Circumbinary planet
A circumbinary planet is a planet that orbits the common center of mass of a binary star system, encircling both stars rather than one individually.[1] These planets, also known as P-type planets in binary systems, must maintain stable orbits beyond a critical minimum separation from the binary to avoid gravitational disruptions.[2] The first unambiguous detection of a circumbinary planet came in 2011 with Kepler-16b, a Saturn-mass world orbiting a binary pair of Sun-like and red dwarf stars approximately 200 light-years away, identified through the transit method by NASA's Kepler space telescope.[3] Since then, additional discoveries have been made using various techniques, including radial velocity (e.g., BEBOP-3b in 2025), microlensing, eclipse timing variations, and direct imaging.[4] As of November 2025, approximately 35 confirmed circumbinary planet systems are cataloged, representing a small fraction (less than 1%) of the more than 6,000 confirmed exoplanets, with most detections favoring gas giants in close orbits around binaries with separations of a few astronomical units.[5] Notable examples include Kepler-34b and Kepler-35b, Saturn-sized planets orbiting near-equal-mass binary stars, and TOI-1338b, the first circumbinary planet found by the Transiting Exoplanet Survey Satellite (TESS) in 2020, which experiences multiple eclipses due to its inclined orbit.[2] These systems often feature binaries with periods ranging from days to years, and the planets' orbits are typically coplanar with the binary's, though some show misalignments, including polar configurations near 90 degrees in recent candidates.[6] Circumbinary planets provide insights into planet formation in complex gravitational environments, where protoplanetary disks around binaries may truncate inward, favoring outward formation followed by migration to stable zones near the binary's dynamical edge. Recent evidence for polar circumbinary exoplanets further highlights diverse orbital dynamics in these systems.[7] Ongoing surveys like BEBOP and PLATO aim to detect smaller, Earth-like circumbinary worlds and probe their potential habitability, particularly in systems where the habitable zone overlaps the stable orbital region.[2]Definition and Fundamentals
Definition and Terminology
A circumbinary planet (CBP) is a planet whose orbit encloses both stars of a binary system, revolving around their common center of mass. This configuration distinguishes CBPs from planets in S-type orbits, where the planet circles only one of the binary stars, or from circumstellar planets around isolated single stars. In binary star systems, planetary orbits are classified as P-type or S-type following the nomenclature introduced by Dvorak (1986), where P-type denotes circumbinary configurations and S-type refers to orbits around a single star within the system.[8] Additional terminology includes the binary separation a_\mathrm{bin}, defined as the semi-major axis of the binary stars' relative orbit, which sets the scale for dynamical interactions. The mutual inclination describes the angle between the orbital plane of the planet and that of the binary, influencing long-term stability.[9] Orbital stability for CBPs requires the planet's semi-major axis to exceed a critical radius, typically approximately 2–3 times a_\mathrm{bin} for coplanar, circular binaries, beyond which chaotic perturbations from the stars diminish.[10] This limit arises from numerical simulations of the restricted three-body problem and varies with binary eccentricity, expanding to about 4 times a_\mathrm{bin} for higher values.[10] The dynamics of CBPs are governed by the restricted three-body problem, where the planet's negligible mass allows treatment as a test particle under the gravitational fields of the two stars, whose mutual orbit is prescribed.[8] In this framework, the planet's orbit encloses both stars, leading to periodic perturbations that can excite eccentricity or inclination but permit stable configurations outside the critical radius.[10]Comparison to Other Exoplanet Types
