Oxidatively-induced C(sp³)–C(sp³) bond formation at a tucked-in iron(III) complex

Abstract

Carbon–carbon (C–C) bond formation is a cornerstone of synthetic chemistry, relying on routes such as transition-metal mediated cross-coupling for the introduction of new carbon-based functionality. For {[M]ⁿ⁺–C} (M = metal) structural units, studies that offer well-defined relationships between metal oxidation state, hydrocarbon strain, and {[M]ⁿ⁺–C} bond thermochemistry are thus informative, providing a means to reliably access new product classes. Here, we show that one-electron oxidation of the iron tucked-in complex [(η⁶-C₅Me₄=CH₂)Fe(dnppe)] (dnppe = 1,2-bis(di-n-propylphosphino)ethane) results in C(sp³)–C(sp³) bond formation giving unique {Fe₂} dimers. Freeze-quenched CW X-band EPR spectroscopy allowed for spectroscopic identification of the reactive [(η⁶-C₅Me₄=CH₂)Fe(dnppe)]⁺ intermediate. Density functional theory (DFT) calculations reveal a primarily Fe-centered radical and a weak {[Fe]–C} bond (BDE_[Fe]–C = 24.5 kcal mol⁻¹, c.f. BDE_{C–C(ethane)} = 90 kcal mol⁻¹). For comparison, a structurally analogous Fe(III) methyl complex was prepared, [Cp*Fe(dnppe)(CH₃)]⁺ (Cp* = C₅Me₅⁻), where C(sp³)–C(sp³) coupling was not observed, consistent with a larger calculated BDE_[Fe]–C value of 47.8 kcal mol⁻¹. These data are analogized to the simple hydrocarbons ethane and cyclopropane, where a strain-induced BDE_C–C decrease of 33 kcal mol⁻¹ is witnessed on cyclization.

---

Introduction

Carbon–carbon (C–C) bond forming reactions are a cornerstone of synthetic chemistry. Most pharmaceutically active compounds contain a C–C bond, along with countless materials relevant to industry and everyday life. It thus comes as no surprise that transition metal mediated C–C bond forming reactions have been the focus of two Nobel prizes.^1,2 Using transition metals, one can form C–C bonds through cross-coupling – a sequence that proceeds via stepwise oxidative addition, transmetalation, and reductive elimination.³ In recent years, there has been a surge of reports that have driven such transformations using electro-^4–7 or photochemistry;^8–11 many of these studies have focused on cheaper and more abundant base transition elements as feedstocks.¹²

Reductive elimination is the main step associated with C–C bond formation.¹³ When reductive elimination from {[M]ⁿ⁺–C} is unfavourable and prohibits C–C bond generation, oxidation of the metal-center can be used to induce so called oxidatively-induced reductive elimination (ORE; Fig. 1A).^14–16 In this way, oxidation increases the tendency to undergo elimination (e.g., homolytic bond cleavage), thus returning the metal to a stable d-electron count.^17,18

Fig. 1 (A) Overview of intra/intermolecular oxidatively induced reductive elimination with transition metals. (B) Decrease in BDE_C–C in cyclopropane upon oxidation, and this work.

Sanford and co-workers have shown that one-electron oxidation of [(^tBu₂bpy)Pd^III(CH₃)₂]⁺ (^tBu₂bpy = 4,4′-di-tert-butyl-2,2′-bipyridine) results in C(sp³)–C(sp³) reductive elimination of ethane (CH₃CH₃) from an intermediate {[Pd^IV]-(CH₃)₃} species (Fig. 1A).¹⁹ Similarly, Baik, Chang, and co-workers showed that oxidation of an Ir(III) complex to Ir(IV) led to a 19 kcal mol⁻¹ reduction in the barrier associated with reductive elimination in a C–H arylation reaction.²⁰ These examples illustrate the power of ORE in C–C bond formation, providing opportunities for otherwise unfavourable bond construction.

Of {[M]ⁿ⁺–C} organometallics, sandwich complexes ([Cp/Cp*]₂M; Cp = C₅H₅⁻; Cp* = C₅Me₅⁻) represent a well-studied area owing to robust syntheses and product stability.^21,22 Lesser studied, however, are related “tucked-in” complexes, where a H⁺/H˙/H⁻ from the Cp*–CH₃ ligand has been removed, encouraging high reactivity.²³ This is especially true for first row late transition elements. As a salient example, it was reported that the ferrocenyl carbocation, [((η⁶-C₅Me₄)=CH₂)Fe(Cp*)]⁺ is reactive at temperatures >−82 °C.^24–26

