Reactions of Ph₃Sb=S with Copper(I) Complexes Supported by N-Donor Ligands: Formation of Stable Adducts and S-Transfer Reactivity

Abstract

In the exploration of sulfur-delivery reagents useful for synthesizing models of the tetracopper-sulfide cluster of nitrous oxide reductase, reactions of Ph₃Sb=S with Cu(I) complexes of N,N,N’,N’-tetramethyl-2R,3R-cyclohexanediamine (TMCHD) and 1,4,7-trialkyltriazacyclononanes (R₃tacn; R = Me, Et, iPr) were studied. Treatment of [(R₃tacn)Cu(NCCH₃)]SbF₆ (R = Me, Et, or iPr) with one equivalent of S=SbPh₃ in CH₂Cl₂ yielded adducts [(R₃tacn)Cu(S=SbPh₃)]SbF₆ (1–3), which were fully characterized, including by X-ray crystallography. The adducts slowly decayed to [(R₃tacn)₂Cu₂(µ-η²: η²-S₂)]²⁺ species (4–6) and SbPh₃, or more quickly in the presence of additional [(R₃tacn)Cu(NCCH₃)]SbF₆ to 4–6 and [(R₃tacn)Cu(SbPh₃)]SbF₆ (7–9). The results of mechanistic studies of the latter process were consistent with rapid intermolecular exchange of S=SbPh₃ between 1–3 and added [(R₃tacn)Cu(NCCH₃)]SbF₆, followed by conversion to product via a dicopper intermediate formed in a rapid pre-equilibrium step. Key evidence supporting this step came from the observation of saturation behavior in a plot of the initial rate of loss of 1 versus the initial concentration of [(Me₃tacn)Cu(NCCH₃)]SbF₆. Also, treatment of [(TMCHD)Cu(CH₃CN)]PF₆ with S=SbPh₃ led to the known tricopper cluster [(TMCHD)₃Cu₃(µ₃-S)₂](PF₆)₃ in good yield (79%), a synthetic procedure superior to that previously reported (Brown, E. C.; York, J. T.; Antholine, W. E.; Ruiz, E.; Alvarez, S.; Tolman, W. B. J. Am. Chem. Soc. 2005, 127, 13752–13753).

Introduction

The environmentally important reduction of N₂O to N₂ is catalyzed enzymatically by nitrous oxide reductase,¹ which has been shown on the basis of X-ray crystallography² and other spectroscopic techniques³ to contain a unique tetracopper-sulfide cluster supported by multiple histidine residues in its active site. The novel structure of this cluster and hypotheses for the mechanism of N₂O coordination and reduction at its copper centers⁴ has stimulated extensive efforts to create copper-sulfur model complexes supported by N-donor ligands.^5,6 These efforts have focused on the use of relatively few types of sulfur-containing reagents (e.g., S₈ and Na₂S₂) in reactions with Cu(I) and Cu(II) species. Studies of the copper-sulfur compounds characterized so far have raised interesting bonding questions,^7,8 and, in one instance, have led to the discovery of reactivity with N₂O.⁶ Nonetheless, only a limited array of N-donor ligated copper-sulfur motifs have been characterized and an accurate model of the nitrous oxide active site has yet to be constructed.

In seeking to broaden the scope of available copper-sulfur complexes as a means to address mechanistic and electronic structural issues relevant to the enzyme active site, we are exploring reactions of copper precursors with an expanded array of sulfur-containing reagents. Several studies have shown that triphenyl antimony sulfide (Ph₃Sb=S) is useful for sulfur transfer reactions,⁹ including for the preparation of transition metal sulfide complexes.^10,11 The utility of Ph₃SbS likely stems from the thermodynamic instability of the Sb-S bond, which is weaker than the related bonds in R₃E=S (E = P or As) congeners.¹² Herein, we report the results of an investigation of the reactivity of Ph₃Sb=S with selected Cu(I) complexes of N,N,N’,N’-tetramethyl-2R,3R-cyclohexanediamine (TMCHD) and 1,4,7-trialkyltriazacyclononanes (R₃tacn; R = Me, Et, iPr). Key findings include the isolation and structural characterization of novel LCu(I)-S=SbPh₃ (L = R₃tacn; R = Me, Et, iPr) adducts, which are the first examples of transition metal complexes bound to Ph₃Sb=S to be structurally characterized by X-ray crystallography.¹³ These complexes subsequently decay cleanly to [Cu₂(µ-η²:η²-S₂)]²⁺ species, particularly when treated with additional [(R₃tacn)Cu(CH₃CN)]SbF₆, and mechanistic insights for this process were obtained through kinetics studies. In addition, by using, Ph₃Sb=S an improved synthetic route to the [Cu₃S₂]³⁺ core⁸ was discovered.

