This story was revised 23 October to improve the explanation of GMR physics.
Last week two European researchers received the Nobel Prize in Physics for discovering that a magnetic field significantly reduces the electrical resistance of certain atom-thin metallic sandwiches. Their reports were published in Physical Review Letters and Physical Review B in 1988 and 1989. Since then, technologists have improved the effect, and it is now widely used to detect the tiny patches of magnetism that store information on high-density disk drives.
Beginning in the 1960s, physicists were using tools adapted from the semiconductor industry to create new materials from scratch, building them up layer-by-layer at the atomic scale. Meanwhile, other researchers were studying properties like magnetism. Some realized they could use the new materials tools to test their understanding of the magnetizing process, because they could precisely control layers of both magnetic and nonmagnetic metal atoms.
Pursuing such fundamental questions with their collaborators, Albert Fert, then of the University of Paris-South in Orsay, France, and Peter Grünberg of the Institute for Solid State Physics Research in Jülich, Germany, independently discovered a new magnetic effect. Grünberg’s team created a sandwich of two several-atom-thick layers of magnetic iron separated by a layer of nonmagnetic chromium between them. They measured the electrical resistance parallel to the layers and found that it dropped by about 1.5% when they placed it in a magnetic field. Fert’s team measured a twofold reduction in resistance in a related many-layered stack at low temperatures. Since a magnetic field changes the resistance of ordinary metals by only a fraction of a percent, they dubbed this phenomenon “giant magnetoresistance,” or GMR.
Current-carrying electrons in a metal travel in all directions, with only a slight average tendency toward the direction of current flow. In these high-quality samples, electrons frequently zip across all of the layers, only occasionally scattering from defects (irregularities) in the crystalline layers or from the interfaces between them. The amount of scattering determines the resistance of the sandwich. GMR arises because each electron carries a spin that points in a specific direction. The likelihood of an electron scattering in a magnetized iron layer depends on whether the spin points in the same direction as the iron magnetization or in the opposite direction.
To take an extreme example, imagine that an electron does not scatter in an iron layer unless its spin points in the same direction as the layer’s magnetization. Grünberg’s team prepared their three-layer sample with the magnetization in the two iron layers pointing in opposite directions. In this scenario, all the electrons eventually scatter during their long trip along the sandwich: the spin-up electrons from one iron layer and the spin-down electrons from the other. With the magnetic field turned on, the iron layers both magnetize, say, in the spin-up direction, so all of the spin-down electrons move through the entire sample without scattering. The real effect is smaller because electrons scatter somewhat from both layers.
“Once you see a magnetoresistance effect that’s relatively large,” says Stuart Solin of Washington University in St. Louis, “it’s fairly straightforward to envision that it might be technologically important.” Soon other researchers, notably Stuart Parkin of IBM in Almaden, California, laboriously developed practical structures to sense the small magnetic fields in disk drives, and now GMR-based devices have largely replaced those based on less sensitive materials. GMR also exemplifies two principles that could drive future technology: nanotechnology, modification of materials at ultra-small scales to improve their properties–in this case adding the thin chromium layer between the iron layers–and spintronics, the use of electrons’ spins, as well as their charges, for devices.