Out there in the Universe, there are some incredibly impressive astrophysical entities, including behemoths that dwarf anything we’re familiar with. There are stars hundreds of times more massive, and millions of times as luminous, as our own Sun is. There are galaxies and galaxy clusters thousands of times as massive as our Milky Way and Local Group. And there are black holes out there that not only dwarf our own Sun, but black holes that are so massive, they’re comparable in mass to all of the stars in the Milky Way galaxy combined. Some of them, quite puzzlingly, at first glance, even exist extremely early on in cosmic history: just a few billion years after the hot Big Bang first began.
At those extremes, the largest black holes pose a number of important cosmic questions. How did they grow so big so fast? Can physics even account for the highest-mass ones? And why do we see the biggest ones at such great distances, rather than close by? Is there something missing in our understanding? These questions, among others, inspired Eric Trant to write in and ask about TON 618: the black hole that first challenged our conventional understanding of how these supermassive black holes form and grow up in our Universe.
“My question tonight has to do with TON 618, the largest black hole we’ve observed to date, at ~66 billion solar masses. Seeing as we’re incredibly far away, observing it as it was 10.8 billion years ago… how did it grow to be so large when it’s only [been] 3 billion years [since] the universe itself [began]? It also seems odd that we don’t see more of similar size, closer to home?”
The supermassive black holes that we know of pose many mysteries, and TON 618 is the one that kicked off our current series of questions. Here’s the story of how we got to where we are today.
Back in the 20th century, black holes were much more of a theoretical prediction than they were an observational puzzle. Although the idea of a black hole dates all the way back to the 18th century — originating with the predictions of John Michell from Newtonian considerations — they came into the mainstream with Karl Schwarzschild’s exact solution to Einstein’s general relativity in 1916. Schwarzschild showed that if you start with a distribution of matter at rest in general relativity and you just release those masses and let them gravitate, they’ll collapse to form a non-rotating black hole of a given mass, with an event horizon at a particular distance, R, known as the Schwarzschild radius.
Any mass that crosses over the event horizon adds to the mass of the black hole, and early theoretical work was able to calculate the maximum rate at which black holes could grow by accreting matter. From the core-collapse of massive stars, which occur in supernova events, initial black holes could be formed. And then, over time, these “seed” black holes would gain mass, grow by accretions and mergers, and would sink to the centers of galaxies. There, they could reside as supermassive black holes. Both the stellar mass black holes (from dying, massive stars) and supermassive black holes (at the centers of massive galaxies) would then be expected to exist.
It wasn’t until the 1950s and 1960s, however, that we began to detect objects that served as candidate black holes. Initially, the observations that suggested black holes physically existed came far away, in terms of wavelength, from the light we’re used to seeing: X-ray light and radio light. In the X-ray, the very bright source known as Cygnus X-1 was discovered in 1964, during a rocket flight. When black holes have another star orbiting them, particularly if it’s a very massive star, the extremely dense black hole can siphon mass off of its companion, producing a system known as an X-ray binary, where the black hole emits X-rays from the heated, accelerated matter that accumulates around it. Until the development of gravitational wave astronomy, this was the main method for detecting stellar mass black holes.
On the other end of both the electromagnetic and mass spectrums, far into the radio and corresponding to incredibly large masses, quasars began to be observed. Originally standing for QUAsi-StellAr Radio Sources (QSRS, or quasars), these objects are now known to be located at the centers of galaxies, even though, back when they were first discovered, the host galaxy itself was normally too faint to be seen. Now, we know these to be active, supermassive black holes, surrounded by accretion disks and emitting powerful jets of matter that can extend for millions or even tens of millions of light-years in length.
This was the standard picture at the turn of the century:
- we had a theoretical maximum limit for how big black holes could get,
- we knew of stellar mass black holes and had a theory of how they formed (from core-collapse supernovae),
- and we knew about supermassive black holes at the centers of galaxies, including some very active ones like active galactic nuclei and quasars, normally located very far away.
Sure, nearby, we also had the occasional active galaxy, but since the Universe was denser, experienced more major mergers between galaxies, and formed stars more rapidly in the distant past, it made sense that these quasars would tend to be more likely to be seen early on in cosmic history.
