Electrophilic As-functionalisation of σ-arsolido complexes

Abstract

The σ-arsolido complex [Mo(AsC₄Me₄)(CO)₃(η⁵-C₅H₅)] is alkylated at arsenic by MeOTf to afford the pentamethylarsole complex [Mo(MeAsC₄Me₄)(CO)₃(η⁵-C₅H₅)](OTf) while iodomethane affords a mixture of [Me₂AsC₄Me₄]I, [MoMe(CO)₃(η⁵-C₅H₅)], [MoI(CO)₃(η⁵-C₅H₅)] and the arsole complexes cisoid- and transoid-[MoI(MeAsC₄Me₄)(CO)₂(η⁵-C₅H₅)] and transoid-[Mo{C(=O)Me}(MeAsC₄Me₄)(CO)₂(η⁵-C₅H₅)], The arsole ligand in [Mo(MeAsC₄Me₄)(CO)₃(η⁵-C₅H₅)](OTf) is readily liberated by NaI in acetone to afford free MeAsC₄Me₄ and [MoI(CO)₃(η⁵-C₅H₅)]. In a similar manner, the reaction of [Mo(AsC₄Ph₄)(CO)₃(η⁵-C₅H₅)] with MeI affords MeAsC₄Ph₄ and [MoI(CO)₃(η⁵-C₅H₅)], while [Mo{AsC₄(SiMe₃)-2-Me₂-3,4}(CO)₃(η⁵-C₅H₅)] with MeOTf affords [Mo{MeAsC₄(SiMe₃)-2-Me₂-3,4}(CO)₃(η⁵-C₅H₅)](OTf). The reaction of [Mo(AsC₄Me₄)(CO)₃(η⁵-C₅H₅)] with activated alkynes (RC≡CR: R = CF₃, CO₂Me) does not proceed via [4 + 2] cyclo-addition but rather electrophilic attack at arsenic followed by metallacyclisation with incorporation of a carbonyl ligand in the spirocyclic complexes [Mo{As(C₄Me₄)CR=CRCO}(CO)₂(η⁵-C₅H₅)].

---

Introduction

Although a small number of σ-arsolyl (arsolido) complexes have been reported,¹ little is known concerning their reactivity, beyond a proclivity towards self-condensation to afford homo-bimetallic μ-arsolido complexes (a) or the addition of extraneous metal centres to generate heterobimetallic μ-arsolido complexes.² They appear to be plausible intermediates in the ultimate formation of pentahapto-arsolyl complexes,^1b,3 and in one case redox disproportionation of the chromium σ-arsolyl [Cr(AsC₄Ph₄)(CO)₃(η⁵-C₅H₅)] affords convenient access to a diarsolyl (AsC₄Ph₄)₂.^2c It stands to reason that by analogy with the demonstrated arsenic-centred nucleophilicity of σ-arsenido complexes L_nM–AsR₂,^4–6 σ-arsolyls should display similar arsenic-centred nucleophilicity. This would be supported by the recent computational interrogation of the complexes [Mo(AsMe₂)(CO)₃(η⁵-C₅H₅)] vs. [Mo(AsC₄H₄)(CO)₃(η⁵-C₅H₅)] that showed that the ‘lone’ pair on arsenic is most certainly available for electrophilic attack, rather than being otherwise involved in establishing a Hückel sextet for the pseudo-aromatic arsole ring (Fig. 1).^3c

Fig. 1 Valence orbitals of interest for the complex [Mo(AsC₄H₄)(CO)₃(η⁵-C₅H₅)].^3c

This has however yet to be experimentally explored beyond the simple addition of coordinatively unsaturated metal centres, e.g., ‘AuC₆F₅’, ‘Mn(CO)₂(η⁵-C₅H₄Me)’ and ‘Fe(CO)₂(η⁵-C₅H₅)⁺’ to [Mo(AsC₄Me₄)(CO)₃(η⁵-C₅H₅)] (1a).² Accordingly, we now report an investigation of the reactivity of a series of σ-arsolyl complexes of molybdenum, viz. [Mo(AsC₄Me₄)(CO)₃(η⁵-C₅H₅)] (1a), [Mo(AsC₄Ph₄)(CO)₃(η⁵-C₅H₅)] (1b) and [Mo{AsC₄H(SiMe₃)-2-Me₂-3,4}(CO)₃(η⁵-C₅H₅)] (1c) towards electrophilic alkylating agents and potentially dienophilic electrophilic alkynes.

Results and discussion

The σ-arsolyl complexes 1a–1c are now readily available via facile transmetallation between the stannyl complex [Mo(SnⁿBu₃)(CO)₃(η⁵-C₅H₅)] and the corresponding As–chloro-arsoles ClAsC₄R₄ (R = Me, Ph) or ClAsC₄(SiMe₃)₂-2,5-Me₂-3,4, the latter being accompanied by an unusual monodesilylation of one arsole ring substituent (Scheme 1).^1c

Scheme 1 Synthesis of σ-arsolyl complexes via transmetallation.^2c

The complex 1a is readily alkylated at the heteroatom by methyl trifluoromethanesulfonate (MeOTf, OTf = SO₃CF₃) in Et₂O solution at 0 °C. Bright yellow, X-ray quality crystals of the salt [Mo(MeAsC₄Me₄)(CO)₃(η⁵-C₅H₅)]OTf [2a]OTf rapidly precipitated from solution and the product was isolated in near-quantitative yield (86%, Scheme 2). Care must be taken to avoid too great an excess of MeOTf which results in decomposition to oily black residues.

Scheme 2 Monomethylation of σ-arsolyl complexes.

Spectroscopic data reflect the formal localisation of positive charge at the molybdenum centre with a shift of ν_CO absorptions to higher wavenumbers (CH₂Cl₂: 2057(vs), 1998(sh), 1965(vs) cm⁻¹), and a shift of the ¹H cyclopentadienyl resonance from 5.16 ppm (CD₂Cl₂) in 2a to higher frequency (5.54 ppm) in the cationic [2a]⁺. The As–CH₃ group resonance in the ¹H NMR spectrum (CD₂Cl₂: δ_H = 1.66) is to slightly lower frequency of the ring methyl groups (δ_H = 1.95, 1.82) which are essentially identical in chemical shift to those of the precursor 1a, being seemingly unaffected by alkylation of the arsenic. A shift to higher frequency is observed in the ¹³C{¹H} NMR spectrum for the carbonyl ligands transoid (225.8 ppm) and cisoid (223.2 ppm) to the pentamethylarsole ligand (cf. 235.5 and 224.7 ppm for 1a in CD₂Cl₂) as the efficacy of Mo(d) → CO(π*) back-bonding is reduced; the same argument applies to the cyclopentadienyl signal (93.5 vs. 94.3 ppm). The arsolyl ring-carbon nuclei are similarly displaced to slightly higher frequency by As-methylation (145.8 and 134.5 ppm vs. 149.1 and 141.8 ppm for 1a), a feature which also manifests in binuclear complexes of the μ-AsC₄Me₄ ligand^2b as the modest ring-current within the cyclic system is further diminished by appropriation of the arsenic lone pair. High-resolution mass spectrometry (ESI) reveals extensive fragmentation of the molecular ion, the primary mode of disintegration being loss of the arsenic-bound CH₃ group in addition to sequential products of decarbonylation; single crystals suitable for X-ray diffractometry were collected from the reaction mixture and the molecular structure is shown in Fig. 2.

