Catalytic functions of cubane-type M₄S₄ clusters

Abstract

Catalysis by the cubane-type Fe₄S₄ cluster has been found in metalloproteins such as biotin synthase, aconitase, and IspH, whereas some synthetic M₄S₄ clusters show catalytic activities for biologically related reactions or various types of organic transformations. This minireview summarizes recent progress in the chemistry of catalytic functions of cubane-type M₄S₄ clusters.

---

Hidetake Seino

Hidetake Seino received his doctorate in 1997 from the University of Tokyo under the supervision of Professor M. Hidai, where he studied the synthesis of organonitrogen compounds by using transition metal–dinitrogen complexes. Then he joined the group of Professor Y. Mizobe at the Institute of Industrial Science, the University of Tokyo, investigating the chemistry of metal–sulfide clusters and the conversion of small molecules by using transition metal complexes. His current research interests focus on development of new catalytic systems by learning from the mechanisms of metalloenzymes.

Masanobu Hidai

Masanobu Hidai, born in 1940, received his doctorate from the Department of Industrial Chemistry at the University of Tokyo in 1968. He then joined the staff and became Full Professor in 1986. From 1976 to 1977, he spent time at the University of Sussex, UK, working with Professor J. Chatt. He received the award of the Chemical Society of Japan in 1996. After his retirement from the University of Tokyo in 2000, he continued his research at Tokyo University of Science until 2006. His research interests include the chemistry of nitrogen fixation and organic synthesis catalyzed by transition metal complexes.

---

Introduction

The cubane-type M₄S₄ structure, in which four metal atoms (Mⁿ⁺) and four μ₃-sulfido ligands (S²⁻) are alternately occupying vertices of a cubic-like framework, is one of representative morphologies of metal sulfido clusters and has appreciable thermodynamic stability. These types of clusters are active for multi-step redox and coordination change at the metal sites, while the fairly robust cubane-type cores are retained in these processes. The cubane-type Fe₄S₄ cluster is often observed in metalloproteins such as ferredoxin, hydrogenase, and the Fe protein of nitrogenase, where the major function of the cluster is to facilitate electron transfer.¹ However, the Fe₄S₄ unit in aconitase is well known to catalyze the stereospecific interconversion of citrate and isocitrate, which represents a vital metabolic pathway in the tricarboxylic acid (TCA) cycle. Further, an Fe₄S₄ cluster is combined with another metal unit by a bridging cysteinyl ligand to compose the active sites in metalloenzymes such as sulfite reductase,² [FeFe] hydrogenase (H-cluster),³ and acetyl coenzyme A synthase (A-cluster).⁴ Recent advances in inorganic synthesis have now enabled the synthesis of a variety of homonuclear and heteronuclear cubane-type M₄S₄ clusters containing not only the first-row transition metals, but also the second- and third-row transition metals.^5–9 Their structures and reactivities have extensively been studied partly in relevance to biological functions of metalloproteins.^5–7 Furthermore, unique catalysis by cubane-type clusters containing noble metals in particular, which are not found in metalloproteins, has also been explored.^9,10 This minireview is the first attempt to focus on catalytic functions of cubane-type M₄S₄ clusters covering a wide field of research, including bioinorganic, inorganic, organometallic, organic, and electro- chemistry.

1. Biological organic transformations catalyzed by the cubane-type Fe₄S₄ cluster in enzymes

The biological Fe₄S₄ clusters that play a role in electron transfer are generally coordinated by four cysteine residues in proteins. On the other hand, most of the Fe₄S₄ clusters catalyzing chemical transformations possess a common structure, in which three iron atoms are bound to cysteinyl ligands, and the remaining fourth iron atom serves as the coordination site of substrates (Chart 1).¹¹ Several types of catalytic functions have been revealed for such Fe₄S₄ active sites.

Chart 1 Biological Fe₄ S₄clusters.

The radical SAM superfamily is a group of enzymes that utilize S-adenosylmethionine (SAM) to initiate diverse radical mediated reactions (Chart 2).¹² About 3000 enzymes of this family have been found, and some of them have been determined by X-ray crystallographic structure analysis, e.g.lysine 2,3-aminomutase (LAM), an enzyme for biosynthesis of molybdopterin (MoaA), biotin synthase (BioB), and coproporphyrinogen oxidase (HemN). Extensive spectroscopic studies suggest that the enzymatic mechanisms of this superfamily involve essentially common initial steps (Scheme 1). A [Fe₄S₄]⁺cluster coordinates SAM at the unique iron subsite and transfers one electron to SAM. This causes cleavage of the C–S bond to generate methionine (Met) and 5′-deoxyadenosyl radical (DOA·), and the latter mediates various types of radical reactions. In the reaction of LAM, this radical abstracts a β-hydrogen of the α-lysine that is in a form of Schiff base with pyridoxal 5′-phosphate (PLP), and the generated carbon radical rearranges with migration of the nitrogen to the β-carbon. The resulting α-C radical is donated a hydrogen from DOAH to give the β-lysine–PLP. Reaction of BioB consumes two equivalents of SAM to cleave two unreactive C–H bonds, and the second iron-sulfur center [2Fe-2S] supplies a sulfur atom to the activated substrate. It is proposed that the sulfur atom of the formed Met coordinates to the unique iron when SAM acts as a catalyst (e.g.LAM and MoaA). In contrast, association of the Met sulfur with a cluster sulfide is suggested for the enzymes that use SAM as a substrate (e.g. BioB and HemN).

Chart 2 Examples of reactions catalyzed by radical SAM enzymes.

Scheme 1 Two mechanisms for the reductive cleavage of SAM.

Aconitase catalyzes the stereospecific interconversion of citrate and isocitratevia cis-aconitate, comprising a part of the TCA cycle.¹³ X-ray crystal structure determinations of aconitase in the substrate-free form and with bound isocitrate have shed light on critical aspects of the catalytic mechanism.¹⁴Scheme 2 shows the proposed mechanism for the conversion of isocitrate to citratevia cis-aconitate at the active site of aconitase. In the resting state, the unique iron atom is bound by a hydroxide ion. In the presence of the substrate, the hydroxide ion is protonated, and the resulting water ligand and the substrate are both coordinated to the unique iron. Consequently, the geometry of the iron changes from four-coordinate to six-coordinate. The substrate is then dehydrated to form cis-aconitate at this site through attack by Ser-642 alkoxide and His-101 imidazolium. Citrate is formed via the reorientation of cis-aconitate to the other binding mode (180° rotation) followed by rehydration. The reaction proceeds via a nonredox process, and the unique iron center just acts as a Lewis acid. There are considerable numbers of enzymes that contain the active sites closely related to aconitase in the hydro-lyase family.

Scheme 2 Mechanisms for the conversion of isocitrate to citratevia cis-aconitate.

