Formation of Heterobimetallic Complexes by Addition of d10-Metal Ions to [(Me3P)xM(2-C6F4PPh2)2] (x = 1, 2; M = Ni and Pt): A Synthetic and Computational Study of Metallophilic Interactions

Treatment of the bis(chelate) complexes trans-[M(κ2-2-C6F4PPh2)2] (trans-1M; M = Ni, Pt) and cis-[Pt(κ2-2-C6F4PPh2)2] (cis-1Pt) with equimolar amounts or excess of PMe3 solution gave complexes of the type [(Me3P)xM(2-C6F4PPh2)2] (x = 2: 2Ma, 2Mbx = 1: 3Ma, 3Mb; M = Ni, Pt). The reactivity of complexes of the type 2M and 3M toward monovalent coinage metal ions (M′ = Cu, Ag, Au) was investigated next to the reaction of 1M toward [AuCl(PMe3)]. Four different complex types [(Me3P)2M(μ-2-C6F4PPh2)2M′Cl] (5MM′; M = Ni, Pt; M′ = Cu, Ag, Au), [(Me3P)M(κ2-2-C6F4PPh2)(μ-2-C6F4PPh2)M′Cl]x (x = 1: 6MM′; M = Pt; M′ = Cu, Au; x = 2: 6PtAg), head-to-tail-[(Me3P)ClM(μ-2-C6F4PPh2)2M′] (7MM′; M = Ni, Pt; M′ = Au), and head-to-head-[(Me3P)ClM(μ-2-C6F4PPh2)2M′] (8MM′; M = Ni, Pt; M′ = Cu, Ag, Au) were observed. Single-crystal X-ray analyses of complexes 5–8 revealed short metal–metal separations (2.7124(3)–3.3287(7) Å), suggestive of attractive metal–metal interactions. Quantum chemical calculations (atoms in molecules (AIM), electron localization function (ELF), non-covalent interaction (NCI), and natural bond orbital (NBO)) gave theoretical support that the interaction characteristics reach from a pure attractive non-covalent to an electron-shared (covalent) character.


■ INTRODUCTION
Heterobimetallic compounds are of great interest due to their tuneable metal−metal bonds or interactions. 1 Hence, the chemistry of heterobinuclear complexes featuring two transition metals give rise to various intriguing redox, spectroscopic, and photophysical properties, which lead to a wide range of applications. 2 These features extend the scope of their monometallic counterparts. Such binuclear complexes bearing d 8 −d 8 , d 8 −d 10 , or d 10 −d 10 metal pairs (d 8 : Ir I , Ni II , Pd II , Pt II , Au III ; d 10 : Cu I , Ag I , Au I , Hg II ) have been generated unsupported or supported by bridging ligands. 3 Heterobimetallic systems featuring groups 10 and 11 metals are of current interest due to their key role in cooperative catalysis. 4 Recent success in trapping a reactive intermediate was reported for a Ni-catalyzed cross-coupling reaction, where a Ni−Cu complex could be isolated as a representative snapshot of the catalytic cycle (Chart 1, A). 5 Similarly, transmetalation from methyl groups from a Pt II toward a Au I center was reported to proceed via direct Pt II −Au I bond formation (supported by mass spectrometric data). 6 Structural information of such reactive intermediates is given for the homologue Pt II −Cu I complex (Chart 1, B) and analogue Pt II −Ag I and Pt II −Au I complexes (Chart 1, C) where in solution a dynamic coinage metal− carbon bond behavior was observed. 7 Discrete synthetically prepared d 8 −d 10 heterobimetallic compounds are important for the fundamental understanding of the relation between the nature of the metal−metal interactions and the complex structure. 8 Various M−M′ interactions with late transition metals (M, M′) have been studied during the past decades, and they were found to span between long range dispersion (London) energy 9 and ionic contribution 10 to metal-to-metal charge transfer 11 with significant covalency. 12 The complexes trans-[M(κ 2 -2-C 6 F 4 PPh 2 ) 2 ] (M = Ni, Pd, Pt; trans-1M) 13 were previously used as a d 8 -metal source to generate such bimetallic complexes. 14,15 The four-membered rings of the ortho-metalated C,P-ligand can open by treatment of trans-1M with neutral alkyl phosphines like PMe 3 or dppe ((diphenylphosphino)ethane), leading to a cis-or transorientation of the dangling κC-2-C 6 F 4 PPh 2 ligands about the transition-metal center (Scheme 1).
