NH
3 has attracted significant attention as a potential hydrogen carrier because of its high hydrogen density, ease of storage, and zero carbon emission. Developing efficient catalysts for NH
3 decomposition to N
2 and H
2 is crucial to achieve the NH
3-mediated hydrogen economy. Ru-exchanged 13X zeolites exhibit extraordinarily high specific activity (⥠3,000 h
-1) for NH
3 decomposition at a low Ru-loading (⤠1.4 wt.%), at which Ru is atomically dispersed
1. In this work, we study the nature of the Ru-13X catalysts using computational chemistry calculations. The catalytic systems are constructed by adding one or two Ru atom(s) to the 13X structure at the experimentally identified sites. First, the N-H bond dissociation step is investigated on a single Ru center varying formal oxidation states of Ru (+1-+3). The Ru oxidation state was found to play a key role; the higher the oxidation state, the stronger the NH
3 binding and the lower the N-H bond activation energy. Interestingly, the Ru oxidation state dictates the N-H activation mode, which occurs either homolytically (Ru(+1)) or heterolytically (Ru(+2 or +3)), where the latter is more facile. The heterolytic dissociation resembles the functionality of Frustrated Lewis pairs (FLP) as Ru accepts electrons from NH
2- (Lewis acid) and framework oxygen donates an electron pair to H
+ (Lewis base). The Ru oxidation state also influences the thermodynamic stability of Ru-13X, with Ru(+2) being the most stable. These suggest that Ru(+3) is active but unstable, whereas Ru(+2) is less active but stable. This is further confirmed by energetics of the full NH
3 decomposition reaction pathway (2NH
3âN
2+3H
2) on the catalysts with two Ru centers. Our computational results are in excellent agreement with experimental observations demonstrating how experiments and theory can synergistically work towards elucidating complex ammonia decomposition chemistry. The identified unique FLP functionality can provide further guidance for designing zeolite-based single-atom catalysts.
