Mechanism and Models for Collisional Energy Transfer in Highly Excited Large Polyatomic Molecules
Australian Journal of Chemistry
48(11) 1787 - 1817
Collisional energy transfer in highly excited molecules (say, 200-500 kJ mol-1 above the zero-point energy of reactant, or of product, for a recombination reaction) is reviewed. An understanding of this energy transfer is important in predicting and interpreting the pressure dependence of gas-phase rate coefficients for unimolecular and recombination reactions. For many years it was thought that this pressure dependence could be calculated from a single energy-transfer quantity, such as the average energy transferred per collision. However, the discovery of 'supercollisions' (a small but significant fraction of collisions which transfer abnormally large amounts of energy) means that this simplistic approach needs some revision. The 'ordinary' (non-super) component of the distribution function for collisional energy transfer can be quantified either by empirical models (e.g., an exponential-down functional form) or by models with a physical basis, such as biased random walk (applicable to monatomic or diatomic collision partners) or ergodic (for polyatomic collision partners) treatments. The latter two models enable approximate expressions for the average energy transfer to be estimated from readily available molecular parameters. Rotational energy transfer, important for finding the pressure dependence for recombination reactions, can for these purposes usually be taken as transferring sufficient energy so that the explicit functional form is not required to predict the pressure dependence. The mechanism of 'ordinary' energy transfer seems to be dominated by low-frequency modes of the substrate, whereby there is sufficient time during a vibrational period for significant energy flow between the collision partners. Supercollisions may involve sudden energy flow as an outer atom of the substrate is squashed between the substrate and the bath gas, and then is moved away from the interaction by large-amplitude motion such as a ring vibration or a rotation; improved experimental and theoretical understanding of this phenomenon is seen as an important area for future development.
© CSIRO 1995