Dark matter may not have needed a calm, cold start to help build the universe.

For decades, that idea would have sounded backward. Cosmologists have generally held that dark matter had to be born cold, meaning slow-moving, because fast particles would wash out small structures before gravity could pull matter together. In that picture, anything too hot at birth would blur the seeds of galaxies instead of helping shape them.

The new analysis reopens that assumption by looking closely at a period just after inflation, the brief burst of rapid expansion thought to have taken place in the infant universe.

Instead of assuming the universe heated up instantly after inflation ended, the authors focused on reheating, a transition period in which the inflaton field slowly decayed and transferred energy into particles and radiation. That timing turns out to matter.

“Dark matter is famously enigmatic. One of the few things we know about it is that it needs to be cold,” Stephen Henrich, a graduate student in Minnesota’s School of Physics and Astronomy, said in a statement. “As a result, for the past four decades, most researchers have believed that dark matter must be cold when it is born in the primordial universe. Our recent results show that this is not the case; in fact, dark matter can be red hot when it is born but still have time to cool down before galaxies begin to form.”

The mechanism at the center of the study is called ultrarelativistic freeze-out, shortened to UFO. In ordinary language, it describes dark matter stopping its interactions with ordinary matter while it is still moving extremely fast. Even after that decoupling, however, the universe keeps expanding. As space stretches, particle momenta drop. By the time cosmic structure begins to grow, the once-fast dark matter can behave as if it were cold.

That possibility depends on when freeze-out happens. If it occurs during reheating, before the universe settles into the more familiar radiation-dominated phase, the particles get extra time to cool.

The idea carries echoes of an older problem in cosmology. Neutrinos, for example, decoupled while moving close to light speed. Because they remained too fast for too long, they became the textbook example of hot dark matter, the kind that would erase galactic-scale structure rather than seed it.

“The simplest dark matter candidate, a low mass neutrino, was ruled out over 40 years ago since it would have wiped out galactic size structures instead of seeding it,” Keith Olive said. “The neutrino became the prime example of hot dark matter, where structure formation relies on cold dark matter. It is amazing that a similar candidate, if produced just as the hot big bang universe was being created, could have cooled to the point where it would in fact act as cold dark matter.”

That distinction rests on the universe’s changing expansion history. Under the standard assumption of instantaneous reheating, ultrarelativistic freeze-out usually leaves dark matter too warm. But once that shortcut is dropped, the authors found a broad region of parameter space where the same basic kind of fast freeze-out can still end in cold dark matter.

Their calculations show that the key lies in how quickly the dark matter interaction rate falls with temperature compared with how the universe’s expansion rate changes. For some kinds of interactions, especially those involving heavy mediator particles, there is a large intermediate zone between two long-discussed dark matter ideas: WIMPs, or weakly interacting massive particles, and FIMPs, or feebly interacting massive particles.

That middle ground is one reason the work stands out.

WIMPs have been a leading dark matter candidate for years, but direct detection experiments have steadily tightened the limits on them. FIMPs sit at the other extreme, interacting so weakly that they are very hard to detect directly or indirectly. The new study describes UFO as a robust production mechanism occupying a broad space between those two categories.

In some interaction models, the transition from WIMP-like behavior to FIMP-like behavior does not skip cleanly from one to the other. Instead, it passes through this ultrarelativistic freeze-out regime. In other cases, the authors write, WIMP-style freeze-out may not happen at all because of theoretical bounds, while UFO and freeze-in remain allowed.

The paper also spells out conditions for when this can happen. Interactions with a steep enough temperature dependence can drive ultrarelativistic freeze-out during reheating. By contrast, some simpler contact interactions do not support that route.

Just as important, the team tested whether this reheating-era dark matter would still be too warm by the time structures began to form. Their conclusion was strikingly direct: if ultrarelativistic freeze-out happens during reheating, dark matter with a mass above about 5 kiloelectron volts is automatically cold enough by the onset of structure formation. The study notes that this stands in sharp contrast to the standard assumption that neutrino-like freeze-out must always produce hot or warm dark matter.

The result does more than expand the dark matter menu. It also points back toward a poorly understood chapter of cosmic history.

“With our new findings, we may be able to access a period in the history of the Universe very close to the Big Bang,” Yann Mambrini, a professor at Université Paris-Saclay, said in a statement to The Brighter Side of News.

That is one of the study’s bigger implications. Many dark matter scenarios are framed in ways that erase most memory of inflation and reheating. This one does not. If the relic abundance of dark matter was set during reheating, then observations or experiments that narrow the properties of dark matter might also tell physicists something about conditions in the universe before the hot big bang fully took shape.

The authors say the idea applies not just to the specific reheating model used in the paper, but to a wider class of scenarios. They also point to several well-motivated beyond-standard-model settings where this production route could matter, including heavy-mediator interactions and portal models.

For a field that has spent decades circling a few familiar candidates, that opens new territory.

The findings widen the search map for dark matter by reviving models that had often been dismissed as too hot to match the observed universe. If dark matter could freeze out while still moving ultrarelativistically during reheating and later cool into a cold relic, theorists gain a larger set of viable candidates to test.

That could affect how experiments are designed and interpreted. Searches in colliders, scattering experiments, and cosmological observations may need to pay closer attention to dark matter models that sit between standard WIMP and FIMP pictures, especially those involving heavy mediators and early-universe reheating effects.

The work also gives cosmologists a new way to connect dark matter physics to one of the least understood stages of cosmic history. If future evidence supports this mechanism, it could sharpen models of how the universe transitioned out of inflation and how the matter that later shaped galaxies first emerged.