Fig. 2 The molecular structure of [2a]⁺ in a crystal of [2a]OTf (50% displacement ellipsoids, hydrogen atoms and triflate anion omitted).

Crystallographic inspection of the reaction product confirms As-methylation. Reflecting the formal positive charge on the Mo(II) atom and the change in bonding description (X to L⁷) between the metal and pnictogen, a contraction of the Mo–As distance by almost 0.15 Å to 2.5844(4) Å is noted (cf. 2.7267(7) Å for 1a). A similar contraction (though of lesser magnitude) is observed upon coordination of Lewis acidic metal centres to 1a.^2b Apparently, sequestration of the arsenic lone pair alleviates a repulsive interaction with the electronically saturated molybdenum centre, permitting closer approach of the two centres – the so-called transition metal gauche effect described by Gladysz.⁸ The arsenic atom adopts a deformed tetrahedral geometry with external bonding angles of 105–117° and an internal C–As–C ring-angle of ca. 89°. The pentamethylarsole ligand is swivelled about the Mo–As bond to lie in an unsymmetrical disposition about the molecule, similar to that found in 1b,²² at the expense of the intramolecular C–H⋯π contact found in the solid state structure of 1a. Interatomic distances within the arsole ring are reflective of decreased aromatic character with a slight compression of the C_α=C_β bonds (mean 1.327(5) Å) and elongation of the C_β–C_β bond (1.496(5) Å) compared to the neutral arsolyl complex (1.351(7) and 1.469(8) Å, respectively), though these variations lie at the precision limits of the data.

Alkylation of 1c was explored similarly: treatment with MeOTf in Et₂O solution causes the product [2c]OTf to rapidly precipitate from solution in excellent yield (81%) as bright yellow crystals suitable for X-ray diffraction. Spectroscopic features are comparable to those of the pentamethylarsole complex [2a]⁺ and accordingly, only salient features are noted. The solution IR spectrum reveals the usual trio of carbonyl vibrations (CH₂Cl₂: 2065(vs), 1992(sh), 1971(vs) cm⁻¹) shifted to higher wavenumbers from 1c (2007(vs), 1938(sh), 1922(vs) cm⁻¹), and the resonance for the olefinic hydrogen α to the arsenic atom appears at 6.97 ppm in CD₂Cl₂ solution. The cyclopentadienyl ligand resonances appear at 5.82 and 92.4 ppm in the ¹H and ¹³C{¹H} NMR spectra, respectively, and the carbonyl ligands at 235.7 (transoid), 225.6 and 223.6 ppm (diastereotopic cisoid) in the latter. The four chemically inequivalent arsole ring-carbon nuclei are found at 158.4, 157.1, 149.7 and 146.7 ppm with the resonance at δ_C = 149.7 corresponding to the methylidyne carbon as indicated by both HSQC NMR measurements, and by nOe enhancement. The molecular structure was determined by X-ray diffractometry and is shown in Fig. 3.

Fig. 3 The molecular structure of [2c]⁺ in a crystal of [2c]OTf (50% displacement ellipsoids, Si-methyls simplified and most hydrogen atoms and triflate anion omitted). Inset: Space-filling representation.

The molecular structure of [2c]⁺ in [2c]OTf is largely similar to that of [2a]⁺. The arsole ligand lies “sideways” against the Mo(CO)₃(η⁵-C₅H₅) centre with the SiMe₃ group occupying the region of space between the cisoid carbonyl ligands. The As→Mo dative bond length of 2.6016(6) Å is slightly longer than that in [2a]⁺ (2.5844(4) Å), though it is noted that the Mo–As covalent bond in 1c is also ca. 0.02 Å longer than in [2a]⁺ no doubt due to the intrinsic proximal steric influence of the SiMe₃ residue. Dimensions of the arsole ring follow the same pattern as for [2a]⁺ with a compression of the inequivalent C_α=C_β′ bonds (1.326(6), 1.342(6) Å) and elongation of the C_β–C_β bond (1.489(6) Å) upon methylation of the arsenic atom, compared to neutral 1c.

The tetraphenylarsolyl complex 1b undergoes alkylation with reluctance. Treatment with MeOTf in toluene solution gives no immediate precipitation of a salt, and only upon standing for several days at ambient temperature do a very small quantity of yellow crystals form which were regrettably unsuitable for X-ray diffractometry; accompanying these crystals were a larger quantity of oily brown residues. Dissolution of the isolated crystals precipitated more of these insipid residues and MeAsC₄Ph₄ (3, Fig. 5) was the only identifiable species. The same outcome was obtained in CH₂Cl₂ solution which resulted in 3 being the only isolable product. The observation that tetraphenylarsoles are generally hesitant to quarternisation suggests a very poorly nucleophilic arsenic centre, in line with observations to follow involving methyl iodide as the alkylating agent. Geometrical minimisation at the ωB97X-D/6-31G*/LANL2Dζ level of density functional theory^2c returns rather similar natural charges (+0.678 a.u. for 1a; +0.677 a.u. for 1b) and negligible atom-condensed Fukui functions (f⁺ = 0.005 for 1a; 0.002 for 1b). The latter are somewhat misleading because for 1a the arsenic lone pair lies only 0.1 eV below the HOMO which is associated with the butadiene component of the arsolyl group. For 1b, the energy gap is larger (0.4 eV) and the HOMO is delocalised out to the 2,5-phenyl substituents. This suggests that the low nucleophilicity of 1b is primarily due to the steric bulk surrounding the arsenic (Fig. 4) rather than any significant electronic impact. The tetraphenylarsolyl complex 1b is, however, cleaved by an excess of CH₃I in solution overnight to provide [MoI(CO)₃(η⁵-C₅H₅)] (¹H NMR (C₆D₆): δ 4.44 ppm. IR (CH₂Cl₂): 2043(vs), 1961(vs) cm⁻¹) and 1-methyl-2,3,4,5-tetraphenylarsole (3) as the only products by NMR spectroscopy. Repeating the reaction in CH₂Cl₂ on a preparative scale allowed isolation of the dark red molybdenum iodide complex (83%) and the light yellow MeAsC₄Ph₄ arsole (3, 70%, Fig. 5) in high yields.

Fig. 4 Corey–Pauling–Kolton models for (a) [Mo(AsC₄Me₄)(CO)₃(η⁵-C₅H₅)] and (b) [Mo(AsC₄Ph₄)(CO)₃(η⁵-C₅H₅)].

Fig. 5 The molecular structure of MeAsC₄Ph₄ (3) in a crystal (50% displacement ellipsoids, phenyl groups simplified and hydrogen atoms omitted for clarity). The As–Me moiety is disordered over two positions and only the major contributor is shown.

The reaction almost certainly proceeds via initial methylation of the arsenic atom (which may or may not follow a radical pathway, vide infra) followed by rapid and irreversible nucleophilic substitution of the arsole ligand from the [Mo(MeAsC₄Ph₄)(CO)₃(η⁵-C₅H₅)]⁺ [2b]⁺ complex by the iodide anion. This process is presumably dissociative in nature since the Mo(II) centre is both electronically-saturated and seven-coordinate, and the displacement of the arsole ligand from the cationic complex must be at least as rapid as its formation since at no point during the reaction were any other cyclopentadienyl signals observed in the ¹H NMR spectrum (Scheme 3).

