A publication of the Foresight Institute
In 1991 J.A. Sidles at the University of Washington proposed an "NMR microscope" that would exploit nuclear magnetic resonance to image a sample in three dimensions at atomic resolution, determining not only the locations of atoms, but their chemical identities as well. A rough prototype device was quickly built to test the idea by detecting electron spins rather than nuclear spins, and in 1992 researchers successfully mapped electron spin densities at micrometer resolution [see Progress in Update 16].
A more sophisticated instrument recently achieved 2.6-micrometer resolution while detecting protons instead of electrons, thus proving Sidles's approach as a way to map specific chemical elements. Although substantial improvements must be made in resolution and sensitivity if the instrument is to single out individual atoms, the resolution already obtained is better than that of conventional NMR devices by about an order of magnitude. Rapid progress continues unabated. [Science 264:1560-1563, 10Jun1994]
A high-resolution NMR microscope would reduce the time required for mapping a molecule from months or years to perhaps hours or days. Some obvious applications include: determining the structure and function of biological polymers (proteins, carbohydrates, nucleic acids), sequencing DNA, and studying inorganic materials. Less obvious, but equally important, would be a speed-up of the research cycle in areas such as protein design, where currently a great deal of effort is required to characterize an experimental molecule after it is synthesized. This, in turn, would make proteins and other polymers more attractive as nanotechnological materials.
Much biological research is done with tissue cultures rather than with whole organs or organisms. These typically consist of cells that are made to grow and reproduce in a stirred vessel or on an undifferentiated surface. The cells grow willy nilly, lose synchrony, randomly contact and influence each other, and become inhomogeneous in size and behavior. This nonuniformity acts as statistical noise in biological experiments and causes the loss of valuable information.
A research group at MIT and Harvard sought to eliminate these problems by controlling the shapes, sizes, and locations of cells grown on a culture plate. A way was needed to force cells to grow in a rectangular array, each cell separated from its neighbors. To accomplish this feat the researchers made a plastic die bearing an array of rectangular patches, using it as a "rubber stamp" to thinly coat a culture plate with rectangular patches of linker molecules. Later treatment with a solution of ECM protein, which has a specific affinity for cells, resulted in an array of cell-adhesive rectangles. When cells are plated onto such a surface they stick to the rectangles but nowhere else. If the rectangles are small enough and spaced correctly, most rectangles will be occupied by just one cell.
A cell attached to a plate in this manner can be made to grow until it fills its rectangle and then stops growing. The resulting array of rectangular cells has several useful properties: individual cells can be assigned coordinates for easy access; cells are effectively isolated from each other and can be experimented upon individually; when the cells stop growing they become synchronized at a particular phase in their life-cycle. [Science 264:696-698, 29Apr1994]
What does this "rubber stamp" technique have to do with nanotechnology? It is an early step toward turning cell cultures into nanosystems. Cells don't qualify as nanodevices just because someone uses them as factories for producing molecular productscells grown helter skelter in tanks or in petri dishes are no more examples of nanotechnology than are molecules synthesized in flasks by traditional chemistry. Cell culture becomes nanotechnology only when we learn to control every aspect of cellular behavior. Getting control over location, shape, and size is a good way to begin.
Julius Rebek at MIT and his colleagues made news recently with a fairly simple molecule that self-assembles pairwise into a roughly spherical shapelike the two identical shapes that form a tennis ball. The basic idea is simple: each half of the final structure should be a symmetrical molecule with substantial curvature, and with ends that are complementary in shape to the middle and are attracted to it. Rebek's minimalist approach led to a molecule having only 88 atoms. This monomer, said to be easy to synthesize, dimerizes to make a hollow "spherical" complex held together by eight hydrogen bonds. The internal cavity readily encapsulates certain small molecules like chloroform. [Science 263:1267-1268, 4Mar94]
Rebek's group is also working on a more ambitious goal: to design and build monomers that would self-assemble around a template of synthetic self-replicating moleculesrather like a stripped-down version of a virus. [C&EN 3Jan93: 23-24].
