Metal-metal interactions in multinuclear systems determine the course of their chemical reactions, with implications in industrial and biological catalysis, the construction of functional materials, as well as in the understanding of fundamental chemical principles (e.g., metal-metal bonding). In addition to metal-metal distance and orientation, the role of peripheral ligands is often critical to the reaction outcome, as the steric bulk of groups proximal to the metal allows or prevents the formation of certain products, and the energy of the metal-based frontier orbitals is tuned by the resonance and inductive effects exerted by these groups. Two-electron oxidative addition/reductive elimination reactions across dinuclear or trinuclear late transition metal centers continue to attract considerable interest [1][2][3]; they can lead to either homovalent or mixed-valent products, depending on the distance between and relative orientation of the redox centers. For example, symmetric two-electron-two-center (2e-2c) oxidative addition across "face-to-face" Au I 2 , Pt II 2 , or Au I Pt II Au I , has led to homovalent products, Au II 2 , Pt III 2 , and Au II Pt II Au II , respectively, containing metal-metal bonds (Scheme 1A-C) [4][5][6][7]. On the other hand, when the bridging ligands tilt the coordination planes of the metal centers relative to each other, stepwise 2e-1c reduction of a triangular Au III 3 Cl 6 complex led to unsymmetrical, mixedvalent Au I Au III 2 Cl 4 and Au I 2 Au III Cl 2 products (Scheme 1D). At the same time, a similar photochemical process has been reported for an Au III 2 Cl 4 generating Au I Au III Cl 2 and Au I 2 [8][9][10][11][12]. Mixed-valent 2e-1c oxidative addition products also result in the "face-to-face"
In 2006, Umakoshi et al. [18] prepared a delocalized mixed-valent Pt III 2Pt II comple namely, pyrazolato-bridged platinum cyclic trimer [Pt3(μ-pz)6Br2], by the two-electron ox dation of its yet not structurally characterized homovalent Pt II 3 precursor. This findin prompted us to reinvestigate dinuclear and trinuclear Pt II -complexes with tilted coordin tion planes, maintained by bridging pyrazolates [19]. We have employed two pyrazole lig ands containing alkyl substituents: one presenting no steric hindrance to the donor N atom and one containing bulky tert-butyl groups, the latter forcing a close contact between meth groups and the N-coordinated metal (Scheme 2). Here, we present the synthesis, structur and spectroscopic characterization, and theoretical studies of three new complexes involv ing two pyrazole ligands: 4-Me-pzH and 3,5t Bu2-pzH (pzH = pyrazole), the latter capab of cyclometallating via its t Bu group. Specifically, the triangular complex [Pt II (μ-4-M pzH)2]3, 1, the dinuclear orthometallated complex [Pt II (μ-3,5t Bu2-pz)(κ 2 -N,C-1-H-5t Bu-CMe2CH2-pzH)]2, 2, and its two-electron oxidation product [Pt III 2(μ-3,5-t Bu2-pz)2(κ 2 -N,C-H-3-CH2Me2CH2-5t Bu-pzH)(κ 2 -N,C-3-CCH2Me2-5t Bu-pz)Cl], 3, are discussed. Scheme 1. Oxidative addition/reductive elimination across dinuclear and trinuclear complexes.
In 2006, Umakoshi et al. [18] prepared a delocalized mixed-valent Pt III 2 Pt II complex, namely, pyrazolato-bridged platinum cyclic trimer [Pt 3 (µ-pz) 6 Br 2 ], by the two-electron oxidation of its yet not structurally characterized homovalent Pt II 3 precursor. This finding prompted us to reinvestigate dinuclear and trinuclear Pt II -complexes with tilted coordination planes, maintained by bridging pyrazolates [19]. We have employed two pyrazole ligands containing alkyl substituents: one presenting no steric hindrance to the donor N atoms, and one containing bulky tert-butyl groups, the latter forcing a close contact between methyl groups and the N-coordinated metal (Scheme 2). Here, we present the synthesis, structural and spectroscopic characterization, and theoretical studies of three new complexes involving two pyrazole ligands: 4-Me-pzH and 3,5-t Bu 2 -pzH (pzH = pyrazole), the latter capable of cyclometallating via its t Bu group. Specifically, the triangular complex [Pt II (µ-4-Me-pzH) 2 ] 3 , 1, the dinuclear orthometallated complex [Pt II (µ-3,5-t Bu 2 -pz)(κ 2 -N,C-1-H-5-t Bu-3-CMe 2 CH 2 -pzH)] 2 , 2, and its two-electron oxidation product [Pt III 2 (µ-3,5-t Bu 2 -pz) 2 (κ 2 -N,C-1-H-3-CH 2 Me 2 CH 2 -5-t Bu-pzH)(κ 2 -N,C-3-CCH 2 Me 2 -5-t Bu-pz)Cl], 3, are discussed. Scheme 2. Two pyrazole ligands employed here and cyclometallation mode.
