Skip to main content
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2021 May 30.
Published in final edited form as: Eur J Inorg Chem. 2020 Sep 4;2020(36):3460–3465. doi: 10.1002/ejic.202000736

Reactions of [Fe6C(CO)14(S)]2−: Cluster Growth, Redox, Sulfiding

Liang Liu [a], Toby J Woods [a], Thomas B Rauchfuss [a]
PMCID: PMC8054992  NIHMSID: NIHMS1664326  PMID: 33883972

Abstract

Redox reactions, substitutions, and metalations are reported for the iron carbido sulfide [Fe6C(CO)14(S)]2− ([1]2−). Dianion [1]2− oxidized to [Fe6C(CO)16(S)]0 ([2]0) upon treatment with of [Fe(C5H5)2]BF4 or HBF4 (H2 formation) under an atmosphere of CO. Reaction of [2]0 with tBuNC gave [Fe6C(S)(CO)13(tBuNC)5], consisting of Fe5C(CO)13 and [Fe(tBuNC)5]2+ subunits linked by a μ3-S2−. The Fe7CS cluster [Fe7C(CO)17(S)]2− formed upon treatment of (Ph4P)2[1] with Fe(benzylideneacetone)(CO)3. The Fe7 species is an edge-fused cluster with [Fe6C(CO)10(μ-CO)4] and Fe(CO)3 subunits joined by μ3-S and two Fe–Fe bonds. The analogous reaction using Mo(CO)4(norbornadiene) gave [MoFe6C(CO)18(S)]2−. In this cluster, the Mo center is located in the octahedral subunit. Treatment of [1]2− with SO2 afforded [Fe6C(S)(SO2)(CO)13]2−. This cluster features an Fe6C core decorated with μ3-S and μ2-SO2 ligands. These experiments were undertaken in an effort to connect organometallic clusters to FeMoco.

Keywords: Carbides, Cluster compounds, Sulfides, Nitrogenases, Iron

Introduction

The cluster [Fe6C(CO)16]2− has been well known for several decades. It has received only intermittent attention probably because it resists substitution reactions, which precludes catalytic applications. It does exhibit rich redox behavior. The dianion can be oxidized to a somewhat fragile neutral derivative [Fe6C(CO)18].[1] Reduction gives [Fe6C(CO)15]4−. [2]

We have started to investigate the possibility that [Fe6C(CO)16]2− could be developed as a precursor to replicas for FeMoco and related clusters that are responsible for biological nitrogen fixation.[3] Arguably the most distinctive cofactor found in nitrogenase, hypervalent carbide ligands are absent from all other synthetic iron-sulfide clusters.[4] As summarized in a recent review, the modeling of FeMoco remains of intense interest.[5] In place of carbide, some models employ organic carbon substituents.[6]

With reference to FeMoco, we recently reported the synthesis of [Fe6C(CO)14(S)]2− ([1]2−), which is the first Fe-C-S cluster. The presence of the sulfide achieves several incremental improvements in the modeling, including raising the average oxidation state of the Fe6C core. The deficiencies in [Fe6C(CO)14(S)]2− ([1]2−) are however numerous. The overarching issue is the absence of CO ligands in the biological cluster. In this report we describe some attempts to minimize the structural differences between FeMoco and [1]2−. These efforts followed two main themes, (i) increasing the S/Fe ratio and (ii) increasing the metal nuclearity. Additionally, relevant to rearrangements in FeMoco and related clusters,[7] we describe redox-induced rearrangements of [1]2−.

Results and Discussion

Redox Transformations of [Fe6C(CO)14(S)]2−

This work began by examining redox reactions of [Fe6C(CO)14(S)]2− ([12−]). In CH2Cl2 solution, (Ph4P)2[1] quantitatively transformed into [Fe6C(CO)16(S)]0 ([2]0) upon treatment with two equiv. of [Cp2Fe]BF4 under an atmosphere of CO (Scheme 1) as confirmed by FT-IR spectrum. After purification by column chromatography, [2]0 was isolated in 83 % yield. Two equiv. of CO are required by the stoichiometry of the reaction, one in effect compensating for the change in cluster charge and the other allowing the cluster to change from closo (86e) to nido (88e).[8] The oxidative carbonylation of [1]2− to give [2]0 can also be effected by acids. Addition of excess degassed HBF4·Et2O under an atmosphere of CO leads cleanly to [2]0 with an 81 % isolated yield. The coproduct H2 was detected by gas chromatography (Figure S6).

Scheme 1.

Scheme 1.

Reactions described in this work.

