Researchers are using a variety of new nanofabrication techniques to introduce pillars, columns, rods and other nanostructures on polymers, ceramics and metals.
Our increasing ability to access the nanoworld is profoundly affecting the way humans communicate, utilize energy and combat disease.
One way to create a nanostructure is to ‘cut it out’ of a larger piece of material; this is known as the ‘top-down’ approach. Two techniques commonly used to do this are electron-beam and ion-beam lithography.
Scientists are also having some success in scaling down manufacturing techniques, such as injection molding, to introduce nanoscale features onto a surface.
In three NSF-funded projects, Lehigh scientists are using nanofabrication techniques to introduce nanostructures onto the surface of a metal, a ceramic and a polymer.
Their goal is to develop technologies that could one day have an impact on our everyday lives.
Moving closer to an all-optical network
As consumers demand ever-faster communications and data processing, electrical engineers are seeking to replace today's electronics with photonic (or light-based) circuitry.
Electronic devices, however, contain nanoscale components, while the optical fibers in conventional microscale photonic circuits are orders of magnitude larger.
As photonic circuitry decreases in size, the dimensions of their optical components approach the wavelength of light (450 to 700 nanometers for visible light). This causes the light waves to diffract and the signal to be lost.
One way around this problem is to exploit a type of electromagnetic wave called a surface plasmon polariton (SPP), which forms when a light wave couples (or sets up a resonance) with collective oscillations of the electrons in the surface of a metal.
Because of the high electromagnetic fields associated with this resonance, these electromagnetic waves are confined to a very thin region along the metal surface.
“In order to build photonic devices that can store, reroute and transmit data, we must first find a way to trap these electromagnetic waves,” says Filbert J. Bartoli, professor of electrical and computer engineering. “One way to do this is to slow them down to a standstill.”
Light can be slowed down by passing it through a material with a high refractive index. And because different colors, or wavelengths of light, travel at different speeds, a prism can split a beam of white light into its separate colors.
Could this principle be applied to plasmonic waves?
Bartoli and Qiaoqiang Gan ‘10 Ph.D., now assistant professor of electrical engineering at the State University of New York in Buffalo, set out to find a way to control the refractive index of the interface along which a plasmonic wave travels.
After performing theoretical simulations, Gan predicted that the lightwaves could be trapped with a nanograting – a structure consisting of a series of nanogrooves cut into a metal surface. By gradually increasing the depth of these grooves along the grating, the frequency of the collective oscillations in metal's surface electrons decreases.
“At some point along the grating, light of a certain wavelength can no longer propagate, bringing the SPP to a standstill,” says Bartoli. “SPPs of different wavelengths would be trapped at different positions along the grating, enabling white light to be separated out into its separate colors.”
Using a focused ion beam generated in the FEI 235 dual-beam microscope in Lehigh’s Nanocharacterization Laboratory, Gan fabricated a series of nanogratings with grooves 150nm wide and 475nm apart.
The milling time for each successive groove was increased gradually to achieve a grating with a linearly graded groove depth. A nanoslit was fabricated 13 microns from the edge of the graded nanograting to allow a beam of white light to launch surface plasmon modes on the top surface containing the nanogrooves.
Under normal circumstances, observing a beam of light that is tightly confined at the metal surface would be impossible with an optical microscope.
“Fortunately for us, imperfections in the surface of the nanograting cause some of the light to be scattered into the far-field,” says Bartoli, “and it is this scattered light that we hoped to observe with our optical microscope.”
To simplify matters, Bartoli and Gan tested the grating on incident light comprised of two colors, red and green. This was done by first passing the incident beam of white light through a filter with transmission bands centered at red and green wavelengths.
“To our delight, we observed a colorful emission of red and green light at different positions along the grating,” says Bartoli. “This is an unambiguous experimental demonstration of ‘rainbow trapping’ in plasmonic nanostructures.”
“Now that we’ve managed to trap these electromagnetic waves, the next step is to find a controllable way of releasing them.”
Improving the performance of LEDs
According to the U.S. Department of Energy, Americans spend more than $37 billion a year on residential and commercial lighting, accounting for almost 22 percent of total national electricity consumption. The government hopes to phase out incandescent lighting in favor of fluorescent, but the search is on for better alternatives.