Complex 1 was prepared by the same method used for the synthesis of the analogous Pd-complex, [Pd II (μ-3-Ph-pz)2]3 [20]. Deprotonation of the pyrazole ligands of trans-[PtCl2(4-Me-pzH)2] initiated the cyclization of the homoleptic trimer (Scheme 3A). However, the analogous reaction involving 3,5t Bu2-pzH resulted in the cyclometallated dimer 2 (Scheme 3B). Complex 3 was prepared by the oxidation of 2 by one equivalent of the oxidizing agent (Scheme 3C). The orientation of pyrazolido anion electron donor orbitals favors the formation of triangular species, as long as bulky 3,5-pyrazole substituents do not sterically hinder these species; these tendencies are well documented in the literature [21][22][23]. In that light, the formation of 1 upon pyrazole deprotonation by a base is unexceptional. In contrast, whereas non-triangular products were expected upon deprotonation of trans-[PtCl2(3,5t Bu-pzH)2], the cyclometallation of a butyl group is noteworthy: cyclometallation reactions involving activated C-H bonds (typically of aromatic rings or heterocycles) have been reported for 4d and 5d transition metals [24][25][26], including platinum [27][28][29]. However, to the best of our knowledge, cyclometallation of a saturated aliphatic group has not been hitherto reported, even though the activation of C-H bonds by platinum is well established in the literature [30,31]. Oxidative addition to a diplatinum(II) complex containing bridging pyrazolates and chelating/orthometallated ligands, similar to the oxidation of 2 to 3 here, has recently been reported also by others [14]. Complexes 1-3 were structurally characterized by single-crystal X-ray crystallography. Selected distances and angles pertaining to 1-3 are listed in Table 1. Scheme 2. Two pyrazole ligands employed here and cyclometallation mode.
Complex 1 was prepared by the same method used for the synthesis of the analogous Pd-complex, [Pd II (µ-3-Ph-pz) 2 ] 3 [20]. Deprotonation of the pyrazole ligands of trans-[PtCl 2 (4-Me-pzH) 2 ] initiated the cyclization of the homoleptic trimer (Scheme 3A). However, the analogous reaction involving 3,5-t Bu 2 -pzH resulted in the cyclometallated dimer 2 (Scheme 3B). Complex 3 was prepared by the oxidation of 2 by one equivalent of the oxidizing agent (Scheme 3C). The orientation of pyrazolido anion electron donor orbitals favors the formation of triangular species, as long as bulky 3,5-pyrazole substituents do not sterically hinder these species; these tendencies are well documented in the literature [21][22][23]. In that light, the formation of 1 upon pyrazole deprotonation by a base is unexceptional. In contrast, whereas non-triangular products were expected upon deprotonation of trans-[PtCl 2 (3,5-t Bu-pzH) 2 ], the cyclometallation of a butyl group is noteworthy: cyclometallation reactions involving activated C-H bonds (typically of aromatic rings or heterocycles) have been reported for 4d and 5d transition metals [24][25][26], including platinum [27][28][29]. However, to the best of our knowledge, cyclometallation of a saturated aliphatic group has not been hitherto reported, even though the activation of C-H bonds by platinum is well established in the literature [30,31]. Oxidative addition to a diplatinum(II) complex containing bridging pyrazolates and chelating/orthometallated ligands, similar to the oxidation of 2 to 3 here, has recently been reported also by others [14]. Complexes 1-3 were structurally characterized by single-crystal X-ray crystallography. Selected distances and angles pertaining to 1-3 are listed in Table 1.