Experimental Section

General Considerations

All solvents and reagents were obtained from commercial sources and used as received unless otherwise noted. The solvents CH₂Cl₂, pentane, and Et₂O were dried over CaH₂ and distilled under vacuum or passed through solvent purification columns (Glass Contour, Laguna, CA). The complexes [(R₃tacn)Cu(CH₃CN)]SbF₆ (R = Me,¹⁴ Et,¹⁵ iPr1¹⁶), [(R₃tacn)Cu(CH₃CN)]BPh₄ (R = Et¹⁵ or iPr¹⁷), [(Me₃tacn)₂Cu₂(S₂)](SbF₆)₂,¹⁴ and [(TMCHD)Cu(CH₃CN)]PF₆ ¹⁸ were prepared according to published procedures. All metal complexes were prepared and stored in a glovebox under a dry N₂ atmosphere. Triphenylantimony sulfide (Ph₃Sb=S) and 2,3-dimethylbutadiene were purchased from Strem and Aldrich, respectively, and were used without purification. NMR spectra were recorded on either Varian VI-300 or VXR-300 spectrometers at ~20 °C. Chemical shifts (δ) were referenced to residual solvent signal and integrated intensities compared to 1,3,5-trimethyloxybenzene added as an internal standard. UV-vis spectra were recorded on an HP8453 (190–1100 nm) diode-array spectrophotometer. Elemental analyses were performed by Robertson Microlit Laboratory (Ledgewood, NJ). Electrospray ionization mass spectra (ESI-MS) were recorded on a Bruker BioTOF II instrument. IR spectra were obtained using a ThermoNicolet Avatar 370 FT-IR.

[(R₃tacn)Cu(S=SbPh₃)]SbF₆ (R = Me (1), Et (2), or iPr (3))

All three complexes were prepared according to this illustrative procedure: In an inert atmosphere, to a solution of the S=SbPh₃ (32 mg, 0.083 mmol) in CH₂Cl₂ (4 mL) was added [(Me₃tacn)Cu(NCCH₃)]SbF₆ (43 mg, 0.083 mmol) in CH₂Cl₂ (1 mL). After stirring for 1 h, the mixture was filtered and the volume of the filtrate was reduced to ~1 mL, and Et₂O (15 mL) was added, resulting in formation of a yellow precipitate. The supernatant was decanted and the yellow powder was washed with Et₂O (3 × 6 mL). The product was isolated in crystalline form by layering Et₂O into a concentrated THF solution at −20 °C (65 mg, 92%). Analogous procedures were used to isolate 2 and 3 as yellow crystals in 41% and 32% yields, respectively. 1: ¹H NMR (300 MHz, CD₂Cl₂): δ = 7.77-7.61 (m, 15H), 2.59-2.53 Hz (m, 12H), 2.38 (s, 9H) ppm; ¹³C{¹H} NMR: (75 MHz, CD₂Cl₂): δ = 134.37, 133.44, 130.97, 55.03, 49.02 ppm. UV-vis [λ_max, nm (ε, M⁻¹ cm⁻¹) in CH₂Cl₂]: 356 (2100). Anal. Calcd for C₂₇H₃₆CuF₆N₃SSb₂: C, 37.90; H, 4.24; N, 4.91. Found: C, 37.77; H, 4.24; N, 4.99. ESI-MS: [Cu(Me₃tacn)(S=SbPh₃)]⁺ calc. m/z 620.0, found 620.3. FT-IR: 2859.2, 1480.5, 1460.1, 1437.5, 1363.4, 1300.0, 1152.8, 1130.2, 1089.4, 1065.5, 1017.0, 966.3, 984.1, 889.5, 773.5, 751.3, 735.4, 692.4, 656.4 cm⁻¹. 2: ¹H NMR (300 MHz, CD₂Cl₂): δ = 7.74-7.60 (m, 15H), 2.60-2.53 Hz (m, 18H), 1.10 (t, J = 6.0 Hz, 9H) ppm; ¹³C{¹H} NMR (75 MHz, CD₂Cl₂): δ = 134.6, 133.4, 130.9, 55.3, 53.3, 13.0 ppm. UV-vis [λmax, nm (ε, M⁻¹ cm⁻¹) in CH₂Cl₂]: 356 (2300). Anal. Calcd for C₃₀H₄₂CuF₆N₃SSb₂: C, 40.13; H, 4.72; N, 4.68. Found: C, 39.76; H, 4.70; N, 4.65. ESI-MS: [Cu(Et₃tacn)(S=SbPh₃)]⁺ calc. m/z 662.0, found 662.2. FT-IR: 2972.4, 1479.3, 1437.6, 1378.9, 1347.9, 1315.0, 1136.6, 1124.9, 1067.4, 1041.1, 1031.4, 995.7, 928.8, 897.9, 876.2, 860.3, 851.7, 827.1, 812.7, 796.8, 751.0, 739.1 cm⁻¹. 3: ¹H NMR (300 MHz, CD₂Cl₂): δ = 7.73-7.59 (m, 15H), 2.66-2.61 Hz (m, 9H), 2.47-2.44 Hz (m, 6H), 1.13 (d, J = 6.0 Hz, 18H) ppm; ¹³C{¹H} NMR (75 MHz, CD₂Cl₂): δ = 138.75, 134.69, 133.38, 130.91, 58.29, 50.83, 19.70 ppm. UV-vis [λ_max, nm (ε, M⁻¹ cm⁻¹) in CH₂Cl₂]: 356 (2300). Anal. Calcd for C₃₃H₄₈CuF₆N₃SSb₂: C, 42.17; H, 5.15; N, 4.47. Found: C, 42.11; H, 5.23; N, 4.46. ESI-MS: [Cu(iPr₃tacn)(S=SbPh₃)]⁺ calc. m/z 704.1, found 704.3. FT-IR: 2965.7, 1491.5, 1478.4, 1437.0, 1386.8, 1368.5, 1351.5, 1299.1, 1265.1, 1167.0, 1129.4, 1067.2, 1020.3, 995.7, 968.1, 856.7, 841.0, 750.3, 737.9, 721.3, 692.6, 656.5 cm⁻¹.