Today, we assumed, the black holes are still there, but they’re more difficult to observe, since they don’t emit those same high levels of radiation that they do in the aftermath of events like major galactic mergers or intense star-formation events. Today’s galaxies are mostly quiet places, but the occasional active galaxy that we do see nearby often has both substantial X-ray and radio emissions, and supermassive black holes can often be inferred by the presence of their jets. Even before we were able to track the motions of individual stars around our own galactic center or directly image event horizons around M87’s or our own Milky Way’s black holes, we knew that these supermassive black holes were found at the centers of practically every galaxy: of millions or even billions of solar masses.
And then, in 2004, TON 618 came along. There’s a correlation that astronomers were putting together between various emission lines in quasars, and in a 2004 paper, they applied this correlation to the active galactic nucleus in the TON 618 system for two emission lines: the emission line from triply-ionized carbon and the cyan-colored emission line in neutral hydrogen. They looked at more than 30 different systems, took spectroscopy, and refuted older notions that metallicity was important in determining luminosity or black hole mass. And when they looked at the implications for TON 618, they wound up with a shocking result: it implied TON 618 had a mass of 66 billion solar masses.
This was very bizarre for two reasons.
- There was a theoretical limit, as we’ve already discussed, for how massive a black hole could have grown to in the time since the Universe was birthed some 13.8 billion years ago. That limit — for luminous, accreting black holes — is around 50 billion solar masses. That’s lower in mass than TON 618 was measured to be.
- And there was the fact that the TON 618 system is not a modern-day system in the nearby Universe, but rather exists far away: its light is redshifted to be more than three times as long as its rest-frame wavelength would be. Located at a redshift of z = 2.219, that implies its light comes to us from a time when the Universe was merely 3 billion years old, or from some 10.8 billion years ago.
If you’ve been paying attention to the more recent results from the James Webb Space Telescopes (JWST), the question you’ll want to ask might sound familiar: how did this black hole form, and get so big, so fast, so early in cosmic history? After all, JWST has brought us hundreds of a new class of object: little red dots, which are extremely luminous, compact, distant sources of light that seem to be extremely massive supermassive black holes. Coming from far earlier in the Universe than TON 618 did and while still being considerably massive, at hundreds of millions or even billions of solar masses, they also compel us to challenge the standard scenario for how black holes form, grow up, and become so massive so early.
The key piece of evidence that changed the story was the discovery that there are black holes out there that are just as massive as all of the stars within their host galaxies combined. In the late-time Universe, including for TON 618, even the heaviest supermassive black holes only max out at about 0.1% (about 1-part-in-1000) of the total “mass in the form of stars” within a galaxy. But the presence of black holes with 1%, 10%, or even 100% of the mass of all the stars combined in their host galaxy gives us a new puzzle: overmassive black holes. It tells us that the story we initially told ourselves — that black holes originate from the deaths of massive stars — can’t be the full story.
When stars form, they come in a wide variety of masses, and it’s the most massive (and also, the rarest) stars that die first, and become stellar mass black holes. Those black holes can then grow, merge, and sink to the centers of galaxies. But if that was how supermassive black holes formed, then you’d never have a supermassive black hole that was equal to or greater than 100% of the mass of stars within a galaxy.
The fact that we find these overmassive black holes tells us that there must have been a different mechanism in place to create large, heavy, high-mass seed for the supermassive black holes that exist today. If you can start with a much more massive seed — not of tens or even hundreds of solar masses but of tens of thousands of solar masses or more — then it becomes a lot easier to have a black hole grow up to become massive early on in cosmic history, and to even lead to higher mass supermassive black holes at late times. Although we haven’t fully solved this particular mystery of where the earliest seeds for black holes come from, leading candidates include:
- converging cold streams of gas that lead to direct collapse black holes (approximately of 30,000-100,000 solar masses),
- enormous, giant, early protostars stars that get hot enough in their cores to convert photons to particle-antiparticle pairs, leading to the direct collapse of the entire protostar (potentially up to 1 million solar masses),
- or even primordial black holes, produced by some exotic mechanism, raising the possibility that black holes, including potentially massive black holes, existed long before even the first stars themselves did.
