When a galaxy burns through its star-forming fuel at 100 times the rate of the Milky Way, you might expect it to simply run out of gas. But new research published in 2025 and early 2026 reveals something stranger: many of these so-called starburst galaxies still have plenty of raw material on hand when they go dark. The real problem, according to a growing body of direct observations, is that something breaks the pipeline that converts gas into stars.
The most striking clue comes from timing. A study led by astronomer Kate Rowlands and collaborators, examining massive post-starburst galaxies at intermediate redshift, found that star formation shuts down roughly 100 million years before the molecular gas reservoir disappears. Stars stop forming first. The fuel vanishes later. That sequence upends the simplest explanation and points to an active suppression mechanism, something that chokes off star birth while raw material is still available.
Now, results from the EMBERS I survey, published in Monthly Notices of the Royal Astronomical Society, are filling in the picture with the most detailed molecular gas census of nearby post-starburst galaxies to date.
The gas is there, but it is in the wrong form
Post-starburst galaxies, often called PSBs, are systems that recently experienced an intense burst of star formation and then abruptly stopped. The EMBERS I team measured molecular gas levels in a sample of low-redshift PSBs and compared them against a carefully matched control group of actively star-forming galaxies. The result was unambiguous: PSBs are systematically depleted in molecular hydrogen (H2), the dense gas that directly fuels new stars.
But a separate study, currently available as a preprint on arXiv, complicates the picture in a revealing way. Many of these same quenched galaxies still hold substantial reserves of atomic hydrogen (HI), the diffuse raw material that must cool and condense into molecular clouds before stars can form. The gas has not been blown out of the galaxy. It is sitting there, unable to make the transition to its denser, star-forming state.
An ALMA-based study of the physical conditions inside PSBs adds another dimension. Researchers examining dense-gas tracers and molecular line ratios found that even when some molecular gas persists, the fraction of the densest gas, the kind that directly collapses into stars, is reduced. The interstellar medium in these galaxies appears to have been stirred up, heated, or pressurized in ways that prevent clouds from collapsing under their own gravity.
A timeline that rules out the obvious answer
The 100-million-year gap between the end of star formation and the disappearance of molecular gas, as established by Rowlands and collaborators studying massive PSBs at around redshift 0.6, is one of the most important constraints researchers have identified. If galaxies simply exhausted their fuel, you would expect gas depletion and the end of star formation to happen simultaneously. Instead, the shutdown comes first, and the gas lingers.
That sequence strongly suggests an active quenching process, something that poisons the conditions for star formation before the raw ingredients are gone. Only later, over tens of millions of years, does the remaining molecular gas disperse, heat up, or get stripped away.
Baseline surveys like xCOLD GASS (for molecular gas) and xGASS (for atomic gas) provide the comparison data that make these conclusions robust. By establishing how much gas normal star-forming galaxies hold at a given stellar mass, these reference datasets confirm that the depletion seen in PSBs is real and not an artifact of how the samples were selected.
More recent work on the EMBERS I sample (a 2026 preprint expanding on the published survey results), combining CO emission measurements with stellar population diagnostics and tightly matched control galaxies, has strengthened the statistical foundation. By tying gas content directly to well-dated starburst histories, the team can separate genuinely post-starburst systems from dusty star-forming galaxies that might otherwise be misidentified as quenched.
The suspects: what could shut down the pipeline?
Identifying the specific mechanism responsible remains the central open question. Several candidates are on the table:
Turbulence from mergers or the starburst itself. A violent burst of star formation can inject enormous energy into the surrounding gas, heating it and preventing it from settling into the dense clouds needed for further star birth. Galaxy mergers, which are thought to trigger many starbursts, can have a similar effect by scrambling gas orbits and creating shear that tears apart nascent clouds.
Active galactic nuclei (AGN). A supermassive black hole at the galaxy’s center can drive winds or bathe the interstellar medium in radiation, destabilizing molecular clouds or heating gas across kiloparsec scales. AGN feedback is a leading candidate in theoretical models of galaxy quenching, but the observational studies in this body of work have not yet systematically measured AGN properties alongside gas conditions in the same PSB samples.
Geometric redistribution. Gas may be spread into extended, low-density configurations where it cannot reach the densities required for gravitational collapse. In this scenario, the total gas mass might look adequate on paper, but its spatial distribution makes star formation impossible.
The 100-million-year delay favors feedback-driven explanations over simple exhaustion, but no single study has isolated which process dominates in typical post-starburst galaxies. It is also possible that different mechanisms operate in different systems, or that several act in concert.
How far these results reach, and where the gaps remain
The EMBERS I molecular gas findings apply to nearby, low-redshift PSBs. The 100-million-year delay was measured in massive galaxies at intermediate redshift (around z of 0.6). Whether the same shutdown sequence plays out in lower-mass galaxies, or at higher redshifts closer to the peak of cosmic star formation roughly 10 billion years ago, has not been established.
The atomic hydrogen results add another layer of complexity. Not all PSBs retain HI, and the amounts vary widely across the population. Environmental factors, such as whether a galaxy sits inside a dense cluster or in relative isolation, likely influence how much gas is stripped or re-accreted, but current samples are too small to map those trends with confidence.
There is also the question of what happens next. In some galaxies, both molecular and atomic gas appear to be genuinely depleted or expelled over time. In others, gas persists indefinitely but in a state that no longer forms stars efficiently. Whether these represent different stages of a single evolutionary pathway or fundamentally different quenching channels is not yet clear. Spatially resolved observations, mapping exactly where gas sits relative to stellar structures and any AGN signatures, will be essential to sorting this out.
What multi-tracer surveys must resolve about post-starburst quenching
The picture emerging from these studies as of early 2026 marks a genuine shift in how astronomers think about galaxy death. Post-starburst galaxies do not simply burn through their fuel and fade. Their gas supply is reconfigured, heated, or otherwise altered so that the usual pipeline from diffuse atomic hydrogen to dense, star-forming molecular clouds breaks down. The fuel is there. The factory is offline.
Pinpointing the agents responsible, whether supermassive black holes, stellar feedback, environmental stripping, or some combination, is now the central challenge. Larger PSB samples with simultaneous measurements of molecular gas, atomic gas, AGN activity, and resolved gas kinematics will be needed to move from correlation to cause. Upcoming survey programs and continued ALMA observations are expected to provide exactly that kind of multi-tracer data, turning what is currently a well-documented mystery into a testable, mechanistic explanation for how galaxies transform from brief, furious starbursts into long-lived, quiescent systems.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
