An international team has published simulations indicating that the solid precursors of rocky planets could have assembled in the debris of the universe's first stellar explosions, far earlier than standard models of cosmic chemical enrichment allow. The paper appeared in The Astrophysical Journal Letters and is available through its journal record, with the University of Portsmouth announcing the result in August.

The claim is specific. In one cosmological simulation, a dense cloud core enriched by a single pair-instability supernova collapsed into a protoplanetary disk around a star roughly 70 percent the mass of the Sun, and several Earth masses of planetesimals formed between about 0.46 and 1.1 astronomical units from that star. For scale, Earth orbits at one astronomical unit.

This is a modeling result, not an observation. No instrument has imaged a protoplanetary disk from that era, and none currently can. What the work changes is the plausibility of an idea that had been widely dismissed: that the raw material for terrestrial worlds existed billions of years before Earth formed.

Inside the Simulation Chain

The physics turns on a specific kind of death. The first stars, called Population III, formed from gas containing essentially no elements heavier than helium. The most massive of them ended in pair-instability supernovae, explosions violent enough to unbind the entire star and eject more than 100 solar masses of heavy elements in a single event.

Conventional large box cosmological simulations average metal enrichment across enormous volumes, and in that picture the universe does not reach the chemical threshold for widespread planet formation until roughly half its present age. The team argues those simulations cannot resolve metal mixing, cooling and clump formation at the small scales where planetary systems actually emerge.

Working at those smaller scales, the researchers found supernova debris producing cloud cores enriched to metallicities above about a tenth of the Sun's. One such core, with a Jeans mass of only one to two solar masses, collapsed into a disk where dust could concentrate and stick.

Daniel Whalen of the University of Portsmouth's Institute of Cosmology and Gravitation said the paper shows that the precursors of terrestrial planets can form around low-mass, long-lived stars in the debris of the first cosmic explosions, and noted that against a universe about 13.8 billion years old this is remarkably early. Whalen has said his doctoral student Christopher Jessop ran the first stage of the simulation chain. The lead author is Eduard Vorobyov of the University of Innsbruck, and the eight co-authors are based at institutions in Austria, the United Kingdom, the United Arab Emirates, Russia, Japan and the United States.

A Figure That Shifted Between Preprint and Publication

Readers comparing coverage will encounter two different numbers, and the discrepancy is worth stating plainly. The preprint version posted to arXiv, last revised in July 2025, places planetesimal formation at 200 million years after the Big Bang. The university announcement accompanying the publication, and Whalen's quoted description of the work, give 100 million years.

Earlier work by the same group found water forming in supernova-enriched clouds in roughly the same era, which suggests the team is describing a window rather than a single instant. Either figure sits within the first two percent of cosmic history, so the qualitative claim holds regardless. The precise number, however, should be read as approximate.

The disk's water content was the second headline result. The simulation produced water mass fractions only a few times below those in the present-day solar system, and planetesimal formation occurred inside the star's water snowline, meaning those particular solids are expected to be water deficient themselves. The prospect the authors raise is delivery, water reaching early planets in roughly the way it is thought to have reached Earth.

Modeling Limits That Keep This Short of Discovery

Several constraints deserve to sit near the top rather than buried at the end. This is a single simulated system, not a population study, so it establishes possibility rather than frequency. The work models planetesimal formation and stops there.

Planetesimals are starting material. Many processes can interrupt the path from a field of them to a finished rocky planet, and the simulation does not follow that path. Separate modeling work by other groups has explored whether vortices in early disks could allow planetesimals to form at lower metallicities, a different mechanism reaching a related conclusion.

The result also inherits the assumptions of its inputs, including the mass distribution of the first stars and how efficiently supernova metals mix into surrounding gas. Both remain active areas of disagreement in the literature.

Observations That Could Test the Idea

The honest answer about verification is that direct confirmation is out of reach. Detecting a protoplanetary disk at those redshifts is far beyond current capability. What is testable is the chain of assumptions.

Pair-instability supernovae themselves are a target. Modeling work has suggested these explosions should be detectable in the near-infrared at high redshift by the James Webb Space Telescope and next-generation ground-based instruments, which would put the population of first stars on firmer observational footing. Water detections in distant star-forming galaxies also bear on whether early enrichment behaved as the simulations require.

For readers who follow exoplanet science, the practical takeaway is a shift in framing rather than a new object to look up, as coverage of the result has emphasized. The population of stars considered plausible planet hosts has been expanding for more than a decade, and this work pushes on the same question from the modeling side. It does not establish that early planets existed, and it certainly does not establish that any were habitable.

The next developments to watch are follow-up simulations that carry planetesimals forward into planet formation, and any JWST detection that constrains the first stellar generation. Until then, the finding stands as a well-argued case that the universe may have had the ingredients much sooner than the textbooks say. The university announcement frames it the same way.

What Readers Want to Know

What did the study actually show? A cosmological simulation produced a protoplanetary disk around a low-mass star in the debris of an early supernova, containing several Earth masses of planetesimals and a substantial amount of water.

Does this mean planets existed that early? No. It shows that the solid building blocks could have formed. The simulation does not follow those planetesimals through to finished planets, and no such system has been observed.

Why do some reports say 100 million years and others 200 million? The arXiv preprint states 200 million years, while the university announcement and the lead scientist's quoted description state 100 million. Related work by the same team describes a window spanning both figures.

Is the research peer reviewed? Yes. It was published in The Astrophysical Journal Letters after review. The preprint has been publicly available since January 2025.

Could early planets have supported life? The paper does not make that claim. It raises the possibility that early planets could have been enriched with water, which is a precondition rather than evidence of habitability.

How would anyone ever confirm this? Not directly with current telescopes. The testable pieces are the first stars and their supernovae, which modeling suggests could be detected in the near-infrared at high redshift, and water in early galaxies.

Who conducted the work? The team was led by Eduard Vorobyov at the University of Innsbruck with Daniel Whalen at the University of Portsmouth, alongside co-authors in the United Arab Emirates, Russia, Japan, and the United States.