[(Et₃tacn)₂Cu₂(µ-S₂)](BPh₄)₂ (5)

Elemental sulfur (1.8 mg, 0.007 mmol) was added to a solution of [(Et₃tacn)Cu(NCCH₃)]BPh₄ (36 mg, 0.056 mmol) in CH₂Cl₂ (4 mL). After stirring for 2 h, solvent was removed under reduced pressure to yield a brown solid. The brown solid was washed with Et₂O (2 × 6 mL), extracted with DMF (2 mL), and then filtered. Slow diffusion of Et₂O into the dark brown filtrate at room temperature afforded the product as deep green crystals (16 mg, 46%). ¹H NMR (300 MHz, d₇-DMF): δ = 7.32 (s, 16H), 6.96 (t, 16H), 6.81 (t, 8H), 3.16-3.09 Hz (m, 36H), 1.36 (t, J = 6.0 Hz, 18H) ppm; ¹³C{¹H} NMR (75 MHz, d₇-DMF): δ = 137.11, 135.71, 126.33, 122.62, 93.84, 56.04, 54.64 and 12.39 ppm. UV-vis [λ_max, nm (ε, M⁻¹ cm⁻¹) in DMF]: 410 (7400), 378 (7600). Anal. Calcd for C₇₂H₉₄B₂Cu₂N₆S₂: C, 68.83; H, 7.54; N, 6.69. Found: C, 68.45; H, 7.42; N, 6.70. FT-IR: 3052.7, 2978.9, 1579.6, 1478.5, 1465.9, 1386.9, 1270.4, 1258.5, 1144.8, 1122.9, 1069.9, 1033.3, 1018.3, 999.5, 921.0, 862.4, 842.3, 821.8, 792.9, 771.5, 742.3, 732.0, 710.1, 700.6, 668.2 cm⁻¹.

[(iPr₃tacn)₂Cu₂(µ-S₂)](BPh₄)₂ (6)

A similar procedure to that used for the preparation of 5 was followed, except THF was used as the reaction solvent and the product was isolated as dark red crystals from CH₂Cl₂ at −20 °C (37% yield). ¹H NMR (300 MHz, CD₂Cl₂): δ = 7.33 (s, 15H), 7.04 (t, J = 6.0 Hz, 14H), 6.89 (t, J = 6.0 Hz, 11H), 3.15-3.06 Hz (m, 5H), 2.95-2.74 (m, 4H), 2.65-2.46 (m, 11H), 2.33-2.20 (m, 10H), 1.26-1.08 (d, J = 6.0 Hz, 36H) ppm; ¹³C{¹H} NMR (75 MHz, CD₂Cl₂): δ = 136.44, 126.22, 122.42, 60.47, 51.75, 45.10, 19.79 and 18.46 ppm. UV-vis [λ_max, nm (ε, M⁻¹ cm⁻¹) in CH₂Cl₂]: 476 (7200), 380 (11000). Anal. Calcd for C₇₈H₁₀₆B₂Cu₂N₆S₂: C, 69.88; H, 7.97; N, 6.27. Found: C, 69.91; H, 7.78; N, 6.24. FT-IR: 3053.5, 2978.0, 1579.8, 1480.7, 1467.0, 1451.0, 1426.7, 1390.1, 1369.2, 1347.1, 1291.9, 1268.7, 1166.3, 1141.9, 1129.4, 1067.3, 1046.6, 1033.5, 962.4, 943.1, 841.7, 761.0, 749.3, 734.1, 705.6, 680.6, 668.1 cm⁻¹.

