Diiron µ2-N2 complexes in bimetallic, four-fold bond activations of SiH4 to produce µ2-silicide complexes Authors: Rhett T. Bryant, Rex C. Handford, Suhashini Handunneththige, Simon J. Teat, Michael B. Hall, T. Don Tilley Published: 2026 ABSTRACT Fe(0) dinitrogen complexes [SiP3R]Fe–N Created by potrace 1.16, written by Peter Selinger 2001-2019 N–Fe[SiP3R] (1-R; [SiP3R] = PhSi(CH2PR2)3; R = iPr, Ph) are accessed by one- or two-electron reduction of Fe(i) and Fe(ii) chloride complexes [SiP3iPr]FeCl or κ2-[SiP3R]FeCl2 (R = iPr, Ph), respectively, and fully characterized by means of multinuclear NMR spectroscopy, single-crystal X-ray diffraction (SC-XRD), and Mössbauer spectroscopy. Complexes 1-R represent rare examples of bimetallic Fe(0) compounds with dinitrogen as a bridging ligand and have an overall spin of S = 0. Heating 1-R in the presence of SiH4 gas resulted in four-fold Si–H activation to furnish diiron complexes [SiP3R](H)2Fe Created by potrace 1.16, written by Peter Selinger 2001-2019 SiFe(H)2[SiP3R] (2-R; R = iPr, Ph), which were characterized by multinuclear NMR spectroscopy, SC-XRD, Mössbauer spectroscopy, and density functional theory calculations. Complexes 2-R exhibit the shortest iron-silicon bonds to date and nearly linear FeSiFe cores. This work extends a previously developed strategy for SiH4 activation from cobalt to iron, while offering insight into the silicon-atom extrusion from SiH4 in molecular systems. Introduction The activation of Si–H bonds by transition metals is a key method for obtaining transition metal–silicon complexes, which display a remarkably diverse range of bonding motifs and reactivity.1–3 In homogeneous catalysis, hydrosilation of olefins by the Chalk–Harrod and modified Chalk–Harrod mechanisms involve oxidative addition of a Si–H bond as a crucial mechanistic step, yielding a metal silyl complex (Ln(H)M–SiR3) as a catalytic intermediate.4 Relatedly, two Si–H bonds of a hydrosilane can also be cleaved in a concerted manner at a single metal center, yielding a metal silylene complex (Ln(H)2MSiR2).5 Investigations have shown that such a concerted, two-fold Si–H activation appears to be a key mechanistic step in some catalytic systems.5 More recently, this laboratory has demonstrated that certain bimetallic systems (vide infra) can mediate the four-fold Si–H bond cleavage of SiH4 to furnish µ2-Si molecular silicides (Ln(H)2MSiM(H)2Ln).6–8 Such a transformation is the bimetallic analogue of silylene formation from two-fold Si–H activation and offers fundamental insight into silicon-atom extrusion from hydrosilanes mediated by transition metals. Solid-state transition metal-silicides are binary materials with MxSiy stoichiometries. These materials find critical applications as intermediates in the direct (Müller–Rochow) process and in the formation of silicon nanowires;9–11 however, their molecular chemistry is much less well developed. Beyond a handful of µ4-Si clusters reported in the literature,12–18 most work has focused on µ2-Si silicides because of their structural simplicity and potential for higher reactivity, due to the low coordination number at silicon. The first µ2-Si silicides, Tp*(CO)2MSi–M(CO)2(PMe3)Tp* (M = Mo, W; Tp* = HB(3,5-dimethyl-1-pyrazolyl)3−), were accessed by treatment of (SIPr)SiBr2 (SIPr = 1,3-bis(2,6-diisopropylphenyl)-4,5-dihydroimidazol-2-ylidene) with two equivalents of Na[Tp*M(CO)2(PMe3)] (M = Mo, W).19,20 Soon after, this laboratory found that treatment of