Recently, we reported the first example of an Fe diphosphine tucked-in complex, [((η⁶-C₅Me₄)=CH₂)Fe(dnppe)] (dnppe = 1,2-bis(di-n-propylphosphino)ethane) (1) (Fig. 1B) and its reactivity with electrophiles including CO₂, HBCy₂ (Cy = cyclohexyl), Br–Au^I-PPh₃, benzaldehyde, and B(C₆F₅)₃.²⁷ Herein, we examine the redox behaviour of 1 using electrochemical, photochemical, and chemical means (Fig. 2A). Oxidation of the Fe-center in 1 led to rapid C(sp³)–C(sp³) formation between CH₂ groups of the η⁶-C₅Me₄=CH₂ ligand. Model studies and theoretical calculations were used to rationalize iron–carbon {[Fe]ⁿ⁺–C} bond strength (i.e., bond dissociation energy (BDE)) as a function of both iron oxidation state and ring-strain. These observations are analogized to cyclic organic compounds (e.g., cyclopropane) that also bear weak C–C bonds due to strain – a characteristic that has been leveraged for oxidatively-induced functionalization.^28,29

Fig. 2 (A) Synthesis of [1]⁺ and C(sp³)–C(sp³) coupled products. (B) Cyclic voltammograms for 1 and 2.

Results and discussion

An investigation into the oxidative chemistry of 1 was commenced by means of cyclic voltammetry (CV). Scanning from −1.90 V to +0.50 V in THF showed a quasi-reversible feature at E_1/2 = −1.23 V (i^red_P/i^ox_P = 0.89 at 200 mV s⁻¹; Fig. S43†)³⁰ and an irreversible feature at E_p,a = +0.15 V vs. Fc/Fc⁺ (Fc = ferrocene) (Fig. 2B); these are assigned as Fe(II/III) and Fe(III/IV) events, respectively.³¹ Varying the scan rate from 50 to 500 mV s⁻¹ did not change voltammogram shape (Fig. S44†). For other known Fe(II) half-sandwich complexes, [Cp*Fe^II(dppe)(X)] (dppe = 1,2-bis(diphenylphosphino)ethane; X = H, CH₃, F, Cl, Br, I), the Fe(II/III) couple was found to be −0.54 > E_1/2 > −0.88 V, indicating the E_1/2 value for 1 to be comparably cathodically shifted.^32,33 To probe the effect of hydrocarbon strain on electrochemical oxidation, an analogous {[Fe^II]–CH₃} complex [Cp*Fe^II(dnppe)(CH₃)] (2) (Fig. 2B) was examined.³⁴ Compound 2 showed a reversible feature at E_1/2 = −1.16 V with i^red_P/i^ox_P near 1.00 at 200 mV s⁻¹ (Fig. S45†), and an irreversible feature at E_p,a = +0.35 V vs. Fc/Fc⁺ in THF, indicating that the tucked-in η⁶-C₅Me₄=CH₂ ligand contributes to a 70 mV cathodic shift in the Fe(II/III) couple. For both 1 and 2, the Fe(III/IV) couple was irreversible – a reported outcome for [Cp*Fe(diphosphine)(CH₃)]⁺-type complexes.³²

As a result of our observations from the CV data, complex 1 was next exposed to 1 equiv. of [Cp₂Fe]B(C₆F₅)₄ at 25 °C (Fig. 2A) in CH₃CN. Monitoring this reaction by ³¹P{¹H} NMR spectroscopy showed the formation of diamagnetic products in the range of ca. δ_P = 78 ppm, consistent with complexes of the type [Cp*Fe(diphosphine)(NCCH₃)]⁺.³³ This hypothesis was further supported by FT-IR spectroscopy, providing a band at ν(C≡N) = 2252 cm⁻¹ (c.f., ν(C≡N) = 2234 cm⁻¹) for [Cp*Fe(dnppe)(NCCH₃)]⁺, prepared independently (vide infra; [7]⁺) (Fig. S28†).

Supposing the main product from oxidation of 1 to be the C(sp³)–C(sp³) coupled {μ-H₂CCH₂⁻} dimer, [{η⁵-C₅Me₄–(μ-CH₂)}Fe^II(dnppe)(NCCH₃)]₂²⁺ ([5]²⁺), the redox behaviour of 1 was probed in the presence of 4 mL of THF with 100 μL CH₃CN added.³⁵ Consistent with an electrochemical–chemical (EC) reaction (onwards reactivity), scanning oxidatively from −2.00 to +1.00 V showed the previously discussed quasi-reversible couple at E_1/2 = −1.23 V, with a decreased value of i^red_P/i^ox_P = 0.62 (c.f., 0.89 in THF at 200 mV s⁻¹). In addition to the observed signature for 1, an additional wave at E_1/2 = −0.16 V was assigned to the Fe(II/III) couple of [5]²⁺, confirmed through independent preparation (vide infra; Fig. S51†).