Complex 1 crystallized in the triclinic space group P-1 with two molecules of 1 and onehalf interstitial acetone solvent molecule per asymmetric unit. The two crystallographically independent molecules of 1 do not differ statistically from each other; both show minor deviation from ideal D 3h symmetry (Figure 1). The existence of a single set of resonances for the six pyrazolido ligands in its 1 H-NMR spectrum shows that the trimeric structure persists in solution (Figure S1). The Pt centers are in an approximate square planar N 4 environment with the Pt atoms deviating by 0.14-0.15 Å from the best-fit planes of the four N atoms in the direction away from the center of the metallacyclic ring. The intramolecular Pt ... Pt distances average 3.0511(6) Å, being statistically indistinguishable from the 3.048(1) Å distance of the corresponding unsubstituted pyrazole complex, [Pt(µ-pz) 2 ] 3 [32], and quite similar to the average Pd ... Pd distances of 3.054(1) Å in [Pd(µ-3-Ph-pz) 2 ] 3 , 3.0471(3) Å in [Pd(µ-pz) 2 ] 3 , and 3.0458(4) Å in [Pd(µ-4-Me-pz) 2 ] 3 (the ionic radii of Pt II and Pd II differ by 0.06 Å) [20,33,34]. nar N4 environment with the Pt atoms deviating by 0.14-0.15 Å from the best-fit planes of the four N atoms in the direction away from the center of the metallacyclic ring. The intramolecular Pt ... Pt distances average 3.0511(6) Å, being statistically indistinguishable from the 3.048(1) Å distance of the corresponding unsubstituted pyrazole complex, [Pt(μpz)2]3 [32], and quite similar to the average Pd ... Pd distances of 3.054(1) Å in [Pd(μ-3-Phpz)2]3, 3.0471(3) Å in [Pd(μ-pz)2]3, and 3.0458(4) Å in [Pd(μ-4-Me-pz)2]3 (the ionic radii of Pt II and Pd II differ by 0.06 Å) [20,33,34]. Complex 2 crystallized in the chiral orthorhombic space group P212121 with a whole molecule of C2 symmetry per asymmetric unit. The structure consists of two Pt atoms, two bridging 3,5t Bu2-pyrazolato groups, and two chelating cyclometallated 1-H-3-CMe2CH2-5t Bu-pzH ligands, the latter forming five-membered chelates with each Pt center (Figure 2). The Pt atoms are approximately square-planar with an N3C-coordination environment. The μ-3,5t Bu2-pz ligands bridge the metals unsymmetrically, one pyrazole leaning towards one Pt atom (Pt-N = 2.026(7), 2.157(7) Å) and the other leaning the opposite way (Pt-N = 2.151(7), 2.034(7) Å). The solution 1 H-NMR spectrum of 2 (Figure 3) is consistent with its solid-state structure. There are four resonances for the cyclometallated butylgroups: two singlets for the diastereotopic Me groups (1.26 and 0.84 ppm) and two doublets for the diastereotopic, geminal H atoms of CH2 groups (2.28 and 1.65 ppm); the 195 Pt satellites are not observed in this ambient temperature spectrum due to the broadening Complex 2 crystallized in the chiral orthorhombic space group P2 1 2 1 2 1 with a whole molecule of C 2 symmetry per asymmetric unit. The structure consists of two Pt atoms, two bridging 3,5-t Bu 2 -pyrazolato groups, and two chelating cyclometallated 1-H-3-CMe 2 CH 2 -5-t Bu-pzH ligands, the latter forming five-membered chelates with each Pt center (Figure 2). The Pt atoms are approximately square-planar with an N 3 C-coordination environment. The µ-3,5-t Bu 2 -pz ligands bridge the metals unsymmetrically, one pyrazole leaning towards one Pt atom (Pt-N = 2.026(7), 2.157(7) Å) and the other leaning the opposite way (Pt-N = 2.151(7), 2.034(7) Å). The solution 1 H-NMR spectrum of 2 (Figure 3) is consistent with its solidstate structure. There are four resonances for the cyclometallated butyl-groups: two singlets for the diastereotopic Me groups (1.26 and 0.84 ppm) and two doublets for the diastereotopic, geminal H atoms of CH 2 groups (2.28 and 1.65 ppm); the 195 Pt satellites are not observed in this ambient temperature spectrum due to the broadening attributed to the coordination of three quadrupolar N atoms. The Pt ... Pt distance of 2.9290(5) Å in 2 is significantly longer than the one determined in a related Pt-(µ-3,5-t Bu 2 -pz) 2 -Pt complex, 2.8343(6) Å, containing also 2-(2,4-difluorophenyl)pyridyl chelating ligands [35]. Inspection of a molecular model of 2 shows that the platinum coordination planes are slightly bent to bring the chelating ligands closer to each other than they might have been in an ideal square planar arrangement. In contrast, a much shorter approach between the chelating ligands is found in complex 3 (vide infra). Both observations point to Coulombic repulsion between the Pt centers as the more likely explanation for this distortion, rather than the steric repulsion between the chelating cyclometallated ligands.