To elucidate aspects of the conversion of [1]2− into [2]0, we examined the oxidation in the presence of tBuNC instead of CO. The results were however poorly reproducible. The apparent complication arises because both [1]2− and tBuNC react with [Cp2Fe]BF4. To circumvent these problems, the oxidized cluster [2]0 was first generated and then treated with tBuNC. Iron carbonyl clusters are characteristically susceptible to ligand substitution.[9] In the case of Fe-S-CO clusters such reactions often proceed by associative mechanisms.[10] The reaction with 5 equiv. of the isocyanide in CH2Cl2 solution produced electron-rich products. Thus, the FT-IR spectra revealed shifts in νCO from 2097–2020 cm−1 to 1989–1962 cm−1. The FT-IR spectrum of the crude product exhibits intense bands in the region (2209–2148 cm−1), which is typical for νNC. Thin layer chromatography (TLC) showed several deeply colored products. After separation by column chromatography in the glovebox, the first and second bands were obtained as oily residue and fine powders, respectively. The second band was amenable to further purification by column chromatography (CH2Cl2/pentane = 1:1) to afford [Fe6C(S)(CO)13(tBuNC)5] ([3]0) as a dark brown solid in 20 % yield.

The structure of the [3]0 was determined by X-ray crystallography (Figure 1). Overall, the product consists of Fe5C(CO)13 and Fe(tBuNC)5 subunits linked by a μ3-sulfide. The octahedral SFe(tBuNC)5 center is assigned as FeII because it structurally resembles simpler derivatives such as [Fe(SR)(tBuNC)5]+.[11] The complex can be viewed either as the adduct of [FeII(S−II)-(tBuNC)5]0 and [Fe5C(CO)13]0 or as the adduct of the Lewis acid [Fe(tBuNC)5]2+ and the Lewis base [Fe5C(CO)132-S)]2−.

Figure 1.

Figure 1.

Structure of [3]0 with thermal ellipsoids drawn at the 50 % probability level. Selected distances in ranked order [Å]. Fe–C1: Fe5–C1, 1.857(3); Fe3–C1, 1.866(3); Fe4–C1, 1.869(3); Fe2–C1, 1.877(3); Fe1–C1, 1.975(3). Fe–S1: Fe5–S1, 2.2145(8); Fe2–S1, 2.2148(8); Fe6–S1, 2.3155(8). Fe6–C: Fe6–C35, 1.878(3); Fe6–C30, 1.882(4); Fe6–C15, 1.883(4); Fe6–C25, 1.886(3); Fe6–C20, 1.888(3).

The 1H NMR spectrum of [3]0 is consistent with the presence of two types of tBuNC ligands in a ratio of 4:1. The 13C NMR spectrum also shows two sets of CH3 (30.6, 31.0 ppm), tert-C (58.7, 58.9 ppm) and NCFe signals (152.9, 153.5 ppm). Three CO signals (217.8, 220.9, 221.3 ppm) can be assigned to 2Fe(CO)3, 2Fe(CO)2 at basal positions and Fe(CO)3 at axial position. The μ5-carbide signal appears at 467.7 ppm, which is 11–16 ppm upfield of μ6-carbide in dianionic clusters (vide infra). Overall, the NMR data indicate that the structure observed crystallographically is retained in solution.

Redox properties of [3]0 were investigated by cyclic voltammetry experiments on CH2Cl2 solutions. It exhibits a quasi-reversible redox couple at E1/2 = −0.39 V and four irreversible events at Ep,a = 0.26, 0.57 V and Ep,c = −1.97, −2.25 V vs. Fc+/0.

Throughout this work, we have been attentive to opportunities to introduce molecular molybdenum sulfides into the Fe6C cluster. Mo-Fe5-C clusters are precedented,[12] although sulfide derivatives had not been reported prior to this work. Since the oxidized cluster [2]0 reacts with tBuNC to give adducts, nucleophilic Mo-S reagents were also tested in place of tBuNC. When treated with (Et4N)2MoS4, [2]0 underwent a reduction to [2].

MFe6CS Clusters (M = Fe, Mo)

A Fe7CS cluster was first detected while investigating the reaction of [Fe6C(CO)16(S)]0 ([2]0) with Na2Fe(CO)4. This reaction was modeled after the synthesis of [Fe6C(CO)16]2− from [Fe5C(CO)15]0.[13] Treatment of [2]0 with one equiv. of Na2Fe(CO)4 however mainly produced the one-electron reduced cluster [2]. This conversion is consistent with the facility of cluster redox reactions and the reducing tendency of Na2Fe(CO)4.[14] In addition to intense signals for [2], the negative-ion ESI-MS spectrum of the crude product showed weak signals (m/z = 991.7 and daughter peaks) corresponding to [Fe7C(CO)17(S)]. These clues induced us to devise a more rational synthesis of the Fe7C cluster.

An efficient synthesis of the Fe7 cluster involved treatment of (Ph4P)2[1] with one equiv. of Fe(bda)(CO)3 (bda = benzylideneacetone). Initial experiments were conducted in MeCN solution, but the yields were low. When conducted in THF solution, the reaction was slower but cleaner. Over the course of 48 h at room temperature, [1]2− converted to [Fe7C(CO)17(S)]2− ([4]2−). An efficient conversion was indicated by TLC and IR analyses. New νCO bands appeared in the FT-IR spectrum. Compared to [1]2−CO 2020(vw), 1958(vs), 1949(vs), 1925(vw) cm−1), [4]2− exhibits νCO bands shifted by about 15 cm−1 toward higher energies (νCO 2037(vw), 1991(m), 1964(vs) cm−1). This shift can be rationalized because the negative charge is delocalized over more Fe(CO)n centers. The presence of the extra band at 1991(m) cm−1 is consistent with [4]2− having a less symmetrical structure than [1]2−. A slight excess (1.2 equiv.) of Fe(bda)(CO)3 is better for a full conversion of [1]2− to [4]2− and did not further react with [4]2−, e.g. to give Fe8 clusters. Unreacted Fe(bda)(CO)3 was easily removed by washing the crude product with Et2O. (Ph4P)2[4] was isolated in up to 85 % yield after workup. A black solid, (Ph4P)2[4] is relatively stable in air, surviving for 1 day at room temperature. Its solutions however decomposed to a tan-orange suspension quickly when exposed to air.