Na(THF)6{[BP3iPr]CoI} ([BP3iPr] = PhB(CH2PiPr2)3−) with SiH4 or PhSiH3 resulted in cleavage of all bonds to silicon with formation of silicides [BP2iPr](RSiH2)(H)2CoSiCo(H)2[BP3iPr] ([BP2iPr] = PhB(CH2PiPr2)2; R = Ph, H) by way of a complex mechanism.6 Further studies found that the base-stabilized SiH2 complex [BP3iPr](H)2CoSiH2(DMAP) (DMAP = 4-dimethylaminopyridine) provides modular and rational syntheses of a range of silicides taking the general form [BP3iPr](H)2CoSiM(H)2Ln (M = Co, Fe).8 In addition, the dinitrogen complex [BP2tBuPz]Co–NN–Co[BP2tBuPz] ([BP2tBuPz] = PhB(CH2PtBu2)2(pyrazolyl)−) mediates the four-fold Si–H activation of SiH4 to yield the µ2-Si silicide [BP2tBuPz](H)2CoSiCo(H)2[BP2tBuPz], which features a MSiM core (Fig. 1B).7 This transformation is possible due to the sterically protected, pseudo-tetrahedral Co(i) centers in [BP2tBuPz]Co–NN–Co[BP2tBuPz], which act as a synthon for the 14e− metal fragment [BP2tBuPz]CoI. In contrast, higher-valent metal centers in different geometries resulted in incomplete Si–H activation of SiH4, as exemplified by the µ2-η3:η3-SiH4 binding mode (Fig. 1A) and intact Si–H bonds of (PR3)2(H)2Ru(SiH4)Ru(H)2(PR3)2 (R = Cy, iPr).21,22 (A–C) Examples of bimetallic, four-fold Si–H bond activations of SiH4, ranging from µ2-η3:η3-SiH4 complexes (A) to full Si–H cleavage in silicide-tetrahydride complexes (B and C). Further developments in the chemistry of molecular silicides require elucidating structure–function relationships by expanding chemical diversity. This contribution describes bimetallic, formally Fe(0) dinitrogen complexes supported by neutral tripodal ligands [SiP3R] ([SiP3R] = PhSi(CH2PR2)3; R = iPr, Ph) which mediate the four-fold Si–H bond activation of SiH4 to furnish molecular diiron silicides. The strong-field nature of the ligands, coupled with their steric bulk, is thought to be critical in suppressing one-electron processes while providing adequate steric protection for the reactive FeSiFe cores. Results and discussion Synthesis and characterization of bimetallic iron dinitrogen complexes Treatment of an emerald-green tetrahydrofuran (THF) solution of [SiP3iPr]FeCl23 with KC8 (1.05 equiv., −35 °C, 1 atm N2) resulted in an immediate color change to an intense red. After stirring at 24 °C for 1 h and subsequent work-up, the Fe(0) dinitrogen complex [SiP3iPr]Fe–NN–Fe[SiP3iPr] (1-iPr) was crystallized in 56% yield by diffusion of hexamethyldisiloxane (HMDSO) into a toluene solution of 1-iPr at −35 °C (Scheme 1). Further investigation revealed that 1-iPr could be conveniently accessed in a single step from κ2-[SiP3iPr]FeCl2,23 under analogous conditions with addition of 2.2 equiv. of KC8, in a yield of 30%. This overall yield is similar to that of the two-step process, which has a 55% yield for the first reduction.23 Similarly, reduction of a THF solution of κ2-[SiP3Ph]FeCl2 (ref. 23) with KC8 (2.2 equiv., −35 °C, 1 atm N2) gave access to the analogous dinitrogen complex [SiP3Ph]Fe–NN–Fe[SiP3Ph] (1-Ph) in 73% yield after crystallization by diffusion of n-pentane into a THF solution of 1-Ph at −35 °C. Complexes 1-R (R = iPr, Ph) represent rare examples of dinuclear Fe(0) complexes containing a bridging N2 ligand, with the only known isoelectronic analogues being [P2PPh]Fe–NN–Fe[P2PPh]24 ([P2PPh] = PhP(C6H4-ortho-PiPr2)2 and (Triphos)Fe–NN–Fe(Triphos)25 (Triphos = MeC(CH2PPh2)3). Synthesis of iron dinitrogen complexes 1-iPr and 1-Ph. The IR spectra of complexes 1-R lack an NN stretching band, which is