Scheme 3 Attempted methylation of a tetraphenylarsolyl complex.

The more nucleophilic 1a also reacts with excess CH₃I in C₆D₆ solution, though the reaction was considerably more complex than anticipated (Scheme 4). Multiple products are generated, including large colourless crystals of [Me₂AsC₄Me₄]I [4]I which separate in 62% yield from solution. This salt is the first example of an authentic cationic arsole (henceforth arsolylium), although there are examples of annulated arsoles undergoing quaternisation.⁹ These are excluded from the present discussion given the ring system is not based on a localised butadiene moiety, one of the major identifying features of arsoles. The salt [4]I was structurally characterised and the molecular geometry of the cation is shown in Fig. 6 in addition to the unit cell packing which indicates weak hydrogen bonding (≈3.0 Å) between the iodide and protons on four methyl groups.

Scheme 4 The reaction of 1a with excess iodomethane.

Fig. 6 The molecular structure of [4]I in a crystal (50% displacement ellipsoids, methyl groups simplified and hydrogen atoms omitted). Inset: The unit cell contents (P2₁/c) viewed along the crystallographic b axis highlighting extended H⋯I⁻ interactions.

The arsolylium salt [4]I was isolated from the reaction mixture in good yield and was fully characterised. The internal symmetry of the cation is reflected in the ¹H NMR spectrum (CDCl₃) which finds only three equally integrating signals at 2.58, 2.32 and 2.00 ppm of which the latter is assigned to the arsenic-bound methyl groups by comparison with the As–methyl resonance in [2a]OTf. In the ¹³C{¹H} NMR spectrum the arsolyl ring-carbon nuclei manifest at 150.3 and 123.5 ppm (α and β to the heteroatom, respectively) with three methyl environments at 15.0, 14.2 and 9.2 ppm. The latter is identified as the arsenic-bound methyl carbon atoms on the basis of HSQC and HMBC NMR experiments. High-resolution mass spectrometry reveals the molecular ion in high intensity (m/z ≈ 213.0628). The formation of [4]I was confirmed in a separate experiment to arise from simple alkylation of MeAsC₄Me₄ by methyl iodide. Methyl triflate will also alkylate pentamethylarsole in Et₂O at 0 °C, though the [4]OTf salt formed degraded over several weeks in the solid state (under argon) to oily brown residues. Conversely, the iodide salt retained its integrity for >2 years. Other methylating agents such as [Me₃O]BF₄ were not tested. Methyltetraphenylarsole MeAsC₄Ph₄ (3) did not undergo such a reaction, recalling the resistance of 1b to alkylation. A reasonable course of events involves initial alkylation of the arsenic in [Mo(AsC₄Me₄)(CO)₃(η⁵-C₅H₅)] to form a salt [2a]I, from which dissociation of the arsole is followed by irreversible coordination of the iodide counter anion.

Hexamethylarsolylium iodide [4]I crystallises (monoclinic P2₁/c) with four separated cation–anion pairs contained in the unit cell. Incidentally, the iodide anions have distorted bisphenoidal coordination geometry (τ₄ = 0.86, τ′₄ = 0.57) in the solid-state since only four hydrogen bonds operate (per iodide) with approximate valence angles of ca. 165° (pseudo-trans axial) and 82° (pseudo-equatorial). The [4]⁺ cation is approximately though not crystallographically laterally-symmetric (C_2v), and the quaternary arsenic atom exists in a distorted tetrahedral environment with external (local) bonding angles of 109–118°, though the internal C_α–As–C_α vertex is regular within this work at 88°. Compared to other RAsC₄Me₄ arsoles which have been structurally characterised,¹⁰ interatomic distances within the [Me₂AsC₄Me₄]⁺ cation reflect a true 2,4-diene localisation of the π-electron density about the ring with a contraction of the As–C_α (mean 1.903(3) Å) and C_α=C_β bonds (mean 1.327(4) Å), and an elongation of the C_β–C_β′ bond (1.500(3) Å). That said, exploration of the molecular orbital manifold for the model cation [Me₂AsC₄H₄]⁺ (DFT:ωB97X-D/6-31G*, Fig. 7) does suggest a modest contribution from arsenic to the π-system (HOMO−1, HOMO−4). It is also noteworthy that rather than the arsenic, as is typical for conventional arsonium cations, C_α and C_α′ would be expected to be the preferred sites for frontier orbital-controlled nucleophilic attack considering the topologies of the LUMO, as well as the atom-condensed Fukui functions (f⁻)¹¹ for arsenic (f⁻ = 0.024) and Cα/α′ (f⁻ = 0.184). This is despite the arsenic carrying a substantial natural positive charge (+1.63 a.u.). Such processes would be expected to yield dimethyl(butadienyl)arsines (Me₂AsCH=CHCH=CH–Nu), however this propensity would be somewhat curtailed in the real cases of arsolyliums bearing C_α-substituents, as most do.

Fig. 7 Frontier orbitals of interest for the hypothetical dimethylarsolylium cation [Me₂AsC₄H₄]⁺ (DFT:ωB97X-D/6-31G*/gas phase).

As an aside, the colourless arsolylium [4]I did appear to react with LiNⁱPr₂ in Et₂O solution at −78 °C to provide a cherry-red solution, presumed to contain an arsenic ylide H₂CAsMeC₄Me₄ (by analogy with H₂CAsPh₃ ¹²) which, however, decomposed to poorly-soluble light brown materials upon warming. The addition of labile [W(THF)(CO)₅] or [Au(C₆F₅)(THT)] (THT = tetrahydrothiophene) to this red solution failed to afford isolable ylide complexes upon workup, and no trapping experiments with organic electrophiles (e.g., ketones) were attempted due to limited quantities of the arsolylium on-hand.

The remainder of the C₆D₆ reaction solution contained the expected [MoI(CO)₃(η⁵-C₅H₅)] as the major cyclopentadienyl-containing species. Interestingly, [MoMe(CO)₃(η⁵-C₅H₅)] is also observed, which speaks to the possible occurrence of radical reaction pathways, since the addition of NaI to [2a]OTf gave [MoI(CO)₃(η⁵-C₅H₅)] (¹H NMR (acetone-d₆): δ_H = 5.88 ppm cf. an authentic sample) and MeAsC₄Me₄exclusively; the presence of a fleeting pentavalent As(V) species i.e., [Mo{AsMe(I)C₄Me₄}(CO)₃(η⁵-C₅H₅)] which then dismutates to [MoI(CO)₃(η⁵-C₅H₅)] and [MoMe(CO)₃(η⁵-C₅H₅)], is therefore discounted. Pentavalent (λ⁵) arsoles¹³ are a subject to which we will return in detail in a subsequent paper. Repeating the reaction in n-hexane solution followed by storage at −30 °C provides an abundance of crystalline specimens comprising no fewer than five molybdenum-containing compounds which were identified by X-ray diffraction. In addition to [4]I (Fig. 6), they are: (a) [MoI(CO)₃(η⁵-C₅H₅)] (Fig. 8), which despite being a septuagenarian compound¹⁴ had not been previously structurally characterised; (b) [MoMe(CO)₃(η⁵-C₅H₅)], a similarly venerated organometallic,¹⁴ (c) cisoid-[MoI(MeAsC₄Me₄)(CO)₂(η⁵-C₅H₅)] (cisoid-5, Fig. 8a), arising from carbonyl rather than arsole substitution, and (d) its isomer transoid-[MoI(MeAsC₄Me₄)(CO)₂(η⁵-C₅H₅)] (transoid-5, Fig. 9b) as a co-crystallite with (e) the novel acyl complex transoid-[Mo{C(O)Me}(MeAsC₄Me₄)(CO)₂(η⁵-C₅H₅)] (6, Fig. 10).

