Transition‐Metal Complexes Featuring Z‐Type Ligands: Agreement or Discrepancy between Geometry and d^<i>n</i> Configuration?

The value of gold: The coordination of ambiphilic diphosphanylborane ligands to AuCl provides unusual square-planar gold(I) complexes. Insight is gained on the nature of the gold→borane interactions in these complexes through natural bond orbital (NBO) analysis and 197Au Mössbauer spectroscopy. The seminal formalism MLlXx (M=transition metal, L=2e-donor ligand, X=1e-donor ligand) provides a unified description for transition-metal complexes. Besides the well-known L- and X-type ligands, the ability of Lewis acids to act as zero-electron donors/two-electron acceptors was recognized early on,1 and these ligands were referred to as Z-type ligands in MLlXxZz complexes. Although such σ-acceptor ligands remain considerably less developed than their σ-donor counterparts, significant advances have been achieved over the last decade with Group 13 Lewis acids of the type ER3 (E=B, Al).2–4 In particular, an increasing number and variety of transition-metal–borane complexes (M→BR3) have been unambiguously authenticated.5 Following the pioneering contribution of Hill et al.,6 metallaboratranes A have become general scaffolds for supporting M→BR3 interactions (M=Fe, Ru, Os, Co, Rh, Ir, Ni, Pd, Pt; Scheme 1).7 The related iridium complex B further expanded the variety of such interactions to complexes featuring only two methimazolyl buttresses.8 In addition, we have demonstrated that M→B interactions9 are readily accessible by coordination of preformed ambiphilic phosphanylborane ligands.17–19 The ensuing square-pyramidal complexes C exemplified the possibility for M→B interactions to exist in the absence of σ-donor ligands in the position trans to the Lewis acid17a and substantiated the marked influence that the metal (M=Rh, Pt, Pd) may have on such interactions.17d Furthermore, the T-shaped complexes D provided the first evidence for M→B interactions in 14-electron complexes supported by a single phosphane buttress.17b Structurally characterized complexes featuring M→BR3 interactions. The increasing number and variety of complexes featuring dative M→B bonds have raised fundamental questions as to the very nature of such interactions. Accordingly, Hill20a and Parkin20b proposed two conflicting bonding situations to describe complexes featuring such Z-type ligands: 1) retention of the original dn configuration of the metal center and a coordinated neutral BR3 ligand, and 2) two-electron oxidation of the metal center resulting in a dn−2 configuration and a dianionic BR32− ligand (Scheme 2). At first glance, these two descriptions may be considered as just different representations of dative M→B bonds, but they in fact reflect two extreme bonding situations that are intimately related to the extent of electron-density transfer from the metal atom to the boron center. In this context, we report herein a combined experimental and theoretical investigation of [AuCl(diphosphanylborane)] complexes featuring short Au–B contacts. The observed square-planar coordination geometries of the tetracoordinate gold centers could be considered as indicating d8 gold(III) configurations, but natural bond orbital (NBO) analyses and Mössbauer spectroscopic measurements unambiguously established that these complexes retain d10 gold(I) configurations. These results provide the first evidence that basic principles usually dictating the geometry of transition-metal complexes may be challenged with Z-type ligands. The two limiting bonding situations for dative M→BR3 interactions proposed by Hill20a and Parkin,20b respectively. The diphosphanylborane 1 a17a,17c readily displaced the labile dimethyl sulfide ligand of [AuCl(SMe2)] in dichloromethane (DCM) at room temperature (RT; Scheme 3). The resulting complex 2 a was isolated in 90 % yield as a white, air-stable powder. The 31P NMR spectrum of complex 2 a exhibited a single signal at δ=73.2 ppm, indicating symmetric coordination of the phosphorus atoms. In addition, the broad resonance observed at δ=24.6 ppm in the 11B NMR spectrum of 2 a is very similar to that encountered for related rhodium complexes (δ=19.4–26.7 ppm), suggesting the presence of an Au→B interaction. The precise structure of 2 a was established by an X-ray diffraction study (Figure 1).21 The gold center is tetracoordinate and had a slightly distorted square-planar coordination geometry (sum of angles (∑Auα)=362.2°). The two phosphane moieties span trans sites with a significantly bent P-Au-P arrangement (160.2°). The B-Au-Cl arrangement is closer to linear (168.7°). The Au–B distance (2.309 Å) is much shorter than those observed in related gold complexes with monophosphanylborane ligands (2.66–2.90 Å),17b but falls within the same range as those encountered in rhodium complexes with diphosphanylborane ligands (2.29–2.37 Å).17a,17d The presence of a rather strong Au→B interaction22 is further supported by the noticeable pyramidalization of the boron environment (∑Bα=341.2°). 