Dinuclear copper ( I ) complexes with N-heterocyclic thione and selone ligands : synthesis , characterization , and electrochemical studies †

The synthesis, characterization, and structures of a series of homoleptic and heteroleptic copper(I) complexes supported by N-heterocyclic chalcogenone ligands is reported herein. The quasi-reversible Cu(II/I) reduction potentials of these copper complexes with monodentate (dmit or dmise) and/or bidentate (Bmm, Bsem, Bme, Bsee) chalcogenone ligands are highly dependent upon the nature and number of the donor groups and can be tuned over a 470 mV range (−369 to 102 mV). Copper–selone complexes have more negative Cu(II/I) reduction potentials relative to their thione analogs by an average of 137 mV, and increasing the number of methylene units linking the heterocyclic rings in the bidentate ligands results in more negative reduction potentials for their copper complexes. This ability to tune the copper reduction potentials over a wide range has potential applications in synthetic and industrial catalysis as well as the understanding of important biological processes such as electron transfer in blue copper proteins and respiration.


Introduction
The chemistry of monodentate and bidentate sulfur and selenium Lewis donor ligands towards soft and borderline metals has recently received much attention due to their potential applications in catalysis, 1,2 the preparation of radiopharmaceuticals, 3 and in supramolecular, bioinorganic, organometallic, and coordination chemistry. 4,5 Thus, great strides have been made in understanding the coordination chemistry of bis-(mercaptoimidazolyl)borate (Bm R ) and bis(mercaptoimidazolyl)-methane (Bmm R ) ligands, first pioneered by Parkin 6,7 and Williams, 8 respectively. In contrast, the reactivity of the corresponding selenium analogs, the bis(selenoimidazolyl)borates (Bse R ), 9,10 bis(selenoimidazolyl)methanes (Bsem R ), 1,11 and related derivatives, 5 remains markedly underdeveloped.
We are interested in the coordination chemistry of the aforementioned bidentate neutral ligands as well as that of the closely related bis(mercaptoimidazolyl)ethanes (Bme R ) and bis(selenoimidazolyl)ethanes (Bsee R ) with copper(I) to understand the fundamentals of the copper-sulfur and copper-selenium interactions and their effect on Cu(I)/Cu(II) redox potentials. The high propensity for sulfur-and selenium-containing ligands to bridge metal centers also results in diverse coordination frameworks 12 and these groups are also potential synthons for the formation of heterocyclic carbenes via potassium metal reduction. 13 There is also increased interest in copper chalcogenolates as single-source precursors in the synthesis of semiconductor materials via metal organic chemical vapor deposition. 14 Although coordination complexes of the Bmm Me ligand with rhenium(I), 3 iron(II), 15 cobalt(II), 11 rhodium(I), 1,16 iridium(I), 17 nickel(II), 11 silver(I), 18,19 gold(I/III), 19 zinc(II), 20 tin(II), 21 lead(II) 22,23 and antimony(III) 8 have been isolated, it is rather surprising that only one report of copper(I) derivatives has been published, 19 particularly given the reported affinity of copper for sulfur-and selenium-containing ligands. 24 In this work, we report the synthesis and crystal structures of a series of dinuclear, three-and four-coordinate copper(I) complexes with the aim of understanding the effect of the methylene linkers and chalcogenone donor groups on the redox potentials of the Cu(I)/Cu(II) couple. These reduction potentials are highly dependent upon S/Se ligand coordination and can be tuned in a wide potential range using a variety of monodentate and bidentate thione and selone ligands. Such redox tuning has practical applications ranging from understanding biological processes such as electron transfer in blue copper proteins and respiration, 25 to industrial and synthetic applications in catalysis. 2,26 Homoleptic and heteroleptic copper(I) complexes bearing monodentate (dmit or dmise) or bidentate (Bmm Me , Bsem Me , Bme Me , Bsee Me ) chalcogenone ligands ( Fig. 1) have been synthesized and characterized using elemental analysis, infrared (IR) and multinuclear ( 1 H, 13 C, 19 F, 77 Se) NMR spectroscopies, single-crystal X-ray diffraction, electrospray ionization mass spectrometry, and cyclic voltammetry.
