Using Special Gases as Brake, They Tame the Fastest Thing in Universe
NEW YORK Researchers say they have slowed light to a dead
stop, stored it and then released it as if it were an ordinary
The achievement is a landmark feat that, by reining in nature's
swiftest and most ethereal form of energy for the first time,
could help realize what are now theoretical concepts for vastly
increasing the speed of computers and the security of
Two independent teams of physicists have achieved the result,
one led by Lene Vestergaard Hau of Harvard University and the
Rowland Institute for Science in Cambridge, Massachusetts,
and the other by Ronald Walsworth and Mikhail Lukin of the
Harvard-Smithsonian Center for Astrophysics, also in
Light normally moves through space at 186,000 miles a
second. Ordinary transparent media - such as water, glass and
crystal - slow light slightly, an effect that causes the bending of
light rays, which allows lenses to focus images and prisms to
Using a distantly related but much more powerful effect, the
Walsworth Lukin team first slowed and then stopped the light in
a medium that consisted of specially prepared containers of
gas. In this medium, the light became fainter and fainter as it
slowed and then stopped. By flashing a second light through
the gas, the team could essentially revive the original beam.
The beam then left the chamber carrying nearly the same
shape, intensity and other properties it had when it entered.
The experiments led by Ms. Hau achieved similar results with
closely related techniques.
"Essentially, the light becomes stuck in the medium, and it can't
get out until the experimenters say so," said Seth Lloyd, an
associate professor of mechanical engineering at the
Massachusetts Institute of Technology who is familiar with the
work. He added, "Whoever thought that you could make light
He said the work's biggest impact could come in futuristic
technologies called quantum computing and quantum
communication. Both concepts rely heavily on the ability of light
to carry so-called quantum information, involving particles that
can exist in many places or states at once.
Quantum computers could crank through certain operations
vastly faster than existing machines; quantum communications
could never be eavesdropped upon. For both these systems,
light is needed to form large networks of computers. But those
connections are difficult without temporary storage of light, a
problem that the new work could help solve. A paper by Mr.
Walsworth, Mr. Lukin and three collaborators - David Phillips,
Annet Fleischhauer and Alois Mair, all at Harvard-Smithsonian -
is scheduled to appear in the Jan. 29 issue of Physical Review
Citing restrictions imposed by the journal Nature, where her
report is to appear, Ms. Hau refused to discuss her work in
Two years ago, however, Nature published Ms. Hau's
description of work in which she slowed light to about 38 miles
(61 kilometers) an hour in a system involving beams of light
shone through a chilled sodium gas.
Mr. Walsworth and Mr. Lukin mentioned Ms. Hau's new work in
their paper, saying she achieved her latest results using a
similarly chilled gas. Mr. Lukin cited her earlier work, which Ms.
Hau produced in collaboration with Stephen Harris of Stanford
University, as the inspiration for the new experiments.
Those experiments take the next step, stopping the light's
"We've been able to hold it there and just let it go, and what
comes out is the same as what we sent in," Mr. Walsworth said.
"So it's like a freeze frame."
Mr. Walsworth, Mr. Lukin and their team slowed light in a gas
form of rubidium, an alkaline metal element.
The deceleration of the light in the rubidium differed in several
ways from how light slows through an ordinary lens.
For one thing, the light dimmed as it slowed through the
rubidium. Another change involved the behavior of atoms in the
gas, which developed a sort of impression of the slowing wave.
This impression, actually consisting of patterns in a property of
the atoms called their spin, was a kind of record of the light's
passing and was enough to allow the experimenters to revive or
reconstitute the original beam.
Both Ms. Hau's original experiments on slowing light, and the
new ones on stopping it, rely on a complex phenomenon in
certain gases called electromagnetically induced transparency,
This property allows certain gases, like rubidium, that are
normally opaque to become transparent when specially treated.
For example, rubidium would normally absorb the dark red
laser light used by Mr. Walsworth and his colleagues, because
rubidium atoms are easily excited by the frequency of that light.
But by shining a second laser, with a slightly different
frequency, through the gas, the researchers rendered it
The reason is that the two lasers create the sort of "beat
frequency" that occurs when two tuning forks simultaneously
sound slightly different notes. The gas does not easily absorb
that frequency, so it allows the light to pass through it; that is,
the gas becomes transparent.
But another property of the atoms, called their spin, is still
sensitive to the new frequency. Atoms do not actually spin; the
property is a quantum mechanical effect analogous to a tiny
bar magnet that can be twisted by the light. As the light passes
through, it alters those spins, in effect flipping them. Though
the gas remains transparent, the interaction serves as a sort of
friction or weight on the light, slowing it.
Using that technique, Ms. Hau and Mr. Harris in the earlier
experiment slowed light to a crawl. But they could not stop it,
because the transparent "window" in the gas became
increasingly narrower, and more difficult to pass through, as
the light moved slower and slower.
In a recent theoretical advance, Mr. Lukin, with Suzanne Yelin
of Harvard Smithsonian and Michael Fleischhauer of the
University of Kaiserslautern in Germany, discovered a way
around this constraint.
They suggested waiting for the beam to enter the gas
container, then smoothly reducing the intensity of the second
The three physicists calculated that this procedure would
narrow the window, slowing the first beam, but also "tune" the
system so that the beam always passes through.
The first beam, they theorized, should slow to an infinitesimally
slow speed, finally remaining only as an imprint on the spins.
Turning the second beam back on should reconstitute the first
The new experiments bore those ideas out.
"The light is actually brought to a stop and stored completely in
the atoms," Mr. Harris said. "There's no other way to do that.
It's been done - done very convincingly, and beautifully."