In contrast to observations along the nickel triad of the reaction of trans-1M with excess PMe 3 , the reaction of trans-1M with 1 equiv PMe 3 shows a different behavior depending on the transition metal. The 31 P NMR spectrum of the crude reaction mixture of trans-1Pt with 1 equiv PMe 3 shows the characteristic signal pattern of 2Pt a and unreacted trans-1Pt beside traces of unknown impurities. The 31 P NMR spectrum of the crude reaction mixture of trans-1Pd with 1 equiv PMe 3 shows a 1:1:1 ratio of trans-1Pd (−55.4 ppm), syn/anti-2Pd a (−18.7, −8.4, −19.2, −2.9 ppm), and 3Pd a (−35.5 ppm, 2 J P,P = 228.7 Hz, doublet, 2P; −13.8 ppm, 2 J P,P = 228.7 Hz, triplet, 1P). The presence of the symmetric isomer 3Pd a could be confirmed by 19 F NMR spectroscopy (−117.2, −128.6, −152.1, −159.5; Scheme 4). Attempts to isolate 3Pd a from the reaction mixture failed. The crude reaction mixture of trans-1Ni with 1 equiv PMe 3 shows the formation of 3Ni a with a small amount of 3Ni b (confirmed by 19 F and 31 P NMR spectroscopy). Rocamora et al. reported a similar nickel complex trans-[(Et 3 P)Ni(κ 2 -2-C 6 Cl 4 PPh 2 ) 2 ] to 3Ni a as the final product. 17 Stirring a dichloromethane (DCM) solution of 3Ni a for 4 days at ambient temperature resulted in ring opening and slow formation of 3Ni b . The obtained 1:3 mixture of the 3Ni a and 3Ni b does show eight additional signals in the 19 F and two signals in the 31 P NMR spectra. Through fractional crystallization, it was possible to separate and characterize the byproduct as [Ni 2 (κ 2 -2-C 6 F 4 PPh 2 ) 2 (μ-2-C 6 F 4 PPh 2 ) 2 ] (4Ni). It was only possible to isolate an isomeric mixture of 3Ni a and 3Ni b as a second fraction. Counterintuitively, the homologous platinum compound [Pt 2 (κ 2 -2-C 6 F 4 PPh 2 ) 2 (μ-2-C 6 F 4 PPh 2 ) 2 ] (4Pt) was accessible by quickly treating trans-1Pt with an excess of PMe 3 (ca. 2.8 equiv: 11% Scheme 3. Formation of Intermediate Compounds cis-[(Me 3 P) 2 Pt(κC-2-C 6 F 4 PPh 2 ) 2 ] (2Pt b ) and cis-[(Me 3 P)Pt(κ 2 -2-C 6 F 4 PPh 2 )(κC-2-C 6 F 4 PPh 2 )] (3Pt b ) Scheme 4. Reaction of trans-1M with 1 equiv of PMe 3 can act as a dimerization promoter. Starting from trans-1Ni or cis-1Pt and reacting it with 1 equiv of PMe 3 resulted in compounds of the type 3M b with cis-C-M-C arrangement. The same cis-C-M-C configuration is visible in both compounds of the type 4M. Our quantum chemical calculations reveal that the relative energies of 3M a and 3M b for M = Ni and Pt are similar (3M b is slightly more stable by <1.6 kcal mol −1 , respectively). The molecular structures of syn-2M a , 3M b , and 4M (M = Ni, Pt) were confirmed by single-crystal X-ray diffraction ( Figure 1 and Table S1). A detailed structural description is given in the Supporting Information.

C o m p l e x e s o f t h e T y p e [ ( M e 3 P ) 2 M ( μ -2 -C 6 F 4 PPh 2 ) 2 M′Cl].
Reacting cis-1Pt with excess PMe 3 resulted in the formation of syn/anti-2Pt b (Scheme 3), which was impossible to isolate. An in situ reaction of syn/anti-2Pt b with M′Cl (M′ = Cu, Ag, Au(tht); tht = tetrahydrothiophene) was not suitable to form the homologue cis-5PtM′ due to the presence of unreacted PMe 3 . The excess PMe 3 can easily react with the coinage metal chlorides to form various species of the form [M′Cl(PMe 3 ) x ] n (x = 1−4, n = 4−1). 19 To prevent byproduct formation, the synthesis was started from 3Pt b , which was reacted with 1 equiv M′Cl (M′ = Cu, Ag) to form complexes of type cis-6PtM′ (vide infra). The isolated compounds cis-6PtM′ were treated with 1 equiv PMe 3 in order to obtain complexes of type cis-5PtM′ (Scheme 5). Compound cis-6PtAu was accessible by reacting cis-1Pt with [AuCl(PMe 3 )] (Scheme 6).
