Early Universe's supermassive black holes grew in cocoons like butterflies

During cocoon phase, young, supermassive black holes are surrounded by high-density gas.

When the James Webb Space Telescope sent its first high-definition infrared images back to Earth, astronomers noticed several tiny, glowing, crimson stains. These objects, quickly named “Little Red Dots,” were too bright to be normal galaxies, and too red to be simple star clusters. They appeared to house supermassive black holes that were far more massive than they had any right to be.

But now a new study published in Nature suggests a solution to the Little Red Dots mystery. Scientists think young supermassive black holes may go through a “cocoon phase,” where they grow surrounded by high-density gas they feed on. These gaseous cocoons are likely what the JWST saw as the Little Red Dots.

The overmassive black hole problem

The first explanation scientists had for the Little Red Dots was that they were compact, distant galaxies, but something felt off about them right from the start. “They were too massive, since we saw they’d have to be completely filled with stars,” says Vadim Rusakov, an astronomer at the University of Manchester and lead author of the study. “They would need to produce stars at 100 percent efficiency, and that’s not what we’re used to seeing. Galaxies cannot produce stars at more than 20 percent efficiency, at least that’s what our current knowledge is.”

The next possible solution for the Little Red Dots was that they were supermassive black holes, but this didn’t fit the data either. For decades, astronomers have observed a consistent relationship between a galaxy and its central supermassive black hole: Usually, the black hole accounts for about 0.1 percent of the galaxy’s total mass. It’s a tidy cosmic ratio that suggests the two grow together in a tightly coordinated manner. The Little Red Dots broke that rule.

Initial analyses suggested if the dots were in fact supermassive black holes, they’d have to be nearly as massive as their host galaxies, varying between 10 and 100 percent of their total mass. The problem was that the dots were visible at very high red shift, which means astronomers saw them as they were when the Universe was roughly 1 billion years old. An “overmassive” black hole as heavy as its entire galaxy in a 1-billion-year-old Universe begged the question how something could grow that big, that fast—we didn’t have any answers for that.

But then Rusakov and his colleagues started noticing odd things in the JWST data. “You normally expect other signals from supermassive black holes, like X-rays, and we didn’t see those signals,” Rusakov says. The oddities didn’t end with the absence of X-rays.

The wide lines

Because black holes can’t be observed directly, astronomers measure their mass by looking at gas orbiting around them. As the gas swirls down into the black hole, it heats up and glows. The gravity of a supermassive black hole pulls that gas at incredible speeds, with the material reaching thousands of kilometers per second. This speed causes what’s known as the Doppler effect, broadening where the light from the gas moving toward the observer on Earth shifts to blue, and the gas moving away shifts to red, stretching the spectral lines into a wide, flat shape. By measuring the width of these lines, we calculate the velocity of the gas and, by extension, the mass of the black hole.

In the case of the Little Red Dots, the lines appeared incredibly wide, leading to those staggering mass estimates. The shape of the lines, though, looked strange. It wasn’t a typical rounded bell-curve but rather a sharp triangle sitting on top of broad, wing-like tails.

The breakthrough came when the team realized they weren’t looking at gas moving fast. They were looking at light getting lost in the fog.

Scaling the giants down

In the case of Little Red Dots, the fog was a dense cocoon of ionized gas—specifically, a thick cloud of free electrons surrounding the black holes. The unusual shape of the spectral lines, Rusakov’s team reports in the paper, is due to the process called Thomson scattering. Photons emitted by gas near the black hole are likely hitting these free electrons, and their direction and energy gets shifted with each collision. After billions upon billions of such collisions, a spectral line that was originally quite narrow becomes broad, mimicking the look of high-velocity gas.

By applying a scattering model to data from the Little Red Dot galaxies, Rusakov’s team found that the intrinsic velocity of the gas was actually much lower than we thought. Scientists concluded that the black holes are likely 100 times smaller than previous estimates. Instead of being “overmassive” monsters that defy our understanding of physics, they are likely “young” supermassive black holes, roughly 10 million to 100 million times the mass of our Sun. This puts them much closer to the standard galaxy-to-black-hole mass ratio we see in the local Universe.

The JWST apparently caught these black holes at the phase in their lifecycle we’ve never seen before.

The cocoon phase

The study suggests that the Little Red Dots represent a previously unknown stage of supermassive black holes’ evolution. “They look like a [developing] butterfly or something in this young state that kind of grows wrapped in some sort of gas that also feeds it,” Rusakov says. “It’s definitely new in the sense people didn’t predict there should be such a cocoon phase in the supermassive black holes’ lifecycle.”

During this phase, a young supermassive black hole is growing rapidly, buried deep within a dense shell of gas and dust. This cocoon is so thick that it acts as a cosmic shield that blocks the high-energy X-rays and radio waves that typically signal the presence of an active black hole. Rusakov’s team thinks their cocoon hypothesis explains one of the most persistent mysteries of Little Red Dots—why they are so bright in infrared light but virtually invisible to X-ray telescopes like Chandra. The X-rays are being absorbed by the same dense material that is scattering the light. “This was a neat solution in that sense,” Rusakov says.

But while the scattering model elegantly solves the mass problem and the missing X-rays, questions remain. Astronomers still need to determine exactly how long this cocoon phase lasts and how common it is in the early Universe. So far, the team studied 12 Little Red Dot objects. As more high-resolution data pours in from JWST, researchers will be able to see if every Little Red Dot follows this pattern. And this, Rusakov thinks, should give us some insights into how our own galaxy came to be.

Based on the JWST observations, it seems like most of the signal we see coming from early-stage galaxies like Little Red Dots is the light from supermassive black holes. “There’s this big, simplified question: Does the galaxy start with the supermassive black hole or with the stars? Is that a chicken or the egg?” Rusakov says. “We don’t know exactly what happens in this first sort of stage of galaxy formation. But our model gives us a new way to look at this kind of object.”

Nature, 2026. DOI: 10.1038/s41586-025-09900-4