Fig. 8 Molecular structure of [MoI(CO)₃(η⁵-C₅H₅)] in a crystal (50% displacement ellipsoids, most hydrogen atoms omitted).

Fig. 9 Isomeric pentamethylarsole complexes. Molecular structures of (a) cisoid-[MoI(MeAsC₄Me₄)(CO)₂(η⁵-C₅H₅)] (cisoid-5) in a crystal (one of two crystallographically unique molecules) and (b) transoid-[MoI(MeAsC₄Me₄)(CO)₂(η⁵-C₅H₅)] (transoid-5) in a co-crystallite of (transoid-5)_0.11(6)_0.89. (50% displacement ellipsoids, methyl groups simplified and most hydrogen atoms omitted).

Fig. 10 Molecular structure of trans-[Mo{C(=O)Me}(MeAsC₄Me₄)(CO)₂(η⁵-C₅H₅)] (6) in a co-crystallite of (trans-5)_0.11(6)_0.89. (50% displacement ellipsoids, arsole methyl groups simplified and most hydrogen atoms omitted).

Despite being first reported in 1956 by Wilkinson,¹⁴ and having been utilised as a synthon in a variety of settings in the interregnum,¹⁵ the crystal structure of red [MoI(CO)₃(η⁵-C₅H₅)] would appear to have been neglected. It transpires however that the structure had in fact been previously determined, but incorrectly identified as the diamagnetic (sic., d³) compound [Mo(Te)(CO)₃(η⁵-C₅H₅)] (CCDC 216750†). This was purported to arise from the reaction of [Mo₂(CO)₆(η⁵-C₅H₅)₂] with diphenylditelluride and [ⁿBu₄N]I,^16a but is clearly [MoI(CO)₃(η⁵-C₅H₅)]. Understandably, molecular features are rather unexciting given the simplistic nature of the complex and only salient crystallographic aspects are noted here. A seven-coordinate molybdenum(II) centre is found which approximates a distorted square-based pyramid, where the three carbonyls and the iodide ligand comprise the basal plane and the centroid of the cyclopentadienyl ring occupies the apical site. The Mo–I bond length of 2.8419(6) Å is typical among similar complexes,¹⁶ and the Mo–C bond transoid to the iodide (1.992(6) Å) is marginally shorter than the pair cisoid (mean 2.035(7) Å) due to the π-donor character of the halide. The unit cell (monoclinic P2₁/n) contains four symmetry-related molecules which display two intermolecular H⋯I interactions (per molecule) of ca. 3.1 Å in length that likely assist with relative orientations during lattice formation. The yellow organometallic complex [MoMe(CO)₃(η⁵-C₅H₅)] first prepared by Wilkinson^14a was structurally characterised much later by Valente and co-workers^14b who sought to utilise it for the catalytic epoxidation of cyclic alkenes in the presence of tert-butylperoxide. The Mo–C(sp³) distance was 2.326(3) Å and no unusual intermolecular interactions were found.

Dark red cisoid-[MoI(MeAsC₄Me₄)(CO)₂(η⁵-C₅H₅)] (cisoid-5, Fig. 9a) crystallises in triclinic P1̄ and the unit cell contains four molecules comprising two pairs of enantiomers, which differ in the relative position of the arsole ligand to the iodide (either to the “left” or “right”) when viewed along the molybdenum-cyclopentadienyl centroid axis. The pair of molecules within the asymmetric unit are crystallographically unique though interatomic distances and angles are statistically indistinguishable between the two entities. The molybdenum atom adopts the usual seven-coordinate geometry which approximates a square-based pyramid. The As→Mo dative bond measures 2.5640(6) (molecule A) and 2.5717(6) (B) Å and the Mo–I covalent bonds are unremarkable at 2.8524(4) (A) and 2.8442(4) (B) Å in length. The pentamethylarsole ligand lies approximately orthogonal to the vertical plane bisecting the Mo(CO)₃(η⁵-C₅H₅) moiety. Distances between arsole ring atoms are typical of σ-dative coordination complexes for this ligand (which are measurably different from [4]I above), with mean As–C_α, C_α=C_β and C_β–C_β′ bond lengths of ca. 1.93, 1.35 and 1.50 Å, respectively.

The third set of red crystals contained two compounds in an unequal occupancy ratio which are apparently (or approximately) isosteric and superimposed within the lattice. They are transoid-[MoI(MeAsC₄Me₄)(CO)₂(η⁵-C₅H₅)] (transoid-5, minor component, 11% occupancy) and transoid-[Mo{C(O)Me}(MeAsC₄Me₄)(CO)₂(η⁵-C₅H₅)] (transoid-6, major component, 89% occupancy). Refinement (including occupancy) led to a satisfactory structural model solution (R₁ = 0.0398), albeit with some eccentric displacement ellipsoids though no restraints/constraints to correct these were applied. The iodide complex is again presumed to arise from carbonyl rather than arsole substitution, and the simplest explanation for the occurrence of the acyl complex would superficially seem to involve migratory insertion of [MoMe(CO)₃(η⁵-C₅H₅)] (vide infra) trapped by free pentamethylarsole.

The two compounds co-crystallise in monoclinic P2₁/n and in both species the pentamethylarsole ligand approximately straddles the (non-crystallographic) vertical symmetry plane of the Mo(CO)₂(η⁵-C₅H₅) centre with the As–methyl substituent projected between the pair of cisoid carbonyl ligands. Since the two complexes are superimposed within the crystal lattice, the positions of the Mo(MeAsC₄Me₄)(CO)₂(η⁵-C₅H₅) atoms are modelled as a weighted average of both and all interatomic dimensions therein are thus identical viz. the mean As→Mo dative bond for both complexes in this particular crystal is 2.5226(6) Å. All crystallographic uniqueness is observed in the region of the overlaid acyl and iodide substituents, with Mo–C(sp²), C=O and Mo–I distances of 2.230(9), 1.215(9) and 2.898(4) Å, respectively. The former two are typical¹⁷ whereas the latter is elongated compared the other Mo–I bond lengths observed previously (ca. 2.4–2.5 Å) since the presence of the arsenic atom in a transoid position is presumably a repulsive influence. It is noted that a transoid configuration appears to be the preferred condition for Group 6 acyl complexes of the form [M{C(=O)R}L(CO)₂(η⁵-C₅H₅)]¹⁸ with which [Mo{C(O)Me}(MeAsC₄Me₄)(CO)₂(η⁵-C₅H₅)] also complies. For [MoIL(CO)₂(η⁵-C₅H₅)] it is reasonable to expect that, in general, ligands of superior π-acidity would favour the coordination site transoid to the iodide though pentamethylarsole is not expected to be particularly outstanding in this respect, possibly explaining why both cisoid and transoid isomers have crystallised. No general trend emerges from a survey of the literature and any regio preferences appear to be highly contextual e.g., steric demands.¹⁹ In many cases the cisoid and transoid isomers interconvert depending on inter alia temperature and solvent polarity^15f,20 and in others it is not possible to unambiguously distinguish isomers from the spectroscopic data reported.

