Recent experiments have shown that thermal conductivity of carbon nanotubes can be more than twice that of diamond. It should be noted that high mechanical strength often comes with high thermal conductivities. Recent experiments have shown that the thermal conductivity of carbon nanotubes can be as high as 3000 to 6000 W/m K at room temperature, which is more than twice that of diamond. It was recently shown by Alex Zettl and his group at the University of California, Berkeley that the relative motion between different shells of multiwall carbon nanotubes has some unique properties and can serve as excellent mechanical bearings that do not undergo any wear. Recent work has led to multifunctional probes, which, besides topography, can detect thermal, electrical, magnetic, and optical signals at nanoscales. The engineering challenge now is to develop microelectromechanical systems (MEMS)-based probes that integrate multiple functions on a single tip.


Imagine yourself shrunk about a billion times and having the ability to swim around in a gas, liquid, or solid. The trip will not be smooth, but rather quite bumpy. What you will encounter is not the continuum that we usually associate with matter, but the discreteness of the atoms and molecules that form the building blocks of nature.

It was more than 40 years ago that Richard Feynman envisioned a technology using the ultimate toolbox of nature, building objects atom. by atom or molecule by molecule. The last two decades have witnessed the first glimpses of this vision, sparked by some key inventions and discoveries-for instance, the scanning tunneling microscope, which can image and manipulate single atoms, or the Fullerene family of molecules such as buckyballs and carbon nanotubes, which possess unique physical properties not found in bulk matter.

The nanometer scales also provide a natural overlap in size between biological molecules, such as DNA and proteins, and inorganic materials, including nanostructured semiconductors, ceramics , and metals. This overlap is already enabling formation of hybrid nanoscale objects, which also have unique properties and functions.

If we stretch our imaginations and think what all this will lead to, we can envision technologies that are very different from what we currently have. For example, information processing and storage may eventually be based on information bits defined by single particles (electrons or photons) as opposed to millions of them currently used, thus providing enormous changes in density and power consumption. Health care and biotechnology could rely on nanostructured molecular assemblies that would deliver drugs at precise locations inside our body or detect the inception of cancerous tumors.

One could also envision solid-state engines and refrigerators based on nanostructured thermoelectric materials that not only have higher performance than devices now, but are environmentally much more benign. Nanostructuring also allows tailoring of mechanical properties of metals and ceramics, which could have a huge impact on the transportation and space industries. There is little doubt that if all these could be achieved, it will create a new technological base and infrastructure that will have major social and economic impact. It is for these reasons that the National Nanotechnology Initiative was recently announced by the federal government as a major research thrust.

It is worth noting that at present the technology in almost all areas is nonexistent. The key to technology is not in the science alone, but in the combination of science and engineering. Nanoengineering, therefore, is one of the most important ingredients of nanotechnology's success.

Eye On The Future: Nanotechnology


Recent experiments have shown that thermal conductivity of carbon nanotubes, shown here in a computer illustration, can be more than twice that of diamond.

Nanoengineering Challenges

So what exactly is nanoengineering? Let us try to filter through much of the hype and identify the key engineering issues that must be addressed for nanotechnology to be developed.

One of the biggest challenges in nanoengineering is the issue of how to integrate objects across different scales. In microelectronics and micro electromechanical systems, one can arrange multiple objects precisely on a single platform using optical lithography. This is often called the "top-down" approach, in which one uses lithography to cut and shape a thin solid film. The success of optical lithography as a manufacturing technique lies in its speed and parallel nature, which result in low cost. However, it is limited in its resolution to feature sizes larger than 100 nanometers.

Nanotechnology, by contrast, is likely to use structures that are 1 to 100 nm in size.

Clearly, the top-down approach of optical lithography is part of the answer, because it interfaces the upper limit of the smaller range to the larger scale. But one must develop "bottom-up" techniques of self-assembly that can integrate in the 1-100 nm size range and overlap with top-down approaches. The self-assembly, however, must be encoded; that is, one must be able to precisely assemble one object next to another to form a designed pattern.

How does one encode self-assembly? One does not have to look very far. In biology, the manufacturing of proteins, for example, occurs by expressing the DNA code into a precise assembly of amino acids through a series of enzymatic reactions. Growth of inorganic crystals on certain facets and patterns is also a form of coded assembly of atoms, although perhaps not as precise as biology.

Production requires one to direct the coded self-assembly of nanostructures using optical lithography and thus form a hybrid manufacturing technique. Many novel techniques have been developed for this purpose, but there is none that can be called a technology yet. Much work remains to be done.

