We discuss the control of the kinetics and dynamics of chemical reactions by the solvent, from a molecular point of view. The kinetics are discussed using a transition state theory (TST) approach, applied to the reactants and their surrounding solvent as one supramolecule. The topics discussed include a molecular interpretation for the changes that take place when one solvent is being replaced by another; the use of local against normal vibrational modes and/or joint description, i.e., local modes for part of the system and normal modes for the other part; and the effect of pressure on the rate in solution. The notion of free volume and volume of activation is extended to a more general phase space in which geometrical volumes may overlap, the approximations that are inherent to cell theory are examined and a molecular interpretation for internal and chemical pressures is suggested. The link to the dynamics is provided by an analysis of the breakdown of TST due to diffusion/cage control of the rate of the reaction. A unified description which interpolates from activation control to diffusion control is presented with a special emphasis on the motion within the solvation cage. Results of molecular dynamics simulations for both activated and activationless reactions are presented. The very detailed computer experiment is interpreted using a reduced mechanical description and the separation of time-scales is discussed using an adiabatic separation of variables. Spectroscopic methods for probing the different time epochs are suggested. The rather short duration typical of the motion within the solvent cage is emphasized, and the possibilities that this affords for studying the short-time dynamical role of the solvent via experiments in clusters or in glasses are noted.

Unimolecular processes can be described as the decay of an ensemble of N excited resonances coupled to K decay channels. Resonances are metastable states characterized by a complex energy whose real part is the position of the state along the energy axis while the imaginary part gives the individual decay rate of the state. Resonances usually overlap in the RRKM regime. The degree of overlap is measured by the parameter R = /dE where is the average of the individual decay rates of the excited resonances and dE is the average spacing between their position. In the exact degeneracy limit, that is, for an infinite value of R, (N-K) resonances have a zero width, so that a fraction of the initial excitation remains permanently trapped in the bound subspace. This trapping effect subsists in the non degenerate case but is not complete. We use a random coupling effective Hamiltonian model to discuss the effect of the degree of overlapping R, and of the number of resonances N and decay channels K, on the temporal evolution laws of the bound subspace and of the fragments. The decay law of the bound subspace and the temporal evolution of the yields in fragments exhibit several time regimes. This is due to the fact that after the diagonalization of the effective Hamiltonian, the decay widths of the resonances cluster into one group of K large widths and one group of(N-K) small ones. The trapping effect is due to the (N-K) small widths. The amount of trapping depends on the value of the degree of overlapping R, and for a given value of R, on the ratio N/K: large values of R and of N/K correspond to a large amount of trapping in the bound subspace for times long when compared to h/. The temporal evolution laws of the yields in fragment are also strongly affected by the degree of overlapping and the value of the ratio N/K. Due to the reorganization of the partial widths which follows the diagonalization of the effective Hamiltonian, we show that the nature of the dominant product can change while increasing the value of R and N/K. We also discuss the time evolution of the sampling of phase space for a specific preparation in terms of these two parameters. The volume sampled is computed using an entropic measure. When the resonances overlap, there is not enough time to completely sample phase space prior to dissociation. The fraction sampled decreases as the amount of trapping in the bound phase space increases.

An iterated map which mimics the dynamics of a high Rydberg electron revolving around an anisotropic ionic core is described. The map specifies the change in the quantum numbers of the electron due to its passage near to the rotating core. Attention is centered on the limiting case of physical interest where the rotation of the core is faster than the orbital motion of the electron. While the map does provide for a very efficient way to numerically simulate the motion, its main advantage is in that it delineates the various dimensionless coupling parameters that govern the dynamics. To make contact with many experiments, external electrical and magnetic fields are included in the Hamiltonian. The stretch of the kinetic time axis due to the presence of external fields is discussed. The full map can be further approximated by a one-dimensional map which captures the essence of the dynamics. The primary aspects having to do with the three-dimensional character of the actual motion are incorporated in the magnitude of the dimensionless coupling parameters. A simple but realistic limit of the one-dimensional map is discussed which can be considered as the electron undergoing diffusion in its energy. The mean first passage time out of the detection window and the branching fractions for ionization vs. stabilization of the electron are computed in the diffusion approximation. As is experimentally observed, the lifetime of the high Rydberg states exhibits a maximal value when plotted vs. the energy.