Circumbinary planets, or P-type planets, differ fundamentally from S-type planets in binary star systems, where the planet orbits only one of the two stars. S-type configurations are far more prevalent, with over 800 confirmed systems cataloged, compared to approximately 35 P-type systems. This disparity arises because S-type orbits are feasible in wider binaries where the companion star's influence is minimal, allowing stable planetary systems around individual stars, whereas P-type orbits require the planet to encircle both stars, which is restricted to closer binaries with specific stability zones. In binary systems overall, S-type planets dominate detections, reflecting both formation preferences and observational accessibility in systems where the binary separation exceeds several astronomical units.[11][5] When compared to planets in single-star systems, circumbinary planets exhibit greater dynamical challenges due to the perturbing gravitational influence of the binary pair, leading to reduced long-term stability and lower overall occurrence rates. Close binary companions suppress planet formation around individual stars (S-type) and further complicate circumbinary (P-type) disk evolution, resulting in fewer stable architectures than the more quiescent environments of single stars. Surveys like NASA's Kepler mission underscore this rarity, confirming only about 12 circumbinary planets amid thousands of exoplanet detections, a fraction less than 1% even accounting for binary star prevalence in the galaxy. This scarcity stems primarily from enhanced instabilities, such as orbital disruptions and ejections, rather than solely detection limitations.[12][13] A distinctive feature of circumbinary planets is their potential for higher orbital eccentricities induced by resonant perturbations from the central binary, contrasting with the generally lower eccentricities in single-star systems where fewer forcing mechanisms exist. These perturbations can drive forced eccentricity components that scale with binary parameters, enabling more varied orbital behaviors. Additionally, the protoplanetary disks from which circumbinary planets form—known as circumbinary disks—differ markedly from single-star protoplanetary disks; the binary torque creates an inner cavity and asymmetric structures, altering dust and gas dynamics and potentially hindering efficient planet assembly compared to the more uniform, untruncated disks around solitary stars.[14][15] Detection of circumbinary planets via transit photometry faces unique biases, including their larger orbital separations, which generally reduce transit probabilities relative to closer-in single-star planets, though this is partially offset by the edge-on geometry of observed eclipsing binaries and orbital precession that enhances transit visibility over time.[16]History of Discovery
Theoretical Predictions
Theoretical interest in circumbinary planets emerged in the 1970s and 1980s through studies of circumbinary dust disks observed around binary stars, which suggested environments conducive to planet formation. For instance, models of material distribution in systems like V471 Tauri indicated the presence of stable circumbinary structures capable of supporting planetary accretion. Similarly, analyses of epsilon Aurigae and Beta Lyrae proposed circumbinary disks as sites for potential planetary systems. These early works laid the groundwork for anticipating planets orbiting binary centers of mass, though direct stability assessments were limited. N-body simulations in the late 1980s began to quantify orbital stability for circumbinary (P-type) configurations. Dvorak (1986, 1989) conducted pioneering numerical integrations, demonstrating that test particles could maintain stable orbits beyond approximately 2.3 times the binary semi-major axis a_\mathrm{bin} for circular, equal-mass binaries, with the limit increasing to about 4 a_\mathrm{bin} for eccentric binaries. These results highlighted the dynamical challenges near the binary, where resonant perturbations could lead to ejections or collisions. Building on this, Holman and Wiegert (1999) performed extensive simulations across a range of binary eccentricities ($0 \leq e \leq 0.7) and mass ratios ($0.1 \leq q \leq 0.5, where q = m_2 / (m_1 + m_2) with m_1 \geq m_2), deriving an empirical criterion for the minimum stable planetary semi-major axis a_p based on Hill-like stability in the three-body problem. The formula is \frac{a_p}{a_\mathrm{bin}} > 1.60 + 5.10 e - 2.22 e^2 + 4.12 q - 4.27 e q - 5.09 q^2 + 4.61 e^2 q^2, which approximates to roughly 2–3 a_\mathrm{bin} for low-eccentricity, near-equal-mass binaries, ensuring long-term survival against chaotic perturbations.