Brave New Nanoworld

This visualization of an electrically conducting, iodine-doped, carbon nanotube in which purple balls form an iodine chain is based on a Z-contrast scanning transmission electron microscope image.

Like toddlers learning to build complex castles from toy blocks, a growing number of researchers are studying ways to assemble materials and devices from atoms. If their products stand up, the applications could be truly astonishing. Already, the growth of tiny crystals from a relatively small number of atoms in the 1980s has brought us more protective sunscreens and better cosmetics. Improved electronic devices ranging from cheaper flat-screen televisions to palm-size computers that recognize speech could be next. Thanks to the use of new techniques that allow scientists to "see" and manipulate atomic and molecular building blocks, such new technologies may be possible.

Click here to see similar images

"It's amazing what one can do just by putting atoms where you want them," says Richard Smalley, co-discoverer of the buckyball in 1985 and winner of a 1996 Nobel Prize in Chemistry. Smalley, a Rice University chemist, is talking about nanotechnology, a new field focused on the very small that could become very big. It's a field that has been explored by just a few ORNL researchers for several years; however, more are expected to become involved because supporting nanotechnology and nanoscience research projects using ORNL's internal funds in the Laboratory Directed Research and Development (LDRD) program is a new initiative at the Laboratory. The more successful projects may qualify for federal funding that will likely come later. In FY 2001 budget guidance given to heads of federal departments and agencies, the White House Office of Science and Technology Policy identified nanotechnology as one of 11 R&D areas that are "important national efforts requiring coordinated investments across several agencies."

Nanotechnology is the study and use of materials, devices, and systems on the scale of a nanometer (a billionth of a meter, or 10^-9m). If researchers can learn to manipulate individual atoms at this scale, some experts believe the results could lead to a revolution in computing, electronics, energy, materials design, manufacturing, medicine, and numerous other fields. In testimony given on June 22, 1999, before the House Science Subcommittee on Basic Research, Eugene Wong, a National Science Foundation assistant director, called the nanometer "truly a magical unit of length. It is the point where the smallest manmade things meet nature." He suggested that, using atom-by-atom manipulation, scientists could change the properties of a material without altering its chemical composition. Instead of discovering a new phenomenon by accident, he said, scientists can now look for one systematically and soon may be able to design it to order.

Why Go Nano?

For most of this century, scientists have practiced "top-down science," a reductionist approach in which the goal is to simplify our understanding of matter by breaking it into its basic building blocks, ranging from quarks to neutrinos. But, in recent years, interest has arisen in complexity. Scientists want to know how simple atoms and molecules come together and arrange themselves to form complex systems, such as living cells that make life possible on earth. This "bottoms-up" science, which deals with how complex systems are built from simple atomic-level constituents, spawned nanoscience. It is the study of the properties of tens of or hundreds of atoms or molecules in a space with a diameter of less than 50 nanometers (a nanometer is about the size of two large atoms or four small ones).

Dave Geohegan and Alex Puretzky use laser ablation to form carbon nanotubes for potential use in improving electronic devices.

"This field really took off in 1986 after the discovery of the buckyball, which is a stable cluster of 60 carbon atoms," says John C. Miller, a section head and physicist in ORNL's Life Sciences Division (LSD), who has been doing research in nanoscience for 10 years. In 1993 Bob Compton, then in LSD, started a project to produce and study buckyballs using laser ablation, the method used to produce the first buckyballs. ORNL is a world leader in laser ablation, which is the most versatile and powerful technique available for synthesizing many nanomaterials.

As part of this project, in 1994, Dave Geohegan of ORNL's Solid State Division (SSD) and Alex Puretzky, a visiting scientist working with Compton, obtained the world's first digital photographs of how buckyballs form in a bubble of carbon atoms confined in argon gas. In 1995 ORNL researchers Don Noid, Bobby Sumpter, and others working in nanoscience developed a computer visualization of helium flow pushing a "buckyball piston" in a carbon nanotube. This image was published on the cover of the Nos. 1 & 2, 1996 issue of the ORNL Review.

Besides the need to build complex materials and devices from atoms and molecules, a major driving force behind the interest in nanotechnology is the desire to build faster computers that will more quickly handle complex calculations. Fortunately, nanofabrication techniques are enabling the construction of smaller and faster logic devices, which will be needed for such difficult tasks as speech recognition and voice synthesis.

Computers have been getting faster and cheaper in large part because of the silicon revolution. According to an observation in the 1960s by Gordon Moore, co-founder of Intel, the number of transistors being packed into integrated circuits (semiconductor chips) doubled every year. He predicted this trend would continue, and the observation became known as Moore's Law. Actually, since 1975, the number of transistors on a semiconductor chip has doubled roughly every 18 months, enabling microprocessors to work faster and memories to grow larger while computer prices plummet. The key to this trend is the incredible shrinking of transistors with each chip generation. By squeezing more transistors and other features into integrated circuits, the travel distances of electrons are shortened so calculations can be made faster using less power.

To keep up this current rate of improvement in computing power, the Semiconductor Industry Association declares that the minimum size of transistors, which is now around 250 nanometers (nm), must reach 130 NM by 2003 and 50 NM by 2012. Mike Simpson, an electrical engineer in ORNL's Instrumentation and Controls (I&C) Division, sees the problem this way:

"If you make devices too small using silicon, they will run up against physical limits such as unwanted quantum effects. For example, electrons could leak, or 'tunnel,' through insulating layers, causing circuits to short. In addition, to make smaller silicon devices, a fabrication facility costing as much as $30 billion may be required. Semiconductor chip fabrication could become prohibitively expensive."