[(R₃tacn)Cu(SbPh₃)]SbF₆ (7, R = Me; 8, R = Et; 9, R = iPr)

These complexes were prepared similarly, according to the following representative procedure for 7. In an inert atmosphere, to a solution of SbPh₃ (40.0 mg, 0.113 mmol) in CH₂Cl₂ (4 mL) was added [(Me₃tacn)Cu(NCCH₃)]SbF₆ (58.0 mg, 0.113 mmol) in CH₂Cl₂ (1 mL). The mixture was stirred 3 h, filtered, and the volume of the filtrate was reduced to ~1 mL under reduced pressure. A portion of Et₂O (15 mL) was added to yield a white precipitate. The supernatant solution was decanted and the white powder was washed three times with Et₂O (3 × mL). The white product was recrystallized by diffusion of Et₂O into a concentrated CH₂Cl₂ solution at room temperature to generate the product as colorless crystals (71 mg, 76 %). 7: ¹H NMR (300 MHz, CD₂Cl₂): δ = 7.48-7.43 (m, 15H), 2.90 Hz (s, 12H), 2.73 (s, 9H) ppm; ¹³C{¹H} NMR (75 MHz, CD₂Cl₂): δ = 135.87, 130.89, 130.27, 55.98, 50.69 ppm. Anal. Calcd for C₂₇H₃₆CuF₆N₃Sb₂: C, 39.37; H, 4.41; N, 5.10. Found: C, 38.87; H, 4.61; N, 4.86. ESI-MS: [Cu(Me₃tacn)(SbPh₃)]⁺ calc. m/z 588.0, found 588.1. FT-IR: 1463.1, 1434.1, 1363.8, 1299.2, 1153.8, 1127.8, 1085.8, 1069.8, 1057.1, 1015.0, 1000.3, 985.9, 768.3, 737.6, 697.7 cm⁻¹. 8 (68% yield): ¹H NMR (300 MHz, CD₂Cl₂): δ = 7.48-7.41 (m, 15H), 2.90-2.85 Hz (m, 18H), 1.15 (t, J = 6.0 Hz, 9H) ppm; ¹³C{¹H} NMR (75 MHz, CD₂Cl₂): δ = 135.85, 130.88, 130.26, 56.87, 54.68, 14.46 ppm. Anal. Calcd for C₃₀H₄₂CuF₆N₃Sb₂: C, 41.62; H, 4.89; N, 4.85. Found: C, 41.50; H, 5.20; N, 4.92. ESI-MS: [Cu(Et₃tacn)(S=SbPh₃)]⁺ calc. m/z 662.0, found 662.2. FT-IR: 1434.2, 1375.8, 1338.5, 1121.2, 1067.2, 1037.6, 999.7, 931.38, 768.2, 737.3, 699.67 cm⁻¹. 9 (70 % yield): ¹H NMR (300 MHz, CD₂Cl₂): δ = 7.49-7.41 (m, 15H), 3.15-3.11 (m, 3H), 2.93-2.89 Hz (m, 6H), 2.80-2.75 Hz (m, 6H), 1.13 (d, J = 6.0 Hz, 18H) ppm; ¹³C{¹H} NMR (75 MHz, CD₂Cl₂): δ = 135.9, 130.85, 130.25, 59.69, 51.77, 20.13 ppm. Anal. Calcd for C₃₃H₄₈CuF₆N₃Sb₂: C, 43.66; H, 5.33; N, 4.63. Found: C, 43.30; H, 5.72; N, 4.62. ESI-MS: [Cu(iPr₃tacn)(SbPh₃)]⁺ calc. m/z 672.1, found 672.2. FT-IR: 1434.4, 1389.4, 1367.0, 1347.1, 1161.7, 1123.2, 1066.5, 997.3, 964.31, 756.2, 739.2, 720.6, 699.6 cm⁻¹.

[(TMCHD)₃Cu₃(S)₂](PF₆)₃ (12)

In an inert atmosphere, to a solution of the [(TMCHD)Cu(NCCH₃)]PF₆ (31 mg, 0.073 mmol) in CH₂Cl₂ (3 mL) was added S=SbPh₃ (29 mg, 0.049 mmol) in CH₂Cl₂ (1 mL). After the mixture was stirred for 2 h, it was filtered and the solvent was removed from the filtrate under vacuum to yield a deep green solid, which was washed with Et₂O (3 × 6 mL). The deep green powder obtained was crystallized from CH₂Cl₂ at −20 °C to form dark amber crystals of the product (28 mg, 79%). The product was identified by its X-ray crystal structure and the similarity of its UV-vis spectrum to previously reported data.^8a

