The non-statistical dynamics of the ¹8O + ³²O2 isotope exchange reaction at two energies.

The dynamics of the (18)O((3)P) + (32)O2 isotope exchange reaction were studied using crossed atomic and molecular beams at collision energies (E(coll)) of 5.7 and 7.3 kcal/mol, and experimental results were compared with quantum statistical (QS) and quasi-classical trajectory (QCT) calculations on the O3(X(1)A') potential energy surface (PES) of Babikov et al. [D. Babikov, B. K. Kendrick, R. B. Walker, R. T. Pack, P. Fleurat-Lesard, and R. Schinke, J. Chem. Phys. 118, 6298 (2003)]. In both QS and QCT calculations, agreement with experiment was markedly improved by performing calculations with the experimental distribution of collision energies instead of fixed at the average collision energy. At both collision energies, the scattering displayed a forward bias, with a smaller bias at the lower E(coll). Comparisons with the QS calculations suggest that (34)O2 is produced with a non-statistical rovibrational distribution that is hotter than predicted, and the discrepancy is larger at the lower E(coll). If this underprediction of rovibrational excitation by the QS method is not due to PES errors and/or to non-adiabatic effects not included in the calculations, then this collision energy dependence is opposite to what might be expected based on collision complex lifetime arguments and opposite to that measured for the forward bias. While the QCT calculations captured the experimental product vibrational energy distribution better than the QS method, the QCT results underpredicted rotationally excited products, overpredicted forward-bias and predicted a trend in the strength of forward-bias with collision energy opposite to that measured, indicating that it does not completely capture the dynamic behavior measured in the experiment. Thus, these results further underscore the need for improvement in theoretical treatments of dynamics on the O3(X(1)A') PES and perhaps of the PES itself in order to better understand and predict non-statistical effects in this reaction and in the formation of ozone (in which the intermediate O3* complex is collisionally stabilized by a third body). The scattering data presented here at two different collision energies provide important benchmarks to guide these improvements.