Outlining an alternative preparation of [5]²⁺, we envisaged treatment of 1 with a halide-based oxidant to initiate C(sp³)–C(sp³) coupling, followed by halide abstraction in CH₃CN (Fig. 2A). Diiodoethane (ICH₂CH₂I), previously shown to oxidize {Fe(diphosphine)_x} complexes,³⁶ was selected as the candidate oxidant, supplying iodide groups that could later be removed. In accord, exposure of 1 to 1 equiv. of ICH₂CH₂I showed an immediate colour change from orange to dark red (Fig. 2A). Analysis by ³¹P{¹H} NMR spectroscopy, however, showed an absence of signals, suggestive of paramagnetism. Growth of crystals from a THF solution layered with pentane at −35 °C overnight provided material appropriate for an Evans method measurement, giving μ_eff = 2.66 μ_B (298 K, CDCl₃), falling within the expected range of a molecule containing two unpaired electrons. Further analysis by single-crystal X-ray diffraction (scXRD) revealed a connectivity map that supported a dimeric structure of [{η⁵-C₅Me₄–(μ-CH₂)}Fe^III(dnppe)(I)]₂(I)₂ ([3](I)₂) (Fig. S54†). This molecule features a bridging {μ-H₂CCH₂−} unit that links two {(C₅Me₄–(CH₂))Fe^III(dnppe)(I)} moieties. Consistent with the μ_eff value obtained from Evans method and the two exogeneous iodide counteranions observed by scXRD, both Fe-centers are formally Fe(III). High-resolution electrospray ionization mass-spectrometry (HRMS-ESI) additionally provides a signal for [3]⁺ at m/z = 1158.299 (calcd. m/z = 1158.293).

Formation of [3](I)₂ is a four-electron process involving two successive oxidations of 1. Considering the Fe(II) resonance contributor for 1 (Fig. 1B), oxidation leads to homolytic bond cleavage, forming an S = 1 Fe(II) center³⁷ and a {C₅Me₄–(·CH₂)} radical (vide infra). Recombination of two {C₅Me₄–(·CH₂)} radicals results in C(sp³)–C(sp³) bond formation. Iodide coordination and subsequent oxidation to Fe(III) (by excess ICH₂CH₂I) thus provides the observed product [3](I)₂ (Fig. 2A). To provide the Fe(II) complex [{η⁵-C₅Me₄–(μ-CH₂)}Fe^II(dnppe)(I)]₂ (4), 0.5 equiv. of ICH₂CH₂I was added to 1 (to preclude oxidation from Fe(II) to Fe(III)). However, this reaction resulted in mixtures of products, with at least two ³¹P{¹H} resonances appearing in the range of 4 (δ_P = ca. +80 ppm), unreacted 1 (δ_P = +89.6 ppm), and free dnppe (δ_P = −28.3 ppm) (Fig. S30†).

As an alternative route towards 4, reduction of [3](I)₂ was explored using cobaltocene ([Cp₂Co]). Addition of 2 equiv. of [Cp₂Co] to a THF solution of [3](I)₂ produced an insoluble yellow precipitate and a purple supernatant over 3 h. Extraction of the purple product into toluene and analysis by NMR spectroscopy revealed a diamagnetic product. By ³¹P{¹H} NMR spectroscopy, one signal at δ_P = +78.5 ppm was observed, standard of [Cp*Fe^II(dnppe)(halide)]-type compounds (e.g., for halide = Cl, δ_P = +79.4 ppm).³⁴ Analysis of the ¹H NMR spectrum also revealed a characteristic signal for the bridging {μ-H₂CCH₂−} moiety (δ_H = 2.57 ppm) as well as two pairs of signals for the unsymmetric methyl groups (δ_H = 1.89 and 1.82 ppm). Crystals of 4 were grown from slow evaporation of a benzene solution overnight; analysis by scXRD confirmed its structural identity as [{η⁵-C₅Me₄–(μ-CH₂)}Fe^II(dnppe)(I)]₂ (4) (Fig. 3). Offering context, a Cambridge Structural Database (CSD) search provides fewer than 5 examples of structurally-authenticated Cp*-based compounds featuring a bridging {–H₂CCH₂−} unit.^24,38–43 Using pre-formed ligand constructs, Wang and co-workers reported a set of polymeric Ir complexes bridged by {C₅Me₄–(CH₂)_n–C₅Me₄} (n = 2, 3, 4);⁴⁰ while Brintzinger and co-workers synthesized monometallic complexes of the type [(C₅Me₄–(CH₂)₂–C₅Me₄)MCl₂] (M = Ti, Zr).⁴¹ Another example has been accessed by Cp*{H˙} removal – Severin and co-workers utilized [OAl(C₆F₅)₃]˙⁻ in the presence of Al(C₆F₅)₃ to perform H-atom abstraction of , generating a [(η⁵-C₅Me₄)(·CH₂)]Fe(Cp*)] radical,³⁸ which dimerized to give [{η⁵-C₅Me₄–(μ-CH₂)}Fe^II(Cp*)]₂. This example, however, does not require Fe oxidation nor does it inform on the relationship between iron-ring-strain and C–C coupling.