Complex 3 crystallized in the triclinic space group P-1 with two molecules per asymmetric unit, accompanied by four interstitial H 2 O molecules at chemically insignificant sites. The structure of 3 retains the basic features of 2, but with a Pt-Pt separation of 2.584(3) Å and 2.586(2) Å, corresponding to a formal single metal-metal bond and one chloride coordinated trans to it (Figure 4). The C 1 molecular symmetry of 3 is reflected in its 1 H-NMR spectrum (Figure 5) showing a doubling of the number of resonances recorded for 2, in addition to a downfield shift of all resonances, consistent with the increase in its oxidation state. Electroneutrality requires the presence of a crystallographically invisible proton on one of the two non-coordinated N atoms of 3; this proton is evident in the 1 H-NMR by a broad resonance at 8.63 ppm whose integrated area corresponds to one H atom per molecule of 3. The absence of paramagnetically shifted resonances in 3 is consistent with its Pt III 2 assignment and the presence of a Pt-Pt bond of 2.585 Å. The latter bond length is shorter than the corresponding unsupported bonds of 2.694(1) Å, 2.6964(5) Å, and 2.726 Å [36][37][38] reported earlier, but within the range of several ligand-bridged diplatinum(III) species [39,40]. The Pt III 2 oxidation state assignment is further supported by a comparison of the 4f electron binding energies of 2 and 3 determined by X-ray photoelectron spectroscopy (Figure S2) and a comparison with the corresponding binding energies of Pt IV species reported in the literature (Table 2). The experimental XPS peaks of 3 are deconvoluted into two equal components, attributed to its two distinct Pt sites, both with higher binding energies than those of 2 and lower than the literature values for Pt IV compounds [41,42].
Chemistry 2023, 5, FOR PEER REVIEW 6 attributed to the coordination of three quadrupolar N atoms. The Pt ... Pt distance of 2.9290(5) Å in 2 is significantly longer than the one determined in a related Pt-(μ-3,5t Bu2pz)2-Pt complex, 2.8343(6) Å, containing also 2-(2,4-difluorophenyl)pyridyl chelating ligands [35]. Inspection of a molecular model of 2 shows that the platinum coordination planes are slightly bent to bring the chelating ligands closer to each other than they might have been in an ideal square planar arrangement. In contrast, a much shorter approach between the chelating ligands is found in complex 3 (vide infra). Both observations point to Coulombic repulsion between the Pt centers as the more likely explanation for this distortion, rather than the steric repulsion between the chelating cyclometallated ligands. Complex 3 crystallized in the triclinic space group P-1 with two molecules per asymmetric unit, accompanied by four interstitial H2O molecules at chemically insignificant sites. The structure of 3 retains the basic features of 2, but with a Pt-Pt separation of 2.584(3) Å and 2.586(2) Å, corresponding to a formal single metal-metal bond and one chloride coordinated trans to it (Figure 4). The C1 molecular symmetry of 3 is reflected in its 1 H-NMR spectrum (Figure 5) showing a doubling of the number of resonances recorded for 2, in addition to a downfield shift of all resonances, consistent with the increase in its oxidation , FOR PEER REVIEW 6 attributed to the coordination of three quadrupolar N atoms. The Pt ... Pt distance of 2.9290(5) Å in 2 is significantly longer than the one determined in a related Pt-(μ-3,5t Bu2pz)2-Pt complex, 2.8343(6) Å, containing also 2-(2,4-difluorophenyl)pyridyl chelating ligands [35]. Inspection of a molecular model of 2 shows that the platinum coordination planes are slightly bent to bring the chelating ligands closer to each other than they might have been in an ideal square planar arrangement. In contrast, a much shorter approach between the chelating ligands is found in complex 3 (vide infra). Both observations point to Coulombic repulsion between the Pt centers as the more likely explanation for this distortion, rather than the steric repulsion between the chelating cyclometallated ligands. Complex 3 crystallized in the triclinic space group P-1 with two molecules per asymmetric unit, accompanied by four interstitial H2O molecules at chemically insignificant sites. The structure of 