In turn, heteroleptic dinuclear complexes of copper(I) were synthesized via a convenient two-step, one-pot synthesis by treating equimolar amounts of [Cu(NCMe) 4 ]BF 4 and dmit or dmise in acetonitrile, followed by cannula addition of Bmm Me or Bsem Me in dichloromethane (eqn (3)). Similarly, treating equimolar amounts of [Cu(NCMe) 4 ]BF 4 and dmit in acetonitrile followed by addition of one molar equivalent of Bmm Me in dichloromethane afforded a polynuclear copper(I) complex (eqn (4)).

Structural analyses of dinuclear copper complexes
The molecular structures of several complexes have been obtained using X-ray crystallography. More specifically, single crystals suitable for X-ray diffraction studies were obtained for [(dmise) 2 The X-ray crystal structure of [(dmise) 2 Cu(μ-dmise)Cu-(dmise) 2 ](BF 4 ) 2 ·CH 3 CN (2), is shown in Fig. 2, and selected bond lengths (Å) and angles (°) are given in Table 1. The structural unit of [(dmise) 2 Cu(μ-dmise)Cu(dmise) 2 ](BF 4 ) 2 is made up of two copper(I) centers, with the Se atom of the dimethylimidazole selone (dmise) ligands bridging the two copper atoms, forming a bent CuSeCu core. Each copper atom is further bonded to two dmise ligands and thus each copper adopts a distorted trigonal planar geometry. The average of the Fig. 2 The crystal structure diagram of the cation in [(dmise) 2 Cu-(μ-dmise)Cu(dmise) 2 ](BF 4 ) 2 ·CH 3 CN (2) showing 50% probability ellipsoids. Hydrogen atoms, counterions, and the solvent molecules are omitted for clarity. four Cu-Se distances involving terminal dmise ligands (2.35 Å) is shorter than those involving the bridging dmise ligand (2.42 Å) but is slightly longer than those in the monomeric copper selone complexes (∼2.30 Å) reported by Kimani et al. 27 In a similar vein, these values are comparable to those observed in the three-coordinate copper selone complexes Cu(dmise) 2 X, (X = Cl, Br, I) 28 and the diphosphine selenide derivative [Cu 3 I 3 {Ph 2 P(Se)-(CH 2 ) 3 -P(Se)Ph 2 } 2 ] n . 29 The molecular structures of the isostructural complexes  Tables 2 and 3, respectively. The dinuclear complexes feature two terminal and one bridging bis(chalcogenone) ligands, forming "butterfly" shape [Cu 2 E 2 ] cores (E = S, Se). Each copper(I) ion adopts a distorted tetrahedral geometry, with angles ranging from 96.45 to 123.86°for 3 and from 100.50 to 123.36°for 4. The Cu⋯Cu distances (2.96 and 2.97 Å for 3 and 4, respectively), significantly longer than twice the covalent radius of copper(I) (2.34 Å), precludes the existence of a copper-copper bonding interaction in these complexes. As expected, the terminal Cu-S and Cu-Se bond distances in 3 and 4 (averages 2.29 and 2.42 Å, respectively) and shorter than those involving the corresponding values involving bridging ligands (averages 2.44 and 2.52 Å, respectively).
The centrosymmetric copper complex [(Bme Me )Cu-(μ-Bme Me      centers, each arranged in a distorted trigonal planar geometry arising from the coordination of a terminal bidentate Bme Me ligand and one of the thione moieties from a bridging bis-(monodentate) Bme Me ligand. As summarized in Table 4, the sum of angles around each copper center is 354.91°and the average C-S bond distance is 2.29 Å.
The molecular structures of [(dmit)Cu(μ-Bsem Me ) 2 Cu(dmit)]-(BF 4 ) 2 (7) and [(dmise)Cu(μ-Bsem Me ) 2 Cu(dmise)](BF 4 ) 2 (9) are shown in Fig. 6, with selected bond length and angles for the isostructural complexes given in Table 5. The two dinuclear complexes are centrosymmetric and exhibit rhombic Cu 2 Se 2 cores, with all the bis(selone) ligands exhibiting the unusual bridging monodentate:bidentate (μ-κ 1 :κ 2 ) coordination mode. Each copper center is coordinated to a terminal dmit or dmise ligand and three selone moieties from Bsem Me ligands (one terminal and two bridging), with an overall distorted tetrahedral geometry in each case. The angles surrounding the copper centers in the two complexes are very similar, ranging from 95.38 to 118.61°for 7 and from 94.97 to 118.58°for 9. The Cu⋯Cu distances (2.73 and 2.74 Å for 7 and 9, respectively) are slightly shorter than the sum of the van der Waals radii of copper, suggesting the presence of weak Cu-Cu interactions.

