Search Magazine 
The fastest, most powerful, and smallest computers and sensors ever built will be possible through materials developed to enable the use of the electron’s spin, as well as its charge.

Unlocking Mysteries of the Nanoscale

Besides their negative electrical charge, electrons offer two other positive benefits—they have spin and act as tiny magnets. In the random access memory (RAM) of today’s computers, information is stored by electronic charges, but all the information is lost each time the device’s power is cut off, so the computer must be rebooted, which takes time. If, however, researchers can both use magnetic material to construct the computer’s RAM (MRAM) and control the spins of electrons, which create magnetism in a material when pointed in the same direction—like tops turning clockwise, then all stored information in a laptop computer could be permanently retained, even if the battery fails. The MRAM-based computer will be turned on instantly.

Researchers today are working to develop the “spintronic” device, which would exploit both electron charge and spin. The holy grail of spintronics research is a pocket-sized “quantum computer” that can replace a massively parallel supercomputer that fills a large room. Such a robust, versatile computer would need very little power. It would easily outpace today’s supercomputers in factoring any number down to its primes, which would, for example, help security agencies rapidly break the encrypted codes of hostile nations and terrorist cells.

The heart of a spintronic device is a chip containing nanosized “spin transistors,” which would operate using a “spin current” that flows only when the spins in the transistor’s source and drain electrodes have the same orientation. Such a spin transistor could be used as either a very fast on-off switch or to store digital information.

Two very important areas in nanoscience are spin-dependent transport of electrical current and magnetism. ORNL researchers are conducting research relevant to these areas.


Imagine a disk of copper or a single salt crystal about 2 cm (1 in.) across. Then picture three tiny structures made of magnetic iron lying side by side on one of these surfaces: a square film, snips of wire laid down parallel, and an array of dots. Amazingly, all three “nanostructures” were deposited on a single substrate the size of a thumbnail, using several different processes.

Jian Shen, a researcher in ORNL’s Solid State Division (SSD), has led a team that has accomplished this feat. “We have put a two-dimensional film, one-dimensional wires, and zero-dimensional dots together on a very small substrate,” he says.

Working with him have been Thomas Schulthess of ORNL’s Metals and Ceramics (M&C) Division, Lee Robertson of SSD, and Frank Klose of the Department of Energy’s Spallation Neutron Source (SNS) Project at ORNL. DOE’s Office of Basic Energy Sciences (BES) has supported this research.

Shen and his colleagues have compared the magnetic properties of all three nanostructures. They found that the magnetic properties are dramatically different for all three—film, wires, and dots—and also in comparison with bulk iron.

“We are interested in studying these nanostructures’ magnetic properties and also the spin-dependent transport of current through them,” Shen says. A researcher in SSD’s Low-Dimensional Materials by Design Group, Shen has been studying the effects of special confinement on magnetic properties of materials.

“One goal is to build an integrated surface semiconductor that combines magnetic and semiconductor elements to make a spin transistor that takes advantage of both the spin and charge of electrons in a current,” Shen says. “But we wondered how small these magnetic elements could be as electronic devices are reduced in size. If we go to smaller dimensions, do we reach a critical point where the magnetic properties differ from those of the bulk horseshoe magnet?”

ORNL researchers have deposited three iron nanostructures together on a very small substrate.
ORNL researchers have deposited three iron nanostructures—a two-dimensional film, one-dimensional wires, and zero-dimensional dots—together on a very small substrate, as shown in the atomic force microscope images and schematic above. To get a smooth iron film two atoms thick, they used an excimer laser to carry out molecular beam epitaxy (MBE), depositing iron atoms at a rate seven orders of magnitude higher than ordinary MBE. To lay down one-dimensional nanowires pointed in the same direction, electrochemical polishing was used to cut parallel steps in the surface on which iron was deposited using conventional MBE. The zero-dimensional nanodots were created by buffer-layer-assisted growth. The inert gas xenon was induced to adsorb to the substrate surface where it was frozen into a solid crystalline film. Iron particles evaporated by MBE formed an array of dots on the xenon layer, which was evaporated, leaving the dots clinging to the substrate.(Schematic redrawn by Judy Neeley.)

Theory predicts that thermal effects from temperature changes will destroy the directional stability of electron spins in nanosized magnetic material. But theorists also predict that some magnetic objects that are very small will have enhanced “magnetic antisotropy”—a property in which electron spins preferentially align in one direction.

“Magnetic ultrathin films or nanowires may have high magnetic antisotropy,” Shen says, “but theory also says that nanowires will have no ferromagnetic ordering—that is, no electron spin alignment at finite temperatures.”

To test these theories, Shen and his colleagues learned how to place three iron nanostructures—a film, wires, and dots—on a single substrate of copper, which proved to be no small task. Scanning tunneling microscope images of the copper substrate clearly showed the three iron nanostructures. Then, because copper conducts electricity, Shen and his colleagues decided to deposit these same iron nanostructures on a nonconductive, insulating substrate—a single crystal of table salt, or sodium chloride.

To compare the magnetic properties of all three nanostructures on the salt crystal, they subjected the substrate to an external magnetic field that was decreased and whose direction was changed. They found that the amplitude and direction of magnetization, preferred electron spin orientation, and stability of magnetic ordering were dramatically different for all three nanostructures. They also found that all three nanostructures lose their magnetism at different temperatures, none of which is the same as the Curie temperature for the horseshoe magnet.

“We found that the film stays magnetized, like a horseshoe magnet,” Shen says. “But the wire won’t stay magnetized in the direction imposed by the magnetic field—which is not a good effect. Also, half the dots point in the direction dictated by the magnetic field but the other half switch direction.” Shen’s group found that their experimental observations of the film agree with theoretical predictions. As for the array of dots, “There is some agreement but we don’t understand all the physical effects,” he says. “We found the net spins on wires are less than those on the dots. We could improve iron wires by alloying them with cobalt to enhance their magnetization, to make them more useful for spin transistors.”