Crystals of (Ph4P)2[4], obtained from our initial reactions, were contaminated with ca. 6 % of (Ph4P)2[1]. The crystal data of (Ph4P)2[4] were however of sufficient quality to confirm its structure (Figure 2). Like [2]0, [4]2− is an edge-fused cluster whereby [Fe6C(CO)10(μ-CO)4] and Fe(CO)3 are joined by a μ3-S and a pair of Fe–Fe bonds. Described differently, [4]2− consists of an edge-fusion of an octahedral Fe6C and tetrahedral Fe3S clusters. In terms of core skeleton, [4]2− arises by the addition of an [Fe(CO)3]2− vertex to the square face of [2]0.

Figure 2.

Figure 2.

Structure of the dianion in (Ph4P)2[4] drawn at the 50 % probability level. Selected distances in ranked order [Å]. Fe–C1: Fe6–C1, 1.866(6); Fe1–C1, 1.879(6); Fe2–C1, 1.896(6); Fe3–C1, 1.909(6); Fe4–C1, 1.909(6); Fe5–C1, 1.907(6). Fe–S1: Fe7–S1, 2.1612(19); Fe6–S1, 2.174(2); Fe3–S1, 2.1998(19).

The 13C NMR spectrum of [4]2− exhibits three CO signals, at 227.2, at 229.0, and 229.8 ppm, the latter being intense. The first two signals are assigned to the SFe(CO)3 and the S[Fe(CO)]2 centers. The intense band is assigned to ten remaining CO centers, which are rapidly scrambling. The 13C NMR signal for the carbide in [4]2− is shifted only slightly (4 ppm) toward downfield compared to [1]2−, which is consistent with the tendency of νCO bands shift in FT-IR spectrum.

The redox behavior of (Ph4P)2[4] was examined in CH2Cl2 solution. The cyclic voltammogram (CV) displays three redox couples at highly negative potentials: E1/2 = −1.46, −1.99, and −2.45 V vs. Fc+/0.

Similar to its reaction with Fe(bda)(CO)3, [1]2− reacts with neutral Mo-carbonyl precursors to afford an MoFe6 cluster. Initial tests examined Mo(CO)3(mesitylene) and the negative-ion ESI-MS displayed a signal at m/z = 1321, corresponding to {Ph4P[MoFe6C(CO)18(S)]}. This result suggested that a “Mo(CO)4” precursor would be better for stoichiometry. Indeed, treatment of [1]2− with 1.2 equiv. of Mo(CO)4(norbornadiene) in CH2Cl2 gave [MoFe6C(CO)18(S)]2− ([5]2−). The reaction proceeds gradually in 48 h, just like the Fe(CO)3(bda) reaction. Reaction progress was monitored by IR spectroscopy. The FT-IR spectrum of [5]2−CO 2000(m), 1968(vs) cm−1, similar to that of [5]2−) is quite distinct from that of the precursor. Although anaerobic TLC and IR indicated a clean conversion, the 13C NMR spectrum of the crude product showed two carbide signals at 479.1 and 477.9 ppm, the latter assigned to unreacted [1]2−. The two dianionic clusters could not be separated by column chromatography. The purification of [5]2− was realized based on the stability differences. The anion [5]2− is more air-stable than [1]2−, especially in solid state. In practice, the purification entailed stirring a slurry of (Ph4P)2[5] in CH2Cl2/Et2O (1:5 to 1:10) in air at room temperature overnight. After removal of the yellow solution, the unreacted black residue of (Ph4P)2[5] was analyzed. The 13C NMR spectrum of this fresh sample revealed only one carbide signal at 479.2 ppm.[21]

It appears, however, that in solution, [5]2− degrades to the parent cluster [Fe6C(CO)16]2−. Thus, single crystals of (Ph4P)2[5], which were analyzed by X-ray crystallography, cocrystallized with ≈ 10 % of (Ph4P)2[Fe6C(CO)16] (Figure 3).

Figure 3.

Figure 3.

Structure of the dianion in (Ph4P)2[5] drawn at the 50 % probability level. Selected distances in ranked order [Å]. M–C1: Fe5–C1, 1.873(4); Fe1–C1, 1.890(4); Fe2–C1, 1.909(4); Fe4–C1, 1.919(4); Fe3–C1, 1.923(4); Mo1–C1, 2.072(4). M–S1: Fe6–S1, 2.1546(11); Fe1–S1, 2.2021(11); Mo1–S1, 2.3924(10).