Compound cis-5PtCu was accessible in a pure state by the mentioned synthesis route. Because of difficulties in handling the PMe 3 -toluene solution (air sensitivity and slow oxidation during storage), the addition of exactly 1 equiv PMe 3 to form complex cis-5PtAg in a pure state was not possible. Isolated   ]. The reaction of cis-6PtAu with 1 equiv PMe 3 only led to product mixtures. It was possible to obtain crystals suitable for single-crystal X-ray diffraction of cis-5PtCu and cis-5PtAg from dichloromethane/n-hexane solution ( Figure 2, Table S2).
By changing the coinage metal and keeping the d 8 -metal equal, the chemical shift of the 31 P NMR signal of the PMe 3 group is changing in a narrow range (Δδ P = 1.1−4.3 ppm) as expected. In contrast, by changing the d 8 -metal of the nickel triad and keeping the coinage metal equal, the chemical shift of the 31 P NMR signal of the bridging μ-2-C 6 F 4 PPh 2 shows a high field shift along the series Ni > Pd > Pt in a range of Δδ P = 8.1 (Cu), 9.5 (Ag), and 6.6 ppm (Au). For the series trans-5MCu (M = Ni, Pd, 14 Pt), it was possible to determine all three molecular structures, which are isomorphous (orthorhombic, space group Pnma). The molecular structures of trans-5NiCu (left) and trans-5PtCu (middle) are shown in Figure 2, and selected interatomic distances and angles are given in Table S2.
The structure overlay ( Figure S65) of complexes of the type trans-5MCu (M = Ni, Pd, 14 43(1) (Pd) < 3. 45(1) Å (Pt), which would suggest an increasing diamagnetic shielding contribution of heavier metal atoms in the 31 P NMR shifts of the PPh 2 group as observed for PMe 3 . However, in the monometallic complexes of type syn-2M a , the M···PPh 2 distances display a higher deviation [3.33(1) (Ni) < 3.40(1) (Pd) < 3.44(1) Å (Pt)] and are closer to the d 8 -metal with a smaller magnitude in Δδ P in the 31 P NMR spectra [1.7 ppm (syn-2M a ) vs 8.1 ppm (trans-5MCu)]. This would suggest that the diamagnetic shielding contribution of heavier metal atoms only plays a minor role in the 31 P NMR shifts of the bridging μ-2-C 6 F 4 PPh 2 . Calculation of the 31 P NMR shift difference of syn-2M a (2.6 ppm) vs trans-5MCu (6.6 ppm) supports this claim (see Supporting Information for details).
Due to the absence of major structural differences among the series of complexes of type trans-5MCu, we were interested    if the electronic structure has an impact on the formal large Δδ P in the 31 P NMR spectra. The non-covalent interaction (NCI) descriptor 21 shows clearly an increasing non-covalent attractive interaction along the series Ni < Pd < Pt ( Figure 3). Natural localized molecular orbital (NLMO) calculations 22 of the series trans-5MCu exhibit a decrease Cu contribution in both Cu−P bonds in the order Ni < Pd < Pt by about 6% in total. In the same order, Cu contributions to the d z 2 -orbital of Ni (0.06%), Pd (0.14%), and Pt (0.24%) increase similarly the (main) M→Cu interaction energy (second-order perturbation theory: Ni: 1.12; Pd: 7.67; Pt: 38.73 kcal mol −1 ; Table 3 and Figure 4). In the electron localization function (ELF), 23 an increase localized electron density is visible in the order Pd Topological analyses (atoms in molecules, AIM) have been performed to get a better understanding of the covalency of the M···M′ interactions. 24 The results of the topological parameters of compounds of the type 5MM′ are listed in Table 4. For all complexes of type 5MM′, critical points with (3, −1) characteristic (bond critical point, bcp) were found along the bond path between M and M′. The definition of a covalent bond "is based on two conditions, namely (i) the existence of a critical point r b and its associated maximum electron density path linking the nuclei in question (necessary condition) and (ii) H(r b ) < 0 which indicates that the accumulation of electron charge in the internuclear region is stabilizing (sufficient condition)" 25 (H(r b ) = electron energy density), which is the case for all complexes of type 5MM′.
