A series of iron­(III) chloride and iron­(III) μ-oxo compounds supported by tetradentate amino-bis­(phenolate) ligands containing a homopiperazinyl backbone were prepared and characterized by electronic absorption spectroscopy, magnetic moment measurement, and MALDI-TOF mass spectrometry. The solid-state structures of three iron­(III) μ-oxo compounds were determined by single crystal X-ray diffraction and revealed oxo-bridged bimetallic species with Fe–O–Fe angles between 171.7 and 180°, with the iron centers in distorted square pyramidal environments. Variable temperature magnetic measurements show the oxo complexes exhibit strong antiferromagnetic coupling between two high-spin S = 5/2 iron­(III) centers. The oxo complexes exhibit poor activity for the reaction of carbon dioxide and epoxides in the presence of a cocatalyst, under solvent free conditions to yield cyclic carbonates. The least active iron oxo compound bears tert-butyl groups on the phenolate donors, and we propose that steric congestion around the iron center reduces catalytic activity in this case. We provide evidence that an epoxide deoxygenation step occurs when employing monometallic iron­(III) chlorido species as catalysts. This affords the corresponding μ-oxo compounds which can then enter their own catalytic cycle. Deoxygenation of epoxides during their catalytic reactions with carbon dioxide is frequently overlooked and should be considered as an additional mechanistic pathway when investigating catalysts.

A family of 17 iron­(III) aminobis­(phenolate) complexes possessing different phenolate substituents, coordination geometries, and donor arrangements were used as catalysts for the reaction of carbon dioxide (CO2) with epoxides. Matrix-assisted laser desorption/ionization time-of-flight mass spectrometry of the iron complexes with a bis­(triphenylphosphine)­iminium chloride cocatalyst in negative mode revealed the formation of six-coordinate iron “ate” species. Under low catalyst loadings (0.025 mol % Fe and 0.1 mol % chloride cocatalyst), all complexes showed good-to-excellent activity for converting propylene oxide to propylene carbonate under 20 bar of CO2. The most active complex possessed electron-withdrawing dichlorophenolate groups and for a 2 h reaction time gave a turnover frequency of 1240 h–1. Epichlorohydrin, styrene oxide, phenyl glycidyl ether, and allyl glycidyl ether could also be transformed to their respective cyclic carbonates with good-to-excellent conversions. Selectivity for polycarbonate formation was observed using cyclohexene oxide, where the best activity was displayed by trigonal-bipyramidal iron­(III) complexes having electron-rich phenolate groups and sterically unencumbering tertiary amino donors. Those containing bulky tertiary amino ligands or those with square-pyramidal geometries around iron showed no activity for polycarbonate formation. While the overall conversions declined with decreasing CO2 pressure, CO2 incorporation remained high, giving a completely alternating copolymer. The difference in the optimum catalyst reactivity for cyclic carbonate versus polycarbonate formation is particularly noteworthy; that is, electron-withdrawing-group-containing phenolates give the most active catalysts for propylene carbonate formation, whereas catalysts with electron-donating-group-containing phenolates are the most active for polycyclohexene carbonate formation. This study demonstrates that the highly modifiable aminophenolate ligands can be tailored to yield iron complexes for both CO2/epoxide coupling and ring-opening copolymerization activity.