An Fe₄S₄ cluster containing protein, called IspH, which is found in pathogenic microorganisms, catalyzes the formation of isoprenoid building blocks, isopentenyl diphosphate (IPP) and dimethylallyl diphosphate (DMAPP) by the 2H⁺/2e⁻reduction of (E)-1-hydroxy-2-methyl-2-butenyl 4-diphosphate (HMBPP). Quite recently, a novel bioorganometallic mechanism for IspH catalysis has been proposed on the basis of X-ray structure analysis¹⁵ as well as detailed spectroscopic investigations of action and inhibition of IspH.¹⁶ The proposed mechanism of action of IspH is shown in Scheme 3. HMBPP initially binds to the unique iron in a [Fe₄S₄]²⁺ cluster to form a hydroxyl complex. Reduction of the cluster by one electron gives a π-complex of HMBPP, which is followed by protonation to remove the 4-OH as H₂O and afford an η¹-allyl complex. The cluster is then reduced by the second electron to form an η³-allyl complex. Finally, protonation at C2 or C4 gives rise to the formation of IPP or DMAPP, respectively. It has been demonstrated that IspH is effectively inhibited by alkynyl diphosphates, which probably form more stable π-complexes than alkenyl compounds. Another organometallic mechanism has quite recently been proposed for action and inhibition of the Fe₄S₄ isoprenoid biosynthesis protein, called GcpE, which catalyzes the 2H⁺/2e⁻reduction of 2-C-methyl-D-erythritol-2,4-cyclo-diphosphate (MEcPP) to HMBPP.¹⁷ It is supposed that MEcPP is converted at the unique iron atom into the intermediate having a ferraoxetane structure, which is then deoxygenated to an η²-alkene complex of HMBPP. An alkyne analogue of HMBPP has been shown to inhibit GcpE. With regard to binding and conversion of MEcPP, however, the possibility of alternative pathways has been suggested, and the detailed mechanism is inconclusive at present.¹⁸

Scheme 3 Two steps in the isoprenoid biosynthesis and the proposed IspH mechanism.

2. Catalysis by synthetic Fe₄S₄ clusters as biological models and its applications

In relevance to the redox functions and the above catalytic roles of the metalloproteins containing the cubane-type Fe₄S₄ clusters, many studies have been performed by using the synthetic analogues [Fe₄S₄(SR)₄]ⁿ⁻ (n = 1–3) since the first preparation by Holm and co-workers (Chart 3, top).¹⁹ They demonstrated the reductive cleavage of sulfonium cations by [Fe₄S₄(SR)₄]³⁻ as the reactivity analogue of SAM enzymes.^20,21 The Fe₄S₄ clusters, whose iron subsites are differentiated in the ratio 3 : 1, were also developed by use of a tridentate thiolate ligand bound to the three iron atoms. The unique iron atom in these site-differentiated clusters binds various ligands (hydroxide, carboxylates, carbon donors, etc.) and adopts four-, five-, and six-coordinate geometries via the site-specific substitution reactions (Chart 3, bottom).^21,22 Thus, these synthetic clusters provide good structural models for the active site clusters in enzymes including aconitase. Catalytic hydrolysis of aryl esters by [Fe₄S₄(SPh)₄]²⁻ (1a) was examined by another group in relation to the function of aconitase.²³

Chart 3 Synthetic Fe₄ S₄clusters.

Similarly to the biological role of [Fe₄S₄(S-Cys)₄]ⁿ⁻, the synthetic Fe₄S₄ clusters are utilized as catalyst precursors in electron transfer reactions. The groups of Tezuka and Hidai reported that controlled-potential electrolysis in the presence of [Fe₄S₄(SCH₂Ph)₄]²⁻ resulted in reduction of carbon dioxide at more positive potentials (∼ −1.7 V vs.SCE) than those without catalysts (–2.4 V vs.SCE).²⁴Formate was predominantly produced instead of oxalate. Analogous reactions catalyzed by the Fe₄S₄ clusters bearing a macrocyclic tetrathiolate ligand have been reported.²⁵ Other electrochemical reductions that convert nitrogenase substrates are described in the next section.

Redox reactions between electron-donor and -acceptor molecules are facilitated by 1 and analogues as summarized in Chart 4. In the presence of a catalytic amount of 1a, thiols are smoothly oxidized to disulfide by O₂, quinones, and other organic electron acceptors.²⁶Reduction of organic nitro compounds with thiols to give amines was demonstrated by this method.^27,28Oxidation of benzoin with benzoquinone was examined by using analogues of 1 with bulky thiolate ligands or cysteine-containing peptides.²⁹ It has been proposed that the oxidized Fe₄S₄³⁺ core formed viaelectron transfer to quinone binds benzoin at the iron subsite by replacing the thiolate ligand and is reduced to the Fe₄S₄⁺ state concurrent with the fomation of benzil. The agglomerates, in which the Fe₄S₄ cores were cross-linked by dithiolate ligands had similar catalytic activity.³⁰

Chart 4 Redox-reactions catalyzed by synthetic Fe₄S₄ clusters.

When 1a and NaBH₄ are used as the catalyst and the electron donor respectively in protic media, phenylacetylene and diphenylacetylene are hydrogenated predominantly to styrene and cis-stilbene respectively, and azobenzene is reduced to hydrazobenzene and aniline.^31,32 This method has also been applied to deoxygenation of oxiranes into olefins, which is assumed to be the crucial function of GcpE enzyme (vide supra).³³ Stoichiometric studies on reduction of proton and acetylene to H₂ and ethylene, respectively, have confirmed that the reduced cluster of Fe₄S₄⁺ core acts as the one electron donor and that a thiolate ligand in this cluster presumably dissociates to open the iron subsite for substrate binding.^34–36

Treatment of [Fe₄S₄Cl₄]²⁻ with phenyllithium under H₂ gave a catalyst for hydrogenation of alkenes, which operated under 1 atm of H₂ at room temperature.³⁷ Addition of a mixture of PhSH and PhSSPh after reaction allowed recovery of cluster species as 1a. In contrast, treatment of 1 with phenyllithium exerted no such hydrogenation activity under H₂.

3. Nitrogenase-related transformations catalyzed by cubane-type MFe₃S₄ clusters (M = Mo, V, Fe)

Nitrogenase is an enzyme, which converts atmospheric N₂ into ammonia by use of electrons and protons. Three types of nitrogenases are known, which possess MoFe, VFe, or Fe-only sulfide clusters at their active sites, and the efficiencies of the enzymes decrease in this order. The recent single crystal X-ray analysis of the Mo-containing nitrogenase has revealed that the active site (FeMo-co) is an unprecedented MoFe₇S₉X cluster with two incomplete cubane-type MoFe₃S₃ and Fe₄S₃ units bridged by three μ₂-sulfido ligands and the interstitial μ₆-X atom (X = N, O, or C) (Chart 5).³⁸ In the late 1970s, however, the active site of the nitrogenaseMoFe protein was presumed to be a cubane-type MoFe₃S₄ cluster on the basis of Mo EXAFS analysis of the MoFe protein.³⁹ A similar cubane-type VFe₃S₄ core has been also suggested for the active site of the V-containing nitrogenase,^40,41 although its crystallographic structure is not yet available. In this context, intensive studies had been focused on the synthesis and reactivity of cubane-type MFe₃S₄ (M = Mo, V) clusters (Chart 6).

Chart 5 Structure of FeMo-co.

Chart 6 Synthetic cubane-tube MoFe₃S₄ and VFe₃S₄ clusters investigated for nitrogenase-related catalysis.

The groups of Holm and Garner almost simultaneously reported synthesis of the first synthetic cubane-type MoFe₃S₄ clusters in the dimeric forms such as [Mo₂Fe₆S₈(SR)₉]³⁻ (2), in which the Mo sites of two MoFe₃S₄ cores were connected by bridging thiolate ligands.^42,43 The monomeric clusters [(cat)(L)MoFe₃S₄(SR)₃]²⁻ (3: cat = catecholate; L = 2e donors) were prepared later⁴⁴ and proven to coordinate various pseudosubstrates of nitrogenase, such as CN⁻, N₃⁻, N₂H₄, NH₃, and MeCN, as L at the Mo site.⁴⁵ These clusters undergo reversible 1e and 2e (only for 2) reductions, and the reduced clusters revert to the initial oxidation level on treatment with proton sources (e.g., PhSH or Et₃NH⁺) accompanied with formation of H₂.⁴⁶Proton reduction is known as one of the functions of nitrogenase.