Fig. 3 Molecular structure of 4 with ellipsoids drawn at 50% probability. Hydrogen atoms omitted except for those on C(1) and C(2). Solvent molecule (benzene) has been omitted for clarity.

Compound 4 is a viable candidate for halide abstraction in CH₃CN. Reaction of a CH₃CN solution of 4 with 2 equiv. K[B(C₆F₅)₄] (Fig. 2A) produced an immediate colour change from purple to orange. After workup, FT-IR (ATR) spectroscopic analysis showed a stretch at ν(N≡C) = 2252 cm⁻¹, corroborating formation of [{η⁵-C₅Me₄-(μ-CH₂)}Fe^II(dnppe)(NCCH₃)]₂[B(C₆F₅)₄]₂ ([5](B(C₆F₅)₄)₂). The ¹H NMR spectrum of [5]²⁺ also showed the expected inequivalent signals (δ_H = 1.63 and 1.48 ppm), a {μ-CH₂} signal (δ_H = 2.22 ppm), and a peak for [Fe]-NCCH̲₃ (δ_H = 2.41 ppm; c.f., δ_H = 1.91 ppm for [Cp*Fe(dnppe)(NCCH₃)]⁺ ([7]⁺)). Comparison of the FT-IR spectra for the reaction between 1 and [Cp₂Fe]B(C₆F₅)₄, [5]²⁺, [7]⁺, and neat CH₃CN, confirmed that [5]²⁺ was formed from oxidation of 1 (Fig. 4) with an observed stretch at ν(N≡C) = 2252 cm⁻¹. Notably, the combination band for neat CH₃CN (ν = 2300 cm⁻¹) also disappears upon CH₃CN coordination.^44,45 In addition to our observations from IR spectroscopy, a THF solution of 4 exposed to 100 μL of CH₃CN and analyzed by CV displayed a predominant wave at E_1/2 = −0.16 V, akin to the voltammogram acquired from the addition of 100 μL CH₃CN to 1, noted above (Fig. S51†).

Fig. 4 Enhanced view of the region between 2400–2150 cm⁻¹ for the FT-IR (ATR) spectra for the reaction between 1 and 1 equiv. of [Cp₂Fe]B(C₆F₅)₄ (red), [5]²⁺ (black), [7]⁺ (dark grey), and neat CH₃CN (light grey).

Compound 4 is also a valuable synthon for clean access to dimeric iron hydrides. For example, exposure of a THF solution of 4 to 2 equiv. of K[HBEt₃] led to clean formation of the bis-hydride complex [{η⁵-C₅Me₄–(μ-CH₂)}Fe^II(dnppe)(H)]₂ (6) (Fig. 2A).⁴⁶ By ¹H NMR spectroscopy, the bridging {μ-H₂CCH₂−} group was located at δ_H = 2.62 and the inequivalent signals were observed at δ_H = 2.05 and 1.96 ppm. Successful [Fe]–H formation was further established by locating the [Fe]–H̲ signal as a triplet at δ_H = −17.81 ppm (²J_H–P = 70.2 Hz) (c.f., δ_H = −17.90 ppm (²J_H–P = 70.2 Hz) for the monomeric [Cp*Fe(dnppe)(H)]).³⁴ FT-IR spectroscopy also revealed a characteristic ν(Fe–H) stretch at 1847 cm⁻¹.

