摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1998, Pages 475–481 475 4-Aminobenzylidyne: a versatile precursor for extended unsaturated alkylidyne ligands † Marie Pui Yin Yu, Andreas Mayr * and Kung-Kai Cheung Department of Chemistry, The University of Hong Kong, Pokfulam Road, Hong Kong The 4-aminobenzylidyne tungsten complexes [W(CC6H4NH2-4)X(CO)2(pic)2] 1 (X = Cl a or Br b) have been prepared by sequential reaction of [W(CO)6] with LiC6H4N(SiMe3)2-4 in diethyl ether and C2O2Cl2 or C2O2Br2 and 4-methylpyridine (pic) in CH2Cl2.Substitution of the picoline ligands by tmen (Me2NCH2CH2NMe2) and dppe (Ph2PCH2CH2PPh2) afforded the complexes [W(CC6H4NH2-4)X(CO)2(L2)] (L2 = tmen 2, X = Cl a or Br b; dppe 3, X = Cl a or Br b). The amino group of the new alkylidyne complexes undergoes typical functional group transformations. Treatment of complexes 2 with pyridine-2-carbaldehyde afforded the Schiff-base derivatives [W{CC6H4(NCHC5H4N-2)-4}X(CO)2(tmen)] 5 (X = Cl a or Br b).Formylation of complexes 2 with acetic formic anhydride afforded the formamides [W(CC6H4NHCHO-4)X(CO)2(tmen)] 6 (X = Cl a or Br b). The isocyanide derivatives [W(CC6H4NC-4)X(CO)2(tmen)] 11 (X = Cl a or Br b) were obtained by dehydration of complexes 6 with triphosgene–NEt3. The molecular structures of 3a, 4b, 8a, 9a and 10b were determined by X-ray crystallography. Molecules in which metal complex fragments and extended organic p systems are connected via metal–carbon multiple bonds1 have received increasing attention in recent years.Owing to their special electronic properties, such metalla-p systems are of potential interest as components in molecular materials. In this context, transition-metal alkylidyne complexes 2 are promising candidates 3,4 as they possess strong metal–carbon triple bonds.5 We are especially interested in exploring the potential of Fischer-type alkylidyne metal complexes. 4,6 Most representatives of this class of compound are electronically and co-ordinatively saturated, which affords them high stability, and the methods for their synthesis are highly developed, at least as far as the generation of the metal–carbon triple bonds and the variation of the ancillary ligands are concerned. 1,7 However, specific methods to extend the p systems of the alkylidyne ligands and to establish conjugated links with other functionalities across unsaturated alkylidyne ligands have not yet been developed. This situation is, at least in part, a consequence of the particular reaction conditions involved in the preparation of the metal–carbon triple bonds. The employment of highly reactive nucleophiles as well as electrophiles at different stages of alkylidyne metal complex synthesis 7,8 precludes the presence of many types of functional group.Owing to these limitations, it is necessary to develop procedures which allow the modification of unsaturated alkylidyne ligands after formation of the metal–carbon triple bonds.As a possible system for this endeavor, we have identified the 4-aminobenzylidyne ligand. The presence of the amino group is expected to provide a variety of opportunities to extend the benzylidyne ligand in an electronically conjugated manner. Results and Discussion The 4-aminobenzylidyne tungsten complexes 1 are synthesized by reaction of [W(CO)6] with LiC6H4N(SiMe3)2-4 9 in diethyl ether, followed by addition of oxalyl halide and 4-methylpyridine (pic) in CH2Cl2 (Scheme 1).After removal of the solvent from the reaction mixture and recrystallization from methylene chloride–hexane, the products 1 are isolated in 25– 40% yield. No special procedures are required for the desilylation of the amino group. Substitution of the two pic ligands in † Dedicated to Professor Walter Siebert on the occasion of his 60th birthday. 1 by tmen (Me2NCH2CH2NMe2) and dppe (Ph2PCH2CH2- PPh2) affords the derivatives 2 and 3.The amino group of the complexes 1–3 undergoes typical functional group transformations.10 For example, Schiff-base formation of complexes 2 with pyridine-4- and -2-carbaldehyde gives the imines 4 and 5 and reaction of 2 and 3 with acetic formic anhydride and with acetyl, benzoyl as well as isonicotinoyl chloride affords the formyl derivatives 6 and 7 and the acyl derivatives 8–10. The isocyano derivatives 11 and 12 form upon Scheme 1 thf = Tetrahydrofuran [W(CO)6] NH2 C W X CO pic CO pic 1.LiC6H4N(SiMe3)2-4,Et2O 2. C2O2X2, CH2Cl2 3. NC5H4Me-4 (pic) 1a: X = Cl 1b: X = Br 1 L L NH2 C W X CO L CO L 2a: X = Cl, L2 = tmen 2b: X = Br, L2 = tmen 3a: X = Cl, L2 = dppe 3b: X = Br, L2 = dppe 2 RCHO thf thf N C W X CO N CO N C H R 4a: X = Cl, R = C5H4N-4 4b: X = Br, R = C5H4N-4 5a: X = Cl, R = C5H4N-2 5b: X = Br, R = C5H4N-2 2 R¢COCl thf N C W X CO N CO N C O R¢ H 8a: X = Cl, R¢ = Me 9a: X = Cl, R¢ = Ph 10a: X = Cl, R¢ = C5H4N-4 10b: X = Br, R¢ = C5H4N-4 2,3 MeCOOCHO thf NHCHO C W X CO L CO L 6a: X = Cl.L2 = tmen 6b: X = Br, L2 = tmen 7a: X = Cl, L2 = dppe triphosgene, CH2Cl2 NC C W X CO L CO L 11a: X = Cl, L2 = tmen 11b: X = Br, L2 = tmen 12a: X = Cl, L2 = dppe476 J. Chem. Soc., Dalton Trans., 1998, Pages 475–481 treatment of complexes 6 and 7 with triphosgene [(Cl3CO)2- CO]–NEt3.11 The transformations of the amino group in complexes 2 into imine, amide and isocyanide functionalities are accompanied by noticeable changes of the spectroscopic parameters of the tungsten alkylidyne fragment.For example, the stretching frequencies of the carbonyl ligands, which are sensitive to the electron density of the tungsten center, increase by about 10 wavenumbers upon Schiff-base formation and formylation, and by about 15 wavenumbers upon formation of the isocyanide functionality. Considering that the carbonyl ligands on the tungsten center are primarily interacting with the d orbital which is orthogonal to the metal alkylidyne p system,4 these data indicate the presence of significant electronic coupling between the tungsten atom and the remote functional groups.The 13C NMR resonances of the alkylidyne carbon atom shift only slightly in response to the functional group transformations taking place on the opposite side of the arene group. Fig. 1 Molecular structure of complex 3a Fig. 2 Molecular structure of complex 4b Fig. 3 Molecular structure of complex 8a The crystal structures of compounds 3a, 4b, 8a, 9a and 10b have been determined by X-ray crystallography.Selected bond distances and angles are listed in Table 1, and the molecular structures are shown in Figs. 1–5, respectively. The amino group in 3a is strongly conjugated with the phenyl group, as indicated by the short N]C(7) bond length of 1.373(9) Å,12 but this feature is not mirrored in the bonding parameters of the W]C(3)]C(4) fragment. The W]C(3) distance of 1.812(7) Å and the C(3)]C(4) distance of 1.453(9) Å are within the established range for metal–carbon triple and C(sp)]C(sp2) bonds.1,13 There are also no signs of bond localization within the phenyl ring.The same situation is found in the other structures determined in this study. The distances within the W]] ] C]C fragments and the phenyl rings of compounds 4b, 8a, 9a and 10b are unexceptional.1 However, compared with that in 3a, the N]C (phenyl) distance, N(2)]C(9), has lengthened to 1.41(1) Å in the Schiff-base derivative 4b, and to almost the same value in the amide derivatives 8a, 9a and 10b, reflecting a reduction or elimination of the p interaction between the nitrogen atom and the phenyl ring in these compounds.13 Thus, the functional group transformations cause the expected changes of the local bonding parameters, but apparently do not have a significant effect on the bond distances involving the alkylidyne carbon atom.In conclusion, the 4-aminobenzylidyne ligand is a versatile unit for the design of extended unsaturated alkylidyne ligands.Several of the new alkylidyne ligands derived from it, in particular the isocyanide derivatives, are designed for the attachment of additional metal centers which are spacially separated from, yet electronically conjugated with, the alkylidyne metal center. There is spectroscopic evidence for electronic coupling between the metal center and the nitrogen functionality across the unsaturated alkylidyne p system.Experimental General Standard inert-atmosphere techniques were used throughout. Diethyl ether, hexane and tetrahydrofuran were purified by reflux over sodium and distilled under nitrogen. Methylene chloride was heated to reflux over calcium hydride and distilled Fig. 4 Molecular structure of complex 9a Fig. 5 Molecular structure of complex 10bJ. Chem. Soc., Dalton Trans., 1998, Pages 475–481 477 Table 1 Selected bond lengths (Å) and angles (8) under nitrogen.Tungsten hexacarbonyl, oxalyl chloride, oxalyl bromide, 4-methylpyridine, N,N,N9,N9-tetramethylethane-1,2- diamine (tmen) and 1,2-bis(diphenylphosphino)ethane (dppe) were obtained from commercial sources and used as received. Proton, 13C and 31P NMR spectra were recorded on Fourier-transform 270 MHz JEOL JNMGSX270, 300 MHz Bruker DPX300 and 500 MHz Bruker DRX500 spectrometers. Chemical shifts of 1H and 13C are given in parts per million (d) relative to tetramethylsilane, 31P to 85% H3PO4 and proton decoupled.The IR spectra were recorded on a Shimadzu FTIR-8201PC spectrometer. Melting points were recorded on a Stuart Scientific SMP1 instrument under nitrogen. Elemental analyses were performed by Butterworth Laboratories Ltd. Preparations [W(CC6H4NH2-4)Cl(CO)2(pic)2] 1a. A solution of LiC6H4N- (SiMe3)2-4 was prepared by the addition of LiBun in hexane (1.6 M, 8.625 cm3) to BrC6H4N(SiMe3)2-4 9 (11.5 mmol, 3.634 g) in ether (30 cm3) and stirring for 2 h at 0 8C.The resulting solution was transferred to a suspension of [W(CO)6] (10 mmol, 3.52 g) Complex 3a W]Cl W]C(1) W]P(1) N]C(7) C(4)]C(5) C(5)]C(6) C(7)]C(8) Cl]W]C(3) P(1)]W]P(2) P(2)]W]C(2) 2.584(2) 2.013(8) 2.544(2) 1.373(9) 1.376(10) 1.38(1) 1.38(1) 173.9(2) 79.91(6) 93.7(2) W]C(3) W]C(2) W]P(2) C(3)]C(4) C(4)]C(9) C(6)]C(7) C(8)]C(9) W]C(3)]C(4) P(1)]W]C(1) C(1)]W]C(2) 1.812(7) 1.986(9) 2.537(2) 1.453(9) 1.402(9) 1.39(1) 1.38(1) 170.7(6) 95.7(2) 89.9(3) Complex 4b W]C(5) W]Br C(1)]O(1) C(6)]C(7) C(8)]C(9) N(2)]C(10) C(11)]C(12) C(13)]N(3) Br]W]C(5) C(1)]W]C(1) C(1)]W]N(1) N(2)]C(10)]C(11) Br]W]N(1) 1.799(9) 2.696(1) 1.141(8) 1.377(9) 1.373(10) 1.15(1) 1.39(2) 1.28(2) 168.3(3) 87.6(5) 96.8(3) 124(1) 90.9(2) W]C(1) W]N(1) C(5)]C(6) C(7)]C(8) C(9)]N(2) C(10)]C(11) C(12)]C(13) N(3)]C(15) W]C(5)]C(6) N(1)]W]N(1) C(9)]N(2)]C(10) Br]W]C(1) 1.986(9) 2.318(6) 1.45(1) 1.396(10) 1.41(1) 1.52(2) 1.36(2) 1.33(2) 173.6(7) 78.6(4) 122(1) 87.3(2) Complex 8a W]Cl W]N(2) W]C(1) C(3)]C(4) C(4)]C(9) C(6)]C(7) C(8)]C(9) N(1)]C(10) C(10)]O(3) C(2)]O(2) Cl]W]C(3) C(1)]W]C(2) C(2)]W]N(2) Cl]W]C(1) Cl]W]N(2) C(3)]W]C(1) C(3)]W]N(2) C(7)]N(1)]C(10) N(1)]C(10)]O(3) 2.585(2) 2.312(6) 1.972(9) 1.467(9) 1.398(9) 1.393(9) 1.375(9) 1.340(9) 1.222(8) 1.155(9) 168.4(2) 85.9(3) 99.3(3) 88.5(2) 89.3(2) 84.8(3) 98.1(2) 123.8(6) 122.8(7) W]C(3) W]N(3) W]C(2) C(4)]C(5) C(5)]C(6) C(7)]C(8) N(1)]C(7) C(10)]C(11) C(1)]O(1) W]C(3)]C(4) C(1)]W]N(3) N(2)]W]N(3) Cl]W]C(2) Cl]W]N(3) C(3)]W]C(2) C(3)]W]N(3) N(1)]C(10)]C(11) C(11)]C(10)]O(3) 1.799(7) 2.295(5) 1.980(8) 1.390(9) 1.363(9) 1.380(9) 1.424(9) 1.51(1) 1.143(9) 167.7(5) 96.3(2) 78.4(2) 88.8(2) 87.9(2) 81.2(3) 102.3(2) 114.5(7) 122.7(7) in ether (20 cm3) at room temperature (r.t).The mixture was stirred for 1 h at r.t., then concentrated (to about 5 cm3), and hexane added to precipitate the acyltungsten intermediate, which was filtered off and washed with hexane (5 × 20 cm3).(If the solvent is simply removed at this stage, the product contains up to 10% of the bromo analogue 1b.) The solid residue was redissolved in CH2Cl2 (80 cm3) and filtered. After cooling to 278 8C, C2O2Cl2 (10.1 mmol, 0.88 cm3) was added. The resulting mixture was allowed to warm up to 220 8C and 4- methylpyridine (5 cm3) added. The mixture was stirred at r.t. for 2 h and the solvent then removed in vacuo. The residue was washed with hexane, dried and redissolved in CH2Cl2.After filtration, hexane was added to the solution to afford orangeyellow crystals. Yield: 1.46 g, 26%, m.p. 114–117 8C (decomp.). 1H NMR (CDCl3): d 8.86 (d, J = 6.48, 4 H, C5H4N), 7.17 (d, J = 8.51, 2 H, C6H4NH2), 7.06 (d, J = 6.13, 4 H, C5H4N), 6.52 (d, J = 8.53 Hz, 2 H, C6H4NH2), 3.6 (br, 2 H, NH2) and 2.35 (s, 6 H, NCH3). 13C NMR (CDCl3): d 265.2 (W]] ] C), 221.5 (CO), 152.4, 150.0, 146.4, 140.8, 131.1, 125.8, 114.0 (C5H4N, C6H4NH2) and 21.1 (NCH3). IR (CH2Cl2, cm21): 1979s (nCO) and 1890s (nCO) [Found (Calc.): C, 44.42 (44.59); H, 3.55 (3.56); N, 6.99 (7.43)%] Complex 9a W]Cl W]C(1) W]N(1) N(3)]C(7) O(3)]C(10) C(4)]C(5) C(5)]C(6) C(7)]C(8) C(10)]C(11) C(11)]C(16) C(13)]C(14) C(15)]C(16) Cl]W]C(3) N(3)]C(10)]C(11) O(3)]C(10)]C(11) N(1)]W]N(2) N(2)]W]C(1) 2.579(2) 2.00(1) 2.292(7) 1.42(1) 1.23(1) 1.40(1) 1.38(1) 1.38(1) 1.49(1) 1.40(1) 1.40(2) 1.40(1) 170.7(3) 117.2(8) 121.5(9) 78.9(3) 97.2(4) W]C(3) W]C(2) W]N(2) N(3)]C(10) C(3)]C(4) C(4)]C(9) C(6)]C(7) C(8)]C(9) C(11)]C(12) C(12)]C(13) C(14)]C(15) W]C(3)]C(4) O(3)]C(10)]N(3) C(7)]N(3)]C(10) N(1)]W]C(2) C(1)]W]C(2) 1.810(9) 1.98(1) 2.315(7) 1.35(1) 1.45(1) 1.42(1) 1.39(1) 1.37(1) 1.38(1) 1.39(1) 1.35(2) 173.5(7) 121.2(9) 124.9(8) 98.8(4) 85.0 Complex 10b W]C(3) W]C(1) W]N(1) C(1)]O(1) C(3)]C(4) C(4)]C(9) C(6)]C(7) C(8)]C(9) N(3)]C(10) C(10)]C(11) C(12)]C(13) N(4)]C(14) C(15)]C(11) Br]W]C(3) C(1)]W]C(2) N(1)]W]N(2) Br]W]C(1) Br]W]N(1) C(7)]N(3)]C(10) O(3)]C(10)]C(11) 1.807(7) 1.974(9) 2.292(7) 1.16(1) 1.440(10) 1.39(1) 1.38(1) 1.39(1) 1.339(9) 1.50(1) 1.39(1) 1.33(1) 1.40(1) 169.6(2) 86.2(4) 78.8(3) 89.2(2) 89.2(2) 123.9(6) 120.5(7) Br]W W]C(2) W]N(2) C(2)]O(2) C(4)]C(5) C(5)]C(6) C(7)]C(8) N(3)]C(7) C(10)]O(3) C(11)]C(12) C(13)]C(14) C(14)]C(15) W]C(3)]C(4) C(1)]W]N(2) N(1)]W]C(2) Br]W]C(2) Br]W]N(2) N(3)]C(10)]O(3) N(3)]C(10)]C(11) 2.6975(9) 1.96(1) 2.291(6) 1.14(1) 1.39(1) 1.375(10) 1.37(1) 1.426(9) 1.208(10) 1.37(1) 1.31(1) 1.38(1) 173.2(6) 98.4(3) 96.6(3) 88.1(3) 91.9(2) 122.3(7) 117.2(7)478 J.Chem. Soc., Dalton Trans., 1998, Pages 475–481 [W(CC6H4NH2-4)Br(CO)2(pic)2] 1b. The synthesis followed the procedure described for complex 1a, whereby C2O2Br2 was used instead of C2O2Cl2. (The solvent was removed directly from the acyltungsten intermediate.) Orange-yellow crystals. Yield: 37%, m.p. 115–120 8C (decomp.). 1H NMR (CDCl3): d 8.92 (d, J = 6.35, 4 H, C5H4N), 7.18 (d, J = 8.54, 2 H, C6H4NH2), 7.07 (d, J = 6.11, 4 H, C5H4N), 6.52 (d, J = 8.54 Hz, 2 H, C6H4NH2), 3.52 (br, 2 H, NH2) and 2.36 (s, 6 H, NCH3). 13C NMR (CDCl3): d 265.1 (W]] ] C), 220.8 (CO), 153.0, 150.0, 146.4, 140.3, 131.0, 125.8, 114.0 (C6H4NH2, C5H4N) and 21.2 (CH3). IR (CH2Cl2, cm21): 1979s (nCO) and 1890s (nCO) [Found (Calc.): C, 40.95 (41.34); H, 3.37 (3.30); N, 6.92 (6.89)%]. [W(CC6H4NH2-4)Cl(CO)2(tmen)] 2a. Complex 1a (1 mmol, 0.565 g) was dissolved in CH2Cl2 (50 cm3) and tmen (1 cm3) added.The resulting mixture was stirred at 50 8C for 2 h and the solvent then removed in vacuo. The residue was washed with hexane, dried and redissolved in CH2Cl2. After filtration, hexane was added to the solution to afford orange-yellow crystals. Yield: 0.28 g, 56%, m.p. 160–165 8C (decomp.). 1H NMR (CDCl3): d 7.09 (d, J = 8.79, 2 H, C6H4NH2), 6.50 (d, J = 8.79 Hz, 2 H, C6H4NH2), 3.80 (br, 2 H, NH2), 3.18 (s, 6 H, NCH3), 3.01–2.81 (br, 4 H, NCH2) and 2.91 (s, 6 H, NCH3). 13C NMR (CDCl3): d 264.8 (W]] ] C), 221.5 (CO), 146.4, 140.3, 131.1, 114.0 (C6H4NH2), 60.9, 58.0, 52.0 [CH2N(CH3)2].IR (CH2Cl2, cm21): 1976s (nCO) and 1884s (nCO) [Found (Calc.): C, 36.31 (36.35); H, 4.44 (4.47); N, 8.46 (8.48)%]. [W(CC6H4NH2-4)Br(CO)2(tmen)] 2b. Orange-yellow crystals. Yield: 77%, m.p. 150–160 8C (decomp.). 1H NMR (CDCl3): d 7.11 (d, J = 8.44, 2 H, C6H4NH2), 6.50 (d, J = 8.46 Hz, 2 H, C6H4NH2), 3.83 (br, 2 H, NH2), 3.22 (s, 6 H, NCH3), 3.02 (s, 6 H, NCH3) and 2.90 (m, 4 H, NCH2). 13C NMR (CDCl3): d 264.6 (W]] ] C), 220.8 (CO), 146.4, 139.7, 131.0, 114.0 (C6H4- NH2), 61.1, 58.3, 53.4 [CH2N(CH3)2]. IR (CH2Cl2, cm21): 1979s (nCO) and 1886s (nCO) [Found (Calc.): C, 33.58 (33.36); H, 4.03 (4.11); N, 7.78 (7.78)%]. [W(CC6H4NH2-4)Cl(CO)2(dppe)] 3a. Complex 1a (1 mmol, 0.565 g) was dissolved in CH2Cl2 (50 cm3) and dppe (1.1 mmol, 0.438 g) added. The resulting mixture was stirred at 50 8C for 2 h and the solvent then removed in vacuo. The residue was washed with hexane, dried and redissolved in CH2Cl2.After filtration, hexane was added to the solution to afford yellow crystals. Yield: 0.46 g, 59%, m.p. 154–160 8C (decomp.). 1H NMR (CDCl3): d 7.74–7.17 (m, 20 H, PPh2), 6.36 (d, J = 8.54, 2 H, C6H4NH2), 6.20 (d, J = 8.79 Hz, 2 H, C6H4NH2), 3.72 (br, 2 H, NH2) and 2.97–2.50 (4 H, CH2PPh2). 13C NMR (CDCl3): d 270.1 (W]] ] C, 1Jcis PC = 10), 213.1 (CO, 1Jcis PC = 7, 1Jtrans PC = 45 Hz), 146.2, 140.5, 136.0, 135.4, 133.0, 132.8, 132.7, 131.5, 130.0, 129.9, 128.5, 128.4, 113.3 (PPh2, C6H4NH2), 27.6, 27.4, 27.2, 27.0 (CH2PPh2). 31P NMR (CDCl3): d 39.0 (1JWP = 229 Hz). IR (CH2Cl2, cm21): 1998s (nCO) and 1929s (nCO) [Found (Calc.): C, 54.03 (54.04); H, 3.80 (3.89); N, 1.96 (1.80)%]. [W(CC6H4NH2-4)Br(CO)2(dppe)] 3b. Yellow crystals. Yield: 80%, m.p. 175–178 8C (decomp.). 1H NMR (CDCl3): d 7.72– 7.19 (m, 20 H, PPh2), 6.50 (d, J = 8.46, 2 H, C6H4NH2), 6.26 (d, J = 8.51 Hz, 2 H, C6H4NH2), 3.70 (br, 2 H, NH2) and 3.03–2.51 (4 H, CH2PPh2). 13C NMR (CDCl3): d 269.6 (W]] ] C, 1Jcis PC = 10), 211.8 (CO, 1Jtrans PC = 43, 1Jcis PC = 7 Hz), 145.9, 140.3, 135.9, 135.6, 133.2, 132.9, 132.8, 132.7, 131.4, 130.0, 128.4, 128.3, 128.1, 113.6 (PPh2, C6H4NH2), 27.4, 27.3, 27.2, 27.1 (CH2- PPh2). 31P NMR (CDCl3): d 36.4 (1JWP = 229 Hz). IR (CH2Cl2, cm21): 1998s (nCO) and 1929s (nCO) [Found (Calc.): C, 50.62 (51.12); H, 3.50 (3.68); N, 1.77 (1.70)%]. [W{CC6H4(NCHC5H4N-4)-4}Cl(CO)2(tmen)] 4a. Complex 2a (1 mmol, 0.496 g) was dissolved in thf (100 cm3) and pyridine-4-carbaldehyde (0.2 cm3) added with tmen (0.5 cm3).The resulting mixture was stirred under reflux overnight and the solvent then removed in vacuo. The residue was washed with hexane, dried and redissolved in CH2Cl2. After filtration, hexane was added to the solution to afford light orange crystals. Yield: 0.47 g, 81%, m.p. 180–188 8C (decomp.). 1H NMR (CDCl3): d 8.76 (d, J = 6.10, 2 H, C5H4N), 8.42 (s, 1 H, NCH), 7.74 (d, J = 6.11, 2 H, C5H4N), 7.29 (d, J = 8.31, 2 H, C6H4NCH), 7.13 (d, J = 8.30 Hz, 2 H, C6H4NCH), 3.24 (s, 6 H, NCH3), 3.06–2.92 (br, 4 H, NCH2) and 2.96 (s, 6 H, NCH3). 