General Procedures for NMR Kinetics

In a glovebox, appropriate volumes of starting materials in CD₂Cl₂ were mixed in a vial and the volumes were quickly adjusted to 1 mL so that the concentrations of adducts 1–3 and [(R₃tacn)Cu(CH₃CN)]SbF₆ were 4.7 mM and 47 mM, respectively. The solution was then quickly transferred to a J. Young NMR tube that was removed from glovebox and placed in the spectrometer probe. The progress of the reaction was monitored by ¹H NMR spectroscopy at room temperature with 1,3,5-trimethoxybenzene as an internal standard. The initial rates were determined in experiments in which the first 5–10% of the reaction was followed; the rate constants were obtained by linear fitting of the initial rate change. In the experiments with 2,3-dimethylbutadiene, identical procedures were used except 20 equivalents of 2,3-dimethylbutadiene was added to the mixture. Data analysis and graphical representations were performed using the program Kaleidagraph.

Results and Discussion

Synthesis and Characterization of LCu(I)-S=SbPh₃ Adducts and [L₂Cu₂(S₂)]²⁺ Decay Products

Reaction of [(R₃tacn)Cu(NCCH₃)]SbF₆ (R = Me, Et, or iPr) with one equivalent of S=SbPh₃ in CH₂Cl₂ yielded adducts 1–3, respectively, as yellow crystalline solids (Scheme 1). The complexes are stable in the solid state when stored under nitrogen, but in solution they decompose slowly (see below). The formulations of 1–3 are based on NMR, UV-vis, and FT-IR spectroscopy, CHN analysis, ESI-MS, and X-ray crystallography. Notable identifying features include (a) ¹H NMR spectra with sharp peaks in the diamagnetic region and multiplets for the Ph₃Sb=S hydrogen atoms shifted ~0.2–0.3 ppm downfield from uncomplexed Ph₃Sb=S (Figure S1), (b) a shoulder in UV-vis spectra with λ_max at ~350 nm (ε = ~2100–2300 M⁻¹cm⁻¹), and (c) parent ions [(R₃tacn)Cu(SSbPh₃)]⁺ with appropriate isotope patterns in ESI mass spectra (Figure S2).

Scheme 1.

Reactions of Cu(I)-S=SbPh₃ adducts.

The X-ray crystal structures of complexes 1–3 are shown in Figures 1 (1) and S3 (2 and 3), with selected bond distances and angles listed in Table 1. To our knowledge, they represent the first such structures with Ph₃Sb=S coordinated to a metal center.¹⁹ They contain similar four-coordinate Cu(I) centers with highly distorted tetrahedral geometries characterized by τ₄ values: 0.640 (1), 0.657 (2), and 0.634 (3).²⁰ Essentially, the distortion involves perturbation of the Cu-S bonds from the idealized trigonal axis towards coplanarity with two of the N-donor atoms on the supporting ligand (N2 and N3 for 1, Figure 1), presumably as a result of steric interactions between the N-donor ligand substituents and the phenyl rings of the coordinated Ph₃Sb=S moiety. These steric effects also appear to influence the Cu⋯Sb distance, which increases as the size of the R group of the ligand increases from 3.411 Å (1), 3.490 Å (2), to 3.547 Å (3). The Cu–S–Sb moiety adopts a ‘bent’ conformation with Cu–S–Sb bond angles between 100.1–104.6°. The average Cu–N bond lengths range between 2.06–2.17 Å, comparable to those in other copper(I) complexes of R₃tacn ligands.^17,21 The Cu–S distances of complexes 1–3 (2.167–2.203 Å) are shorter than typical copper(I)-thioether sulfur interactions (~ 2.30 Å), ²² and longer than copper(I)-thiolate interactions (2.13 ~ 2.18 Å),²³ but comparable to known Cu(I)–S=PPh₃ complexes (2.21 ~ 2.27 Å).²⁴ Complexation of Ph₃Sb=S to the copper center induces little if any change in the S–Sb bonding, as the S–Sb distances in 1–3 (2.281–2.283 Å) are only slightly longer than that in free S=SbPh₃ (2.244(1) Å).²⁵

Figure 1.

Representation of the X-ray crystal structure of 1, showing the cationic portion with all non-hydrogen atoms as 50% thermal ellipsoids (H atoms omitted for clarity, only heteroatoms labeled).

Table 1.