Having characterized the terminal C(sp³)–C(sp³) coupled product, we next sought to characterize the reactive oxidized precursor, [1]⁺. Oxidation of 1 was thus carried out in 2-MeTHF using [Cp₂Fe]⁺ at −78 °C. Following 5 min of reaction time, freeze-quenched continuous wave (CW) X-band EPR spectroscopy provided a spectrum with rhombic symmetry (g = [2.0004, 2.0526, 2.2134]) (Fig. 5A) – this represents the first example of a spectroscopically characterized Fe(III) tucked-in complex by EPR. Compound [1]⁺ can likewise be photochemically generated in a 2-MeTHF glass using 1 and a 405 nm laser (350 mW mm⁻²) following 10 min of irradiation (Fig. 5B). To provide a comparative basis, the {[Fe^III]–CH₃} complex [2]⁺ was also prepared by reaction of 2 with 1 equiv. of [Cp₂Fe]⁺ in 2-MeTHF at −78 °C. By CW X-band EPR spectroscopy, a rhombic signal was also observed (g = [1.9827, 2.0450, 2.3968]) (Fig. 5A), again consistent with an S = 1/2 Fe(III) radical species ([2]⁺).⁴⁷ The g_max shift between compounds [1]⁺ and [2]⁺ is attributed to a difference in geometry about Fe. For [1]⁺, g-values are all greater than 2.0 with a spread of 0.05 to 0.16 – characteristic of the unpaired electron predominately residing in a d_z² orbital.⁴⁸ For [2]⁺, a larger value of g_max is consistent with an unpaired electron residing in a d_x²−_y² orbital.⁴⁹ Calculated spin density plots for [1]⁺ and [2]⁺ additionally corroborate that the unpaired electron predominately resides on Fe, with populations of 0.917e⁻ for [1]⁺ and 0.987e⁻ for [2]⁺ (Fig. 5C). The marked geometry change noted above (between [1]⁺ and [2]⁺) is best visualized from an overlay plot of their DFT-optimized structures (with coincident Fe and P atoms) (Fig. 6A). This data also showcases an elongated Fe–C bond in [1]⁺ (d(Fe–C) = 2.24 Å) as compared to [2]⁺ (d(Fe–C) = 2.01 Å), indicating weaker Fe–C overlap in [1]⁺. The ring-strain inherent to [1]⁺ is additionally described by an Fe–C–C triangle with angles of 38.4°, 60.5°, and 81.1° (Fig. 6A).

Fig. 5 A) Freeze-quenched CW X-band EPR spectra (8.885 GHz) recorded in 2-MeTHF glass at 77 K for [1]⁺ (black) and [2]⁺ (red). (B) Freeze-quenched CW X-band EPR spectra for [1]⁺ (top) and a photooxidized sample of 1 under 405 nm (350 mW mm⁻²) irradiation (bottom). (C) DFT-calculated spin-density plots for [1]⁺ and [2]⁺ (all C and H atoms from the dnppe ligand, except those C atoms directly attached to P, have been omitted for clarity). Fc = Cp₂Fe.

Fig. 6 (A) Structural overlay (capped sticks) of [1]⁺ (black) and [2]⁺ (red). Select angles (°) and select bond lengths (Å) are provided. (B) Mayer bond order (MBO) and natural bond orbital (NBO) analysis providing E⁽²⁾ values between [Fe]–C. (C) Reaction of [1]⁺ with (2,2,6,6-tetramethylpiperidin-1-yl)oxyl (TEMPO) forming [8]⁺.

Mayer bond order (MBO) and natural bond orbital (NBO) analysis provide further insight that helps to rationalize C(sp³)–C(sp³) coupling in [1]⁺ (Fig. 6B).⁵⁰ To determine the change in bond order upon oxidation, MBOs were calculated. For 1/[1]⁺, the MBO decreased by 0.06 on oxidation (1: 0.66/[1]⁺: 0.60), while for 2/[2]⁺, the MBO was found to slightly increase (2: 0.73/[2]⁺: 0.80). NBO calculations provided second order perturbation energies (E⁽²⁾) which showed a similar trend. For the C(lp) → Fe(lv) (lp = lone pair, lv = lone valence) interaction, energies were found to be 43.4 and 53.8 kcal mol⁻¹ for [1]⁺ and [2]⁺, respectively (Fig. 6B). This data indicates that (1) the Fe–C bond becomes weaker on going from 1 → [1]⁺ and (2) that the Fe–C bond of 1/[1]⁺ is weaker than 2/[2]⁺ – a consequence of ring-strain.

The radical character intrinsic to [1]⁺ is also evidenced by its chemical reactivity with 10 equiv. of the radical trap, (2,2,6,6-tetramethylpiperidin-1-yl)oxyl (TEMPO), which gives [{η⁵-C₅Me₄–(CH₂–ONC₉H₁₈)}Fe^II(dnppe)(NCCH₃)]B(C₆F₅)₄ ([8]B(C₆F₅)₄) (Fig. 6C). Compound [8]⁺ was characterized by signals in the ¹H NMR spectrum corresponding to {CH̲₂–ONC₉H₁₈} at δ_H = 4.18 ppm and Fe-bound NCCH₃ at δ_H = 2.40 ppm. Confirmation of an Fe-bound NCCH₃ group was determined via FT-IR spectroscopy, with a characteristic ν(N≡C) stretch at 2247 cm⁻¹. The ¹³C{¹H} NMR spectrum also showed a signal at δ_C = 71.0 ppm, corresponding to the TEMPO-bound CH₂ from the η⁵-C₅Me₄–(CH₂–ONC₉H₁₈) ligand. HRMS-ESI also provided a signal corresponding to [8]⁺ at 649.407 (calcd; 649.407). These data corroborate a radical mechanism, whereby the generated {C₅Me₄–(·CH₂)} group has been trapped by TEMPO.⁵¹

