The scanning tunneling microscope is less than 10 years old but has already
given rise to at least nine other kinds of microscopes, each designed to
collect a different kind of information from a sample. Robert Pool has provided
a nice overview of these devices, which all share the the same general mechanical
and imaging technology: samples are placed on a surface that can be moved
by the expansion and contraction of piezoelectric crystals; images are made
by scanning with a pointed probe brought very close to the sample.
The devices differ in what they measure or do to the sample. The STM sends
an electric current through the sample and, in effect, measures the electrical
resistance. The atomic force microscope measures the longitudinal force
between the probe tip and the sample (i.e., the force along the line
between probe and sample), whereas the friction force microscope measures
the transverse force. The magnetic force microscope measures the magnetic
field around the sample; the electric force microscope measures the electric
field. The scanning thermal microscope measures the temperature variations
along a sample. The optical absorption microscope does spectroscopic measurements.
The scanning ion-conductance microscope maps ion flows passing through the
sample. The scanning near-field optical microscope uses an optical probe
to make images with visible light at molecular-scale resolution. The scanning
acoustic microscope uses sound waves to image samples, permitting a view
beneath the surface. Some of these devices can be used to alter specimens
as well as to view them.
[Science 247:634-247, 9Feb90]
Rather suddenly, science finds itself able to obtain a wealth of information
at atomic or near-atomic resolution. This should bring on an avalanche of
useful structural information, much of it immediately applicable to problems
in molecular engineering. The direct impact should be felt first in biology
and materials science; and soon after in fields that depend on these, such
as biotechnology, electronics, medicine, and chemistry.
Although scanning probe devices (like the scanning tunneling microscope)
can image objects of atomic size, they lack reliable methods for positioning
and measuring samples. Researchers at the University of Tokyo are trying
to remedy the situation by using the lattices of crystals as a calibration
standard. A dual-probe scanning device is being developed in which one probe
scans the sample while the other scans a reference crystal. Rough positioning
of the sample would be performed by an impact drive mechanism in which a
piezoelectric element strikes the block holding the sample, knocking it
forward several nanometers. An impact rate of 80 impacts/sec is being investigated.
[Paper by Hideki Kawakatsu, et al., IEEE document #CH2832-4/90:197-201]
Scanning probe devices are thus being transformed from novelty items
into reliable instruments for routine research.
The Diehls-Alder reaction, discovered in the mid-1800s, is a chemical reaction
with broad application for building larger molecules from smaller ones.
In the past few years chemists have begun using it for molecular construction
on the basis of shape--to make molecules designed to look like gears, for
example. In an intriguing review of the subject, Franz H. Kohnke at Univ.
di Messina suggests that the next few years will see rapid development of
such "structure-directed synthesis," giving rise to molecules
"that look like ball bearings, beads and threads, belts, cages, chains,
chimneys, clefts, coils, collars, knots, ladders, nets, springs, stacks,
strips, washers, and wires--and concurrently and subsequently for molecules
with function--that work like abacuses, capacitors, catalysts, circuits,
clocks, conductors, dynamos, membranes, motors, nuts and bolts, resistors,
screws, semiconductors, sensors, shuttles, superconductors, and switches."
Julius Rebek and co-workers at MIT have designed and built a replicator:
a molecule that produces copies of itself, given appropriate raw materials.
The raw materials are quite specialized: one of them is a variant of adenosine
(one of the four building blocks of DNA); the other is a fluorinated ester
having features complementary to the adenosine. The ester component includes
a catalytic group that promotes bonding between adenosine and ester.
This primitive replicator contains about two hundred atoms. The initial
copies were made by ordinary chemical synthesis. When a solution of adenosine
and the fluorinated ester was seeded with replicators, the replicator molecules
paired up with adenosine and ester molecules and catalyzed bond formation
between them, forming more replicators.
[Article by I. Amato in Science News, 3Feb90: 69]
So, structures that foster their own formation need not be very complicated.
Replication appears to derive from the complementarity of the components
along with the inclusion of an appropriately situated catalytic group.
An interesting question left unanswered by this research is whether it
is easier to satisfy the conditions for replication with a single molecule
or with a set of molecules that catalyze each others' formation.
The "folding problem" for proteins is to figure out what spatial
configuration will be taken by a given chain of amino acids. Researchers
at MIT started with the repressor protein from the lambda phage (a virus
that attacks bacteria). They made numerous versions of the repressor, each
with different amino acid substitutions at one or several locations, and
tested the activity of phages containing these. It was found that at certain
locations almost any substitution can be made with little or no effect on
activity. At other locations, the protein is intolerant to any change at
all. Generally speaking, amino acids located in the core of the protein
could often be substituted, but only with other amino acids of similar (hydrophobic)
type. Surface amino acids were usually tolerant of a wider variety of substitutions
except at a few functionally important sites.
[Paper by James U. Bowie, et al. in Science 247:1306-247,
Designing proteins for given roles may require less computation than
one might think, since many different amino acid sequences can give rise
to the same functionality.
Small diamonds (3 to 5 nm in diameter) constitute about 20% by weight of
soot formed during detonations, and are also found in meteorites and in
nucleation experiments. A group of investigators has shown by thermodynamic
arguments that in this size range, diamond is as stable as graphite and
its relatives, and that extremely high pressure is not necessarily required
to make diamond.
[Paper by P. Badziag at Univ. of South Africa, et al., in Nature
Diamond is being considered as a construction material for assemblers
and other nanotechnological devices. Badziag's work supports the thesis
that small structures can be built with diamond at low pressure. It is also
comforting to know that such devices would be unlikely to disintegrate or
turn spontaneously into graphite.
