The diiron active site plays a crucial role in the transport and activation of oxygen in biological systems [[1], [2], [3]]. Upon the activation of oxygen, non-heme diiron metalloenzymes are capable of executing a wide range of oxidative transformations. These include CH activations as seen in sMMO [[4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16]], BesC [17,18], UndA [[19], [20], [21]], PtmU3 [22], and Δ9 Desaturase [[23], [24], [25], [26]], as well as the epoxidation of observed in BoxB [27,28], N‑oxygenation occurring in AurF [[29], [30], [31], [32]] and CmlI [31,[33], [34], [35], [36], [37]], and N-hydroxylation in SznF [[38], [39], [40]]. In non-heme diiron metalloenzymes, the activation of molecular oxygen can give rise to diverse diiron-(hydro)peroxo species, as delineated in Scheme 1. These species can either directly facilitate oxidative transformations or undergo OO bond cleavage to generate a range of oxo intermediates [1,[41], [42], [43], [44]].

Inspired by the catalysis of non-heme diiron metalloenzymes, various ligand supported diiron active site have been artificially synthesized, which were mainly intended to mimick the structure and reactivity of the native metalloenzymes [[45], [46], [47]], or even expand the substrate scope and catalytic efficiency of the corresponding metalloenzymes [48]. Particularly, numerous diiron complexes have been engineered as synthetic analogs to emulate the catalytic activity of sMMO, a pivotal enzyme facilitating the transformation of methane into methanol [13]. In 2001, Jacobsen and coworkers reported the first example of a methane monooxygenase (MMO) model system capable of catalyzing the epoxidation of alkenes [49]. In more recent research, Que. and colleagues have meticulously characterized model complexes that feature [Fe(III)Fe(IV)(μ-O)2] and [Fe(IV)2(μ-O)2] core structures, thereby providing synthetic analogs for the intermediate Q stage in sMMO [50]. Recently, Liu and coworkers reported a series of the carboxylate and hydroxo-bridged diiron(III) complexes, which feature both N and O ligands and are structurally similar to the resting state of sMMO [51]. These synthetic sMMO analogs exhibit efficient catalytic performance for the dehydrogenation of indolines using hydrogen peroxide as the oxidant (Fig. 1A). A KIE study revealed a KIE value of 2.7 at the C1H site. This suggests that the rate-determining step might encompass the cleavage of the C1H bond, as depicted in Fig. 1B.

Then, a plausible mechanism has been suggested for diiron(III)-catalyzed dehydrogenation of indolines in presence of H2O2 (Fig. 1). The reaction commences with the interaction between diiron(III)-OH and H2O2, resulting in the formation of a μ-1,2-(hydro)peroxo diiron intermediate 1. The subsequent OO cleavage of 1 forms the [Fe(IV)2(μ-O)(μ-OH)]complex 2. Within this complex, the bridging oxo moiety is capable of executing a hydrogen atom abstraction (HAA) from the substrate, leading to the substrate radical along with the mixed-valent diiron(III,IV) 3. A subsequent single electron transfer (SET) from the substrate radical to the diiron center in complex 3 yields a substrate cation and a di(μ-hydroxo)diiron(III) complex, which is denoted as complex 4. The concluding step involves the deprotonation of the substrate cation, culminating in the synthesis of the product imine.

Nevertheless, the mechanisms of diiron(III)-catalyzed dehydrogenation remain elusive regarding: (1) The diiron(III)-OOH species may exist in different forms, but it is unclear which specific form actively participates in catalytic reactions. (2) Though the biomimetic diiron catalysts shares the similar active architecture with sMMO, the presence of phenolate anion ligands in the biomimetic complex introduces a variable that could potentially alter the reactivity of the diiron cores. (3) The biomimetic complex features a crowded coordination environment, which may also affect the structure and reactivity of diiron‑oxygen species. To address the aforementioned challenges, extensive DFT calculations have been conducted. Our computational analysis has identified that a Fe(III)Fe(III)-1,1-μ-hydroperoxy active species plays a pivotal role in the dehydrogenation of indolines. Contrary to a radical-mediated pathway, our findings suggest that the reaction proceeds through a hydride transfer mechanism. Specifically, this transfer occurs from the C1 position of the substrate to the Fe(III)Fe(III)-1,1-μ-hydroperoxy active species, following a non-redox pathway.