NMR spectroscopy of dinuclear copper thione and selone complexes
The dinuclear copper complexes were characterized by 1 H, 13 27 The 13 C{ 1 H} NMR resonances for the complexed and uncomplexed thione and selone and thione ligands are given in Table 7. Substantial shifting of the CvS/CvSe resonances of the dmit, dmise, Bmm Me , Bsem Me carbon atoms are observed upon copper complexation relative to the free ligands. Coordination of the thiones and selones via the sulfur and selenium atoms results in upfield shifts of δ 5-8 ppm for both the CvS and CvSe carbons, in agreement with previous reports. 34 (2) 147. Electrochemical studies of the dinuclear copper complexes Cyclic voltammetry studies of the chalcogenones and their dinuclear copper complexes were conducted to determine the influence of the methylene linkers on the redox potential of the chalcogenone ligands and the change in Cu(II/I) reduction potential upon coordination of the chalcogenone ligands to copper. All the uncoordinated chalcogenone ligands exhibit chemically reversible and quasi-reversible electrochemical behavior, with the selone ligands having more negative reduction potentials relative to the analogous thione ligands (Fig. 8 and Table 8). The unbound bidentate ethylene-bridged ligands (Bme Me and Bsee Me ) have larger peak separations between the oxidized and reduced products relative to the methylenebridged ligands (Bmm Me and Bsem Me ), suggesting faster electron transfer in the latter. 43 The reduction potentials of the unbound selone ligands are: dmise −367 mV < Bsee Me Table 8). The reduction potentials of the free bidentate chalcogenones indicate that increasing the length of the linker from methylene to ethylene results in more negative reduction potentials.
The Cu(II/I)and Cu(I/0) redox potentials of the complexes versus NHE are given in Table 8. The cyclic voltamograms (CV) of the copper complexes 1, 2, 3, 4, 5, 6, 9, and 10 exhibit two, one-electron redox potential waves belonging to the Cu(II/I)and Cu(I/0) couples, with the exception of complexes 7 and 8 which exhibit three, one-electron redox potential waves. The Cu(I/0) redox couple commences at potentials more than −1000 mV vs. NHE and after switching the scan direction at potentials close to 750 mV, Cu(0) is stripped off the electrode (Fig. 9). All the dinuclear copper thione and selone complexes exhibit oneelectron Cu(II/I) oxidation and reduction waves with large ΔE values, indicating that these redox processes are not fully reversible (ESI, Fig. S2 †).
Upon examination of the reduction potentials for the copper complexes 1, 2, 3, 4, 5, and 6, it is clear that the selonecontaining complexes exhibit more negative Cu(II/I) reduction potentials relative to the analogous thione complexes regardless of whether the thione and selone ligands are bridging. A similar trend was reported by Kimani et al. for the electrochemistry of only monodentate [Tpm R Cu(X)] + complexes (X = dmise or dmit). 27 Interestingly, increasing the length of the linker in the bidentate ligands from methylene to ethylene results in lower Cu(II/I) reduction potentials for [Cu 2
The heterogeneous dinuclear complex [(dmit) 2 Fig. S2H †) exhibits three oxidation and reduction waves. One reduction and oxidation wave in the dinuclear copper complex 8 likely corresponds to the reduction potential of the bidentate Bmm Me ligand (E 1/2 = −51 mV), whereas the remaining two waves correspond to Cu(II/I) reduction potentials, similar to those observed for complex 7. These two different Cu(II/I) reduction potentials are only observed for the dinuclear copper complexes with mixed thione and selone ligands, and effect which has not been previously reported for copper complexes ( Table 8).