Coucouvanis and co-workers demonstrated nitrogenase-related transformations by using the analogue of 3, [(Cl₄-cat)-(MeCN)MoFe₃S₄Cl₃]²⁻ (4: Cl₄-cat = tetrachlorocatecholate).⁴⁷ The reduction of hydrazine to ammonia in the system utilizing cobaltocene (Cp₂Co) and 2,6-lutidinium chloride (LutH·Cl) as external sources of electrons and protons, respectively (Eq 1), was catalyzed by 4 at ambient temperature in MeCN.^48–50 This cluster also promoted the disproportionation reaction of hydrazine (Eq 2) at a rate slower than the above reduction by one order of magnitude. The iron-only cluster [Fe₄S₄Cl₄]²⁻ was inactive for these reactions. Reduction of hydrazine is considered to proceed at the single Mo site and is inhibited remarkably by PEt₃ and CO, which bind tightly to the Mo site. Although several double-cubane clusters containing a hydrazine ligand as the bridging unit between the Mo atoms were isolated,⁵¹ these showed quite poor activities. An analogous cluster [(Hcitr)MoFe₃S₄Cl₃]³⁻ (Hcitr = κ³-citrate trianion), which was prepared in relevance to the homocitrate-binding Mo site in FeMo-co, exhibited an even higher catalytic rate for hydrazine reduction (about 80 cycles in the initial 5 min under certain conditions) than 4 but had no activity for disproportionation. It is presumed that the coordination site for hydrazine is formed via protonation at the tridentate Hcitr ligand. Similar catalytic activities were also observed for other MoFe₃S₄ clusters binding carboxylate ligands at the Mo site.⁵²

N₂H₄ + 2Cp₂Co + 4LutH·Cl→2NH⁺₄ + 2[Cp₂Co]Cl + 4Lut (1)

3N₂H₄ → 4NH₃ + N₂ (2)

Employment of Cp₂Co and LutH·Cl as sources of electrons and protons, respectively, in the presence of 4 as a catalyst has further enabled the reduction of acetylene into ethylene,⁵³cis-dimethyldiazene into methylamine,⁵⁴ or phenylhydrazine to aniline and ammonia (Chart 7).⁴⁹Methylhydrazine was converted to methylamine and ammonia, though in competition with the reduction of proton to evolve H₂.⁴⁸ The vanadium containing cubane-type clusters [(L)₃VFe₃S₄Cl₃]⁻ ((L)₃ = (dmf)₃, (dmf)(bpy): dmf = N,N-dimethylformamide; bpy = 2,2′-bipyridyl: Chart 6)⁵⁵ were also shown to be catalytically effective for the reduction of hydrazine or phenylhydrazine.^48,56

Chart 7 Reductions relevant to nitrogen fixation catalyzed by 4.

Cluster 3 (cat = Cl₄-cat; L = Me₂CO; R = p-C₆H₄-n-C₈H₁₇) catalyzed the reduction of n-C₅H₁₁N₃ by Na₂S₂O₄ in aqueous Triton X-100 micellar solutions, giving equal amount of n-C₅H₁₁NH₂ and N₂ (Eq 3, R = n-C₅H₁₁).⁵⁷ Addition of methyl viologen to these system not only increased reaction rates but also gave ammonia and hydrazine, whose production required larger numbers of electrons per substrate. The iron cluster 1 (R = p-C₆H₄-n-C₈H₁₇) catalyzed only Eq 3 under the same conditions.

RN₃ + 2e⁻ + 2H⁺ → RNH₂ + N₂ (3)

Tanaka and co-workers extensively investigated the catalytic activities of the cubane-type MoFe₃S₄ and Fe₄S₄ clusters for nitrogenase-related transformations under controlled potential electrolysis conditions (Hg working electrode). The double cubane-cluster 2a was effective for the reduction of hydrazine into ammonia in methanol/THF solution at the potential that attained 2e reduction of 2a (–1.25 V vs.SCE).⁵⁸Reduction of proton to form H₂ competitively took place (vide supra) only in a small ratio (∼3%). The iron-only cubane cluster 1a had higher catalytic activity for H₂ evolution than 2a⁵⁹ and exhibited inferior selectivity for reduction of hydrazine against that of proton (44 : 56). The proportion of H₂ evolution became greater in aqueous media for both 1a and 2a, especially at low pH values. Acetylene was also preferentially reduced in these systems into ethylene.^59,60 In contrast to single cubane clusters 3 and 4, double cubane clusters 2 are proposed to bind substrates not at the Mo atom but at the Fe subsite that has an easily dissociable terminal thiolate ligand. Observations that acetylene³⁶ or methyl azide⁶¹ coordinate to the iron subsite in the 1e-reduced species of 2a were claimed. Reduction of CH₃NC by electrolysis in the presence of 1a and 2a gave CH₃NH₂, CH₄, C₂H₆, and a small amount of C₂H₄.⁶² These hydrocarbons are considered to originate from the isocyanide carbon, and it is explained that C₂H₆ is produced via the insertion of CH₃NC into the CH₃–metal bond that is formed at an intermediary stage of reduction. Acetonitrile was also reduced to C₂H₆, C₂H₄, and NH₃ in these systems. Electrolysis with 2a converted CH₃N₃ into CH₃NH₂ and N₂via a 2e-transfer process (Eq 3, R = CH₃) with nearly 100% current efficiency at high substrate concentration, while hydrazine and ammonia were also formed in dilute solution as multielectron reduction products.^61,63

In similar systems using 1 or 2, N₃⁻ was transformed into N₂ and ammonia in about a 2 : 1 ratio, concomitantly forming large amounts of H₂ (more than 10-fold of NH₃) and a little hydrazine.⁶⁴ Depression of H₂ evolution and improvement of the ammonia yield was achieved by using the working electrodes, which were prepared by modifying glassy-carbon plates with 2a.⁶⁵ Superiority of this electrode in multielectron reduction of HOCH₂CH₂N₃ was demonstrated in aqueous solution to give large amounts of ammonia with high catalytic rates (Eq 4).^61,63Electrodes modified with 1a and the double-cubane cluster [{(Cl₄-cat)MoFe₃S₄(SPh)₂}₂(μ₂-SPh)₂]⁴⁻ (5: Chart 6), which was regarded as the dimer form of 3, were also tested for the reduction of HOCH₂CH₂N₃ in water (Eq 5).⁶⁶

Electrolysis using the electrode modified with 2a or 5 selectively reduced NO₂⁻ and NO₃⁻ to ammonia at −1.25 V vs.SCE in aqueous solution.⁶⁷ When a CO₂-saturated MeCN solution containing NO₂⁻, PhCOCH₃, and 3Å molecular sieves was electrolyzed in the presence of 2a, PhCOCH₃ acted as the proton source in reduction of NO₂⁻ into N₂ and was subsequently carboxylated to give PhCOCH₂CO₂⁻ (Eq 6).⁶⁸ Similar activity was also observed for 1a, and other organic compounds having active hydrogens (pK_a = 18–21) could replace PhCOCH₃. The CO₂ fixation via 2e reduction was performed in the system that utilized thioesters R′COSEt (R′ = CH₃, C₂H₅, Ph) as the CO₂ acceptor and 2 (R = Et) as the electrocatalyst, giving α-ketocarboxylates R′COCOO⁻ and EtS⁻ (Eq 7).⁶⁹Methyl acrylate underwent similar reductive addition of CO₂ under electrolysis conditions with 2.⁷⁰

4. Nitrogenase-related transformations catalyzed by other cubane-type M₄S₄ clusters

Several synthetic routes have been developed to prepare various heteronuclear cubane-type M₄S₄ clusters with metal compositions other than MFe₃S₄ (M = Fe, Mo, V). The catalytic activities of those clusters for nitrogenase-related transformations have been investigated by our group.

