2020ApJ...891..158M

Several models have been introduced in order to explain the radius distribution in exoplanet radii observed by Fulton et al. with one peak at ∼1.3R_⊕, the other at ∼2.4R_⊕, and the minimum at ∼1.75R_⊕. In this paper we focus on the hypothesis that the exoplanet size distribution is caused by stellar X-ray and ultraviolet (XUV)-induced atmospheric loss. We evolve 10⁶ synthetic exoplanets by exposing them to XUV irradiation from synthetic zero-age main-sequence stars. For each planet we set a different interior composition, which ranged from 100 wt% Fe (very dense), through to 100 wt% MgSiO₃ (average density), and to 100 wt% H₂O ice (low density), with varying hydrogen envelope sizes that varied from 0 wt% (a negligible envelope) to 100 wt% (a negligible core). Our simulations were able to replicate the bimodal distribution in exoplanet radii. We argue that in order to reproduce the distribution by Fulton et al. it is mandatory for there to be a paucity of exoplanets with masses above ∼8M_⊕. Furthermore, our best-fit result predicts an initial flat distribution in exoplanet occurrence for M_P≲8M_⊕ with a strong deficiency for planets with ≲3M_⊕. Our results are consistent with the ∼1.3R_⊕ radius peak mostly encompassing denuded exoplanets, while the ∼2.4R_⊕ radius peak is mainly comprised of exoplanets with large hydrogen envelopes.