Professor Lloyd Harriott (ECE)
Smaller. Faster. Cheaper. Better. For more than forty years, the steady stream of innovation in integrated circuit design and production has transformed Gordon Moore’s prophecy — that semiconductor circuit density would double every 18 months — into Moore’s Law. The dramatic surge in computational power unleashed by this exponential growth has transformed every aspect of life. It allows us to pursue our daily routines more productively, while enabling great leaps forward in such diverse fields as healthcare, communications and transportation.
Yet we are far from satisfied. Driven by the promise of ubiquitous computing and the desire to solve ever more complex research problems, our appetite for even more efficient and powerful semiconductors has only increased. There are signs, however, that the inherent limits, both in integrated circuits themselves and in the tools we use to produce them, will force us off pace in coming years. The work of Lloyd Harriott and an interdisciplinary team of researchers from the School of Engineering and Applied Science and the College and Graduate School of Arts & Sciences is designed to give Moore’s Law a new lease on life.
There are a number of obstacles to continued exponential improvements in integrated circuits. For one thing, as transistors become smaller, they behave differently. Considerable effort has been devoted to modifying the design of these transistors to avoid the short channel effects that undermine their efficiency. At the same time, it has become increasingly difficult and expensive to wring the necessary improvements from the photolithographic process that is at the heart of semiconductor production.
Harriott and his colleagues are taking a different approach — building electronic devices out of molecules or small groups of molecules. Molecular electronics are smaller and in many cases more efficient than conventional devices, they require much less power and generate less heat, and they can be designed with completely novel characteristics. Molecular electronics also opens the way to new methods of production. As Harriott points out, “Chemists are really good at making designer molecules, and with nanofabrication techniques, we can produce self-assembling circuits.”
Harriott, chair of the Charles L. Brown Department of Electrical and Computer Engineering, is working with a team of engineers and chemists on this enterprise. His colleagues include department members John Bean, Mircea Stan, Avik Ghosh and Nathan Swami; chemical engineer Matthew Neurock; and chemist Lin Pu. Mark Reed at Yale University and James Tour at Rice University are also part of the group. One of the essential challenges they faced was developing a test bed, essentially a physical structure to enclose the molecule and enable it to function as part of a circuit. Using planar processes similar to those used for conventional semiconductor devices, they succeeded in embedding molecules in nanowells just tens of nanometers in diameter.
They have also been exploring a number of different types of organic molecules, assessing their functionality and their potential for self assembly. “One of the devices you can make with molecules possesses negative differential resistance or NDR,” says Harriott. “For such basic functions as adding numbers, NDRs are incredibly efficient.”
The U.Va. group has also succeeded in self-assembling a monolayer in a regular dense array using molecules that have a thiol group at one end of an alkane chain. “It is an important first step to be able to assemble a single, uniform molecular layer,” Harriott says.
Over the last five years, Harriott and his colleagues have used funding from the National Science Foundation and the Defense Advanced Research Projects Agency to create a world-class initiative in molecular electronics. One reason for growing interest on the part of industry as well as funding agencies is a decision that Harriott and his colleagues made to produce molecular integrated circuits that are compatible with existing devices. For instance, they use vapor phase assembly, a method that is analogous to processes currently used in integrated circuits. This enables them to avoid the oxidation that might otherwise cause conventional semiconductor metals like aluminum and copper to react with volatile organic molecules and change their properties. “Our goal is not to replace conventional technology, but to add the specialized functionality that can be created with molecular circuits,” Harriott says.
|