The unbound dmit and dmise ligands have more negative reduction potentials than the bidentate chalcogenones (Bmm Me , Bsem Me , Bme Me and Bsee Me ). The reduction potentials from the bidentate chalcogenones indicate that increasing the length of the linker from methylene to ethylene results in more negative reduction potentials. All the synthesized copper-selone complexes have more negative Cu(II/I) reduction potentials relative to the analogous copper-thione complexes. The copperselone complexes stabilize the Cu(II) oxidation state more effectively than the copper-thione complexes by an average of 144 mV, consistent with previously observed results. 27,28  Notably, the Cu(II/I) reduction potential of the dinuclear copper chalcogenone complexes 1 to 10 can be tuned in a 470 mV window from 102 mV to −369 mV by simply changing the nature of the chalcogen donor and the denticity of thione and selone ligands. This ability to tune the copper redox potentials could have potential applications in copper-based catalysis. Compared to naturally occurring cupredoxins with a Cu(II/I) reduction potential range of 90 to 670 mV, 44 the synthesized copper chalcogenone complexes have significantly more negative Cu(II/I) reduction potentials.

Conclusions
Dinuclear homoleptic and heteroleptic copper(I) complexes with monodentate and bidentate chalcogenone ligands have been synthesized and characterized, and the electrochemistry of the resulting species has been investigated and compared. Treating the copper(I) starting material [Cu(NCMe) 4 ]BF 4 with bidentate (Bmm Me , Bsem Me , Bme Me , Bsee Me ) and monodentate chalcogenone ligands (dmit and dmise) results in the formation of dinuclear copper complexes (1, 2, 3, 4, 5, and 6). The dinuclear copper complexes adopt either trigonal or tetrahedral geometries with both terminal and bridging thione or selone ligands. The heteroleptic dinuclear copper complexes [(dmit)Cu(μ-Bsem Me ) 2 Cu(dmit)](BF 4 ) 2 (7) and [(dmise)Cu(μ-Bsem Me ) 2 Cu(dmise)](BF 4 ) 2 (9) adopt distorted tetrahedral geometry where each copper is coordinated to three selenium atoms from Bsem Me ligands and one sulfur atom from dmit for 7 and one selenium atom from dmise for 9. Interestingly, the mixed ligand complex 10 consists of infinite chains of tetrahedrally coordinated Cu(I) ions bound to two sulfur atoms from a Bmm Me ligand and a bridging sulfur atom from a dmit ligand.
The copper selone complexes 2, 4, 6, and 9 have more negative Cu(II/I) reduction potentials relative to their sulfur analogs (1, 3, 5, and 10), and increasing the length of the methylene linker in the bidentate chalcogenone ligands results in more negative reduction potentials for their copper complexes. This study provides detailed comparative coordination chemistry of thiones and selones with copper and its effect on the Cu(II/I) reduction potentials. Simply changing the chalcogens and denticity of the thione and selone ligands results in Cu(II/I) reduction potentials of the synthesized copper chalcogenone complexes that can be tuned in a range of 471 mV, a difference Table 8 Redox potentials (mV) of chalcogenone ligands and Cu(II/I) and Cu(I/0) couples for dinuclear copper complexes vs. NHE all data were collected with 1 mM compound in acetonitrile with n-butylammonium phosphate as the supporting electrolyte (0.1 M) at a scan rate of 100 mV s −1  that would have significant effects in redox-mediated reactions.
Electrochemical experiments were performed with a BAS 100B potentiostat. A three-compartment cell was used with an Ag/AgCl reference electrode, Pt counter electrode, and a glassy carbon working electrode. Freshly-distilled acetonitrile was used as the solvent with tetra-n-butylammonium phosphate as the supporting electrolyte (0.1 M). Solutions containing 1 mM analyte were deaerated for 2 min by vigorous nitrogen purge. The measured potentials were corrected for junction potentials relative to ferrocenium/ferrocene (0.586 mV vs. Ag/AgCl 50 ) and adjusted from Ag/AgCl to NHE (−0.197 V (ref. 51)) All E 1/2 values were calculated from (E pa + E pc )/2 at a scan rate of 100 mV s −1 , and ΔE = E pa − E pc .
