The oxidation of n-octane by iridium and cobalt PNP complexes.

Abstract

Paraffin activation has practical implications in the replacement of current petrochemical feedstocks (olefins), by utilizing economical and easily accessible alkanes, which may result in more efficient strategies for fine chemical synthesis and the proficient use of energy. However, the chemical inertness of paraffins limits their conversion to more valuable products. Several pincer chelate complexes are utilized in stoichiometric and catalytic C–H activation. These pincer ligands have attained much interest in that they are part of a system, which displays high stability, activity and variability. In this study four aminodiphosphine (PNP) pincer ligands were successfully synthesized and characterized by NMR, IR and HRMS. To investigate the steric effects on the metal center, four different functional groups on the nitrogen atom were used, a cyclic ring (cyclohexyl (3.1)) branched chain (iso-propyl (3.2)); straight chain (pentyl (3.3)); and aromatic ring (benzyl (3.4)). The ligands were successfully complexed to the transition metals iridium and cobalt and characterized by elemental analyses, IR, HRMS and thermogravimetric measurements. The thermal behaviour of the ligands showed that ligands 3.1-3.3 displayed similar decomposition patterns. Similar fragmentation patterns were observed for the iridium and cobalt complexes containing ligands 3.1 and 3.3. The complexes were tested in the oxidation of n-octane in two solvent systems, DCM and MeCN with H2O2 and t-BuOOH as the oxidants. The optimum substrate to oxidant ratio was found to be 1:5. No conversion was observed with H2O2. The conversion in DCM for the iridium catalysts was much higher than that of the cobalt catalyst. However, higher conversion was obtained in MeCN for the cobalt catalysts. No conversion was observed for the iridium catalyst in MeCN. The selectivity to ketones was much higher than to the alcohols, with only the C(1) position being most selective to the alcohols. The in situ, single pot testing of n-octane using a ruthenium precursor and ligand 3.1- 3.4 undertaken in DCM showed no conversion, whilst in MeCN a conversion of 17% was observed. The selectivity was similar to that obtained by the cobalt catalysts in MeCN. All testing showed that the catalyst containing ligand 3.1 was the most active giving the highest conversions in different solvent systems, which is attributed to the bite angle effect.