Density Functional Theory and Car-Parrinello Molecular Dynamics for Metal Clusters

Density-functional theory (DFT) has provided the basis for many computations of physical and chemical properties of clusters made up of metal atoms ("metal clusters''). The chapter is divided into two parts. In part 1, comprising Sections 2 to 5, we give an outline of DFT: we describe various
implementations, comment on the accuracy of several approximations, and discuss recent advances and perspectives. In particular, we describe in some detail the Car--Parrinello method, which unifies
molecular dynamics techniques and DFT, and discuss several methods currently used to optimize the structural properties of an atomic aggregate. In part 2, comprising Sections 6 to 11, we review the applications made over the past 15 years or so on metal clusters of all different classes that have so
far been studied. Particular emphasis is put on the main issues (solved or not) that are specific to each class, and also on the results of ab initio molecular dynamics. Comparison with experiment is made whenever possible. Finally, an outline is given of future perspectives for DFT computations of metal clusters.