The technology of nano-to-macro LSI will offer the possibilities of synthesizing integrated nanosystems. The questions one should ask are: What advantage does nanostructuring provide for a system? What is unique about nanoscales? Let us explore this through a few examples.

The ability to convert energy efficiently among different forms-thermal, mechanical, electrical, optical, chemical-creates an important part of any modern society's infrastructure. The magnitude of societal energy consumption suggests that even marginal improvements in efficiency and conversion methods can have an enormous impact on the economy, energy reserves, and the environment.

It was recently shown by Mildred S. Dresselhaus and her group at MIT that if thermoelectric materials such as alloys of bismuth are nanostructured, quantum confinement of electrons and phonons allow tailoring of their thermal , electrical, and thermoelectric properties in a way that their thermoelectric figure of merit can be dramatically increased.

This means that one then would be able to design and build solid-state energy conversion devices with better performance than internal combustion engines and vapor compression refrigerators. Although electron transport in these confined nano- structures is relatively well understood , understanding of heat transport has remained a bottleneck.

Chemomechanical Engines

Engines can be thought of as devices that convert energy among chemical, thermal, and mechanical forms. Direct chemomechanical energy conversion is rare at macroscopic scales. In contrast, motion at nanoscales is generally by such direct conversion. For example, biological motors such as ATPsynthase and flagellar motors create rotary motion by extracting energy from an ionic current (H + ,Na+) . A subunit of ATPsynthase can rotate at up to 1,000 rpm, and it can generate 100pN'nm of work. Flagellar motors, found on the cell membranes of bacteria such as E.coli, are about 45 run in diameter and are even more powerful. They can generate speeds of 6,000 to 60,000 rpm at 10-9 hp and a power-to-weight ratio of 10 hp/1b.

Biologists have investigated these motors. If, however, one looks at them from the engineer's viewpoint, one could envision isolating the motors to propel nanoscale robots or using them in devices that could convert light energy into mechanical power through the use of light harnessing molecules.

Nanostructured solids can be far superior to bulk ones in terms of mechanical properties. For example, carbon nanotubes have elastic modulus on the order of 1 TPa and can sustain critical strain of about 5 percent before yielding. What is interesting, however, is that the nanomechanics of plasticity and dissipation in such solids can be very different from those in bulk solids, where dislocations play a dominant role.

The high strength combined with the small mass of nanostructured solids also leads to mechanical resonant structures with frequencies of 100 MHz to 10 GHz. The fact that much of wireless communication occurs in this frequency range suggests the prospect of a wireless technology based on nanomechanical structures for signal processing.

It should be noted that high mechanical strength often comes with high thermal conductivities . Recent experiments have shown that the thermal conductivity of carbon nan o tubes can be as high as 3,000 to 6,000 W/m K at room temperature, which is more than twice that of diamond.

Because heat transport along the length of the tube is much higher than in other directions, carbon nanotubes can truly be called thermal wires, analogous to electrical wires . Therefore, by patterning carbon nanotubes in a composite material, one would be able to channel heat flow along a desired path, a feat that cannot be achieved in bulk solids because of diffusion in three dimensions.

It was recently shown by Alex Zettl and his group at the University of California, Berkeley, that the relative motion between the different shells of multi wall carbon nanotubes have some unique properties and can serve as excellent mechanical bearings that do not undergo any wear.

Because the spacing between the different shells is on the order of atomic scale, there is also no possibility of contaminants getting in the b ea ring, so it is ultraclean. Whereas MEMS-based microactuators have often suffered from high friction and stiction, this may not be a problem at nanoscales.

The fact that the tiny biological motors are the only natural objects that generate rotary motion suggests that evolution may have selected rotary motion at nanoscales due to its efficiency. It will indeed be truly remarkable if one could integrate carbon- nanotube or molecular bearings on MEMS devices so that efficient rotary or linear motion can be achieved.


A scanning electron micrograph of carbon nanotubes: The structures have elastic modulus on the order of 1 TPa and can sustain critical strain of approximately 5 percent before yielding.

Important Areas for Biotechnology

Although nanomechanics is generally associated with solids, the nanomechanics of biomolecules such as DNA and proteins is emerging as an important area for biotechnology. The structure and chemistry of biomolecules make reactions such as antigen-antibody, DNADNA, or DNA-protein binding highly specific.

