Researchers find collisions can help young rocky planets keep atmospheres

Jun Korenaga of Yale University and Darius Modirrosta-Galian of Shanghai Jiao Tong University published a study in Nature Astronomy in 2026 concluding that collisions during rocky planet formation can help young worlds retain hydrogen-rich primordial atmospheres (the gas envelopes planets acquire during formation) instead of losing them.

The paper Retention of primordial atmospheres through giant impacts appeared in Nature Astronomy in 2026 with DOI 10.1038/s41550-026-03006-1. The team tracked how rocky planets evolve after their biggest impact following the loss of the gas disk, handling gravitational binding energy and mantle thermal energy at the same time and in a self-consistent way. Korenaga said the result rejects the long-standing idea that giant impacts nearly always strip atmospheres completely.

Following a collision, added mass strengthens the planet’s gravitational binding energy (the energy holding the planet together) and its hold on the atmosphere. This effect operates alongside impact heating. The impactor’s metallic core remains superheated at depth and passes little energy to the rocky middle layer, so only some of the impact’s heat goes into the mantle. Melting the mantle also consumes energy as latent heat, taking up part of the impact’s thermal budget.

Since every bit of added mass increases binding energy while only part of the impact’s heat warms the mantle, a collision shifts the planet’s energy balance toward holding on to its atmosphere. The scientists handled gravitational binding energy and mantle thermal energy together in their models. That imbalance decides if the planet keeps its gas following a collision. Mantle melting is one of several ways impact energy is divided rather than all warming the mantle.

Planets at least the size of Earth are expected to experience at least one giant collision as they finish assembling. Several of these impacts take place after the gas disk disappears, so they can shape primordial atmospheres that survive past that point. The models show that large sub-Neptunes (planets larger than Earth with primordial atmospheres) holding thick atmospheres arise more frequently through collisions, and in population terms the research predicts that sub-Neptunes generally have heavier rocky interiors compared with super-Earths.

Super-Earths that never experience a giant impact generally lose their primordial atmospheres under long-lasting irradiation from their host stars. For low-mass worlds, stellar energy slowly erodes gas that is only weakly held. That irradiation gradually strips away weakly held gas from small planets. This contrasts with the outcome for larger atmosphere-bearing planets that emerge through collisional growth.

Korenaga said giant impacts drive the growth of rocky planets, and worlds at least as large as Earth are expected to form through collisions between smaller planetary bodies. He explained that such collisions add mass and release enormous heat that can drive atmospheric escape, and that the fate of the atmosphere depends on how those two changes work together.

The lead author called the abundance of super-Earths and sub-Neptunes a fundamental puzzle in exoplanetary science. Super-Earths are slightly larger than Earth without primordial atmospheres, while sub-Neptunes are larger and with primordial atmospheres. The lead author said the group set out to explain why both super-Earths and sub-Neptunes occur so frequently. Modirrosta-Galian worked as a postdoctoral researcher in Korenaga’s lab at Yale.

The two scientists wrote several papers in recent years laying the groundwork for how atmospheric escape and planetary interiors should coevolve. The Nature Astronomy study brings those earlier studies together to model the aftermath of a giant impact.

Impacts heat surfaces and encourage gas loss, while added mass strengthens the gravitational well that holds the atmosphere. Before the study, it remained unclear which of these competing effects would win. Prior to this work, researchers believed a late giant impact would strip a rocky planet’s original hydrogen-rich atmosphere. Earth-sized and larger planets are expected to face giant collisions after the disk dissipates, creating competing effects whose relative strength had not been resolved.

Collisions that take place after the gas disk vanishes can heat surfaces enough to drive away primordial hydrogen-rich atmospheres. Collisions increase gravitational binding energy more than they heat the mantle, so planets generally keep their atmospheres. The modelled split in energy means a collision pushes the planet toward retaining its atmosphere rather than losing it.

The results rest on simulation assumptions, particularly how impact energy is divided among the mantle, the impactor’s core, and the atmosphere. The paper’s figures look at post-impact surface temperatures, the chance of total atmospheric loss, and how different giant-impact assumptions change predicted radius distributions.

The research addresses the observed split in which nearby planets cluster into two size ranges with few objects between. Learning how planets gain or lose primordial atmospheres through impacts, instead of only through stellar heating, adds another mechanism that could help explain why the exoplanet radius gap exists. Should the models prove correct, collisions that appeared likely to strip young worlds may have allowed them to keep their first atmospheres. That conclusion follows from the energy balance tilting toward retention as impact energy is divided in the manner described.