摘要:

DALTON J. Chem. Soc., Dalton Trans., 1997, Pages 2221–2225 2221 [HN(CH2CH2)3NH]3[Fe8(HPO4)12(PO4)2(H2O)6]: an organically templated iron phosphate with a pillared layer structure Kwang-Hwa Lii * and Yuh-Feng Huang Institute of Chemistry, Academia Sinica, Taipei, Taiwan An organically templated iron phosphate, [HN(CH2CH2)3NH]3[Fe8(HPO4)12(PO4)2(H2O)6], has been synthesized under solvothermal conditions and characterized by single-crystal X-ray diffraction, Mössbauer spectroscopy and thermogravimetric analysis.The compound crystallizes in the trigonal space group P3� c1 (no. 165) with a = 13.5274(5), c = 19.2645(6) Å, U = 3052.9(3) Å3 and Z = 2. The structure consists of layers of corner-sharing FeO6 and FeO5(OH2) octahedra and PO4 and PO3(OH) tetrahedra which are pillared through additional FeO6 octahedra to form a three-dimensional framework structure. The framework contains a two-dimensional array of intersecting channels in which the charge compensating diprotonated 1,4-diazabicyclo[2.2.2]octane cations reside.The framework is closely related to that of an imidazole encapsulating indium phosphate, [H3O][C3N2H5]3[In8(HPO4)14(H2O)6]?5H2O. Recently we have synthesized a large number of ternary iron phosphates by high temperature, high pressure hydrothermal methods. These compounds present a variety of complex crystal structures and thus interesting magnetic properties and are a challenge to complete characterization. Their structures cover discrete FeO6 octahedra, FeO5 trigonal bipyramids, dimers of corner-sharing, edge-sharing or face-sharing FeO6 octahedra, trimeric, tetrameric units of Fe]O polyhedra and infinite chains of FeO6 octahedra sharing either trans or skew edges.They include iron(II), iron(III) and mixed-valence compounds. 1 However, syntheses performed under high temperature, high pressure hydrothermal conditions can only yield dense phases. To generate large internal micropore volumes within the inorganic oxide frameworks, the syntheses are usually carried out under mild hydrothermal conditions (<200 8C) with a variety of organic cationic templates. Some of the microporous materials were synthesized from predominantly non-aqueous systems.Recently, an organically templated iron phosphate [H3NCH2CH2NH3]0.5[Fe(OH)(PO4)] was reported.2,3 It adopts a layered structure with the ethylenediammonium cations in the interlayer region, which is isotypic with the layered gallophosphate obtained in ethylene glycol (HOCH2CH2OH).4 A fluorinated iron phosphate with an open structure, [C6H14N2][Fe4(PO4)4F2(H2O)3] was reported recently.5 The framework consists of large cages limited by three-, four-, six- and eight-membered windows in which the diprotonated 1,4-diazabicyclo[2.2.2]octane cations are located.We have also become interested in the synthesis of microporous metal phosphates under mild solvothermal conditions because of their rich structural chemistry and potential applications as molecular sieves, ion-exchange materials and catalysts.Here we describe the solvothermal synthesis and characterization of a 1,4-diazabicyclo[2.2.2]octane encapsulating iron phosphate, [HN(CH2CH2)3NH]3[Fe8(HPO4)12(PO4)2(H2O)6], which exhibits a pillared layer structure. Its inorganic oxide framework is closely related to that of an indium phosphate with encapsulated imidazolium cation as the template, [H3O]- [C3N2H5]3[In8(HPO4)14(H2O)6]?5H2O.6 Experimental Synthesis and initial characterization The synthesis was carried out in a Teflon-lined acid digestion * E-Mail: lii@chem.sinica.edu.tw bomb (23 cm3) under autogenous pressure.The reaction of FeCl3?6H2O (2.5 mmol), 1,4-diazabicyclo[2.2.2]octane (DABCO) (7.5 mmol), H3PO4 (7.5 mmol), n-butanol (3 cm3) and water (7 cm3) at 180 8C for 3 d followed by slow cooling at 10 8C h21 to room temperature produced [HN(CH2CH2)3- NH]3[Fe8(HPO4)12(PO4)2(H2O)6] 1 as colourless crystals and a small amount of a green material.The colourless crystals were manually separated from the green material to give pure compound 1 as judged by visual microscopic examination and by comparison of the X-ray powder pattern to the pattern simulated from the atomic coordinates derived from a single-crystal study. The yield was 70% based on iron. Energy-dispersive X-ray fluorescence analysis using a JEOL analytical electron microscope confirmed the presence of Fe and P but no Cl in the colourless crystals {Found: C, 9.60; H, 2.98; N, 3.73%.Calc. for [HN(CH2CH2)3NH]3[Fe8(HPO4)12(PO4)2(H2O)6]: C, 9.66; H, 2.97; N, 3.75%, confirming that 1,4-diazabicyclo[2.2.2]octane is present in the compound}. The pure sample was used for thermogravimetric analysis and Mössbauer spectroscopy measurements. Thermogravimetric analysis of 1 was performed on a Perkin- Elmer TGA7 thermal analyser: the sample was heated to 950 8C at 10 8C min21 in air.In order to characterize the decomposition products, an experiment was performed in which 1 was heated at 325, 400, 600 and 800 8C for 8 h in a platinum crucible in air. The product of each heat treatment was analysed by powder X-ray diffraction at room temperature. The 57Fe Mössbauer measurements were made on a constant-acceleration instrument at 300 K. Isomer shift is reported with respect to an iron foil standard. Single-crystal X-ray diVraction Most of the colourless crystals were not suitable for singlecrystal X-ray structure analysis as indicated from peak profile analysis.Many were selected before a satisfactory crystal was obtained. A small tabular crystal of dimensions 0.1 × 0.1 × 0.025 mm was mounted on a Siemens Smart-CCD diffractometer equipped with a normal focus, 3 kW sealed tube X-ray source. Intensity data were collected in 1200 frames with increasing w (width of 0.38 per frame). The orientation matrix and unit cell dimensions were determined by a least-squares fit of 4881 reflections with 2.5 < 2q < 508.Absorption correction was based on 4356 symmetry-equivalent reflections using the SHELXTL PC program package (Tmin, Tmax = 0.751, 0.884).7 On the basis of systematic absences, statistical analysis of the2222 J. Chem. Soc., Dalton Trans., 1997, Pages 2221–2225 intensity distribution, and successful solution and refinement of the structure, the space group was determined to be P3� c1 (no. 165). The structure was solved by direct methods: the iron and phosphorus atoms were first located and the oxygen, carbon and nitrogen atoms were found in Fourier-difference maps. The hydrogen atoms were not located. Bond-valence calculations 8 indicated that O(2), O(6), O(9) and O(11) had valence sums of 1.10, 1.09, 1.18 and 0.36, respectively, and all other oxygen atoms had values between 1.75 and 1.93. The value 0.36 indicates that O(11) is a water oxygen. Two more hydrogen atoms at general positions must be included to balance the charge.Valence sums of 1.10 and 1.09 suggest that O(2) and O(6) are hydroxo oxygens. Atom O(9) sits on a three-fold axis and is hydrogen bonded to three DABCO cations with the O? ? ?N distances at 2.78 Å. The final cycles of least-squares refinement on F including atomic coordinates and anisotropic thermal parameters for all atoms converged at R = 0.0634 and R9 = 0.0589. The final Fourier-difference maps were flat to (Dr)max,min = 0.92, 20.84 e Å23.Neutral-atom scattering factors were used for all atoms. Anomalous dispersion and secondary extinction corrections were applied. Structure solution and refinement were performed on SHELXTL PC programs. Atomic coordinates, thermal parameters, and bond lengths and angles have been deposited at the Cambridge Crystallographic Data Centre (CCDC). See Instructions for Authors, J. Chem. Soc., Dalton Trans., 1997, Issue 1. Any request to the Fig. 1 Thermogravimetric analysis of [HN(CH2CH2)3NH]3[Fe8- (HPO4)12(PO4)2(H2O)6] in flowing air at 10 8C min21 Table 1 Crystallographic data for [HN(CH2CH2)3NH]3[Fe8(HPO4)12- (PO4)2(H2O)6] Formula M Crystal system Space group a/Å c/Å U/Å3 Z Dc/g cm23 F(000) m(Mo-Ka)/cm21 T/8C l/Å Maximum 2q/8 Reflections collected Unique reflections Observed unique reflections [I > 3s(I )] Weighting scheme Number of parameters Rint R9 b Goodness of fit (Dr)max,min/e Å23 C18Fe8H66N6O62P14 2239.15 trigonal P3� c1 13.5274(5) 19.2645(6) 3052.9(3) 2 2.436 2260 23.6 23 0.710 73 53.5 17 333 2516 1335 w21 = s2(F) 1 0.000 322F 2 165 0.0760 0.0634 0.0589 1.90 0.92, 20.84 a R = S||Fo| 2 |Fc||/S|Fo|.b R9 = [Sw(|Fo| 2 |Fc|)2/Sw|Fo|2]� �� . CCDC for this material should quote the full literature citation and the reference number 186/520. Results and Discussion Physical measurements The TG analysis showed weight loss in three steps (Fig. 1). The step in the temperature range from ª350 to ª500 8C is not resolved from the other two steps.On the basis of powder X-ray analysis, the decomposition products at 325, 400 and 600 8C are amorphous. The product at 800 8C is a mixture of Fe4(P2O7)3 and Fe(PO3)3.† At present we are unable to rationalize the decomposition mechanism of each step. The observed total weight loss of 27.7% between room temperature and 800 8C agrees well with that calculated for the loss of 15 H2O and 3 DABCO molecules (27.1%), as indicated from equation (1).[HN(CH2CH2)3NH]3[Fe8(HPO4)12(PO4)2(H2O)6] æÆ Fe4(P2O7)3 1 8/3 Fe(PO3)3 1 2/3 Fe2O3 1 15 H2O 1 3 N(CH2CH2)3N (1) Fig. 2 Mössbauer spectrum of [HN(CH2CH2)3NH]3[Fe8(HPO4)12- (PO4)2(H2O)6] at 300 K Table 2 Atomic coordinates for [HN(CH2CH2)3NH]3[Fe8(HPO4)12- (PO4)2(H2O)6] Atom x y z Fe(1) Fe(2) Fe(3) P(1) P(2) P(3) O(1) O(2) O(3) O(4) O(5) O(6) O(7) O(8) O(9) O(10) O(11) N(1) C(1) C(2) C(3) 0.855 6(1) 00 0.963 5(2) 0.632 1(2) 0.666 7 0.880 9(5) 0.081 2(6) 0.923 4(6) 0.983 6(9) 0.713 5(5) 0.702 5(5) 0.580 0(5) 0.542 8(6) 0.666 7 0.772 2(5) 0.939 0(5) 0.602 9(8) 0.625 8(9) 0.476 4(9) 0.645(1) 0.565 5(1) 00 0.812 1(2) 0.505 1(2) 0.333 3 0.686 0(5) 0.824 9(6) 0.859 7(5) 0.874 8(7) 0.564 1(5) 0.495 0(6) 0.574 1(5) 0.382 0(5) 0.333 3 0.443 2(5) 0.691 7(5) 0.496 2(8) 0.571 2(9) 0.411 7(9) 0.569(1) 0.450 37(7) 0.25 0.5 0.371 2(1) 0.543 9(1) 0.362 5(2) 0.383 1(3) 0.347 5(4) 0.312 8(3) 0.440 2(3) 0.483 7(3) 0.606 6(3) 0.571 1(3) 0.527 3(3) 0.281 8(5) 0.386 2(3) 0.529 3(3) 0.257 7(4) 0.318 9(5) 0.252 8(6) 0.192 5(6) † Fe4(P2O7)3, file number 36-318; Fe(PO3)3, 38-109; Joint Committee on Powder Diffraction Standards, International Center of Diffraction Data, Swarthmore, PA.J.Chem. Soc., Dalton Trans., 1997, Pages 2221–2225 2223 Table 3 Interatomic distances and bond valence sums (Ss) for [HN(CH2CH2)3NH]3[Fe8(HPO4)12(PO4)2(H2O)6] Fe(1)O6 octahedron* Fe(1) O(1) O(5) O(10) O(11) O(7a) O(8a) O(1) 1.973(7) 89.4(3) 94.2(3) 90.2(3) 92.0(3) 173.3(3) O(5) 2.806(8) 2.016(8) 92.9(3) 84.3(3) 172.1(2) 90.0(2) O(10) 2.850(9) 2.852(10) 1.916(6) 174.8(2) 94.7(3) 92.5(3) O(11) 2.915(9) 2.790(8) 2.138(6) 87.9(3) 83.1(3) O(7a) 2.849(9) 2.871(9) 2.865(9) 1.987(4) 87.7(2) O(8a) 2.853(8) 2.841(9) 2.758(9) 2.774(11) 2.017(6) Ss[Fe(1)]O] = 3.11 Fe(2)O6 octahedron* Fe(2) O(3a) O(3b) O(3c) O(3d) O(3e) O(3f) O(3a) 2.042(6) 88.5(2) 88.5(2) 88.2(4) 175.1(4) 95.0(4) O(3b) 2.851(7) 2.042(6) 88.5(2) 175.1(4) 95.0(4) 88.2(4) O(3c) 3.011(12) 2.042(3) 95.0(4) 88.2(4) 175.1(4) O(3d) 3.011(12) 2.843(14) 2.851(7) 2.042(6) 88.5(2) 88.5(2) O(3e) 2.843(14) 2.851(7) 2.851(7) 2.042(6) 88.5(2) O(3f) 2.851(7) 2.851(7) 2.843(14) 3.011(12) 2.042(3) Ss[Fe(2)]O] = 2.79 Fe(3)O6 octahedron* Fe(3) O(4a) O(4b) O(4c) O(4d) O(4e) O(4f) O(4a) 1.967(9) 180.0(0) 89.2(2) 90.8(2) 89.2(2) 90.8(2) O(4b) 2.802(12) 1.967(9) 90.8(2) 89.2(2) 90.8(2) 89.2(2) O(4c) 2.761(12) 1.967(7) 180.0(0) 89.2(2) 90.8(2) O(4d) 2.802(12) 2.761(12) 2.802(12) 1.967(7) 90.8(2) 89.2(2) O(4e) 2.761(12) 2.802(12) 2.761(12) 1.967(5) 180.0(0) O(4f) 2.761(12) 2.802(12) 2.761(12) 2.802(12) 1.967(5) Ss[Fe(3)]O] = 3.42 P(1)O4 tetrahedron P(1) O(1) O(2) O(3) O(4) O(1) 1.518(6) 107.6(5) 111.7(4) 108.6(4) O(2) 2.500(8) 1.581(9) 107.0(4) 107.5(5) O(3) 2.517(10) 2.496(13) 1.524(8) 114.2(6) O(4) 2.473(9) 2.506(14) 2.562(9) 1.527(8) Ss[P(1)]O] = 4.97 P(2)O4 tetrahedron P(2) O(5) O(6) O(7) O(8) O(5) 1.521(6) 108.5(4) 112.2(4) 112.5(4) O(6) 2.522(9) 1.585(8) 105.9(4) 104.4(4) O(7) 2.522(10) 2.475(12) 1.516(9) 112.8(4) O(8) 2.533(7) 2.456(8) 2.532(10) 1.524(6) Ss[P(2)]O] = 4.99 P(3)O4 tetrahedron P(3) O(9) O(10) O(10a) O(10b) O(9) 1.556(11) 107.4(3) 107.4(3) 107.4(3) O(10) 2.484(10) 1.528(5) 111.5(3) 111.5(3) O(10a) 2.484(10) 2.525(8) 1.528(8) 111.5(2) O(10b) 2.484(10) 2.525(8) 2.525(8) 1.528(4) Ss[P(3)]O] = 5.00 [HN(CH2CH2)3NH]21 cation N(1)]C(1) N(1)]C(3) C(2)]C(2a) C(1)]N(1)]C(2) C(2)]N(1)]C(3) N(1)]C(2)]C(2a) 1.48(1) 1.52(1) 1.52(3) 109.6(9) 110.7(9) 109.2(7) N(1)]C(2) C(1)]C(3a) C(1)]N(1)]C(3) N(1)]C(1)]C(3a) N(1)]C(3)]C(1a) 1.51(1) 1.54(2) 109.2(9) 109.0(9) 108.5(10) O(9) ? ? ? N(1) 2.78(1) (3×) * The distances between trans oxygen atoms are not shown.The room-temperature Mössbauer spectrum of 1 (Fig. 2) can be least-squares fitted by one doublet and does not show three Fe components as observed in the crystal structure. The obtained parameters are d (isomer shift) = 0.45 mm s21, DEQ (quadrupole splitting) = 0.35 mm s21 and G (full width at halfheight) = 0.47 mm s21.The peaks are broader than those for a thin iron calibration foil (G = 0.30 mm s21). The isomer shift is characteristic of high-spin FeIII. According to Menil,9 the usual ranges of isomer shifts in oxides are 0.29–0.50 and 1.03–1.28 mm s21 for FeIII and FeII in six-co-ordination, respectively. Therefore, the composition of 1 is further defined by Mössbauer spectroscopy and TG analysis. Crystal structure The crystallographic data are listed in Table 1.Atomic coordinates, interatomic distances, bond angles and bond-valence sums are given in Tables 2 and 3, respectively. The iron atoms Fe(2) and Fe(3) are at special positions with local symmetries D3 and S6, respectively. The atoms P(3) and O(9) sit on three-fold axes and all other atoms are at general positions. The Fe and P atoms are six- and four-co-ordinate, respectively. Both the Fe(2)O6 and Fe(3)O6 octahedra are quite regular.The Fe(1)O6 octahedron is considerably more distorted. The Fe(1)]O(11) distance is the longest because O(11) is the water oxygen. The2224 J. Chem. Soc., Dalton Trans., 1997, Pages 2221–2225 Fe(2)]O distances are significantly longer than those of Fe(3)]O. Both P(1) and P(2) have a terminal P]OH group as shown by the unsatisfied valence sums for O(2) and O(6) and the longer P]O distances [P(1)]O(2) 1.581 and P(2)]O(6) 1.585 Å].The P(3)]O(9) distance is also longer at 1.556 Å. The valence sum of O(9) is satisfied by forming three O? ? ?H]N bonds with three different protonated DABCO cations. In the indium phosphate [H3O][C3N2H5]3[In8(HPO4)14(H2O)6]?5H2O all three distinct phosphorus atoms form hydrogen phosphate groups. The structure of [HN(CH2CH2)3NH]3[Fe8(HPO4)12(PO4)2- (H2O)6], viewed along the [100] direction, is shown in Fig. 3. It consists of macroanionic sheets normal to [001], in which the Fe(3)O6 and Fe(1)O5(OH2) octahedra share corners with the PO4 and PO3(OH) tetrahedra in an alternating manner to form four- and six-membered rings (Fig. 4). The connectivity is the same as that in [H3O][C3N2H5]3[In8(HPO4)14(H2O)6]?5H2O. The sheets are linked through Fe(2)O6 pillars to generate inter- Fig. 3 Polyhedral representation of the [HN(CH2CH2)3NH]3[Fe8- (HPO4)12(PO4)2(H2O)6] structure viewed along the a axis. Open circles, C atoms; solid circles, N atoms secting channels parallel to the ·100Ò directions in which the diprotonated DABCO cations reside.Within the channels are 14-membered rings formed by seven iron–oxygen octahedra and seven phosphate tetrahedra (Fig. 5). The terminal P(3)]O(9) groups are directed towards the interlayer region and receive three hydrogen bonds from neighbouring DABCO cations. Intralayer hydrogen bonds also exist as inferred from the O? ? ? O distances. In the indium compound the channels contain positionally disordered imidazolium cations and water molecules.One-sixth of the non-framework water molecules are protonated to achieve charge balance with the framework. The compound described here is a second example of an organic molecule encapopen-framework structure of iron phosphate. Its inorganic oxide framework is similar to that of an indium phosphate. It provides another example of a large class of metastable phases of open-framework structures which may be accessible within hydrothermal reaction domains.The introduction of different organic templates into the Fe]P]O system leads to different structures which are very sensitive not only to the nature of the templates incorporated but also to reaction conditions. Other hydrothermally prepared iron phosphates with open-framework structures have been prepared and will be presented in future publications. Fig. 4 A layer of the [HN(CH2CH2)3NH]3[Fe8(HPO4)12(PO4)2(H2O)6] structure viewed along the c axis Fig. 5 A 14-membered ring containing a diprotonated 1,4-diazabicyclo[2.2.2]octane cation viewed along the a axis. Hydrogen-bonding interactions, O(9) ? ? ?H]N, are shown as dotted lines. Thermal ellipsoids are shown at 50% probabilityJ. Chem. Soc., Dalton Trans., 1997, Pages 2221–2225 2225 Acknowledgements We thank the Institute of Chemistry, Academia Sinica and the National Science Council (NSC86-2113-M-001-014) for support, Professor S.-L. Wang and Ms. F.-L. Liao at the National Tsing Hua University for X-ray intensity data collection, and Professor T.-Y. Dong at the National Sun Yat-Sen University for Mössbauer spectroscopy measurements. References 1 K.-H. Lii, T.-Y. Dong, C.-Y. Cheng and S.-L. Wang, J. Chem. Soc., Dalton Trans., 1993, 577; K.-H. Lii, P.-F. Shih and T.-M. Chen, Inorg. Chem., 1993, 32, 4373; E. Dvoncova and K.-H. Lii, Inorg. Chem., 1993, 32, 4368; K.-H. Lii, J. Chem. Soc., Dalton Trans., 1994, 931; K.-H. Lii and C.-Y. Huang, J. Chem. Soc., Dalton Trans., 1995, 571; Eur. J. Solid State Inorg. Chem., 1995, 32, 225; K.-H. Lii, Eur. J. Solid State Inorg. Chem., 1995, 32, 225; K.-H. Lii, J. Chem. Soc., Dalton Trans., 1996, 819. 2 M. Cavellec, D. Riou and G. Ferey, Acta Crystallogr., Sect. C, 1995, 51, 2242. 3 J. R. D. DeBord, W. M. Reiff, R. C. Haushalter and J. Zubieta, J. Solid State Chem., 1996, 125, 185. 4 R. H. Jones, J. M. Thomas, H. Qisheng, R. Xu, M. B. Hursthouse and J. Chen, J. Chem. Soc., Chem. Commun., 1991, 150. 5 M. Cavellec, D. Riou, C. Ninclaus, J.-M. Greneche and G. Ferey, Zeolites, 1996, 17, 250. 6 A. M. Chippindale, S. J. Brech, A. R. Cowley and W. M. Simpson, Chem. Mater., 1996, 8, 2259. 7 G. M. Sheldrick, SHELXTL PC, Version 5, Siemens Analytical X-Ray Instruments Inc., Madison, WI, 1995. 8 I. D. Brown and D. Altermatt, Acta Crystallogr., Sect. B, 1985, 41, 244. 9 F. Menil, J. Phys. Chem. Solids, 1985, 46, 763. Received 21st January 1997; Paper 7/00472I

