Black Holes

Priya Natarajan is in in the news for discovering that black holes have an upper limit of growth and dont just keep growing out of control.

By Charles PetitOctober 25th, 2008; Vol.174 #9 (p. 18) Text Size A black hole can consume anything in its path. These monsters can become huge — but perhaps only so huge.BLACK HOLE TAXONOMYENLARGE | Click to enlarge for a primer on black hole taxonomy.Design: J. Korenblat

If asked to name stupendously amazing things in space, most people would probably pick black holes. These evil-tinged clowns of the universe are definite wows. Insatiable is their middle name.


Grand and merciless, voracious and monstrous, pure appetite and deep mystery. The biggest fatten themselves in galaxy cores mainly via a seemingly limitless hunger for a main source of sustenance: fat, circular wads of gas that gather around the black holes and are sometimes given a name to delight any glutton, Polish doughnuts. Black holes cloak their innards behind an “event horizon,” from inside which no message can be sent (which explains the one-liner physics joke: “Two protons walk into a black hole”).

What a parade of jaw-droppers that is. Well listen up, this just in: It looks like there is a limit to the superlatives. Black holes can’t eat everything. If a new analysis from a Yale astronomer is correct, even black holes run out of steam, and at a fairly precise point. The biggest black holes may reach only a few tens of billions of times the mass of the sun.


To be sure, that’s huge. Most galaxies harbor central black holes of a few million solar masses (about 4 million for the Milky Way). Fifty million light-years away in Virgo, the giant elliptical galaxy M87 is believed to harbor one having about 3 billion solar masses. The record heft for a suspected black hole, 3.5 billion light-years away and part of a double–black-hole system with a partner’s orbit that reveals its mass with some precision, is 18 billion solar masses.

Any possible cap on the size of these monsters occupying galactic centers shouldn’t diminish the place of black holes in popular imagination. And for astronomers, the newly proposed mass limit illustrates how the status of black holes, as both scientific challenge and principal player in the universe’s appearance, is on the rise.


Astrophysicists and cosmologists thought they had black holes pretty well pegged about 10 years ago. Black holes eat, they grow and they can sure produce a bright light from X-ray to radio wavelengths while on a binge. Their quasar-pumping conversion of matter to outward-beamed energy as they consume gas, dust and the occasional unlucky star is believed to reach about 40 percent efficiency. It’s not only E=mc2 at which black holes excel. They also provide wonderful playgrounds for a panoply of other Einsteinian gymnastics. They bend time, warp space and, along their borders, they spawn a fizz of evanescent virtual particles popping in and out of space’s fabric.


But all in all, to many pros interested in the big picture, black holes have been seen as intriguing and flashy character actors, bit players in the grand story of galaxy evolution and in the overall distribution of ordinary matter in the universe. Even supermassive black holes’ gravity, after all, dominates only a few parsecs radius in the crowded hearts of galaxies many thousands of parsecs across.

A consuming influence


A budding new paradigm is that black holes—in a dance of mutual self-regulation—may influence almost everything about galactic origins, growth, form and ultimate fates. They are not just the overstuffed kernels in the middle of galaxies. For reasons not fully understood, it appears that the sizes of central black holes and the masses of their galaxies, especially the central bulges, are almost perfectly in step.

The relation has become clear only since the late 1990s. Even the halo mass of dark matter—the mysterious invisible stuff that seems to make up more than 80 percent of all matter—around galaxies seems correlated with the size of supermassive black holes in galactic centers. That is a surprise. And when the black holes stop growing, galaxies themselves appear to stop evolving. “Now, we think we cannot understand galaxies without understanding black holes,” says Abraham “Avi” Loeb, director of the Institute for Theory and Computation at the Harvard-Smithsonian Center for Astrophysics in Cambridge, Mass.


The proposed limit on black hole mass comes from Yale University cosmologist Priyamvada Natarajan and Chilean astronomer Ezequiel Treister of the European Southern Observatory. Their paper, to appear in the Monthly Notices of the Royal Astronomical Society, was posted online in August.

Declaring an upper mass limit to black holes is notable, even were such a limit not part of a bigger relationship to overall galactic physics. For one thing, it would give bounds to the specs of the black hole bestiary.


Ignoring hypothetical mini–black-holes of subatomic size that might briefly form under exotic conditions, astronomical black hole taxonomy would go like this, from smallest to largest: Substellar-mass or primordial black holes, still unproven, proposed by physicist Stephen Hawking to have formed in the dense soup of particles shortly after the Big Bang. A stellar-mass black hole is what remains after some supernovas. Intermediate-mass black holes, conjectured to form from runaway mergers of stars into dense clusters that undergo gravitational collapse, would be 100 to a million times as massive as the sun. Next up are supermassive black holes, which can grow as gas accretes into galactic centers and when galaxies hosting central black holes merge. The Milky Way’s central black hole, at 4 million solar masses, is supermassive. And at the top of the scale are ultramassive black holes, the name Natarajan gives those with 10 billion to a few tens of billion solar masses.