The Mo₂Ir₂S₄ cluster [(MoOCl₂){MoCl₂(dmf)}(IrCp*)₂S₄] (6: Cp* = η⁵-C₅Me₅) catalyzed the reduction of organohydrazine MePhNNH₂ at room temperature in the presence of Cp₂Co and LutH·Cl to give MePhNH.⁷¹ The Mo=O moiety in 6 smoothly condensed with the NH₂ group of MePhNNH₂ to form the hydrazido(2−) ligand as the possible intermediate of the reductive N–N bond cleavage (Scheme 4). The isolated cluster [{Mo(NNMePh)Cl₂}{MoCl₂(dmf)}(IrCp*)₂S₄] liberated PhMeNH (21% yield) by treatment with Cp₂Co and LutH·Cl (each two equiv). Replacement of the two Ir sites in 6 by Rh and/or of two of the sulfide ligands by selenides altered the catalytic activity to some extent (TON = 3–7).

Scheme 4 Reduction of organohydrazine via formation of hydrazido(2−) ligand on the Mo₂Ir₂S₄ cluster.

Disproportionation of hydrazine (Eq 2) was investigated by using the RuMo₃S₄ clusters in THF solution.⁷² In the reaction of [{RuH₂(PPh₃)}(MoCp*)₃S₄][PF₆] (7: Chart 8) with 20 equiv of anhydrous hydrazine at 60 °C, 11.5 equiv of hydrazine was converted after 48 h to give ammonia and N₂ in an almost 4 : 1 ratio. The conversion rate of hydrazine increased to 15.2 equiv after 18 h by changing the phosphine ligand for P(cyclohexyl)₃. From the reaction mixture using 7, the ammonia cluster [{Ru(NH₃)(PPh₃)}(MoCp*)₃S₄][PF₆] (8) was isolated, which exhibited a catalytic activity comparable to 7. In contrast, [{Ru(CO)(PPh₃)}(MoCp*)₃S₄][PF₆] was inactive, presumably because of the difficulty in replacing the CO ligand by hydrazine. Treatment of 7 with 20 equiv of phenylhydrazine at 60 °C for 18 h produced aniline (2.3 equiv), ammonia (1.2 equiv), N₂ (1.6 equiv), and benzene (2.1 equiv), in addition to cluster 8.

Chart 8 RuMo₃S₄ clusters active for disproportionation of hydrazine.

Recently, Mizobe and co-workers have synthesized the RuIr₃S₄ cluster binding molecular N₂ at its Ru site, [{Ru(N₂)(tmeda)}(IrCp*)₃S₄] (tmeda = Me₂NCH₂CH₂NMe₂: Chart 9).⁷³ At present, this is the only example of the stable coordination of N₂ to transition metal clusters containing bridging sulfide ligands. Although the N₂ ligand was not transformed into ammonia, the N₂cluster catalytically reduced proton to H₂ by using various acids ([LutH][BF₄], [Et₃NH][BF₄], etc.) and Cp₂Co as the proton and the electron sources.⁷⁴

Chart 9 The cubane-type cluster binding molecular N₂.

5. Organic transformations catalyzed by cubane-type M₄S₄ clusters

Hidai and co-workers have investigated the synthesis and reactivity of cubane-type M₄S₄ clusters containing noble metals in particular, which are not found in natural enzymes, because these clusters are expected to exhibit unique catalytic activities for organic synthesis.¹⁰

The PdMo₃S₄ cluster [(PdCl){Mo(tacn)}₃S₄][PF₆]₃ (9: tacn = 1,4,7-triazacyclononane) was found to be a highly efficient catalyst for the stereoselective addition of alcohols to electron-deficient alkynes. Treatment of HC≡CCOOMe with 0.06 mol% 9 in a MeOH solution at 40 °C for 24 h produced (Z)-MeOCH=CHCOOMe in 95% yield.⁷⁵ The product selectivity was so high as that concomitant formations of (E)-MeOCH=CHCOOMe and (MeO)₂CHCH₂COOMe were inhibited to each less than 1%. This reaction was applicable to various combinations of terminal and internal alkynoic acid esters R¹C≡CCOOR² (R¹ = H, Me, Ph, COOMe; R² = alkyl, aryl) with alcohols ROH (R = Me, Et, CH₂Ph), and the trans addition products (Z)-ROCR¹=CHCOOR² were selectively obtained. Preservation of the cluster structure during the course of catalysis was confirmed by monitoring UV-vis spectra of the reaction solution.

A plausible mechanism for this catalysis is shown in Scheme 5. Initially, the alkynoic acid ester coordinates to the Pd site of the cluster to form the π-alkyne complex. Successive nucleophilic attack of alcohol to the electron-deficient carbon atom from the outer coordination sphere leads to the formation of the vinylpalladium intermediate. Protonolysis of the Pd–C bond with retention of the stereochemistry around the double bond affords the trans addition product. Although the π-alkyne intermediate can not be clearly observed, it is evidenced from the π-coordination of alkenes to the Pd site of 9 and the well characterized alkyne adduct of the related NiMo₃S₄ cluster (vide infra). It is noteworthy that conventional mononuclear Pd complexes do not work as effective catalysts in this reaction. The unique activity of 9 may originate from the tetrahedral Pd(II)-like site, which is surrounded by three firmly bonded sulfido ligands and provides only a single vacant site. The electron-poor nature of the Pd atom is deduced from the high CO stretching frequency of [{(Pd(CO)}{Mo(tacn)}₃S₄][Cl][PF₆]₃ (2085 cm⁻¹). The Fenske-Hall molecular orbital calculations of the PdMo₃S₄ motif, where a zero oxidation state is assigned to the Pd atom, show that the Pd atom shares electron density through the formation of metal–metal bonds with the Mo(IV)₃ triangle.⁷⁶

Scheme 5 Proposed mechanism for stereoselective addition of alcohol to electron-deficient alkyne.

Cluster 9 also catalyzed the stereoselective addition of carboxylic acids to electron-deficient alkynes. Reactions of HC≡CCOOMe with 3 equiv of acetic acid in an acetonitrile solution proceeded in the presence of catalytic amounts of 9 and NEt₃ to produce MeCOOCH=CHCOOMe with a stereoselectivity of Z/E = 98/2 (Eq 8).⁷⁷ Various aliphatic and aromatic carboxylic acids as well as substituted acrylic acids added to HC≡CCOOMe stereoselectively. Other acetylenes with electron-withdrawing groups (including ester, amide, or sulfone) underwent Z-selective addition of acetic acid. Only the reaction between HC≡CCOPh and acetic acid exceptionally afforded the E-vinyl ketone probably due to the facile Z to Eisomerization of the product. A mechanism analogous to Scheme 5 has been proposed.

