| Experiments and computational chemistry have been used to explore the identity of the active Fe/ZSM-5 surface oxygen which participates in redox reactions and the possible kinetics pathways involving it. The experiments entail monitoring the mass of the catalyst during reactions using a Tapered Element Oscillating Microbalance (TEOM), measuring kinetics in a tubular reactor as well as in the TEOM, and characterizing the catalysts using Fourier Transform Infrared (FTIR) spectroscopy, X-ray Diffraction (XRD) and Temperature Programmed Reduction (TPR). Computer studies involve kinetics modeling using microkinetic modeling theory, and probing the structures and thermochemistry of possible catalytic intermediate species using Density Functional Theory (DFT).; In-situ gravimetry results show that the redox capacity of Fe/ZSM-5 using H2/O2 probe is high (0.6 ~1.25), additionally two peaks are observed in H2-TPR. This indicates that following reduction, some of the iron may exist in an oxidation state lower than +2. However, for the redox reactions studied herein, the working state of the catalyst is fully oxidized and corresponds to ferric cations.; The catalyst mass was monitored while making step changes in the gas phase composition at N2O decomposition conditions. The results suggest that NOx surface species do not form in N2O/He, and if they form in N2O/NO/He, their surface concentration is exceedingly small.; A single mechanistic framework was proposed and used to model a variety of redox reactions; the microkinetic model is consistent with experimental observations of conversion and catalyst mass change as well as DFT calculations. Water Gas Shift (WGS) can not proceed over Fe/ZSM-5 up to 500°C. Unlike NO, trace CO and H2 don't have a promotional effect beyond stoichiometric reaction when introduced at trace levels into N2O decomposition. Similarly, trace NO does not promote CO/H2 oxidation. It appears that NO produces nitrite/nitrate intermediates and creates a fast pathway for O2 desorption. The latter species' coverage is under 1% both in modeling as well as experimental results. However, the modeling was successful only if the surface bond energy of oxygen on the active site is different, when generated from N2O decomposition than when generated from oxygen.; Density functional calculations were performed on the surface species involved. The computational chemistry shows that when this site is reduced, one of the terminal hydroxyl groups moves into a bridging position. In this way there exists a pool of 1.5 oxygen atoms per iron cation which participate in the N2O decomposition reaction whereas the redox capacity of the catalyst is only 0.5 oxygen atoms per iron cation. |