Natarajan, a native of New Delhi, went in 1997 from MIT to the University of Cambridge in England as a graduate student during a transition time in black hole and cosmology studies. Experts were already suspecting that extremely massive galactic black holes in the current universe are not as common as one would expect. The fast growth of numerous quasars—galactic core black holes glowing fiercely as matter falls into them—seen at great distances and as they were long ago, implied that many were bound to reach masses exceeding 10 billion suns. There is no way to see how those black holes turned out at the end of their quasar days, but astronomers can check nearby galaxies that presumably went through similar youths. And the current universe seems to have a shortage of the fatties that it appears should have grown from earlier epochs.


A basic picture of black hole growth had been worked out in the 1970s and 1980s by Bohdan Paczynski of Warsaw University (and later Princeton) and others. When Paczynski died in 2007, his obituaries all mentioned Polish doughnuts. That was his name for the fat rings of gas that ought to form in any gas-rich region around a large black hole. These torus-shaped rings would feed a steady stream of matter into a hot, brilliantly glowing flat disk of plasma spiraling down—the inner accretion disk. Most of the matter spirals down to its doom, while some gets ejected as powerful polar jets—gouts of radiation.

The result can be a quasar that shines from a region smaller than Earth’s orbit of the sun with a brilliance 100 times that of the rest of the quasar’s host galaxy. To achieve such power, the quasar must be bumping up against a barrier called the Eddington Limit. The limit’s namesake, English astronomer Arthur Stanley Eddington, in the early 20th century worked out how brightly a star can shine before its radiation pressure starts blowing its outer layers into space. Turned around and applied to black holes, that limit is the brightness at which a black hole’s accretion disk is so great that it stops more gas from falling in. And to reach that, a quasar of a million solar masses must nearly triple its mass every 10 million to 100 million years. By the time it reaches a billion solar masses, it consumes 20 suns’ worth of gas every year.


A quasar’s brightness is related to how much matter the black hole is consuming. When matter stops falling in, the light goes out. Each quasar shines for only a few hundred million years. But there was no obvious reason why galaxies should run short of gas to feed into Polish doughnuts that quickly.

Stunting growth


Working in a Cambridge group headed by Great Britain’s Astronomer Royal, Martin Rees, Natarajan first decided 10 years ago to calculate how a supermassive black hole might shut off its own food supply and stop growing. Rees, in partnership with University of Oxford cosmologist Joseph Silk, at about the same time worked out one plausible way. “As the black hole grows, we felt it would expel a lot of energy in a jet. It sort of fans out and clears a bubble in surrounding gas,” Silk says.

For her thesis, Natarajan worked out another plausible way: A quasar, fueled by a growing, supermassive black hole, reaches a point at which its radiation not only slows the infall of more gas, but also turns the gas around and clears out a large region around itself—leaving a nearly gas-free or “dry” galaxy. This, she estimated, would occur as the black hole reached about 10 billion solar masses.


With this theoretical exercise complete, Natarajan a few years ago tackled another aspect of galactic behavior that would eventually lead her back to how black holes might stunt their own growth. She worked with Marta Volonteri—a former fellow Cambridge postdoc now at the University of Michigan in Ann Arbor—who had developed a model for how the mysterious dark matter would behave early in the universe. Specifically, the astronomers wanted to see how dark matter’s clumping under gravity shapes evolution of galaxies that form from the regular matter accompanying them.

Observations with space telescopes had shown that quasars started to pop off when the universe was less than a billion years old, and at immense power. Small black holes cannot do the job. That takes black holes of around a billion solar masses.


Earlier theorists had thought the seeds of galactic black holes were sown by the collapse of the first, immense “Generation III” stars, but those looked too puny to grow fast enough to get quasars going so soon. The two women joined a cadre of cosmologists imagining a direct-collapse model. In it, the first galaxies would form mostly from hydrogen and early stars within blobs of cold dark matter. And in these galaxies’ dense centers, gas would congregate so fast it would spiral directly into multimillion-mass black holes, not stopping to form stars first.

With their primordial dark matter blobs set up in their model—each with one or several galaxies and each of those equipped with sizable, often quasar-worthy black holes—the two scientists ran the process to the present time. Out came a universe with, sure enough, galaxies, galaxy clusters and black holes in the middles. But, as others have found, the model predicted more immense galaxies and more black holes of 10 billion solar masses and beyond than are actually evident in nearby (and therefore current) regions.