13C NMR (CDCl3): d 261.0 (W]] ] C), 220.9 (CO), 157.5 (NCH), 150.6, 149.5, 148.0, 142.7, 130.4, 122.2, 120.9, 114.0 (C6H4N, C5H4N), 61.0, 58.2, 52.2 [CH2N(CH3)2]. IR (CH2Cl2, cm21): 1985s (nCO) and 1894s (nCO) [Found (Calc.) (with 0.25 mol CH2Cl2): C, 42.02 (42.12); H, 4.20 (4.24); N, 9.18 (9.25)%]. [W{CC6H4(NCHC5H4N-4)-4}Br(CO)2(tmen)] 4b.Redorange crystals. Yield: 64%, m.p. 185–190 8C (decomp.). 1H NMR (CDCl3): d 8.76 (d, J = 5.98, 2 H, C5H4N), 8.42 (s, 1 H, NCH), 7.74 (d, J = 6.11, 2 H, C5H4N), 7.31 (d, J = 8.54, 2 H, C6H4NCH), 7.16 (d, J = 8.35 Hz, 2 H, C6H4NCH), 3.28 (s, 6 H, NCH3), 3.07 (s, 6 H, NCH3) and 2.96 (m, 4 H, NCH2). 13C NMR (CDCl3): d 260.8 (W]] ] C), 220.2 (CO), 157.6 (NCH), 150.6, 149.7, 147.4, 142.7, 131.0, 130.3, 122.2, 121.0, 114.0 (C6H4NCH, C5H4N), 61.2, 58.6, 53.5 [CH2N(CH3)2].IR (CH2Cl2, cm21): 1987s (nCO) and 1896s (nCO) [Found (Calc.): C, 39.91 (40.09); H, 3.86 (4.00); N, 8.72 (8.90)%]. [W{CC6H4(NCHC5H4N-2)-4}Cl(CO)2(tmen)] 5a. The synthesis followed the procedure described for complex 4a, whereby pyridine-2-carbaldehyde was used. Red-orange crystals. Yield: 43%, m.p. 115–120 8C (decomp.). 1H NMR (CDCl3): d 8.72 (d, J = 4.52, 1 H, 2-C5H4N), 8.57 (s, 1 H, NCH), 8.17 (d, J = 7.81, 1 H, 2-C5H4N), 7.82 (dt, J = 7.76, 1.47, 1 H, 2-C5H4N), 7.39 (ddd, J = 7.51, 4.82, 1.22, 1 H, 2-C5H4N), 7.30 (d, J = 8.54, 2 H, C6H4NCH), 7.18 (d, J = 8.54 Hz, 2 H, C6H4NCH), 3.24 (s, 6 H, CH3), 3.06–2.87 (br, 4 H, NCH2) and 2.96 (s, 6 H, NCH3). 13C NMR (CDCl3): d 261.5 (W]] ]C), 221.0 (CO), 160.2 (NCH), 154.3, 149.8, 147.7, 136.8, 131.1, 130.5, 125.3, 122.2, 121.1, 114.0 (C6H4NCH, C5H4N), 61.0, 58.2, 53.5 [CH2N(CH3)2]. IR (CH2Cl2, cm21): 1985s (nCO) and 1894s (nCO) [Found (Calc.): C, 43.03 (43.13); H, 4.37 (4.31); N, 9.37 (9.58)%].[W{CC6H4(NCHC5H4N-2)-4}Br(CO)2(tmen)] 5b. Redorange crystals. Yield: 0.274 g, 24%, m.p. 115–120 8C (decomp.). 1H NMR (CDCl3): d 8.72 (d, J = 3.91, 1 H, 2- C5H4N), 8.57 (s, 1 H, NCH), 8.18 (d, J = 7.81, 1 H, 2-C5H4N), 7.82 (td, J = 7.69, 1.22, 1 H, 2-C5H4N), 7.39 (ddd, J = 7.39, 4.38, 1.22, 1 H, 2-C5H4N), 7.32 (d, J = 8.54, 2 H, C6H4NCH), 7.17 (d, J = 8.79 Hz, 2 H, C6H4NCH), 3.28 (s, 6 H, NCH3), 3.06 (s, 6 H, NCH3) and 2.96 (m, 4 H, NCH2). 13C NMR (CDCl3): d 261.4 (W]] ] C), 220.4 (CO), 160.3 (NCH), 154.4, 149.8, 147.1, 136.8, 130.3, 125.3, 122.1, 121.1, 114.0 (C6H4NCH, C5H4N), 61.2, 58.5, 53.5 [CH2N(CH3)2].IR (CH2Cl2, cm21): 1987s (nCO) and 1896s (nCO) [Found (Calc.): C, 39.78 (40.09); H, 3.70 (4.00); N, 8.64 (8.90)%]. [W(CC6H4NHCHO-4)Cl(CO)2(tmen)] 6a. Complex 2a (6 mmol, 2.974 g) was dissolved in thf (50 cm3) and acetic formic anhydride (0.6 cm3) was added at 0 8C. The resulting mixture was stirred at 0 8C for 15 min and the solvent then removed in vacuo. The residue was washed with anhydrous ether, dried and redissolved in CH2Cl2. After filtration, hexane was added to the solution to afford yellow-orange crystals. Yield: 2.86 g, 91%, m.p. 150–158 8C (decomp.). 1H NMR (CDCl3), two isomers: major isomer, d 8.36 (d, J = 1.71, 1 H, NHCHO), 7.33 (br, 1 H, NHCHO), 7.44 (d, J = 8.55, 2 H, C6H4NH), 7.21 (d, J = 8.55, 2 H, C6H4NH); minor isomer, 8.67 (d, J = 11.47, 1 H, NHCHO), 7.55 (d, J = 11.23, 1 H, NHCHO), 7.23 (d, J = 8.30, 2 H, C6H4NH), 6.94 (d, J = 8.55 Hz, 2 H, C6H4NH), 3.21 (s, 6 H, NCH3), 3.00–2.83 (br, 4 H, NCH2) and 2.94 (s, 6 H, NCH3). 13CJ. Chem. Soc., Dalton Trans., 1998, Pages 475–481 479 NMR (CDCl3), two isomers: major isomer, d 261.6 (W]] ] C), 221.0 (CO), 158.8 (NHCHO), 145.7, 136.0, 130.3, 119.4 (C6H4NHCHO); minor isomer, 260.1 (W]] ] C), 220.8 (CO), 161.7 (NHCHO), 146.2, 135.5, 130.9, 118.0 (C6H4NHCHO), 61.0, 58.1, 52.5 [CH2N(CH3)2]. IR (CH2Cl2, cm21): 1985s (nCO), 1894s (nCO) and 1705m (nC]] O) [Found (Calc.) (with 0.25 mol CH2Cl2): C, 35.92 (35.82); H, 4.13 (4.16); C, 7.73 (7.71)%].[W(CC6H4NHCHO-4)Br(CO)2(tmen)] 6b. Orange-yellow crystals. Yield: 76%, m.p. 165–167 8C (decomp.). 1H NMR (CDCl3), two isomers: major isomer, d 8.37 (d, J = 1.55, 1 H, NHCHO), 7.43 (d, J = 8.56, 2 H, C6H4NH), 7.25 (d, J = 8.73, 2 H, C6H4NH); minor isomer d 8.69 (d, J = 11.36, 1 H, NHCHO), 7.56 (d, J = 9.61, 1 H, C6H4NH), 7.23 (d, J = 8.70, 2 H, C6H4NH), 6.93 (d, J = 8.51 Hz, 2 H, C6H4NH), 3.25 (s, 6 H, NCH3), 3.04 (s, 6 H, NCH3) and 2.94 [m, 4 H, CH2N(CH3)2]. 13C NMR (CDCl3), two isomers: major isomer, d 261.1 (W]] ] C), 220.3 (CO), 158.6 (C]] O), 136.0, 130.1, 119.4 (C6H4NH); minor isomer, 259.8 (W]] ] C), 220.1 (CO), 161.5 (C]] O), 135.5, 130.7, 118.0 (C6H4NH), 61.2, 58.5, 53.5 [CH2N(CH3)2]. IR (CH2Cl2, cm21): 1987s (nCO), 1896s (nCO) and 1707m (nC]] O) [Found (Calc.): C, 34.12 (33.83); H, 3.90 (3.90); N, 7.40 (7.40)%]. [W(CC6H4NHCHO-4)Cl(CO)2(dppe)] 7a.Yellow crystals. Yield: 48%, m.p. 150–152 8C (decomp.). 1H NMR (CDCl3): d 8.61 (d, J = 11.23 Hz, 1 H, major isomer NHCHO), 8.16 (s, 1 H, minor isomer NHCHO), 7.74–6.44 (25 H, PPh2, C6H4NHCHO, NHCHO), 3.52–2.91 (m, 2 H, CH2PPh2) and 2.8–2.60 (m, 2 H, CH2PPh2). 13C NMR (CDCl3): d 267.1 (W]] ] C), 212.6 (CO, major isomer, 1Jcis PC = 8, 1Jtrans PC = 44), 212.4 (CO, minor isomer, 1Jcis PC = 8, 1Jtrans PC = 45 Hz), 161.5 (NHCHO, major isomer), 158.8 (NHCHO, minor isomer), 146.2, 145.9, 145.6, 136.3, 136.1, 136.0, 135.8, 135.3, 135.2, 132.9, 132.8, 132.7, 132.5, 132.2, 132.1, 131.9, 131.5, 131.1, 130.8, 130.7, 130.5, 130.2, 129.2, 128.9, 128.6, 128.5, 128.2, 127.9, 118.5, 118.2, 116.8 (PPh2, C6H4NHCHO), 27.5, 27.3, 27.1, 26.9 (CH2PPh2). 31P NMR (CDCl3): d 38.9 (1JWP = 230, major isomer) and 38.7 (1JWP = 231 Hz, minor isomer). IR (CH2Cl2, cm21): 2006s (nCO), 1936s (nCO) and 1705m (nC]] O). [W(CC6H4NHCOMe3-4)Cl(CO)2(tmen)] 8a. Complex 2a (0.5 mmol, 0.248 g) was dissolved in thf (50 cm3) and acetyl chloride (1 mmol, 0.07 cm3) was added at 0 8C.The resulting mixture was stirred at 0 8C for 15 min and the solvent then removed in vacuo. The residue was washed with anhydrous ether, dried, and redissolved in CH2Cl2. After filtration, hexane was added to the solution to afford yellow-orange crystals. Yield: 0.207 g, 77%, m.p. 180–188 8C (decomp.). 1H NMR (CDCl3): d 7.38 (d, J = 8.54, 2 H, C6H4NH), 7.23 (br, 1 H, NH), 7.20 (d, J = 8.54 Hz, 2 H, C6H4NH), 3.20 (s, 6 H, NCH3), 3.04–2.86 (br, 4 H, NCH2), 2.94 (s, 6 H, NCH3) and 2.17 (s, 3 H, COCH3). 13C NMR (CDCl3): d 261.7 (W]] ] C), 221.0 (CO), 168.1 (NHCO), 136.9, 130.2, 119.1 (C6H4NH), 61.0, 58.1, 52.1 [CH2N(CH3)2] and 24.7 (COCH3). IR (CH2Cl2, cm21): 1985s (nCO), 1892s (nCO) and 1697w (nC]] O) [Found (Calc.) (with 0.25 mol CH2Cl2): C, 37.09 (37.07); H, 4.27 (4.42); N, 7.57 (7.52)%]. [W(CC6H4NHCOPh-4)Cl(CO)2(tmen)] 9a. The synthesis followed the procedure described for complex 8a, whereby benzoyl chloride was used instead of acetyl chloride.Yellow-orange crystals. Yield: 67%, m.p. 195–199 8C (decomp.). 1H NMR (CDCl3): d 7.78 (br, 1 H, NH), 7.88–7.24 (9 H, C6H4NH, Ph), 3.23 (s, 6 H, NCH3), 3.06–2.88 (br, 4 H, NCH2) and 2.95 (s, 6 H, NCH3). 13C NMR (CDCl3): d 261.5 (W]] ] C), 221.0 (CO), 165.4 (NHCO), 137.0, 134.7, 132.1, 130.3, 128.9, 127.0, 120.0 (C6H4NH, Ph), 61.0, 58.2, 52.2 [CH2N(CH3)2]. IR (CH2Cl2, cm21): 1985s (nCO), 1894s (nCO) and 1682w (nC]] O) [Found (Calc.): C, 43.77 (44.06); H, 4.29 (4.37); N, 6.98 (7.01)%].[W{CC6H4NHCO(C5H4N-4)-4}Cl(CO)2(tmen)] 10a. Complex 2a (0.5 mmol, 0.248 g) was dissolved in thf (50 cm3) and NEt3 (1 cm3) added. Then isonicotinoyl chloride hydrochloride (0.1 g) was added. The resulting mixture was stirred at 50 8C for 2 h, then filtered and the solvent removed in vacuo. The residue was washed with anhydrous ether, dried, and redissolved in CH2Cl2. After filtration, hexane was added to the solution to afford yellow-orange crystals.Yield: 0.1636 g, 54%, m.p. 170– 175 8C (decomp.). 1H NMR (CD3CN): d 8.92 (1 H, s, NHCO), 8.76 (d, J = 6.11, 2 H, C5H4N), 7.66 (d, J = 8.63, 2 H, C6H4NH), 7.32 (d, J = 8.53 Hz, 2 H, C6H4NH), 7.27 (d, J = 5.73, 2 H, C5H4N), 3.19 (s, 6 H, CH3), 2.94 [m, 4 H, CH2N(CH3)2] and 2.84 (s, 6 H, CH3). 13C NMR (CD3CN): d 263.4 (W]] ] C), 223.5 (CO), 165.0 (NHCO), 151.5, 150.6, 146.5, 142.9, 138.7, 131.1, 122.3, 120.9 (C6H4NH, C5H4N) 61.8, 58.6, 53.8 [CH2N(CH3)2].IR (CH2Cl2, cm21): 1985s (nCO), 1894s (nCO) and 1686w (nC]] O) [Found (Calc.): C, 41.70 (41.99); H, 4.18 (4.20); N, 9.07 (9.33)%]. [W{CC6H4NHCO(C5H4N-4)-4}Br(CO)2(tmen)] 10b. Yelloworange crystals. Yield: 72%, m.p. 180–183 8C (decomp.). 1H NMR (CDCl3): d 8.93 (s, 1 H, NHCO), 8.75 (d, J = 6.08, 2 H, C5H4N), 7.76 (d, J = 6.10, 2 H, C5H4N), 7.65 (d, J = 8.67, 2 H, C6H4NH), 7.34 (d, J = 8.68 Hz, 2 H, C6H4NH), 3.22 (s, 6 H, NCH3), 3.01–2.90 (br, 4 H, NCH2) and 2.96 (s, 6 H, NCH3). 13C NMR (CDCl3): d 263.7 (W]] ] C), 222.6 (CO, 1Jcis WC = 171 Hz), 164.9 (NHCO), 151.3, 145.7, 142.7, 138.7, 130.9, 130.8, 122.1, 120.8 (C6H4NHCO), 61.6, 58.7, 53.6 (CH2NCH3). IR (CH2Cl2, cm21): 1987s (nCO) and 1896s (nCO). [W(CC6H4NC-4)Cl(CO)2(tmen)] 11a. Complex 6a (1 mmol, 0.524 g) was dissolved in CH2Cl2 (50 cm3) and NEt3 (0.56 cm3) added. After cooling to 278 8C, a solution of triphosgene (0.2 g) in CH2Cl2 (10 cm3) was added. The resulting mixture was allowed to warm up to 0 8C and was stirred at 0 8C for 30 min.The solvent was then removed in vacuo. The residue was washed with hexane, redissolved in thf (30 cm3) and then filtered. The solvent was again removed in vacuo. The residue was redissolved in CH2Cl2. After filtration, hexane was added to the solution to afford red-orange crystals. Yield: 0.20 g, 40%, m.p. 80–86 8C (decomp.). 1H NMR (CDCl3): d 7.17–7.25 (br, 4 H, C6H4NC), 3.21 (s, 6 H, NCH3), 3.01–2.85 (br, 4 H, NCH2) and 2.95 (s, 6 H, NCH3). 13C NMR (CDCl3): d 257.5 (W]] ] C), 220.5 (CO), 164.3 (NC), 149.6, 130.1, 126.3 (C6H4NC), 61.1, 58.2, 52.3 [CH2N(CH3)2]. IR (CH2Cl2, cm21): 2124m (nCN), 1992s (nCO) and 1902s (nCO) [Found (Calc.) (with 0.5 mol CH2Cl2): C, 35.95 (36.16), H, 3.80 (3.86); N, 7.62 (7.67)%]. [W(CC6H4NC-4)Br(CO)2(tmen)] 11b. Red-orange crystals. Yield: 54%, m.p. 95–100 8C (decomp.). 1H NMR (CDCl3): d 7.26 (br, 4 H, C6H4NC), 3.25 (s, 6 H, NCH3), 3.05 (s, 6 H, NCH3) and 2.95 (m, 4 H, CH2NMe2). 13C NMR (CDCl3): d 257.3 (W]] ] C), 219.7 (CO), 165.6 (NC), 149.1, 129.9, 126.3 (C6H4NC), 61.2, 58.6, 53.6 [CH2N(CH3)2]. IR (CH2Cl2, cm21): 2124m (nCN), 1992s (nCO) and 1904s (nCO) [Found (Calc.) (with 0.25 mol CH2Cl2): C, 33.96 (34.16); H, 3.60 (3.62); N, 7.03 (7.36)%]. [W(CC6H4NC-4)Cl(CO)2(dppe)] 12a. Yellow microcrystals. Yield 30%, m.p. 115–118 8C (decomp.). 1H NMR (CDCl3): d 7.74–7.19 (20 H, PPh2), 6.91 (d, J = 8.56, 2 H, C6H4NC), 6.39 (d, J = 8.48 Hz, 2 H, C6H4NC), 3.03–2.84 (m, 2 H, CH2PPh2) and 2.77–2.58 (m, 2 H, CH2PPh2). 13C NMR (CDCl3): d 261.7 (W]] ] C), 212.0 (CO, 1Jcis PC = 7, 1Jtrans PC = 46 Hz), 165.5 (N]] ] C), 149.7, 135.9, 133.0, 132.9, 132.5, 132.4, 130.4, 130.2, 128.7, 128.6, 128.5, 125.5 (PPh2, C6H4NC), 27.5, 27.4, 27.2, 27.0 (CH2PPh2). 31P NMR (CDCl3): d 38.2 (1JWP = 230 Hz). IR (CH2Cl2, cm21): 2124w (nCN), 2010s (nCO) and 1944s (nCO) [Found (Calc.): C, 54.66 (54.88); H, 3.22 (3.58); N, 1.82 (1.78)%].480 J.Chem. Soc., Dalton Trans., 1998, Pages 475–481 Table 2 Crystal data and collection parameters for complexes 3a, 4b, 8a, 9a and 10b Molecular formula M Crystal system Space group a/Å b/Å c/Å b/8 U/Å3 Z Dc/g cm23 m/cm21 F(000) T/K Crystal dimensions/mm Diffractometer Total reflections measured Unique reflections Reflections used Parameters R R9 Goodness of fit (D/s) Maximum, minimum, peak in final Fourier map/e Å23 3a C35H30ClNO2P2W?CH2Cl2 862.81 Monoclinic P21/c (no. 14) 19.851(2) 9.089(3) 21.874(2) 113.16(1) 3628.6(9) 4 1.579 35.28 1704 301 0.25 × 0.15 × 0.45 Enraf-Nonius CAD4 6288 6204 4409 406 0.033 0.041 2.51 0.03 1.30, 20.66 4b C21H25BrN4O2W 629.21 Orthorhombic Pnma (no. 62) 15.563(6) 10.855(6) 13.592(4) 2296(2) 4 1.820 68.05 1216 301 0.25 × 0.15 × 0.35 Rigaku AFC7R 2320 2320 1525 151 0.030 0.034 1.67 0.00 0.79, 20.61 8a C17H24ClN3O3W 537.70 Monoclinic P21/n (no. 14) 11.756(4) 12.972(5) 14.510(4) 112.17(2) 2049(1) 4 1.743 57.9 1048 301 0.20 × 0.15 × 0.35 Rigaku AFC7R 3548 3368 2722 230 0.029 0.036 2.13 0.04 1.13, 21.29 9a C22H26ClN3O3W 599.77 Monoclinic P21/c (no. 14) 10.642(5) 13.764(6) 15.915(5) 95.71(3) 2320(1) 4 1.717 51.28 1176 301 0.20 × 0.15 × 0.40 Rigaku AFC7R 3436 3237 2621 271 0.031 0.046 2.50 0.01 0.98, 20.47 10b C21H25BrN4O3W 645.21 Monoclinic P21/c (no. 14) 10.680(3) 13.549(4) 16.015(4) 95.73(2) 2305.8(10) 4 1.858 67.82 1248 301 0.20 × 0.15 × 0.10 Mar IPDS 14 592 4397 3210 271 0.038 0.047 1.63 0.00 1.07, 22.38 Crystallography Complex 3a.Enraf-Nonius CAD4 diffractometer with graphite-monochromatized Mo-Ka radiation (l = 0.71 073 Å), w–2q scans, 6288 reflections measured (2qmax = 488), 6204 independent, 4409 with I > 3s(I) considered observed. Structure solution by Patterson and Fourier methods (PATTY14) and refinement using the software package TEXSAN15 on a Silicon Graphics Indy computer. One formula unit constitutes a crystallographic asymmetric unit. All 45 non-H atoms were refined anisotropically.Atoms H(1) and H(2) bonded to N were located in a Fourier-difference synthesis, and the other 30 H atoms placed at calculated positions with thermal parameters equal to 1.3 times that of the attached C atoms were not refined. Convergence for 406 parameters by full-matrix leastsquares refinement on F with w = 4Fo 2/s2(Fo 2), where s2(Fo 2) = [s2(I) 1 (0.013Fo 2)2] for 4409 reflections with I > 3s(I), was reached at R = 0.033 and R9 = 0.041 with a goodness of fit of 2.51.(D/s)max = 0.03. The final Fourier-difference map was featureless, with maximum positive and negative peaks of 1.30 and 0.66 e Å23, respectively. The crystal structures of compounds 4b, 8a, 9a and 10b were solved by a similar procedure using other diffractometers. Details are given in Table 2. The thermal ellipsoids in the ORTEP16 drawings of Figs. 1–5 are drawn at the 40% probability level. CCDC reference number 186/804. See http://www.rsc.org/suppdata/dt/1998/475/ for crystallographic files in .cif format.Acknowledgements Support for this work by the Committee on Research and Conference Grants (CRCG) is gratefully acknowledged. M. P. Y. Y. acknowledges the receipt of a Postgraduate Studentship, administered by The University of Hong Kong. References 1 M. H. Chisholm, Angew. Chem., 1991, 103, 690; Angew. Chem., Int. Ed. Engl., 1991, 30, 673; M. B. Sponsler, Organometallics, 1995, 14, 1920. 2 E. O. Fischer, Angew. Chem., 1974, 86, 651; Adv.Organomet. Chem., 1976, 14, 1; R. R. Schrock, Acc. Chem. Res., 1986, 19, 342; M. A. Gallop and W. R. Roper, Adv. Organomet. Chem., 1986, 25, 121; H. P. Kim and R. J. Angelici, Adv. Organomet. Chem., 1987, 27, 51; H. Fischer, P. Hofmann, F. R. Kreissl, R. R. Schrock, U. Schubert and K. Weiss, in Carbyne Complexes, VCH, Weinheim, 1988; A. Mayr and H. Hoffmeister, Adv. Organomet. Chem., 1991, 32, 227; A. Mayr and S. Ahn, Adv. Trans. Met. Coord. Chem., 1996, 1, 1. 3 M.L. Listemann and R. R. Schrock, Organometallics, 1985, 4, 74; S. A. Krouse and R. R. Schrock, J. Organomet. Chem., 1988, 355, 257; Macromolecules, 1989, 22, 2569; M. H. Chisholm, J. C. Huffman and J. A. Klang, Polyhedron, 1990, 9, 1271; T. M. Gilbert and R. D. Rogers, J. Organomet. Chem., 1991, 421, C1; T. P. Pollagi, S. J. Geib and M. D. Hopkins, J. Am. Chem. Soc., 1994, 116, 6051; T. P. Pollagi, J. Manna, S. J. Geib and M. D. Hopkins, Inorg. Chim. Acta, 1996, 243, 177; H.A. Brison, T. A. Pollagi, T. C. Stoner, S. J. Geib and M. D. Hopkins, Chem. Commun., 1997, 1263. 4 E. O. Fischer, V. N. Postnov and F. R. Kreissl, J. Organomet. Chem., 1977, 127, C19; E. O. Fischer, M. Schluge and J. O. Besenhard, Angew. Chem., Int. Ed. Engl., 1976, 15, 2265; E. O. Fischer, M. Schluge, J. O. Besenhard, P. Friedrich, G. Huttner and F. R. Kreissl, Chem. Ber, 1978, 111, 3530; E. O. Fischer, F. J. Gammel, J. O. Besenhard, A. Frank and D. Neugebauer, J. Organomet.Chem., 1980, 191, 261; E. O. Fischer, W. Roll, N. H. T. Huy and K. Ackermann, Chem. Ber., 1982, 115, 2951; E. O. Fischer, F. J. Gammel and D. Neugebauer, Chem. Ber., 1983, 113, 1010; N. A. Ustynyuk, V. N. Vinogradova, V. G. Andrianov and Y. T. Struchkov, J. Organomet. Chem., 1984, 268, 73; J. R. Fernandez and F. G. A. Stone, J. Chem. Soc., Dalton Trans., 1988, 3035; W. Weng, J. A. Ramsden, A. M. Arif and J. A. Gladysz, J. Am. Chem. Soc., 1993, 115, 3824; S. Anderson and A. F.Hill, J. Chem. Soc., Dalton Trans., 1993, 587. 5 N. M. Kostíc and R. F. Fenske, Organometallics, 1982, 1, 489; J. M. Poblet, A. Strich, R. Weist and M. Bénard, Chem. Phys. Lett., 1986, 126, 169. 6 E. O. Fischer and U. J. Schubert, Organomet. Chem., 1975, 100, 59. 7 A. Mayr, G. A. McDermott and A. M. Dorries, Organometallics, 1985, 3, 608; A. Mayr, A. M. Dorries, G. A. McDermott and D. Van Engen, Organometallics, 1986, 5, 1504; A. Mayr, M. F. Asaro, M. A. Kjelsberg, K. S. Lee and D. Van Engen, Organometallics, 1987, 6, 432; G. A. McDermott, A. M. Dorries and A. Mayr, Organometallics, 1987, 6, 925; P. Steil and A. Mayr, Z. Naturforsch., Teil B, 1992, 47, 656. 8 E. O. Fischer and G. Kreis, Chem. Ber., 1976, 109, 1673; H. Fischer and E. O. Fischer, J. Organomet. Chem., 1974, 69, C1; S. Anderson and A. F. Hill, J. Organomet. Chem., 1993, 463, C3; D. S. Williams and R. R. Schrock, Organometallics, 1994, 13, 2101.J. Chem. Soc., Dalton Trans., 1998, Pages 475–481 481 9 J. R. Pratt, W. D. Massey, F. H. Pinkerton and S. F. Thames, J. Org. Chem., 1975, 40, 1090. 10 S. R. Sandler and W. Karo, in Organic Functional Group Transformations, Academic Press, New York, 1968. 11 I. Ugi, U. Fetzer, U. Eholzer, H. Knupfer and K. Offermann, Angew. Chem., 1965, 77, 492; H. Eckert and B. Forster, Angew. Chem., 1987, 99, 922; Angew. Chem., Int. Ed. Engl., 1987, 26, 894. 12 G. R. Clark, N. R. Edmonds, R. A. Pauptit, W. R. Roper, J. M. Waters and A. H. Wright, J. Organomet. Chem., 1983, 244, C57. 13 A. D. Mitchell and L. C. Cross, in Tables of Interatomic Distances and Configuration in Molecules and Ions, The Chemical Society, London, 1958; Tables of Interatomic Distances and Configuration of Molecules and Ions, ed. L. E. Sutton, The Chemical Society, London, 1965. 14 PATTY, P. T. Beurskens, G. Admiraal, G. Beurskens, W. P. Bosman, S. Garcia-Granda, R. O. Gould, J. M. M. Smits and C. Smykalla, The DIRDIF program system, Technical Report of the Crystallography Laboratory, University of Nijmegen, The Netherlands, 1992. 15 TEXSAN, Crystal Structure Analysis Package, Molecular Structure Corporation, Houston, TX, 1985 and 1992. 16 C. K. Johnson, ORTEP, Report ORNL-5138, Oak Ridge National Laboratory, Oak Ridge, TN, 1976. Received 20th October 1997; Paper 7/07519G

