Discovery of a Planet-Scale Moon around a Brown Dwarf ✛ ✎ ✦
Context
Since the first exoplanet around a Sun-like star was discovered by Michel Mayor and Didier Queloz in 1995, the search for moons around exoplanets—exomoons—has been a top priority in observational astronomy. These objects can hold key information about planet formation mechanisms and dynamical evolution, and could potentially expand habitable zones. Despite a few candidates, no confirmed detection had been made—until now. CD-35 2722 B is a young brown dwarf on a well-separated orbit around its host M-dwarf star, making it an ideal target: the large angular separation and favorable contrast avoid contamination from stellar light. Applying the radial velocity method, based on the Doppler effect, to the brown dwarf's own spectrum led to the first detection of a periodic signal caused by the orbital motion of a moon.
Methods
Observations were carried out in 2023–2025 using the CRIRES+ spectrograph (spectral resolution ~100,000) on the VLT in AB-nodding mode for sky subtraction. The spectrograph slit was oriented perpendicular to the star's direction, effectively preventing stray light. Radial velocities were extracted with the viper code, which builds an empirical spectral template of the object without relying on theoretical models, and were averaged over nod pairs to improve precision (typical error ~31 m/s). Orbital analysis was performed with the EMPEROR code, which uses Markov chain Monte Carlo (MCMC) numerical modeling of orbits. Keplerian models were fitted iteratively, starting with a single moon and adding components until they significantly improved the Bayesian evidence. Additionally, the Roche limit and Hill radius were computed, accounting for the variable distance to the star, to confirm the dynamical stability of the orbits within the law of universal gravitation.
Results
Lomb–Scargle periodogram analysis revealed a highly significant peak at a period of ~170 days (false alarm probability <0.1%). The best-fit model includes two moons on nearly circular orbits. The first has a period of 169.5 days, a minimum mass of 0.74 Jupiter masses, and a semi-major axis of 0.20 AU. The second, with a period of 87.5 days and a minimum mass of 0.28 Jupiter masses (semi-major axis 0.13 AU), is near a 2:1 resonance with the first. Both periods are determined with high precision, but the second moon has alternative solutions (periods around 70 and 115 days) due to an incomplete phase curve. A single-moon model requires a high eccentricity (0.29), which is mathematically degenerate with two resonant bodies. Comparison with the expected rotation period of the brown dwarf (<0.65 days) rules out rotational modulation as the signal source. The calculated Roche limit (8.4 brown dwarf radii) and Hill radius (>1 AU) confirm that the orbits lie well within the stability zone.
Implications
This discovery provides the first direct evidence of planet-scale exomoons and has profound implications for planet formation theories. The system likely formed through gravitational instability in the disk around the brown dwarf, explaining the anomalously high moon-to-host mass ratio (up to 2%). This contrasts with core accretion, which dominates in our solar system and produces much smaller moons. The presence of a 2:1 orbital resonance points to significant migration early in the system's evolution. It also calls into question the definitions of planets and moons in hierarchical systems, where planetary-mass objects orbit substellar bodies.
Future development
Full characterization of the system requires additional spectroscopic epochs to unambiguously determine the second moon's period and measure eccentricities. In the future, direct imaging of these moons with next-generation instruments (e.g., METIS on the ELT) will provide their radii and temperatures. New hydrodynamical models of disks around brown dwarfs will help understand the conditions that birth such massive moons. Advancements in high-resolution spectroscopy of exomoons will pave the way for searching for smaller, potentially habitable bodies.
Impact
These findings will impact the physics of planet formation, dynamics of gravitationally bound systems, astrobiology (by expanding the concept of habitable zones), and observational astronomy (boosting moon searches around wide-orbit companions).
Next steps
Immediate steps include scheduling dedicated observations of CD-35 2722 B to fill in the phase curve and mining archival data from other instruments. On the theoretical side, incorporating moon systems into population synthesis models is a priority.
Key open problems
The discovery sharpens the problem of massive moon formation: why does our solar system lack anything above 0.1 Jupiter masses, even though gravitational instability models predict such objects should be common? It also forces a reexamination of the boundary between planets and moons in light of recent findings of free-floating planets and binary systems with extreme mass ratios.
The knowledge machine's view
This section is addressed to the author of the paper and their colleagues. Neighbouring works were found by meaning-based search across our archive, not by word overlap. We do not review: we show the field around the work.
What the work does
For the first time, a planet-scale companion has been detected around the brown dwarf CD-35 2722 B using the radial velocity method. Observations with CRIRES+ revealed a periodic signal with a period of ~170 days, which is best explained by two companions with masses of 0.74 and 0.28 Jupiter masses, near a 2:1 resonance. The authors ruled out alternative sources of the signal, such as the dwarf's rotation, and confirmed the dynamical stability of the orbits.
Where its strength is
This work demonstrates the feasibility of directly detecting exomoons via radial velocities around objects with favorable contrast, paving the way for systematic searches. The use of an empirical spectral template (viper) and Bayesian orbital modeling (EMPEROR) sets a high standard for analyzing weak signals.
Where it could go next
- It might be interesting to compare the derived system parameters with the results of the first spectroscopic detection of an exomoon described in a neighboring paper, to identify common patterns in moon formation around brown dwarfs.based on: Waltz of Invisible Moons: The First Spectroscopic Find arXiv:2607.05193v1
- Nearby is an approach to finding planets around white dwarfs via spectral feature analysis; similar techniques might help refine the properties of the host star and the dynamics of the CD-35 2722 system.based on: How to Hear a Planet Near a Dead Star arXiv:2605.14077
- One could look at work on the brown dwarf 'desert' from Gaia data to assess how typical this system is and how it fits into the overall population.based on: Dance of the Invisible Giants: Gaia Finds a 'Desert' of Brown Dwarfs arXiv:2601.03539v1
Worth trying
- Check whether the second companion's signal appears in archival data from other instruments to break the period degeneracy.
- Apply a joint analysis of radial velocities and astrometry if Gaia can resolve the dwarf's orbital motion.
- Search for additional, lower-mass companions using longer observation baselines and algorithms like EMPEROR.
Why this may matter
Confirming exomoons around brown dwarfs is important for understanding planet formation in environments different from our solar system, and could expand the search for habitable worlds beyond traditional zones.
Nearby works in our archive
- Waltz of Invisible Moons: The First Spectroscopic Findsimilarity 0.67
- How to Hear a Planet Near a Dead Starsimilarity 0.61
- Dance of the Invisible Giants: Gaia Finds a 'Desert' of Brown Dwarfssimilarity 0.61
- Catching a Sunset on a Distant Planetsimilarity 0.60
- The Hunt for Moons of Rogue Planetssimilarity 0.59
- A Sun-Sized Lens Searches for Life Around Distant Starssimilarity 0.59
The closest work in the archive sits at 0.67 by meaning — the usual level for our corpus.
The mass ratio of the largest moon to its host is ~2%, nearly twice the record in our solar system — the Earth-Moon system (1.2%). For comparison, Jupiter's largest moon, Ganymede, has a ratio of just 0.04%.
In science fiction, gas giant moons often serve as exotic settings — think Pandora in 'Avatar.' The discovery of a real world with an unusually massive moon makes such scenarios a little less hypothetical.
Key numbers
- period of large moon: 169 days
- minimum mass of large moon: 0.74 M_Jup
- period of small moon (candidate): 87.5 days
- minimum mass of small moon: 0.28 M_Jup
- mass of host brown dwarf: 37 M_Jup
- semi-major axis of large moon's orbit: 0.2 AU