Foresight Update 16

page 2

A publication of the Foresight Institute

Recent Progress: Steps Toward Nanotechnology

by Russell Mills

Molecular construction sets

dendrimers

Dendritic polymers -- also called "dendrimers" -- are molecules that branch successively from a central core. They have engaged the interest of chemists in recent years because, being a generalization of traditional polymers, they are a huge class of chemical substances that could have valuable new properties. And, being polymeric, a given dendrimer can be built up from a small number of kinds of simple components (monomers) through repetition of a few basic chemical reactions. Traditional polymers are usually simple chains because they are made of monomers having just two reactive sites. Dendrimers, however, are made of "branched" monomers having three or more reactive sites.

Until now chemists have sought to synthesize dendrimers by starting at the core of the molecule and working outward. Monomers are bonded to the growing molecular surface in layers until the branching chains become too dense for further layers to be added. At least that was what was supposed to happen. In practice there were problems that led to inaccuracies and low yields of pure dendrimers with the desired structure.

Now C.J. Hawker and J.M.J. Frechet of Cornell University have developed a versatile new approach to dendrimer synthesis in which the molecules are built from the outside inward. This is accomplished by fabricating the molecule out of several wedge-shaped sections; the sections themselves are built starting from the outside of each wedge and working toward the apex. The sections, when completed, are assembled onto a core molecule with the apices facing inward. Not only does this method have the advantage of providing better yields and purity with less effort, it also permits dendrimers to be put together from sections of quite different structure -- for example, sections bearing different functional groups on their surfaces. [Nature 362:18-19,4Mar93]

Dendrimers possess some of the properties we expect of nanodevices: they are eutactic (i.e., have atomically precise structure), are relatively large (about 10⁵ atoms), and can have nontrivial functionality. They may play a role in developing nanotechnology: for example, they could be worth investigating as tips for atomic force microscopes, or as platforms for attaching functional protein and nucleic acid structures to make multifunctional nanosystems.

It is even possible that dendrimers represent the beginnings of a new approach to nanotechnology. A suitably designed set of dendrimers might serve as a set of structural components for building larger devices, much as the parts of an Erector set do.

carborods

More news on the construction set front: a new family of polymers called "carborods" has been developed by M.F. Hawthorne and J. Michl and their coworkers. Carborods are chains of para-carborane monomers -- molecules consisting of a cage of boron atoms with a carbon atom at either end. Carbon-carbon bonds form the links between the monomers.

The researchers report that carborods have exceptional chemical and thermal stability and are transparent to ultraviolet and visible light. For geometrical reasons, they are stiffer than other types of polymers. Carborane chemistry provides more control over length and end-group functionality than is usually the case for polymeric materials. [Nature 361:118-119,14Jan93]

For some of us, at least, it comes as a surprise that boron might serve better than carbon as the basic element in nanosystems. A systematic study could turn up still more possibilities.

Another type of cage molecule that has been in the news a lot lately is the fullerene family. These generally have a tubular framework of carbon atoms capped at both ends. (In the special case of a tubule of zero length, the molecule consists only of the end-caps and is called C₆₀ or buckminsterfullerene.) Chemists have been exploring the properties of fullerenes in hopes of turning them into something useful.

fullerene nanotubes

Last year J.Q. Broughton and M.R. Pederson, two physicists at the U.S. Naval Research Laboratory, did a computer simulation of fullerene tubules in the presence of various liquids. They concluded that capillary action ought to cause uncapped tubules to draw some kinds of molecules into their interiors. [Science News 14Nov92:327]
Broughton and Pederson's prediction has been confirmed in a report by P.M. Ajayan and Sumio Iijima of NEC. They found that when lead particles were deposited on fullerene tubules and the tubules were then annealed in air at 400 degrees C, the tubules would sometimes open at the end and suck up some molten lead. Electron microscope examination showed such tubules undamaged and solidly filled with lead for considerable distances. The researchers speculate about using the tubules as molds to make nanowires. [Nature 361:333-334,28Jan93]

From the standpoint of nanotechnology, the instability of the end-caps relative to the cylindrical part of the tubules is of interest. Unlike the cylindrical body of a tubule, the end-caps contain a number of five-membered rings of carbon atoms which apparently make the structure less stable.

