Computationally designed tandem direct selective oxidation using molecular oxygen as oxidant without coreductant†
Abstract
Developing greener technologies to produce chemicals has attracted much recent attention. For the selective oxidation of organic compounds, direct selective oxidation in which molecular O2 is utilized as oxidant without using a co-reductant or co-oxidant is desirable to avoid forming waste co-products; however, existing catalysts are limited in the types of substrates that can be used to achieve this. To address this challenge, we introduce a tandem direct selective oxidation process that separates the molecular oxygen activation step from the substrate oxidation step. Specifically, two reactions occur in separate reactors over two different catalysts: (1) molecular oxygen activation via reaction with an oxygen acceptor molecule to produce an oxygen transfer intermediate, and (2) substrate oxidation via reaction with the oxygen transfer intermediate to produce substrate oxide and regenerate the oxygen acceptor molecule. The oxygen acceptor molecule should be recycled back to the first reactor to achieve a net reaction of 2 substrate + O2 → 2 substrate oxide. This separation of molecular oxygen activation and substrate oxidation steps reduces by-product formation by avoiding some side reactions. We use density functional theory (DFT) to study propene epoxidation as an important example. This reaction is of interest because: (a) propylene oxide (PO) is one of the leading commodity chemicals worldwide, (b) all current commercial PO production processes produce co-products, and (c) propene epoxidation illustrates the class of difficult terminal alkene epoxidations for substrates containing allylic hydrogen atoms. Using DFT calculations, we identify plausible candidates for the oxygen transfer intermediate and catalysts for the molecular oxygen activation and substrate oxidation reactions. The Zr(C6H4-1,2-(N(C6H3-2′,6′-(CH3)2)O)2)2 [DMZB] and the Ru(meso-tetrakis(2,6-dichlorophenyl)porphyrin) [RuTDCPP] catalysts were chosen for reactions (1) and (2), respectively. Several pyridine based N-oxides were tested as oxygen transfer intermediates. Our DFT computations indicate 2,6-dimethylpyridine N-oxide should perform well. The RuTDCPP catalyst was prior experimentally demonstrated to oxidize organic substrates (e.g., 1-octene) using aromatic N-oxides as oxidants with above 90% selectivity towards the desired product under mild conditions. For molecular O2 activation (reaction (1)) and propene epoxidation (reaction (2)), our computed enthalpic energetic spans are 33.3 and 31.6 kcal mol−1, respectively, predicting decent activities for both catalysts.