ISSN:1477-9226    

DOI:10.1039/a700472i

出版商:RSC    

年代:1997

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1999, 2225–2230 2225 The reaction of the tertiary phosphine sulfides R3PS (R 5 Ph, Me2N or C6H11) with X2 (X2 = I2, Br2, IBr or ICl); structural characterisation of the CT complexes (Me2N)3PSI2 and Ph3PS(I0.89Br0.11)Br and the ionic compound [{(Me2N)3PS}2S]21 2[Br3]2 Wendy I. Cross, Stephen M. Godfrey,* Sheena L. Jackson, Charles A. McAuliVe and Robin G. Pritchard Department of Chemistry, University of Manchester Institute of Science and Technology, Manchester, UK M60 1QD.E-mail: stephen.m.godfrey@umist.ac.uk Received 26th March 1999, Accepted 18th May 1999 The reactions of the tertiary phosphine sulfides R3PS (R = Ph, Me2N or C6H11) with X2 (X = I or Br) and IX (X = Br or Cl) have been studied. Reaction of R3PS with I2 or IX results in quantitative isolation of the CT complexes R3PSIX (X = I, Br or Cl), except for Ph3PS with I2 which produces the unusual compound (Ph3PSI2)2I2, which has been crystallographically characterised by earlier workers.The crystal structure of (Me2N)3PSI2 has been determined and compared to Ph3PSI2, previously described. The greater d(I–I) for (Me2N)3PSI2, 2.856(1) Å, compared to that of 2.823(1) Å for Ph3PSI2 clearly illustrates that d(I–I) in R3PSI2 compounds is sensitive to R, although this eVect is less pronounced when compared to that of analogous R3PSeI2 compounds. The crystal structure of the product from the reaction of Ph3PS with IBr has also been determined and represents the first example of a tertiary phosphine sulfide interhalogen CT complex.It has the formula Ph3PS(I0.89Br0.11)Br and is isomorphous with Ph3PSI2. The reaction of R3PS with Br2 is complex. In the case of Ph3PS, phosphorus–sulfur bond cleavage occurs quantitatively to produce Ph3PBr2 and elemental sulfur. Reaction of (Me2N)3PS with Br2 gives, as one product, [{(Me2N)3PS}2S][Br3]2 in moderate (ca. 30%) yield. This unusual dication is compared to the previously reported [(But 3PTe)2Te][SbF6]2.Tricyclohexylphosphine sulfide reacts with Br2 in solvents of low relative permittivity (Et2O) to produce the 1 : 1 addition complex (C6H11)3PSBr2; however, dissolution of this material in solvents of higher relative permittivity results in phosphorus–sulfur bond cleavage to produce (C6H11)3PBr2 and elemental sulfur. Introduction The ability of tertiary phosphine chalcogenides to form addition compounds with dihalogens and interhalogens was first recognised by Zingaro and co-workers 1–5 in the early 1960s.In the case of tertiary phosphine selenides, reaction with IX (X = Cl, Br or I) appeared to produce the CT compounds R3PSe–X–X where, in the case of the interhalogen complexes, the heavier halogen binds directly to the selenium atom, according to IR and UV/VIS spectroscopy.1–5 Solution studies concerning trialkylphosphine sulfides indicated similar results, i.e. 1 : 1 CT complex formation upon reaction with IX (X = Cl, Br or I).However, in the case of the reaction of triphenylphosphine sulfide with diiodine, an unusual 2 : 3 (Ph3PS:I2) adduct was isolated.6 The crystal structure of this adduct, the first reported for a tertiary phosphine sulfide–dihalogen complex, revealed two Ph3PSI2 units linked into pairs by a supporting I2 molecule. The d(I–I) is significantly increased (2.85(1) Å) when compared to d(I–I) in solid diiodine (2.71 Å) indicating that electron density is being donated to the s* antibonding orbitals of the I2 by the two Ph3PSI2 moieties.The conclusion of these workers was that, despite the fact that a 1 : 1 triphenylphosphine sulfide–diiodine adduct, Ph3PSI2, could be identified in solution, it could not be isolated in the solid state. It was therefore reasoned that, due to the poor donor ability of triphenylphosphine sulfide towards diiodine, a 2 : 3 adduct formed consisting of two Ph3PSI2 moieties and a supporting diiodine molecule.The interaction of triphenylphosphine sulfide with diiodine has been independently re-examined by Sobczky and coworkers 7 and Kaur and Lobana.8,9 Both groups reported the formation of a 1 : 1 CT complex, although it was noted by Kaur and Lobana8 that Ph3PS formed a 2 : 3 (Ph3PS:I2) complex in CCl4 and a 1 : 1 complex in CH2Cl2, suggesting that the formation of the 1 : 1 complex is solvent dependent. Very recently, the interaction of triphenylphosphine sulfide with diiodine has been studied by Bricklebank and coworkers 10 using 31P-{1H} NMR and UV/VIS spectroscopy.More significantly, the crystal structure of the 1 : 1 CT complex, Ph3PSI2, prepared from dichloromethane, was reported thus confirming the existence of the 1 : 1 CT adduct in the solid state. The d(P–S) for this compound, 1.998(2) Å, is rather short suggesting some retention of double bond character. However, d(I–I) for the complex, 2.823(1) Å, clearly illustrates lengthening of the I–I bond upon adduct formation when compared to that of free diiodine, as expected.We are currently engaged in studying the interaction of a variety of tertiary phosphine chalcogenides with dihalogens and interhalogens and have found that, in the case of the reaction of R3PSe with I2 (R = Ph, Me2N or Et2N), 1 : 1 CT complexes result both in the solid state and in solution.11 All three compounds have been studied crystallographically. The structural features of Ph3PSeI2 are generally very similar to those exhibited by Ph3PSI2.One notable diVerence however is d(I–I) for the two complexes, 2.823(1) and 2.881(3) Å for Ph3PSI2 and Ph3PSeI2, respectively. This greater lengthening of the diiodine bond in the tertiary phosphine selenide complex compared to the tertiary phosphine sulfide reflects the greater donor power of selenium compared to sulfur towards diiodine. The iodine– iodine bond length in the CT complexes R3PSeI2 is also sensitive to the nature of R.For example, as previously stated, d(I–I)2226 J. Chem. Soc., Dalton Trans., 1999, 2225–2230 for Ph3PSeI2 is 2.881(3) Å whereas d(I–I) for the complex (Et2N)3PSeI2 is 2.985(2) Å. This illustrates that the R groups on the parent tertiary phosphine significantly aVect the donor power of the selenium atom towards diiodine, despite the fact that they are not directly bound to this atom. We have also investigated the reaction of dibromine with certain tertiary phosphine selenides.In the case of R3PSe (R = Me2N or C6H11), T-shaped 1 : 1 adducts result, R3PSeBr2.12 However, in the case of Ph3PSe, whilst reaction of this tertiary phosphine selenide with dibromine in diethyl ether produces the T-shaped Ph3- PSeBr2, the reaction product proved to be solvent sensitive; the same reaction performed in dichloromethane produced the unusual ionic dinuclear compound [Ph3PSeBrSePPh3]Br? 2CH2Cl2.13 A further anomaly was noted for the reaction of (Prn 2N)3PSe with dibromine, which results in the isolation of equimolar quantities of (Prn 2N)3PSe2Br2 and (Prn 2N)3PBr2. The former compound contains both bent and T-shaped geometries for the two selenium atoms, respectively.13 We now report the reaction of some tertiary phosphine sul- fides with X2 (X = I or Br) and IX (X = Br or Cl).Considering the variety and, in some cases, unexpected products obtained from the analogous reactions of the tertiary phosphine selenides, we felt the present study was certainly worthwhile.Results and discussion Analytical and spectroscopic data for the tertiary phosphine sulfides and the 1 : 1 addition compounds R3PSX2 (X2 = I2 or IBr) are displayed in Table 1. The tertiary phosphine sulfides were easily prepared from the direct reaction of the tertiary phosphine with elemental sulfur at room temperature (RT) in diethyl ether according to literature methods.4,14 They were treated with X2 (X2 = I2 or IBr) in a 1 : 1 stoichiometric ratio under anhydrous conditions according to eqn.(1). (R = Ph, R3PS 1 X2 N2, ca. 2 d Et2O, RT R3PSX2 (1) Me2N or C6H11). All of the compounds were isolated in quantitative yield and proved to be air-stable. In order to investigate the eVects of the R groups on d(I–I) for a given R3PSI2 compound (we have previously reported the sensitivity of d(I–I) on R for analogous R3PSeI2 compounds11) we decided crystallographically to characterise (Me2N)3PSI2 to compare with Ph3PSI2 which was recently reported by Bricklebank and co-workers.10 Crystals of (Me2N)3PSI2 were easily grown from diethyl ether–dichloromethane (1 : 1) solution at 50 8C on cooling to room temperature.From the large crop of dark red crystals one was chosen for analysis by single crystal X-ray diVraction. The structure of (Me2N)3PSI2 is illustrated in Fig. 1 and selected bond lengths and angles are given in Table Fig. 1 The crystal structure of (Me2N)3PSI2. 2. In common with Ph3PSI2, (Me2N)3PSI2 adopts the CT structure with an approximately linear S–I–I bond (177.98(6)8).The d(I–I), 2.856(1) Å, is significantly increased compared to that of solid diiodine, 2.71 Å, as expected since electron density is passed from the sulfur atom to the s* antibonding orbitals of the diiodine molecule. Significantly, it is also longer than that reported for Ph3PSI2 (2.823(1) Å), clearly illustrating that d(I–I) in R3PSI2 compounds is sensitive to R.However, the eVect is less pronounced compared to that in the analogous R3PSeI2 (the diVerence in d(I–I) for (Me2N)3PSI2 and Ph3PSI2 is 0.033(1) Å; the diVerence for (Me2N)3PSeI2 and Ph3PSeI2 is 0.081(2) Å). In addition to the longer d(I–I) in (Me2N)3PSI2 compared to Ph3PSI2, there is also a slight lengthening of the phosphorus– sulfur bond, 2.014(4) Å compared to 1.998(2) Å for Ph3PSI2. Although the ability of Ph3PS to form a stable 1 : 1 adduct with iodine monobromide was reported by Zingaro and Meyers,3 no compound of this stoichiometry has been crystallographically characterised.We therefore decided to prepare crystals of Ph3PSIBr with a view to comparing the resultant crystal structure with that recently reported for Ph3PSI2. Crystals of Ph3PSIBr were prepared from diethyl ether–dichloromethane solution (1 : 1) by dissolution of the orange powder in the solvent at ca. 50 8C and subsequently allowing the solution to cool slowly to RT. After ca. 3 d, large orange crystals formed in the reaction vessel, one of which was selected for analysis by single crystal X-ray diVraction. The crystal structure of the resultant complex is illustrated in Fig. 2. Selected bond lengths and angles are displayed in Table 3. In fact the compound has the formula Ph3PS(I0.89Br0.11)Br and is isomorphous with the previously reported Ph3PSI2. The d(I–Br), 2.6832(6) Å, is increased with respect to that of solid IBr (2.52 Å) as expected with the formation of a CT complex. Unlike Ph3PI1.27Br0.73, which shows dual occupancy of the halogen sites and is rich in diiodine,15 Ph3PSIBr exists as Ph3PS(I0.89Br0.11)Br, i.e.although predominantly the heavier halogen is bound to the sulfur atom, the molecule overall is rich in bromine with respect to iodine. The d(S–I) for Ph3PSIBr is 2.656(1) Å, less than that observed in Ph3PSI2, 2.753(2) Å. The d(P–S) 2.007(1) Å, is greater than that observed for Ph3PSI2, 1.998(2) Å, although clearly this diVerence is very slight.Thus Ph3PSIBr represents the first crystallographically characterised R3PSIBr compound and verifies the spectroscopic data reported by earlier workers.3 The reaction of R3PS with Br2 Unlike the reaction of R3PS with X2 (X2 = I2 or IBr), which produce the 1 : 1 addition compounds R3PSX2 in quantitative yield, the reaction of R3PS with dibromine is complex. In the case of the reaction of triphenylphosphine sulfide with dibromine two reaction products appear to be formed.Addition of the two reactants in diethyl ether produces after ca. 2 d a large quantity of a white solid. The 31P-{1H} NMR spectrum of this material, recorded in CDCl3, showed a single resonance at d 51.8, very similar to that previously recorded for Ph3PBr2, Fig. 2 The crystal structure of Ph3PS(I0.89Br0.11)Br.J. Chem. Soc., Dalton Trans., 1999, 2225–2230 2227 Table 1 The reaction of R3P with S and R3PS with X2 (R = Ph, Me2N or C6H11; X2 = I2, IBr, ICl or Br2); analytical and spectroscopic data for the products formed Analysis (%), found (calculated) 31P-{1H} Reactants Product(s) Colour Mp/8C C H N S I Br Cl NMR, d a n & (P–S)/cm21 Ph3PS 1 I2 Ph3PS 1 IBr Ph3PS 1 ICl Ph3PS 1 Br2 (Me2N)3P 1 S (Me2N)3PS 1 I2 (Me2N)3PS 1 IBr (Me2N)3PS 1 ICl (Me2N)3PS 1 Br2 (C6H11)3P 1 S (C6H11)3PS 1 I2 (C6H11)3PS 1 Br2 (Ph3PSI2)2I2 1 Ph3PS Ph3PSIBr Ph3PSICl Ph3PBr2 1 S8 (Me2N)3PS (Me2N)3PSI2 (Me2N)3PSIBr (Me2N)3PSICl [{(Me2N)3PS}2S]21 2[Br3]2 (C6H11)3PS (C6H11)3PSI2 (C6H11)3PSBr2 Red Yellow Yellow White White Red Yellow Yellow Yellow White Dark Red Yellow 140 (decomp.) 145 (decomp.) 138–139 30–31 49–50 99–100 90–91 66–68 183–184 173–174 121–122 32.0 (32.0) 43.0 (43.1) 47.1 (47.3) 51.5 (51.2) 36.1 (36.9) 16.3 (16.0) 18.3 (17.9) 20.9 (20.1) 16.4 (16.0) 70.0 (69.2) 38.1 (38.2) 44.2 (45.8) 2.5 (2.2) 3.4 (3.0) 3.0 (3.3) 3.7 (3.6) 10.0 (9.2) 4.3 (4.0) 4.8 (4.5) 5.3 (5.0) 4.2 (4.0) 10.7 (10.6) 6.1 (5.8) 7.0 (7.0) ———— 20.9 (21.5) 9.5 (9.4) 10.6 (10.4) 11.9 (11.7) 9.3 (9.3) ——— 4.7 (4.7) 6.0 (6.4) 6.7 (7.0) — 15.3 (16.4) 8.0 (7.1) 8.6 (8.0) 9.1 (9.0) 10.0 (10.6) 9.9 (10.3) 5.6 (5.7) 5.7 (6.8) 55.2 (56.4) 23.8 (25.3) 27.8 (27.8) —— 55.4 (56.6) 30.9 (31.6) 34.8 (35.5) —— 44.8 (44.9) 33.6 (33.9) — 18.9 (16.0) — 37.8 (37.9) —— 20.6 (19.9) — 50.0 (53.4) ——— —— 7.7 (7.8) ———— 9.5 (9.9) ———— 42.3, 44.0 42.0 42.7 51.8 82.4 73.8 70.1 69.6 Insoluble 62.7 58.4 103.9 b 592, 638 588 586 — 565 544 538 540 629 583 573 a Shifts recorded in CDCl3 relative to concentrated phosphoric acid as standard.b NMR resonance due to (C6H11)3PBr2, see text.2228 J. Chem. Soc., Dalton Trans., 1999, 2225–2230 d 49.2.16 Elemental analysis of this solid confirms its identity as Ph3PBr2 [Found (Calc.): C, 51.5 (51.2); H, 3.7 (3.6); Br, 37.8 (37.9)%]. Concentration of the resultant filtrate produced some pale yellow crystals, one of which was selected for analysis by single crystal X-ray diVraction.A unit cell determination of this material revealed it to be S8. Clearly, reaction of Ph3PS with dibromine results in the cleavage of the phosphorus–sulfur bond, eqn. (2). 8 Ph3PS 1 8 Br2 N2, ca. 2 d Et2O, RT 8 Ph3PBr2 1 S8 (2) This result is in direct contrast to the analogous reaction of triphenylphosphine selenide with dibromine13 which, in solvents of low relative permittivity, produces the T-shaped adduct Ph3PSeBr2, analogous to (Me2N)3PSeBr2 and (C6H11)3PSeBr2 previously described.However, reaction of Ph3PSe with Br2 in solvents of higher relative permittivity, e.g. CH2Cl2, produces the unusual dinuclear complex [Ph3PSeBrSePPh3]Br?2CH2Cl2, thus illustrating that the phosphorus–sulfur bond is more susceptible to cleavage upon reaction with dihalogens than the phosphorus–selenide bond. In order to gain further information concerning the reaction of R3PS with dibromine, we also investigated the reaction of (C6H11)3PS with dibromine in diethyl ether solution.Reaction of dibromine with this triorganophosphine sulfide appeared to proceed in a diVerent way than the analogous reaction with Ph3PS, described above. Tricyclohexylphosphine sulfide reacts with dibromine over ca. 2 d to produce a yellow solid. Elemental analysis of this material suggests the formation of a 1 : 1 adduct, (C6H11)3PSBr2, Table 1. Further evidence for this adduct formation may be inferred from its IR spectrum, clearly illustrating a band due to the phosphorus–sulfur stretch, thus confirming that, in contrast to the reaction of Ph3PS with Br2, which results in cleavage of the P–S bond to produce Ph3PBr2, reaction of (C6H11)3PS does not result in cleavage of the P–S bond.Moreover, n(P–S) for (C6H11)3PSBr2, 573 cm21, is shifted downfield compared to n(P–S) for the parent tricyclohexylphosphine sulfide, 629 cm21. Both we 11–13 and other workers 3,10 have previously noted both in the present and previous studies that this is a good indication of adduct formation since n(P–S) shifts to lower frequency upon co-ordination of a halogen atom to the sulfur donor, as expected.Final confirmation of the formation of (C6H11)3PSBr2 should be provided by its 31P-{1H} NMR spectrum, which would be expected to exhibit a single peak, shifted from that observed from the parent tertiary phosphine selenide. Unfortunately, (C6H11)3PSBr2 is insoluble in non-polar solvents. Dissolution of the material in polar solvents such as CDCl3 results in cleavage of the P–S bond to produce (C6H11)3PBr2 and, presumably, elemental sulfur, since a single resonance at d 103.9 is observed which is very close to the reported value for tricyclohexylphosphine dibromide.16 This behaviour mirrors the triphenylphosphine sulfide–dibromine system, although in this case P–S bond cleavage occurs regardless of the relative permittivity of the solvent.Table 2 Selected bond lengths (Å) and angles (8) for (Me2N)3PSI2 I(1)–I(2) I(1)–S(1) S(1)–I(1)–I(2) N(2)–P(1)–N(1) 2.856(1) 2.705(3) 177.98(6) 102.6(3) P(1)–S(1) N(2)–P(1)–S(1) 2.014(4) 118.4(4) Table 3 Selected bond lengths (Å) and angles (8) for Ph3PS(I0.89- Br0.11)Br S(1)–I(1) I(1)–Br(2) S(1)–I(1)–Br(2) 2.656(1) 2.6832(6) 175.13(2) P(1)–S(1) P(1)–S(1)–I(1) 2.007(1) 107.63(5) In a final attempt crystallographically to characterise the elusive R3PSBr2, we decided to treat (Me2N)3PS with Br2 in diethyl ether solution.This tertiary phosphine sulfide was chosen since it contains a very basic parent tertiary phosphine and we have previously reported that the analogous compound, (Me2N)3- PSe, reacts with dibromine to produce (Me2N)3PSeBr2 quantitatively. 12 In the reaction of (Me2N)3PS with Br2, after ca. 2 d a large quantity of yellow powder was produced which was isolated by standard Schlenk techniques. Recrystallisation of the product from diethyl ether solution (dichloromethane was avoided since the use of this solvent may have resulted in cleavage of the phosphorus–sulfur bond) at 50 8C produced, on standing at room temperature for ca. 5 d, a small crop of yellow-orange crystals which we assumed to be (Me2N)3PSBr2. The crystals were plunged into an inert oil under anaerobic conditions and examined under the microscope. From these, one was chosen for analysis by single crystal X-ray diVraction. Surprisingly, the material proved to be the unusual ionic compound [{(Me2N)3PS}2S]21 2[Br3]2, Fig. 3, and not the expected 1 : 1 addition compound (Me2N)3PSBr2. Selected bond lengths and angles are in Table 4. Clearly, this material cannot be considered as representative of the only bulk product from the reaction of (Me2N)3PS with Br2, but it is nevertheless isolated in significant yield (ca. 30%) and provides an interesting insight into the complex reaction of certain R3PS compounds with dibromine. One possible other product is the free phosphine (Me2N)3P, although this was not observed in the 31P-{1H} NMR spectrum of the bulk material.It is possible to speculate that during the reaction phosphorus–sulfur bond cleavage has again occurred, but only for some of the (Me2N)3PS molecules. The free elemental sulfur thus produced may then react with dibromine to produce transient dications (e.g. SBr2) which then react with other (Me2N)3PS moieties producing the dipositive cation [{(Me2N)3PS}2S]21, the charge being balanced by tribromide anions.Again, no evidence for a sulfur–bromine bond is observed. No cation of the formula [(R3PS)2S]21 has previously been crystallographically characterised; however, the analogous tellurium containing cation, [(But 3PTe)2Te]21, has been described by Kuhn et al.17 This cation may be considered as a tellurophosphorane Te21 complex or as a phosphine stabil- Fig. 3 The crystal structure of [{(Me2N)3PS}2S][Br3]2. Table 4 Selected bond lengths (Å) and angles (8) for [{(Me2N)3- PS}2S]21 2[Br3]2 Br(1)–Br(2) Br(2)–Br(3) Br(4)–Br(5) Br(5)–Br(6) Br(1)–Br(2)–Br(3) Br(6)–Br(5)–Br(4) S(2)–S(1)–P(1) 2.506(1) 2.538(1) 2.550(1) 2.521(1) 176.51(5) 177.38(5) 104.2(1) S(1)–S(2) S(1)–P(1) S(2)–S(3) S(3)–P(2) S(1)–S(2)–S(3) S(2)–S(3)–P(2) 2.032(3) 2.119(3) 2.053(3) 2.079(3) 104.7(1) 100.7(1)J. Chem.Soc., Dalton Trans., 1999, 2225–2230 2229 Table 5 Crystal data and details of refinement for Ph3PS(I0.89Br0.11)Br, R3PSI2 and [(R3PS)2S][Br3]2 (R = Me2N) Ph3PS(I0.89Br0.11)Br (Me2N)3PSI2 [{(Me2N)3PS}2S][Br3]2 Formula MT /K Crystal system Space group a/Å b/Å c/Å a/8 b/8 g/8 U/Å3 Z m/cm21 Reflections collected Observed reflections Final R1, wR2 [I > 2s(I)] (all data) C18H15BrIPS 501.14 203(2) Monoclinic P21/c 12.352(2) 9.386(1) 15.298(2) — 95.47(2) — 1765.5(4) 4 42.80 3249 3249 0.0295, 0.0724 0.0410, 0.0779 C6H18I2N3PS 449.06 203(2) Orthorhombic Cmc21 10.878(1) 9.0848(9) 14.213(2) ——— 1404.6(3) 4 47.10 688 688 0.0323, 0.0827 0.0326, 0.0830 C12H36Br6N6P2S3 902.04 203(2) Triclinic P1� 8.429(2) 9.972(2) 19.282(4) 80.66(2) 81.18(2) 74.33(2) 1529.5(6) 4 81.98 5871 5361 0.0544, 0.1176 0.1048, 0.1386 ised Te3 dication.This description is equally valid for [{(Me2N)3- PS}2S]21 described here, which could be considered as either a tertiary phosphine sulfide S21 complex or as a phosphine stabilised S3 dication. The sulfur–sulfur bond distances, 2.032(3) and 2.053(3) Å, are fairly typical for a single bond, 2.05 Å, thus indicating that little or no S–S double bond character is observed in this dication.A similar situation is observed for the tellurium analogue [(But 3PTe)2Te]21, d(Te–Te) = 2.713(1), 2.715(2) Å; d(Te–Te) for organic ditellurides = 2.70 Å. Experimental The compounds R3PS were either obtained commercially (R = Ph) (Lancaster) or easily prepared from the direct reaction of the appropriate tertiary phosphine with elemental sulfur, according to literature methods.14 Reaction time was approximately 1 d.Reaction of the tertiary phosphine sulfides with dihalogens or interhalogens was carried out under anaerobic and anhydrous conditions, although it was later noted that the complexes R3PSI2 and R3PSIBr are moisture-stable. All manipulations of the compounds were performed inside a Vacuum Atmospheres HE-493 glove-box. Diethyl ether (BDH) was dried by standing over sodium wire for ca. 1 d, refluxed over CaH2 in an inert atmosphere (N2) and distilled directly into the reaction vessel.Anhydrous CH2Cl2 was obtained commercially and used as received, as were the dihalogens (I2, Br2) and iodine monobromide (Aldrich). The R3PSX2 compounds (X2 = I2 or IBr) were synthesized in the same way, that of Ph3PSIBr being typical. Triphenylphosphine sulfide (2.00 g, 6.80 mmol) was suspended in Et2O (ca. 75 cm3) and subsequently iodine monobromide (1.41 g, 6.80 mmol) added. After ca. 2 d the resultant dark red (R3PSI2) or orange (R3PSIBr) solid was isolated using standard Schlenk techniques.The solids were then transferred to pre-dried argon- filled ampoules which were flame sealed. Elemental analyses were performed by the analytical laboratory of this department. The 31P-{1H} NMR spectra were recorded as CDCl3 solutions on a Brüker AC200 high resolution multiprobe spectrometer relative to concentrated phosphoric acid as standard, IR spectra on a Nicolet 5PC FT spectrometer. Crystallography Crystals of all three compounds were independently submerged in an inert oil under anaerobic conditions and suitable ones were chosen by examination under a microscope.The crystals, with their protective coating of oil, were then independently mounted on glass fibres and transferred to the diVractometer and cooled to 203(2) K in the cold gas stream derived from liquid nitrogen. All measurements were performed on a Nonius MAC 3 CAD 4 diVractometer employing graphite-monochromated Mo-Ka radiation (l = 0.71069Aring;) and w–2q scans.The structures were solved by direct methods. Unit-cell dimensions were derived from the setting angles of 25 accurately centred reflections. Lorentz-polarisation corrections were applied. Details of the X-ray measurements and subsequent structure determinations are presented in Table 5. During refinement of Ph3PSIBr it was noticed that the iodine vibrational ellipsoid was significantly larger than those of the surrounding atoms.The iodine site was therefore refined as a mixture of I and Br atoms, which converged to give a crystal composition of 89% Ph3PSIBr and 11% Ph3PSBrBr. A similar treatment of the terminal Br indicated that it should remain a purely Br site. Hydrogen atoms were confined to chemically reasonable positions. Neutral atom scattering factors were taken from ref. 18, anomalous dispersion eVects from ref. 19. The structure determinations were performed using SHELXS 86 and refinement based on F2 by using SHELXL 93 crystallographic software packages.20,21 CCDC reference number 186/1468.See http://www.rsc.org/suppdata/dt/1999/2225/ for crystallographic files in .cif format. Acknowledgements We are grateful to the EPSRC for a research studentship to S. L. J. References 1 R. A. Zingaro and R. M. Hedges, J. Phys. Chem., 1961, 65, 1132. 2 R. A. Zingaro, Inorg. Chem., 1963, 2, 192. 3 R. A. Zingaro and E. A. Meyers, Inorg. Chem., 1962, 1, 771. 4 W. Tefteller and R. A. Zingaro, Inorg. Chem., 1966, 5 2151. 5 R. A. Zingaro, R. E. McGlothlin and E. A. Meyers, J. Phys. Chem., 1962, 66, 2579. 6 W. W. Schweikert and E. A. Meyers, J. Phys. Chem., 1968, 72, 1561. 7 F. Lux, R. Paetzold, J. Danel and L. Sobczky, J. Chem. Soc., Faraday Trans. 2, 1975, 1610. 8 S. Kaur and T. S. Lobana, J. Inorg. Nucl. Chem., 1981, 43, 2439. 9 S. Kaur and T. S. Lobana, J. Indian Chem. Soc., 1983, 60, 126. 10 D. C. Apperley, N. Bricklebank, S. L. Burns, D. E. Hibbs, M. B. Hursthouse and K. M. Abdul Malik, J. Chem. Soc., Dalton Trans., 1998, 1289. 11 S. M. Godfrey, S. L. Jackson, C. A. McAuliVe and R. G. Pritchard, J. Chem. Soc., Dalton Trans., 1997, 4489. 12 S. M. Godfrey, S. L. Jackson, C. A. McAuliVe and R. G. Pritchard, J. Chem. Soc., Dalton Trans., 1998, 4201. 13 S. M. Godfrey, S. L. Jackson, C. A. McAuliVe and R. G. Pritchard, unpublished work.2230 J. Chem. Soc., Dalton Trans., 1999, 2225–2230 14 D. W. Allen and B. F. Taylor, J. Chem. Soc., Dalton Trans., 1982, 51. 15 N. Bricklebank, S. M. Godfrey, C. A. McAuliVe and R. G. Pritchard, J. Chem. Soc., Dalton Trans., 1993, 2261. 16 S. M. Godfrey, C. A. McAuliVe, I. Mushtaq, R. G. Pritchard and J. M. SheYeld, J. Chem. Soc., Dalton Trans., 1998, 3815. 17 N. Kuhn, H. Schumann and R. Boese, J. Chem. Soc., Chem. Commun., 1987, 1257. 18 International Tables for X-Ray Crystallography, Kynoch Press, Birmingham, 1974, vol. 4, Table 2.2A. 19 International Tables for X-Ray Crystallography, Kynoch Press, Birmingham, 1974, vol. 4, Table 2.3.1. 20 G. M. Sheldrick, SHELXS 86, in Crystallographic Computing 3, ed. G. M. Sheldrick, Oxford University Press, 1985, p. 175. 21 G. M. Sheldrick, SHELXL 93, University of Göttingen, 1993. Paper 9/02433F

