Detection of 27 Candidate Circumbinary Planets Through Apsidal Precession of Eclipsing Binaries Observed by TESS

TL;DR

This study develops a transit-independent method to detect circumbinary planets by measuring apsidal precession in eclipsing binaries observed by TESS, using a large sample of Gaia DR3 EBs. By calculating the observed precession and subtracting general relativistic and tidal/rotational contributions, the authors infer the presence of a third body and constrain the mass–separation parameter space, identifying 27 planet-candidate companions (with some allowing sub-Jupiter masses) and highlighting degeneracies that require radial-velocity confirmation. The approach expands the CBP census to hotter stars and wider orbits, reducing biases inherent to transit searches and offering a complementary pathway to study CBP demographics and formation. Future RV measurements, extended eclipse timing baselines, and Gaia astrometric data will be essential to confirm these candidates and refine their masses and orbital architectures, enabling robust tests of circumbinary planet formation and evolution.

Abstract

Most circumbinary planets have been discovered by their transits, limiting our understanding of such systems to those with mutually coplanar architectures. This bias makes it difficult to infer the true circumbinary planet population, highlighting the need for alternative detection methods that do not rely on transits. In this work, we explore one such approach by leveraging apsidal precession as a dynamical signature of planetary companions. We analyse TESS photometry of a sample of 1,590 eclipsing binaries from the Gaia DR3 Catalogue of Eclipsing Binary Candidates to identify systems exhibiting detectable apsidal precession. We rule general relativistic, tidal, and rotational contributions as insufficient to account for the measured apsidal precession, demonstrating that an additional gravitational perturber is required. This enables us to constrain the possible masses and orbital separations of a companion that would cause the observed precession. We present a new set of 27 candidate circumbinary planets identified through this precession-based method as well as 6 candidate companions with a higher minimum mass. Their inferred properties remain degenerate, as the same dynamical signatures can arise from lower-mass planets at less than 1 AU or from more massive companions on wider, few-AU orbits, reflecting the current uncertainty in characterising these systems. Radial velocities can help break this degeneracy and provide direct confirmation.

Figures (12)

• Figure 1: The distribution of the semimajor axis of the known CBPs around main sequence binaries with reference to the critical stability radius of the binary (red "x"). For the multiple planet systems, Kepler-47 kepler47 and TOI-1338 toi1338bbebop1, only the innermost planet is shown with the size of the circle corresponding to the number of planets in the system. The CBP discovered via microlensing microlensing has been omitted from this diagram due to its imprecise orbital parameter determinations. In a vast majority of the systems, the innermost planet exists just outside of the critical radius.
• Figure 2: The phase-folded light curve of TIC 343127696, a $\sim4.67$-day eclipsing binary, showing seven sectors ($\sim 5$ years) of data. The primary eclipses are in the left two panels, while the secondary eclipses are in the right two panels. On top, the light curve has been folded on the period of the primary eclipses while on the bottom, it has been folded on the period of the secondary eclipses. When folded on the primary period, only the primary eclipses align cleanly; when folded on the secondary period, only the secondary eclipses align clearly. This systematic offset between the two eclipse types demonstrates that their inferred periods differ, which is direct evidence of apsidal precession in the binary.
• Figure 3: The observed minus calculated (O-C) plot based on a common period for TIC 286310830, a $\sim8.37$-day eclipsing binary. The primary eclipses in blue are occurring later over time while the secondary eclipses in red occur earlier than expected. This divergence in eclipse timing is a sign that the system is precessing.
• Figure 4: The parameter space for the companion to TIC 286310830, represented by the blue line. This precession could be due to a $\sim0.15 M_{Jupiter}$ companion at $\sim0.25$ AU. The region within the critical stability radius is shaded in grey.
• Figure 5: Same as Figure \ref{['fig:TIC286310830_OCwErr']}, but for the first dozen systems in Table \ref{['tab:tic_omega_dot']}.