Viewing crystal growth on an atomic scale.

How does a crystal grow?

Precisely which processes permit random, energetic atoms, bouncing chaotically on a surface, to organize themselves into a neatly ordered lattice has never been entirely clear.

Do the atoms roll easily into neat rows on the forming crystalline surface? Or do they bounce around randomly until they lodge in an available slot? And once an atom has settled near the edge of an emerging layer, is it attracted to, or repulsed by, the atomic forces generated there?

Such questions have led Gert Ehrlich, a materials scientist at the University of Illinois at Urbana-Champaign, and his colleagues to study the behavior of metal atoms adsorbed onto metal surfaces from low-temperature vapors. Using a field ion microscope, they have managed to observe single iridium atoms settling into position on closely packed iridium planes, revealing, among other things, a previously unseen "empty zone" near the lattice's edge, Ehrlich said last week in Boston at a meeting of the Materials Research Society.

"The appearance of this empty zone was unexpected," Ehrlich says. "It may have important implications for the growth of crystals at low temperatures."

In the standard model of crystal growth, layers of atoms form in regular, stepped planes, Ehrlich says.

In the case of metal lattices, atoms from a vapor are thought to settle onto a layered surface and diffuse. Eventually, those diffusing atoms strike a lattice step and become incorporated into the edge of that layer.

However, Ehrlich reports that observations of vaporized iridium atoms incorporating themselves into iridium planes reveal "a much more diverse" picture. Along the edge of ascending steps, an empty zone more than two atoms wide regularly emerges; diffusing atoms will not settle there. That empty zone forms, Ehrlich believes, because of attractive forces that "suck in" atoms near the step's edge.

His group's observations also show that atoms tend to diffuse to the edge of descending steps of crystalline planes, then stop.

"They won't just roll over the edge, as originally envisioned," he says. Rather, the atoms get trapped at the top of the step and remain stuck there. Only after thermal excitation do they become incorporated into the crystal structure."

Under rare conditions, an atom will burrow into the lattice by pushing nearby atoms aside, the group finds. In one case, a rhenium atom condensing atop a cluster of iridium atoms appeared to move an iridium atom out from the edge and take its place. Such atomic behavior does not fit predictably into the current model of lattice formation, Ehrlich notes.

As a result of these observations, Ehrlich believes more detailed experimental work is needed to understand how atoms become incorporated into lattice steps.

"We're developing a new view that is changing our understanding of how crystals actually grow."