Active Sites of the Selective Catalytic Reduction of NO by NH 3 over Fe-ZSM-5: Combining Reaction Kinetics with Postcatalytic Mössbauer Spectroscopy at Cryogenic Temperatures

57 Fe-ZSM-5 catalysts were prepared by solid-state ion exchange of 57 FeCl 3 into H-ZSM-5, the exchange capacity of which had been blocked by Na + or Ca 2+ to ≈50, ≈25, and 0%. Their Fe site structure was investigated by UV–vis, electron paramagnetic resonance (EPR), and Mössbauer spectroscopy at different temperatures including cryogenic conditions, and their selective catalytic reduction (SCR) activity was studied between 400 and 870 K. Mössbauer analysis was performed by establishing consistent models that well represent cryogenic measurements at low and high magnetic fields (10 mT and 5 T, respectively) but also low-field measurements at higher temperatures, which were performed in order to evaluate their analytical potential. It was found that a reliable analysis in particular of species in small structures (binary, oligomeric sites) requires modeling of cryogenic Mössbauer data with spin-Hamiltonian formalism, which needs to be combined with input from EPR spectroscopy for specifying/confining the type of paramagnetic Fe III site to be included in the model. Spin-Hamiltonian formalism (combined with input from EPR) can be applied also to measurements at higher temperatures where reliable results for paramagnetic Fe III and Fe II sites and for the sum of superparamagnetic Fe III species and particles can be obtained from spectra measured at 77 K. While qualitative observations in EPR spectra of initial samples were in good agreement with quantitative site abundances obtained from the Mössbauer spectra, there were significant deviations between the latter and results from UV–vis measurements. Mössbauer-derived site distributions in postcatalytic samples (after reaction at 670 K) showed that the standard SCR rate correlates well with the abundance of oligomeric Fe III oxo sites, in particular with binary Fe III –O–Fe III structures, although an intercept may indicate a minor relevance of an additional species. The clear correlation between rates and abundance of Fe III sites rejects reoxidation as a rate-limiting step for the given catalysts and reaction conditions.