After 20 years of exploring the esoteric nature of liquid helium when it is cooled to ultra-low
temperatures in zero gravity, physicist John Lipa suddenly finds that his work could have important ramifications for the miniaturization in the
microelectronics industry.
His latest experiment — scheduled to launch in the space
shuttle Columbia on Nov. 19 — is called the Confined
Helium Experiment (CHeX). Its purpose is to determine
what happens to a material when it is confined to such
narrow dimensions that it begins to behave as if it has
only two dimensions, rather than three. Lipa is the
principal investigator on the experiment. His
co-investigators are Ulf Israelsson, Talso Chui and
Frank Gasparini at the Jet Propulsion Laboratory.
In most materials, this confinement effect
surfaces at extremely small dimensions,
thicknesses of a few atomic widths. It arises from the
fact that fundamental particles have a dual nature,
acting sometimes like solid objects and sometimes like a
packet of waves. A particle contained within a layer
that is so thin that the waves associated with it come
in contact with both edges is restricted to moving in
only two dimensions. This constraint can change the
physical properties of the material. If the particle in
question is an electron, for example, then the
electrical properties of the material are affected.
"The size of the transistors in today's integrated
circuits is about two tenths of a micron. Intel and the
other semiconductor manufacturers are talking about
reducing this by a factor of 10 or more in the next
decade," Lipa said. "That is about the size where we
expect these confinement effects to appear in metals and
semiconductors. The preliminary indications are that
this effect tends to have a depressing effect on
properties like electrical conductivity, so it looks as
if it might present a roadblock to the miniaturization
process."
Such a roadblock could have serious consequences for the
microelectronics industry. The ability to continually
miniaturize the circuitry printed on silicon chips has
been the primary reason that the industry has been able
to simultaneously reduce the cost and increase the
performance of everything from computers to telephones.
If the confinement effect proves to be relatively small,
and reduces the conductivity of silicon only slightly,
then the process of miniaturization can continue until
some other factor intervenes. If the confinement effect
is large, however, it could slow or block further size
reductions. In that case, the industry will be forced to
develop a new technology to reach smaller size scales,
Lipa said.
Scientists have several competing theories for how the
confinement effect might work, but there is little
direct evidence of its exact nature and magnitude. That
is where helium comes in. It has some unique qualities
that make it an ideal substance in which to observe this
effect. It is the only substance that remains a liquid
at absolute zero, a temperature of 273 degrees Celsius
below zero. At about 2 degrees Celsius above absolute
zero, helium becomes a superfluid, a material without
resistance to current flow.
As helium is cooled to the point where it turns from an
ordinary liquid into a superfluid, its confinement
effect increases by a factor of 10,000 or more. As the
effect increases, the distance at which helium atoms
sense boundaries increases from a few atomic widths to
thousands of atomic widths. This makes it possible for
Lipa and his colleagues to measure the confinement
effect cleanly and directly with current technology.
Lipa and his colleagues designed an experiment that
consists of more than 400 silicon wafers. The thin
wafers, which are two inches in diameter, are stacked
together in a column. The surface of each wafer contains
a micromachined recess with a depth of 50 microns, about
twice the width of a human hair. When the column is
immersed in about two cubic inches of chilled helium,
the liquid forms thin layers between the wafers.
Confinement is expected to affect a number of a
material's physical properties. The specific
manifestation that CHeX is designed to measure is its
impact on helium's heat capacity. Heat capacity is the
amount of heat it takes to raise the temperature of the
substance by a set amount. To make these measurements,
the scientists have developed some of the world's most
precise thermometers. They can measure temperature
changes in liquid helium of less than a billionth of a
degree. As a result, they can record changes in energy
as small as a fly's landing on a table.
Lambda Point Experiment
The thermometers and much of the other hardware
originally were developed for an experiment that flew on
the shuttle in 1992. Called the Lambda Point Experiment,
the original investigation determined the way in which
bulk helium's heat capacity changes as the material
makes the transition from normal to superfluid state. As
they cool the confined helium to the superfluid
transition point, the researchers expect its heat
capacity to diverge from that of the three-dimensional
helium. The direction and magnitude of that divergence
will provide them with a direct measurement of the
strength and nature of the confinement effect. The
experiment must be done in zero gravity. On earth, the
variations in pressure caused by gravity are enough to
obscure the divergence.
There are three leading theories that attempt to predict
the confinement effect: renormalization group theory by
Volker Dohm of the University of Aachen in Germany; a
Monte Carlo-based theory by Efstratios Manousakis at the
University of Florida; and a vortex ring dynamics theory
by Gary Williams at the University of California-Los
Angeles. Each makes slightly different predictions for
the size of the effect and how it varies in different
materials. The results of the CHeX experiment should
help refine these theories, Lipa said.
Conducting an experiment that has some important
economic implications has provided an added element of
excitement, the physicist acknowledged. "Several months
ago, I read a newspaper story about Intel's plan to
invest $250 million in a plant to reduce the size of the
transistors by a factor of 10. Here I've been, sitting
in the ivory tower doing esoteric science, and now these
guys are getting down to sizes that are relevant to what
we're measuring."