So what are the answers? "One solution," Simpson says, "is to switch to materials other than silicon that take advantage of quantum effects to increase the speed and density of features on chips. Examples of alternative materials being explored at ORNL are spinach proteins for diodes and carbon nanotubes for pins and molecular wires. Then we must find ways to get nano-objects made of these materials to assemble themselves to bring down the costs of device fabrication."

Carbon Nanotubes: The Strongest Material

Since 1991, many organizations, including ORNL, have developed and tested various ways to make carbon nanotubes and put them to use. Carbon nanotubes are tiny strips of graphite sheet rolled into tubes a few nanometers in diameter and up to hundreds of micrometers (microns) long. The graphite has a network of hexagonal rings, leaving it with many unpaired electrons. Carbon nanotubes conduct electricity and heat amazingly well, so they are being considered for use as wires for nanosized electronic devices in future computers, charge-storage devices in batteries, and electron guns for semiconductor chip etching and flat-screen televisions and computer monitors. They also could store hydrogen gas to power fuel cells. Furthermore, carbon nanotubes are the world's strongest material in terms of tensile strength and they are lightweight and flexible. Consequently, they could be useful as ultralight structural materials for wings of advanced aircraft and space probes to make them stronger and more energy efficient. Thus, there is intense interest in finding a way to produce nanotubes in large quantities.

Geohegan and fellow SSD researchers Puretzky, Xudong Fan, and Steve Pennycook have synthesized these long, thin tubes through laser ablation and then studied the results. They use a pulsed ultraviolet or visible laser to vaporize a graphite target containing a small amount of metal catalyst in a quartz tube that is heated in an oven to over 1000°C. Streaming from the target is a "bubble" of 10¹⁶ carbon and metal atoms that is confined by a low-pressure background gas of argon. If the argon gas is cool, it forces the carbon and metal atoms to collide frequently with each other so that they form nanoclusters (up to 2 NM in diameter) and nanoparticles (up to 100 NM in diameter). But when the argon gas is hot, the metal transforms the carbon with astonishing efficiency into single-walled nanotubes, which consist of one layer of carbon atoms.

Until this past year, no one had witnessed this synthesis process. But in 1999 the SSD researchers used special imaging and diagnostic techniques (digital imaging, laser-induced fluorescence, spectroscopy) to determine what forms when during nanotube synthesis. Transmission electron microscope (TEM) images reveal the product of laser ablation: bundled ropes of single-walled nanotubes that resemble a pile of spaghetti.

"One of our goals," Geohegan says, "is to learn how to use laser ablation to control the length, structure, and electrical properties of carbon nanotubes we make for nanoelectronic devices being constructed at ORNL. Another goal is to understand how carbon nanotubes form and grow. This information could lead to methods for scaling up production to meet the need for ultralight structural materials that are one hundred times stronger than steel but have only one-sixth the weight."

The secret to making carbon nanotubes lies in the target — a pellet made of graphite powder and cement mixed with particles of a metal, such as cobalt, nickel, or iron. During laser ablation, the metal forms tiny nanoparticles that serve as seeds to catalyze the formation of the carbon nanotubes.

By obtaining a sequence of stop-action images of the laser ablation process and the carbon-catalyst plume, the ORNL researchers are addressing these questions: When do the metal catalyst atoms and carbon atoms disappear and form nanoparticles? Do nanotubes grow from atoms or nanoparticles? When does growth begin, and how fast do the nanotubes grow? What role does catalyst particle size play in the growth?

Using digitized snapshots and spectroscopy of the vaporized plume during nanotube formation, Geohegan and Puretzky are removing some of the mystery of nanotube synthesis, controlling the time available for growth, and producing short nanotubes for electronic applications. Together with Fan's TEM analysis of nanotubes produced under various controlled conditions, the researchers have measured the first growth rates of nanotubes made by laser ablation and gleaned clues as to how to scale up the process. This research led to a seed money project with Roland Seals of the Oak Ridge Y-12 Plant's Development Division to come up with a manufacturing process that would produce large quantities of carbon nanotubes for direct use in structural materials.

The Laboratory's initials have been constructed from carbon nanotubes using plasma-induced CVD on pre-deposited catalyst patterns. The letters are started by forming a pattern using electron beam lithography followed by electron-gun evaporation of the iron catalyst.

Doug Lowndes, Gyula Eres, and postdoctoral researchers Vladimir Merkulov and Yayi Wei, all of SSD, are using chemical vapor deposition (CVD) and plasma-enhanced CVD to make arrays of carbon nanotubes for use as field emitters of electrons for an ORNL-developed electron lithography method of making faster silicon chips (see "Carbon Nanotubes for Chip Writing?"). The first tubes grown in the summer of 1999 using CVD produced multiwalled tubes (tubes within tubes) that are microns in length and 20 to 40 NM in diameter. Lowndes and Merkulov also evaporated metal catalyst particles, such as nickel or iron, onto a silicon substrate and then exposed it to a carbon-bearing gas mixture to produce arrays of vertically oriented carbon nanotubes on silicon. The process is being fine tuned to produce multiwalled nanotubes as small as 10 NM in diameter.

The pink glow enveloping the square sample results from plasma-induced chemical vapor deposition, which is used to grow carbon nanotubes. The sample has a layer of a metal catalyst, such as iron or nickel, to promote nanotube growth.

