Juan Maldacena

The Big Bang Machine

A Long Island particle smasher re-creates the moment of creation.

by Tim Folger

published online February 27, 2007

"Here is where the action takes place. This is where we effectively try to turn the clock back 14 billion years. Right above your head, about 13½ feet in the air."

Looking up, I try to imagine the events Tim Hallman is describing—atoms of gold colliding at 99.99 percent the speed of light; temperatures instantly soaring to 1 trillion degrees, 150,000 times hotter than the core of the sun. Then I try to picture a minuscule five-dimensional black hole, which, depending on your point of view, may or may not have formed at that same spot over my head. It's all a little much for an imagination that sometimes struggles with the plot of Battlestar Galactica.

Particles unleashed by the high-energy
collisions at the RHIC collider offer a
peek at the freekish far end of physics.

I'm standing in a Battlestar-scale room at Brookhaven National Laboratory in Upton, New York, where Hallman and about 1,200 other physicists labor away as latter-day demiurges. Here in the middle of Long Island, they are re-creating the opening microseconds of the universe's existence, the time when the first particles of matter—of everything—appeared.

The building we're in straddles a small segment of the Relativistic Heavy Ion Collider, or RHIC (pronounced "Rick"), an ultrapowerful particle accelerator with a 2.4-mile circumference. For nine months of the year, a 1,200-ton detector as big as a house fills most of the room. But technicians have hauled the detector to an adjoining hangar-size area for maintenance, leaving me and Hallman free to amble just below the spot where a new form of matter exploded into being during the accelerator's recent runs. New, that is, in that it hasn't existed since the very beginning of time, or by the transcendently precise reckoning of physicists, since 10 millionths of a second after the Big Bang, 13.7 billion years ago.

That was the last time particles called quarks and gluons—the most fundamental constituents of matter—roamed free in the cosmos, and it was a brief run. After just a hundred-millionth of a second, all the universe's quarks combined in triplets—held together by gluons—to form protons and neutrons. They have been locked inside the hearts of atoms ever since, until RHIC set them loose once again.

When the gold nuclei collide in the accelerator, they explode in a fireball just a trillionth of an inch wide. Inside that nanoscale fireball, temperatures exceed a trillion degrees, mimicking conditions in the immediate aftermath of the Big Bang. The nuclei literally melt back into their constituent quarks and gluons. Then, 50 trillionths of a trillionth of a second later, the fireball cools, just as the infant universe did as it expanded, and the quarks and gluons merge once again to form protons and neutrons.

With these experiments, Hallman and his Brookhaven colleagues are discovering something extraordinary about the early universe. The quarks and gluons that coursed through the newborn cosmos—and considerably more recently, through RHIC—took the form not of a gas, as physicists expected, but of a liquid. For a few instants, a sloshing soup of quarks and gluons filled the universe.

"I like to say that our theory of the early universe is now all wet," says Bill Zajc, a physicist at Columbia University and the leader of one of the experimental teams at RHIC.

In the circular tunnel at RHIC (above), two rings of superconducting magnets accerlate heavy ions, like gold, to high speeds in opposite directions. At six points along the way, the rings cross, causing collisions between the ions. Gigantic detectors (below) track the spray of particles released in the smashups.

He might have added that the theory is full of holes, little black ones from the fifth dimension, because it turns out that in a strange mathematical sense, the quarks and gluons at RHIC are equivalent to microscopic black holes in a higher-
dimensional space. Understanding just why that is so involves navigating a labyrinth of strange, heady, and heretofore seemingly unrelated theories of physics. In addition to challenging the conventional model of how the universe behaved in its earliest instants, the RHIC data also provide the first empirical support for a theory so enthralling it once had physicists dancing at a major conference. Moreover, the accelerator's results hint that string theory—the much-
ballyhooed "theory of everything," which has lately come under attack as being little more than a fanciful, if elegant, set of equations—may have something to say about how the universe works after all.

Before the physicists at Brookhaven could begin their pursuit of quarks, gluons, and hyperdimensional holes in space-time, they first had to prove that they wouldn't destroy the planet in the process. The doomsday risk never really existed, but making that clear to a worried public occupied the time of some of the world's leading physicists (see "Could a Man-Made Black Hole Swallow Long Island?" opposite page).