ISSN:1477-9226    

DOI:10.1039/a902433f

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

DALTON J. Chem. Soc., Dalton Trans., 1997, Pages 2227–2240 2227 X-Ray–neutron diVraction study of the electron-density distribution in trans-tetraaminedinitronickel(II) at 9 K: transition-metal bonding and topological analysis† Bo B. Iversen,*,a Finn K. Larsen,a Brian N. Figgis b and Philip A. Reynolds c a Department of Chemistry, Aarhus University, DK-8000 Aarhus C, Denmark b Department of Chemistry, University of Western Australia, Nedlands, WA 6907, Australia c Research School of Chemistry, Australian National University, Canberra, ACT 0200, Australia A quantitative description of transition-metal bonding has been obtained through combined analysis of 9(1) K X-ray and 13(1) K time-of-flight neutron diffraction data.It is shown that a simple valence-orbital model is too crude an approximation adequately to describe the electron-density distribution of Ni(ND3)4(NO2)2. To exhaust more fully the information present in the very-low-temperature diffraction data, a more flexible electron-density model was used.Quantitative measures describing the bonding in the complex have been achieved through topological analysis of the derived static model density. To study the effects of co-ordination and intermolecular interactions, comparisons were made with good-quality wavefunctions calculated for free nitrite and ammonium ions. Both ligands appear co-ordinated through predominantly electrostatic interactions. Contrary to previous studies of Ni(ND3)4(NO2)2, the topological analysis revealed that the metal–ligand interactions, besides cylindrical s contributions, also have non-cylindrical p contributions to the covalent part of the bonding.Plots of the Laplacian of the electron density were used to locate regions of charge concentration and charge depletion in the valence regions of the atoms in the molecule. For all atoms, maxima in the valence-shell charge concentration are found in accord with the simple Lewis electron-pair concept of bonded and non-bonded charge concentrations.The study demonstrates that X-ray diffraction data measured carefully at very low temperatures have sufficient precision to allow for a reliable and detailed topological analysis of transition-metal electron-density distributions. Quantitative descriptions of metal–ligand bonding in transition-metal complexes can be obtained through analysis of molecular electron-density distributions (EDDs).1 These can be determined from experiment by accurate X-ray diffraction measurements preferably in combination with neutron diffraction measurements on the same compound (X–N method).Previously we have carried out a combined X-ray and neutron diffraction study of diammonium hexaaquacopper(II) disulfate, 2 (ND4)2Cu(OD2)6(SO4)2 (ammonium copper Tutton salt, 1), similar to the present one on Ni(ND3)4(NO2)2 2. In that study considerable improvements were obtained by collecting data at 9 K relative to a study done with liquid-nitrogen cooling 3 (ª85 K).At 9 K usual systematic errors such as thermal diffuse scattering (TDS) and anharmonic motion are suppressed to a point where they are negligible.4 For the lowtemperature data for 1 extra radial flexibility, besides the k parameter normally used in experimental charge-density studies,5 was necessary in order to obtain a satisfactory model. The very flexible model used in that study resulted in the coincidence of both the positional and thermal parameters obtained separately from the X-ray and neutron diffraction data.At higher temperatures discrepancies between X-ray and neutron parameters are observed due to temperaturedependent differences in systematic errors between the two experiments, and such discrepancies have often compromised the accuracy of X–N studies. In the present study of compound 2 we have carried out 9(1) K X-ray diffraction and 13(1) K time-of-flight neutron diffraction measurements.In a preceding paper 6 details about the data acquisition, data reduction and the refined structural parameters were described.‡ It was shown that for 2 excellent agreement between positional and thermal parameters for the two experiments can be obtained. The root mean square (r.m.s.) † Non-SI unit employed: au ª 4.36 × 10218 J. difference between the U values (·(DUij)2Ò� �� ) was as low as 0.000 50 Å2 and ·(DUij/sij)2Ò� �� = 1.92. Owing to the very low temperatures of the experiments the absolute differences between the thermal parameters are much smaller than what is normally obtained in studies using nitrogen cooling.For complex systems the faster but more complex time-of-flight neutron diffraction technique can produce structural parameters of a quality comparable to experiments with monochromatic neutrons. The excellent correspondence between the parameters derived separately from the X-ray and neutron data demonstrates that systematic errors in the data are small and this gives confidence in the deconvolution of the thermal motion from the X-ray data.Recently it has been shown that, for systems as large as transition-metal complexes, neither current theoretical methods nor empirical modelling techniques are able to produce EDDs that fully do justice to the quality of the experimental data which can be produced by careful diffraction measurements at very low temperatures (ª10 K).7 For the theoretical results this may be due to inadequate treatment of both e2–e2 correlation and intermolecular effects, such as charge transfer and polarisation.In the case of the empirical EDD models it is clear that, although very flexible, they are still too rigid to fit all subtle density features. Much knowledge can therefore be gained ‡ Time-of-flight neutron diffraction data: 13(1) K, 40 threedimensional data histograms, 2436 reflections for least squares, wavelength range 0.7 < l < 4.2 Å, (sin q/l)max = 1.17 Å21, a = 10.580(2), b = 6.720(1), c = 5.863(1) Å, b = 114.82(1)8, space group C2/m, Z = 2, ma(true absorption at l = 1.8 Å) = 0.549 cm21, mS(total scattering) = 1.202 cm21, 78.5% deuteriation (refined), box-shaped crystals of side length 2.5 mm.X-Ray diffraction data: 9(1) K, Ag-Ka radiation (l = 0.5603 Å), 9231 reflections measured, 4016 unique for least squares, RI = 0.023, (sin q/l)max = 1.40 Å21, a = 10.647(2) Å, b = 6.799(1) Å, c = 5.891(1) Å, b = 114.82(1)8, m1(Ag-Ka) = 13.11 cm21, crystal boundaries ±(001) 0.13, ±(2021) 0.09, (12121) 0.12, (22241) 0.14, (010) 0.32 mm.2228 J.Chem. Soc., Dalton Trans., 1997, Pages 2227–2240 through detailed comparison of advanced empirical EDD models with results of high quality ab-initio calculations. We are presently carrying out extensive ab-initio calculations on 2 and in a forthcoming paper a full account of this aspect will be given. The present compound was chosen further to probe the nature of the deficiencies in both theory and experimental models on a chemically more interesting system having ligands of quite different positions in the electrochemical series.It is found (Fig. 1) as a single molecular entity in the crystal and so it is expected that intermolecular effects on the molecular EDD are reduced and comparisons between theory and experiment can be enhanced relative to more ionic systems. Furthermore compound 2 has been examined by a wealth of different methods including X-ray diffraction,9 magnetochemical, 9b spectroscopic,9b neutron diffraction,10 polarised neutron diffraction experiments 11 and by ab-initio theoretical calculations. 12 This allows for a comparison with other experimental and theoretical results. The geometry and structural characteristics of the complex were discussed by Figgis et al.9b based on X-ray diffraction data collected at 130 K, and based on a limited set of neutron diffraction data collected at 4.2 K.10 Overall the present study agrees well with the previous results.The most noteworthy difference is the more precise N]D bond lengths found in the present study. In the previous neutron diffraction study N]D 0.996(8), 1.045(12) and 1.005(11) Å comparesent values of 1.015(1), 1.015(1) and 1.014(1) Å. The complex has a small but significant orthorhombic distortion [N(2)]Ni]N(2) 87.46(2)8]. Furthermore it has consistently been observed that the two N]O bonds are slightly different in length [N(1)]O(1) 1.244(1) and N(1)]O(2) 1.255(1) Å].In the crystal structure there are hydrogen bonds between the ammonia deuterium atoms and the nitrite oxygen atoms. One oxygen atom [O(1)] has two hydrogen bonds [D(3) ? ? ? O(1) 2.159(1) Å] and the other [O(2)] has four hydrogen bonds [D(1) ? ? ? O(2) 2.175(1) and D(2) ? ? ? O(2) 2.126(1) Å]. Figgis et al.9c further studied compound 2 by X-ray diffraction methods at 110 K, and found a valence-orbital model to be an adequate description of the improved data.The new 9 K data reveal that this is not really true. Comparison between the 110 and 9 K models directly shows the increased resolution gained by lowering the temperature of the experiment. At the Fig. 1 An ORTEP8 drawing of Ni(ND3)4(NO2)2 showing 90% ellipsoids and the atom-numbering scheme. Also shown are the bond critical points found in topological analysis of the static model density.Note the large zero-point thermal motion of the deuterium atoms relative to the heavier atoms same time we also now use X-ray diffraction in combination with neutron diffraction results. With the high quality of the low-temperature data and the inclusion of many high-order data, the valence-orbital model is now too inflexible. Chandler et al.12 compared the 110 K experimental charge-density to abinitio calculations and found broad agreement, although some notable differences were observed.The 110 K X-ray experiment yielded a very low nickel 3dxy population and high 4p populations which were not reproduced by theory. It was speculated that this was due to mixing of a doubly excited state into the ground state through configuration interaction (CI), and to resolve this point a rigorous CI calculation seemed to be needed. The higher precision of the present data makes it possible partly to resolve this question. However, because it is necessary to introduce a very flexible electronic model properly to describe the data, it is no longer really meaningful to speak of pseudo-atom fragments and their associated orbital populations. We have instead attempted to use topological analysis on the static model total density derived from the X-ray data inspired by Bader’s theory of atoms in molecules.13 The theory of atoms in molecules has mostly been applied to fairly simple organic molecules which can be well described by theoretical methods. In recent years it has been shown that valuable topological information can also be obtained from experimentally derived EDDs of small organic molecules and simple metals.14 However, because of the much greater complexity of transitionmetal systems, little is known about the topological features of this important class of compounds.Numerous bonding models are in general use to describe the diversity found in transitionmetal bonding. It would therefore be interesting to obtain a new view of transition-metal bonding through use of topological methods.This cannot be achieved from a study of a single complex, but if precise EDDs can be derived for a number of complexes, trends and characteristics of the topology of various bonding types may be found. In the present study we probe what kind of topological information can be reliably retrieved for transition-metal complexes from conventional diffraction data measured at very low temperatures (ª10 K).Multipole Modelling In this study we used the program XD.15 It employs the Hansen and Coppens16 multipole formalism in the description of the charge density. The atomic density contributions are parameterised into a core term, rcore, a spherical valence term, rvalence, and a set of multipolar functions as shown in equation (1). In the ratom(r) = Pcorercore 1 Pvalencek3rvalence(kr) 1 lmax l = 0 o kl93Rl(kl9r) l m = 0 o Plm±dlm±(q, f) (1) standard model two k expansion/contraction parameters are included for each atom: one modifies the radius of the spherical valence term and the second modifies the radius of the multipolar terms.The XD program also allows extra radial flexibility to be introduced in the modelling by use of separate k parameters for each multipole order. With the results of our study 2 of compound 1 in mind it was clear that we had to construct a very flexible electronic model properly to describe the present data.A number of models were refined but we only present detailed results for what we believe to be the best multipole model. This model contained full multipole expansions up to fourth order on Ni, up to third order on N(1), N(2), O(1) and O(2), and to second order on D(1), D(2) and D(3). On Ni a very flexible model containing two sets of multipoles was used. For N(1), the nitrite nitrogen atom, separate k parameters were refined for all multipole orders. The other atoms only had two k parameters refined.The two k parameters were constrained to be the same on all threeJ. Chem. Soc., Dalton Trans., 1997, Pages 2227–2240 2229 Table 1 Electron-density parameters obtained from a refinement with the XD program: M refers to monopoles, D to dipoles, Q to quadrupoles, O to octapoles and H to hexadecapoles; for Ni two sets of multipoles (mul), with radial dependencies derived from 3d and 4s orbital products respectively, were used Ni: M13d 2.21(3), Q03d 20.33(7), Q213d 20.03(4), Q223d 20.05(4), H03d 0.116(6), H213d 0.151(5), H223d 20.179(5), H413d 0.012(4), H423d 20.124(4), M14s 0.75(24), Q04s 0.25(7), Q214s 0.02(4), Q224s 0.06(4), H04s 0.127(6), H214s 20.09(5), H224s 0.17(5), H414s 20.04(4), H424s 0.122(4), k3d 1.09(1), k4s 1.00 O(1): M1 3.10(4), D11 20.046(7), D12 0.002(7), Q0 20.054(9), Q21 20.062(8), Q22 20.020(7), O11 0.008(7), O12 20.009(8), O31 0.004(6), O32 0.002(6), k(M1) 1.028(7), k(mul) 1.38(12) O(2): M1 3.14(4), D11 20.043(8), D12 20.019(8), Q0 20.041(10), Q21 20.034(8), Q22 20.000(7), O11 20.004(8), O12 20.006(8), O31 0.011(6), O32 0.001(6), k(M1) 1.109(7), k(mul) 1.27(13) N(1): M1 2.43(5), D11 0.056(9), D12 20.011(7), Q0 20.11(2), Q21 0.01(1), Q22 0.01(1), O11 0.06(2), O12 0.01(2), O31 0.17(3), O32 0.01(2), k(M1) 1.05(1), k(D) 1.36(15), kQ 0.98(10), k(O) 0.75(5) N(2): M1 4.13(18), D0 0.21(5), D11 0.04(4), D12 0.06(4), Q0, 0.11(3), Q11 20.03(3), Q12 0.06(3), Q21 20.07(3), Q22 0.00(3), O0, 0.17(4), O11 0.05(3), O12 20.05(3), O21 20.02(3), O22 20.08(4), O31 0.17(4), O32 20.02(4), k(M1) 1.09(2), k(mul) 0.75(5) D(1): M1 0.99(9), D0 0.36(5), D11 20.02(4), D12 0.07(5), Q0 0.23(7), Q11 0.03(5), Q12 0.07(5), Q21 0.02(5), Q22 0.00(5), k(M1) 0.94(3), k(mul) 1.00(5) D(2): M1 1.10(10), D0 0.40(6), D11 20.02(4), D12 0.11(4), Q0 0.27(7), Q11 0.07(6), Q12 0.07(6), Q21 20.00(5), Q22 20.01(4), k(M1) 0.94(3), k(mul) 1.00(5) D(3): M1 1.13(10), D0 0.42(6), D11 20.04(4), D12 0.02(4), Q0 0.20(7), Q11 0.06(5), Q12 0.01(5), Q21 0.00(4), Q22 0.07(5), k(M1) 0.94(3), k(mul) 1.00(5) Table 2 Refinement residuals for the XD multipole model No 4006 Nv 140 RF 0.0234 RwF 0.0273 RF2 0.0303 RwF2 0.0539 Goodness of fit 1.257 deuteriums.The model contained an isotropic extinction parameter (type I, Becker and Coppens,17 Lorentzian distribution) and two scale factors, the last because the data set consists of two blocks collected slightly separated in time (see ref. 6). In all refinements full-matrix least-squares minimisation was used. The maximum shift/e.s.d. was less than 0.005 and the maximum amount of extinction is for the 001 reflection (y = 0.67) with only 11 other reflections having more than 10% extinction. The refinement was constrained to maintain overall electroneutrality of the unit cell. No other constraints were introduced besides those mentioned above and those imposed by the choices of the local coordinate systems.§ For all non-deuterium atoms both the core and the valence scattering function were calculated from Hartree–Fock atomic wavefunctions.18 For the nickel site, functions for Ni0 were used with the radial form of the valence functions derived from 3d atomic orbital products for the contracted set of functions and from 4s orbital products for the diffuse set of functions.For deuterium the scattering factors calculated by Stewart et al.19 were used. A number of other models were tested before deciding on the above model.For instance we tried to refine separate k parameters for all multipole orders on all atoms. While such refinements achieve convergence they do not improve the residuals nor lead to differences in the topological features discussed below. The extra k parameters were therefore deemed unimportant. Only for N(1) the use of separate k parameters for each multipole order was found to be important for the topology of the static model density.Fourth-order multipoles were tested on the N and O atoms, but were found to refine to insignificant values and were therefore omitted. Introduction of anharmonic thermal parameters was also tried on all atoms, but none of them refined to significant values. The study therefore shows that anharmonic motion appears to be negligible at temperatures close to absolute zero. § The following local coordinate systems were adapted: Ni [Z axis Ni æÆ (0, 1, 0), X axis Ni æÆ N(1)], O(1) [Z axis O(1) æÆ (0.3082, 1, 0.1924), X axis O(1) æÆ N(1)], O(2) [Z axis O(2) æÆ (0.2135, 1, 20.2096), X axis O(2) æÆ N(1)], N(1) [Z axis N(1) æÆ (0.2005, 1, 20.0063), X axis N(1) æÆ Ni], N(2) [Z axis N(2) æÆ Ni, X axis N(2) æÆ (20.0164, 0.3098, 0.2573)], D(1) [Z axis D(1) æÆ N(2), X axis D(1) æÆ (1, 0, 1)], D(2) [Z axis D(2) æÆ N(2), X axis D(2) æÆ (1, 0, 1)], D(3) [Z axis D(3) æÆ N(2), X axis D(3) æÆ (1, 0, 1)].In Fig. 2 residual maps are shown for two sections through the molecule.In general the residual features are less than 0.2 e Å23. The small peak relatively near the nickel atom in the Ni]N(1)]O(2) plane could not be modelled even with use of the extra parameters described above. It may be due to a small systematic error in the data. In Table 1 the refined electrondensity parameters are listed, and in Table 2 refinement residuals for the XD multipole model are shown. The positional and thermal parameters were given in ref. 6. The monopole populations obtained from more restricted multipole refinements have in many previous studies been used to assign pseudo-atom charges.20 However, when using a diffuse and necessarily flexible model such assignments are not meaningful.In the present model it is especially the Ni- and the ligating nitrogen-centred functions which are very diffuse. This leads to unrealistic charges. Indeed the pseudo-atom monopole populations suggest a negative nickel atom. At the end of the paper we will briefly discuss simplified and less diffuse models which qualitatively give ‘better’ pseudo-atom charge assignments.However such models are no longer adequate for detailed description of the data. Therefore quantitative discussions of the atoms in the molecule are in this paper carried out by analysis of the total electron density using rigorous topological methods. Theoretical Calculations For comparison with the experimental ammonia and nitrite critical point analysis and deformation density we calculated good-quality wavefunctions for the free ammonia molecule and for the nitrite ion.Since our objective is to examine any changes due to bonding in molecule 2, we require calculations which are chemically indistinguishable from exact results. Accordingly we followed the prescription of Schaefer et al.21 Using their triplezeta plus two polarisation function Gaussian basis set (TZ2P) we optimised the nuclear geometry in the Hartree–Fock framework using the ab-initio package GAMESS22 implemented on a DEC alpha workstation with 8 Gbyte disk storage.The resulting energies and geometries are: NH3, E = 256.220 26 au, r(N]H) = 0.9980 Å, q(H]N]H) = 107.88; NO2 2, E = 2204.154 68 au, r(N]O) = 1.2208 Å, q(O]N]O) = 116.98. At this geometry we introduced configuration interaction, using all single and double excitations (29 161 configurations in Cl sym-2230 J. Chem. Soc., Dalton Trans., 1997, Pages 2227–2240 metry for NH3 and 314 821 in Cl symmetry for NO2 2) giving final energies of 256.455 73 and 2204.751 38 au.This calculation duplicates that for NH3 of Schaefer et al.21 so that appropriate properties of the wavefunction can be derived. That for NO2 2 is an improvement on earlier studies 23 which used at best only double zeta plus single polarisation, included no electron correlation and reached a lowest energy of 2204.1318 au. We also carried out the same calculations at the experimental geometry.For NH3 the results are almost unchanged and only the optimised geometry results are discussed. Deformation densities The improvement in the 9 K data over the 110 K data is immediately visible when mapping the difference between observed and spherical atom model densities, Fig. 3. Compared to the 110 K data of Figgis et al.9c the 9 K data reach much higher contour levels, mainly because of less thermal motion. In Fig. 4 dynamic model deformation maps are shown.Holes in the 3d-like distribution along the six nickel–ligand bonds are observed as predicted by the crystal-field model for a Ni21 ion in tetragonal symmetry. The holes directed towards the Fig. 2 Residual density in the Ni]N(1)]O(2) plane (a) and in the Ni]N(1)]N(2) plane (b). The resolution of the maps is 0.8 Å21 and the contour interval is 0.1 e Å23. Solid lines represent positive contours and broken lines negative contours ammonia groups are deeper than the holes towards the nitrite groups.Lone pairs as well as bonding densities are clearly visible for both the nitrogen and the oxygen atoms. The differences in the crystal environment between the two oxygen atoms are reflected in the model deformation maps. The lone-pair density of O(1), which only accepts two hydrogen bonds, reach a peak height of more than 1.95 e Å23. The lone-pair density of O(2), which is involved in four hydrogen bonds, only reaches 1.50 e Å23. For O(1) a small hole near the nucleus is observed, which for O(2) only is a dip in the positive density.The qualitative features present in the dynamic model deformation maps are virtually unchanged in the static model deformation maps shown in Fig. 5. If we assume the deconvolution of the thermal motion to be adequate then these maps are directly comparable to theoretical deformation maps. In Fig. 6 deformation maps based on the optimised geometry calculations are shown. The deformation density in the N]H bond resembles that for an uncorrelated TZ1P NH4 1 calculation,2 while that for NO2 2 resembles those calculated by Cruickshank Fig. 3 Experimental deformation densities in the Ni]N(1)]O(2) plane: (a) copied from Figgis et al.,9c (b) present study. Contours and resolution as for Fig. 2J. Chem. Soc., Dalton Trans., 1997, Pages 2227–2240 2231 and Eisenstein 23b in DZ1P. They are, however, experimentally distinguishable. For example examination of our maps and those of ref. 23(b) shows that our calculation has produced a significant diminution in the hole along the N]O bond from greater than 21.0 to 20.4 e Å23. For compound 1 we have previously shown that triple-zeta calculations do indeed fit X-ray data better than do double-zeta ones.7 In the crystal the two N]O bonds are longer than the value obtained by theory. In Fig. 7 the deformation density is shown for a calculation using the experimental geometry. The changes include deeper holes near the oxygens as well as a lowering of the N]O bond peaks.There is good qualitative agreement between the theoretical and the experimental maps, but notable differences are also present. The differences are presumably due to the effect of the crystal environment on the experimental density. The nitrite ion has been examined in considerable detail both by theoretical methods,23 but not to a level currently easily obtainable, and by diffraction studies of NaNO2 24 and K2- NaCo(NO2)6.25 Deformation maps for the plane containing the N]O bond but perpendicular to the NO2 plane have been suggested to reveal p-bonding features.24 In Fig. 8 deformation Fig. 4 Dynamic model deformation densities in the Ni]N(1)]O(2) plane (a) and the Ni]N(1)]N(2) plane (b). The resolution of the maps is 1.4 Å21 and the contour interval is 0.15 e Å23 in order more clearly to illustrate the height of the oxygen lone-pair peaks. Solid lines represent positive contours and broken lines negative contours densities along the N]O bonds in this plane are shown for both the experimentally derived density and for the theoretical calculation employing the optimised geometry.The maps closely resemble the maps obtained for NaNO2 and K2NaCo(NO2)6. Fig. 5 Static model deformation density in the Ni]N(1)]O(2) plane (a), the Ni]N(1)]N(2) plane (b), and the N(2)]H(2)]H(3) plane (c). For plots (a) and (b) contours are as for Fig. 4. For (c) the contour interval is 0.1 e Å232232 J.Chem. Soc., Dalton Trans., 1997, Pages 2227–2240 The differences between the two oxygen atoms due to intermolecular interactions are clearly seen. Since deformation maps are model dependent, conclusions from them about p contributions are ambiguous. Instead, we examine these features below using topological analysis of the total electron density. The previous theoretical studies of the nitrite ion were particularly concerned with the basis-set dependence of the properties.They showed that polarisation functions are vital in theoretical calculations in order properly to describe bonding effects even for first-row structures. This is confirmed by the present experimental study where we are forced to use thirdorder multipoles on the O and N atoms in the least-squares fit. Topology The topological analysis of electron densities developed by Bader13 has been extensively presented in the literature. Critical points (CPs) in the density have —Ò = 0 and they can in Bader’s scheme be classified in terms of the properties of the eigenvalues of the Hessian matrix (second derivatives) at the critical point.The CPs are characterised by (rank, signature) where Fig. 6 Static deformation density based on TZDP/CISD calculations at optimised geometry in the O]N]O plane (a) and a H]N]H plane (b).The contour interval is 0.15 e Å23 in (a) and 0.1 e Å23 in (b). Solid lines represent positive contours and broken lines negative contours rank is the number of non-zero eigenvalues and signature is the algebraic sum of their signs.Atomic entities in the density peak at (3,23) points and chemical bonds go through (3,21) points. The (3,1) and (3,3) points define the two other structural concepts, namely rings and cages. Different chemical bonds can be characterised depending on the sign of the Laplacian (—2r) at the bond CP. The Laplacian shows where charge is locally concentrated (—2r < 0) or depleted (—2r > 0).Closed-shell interactions have positive Laplacians (mainly ionic or electrostatic interactions such as hydrogen bonds and van der Waal bonds), whereas electron-sharing interactions have negative Laplacians (covalent bonds). The p character as well as the stability of a bond can be described by the ellipticity of the bond, e = l1/ l2 2 1, where l1 and l2 are the negative eigenvalues of the Hessian matrix at the bond CP. In Table 3 the bond CPs found in the static model density by the XD program are listed together with values obtained from the theoretical densities (optimised geometry) of NO2 2 and NH3.Also listed are the eigenvalues of the Hessian matrix, and their sum (—2rc) calculated at the CPs, the lengths of the bond paths and the internuclear distances. The critical points were located as points in the density where the gradient attains a value smaller than 1 × 1025 e Å24. The static model density was calculated by including all density functions centred less than 5 Å from the point of consideration. It may be noted that the XD program at present only allows calculation of leastsquares error estimates on —2rc and rc but not on the individual eigenvalues of the Hessian matrix.Such error estimates rely, however, on the adequacy of the model in describing the density. It has recently been shown that even the very flexible models used in experimental EDD analysis do not fully exhaust the information in the data,7 and that the radial form of the fitting functions is important for the fine details of the model density.It is probably therefore a fairer estimate of the uncertainty in the topological features derived from experimental data to state the values obtained with different models having very similar values of c2. In the discussion below we state the values obtained with different models whenever there is a notable difference. The nitrite ion Results of the topological analysis of the nitrite group are shown in Fig. 9 and in Table 3.The first two bond CPs listed in Fig. 7 Static deformation density based on TZDP/CISD calculations at the experimental geometry in the O]N]O plane. Contours as for Fig. 6J. Chem. Soc., Dalton Trans., 1997, Pages 2227–2240 2233 Table 3 correspond to the N]O bonds of the nitrite group. It is striking that both rc, —2rc and the individual values of the Hessian matrix are so similar for the two bonds, when considering the differences found in the experimental deformation Fig. 8 Static model deformation density along the N(1)]O(1) bond (a) and the N(1)]O(2) bond (b) in the plane perpendicular to the molecular plane. In (c) the same plane is shown for the TZDP/CISD calculations employing the optimised geometry. The contour level is 0.15 e Å23. Solid lines represent positive contours and broken lines negative contours density. The fact that the two N]O bonds appear topologically equivalent suggests that the effect of the crystal environment is confined to the lone-pair regions and to near the oxygen nuclei.In the N]O bonds —2rc is large and negative as expected for a covalent-type interaction. If we do not use extra radial flexibility in the modelling of N(1) the values of rc drop to 3.18 and 3.26 e Å23 for the N(1)]O(1) and N(1)]O(2) bonds respectively; —2rc grows to 23.7 and 23.4 e Å25. Since we expect a large degree of covalency in the nitrite bonds we deemed the extra radial flexibility important, even though it only gives a very slight improvement in the refinement residual.The large numerical increase in —2rc when introducing extra radial flexibility is both due to a smaller positive curvature and larger negative curvatures at the bond critical points. In general the positive curvature is especially sensitive to the radial form of the fitting functions. Recently Bianchi et al.26 studied lithium bis(tetramethylammonium) hexanitrocobaltate(III) using both theoretical methods and multipole modelling of 100 K X-ray diffraction data.In this study rc in the N]O bonds are 3.19(4) and 3.44(4) e Å23 for the experimental density, and 3.28 and 3.39 e Å23 for the theoretical calculations employing the Hartree–Fock periodic approach.27 For —2rc Bianchi et al.26 obtained 27.5 and 211.9 e Å25. When comparing the experimental Laplacians to theory it is clear that theory gives a smaller value of the positive curvature than does the multipole analysis.Since the effect of employing separate k values on N(1) was to lower the value of the positive curvature, comparison between theory and experiment indicates that even the radially flexible model used in this study may still be too rigid. A simple Lewis electron-pair model predicts for a NO2 2 ion that the N]O bond is intermediate between a single and a double bond. In other words, p contributions to the bonding are expected. The p contributions are reflected in the non-cylindrical shape of the electron density at the bond critical points.The ellipticity at the bond CP measures this shape and we obtain values of 0.15 and 0.12 for the two N]O bonds respectively. The ellipticity of the bonds is also affected by the extra k parameters on N(1) [0.06 and 0.05 without extra k parameters on N(1)]. For comparison the ellipticity in carbon–carbon bonds is 0.23 for benzene and 0.45 for ethene.13 The two bond paths between the N and O nuclei fall almost exactly on the internuclear axis as seen in the identical values of Rij and Dij.Figgis et al.9b noted that the difference in the number of hydrogen bonds to the two oxygen atoms presumably causes the observed difference in the N]O internuclear distance [1.244(1) and 1.255(1) Å]. The bond critical point is almost exactly at the same distance from the nitrogen atom in the two N]O bonds. This means that the difference mainly resides inside the atomic basins of the oxygen atoms.The ammonia molecule Results of the topological analysis of the ammonia molecule are shown in Figs. 10 and 11 and in Table 3. For the ammonia molecule the densities at the bond CPs are smaller than for the nitrite ion. They are in good agreement with values found in other studies of N]H bonds.13,14a,e28 The three N]D bonds have similar values of rc. This need not be so since each deuterium atom is also involved in hydrogen bonding in which charge migration takes place.For all three bonds very large and negative values of the Laplacian are obtained indicating a large degree of covalency in the bonds. If we introduce additional flexibility through use of separate k parameters on each multipole order on N(2) the values of —2rc are 288.8, 256.5 and 251.2 e Å25 for the three N]D bonds respectively. The values of —2rc are somewhat larger than those reported in the literature.13,14a,c When examining the values of the Hessian matrix it is clear that it is the positive eigenvalue, especially, which is smaller in the present study.The positive eigenvalue reflects movement of charge away from the bond CP and into the atomic basins. As explained above it is intimately related to2234 J. Chem. Soc., Dalton Trans., 1997, Pages 2227–2240 Table 3 Bond critical points in the static XD model density: the first entry is the experimental density, the second the theoretical density; rc (e Å23) is the electron density, —2rc (e Å25) the Laplacian at the CP, l1, l2, l3 (e Å25) are the eigenvalues of the Hessian matrix, e is the ellipticity, d1 and d2 are the distances to the bond path attractors, Rij (Å) is the sum of d1 and d2 and Dij (Å) is the internuclear distance as obtained from the neutron diffraction data N(1)]O(1) N(1)]O(2) N(2)]H(1) N(2)]H(2) N(2)]H(3) Ni]N(1) Ni]N(2) N(2)]D(1) ? ? ? O(2) N(2)]D(2) ? ? ? O(2) N(2)]D(3) ? ? ? O(1) rc 3.57(7) 3.433 3.61(7) 3.433 2.10(1) 2.403 2.13(1) 2.403 2.21(1) 2.403 0.65(2) 0.76(5) 0.10(2) 0.12(2) 0.09(2) —2rc 218.6(3) 250 217.8(3) 250 269.2(6) 240 254.3(5) 240 253.5(6) 240 7.1(2) 9.9(3) 0.56(3) 0.45(3) 0.71(3) l1 237.1 241 237.5 241 239.6 230 233.9 230 233.9 230 23.4 23.0 20.5 20.7 20.5 l2 232.2 228 233.4 228 239.0 230 233.2 230 232.1 230 22.3 22.9 20.5 20.6 20.4 l3 50.7 19 53.1 19 9.5 20 12.8 20 12.5 20 12.8 15.8 1.5 1.7 1.6 e 0.15 0.46 0.12 0.46 0.01 0 0.02 0 0.06 0 0.45 0.04 0.07 0.10 0.27 Rij 1.24 1.26 1.02 1.01 1.01 2.15 2.11 2.19 2.14 2.19 d1 0.647 0.665 0.649 0.665 0.859 0.730 0.812 0.730 0.806 0.730 1.046 0.010 0.841 0.815 0.842 d2 0.595 0.563 0.609 0.563 0.158 0.267 0.198 0.267 0.205 0.267 1.099 1.101 0.354 1.328 1.353 Dij 1.244(1) 1.255(1) 1.015(1) 1.015(1) 1.014(1) 2.142(1) 2.106(1) 2.175(1) 2.126(1) 2.159(1) Fig. 9 Plots of the total static model electron density (r), and the negative Laplacian of the total static model electron density (2—2r) in the nickel– nitrite plane. (a) Relief plot of r.The density is truncated at 256 e Å23. (b) Contour plot of r. The contours are drawn on a logarithmic scale, 1.0 × 2N e Å23. (c) Relief plot of 2—2r. Negative and positive regions are truncated at 2256 and 256 e Å25. (d) Contour plot of 2—2r. The contours are drawn on a logarithmic scale, 4.0 × 2N e Å25J. Chem. Soc., Dalton Trans., 1997, Pages 2227–2240 2235 Fig. 10 Plots of 2—2r for the nitrite ion based on the total static model density. (a) Relief plot in the Ni]N(1) plane perpendicular to the molecular plane. Truncation as in Fig. 9(c). (b) Contour plot of the Ni]N(1) plane. Contour interval as in Fig. 9(d). (c) Relief plot in the N(1)]O(1) plane perpendicular to the molecular plane. Truncation as in Fig. 9(c). (d) Contour plot of the N(1)]O(1) plane. Contour interval as in Fig. 9(d) the radial form of the density functions. It should again be stressed that the present study employs more radial flexibility than is normally used in experimental charge-density studies, and we also have an unusually precise deconvolution of the thermal motion.We are therefore inclined to believe in the larger negative values found for ND3. As expected the shapes of the electron density at the N]D bond CPs are almost cylindrical, which is reflected in the small bond ellipticities (0.01, 0.02 and 0.06). These values are unaffected by introduction of extra radial flexibility on the ammine group. The bond paths are close to the internuclear axis as seen in the equal values of Rij and Dij.Figgis et al.9b observed a weak band at 1850 cm21 in the IR spectrum which they assigned to a weaker N]D(1) bond compared to N]D(2) and N]D(3). The monopole population on D(1) is somewhat smaller than the corresponding D(2) and D(3) monopole populations, and rc is slightly smaller for N]D(1) than for N]D(2) and N]D(3). On the other hand the Laplacian shows, if anything, a more covalent N]D(1) bond.Furthermore, we obtain almost equivalent N]D internuclear distances. Also the O ? ? ? D distances are almost equivalent. The geometries of the hydrogen bonds are slightly different; for D(2) and D(3) the N]D ? ? ? O angle is close to 1808 [179.1(1) and 175.4(1)] compared to 165.9(1)8 for D(1). The value of rc is very similar for all three hydrogen bonds which have small positive values of the Laplacian at the bond CP reflecting closed-shell interactions. For all three bonds the bond paths follow the internuclear axis with Rij being only about 0.01–0.03 Å longer than the internuclear separation.In conclusion the topological analysis is not conclusive about the proposed weakening of N]D(1). The band at 1850 cm21 in the IR spectrum may have another origin. Metal–ligand interaction The topological analysis shows that both the nitrite ion and the ammonia ligands have positive Laplacians at the metal–ligand bond CP. These values are positive irrespective of the choice of model.Thus both ligand types are co-ordinated to nickel in predominantly ionic interactions. Figgis et al.9c showed that the crystal-field model is a reasonable first approximation to2236 J. Chem. Soc., Dalton Trans., 1997, Pages 2227–2240 describe the bonding in compound 2 at a qualitative level. This is faithfully mapped in the Laplacian of the density. In all models the nickel–ammine bond consistently has a larger positive value of the Laplacian compared to the nickel–nitrite bond.This indicates that the nitrite ligand has a greater degree of covalency in the bond compared to the nickel–ammine bond, as expected from the relative positions of the ligands in the spectrochemical series. In the previous 110 K study Figgis et al.9c found that the nitrite ligand is a stronger s donor than the ammonia ligand. No evidence was found for a p contribution to the metal–ligand bonding for either the nickel–nitrite or –ammine bond.In that study a valence-orbital model gave an adequate description of the data, but as will be shown below it is not adequate in the description of the more accurate less thermally affected 9 K data. There is a clear difference between the ellipticities in the two metal–ligand bonds. The nickel–nitrite bond has an ellipticity of 0.45 whereas the nickel–ammine bond only shows weak ellipticity (e = 0.04). For the nitrite bond the ellipticity is perpendicular to the mirror plane.Even though there is some Fig. 11 Plots of 2—2r for the ammonia molecule in the N(2)] H(1)]H(2) plane. (a) Relief plot with truncation as in Fig. 9(c). (b) Contour plot with contour interval as in Fig. 9(d) uncertainty about the exact values of the ellipticities, this study suggests that p interaction takes place in the nickel–nitrite bond, which may be regarded as having a p as well as a s component in the covalent part. The ammine ligand is mostly bonded through s interaction for the covalent part of the bonding.This conforms with the conventional ligand-field angularoverlap- model contention that primary amine ligand bonding to metals does not contain a p component. If we omit the extra radial flexibility on N(1) the ellipticity in the nickel–nitrite bond drops to 0.06. In the simple ligand-field picture we expect p-back donation from the nickel ion into the p* antibonding orbital of the nitrite group since we are formally dealing with an electron-rich 20-electron complex.The archetype examples of metal to ligand p-back donation are carbonyl complexes. Evidence for p-back bonding has earlier been given based on deformation densities in a study of [Cr(CO)6].29 Furthermore a large amount of indirect evidence has accumulated supporting the idea of p-back donation in the bonding. Among this indirect evidence is the observed lowering of the CO stretching frequencies in the IR spectrum.30 Other evidence includes the fact that the CO internuclear distance is increased by 0.01–0.06 Å for coordinated carbonyls compared to free CO.31 However, unlike considerations about the internuclear distances, the topological evidence is a direct experimental validation of the p-backdonation scheme, which is a generally accepted simple bonding model for transition-metal complexes.Laplacian of the Electron Density In the preceeding paragraph it was shown that the values of the second derivatives of r can be used to describe chemical bonding in transition-metal complexes. In general the Laplacian shows where charge is locally concentrated (—2r < 0) or depleted (—2r > 0).This is in contrast to the total density which is a smoothly decaying function away from the nuclei. In Fig. 9 plots of r and 2—2r are shown for the nickel–nitrite plane. Note that the negative of the Laplacian, 2—2r, is plotted in order that positive regions correspond to regions of charge concentration.Atoms in molecules have a shell structure of alternating charge concentrations and charge depletions corresponding to the quantum shells. This shell structure is seen as positive and negative spikes in 2—2r. The outermost region of charge concentration has been termed the valence-shell charge concentration (VSCC).13 It has been shown that the maxima in the VSCC correspond in number and positions to the electron pairs of the Lewis model.31 The local maxima in our experimental VSCC are clearly seen in Fig. 9. For the oxygen atoms there are maxima in the bond directions and for lone-pair charge concentrations. If we map the Laplacian in the plane perpendicular to the mirror plane we observe new features of the VSCC for the nitrite group atoms. In Fig. 10 maps of the negative of the Laplacian for the Ni]N(1) and N(1)]O(1) bonds perpendicular to the molecular plane are shown. While the oxygen atom clearly has strong charge concentrations in all directions, the nitrogen atom shows much more structure in the VSCC.For the nitrogen atom there are clear regions in charge depletion in the valence shell. The regions of charge depletion in the VSCC of N(1) confirm the simple view of the nitrite group with a ‘negative’ oxygen and a ‘positive’ nitrogen. The Ni atom shows even greater charge depletion in the valence shell, and it appears totally stripped for its outer charge concentration. This corroborates the expectation of a nickel ion which has lost the 4s electrons.In Fig. 11 the negative of the Laplacian is plotted for a section through the ND3 group. Again the VSCC faithfully shows the charge concentrations expected in the N]D bonds. The nitrogen atom in ND3 has two other maxima in the VSCC besides the ones shown in Fig. 11. These are in the directions ofJ. Chem. Soc., Dalton Trans., 1997, Pages 2227–2240 2237 Ni and D(1). As can be seen the N]D bonds appear very similar even though the value of the Laplacian is greater in the N]D(1) bond than in N]D(2) and N]D(3).The ammonia group has clear regions of charge depletion on the faces of the NiND3 tetrahedra opposite to the atoms. There are no non-bonded charge concentrations in the VSCC. In Fig. 12 the negative of the Laplacian is plotted in the Ni]N(1)]N(2) plane. In this plot both the donating nitrogen lone pairs are clearly seen. The hydrogen-bond interactions are also reflected in the VSCC of the atoms. In Fig. 13 the negative of the Laplacian is plotted for the representative N(2)]H(1) ? ? ? O(1) plane. A small maximum in the VSCC of O(1) is seen in the direction of the hydrogen atom. The numerical value of the Laplacian is relatively small in the hydrogen-bonding region. The important point is that the interaction can be classified as a normal hydrogen bond. Recent topological analyses of small organic molecules with very short intramolecular hydrogen bonds have shown that in special cases the interaction can change character and become predominantly covalent if the O ? ? ? H distance is sufficiently short.14c,f Fig. 12 Plots of 2—2r in the Ni]N(1)]N(2) plane. Details as in Fig. 11 Valence Orbital Models The topological analysis presented above was based on a very flexible electron-density model. Such a model is necessary in order to exhaust the information in the data. With the flexible model only the total electron density is well modelled, and atomic properties can be derived only through topological analysis.Conventionally, more restricted models are used in experimental charge-density analysis. Restricted models allow better pseudo-atom properties such as orbital populations to be derived, but at the expense of only being semiquantitative. The most chemically inspired restricted model is the valence-orbital model advanced by Figgis et al.32 In order to get a qualitative, but also more chemically intuitive, understanding of the chemistry in the complex in conventional terms, we have also refined various valence-orbital models. The valence-orbital approach can be viewed as a restricted multipole refinement where the fitting functions are limited to those deemed important based on chemical intuition.For a discussion of different electrondensity modelling techniques readers are referred to our paper on compound 1.2 In the valence-orbital refinements the program ASRED30 Fig. 13 Plots of 2—2r in the O(1) ? ? ? H(1)]N(2) plane.Details as in Fig. 112238 J. Chem. Soc., Dalton Trans., 1997, Pages 2227–2240 was used. This minimises the function Sw(Iobs 2 Icalc)2 where w = 1/s2(I). Initially we refined a model similar to the aspherical valence model of the 110 K X-ray study 9c using only the new 9(1) K X-ray data. This allows direct comparison with the results obtained at 110 K and thus demonstrates the advantages of doing crystallography at very low temperatures. The model contained, besides two scale factors, positional and anisotropic thermal parameters on all atoms except for deuterium where isotropic thermal parameters were used.Corrections were made for type II extinction 17 and for multiple scattering by a twoparameter function.33 Furthermore the cell content was constrained to the correct number of electrons (228.6). The electron-density model included core, 3d and 4p functions on Ni, cores, three sp2 hybrids and one pp function on O(1), O(2) and N(1), core and four sp3 hybrid functions on N(2) and deuterium 1s functions.For all heavy atoms a radial k parameter was refined for the valence shell. The local atomic coordinate systems were chosen as in the 110 K study. The d orbitals on the nickel atom are labelled according to the Cartesian axis system appropriate to the molecular point group, x||b, y||c*, z||a. Note that this is not the ‘natural’ octahedral system, and our axis system gives dxy and dz2 not dx22y2 and dz2 as the octahedral eg set.The refinement results are summarised in Table 4. Scattering factors were calculated for core and valence functions from the atomic wavefunctions of Clementi and Roetti 18 using the program JCALC.34 The valence-orbital model provides, contrary to the 110 K study, a quite poor fit to the data and the residual map, Fig. 14, reveals distinct density features which are not fitted by the model. Furthermore many parameters refine to values that are not meaningful.However if we use a cut-off in (sin q)/l equal to the resolution of the 110 K study [(sin q)/ l < 1.08 Å21] much improved residual maps are obtained, Fig. 15. The use of a large amount of accurate high-order data reveals that the valence-orbital model is too rigid to both fit the detailed core information present in the highorder data and describe the reorganisation of charge in the valence regions. The model is too simple an approximation to separate electronic and vibrational effects properly when very precise infor- Fig. 14 Residual density in the Ni]N(1)]O(1) plane based on the valence-orbital model and including all data in the refinement. The resolution of the map is 0.8 Å21 and the contour interval is 0.1 e Å23. Solid lines represent positive contours and broken lines negative contours mation about the thermal motion is present. To demonstrate this even further we compare the values of the thermal parameters obtained with the valence-orbital model refinement to the values from the 13 K time-of-flight neutron diffraction data.The r.m.s. difference between the thermal parameters is ·DU2Ò� �� = 0.001 14 Å2. This is more than twice that obtained with the very flexible multipole model. For the valence-orbital refinement with a cut-off in (sin q)/l of 1.08 Å21 the pseudo-atom charges are Ni(11.55), O(1)- (20.40), O(2)(20.44), N(1)(10.29), N(2)(20.48), D(1)(10.16), D(2)(10.17) and D(3)(10.03). These are in good correspondence with what we would qualitatively expect for the system, but as explained above the valence-orbital model does not fully describe the data.With regard to the orbital populations it is clear that the very low 3dxz population on Ni obtained by Figgis et al.9c from 110 K X-ray data is not reproduced by the present more precise data. In general the orbital populations agree better with theory for the 9 K data. It is notable that in all valence-orbital refinements we obtain significantly different Table 4 Valence-orbital parameters No Nv RF (3s level) (0s level) RF2 (0s level) Goodness of fit Scale factor 1 Scale factor 2 Ni 3dxy 3dxz 3dyz 3dx22y2 3dz2 4px 4py 4pz k(3d) N(1) pp sp2(1) sp2(2) sp2(3) k(sp2) N(2) sp3(1) sp3(2) sp3(3) sp3(4) k(sp2) O(1) pp sp2(1) sp2(2) sp2(3) k(sp2) O(2) pp sp2(1) sp2(2) sp2(3) k(sp2) H(1) 1s H(2) 1s H(3) 1s Model 1 4016 c 80 0.0314 0.0493 0.0722 1.336 1.670(6) 1.461(10) 1.13(6) 1.61(6) 2.02(5) 1.69(6) 1.33(5) 2.1(3) 0.4(3) 20.6(3) 0.99(1) 0.24(3) 1.25(4) 1.12(4) 1.46(4) 0.9810(1) 0.49(5) 0.14(6) 0.76(5) 0.80(5) 0.9809(1) 1.01(6) 1.84(4) 1.53(4) 1.88(4) 0.0795(2) 0.98(6) 1.84(4) 1.59(4) 1.84(4) 0.9798(2) 1.14(7) 1.50(10) 3.43(14) Model 1 (sin q)/l < 1.08 2083 80 0.0173 0.0309 0.0430 0.998 1.620(10) Not refined 1.09(6) 1.68(5) 2.00(5) 1.72(6) 1.27(6) 20.18(26) 0.26(23) 0.61(27) 0.88(2) 0.66(6) 1.55(5) 1.11(5) 1.39(5) 0.96(1) 1.50(4) 1.33(6) 1.33(5) 1.32(4) 1.01(1) 1.31(6) 1.76(4) 1.58(4) 1.75(4) 0.99(1) 1.39(6) 1.76(4) 1.54(4) 1.75(4) 1.00(4) 0.84(5) 0.83(5) 0.97(6) 110 Ka 2109 45 d 0.023 0.035 2.1 1.28(7) 1.31(6) 1.80(6) 1.84(6) 1.36(8) 1.59(20) 0.09(29) 20.19(20) 1.013(6) 0.86(4) 1.66(4) 1.32(4) 1.26(4) 1.59(3) 1.21(3) 1.21(3) 1.24(3) 1.052(5) 1.61(4) 1.61(4) e 1.43(3) 1.62(3) 1.046(4) 1.43(4) 1.74(4) 1.42(3) 1.64(3) 1.046(4) 0.77(4) 0.92(3) 0.93(3) Theoryb 1.11 1.94 1.99 1.97 1.31 0.06 0.04 0.31 1.07 1.70 1.17 1.17 1.64 1.57 1.57 1.57 1.47 1.61 1.69 1.69 1.47 1.61 1.69 1.69 0.55 0.55 0.55 a Figgis et al.9c b Chandler et al.7 c Ten reflections had negative intensities and were included in the ASRED refinements. d Positional and thermal parameters fixed at values from a high-order refinement.e pp orbital.J. Chem. Soc., Dalton Trans., 1997, Pages 2227–2240 2239 values of 3dxz compared to 3dyz. This is in accord with the significant orthorhombic distortion of the complex. The theoretical calculations by Chandler et al.12 are much less convincing in showing this difference.The diffuse 4p functions attain appreciable values but also have quite high uncertainties. However based on the results of the flexible multipole model it is reasonable to conclude that the 4p mixing is significant and that the theoretical calculations probably underestimate this component. Conclusion It has been shown that diffraction data measured on Ni(ND3)4- (NO2)2 at very low temperatures cannot be explained by use of a simple valence-orbital model.This is consistent with results of a previous analysis of 9 K X–N diffraction data on (ND4)2Cu(SO4)2?6D2O.2,7 This study shows that detailed understanding of transition metal to ligand bonding requires a higher level of theory. To model the data adequately a very flexible electron-density model must be used. The chosen atom-centred multipole model describes the total electron-density distribution well, but it cannot express individual pseudo-atom properties.Owing to the very low temperature of the experiment a good deconvolution of the thermal motion has been achieved. The resulting static model density derived from the diffraction data has sufficient accuracy to allow topological analysis of the electron density. The topological analysis gives a detailed quantitative account of the metal–ligand interactions in the complex. The fact that precise topological information can be obtained even for transition-metal complexes opens up new possibilities for quantitative understanding of transition metal to ligand bonding. If topological information can be obtained on a number of complexes, trends and characteristics of the topology of various bonding types may be discovered. Acknowledgements B.B. I. gratefully acknowledges support for this work from the Danish Natural Science Research Council and the Carlsberg Foundation. The latter is thanked for the low-temperature diffractometer in Århus.The US Department of Energy is Fig. 15 Residual density in the Ni]N(1)]O(1) plane based on the valence-orbital model and including only data with (sin q)/l < 1.08 Å21 in the refinement. The resolution of the map is 0.8 Å21. Contours as for Fig. 14 thanked for granting beamtime at the Intense Pulsed Neutron Source, Argonne National Laboratory. Work at Argonne is supported by the US Department of Energy, BES-Materials Science, under contract No. W-31-109-ENG-38.B. N. F. and P. A. R. acknowledge support from the Australian Research Council and Department of Industry. References 1 B. N. Figgis and P. A. Reynolds, Int. Rev. Phys. Chem., 1986, 5, 265; P. Coppens, Annu. Rev. Phys. Chem., 1992, 43, 663. 2 B. N. Figgis, B. B. Iversen, F. K. Larsen and P. A. Reynolds, Acta Crystallogr., Sect. B, 1993, 49, 794. 3 B. N. Figgis, L. Khor, E. S. Kucharski and P. A. Reynolds, Acta Crystallogr., Sect. B, 1992, 48, 144. 4 See, for instance, F. K. Larsen, Acta Crystallogr., Sect.B, 1995, 51, 468. 5 R. F. Stewart, Acta Crystallogr., Sect. A, 1976, 32, 565; N. K. Hansen and P. Coppens, Acta Crystallogr., Sect. A, 1978, 34, 909. 6 B. B. Iversen, F. K. Larsen, B. N. Figgis, P. A. Reynolds and A. J. Schultz, Acta Crystallogr., Sect. B, 1996, 52, 923. 7 G. S. Chandler, B. N. Figgis, P. A. Reynolds and S. K. Wolff, Chem. Phys. Lett., 1994, 225, 421. 8 C. K. Johnson, ORTEP, Report ORNL-5138, Oak Ridge National Laboratory, Oak Ridge, TN, 1976. 9 (a) M. A. Porai-Koshits and L. M. Dikareva, Sov. Phys.-Crystallogr. (Engl. Transl.), 1960, 4, 611; (b) B. N. Figgis, P. A. Reynolds, A. H. White, G. A. Williams and S. J. Wright, J. Chem. Soc., Dalton Trans., 1981, 997; (c) B. N. Figgis, P. A. Reynolds and S. J. Wright, J. Am. Chem. Soc., 1983, 105, 434. 10 B. N. Figgis, P. A. Reynolds and G. A. Williams, Aust. J. Chem., 1981, 34, 993. 11 B. N. Figgis, P.ynolds and R. Mason, J. Am. Chem. Soc., 1983, 105, 440. 12 G. S. Chandler, R. J. Deeth, B. N. Figgis and P. A. Reynolds, J. Chem. Soc., Dalton Trans., 1990, 1417. 13 R. F. W. Bader, Atoms in Molecules. A Quantum Theory, Cambridge University Press, 1991. 14 (a) R. F. Stewart, in The application of charge density research to chemistry and drug design, eds. G. A. Jeffrey and J. F. Pinielle, NATO ASI ser., Ser. B, 1991, 250; (b) B. B. Iversen, F. K. Larsen, M. Souhassou and M. Takata, Acta Crystallogr., Sect. B, 1995, 51, 580; (c) C. Flensburg, S. Larsen and R. F. Stewart, J. Phys. Chem., 1995, 99, 10 130; (d) R. Destro, R. Bianchi, C. Gatti and F. Merati, Chem. Phys. Lett., 1991, 186, 47; (e) R. Destro and F. Merati, Z. Naturforsch., Teil A, 1993, 48, 99; ( f ) G. K. H. Madsen, B. B. Iversen, F. K. Larsen, M. Kapon, G. Reisner and F. Herbstein, unpublished work; (g) R. Destro and F. Merati, Acta Crystallogr., Sect. B, 1995, 51, 559. 15 T. Koritzansky, S. Howard, P. R. Mallison, Z. Su, T. Richter and N. K. Hansen, XD, a computer program package for multipole refinement and analysis of charge densities from diffraction data, Institute for Crystallography, Berlin, 1995. 16 N. K. Hansen and P. Coppens, Acta Crystallogr., Sect. A, 1978, 34, 909. 17 P. Becker and P. Coppens, Acta Crystallogr., Sect. A, 1974, 30, 129. 18 E. Clementi and C. Roetti, At. Data Nucl. Data Tables, 1974, 14, 177. 19 R. F. Stewart, E. R. Davidson and K. T. Simpson, J. Chem. Phys., 1965, 42, 3175. 20 P. Coppens, T. N. Guru Row, P. Leung, E. D. Stevens, P. J. Becker and Y. W. Yang, Acta Crystallogr., Sect. A, 1979, 35, 63. 21 H. F. Schaefer, J. R. Thomas, Y. Yamaguchi, B. J. DeLeeuw and G. Vacek, Modern Electronic Structure Theory, ed. D. R. Yarkony, World Scientific, Singapore, 1995, p. 1. 22 M. W. Schmidt, K. K. Baldridge, J. A. Boatz, J. H. Jensen, S. Koseki, M. S. Gordon, K. A. Nguyen, T. L. Windus and S. T. Elbert, QOPH Bull., 1990, 10, 52. 23 (a) T. Kikkawa, S. Ohba, T. Saito, S. Kamata and S. Iwata, Acta Crystallogr., Sect. B, 1987, 43, 83; (b) D. W. J. Cruickshank and M. J. Eisenstein, J. Comput. Chem., 1987, 8, 1, 6. 24 M. Okuda, S. Ohba, Y. Saito, T. Ito and I. Shibuya, Acta Crystallogr., Sect. B, 1990, 46, 343. 25 S. Ohba, K. Toriumi, S. Sato and Y. Saito, Acta Crystallogr., Sect. B, 1978, 34, 3535. 26 R. Bianchi, C. Gatti, V. Adovasio and M. Nardelli, Acta Crystallogr., Sect. B, 1996, 52, 471. 27 R. Dovesi, V. R. Sauders and C. Roetti, CRYSTAL 92 User Documentation, University of Torino, 1992.2240 J. Chem. Soc., Dalton Trans., 1997, Pages 2227–2240 28 W. T. Klooster, S. Swaminathan, R. Nanni and B. M. Craven, Acta Crystallogr., Sect. B, 1992, 42, 217. 29 B. Rees and A. Mitschler, J. Am. Chem. Soc., 1976, 98, 7918. 30 J. P. Collman, L. S. Hegedus, J. R. Norton and R. G. Finke, Principles and applications of organotransition metal chemistry, University Science Books, Mill Valley, CA, 1987. 31 R. F. W. Bader and H. Essen, J. Chem. Phys., 1984, 93, 2946; R. F. W. Bader, P. J. MacDougall and C. D. H. Lau, J. Am. Chem. Soc., 1984, 106, 1594. 32 B. N. Figgis, P. A. Reynolds and G. A. Williams, J. Chem. Soc., Dalton Trans., 1980, 2339. 33 B. N. Figgis, E. S. Kucharski and P. A. Reynolds, Acta Crystallogr., Sect. B, 1989, 45, 232. 34 B. N. Figgis, P. A. Reynolds and A. H. White, J. Chem. Soc., Dalton Trans., 1987, 1737. Received 21st March 1997; Paper 7/01978E