In 1999, the researchers developed a nanotube transistor using a carbon nanotube to bridge electrically conducting catalyst electrodes fabricated by electron beam lithography. The multiwalled tube was grown between the electrodes using selective area thermal CVD.

"In addition to its great mechanical strength," Lowndes says, "the nanotube is capable of ballistic conduction, similar to photons flying through optical fibers. Electrons flow through a nanotube without scattering off any atoms, so they encounter hardly any resistance and lose virtually no energy. In fact, a superconducting current was recently induced in a carbon nanotube at low temperatures." Carbon nanotubes can carry an electric current as high as 10⁹ amps/cm².

Geohegan, Lowndes, Simpson, Phillip Britt of the Chemical and Analytical Sciences Division (CASD), and David Joy of ORNL's Metals and Ceramics (M&C) Division and the University of Tennessee have received LDRD funding to study carbon nanotubes. They seek to determine the effect of a nanotube's atomic-scale structure on its electronic properties. They hope to understand better how carbon nanotubes grow and how they operate as electronic wires and semiconductors. Such knowledge is needed to move nanotubes toward commercial application in electronic devices.

Nanoelectronic Devices:
Made of Spinach Proteins and Carbon Nanotubes?

In the 1990s, three researchers in ORNL's Chemical Technology Division — Eli Greenbaum, James Lee, and Ida Lee — gathered evidence that spinach is a plant with potential. They became particularly interested in a chlorophyll-containing protein in spinach called Photosystem I (PSI, pronounced PS One). They knew it performed photosynthesis using the energy of the sun to make plant tissue. But another amazing feature of this photosynthetic reaction center was that when it receives a photon of light, electrical current flows through it in one direction in 10 to 30 picoseconds — 100 times faster than in a silicon photodiode. Thus, spinach proteins could be used as a photo battery or solar electric cell.

The ORNL researchers' challenge was to show that PSIs could be lined up perpendicular to a metal surface such as that of an electrode. Then these 10-NM spinach proteins could be used as nanodiodes to transmit current in one direction and block it in the other in the absence of light, making possible the fabrication of biomolecular electronic devices. The researchers met the challenge, suggesting that the next generation of opto-electronic devices may be based on spinach, not silicon.

After isolating PSIs from spinach leaves, they learned how to make current flow along spinach proteins by depositing a platinum electrical contact on the end of each PSI. Next they showed that platinum anchors PSIs to a gold surface.

A major challenge was to orient the PSIs so that all the same ends point in the same direction either perpendicular or parallel to a metal surface. In 1997 the researchers found an effective way to achieve preferred orientation of PSIs: chemical treatment of the atomically flat gold surface on a mica substrate. They coated gold surfaces chemically with 2-mercaptoethanol and found that most of the PSIs point up, like lawn grass. The sulfur atom in this molecule attaches strongly to gold, and the molecule's other end selectively binds to the positively charged free end of a PSI, causing it to be perpendicular to the surface.

The next step was to build a biomolecular device. Greenbaum met with Simpson to plan the design of a hybrid nanodevice using spinach reaction centers as the diodes. Other ORNL participants in the project are Lowndes, Geohegan, and Britt. Their expertise is needed because the nanodiodes made of oriented spinach proteins are to be connected by nanowires made of carbon nanotubes.

Simpson and two University of Tennessee graduate students have been building molecular electronic devices that do signal processing and logic functions using spinach protein nanodiodes. Eventually, these devices will have wiring made from carbon nanotubes. These tubes are produced by laser ablation in SSD and delivered to the I&C Division as black soot in a jar.

ORNL's first biomolecular electronic device using spinach proteins as diodes was built in the spring of 1999. PSIs, each measuring 10 NM by 6 NM, stood like soldiers in a tunnel with their feet on the floor and their heads against the ceiling. The gold electrode was treated with a chemical that binds strongly to both gold and the positively charged free ends of the PSIs, making them point up to the platinum electrode to which they adhere. The PSIs were illuminated by pulses of light from the device, causing them to conduct a current in one direction and block it in the other. This current was measured on the device, showing that the spinach biochip worked.

"We made the first device with two electrodes only 15 nanometers apart and with PSIs oriented between the electrodes," Simpson says. "We have atomic force microscope images that show the PSIs are perpendicular to the electrodes."

At the Cornell University Nanofabrication Facility, Michael Guillorn, a UT graduate student who works with Simpson, recently constructed a molecular logic gate in which a PSI is placed in a 15-NM gap among three electrodes — two gold and one platinum.

While PSIs are building blocks for logic gates, it is hoped that carbon nanotubes can be used on the ORNL hybrid biochips as molecular wiring to connect the logic gates. "Carbon nanotubes," Simpson says, "also could be used for molecular components such as diodes, transistors, logic gates, or field emitters of electrons for supercomputer chips that will perform molecular computing."

Simpson sees carbon nanotubes as the solution for the input-output problem involved in setting up communications between nanodevices and the larger world outside. "Right now each chip is nestled on a larger block of pins that communicate through attached wires with the outside world," Simpson says. "If you have 18 million transistors or 10¹² gates on a chip, the device would have to be much bigger just to make space for that many more pins. So, we're proposing to replace the pins and wires with electron beams emitted by arrays of carbon nanotubes placed in an electric field both on and off the chip. Carbon nanotubes in arrays outside the chip would serve as field emitters of electron beams that 'talk' with an array of beam-emitting carbon nanotubes on the chip."

