416 J.C.S. DaltonCrystal and Molecular Structure of the Polymeric Complex Chloro(2,5-d ith ia hexane)copper( I)By Edward N. Baker * and Peter M. Garrick, Department of Chemistry, Biochemistry, an4 Biophysics, MasseyUniversity, Palmerston North, New ZealandThe crystal and molecular structure of the title complex has been determined by single-crystal X-ray diffraction tech-niques, from diffractometer data. The crystals are monoclinic, with a = 7.673(2), b = 6.727(2), c = 15.700(5) A,p = 104.8". Z = 4, and space group P2Jc. The structure has been refined by full-matrix least-squares methodsto a final R factor of 0.035 for 1 71 0 independent reflections. The complex is polymeric, with adjacent molecules inthe crystal joined by a bond between the copper atom of one molecule and one of the sulphur atoms of the dithia-hexane ligand of a neighbouring molecule.Each copper atom has a distorted-tetrahedral geometry, with two bondsto the sulphur atoms of the dithiahexane ligand [Cu-S 2.336(1) and 2.342(1) A], one to a terminal chlorine atom[Cu-CI 2.239(1) A], and one to one sulphur atom of a neighbouring molecule [Cu-S 2.315(1) A].RECENT interest in copper complexes of thioetherligands has arisen for several reasons. Preparative andspectroscopic studies have shown that stable complexesof both Cul and CuIZ can be prepared, allowing thepossibility of a detailed comparison of the bonding ofcopper(') and copper(I1) atoms in similar environments.Furthermore, the copper( I) complexes display a varietyof stoicheiometries, with some simple four-co-ordinatetetrahedral species and others which are apparentlythree-co-ordinate or polynuclear.A third point ofinterest lies in the possible relevance of such complexesas bonding models for biological molecules. The thio-ether ligands resemble the side chain of the amino-acidmethionine, a ligand for iron in cytochrome C, and apotential metal ligand in other proteins. In the ' blue 'copper redox proteins, for example, some of the copperatoms, either singly (Type I) or in pairs (Type 111)apparently co-ordinate to one or more sulphur l i g a n d ~ . ~ - ~We have undertaken X-ray structure analyses ofseveral of these copper-thioether complexes in order toexamine the influence of the metal oxidation state andthe co-ordination environment on the Cu-S bonding6The structure analysis of chloro(2,5-dithiahexane)-copper('), [{Cu(dth)Cl},] was undertaken because spectro-scopic studies1 indicated a terminal Cu-Cl bond, andsuggested a three-co-ordinate trigonal-planar complexdifferent in structure from the other complexes investi-gated.The crystal structure reported here shows that1 E. W. Ainscough, A. M. Brodie, and K. C. Palmer, J.C.S.Dalton, 1976, 2375.D. R. McMillin, R. A. Holwerda, and H. B. Gray, Proc. Nut.Acud. Sci. U.S.A., 1974, 71, 1339.J . A. Fee, Structure and Bonding, 1975, 23, 1.the complex is, in fact, four-co-ordinate, distortedtetrahedral, with a terminal Cu-C1 bond, as predicted,but with rather unusual bridging through one sulphuratom of the dth ligand to give a polymeric structure.EXPERIMENTALThe complex [{ Cu(dth)Cl},], prepared as describedpreviously,l was crystallised from methanol as clear needle-shaped crystals.The space group and approximate celldimensions were obtained from oscillation and Weissenbergphotographs. More accurate cell dimensions were obtainedfrom a least-squares analysis of the positions of 12 generalreflections on a four-circle X-ray diffractometer.Crystal Data.