Nadrian Seeman of New York University and his colleagues have been mentioned in this column before for the construction of complex molecular shapes out of DNA. (See Update 12.) While their previous efforts (such as the synthesis of a DNA cube, reported in 1991) seemed like impressive tours-de-force, they didn't at the time strike me as a particularly promising approach to nanotechnology. The problem, I thought, is that nucleic acids are not a rigid enough material for use in nanomachinery. (I should have known better: after all, are not ribosomes biology's most ancient and fundamental molecular machinerymade largely of nucleic acids?) Dismissing DNA in this manner, I let the true significance of Seeman's work pass me by, unrecognized.
The latest DNA construct, however, is an eye-opener. This time Seeman and Yuwen Zhang have assembled a molecular complex having the topology of the edges of a truncated octahedron. Each of the 14 "faces" of the object is defined by a closed strand of DNA that intertwines with the strands associated with adjacent faces. The entire complex contains 2550 nucleotides and has a molecular weight of about 790,000 daltons.
Each of the 24 vertices of the object has an extra loop of DNA for future use as a connectorthe aim here is to connect an array of these truncated octahedrons together to form periodic matter in one, two, or three dimensions. (Seeman stated in a 1988 interview that his aim was to build periodic lattices out of DNA. As a protein crystallographer he wanted to trap hard-to-crystallize proteins within a periodic framework so that their structures could be solved by x-ray diffraction. It appears that he is closing in on his goal.) [J. Am. Chem. Soc. 116:1661-1669, 1994]
The significance of this work for nanotechnology lies not in x-ray crystallography, but rather in its application to building nanomachinery. While protein engineers laboriously puzzle out the relationships between sequence, 3D structure, and function in proteins, DNA engineers can already design complicated structures and reliably synthesize them!
This is not to say that DNA is an all-purpose material. Some nanomachine components (such as those that carry out chemical reactions) will need to be made of protein or diamond or some other material. But DNA might make a dandy framework for holding active components in the proper relative positionsa task that is as important in a nanodevice as it is in a car. DNA might also be used as transport linesi.e., serve as tracks for guiding molecular motors as they haul loads within a nanodevice; in cells this role is filled by proteins.
Maybe we don't have to wait for protein engineering or STM techniques to mature in order to start experimenting with nanomachines. With current knowledge and techniques, it may be possible to choose groups of enzymes that normally function together on a cell membrane and attach themor the active portions of themto DNA frameworks to see if they can be made to function together there.
One possible path to nanotechnology uses scanning probe devices like the atomic force microscope (AFM) as robot arms for manipulating individual atoms and reactive groups. The probe tips in existing AFMs, being merely pointed bits of ceramic, have too little structure for any but the simplest tasks. In order to hold reactive groups in specific orientations while they are being bonded to a workpiece, an AFM tip should be functionalizedprovided with a cap of some sort that would have the specific shape and charge distribution needed for it to serve as a tool for the task at hand.
Preliminary steps in this direction include recent work by Ernst-Ludwig Florin and others at the Technical University in Munich, who studied the interaction of two proteins: avidin and biotin. Avidin molecules, which contain a biotin-binding pocket, were used as functionalizers of an AFM tip. Biotin-coated beads served as a "sample" for the tip to interact with. Instead of scanning the tip across the sample as would be done in normal AFM operation, the researchers simply brought the tip into contact with the sample and then pulled it away, measuring the adhesive force as they did so. Since the adhesion arose from the binding of avidin-biotin pairs that happened to align properly, the force fell in decrements of 160 piconewtons as the tip's withdrawal caused individual pairs to unbind. [Science 264:415-417, 15Apr94]
The functionalized AFM tip in this experiment had many avidin molecules bound to it in random locations. Researchers will now have to find a way to limit the functionalization to a single molecule located near the apex of the tip.
Biological evolution has been dabbling in electronics for more than a billion years, responding to the need for control over the movement of electrons in such biochemical processes as respiration and photosynthesis. A variety of proteins have evolved to carry out electron transport, the best known being the cytochromes and chlorophyll. These proteins are closely associated with a non-protein structure called hemea ring of four 5-membered rings with various side-chains. The electrical and mechanical details of how heme-containing proteins manipulate electrons are starting to be understood. (See this column in Update 9)
A research group at the University of Pennsylvania and the Du Pont Co. has been investigating a designed protein that self-assembles into a bundle of four helices with hemes bound into the cavity running down the center of the bundle. The design is a drastic simplification of the heme-binding portion of cytochrome bc1, a biological protein involved in transporting electrons generated by respiration. Several versions of the model protein were studied in order to tease out the contribution made by each of the four hemes that are found in cytochrome bc1. It was found that the synthetic proteins stably bind hemes in the expected locations and that the hemes accept electrons in the theoretically predicted order. [Nature 368:425-432, 31Mar94]
This work illustrates the steady progress being made toward designing proteins that do more than just float around in solution. The fun won't really begin, though, until someone genetically modifies a cell so as to replace a natural protein with a radically redesigned one. When will that happen? I expect to be taken by surprise.
