Supermassive black holes came before stars in ancient galaxies

What would you see if you could look into the past of every human being on Earth and see them as they were when they were 5 years old? You’d expect to see people possessing a wide variety of traits: some short and some tall, some heavy and some light in weight, some with larger feet and some with smaller feet, etc. Despite this variety, however, you’d fully expect that they’d all look like 5 year olds; you’d be shocked to find one that looked like a teenager, a young adult, or even a middle-aged adult. If this were the case, it would make you question — and rightly so — if what you were seeing was actually reflective of reality.

But in the Universe, this is precisely what occurs when we look at the earliest, brightest, most active galaxies that contain black holes. In theory, there should be a limit to:

• how early you can form the first stars (and, hence, the first black holes),
• how large of a “seed” black hole you can create from those first stars,
• and how quickly those black holes can grow in mass.

And yet, somehow, even in the first few hundred million years after the Big Bang, we see evidence for supermassive black holes that are far more massive than these limits would allow. Although it’s taken 19 years, a team of researchers, in a new paper in Nature, claims to have it all figured out. Here’s how we think the Universe actually does it!

The overdense regions that the Universe was born with grow and grow over time, but are limited in their growth by the initial small magnitudes of the overdensities, the cosmic scale on which the overdensities are found (and the time it takes the gravitational force to traverse them), and also by the presence of radiation that’s still energetic, which prevents structure from growing any faster. It takes tens-to-hundreds of millions of years to form the first stars; small-scale clumps of matter exist long before that, however. Until stars form, the atoms in these clumps remain neutral, requiring ionizing, ultraviolet photons to render them transparent to visible light.

The first thing you have to realize is that, at the very beginning of the hot Big Bang, the Universe is almost perfectly uniform. If you were to draw a sphere around any region of space — regardless of the size of your sphere — you’d find that you had enclosed a certain amount of mass. If you then took that same-sized sphere and enveloped 1000 different regions of exactly the same size, you’d find that:

• about 683 regions had between 99.997% and 100.003% of the average density,
• about 954 regions had between 99.994% and 100.006% of the average density,
• about 997 regions had between 99.991% and 100.009% of the average density,
• and all 1000 regions had between 99.988% and 100.012% of the average density.

In other words, at the start of the hot Big Bang, even the densest regions out there contain only ever-so-slightly more mass than the average.

As a result, along with the physics of how matter and radiation evolve in the early Universe, we can calculate how long we expect it will take for even the initially most dense regions to collect enough matter to form the first collapsed objects: stars and black holes among them. Although there are, of course, uncertainties here, we all agree that it takes tens of millions of years for the very first stars to form, and that the first major wave of star formation in the Universe won’t occur until between 100 and 200 million years have elapsed.

The very first stars to form in the Universe were different than the stars today: metal-free, extremely massive, and nearly all destined for a supernova surrounded by a cocoon of gas. There was a time, prior to the formation of stars where only clumps of matter, unable to cool and collapse, remained in large, diffuse clouds. It is possible that clouds that grow slowly enough may even persist until very late cosmic times.

Once they do form, the most massive among these first stars are expected to live only for a short amount of time: 2 million years or less, before their cores collapse and they form black holes. Given that the most massive known, single star in the Universe today — R136a1 in the Large Magellanic Cloud’s Tarantula Nebula — weighs in at about ~260 times the mass of the Sun, it’s reasonable to assume, as a starting point, that the earliest black holes would also come in at perhaps a few hundred solar masses.

The big questions, then, are how early can we form these seed black holes, and how fast can they grow and gain mass as the Universe evolves?

If we stick to the early end of the time spectrum, we might be able to form them just 100 million years after the Big Bang. Then, if you ask how they grow, there’s a limit set by the laws of physics for how quickly a single mass can grow by accumulating the matter surrounding it: the Eddington limit. That allows us to come up with a prediction for how massive black holes should be allowed to get in a given amount of time. This is great, because then all we need to do is to go out and measure the masses of the supermassive black holes that power the earliest, most energetic quasars, and to see if our observations are consistent with what we predict.

If you begin with an initial, seed black hole when the Universe was only 100 million years old, there’s a limit to the rate at which it can grow: the Eddington limit. If seeds of several tens-of-thousands of solar masses arise early on and these SMBH seeds grow rapidly thereafter, there may be no conflict with what’s observed, after all.

And this is where the puzzle arises. We’ve now found quasars with supermassive black holes anywhere from 500 million to slightly over 1 billion solar masses that date back to a time where the Universe was just around ~700 million years old: a mere 5% of its current age.

When we do this, we find that we’re not just seeing the proverbial 5-year-old humans with a variety of heights, weights, foot sizes, and other properties. It’s as though some of those 5-year-olds are as tall as full-grown basketball players, reaching sizes that we didn’t expect them to reach until they were much, much older.

