We've gotten a charge out of the electron
for years in all kinds of digital devices. Now it's time to take the electron
for a spin.Spin-based electronics —— a.k.a. spintronics —— is a growing
field of science that seeks to use the spin of electrons to create a new
generation of computers and a host of new microelectronic devices that will
be smaller, faster and more robust than their predecessors."Ideally,
if the plans unfold, spintronics will render classical computers obsolete,"
said Henry van Driel, chair of the physics department at the University of
Toronto. "It's almost mind-boggling when you think about what some of the
capabilities might be."Van Driel has been working on spintronics research
for the past two years, slowly watching those plans unfold through his own
research and the work of countless other physicists around the world.
At first glance, Van Driel's spintronics lab looks like the backroom of an
insane optometrist's office. There are lenses on black and silver stands,
seemingly scattered all over a couple of metal tables and a couple of glowing,
green beams of light. But their placement is far from haphazard. Each
lens is actually screwed into the table and angled just so. This is not a
place for small children. Move one of those lenses just ever so slightly
and nothing will work.The lights and lenses are all part of an elaborate
spintronics experiment, in which each of the lenses plays a critical role
in getting just the right kind of light to shine on the electrons and control
their spin.Thinking back to your high school science, you'll probably
remember that electrons —— those subatomic particles that are sometimes found
orbiting the nucleus of an atom —— are negatively charged. That negative
charge makes up the electric current that runs through the circuit boards
of your Walkman or VCR. "All electronics today is based on using electrons
for their charge, but the fact is that electrons don't have only charge,
they also have spin," said Peter Nemec, a visiting physics professor at the
University of Toronto, who is part of the spintronics team.Spin is
what's called a quantum property, something that governs the way the parts
of atoms interact with each another. Electrons don't really spin in the normal
sense of the word, but it's a decent analogy for their angular momentum.
There are two basic states of spin: spin up and spin down. This is akin to
spinning a top clockwise or counter-clockwise. In a normal electronic circuit,
the electron's spin doesn't have any effect on the circuit because the different
spins are randomly distributed. The trick to spintronics is to be able to
organize electrons according to their spin and get them to move from one
place to another. "In future, new devices may be purely based on the
spin properties of electrons, but most likely there will be a hybrid phase
where in some cases it's the electrical properties that are used and in some
cases it will be the spin properties," said van Driel.Why do we need
to use spin, when we've already got the charge to play with? Researchers
say there's a limit to what can be done using an electron's charge alone.
"Electronics based on charge is hitting a technology barrier," said
Andrew Sachrajda, a physicist and spintronics researcher for the National
Research Council (NRC) of Canada.
`Ideally, if the plans unfold, spintronics will render classical computers
obsolete. It's almost mind-boggling when you think about what some of the
capabilities might be.'Henry van Driel chair of the physics department, University of Toronto
At some point, using traditional electronics, we will hit the end
of Moore's Law, which states that computer performance will double every
18 months. That increase is based on our ability to pack more and more transistors
in the same space. While Sachrajda notes that Moore's Law has held
true for about three decades, there are limits to just how small we can make
traditional electronic components. When we get down to making things really
small, there are new laws that govern how particles interact. These are the
laws of quantum mechanics. So, to really make things work on a small scale,
we need to work with the quantum properties of the electron, such as spin.
Working with spin has other potential advantages. Researchers believe that
one day the up and down states of spin could be used to represent encoded
information —— like the ones and zeros that make up the binary computer language.
Spin was actually discovered about 80 years ago, but it's only recently
that it really seemed like something that humans could harness for their
own use.In the late 1980s, it was discovered that when electron spins
were organized so that they were all spin up or all spin down, a magnetic
effect called giant magnetoresistance (GMR) was produced. Magnetoresistance
had been used in computer hard disks to access data, but this newly discovered
variety was 200 times stronger. In practical terms, it meant that far more
data could be crammed into a hard disk if it was read using materials with
GMR. "That's what made laptops so small," said Sankar Das Sarma, a University of Maryland professor, who researches spintronics.
Das Sarma calls this kind of spintronics —— which works in metals —— the
first generation of spintronic devices. But he said that's only a small part
of the future of spintronics. While the first generation of spintronics devices
made computer memory more efficient, the next generation will combine memory
with the computer processor.The two parts of today's computer —— the
hard disk memory and the central processing unit —— are separate. Second-generation
spintronics would bring the logic and memory functions together in a single
semiconductor chip."As a trivial benefit, imagine your computer switching
on almost immediately instead of waiting during that annoying, long booting-up
time," said Sachrajda of the NRC. He and his colleagues recently announced
they have created a spintronic transistor. It's a landmark finding, but it's
still far from ready for commercial use. It's really just a prototype to
show what can be done, once some of the other issues with spintronics have
been solved.A critical issue in advancing spintronics is finding the
right materials for the job. For a long time, the trouble with this second-generation
spintronics research was finding a semiconductor with magnetic properties.