To assess the effect of ring-strain on {[Fe]–C} bond homolysis, we next sought information regarding the BDE_[Fe]–C of [1]⁺. For reference, the BDE_C–C associated with the simplest cyclical saturated hydrocarbon (cyclopropane; (CH₂)₃) is substantially lower than ethane due to ring strain (BDE_{C–C(cyclopropane)} = 57 kcal mol⁻¹vs. BDE_{C–C(ethane)} = 90 kcal mol⁻¹).^52–54 Given this, and assuming CH₃CH₃ and RCH₂CH₂R (R = C₅Me₄⁻) to have similar bond strengths, C–C coupling should be favoured for {[Fe]–CH₂R} when BDE_[Fe]–C ≤ 45 kcal mol⁻¹.

Computations show the difference in BDE_[Fe]–C for [1]⁺ and [2]⁺ to be 23.3 kcal mol⁻¹ (Fig. 7). The absolute values of BDE_[Fe]–C: 24.5 and 47.8 kcal mol⁻¹ for [1]⁺ and [2]⁺, respectively are also consistent with observation of C(sp³)–C(sp³) coupling for the former. These calculated values are benchmarked against ethane, which provides BDE_C–C(calcd) = 88.4 kcal mol⁻¹c.f., the experimental BDE_C–C = 90 kcal mol⁻¹.⁵² Notwithstanding differences in BDE_C–C, both processes were calculated to be mildly exergonic (Fig. 7): ΔG° = −2.1 kcal mol⁻¹ and −4.3 kcal mol⁻¹ for C–C coupling from [1]⁺ or [2]⁺, respectively, suggesting a difference in kinetic barrier. Based on the above discussion, the C(sp³)–C(sp³) coupling associated with [1]⁺ can thus be ascribed to a minimum of two features: (1) a strained three-membered {–Fe–CH₂–C–} ring and (2) a weakened {[Fe^III]–CH₂R} bond c.f., 1.³² This is contrasted with [2]⁺, where experimentally, CH₃CH₃ is not evolved during synthesis (Fig. 7B).

Fig. 7 (A) Thermodynamic calculations showing energies associated with the dimerization of [1]⁺. (B) Analogous calculations for the elimination of ethane (CH₃CH₃) from [2]⁺. All energies are in kcal mol⁻¹.

Conclusions

Metal-mediated cross-coupling has emerged as a powerful means to construct C–C bonds. In recent years, such studies have focused predominantly on base transition elements, such as iron, cobalt, and nickel, as cheaper and more sustainable sources. Here, we have studied an oxidized iron tucked-in complex [(η⁶-C₅Me₄=CH₂)Fe^III(dnppe)]⁺ ([1]⁺), which due to both ring-strain and iron oxidation state, promotes C(sp³)–C(sp³) bond coupling. The product {Fe₂} diphosphine dimers invite reactions to give functionalized {Fe^III₂} or {Fe^II₂} products having coordinated halide, hydride, or solvent equivalents. Characterization of the reactive oxidized “tucked-in” {Fe^III} precursor was performed using CW X-band EPR spectroscopy, with computations revealing a weak {[Fe]^III–C} bond with BDE_[Fe]–C = 24.5 kcal mol⁻¹. The observed C(sp³)–C(sp³) coupling, however, is not simply an artifact of iron oxidation state. The model, [Cp*Fe^III(dnppe)(CH₃)]⁺ [2]⁺, for instance, does not undergo C(sp³)–C(sp³) coupling, consistent with a higher calculated BDE_[Fe]–C of 47.8 kcal mol⁻¹. This pair of compounds and their contrasting thermochemistry/reactivity pave the way towards better understanding the intricacies of {M–C} strain and metal oxidation state on reductive elimination, offering insight into an important elementary transformation of direct relevance to C–C bond coupling.

Data availability

The data supporting this article have been included as part of the ESI.† Crystallographic data for 4 has been deposited at the CCDC under 2341812. This data can be obtained free of charge from the cambridge crystallographic data centre viahttps://www.ccdc.cam.uk/data_request/cif.

