Planetary-Mass Satellites: First Discovery around a Brown Dwarf via Spectroscopy
Convincing evidence for the existence of moons around the brown dwarf CD-35 2722 B has been presented, obtained via the radial velocity method using VLT/CRIRES+ spectra. A periodic signal was detected, indicating at least one moon. A candidate with a minimum mass of 0.743 Jupiter masses and an orbital period of 169 days has been reliably identified. The best-fit model also includes a second, closer moon with a minimum mass of 0.277 Jupiter masses and a period of 87 days, though its parameters are less certain. The orbital periods suggest a near 2:1 mean motion resonance, reminiscent of the configuration of the Galilean moons. This first successful application of the radial velocity method for detecting exomoons opens up possibilities for studying planet formation, system dynamics, and even the search for extraterrestrial life.
Context
The discovery of exoplanets revolutionized astrophysics, but the search for their moons—exomoons—remained elusive for a long time. Why does it matter? Moons not only help unravel the history of dynamical evolution of systems but can also serve as potential havens for life thanks to tidal heating. The traditional transit method, so successful for planets, has not yet yielded unambiguous confirmations for moons. This study employed a different approach—high-precision spectroscopy, which measured tiny wobbles in the motion of the brown dwarf caused by the gravity of unseen satellites. This became possible thanks to recent upgrades of instruments that split light into the finest details, using the constancy of the speed of light as a reference. These studies allow us to peer into the distant past, reminding us that the entire diversity of present-day objects stems from primordial fluctuations after the Big Bang.
Methods
The methodology resembles the hunt for exoplanets, but with unprecedented precision. From October 2023 to January 2025, the team led by Alice Zurlo conducted 21 observing sessions of the brown dwarf CD-35 2722 B at the VLT. The CRIRES+ spectrograph, with a resolution of about 100,000, captured infrared spectra based on the laws of electromagnetism formulated by James Clerk Maxwell. Each session included nodding (AB sequence) to subtract the sky background. The spectra were calibrated using telluric lines, whose stability relies on the speed of light as an absolute standard. Although molecular bands dominate the spectrum, hydrogen lines are also present. Radial velocities were extracted using the viper code, which matches observed spectra to an empirical template built iteratively without theoretical models.
Results
The analysis revealed a clear periodic signal with a period of about 170 days, exceeding the 0.1% false-alarm threshold in the Lomb-Scargle periodogram. The best-fitting two-satellite model, obtained with the EMPEROR code, predicts a large satellite with a minimum mass of 0.743 Jupiter masses (M sin i) on a nearly circular orbit (eccentricity <0.01) and a period of 169.45 days. The semi-major axis is about 0.2 AU, well within the Hill sphere and Roche limit. A second candidate satellite has a period of 87.46 days, a minimum mass of 0.277 Jupiter masses, and also moves on a circular orbit at a distance of 0.13 AU. The periods of these two bodies are in a 2:1 resonance, reminiscent of the Galilean moons of Jupiter. The model with a single eccentric satellite (eccentricity 0.29) fits the data worse, with a Bayesian evidence log difference ΔlogZ = 6.9 in favor of the two-satellite scenario. Crucially, the rotation of the brown dwarf itself, with a maximum period of 0.65 days, cannot account for the observed variations. This success was made possible by the high precision of spectroscopy, which, unlike the transit method, does not depend on geometric alignment. The Doppler shifts of lines are proportional to the radial velocity, which directly follows from the constancy of the speed of light.
Implications
The discovery of massive satellites around a brown dwarf has profound implications for formation theories. Massive satellites are more naturally formed via gravitational instability in the protoplanetary disk, consistent with how CD-35 B itself likely formed as a wide companion. The mass ratios of the satellites to the host (2% and 0.7%) far exceed those in the Solar System, where even the Earth-Moon system reaches only 1.2%. This supports the idea that moon formation mechanisms can vary drastically depending on the mass of the central body and its environment. Moreover, the 2:1 resonance points to dynamical evolution involving migration and resonant capture.
Future development
Future observations at even higher resolution, perhaps with upcoming telescopes like the ELT, will not only confirm the second satellite but possibly detect smaller bodies. Advances in direct imaging and interferometry could yield direct separation of the satellites and their spectra. Of particular interest is the study of the atmospheres of these satellites: if they are cold enough, they may host lines such as the hydrogen Balmer series, discovered by Johann Balmer in the 19th century. Next-generation gravitational wave detectors might even catch mergers of such massive objects if they spiral together.
Impact
This discovery will impact several areas: the dynamics of hierarchical systems, models of planet and moon formation, astrobiology (the search for habitable moons), and the instrumental development of high-precision spectroscopy. Hubble and its successors will likely include such targets in their programs to search for transits and study atmospheres.
Next steps
Key steps include additional observations at other epochs to resolve the period degeneracy of the second satellite, detailed characterization of the atmospheres of the brown dwarf and satellites, and searches for similar signals in other systems with wide substellar companions.
Key open problems
This research directly ties into unresolved questions: How do substellar companions with wide orbits form? What are the universal patterns of satellite formation? What is the maximum mass of a moon, and where is the boundary between a planet and a satellite? Determining the nature of objects in hierarchical systems touches on fundamental problems in the evolution of the Universe from the Big Bang to the formation of chemical elements.
🎯 The brown dwarf CD-35 B has a mass of 37 Jupiters, making it almost a star. And the satellite with a minimum mass of 0.743 Jupiter masses is itself heavier than some known exoplanets!
🎬 Fictional worlds like Pandora from 'Avatar' or Europa from '2001: A Space Odyssey' have long portrayed moons as habitable oases. The discovery of real massive exomoons brings these dreams a bit closer.
Key numbers
- minimum mass of the large satellite: 0.74 Jupiter masses
- orbital period of the large satellite: 169 days
- mass of the brown dwarf: 37 Jupiter masses
- semi-major axis of the large satellite's orbit: 0.2 AU
- period of the second satellite: 87 days