Molecular dynamics simulations are presented showing that four-center reactions with high activation barriers occur readily, and in a concerted manner, under conditions of cluster impact. At the high velocities which prevail inside an impact heated cluster, we show that one can conveniently examine the reactive event as a sequence of elementary, suddenlike steps. The importance of the timing of these steps is emphasized. The bond-switching step is described by a kinematic model, which is shown to account for the energetic and steric requirements and for energy disposal of the concerted four-center reaction. Three different cases are examined in detail: the bimolecular N-2 + O-2 and H-2 + I-2 reactions and the unimolecular norbornadiene –> quadricyclane isomerization. The cluster is shown to have a significant role not only in activating the reactants but, equally important, in stabilization of the energy-rich nascent products. The event is over when the cluster beings to fragment, which occurs after a small number of collisions of its constituents. Comparison is made between clusters of different rare gases. In terms of the overall reactivity, all rare gases are similar and the role of the mass can be compensated by a suitable scaling of the initial velocity of impact. There are, however, differences related to the mass ratio of the reactants to that of the rare-gas atom. The high yield of the four center concerted mechanism in impact heated clusters and the very essential role of the cluster in overcoming the constraints on such a mechanism suggest that the designation `'cluster catalyzed reaction'' is appropriate.

A computationally tractable approximation for both interstate and intrastate dynamics is derived and applied. The correlation between the electronic and nuclear degrees of freedom is explicitly allowed for in that there is an equation of motion for the nuclear dynamics on each electronic state. These equations for the intrastate dynamics are coupled due to the interstate interaction. The exact equations are derived from a quantum mechanical Hamiltonian and are then simplified by assuming that the coupling between the different electronic states is localized and that, in the absence of interstate coupling, the nuclear motion on each electronic state is classical-like. Equations for the populations and the phases of the different electronic states are also derived. Coupling of the nuclear modes to a classical solvent is included in the formalism and the main computational effort is in the mechanical description of the solvent. As a computational example, a simulation of a fast pump-fast probe for an iodine X –> B (bound) transition, in rare gas solvents, is presented and discussed. Despite the long range of the B state potential of iodine, which enhances the effect of the solvent on the excited state dynamics, there is a finite delay before the coupling to the solvent is manifested. The delocalization of the optically prepared state markedly slows down as the density is lowered. At longer times there is considerable energy exchange with the solvent. As a result many molecules either gain enough energy to dissociate or are cooled down, depending on the temperature and density of the solvent. At the higher densities, many molecules which attempt to dissociate are caged.

Solvent-induced electronic predissociation (X –> B –> al(g)((3) Pi) state) of molecular iodine is discussed using a classical ensemble representation of Heisenberg's equations of motion. Excitation of the intermediate B state by an ultrafast pulse creates a coherent vibrational motion in this bound state. The localized solvent-induced coupling to the a state results in the spawning of dissociation products which occurs in bursts, twice per vibrational period. Equations of motion for both the electronic and nuclear degrees of freedom in each electronic state are derived from a quantum mechanical Hamiltonian. These equations are coupled whenever two electronic states are interacting. The formalism includes coupling to the surrounding classical solvent. Comparison with a pump-probe experiment is provided.

Air is shown to burn (in theory, using two complementary procedures) under the unusual combination of conditions made possible within a cluster heated by a supersonic impact at an inert surface. Both clusters of neat N-2/O-2 and clusters containing several N-2 and O-2 molecules within a rare gas envelope have been studied. The principal reaction is N-2+ O-2 –> 2NO which proceeds via a four-center mechanism. The four-center reaction N-2 + O-2 –> N2O + O leads to N2O which is quite hot, as is to be expected on the basis of kinematic considerations, and is efficiently destroyed as the cluster expands. During the early, compression, stage multi- (>4) center reactions readily occur. The results of molecular dynamics simulations using a many-body potential are well accounted for by a distribution of products of maximum entropy subject to conservation of energy, matter and charge.