include "../includes/header.php"; ?>
A researcher at Intel recently acknowledged that he saw no way to continue the expected progress in chip miniaturization without a major change in technology (Ref. 1). While semiconductor technology regularly comes up with astonishing advances, Moore's Law predicts that by about 2015, we must be working at the molecular level to stay competitive. Molecular nanotechnology is the likeliest candidate for providing this ultimate precision in manufacturing. This brief survey, while far from comprehensive, attempts to sketch the state of the field and speculate on its future.
For millennia our species has worked to improve its control over the structure of matter—pounding, heating, and cutting to approximate the right combinations of atoms, in the right ratios, formed into the right shapes. The frontier of the small was envisioned in 1959 by Richard Feynman in a famous after-dinner speech, in which he pointed out that physics does not rule out the control of individual atoms (Ref. 2). He even pointed to a development pathway: machines that would make smaller machines, which would make smaller machines, and so on—what we refer to today as the "top-down" path of miniaturization.
Since then developments on or near the nanoscale have spread in all directions: nanoparticles, nanostructured materials, and nanodevices; a worldwide study of this broad field has been done by the International Technology Research Institute at Loyola College (Ref. 3). The trend across the board, however, is clear—we continue to pursue better control, and this will go on until we hit the limits imposed by physical law. While this effort is being carried out all over the world, it is in the U.S. that the more specific goal of molecular nanotechnology per se has received the most attention and is the farthest along experimentally.
In 1981, a paper in the Proceedings of the National Academy of Sciences (Ref. 4) pointed in a new direction: construction of materials and devices from the "bottom-up" with every atom in a designed location, now termed molecular nanotechnology. The new goal aims at the construction of systems of molecular machines, and biology is viewed as an existence proof for the feasibility of this. Controversial for a number of years, the goal's acceptance benefited from a series of conferences (Ref. 5) and a graduate-level textbook (Ref. 6), combined with major advances in the laboratory including the imaging and movement of individual atoms via scanning tunneling microscope (http://www.almaden.ibm.com/vis/stm/atomo.html#stm10).
The pursuit of molecular nanotechnology (often referred to simply as nanotechnology) comprises a wide variety of disciplines: chemistry, physics, mechanical engineering, materials science, molecular biology, and computer science. While participation from the natural sciences is needed, the field is primarily one of technology development—of engineering—rather than a scientific endeavor per se.
Key to enabling these systems to make macro-scale objects is the same principle used by nature in making large products: parallelism. Very small machines can make very large objects, but only if many of them work together, as when plant cells build large trees. Producing these large numbers of small machines requires that at least some of them be able to make copies of themselves, an ability called self-replication. Normally only found in nature and in software, self-replication will need to be made to work for human-designed nanoscale machinery.
In the media, nanotechnology is often confused with related fields such as MicroElectroMechanical Systems (MEMS) and molecular electronics. Table 1 illustrates the most basic differences among these various efforts, which do have some overlap. In the case of MEMS, it helps to remember that while the two technologies differ by a factor of about 1000 in linear dimension, this translates to a factor of a billion in volume—very different indeed. Also, as MEMS researchers point out, MEMS is not a goal but a working technology, rapidly growing into a major industry.
in a designed location)?
|Controlled movement of
atoms in 3D space?
|Chemistry||Yes, local only||Local, yes; large scale, no|
|Molecular electronics||No||No large-scale movement|
An early concern regarding the feasibility of nanotechnology involved quantum uncertainty: would it make these systems impossibly unreliable? Despite the molecular machine systems seen in nature—the bacterial flagellar motor, for example—it took a while for understanding of this area to spread. Meanwhile, a far more relevant limitation was less often discussed: the problem of thermal noise, the continual jiggling movement of molecules due to heat. It is this phenomenon which most seriously limits what can be done mechanically at the molecular scale.
This challenge of building machines which function correctly despite thermal noise drives nanotechnologists to maximize the stiffness of the materials they design. A soft floppy molecule is simply battered about, while a stiff molecule is better able to do accurate, reliable positioning. These types of calculations are not usually done by chemists—who rarely attempt to control forces exerted at the molecular scale—and instead require techniques drawn from mechanical engineering or materials science.