[17] This Hill stability analogy treats the binary as a central perturber, preventing the planet's orbit from overlapping unstable regions dominated by mean-motion resonances. Disk evolution models from the 1990s and early 2000s predicted circumbinary planet occurrence rates of approximately 1–10% around close binaries (a_\mathrm{bin} \lesssim 1 AU), based on simulations of protoplanetary disk truncation and migration in binary environments. These models accounted for disk mass transfer to the binary and cavity formation, suggesting efficient in situ formation of gas giants near the stability limit despite dynamical heating. Pre-Kepler era hints of circumbinary planets came from pulsar timing observations in 1993, where variations in the pulse arrival times of the neutron star-white dwarf binary PSR B1620-26 were interpreted as gravitational perturbations from a ~2.5 Jupiter-mass planet at ~23 AU, implying stable orbits around compact binaries.Key Observational Milestones
The earliest candidate for a circumbinary planet was identified in 1993 around the binary pulsar system PSR B1620-26 in the globular cluster M4, based on timing variations in pulsar signals suggesting a low-mass companion orbiting the pulsar-white dwarf binary.[18] This candidate, PSR B1620-26 b, was confirmed in 2003 through Hubble Space Telescope observations that resolved the white dwarf companion and refined the system's dynamics, marking the first verified circumbinary exoplanet despite its unusual pulsar environment.[19] The Kepler mission revolutionized circumbinary planet detection starting in 2011, when it announced Kepler-16b, the first unambiguous circumbinary planet around main-sequence stars, confirmed via transit photometry revealing eclipses of both stars by the Saturn-mass planet. This breakthrough was followed by additional discoveries during Kepler's primary mission (2011-2015), including Kepler-34b and Kepler-35b in 2012, which demonstrated that circumbinary planets could orbit a diverse range of binary configurations, from Sun-like pairs to cooler red dwarfs.[20] By the end of the Kepler era, the mission had confirmed 10 circumbinary planets, establishing their prevalence and providing key insights into their formation around close binaries.[21] A notable late addition came from reanalysis of Kepler data, with Kepler-1647b validated in 2019 as the longest-period transiting circumbinary planet known at the time, orbiting at over twice Earth's distance from the Sun and highlighting the mission's archival value.[22] Following Kepler's conclusion, the Transiting Exoplanet Survey Satellite (TESS) extended transit-based searches, yielding its first circumbinary planet, TOI-1338 b, in 2020—a Neptune-sized world orbiting a binary with stars of unequal brightness, detected through irregular transit timings.[23] This discovery underscored TESS's capability to probe wider fields for such systems. In 2023, the Binaries Etcetera Beyond the Origin of Planets (BEBOP) survey achieved a milestone by detecting BEBOP-1c (also known as TOI-1338 c) via radial velocity measurements with HARPS and ESPRESSO spectrographs, the first circumbinary planet confirmed without relying on transits and revealing a gas giant in the same system as TOI-1338 b.[24] As of 2025, ongoing surveys continue to diversify detection methods and expand the sample. The BEBOP project reported BEBOP-3 b, a giant circumbinary planet on an eccentric orbit detected via radial velocities from the SOPHIE spectrograph, demonstrating the technique's sensitivity to longer-period worlds around evolved binaries.[25] Concurrently, evidence emerged for a polar circumbinary exoplanet candidate around the brown dwarf binary 2M1510 (AB), with an orbit inclined at approximately 90 degrees to the binary plane, inferred from radial velocity variations and challenging models of stability in misaligned systems.[6] These findings, alongside TESS's ongoing contributions, illustrate the growing role of radial velocity methods in complementing photometry and broadening the observational landscape beyond Kepler's legacy.[26]Detection Methods
Transit Photometry