To be certain, Natarajan needed a more complete history of quasars over the lifetime of the universe for closer comparison with the model, so she could see better where reality and mathematical simulations had parted ways. Her coauthor of the recent paper, Chilean astronomer Treister, gathered the necessary stats from the ground-based Sloan Digital Sky Survey and from some of the most powerful new telescopes in the heavens, including the Chandra X-ray Observatory and Europe’s Integral, a gamma-ray observatory. These data informed her not only on the optically obvious quasars shining at visible wavelengths and first identified in the 1960s, but also on roughly twice as many others cloaked by the belts of dust and gas feeding them.

“This was the aha moment,” Natarajan says. Early models showing that black holes can turn off their own feeding station were combined with models of galaxy evolution and the populations of quasars and other active galactic nuclei over time. “The only way to fit the data is to physically cut off the ability of black holes to grow beyond some point, and that is at about 10 billion solar masses.”


Physically, she explains, the largest black holes reach the end of the line by heating gas not only in their own vicinity but, in a final stage of frenzied luminosity, heating gas throughout their enormous host galaxies and often among the galaxies of the clusters where they reside. Furthermore, it appears that black holes can keep the gas too hot to settle in large quantities back to the galaxy’s nucleus or to form stars through most of the galaxy’s bulk. Only in the past 10 years have other observations, in fact, revealed that the thin gas permeating massive galactic clusters is heated to tens of millions of degrees. “Nobody expected that,” says Harvard’s Loeb. “So galaxies reach the point where you don’t make stars. This must be intimately related to black hole growth and why it stops.”

Case closed? Not likely. Oxford’s Silk, one of the grand figures in contemporary cosmology, calls the paper “very nicely done, very competent,” but also says that “this is pretty speculative territory.” He continues: “She starts with a weak set of assumptions. You don’t really know how to make the first, seed galactic black holes in the first place. The first galaxies and the first halos of dark matter were not so big. How exactly did billion-mass black holes form? It is one thing to say that, if you have the right ingredients, you can make the cake. But these ingredients are not so natural, I think.”


Natarajan expresses similar concern about those original seeds. “The big question that remains is the early merging history of dark matter halos. This has opened up an absolutely new theoretical simulation to see if we can understand the formation of those black hole seeds.” New instruments may help explore that question. Some answers may come in 10 years or so when a joint NASA and European trio of widely spaced satellites, called the Laser Interferometer Space Antenna or LISA, may detect the gravitational waves from black holes forming and coalescing in distant galaxies. That could provide vital info on the origin of the seeds for eventual, supermassive black holes.

While the scaffolding of a coherent hypothesis linking galaxy evolution and massive black hole behavior is rising, it is not a monument yet. Other questions loom as well. It remains a puzzle that objects of such enormous difference in scale—gigantic galaxies and tiny (if massive) black holes in their centers—seem to move in smooth coordination of growth and evolution. Says Michigan’s Volonteri, “Yes, black hole growth has to stop at some point. Priya [Natarajan] suggests black holes stop their own growth.”

Then Volonteri adds, “Are black holes stopping the galaxies too? Or are the galaxies stopping the black holes?”








SIDEBAR: Black Hole Taxonomy

Black holes can be classified into categories based on size, which depends directly on mass. Apart from small “primordial” black holes that possibly formed in the early universe, the least massive are the size of a large city; the largest are huge enough to reach from the sun out beyond Neptune. New research suggests that there is a limit to how massive a black hole can become.

Stellar-Mass Black Holes

About 5 to 10 solar masses, formed when a massive star exhausts its fuel, central pressure falls and the core collapses to black hole density. (A shock wave blasts the rest of the star off in a supernova). The Hubble Space Telescope image above is of a supernova remnant in the constellation Cassiopeia.


SIZE: Roughly 30 kilometers across, or about 10 km longer than Manhattan.

MASS: 5 suns\


Intermediate-Mass Black Holes

About 100 to a million solar masses, conjectured to form in dense star clusters from a merger of stars into a giant mass that then undergoes runaway gravitational collapse.

SIZE: About 60,000 km across, or almost five times Earth’s diameter. If a stellar-mass black hole were the size of the period at the end of this sentence, this black hole would be about 2 feet across.


MASS: 10,000 suns

Supermassive Black Holes

From a million to a few billion solar masses, formed by accretion of gas in galactic centers and by mergers of black holes as their host galaxies collide. The Milky Way’s central black hole is in this group. The above Hubble image shows the collision of two galaxies.


SIZE: About 25 million km across, it would fit within Mercury’s orbit around the sun. If a stellar-mass black hole were period-sized, this black hole would be 250 meters across.

MASS: 4 million suns

(central black hole in the Milky Way)


Ultramassive Black Holes

Newly proposed category for black holes from 10 billion to tens of billions of solar masses. At such sizes, the event horizon diameter can reach hundreds of billions of kilometers.

SIZE: 60 billion km across, it would stretch from the sun to far past Neptune, even beyond some distant comets. If a stellar-mass black hole were a period, this black hole would stretch from Cleveland to Washington, D.C.


MASS: 10 billion suns