ISSN:1477-9226    

DOI:10.1039/a701978e

出版商:RSC    

年代:1997

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1999, 2231–2236 2231 New P–S–N containing ring systems. Reaction of 2,4-(naphthalene- 1,8-diyl)-1,3,2,4-dithiadiphosphetane 2,4-disulfide with methylbis(trimethylsilyl)amine Petr Kilián,a Jaromír Marek,a Radek Marek,a Jaromír Tousek,a Otakar Humpa,a Alexandra M. Z. Slawin,b Jirí Touzín,a Josef Novosad a and J. Derek Woollins *b a Department of Chemistry, Masaryk University, Kotlárská 2, Brno 611 37, Czech Republic b Department of Chemistry, University of St Andrews, St Andrews, Fife, UK KY16 9ST.E-mail: J.D.Woollins@st-andrews.ac.uk Received 3rd March 1999, Accepted 18th May 1999 From the reaction of 2,4-(naphthalene-1,8-diyl)-1,3,2,4-dithiadiphosphetane 2,4-disulfide 1 with methylbis- (trimethylsilyl)amine 2 in dichloromethane a product of the P2S2 ring cleavage, (C10H6)P(S)(SSiMe3)(m-S)P(S)- (NMeSiMe3) 3, was isolated, existing in an equilibrium between two diastereomeric forms in solution. Compound 3 was also obtained when toluene was used as a solvent. Its subsequent reaction with pyridine (py) led to the desilylated ionic product [Hpy1][(C10H6)P(S)(NHMe)(m-S)PS2 2] 4, which reacts with tetraphenylphosphonium chloride to give [PPh4 1][(C10H6)P(S)(NHMe)(m-S)PS2 2] 5.When acetonitrile was used as a solvent a cage compound 2,6-(naphthalene- 1,8-diyl)-3,4-dimethyl-1,3,5,2l5,6l5-thiadiazadiphosphinine 2,6-disulfide 6 containing a six-membered CN2P2S heterocycle was obtained. The new compounds were studied spectroscopically (1- and 2-D high resolution NMR, IR, MS) and in the cases of 3, 5 and 6 by X-ray crystallography.The reaction of trans-diorganodithiadiphosphetane disulfides (RPS2)2 (R = Me, Et or C6H4OMe-p) with methylbis(trimethylsilyl) amine when performed in dichloromethane leads to N-alkylated thiazadiphosphetidine disulfides (Scheme 1), and the unstable intermediate in square brackets has been proposed.1 We have previously reported on the synthesis of cis-2,4-(naphthalene-1,8-diyl)-1,3,2,4-dithiadiphosphetane 2,4- disulfide 2,3 1 from the reaction of 1-bromonaphthalene with P4S10 and have noted some striking diVerences in reactivity between this cage compound and the more well established trans systems.Here we describe some studies to determine how the structural anomaly of this compound (a bridge created by organic substituent naphthalene-1,8-diyl, joining both phosphorus atoms) aVects the reaction and structure of the products.We have found that the reaction of 1 with NMe(SiMe3)2 2 gives a product 3 which has an analogous structure to that of the unstable intermediate proposed in the case of the above described reactions of (RPS2)2 with 2 (Scheme 1). Even prolonged heating of 3 in boiling toluene or addition of base (pyridine) did not lead to elimination of bis(trimethylsilyl) sulfide and formation of the thiazadiphosphetidine disulfide; in the latter case a desilylated pyridinium salt 4 was obtained instead, which was for easier purification converted into its tetraphenyl- Scheme 1 P S P S S R R S P S P S R R S S N Me SiMe3 SiMe3 P S P N S R R S Me (Me3Si)2NMe 2 - (Me3Si)2S phosphonium salt.Moreover we studied the reaction of 1 with 2 in acetonitrile, leading to a new cage compound 2,6-(naphthalene- 1,8-diyl)-3,4-bis(methyl)-1,3,5,2l5,6l5-thiadiazadiphosphinine 2,6-disulfide containing a six-membered unsaturated heterocycle CN2P2S. Experimental All manipulations were performed under dried nitrogen gas in Schlenk vessels.Compound 1 was prepared by the reaction of P4S10 with 1-bromonaphthalene 3 and recrystallized from dichloromethane. Anhydrous pyridine, tetraphenylphosphonium chloride and 2 were obtained from Aldrich; 2 was distilled before use. Solvents were purified and/or dried using standard methods. The IR spectra were recorded in Nujol mull in cells equipped with KBr windows or as pellets with KBr on a Bruker IFS 28 spectrometer or on a Perkin-Elmer system 2000, mass spectra by the EPSRC National Mass Spectrometry Service Centre, University of Wales, Swansea.Microanalyses were carried out by the Department of Chemistry, Palacky University, CZ, and Loughborough Chemistry Departmental service, UK. Preparations (C10H6)P(S)(SSiMe3)(Ï-S)P(S)(NMeSiMe3) 3. A suspension of compound 1 (0.50 g, 1.58 mmol) in 5 cm3 CH2Cl2 and 2 (0.52 cm3, 2.37 mmol) was stirred at room temperature for 8 h. The resulting clear yellow solution was concentrated in vacuo to 3 cm3 and cooled to 220 8C for 1d.The resulting yellow crystals of 3 (some of them suitable for X-ray crystallography) were filtered oV, washed with hexane and dried in vacuo. A further portion of less pure 3 was obtained from the mother-liquor: the oily product obtained after evaporation of the solvent in vacuo was stirred for several hours with 10 cm3 n-hexane, the resulting suspension was filtered oV, washed with 5 cm3 of n-hexane and dried in vacuo.Compound 3 decomposes very fast when exposed to air moisture. Yield of the 1st fraction 228 mg (29.3%), mp 156–158 8C (Found: C, 41.8; H, 5.5; N, 2.7; S, 25.3. C17H27NP2S4Si2 requires C, 41.5; H, 5.5; N, 2.8; S, 26.1%). IR2232 J. Chem. Soc., Dalton Trans., 1999, 2231–2236 (n& max/cm21) 682vs and 645s [n(P]] S)]. NMR (500 MHz, 31P-{1H} and 1H in CDCl3, 13C-{1H} in CH2Cl2, for numbering of atoms see Fig. 1): major isomer, 31P-{1H} d 67.8 [d, P(9)], 55.1 [d, P(1)], 2J(PP) = 13.0 Hz; 1H d 8.82 [1 H, dd, 3J(PH) = 21.8, 3J(HH) = 7.2, H2], 8.50 [1 H, dd, 3J(PH) = 19.0, 3J(HH) = 7.2, H8], 8.10 (2 H, m, H4 and H6), 7.69 (2 H, m, H3 and H7), 2.45 [3 H, d, 3J(PH) = 17.3 Hz, NMe], 0.61 (9 H, s, SSiMe3) and 0.26 (9 H, s, NSiMe3); 13C-{1H} d 135.8–125.1 (naphthalene ring carbons), 33.9 (s, NMe), 2.1 [d, 4J(PC) = 3.5 Hz, SSiMe3] and 1.8 (s, NSiMe3); 15N d 52.1; minor isomer, 31P-{1H} d 66.1 [d, P(9)], 57.2 [d, P(1)], 2J(PP) = 15.5 Hz; 1H d 8.89 [1 H, m, H8], 8.86 [1 H, m, H2], 8.10 (2 H, m, H4 and H6), 7.69 (2 H, m, H3 and H7), 2.78 [3 H, d, 3J(PH) = 16.9 Hz, NMe], 0.71 (9 H, s, SSiMe3) and 20.12 (9 H, s, NSiMe3); 13C-{1H} d 135.8–125.1 (naphthalene ring carbons), 33.6 (s, NMe), 2.4 [d, 4J(PC) = 3.5 Hz, SSiMe3] and 0.7 (s, NSiMe3); 15N d 50.6.[Hpy1][(C10H6)P(S)(NHMe)(Ï-S)PS2 2] 4. A suspension of compound 1 (1.0 g, 3.16 mmol) in 10 cm3 toluene and 2 (0.80 cm3, 3.64 mmol) was stirred at room temperature for two days.Pyridine (0.38 cm3, 4.74 mmol) was slowly added to the resulting clear yellow solution with stirring. The resulting suspension was stirred for 5 h, the solid was filtered oV, washed with cold toluene (5 cm3) and dried in vacuo. Yield 1.135 g (84.1%), decomp. above 120 8C (Found : C, 44.6; H, 4.3; N, 5.7. C8H8NPS2 requires C, 45.1; H, 3.8; N, 6.6%). IR (n& max/cm21) 1607m, 1485m [nring(py)], 1589 (sh) [d(NH)], 655vs and 673vs [n(P]] S)].NMR [400 MHz, 31P-{1H} in py, 1H and 13C-{1H} in d6-dimethyl sulfoxide (d6-dmso)]: 31P-{1H} (d values calculated for AB system) d 70.47 (d), 70.09 (d), 2J(PP) = 12.0 Hz; 1H d 8.88 (2 H, m, H13), 8.69 (1 H, m, H2), 8.58–8.51 (2 H, two overlapping multiplets, H8 and H15), 8.19 (1 H, m, H6), 8.04– 7.97 (3 H, two overlaping multiplets, H4 and H14), 7.70 (1 H, m, H7), 7.60 (1 H, m, H3), 6.2 (2 H, br s, H12 and H16), 2.61 [3 H, d, 3J(PH) = 20.0 Hz, H11]; 13C-{1H} d 146.1 (br s, C13), 143.6 (br s, C15), 142.7 [d, 1J(PC) = 81.5, C1], 135.2–134.9 (m, C6 and C8], 134.2 [dd, 3J(PC) = 9.5, 11.3, C5), 132.5 (m, C4), 132.0 [dd, 1J(PC) = 110, 3J(PC) = 2.1, C9], 129.8 [d, 2J(PC) = 13.3 Hz, C2], 128.6 [dd, 2J(PC) = 8.0, 9.8, C10], 127.7 (br s, C14), 125.9–125.5 (m, C3 and C7) and 27.0 [d, 2J(PC) = 4.0 Hz, C11].Mass spectrum (FAB1): m/z 427 (M 1 H1), 370 (M 2 py 1 Na), 348 (M 2 py 1 H) and 317 (M 2 Hpy 2 S). [PPh4 1][(C10H6)P(S)(NHMe)(Ï-S)PS2 2] 5. A solution of compound 4 (0.25 g, 0.58 mmol) in 3 cm3 pyridine was added to a stirred solution of PPh4Cl (0.24 cm3, 0.64 mmol) in 4 cm3 pyridine.The volume of the solvent was reduced to half by its evaporating in vacuo and 60 cm3 of cool water were added with vigorous stirring. The resulting white solid was filtered oV, washed with cool water and dried in vacuo. Recrystallized from dry CHCl3–hexane. Yield 0.297 g (74.0%), mp 195–197 8C (Found : C, 61.1; H, 4.5; N, 1.9. C35H30NP3S4 requires C, 61.3; H, 4.4; N, 2.0%).IR (n& max/cm21) 3133w [n(NH)], 1586w [d(NH)], 689s, 659vs [n(P]] S)]. NMR [400 MHz, 31P-{1H} in d6-dmso, 1H and 13C-{1H} in CDCl3; for numbering of atoms see Fig. 2]: 31P-{1H} (d values of anion calculated for AB system) d 73.03 (d), 72.23 (d), 2J(PP) = 11.6 Hz, 25.6 (s, PPh4 1); 1H d 8.79 (1 H, m, H2), 8.59 (1 H, m, H8), 7.82 (1 H, m, H6), 7.72–7.41 (21 H, three partially overlapping multiplets, H4 and hydrogens of PPh4 1 cation), 7.38 (1 H, m, H7), 7.27 (1 H, m, H3), 6.19 (1 H, m, NH), 2.60 [3 H, dd, 3J(PH) = 20.1, 3J(HH) = 5.60 Hz, H11]; 13C-{1H} d 142.8–124.7 (carbons of naphthalene entity), 136.1– 117.2 (carbons of PPh4 1 cation) and 27.2 (s, C11).Mass spectrum: (ES1) m/z 339 (PPh4 1) and 261 (PPh3 1 H); (ES2) m/z 346 (anion). C13H12N2P2S3 6. A suspension of compound 1 (0.50 g, 1.58 mmol) in 40 cm3 CH3CN and 2 (0.38 cm3, 1.74 mmol) was heated under reflux for 5 h. The resulting clear brown solution was concentrated in vacuo to 15 cm3, placed in a closed vessel in a Dewar flask with hot water and allowed to cool slowly (10 d) to ambient temperature. Yellow clear needles of 4 (some of them suitable for X-ray crystallography) were filtered oV, washed with 2 × 2 cm3 CH3CN and dried in vacuo.Yield 195 mg (34.8%), mp 218–220 8C (Found: C, 43.9; H, 3.4; N, 7.7; S, 27.9. C13H12N2P2S3 requires C, 44.1; H, 3.4; N, 7.9; S, 27.1%). IR (n& max/cm21) 1582vs [n(C]] N)], 655s, 639s [n(P]] S)].NMR (500 MHz, in CD2Cl2; for numbering of atoms see Fig. 3): 31P-{1H} d 56.9 (d, P9), 28.6 (d, P1), 2J(PP) = 11.2 Hz; 1H d 8.70 (2 H, m), 8.10 (2 H, m), 7.66 (2 H, m, H3 and H7), 3.48 [3 H, d, 3J(PH) = 11.0, H13] and 2.28 [3 H, d, 4J(PH) = 2 Hz, H12]; 13C-{1H} d 165.1 [d, 2J(PC) = 19.8, C11], 135.2 (s), 134.3 [d, J(PC) = 13.5], 134.1 [d, J(PC) = 15.3], 133.9 (s), 133.7 [t, J(PC) = 10.8], 131.2 [dd, 1J(PC) = 97, 3J(PC) = 8.1], 130.7 [d, 1J(PC) = 102], 130.4 [t, J(PC) = 9.0], 125.7 [d, J(PC) = 17.1], 125.5 [d, J(PC) = 18.0], 33.8 [d, 3J(PC) = 8.1, C13] and 28.3 [d, 3J(PC) = 18.9 Hz, C12].Mass spectrum (FAB1): m/z 377 (M 1 Na) and355 (M 1 H). Crystallography Details of the data collections and refinements are summarised in Table 1. Data for compounds 3 and 6 were collected using graphite-monochromatized Mo-Ka (l = 0.71073 Å) radiation on a KUMA KM-4 four circle k-axis diVractometer equipped with an Oxford Cryostream Cooler. Data were collected in the 2q range 4–508 with w–2q scan techniques.The application of semiempirical correction of intensities (program DIFABS)4 was in the case 3 negligible, whilst in the case 6 led to increased R factors. Data for 5 were collected using graphite-monochromatized Mo-Ka radiation and a 0.7 mm collimator on a KUMA KM4CCD four circle diVractometer. A total of 830 frames were recorded. The exposure time was 30 s for each 0.58 step. Data reduction was processed by 3D profile analysis software (KM4RED)5 which corrects intensities for Lorentzpolarization eVects too.All structures were solved by direct methods (SHELXS 86).6 Non-hydrogen atoms were refined anisotropically by the full-matrix least-squares procedure based on F 2 (SHELXL 93).7 All hydrogen atoms were found from the Fourier-diVerence synthesis and refined isotropically, except in the case of 5 where the hydrogen atoms on C(11), as well as hydrogens on the disordered carbon atoms in one of the phenyl rings in the cation, were fixed in idealized CH3 and aromatic positions respectively; their U and site occupation factor (s.o.f.) were driven by riding C atoms.CCDC reference number 186/1471. See http://www.rsc.org/suppdata/dt/1999/2231/ for crystallographic files in .cif format. NMR NMR Spectra were recorded either on a Bruker Avance DRX- 500 or DPX-400 spectrometer. Direct measurement of 1H, 13C and 31P spectra was carried out on a 5 mm (13C/19F/31P-{1H}) probehead. Homonuclear and heteronuclear chemical shift correlation spectra (DQF-COSY,8 NOESY,9 GHMBC,10 GSQMBC,11 31P–15N GHMQC12) were recorded either on a 5 mm triple resonance inverse probehead (1H-{13C} {BB} z-grad) equipped with a z-gradient coil or a 5 mm inverse probehead (1H-{13C} {15N} {31P} x,y,z-grad). For 1H and 13C NMR spectroscopy TMS was used as an internal standard; for 31P NMR 85% H3PO4 was used as an external standard.Gradient-enhanced heteronuclear multiple bond correlation (GHMBC) experiments used the following parameters: sequence published elsewhere; 13 for 1H–15N correlation, gradient ratio G1 :G2 :G3 = 42 : 18 :30 G cm21; for 1H–31P correlation, G1 :G2 :G3 = 36:6:14.9 G cm21; delay for evolution of long-range coupling constants, 60–120 ms.Gradient-enhanced single-quantum multiple bond correlation (GSQMBC) experiments: sequence; 11,13 for 1H–15N correlation, gradient ratioJ. Chem. Soc., Dalton Trans., 1999, 2231–2236 2233 G1:G2:G3 = 4.8 : 52.8 :±4.8 G cm21; for 1H–31P, G1:G2:G3 = 12 : 41.6 :±12 G cm21; evolution delays were as for GHMBC spectra. 31P–15N Gradient-enhanced heteronuclear multiple quantum correlation (31P–15N GHMQC) experiment: sequence; 14 gradient ratio, G1:G2 :G3 = 30 : 18 :24 G cm21; evolution delay, 20 ms. Results and discussion The 31P-{1H} NMR spectrum of the reaction mixture of compound 1 with 2 in dichloromethane showed two distinct AX signals. The spectra indicate asymmetrical substitution of the phosphorus atoms, whilst the similar chemical shifts and coupling constants of both sets of doublets indicates the presence of two diastereomers.The 1H NMR spectrum was in accord with these facts. The two stereogenic phosphorus centers and the existence of planar chirality in the proposed structure give rise to 4 possible diastereomers (Schemes 2 and 3). We performed semiempirical quantum chemical computations (AM1 method 15) which revealed that the isomers iii and iv converge to i and ii when minimized.The DHf of the thermo- Scheme 2 P P S S S S N SiMe3 Me SiMe3 P P S S S S N SiMe3 Me SiMe3 P P S S S N SiMe3 Me SiMe3 S P P S S S N SiMe3 Me SiMe3 S ii iii iv i Scheme 3 P P S S S P P S P P S N P P S S S S SiMe3 N Me SiMe3 Me S S S S N N C Me Me S P P S S S S NH Me P P S S S S NH Me 1 80ºC, MeCN - (Me3Si)2S (Me3Si)2NMe 2 + py - 2 MeC6H4SiMe3 PPh4Cl toluene - py.HCl 8 3 6 4 pyH+ PPh4 + 5 dynamically stable isomers i and ii are almost equal (220.432 and 220.436 kcal mol21).The proposed structure of the diastereomer i was confirmed by a crystal structure determination (Table 2, Fig. 1). Several crystals of the same type as these used for X-ray measurement (diastereomer i) were examined by 31P-{1H} and 1H NMR. After their dissolution in various solvents (e.g. dichloromethane, chloroform, diethyl ether, benzene) a mixture of both diastereomers was observed in solution. We suppose that chemical migration of the trimethylsilyl group bonded to the sulfur atom in the solution leads to an equilibrium mixture of diastereomers i and ii, as the stability of S–Si bond is rather low16 and the existence of the related anion 7 has been demonstrated by an X-ray diVraction determination.13 Moreover the presence of 1 and 2 was detected by 31P-{1H} and 1H NMR in the solution prepared from analytically pure crystals of 3, suggesting a second equilibrium between starting compounds and products.This equilibrium is pushed to the right hand products side by use of an excess of 1 during the synthesis.Distinct signals due to both diastereomers were found in the NMR spectra at ambient temperature. On the other hand, exchange “cross-peaks” found in standard phase-sensitive 1H–1H NOESY spectrum (700 ms mixing time) as well as in the 31P–31P NOESY spectrum (700 and 300 ms mixing time) indicate relatively fast interconversion. We found that the molar ratios of diastereomers depend on the solvent, e.g.in dichloromethane 1: 3 and in decalin 1 : 5 molar ratios were observed by 31P NMR . A set of 2-D NMR experiments (1H–1H COSY, 1H–15N, 1H–31P GHMBC, 1H–31P, 1H–15N GSQMBC, 31P–15N GHMQC) was made allowing us to assign 1H, 31P-{1H} and 15N NMR signals of both major and minor isomers and to find their connectivities. However it was impossible to distinguish which arrangement (i, ii) corresponds with the major and minor isomer. Correlations of major and minor isomer observed in 1H–31P, 1H–15N and 31P–15N 2-D spectra are demonstrated in Fig. 4. Four-bond correlations of the P atom with methyl hydrogens of its SSiMe3 group found for both isomers confirm that neither of them is present as the anion (with analogous structure to 7) in solution. The crystal structure of compound 3 (diastereomer i) reveals that the naphthalene part of the molecule and phosphorus atoms lie very close to the mean plane fitted to these atoms [maximum deviation 0.03 Å for C(8)].Exocyclic substituents lie in a cis position relative to the naphthalene ring, atoms S(1), S(9) and S(19) lie above the plane, S(SiMe3) and N(Me)(SiMe3) groups lie below the plane. The C3P2S ring is hinged with the C3P2 and P2S planes inclined by 47.38 with respect to each other. The opening of the P2S2 ring results in substantial lengthening of the P ? ? ? P distance vs. that in 1 2 (3.27 vs. 2.73 Å) and broad- Fig. 1 Molecular diagram of compound cis-3 (diastereomer i).2234 J.Chem. Soc., Dalton Trans., 1999, 2231–2236 Table 1 Details of the data collections and refinements for compounds 3, 5 and 6 3 5 6 Empirical formula M Crystal system Space group a/Å b/Å c/Å a/8 b/8 g/8 U/Å3 T/K Z m/mm21 Reflections measured Reflections independent (Rint) Final R1, wR2 [I > 2s(I)] C17H27NP2S4Si2 491.8 Triclinic P1� 7.178(2) 10.502(3) 17.184(5) 105.81(2) 93.25(2) 100.94(2) 1215.6(6) 150(2) 2 0.625 4439 4289 (0.0821) 0.0430, 0.1140 C35H30NP3S4 685.75 Triclinic P1� 9.8665(7) 10.4165(8) 17.131(2) 105.217(8) 102.994(7) 90.777(6) 1650.4(2) 150(2) 2 0.460 8096 5459 (0.0372) 0.0466, 0.1196 C13H12N2P2S3 354.4 Monoclinic P21/n 7.038(3) 27.939(3) 7.891(2) 103.81(3) 1506.8(8) 293(2) 4 0.693 2888 2668 (0.0318) 0.0347, 0.0874 ening of the P(1)–S(19)–P(9) angle [102.69(4) vs. 80.0(1)8]. However, the value of the P–S–P angle in 3 is comparable to that in the related molecules C10H6P(S)(SMe)(m-S)P(S)(OMe) 17 [103.0(1)8], C10H6P(S)(OCH2CH2OH)(m-S)P(S)(OCH2CH2- OH)18 [ 101.8(1)8], [C10H6P(S)(NHSiMe3)(m-S)P(S)2]2 anion 7 13 [100.44(6)8] as well as that in the anion of 5 [100.21(4)8].As expected, the terminal P]] S bond lengths are significantly shorter than those to the bridging sulfur. The molecules pack with partial face-to-face overlap of the naphthyl rings to form stacks along the a axis. The 31P-{1H} NMR spectrum of a mixture of compound 1 with 2 in toluene after prolonged heating to reflux temperature does not show the presence of any product other than 3.The ring-closure reaction leading to the thiazadiphosphetidine disulfide does not occur. Even addition of base (anhydrous pyridine) to the solution of 3 in toluene did not lead to the four membered PSPN ring-closure reaction with elimination of (Me3Si)2S, but product 4 was isolated instead. The structure of 4 was determined by 31P-{1H}, 1H and 13C-{1H} NMR, IR and P P S S O O Me Me S PA PB S S NH SiMe3 S S P P S S S N CH3 N H S H 9 7 9 2 1 3 6 12 10 4 5 8 11 13 14 15 16 7 4 Table 2 Selected bond lengths (Å) and angles (8) in compound 3 (diastereomer i) P(1)–S(19) P(1)–S(1) P(9)–S(9) P(1)–C(1) S(19)–Si(1) N(9)–C(11) P(1)–S(19)–P(9) S(1)–P(1)–S(19) S(9)–P(9)–S(19) P(1)–S(19)–Si(1) C(1)–P(1)–S(19) C(1)–P(1)–S(1) C(9)–P(9)–S(9) S(1)–P(1)–S(19) C(11)–N(9)–P(9) 2.084(1) 1.936(1) 1.940(1) 1.815(3) 2.187(1) 1.456(4) 102.69(4) 106.58(5) 104.76(5) 106.23(5) 105.85(9) 114.57(9) 114.55(9) 116.31(5) 120.1(2) P(9)–S(19) P(1)–S(19) P(9)–N(9) P(9)–C(9) N(9)–Si(9) S(19)–P(1)–S(19) N(9)–P(9)–S(19) P(9)–N(9)–Si(9) C(9)–P(9)–S(19) C(1)–P(1)–S(19) C(9)–P(9)–N(9) N(9)–P(9)–S(9) C(11)–N(9)–Si(9) 2.109(1) 2.074(1) 1.647(2) 1.815(3) 1.794(2) 110.59(5) 110.35(9) 123.0(1) 104.62(9) 102.53(9) 107.1(1) 115.00(9) 115.0(2) mass spectroscopy; its purity was assessed by elemental analysis.No signals due to SiMe3 groups were found in the 1H and 13C-{1H} NMR spectrum. Pyridine thus acted as a protontransfer catalyst; the protons are transferred from the solvent to 3, providing desilylated ionic compound 4 (which precipitated as a solid from the reaction mixture) and a side product, trimethylsilylated toluene. The 31P NMR spectrum of 4 in pyridine is an AB system [d 70.47 and 70.09, 2J(PP) = 12.0 Hz], whilst in dmso the diVerence in chemical shifts of both phosphorus atoms is so small, that only a singlet at d 67.1 is observed; we have no ready explanation for the solvent dependence of the AB spectrum.For comparison, the chemical shifts of phosphorus atoms in anion 7 13 are d 66.1 (PA) and 57.1 (PB), 2J(PP) = 13.6 Hz. The 1H and 13C NMR signals of the naphthalene entity of 4 are assigned on the basis of its analogy with anion 7.13 In the IR spectrum of 4 a broad unresolved multiplet of overlapping bands between 3700 and 2300 cm21 is present, which makes assignment of vibrations in this region diYcult. Compound 4 is very soluble in pyridine, dmso and dmf; it decomposes slowly in the latter two.We were unable to prepare single crystals of it suitable for X-ray analysis. To improve the solubility of compound 4 we substituted its pyridinium cation by a larger organic cation, tetraphenylphosphonium, to give 5 which is soluble in more organic solvents, e.g. in chlorinated alkanes. This feature allowed for easier puri- fication of 5 by crystallization, and also allowed us to obtain crystals suitable for structure analysis by diVusion of hexane into a diluted solution of 5 in dichloromethane.The 31P-{1H} NMR spectrum of the anionic part of 5 in dmso is an AB system [d 73.03 and 72.23, 2J(PP) = 11.6 Hz], whilst in CDCl3 or py it shows only a singlet, as a result of the very similar chemical shifts of its two obviously magnetically non-equivalent phosphorus atoms. The 1H NMR signals of the naphthalene entity were assigned on the basis of its analogy with anion 7.13 The crystal structure of compound 5 (Table 3, Fig. 2) reveals the close similarity of its anionic part to the structure of the anion 7, determined in the form of its hexamethyldisilazan-2- onium salt.13 The only significant diVerences between these two structures are in the environment of P(9) and N(9), due to change of the bulky SiMe3 group bonded to the nitrogen atom in 7 to a methyl group in 5. The C10H6P2 part of anion of 5 is significantly distorted from planar; atoms P(1) and P(9) lie 0.34 and 0.29 Å above and below the mean plane respectively, with additional small deviation for the sulfur atom S(1) which lies 0.36 Å below this plane.The C3P2S ring is hinged with the C3P2 and P2S planes inclined by 49.58 with respect to each other, corresponding to a value in 7 of 51.28. The transanular P ? ? ?P distance and internal P–S–P angle in 5 [3.24 Å, 100.21(4)8] are quite similar to the corresponding values in 7 [3.22 Å, 100.44(6)8]. The P(1)–S(1) [1.967(1) Å] and P(1)–S(19) [1.981(1)J.Chem. Soc., Dalton Trans., 1999, 2231–2236 2235 Å] distances indicate delocalization of the negative charge in the PS2 group; these distances are slightly, but significantly, longer than P(9)–S(9) [1.946(1) Å], due to their lower bond order. The hydrogen atom bonded to nitrogen N(9) is not involved in any intra- or inter-molecular hydrogen bonds. Use of hot acetonitrile as a solvent for the reaction of compound 1 with 2 gives rise to 6 containing in its cage structure the unsaturated CN2P2S heterocycle. We suppose that the first two reaction steps of formation of 6 are analogous to that in the reaction of trans-diorganodithiadiphosphetane disulfides (RPS2)2 (R = Me, Et or C6H4OMe-p) with 2 (Scheme 1).1 Thus the first reaction produs 3, which in hot acetonitrile undergoes a ring closure reaction giving compound 8 containing a four-membered heterocycle NP2S and (Me3Si)2S as a coproduct. Consequent nucleophilic attack of the acetonitrile nitrogen atom at a phosphorus centre results in formation of a new heterocycle CN2P2S and thus gives 6 as a final product (Scheme 3).Yellow compound 6 is indefinitely stable at room temperature and shows a good solubility in hot acetonitrile. However in cold acetonitrile it is much less soluble, which enabled us to isolate the compound from a crude mixture of products by crystallization. The infrared spectrum of 6 contains a characteristic band due to the n(C]] N) at 1582 cm21 as well as two n(P]] S) bands at 655 and 639 cm21.The crystal structure of compound 6 (Table 4, Fig. 3) reveals that the planarity of the C10H6P2 part of molecule is noticeably distorted; atoms P(1) and P(9) lie 0.20 and 0.15 Å above and below the mean plane respectively, with additional small deviation for the sulfur atom S(1), which lies 0.16 Å below this plane. Sulfur atoms S(9) and S(19) lie 1.02 and 1.16 Å below the mean plane. The CN2P2 part of the CN2P2S ring is almost planar [maximum deviation from mean plane 0.09 Å for N(9)]; the internal sulfur atom S(19) lies 1.09 Å below this plane, Fig. 2 Molecular diagram of the anionic part of compound 5; PPh4 1 cation omitted for clarity. Table 3 Selected bond lengths (Å) and angles (8) in compound 5 P(1)–S(19) P(1)–S(1) P(9)–S(9) P(1)–C(1) N(9)–C(11) P(1)–S(19)–P(9) S(1)–P(1)–S(19) S(9)–P(9)–S(19) C(1)–P(1)–S(19) C(1)–P(1)–S(1) C(9)–P(9)–S(9) S(1)–P(1)–S(19) 2.132(1) 1.967(1) 1.946(1) 1.819(3) 1.450(6) 100.21(4) 103.18(4) 107.04(5) 101.84(9) 112.19(9) 115.4(1) 118.79(5) P(9)–S(19) P(1)–S(19) P(9)–N(9) P(9)–C(9) P(9)–N(9)–C(11) S(19)–P(1)–S(19) N(9)–P(9)–S(19) C(9)–P(9)–S(19) C(1)–P(1)–S(19) C(9)–P(9)–N(9) N(9)–P(9)–S(9) 2.087(1) 1.981(1) 1.651(4) 1.804(3) 118.6(4) 111.30(5) 111.8(1) 107.67(9) 108.18(9) 102.6(2) 112.3(2) whilst C(12) and C(13) are coplanar with this plane.The planes fitted to CN2P2 and C10H6P2 atoms are inclined by 74.68 with respect to each other.The internal P–S–P angle in 6 [93.06(4)8] is as expected significantly reduced vs. that in cis-3 [102.69(4)8] and relative to that in the seven-membered C2O2P2S heterocycle 9 19 (about 97.18), whilst in comparison with that angle in the P2S2 ring of 12 [80.0(1)8] it is substantially enlarged. The N(1)– C(11) distance [1.284(3) Å] is reasonable for a formal C]] N double bond. The transannular P ? ? ? P distance is 3.02 Å. Molecules pack with naphthalene rings parallel to each other along the b axis.To our best knowledge, 6 is the first reported example of a six membered heterocycle containing the internal atom sequence PSPNCN; no example of it even in its saturated form has been previously reported, although several compounds containing related six membered rings with atom sequence POPNCN have been reported.20 Fig. 3 Molecular diagram of compound 6. Fig. 4 Heteronuclear interactions observed for major and minor isomers (dashed lines are for weak correlations).P P S S N CH3 S S P P S S N CH3 S S Si(CH3)3 (CH3)3Si (CH3)3Si Si(CH3)3 minor isomer major isomer Table 4 Selected bond lengths (Å) and angles (8) in compound 6 P(1)–S(19) P(1)–N(1) N(1)–C(11) P(1)–S(1) P(1)–C(1) C(11)–C(12) P(1)–S(19)–P(9) C(11)–N(1)–P(1) N(9)–P(9)–S(19) N(1)–P(1)–C(1) N(1)–P(1)–S(1) C(9)–P(9)–S(19) C(9)–P(9)–S(9) S(1)–P(1)–S(19) N(1)–C(11)–C(12) C(11)–N(9)–C(13) 2.105(1) 1.626(2) 1.284(3) 1.919(1) 1.793(3) 1.503(4) 93.06(4) 133.4(2) 106.11(9) 105.4(1) 114.48(9) 106.69(9) 115.2(1) 109.57(5) 116.3(3) 119.4(3) P(9)–S(19) P(9)–N(9) N(9)–C(11) P(9)–S(9) P(9)–C(9) N(9)–C(13) N(1)–C(11)–N(9) C(11)–N(9)–P(9) N(1)–P(1)–S(19) N(9)–P(9)–C(9) N(9)–P(9)–S(9) C(1)–P(1)–S(19) C(1)–P(1)–S(1) S(9)–P(9)–S(19) N(9)–C(11)–C(12) P(9)–N(9)–C(13) 2.060(1) 1.714(2) 1.365(4) 1.918(1) 1.809(3) 1.470(4) 126.2(2) 122.3(2) 106.95(9) 105.1(1) 112.91(9) 104.13(9) 115.5(1) 110.21(5) 117.5(3) 117.9(3)2236 J.Chem. Soc., Dalton Trans., 1999, 2231–2236 Acknowledgements P.K. is grateful to the Royal Society of Chemistry for a RS/ NATO fellowship tenable in Loughborough University. This work was supported by grant No. VS 96 095 of the Department of Education of the Czech Republic and by grants No. 203/95/ 1190, 203/96/0111, 203/96/1513, 203/97/0955 and 203/99/0067 of the Grant Agency of the Czech Republic. We are grateful to JREI (Joint Research Equipment Initiative) for an equipment grant. References 1 W.Zeis, H. Henjes, D. Lux, W. Schwarz and H. Hess, Z. Naturforsch., Teil B, 1979, 34, 1334. 2 A. M. Z. Slawin, D. J. Williams, P. T. Wood and J. D. Woollins, J. Chem. Soc., Chem. Commun., 1987, 1741. 3 M. R. St. J. Foreman, J. Novosad, A. M. Z. Slawin and J. D. Woollins, J. Chem. Soc., Dalton Trans., 1997, 1347. 4 N. Walker and D. Stuart, Acta Crystallogr., Sect. A, 1983, 39, 158. 5 KM4RED, data reduction program, KUMA DiVraction, Wroclaw, 1998. 6 G. M. Sheldrick, SHELXS 86, University of Göttingen, 1986. 7 G. M. Sheldrick, SHELXL 93, University of Göttingen, 1993. 8 A. L. Davis, E. D. Laue, J. Keeler, D. Moskau and J. Lohman, J. Magn. Reson., 1991, 94, 637. 9 J. Jeener, B. H. Meier, P. Bachmann and R. R. Ernst, J. Chem. Phys., 1979, 71, 4546. 10 A. Bax and M. F. Summers, J. Am. Chem. Soc., 1986, 108, 2093. 11 R. Marek, L. Králík and V. Sklenár, Tetrahedron Lett., 1997, 38, 665. 12 E. Pelaez-Arango, F. J. Garcia-Alonso, G. Carriedo and F. Lopez- Ortiz, J. Magn. Reson., Sect. A, 1996, 121, 154. 13 P. Kilián, J. Marek, R. Marek, J. Touzín, O. Humpa, J. Novosad and J. D. Woollins, J. Chem. Soc., Dalton Trans., 1998, 1175. 14 R. Marek, O. Humpa and V. Sklenár, unpublished results. 15 M. J. S. Dewar, E. G. Zoebisch, E. F. Healy and J. J. P. Stewart, J. Am. Chem. Soc., 1985, 107, 3902. 16 H. W. Roesky and G. Remmers, Z. Anorg. Allg. Chem., 1977, 431, 221. 17 M.-E. Eleftheriou, J. Novosad, D. J. Williams and J. D. Woollins, J. Chem. Soc., Chem. Commun., 1991, 116. 18 P. Kilián, J. Touzín, J. Marek, J. D. Woollins and J. Novosad, Main Group Chem., 1996, 1, 425. 19 M. R. St. J. Foreman, A. M. Z. Slawin and J. D. Woollins, J. Chem. Soc., Chem. Commun., 1995, 2217. 20 J. Breker and R. Schmutzler, Chem. Ber., 1990, 123, 1307; W. S. Sheldrick, S. Pohl, H. Zamankhan, M. Banek and D. Amirzadeh- Asl, Chem. Ber., 1981, 114, 2132; H. W. Roesky, K. Ambrosius, M. Banek and W. S. Sheldrick, Chem. Ber., 1980, 113, 1847; D. W. U. Schomburg and R. Schmutzler, Z. Naturforsch., Teil B, 1986, 41, 207. Paper 9/01698H

