include "../../includes/header.php"; ?>
A publication of the Foresight Institute
Progress in developing nanotechnology takes place across a broad front. I find it convenient to factor this progress into several components:
Let us look at a few recent technological developments with these categories in mind.
One approach to nanosensing is provided by microscopy. Here we
are concerned solely with structure, rather than motions, since
molecular motion is far too rapid to be resolved with microscopy
techniques. A new instrument, the atomic force microscope (AFM),
recently made its debut by producing images with a resolution
smaller than 0.5 nanometers. The AFM, though related in a general
way to the scanning tunneling microscope, is not limited to
conductive or semiconductive specimens. Initial tests have used
polymerized monolayers of a simple organic compound. These are
dry, fairly rigid specimens. While the developers are confident
that the AFM will operate on samples in fluids, it remains to be
seen whether loose molecules, such as biological specimens in
their natural state, can be used.
The AFM "scans" a specimen by dragging a diamond stylus over it very lightly. The stylus and the cantilever that holds it are deflected as they move across the specimen (it is actually the specimen that does the moving). This deflection, sensed by optical means or by measuring the tunneling current between the cantilever and a platinum-iridium point, is maintained at a constant level by raising and lowering the specimen using a piezoelectric actuator. A record of the voltage required for this specimen movement gives rise to a picture. The AFM is being developed by O. Marti, H.O. Ribi, and others at Stanford and the Univ. of Calif. at Santa Barbara (Science, 1Jan88, p50; Sci News, 9Jan88, p25).
[A web primer on atomic force microscopy] [A few AFM images of DNA (not atomic resolution)]
While microscopy pursues spatial resolution at the expense of time resolution, laser spectroscopy does just the opposite. Ahmed Zewail's team at Caltech has been probing chemical reactions with laser pulses of femtosecond duration. The reactant(s), sprayed into an evacuated chamber, are exposed to a laser pulse of appropriate frequency to initiate the desired reaction; a subsequent pulse of a different frequency elicits a fluorescence response containing detailed information about the reaction mechanism. By varying frequency and timing, the researchers can obtain a series of "snapshots" of a chemical reaction from which they have deduced the precise movements of individual atoms as they go through the transition states of the reaction.
This new "femtosecond chemistry" provides the same kind of information we may someday obtain from nanosensors, but does so by carefully combining bulk technology and chemical theory. From this work will emerge an exact understanding of the forces and motions experienced by atoms and molecules when they interact, and this knowledge should play an important role in the design of assemblers. Femtosecond light pulses might also form the basis of a downscale communication channel (Science, 11Dec87, p1512).
Atomic-scale mapping of the structures of existing enzymes and
other biological nanomachines is an essential part of learning
how to design new ones. Such mapping has depended largely on
X-ray diffraction techniques applied to crystalline samples of
the materials. But getting the materials to crystallize into a
usable form has often proved difficult or impossible.
Alexander McPherson of the Univ. of Calif. at Riverside and Paul Schlichta of JPL now report that the surfaces of some minerals can greatly facilitate the crystallization of proteins and, in some cases, even cause the proteins to crystallize in forms better suited to X-ray diffraction mapping than their usual ones. The technique relies on the ability of the mineral's crystal lattice to influence the deposition and spacing of protein molecules as they deposit onto a mineral face from a supersaturated solution (Science, 22Jan88, p385).
A recent report on enzymatic catalysis in supercritical carbon dioxide should remind us that nanomachinery can be designed for operation in nonaqueous environments. In fact, many biological nanomachines already do operate at least partially in such environments: enzymes that make their homes in membranes are examples. In principle, just about any fluid should be able to host properly designed nanomachines. And, as T.W. Randolph and collaborators at the Univ. of Calif. at Berkeley point out, nonaqueous solvents offer higher solubilities for the compounds that certain enzymes operate upon. They also may provide an escape from kinetic or equilibrium restraints imposed by the use of water.
Randolph's experiments on the enzyme cholesterol oxidase are conducted in an environment of carbon dioxide at a variety of temperatures and pressures near the critical point of that solvent. The enzyme functions under such conditions, and its performance is improved by the addition of certain cosolvents such as tert-butyl alcohol (Science, 22Jan88, p389).
An upsurge of interest in ribosomes is underway, thanks to the discovery of RNA enzymes and to the application of cloning and sequencing techniques to ribosomal RNAs. Ribosomes--the molecular devices that fabricate proteins from genetic instructions--are nature's best approximation to our notion of an assembler. Some think they preserve the basic structure and function that primordial replicators must have possessed before the evolution of cellular organisms. In this view, the ribosome is the central actor in the biological drama; everything else is a set of supporting actors, props and assistants.
