[Pd 4 ( μ 3 ‑ SbMe 3 ) 4 (SbMe 3 ) 4 ]: A Pd(0) Tetrahedron with μ 3 ‑ Bridging Trimethylantimony Ligands

: The palladium(II) chlorostibine complex [PdCl 2 (SbMe 2 Cl) 2 ] 2 has a dimeric structure in the solid state, stabilized by hyper-coordination at the Lewis amphoteric Sb centers. Reaction with 8 equiv of MeLi forms [Pd 4 ( μ 3 -SbMe 3 ) 4 (SbMe 3 ) 4 ], whose structure comprises a tetrahedral Pd(0) core with four terminal SbMe 3 ligands and four μ 3 -SbMe 3 ligands, one capping each triangular Pd 3 face. Density functional theory calculations, supported by energy decomposition analysis and the natural orbitals for chemical valence scheme, highlight signi ﬁ cant donor and acceptor orbital contributions to the bonding between both the terminal and the bridging SbMe 3 ligands and the Pd 4 core.

D espite their ubiquity in modern coordination chemistry, it was long thought that phosphine ligands and their heavier pnictine congeners (PnR 3 ; Pn = P, As, Sb, Bi) were terminal donors only, while other π-acceptor ligands such as CO are frequently found to bridge more than one metal center. Werner was the first to challenge this concept, with the isolation of a Rh 2 dimer bridged by Sb i Pr 3 (Figure 1, i), its subsequent ligand metathesis giving rise to the first examples of μ 2 -bridging PR 3 and AsR 3 . 1,2 Despite this breakthrough, very few other systems with bridging pnictines have since been characterized; Balch isolated a few examples of PF 3 triply bridging a Pd 3 triangle (Figure 1, ii), 3,4 and Gabbaı̈recently reported the complexation of a tetradentate P 3 Pn ligand (Pn = Sb, Bi) with a M 3 triangle (M = Cu, Ag) in which the heavy pnictine donor is supported centrally above the M 3 face ( Figure  1, iii). 5 Because of the rarity of such species, little is known about the nature of the bonding in these complexes. Based on the limited examples, the bridging mode seems to feature a significant component of acceptance by the ligand and is best stabilized by late transition metals in low oxidation states and strongly π-accepting or heavier, more Lewis acidic pnictines. 6 There has been a surge of recent interest in the "noninnocent" behavior of coordinated heavy pnictines, which in several cases demonstrate redox reactivity or anion exchange at Pn in preference to the transition metal center. 7,8 They are also prone to hyper-coordination, forming intra-or intermolecular secondary acceptor interactions with electronegative donor atoms; this behavior is enhanced by electronegative substituents on the pnictine, which increase the Lewis acidity of the Pn center. 9 We have previously demonstrated that increasing the number of halide substituents in SbBr n Me 3−n (n = 0−2) increases the π-acceptor capacity of the stibine ligand. 10 While triorganopnictines are σ-donor/π-acceptor ligands, halidesubstituted Sb and Bi centers have been seen to act as σacceptors toward electron-rich transition metals, giving rise to complexes with highly unusual electronic strutures. 11−14 We report here an unusual dimeric Pd(II) complex of SbMe 2 Cl which demonstrates significant Lewis acidity of the bound halostibine. The reaction of this complex with MeLi leads to formation of an unexpected Pd(0) cluster featuring both terminal and triply bridging SbMe 3 ligands; the first example of an unsupported μ 3 -organopnictine ligand.
The reaction of [PdCl 2 (MeCN) 2 ] with 2 equiv of SbMe 2 Cl resulted in the formation of [PdCl 2 (SbMe 2 Cl) 2 ] as a red solid in good yield, which appears stable in air for several hours (Scheme 1). The expected singlet was observed in the 1 H NMR spectrum as well as a single broad 13 C{ 1 H} NMR resonance. The solid-state far IR spectrum shows two bands corresponding to Pd−Cl stretches (C 2v : A 1 + B 1 ). Crystals were grown from the benzene filtrate and analyzed by X-ray crystallography. The solid-state structure comprises the centrosymmetric dimeric unit [PdCl 2 (SbMe 2 Cl) 2 ] 2 , consisting of two distorted square planar Pd centers with cis chloride and chlorostibine ligands, connected by a fairly short Pd(II)−Pd(II) interaction (2.9143(4) Å) ( Figure 2).
