Klaus‐Peter Dinse

Doping of Li/MgO with additional metal ions is suggested leading to an improved system with respect to the catalytic performance and stability when used for oxidative coupling of methane. We used Gd and Fe as dopants and characterized the resulting materials, showing that Fe seems to be completely and Gd partly incorporated into the MgO lattice. The catalytic performance is improved in most cases, but all materials still suffer from severe deactivation. A loss of Li is observed when being used under reaction conditions, but this Li loss is retarded for Fe‐Li/MgO as compared to undoped Li/MgO.

The role of Li in Li/MgO as a catalyst for oxidative coupling of methane (OCM) is to promote MgO surface morphology change rather than serve as a constituent of catalytically active sites. While Li/MgO is unstable at realistic conditions with respect to loss of Li, the resulting samples show enhanced selectivity towards C₂ hydrocarbons versus CO₂, although activity is low and close to pristine MgO. The way (co‐)doping with alio‐valent metal ions affects the catalytic performance of Li/MgO has now been explored. To analyze the structure and the stability of the samples, catalysts with well‐defined stoichiometry were prepared using a co‐precipitation method with freeze‐drying and subsequent annealing. Gd and Fe were used as dopants. Apart from their potential direct role in catalysis, these dopants are anticipated to stabilize Li in the catalyst under the reaction conditions, allowing further clarification of the role of Li. In the case of Gd/Li co‐doping, changes observed in EPR and ⁷Li‐NMR spectra indicate the formation of correlated, next‐neighbor Li⁻_Mg⋅⋅⋅Gd⁺_Mg pairs co‐existing with “isolated” Gd³⁺ ions at octahedral Mg lattice sites. For Li/Fe co‐doping, no significant change in the EPR pattern is observed in the presence of Li⁺ ions, indicating a larger distance between the Li⁺ and Fe³⁺ cations in the MgO lattice. Hybrid DFT calculations explain the difference between Fe and Gd co‐doping by a less efficient screening of the Coulomb repulsion between Gd³⁺ and neighboring cations in Gd doped samples, leading to the stabilization of Li_Mg near Gd_Mg.

VO_x (1.4–1.7 V nm⁻²) supported on SBA‐15, Al₂O₃, or TiO₂ was studied before and after exposure to oxidative dehydrogenation of propane (ODP), and pure hydrogen or propane. After treatment, samples were quenched and frozen in quartz vials and characterized by using high‐frequency electron paramagnetic resonance (HF‐EPR). For SBA‐15‐ and Al₂O₃‐supported vanadia, V⁴⁺ sites were the most abundant paramagnetic species, whereas Ti³⁺ was dominant in TiO₂‐supported V₂O₅. For the quantification of paramagnetic reduced sites, Mn²⁺ was used as reference. The maximum relative numbers of reduced V⁴⁺ or Ti³⁺ sites were found to increase in the sequence SBA‐15 (11 % V⁴⁺/V)2O₃ (16 % V⁴⁺/V)2 (21 % Ti³⁺/V) if hydrogen was used as a reducing agent. Using propane as a reductant, the percentage ratios were found to be SBA‐15 (8 % V⁴⁺/V), Al₂O₃ (18 % V⁴⁺/V), and TiO₂ (27 % Ti³⁺/V). The corresponding V⁴⁺/V or Ti³⁺/V percentage ratios for samples exposed to ODP were lower, at 0.08, 9, and 17 %, respectively. The number of paramagnetic non‐V⁴⁺ or non‐Ti³⁺ sites was negligible for Al₂O₃ and SBA‐15 and in the range of 12–37 % for TiO₂‐supported vanadia. The number of reduced sites after oxidation in air and storage in sealed vials is negligible, which indicates that all reduced sites can generally be reoxidized. After reaction in an oxygen‐lean ODP atmosphere, a structureless HF‐EPR pattern was observed for SBA‐15‐supported vanadia, which is tentatively attributed to the formation of two‐dimensional polymeric structures of vanadia. For Al₂O₃‐supported vanadia, this loss of spectral resolution was less pronounced, which indicates a more defined environment of V⁴⁺ sites. The results are discussed in terms of a two‐parameter model, rationalizing the dual influences of reducibility and local structure of redox‐active sites on the ODP reaction.