A hard step toward diamond circuitry.

A hard step toward diamond circuitry

Standard microelectronic chips work fine for everyday PC tasks such as word processing. But more demanding jobs -- such as regulating lots of current, operating in hot engines or computing at ever-faster speeds -- would reduce these silicon wonders of miniaturization to melted jumbles of atoms. That's why scientists have been eyeing melt-resistant synthetic diamond as a semiconducting material that could go where no silicon chip has gone before.

Amidst some skepticism, Ken Okano of Tokai University in Japan claimed this week that his team has taken a long-awaited step toward that goal. The step involves making "n-type" semiconducting diamond, in which impurities such as phosphorus ions carry current through a microscopic electronic gate.

In the past, researchers have synthesized thin films of only "p-type" diamond, which carries current via positively charged "holes" created by inserting atomic impurities, such as boron ions. These ions readily assemble within a growing diamond lattice. In principle, circuit makers could use such boron-doped diamond for making novel types of electronics devices. But since standard silicon-based components comprise thin layers of both n- and p-type silicon, diamond chips featuring both types of materials are preferred.

At the Second International Conference on the New Diamond and Technology, held in Crystal City, Va., Okano outlined his group's method for depositing 10-micron layers of n- and p-type semiconducting diamond onto a silicon base. The team uses a technique called chemical vapor deposition, in which a heat source such as a hot metal filament breaks hydrocarbon gas molecules into fragments. Liberated carbon atoms then systematically rearrange into diamond films as they deposit on materials such as silicon placed nearby within the deposition chamber (SN: 8/4/90, p.72).

By including certain boron or phosphorus compounds in the gas, the researchers were able to lay down both the p- and n-type diamond layers on silicon, Okano says. Moreover, their data suggest that the triple-layer sandwich behaved as a rudimentary electronic gate.

Some scientists argue that this behavior could arise from crystal imperfections, such as tiny lattice voids, rather than from the controlled implantation of phosphorus ions, which has proved so elusive in the past. And diamond isn't the only candidate for heat-resistant chips, they note. Other experimental semiconducting materials, such as boron nitride or silicon carbide, could sneak into the same market niches.