ISSN:1477-9226    

DOI:10.1039/a901698h

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1998, Pages 2233–2239 2233 Dichloro(1,4,8,11-tetraazacyclotetradecane)manganese(III) chloride: cis–trans isomerisation evidenced by infrared and electrochemical studies Fabien Létumier,a Grégory Broeker,a Jean-Michel Barbe,a Roger Guilard,*,†,a Dominique Lucas,b Valérie Dahaoui-Gindrey,c Claude Lecomte,*,c Laurent Thouin d and Christian Amatore *,d a Laboratoire d’Ingénierie Moléculaire pour la Séparation et les Applications des Gaz (LIMSAG, UMR 5633), Faculté des Sciences ‘Gabriel’, 6, Bd Gabriel, 21100 Dijon, France b Laboratoire de Synthèse et d’Electrosynthèse Organométallique (LSEO, UMR 5632), Faculté des Sciences ‘Gabriel’, 6, Bd Gabriel, 21100 Dijon, France c Laboratoire de Cristallographie et Modélisation des Matériaux Minéraux et Biologiques (LCM3B), UPRESA 7036, Université Henri Poincaré, Nancy 1, Faculté des Sciences, B.P. 239, 54506 Vandoeuvre lès Nancy Cedex, France d Ecole Normale Supérieure, Département de Chimie, URA CNRS 1679, 24, rue Lhomond, 75231 Paris, Cedex 05, France The isomers cis- and trans-dichloro(1,4,8,11-tetraazacyclotetradecane)manganese(III) chloride have been synthesized and characterised.An isomerisation of the type cis-MnIII æÆ trans-MnIII was observed by IR spectroscopy. Recrystallisation of the cis species gave systematically crystals of the trans derivative. The crystal structure of the latter complex has been determined. An electrochemical study of the trans-manganese(III) complex in dimethyl sulfoxide revealed a rare trans æÆ cis isomerisation reaction which proceeds through an electrochemical step–chemical step mechanism. The overall first-order rate constant, kiso, for this isomerisation was determined by simulation of the chronoamperometric data.Supplementary electrochemical information for the trans-manganese(III) isomer was obtained through simulation of the cyclic voltammograms over a large range of scan rates.No influence of chloride ion concentration was observed as determined by cyclic voltammetry using an internal chloride standard, [N(PPh3)2]Cl. The cyclic voltammogram of the cis-dichloro(1,4,8,11-tetraazacyclotetradecane) manganese(II) generated in situ displays the usual cis æÆ trans isomerisation which also occurs via an electrochemical step–chemical step mechanism. Manganese is involved in many biological processes, in particular it is the active site of several enzymes.1–3 In order to mimic these enzymes, many manganese polyazamacrocyclic complexes have been synthesized and studied in connection with oxidation state,4 co-ordination scheme 5 and number of manganese sites present in these biological catalysts.6,7 As an example, manganese pentaazamacrocyclic complexes have been prepared and characterised in eVorts to determine the mechanism of action of superoxide dismutase.8 Furthermore, parallel syntheses of porphyrin and triazamacrocycle manganese complexes have been carried out and these complexes have been shown to be catalytically active in oxidation reactions.9–14 In these latter derivatives the oxidation state of the metal centre is II, III, or IV, their structure varying from monomer up to tetramer.15–17 Two types of manganese cyclam complexes have been found to be important in catalytic oxidation reactions: a mixed-valence dinuclear complex of the type MnIII]MnIV 18–20 and a transdichlorocyclam manganese(III) species.21 However, the active form of the 1,4,8,11-tetraazacyclotetradecane (cyclam) manganese complex involved in these oxidation reactions remains unknown.22,23 In the present work, we have synthesized and characterised a third type of cyclam manganese complex, the cis-dichlorocyclam manganese(III) chloride.In addition, a cis æÆ trans isomerisation reaction has been observed and monitored by IR spectroscopy. Complementary information has been provided through electrochemical studies where both cis æÆ trans 24,25 and, a more rarely observed, trans æÆ cis isomerisation reactions 26,27 occur as the precursors are oxidised or reduced respectively.EVorts have been made to crystallise † E-mail: rguilard@u-bourgogne.fr the cis complex, but systematically crystals of the trans isomer have been produced via the above mentioned cis æÆ trans isomerisation reaction. The crystal structure of this latter derivative is reported. Experimental Instrumentation Cyclic voltammetry experiments were performed using a model 273A EG&G potentiostat controlled by a personal computer using the EG&G 270/250 Research Electrochemistry Software (version 4.23).These experiments were carried out in a threeelectrode cell using a glassy carbon disc (r = 1.5 mm) as the working electrode, a platinum spiral as the counter electrode, and a saturated calomel electrode (SCE) as the reference. In experiments at higher scan rate the working electrode was a platinum disc (r = 0.025 mm).The cyclic voltammetry was simulated using the DIGISIM program (version 2.1) distributed by the Bioanalytical Systems Corporation.28 For the coulometric measurements a IG6-N Tacussel instrument was used. The potentials were held constant with a model 362 EG&G potentiostat. In the electrolyses a carbon tissue was used as the working electrode, and the counter electrode was either a platinum spiral or magnesium ribbon. The counter electrode compartment was separated from the working electrode with a porous glass sintered disc.Experiments using a rotating disc electrode (RDE) were accomplished with a gold electrode (r = 1 mm, Tacussel). Rotation speeds of 900 revolutions min21 yielded stationary-state voltammograms. Double-step chronoamperometry was performed using the same electrodes as in the experiments at the elevated scan rates. Three diVerent concen-2234 J. Chem. Soc., Dalton Trans., 1998, Pages 2233–2239 trations were analysed: 0.5, 1.0 and 2.0 mM over time intervals of 5 to 200 ms.The programmed potential ramp began and ended approximately 200 mV before and after the peak of the corresponding reduction wave in the cyclic voltammetry for the equivalent time interval. All potentials are referred to the SCE; electrochemical experiments were conducted at room temperature. The IR spectra were recorded on a Bruker IFS 66v FTIR spectrometer and samples were prepared as 1% dispersions in KBr pellets.Chemicals The cyclam ligand was prepared in our laboratory following literature methods.29,30 For all syntheses, chemicals were commercially available and used as received without further purifi- cation. For the electrochemical studies, dimethyl sulfoxide (dmso) was dried over activated 3 Å molecular sieves. Bis- (triphenylphosphoranylidene)ammonium chloride [N(PPh3)2]- Cl was from Aldrich. The electrolyte support, tetra-n-butylammonium hexafluorophosphate (Fluka), was used as received. Di-Ï-oxo-bis(1,4,8,11-tetraazacyclotetradecane)dimanganese( III,IV) perchlorate. This compound is formed during the preparation of the cis-[Mn(cyclam)Cl2]Cl (see below).It was synthesized using a slightly diVerent procedure than that described.19,20 A solution of Mn(ClO4)2?6H2O (1.8 g, 5 mmol) in methanol (50 cm3) and water (10 cm3) was added dropwise to cyclam (1 g, 5 mmol) in methanol (20 cm3). The resultant mixture turned olive green.After stirring for 1 h, it was filtered and a green precipitate recovered after several washings with cold methanol. The precipitate was then air-dried. CAUTION: care should be taken in isolation of the solid product; transition metal perchlorates are potentially explosive and must be prepared in small amounts. Yield 1.43 g (33%) (Found: C, 27.7; H, 5.8; N, 12.9. Calc. for C20H48Cl3Mn2N8O14?H2O: C, 27.9; H, 5.8; N, 13.0%). trans-Dichloro(1,4,8,11-tetraazacyclotetradecane)manganese( III) chloride.A solution of Mn(CH3CO2)3?2H2O (6.7 g, 25 mmol) in methanol (200 cm3) was added to cyclam (5 g, 25 mmol) dissolved in methanol (50 cm3). The reaction mixture was then stirred for 3 h at room temperature. Concentrated HCl (1 cm3) was then added to allow the formation of a clear green precipitate which was filtered oV and recrystallised from water (7.33 g, 65%) (Found: C, 26.8; H, 7.6; N, 12.3. Calc. for C10H24Cl3MnN4?5H2O: C, 26.6; H, 7.6; N, 12.4%). trans- and cis-Dichloro(1,4,8,11-tetraazacyclotetradecane)- manganese(III) chloride (one-pot synthesis).A solution of MnCl2?4H2O (1.98 g, 10 mmol) in methanol (100 cm3) was added to a solution of cyclam (2 g, 10 mmol) in methanol (25 cm3). The resultant solution was olive green. After evaporation of three-fourths of the solvent, a clear green precipitate was collected and, after filtration, identified as the transmanganese( III) complex previously described (see above).Treatment of the filtrate [containing the mixed-valence m-oxodimanganese( III, IV) species] by concentrated HCl (0.4 cm3, 1 equivalent) led to the formation of a red precipitate. After filtration and washing several times with methanol, the cis-dichloro(1,4,8,11-tetraazacyclotetradecane)manganese(III) chloride was obtained. (1.3 g, 35%) (Found: C, 33.0; H, 6.7; N, 14.9. Calc. for C10H24Cl3MnN4?0.33CH3OH: C, 33.3; H, 6.8; N, 15.0%). Crystallography Crystal data for trans-[Mn(cyclam)Cl2]Cl?5H2O.C10H24Cl3- MnN4?5H2O, M = 451.7, monoclinic, space group P21/n, a = 9.876(1), b = 6.501(1), c = 16.651(2) Å, b = 107.39(1)8, U = 1020.3(2) Å3, room temperature, Z = 2, m(Mo-Ka) = 1.09 mm21, 4421 reflections measured {including substructure, see below, [sin q/l]max = 0.6 Å21}, 2105 structure reflections of which 1479 [I > 3s(I), Rint = 0.005] were used for the refinement. Final R(F) = 0.036 and R9(F) = 0.042. A first quick and short data collection had shown that this manganese(III) complex crystallises in the monoclinic system with lattice type C (systematic absences: hkl, h 1 k � 2n).A total of 4421 reflections was collected. During the data treatment it became evident that reflections with k odd (or h odd) were much weaker than those with k even (or h even), and furthermore that the reflections with k odd may not be averaged according to 2/m symmetry, indicating the existence of a substructure of lower symmetry. Then, in order to find the average structure, we only used the reflections with k (or h) even, the new unit cell being four times smaller than the old one.This transformation led to systematic absences (h0l, h 1 l � 2n); thus this manganese complex crystallises in the monoclinic system, space group P21/n. Lorentz-polarisation corrections, intensity scaling (2% decay), and data reduction were carried out using the DREAR package.31 No absorption correction was applied. The structure was solved by Patterson methods and successive Fourier synthesis, then refined by full-matrix least squares.32 After location of all non-H atoms, diVerence electron density maps revealed three residual peaks of approximately the same weight in the asymmetric unit.Two have been attributed to oxygen atoms of water solvate molecules [w(1) and w(2)]. The third peak corresponds to a site occupied by a chloride ion [Cl(2)] and by an oxygen atom of a water molecule [w(3)], each with an occupancy factor of 50%.The positions of the Cl(2) and w(3) atoms have been alternatively refined. The two atoms are separated by 0.552(6) Å. This disorder refined in the P21/n space group explains the weak substructure reflections due to the solvent and counter-ion structure. All attempts to solve the substructure failed. The refined parameters included anisotropic mean-square displacements for the non-H atoms, and positions for the macrocycle H atoms, isotropic mean-square displacements being B(H) = 1.3 B(X), where B(X) is the equivalent isotropic mean-square displacement for atom X to which the H atom is covalently bound. The ORTEP33 program was used to draw the crystal structure views (Figs. 1 and 2). Selected bond distances and angles are reported in Table 1, and the geometry of the intra- and inter-molecular hydrogen bonds in Table 2. CCDC reference number 186/990. Results and Discussion Synthesis The complexes were obtained starting from stoichiometric amounts of cyclam and manganese dichloride.Under these experimental conditions, the mixed-valence dinuclear complex MnIII]MnIV and the trans-dichlorocyclam manganese(III) complex are formed simultaneously at room temperature. We observed that hydrolysis of the mixed-valence dinuclear complex, in methanol, by addition of concentrated HCl led to the formation of cis-[Mn(cyclam)Cl2]Cl (see below). The reaction is visibly apparent, as the initial solution of MnIII]MnIV is olive green, and upon addition of HCl the solution gradually becomes red.Interestingly, the reaction with HCl seems to break the m-oxo bridges but maintains the cis arrangement. In aqueous medium the cis isomer rapidly and irreversibly converts into the trans species at room temperature as has been previously observed for cobalt(III) complexes.34,35 However, the present reaction is diVerent from that reported for corresponding iron(III) complexes where the cis æÆ trans conversion requires heating.36 Crystal structure of trans-dichloro(1,4,8,11-tetraazacyclotetradecane) manganese(III) chloride An ORTEP view of the crystal structure of the trans-J.Chem. Soc., Dalton Trans., 1998, Pages 2233–2239 2235 [Mn(cyclam)Cl2]1 complex is given in Fig. 1 with the numbering scheme used. The manganese ion lies on one 1� inversion centre. The structure consists of centrosymmetric [Mn- (cyclam)Cl2]1 cations which are linked by hydrogen bonds to the chloride anions and to the water molecules.The geometry at manganese is trans-pseudo-octahedral with the four nitrogen atoms of the ligand in equatorial position [average Mn]N 2.033(3) Å], and the two chloride ions in axial positions with a very long Mn]Cl bond length [2.5269(7) Å]. This bond distance is much longer than that found in the similar complex trans- [Co(cyclam)Cl2]Cl?4H2O?0.47HCl37 where the distance Co]Cl is equal to 2.2524(6) Å. The Mn]N and Mn]Cl bond lengths are very similar to those of the trans-[Mn(cyclam)Cl2]NO3 complex,21 as are the cis N]Mn]N bond angles of 85.5(1) and 94.5(1)8, the smaller value being associated with the fivemembered ring [Mn, N(1), C(1), C(2), N(2)] as expected.In the same way, the bond distances and angles in the cyclam ligand are thoroughly consistent with those found in the literature.21,37–39 The metal ion is located in the plane of the four nitrogen atoms. Hydrogen atoms bonded to N(19) and N(2) are above the four-nitrogen plane, whereas the hydrogen atoms bonded to N(1) and N(29) are below, giving rise to the expected trans-III conformation according to the nomenclature of Bosnich et al.40 Based on the N]H bond direction relative to the four-nitrogen atom plane, five energetically distinct nonenantiomeric cis and trans isomers, denoted I to V, are conceptually possible since each co-ordinated nitrogen atoms is chiral.In the trans-III isomer the five-membered chelate rings adopt twist conformations, and the six-membered chelate rings are in chair conformations.Fig. 2 shows the crystal packing of trans-[Mn(cyclam)Cl2]Cl?5H2O as a projection in the (Æb , Æc ) plane, and the geometry of the intra- and inter-molecular hydrogen bonds is given in Table 2. The HN(2) hydrogen atom participates in a three-centre hydrogen bond,41 involving the N2]HN(2) bond and two hydrogen bonds with the Cl(1) Fig. 1 An ORTEP view of the molecular structure of the trans- [Mn(cyclam)Cl2]Cl complex with 50% probability thermal ellipsoids for non-H atoms.Only hydrogen atoms bonded to the nitrogen atoms are shown. Atoms obtained by inversion [0, 0, 0] have primed labels Table 1 Selected bond distances (Å) and angles (8) with estimated standard deviations (e.s.d.s) in parentheses for trans-[Mn(cyclam)- Cl2]Cl?5H2O Mn]N(1) Mn]N(2) N(1)]Mn]N(2) N(1)]Mn]Cl(1) N(2)]Mn]Cl(1) N(1)]Mn]N(2I) 2.036(3) 2.031(2) 85.5(1) 88.74(7) 88.26(7) 94.5(1) Mn]Cl(1) N(1)]Mn]Cl(1I) N(2)]Mn]Cl(1I) Cl(1)]Mn]Cl(1I) 2.5269(7) 91.26(7) 91.74(7) 180 Symmetry code: I 2x, 2y, 2z.chloride ion eptor, one being an intramolecular bond and the other an intermolecular bond. That three-centre hydrogen bond gives rise to infinite chains of the [Mn(cyclam)] complex which extend along the Æb axis direction. The sum of the angles involving HN(2) as central atom is exactly 3608. A Mn]N(2)]NH(2) ? ? ? Cl(1)]Mn interaction has also been found in the trans-[Mn(cyclam)Cl2]NO3 complex,21 which also crystallises in the space group P21/n, but in that case the chain structure runs parallel to the Æa axis, the Mn ? ? ? Mn separation (i.e. the Æa axis length) being 6.547(2) Å compared to 6.501(1) Å for the Æb axis length in our case.These infinite chains are linked to one another via the Cl(2) chloride ion and the water molecules. The HN(1) hydrogen atom is also involved in a three-centre hydrogen bond, on the one hand from an intramolecular contact with the Cl(1) chloride ion and on the other from an intermolecular contact with the Cl(2) ion and the w(3) water molecule.The sum of the characteristic angles of the three-centre hydrogen bond is 360 and 3578, for w(3) and Cl(2) as acceptor respectively. The intermolecular contact of 2.875(7) Å between w(1) and w(3) seems to correspond to an hydrogen bond where w(3) would be the donor. Characterisation of the isomerisation reaction by IR spectroscopy The cis-MnIII æÆ trans-MnIII isomerisation reaction can be monitored by IR spectroscopy in the region 800–900 cm21 where the N]H and CH2 vibrations of the cyclam moiety are sensitive to the geometric nature of the ligand co-ordinated to the metal.The cis complexes have a distinct fingerprint pattern which is practically insensitive to the nature of the metal. Previously reported cis complexes of the type MIII(cyclam)Cl2 exhibit two CH2 (794–824 cm21) and three to four N]H absorption bands (841–926 cm21) (see Table 3).35,36 In agreement with those studies, the cis-[Mn(cyclam)Cl2]Cl complex presents also two bands in the region 790–830 cm21 and two broad and intense bands in the region 840–890 cm21. Another band of weaker intensity is noted at 923 cm21.In a short period of time the cis isomer dispersed in a KBr pellet transformed into the trans complex. We attribute this isomerisation to trace amounts of water contained in the KBr salt. This isomerisation reaction is demonstrated through IR spectroscopy by the disappearance of the four bands at 923, 858, 844 and 807 cm21 and the concomitant appearance of one intense band at 880 cm21.Conversely, the band at 795 cm21 is unaVected. These two remaining bands are characteristic of the trans isomer (see Table 3). The evolution of the IR spectrum during the cis æÆ trans isomerisation reaction is depicted in Fig. 2 An ORTEP view of the crystal packing in trans-[Mn- (cyclam)Cl2]Cl?5H2O as a projection in the ( Æb , Æc ) plane, with 25% probability thermal ellipsoids for non-H atoms.Hydrogen bonds are indicated by single thin lines2236 J. Chem. Soc., Dalton Trans., 1998, Pages 2233–2239 Table 2 Geometry of the intra- and inter-molecular hydrogen bonds (distances in Å, angles in 8) in the trans-[Mn(cyclam)Cl2]Cl?5H2O complex Intramolecular hydrogen bonds Cl(1I) ? ? ? N(2) Cl(1) ? ? ? N(1) 3.193(2) 3.210(3) N(2)]HN(2) N(1)]HN(1) 0.86(4) 0.76(4) HN(2) ? ? ? Cl(1I) HN(1) ? ? ? Cl(1) 2.72(3) 2.83(4) N(2)]HN(2) ? ? ? Cl(1I) N(1)]HN(1) ? ? ? Cl(1) 116(3) 114(3) Intermolecular hydrogen bonds w(1) ? ? ? w(2) w(1) ? ? ? w(3II) w(1) ? ? ? Cl(2II) w(1) ? ? ? w(3) w(1) ? ? ? Cl(2) w(2) ? ? ? Cl(2III) w(2) ? ? ? w(3III) N(1) ? ? ? w(3I) N(1) ? ? ? Cl(2I) N(2) ? ? ? Cl(1IV) 2.783(6) 2.773(6) 3.231(4) 2.875(7) 3.024(5) 2.859(5) 3.174(6) 3.050(6) 3.238(3) 3.233(3) w(1)]Hw(12) w(1)]Hw(11) w(2)]Hw(21) N(1)]HN(1) N2]HN(2) 0.780(4) 1.058(4) 0.764(5) 0.76(4) 0.86(4) Hw(12) ? ? ? w(2) Hw(11) ? ? ? w(3II) Hw(11) ? ? ? Cl(2II) Hw(21) ? ? ? Cl(2III) Hw(21) ? ? ? w(3III) HN(1) ? ? ? w(3I) HN(1) ? ? ? Cl(2I) HN(2) ? ? ? Cl(1IV) 2.015(5) 1.717(5) 2.174(2) 2.240(2) 2.632(5) 2.41(4) 2.56(4) 2.49(4) w(1)]Hw(12) ? ? ? w(2) w(1)]Hw(11) ? ? ? w(3II) w(1)]Hw(11) ? ? ? Cl(2II) w(2)]Hw(21) ? ? ? Cl(2III) w(2)]Hw(21) ? ? ? w(3III) N(1)]HN(1) ? ? ? w(3I) N(1)]HN(1) ? ? ? Cl(2I) N(2)]HN(2) ? ? ? Cl(1IV) 168.4(4) 175.1(3) 175.7(2) 138.8(3) 129.6(3) 143(4) 150(4) 146(3) Symmetry codes: I 2x, 2y, 2z; II 0.5 2x, 20.5 1y, 20.5 2z; III x, 21 1y, z; IV x, 1 1y, z.Table 3 Infrared absorption bands in the 800–900 cm21 region of selected cis- and trans-[M(cyclam)Cl2]X complexes Geometry cis cis cis cis trans trans trans trans M Cr Mn Fe Co Cr Mn Fe Co X Cl Cl Cl Cl Cl Cl ClO4 Cl N]H vibration/cm21 872m 862m (sh) 854m 923m 858s 844s 866s 858m 850s 890w 872s 859s 841w 890s 882s 880s 890s 888s (sh) 906s 888s CH2 vibration/cm21 815w 805m 807w (sh) 795m 808m 794s 824w 808s 804s 795m 810m 818s Ref. 35 * 36 36 35 * 36 36 * This work. Fig. 3. This decrease in band number is consistent with the higher symmetry of the trans compared to the cis isomer. Electrochemistry Evidence of the reductive isomerisation trans-MnIII æÆ cis-MnII has been observed by cyclic voltammetry experiments starting from the trans-[MnIII(cyclam)Cl2]Cl complex. As shown in Fig. 4(a), the first positive scan displays only a capacitive current (no electrochemically active species being present and thus no oxidation to MnIV takes place up to 0.6 V vs.SCE). However, after the chemically irreversible reduction of the trans-manganese(III) isomer, a reversible oxidation wave appears at E8c = 0.3 V and a parallel decrease in cathodic current is noted during the second cycle of the voltammogram. This diminution in cathodic current is consistent with the formation of the cis-manganese(II) isomer produced and characterised by its reversible oxidation wave signature during the first cycle.However, the irreversible reduction of the transmanganese( III) complex becomes electrochemically quasi- Fig. 3 Evolution of IR spectra of cis- to trans-[Mn(cyclam)Cl2]Cl reversible at scan rates approaching 20 V s21. Exclusive formation of the cis-manganese(II) complex was observed during bulk reductive electrolysis of the trans-manganese(III) isomer. Furthermore, the cathodic stationary current for the reduction of the trans-[MnIII(cyclam)Cl2]Cl complex before the coulometry experiment and the resulting anodic stationary current Fig. 4 Cyclic voltammograms of (a) trans-[Mn(cyclam)Cl2]Cl and (b) cis-[Mn(cyclam)Cl2] in dmso containing 0.1 M NBun 4PF6. [Complex] = 2 mM, scan rate = 100 mV s21, glassy carbon working electrode (r = 1.5 mm)J. Chem. Soc., Dalton Trans., 1998, Pages 2233–2239 2237 for the cis-[MnII(cyclam)Cl2] complex produced are equal in the RDE experiments, indicative of a quantitative conversion.Since the chemical trans-MnIII æÆ cis-MnII reductive isomerisation was of particular interest, double-step chronoamperometry experiments were performed on solutions of the trans- [MnIII(cyclam)Cl2]Cl complex. The normalised current ratios were plotted as a function of log kisot for a mechanism where one molecule of trans-manganese(II) species isomerises to the cis complex. A good agreement between the simulated normalised current ratio Rnorm for the electrochemical step–chemical step mechanism (Rnorm vs.log kisot) and the experimental data was obtained over a concentration range from 0.5 to 2.0 mM. Similarly, neither addition of free chloride ions {40 equivalents [N(PPh3)2]Cl nor deliberately added water (150 equivalents) aVects the variation of Rnorm. Therefore, the overall kinetics of the chemical isomerisation is not bimolecular and does not involve chloride ions or water molecules. The results of the chronoamperometric analysis are depicted on Fig. 5.The rate constant, kiso, is equal to 5.8 ± 0.6 s21. Cyclic voltammetry was also performed upon the cis-[MnII- (cyclam)Cl2] generated in situ by addition of cyclam to a solution of MnCl2?4H2O in dmso. The electrochemical oxidation potential (E2� 1 ) of this reduced cis complex is identical to the reduction potential of the cis-manganese(III) complex the preparation of which is described herein [Fig. 4(b)].pon oxidation of this cis-manganese(II) complex the resulting cis-manganese( III) complex undergoes a chemical isomerisation reaction to form a small quantity of the corresponding trans isomer during a voltammetric scan at scan rate of 100 mV s21.Simulation of the corresponding set of voltammograms allows one to evaluate the corresponding rate constant as ca. k2iso = 0.1 s21. However, the one-electron oxidation of cis-[MnII(cyclam)Cl2] is electrochemically reversible and becomes chemically reversible when scan rates approach 1 V s21 in the cyclic voltammetry [Fig. 6(a)]. It must be emphasised that this cis- MnIII æÆ trans]MnIII isomerisation between the oxidised states does not interfere at all with our above kinetic determinations (i.e. that relative to the trans-MnII æÆ cis]MnII isomerisation occurring between the reduced states) because of the design of the experiments and because of the largely diVerent half-lives of trans-MnII (kiso = 5.8 s21) and cis-MnIII (k2iso = 0.1 s21).Fig. 5 Variation in the normalised current ratios Rnorm with the parameter log kisot for the electrochemical step–chemical step mechanism in double-step chronoamperometry. The initial potential was imposed at 10.1 V changed to 20.4 V for a time t and returned to 10.1 V. Comparison to the theoretical curve for various concentrations of trans-[Mn(cyclam)Cl2]Cl : 0.5 (n), 1.0 (s) and 2.0 mM (d) Simulation of the cyclic voltammograms was achieved using the DIGISIM program.28 From the cyclic voltammetry data of cis-[MnII(cyclam)Cl2] generated in situ, the E8c (cis- MnIII æÆ cis-MnII) = 0.305 V was calculated over several scan rates (100, 500, and 700 mV s21).Other electrochemical parameters of interest such as the transfer coeYcient (a), the heterogeneous rate constant (ks), and the diVusion coeYcient (D) were also optimised over these scan rates (see Table 4). The correspondence between the simulation and the experimental voltammograms is shown in Fig. 6(a). Taking into account these results for the cis-manganese(II) complex and the rate constant, kiso, given by the chronoamperometry experiments, the voltammograms of trans-[MnIII(cyclam)Cl2]Cl were also simulated over a large range of scan rates (0.5, 1.3 and 5.1 V s21). The good agreement between the experimental data and simulations led to the determination of all electrochemical parameters (see Table 4), among which E8t (trans-MnIII æÆ trans-MnII) = 20.090 V was calculated.Fig 6(b) illustrates the good agreement between simulated and experimental voltammograms. This trans-MnII æÆ cis-MnII isomerisation is of particular interest since previous studies performed with the iron analogue, trans-[FeIII(cyclam)Cl2]Cl, in dmso, do not reveal such a transformation.42 Moreover, only the isomerisation cis- FeIII æÆ trans-FeIII is noted in aqueous media; little transformation is observed in dmso solution even when heating up to 90 8C.42 In aqueous media cis-[MnIII(cyclam)Cl2]Cl also isomerises to form the trans complex but the reverse reaction is not Fig. 6 Experimental and simulated cyclic voltammograms for (a) a 2 mM solution of MnCl2?4H2O 1 cyclam at 500 mV s21 and (b) a 2 mM solution of trans-[Mn(cyclam)Cl2]Cl at 1.28 V s21. Experimental data (—); Simulation (s) according to the set of parameters reported in Table 4 and kiso = 5.8 s21, k2iso = 0.1 s21 (see text) Table 4 Electrochemical data for cis- and trans-[MnIII(cyclam)Cl2]Cl obtained through simulation of their cyclic voltammograms Parameter Standard potential E8/V vs.SCE Heterogeneous rate constant, ks/cm s21 DiVusion coeYcient, D/cm2 s21 Transfer coeYcient a cis-[Mn(cyclam)- Cl2]Cl 0.305 2 0.004 9.9 × 1027 0.5 trans-[Mn(cyclam)- Cl2]Cl 0.090 0.002 1.1 × 1026 0.52238 J. Chem. Soc., Dalton Trans., 1998, Pages 2233–2239 observed. The equilibrium constant ratio Kiso :K9iso [where Kiso is the equilibrium constant cis : trans at the manganese(II) level and K9iso that at the manganese(III) one] can be determined by considering the thermodynamic cycle in Scheme 1.It has to be noted that for the nickel cyclam derivative a similar cycle has been proposed upon oxidation of NiII to NiIII.43 Based on the experimental values of E8t and E8c determined above, one obtains Kiso :K9iso < 5 × 106. Such a value explains why the trans to cis isomerisation is irreversible at the manganese(II) level while it is irreversible in the cis to trans direction at the manganese( III) stage.The diVerent rate constants for MnII and MnIII are typical, as manganese(II) complexes give much faster ligand exchange rates than manganese(III) species.44 trans-MnIIICl2 2FE8t trans-MnIICl2 2RTln K9iso 2RTln Kiso cis-MnIIICl2 1FE8c cis-MnIICl2 Scheme 1 Kiso/K9iso = exp[(F/RT)(E8c 2 E8t)] Furthermore, no kinetically controlled predissociation of the chloride ligand from trans-[MnIII(cyclam)Cl2]Cl or cis- [MnII(cyclam)Cl2] isomers occurs as determined by cyclic voltammetry using addition of diVerent excesses of an internal chloride standard [N(PPh3)2]Cl.Despite this lack of any kinetic signature of a chloride ion dissociation/recombination, such a reaction may well occur whenever the dissociation is fast and reversible, equations (1) and (2). Indeed, considering a steady trans-MnIICl2 k1 k21 MnIICl1 1 Cl2 (1) Cl2 1 MnIICl1 k2 k22 cis-MnIICl2 (2) state kinetic behaviour of the transient MnIICl1 intermediate, the overall kinetic behaviour of the sequence (1) 1 (2) is equivalent to that of a direct reversible reaction (3) with kiso = k1k2/ trans-MnIICl2 kiso k-iso cis-MnIICl2 (3) (k21 1 k2) and k2iso = k21k22/(k21 1 k2).Thus, despite the real mechanism involving most likely dissociation and recombination of the Cl2 ligand, this is not apparent kinetically and no eVect of added chloride ions can be observed. The same situation obviously may occur for the reverse overall process when an authentic solution of cis-[MnII(cyclam)Cl2] is oxidised [Fig. 4(b)]. Conclusion A novel synthesis of a new type of cyclam manganese complex, cis-dichlorocyclammanganese(III) chloride, has been described. The cis and trans isomers could be conveniently distinguished by IR spectroscopy. The cis-MnIII æÆ trans-MnIII isomerisation reaction is enhanced by the presence of water as was previously reported for the corresponding iron and cobalt compounds36 as deduced by IR spectroscopy.The two overall isomerisation reactions were also characterised by cyclic voltammetry experiments: trans-MnIII æÆ cis-MnII and cis- MnII æÆ trans-MnIII. No kinetically limiting predissociation of the chloride ligand from trans-[MnIII(cyclam)Cl2]Cl or cis- [MnII(cyclam)Cl2] isomers is associated with these electrochemical step–chemical step mechanisms occurs as determined in experiments with an internal chloride standard, [N- (PPh3)2]Cl. The overall unimolecular trans-MnIII æÆ cis-MnII isomerisation reaction was characterised by double-step chronoamperometry and modelling of its cyclic voltammogram was made over a large timescale.The rate constant, kiso, is equal to 5.8 ± 0.6 s21. No such electrochemically induced trans-MIII æÆ cis-MII in dmso has been noted for iron or cobalt complexes of the type trans-[MIII(cyclam)Cl2]Cl. Thus, our current research eVorts are focussed upon how the size of the metal and the macrocyclic cavity can aVect this reaction (k2iso = 0.1 s21).Acknowledgements Centre National de la Recherche Scientifique is gratefully acknowledged for financial support of this research in Dijon (UMR 5633 and 5632), Nancy (UPRESA 7036), and Paris (URA 1679). Financial support from Ecole Normale Superieure is also gratefully appreciated. G. B. acknowledges the Région-Bourgogne for a post-doctoral fellowship. References 1 Y. Kono and I. Fridovich, J. Biol. Chem., 1983, 258, 13 646. 2 M.L. Ludwig, K. A. Pattridge and W. C. Stallings, Metabolism and Enzyme Function, New York, 1986. 3 A. Willing, H. Follmann and G. Auling, Eur. J. Biochem., 1988, 175, 167. 4 J. H. Koek, S. W. Russell, L. Van der Wolf, R. Hage, J. B. Warnaar, A. L. Spek and L. DelPizzo, J. Chem. Soc., Dalton Trans., 1996, 353. 5 N. Kitajima, M. Osawa, S. Imai, K. Fujisawa, Y. Morooka, K. Heerwegh, C. A. Reed and P. D. W. Boyd, Inorg. Chem., 1994, 33, 4613. 6 Y. Naruta and M. Sasayama, J. Chem.Soc., Chem. Commun., 1994, 2667. 7 V. L. Pecoraro, M. J. Baldwin and A. Gelasco, Chem. Rev., 1994, 94, 807. 8 D. P. Riley and R. H. Weiss, J. Am. Chem. Soc., 1994, 116, 387. 9 D. E. De Vos, J. L. Meinershagen and T. Bein, Angew. Chem., Int. Ed. Engl., 1996, 35, 2211. 10 D. E. De Vos and T. Bein, J. Organomet. Chem., 1996, 520, 195. 11 R. Hage, Eur. Pat. 0549 272 A1, 1993. 12 R. Hage, J. E. Iburg, J. Kerschner, J. H. Koek, E. L. M. Lempers, R. J. Martens, U. S. Racherla, S. W.Russell, T. SwarthoV, M. R. P. Vanvliet, J. B. Warnaar, L. Vanderwolf and B. Krijnen, Nature (London), 1994, 369, 637. 13 S. M. Harriot-Jureller and J. L. Kerschner, Eur. Pat. 0544 519 A2, 1993. 14 R. J. Martens and S. Ton, Eur. Pat. 0544 490 A1, 1993. 15 V. C. Quee-Smith, L. Delpizzo, S. H. Jureller and J. L. Kerschner, Inorg. Chem., 1996, 35, 6461. 16 K. Wieghardt, U. Bossek, D. Ventur and J. Weiss, J. Chem. Soc., Chem. Commun., 1985, 347. 17 K. Wieghardt, U. Bossek, L. Zsolnai, G.Huttner, G. Blondin, J.-J. Girerd and F. Babonneau, J. Chem. Soc., Chem. Commun., 1987, 651. 18 K. J. Brewer, A. Liegeois, J. W. Otvos, M. Calvin and L. O. Spreer, J. Chem. Soc., Chem Commun., 1988, 1219. 19 K. J. Brewer, M. Calvin, R. S. Lumpkin, J. W. Otvos and L. O. Spreer, Inorg. Chem., 1989, 28, 4446. 20 P. A. Goodson and D. J. Hodgson, Inorg. Chim. Acta, 1990, 172, 49. 21 P. A. Daugherty, J. Glerup, P. A. Goodson, D. J. Hodgson and K. Michelsen, Acta Chem. Scand., 1991, 45, 244. 22 W. Nam and J. S. Valentine, J. Am. Chem. Soc., 1993, 115, 1772. 23 Y. Park, S. Kim and H. Na, J. Korean Chem. Soc., 1993, 37, 648. 24 A. M. Bond, R. Colton and J. J. Jackowski, Inorg. Chem., 1975, 14, 274. 25 M. F. C. Guedes da Silva, J. J. R. Frausto da Silva, A. J. L. Pombeiro, C. Amatore and J. N. Verpeaux, Organometallics, 1994, 13, 3943. 26 F. L. Wimmer, M. R. Snow and A. M. Bond, Inorg. Chem., 1974, 13, 1617. 27 R. A. Rader and D. R. McMillin, Inorg. Chem., 1979, 18, 545. 28 M. Rudolph, D. P. Reddy and S. W. Feldberg, Anal. Chem., 1994, 66, 589A. 29 E. K. Barefield, Inorg. Chem., 1972, 11, 2273. 30 E. K. Barefield, F. Wagner, A. W. Herlinger and A. R. Dahl, Inorg. Synth., 1976, 16, 220. 31 R. H. Blessing, Crystallogr. Rev., 1987, 1, 3.J. Chem. Soc., Dalton Trans., 1998, Pages 2233–2239 2239 32 G. M. Sheldrick, SHELX 76, Program for Crystal Structure Determinations, University of Göttingen, 1976. 33 C. K. Johnson, ORTEP, Report ORNL-3794, 2nd revision, Oak Ridge National Laboratory, TN, 1970. 34 C. K. Poon and M. L. Tobe, J. Chem. Soc. A, 1968, 1549. 35 C. K. Poon and K. C. Pun, Inorg. Chem., 1980, 19, 568. 36 P. K. Chan and C. K. Poon, J. Chem. Soc., Dalton Trans., 1976, 858. 37 M. E. Sosa-Torres and R. A. Toscano, Acta Crystallogr., Sect. C, 1997, 53, 1585. 38 N. W. Alcock, A. Berry and P. Moore, Acta Crystallogr., Sect. C, 1992, 48, 16. 39 A. Bakac and J. H. Espenson, Inorg. Chem., 1987, 26, 4353. 40 B. Bosnich, C. K. Poon and M. L. Tobe, Inorg. Chem., 1965, 4, 1102. 41 G. A. JeVrey and J. Mitra, Acta Crystallogr., Sect. B, 1983, 39, 469. 42 R. Guilard, O. Siri, A. Tabard, G. Broeker, P. Richard, D. J. Nurco and K. M. Smith, J. Chem. Soc., Dalton Trans., 1997, 3459. 43 D. T. Pierce, T. L. Hatfield, E. J. Billo and Y. Ping, Inorg. Chem., 1997, 36, 2950. 44 V. L. Pecoraro, Manganese Redox Enzymes, VCH, New York, 1992. Received 30th January 1998; Paper 8/00824H

