Planets are built against a deadline. The gas that gives Jupiter and Saturn their enormous envelopes is not a permanent feature of a young star system - it is a dwindling resource. Once it drains away, the opportunity to assemble a gas-rich planet is essentially over.
A new survey with NASA's James Webb Space Telescope has now watched that drain in action across 72 young, Sun-like stars, and it shows the escape route itself changing as a system ages.
The work, led by Naman Bajaj of the University of Arizona with SETI Institute scientist Uma Gorti among the co-authors, was published in The Astronomical Journal in August 2026 under the title "JWST/MIRI Reveals the Evolution from Molecular to Atomic Disk Winds." It is one of the largest planet-formation studies Webb has produced to date.
The clock that governs planet formation
Around 4.5 billion years ago, the young Sun sat inside a thick protoplanetary disk holding roughly a hundred times more gas than dust. Almost all of that gas is gone now. Where it went, and how quickly, is not a bookkeeping detail - it is one of the most important constraints on what kind of planets a system can build.
Rocky planets can keep assembling from solids long after the gas has cleared. Gas giants cannot. A Jupiter needs to reach a critical core mass and then pull down a massive hydrogen-helium envelope while that envelope is still available. If the disk disperses first, the system is left with cores that never became giants.
"Disk dispersal sets a fundamental clock for planet formation. Once the gas is gone, the opportunity to build gas-rich planets is essentially over." - Uma Gorti, SETI Institute
Two tracers, seventy-two systems
To read that clock, Bajaj's team went to the archive rather than the telescope queue, pulling existing observations from Webb's Mid-Infrared Instrument (MIRI) for 72 young systems. No single disk can show you an evolutionary sequence - each one is a snapshot. But 72 disks caught at different ages behave like frames of a film: line them up and the process becomes visible.
The team followed two signatures of escaping gas. The first is molecular hydrogen, the most abundant molecule in a protoplanetary disk and, until Webb, very difficult to trace directly in these outflows. The second is ionized neon, which marks hotter, faster, more collimated material - the jets.
MIRI's sensitivity and spatial resolution were the enabling factors. With both tracers resolved, the researchers could separate the broad molecular winds from the narrow atomic jets threading through them, rather than lumping all outflowing gas together.
Extended emission from molecular hydrogen or neon turned up in 66 of the 72 disks. Conical molecular-hydrogen winds appeared in 46 systems, and fast neon jets in 40. Notably, every system showing a neon jet also showed a slower wind traced by molecular hydrogen or oxygen - the jets never appear alone.
From magnetic winds to photoevaporation
The pattern across the sample is a handover between two very different physical mechanisms.
In the youngest systems, still actively funneling material onto their central stars, Webb sees strong jets alongside broad winds containing both molecular and atomic gas. That combination is the expected signature of magnetically driven winds: gas threaded onto magnetic field lines running through the disk, flung outward, carrying away mass and, critically, angular momentum. Removing angular momentum is part of what allows the remaining disk material to spiral inward and feed the star.
As accretion slows and the disk thins, the picture changes. The jets weaken. The winds become predominantly atomic. And the star's own high-energy ultraviolet and X-ray radiation, previously absorbed by dense overlying material, can now reach and heat the disk gas directly until it escapes. This is photoevaporation, and the survey shows it taking on a larger role as disks age.
There is a satisfying loop closed here. In 2020, a team led by Ilaria Pascucci - Bajaj's adviser and second author on the new paper - predicted from jet and wind evolution that massive molecular winds should exist at early ages, dense enough to block stellar X-ray photons from reaching the disk. At the time, molecular hydrogen could not be observed directly in these systems. Webb has now confirmed the prediction by seeing it.
No single culprit
The headline conclusion is that disk dispersal has no single mechanism. Systems appear to migrate from an early, magnetically driven regime into a later one where atomic winds, including photoevaporative winds, play a bigger role, with the balance shifting gradually rather than switching over at a sharp boundary.
"Planet formation is therefore a race against time. Gas giants like Jupiter must assemble their massive atmospheres while the disk is still substantial enough to supply them, before winds and jets carry that raw material away into space." - Naman Bajaj, University of Arizona
The study extends earlier work by the same group. In 2024, Bajaj, Gorti and colleagues used Webb to image the disk wind streaming away from the young star T Cha, proof that dispersal could be studied directly in a single system. The new survey scales that approach to dozens.
What comes next
The obvious follow-up is quantitative. Detecting a wind is one thing; measuring how much mass it removes per year, and from which radii in the disk it is launched, is another. Those numbers would convert a qualitative sequence into an actual timetable - not just how fast the planet-forming window closes, but where in the disk it closes first, and therefore which orbital neighborhoods can still produce gas giants when the clock runs down.
Sources
- SETI Institute, "JWST Reveals a Race Against Time for Forming Planets" (25 August 2026).
- N. S. Bajaj et al., "JWST/MIRI Reveals the Evolution from Molecular to Atomic Disk Winds," The Astronomical Journal (2026), DOI 10.3847/1538-3881/ae9089.
- Preprint: arXiv:2607.21733.