As for [4]2−, the structure of [5]2− can be described as the edge-fusion of an M6C octahedron and a M3S tetrahedron. The surprising feature of [5]2− is the position of the Mo atom, which is located in the octahedral subunit, as indicated by the refinement (R1 = 0.0497). Since cluster [5]2− contains one more CO ligand than does [4]2−, the location of this “extra” CO ligand confirms the presence of the Mo atom within the octahedron. As shown in Figure 3, the extra CO is on the M6 core structure rather than the external M(CO)3 moiety. The Mo atom is bonded with five non-metal ligands (carbide, sulfide, two terminal CO and one bridging CO) whereas all the Fe atoms are attached with four non-metal ligands. Thus, the incorporation of the Mo(CO)x vertex by [1]2− entails the breaking of at least three Fe–Fe, one Fe–C (carbide), and one Fe–S bonds, while five Mo–Fe, one Mo–C, and one Mo–S bonds are formed. This drastic rearrangement suggests that the reaction produced the thermodynamic product, despite the mild conditions. The pathway from [1]2− to [5]2− is proposed to involve successive opening and closing process of the core cluster (see Figure S20 in Supporting Information for a proposed mechanism).

Fe6CS(SO2) Clusters

Qualitative experiments revealed that [1]2− reacts with SO2 at room temperature as indicated by a color change from dark brown to greenish dark brown. Initial preparations involved addition of 100–150 equiv. of SO2 to a MeCN solution of (Ph4Ph)2[1]. This approach cogenerated significant amounts of by-products as indicated by FT-IR and TLC analyses. Yields improved when the reaction was conducted at 0 °C in THF. In this way, a clean conversion required only 7–10 equiv. of the SO2, which was introduced as a gas by syringe. This modified method exploits the high solubility of SO2 in cold THF. A simple flash column eluting with CH2Cl2 followed by Et2O/MeCN (2:1) was sufficient for purification, affording (Ph4P)2[Fe6C(S)(SO2)(CO)13] {(Ph4P)2[6]} as a greenish black solid in > 65 % isolated yield. In MeCN solution, (Ph4P)2[6] is only stable for a few hours. The νCO bands of [6]2−CO 2029(vw), 1975(sh), 1968(vs), 1941(vw) cm−1) shift about 19 cm−1 towards high energies compared to those of the precursor [1]2−.

The salt (Ph4P)2[6] was characterized by elemental analysis and 13C NMR spectroscopy. Single crystals were obtained for the mixed cation salt (Ph4P)Na[6], solvated by both ether and MeCN. X-ray crystallography confirmed the expected structure consisting of an Fe6C core decorated with S and SO2 (Figure 4). The structures of [1]2− and [6]2− are related by the interchange of μ-CO and μ-SO2 ligands, which occupy an Fe···Fe edge that is opposite to the μ3-S-capped face. The remaining CO ligands are terminal. The accumulation of negative charge on the SO2 centers is reflected by the coordination to sodium cations, resulting in an eight-membered [SO2Na]2 ring.

Figure 4.

Figure 4.

Structure of [(Et2O)2Na][(Et2O)(MeCN)2Na](Ph4Ph)2[6]2 drawn at the 50 % probability level. Selected distances in ranked order [Å]. Fe–C1: Fe6–C1, 1.877(2); Fe1–C1, 1.881(3); Fe2–C1, 1.884(3); Fe5–C1, 1.891(2); Fe4–C1, 1.907(3); Fe3–C1, 1.911(3).

The basic character of the μ-SO2 ligands in [6]2− was exploited by O-methylation. Thus, addition of one equiv. of MeOTf to a solution of (Ph4P)2[6] in dimethoxyethane (DME) resulted in a rapid reaction as indicated by precipitation of white solids (Ph4POTf). In the FT-IR spectrum, the νCO bands shifted by ≈ 28 cm−1 (Figure S27). The stoichiometry of the methylated product [5Me] was confirmed by the negative-ion ESI-MS spectrum, which revealed an intense signal at m/z 822.8 (Figure S28). Values for νCO for [6Me] and [Fe6C(CO)15(SO2Me)] are similar, respectively 1997 vs. 2002 cm−1. Solutions of Ph4P[6Me], however, proved to be unstable at room temperature, and its reactions were not pursued.

Conclusions

The goal of this project was to evaluate [1]2− as a precursor to clusters with stoichiometries approximating Fe7MoCS9, as seen in FeMoco. Some progress is reported in terms of (i) increasing the metal nuclearity and (ii) increasing the S/M ratio. Both approaches however suffer from limitations. Experiments are also described that probe the redox behavior of [1]2− since evidence exists that Fe7MoCS9 and the related vanadium cluster undergo rearrangements that affect the core structure.[7]

Increasing the M/C Ratio.