Vibrations in chemical bonds occur on timescale of about 10-13
sec, and rotations in about 10-10 sec. Laser observations with
a resolution of a few tens of femtoseconds have enabled researchers to follow
these processes in detail. To study vibrations, a "pump" pulse
at a wavelength of 620 nm prepared iodine molecules by sending them to higher
energy. A probe pulse at 310 nm then sent these excited molecules into a
still higher energy state from which they decayed by fluorescence. By varying
the delay between the pump and probe pulses, the fluorescence intensity
was seen to vary as the chemical bond stretched and contracted.
Rotations were followed with pulses of polarized laser light. Only those
molecules with axes aligned with electric field of the pulse were excited.
As the delay between pump and probe pulses was increased, the observed fluorescence
declined since the molecules' axes were rotating out of alignment with the
probe's electric field. The results closely matched the predictions made
from quantum theory.
[Paper by M. Dantus at Cal Tech, et al., in Nature 343:737-739,
Assemblers and other nanomachines will have to deal with vibrational
and rotational effects every time a bond is made or broken. Polarized femtosecond
spectroscopy has now provided a direct source of information about these
Electrons behave more like particles when free to move in regions much larger
than their wavelength, and more like waves when confined into regions comparable
to their wavelength. Fabrication techniques now being developed will enable
electronics to exploit the wave properties of electrons.
Today's most advanced commercial transistors have features as small as .75
micrometers (a hundredth of the diameter of a hair); quantum effect devices
being experimented with today are substantially smaller, having features
about 25 nanometers (about 100 atoms) across--and researchers are aiming
for 10 nanometers in the near future. Devices of this scale can be made
by x-ray lithography, analogous to the photolithographic technique that
the electronics industry has relied on for a quarter century.
To study electron wave effects, researchers have etched electron waveguides
about 30 nanometers across into a block of aluminum gallium arsenide (AlGaAs).
Transistor-like behavior can be obtained from a block of AlGaAs after etching
a rectangular pattern of holes.
One intriguing way to utilize quantum devices would be to arrange them as
arrays of "quantum dots" on a surface, each dot storing a small
amount of information and interacting with its neighbors according to prescribed
rules. Such arrays, called "cellular automata," can perform computations
without a network of wires to shunt information around. Arrays of 200 million
quantum dots per cm2 have been made, but are not yet programmable.
[Review article by Henry I. Smith at MIT, et al., in Technology
While this kind of electronics exploits only a few basic quantum mechanical
effects in simple crystals, it nevertheless promises major advances in speed
and miniaturization. Even greater improvements will come from more sophisticated
materials, in which complex structures process information at the molecular
level. The following two reports provide hints of what is to come.
Molecular sieves are lattice structures
containing regular patterns of "host" cavities--channels, cups,
or cages. A variety of molecular sieves can be made, each with its own characteristic
size, shape, and pattern of host cavities.
The host cavities can be used to trap and hold "guests"--small
molecules or clusters of semiconductor atoms--forcing them into regular
arrays called "superlattices" and constraining their internal
Molecular sieves are attracting the interest of researchers in a number
of fields because they provide new ways to control and obtain information
about guest molecules. Semiconductor scientists are interested in them because
confinement and regular spacing of clusters of semiconductor atoms gives
rise to quantum phenomena not seen in ordinary crystals or solutions of
the same substances.
Superlattices should find early application in optoelectronics because of
the ease with which their optical properties can be manipulated. Some superlattices,
for example, undergo color changes as the temperature is varied; others
respond to pressure, humidity, light, pH, or electric fields.
[Review article by Galen D. Stucky at UC Santa Barbara, et al., in
Science 247:669-678, 9Feb90]
Since superlattices are generalizations of the crystalline state, they
have a wider range of bulk properties than conventional crystals do, and
offer more opportunities for the control of these properties. They are early
examples of the kind of atomically precise manmade materials we will see
more of when molecular assembly machines become available.
Ordinary photosynthesis takes place within a complex structure embedded
in a membrane of a plant or bacterium. In this process a photon is captured
by "antenna" molecules and its energy transferred to a pigment
molecule where it is absorbed by an electron. The electron moves quickly
to nearby quinone molecules, leaving behind a positive charge. From the
quinone the electron passes along a chain of other structures to the outer
side of the membrane and is transferred to other molecules. Meanwhile, the
positive charge left behind is passed to the inner side of the membrane
where it is neutralized by an electron pulled from a suitable electron donor.
In plants, the electron donors are water molecules which are converted to
oxygen and protons; the protons are used in the manufacture of adenosine
triphosphate (ATP), the basic energy coinage of the biological realm.
At Arizona State University researchers have made a vastly simplified version
of a significant part of this photosynthetic apparatus. They have designed
and synthesized an impressive-looking molecule of approximately two hundred
atoms that can absorb photons, transfer the energy to electrons, and send
the electrons down one arm of the molecule and the positively charged "hole"
down the other arm. The charge-separated state has a lifetime of about 55
microseconds and preserves 83% of the original photon energy.
[Paper by Devens Gust, et al., in Science 248:199-201,
The design of this artificial photosynthetic device was no exercise in
trial and error--it proceeded from a detailed understanding of how energy
is transferred between quantum states of molecules. The methods used by
Gust's team provide a glimpse of the awesome capabilities that will soon
be routinely available to molecular engineers.