consistent with bridging rather than terminal dinitrogen ligands. This was corroborated by a single crystal X-ray diffraction (SC-XRD) analysis, as both 1-iPr and 1-Ph feature near linear arrangements of the two iron centers and a µ2-η1:η1-N2 ligand (1-iPr: ∠(Fe–N–N) = 177.4(3)°; 1-Ph: ∠(Fe–N–N) = 176.9(2)°; Fig. 2). In the solid state, both complexes possess short Fe–N bond distances of 1.742(2) Å (1-iPr) and 1.721(2) Å (1-Ph), along with relatively long N–N distances of 1.178(4) Å (1-iPr) and 1.202(3) Å (1-Ph). Together, these data are consistent with significant activation of the N2 ligand, although to a lesser extent than in the Fe(0) ß-diketiminate dinitrogen complexes reported by Holland et al. (Fe–N = 1.741(5)–1.773(7) Å; N–N = 1.215(6)–1.241(7) Å).26,27 The Fe–N and N–N bond distances of complexes 1-R are comparable to those found in (Triphos)Fe–NN–Fe(Triphos) (Fe–N = 1.736(2), 1.725(2) Å; N–N = 1.208(3), 1.207(3) Å; note that two unique molecules are present in the asymmetric unit) and thus 1-R exhibit similar degrees of N2 activation as in the Triphos analogue.25 Further, Raman spectra revealed similar stretching frequencies (1-iPr: νNN = 1491 cm−1; 1-Ph: νNN = 1501 cm−1) to that observed for (Triphos)Fe–NN–Fe(Triphos) (νNN = 1501 cm−1).25 Compared to the isostructural Fe(i) complex [BP3iPr]Fe–NN–Fe[BP3iPr] (Fe–N = 1.811(5)–1.818(5) Å; N–N = 1.138(6) Å),28,29 the Fe(0) complexes 1-R display a slight shortening of the Fe–N bond and lengthening of the N–N bond, consistent with a moderately higher degree of N2 activation due to the lower valency of the metal. Solid-state molecular structures of 1-iPr (A), 1-Ph (B), 2-iPr (C), and 2-Ph (D) with 50% probability ellipsoids. Hydride ligands could not be located in the case of 2-iPr. Carbon-bound hydrogen atoms and positional disorder are omitted, and substituents on phosphorus are drawn as wireframe, for clarity. Although complexes 1-R both contain pseudo-tetrahedral formally Fe(0) centers, they are diamagnetic and display C3 symmetry in solution, as evidenced by observation of a single resonance in the 31P{1H} NMR spectrum of each complex (1-iPr: δP = 75.7; 1-Ph: δP = 64.1). Isoelectronic complexes (Triphos)Fe–NN–Fe(Triphos)25 and [P2PPh]Fe–NN–Fe[P2PPh]24 are also diamagnetic, which in both cases was attributed to antiferromagnetic coupling of two S = 1 Fe(0) centers. To probe the bonding in complexes 1-R further, 57Fe Mössbauer spectra were recorded, which revealed similar isomer shifts (1-iPr: δ = 0.21 mm s−1; 1-Ph: δ = 0.17 mm s−1; Fig. 3) and quadrupole splitting (1-iPr: |ΔEQ| = 0.36 mm s−1; 1-Ph: |ΔEQ| = 0.36 mm s−1). The isomer shifts of complexes 1-R are lower than that of the isostructural complex [BP3iPr]Fe–NN–Fe[BP3iPr] (δ = 0.53 mm s−1),30 which is contrary to the trend generally seen in 57Fe Mössbauer spectroscopy that a higher valency results in a lower isomer shift. However, it is well known that for low-valent iron complexes this trend can be reversed due to an increase in metal–ligand covalency (i.e., back-bonding).31–34 In this case, the high degree of back-bonding is evidenced by short Fe–N (vide supra) and Fe–P bond distances (1-iPr: Fe–P = 2.10(1)–2.224(4) Å; 1-Ph: Fe–P = 2.1555(6)–2.1641(6) Å). Moreover, complexes 1-R have isomer shifts comparable to other Fe(0) complexes that display substantial π back-bonding to the supporting ligands, such as the isoelectronic complex (Triphos)Fe–NN–Fe(Triphos) (δ = −0.09 mm s−1)25 and Fe(CO)5 (δ = −0.08 mm s−1).34 Further, complexes 