Infrared spectra were obtained using Nujol mulls on KBr salt plates with a Magna 550 IR spectrometer. Abbreviations used in the description of vibrational data are as follows: vs, very strong; s, strong; m, medium; w, weak; b, broad. Electrospray ionization mass spectrometry (ESI-MS) was conducted using a QSTAR XL Hybrid MS/MS System from Applied Biosystems via direct injection of sample (0.05 mL min −1 flow rate) into a Turbo Ionspray ionization source. Samples were run under positive mode, with ionspray voltage of 5500 V, and TOF scan mode. MALDI-TOF-MS was conducted on a Bruker Microflex. trans-2-[3-(4-tert-Butylphenyl)-2-methyl-2-propenylidene]malononitrile was used as a matrix for co-crystallization of the copper complex characterized. All the peak envelopes matched their calculated isotopic distributions. Melting points were determined using a Barnstead Electrothermal 9100 apparatus in silicon-grease-sealed glass capillary tubes. Absorption spectra were collected using a Varian Cary-50 Bio spectrophotometer in quartz cuvettes with a path length of 1 cm. Elemental analysis (EA) was performed using PerkinElmer Series II CHNS/O Analyzer 2400.

X-ray data collection and structural determination
Single crystals grown from vapor diffusion were mounted on a glass filament with silicon grease and immediately cooled to 168 K in a cold nitrogen gas stream. Single crystals suitable for X-ray analysis were obtained by slow diffusion of diethyl ether into an acetonitrile solution of [(dmise) 2 (10). Intensity data were collected using a Rigaku Mercury CCD detector and an AFC8S diffractometer. The space group P2 1 /c for 9 was determined from the observed systematic absences. No symmetry higher than triclinic was observed for 2, 4, 5, 7, and 10 and assignment of the centrosymmetric space group option, P1, provided chemically reasonable refinement results. Data reduction including the application of Lorentz and polarization (Lp) effects and absorption corrections used the CrystalClear program. 52 The structures were solved by direct methods and subsequent Fourier difference techniques, and refined anisotropically, by full-matrix least squares, on F 2 using SHELXTL 6.10. 53 In the final cycle of least squares, independent anisotropic displacement factors were refined for the non-hydrogen atoms and the methyl hydrogen atoms were fixed in "idealized" positions with C-H = 0.96 Å. Their isotropic displacement parameters were set equal to 1.5 times U eq of the attached carbon atom. For complex 2, the largest peak in the final Fourier difference map (1.08 e A −3 ) was located 0.83 Å from Se(4) and the lowest peak (−0.81 e A −3 ) was located at a distance of 0.86 Å from Se(4). The largest peak for complex 4 in the final Fourier difference map (0.82 e A −3 ) was located 0.08 Å from Se(4) and the lowest peak (−0.79 e A −3 ) was located at a distance of 0.77 Å from Se (5). The largest peak for 7 in the final Fourier difference map (1.10 e A −3 ) was located 1.23 Å from N(5) and the lowest peak (−0.78 e A −3 ) was located at a distance of 0.88 Å from Se(1). The largest peak for 9 in the final Fourier difference map (1.16 e A −3 ) was located 1.19 Å from H(6C) and the lowest peak (−0.74 e A −3 ) was located at a distance of 0.92 Å from Se(1). The largest peak for 10 in the final Fourier difference map (0.42 e A −3 ) was located 1.73 Å from S(1), and the lowest peak (−0.42 e A −3 ) was located at a distance of 0.76 Å from Cu(1).
For complex 3, a suitable crystal was mounted using viscous oil onto a plastic mesh, and cooled to the data collection temperature. Data were collected on a Bruker-AXS APEX CCD diffractometer with graphite-monochromated Mo-Kα radiation (λ = 0.71073 Å). The systematic absences in the diffraction data were consistent with Pna2 1 and Pnma. The absence of a molecular mirror or inversion point, and the observed occupancy, Z = 4, were consistent with Pna2 1 , the noncentrosymmetric option. The Flack parameter refined to zero, indicating that the true hand of the data was determined. This data set was treated with absorption corrections based on redundant multiscan data. The structures were solved using direct methods and refined with full-matrix, least-squares procedures on F 2 . All non-hydrogen atoms were refined with anisotropic displacement parameters. All hydrogen atoms were treated as idealized contributions. Scattering factors are contained in the SHELXTL 6.12 program library. 54 Final refinement parameters for the structures of 2, 3, 4, 5, 7, 9, and 10 are provided in Tables 9 and 10.