Electron density (ρ(r b ) in au), Laplacian of electron density (∇ 2 ρ(r b ) in au), Lagrangian kinetic energy density (G(r b ) in au), potential energy density (V(r b ) in au), ratio |V(r b )|/G(r b ), ratio G(r b )/ρ(r b ) in au, electron energy density (H(r b ) in au), the product of sign of second largest eigenvalue of Hessian matrix of electron density (λ 2 (r b )), and ρ(r b ) in au.

Inorganic Chemistry
pubs.acs.org/IC Article electron-shared (covalent) character. Also, the relatively low electron density ρ(r b ) at the bcps together with a positive Laplacian of electron density (∇ 2 ρ(r b ) > 0) and similar modulus of Lagrangian kinetic energy density and potential energy density (G(r b ) ≙ |V(r b )|) are in support of the presence of a closed shell bonding with donor−acceptor characteristics. 26,27 The electron density at the bcp ρ(r b ) in the series trans-5MCu does increase in the expected order (M = Ni < Pd < Pt), which is in good agreement with the observations of ELF ( Figure 5). Closed shell interactions were previously categorized by Macchi et al. as metallic (shared) bonding and donor− acceptor interactions. 27 We will use these terms as described in the literature. To differentiate between metallic (shared) bonding behavior and donor−acceptor interactions, Macchi et al. suggested for heavy metal interactions that the G(r b )/ρ(r b ) ratio can be used as a classification criterion (G(r b )/ρ(r b ) < 1: metallic (shared) bonding vs G(r b )/ρ(r b ) ∼ 1: closed shell donor−acceptor interaction). 27 In the series trans-5MCu, the magnitude of G(r b )/ρ(r b ) is in the expected range and does increase (closer to 1) in the order trans-5NiCu < trans-5PdCu < trans-5PtCu. The increase closed shell donor−acceptor interaction down the nickel triad is supported by NBO analysis. Besides the main donor−acceptor interactions listed in Table 3 Due to a lack in structural difference, we assume that the 31 P NMR shift of the phosphorous atom of the bridging ligand μ-2-C 6 F 4 PPh 2 toward higher field (Table 2) is originated in electronic differences within the complex series caused by metal−metal interactions. For the series trans-5MAg and trans-5MAu (M = Ni, Pd, Pt), the same high field shift of the 31 P NMR resonance of the bridging ligand μ-2-C 6 F 4 PPh 2 down the nickel triad was detected. The trends found by AIM and NBO calculations (M = Pd, Pt; M′ = Ag, Au) are in accord with the results found for the series trans-5MCu (M = Ni, Pd, Pt). Within the respective series trans-5MCu, trans-5MAg, and trans-5MAu, increasing metal−metal interaction energies correlate with a high field shift at the 31 P nuclei of the bridging ligands μ-2-C 6 F 4 PPh 2 in the order Ni < Pd < Pt.
Compound trans-5PdAu shows similar BD*(Pd−C/P)← LP(Au) (10.17 kcal mol −1 ) and LP(Pd)→LV(Au) interactions (10.62 kcal mol −1 ). Therefore, the main character of the interaction should be described as predominantly metallic (shared) bonding. The complex cis-5PtCu and cis-5PtAg show similar metal−metal interactions as observed in the corre-sponding compounds trans-5PtCu and trans-5PtAg, respectively. The magnitude of |V(r b )|/G(r b ) is slightly lower for cis-5PtM′ as found in the respective trans-5PtM′ complex, which is indicative of a more pronounced ionic contribution in the cis configuration over trans. This is reflected in a higher NC at Cu over Ag in complexes with cis arrangement over trans leading to a higher coulomb repulsion. In conclusion, the metallophilic interactions can be described as predominantly metallic (shared) bonding (Ni−Cu, Pd−Au) with additional donor− acceptor bonding characteristic (Pd→Cu, Pd→Ag, Pt→Cu, Pt→Ag, Pt→Au).
