Most High-density Exoplanets Are Unlikely to Be Remnant Giant Planet's Cores

Some exoplanets have much higher densities than expected from stellar abundances of planet-forming elements. There are two theories—metal-rich formation hypothesis and naked core hypothesis—that explain how formation and evolution can alter the compositions and structures of rocky planets to diverge from their primordial building blocks. Here we revisit the naked core hypothesis, which states that high-density planets are remnant cores of giant planets that remain in a fossil-compressed state, even after envelope loss. Using a planetary interior model and assuming energy-limited atmospheric escape, we show that a large fraction, if not all, of the iron–silicate core of a giant planet is molten during the planet's early evolution. Upon envelope loss, the molten part of the planets can rapidly rebound owing to low viscosity, resulting in a decrease in radius by at most 0.06%, if they had hydrogen/helium envelopes, or by at most 7%, if they had H2O envelopes, compared to self-compressed counterparts with the same core mass fraction. Based on our findings, we reject the hypothesis that all high-density exoplanets are naked cores with Kolmogorov–Smirnov p-value ≪0.05 for both envelope compositions. We find that some high-density exoplanets can still possibly be naked cores, but the probabilities are lower than ∼1/2 and ∼1/3 for the ice giant and gas giant scenario, respectively, in 95% of the cases. We conclude that most high-density exoplanets are unlikely to be remnant giant planet cores.