Single-atom catalysts supported on metal−organic frameworks have attracted growing attention over the past few decades, as have theoretical studies on these systems. While accurate predictions are highly desirable for improved mechanistic understanding and catalyst design, benchmarks on methodologies employed for such systems remain limited. In this work, we present a comprehensive evaluation of computational models and various density functionals for catalytic ethylene dimerization on a Ni single-atom catalyst supported on NU-1000. We elucidate the nature of the mononuclear Ni sites and their catalytic transformations along the reaction pathway via the Cossee-Arlman mechanism. We assess the adequacy of truncated cluster models and the accuracy of different density functionals by comparing them with the highly accurate yet computationally demanding periodic model and the DLPNO−CCSD(T) method, respectively.
Huynh et al. (Wed,) studied this question.