Investigators studying ribosomes have sequenced all of the 50+ ribosomal proteins of E. coli, and all three of the ribosomal RNAs. The arrangement of the proteins in the smaller of the two ribosomal subunits is now known, as well as the position of 60% of the corresponding RNAs. The secondary structure (that is, self-pairing) of the RNAs has been worked out, and experiments are already being performed to study the effects of sequence changes on the functions of ribosomal RNA. A great deal is known about the mechanics of ribosomal translation, but not yet at the nanometer scale (Nature, 21Jan88, p223; Science, 4Dec87, p1403).
[The RNA World monograph]
The immune system provides us with a set of molecular devices that are easily transformed into nanoeffectors: these are the antibodies. The trick, as described by Richard Lerner, et al., of Scripps Clinic, is to choose a stable molecule that resembles in form an unstable, high-energy transitional state of the chemical reaction one wants to catalyze. This choice depends upon having a detailed theoretical understanding of the reaction in question. The molecular "stand-in" is injected into an animal, where it elicits antibodies. Among these antibodies are some that bind not only to the "stand-in" but also to the transitional state of the desired chemical reaction. Using the antibodies in the presence of the reaction's precursors lowers the activation energy for the reaction--which is the essence of catalysis.
[An example of antigen design to obtain catalytic antibodies]
This technique is limited to those reactions which are well understood, but may provide powerful tools for breaking nucleic acids and proteins at specific sites or for linking them together in specific patterns (Sci. Am., Mar88, p58).
A molecular-based transistor being developed by Mark S.
Wrighton, Tracy T. Jones, and Oliver M. Chyan at MIT links the
sensitivity and selectivity of certain "redox" polymers
to the signal-carrying abilities of electronics. A polymer bridge
replaces the gate electrode found in traditional transistors; the
polymer changes conductivity in response to environmental
conditions, such as pH, thereby causing this device to act as a
sensor. The polymer bridge is about 50 nanometers across. To
develop sensors for other purposes would entail the substitution
of different polymers, each tailored for a specific task. Their
size notwithstanding, these devices are not nanosensors: they are
fabricated by bulk-technological methods, and they are not
intended to provide data on the activities of individual atoms.
They do, however, demonstrate the upscale transport of
information from fairly deep in the microworld (Sci News,
[Recent achievement: a functioning molecular wire]
To predict the outcome of chemical experiments before
performing them has long been a goal of theoretical chemists.
This difficult problem requires large amounts of computation, but
promises to make chemical research a faster and more productive
endeavor. Applied to the development of nanotechnology, it will
make the difference between being able to design
nanomachines and having to construct them by trial-and-error.
An indication of progress along these lines is the work of W. Koch and collaborators, who have predicted from purely theoretical considerations that helium should be able to combine with beryllium oxide to form the molecule HeBeO, and that HeBeO should be stable with respect to dissociation back into helium and BeO (Nature, 11Feb88, p487).
The next revolution in electronic miniaturization is being pursued at both industrial and academic laboratories today, suggests Robert T. Bate of Texas Instruments. Theoretical work on quantum-effect devices shows that a 100-fold reduction in (linear) size of electronic components should be achievable, with corresponding improvements in speed. Reliability would markedly increase as well, thanks to the stability of quantum phenomena as compared with today's larger devices that operate according to noisier principles. The quantum-effect device is based on controlling the "tunneling" current in 20-nanometer thick layers of doped AlGaAs. A small voltage applied to one terminal of the device shifts quantum energy levels there so that they match levels in another part of the device; this allows electrons to tunnel through the intervening layer. A very small voltage change can eliminate the tunneling current. A working prototype of a quantum-effect device is still a year or more away. The theory is well understood, but formidable fabrication problems remain to be solved since the devices will (initially) be made by bulk methods rather than nanotechnological ones (Sci. Am., Mar88, p96).
[A Web page about quantum dots]
By studying small clusters of atoms, K. Rademann, C. Brechignac, and others are learning how the atomic and molecular properties of substances scale up into the properties of bulk materials. For example, the ionization potential for mercury scales up into what is known as the "work function," as the number of atoms in a cluster grows from about 5 to about 70. Such transformations of atomic properties into their corresponding bulk properties, when examined in detail, reveal atomic and molecular structure and behavior that has been inaccessible until now. Most work so far has dealt with metals; the results cannot be extrapolated to organic materials since the bonding is quite different (Nature, 14Jan88, p116).
[A web page on clusters]
Dr. Mills has a degree in biophysics and runs a business in Palo Alto, California. He is also the designer and assembler of issues 2 and 3 of Foresight Update.
Many of you have been sending in journal articles and magazine clippings relevant to nanotechnology, major AI advances, hypertext publishing, and other topics of interest to FI. To ensure broader coverage and avoid duplication, Jerry Fass will be coordinating these efforts. If you would like to participate, please contact Jerry to discuss which journal(s) or magazine(s) you would like to monitor: 2975 N. Oakland Ave., Milwaukee, WI 53211; phone 414-332-6387.
From Foresight Update 3, originally published 30 April 1988.
Foresight thanks Dave Kilbridge for converting Update 3 to html for this web page.