Theoretically, it’s also important to recognize that the traditional limit of “50 billion solar masses” is only a limit under two assumptions: both of which are known not to be true. First, that estimate assumes that black holes are spherically symmetric and not rotating, whereas the evidence we’ve collected instead suggests that realistic black holes rotate at close to the theoretical limit: over 50% and up to 95% of the speed of light, meaning that they depart from perfect sphericity in terms of their event horizons quite significantly. This, on its own, ups the theoretical limit that a black hole can achieve from 50 billion to 270 billion solar masses: well above anything that’s ever been observed.
Second, the estimate for maximum mass is based on a scenario where we have the continuous accretion of matter at a maximal, sustainable rate onto a black hole: the Eddington rate. But we’ve long known — for several decades — that super-Eddington accretion is very common, often at a rate that’s 10 times the theoretical limit. These “bursts” of high levels of accretion may not be sustainable over long periods of time, but they can lead to black holes growing more massive, more quickly than you’d expect under the slow, continuous, steady accretion scenario.
Combined, these two elements enable more massive black holes to exist earlier in cosmic history than in the standard, original “theoretical limit” scenario.
Then, on the observational front, it turns out that there are plenty of quite massive black holes out there, but they all have large uncertainties attached to them, and that TON 618, in particular, is probably not as massive as we once thought. The same triply-ionized carbon line that led to the mass estimate of 66 billion solar masses for TON 618’s black hole in 2004 was found to have a bias in it, and that when you account for that bias, the mass estimate drops to “only” about 40-41 billion solar masses: a substantial decrease.
Although there are 10 other black hole systems, in addition to TON 618, that have an estimated mass of at least 30 billion solar masses, the only three that are larger:
all have enormous uncertainties attached to them, are much closer than TON 618 is, and were estimated with crude methods by comparison with TON 618. A black hole located 5 billion light-years away with a mass of 36 billion solar masses was just confirmed last year, in 2025, further suggesting that TON 618 may be a stupendous object, but not a unique one. There are others like it, and they might simply be difficult to find at late times, because quasar-like activity is rarer here in the late-time Universe.
In summary, if you want to know how an ancient black hole, like TON 618, came to be so massive at such early times in cosmic history, it’s vital that we put all of these pieces of information together, and that we do it correctly. When we synthesize together everything we just went through, we learn that:
- TON 618 is very massive, but it’s far less massive — about 25 billion solar masses less massive — than we originally thought when it was first well-measured back in 2004.
- The original theoretical assumptions that went into the theoretical mass limit for a supermassive black hole, that black holes are spherical and accrete matter at a continuous, maximal rate, have both been demonstrated to be wrong, raising the theoretical limit from 50 to 270 billion solar masses, and enabling higher mass black holes at earlier times.
- The original scenario for how supermassive black holes would form and grow, from stellar mass black hole seeds, has been ruled out, instead suggesting one of several scenarios that begin with significantly larger black hole seeds at extremely early times.
- And the reason that we don’t see ultramassive black holes in the nearby Universe is because we primarily detect supermassive black hole through their central activity: when they feed on matter. Nearby galaxies are far less active, in general, than ultra-distant galaxies from earlier on in cosmic history, and so just like you don’t hear a tiger that doesn’t roar, we don’t see black holes of similar sizes in the nearby Universe because of their lack of activity.
There’s been a lot to learn about black holes in recent years, and hopefully there will be even more to come in the years ahead. While it’s safe to say that cosmology isn’t “broken” as far as these very massive black holes are concerned, there are certainly some significant, unanswered questions that we’d love to gain additional insight into. With new data and new observatories on the horizon, perhaps we’ll finally develop a clear picture about the beginning-to-end pathways of growing a supermassive black hole.
Send in your Ask Ethan questions to startswithabang at gmail dot com!
This article Ask Ethan: Can we solve the mysteries of the largest black holes? is featured on Big Think.