Circuitry in a nanowire: Novel growth method may transform chips

Peter Weiss

In a feat of nanometer-scale engineering, researchers have produced semiconductor filaments that are as thin as viruses but contain working electronic and optical devices. Alternating bands of different semiconductor materials in the superthin wires serve as the electron and photon manipulators. Someday, such striped strands may form the basis of a new type of circuitry that is far tinier, faster, and more energy efficient than conventional chips will ever be, the scientists say.

Last year, a Harvard University team led by Charles M. Lieber demonstrated nanowire-based electronic devices and rudimentary logic circuits, but those were composed of wires of uniform composition (SN: 11/10/01, p. 294: Available to subscribers at http://www.sciencenews.org/articles/20011110/fob5.asp). By crossing different nanowires over one another, the researchers made them behave as transistors and diodes.

Now, Lieber's group and two other teams—one led by Peidong Yang at the University of California, Berkeley and the other led by Lars Samuelson of Lund University in Sweden—have unveiled striped nanowires resembling submicroscopic barber poles. Each stripe has a different composition, and thereby different electronic properties.

Electrical measurements by the Harvard and Lund groups show that the junction of just two adjacent stripes within one wire can be a diode that guides electrons. In the Feb. 7 Nature, Lieber and his colleagues also report making within a single wire a type of diode that emits light. What's more, the Harvard investigators constructed a prototype, one-wire "nano-bar code" that fluoresces under green light in alternating dark and bright stripes. It's possible, they claim, to make stacks of multiple colors that would be leaner than any microscopic bar code rods created so far (SN: 10/6/01, p. 212: Available to subscribers at http://www.sciencenews.org/articles/20011006/fob2.asp). Such nano-bar codes might label and track individual proteins and other biomolecules.

The Berkeley and Lund groups each report their striped-nanowire work in the February Nano Letters. The Lund team presents further details of their approach in the Feb. 11 Applied Physics Letters.

All three groups use similar, high-temperature methods to create their striped nanowires. They start with a wafer of silicon, or another substrate, sprinkled with nanometer-scale blobs of gold. In a furnace, a vapor of a semiconductor material, such as indium phosphide, settles on and dissolves into the molten blobs. When the dissolved material reaches a sufficient concentration, it crystallizes and the blob exudes a shaft of semiconductor about the same diameter as the blob. As the shaft lengthens, the researchers change vapors, thereby producing successive stripes of different materials.

"This is an important milestone in a very fast-moving field," comments Thomas E. Mallouk of Pennsylvania State University in State College.

Samuelson notes that the three teams have partially retraced steps taken a decade ago by Japanese researchers who demonstrated diode function in nanowires with one band each of two different semiconductors.

A major impetus for making nanowires is the expectation that miniaturization of conventional microcircuits, which has continued unabated for 30 years, will soon bottom out. As the dimensions of conventional transistors and other devices shrink to only a few tens of nanometers, the laws of quantum mechanics and other effects will prevent those devices from working properly, many scientists predict.

Small-scale devices based on nanowires or their rival circuit-building components called carbon nanotubes (SN: 5/26/01, p. 335: Available to subscribers at http://www.sciencenews.org/articles/20010526/note16.asp) are among many technologies being developed as possible successors to today's circuitry (SN: 11/25/00, p. 350: http://www.sciencenews.org/articles/20001125/bob10.asp). However, now that nanowires can be grown with stripes of different semiconducting materials, the wires may pull ahead of tubes in that race, Lieber notes.

There are differences among the nanowires coming out of the Harvard, Berkeley, and Lund laboratories. The Harvard and Berkeley wires grow 100 times faster than those made by the Lund group, Samuelson notes. The slow growth of the Lund wires enables the Swedish researchers to orchestrate transitions between different semiconductor materials within just a few atomic layers, he claims. Such sharp boundaries make for superior electronic and optical properties in the resulting devices, says Samuelson.

Lieber contends that many electronic devices don't require such crisp boundaries. Yang says that the slow growth rate could make the Lund process impractical.

Despite such disputes, one advantage of all striped nanowires is that different semiconductors can be neighbors even when their crystal structures don't match very well. Because the width of the wire can vary, the semiconductor stripes have room to accommodate each other. Such adjustments are not possible in conventional chips. This advantage could make it possible to build nanowires with novel combinations of materials and, therefore, build novel devices.

The advent of striped nanowires should also open up new options for tiny circuits beyond the crossing of wires, says Harvard's Mark S. Gudiksen. For instance, if segments of a single wire can be interconnected, a striped nanowire could serve as a compact, elongated circuit. Alternatively, suggests Samuelson, forests of striped nanowires might share space with conventional semiconductor technology within hybrid chips.

********

References:

Björk, M.T., et al. 2002. One-dimensional heterostructures in semiconductor nanowhiskers. Applied Physics Letters 80(Feb. 11):1058–1060. Abstract available at http://dx.doi.org/10.1063/1.1447312.

Björk, M.T., … and L. Samuelson. 2002. One-dimensional steeplechase for electrons realized. Nano Letters 2(February):86–90. Available at http://dx.doi.org/10.1021/nl010099n.

Gudiksen, M.S., … and C.M. Lieber. 2002. Growth of nanowire superlattice structures for nanoscale photonics and electronics. Nature 415(Feb. 7):617–620. Abstract available at http://dx.doi.org/10.1038/415617a.

Lieber, C.M. 2002. Nanowire superlattices. Nano Letters. 2(February):81–82. Available at http://dx.doi.org/10.1021/nl020289d.

Wu, Y., R. Fan., and P. Yang. 2002. Block-by-block growth of single-crystalline Si/SiGe superlattice nanowires. Nano Letters 2(February):83–86. Available at http://dx.doi.org/10.1021/nl0156888.

Further Readings:

Gorman, J. 2001. Molecules get microscopic bar code labels. Science News 160(Oct. 6):212. Available to subscribers at http://www.sciencenews.org/articles/20011006/fob2.asp.

______. 2001. Nanotubes form dense transistor array. Science News 159(May 26):335. Available to subscribers at http://www.sciencenews.org/articles/20010526/note16.asp.

Sivitz, L. 2000. When the chips are down. Science News 158(Nov. 25):350–351. Available at http://www.sciencenews.org/articles/20001125/bob10.asp.

Wang, L. 2001. Getting nanowired. Science News 159(May 5):286–287. Available to subscribers at http://www.sciencenews.org/articles/20010505/bob18.asp.

Weiss, P. 2001. Wiring teensy tubes, strands into circuits. Science News 160(Nov. 10):294. Available to subscribers at http://www.sciencenews.org/articles/20011110/fob5.asp.

Sources:

Mark S. Gudiksen
Department of Chemistry and Chemical Biology
Harvard University
Cambridge, MA 02138

Charles M. Lieber
Department of Chemistry and Chemical Biology
Harvard University
Cambridge, MA 02138

Thomas E. Mallouk
Department of Chemistry
Pennsylvania State University
University Park, PA 16802

Lars Samuelson
Solid State Physics/Nanometer Consortium
Lund University
Box 118
S-221 00 Lund
Sweden

Zhong Lin Wang
Georgia Institute of Technology
Materials Science and Engineering
771 Ferst Drive, N.W.
Atlanta, GA 30332-0245

Peidong Yang
Department of Chemistry
University of Califoria, Berkeley
Berkeley, CA 94720

From Science News, Volume 161, No. 6, February 9, 2002, p. 83.