Features:

NANOPARTICLE RESEARCH PROVIDES FUEL FOR CATALYST DESIGNS

The new year offers promising discoveries in a relatively new science, nanotechnology. While sci-fi writers predict nanorobots will be able to invade our cells in order to conquer the world for the forces of evil, physicists in the University of Pennsylvania's chemistry department and Laboratory for Research on the Structure of Matter apply complex nanotechnology simulations to pollution control for the forces of good. They predict they will be able to help clean our environment by continuing research on the properties of metal nanoparticles, also known as "nanoislands" or "clusters," which could someday soon lower emission rates in automobiles' catalytic converters. Metal nanoparticles might also prove useful in hydrogen fuel cells, systems proposed by many as a future alternative to combustion engines.

Andrew Rappe, who leads this research project, is no stranger to this realm. Since 1995, he and his team have been conducting computational experiments on nanosurfaces using NCSA resources. His team includes graduate students Valentino Cooper, Sara Mason, and Myung-Won Lee; post-docs Yashar Yourdshahyan, Na Sai, and Ilya Grinberg; and Russel Kauffman, a physics professor at Muhlenberg College in Allentown, Pennsylvania.

"Without NCSA resources, we would not be as effective," Rappe says.

Through their first-principles computations, Rappe's team explores the effect of materials modification on the properties of metal nanoparticles supported on oxide substrates. The main goal is to gain a fundamental understanding of how particle size and composition influence the structural, electronic, and chemical properties of the supported particle. Supported nanoparticles are an important subject because they can be used to gain insight into complex real- world systems, bridging the gap between fundamental research and future catalyst applications. Currently, the team can model the fundamental chemical reactions that a catalytic converter must perform, in the presence of a simplified but fairly realistic model of a nanoparticle catalyst. This process allows the team to predict with confidence which particle sizes and compositions could lead to better catalysts than exist today.

More than just empty space

In studying supported nanoparticles, the Rappe team's work on vacancies below the surface of palladium has already provided significant results. Transition metals like palladium are the active ingredient in the catalytic converter. Pollutants adsorb, or stick, to the surface of the transition metal. The special properties of the transition metal also lower barriers to reactions that form gases that are not pollutants. The non-polluting gases desorb and leave the converter. In their research project, the team calculated what happens to a palladium surface when an atom is removed from one of the layers, creating a vacancy. The team wanted to know how the properties of the vacancy change with its depth and with the concentration of the vacancies in each layer.

They found that the energy required to create a vacancy increases with the depth of the vacancy. They also found that the atoms surrounding the vacancy relax toward it. Rappe described this phenomenon as being like teeth moving in toward the space in your mouth where a tooth has been lost or pulled. This relaxation is bigger for vacancies near the surface than for vacancies in the bulk of the material. Subsequently, Rappe and Kauffman found that the bonds between the atoms surrounding the vacancy are strengthened as a result of this inward relaxation. These results are important because vacancies are fundamental defects in materials, and on surfaces, and this project studies how they affect each other. This work demonstrates how changing the structure of a metal changes its properties, opening the door to designing metal structures with the ability to process pollutants more effectively. Detailed results of these calculations have been accepted for publication in Physical Review B.

The main computational tool used in this research is a computer code that finds the configuration of electrons that minimizes the total energy of a system of atoms. The code then moves the atoms and minimizes the total energy with respect to their positions. Over the course of the team's studies, they consumed about 500,000 hours of computing time on NCSA's SGI Origin2000.

"These results," said Kauffman, "are both solid and interesting. [They] help to bridge the gap between the idealized surfaces that theorists study and the real surfaces with defects that are actually used." The results also open the door to further investigation. Questions yet to be answered include: How do vacancies change chemical properties? More specifically, how is the reactivity of a surface affected by the presence of surface and subsurface vacancies? What happens when the material is deformed or strained?

Possible islands of paradise

A challenging research project looking at platinum nanoclusters on aluminum oxide (alumina) by Rappe, Yourdshahyan, and Cooper is in a more nascent stage. The goal of this research is to understand the growth process of platinum clusters on supported oxide material and their catalytic behavior for the design of nanocatalysts. By using a modified alumina surface for the manufacture of platinum nanoclusters, engineers might someday be able to create a more active catalyst that is less susceptible to poisoning by the sulfur often created by combustion.

Ceramic materials like alumina have many distinguishing properties. They are reliable and widely used as support material in many technologies, from catalytic converters in our cars to dental and bone implants. Another reason for the vast technological importance of alumina is its abundance--only oxygen and silicon are more abundant in the earth's crust. Finally, it's relatively cheap. The fact that it also appears in nature as sapphire and ruby may give it a pricey reputation, however it is far less expensive than other materials such as platinum.

Studying the adsorption of platinum atoms and growth processes on alumina surfaces presents a challenging case, since the structure of the form of alumina used in automotive catalysts is still a matter of controversy. The Rappe group, therefore, uses the well-known, stable phase of alumina, known as a-alumina, as a model for their study. By mapping the potential energy surface they are able to identify the most energetically stable adsorption sites of platinum atoms on the alumina surface, knowing that atoms search for areas of lowest potential energy. They study surface modification of alumina to enhance the three-dimensional cluster formation of platinum atoms on the surface. This work demonstrates that each different platinum cluster size has different structural and chemical properties. The Rappe group plans to examine how particular nanoislands of platinum can be selected for desired pollution control reactions.

Yourdshayan uses the SGI Origin2000 for this project, as well.

A new approach

In addition to these computational research projects, the Rappe group is experimenting with computational theory itself. Traditionally they have used, and will continue to use, density functional theory (DFT), a computational method that can model between 50 and 100 atoms. This technique has proven itself repeatedly as a powerful, efficient, and accurate method of analyzing and predicting phenomena in metallic, semiconducting, insulating, and molecular systems.

For certain fundamental systems, it is important to achieve higher accuracy than DFT can provide. Therefore, Rappe's team has begun using the Quantum Monte Carlo (QMC) technique, a relatively new computational method. QMC's near-perfect accuracy is impressive, but currently it is less efficient and more computationally expensive than DFT. Despite these drawbacks, the team finds the opportunity to use QMC in conjunction with DFT "exciting and promising."

There is little doubt nanotechnology will become an increasing part of industry and our everyday lives this year and in the future. The next time we read an article about nanotechnology, or more likely, the next time we find ourselves stuck in traffic, we might think not of evil. Instead, we might remember the team of computational scientists in Pennsylvania working for a better environment, one atom at a time.