All this design effort is intimately bound up with work drawn from computer science. Designing molecular machine systems, like designing pharmaceuticals, is done using computational chemistry software, some of the best of which is available in commercial software packages. This software based on force field models which predict how atoms will be most stably bonded and positioned in a given structure. Designs that survive analysis with this software are structurally reasonable, at least in theory.
The most complex molecular machines designed to date include a differential gear (Fig. 1), a simple pump (Fig. 2), and a fine-motion controller (Fig. 3), most of which result from collaboration between the Institute for Molecular Manufacturing (Ref. 7) and Xerox Palo Alto Research Center. Animations of some machine designs have been done at NASA Ames (Ref. 8). One common format for publication of these atomic patterns is called PDB, for Protein Databank; designs for the above machines are published online for downloading and testing by other laboratories.
|Figure 1. A molecular differential gear.
(Image courtesy of Institute for Molecular Manufacturing)
|Figure 2. A simple pump selective for neon.
(Image courtesy of Institute for Molecular Manufacturing)
|Figure 3. A fine-motion controller for molecular assembly
(Image courtesy of Institute for Molecular Manufacturing)
The designs can also be analyzed using molecular mechanics software that models the dynamics; however, these tests involve such large amounts of supercomputer time that testers can be tempted to increase the modeled device operation speed far beyond specification, resulting in catastrophic breakdowns which are interesting but unrealistic.
Some of the most advanced work in modeling molecular machines has been done at Caltech's Materials and Process Simulation Center under the direction of William Goddard, whose team won the 1999 Feynman Prize in Nanotechnology (Theory) (Ref. 9).
The building of these molecular machine designs requires an analysis of specific chemical reactions used by specific molecular tools. One example, a hydrogen abstraction tool, has been analyzed using both molecular mechanics and ab initio quantum chemistry methods by several different groups, the first of which was awarded the 1993 Feynman Prize in Nanotechnology (Theory) to Charles Musgrave, then in Goddard's group at Caltech, now at Stanford's Department of Chemical Engineering.
In the world of physical, rather than computational, experiments, perhaps the most exciting tools have been the various scanning probes including the classic Scanning Tunneling Microscope (STM) and the later Atomic Force Microscope (AFM), followed by many others. Early work with these involved moving atoms around on a surface in vacuum; now they can be used in liquid and can bond a molecule to a surface. Recently, an antibody was used in conjunction with an SPM tip (Ref. 10, 11); it is this kind of interdisciplinary work which is most exciting to nanotechnologists.
Perhaps the easiest way to track the highlights of progress toward molecular nanotechnology is to monitor the Feynman Prizes in Nanotechnology (Ref. 12), begun in 1993 and now awarded annually in both experimental and theoretical work. A Feynman Grand Prize of $250,000 is also posted for specific, very ambitious technical achievements in molecular nanotechnology (Ref. 13).
Work honored with this prize includes the DNA constructions made by Prof. Nadrian Seeman, a chemist at New York University. Seeman uses DNA not as an information-embodying molecule, but as a structural member with which to build three-dimensional objects (Example: http://seemanlab4.chem.nyu.edu/nano-oct.html; see also "New Motifs In DNA Nanotechnology"). One of his long-term goals is to build general-purpose computational devices. Note that this work, while aimed at DNA-based computing, is quite different in approach from Leonard Adleman's at UCLA, who solves specific search algorithms by matching DNA strands moving loosely in liquid—fast at some narrowly-defined search problems, but probably not practical for general-purpose computing, though this possibility continues to be examined.