One goal was to prepare Fe-C-S clusters with nuclearities > Fe6. At the outset, it must be noted that [Fe6C(CO)16]2− does not react with Fe(CO)x sources. Instead, metalation is only observed with electrophilic reagents such as [Au(PPh3)]+.[2] Thus, the easy metalation of [Fe6C(CO)14S]2− demonstrates the advantages of the sulfide ligand in attracting a seventh metal. Both Fe(CO)3 and Mo(CO)4 vertices could be introduced. The dynamic character of the carbide cluster is illustrated by the easy incorporation of the Mo within the carbide-centered octahedron.

Increasing the S/M Ratio.

This subproject targeted clusters with the stoichiometry Fe6C(S)x, where x > 1. Since no reactions were detected between [1]2− and traditional S0 sources, we turned to SO2 as a potential source of sulfido ligands. We were mindful that the μ-SO2 ligands in [Fe6C(CO)15(SO2)]2− and [Ru6C(CO)15(SO2)]2−[15] can be deoxygenated.[15b,16] The deoxygenation strategy involves O-methylation of the μ-SO2 ligand.[17] As indicated by the nucleophilic character of the μ-SO2 ligands, the addition of SO2 across a Fe–Fe bond, which occurs concomitantly with loss of CO, represent a 2e oxidation of the cluster. We thus aspired to demonstrate the sequence [Fe6C(CO)14S]2− → [Fe6C(CO)13S(SO2)]2− → [Fe6C(CO)13S-(SO2Me)] → [Fe6C(CO)12(S)2]2−. Such a sequence would increase both the sulfur content and the average oxidation state of the cluster. The addition of the SO2 to [1]2− proceeded well. Unfortunately, the O-alkylation step yielded an unstable product.

Redox.

The redox behavior of [Fe6C(CO)14S]2− tested the resilience of these low-spin clusters as electron reservoirs. It is known that the 2e oxidation of [Fe6C(CO)16]2− causes degradation to [Fe5C(CO)15]0, a “decapitation” process.[18] Related decapitation reactions occur for [Fe6C(CO)14S]2− except that the ejected Fe(CO)3 vertex remains tethered to the Fe5C core by the μ3-S ligand. This pattern was seen for both [Fe6C(CO)16S]0 and [Fe6C(S)(CO)13(tBuNC)5]0. The oxidation of [Fe6C(CO)14S]2− can be effected by protonation with strong acids. H2 evolution and clean conversion to [Fe6C(CO)16S]0 were established. The catalytic properties of carbide clusters for hydrogen evolution has been reported.[19]

Prospectus.

The convertibility of [Fe6C(CO)14S]2− into clusters that more closely resemble FeMoco remains an intriguing but challenging opportunity.[5] One potential goal for future research could be the synthesis of Fe–C clusters with fewer CO ligands. The cluster [Co6C(S)2(CO)12]0 provides precedent not only for the increased degree of sulfiding but also for a trigonal prismatic M6C core,[20] as observed in FeMoco. The oxidation of [Fe6C(CO)14S]2− appears however to consistently lead to loss of a vertex, despite the presence of the interstitial carbide atom. It should be noted that low-spin metal centers comprise [Fe6C(CO)14S]2− and its derivatives, but high spin metal centers comprise FeMoco.

Experimental Section

Manipulations were performed with standard Schlenk techniques or in a N2 atmosphere glovebox. SO2 (gas in steel cylinder), tBuNC and MeOTf were purchased from Sigma-Aldrich and used as received. Solvents were degassed and dried by an MBraun SPS system. Dimethoxyethane was distilled from sodium. Fourier Transform infrared (FT-IR) spectra were recorded on a Perkin-Elmer Spectrum 100 FT-IR spectrometer. 1H and 13C NMR spectra were recorded on a Bruker AVANCE III spectrometer (FT; 500 MHz for 1H, 126 MHz for 13C). Chemical shifts (δ/ppm) were referenced to signals of deuterated solvents (CD2Cl2: 5.32 for 1H, 54.0 for 13C; CD3CN: 1.94 for 1H, 118.26 for 13C). ESI-MS data were acquired on a Waters Micromass Quattro II spectrometer. Elemental analysis was performed at the School of Chemical Sciences Microanalysis Laboratory utilizing a Model CE 440 CHN Analyzer or Robertson Microlit Laboratories, Inc. (Ledgewood, NJ, US). Cyclic voltammetry (CV) experiments were carried out under an atmosphere of N2 (in the glovebox) or CO on a CHI 630D electrochemical analyzer from CH Instruments (Austin, TX, US). In the experiments under CO, an electrochemical cell was sealed with a rubber septum and the solution was sparged with CO for 2 min. The working electrode, counter electrode and pseudo-reference electrode are glassy carbon, Pt wire, and Ag wire [or “no leak” Ag/AgCl (3 M KCl)], respectively. All potentials were referenced to the internal Fc+/0. H2 was detected by gas chromatography on a column packed with 5 Å molecular sieves (carrier gas: Ar) using Agilent 7820A instrument equipped with a thermal conductivity detector (TCD). Unless otherwise noted, column chromatography was performed using silica gel (60 Å, 230–400 mesh) in the glovebox and reactions were conducted at room temperature.