It was recently shown that when such specific biomolecular reactions occur between receptor and target molecules on one surface of a microcantilever, the cantilever bends. What is remarkable is that this nanomecharucal technique is capable of detecting single errors in the DNA code as well as protein concentrations in the blood serum that are relevant for cancer detection. The ability of biomolecular reactions to generate motion also offers the possibility of biomolecule-activated mechanical switches and motors.

It was found that the motion is due to changes in intermolecular energetics and the entropy induced by the reactions. Hence, the deeper lesson that one learns from these observations is that, in contrast to experience at macroscales, mechanics and thermodynamics may be very closely related at nanoscales.

Since biology requires an aqueous environment, it is interesting to investigate the scales of intermolecular forces in liquids. It turns out that most of the dominant forces-electrostatic, van der Waals, steric-operate at scales generally less than 50 nm.

The mechanics of fluids at these scales are likely to be very different from those at macroscopic scales. Indeed, electrostatic forces can be l11.anipulated to produce fluid pumps and valves that have strong relevance in microfluidics. Since microfluidic devices are often used for biomolecular assays, mass transport in nanoscale liquids is an important area of investigation.

One of the biggest challenges in nanoengineering is how to integrate objects across different scales.

Without doubt, the fastest growing area of technology is currently information processing and storage. In the area of computing, the role of nanotechnology will probably be felt in the long term, because it will be difficult to beat silicon-based technology for the next 10 to 15 years.

Perhaps the short-term impact will be felt more in photonics and data storage. For example, by modulating the refractive index of a transparent medium at the nanoscale, one can synthesize photonic band gap structures that can block light completely or bend light through sharp turns or produce microlasers. This could help in miniaturizing optical communication devices.

In addition, by nanostructuring optically active semiconductors, one could produce lasers in the blue and the ultraviolet spectral range. This could dramatically increase the density of optical data storage. However, one needs to always consider competition from magnetic data storage, which currently has a density of about 50 Gbitslin2 using 100 nm magnetic domains. The challenge right now is to achieve 1 Tbitlin2 using bits 10 to 20 nm in size. It is unlikely that the current thin film magnetic medium will allow this level of density.

One approach is to pattern the storage medium with 10 to 20 nm magnetic nanoparticles, which will determine the bit size. But the challenge in this area is largescale integration of nanoparticles. Another promising approach, being developed at IBM Zurich, relies on an array of cantilever beams with sharp tips mounted at the end. When they are heated electrically, the tips make indents on a polymer film and thereby write bits of information. Readout is achieved by scanning the tip and recording the cooling when the tip enters an indent. The bit size is governed by the tip sharpness, which is generally on the order of 10 m11.

This technique falls under the general category of scanned probe storage devices that rely on some near-field tip-sample interaction. Although scanned probe data storage is known to be slow because of the low-frequency mechanical resonance of the cantilever beams, the use of multiple cantilevers in parallel can alleviate the speed problems. It is likely that several other scanned probe techniques will be developed in the future, which will make 1 Tbitlin2 certainly within reach. While the science of near-field interactions is relatively well-known, the bulk of the work remains in the realm. of nanoscale engineering of the novel data storage systems.

Finally, regardless of the area of nanotechnology, its success depends critically on the ability to observe, detect, and manipulate nanostructures and nanoscale phenomena. Design and fabrication of instruments that are low-cost and readily available are extremely important. Scanning probe microscopes have revolutionized the infield of nanoscale imaging and are likely to play a significant role in the future. However, the• key to SPMs is not in the microscope but in micro cantilever probes with nanoscale tips.

Recent work has led to multifunctional probes, which, besides topography, can detect thermal, electrical, magnetic, and optical signals at nanoscales. The engineering challenge now is to develop MEMS-based probes that integrate multiple functions on a single tip.

In addition, one must recognize two shortcomings of SPMs, namely: They cannot image in three dimensions and they cannot provide chemical information about the sample. Any technique that can achieve this with nanometer-scale resolution is likely to generate another revolution in imaging.

Although I have identified only a few nanoengineering issues, it is likely that as progress is made, new ones will emerge to be addressed. It is probably clear by now that the engineering at nanoscales must transcend disciplinary boundaries.

Progress cannot be made in isolation, but only through collaboration and synergy between engineers and scientists of different backgrounds. For engineers, understanding the fundamental science is critical, whereas scientists interested in nanotechnology must appreciate the engineering issues. This synergy must also be reflected in the educational programs that train the next generation of engineers and scientists.