ISSN:1477-9226    

DOI:10.1039/a707519g

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1999, 479–486 479 Migration of a phenyl group from co-ordinated CH2(PPh2)2 to an acetylide on an Ru3 cluster: crystal structure of [Ru3(Ï-H)- (Ï3-PPhCH2PPh2)(Ï3-PhC2But)(CO)6] † Michael I. Bruce,*a Paul A. Humphrey,a Brian W. Skelton,b Allan H. White,b Karine Costuas c and Jean-François Halet c a Department of Chemistry, University of Adelaide, Adelaide, South Australia 5005, Australia. E-mail: mbruce@chemistry.adelaide.edu.au b Department of Chemistry, University of Western Australia, Nedlands, Western Australia 6907, Australia c Laboratoire de Chimie du Solide et Inorganique Moléculaire, UMR CNRS 6511, Université de Rennes 1, 35042 Rennes Cedex, France Received 8th September 1998, Accepted 18th November 1998 Thermolysis (refluxing toluene, 60 h) of [Ru3(m-H)(m3-C2But)(m-dppm)(CO)7] resulted in phenyl transfer from co-ordinated dppm to the m3-acetylide to give [Ru3(m-H)(m3-PPhCH2PPh2)(m3-PhC2But)(CO)6] in 41% yield, fully characterised by X-ray determinations of thf and CH2Cl2 monosolvates.The alkyne is co-ordinated to the Ru3 cluster such that the C(1)–C(2) vector forms an angle of 238 with the Ru(1)–Ru(2) vector. This distortion brings two C atoms of the alkyne Ph group close to Ru(2). Density functional and extended Hückel calculations carried out on the new compound indicated that the unusual co-ordination of the alkyne ligand can be attributed to the stereoelectronic asymmetry of the metallic fragment. Introduction The chemistry of [Ru3(m-dppm)(CO)10] 1 (Scheme 1) has been well developed in recent years.1 One of its characteristic reactions is the ready loss of a phenyl group by cleavage of a P–C bond.Thus, hydrogenation of 1 aVorded [Ru3(m-H)(m3-PPhCH2- PPh2)(CO)9] 2, probably via an intermediate cluster hydride, the phenyl group combining with one H atom to give benzene.2 This process was observed directly during the pyrolysis of [Ru4(m-H)4(m-dppm)(CO)10], which aVorded [Ru4(m-H)3(m3- Scheme 1 † Dedicated to Warren Roper on the occasion of his 60th birthday, in recognition of his outstanding contributions to organometallic chemistry.PPhCH2PPh2)(CO)9].3 An alternative source of the hydrogen atom is a phenyl group on the second P atom, which becomes metallated in the complex [Ru3{m3-PPhCH2PPh(C6H4)}(CO)9] 3.2 It is likely that ortho-metallation of this phenyl group occurs prior to elimination of benzene, as found for several arylphosphine cluster complexes.4 An alternative product, formed under CO, is 4, which contains the unusual chelating bridged ditertiary phosphine C6H4(PPhCH2PPh).2 In all of these reactions the phenyl group has been eliminated from the precursor complex, no phenylated derivatives having been isolated.We have recently described the formation of the cluster phenyl complexes [Ru3(m3-PPhCH2PPh2)(m3-C8H8)(Ph)- (CO)5] 5 and [Ru3(m3-PPhCH2PPh2)(m3-C2PPh2)(m-PPh2)(Ph)- (CO)6],6 in both of which the Ph group is trapped on the cluster.In the course of studies of the reactions of 1 with various alkynes,7 we have found an example of migration of a phenyl group to a cluster-bound alkynyl group to give an alkyne which is attached to the cluster in an unusual way. Results We have described elsewhere 5 the ready addition of terminal alkynes to 1 to give the complexes [Ru3(m-H)(m3-C2R)(m-dppm)- (CO)7] 5 (Scheme 2). On heating the tert-butyl complex 5 (R = But) in refluxing toluene for an extended period a new red complex was isolated in 41% yield and identified as [Ru3(m-H)- (m3-PPhCH2PPh2)(m3-PhC2But)(CO)6] 6 by a single-crystal X-ray determination.Scheme 2 Ru(CO)3 Ru(CO)2 PPh2 Ph2P (OC)2Ru But H C C But PPh (OC)2Ru1 Ru2 C1 C2 Ph2P Ru3 H 5 (CO)2 (CO)2 6 C6H5Me 110°C/60 h480 J. Chem. Soc., Dalton Trans., 1999, 479–486 Crystal and molecular structures of [Ru3(Ï-H)(Ï3-PPhCH2PPh2)- (Ï3-PhC2But)(CO)6] 6 As described at length below, crystals of compound 6 (Fig. 1) were obtained monosolvated with thf and CH2Cl2, a study of a less completely solvated form of the latter also being recorded.All forms crystallize in the orthorhombic space group Pbca; unit cell projections are given for the fully solvated thf monosolvate and the fractional [0.366(3)] dichloromethane solvate, projected down a and b in Fig. 2, cell volumes of these two extreme forms diVering by more than 10%. Nevertheless, as the projections down b show, the x and z coordinates of the two forms are very similar, and it is also seen that the molecules may be considered to be disposed as layers about c = 1/8, 3/8, 5/8, 7/8.The relative alignments of successive layers change with respect to each other in y, however, and, in fact, the y coordinates of the defining molecule of the asymmetric unit are displaced between the two types of solvate by ca. a quarter of a cell in that dimension. Consideration has been given as to Fig. 1 Projections of [Ru3(m-H)(m3-PPhCH2PPh2)(m3-PhC2But)(CO)6] 6 (thf solvate), (a) normal to and (b) ‘through’ the Ru3 plane; 20% thermal ellipsoids are shown for the non-hydrogen atoms, hydrogen atoms having arbitrary radii of 0.1 Å.whether this change impacts on the structure of 6; comparative geometries are given for the diVerent forms in Tables 1 and 2, no non-trivial diVerence being observed in the associated bond lengths and angles. In terms of broader conformation, however, Table 2 shows the parameter most aVected to be the pitch of the 22n ring plane vis-à-vis the rest of the molecule.Discussion of the individual molecule of 6 is now conducted in terms of the most precisely determined example in the thf solvate. The complex contains a triangular cluster of three ruthenium atoms [Ru–Ru 2.7311(4)–2.8936(4) Å], each of which bears two CO groups. One face of the triangle is capped by a dephenylated dppm ligand, similar to those found in the other related complexes mentioned above.The Ru(1)–Ru(2) vector is bridged by the phosphido P atom [Ru(1,2)–P(1) 2.2901(8), 2.3254(8) Å] while P(2) is attached to Ru(3) [2.3235(8) Å], while the longer Ru(2)–Ru(3) vector is bridged by the hydride ligand [Ru(2,3)– H(23) 1.71(2), 1.80(2) Å]. The other face is capped by a PhC2But ligand in a m3-h2 mode with further interaction between C(201) and C(206) of the phenyl ring and Ru(2). The arrangement is shown in Figs. 1 and 3. The alkyne deviates somewhat from the symmetrical m3-h2-(^) mode8 [the projected angle between the Ru(1)–Ru(3) and C(1)– C(2) bonds is 238]. This is reflected in the bond lengths Ru(1)– C(1) [2.035(3) Å] and Ru(1)–C(2) [2.426(3) Å] which are rather diVerent from Ru(3)–C(1) [2.433(3) Å] and Ru(3)–C(2) [2.162(3) Å], respectively. Such a twist is accompanied by quite a large C(1)–C(2)–C(201) angle [142.1(3)8] and some C(101)– C(1)–C(2)–C(201) torsion [234.8(6)8] which brings two carbon atoms of the Ph group, C(201) and C(206), close to Ru(2) [2.404(3) and 2.557(3) Å, respectively].A rather strong interaction between Ru(2) and C(2) [2.273(3) Å] is also observed. The C(1)–C(2) bond length is 1.355(4) Å. We note also that there is some degree of localization of p-electron density in the intraring C–C bonds such that two [C(202)–C(203) 1.366(4), C(204)–C(205) 1.348(5) Å] are shorter than the other four [all greater than 1.388(5) Å], suggesting interaction of the ring p-electron density in C(201)–C(206) with Ru(2).The C(2)– C(201) separation is 1.455(4) Å. These tendencies are reproduced in the geometries of the CH2Cl2 solvate, albeit at a lower level of precision. The spectroscopic properties of compound 6 are consistent with the solid-state structure. The solution IR spectrum contains six terminal n(CO) bands while the 1H NMR spectrum contains a complex multiplet at d 217.56 which is assigned to the Ru–H proton.The CH2 protons of the phosphorus ligand resonate between d 4.0 and 4.5. A doublet at d 5.69 can be assigned to a proton adjacent to the Ru-bonded phenyl carbon. The FAB mass spectrum contains M1 centred around m/z 939, which decomposes by loss of H, CO, PPhCH2PPh2 and alkyne fragments. The electron count for this cluster is interesting. A count of 46 “metallic” valence electrons (MVEs) is achieved assuming that the alkyne ligand donates four electrons to the cluster [24 (3 × Ru) 1 1 (H) 1 5 (PPhCH2PPh2) 1 12 (6 × CO) 1 4 (PhC2But) = 46].Such a count precludes any electron donation from the phenyl group attached to C(2) to the metallic array. The Polyhedral Skeletal Electron Pair (PSEP) theory supported by molecular orbital calculations has proven to be very helpful for understanding the structural chemistry of trimetallic alkyne cluster complexes.9 Such compounds are mainly encountered in two distinct geometries depending on their electron counts.Species characterized by 46 MVEs adopt a closo-trigonalbipyramidal structure with the alkyne moiety lying perpendicular to one metal–metal bond [m3-h2-(^) mode],8 such as [Fe3{m3-(^)-C2Ph2}(CO)9] 7.10 Those having 48 MVEs are generally found with a nido-square-pyramidal geometry with the acetylenic ligand positioned parallel to a metal–metal vector [m3-h2-(||) mode] 8 as exemplified by [Co2Fe{m3-(||)-C2Et2}(CO)9] 8.11 A third arrangement which can regarded as a skeletal isomer of the nido form, unexpected according to the PSEPJ. Chem.Soc., Dalton Trans., 1999, 479–486 481 Table 1 Selected molecular geometries (distances in Å, angles in 8) for [Ru3(m-H)(m3-PPhCH2PPh2)(m3-PhC2But)(CO)6] 6. The three values in each entry are for 6?thf, ?CH2Cl2 and ?0.366(3)CH2Cl2 respectively Ru(1)–Ru(2) Ru(1)–Ru(3) Ru(2)–Ru(3) Ru(1)–P(1) Ru(2)–P(1) Ru(3)–P(2) Ru(1)–C(1) Ru(1)–C(2) Ru(2)–C(2) Ru(2)–C(201) Ru(3)–C(1) Ru(3)–C(2) P(1)–C(0) P(2)–C(0) C(1)–C(2) C(1)–C(101) C(2)–C(201) C(201)–C(202) C(201)–C(206) C(202)–C(203) C(203)–C(204) C(204)–C(205) C(205)–C(206) Ru(2) ? ? ? C(206) Ru(2)–H(23) Ru(3)–H(23) 2.8716(4), 2.866(2), 2.8701(6) 2.7311(4), 2.721(1), 2.7253(5) 2.8936(4), 2.889(1), 2.8928(6) 2.2901(8), 2.297(2), 2.293(1) 2.3254(8), 2.317(2), 2.315(2) 2.3235(8), 2.319(2), 2.327(1) 2.035(3), 2.040(7), 2.022(3) 2.426(3), 2.411(7), 2.412(3) 2.273(3), 2.258(6), 2.265(3) 2.404(3), 2.407(7), 2.409(3) 2.433(3), 2.415(7), 2.433(3) 2.162(3), 2.149(7), 2.162(3) 1.833(3), 1.839(7), 1.839(4) 1.836(3), 1.837(7), 1.831(4) 1.355(4), 1.323(9), 1.344(5) 1.526(4), 1.52(1), 1.534(5) 1.455(4), 1.48(1), 1.457(5) 1.415(4), 1.41(1), 1.417(6) 1.403(4), 1.40(1), 1.392(6) 1.366(4), 1.35(1), 1.364(8) 1.388(5), 1.42(1), 1.381(9) 1.348(5), 1.34(2), 1.333(9) 1.402(4), 1.39(1), 1.412(7) 2.557(3), 2.614(8), 2.601(4) 1.71(2), [1.48(est.)], 1.52(2) 1.80(2), [1.80(est.)], 1.87(2) Ru(1)–Ru(2)–Ru(3) Ru(2)–Ru(3)–Ru(1) Ru(3)–Ru(1)–Ru(2) Ru(1)–P(1)–Ru(2) Ru(1)–P(1)–C(0) Ru(2)–P(1)–C(0) Ru(3)–P(2)–C(0) Ru(1)–C(1)–Ru(3) Ru(2)–C(2)–Ru(3) P(1)–C(0)–P(2) Ru(1)–C(1)–C(2) Ru(2)–C(2)–C(1) Ru(2)–C(201)–C(2) C(2)–C(1)–C(101) C(1)–C(2)–C(201) C(202)–C(201)–C(206) C(203)–C(202)–C(201) C(204)–C(203)–C(202) C(205)–C(204)–C(203) C(206)–C(205)–C(204) C(201)–C(206)–C(205) Ru(2)–H(23)–Ru(3) 56.551(8), 56.44(2), 56.45(1) 61.32(1), 61.35(4), 61.36(1) 62.132(9), 62.22(4), 62.20(1) 76.94(3), 76.79(8), 77.05(3) 120.6(1), 121.3(2), 120.8(1) 108.4(1), 107.5(2), 107.9(1) 108.9(1), 107.8(2), 107.5(1) 74.78(8), 74.8(2), 74.8(1) 81.41(9), 81.9(2), 81.5(1) 109.6(2), 108.5(3), 109.2(2) 89.1(2), 89.0(4), 89.2(2) 132.3(2), 133.4(5), 132.5(2) 67.0(1), 66.1(3), 66.5(2) 129.7(2), 131.1(6), 128.9(3) 142.1(3), 141.3(5), 142.1(3) 117.2(3), 117.9(7), 116.4(4) 121.5(3), 121.1(8), 122.3(5) 119.8(3), 120.0(9), 119.5(5) 120.5(3), 119.9(9), 120.5(6) 120.9(3), 121.1(8), 121.1(5) 120.0(3), 119.9(8), 120.1(4) 111(1), [123(est.)], 117(1) Fig. 2 Unit cell projections down (a) a, (b) b and (c) c (the latter of the layers at c = 0.13) of (i) compound 6. 0.366(3)CH2Cl2 and of (ii) 6?thf [(c) only; views corresponding to (a) and (b) are very similar]. rules and called “basket-like”, has been reported for the 48- MVE compound [Os3(m-H)2(m3-HC2NEt2)(CO)9] 9 and related species.12 Theoretical calculations have demonstrated that the unusual co-ordination of the aminoacetylene ligand to the trimetallic array in 9 is mainly due the p-donor eVect of the amino substituent on the alkynyl grouping.13482 J.Chem. Soc., Dalton Trans., 1999, 479–486 With a cluster electron count of 46, compound 6 is unique in featuring an alkyne ligand twisted relative to an M–M bond that is bound with a distorted m3-h2-(^) mode to the cluster framework. Extended Hückel Theory (EHT) and Density Functional Theory (DFT) calculations were carried out on 6 and related models in order to understand why such a distorted geometry for the alkyne ligand is favored in this particular Ru3 environment (see the Experimental section for details).DFT Calculations were first performed on the model [Ru3- (m-H)(m3-PHCH2PH2)(m3-MeC2Ph)(CO)6] 69 based on the crystal structure of 6 in order to reduce computational eVort. A large HOMO 2 LUMO gap (2.13 eV) is computed for the count of 46 MVEs. Comparable results were obtained with EHT calculations: the same electron configuration was obtained, with a HOMO 2 LUMO gap of 1.73 eV.The composition of the MOs in the HOMO 2 LUMO region is similar for both DFT and EHT results. Therefore, because of its structural complexity (size of the molecule, lack of symmetry) the detailed analysis of the bonding in 6 was carried out using EHT calculations. Fig. 3 Detailed geometry of the alkyne–Ru3 cluster interaction in compound 6. Ru1 Ru2 Ru3 C1 C2 C201 C206 But 1.355 1.455 2.273 2.404 2.557 2.894 2.872 2.731 2.035 2.162 1.388 1.366 1.402 1.403 1.415 1.348 H 1.526 2.433 2.426 (OC)3Fe Fe(CO)3 C C (OC)3Co Ph Ph Fe(CO)3 C C Et Et C C H Et2N Os(CO)3 (OC)3Os 7 H H 9 8 Fe(CO)3 Co(CO)3 Os(CO)3 Table 2 Ru3, phenyl C6 interplanar dihedral angles (8) (1 thf; 1, 0.366 CH2Cl2 solvates, respectively) Plane Ru3 11n 21n 22n 11n 48.27(9) 51.7(2) 51.5(1) 21n 27.0(1) 31.6(2) 32.0(1) 69.6(1) 70.8(3) 70.1(2) 22n 64.62(9) 47.2(2) 46.3(1) 75.8(1) 87.2(3) 86.8(2) 63.9(1) 63.7(3) 63.6(2) 20n 45.08(8) 43.7(2) 44.1(1) 82.1(1) 84.4(3) 85.3(2) 53.0(1) 60.7(3) 61.6(2) 22.2(1) 3.5(3) 2.2(2) Deviations of Ru(2) from the C(20n) plane are 2.083(3), 2.107(8), 2.112(4) Å.Torsion angles: C(2,1,101,102) 2149.9(3), 2158.0(8), 2156.3(4); Ru(1),P(1),C(111,112) 27.2(3), 24.4(7), 24.5(4)8 The first step was to determine if our results mirror the experimental data. We started by simply rotating the MeC2Ph ligand relative to a frozen Ru3(m-H)(m3-PHCH2PH2)(CO)6 fragment. Indeed, starting from the closo arrangement 10, a slight rotation of ca. 208, bringing C(2) close to Ru(3) and accompanied by the bending of the phenyl group, is suYcient to generate the distorted closo (DC) geometry observed in compound 6.Alternatively, a rotation of the alkyne ligand on the same core bringing C(1) and C(2) close to Ru(3) and Ru(1), respectively, can be envisaged, leading to another DC geometry (11), which is not observed experimentally (see Scheme 3). Calculations were carried out on the closo model 10 with the C(1)–C(2) vector strictly perpendicular to the Ru(1)–Ru(3) bond and compared to those obtained for 69 and 11.The alkyne ligand was placed on the top of the Ru3 triangle similarly to that encountered in the related closo cluster [Ru3(m3-PhC2Ph)- (m-dppm)(CO)7] 12.14 According to EHT calculations the closo structure 10 is less stable than the DC structure 69 by 1.24 eV. The computed bonding energies between the MeC2Ph ligand and the [Ru3(m-H)(m3-PHCH2PH2)(CO)6] fragment are ca. 3.0 and 4.0 eV for 10 and 69, respectively. Although optimization of the closo geometry would somewhat lower its energy, we think nevertheless that the DC arrangement 69 will have a still lower energy.The EHMO diagrams for the closo 10 and DC 69 arrangements are compared in Fig. 4. They were built up from the interaction of the frontier molecular orbitals (FMOs) of the metallic Ru3(m-H)(m3-PHCH2PH2)(CO)6 fragment with the frontier orbitals of the MeC2Ph ligand. The metallic fragment exhibits a set of two-below-three FMOs characteristic of triangular M3L9 units with an edge bridged by a hydride ligand.15 The FMO (2a) extends mainly in the Ru3 plane, whereas the four others (1a, 3a, 4a, and 5a) extend predominantly above the metal plane (see the middle of Fig. 4). The linear C2H2 ligand has two p and two p* orbitals. Upon co-ordination to a metal framework the H atoms bend back and these two sets lose their degeneracy. Mixing with s orbitals, one p component (s/p) is destabilized, while one p* component (s/p*) is stabilized.Replacement of the H substituents by Me and Ph groups modifies the shape and the energy of these FMOs. The C–C bonding p and s/p FMOs are predominantly localized on C(1) attached to the Me group, and the C–C antibonding p* and s/p* FMOs are slightly more localized on C(2) to which is tethered the Ph group. In the closo structure the main bonding interactions between the two fragments occur between the metallic 3a and 4a FMOs and the alkyne s/p and p FMOs, respectively, and between the metallic 2a FMO and the s/p* and p* FMOs of the MeC2Ph ligand. The metallic 1a FMO hardly interacts with the alkyne ligand and remains almost unperturbed after interaction.The Scheme 3 Me Me Me 1 3 1 2 2 2 1 1 3 2 2 1 1 3 2 6' 10 11 Ph2P P Ph2 (OC)2Ru C C Ru(CO)3 Ph Ph 12 Ru(CO)2J. Chem. Soc., Dalton Trans., 1999, 479–486 483 Fig. 4 EHMO Interaction diagram of the model [Ru3(m-H)(m3-PHCH2PH2)(m3-MeC2Ph)(CO)6] in the closo arrangement 10 (on the left) and the DC arrangement 69 (on the right).FMO occupations after interaction are given in parentheses. resulting MO diagram of 10 is shown on the left-hand side of Fig. 4. The MO diagram of the DC structure 69 shown on the right-hand side resembles quite strongly that of the closo structure 10 except in the LUMO region. Indeed, the 4a LUMO is pushed up in energy in the DC arrangement and the HOMO 2 LUMO gap rises to 1.70 eV for the count of 46 MVEs (it is 1.22 eV in the closo structure).Such a large HOMO 2 LUMO gap ensures the stability of this DC arrangement. Examination of the LUMO reveals a quite strong antibonding character between Ru(2) and carbons of the Ph group tethered to C(2), particularly C(206). Upon bending of the Ph group towards Ru(2), interaction between the metallic fragment and the alkyne ligand increases, in particular between the metallic 2a FMO and the alkyne s/p* FMO, due to a better overlap, leading to some stabilization of occupied MOs and destabilization of vacant MOs in particular the LUMO.A second-order Jahn–Teller distortion is suspected in the closo structure 10, leading to the more stable DC structure 69. The destabilization of the LUMO is accompanied by the stabilization of occupied orbitals (because of the lack of symmetry, we were not able to identify which particular orbitals are stabilized). We tried to quantify the gain of energy due to the bending of the phenyl ring towards Ru(2) in structure 69.A DC arrangement with the Ph group pointing away from Ru(2) is less stable by 1.05 eV and the HOMO 2 LUMO gap decreases from 1.70 to 1.22 eV, comparable to that computed for the closo arrangement 10. The bonding energy between the two fragments drops from ca. 4.0 to 3.2 eV. Clearly the bending of the phenyl group of the alkyne ligand towards Ru(2) with C(1) and C(2) close to Ru(1) and Ru(3), respectively, is essential in stabilizing structure 69.Nevertheless, a close look at the atomic net charges indicates almost no change in the electron transfer between the alkyne ligand and the metal framework upon bending of the phenyl ring. The metallic fragment is very slightly negatively charged before and after bending (20.21 vs. 20.18), indicating that in both cases donation from the alkyne ligand to the cluster is slightly more important than back donation from the cluster to the C2 ligand. The EH atomic net charges of the metal atoms are 20.49, 20.01 and 20.32 for Ru(1), Ru(2) and Ru(3), respectively.The corresponding charges before approach of C(201) and C(206) were 20.47, 20.17 and 20.28, respectively. Thus, the less electron-rich Ru(2) centre loses some electron density upon bending of the phenyl ring. The charge on atom C(1) of the alkyne, nearly neutral before (10.02), becomes slightly negative (20.06) after bending of the phenyl ring, whereas the charge on C(2) remains almost unchanged (20.04 vs. 20.05). The total charge on the phenyl ring becomes slightly more positive after bending (10.09 vs. 10.06 before). We conclude that upon bending of the phenyl ring there is a very weak additional electron transfer from the phenyl ring towards the metal fragment, which in turn transfers some electron density to atom C(1) of the alkyne ligand. Clearly, the “side on” Ru(2)–C(201, 206) interaction occurring in compound 6 should be regarded as a covalent interaction between metallic orbitals and alkyne FMOs containing some admixture of phenyl carbon atoms.The delocalization of the latter FMOs on the phenyl ring allows the Ru(2) center to establish bonding interactions with C(201) and C(206). The EH overlap populations (OPs) computed between Ru(2) and C(201) and C(206) (0.05 and 0.06, respectively) suggest a nonnegligible metal–carbon bonding interaction. For comparison the OP corresponding to the Ru(2)–C(2) bond is 0.22.This is supported by the DFT electron density distribution in the Ru(2)–C(201)–C(206) plane illustrated in Fig. 5. Some electron density is present between Ru(2) and the two close carbon atoms of the phenyl ring. It is interesting that the alternative DC structure 11 with C(1) and C(2) close to Ru(3) and Ru(1), respectively, and the Ph group bent towards Ru(2) (see Scheme 3) is computed to be 2.01 eV less stable than structure 69, with a bonding energy between the alkyne ligand and the metallic fragment of only 2.21 eV.The diVerent ligand environment around the Ru atoms in the Ru3(m-H)(m3-PHCH2PH2)(CO)6 fragment (see above) renders them electronically non-equivalent. This indicates that the slight twist of the alkyne ligand accompanied by the bending of484 J. Chem. Soc., Dalton Trans., 1999, 479–486 the Ph group in 6 is mainly dictated by the stereoelectronic asymmetry of the metallic fragment. Note that a regular closo arrangement is observed in [Ru3(m3-PhC2Ph)(m-dppm)(CO)7] 12 even though it contains a related C2Ph2 ligand bound to a symmetrical metal fragment.14 In summary, the asymmetry of the metallic fragment in compound 6 obliges the alkyne to rotate with respect to the Ru(1)– Ru(2) vector, somewhat weakening the M–C bonding.This is reflected in the C–C overlap population which is slightly stronger in 69 than in 10, indicating that the C2 unit is less strongly bound to the metal triangle in the former.This M–C bonding weakening is compensated by the approach of two carbon atoms of the Ph group towards Ru(2), somewhat resembling an “agostic” interaction.16 Comparable “side-on” M–C interactions involving a double bond of the phenyl ring have been previously encountered in electron-deficient metal complexes such as [Fe2(m-C3Ph3H)(CO)6],17 [Fe2{m-C(OEt)CPhCHPh}- (CO)6] 18 or [Fe2(m-CHCHCMePh)(CO)6].19 Analogously to these complexes, this interaction in 6 induces a partial localization of the C–C bonds of the phenyl ring, which become alternatively short and long (see Table 1).Such a “side-on” M–C interaction is diVerent from those involving co-ordination between a metal center and a C–C (Ph) bond, as found in [Ru3{m3-CHCPhC(O)CPhCPh}(m-dppm)(CO)6] 14 or a P–C (Ph) bond, as found in [Ru3(m-H)(m-PPh2)(CO)9].20 Discussion The major point of interest in the chemistry reported here is the unusual attachment of the alkyne ligand in complex 6. Perhaps the closest analogy is to be found in the recently described [Ru3{m3-CHCPhC(O)CPhCPh}(m-dppm)(CO)6], obtained from the reaction of phenylethyne with the 46e complex [Ru3{m3-(^)- C2Ph2}(m-dppm)(CO)7].14 The former complex is also coordinatively unsaturated, the side-on bonding of the C–C (Ph) bond apparently compensating for the loss of a CO group during its formation.The long-known phosphide derivative [Ru3(m-H)(m-PPh2)(CO)7], with a similar P–C (Ph) interaction with a cluster Ru atom is another example of these ‘agostic’ phenyl groups.20 Complex 6 has been formed by migration of a P-bonded phenyl group to the s-bonded carbon of the m3-C2But ligand in the precursor 5 (R = But).This may occur by prior migration of the Ph group to the cluster with displacement of CO, followed by further migration to the s-bonded carbon of the m3-acetylide ligand. The actual molecular structure of 6 can be envisaged as an intermediate stage in the transfer of the Ph group from the dppm ligand to the acetylide via the cluster; alternatively, it Fig. 5 Theoretical electron density contour map in the Ru(2)–C(201)– C(206) plane of structure 69. Atoms P(1) and C(22) are 0.81 and 0.58 Å above the plane, respectively. Contour values are 0.03, 0.05, 0.07, 0.09, and 1.00 e bohr23 (bohr ª 5.29 × 10211 m). could be considered to illustrate the activation of the stable alkyne towards C–C bond cleavage. In larger clusters, however, the migration of Ph from diphenylphosphinoacetylide ligands has also been observed, the net result being the elimination of PPh groups to the cluster.21 The unusual co-ordination of the alkyne fragment is found to be a rather distorted m3(^) mode.Conclusion Pyrolysis of compound 2 has resulted in phenyl transfer from co-ordinated dppm to the m3-acetylide ligand to give the alkyne PhC]] ] CBut which, in 6, is co-ordinated with an intermediate geometry, with the Ph group involved in a side-on ‘agostic’ bonding.The growing number of examples of this type of bonding suggests that it represents an intermediate or transition state in the overall phenyl transfer or rearrangement reactions. Experimental Instrumentation IR: Perkin-Elmer 1700X FT IR. NMR: Bruker CXP300 or ACP300 (1H at 300.13 MHz, 13C at 75.47 MHz). FAB MS: VG ZAB 2HF (using 3-nitrobenzyl alcohol as matrix, exciting gas Ar, FAB gun voltage 7.5 kV, current 1 mA, accelerating potential 7 kV). General reaction conditions Reactions were carried out under an atmosphere of nitrogen, but no special precautions were taken to exclude oxygen during work-up.Starting materials Complexes 122 and 5 (R = But) 7 were prepared as previously described. Preparation of [Ru3(Ï-H)2(Ï3-PPhCH2PPh2)(Ï3-PhC2But)- (CO)6] 6 A solution of [Ru3(m-H)(m3-C2But)(m-dppm)(CO)7] (200 mg, 0.207 mmol) was heated in refluxing toluene (50 ml) for 60 h. Evaporation and separation of the products by preparative TLC (acetone–light petroleum, b.p.range 60–80 8C, 3 : 7) gave the major product as an orange band (Rf 0.69). Crystallisation (CH2Cl2–MeOH) gave red crystals of [Ru3(m-H)(m3-PPhCH2- PPh2)(m3-PhC2But)(CO)6] 6 (80 mg, 41%), mp >150 8C (decomp.). [Found: C, 47.08; H, 3.54%; M (mass spectrometry) 939; C37H33O6P2Ru3 requires C, 47.39; H, 3.44%; M 939]. IR: n(CO) (cyclohexane) 2034m, 2024vw, 2008vs, 1992m, 1977vs, 1954m (br), 1942 (sh) and 1921m cm21. 1H NMR: d(CDCl3) 217.56 [2 × m (7 peaks), 1 H, Ru–H], 4.06–4.53 (m, 2 H, CH2), 5.69 (d, JPH = 7 Hz, 1 H) and 7.16–7.77 (m, 19H, Ph).FAB MS (m/z, relative intensity): 939, M1, 16; 910, [M 2 H 2 CO]1, 10; 882, [M 2 H 2 2CO]1, 10; 854, [M 2 H 2 3CO]1, 50; 826, [M 2 H 2 4CO]1, 75; 798, [M 2 H 2 5CO]1, 42; 781, [M 2 PhC2Bu]1, 20; 770, [M 2 H 2 6CO]1, 62; 750, [M 2 4CO 2 Ph]1, 25; 691, [Unknown], 40; 634, [M 2 PPhCH2PPh2]1, 35; 611, [M 2 2H 2 6CO 2 PhC2Bu]1, 65; 531, [M 2 C2Ph 2 PPhCH2PPh2]1, 100. Structure determination of compound 6 The execution of this exercise has occurred over a period of ten years.Initial structure determination studies were undertaken using unique room-temperature single counter/four-circle diffractometer data sets (2q–q scan mode; monochromatic Mo-Ka radiation, l = 0.71073 Å, T ª 295 K) yielding N reflections, No with I > 3s(I) being considered ‘observed’ and used in the full matrix least squares refinement after gaussian absorption correction. Anisotropic thermal parameters were refined for theJ.Chem. Soc., Dalton Trans., 1999, 479–486 485 non-hydrogen atoms, (x, y, z, Uiso)H for the ligand and solvent hydrogen atom parameters being constrained at estimated values. Conventional residuals R, R9 at convergence are quoted, statistical weights being derivative of s2(I) = s2(Idiff) 1 0.0004 s4(Idiff). The initial study was undertaken on an inferior specimen of the thf monosolvate of compound 6 and did not yield a definitive location for the core hydrogen atom component. In an eVort to resolve the latter, material recrystallized from dichloromethane was studied; the structure is disposed quasiisomorphously to the thf solvate in a unit cell of similar symmetry and dimensions, but with a general displacement of y coordinates by ca. 1/4 and the cell volume diminished by ca. 10%. DiVerence map residues were modelled in terms of one molecule of solvation (CH2Cl2), with the site occupancy set at unity after trial refinement.This determination, although of more useful precision, again did not yield a definitive core hydrogen atom description. A more suitable specimen of the thf solvate obtained from fresh material yielded extensive data measured to 2qmax = 708, the site occupancy of the solvent refining to unity and a core hydrogen atom being located. At this stage the novelty of the compound was apparent, as was the desirability of some attempt at theoretical modelling of the associated bonding. After this had been carried out, a new generation of equipment was available in the form of a Bruker AXS CCD detector instrument; the latter thf solvate sample was still available, seemingly having preserved its integrity, and it was decided to reexamine the material in this manner to improve, if possible, parameters for incorporation in/comparison with the theoretical model.This was done, the site occupancy of the solvent thf being set at unity after trial refinement and all hydrogen atoms other than those of the solvent (which were set constrained with estimated values) being refined in (x, y, z, Uiso), with the core hydrogen atom complement as shown.The proximity of the phenyl ring defined by C(20n) to Ru(2) was of interest and, given the somewhat diVerent nature of the structure of the dichloromethane solvate, it was decided to revisit that also in case the rather diVerence lattice packing impacted at all significantly on molecular conformation.It was found that that material also had substantially retained its integrity, yielding useful data; however, refinement of the solvate site occupancy showed that over the years that had diminished to 0.366(3). Again, all hydrogen atoms other than the solvent were refinable in (x, y, z, Uiso), with a similar core hydrogen component. Accordingly, we present, hereunder, details for the following determinations (a)–(c): (a) the CCD instrument study of the thf solvate; (b) the single counter instrument study of the CH2Cl2 solvate, site occupancy 1; (c) the CCD instrument study of the CH2Cl2 solvate, site occupancy 0.366(3).For the CCD data, full spheres were measured by w scans (0.38, 15 s frames) to 2qmax 588 in a T = 299 K ambience [Ntot data merging to N (unique), Rint as cited], data being processed with the proprietary software SAINT and SADABS the latter encompassing an ‘empirical absorption correction’. Structure solutions and refinements were carried out using the XTAL 3.4 program system.23 Crystal/refinement data for [Ru3(Ï-H)(Ï3-PPhCH2PPh2)- (Ï3-PhC2But)(CO)6] º C37H32O6P2Ru3.Orthorhombic, space group Pbca (D15 2h, No. 61), Z = 8. (a) 6?thf º C41H40O7P2Ru3, M = 1009.9, a = 14.1102(7), b = 19.006(1), c = 32.422(2) Å, V = 8695 Å3, Dc = 1.543 g cm23; F(000) = 4032, mMo = 11.5 cm21, specimen 0.39 × 0.35 × 0.23 mm, ‘Tmin,max’ = 0.73, 0.90, Ntot = 94246, N = 11204 (Rint = 0.032), No = 9410, R = 0.039, R9 = 0.025 (statistical weights), Drmin,max = 20.55, 0.74 e Å23.(The volume of the unit cell in the more precise of the single counter instrument studies was 8667 Å3; the change may be consequent upon the slightly diVerent temperature, or crystal ‘aging’.) (b) 6?CH2Cl2 º C38H34Cl2O6P2Ru3, M = 1022.7, a = 14.218(8), b = 18.512(15), c = 30.105(13) Å, V = 7924 Å3, Dc = 1.714 g cm23, F(000) = 4048, mMo = 12.5 cm21, specimen 0.12 × 0.12 × 0.40 mm, Tmin,max = 0.85, 0.87, 2qmax = 558, N = 9093, No = 5089, R = 0.043, R9 = 0.041 (statistical weights), Drmin,max = 0.86, 1.21 e Å23.(c) 6?0.366(3)CH2Cl2 º C37H32O6P2Ru3. ª0.366CH2Cl2, M = 969.2, a = 14.146(3), b = 18.476(3), c = 29.908(6) Å, V = 7816 Å3, Dc = 1.647 g cm23, F(000) ª 3836.3, mMo = 13.2 cm21, specimen 0.35 × 0.13 × 0.12 mm, ‘Tmin,max’ = 0.69, 0.85, Ntot = 86234, N = 10114 (Rint = 0.026), No = 7030, R = 0.039, R9 = 0.022, Drmin,max = 20.57, 0.98 e Å23. (The diminished cell volume cf. 6?CH2Cl2 is consistent with the diminution of solvent site occupancy.) CCDC reference number 186/1257.See http://www.rsc.org/suppdata/dt/1999/479/ for crystallographic files in .cif format. Theoretical calculations Density functional calculations were carried out on model 69 using the Amsterdam Density Functional (ADF) program24 developed by Baerends and co-workers 25 using non-local exchange and correlation corrections.26 The atom electronic configurations were described by a double-z Slater-type orbital (STO) basis set for H 1s, C 2s and 2p, O 2s and 2p, P 3s and 3p.A triple-z STO basis set was used for Ru 4d and 5s, augmented with a single-z 5p polarization function. A frozen-core approximation was used to treat the core electrons of C, O, P, and Ru. Extended Hückel calculations were carried out within the extended Hückel formalism27 using the program CACAO.28 The exponents (z) and the valence shell ionization potentials (Hii in eV) were respectively: 1.3, 213.6 for H 1s; 1.625, 221.4 for C 2s; 1.625, 211.4 for C 2p; 2.275, 232.4 for O 2s; 2.275, 214.8 for O 2p; 1.6, 218.6 for P 3s; 1.6, 214.0 for P 3p; 2.078, 28.6 for Ru 5s; 2.043, 25.1 for Ru 5p.The Hii value for Ru 4d was at 212.2. A linear combination of two Slater-type orbitals with exponents z1 = 5.378 and z2 = 2.303 with the weighting coeYcients c1 = 0.5340 and c2 = 0.6365 was used to represent the Ru 4d atomic orbitals. The diVerent molecular models used were based on the experimental structure 6.Acknowledgements We thank the Australian Research Council for financial support and Johnson Matthey plc for a generous loan of RuCl3?nH2O. P. A. H. acknowledges receipt of a University of Adelaide Post-graduate Research Scholarship. J.-F. H. thanks the University of Adelaide and the Centre National de la Recherche Scientifique for his stay at Adelaide in August 1997. K. C. and J.-F. H. thank the Centre de Ressources Informatiques (CRI) of Rennes and the Institut de Développement et de Ressources en Informatique Scientifique (IDRIS-CNRS) of Orsay (project 970649) for computing facilities.Professor J.-Y. Saillard (Rennes) is thanked for helpful discussions. References 1 A. J. Deeming, in Comprehensive Organometallic Chemistry, eds. E. W. Abel, F. G. A. Stone and G. Wilkinson, Elsevier, Oxford, 2nd edn., 1995, vol. 7, ch. 12, p. 683. 2 N. Lugan, J.-J. Bonnet and J. A. Ibers, J. Am. Chem. Soc., 1985, 107, 4484. 3 M. I.Bruce, E. Horn, O. bin Shawkataly, M. R. Snow, E. R. T. Tiekink and M. L. Williams, J. Organomet. Chem., 1986, 316, 187. 4 See, for example: M. I. Bruce, G. Shaw and F. G. A. Stone, J. Chem. Soc., Dalton Trans., 1982, 2094; A. J. Deeming, S. E. Kabir, N. I. Powell, P. A. Bates and M. B. Hursthouse, J. Chem. Soc., Dalton Trans., 1987, 1529. 5 M. I. Bruce, P. A. Humphrey, B. W. Skelton and A. H. White, J. Organomet. Chem., 1996, 526, 85.486 J. Chem. Soc., Dalton Trans., 1999, 479–486 6 M.I. Bruce, P. A. Humphrey, B. W. Skelton and A. H. White, J. Organomet. Chem., 1997, 539, 141. 7 M. I. Bruce, P. A. Humphrey, E. Horn, B. W. Skelton, E. R. T. Tiekink and A. H. White, J. Organomet. Chem., 1992, 429, 207. 8 M. G. Thomas, E. L. Muetterties, R. O. Day and V. W. Day, J. Am. Chem. Soc., 1976, 98, 4645. 9 See, for example: M. J. McGlinchey, in Topics in Physical Organometallic Chemistry, ed. M. Gielen, Freund Publishing House, London, 1992, vol. 4, p. 41; J.-F. Halet, Coord. Chem. Rev., 1995, 143, 637 and refs. therein. 10 J. F. Blount, L. F. Dahl, C. Hoogzand and W. Hübel, J. Am. Chem. Soc., 1966, 88, 292; A. J. Carty, N. J. Taylor and E. Sappa, Organometallics, 1988, 7, 405; D. Osella, L. Pospisil and J. Fiedler, Organometallics, 1993, 12, 3140. 11 S. Aime, L. Milone, D. Osella, A. Tiripicchio and A. M. Manotti Lanfredi, Inorg. Chem., 1982, 21, 501. 12 R. D. Adams and J. T. Tanner, Organometallics, 1988, 7, 2241; A. J. Deeming, S. E. Kabir, D. Nuel and N. I. Powell, Organometallics, 1989, 8, 717. 13 Z. Nomikou, J.-F. Halet, R. HoVmann, J. T. Tanner and R. D. Adams, Organometallics, 1990, 9, 588. 14 S. Rivomanana, C. Mongin and G. Lavigne, Organometallics, 1996, 15, 1195. 15 B. E. R. Schilling and R. HoVmann, J. Am. Chem. Soc., 1979, 101, 3456; J.-F. Halet, J.-Y. Saillard, R. Lissillour, M. J. McGlinchey and G. Jaouen, Inorg. Chem., 1985, 24, 218. 16 M. Brookart and M. L. H. Green, J. Organomet. Chem., 1983, 250, 395. 17 G. Gervasio and E. Sappa, Organometallics, 1993, 12, 1458. 18 I. Ross, R. Mathieu, X. Solans and M. Font-Altaba, J. Organomet. Chem., 1984, 260, 240. 19 G. Dettlaf, U. Behrens and E. Weiss, Chem. Ber., 1979, 110, 3019. 20 S. A. MacLaughlin, A. J. Carty and N. J. Taylor, Can. J. Chem., 1982, 60, 87. 21 M. I. Bruce, B. W. Skelton and A. H. White, J. Organomet. Chem., 1991, 420, 87. 22 M. I. Bruce, B. K. Nicholson and O. bin Shawkataly, Inorg. Synth., 1989, 26, 325. 23 S. R. Hall, G. S. D. King and J. M. Stewart (Editors), XTAL Reference Manual, Version 3.4, University of Western Australia, Lamb Press: Perth (1995). 24 Amsterdam Density Functional (ADF) Program, release 2.0.1, Vrije Universteit, Amsterdam, 1996. 25 E. J. Baerends, D. E. Ellis and P. Ros, Chem. Phys., 1973, 2, 41; E. J. Baerends and P. Ros, Int. J. Quantum. Chem., 1978, S12, 169; P. M. Boerrigter, G. te Velde and E. J. Baerends, Int. J. Quantum Chem., 1988, 33, 87; G. te Velde and E. J. Baerends, J. Comput. Phys., 1992, 99, 84. 26 A. D. Becke, Phys. Rev A., 1988, 38, 3098; J. P. Perdew, Phys. Rev B., 1986, 33, 8822; B34, 7406 (erratum). 27 R. HoVmann, J. Chem. Phys., 1963, 39, 1397. 28 C. Mealli and D. Proserpio, J. Chem. Educ., 1990, 67, 399. Paper 8/07008C