ISSN:1477-9226    

DOI:10.1039/a800824h

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1999, 2237–2241 2237 Chelate complexes of cobalt(III) with bis(dithiolate) ligands: backbone influence on the electronic properties and the reactivity of the metal center Wolfram W. Seidel and F. Ekkehardt Hahn * Anorganisch-Chemisches Institut der Westfälischen Wilhelms-Universität Münster, Wilhelm-Klemm-Strasse 8, D-48149 Münster, Germany Received 21st January 1999, Accepted 6th May 1999 The tetradentate bis(dithiolate) ligands 1,4-bis[(2,3-disulfanylbenzamido)methyl]benzene H4L1 and 1,7-bis(2,3- disulfanylbenzamido)heptane H4L2 were prepared and converted into the dinuclear titanocene complexes [(TiCp2)2(L1)] 1 and [(TiCp2)2(L2)] 2, respectively. A ligand transfer reaction of 1 and 2 with [NR4]2[CoCl4] (R = Et or Me) led to the formation of the bis(dithiolato)cobaltate(III) complexes [NEt4][Co(L1)] 3 and [NR4][Co(L2] (R = Et 4a or Me 4b).The anion [Co(L2)]2 adopts an approximately square-planar co-ordination geometry.Two anions are connected by intermolecular Co–S contacts at the apical position of the metal to form [Co(L2)]2 22. In contrast, mononuclear 3 turned out to be unstable with respect to its co-ordination polymer. This phenomenological diVerence between 3 and 4a, 4b is reflected in the UV-VIS and cyclic voltammetry data of the complexes, which are compared to the data of the corresponding prototype complex tetraethylammonium [bis(3-N-benzylcarbamoylbenzene-1,2- dithiolato)cobaltate(III)] 5 with two unbridged dithiolato-ligands (L3)22.Polythiolate chelate ligands are able to force either unusual co-ordination geometries 1 or to eVect subtle changes in the complex geometry 2 due to the steric demand of the ligand backbone. These structural changes often cause desirable consequences for the electronic properties 1 as well as for the reactivity at the metal center.1,2 By comparison of such complexes with their prototypes 3 exhibiting unbridged thiolates, structure behaviour relationships can be inferred, which eventually are expected to shed light on structure function relationships in certain metalloenzymes with sulfur dominated co-ordination in a cofactor.Multidentate, sulfur-rich ligands of type A based on benzene-1,2-dithiol have been used successfully in the preparation of model compounds for the active sites in nitrogenase 4 and for certain nickel enzymes.5 In these ligands two aromatic dithiolate units are bridged via a thioether link which leads to a tetradentate ligand with two thiolate and two thioether sulfur donor atoms.The thioether links lead, in spite of the superb properties of ligands of type A, to a loss of the dithiolene character in the bidentate binding unit. Our approach to sulfur-rich multidentate ligands was directed at the utilization of the dithiolene-like metal binding capacity of benzene-1,2-dithiolates, that diVers from the benzene thiolate-thioether arrangement in A.We synthesized tetradentate ligands with two benzene-1,2-dithiolate donors of type B in which the bidentate binding unit might retain some dithiolene character upon co-ordination. Our research was particularly prompted by the intriguing "non innocent" behaviour of dithiolene ligands 6 including the stability of their complexes S SH S HS SH NH SH O HN O A B HS HS in two or more oxidation states and the (minor 7 or major 8) dependence of the complex co-ordination geometry on the oxidation state of the metal.The geometry of [Mn(tdt)2]n2 (tdt22 = toluene-3,4-dithiolate) for instance changes from planar (n = 1) to distorted tetrahedral (n = 2) upon reduction.8 A suitably bridged bis(dithiolate) ligand of type B might not be able to accommodate both co-ordination geometries and thus can destabilize a specific oxidation state possibly generating an unprecedented reactivity. Thus, ligands of type B oVer not only the opportunity for the preparation of new model complexes for the active sites in sulfur containing metalloenzymes but might also allow an investigation of how geometric constraints originating comparatively far from the metal centre in the organic backbone of the ligand aVect the properties of the co-ordinated metal centre.A first report on the synthesis and co-ordination chemistry of a tripodal tris(dithiolate) ligand appeared in 1995.9 A report on a bis(dithiolate) ligand and its chelate complex with the CpTi fragment was published recently.10 In this contribution we report on cobalt(III) complexes of the bis(dithiolate) ligands H4L1 and H4L2.We describe the synthesis of the complex anions [Co(L1)]2 and [Co(L2)]2 by a dithiolate transfer reaction as well as the prototype complex with an unbridged dithiolate ligand [Co(L3)2]2. In addition, a comparison of the UV-VIS and cyclic voltammetry data of all bis(dithiolato)cobaltate complexes is presented. Results and discussion Preparation of the complexes The investigation of the co-ordination chemistry of H4L1 and H4L2 has been hampered by the poor solubility of these ligands in most organic solvents (exceptions dmf and dmso).In addition, the sensitivity of the thiol functions to aerial oxidation restricted possible purification methods for them. The solubility in protic solvents can be improved considerably upon complete deprotonation with NaOMe. However, simple metathesis reactions of Na4L1 or Na4L2 with metal halides like MCl2(H2O)6 (M = Ni or Co) proved unsuccessful at least in our hands,2238 J.Chem. Soc., Dalton Trans., 1999, 2237–2241 apparently because the high reactivity of these salts promotes the formation of polynuclear species. In contrast, we found that ligand transfer reactions utilizing the halophilicity of the Cp2Ti unit constitutes a useful tool for the synthesis of chelate complexes with H4L1 and H4L2. The dinuclear bis(titanocene) complexes of (L1)42 and (L2)42 are easily accessible by reaction of [TiCp2Cl2] with H4L1 and H4L2 and triethylamine in thf.The intensly green complexes [(TiCp2)2(L1)] 1 and [(TiCp2)2(L2)] 2 are air stable and highly soluble in aprotic solvents like thf and CH2Cl2. In addition, they can be subjected to chromatographic purification. This advantage turned out to be crucial due to the restricted opportunities to purify unco-ordinated H4L1 or H4L2. In our previous work9,10 we found complex 1 to be reactive towards certain inorganic halides like HCl and NR4Cl.The halophilicity of the Cp2Ti unit has been exploited previously in transmetallation reactions between titanocene dithiolate complexes and late transition metal halides to give dithiolate complexes of the late transition metals and [TiCp2Cl2].11 This type of transfer has also been extensively demonstrated with non-metal halides, which allowed for the preparation of cyclic sulfur derivatives from [TiCp2S5] and SxCl2.12 We decided to investigate the straightforward reaction of 1 and 2 with tetrachlorometalates invoking the simultaneous transfer of two dithiolate units.A CH3CN–thf solvent mixture had to be used to keep both the salts [NR4]2[CoCl4] and the uncharged complexes 1 or 2 in solution. Whereas at ambient temperature no reaction could be observed, under reflux conditions a slow change over the course of hours from dark green towards eventual intense blue indicated the formation of bis(dithiolato)- cobaltate(III) complexes.13 Thus, [NR4]2[CoCl4] and 1 or 2 reacted under elimination of two equivalents of [TiCp2Cl2] and aerial cobalt(II) oxidation to yield [NEt4][Co(L1)] 3 and [NR4]- [Co(L2)] (R = ethyl 4a or methyl 4b), respectively (Scheme 1).Both the solubility of 3 and 4a, 4b in thf, CH2Cl2 and CH3CN and the detection of the parent anions in the FAB mass spectra are in accord with the assumption of mononuclear chelate complexes for 3 and 4a, 4b.Both salts containing the [Co(L2)]2 complex anion could be crystallized by either diVusion of Et2O in a CH3CN solution (4a) or by cooling of a toluene–acetone solution from 50 8C to ambient temperature (4b). The determination of the molecular structure of complex 4b by X-ray diVraction was hampered by the small size of the thin, needle shaped crystals. The results obtained from three independent data sets are unsuYcient for publication but they confirm the identity of 4b as chelate complex containing the two dithiolate binding units of (L2)42.A number of structure determinations of bis(benzenedithiolato)cobaltate(III) complexes have been published.14–16 The complex anions [Co(tdt)2]214 and [Co(S2C6Cl4)2]215 exhibit an approximately square-planar co-ordination geometry with Co–S distances of 2.15–2.20 Å and they share this feature with the [Co(L2)]2 SH SH N Ph O H SH SH N N HS HS H H O O SH SH N N HS HS H H O O H4 L1 H4 L2 H2 L3 anion. Moreover, two molecules of [Co(L2)]2 are connected by cobalt–sulfur contacts. Remarkably, [Co(S2C6Cl4)2]2 shows the same structural characteristics, whereas mononuclear [Co(tdt)2]2 does not.Apparently the electron poorer nature of (L2)42 and C6Cl4S2 22 compared to tdt22 enhances the electrophilicity at the cobalt centre thereby leading to the increase of the co-ordination number in the solid state. Reactivity of complex 3 Solutions of complexes 4a and 4b in aprotic solvents can be kept without any sign of decomposition.Surprisingly, dark blue solutions of 3 in CH2Cl2 or CH3CN decolourize over the course of days with irreversible precipitation of a blue solid. Microanalytical and UV-VIS spectroscopic data (in dmf, see below) of this secondary product 3* are identical to those calculated for 3, which led us to the assumption of a polymeric structure for 3*. We assume, that a gradual conversion of the chelate complex anion [Co(L1)]2 into a complex polymer by way of a chelate ring opening polymerization takes place.This would lessen the strain in the small backbone of the complex anion [Co(L1)]2. This assumption is corroborated by inspection of the structural parameters of [NMe4][TiCp(L1)], the only other known complex of the (L1)42 ligand.10 The [TiCp(L1)]2 anion shows bent chelate rings in an exo/ endo fashion (Fig. 1). This exo/endo bend reduces the mutual distance of the amide groups to be bridged and allows the amide functions to assume the favored twist out of the aromatic planes.This form of strain reduction is unlikely for complex 3, because bent chelate rings are so far unknown for bis- (dithiolato)-cobaltate(III) complexes. A dinuclear species with cobalt sulfur contacts between the CoS4 planes as observed in the solid state structure of 4b could be considered as a Fig. 1 Molecular structure of the anion in [NMe4][TiCp(L1)]. S S S S N H Ti N O O H – Scheme 1 Synthesis of the complexes 3 and 4a.Ti N S S H N H S Ti S R O O S S N H S S N R H Co O O [NEt4]2[CoCl4], O2, CH3CN / THF, -Cp2TiCl2 1 R = p-C6H4 2 R = (CH2)5 3 R = p-C6H4 4a R = (CH2)5 [NEt4]J. Chem. Soc., Dalton Trans., 1999, 2237–2241 2239 reasonable transition state for the polymerization process of 3 to 3*. Electronic properties of the complexes 3–5 In order to evaluate the influence of the bridge on the electronic properties at the cobalt centre we carried out UV-VIS and cyclic voltammetry studies. The complex [NEt4][Co(L3)2] 5, with two unbridged ligands, served as a prototype for comparison of bridged and unbridged benzamide 1,2-dithiolate complexes. The UV-VIS spectroscopic data of 3–5 are depicted in Fig. 2 and compared in Table 1. The spectra of 3, 4a and 5 are very similar indicating the high preference of a squareplanar co-ordination geometry. However, while the spectra for 4a and 5 are almost identical (in spite of the unresolved question of cis/trans isomerism for 5), the changed intensity relations and the hypsochromic shift of the bands in the spectra of 3 indicate a slightly diVerent situation at this cobalt(III) centre in accordance with the previously described reactivity pattern.Fig. 2 The UV-VIS spectra of complexes 3–5 (* impurity). Table 1 UV-VIS Data a of bis(benzenedithiolato)cobaltate(III) complexes Complex l1/nm l2/nm l3/nm 3 [ NEt4][Co(L1)] b 4a [ NEt4][Co(L2)] b 5 [ NEt4][Co(L3)2] b [NBu4][Co(tdt)2] 13,c 320 (2) 325 (11200) 323 (10700) 315 (12700) 363 (2) 368 (11400) 367 (11100) 361 (13700) 653 (2) 660 (6500) 653 (6300) 660 (12000) a Values in parentheses: e/M21 cm21.b Solvent CH2Cl2. c Solvent CH3CN. Cyclic voltammetry revealed in a more distinct manner differences between complexes 3 and 4a (Fig. 3, Table 2). In the range between 0 and 21 V vs. Ag–AgCl the redox process [CoL2]2–[CoL2]22 could be detected by means of a signal, which shows all the features of a reversible one electron transfer (DEp = 65 mV, ipa/ipc ª 1).The half-wave potential of 5 in dmf was determined to be E1/2 = 2510 mV. This result can be correlated very well with data 17 regarding the electron density in the ligands. The half-wave potential of 5 with benzamide 1,2- dithiolate ligands falls between that of the corresponding complexes with the electron rich C6H4S2 22 and the electron deficient C6Cl4S2 22 (Table 2). Complex 4a with the flexible backbone and the prototype complex 5 exhibit an almost identical half-wave potential (E1/2 = 2510 mV).In contrast, the half-wave potential of 3 with the shorter and more rigid bridge was determined to be E1/2 = 2570 mV. In addition, this electron transfer appears to be less reversible (Figure 3). The reduction of 3, which is expected to lengthen the metal–sulfur bonds and therefore to add stress to the ligand backbone is more diYcult by 60 mV relative to the unbridged and flexibly bridged bis(dithiolate) complexes 4a and 5.This result corroborates the assumption that the short and rigid xylylene bridge in (L1)42 is too small to permit the existence of a thermodynamically stable chelate complex [Co(L1)]2. Concluding remarks We have shown that the electronic properties and the reactivity of the cobalt centre in bis(dithiolato)cobaltate complexes with the new bridged bis(dithiolate) ligands (L1)42 and (L2)42 can be manipulated by the character of the bridge, thus by eVects originating relatively far from the metal centre in the organic backbone of the ligand.Our experiments indicate that large multidentate ligands with many degrees of freedom do not favor the formation of chelate complexes in a fast reaction like Fig. 3 Cyclic voltammograms of complexes 3–5. Table 2 Half-wave potentials of bis(benzenedithiolato)cobaltate complexes: [Co(L)2]2 1 e2 [Co(L)2]22 Complex E1/2 a/mV [NEt4][Co(S2C6H4)2] 17 3 [ NEt4][Co(L1)] 4a [ NEt4][Co(L2)] 5 [ NEt4][Co(L3)2] [NEt4][Co(S2C6Cl4)2] 17 2880 b 2570 2510 2510 2350 b a vs.Ag–AgCl (3 M KCl solution), solvent dmf. b Potential in the original reference vs. Ag–AgClO4.2240 J. Chem. Soc., Dalton Trans., 1999, 2237–2241 metal halogenide–sodium thiolate metathesis owing to the necessary rearrangement of the ligand during complex formation. In contrast to the fast metathesis reaction, the ligand transfer reactions between the bis(titanocene dithiolate) complexes 1 and 2, respectively, and tetrachlorocobaltate(II) salts described for the preparation of 3–5 proceed under reflux conditions in the course of hours.The comparatively low reaction rate suggests a high activation enthalpy which turns out to be advantageous for the formation of the chelate complexes 3 and 4a, 4b, because the rearrangement barriers for the ligand become negligible under these conditions. Experimental If not noted otherwise, all manipulations were performed in an atmosphere of dry argon by using standard Schlenk techniques.Solvents were dried by standard methods and freshly distilled prior to use. The 1H and 13C NMR spectra were recorded on a Bruker AM 250 spectrometer. Elemental analyses (C, H, N) were performed at the Freie Universität Berlin on a Perkin- Elmer 240 C elemental analyzer. Mass spectra (FAB) were recorded on a Varian MAT CH 5 DF instrument, UV-VIS spectra on a Perkin-Elmer Lambda 9 spectrometer. Cyclic voltammetric data were acquired on Bank High Power Potentiostat Wenking HP 72 and Bank Scan Generator Wenking Model VSG 83 instruments using a platinum working electrode, a Ag– AgCl double junction electrode (3 M KCl solution) as reference and NBu4ClO4 as supporting electrolyte; Fc was used as an internal standard and exhibited E1/2 = 480 mV under these conditions.The complex salts [NEt4]2[CoCl4] and [NMe4]2[CoCl4] were prepared according to ref. 18, [TiCp2Cl2] was used as purchased, H4L1, [(TiCp2)2(L1)], H2L3 and [TiCp2(L3)] were prepared as published 9,10 and H4L2 was prepared by an identical procedure to that described for H4L1 10 using 1,7-diaminoheptane.Preparations [(TiCp2)2(L2)] 2. Solid [TiCp2Cl2] (293 mg, 1.18 mmol) and subsequently 50 ml thf were added to 0.59 mmol of crude H4L2. The addition of 400 ml of triethylamine (2.87 mmol) to the red suspension (undissolved H4L2) resulted in a fast change to intense green. After 2 h of stirring the solvent was removed in vacuo and the residue redissolved in a little amount of CH2Cl2 was put on a chromatographic column (SiO2). The dinuclear complex 2 was eluted with CH2Cl2–CH3OH (25: 1).Thorough drying of the dark green powder in vacuo yielded 354 mg of 2 (73%, based on Pri 4L2) (Found: C, 60.15; H, 5.13; N, 3.42. C41H42N2O2S4Ti2 requires C, 60.14; H, 5.17; N, 3.42%). NMR: dH (250 MHz, CDCl3) 7.61 (d, 2 H, aryl H), 7.48 (d, 2 H, aryl H), 7.22 (t, 2 H, NH), 7.10 (t, 2 H, aryl H), 5.97 (s, br, 20 H, C5H5), 3.38 (dt, 4 H, NCH2), 1.53 (m, 4 H, NCH2CH2) and 1.32 [m, 6 H, (CH2)3]; dC (62.90 MHz, CDCl3) 168.2 (CO), 158.9, 152.6, 135.4, 131.5, 125.6, 124.4 (aryl C), 113.0 (C5H5), 39.8 (NCH2), 29.3, 28.8 and 26.9 (CH2).[NEt4][Co(L1)] 3. A solution of 82 mg (0.1 mmol) complex 1 in 70 ml thf was added by syringe to a solution of 46 mg (0.1 mmol) [NEt4]2[CoCl4] in 70 ml CH3CN. While refluxing this mixture for 12 h it changed slowly from green to blue. After cooling the solvents were removed in vacuo.The residue was washed two times with thf (5 ml) in order to remove [TiCp2Cl2]. Extraction of the crude product 3 with CH2Cl2 or dmf and subsequent filtration yielded solutions used for UV-VIS spectroscopy (CH2Cl2) and CV (dmf). Keeping the filtered blue solutions in CH2Cl2 or CH3CN resulted in the irreversible precipitation of polymer 3* in approximate composition concomitant with a decolorisation of the solution. Complex 3: FABMS: m/z (M2), 527, [Co(L1)]2 requires 527.Complex 3*: (Found: C, 54.30; H, 5.64; N, 6.24. C30H36CoN3O2S4 requires C, 54.78; H, 5.52; N, 6.39%): UV-VIS lmax/nm (dmf) 321, 366 and 657. [NR4][Co(L2] (R 5 Et 4a or Me 4b). A solution of 82 mg (0.1 mmol) complex 2 in 70 ml thf was added by syringe to a solution of 46 mg [NEt4]2[CoCl4] {35 mg [NMe4]2[CoCl4], 0.1 mmol} in 70 ml CH3CN. While refluxing this mixture for 12 h it changed slowly from green to blue. After cooling the solvents were removed in vacuo. The residue was washed two times with thf (5 ml) in order to remove [TiCp2Cl2].DiVusion of Et2O into a solution of crude 4a in 10 ml CH3CN resulted at first in the precipitation of by-products. Filtration and cooling the solution to 230 8C yielded 20 mg (30%) 4a in the form of large thin sheets. FAB-MS: m/z 521 (M2), [Co(L2)]2 requires 521 (Found: C, 53.63; H, 6.52; N, 6.38. C29H42CoN3O2S4 requires C, 53.44; H, 6.49; N, 6.45%). In contrast to 4a, the addition of 10 ml THF to the product mixture 4b led to a blue solution.Addition of 10 ml toluene, subsequent filtration and cooling to 230 8C resulted in the precipitation of crude 4b. Thin, needle shaped crystals were obtained by slow cooling of a toluene–acetone (5 : 1) solution of 4b from 50 8C to ambient temperature (Found: C, 50.56; H, 5.73; N, 6.99. C25H34CoN3O2S4 requires C, 50.40; H, 5.75; N, 7.05%). [NEt4][Co(L3 2] 5. A mixture of 190 mg (0.42 mmol) [TiCp2- (L3)] and 97 mg (0.21 mmol) [NEt4]2[CoCl4] in 30 ml CH3CN was refluxed for 12 h.Cooling this solution to 230 8C resulted in the precipitation of a blue powder, which was isolated as complex 5. DiVusion of Et2O into a CH3CN solution of 5 yielded 98 mg (63%) of needle shaped blue crystals. FAB-MS: m/z 605 (M2), [Co(L3)2]2 requires 605; (Found: C, 58.69; H, 5.78; N, 5.71. C36H42CoN3O2S4 requires C, 58.76; H, 5.75; N, 5.71%). Acknowledgements We thank the Deutsche Forschungsgemeinschaft and the Fonds der Chemischen Industrie for financial support.References 1 J. D. Franolic, W. Y. Wang and M. Millar, J. Am. Chem. Soc., 1992, 114, 6587; D. H. Nguyen, H.-F. Hsu, M. Millar, S. A. Koch, C. Achim, E. L. Bominaar and E. Münck, J. Am. Chem. Soc., 1996, 118, 8963; K. K. Nanda, E. Sinn and A. W. Addison, Inorg. Chem., 1996, 35, 1; T. Beissel, F. Birkelbach, E. Bill, T. Glaser, F. Kesting, C. Krebs, T. Weyhermüller, K. Wieghardt, C. ButzlaV and A. X. Trautwein, J. Am. Chem. Soc., 1996, 118, 12376; T.Beissel, T. Glaser, F. Kesting, K. Wieghardt and B. Nuber, Inorg. Chem., 1996, 35, 3936; P. J. Blower, J. R. Dilworth, J. P. Hutchinson and J. Zubieta, Transition Met. Chem., 1982, 7, 353. 2 N. de Vries, J. Cook, A. G. Jones and A. Davison, Inorg. Chem., 1991, 30, 2662; H. Spies, M. Glaser, H.-J. Pietzsch, F. E. Hahn, O. Kintzel and T Lügger, Angew. Chem., Int. Ed. Engl., 1994, 33, 1418. 3 D. Swenson, N. C. Baenziger and D. Coucouvanis, J. Am. Chem. Soc., 1978, 100, 1932; N.de Vries, J. C. Dewan, A. G. Jones and A. Davison, Inorg. Chem., 1988, 27, 1574; T. Maina, A. Pecorale, A. Dolmella, G. Bandoli and U. Mazzi, J. Chem. Soc., Dalton Trans., 1994, 2437. 4 D. Sellmann, W. Soglowek, F. Knoch and M. Moll, Angew. Chem., Int. Ed. Engl., 1989, 28, 1271; D. Sellmann, W. Soglowek, F. Knoch, G. Ritter and J. Dengler, Inorg. Chem., 1992, 31, 3711. 5 D. Sellmann, G. H. Rackelmann and F. W. Heinemann, Chem. Eur. J., 1997, 3, 2071; D. Sellmann, D. Heusinger, F.Knoch and M. Moll, J. Am. Chem. Soc., 1996, 118, 5368. 6 J. A. McCleverty, Prog. Inorg. Chem., 1968, 10, 49; G. N. Schrauzer, Acc. Chem. Res., 1969, 2, 72; R. Eisenberg, Prog. Inorg. Chem., 1970, 12, 295; R. P. Burns and C. A. McAuliVe, Adv. Inorg. Chem. Radiochem., 1979, 22, 303; C. Mahadevan, J. Crystallogr. Spectrosc. Res., 1986, 16, 347. 7 For example: C. Mahadevan, M. Seshasayee, P. Kuppusamy and P. T. Manoharan, J. Crystallogr. Spectrosc. Res., 1985, 15, 305; D. Sellmann, S. Fünfgelder, F. Knoch and M. Moll, Z. Naturforsch. Teil B, 1991, 46, 1601.J. Chem. Soc., Dalton Trans., 1999, 2237–2241 2241 8 G. Henkel, K. Greiwe and B. Krebs, Angew. Chem., Int. Ed. Engl., 1985, 24, 117. 9 F. E. Hahn and W. W. Seidel, Angew. Chem., Int. Ed. Engl., 1995, 34, 2700. 10 W. W. Seidel, F. E. Hahn and T. Lügger, Inorg. Chem., 1998, 37, 6587. 11 C. M. Bolinger and T. B. Rauchfuss, Inorg. Chem., 1982, 21, 3947; K. Osakada, Y. Kawaguchi and T. Yamamoto, Organometallics, 1995, 14, 4542; T. A. Wark and D. W. Stephan, Can. J. Chem., 1990, 68, 565. 12 M. Schmidt, B. Block, H. D. Block, H. Köpf and E. Wilhelm, Angew. Chem., Int. Ed. Engl., 1968, 7, 632; H. W. Roesky, H. Zamankhan, J. W. Bats and H. Fuess, Angew. Chem., Int. Ed. Engl., 1980, 19, 125. 13 D. T. Sawyer, G. S. Srivatsa, M. E. Bodini, W. P. Schaefer and R. M. Wing, J. Am. Chem. Soc., 1986, 108, 936. 14 R. Eisenberg, Z. Dori, H. B. Gray and J. A. Ibers, Inorg. Chem, 1968, 7, 741. 15 M. J. Baker-Hawkes, Z. Dori, R. Eisenberg and H. B. Gray, J. Am. Chem. Soc., 1968, 90, 4253. 16 C. G. Pierpont and R. Eisenberg, Inorg. Chem, 1970, 9, 2218; B. Kang, J. Peng, M. Hong, D. Wu, X. Chen, L. Weng, X. Lei and H. Liu, J. Chem. Soc., Dalton Trans., 1991, 2897. 17 M. J. Baker-Hawkes, E. Billig and H. B. Gray, J. Am. Chem. Soc., 1966, 88, 4870. 18 J. R. Wiesner, R. C. Srivastva, C. H. L. Kennard, M. DiViara and E. C. Lingafelter, Acta Crystallogr., 1967, 23, 565. Paper 9/00578A