Simpson envisions a supercomputer the size of a grain of rice that is wired with carbon nanotubes. He sees it surrounded by arrays of carbon nanotubes that would beam data into the nanocomputer and retrieve from it any needed information.


Using a model of a carbon nanotube, Phil Britt shows how he and Shane Bromley chemically alter a nanotube with modified thiol molecules to get it to bind to the surface of a gold electrode.	Carbon nanotubes are chemically modified by heating them in a solution of alkanethiol, which enables them to assemble themselves in the proper orientation on a gold pad. Chemically treated nanotubes may someday be used as molecular wires in newly developed electronic devices.

To build a device with nanowires, ORNL staff members purify carbon nanotubes, cut them to the appropriate size range, and separate them by size. Britt and Shane Bromley, a UT graduate student who works with Simpson, treat them chemically by attaching molecules at their ends and on electrode surfaces, causing the nanotubes to assemble themselves in place in proper orientation.

Shane Bromley, a University of Tennessee graduate student, displays images of single-walled carbon nanotubes (yellow features) attached by sulfur-bearing (thiol) molecules bonded to a gold surface (reddish area) on a silicon-titanium substrate, as seen through the atomic force microscope shown in the photograph.

Carbon Nanotubes for Chip Writing?

Today's circuit patterns are etched on chips on silicon wafers by use of light — optical lithography. A mask containing a circuit pattern is imaged on each layer of every silicon chip, and a beam of light carves a circuit in the chip part not shielded by the mask. The size of the circuits made by optical lithography is limited by the wavelength of light.

Because the wavelength of electrons is so much shorter than that of light photons, an electron beam could etch a much narrower winding path in the chip, creating a finer, more closely packed circuit. Today's state-of-the-art circuits are about 200 NM wide, but ORNL researchers led by C. E. (Tommy) Thomas of the I&C Division think their chip-writing technique using precise electron beams from carbon nanotube-tipped field emitters could make a circuit only 100 NM across or even as small as 10 NM They believe that by 2004, they will meet the semiconductor industry's goal of making production chips whose circuits are 8 times denser and up to 16 times faster than chips of the same size currently being etched by optical lithography.

ORNL's proposed addressable-field-emitter array (AFEA) is a two-dimensional array of miniature electron beam sources. The sources will use carbon nanotubes made by Lowndes and Merkulov using plasma-enhanced CVD. Early tests by Larry Baylor of the Fusion Energy Division (FED) show that when the Oak Ridge carbon nanotubes are placed under a certain voltage, they are 10 times more efficient than amorphous diamonds as field emitters of electrons.

When a computer-controlled bias grid places an array of nanotubes under a voltage, it will emit electrons. When the voltage is dropped to zero, the nanotube array will stop emitting electrons. Each nanotube array is addressable by a computer, enabling the programming into the AFEA of the desired circuit patterns to be written onto the chip.

The chip-writing program will be allocated to a network of 100 parallel computers that will send turn-on and turn-off signals to the AFEA logic and memory circuits connecting the cathodes. This digital "mask" can be reprogrammed to create different circuit patterns on new layers within milliseconds.

Making New Smart-Surface Materials for Miniature Devices
Using Ion Beams

A team led by Lynn Boatner of SSD has demonstrated the concept of creating "smart nanocomposite surfaces" on inactive materials to make them both "sensors" and "actuators." The scientists used accelerators to implant ions of vanadium and oxygen in a single crystal of aluminum oxide (Al₂O₃), or sapphire. Then the sapphire host was appropriately heated in the correct atmosphere to remove ion-implantation damage and induce the growth of dispersed rod-like crystallites of vanadium dioxide (VO₂). Because the sapphire has a smart nanocomposite surface that can both detect and react to light or heat, this patented material could be a candidate for a number of optical-device applications, including "smart windows."

In a smart window application, intense light like that from a powerful laser strikes the nano-composite surface and rapidly heats up the embedded VO₂ particles, transforming them into a metallic "mirror-like" state. Instead of letting the incoming light through, the window now reflects it back in the direction of its source. Such smart windows might be used, for example, to shield a satellite's sensitive internal optical components from either accidental damage or intentional sabotage.

This interference color optical micrograph shows voids and defects on the surface of an oriented single crystal of sapphire implanted with vanadium and oxygen ions and then subjected to rapid heating.

"Our patented sapphire surface nano-composite contains nanometer-thick crystallites of vanadium oxide that can be up to microns in length," Boatner says. "The vanadium oxide crystallites are embedded nanophase precipitates that effectively 'activate' the near-surface region of an inactive material like Al₂O₃. The VO₂ crystallites can both sense a signal and respond to it, making the surface nanocomposite a classic smart material."

Boatner and Tony Haynes of SSD lead a team of SSD and Engineering Technology Division (ETD) researchers who recently received LDRD funding to apply this smart surface-layer concept to the development of new miniature devices. Because the surface nanocomposites are formed by ion implantation and thermal processing, the research will require the development of appropriate ion implantation conditions as well as heating and cooling rates for the thermal processing of a wide variety of host materials.

In this art-glass vase crafted during the reign of the Chinese Emperor Chi'en Lung (1736-96), very small single crystals of either gold or copper produce the beautiful red-colored glass by scattering light through a process first explained in G. Mie's 1908 theoretical work.

Four-nanometer-diameter gold particles like those shown in the transmission electron microscope image would produce a light pink color (top right), while larger ~30-NM-diameter gold particles would produce a deeper red color (bottom right). Though unaware of it, artisans were applying nanotechnology for hundreds of years before the electron microscope revealed the presence of the gold or copper nanoparticles responsible for the red coloration of art glass.

