A new method of using quasiclassical trajectories to study the dynamics of elementary reactions is described. Trajectories are initiated in the phase space of a suitably chosen transition state and run forwards and backwards in time from the same starting point to simulate a complete collision. Calculations on a wide range of collinear A+BC reactions involving vibrationally excited reagents reveal that the optimum choice of transition state is a periodic orbiting dividing surface (pods) for which the action over one cycle of the pods is (v+0.5)h The method is extended to three dimensional reactions using the adiabatic periodic reduction scheme to find pods on fixed angle potential surfaces. The complete transition state is defined by joining these pods together. Methods for pseudorandomly sampling the transition state are described and the combined transition state theory-quasiclassical trajectory (TST-QCT) method is applied to the H+H2(v), N+N2(v) and F+H2(v = O) reactions at constant temperature. The TST-QCT method produces relative quantities directly, absolute values are readily obtained using transition state theory. The results of the new method are compared with conventional quasiclassical trajectory studies in the literature. Agreement is very good and the combined method brings about a very great saving in computer time by eliminating trajectories which fail to reach the strong interaction zone as well as revealing the extent of vibrational adiabaticity between reagents and the transition state. Finally, a modification to the TST-QCT method to allow the simulation of fixed collision energy reactions is described and tested on the F+H2 reaction.

No description available.

No description available.

No description available.

No description available.

No description available.

No description available.

No description available.

No description available.

The prediction of an atomistic system's macroscopic observables from microscopic physical characteristics is often intractable, either by theory or computation, due to the intrinsic complexity of the underlying dynamical rules. This complexity can be simplified by identifying key mechanisms that drive behavior and considering the system in a reduced representation that captures these mechanisms. Through theory, this thesis examines complex relationships in structured assembly and reaction mechanisms that occur when effective interactions are applied to mesoscale structures. In the first part of this thesis, the structure and assembly of soft matter systems are characterized while varying the interpenetrability of the constituent particles. The nature of the underlying softness allows these systems to be packed at ever higher density, albeit with an increasing penalty in energy. Stochastic equations of motion are developed in which mesoscopic structures are mapped to single degrees of freedom through a coarse-graining procedure. The effective interactions between these coarse-grained sites are modeled using stochastic potentials that capture the spatial behavior observed in systems governed by deterministic bounded potentials. The second part of this thesis presents advancements in time-dependent transition state theory, focusing on chemical reactions that are induced by oscillatory external forces. The optimal dividing surface for a model driven reaction is constructed over a transition state trajectory. The stability of the transition state trajectory is found to directly dictate the reaction rate, and it is thus the fundamental and singular object needed to predict barrier-crossing rates in periodically driven chemical reactions. This thesis demonstrates that using minimalist models to examine these complex systems can provide valuable insight into the dynamical mechanisms that drive behavior.