2 a is the first example of a complex of an ambiphilic phosphanylborane ligand featuring a coligand in a position trans to the boron atom. Molecular structure (thermal ellipsoids set at 50 % probability) of complex 2 a in the solid state. Hydrogen atoms are omitted for clarity. Coordination of the ambiphilic diphosphanylborane ligands 1 a,b to AuCl. With regards to the bonding description of the M→B interactions, the square-planar geometry observed in 2 a may indicate a d8 gold(III) configuration.23 Indeed, complexes of tetracoordinate d10 gold(I) adopt tetrahedral and not square-planar arrangements.24 To probe substituent effects on M→B interactions, the isopropyl groups at the phosphorus atoms were replaced by phenyl moieties. The ligand 1 b and ensuing complex 2 b were prepared in a similar manner to 1 a and 2 a. X-ray diffraction analysis of 2 b revealed an overall geometry very similar to that of 2 a.21 Notably, the lower steric hindrance and electron-donating character of the PPh2 groups does not affect the Au–P and Au–Cl distances, but results in a slightly elongated Au–B distance (2.335 Å) and a slightly less pronounced pyramidalization of the boron environment (∑Bα=343.8°). To address the bonding description of the Au→B interactions in more detail, DFT calculations were carried out on the actual complexes 2 a,b*.25 As previously observed for related complexes of Group 9 and 10 transition metals, the B3PW91/SDD(Au,P),6-31G**(other atoms) level of theory was found to reproduce the experimental geometric features particularly well (Table 1). The corresponding frontier molecular orbitals reveal three-center B-Au-Cl interactions. These interactions involve the d orbital of gold and the py orbitals of boron and chlorine, with bonding gold–boron and antibonding gold–chlorine interactions in the highest occupied molecular orbital (HOMO), and antibonding gold–boron and gold–chlorine interactions in the lowest unoccupied molecular orbital (LUMO; Figure 2). Parkin et al. reported similar three-center four-electron interactions for iridaboratranes and argued for a formal two-electron oxidation/dn−2 configuration of the metal atom on the basis of its negligible contribution to the nonbonding B-Ir-Cl orbital.7g Second-order perturbative NBO analyses of 2 a,b* also provided evidence for Au→B dative interactions, with NBO delocalization energies of about 55 kcal mol−1. Notably, the coordination of the diphosphanylborane ligand induces only a moderate depletion of the NBO charge at boron (from +0.85 in the open form of the ligand 1 a17d to +0.37 in the ensuing complex 2 a), which is accompanied by a slight increase of the charge at gold (from +0.30 in the model borane-free T-shaped complex [AuCl(PMe3)2] to +0.64 in 2 a). This result suggests that complexes 2 a,b would be more appropriately described as gold(I) than as gold(III) complexes, as initially proposed on the basis of geometric considerations. This hypothesis was further corroborated by the computed natural electron configuration26 of the metal center ([Xe]6s(0.70)5d(9.63)6d(0.01)7p(0.01) for 2 a), which deviates only marginally from the d10 configuration expected for a gold(I) center ([Xe]6s(0.88)5d(9.79)6p(0.03) computed for [AuCl(PMe3)2]). Frontier orbitals for complex 2 a*. Hydrogen atoms are omitted for clarity. P–Au Au–B Au–Cl ∑αB ∑αAu P-Au-P B-Au-Cl 2 a 2.313(2) 2.328(2) 2.309(8) 2.522(2) 341.2 362.2 160.2(1) 168.7(2) 2 a* 2.35 2.37 2.32 2.56 340.1 362.5 158.0 170.8 2 b 2.307(1) 2.329(1) 2.335(5) 2.524(1) 343.8 364.2 157.2 (1) 162.0(2) 2 b* 2.34 2.36 2.35 2.56 343.5 362.9 158.2 168.8 To support such an unprecedented combination of a square-planar geometry with a d10 gold(I) configuration, further experimental evidence was necessary. 197Au Mössbauer spectroscopy, which has previously been used to determine the structure and bonding of a variety of inorganic gold compounds,27 was considered a particularly valuable probe. These measurements were performed on both complexes 2 a,b to avoid any ambiguity arising from substituent effects. The resulting spectra consist of well-resolved quadrupole doublets, with isomer shifts (ISs) of about 3.35 mm s−1 (relative to a 196Pt/Pt source, corresponding to 4.57 mm s−1 relative to gold metal) and quadrupole splittings (QSs) of about 7.6 mm s−1 (Figure 3 a). Although these values are hardly comparable with those observed for gold complexes featuring only σ-donor ligands, the well-known IS/QS relationship28 unambiguously positions complexes 2 a,b among gold(I) complexes (Figure 3 b). These data, thus, definitely confirm that the square-planar complexes 2 a,b should be considered as complexes of d10 gold(I) rather than d8 gold(III). a) 197Au Mössbauer spectra for complexes 2 a,b at 12 K. T=Relative transmission, v=Doppler velocity. b) QS/IS correlation for complexes 2 a,b (red triangles) and representative phosphane complexes of gold(I) (green squares) and gold(III) (blue circles). IS values are given relative to gold metal. In conclusion, the square-planar gold(I) complexes 2 a,b with ambiphilic diphosphanylborane ligands provide a better understanding of the precise nature of M→BR3 interactions. It is demonstrated that the coordination of σ-acceptor Z-type ligands to transition metals is not necessarily accompanied by formal two-electron oxidation. This type of coordination may eventual challenge the basic rules usually governing the geometry of transition-metal complexes. Further investigations in this area are currently in progress with lighter Group 11 elements, as well as triphosphanylborane ligands. All reactions and manipulations were carried out under an atmosphere of dry argon, using standard Schlenk techniques. 2 a: A solution of 1 a (600 mg, 1.26 mmol) in CH2Cl2 (5 mL) was added at room temperature to a suspension of [AuCl(SMe2)] (371 mg, 1.26 mmol) in CH2Cl2 (5 mL). After stirring for 15 minutes, filtration of the reaction mixture over SiO2 or Al2O3 and evaporation of the volatile components afforded 2 a as an air-stable solid (899 mg, 90 % yield). Colorless crystals were obtained from a saturated pentane solution at room temperature. M.p. 183–185 °C; 31P NMR (202.5 MHz, CDCl3, 23 °C): δ=73.2 ppm; 11B NMR (160.5 MHz, CDCl3, 23 °C): δ=24.6 ppm; MS (ESI+) m/z (%): 707.5 [M+H]+ (<1), 671.3 [M−Cl]+ (100); HRMS (ESI+) m/z calcd for [C30H41AuBP2]+: 671.2442; found: 671.2451. Supporting information for this article is available on the WWW under http://www.wiley-vch.de/contents/jc_2002/2007/z703518_s.pdf or from the author. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.