ISSN:1477-9226    

DOI:10.1039/a900578a

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1998, Pages 2241–2246 2241 Mechanistic studies on the epoxidation of alkenes with molecular oxygen and aldehydes catalysed by transition metal–‚-diketonate complexes † Bastienne B. Wentzel, Patricia A. Gosling, Martin C. Feiters and Roeland J. M. Nolte* Department of Organic Chemistry/Nijmegen SON Research Center, University of Nijmegen, 6525 ED Nijmegen, The Netherlands The scope, mechanism and kinetics of the aerobic epoxidation of alkenes with an aldehyde and substituted b-diketonate–transition metal complexes as catalysts were studied.b-Diketonate complexes of nickel(II) proved to be among the best catalysts for this reaction. The epoxidation is not dependent on substrate concentration and is first order in aldehyde, catalyst concentration and oxygen partial pressure. It was shown by reactivity studies and EPR experiments that the reaction is radical in nature. Additional evidence for this was obtained from stereochemical investigations.The metal catalyst is not only an eYcient initiator of the reaction, but is also believed to enhance the reactivity of intermediate species in the oxidation process by allowing these to co-ordinate to the metal center. A mechanism is proposed for the catalytic reaction. Molecular oxygen as a cheap, clean and readily available oxidant has received much attention in recent years.1 Mukaiyama and co-workers 2–7 and others 8–11 have reported that molecular oxygen can be used as the terminal oxidant in the epoxidation of alkenes with an aldehyde or primary alcohol as coreactant and a metal b-diketonate as a catalyst (Scheme 1).There has been discussion in the literature about the mechanism of the ‘Mukaiyama’ catalytic system and the role of the transitionmetal catalyst in it, which can be omitted as was shown by Kaneda et al.12 Since peroxyacids, which are formed in the autoxidation of aldehydes, are powerful epoxidizing reagents, the reaction in Scheme 1 might proceed through the peroxyacid as the actual epoxidizing agent.The only role of the transitionmetal catalyst in this scenario is to catalyse the formation of the peracid as shown in Scheme 2. Another possible mechanism proposed in the literature 10 is the formation of a metal–oxygen complex which reacts to form an oxometal species, similar to species described for manganese or vanadium.13,14 In Scheme 3 this mechanism is outlined for a transition-metal(II)–b-diketonate complex.A combination of the mechanisms in Schemes 2 and 3 was considered by Nam et al.15 They investigated the ‘Mukaiyama’ system using cyclam-type transition-metal complexes and concluded from indirect evidence that the epoxidation reaction in their system is radical in nature. The peroxyacid and the oxometal mechanisms in Schemes 2 and 3 were believed to play no role. An acylperoxy radical, rather than a peroxyacid, was proposed to react with the alkene to form an epoxide.Alternatively, the acylperoxy radical could co-ordinate to the metal first, and this complex subsequently epoxidizes the alkene. The same authors reported shortly after 16 that cyclam complexes of NiII are inhibitors of this radical reaction. These complexes were believed to be suYciently good reducing agents to react with an acylperoxy radical and form an unreactive acylperoxy anion and a nickel(III) complex. In the present paper we further explore the scope, kinetics and mechanism of the epoxidation of alkenes by the ‘Mukaiyama’ system.The most eVective catalysts for this reaction reported so far, viz. second-row transition-metal complexes of † Supplementary data available: epoxidation results. For direct electronic access see http://www.rsc.org/suppdata/dt/1998/2241/, otherwise available from BLDSC (No. SUP 57387, 3 pp.) or the RSC Library. See Instructions for Authors, 1998, Issue 1 (http://www.rsc.org/dalton).Non-SI units employed: atm = 101 325 Pa, G = 1024 T. b-diketonates, are used. In spite of the extensive discussions in the literature the role of the metal complex is still not entirely clear. This issue will be addressed as well. Experimental Materials Dichloromethane was dried over CaCl2, distilled from CaH under dry nitrogen and stored over molecular sieves. Acetonitrile was HPLC grade. All other solvents and isobutyraldehyde were distilled before use. Oxygen was obtained from Hoek-Loos and dried over calcium chloride.All alkene substrates were commercial samples (Aldrich) and were purified by column chromatography over basic alumina with CH2Cl2 as eluent or by vacuum distillation. An exception is S-limonene [1-methyl-4-(1-methylethenyl)cyclohexene] (Aldrich, 96%) which was used as received. Epoxide products were identified Scheme 1 O O2, (catalyst) RCHO RCO2H Scheme 2 Peracid epoxidation mechanism O RCH O RC OOH O O RC OH + (transitionmetal catalyst) O2 Scheme 3 Oxo–metal epoxidation mechanism MII MIII O• MII O + RCHO, O2 RCO2H MIV O2242 J.Chem. Soc., Dalton Trans., 1998, Pages 2241–2246 with gas chromatography. The metal complexes were commercial products or were synthesized according to literature procedures.17,18 Their physical properties were consistent with their structures and with literature values.17,18 Instrumentation The GC analyses were performed on a Varian 3700 instrument with a fused-silica capillary column (25 m length, 25 mm diameter) with a CP-sil stationary phase or a 15 m × 35 mm diameter column with an FFAP stationary phase.The instrument was equipped with a flame-ionization detector and coupled to a Hewlett-Packard 3395 integrator. The UV/VIS spectra were taken on a Perkin-Elmer Lambda 5 spectrometer, IR spectra on a Bio-Rad FTS-25 spectrometer, low-temperature EPR spectra on a Bruker Electron Spin Resonance ER-220D-LR spectrometer and room-temperature spectra on a Bruker ESP-300 instrument.The NMR analyses were performed on a Bruker AC-300 or WH-90 instrument; the solvent was CDCl3. The catalytic system The standard conditions used in the epoxidation of alkenes by nickel(II)–b-diketonate complexes in the presence of an aldehyde were as follows. 0.1 mol l21 Alkene, 0.3 mol l21 aldehyde and 1 mmol l21 catalyst were stirred (1000 revolutions min21) in CH2Cl2 at 25.0 ± 0.5 8C under 1.0 atm of oxygen. Unless indicated otherwise, kinetic experiments were carried out with a-pinene as the alkene substrate, isobutyraldehyde as the coreagent and bis[3-(p-tert-butylbenzyl)pentane-2,4-dionato]- nickel(II) 1c as the catalyst.Errors were estimated to be less than 5%. The catalytic reaction was followed by monitoring the disappearance of the substrate and the appearance of product(s) as a function of time with gas chromatography; the internal standard was dmf (100.0 ml). Adding dmf in small quantities did not aVect the reaction.Determination of CO2 evolved from the epoxidation reaction 19 A standard reaction mixture (see above) was prepared with [Ni(acac)2] 1a as the catalyst and S-limonene as the substrate. The exhaust gas of the reaction was bubbled though a BaCl2 solution [in 4 mol l21 NaCl (aq)–ethanol–glycol (1.5:2:1 v/v/v) at pH 11], and the CO2 formed precipitated as BaCO3. The turbidity of regularly taken samples from the barium solution was measured with UV spectroscopy at 360 nm, and the amount of CO2 that had evolved was calculated from a calibration curve.Results Scope of the epoxidation reaction A number of substrates were tested in the epoxidation reaction using the ‘Mukaiyama’ conditions and compound 1c (MII = NiII) as the catalyst. The results are collected in Table 1. Substituted alkenes, especially a-pinene, limonene and norbornene (bicyclo[2.2.1]hept-2-ene) gave very good yields as was expected based on other studies (see for example Yamada et al.4 and Fdil et al.11).Also styrene is a very good substrate for this epoxidation reaction. Doubly substituted and electron-rich alkenes such as cis-stilbene, cyclohexene, b-pinene, trans-b-methylstyrene and camphene gave poorer but still respectable yields of epoxides between 23 (b-pinene) and 61% (camphene). We also tested a variety of other metal complexes and metal salts in the epoxidation of a-pinene and S-limonene (see SUP 57387). In agreement with literature studies, nickel and cobalt complexes gave the highest epoxide yields (e.g. Fdil et al.11).It is noteworthy in this respect that Nam et al.15,16 found nickel cyclam-type complexes to be inactive in their epoxidation reactions. These complexes were shown to reduce the generated acyl peroxy radical to the peroxy anion which is inactive as an oxidant. Electronic and steric eVects In order to study electronic eVects a new series of 3-substituted nickel(II)–b-diketonate complexes (1a–1h) was synthesized and tested as epoxidation catalysts.The results for the epoxidation of a-pinene are shown in Table 2. As can be seen in Table 2 there is a small but significant eVect on the turnover rate of a-pinene by the catalyst when the para position of the benzene ring of the nickel complex is altered. Substitution of the benzene ring with an electron-withdrawing nitro group (entry 6) yields a catalyst with a relatively high turnover number (39) when compared to 11 turnovers found for the catalyst unsubstituted at the aromatic ring (entry 4).Remarkably, the catalyst substituted with an electron-rich methoxybenzyl group (entry 5) gives a high turnover number as well (32). Apparently, electronic eVects do not play a major role in the reaction catalysed by the nickel(II) complexes 1. Aldehyde reactivity The reactivity of a variety of aldehydes as coreactants in Scheme 1 was tested under standard conditions (see SUP 57387).Straight-chained and branched aldehydes such as pivaldehyde (68% a-pinene epoxide after 4 h) and isobutyraldehyde (91% epoxide) were the most active coreactants. Aromatic or conjugated aldehydes such as benzaldehyde and cinnamaldehyde were completely inactive under the reaction conditions. Investigations into a heterogeneous epoxidation system by Laszlo and O R R¢ O R¢ O R R¢ O R¢ MII 1 a R = H, R¢ = CH3 b R = H, R¢ = p-methoxyphenyl c R = p-tert-butylbenzyl, R¢ = CH3 d R = benzyl, R¢ = CH3 e R = p-methoxybenzyl, R¢ = CH3 f R = p-nitrobenzyl, R¢ = CH3 g R = p-fluorobenzyl, R¢ = CH3 h R = ethyl, R¢ = CH3 i R = H, R¢ = CF3 Table 1 Epoxidation of alkene substrates * Substrate a-Pinene b-Pinene S-Limonene Norbornene Styrene trans-b-Methylstyrene Camphene Allylbenzene cis-Stilbene trans-Stilbene Cyclohexene Oct-1-ene Conversion (%) 93 27 77 100 98 51 68 16 41 6 49 23 Yield epoxide (%) 86 23 72 96 79 47 61 12 36 5 45 19 * Reaction conditions: 0.1 mol l21 alkene, 0.3 mol l21 isobutyraldehyde, 1.0 × 1023 mol l21 catalyst 1c, 5.0 cm3 CH2Cl2, 1.0 atm O2, 25 8C, 4 h.Table 2 Epoxidation of a-pinene catalysed by substituted nickel(II)– b-diketonate complexes a Entry 123456789 Complex 1 abc def ghi Turnover numberb 15 c 21 11 32 39 24 81 a Reaction conditions as in Table 1. b Estimated error: 5%. c 74% Yield in 4 h.J. Chem. Soc., Dalton Trans., 1998, Pages 2241–2246 2243 Levart,20 using kaolinite and an aldehyde in the presence of O2, yielded similar aldehyde reactivities.If the acylperoxy radical is the active oxidizing agent, this radical might yield a carboxyl radical after epoxidation. The latter radical can decompose into CO2 and an alkyl radical. As Lassila et al.21 have suggested, epoxidation and decomposition may occur in a concerted process. A relatively stable alkyl radical is formed in the case of isobutyraldehyde and pivaldehyde, whereas the alkyl or aryl radicals generated from the other aldehydes will be less stable.This decomposition could create a driving force for the epoxidation reaction. If decomposition of the carboxyl radical into carbon dioxide and an alkyl radical plays a major role in the reaction, it should be possible to detect this by measuring the amount of CO2 evolving from the reaction; CO2 was determined as described in the Experimental section. The results are shown in Fig. 1. After 3 h 67% of the aldehyde had reacted and only 10% had evolved as CO2.Isopropyl hydroperoxide might be anticipated as an oxidation product of isobutyraldehyde.21 Decomposition of this hydroperoxide would yield acetone or isopropyl alcohol, neither of which was detected by GC. Finally, only a part of the converted aldehyde was retrieved as isobutyric acid (approximately 10% by GC), indicating a diVerence in reaction mechanism of the present catalytic system and the systems reported by, for example, Mizuno et al.,22 Yanai et al.10 (the ‘Mukaiyama’ system) and Nam et al.15 (see also Discussion).Stereochemistry In order to obtain information about the stereochemistry of the reaction we studied the epoxidation of cis-stilbene with various nickel(II) complexes and several oxidants. The results are shown in Table 3. From a comparison of the products obtained with m-chloroperbenzoic acid (entry 4) with the products obtained with the nickel(II) catalysts (other entries) it can be concluded that a free peroxyacid cannot be the main oxidizing species in our system; in that case the stereochemistry of the epoxide would have been retained, which is not observed.In the presence of a small quantity of pyridine the cis : trans ratio was shifted from 1: 13 (entry 1) to 1 : 45 (entry 2), indicating that, although the main pathway for epoxidation is not concerted, a small fraction of the products are formed via a concerted pathway, which is inhibited by the presence of a co-ordinating ligand such as pyridine.We found that it is not possible to induce chirality in the epoxidation products by using chiral nickel(II)–b-diketonate complexes such as camphor {(1R)-1,7,7- trimethylbicyclo[2.2.1]heptan-2-one} and carvone [2-methyl- 5-(1-methylethenyl)-2-cyclohexen-1-one] derived complexes synthesized by Fdil et al.,11 in agreement with their results. Fig. 1 Generation of CO2 in the epoxidation of S-limonene. Reaction conditions: (a) 0.01 mmol [Ni(acac)2] in CH2Cl2 (5 cm3); (b) 0.01 mmol [Ni(acac)2] and 3 mmol isobutyraldehyde in CH2Cl2 (5 cm3); (c) 1 mmol S-limonene, 3 mmol isobutyraldehyde, 0.01 mmol [Ni(acac)2], CH2Cl2 (5 cm3), 1 atm O2, 25 8C Kinetics The order in substrate concentration was determined for a-pinene in dichloromethane at 25 8C using complex 1c as the catalyst and isobutyraldehyde as the coreactant.For aldehyde : alkene ratios >2 : 1 the order was zero. Below this ratio the reaction was first order. The order in aldehyde concentration was calculated from the initial epoxidation rates measured at various aldehyde concentrations (0 to 0.6 mol l21 isobutyraldehyde) and the results are shown in Fig. 2. When the concentration of aldehyde was under 2.0 mol equivalents with respect to the substrate alkene little epoxide was formed. When it was equal to or greater than this a first-order dependence on the aldehyde was found. These results were observed regardless of the concentrations of the reactants.Therefore, it appears that the aldehyde must be present in the reaction mixture at a concentration of approximately twice that of the substrate for epoxidation to occur. The origin of this eVect is not yet clear. The mechanism outlined in Scheme 4 does not, for instance, explain the need for ca. 2 equivalents of aldehyde in the reaction. A tentative explanation might be the following. For epoxidation to occur the alkene must be close to the active epoxidizing species, which we propose to be the acylperoxy metal complex. It is conceivable that the alkene first coordinates to the nickel center.In that case its position will be trans to that of the co-ordinated peroxy radical. When enough acid has been formed from the aldehyde one of the acetylacetonate ligands of the acylperoxy–nickel–alkene complex can dissociate allowing the alkene to move cis to the active oxidizing species. This displacement will not be possible when not enough acid formed by autoxidation of the aldehyde is present, and thus will not take place at low aldehyde concentrations.For a Fig. 2 EVect of aldehyde concentration on the initial epoxidation rate under standard conditions Table 3 Stereoselectivity of the epoxidation of cis-stilbene with diVerent oxidants and nickel complexes a Entry 12 3 4567 Catalyst (%) Complex 1c Complex 1c and pyridine Complex 1c and m-ClC6H4CO3Hb m-ClC6H4CO3Hc [Ni(salophen)] d [Ni(tpp)] e [Ni(O2CMe)2] Conversion (%) 40 47 63 60 70 80 71 Epoxide (%) 36 46 65 60 53 68 69 cis : trans ratio 1:13 1:45 3:1 100% cis 1:10 1:12 1:22 a Reaction conditions as in Table 1.b 0.1 mol l21 m-ClC6H4CO3H under N2. c 0.1 mol l21 m-ClC6H4CO3H, no catalyst. d H2 salophen = N,N9-Bis(salicylidene)-o-phenylenediamine. e H2tpp = 5,10,15,20-Tetraphenylporphyrin.2244 J. Chem. Soc., Dalton Trans., 1998, Pages 2241–2246 related situation involving the dissociation of an acetylacetonate ligand from [Ni(acac)2] see ref. 23. The UV/VIS experiments revealed that isobutyraldehyde binds to the nickel center of complex 1c as could be concluded from the change from pink to green and the disappearance of the broad band at 520 nm in the spectrum of 1c. From a UV/VIS titration experiment the binding constant of the 1 : 1 1c–isobutyraldehyde complex was calculated to be Kb = 0.68 ± 0.08 l mol21. The epoxide reaction could be inhibited by adding competing Lewis bases to the reaction mixture. Reactions carried out in the presence of pyridine (a strong Lewis base relative to isobutyraldehyde) slowed the reaction and changed the order in substrate concentration from one to zero if the aldehyde : alkene ratio was 2 : 1.This suggests that coordination of the aldehyde to the nickel complex is an important step in the reaction sequence. The concentration of the catalyst was varied and the eVect on the initial rate was determined, as shown in Fig. 3. The rate increases linearly with the concentration of catalyst up until a concentration of 5.0 × 1024 mol l21.At higher than 10.0 × 1024 mol l21 (which is 1 mol% with respect to the alkene concentration) the rate decreased dramatically. Since this catalyst (1c) can only exist in the monomeric form under the reaction conditions used,24 aggregation of the nickel complex into a trimer (as in the case of 1a) is not a factor in the decrease in activity. It is possible that an increased amount of nickel catalyst acts as a radical trapping compound (see for example Nam et al.16), which would inhibit the reaction by preventing the formation of the nickel–peroxo radical species or converting it into a nonradical peroxy anion.The epoxidation reaction was investigated at two diVerent oxygen pressures, viz. 0.21 and 1.0 atm. No eVect on the selectivity for epoxide was found. At both oxygen concentrations the reaction followed zero-order kinetics in substrate concentration with rate constants k0 = 1.73 × 1025 and 6.32 × 1025 mol l21 s21, respectively.Without O2 the reaction did not proceed. Based on these data it can be concluded that the reaction is approximately first order in oxygen concentration. The epoxidation of a-pinene was carried out at various temperatures between 18 and 32 8C and found to follow Arrhenius behavior, with an activation energy Ea = 48 ± 6 kJ mol21. Using the Eyring relationship the parameters DH‡ and DS‡ were calculated to be 46 ± 6 kJ mol21 and 2116 ± 6 J K21 mol21, respectively.The large negative value of DS‡ points to a rigid transition state for the rate-determining step. Probes for a radical reaction mechanism When a radical trapping compound such as 2-tert-butyl-4- NiII O NiI OCR H NiII NiI OCOO• R H+ RCHO O2 * H+ RCO2H OCR • Scheme 4 Proposed mechanism for the epoxidation of alkenes with O2 and an aldehyde, catalysed by nickel(II)–b-diketonate complexes; * = rate-determining step methylphenol was added to the reaction mixture during the reaction epoxidation stopped immediately.When it was added at the beginning no epoxide was formed. These results indicate that the formation of a radical species in the reaction mixture is crucial for epoxidation to occur. Furthermore, in the presence of a radical inhibitor, no conversion of the substrate into other oxidation products was observed. When cyclobutanol was used as the substrate in a reaction with complex 1c as catalyst and isobutyraldehyde as reductant only 4-hydroxybutyraldehyde was produced, indicating an oxidizing species of a radical nature as opposed to a twoelectron oxidant.25 Kaneda et al.12 have found conditions (using temperatures slightly above ambient) under which a metal catalyst is not necessary in the epoxidation of alkenes with O2 and aldehyde.Initiation in the absence of metal complex can only take place by light. We investigated these conditions and found that the observations of Kaneda were not valid for our reactions.The epoxidation under standard conditions (see Experimental section) did not proceed without a catalyst, even at higher temperatures (80 8C). In neat (freshly distilled) isobutyraldehyde the reaction did proceed but slower than with nickel catalyst present. For comparison reactions with conventional initiators were also performed. The exceptionally reactive radical initiator di-tert-butyl peroxalate 26 was tested in the epoxidation of S-limonene without catalyst under otherwise standard conditions.‡ The reaction proceeded much more slowly than with 1a as the catalyst.The product distribution was identical, i.e. the cis : trans ratio of the epoxide was in both cases 2 : 3. The selectivity for epoxide was, however, much lower: 67% as opposed to 93% when 1a was used. The epoxidation reaction did not proceed when either an initiator or a catalyst was present, although it was determined by iodide–thiosulfate titration 27 that the aldehyde used contained a small amount of peroxide (ca. 0.5%). EPR studies using a spin trap Electron spin spectroscopy using ‘spin traps’ can provide evidence for the presence of radicals in a reaction. We investigated the epoxidation of diVerent alkenes by the nickel complexes 1a and 1c with the radical trap 5,5-dimethyl-1-pyrroline N-oxide 2.28 Fig. 3 Influence of the catalyst concentration on the epoxidation rate under standard conditions ‡ Di-tert-butyl peroxalate was synthesized according to a literature procedure. 26 For use in the epoxidation reaction a 16.4 mmol l21 solution in CH2Cl2 was prepared, and 5 cm3 of it was used as the solvent for the reaction. The amount of peroxide present was checked by iodide– thiosulfate titration.27 Thus, in the reaction mixture 1.5% of peroxide with respect to the alkene was present.J. Chem. Soc., Dalton Trans., 1998, Pages 2241–2246 2245 Table 4 The EPR hyperfine splitting constants from spin-trap experiments with complexes 1a and 1c and compound 2 in CH2Cl2 at 25 8C Entry 12345678 Sample No substrate a a-Pinene a Stilbene a S-Limoneneb Isobutyraldehyde Yes Yes Yes No No Yes Yes No O2 present Yes No Yes Yes Yes Yes No Yes aN/G 12.7 13.35 ± 0.17 13.3 13.1 No signal 12.69 ± 0.35 13.30 ± 0.20 No signal aHb/G 9.8 7.37 ± 0.12 8.7 8.4 8.03 ± 0.37 7.16 ± 0.12 aHg/G 0.3 1.5 0.3 ± 0.1 a Catalyst 1c.b Catalyst 1a. A solution containing compound 2 and complex 1a (MII = NiII) in dichloromethane was prepared.This did not give any signal (other than that of the cavity). To it were added the components of the reaction, viz. isobutyraldehyde and/or alkene substrate, so that the concentrations were identical to those of the standard reaction mixture (see Experimental section). No signal or only very broad signals were observed in a glass at 15 K. At 298 K signals were found when at least 2, NiII and aldehyde or alkene were present. No signals were observed without 2, strongly suggesting the presence of spin adducts 3 in the EPR-active samples.The nitrogen and hydrogen hyperfine splittings aN and aH observed for these samples are summarized in Table 4. Comparison of the aN and aH values with those reported in the literature 28 with benzene as the solvent allows a tentative identification of some of the spin adducts and therefore of the radicals trapped. Keeping the diVerence in solvent in mind, the resemblance of the set of hyperfine splittings of the signals in entry 1 to those reported for a benzoyloxy radical adduct with 2 in benzene (aN and aHb of approximately 12.24 and 9.63 G, respectively) points to the trapping of an acyloxy radical 28 in our case.Our trapping experiment cannot distinguish between an adduct derived from an acyloxy and an alkoxy radical. Alkoxy radicals are reported to have an aN value of 13–13.5 G and an aHa value of 7–8 G.28 The observed splitting parameters in entries 2, 3, 6 and 7 are very similar to these values, even without any O2 present (entries 2 and 7), indicating the trapping of a radical derived from aldehyde, as an O- rather than a C-centered radical.The hyperfine splitting constants of the EPR signals of the radical adducts in entries 1, 3, 4 and 6 are suYciently diVerent from each other to exclude the possibility that the same radical, derived, for example, from the acetylacetonate ligand, is trapped in these cases.The radical in entry 1 is likely to come from the isobutyraldehyde, i.e. it is probably the trapped acyl radical or, more likely, the trapped acylperoxy radical. The signals in entries 3, 4 and 6 might result from the alkene. In the hematin/cumene hydroperoxide system reported previously 29 no radicals were trapped; all the EPR signals observed were due only to oxidation of the spin trap, resulting in 5,5-dimethylpyrrolidin-2-one N-oxyl, 4.30 The possibility that this compound was formed in our system could be ruled out, as the aN and aH values reported for this radical in solvents related in polarity to dichloromethane are diVerent from those observed by us (benzene, aN = 6.45 and aH = 3.28; chloroform, aN = 6.58 and aH = 3.60 G 30).It should be mentioned here that it was recently shown31 that 4 is always formed when oxometal complexes are present as intermediate species, e.g. in the case of Mn. The absence of 4 in our experiments provides additional evidence against an oxometal epoxidation mechanism as shown in Scheme 3.+ N Me Me O – H N Me Me O• O N Me Me O• Hb R Hg Hg 2 3 4 Discussion From our results, and from those in the literature, we may conclude that the [Ni(acac)2]–isobutyraldehyde–alkene system is a very useful catalytic system for the epoxidation of alkenes. The mechanism of the epoxidation reaction seems to be diVerent from that of other epoxidations, such as peroxide-initiated reactions or reactions catalysed by other transition-metal complexes (e.g. salen or cyclam-type complexes).First, an oxometal mechanism (Scheme 3) can probably be ruled out. No spectroscopic evidence was found for oxometal complexes in our reactions, and b-diketonate complexes do not form oxo-complexes easily. There is no evidence that NiII, when co-ordinated by oxygen ligands, is capable of directly binding molecular oxygen and forming an oxonickel species, based on the extensive studies carried out in this area.14,32 In addition, as reported by Nam et al.,15 the epoxidation reaction can proceed when it is initiated with a perester giving the same product distributions as the nickel(II)-catalysed reaction.If an oxometal species were to play an important role in the epoxidation reaction the perester-initiated reaction would most probably yield a diVerent product distribution. Finally, no 4 was detected by EPR spectroscopy which can be taken as additional evidence that no oxometal complexes were generated in solution.A second possibility is a mechanism in which peracid is formed in situ, which then oxidizes the alkene, as shown in Scheme 2. However, this would be at variance with the reactivity of the aldehydes as observed in our investigations as well as in those of Mukaiyama and co-workers.2–7 Benzaldehyde, which has been reported to be easily converted into its corresponding peroxyacid in the presence of a number of transition metals,33–37 is unreactive under the reaction conditions described here.Examination of the stereochemistry of the products (see Table 3) provides even stronger evidence discounting an in situ formed peroxyacid as the active epoxidizing agent. Peroxyacids normally carry out the epoxidation of alkenes by a concerted oxygen-transfer step,33,34,38 so that the stereochemistry of the substrate is conserved in the product. We found that with cisstilbene the opposite occurs: trans-stilbene oxide is formed almost exclusively.Thus peroxyacid epoxidation is not a major oxidation pathway in the system we studied, though it might very well be in the system Kaneda et al.12 investigated. Their results regarding the stereochemistry and kinetics of the reaction are reminiscent of an autoxidation process. The combined results of our kinetic and mechanistic studies lead us to conclude that the epoxidation reaction is radical in nature with the formation of a nickel-bound acyl radical as the first step and the formation of a nickel–acylperoxy intermediate, which might be cyclic for stability reasons, as the second important step.When compared to the literature (Nam et al.15), the most compelling, new evidence for a radical nature of the reaction intermediates was found with EPR spectroscopy. Our investigations pointed to the presence of two radical species. One is formed when the catalyst and isobutyraldehyde are present but oxygen is absent. We propose that this is the trapped acyl radical bound to the nickel center of the catalyst.The second radical species is observed when oxygen is added to the2246 J. Chem. Soc., Dalton Trans., 1998, Pages 2241–2246 reaction mixture; this may be the nickel–acylperoxy radical. Both radicals were trapped with 2 as oxygen-centered radicals. Based on the data presented in the Results section the rate equation (1) holds, provided that the molar ratio of aldehyde to r = kobs[O2][Sub]0[RCHO][Cat] (1) substrate is greater than 2 : 1.Here, kobs is the observed rate constant, [Sub] the concentration of alkene substrate, [RCHO] the concentration of isobutyraldehyde and [Cat] the concentration of nickel(II) complex. The reaction is first order in substrate concentration when the ratio of aldehyde to substrate is equal to or less than 2 : 1, changing to zero order when a Lewis base (pyridine) is added. The overall rate law that usually applies to the autoxidation of aldehydes is given by equation (2),39,40 provided that the r = k S ki 2kt D� �� [In]� �� [RCHO] (2) oxygen and aldehyde concentrations are suYciently high.Here k is the rate constant of the rate-limiting propagation reaction, ki that of the initiation reaction, kt that of the termination reaction and [In] is the initiator concentration. The rate-limiting step is hydrogen abstraction from the aldehyde. We assume that our epoxidation reaction (Scheme 1) is suYciently fast as not to interfere with the autoxidation reaction.Our results do not conform to the rate law (2), because the order in catalyst concentration (assuming that its only role is initiating the reaction) is 1 instead of ��� . The rate law (2) depends on the type of termination step that is operative which may explain the dependence on oxygen pressure 40 of our reaction. The role of our metal catalyst is likely to be more than just an initiator of the reaction. If a radical chain mechanism, initiated by the nickel(II) complex, takes place an order of ��� is expected, according to equation (3) where [Ni(acac)2] is taken as an example. [NiII(acac)2] æÆ [NiI(acac)] 1 acac? (3) The mechanism we propose for the epoxidation of substituted alkenes with molecular oxygen and isobutyraldehyde, catalysed by nickel(II)–b-diketonate complexes, is a catalytic cycle in which the active oxidizing species is an acylperoxy radical which stays bound to the metal complex for stabilization (Scheme 4).Based on the observed rate equation, we may tentatively conclude that the rate-limiting step is the formation of the nickel acylperoxy species, formally a nickel(I) species (see Scheme 4). The products evolving from the aldehyde are small amounts of carboxylic acid and CO2 (both ca. 10% with respect to converted aldehyde). Other aldehyde oxidation products could not be identified. It should be noted that Scheme 4 probably is a simplified picture of the actual prounravel the details of the reaction are currently in progress. The role of the metal complex is to promote both hydrogen abstraction from the aldehyde and acceleration of the oxidation reaction. The nickel(II) complex is proposed to take up an electron from the co-ordinated aldehyde which then loses a proton. Alternatively, one could imagine the formation of a nickel(III) hydride complex. These possibilities are currently under investigation.Conclusion We have proposed further details of the mechanism for the epoxidation of alkenes with molecular complex and an aldehyde, catalysed by b-diketonate complexes of NiII. It is shown that the mechanism is radical in nature, with the metal complex acting as an initiator of the reaction and a promoter of the oxidation. Our mechanism is in this respect diVerent from that of Mukaiyama and others. The catalytic system is very useful to prepare a variety of multiply substituted epoxides from alkenes under mild conditions.It is not yet clear what the fate of the aldehyde is, since neither large amounts of carboxylic acid nor of CO2 could be detected. The proposed mechanism is consistent with the kinetic and EPR data and with the observed stereochemistry of the reaction. Acknowledgements The authors would like to thank Gerrit Jansen and Dr. Paul van Kan for performing the EPR measurements. This research is financed by the Innovation Oriented research Programmes of the Ministry of Economic AVairs (project nos.IKA 90037 and 94025). R. J. M. N. and M. C. F. thank B. B. W. and P. A. G. for their equal contribution to this paper. References 1 See Proc. 6th Int. Symp. Activation of Dioxygen, J. Mol. Catal. A, Chemical, 1997, 117. 2 T. Yamada, T. Takai, O. Rhode and T. Mukaiyama, Bull. Chem. Soc. Jpn., 1991, 64, 2109. 3 T. Takai, E. Hata, T. Yamada and T. Mukaiyama, Bull. Chem. Soc. Jpn., 1991, 64, 2513. 4 T. Yamada, K.Imagawa and T. Mukaiyama, Chem. Lett., 1991, 1. 5 T. Yamada, K. Imagawa and T. Mukaiyama, Chem. Lett., 1992, 2109. 6 T. Mukaiyama, T. Yamada, T. Nagata and K. Imagawa, Chem. Lett., 1993, 327. 7 T. Mukaiyama and T. Yamada, Bull. Chem. Soc. Jpn., 1995, 68, 17. 8 R. Irie, Y. Ito and T. Katsuki, Tetrahedron Lett., 1991, 32, 6891. 9 S. I. Murahashi, Y. Oda and T. Naota, J. Am. Chem. Soc., 1992, 114, 7913. 10 K. Yanai, R. Irie, Y. Ito and T. Katsuki, Mem. Fac. Sci., Kyushu Univ., Ser.C, 1992, 18, 213. 11 N. Fdil, A. Romane, S. Allaoud, A. Karim, Y. Castanet and A. Mortreux, J. Mol. Catal. A: Chemical, 1996, 108, 15. 12 K. Kaneda, S. Haruna, T. Imanaka and M. Hamamoto, Tetrahedron Lett., 1992, 45, 6827. 13 D. Srinivasan, P. Michaud and J. K. Kochi, J. Am. Chem. Soc., 1986, 108, 2309. 14 R. H. Holm, Chem. Rev., 1987, 87, 1401. 15 W. Nam, H. J. Kim, S. H. Kim, R. Y. N. Ho and J. S. Valentine, Inorg. Chem., 1996, 35, 1045. 16 W. Nam, S. J. Baek, K. A. Lee, B. T. Ahn, J. G. Muller, C. J. Burrows and J. S. Valentine, Inorg. Chem., 1996, 35, 6632. 17 D. P. Graddon, Coord. Chem. Rev., 1969, 4, 1. 18 J. P. Fackler, jun., Prog. Inorg. Chem., 1966, 7, 361. 19 A Textbook of Quantitative Inorganic Analysis, ed. A. I. Vogel, Longmans, New York, 3rd edn., 1961. 20 P. Laszlo and M. Levart, Tetrahedron Lett., 1993, 34, 1127. 21 K. R. Lassila, F. J. Waller, S. E. Werkheiser and A. L. Wressell, Tetrahedron Lett., 1994, 35, 8077. 22 N. Mizuno, H. Weiner and R. G. Finke, J. Mol. Catal. A: Chemical, 1996, 114, 15. 23 R. J. M. Nolte and W. Drenth, Recl. Trav. Chim. Pays-Bas, 1973, 92, 788. 24 J. P. Fackler, jun. and F. A. Cotton, J. Am. Chem. Soc., 1961, 83, 3775. 25 J. Rocek and A. E. Rodkowsky, J. Am. Chem. Soc., 1973, 95, 7123. 26 N. A. Milas and P. C. Panagiotakos, J. Am. Chem. Soc., 1946, 68, 534. 27 P. D. Bartlett, E. P. Benzing and R. E. Pincock, J. Am. Chem. Soc., 1960, 82, 1762. 28 E. G. Janzen and C. A. Evans, J. Magn. Reson., 1973, 91, 510. 29 R. A. Floyd and L. M. Soong, Biochem. Biophys. Res. Commun., 1977, 74, 79. 30 H. G. Aurich and J. Trosken, Liebigs Ann. Chem., 1971, 745, 159. 31 T. Sciarone and J. Reek, unpublished work. 32 J. S. Valentine, Chem. Rev., 1973, 73, 235. 33 D. Swern, Chem. Rev., 1949, 45, 1. 34 D. Swern, Org. React., 1953, 378. 35 W. P. Jorissen and P. A. A. van der Beek, Recl. Trav. Chim. Pays-Bas, 1926, 45, 245. 36 P. A. A. van der Beek, Recl. Trav. Chim. Pays-Bas, 1928, 47, 286. 37 T. Mlodnicka, J. Mol. Catal. A: Chemical, 1986, 36, 205. 38 D. Swern and T. W. Findley, J. Am. Chem. Soc., 1950, 72, 4315. 39 Free Radical Chain Reactions, ed. E. S. Huyser, Wiley-Interscience, New York, 1970. 40 L. Bateman, Q. Rev. Chem. Soc., 1954, 8, 147. Received 10th February 1998; Paper 8/01175C