The SSD and ETD researchers will look at ways of applying vanadium oxide nanocomposites to integrated optical devices, including switches and attenuators. ETD's Steve Allison has envisioned several configurations of fiber optics and thin-film integrated optics in which these nanocomposites may be used to modify and control light signals in new and useful ways. Such materials can also store data recorded on them by a laser beam, because at each point heated above the VO₂ transition temperature by the beam, the surface reflectivity not only changes but also does not return to its original value. Optical-transmission effects resulting from the phase transition in a VO₂ nanocomposite surface can also be controlled by overlying the surface with a transparent thin film of tin oxide as a resistive heater. When an applied current heats up the tin oxide film, the VO₂-host nanocomposite surface changes from the transparent to the reflective state, creating an optical switch.

Haynes and Boatner and their SSD associates, including Al Meldrum and John Budai, have been studying iron nanocrystals implanted in yttrium-stabilized zirconia (YSZ) for their magnetic and ferromagnetic properties. "The beauty of having iron particles embedded just below the surface," Haynes says, "is that they don't oxidize — that is, rust — because they are protected from the atmosphere by the surrounding host material. In addition, crystalline YSZ is an ideal substrate for iron particles because it lines them up crystallographically. The result is a smart ferromagnetic material that senses an applied magnetic field and responds by changing its state and retaining a magnetic field proportional to the field that was originally applied."

SSD's Frank Modine and Shinichi Honda showed that the intensity of specially polarized light passing through iron nanocrystals in YSZ changes when a magnetic field is applied (an effect known as magnetic circular dichroism). They plan to study nanocrystals of three other magnetic materials — nickel, cobalt, and magnetite (Fe₃O₄) — embedded in zirconia. "If we could make a material with evenly spaced 10-nanometer magnetic particles," Haynes says, "we might be able to develop the highest-density and most stable data storage disk on which data could be read and written by using magnetic fields." Most recently, Jim Thompson, a visiting scientist from the University of Tennessee and one of his doctoral students, K. J. Song, have been using the ORNL superconducting quantum interference device magnetometer to characterize new ferromagnetic nanocomposite surfaces.

ORNL researchers also plan to use the ion-implantation-thermal processing technique to make "shape memory" nanocomposites in which the precipitated particles can change and restore their shape in response to temperature variations. Such shape-memory surface nanocomposites might find applications as actuators in microelectro-mechanical systems (MEMS) devices. Magnetostrictive particles that change their size when a magnetic field is applied could also be formed in appropriate hosts for similar micromechanical applications.

SSD researchers led by Woody White, a co-principal investigator in this LDRD project, have also created other types of surface nanocomposites by using ion implantation and thermal processing to form semiconducting and metal precipitates embedded in a variety of host materials. "We are learning how to control both the size and location of metallic and semiconducting nanocrystals in a material to obtain desired optical and electrical properties," White says. "Size is important because it determines a semiconducting nanocrystal's ability to absorb and emit light. Because size determines the color of emitted light, full-color flat panel displays for computers may someday be made of appropriately sized semiconductor nanocrystals formed by ion implantation."

To have even better control over nanocrystal precipitate size and location in a matrix and to develop new devices based on fiber-optic, thin-film, and integrated chip configurations, ORNL scientists will use Vanderbilt University's new finely focused ion beam facility in collaboration with Len Feldman, an ORNL Distinguished Visiting Scientist from Vanderbilt. This finely focused ion beam can implant atoms into dots as small as 5 NM at precisely controlled locations. "Ion implantation," White says, "could be useful for making modulators and optical switches and as a good prototyping tool to verify that nanocrystals of the desired size and spacing can be made to produce materials with desired properties."

Molecular Broom and Quantum Dots

John Wendelken and Zhenyu Zhang's first foray into nanoscience resulted in a story in the February 16, 1998, issue of Chemical & Engineering News. The work involved using a scanning tunneling microscope (STM) probe to induce molecules to move in a specific direction. In an STM, a probe scans across a surface, holding a gap of less than 1 NM between the STM tip's last atom and the surface atoms. By maintaining a constant current with a sensitive feedback mechanism, the undulations of the STM tip as it is scanned across the surface are interpreted as an atomic-level image of the surface.

In one set of experiments performed by Wendelken and his former postdoctoral researcher Larry Pai in late 1997, the STM was used to image a silver substrate exposed to low-pressure ferrocene gas [Fe(C₅H₅)₂]. The STM tunneling current decomposed the gas, separating iron from the hydrocarbon molecules. In this same way, the STM could be used to fabricate iron-containing nano-scale structures having arbitrary shapes. At the same time, it was found that when ferrocene gas was decomposed on another surface, ring-shaped cyclopentadienyl (C₅H₅) radicals were deposited on the silver substrate. The ORNL researchers were the first to show that these C₅H₅ rings could be imaged and manipulated with the STM at room temperature.

Pai discovered that the STM probe could be used as a "molecular broom" that "sweeps" C₅H₅ rings to the side, perpendicular to the probe's scan direction at room temperature. The discovery came shortly after SSD theorist Zhang predicted that the C₅H₅ rings would make quantum jumps perpendicular to the scan direction on the silver surface. Indeed, Pai's STM images show the rings jumping across the silver surface, leaving a zigzag track.

