Particle Clumping and Planetesimal Formation Depend Strongly on Metallicity

We present three-dimensional numerical simulations of particle clumping and planetesimal formation in protoplanetary disks with varying amounts of solid material. As centimeter-size pebbles settle to the mid-plane, turbulence develops through vertical shearing and streaming instabilities. We find that when the pebble-to-gas column density ratio is 0.01, corresponding roughly to solar metallicity, clumping is weak, so the pebble density rarely exceeds the gas density. Doubling the column density ratio leads to a dramatic increase in clumping, with characteristic particle densities more than ten times the gas density and maximum densities reaching several thousand times the gas density. This is consistent with unstratified simulations of the streaming instability that show strong clumping in particle dominated flows. The clumps readily contract gravitationally into interacting planetesimals of order 100 km in radius. Our results suggest that the correlation between host star metallicity and exoplanets may reflect the early stages of planet formation. We further speculate that initially low metallicity disks can be particle enriched during the gas dispersal phase, leading to a late burst of planetesimal formation.

💡 Research Summary

This paper investigates how the solid‑to‑gas mass ratio (often referred to as metallicity) controls particle clumping and the subsequent formation of planetesimals in protoplanetary disks. Using high‑resolution three‑dimensional hydrodynamic simulations (Athena++), the authors model a local shearing‑box region populated by centimeter‑sized “pebbles” that settle toward the mid‑plane. Two key column‑density ratios are explored: Σₚ/Σ_g = 0.01, which approximates the solar metallicity of the early Solar Nebula, and Σₚ/Σ_g = 0.02, a modest doubling of the solid content.

In the 0.01 case, the pebbles generate vertical shear that triggers Kelvin‑Helmholtz instability (KHI) and a modest streaming instability (SI). However, because the solid mass is insufficient to dominate the gas dynamics, the SI remains in a linear or weakly nonlinear regime. Particle densities rise only to about one to two times the gas density, never reaching the threshold required for self‑gravity to become important. Consequently, no gravitational collapse occurs and planetesimal formation is effectively stalled.

When the solid‑to‑gas ratio is increased to 0.02, the dynamics change dramatically. The higher pebble mass fraction allows the particle component to exert a non‑negligible back‑reaction on the gas, pushing the streaming instability into a strongly nonlinear state. The simulations show average particle densities exceeding ten times the gas density, with peak overdensities of 10³–10⁴. These dense clumps rapidly contract under their own gravity, producing bound objects with radii of order 50–150 km. The resulting planetesimals interact, merge, and form a modest population of roughly 100‑km‑scale bodies, reminiscent of the primordial planetesimal size distribution inferred for the asteroid and Kuiper belts.

A further set of experiments introduces a gradual exponential decay of the gas column density, mimicking the late‑stage dispersal of the nebular gas. Even disks that start with Σₚ/Σ_g = 0.01 can, during gas loss, see their solid‑to‑gas ratio climb to the 0.02 regime. At that point, the same burst of strong SI‑driven clumping occurs, leading to a “late‑time planetesimal formation event.” This mechanism provides a natural explanation for the observed correlation between host‑star metallicity and the occurrence rate of exoplanets: higher metallicity disks reach the critical solid fraction early, producing planetesimals promptly, while lower‑metallicity disks may wait until gas dispersal to trigger a delayed planetesimal surge.

The paper’s contributions are threefold. First, it quantifies the nonlinear sensitivity of particle clumping to a modest (factor‑two) change in solid content. Second, it demonstrates that the streaming instability’s outcome is bifurcated: weak clumping in gas‑dominated flows versus runaway collapse in particle‑dominated flows. Third, it highlights gas dispersal as a secondary pathway for low‑metallicity disks to achieve the critical solid fraction, suggesting that planet formation can be a two‑stage process.

Limitations include the neglect of magnetic fields, the assumption of a single pebble size, and the use of a local shearing‑box rather than a global disk model. Future work should incorporate magnetohydrodynamic effects, a realistic size distribution, and radial gradients to test the robustness of the metallicity threshold across the full disk. Observationally, high‑resolution ALMA measurements of dust‑to‑gas ratios in young disks, combined with stellar metallicity surveys, could directly test the predicted link between Σₚ/Σ_g and the onset of planetesimal formation.

In summary, the study provides a clear physical pathway connecting metallicity to the early stages of planet formation: a modest increase in solid content pushes the streaming instability into a regime of strong clumping, enabling rapid gravitational collapse into planetesimals. This mechanism not only explains the empirical metallicity‑planet correlation but also predicts a possible late‑burst of planetesimal creation during gas dispersal, offering fresh avenues for both theoretical and observational exploration.