Cadmium germanium disilicide crystals grown in a tin flux.
Cadmium germanium disilicide crystals grown in a tin flux. (Photo by Curtis Boles; enhanced by Jane Parrott.)


In one theory of high-temperature superconductivity supported by results of neutron scattering experiments in which ORNL researchers participated, electron spins can be separated from electron charges in one-dimensional regions called stripes. In this “electronic phase separation,” the spins can form superconducting pairs without their charges keeping them apart through mutual repulsion.

Today the major intellectual challenges of condensed-matter physics are to understand and control both complex, self-organizing behavior that emerges on the nanoscale and that involves collective phenomena (e.g., correlated electrons in which spins of all electrons in one microscopic region, or domain, are aligned) and competing states (e.g., charge and spin, electrical conductivity and magnetism). This self-organizing behavior, which occurs in new and some old materials, cannot be understood using traditional textbook concepts.

Double-layered strontium ruthenate crystals grown in an optical floating zone furnace at ORNL.
Double-layered strontium ruthenate crystals grown in an optical floating zone furnace at ORNL. (Photo by Curtis Boles.)

According to Doug Lowndes, a corporate fellow in SSD, ORNL researchers who were awarded funding in 2001 from BES’s Nanoscience, Engineering, and Technology (NSET) Program are focusing on understanding and controlling the highly correlated electronic behavior that results in spontaneous electronic-phase separation on the nanometer scale in transition metal oxides (TMOs).

TMOs, which include high-temperature superconductors, exhibit an astonishing variety of possible ground states very close together in energy, where the balance between competing phases is very subtle and small changes can create new phenomena or big effects. The properties of these materials traditionally have been “tuned” by doping. ORNL researchers plan to explore and control the collective phenomena exhibited by TMOs by growing TMO crystals and artificial TMO structures consisting of closely coupled alternating layers. They will use ORNL’s advanced characterization tools, including neutron scattering at the High Flux Isotope Reactor and eventually the SNS (under construction at ORNL), to determine if the exotic properties of these TMOs respond to dimensional confinement, strain, or the interaction of distinct adjacent nanoscale regions that arise spontaneously or as a result of careful engineering.

Rongying Jin of ORNL’s Solid State Division grows and characterizes crystals of transition metal oxides, which are correlated electron materials.
Rongying Jin of ORNL’s Solid State Division grows and characterizes crystals of transition metal oxides, which are correlated electron materials. She uses a Japanese-made optical floating zone furnace at ORNL to grow crystals of extreme purity. (Photo by Curtis Boles; enhanced by Judy Neeley.)

David Mandrus, leader of SSD’s Correlated Electron Materials Group and associate professor of physics at the University of Tennessee at Knoxville (UTK), ORNL-UTK Distinguished Scientist Ward Plummer, SSD’s Brian Sales, and several postdoctoral scientists are growing and characterizing crystals of correlated electron materials.

“These materials are very strongly coupled, meaning the charge, spin, and structure are all tied together,” Mandrus says. “A small change in the external environment—such as an applied magnetic field—can produce a large change in the material’s properties, such as its structure or electrical resistance. Thus, these exciting materials may have potential for use in sensors or active electronic components.

“Our focus is the discovery of new types of order—new phase transitions—in complex TMOs,” he continues. “For example, we recently discovered a unique charge density wave transition in the cadmium pyrorhenate superconductor crystals we have grown. Our discovery has attracted a lot of attention around the world, and various research groups are using our crystals for their research.

Single-layered strontium ruthenate crystals grown in an optical floating zone furnace at ORNL.
Single-layered strontium ruthenate crystals grown in an optical floating zone furnace at ORNL.(Photo by Curtis Boles.)

“In a charge density wave (CDW) transition, the electron density ceases to be uniform but instead displays a periodic spatial variation. CDWs are usually found in electronically anisotropic materials, but cadmium pyrorhenate is electronically isotropic, and it is not understood how a CDW transition can occur in this material.

“Neutron scattering is essential to our research because neutrons probe both atomic and magnetic order in a solid. Neutrons are subatomic magnets that scatter differently when spin alignments change.”

ORNL’s synthesis and characterization capabilities are being combined with the modeling talents of ORNL Corporate Fellow Malcolm Stocks (M&C) and Schulthess to better understand why TMOs behave as they do.

Pyrochlore cadmium rhenium oxide crystals made using a vapor-transport technique.
Pyrochlore cadmium rhenium oxide crystals made using a vapor-transport technique.(Photo by Curtis Boles.)

SSD’s Hans Christen plans to grow and stack alternating layers of electrically conducting and insulating perovskites, bringing together such dissimilar physical properties as magnetism and ferroelectricity on the length-scale of a few nanometers.

“Will putting the layers close together cause the electrons in adjacent layers to ‘feel’ each other and act collectively, forming correlated regions across a barrier made of a different material?” Christen asks. “If so, the material would be highly susceptible to external parameters, such as temperature, mechanical stresses, or electric or magnetic fields. One layer’s response could lead to measurable changes of, say, electrical resistivity or voltages, in the other layers. This type of nanoscale-coupled behavior may lead to interesting new sensors.”

Beginning of Article

Related Web sites

Search Magazine 

Web site provided by Oak Ridge National Laboratory's Communications and External Relations
ORNL is a multi-program research and development facility managed by UT-Battelle for the US Department of Energy
[ORNL Home] [Communications] [Privacy and Security Disclaimer]