By Peter Weiss
From Washington, D.C., at the 166th annual meeting of the American Association for the Advancement of Science
Typical microchips have more layers than a club sandwich. Circuit fabricators stack up thin films of semiconductors and other materials to create devices such as transistors and the connections between them.
Those stacks make the pioneers of a new branch of circuit technology nervous. Spin electronics, or spintronics, relies not on electric charge but on another property called spin. It’s roughly analogous to electron rotation. Developers have worried that electrons’ movement between layers might change the direction of spins—the data in spintronic devices.
New experiments indicate, however, that electron spins can breeze between different semiconductors with little or no mussing of their direction. “If you’re going to build anything useful, you’re going to have a series of interfaces. So, this is a very crucial piece of information,” says David D. Awschalom of the University of California, Santa Barbara, who led the research.
The Santa Barbara team, which collaborated with scientists at Pennsylvania State University in State College, previously described its experiments in the Jan. 31 Physical Review Letters.
Manufacturers already make commercially successful spintronics devices from metals, most notably the read-write heads in disk drives. However, researchers are still trying to figure out how to make the technology work in the semiconductors on which mainstream electronics is based (SN: 1/16/99, p. 39).
In the recent experiments, Awschalom and his colleagues deposited a thin layer of the semiconductor zinc selenide on another semiconductor, gallium arsenide. Using light pulses from a laser, they created, within the lower layer, clouds of electrons with identically oriented spins.
Then, they scanned the selenide surface with a second laser for signs of spins that had diffused through the boundary. From 2.5 percent to 10 percent of the spins made the crossing, the researchers estimate. That count matches rough calculations of how many spins would reach the interface if the particles diffuse randomly in all directions. Experiments now under way should clarify how permeable to spin the interface is, Awschalom says. They may also determine whether spin-polarized electrons cross the border or smuggle their spin alignments to electrons on the other side.