Only a few universities have centers explicitly aimed at molecular nanotechnology. The first was Rice University's Center for Nanoscale Science and Technology, thanks to Nobel chemist Richard Smalley, co-discoverer of C60 (known as the "buckyball"). Smalley is very clear about his goals: "The impact of nanotechnology on health, wealth, and lives of people will be at least the equivalent of the combined influences of microelectronics, medical imaging, computer-aided engineering, and man-made polymers developed in this century...I believe at the moment our weakness is the failure so far to identify nanotechnology for what it is: a tremendously promising new future which needs to have a flag," he said. "Somebody needs to go out, put a flag in the ground, and say: 'Nanotechnology: This is where we're going to go.' We should have a serious national initiative in this area." (Ref. 14)
Another targeted group is the Center for Nanotechnology at the University of Washington, a group of 38 faculty headed by Prof. Viola Vogel, whose own work focuses on molecular assembly at interfaces. Yet another player to watch is the Center for Nanostructured Materials and Interfaces at University of Wisconsin-Madison.
Currently, by far the largest amount of relevant research is done in labs not specifically targeted on nanotechnology. The Beckman Institute for Advanced Science and Technology at University of Illinois at Urbana-Champaign, for example, has a broader focus but includes many projects on the pathway to nanotechnology, using enabling technologies ranging from scanning probes to self-organizing systems. Their Molecular and Electronic Nanostructures group, one of three main areas at the Institute, includes 25 faculty members.
Relevant work is also being done in the private sector, with IBM being the company most often mentioned. The original work on scanning probes was done partially at IBM Zurich, also the base of physicist James Gimzewski, whose work on the molecular abacus (Molecular abacus from IBM Zurich as shown in http://www.zurich.ibm.com/pub/hug/PR/Abacus/) and other achievements earned his team the 1997 Feynman Prize (Experimental).
A small number of startup companies have begun to enter the field, the best-known of which is Zyvex of Richardson, Texas. The recent departure of computational nanotechnologist Ralph Merkle from Xerox PARC in favor of Zyvex indicates that these smaller companies are beginning to attract leaders in the field.
Most computer companies are at the stage of beginning to track nanotechnology, rather than doing active research. Sun Microsystems is farther along than most, now on its third year of sponsorship of the main nanotech conference series.
Probably the best-known corporate lab targeted on nanotechnology is Hewlett Packard's basic research lab headed by R. Stanley Williams. In July 1999 they announced the construction of a molecular switch, made in collaboration with a team at UCLA under the direction of Prof. James Heath. Using rotaxanes, a molecule type invented by UCLA chemist Prof. J Fraser Stoddart which changes shape and conductivity when oxidized, the HP/UCLA team constructed a one-molecule-thick layer of molecules all pointed in the same direction. They then oxidized groups of these, forming small spots of about 1 micron diameter where the conduction of current was blocked compared to the surrounding, unoxidized film.
As was stressed by the researchers in their Science article, but omitted in most media coverage, the switch is not reconfigurable: once oxidized, it stays open and cannot be closed. Thus, it would be useful in programmable read-only memory (PROM), but not random-access memory (RAM) applications. Williams' group will be pursuing a reconfigurable switch in future work.
HP's non-reconfigurable switch construction process combines a feature of molecular nanotechnology—physical movement—with the two-dimensional films typical of molecular electronics, to make a final product in the category of molecular electronics. The physical movement is important to the building, but not the operation, of the device, since it cannot yet be reversed. Still, it's a step toward the routine use of molecular movement in computing, and though the device is molecular-scale in one dimension only, its success could bring great attention—and funding—to the field of molecular computing in general.
Not surprisingly, funding is the limiting factor for nanotechnology R&D. Until the major U.S. federal research funding agencies step up to the job, the field doesn't look "real." Fortunately for many who would like to work in this area, significant federal funding appears to be on the way—which should stimulate private industry budget commitments as well.
One of the first agencies to take a strong interest in molecular nanotechnology was NASA, stimulated by the prospects of super-strong materials for spacecraft. Their modeling work at NASA Ames' Numerical Aerospace Simulation facility has already resulted in two Feynman Prizes.
Both the U.S. Navy, through its Strategic Studies Group, and the Army at its Nanotechnology for the Soldier System Conference, have been tracking molecular nanotechnology work of relevance to military systems. To public knowledge, no formal U.S. military R&D program has yet been established, other than the scanning probe work in place at the Naval Research Lab for a number of years already.
