Although there are only a few known examples of supported single-atom catalysts, they are unique because they bridge the gap between homogeneous and heterogeneous catalysis. Here, we report the CO oxidation activity of monodisperse single Pt atoms supported on an inert substrate, θ-alumina (Al2O3), in the presence of stoichiometric oxygen. Since CO oxidation on single Pt atoms cannot occur via a conventional Langmuir-Hinshelwood scheme (L-H scheme) which requires at least one Pt-Pt bond, we carried out a first-principles density functional theoretical study of a proposed pathway which is a variation on the conventional L-H scheme and inspired by the organometallic chemistry of platinum. We find that a single supported Pt atom prefers to bond to O2 over CO. CO then bonds with the oxygenated Pt atom and forms a carbonate which dissociates to liberate CO2, leaving an oxygen atom on Pt. Subsequent reaction with another CO molecule regenerates the single-atom catalyst. The energetics of the proposed mechanism suggests that the single Pt atoms will get covered with CO3 unless the temperature is raised to eliminate CO2. We find evidence for CO3 coverage at room temperature supporting the proposed mechanism in an in situ diffuse reflectance infrared study of CO adsorption on the catalyst's supported single atoms. Thus, our results clearly show that supported Pt single atoms are catalytically active and that this catalytic activity can occur without involving the substrate. Characterization by electron microscopy and X-ray absorption studies of the monodisperse Pt/θ-Al2O3 are also presented.

The mechanism of the catalytic oxidation of water by cis,cis-[(bpy)(2)Ru(OH(2))](2)O(4+) to give molecular dioxygen was investigated using Density Functional Theory (DFT). A series of four oxidation and four deprotonation events generate the catalytically competent species cis,cis-[(bpy)(2)Ru(V)O](2)O(4+), which breaks the H-OH bond homolytically at the rate determining transition state to give a hydroperoxo intermediate. Our calculations predict a rate determining activation barrier of 25.9 kcal/mol in solution phase, which is in reasonable agreement with the previously reported experimental estimate of 18.7-23.3 kcal/mol. A number of plausible coupling schemes of the two metal sites including strong coupling, weak ferromagnetic and weak antiferromagnetic coupling have been considered. In addition, both high-spin and low-spin states at each of the Ru(V)-d(3) centers were explored and we found that the high-spin states play an important mechanistic role. Our calculations suggest that cis,cis-[(bpy)(2)Ru(V)O](2)O(4+) performs formally an intramolecular ligand-to-metal charge transfer when reacting with water to formally give a cis,cis-[(bpy)(2)Ru(IV)O*](2)O(4+) complex. We propose that the key characteristic of the diruthenium catalyst that allows it to accomplish the most difficult first two oxidations of the overall four-electron redox reaction is directly associated with this in situ generation of two radicaloid oxo moieties that promote the water splitting reaction. A proton coupled metal-to-metal charge transfer follows to yield a Ru(V)/Ru(III) peroxo/aqua mixed valence complex, which performs the third redox reaction to give the superoxo/aqua complex. Finally, intersystem crossing to a ferromagnetically coupled Ru(IV)/Ru(III) superoxo/aqua species is predicted, which will then promote the last redox event to release triplet dioxygen as the final product. A number of key features of the computed mechanism are explored in detail to derive a conceptual understanding of the catalytic mechanism.

The resting state of the recently reported water oxidation catalyst [tpyRu(II)-OH(2)](2)L(3+) (tpy = terpyridine; L = bipyridylpyrazolylic anion) ([2,2](3+)) must be activated by a series of proton-coupled oxidations in which four protons and four electrons are removed overall to afford the catalytically competent species [tpyRu(IV)O](2)L(3+) ([4,4](3+)). We have examined all of the plausible redox intermediates utilizing density functional theory coupled to a continuum solvation model. Our calculations reproduce well the first three redox potentials under pH = 1 conditions, and a reasonable correlation between theory and experiment is found for the fourth irreversible redox process that accompanies O(2) generation. The computed oxidation potentials to access [5,4](4+) and [5,5](5+), 1.875 and 2.032 V vs NHE, respectively, exclude the otherwise plausible possibilities of the catalytically active species having a higher oxidation state. [4,4](3+) has an antiferromagnetically coupled ground state in which one ruthenium has two unpaired electrons antiparallel to those of the other ruthenium. As we found in our previous work, two radicaloid terminal oxygen moieties with different spin orientations that are induced by spin polarization from the electron-deficient Ru(IV) centers are found. Two mechanistic scenarios are relevant and interesting for the key O-O bond formation event: intramolecular oxo-oxo coupling and coupling between one terminal oxo and the oxygen atom of the incoming water substrate. The intramolecular oxo-oxo coupling is facile, with a low barrier of 13.9 kcal mol(-1), yielding a peroxo intermediate. The necessary subsequent addition of water in an associative substitution mechanism to cleave one of the Ru-peroxo bonds, however, is found to be impractical at room temperature, with a barrier of DeltaG(double dagger) = 30.9 kcal mol(-1). Thus, while plausible, the intramolecular oxo-oxo coupling is unproductive for generating molecular dioxygen. The intermolecular O-O coupling is associated with a high barrier (DeltaG(double dagger) = 40.2 kcal mol(-1)) and requires the assistance of an additional proton, which lowers the barrier dramatically to 24.5 kcal mol(-1).