Author contributions

Joseph A. Zurakowski: investigation, methodology, formal Analysis, visualization, writing – original draft, writing – review & editing. Connor S. Durfy: investigation, validation. Noah B. Stocek: investigation, methodology. Giovanni Fanchini: resources, methodology, funding acquisition. Marcus W. Drover: project administration, funding acquisition, visualization, supervision, formal analysis, conceptualization, methodology, writing – review & editing.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

The authors are grateful to Western University, the Council of Ontario Universities for a John C. Polanyi award to M. W. D., the Canadian Foundation for Innovation (LOF-212442), and the Natural Sciences and Engineering Research Council of Canada (Discovery Grant, RGPIN-2020-04480 (M. W. D.), RGPIN-2020-05008 (G. F.), Discovery Launch Supplement, DGECR-2020-00183), and graduate award (CGS-D/NSERC Vanier to J. A. Z.) for funding.

References

1. R. H. Grubbs, Angew. Chem., Int. Ed., 2006, 45, 3760–3765.
2. C. C. C. Johansson Seechurn, M. O. Kitching, T. J. Colacot and V. Snieckus, Angew. Chem., Int. Ed., 2012, 51, 5062–5085.
3. A. Biffis, P. Centomo, A. Del Zotto and M. Zecca, Chem. Rev., 2018, 118, 2249–2295.
4. B. Zhang, Y. Gao, Y. Hioki, M. S. Oderinde, J. X. Qiao, K. X. Rodriguez, H.-J. Zhang, Y. Kawamata and P. S. Baran, Nature, 2022, 606, 313–318.
5. Y. Liu, P. Li, Y. Wang and Y. Qiu, Angew. Chem., Int. Ed., 2023, 62, e202306679.
6. P. Li, G. Kou, T. Feng, M. Wang and Y. Qiu, Angew. Chem., Int. Ed., 2023, 62, e202311941.
7. Y. Wang, S. Dana, H. Long, Y. Xu, Y. Li, N. Kaplaneris and L. Ackermann, Chem. Rev., 2023, 123, 11269–11335.
8. C. K. Prier, D. A. Rankic and D. W. C. MacMillan, Chem. Rev., 2013, 113, 5322–5363.
9. C.-S. Wang, P. H. Dixneuf and J.-F. Soulé, Chem. Rev., 2018, 118, 7532–7585.
10. M. De Abreu, P. Belmont and E. Brachet, Eur. J. Org Chem., 2020, 2020, 1327–1378.
11. L. K. G. Ackerman, J. I. Martinez Alvarado and A. G. Doyle, J. Am. Chem. Soc., 2018, 140, 14059–14063.
12. M. L. Clapson, C. S. Durfy, D. Facchinato and M. W. Drover, Cell Rep. Phys. Sci., 2023, 4, 101548.
13. X. Chen, K. M. Engle, D. Wang and J. Yu, Angew. Chem., Int. Ed., 2009, 48, 5094–5115.
14. M. B. Watson, N. P. Rath and L. M. Mirica, J. Am. Chem. Soc., 2017, 139, 35–38.
15. J. Kim, K. Shin, S. Jin, D. Kim and S. Chang, J. Am. Chem. Soc., 2019, 141, 4137–4146.
16. A. Cizikovs, E. E. Basens, P. A. Zagorska, A. Kinens and L. Grigorjeva, ACS Catal., 2024, 14, 1690–1698.
17. A. Pedersen and M. Tilset, Organometallics, 1993, 12, 56–64.
18. W. Lau, J. C. Huffman and J. K. Kochi, Organometallics, 1982, 1, 155–169.
19. M. P. Lanci, M. S. Remy, W. Kaminsky, J. M. Mayer and M. S. Sanford, J. Am. Chem. Soc., 2009, 131, 15618–15620.
20. K. Shin, Y. Park, M.-H. Baik and S. Chang, Nat. Chem., 2018, 10, 218–224.
21. T. J. Kealy and P. L. Pauson, Nature, 1951, 168, 1039–1040.
22. P. J. Chirik, Organometallics, 2010, 29, 1500–1517.
23. P. Preethalayam, K. S. Krishnan, S. Thulasi, S. S. Chand, J. Joseph, V. Nair, F. Jaroschik and K. V. Radhakrishnan, Chem. Rev., 2017, 117, 3930–3989.
24. A. Z. Kreindlin, F. M. Dolgushin, A. I. Yanovsky, Z. A. Kerzina, P. V. Petrovskii and M. I. Rybinskaya, J. Organomet. Chem., 2000, 616, 106–111.
25. E. I. Fedin, A. L. Blumenfeld, P. V. Petrovskii, A. Z. Kreindlin, S. S. Fadeeva and M. I. Rybinskaya, J. Organomet. Chem., 1985, 292, 257–268.
