Toward Predictive Theory in Single-Atom Catalysis.

Single-atom catalysis has become a central framework for experiment-theory integration, as catalytic performance is highly sensitive to the environment of individual metal atoms, a feature that electronic structure calculations are well-suited to analyze. Yet much of current theoretical practice relies on simplified single-site models and narrow reactivity windows, overlooking the intrinsic site diversity and evolution of single-atom catalysts (SAC). This Perspective discusses how SAC modeling can be reframed through a lifecycle-oriented view that integrates synthesis, activity, stability, and safety. By adopting ensemble-based descriptions and modular thermodynamic descriptors, we show how theory can be used systematically in line with the level of structural definition accessible experimentally. Using acetylene hydrochlorination as a prominent SAC application with exceptional data coherence for examining the theory-experiment interplay, wedemonstrate that site formation and evolution under synthesis and reaction conditions, as well as ensemble-driven activity trends consistent with experimental yields, can be treated quantitatively. In contrast, stability and safety are more effectively addressed through comparative, pathway-resolved analyses. More broadly, this perspective points toward a shift in how SAC modeling is framed across reactions, enabling theory to move beyond post-rationalization toward disciplined prediction.