Prevalence of Bimolecular Routes in the Activation of Diatomic Molecules with Strong Chemical Bonds (O2, NO, CO, N2) on Catalytic Surfaces.

Dissociation of the strong bonds in O2, NO, CO, and N2 often involves large activation barriers on low-index planes of metal particles used as catalysts. These kinetic hurdles reflect the noble nature of some metals (O2 activation on Au), the high coverages of co-reactants (O2 activation during CO oxidation on Pt), or the strength of the chemical bonds (NO on Pt, CO and N2 on Ru). High barriers for direct dissociations from density functional theory (DFT) have led to a consensus that "defects", consisting of low-coordination exposed atoms, are required to cleave such bonds, as calculated by theory and experiments for model surfaces at low coverages. Such sites, however, bind intermediates strongly, rendering them unreactive at the high coverages prevalent during catalysis. Such site requirements are also at odds with turnover rates that often depend weakly on cluster size or are actually higher on larger clusters, even though defects, such as corners and edges, are most abundant on small clusters. This Account illustrates how these apparent inconsistencies are resolved through activations of strong bonds assisted by co-adsorbates on crowded low-index surfaces. Catalytic oxidations occur on Au clusters at low temperatures in spite of large activation barriers for O2 dissociation on Au(111) surfaces, leading to proposals that O2 activation requires low-coordination Au atoms or Au-support interfaces. When H2O is present, however, O2 dissociation proceeds with low barriers on Au(111) because chemisorbed peroxides (*OOH* and *HOOH*) form and weaken O-O bonds before cleavage, thus allowing activation on low-index planes. DFT-derived O2 dissociation barriers are much lower on bare Pt surfaces, but such surfaces are nearly saturated with CO* during CO oxidation. A dearth of vacant sites causes O2* to react with CO* to form *OOCO* intermediates that undergo O-O cleavage. NO-H2 reactions occur on Pt clusters saturated with NO* and H*; direct NO* dissociation requires vacant sites that are scarce on such surfaces. N-O bonds cleave instead via H*-assistance to form *HNOH* intermediates, with barriers much lower than for direct NO* dissociation. CO hydrogenation on Co and Ru occurs on crowded surfaces saturated with CO*; rates increase with increasing Co and Ru cluster size, indicating that low-index surfaces on large clusters can activate CO*. Direct CO*dissociation, however, occurs with high activation barriers on low-index Co and Ru surfaces, and even on defect sites (step-edge, corner sites) at high CO* coverages. CO* dissociation proceeds instead with H*-assistance to form *HCOH* species that cleave C-O bonds with lower barriers than direct CO* dissociation, irrespective of surface coordination. H2O increases CO activation rates by assisting H-additions to form *HCOH*, as in the case of peroxide formation in Au-catalyzed oxidations. N2 dissociation steps in NH3 synthesis on Ru and Fe are thought to also require defect sites; yet, barriers on Ru(0001) indicate that H*-assisted N2 activation - unlike O2, CO, and NO - is not significantly more facile than direct N2 dissociation, suggesting that defects and low-index planes may both contribute to NH3 synthesis rates. The activation of strong chemical bonds often occurs via bimolecular reactions. These steps weaken such bonds before cleavage on crowded low-index surfaces, thus avoiding the ubiquitous kinetic hurdles of direct dissociations without requiring defect sites.