There's nothing new about nanoparticles. Artisans used them as far back as 9th century Mesopotamia to generate glittering optical effects on the surface of pots. What is new is our ability to produce, control, and design new nanoparticles for industry.
One of the most promising areas of nanotechnology innovation lies in new energy applications, where nanoparticles can perform nano-processes — like separating hydrogen from oxygen in water or transforming photons of light into useable power — with far greater efficiency than larger molecules.
Over the last decade, researchers have developed a cookbook of recipes to bake up useful nanoparticles for all sorts of industrial applications. However, a comprehensive understanding of the "design principles" behind nanotechnology continues to develop. This is largely because nanoparticles are too small and form too quickly to be captured by microscopes or other imaging devices. For that reason, one of the best ways to understand how nanoparticles form and operate is through computer models and simulations.
Richard Hennig, a materials scientist and engineering researcher at Cornell University, is working to uncover the design principles at play in the formation of nanoparticles - in particular, nanocrystals relevant to energy applications.
Using the high performance computing systems of the National Science Foundation-supported eXtreme Science and Engineering Discovery Environment (XSEDE), as well as those at the Computation Center for Nanotechnology Innovation at Rensselaer Polytechnic Institute, Hennig and his research team showed that the concentration and location of small molecules (ligands) on the surface of lead-selenium nanoparticles cause the particles to form different shapes with different energy potentials. The results of the study were published in ACS Nano in February 2012.
Lead-selenium and other lead salts are a common and well-studied system used in photovoltaic cells. When the individual nanocrystals line up into periodic superstructures with long-range order, they maintain the ideal band gap and electronic properties to produce electricity from the sun. However, the nanocrystals can form a range of shapes and assemble into different superstructures that are more or less efficient. Hennig's study focused on two questions: what controls the shape of the nanocrystals and what controls their assembly?
By altering the concentration of ligands present when the nanocrystals form, Hennig and his team produced a range of shapes from octahedrons to cubes with cut-off corners. When the nanocrystals age in ligand-free solvent, the anisotropic change in surface coverage affects the assembly of the nanocrystals, changing the order of the superstructure from a face-centered cubic to a body-centered cubic assembly.
"Experimentally, you're a little bit at a loss because we don't really know how these nanoparticles interact with each other at that scale," Hennig said. Simulations help understand what actually controls the assembly of nanoparticles into different crystal structures. "We've been asking: What are the surfaces of these nanoparticles really like?"
According to Hennig's research, the nanoparticles are almost never bare. Small ligands attach to their surface, like hairs on a head. Their overall concentration and specific placement (including "bald spots") appear to control the shape of the larger crystal. These results, determined by computer simulations on the Ranger supercomputer at the Texas Advanced Computing Center using density functional calculations, were confirmed by laboratory experiments. By capturing the dynamics of multiple, interacting nanoparticles over time, with atomic resolution, the simulations provide additional detail on how nanoparticles behave.
"Ranger is one of the lead machines that has helped us over the last few years," Hennig said. "On Ranger, we can actually do the calculations on the scale that we need to understand these nanoparticles."
In their next round of simulations, the researchers are working with various experimental groups to select different ligands to place on the surface of nanoparticles.
Hennig likens the work to playing with Lego models. "The nanoparticles are like your Lego pieces and we change how these Lego pieces can stick together," he said.
Understanding the design principles that govern the formation of nanostructures, and designing these structures more rationally will ultimately speed up how scientists develop materials for photovoltaic cells, bio-medical applications and catalysts for batteries (another area where Hennig's group is active).
"We need to make sure we understand how nanoparticles assemble, and what shapes they take. If we understand how that works, we can design a shape that is preferred for certain applications," Hennig said. "Computation really speeds up that process."
After World War I, chemists learned how organic polymers derived from petrochemicals could be transformed through industrial processes to make a broad range of plastic products. Today, materials scientists like Hennig are working on a similar problem (albeit at a scale hundreds of times smaller) with the potential to provide an even greater benefit to society.
"There's a whole world out there of different structures that you can assemble by modifying what's on the surface of the nanoparticles," Hennig observed. "The open question is: What are the useful structures and what are the structures that are just interesting? That's what the simulations can help answer."
A version of this story first appeared on the Texas Advanced Computing Center website.