-C,H,,ClCnS,, M = 221, Monoclinic, a =784 A3, F(000) = 448, D, = 1.89(1) g cmp3 (by flotation),2 = 4, D, = 1.88 g c ~ n - ~ , Mo-K, radiation, p(Mo-K,) =37.0 cm-], space group P2Jc from systematic absences.Data Collection and Heduction.-lntensity measurementswere made with a computer-controlled Hilger and Wattsfour-circle X-ray diffractometer, using zirconium-filteredMo-K, X-radiation, on a crystal of dimensions ca.0.04 x0.02 x 0.02 cm niounted along the longest dimension.The orientation of the crystal was defined by a least-squares treatment of the positions of 12 general reflections.Intensities were measured by means of a 8-28 scan,consisting of 35 steps of 0.02", through each reflection, witha count lasting 1.4 s a t each step and a background countof 7 s at the beginning and end of each scan. Threestandard reflections, chosen from diff erent regions ofE. I. Solomon, P. J . Glendening, H. 33. Gray, and F. J.Grunthaner, J . Amer. Chem. Soc., 1975, 97, 3878.W. Byers, C;. Curzon, I<. Garbett, B. E. Speyer, S.N. Young,and K. J. P. Williams, Biochim. Biophys. Acta, 1973, 310, 38.E. N. Baker and G. E. Norris, J.C.S. Dalton, 1977, 877.7.672(2), b = 6.727(2), G = 15.700(5) A, p = 104.8", U 1978 417reciprocal space, were monitored every 100 general reflec-tions as a check against possible crystal deterioration ormisalignment. No significant intensity reduction wasnoted during data collection. Reflections hkE, hk2, hki, andhkl were measured, up to 8 27". Equivalent reflectionswere merged, t o give a total of 1 5 2 4 observed [for whichIhkl > ~ ( l h k l ) ] and 186 unobserved reflections. The usualLorentz and polarisation corrections were applied, but nocorrection was made for absorption.Structure Determination.--The position of the copperatom was determined from a three-dimensional Pattersonsynthesis, and a subsequent electron-density map phasedby the copper atom led to the identification of the positionsof the remaining non-hydrogen atoms.The structure wasseen, unexpectedly, to be polymeric, and the connectivityof the atoms established that the bridging atoms weresulphur rather than chlorine. The structure was refinedby full-matrix least-squares methods, using a local versionof the program ORFLS7 adapted for the BurroughsB6700 computer by the University of Canterbury. Thequantity minimised was C W ( A F ) ~ , where AF = klF,I -lFcl and the weighting ze, = 4 F , 2 / [ ~ ( F , ) 2 ] 2 . Four cycles ofleast-squares refinement with isotropic temperature factorsfor all the atoms reduced the conventional R factor to0.082. Two further cycles, with anisotropic temperaturefactors, reduced R to 0.046, and a difference electron-density map then revealed the positions of all the hydrogenatoms.For the two final refinement cycles, hydrogenatoms were included in the structure-factor calculation(with temperature factors cu. 1.2 times those of their carbonatoms) but were not refined. This reduced the I? factorfrom 0.038 to convergence at I2 0.035 (for all the 1 7 1 0independent reflections, observed and unobserved) ; R'was 0.039. A subsequent difference electron-density mapcontained no feature higher than 0.5 e A-3 or lower than-0.5 e k3.Final atomic co-ordinates are listed in Table 1 and bondlengths and angles in Table 2.The numbering system forTABLE 1Final atomic co-ordinates, with standard deviations inparenthesesc u 0.131 5(1) 0.116 3(1) 0.240 5( 1)c1 0.222 4(1) 0.064 6( 1) 0.117 6(1)S(1) 0.141 l(1) 0.453 3(1) 0.277 3(1)S(2) 0.313 2(1) 0.033 2(1) 0.379 8(1)0.310 4(5) 0.562 6(5) 0.229 5(2)0.255 6(5) 0.443 4(5) 0.393 5(2)0.394 7(4) 0.279 9(5) 0.417 l(2)0.161 4(5) -0.010 