Transit photometry is the primary method for detecting circumbinary planets (CBPs), relying on the periodic dimming of the binary stars' combined light as the planet passes in front of one or both stars during its orbit around the binary barycenter. For a transit to be observable, the planetary orbit must be nearly coplanar with the binary's orbit, allowing the planet to cross the line of sight to both stars; this configuration produces characteristic light curve dips that vary in depth, duration, and timing due to the stars' mutual orbital motion. These circumbinary transit timing variations (CBTTV) arise from the planet's interaction with the moving binary, causing transits to occur earlier or later than expected from a constant period, which serves as a key diagnostic for confirming CBP candidates.[27] The method's success stems from the high sensitivity of space-based observatories like Kepler and TESS, which have identified nearly all confirmed CBPs through repeated transits amid the binary's eclipses. Transit probability for CBPs is influenced by their typically larger semi-major axes compared to single-star planets of similar orbital periods, but dynamical stability often enforces near-edge-on inclinations, enhancing detectability; an approximate geometric probability is given by P_{\trans} \approx \frac{R_{\star}}{a_p} \left(1 + \frac{a_{\bin}}{a_p}\right), where R_{\star} is the effective stellar radius, a_p the planetary semi-major axis, and a_{\bin} the binary semi-major axis, reflecting the extended "target" size of the binary system. This approach has yielded robust detections, such as the first confirmed CBP, Kepler-16b, announced in 2011 from Kepler photometry showing transits across both stars, later corroborated by radial velocity measurements.[16] Detecting CBPs via transits faces significant challenges from the binary eclipses, which produce deeper, more frequent flux drops that can mimic or obscure planetary signals in the light curve, necessitating precise modeling of the binary orbit to isolate planetary transits. The non-periodic nature of CBP transits—due to CBTTV and transit duration variations (TDV) from the changing sky-projected separation of the stars—requires observations spanning multiple planetary orbits (often years) to establish the superperiod and rule out false positives like stellar spots or instrumental artifacts. Automated detection algorithms must therefore incorporate dynamic transit predictions, accounting for the binary's reflex motion to achieve reliable identifications.[28][29]Radial Velocity Spectroscopy
Radial velocity (RV) spectroscopy detects circumbinary planets by observing the periodic Doppler shift in the spectral lines of the binary stars, caused by the planet's gravitational tug on the system's center of mass. In these systems, the planet orbits the binary barycenter, inducing a reflex motion that manifests as an RV variation superimposed on the stars' velocities. This method requires high-precision measurements to isolate the planetary signal from the dominant binary orbital motion.[30] The semi-amplitude K_p of the planetary RV signal, assuming a circular orbit, is given by K_p = \left( \frac{2\pi G}{P} \right)^{1/3} \frac{m_p \sin i}{(m_1 + m_2)^{2/3}}, where P is the planet's orbital period, m_p is the planet's mass, i is the orbital inclination relative to the line of sight, m_1 and m_2 are the stellar masses, and G is the gravitational constant. For non-circular orbits, the formula includes an additional factor of $1 / \sqrt{1 - e^2}, where e is the eccentricity, and the planet mass term in the denominator is often negligible. However, extraction of this signal is complicated by the binary's own RV curve, which has a much larger amplitude (typically tens to hundreds of m/s) compared to the planetary signal (often <10 m/s for Jupiter-mass planets), necessitating advanced modeling techniques like multi-Keplerian fits or Gaussian process regression to subtract the binary orbit.[30][31] Historically, RV detection of circumbinary planets lagged behind transit methods due to these modeling challenges and the need for stable, long-term spectrographic monitoring. The first such detection occurred in 2023 with BEBOP-1c, a super-Earth-mass planet (approximately 65 Earth masses) orbiting the binary TOI-1338 at a period of 215 days, identified using data from the HARPS and ESPRESSO instruments as part of the BEBOP survey. This marked the inaugural RV discovery of a circumbinary planet, following transit-based findings. In 2025, the BEBOP survey announced BEBOP-3b, a Jupiter-mass planet (0.56 Jupiter masses) on an eccentric orbit (e ≈ 0.25) with a 550-day period, representing the first circumbinary planet detected solely via RV without prior transit observations, using SOPHIE spectrograph data.[32][26] A key advantage of RV spectroscopy for circumbinary planets is its ability to measure minimum masses (m_p \sin i) for non-transiting systems, providing dynamical constraints that complement transit-derived radii and enabling full mass-radius characterization when combined with other methods. It is particularly suited to detecting massive gas giants, which produce stronger signals and are less likely to transit due to their wider orbits, thus probing a distinct population missed by transit surveys biased toward close-in planets.