Algorithms for predicting the structural properties of clusters

Predicting the structure of atomic clusters is one of the outstanding problems in condensed matter physics. Traditional theoretical approaches are hindered by the large number of degrees of freedom, and the lack of symmetry in these systems. Some new computational techniques for predicting the structural properties of small silicon clusters will be illustrated. The emphasis of this effort is on simulated-annealing procedures based on Langevin dynamics. Quantum forces, derived from ab initio pseudopotential calculations, are incorporated in these simulations. These forces can be efficiently calculated using higher-order finite difference methods.