Acyl complexes of the mid-transition metals are well studied: their deliberate method of synthesis entails treatment of a metal carbonyl anion with an acyl halide or by migratory insertion of an alkyl complex,²¹ often in the presence of a Lewis base^18a,b,20,22 although some Lewis acids have also been found to promote this process.²³ For acyl complexes prepared from the corresponding alkyl–carbonyl in the presence of a neutral two-electron ligand, the extraneous ligand is usually of reasonable basicity (e.g., isocyanides, phosphines or phosphites) and weaker nucleophiles (amines, nitriles, arsines, stibines etc.) either provide the product in low yield²⁴ or afford products of decarbonylation, especially for Group 6 centres. It is therefore intriguing that an acyl complex possessing an arsole ligand was observed, since (a) arsines appear less inclined to affect migratory insertion reactions upon exposure to [MoR(CO)₃(η⁵-C₅H₅)],²⁵ (b) the reaction was carried out at −30 °C whereas thermal impetus (generally ≫25 °C) is often required, and (c) the reaction was carried out in n-hexane medium, which is considered to disfavour the migration process, being unable to effectively stabilise the coordinatively unsaturated and dipolar transition state.^22a Attempts were made to deliberately synthesise [Mo{C(O)Me}(MeAsC₄Me₄)(CO)₂(η⁵-C₅H₅)] (6) by heating [MoMe(CO)₃(η⁵-C₅H₅)] to 80 °C in CD₃CN with one equivalent of MeAsC₄Me₄. The metal alkyl signals were unaffected and only gradual decomposition of the arsole was observed; a similar outcome was obtained with PhAsC₄Me₄, an arsole of superior thermal stability, wherein the spectrum was totally unchanged after heating. It is also noted that these arsoles did not proceed to carbonyl substitution products either (i.e., cisoid- or transoid-[MoMe(RAsC₄Me₄)(CO)₂(η⁵-C₅H₅)], R = Me, Ph), speaking to their poor nucleophilicity. The presence of 6 among the crystalline products therefore remains inexplicable since the reaction conditions would be expected to disfavour formation of such a compound (vide supra); like the occurrence of [MoMe(CO)₃(η⁵-C₅H₅)] in the mixture, it is possible that radical processes may be operative though not via any photochemical route since this reaction was performed in the dark.

Reactions with electrophilic alkynes

Both metal-complexed phosphole and arsole ligands are known to undergo [4 + 2] cycloadditions upon treatment with dienophiles.^26–28 For the former, this led to the formation of stable 9-phosphanorbornene and -norbornadiene complexes en route to the first examples of saturated phosphirane (RPC₂R′₄) and unsaturated phosphirene (RPC₂R′₂) complexes.²⁶ Voβ and Schenk treated the Group 6 pentacarbonyl arsole complexes [M(RAsC₄Me₄)(CO)₅] (M = Cr, W; R = Me, ^tBu, Ph) with the electrophilic alkyne dimethylacetylene dicarboxylate (DMAD, MeO₂CC≡CCO₂Me) though only the corresponding tetramethyl-ortho-phthalate dimethylester was isolated and no 9-arsanorbornadiene complex was obtained.²⁷ Leung has, however, demonstrated the enantioselective [4 + 2] cyclo-additions of alkenes to palladium-coordinated arsoles.²⁸ It was therefore of interest whether arsoles bearing metal substituents might enter into pseudo-Diels–Alder processes with activated alkynes.

The nucleophilic arsolyl 1a reacts rapidly with a slight excess of DMAD in CH₂Cl₂ solution at ambient temperature and complete consumption of 1a occurs within only a few minutes of mixing as confirmed by IR spectroscopy. Purification via column chromatography on Florisil® provided an air-stable, reddish-brown solid as the only isolable compound in good yield (85%). High-resolution mass spectrometry confirmed the formation of a 1 : 1 1a/DMAD cycloadduct (m/z ≈ 572.9797) though both the ¹H NMR (C₆D₆) and IR data were not consistent with a C_s-symmetric [4 + 2] cycloaddition 9-arsanorbornadien-9-yl product of the form [Mo{AsC₆Me₄(CO₂Me)₂}(CO)₃(η⁵-C₅H₅)] (Scheme 5).

Scheme 5 [3 + 2]-cycloaddition of a σ-arsolyl complex 1a with activated alkynes (R = CF₃, CO₂Me).

Neither the arsolyl nor alkyne moieties eventuate in symmetrical environments: a pair of methyl ester (3.45, 3.23 ppm, relative integration 3 : 3) and trio of methyl (1.88, 1.59 and 1.54 ppm, relative integration 3 : 3 : 6) signals in the ¹H NMR spectrum, in addition to three equal intensity carbonyl resonances (δ_C = 263.2, 245.5 and 234.7) indicate the molecule has no element of symmetry. The 9-arsanobornadienyl formulation would allow a time-averaged molecular plane of symmetry on the ¹³C NMR timescale through rotation around the Mo–As bond and may therefore be discarded. Furthermore, the solution IR spectrum (CH₂Cl₂) exhibits a ν_CO intensity profile that is not suggestive of a C_s-symmetric Mo(CO)₃(η⁵-C₅H₅) centre. Instead, only two terminal carbonyl modes are observed at 1962(vs) and 1889(s) cm⁻¹ in addition to a lower-energy absorption at 1732(s) cm⁻¹ indicative of a ketonic carbonyl. A pair of medium-intensity methyl ester carbonyl absorptions are found at 1605 and 1580 cm⁻¹.

The compound crystallised from Et₂O at −20 °C and X-ray diffraction revealed its identity as a novel spirocyclic 3-(arsolyl)propenoyl chelate complex [Mo{C(O)CR=CRAsC₄Me₄}(CO)₂(η⁵-C₅H₅)] (R = CO₂Me, 7a). The Z-propenoyl linkage arises from insertion of a DMAD molecule into the region of space between the nucleophilic arsenic atom and an electrophilic cisoid carbonyl ligand, forming a new five-membered molybda-arsacyclopentenone (Scheme 5).

The molecular structure is shown in Fig. 11, and a proposed mechanism of formation is given in Scheme 4. Consistent with the nucleophilicity of the arsenic centre discussed above, the mechanism commences with DMAD acting as an electrophile towards arsenic, rather than as a dienophile to the arsolyl butadiene component. It should be noted that a not dissimilar process has been suggested to occur in the reaction of triphenylarsine with DMAD, for which a somewhat questionable and yet to be substantiated, λ⁵-arsenole Ph₃AsC₄(CO₂Me)₄ has been proposed to arise.^29,30

Fig. 11 The molecular structure of one enantiomer 7a in a crystal (50% displacement ellipsoids, methyl and ester groups simplified and hydrogen atoms omitted, alternative enantiomer generated by P1̄ crystallographic symmetry). Inset: Alternative perspective of the molecule.