A group at Boston College has designed and built a "paddlewheel" molecule with a built-in "brake." The paddlewheel portion of the molecule is a simple 3-ring structure known as tryptycene; the brake has two rings and is connected to the "axle" of the paddlewheel. In solution, the wheel tends to spin relative to the brake. But when mercury(II) ions are added, the brake changes position and interferes with the wheel's rotation. [C&EN 25Apr94: 6-7]
A reliable method has been developed for making dense arrays of nanowires. At the Naval Surface Warfare Center in Maryland, C.A. Huber and colleagues used injection molding to force molten metals into 250 nm-diameter channels in an aluminum oxide plate. A study of the resulting nanowire array showed that most of the wires were conductive. The metals used in this study included indium, tin, and aluminum; several semiconductor materials were also successfully injected. In theory, at least, it should be possible to fill channels as narrow as 3 nm by this technique. [Science 263:800-802, 11Feb94]
Dr. Russell Mills is research director at KAH Sciences in California.
This meeting covers a wide range of nanoscale science and technology, including both top-down and bottom-up approaches to miniaturization, without an emphasis on molecular nanotechnology. Readers with technical backgrounds can use their knowledge of the most promising development paths to select sessions of relevance to molecular nanotechnology and molecular manufacturing. Note that the dates of this meeting combine well with our fall Senior Associates meeting.
The recently formed Nanometer-scale Science and Technology Division of the American Vacuum Society (AVS) will host the Third International Conference on Nanometer-scale Science and Technology (NANO 3) at the 41st AVS National Symposium. The conference, which will be held at the Colorado Convention Center, Denver, Colorado, during the week of October 24-28, 1994, is being co-sponsored by the AVS, the Office of Naval Research, and the National Science Foundation.
NANO 3 will be of interest to those who wish to keep abreast of the rapid pace of development in nanometer-scale science and technology, which is now underway at a growing number of laboratories throughout the world. The conference will highlight recent technical achievements, with major results reported from such disciplines as biology, chemistry, electronics, engineering, fabrication, materials science, medicine, metrology, micro-instrumentation, optics, and physics.
The central theme of the conference is the demonstration of novel laboratory capabilities for the deliberate creation of atomic- and molecular-scale functionality, and the development of systems and processes which will make control and measurement of new physical and biological phenomena on these scales possible. The three plenary lectures scheduled to begin the conference embrace this theme:
Prof. Gopel is Director of the Institute of Physical Chemistry at the University of Tubin. His work deals with the development of novel sensors based on the self-assembly of molecules at interfaces. Dr. Eigler is an IBM Fellow at the IBM Almaden Research Center. Using a cryogenic scanning tunneling microscope, he has shown that it is possible to form individual atoms into structures and map changes in the electronic structure which occur as a consequence. Dr. Sugiyama is at the Electrotechnical Laboratory of the Ministry of International Trade and Industry in Japan. He has worked extensively on semiconductor magnetic sensors including the Quantum Hall probe.
The NANO 3 program consists of 16 oral and four poster sessions, featuring 19 invited speakers and over 225 contributed papers in addition to the plenary session. A listing of just a few of the session titles and invited talks will give a sense of the topics to be covered during the meeting:
To tie in with NANO 3, the AVS is organizing a special tutorial on Sunday, October 23, entitled "Nanostructures: Fabrication and Characterization."
For a preliminary program that contains a registration form for the AVS 41st National Symposium and NANO 3, contact the American Vacuum Society at: 120 Wall St., 32nd Floor, New York, NY 10005 tel: (212) 248-0200, fax: (212) 248-0245, e-mail firstname.lastname@example.org.
From Foresight Update 19, originally published 15 September 1994.