(Ph4P)2[Fe6C(CO)14(S)] {(Ph4P)2[1]}:

A 200-mL flask was fitted with Ph4P[Fe6C(CO)15(SO2Me)] (3 mmol, 3.557 g) and THF (100 mL), followed by a solution of Na-naphthenide (6.7 mmol, 35 mL, 0.19 M) in THF. After 1 h, the reaction mixture was treated with Ph4PBr (6.7 mmol, 2.810 g), added as solid. After stirring overnight, the reaction mixture was concentrated to dryness. The residue was washed with Et2O/THF (10:1) to remove naphthalene. After purification by flash column chromatography (silica gel; gradient elution: CH2Cl2/Et2O = 1:1, CH2Cl2/Et2O = 2:1, CH2Cl2/MeCN = 10:1) in the glovebox, the second band was collected, and black solids were obtained after evaporation of solvents. Yield: 2.871 g (66 %). IR (THF): υCO 2020(vw), 1958(vs), 1949(vs), 1925(vw) cm−1; IR (MeCN): υCO 2024(vw), 1962(sh), 1956(vs), 1928(vw) cm−1. 13C NMR (126 MHz, CD3CN): δ = 118.5 (C-P), 119.2 (C-P), 131.3 (CH), 131.4 (CH), 135.6 (CH), 135.7 (CH), 136.4 (CH), 136.4 (CH), 223.3 (CO), 227.2 (CO), 229.0 (CO), 478.4 (μ6-C). ESI-MS: m/z 1110.7 for [M − Ph4P] ({Ph4P[Fe6C(CO)14(S)]} calcd. 1110.6). Anal. Calcd for C63H40Fe6O14P2S: C, 52.18; H, 2.78; N, 0.00; found C, 52.18; H, 2.76; N, 0.17.

Oxidation of (Ph4P)2[1] to Fe6C(CO)14(S) ([2]0):

A 100-mL round flask was charged with (Ph4P)2[1] (0.06 mmol, 87.0 mg) and CH2Cl2 (30 mL) in the glovebox. The solution was sparged with CO for 0.5 h, followed by addition of [Cp2Fe]BF4 (0.13 mmol, 35.5 mg) in CH2Cl2 (10 mL) at room temperature. The reaction was monitored by FT-IR. After 10 min, clean CO bands of 2 were observed and did not change in 6 h. After column chromatography (silica gel; gradient elution: Et2O, CH2Cl2/pentane = 1:1) in the ventilation hood, [2]0 was obtained as brownish black solids. Yield: 41.2 mg (83 %). The same procedure was applied to the protonation reaction. A 100-mL round flask was charged with (Ph4P)2[1] (0.06 mmol, 87.0 mg) and CH2Cl2 (30 mL) in the glovebox. The solution was sparged with CO for 0.5 h, followed by addition of HBF4·Et2O (degassed, 1.2 mmol, 195 mg) in CH2Cl2 (10 mL) at room temperature. The reaction was monitored by FT-IR. After 10 min, clean CO bands of 2 were observed and did not change in 6 h. H2 was detected by gas chromatography. After column chromatography (silica gel; CH2Cl2/pentane = 1:1) in the ventilation hood, [2]0 was obtained as brownish black solids. Yield: 40.3 mg (81 %).

Fe6C(CO)13(tBuNC)5(S) ([3]0):

A solution of Fe6C(CO)16(S) (0.155 mmol, 128 mg) in 10 mL of CH2Cl2 was treated with tBuNC (0.775 mmol, 64.3 mg) at room temperature. The solution was allowed to stir overnight. After purification by column chromatography (silica gel; gradient elution: CH2Cl2/pentane = 1:1, CH2Cl2/pentane = 5:1) in the glovebox, the second band was collected. Removal of solvent gave a crude product that was further purified by column chromatography (silica gel; gradient elution: CH2Cl2/pentane = 1:1, CH2Cl2; second band) to afford brownish black solids. Yield: 36 mg (20 %). IR (CH2Cl2): υNC 2210(vw), 2165(m); υCO 2042(m), 1990(s), 1981(s), 1973(sh), 1939(vw), 1916(vw) cm−1. 1H NMR (500 MHz, CD2Cl2): 1.49 (s, 9H, CH3), 1.55 (s, 36H, CH3). 13C NMR (126 MHz, CD2Cl2): δ = 30.6 (CH3), 31.0 (CH3), 58.7 (quat. C), 58.9 (quat. C), 152.9 (NC), 153.5 (NC), 217.8 (CO), 220.9 (CO), 221.3 (CO), 467.7(μ6-C). Anal. Calcd for C39H45Fe6N5O13S: C, 40.42; H, 3.91; N, 6.04; found C, 39.60; H, 3.65; N, 5.84.

