For decades, astronomers assumed that all young star clusters shake off their birth clouds on roughly the same schedule. A massive survey using the James Webb Space Telescope and the Hubble Space Telescope has now shown that assumption was wrong. The heaviest clusters blast free of their natal gas in about 5 million years, while lighter clusters stay wrapped in dense cocoons for 7 to 8 million years. That gap, reported in June 2026 in Nature Astronomy (note: this DOI is forward-dated and has not yet been independently verified), rewrites a foundational piece of star-formation theory and carries consequences for how galaxies themselves evolve.
The result comes from the FEAST program (Feedback in Emerging extrAgalactic Star clusTers), a multi-galaxy observing campaign led by astronomer Angela Adamo at Stockholm University. Her team cataloged nearly 9,000 star clusters across four galaxies: the grand-design spirals M51 and NGC 628, the barred spiral M83, and the irregular dwarf NGC 4449. By combining Webb’s infrared vision, which can pierce thick dust, with Hubble’s sharp optical imaging, the researchers tracked each cluster’s transition from a gas-shrouded cocoon to an exposed stellar group.
A mass-dependent clock
The physical story is intuitive once you see the data. When stars ignite inside a dense molecular cloud, their ultraviolet radiation, stellar winds, and eventually supernova explosions push the leftover gas outward. Astronomers call the interval between a cluster’s birth and the moment its surrounding gas fully disperses the “emergence timescale.” Previous models placed that interval somewhere between 5 and 10 million years, with no strong dependence on how massive the cluster was.
The FEAST data reveal a clear dependence. The most massive clusters, packed with the hottest, most luminous stars, generate fiercer radiation fields and more powerful winds. That stronger feedback ionizes, heats, and expels surrounding gas more efficiently, cutting the embedded phase short. Lower-mass clusters, with fewer high-energy stars, push back against their birth material more weakly, so the gas lingers for millions of years longer.
A difference of 2 to 3 million years may sound modest, but in astrophysical terms it is significant. That extra time changes how long a cluster can gravitationally sweep up additional material, how it interacts with neighboring clouds, and how quickly the heavy elements forged inside its stars get recycled into the surrounding interstellar medium.
Four galaxies, one pattern
One of the study’s strengths is its breadth. The four FEAST galaxies differ in structure, metallicity, gas density, and star-formation rate, yet all four show the same qualitative pattern: heavier clusters emerge faster. Earlier work on individual galaxies had hinted at this trend. An analysis focused on M83 described the specific Webb/NIRCam color diagnostics used to sort embedded from exposed clusters, relying on tracers such as hydrogen recombination emission (Paschen-alpha) and 3.3-micron polycyclic aromatic hydrocarbon excess. A separate study of NGC 4449 explored whether pre-supernova feedback alone could explain the clearing or whether supernova explosions were also required. The full four-galaxy paper now synthesizes those threads into a statistically robust result.
Previous studies typically examined individual galaxies or small samples, making it hard to separate real physical trends from local quirks. By assembling nearly 9,000 clusters across four systems, the FEAST team could average over local variations and isolate a mass-timescale correlation that holds across very different galactic environments.
What remains unresolved
The result raises as many questions as it answers. All four galaxies are relatively nearby and represent fairly typical star-forming environments in the local universe. Whether the same mass-dependent clearing timescale holds in more extreme settings, such as compact starburst galaxies or systems in the early universe where gas densities were far higher, has not been tested with comparable statistics.
There is also scatter. Individual clusters of similar mass can emerge at noticeably different times, likely reflecting local differences in gas structure, turbulence, magnetic fields, and the chance presence of especially massive stars. Parsing how much of that spread comes from environment versus the inherent randomness of star formation will require larger samples and more detailed simulations.
A subtler issue involves measurement itself. Identifying exactly when a cluster “emerges” depends on the wavelength and sensitivity of the telescope. Webb’s infrared instruments can detect residual dust and gas around clusters that look fully exposed in Hubble’s optical images. Conversely, some very young clusters may be invisible in optical light while still glowing brightly in the mid-infrared. The FEAST team’s method of combining both datasets strengthens confidence in the timescales, but future instruments with finer spatial resolution and broader wavelength coverage could sharpen the estimates further.
The relative contribution of different feedback mechanisms at different mass scales is still debated, too. Radiation pressure dominates in the most massive clusters, but at lower masses the balance between photoionization, winds, and supernovae is less clear. A companion preprint (note: this arXiv identifier is forward-dated and has not yet been independently verified) provides the full sample description and numerical results in open-access form, offering a starting point for theorists looking to calibrate their models against the new data.
Why the timeline matters beyond clusters
Gas clearing timescales ripple outward through an entire galaxy’s ecosystem. If massive clusters evacuate their birth environments faster, they shut down further local star formation sooner, carving feedback “bubbles” whose sizes and lifetimes depend on cluster mass. That, in turn, affects how efficiently a galaxy converts its gas reservoir into stars over cosmic time.
The speed of gas clearing also governs how quickly heavy elements, oxygen, carbon, iron, and the other products of stellar nucleosynthesis, get mixed into the interstellar medium and become available for the next generation of stars and planets. A faster clearing time means those elements disperse sooner, potentially altering the chemical enrichment history of an entire galactic neighborhood.
There are implications for cluster survival as well. Rapid gas loss can gravitationally unbind a young cluster, scattering its stars into the galactic field. Slower gas loss gives a cluster more time to settle into a gravitationally bound state. Understanding which clusters survive and which dissolve is central to explaining the origin of ancient globular clusters, some of the oldest structures in the universe.
How FEAST reshapes the mass-dependent picture of cluster emergence
For now, the FEAST result establishes a new empirical benchmark: not all star clusters grow up at the same pace. The heaviest ones blast free in about 5 million years; lighter ones take several million years longer. As Webb continues to observe more galaxies and push to greater distances, astronomers will be able to test whether this mass-dependent clock ticks the same way everywhere, or whether the universe has more surprises buried in the dust.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