A page on carbon nanotubes

Foresight Update 16 - Table of Contents

Rotaxanes

Researchers working with Fraser Stoddart at Birmingham University have made a molecular structure in which a ring-shaped molecule is confined to a molecular "string" having enlarged ends (to keep the ring from coming off). Two "stations" are built into the string; the ring randomly shuttles back and forth between these stations several hundred times per second. The group hopes soon to be able to gain electrochemical control over this shuttling motion, the ultimate aim being to develop molecular-sized switching devices. [Science 259:890-892,12Feb93]

Foresight Update 16 - Table of Contents

Catalysis

Nanotechnology evokes images of tiny robots building things out of atoms, and that seems like a radical departure from the past. However, nanotechnology can equally well be viewed as an improved version of a familiar chemical tool: catalysis. The basic underlying process in both cases is the act of forming chemical bonds by guiding, orienting, and distorting the reacting molecular groups. Catalysts are molecular manipulators -- primitive predecessors of tomorrow's molecular assemblers.

The search for new catalysts is a central part of chemistry. If only we had a good specific catalyst for each chemical reaction we want to carry out, chemistry would be simple and nanotechnology would be a straightforward development. Unfortunately, existing catalysts are far too few and their specificities leave much to be desired. Even enzymes, biology's catalysts, will generally act upon a range of different molecules rather than on a single type.

Catalysts vary greatly in size, from hydrogen ions to enzymes consisting of thousands of atoms. As one would expect, there is a correlation between size and specificity. The smallest catalysts, being single atoms, have few appendages with which to manipulate their substrates: they have only a few fuzzy lobes of electric charge -- their electron clouds. Large catalysts, in contrast, have complex shapes that trap and hold their substrates; they bring groups of atoms to bear that probe and distort the substrates, enhancing their reactivity; many catalysts even have moving parts that drive the substrates along optimum paths during the reaction.

Current work focuses mainly on catalytic molecules in two size ranges -- small and large.

Catalysts in the small size range tend to be recognizably traditional in nature: they are usually based on metal atoms and are not large enough to surround their substrates. A good example is the recent work at Du Pont by William A. Nugent and coworkers, who have developed a series of catalysts up to about 90 atoms in size, containing an atom of rhodium. These catalysts can be used in the stereospecific synthesis of amino acids from non-stereospecific starting materials. [Science 259:479-483,22Jan93]
Catalysts in the large size range include modified enzymes with specificities different from the original enzyme, and modified antibodies that catalyze chemical reactions. The latter, known as "catalytic antibodies", are illustrated by an advance made by Kim D. Janda, and others, at Scripps Research Institute. These workers have developed a catalytic antibody that causes certain alcohols to undergo a transformation that is not energetically favored. Without this catalyst, the alcohols will form a six-membered ring; with it, they form a five-membered ring instead -- a structure found in many important natural products. [Science 259:469-470,22Jan93; Science 259:490-493,22Jan93]

How is this result to be explained? After all, generations of chemistry students have been taught that catalysts accelerate chemical reactions but do not change their eventual outcome -- i.e., the relative proportions of the various possible products and byproducts. At first take, this seems to imply that catalysts should be helpless to selectively enhance or suppress one reaction path over another, for a given set of reactants.

In the long run (which might mean billions of years), a set of reactants will indeed interact to give the most energetically favored set of products -- i.e., the set with the lowest overall energy -- with or without catalysis. But in the short run, matters are different. Catalysts can create new reaction paths to products otherwise impeded by high activation energy barriers. In effect, they can selectively lower these barriers while raising barriers along the paths that otherwise would be favored by low activation energies. If the products are removed from the system before equilibrium is reached, a disfavored reaction will have become the favored one. This is not to say that the catalysts make the reactions "run uphill" from lower to higher energy, but rather that the reaction path promoted by a given catalyst need not be the one that originally had the lowest activation energy barrier.