3 retains the basic features of 2, but with a Pt-Pt separation of 2.584(3) Å and 2.586(2) Å, corresponding to a formal single metal-metal bond and one chloride coordinated trans to it (Figure 4). The C1 molecular symmetry of 3 is reflected in its 1 H-NMR spectrum (Figure 5) showing a doubling of the number of resonances recorded for 2, in addition to a downfield shift of all resonances, consistent with the increase in its oxidation A structural comparison of 2 and 3 shows that the shortening of the separation between the two metal centers, brought about by the formation of a Pt-Pt bond, is accompanied by a decrease in the dihedral angles formed between the µ-3,5-t Bu 2 -pz ligands from 102.5 • in 2 to 92.1 • and 95.9 • in 3. The pyrazole-pyrazole dihedral angles of both 2 and 3 are more acute than the 110.6 • (average) angle of the less sterically hindered 1. This agrees with the earlier observation that the Pt-Pt separation in a series of Pt-(µ-3,5-R 2 -pz) 2 -Pt complexes increases as the steric bulk of the bridging ligands decreases [35]. The electronic spectra of 1, 2, and 3 each contain an intense UV band with λ max at 225-230 nm, attributed to π-π* transitions. However, compound 3 shows five additional bands spanning the UV to NIR range-361 nm, 531 nm, 587 nm, 761 nm, and 818 nm-attributed to states arising from the Pt-Pt bonding manifold (Figure S4). To further probe the bonding in the complexes, density functional theory (DFT) calculations were carried out.
The optimized geometries agree with the corresponding experimental X-ray structural values of 1-3 (Table 2). We ascertained that all optimized geometries exhibited no imaginary frequency. The computed bond length is slightly larger (0.02-0.10 Å) than the experimental data, since the molecules were optimized in the gas phase, and there is no interaction with other complex units as in the crystalline phase.
The DFT-simulated infrared spectra (IR) of the three complexes are shown in Figure S3. The infrared (IR) spectrum of complex 1 is clearly distinguished from those of 2 and 3 due to its different structure. Complexes 2 and 3 have approximately the same IR distribution, as they share similar structural features, except for the additional Cl atom in 3. A calculated mode of 286.6 cm -1 , assigned to Pt-Cl stretching in complex 3, falls outside the experimentally accessible spectral window. No Pt-Pt interaction mode is identified in complex 2.
The HOMO-LUMO gaps of the Pt complexes computed by BP86 functional are 0.26, 2.74, and 0.94 eV for 1, 2, and 3, respectively. The frontier molecular orbitals of Pt complex 1, 2, and 3 are presented in Figure S5. The d z 2 orbitals of Pt atoms dominate the HOMO of complexes 1 and 2, and Pt d-orbitals also contribute to their HOMO-1, HOMO-2, LUMO, LUMO+1, and LUMO+2. However, for complex 3, the d z 2 orbitals of Pt and the p orbital of Cl contribute to the LUMO, the π orbital of pyrazole, d orbital of Pt, and p orbital of Cl contribute to the HOMO, and the π orbital of pyrazole mainly contribute to the HOMO-1, HOMO-1, LUMO+1, and LUMO+2. The LUMO of 2 consists of d xy and d xz of Pt atoms, and the overlap of the two orbitals exhibit a σ-bonding character, resulting in a shorter Pt-Pt distance than 1.
To help understand the bonding characteristics of these complexes, we calculated the natural population analysis (NPA) charges and the Wiberg bond index (WBI) based on natural bonding orbital (NBO) computations at the BP86/6-31G*~dz level of theory. The Pt atoms are positively charged (ca. 0.43, 0.44, and 0.47 |e| in 1; 0.35 |e| in 2; and 0.46, 0.56 |e| in 3), and the N/C/Cl atoms are negatively charged (for details see Table S4). Correspondingly, the natural electron configurations for Pt, N, Cl, and C (bonded to Pt) atoms are listed in Table S4. Pt atoms transfer considerable charge (ca. ~0.5 e) to the N/C/Cl atoms of all three complexes. The computed WBIs (Table S5) for Pt-Pt in 3 (0.28) are strikingly larger than that in 1 (0.08~0.09) and 2 (0.06), contributing to the shorter Pt-Pt distance in 3; the WBIs for Pt-N are comparable in the three complexes. These indicate that partial bonds form between Pt atoms and their surrounding Pt/N/C/Cl atoms, along with a weak σ-interaction in 2 and a formal Pt-Pt single bond in 3.