ORNL researchers showed that cyclopentadienyl (C₅H₅) rings could be imaged and manipulated with a scanning tunneling microscope (STM) at room temperature. They discovered that the STM probe could be used as a "molecular boom" that "sweeps" C₅H₅ rings to the side, perpendicular to the probe's scan direction. As the probe moves, the rings make quantum jumps across the silver surface, leaving a zigzag track.

We don't know exactly why the rings move in steps the way they do," Zhang says, "but we believe the repulsive force between the tip and the negatively charged radicals perturbs the rings, giving them enough energy to jump over the natural barrier to their forward motion as they try to diffuse."

This process of using an STM to make iron-containing nanoclusters and depositing C₅H₅ rings represents an inefficient, serial approach to nanofabrication, or self-assembly of nanoscale structures. The SSD group has received LDRD funding to develop a parallel approach to nanofabrication to speed up the production of useful devices. Zhang, Wendelken, and Jian Shen hope to make a large number of quantum dots — each a nanocluster of a few hundred atoms — that have the same size and are spaced equally. The idea is to take advantage of natural Coulomb repulsion in which two particles with electrical charges of the same sign repel each other.

The researchers are developing a "buffer layer assisted growth" process to enable self-assembly of nanoclusters with a uniform size and spacing. They have been depositing iron atoms on a solid buffer layer of frozen xenon gas on copper and silicon substrates at low temperature. Xenon was chosen as a buffer layer because it does not absorb electrons, provides a high-mobility surface for atom diffusion, and isolates the growing quantum dots from the substrate. Zhang predicted that the iron atoms would slide around on the xenon layer and form three-dimensional islands, or nanoclusters. The researchers use a low-energy electron source to deposit electrical charges on the nanoclusters. According to Zhang's theory, the charges will limit the growth of the nanoclusters, keeping them at about the same size because additional charged atoms moving their way would be repelled. In addition, because the charged nanoclusters are about the same size, they will repel each other the same amount, resulting in an array of equally spaced, immobilized atom clusters.

In a preliminary test of Zhang's prediction, Wendelken and Pai used the UT-ORNL variable temperature STM, which they operated at 30K. After they warmed the sample to remove the buffer gas, they found that the iron clusters on the copper substrate were substantially smaller and more uniform in size when the clusters were charged than when they were not. However, these clusters were not equally spaced. Zhang says, nevertheless, that even this improved size distribution is enough to enhance the optical properties of the material, suggesting it might have potential for quantum dot lasers as nanoscale sources of light for optical circuits.

Why was the ordering not as good as Zhang predicted? "We had to remove the electrically insulating buffer gas so the clusters could be imaged by the STM," Wendelken says. "It may be that the desorption of the gas disturbed the nanocluster pattern. When the variable-temperature STM atomic-force microscope we have ordered arrives, we will be able to image an insulating surface and see if a highly ordered pattern exists before the xenon is desorbed."

One likely use for perfectly ordered patterns of magnetic quantum dots (such as iron or cobalt) may be in magnetic high-density storage disks that could hold 100,000 times more data than current disks.

Quantum Drops

Materials that normally don't mix can now be blended within single microparticles and nanoparticles, thanks to a droplet technique discovered at ORNL for producing arbitrarily sized particles. The technique may lead to new materials with tunable properties with applications to drug delivery systems, improved coatings, and faster components for electronic devices.

In this molecular dynamics simulation, a nanosized polymer droplet is formed in a solvent and forced through a micron-sized orifice. Submicron polymer particles exhibit unique properties that could make them extremely useful for optical displays and industrial coatings.

In 1998 Don W. Noid, Bobby Sumpter, and Michael D. Barnes, all CASD researchers, developed a novel way to prepare electrically charged, submicron-diameter, spherical composite particles of organic polymers of nearly uniform size. They started by forming droplets of a dilute solution of two polymers that do not ordinarily mix together. For droplets less than 10 microns in diameter, the solvent evaporates faster than the polymer molecules could disentangle, resulting in homogeneous mixed-polymer particles. Optical probes were used to determine the material homogeneity, size, and dielectric constant of the particles. The work showed that material properties of the particles could be tuned simply by adjusting the mixture of polymers in solution.

Barnes, Noid, Sumpter, Thomas Thundat of the Life Sciences Division (LSD), and M. Alfred Akerman of ETD have received LDRD funding to refine this technique to make "quantum drops," clusters of one or more different types of charged polymer molecules predicted to act like "artificial atoms" with tunable electronic properties. Electrons, like photons, possess both "wave" and "particle" properties; the characteristic (DeBroglie) wavelength of a particle (usually very nearly zero for macroscopic objects) is inversely proportional to its momentum. When electrons are confined to a "box" whose dimensions are comparable to their wavelength, discrete, or "quantized," energy levels are observed whose spacing increases with decreasing "box" size. Thus, such systems are termed artificial atoms because the colors of light they absorb or emit depend on the size of the particle. The researchers will investigate whether these spherical 2- to 10-NM drops (which may contain more than 10,000 atoms) exhibit size quantization behavior similar to that of more familiar semiconductor quantum dots, as predicted by computational simulations.

"When we model quantum drops," Noid says, "we can predict their electron energy levels and their chemical potential, or electron affinity — that is, whether they have the correct symmetry to make it easy to attach electrons. We can predict the effects of electric and magnetic fields on the drops. We can predict the melting points of the particles in the drops, which become lower as particle size gets smaller."

"We think our drop technology will enable us to tune, or select, the size of the drop particles and the number of electrons on each particle," Barnes says. "The particle size is determined by the droplet size and the concentration of polymers in the droplet solution. The material properties of the quantum drops are determined by the polymer composition.

