Selective oxidation with molecular oxygen is one of the most appealing approaches to functionalization of organic molecules and, yet at the same time, one of the most challenging reactions facing organic synthesis due to poor selectivity control. Molecular oxygen is a green and inexpensive oxidant, producing water as the only byproduct in oxidation. Not surprisingly, it has been used in the manufacturing of many commodity chemicals in the industry. It is also nature’s choice of oxidant and drives a variety of oxidation reactions critical to life and various other biologic processes. While the past decades have witnessed great progress in understanding, both structurally and mechanistically, how nature exploits metalloenzymes, i.e., monooxygenases and dioxygenases, to tackle some of the most challenging oxidation reactions, e.g., methane oxidation to methanol, there are only a small number of well-defined, man-made metal complexes that have been reported to enable selective oxidation with molecular oxygen of compounds more relevant to fine chemical and pharmaceutical synthesis.
In the past 10 years or so, our laboratories have developed several transition metal complexes and shown that they are capable of catalyzing selective oxidation under 1 atm of O2. Thus, we have shown that an Fe(II)-bisimidazolidinyl-pyridine complex catalyzes selective oxygenation of C–H bonds in ethers with concomitant release of hydrogen gas instead of water, and when the iron center is replaced with Fe(III), selective oxidative cleavage of C═C bonds of olefins becomes feasible. To address the low activity of the iron complex in oxidizing less active olefins, we have developed a Mn(II)-bipyridine complex, which catalyzes oxidative cleavage of C═C bonds in aliphatic olefins, C–C bonds in diols, and carboxyl units in carboxylic acids under visible light irradiation. Light is necessary in the oxidation to cleave an off-cycle, inactive manganese dimer into a catalytically active Mn═O oxo species. Furthermore, we have found that a binuclear salicylate-bridged Cu(II) complex enables the C–H oxidation of tetrahydroisoquinolines as well as C═C bond cleavage, and when a catalytic vitamin B1 analogue is brought in, oxygenation of tetrahydroisoquinolines to lactams takes place via carbene catalysis. Still further, we have found that a readily accessible binuclear Rh(II)-terpyridine complex catalyzes the oxidation of alcohols, and being water-soluble, the catalyst can be easily separated and reused multiple times. In addition, we recently unearthed a simple protocol that allows waste polystyrene to be depolymerized to isolable, valuable chemicals. A cheap Brønsted acid acts as the catalyst, activating molecular oxygen to a singlet state through complexation with the polymer under light irradiation, thereby depolymerizing the polymer.
Millions of chiral compounds contain a stereogenic sp3-hybridized carbon center with a hydrogen atom as one of the four different substituents. The stereogenic center can be edited in an increasing number of cases by selective hydrogen atom transfer (HAT) to and from a photocatalyst. This Account describes the development of photochemical deracemization reactions using chiral oxazole-annulated benzophenones with a bonding motif that allows them to recognize chiral lactam substrates by two-point hydrogen bonding. The backbone of the catalysts consists of a chiral azabicyclo[3.3.1]nonan-2-one with a U-shaped geometry, which enables substrate recognition to occur parallel to the benzoxazole part of the aromatic ketones. The photocatalysts facilitate a catalytic photochemical deracemization of several compound classes including hydantoins, N-carboxyanhydrides, oxindoles, 2,5-diketopiperazines, and 4,7-diaza-1-isoindolinones. In addition, if more than one stereogenic center is present, the editing delivers a distinct diastereoisomer upon the appropriate selection of the respective photocatalyst enantiomer. The chiral photocatalysts operate via the benzophenone triplet that selectively abstracts a properly positioned hydrogen atom in exclusively one of the two substrate enantiomers. The photochemical step creates a planar carbon-centered radical and erases the absolute configuration at this position. While returning HAT to the same position would likely recreate the stereogenic center with the same absolute configuration, spectroscopic and quantum chemical studies suggest that the hydrogen atom is delivered from the photocatalyst to a heteroatom that is in conjugation to the radical center. Two scenarios can be distinguished for the hydrogen atom shuttling process. For hydantoins, N-carboxyanhydrides, and 4,7-diaza-1-isoindolinones, the back HAT occurs to a carbonyl oxygen atom or an imine-type nitrogen atom which is not involved in binding to the catalyst. For oxindoles and 2,5-diketopiperazines, a single lactam carbonyl group in the substrate is available to accept the hydrogen atom. It is currently assumed that back HAT occurs to this group, although the carbonyl oxygen atom is involved in hydrogen bonding to the catalyst. In comparison to the former reaction pathway, the latter process appears to be less efficient and more prone to side reactions. For both cases, an achiral enol or enamine is formed, which delivers upon dissociation from the catalyst statistically either one of the two stereoisomers of the substrate. Since only one substrate enantiomer (or diastereoisomer) is processed, a high enantioselectivity (or diastereoselectivity) results. Even though the editing is a contra-thermodynamic process, the described decoupling of a photochemical and a thermal step allows the usage of a single catalyst in loadings that vary between 2.5 and 10 mol % depending on the specific mode of action.