The cycloadduct crystallises in triclinic P1̄ and a pair of symmetry-related enantiomeric molecules (Fig. 11) are found in the unit cell. The net result of the reaction with DMAD is the formation of a new Mo–C covalent bond and conversion of the Mo–As bond to a As→Mo dative interaction. Thus, the molecular structure of the spirocyclic product features a pair of vertex-sharing pentagons united at the deformed tetrahedral arsenic atom – their angle of intersection (defined by the separate mean planes of each five-membered ring) is nearly perpendicular at 87°. This indicative geometric parameter comes with some imprecision as only the arsole ring is approximately planar (envelope angle ca. 7°) whereas the molybda-arsacyclopentenone is puckered at the arsenic and acyl carbon atoms by ca. 20°. Reflecting the fact that all interatomic bonds within the ring are unique, the dimensions of the pentagonal metallacyclic system are expectedly distorted from the ideal such that the molybdenum atom is skewed toward the acyl carbon by over 0.25 Å (Mo–C 2.244(2) vs. As→Mo 2.5110(2) Å). The latter compares to the marginally longer As→Mo dative interaction of 2.5844(4) Å found in [2a]⁺ (Fig. 2). Ring-atom distances within the arsolylpropenoyl chelate are typical of the individual bonding partners, comprising regular localised C=O, C–C, C=C and C–As covalent bonds. Reflecting the appreciable adjustment to bonding patterns surrounding the Mo(II) centre, the ¹³C{¹H} NMR spectrum (C₆D₆) of 7a reveals two resonances for the diastereotopic terminal carbonyl ligands, to low frequency of 1a (263.2, 245.5 ppm) with the acyl carbonyl resonance at 234.7 ppm. The chemically unique methyl ester carbon signals appear at 167.1 and 159.9 ppm to slightly higher frequency of the propenoyl olefin carbon atoms at 161.0 (acyl-bound) and 149.7 (arsenic-bound) ppm. The arsole ring atoms are all inequivalent, manifest as four distinct signals in the range 146.5–132.9 ppm while the remaining cyclopentadienyl and methyl resonances are unremarkable.

Given the formation of 7a from 1a and DMAD, the generality of the reaction with respect to alkyne functionalisation was explored. Combining 1a with hexafluorobut-2-yne (HFB, F₃CC≡CCF₃) in CH₂Cl₂ solution at −78 °C followed by warming to ambient temperature was accompanied by a colour change from orange to a deep red, and purification by column chromatography on Florosil® provided the second example of a 3-(arsolyl)propenoyl species, viz. [Mo{C(=O)CR=CRAsC₄Me₄}(CO)₂(η⁵-C₅H₅)] (R = CF₃7b), in high yield (90%). Alkyne insertion is confirmed by comparison of IR, NMR and MS data with those for 7a in addition to the ¹⁹F{¹H} NMR spectrum (C₆D₆) which confirmed the two chemically unique CF₃ groups (δ_F = –56.7, −59.1 ppm, ⁵J_FF = 10.2 Hz). Single crystals of 7b were grown by evaporation of a Et₂O solution at 0 °C and the molecular structure, determined by X-ray diffraction, is shown in Fig. 12.

Fig. 12 The molecular structure of one enantiomer of 7b in a crystal (50% displacement ellipsoids, methyl groups simplified and hydrogen atoms omitted, alternative enantiomer generated by crystallographic P2₁/n symmetry).

The complex crystallises in centrosymmetric monoclinic P2₁/n and the centric unit cell therefore contains both enantiomers, though only one is shown in Fig. 12. The gross molecular configuration of 7b is analogous to that of 7a (Fig. 10) and most salient features of the latter are applicable to the former. Thus, the formal [3 + 2] addition of one HFB molecule to the region between the arsenic atom and a cisoid carbonyl provides a new five-membered ring encompassing the arsolylpropenoyl chelate ligand. The respective As→Mo and Mo–C bonds measure 2.5161(5) and 2.216(3) Å, respectively, and the mean planes of the vertex sharing five-membered rings intersect at an angle more acute than in the previously discussed DMAD [2 + 3] cycloadduct, approximately 78°, though a similar puckering of the molybda-arsacyclopentenone ring (ca. 17°) is measured.

The bright yellow/orange colours of 7a and 7b may be traced to primarily MLCT transitions from metal-based orbitals (HOMO−0,1,2) to unoccupied orbitals located primarily on the metallacycle (LUMO), the arsole (LUMO+1) and carbonyl ligands (LUMO+2) on the basis of computational analysis (TD-DFT:ωB97X-D/6-31G*/LANL2Dζ/gas-phase) of 7b (see ESI†).

In contrast to the facile replacement of pentamethylarsole by iodide from [2a]⁺, the arsolylpropenoyl complexes were surprisingly resistant toward displacement of the arsenic donor: no reaction whatsoever was observed upon treatment of 7a with an excess of PMe₃, P(OMe)₃ or ^tBuNC at elevated temperatures (C₆D₆, 80 °C). The NMR spectrum of 7b was unchanged after being sealed under an atmosphere of carbon monoxide for one week, and the addition of [NO]PF₆ to a CD₃CN solution of the same led to complete decomposition with gas evolution. No reaction was observed between 1a and electron rich alkynes such as 2-butyne and diphenylacetylene, nor between [2a]⁺ or [MoAu(μ-AsC₄Me₄)(C₆F₅)(CO)₂(η⁵-C₅H₅)] (from 1a and [Au(C₆F₅)(THT)]^2b) and DMAD. The addition of other typically dienophilic substrates such as ortho-benzyne (via the Kobayashi protocol³¹), tetracyanoethylene and N-phenylmaleimide did not afford similar spirocyclic cycloadducts (by mass spectrometry) and instead yielded chromatographically immobile compounds, except for the former which afforded a very small quantity of [Mo₂(CO)₆(η⁵-C₅H₅)₂]. The course of these latter reactions was not reliably deduced from spectroscopic data.

In light of the above reactions between 1a and DMAD or HFB, it is unsurprising that other transition metal arsenidos undergo analogous ring forming process with electron-poor alkynes: Davidson et al., have treated the cacodyl complexes [M(AsMe₂)(CO)_n(η⁵-C₅H₅)] (M = Fe, n = 2; M = W, n = 3) with the same alkynes to produce the metallacycles [M{C(O)CR=CRAsMe₂}(CO)_n−1(η⁵-C₅H₅)]³² and in this regard the behaviour of the AsC₄Me₄ ligand resembles that of a conventional arsenido ligand despite the nucleophilic diene function in combination with the demonstrated dienophilic proclivities of DMAD and HFB. Notably, a large excess of the alkyne was employed yet the arsole butadiene remained intact, even when the arsenic lone pair of electrons is already sequestered in As:→M bonding to reinforce the diene (cf. aromatic elementole) character. Mathey has noted that the disubstituted phospholyl complexes [W(PC₄R₂R′₂)(CO)₃(η⁵-C₅H₅)] (R = H, R′ = Me; R = Ph, R′ = H) are also reluctant to undergo [4 + 2] cycloaddition reactions with dienophiles (despite the same phosphole rings entering into cycloaddition reactions when coordinated to W(CO)₅ ³³), instead yielding phospholylpropenoyl complexes.³⁴ Cullen has reported an arsinopropenoyl complex of rhenium(I) obtained by photolysis of [Re₂(CO)₁₀] with the bis(arsino)butene Z-Me₂AsC(CF₃)=C(CF₃)AsMe₂, though the low yields discourage mechanistic conjecture.³⁵

Conclusions

The susceptibility of σ-arsolyl ligands towards electrophilic attack at arsenic has been demonstrated, recalling similar behaviour by more conventional arsolyl ligands devoid of the unsaturated elementole ring. This adds to the already documented propensity of σ-arsolyls to coordinated to extraneous Lewis-acidic metal centres.^2b Even reactions that might be expected to involve the 2,4-diene unit, viz. pseudo-Diels–Alder [4 + 2] cycloadditions either fail or alternatively proceed again via direct electrophilic attack at the arsenic. It remains to be seen if this is a specific feature of the ‘Mo(CO)₃(η⁵-C₅H₅)’ metal centre or arsolyl substitution pattern chosen for study.

