For decades researchers have tried to wrest testable predictions from string theory, the leading candidate for a more fundamental understanding of the universe. Now physicists say they have used one of the most sophisticated pieces of string theory to predict properties of the ultradense matter created in an atom smasher in Long Island, N.Y. If confirmed, however, the prediction would not offer evidence for string theory, which requires the existence of extra dimensions of space full of higher-dimensional stringlike objects and other widgets. Instead, it would establish that some of string theory's mathematics could be used to study the forces at work inside an atom's nucleus.

The Relativistic Heavy Ion Collider (RHIC) at the Brookhaven National Laboratory smashes together the nuclei of gold atoms to create hot, ultradense sprays of particles, the likes of which have not occurred since the first few microseconds of the big bang—the intense burst of energy that kicked off the universe. These mini-explosions temporarily liberate quarks, the particles normally trapped inside protons and neutrons, along with gluons, the particles that hold quarks together. Researchers have dubbed this state of matter the quark-gluon plasma.

But studying even its general properties is tricky. "You're trying to get both insight and quantitative prediction, and we got both," says nuclear physicist Krishna Rajagopal of the Massachusetts Institute of Technology.

In a paper scheduled to appear in Physical Review Letters, Rajagopal and colleagues report they have calculated the rate at which particles called J/psi mesons would form in the heat of the quark-gluon plasma, which dissolves them in an effect called screening. They find that the screening would vary with J/psi velocity in a particular way. "We're making a qualitative prediction that as you turn up the [velocity] you should get more effective screening," he says, resulting in fewer particles at high velocities. "Time will tell whether at a quantitative level we do as well."

Normally, researchers use a theory called quantum chromodynamics (QCD) to calculate forces between quarks. To do so, they have to make estimates that are only accurate when the force is weak, but the force between quarks in the plasma can become rather strong. In fact, physicists now believe the plasma, which was first conceived as a kind of gas, is more like a liquid.

Rajagopal and his cohorts took advantage of a 1997 discovery by particle physicist Juan Maldacena at the Institute for Advanced Study in Princeton, N.J., that a form of string theory in four dimensions of space—one more than our own—is equivalent to a QCD-like theory in three space dimensions. When forces become strong in this QCD-like theory, they become weak in string theory, offering an easier way to do calculations related to quarks.

Although the stringy half of Maldacena's coin includes gravity and other forces, the QCD-like half contains only gluons and similar particles that do not exist in the real world. "It's like saying you are trying to study water, but instead you are studying alcohol," Maldacena says. "We certainly know it's not the correct theory, but maybe it behaves in the same way."

Trying to fit the QCD-like theory to reality makes the results only semi-precise, Rajagopal says. But last year, the same group used this piece of string theory to estimate the degree to which quarks rocketing through the plasma would be jiggled from side to side. "What we get is in better agreement with the data than what you get with conventional methods," Rajagopal says.

Other groups have used Maldacena's result to more directly estimate the slowdown of quarks in the plasma and to gauge the plasma's viscosity.

"The idea that there might be some connection between the string theory mathematics in higher dimensions and real-world phenomena being studied at RHIC—that to me is just fascinating," says physicist William Zajc of Columbia University, who is working on an experiment at RHIC to detect J/psi particles.

Such attempts are exciting, agrees Ulrich Heinz, a theoretical physicist at Ohio State University, but he says the QCD-like theory is still too far removed from reality to be convincing. "Even if any of the numbers worked out by accident, I don't think it would validate the approach," he says. "If they predict the color of an apple, and somebody looks at a pear and finds it has the same color, would you say that the prediction was correct?"

Zajc says RHIC should be able to gather enough data in the next two years to evaluate the J/psi prediction. "It would require a fairly spectacular numerical agreement before you would argue it establishes the superiority of these methods" over more conventional calculations, he says. Finding that agreement might be tough, he notes, because J/psi screening is not measured directly but has to be inferred, leaving room for argument about its true value.