The host star is running on fumes. HR 8799 — 129 light-years away in the constellation Pegasus — has only 35% of the sulfur found in our Sun, according to Wikipedia's entry on the system. Yet all four gas giants orbiting it are enriched in sulfur, and the enrichment gets stronger the farther you look from the star: roughly 2 times the solar level at the inner edge of the system, climbing to 5 times at the outer edge. Nobody has explained why.
The finding, from JWST observations published this month in the Astrophysical Journal by a team led by Jerry Xuan of UCLA, is the first detection of hydrogen sulfide — the compound that gives rotten eggs their smell — in the atmospheres of these four planets. The team also found carbon monoxide, methane, water vapor, and carbon dioxide. But the sulfur is what matters, because it shouldn't be there in these quantities, and the gradient makes no sense at all.
"You would expect a star's composition to set the ceiling for what its planets contain," Xuan said in a statement accompanying the paper. "The fact that we're seeing the opposite — planets significantly enriched where the star is depleted — suggests something in the formation process is concentrating sulfur in ways our models don't predict."
The standard picture of how gas giants form is called core accretion: a rocky core forms first, then pulls in surrounding hydrogen and helium gas. The competing theory, gravitational instability, suggests gas giants can form rapidly from direct collapse of the protoplanetary disk. Both frameworks struggle with what JWST found in the HR 8799 system.
The team's analysis points toward core accretion with an important addition: pebble accretion. This is the process where small particles drift inward through the protoplanetary disk, evaporate as they cross various snowlines (distances from the star where specific compounds transition from solid to gas), and deliver their contents to growing planets. The data implies a total of roughly 750 Earth masses of material flowed through the disk over the star's 30-million-year lifetime. That is a large number, and it fits the pebble accretion model reasonably well. But the sulfur gradient does not follow the pattern the model would predict, and the authors describe the trend as tentative.
The broader significance is in what this might mean for our own solar system. Jupiter and Saturn are enriched in heavy elements relative to the Sun — a fact that has long been taken as evidence for core accretion. If the HR 8799 system, one of the few multi-planet systems directly imaged by ground-based telescopes, follows the same process, it suggests our own system's chemistry may be typical rather than exceptional. The uniform enrichment in carbon and oxygen across all four planets — 3 to 5 times the stellar abundance — is consistent with this reading.
"We may be seeing the same mechanism operating in a different planetary system," said co-author Jean-Baptiste Ruffio of UC San Diego. "If the enrichment patterns are universal, it constrains formation models significantly."
But the sulfur paradox is the piece that doesn't fit. If it holds up — if future observations confirm the gradient — it means our solar system's chemistry is not guaranteed to be the galactic default. It means there are formation pathways we haven't accounted for, operating in disks we haven't fully modeled. That is a significant thing to learn from four planets 129 light-years away.
The data is available for independent verification. Xuan's team released the full spectral dataset on Zenodo, DOI 10.5281/zenodo.19355669. The measurement requires JWST's NIRSpec instrument — ground-based telescopes cannot do this work at this wavelength range. Whether other teams can reproduce the sulfur gradient with the same data will be one of the first tests.
The next telescope mission that matters here is Ariel, Europe's Atmospheric Exoplanets Survey mission scheduled for launch in 2029. It will characterize the atmospheres of roughly 1,000 exoplanets in a way JWST is not designed to do at scale. If the sulfur paradox is real, Ariel will find more examples. If it isn't — if the gradient is an artifact or a statistical fluctuation — then HR 8799 remains a fascinating system but not a paradigm shift.
For now, the finding sits in the uncomfortable place where new data creates more questions than answers. The scientists said so themselves in the paper's conclusion, which is the right place for that kind of honesty.