Hearing Eugene
Haller talk about his research is like reading a spy novel written
by Albert Einstein. The Materials Science and Engineering professor
studies semiconductors, the stuff used to make computer chips. But
instead of plain old silicon, Haller's semiconductors are made from
enriched isotopes - natural constituents of an element that have
the same atomic number but different atomic weights. The unique
physical properties of isotopically enriched silicon, germanium,
gallium arsenide, or other elements might lead to ultra-fast computer
chips or even quantum computers. The cloak-and-dagger part lies
in where Haller acquires his isotopes: former nuclear weapon laboratories
in Russia.
"The Russians have enormous stashes of this stuff," Haller explains.
"In the 1980s, everyone there knew that the need for the uranium
they were producing was going down because fortunately, fewer
bombs were being made."
So as the Cold War began to thaw, the Soviet scientists started
using their centrifuges to crank out stable isotopes for which
they hoped Western scientists might someday be in the market.
The Russians' foresight paid off.
As it turns out, isotopically enriched semiconductors are far
better at conducting heat than are their natural elemental kin
used in today's computer chips. Packing more transistors onto
a silicon chip increases the speed of a processor, but the faster
the transistors switch on and off, the hotter they become. That
heat issue contributes to a brick wall facing the shrinkage of
integrated circuits. But according to Haller, fabricating the
transistors from isotopically enriched silicon - with the material's
measured 60% increase in thermal conductivity - could enable chips
to run much faster without their components melting.
Haller outlined advantages of working with isotopically enriched
semiconductors in a seminal 1990 paper. He proposed research that
would only be possible if certain enriched isotopes of germanium,
the heavier cousin of silicon, became affordable. Two years later,
one of his colleagues returned from a trip to Russia with a gift
from Professor Valeri Ozhogin of the Kurchatov Institute: two
bottles, each containing 100 grams of germanium isotopes. They
were the first of a steady stream of samples received for collaborative
experiments.
During the last few years, UC Berkeley has developed a direct
and nearly exclusive line to the Russian isotope empire via a
Department of Energy-sponsored Initiatives for Proliferation Prevention
(IPP) project involving Haller's research group, the Krasnoyarsk
Electrochemical Plant in Siberia, and Isonics Corporation, which
imports isotopes from Russia. A focused effort has been launched
to create large quantities of highly enriched silicon 28, silicon
29, and silicon 30 in the hopes of commercializing isotopically
enriched silicon for broad industrial applications.
David
Pescovitz photo
Haller
explains the operation of LBNL's ion implantation machine,
which enables him to control the electrical properties of
silicon nanocrystals. (Click for larger
image.)
|
"Having sizeable quantities of all silicon isotopes on US soil is
a huge step forward for us," Haller says.
A key enabler in Haller's research is a specialized device that
the Berkeley team uses to produce its own isotopically enriched
polycrystalline silicon feedstock for research at Berkeley and
worldwide. (In the private sector, polycrystalline silicon feedstock
is used to grow bulk silicon crystals that are sliced into wafers
on which millions of electronic elements forming integrated circuits
are fabricated.) The $200,000 polysilicon deposition reactor at
Lawrence Berkeley National Laboratory had to be custom built because
commercially available reactors are designed to handle hundreds
of tons of material, far more than needed for experimentation.
"If you wanted a small quantity of specialized silicon, you
couldn't get it," Haller explains. "The private companies' products
are so standardized that they only deal in hundreds of kilos at
a time. Now we can make whatever we need in house."
Haller's latest effort with enriched silicon isotopes is to
synthesize semiconductor nanostructures and control their electrical
properties by adding impurities, a process known as doping. Today,
integrated circuit manufacturers dope transistors of 100-nanometer
dimension with millions of dopant atoms. (A nanometer is one-billionth
of a meter.) But Haller hopes to precisely control the doping
of silicon nanocrystals just 5 nanometers in size with as little
as a single dopant atom.
According to Haller, creating novel device designs that use
the Russian-supplied isotopes to control atomic spin and charge
in semiconductors could perhaps lead to the quantum bit (q-bit),
the basic unit of information in what may become an enormously
powerful computer architecture of the future.
"This story is really about politics playing right into the
hands of science," Haller says.
Eugene
Haller's home page
Lawrence Berkeley National Laboratory
Isonics Corporation
