Insertion vs. abstraction in the atomic hydrogen + molecular hydrogen .fwdarw. molecular hydrogen + atomic hydrogen exchange reaction

Abstract

At collision energies above 1 eV an insertion mechanism is shown to dominate in the hydrogen exchange reaction. The cone of acceptance for reaction is found to be made up of an inner cone (Le., for more nearly collinear collisions) where exchange proceeds by abstraction and an outer spherical sector where the mechanism is by insertion. The cross section for reaction, computed by classical trajectories, declines at energies above ca. 1 eV due to a recrossing of the transition state after a collision with an inner hard core. Thus, while the barrier to insertion is higher, this mechanism dominates for such hot H atoms as are currently available from photodissociation. For the H + HD reaction with rotationally cold HD, the cone of acceptance about the D atom is significantly wider. The hydrogen exchange reaction] is usually assumed to proceed via preferentially nearly collinear collisions. Recently, there has been considerable progress in the study of the dynamics of this reaction using translationally “hot” H atoms produced by photodi~sociation.~ -~ Examination of the potential energy surface for the H3. system’ suggested to us that, for hot H atoms, the reaction will also proceed by insertion. By this we mean that the attacking H atom inserts between the two initially bound atoms while these two move apart to accommodate the incident atom. The transition state is then an equilateral triangle with the inserting atom at its apex. The purpose of this letter is to present the argument using the ab-initio potential energy surface’ and then to demonstrate that dynamical (classical trajectory) computations lead to the same conclusion. The most direct experimental test that we could find is as follows. For exchange reactions with HD which are not collinearly dominated, one expects that ejection of the H atom will be preferred.6,7 That is, in an X + HD reaction, HD will be the preferential product. Trajectory computations for H + HD do show an HD/H2 branching ratio above unity (cf. Figure 4) for hot H atoms. To consider the static steric requirements of simple exchange reactions it proves convenient to examine the recently introduced* “reaction surface”. This representation of the potential energy hypersurface is similar to the familiar “polar” representation’ with one key difference. In the usual polar plot the (old) bond distance is held constant and one plots the potential energy as a function of the distance and angle of orientation of the incident atom. The problem is that upon changing the (old) bond distance it is necessary to make a new plot and there can be significant and even qualitative changes in the shape of the resulting surface.]’ Hence, we proceed as follows. The old bond is placed (as in the ordinary polar plot) along the x axis. Then the incident atom is placed at a given distance (from the diatomic center of mass) and given