Greater selectivity generally requires more intricate catalysts. Couched in mechanical terms: objects with lots of springs and gears can't be assembled with a hammer. A molecular assembly robot represents the ultimate catalyst because it is intricate enough to perform the manipulations needed to assemble a complicated product.

Foresight Update 16 - Table of Contents

Imaging

Magnetic Resonance Force Microscope

About a year ago, J.A. Sidles of the University of Washington proposed a method for using magnetic resonance to detect and locate single protons on a surface. Now he and D. Rugar and C.C. Yannoni of IBM have built a device to prove the validity of the basic concept. In this initial effort their goals were modest: to detect electrons rather than protons, and to measure position with low resolution in only one dimension. These goals were were achieved last fall.

In the device, a few grains of a crystalline sample are glued to a small cantilever. The cantilever is then subjected to the combined effects of three different magnetic fields: a field A varying strongly in space, a field B oscillating at the resonant frequency of the cantilever, and a field C to excite the electron spins in the sample. The forces generated by these fields cause the cantilever to vibrate at its resonant frequency. The strength of the field B is systematically varied while the vibrational amplitude of the cantilever is measured optically. In a plot of the amplitude versus field strength, peaks occur where there is resonance with the electron spins. Because the field A is spatially inhomogeneous, different parts of the sample may be subject to different field strengths and may resonate at different values of field B. Thus, the peaks in the resonance curve can be interpreted as images of these different parts of the sample.

The researchers believe that these techniques can be used to build magnetic resonance microscopes capable of forming images of molecules in three dimensions at atomic resolution. [Science News 27Mar93:199; Nature 360:563-566,10Dec92]

News from March 1997 ; Pictures

AFM imaging of actin molecules in living cells

E. Henderson, et al., at Iowa State University have imaged actin filaments in living cells with an atomic force microscope (AFM). Glial cells adhering to a coated glass surface have thin flat borders in which actin filaments lie close to the cell membrane. It unclear whether the AFM was detecting bulges in the membrane caused by the underlying actin filaments or was penetrating the surface and imaging the filaments directly. [Science 257:1944-1946,25Sep92 --MEDLINE Abstract]

AFM imaging of soft samples

AFMs are routinely used to image hard samples with atomic resolution, but soft samples are often problematic. The mechanism of soft sample imaging has been studied by M. Radmacher, and others, at the Technische Universität München. Their analysis helps to explain why AFM images of soft materials have lower resolution and spurious contrast. For example, when an AFM tip crosses a region of higher friction, the resulting deflection of the AFM cantilever will produce an erroneous value for the height of the region. However, the authors show that these effects give AFMs the ability to measure local properties of surfaces, such as viscosity and elasticity. [Science 257:1900-1905,25Sep92 --MEDLINE Abstract]

Russell Mills is research director at a company in California.

Foresight Update 16 - Table of Contents

Media Watch

Due to lack of space in this issue, coverage of media articles will be postponed. Briefly, nanotechnology has been discussed in a myriad publications including Technology Review, Nature, Popular Science, Chemical & Engineering News, New York Times, World Monitor, Government Technology, Utne Reader, Hemispheres (United Airlines), Develop: Apple Technical Journal, Pittsburgh Post-Gazette. As usual, the level of accuracy varied tremendously.

Foresight Update 16 - Table of Contents

Drexler Interview: Japan Launches
New Nanotechnology Research Center

Foresight Update interviewed Chairman Eric Drexler on the subject of his spring visit to Japan. Watch for further news after his next trip this fall.

Update: What was the purpose of your most recent trip to Japan?

Drexler: MITI organized a meeting to launch its new nanotechnology research program and invited me as the foreign keynote speaker for the molecular machine workshop. This was the third time a Japanese research institution brought me to Japan and the second time by MITI. I enjoyed the trip in part because it was an opportunity to find out where Japanese researchers are going.

Update: Tell us a little bit about the meeting and the bigger picture of what MITI is doing in nanotechnology.

Drexler: The meeting had three sections lasting a total of five days. The first concentrated on technologies relevant to mechanosynthesis, primarily the use of scanning tunneling microscopes to manipulate atoms and molecules and the use of quantum chemistry programs to model the behavior of reacting atoms and molecules.