QTAIM topological analysis of the electronic density [43][44][45] gave further details of the bonding in the three Pt complexes. For simplicity, we substituted the methyl groups by H, and the BP86/6-31G*~dz optimized results provide almost the same structural parameters as the initial configurations. Figure 6 depicts the simplified complexes' molecular graphs (at BP86/6-31G*~dz) representing Pt-Pt/Pt-N/Pt-C interactions. The bond critical points (BCPs) between Pt atoms for all three complexes lead to 3, 0, and 1 bond paths for the 1, 2, and 3, respectively. The larger 0.05 au ρ bcp electron densities at the Pt-Pt bond critical points (BCPs) for complex 3 compared to the value of 0.02 au for complex 1 suggest stronger bonds, consistent with the shorter Pt-Pt distance in complex 3. Note that there is no BCP between Pt and Pt in complex 2, leading to the longer Pt-Pt distance compared to complex 3. In addition, the low electron densities ρ bcp (0.11~0.13 au, 0.09~0.14 au, and 0.08~0.14 au in 1, 2, and 3, respectively), as well as negative Laplacian ∇ 2 ρ bcp (-0.11~-0.14 au, -0.05~-0.15 au, and -0.04~-0.13 au for 1, 2 and 3, respectively), located at the Pt-N/Pt-C/Pt-Cl BCPs, indicate ionic interactions and limited contributions to the total stability. stronger bonds, consistent with the shorter Pt-Pt distance in complex 3. Note that there is no BCP between Pt and Pt in complex 2, leading to the longer Pt-Pt distance compared to complex 3. In addition, the low electron densities ρbcp (0.11 ~ 0.13 au, 0.09 ~ 0.14 au, and 0.08 ~ 0.14 au in 1, 2, and 3, respectively), as well as negative Laplacian ∇ 2 ρbcp (-0.11 ~ -0.14 au, -0.05 ~ -0.15 au, and -0.04 ~ -0.13 au for 1, 2 and 3, respectively), located at the Pt-N/Pt-C/Pt-Cl BCPs, indicate ionic interactions and limited contributions to the total stability.
Commercial reagents-K2PtCl4, 4-Me-pyrazole, pivaloylmethane, and hydrazinewere used as received. 3,5-Di-tert-butyl-pyrazole (3,5t Bu2-pzH) was prepared by refluxing equivalent amounts of dipivaloylmethane and hydrazine in 95% EtOH. Trans-[PtCl2(4-Me-pzH)2] and trans-[PtCl2(3,5t Bu2-pzH)2] were prepared quantitatively by stoichiometric addition of two equivalents of 4-Me-pzH, or 3,5t Bu2-pzH, to K2PtCl4 in MeOH/H2O and characterized by X-ray structure determination (Tables S1-S3). Solid p-Cl-C6H4-ICl2 was prepared by bubbling gaseous Cl2 through a solution of p-Cl-C6H4-I in toluene and collecting the product by filtration after washing with toluene and diethylether (Caution! The reaction should be carried out under a fume hood with a solution of a base trapping excess Cl2). Solvents were purified by standard methods [46]. 1 H-NMR spectra were recorded with a Bruker Avance DPX-400 spectrometer. 13 C-NMR resonances could not be safely distinguished from baseline noise due to solubility limitations. The electronic spectra of the complexes in solution were recorded on a Varian CARY 500 spectrophotometer in the 40,000-4000 cm -1 (250-2500 nm) range. Elemental analyses were performed by Galbraith Laboratories, Inc., Knoxville, TN.
