Broad terms in the literature, such as nonstatistical reactivity or nontraditional luminescence, emerge when standard theories fail to explain experimental results. In the case of nonstatistical and dynamic effects, reaction rates and product ratios may vary wildly from transition state theory (TST) predictions and are commonly accompanied by a lack of temperature dependence. In this Tutorial, we explain how to use modern and freely available computational chemistry tools to model apparently nonstatistical reactions in relatively large organic molecules, using a concrete example of the thermal Garratt–Braverman/1,5-H shift of an ene-diallene, such that a non-expert can easily replicate our approach by following our steps. As a team of synthetic organic chemists and computational chemists, we aim to promote the use of preparatory computational work that can aid in reaction design during the experimental process rather than merely serving as complementary data to finished experimental studies. For this reason, we also offer an overview of the literature leading into more advanced computational techniques. Working through the example of the thermal Garratt–Braverman/1,5-H shift, we illustrate how theory and experiment can be used together to investigate the possibility of parallel light-enabled reactions and the role of tunneling effects and to perform careful variable temperature analysis when it becomes necessary. As the motifs of reactive π–π* absorptions, hydrogen transfers, and diradical intermediates are quite common, the points presented in this Tutorial are broadly applicable and indicative of the underlying complexity behind many chemical reactions that exhibit unexpected rates and ratios. The failure of TST also underscores a critical limitation of massive reaction network discovery schemes that heavily rely on calculated ground-state activation energies as well as the potential pitfalls of the conventional free energy diagram mindset.
Rožić et al. (Sun,) studied this question.