Complexes of the Type [(Me 3 P)M(2-C 6 F 4 PPh 2 ) 2 M′Cl]. As described above, trans-1Ni and cis-1Pt show the ability to form complexes of the form 3M b (M = Ni, Pt), where one 2-C 6 F 4 PPh 2 ligand remains chelating and the other is dangling. Treatment of 3Pt b with 1 equiv CuCl or AgCl gave the expected complex cis-6PtCu and cis-6PtAg (Scheme 6), that proves the ability of the dangling PPh 2 unit to coordinate a coinage metal chloride.
The 31 P NMR spectra of both complexes show the expected three equally intense resonances, corresponding to the three inequivalent phosphorus nuclei. In both complexes, a respective signal appears at ca. −30 ppm, flanked by 195 Pt satellites of ca. 2200 Hz, and can be assigned to the PMe 3 ligand. The signals at −60 ppm, flanked by 195 Pt satellites of ca. 1600 Hz, can be assigned to the chelating κ 2 -2-C 6 F 4 PPh 2 ligand. The third resonance appears at 16.5 ppm for cis-6PtCu (flanked by 195 Pt satellites of ca. 340 Hz) and at 21.6 ppm for cis-6PtAg (split into a doublet of doublets; 600 Hz due to coupling with 107 Ag and 690 Hz due to coupling with 109 Ag, flanked by 195 Pt satellites of ca. 395 Hz) can be assigned to the bridging μ-2-C 6 F 4 PPh 2 ligand. In both cases, the 19 F NMR spectra show eight equally intense signals, confirming two different 2-C 6 F 4 PPh 2 ligand environments. The spectroscopic results show the same pattern as observed for the starting material 3Pt b (−6.8, 3 J Pt,P = 207 Hz, κC-PPh 2 ; −29.3, 1 J Pt,P = 2350 Hz, PMe 3 ; −59.9, 1 J Pt,P = 1540 Hz, κ 2 -PPh 2 ), supporting that the reaction only takes place at the dangling κC-2-C 6 F 4 PPh 2 ligand and no isomerization takes place.
The molecular structure was confirmed by single-crystal Xray diffraction analysis for cis-6PtCu and cis-6PtAg ( Figure 6 and Table S3; detailed structure description is given in the Supporting Information). The homologue gold compound cis-6PtAu was accessible from the reaction of cis-1Pt with 1 equiv [AuCl(PMe 3 )], in a similar manner to its PPh 3 analogue cis- do show a head-to-tail arrangement at both metal centers with a trans orientation about the Pt atom. Complex trans-7PtAu also undergoes isomerization to cis-6PtAu in solution, indicating the presence of an equilibrium between both isomers. This observation stands in contrast to the behavior of cis-[(Ph 3 P)Pt(κ 2 -2-C 6 F 4 PPh 2 )(μ-2-C 6 F 4 PPh 2 )AuCl], which did not show any isomerization product even after stirring the solution for 5 days. 18b This can be explained by PMe 3 being less sterically demanding as PPh 3 , which appears to favor isomerization. Also, upon mixing trans-1Pt with [AuCl(PPh 3 )], no reaction was observed. 18b In contrast, trans-1Pt with 1 equiv [AuCl-(PMe 3 )] in CH 2 Cl 2 does show slow reaction to trans-6PtAu, which immediately reacts further to trans-8PtAu. The reaction was followed by 31 P NMR spectroscopy. The 31 P NMR spectrum of the crude reaction mixture (Figure 7) shows the characteristic signals of the starting materials ([AuCl(PMe 3 )],

Inorganic Chemistry pubs.acs.org/IC
Article trans-1Pt) besides the signal sets for trans-6PtAu and trans-8PtAu. By addition of n-hexane to the crude reaction mixture, it was possible to grow crystals of trans-6PtAu and trans-8PtAu suitable for single-crystal X-ray diffraction (vide infra). A pure sample of trans-8PtAu was accessible by refluxing a mixture of trans-1Pt with 1 equiv [AuCl(PMe 3 )] in CH 2 Cl 2 for 4 days. The formation of trans-8PtAu from the reaction of trans-1Pt with 1 equiv [AuCl(PMe 3 )] is in agreement with the reaction behavior found for the palladium homologue; however, an intermediate species trans-6PdAu was not observed. 14 It was possible to determine molecular structures using single-crystal X-ray diffraction analysis for all four isomers cis-6PtAu, trans-6PtAu, trans-7PtAu, and trans-8PtAu ( Figure 8 and Table S3; detailed structure description is given in the Supporting Information). The 31 P NMR spectra of the corresponding reactions of 3Ni b with either CuCl or AgCl show two signals in a 2:1 ratio (Cu: 4.2, −4.7 ppm; Ag: 14.6 ( 1 J (109)AgP = 566 Hz, 1 J (107)AgP = 490 Hz), −6.1 ppm). For both complexes, the 19 F NMR spectra show four equally intense signals, which indicates a symmetric μ-2-C 6 F 4 PPh 2 environment. Therefore, the reaction led to the formation of the nickel homologous trans-8NiM′ (M′ = Cu, Ag, Scheme 7), which stands in contrast to the reaction behavior of 3Pt b with either CuCl or AgCl. The 31 −160.1 ppm). The 2 J P,P coupling constant of 330 Hz stays in good agreement with the trans Me 3 P−Ni-PPh 2 arrangement, indicating the main complex to be cis−trans-7NiAu, which was separated from the byproducts by fractional crystallization. It was possible to analyze the molecular structure of trans-8NiCu·acetone and cis−trans-7NiAu with single-crystal X-ray diffraction (Figure 9; detailed structure description is given in the Supporting Information).
