Astronomers report possible first exomoon around brown dwarf CD-35 2722 B

The system lies about 73 light-years from Earth. A star with roughly half the Sun's mass sits at its centre. Circling that star is a brown dwarf catalogued as CD-35 2722 B, an object too massive to be a planet and too light to burn hydrogen like a star. The candidate satellite orbits the brown dwarf, not the star. That makes the arrangement a three-tier hierarchy, like the Earth-Moon-Sun chain. The companion completes an orbit roughly every 170 days. The brown dwarf itself is more than 30 times as massive as Jupiter, so the companion is far lighter than the object it circles. The study was published in Nature, and ESO announced the result in October.

Brown dwarfs occupy a narrow band of possibilities. Above roughly 80 Jupiter masses an object can sustain hydrogen fusion and becomes a star; at about 13 Jupiter masses it can burn deuterium. Everything between is neither. That is why the vocabulary strains here, and why the team reached for the deliberately generic term "exosatellite". The term says only that something orbits something larger, without ruling on whether it counts as a moon.

Astronomers have catalogued more than 6,000 exoplanets and have not confirmed a single moon around any of them. The search has a structural handicap. The transit and radial velocity methods both work best on planets sitting close to their stars, and simulations suggest those close-in planets lose their moons quickly. The planets easiest to find are therefore the ones least likely to have a moon. Our own solar system offers the counterpoint: six of its eight planets host known moons. Decades of searching have produced candidates, but no confirmation.

There is also no universally accepted definition of an exomoon.

The definition gap affects how the object is classified. The CD-35 2722 candidate is massive enough to count as a planet under some classification schemes, yet it does not orbit a star directly. A confirmed detection would force astronomers to draw the planet-moon line somewhere. Whether the signal is a genuine satellite or a misreading of a complex three-body system remains open.

How the team detected it

Nobody has photographed the companion. The evidence comes from 23 usable radial velocity observations gathered between October 2023 and February 2026 with CRIRES+, a high-resolution infrared spectrograph on ESO's Very Large Telescope. As the companion swings around the brown dwarf, it pulls the brown dwarf towards Earth and then away, shifting lines in its spectrum by small, measurable amounts.

The brown dwarf sits about two arcseconds from its host star, far enough for the telescope to isolate its light.

Radial velocity works unusually well here. The brown dwarf is much less massive than a normal star, so a planet-sized companion moves it more. CD-35 2722 B also orbits about 60 times the average Earth-Sun distance from its star, which keeps the two light sources apart. Brown dwarfs are cool enough to emit most of their light in the infrared, where CRIRES+ operates.

A system that resists labels

An exosatellite is any natural satellite beyond the solar system. Exomoon usually implies a body orbiting a planet. Because this companion orbits a brown dwarf, the second term fits only loosely, and astronomers have not settled on a rule that covers it.

The masses are the first puzzle. ESO describes the companion as at least as massive as Jupiter, while the brown dwarf is more than 30 times as massive as Jupiter. The brown dwarf is only about 10 percent as massive as its host star.

Both orbits are eccentric. The brown dwarf's eccentricity is about 0.8, down from an earlier estimate near 0.9, and the candidate's is about 0.4.

How such a system formed is unclear. "We really have almost no clue actually how we form the system," Hoy said. The team considered gravitational instability, in which a massive disc breaks into fragments under its own gravity. In that picture a large fragment became the brown dwarf and a smaller one became the Jupiter-mass companion. Eccentric orbits sit awkwardly with that story, since disc formation generally produces more circular paths.

The scepticism and the open question

Hoy has been careful about the headline claim. "For now, we're not sure that we found an exomoon," he said in a SETI Live conversation. "I would say we probably haven't." He said the team is confident the object is real but is unsure what to call it.

The paper's language follows that caution. The researchers describe a possible first exomoon rather than a proven discovery, and say the detection requires further validation.

Astronomers now have a demonstrated way to hunt for satellites. It works around young, self-luminous objects that sit far enough from their stars to be studied on their own.