1-R have comparable isomer shifts to those observed in pseudo-tetrahedral Fe(0) (S = 1) complexes with terminal N2 ligands ((ICy)3FeN2: δ = 0.47 mm s−1; ICy = 1,3-dicyclohexylimidazol-2-ylidene; [PhB(IAd)3FeN2]K(18-crown-6): δ = 0.13 mm s−1; IAd = 3-adamantylimidazol-2-ylidene).35,36 Zero-field 57Fe Mössbauer spectra of 1-iPr (A), 1-Ph (B), 2-iPr (C), and 2-Ph (D). Activation of SiH4 by bimetallic iron dinitrogen complexes With complexes 1-R in hand, their propensity towards SiH4 activation was investigated. When a THF solution of 1-iPr was heated to 60 °C in the presence of SiH4 (3.0 equiv.) for 3 h, a gradual color change from dark red to magenta was observed. After work-up, the symmetrical diiron silicide [SiP3iPr](H)2FeSiFe(H)2[SiP3iPr] (2-iPr) was isolated in moderate yield (36%) as a magenta solid (Scheme 2). Similarly, heating a toluene solution of 1-Ph to 60 °C for 16 h in the presence of SiH4 (1.1 equiv.) produced [SiP3Ph](H)2FeSiFe(H)2[SiP3Ph] (2-Ph) in a significantly higher yield (82%) after crystallization induced by diffusion of n-pentane into a THF solution of 2-Ph at −35 °C. Synthesis of complexes 2-iPr and 2-Ph. In the solid state, complex 2-iPr displays positional disorder of the µ2-Si atom over a special position, likely due to crystal packing forces as bending of the MSiM core is expected to have a minimal energy penalty.19 When considering the major disordered component, complexes 2-iPr and 2-Ph (Fig. 2) feature Fe–Si–Fe angles of 165.4(2)° and 180.0°, respectively, and extremely short Fe–Si bond distances of 2.1185(6) Å and 2.0937(3) Å, respectively. To the best of our knowledge, these Fe–Si distances represent the shortest reported to-date (cf., Cp*(MePiPr2)Fe(µ2-H)Si(H)Trip: Fe–Si = 2.129(1) Å),37 and are significantly shorter than the sum of the double-bond covalent radii of iron and silicon (2.16 Å).38 Such short iron-silicon distances are consistent with multiple bonding within the FeSiFe core. The hydrides in the solid-state structure of complex 2-Ph (but not in that of 2-iPr) were located in the Fourier difference map and refined isotropically.39 The four hydrides exhibit positional disorder over six positions; consistently, computational modeling suggests that two conformations of similar energy are thermally accessible by rotation about the Fe–Si–Fe axis (vide infra). In one conformer, the hydrides are staggered about the µ2-Si atom in a manner analogous to that observed for previously reported cobalt silicide-tetrahydride complexes (Fig. 4, middle).6–8 In contrast, the other conformer displays hydrides that form a square plane (Σ∠(H–Si–H) = 360°) about the µ2-Si atom (Fig. 4, right). In both conformers, considering the resulting hydride positions, each iron center exhibits a distorted octahedral geometry. Critically, the non-tetrahedral array of hydrides about the µ2-Si atom is not consistent with a µ2-η3:η3-SiH4 binding mode, as seen in (PR3)2(H)2Ru(SiH4)Ru(H)2(PR3)2 (R = Cy, iPr).21,22 The Si–H bond lengths in 2-Ph (1.60(4)–1.77(3) Å) are significantly longer than those observed for the Fe–H bonds (1.52(4)–1.59(4) Å). Although hydride positions determined by SC-XRD can be dubious, shorter Fe–H bond lengths, compared to that of Si–H, is consistent with more Fe–H bonding character, given that iron and silicon have the same single-bond covalent radii (1.16 Å).38 The computed structure 2-iPrcalc (B3LYP/6-311G(d)) closely resembles 2-iPr and 2-Ph in terms of bond distances (Δd(Fe–Si) = <0.035 Å); however, the