Electronic structure calculations at the density functional theory/ B3LYP level (selectively benchmarked by CCSD(T)) were performed on neutral and protonated monomer and dimer clusters of vanadium oxide (V x O y ) on a cluster model of a TiO 2 support to predict the first steps in the mechanism of the selective catalytic reduction (SCR) of nitric oxide by ammonia. The vanadium cluster structures are based on experimental NMR measurements. The first step is Lewis acid−base addition of NH 3 to a vanadium site for the neutral and formation of an "NH 4+ " site on the protonated surface. Different proton transfer pathways, which depend on the initial neutral or protonated sites coupled with addition of NO lead to the formation of NH 2 NO surface species. The mechanisms can be complicated involving many different pathways for the proton transfers, especially for the initial protonated surface. The addition of the doublet NO leading to the formation of NH 2 NO leads to a reduction of a vanadium and transfer of the spin to this site. NH 2 NO desorbs and then undergoes gas-phase rearrangements to form the final products N 2 + H 2 O. The barrier heights for the gas-phase rearrangement process leading to final product formation are comparable in a number of cases to the barrier heights on the catalyst and may represent the rate-determining step. The reaction on the cluster model with a reduced vanadium(+4) proceeds by different paths with addition of NO leading to a second reduced vanadium site. The predicted pathways are consistent with the available experimental data and show that the complete SCR mechanism is very complicated as additional H 2 O molecules will be removed from the surface by the addition of O 2 to fully regenerate the catalyst and oxidize V back to the formal +5 state.

The chemiluminescent reactions of the group 3 metals Sc and Y with F 2 , Cl 2 , Br 2 , ClF, ICl (Sc), IBr (Y), and SF 6 and La with F 2 , SF 6 , Cl 2 , and ClF have been studied at low pressures (6 × 10 −6 to 4 × 10 −4 Torr) using a beam-gas arrangement and extended to the 10 −3 Torr multiple collision pressure range. Contrary to previous reports, the observed chemiluminescent spectra are primarily attributed to emission from the metal monohalides. Extensive pressure and temperature dependence studies and high-level correlated molecular orbital theory calculations of the bond dissociation energies support this conclusion and the attribution of the chemiluminescence. Evidence for the "selective" production of a monohalide excited electronic state is obtained for several of the Sc and Y reactions. All reactions producing the metal monofluorides are first order with respect to the oxidant, while reactions producing the monochlorides and monobromides are found to be "faster than first order" with respect to the oxidant. This difference is associated with the metal halide bond dissociation energies and the metal halide product internal density of states. Analysis of the temperature dependence for six representative reactions indicates that the "selective" excited-state formation of the metal monohalides proceeds via a direct mechanism with negligible activation energy. We compare and contrast the present results with previous experiments and interpretations which have assigned the selective emission from these systems to the group 3 dihalides produced in a two-step reaction sequence analogous to an electron jump process. The current results suggest a distinctly different interpretation of the observed processes in these systems. The observed selectivity observed in these studies is remarkable given the significant number of known and potential excited states in the scandium and yttrium halides as well as their different electronic configurations.

The effect of frustrated Lewis donors on metal selectivity between actinides and lanthanides was studied using a series of novel organic ligands. Structures and thermodynamic energies were predicted in the gas phase, in water, and in butanol using 9-coordinate, explicitly solvated (H2O) Eu, Gd, Am, and Cm in the +III oxidation state as reactants in the formation of complexes with 2-(6-[1,2,4]-triazin-3-yl-pyridin-2-yl)-1H-indole (Core 1), 3-[6-(2H-pyrazol-3-yl)­pyridin-2-yl]-1,2,4-triazine (Core 2), and several derivatives. These complexations were studied using density functional theory (DFT) incorporating scalar relativistic effects on the actinides and lanthanides using a small core pseudopotential and corresponding basis set. A self-consistent reaction field approach was used to model the effect of water and butanol as solvents. Coordination preferences and metal selectivity are predicted for each ligand. Several ligands are predicted to have a high degree of selectivity, particularly when a low ionization potential in the ligand permits charge transfer to Eu­(III), reducing it to Eu­(II) and creating a half-filled f7 shell. Reasonable separation is predicted between Cm­(III) and Gd­(III) with Core 1 ligands, possibly due to ligand donor frustration. This separation is largely absent from Core 2 ligands, which are predicted to lose their frustration due to proton transfer from the 2N to the 3N position of the pyrazole component of the ligands via tautomerization.

No abstract

NO 2 and NO, which are generated in combustion processes, binding to vanadium oxide clusters including TiO 2 -supported catalyst models in the selective catalytic reduction (SCR) of NO has been studied by density functional theory and coupled cluster methods. NO x binding on vanadium oxides is predicted to depend on several factors, including the excitation energy of the oxide, ionization energies of both the unbound oxide and the deoxygenated reduced oxide, and the strength of the molecular V−O bonds. NO 2 chemisorption occurs either through covalent bond formation in a HONO-like pattern or through abstraction of a metal oxide oxygen leading to the formation of NO 3 − . Nitrate formation is more favorable than what was predicted for group IVB or group VIB oxides [except (CrO 3 ) n ] and is either the lowest energy binding mode or within a few kcal/ mol of the lowest mode in all clusters, likely due to the stability of V in the +4 oxidation state. Physisorption on V oxides is very weak. V with 2 oxo groups have a lower excitation energy and a more sterically open geometry which results in strong chemisorption as predicted for group IVB oxides. Tetrahedrally coordinated vanadia with a single oxo group and 3 V−O single bonds are predicted to have significantly higher excitation energies and behave like group VIB oxides such that chemisorption is unlikely and weak physisorption dominates the interaction. In larger clusters, including SCR catalyst models, only tetrahedrally coordinated vanadia are present and NO 2 binding is not expected to occur. NO adsorption is weaker overall than NO 2 binding and occurs either as physisorption or as chemisorption through the formation of NO 2 − analogous to nitrate formation in NO 2 binding. The ability of NO to bind reflects the patterns predicted for NO 2 , such that NO is strongly bound vanadia with two V�O groups and only weakly physisorbed when there is a single V�O or none at all.