Twenty-two years ago, physicists discovered that certain materials containing copper and oxygen could carry electricity without any resistance at inexplicably high temperatures--well above absolute zero although still far colder than winter in Fargo, North Dakota. Such "high-temperature superconductivity" remains the biggest mystery in the physics of materials, and some scientists are trying to solve it in an unusual way, by simulating the crystals with patterns of laser light and puffs of ultracold atoms. Now, a team in Switzerland has taken a key step in that effort by replicating the basic starting point from which the superconductivity is thought to emerge--a sort of traffic jam in which nothing moves.
High-temperature superconductors carry electricity without any resistance at temperatures as high as 138 kelvin--three times the highest temperature for an ordinary superconductor. The materials contain planes of copper and oxygen ions. The electrons in the planes pair up, and this pairing allows the electrons to glide along the planes without any of the electrical drag found in normal materials. However, physicists aren’t sure just what glues the pairs together in the first place. What seems to happen is this: At low temperatures, the electrons in the planes push against each other mightily. That causes them to get stuck in a massive traffic jam called a "Mott insulator" state. Then, if the material's composition is adjusted to take out a few of the electrons, the rest somehow pair up to flow unimpeded. Or so the idea goes. Nobody has proved mathematically that this "Hubbard model" of strongly repelling electrons hopping around in planes produces superconductivity.
So, like others around the world, Henning Moritz, Tilman Esslinger, and colleagues at the Swiss Federal Institute of Technology in Zurich hope to skip the math by making a Hubbard model using laser light and atoms. The idea is to use intersecting laser beams to generate a three-dimensional pattern of bright spots that mimic the patterned arrangement of ions in a crystal. Atoms chilled to less than a millionth of a kelvin get trapped in these bright spots in this "optical lattice" and can hop between them to mimic the movement of electrons. Now, Esslinger, Moritz, and colleagues have reached the first major milestone on the path to their goal, the jammed-up Mott insulator state with one atom per bright spot, as they report this week in Nature.
The researchers loaded their lattice with potassium-40 atoms--chosen because they spin in a way that makes them particularly similar to electrons. The researchers devised a way to tally the fraction of spots occupied by two atoms. They then adjusted a magnetic field, which controlled the repulsion between atoms. If the atoms pushed against one another hard enough, the number of doubles plummeted to essentially zero, indicating that the atoms were spreading out and that Mott state had been achieved. "When I started here 6 years ago, that was the goal, so we are extremely happy to have achieved it," Esslinger says.
But Immanuel Bloch, an experimenter at the University of Mainz, Germany, isn't sure that the Swiss team has. The Mott state requires precisely one atom per site, he notes, and although Esslinger and colleagues have essentially no sites with two atoms, they may still have plenty of sites that don't have any. "You could still have a lot of holes" in the pattern of atoms, Bloch says. However, Henk Stoof, a theorist at the University of Utrecht in the Netherlands, says the result "looks to me pretty convincing."
Researchers still have two big steps to take to fully imitate a high-temperature superconductor in cold atoms. First, they must coax the neighboring atoms to spin in opposite directions, mimicking a condition in the copper-and-oxygen planes known as "antiferromagnetism." That step may come within a year, researchers say. The last step--making the atoms pair--may take longer, as it requires making the atoms much colder still.