The most promising contenders are gallium nitride-based light-emitting diodes (GaN-based LEDs). These small solid state devices contain several extremely thin layers (15-500nm) of gallium nitride, aluminum-gallium nitride and indium-gallium nitride deposited onto a sapphire substrate using metal-organic chemical vapor deposition. The wavelength of light they produce can be controlled by changing the ratio of In or Al to Ga in each layer, providing access to the whole spectrum of visible light.
Several challenges stand in the way of widespread adoption of LEDs. Improvements are needed to the internal quantum and light-extraction efficiencies of GaN-based LEDs, and manufacturing costs must come down.
Most commercial GaN-based LEDs are grown at high temperature, typically 1,000 degrees C, on a highly polished sapphire substrate.
“Direct growth of GaN on sapphire at such temperatures often results in a poor quality film containing numerous defects known as threading dislocations that grow right through the GaN layer,” says Helen Chan, professor of materials science and engineering. “This is due to stresses generated in the film which arise from the large lattice mismatch (16 percent) between the crystal structures of GaN and sapphire.”
To overcome this, manufacturers deposit a buffer layer of GaN at lower temperature prior to growing the high temperature GaN layer. This process usually adds 30 to 45 minutes to the processing time and has a significant impact on the cost of fabrication.
Chan, Richard Vinci, professor of materials science and engineering, and Nelson Tansu, associate professor of electrical and computer engineering and a faculty member in the Center for Optical Technologies, have developed a cost-effective method of fabricating GaN-based LEDs on nanopatterned sapphire substrates.
“Rather than use a GaN buffer layer, we fabricate an ordered array of single-crystal sapphire nano-islands onto the sapphire substrate,” says Chan. “These modify the nucleation and growth mechanism of the GaN, resulting in a defect structure that is less detrimental to the optical properties of the device.”
Producing a high-quality array of single-crystal nano-islands would be difficult using standard etching techniques. The team adopted a novel approach, using electron beam lithography to create the nanopatterned array in a thin film of aluminum metal that had been evaporated onto the sapphire substrate. The aluminum nano-islands were oxidized at about 450 degrees C and converted into polycrystalline Al2O3, and then heated to 1,200 degrees C to induce grain growth, converting the structures into single-crystal sapphire.
“While this sounds like a long drawn-out process, it must be remembered that one of the major contributions to the cost of fabricating GaN LEDs is the time spent in the growth chamber,” says Chan. “All of our surface nanopatterning is done beforehand, which could lead to significant cost benefits.”
GaN-based LEDs grown on these nanopatterned substrates show a 24-percent improvement in output power over LEDs grown on conventional GaN templates. The increase is attributed to improvements in the device's internal quantum efficiency.
Lessons learned from fabricating GaN-based LEDs on nanopatterned sapphire substrates could contribute to the development of low-dislocation GaN material for solar cells, thermoelectric devices and smart-grid power electronics.
Fine-tuning the development of adult stem cells
Nanopatterning a polymeric surface could help scientists control the growth of adult stem cells and develop transplantation-based therapies.
Some researchers have learned that the mechanical stiffness of a flat substrate on which stem cells are placed has a profound effect on their subsequent structure and function. Nerve cells thrive on soft surfaces, while cartilage cells prefer harder surfaces.
“Instead, they are confronted by topographies that vary on the nanoscale. Mimicking the cellular environment during early stem cell differentiation may provide a way to control the process and determine the type of cells that grow.”
Jedlicka and John Coulter, professor of mechanical engineering and mechanics, are trying to develop ‘off-the-shelf’ nanostructured polymeric surfaces, comprised of an ordered array of nanopillars or nanogrooves and designed to support specific types of stem cell differentiation.
“The mechanical properties of each surface will depend on the height or depth of these features and the spacing between them,” says Coulter, who chairs the International Micro/Nano Molding Technical Group of the Society of Plastics Engineers.
Coulter’s group was one of the first in the world to develop an injection molding process to fabricate nanostructures on a thermoplastic polymer.
“The trick is to create a suitable mold,” says Coulter. “In this case, we introduced the desired nanofeatures into a silicon mold using a combination of electron beam lithography and ion etching.”
To ensure that none of the thermoplastic polymer remained stuck to the mold, a thin film of plasma-polymerized ‘release’ film was then deposited onto the surface of the mold. This mold was then attached onto a micro-injection molding system.
“If you want to produce off-the-shelf nanostructured polymer surfaces at low cost, you have to produce them in large numbers,” says Coulter. “Injectionmolding in this respect is the only way to go, as the cycle time is around 15 seconds.”