Intramolecular addition of acids smoothly takes place even to the non-activated alkyne moieties. Cyclization of alkynoic acids into enol lactones proceeded under mild conditions (20–40 °C, in acetonitrile) in the presence of 0.33 mol% 9 and 3.3 mol% NEt₃ (Chart 10).⁷⁸ Several complexes or salts of Rh, Pd, Ag, and Hg also catalyze these reactions, but the activities of these are much lower than 9. For example, catalytic rates observed for 9 were 12–17 times higher than the mononuclear complex [PdCl₂(PhCN)₂]. 4-Pentynoic acid and 5-hexynoic acid were selectively converted to γ- and δ-lactones, respectively. Catalytic rates for γ-lactonization were remarkably high, and the turnover number of the cyclization of 4-pentynoic acid reached 100 000 after 19 h at 40 °C (300 times within 6 min at 20 °C). Reaction of an internal alkyne 4-hexynoic acid gave a mixture of γ- and δ-lactones, while only γ-lactone was formed from 3-decynoic acid. The cyclization of 6-heptynoic acid to the corresponding ε-lactone was very slow.

Chart 10 Cyclization of alkynoic acids catalyzed by 9.

Although less active than 9, the NiMo₃S₄ clusters [{Ni(L)}(MoCp*)₃S₄][PF₆] (L = 1/2 1,5-cyclooctadiene (10a), dimethyl acetylenedicarboxylate (10b): Chart 11) also catalyze these lactone formations.⁷⁹ The clusters that contain no noble metals work efficiently as catalysts under conditions similar to the above reactions with 9 (0.33 mol% NiMo₃S₄, 3.3 mol% NEt₃, room temperature) in CH₂Cl₂ solution. The activity of 10 was much superior than [PdCl₂(PhCN)₂] in all cases, and the reaction was scarcely promoted by mononuclear Ni compounds or by the incomplete cubane-type cluster [(MoCp*)₃S₄][PF₆]. It is notable that the π-coordination of alkyne to the Ni site has been crystallographically confirmed for the alkyne adduct 10b. This supports the validity of the catalytic mechanism shown in Scheme 5 for both 9 and 10, which have the same electronic configurations.

Chart 11 NiMo₃S₄ clusters active for cyclization of alkynoic acids.

The related PdMo₃S₄ clusters [{Pd(L)}(Cp*Mo)₃S₄][PF₆] (L = dibenzalacetone (11), maleic anhydride) catalyze cyclization of aminoalkynes to cyclic imines.⁸⁰ In the presence of 1 mol% 11, 5-amino-1-phenyl-1-pentyne was quantitatively converted into 2-benzyl-1-pyrroline at 60 °C for 0.5 h in THF (Eq 9). Five- and six-membered cyclic imines were prepared in high yields.

Recently, 11 was found to be effective for direct amination of allylic alcohols.⁸¹ When an equimolar mixture of allyl alcohol and secondary amine were refluxed in CH₂Cl₂ with 0.5 mol% 11 and 0.5 equiv of boric acid (or PhB(OH)₂, B(OMe)₃), N-allylamine was obtained in high yield. Only [Pd(PPh₃)₄] promotes such reactions among conventional Pd(0) and Pd(II) complexes, but this as well as 10a are less active than 11. Even amide was allylated by this method, while allylation of primary amines afforded some amounts of diallylamines. From the reactions of both α- and γ-monosubstituted allyl alcohols, only the γ-monosubstituted allyl amines were obtained. This suggests the catalytic cycle involving the formation of a π-allyl Pd complex as shown in Scheme 6, in which amines may attack the less hindered carbon atom of the π-allyl intermediate. This mechanism has certain resemblance to the IspH mechanism shown in Scheme 3. In fact, the π-allyl cluster [{Pd(η³-C₃H₅)}(MoCp*)₃S₄][PF₆]₂ (12) was produced by treatment of 11 with allyl chloride and AgPF₆, and was fully characterized by X-ray crystallography.⁸² Further, 12 efficiently catalyzed the above allylamine formation from allyl alcohols without addition of boron reagents. Allylation of active methylene compounds was also facilitated by 12.

Scheme 6 Proposed mechanism for the direct amination of allyl alcohol catalyzed by 11.

Our preliminary studies have recently revealed that the cubane-type cluster [(RuCl₃)(MoCp*)₃S₄]⁸³ without any additives catalyzes the polymerization of methyl methacrylate (MMA).⁸⁴ Thus, heating a solution of MMA (5.4 mL) in THF (17.5 mL) in the presence of the cubane-type cluster (10 mmol) at 80 °C for 6 h gave a polymer of MMA (1.25 g) with M_w = 360 000 and M_w/M_n = 2.9 (M_w = weight-averaged molecular weight, M_n = number-averaged molecular weight). Detailed investigations are now in progress to elucidate the mechanism of this polymerization, in comparison with the living radical polymerization of MMA by the combination of Ru(Ind)Cl(PPh₃)₂ (Ind = η⁵-C₉H₇) with an alkyl halide as an initiator.⁸⁵

The groups of Llusar and Pérez-Prieto examined the addition reaction of carbenoids generated from diazo compounds by using the CuMo₃S₄ cluster [(CuCl){MoCl(dmpe)}₃S₄][PF₆] (13: dmpe = Me₂PCH₂CH₂PMe₂: Chart 12).⁸⁶ Intra- and intermolecular cyclopropanation of alkenes in the presence of 1 mol% 13 proceeded with high yields under reflux in CH₂Cl₂ (Eqs 10 and 11). The cluster remained intact after reaction. Optically active diphosphine ligands, 1,2-bis(2,5-dimethylphospholan-1-yl)ethane (Me-bpe), exert a chiral coordination environment on the CuMo₃S₄ core (Chart 12), and one of the enantiomers 14 have exhibited some asymmetric induction (up to 25% ee) in cyclopropanation. Based on the catalytic activities of Br- or Se-congeners of 14, the active site is considered to be generated by cleavage of one Cu–chalcogen bond.⁸⁷

Chart 12 CuMoS₄ catalysts for carbenoid addition.

Sulfido clusters containing molybdenum and cobalt have been investigated in relevance to heterogeneous catalysts used for the industrial hydrodesulfurization (HDS) process.⁸⁸ Curtis and co-workers studied the reactivities of Mo₂Co₂S₃ and Mo₂Co₂S₄ clusters toward desulfurization of organosulfur compounds. When [{Mo(η⁵-C₅Me₄Et)}₂{Co(CO)}₂S₄] was heated with excess PhSH in toluene at 150 °C under CO pressure (1000 psi), PhSSPh and PhSCOPh were produced in addition to the cluster compound [{Mo(η⁵-C₅Me₄Et)}₂{Co(SPh)}₂S₄] (Eq 12).⁸⁹ Amount of PhSCOPh, which was produced via desulfurization of PhSH, was more than 100% per mole of the starting cluster.

Several groups have prepared heterogeneous catalysts by incorporation of the cubane-type cluster [{Mo(H₂O)₃}₃(NiCl)S₄]³⁺ into zeolites by ion exchange⁹⁰ or impregnation of the clusters [{Mo(η⁵-C₅H₄Me)}₃(ML_m)S₄][p-MeC₆H₄SO₃]_n (M = Ru, Rh, Ir, Ni, Pd, Pt) on γ-alumina⁹¹ and employed the catalysts for CO hydrogenation, HDS, or hydrodenitrogenation reactions at high temperatures. However, since the structures of the catalytically active sites are not clear, the details are not described here.

