Compositional Links Between Stars and Their Planets

A composition link between planets and stars is predicted by formation theories given that they are born from the same disk. It is unclear however, if the planets that are formed preserve the primordial composition or exhibit reprocessing during formation. Rocky super-Earths offer a unique opportunity to test these theories because they are spared the compositional degeneracy of other planets. Different studies so far have used rocky populations that suffer from a lack of consistency biasing their results. The main culprit is the inconsistent analysis of stellar mass, radius and composition, thus impacting the inferred planetary mass and radius. To overcome this shortcoming, we homogeneously re-analyze a suitable population of stars and revise mass-radius data for their compact exoplanets (N = 21), which changes their composition. To compare the stellar host and planetary refractory compositions, we use a sophisticated interior structure model using consistent chemical priors. We also explore different statistical metrics to investigate the robustness of any correlations between the stellar refractory component and their stars. We find that using the whole sample yields no correlation. Only if we group different categories of planets (iron-rich, Earth-like and volatile planets), there appears to be a weak correlation, albeit with high uncertainty, among the super-Earths. Very interestingly and at odds with a primordial origin, there seems to be three stars with low to normal equivalent iron-mass fraction (FeMF) that host three high iron-rich planets (K2-38b, K2-229b and Kepler-107c, two of which are beyond one σ). We propose these as high value observational targets as they test the parameter space where formation theories seem to fail.