Molecular Electronic Devices

Forrest L. Carter
Naval Research Laboratory, Washington, D.C.
Marcel Dekker, Inc., New York and Basel, 1983.

Chapter II
Molecular Memory and Hydrogen Bonding
R.C. Haddon and F.H. Stillinger
Bell Laboratories
Murray Hill, NJ 07974

INTRODUCTION

Characteristic sizes of electronic components in computer hrdware have displayed a relentless shrinkage with the passage of time. While 10 cm might be an appropriate length scale to describe vacuum tube equipment available in 1940, 10 mm spacings on semiconductor chips has become routine in 1980. Each passing decade thus has witnessed reduction in linear dimension by a factor of ten. By obvious extrapolation one concludes that 2020 will be "the year of the nanometer," with computing elements reduced to the size of individual molecules.

Does this kind of extrapolation make sense in useful technological terms? Can single molecules actually serve as reliable computing elements? Intriguing support for affirmative answers to these questions emerges from established fact. Certainly molecular biology demonstrates by concrete example that individual molecules (specifically DNA and RNA) can serve as information storage, replication, and transmittal media. But information processing in the regime of molecular biology is notoriously slow, perhaps having had no eveolutionary compulsion to be otherwise. Furthermore it is chemically uncertain that the class of compounds produced and selected by terrestrial chemical evolution could ever become substantially more rapid, efficient, or versatile by conservative structural modifications. Fortunately synthetic chemistry offers a much wider set of opportunities, one aspect of which we explore in this paper. Another approach, based on photochemical hole burning, has been previously explored. (1)

If indeed some part of computer technology is headed for the molecular regime it is time to turn creative intellectual effort in chemistry and physics to the question of how best that can be achieved.