Conclusion

Progress in catalysis by cubane-type M₄S₄ clusters has been summarized here. Intriguing catalysis by the Fe₄S₄ cluster will further be found in other metalloproteins, while synthetic M₄S₄ clusters would serve as unique catalytic precursors for organic synthesis that are not attainable by conventional metal complexes. It should be noted that only a few papers describing catalysis by synthetic M₄S₄ clusters have presented evidence to rule out the possibility that a small amount of active species formed via fragmentation of the cluster cores is responsible for the catalysis. However, the reactions generally proceed under mild conditions, so the relatively robust cubane-type structures are presumed to be fundamentally retained during the course of catalysis in most cases. We must await further investigations by both experimental and theoretical methods to elucidate the detailed reaction mechanisms and advantages of the polynuclear complexes.

Acknowledgements

M.H. expresses his hearty thanks to Professor Yasushi Mizobe (1953–2010) for working together and collaborating with him for almost 30 years. H.S. expresses deep gratitude to Professor Y. Mizobe, who directed him to this research field, provided a fulfilling and comfortable environment for research, and had been giving warm advises and supports. His absence is a great loss to our family. The recent research by Y.M. and H.S. was supported by a Grant-in-Aid for Scientific Research (No. 18065005 on Priority Area “Chemistry of Concerto Catalysis” and No. 21350033 (B)) from the Ministry of Education, Culture, Sports, Science and Technology, Japan.