ISSN:1477-9226    

DOI:10.1039/a807008c

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1998, Pages 483–488 483 Syntheses and electrochemical characterization of heteroleptic cyclopentadienyl–dithiolene d2 tungsten complexes. Structures and magnetic properties of charge-transfer salts † Isabelle V. Jourdain,a Marc Fourmigué,*,b Fabrice Guyon *,a and Jacques Amaudrut a a Laboratoire de Chimie et Electrochimie Moléculaire, UFR des Sciences et Techniques, 25030 Besançon Cedex, France b Institut des Matériaux de Nantes (IMN), UMR 6502 CNRS-Université de Nantes, 2, rue de la Houssinière, BP 32229, 44322 Nantes cedex 03, France Novel diamagnetic tungsten(IV) complexes of general formula W(h-C5H4R)2(dithiolene) [R = H, SiMe3 or But; dithiolene = C3S5 22 (4,5-disulfanyl-1,3-dithiole-2-thionate), C3OS4 22 (4,5-disulfanyl-1,3-dithiole-2-onate) or dddt22 (5,6-dihydro-1,4-dithiine-2,3-dithiolate)] have been synthesized and their redox properties investigated by cyclic voltammetry. Two reversible oxidation waves are observed for each complex.The redox half-wave potentials allow [W(cp)2(dddt)] (cp = h-C5H5) to react with the organic acceptor tcnq (tetracyanoquinodimethane) while the C3S5 22 and C3OS4 22 compounds reduce tcnqf4 (7,7,8,8-tetracyano-1,2,4,5-tetrafluoroquinodimethane). X-Ray crystallographic studies were carried out on [W(cp)2(dddt)]~1[tcnq]~2 and [W(h-C5H4But)2(C3S5)]~1[tcnqf4]~2. The structural analyses and extended-Hückel calculations showed that the radical anions are strongly dimerised into diamagnetic moieties.The cations form centrosymmetrical dimers and exhibit antiferromagnetic interactions, as deduced from the temperature dependence of the magnetic susceptibility, with T(cmax) = 18 and 22 K for [W(cp)2(dddt)]~1[tcnq]~2 and [W(h-C5H4But)2(C3S5)]~1[tcnqf4]~2 respectively. Square-planar M(dithiolene)2 complexes have received much attention as precursors for molecular metals and superconductors over the last 15 years.1 The importance in this chemistry of sulfur-rich dithiolene ligands such as C3S5 22 (4,5- disulfanyl-1,3-dithiole-2-thionate) or dddt22 (5,6-dihydro-1,4- dithiine-2,3-dithiolate) 2 is largely due to their ability to overlap with each other and to form extended networks with strong electronic interactions, a prerequisite for collective electronic properties such as conductivity, ferro- or antiferro-magnetism.We have been interested in exploring the co-ordination of these ligands on metallocenes and especially in developing the synthesis of paramagnetic species.Recent reports on d0 complexes M(cp)2(dithiolene) (cp = h-C5H5) with M = Ti 3 or Zr4 show that radical formation is difficult: the loss of one electron leads to very unstable species and the reduction requires a low potential. Group 5 metals appeared to be more interesting since paramagnetism is found in neutral compounds (d1). Recently works were published on vanadium and niobium complexes.4,5 However solution EPR spectra indicated a strong localisation of spin density on the metal centre and SQUID magnetic studies on several Nb(cp)2(dithiolene) complexes showed Curietype behaviour.This leads to the conclusion that paramagnetic species do not interact in the solid state. Another way to obtain d1 complexes consists of oxidising diamagnetic Mo(cp)2(dithiolene) compounds where reversible oxidation waves were evidenced.6 We reported on the presence of strong antiferromagnetic interactions in [Mo(cp)2(dddt)]~1[tcnq]~2 (tcnq = tetracyanoquinodimethane) which demonstrated the ability of the molecules to interact with each other in the solid state.7 Thus d2 complexes M(cp)2(dithiolene) seem to be good precursors of novel materials with interesting electronic properties. In addition, Green et al.8 have shown that changing the metal centre from molybdenum to tungsten in compounds [M(h- C7H7)LI2] (L = MeCN or PMe3) increases the exchange constant in the one-dimensional antiferromagnets.For these reasons we decided to investigate the tungsten chemistry with † Non-SI units employed: emu = SI × 106/4p; eV ª 1.60 × 10219 J. the aim to establish stronger interactions by changing the nature of the metal. The co-ordination of dithiolene ligands on tungsten is not rare 9,10 but little attention has been devoted to the chemistry of complexes having both cp and sulfur-rich ligands.11 In this work we report the preparation and characterisation of some W(h-C5H4R)2(dithiolene) (R = H, SiMe3 or But) complexes.Note that substituted cyclopentadienyls were chosen because we expected to take advantage of their higher solubility to grow single crystals. We describe herein the structures and magnetic properties of [W(cp)2(dddt)]~1[tcnq]~2 and [W(h-C5H4But)2(C3S5)]~1[tcnqf4]~2 (tcnqf4 = 7,7,8,8-tetracyano- 1,2,4,5-tetrafluoroquinodimethane). These results are compared to those obtained with [Mo(cp)2(dddt)]~1[tcnq]~2 and the influence of metal and ligand changes on the structural and magnetic properties of the salts is discussed.Results and Discussion Syntheses and electrochemical behaviour The preparation of [W(h-C5H4R)2(C3S5)] (R = H 1, SiMe3 2 or But 3), [W(cp)2(C3OS4)] 4 (C3OS4 22 = 4,5-disulfanyl-1,3- dithiole-2-onate) and [W(cp)2(dddt)] 5 involves the conventional metathesis reaction of the dithiolene dianion with the corresponding metallocene dichloride (Scheme 1). In all cases pink-purple air-stable crystalline solids are isolated.In solution 5 is very sensitive to oxidation and the C3S5 and C3OS4 compounds slowly decompose in chlorinated solvents such as CH2Cl2 and CHCl3. Indeed the UV absorption spectra measured in CHCl3 display a decrease of the absorption at l = 500 nm for the C3S5 compounds while near-infrared transitions appear. The diamagnetic complexes were identified by elemental analysis and 1H NMR spectroscopy (Table 1). As expected, the more electron-rich character of the metal results in a slight shielding of all signals compared to those of analogous molybdenum complexes.Furthermore the characteristic absorption bands for the dithiolene are present in the infrared spectra: that arising from the C]] C bond at low frequency484 J. Chem. Soc., Dalton Trans., 1998, Pages 483–488 Table 1 Proton NMR data Complex [M(cp)2(C3S5)] [M(h-C5H4But)2(C3S5)] [M(h-C5H4SiMe3)2(C3S5)] [M(cp)2(C3OS4)] [M(cp)2(dddt)] M = Mo7,12 5.40 (s, C5H5) 5.44 (m, 4 H, C5H4), 5.19 (m, 4 H, C5H4), 1.22 (s, 18 H, But) 5.58 (m, 4 H, C5H4), 5.07 (m, 4 H, C5H4), 0.25 (s, 18 H, SiMe3) 5.37 (s, C5H5) 5.37 (s, C5H5), 3.13 (s, 4 H, dddt) M = W 5.36 (s, C5H5) 5.46 (m, 4 H, C5H4), 5.00 (m, 4 H, C5H4), 1.26 (s, 18 H, But) 5.61 (m, 4 H, C5H4), 4.86 (m, 4 H, C5H4), 0.29 (s, 18 H, SiMe3) 5.35 (s, C5H5) 5.26 (s, C5H5), 3.03 (s, 4 H, dddt) (ª1430–1450 cm21) and an absorption at 1050 cm21 due to the C]] S bond of the C3S5 22 ligand.Cyclic voltammetry experiments performed in MeCN at scan rate 100 mV s21 (Fig. 1) evidence the same behaviour as with molybdenum dithiolene compounds which consists of two reversible oxidation waves. The obtention of d1 and d0 species demonstrates the electron-rich character of dithiolene ligands since the second wave of thiolate complexes [M(cp)2(SR)2] (M = Mo or W, R = Me or Ph) was described by Kotz et al.13 as an irreversible oxidation. No significant differences were observed for the redox potentials of the first oxidation between molybdenum and tungsten compounds (Table 2).This is in contrast with the results reported by Kotz for thiolate and dihalide complexes for which the potentials for tungsten were shifted toward cathodic potentials by respectively ª 0.1 and 0.2 V. Thus, dithiolene ligands appear to be involved in the first oxidation and they level out the influence of the electron-density difference between Mo and W. Comparison of the oxidation potentials with the reduction potential of the couple tcnq–tcnq~2 (E2� 1 = 20.31 V vs.Ag–0.1 M AgClO4) shows that only [W(cp)2(dddt)] can reduce this electron acceptor. Despite the substitution of tungsten for molybdenum and the presence of alkyl-substituted cyclopentadienyl Scheme 1 (i) NaOMe (2 equivalents) in MeOH; (ii) Et2O; (iii) [W(h- C5H4R)2Cl2], tetrahyofuran (thf), reflux S S PhCOS PhCOS S S S –S –S S S S S S S (RC5H4)2W ( i) ( ii) (iii) 1 R = H 2 R = SiMe3 3 R = But S S S S O S S –S –S O S S S S O (C5H5)2W ( i) ( ii) (iii) O 4 S S –S –S S S (C5H5)2W ( i) ( ii) (iii) O 5 S S S S S S Fig. 1 Cyclic voltammogram of [W(h-C5H4But)2(C3S5)]. Scan rate 100 mV s21 rings, neither compounds derived from C3S5 22 nor C3OS4 22 undergo a reaction with tcnq. However, oxidation of these novel electron donors can be investigated with the stronger oxidiser tcnqf4 (E2� 1 = 10.07 V vs. Ag–0.1 M AgClO4). Another interesting organic acceptor is tcne (tetracyanoethylene) which can be a smaller anionic counterpart and is presumed to favour interactions in the solid state compared to the cyclic tcnq and tcnqf4.Tetracyanoethylene (E2� 1 = 20.14 V vs. Ag–0.1 M AgClO4) oxidises only [W(cp)2(dddt)] to form a salt of 1 : 1 stoichiometry, as determined by elemental analysis. Reaction with tcnq occurs in the same way as for [Mo(cp)2(dddt)]:7 at first a green solid, of 2 : 3 stoichiometry, precipitates and after cooling we obtained black crystals of 1 : 1 stoichiometry (6) the structure of which was solved.With the C3S5 and C3OS4 compounds only tcnqf4 salts of 1 : 1 stoichiometry appear as product {[W- (h-C5H4R)2(C3S5)]~1[tcnqf4]~2 (R = But 7 or H 8) and [W(cp)2- (C3OS4)]~1[tcnqf4]~2 9}. As expected, the use of substituted cyclopentadienyl favours the development of single crystals and allows the crystal structure determination of 7. Solid-state structure The complex [W(cp)2(dddt)]~1[tcnq]~2 crystallises in the triclinic system, space group P1� ; the structure is similar to that of the analogous [Mo(cp)2(dddt)]~1[tcnq]~2. Molecules are located in general positions in the unit cell (Fig. 2). Note that data were collected at 200 K due to the high thermal parameters which occur at ambient temperature. The 1: 1 salt [W(h- S W S S S S W S But But S S S NC NC CN CN NC NC CN CN F F F F 6 7 Table 2 Electrochemical data (in V vs. Ag–0.1 M AgClO4) Complex [W(cp)2(C3S5)] [Mo(cp)2(C3S5)] [W(h-C5H4But)2(C3S5)] [Mo(h-C5H4But)2(C3S5)] [W(h-C5H4SiMe3)2- (C3S5)] [Mo(h-C5H4SiMe3)2- (C3S5)] [W(cp)2(C3OS4)] [Mo(cp)2(C3OS4)] [W(cp)2(dddt)] [Mo(cp)2(dddt)] E2� 1 (MV]MIV) 20.06 20.01 20.05 20.04 20.03 20.04 20.11 20.08 20.36 20.34 E2� 1 (MVI]MV) 0.42 0.57 0.46 0.53 0.50 0.56 0.43 0.55 0.15 0.26 Ref.This work 7 This work 12 This work 12 This work 7 This work 7J. Chem. Soc., Dalton Trans., 1998, Pages 483–488 485 Table 3 Selected bonds (Å) and angles (8) of M(cp)2(dithiolene) complexes Complex [W(cp)2(S2C6H4)] [V(cp)2(S2C6H4)] (A) (B) [Nb(C5H4SiMe3)2(C3S5)] [Mo(cp)2(dddt)][tcnq] [W(cp)2(dddt)][tcnq] [W(C5H4But)2(C3S5)][tcnqf4] M]S 2.421(2) 2.431(6) 2.426(6) 2.51(1) 2.425(7) 2.413(2) 2.4335(2) C]S 1.753(8) 1.70(2) 1.77(2) 1.78(1) 1.725(7) 1.74(2) 1.7275(8) C]C — — — 1.25(3) 1.373(1) 1.373(11) 1.372(10) Cp(1)]M 2.004 1.992 2.001 2.106 1.996 2.011 2.012 Cp(2)]M 2.004 1.977 1.989 2.095 1.995 2.007 2.017 S]M]S 82.3(3) 79.7(2) 79.9(2) 81.3(6) 80.38(2) 80.66(6) 82.61(7) Cp]M]Cp 137 135.4(6) 133.4(6) 131(2) 135.5(3) 135.3 133.5 q * 8.1 38.3 38.4 34 32.1 33.22 29.28 Ref. 15 16 5 7 This work This work * Folding angle between S]M]S and dithiolene planes. C5H4But)2(C3S5)]~1[tcnqf4]~2 7 crystallises with a solvent molecule (MeCN), in the monoclinic space group P21/n. Both the cation and the anion are located in general position in the unit cell (Fig. 3). The important geometrical features of the cation are compared on the one hand with values reported for d1 complexes and on the other hand with the characteristics of neutral tungstenocene compounds in Table 3.Both cations of 6 and 7 present a folding angle (q) between the ligand plane and the S]M]S plane in the range of those observed for d1 compounds (30–408) (Fig. 4). This behaviour is now well known and experimental results have been rationalised on the basis of extended-Hückel calculations.7,17 The lower value observed in the case of 7 is attributable to the steric crowding induced by the substituted cyclopentadienyl as already demonstrated for [Ti(h-C5Me5)2- (C3S5)].18 Note also that d1 neutral compounds present a more important folding than do the d1 cations as a consequence, may Fig. 2 An ORTEP14 view of the unit cell of [W(cp)2(dddt)]?1[tcnq]?2 6; H atoms are omitted for clarity Fig. 3 An ORTEP14 view of the unit cell of [W(h-C5H4But)2- (C3S5)]?1[tcnqf4]?2 7; H atoms and MeCN molecules are omitted for clarity be, of the dithiolene contribution upon oxidation.Indeed, the higher the electron density on the metal, the less is the folding of the ligand. Furthermore, we have previously noticed the participation of dddt22 ligand in oxidation, the formation of [Mo(cp)2(dddt)]~1 leading to a more diketonic structure for the dithiolene.7 The reduction of the organic acceptor induces a slight shift from a quinonoid to a benzenoid form in both 6 and 7. The extracyclic C]] C bond lengthens especially while the C]C bond lengths shorten.In addition, a comparison of the geometrical features (Table 4) confirms the association into dimers observed in Fig. 2 for salt 6. Note also that Fig. 3 reveals the same behaviour for the anionic species tcnqf4~2. The intradimer plane-to-plane distances are similar to those reported for fully reduced dianionic (tcnq)2 22 moieties (Table 4). Extended- Hückel calculations were performed in order to calculate the energy difference between the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO) of the dimeric dianionic moieties.The values are 0.896 (tcnq) and 0.888 eV (tcnqf4), thus confirming a strong overlap which allows one to consider these diradical moieties as being in a singlet state at room temperature and below; for comparison we calculate a value of 0.837 eV for [Mo(cp)2- (dddt)]~1[tcnq]~2. As for the anionic species, the organometallic dimers are organised centrosymmetrically with parallel dithiolene ligand planes.The plane-to-plane distances are 2.19 (6) and 2.27 Å (7) and the intermolecular M ? ? ? M distance increases from 7.07 to 9.54 Å when the substituted cyclopentadienyl is present. Contrary to the anionic components which tend to a face-to-face arrangement, the dithiolenes lie side by side. In the case of [W(cp)2(dddt)]~1[tcnq]~2 the cationic species are stacked along the a axis within a layer structure parallel to the ab plane, the layers being separated from each other by the tcnq dimers (Fig. 5). The situation is somewhat different with the salt 7 since only columns of dimers are identified along the a axis (Fig. 6). As expected from the intermolecular M ? ? ? M and plane-to-plane distances (molecules 1 and 2 in Figs. 5 and 6), the intradimer interaction calculated by extended-Hückel calculations is found to be stronger in [{W(cp)2(dddt)}2]21 than in [{W(h-C5H4- But)2(C3S5)}2]21, 0.16 and 0.08 eV respectively.These interactions rely mainly on S ? ? ? S overlap due to the following short contacts: S(2)1 ? ? ? S(2)2 (3.730 Å) in [W(cp)2(dddt)][tcnq], Fig. 4 The ORTEP14 diagrams showing eclipse views of [W(cp)2- (dddt)]?1 (left) and [W(h-C5H4But)2(C3S5)]?1 (right); H atoms are omitted for clarity486 J. Chem. Soc., Dalton Trans., 1998, Pages 483–488 Table 4 Experimental geometric parameters (Å) for tcnqn and tcnqf4 n (n = 0 or 21) N N N N 1 2 3 4 5 Distance C(1)]C(2) C(2)]C(3) C(3)]C(4) C(4)]C(5) C(5)]N * tcnq 19 1.345 1.445 1.373 1.438 1.139 — tcnq in 6 1.353 1.428 1.412 1.424 1.152 3.16 [Mo(cp)2(dddt)][tcnq] 7 1.352 1.425 1.420 1.418 1.143 3.20 [Fe(C5Me5)2][tcnq] 20 1.365 1.420 1.410 1.421 1.146 3.15 tcnqf4 21 1.335 1.439 1.373 1.440 1.142 — tcnqf4 in 7 1.340 1.421 1.393 1.423 1.145 3.25 * Plane to plane distance.S(4)1 ? ? ? S(4)2 (3.793) and S(4)1 ? ? ? S(2)2 (3.811 Å) in [W(h- C5H4But)2(C3S5)][tcnqf4]. In the case of [W(h-C5H4But)2- (C3S5)]~1 cations, there is no evidence of interdimer interactions while short contacts S ? ? ? S or S ? ? ? C between molecules are present in the layer of the cationic entities of 6.However extended-Hückel calculations indithe interdimer interactions remain weak (<0.06 eV) compared to the intradimer interaction. Finally, comparison of the analysis above with the results obtained in the case of [Mo(cp)2(dddt)]~1[tcnq]~2 leads to the following preliminary conclusions: (i) acceptors are organised as dimers with strong interactions, contrary to the situation observed in most organic tcnq salts where the anions exhibit a strong tendency to stack in the solid state; (ii) changing the metal centre from molybdenum to tungsten in [M(cp)2(dddt)]~1- [tcnq]~2 gives an isostructural salt in which intermolecular contacts are a little shorter (Table 5); (iii) isolated centrosymmetrical dimers are present with [W(h-C5H4But)2(C3S5)]~1 but the introduction of the large But substituent weakens the interactions. Fig. 5 Intermolecular contacts between [W(cp)2(dddt)]?1 moieties in complex 6 Fig. 6 Intermolecular contacts between [W(h-C5H4But)2(C3S5)]?1 moieties in complex 7 Antiferromagnetic interactions The magnetic susceptibilities of [W(h-C5H4But)2(C3S5)]~1- [tcnqf4]~2 and [W(cp)2(dddt)]~1[tcnq]~2 have been measured in the temperature range 5–300 K. The data are collected in Figs. 7 and 8. The shape of the curves is characteristic of antiferromagnetic exchange interactions with T(cmax) = 22 for 7 and 18 K Fig. 7 Temperature dependence of magnetic susceptibility for complex 6 Fig. 8 Temperature dependence of magnetic susceptibility for complex 7 Table 5 Selected intermolecular distances (Å) in [M(cp)2(dddt)][tcnq] Distance Intradimer M ? ? ? M Intradimer plane to plane dithiolene Interdimer plane to plane dithiolene S(2)1 ? ? ? S(2)2 (Fig. 5) S(3)1 ? ? ? S(3)3 (Fig. 5) S(1)1 ? ? ? C(7)4 (Fig. 5) M = Mo 7.15 2.25 4.50 3.739 3.598 3.684 M = W 7.07 2.19 4.37 3.730 3.569 3.616J.Chem. Soc., Dalton Trans., 1998, Pages 483–488 487 for 6. Since the structural analysis and extended-Hückel calculations predicted isolated dimers for [W(h-C5H4But)2- (C3S5)]~1, we attempted to fit the data as a sum of a Bleaney– Bower law which assumes exchange (J) to be significant only within a dimer 22 and a Curie tail which takes traces of paramagnetic impurities into account, expression (1). cmol = C0 1 C1 T 1 Ng2b2 kT 1 3 1 exp(22J/kT) (1) Fig. 7 shows that in this way the calculated curve corresponds satisfactorily to the experimental data with the following parameters: C0 = 2.31 × 1023 emu, C1 = 1.07 × 1022 K21 and J/k = 218.6 K. In contrast the model of isolated dimers did not give a completely satisfactory fit in the case of [W(cp)2- (dddt)]~1[tcnq]~2 (Fig. 8). This result suggests that the interdimer interactions are not insignificant in this complex. Also, it was tempting to simulate the data by an equation including those interactions. The numerical values of magnetic susceptibility for an alternating chain with two interaction parameters J and aJ (0 < a < 1) have been fitted by Hatfield 23,24 using the analytical expression (2) where A, B, C, D, E and F are polynomial cmol = C0 1 C1 T 1 Ng2b2 kT A 1 Bx 1 Cx2 1 1 Dx 1 Ex2 1 Fx3 (2) functions of a and x = |J|/kT.This model allows a better simulation of the experimental data especially with a = 0.5–0.6 (R = 0.998) and the different fits carried out indicate for J/k a value of the order of 216 K.Note, however, that the fits proposed by Hatfield correspond to a one-dimensional model and therefore these results should be analysed cautiously since layers of cations were identified in the solid-state arrangement. Nevertheless the value of a confirms that the magnetic exchange is not only present within a dimer. This complex exhibits T(cmax) and J/k values suggesting that the antiferromagnetic interaction is of the same order of magnitude as that in complex 7 and half that in [Mo(cp)2(dddt)]~1- [tcnq]~2 [T(cmax) = 32 K, J/k = 226 K].Surprisingly the experimental data [T(cmax)] do not confirm the extended- Hückel calculations which predict comparable interdimer interactions in the two isostructural salts 6 and [Mo(cp)2(dddt)]- [tcnq].‡ In conclusion we have prepared a series of W(h- C5H4R)2(dithiolene) complexes which could be precursors of materials with interesting properties.Further investigations are in progress with the aim to rationalise the dimensionality of the interaction according to the nature of the metal and dithiolene ligand. Experimental All reactions were carried out under argon using standard Schlenk techniques. Solvents were dried over molecular sieves and distilled just prior to use. The compound [W(cp)2Cl2] was prepared according to the literature method;25 [W(h-C5H4R)2Cl2] (R = SiMe3 or But) were synthesised with a similar procedure from the reaction of [WCl4(dme)] 26 (dme = 1,2-dimethoxyethane) with the corresponding salt Li(C5H4R) in diethyl ether.Both tcnq and tcne are commercially available and were used without purification. The method described by Wheland and Martin27 was employed for the synthesis of tcnqf4. The NMR spectra were recorded on a 200 MHz Bruker spectrometer in CDCl3 with SiMe4 as the internal reference. Elemental analyses were performed at the Service Central d’Analyse, CNRS, Vernaison, France.‡ The intradimer interaction calculated by extended-Hückel calculations is 0.15 eV in [{Mo(cp)2(dddt)}2]21. Preparations Bis(Á5-cyclopentadienyl)(4,5-disulfanyl-1,3-dithiole-2-thionato) tungsten(IV), [W(cp)2(C3S5)] 1. Equimolar quantities (1 mmol) of [W(cp)2Cl2] and Na2C3S5 28 were refluxed in thf (50 cm3) overnight. After filtration, the solvent was removed under reduced pressure and the residue extracted with CH2Cl2.The complex [W(cp)2(C3S5)] was isolated as a violet powder after concentration. Yield 0.214 g (42%) (Found: C, 31.0; H, 2.2; S, 29.3. C13H10S5W requires C, 30.6; H, 1.95; S, 31.35%); n& (KBr)/ cm21 3100w (C]H), 1450s (C]] C), 1050s (C]] S) and 835s (cp). (4,5-Disulfanyl-1,3-dithiole-2-thionato)bis(Á5-trimethylsilylcyclopentadienyl) tungsten(IV), [W(Á-C5H4SiMe3)2(C3S5)] 2. As above from equimolar quantities (1 mmol) of [W(h-C5H4- SiMe3)2Cl2] and Na2C3S5. Yield 0.196 g (30%) (Found: C, 35.15; H, 4.05; S, 23.75.C19H26S5Si2W requires C, 34.85; H, 4.0; S, 24.45%); n& (KBr)/cm21 3100w (C]H), 2950w (C]H), 1465s (C]] C), 1055s (C]] S) and 835s (C5H4SiMe3). Bis(Á5-tert-butylcyclopentadienyl)(4,5-disulfanyl-1,3-dithiole- 2-thionato)tungsten(IV) [W(Á-C5H4But)2(C3S5)] 3. This was prepared as above from equimolar quantities (1 mmol) of [W(h- C5H4But)2Cl2] and Na2C3S5. Yield 0.205 g (33%) (Found: C, 39.2; H, 4.25; S, 25.85. C21H26S5W requires C, 40.5; H, 4.2; S, 25.7%); n& (KBr)/cm21 3100w (C]H), 2950w (C]H), 1450s (C]] C), 1050s (C]] S) and 840s (C5H4But). Bis(Á5-cyclopentadienyl)(4,5-disulfanyl-1,3-dithiole-2-onato)- tungsten(IV), [W(cp)2(C3OS4)] 4.Equimolar quantities (1 mmol) of [W(cp)2Cl2] and Na2C3OS4 were employed.29 Yield 0.173 g (35%) (Found: C, 31.0; H, 2.2; S, 27.3. C13H10OS4W requires C, 31.6; H, 2.0; S, 25.9%); n& (KBr)/cm21 3100w (C]H), 1655 and 1600s (C]] O), 1420s (C]] C) and 835s (cp). Bis(Á5-cyclopentadienyl)(5,6-dihydro-1,4-dithiine-2,3-dithiolato) tungsten(IV), [W(cp)2(dddt)] 5.As for complex 1 from equimolar quantities (1 mmol) of [W(cp)2Cl2] and Na2dddt.30 Yield 0.247 g (50%) (Found: C, 34.1; H, 2.95; S, 25.85. C14H14S4W requires C, 34.0; H, 2.85; S, 25.9%); n& (KBr)/cm21 2900w (C]H), 1450s (C]] C) and 810s (cp). [W(cp)2(dddt)][tcnq] 6. Warm solutions of [W(cp)2(dddt)] (50 mg, 0.1 mmol) in MeCN (30 cm3) and tcnq (21 mg, 0.1 mmol) in MeCN (15 cm3) were mixed and cooled. After filtration of a green precipitate (2 : 3 stoichiometry) and concentration, dark crystals of product 6 were isolated: [W(cp)2(dddt)][tcnq], yield ca. 0.045 g (65%) (Found: C, 43.3; H, 2.9; N, 7.35. C26H18N4S4W requires C, 44.7; H, 2.6; N, 8.0%); [W(cp)2- (dddt)2][tcnq]3 (Found: C, 47.65; H, 2.3; N, 10.5. C32H20N6S4W requires C, 48.0; H, 2.5; N, 10.5%). [W(Á-C5H4R)2(C3S5)][tcnqf4] (R 5 But 7 or H 8). A warm solution of [W(h-C5H4R)2(C3S5)] (0.05 mmol) in CH2Cl2 (10 cm3) was added to an equimolar proportion of tcnqf4 (14 mg, 0.05 mmol) in MeCN (10 cm3).Concentration and cooling afforded black crystals in quantitative yield: [W(h-C5H4But)2- (C3S5)][tcnqf4] (Found: C, 44.3; H, 3.3; N, 7.2. C33H26F4N4- S5W?CH3CN requires C, 44.7; H, 3.1; N, 7.45%); [W(h-C5H5)2- (C3S5)][tcnqf4] (Found: C, 37.85; H, 1.65; N, 6.85. C25H10- F4N4S5W requires C, 38.15; H, 1.25; N, 7.1%). [W(cp)2(C3OS4)][tcnqf4] 9. As above from equimolar quantities of [W(cp)2(C3OS4)] and tcnqf4.Yield ca. 0.030 g (80%) (Found: C, 38.4; H, 1.3; N, 6.95. C25H10F4N4S5W requires C, 38.95; H, 1.3; N, 7.25%). Electrochemical measurements Cyclic voltammetry experiments were performed with a Radiometer PGP 201 potentiostat. The electrolyte consisted of a488 J. Chem. Soc., Dalton Trans., 1998, Pages 483–488 0.1 M NBun 4PF6 solution in acetonitrile dried on molecular sieves. A three-compartment cell was used with a platinum working electrode (diameter 2 mm), a platinum counter electrode, and an Ag–0.1 M AgClO4 reference electrode.Experiments were performed under an argon flow. After each measurement the reference electrode was checked against the ferrocene–ferrocenium couple (10.025 V). Crystallography Crystals were mounted in glass capilleries using araldite glue. Data were collected on a Stoe-IPDS diffractometer, with graphite-monochromated Mo-Ka radiation (l 0.710 73 Å). Details are given in Table 6. Empirical absorption corrections were made for complex 7 (multi-scan, X-PREP, SHELXTL 5.04 31) and a numerical one for 6 (FACEIT, Stoe).Structures were solved by direct methods and refined anisotropically by full-matrix least squares on F2 (program SHELXTL 5.04). Hydrogen atoms were placed in calculated positions (C]H 0.93 Å), included in structure-factor calculations but not refined. CCDC reference number 186/807. See http://www.rsc.org/suppdata/dt/1998/483/ for crystallographic files in .cif format.Extended-Hückel calculations Extended-Hückel calculations were performed with double-z quality orbitals for C, W, S, F, N and with single-z orbitals for H taken from ref. 32. Magnetic susceptibility measurements Magnetic susceptibility data were collected on a commercial Quantum Design MPMS5 SQUID susceptometer. Data were Table 6 Crystallographic data for complexes 6 and 7a Formula M Crystal size/mm T/K Crystal system Space group a/Å b/Å c/Å a/8 b/8 g/8 U/Å3 Z Dc/g cm23 m/mm21 Tmax, Tmin Minimum, maximum, h, k, l F(000) Collected reflections Independent reflections Goodness of fit R, R9 Minimum, maximum residual peak/e Å23 6 C26H18N4S4W 698.53 0.18 × 0.16 × 0.024 200 Triclinic P1� 9.5090(14) 10.5400(14) 12.385(2) 84.310(14) 81.45(2) 89.32(2) 1221.4(3) 2 1.899 5.095 0.86, 0.44 210, 10; 212, 12; 214, 14 680 10 218 3624 1.019 0.0366, 0.0865 23.075, 1.627 b 7 C35H29F4N5S5W 939.78 0.21 × 0.12 × 0.09 293 Monoclinic P21/n 7.9560(13) 21.020(3) 21.722(4) 96.390(14) 3610.1(10) 4 1.729 3.544 0.277, 0.203 29, 9; 223, 23; 224, 24 1856 22 608 5715 0.848 0.0358, 0.0667 21.522, 0.560 b a Details in common: black parallelepiped crystal; qmax 24.28; R = S Fo| 2 |Fc /SFo; R9 = {Sw(Fo 2 2 Fc 2)2]/S[w(Fo 2)]2}� �� .b Found near the W atom at a distance less than 1 Å. corrected for molecular diamagnetism and holder contribution. References 1 P. Cassoux, L. Valade, H. Kobayashi, R. A. Clar and A. E. Underhill, Coord. Chem. Rev., 1991, 110, 115; S.Q. Sun, B. Zhang, P. J. Wu and D. B. Zhu, J. Chem. Soc., Dalton Trans., 1997, 277. 2 R. M. Olk, B. Olk, W. Dietzsch, R. Kirmse and E. Hoyer, Coord. Chem. Rev., 1992, 177, 99. 3 H. Ushijima, S. Sudoh, M. Kajitani, K. Shimizu, T. Akiyama and A. Sugimori, Appl. Organomet. Chem., 1991, 5, 221; F. Guyon, C. Lenoir, M. Fourmigué, J. Larsen and J. Amaudrut, Bull. Soc. Chim. Fr., 1994, 131, 217. 4 S. Zeltner, W. Dietzsch, R. M. Olk, R. Kirmse, R. Richter, U. Schröder, B. Olk and E.Hoyer, Z. Anorg. Allg. Chem., 1994, 620, 1768. 5 F. Guyon, M. Fourmigué, R. Clérac and J. Amaudrut, J. Chem. Soc., Dalton Trans., 1996, 4093. 6 M. L. H. Green, W. B. Heuer and G. C. Saunders, J. Chem. Soc., Dalton Trans., 1990, 3789. 7 M. Fourmigué, C.Lenoir, C. Coulon, F. Guyon and J. Amaudrut, Inorg. Chem., 1995, 34, 4979. 8 M. L. H. Green, A. Harrison, P. Mountford and D. K. P. Ng, J. Chem. Soc., Dalton Trans., 1993, 2215. 9 P. K. Baker, M. G. B. Drew, E. E. Parker, N.Robertson and A. E. Underhill, J. Chem. Soc., Dalton Trans., 1997, 1429. 10 G. Matsubayashi, K. Douki, H. Tamura, M. Nakano and W. Mori, Inorg. Chem., 1993, 32, 5990. 11 M. L. H. Green and W. E. Lindsell, J. Chem. Soc. A, 1967, 1455. 12 F. Guyon, Ph.D. Thesis, Université de Besançon, 1994. 13 J. C. Kotz, W. Vining, W. Coco, R. Rosen, A. R. Dias and M. H. Garcia, Organometallics, 1983, 2, 68. 14 C. K. Johnson, ORTEP, Report ORNL-5138, Oak Ridge National Laboratory, Oak Ridge, TN, 1976. 15 T. Debaerdemacker and A. Kutoglu, Acta Crystallogr., Sect. B, 1973, 29, 2664. 16 D. W. Stephan, Inorg. Chem., 1992, 31, 4218. 17 J. W. Lauher and R. Hoffmann, J. Am. Chem. Soc., 1976, 98, 1729. 18 F. Guyon, M. Fourmigué, P. Audebert and J. Amaudrut, Inorg. Chim. Acta, 1995, 239, 117. 19 R. E. Long, R. A. Sparks and K. N. Trueblood, Acta Crystallogr., 1965, 18, 932. 20 J. S. Miller, J. H. Zhang, W. M. Reiff, D. A. Dixon, L. D. Preston, A. H. Reis, E. Gebert, M. Extine, J. Troup, A. J. Epstein and M. D. Ward, J. Phys. Chem., 1987, 91, 4344. 21 T. J. Emge, M. Maxfield, D. O. Cowan and T. J. Kistenmach, Mol. Cryst. Liq. Cryst., 1981, 65, 161. 22 J. C. Bonner, H. W. J. Blöte, J. W. Bray and I. S. Jacobs, J. Appl. Phys., 1979, 50, 1810. 23 W. E. Hatfield, J. Appl. Phys., 1981, 52, 1985. 24 O. Kahn, Molecular Magnetism, VCH, Weinheim, 1993, ch. 11. 25 C. Persson and C. Andersson, Organometallics, 1993, 12, 2370. 26 C. Persson and C. Andersson, Inorg. Chim. Acta, 1993, 203, 235. 27 R. C. Wheland and E. L. Martin, J. Org. Chem., 1975, 40, 3101. 28 K. S. Varma, A. Bury, N. J. Harris and A. Underhill, Synthesis, 1987, 837. 29 R. R. Shumaker and E. M. Engler, J. Am. Chem. Soc., 1977, 99, 5521. 30 J. Larsen and C. Lenoir, Synthesis, 1989, 134. 31 G. M. Sheldrick, SHELXTL, version 5, Siemens Analytical X-Ray Division, Madison, WI, 1994. 32 M.-H. Whangbo, J. M. Williams, P. C. W. Leung, M. A. Beno, T. J. Emge, H. H. Wang, K. D. Carlson and G. W. Crabtree, J. Am. Chem. Soc., 1985, 105, 5815; E. Clementi and C. Roetti, At. Data Nucl. Data Tables, 1974, 14, 177. Received 5th September 1997; Paper

ISSN:1477-9226    

DOI:10.1039/a706495k

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1999, 487–488 487 Simulation of ethylene insertion in an aluminium catalyst Meike Reinhold,a John E. McGrady a and Robert J. Meier a,b a Chemistry Department, University of York, Heslington, York, UK YO10 5DD b DSM Research, PO Box 18, 6160 MD Geleen, The Netherlands. E-mail: rob.meier@dsm-group.com Received 7th October 1998, Accepted 2nd December 1998 Ethylene insertion into [AlMe{MeC(NMe)2}]1 has been studied using quantum mechanical simulations.Both planewave Car–Parrinello and localised basis set DFT calculations predict an insertion barrier of approximately 25 kcal mol21, in close agreement with estimates derived from experimental data. The calculated barrier for insertion into a methyl bridged dinuclear aluminium species is over 10 kcal mol21 higher, suggesting that the monomeric species is the active catalyst. Simple aluminium compounds, AlR3, are known to be active in olefin oligomerisation, but their use in catalytic polymerisation has proved more problematic, due to rapid chain termination.The inactivity of these main group complexes is in marked contrast to the numerous catalytic pathways involving complexes of metals of the transition and lanthanide series. However, recent work by Coles and Jordan1 has shown that simple cationic aluminium complexes [AlMe{RC(NR9)2}]1 (see 1 where R = R9 = Me) can catalyse ethylene polymerisation with activities of several thousand gram polyethylene (PE) mol21 h21 (atm C2H4)21, yielding PE with molecular masses up to 105 Daltons.Since this initial report, a range of aluminium-based compounds has been synthesized and tested for ethylene polymerisation activity, but the nature of the active species remains unknown. Given the propensity of the electron deficient aluminium centres to form bridged dinuclear species, it is possible that the observed catalytic activity is not due to the mononuclear complex, [AlMe{RC(NR9)2}]1, but instead a methyl-bridged dinuclear adduct such as [{[AlMeRC- (NR9)2]}2(m-Me)]1 (2, R = R9 = Me).Spectroscopic investigations 2 into the equilibrium distribution of species in solution are currently underway in this laboratory, but in the context of the current paper it is suYcient to note that if several species are present the overall polymerisation rate will be determined by their relative concentrations, as well as the catalytic activity of each one. In recent years quantum mechanical simulations have greatly enhanced our understanding of the mechanisms of catalytic cycles.Static calculations have been used in a number of cases to construct potential energy profiles, from which the energy barrier for the reaction can be abstracted (see, e.g., Margl et al.3 and refs. therein). More recently, first principles molecular dynamics 4 simulations using plane-wave based methods such as the Car–Parrinello (CP) technique have been reported for catalytic cycles.5–7 The computational cost of the plane-wave calculations is less size-dependent than the corresponding calculations using localised basis sets, and so the former are more suited to the study of larger systems such as the dinuclear species noted above.There are, however, relatively few papers in the literature providing a direct comparison of the two methodologies, and so here we employ both techniques to study the insertion of ethylene into the mononuclear species [AlMe- {MeC(NMe)2}]1 1.Having established that the CP code gives similar results to static DFT for the monomeric system, we then apply the plane-wave technique to compare ethylene insertion into the monomer 1 with the corresponding process in the model dinuclear species 2. Computational details Density functional calculations were performed using the Amsterdam Density Functional code ADF, version 2.3.0.8–10 Carbon, nitrogen and hydrogen atoms were described by a double zeta 1 polarisation basis set, while a triple zeta 1 polarisation basis was used for aluminium.The local density approximation was employed throughout, along with the correlation functional of Vosko, Wilk and Nusair 11 and gradient corrections to exchange (Becke) 12 and correlation (Perdew)13. The first principles molecular dynamics calculations are based on the original Car–Parrinello code.4 The Perdew–Zunger parametrisation 14 of the exchange-correlation function was used, along with non-local corrections according to Perdew and Becke 12,13.Soft Vanderbilt 15 pseudo-potentials were employed, except for Al. The cell size chosen ensured empty space left between images in neighbouring cells in order to minimise unwanted interaction between images. The common geometry optimisation step by steepest descent or conjugate gradient techniques was replaced by a short series of low temperature dynamics simulations, at T = 50, 20 and finally 10 K.In the CP simulations a 15 Rydberg cut-oV energy was found to be suYcient by comparing to the entire set of equivalent data obtained for a 25 Rydberg cut-oV. Results and discussion Potential energy curves for ethylene insertion into the mononuclear aluminium complex 1, obtained using both DFT and CP calculations, are shown in Fig. 1. All parameters are freely optimised, with the exception of the distance between the methyl carbon attached to Al (C3) and one carbon of the ethylene molecule (C2), which is used to define the reaction coordinate.The energy profiles derived from the two methodologies are very similar, both predicting a barrier to insertion of around 25 kcal mol21. For the aluminium catalyst with R = tBu and488 J. Chem. Soc., Dalton Trans., 1999, 487–488 R9 = iPr, the highest experimentally determined activities1 at 60 and 85 8C are 2480 and 3050 g PE mol21 h21 (atm C2H4)21 respectively (details are given in the Supplementary Material to ref. 1). Assuming that the pressure applied is 1 atm, these values correspond to reaction rates of approximately one ethylene insertion every 30 s. From the Arrhenius rate equation, k = A?exp(2Eact/RT), and assuming A = 1013 s21, the estimated activation energy lies in the range 22–24 kcal mol, in excellent agreement with our predicted values. The structures of three key points along the curve are shown in the figure, corresponding to (a) the van der Waals complex between catalyst and ethylene, (b) the transition state,† (c) the final product (the structures shown are generated from the static DFT calculations, but those from the CP code are very similar).The van der Waals complex between the catalyst and ethylene complex (C2–C3 3.0 Å) lies in a shallow minimum in the potential energy curve, with Al–C1 and Al–C2 separations of 2.58 and 2.42 Å respectively. In the transition state (C2–C3 2.15 Å) the aluminium centre is approximately symmetric with respect to the terminal carbon atoms (Al–C1 2.02, Al–C3 2.05 Å), and the methyl group is forced below the plane of the amidinate ring, causing a distinct pyramidalisation at the aluminium centre. In the product (C2– C3 1.53 Å) the Al–C3 bond is completely cleaved (2.97 Å), and the co-ordination about the aluminium centre reverts to approximately trigonal planar, with the metal-bonded carbon of the newly formed propyl group lying in the amidinate plane.Fig. 1 Energy profile for the insertion of ethylene into the Al–CH3 group of mononuclear complex {MeC(NMe)2}AlMe1 (results from ADF calculations are shown as triangles, those from the CP method as diamonds). The reaction coordinate is defined by the C2–C3 distance.ADF Calculated structures of the van der Waals complex (C2–C3 3.0 Å), the transition state (C2–C3 2.15 Å) and the product (C2–C3 1.53 Å) are also shown. † Here taken as the maximum along the reaction coordinate, which was located to within 0.02 Å. Preliminary studies, using the CP methodology, suggest that the energetics of insertion into one of the non-bridging methyl groups of the dinuclear aluminium complex 2 are rather diVerent from those discussed above.The overall reaction is thermochemically neutral, and shows a substantially larger barrier (40 kcal mol21) to insertion. We are currently investigating the electronic origin of these diVerences, but initial results suggest that it is the mononuclear complex which is the active species in the polymerisation.Conclusion Plane-wave based Car–Parrinello techniques, along with static density functional methods using localised basis sets, have been used to calculate the barrier to insertion of ethylene into a model aluminium catalyst, [AlMe{MeC(NMe)2}]1. The two techniques predict almost identical values of approximately 25 kcal mol21, in excellent agreement with the estimate of 22–24 kcal mol21 extracted from experimental polymer yields.Initial calculations indicate that the barrier to insertion in a dinuclear methyl-bridged species is much higher, suggesting that the mononuclear species is the active catalyst. Acknowledgements The CP code employed in this work is under development at the Forum-INFM laboratory in Pisa, Italy, by Dr. Franco Buda and co-workers, who we thank for their strong support. We gratefully acknowledge Robin Perutz (York) for very valuable discussions and for critically reading the manuscript. References 1 M. P. Coles and R. F. Jordan, J. Am. Chem. Soc., 1997, 119, 8125. 2 E. Koglin, D. Koglin, R. J. Meier and J. van Heel, Chem. Phys. Lett., 1998, 290, 99. 3 P. Margl, L. Deng and T. Ziegler, J. Am. Chem. Soc., 1998, 120, 5517. 4 R. Car and M. Parrinello, Phys. Rev. Lett., 1985, 55, 2471. 5 R. J. Meier, G. H. J. VanDoremaele, S. Iarlori and F. Buda, J. Am. Chem. Soc., 1994, 116, 7274. 6 S. Iarlori, F. Buda, R. J. Meier and G. H. J. VanDoremaele, Mol. Phys., 1996, 87, 801. 7 O. M. Aagaard, R. J. Meier and F. Buda, J. Am. Chem. Soc., 1998, 120, 7174. 8 E. J. Baerends, D. E. Ellis and P. Ros, Chem. Phys., 1973, 2, 41. 9 E. J. Baerends and P. Ros, Chem. Phys., 1973, 2, 52. 10 G. te Velde and E. J. Baerends, J. Comp. Phys., 1992, 99, 84. 11 S. H. Vosko, L. Wilk and M. Nussair, Can. J. Phys., 1980, 58, 1200. 12 A. Becke, Phys. Rev. A, 1988, 38, 3098. 13 J. P. Perdew, Phys. Rev. B, 1986, 33, 8822. 14 J. P. Perdew and A. Zunger, Phys. Rev. B, 1981, 23, 5048. 15 D. Vanderbilt, Phys. Rev. B, 1990, 41, 7892. Paper 8/07809B

