Colloidal Iron Nanoparticles Provide Size-Control in the Catalytic CO Hydrogenation Reaction

Recent advances in synthetic nanotechnology to produce uniform metal nanoparticle catalysts and to use such “monodisperse” particles in catalytic reactions clearly allowed establishing correlations between the catalyst activity/selectivity patterns and the size and shape of metal catalyst particles. The rationale behind these efforts is the quest for high activity and selectivity in multi-path catalytic reactions, with little to no waste or polluting by-products. Among the various methods employed to prepare well-defined metal nanoparticles, those based on colloidal recipes have gained major importance. While the use of stabilizers (“capping agents”, sometimes simply “ligands”) to obtain stable colloidal dispersions with narrow particle size distribution is considered mandatory, there is ample evidence that these stabilizers cause alterations in the catalytic properties of the metal nanoparticles. On the other hand, an intriguing reaction for various green feedstock production on an industrial scale is the hydrogenation of carbon monoxide according to Fischer-Tropsch. Fe-, Co- and Ru-based catalysts have been most commonly investigated in the past for their capability to produce paraffins, olefins, and oxygenates with varying hydrocarbon chain length. Tuning these catalysts for high activity and selectivity to specific product classes is therefore of utmost interest. In this research we demonstrated that monodisperse Fe(0) particles in silica could be successfully prepared and tested in the catalytic CO hydrogenation (so-called Fischer-Tropsch process) without ligand effects to play any visible influence.¹ Different from all of the previous studies, which use iron oxide as precursor to prepare catalysts, we applied metallic (0) iron particles. This Fe(0) metal can be easily converted into iron carbide during Fischer-Tropsch synthesis since it is much more active than iron oxide. Palmitic acid and hexadecylamine used as capping agents in suitable ratios, were quantitatively eliminated through the reaction with hydrogen. We further reveal the occurrence of pronounced particle size effects with respect to catalytic activity and selectivity. In particular, an increase in the particle size causes increasing TOF (activity). TOF values of the same magnitude and particle size dependence have been found when relating the geometry-derived TOF to the active Fe(0) metal surface as determined through hydrogen-deuterium exchange. As to the selectivity signatures of Fe/MCF-17, methane and short-chain (terminal) olefin formation significantly increase with decreasing Fe particle size. Quite differently, both the long-chain C₅₊ olefin and (total) oxygenate slates show the opposite behavior. As to the C₂₊ paraffins, it seems that small particles are less efficient in chain lengthening than larger ones. According to XRD results, bulk Fe(0) of the as-prepared catalyst is chemically reconstructed into iron-carbides, ε- and χ-Fe₅C₂, after reaction. The occurrence of iron-oxide phases, though not clearly detectable here, cannot be ruled out either. Reaction-induced nanoparticle sintering was limited according to TEM and microtome studies.

Summarizing, the methodological approach to study the catalytic CO hydrogenation over monodisperse Fe nanoparticles intercalated in the mesopores of silica provided new insights into the structure-sensitivity of the reaction and is anticipated to help design catalysts with specific selectivity signatures at high reaction rates.

1.       V. Iablokov, Y. Xiang, A. Meffre, P.F. Fazzini, B. Chaudret, N. Kruse ACS Catalysis 2016, 6, 2496-2500