Semiconducting Material May Have The Right Stuff To Develop Spintronic Devices

Scheduled for publication in the May 27 issue of Applied Physics Letters, the research demonstrates that development of new semiconductor spintronic devices, including a prototype magnetic semiconductor, is not very far off, according to the UB physicists.

"After years of struggling to create materials that would make spintronic devices possible, we now have them in hand," said Hong Luo, Ph.D., UB physics professor and a co-author on the paper.

"Based on what we know now about this material, we believe we have satisfied the most important requirements for spintronic devices that function at room temperature, such as a spin transistor," he said.

The UB team now is shifting its focus to begin work on such devices.

Spintronics, the emerging field of technology in which not just the charge, but the spin, of electrons is exploited, is expected to lead to dramatic improvements in electronic systems and devices, such as memory elements, logic elements, spin transistors and spin valves.

These improvements include faster processing speeds with less power consumption; non-volatility, where turning off the power doesn't "turn off" the information, and possibly the development of quantum computers.

The UB research was supported by a $10 million grant made in 2000 by the U.S. Defense Advanced Research Projects Agency's SpinS program to a consortium of institutions led by the University at Buffalo.

If certain classes of spintronic devices become a reality, they will far outperform conventional electronic devices because instead of relying on one of two binary digits to encode information, they could process data using any of an infinite variety of spin states of electrons.

This property would allow them to process at once millions or billions of bits of information, which conventional computers have to sift through one at a time.

"Based on the new functionalities associated with this 'spin degree of freedom,' there are entirely new circuit possibilities which have not even been envisioned yet," said Bruce McCombe, Ph.D., SUNY Distinguished Professor in the UB Department of Physics, associate dean of the UB College of Arts and Sciences and a co-author on the paper.

The UB materials are digital alloys in which Gallium Antimonide/Manganese is layered in very thin slices, measuring just a few atoms, with some of the layers containing controlled mixtures of the two.

These digital alloys take just a few hours to fabricate using a sophisticated technique called molecular beam epitaxy, in which an ultra-high-vacuum environment creates new combinations of atoms, which don't exist in nature.

Because the material developed at UB, Gallium Antimonide/Manganese (GaSb/Mn), is a modification of a well-studied semiconducting material, it should be comparatively easy to integrate with existing electronic systems, an important advantage.

The semiconductors developed at UB are the first to demonstrate at room temperature a phenomenon called hysteresis, which the researchers say is an unambiguous signature of ferromagnetism, another prerequisite for some types of spintronic devices, in which a lasting magnetic effect does not disappear when an applied magnetic field is withdrawn.

"This is the first report in the literature of any material that exhibits hysteresis in ferromagnetic semiconductors at and above room temperature," said Luo. "For real-world applications in spintronics, this has to be seen. It's not optional."

For most practical applications, the physicists explained, functionality at room-temperature and even higher temperatures is critical.

"The temperature hurdle has been the biggest challenge up to this point," explained Luo.

The Gallium Antimonide/Manganese showed ferromagnetism up to 400 Kelvin (about 260-degrees Fahrenheit), which is the upper limit of the magnetometer, so the physicists expect to see the phenomena at even higher temperatures.

By modifying their semiconductor and combining it with a non-magnetic semiconductor in a heterostructure, a type of structure in which two semiconductors are sandwiched together, the UB team expects to be able to manipulate spin-polarized electrons, an important goal of creating materials with spintronic properties.

"Electrons move more readily in these non-magnetic materials and also are more likely to maintain their spin properties, which permits them to store and transmit information," explained McCombe.

Spintronic materials would enable the storage and processing of data on the same material, a kind of "computer on a chip," which could be a vast improvement over current electronics, where permanent data storage requires magnetic media, physically separate from the semiconductors, which do the processing.

In addition to Luo and McCombe, the paper was co-authored by X. Chen, M. Na, M. Cheon and S. Wang, all doctoral candidates in the UB Department of Physics; W. Liu, Y. Sasaki, T. Wojtowicz and J.K. Furdyna in the Department of Physics at the University of Notre Dame, and S. J. Potashnik and P. Schiffer in the Department of Physics and Materials Research Institute at Pennsylvania State University.