When the Transiting Exoplanet Survey Satellite (TESS) scanned a faint red-dwarf pair in 2019, astronomers noticed periodic dips in brightness that hinted at planets. The binary, catalogued as TOI-2267, lies within the constellation Canis Minor and consists of two cool M-type stars.<\/p>\n
Analysis of the TESS light curves revealed three repeating signals, suggesting the presence of planets roughly the size of Earth. Two signals, now named TOI-2267 b and TOI-2267 c, repeat every 2.28 days and 3.49 days. A third signal at 2.03 days, labelled TOI-2267.02, remains a strong but unconfirmed candidate.<\/p>\n
\u201cOur analysis reveals a unique planetary configuration: two planets pass through one of the stars, and the third passes through its companion,\u201d explained Sebasti\u00e1n Z\u00fa\u00f1iga-Fern\u00e1ndez of the University of Li\u00e8ge. \u201cThis makes TOI-2267 the first known binary system hosting planets in transit around each of its stars.\u201d<\/p>\n
If verified, this geometry would mean that both stars in the system are orbited by planets crossing their faces from Earth\u2019s perspective, a configuration astronomers had never before observed.<\/p>\n
How the team proved the planets were real<\/p>\n
Initial signals alone were not enough. TESS pixels are wide, each spans about 21 arcseconds, making it easy for background stars to mimic planetary transits. To confirm the findings, the researchers combined space data with precision observations from ground observatories.<\/p>\n
The team used SHERLOCK, a detection software developed by ULi\u00e8ge and the Instituto de Astrof\u00edsica de Andaluc\u00eda (IAA-CSIC), to verify the periodic patterns. Ground-based telescopes, including SPECULOOS, TRAPPIST, SAINT-EX, and the Las Cumbres Observatory Network, recorded follow-up transits in different filters. These instruments are robotic and optimized for faint, cool stars.<\/p>\n
\u201cSPECULOOS and TRAPPIST, run by ULi\u00e8ge (Main Investigator Micha\u00ebl Gillon), played a central role in confirming the planetary nature of the signals,\u201d said the team. Their campaigns observed more than twenty separate transits between 2021 and 2024.<\/p>\n
To rule out impostors, the astronomers used high-resolution imaging with the Gemini North 8-m telescope and the SAI 2.5-m telescope in Russia. These observations resolved the binary itself, revealing two stars only 0.384 arcseconds apart, about 8 AU at TOI-2267\u2019s distance, and excluded any background eclipsing binaries. Statistical analysis using TRICERATOPS produced false-positive probabilities below 0.01 percent, confirming the planetary nature of TOI-2267 b and c.<\/p>\n
Inside a world of two suns<\/p>\n
TOI-2267\u2019s stars are among the smallest known to host planets. The primary is an M5 V dwarf with a radius of 0.21 solar radii and a surface temperature near 2 757 \u00b0C (4 995 \u00b0F). The secondary is an even cooler M6 V dwarf, 0.13 solar radii across and about 2 657 \u00b0C (4 814 \u00b0F).<\/p>\n
Despite their low luminosity, the stars\u2019 close orbits create intense gravitational interactions that would normally strip or destabilize forming planets. Yet the data show two stable, nearly circular planetary orbits with sizes of 1.0 \u00b1 0.1 Earth radii and 1.14 \u00b1 0.13 Earth radii. The candidate planet is similar in size at 0.95 \u00b1 0.12 Earth radii.<\/p>\n
The planets likely circle one of the stars in tight orbits inside 0.03 astronomical units, closer than Mercury\u2019s orbit by a factor of ten. Their periods suggest mild mutual gravitational resonance, a pattern common in compact systems like TRAPPIST-1.<\/p>\n
\u201cOur discovery breaks several records,\u201d said Francisco J. Pozuelos of IAA-CSIC. \u201cIt is the most compact and coldest pair of stars known to date with planets, and the first system where transiting planets have been detected around both components. This system is a real natural laboratory for understanding how rocky planets can emerge and survive in extreme dynamic conditions, where it was previously thought their stability would be compromised.\u201d<\/p>\n
Balancing on the edge of chaos<\/p>\n
After confirming the two planets, the team ran extensive numerical simulations to test the system\u2019s stability. They used machine-learning and N-body integration tools such as SPOCK and REBOUND to model millions of orbital possibilities over 10 million years of simulated time.<\/p>\n
Results showed that a configuration where all three planets orbit the same star becomes unstable in less than 100 000 years. The stable architecture likely places planets b and c around the primary star, while the candidate .02 orbits the secondary. This arrangement naturally explains the observed transit spacing and light-curve differences.<\/p>\n
The two confirmed planets are close to a 3:2 mean-motion resonance, meaning that for every three orbits of planet b, planet c completes two. Such resonances act like cosmic shock absorbers, preventing destructive encounters. Simulations suggest their orbits oscillate gently rather than collide, locked in a resonant pattern that maintains long-term stability.<\/p>\n
This finding shows that compact, low-mass binaries can host dynamically stable terrestrial worlds if the orbital geometry forms early and remains aligned with the stars\u2019 disks.<\/p>\n
A new view of planet formation<\/p>\n
For decades, astrophysicists assumed close binaries were barren. The combined gravity of two stars was thought to truncate or evaporate the disks where planets grow. Recent high-resolution observations from ALMA and the VLT have shown that many binaries retain circumstellar disks around each component.<\/p>\n
TOI-2267 strengthens this picture. The system\u2019s architecture suggests that two separate disks formed planets almost simultaneously. Dust and gas may have been sculpted by resonant interactions between the stars, funneling material inward to create rocky worlds on tight orbits.<\/p>\n
\u201cDiscovering three planets the size of the Earth in such a compact binary system is a unique opportunity,\u201d Z\u00fa\u00f1iga-Fern\u00e1ndez said. \u201cThis allows us to test the boundaries of planetary formation models in complex environments and to better understand the diversity of possible planetary architectures in our galaxy.\u201d<\/p>\n
The result challenges the notion that planet formation efficiency declines in multi-stellar environments. Instead, it points toward a broader diversity of planetary systems than previously recognized, even around the faintest stars in the Milky Way.<\/p>\n
What comes next for these twin suns<\/p>\n
The discovery of TOI-2267\u2019s planets opens new frontiers for observational astronomy. The system is bright enough for future monitoring by the James Webb Space Telescope (JWST) and upcoming extremely large telescopes (ELTs) on Earth. These facilities could measure planetary masses, densities, and possibly even atmospheric composition.<\/p>\n
The near-resonant configuration of planets b and c may produce subtle transit-timing variations. These shifts in the exact timing of each transit could reveal planetary masses. Early modelling predicts variations of only one to four minutes, a precision achievable with next-generation instruments such as HiPERCAM on the Gran Telescopio Canarias.<\/p>\n
If the third planet is confirmed, TOI-2267 would become the first binary system known to host transiting planets around both stars. Such a finding would offer a direct test of how two separate disks can each yield terrestrial planets within one shared gravitational field.<\/p>\n
Future campaigns aim to monitor the system across multiple wavelengths, seeking subtle color differences between the two stars\u2019 transits. Combined with radial-velocity data, these efforts may reveal whether each planet orbits the primary or secondary star and how the pair evolved together.<\/p>\n
References: <\/p>\n
1 Discovery of three Earth-sized planets in a compact binary system \u2013 University of Liege <\/a>\u2013 October 24, 2025<\/p>\n