Technique Tethers Molecules To Silicon With Atomic Precision

"The semiconductor industry is fast approaching a fundamental limit on how many components can be crammed onto a conventional chip," said Joseph Lyding, a UI professor of electrical and computer engineering and a researcher at the university's Beckman Institute for Advanced Science and Technology. "We are exploring what we can make at the atomic level, and how we might merge that with what we perceive as end-of-the-road silicon technology."

To selectively bind molecules to a silicon surface, Lyding and graduate students Mark Hersam and Nathan Guisinger first passivate the silicon bonds with hydrogen. Then they use an ultra-high vacuum scanning tunneling microscope to break individual silicon-hydrogen bonds and dislodge hydrogen atoms from selected sites.

"We take advantage of the difference in chemical reactivity between clean silicon and hydrogen-passivated silicon," Hersam said. "By removing individual hydrogen atoms, we create holes in the clean silicon surface. Since these holes -- or dangling bonds -- serve as effective binding sites, molecules injected in the gas phase will spontaneously self-assemble into the predefined patterns."

A technique called feedback-controlled lithography gives the patterning process an atomic precision. "Feedback-controlled lithography works by actively monitoring the microscope feedback signal and the tunneling current during patterning, and immediately terminating the patterning process when a bond is broken," Lyding said. "By operating the microscope under feedback control, a carefully controlled dose of electrons can be written along a line or over an area to locally depassivate the surface and create templates of individual dangling bonds."

The researchers demonstrated the feasibility of their technique with three organic molecules: norbornadiene, copper phthalocyanine and carbon-60 "buckyballs." One advantage of organic molecules is that their end groups can be functionalized for potential electronic or mechanical switching properties.

"Now the fun can begin as we attempt to take advantage of what chemists can make to add functionality in the form of data storage or processing," Lyding said. "It may be possible to make molecular devices and switching elements that operate at a hundred trillion times a second."

While the technology for economically tethering billions of molecular devices on a chip surface does not yet exist, "we can make small numbers of the devices and test their function," Hersam said. "This is an important step in bridging the gap between molecular electronics and silicon technology."