Whether a reaction is occurring in the gas phase or condensed phase, electronically nonadiabatic effects can become important if the adiabatic reaction coordinate requires a considerable change in the electronic wavefunction. The experiments and analysis presented here seek to make progress on a difficult and important problem, that of developing a back-of-the-envelope method to predict which energetically allowed products are favoured or disfavoured when significant electronic configuration changes are required to access one or all of the possible product channels. By examining the off-diagonal matrix elements responsible for coupling electronic configurations in the initially excited molecule with those of the products, we begin to formulate a hierarchy of what electronic configurations are strongly vs. weakly coupled. Hence, the paper focusses on understanding how an electronic wavefunction is most likely to change during a chemical reaction when it cannot adjust adiabatically during the nuclear dynamics.

We begin by analyzing the results of two prior series of experiments in order to develop a hierarchy of propensity rules for electronic configuration changes from reactant to products. Analysis of experimental and computational results on the competition between C–Br fission and C–Cl fission in n_Oπ*_C=O excited Br(CH₂)₂COCl and on the ππ* photofragmentation channels of nitric acid suggest the following. If the one-electron configuration interaction matrix elements between the reactant electronic configuration and a product electronic configuration are zero, then the reaction is strongly susceptible to nonadiabatic suppression of the reaction rate and/or appearance of nonadiabatic asymptotic products. One must then analyze the remaining two-electron configuration interaction (Förster- and Dexter-type) matrix elements. If the two-electron change required to couple the reactant and product electronic configurations involves simultaneous configuration changes on two spatially/electronically isolated functional groups, then that product channel is strongly disfavoured. We show why this is the case by examining the two-electron integrals for C–Br fission in Br(CH₂)₂COCl and for the forbidden NO₂(1²B₁)+OH(A″) channel from ππ* excited nitric acid, comparing them to those for the NO₂(1²B₂)+OH(A′) channel where the orbitals involved are localized on the same functional moiety. This hierarchy in electronic coupling motivates the introduction of a ‘restricted adiabatic’ correlation diagram to predict which product channels are electronically accessible.

In the final section of this paper we present new results on the photodissociation of N,N-dimethylformamide following π_nbπ* excitation at 193 nm, where we test the ideas developed from analysis of the previous work. Our measurement of the photofragment velocity and angular distributions of the dissociation products reveals that dissociation pathway to form HCO +N(CH₃)₂ results in formation of HCO()+N(CH₃)₂(Ã) but not HCO(Ã) +N(CH₃)₂(). As both are energetically allowed product channels in the singlet A′ manifold, the selectivity may be analyzed with respect to the required change in electronic configuration to access each asymptotic product channel. To understand the experimental results in the context of the model developed from the prior work, we consider both one-electron and two-electron contributions to the configuration interaction matrix elements between the reactant and product electronic configurations to determine which product channels are most likely to be accessed.