Quantum chemical calculations were carried out to shed some light into the relative energy levels of the seven possible monomeric isomers of complexes of the type 6, 7, and 8 (Chart 2 and Table 5). For M = Ni, the isomers trans-8NiCu, trans-8NiAg, and cis−trans-7NiAu were experimentally ob-Scheme 7. Synthesis of Compounds of the Composition [(Me 3 P)NiCl(2-C 6 F 4 PPh 2 ) 2 M′] Alternatively, more complex scenarios (e.g., by coordinationdissociation effects of discrete solvent molecules or dimerization with the μ-Cl bridging mode) might also play reasonable roles and could potentially further lower the energy levels of the experimentally observed isomers. The isomer cis−trans-7MAu was predicted to be the thermodynamically most stable isomer for M = Ni, Pd, 14 and Pt. However, attempts to isolate the complex cis−trans-7PtAu by refluxing a toluene solution of cis-6PtAu or trans-7PtAu failed because of decomposition ( Figure S66). The metal−metal interactions in complexes of type 6, 7, and 8 (M = Ni, Pt; M′ = Cu, Au) were investigated in an analogous manner to the complexes of type 5 (vide supra). According to the NBO/NLMO calculations, the energy levels of the main donor-orbital of Pt with one PMe 3 ligand in their coordination sphere are ranging between −0.22 and −0.24 au (Table 6, Figure 10) and are very similar to the energy levels of the complexes of type 5 (−0.21 to −0.22 au, Table 3).
The Pt···Cu separation in cis-6PtCu is significant longer, by approximately 0.06 Å, in comparison to trans-5PtCu and cis-5PtCu, caused by steric repulsion of the chloride atom toward the ligand backbone. The topologic analysis (Table 7) of cis-6PtCu reveals a bcp between Pt and Cu with the same characteristics found in trans-5PtCu and cis-5PtCu , categorizing the Pt···Cu interaction in cis-6PtCu as predominantly metallic (shared) with an additional donor−acceptor bonding characteristic. However, the copper contribution in the main donor lone pair located at Pt in cis-6PtCu is about 0.29% and is similar to trans-5PtCu (0.24%) and cis-5PtCu (0.34%). Also, the lower acceptor orbital energy at copper (linear Cl−Cu−P arrangement) in cis-6PtCu (by about 0.1 au) in comparison to trans-5PtCu and cis-5PtCu (trigonal planar Cl−Cu−P 2 arrangement) leads to a less intense interaction energy (ΔE 2 ≈ 13 kcal mol −1 ). Therefore, a formal loss of Pt···Cu interaction intensity is present going from cis-6PtCu to either trans-5PtCu or cis-5PtCu, which is also reflected in lower ρ(r b ) in cis-6PtCu. This behavior of Pt··· Cu interaction intensity is visible by a shortening of the Pt···Cu separation in complexes of type 5 over 6 due to the absence of steric repulsion. In contrast, loss of interaction intensity between Pt···Cu caused by an increase in the coordination number of copper was observed between [(dppe)Pt(κC-2-C 6 F 4 PPh 2 )(μ-2-C 6 F 4 PPh 2 )CuCl] (linear Cl−Cu−P arrangement with significant donor−acceptor interactions) and [(dppe)Pt(μ-2-C 6 F 4 PPh 2 ) 2 CuCl] (trigonal planar Cl−Cu−P 2 arrangement absence of significant metallophilic interaction), where steric repulsion is present in both configurations. 15 Interestingly, in complexes with a trigonal planar ClM′(PR 3 ) 2 unit (type 5) and complexes with a linear ClM′PR 3 unit (type 6) a trend of stronger Pt→M′ donor−acceptor interaction with M′ = Au < Cu is observed.