26. A. Z. Kreindlin, S. S. Fadeeva and M. I. Rybinskaya, Izv. Akad. Nauk SSSR, Ser. Khim., 1984, 2, 403–406.
27. J. A. Zurakowski and M. W. Drover, Chem. Commun., 2023, 59, 11349–11352.
28. J. P. Dinnocenzo, W. P. Todd, T. R. Simpson and I. R. Gould, J. Am. Chem. Soc., 1990, 112, 2462–2464.
29. P. Peng, X. Yan, K. Zhang, Z. Liu, L. Zeng, Y. Chen, H. Zhang and A. Lei, Nat. Commun., 2021, 12, 3075.
30. T. Krämer, M. R. Gyton, I. Bustos, M. J. G. Sinclair, S. Tan, C. J. Wedge, S. A. Macgregor and A. B. Chaplin, J. Am. Chem. Soc., 2023, 145, 14087–14100.
31. Given the similarity of E_1/2 for both 1 and 2, along with the lower reduction potentials (E_1/2 ≪ −1.23 V vs. Fc/Fc⁺) observed for related Fe(0) species, a formal Fe(II) oxidation state seemingly better describes the observed electrochemical behaviour of 1.
32. M. Tilset, I. Fjeldahl, J.-R. Hamon, P. Hamon, L. Toupet, J.-Y. Saillard, K. Costuas and A. Haynes, J. Am. Chem. Soc., 2001, 123, 9984–10000.
33. D. Patel, A. Wooles, A. D. Cornish, L. Steven, E. S. Davies, D. J. Evans, J. McMaster, W. Lewis, A. J. Blake and S. T. Liddle, Dalton Trans., 2015, 44, 14159–14177.
34. J. A. Zurakowski, K. R. Brown and M. W. Drover, Inorg. Chem., 2023, 62, 7053–7060.
35. Complex 1 is insoluble in pure acetonitrile and was found to react with CH₂Cl₂.
36. F. S. Pick, D. B. Leznoff and M. D. Fryzuk, Dalton Trans., 2018, 47, 10925–10931.
37. P. Hamon, L. Toupet, J.-R. Hamon and C. Lapinte, Organometallics, 1996, 15, 10–12.
38. Y. Liu, E. Solari, R. Scopelliti, F. Fadaei Tirani and K. Severin, Chem.–Eur. J., 2018, 24, 18809–18815.
39. M. Kessler, S. Hansen, C. Godemann, A. Spannenberg and T. Beweries, Chem.–Eur. J., 2013, 19, 6350–6357.
40. X. Tan, B. Li, S. Xu, H. Song and B. Wang, Organometallics, 2013, 32, 3253–3261.
41. F. Wochner, L. Zsolnai, G. Huttner and H. H. Brintzinger, J. Organomet. Chem., 1985, 288, 69–77.
42. D. V. Muratov, A. S. Romanov, M. Corsini, A. R. Kudinov, F. F. de Biani and W. Siebert, Chem.–Eur. J., 2017, 23, 11935–11944.
43. X. Wang, M. Sabat and R. N. Grimes, Organometallics, 1995, 14, 4668–4675.
44. Y. Kwon, C. Lee and S. Park, Chem. Phys., 2014, 445, 38–45.
45. B. Dereka, N. H. C. Lewis, J. H. Keim, S. A. Snyder and A. Tokmakoff, J. Phys. Chem. B, 2022, 126, 278–291.
46. E. S. Wiedner, M. B. Chambers, C. L. Pitman, R. M. Bullock, A. J. M. Miller and A. M. Appel, Chem. Rev., 2016, 116, 8655–8692.
47. J. Morrow, D. Catheline, M. H. Desbois, J. M. Manriquez, J. Ruiz and D. Astruc, Organometallics, 1987, 6, 2605–2607.
48. D. Chong, W. E. Geiger, N. A. Davis, A. Weisbrich, Y. Shi, A. M. Arif and R. D. Ernst, Organometallics, 2008, 27, 430–436.
49. F. Paul and C. Lapinte, Coord. Chem. Rev., 1998, 178–180, 431–509.
50. A. J. Bridgeman, G. Cavigliasso, L. R. Ireland and J. Rothery, J. Chem. Soc., Dalton Trans., 2001, 2095–2108.
51. Of note, complexes 1 and 2 exhibited no reactivity with TEMPO.
52. B. Ruscic, J. Phys. Chem. A, 2015, 119, 7810–7837.
53. F. H. Seubold, J. Phys. Chem., 1954, 22, 945–946.
54. X. Lei, X. Cao, J. Wang and X. Li, Combust. Flame, 2022, 237, 111881.

---

Footnote

† Electronic supplementary information (ESI) available: Experimental procedures, spectroscopic data, computational details. CCDC 2341812. For ESI and crystallographic data in CIF or other electronic format see DOI: https://doi.org/10.1039/d4sc03292f

---