Chemistry—the study of molecules— is the science of the everyday world. For the world at large, molecules are the fundamental units of matter. An understanding of the structure, interactions, and reactions of molecules is thus of critical importance to a wide range of phenomena, from the fate of contaminants in the environment, through the production of plastics from crude oil, to the occurrence and treatment of genetic diseases.
Chemists are now about to cross a remarkable threshold and expect a dramatic expansion in their ability to make reliable predictions about molecular structure and processes. This sea change is due to the confluence of advances in theoretical, computational, and experimental capabilities, allowing chemists to understand and characterize matter at detailed atomic and molecular levels. By integrating chemists' capabilities of synthesis and characterization with computational modeling and simulation, it will soon be possible to use computation to design molecules to do what we want, and to control how we make them.
The computational chemical sciences group in the Department of Energy's Center for Computational Sciences at ORNL is working to fulfill this vision. Real-world chemical problems are complex and a multi-disciplinary approach embracing chemistry, materials science, solid-state physics, mathematics, and computational science is essential, as is a strong connection with experiment. Excellent examples of the interplay of theory and experiment are advances being made at ORNL in molecular optics and electronics. Using ink-jet printing techniques proposed by computational chemists, an experimental group led by Mike Barnes of ORNL's Chemical Sciences Division, including researchers from the University of Tennessee (UT) and Georgia Tech, coaxed stretched molecules of polyphenylene vinylene to form an array of glowing antennas on a glass substrate. These semiconducting polymer nanostructures have many potential applications including optical wires. Members of the computational chemical sciences group, including Bobby Sumpter, Bill Shelton, and Jack Wells, are performing detailed simulations to better understand this newly discovered nanostructure.
is led by Robert J. Harrison, the principal architect of the Northwest
Computational Chemistry Software (NWChem) code for massively parallel
computers, for which he received the IEEE Computer Society's 2002 Sidney
The other SciDAC project, entitled "An Integrated Approach to Multi-scale Modeling of Molecular Electronic Devices," is led by Peter Cummings of Vanderbilt University. This multidisciplinary effort involves researchers from four universities (including UT) and ORNL. The goals of this computational project are to develop and apply tools to understand how to construct and control molecular electronic devices by self-assembly of molecules on surfaces. Molecular electronic devices may someday replace silicon-based devices in future computers. High-performance computer programs are being developed by Vincent Meunier to use quantum theory to predict, for instance, the current across a single molecule. Fundamental new theoretical models are also being developed to confront a profound challenge to chemists—the wide range of time and length scales, which cause severe problems as molecular devices are extended to the nanoscale and beyond.
To probe nanoscale devices, design new catalysts, and develop clean energy sources, chemists nationwide see a critical need for multi-scale methods and leadership-class computing to provide them with a "computational microscope."
Web site provided by Oak Ridge National Laboratory's Communications and External Relations