Scientists Propose Growing Better Semiconductor Crystals In Space

Scientists and engineers who are trying to develop semiconductor "alloy crystals" -- special blends of germanium and silicon -- have a big problem on their hands. The crystals possess highly desirable thermoelectric and electro-optic properties, but they are nearly impossible to grow on Earth because of the effects of gravity.

"Germanium is about three times heavier than silicon, so it generally sinks to the bottom of the melt in the crucible, destroying the desired homogenous concentration in the crystal," said John Walker, a professor of mechanical and industrial engineering at the University of Illinois. "On Earth, gravity also presses the liquid against the walls of the container, resulting in the formation of numerous faults, dislocations and contact stresses in the growing crystal."

In the absence of gravity, however, the ingredients don't separate as readily and the molten material tends to pull away from the container shortly before solidifying, Walker said. "In experiments performed on the space shuttle, this 'detached growth' process has produced much better crystals."

Walker is working with scientists at Marshall Space Flight Center -- NASA's lead center for microgravity research in materials science -- to explore physical processes in space that are difficult to study on Earth. The group has proposed growing alloy crystals on the International Space Station.

"The pencil-thin crystals would be grown in special ampuls within magnetic damping furnaces on the space station," Walker said. "The magnetic fields would act as a brake, suppressing all movement in the molten material and thereby preventing the mixture from separating."

Walker has developed models to help optimize the benefits of using magnetic fields to control crystal growth. He also has devised methods for determining the distributions of the magnetic field and the electromagnetic force at different frequencies, and their effects on the melt motion.

"These crystals take up to 14 days to grow," Walker said. "It's a very slow and delicate process that must be precisely monitored and controlled. By controlling the melt motion with an externally applied magnetic field, we can produce a uniform distribution in the crystal."

While growing crystals in space will probably never be commercially viable, Walker and his colleagues hope to show that space-grown crystals consistently create better semiconductor materials.

"Then, once we understand the fundamental materials science, we can search for a way to reliably reproduce these crystals on Earth, in the presence of gravity," he said.

Walker will present his results at the NASA Microgravity Materials Science Conference, to be held June 6-8, at the Marshall Space Flight Center in Huntsville, Ala.