computed Fe–Si–Fe bond angle (176.2°) differs somewhat from that observed in 2-iPr (165.4(2)°), although it closely resembles that in 2-Ph (180.0°; note that the µ2-Si atom is located on a special position). This discrepancy is thought to arise from crystal packing forces (vide supra). The computed structure 2-iPrcalc exhibits a square plane of hydrides about the µ2-Si atom (Σ∠(H–Si–H) = 360°) and H–Si distances (1.84 Å, 1.89 Å) that are comparable to that of 2-Ph (considering that SC-XRD underestimates metal-hydride distances); however, a conformation search involving rotation about the Fe–Si–Fe axis yields an additional conformer of similar energy which instead exhibits staggered hydrides (vide supra). From the square planar conformer, rotation clockwise yields one enantiomer of the staggered hydride conformation, while rotation counterclockwise yields the other enantiomer. Preliminary results suggest that the difference in energy between these conformers in the gas-phase is minimal (∼3 kcal mol−1) and that the barrier to rotation is easily accessible at room temperature (∼8 kcal mol−1). Given such a small calculated difference in energy between these gas-phase conformers, both would seem to be energetically accessible in solution. Additionally, the low barrier to interconversion is further evidence for weak Si⋯H interactions. Comparison of the solid-state molecular structures of (PiPr3)2(H)2Ru(SiH4)Ru(H)2(PiPr3)2 (left), [BP3iPr](H)2CoSiCo(H)2[BP3iPr] (middle), and [SiP3Ph](H)2FeSiFe(H)2[SiP3Ph] (right) showing the different arrangement of hydrides about the µ2-Si atom (only metal-bound atoms are shown). Complex 2-Ph is shown in the conformation with hydrides in a square plane about the µ2-Si atom; the other conformation of 2-Ph with similar energy is analogous to that of [BP3iPr](H)2CoSiCo(H)2[BP3iPr] (middle). In solution, complexes 2-R display pseudo-C3 symmetry, as evidenced by 31P{1H} NMR spectroscopy which reveals a single resonance in each case (2-iPr: δP = 69.7; 2-Ph: δP = 56.9). In line with this, the 1H NMR spectra of complexes 2-R each exhibit only one hydride resonance (2-iPr: δH = −14.11; 2-Ph: δH = −11.77), which appear as a quartet due to coupling to three 31P nuclei that are equivalent on the NMR time scale (2-iPr: 2JHP = 17.3 Hz; 2-Ph: 2JHP = 16.3 Hz). These experimental findings suggest stereochemical nonrigidity for complexes 2-R in solution. Importantly, the hydride 1H NMR resonance of complexes 2-R integrate to four, as expected for a Ln(H)2MSiM(H)2Ln silicide. In contrast, the hydride 1H NMR resonance of a SiH62− ligand, such as that of the closely related complex [BP3Ph]Ru(SiH6)Ru[BP3Ph] ([BP3Ph] = PhB(CH2PPh2)3−),40 would integrate to six. Satellites due to coupling between hydrides and the µ2-Si atom could not be observed in 1H or 1H{31P} NMR experiments, nor were cross-peaks observed in 1H–29Si HMBC experiments. These observations are consistent with minimal Si⋯H interactions. The 29Si{1H} NMR spectra of complexes 2-R reveal remarkably downfield 29Si resonances for the µ2-Si atom of 405 ppm (2-iPr) and 403 ppm (2-Ph), in line with significant multiple bonding in the FeSiFe core. These chemical shift values are at lower-field than those of the closely related cobalt and cobalt-iron silicide complexes [BP3iPr](H)2CoSiCo(H)2[BP3iPr] (δSi = 333) and [BP3iPr](H)2CoSiFe(H)2[SiP3iPr] (δSi = 358),8 but lie between those reported for the metallasilylidyne complexes Tp*(CO)2MSi–M(CO)2(PMe3)Tp* (M = Mo: δSi = 439; M = W: δSi = 