The second concentrated on experiments and experimental techniques for characterizing a class of nanometer scale objects made with present technologies, termed "clusters." The third was divided into two sections; one on biological processes that lay down solid materials, called "biocrystallization," and the other concentrating on molecular machine systems.

This series of meetings mixed talks from Japanese and foreign researchers including several Americans and several Europeans.

Since mechanosynthesis and molecular machine systems are basic to molecular manufacturing, I found the combinations of these two topics in one meeting quite striking. US research into molecular machines systems -- at least, research with an engineering orientation -- is almost absent.

The broader context of this meeting was the inauguration of MITI's ten-year $200 million nanotechnology research program. This will be housed in a building just a short walk away from this meeting site, on the campus of the Agency for Industrial Science and Technology, a large complex of MITI research laboratories at the Tsukuba science city north of Tokyo.

Update: Is your new book, Nanosystems, available in Japan yet? Did you find that researchers are familiar with those concepts?

Drexler: Researchers in Japan seem quite familiar with the concept of molecular manufacturing, at least, relative to researchers in the United States. Maybe it was just an excess of Japanese politeness, but I was repeatedly told by different researchers that efforts including the Protein Engineering Research Institute in Osaka and the new MITI initiative were in significant measure inspired by my papers, and books published during the 1980s.

Nanosystems at that time had been on the US market for only a few months, and I saw no evidence that it had yet reached Japan, though one researcher had ordered it through his company's technical awareness service, which had brought it to his attention.

In Books Sanseido, a large bookstore in downtown Tokyo, I found several new nanotechnology titles in Japanese on the same shelves with the Japanese translation of Engines of Creation.

Update: Did you visit any other laboratories on this trip?

Drexler: I visited several groups and gave a number of talks, but the most interesting laboratory visit was at the Aono Atomcraft Project, which has its main research laboratory in Tsukuba. They are working to use a scanning tunneling microscope to lay down individual metal atoms on a semiconductor surface, with the goal of making quantum wires and a four-point probe for measuring the resistance of these wires. They have a novel and promising approach that they liken to dipping a quill pen in ink and then writing with it. This should be tested experimentally sometime this year. Their laboratory has already succeeded in removing individual silicon atoms and moving silicon atoms on a surface with considerable control.

After the workshop I also went on to Matsushita, a company that makes everything from washing machines to next-generation computer chips. After a reorganization this fall, two of their six R&D divisions will be working in areas on the path to nanotechnology, One is aimed at the manipulation of atoms and molecules, and the other has a biological focus. When I asked for more details on the latter, I was told that it would work with molecular machines, such as the bacterial flagellar motor.

Foresight member Tom McKendree, who is also president of the National Space Society's Molecular Manufacturing Shortcut Group, will be in Tsukuba this summer as part of an NSF program, I expect he can give us a good report this fall.

Update: You and the Foresight Institute advocate international development of nanotechnology. When you picture this, do you picture many independent efforts all of which are international and collaborative, or do you picture one national program which is then done jointly, and what kind of European participation do you picture?

Drexler: At this point I think the sensible objective is to have multiple efforts pursuing multiple directions in multiple countries. The international goal should be to pursue this research in an atmosphere in which there is a consensus that extensive international exchange of information is appropriate and actively encouraged by national governments and their funding agencies. As we draw closer to real breakthrough capabilities, we may be able to build a consensus that a more closely coupled effort is appropriate.

Foresight Update 16 - Table of Contents

Offprints Available

Copies of four articles are still available from Foresight in the form of offprints:

"Molecular Directions in Nanotechnology," K.E. Drexler, Nanotechnology 2 (1991) 113-118.

"Molecular Manufacturing for Space Systems: an Overview," K.E. Drexler, JBIS 45 (1992) 401-405.

"Self Replicating Systems and Molecular Manufacturing," R. Merkle, JBIS 45 (1992) 407-413.

"Nanotechnology: Evolution of the Concept," C.L. Peterson, JBIS 45 (1992) 395-400.

To obtain these, send $3 each to Foresight Institute, PO Box 61058, Palo Alto, CA 94306 USA. (To conserve our supply, please request only one copy of each paper.)