Just recently, though, federal attention has increased: the National Science Foundation has announced that it will be coordinating federal R&D funding in nanotechnology. While not aimed specifically at molecular nanotechnology, it appears that at least some funds should flow in that direction. Certainly they will if NSF heeds the advice of Richard Smalley, whose Congressional testimony dramatically made the case for such funding (Ref. 14).
There is strong political appeal in funding nanotechnology, due to its anticipated applications in such areas of popular interest as medicine, space, and the environment. Early Senate testimony on its potential usefulness in clean manufacturing (Ref. 15) made clear the power of nanotechnology to virtually eliminate unwanted chemical byproducts—pollution—from our air and water. Beyond making future manufacturing processes extremely clean, the technology would be useful in cleaning up wastes produced in the past: a true ability for environmental restoration. Enabling a clean environment, without sacrificing standard of living, appeals to researchers and politicians alike, and should result in substantial funding as this understanding spreads. The dream of being able to make our own products as elegantly and cleanly as nature's are made now is a powerful draw.
The strong materials studied by NASA made an appearance in Smalley's testimony, in the form of the "space elevator," a design involving a physical structure stretching all the way from Earth's surface to geosynchronous orbit, along which cargo and humans could move. This would require an extremely strong material—as strong as nanotubes—but calculations indicate such a material could exist. With nanotechnology, we could build it.
By far the most popular application mentioned for nanotechnology is in medicine, where extremely small machines are envisioned that would enter the human body to destroy viruses, remove arterial plaque (see Nanomedicine Art Gallery, Image 104), and get rid of cancer cells. In his House testimony, Richard Smalley made a dramatic case for nanotechnology based on its expected use in being able to thoroughly destroy cancers. (Ref. 14)
Even more extreme and surprising medical applications are discussed (Ref. 16), from preventing aging, to reviving patients from low-temperature suspended animation (cryonics), to making significant structural changes in the body itself.
These extreme scenarios are not expected to arrive in the short term. Even the cancer treatments advocated by Smalley were estimated by him as taking ten years of serious effort; however, the likelihood that such an effort will occur during the working lifetime of many of today's researchers appears quite high. Nanomedicine, a new text published in October 1999, may help in making these ambitious goals more understandable (Ref. 17).
Many feel that the wide-ranging applications of such a potentially powerful tool as nanotechnology—one which raises key public policy issues—should be discussed beyond the research community as well as within it. A 1986 book, Engines of Creation (Ref. 18), in which the present author was involved, ensured that this concern was addressed from the earliest days of nanotechnology research.
Safety concerns have been raised regarding the proposed ability of some molecular machine systems to self-replicate—to make copies of themselves. Constructing an artificial self-replicating system at the molecular level will be difficult, but once made, it could make many copies: perhaps too many. We have sad experience with what happens when nature's molecular machine systems decide to over-replicate, especially when they do so by pirating our own bodies' machines, as viruses do.
As in biotechnology, a number of fairly obvious safety rules can reduce the potential for problems here. Perhaps the most fundamental is that replicators should not be built which could survive and make copies "in the wild"; they should always require unusual lab-supplied materials not found outside. Another rule would be to encrypt the replication instructions so that even a one-bit accidental change would render the device useless. Foresight Institute is preparing a set of suggested safety guidelines for discussion by the research community.
While safety remains an issue to address, a more difficult problem is that of deliberate abuse of the technology. Just as in biotech, where concerns are mounting that the technology could be used for weapons of mass destruction, nanotechnology has the potential for serious abuse by hostile entities ranging from governments to terrorists. And as with biotech, such weapons cannot be seen from space-based satellites, a technology which has played a key role in monitoring the proliferation of nuclear weapons. It is this concern which stimulated the early publication of public policy issues in Engines of Creation and Unbounding the Future. Debate on this intimidating topic continues within both Foresight and the Navy's Strategic Studies Group.
