The New Generation Planetary Population Synthesis (NGPPS) VIII. Impact of host star metallicity on planet occurrence rates, orbital periods, eccentricities, and radius valley morphology

Abstract

Context. The dust-to-gas ratio in the protoplanetary disc, which is likely imprinted into the host star metallicity, is a property that plays a crucial role during planet formation. On the observational side, statistical studies based on large exoplanet datasets have determined various correlations between planetary characteristics and host star metallicity. Aims. We aim to constrain planet formation and evolution processes by statistically analysing planetary systems produced at different metallicities by a theoretical model, and we compare them with the correlations derived from observational samples. Methods. We used the Generation III Bern model of planet formation and evolution to generate synthetic planetary systems at different metallicities. This global model incorporates the accretion of planetesimals and gas, planetary migration, N-body interactions between embryos, giant impacts, and protoplanetary disc evolution, as well as the planets’ long-term contraction and atmospheric loss of gaseous envelopes. Using synthetic planets biased to observational completeness, we analysed the impact of stellar metallicity on planet occurrence rates, orbital periods, eccentricities, and the morphology of the radius valley. Results. Based on our nominal model, we find that (1) the occurrence rates of large giant planets and Neptune-sized planets are positively correlated with [Fe/H], while small sub-Earths exhibit an anti-correlation. In between, at radii of 1 to 3.5 R, the occurrence rate first increases and then decreases with increasing [Fe/H], with an inflection point at ∼0.1 dex. (2) Planets with orbital periods shorter than ten days are more likely to be found around stars with a higher metallicity, and this tendency weakens with increasing planet radius. (3) Both giant planets and small planets exhibit a positive correlation between the eccentricity and [Fe/H], which could be explained by the self-excitation and perturbation of outer giant planets. (4) The radius valley deepens and becomes more prominent with increasing [Fe/H], accompanied by a lower super-Earth-to-sub-Neptune ratio. Furthermore, the average radius of the planets above the valley (2.1–6 R) increases with [Fe/H]. Conclusions. Our nominal model successfully reproduces many observed correlations with stellar metallicity either quantitatively or qualitatively, and supports the description of physical processes and parameters included in the Bern model. Quantitatively, the dependence of orbital eccentricity and period on [Fe/H] predicted by the synthetic population, however, is significantly weaker than observed. This discrepancy likely arises because the model only accounts for planetary interactions for the first 100 Myr and neglects the effects of the stellar environment (e.g. clusters, binaries). This suggests that long-term dynamical interactions between planets, along with the impact of binaries and/or companions, can drive the system towards a dynamically hotter state.