Selected Bond Distances (Å) and Angles (deg) for the X-ray structures of 1–3.^a

	1	2	3
Cu1-N1	2.160(2)	2.006(6)	2.178(2)
Cu1-N2	2.107(2)	2.030(8)	2.161(2)
Cu1-N3	2.199(2)	2.157(4)	2.166(2)
Cu1-S1	2.1671(8)	2.1813(10)	2.2027(7)
S1-Sb1	2.2832(7)	2.2735(9)	2.2812(7)
Cu1⋯Sb1	3.411	3.490	3.547
N1-Cu1-N2	84.96(9)	91.3(3)	84.52(8)
N1-Cu1-N3	83.41(8)	71.1(2)	84.08(8)
N2-Cu1-N3	83.96(9)	82.2(3)	84.39(8)
N1-Cu1-S1	114.17(6)	115.8(2)	127.05(6)
N2-Cu1-S1	148.91(7)	150.2(2)	143.48(6)
N3-Cu1-S1	120.92(6)	117.09(10)	113.59(6)
Cu1-S1-Sb1	100.06(3)	103.14(4)	104.55(3)

^aEstimated standard deviations indicated in parentheses.

By monitoring solutions of 1 in CD₂Cl₂ by ¹H NMR spectroscopy with an internal standard (Figure S4) at 20 °C the slow decay (t_1/2 ~12h) of 1 to SbPh₃ and the disulfido-dicopper( II) species [(Me₃tacn)₂Cu₂(µ-η²:η²-S₂)](SbF₆)₂ (4) was observed in accordance with the stoichiometry shown in Scheme 1. A similar decay reaction of 2 occurred to yield the respective disulfido-dicopper(II) complex (5, with SbF₆⁻ counterion), but at a significantly slower rate (t_1/2 ~2d). Complex 3 decayed at a similar rate as 2, but the nature of the product(s) in this case was not clear. The disulfido-dicopper(II) complexes 4–6 formed much more rapidly and with higher yields/conversions upon addition of [(R₃tacn)Cu(CH₃CN)]SbF₆ to solutions of 1–3, but in these reactions the adducts [(R₃tacn)Cu(SbPh₃)]SbF₆ (7, R = Me; 8, R = Et; 9, R = iPr) formed instead of free SbPh₃ (Scheme 1).

The disulfido-dicopper(II) complexes 4–6 were identified by comparison of ¹H NMR spectra with data reported previously (4)¹⁴ or obtained from independently prepared samples of the variants 5 and 6. These latter complexes were isolated as BPh₄⁻ salts by treating [(R₃tacn)Cu(CH₃CN)]BPh₄ with S₈ and were fully characterized by CHN analysis and NMR, FTIR, and UV-vis spectroscopy. Complexes 4–6 share similarly sharp ¹H NMR features in the diamagnetic region, consistent with singlet ground states arising from disulfide-mediated antiferromagnetic coupling between the Cu(II) ions. They also share diagnostic S₂²⁻ → Cu(II) LMCT transitions in UV-vis spectra at λ_max 380–400 nm (ε ~8000–14,000). ^26,7b In addition, the X-ray structure of 5 was solved (Figure 2). The X-ray structures of 5 is similar to that previously reported for 4,¹⁴ as illustrated by analogous S-S (4: 2.165(4) Å; 5: 2.150(1) Å) and Cu-Cu (4: 3.84 Å; 5: 3.88 Å) distances and square pyramidal coordination geometries (τ₅ = 0.03 for 4 and 0.01 for 5).²⁸

Figure 2.

Representation of the X-ray structure of 5 (BPh₄⁻ salt), showing the dicationic portion with all non-hydrogen atoms as 50% thermal ellipsoids (H atoms omitted for clarity, only heteroatoms labeled). Selected bond distances (Å) and angles (deg) are as follows: Cu1-N1, 2.202(3); Cu1-N2, 2.017(2); Cu1-N3, 2.003(2); Cu1-S1, 2.2152(8); S1-S1´, 2.1501(14); Cu1⋯Cu1´, 3.876; N1-Cu1-N2, 86.17(10); N1-Cu1-N3, 86.13(10); N2-Cu1-N3, 88.50(10); N1-Cu1-S1, 108.49(7); N2-Cu1-S1, 160.58(7); N3-Cu1-S1, 104.76(8); S1-Cu1-S1´, 58.03(3).

The adducts [(R₃tacn)Cu(SbPh₃)]SbF₆ (7, R = Me; 8, R = Et; 9, R = iPr) formed in the reactions of 1–3 with [(R₃tacn)Cu(CH₃CN)]SbF₆ also were identified by comparison of ¹H NMR spectra with data obtained from independently prepared samples. These samples were synthesized in good yield (~70%) from the reaction of SbPh₃ with [(R₃tacn)Cu(CH₃CN)]SbF₆. They were isolated as colorless crystalline solids and were fully characterized by CHN analysis, NMR and FTIR spectroscopy, and ESI-MS. Notably, the mass spectrum for each complex exhibits a parent ion with the appropriate isotope pattern for [(R₃tacn)Cu(SbPh₃)]⁺ (illustrated for R = Me in Figure S5).