ISSN:1477-9226    

DOI:10.1039/a807809b

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1999, 489–496 489 On the bonding isomerism in three-co-ordinated copper(I) thiocyanates José A. Dobado,a Rolf Uggla,a Markku R. Sundberg a and José Molina b a Laboratory of Inorganic Chemistry, Department of Chemistry, P.O. Box 55 (A.I. Virtasen aukio 1), FIN-00014 University of Helsinki, Finland. E-mail: sundberg@cc.helsinki.fi b Grupo de Modelización y Diseño Molecular, Instituto de Biotecnología, Campus Fuentenueva, Universidad de Granada, E-18071 Granada, Spain.E-mail: jmolina@goliat.ugr.es Received 14th August 1998, Accepted 4th December 1998 Density functional theory (DFT) and ab initio MP2 methods have been applied to characterize the structural features of seven diVerent bonding isomers of copper(I) thiocyanate dianion complexes (S- and/or N-bonded). The DFT calculations were carried out by means of the hybrid Becke 3LYP functional, using the 6-3111G* basis set. The ab initio calculations were done at the MP2/6-3111G* theoretical level.The results indicate that in the gas phase N-bonding is preferred to S-bonding. The Atoms in Molecules theory was also employed to study the electronic properties in these isomers. The co-ordination bond between the copper(I) cation and the donor atoms is strongly polarized, almost ionic. The charge depletion around the copper(I) cation is in accordance with sp2 hybridization. Moreover, the canonical form for the non-co-ordinated as well as S-co-ordinated thiocyanates is mainly 2S–C]] ] N, whereas the N-bonded thiocyanates have also 2N]] C]] S contribution. 1 Introduction Three-co-ordination in metal complexes is relatively uncommon.It has been suggested that steric hindrance of the coordinating ligand itself is the main factor in limiting the co-ordination number to three.1 Surprisingly, there are two structural reports of discrete three-co-ordinated complexes containing a transition metal and thiocyanate anion as ligand,2,3 despite of the lack of steric hindrance.Generally, structural studies on mononuclear copper(I) complexes containing merely monodentate S-donor ligands are scarce 4–9 owing to the pronounced tendency of copper(I) ions to form polynuclear complexes with sulfur containing ligands.7,10–18 The available structural information on copper(I) thiocyanate systems is limited to polymeric copper thiocyanate itself 19,20 and the pyridinium salt of the polyanion [Cu(SCN)3 2]•,21 broken to give single and double stranded polymeric chains 2,23 in which the anion still bridges the metal nuclei.There is a basic question connected with the possible coordination of a thiocyanate anion to copper(I) cation: is the ligand S- or N-co-ordinated? The hard–soft acid–base (HSAB) principle simply states that hard acids prefer to co-ordinate to hard bases and soft acids to soft bases.24,25 Since the copper(I) cation is classified as a soft Lewis acid, its co-ordination to sulfur of the thiocyanate anion would be favoured.In general, bending of a thiocyanate anion is connected with concomitant change in the electronic configuration. If the carbon atom has a pure sp hybridization the anion is linear. A change of the hybridization towards sp2 results in bending in the S–C–N angle, which may lead even to an h (sideways) co-ordination mode.26 There are various factors influencing the co-ordination of thiocyanate anion.27,28 These are the electronic and steric eVects of ancillary ligands as well as solvent and counter ion eVects.The N- and S-bonding in copper(I) complexes is also of importance in photosynthesis. Plastocyanin is a water-soluble protein that receives electrons from a cytochrome complex. The redox center of plastocyanin consists of a copper ion coordinated to the side chains of a cysteine, a methionine, and two histidine residues.30,31 Thiocyanate anion is also a strong protein-stabilizing and structure-destabilizing agent owing to its interactions with the protein involved.32 Density Functional Theory (DFT) 33,34 has generated a lot of interest, and it can potentially be applied to larger systems than any accurate ab initio method currently in use.Moreover, DFT methods also include electron correlation and they have been applied successfully for the study of transition metal complexes.35–42 The aim of the present work is to study the competition between sulfur and nitrogen of a thiocyanate anion in coordination with a copper(I) ion, and to characterize the unusual three-co-ordination mode for these complexes.To achieve this, the energetics, structure and harmonic vibrations of the seven diVerent copper(I) thiocyanate dianion complexes (S and/or N bonded) have been investigated using quantum chemical methods. We employed both B3LYP and MP2 methods to compare their applicability to copper(I) complexes. Although the HSAB principle and simple valence bond models are able to give a qualitative picture of the bonding in many cases, we use here the Atoms in Molecules (AIM) theory to produce a detailed quantitative description for the bonding properties.It has been used in the electronic description of many types of compounds.43–46 2 Methods of calculation All calculations have been carried out using the GAUSSIAN 94 package 47 of programs. The standard grids have been used for the integration of the DFT electron density. All geometries have been fully optimized at the B3LYP/6-3111G* and MP2(full)/6- 3111G* levels, and all stationary points on the hypersurface have been characterized as true minima by harmonic frequency analysis at the B3LYP/3-211G*//B3LYP/3-211G* level.Bader analyses have been performed by the AIMPAC series of programs48 using the wavefunction at B3LYP and MP2 levels as input, as described in AIM theory.49,50 An overview of the AIM theory can be obtained elsewhere.46490 J.Chem. Soc., Dalton Trans., 1999, 489–496 3 Results and discussion A Geometries and energetics To gain an insight into the geometries in N- and S-bonded copper(I) compounds displaying three-co-ordination we carried out a survey on the structures in CSDS (Cambridge Structural Database).51,52 The search was restricted to the discrete monomeric complexes with three identical ligands. In total 10 N-coordinated and 14 S-co-ordinated complex units were found. It is noteworthy that no mixed monomeric complexes exist.The maximum out-of-the-plane deviations for the central copper cation in the N- and S-co-ordinated complexes were 0.205 and 0.249 Å, respectively. However, in most of the complexes the deviation was very small. This can be seen in the average values for the deviations, which were for the N- and S-bonded complexes 0.05(6) and 0.07(7) Å, respectively. Three N-bonded complexes displayed zero deviation, whereas no such planarities were found for the S-bonded moieties.The complexes are characterized by significant variation in the bond lengths and in the bond angles around the central copper(I) cation. The ranges for the Cu–S and Cu–N bond lengths are 2.213–2.338 and 1.934– 2.096 Å, respectively. The wide variation is even more pronounced in the ranges for the S–Cu–S and N–Cu–N angles, which are 108.9–139.4 and 106.1–141.58, respectively. However, there was one N-bonded and two S-bonded complex units, where a threefold rotation symmetry axis could be found.Based upon the results discussed above we constructed seven possible bonding isomers assuming planarity for the moieties. The results of the optimizations for the copper(I) trithiocyanate at the diVerent theoretical levels are listed in Table 1, for structure 1, and in Figs. 1 and 2 for structures 2–7. All of the structures 1–7 display at least a symmetry plane (Cs); 2 and 3 had an even higher symmetry of C3h and D3h, respectively.There are some basic features that are characteristic for the optimized structures. In the S-bonded isomers the C–S–Cu angle is always bent, whereas in the N-bonded complexes the corresponding C–N–Cu angle is always 1808. Moreover, there is always a slight bending involved for the S–C–N angle in the former isomers, whereas the angle is always linear in the latter ones. Fig. 1 Copper(I) trithiocyanate anion structures 1–4, with the corresponding geometrical parameters (B3LYP and in parentheses MP2 values, all in Å and 8).In general, the MP2 level of theory seems to give more reliable geometrical parameters, when the experimental structure and the optimized structure 1 are compared (see Table 1). Usually the B3LYP method tends to give too long Cu–S bond lengths. The biggest deviations between the experimental and calculated structures occur at the parameters concerning one of the thiocyanate groups (with the subscript 2 in our notation). However, all of the experimental parameters (including the thermal displacement values) are anomalous.Obviously there must be either a disorder in the structure or the anomalies are mathematical artefacts due to the modest quality of the diVraction data [the crystallographic R(F) value was 0.079].3 As stated in the Introduction, there are only two reports concerning discrete trithiocyanate complexes of copper. Both of the papers describe the structure determination for bis[6- amino-5-(2-ethylphenylazonium)-1,3-dimethyluracil] tris(thiocyanato- S)cuprate(I).2,3 In the former report the single-crystal structure determination was carried out at room temperature and in the latter at 193 K. In the subsequent discussion we Fig. 2 Copper(I) trithiocyanate anion structures 5–7, with the corresponding geometrical parameters (B3LYP and in parentheses MP2 values, all in Å and 8).J. Chem. Soc., Dalton Trans., 1999, 489–496 491 Fig. 3 A perspective view of the two crystallographically independent complex units in bis[6-amino-5-(2-ethylphenylazonium)-1,3-dimethyluracil] tris(thiocyanato-S)cuprate(I). The input co-ordinates are taken from ref. 3 . The short contacts between the anions and cations are depicted by dashed lines. The hydrogen atoms are omitted for clarity. The orientation for each complex unit is chosen to give the best possible illustration of the short contacts for each nitrogen atom of the thiocyanate groups. refer to the latter paper only.The experimental and theoretical structural data for this compound are listed in Table 1. Two crystallographically independent three-co-ordinated tri(thiocyanato- S)cuprate(I) moieties (Exptl. I and Exptl. II, respectively) were found in the same unit cell (see Fig. 3). The geometrical parameters are reproduced reasonably well by the computational methods, especially at the MP2 level. The average value for the experimental Cu–S bond lengths is 2.254 Å and the corresponding calculated value is 2.253 Å (MP2).There is notably more deviation in the S–Cu–S angles, however. Obviously the angles show a propensity to deform easily, as is seen also in the statistical variation for the angles in the structures found from CSDS. Both anions in the experimental structure present bent SCN groups (ca. 1768). Furthermore, both coordination moieties in the unit cell display non-planarity with one of the three S–Cu–S–C torsion angles of one SCN group out of the plane constructed through three sulfur atoms, ca. 19 and 258, respectively. According to a recent review thiocyanate anions are very susceptible to participate in hydrogen bonding.54 A closer inspection of the experimental structure reveals that there are relatively short intermolecular distances between each of the N Table 1 Geometrical parameters (Å and 8) for thiocyanate ion and for the structure 1, at diVerent theoretical levels B3LYP MP2 Exptl. I Exptl. II Thiocyanate ion (C•v) S–C C–N 1.669 1.175 1.657 1.195 1.689 ± 0.013 a 1.149 ± 0.014 a 1.63 b 1.15 b Structure 1 (Cs) c Cu–S(1) Cu–S(2) Cu–S(3) S(1)–C(1) S(2)–C(2) S(3)–C(3) C(1)–N(1) C(2)–N(2) C(3)–N(3) S(1)–Cu–S(2) S(2)–Cu–S(3) S(3)–Cu–S(1) Cu–S(1)–C(1) Cu–S(2)–C(2) Cu–S(3)–C(3) S(1)–C(1)–N(1) S(2)–C(2)–N(2) S(3)–C(3)–N(3) 2.438 2.303 2.340 1.680 1.682 1.678 1.172 1.172 1.172 115.3 147.8 97.0 114.3 112.9 115.4 176.5 175.2 175.5 2.312 2.205 2.240 1.671 1.679 1.673 1.191 1.189 1.188 115.4 144.6 100.1 108.6 110.3 112.6 176.3 174.5 174.5 2.264 2.228 2.227 1.62 1.69 1.65 1.12 1.16 1.18 113.1 135.7 111.1 100 102 108 177 175 178 2.316 2.271 2.253 1.63 1.41 1.64 1.16 1.30 1.15 111.0 139.0 109.0 104 89 107 176 158 175 a X-Ray data from ref. 53 for potassium thiocyanate. b X-Ray data from ref. 53 for ammonium thiocyanate. c Exptl. I and II data from ref. 3 correspond to two crystallographically independent isomers. The atom numbering is depicted in Fig. 1. atoms of the thiocyanate anions and the amino groups of the cations.These interactions are seen for both of the crystallographically independent complex units (Fig. 3). There are bifurcated contacts with one of the amino groups to two of the thiocyanate groups to form an almost coplanar system. The third thiocyanate group is bent towards an amino group of the other cation. It is readily seen that the isomer formation of this compound is governed by the ratio of two cations to one anion. Accordingly, the existence of structure 2 is highly improbable in the solid state.We checked the torsional barrier of one of these SCN groups (structures 1 and 2). The values for the barrier were very small (ca. 1.6 kcal mol21, at the B3LYP/LANL2DZ level), indicating an almost free rotation around the S(2)–Cu–S(1)–N(1) angle. Therefore, the intermolecular interactions are very likely playing a marked role in defining the conformations in the solid state. It is also very interesting that the solid state structure corresponds to the isomer with the least favourable energy.We suggest that hydrogen bonding must play an important role here. On the other hand, the energy diVerences are small, especially at the MP2 level (<7 kcal mol21). There are several examples of how the co-ordination mode varies depending on the solvent. So, in the series of ML2(CNS)2 complexes (where M = PdII or PtII, CNS = the thiocyanate ligand without any reference to the co-ordination mode and L = PPh3, AsPh3 or SbPh3) the only co-ordination mode is the S-bonding, when a solvent with a high relative permittivity is used.However, when solvents like C6H6 or CHCl3 with lower values are applied, the complexes display either a mixture of S- and N-bonding or even N-bonding alone.55 In another example, [Co(CN)5(CNS)]32 exists in an equilibrium of the S-bonded (70%) and N-bonded (30%) isomers in an aqueous solution. Yet the N-bonded isomer is more stable, if the solvent is CH2Cl2, PhNO2, 2- furaldehyde or Me2CO.56 Accordingly, it seems that a M–SCN fragment will form stronger hydrogen bonds than a M–NCS moiety.Indeed, thiocyanic acid displays a structure where the proton is attached to the nitrogen atom.56 Therefore, we suggest that the seemingly reversed order of the stability may well be explained by hydrogen bonding. A similar suggestion was recently made by Fukushima et al.57 who discussed the eVect of solvents on a series of thiocyanate complexes of ZnII, CdII and HgII.The theoretical S–C and C–N bond lengths at both levels (ca. 1.675 and 1.18 Å, respectively) are within the experimental values 1.64(4) and 1.14(5) Å, respectively.† With respect to the other geometrical parameters for structure 1, the agreement for bond angles is poorer. The largest deviation for the Cu–S–C angles is ca. 148. As will be discussed later on, the bonds between the central copper(I) ion and the ligands are very † Experimental values obtained for 68 fragments found from CSDS.Each fragment contained a transition metal and a monoco-ordinated thiocyanate group.492 J. Chem. Soc., Dalton Trans., 1999, 489–496 Table 2 Relative energies Er/kcal mol21 and dipole moments m/D for structures 1–7 at diVerent theoretical levels, taking the calculated total energies of 3 at each level as reference Level of 1 (Cs) 2 [C(3)h] 3 (D3h) 4 (Cs) 5 [C(2)v] 6 (Cs) 7 [C(2)v] theory B3LYP MP2 Er 18.83 6.83 m 3.40 3.86 Er 17.45 3.81 m —a —a Er 0.00 b 0.00 c m —a —a Er 5.39 1.42 m 2.35 1.97 Er 12.85 4.84 m 3.97 3.29 Er 10.99 2.57 m 1.80 1.81 Er 11.43 3.58 m 2.60 3.34 a Dipole moment equal to zero by symmetry.b Calculated total energy 23114.0036052 au. c Calculated total energy 23111.2932200 au. Table 3 Charge density r(r) (e/a0 3), Laplacian of the charge density ,2r(r) (e/a0 5), ellipticity e and local energy density Ed(r) (hartree/a0 3) for BCPs for structures 1, 2, 3 and 5 r(r) ,2r(r) e Ed(r) Structure (bond) 1 Cu–S(1) Cu–S(2) Cu–S(3) S(1)–C(1) S(2)–C(2) S(3)–C(3) C(1)–N(1) C(2)–N(2) C(3)–N(3) 2 Cu–S S–C C–N 3 Cu–N N–C C–S 5 Cu–S(1) Cu–N(2) C(1)–N(1) N(2)–C(2) S(1)–C(1) C(2)–S(2) HSCN S–C C–N S–H SCN2 S–C C–N SCNH S–C C–N N–H SCNLi S–C C–N N–Li MP2 0.068 0.083 0.077 0.204 0.203 0.203 0.433 0.434 0.434 0.077 0.203 0.433 0.098 0.425 0.207 0.080 0.088 0.434 0.425 0.203 0.207 0.208 0.452 0.210 0.205 0.443 0.222 0.413 0.325 0.219 0.435 0.050 B3LYP 0.053 0.068 0.063 0.205 0.204 0.204 0.458 0.459 0.459 0.067 0.205 0.459 0.087 0.450 0.208 0.064 0.079 0.459 0.451 0.204 0.209 0.207 0.469 0.203 0.206 0.455 0.226 0.422 0.328 0.221 0.443 0.054 MP2 0.170 0.219 0.202 20.053 20.089 20.082 20.390 20.371 20.371 0.205 20.065 20.388 0.511 20.289 0.111 0.213 0.448 20.369 20.301 20.088 0.105 20.073 20.045 20.582 0.292 20.446 0.749 20.146 22.059 0.611 20.681 0.388 B3LYP 0.120 0.170 0.154 20.163 20.198 20.190 20.395 20.372 20.374 0.172 20.167 20.392 0.440 20.336 20.018 0.157 0.382 20.377 20.346 20.191 20.014 20.451 20.120 20.519 20.104 20.451 0.373 20.220 21.854 0.226 20.644 0.392 MP2 0.018 0.010 0.004 0.131 0.163 0.158 0.004 0.006 0.006 0.022 0.147 0.004 0.061 0.022 0.004 0.010 0.048 0.006 0.013 0.155 0.004 0.599 0.048 0.099 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 B3LYP 0.011 0.021 0.024 0.106 0.135 0.130 0.006 0.008 0.008 0.003 0.123 0.007 0.048 0.019 0.003 0.016 0.033 0.007 0.010 0.122 0.002 0.309 0.018 0.086 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 MP2 20.020 20.026 20.024 20.241 20.240 20.241 20.755 20.757 20.757 20.024 20.240 20.755 20.020 20.732 20.243 20.025 20.017 20.758 20.733 20.240 20.243 20.252 20.809 20.177 20.232 20.782 20.239 20.695 20.548 20.240 20.758 0.012 B3LYP 20.014 20.019 20.018 20.237 20.235 20.236 20.807 20.811 20.810 20.019 20.238 20.808 20.015 20.784 20.243 20.018 20.013 20.810 20.785 20.236 20.243 20.214 20.839 20.164 20.239 20.799 20.260 20.705 20.502 20.258 20.760 0.011 polarized.Therefore, the orientation of the ligands is very likely dependent on the intermolecular interactions. Structures 2 and 3 are the S- and N-bonded ones with C(3)h and D3h symmetries, respectively. The Cu–S bond lengths are shorter than for structure 1, independent of the level of theory. The S–C bond length for structure 3 becomes shorter (ca. 0.03 Å) than for structure 2 and the thiocyanate ion. Structure 4 presents two N- and one S-bonding to the copper(I) cation.The most characteristic structural feature is the increase for the S–Cu bond length (ca. 0.08 Å) at the B3LYP and N–S–Cu angle (48) at the MP2 level compared with structure 2. Moreover, a decrease in the N–Cu distance (ca. 0.03 Å) is also observed compared with structure 3. Structures 5–7 are the three remaining isomers with two S and one N bonded to copper. Structure 6 shows two signifi- cantly diVerent Cu–S bond lengths (the diVerence is ca. 0.1 Å at both levels), while 5 and 7 present Cu–S distances with intermediate values ca. 2.33 and 2.38 Å at the B3LYP and 2.22 and 2.27 Å at the MP2 levels. The energy values for the structures 1–7 are listed in Table 2, together with the respective dipole moments. At both theoretical levels, 3, the purely N-bonded isomer, was the most stable structure, whereas 1 and 2 with three S bonded to copper had the highest energy (in the range 3.8–6.9 kcal mol21 at MP2 level). However, structures 5–7, with two S and one N bonds, present similar energies (in the range 2.5–4.9 kcal mol21 at the MP2 level).The remaining mixed complex (structure 4) has one S and two N bonded to copper, and is closer in energy to structure 3 (ca. 1.4 kcal mol21). The same trend was observed at the B3LYP level but with an increase in the relative barriers by ca. 4–14 kcal mol21. Based upon the discussion above, the theoretical results favour N- to S-bonding in the isolated isomers.B Bonding nature Since a copper(I) cation has a d10s0 electron configuration there are no empty d orbitals available to form covalent bonds with either sulfur or nitrogen atoms. Instead, the s and p orbitals are conceivable. In terms of the valence-bond approach, the planar three-co-ordination can be described as a result of sp2 hybridization. To get an insight into the interactions between the central copper(I) cation and a ligand we made use of the AIM theory.49,50 The bond critical points (BCPs) on the charge density r(r) for the structures 1, 2, 3 and 5 have been calculated, and the numerical values are listed in the Table 3.The diVerent BCPs are further characterized by their values of charge density r(r), Laplacian of charge density ,2r(r), ellipticity e and local elec-J. Chem. Soc., Dalton Trans., 1999, 489–496 493 Table 4 AIM Summary of the critical points (3,13) for Cu and (3,23) for S, C and N, in 2,2r(r) for the structures 1, 2, 3 and 5 at the B3LYP theoretical level indicating the values of r(r), its Laplacian and its relative orientation Structure 1 2 3 5 Central atom (type) Cu (3,13) Cu (3,13) S(1) (3,23) S(1) (3,23) S(1) (3,23) C(1) (3,23) N(1) (3,23) Cu (3,13) Cu (3,13) S (3,23) S (3,23) S (3,23) C (3,23) N (3,23) Cu (3,13) Cu (3,13) N (3,23) N (3,23) C (3,23) S (3,23) Cu (3,13) Cu (3,13) Cu (3,13) S(1) (3,23) S(1) (3,23) S(1) (3,23) S(2) (3,23) N(1) (3,23) N(2) (3,23) N(2) (3,23) C(1) (3,23) C(2) (3,23) Bond Cu–S(1) Cu S(1)–C(1) S(1)– S(1)–Cu C(1)–N(1) N(1)– Cu–S Cu S–C S–Cu S C–N N– Cu–N Cu N–Cu N–C C–S S– Cu–S(1) Cu–N(2) Cu S(1)–C(1) S(1)–Cu S(1) S(2) N(1)– N(2)–Cu N(2)–C(2) C(1)–N(1) C(2)–N(2) r(r) (e/a0 3) 1.61 1.65 0.29 0.18 0.20 0.51 0.54 1.61 1.65 0.29 0.20 0.18 0.51 0.54 1.61 1.64 0.53 0.51 0.30 0.18 1.62 1.61 1.64 0.29 0.20 0.18 0.18 0.54 0.53 0.51 0.51 0.51 ,2r(r) (e/a0 5) 19.01 18.37 20.86 20.53 20.64 21.74 22.29 19.02 18.36 20.86 20.64 20.53 21.74 22.28 19.29 18.23 22.06 21.68 20.90 20.51 18.97 19.55 18.30 20.86 20.63 20.53 20.51 22.30 22.02 21.69 21.75 21.69 Orientation a Plane 0.883 au Vertical 0.877 au Plane 2.264 au Out of plane 1.296 au Plane 1.294 au Linear 1.468 au Linear 0.737 au Plane 0.883 au Vertical 0.877 au Plane 2.266 au Plane 1.294 au Out of plane 1.296 au Linear 1.467 au Linear 0.737 au Plane 0.883 au Vertical 0.875 au Linear 0.747 au Linear 0.828 au Linear 0.920 au Plane 1.300 au Plane 0.881 au Plane 0.878 au Vertical 0.877 au Plane 2.263 au Plane 1.294 au Out of plane 1.296 au Plane 1.300 au Linear 0.737 au Linear 0.748 au Linear 0.828 au Linear 1.465 au Linear 1.448 au NECPb 32 2 32 2 32 2212 22 a Refers to a position of a critical point and its distance (in atomic units) from a central atom. “Plane” refers to a molecular plane.“Vertical” means a perpendicular direction from a molecular plane with a central atom as starting point. An “out-of-plane” critical point is situated at a nonperpendicular position at a given distance from a central atom.A “linear” position refers to a critical point pointing away from a central atom along a given bond. b Number of equivalent critical points. tronic energy density Ed(r), all in atomic units. Three BCPs are found in the surroundings of the copper atom, located nearby the ligand atoms in every structure (see Table 3). These BCPs display similar overall characteristics. They present small r(r) values (0.05–0.10 e/a0 3), and medium and positive values of ,2r(r) (ca. 0.1–0.5 e/a0 5), which is expected for a closed shell bonding interaction (ionic). However, the BCPs present signifi- cant diVerences depending on their bonding mode (see Table 3). Table 4 gives a summary of the critical points (3,13) for Cu and (3,23) for S, C and N in ,2r(r) for the structures 1, 2, 3 and 5. Each point is characterized by its r(r) value, its Laplacian and its relative orientation. The copper atoms have five (3,13) critical points (minimum of charge concentration), three of them in the molecular plane and directed towards the ligands and two in the axial positions, compatible with the sp2 hybridization and with the non-hybridized pz orbital of the copper(I) ion.The charge concentration maxima (3,23) corresponding to the ligand atoms bonded to copper are directed towards the copper charge depletion. The nature of the bonds to S and N is characterized by the number and orientation of the charge concentration maxima of the ligand atoms bonded to copper.The sulfur atoms present four critical points (3,23). Two of them are on the molecular plane and directed towards copper and carbon, respectively (although the second one is closer to C than S). The remaining two are out of the co-ordination plane completing a distorted tetrahedron around sulfur. These latter critical points correspond to the electron lone pairs of sulfur. The orientation of the four critical points on sulfur restrains direct linear interaction between sulfur and copper.As for the bonding at N, there are two (3,23) charge concentration maxima in a linear direction, one towards Cu and the other one towards C forming a straight line Cu–BCP–N–BCP– C, also compatible with a larger charge concentration towards the copper (see Table 4). Furthermore, the S-bonded compounds have extra charge concentration in the axial positions over and below the molecular plane.In structure 5 the charge ‡ for the central copper atom is 0.72, which means that the coordination bonds between the donor atoms and the acceptor are almost ionic. The charges for the co-ordinated and non-coordinated sulfur atoms are 20.09 and 20.07, respectively. The corresponding values for nitrogen are 21.24 and 21.10. These values show clearly that the co-ordination has only a very marginal eVect on the charge distribution in the ligands. Moreover, it seems that the bond between copper and a donor atom is essentially electrostatic.Structures 1–7 have thiocyanate anions N- or S-bonded to copper. The electronic properties of the thiocyanate anion can be described with the help of two canonical forms 2N]] C]] S and 2S–C]] ] N. From the electronic characteristics calculated and presented in Tables 3 and 4, the latter canonical form is the main one in both N- and S-bonded ligands. This is compatible with the existence of only one maxima concentration on the nitrogen atom in a linear disposition, irrespective of the co-ordination mode.The electronic properties of the S–C BCPs in the N-bonded complexes show small deviations compatible with some double bond character (see Table 3). This is in accord with the calculated S–C bond length (ca. 1.67 Å for S-bonded and 1.64 Å for the N-bonded complexes). ‡ Bader atomic charges were calculated with the PROAIM program within the AIMPAC series of programs, see ref. 48.494 J.Chem. Soc., Dalton Trans., 1999, 489–496 Table 5 Comparison of the most important vibrational frequencies for the thiocyanate ion and structures 1–7 at the B3LYP/3-211G*//B3LYP/3- 211G* level 1 (Cs) 2 [C(3)h] 3 (D3h) 4 (Cs) No. 123456789 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Frequency a 27(A0) 39(A9) 40(A9) 45(A0) 66(A0) 73(A9) 98(A9) 102(A0) 109(A9) 193(A9) 209(A9) 271(A9) 495(A9) 497(A0) 503(A0) 508(A0) 511(A9) 513(A9) 720(A9) 723(A9) 729(A9) 2116(A9) 2123(A9) 2130(A9) Intensity b 193 11 0 19 84 12 15 1 16 1115125 11 4 579 80 468 Frequency a 40(E9) 40(E9) 41(A0) 42(E0) 42(E0) 75(A9) 89(A0) 104(E9) 104(E9) 205(A9) 241(E9) 241(E9) 500(E0) 500(E0) 500(A0) 501(A9) 505(E9) 505(E9) 720(E9) 720(E9) 721(A9) 2119(E9) 2119(E9) 2122(A9) Intensity b 66 13 0004 15 15 0 16 16 007022990 559 559 0 Frequency a 22(E9) 22(E9) 39(A20) 89(A29) 116(E0) 116(E0) 132(E9) 132(E9) 197(A19) 228(A20) 252(E9) 252(E9) 514(A29) 518(E9) 518(E9) 531(E0) 531(E0) 540(A20) 798(E9) 798(E9) 805(A19) 2135(E9) 2135(E9) 2147(A19) Intensity b 2250009909 31 31 033005 62 62 0 753 753 0 Frequency a 24(A9) 32(A9) 43(A0) 57(A0) 78(A9) 129(A9) 131(A0) 139(A9) 194(A9) 210(A0) 217(A9) 288(A9) 497(A9) 501(A0) 511(A9) 515(A9) 533(A0) 544(A0) 724(A9) 803(A9) 813(A9) 2115(A9) 2137(A9) 2145(A9) Intensity b 4380 11 40 19 06 17 35 0206136 46 34 467 814 220 5 [C(2)v] 6 (Cs) 7 [C(2)v] Thiocyanate ion (C•v) c 123456789 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Frequency a 27(A9) 39(A9) 43(A0) 44(A0) 74(A0) 79(A9) 106(A9) 135(A9) 161(A0) 202(A9) 231(A9) 248(A9) 497(A0) 499(A0) 504(A9) 505(A9) 513(A9) 528(A0) 722(A9) 722(A9) 800(A9) 2120(A9) 2127(A9) 2137(A9) Intensity b 71092 15 8350 35 18 40014191 32 98 949 287 Frequency a 29(A9) 30(A9) 44(A0) 50(A0) 78(A0) 78(A9) 100(A9) 142(A9) 176(A9) 192(A9) 218(A9) 287(A9) 498(A9) 502(A0) 508(A0) 510(A9) 515(A9) 534(A0) 719(A9) 725(A9) 812(A9) 2114(A9) 2122(A9) 2142(A9) Intensity b 467417 17 10 48 11 26 12204267 36 526 337 496 Frequency a 27(A9) 33(A0) 38(A9) 66(A0) 68(A9) 70(A0) 99(A9) 135(A9) 184(A0) 200(A9) 202(A9) 289(A9) 492(A9) 505(A9) 506(A0) 508(A0) 515(A9) 543(A0) 722(A9) 727(A9) 819(A9) 2117(A9) 2121(A9) 2145(A9) Intensity b 1890 24 4020 409 29 100532 14 1 33 804 2 557 Frequency a 735 465 2058 Description nSC dSCN nCN a Wavenumber in cm21 and symmetry in parentheses.b Intensity in kM mol21. c Experimental data from refs. 58 and 59. All of the above considerations are in accordance with a small contribution of the 2N]] C]] S canonical form.It is also possible to draw some conclusions from Table 4 about the stability of the S- or N-bonded isomers in the presence of a polar solvent (hydrogen bonding). For the structure 3 the absolute maximum of non-bonding concentration [r(r) = 0.5 e/a0 3 and ,2r(r) = 22.1 e/a0 5] of the ligand (corresponding to the nitrogen lone electron pair) is oriented towards the charge concentration minimum of Cu producing a strong electrostatic stabilization.Moreover, the extra negative charge is located on the sulfur atoms (large and polarizable ones), in two small charge concentration maxima [r(r) = 0.18 and ,2r(r) = 20.51 e/a0 5] on the molecular plane. This results in a more dispersed charge concentration which is also more stable under apolar conditions with less propensity for forming strong hydrogen bonds. On the other hand, the S-bonded structures present the absolute charge concentration maxima (corresponding to the non-bonding lone electron pair on the N) in only one linear CP [r(r) = 0.5 e/ao 3 and ,2r(r) = 22.3 e/a0 5], which favours the formation of stable hydrogen bonds.C Vibrational spectra The vibrational spectrum aVords another window into what is happening to the electronic structure on complexation. Values for the nSC and nCN stretching and dSCN bending modes of theJ. Chem. Soc., Dalton Trans., 1999, 489–496 495 thiocyanate anion alone and within the copper(I) complexes, are listed in Table 5. Kivekäs et al.3 reported two experimental IR bands at 2107 and 2093 cm21 for bis[6-amino- 5-(2-ethylphenylazonium)-1,3-dimethyluracil] tri(thiocyanato- S)cuprate(I). As noted before, their structure resembles the structure 1.There are two calculated strong bands for this structure at 2130 and 2116 cm21. The diVerence between the calculated and observed frequencies may well be due to hydrogen bonding in the solid state.Hydrogen bonding is known to cause shifting to lower wavenumbers.60 In their earlier paper, Kivekäs et al. suggested that the origin of the two bands is in the interaction between the central copper(I) cation and a neighbouring carbon atom of a thiocyanate group. However, we claim that the existence of two strong bands is due to two diVerent orientations of the thiocyanate groups. The thiocyanate anions S(2)– C(2)–N(2) and S(3)–C(3)–N(3) are mutually similar in contrast to the remaining S(1)–C(1)–N(1) (see Fig. 1). Indeed, when the thiocyanate anions are identical, as required by symmetry, there is only one C–N stretching band, as for structures 2 and 3. However, the proposed interaction between a copper(I) cation and a neighbouring carbon atom cannot be totally ruled out. 4 Conclusion The structures of seven diVerent copper(I) thiocyanate complexes (S and/or N bonded) were optimized by MP2 and B3LYP methods. The resulting geometry of structure 1 is in agreement with the corresponding crystallographic data at both levels of theory (including polarization functions on carbon, nitrogen and sulfur atoms).All the theoretical results yielded planar geometries with slight bending of the S–C–N angle (ca. 1758) for the S-bonded thiocyanate anions. The results also indicate that in the gas phase N-bonding is preferred to S-bonding. The existence of the pure S-bonded isomer in the solid state may be explained by preferred hydrogen bonding to the N atoms.The existence of hydrogen bonding in the solid state is also supported by the comparison of the observed and calculated nCN frequencies. The charge densities at the BCPs fall merely into three categories representing the Cu–X, S–C and C–N bonds. The ranges for these bonds are 0.05–0.09, 0.20–0.21 and 0.43–0.46 e/a0 3, respectively. In the S-bonded isomers there is a formal co-ordinate bond between copper and sulfur. However, the bond is strongly polarized and almost of ionic nature.This suggests easy deformations in the bond lengths and angles around a copper(I) cation. These deformations manifest themselves in the wide ranges of the experimental parameters obtained from CSDS. According to the AIM analyses, the thiocyanate anion prefers the canonical form of 2S–C]] ] N in both S- and N-bonded forms, however when it is N-bonded some 2N]] C]] S contribution is expected. The charge density depletion around the central copper(I) cation is in accordance with sp2 hybridization.Finally, we note that the geometrical parameters are better reproduced by MP2 than B3LYP methods. Furthermore, the AIM theory is a convenient tool to analyse quantitatively the electronic properties in metal complexes. 5 Acknowledgements Support by CIMO and Magnus Ehrnrooth fellowships to J. A. D. and the Neste foundation to J. M. are gratefully acknowledged. Computing time has been provided by the Centre of Scientific Computing, Espoo (Finland) and by the Universidad de Granada, Granada (Spain).We are grateful to Professor R. W. F. Bader for a copy of the AIMPAC package and Professor P. L. A. Popelier for a copy of the MORPHY program. References 1 P. G. Eller, D. C. Bradley, M. B. Hursthouse and D. W. Meek, Coord. Chem. Rev., 1977, 24, 1. 2 R. Kivekäs, J. Ruiz and E. Colacio, Acta Chem. Scand., 1994, 48, 382. 3 R. Kivekäs, M. Klinga, J. Ruiz and E. Colacio, Acta Chem. Scand., 1995, 49, 305. 4 M. S. Weininger, G. W. Hunt and E. L. Amma, J. Chem. Soc., Chem. Commun., 1972, 1140. 5 L. P. Battaglia, A. B. Corradi, M. Nardelli and M. E. V. Tani, J. Chem. Soc., Dalton Trans., 1976, 143. 6 M. B. Cingi, A. M. M. Lanfredi, A. Tiripicchio and M. T. Camellini, Acta Crystallogr., Sect. B, 1977, 33, 3772. 7 D. Coucouvanis, D. Murphy and S. K. Kanodia, Inorg. Chem., 1980, 19, 2993. 8 S. C. Kokkou, S. Fortier, P. J. Rentzeperis and P. Kara-Giannidis, Acta Crystallogr., Sect. C, 1983, 39, 178. 9 E. R. Atkinson, D. J. Gardiner, A. R. W. Jackson and E. S. Raper, Inorg. Chim. Acta, 1988, 98, 35. 10 E. S. Raper, Coord. Chem. Rev., 1985, 61, 115. 11 L. E. McCanlish, E. C. Bissell, D. Coucouvanis, J. P. Fackler and K. Knox, J. Am. Chem. Soc., 1968, 90, 7357. 12 A. I. Crumbliss, L. J. Gestaut, R. C. Rickard and A. T. McPhail, J. Chem. Soc., Chem. Commun., 1974, 545. 13 I. G. Dance, J. Chem. Soc., Chem. Commun., 1976, 68. 14 I. G. Dance, J. Chem. Soc., Chem. Commun., 1976, 103. 15 E. H. GriYth, G. H. Hunt and E. L. Amma, J. Chem. Soc., Chem. Commun., 1976, 432. 16 F. J. Hollander and D. Coucouvanis, J. Am. Chem. Soc., 1977, 99, 6268. 17 D. M. J. Goodgame, G. A. Leach, A. C. Skapski and K. A. Woode, Inorg. Chim. Acta, 1978, 31, 375. 18 J. R. Nicholson, I. L. Abrahams, W. Clegg and C. D. Garner, Inorg. Chem., 1985, 24, 1092. 19 M. Kabesova, M. Dunaj-Jurco, M. Serator, J. Gazo and J. Garaj, Inorg. Chim. Acta, 1976, 17, 161. 20 D. L. Smith and V.I. Saunders, Acta Crystallogr., Sect. B, 1982, 38, 907. 21 C. L. Raston, B. Walter and A. H. White, Aust. J. Chem., 1979, 32, 2757. 22 P. C. Healy, C. Pakawatchai, R. I. Papasergio, V. A. Patrick and A. H. White, Inorg. Chem., 1984, 23, 3769. 23 G. O. Morpurgo, G. Dessy and V. Fares, J. Chem. Soc., Dalton Trans., 1984, 785. 24 R. G. Pearson, J. Am. Chem. Soc., 1963, 85, 3553. 25 R. D. Hancock and A. E. Martell, J. Chem. Educ., 1996, 73, 654. 26 S. Gambarotta, M. L. Fiallo, C.Floriani, A. Chiesi-Villa and C. Guastini, Inorg. Chem., 1984, 23, 3532. 27 S. Kawaguchi, Variety in Co-ordination Modes of Ligands in Metal Complexes in Inorganic Chemistry Concepts, 11, Springer, Berlin, 1988. 28 M. Kabesova, R. Boca, M. Melnik, D. Valigura and M. Dunaj- Jurco, Coord. Chem. Rev., 1995, 140, 115. 29 A. H. Lewin, R. J. Michl, P. Ganis and U. Lepore, Chem. Commun., 1976, 312. 30 B. A. Fields, H. H. Bartsch, H. D. Bartunik, F. Cordes, J. M. Guss and H. C. Freeman, Acta Crystallogr., Sect.D, 1994, 50, 709. 31 L. Stryer, Biochemistry, W. H. Freeman, New York, 3th edn., 1988, pp. 524–526. 32 R. Sousa, Acta Crystallogr., Sect. D, 1995, 51, 271. 33 R. G. Parr and W. Yang, Density functional theory of atoms and molecules, Oxford University Press, Oxford, 1989. 34 J. Labanowski and J. Andzelm, Theory and Applications of Density Functional Approaches to Chemistry, Springer, Berlin, 1990. 35 T. Ziegler, Chem. Rev., 1991, 91, 651. 36 C. Daul, E.J. Baerends and P. Vernooijs, Inorg. Chem., 1994, 33, 3558. 37 T. Ziegler and J. Li, Organometallics, 1995, 14, 214. 38 J. Li, G. Schreckenbach and T. Ziegler, J. Am. Chem. Soc., 1995, 117, 486. 39 D. A. Estrin, O. Y. Hamra, L. Paglieri, L. Slep and J. A. Olabe, Inorg. Chem., 1996, 35, 6832. 40 J. A. R. Navarro, M. A. Romero, J. M. Salas, M. Quiros, J. E. Bahraoui and J. Molina, Inorg. Chem., 1996, 35, 7829. 41 M. R. Sundberg, R. Sillanpää and R. Uggla, Inorg. Chim. Acta, 1996, 245, 35. 42 M. R. Sundberg and R. Uggla, Inorg. Chim. Acta, 1997, 254, 259. 43 M. F. Fan, G. C. Jia and Z. Y. Lin, J. Am. Chem. Soc., 1996, 118, 9915. 44 J. A. Platts, S. T. Howard and B. R. F. Bracke, J. Am. Chem. Soc., 1996, 118, 2726.496 J. Chem. Soc., Dalton Trans., 1999, 489–496 45 C. Heinemann, T. Muller, Y. Apeloig and H. Schwarz, J. Am. Chem. Soc., 1996, 118, 2023. 46 J. A. Dobado, H. Martínez-García, J. Molina and M. R. Sundberg, J. Am. Chem. Soc., 1998, 120, 8461. 47 M. J. Frisch, G. W. Trucks, H. B. Schlegel, P. M. W. Gill, B. G. Johnson, M. A. Robb, J. R. Cheeseman, T. Keith, G. A. Petersson, J. A. Montgomery, K. Raghavachari, M. A. Al-Laham, V. G. Zakrzewski, J. V. Ortiz, J. Foresman, B. B. Cioslowski, A. Stefanov, M. Nanayakkara, J. B. Challacombe, C. Y. Peng, P. Y. Ayala, W. Chen, M. W. Wong, J. L. Andres, E. S. Replogle, R. Gomperts, R. L. Martin, D. J. Fox, J. S. Binkley, D. J. Defrees, J. Baker, J. P. Stewart, M. Head-Gordon, C. Gonzalez and J. A. Pople, GAUSSIAN 94, revision C.2, Gaussian Inc., Pittsburgh, PA, 1995. 48 F. W. Biegler-Köning, R. F. W. Bader and T. H. Tang, J. Comput. Chem., 1982, 3, 317. 49 R. F. W. Bader, Atoms in Molecules: a Quantum Theory, Clarendon Press, Oxford, 1990. 50 R. F. W. Bader, Chem. Rev., 1991, 91, 893. 51 F. H. Allen and O. Kennard, Chem. Des. Automat. News, 1993, 8, 31. 52 F. H. Allen and O. Kennard, Chem. Des. Automat. News, 1993, 8, 1; CSDS V5.15, April 1998 release. 53 M. N. Hughes, Chemistry and Biochemistry of Thiocyanate Acid and its Derivatives, Academic Press, New York, 1975. 54 L. Tchertanov and C. Pascard, Acta Crystallogr., Sect. B, 1996, 52, 685. 55 J. L. Burmeister, R. L. Hassel and R. L Phelan, Inorg. Chem., 1971, 10, 2032. 56 D. F. Gutterman and H. B. Gray, J. Am. Chem. Soc., 1975, 15, 91. 57 N. Fukushima, G. Iisaka, M. Saito and K. Waizumi, Inorg. Chim. Acta, 1997, 255, 211. 58 N. B. Colthup, L. H. Daly and S. E. Wiberley, Introduction to Infrared and Raman Spectroscopy, Academic Press, Boston, 1990. 59 K. Nakamoto, IR and Raman Spectra of Inorganic and Coordination Compounds, Wiley, New York, 1978. 60 M. R. Sundberg, M. Klinga and R. Uggla, Inorg. Chim. Acta, 1994, 216, 57. Paper 8/06424E