[30][26] Despite these strengths, limitations persist: the binary's RV noise demands extended baselines (often years) of precise observations to phase the planetary signal accurately, and double-lined binaries further complicate line disentangling. As of November 2025, four circumbinary planets, including Kepler-16b (RV confirmation of its transit detection), TOI-1338b (mass refinement via RV), BEBOP-1c (discovered via RV), and BEBOP-3b (discovered via RV), have been confirmed or detected via RV, underscoring the method's nascent application to these systems.[26]Known Circumbinary Systems
Confirmed Planets
As of November 2025, 41 circumbinary planets have been confirmed in 35 systems through various detection methods, primarily transit photometry from space-based telescopes like Kepler and TESS, as well as radial velocity spectroscopy, pulsar timing, eclipse timing variations, and microlensing. These discoveries span a range of stellar types, including main-sequence binaries, evolved systems, and cataclysmic variables, highlighting the diversity of environments where planets can form and survive around binary stars. Many additional systems have been identified via eclipse timing variations in post-common-envelope binaries (e.g., NN Ser, DP Leo) and microlensing events (e.g., OGLE-2007-BLG-349L, OGLE-2023-BLG-0836). The first detection, PSR B1620-26 b, was identified in 1993 via pulsar timing observations of the millisecond pulsar PSR B1620-26 in the globular cluster M4. Subsequent breakthroughs came with the Kepler mission, which revealed multiple systems through precise light curve monitoring that captured planetary transits across both stars. More recent additions include radial velocity detections from the BEBOP survey and eclipse timing evidence for inclined orbits.[33][5] The confirmed planets are predominantly gas giants or close to super-Earth sizes, with masses typically between 0.2 and several Jupiter masses where measured, and radii up to about 1.5 times Jupiter's for transiting examples. Their orbits have semi-major axes generally between 0.7 and 2.5 AU, while the host binaries exhibit periods from 7 to 2000 days, allowing stable planetary configurations outside the binary's dynamical influence zone. Recent discoveries include BEBOP-3 b, a Jupiter-mass planet on an eccentric orbit detected via radial velocities with the SOPHIE spectrograph, and V808 Aur b, a polar planet inclined approximately 90° relative to the binary plane, providing evidence for misaligned architectures around cataclysmic variables. Orbital stability in these systems is supported by dynamical simulations showing long-term survivability for planets beyond about 2-3 times the binary separation.[25][22] Key parameters for selected confirmed circumbinary planets are summarized in the table below, drawn from peer-reviewed analyses and updated archival data. Note that masses are often minimum values (m sin i) for radial velocity detections, radii are for transiting planets, and binary separations refer to the semi-major axis of the binary orbit.| System Name | Discovery Year/Method | Binary Separation (AU) | Planet Mass (M_J) / Radius (R_J) | Orbital Period (days) | Eccentricity |
|---|---|---|---|---|---|
| PSR B1620-26 b | 1993 / Pulsar timing | 0.008 | ~2.5 / - | ~36,525 | ~0.00 |
| Kepler-16 b | 2011 / Transit | 0.22 | ~0.76 (min) / 0.76 | 229 | 0.007 |
| Kepler-34 b | 2012 / Transit | 0.23 | ~0.22 (min) / 0.79 | 289 | 0.182 |
| Kepler-35 b | 2012 / Transit | 0.18 | ~0.13 (min) / 0.72 | 131 | 0.042 |
| Kepler-38 b | 2012 / Transit | 0.15 | ~0.02 (min) / 0.27 | 88 | 0.032 |
| Kepler-47 b | 2012 / Transit | 0.08 | ~0.02 (min) / 0.28 | 49 | 0.014 |
| Kepler-47 c | 2012 / Transit | 0.08 | ~0.12 (min) / 0.34 | 307 | 0.014 |
| Kepler-64 b (PH1) | 2013 / Transit | 0.18 | ~0.40 (min) / 0.90 | 138 | 0.054 |
| Kepler-413 b | 2013 / Transit | 0.10 | ~0.22 (min) / 0.36 | 66 | 0.118 |
| Kepler-453 b | 2015 / Transit | 0.18 | ~0.10 (min) / 1.01 | 240 | 0.036 |
| Kepler-1647 b | 2019 / Transit | 0.13 | ~0.20 (min) / 1.06 | 1,048 | 0.058 |
| TOI-1338 b | 2020 / Transit | 0.13 | ~0.38 (min) / 1.32 | 95 | 0.088 |
| BEBOP-1 c | 2023 / Radial velocity | 0.50 | ~4.0 / - | 1,000 | 0.10 |
| BEBOP-3 b | 2025 / Radial velocity | 0.12 | ~0.56 (min) / - | 550 | 0.25 |
| V808 Aur b | 2024 / Eclipse Timing Variations | ~0.01 | ~6.8 (min) / - | 4,015 | 0.805 |