|
|
|
|
|
|
New supercomputers at ORNL are unlocking the mysteries of superconducting materials.
Biofuels produced from biomass, such as cellulosic ethanol, could replace perhaps as much as a third of the current U.S. demand for transportation fuels with a homegrown, renewable energy source without affecting food production," according to Department of Energy Under Secretary for Science Raymond Orbach. While physicists rarely have the image of the 1960s counterculture, the first American Physical Society session on high-temperature superconductors, held in 1987 at New York City's Hilton Hotel, has been termed by some the "Woodstock of Physics." Excited by the discovery of high-temperature superconductors (HTSCs) the previous year, more than 2,000 physicists packed a hotel meeting room and spilled out into the corridors, fighting for the privilege of hearing the technical papers first-hand. The session lasted all night. The conference is perhaps the only meeting of physicists ever characterized as "a riot." Twenty years after "Woodstock," the potential of high temperature superconductors both tantalizes and taunts the scientific community. Not unlike Woodstock, the jubilation of 1987 eventually encountered reality. Fulfilling the promise has proved to be much harder than predicted by many of those starry-eyed physicists two decades ago. HTSCs are being used in some applications, and the technology is advancing, but superconductivity is far from being a part of daily life. New tools The technological breakthroughs needed to accelerate the pace of superconductivity research may now be possible with a new generation of supercomputers. In calculations conducted during 2006, a team of scientists used the unparalleled computing resources at Oak Ridge National Laboratory to identify a mechanism that provides clues to the mystery of how HTSC materials work. The discovery may eventually help to realize the potential of these highly energy-efficient materials. HTSCs are materials that conduct electricity without resistance at temperatures as high as 150 K (roughly -190°F). While extremely cold, the temperature is relatively balmy compared with the level at which conventional superconductors must operate—near absolute zero, or below -400°F. Because the higher temperatures are more manageable, HTSCs have much more technological potential than conventional superconductors. A theoretical understanding of why HTSCs lose their resistance to electricity remains elusive. "There has been a huge amount of theoretical and experimental work on high-temperature superconducting systems, but no complete understanding," says ORNL's Thomas Maier. "One especially would like to understand what causes the pairing interaction and why these systems become superconducting." The question is a very important one, particularly in a world confronting rising temperatures and looming energy shortages: Understanding why HTSCs behave as they do would open the door to the development of new materials operating at higher temperatures, perhaps even at room temperature. These new materials would pave the way for breakthrough applications that could save enormous amounts of energy. Researchers envision power cables that transmit electricity with little or no losses, mass production of consumer electric vehicles, super-efficient high-speed trains and leaps in energy efficiency for a variety of electrical machinery. No resistance Maier is part of an Oak Ridge team working to develop a theoretical description of HTSCs. One mystery they are trying to solve is why the electrons in some materials bond to form pairs called "Cooper pairs" that settle into a state in which they conduct electricity without resistance. "The ultimate reason a material becomes superconducting is that the electrons join into Cooper pairs," Maier explained. "If a current is driven through the system, the superfluid phase formed by the Cooper pairs does not resist it." As a basis for their calculations, the scientists are studying the two-dimensional Hubbard model, believed by most physicists to provide an appropriate framework for describing the physics underlying HTSCs. The team resolved a key question in 2005 with simulations showing that superconductivity emerges as a result of strong interactions among electrons. Those calculations, conducted on the Cray X1E Phoenix supercomputer, were the first solution of the Hubbard model ever to include a large enough atom cluster to provide confidence in the results. The 2006 calculations addressed the next step in developing the theory: uncovering the mechanism that underlies the Cooper pairing. Scientists know that some force causes electrons to form Cooper pairs via a pairing interaction. The 2006 simulations, also run on Phoenix, were aimed at identifying the force. They revealed that the interaction behind Cooper pairing is driven by a mechanism called "spin fluctuation," a magnetic effect associated with the rotation of electrons. "All the structure in this pairing interaction comes from the spin fluctuation contribution. We have shown that in the Hubbard model, commonly believed to be a description of high-temperature superconductors, spin fluctuation mediates the pairing that leads to superconductivity," says Maier. All of the known HTSCs are cuprates, materials with copper-oxide planes separated by atoms of other elements. Maier says that theorists have long speculated that spin fluctuations could lead to the type of symmetry found in Cooper pairs in the cuprates. But the calculations conducted at Oak Ridge are the first conducted at a scale adequate to analyze a microscopic model for the source of the pairing interaction and confirm the theorists' understanding.
Doug Scalapino of the University of California–Santa Barbara was among the first to propose that spin fluctuations underlie the pairing mechanism in HTSCs. Scalapino, a key figure in the HTSC research community and a contributor to the Oak Ridge project, said the new findings help relate the Hubbard model to lab experiments on cuprates, particularly neutron scattering and angle-resolved photoemission scattering. The Oak Ridge supercomputers revealed ways to address the pairing mechanism by using data obtained from actual materials by these techniques. Researchers do not know how to measure the pairing mechanism directly in materials, Scalapino explains, but the Hubbard model used to reveal the pairing mechanism can be used to obtain neutron and angle-resolved photoemission scattering results. Once that is done, "we can find the particular combination, the certain way of putting those together that does a good job of predicting what we know the pairing interaction is." Researchers then can use that formula to do the same thing for the actual material—use neutron scattering and angleresolved photoemission data obtained directly from materials to calculate the pairing interaction in actual materials. The team will build on the 2005 and 2006 breakthroughs to answer other critical questions. For example, what causes different cuprates to have different transition temperatures (i.e., to become superconducting at different temperatures)? What mechanism besides spin fluctuations might enhance the pairing interaction and affect transition temperatures? One possibility suggested by experiment is inhomogeneities in a material, says Maier. In 2007, his team added inhomogeneities to their model to see how they affected the pairing interaction and the transition temperature. The calculations for the inhomogeneities probably were conducted on ORNL's Cray XT4 Jaguar supercomputer because, whereas the previous simulations required a smaller number of faster processors, the new problem required a large number of processors. "The interest ultimately is in how to increase the transition temperature to room temperature," Maier says. "If we can understand why a material is superconducting at 150 K, then we can ask what we must do to raise that temperature. There is no guarantee once we understand the process that we can raise the transition temperature, but it's a big step." Had Maier's colleagues at the 1987 "Woodstock" conference had access to modern supercomputers, one can imagine that the path to understanding the mysteries of high-temperature superconductors would have been an easier one.
|
 |
 |
 |
 |
 |
Web site provided by Oak Ridge National Laboratory's Communications and External Relations |