6(6) 0.448 2(2)0.435 0.530 0.2780.270 0.530 0.1640.300 0.710 0.2200.150 0.435 0.4230.300 0.580 0.4140.438 0.295 0.4810.505 0.310 0.3900.058 0.070 0.4420.207 - 0.035 0.5080.100 -0.130 0.430Atom xla Ylb ZlGC(1)C(2)C(3)(74)H ( WH(1b)H(lc)H(2a)H (2b)H(34H(3b)H (44H(4.b)H(4c)non-hydrogen atoms is shown in the Figure, hydrogenatoms being numbered according to the carbon atoms towhich they are attached.Observed and calculated struc-* For details see Notices t o Authors No. 7, J.C.S. Dalton, 1977,Index issue.TABLE 2Bond lengths (A) and angles (") with standard deviationsin parentheses(a) Lengths *cu-Cl 2.239( 1) S(1)-C(1) 1.814(3)cu-s (1) 2.336( 1) S(l)-C(2) 1.813(3)cu-s ( 11) 2.315( 1) S( 2)-C( 3) 1.8 17 (3)cu-s (2) 2.342( 2) S ( 2)-C( 4) 1.799 (3)1.51 1 (5) C( 2)-C( 3)C-H bonds range from 0.93 t o 1.08 A (mean 1.00 A)( b ) AnglesS ( 1)-Cu-S(2)S ( l)-Cu-ClS( 1)-cu-s (11) s ( 2)-Cu-C1 s (2)-cu-s ( 11) s (lI)-Cu-Clcu-S( 1)-C( 1)cu-s ( 1)-C( 2)cu-s ( Ij-Cu'l91.54( 3)111.53(3)11 7.37 (2)121.70(3)107.69(3)107.04( 3)105.8( 1)117.36(3)101.0( 1)CuII-S( 1)-C(1)CuI'-S( 1)-C( 2)C( 1 )-s ( 1 )-C( 2)cu-s ( 2)-c (3)cu-s (2)-c (4)C( 3)-s (2)-C( 4)S(l)-C(2)-C(3) s (2)-C( 3)-c (2)118.6(1)109.3(1)102.5( 2)106.0( 1)100.3 (2)11342)115.0(2)99.1(1)Bond angles a t the carbon atoms, involving hydrogen atoms,* Roman numeral superscripts refer t o the following equivalentall have approximately tetrahedral valuespositions with respect to the reference molecule a t (x, y. z) :I -x, -; + y , & - 2; I1 -x, 3 + y , - 2:'PcuA \ 'i \ \ hDiagram showing the polymeric structure of the complex.TheCu-S bond linking neighbouring Cu(dth)Cl entities is shown asa broken lineture factors and anisotropic thermal parameters are listedin Supplementary Publication No. SUP 22195 (6 pp.).*DISCUSSION' Molecules ' of Cu(dth)Cl (dth = 2,5-dithiahexane)form infinite chains in the crystal, as shown in theFigure.Each copper atom makes four bonds: two toW. R. Busing, K. P. Martin, and H. A. Levy, ProgramORFLS, Oak Ridge National Laboratory, Oak Ridge, Tennessee418 J.C.S. Daltonthe sulphur atoms of the chelating dth ligand, one tothe terminal chlorine atom, and one to one of the sulphuratoms of the dth ligand of a neighbouring molecule.The Cu(dth)Cl moieties are thus formed into polymericchains through linking Cu-S bonds.Such a structure, with bridging through sulphur inpreference to chlorine atoms, was unexpected, sincespectroscopic studies suggest that halide bridging occursin other copper(1)-thioether complexes,l and the struc-tures of many other halide-bridged copper( I) complexeshave been reported (e.g.refs. 8 and 9). In contrast,although some complexes are known in which copper(1)atoms are bridged by the sulphur atoms of ligands suchas substituted thioureas lo* l1 and trimethylphosphinesulphide,12 where the S atoms then become three-co-ordinate, to our knowledge only one example of bridgingby a thioether ligand has previously been reported, vix.[Cu414(SEt2) 31 -13Two features of interest in the bonding in [(Cu(dth)-Cl),] are the shortness of the bridging Cu-S bond [2.315( 1)A] compared with the two Cu-S bonds within the chelatering [2.336(1) and 2.342(1) A], and the equivalence of thelatter two bonds. The usual trend in complexes whereboth bridging and terminal bonds exist is for the bridgingto be longer than the terminal Cu-L bonds.Forexample, in [CU,I,(SE~,)~] the bridging Cu-S bonds are2.337(6) and 2.331(8) A while the terminal Cu-S bondsare 2.297(10) and 2.298(9) A. The longer Cu-S bondsin the chelate ring of [(Cu(dth)Cl},] probably