Most examples of Pd(II) dimers feature bidentate bridging ligands supporting the Pd−Pd interaction. One rare counterexample is the diaminosugar complex [Pd(C 7 H 16 N 2 O 2 )Cl 2 ] 2 (Pd−Pd = 3.284 Å), in which dimerization is supported by H-bonding between amine and Cl ligands. 15 A similar conformation is adopted by [PdCl 2 (SbMe 2 Cl) 2 ] 2 ; each chlorostibine ligand eclipses a Cl ligand on the opposite Pd center when viewed down the Pd−Pd vector (torsion angles: Cl4−Pd1−Pd1a−Sb1a = 3.94(3)°Cl3−Pd1−Pd1a−Sb2a = 5.68(3)°), leading to very short intermolecular Sb−Cl distances (mean 2.96 Å, cf. mean Sb−Cl covalent = 2.39 Å, Σ vdW = 4.29 Å). 16 The propensity of coordinated Sb or Bi donors to act simultaneously as acceptors, forming intra-or intermolecular "hypervalent" interactions with electronegative atoms, is a current area of interest, 8,9 and it has been demonstrated that these interactions can strongly direct the solid-state structure of a complex. 17 It appears that in [PdCl 2 (SbMe 2 Cl) 2 ] 2 , four such Sb···Cl interactions support the formation of the dimeric species. These interactions form approximately trans to the covalently bonded halide substituent on Sb (mean ∠ Cl···Sb− Cl = 169.0°), the Sb−Cl σ* being the most accessible acceptor orbital on Sb. Consistent with this, natural bond orbital (NBO) analysis identifies a notable 3p Cl → σ* Sb−Cl interaction (see SI for further details). The geometry at Sb is near trigonal bipyramidal, severely distorted from the expected pseudotetrahedral. The structure is reminiscent of that of the Pt(II)− Pt(II) dimer [PtCl 2 {CH 2 (o-C 6 H 4 CH 2 SbMe 2 )} 2 ] 2 , which contains weak intermolecular Sb···Cl contacts (mean 3.48 Å); 18 the considerably shorter Sb···Cl distances found in [PdCl 2 (SbMe 2 Cl) 2 ] 2 can be accounted for by the increased acceptor power of the halostibine in comparison to the triorganostibine. In each case it is difficult to separate the magnitude of the secondary Sb···Cl interaction from that of the metallophilic interaction between the group 10 metals.
There are very few previous reports of halostibine complexes with transition metal halides. In view of the recent interest surrounding the "noninnocent" behavior of coordinated stibines, we investigated the reactivity of [PdCl 2 (SbMe 2 Cl) 2 ] 2 with reagents which specifically have the potential to target both the Pd and Sb metal centers. Treatment of [PdCl 2 (SbMe 2 Cl) 2 ] 2 with 8 equiv of MeLi (equimolar with Cl) in tetrahydrofuran resulted in the formation of an intensely violet solution, from which a dark purple solid was isolated (Scheme 1). The product is stable over several weeks when stored under an N 2 atmosphere, but slowly becomes black/ brown in contact with air. It is remarkably soluble in n-hexane and less soluble in chlorinated solvents. Small purple crystals were analyzed by X-ray diffraction, giving the solid-state structure shown in Figure 3, formulated as [Pd 4 (μ 3 - The structure comprises a central tetrahedron of four Pd(0) atoms, with an average Pd−Pd distance of 2.805 Å. Each Pd is coordinated to one terminal SbMe 3 ligand (mean Pd−Sb = 2.520 Å) and one SbMe 3 ligand caps each face of the tetrahedron, bridging three Pd atoms (mean Pd−Sb = 2.773 Å). A mirror plane bisects the tetrahedron, passing through two Pd atoms, two bridging Sb, and two terminal Sb atoms. There is symmetry-related disorder of the Me substituents on the terminal Sb1. Each bridging SbMe 3 ligand is almost equidistant from the three Pd atoms it caps; the least symmetrical is Sb5, with a difference of 0.07 Å between Sb5−Pd1 and Sb5−Pd2. The most symmetrical, Sb4, has <0.01 Å difference between Sb4−Pd1, Sb4−Pd2, and Sb4−Pd3. The molecule has near C 3v symmetry, but attempts to solve the diffraction data in higher symmetry space groups were unsatisfactory; the Cmc2 1 solution is correct. Figure 3b,c shows alternative views of the structure in which this pseudosymmetry can be clearly discerned.
The 1 H and 13

Journal of the American Chemical Society
Communication distinct SbMe 3 environments. This is consistent with the conservation of the tetramer in solution, the broadening of the peaks being most likely due to the proximity of the quadrupolar Sb nuclei ( 121 Sb I = 5/2; 123 Sb I = 7/2). The identity of the product is supported by elemental analysis.