Notes and references

1. Reviews: R. Cammack, Adv. Inorg. Chem., 1992, 38; H. Beinert, R. H. Holm and E. Münck, Science, 1997, 277, 653–659; M. K. Johnson, Curr. Opin. Chem. Biol., 1998, 2, 173–181; J. B. Broderick, in Comprehensive Coordination Chemistry II, ed. J. A. McCleverty and T. J. Meyer, Elsevier, Oxford, 2004, vol. 8, ch. 8.27, pp. 739–757; M. S. Koay, M. L. Antonkine, W. Gärtner and W. Lubitz, Chem. Biodiversity, 2008, 5, 1571–1587.
2. B. R. Crane, L. M. Siegel and E. D. Getzoff, Science, 1995, 270, 59.
3. J. W. Peters, W. N. Lanzilotta, B. J. Lemon and L. C. Seefeldt, Science, 1998, 282, 1853–1858; Y. Nicolet, C. Piras, P. Legrand, C. E. Hatchikian and J. C. Fontecilla-Camps, Structure, 1999, 7, 13–23.
4. T. I. Doukov, T. M. Iverson, J. Seravalli, S. W. Ragsdale and C. L. Drennan, Science, 2002, 298, 567–572.
5. Fe–S clusters: R. H. Holm, Acc. Chem. Res., 1977, 10, 427–434; A. Nakamura and N. Ueyama, Adv. Inorg. Chem., 1989, 33, 39–67; P. Venkateswara Rao and R. H. Holm, Chem. Rev., 2004, 104, 527–559.
6. Fe–Mo–S clusters: D. Coucouvanis, Acc. Chem. Res., 1991, 24, 1–8; S. M. Malinak and D. Coucouvanis, Prog. Inorg. Chem., 2001, 49, 599–685.
7. Fe–S and Fe–Mo–S clusters: R. A. Henderson, Chem. Rev., 2005, 105, 2365–2437; R. A. Henderson, Coord. Chem. Rev., 2005, 249, 1841–1856.
8. Non-Fe based clusters: S. Harris, Polyhedron, 1989, 8, 2843–2882; T. Shibahara, Adv. Inorg. Chem., 1991, 37, 143–173; T. Shibahara, Coord. Chem. Rev., 1993, 123, 73–147; G. Sakane and T. Shibahara, in Transition Metal Sulfur Chemistry: Biological and Industrial Significance, ed. E. I. Stiefel and K. Matsumoto, American Chemical Sociey, Washington DC, 1996, ch. 13, pp. 225–239; R. Hernandez-Molina and A. G. Sykes, Coord. Chem. Rev., 1999, 187, 291–302; R. Hernandez-Molina and A. G. Sykes, J. Chem. Soc., Dalton Trans., 1999, 3137–3148; R. Hernandez-Molina, M. N. Sokolov and A. G. Sykes, Acc. Chem. Res., 2001, 34, 223–230.
9. M. Hidai, S. Kuwata and Y. Mizobe, Acc. Chem. Res., 2000, 33, 46–52; R. Llusar and S. Uriel, Eur. J. Inorg. Chem., 2003, 1271–1290.
10. M. Hidai and Y. Mizobe, in Transition Metal Sulfur Chemistry: Biological and Industrial Significance, ed. E. I. Stiefel and K. Matsumoto, American Chemical Society, Washington DC, 1996, ch. 19, pp. 310–323; M. Hidai, in Perspectives in Organometallic Chemistry, ed. C. G. Screttas and B. R. Steele, The Royal Society of Chemistry, Special Publication No.287, Cambridge, 2003, pp. 62–73; Y. Ishii and M. Hidai, in Multimetallic Catalysts in Organic Synthesis, ed. M. Shibasaki and Y. Yamamoto, Wiley-VHC, Weinheim, 2004, ch. 9, pp. 201–223; M. Hidai and Y. Mizobe, Can. J. Chem., 2005, 83, 358–374; Y. Mizobe, in Handbook of Chalcogen Chemistry: New Perspectives in Sulfur, Selenium and Tellurium, ed. F. A. Devillanova, The Royal Society of Chemistry, Cambridge, 2007, ch. 11.2, pp. 714–730.
11. Although most of the cubane-type clusters described here have metal–metal bonds in the cluster core, these bonds are omitted for clarity from illustrations of molecular structures throughout this report.
12. Reviews: P. A. Frey and O. Th. Magnusson, Chem. Rev., 2003, 103, 2129–2148; P. A. Frey, A. D. Hegeman and F. J. Ruzicka, Crit. Rev. Biochem. Mol. Biol., 2008, 43, 63–88; A. Marquet, B. T. S. Bui, A. G. Smith and M. J. Warren, Nat. Prod. Rep., 2007, 24, 1027–1040; K. S. Duschene, S. E. Veneziano, S. C. Silver and J. B. Broderick, Curr. Opin. Chem. Biol., 2009, 13, 74–83.
13. Reviews: H. Beinert, M. C. Kennedy and C. D. Stout, Chem. Rev., 1996, 96, 2335–2373; D. H. Flint and R. M. Allen, Chem. Rev., 1996, 96, 2315–2334.
14. A. H. Robbins and C. D. Stout, Proc. Natl. Acad. Sci. U. S. A., 1989, 86, 3639–3643; H. Lauble, M. C. Kennedy, H. Beinert and C. D. Stout, Biochemistry, 1992, 31, 2735–2748; H. Lauble, M. C. Kennedy, M. H. Emptage, H. Beinert and C. D. Stout, Proc. Natl. Acad. Sci. U. S. A., 1996, 93, 13699–13703.
15. T. Gräwert, I. Span, W. Eisenreich, F. Rohdich, J. Eppinger, A. Bacher and M. Groll, Proc. Natl. Acad. Sci., 2010, 107, 1077–1081.
16. W. Wang, K. Wang, Y.-L. Liu, J.-H. No, J. Li, M. J. Nilges and E. Oldfield, Proc. Natl. Acad. Sci. U. S. A., 2010, 107, 4522–4527; K. Wang, W. Wang, J.-H. No, Y. Zhang, Y. Zhang and E. Oldfield, J. Am. Chem. Soc., 2010, 132, 6719–6727.
17. W. Wang, J. Li, K. Wang, C. Huang, Y. Zhang and E. Oldfield, Proc. Natl. Acad. Sci. U. S. A., 2010, 107, 11189–11193.
18. W. Xu, N. S. Lees, D. Adedeji, J. Wiesner, H. Jomaa, B. M. Hoffman and E. C. Duin, J. Am. Chem. Soc., 2010, 132, 14509–14520.
19. T. Herskovitz, B. A. Averill, R. H. Holm, J. A. Ibers, W. D. Phillips and J. F. Weiher, Proc. Natl. Acad. Sci. U. S. A., 1972, 69, 2437–2441.
20. C. J. A. Daley and R. H. Holm, Inorg. Chem., 2001, 40, 2785–2793.
21. C. J. A. Daley and R. H. Holm, J. Inorg. Biochem., 2003, 97, 287–298.
22. T. D. P. Stack and R. H. Holm, J. Am. Chem. Soc., 1988, 110, 2484–2494; J. A. Weigel and R. H. Holm, J. Am. Chem. Soc., 1991, 113, 4184–4191; L. Cai and R. H. Holm, J. Am. Chem. Soc., 1994, 116, 7177–7188; P. V. Rao and R. H. Holm, Chem. Rev., 2004, 104, 527–559 , and references therein.
23. A. K. Yatsimirsky and A. I. Diky, Inorg. Chim. Acta, 1991, 186, 161–163.
24. M. Tezuka, T. Yajima, A. Tsuchiya, Y. Matsumoto, Y. Uchida and M. Hidai, J. Am. Chem. Soc., 1982, 104, 6834–6836; M. Nakazawa, Y. Mizobe, Y. Matsumoto, Y. Uchida, M. Tezuka and M. Hidai, Bull. Chem. Soc. Jpn., 1986, 59, 809–814.
25. T. Tomohiro, K. Uoto and H. Okuno, J. Chem. Soc., Chem. Commun., 1990, 194–195.
26. T. Nagano, K. Yoshikawa and M. Hirobe, Tetrahedron Lett., 1980, 21, 297–300; I. Tabushi, Y. Kuroda and Y. Sasaki, J. Mol. Catal., 1988, 49, 1–7.
27. T. Ito, T. Nagano and M. Hirobe, Chem Pharm. Bull., 1986, 34, 2013–2017.
28. H. Inoue, M. Shirai and E. Haruki, J. Chem. Soc., Chem. Commun., 1987, 674–675.
29. N. Ueyama, T. Sugawara, A. Kajiwara and A. Nakamura, J. Chem. Soc., Chem. Commun., 1986, 434–435; T. Sugawara, N. Ueyama and A. Nakamura, J. Chem. Soc., Dalton Trans., 1991, 249–253.
30. N. Ueyama, T. Sugawara, A. Nakamura, K. Yamaguchi and T. Fueno, Chem. Lett., 1988, 223–224.
31. A. Nakamura, M. Kamada, Sugihashi and S. Otsuka, J. Mol. Catal., 1980, 8, 353–367.
32. T. Ito, T. Nagano and M. Hirobe, Tetrahedron Lett., 1980, 21, 1343–1246.
33. T. Ito, T. Nagano, M. Sato and M. Hirobe, Tetrahedron Lett., 1989, 30, 6387–6388.
34. R. S. McMillan, J. Renaud, J. G. Reynolds and R. H. Holm, J. Inorg. Biochem., 1979, 11, 213–227.
35. K. L. C. Grönberg, R. A. Henderson and K. E. Oglieve, J. Chem. Soc., Dalton Trans., 1998, 3093–3104.
36. K. Tanaka, M. Nakamoto, M. Tsunomori and T. Tanaka, Chem. Lett., 1987, 613–616.
37. H. Inoue and M. Suzuki, J. Chem. Soc., Chem. Commun., 1980, 817–818; H. Inoue and M. Sato, J. Chem. Soc., Chem. Commun., 1983, 983–984.