In other words, something doesn’t add up. These earliest supermassive black holes have grown larger than we expect that they should have given what we know about the Universe and how the laws of physics work over time.

In the recent past, many ideas have been floated to explain why, but they all fall into three categories.

1. Perhaps we’ve got the seed-size and time right, but the growth wrong, and that black holes masses grow faster than we realize.
2. Perhaps we have the seed-size all wrong, and that larger cosmic seeds are possible, driven by processes like structure formation or a much greater initial mass for the heaviest stars.
3. Or, perhaps our picture is way off base, and the Universe was actually born with black holes that formed before any stars ever could: a set of either primordial or pre-stellar black holes.

If the Universe was born with primordial black holes, a completely non-standard scenario, and if those black holes served as the seeds of the supermassive black holes that permeate our Universe, there will be signatures that current and future observatories, like JWST and LISA, will be sensitive to. Measuring the growth rate of black holes over time is one key test; looking for fully evaporating black holes (for which there is no evidence) is a second.

For the first option, it’s true that non-spherical accretion can cause black holes to grow faster than the Eddington limit, but that type of accretion struggles to persist for long periods of time. Even if we allow for bursts of faster-than-expected growth, it remains very difficult to explain how so many supermassive black holes — as there are now some ~200 of them discovered at very early cosmic times — all experienced what should be rare and transient conditions for such persistent epochs.

For the third option, this starts off as a remarkably unattractive proposition: we’d have to postulate some new, never-before-seen physics to create a “spike” in the mass spectrum at a particular value. In order to make a primordial black hole, you need to have a region of space that’s more dense than 168% of the cosmic average in the early stages of the expanding Universe. But remember: we just said that about ~100% of all early regions in space are between 99.988% and 100.012% of the average density. Unless you invent some new way to have ultra-large-magnitude fluctuations on a very small and specific cosmic scale (and no other), this scenario cannot come to pass.

The largest group of newborn stars in our Local Group of galaxies, cluster R136, contains the most massive stars we’ve ever discovered: over 250 times the mass of our Sun for the largest. The brightest of the stars found here are more than 8,000,000 times as luminous as our Sun. And yet, these stars only achieve temperatures of up to ~50,000 K, with white dwarfs, Wolf-Rayet stars, and neutron stars all getting hotter.

However, there’s still a hope that this problem — or rather, this apparent problem — may simply turn out to result from ordinary, mundane, run-of-the-mill astrophysics. Sure, a seed black hole of a few hundred solar masses at an epoch of ~100 million years after the Big Bang won’t quite do the job, but if we could make seed black holes that were just a hundred times more massive at that same, early epoch, that would provide a way out. If the Universe could give birth to a seed black hole that was a few tens of thousands of solar masses just ~100 million years after the Big Bang, that would resolve the tension.

The composition of the Universe, back before the very first stars ever formed, provides a tremendous clue that this might be possible. Today, stars form from the collapse of gas clouds, made mostly of hydrogen and helium, but peppered with a small amount of heavier elements: oxygen, carbon, nitrogen, neon, silicon, sulfur, calcium, magnesium and iron, among others. It’s primarily those heavy elements that allow those gas clouds to cool and collapse, giving rise to stars with an average mass of 40% that of the Sun. Sure, we might be able to form a few very massive stars with masses of 50, 100, or even 200 solar masses or more, but just a tiny fraction of them will be massive enough to go supernova and/or leave a black hole remnant behind.

The visible/near-IR photos from Hubble show a massive star, at least 25 times the mass of the Sun, that has winked out of existence, with no supernova or other explanation. Direct collapse is the only reasonable candidate explanation, and is one known way, in addition to supernovae or neutron star mergers, to form a black hole for the first time. The direct collapse of this particular object, while still under investigation, may have been triggered by a stellar companion.

Early on, however, star-formation would have had to proceed differently. Without any of those heavy elements present, the most efficient way to cool clouds of gas that are attempting to gravitationally contract is still tremendously inefficient: to have neutral, molecular hydrogen radiate that heat away. With such an inefficient means of cooling at play, molecular clouds have to get much more massive in order to collapse, and the expected stars they’ll form are ~25 times more massive than today’s stars, on average.

It might be quite common to form stars of many hundreds or even a few thousand solar masses in these very early stages, and those stars might not even go supernova; it’s plausible that they’ll directly collapse, leading to a scenario where 100% of the star’s mass becomes a black hole.

Given that:

• we’ve seen stars directly collapse,
• in the environments where the most massive stars reside, we find many stars of comparably high masses,
• and that the first stars formed from ultra-massive molecular clouds with (likely) many stars of ~1000 solar masses,

it seems reasonable that, perhaps, many “seed” black holes cloud form, and those seeds could then merge together, leading to a starting black hole that just might, potentially, start off with enough mass for us to wiggle our way out.

