System breaks individual molecular bonds.

For scientists seeking the ultimate control of matter--the ability to manipulate one atom at a time--molecular bonds have long presented a formidable obstacle.

How, researchers wonder, does one make or break a specific chemical bond? Or energize a molecule precisely enough to disrupt one bond and no others?

Tsung-Cheng Shen, a physicist at the University of Illinois at Urbana-Champaign, and his colleagues describe a method for achieving this long-sought goal. They have devised a system, using a scanning tunneling microscope, that enables them to break individual chemical bonds between silicon and hydrogen atoms. Their report appears in the June 16 Science. "This is the first time that bond breaking has been achieved with this level of accuracy," says coauthor Joseph W. Lyding, an electrical engineer at the University of Illinois. "We can actually get down to individual rows of molecules on a silicon surface and selectively remove hydrogen along those rows. That hasn't been demonstrated before."

Starting with a thin film of silicon coated with a single-atom layer of hydrogen, the researchers fine-tuned a scanning tunneling microscope to pluck single hydrogen atoms off the surface in neat rows. They achieved this delicate feat using a narrow beam of low-energy electrons, which emanate from the microscope's single-atom tip as it hovers over the silicon-hydrogen surface.

The electrons "pump" energy into the hydrogen atoms, causing them to vibrate at just the right rate to fracture the bonds holding them in place on the silicon surface.

Once the bonds have broken, the atoms settle into a stable state at room temperature, the scientists note. The technique has proved relatively simple, requires no special conditions--such as low temperatures--and moves fairly briskly, despite its exacting nature. Robert E. Walkup, a physicist at IBM's Thomas J. Watson Research Center in Yorktown Heights, N.Y., says the team has reached "the limit of control of silicon surface chemistry."

"In the long run, we'd like to make small, electronic devices on the nanometer scale," says Lyding. "Or at least pieces of devices, the simplest being a metal conductor.

"When you're making electronic structures below one-tenth of a micrometer in size, you need to use a different strategy. The conventional techniques of microlithography just don't work well at that scale."

While stressing that their results represent the earliest stages of this work, Lyding sees potential applications down the road for the manufacture of transistors, computer chips, and electronic memory devices. "We've already made extremely small features on silicon that are similar to those seen in state-of-the-art microprocessor chips."