Some features of the structure, the dynamics of nuclear motion, the nature of electronic states, excited-state energetics and relaxation phenomena in "isolated" nonmetallic clusters are reviewed, with an emphasis on the interrelationship between the characteristics of molecular and condensed phase systems.
PrologueThe search for unifiable concepts in chemistry and physics constitutes a central scientific endeavour. Each of the traditional areas of solid-state physics and molecular physics rests on a set of such unified concepts. In traditional solid-state physics one encounters the ubiquity of ordered structures with (approximate) symmetry of a periodic space group [I]. In addition, and of considerable interest, positional and compositional disordered materials [2] also exist. The density of states for an elementary excitation, e.g., phonons, electrons and excitons, is continuous spanning a finite energy range. Finally, the characteristic length of elementary excitations, e. g., phonon wavelength, electron mean free path and coherence lengths, are considerably shorter than the sample dimensions. On the other hand, molecular physics deals with finite sets of bound atoms, whose structure can be specified in terms of permutation symmetry [3] and (approximate) point group symmetry [4]. The fundamental vibrational excitations are discrete, being classified in terms of normal and local modes. The electronic excitations are also discrete below the first ionization potential. These discrete features of the low-lying excitations in molecular systems are qualitatively different from the continuous level structure in solids. The atomic structure and the level structure of elementary excitations in clusters, i. e., finite aggregates [ 5 , 61, cannot be characterized in terms of the conceptual framework, which rests solely on one of these two traditional disciplines. The area of the physics and chemistry of clusters, which is the subject matter of this Discussion Meeting, bridges the gap between solid-state and molecular physics. I would like t o dwell on some features of level structures and dynamics of "isolated" clusters. In 1977, when modern cluster physics was in the embryonic stages, Friedel wrote [7]: "Clean and isolated clusters in vacuum are a theoretician's dream". This dream has come true! The remarkable progress in the areas of supersonic jets and cluster beams [8, 91 makes clean and isolated clusters amenable t o experimental interrogation, providing the basis for intensive and extensive experimental and theoretical effort. There are some compelling reasons for the study of isolated clusters.A . Exploration of new basic physical phenomena. These are finite systems with a congested spectrum of energy levels (for electronic states, phonons, etc.) [lo]. The level structure can be varied continuously by changing the cluster size. Accordingly, one can accomplish a "continuous transition" from molecular to solid-state systems.B. Microscopic approach to macroscopic phenomena, such as nucleation, adsorption and desorpti...