38. J. Kim and D. C. Rees, Science, 1992, 257, 1677–1682; O. Einsle, F. A. Tezcan, S. Andrade, B. Schmid, M. Yoshida, J. B. Howard and D. C. Rees, Science, 2002, 297, 1696–1700.
39. S. P. Cramer, K. O. Hodgson, W. O. Gillum and L. E. Mortenson, J. Am. Chem. Soc., 1978, 100, 3398–3407.
40. J. M. Arber, B. R. Dobson, R. R. Eady, P. Stevens, S. S. Hasnain, C. D. Garner and B. E. Smith, Nature, 1987, 325, 372.
41. G. N. Goerge, C. L. Coyle, B. J. Hales and S. P. Cramer, J. Am. Chem. Soc., 1988, 110, 4057–4059.
42. G. Christou, C. D. Garner and F. E. Mabbs, J. Chem. Soc., Chem. Commun., 1978, 740–741; G. Christou, C. D. Garner and F. E. Mabbs, Inorg. Chim. Acta, 1978, 28, L189–L190.
43. T. E. Wolff, J. M. Berg, C. Warrick, K. O. Hodgson, R. H. Holm and R. B. Frankel, J. Am. Chem. Soc., 1978, 100, 4630–4632; T. E. Wolff, J. M. Berg, K. O. Hodgson, R. B. Frankel and R. H. Holm, J. Am. Chem. Soc., 1979, 101, 4140–4150.
44. W. H. Armstrong, P. K. Mascharak and R. H. Holm, Inorg. Chem., 1982, 21, 1699–1701; W. H. Armstrong, P. K. Mascharak and R. H. Holm, J. Am. Chem. Soc., 1982, 104, 4373–4383.
45. R. E. Palermo, R. Singh, J. K. Bashkin and R. H. Holm, J. Am. Chem. Soc., 1984, 106, 2600–2612.
46. T. Yamamura, G. Christou and R. H. Holm, Inorg. Chem., 1983, 22, 939–949; Y. Mizobe, P. K. Mascharak, R. E. Palermo and R. H. Holm, Inorg. Chim. Acta, 1983, 80, L65–L67.
47. R. E. Palermo and R. H. Holm, J. Am. Chem. Soc., 1983, 105, 4310–4318.
48. D. Coucouvanis, K. D. Demadis, S. M. Malinak, P. E. Mosier, M. A. Tyson and L. J. Laughlin, J. Mol. Catal. A: Chem., 1996, 107, 123–135.
49. D. Coucouvanis, P. E. Mosier, K. D. Demadis, S. Patton, S. M. Malinak, C. G. Kim and M. A. Tyson, J. Am. Chem. Soc., 1993, 115, 12193–12194.
50. K. D. Demadis, S. M. Malinak and D. Coucouvanis, Inorg. Chem., 1996, 35, 4038–4046.
51. P. R. Challen, S.-M. Koo, C. G. Kim, W. R. Dunham and D. Coucouvanis, J. Am. Chem. Soc., 1990, 112, 8606–8607; P. E. Mosier, C. G. Kim and D. Coucouvanis, Inorg. Chem., 1993, 32, 2620–2621.
52. K. D. Demadis and D. Coucouvanis, Inorg. Chem., 1994, 33, 4195–4197; K. D. Demadis and D. Coucouvanis, Inorg. Chem., 1995, 34, 3658–3666.
53. L. J. Laughlin and D. Coucouvanis, J. Am. Chem. Soc., 1995, 117, 3118–3125.
54. S. M. Malinak, A. M. Simeonov, P. E. Mosier, C. E. McKenna and D. Coucouvanis, J. Am. Chem. Soc., 1997, 119, 1662–1667.
55. J. A. Kovacs and R. H. Holm, J. Am. Chem. Soc., 1986, 108, 340–341; J. A. Kovacs and R. H. Holm, Inorg. Chem., 1987, 26, 702–711.
56. S. M. Malinak, K. D. Demadis and D. Coucouvanis, J. Am. Chem. Soc., 1995, 117, 3126–3133.
57. K. Tanaka, M. Moriya and T. Tanaka, Chem. Lett., 1987, 373–376; K. Tanaka, M. Moriya, S. Uezumi and T. Tanaka, Inorg. Chem., 1988, 27, 137–143.
58. Y. Hozumi, Y. Imasaka, K. Tanaka and T. Tanaka, Chem. Lett., 1983, 897–900.
59. K. Tanaka, M. Tanaka and T. Tanaka, Chem. Lett., 1981, 895–898 . The catalyst used in this report is probably 2a despite description as [Mo₂Fe₆S₉(SPh)₈]³⁻.
60. K. Tanaka, M. Honjo and T. Tanaka, J. Inorg. Biochem., 1984, 22, 187–199; K. Tanaka, M. Nakamoto, Y. Tashiro and T. Tanaka, Bull. Chem. Soc. Jpn., 1985, 58, 316–321.
61. S. Kuwabata, K. Tanak and T. Tanaka, Inorg. Chem., 1986, 25, 1691–1697.
62. K. Tanaka, Y. Imasaka, M. Tanaka, M. Honjo and T. Tanaka, J. Am. Chem. Soc., 1982, 104, 4258–4260.
63. S. Kuwabata, Y. Hozumi, K. Tanaka and T. Tanaka, Chem. Lett., 1985, 401–404.
64. Y. Imasaka, K. Tanaka and T. Tanaka, Chem. Lett., 1983, 1477–1480.
65. K. Tanaka, S. Kuwabata, S. Denno and T. Tanaka, Bull. Chem. Soc. Jpn., 1989, 62, 1561–1566.
66. K. Tanaka, S. Uezumi and T. Tanaka, J. Chem. Soc., Dalton Trans., 1989, 1547–1554.
67. S. Kuwabata, S. Uezumu, K. Tanaka and T. Tanaka, J. Chem. Soc., Chem. Commun., 1986, 135–136; S. Kuwabata, S. Uezumi, K. Tanaka and T. Tanaka, Inorg. Chem., 1986, 25, 3018–3022; K. Tanaka, T. Matsui and T. Tanaka, Chem. Lett., 1989, 1827–1830; K. Tanaka, N. Komeda and T. Matsui, Inorg. Chem., 1991, 30, 3282–3288.
68. K. Tanaka, R. Wakita and T. Tanaka, Chem. Lett., 1987, 1951–1954; K. Tanaka, R. Wakita and T. Tanaka, J. Am. Chem. Soc., 1989, 111, 2428–2433.
69. K. Tanaka, T. Matsui and T. Tanaka, J. Am. Chem. Soc., 1989, 111, 3765–3767; N. Komeda, H. Nagao, T. Matsui, G. Adachi and K. Tanaka, J. Am. Chem. Soc., 1992, 114, 3625–3630.
70. H. Nagao, H. Miyamoto and K. Tanaka, Chem. Lett., 1991, 323–326.
71. T. Masumori, H. Seino, Y. Mizobe and M. Hidai, Inorg. Chem., 2000, 39, 5002–5003; H. Seino, T. Masumori, M. Hidai and Y. Mizobe, Organometallics, 2003, 22, 3424–3431.
72. I. Takei, K. Dohki, K. Kobayashi, T. Suzuki and M. Hidai, Inorg. Chem., 2005, 44, 3768–3770.
73. H. Mori, H. Seino, M. Hidai and Y. Mizobe, Angew. Chem., Int. Ed., 2007, 46, 5431–5434.
74. H. Seino, H. Mori, K. Hirata and Y. Mizobe, unpublished results.
75. T. Murata, Y. Mizobe, H. Gao, Y. Ishii, T. Wakabayashi, F. Nakano, T. Tanase, S. Yano, M. Hidai, I. Echizen, H. Nanikawa and S. Motomura, J. Am. Chem. Soc., 1994, 116, 3389–3398.
76. C. S. Bahn, A. Tan and S. Harris, Inorg. Chem., 1998, 37, 2770–2778.
77. T. Wakabayashi, Y. Ishii, T. Murata, Y. Mizobe and M. Hidai, Tetrahedron Lett., 1995, 36, 5585–5588.
78. T. Wakabayashi, Y. Ishii, K. Ishikawa and M. Hidai, Angew. Chem., Int. Ed. Engl., 1996, 35, 2123–2124.
79. I. Takei, Y. Wakebe, K. Suzuki, Y. Enta, T. Suzuki, Y. Mizobe and M. Hidai, Organometallics, 2003, 22, 4639–4641.
80. I. Takei, Y. Enta, Y. Wakebe, T. Suzuki and M. Hidai, Chem. Lett., 2006, 35, 590–4591.
81. Y. Tao, Y. Zhou, J. Qu and M. Hidai, Tetrahedron Lett., 2010, 51, 1982–1984.
82. Y. Tao, B. Wang, B. Wang, L. Qu and J. Qu, Org. Lett., 2010, 12, 2726–2729.
83. I. Takei, K. Kobayashi, K. Dohki, S. Nagao, Y. Mizobe and M. Hidai, Chem. Lett., 2007, 36, 546–547.
84. M. Nabika and M. Hidai, unpublished results.
85. H. Takahashi, T. Ando, M. Kamigaito and M. Sawamoto, Macromolecules, 1999, 32, 3820–3823.
86. M. Feliz, E. Guillamón, R. Llusar, C. Vicent, S.-E. Stiriba, J. Pérez-Prieto and M. Barberis, Chem.–Eur. J., 2006, 12, 1486–1492.
87. E. Guillamón, R. Llusar, J. Pérez-Prieto and S.-E. Stiriba, J. Organomet. Chem., 2008, 693, 1723–1727.
88. M. D. Curtis, in Transition Metal Sulfur Chemistry: Biological and Industrial Significance, ed. E. I. Stiefel and K. Matsumoto, American Chemical Sociey, Washington DC, 1996, ch. 8, pp. 154–175.
89. M. A. Mansour, M. D. Curtis and J. W. Kampf, Organometallics, 1995, 14, 5460–5462 ; 1997, 16, 3363–3370.
90. M. Taniguchi, Y. Ishii, T. Murata, T. Tatsumi and M. Hidai, J. Chem. Soc., Chem. Commun., 1995, 2533–2534; T. Tatsumi, M. Taniguchi, S. Yasuda, Y. Ishii, T. Murata and M. Hidai, Appl. Catal., A, 1996, 139, L5–L10; M. Taniguchi, D. Imamura, H. Ishige, Y. Ishii, T. Murata, M. Hidai and T. Tatsumi, J. Catal., 1999, 187, 139–150.
91. K. Herbst, M. Brorson and A. Carlsson, J. Mol. Catal. A: Chem., 2010, 325, 1–7.

---

Footnote

† In memory of Professor Yasushi Mizobe

---