[Pt II (μ-4-Me-pz)2]3, 1: To a CH3CN solution (50 mL) of trans-[PtCl2(4-Me-pzH)2] (230 mg, 0.53 mmol) was added Et3N (162 mg, 1.60 mmol), and the solution was refluxed for 8 h. Yellow solid precipitated and was removed by filtration while the solution was still hot. The colorless filtrate was concentrated under air, yielding a white microcrystalline solid, which was collected and air-dried; Yield, 55 mg (29%). Colorless crystals of 1, suitable for X-ray analysis, were grown from MeOH/CH3COCH3. Anal. Calcd. for C24N12H30Pt3: C, 26.89; H, 2.82; N, 15.68%. Found: C, 27.14; H, 2.79; N, 15.93%. S1-S3). Solid p-Cl-C 6 H 4 -ICl 2 was prepared by bubbling gaseous Cl 2 through a solution of p-Cl-C 6 H 4 -I in toluene and collecting the product by filtration after washing with toluene and diethylether (Caution! The reaction should be carried out under a fume hood with a solution of a base trapping excess Cl 2 ). Solvents were purified by standard methods [46]. 1 H-NMR spectra were recorded with a Bruker Avance DPX-400 spectrometer. 13 C-NMR resonances could not be safely distinguished from baseline noise due to solubility limitations. The electronic spectra of the complexes in solution were recorded on a Varian CARY 500 spectrophotometer in the 40,000-4000 cm -1 (250-2500 nm) range. Elemental analyses were performed by Galbraith Laboratories, Inc., Knoxville, TN.
[ X-ray diffraction data were collected with a Bruker AXS SMART 1K CCD diffractometer [47], using graphite-monochromated Mo-Kα radiation at ambient temperature from single crystals mounted atop glass fibers at random orientation. Data were corrected for Lorentz and polarization effects [48]. The structures were solved employing the SHELXTL-direct methods program and refined by full-matrix least-squares on F 2 [49]. Crystallographic details for 1, 2, and 3 are summarized in Table 3. The Gaussian 09 software package was employed throughout our density functional theory (DFT) computations [50]. Full geometry optimizations for the three complexes were carried out using the BP86 functional [51,52]. The 6-31G* basis set for C, N, H, and Cl atoms and a double-ζ basis set (LanL2DZ) with the effective core potential (ECP) for Pt (denoted here by 6-31G*∼dz) were used. All the optimized geometries were characterized as true local minima by harmonic vibrational frequency analysis at the same theoretical level. Atomic charges were based on the Natural Population Analysis (NPA) of Weinhold et al. [53]. To gain more insights into the chemical bonding, we performed a quantum theory of atoms in molecules (QTAIM) [43][44][45] study, using the all-electron basis set (6-31G* for C, N, Cl, and H; double zeta plus polarization function basis set, Douglas-Kroll-Hess for Pt) [54] by AIM2000 software [55]. Natural bond orbital NBO population analysis was used to describe the details of chemical bonding in the systems studied.
Differences in the steric bulk of peripheral substituents between the two pyrazoles employed here determine the topology of the resulting pyrazolato products, yielding upon deprotonation the new trinuclear homoleptic complex 1, or the dinuclear 2. Platinum(II) complexes involving less sterically crowded, 3,5-Me2-pzH and 3-t Bu-pzH ligands have been employed in the stepwise construction of multinuclear heterometallic complexes via straightforward coordination chemistry [56,57]. In contrast, the bulky 3,5-t Bu 2 groups employed here apparently prevent the formation of a trimeric ring, leaving the Pt center coordinatively unsaturated. The latter satisfies the four-coordination requirement of Pt II via the unprecedented cyclometallation of a tert-butyl group (C-H BDE of ~100 kcal/mol), suggesting possible further applications of 3,5-t Bu 2 -pzH in C-H activation chemistry and catalysis. Dinuclear half-lantern Pt II complexes, with even longer Pt . . . Pt separation than 2, have been studied in detail with regard to their tunable visible luminescence [35,[58][59][60][61][62][63]. In contrast, compound 2 does not luminesce; this is tentatively attributed to the proximity to Pt atoms of the (cyclometallated) C-H group, whose vibrational modes can quench the excited state. The facile oxidative addition of complex 2 to a Pt III 2 product was expected. However, the unsymmetrical addition of chloride across the Pt-Pt single bond of 3, while not unprecedented [14], is a rare example of this type of reactivity. The chemistry of fivecoordinate (i.e., coordinatively unsaturated) Pt III centers, such as one of the two Pt centers of 3, remains unexplored, to date.