Once the doubts about Earth's safety had been laid to rest, the physicists at Brookhaven fired up their $500 million accelerator for the first time, in the summer of 2000. For Nick Samios, it was the culmination of two decades of work. Samios, who is now director of the RIKEN-BNL Research Center, headed Brookhaven from 1982 until 1997 and was the driving force behind the effort to build RHIC. "I'll tell you a story," he says over lunch at Brookhaven's staff cafeteria, leaning forward. The story's principals are Stalin; his chief of secret police, Lavrenty Beria; and Igor Vasilyevich Kurchatov, a leading Soviet nuclear physicist. "Stalin and Beria are discussing the Soviet Union's first atomic-bomb test. 'Who gets the award if the test is a success?' Beria asks Stalin. 'Kurchatov.' So then Beria asks, 'Who gets shot if it doesn't work?' 'Kurchatov.' I feel like Kurchatov. Anyone else could disassociate themselves from the project. I couldn't."

One gamble Samios and his colleagues made 20 years ago was to trust in Moore's law, first formulated in 1965 by one of Intel's founders, which holds that computing power doubles roughly every 18 months. The type of accelerator Samios wanted to design would generate a petabyte of data—a million gigabytes—during each run, a rate that would fill the hard drive of one of today's typical desktop PCs every few minutes. In 1985 there were no computers that could handle anything close to that. But the "if we build it, the computers will be there" strategy paid off, and Samios's dream of a Big Bang machine became a reality.

To re-create the immediate aftermath of the Big Bang, RHIC reaches higher energies than any other collider in the world. Unlike most accelerators, which smash together simple particles like individual protons, RHIC accelerates clusters of hundreds of gold atoms—with 79 protons and neutrons in each gold nucleus—to 99.99 percent the speed of light. In the resulting multiatom collisions, a melee of tens of thousands of quarks and gluons is released. They in turn form thousands of ordinary particles that can be tracked and identified.

"The physics at RHIC is complicated," Samios says with a touch of understatement. "Two big nuclei are hitting each other. Physicists are used to calculating a proton hitting a proton. We're hitting 200 nucleons with 200 nucleons [a nucleon is a proton or neutron]. With each collision we get thousands of particle tracks coming out. We had to build detectors that could count all of them. People weren't used to that. They were used to counting 50 in a collision.

"We hoped that RHIC would make great discoveries. We hoped that we'd break nuclei into quarks—the early universe was quarks and gluons, and then it cooled off and you got protons and us. We've done that. The question is–Is there something new going on? And the answer is yes."

Until the first gold atoms started making their 13-microsecond laps around RHIC's 2.4-mile-long perimeter, physicists thought they had a pretty good idea of what to expect from the collisions. The gold nuclei were supposed to shatter and form a hot gas, or plasma, of quarks and gluons. For physicists, watching the collisions in RHIC would be like watching the Big Bang unfolding before their eyes, but running in reverse—instead of a seething cloud of gluons and quarks settling down to form protons and neutrons, they would see the protons and neutrons burst open in sprays of quarks and gluons.

The universe has been around now for about 400 quadrillion seconds, but for physicists like Tom Kirk, the former associate lab director for high-energy and nuclear physics at Brookhaven, the first millionth of a second was more intriguing than any that followed.

"You may remember Steven Weinberg's book The First Three Minutes," says Kirk, referring to a classic account of the physics of the early universe. "Steve said after those first three minutes, the rest of the story is boring. Well, we could say after that first microsecond, everything else was pretty boring."

Kirk smiles and raises his eyebrows slightly, gauging whether I'm sympathetic to his remark, made, perhaps, only half in jest. When that first microsecond of eternity ended, the remainder of cosmic history unfolded with stolid inevitability. Once quarks finished clumping together as protons and neutrons, it was only a matter of time—and gravity—before the first simple atoms gathered in vast clouds to form stars and galaxies, which eventually begot us. (For an elaboration of your personal relationship with quarks, see "The Big Bang Within You," below.)

RHIC was designed to observe directly, for the first time, how quarks behave when freed from their nuclear prisons. The initial results, announced in 2005, stunned physicists everywhere. The particles released by the high-speed smashups were not bounding around freely the way atoms in a gas do but moving smoothly and collectively like a liquid, responding as a connected whole to changes in pressure within the fireball. The RHIC physicists describe their creation as a near "perfect" fluid, one that has extremely low internal friction, or viscosity. By the standards physicists use, the quarks and gluons make a much better liquid than water.