ISSN:1477-9226    

DOI:10.1039/a806424e

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1999, 497–507 497 Formation of organometallic hydroxo and oxo complexes by oxidation of transition metal hydrides in the presence of water. X-Ray structures of [CpMo(OH)(PMe3)3][BF4] and [CpMo(O)(PMe3)2][BF4] James C. Fettinger,a Heinz-Bernhard Kraatz,a† Rinaldo Poli *b and E. Alessandra Quadrelli a a Department of Chemistry and Biochemistry, University of Maryland, College Park, Maryland 20742, USA b Laboratoire de Synthèse et d’Electrosynthèse Organometalliques, Faculté des Sciences “Gabriel”, Université de Bourgogne, 6 Boulevard Gabriel, 21100 Dijon, France.E-mail: Rinaldo.Poli@u-bourgogne.fr Received 11th September 1998, Accepted 1st December 1998 The decomposition of complex [CpMoH(PMe3)3]1, 11, in wet, weakly coordinating solvents (THF and acetone) leads to the formation of the Mo(III) hydroxo compound [CpMo(OH)(PMe3)3][BF4], 2, and the Mo(IV) oxo compound [CpMo(O)(PMe3)2][BF4], 3. Both of these have been characterized by X-ray crystallography. The formation of these products is rationalized by a mechanism which involves water coordination followed by multiple steps of tandem oxidation and deprotonation reactions.The proposed [CpMoH(PMe3)3(H2O)]21 intermediate is observed by NMR spectroscopy. EPR monitoring of the oxidation of 1 with ferrocenium shows also the formation of a species characterized by a binomial triplet resonance and interpreted as the neutral Mo(III)–oxo complex CpMo(O)(PMe3)2, 4.The proposed mechanism is in harmony with the previously investigated mechanism of decomposition of 11 in dry solvents and with the coupled proton and electron transfer processes that relate aqua, hydroxo, and oxo species in the chemistry of Mo enzymes. Computational studies at the DFT level were carried out on a model system where the PMe3 ligand was replaced by PH3. Their results are consistent with the proposed mechanism. Introduction The oxidation of transition metal hydride complexes and the reactivity of the resulting oxidation products has recently been the subject of intensive and detailed investigations.1–36 Oxidation of typical 18-electron neutral hydride complexes, M–H, dramatically increases the acidity of the hydride ligand in the resulting 17-electron products [M–H]1, resulting in its rapid decomposition by deprotonation (see Scheme 1).16 The cationic hydride can be deprotonated either by an external base or by the unoxidized starting compound, resulting in diVerent oxidation stoichiometries [equations (1)–(3)].In particular, M–H 1 S 1 B æÆ [M(S)]1 1 BH1 2 2e2 (1) 2 M–H 1 S æÆ [M(S)]1 1 [MH2]1 2 2e2 (2) 2 M–H 1 2S æÆ 2[M(S)]1 1 H2 2 2e2 (3) deprotonation by an external base [path a, equation (1)] leads to the consumption of 2 oxidizing equivalents per M–H molecule, while deprotonation by M–H [path b, equations (2) and (3)] leads to the consumption of 1 oxidizing equivalent per M–H Scheme 1 M-H [ox] [M-H]+ M.[ox], S [M(S)]+ B BH+ M-H [MH2]+ +S, -H2 (a) (b) † Current address: Department of Chemistry, University of Saskatchewan, Saskatoon, SK S7N 5C9, Canada. molecule. The diVerence between equations (2) and (3) is related to the susceptibility of the protonated [MH2]1 product to lose H2 in favor of coordination of solvent (S). In some cases, the oxidation stoichiometry for the same hydride material depends upon whether the complex is oxidized chemically or electrochemically.16,26 This phenomenon has been attributed to the presence of adventitious water as an external base in the electrochemical cell.We have recently studied in detail the influence of water on the oxidation and protonation of compounds CpMoH(CO)2L (L = PMe3, PPh3) and described the general phenomenon of the action of an external base as a “proton shuttle”.37 In essence, when B is a stronger base than M–H, equation (1) is thermodynamically favored over equations (2) and (3), but if the less favored protonation of M–H is followed by irreversible H2 loss, then the BH1 product delivers its proton to unoxidized M–H, changing the stoichiometry to that of equation (3).However, deprotonation is not the only decomposition pathway available to oxidized hydride complexes. When good donor and/or sterically encumbering ligands are coordinated to the metal center in the precursor hydride complex M–H, the acidity (either kinetic or thermodynamic) of [M–H]1 is not suYciently high, allowing further oxidation with concomitant solvent (S) coordination, to yield [M–H(S)]21 as an intermediate of a decomposition mechanism by disproportionation (Scheme 2).30 This dicationic hydride can then be deprotonated, leading to Scheme 2 M-H [ox], S [M(S)H]2+ [M(S)]+ B BH+ M-H [MH2]+ +S, -H2 (a) (b) [ox] [M-H]+498 J.Chem. Soc., Dalton Trans., 1999, 497–507 Table 1 NMR data Complex [CpMoO(PMe3)2]1 [CpMoH(PMe3)3(H2O)]21c [CpMoH(PMe3)3(H2O)]21c [CpMoH(PMe3)3(CD3COCD3)]21 [CpMoH(PMe3)3(CD3COCD3)]21 Solvent CD3CN CD3COCD3 CD3CN CD3COCD3 CD3CN 1H NMR (d) 5.53 (s, 5 H, Cp) 1.63 (d, 18 H, JPH = 9.6 Hz, PMe3) 5.26 (m, 5 H, Cp) 1.59 (d, 27 H, JHP = 8.8 Hz, PMe3) 5.09 (s, 5 H, Cp) d 5.26 (m, 5 H, Cp) 1.67 (d, 27 H, JHP = 5.6 Hz, PMe3) 5.06 (s, 5 H, Cp) d 31P{1H} NMR (d) 13.3 (s) a 10.6 (t, 1 P, JPP = 25.9 Hz) b 2.8 (d, 2 P, JPP = 24.3 Hz) b 10.7 (t, 1 P, JPP = 25.1 Hz) e 2.8 (d, 2 P, JPP = 24.3 Hz) e 10.4 (t, 1 P, JPP = 22.7 Hz) 4.8 (d, 2 P, JPP = 24.3 Hz) a d 13.2 in CD3COCD3; the 1H NMR of [CpMoO(PMe3)2]1 has been previously reported.41 b Splits further into a doublet (JPH = 12 Hz) in the 31P{selective-1H} NMR.c The hydride and water resonances are not visible in the 1H NMR. d The PMe3 resonance could not be identified because of extensive overlap with the other reaction products (see Experimental section). e Splits futher into a doublet (JPH = 14 Hz) in the 31P{selective-1H} NMR.the same possibilities of oxidation stoichiometries encountered for the alternative decomposition mechanism of Scheme 1. While, in most cases, the [M–H]1 and [M(S)H]21 intermediates are too unstable to be spectroscopically observed, the oxidation of the electron-rich and sterically crowded CpMoH(PMe3)3 complex, 1, has allowed the formation and spectroscopic observation of the relatively stable one-electron oxidation product, 11, and the isolation and single crystal X-ray analysis of [CpMoH(PMe3)3(MeCN)]21.38 The steric bulk and donor power of the ancillary ligands in this system are suf- ficiently great to slow down the deprotonation process of both 11 and [1(MeCN)]21 by either 1 or NEt3.In this contribution, we examine the intervention of water as a coordinative molecule (S in Scheme 2) during the oxidation of compound 1. Since free water is not suYciently basic to engage in proton transfer reactivity for this system (i.e. as B in Schemes 1 and 2), new reactivity pathways are made possible by the presence of acidic protons on the coordinated water ligand.An unprecedented paramagnetic organometallic hydroxo derivative of Mo(III), its conjugate base, and a previously reported organometallic oxo derivative of Mo(IV) are identified as the products of this process. The geometry and energies of model compounds for the Mo(III) and Mo(IV) complexes have been calculated through geometry optimizations and have been used to draw the energetic profile of the transformations.Part of this work has previously been communicated.39 Experimental General All operations were carried out under an atmosphere of dinitrogen using standard glove-box and Schlenk-line techniques. Solvents were dehydrated by standard methods (THF from Na/K/benzophenone, heptane from Na and acetone from MgSO4), deoxygenated, and distilled directly from the dehydrating agent under dinitrogen prior to use. The amount of residual water in the solvents was assessed by titration with a coulomatic Karl–Fischer titrimeter (Fisher Scientific).Complex CpMoCl- (PMe3)3 was prepared as previously described.40 Spectroscopic analyses Samples for 1H and 31P NMR in thin-walled 5 mm glass tubes were measured on a Bruker WP200 spectrometer. The 1H NMR spectra were calibrated against the residual proton signal of the deuterated solvents. The 31P NMR were calibrated against 85% H3PO4 in a capillary tube which was placed in a diVerent 5 mm glass tube, containing the same deuterated solvent used for the measurement.All NMR spectra are collected in Table 1. EPR measurements were carried out at the X-band microwave frequency on a Bruker ER 200 D spectrometer upgraded to ESP 300, equipped with a cylindrical ER/4103 TM 110 cavity. Oxidation of 1 by Fc1PF6 2 in CD3CN in presence of water The CD3CN solution (0.753 mL) resulting from the addition of a solution (0.2 mL) of FcPF6 (18.2 mg, 55 mmol) to a solution (553 mL) containing 1 (56 mmol) and H2O (3 mL, 160 mmol) was monitored over time and compared to the analogous reaction performed without the addition of water.38 The 1H and 31P NMR spectra confirmed the previously reported formation of [CpMoH(CD3CN)(PMe3)3]21, [CpMoH2(PMe3)3]1 and [CpMo- (CD3CN)(PMe3)3]1.38 In addition, however, the spectra also revealed Cp resonances of two additional diamagnetic products whose formation does not occur under dry conditions.These were attributed to [CpMoO(PMe3)2]1 (ca. 15% of all diamagnetic products by integration of the Cp resonances), and [CpMoH(PMe3)3(H2O)]21 (ca. 20%). Decomposition of 11 in THF (a) Isolation of [CpMo(PMe3)3(OH)][BF4], 2, and [CpMo- O(PMe3)2][BF4], 3. Complex 11 was generated in situ according to the literature:38,40 to a stirred solution of CpMoCl(PMe3)3 (149 mg, 0.351 mmol) in THF (10 mL) was added via microsyringe a 1 M solution of LiBHEt3 (351 mL, 0.351 mmol). After 30 min, the reaction mixture was evaporated to dryness and extracted with n-heptane (2 × 10 mL), yielding a spectroscopically pure solution of CpMoH(PMe3)3.After filtration and evaporation to dryness, the residue was dissolved in THF and cooled to 278 8C. Solid AgBF4 (68 mg, 0.350 mmol) was added, resulting in the immediate formation of a dark brown precipitate. An aliquot of this reaction mixture was transferred into an EPR tube via cannula and frozen in liquid nitrogen until the introduction into the EPR probe, which was preset at 278 8C.The resulting EPR spectrum (doublet of quartets [g = 2.017; a(P) = 28.5 G; a(H) = 13 G]) confirmed the formation of complex 11, by comparison with the previously reported spectrum.38 The reaction mixture was allowed to warm to room temperature, filtered and evaporated to dryness under reduced pressure. This solid was extracted with acetone (3 × 10 mL) and the resulting orange solution was filtered, reduced in volume to about 10 mL, and the product was precipitated out by the addition of heptane (20 mL).Recrystallization of the dark solid from acetone–heptane led to the deposition of three types of crystals: red prisms, yellow blocks and yellow needles, which were all used for single crystal X-ray analyses. The yellow blocks and yellow needles correspond to [CpMo- (PMe3)3(OH)][BF4], 2, and to [CpMo(PMe3)3(OH)1 2 xClx]- [BF4], respectively, whereas the red prisms correspond to [CpMoO(PMe3)2][BF4], 3.A few red crystals were separated by hand-picking and investigated by NMR spectroscopy. The 1H NMR spectrum in CD3COCD3 is identical with that previously reported for compound 3.41 (b) Spectroscopic study of the reaction mixture. A reactionJ. Chem. Soc., Dalton Trans., 1999, 497–507 499 Table 2 Crystal data for all compounds a Compound Formula Mw Space group a/Å b/Å c/Å V/Å3 Reflections collected, unique Rint m(Mo-Ka)/mm21 R indices [I > 2s(I)] R1b wR2c R indices (all data) R1 wR2 2 C14H33BF4MoOP3 493.06 Pna21 20.525(5) 13.380(4) 7.9740(7) 2189.9(9) 3712, 3712 — 0.851 0.0267 0.0567 0.0353 0.0604 0.79(2)?0.21[CpMoCl(PMe3)3][BF4] C28H65.58B2Cl0.42F8Mo2O1.58P6 993.78 P212121 8.7140(9) 13.300(4) 37.939(6) 4397(2) 3488, 3488 — 0.872 0.0509 0.1139 0.0807 0.1292 3 C11H23BF4MoOP2 415.98 Ama2 17.29(2) 15.201(2) 6.6459(6) 1747(2) 969, 692 0.0336 0.963 0.0391 0.0715 0.0782 0.0801 a Details in common: orthorhombic, Z = 4, T = 293(2) K.b R1 = S Fo| 2 |Fc /S|Fo|.c wR2 = [Sw(|Fo| 2 |Fc|)2/Sw|Fo|2]� �� . identical to that described in the previous section was carried out from 1 (0.86 mmol) and AgBF4 (164 mg, 0.84 mmol). After filtration and evaporation to dryness, the brown solid was washed with Et2O and dried under vacuum. Yield 332 mg. This solid was used for NMR and EPR investigations. EPR (CD3COCD3, 250 8C): dq (g = 2.032, ad(H) = 14.1 G; aq(P) = 28.8 G, aMo = 29 G). This spectrum is attributed to complex 11 (cf.EPR spectrum in THF).38 The 1H NMR spectrum (CD3- COCD3, d) showed four diVerent products in an approximate relative ratio of 10 : 60 : 15 : 15, as follows. [CpMoO(PMe3)2]1 (relative intensity 10),41 [CpMoH2(PMe3)3]1 (relative intensity 60):41 [CpMoH(PMe3)3(CD3COCD3)]21 (relative intensity 15), [CpMoH(PMe3)3(H2O)]21 (relative intensity 15). The CD3- COCD3 solution was evaporated to dryness and redissolved in CD3CN. The 1H NMR spectrum showed five Cp resonances, whose relative intensities changed with time. The initial relative intensities and assignments are 23 ([CpMoO(PMe3)2]1) : 55 (d 4.80, overlap of [CpMoH2(PMe3)3]1 and [CpMo(CD3CN)- (PMe3)3]1) 38:7 (d 5.31, [CpMoH(PMe3)3(CD3CN)]21) : 7 ([Cp- MoH(PMe3)3(CD3COCD3)]21) : 8 ([CpMoH(PMe3)3(H2O)]21). Over 24 h, there is no significant change in the relative intensities of the [CpMoO(PMe3)2]1, [CpMoH2(PMe3)3]1, [CpMo- (CD3CN)(PMe3)3]1 and [CpMoH(PMe3)3(H2O)]21 resonances, while the [CpMoH(PMe3)3(CD3COCD3)]21 resonance disappeared and that of [CpMoH(PMe3)3(CD3CN)]21 correspondingly increased. NMR and EPR monitoring of the decomposition of 11 in wet CD3COCD3.Formation of CpMoO(PMe3)2, 4 To a CD3COCD3 solution (0.5 mL) of CpMoH(PMe3)3 (17 mmol, generated in situ as above) was added water (1 mL, 56 mmol) and AgBPh4 (4.3 mg, 10 mmol, Mo:Ag = 1 : 0.6). The 1H NMR monitoring showed a slow reaction that eventually consumed all of the starting hydride. After 24 h, an aliquot of the solution was investigated by EPR spectroscopy: binomial triplet with Mo satellites, g = 2.000, aP = 20 G, aMo = 30 G.This resonance is assigned to CpMoO(PMe3)2, 4. At the same time, the 31P NMR spectrum revealed a broad (w1/2 = 360 Hz) resonance at d 236 and a single organometallic diamagnetic product (d 7.9) , identified as [CpMoH2(PH3)3]1 by comparison with an authentic sample.38,41 An analogous experiment was carried out with a CD3- COCD3 solution (1.0 mL) of CpMoH(PMe3)3 (100 mmol, generated in situ as above), containing 5 mL (0.25 mmol) of water.In a separate Schlenk, Ag1BF4 2 (65 mg, 0.333 mmol) was dissolved in CD3COCD3 (200 mL). A first aliquot (66 mL, 0.11 mmol, ca. 1 equiv.) was added to the hydride solution, generating the EPR spectrum assigned to 4 (see previous section). A second aliquot of silver solution (50 mL, 83 mmol) was added to the residual solution (Mo :Ag = ca. 1 : 2). The resonances of complex [CpMoO(PMe3)2]1 were present in the 1H NMR spectrum, while a subsequent EPR investigation showed no residual EPR activity.Protonation of 1 in the presence of PMe3 Upon addition of HBF4?Et2O (10 mL, 70 mmol) to a CD3COCD3 solution (0.5 mL) of 1 (70 mmol) and PMe3 (10 mmol), no changes in the 1H and 31P NMR resonances of the trimethylphosphine were detected, while the resonances of 1 were replaced by those of [1–H]1. Protonation of 1 in the presence of H2O Upon addition of HBF4?Et2O (10 mL, 70 mmol) to a CD3- COCD3 solution (0.6 mL) of 1 (79 mmol) and H2O (2 mL, 110 mmol), no major change in the 1H resonance of the water peak was detected (d 2.2), while the resonances of 1 were replaced by those of [1–H]1.X-Ray crystallography (a) [CpMo(OH)(PMe3)3][BF4]. Data collection and reduction and structure solution were routine. The structure was refined with an initial Flack parameter refining to 0.87(10) indicating that the structure needed to be inverted. Following inversion and further refinement, one of the three PMe3 groups was found to be disordered with major : minor contributors of 0.631 : 0.369.The hydroxide hydrogen was initiacated in the diVerence Fourier map at a distance of 0.88 Å from the O atom and a Mo–O–H angle of 1408, but free refinement led to instability. Thus, it was eventually refined with DFIX (restrained O–H distance and free Mo–O–H angle). Crystal data and refinement parameters are collected in Table 2, while selected bond distances and angles are listed in Table 3.(b) [CpMo(OH)0.79Cl0.21(PMe3)3][BF4]. Routine structure solution revealed two independent molecules in the asymmetric unit. During the refinement, it became apparent that one of the two molecules had disorder in the hydroxo position. Further analysis revealed that this position was composed of two diVerent ligands, OH and Cl. The ratio was determined to be OH :Cl 0.58 : 0.42 by refinement of the X-ray data. The overlap on this site required the use of restraints.The Mo–O–H moiety was imposed to be identical with that in the other independent molecule. This allowed for both the hydroxyl oxygen and chlorine atoms to refine to acceptable positions. Crystal data and refinement parameters are collected in Table 2, while selected bond distances and angles are listed in Table 4.500 J. Chem. Soc., Dalton Trans., 1999, 497–507 Table 3 Selected bond lengths (Å) and angles (8) for [CpMo(OH)(PMe3)3][BF4] 2 and comparison with B3LYP-optimized [CpMo(OH)(PH3)3]1a Mo–CNT Mo(1)–C(1) Mo(1)–C(2) Mo(1)–C(3) Mo(1)–C(4) Mo(1)–C(5) Mo(1)–P(1) Mo(1)–P(2) Mo(1)–P(3) Mo(1)–O(1) O(1)–H(1A) X-Ray 1.980(2) 2.302(4) 2.241(5) 2.271(4) 2.348(4) 2.369(4) 2.4533(12) 2.4989(12) 2.4749(11) 2.080(3) 0.831(10) B3LYP 2.054 2.025 2.302 2.474 2.572 2.632 2.557 2.050 0.979 O(1)–Mo(1)–CNT P(1)–Mo(1)–CNT P(2)–Mo(1)–CNT P(3)–Mo(1)–CNT P(1)–Mo(1)–P(2) P(1)–Mo(1)–P(3) P(2)–Mo(1)–P(3) P(1)–Mo(1)–O(1) P(2)–Mo(1)–O(1) P(3)–Mo(1)–O(1) Mo(1)–O(1)–H(1A) CNT–Mo–O–H X-Ray 112.40(13) 108.52(10) 110.63(10) 110.81(10) 94.22(5) 93.21(4) 132.83(4) 139.07(10) 71.03(8) 73.15(8) 133(4) 220(5) B3LYP 121.35 112.22 110.31 114.38 85.97 85.20 136.40 123.70 68.64 81.40 126.60 83.00 a CNT = centroid of cyclopentadienyl ring.(c) [CpMoO(PMe3)2][BF4]. Two of the PMe3 carbon atoms, C4 and C5, were found to suVer from disorder that was modeled resulting in a final ratio C4 :C4A = C5:C5A = 0.584: 0.416.The BF4 molecule was also directly located and displayed large librational motion. The final model included two orientations for the fluorine atoms with equal occupancy (50 : 50). Crystal data and refinement parameters are collected in Table 2, while selected bond distances and angles are listed in Table 5. CCDC reference number 186/1270. See http://www.rsc.org/suppdata/dt/1999/497/ for crystallographic files in .cif format. Theoretical calculations All calculations were performed using Gaussian-9442 on a SGI Origin 200 at the Université de Bourgogne and on a DEC/Alphastation 250 at the University of Maryland.The LanL2DZ set was employed to perform complete geometry optimizations with a Density Functional Theory (DFT) approach. The three-parameter form of the Becke, Lee, Yang and Parr functional (B3LYP),43 was employed. The LanL2DZ basis set includes both Dunning and Hay’s D95 sets for H and C,44 and the relativistic Electron Core Potential (ECP) sets of Hay and Wadt for the heavy atoms.44–47 Electrons outside the core were all those of H, C and O atoms, the 3s and 3p electrons in P, and the 4s, 4p, 4d and 5s electrons in Mo.The mean value of the spin of the first order electronic wavefunction, Table 4 Selected bond lengths (Å) and angles (8) for [CpMo(OH)0.79- Cl0.21(PMe3)3][BF4] Mo(1)–P(1) Mo(1)–P(2) Mo(1)–P(3) Mo(1)–O(1) Mo(1)–C(1) Mo(1)–C(2) Mo(1)–C(3) Mo(1)–C(4) Mo(1)–C(5) Mo(1)–CNT(1) P(1)–Mo(1)–P(2) P(1)–Mo(1)–P(3) P(1)–Mo(1)–O(1) P(1)–Mo(1)–CNT(1) P(2)–Mo(1)–P(3) P(2)–Mo(1)–O(1) P(2)–Mo(1)–CNT(1) P(3)–Mo(1)–O(1) P(3)–Mo(1)–CNT(1) O(1)–Mo(1)–CNT(1) Cl(2)–Mo(2)–P(4) Cl(2)–Mo(3)–P(5) 2.484(4) 2.481(4) 2.456(4) 2.078(9) 2.25(2) 2.25(2) 2.31(2) 2.36(2) 2.31(2) 1.98(2) 135.4(2) 92.6(2) 75.6(2) 108.9(5) 92.0(2) 72.5(2) 111.5(2) 139.6(3) 108.3(5) 112.1(6) 76.8(11) 70.9(10) Mo(2)–Cl(2) Mo(2)–P(4) Mo(2)–P(5) Mo(2)–P(6) Mo(2)–O(2) Mo(2)–C(15) Mo(2)–C(16) Mo(2)–C(17) Mo(2)–C(18) Mo(2)–C(19) Mo(2)–CNT(2) Cl(2)–Mo(2)–P(6) Cl(2)–Mo(2)–O(2) Cl(2)–Mo(2)–CNT(2) P(4)–Mo(2)–P(5) P(4)–Mo(2)–P(6) P(4)–Mo(2)–O(2) P(4)–Mo(2)–CNT(2) P(5)–Mo(2)–P(6) P(5)–Mo(2)–O(2) P(5)–Mo(2)–CNT(2) P(6)–Mo(2)–O(2) P(6)–Mo(2)–CNT(2) O(2)–Mo(2)–CNT(2) 2.33(3) 2.488(4) 2.499(4) 2.464(4) 2.09(3) 2.36(2) 2.348(13) 2.265(14) 2.22(2) 2.30(2) 1.98(2) 138.4(9) 7(3) 113.8(11) 135.70(13) 93.7(2) 70(2) 108.1(5) 91.0(2) 78(2) 112.3(5) 139(2) 107.6(5) 113(2) which is not an exact eigenstate of S2 for unrestricted calculations on open-shell systems, was considered suitable to unambiguously identify the spin state.Spin contamination was carefully monitored and the energies shown in the Results section correspond to unrestricted B3LYP (UB3LYP) calculations. Results Chemical oxidation of CpMoH(PMe3)3 in the presence of water The oxidation of 1 in dry CD3CN has been previously investigated. 38 We have now reexamined this process in the presence of water. The NMR monitoring shows the formation of the same products that are obtained when the solvent is dry, i.e.[1(CD3CN)]21, [1–H]1 and [CpMo(PMe3)3(CD3CN)]1. In addition, small amounts of the previously described 41 complex [CpMoO(PMe3)2]1 and another species that has very similar NMR properties to those of [1(MeCN)]21, were also present. The latter species could not be isolated, but its assignment to [1(H2O)]21 appears most reasonable. The two species [1(MeCN)]21 and [1(H2O)]21 have the same pattern of 1H and 31P NMR resonances (one Cp resonance and two PMe3 resonances in a 2 : 1 ratio with similar chemical shifts and coupling constants), and both show coupling of the 31P NMR resonances to a single hydride ligand upon selective decoupling (JHP = 10 Hz for the CD3CN adduct 38 and 14 Hz for the H2O adduct, see Table 1).Complex [1(H2O)]21 does not decompose readily in CD3CN (24 h), nor does the ratio of complexes [1(MeCN)]21 and [1(H2O)]21 change with time. On a longer time scale, however (4 days), only the oxo and dihydrido Mo(IV) complexes are observed in solution.The oxidation was also carried out in less coordinating solvents such as THF and acetone. The solid that could be isolated when the oxidation was carried out with one equivalent of oxidizing agent (Ag1) in THF shows, after dissolution in CD3COCD3, the presence of 11 by EPR spectroscopy38 and complexes [CpMoO(PMe3)2]1, [1–H]1, [1(H2O)]21 (vide supra), and a new complex assigned as [1(CD3COCD3)]21 by 1H and 31P NMR spectroscopy.Complex [1(CD3COCD3)]21 shows 1H and 31P NMR patterns similar to those of the related CD3CN and H2O derivatives (see Table 1). The corresponding complex [1(THF)]21 was previously obtained by oxidation of 1 with two equivalents of oxidant in THF,38 but its sparing solubility in THF did not permit its NMR characterization. When the isolated solid was dissolved in CD3CN, the smooth transformation of [1(CD3COCD3)]21 to [1(CD3CN)]21 could be witnessed over a few hours by NMR spectroscopy, while the resonances attributed to [1(H2O)]21 remained unchanged over the same time scale.A spectroscopic monitoring of the oxidation reaction of 1 in CD3COCD3 carried out in the presence of small amounts of water and with a substoichiometric amount of oxidant shows the formation of an EPR active species characterized by a tripletJ. Chem. Soc., Dalton Trans., 1999, 497–507 501 Table 5 Selected bond lengths (Å) and angles (8) for [CpMo(O)(PMe3)2][BF4] 3 and comparison with B3LYP-optimized [CpMo(O)(PH3)2]1 Mo(1)–CNT Mo(1)–C(1) Mo(1)–C(2) Mo(1)–C(3) Mo(1)–P(1) Mo(1)–O(1) X-Ray 2.05(2) 2.20(2) 2.298(13) 2.521(12) 2.445(4) 1.674(13) B3LYP 2.122 2.265 2.363 2.620 2.547 1.714 O(1)–Mo(1)–CNT P(1)–Mo(1)–CNT P(1)–Mo(1)–P(1A) P(1)–Mo(1)–O(1) P(1)–Mo(1)–CNT–P(1A) X-Ray 149.1(6) 106.9(3) 94.4(2) 93.7(3) 100.01 B3LYP 147.41 107.44 93.25 94.60 99.26 a CNT = centroid of cyclopentadienyl ring.resonance (aP = 20 G) flanked by Mo satellites (aMo = 28 G) in the EPR spectrum (see Fig. 1). This resonance disappears after the addition of an excess of oxidant. This EPR active intermediate is formulated as CpMo(O)(PMe3)2, 4, see Discussion. This EPR triplet must be due to an odd-electron Mo complex that contains two PMe3 ligands and no hydride ligand. Hydride ligands on Mo typically lead to large, observable coupling patterns in the EPR spectra, for instance aH = 12.9 G in 11.38 A simultaneous NMR investigation shows the formation of the dihydride cation [1–H]1 and a very broad resonance at d 236 in the 31P NMR spectrum.The latter is assigned to free PMe3, the broadness and shift relative to the usual 31P NMR resonance (d ca. 260) being attributed to chemical exchange with the paramagnetic compound 4. This is confirmed by the upfield shift and sharpening of this resonance upon addition of free PMe3 to this solution. After further oxidation, the 1H and 31P NMR resonances of [CpMoO(PMe3)2]1 become observable, in concomitance with the disappearance of the signal of 4 from the EPR spectrum.The recrystallization of 11BF4 2 from wet acetone yields three diVerent types of crystals. These were shown by X-ray crystallography (vide infra) to correspond to the BF4 2 salt of the [CpMo(O)(PMe3)2]1 complex, compound 3, the Mo(III) hydroxo species [CpMo(OH)(PMe3)3]BF4, 2, and a solid solution of the latter with [CpMoCl(PMe3)3]BF4, corresponding to the composition [CpMo(OH)0.79Cl0.21(PMe3)3]BF4.The X-ray structure of a pure sample of the chloro species was reported previously.48 The presence of the chloride impurity in this material can be rationalized on the basis of the preparative chemistry, as follows. Compound 1 is very soluble in all nonpolar solvents and quite diYcult to crystallize.40 For that reason, it was generated in situ by the action of LiBEt3H on CpMoCl(PMe3)3, this being a facile and selective reaction 40 so long as care is taken to assure a complete conversion.In the present case, a residual chloride impurity probably contaminated the isolated product 1, and its subsequent oxidation gener- Fig. 1 Room temperature EPR spectrum of a solution of 4 in CD3COCD3. ated, as previously established,40 the stable [CpMoCl(PMe3)3]1 cation which co-crystallized with compound 2. Compound 2 has also been recently obtained in our laboratory by an alternative procedure, namely one-electron oxidation of the 16-electron Mo(II) hydroxo complex, CpMo(OH)- (PMe3)2, in the presence of PMe3.39,49 A unit cell determination on single crystals confirms that the two products are identical.This compound does not lead to detectable EPR or NMR resonances. Compound 3, in turn, has previously been obtained by protonation of CpMoH(PMe)3 with aqueous HBF4.41 Other similar CpMo(IV) oxo cations, i.e. [CpMo(O)(dppe)]1PF6 2 and [CH3C(CH2–h5-C5H4)(CH2PPh2)2MoO]1,50,51 have also been previously described in the literature.X-Ray structures The previously communicated 39 X-ray structure of compound 2 is reported here in full. A top view of the cation is shown in Fig. 2. The geometry of the complex is the typical four-legged piano stool, the stool being the Cp ligand and the four legs being identified by the three PMe3 ligands and the OH group. The hydroxo hydrogen atom could not be freely refined and restraints were applied to the O–H distance, but the Mo–O–H angle was refined freely and converged to 133(4)8.The Mo– CNT and Mo–P distances are the same, within experimental error, to the corresponding distances in the isostructural complex [CpMoCl(PMe3)3]1.48 The shorter Mo(1)–P(1) distance relative to the Mo(1)–P(2) and Mo(1)–P(3) distances reflects the weaker trans influence of the OH group relative to PMe3. The Cp ligand is somewhat asymmetrically disposed on top of the metal center, the diVerence between the longest and the shortest Mo–C distance being 0.130(6) Å.As is clear from Fig. 2, the two relative trans PMe3 ligands are distorted toward the OH ligand and away from the third PMe3 ligand. This is probably the result of the steric repulsion betweeen the bulkier PMe3 ligands. The same eVect, although less pronounced, was observed for the chloro analogue.48 Compound 2 is, to the best of our knowledge, the first known compound with a terminal Fig. 2 A top view of the [CpMo(OH)(PMe3)3]1 cation in compound 2.Only the major orientation of the disordered PMe3 ligand (P1 donor atom) is shown for clarity.502 J. Chem. Soc., Dalton Trans., 1999, 497–507 MoIII–OH bond. The Mo–O distance, therefore, can only be compared with those found in hydroxide-bridged complexes. These comprise (averaged Mo–O distances in parentheses) [Mo2(m-OH)2(m-O2CMe)(edta)]2 (H4edta = ethylenediaminetetraacetic acid) [2.04(1) Å],52 [Mo2L2Cl2(m-OH)2]21 [2.13(1) Å],53 [Mo2L2(m-O2CMe)(m-OH)2]31 [2.08(1) Å],53 and [Mo2L2- Br2(m-OH)(m-Br)]21 (L = 1,4,7-triazacyclononane) [1.84(2) Å].54 The distance in 2 compares relatively well with some of those reported above.A compensation of diVerent eVects might be responsible for this similarity: compound 2 is electronically more saturated than the above edge-bridged bioctahedral examples (less oxygen p bonding is allowed), and the OH ligand in 2 experiences a stronger trans influence from the PMe3 ligand. Both these eVects should lengthen the Mo–O bond in 2.On the other hand, the terminal bonding mode in 2 will in part counter those eVects. The angular distortions from the Cp ring are similar for all four monodentate ligands [CNT–Mo–X angles = 108.5(1), 110.7(1), 110.8(1), 112.2(2)8 for P(1), P(2), P(3) and O(1), respectively], as found for the analogous chloro derivative.48 These angular parameters have been shown to be diagnostic of the electronic configuration at the metal and the s/p bonding capabilities of the ligands.55,56 A second crystal from the batch that contains crystals of 2 turned out to correspond to a solid solution of 2 and the corresponding chloro complex.The structure contains two molecules in the asymmetric unit, the first one corresponding to a fully ordered hydroxide derivative (Fig. 3) and the second one to a compositional disorder of the hydroxide and the chloride in a 58 : 42 ratio (Fig. 4). It is somewhat curious that the OH/Cl disorder is only observed in one of the two crystallographically independent molecules.The metric parameters of both molecules compare very well with those of the crystal structure of pure 2 described above (cf. Tables 3 and 4). The closeness of all bonding and angular parameters between the two complexes rationalizes the ease with which they form a solid solution. However, the molecular packing is not quite identical for the two complexes, since the pure chloride