On a tabletop in Kazuo Hosokawa’s lab is a small dish of water. Sprinkled on the water’s surface are a hundred or so small, flat pieces of silicon. When Hosokawa shakes the dish, some of the little pieces begin to clump together into shapes that look something like four-leaf clovers. This humble-looking experiment is a step toward an ambitious goal. It would be nice, says Hosokawa, a mechanical engineer at the Institute of Physical and Chemical Research in Saitama, Japan, to be able to mix some silicon and metal bits and have them spontaneously assemble themselves into electronic components.
Hosokawa and his colleagues at the University of Tokyo are nowhere near that goal yet. But they have made an intriguing start, using the surface tension of water to push pieces of silicon together. Each piece, which is shaped like a semicircle with a triangle jutting out on the flat side, is about .016 inch wide. The researchers’ goal was to get four of these shapes to link, forming a clover shape, but they weren’t relying on chance to get the pieces to join; they chose the shape of the pieces for a good reason.
On a small scale, the surface tension of water is a force to be reckoned with, far more powerful than gravity. It acts on the silicon pieces somewhat like static electricity, making them stick together. But the amount of force it exerts depends on the shape of the object. Surface tension tends to minimize the surface area of the water in a confined space, and one of the ways to minimize surface area is to minimize the deformation of the water surface around floating objects.
The water surface gets most deformed around sharp points, which is why Hosokawa and his colleagues designed their silicon chips with projecting triangles. The idea was that surface tension, in its effort to minimize surface distortion, would push the silicon pieces into the desired clover shape. Surface distortion would be minimized, the researchers calculated, if the rounded edges of four linked pieces faced outward, with the triangles on each piece all pointing toward a common center.
By gently shaking the silicon, Hosokawa’s team has so far gotten about 30 out of 100 pieces to spontaneously link up into clovers. Hosokawa thinks he may get better results by using pieces of several different shapes, or even by building three-dimensional structures. Making more complicated structures might reveal the unknown principles of self- assembly, says Hosokawa. In the long run, he says, self-assembling techniques would enable us to fabricate many complicated microstructures at a high rate.