Then about six years ago, it was discovered that a common semiconductor had
magnetic properties at low temperatures. That brought new hope to the field
and research in the area took off. But for researchers such as Das Sarma,
the hunt is still on for semiconductors that have magnetic properties at
room temperature.Van Driel's lens-ridden experiments get around this
problem by avoiding the use of magnetism. Instead, van Driel used polarized
light waves that vibrate in a very specific way to generate electrons with
a certain spin state, and make them move in a certain direction, thus producing
a spin current."If you're going to make any kind of a device, then you have to be able to transport those electrons," van Driel said.
While researchers focused on magnetism deal largely with up and down spins,
this all-optical solution is also concerned with the full range of spin states
in between. Van Driel is actually trying to generate these mixed spin states
(a.k.a. superpositions) in electrons because they will be able to represent
even more information.
`Viable technology based on spintronics will be available in the next five years. Of that I'm absolutely confident.'Sankar Das Sarmaprofessor and researcher at the University of Maryland
"With respect to new ways of processing and storing information,
the key is to take advantage of these in-between states," he said. "This
provides extra information-carrying capacity."Just how these mixed
states can represent extra data is a bit confusing. Like much of quantum
physics, the logic behind this theory sounds like it comes from the Wonderland.
Everything is so different behind the looking glass. Similarly, the laws
of physics seem to change into something beyond our everyday experience when
it comes to the very small.But the basic, though seemingly unbelievable
reason is that mixed spins actually represent several different spin states.
It's like one object that's really two. As electrons in a mixed spin
state interact with other particles, they tend to fall back into either a
spin-up or spin-down state in a fraction of a second. But van Driel said
that's likely long enough to use them to briefly hold data used in rapid
computer processing.Right now, the University of Toronto team is verifying
its work, checking the strength of the currents and seeing how long they
will retain certain spin characteristics."This is a very pioneering
area and there is an awful lot of fundamental work that has to be done before
one can talk about devices," van Driel said.Spintronics will also
be an important part of the much-hyped, though far from realized, quantum
computer. Such a computer would use the laws of quantum physics to achieve
significant improvements in processing power. But Das Sarma said such a machine
is at least 50 years away because it would require reading the spin of individual
electrons."You can think of each spin as a little bar magnet, but
it's also an incredibly weak bar magnet, so measuring it is a very tall order,"
he said. "And for quantum applications, that needs to be done."Fortunately,
there are simpler spintronics applications that will be perfected much sooner.
Both van Driel and Das Sarma are confident that other spintronics technologies
will emerge in the next couple of years."Viable technology based on
spintronics will be available in the next five years. Of that I'm absolutely
confident," said Das Sarma. "Whether that will actually go into any consumer
applications or not, I simply don't know. That involves a business model."
And Das Sarma admits he's no expert on profit models. But there's a growing
sense that the technology will prove valuable. The Institute of Physics,
an international professional association, estimates that the potential market
for spintronics will be worth hundreds of billions of dollars a year.
Das Sarma's own work is funded in large part by the U.S. military. He said
he gets about 80 per cent of his $1 million budget from the U.S. Department
of Defence."The big word within the Department of Defence now is multi-functional,"
he said. "You have one device that can do 50 different things. It needs to
be small, robust and flexible."And he believes that spintronics will
likely be able to meet those needs, by combining traditional electronics,
magnetics and optics. It would be like having your electronic central processing
unit, magnetic hard drive and optical CD player all in one piece, on one
medium —— instead of the current computer system which connects separate
devices that lose precious processing time sending and converting data between
one unit and the next.It might sound like a fanciful idea, but Das
Sarma notes that much of science seemed ridiculous at some point. When lasers
first came out, he said, no one thought they would have any use at all. They
were big and cumbersome and took a lot of time and effort to operate. Today,
they are used in all kinds of home electronics and medical applications.
"Just look at the history of technology," he said. "When a new fundamental
aspect of nature is conquered, invariably there are applications. And often
those applications come from a direction that nobody knows of."
Legal Notice:
- Copyright 1996-2002. Toronto Star Newspapers Limited. All rights reserved.
Distribution, transmission or republication of any material from www.thestar.com
is strictly prohibited without the prior written permission of Toronto Star
Newspapers Limited. For information please contact us using our webmaster form.
