Doing the Spintronics Dance
Carlos Andres Meriles | City College of the City University
of New York
Picture a professor about to lecture to his physics majors. Like characters
in a Woody Allen movie, they will soon be off zipping around Manhattan,
relaxing, bumping into things, losing information. In a way, they are
like the free-spirited electrons in Carlos Andres Meriles’ research.
But, says Meriles, of City College of the City University of New York,
that stable professor, like the fixed nucleus in his work, will remember.
He will be there, same place, same hour, next week. Happily, new interactions
will refresh those memories, until final exams, he notes wryly.
(A) Diagram of the magnetic field created by a saddle (Golay) coil. (B) Calculated image of a spin system distributed as shown in the insert. Evolution takes place in the magnetic field described in (A). An image can be retrieved by creating an average Hamiltonian that resembles the situation at high-field
Those patterns — remembering and forgetting by electrons and nuclei — and
how to control them are central to his work on magnetization, semiconductors
and spin, which is at the heart of a new electronics field called spintronics.
“It’s a big trend in physics and engineering to try to control the
spin property of electrons,” he said. “Now, digital devices
work on the basis of charge — current or the absence of it, defining
a zero or a one. But we know that in spin-based logic a transistor could
be more efficient, use less energy, run faster.”
In spintronics, dissipated heat will diminish, as will the size of circuits,
down to a billionth of a meter, or one nanometer.
So far, there’s no Alexander Graham Bell for the spintronics era,
but the industry has experts in physics, engineering and materials science
hard at work on how spin behaves, seeking a way to make spin-based electronics
a viable technology.
Electron spin has its own problems. Due to interactions with other particles,
the once neatly aligned spin loses its memory and realigns randomly, in
what’s
called relaxation. Finding a way to postpone that relaxation, or decay,
is vital in a semiconductor to retaining information.
The strong interaction in a semiconductor’s nanostructures between nuclear
and electron spins can be exploited several ways. For example, volatile information
encoded through the electron spin could be transferred to longer-lived nuclear
spins for storage. Alternatively, electron spins aligned by external means — like
light pulses — can be exploited to increase the nuclear spin magnetization
beyond what’s possible with the strongest magnets available today.
And the nucleus poses its own spin issues. In semiconductor materials — say,
cadmium telluride, selenium, cadmium selenide or indium phosphide — nuclear
spin is the same as electron spin, which Meriles said, “makes your
life simpler.”
Meriles is studying these factors in spintronics in search of ways to extend
the time information can be stored. Nuclear spins live longer but they
are hard to align.
“By manipulating the electrons, you can do things in the nucleus,” he
said. “We can store information relatively quickly, and longer, in
the nuclei.”
On a far horizon are more efficient devices — memory chips, cellphones,
even a TV or a watch. “There are many things we don’t do because
of the time it takes,” he said. “If we can do them in a millionth
of the time, we’d do a lot more.”
He knows the slow pace. As a student in Cordoba, Argentina, in 1990, he
wrote on an early computer consisting of a keyboard, a tape recorder and TV screen. “I
hated it,” Meriles recalled, “because it took so long.”
Nov. 1, 2007
