COMPUTATIONAL MATERIALS DESIGN: SYNERGY OF FIRST-PRINCIPLES CALCULATIONS AND EXPERIMENTAL DATABASES. E. Wimmer, Molecular Simulations, Inc., Orsay, France

Progress in computational methods combined with the development of novel experimental techniques have created exciting opportunities for materials design. Both computations and experiments are generating data at an increasing rate. Thus, it is of utmost importance to define and implement comprehensive data models to accommodate these results and to create software systems which can operate on this wealth of information. The first part of this contribution provides an overview of current theoretical and computational methods for the prediction of materials properties [1,2]. A particular emphasis is on first-principles density functional methods, which have become truly remarkable tools for structural predictions for a wide variety of systems including organic molecules, organometallic compounds, semiconductors, metals, and ionic compounds. In addition to giving structural information with an accuracy of a few hundredths of one Ångstrom, quantum mechanical methods generate detailed information on the electronic structure and related properties, which is usually complementary to the results of experimental methods such as x-ray diffraction methods, x-ray photoemission spectroscopy, scanning tunneling microscopy, and vibrational spectroscopies. This complementarity can be exploited for analytical purposes, for example in solving crystal structures. In the second part, the capability of present computational methods are illustrated in the context of semiconductors, inorganic pigments, and materials for energy storage. The third part provides an outline of a data model to capture both experimental and computational information. A perspective on emerging computational methods, computer technologies, and communications will conclude this contribution.

1. E. Wimmer, J. Comp.-Aided Mat. Design 1, 215 (1993).

2. E. Wimmer, Science 269, 1397 (1995).