Experimental

Experimental protocols, instrumentation, crystallographic and computational methods are provided in the accompanying ESI.†

Author contributions

RMK was responsible for the design and execution of the experimental research, the acquisition and critical analysis of the characterisation data and compilation of the original draft. AFH was responsible for funding acquisition, project conceptualisation and administration, validation and refinements to the manuscript.

Data availability

The data supporting this article have been included as part of the ESI.†

Crystallographic data have been deposited at the CCDC under accession numbers 2145381–2145383, 2149526, 2145364, 2145367, 2145351 and 2145459 and can be obtained from https://www.ccdc.cam.ac.uk.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

We gratefully acknowledge the Australian Research Council (DP200101222 and DP230199215) for funding.

References

1. (a) E. W. Abel and C. Towers, J. Chem. Soc., Dalton Trans., 1979, 814–819; (b) E. W. Abel, I. W. Nowell, A. G. J. Modinos and C. Towers, J. Chem. Soc., Chem. Commun., 1973, 258–259; (c) G. Thiollet, F. Mathey and R. Poilblanc, Inorg. Chim. Acta, 1979, 32, L67–L68.
2. (a) R. M. Kirk and A. F. Hill, Dalton Trans., 2023, 52, 13235–13243; (b) R. M. Kirk and A. F. Hill, Dalton Trans., 2023, 52, 10190–10196; (c) R. M. Kirk and A. F. Hill, Dalton Trans., 2024, 53 10.1039/d4dt01308e , accepted.
3. (a) A. J. Ashe, S. Mahmoud, C. Elschenbroich and M. Wunsch, Angew. Chem., Int. Ed. Engl., 1987, 26, 229–230; (b) L. Chiche, J. Galy, G. Thiollet and F. Mathey, Acta Crystallogr., Sect. B: Struct. Crystallogr. Cryst. Chem., 1980, 36, 1344–1347; (c) G. De Lauzon, F. Mathey and M. Simalty, J. Organomet. Chem., 1978, 156, C33–C36; (d) F. Mathey, Tetrahedron Lett., 1976, 17, 4155–4158; (e) A. J. Ashe, J. W. Kampf, S. Pilotek and R. Rousseau, Organometallics, 1994, 13, 4067–4071; (f) R. M. Kirk and A. F. Hill, Chem. Sci., 2022, 13, 6830–6835.
4. (a) E. Keller, A. Trenkle and H. Vahrenkamp, Chem. Ber., 1977, 110, 441–448; (b) R. Janta, W. Albert, H. Rossner, W. Malisch, H.-J. Langenbach, E. Rottinger and H. Vahrenkamp, Chem. Ber., 1980, 113, 2729–2738; (c) F. Richter and H. Vahrenkamp, Chem. Ber., 1982, 115, 3224–3242; (d) M. Muller and H. Vahrenkamp, Chem. Ber., 1983, 116, 2748–2764; (e) H.-J. Langenbach, E. Rottinger and H. Vahrenkamp, Chem. Ber., 1980, 113, 42–54; (f) F. Richter, H. Beurich, M. Muller, N. Gartner and H. Vahrenkamp, Chem. Ber., 1983, 116, 3774–3793; (g) H. J. Langenbach, E. Keller and H. Vahrenkamp, J. Organomet. Chem., 1979, 171, 259–271; (h) E. Keller and H. Vahrenkamp, Chem. Ber., 1978, 111, 65–71; (i) E. Keller and H. Vahrenkamp, Chem. Ber., 1976, 109, 229–236; (j) E. Rottinger, A. Trenkle, R. Muller and H. Vahrenkamp, Chem. Ber., 1980, 113, 1280–1289; (k) H. J. Langenbach, E. Keller and H. Vahrenkamp, J. Organomet. Chem., 1980, 191, 95–106; (l) H. Vahrenkamp, Chem. Ber., 1974, 107, 3867–3873; (m) M. Muller, H.-T. Schacht, K. Fischer, J. Ensling, P. Gutlich and H. Vahrenkamp, Inorg. Chem., 1986, 25, 4032–4038; (n) F. Richter and H. Vahrenkamp, Organometallics, 1982, 1, 756–757; (o) E. Gross, C. Burschka and W. Malisch, Chem. Ber., 1986, 119, 378–382.
5. (a) A. M. Arif, R. A. Jones, M. H. Seeberger, B. R. Whittlesey and T. C. Wright, Inorg. Chem., 1986, 25, 3943–3949; (b) R. A. Jones and B. R. Whittlesey, Inorg. Chem., 1986, 25, 852–856; (c) A. M. Arif, R. A. Jones, S. T. Schwab and B. R. Whittlesey, J. Am. Chem. Soc., 1986, 108, 1703–1705; (d) R. A. Jones and B. R. Whittlesey, Inorg. Chem., 1986, 25, 852–856; (e) A. M. Arif, D. E. Heaton, R. A. Jones, K. B. Kidd, T. C. Wright, B. R. Whittlesey, J. L. Atwood, W. E. Hunter and H. Zhang, Inorg. Chem., 1987, 26, 4065–4073; (f) R. A. Jones, S. T. Schwab, A. L. Stuart, B. R. Whittlesey and T. C. Wright, Polyhedron, 1985, 4, 1689–1695.
6. (a) G. Conole, J. E. Davies, J. D. King, M. J. Mays, M. McPartlin, H. R. Powell and P. R. Raithby, J. Organomet. Chem., 1999, 585, 141–149; (b) G. Conole, J. E. Davies, J. D. King, M. J. Mays, M. McPartlin, H. R. Powell and P. R. Raithby, J. Organomet. Chem., 1999, 585, 141–149; (c) J. E. Davies, N. Feeder, C. A. Gray, M. J. Mays and A. D. Woods, J. Chem. Soc., Dalton Trans., 2000, 1695–1702.
7. (a) M. L. H. Green, J. Organomet. Chem., 1995, 500, 127–148; (b) M. L. H. Green and G. Parkin, J. Chem. Ed., 2014, 91, 807–816.
8. W. E. Buhro, B. D. Zwick, S. Giorgiou, J. P. Hutchinson and J. A. Gladysz, J. Am. Chem. Soc., 1988, 110, 2427–2439.
9. (a) G. Wittig and D. Hellwinkel, Angew. Chem., 1962, 74, 782–783; (b) G. Wittig and D. Hellwinkel, Chem. Ber., 1964, 93, 769–783.
10. (a) H. Imoto, A. Urushizaki, I. Kawashima and K. Naka, Chem. – Eur. J., 2018, 24, 8797–8803; (b) M. Ishidoshiro, Y. Matsumura, H. Imoto, Y. Irie, T. Kato, S. Watase, K. Matsukawa, S. Inagi, I. Tomita and K. Naka, Org. Lett., 2015, 17, 4854–4857.
11. (a) R. G. Parr and W. Yang, J. Am. Chem. Soc., 1984, 106, 4049–4050; (b) W. Yang and W. J. Mortier, J. Am. Chem. Soc., 1986, 108, 5708–5711.
