onwards, 9 optical spectra from LAMOST, and two more from the Nordic Optical Telescope in December 2025. In each spectrum, the Doppler shift of the light gives the star’s speed towards or away from us. That speed rises and falls on a cycle of about 1,218 days, swinging 16.75 km/s either side of its average, on an orbit that is nearly a circle (eccentricity 0.09).
Weighing something that cannot be seen
The speeds alone give the “mass function”: 0.587, a number written in solar masses but not itself the mass of anything. It is a fixed combination of the companion’s mass, the giant’s mass and the tilt of the orbit. The Doppler shift only records the part of the motion aimed at us, so a tilted orbit hides part of the real speed and calls for a heavier companion. Edge-on is therefore the lightest case: the companion would weigh at least 1.86 to 2.23 solar masses, depending on which estimate of the giant’s mass is used.
To measure the tilt, the team added Gaia. Its second and third data releases tracked the star’s position on the sky for about 1,000 days — roughly 82% of one orbit. Fitting speeds and positions together gives an inclination of about 73°, give or take 11°.
The giant’s own mass comes from its surface gravity and its radius: 1.67 ± 0.15 solar masses, although the different surveys spread it from 1.2 to 2.7. Put together, the companion weighs between 1.8 and 2.8 solar masses, with a best estimate of 2.20 solar masses.
Not a star, and silent
Could the companion be an ordinary star? A main-sequence star of 1.8 to 2.8 solar masses would shine 11 to 81 times brighter than the Sun — at least comparable to the giant itself. Nothing of the sort appears in its light. The team also split the infrared spectra into two possible sources, a technique called spectral disentangling: only the giant is there. A pair of white dwarfs would need an improbably narrow range of temperatures, and no close white-dwarf pair heavier than 1.4 solar masses has ever been observed.
The companion is also silent. The Einstein Probe X-ray telescope saw nothing during 9,421 seconds of observation in June 2025. The FAST radio telescope found no pulsar in four observations between February and April 2025. The authors show that both silences make sense: the giant’s wind would feed only a trickle onto a neutron star, a neutron star of 2.2 solar masses cools quickly, and a pulsar’s beam may simply never sweep past Earth.
So the companion is most likely a compact object: a neutron star at or beyond the theoretical ceiling, or a very light black hole. A tight pair of smaller compact objects hidden inside the orbit cannot be fully excluded.
An orbit nobody knows how to build
The second surprise is the orbit itself. Among the neutron-star systems the team compared, J0008+0128 has the longest orbital period. In a large population of binaries simulated with the POSYDON code, not one system ever passed through a similar state.
A simple estimate from the paper shows the problem. Start from a circular orbit and let the neutron star’s parent lose mass symmetrically in its explosion: the eccentricity then equals the mass thrown off divided by the mass the pair keeps, giant plus neutron star. An eccentricity of 0.09 means only about 0.4 solar masses were lost, so the parent was a helium star of at most about 2.6 solar masses. Recent stellar models instead make a 2.2-solar-mass neutron star from helium stars of 6 to 8 solar masses. The shove the explosion gives the newborn neutron star — its “kick” — would also have to be tiny, below about 20 km/s.
The authors also consider a three-body origin. The giant may once have circled an inner pair of stars that later merged — two neutron stars, or a neutron star and a white dwarf. A merger gives little kick, which would keep the outer orbit round. The companion’s mass is comparable to what remained after the neutron-star merger GW170817, at most about 2.3 solar masses.
The giant’s next chapter
Fast, repeated spectroscopy could reveal a hidden inner pair, through tiny short-period wobbles that the authors describe as extremely challenging to detect. Meanwhile, the giant keeps evolving. One day it may swell enough to spill matter onto its companion, and the system could end as a wide pair made of a neutron star and a white dwarf. The work has been accepted for publication in Science Advances.