The formal cluster electron count of 56 for [Pd 4 (μ 3 -SbMe 3 ) 4 (SbMe 3 ) 4 ] is 4 electrons fewer than the predicted valence electron count of 60 for a tetrahedron with localized bonding. Such an ideal count is exemplified in [Ni 4 (CO) 6 (P-(C 2 H 4 CN) 3 ) 4 ] (T d symmetry), 22 which contains 6 edgebridging carbonyls and 4 terminal phosphine ligands at the vertices. However, stable electron-deficient clusters are not uncommon for the heavier group 10 metals, 20,23−26 which often form stable compounds that do not conform to the 18-electron rule, a fact attributed to the increased energy gap between the valence d and p orbitals in these late transition metals. Mingos discussed the electronic structure of the hypothetical [Pt(PH 3 ) 2 ] 4 clusters, in which terminal PH 3 bonding was assumed, and predicted counts of 56 or 54 electrons depending on the orientation of the ligands. 27 Both the stability and diamagnetism of the title cluster are borne out in the molecular orbital (MO) diagram ( Figure S2). There are 40 electrons that occupy all orbitals of the d manifold, forming a band of MOs centered on the edges and faces of the Pd 4 core. Overlap between symmetry-adapted (4a 1 + 2e) ligand group donor orbitals with combinations of metalbased σ-type 5s/5p cluster acceptor orbitals of matching symmetry leads to the formation of bonding MOs, lying energetically below the d block. These orbitals accommodate the eight donor electron pairs and account for metal−ligand bonding. A considerable gap of ∼2 eV separates the LUMO (e 1 symmetry) from the HOMO.
An energy decomposition analysis (EDA) 28,29 was carried out in order to compare the donor−acceptor capabilities of the terminal and face-capping stibine ligands and their interactions with the remaining {Pd 4 (SbMe 3 ) 7 } fragment ( Table 1). The terminal SbMe 3 ligand exhibits a larger fragment binding energy, −D e , (−17.2 kcal mol −1 ) compared to the face-capping motif (−10.3 kcal mol −1 ). The face-capping location of the μ 3 -SbMe 3 ligand over a {Pd 3 } face results in enhanced interactions relative to the terminal SbMe 3 interacting with a single Pd center. Thus, μ 3 -SbMe 3 has a larger ΔE steric (+47.2 vs +15.5 kcal mol −1 ), but this is offset by a greater ΔE orb (−65.5 vs −37.8 kcal mol −1 ). The individual contributions to ΔE orb are dominated by the A 1 and E 1 components equating to σdonation (Sb → Pd) and π-back-donation (Sb ← Pd), respectively; both components are again more significant for the μ 3 -SbMe 3 ligand. It is striking that for both binding modes the electrostatic term ΔE elstat makes a significantly larger contribution (71−76%) to the total metal−ligand bonding than the orbital term ΔE orb (24−29%). A similar observation has been reported for terminal phosphines, 30 indicating that focusing on orbital interactions alone may be misleading.
Importantly, the ligand−cluster interactions are further stabilized by dispersion effects, yielding total fragment binding energies of −53.6 and −43.8 kcal mol −1 for μ 3 -SbMe 3 and terminal SbMe 3 , respectively, reversing the order of ligand binding relative to the electronic term alone.
The natural orbitals for chemical valence (NOCV) 31 scheme allows for further insight into the ΔE orb term by highlighting the dominant deformation density channels. Isocontour plots of these channels aid visualization of σ-donation and π-backdonation in the cluster (Figure 4). For the terminal case, electron σ-donation from the Sb lone pair (5s) into the vacant Pd acceptor orbital (5s) makes a strong contribution to bonding (ΔE 1 orb = −14.8 kcal mol −1 ). Two components of πback-donation from Pd to Sb can also be clearly identified and are characterized by energies of ΔE 2 orb = ΔE 3 orb = −6.0 kcal mol −1 . A similar analysis for the μ 3 -SbMe 3 fragment reveals an increase in the stabilization energies and associated charge flows for both the σ (ΔE 1 orb = −24.5 kcal mol −1 ) and π channels (ΔE 2 orb = ΔE 3 orb = −10.5 kcal mol −1 ) due to the larger overlap area provided by the Pd 3 face, in line with the EDA analysis.
To summarize, a rare example of a halostibine complex with a transition metal halide has been synthesized and characterized as a dimer in the solid state, supported by secondary Sb···Cl interactions. This complex demonstrates unexpected reactivity with MeLi, resulting in isolation of a highly unusual Pd(0) 4 cluster with μ 3 -SbMe 3 ligands; the first example of triple bridging by a monodentate organopnictine. Computational modeling of the cluster reveals that both bridging and terminal SbMe 3 ligands can be efficient acceptors in metal-to-ligand πback-donation, helping to stabilize the electron rich {Pd 4 } core. Investigation of potential phosphine and arsine analogues is underway in our group. The effect of this new pnictine bonding mode on the electronic environment of the transition metal could have considerable impacts in organometallic chemistry, including in the design of new homogeneous catalysts.