Open-Shell Iron Dipyrrin Complexes: Correlating Electronic Structure and Reactivity

Abstract

Despite significant advances in the field of homogeneous iron catalysis, systematic design of new catalyst classes remains challenging due to limited knowledge concerning the interplay between the electronic structure of a catalyst and its reactivity. This thesis presents our examination of the relationship between coordination environment, imparted electronic structure and resultant reactivity in weak-field iron dipyrromethene compounds. Three ferric dipyrromethene complexes [(ArL)FeX2] (ArL = 1,9-(2,4,6-Ph3C6H2)2-5-mesityldipyrromethene) bearing either chloride or tert-butoxide ligands were synthesized and the effect of the ancillary ligand was investigated. Extensive spectroscopic and magnetic characterization supported a high-spin (S = 5/2) ground state that was independent of ancillary ligand identity. Electrochemical experiments revealed, however, a significant effect of ancillary ligand substitution on the reduction potential of the ferric species. Additionally, the optical properties of the complexes undergo a pronounced change upon anion substitution, demonstrating how ancillary ligand identity can affect both ground state and excited state properties. We further showcased the influence of the X-type ligand on the propensity of the ferric complexes to initiate C–H bond activation and a correlation between hydrogen atom abstraction (HAA) reactivity and residual spin density on the ancillary ligand was established. Having observed C–H bond activation by a ferric precursor, we set out to target ferrous dipyrrin complexes featuring anionic nitroxido ligands to study their reactivity towards C–H bonds. The reaction of nitroxyl radicals, TEMPO (2,2’,6,6’-tetramethylpiperidinyloxyl) and AZADO (2-azaadamantane-N-oxyl), with an iron(I) synthon affords iron(II)-nitroxido complexes (ArL)Fe(κ1-TEMPO) and (ArL)Fe(κ2-N,O-AZADO). Both high-spin compounds are stable in the absence of weak C–H bonds, but decay to ferrous or ferric iron hydroxides in the presence of 1,4-¬cyclohexadiene. Mechanistic experiments reveal saturation behavior in C–H substrate and are consistent with rate-determining HAA at low substrate concentration. While the intermediacy of an iron-oxo cannot be fully excluded, direct participation of the iron(II)-nitroxido species in C–H activation is proposed. Diiron units featuring bridging oxygen ligands are key motifs found in metalloenzymes and commonly undergo extensive redox changes. To emulate these enzymatic systems, we prepared a diiron µ-oxo species (tBudmx)Fe2(µ-O) using a dinucleating Pacman dipyrrin ligand platform (tBudmx)H2 (1,9-(tert-butyl)2-5-dipyrromethene units bridged by 9,9-dimethyl-9H-xanthene) and explored its redox chemistry, resulting in isolation of a diferric complex (tBudmx)Fe2(µ-O)Cl2 and a mixed-valent salt [Cp2Co][(tBudmx)Fe2(µ-O)Cl2]. Both (tBudmx)Fe2(µ-O) and [Cp2Co][(tBudmx)Fe2(µ-O)Cl2] exhibit bridging oxo basicity and form diferrous µ-hydroxide species upon exposure to weak acids. In contrast, attempts to synthesize a diferric µ-hydroxide resulted in isolation of (tBudmx)Fe2(µ-O)Cl2 with concomitant loss of a proton. The results highlight the intricate interplay between oxidation state and reactivity in diiron µ-oxo units. Complementing our electronic structure investigations, we developed a novel iron-based catalyst that permits intramolecular amination of a full suite of C–H bond types without requiring activating or protecting groups. Substituted bicyclic, spiro, and fused heterocycles are all easily constructed using this method and thermally induced catalyst aggregation leads to facile product separation and catalyst recovery, enabling catalyst recycling.