The origins of nearly coplanar, non-resonant systems of close-in super-Earths

Abstract

Some systems of close-in 'super-Earths' contain five or more planets on non-resonant but compact and nearly coplanar orbits. The Kepler-11 system is an iconic representative of this class of system. It is challenging to explain their origins given that planet-disc interactions are thought to be essential to maintain such a high degree of coplanarity, yet these same interactions invariably cause planets to migrate into chains of mean motion resonances. Here, we mine a large data set of dynamical simulations of super-Earth formation by migration. These simulations match the observed period ratio distribution as long as the vast majority of planet pairs in resonance become dynamically unstable. When instabilities take place resonances are broken during a late phase of giant impacts, and typical surviving systems have planet pairs with significant mutual orbital inclinations. However, a subset of our unstable simulations matches the Kepler-11 system in terms of coplanarity, compactness, planet-multiplicity, and non-resonant state. This subset has dynamical instability phases typically much shorter than ordinary systems. Unstable systems may keep a high degree of coplanarity post-instability if planets collide at very low orbital inclinations (<1°) or if collisions promote efficient damping of orbital inclinations. If planetary scattering during the instability takes place at low orbital inclinations (i < 1°), orbital inclinations are barely increased by encounters before planets collide. When planetary scattering pumps orbital inclinations to higher values (>1°) planets tend to collide at higher mutual orbital inclinations, but depending on the geometry of collisions mergers' orbital inclinations may be efficiently damped. Each of these formation pathways can produce analogues to the Kepler-11 system.