Since the similarity of quarks and gluons to water is not readily apparent to me, I take the subway to Columbia University to meet Zajc, the leader of one of RHIC's main experimental groups, hoping he'll enlighten me. "So how does one calculate the viscosity of those quarks and gluons?" he asks rhetorically. I sit silently, clueless, hiding my befuddlement behind a vigorous show of note taking. "It turns out there's a connection here to black-hole physics." That connection could be the first, long-awaited sign that string theory—which is in desperate need of evidence, any evidence, to support its ambitious claims to truth—is on the right track. The implausible-sounding connection between droplets of quarks and black holes may also vindicate a theory that once had 200 of the world's leading theorists jubilantly dancing the macarena.

Ehhhh! Maldacena!

M-theory is finished,

Juan has great repute.

The black hole we have mastered,

QCD we can compute.

Too bad the glueball spectrum

Is still in some dispute.

Ehhhh! Maldacena!

So goes Jeffrey Harvey's über-geek version of the once-ubiquitous 1996 hit. Harvey, a theoretical physicist at the University of Chicago, wrote the lyrics to honor Juan Maldacena, a young Argentine string theorist now at the Institute for Advanced Study in Princeton.

It was the summer of 1998, and Maldacena had just published a paper that to physicists bordered on the miraculous. He proposed an unexpected link between two ostensibly different theories of fundamental physics: string theory and quantum chromodynamics. String theory purports to describe all the elementary components of matter and energy not as particles but as vanishingly small vibrating strings. Photons, protons, and all the other particles are, according to this theory, just different "pitches" of vibration of these strings. If it is right, string theory would unify gravity and quantum mechanics in a single overarching framework—a goal that physicists have pursued for more than half a century. The problem is that there is no shred of experimental evidence that string theory is correct; all the arguments in its favor have been made entirely on the basis of its sophisticated mathematical structure. Direct experimental tests of string theory have thus far proved impossible, in part because strings are predicted to be so small that no conceivable particle accelerator could ever reach the energies needed to produce them.

Quantum chromodynamics, or QCD, on the other hand, is backed by decades of experiments. It describes the interactions of quarks and gluons. (Quarks come in three "colors," analogous to electric charge; hence the "chromo" in chromodynamics.) Unfortunately, unlike string theory, QCD says nothing about gravity, so physicists know they need a broader, more complete theory if they want to explain all of physics. Moreover, the equations of QCD are notoriously difficult to work with.

Enter Juan Maldacena. He developed a theory nearly identical to standard QCD, with the major difference being that in his version quarks come in many different colors rather than the usual three. Even though his theory does not fully apply to the universe we know—it doesn't correctly describe quarks—physicists frequently use such so-called toy models to get a handle on otherwise impossibly difficult problems. And Maldacena's theory had a remarkable feature, the one that inspired his colleagues to hit the dance floor. He proved that his pseudo-QCD and string theory are not in fact different theories at all; mathematically, they are entirely equivalent.

The site of the RHIC collider.

Maldacena's realization raised an enormous question: If string theory and his slightly altered take on QCD are essentially one and the same beast, does that mean there is a way to connect string theory to the physics of the real world? For the next few years, Maldacena's tour de force remained largely a plaything for theorists, who almost immediately found intriguing ways to use it. Most important, his theory simplified the grueling calculations of QCD by offering a way to translate certain QCD problems into the more tractable mathematics of string theory. "Things that are hard to calculate in QCD are easy in string theory, and vice versa," says Horatiu Nastase, a string theorist at Brown University.

It was this power to shift problems from the QCD perspective to a string theory view that first led some physicists to see a link between the quarks and gluons at RHIC and the equations describing a black hole. Dam Thanh Son, a physicist at the University of Washington in Seattle, was one of them. I called him to ask about what seems, on the face of it, an extraordinarily unlikely comparison. What could quarks and gluons possibly have in common with nature's ultimate trash compactors—ultradense concentrations of matter whose gravitational field is so powerful it curves space-time around itself, trapping anything that crosses its surface?

Son insists that black holes, quarks, and gluons really do have a big thing in common: They can be described by equations that govern the behavior of liquids. Then he explains that black holes—and quarks and gluons—are really no stranger than a cup of water.