"The electronic properties of quantum drops, such as their ability to emit light of a particular color, can be tuned by adjusting or selecting the particle size or the number of excess electrons. The resulting quantum drops could be a new class of luminescent particle."

In this computer simulation, electron orbits sit on a polymer droplet. Such computations could guide the development of new materials for possible use in flat panel displays and new storage media.

Sumpter notes that his simulations predict that the hardness of a polymer can be tuned by mixing the hard material with a squishy polymer using the droplet technique. Mixing polymers to make particles can also alter the index of refraction or strength and compressibility of each material.

Quantum drops could be used for applications ranging from catalysis to quantum computing. In the gas phase these particles could speed up reactions because their high surface area increases the opportunity for contact and electron exchange between the particles and the reactant species diffusing throughout the reaction chamber.

Quantum dots theoretically could be tuned to make each electron represent a bit of information: Each electron with an up spin could be a "1" and each electron with a down spin could be a "0." "We propose to use the droplet technique to encode several memory elements on each nanoparticle," Barnes says. "If we are successful, this technique could be useful for the proposed development of a quantum computer."

"We have to learn how to make a variety of mixed-polymer particles and sort them by size," Noid says. "Then we must use spectroscopic tools to measure their electron energy levels experimentally and verify that the mixed-polymer particles behave like artificial atoms, as predicted by our simulations."

New Semiconductor Devices Using Micromechanical Quantum Wells

Imagine going outdoors one dark night and seeing the vivid colors of the scene as if the sun were shining. A hand-held night-vision device to make this possible may emerge someday from the LDRD project of a team led by Panos G. Datskos of ETD. Datskos, Slo Rajic (ETD), H. M. Meyer (ETD),Thomas Thundat (LSD), Ray Zuhr (SSD), John Wendelken (SSD), David Zehner (SSD), Bill Butler (M&C), and Don Nicholson (Computational Physics and Engineering Division) will be developing and studying a new class of tunable semiconductor devices that rely on nanoelectromechanical systems (NEMS). These devices may be used to generate power and detect electromagnetic fields and toxic chemicals.

At the microscopic level, a slice of silicon or some other semiconducting material will bend when zapped with high-energy photons of light. This ORNL-discovered photo-induced effect that quickly results in a microscopic mechanical stress in the material also causes changes in it at the nanoscale level.

Joe Cunningham uses a diamond-turning machine for one of the steps of making a device based on a microelectronic mechanical system.

For the NEMS project, ORNL researchers will apply their experience in making MEMS devices using focused ion beam milling and diamond turning, their theoretical and modeling tools, their materials analysis techniques, and their ability to form nanostructures and thin films. Relying on semiconducting materials, such as silicon, germanium, silicon nitride, gallium arsenide, and indium antimonide, they will fabricate thin-film nanostructures that confine electrons in microscopic springboards (microcantilevers).The team hopes to show that by using photons to bend microcantilevers and applying an electric field, the electrons trapped in or blocked by these nanostructures — quantum wells, quantum barriers, and quantum contacts — will be permitted to flow.

One of the team's goals is to begin developing a revolutionary class of uncooled photon detection devices by building an infrared detector that will operate at room temperature. This device can be tuned to specific far infrared wavelengths. It will be smaller and use less energy than today's detectors, which are cooled by liquid nitrogen. It will use photo-induced electronic stresses in quantum wells. When incoming photons stress the device's microcantilevers, its quantum wells will shrink in width, causing the energy levels of electrons inside the wells to rise enough for the trapped electrons to hop out and flow. The size of the resulting electrical current will be proportional to the intensity of the photons.

This series shows a metal-oxide-oxide-semiconductor photovoltaic device in which an electron, energized by a photon of sunlight (below) hops over a multiple quantum barrier after one barrier is lowered by an electric field.

The team believes that tunable pulsed-power devices can be fabricated using quantum barriers made of layers of insulating material. An electric field can be used to raise and lower quantum barriers, which block electron flow. When the field lowers a quantum barrier, electrons energized by photons start to flow. Such devices could produce power using sunlight.

Another of the team's goals is to create a new kind of semiconductor device — a nanomechanical single-electron transistor (SET). A SET might replace today's field-effect transistor, which relies on a gate electrode to pass or block current flowing from a source electrode toward a drain electrode. In a SET, incoming photons will bend two microcantilevers meeting each other at a quantum point contact, where electron flow is blocked. The mechanical stress will raise the energy level of the electrons enough to push them across the quantum contact. SETs will be useful in smart nanosensors needing on-off switches and in data storage.

Quantum Computing by Connecting the Dots

The world's first practical computation on a fabricated nanoscale device may be achieved by a team led by Jacob Barhen of ORNL's Center for Engineering Science Advanced Research (CESAR) and the Computer Science and Mathematics Division (CSMD). The team has received LDRD funding to develop a quantum dot array to perform this task. For this project, quantum dots are clusters of atoms, a few nanometers in diameter, surrounded by an insulating layer. The relatively small size of the metallic dots planned for the ORNL array will enable operation at room temperature (because no cooling will be needed).

Using colloid chemistry, Leon Maya fabricated these 2-NM platinum quantum dots, which appear as round yellow spots in this atomic force microscope image.

Leon Maya, a team member from CESAR and CASD, has already fabricated platinum particles about 2 NM in diameter. For this project, gold particles approximately 1.5 NM in diameter will be produced by trapping them in tiny cavities of polymer molecules.