Mechanistic Studies

A series of experiments were performed in order to gain insight into the reactions of the adducts 1–3 with [(R₃tacn)Cu(CH₃CN)]SbF₆ to yield the disulfido-dicopper( II) complexes 4–6 and the SbPh₃ adducts 7–9 (Figure S6). First, the reactions of the complexes with identical supporting ligands under pseudo-first-order conditions (i.e., 1 + 10 equiv. [(Me₃tacn)Cu(CH₃CN)]SbF₆, 2 + 10 equiv. [(Et₃tacn)Cu(CH₃CN)]SbF₆, and 3 + 10 equiv. [(iPr₃tacn)Cu(CH₃CN)]SbF₆) in CD₂Cl₂ were monitored as a function of time by ¹H NMR spectroscopy. At initial concentrations [1–3]₀ = 4.7 mM at 20 °C, the consumption of 1–3 followed first-order kinetics (Figures S6 and S7). The rates measured for the reactions of 2 and 3 are similar, with both being >~10 times slower than that of 1, as reflected by the measured k_obs values (averages from 3 replicate runs) of 6.6(5) × 10⁻⁴ s⁻¹ (1), 8.4(1) × 10⁻⁵ s⁻¹ (2), and 6.0(4) × 10⁻⁵ s⁻¹ (3). The results are roughly consistent with the relative steric profiles of the reactant pairs (1 < 2 ~ < 3) and support a mechanism wherein steric interactions among the reactant pairs influence the rate (e.g., involving interaction of the Cu(I)-S=SbPh₃ adduct with the added Cu(I) reactant).

To further test this hypothesis, reactions of [(Me₃tacn)Cu(CH₃CN)]SbF₆ with 1 were examined by measuring initial reaction rates of disappearance of 1 as a function of [[(Me₃tacn)Cu(CH₃CN)]SbF₆]₀ (Figure 3). The initial rate saturates as the initial concentration of the added Cu(I) reagent increases. This finding is consistent with a mechanism (Scheme 2) involving an initial rapid pre-equilibrium (K_eq) involving formation of a dicopper intermediate (A) followed by a rate-determining product formation step (k₂). The fit of the data to the corresponding eq. 1 is shown in Figure 3 (solid line), yielding K_eq = 200 M⁻¹ and k₂ = 1.2 × 10⁻⁴ s⁻¹. The proposed structure for A is speculative, as it was not observed directly. While the product formation process from A (k₂) must involve multiple steps, including a step in which a second sulfur atom is added, the decay of 1 is first order in [1] (see above) which implies that a unimolecular reaction of A is rate-controlling (e.g., cleavage of the S-Sb bond).

$$\mathit{\text{rate}}=\frac{k_2[\mathbf{1}][\mathit{\text{Cu}}(I)]}{K_{\mathit{\text{eq}}}^{-1}+[\mathit{\text{Cu}}(I)]}$$ eq. (1)

Figure 3.

Plot of the initial rate of decay (¹H NMR) of 1 vs. [[(Me₃tacn)Cu(CH₃CN)]SbF₆]₀ for the reaction of 1 with [(Me₃tacn)Cu(CH₃CN)]SbF₆ to yield 4 and 7 in CD₂Cl₂ at 20 °C. Each data point is an average of data for 3 replicate runs. Error bars span the range of values for the replicate runs. The line is a fit of the data to eq. 1 (R = 0.95).

Scheme 2.

Proposed mechanism for the formation of 4 from the reaction of 1 with [(Me₃tacn)Cu(CH₃CN)]SbF₆.

We also considered the possibility that the reactions of 1–3 to form the disulfido-dicopper(II) complexes 4–6 might involve decomposition of S=SbPh₃ to S₂.^9a This decomposition was reported to occur in CS₂ via a second-order process with a rate constant of 0.014(2) M⁻¹s⁻¹ at 35 °C, with the release of S₂ suggested by the formation of cyclic sulfides 10 and 11 when 2,3-dimethylbutadiene was used as a trapping reagent (Scheme 3). We found that mixtures of S=SbPh₃ (4.7 mM) and 2,3-dimethylbutadiene (20 equiv.) in CD₂Cl₂ at 20 °C were unchanged after ~3 d, with no evidence for formation of 10 or 11. In a second experiment, ¹H NMR spectroscopic monitoring of a mixture of 1 and 2,3-dimethylbutadiene (20 equiv.) revealed slow generation of 10 and loss of 1 via a first-order process with a rate constant equal to 1.4(4) × 10⁻⁵ s⁻¹ (t_1/2 = ~14 h). There was no evidence for the presence of the disulfido-dicopper(II) complex 4 during this process. Interestingly, under identical conditions 4 also decayed in the presence of 2,3-dimethylbutadience to yield 10. This reaction of 4 is characterized by a first-order rate constant of 1.7(3) × 10⁻⁴ s⁻¹, corresponding to a rate approximately 10 times faster than that of the reaction of 1 to yield 10. The rate of decay of 1 to yield 10 in the presence of 2,3-dimethylbutadiene is similar to that observed for the decay of 1 to 4 in its absence, suggesting that S₂ formation cannot be ruled out in the pathways of both reactions. However, these reactions are significantly slower than that for the reaction of 1 with [(Me₃tacn)Cu(CH₃CN)]SbF₆ to give 4 and 7. On this basis and in view of the kinetic data described above, it appears unlikely that S₂ formation is important in the reaction of 1 with [(Me₃tacn)Cu(CH₃CN)]SbF₆.