"If you have a cup of water, and you disturb the water—say, you drop a pebble into it—the disturbance will not last forever. The water will come to rest. If you take a cup of honey, the motion ceases more quickly than in water; the more viscous the fluid, the quicker the perturbation of the system decays with time."

When something falls into a black hole, Son says, the surface of the black hole is disturbed, just like the water in a cup. "The black hole will wiggle for some time and come to rest. In these two processes"—disturbances in black holes and in water—"there is a connection at the mathematical level. The equation that describes the evolution of the stirring of water in a cup is similar in form to the equations that describe the evolution of the surface of a black hole. When I deform a black hole, it goes back and forth and then comes to rest. To describe that I use equations that are similar to equations used for any fluid."

As word spread that RHIC had created a quark-gluon fluid, Son and a number of other theorists began to wonder if they could use Maldacena's sleight of hand and substitute the equations of a black hole for the ones normally applied to quarks and gluons. The switch would make calculating the properties of the primordial particle soup much easier. Compared with a trillion-degree ruck of quarks and gluons, black holes are simple objects. (Which is why the lyrics to the Maldacena macarena go: "The black hole we have mastered, QCD we can compute.")

One property of the quarks and gluons that Son and his colleagues wanted to calculate was viscosity. Using a black-hole model, they predicted that quarks and gluons should have almost zero viscosity. When experimentalists at RHIC finally crunched through all their data, they confirmed that the quark-gluon fluid indeed had a low viscosity, at or near the theoretical minimum value predicted by the five-
dimensional black-hole model.

"Talk about a shot out of the blue," Zajc says. "Who would have thunk it? It is the most fascinating thing I've been involved with, to see this completely unexpected connection emerge and start having an impact on our field."

So does this success bolster the idea that string theory is the right way to unify all of physics?

"Absolutely," says Horatiu Nastase of Brown, who has also sought to understand RHIC's results in terms of a black hole. "At least that's my interpretation and the interpretation of other people. My understanding is that one is experimentally testing, in this indirect way, string theory."

Zajc and many other physicists aren't so sure. "I've thought an awful lot about this," he says. "But I'm not ready yet to claim that this validates string theory. Even the string theorists will tell you the viscosity result depends only on ordinary quantum mechanics—it's just that string theory gives you a snazzy way to calculate it."

In any event, the black hole under consideration is not the sort that could swallow Long Island. It's an entirely different animal. According to string theory, the universe may contain as many as 10 dimensions. Most of them are hidden, curled up on scales so small that we cannot sense or even detect them. The black hole in Son's calculation dwells in a theoretical world of five dimensions, where the effects of gravity drop so precipitously with increasing distance that a five-dimensional black hole poses no threat—if it even exists at all. Some physicists consider the five-dimensional black hole to be a mathematical convenience, a way to tackle a complex physical system. Others are open to a far more radical interpretation, however.

"What we think of as atomic nuclei, quarks, and gluons may really be objects that are projections, in a sense, on a screen," says Miklos Gyulassy of Columbia, sounding more like Plato philosophizing than like the theoretical physicist that he is. "We are on the screen. It looks to us like there are photons and these other particles, but they might really be manifestations, projections, from a higher-dimensional space, of objects that are more conveniently described in our world by saying, 'There is a photon,' or 'There is a gluon.' So the very hot quarks and gluons at RHIC may really be a hologram of some nasty black hole somewhere."

All of these issues and more will continue to be studied at RHIC and at an even more powerful accelerator nearing completion in Switzerland. The Large Hadron Collider, as the new accelerator is called, will be almost 17 miles in circumference and will reach energies 27 times higher than RHIC's.

"One question that screams out to be answered is whether we'll see the same sort of perfect fluid that we see at RHIC," Zajc says, "or whether we'll see something like an ideal gas where the quarks and gluons are essentially free. I think it will continue to be a perfect fluid, or very nearly so. But we've been surprised before in this field."

As to whether Maldacena's ideas will further strengthen string theory or prove a theoretical dead end is anyone's guess. The data, says Zajc, are simply too raw.

"This is what new discoveries look like from the inside," he says. "If you'll allow me to mix metaphors, it's sort of a Mixmaster of swirling ideas that may gradually be distilled into something elegant and nice. But at the moment we're watching the sausage-making process."

http://discovermagazine.com/2007/feb/cover/article_view?b_start:int=0&-C=