In the series of Pt−Au complexes, the Pt···Au distances are decreasing in the order cis-6PtAu > trans-6PtAu > trans-5PtAu > trans-7PtAu > trans-8PtAu. Topologic analyses of the 5 complexes reveal the expected characterization at the bcps (1 < |V(r b )|/G(r b ) < 2, ∇ 2 ρ(r b ) > 0, G(r b ) ≙ |V(r b )|, H(r b ) < 0, WBO = 0.30−0.44), categorizing the Pt···Au interactions in PtAu complexes as predominantly metallic (shared) with an additional donor−acceptor bonding characteristic. The NCI descriptor and ELF increases in the expected trend from cis-6PtAu < trans-6PtAu < trans-5PtAu < trans-7PtAu < trans-8PtAu (Figures S67−S70) and is in agreement with the trend of the Pt···Au distances. The magnitudes of ρ(r b ) and ∇ 2 ρ(r b ) and WBO are increasing in the same order, which is indicative for an increase of strength of the metallophilic interactions.  The G(r b )/ρ(r b ) ratio is getting closer to 1, which suggests more intense donor−acceptor type interaction going from cis-6PtAu to trans-8PtAu. The magnitude of E NBO main LP→LV/ BD* interaction does follow the expected trend as observed by AIM analyses for complexes with a linear coordination sphere at Au (complexes type 6PtAu, 7PtAu, and 8PtAu, Figure 10). Complex cis−trans-7NiAu shows similar characteristics at the bcp as observed in the series of Pt−Au-complexes (1 < | V(r b )|/G(r b ) < 2, ∇ 2 ρ(r b ) > 0, G(r b ) ≙ |V(r b )|, H(r b ) < 0, WBO = 0.32). NBO analysis reveals a stronger Ni←Au donor−acceptor interaction over Ni→Au by 6.27 kcal mol −1 , which is also the case for trans-7PtAu (ΣE NBO (Pt→Au) < ΣE NBO (Pt←Au) by 1.18 kcal mol −1 ). In complexes of type 7, the gold atom is a stronger donor over the d 8 -metal (see the Supporting Information for details).
NBO calculations at trans-8NiCu reveal that the Cl atom acts as a lone pair donor toward the copper center (∑E NBO/LP(Cl)→LV(Cu) = 30.78 kcal mol −1 ). The NC at Cl (−0.61) and Cu (0.72) supports the presence of attractive coulomb interaction. We find the Cl atom to be covalently bound to nickel and showing attractive dative bonding modes toward copper (WBO Cu−Cl = 0.77 vs WBO Ni−Cl = 1.01), which is in agreement with the interatomic distances. The Cu1···O1 distance is approximately 2.20(1) Å and significantly longer than the sum of the covalent radii (1.98(4) Å). 20    The presence of formal weak Cu1···O1 interactions was also observed in complexes with O�PPh(C 6 H 4 PPh 2 ) 2 as the donor in the copper coordination sphere. 28 The calculation of the NCI descriptor for trans-8NiCu·acetone underlines the presence of attractive NCI between Cu1···Cl1 and Cu1···O1 ( Figure 11).
The Ni···Cu separation in trans-8NiCu is significantly shorter than in compound trans-5NiCu (by 0.1 Å), which can be explained by an additional chloride bridge in trans-8NiCu. However, AIM calculations did not support a bcp between Ni and Cu in trans-8NiCu. We address this absence of a bcp with three parameters: (i) Cu is distorted tetrahedral coordinated by P, Cl, and O donor atoms and is therefore sterically saturated for weak covalent interactions, (ii) the NCs for Ni and Cu are significantly higher in magnitude and therefore the Ni and Cu are more positively charged in comparison to other complexes presented in this study, and (iii) the donor−acceptor possibility is reduced by a comparative large gap between donor and acceptor orbital energy levels. According to the NCI descriptor, the Ni···Cu interaction should be described as very weak attractive noncovalent intermetallic (dispersion) interaction.