FUNCTIONAL MOLECULAR STRUCTURES AND DEVICES

Mark Reed
Department of Electrical Engineering
Yale University
PO Box 208284
New Haven CT 06520

Within the last few years it has become possible to microfabricate matter on an unprecedented scale. Advances in nanofabrication technology now allow creation of electronic structures that exhibit manifest quantum and single-electron effects. However, proposed solid state device implementations at this level suffer from three problems:

1. Critical dimensional control difficulties. For instance, electron devices that operate in this range must operate by tunneling, since a barrier (heterostructure, oxide, or otherwise) is a prerequisite for isolation in a 3-terminal device that can exhibit gain. However, electron tunneling is exponentially sensitive to atomic-layer fluctuations in the tunneling barriers, resulting in device characteristic variations unacceptable for large scale integration. Similar effects also occur for the position of energy states, either size-quantized or charge-quantized.
2. Reduced operating temperatures. Device embodiments utilizing discrete electron charging (SETs) suffer from reduced operating temperatures for room temperature operation; 1 nm or less size junctions are required, dimensions that imply severe tunnel barrier and size fluctuation problems for solid state or metal embodiments.
3. Interconnection and alignment problems. None of the current nanofabrication approaches to solid state device implementation addresses these problems.

The generic properties of a technology that addresses the critical problems can be detailed:

• the key innovation will be the one that solves the interconnect problem
• the fabrication technology must be predominantly self-aligned, perhaps non-lithographic and self-limiting
• scaling to the atomic level and room temperature operation are desirable

Beyond ULSI, the ultimate device technology (if it exists), independent of device embodiment, must solve the interconnection problem and will be predominantly self-aligned. A technology-dependent issue is where existing ULSI will usefully end (from recent work, it is reasonable that this will occur in the 0.1 micron regime); scaling to just the 100s of Å level may not be cost-effective, as the increase in performance is marginal compared to the development costs of the technology. Thus, identifying an atomic-scale device technology seems the only approach worth the investment.

Conventional contemporary personal-sized computers utilize ca. 10⁸ microfabricated devices per instrument with a coverage of ca. 10⁷ uniquely addressable transistors per cm². Extension of microelectronics beyond these limits requires the exploration of nonconventional electronic structures that scale far beyond these limits. For example, if devices were to be based upon single molecules, using routine chemical syntheses one could prepare 6 x 10²³ (Avogadro's number) devices in a single reaction flask, hence, more devices than are presently in use by all the computational systems combined, worldwide. Equally attractive is the fact that self-assembled monolayers can permit thermodynamically-driven ordering of ~ 10¹⁴ molecules/cm², although unique addressing of such structures in a conventional manner is not presently conceivable. Thus, molecular-based systems can offer distinct advantages in density, uniformity, and potential fabrication cost.

The self-aligned spontaneous assembly of chemically synthesized interconnects, active devices, and circuits is a revolutionary approach for spontaneously assembling atomic scale electronics (Reed 1995). It attacks the interconnection and critical dimension control problems in one step and is implicitly atomic in scale. Concurrently, the approach utilizes an inherently self-aligned batch processing technique that addresses the ultimate fabrication limitations of conventional ULSI. At this point, efforts worldwide are concentrated on obtaining charge transport in and the measurement of the conductance of single organic molecules. Previous measurements have been done on molecular systems by scanning tunneling microscopes (Bumm et al. 1996; Dorogi et al. 1995; Andres et al. 1996; Nejoh 1996) in an attempt to yield detailed direct measurements of molecule conductance. Preliminary results have been reported (Reed et al. 1997) of single molecule conductance in a self-assembled structure, an important first step in creating self-assembled molecular devices. Future work should concentrate on the creation of active device structures with conducting segments (based on (-conjugated oligomers such as oligo[phenylene ethynylene]s) with transport barriers (such as CH₂ [methylene] units), to create the analogues of heterojunction bandgap-engineered devices.

References

Andres, R.P., T. Bein, M. Dorogi, S. Feng, J.I. Henderson, C.P. Kubiak, W. Mahoney, R.G. Osifchin, and R. Reifenberger. 1996. Science 272: 1323.

Bumm, L.A., et al. 1996. Science 271: 1705.

Dorogi, M., J. Gomez, R. Osifchin, R.P. Andres, and R. Reifenberger. 1995. Phys. Rev. B 52: 9071.

Nejoh, H. 1991. Nature 353: 640.

Reed, M.A. 1995. U.S. Patent 5,475,341, issued 12/12/95.

Reed, M.A., et al. 1997. Science. 278: 252.