scaling the depths

Before we can make nanotechnology a reality, we need a better understanding of fundamental properties at the molecular level.

in the late 1950s, there was a revolution under way. Vacuum tubes, long the mainstays of the electronic era, were on the verge of being rendered obsolete. The transistor, co-invented less than a decade earlier by John Bardeen, could perform all the tasks of a vacuum tube at a fraction of the size. Soon, transistor radios would become a fad, and electronic calculators would replace mechanical adding machines.

On one of the last nights of the 1950s, Caltech physicist Richard Feynman delivered a lecture at the American Physical Society meeting in Pasadena, Calif. The talk, entitled "There Is Plenty of Room at the Bottom," took the trend toward miniaturization and carried it to its logical extreme.

Novel approaches to manipulating matter at microscopic scales include a "wire" made up of silver particles 50 nanometers across.

Feynman talked about miniaturization of machines to a scale so small that they would not be visible by the naked eye, about machines that would read and write data by plucking and placing atoms. Feynman then predicted that all the publications of the world could be written "in a cube of material one-hundredth of an inch wide—which is the barest piece of dust that can be made out by the human eye."

With this, Feynman planted the seed for another revolution. Beyond just miniaturizing electronics, this revolution would shrink machines. It was a seed, however, that would need decades for germination.

A neuron grown on a silicon chip.

By 1983, when Feynman gave another talk about microscopic machines, nothing of note had yet been done in the field. Feynman described a variety of possible applications of such small machines and their method of manufacture. But he was growing impatient. "I keep getting frustrated in thinking about these small machines," he said. "I want somebody to think of a good use, so that the future will really have these machines in it."

Today, of course, we are beginning to see the fruits of Feynman's vision. Microelectromechanical machines are cropping up in such everyday items as automobiles and televisions. But as we push further, into the realm of nanoscale devices, we begin to run into a lack of understanding of the fundamental physical forces in play. Before we fulfill the promise of nanotechnology as laid out by Feynman, we must first discover which familiar rules apply—and which don't—at the molecular and atomic scales.

pinpoint accuracy

Just how small is the nanoscale? The tip of a board pin or the cross-section of human hair are similar in size, about 100 micrometers in diameter. An inch is about 25,000 micrometers long. A red blood cell is about 7 micrometers across and just 2 micrometers thick. A complete MEMS device, an electrostatic actuator, can be only 35 micrometers long, with structural beams just 0.4 micrometer wide and 4 micrometers deep. The tip of the probe of a scanning tunneling microscope, used to record surface topography, has a diameter of 10 nanometers—that's 0.01 micrometer, or 10,000 across on the tip of a pin.

Within the cross-section of a human hair one can position several MEMS sensors and actuators today. Compared to such tiny machines, a particle of airborne dust appears to be a mountain.

One of the aims of research labs such as mine, the University of Illinois MEMS/Micromechanics Laboratory, is to study the mechanical properties of material at such a small scale. For example, MEMS can be used to measure the stiction between two moving surfaces brought together by capillarity of liquid between them, or the dynamic characteristics of fluid flowing through micro channels.

Most of these problems are multidisciplinary in nature. In our lab, we have developed an experimental method, using MEMS to study the mechanical behavior of thin films (materials in microelectronics and MEMS) under uniaxial tension. We have, for the first time, tested 99.999 percent pure aluminum films as thin as 30 nanometers. We found that the yield strength of aluminum increases as the film thickness decreases. The strength reaches a peak value of 750 MPa for 100 nanometers thin film, but it decreases with further decrease of thickness. Bulk aluminum of commercial purity yields at only 35 MPa. Our objective is to understand the microstructural mechanism responsible for the increase and decay of strength with thickness.

Using cantilevers to act on living cells (above) and thin sheets of aluminum (below), researchers can examine whether matter at the smallest scale reacts unexpectedly to mechanical stress.

Another difference between the nanoscale world and our familiar macroscale one is in the effect of energy. Light applies a small force on surfaces due to the "impact of photons," a force known as radiation pressure. It's almost negligible, on the order of one piconewton—a little more than two-ten trillionths of a pound-force—from a moderate-intensity laser beam, such as a laser pointer. But concentrated to the micro- and nanoscale, such a tiny force can have a real effect.

Our lab is exploring the use of radiation pressure in operating micro- and nanomachines. We have explored the possibility of developing opto-mechanical computational logic elements using bi-stable micromachines operated by radiation pressure.

Opto-mechanical devices have the ability to change or modulate the path of a light beam. The most common micro-optical elements are those that reflect, diffract, or refract light. Logic elements made out of such parts would process light pulses much the way that semiconductors process electrical signals. They are a necessary component for optical computers, which in theory can be much more powerful than standard electronic computers.

Of course, to make such computers fully optical, the switches themselves should be light activated. Not only will that save power—no photons must be converted into electrical pulses to operate switches—but they are less susceptible to interference from strong magnetic fields or other electromagnetic radiation.

These devices could be used in extreme environments—such as conditions with high temperatures or electromagnetic radiation flux—where conventional microelectronic devices usually fail. There are plenty of unknowns, including basic questions, such as whether the devices suffer progressive damage over time from the exposure to light, and whether the absorption of light creates surface stresses on the minute moving parts.