forms monoclinic crystals (space group P21/c), whereas the pure hydroxide and the solid solution crystallize in diVerent orthorhombic space groups (see Table 2).Compound 3 was previously reported 41 but not crystallographically characterized. A view of the [CpMo(O)(PMe3)2]1 ion is shown in Fig. 5. The short Mo–O distance of 1.674(13) Å is an immediate indication of the multiple bonding between these two atoms. Other terminal Mo(IV)–O(oxo) bonds are 1.710(3) Å for [MeC(CH2–h5-C5H4)(CH2PPh2)2MoO]1,51 1.68(1) Å for [Mo(O)I(dmpe)2]1,57 1.676(7) and 1.801(9) Å for cis,mer-MoOCl2(L)3 (L = PMe2Ph and PEt2Ph, respectively), 58,59 and 1.8184(8) Å in trans-MoO2(dppe)2.60 Some of the Fig. 3 A top view of the ordered [CpMo(OH)(PMe3)3]1 cation in the structure of [CpMo(OH)0.79Cl0.21(PMe3)3]1BF4 2. Ellipsoids are drawn at the 30% probability level. previously reported values might be artificially long because of what has been initially described as “bond-stretch isomerism” and later recognized as a disorder-related artifact.61 The Mo–O distance found for 3, however, is close to the shortest distances ever reported, thus we strongly believe that there are no disorder problems (e.g.O/Cl compositional disorder) in this case. The Mo–P distance is shorter than in compound 2, as expected for the smaller ionic radius associated with the higher oxidation state. The Cp ring is highly distorted for this compound, with the carbon atoms trans to the oxo ligand being the furthest away from the metal center.This phenomenon has also been observed for the isoelectronic Cp*ReOCl2 molecule.62 Computational studies Geometry optimizations have been carried out on model complexes where the PMe3 ligand is replaced with PH3. The systems investigated are the singlet Mo(IV) complexes [CpMoO(PH3)2]1 (Cs), [CpMoH2(PH3)3]1 (Cs), and [CpMoH(OH)(PH3)3]1 (C1), the Mo(II) aqua complex [CpMo(PH3)3(H2O)]1 (C1), and the Mo(III) doublet species [CpMoH(PH3)3]1 (Cs), [CpMo(OH)- (PH3)3]1 (C1) and CpMo(O)(PH3)2 (Cs). For the purpose of estimating reaction energies (see Discussion), optimized geometries and energies were also obtained for the redox Fe(II/III) couple Cp2Fen1 (n = 0, 1; singlet and doublet, respectively) and for the inorganic base PH3 and its conjugated acid PH4 1.All calculations with mirror-imposed symmetry were carried out only with eclipsed Cp and X ligands (i.e. with the unique Cp carbon atom eclipsed with a ligand). Calculations on both eclipsed and staggered geometries for [CpMoO(PH3)2]1 show a diVerence of only 0.9 kcal mol21 (1 cal = 4.184 J), and previous studies on other half-sandwich systems, namely CpMoCl2(PH3) Fig. 4 A top view of the disordered [CpMo(OH)0.58Cl0.42(PMe3)3]1 cation in the structure of [CpMo(OH)0.79Cl0.21(PMe3)3]1BF4 2. Ellipsoids are drawn at the 30% probability level. Fig. 5 A view of the [CpMo(O)(PMe3)2]1 ion in compound 3. Ellipsoids are drawn at the 30% probability level.J.Chem. Soc., Dalton Trans., 1999, 497–507 503 (M = Cr, Mo)63 and CpMo(PH2)(PH3)2 49 have shown analogous small diVerences in energy between the two possible Cs conformations. All energies of relevant systems are collected in Table 6. The optimized bonding parameters for the [CpMo(OH)(PH3)3]1 and [CpMoO(PH3)2]1 species are compared with the X-ray results in Tables 3 and 5, respectively, while those of the other Mo complexes are collected in Table 7. The comparison between experimental and calculated geometries is excellent.The main diVerences consist in the well Fig. 6 Optimized geometries of (a) [CpMoO(PH3)2]1; (b) [CpMo- (OH)(PH3)3]1; (c) CpMo(OH)(PH3)2 (S = 1/2); (d) [CpMoH2(PH3)3]1; (e) [CpMoH(OH)(PH3)3]1; (f) [CpMo(PH3)3(H2O)]1; (g) [CpMoH- (PH3)3]1. Table 6 Spin state, U3BLYP total energies, and UB3LYP S(S 1 1) for various systems System [CpMoO(PH3)2]1 [CpMo(OH)(PH3)3]1 CpMoO(PH3)2 CpMoH(PH3)3 a CpMo(OH)(PH3)3 a CpMo(OH)(PH3)2 a [CpMoH2(PH3)3]1 [CpMoH(OH)(PH3)3]1 [CpMo(PH3)3(H2O)]1 [CpMo(H)(PH3)3]1 PH3 a [PH4]1 Cp2Fe [Cp2Fe]1 Spin state Singlet Doublet Doublet Singlet Singlet Triplet Singlet Singlet Singlet Doublet Singlet Singlet Singlet Doublet Total energy/Eh 2352.6380 2361.5012 2352.8176 2286.4484 2361.6866 2353.4096 2286.8546 2362.0893 2362.0957 2286.2541 28.2699 28.5697 2510.4397 2510.2030 S(S 1 1) 0 0.7556 0.7556 000 0.7560 00 0.8135 a Data from ref. 49. known overestimation of the bond distances by the DFT method,64 while the angular parameters are very similar.Other minor diVerences are attributable to the diVerent steric encumbrance of the PMe3 ligand relative to the PH3 model. Views of all optimized Mo complexes are shown in Fig. 6. The geometry of CpMoO(PH3)2 converged to a pseudotetrahedral coordination environment with one position occupied by a h3-Cp ligand [Fig. 6(c)], similar to the corresponding one-electron oxidation product, [CpMoO(PH3)2]1 [Fig. 6(a)]. The shortening of all the bonds to the Mo atom (especially the Mo–P bonds) upon oxidation is perfectly consistent with the increased eVective charge of the oxidized Mo center.The derivative [ CpMoH2(PH3)3]1 [ Fig. 6(d)] was optimized starting from a pseudo-octahedral geometry with the Cp ring and one hydride occupying the axial positions and the three phosphine ligands adopting a mer arrangement in the equatorial plane. This geometry is indicated for the previously reported 41 [CpMoH2(PMe3)3]1 complex by low-temperature 1H and 31P NMR studies38 and corresponds to that observed by X-ray crystallography for the isoelectronic complex CpMoH3(PMe2- Ph)2.65 A similar geometry is also adopted by the isoelectronic complex [(C6H5Me)WH2(PMe3)3]2 1.66 The input geometry for [CpMoH(OH)(PH3)3]1 was derived from the optimized [CpMoH2(PH3)3]1 geometry, by replacement of the equatorial H ligand with OH. Two diVerent optimizations were carried out, with an initial dihedral CNT–Mo–O–H angle of 458 and 1358, respectively, leading to the same final result [Fig. 6(e)].The equatorial placement of the OH ligand and the axial placement of the hydrido ligand are suggested by the identical stereochemistry of the isoelectronic [CpMoH(PMe3)3- (MeCN)]21.38 The metric parameters correspond rather closely to those of the dihydride cation, the major diVerence being that the axial hydride ligand is now pushed more strongly by the hydroxo ligand, thus bending toward the Ptrans atom.This is an indication that the hydrogen atom has indeed hydridic character, resulting in a repulsive eVect of the oxygen lone pairs. Complex [CpMo(PH3)3(H2O)]1 was generated by adding a proton to CpMo(OH)(PH3)3 49 in C1 symmetry, leading, however, to a pseudo-C2v final geometry as shown in Fig. 6(f). In view of the fact that the metal is formally saturated in this complex, it is somewhat surprising to find a nearly planar environment at oxygen (the sum of bond angles at oxygen is 359.508).The geometry of optimized [CpMoH(PH3)3]1 [Fig. 6(g)] is very similar with that of the neutral precursor. A comparison of the bond distances and bond angles of CpMoH(PH3)3 1 with those previously obtained 49 for CpMoH(PH3)3 shows that, upon oxidation, the Mo–Cp and Mo–H bonds shorten, while the Mo–P bonds lengthen. The same eVects have been noted for the [CpWH3(H2PCH2CH2PH2)]n1 (n = 0, 1) pair.67 In addition, there is an overall increase in the umbrella compression of the “legs” of the four legged piano stool structure.In comparison with the [CpMo(PH3)3(OH)]1 and [CpMo(PH3)3(H2O)]1 structures, the steric repulsion between two adjacent PH3 ligands results in a wider P–Mo–P angle, because the hydride ligand opposes a weaker repulsive eVect relative to the OH and H2O ligands. The optimized ferrocene (eclipsed configuration) has distances (Fe–C 2.120 Å, C–C 1.443 Å) in good agreement with those determined by X-ray68 and neutron diVraction.69 The ferrocenium cation (S = 1/2) gave a slightly distorted eclipsed geometry (138 distortion, Fe–C 2.164 Å, C–C 1.443 Å).Discussion As stated in the Introduction, the decomposition of 11 in dry MeCN has been the subject of a recent report from our Laboratory.38 The new results reported herein show that water can participate in the decomposition scheme of oxidized hydride complexes by acting as a ligand and deviate the decomposition path toward the formation of higher oxidation state organometallic hydroxo and oxo derivatives.504 J.Chem. Soc., Dalton Trans., 1999, 497–507 Table 7 B3LYP-optimized selected geometric parameters for CpMoO(PH3)2, CpMoH2(PH3)3 1, CpMoH(OH)(PH3)3 1, CpMo(H2O)(PH3)3 1 and [CpMoH(PH3)3] 1a Compound CpMoO(PH3)2 [CpMoH2(PH3)3]1 [CpMoH(OH)(PH3)3]1 [CpMo(H2O)(PH3)3]1 [CpMoH(PH3)3]1 Distance/Å Mo–CNTa Mo–C1 Mo–C2 Mo–C3 Mo–P Mo–O Mo–CNT Mo–Hax Mo–Heq Mo–Ptrans Mo–Pcis Mo–CNT Mo–H Mo–O O–H Mo–Ptrans Mo–Pcis Mo–CNT Mo–O O–H Mo–Ptrans Mo–Pcis Mo–CNT Mo–H Mo–Ptrans Mo–Pcis 2.100 2.262 2.372 2.57 2.701 1.746 2.030 1.691 1.718 2.583 2.583 2.084 1.716 2.034 0.980 2.597 2.565 2.540 2.044 2.294 0.975 2.540 2.530 2.021 1.705 2.575 2.564 Angle/8 CNT–Mo–O CNT–Mo–P P–Mo–P O–Mo–P CNT–Mo–Heq CNT–Mo–Hax CNT–Mo–Ptrans CNT–Mo–Pcis Heq–Mo–Pt Heq–Mo–Pc Heq–Mo–Hax Pcis–Mo–Pcis Ptrans–Mo–Pcis CNT–Mo–H CNT–Mo–O CNT–Mo–Ptrans CNT–Mo–Pcis H–Mo–Ptrans H–Mo–Pcis O–Mo–Ptrans O–Mo–Pcis O–Mo–H Ptrans–Mo–Pcis Pcis–Mo–Pcis Mo–O–H CNT–Mo–O CNT–Mo–Ptrans CNT–Mo–Pcis O–Mo–Ptrans O–Mo–Pcis Ptrans–Mo–Pcis Pcis–Mo–Pcis Mo–O–H H–O–H CNT–Mo–H CNT–Mo–Ptrans CNT–Mo–Pcis Pcis–Mo–Pcis Ptrans–Mo–Pcis 157.95 109.50 99.79 84.03 104.03 170.49 141.10 111.26 141.10 78.01 66.46 134.95 87.68 173.50 106.58 107.29 110.57 111.17 66.22 69.70 69.77 145.62 86.25 74.56 79.91 87.82 87.43 137.48 121.02 106.96 110.14 122.58 142.89 79.17 81.07 114.70 123.83 111.84 102.05 115.44 120.96 113.41 87.17 Dihedral angle/8 P–Mo–CNT–P O–Mo–CNT–P CNT–Mo–O–H Ptrans–Mo–CNT–Pcis Ptrans–Mo–CNT–Pcis 108.53 122.73 85.50 92.19 102.83 a CNT = centroid of cyclopentadienyl ring.Water as a ligand and not as a base Water is not expected to act as a deprotonating agent toward 11 in the presence of 1. This is because it has been shown previously that compound 1 can be protonated by strong acids in water 41 (that is, 1 is a stronger base than water). As a matter of fact, compound 1 is even a stronger base than PMe3.In a control experiment, we have shown that the protonation of a 1 : 1 mixture of compounds 1 and PMe3 with a substoichiometric amount of HBF4 gives rise to the NMR resonances (1H and 31P) of complex [1–H]1, while those of free PMe3 remain unchanged. In addition, no evidence for the formation of PHMe3 1 was ever obtained during the oxidation studies of 1, while the formation of [1–H]1 has been verified by NMR spectroscopy. The computational work on the PH3 model systems also indicates that the Mo complex is the better base: from the data in Table 6, we calculate that equation (4) is highly exothermic.PH4 1 1 CpMoH(PH3)3 æÆ PH3 1 [CpMoH2(PH3)3]1 (4) DE = 266.8 kcal mol21 The quantitative result of the energetic calculation should be taken cautiously, because the solvent might significantly alter the energetic situation relative to the gas phase calculation via solvation eVects. The solvation energies on the two sides of the equation, especially those related to the cationic species, might not balance oV exactly.In addition, the replacement of PH3 with PMe3 should render both the free phosphine and the neutral hydride complex more basic. The inductive eVect of a single H/Me replacement will influence the basicity of the P lone pair (only one bond away) in a stronger way relative to the Mo lone pair (two bonds away). However, the Mo hydride complex has 9 H/Me substitutions as opposed to only three for the free phosphine.Once again, these eVects might not balance oV completely. An additional, probably smaller eVect, is related to entropy changes. Considerations such as these should be kept in mind also when analyzing the other computed reaction energies below. On the basis of the above considerations, we conclude that the presence of water (and even PMe3) as a base is irrelevant when compound 1 is present (for instance when part of 1 is left unoxidized by use of a substoichiometric amount of ferrocenium).Therefore, the diVerent outcome of the oxidation under wet conditions (formation of 2, 3 and 4) relative to the oxidation in dry MeCN or THF may be attributed to the coordination of water in place of a solvent molecule and to the follow-up reactivity that results from this event.J. Chem. Soc., Dalton Trans., 1999, 497–507 505 The deliberate addition of water to a MeCN solution does not change the course of the oxidation reaction of compound 1 in a substantial way, relative to the reported procedure in the dry solvent.38 The major products of this reaction are the same ones obtained under dry conditions.However, the formation of some oxo complex [CpMoO(PMe3)2]1 and the water adduct [CpMoH(PMe3)3(H2O)]21 indicates that water is able to compete with the MeCN coordination and partially deviate the course of the decomposition of the 11 intermediate. In the less coordinating solvents THF or acetone, the coordination of water is seen to compete even more favorably with the solvent, although the acetone adduct [CpMoH(PMe3)3(CD3COCD3)]21 has also been spectroscopically observed.Exchange studies carried out in CD3CN show that the acetone adduct is more susceptible than the water adduct toward exchange of the solvent donor. Mechanistic proposal A possible mechanism for the formation of the observed products is shown in Scheme 3, where all boxed species have been observed spectroscopically or analyzed by single crystal X-ray diVraction.This mechanism is consistent with the basic understanding that we have gained for this system by the studies in dry acetonitrile,38 with other related previous observations, and with the results of our computational studies. As the computational work has been used as an energetic guide for the rationalization of mechanistic pathways that also involve a few unobserved intermediates, a number of control calculations were also carried out.The first one involves the oxidation reaction itself. Reaction (5), a model of the oxidation process of 1, is CpMoH(PH3)3 1 Fc1 æÆ [CpMoH(PH3)3]1 1 Fc (5) DE = 226.5 kcal mol21 Scheme 3 CpMoH(PMe3)3 [CpMoH(PMe3)3]+ [ox] 1 [1]+ [CpMoH(PMe3)3(OH2)]2+ [1(OH2)]2+ [CpMoH(PMe3)3(OH)]+ [1-OH]+ CpMo(PMe3)3OH CpMo(PMe3)2(OH) + PMe3 [CpMo(PMe3)3(OH)]+ CpMo(O)(PMe3)2 4 PMe3 [CpMo(O)(PMe3)2]+ [CpMo(PMe3)3(H2O)]+ 1 [1-H]+ [1]+ 1 H2O 1 [1-H]+ 1 [1-H]+ 1 [1-H]+ [1]+ 1 1 [1-H]+ [1]+ 1 strongly exothermic.Equalizing the calculated energy diVerence with the free energy change (both PV and entropy contributions are probably negligible), we arrive at a calculated reduction potential of 21.15 V for complex CpMoH(PH3)3 relative to the ferrocene standard. This compares with the experimentally determined value of 21.46 V for compound 1. Given that the PMe3 should be a better donor than PH3 and should render the hydride complex easier to oxidize, we consider the agreement of our calculations with the experimental data as satisfactory. The mechanism in Scheme 3 parallels that established for 11 in MeCN38 involving disproportionation of 11 according to Scheme 2 (S = H2O).This involves the formation of [1(H2O)]21, analogous to the previously reported [1(THF)]21 and [1(MeCN)]21,38 as an intermediate. NMR resonances that are attributed to complex [1(H2O)]21 have been observed under particular conditions (see Results). There may be an equilibrium between the water adduct and other solvent adducts (e.g.[1(CH3COCH3)]21 in acetone), but only the former proceeds to further decomposition, as discussed below. Water becomes more acidic when coordinated to a Lewisacidic metal center. It is thus reasonable to suppose that the water ligand in the latter complex may be easily deprotonated, leading to complex [1–OH]1, which is still an 18-electron complex of Mo(IV). Hydrido–hydroxo complexes analogous to [1–OH]1 are amply precedented in the literature, for instance Cp*2W(OH)H70 and [IrH(OH)(PMe3)4]1.71 In addition, Mo(IV) complexes that are isoelectronic with [1–OH]1 such as Cp- MoH(OH)(C6D5)(PMe3)2 and CpMoH(OH)(h2-CH2PMe2)- (PMe3) have been implicated as intermediates in C–H activation processes.49 However, an alternative possibility involves deprotonation of [1(H2O)]21 at the metal, to aVord the tautomer complex [CpMo(PMe3)3(H2O)]1.This corresponds to the pathway established in dry MeCN, transforming [1(MeCN)]21 to the final product [CpMo(PMe3)3(MeCN)]1.Either way, the deprotonating agent must be 1, which is generated locally and stoichiometrically by the preceding electron transfer step. Indeed, complex [1–H]1 is one of the observed products of the oxidation of 1 in wet acetone or THF. The calculations on the PH3 model systems indicate that the Mo(II) aquo complex is 4.0 kcal mol21 more stable than the Mo(IV) hydrido–hydroxo tautomer, implying a stronger thermodynamic acidity at the hydride position for [CpMoH(PH3)3- (H2O)]21.The tautomerization reaction corresponds to an oxidative addition/reductive elimination of the water O–H bond and could well be a very rapid process. It is relevant to mention here that a fast Mo(H2O)/MoH(OH) tautomerization process has been invoked for a very similar CpMoH(OH)- (h2-CH2PMe2)(PMe3) system to rationalize the rapid H/D scrambling between the hydroxo and PMe3 positions in compound CpMo(OH)(PMe3)2.49 The alternative proton transfer decomposition pathway, according to the well established mechanism,16,18,19,26 would lead to the same product, viz.[CpMo(PMe3)3(H2O)]1 or to the hydrido–hydroxo tautomer, according to Scheme 1. Up to this point, the oxidation of 1 with 1 equivalent of Fc1 should proceed according to the stoichiometry of equation (6), which 2 1 2 2 e2 1 H2O æÆ [1–H]1 1 [CpMo(PMe3)3(H2O)]1 (6) corresponds to that established for the decomposition of the 11 radical in dry MeCN.However, there is no evidence in the NMR monitoring for the accumulation of an intermediate that could be interpreted as [CpMo(PMe3)3(H2O)]1. Therefore, if this intermediate forms, it must be further transformed rapidly under the reaction conditions. The transformation of [CpMo(PMe3)3(H2O)]1 to the cation of 2 requires loss of one proton and one electron. The order of these events shown in Scheme 3 is suggested by the existence506 J. Chem.Soc., Dalton Trans., 1999, 497–507 of CpMo(OH)(PMe3)2,39,49 and by the energetics of Table 6. CpMo(OH)(PMe3)3 is unstable relative to PMe3 dissociation and formation of the 16-electron CpMo(OH)(PMe3)2 complex, a spin triplet complex.49 The deprotonation process [equation (7)] is predicted as too unfavorable when using PMe3 or H2O as [CpMo(PH3)3(H2O)]1 1 CpMoH(PH3)3 æÆ CpMo(OH)(PH3)3 1 [CpMoH2(PH3)3]1 (7) DE = 11.8 kcal mol21 a base. Thus, this pathway is possible only in the presence of unoxidized 1.Complex CpMo(OH)(PMe3)2 is not spectroscopically observed during the oxidation of 1, thus electron transfer must rapidly follow the proton loss. It is also worth noting here that, as we have previously shown,39,49 the FcBF4 oxidation of isolated CpMo(OH)(PMe3)2 in the presence of PMe3 aVords a mixture of 2 and 3. The oxidation of CpMo(OH)(PMe3)3 (in equilibrium with CpMo(OH)(PMe3)2 and free PMe3) may be accomplished again by the long-lived 11 radical as indicated in Scheme 3.The model PH3 system confirms the feasibility of this electron transfer process [equation (8)]. CpMo(OH)(PH3)3 1 [CpMoH(PH3)3]1 æÆ [CpMo(OH)(PH3)3]1 1 CpMoH(PH3)3 (8) DE = 25.5 kcal mol21 The formation of 2 from 1 requires the consumption of 3 oxidizing equivalents and leads to the release of 2 equivalents of protons. The resulting stoichiometry is that of equation (9). 3 1 2 3 e2 1 H2O æÆ 2 [1–H]1 1 [CpMo(OH)(PMe3)3]1 (9) Note that this equation involves an ox/1 ratio of 1 : 1, just like equation (6).Compound 3 is obtained by further oxidation of 2, probably via deprotonation and loss of PMe3 to aVord the neutral Mo(III) oxo intermediate CpMo(O)(PMe3)2, 4, which can subsequently be oxidized to 3. The formation of 4 and 3 requires the stoichiometries of equations (10) and (11), respectively. 4 1 2 3 e2 1 H2O æÆ 3 [1–H]1 1 4 1 PMe3 (10) 4 1 2 4 e2 1 H2O æÆ 3 [1–H]1 1 [CpMoO(PMe3)2]1 1 PMe3 (11) Equation (11) involves again an ox/1 ratio of 1 : 1, whereas the formation of 4 in equation (10) requires less than 1 equivalent of oxidant.The observation of the EPR triplet resonance when using a small amount of oxidant (see Fig. 1), is in agreement with the above argument. No other species involved in Scheme 3 could give rise to the observed EPR resonance. Our attempts to optimize a neutral CpMoO(PH3)3 model led to direct expulsion of a PH3 ligand and converged to two separate CpMoO(PH3)2 and PH3 units.The experimental evidence of a fast exchange between free PMe3 and the ligand coordinated to compound 4 (see Results) indicates, however, that a CpMoO(PMe3)3 complex may be easily accessible as a transient. The thermodynamic feasibility of the overall deprotonation process is shown in equation (12). Although the process is calculated as slightly [CpMo(OH)(PH3)3]1 1 CpMoH(PH3)3 æÆ [CpMoH2(PH3)3]1 1 CpMoO(PH3)2 1 PH3 (12) DE = 14.7 kcal mol21 endothermic, the reaction is entropically favored by the release of one molecule of free PH3.A spontaneous deprotonation of [CpMo(OH)(PMe3)3]1 with proton capture by the released PMe3 ligand does not seem consistent with the calculations on the model system [equation (13)]. The subsequent oxidation of [CpMo(OH)(PH3)3]1 æÆ CpMoO(PH3)2 1 PH4 1 (13) DE = 171.7 kcal mol21 4 by 11 is also a favorable process, as indicated in the model system by equation (14). CpMoO(PH3)2 1 [CpMoH(PH3)3]1 æÆ [CpMoO(PH3)2]1 1 CpMoH(PH3)3 (14) DE = 29.2 kcal mol21 It is interesting to observe that each oxidation step is followed by the release of a proton, and that each molecule of the oxidizing 11 generates a molecule of 1 that can capture the proton.Thus, all 1 is ultimately consumed by the oxidation reaction with 1 equivalent of ferrocenium, to yield the oxidation products and the protonation product [1–H]1. According to the stoichiometry of equations (9), (10) and (11), compound 4 should only be observed when using a substoichiometric amount of oxidant (experimentally verified), while the use of 1 or more equivalents of oxidant should yield only 2 and 3.In conclusion, the formation of oxo and hydroxo products probably starts with coordination of a water molecule to an unsaturated metal center which is generated by an oxidation process, and continues with tandem oxidation/deprotonation steps. Conclusions With this contribution we have shown that water can intervene in the decomposition of paramagnetic hydride complexes not only as a proton acceptor, as previously established for less electron-rich hydride complexes, but also as a ligand.This event leads to higher oxidation state hydroxo and oxo complexes via a cascade of energetically favorable deprotonation/oxidation steps. The existence of sequential electron and proton tranfer reactions is well documented in coordination chemisry. In particular, these tandem reactions on molybdenum derivatives containing a coordinated water residue have been proven biologically relevant.72,73 When the parent neutral hydride is loaded with electron-releasing ligands, it becomes more basic than water and is the preferred proton acceptor.However, the need for a donor solvent to electronically saturate the products of decomposition of the paramagnetic hydride may make water the preferred substrate when this process is carried out in solvents of limited coordinating ability. The follow-up processes are then a consequence of the increased acidity of water upon coordination to the Lewis-acidic metal center.Acknowledgements We are grateful to the DOE (grant no. DEFG059ER14230) for support of this work and to Dr Raymund C. Torralba for some preliminary observations. References 1 J. R. Sanders, J. Chem. Soc., Dalton Trans., 1973, 748. 2 J. R. Sanders, J. Chem. Soc., Dalton Trans., 1975, 2340. 3 M. Gargano, P. Giannoccaro, M. Rossi, G. Vasapollo and A.Sacco, J. Chem. Soc., Dalton Trans., 1975, 9. 4 G. Pilloni, G. Schiavon, G. Zotti and S. Zecchin, J. Organomet. Chem., 1977, 134, 305. 5 P. M. Treichel, D. C. Molzahn and K. P. Wagner, J. Organomet. Chem., 1979, 174, 191. 6 R. J. Klinger, J. C. HuVman and J. K. Kochi, J. Am. Chem. Soc., 1980, 102, 208. 7 J. D. Allison, C. J. Cameron, R. E. Wild and R. A. Walton, J. Organomet. Chem., 1981, 218, C62.J. Chem. Soc., Dalton Trans., 1999, 497–507 507 8 L. F. Rhodes, J. D. Zubkowski, K.Folting, J. C. HuVman and K. G. Caulton, Inorg. Chem., 1982, 21, 4185. 9 J. W. Bruno and K. G. Caulton, J. Organomet. Chem., 1986, 315, C13. 10 M. R. Detty and W. D. Jones, J. Am. Chem. Soc., 1987, 109, 5666. 11 M. T. Costello and R. A. Walton, Inorg. Chem., 1988, 27, 2563. 12 W. A. Herrmann, H. G. Theiler, E. Herdtweck and P. Kiprof, J. Organomet. Chem., 1989, 367, 291. 13 C. Bianchini, F. Laschi, M. Peruzzini, F. Ottaviani, A. Vacca and P. Zanello, Inorg. Chem., 1990, 29, 3394. 14 L. Chen and J. A. Davies, Inorg. Chim. Acta, 1990, 175, 41. 15 L. Roullier, D. Lucas, Y. Mugnier, A. Antiñolo, M. Fajardo and A. Otero, J. Organomet. Chem., 1990, 396, C12. 16 O. B. Ryan, M. Tilset and V. D. Parker, J. Am. Chem. Soc., 1990, 112, 2618. 17 O. B. Ryan, K.-T. Smith and M. Tilset, J. Organomet. Chem., 1991, 421, 315. 18 O. B. Ryan, M. Tilset and V. D. Parker, Organometallics, 1991, 10, 298. 19 O. B. Ryan and M. Tilset, J. Am. Chem. Soc., 1991, 113, 9554. 20 C. Roger, P. Hamon, L. Toupet, H. Rabaâ, J.-Y. Saillard, J.-R. Hamon and C. Lapinte, Organometallics, 1991, 10, 1045. 21 L. Roullier, D. Lucas, Y. Mugnier, A. Antiñolo, M. Fajardo and A. Otero, J. Organomet. Chem., 1991, 412, 353. 22 D. E. Westerberg, L. F. Rhodes, J. Edwin, W. E. Geiger and K. G. Caulton, Inorg. Chem., 1991, 30, 1107. 23 P. Hamon, L. Toupet, J.-R. Hamon and C. Lapinte, Organometallics, 1992, 11, 1429. 24 M. Tilset, J. Am. Chem. Soc., 1992, 114, 2740. 25 G. Jia, A.J. Lough and R. H. Morris, Organometallics, 1992, 11, 161. 26 K.-T. Smith and M. Tilset, J. Organomet. Chem., 1992, 431, 55. 27 C. Amatore, J. J. R. Fraústo da Silva, M. F. C. Guedes da Silva, A. J. L. Pombeiro and J.-N. Verpeaux, J. Chem. Soc., Chem. Commun., 1992, 1289. 28 V. Skagestad and M. Tilset, J. Am. Chem. Soc., 1993, 115, 5077. 29 A. A. Zlota, M. Tilset and K. G. Caulton, Inorg. Chem., 1993, 32, 3816. 30 K.-T. Smith, C. Rømming and M. Tilset, J. Am. Chem. Soc., 1993, 115, 8681. 31 M. Jiménez-Tenorio, M. C. Puerta and P. Valerga, Organometallics, 1994, 13, 3330. 32 W. Kaim, R. Reinhardt and M. Sieger, Inorg. Chem., 1994, 33, 4453. 33 A. Pedersen and M. Tilset, Organometallics, 1994, 13, 4887. 34 C. A. Blaine, J. E. Ellis and K. R. Mann, Inorg. Chem., 1995, 34, 1552. 35 R. T. Hembre, J. S. McQueen and V. W. Day, J. Am. Chem. Soc., 1996, 118, 798. 36 D. Menglet, A. M. Bond, K. Coutinho, R. S. Dickson, G. G. Lazarev, S. A. Olsen and J. R. Pilbrow, J.Am. Chem. Soc., 1998, 120, 2086. 37 E. A. Quadrelli, H.-B. Kraatz and R. Poli, Inorg. Chem., 1996, 35, 5154. 38 J. C. Fettinger, H.-B. Kraatz, R. Poli, E. A. Quadrelli and R. C. Torralba, Organometallics, 1998, 17, 5767. 39 J. C. Fettinger, H.-B. Kraatz, R. Poli and E. A. Quadrelli, Chem. Commun., 1997, 889. 40 F. Abugideiri, M. A. Kelland, R. Poli and A. L. Rheingold, Organometallics, 1992, 11, 1303. 41 M. Brookhart, K. Cox, F. G. N. Cloke, J. C. Green, M. L. H. Green, P.M. Hare, J. Bashkin, A. E. Derome and P. D. Grebenik, J. Chem. Soc., Dalton Trans., 1985, 423. 42 M. J. Frisch, G. W. Trucks, H. B. Schlegel, P. M. W. Gill, B. G. Johnson, M. A. Robb, J. R. Cheeseman, T. A. Keith, G. A. Petersson, J. A. Montgomery, K. Raghavachari, M. A. Al-Laham, V. G. Zakrzewski, J. V. Ortiz, J. B. Foresman, J. Cioslowski, B. B. Stefanov, A. Nanayakkara, M. Challacombe, C. Y. Peng, P. Y. Ayala, W. Chen, M. W. Wong, J. L. Andres, E. S. Replogle, R. Gomperts, R. L. Martin, D. J. Fox, J. S. Binkley, D. J. Defrees, J. Baker, J. P. Stewart, M. Head-Gordon, C. Gonzales and J. A. Pople, Gaussian 94 (Revision E.1). Gaussian Inc., Pittsburgh, PA, 1995. 43 A. D. Becke, J. Chem. Phys., 1993, 98, 5648. 44 T. H. Dunning, Jr. and P. J. Hay, in Modern Theoretical Chemistry ed. H. F. Schaefer, III, Plenum Press, New York, 1976, p. 1. 45 P. J. Hay and W. R. Wadt, J. Chem. Phys., 1985, 82, 270. 46 P. J. Hay and W. R. Wadt, J. Chem. Phys., 1985, 82, 299. 47 W. R. Wadt and P. J. Hay, J. Chem. Phys., 1985, 82, 284. 48 J. C. Fettinger, H.-B. Kraatz, R. Poli and A. L. Rheingold, Acta Crystallogr., Sect. C, 1995, 51, 364. 49 R. Poli and E. A. Quadrelli, New J. Chem., 1998, 22, 435. 50 G. S. B. Adams and M. L. H. Green, J. Chem. Soc., Dalton Trans., 1981, 353. 51 B. Antelmann, G. Huttner and U. Winterhalter, J. Organomet. Chem., 1998, 555, 119. 52 G. G. Kneale, A. J. Geddes, Y. Sasaki, T. Shibahara and A. G. Sykes, J. Chem. Soc., Chem. Commun., 1975, 356. 53 K. Wieghardt, M. Hahn, W. SwiridoV and J. Weiss, Inorg. Chem., 1984, 23, 94. 54 P. Chauduri, K. Wieghardt, I. Jibril and G. Huttner, Z. Naturforsch., Teil B, 1984, 39, 1172. 55 R. Poli, Organometallics, 1990, 9, 1892. 56 Z. Lin and M. B. Hall, Organometallics, 1993, 12, 19. 57 B. E. Owens and R. Poli, Acta Crystallogr., Sect. C, 1992, 48, 2137. 58 L. Manojlovic-Muir, J. Chem. Soc., Dalton Trans., 1971, 2796. 59 L. Manojlovic-Muir and K. W. Muir, J. Chem. Soc., Dalton Trans., 1972, 686. 60 J. Bendix, H. Birkedal and A. Bøgevig, Inorg. Chem., 1997, 36, 2702. 61 G. Parkin, Acc. Chem. Res., 1992, 25, 455. 62 W. A. Herrmann, E. Herdtweck, M. Flöel, J. Kulpe, U. Küsthardt and J. Okuda, Polyhedron, 1987, 6, 1165. 63 I. Cacelli, D. W. Keogh, R. Poli and A. Rizzo, New J. Chem., 1997, 21, 133. 64 Recent developments and Applications of Modern density Functional Theory, ed. J. M. Seminario, Elsevier, Amsterdam, 1996. 65 F. Abugideiri, J. C. Fettinger, B. Pleune, R. Poli, C. A. Bayse and M. B. Hall, Organometallics, 1997, 16, 1179. 66 M. L. H. Green, A. K. Hughes, P. Lincoln, J. J. Martin-Polo, P. Mountford, A. Sella, L.-L. Wong, J. A. Bandy, T. W. Banks, K. Prout and D. J. Watkin, J. Chem. Soc., Dalton Trans., 1992, 2063. 67 B. Pleune, D. Morales, R. Meunier-Prest, P. Richard, E. Collange, J. C. Fettinger and R. Poli, J. Am. Chem. Soc., submitted. 68 P. Seiler and J. D. Dunitz, Acta Crystallogr., Sect. B, 1979, 35, 1068. 69 F. Takusagawa and T. F. Koetzle, Acta Crystallogr., Sect. B, 1979, 35, 1074. 70 G. Parkin and J. E. Bercaw, Organometallics, 1989, 8, 1172. 71 D. Milstein, J. C. Calabrese and I. D. Williams, J. Am. Chem. Soc., 1986, 108, 6387. 72 E. I. Stiefel, Proc. Natl. Acad. Sci. USA, 1973, 70, 988. 73 E. I. Stiefel, J. Chem. Soc., Dalton Trans., 1997, 3915. Paper 8/07111J © Copyright 1999 by the Royal Society of Chemistry

