Modern computational methods have become so powerful for predicting the outcome for the H + H2 → H2 + H bimolecular exchange reaction that it might seem further experiments are not needed. Nevertheless, experiments have led the way to cause theorists to look more deeply into this simplest of all chemical reactions. The findings are less simple.

Interferences are genuine quantum phenomena that appear whenever two seemingly distinct classical trajectories lead to the same outcome. They are common in elastic scattering but are seldom observable in chemical reactions. Here we report experimental measurements of the state-to-state angular distribution for the H + D2 reaction using the 'photoloc' technique. For products in low rotational and vibrational states, a characteristic oscillation pattern governs backward scattering. The comparison between the experiments, rigorous quantum calculations and classical trajectories on an accurate potential energy surface allows us to trace the origin of that structure to the quantum interference between different quasiclassical mechanisms, a phenomenon analogous to that observed in the double-slit experiment.

Electronic quenching of OH A (2)Σ(+) by Kr was investigated through experimental studies of the collision cross sections and the OH X (2)Π product state distribution. The quenching cross sections decrease with increasing rotational excitation in the excited OH A (2)Σ(+) electronic state. The OH X (2)Π products of quenching exhibit a significant degree of rotational excitation but minimal vibrational excitation. Complementary theoretical studies of the OH (A (2)Σ(+), X (2)Π) + Kr potential energy surfaces (PESs), nonadiabatic coupling, and quasiclassical trajectory calculations were carried out to elucidate the quenching dynamics. Accurate PESs for the two lowest diabatic states of A' symmetry were computed along with the angularly dependent coupling between them. Coupling in nearly linear HO-Kr configurations provides the mechanism for the observed electronic quenching. A deep attractive well on the OH A (2)Σ(+) + Kr PES facilitates access to this region of strong coupling. Surface-hopping quasiclassical trajectory calculations yielded quenching cross sections and a OH X (2)Π product rotational distribution in good accord with experimental observations.

When a hydrogen (H) atom approaches a deuterium (D(2)) molecule, the minimum-energy path is for the three nuclei to line up. Consequently, nearly collinear collisions cause HD reaction products to be backscattered with low rotational excitation, whereas more glancing collisions yield sideways-scattered HD products with higher rotational excitation. Here we report that measured cross sections for the H + D(2) → HD(v' = 4, j') + D reaction at a collision energy of 1.97 electron volts contradict this behavior. The anomalous angular distributions match closely fully quantum mechanical calculations, and for the most part quasiclassical trajectory calculations. As the energy available in product recoil is reduced, a rotational barrier to reaction cuts off contributions from glancing collisions, causing high-j' HD products to become backward scattered.

We present a new trajectory surface hopping study of the rotational energy transfer and collisional quenching of electronically excited OH(A) radicals by Kr. The trajectory surface hopping calculations include both electronic coupling between the excited 2(2)A' and ground 1(2)A' electronic states, as well as Renner-Teller and Coriolis roto-electronic couplings between the 1(2)A' and 1(2)A″, and the 2(2)A' and 1(2)A″ electronic states, respectively. The new calculations are shown to lead to a noticeable improvement in the agreement between theory and experiment in this system, particularly with respect to the OH(X) rotational and Λ-doublet quantum state populations, compared with a simpler two-state treatment, which only included the electronic coupling between the 2(2)A' and 1(2)A' states. Discrepancies between the predictions of theory and experiment do however remain, and could arise either due to errors in the potential energy surfaces and couplings employed, or due to the limitations in the classical treatment of non-adiabatic effects.