Hot Jupiter and ultra-cold Saturn formation in dense star clusters

TL;DR

This study uses high-precision simulations to show that stellar flybys in dense clusters can trigger hot Jupiter formation via high eccentricity migration, also predicting the formation of ultra-cold Saturns.

Contribution

It demonstrates that stellar flybys can significantly enhance hot Jupiter formation rates and introduces the prediction of ultra-cold Saturns in dense star clusters, expanding understanding of planetary system evolution.

Findings

Flybys activate high eccentricity migration mechanisms.

Many hot Jupiters are accompanied by ultra-cold Saturns.

Predicted hot Jupiter formation rate is lower than observed in M67.

Abstract

The discovery of high incidence of hot Jupiters in dense clusters challenges the field-based hot Jupiter formation theory. In dense clusters, interactions between planetary systems and flyby stars are relatively common. This has a significant impact on planetary systems, dominating hot Jupiter formation. In this paper, we perform high precision, few-body simulations of stellar flybys and subsequent planet migration in clusters. A large parameter space exploration demonstrates that close flybys that change the architecture of the planetary system can activate high eccentricity migration mechanisms: Lidov-Kozai and planet-planet scattering, leading to high hot Jupiter formation rate in dense clusters. Our simulations predict that many of the hot Jupiters are accompanied by "ultra-cold Saturns", expelled to apastra of thousands of AU. This increase is particularly remarkable for planetary systems originally hosting two giant planets with semi-major axis ratios $\sim$ 4 and the flyby star approaching nearly perpendicular to the planetary orbital plane. The estimated lower limit to the hot Jupiter formation rate of a virialized cluster is $\sim 1.6\times10^{-4}({\sigma}/{\rm 1kms^{-1}})^5({a_{\rm p}}/{\rm 20 AU})({M_{\rm c}}/{\rm 1000M_\odot})^{-2}$Gyr$^{-1}$ per star, where $\sigma$ is the cluster velocity dispersion, $a_{\rm p}$ is the size of the planetary system and $M_{\rm c}$ is the mass of the cluster. Our simulations yield a hot Jupiter abundance which is $\sim$ 50 times smaller than that observed in the old open cluster M67. We expect that interactions involving binary stars, as well as a third or more giant planets, will close the discrepancy.