2022MNRAS.512..602M

All of the known circumbinary planets are large (R_p >= 3 R_⊕). Whilst observational biases may account for this dearth of small planets, in this paper we propose a theoretical explanation. Most of the known planets are near the stability limit, interspersed between potentially unstable 5:1, 6:1, and 7:1 mean motion resonances with the binary. It is believed that these planets did not form in situ, but rather migrated from farther out in the disc, and hence passed through these resonances. Planets are expected to migrate at a speed proportional to their mass, and a slower rate makes resonant capture and subsequent ejection more likely. Therefore, whilst large planets may be able to successfully 'run the gauntlet', small planets may be imperilled. This hypothesis is tested using N-body integrations of migration in a truncated and turbulent disc. We discover that surprisingly none of the known planets exist interior to a fully unstable resonance. We demonstrate that resonant ejection of migrating planets may occur in nature, and that it does indeed disproportionately affect small planets, but it requires a highly turbulent disc and its efficiency is likely too low to fully explain a dearth of R_p < 3 R_⊕ planets.