ISSN:1477-9226    

DOI:10.1039/a807111j

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1998, Pages 505–510 505 Complexes of p-tert-butylcalix[5]arene with lanthanides: synthesis, structure and photophysical properties † Loïc J. Charbonnière,a Christian Balsiger,a Kurt J. Schenk b and Jean-Claude G. Bünzli *,a a Institute of Inorganic and Analytical Chemistry, BCH, University of Lausanne, CH-1015 Lausanne, Switzerland b Institute of Crystallography, BSP, University of Lausanne, CH-1015 Lausanne, Switzerland Spectrophotometric pKa determination for p-tert-butylcalix[5]arene (H5L) in acetonitrile (pKa1 = 11.5 ± 0.7, pKa2 = 15.4 ± 1.0 at 298 K) evidenced both intra- and inter-molecular stabilisation of the deprotonated forms.Dimeric complexes [Ln2(H2L)2(dmso)4] (Ln = EuIII, GdIII, or TbIII; dmso = dimethyl sulfoxide) were isolated from tetrahydrofuran (thf) in the presence of NaH as base. A single-crystal analysis of [Eu2(H2L)2(dmso)4]? 10thf showed the deformation of the cone conformation of the calixarene upon complexation and co-ordination of dmso molecules by inclusion through the hydrophobic cavity of the ligand. A photophysical investigation revealed a total quenching of the metal luminescence by a ligand-to-metal charge-transfer state in the case of EuIII while luminescence of TbIII is sensitised (quantum yield in thf: 5.1%).The temperature-dependent lifetime of TbIII is analysed in terms of a potential metal-to-ligand back-transfer process. The cyclic framework of calixarenes,1 associated with the presence of phenol oxygen donor atoms, affords an interesting platform for the complexation of metal atoms,2 while the hydrophobic cavity allows the inclusion of charged 3 and neutral 4 organic guests.The relative facility with which calixarenes can be partially or totally functionalised at the upper or lower rims, coupled with their easy large-scale synthesis, at least for the even members of the series (n = 4, 6 or 8), has opened further perspectives for their use in supramolecular chemistry.5 In particular, the hard acid character of the lower rim makes calixarenes interesting potential ligands for the complexation of trivalent lanthanide ions, either for extraction purposes 6 or for the design of efficient lanthanide-based luminescent devices.7–9 Antenna effects 10 can be generated directly by the phenol units or by lower or upper rim substitution.Previous work on the complexation of lanthanides with calix[n]arenes has mainly focused on the four-, six- and eightmembered systems.While p-tert-butylcalix[8]arenes form bimetallic complexes in which the ligand is six times deprotonated, 11 the six-membered parent calixarene is bonded to LnIII by a single phenolate group.12 In the case of p-tert-butylcalix[4]- arene 2 : 2 dimeric complexes were obtained, in which the two metal atoms are separated by only 3.91 Å.13 In our laboratory we have mainly studied the photophysical properties of bimetallic complexes with p-tert-butylcalix[8]arene 14–16 and demonstrated the influence of the low-lying ligand-to-metal charge-transfer state (LMCT) on the EuIII-containing assemblies, which enhances the f–f absorption probabilities through mixing with the 4f orbitals and quenches the 5D0(Eu) excited level.Moreover, the sensitisation of EuIII and/or TbIII can be conveniently tuned by changing the nature of the para substituent.17 Calix[5]arenes appear to be adequately suited for the complexation of lanthanoid(III) ions in view of the oxygen-rich array displayed by the five phenol groups in the cone conformation, 18 and the possibility of forming multiply charged anions.19 To our knowledge, the only study on the interaction between lanthanoid(III) ions and a calix[5]arene has been † Supplementary data available: UV/VIS and IR spectra, crystallographic numbering scheme.For direct electronic access see http:// www.rsc.org/suppdata/dt/1998/505/, otherwise available from BLDSC (No.SUP 57328, 4 pp.) or the RSC Library. See Instructions for Authors, 1998, Issue 1 (http://www.rsc.org/dalton). reported by Steed et al.19 who isolated water-soluble inclusion complexes between p-sulfonatocalix[5]arene with La, Gd, Eu, Tb and Yb and showed by X-ray crystallography that the interaction is either outer sphere or occurs exclusively via one sulfonato functionality, the metal centre acting often as a bridging brick. In this paper we present the isolation and the structural and photophysical properties of lanthanoid(III) (Ln = Eu, Gd or Tb) phenolate-bonded inner-sphere complexes with p-tert-butylcalix[5]arene (H5L).The reported data add valuable information on the interaction between lanthanoid(III) ions and p-tert-butylcalix[n]arenes and demonstrate that multiple bonds between the ligand and the metal ions can be generated only in the presence of a strong base. In addition, the spectrophotometric determination of the first two pKa values of the ligand is reported.Results and Discussion Acid-base behaviour of H5L in acetonitrile The compound H5L was titrated at 298 K by Et3N in acetonitrile with NEt4ClO4 as supporting electrolyte. The evolution of the UV/VIS spectra during titration could be best fitted with the model (standard deviation: 0.007 absorbance unit between calculated and measured values) in equations (1)– (3). The first two acidity constants of p-tert-butylcalix[5]arene, H5L 1 Et3N H4L2 1 Et3NH1 log K11 = 7.0 ± 0.7 (1) 2(H5L) 1 Et3N [H5L?H4L]2 1 Et3NH1 log K21 = 13.1 ± 0.91 (2) OH R R O(3)H R O(2)H O(4) R R R O(1) O(5) n C B A E D R = But506 J.Chem. Soc., Dalton Trans., 1998, Pages 505–510 H5L 1 2 Et3N H3L22 1 2 Et3NH1 log K12 = 10.1 ± 0.7 (3) Ka1 and Ka2, can be evaluated from the relationships K11 = Ka1/ KEt3NH1 and K12 = Ka1Ka2/(KEt3NH1)2. With pKEt3NH1 = 18.46 20 one gets pKa1 = 11.5 ± 0.7 and pKa2 = 15.4 ± 1.0.The distribution curves as a function of the Et3N:H5L ratio are shown in Fig. 1. For a 1 : 2 ratio the major species is an adduct formed from two calixarene molecules with one proton removed [equation (2)]. This can be related to the association process (4) with log Kass = 6.1 ± 1.1. H5L 1 H4L2 [H5L?H4L]2 Kass = K21KEt3NH1/K1 (4) Adding more base leads to the monodeprotonated calixarene while the dianionic species appears for ratios larger than 1 : 1. The observed pKa values can be understood on the basis of the monoanion stabilisation by intramolecular hydrogen bonds forming cyclic arrays, as previously reported for pKa data measured in water-containing solvents 21 or in the same solvent but with different calixarenes.22 The reconstructed electronic spectra (SUP 57328) display an increase of the intensity of the low-energy p æÆ p* transition centred on the ligand at 288 nm upon deprotonation, together with the emergence of a shoulder around 305 nm.We tried to demonstrate the formation of the postulated dimeric adduct [H5L?H4L]2 by measuring the electrospray (ES) mass spectrum of a solution with a 1 : 2 base : calixarene ratio at which the dimer is the major species in solution (88%).The spectrum effectively displays a peak at m/z 1641.1 (15%), which can be attributed to a dimer [2H5L? H2O?H1] (Mcalc = 1640.3). The presence of this species emphasises the possibility of intermolecular hydrogen bonds stabilising the monoanion, a way found by the system to minimise its energy in the absence of highly solvating molecules such as MeOH or water.Further deprotonation leads to Coulombic repulsion between the two parts of the dimer and destabilises it. Synthesis and characterisation of the complexes (Ln 5 Eu, Gd or Tb) Spectrophotometric investigation of the interaction between H5L and lanthanoid(III) ions in acetonitrile and in the presence of Et3N revealed a weak interaction leading to the formation of 1 : 1 and 2 : 1 species, but the analysis of the data did not converge properly.Given the values of pKa reported above, triethylamine is not a strong enough base to produce the anionic species necessary for a stable association with the lanthanoid( III) ions. We therefore switched to NaH to deprotonate the calixarene and to tetrahydrofuran (thf) as solvent. Under Fig. 1 Species distribution in solutions of H5L in acetonitrile vs. the Et3N:H5L ratio these conditions, crystalline complexes were obtained by mixing stoichiometric amounts of H5L with dimethyl sulfoxide (dmso) adducts of the lanthanoid(III) nitrates.The elemental analyses are compatible with [Ln2(H2L)2(dmso)x] with x = 4 for Ln = Gd and Tb and 3 for Eu, which emphasises the tendency of the latter compound to lose 1–2 dmso molecules upon drying. These analyses confirm the absence of nitrate in the isolated compounds. Infrared spectra of the three solids are very similar, showing an absorption band at 1023 cm21, not present in the spectrum of free H5L, and assigned to the S]O stretching vibration of oxygen-bonded dmso.23 Upon complexation the broad n(O]H) pattern shifts from 3305 cm21 for free H5L to a band centred around 3409 cm21 with a shoulder at 3305 cm21 (SUP 57328).The p æÆ p* absorption bands of the ligand are very similar for the three complexes in thf (Fig. 2). They undergo a red shift with respect to H5L, with a maximum at 288 nm (34 200 cm21;e = 25 620, 26 230 and 25 420 dm3 mol21 cm21 respectively for Eu, Gd and Tb) and a shoulder at 295 nm (33 900 cm21).In the case of EuIII an additional absorption band around 409 nm (24 445 cm21, e = 720 dm3 mol21 cm21) is attributed to the ligand-to-metal charge-transfer transition, as previously reported for bimetallic complexes with p-substituted-calix[8]- arene.15,17 The presence of the LMCT transition relaxes Laporte’s rule, so that the 5D0 �æ 7F0 transition is observed at 577.7 nm (17 310 cm21), with an unusually large absorption coefficient (e = 2.4 dm3 mol21 cm21, after correction for MLCT absorption).The ES mass spectra of thf solutions of the complexes of GdIII and TbIII display base peaks at m/z = 1932.6 and 1929.5 respectively, which are typical of the dimeric cations [Ln2(H2L)2]1. No such peak is observed for the europium(III) complex, possibly as a result of reductive reactions on the metal, in accordance with the low-lying LMCT state.Crystal structure of [Eu2(H2L)2(dmso)4]?10thf Slow concentration of a mixture of the europium(III) complex in thf afforded yellow-orange platelets suitable for X-ray diffraction analysis, with formula [Eu2(H2L)2(dmso)4]?10thf. Fig. 2 Top: UV absorption spectra of [Ln2(H2L)2(dmso)4] in thf. Bottom: visible absorption spectrum for Ln = Eu showing the LMCT and 5D0 �æ 7F0 transitionsJ. Chem. Soc., Dalton Trans., 1998, Pages 505–510 507 Table 1 Selected averaged distances (Å) and angles (8) for the co-ordination sphere around the two europium(III) ions in [Eu2(H2L)2(dmso)4]?10thf (a and b refer to the dmso O atoms, b corresponding to the molecule included in the calixarene cavity) Eu ? ? ? Eu9 Eu]O(1) Eu]O(2) 3.891(2) 2.33(1) 2.54(1) Eu]O(3) Eu]O(4) Eu]O(5) 2.56(2) 2.35(1) 2.48(1) Eu]O(59) Eu]O(a) Eu]O(b) 2.32(1) 2.32(1) 2.36(1) O(x)]Eu]O(y) angles y x 2 3 4 5 59 a b 1 79.8(8) 134.9(11) 144.0(5) 73.2(5) 78.5(6) 100.4(14) 92.0(19) 2 55.4(6) 133.8(7) 132.7(6) 138.9(13) 74.8(14) 67.1(7) 3 78.8(9) 132.4(7) 138.7(7) 74.9(11) 67.5(8) 4 73.0(5) 79.5(7) 101.0(15) 91.5(16) 5 71.8(13) 147.2(9) 75.7(8) 59 75.4(10) 147.5(9) a 137.0(6) Fig. 3 Stereoscopic view of a dimeric [Eu2(H2L)2(dmso)4] unit Details of the experimental conditions, cell data, structure solution and refinement are given in the Experimental section. Selected bond lengths and angles for the co-ordination sphere around EuIII are given in Table 1 (the complete numbering scheme is described in SUP 57328).The unit cell contains two crystallographically independent and neutral [Eu2(H2L)2(dmso)4] dimers comprised of two H2L32 anions, two europium(III) ions and four dmso molecules co-ordinated to the metal ions. Within a dimer, an inversion centre is located halfway from the two EuIII. Each eight-coordinated EuIII is bonded to the five oxygen atoms of a calix- [5]arene trianion, two dmso oxygen atoms and a eighth oxygen from the second calix[5]arene anion (see the ORTEP24 view in Fig. 3). The latter, O(59), bridges the two europium ions, as does O(5), O(5) and O(59) being related by the inversion centre. The co-ordination polyhedra of the two ions (SUP 57328) are severely distorted as the analysis of the Eu]O bond distances shows. The Eu]O(2) and Eu]O(3) bond lengths (2.54 and 2.56 Å, respectively) are longer than the other Eu]O (calixarene) bonds. Since two of the five phenol groups remain protonated, we think this lengthening is typical of Eu]OH bonds similarly to what was observed for the parent complex with a calix[4]- arene.13 The average Eu]O distance is 2.41(10) Å resulting in an effective ionic radius, according to Shannon’s definition,25 of 1.10 Å (with rO = 1.31 Å), in reasonable agreement with the literature value of 1.07 Å for eight-co-ordination.26 The two bridging oxygen atoms form a parallelepiped O(5)]Eu(1)] O(59)]Eu(19) in which the metal ions are separated by 3.89 Å.A special feature of the structure is the bonding of one dmso molecule to EuIII through the hydrophobic cavity of the calixarene, thus combining co-ordination to the metal ion and inclusion in the calixarene, in a similar way to that reported for the co-ordination of pyridine N-oxide in Na8[Tb4(C5H5NO)4- (H2O)18L94] (L9 = p-sulfonatocalix[5]arene).19 In the acetone and acetate clathrates of H5L the macrocycle adopts a typical cone conformation,18,27 which is not the case in the presently described structure.The phenol rings B and C are strongly bent, with the oxygen atom pointing inside the hydrophobic cavity, while the methylene group bridging these rings points outward, in contrast to the four other bridging methylene groups. The phenol rings A, D and E remain in the cone conformation and the whole configuration can be described as distorted toward a 1,2-partial cone conformation (Fig. 4). This X-ray analysis confirms that the reaction of H5L with a dmso adduct of a nitrate salt of europium in the presence of a strong base leads to the isolation of a dimeric complex, as for the parent p-tert-butylcalix[4]arene complexes with EuIII 13 or with ZnII, AlIII, FeIII and TiIV.29 The Eu ? ? ? Eu9 distances in the two dimeric moieties with calix[4]- and calix-[5]arenes are very similar, 3.91 Å for the former and 3.89 Å for H5L.These distances are also close to those measured in the bimetallic complexes [Eu2L0(dmf)5] (dmf = dimethylformamide): 3.69 (L0 = hexaanion of p-tert-butylcalix[8]arene 11) and 3.81 Å (L0 = hexaanion of p-nitrocalix[8]arene 30).Luminescent properties Upon excitation through the ligand p æÆ p* transition, solutions of the three complexes in thf display a sizeable emission band in the range 33 000–27 800 cm21, with a maximum at 31 150 cm21, assigned to luminescence from the 1pp* state of Fig. 4 Conformation of the ligand in [Eu2(H2L)2(dmso)4] as drawn with the PACHA program28508 J.Chem. Soc., Dalton Trans., 1998, Pages 505–510 the ligand (Fig. 5). In the case of GdIII the triplet-state emission can also be seen as a weaker band in the range 27 000–20 000 cm21 and with a maximum at 23 640 cm21. Assignment to the 3pp* state relies on its decay time (a few tenths of ms). For the europium(III) compound no metal-centred luminescence is observed. Previous studies on p-tert-butylcalix[8]arene complexes with EuIII 15,17 have pointed to a low-lying LMCT state being responsible for the almost complete quenching of the EuIII-centred luminescence in these compounds and the same explanation probably holds for the calix[5]arene dimeric complex. In contrast to EuIII, the complex of TbIII displays a luminescence pattern characteristic of the metal-centred 5D4 æÆ 7FJ (J = 0–6) transitions (Fig. 5), revealing a ligandto- metal energy-transfer process. The quantum yield of the dimeric complex 1023 M in thf amounts to 5.1%. Given the absence of specifically designed chromophoric groups attached to the calix[5]arene, this figure appears to be quite encouraging for the development of calix[5]arene-based luminescent stains.In the solid state the lifetime of the Tb(5D4) exc state, obtained by direct laser excitation to 5D4, amounts to 1.12 ± 0.04, 0.90 ± 0.04, and 0.21 ± 0.01 ms at 10, 70 and 295 K, respectively. This relatively short lifetime may be partially due to the complexation of the two hydroxyl groups O(2)H and O(3)H.In [Tb2L0(dmf)4] (L0 = hexaanion of p-tert-butylcalix[ 8]arene) where each TbIII is co-ordinated to one phenoxyl group only, t(5D4) = 1.52 ms at 77 K 14 and similar lifetimes have been reported at room temperature for complexes with lowerrim substituted calix[4]arenes with carbamoyloxy groups (1.5 ms in water for R = OCH2CONEt2 31 and 1.79 ms in methanol for R = OCH2CONH2 32). The large temperature dependence observed between 10 and 295 K and leading at room temperature to a lifetime shorter than that observed for the aqua-ion (0.42 ms 33) is noteworthy.Such dependence has been assigned to back energy transfer from the terbium(III) ion to the ligand triplet state,34 a phenomenon often observed for TbIII included in supramolecular edifices.35,36 Analysis of the temperature-dependent t(5D4) lifetime between 50 and 295 K according to an Arrhenius relation of the type ln(t21 2 t0 21) = A 1 (Ea/RT),34 where t is the lifetime at temperature T, t0 that in the absence of quenching (taken here at 10 K) and Ea the activation energy for the quenching process, is shown on Fig. 6.A linear regression leads to Ea = 180 ± 20 cm21, smaller than the values reported 34 for triacetate chelates with 1,10-phenanthroline (900–2000 cm21). If back energy transfer from the 5D4 excited state of terbium to the triplet state occurs in [Tb2(H2L)(dmso)4], such a small activation barrier can be related to vibrational motion in the complex and the deactivation pathway may be phonon assisted, for instance by Ln]O vibrations, which occur at around 220 cm21.37 Fig. 5 Luminescence spectra of 1023 M [Ln2(H2L)2(dmso)4] in thf, at room temperature Conclusion The data presented here point to p-tert-butylcalix[5]arene being an interesting ligand for lanthanide complexation. In a lowpolarity solvent such as acetonitrile, stabilisation of its anionic forms is achieved both by intra- and inter-molecular hydrogen bonds. In the presence of a strong base, such as NaH, deprotonation is favoured and the macrocyclic anion reacts with lanthanide trivalent ions to form dimeric complexes, both in solution and in the solid state, as demonstrated by ES mass spectra and X-ray diffraction data.The energy-level scheme reproduced in Fig. 7 summarises the photophysical properties of the isolated complexes. If the europium dimer proves to be inefficient as a luminescent probe, the metal-centred luminescence being completely quenched by the LMCT state (cf.the respective energy of this state and of the ligand 3pp* state), the terbium assembly conveniently absorbs UV light and transfers its energy from the ligand 3pp* state to the terbium 5D4 excited state. The overall efficiency of this transfer remains modest for two reasons. (i) The occurrence of a back-transfer process: Fig. 7 shows the near overlap between the Tb(5D4) level and the low-energy tail of the 3pp* state, so that a low-energy phonon-assisted back transfer will be easily achieved (cf.the activation energy of 180 cm21 found for this process). (ii) More significantly, the intersystem 1pp* æÆ 3pp* conversion Fig. 6 Arrhenius plot of the Tb(5D4) lifetime in [Tb2(H2L)2(dmso)4] vs. the reciprocal temperature (see text) Fig. 7 Schematic energy-level diagram for H2L32 and its dimeric complexes with EuIII and TbIII (I.C. = intersystem crossing)J. Chem. Soc., Dalton Trans., 1998, Pages 505–510 509 has a poor yield, as exemplified by the persistence of the singlet-state ligand-centred luminescence in the spectrum of the dimeric metal complex.The ligand used in this study does not bear specifically designed chromophoric groups and it may be envisaged that the relatively small quantum yield might be improved by grafting such groups on the calix- [5]arene skeleton. Finally, the inclusion-complexation properties of the calix- [5]arene evidenced for dmso may be of interest to probe the inclusion of organic molecules in the calixarene cavity, since co-ordination to the lanthanoid(III) ions will change the photophysical properties of the latter.Experimental Synthesis and characterisation of the complexes Solvents and starting materials other than p-tert-butylcalix[5]- arene (Acros) were from Fluka (Buchs, Switzerland) and used without further purification unless otherwise stated. Tetrahydrofuran was distilled over Na and acetonitrile was treated with CaH2 and P2O5.38 The dmso adducts of the lanthanoid salts were prepared from the oxides (Rhône- Poulenc, 99.99%) 39 and their lanthanoid content determined by titration with Titriplex III (Merck) in the presence of urotropine (1,3,5,7-tetraazatricyclo[3.3.1.13,7]decane) and xylene orange.40 The complexes were synthesized according to the following general procedure.The compound H5L (1.3 × 1024 mol, 1 equivalent) was dissolved in dry thf (0.12 dm3) under a nitrogen atmosphere.Sodium hydride (4.55 × 1024 mol, 3.5 equivalents, 60% in oil) was added and the solution stirred for 2 h before Ln(NO3)3?xdmso (x = 3.18, 2.97 or 3.25 for Ln = Eu, Gd or Tb) (1.3 × 1024 mol, 1 equivalent) was added and the mixture stirred for 18 h at room temperature. A white precipitate of NaNO3 slowly appeared and was filtered off. The solution was concentrated under reduced pressure until a precipitate formed, which was separated from the mother-liquor by centrifugation, washed with cold thf (2 cm3), centrifuged, and separated from the supernatant liquid. The resulting efflorescent solid was dried under vacuum [12 h, 40 8C, 1022 Torr (ca. 1.33 Pa)] to yield 66, 68 and 58% (Eu, Gd and Tb) of [Ln2(H2L)2(dmso)x] as a white (Tb and Gd) or orange (Eu) powder [Found: C, 65.33; H, 7.12. Calc. for C116H152Eu2O13S3 (x = 3): C, 64.67; H, 7.11. Found: C, 63.17; H, 7.16. Calc. for C118H158Gd2O14S4 (x = 4): C, 63.18; H, 7.10. Found: C, 62.93; H, 6.92.Calc. for C118H158O14S4Tb2 (x = 4): C, 63.08; H, 7.09%]. No nitrogen was found in any compound. Elemental analyses were performed by Dr. H. Eder (Microchemical Laboratory, University of Geneva). Infrared spectra were measured on a Mattson Alpha Centauri FT spectrometer as KBr pellets. The absorption maxima (in cm21) were identical, within experimental error, for all three complexes: 3305, 3409 (nOH); 3049 (nCHaro); 2957, 2868 (nCH3); 1605, 1482, 1465 (dCH3); 1461 (nCH2); 1023 (ndmso).Physicochemical measurements Electrospray mass spectra were measured on a Finnigan SSQ 710C spectrometer with 1024 M solutions in thf or acetonitrile. Spectrophotometric titrations were performed at 298 ± 0.2 K on a UV/VIS Perkin-Elmer Lambda 7 spectrometer with 1 cm quartz cells. In a typical experiment, a 1024 M solution (10 cm3) of H5L in dry acetonitrile containing 7.9 × 1023 M NEt4ClO4 as inert salt was titrated by increasing amounts of a 1023 M solution of Et3N in the same solvent, delivered by a Metrohm Dosimat E 535 and recorded for Et3N:H5L ratios between 0 and 40: 1.The spectra were fitted using the SPECFIT program. 41 Factor analysis revealed the presence of four different species, which were described according to the model given in the Results and Discussion section. Luminescence spectra were recorded on a Perkin-Elmer LS-50 spectrofluorimeter, using a 300 nm excitation filter for the terbium-centred emission. The quantum yield was determined in degassed thf (90 ppm water) at room temperature as previously described42 using [Tb(terpy)3][ClO4]3 (terpy = 2,29:69,20-terpyridine) in degassed acetonitrile (100 ppm water) as a secondary standard {absolute quantum yield 4.7%, as determined by the same method with [Ru(bipy)3][ClO4]2 (bipy = 2,29-bipyridine) in air-saturated water as standard43}.The concentrations used were 1023 M for both the terbium dimer and the reference (to avoid decomplexation), with lexc = 320.0 (sample) and 365.0 nm (reference).The 5D4æÆ7FJ transitions with J = 3–6 only were considered to obtain the integrated luminescence intensity. Neglecting the weak transitions to J = 0, 1 and 2 (<5% of the total integrated intensity) introduces negligible error and avoids corrections for the Rayleigh diffusion band interfering with these transitions. The luminescence spectra were corrected for the residual non TbIII-centred emission. The Tb(5D4) lifetimes were determined on a previously described instrumental set-up 44 using microcrystalline samples and selective laser excitation to the 5D4 level (486 nm); the reported lifetimes are averages of at least three determinations; biexponential fitting of the curves was used, revealing a residual short lifetime due to an instrumental artefact, as confirmed by blank measurements.Crystallography The compound [Eu2(H2L)2(dmso)4]?10thf crystallised as yellow-orange platelets which were incorporated into a drop of Hostinert 216 oil and frozen to 170 K (Oxford Cryostream). Such a manipulation prevented observation of the crystals under orthoscopic conditions but an episcopic inspection revealed thin boundaries subdividing the platelets into rectangular blocks which were warped with respect to each other.Data were collected on a Stoe IPDS system equipped with Mo- Ka radiation (l 0.710 73 Å); 200 images in f intervals of 18 were exposed for 6 min each. The crystal–image plate distance (80 mm) corresponded to a 1.13 Å21 resolution. Other experimental details are reported in Table 2.The indexing program of the IPDS system found the cell parameters from 2000 reflections (20 images), but only 1– 4 of the peaks belonged to the cell, as a result of the poor quality of the crystals (several attempts to produce better crystals failed). The defects of the crystals materialised in diffraction patterns containing large spots and in a large background noise all over the reciprocal space, which made it difficult to evaluate a reasonable effective mosaic spread for the data. Following a referee’s suggestion, we attempted to solve the structure in a more symmetrical monoclinic cell.This resulted in a slightly worse agreement factor and did not change the topology of the dimer. We therefore think that the cell proposed in Table 2 soundly describes the important features of the diffraction pattern, despite the large final R values.Data were corrected for Lorentz-polarisation effects (the intensity decay during the measurement was negligible) and the structure was solved with SHELXTL.45 A first solution was found in space group P1, but after successful refinement the program MISSYM46 revealed an inversion centre between the Eu atoms of the dimer. Therefore, we refined two half dimers related by a pseudo-translation (��� c*). One isotropic displacement parameter was used for all atoms of the phenol rings, one for the atoms from the tert-butyl groups and one for the atoms of the methylene bridges. Benzene rings were restrained to be flat and to have literature C]C bond lengths,47 and tertbutyl groups and dmso molecules were restrained to conventional bond lengths and angles.47 The dmso molecules were refined using both isotropic and anisotropic displacement parameters.510 J.Chem. Soc., Dalton Trans., 1998, Pages 505–510 CCDC reference number 186/812. See http://www.rsc.org/suppdata/dt/1998/505/ for crystallographic files in .cif format.Acknowledgements We gratefully acknowledge Dr. Claude Piguet and Fabien Renaud for measuring the absolute quantum yield of [Tb- (terpy)3][ClO4]3, Cédric Sager for recording ES mass spectra and Ms. Véronique Foiret for her technical assistance in luminescence measurements. We thank the Fondation Herbette (Lausanne) for a gift of spectroscopic equipment. This work is supported through grants from the Swiss National Science Foundation.References 1 C. D. Gutsche, Calixarenes, Royal Society of Chemistry, Cambridge, 1989. 2 L. Giannini, E. Solari, A. Zanottigerosa, C. Floriani, A. Chiesivilla and C. Rizzoli, Angew. Chem., Int. Ed. Engl., 1997, 753. 3 R. Arnecke, V. Böhmer, R. Cacciapaglia, A. D. Cort and L. Mandolini, Tetrahedron, 1997, 53, 4901. 4 T. Haino, M Yanase and Y. Fukazawa, Angew. Chem., Int. Ed. Engl., 1997, 259. 5 Calixarenes: A Versatile Class of Macrocyclic Compounds, eds.J. Vicens and V. Böhmer, Kluwer, Dordrecht, 1991. 6 F. Arnaud-Neu, S. Cremin, S. Harris, M. A. McKervey, M.-J. Schwing-Weill, P. Schwinte and A. Walker, J. Chem. Soc., Dalton Trans., 1997, 329. 7 N. Sabbatini, M. Guardigli, I. Manet, R. Ungaro, A. Casnati, C. Fischer, R. Ziessel and G. Ulrich, New J. Chem., 1995, 19, 137. 8 F. J. Steemers, W. Verboom, D. N. Reinhoudt, E. B. Vandertol and J. W. Verhoeven, J. Am. Chem. Soc., 1995, 117, 9408. 9 H. Matsumoto and S. Shinkai, Chem.Lett., 1994, 901. 10 J.-M. Lehn, Supramolecular Chemistry. Concepts and Perspectives, VCH, Weinheim, New York, Basel, Cambridge, Tokyo, 1995. 11 B. M. Furphy, J. M. Harrowfield, D. L. Kepert, B. W. Skelton, A. H. White and F. R. Wilner, Inorg. Chem., 1987, 26, 4231. 12 L. M. Engelhardt, B. M. Furphy, J. M. Harrowfield, D. L. Kepert, A. H. White and F. R. Wilner, Aust. J. Chem., 1988, 41, 1465. Table 2 Crystal data and structure refinement for [Eu2(H2L)2- (dmso)4]?10thf Formula M Colour, shape Crystal system Space group a/Å b/Å c/Å a/8 b/8 g/8 U/Å3 Z F(000) Dc/g cm23 m/mm21 q Range for data collection/8 hkl Ranges of reflections Reflections collected Independent reflections Refinement method Data, restraints, parameters Goodness of fit on F 2 R1, wR2 [I > 2s(I)] (all data) Extinction coefficient Residuals in final difference map/ e Å23 C118H158Eu2O14S4?10C4H8O 2961.87 Yellow-orange platelets Triclinic P1� 16.442(3) 16.496(3) 32.629(7) 97.97(3) 97.51(3) 114.26(3) 7816(3) 2 3088 1.245 0.912 1.93 to 24.05 218 to 18, 218 to 18, 237 to 37 44 486 22 971 (Rint = 0.2389) Full-matrix least squares on F 2 22 971, 229, 724 0.822 0.112, 0.247 0.281, 0.317 0.000 24(14) 1.106 and 21.179 13 B.M. Furphy, J. M. Harrowfield, J. S. Ogden, B. W. Skelton, A. H. White and F. R. Wilner, J. Chem. Soc., Dalton Trans., 1989, 2217. 14 P. Froidevaux and J.-C. G. Bünzli, J. Phys. Chem., 1994, 98, 532. 15 J.-C. G. Bünzli, P. Froidevaux and J. M.Harrowfield, Inorg. Chem., 1993, 32, 3306. 16 J.-C. G. Bünzli and J. M. Harrowfield, in Calixarenes: a Versatile Class of Macrocyclic Compounds, eds. J. Vicens and V. Böhmer, Kluwer, Dordrecht, 1991, pp. 211–231. 17 J.-C. G. Bünzli and F. Ihringer, Inorg. Chim. Acta, 1996, 246, 195. 18 M. Coruzzi, G. D. Andreetti, V. Bocchi, A. Pochini and R. Ungaro, J. Chem. Soc., Perkin Trans. 2, 1982, 1133. 19 J. W. Steed, C. P. Johnson, C. L. Barnes, R. K. Juneja, J. L. Atwood, S.Reilly, R. L. Hollis, P. H. Smith and D. L. Clark, J. Am. Chem. Soc., 1995, 117, 11 426. 20 L. N. Bykova and S. I. Petrov, Russ. Chem. Rev. (Engl. Transl.), 1972, 41, 975. 21 M. Backes, V. Böhmer, G. Ferguson, C. Grüttner, C. Schmidt, W. Vogt and K. Ziat, J. Chem. Soc., Perkin Trans. 2, 1997, 1193. 22 C. D. Gutsche, M. Iqbal and A. Iftikhar, J. Am. Chem. Soc., 1987, 109, 4314. 23 J.-C. G. Bünzli and C. Mabillard, J. Less-Common Met., 1983, 94, 317. 24 C. K. Johnson, ORTEP, Report ORNL-5138, Oak Ridge National Laboratory, Oak Ridge, TN, 1976. 25 R. D. Shannon, Acta Crystallogr., Sect. A, 1976, 32, 751. 26 G. R. Choppin, in Lanthanide Probes in Life, Chemical and Earth Sciences. Theory and Practice, eds. J.-C. G. Bünzli and G. R. Choppin, Elsevier, Amsterdam, 1989, ch. 1, pp. 1–41. 27 J. L. Atwood, R. K. Juneja, P. C. Junk and K. D. Robinson, J. Chem. Crystallogr., 1994, 24, 573. 28 M. Henry, Program PACHA, Partial Atomic Charge Analysis, Université Louis Pasteur, Strasbourg, 1993. 29 M. M. Olmstead, G. Sigel, H. Hope, X. Xu and P. P. Power, J. Am. Chem. Soc., 1985, 107, 8087; J. L. Atwood, S. G. Bott, C. Jones and C. L. Raston, J. Chem. Soc., Chem. Commun., 1992, 1349; M. G. Gardiner, S. M. Lawrence, C. L. Raston, B. W. Skelton and A. H. White, Chem. Commun., 1996, 2491. 30 J.-C. G. Bünzli, F. Ihringer, P. Dumy, C. Sager and R. D. Rogers, J. Chem. Soc., Dalton Trans., 1998, preceding paper. 31 N. Sabbatini, M. Guardigli, A. Mecati, V. Balzani,ro, E. Ghidini, A. Casnati and A. Pochini, J. Chem. Soc., Chem. Commun., 1990, 878. 32 E. M. Georgiev, J. Clymire, G. L. McPherson and D. M. Roundhill, Inorg. Chim. Acta, 1994, 227, 293. 33 J.-C. G. Bünzli, in Lanthanide Probes in Life, Chemical and Earth Sciences. Theory and Practice, eds. J.-C. G. Bünzli and G. R. Choppin, Elsevier, Amsterdam, 1989, ch. 7, pp. 219–293. 34 M. Kleinerman and Choi Sang-Il, J. Chem. Phys., 1968, 49, 3901. 35 G. Blasse, G. J. Dirksen, D. van der Voort, N. Sabbatini, S. Perathoner, J.-M. Lehn and B. Alpha, Chem. Phys. Lett., 1988, 146, 347. 36 J.-M. Lehn, in Frontier in Supramolecular Organic Chemistry and Photochemistry, eds. H.-J. Schneider and H. Dürr, Verlag Chemie, Weinheim, 1991, pp. 1–28. 37 J.-C. G. Bünzli, J.-R. Yersin and C. Mabillard, Inorg. Chem., 1982, 21, 1471. 38 D. D. Perrin and W. L. F. Armarego, Purification of Laboratory Chemicals, Pergamon, Oxford, 1988. 39 S. K. Ramalingam and S. Soundararajan, J. Inorg. Nucl. Chem., 1967, 29, 1763. 40 G. Schwarzenbach, Complexometric titrations, Chapman & Hall, London, 1957. 41 H. Gampp, M. Maeder, C. J. Meyer and A. D. Zuberbühler, Talanta, 1985, 32, 257. 42 S. Petoud, J.-C. G. Bünzli, K. J. Schenk and C. Piguet, Inorg. Chem., 1997, 36, 1345. 43 K. Nakamaru, Bull. Chem. Soc. Jpn., 1982, 55, 2697. 44 J.-C. G. Bünzli and A. Milicic-Tang, Inorg. Chim. Acta, 1996, 252, 221. 45 SHELXTL 5.05, Siemens Analytical X-Ray Instruments, Madison, WI, 1996. 46 Y. Le Page, J. Appl. Crystallogr., 1988, 21, 983. 47 International Tables for Crystallography, International Union of Crystallography, Dordrecht, 1995. Received 22nd August 1997; Paper 7/06152H © Copyright 1998 by the Royal Society of Chemistry

ISSN:1477-9226    

DOI:10.1039/a706152h

出版商:RSC    

年代:1998

数据来源: RSC