398).19,20 The observed downfield shift of the 29Si resonance upon substituting cobalt for iron may be due to a larger paramagnetic shielding component for iron, or an increase in magnetic anisotropy.41 To further interrogate the bonding within complexes 2-R, 57Fe Mössbauer spectra were collected, revealing similar isomer shifts (2-iPr: δ = 0.13 mm s−1; 2-Ph: δ = 0.10 mm s−1; Fig. 3) and quadrupole splitting (2-iPr: |ΔEQ| = 0.66 mm s−1; 2-Ph: |ΔEQ| = 0.71 mm s−1), consistent with similar bonding environments. An isomer shift near zero for silicides 2-R is consistent with two pseudo-octahedral formally Fe(ii) centers.42 However, these Mössbauer data are also consistent with an Ln(H)2FeSiFe(H)2Ln bonding motif with minimal Si⋯H interactions. Although iron is formally tetravalent in the FeSiFe core of 2-R, previous work has shown that tetravalent iron carbenes can have isomer shifts that are similar to those of Fe(ii) complexes.43 Such disagreements between isomer shift and valency should be expected in highly covalent environments where the oxidation state formalism does not accurately describe bonding.44,45 Given the novelty of the square planar hydride conformation of 2-R, bonding of that conformer was thoroughly investigated. Analysis of Wiberg bond indices46 (WBIs) of 2-iPrcalc offers further insight into such bonding; the Fe–H WBIs (0.72, 0.69) are much higher than that of the Si–H WBIs (0.14, 0.18), indicating a more significant Fe–H bonding interaction. Further, the Fe–Si WBI (1.24) is significantly higher than that of the WBIs of single bonds in 2-iPrcalc (e.g., Fe–P WBIs = 0.88–0.95), consistent with multiple bonding. The WBIs calculated for 2-iPrcalc are in stark contrast to that of (PiPr3)2(H)2Ru(SiH4)Ru(H)2(PiPr3)2 (phosphines truncated to PH3), which reports WBIs of 0.42, 0.21 and 0.58 for Ru–Si, Ru–H (bridging) and Si–H bonds, respectively.22 Taken together, the WBI analysis strongly supports an FeSiFe core with minimal Si⋯H interactions, analogous to that observed for closely related cobalt and cobalt-iron silicides.6–8 Although WBI calculations are very sensitive to the atomic basis set, the WBIs reported here are with the usual standard basis set for WBIs. Additional information was provided by an atoms in molecules47 (AIM) analysis of 2-iPrcalc. No bond critical points could be found between the µ2-Si atom and the flanking hydrides; however, four bond critical points were found between the µ2-Si atom and hydrogen atoms within the ligand isopropyl groups, possibly indicating some stabilization through weak interactions with the ligand substituents. In contrast, a bond critical point was found for all four Fe–H bonds (near 1 Å in each case). Further data from the AIM analysis indicate minimal charge build-up at iron (q(Fe) = 0.18), consistent with a highly covalent bonding situation which is poorly described by the oxidation state formalism (vide supra). In contrast, the µ2-Si atom has a larger build-up of positive charge (q(Si) = 0.70), in line with other closely related silicides.8 AIM delocalization indices (DI) are an alternative bond order index and support bond orders from the WBIs of Fe–Si > Fe–H > Si–H, but with increased Si–H and decreased Fe–Si bond orders. All of the key geometric and bonding features of 2-iPrcalc, the square planar ‘SiH4’ geometry, the high Fe–Si WBI, and the non-zero Si–H WBI, are the result of an sp hybridized silicon with two bent Fe–Si σ-bonds (the bond critical point deviates from the linear bond