(Ph4P)2[Fe7C(CO)17(S)] {(Ph4P)2[4]}:

A solution of (Ph4P)2[1] (0.1 mmol, 145 mg) in THF (10 mL) was treated with Fe(bda)(CO)3 (0.12 mmol, 35.0 mg). After 48 h, the solvent was removed. The solids extracted into a small volume of CH2Cl2 solution. Dilution of this extract with Et2O gave a black solid. Yield: 135 mg (85 %). IR (THF): υCO 2037(vw), 1991(s), 1964(vs) cm−1; IR (CH2Cl2): υCO 2042(vw), 1995(s), 1968(vs) cm−1. 13C NMR (126 MHz, CD3CN): δ = 118.5 (C-P), 119.2 (C-P), 131.2 (CH), 131.3 (CH), 135.6 (CH), 135.7 (CH), 136.3 (CH), 136.4 (CH), 227.2 (CO), 229.0 (CO), 229.8 (CO), 482.4 (μ6-C). ESI-MS: m/z 911.7 for [M − 2Ph4P] ([Fe7C(CO)17(S)] Calcd 911.4). Anal. Calcd for C66H40Fe7N2O17P2S: C, 49.86; H, 2.54; N, 0.00; found C, 50.01; H, 2.70; N, 0.00. Single crystals suitable for crystallographic analysis were grown by vapor diffusion of Et2O into CH2Cl2 solution at −30 °C.

(Ph4P)2[Fe6MoC(CO)18(S)] {(Ph4P)2[5]}:

A CH2Cl2 solution (20 mL) of (Ph4P)2[1] (0.1 mmol, 145 mg) was treated with Mo(CO)4(norbornadiene) (0.12 mmol, 36.0 mg) at RT. After 96 h, solvents were removed under vacuum. The residue was extracted into a small volume of CH2Cl2. Dilution of this extract with Et2O gave a black solid. Yield: 129 mg (78 %). IR (CH2Cl2): υCO 2000(m), 1968(vs) cm−1. 13C NMR (126 MHz, CD2Cl2): δ = 117.6 (C-P), 118.3 (C-P), 130.9 (CH), 131.0 (CH), 134.8 (CH), 134.9 (CH), 136.1 (CH), 213.1 (CO), 222.9 (CO), 226.9 (CO), 228.7 (CO), 479.2 (μ6-C). ESI-MS: m/z 1320.9 for [M − Ph4P]({Ph4P[Fe6MoC(CO)18(S)]} Calcd 1320.5). Anal. Calcd for C67H40Fe6MoO18P2S: C, 48.53; H, 2.43; N, 0.00; found C, 47.83; H, 2.34; N, 0.22. Single crystals suitable for crystallographic analysis were grown by vapor diffusion of Et2O into CH2Cl2 solution at −30 °C.

(Ph4P)2[Fe6C(CO)13(S)(SO2)] {(Ph4P)2[6]}:

A 100-mL round flask was charged with (Ph4P)2[Fe6C(CO)14(S)] (1.0 mmol, 1.45 g) and THF (50 mL). The flask was then brought out of the glovebox and put in an ice bath. An SO2-filled balloon (150–170 mL) was attached to the headspace of the flask. The reaction mixture was warmed to room temperature gradually overnight, forming a greenish black solution. The product was purified by flash column (silica gel; Et2O/MeCN = 10:1 to 2:1) in the glovebox. The greenish fraction was collected. Removal of solvents afforded a black residue. Yield: 965 mg (65 %). IR (THF): υCO 2029(vw), 1975(sh), 1968(vs),1941(w) cm−1; IR (DME): υCO 2030(vw), 1969(vs),1942(w) cm−1. IR (solid): υCO 2031(w), 1952(vs), 1929(vs), 1904(vs), 1882(vs) cm−1; υSO 1041(s) cm−1. 13C NMR (126 MHz, CD3CN): δ = 118.3 (C-P), 119.0 (C-P), 131.1 (CH), 131.2 (CH), 135.4 (CH), 135.5 (CH), 136.2 (CH), 219.1 (CO), 219.9 (CO), 220.5 (CO), 221.2(CO), 223.3 (CO), 479.3 (μ6-C). ESI-MS: m/z 1147.0 for [M – Ph4P]({Ph4P[Fe6C(CO)13(S)(SO2)]} Calcd 1146.6). Anal. Calcd for C62H40Fe6O15P2S2: C, 50.11; H, 2.71; N, 0.00; found C, 50.28; H, 2.76; N, 0.25. Single crystals suitable for crystallographic analysis were grown by vapor diffusion of Et2O into a CH2Cl2 solution at −30 °C.

Deposition Numbers 2019490 {for [3]0}, 2019487 {for (Ph4P)2[4]}, 2019488 {for (Ph4P)2[5]}, and 2019489 {for Na(Ph4P)[6]} contain the supplementary crystallographic data for this paper. These data are provided free of charge by the joint Cambridge Crystallographic Data Centre and Fachinformationszentrum Karlsruhe Access Structures service www.ccdc.cam.ac.uk/structures.

Supplementary Material

Supporting Info

Acknowledgments

This research was supported by the U.S. National Institutes of Health through grant GM61153.

Footnotes

Supporting information and ORCID(s) from the author(s) for this article are available on the WWW under https://doi.org/10.1002/ejic.202000736.

