BEBOP VIII. SOPHIE radial velocities reveal an eccentric, circumbinary brown dwarf

TL;DR

This study reports the detection of BEBOP-4 b, a circumbinary companion with $m_{ m b}\sin i_{ m b}=20.9\,M_{\rm Jup}$ on a $P_{ m b}=1823.5$ d orbit with $e_{ m b}=0.428$, around a close eclipsing binary ($P_{ m bin}=72$ d, $M_A=1.51\,M_\odot$, $M_B=0.46\,M_\odot$). Using 84 high-resolution SOPHIE spectra, the authors derive dynamical mass constraints and demonstrate that the true mass is $<26.3\,M_{\rm Jup}$ via apsidal precession, placing BEBOP-4 b in the planetary-mass brown-dwarf regime. The system, near its stability boundary, provides insight into circumbinary disc-driven formation and the dynamical architecture of high-mass circumbinary companions; Gaia DR4 is expected to yield the true mass and mutual inclination. Together with dynamical stability analyses, these results suggest that high-mass circumbinary companions, though rare, can form and persist in circumbinary discs, informing models of planet formation in perturbed environments.

Abstract

Circumbinary configurations offer a test of planet formation in an altered environment, where the inner binary has perturbed a protoplanetary disc. Comparisons of the physical and orbital parameters between the circumbinary planet population and the population of exoplanets orbiting single stars will reveal how these disc perturbations affect the assembly of planets. Circumbinary exoplanets detected thus far typically have masses $< 3 \,\rm M_{jup}$ raising the question of whether high-mass circumbinary planets are possible, and also whether population features such as the brown dwarf desert would appear in circumbinary configurations like for single star systems. Here, we report observations taken with the SOPHIE high-resolution spectrograph. These observations reveal an $m_{\rm b}\,\sin i_{\rm b} = 20.9 \,\rm M_{jup}$ outer companion, on an eccentric ($e = 0.43$), $1800\,\rm d$ orbit, which we call BEBOP-4 (AB) b. Using dynamical arguments we constrain the true mass $m_{\rm b}< 26.3 \,\rm M_{jup}$. The inner binary's two eclipsing stellar components have masses $M_{\rm A} = 1.51\,\rm M_\odot$, and $M_{\rm B} = 0.46\,\rm M_\odot$. Their orbital period is $72\,\rm d$, and their eccentricity is $0.27$. This system contains the longest period binary surveyed by the BEBOP project. BEBOP-4b is expected to be detectable using Gaia DR4 single epoch astrometric measurements. Despite a large period ratio of $\sim 25:1$, the substantial eccentricities of both orbits mean that the outer orbit is on the edge of orbital stability, and located in between two destabilising secular resonances. Should the outer companion survive, the BEBOP-4 system appears like a precursor to several post-common envelope binaries exhibiting eclipse timing variations where very massive circumbinary companions have been proposed.

Figures (11)

• Figure 1: HRCCS detection of BEBOP-4 B. Left: $K$-focused cross-correlation map. Center: Saltire model for the SOPHIE data, with parameters from MCMC. Right: residual CCF map after subtraction of the model. Data and model have been median-subtracted to allow for a common scale to be shared between the three panels. A secondary maximum is identified in the CCF map with a similar significance at $K_{\rm B}\sim62$ km s$^{-1}$, $v_{\rm sys}\sim-3$ km s$^{-1}$. This false signal, at the incorrect systemic velocity, is actually less significant and only gets highlighted due to the presence of the main signal. In the residual map, the 'false' maximum becomes almost indistinguishable from other noise structures.
• Figure 2: Phased radial velocity Keplerian and residuals for the detected circumbinary companion. The parameters used are those from the maximum-likelihood solution.
• Figure 3: Face-on view of the orbital configuration of the BEBOP-4 system showing the orbits of each star and the brown dwarf companion for 600 posterior draws. The observer is located on the right hand side. For illustrative purposes, the stability limit is calculated using holman_long-term_1999 and the binary orbital parameters from the maximum-likelihood solution is also shown. In magenta we highlight the parts of the companion orbit where it might transit one of the stars (calculated as the orbital phases where the companion crosses in front of the area swept out by the binary orbit, assuming coplanarity).
• Figure 4: Weighted distance between the measured apsidal precession rate (with the GR contribution removed) and the predicted apsidal precession for various mutual inclination values (i.e. inclinations of the circumbinary companion). Dashed lines show the $3\sigma$ and $5\sigma$ constraints.
• Figure 5: Detection limit plot for the BEBOP-4 system. In blue, the posterior density of the kima run forced to fit a signal, with the brown-dwarf companion removed. The posterior distribution for the companion is shown in green. Dashed lines show expected signals of solar-system mass planets. The 99% confidence detection limit as well as its uncertainty are shown in purple.