Abstract

The discovery of fullerenes (Kroto et al. 1985) and the experimental evidence of the so-called magic cluster sizes (Knight et al. 1984, de Heer 1993), more than 25 years ago, were at the origin of unceasing investigations on small clusters and nanoparticles and have contributed to the emergence of nanosciences, at the crossing point of several branches such as physics, chemistry, or even biology. In view of their high surface-to-volume ratio, clusters possess properties, different from those of bulk matter, that are very sensitive to their size and shape, rendering them very attractive from both the fundamental and technological points of view. In particular, the metallic species of a few nanometers in diameter disclose electronic properties intermediate between those of molecular systems and those of bulk matter. Whereas the sparse energy levels are quantized in atoms and molecules and continuously distributed in the energy bands of the crystal (Ashcroft and Mermin 1976), they tend to bunch together in metallic clusters and thus pattern the so-called electronic shells. This shell structure was evidenced experimentally (magic sizes) in the early 1980s in several metals (de Heer 1993) and nicely interpreted in the frame of the jellium model (Brack 1993). However, it was shown that magic sizes can also be correlated with atomic shell closures (Martin 1996) and even that both electronic and geometric structures may compete with each other, depending on the temperature (Martin et al. 1990).