line) and two multi-center, 4-electron interactions, where each of these involves two electrons on one Fe–H bond, an empty silicon 3p orbital (not used in the Fe–Si σ-bonding), and two electrons on a second Fe–H bond (trans to the first one; Fig. S44). It is these two multi-center, 4-electron interactions that cause the two P3H2Fe units that bond to the silicon to distort from an Oh coordination at iron in an idealized P3H2FeSiFeH2P3 dimer by tipping the P3H2Fe units to move the hydrogens toward the silicon. The HOMO (Fig. S38) is a key component of this multi-center FeH–Si–HFe bonding. A Natural Bond Orbital (NBO) calculation (see SI) supports this analysis as it yields four 2-center, 2-electron Fe–H NBOs, two bent 2-center, 2-electron Fe–Si σ NBOs, and two low occupancy π NBOs that correspond to the silicon 3p orbitals perpendicular to each other and to the 3p involved in the Fe–Si σ-bonds (Fig. S38). The electron density in each of these partially occupied silicon 3p NBOs arises from pairs of ‘trans’ Fe–H NBOs using the silicon 3p orbital to form the out-of-phase component of the multi-center, 4-electron FeH–Si–HFe interactions (HOMO in Fig. S38). Additionally, the conformer of 2-iPrcalc with staggered hydrides (vide supra) was briefly investigated and found to similarly exhibit delocalization of electron density from four strong Fe–H bonds into the silicon 3p π orbitals. In summary, although complexes 2-R exhibit extensive Si–H bond cleavage, the electron-deficient silicon 3p orbitals in the FeSiFe cores are stabilized by donation of electron density from the flanking Fe–H bonds. Conclusions Iron dinitrogen complexes 1-R are rare instances of dinitrogen bridging formally Fe(0) centers. The soft, strong-field nature of the [SiP3R] ligands, coupled with their significant steric bulk, is thought to be critical in imparting stability to complexes 1-R. The electron-rich, pseudo-tetrahedral Fe(0) centers in 1-R were further capable of a bimetallic, four-fold SiH4 bond activation to yield diiron silicides 2-R which possess FeSiFe cores. The work described herein offers fundamental insight into the transition metal mediated extrusion of silicon atoms from organosilanes and suggests that four-fold Si–H activation of SiH4 may be a generalizable route to molecular silicides. Further, iron silicides offered the possibility to interrogate fundamental aspects of bonding within molecular silicides by means of 57Fe Mössbauer spectroscopy, revealing that silicides of this type exhibit highly covalent bonding. Future work will focus on understanding how structural changes (e.g., metal identity, ligand donor ability, etc.) affect the bonding and reactivity of molecular silicides. Author contributions The syntheses and characterization of new compounds were performed by R. T. B.; R. C. H. initially synthesized compounds 1-iPr and 2-iPr. R. T. B. conducted the crystallographic analysis of compounds 1-Ph and 2-Ph; R. C. H conducted the crystallographic analysis of compounds 1-iPr and 2-iPr; S. J. T. assisted in the crystallographic refinement of compounds 1-Ph, 2-iPr, and 2-Ph. Mössbauer experiments were conducted by R. C. H. Quantum-chemical investigations were performed by S. H. with supervision from M. B. H. The project was supervised by T. D. T. The manuscript was written by R. T. B., T. D. T., and M. B. H. All authors have given approval to the final version of the manuscript. Conflicts of interest There are no conflicts to declare. Supplementary Material