This masterful balancing act shows how difficult atoms can be. But dealing with them is becoming a top priority in research labs. The reason is simple: as machines get smaller, the parts are smaller still, and they’re getting hard to handle. Computer chips (a kind of electrical machine) already contain wires that are only a few atoms wide. And if scientists ever figure out how to make a quantum computer, which would use atoms or molecules as computing elements, they’re going to need a way to assemble it.
A promising technique for getting atoms to do what you want is self-assembly. It involves taking advantage of the natural tendency of atoms and molecules to arrange themselves in certain ways. Life, in a sense, is an act of self-assembly: DNA, a complex molecule, naturally organizes itself in certain ways because its various components react according to laws of chemistry.
Scientists, of course, aren’t up for so grand a project as creating their own version of DNA. A more appropriate example of what they want to start with is a snowflake. It begins as a speck of dust floating through the air; water collects on the speck and freezes, and as it falls through the air its structure is shaped by subtleties of humidity, temperature and wind. In his lab at Caltech, physicist Kenneth Libbrecht practices making snowflakes, which may one day help engineers find ways of growing computer circuits. “If you want to understand crystal growth,” he says, “you have to start with ice.”
The work of Libbrecht and others is already having an impact in industrial labs. Engineers have been “growing” tiny magnetic memories and supersensitive sensors. At IBM’s Watson Research Center in New York, engineers are developing ultrasmall optical crystals in computer chips that would be able to handle laser signals. “We’re starting to understand machines that nature has made,” says Thomas Theis, physics director at Watson, “but we’re not that clever yet in dealing with complexity.” The old adage that no two snowflakes are alike speaks to the explosion of possibilities engineers face when dealing with the minuscule.