STRUCTURE PREDICTION. Clive M. Freeman, Alan M. Gorman and Steve M. Levine, MSI 9685 Scranton Road, San Diego CA 92121, USA.

Techniques for predicting and solving crystal structures based on powder diffraction data and simple descriptions of interatomic interaction have advanced dramatically in recent years. Such procedures exploit the standard techniques of computer simulation in combination with hybrid potential energy functions to yield structures which are sterically and experimentally plausible. These methods have been especially useful in the investigation of microcrystalline framework structured materials for which powder diffraction may be the primary experimental route to structural information. In contrast to traditional refinement procedures, which use rapidly convergent but locally biased least squares methods, analogy with physical simulations has prompted the use of simulated annealing as an optimization method in the majority of studies. Metropolis Monte Carlo or Molecular Dynamics based annealing procedures provide significant `searching' capabilities which yield global minimization properties. Applications to framework structured solids, condensed metal oxides and molecular crystal structures will be described.

Acknowledgement: The MSI Catalysis and Sorption Project is supported by a consortium of industrial, academic and government institutions. We thank the membership for their ongoing advice, guidance and support.