Another venue where fundamental work needs to be done is in the interface between mechanical and biological entities. Many nanotechnology enthusiasts have looked to nanomachines as the perfect vehicles for repairing damage to the body caused by disease, injury, or simple old age. Tiny machines could fix the damage cell by cell, eliminating the need for many types of drug or radiation therapies.

The natural analogues of such devices are viruses and other microbes that attack animal cells, and the body's own natural response to such attacks. Unfortunately, we still don't know how cells will respond to these sorts of intrusions by microscopic machines.

In a joint project with Harvard Medical School, our lab has been able to measure the mechanical response of endothelial cells when deformation is applied at an anchorage site of the cell. We attached a cell to a petri dish and began to stretch it with a MEMS cantilever beam. The more the cantilever bends, the greater the force applied to the cell.

Such studies will enhance our understanding of the collective behavior of cells in forming organs through mutual biochemical and force interaction, and on the effects of gravity and shear stress induced by blood flow on cell behavior. One day, perhaps, the function of cells can be changed by simple mechanical stimuli, rather than chemical stimuli such as drugs.

molecular parts

Where will all this fundamental research lead? The art that is gaining increased consideration is the so-called "bottom up" approach. MEMS and microelectronic devices are built by imposing the macroscale onto the micro. The structures are "sculpted" through use of photolithography. True nanotechnology, in some way, will be assembled from parts made of individual molecules or even atoms.

One way of doing this is through molecular self-assembly. Suitably designed parts will snap together without direct supervision from human controllers. The self-assembly of devices will be driven by chemistry, just as biology assembles its basic element—the cell, which grows and replicates itself. Futurists envision that one day, computer chips with molecular electronics may be self-assembled in a chemistry beaker with 1024 components and wires.

One such component will no doubt be a carbon nanotube. Just a few nanometers in diameter and up to several micrometers long, nanotubes are molecules made of single or multiple layers of carbon atoms rolled into a cylinder. Nanotubes are under extensive study for their unique mechanical and electronic properties.

Another potential component is a nanoparticle—a fabricated structure less than 100 nanometers in diameter. Such particles could deliver drugs to a specific location in the body or coat the surfaces of implanted devices to reduce the possibility of rejection. They might also be designed as modules able to perform one or two specific functions that can be put together Tinkertoy-style to form complex nanomachines.

The future potential of nanotechnology is enormous, but so are its challenges. How the two balance out to affect our lives is yet to be seen.

By testing ever-smaller quantities of matter—for instance, using the minuscule tensilometer (above) developed at Penn State—researchers are laying groundwork for building atomic-scale machines, such as the actuator envisioned below.

The success and the influence of a technological invention depend on a variety of factors, including the status of competing technology, market demands, economics, and the workforce that can mature the technology. Many inventions with considerable future promise—from the Wankel engine to the Dvorak keyboard—have failed to find the success their promoters thought was assured.

The key to success for a technology is its timely emergence and broad appeal. During the 1980s and early 1990s, MEMS were not immune from skepticism. But they stood the test of time. One reason is that the microelectronics industry had matured, developing techniques that were essential for MEMS fabrication. Micromechanics is like a sister of microelectronics: Both share the same technological infrastructure and similar technical know-how.

In addition, MEMS have three characteristics that have broad appeal. They are small, allowing access to previously inaccessible regions for sensing and actuation with considerably higher resolution than can be found in conventional machines. They bring enormous economies of scale because batch fabrication allows millions of MEMS devices to be made in the same time and with the same effort it would take to make just one. And, most important, the integrability with microelectronics allows MEMS and electronics to be made together and to cohabit the same chip. MEMS can sense the physical world—impact, light, heat, sound—and electronics can send the message to the outside world.

MEMS has found a permanent home in industry. Microscale phenomena offer a broad range of fundamental and technological challenges that demand substantial academic research. Most research universities in the United States, Europe, and Asia have active MEMS research groups.

Nanoscale research on the same fundamental questions is only now beginning. It is an open question whether the answers found at the microscale will hold up as we probe closer and closer to the molecular level. More than likely, there will be some surprises, and these will lead to the development of technologies that no one has yet dreamed of.

The poet Paul Valery once wrote, with a somewhat jaded eye, "The future is not what it used to be." But the continued introduction of MEMS into everyday life and the new world of nanotechnology that appears at the horizon seem like the true fulfillment of future promises.

We envision a world where our everyday usable objects will have MEMS sensors. They will form the eyes, nose, ears, tongue, and fingers to sense the environment, whether it is light, temperature, force, or contact. Together with microelectronics, they will manipulate at an atomic and molecular scale to read and write data in response to the demands of the information age. They will form hand-held diagnostic laboratories—labs on chips—to conduct a variety of pathological tests from a tiny drop of body fluid, and communicate the data to a central database for doctors' inspection. They will monitor the environment for chemicals in ground water, leaks in pipelines, cracks in buildings, and biological agents on the battlefield. The list can be endless.

Combine this with the staggering potential of nanotechnology to forge devices from the smallest building blocks of matter, and we are certainly on the verge of a revolution in the way we sense and control the physical world around us.

Taher Saif is an associate professor in the Department of Mechanical and Industrial Engineering at the University of Illinois at Urbana-Champaign.

home | features | news update | marketplace | departments | about ME | back issues | ASME | site search

© 2004 by The American Society of Mechanical Engineers