12. J. A. Steed, H. R. Sharpe, H. J. Futcher, A. J. Wooles and S. J. Liddle, Angew. Chem., Int. Ed., 2020, 59, 15870–15874.
13. G. Märkl and H. Hauptman, J. Organomet. Chem., 1983, 248, 269–285.
14. (a) T. S. Piper and G. Wilkinson, J. Inorg. Nucl. Chem., 1956, 3, 104–124; (b) M. Abrantes, P. Neves, M. M. Antunes, S. Gago, F. A. Almeida Paz, A. E. Rodrigues, M. Pillinger, I. S. Gonçalves, C. M. Silva and A. A. Valente, J. Mol. Catal. A: Chem., 2010, 320, 19–26.
15. (a) M. Mickiewicz, K. P. Wainwright and S. B. Wild, J. Chem. Soc., Dalton Trans., 1976, 262–269; (b) J. L. Davidson and D. W. A. Sharp, J. Chem. Soc., Dalton Trans., 1975, 2531–2534; (c) G. B. McVicker, Inorg. Chem., 1975, 14, 2087–2092; (d) J. L. Davidson, J. Chem. Soc., Chem. Commun., 1980, 113–114; (e) E. Viola, C. Lo Sterzo and F. Trezzi, Organometallics, 1996, 15, 4352–4354; (f) A. R. Manning, J. Chem. Soc. A, 1967, 1984–1987.
16. (a) Y.-C. Shi, Polyhedron, 2004, 23, 1663–1667; (b) A. Mawby and G. E. Pringly, J. Inorg. Nucl. Chem., 1972, 34, 525–530; (c) J. H. Brownie, M. C. Baird and H. Schmider, Organometallics, 2007, 26, 1433–1443; (d) A. Ricci, F. Angelucci, M. Chiarini, C. Lo Sterzo, D. Masi, G. Giambastiani and C. Bianchini, Organometallics, 2008, 27, 1617–1625.
17. (a) M. R. Anstey, J. L. Bost, A. S. Grumman, N. D. Kennedy and M. T. Whited, Acta Crystallogr., Sect. E: Crystallogr. Commun., 2020, 76, 547–551; (b) M. T. Whited, M. A. Ball, A. Block, B. A. Brewster, L. Ferrer, H. J. Jin-Lee, C. J. King, J. D. North, I. L. Shelton and D. G. Wilson, Acta Crystallogr., Sect. E: Crystallogr. Commun., 2021, 77, 912–918; (c) M. T. Whited, G. E. Hofmeister, C. J. Hodges, L. T. Jensen, S. H. Keyes, A. Ngamnithiporn and D. E. Janzen, Acta Crystallogr., Sect. E: Struct. Rep. Online, 2014, 70, 216–220.
18. (a) K. W. Barnett, D. L. Beach, S. P. Gaydos and T. G. Pollmann, J. Organomet. Chem., 1974, 69, 121–130; (b) R. B. King and M. Saran, Inorg. Chem., 1974, 13, 364–367; (c) G. Grötsch and W. Malisch, J. Organomet. Chem., 1983, 258, 297–305; (d) P. J. Craig and J. Edwards, J. Organomet. Chem., 1972, 46, 335–337.
19. (a) J. M. Smith, D. P. White and N. J. Coville, Polyhedron, 1996, 15, 4541–4554; (b) W. E. Stanclift and D. G. Hendricker, J. Organomet. Chem., 1973, 50, 175–184.
20. (a) P. Kalck, R. Pince, R. Poilblanc and J. Roussel, J. Organomet. Chem., 1970, 24, 445–452; (b) R. J. Mawby and G. Wright, J. Organomet. Chem., 1970, 21, 169–170; (c) G. Wright and R. J. Mawby, J. Organomet. Chem., 1973, 51, 281–287; (d) M. D. Bala, D. C. Levendis and N. J. Coville, J. Organomet. Chem., 2006, 691, 1919–1926.
21. (a) A. Wojcicki, Adv. Organomet. Chem., 1973, 11, 87–145; (b) F. Calderazzo, Angew. Chem., Int. Ed. Engl., 1977, 16, 299–311; (c) K. J. Cavell, Coord. Chem. Rev., 1996, 155, 209–243.
22. (a) J. P. Bibler and A. Wojcicki, Inorg. Chem., 1966, 5, 889–892; (b) K. W. Barnett and P. M. Treichel, Inorg. Chem., 1967, 6, 294–299; (c) O. G. Adeyemi and N. J. Coville, Organometallics, 2003, 22, 2284–2290; (d) M. Kumar, A. J. Metta-Magana, H. K. Sharma and K. H. Pannell, Dalton Trans., 2010, 39, 7125–7131.
23. (a) S. A. Llewellyn, M. L. H. Green and A. R. Cowley, Dalton Trans., 2006, 1776–1783; (b) S. B. Butts, S. H. Strauss, E. M. Holt, R. E. Stimson, N. W. Alcock and D. F. Shriver, J. Am. Chem. Soc., 1980, 102, 5093–5100.
24. A. Harris and A. J. Rest, J. Organomet. Chem., 1974, 78, C29–C30.
25. A. C. Gingell, A. Harris, A. J. Rest and R. N. Turner, J. Organomet. Chem., 1976, 121, 205–210.
26. F. Mathey, Chem. Rev., 1990, 90, 997–1025.
27. W. A. Schenk and E. Vob, J. Organomet. Chem., 1994, 467, 67–73.
28. (a) M. Ma, S. A. Pullarkat, Y. Li and P.-H. Leung, J. Organomet. Chem., 2008, 693, 3289–3294; (b) M. Ma, S. A. Pullarkat, M. Yuan, N. Zhang, Y. Li and P.-H. Leung, Organometallics, 2009, 28, 4886–4889; (c) M. Ma, S. A. Pullarkat, K. Chen, Y. Li and P.-H. Leung, J. Organomet. Chem., 2009, 694, 1929–1933.
29. J. B. Hendrikson, R. E. Spenger and J. J. Sims, Tetrahedron Lett., 1961, 477–480.
30. For a review of heterocycles arising from the nucleophilic addition of heteroatoms to DMAD see: M. V. Georg, S. K. Khetan and R. K. Gupta, Adv. Heterocycl. Chem., 1976, 19, 279–371.
31. Y. Himeshima, T. Sonoda and H. Kobayashi, Chem. Lett., 1983, 12, 1211–1214.
32. L. Carlton, J. L. Davidson and M. Shiralian, J. Chem. Soc., Dalton Trans., 1986, 1577–1586.
33. A. Marinetti, F. Mathey, J. Fischer and A. Mitschler, J. Chem. Soc., Chem. Commun., 1982, 667–668.
34. F. Mercier, L. Ricard and F. Mathey, Organometallics, 1993, 12, 98–103.
35. W. R. Cullen, L. Mihichuk, F. W. B. Einstein and J. S. Field, J. Organomet. Chem., 1974, 73, C53–C55.

---

Footnote

† Electronic supplementary information (ESI) available: Spectroscopic, computational and crystallographic data. CCDC 2145351, 2145364, 2145367, 2145381–2145383, 2145459 and 2149526. For ESI and crystallographic data in CIF or other electronic format see DOI: https://doi.org/10.1039/d4dt01371a

---