Multidisciplinary fields tend to have a larger than average share of "disconnects," and this happens fairly often in nanotechnology. A common problem was pointed out in Chemical & Engineering News, where University of Texas Prof. Allen Bard correctly pointed out that current work on devices rarely makes any attempt to see whether it is even theoretically possible to tie these devices into systems. Too often, the teams assembled are missing those who could provide systems perspective. Natural scientists, whose work is forced to fit together by nature itself, are unused to having to do systems-level design, and may even be entirely unfamiliar with that field.
Phaedon Avouris of IBM, winner of this year's experimental Feynman Prize in Nanotechnology (Ref. 9), has stated a similar, valid criticism: "Any time someone observes an effect, a press release is issued claiming a revolution in computing. There's no revolution imminent. No one is going to abandon a trillion-dollar electronics industry to start from ground zero." He describes his own work modestly: "But carbon nanotubes do have rather unique electrical properties that are likely to lead to device applications. Special-application electronics and hybrid silicon-nanotube devices are the likely initial applications."(Ref. 19)
Very few arguments against the feasibility of molecular nanotechnology itself are made today, but at least one concern is still with us: the problem of complexity. The more advanced systems have huge numbers of parts, and their design and manufacturing involve a series of projects impressive in their complexity. Ralph Merkle has examined the numbers involved, compared to past human-built systems, and comes to the conclusion that this task is not beyond us in principle. (Ref. 20) Still, of all the concerns that stand in the way, complexity is perhaps the most daunting.
Both the complexity aspect and the system engineering challenge potentially fall into the areas of expertise of computer scientists. They are perhaps best positioned to coordinate, and even direct, the work of the other specialties needed in the molecular nanotechnology development task. As the open-source software community has demonstrated repeatedly, software engineers are at the cutting edge of grappling with large, physically dispersed teams working on tasks of boggling complexity, and succeeding.
After a slow start, interest from chemists has been growing. As Science Watch proclaimed in a headline, "Chemists Show Mega Interest in Nanotechnology." In a survey of the top ten "Hot Papers" in chemistry, they reported that "Nanotechnology is fast becoming one of the major areas of chemistry. Its appeal lies in the long-dreamt-of ability to investigate and manipulate matter at the level of individual atoms and molecules...there is the tempting thought that discoveries could play a key role in a future world of nano devices and nano computers." Of the top ten papers, four were on nanotechnology. (Ref. 21)
But there are cultural barriers to overcome. As the September 1999 crash of the Mars Climate Orbiter has shown, even differences in units can cause diaster in an R&D project, and the various disciplines needed for nanotechnology use widely different units, even for measurements as simple as linear dimension (nanometers vs. Angstroms). This problem gets much worse when attempting to compare energies: electron volts vs. kcals/mol vs. Hartrees vs. joules.
Worse is the academic structure still dominating in chemistry, which is based directly on the medieval model of the "master" (professor), the journeyman (post-doc), and the apprentices (graduate students). Unlike computing, young people are not to create in their own name until they have completed their "masterpieces" (PhDs) and eventually attained the status of masters themselves. This credentialist system is so different from computer science that getting the two groups to work well together will be a real challenge.
On occasion one will still hear the comment that molecular nanotechnology is "just chemistry." At one time nascent nanotechnologists might have been put off by this, but now, as the field coalesces and those participating feel more confident, we can choose to see it as an assertion that the field of chemistry is ready, willing, and able to expand to include aspects of areas needed to do molecular nanotechnology, from understanding current molecular machine systems (molecular biology) to calculating forces in the design of moving parts (mechanical engineering).
Two fields of study already try to interact with all these relevant fields needed by nanotechnology: systems engineering and computer science. We should not be surprised to find individuals from these fields in charge of molecular nanotechnology R&D projects in the years to come.
The goal is now clear to all: the ability to inexpensively arrange atoms in most of the patterns permitted by physical law. The pathway is still being debated, and will no doubt continue to be debated until nanotechnology is actually developed and deployed—which isn't that far away.
Christine Peterson is Executive Director of Foresight Institute. She has co-authored two books, Unbounding the Future: the Nanotechnology Revolution and Leaping the Abyss: Putting Group Genius to Work. Contact her at .
include "../includes/footer.php"; ?>