A wave-packet study on nonadiabatic transition dynamics in photodissociation: The importance of parent bending motion

Abstract

The influence of parent molecule bending motion on nonadiabatic transitions during photodissociation is investigated using a model involving two linear excited-state surfaces. The two dissociative states are strongly coupled near the so-called conical intersection. Due to symmetry constraints, the two adiabatic surfaces are not allowed to interact in linear configurations and the coupling is only switched on when the molecule is bent. Therefore, electronic transitions from one adiabatic surface to another are only possible when the dissociation process involves bending motion. A quantum-mechanical model including both the dissociation and bending degrees of freedom is established to assess the importance of parent bending motion on nonadiabatic transition dynamics. The coupling between the two electronic states is explicitly taken into account and the dissociation dynamics is described by a time-dependent wave packet. Although the approach is applicable to any triatomic molecular systems, we choose to model a pseudotriatomic system, i.e., methyl iodide CX3I (X=H, D, etc.). In our calculations, the three X atoms are treated as a single pseudoatom and the C–X3 umbrella bending coordinate is frozen at its equilibrium geometry. The two dynamically active coordinates are the I–CX3 stretch and the I–C–X3 bend. Both the ground and two excited states are represented by linear potential-energy surfaces and the coupling of the two dissociative surfaces is a conical intersection in nature. Several dissociation processes are modeled with different initial bending wave functions and different isotopic substitutions. It is found that parent bending motion has a significant effect on the final electronic branching of dissociation fragments. The calculation generates a larger I* yield from the CD3I dissociation than that from CH3I, in agreement with experimental observations which could not be reproduced by previous theoretical calculations. Our model also predicts that the dissociation of the first bending overtone of methyl iodide gives a smaller I* yield than that of its ground-state counterpart, which is consistent with a recent experiment on the CF3I photodissociation.