Scheme 3.

Decay of S=SbPh₃ to S₂ as determined by trapping with 2,3-dimethylbutadiene.

This latter reaction is further complicated by exchange of the S=SbPh₃ unit between the Cu(I) centers. This exchange was identified by ¹H NMR spectroscopy and ESI-MS analysis of the reaction of 1 with [(R₃tacn)Cu(CH₃CN)]SbF₆ (R = Et or iPr). Shortly after mixing, the ¹H NMR spectrum showed peaks due to all four species in the equilibrium shown in Scheme 4. From the relative integrations after equilibrium was reached (~10 min), K’_eq values of 0.41 (R = Et) or 0.054 (R = iPr) were measured. These results were corroborated by ESI-MS (Figure S8), where parent ion peak envelopes for the cationic portions of 1 and 2 (R = Et, ~1:1 ratio) or 1 and 3 (R = iPr, ~6:1 ratio) were observed immediately after mixing of the respective reagents. The trend in K’_eq values correlates inversely with the degree of steric interactions between the ligand substituents and the bound S=SbPh₃ moiety (K’_eq decreases as the steric interactions increase, R = Et < iPr). Importantly, equilibration is rapid relative to the decay to the disulfido-dicopper(II) complexes, and thus occurs prior to the suggested pathway for the decay reaction shown in Scheme 2.

Scheme 4.

Equilibration of Cu(I)-S=SbPh₃ and Cu(I)-NCCH₃ complexes.

Reaction of S=SbPh₃ with [(TMCHD)Cu(NCCH₃)]PF₆

In contrast to the reactions with the Cu(I) complexes of R₃tacn ligands that led to isolable Cu(I)-S=SbPh₃ adducts, no such adducts were identified when the Cu(I) complex of the bidentate diamine TMCHD was treated with S=SbPh₃. Instead, the known tricopper cluster [(TMCHD)₃Cu₃(S)₂](PF₆)₃ (12) was isolated cleanly in good yield (79%). This procedure for the synthesis of 12 is superior to that previously reported involving use of S₈,^8a facilitating advanced spectroscopic studies of the cluster aimed at addressing contentious bonding and oxidation state issues.⁸ Presumably, an initial Cu(I)-S=SbPh₃ adduct forms in the reaction, but due to its lower coordination number is more prone to oligomerization than the analogs supported by the tridentate R₃tacn ligands.

Summary and Conclusions

In explorations ultimately aimed at preparing copper-sulfur complexes to model the active site of nitrous oxide reductase, we have found that Ph₃Sb=S forms stable adducts [(R₃tacn)Cu(S=SbPh₃)]SbF₆ (1–3), the first examples of which have been structurally characterized by X-ray crystallography. These adducts undergo slow decay in solution to form [(R₃tacn)₂Cu₂(µ-η²:η²-S₂)]²⁺ species (4–6) and SbPh₃. Conversion to 4–6 is accelerated by addition of [(R₃tacn)Cu(NCCH₃)]SbF₆ to 1–3, and yield [(R₃tacn)Cu(SbPh₃)]SbF₆ (7–9) as coproduct instead of free SbPh₃. Mechanistic studies of this reaction revealed rapid exchange of Ph₃Sb=S between the Cu(I) sites and pre-equilibrium formation of a dicopper intermediate. We speculate that the dicopper intermediate contains a bridging Ph₃Sb=S moiety and that the rate-controlling step in the reaction involves loss of Ph₃Sb from that intermediate. Subsequent more rapid events that ultimately result in [Cu₂(µ-η²:η²-S₂)]²⁺ core formation remain unclear. Reaction of [(TMCHD)Cu(CH₃CN)]PF₆ with S=SbPh₃ did not lead to an observable adduct, and instead led to the known tricopper cluster [(TMCHD)₃Cu₃(µ₃-S)₂](PF₆)₃ in good yield. Overall, the results demonstrate the utility of Ph₃Sb=S for delivering sulfur to Cu(I) centers supported by N-donor ligands, cleanly yielding thermodynamically stable [Cu₂(µ-η²:η²-S₂)]²⁺ and [Cu₃S₂]³⁺ cores.