do notarise because of strain in the ring, since shorter bonds[average 2.303(5) A] are seen in the closely relatedcomplex [Cu(dto),][BF,] (dto = 3,6-dithiaoctane) .6Rather it is probable that the two ligands whose stereo-chemical arrangement about the copper atom is notrestricted by the constraints of the chelate ring, vix. theterminal C1 atom and the S atom of the neighbouringmolecule, can more easily be arranged to give maximumoverlap with the copper orbitals.The Cu-C1 bondlength does in fact lie at the lower end of the range (2.24-2.40 A) observed for terminal Cu-C1 bonds in othertetrahedral copper( I) c~mplexes,~*-~~ and is close to thatM. R. Churchill and K. L. Karla, J . Amer. Chem. Soc., 1973,95, 5772.A. Camus, G. Nardin, and L. Randaccio, Inorg. Chim. A d a ,1975, 12, 23.lo I. F. Taylor, M. S. Weininger, and E. L. Amma, Inorg. Chem.,1974, 13, 2835.l1 A. L. Crumbliss, L. J. Gestaut, R. C. Rickard, and A. T.McPhail, J . C . S . Chem. Comm.. 1974, 545.l2 J. A. Tiethof, J. K. Stalick, and D. W. Meek, Inorg. Cham.,1973, 12, 1170.observed in three-co-ordinate c0mp1exes.l~~ l7 Thus astronger interaction with C1 and S(1') may allow weakerinteraction with the chelating S atoms.As a consequence of the sulphur bridging in thestructure, the two sulphur atoms of each dth ligand arenot equivalent, S(1) forming four bonds and S(2) onlythree. Despite this, the three ' intramolecular ' bondsmade by each, including the Cu-S bonds of the chelatering, are closely similar in length, and the angles aboutS(l) and S(2) are also very similar.Thus the formationof an extra bond by S(l) apparently does not affect itsstereochemistry or bonding to other ligands at all.The stereochemistry of the copper atom is distortedtetrahedral, The chelate angle [91.54(3)"] is smallerthan in [Cu(dto),][BF,], where all four Cu ligands are inchelate rings and the chelate angle is expanded to anaverage of 94.8(2)". The internal angles a t the S atomsof [(Cu(dth)Cl},] [average 100.0( I)"] are correspondinglylarger than those in [C~(dto)~][BF,l [average 95.9(6)"].6Other bond lengths and angles are normal, and the S-C-C-S bridge has the usual asymmetric gauche configur-ation, with C(2) 0.22A below the CuS, plane and C(3)0.44A above it.Finally, the observation of bridging by a thioetherligand suggests that the side chain of methionine couldact as a bridging ligand in metalloproteins.In thelatter, metal atoms are thought often to occur as pairsor clusters (e.g. the e.s.r.-non-detectable, or Type 111,copper atoms of copper proteins). Possible bridgingligands include cysteinyl sulphur atoms or disulphidegroups, but the structure of [{Cu(dth)Cl},] suggests themethionine side chain as another possibility.We thank Mr. K. C. Palmer for the sample of [{Cu(dth)-Cl],], the Chemistry Division, D.S.I.R., Wellington, for useof their 4-circle diffractometer, Dr. K. L. Brown for assist-ance with data collection, members of the Computer Unit,Massey University, and the Chemistry Department, Uni-versity of Canterbury, for help with computing, and Drs.A. M. Brodie and E. W. Ainscough for many useful dis-cussions.[7/1342 Received, 25th July, 1977113 J. San Filippo, L. E. Zyontz, and J. A. Potenza, Inorg.l4 R. L. Girling and E. L. Amma, Inorg. Chem., 1971, 10, 335.I5 V. G. Albano, P.'L. Bellon, and G. Ciani, J . C . S . Dalton, 1972,16 F. A. Cotton, B. A. Frenz, D. L. Hunter, and 2. C. Mester,17 E. W. Ainscough, H. A. Bergen, A. M. Brodie, and K. L.Chem., 1975, 14, 1667.1938.Inorg. Chim. Acta, 1974, 11, 111.Brown, J . C . S . Dalton, 1976, 1649