NCSA NEWS

Computational Chemistry Comes of Age With Nobel Prize

by Peter Taylor, San Diego Supercomputer Center and Balaji Veeraraghavan, National Center for Supercomputing Applications

The emergence of computational science has added a third manner of performing scientific research. The existing methods were experiment and theory, closely coupled: without theory, there is no way to systematize and analyze the results of experiments; without experiments there is no point in theorizing--theory becomes philosophy, rather than science. Computational approaches provide a third way, often allowing faster and more elaborate tests of theory than would be possible by experiment, but also allowing a much more sophisticated analysis of experiment than theory alone could provide. In this sense we have a tripod: three legs, which provide a much more stable base than two.

This year's Nobel Prize in Chemistry, awarded to Walter Kohn and John Pople, affirms and legitimizes the status of the third leg of that tripod. As scientists working to extend the uses of computational science, we are very pleased that these chemists have been recognized at the highest level for their theoretical and computational contributions in a field whose outstanding contributors have long been laboratory chemists.

The work of John Pople and of Walter Kohn provided the foundations for what was then a relatively new approach to research in chemistry. During the last two decades, an increasing number of chemists have turned to the computer to predict the results of experiments before they are performed, or to help interpret the results of experiments. Computer simulations have become an essential item in the researcher's toolbox.

One of the most influential strategies for predicting molecular properties based on its electronic structures is referred to as ab initio methods, where no assumptions are made other than the fundamental laws of quantum mechanics and the fewest approximations are used in solving the resulting equations. Such methods can be made sufficiently accurate to rival many experiments, yet they can be applied in practice to chemical reactions and species that are general enough to interest the majority of chemists.

Pople's path to using ab initio methods to do chemical calculations passed through an intermediate stage, in the 1950s and 1960s, of using so-called semi-empirical methods. Experimental data are used as input to the calculations to try to improve the reliability of simpler and less accurate methods. Pople realized early on that such semi-empirical approaches were and are very successful, but the ultimate limits on accuracy mean that they do not always produce reliable results.

Soon Pople started focusing his research efforts on ab initio methods for solving electronic structure of molecules. These research efforts led to development of the now well-known GAUSSIAN program, which includes extensive functionality for solving electronic structure of molecules ab initio at various levels of approximation and for predicting a wide range of molecular properties. His systematic research studies on a wide range of chemical systems and their properties led to wide acceptance of these computational approaches by the entire chemistry community. Also, the simple user interface that allows chemists to focus on molecules and properties of interest, efficient algorithms that permits quick studies on large molecules, and early free distribution of the GAUSSIAN package contributed to the wider acceptance.

No computer program has had such wide-ranging effects on a scientific community as GAUSSIAN has had in the world of chemistry. For the first time, chemists with little or no experience in computing could run calculations themselves to help with the planning and understanding of their experiments and to predict the properties of new chemical species. Since 1970 the GAUSSIAN program has undergone many additions and improvements, and has become a commercial product (distributed by a company Pople helped found, although he is no longer connected with it). It is still the most widely used ab initio electronic structure program, although it now has a host of competitors.

The contributions of Walter Kohn are different, but equally pioneering. In the mid 1960s he was able to establish mathematically that the energy of a quantum-mechanical system, like a chemical species, could be obtained using a much simpler formula than had previously been believed. Further, with other collaborators he suggested a means whereby this simple formula could be used in practical calculations, leading to a class of ab initio methods called "density functional" approaches. In the last two decades, computers grew powerful enough to make density functional methods easily applicable, and work by a number of researchers (including Pople) produced density functional approaches that were both practical and reliable. Such calculations are probably now the most common type performed. The density functional methods have thus contributed in full measure to the widespread use of calculations in chemistry. In addition, the density functional methods bridged the gap between the molecular chemistry and condensed matter physics, which is very timely because of the current wide interest in nanomaterials, which requires knowledge of both molecular chemistry and condensed matter physics.

Both of the NSF PACI Leading Edge Sites at SDSC and NCSA support many remote scientists whose work is made possible by the pioneering research of Pople and Kohn. Ab initio and density functional methods are widely used today in chemistry, materials science, chemical engineering and bimolecular sciences. These scientific fields are among the largest consumers of time on the high-end computers at the two LESs.

Impact of Kohn and Pople's contributions to Alliance Researchers

Researchers at the National Computational Science Alliance say their work would not be possible without the methodologies of Nobel Prize winners John Pople, professor of chemistry at Northwestern University, and Walter Kohn of the University of California at Santa Barbara. Much of the modern electronic structure research is founded on the methodology and algorithms developed by Kohn and Pople.

Within the Alliance, the chemical engineering, nanomaterials, and molecular biology Application Technologies (AT) teams utilize Pople's and Kohn's broad array of techniques to anticipate the behavior of molecules. These approaches to quantum chemistry and molecular behavior allow engineers to take the fundamental properties of molecules, build off them, and accurately estimate the molecular behavior in a physical context, such as a chemical plant. This alternate process of scientific study supplants expensive lab research, yet yields equally satisfying results. Today, Kohn's density-functional system and Pople's methodologies are the backbone of the Alliance's Chemical Engineering Workbench, a collection of chemical engineering tools tied together in a Web-based interface.

About the Nobel Prize Winners

John A. Pople was born in England and remains a British citizen. He is currently Professor of Chemistry at Northwestern University in Evanston, Illinois. He obtained his Ph.D. at Cambridge in 1951 and served as director of Britain's National Physical Laboratory before moving to Carnegie Institute of Technology (later Carnegie-Mellon University) in 1964. He has been at Northwestern since 1986.

Walter Kohn was born in Austria. He is currently a professor in the Institute of Theoretical Physics at the University of California, Santa Barbara, where he has been since 1979. He obtained his Ph.D. from Harvard, and held an appointment at Carnegie Institute of Technology before moving to the University of California, San Diego, in 1960, where he remained for almost 20 years.

 

More News Briefs