It would be tough to stick it to your refrigerator, but an ultra-cold gas magnetizes itself just as do metals such as iron or nickel, a team of atomic physicists reports. That cool trick shows that the messy physics within solids can be modeled with pristine gases, the researchers say. But others are skeptical that the team has actually seen what they claim.
Condensed matter physicists can tell you essentially all there is to know about how common metals carry electricity and heat. Why some of them are magnetic is a trickier question. Physicists know the basics: The electrons that flow through iron, nickel, and other magnetic materials act like little bar magnets. Below a certain temperature the electrons align so that they all point in the same direction, at least within relatively large "domains" in the crystalline material. The question is why do the electrons align themselves?
An answer was proposed in the 1930s by British theorist E. C. Stoner. It depends on a key bit of quantum mechanics called the Pauli exclusion principle, which says that no two electrons can be in exactly the same condition or "quantum state" at the same time. To see how this works, first consider a nonmagnetic metal. The electrons can be thought of as a kind of gas within the solid, with equal numbers of electrons pointing with their north poles up as down, because that would be their lowest-energy state.
Electrons repel each other, which increases the energy of the gas. Stoner argued that if the electrons repel each other hard enough, they could lower their total energy by aligning. The flipping of some of the electrons would agitate the gas and increase its "kinetic" energy a bit. But because of the exclusion principle, no two aligned electrons could be in the same place at the same time, meaning the electrons would avoid each other so that energy from the short-range repulsion would drop even more. Stoner came up with a highly simplified mathematical model that encapsulates this idea. However, no one has ever rigorously proved that the model produces such alignment or "ferromagnetism."
So Gyu-Boong Jo, Wolfgang Ketterle, and colleagues at the Massachusetts Institute of Technology in Cambridge set out to reproduce the mathematical model experimentally in a puff of ultracold atoms. Currently, many physicists are pursuing such "quantum simulations" because the experiments may provide the best hope for solving such intractable mathematical models, which themselves are abstractions of the far messier physics of electrons whizzing around in solids.
Jo and Ketterle studied a puff of lithium-6 atoms, which, because of the way they spin, mimic electrons. The team trapped lithium atoms spinning in two directions in a spot of laser light and cooled them to within a millionth of a degree of absolute zero. By applying a magnetic field, they could make the atoms repel each other more or less vigorously.
It was a tricky experiment, as the atoms tended to undergo three-way collisions that would quickly turn pairs of them into molecules. Nevertheless, the team saw three signs that the atoms were aligning and the gas was becoming magnetized. First, as the strength of the repulsion passed a "critical" level, the rate of molecule formation peaked and began to plummet, suggesting that the atoms were aligning and avoiding each other. Second, the kinetic energy of the gas started to climb as expected. Finally, the size of the cloud peaked at that critical repulsion, too. All of this is consistent with the notion that the Stoner model produces alignment and ferromagnetism, they argue today in Science.
Ketterle and company weren't able to spot individual domains of alignment, which would be incontrovertible proof of ferromagnetism, says Wilhelm Zwerger, a theorist at the Technical University of Munich in Germany. Still, he says, "there is no other plausible explanation for the experiment."
But Tin-Lun "Jason" Ho, a theorist at Ohio State University, Columbus, disagrees. He says that all the data can be explained if subtle correlations between neighboring atoms keep opposite spins from colliding even though there is no overall alignment. The fact that Jo and Ketterle didn't see domains suggests they don't exist, Ho says: "Nature is telling us that the system is not ferromagnetic."