References

  • [1].Kuppuswamy S, Wofford JD, Joseph C, Xie Z-L, Ali AK, Lynch VM, Lindahl PA, Rose MJ, Inorg. Chem 2017, 56, 5998–6012. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [2].Bortoluzzi M, Ciabatti I, Cesari C, Femoni C, Iapalucci MC, Zacchini S, Eur. J. Inorg. Chem 2017, 2017, 3135–3143. [DOI] [PubMed] [Google Scholar]
  • [3].Sickerman NS, Tanifuji K, Hu Y, Ribbe MW, Chem. Eur. J 2017, 23, 12425–12432. [DOI] [PubMed] [Google Scholar]
  • [4].Lee SC, Holm RH, Chem. Rev 2004, 104, 1135–1158. [DOI] [PubMed] [Google Scholar]
  • [5].Tanifuji K, Ohki Y, Chem. Rev 2020, 120, 5194–5251. [DOI] [PubMed] [Google Scholar]
  • [6].a) Creutz SE, Peters JC, J. Am. Chem. Soc 2015, 137, 7310–7313; [DOI] [PMC free article] [PubMed] [Google Scholar]; b) Čorić I, Holland PL, J. Am. Chem. Soc 2016, 138, 7200–7211. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [7].a) Jasniewski AJ, Wilcoxen J, Tanifuji K, Hedman B, Hodgson KO, Britt RD, Hu Y, Ribbe MW, Angew. Chem. Int. Ed 2019, 58, 14703–14707; Angew. Chem. 2019, 131, 14845; [DOI] [PMC free article] [PubMed] [Google Scholar]; b) Spatzal T, Perez KA, Einsle O, Howard JB, Rees DC, Science 2014, 345, 1620–1623; [DOI] [PMC free article] [PubMed] [Google Scholar]; c) Sippel D, Einsle O, Nat. Chem. Biol 2017, 13, 956–960. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [8].Fehlner TP, in Encycl. Inorg. Chem 2006. (eds King RB, Crabtree RH, Lukehart CM, Atwood DA and Scott RA). [Google Scholar]
  • [9].Drance MJ, Mokhtarzadeh CC, Melaimi M, Agnew DW, Moore CE, Rheingold AL, Figueroa JS, Angew. Chem. Int. Ed 2018, 57, 13057–13061; Angew. Chem. 2018, 130, 13241. [DOI] [PubMed] [Google Scholar]
  • [10].Cetini G, Stanghellini PL, Rossetti R, Gambino O, Inorg. Chim. Acta 1968, 2, 427–432. [Google Scholar]
  • [11].Lawrence JD, Rauchfuss TB, Wilson SR, Inorg. Chem 2002, 41, 6193–6195. [DOI] [PubMed] [Google Scholar]
  • [12].a) McGale J, Cutsail GE, Joseph C, Rose MJ, DeBeer S, Inorg. Chem 2019, 58, 12918–12932; [DOI] [PMC free article] [PubMed] [Google Scholar]; b) Joseph C, Kuppuswamy S, Lynch VM, Rose MJ, Inorg. Chem 2018, 57, 20–23. [DOI] [PubMed] [Google Scholar]
  • [13].Tachikawa M, Geerts RL, Muetterties EL, J. Organomet. Chem 1981, 213, 11–24. [Google Scholar]
  • [14].Bockman TM, Cho HC, Kochi JK, Organometallics 1995, 14, 5221–5231. [Google Scholar]
  • [15].a) Bogdan PL, Sabat M, Sunshine SA, Woodcock C, Shriver DF, Inorg. Chem 1988, 27, 1904–1910; [Google Scholar]; b) Chihara T, Kubota H, Fukumoto M, Ogawa H, Yamamoto Y, Wakatsuki Y, Inorg. Chem 1997, 36, 5488–5497. [Google Scholar]
  • [16].Liu L, Rauchfuss TB, Woods TJ, Inorg. Chem 2019, 58, 8271–8274. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [17].Karet GB, Norton DM, Stern CL, Shriver DF, Inorg. Chem 1994, 33, 5750–5753. [Google Scholar]
  • [18].Fiedler J, Nervi C, Osella D, Calhorda MJ, Godinho SSMC, Merkel R, Wadepohl H, J. Chem. Soc., Dalton Trans 2002, 3705–3711. [Google Scholar]
  • [19].Carr CR, Taheri A, Berben LA, J. Am. Chem. Soc 2020, 142, 12299–12305. [DOI] [PubMed] [Google Scholar]
  • [20].Bor G, Dietler UK, Stanghellini PL, Gervasio G, Rossetti R, Sbrignadello G, Battiston GA, J. Organomet. Chem 1981, 213, 277–292. [Google Scholar]
  • [21].13C NMR signals for μ6-carbide ligands fall into a small range (478–483 ppm): 478.4 ([1]2−), 482.4 ([4]2−), 479.2 ppm ([5]2−), 479.3 ppm ([6]2−), 484.1 ppm {[Fe6C(CO)16]2−},[18] and 483.2 ppm {[Fe6C(CO)15(SO2)]2−}[18].

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supporting Info

RESOURCES