Advanced Chip Sparks Electric Car

Jian H. Zhao is developing a technology that could radically improve the efficiency of electric vehicles in the future, but to get where he is today he had to solve a problem that has been confounding scientists and engineers for 35 years.

The heart of an electric car is a computer chip that can convert the direct current from a battery to alternating current to run the motor and at the same time change the current to different frequencies to run everything from power windows to the brake system.

Prototype vehicles rely on silicon semiconductors to do that, but they have severe limitations, such as requiring a substantial cooling system that takes up space, adds weight and consumes power.

Zhao, associate professor of electrical and computer engineering at Rutgers University in New Brunswick, N.J., is at the leading edge of research into a new type of chip fabricated from silicon carbide. His lab has produced a prototype of a silicon carbide chip that “promises to increase power density by at least 100 times compared to a similarly sized silicon chip” and at the same time “reduce power consumption by 100 times.”

Think of his device as “like the chip that runs your PC,” Zhao said. The more efficient the chip, the more efficient the computer because a better chip is better equipped to allocate resources and channel energies.

The ideal chip for an electric vehicle, Zhao said, would operate at several times the temperature at which silicon chips begin to malfunction, thus eliminating the need for the cooling system. And it must help make electric vehicles lighter, more powerful and run longer distances without recharging.

The ideal material is silicon carbide, a synthetically produced substance that occurs naturally only in some meteorites. Almost as hard as diamond, silicon carbide is durable and can withstand extremely high temperatures.

But how do you etch something nearly as hard as diamond to produce a three-dimensional chip with contact points where you can attach electrical leads?

“It’s very difficult,” said Zhao, who is leading the research effort for the Electric Vehicle Consortium, or Electricore, a nonprofit association of private-sector companies, universities and organizations pioneering the development of advanced electric vehicle technologies. Congress recently authorized $300 million to subsidize the research over the next six years, and Zhao’s lab has received $2.2 million in the last three years.

Manufacturers of silicon chips use a “wet chemical” technique to etch their chips. “You take a chip and dip it into a chemical solution and it will etch the pattern for you,” leaving, for example, “a raised circular island where you would put a contact,” Zhao said.

But that process won’t work for silicon carbide. “How do you etch something that is nearly as hard as diamond?” Zhao asked. “That has been studied for 35 years with no solution.”

Until now, Zhao said. His lab has a patent pending for a technique called ion implantation. It uses an accelerator to bombard the silicon carbide with ions, or electrically charged atoms.

“The ions bombard the silicon carbide, knocking out silicon and carbon atoms,” thus etching out the desired three-dimensional structure on an atom-by-atom basis.

“It’s a very difficult process, but we have it under control,” Zhao said.

The technology has been used to create some silicon carbide chips, but they are still in the test phase. But Zhao said he sees no major hurdles in moving from the laboratory to the manufacturing stage.

The stakes, he added, are enormous. “There is a global race to develop silicon carbide devices, and countries like Sweden, Japan and Germany are all heavily funding their own research efforts,” he said. “All the major automobile manufacturers want to be the first to use it in their cars.”

The Electric Power Research Institute, which represents all major utilities, has also joined in the effort, because silicon carbide chips could significantly reduce problems with power overloads, outages, brownouts and delays in relay switching. Zhao said the technology could make it possible to shrink transformers that are now the size of a city block to something about the size of a minivan.

He expects the technology to be ready for the production line in about two years.

*

Lee Dye can be reached via e-mail at leedye@compuserve.com.