current closer to computing
Technology Research News
One way to significantly improve computers
is to use something other than the presence or absence of electric current
to signal the ones and zeros that form the binary logic of computing.
One promising alternative takes advantage of the quantum nature of electrons.
Spintronics is an emerging field that uses the spin of electrons to represent
ones and zeros. Electrons spin in one of two directions, up or down, which
is roughly analogous to a top spinning clockwise or counterclockwise.
In theory, these two states of an electron would allow for ultra low-power
conventional computers and would provide the means for moving information
within and between quantum computers. Proposed schemes for quantum
computers use atoms or subatomic particles to represent ones and zeros
and use quantum mechanics to check every possible answer to a problem
at the same time.
In practice, there are many details to be worked out.
In order to use electron spin to signal a one or zero, the spins of a
group, or current of electrons have to be aligned, and this collective
spin must survive the electrons' transfer from one transistor to another
and then last long enough to be useful.
Researchers from the University of California and Pennsylvania State University
have moved spintronics a significant step forward by demonstrating that
it is possible to efficiently move a current of electrons, with their
collective spin intact, from one semiconductor
material to another. In addition, the research shows that the spin state
can be made to last as long as 100 nanoseconds, which is long enough to
work for traditional computing.
"We have shown that spin lifetimes can exceed 100 nanoseconds and can
be transported over distances exceeding 150 microns. In both cases, this
exceeds the time and length scales used in today's technology," said David
Awschalom, a physics professor at the University of California at Santa
That this was fairly easy to accomplish surprised even the researchers.
The implication of the results is that it should be possible to fabricate
spin transistors, said Awschalom.
To investigate how practical using electron spin for computing could be,
the researchers measured the spin of a current of electrons that was moving
from a gallium arsenide semiconductor to a zinc selenide semiconductor.
"I thought that this would be the simplest laboratory in which to test
the basic idea: an atomically clean interface between two well-studied
semiconductors," said Awschalom.
The researchers started by using polarized laser beams to create in a
layer of gallium arsenide a reservoir of electrons whose spins were aligned.
Ordinarily, only a small number of electrons from this reservoir would
cross the barrier to a layer of zinc selenide and their spins would become
random within a few hundred picoseconds, or trillions of a second. The
researchers found that applying an electric field increased the number
of electrons crossing the barrier by 40 times and also boosted the lifetime
of the spins to usable levels.
Ultimately, the researchers hope to use electron spins for high-density
information technology and fundamentally new methods of information processing
like quantum computation, said Awschalom. If practical quantum computers
can be built, they would be phenomenally fast for solving certain problems
like cracking codes and searching large databases.
The experiments are something of a milestone in the spintronics field,
said Jay Kikkawa, an assistant professor of physics and astronomy at the
University of Pennsylvania.
In the experiments, the spin of the electron acts as an identification
tag, said Kikkawa. "Its response to a magnetic field reveals the electron's
magnetic history, which, in part, includes how long spins have spent in
different layers," he said.
The researchers use this information to distinguish among several different
channels within a spin current flowing across an interface between materials.
"It's a very clever trick that one could never pull off in a purely electrical
system," Kikkawa said. This is because electrical current consists of
electric charge and the spins of its electrons are random.
Electron spins could be used in computing within the decade, Awschalom
Awschalom's research colleagues were Irina Malajovich of the University
of California at Santa Barbara, and Joseph J. Berry and Nitin Samarth
of Pennsylvania State University. They published the research in the June
14, 2001 issue of the journal Nature. The research was funded by the Defense
Advanced Research Projects Agency (DARPA), the Office of Naval Research
(ONR) and the National Science Foundation (NSF).
Timeline: < 10 years
TRN Categories: Semiconductors and Superconductors; Quantum
Story Type: News
Related Elements: Technical paper, "Persistent Sourcing
of Coherence Spins for Multifunctional Semiconductor Spintronics," Nature,
June 14, 2001 were.
Pen and paper networked
closer to computing
choices key for speech software
tales into animations
Watched quantum pot
Research News Roundup
Research Watch blog
View from the High Ground Q&A
How It Works
News | Blog
Buy an ad link