A mathematical simulation of the warped space-time near two merging neutron stars that result in the creation of a black hole. The colored bands are gravitational-wave peaks and troughs, with the colors getting brighter as the wave amplitude increases. The strongest waves, carrying the greatest amount of energy, come just before and during the merger event itself. What occurs outside the event horizon is not practically affected by whether there is a ring singularity at the center, or some other, extended object that is non-singular.

But even this scenario, which gets us quite a bit closer to the desired outcome, poses a difficulty: how do we get these seed black holes to all merge together quickly enough without gravitational interactions either forcing an ejection or mutual interactions clearing out the galactic center from the needed material for accretion?

There needs to be something else. Something else has to come into play if we’re to transform this conundrum from relying on hand-waving to relying on solidly understood physics and astrophysics. And that’s precisely where a remarkable study in 2022, led by Daniel Whalen of the University of Portsmouth, came in to save the day.

By simulating how structure forms in the early Universe, including dark matter, early star clusters, and streams of neutral gas (i.e., normal matter), and watching how proto-galaxies and star clusters merge together amidst the backdrop of the emerging cosmic web, cosmologists can monitor where large, massive collections of matter can gather in one particular location. A few years ago, this method was able to reveal that cold, ultramassive streams of gas would collide at the nexus points of this proto-cosmic web, and that gas could rise to high densities in small volumes, containing as much as ~100,000 solar masses in one particular location.

This snippet from a supercomputer simulation shows just over 1 million years of cosmic evolution between two converging cold streams of gas. In this short interval, just a little over 100 million years after the Big Bang, clumps of matter grow to possess individual stars containing tens of thousands of solar masses each in the densest regions, and could lead to direct collapse black holes of an estimated ~40,000 solar masses. This could provide the needed seeds for the Universe’s earliest, most massive black holes, as well as the earliest seeds for the formation of stars and the growth of galactic structures.

But these environments are rare, and cannot explain the ~200+ very massive quasars we’ve discovered back when the Universe was just ~5% of its current age or so. That’s where the new simulations by the Portsmouth group comes in. The team was able to show that where strong, cold accretion flows occur and converge, they can all of a sudden trigger the all-at-once collapse of dense clouds of normal matter, creating either short-lived stars or direct-collapse black holes that range from 30,000 to 40,000 solar masses.

As the simulation snippet, above, illustrates, there’s no need for any fine-tuning mechanisms to trigger this phenomenon. Whereas previous studies invoked ultraviolet backgrounds, supersonic streaming motions, or some variety of atomic and molecular cooling, this new research showed that these cold gas flows lead to a copious amount of turbulence, which prevents star-formation from occurring entirely until a critical mass is reached. Upon crossing that mass threshold, however, the dense region suddenly collapses, which can trigger the formation of individual objects — stars or black holes — of up to 40,000 solar masses. For the first time, using known, non-exotic physics alone, a team has successfully shown that seed black holes of the needed mass can be created just slightly over 100 million years after the Big Bang.

When the data from a variety of “little red dot” galaxies are broken up into their stellar mass component versus the component arising from an active supermassive black hole, the mass ratios of the galaxy’s total stellar mass compared with the supermassive black hole’s mass can be determined. Many, and perhaps even most, of these black holes are found to be significantly overmassive: at much more than 0.1% of the mass of the stellar component.

This theoretical development couldn’t come at a better time: it came right as JWST’s science operations were about to begin. One of its unprecedented capabilities will be to examine the earliest black holes in their galactic environments, shedding light as to how they formed and grew up. This novel scenario, where cold gas streams collide to produce monster-sized stars that can quickly form black holes up to tens-of-thousands of solar masses, will soon be put to the ultimate test. As Whalen himself noted:

“The only primordial clouds that could form a quasar just after cosmic dawn, when the first stars in the universe formed, also conveniently created their own massive seeds. This simple, beautiful result not only explains the origin of the first quasars but also their demographics: their numbers at early times. The first supermassive black holes were simply a natural consequence of structure formation in cold dark matter cosmologies: children of the cosmic web.”

For nearly two full decades, astronomers have puzzled over just how these supermassive black holes, found in the Universe’s most distant quasars, grew up to be so big so fast. This new result strongly supports a novel, gas-driven scenario that requires no new physics. Over the subsequent three years, we have indeed discovered overmassive black holes, supporting this early “seed” formation scenario for supermassive black holes. We at last think we know where black holes come from, and that indeed, in the chicken-and-egg story of stars vs. black holes in galaxies, it now looks like black holes did indeed come first!

This article was first published in July of 2022. It was updated in November of 2025.