Because of his knowledge of the structure of DNA, which he has studied using an atomic force microscope, Thomas Thundat will use a DNA molecule as an architectural template for positioning quantum dots on a surface between two electrodes to program the device. "By attaching complementary four-base-DNA sequences to the quantum dots," Thundat says, "we should be able to place them at predetermined spatial locations on the DNA backbone with sub-nanometer precision."

Gold electrodes that provide access to the quantum dots from the macroscopic environment will be fabricated. The two electrodes, which will anchor each end of the predesigned DNA sequence, will be laid on a flat substrate and separated by a narrow gap a few nanometers in length. A novel technique will be used to decompose the DNA template without altering the geometrical position of the electrodes and quantum dots.

The practical goal of the ORNL team is to build a device that emulates a neural network. Instead of the neurons and connecting synapses found in the brain, the nanoscale computer will depend on electrically charged quantum dots connected by electrons that tunnel between them at different rates.

Using colloid chemistry, Maya will fabricate and coat the gold quantum dots. Thundat will assemble them using the DNA templates and an atomic force microscope. Jack Wells of CESAR/CSMD and David Dean and Michael Strayer, both of the Physics Division, will develop a first-principles simulation of the device on ORNL's IBM-SP3 parallel supercomputer. "The simulation," Barhen says, "should provide immediate and invaluable feedback to the experimental design and help us accurately specify the parameters needed to use the device as a computer.

Barhen, Yehuda Braiman, Vladimir Protopopescu, and Nageswara Rao, all from CESAR/CSMD, will develop the methodology and algorithms needed to implement neuromorphic computations on the quantum-dot array. This implementation involves innovative techniques that modify the electron tunneling rate between dots. For demonstration purposes, the researchers intend to solve a pattern recognition problem. An example of such a problem would be the analysis of sensor data to identify seismic patterns indicating the presence of a porous sandstone layer that might contain oil.

"This project," Barhen says, "is motivated by the information processing needs of future generations of intelligent systems. On one hand, there is a need to meet the tremendous constraints on power consumption, size, and temperature. On the other hand, novel sensors, such as hyperspectral cameras used for imaging landscapes, require ever more powerful, dedicated processors. Thus, we are targeting applications that can exploit the emergent collective computational properties of an ensemble of nanostructures."

Self-Assembling Triblock Copolymers

Triblock copolymers are made by joining three chemically distinct polymer blocks (large molecules), each a linear series of identical monomers (small molecules). Because the blocks may be thermodynamically incompatible, they will separate on the nanometer scale (5-100 NM), producing complex, ordered nanostructures by self-assembly. Examples of these structures are the lamellar morphology in which the three blocks form alternating sequential layers; a regular arrangement of spheres of one or more blocks in a matrix of another; and similarly ordered rods in a matrix or layered structure. A particularly interesting morphology is the continuous core-shell gyroid structure. An electrically conducting polymer core could be formed in an insulating shell, embedded in a matrix of a third block, providing other mechanical or electrical properties. The properties of these self-assembled nanostructures can be manipulated by independent selection of the three triblock components, the block sequence, and block length.

Block copolymers can be configured into a nearly limitless number of molecular architectures based on two, three, or more monomer types.

Recent advances in synthetic chemistry have made possible the controlled production of compositionally uniform triblock copolymers. Frank Bates and his associates at the University of Minnesota will be synthesizing a variety of triblock copolymers using block sequences based on various existing polymers. As part of this LDRD project, George Wignall of SSD and Tony Habenschuss of CASD will use powerful scattering tools, especially small-angle neutron scattering, to probe the structure of the triblock copolymers, while Bates' group will measure their mechanical, optical, electrical, ionic, barrier, and other physical properties.

Some morphologies (e.g., shapes such as layers, cylinders, and spheres) for linear ABC triblock copolymers, the ability to select block sequence (ABC, ACB, BAC), composition, and molecular weights provides a significant opportunity to create new morphologies.

Core-shell gyroid structure showing a conducting polymer (green) encased in an insulating polymer (yellow) inside a matrix (blue) made of another polymer.

"Our first goal is to understand the effects of A, B, and C homopolymer additives in controlling the morphology in ABC triblock copolymers," Habenschuss says. "Our second goal is to apply multi-continuous three-monomer block copolymer membranes to fuel-cell technology."

In some fuel cells, electrolytic membranes must allow passage of protons between anode and cathode while restricting the flow of fuel, such as methanol. Today's proton-exchange membranes are very expensive and are prone to leaking fuel while allowing protons to pass. An alternative proton-exchange membrane must combine suitable ionic transport with solvent resistance, mechanical strength, and processability at a competitive cost. Because the constituent blocks can be selected independently and several multi-continuous morphologies are known, ABC triblocks and blends represent an attractive opportunity for improving this technology.

Research on these triblock copolymers could lead to other applications. "Nanoscale reinforced, ordered polymer composites" could be made in which one component could be a high-strength reinforcing material such as nylon. Triblocks could be used as "templates" by modifying the copolymers after synthesis so that they retain their morphology but possess properties different from those in the original materials. "Nano-porous materials" could be made by leaching out channels to create filters having controlled channel size (block length), or, conversely, by leaching out the matrix component to get an open aerogel-like structure. The research at ORNL and the University of Minnesota should help generate the scientific knowledge base needed to make practical use of these marvelous materials.

ORNL researchers will continue to explore our brave new nanoworld to find out how best to exploit its out-of-this-world features.