ISSN:1477-9226    

DOI:10.1039/a801175c

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1999, 2243–2248 2243 Reactions of a dinuclear tungsten complex containing an O-co-ordinated bridging ketene with various heterocumulenes Luxti J. J. Wang, Sen-Jen You, Shou-Ling Huang, Yu-Lee Yang, Ying-Chih Lin,* Gene-Hsiang Lee and Shie-Ming Peng Department of Chemistry, National Taiwan University, Taipei, Taiwan 10764, Republic of China Received 9th March 1999, Accepted 14th May 1999 Treatment of the dinuclear O-co-ordinated ketene complex W2Cp2(CO)5(m-h1 :h2-CH2CO) (Cp = h5-C5H5) with PhCH2NCS aVorded the dark red bridging thioketene complex W2Cp2(CO)5(m-h3-SC]] CH2). Reaction of the latter with HBF4 gave the stable cationic product [W2Cp2(CO)5(m-SCMe3)]BF4 by addition of a proton to the terminal carbon of the thioketene group.In the reaction of W2Cp2(CO)5(m-h1 :h2-CH2CO) with CS2 cleavage of one CS bond is accompanied by insertion of the resulting sulfur atom into the tungsten-acyl bond to aVord W2Cp2(CO)5(CS)- (m-CH2COS).The Cp9 analogue (Cp9 = h5–C5H4Me) was also prepared. The reaction of the trithiocarbonate S]] C(SCH2)2 with W2Cp2(CO)5(m-h1 :h2-CH2CO) aVorded the carbene complex W2Cp2(CO)5[C(SCH2)2](m-CH2COS), and with allene gave the allylic complex W2Cp2(CO)5(m-h1 :h3-CH2COC3H4). Those of the complexes have been characterized by single crystal X-ray diVraction analysis. Introduction Metal complexed ketenes have been prepared by a variety of routes, including coupling of alkylidene and carbonyl moieties,1 addition of free ketene to unsaturated metal systems 2 and deprotonation of metal acyls.3 Unlike free ketenes which have been examined for nearly a century,4 metal complexed ketenes have only recently been investigated.Carbonylation of m-methylene complexes, A, not containing a metal–metal bond has been shown to yield polynuclear ketene complexes.5 In contrast, dinuclear m-methylene complexes,6 B, containing a metal–metal bond are not readily carbonylated to m-ketene complexes due to the stability of the dimetallacyclopropane skeleton. Heterobimetallic ketene complexes have been prepared from the acylation of metal anions MLn 2 by Fp-CH2COCl [Fp = FeCp(CO)2].7 We previously reported that reaction of CH2I2 with W2Cp(CO)3 1 in MeOH aVords the dinuclear tungsten complex W2Cp2(CO)5(m-h1 :h2-CH2CO) 2, which contains an O-co-ordinated ketene bridge.8 Formation of 2 is believed to proceed through carbonylation of the bimetallic methylene intermediate W2Cp2(CO)6(m-CH2).The ketene oxygen atom of complex 2 is weakly co-ordinated to the tungsten metal center and easily replaced by donor ligands L to give W2Cp2(CO)5L(m-h1 :h1-CH2CO).9 In the presence of CH3CN, 2 is converted into the acetylide W2Cp2- (CO)6(m-C]] ]C). Herein we report the reactions of 2 with various heterocumulenes. M M M M A B W CCOO CO O W CC O O 2 W CCOO CO W CC O O W CCOO CO CH2I2 CH3OH Results and discussion Reactions of complex 2 with isothiocyanate Treatment of the O-co-ordinated ketene bridged complex 2 with PhCH2NCS in CH2Cl2 at room temperature for 30 min aVords a dark red C,S-co-ordinated thioketene complex W2Cp2(CO)5(m-SC]] CH2) 3 in moderate yield (Scheme 1).9 The thioketene bridges the two metal centers in a m-h1 :h2- bonding mode with the CS portion behaving as a four-electrondonor ligand. Use of other organic isothiocyanates in this reaction led to the same product. The Cp9 analogue 39 (Cp9 = C5H4Me) was also prepared.In the 1H NMR spectrum of 3 two doublet resonances at d 5.93 and 6.32 with JH-H = 0.88 Hz are assigned to the terminal protons of the thioketene ligand. These resonances are downfield relative to those of the ketene ligand of 2 (d 3.41) and are characteristic of olefinic ]] CH2 methylene protons. The two 13C resonances of the thioketene ligand occur at d 117.9 (terminal carbon) and 166.5, and show 1JC-H and 2JC-H coupling with the olefinic protons; assignments were made via two dimensional heteronuclear multiple quantum and multiple bond correlation (HMQC and HMBC) NMR techniques.10 Scheme 1 W C C O O C O O W C C O O W C C O C O W C CO S O O RNCS W C C O O C O S W C C O O O N R W C C O O C O W CC O O S W C C O C O W C C O S O O H W C C O O C O O W CC O O S N R –RNCO Et3N 2 3 42244 J.Chem. Soc., Dalton Trans., 1999, 2243–2248 In order to establish the structure, complex 3 was characterized by an X-ray diVraction analysis.An ORTEP drawing is shown in Fig. 1 and selected bond distances and angles are listed in Table 1. The thioketene ligand bridges the two metal centers with the sulfur atom in a bonding mode that diVers from that seen in the O-co-ordinated ketene complex 2. This may be attributed to higher aYnity of tungsten for sulfur. The W(1)–S and W(2)–S bond distances are 2.530(6) and 2.466(6) Å, respectively, and the W(2)–C(6) distance is 2.07(2) Å. The C(6)–C(7) bond distance (1.31(4) Å) is typical of a carbon– carbon double bond and is consistent with the NMR data. This type of co-ordination diVers from that of the bimetallic S-bridging thioketene complex Mo2Cp2(CO)4(m-SC]] CR2) which contains a metal–metal bond.This complex was prepared from reaction of cyclohexene sulfide with the bridging vinylidene complex Mo2Cp2(CO)4(m-C]] CR2).12 No structure data were provided for it. Unlike the weakly co-ordinated oxygen in 2, the bridging thio ligand in 3 is much more strongly co-ordinated to the tungsten metal centers.No reaction was observed between 3 and PPh3. It has been reported that the reaction of WCp(CO)3H with S(NMe2)2 aVords the dinuclear S-bridging tungsten complex [WCp(CO)3]2(m-S) which, upon reacting with CH2N2, gives the thioformaldehyde complex [WCp(CO)2]2(m-CH2S).13 Conversion of complex 2 into 3 involves substitution of the ketene oxygen with sulfur and a change of co-ordination mode. A possible mechanism for the formation of 3 is depicted in Scheme 1.The reaction is suggested to proceed via a [212] cycloaddition of the CS group of the RNCS with the CO unit with concomitant generation of RNCO and/or (RNH)2CO. It is less likely that the thioisocyanate co-ordinates to the metal center followed by migration of S to ketene because the reac- Fig. 1 An ORTEP11 drawing of W2Cp2(CO)5(m-CH2CS) 3. Table 1 Bond distances (Å) and bond angles (8) of complex 3 W(1)–S W(1)–C(1) W(1)–C(2) W(1)–C(3) W(2)–S W(2)–C(4) W(2)–C(5) W(2)–C(6) S–W(1)–C(1) S–W(1)–C(2) S–W(1)–C(3) C(1)–W(1)–C(2) C(1)–W(1)–C(3) C(2)–W(1)–C(3) S–W(2)–C(4) S–W(2)–C(5) S–W(2)–C(6) C(4)–W(2)–C(5) C(4)–W(2)–C(6) C(5)–W(2)–C(6) 2.530(6) 1.97(3) 2.061(25) 1.995(23) 2.466(6) 1.915(24) 2.01(3) 2.072(23) 132.9(6) 77.6(7) 75.1(6) 80.0(11) 75.5(9) 111.4(11) 90.3(7) 116.0(9) 45.5(7) 76.5(14) 114.5(9) 84.0(14) S–C(6) C(1)–O(1) C(2)–O(2) C(3)–O(3) C(4)–O(4) C(5)–O(5) C(6)–C(7) W(1)–S–W(2) W(1)–S–C(6) W(2)–S–C(6) W(1)–C(1)–O(1) W(1)–C(2)–O(2) W(1)–C(3)–O(3) W(2)–C(4)–O(4) W(2)–C(5)–O(5) W(2)–C(6)–S W(2)–C(6)–C(7) S–C(6)–C(7) 1.793(24) 1.15(3) 1.10(3) 1.12(3) 1.17(3) 1.10(3) 1.31(4) 126.82(25) 115.5(7) 55.6(8) 177.7(20) 172.8(23) 178.6(20) 176.7(20) 172(4) 78.9(8) 158.4(21) 122.6(21) tion of CS2 with 2 (described below) does not involve loss of the oxygen atom.Protonation of complex 3 Protonation of complex 3 with HBF4 takes place at the terminal carbon of the thioketene ligand aVording [W2Cp2- (CO)5(m-SCMe3)]BF4 4.The IR spectrum of 4 displays absorption bands at 2048, 2029, 1968 and 1944 cm21, much higher than those (2036, 1947, 1916 and 1813 cm21) of the neutral complex 3, indicating a cationic character. In the 1H NMR spectrum of 4 three singlet resonances at d 5.95, 5.93 and 2.99 are assigned to the two cyclopentadienyl and the methyl groups, respectively. In the 13C NMR spectrum, a downfield resonance at d 256.6 is attributed to the carbene (or carbocation) carbon center of the CS group.The FAB mass spectrum shows parent peaks assignable to the cation at m/z = 697. Protonation of 3 with CF3CO2H gave the product [W2Cp2(CO)5(m-SC(O2CCF3)- Me3)] which displayed a parent peak at m/z = 810 in its FAB mass spectrum indicating that the trifluoroacetate moiety is bound. The protonation reaction is reversible; addition of triethylamine converts 4 into 3. Protonation of the mononuclear cobalt complex CoCp(PMe3)(h2-SC]] CR2) also occurs at the terminal carbon of the h2-co-ordinated thioketene ligand to give a thioacyl complex.9a Reaction of complex 2 with CS2 Treatment of a red dichloromethane solution of complex 2 with an excess of CS2 for 30 min at room temperature aVorded a scarlet red solution from which the dinuclear thiocarbonyl complex W2Cp2(CO)5(CS)(m-CH2COS) 5 was isolated in 93% yield.The product is air stable and can be stored at room temperature. The 1H NMR analysis of the crude reaction mixture indicated that there were two products in a ratio of 4 : 1.In the spectrum of the crude product the two Cp resonances at d 5.79, 5.54 and the singlet methylene resonance at d 2.86 with a pair of tungsten satellites (JW-H = 5.2 Hz) are assigned to the major product. Two singlet resonances at d 5.71, 5.57 and resonances with an AB pattern centered at d 2.91 are assigned to the minor product. Both sets display a relative intensity ratio of 5:5:2. The resonances at d 2.91 assignable to the diastereotopic CH2 group of the minor product imply asymmetry at the neighboring metal center.The singlet resonance at d 2.86 assignable to the CH2 group of the major product indicates the trans disposition of the CS ligand. In the 13C NMR spectrum the resonances at d 1.5 and 0.9 are attributed to the methylene carbon atoms of the two isomers and those at d 228.8 and 228.9 are assigned to the bridging COS groups. The two isomers are inseparable by recrystallization and the FAB mass spectrum of the mixture gave a parent peak at m/z = 756 with fragmentation peaks due to consecutive losses of CO groups.The Cp9 (h5– C5H4Me) analogue of 5 was also prepared as a mixture of isomers with the same ratio (4 : 1). In the 13C NMR spectrum two downfield resonances at d 348.2 and 332.8 assignable to the CS ligands are comparable to that of many thiocarbonyl complexes. 14 On the basis of these spectroscopic data, we conclude that the two products are a mixture of cis and trans isomers.Single crystals of complex 59 were grown by careful addition of hexane to a CH2Cl2 solution. An X-ray diVraction study gave the structure shown in Fig. 2. Selected bond distances and angles are listed in Table 2. Co-crystallization of the cis and trans isomers causes disorder of the CS and CO ligands. The two W atoms are connected by a m-CH2COS bridge with one W atom bound to the methylene carbon and the other W atom bound to the S atom.The W(1)–C(7) and W(2)–S(1) bond distances are 2.32(2) and 2.489(4) Å, respectively, with the W(1)– C(7)–C(8) and W(2)–S(1)–C(8) bond angles being 114.8(10) and 107.0(5)8, respectively. All the bond distances and angles of the CH2COS bridge are normal. The thiocarbonyl CS ligand is bound to W(2), which is also bonded to the S(1) atom of theJ. Chem. Soc., Dalton Trans., 1999, 2243–2248 2245 SC(O)CH2 bridge. The cis : trans ratio is 33 : 67 for 59 in crystal form, slightly diVerent from that observed by NMR (20: 80).This assignment gives reasonable thermal parameters and bond distances. The two metal centers exist as mutually independent mononuclear states and no evidence for metal–metal interaction is detected. Complex 5 is stable in CDCl3 and C6D6, even at the refluxing temperatures. The CS ligand is not replaced under 1 atm of CO pressure. However, thermolysis of 5 in MeCN caused cleavage of the C–S bond of the bridging ligand to give the mononuclear complex WCp(CO)3(CH2CO2H) in high yield.9 Nucleophilic attack of water at the central carbon atom of the Me2COS bridge took place in the presence of trace water in MeCN.The other half of 5, namely Cp(CO)2(CS)WS, decomposed at the refluxing temperature of MeCN to give an unidentified mixture. Reactivity of 59 is similar to that of 5. Reaction of complex 2 with trithiocarbonate Treatment of complex 2 with trithiocarbonate S]] C(SCH2)2 aVorded the dithiocarbene complex 15 W2Cp2(CO)5[C(SCH2)2]- (m-CH2COS) 6 in 61% yield.In the 13C NMR spectrum a resonance at d 266.3 assignable to the dithiocarbene carbon atom is observed far downfield from the resonance (d 228.0) of the corresponding carbon atom in free ethylene trithiocarbonate, and the resonance at d 229.6 assignable to the bridging COS functionality is consistent with that (d 228.8 and 228.9) of 5. The assignment was confirmed by a 2-D HMBC NMR experiment in which long range 3JC-H coupling was seen between the resonances at d 266.3 (13C) and 3.38 (1H NMR) Fig. 2 An ORTEP drawing of W2Cp92(CO)5(CS)(m-CH2COS) 59. Table 2 Selected bond distances (Å) and angles (8) of complex 59 W(1)–C(1) W(1)–C(2) W(1)–C(3) W(1)–C(7) W(2)–S(1) W(2)–C(4) W(2)–C(5) W(2)–C(6) S(1)–C(8) C(1)–W(1)–C(2) C(1)–W(1)–C(3) C(1)–W(1)–C(7) C(2)–W(1)–C(3) C(2)–W(1)–C(7) C(3)–W(1)–C(7) S(1)–W(2)–C(4) S(1)–W(2)–C(5) S(1)–W(2)–C(6) C(4)–W(2)–C(5) C(4)–W(2)–C(6) C(5)–W(2)–C(6) 1.97(2) 1.98(2) 2.00(2) 2.32(2) 2.489(4) 2.03(2) 1.97(2) 2.00(2) 1.820(14) 78.7(8) 74.7(7) 131.4(7) 108.8(8) 73.7(7) 77.5(7) 76.6(5) 76.1(5) 131.8(4) 108.3(7) 76.2(6) 75.8(7) S(29)–C(5) a S(2)–C(6) C(1)–O(1) C(2)–O(2) C(3)–O(3) C(4)–O(4) C(7)–C(8) C(8)–O(7) W(2)–S(1)–C(8) W(1)–C(1)–O(1) W(1)–C(2)–O(2) W(1)–C(3)–O(3) W(2)–C(4)–O(4) W(2)–C(5)–S(29) W(2)–C(6)–S(2) W(1)–C(7)–C(8) S(1)–C(8)–C(7) S(1)–C(8)–O(7) C(7)–C(8)–O(7) 1.58(3) 1.49(2) 1.16(2) 1.13(2) 1.13(2) 1.13(2) 1.49(2) 1.20(2) 107.0(5) 177(2) 179(2) 179.4(13) 176(2) 172(2) 172.6(13) 114.8(10) 111.2(12) 122.6(11) 126.1(14) a S(2) is 67% S and 33% O and S(29) is 33% S and 67% O.for the M]] C(SCH2)2 ligand and between those at d 229.6 (13C) and 2.83 (1H) for the CH2COS group. The FAB mass spectrum of 6 displays a parent peak at m/z = 817 as well as fragmentations attributed to successive losses of three CO ligands. We believe that opening of the weakly co-ordinated ketene oxygen atom along with a p co-ordination of the S]] C bond of CS2 or S]] C(SCH2)2 to form an adduct (Scheme 2) may occur in the first stage, in analogy to the case of phosphine addition.9 Subsequent cleavage of the co-ordinated C=S bond along with co-ordination of the CS moiety (as in the reaction of CS2) or of the carbene unit (in the reaction of trithiocarbonate) and insertion of an S atom into the W–O bond accounts for the product.The h2-CS2 adduct has been reported for CoCp- (PMe3)(h2-CS2) 16 and several platinum complexes.17 Cleavage of one of the C]] S bonds of CS2 either by a metal cluster or in the process of forming a metal cluster has been previously observed.18 In the reaction of Os3(CO)12 with CS2 the thiocarbonyl cluster Os3(CO)10(CS)(S) was obtained.19 Thermolysis of CoCp(CO)2 in the presence of CS2 aVords the thiocarbonyl sulfide cluster Co3Cp3(m3-CS)(m3-S).20 While there are previous examples of C]] S bond cleavage, S insertion into the metal– oxygen bond is the first example of this type.21 Carbon disulfide is an unsaturated electrophile with an extensive organic and organometallic chemistry.22 Typically, it reacts with metal alkyls or hydrides by insertion, forming dithiocarboxylate or dithioformate complexes.Reaction of complex 2 with allene Reaction of gaseous allene with complex 2 readily occurred at room temperature to yield the dinuclear allylic complex W2Cp2(CO)5[m-h1 :h3-CH2COC(CH2)2] 7 in 92% yield. The reaction likely follows the same route as do other donors reacting with 2 to form, in this case, a p allene-acyl intermediate.Subsequent coupling of the allene and acyl ligands gives the observed product. Coupling of the co-ordinated s-allenyl group with another ligand co-ordinated to the same metal generally occurs at the a-carbon atom but coupling of the p-allene ligand occurs at the central carbon atom.23 This may be due to the proximity of the ligands. At room temperature complex 7 displays fluxionality on the allylic part of the molecule, namely the allylic ligand undergoes interconversion between endo and exo forms.Thus, in the 13C NMR spectrum obtained at room temperature, sharp reson- Scheme 2 W C C O O C O O W C O OC W C C O O C O O W C O O C W C C O O C O O W C O OC S S CS2 W C C O O C O O W C C O O C S S W C C O O C O O W C O OC S S S S S S H2CCCH2 2 7 6 5 ( cis + trans) 22246 J. Chem. Soc., Dalton Trans., 1999, 2243–2248 ances at d 92.4 (Cp) and 211.0 (CH2) are assigned to the Cp(CO)3WCH2 part of 7, and broad resonances at d 87.8 and 25.7 are assigned to the Cp(CO)2W(allyl) part.In the 1H NMR spectrum the resonance of the methylene group of the ketene unit appears at d 2.25 and those of the syn and anti protons of the allylic group at d 3.08 and 1.58, respectively. These assignments have been confirmed by two-dimensional HMBC and HMQC NMR experiments. The 1H NMR spectrum of 7 at 210 8C displays resonances at d 2.99, 2.18 and 1.60, assignable to the allylic anti proton, WCH2 and allylic syn proton, respectively, of the endo isomer.The corresponding resonances for the exo isomer appear at d 3.10, 2.10 and 1.15. The endo : exo ratio is ca. 10 : 1. In the 13C NMR spectrum the broad Cp resonance at d 87.8 resolves into two sharp resonances also with a ratio of 10 : 1. Since 7 is a b-substituted allylic complex, the major isomer is expected to have an endo conformation. No attempt was made to assign the 13C resonances of the minor isomer.The fluxionality is similar to that seen in W2Cp2(CO)5[m-h1 :h2- COC(CH2)2] prepared from the reaction of W2Cp(CO)3 with WCp(CO)2(C(CH2)2COCl).24 Suitable crystals of complex 7 for X-ray diVraction analysis were obtained by recrystallization from hexane, and an ORTEP drawing of the molecule is shown in Fig. 3. Selected bond distances and angles are listed in Table 3. The two metal centers are bridged by a b-substituted allylic unit with the allylic unit in an endo conformation.The three W(1)–C (allyl) bond distances (2.29(1), 2.27(1) and 2.29(1) Å, for W(1)–C(6), W(1)–C(7) and W(1)–C(8), respectively) and the W(2)–C(10) bond distance (2.35(1) Å) are in keeping with the literature for W–h3-C3H5 allyl derivatives and W–C single bonds, respectively. The W–CO bond lengths and the W–Cp distances are all normal for both metal centers. The allyl unit is approximately coplanar with the neighboring carbonyl group.In the presence of Fe(CO)5, allenes and aldehydes give substi- Fig. 3 An ORTEP drawing of W2Cp2(CO)(m-CH2COC(CH2)2) 7. Table 3 Selected bond distances (Å) and angles (8) of complex 7 W(1)–C(1) W(1)–C(2) W(1)–C(6) W(1)–C(7) W(1)–C(8) W(2)–C(3) W(2)–C(4) W(2)–C(5) W(2)–C(10) C(1)–O(1) C(1)–W(1)–C(2) C(3)–W(2)–C(4) C(3)–W(2)–C(5) C(3)–W(2)–C(10) C(4)–W(2)–C(5) C(4)–W(2)–C(10) C(5)–W(2)–C(10) W(1)–C(1)–O(1) W(1)–C(2)–O(2) W(2)–C(3)–O(3) 1.946(13) 1.955(13) 2.291(13) 2.270(10) 2.289(12) 1.974(13) 1.957(14) 1.921(13) 2.346(12) 1.158(17) 75.9(6) 75.1(6) 109.7(6) 82.1(5) 76.9(6) 134.1(5) 74.1(6) 176.5(10) 177.8(10) 175.0(10) C(2)–O(2) C(3)–O(3) C(4)–O(4) C(5)–O(5) C(6)–C(7) C(7)–C(8) C(7)–C(9) C(9)–C(10) C(9)–O(6) W(2)–C(4)–O(4) W(2)–C(5)–O(5) C(6)–C(7)–C(8) C(6)–C(7)–C(9) C(8)–C(7)–C(9) C(7)–C(9)–C(10) C(7)–C(9)–O(6) C(10)–C(9)–O(6) W(2)–C(10)–C(9) 1.161(16) 1.145(16) 1.145(17) 1.182(17) 1.401(18) 1.417(20) 1.536(17) 1.446(20) 1.220(16) 178.5(14) 177.3(11) 116.2(11) 124.4(12) 119.0(12) 117.8(11) 118.2(12) 123.9(11) 111.6(8) tuted trimethylenemethane complexes whose formation has been proposed to proceed via coupling of the carbonyl carbon of the aldehyde with the central carbon of the allene followed by elimination of CO2.25 Recently, migratory insertion of allene into alkyl and acetyl palladium complexes leading to stable h3-allylic compounds has been reported.26 The acetyl group gives a rapid allene insertion, but the alkyl ligand requires a poorly co-ordinating ligand such as BF4 to give the same insertion.The allene insertion proceeds considerably faster than the insertion of alkenes. Concluding remarks We studied the chemical reactivity of the bimetallic tungsten complex 2 containing a m-h1 :h2-ketene bridge; particularly chemical reactions with various heterocumulenes. From the reaction with isothiocyanate, the bridging thioketene complex 3 was isolated which displayed a dissimilar bridging mode from that of the bridging ketene in 2.The reaction of 2 with CS2 induced S insertion and aVorded the thiocarbonyl complex 5. In the case of allene as an incoming ligand, co-ordination of the allene ligand is followed by a subsequent ketene–allene coupling at the central carbon atom of the co-ordinated allene to yield the allylic product 7. Experimental General procedures All manipulations were performed under nitrogen using vacuum line, dry-box and standard Schlenk techniques.The NMR spectra were recorded on a Bruker AM-300WB spectrometer and are reported in units of ppm with residual protons in the solvent as an internal standard (CDCl3, d 7.24), IR spectra on a Bruker Vector-22 instrument and frequencies (cm21) assigned relative to a polystyrene standard and FAB mass spectra on a JEOL SX-102A spectrometer. Diethyl ether was distilled from CaH2 and stored over molecular sieves prior to use, benzene and CH2Cl2 from LiAlH4 and CaH2, respectively and THF from sodium–benzophenone. All other solvents and reagents were reagent grade used without further purification. The compound W(CO)6 was purchased from Strem Chemical, dicyclopentadiene, CS2 and CH2I2 from Merck.Complexes [WCp(CO)3]2,27 WCp(CO)3 2 1,28 W2Cp2(CO)5(m-h1 :h2- CH2CO), 2 and the Cp9 analogue 29 8 were prepared according to the literature methods. Reaction of complex 2 with PhCH2NCS at room temperature A solution of complex 2 (1.20 g, 1.76 mmol) in 20.0 mL of CH2Cl2 under nitrogen was treated with PhCH2NCS (0.28 g, 1.88 mmol) and the resulting solution stirred for 30 min turning from red to dark red.Then the solvent was removed under vacuum, and the residual red oil redissolved in 5 mL of CH2Cl2. Addition of 20 mL of hexane caused precipitation of a dark red product which was filtered oV and washed with 2 × 10 mL of hexane to give W2Cp2(CO)5(m-SC]] CH2) 3, (0.47 g) in 38% yield. IR, (cm21, CHCl3): 2036vs, 1947vs, 1916 (sh) and 1813m [n(CO)]. 1H NMR, (CDCl3): d 6.32 (d, JH-H = 0.88, JW-H = 6.90, 1 H, ]] CH); 5.93 (d, JH-H = 0.88 Hz, 1 H, ]] CH); 5.53, 5.42 (s, Cp). 13C NMR, (CDCl3): d 242.1, 224.6, 214.2 (CO); 166.5 (CS); 117.9 (CH2); 94.7, 92.5 (2 Cp). FAB MS: m/z 696 (M1) and 668 (M1 2 CO). Calc. for C17H12O5SW2: C, 29.34; H, 1.74. Found: C, 29.48; H, 1.68%. Complex W2Cp92(CO)5(m-SC]] CH2) 39, was prepared from 29 and PhCH2NCS in 33% yield using the same procedure. IR (cm21), CHCl3: 2033vs, 1943vs, 1913 (sh) and 1816m [n(CO)]. 1H NMR, (CDCl3): d 6.25 (d, JH-H = 0.66, 1 H, ]] CH); 5.86 (JH-H = 0.66 Hz, 1 H, ]] CH); 5.53–5.14 (m, 8 H, C5H4); 2.18, 1.95 (s, Me). FAB MS: m/z 724 (M1), 696 (M1 2 CO) and 668J. Chem. Soc., Dalton Trans., 1999, 2243–2248 2247 (M1 2 3CO). Calc. for C19H16O5SW2: C, 31.52; H, 2.23. Found: C, 31.68; H, 2.33%. Reaction of complex 3 with HBF4 To a solution of complex 3 (0.20 g, 0.03 mmol) dissolved in 0.50 mL of CDCl3 under nitrogen, HBF4 (54% in diethyl ether, 0.30 mmol) was added.The solution changed from dark red to orange immediately. The solvent was removed under vacuum, and the red oil redissolved in 5 mL of CH2Cl2. Addition of 20 mL of hexane caused precipitation of a dark red product which was filtered oV and washed with 2 × 10 mL of hexane to give [W2Cp2(CO)5(m-SCMe)]BF4 4 (0.21 g) in 95% total yield. IR (cm21, CHCl3): 2048vs, 2029s, 1968s and 1944vs, [n(CO)]. 1H NMR, CDCl3: d 5.95 (s, 5 H, Cp); 5.93 (s, 5 H, Cp) and 2.99 (s, 3 H, CH3). 13C NMR (CDCl3): d 256.6 (CS); 220.3, 215.9, 215.8, 214.8, 210.0 (CO); 94.5, 94.0 (2 Cp); 35.5 (CH3). FAB: MS m/z 697 (M1), 669 (M1 2 CO) and 641 (M1 2 2CO). Calc. for C17H13BF4O5SW2: C, 26.04; H, 1.67. Found: C, 26.29; H, 1.82%. Protonation using CF3CO2H and the same procedure aVorded a similar product [W2Cp2(CO)5(m-SC(O2CCF3)Me)]. The parent peak of this complex in the FAB mass spectrum appeared at m/z = 810. The 1H NMR spectrum is essentially the same as that of 4.IR (cm21, CHCl3): 2051vs, 2031s, 1970s, 1945vs and 1787vs, [n(CO)]. 1H NMR, (CDCl3): d 5.88 (s, 5 H, Cp); 5.86 (s, 5 H, Cp); and 2.96 (s, 3 H, CH3). FAB MS: m/z 810 (M1), 782 (M1 2 CO), 754 (M1 2 2CO) and 697 (M1 2 CF3- COO). Calc. for C19H13F3O7SW2: C, 28.17; H, 1.62. Found: C, 28.44; H, 1.74%. When CF3CO2D is used, a 1H multiplet resonance at d 2.94 (JH-D = 4.91 Hz) was observed indicating protonation at the methylene group. By addition of Et3N to the solution of 4, complex 3 was recovered in greater than 90% NMR yield.Reaction of complex 2 with CS2 To a solution of complex 2 (1.20 g, 1.76 mmol) in 20.0 mL of CH2Cl2, CS2 (0.16 g, 2.11 mmol) was added under nitrogen. The resulting solution was stirred for 30 min turning scarlet red. The solvent and excess of CS2 were removed under vacuum to leave a red oily residue, which was redissolved in 5 mL of CH2Cl2. Addition of 20 mL of hexane caused precipitation of a red product which was filtered oV and washed with 2 × 10 mL of hexane to give an isomeric mixture of cis- and trans-W2- Cp2(CO)5(CS)(CH2COS) 5 (1.24 g) in 93% total yield.IR (cm21, CHCl3): 2024vs, 1927vs and 1594w [n(CO) and n(CS)]. 1H NMR (CDCl3): trans isomer, d 5.79, 5.54 (s, Cp); 2.86 (s, JW-H = 5.2 Hz, 2 H, CH2), cis isomer, d 5.71, 5.57 (s, Cp); 2.91 (AB pattern, JH-H = 3.3 Hz, 2 H, CH2). 13C NMR (CDCl3): d 228.8, 224.8, 217.1, 216.8, 216.6, 215.1, 210.8, 210.3 and 209.1 (CO); trans isomer, 343.7 (CS); 228.9 (CH2CO); 96.2, 91.8 (2 Cp); 1.5 (CH2); cis isomer, 330.2 (CS); 228.4 (CH2CO); 95.0, 91.8 (2 Cp); 0.9 (CH2).FAB MS: m/z 756 (M1), 728 (M1 2 CO), 700 (M1 2 2CO), 628 (M1 2 3CO,CS), 616 (M1 2 5CO) and 600 (M1 2 4CO,CS). Calc. for C9H6O3SW: C, 28.59; H, 1.60. Found: C, 28.31; H, 1.41%. A mixture of cis and trans isomers of complex W2Cp92- (CO)5(CS)(m-CH2COS) 59 in 90% total yield could be similarly prepared. Single crystals of 59 containing both isomers were grown by careful addition of hexane to a CH2Cl2 solution of 59.IR (cm21, CHCl3): 2021vs, 1923vs and 1593w [n(CO) and n(CS)]. 1H NMR, (CDCl3): trans isomer, d 5.69–5.43 (m, 8 H, 2 C5H4); 2.84 (s, JW-H = 5.4 Hz, 2 H, CH2), 2.17 (s, 3 H, Me) 2.11 (s, 3 H, Me); cis isomer, d 5.69–5.48 (m, 8 H, 2 C5H4); 2.88 (AB pattern, JH-H = 5.6 Hz, 2 H, CH2), 2.17 (s, 3 H, Me) 2.15 (s, 3 H, Me). 13C NMR (CDCl3): d 230.1, 225.3, 218.3, 218.0, 217.7, 215.7, 210.8, 210.5, 209.0 (CO); trans isomer, 348.2 (CS); 230.2 (CH2CO); 115.2 (CMe), 107.4 (CMe); 95.2–92.0 (C5H4); 13.9, 13.5 (2Me); 2.2 (CH2); cis isomer, 332.8 (CS); 229.9 (CH2CO); 114.1 (CMe), 108.2 (CMe); 96.2–91.2 (C5H4); 13.7, 13.3 (2Me); 1.7 (CH2).FAB MS: m/z 784 (M1), 756 (M1 2 CO), 728 (M1 2 2CO), 656 (M1 2 3CO,CS), 644 (M1 2 5CO) and 628 (M1-4CO,CS). Calc. for C10H8O3SW: C, 30.63; H, 2.06. Found: C, 30.41; H, 1.84%. Reaction of complex 2 with S]] C(SCH2)2 To a solution of complex 2 (0.70 g, 1.03 mmol) in 20 mL of CH2Cl2 under nitrogen, SC(SCH2)2 (0.21 g, 1.05 mmol melt at 45 8C) was added through a micro-syringe. The resulting solution was stirred for 30 min turning dark yellow.Then the solvent was removed under vacuum yielding oily residue, which was washed with 2 × 10 mL of hexane to give the product. The crude product was recrystallized from CH2Cl2–hexane (1 : 5) to give W2Cp2(CO)5[C(SCH2)2](m-CH2COS) 6 (0.51 g) in 61% yield. IR (cm21, CH2Cl2): 2021vs, 1981m, 1925s and 1903s [n(CO)]. 1H NMR, (CDCl3): d 5.70, 5.57 (s, Cp); 3.38 (s, 4 H, SCH2); 2.83 (s, 2 H, CH2). 13C NMR (CDCl3): d 266.3 (M]] C); 229.6, 216.9, 216.8, 210.3 (CO and CS); 97.5, 91.9 (2 Cp); 44.9 (SCH2); 2.1 (CH2CO). FAB MS: m/z 817 (M1 1 1), 788 (M1 2 CO), 759 (M1 2 2CO) and 732 (M1 2 3CO). Calc. for C20H16O6S3W2: C, 29.43; H, 1.98. Found: C, 30.01; H, 1.81%. Reaction of complex 2 with H2C]] C]] CH2 Gaseous allene was slowly bubbled through a deep red solution of complex 2 (2.30 g, 3.38 mmol) in 30.0 mL of CH2Cl2 at room temperature for 35 min until it turned light yellow.Then the solvent was removed under vacuum to yield a yellow oil which was further purified by recrystallization from CH2Cl2–hexane (2 : 3) at 220 8C to give W2Cp2(CO)5[m-CH2COC(CH2)2] 7, (2.24 g) in 92% yield. IR (cm21, CH2Cl2): 2022m, 1957w (sh), 1918vs and 1613m [n(CO)]. 1H NMR (CDCl3, 210 8C): endo isomer; d 5.58, 5.29 (s, Cp); 2.99 (br, 2 H, anti-CH2); 2.18 (s, 2 H, CH2); 1.60 (br, 2 H, syn-CH2); exo isomer; d 5.61, 5.26 (s, Cp); 3.10 (br, 2 H, anti-CH2); 2.10 (s, 2 H, CH2); 1.15 (br, 2 H, syn- CH2). 13C NMR (CDCl3): d 229.0, 218.6, 208.1 (CO); 106.6 (C); 92.4, 87.8 (2 Cp); 25.7 (CH2); 211.0 (CH2). FAB MS: m/z 720 (M1), 692 (M1 2 CO), 636 (M1 2 3CO), 608 (M1 2 4CO) and 580 (M1 2 5CO). Calc. for C10H8O3W: C, 33.36; H, 2.24. Found: C, 33.51; H, 2.11%. X-Ray analysis Dark red crystals of complex 3 suitable for X-ray diVraction study were grown directly from CH2Cl2.A suitable single crystal of dimensions 0.11 × 0.24 × 0.30 mm was glued to a glass fiber and mounted on an Enraf-Nonius CAD4 diVractometer. Initial lattice parameters were determined from a least-squares fit to 25 accurately centered reflections having 16.88 < 2q < 27.128. Cell constants and other pertinent data are collected in Table 4. An empirical correction for absorption, based on the azimuthal scan data, was applied to the intensities. Crystallographic computations were carried out using the NRCC structure determination package.29 Final refinement using full-matrix, least squares converged smoothly.For complex 59 the data were collected on a Siemens SMART CCD system using 3 kW sealed-tube molybdenum K radiation. Exposure time was 5 s per frame. A SADABS (Siemens area detector absorption) absorption correction 30 was applied, and decay was negligible. Data were processed and the structure was solved and refined by the SHELXTL program.31 The structure was solved using direct methods and confirmed by Patterson methods refining on intensities of all data (3925 reflections) to give RF = 0.0633, R9F2 = 0.1389. In the structure determination the composite scattering factors were used for S(2) and O(5) in the final least squares refinement.The scattering factor for S(2) was set to be the sum of 67% of the scattering factor of S and 33% of that of O. The scattering factor for O(5) was set to be the sum of 33% of the scattering factor of S and 67% of that of O.The procedures for the structure determination of 7 were similar to those for 3.2248 J. Chem. Soc., Dalton Trans., 1999, 2243–2248 CCDC reference number 186/1466. See http://www.rsc.org/suppdata/dt/1999/2243/ for crystallographic files in .cif format. Acknowledgements We are grateful for support of this work by National Science Council, Taiwan, Republic of China. References 1 W. A. Herrmann and J. Plank, Angew. Chem., Int.Ed. Engl., 1978, 17, 525; T. W. Bodner and A. R. Cutler, J. Am. Chem. Soc., 1983, 105, 5926; P. T. Barger, B. D. Santarsiero, J. Armantrout and J. E. Bercaw, J. Am. Chem. Soc., 1984, 106, 5178; M. D. Curtis, L. Messerle, J. J. D’Errico, H. E. Solis, I. D. Barcelo and W. M. Butler, J. Am. Chem. Soc., 1987, 109, 3603; G. L. GeoVroy and S. L. Bassner, Adv. Organomet. Chem., 1988, 28, 1. 2 M. C. Fermin, A. S. Hneihen, J. J. Maas and J. W. Bruno, Organometallics, 1993, 12, 1845; M.C. Fermin and J. W. Bruno, J. Am. Chem. Soc., 1993, 115, 7511; E. Bleuel, M. Laubender, B. Weberndörfer and H. Werner, Angew. Chem., Int. Ed. Engl., 1999, 38, 156. 3 D. A. Straus and R. H. Grubbs, J. Am. Chem. Soc., 1982, 104, 5499; E. J. Moore, D. A. Straus, J. Armantrout, B. D. Santarsiero, R. H. Grubbs and J. E. Bercaw, J. Am. Chem. Soc., 1983, 105, 2068; C. A. Rusik, T. L. Tonker and J. L. Templeton, J. Am. Chem. Soc., 1986, 108, 4652; C. A. Rusik, M. A. Collins, A.S. Gamble, T. L. Tonker and J. L. Templeton, J. Am. Chem. Soc., 1989, 111, 2550. 4 W. T. Brady, in The Chemistry of Ketenes, Allenes and Related Compounds, ed. S. Patai, Wiley, New York, 1980, ch. 8. 5 Y. C. Lin, J. C. Calabrese and S. S. Wreford, J. Am. Chem. Soc., 1983, 105, 1679; Y. C. Lin, J. Chin. Chem. Soc., 1985. 32, 295; M. C. Chen, Y. J. Tsai, C. T. Chen, Y. C. Lin, T. W. Tseng, G. H. Lee and Y. Wang, Organometallics, 1991, 10, 378; E. D. Morrison, G. R. Steinmetz, G.L. GeoVroy, W. C. Fultz and A. L. Rheingold, J. Am. Chem. Soc., 1984, 106, 4783; S. C. H. Ho, D. A. Straus, J. Armantrout, W. P. Schaefer and R. H. Grubbs, J. Am. Chem. Soc., 1984, 106, 2210; E. D. Morrison and G. L. GeoVroy, J. Am. Chem. Soc., 1985, 107, 3541; J. S. Holmgren, J. R. Shapley, S. R. Wilson and W. T. Pennington, J. Am. Chem. Soc., 1986, 108, 508. 6 W. A. Herrmann, Adv. Organomet. Chem., 1982, 20, 159; Angew. Table 4 Crystal and intensity collection data for W2Cp2(CO)5- (m-CH2CS) 3, W2Cp92(CO)5(CS)(m-CH2COS) 59 and W2Cp2(CO)5- [m-CH2COC(CH2)2] 7 3 59 7 Chemical formula M Crystal system Space group a/Å b/Å c/Å b/8 V/Å3 ZT /K m/cm21 Total number of reflections Unique reflections RR 9 C17H12O5SW2 696.03 Monoclinic P21/n 9.755(2) 15.936(4) 12.226(4) 111.81(2) 1764.5(8) 4 298 134.483 3100 2105 (I > 2s(I)) 0.057 0.056 C20H16O6S2W2 784.15 Monoclinic P21/n 7.1019(8) 46.339(4) 7.511(10) 116.44(1) 2213(3) 4 298 203.811 4520 3661 (I > 3s(I)) 0.056 0.063 C20H16O6W2 720.04 Monoclinic C2/c 26.296(5) 7.023(2) 24.226(6) 118.30(2) 3939.0(17) 8 298 119.576 3483 2549 (I > 2s(I)) 0.038 0.038 Chem., Int.Ed. Engl., 1982, 21, 117; R. J. Puddephatt, Comments Inorg. Chem., 1982, 2, 69a; Polyhedron, 1988, 7, 767. 7 M. Akita, A. Kondoh and Y. Moro-oka, J. Chem. Soc., Chem. Commun., 1986, 1296. 8 Y. L. Yang, L. J. Wang, Y. C. Lin, S. L. Huang, M. C. Chen, G. H. Lee and Y. Wang, Organometallics, 1997, 16, 1573. 9 (a) F. Drews, F.Edelmann and U. Behrens, J. Organomet. Chem., 1986, 315, 369; (b) J. Wolf, R. Zolk, U. Schubert and H. Werner, J. Organomet. Chem., 1988, 340, 161; (c) H. Werner and U. Brekau, Z. Naturforsch., Teil B., 1989, 44, 1438; (d) H. Werner, T. Rappert and J. Wolf, Isr. J. Chem., 1990, 30, 377. 10 A. Bax, M. Ikura, L. E. Kay, D. A. Torchia and R. Tschudin, J. Magn. Reson., 1990, 86, 304; S. Roy, M. Z. Papastarros, V. Sanches and A. G. Redfield, Biochemistry, 1984, 23, 4395; A.Bax and M. F. Summers, J. Am. Chem. Soc., 1986, 108, 2093; T. J. Norwood, J. Boyd, J. E. Heritage, N. SoVe and I. D. Campbell, J. Magn. Reson., 1990, 87, 488; A. Bax and S. Subramanian, J. Magn. Reson., 1986, 67, 565. 11 C. K. Johnson, ORTEP II, Report ORNL-5138, Oak Ridge National Laboratory, Oak Ridge, TN, 1976. 12 M. Bamber, S. F. T. Froom, M. Green, M. Schulz and H. Werner, J. Organomet. Chem., 1992, 434, C19. 13 M. Herberhold, W. Jellen and H. H. Murray, J. Organomet.Chem., 1984, 270, 65. 14 A. M. English, K. R. Plowman, I. M. Baibich, J. P. Hickey, I. S. Butler, G. Jaouen and P. Lemaux, J. Organomet. Chem., 1981, 205, 177. 15 F. J. Brown, Prog. Inorg. Chem., 1980, 27, 1. R. J. Angelici, F. B. McCormick and R. A. Pickering, in Fundamental Research in Organometallic Chemistry, eds. M. Tsutsui, Y. Ishii and H. Yaozeng, Van Nostrand, New York, 1982, p. 347; B. E. Boland-Lussier and R. P. Hughes, Organometallics, 1982, 1, 635; S. Myrvold, O. A. Nassif, G. Semelhago, A. Walker and D. H. Farrar, Inorg. Chim. Acta, 1986, 117, 17; G. Beck and W. P. Fehlhammer, Angew. Chem., Int. Ed. Engl., 1988, 27, 1344. 16 H. Werner, K. Leonhard, O. Kolb, E. Rottinger and H. Vahrenkamp, Chem. Ber., 1980, 113, 1654. 17 H. Werner, M. Ebner and H. Otto, J. Organomet. Chem., 1988, 350, 257. 18 J. Qi, P. W. Schrier, P. E. Fanwick and R. A. Walton, J. Chem. Soc., Chem. Commun., 1991, 1737. 19 P. V. Broadhurst, B. F. G. Johnson, J. Lewis and P. R. Raithby, J. Chem. Soc., Dalton Trans., 1982, 1641. 20 J. Fortune and A. R. Manning, Organometallics, 1983, 2, 1719. 21 L. Contreras, A. Pizzano, L. Sánchez, E. Carmona, A. Monge and C. Ruiz, Organometallics, 1995, 14, 589. 22 P. V. YaneV, Coord. Chem. Rev., 1977, 23, 183. 23 T. W. Tseng, I. Y. Wu, J. H. Tsai, Y. C. Lin, D. J. Chen, G. H. Lee, M. C. Chen and Y. Wang, Organometallics, 1994, 13, 3963; S. Doherty, M. R. J. Elsegood, W. Clegg and M. Waugh, Organometallics, 1996, 15, 2688. 24 C. L. Hsing and Y. C. Lin, manuscript in preparation. 25 R. Aumann, H.-D. Melchers and H.-J. Weidenhaupt, Chem. Ber., 1987, 120, 17. 26 R. K. Rulke, D. Kliphuid, C. J. Elsevier, J. Fraanje, K. Goubitz, P. W. N. M. van Leeuwen and K. Vrieze, J. Chem. Soc., Chem. Commun., 1994, 1817. 27 J. E. Thomasson, P. W. Robinson, D. A. Ross and A. Wojcicki, Inorg. Chem., 1971, 10, 2130. 28 I. Y. Wu, J. H. Tsai, B. C. Huang, S. C. Chen and Y. C. Lin, Organometallics, 1993, 12, 3971. 29 E. J. Gabe, F. L. Lee and Y. Lepage, in Crystallographic Computing 3, G. M. Sheldrick, C. Kruger and R. Goddard (eds.), Clarendon Press, Oxford, 1985, p. 167. 30 G. M. Sheldrick, SADABS, Bruker AXS, Madison, WI, 1997. 31 G. M. Sheldrick, SHELXTL-PLUS, Siemenes Analytical X-ray Instruments, Madison, WI, 1990; SHELXL-97, Structure Solution and Refinement Package, University of Göttingen, Göttingen, Germany, 1997. Paper 9/01823I © Copyright 1999 by the Royal Society of Chemistry

ISSN:1477-9226    

DOI:10.1039/a901823i

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1998, Pages 2247–2252 2247 New dioxocyclam ligands appended with 2-pyridylmethyl pendant(s): synthesis, properties and crystal structure of their copper(II) complexes (dioxocyclam 5 1,4,8,11-tetraazacyclotetradecane- 12,14-dione) ‡ Xian He Bu,*,†,a Dao Li An,a Xi Chuan Cao,a Ruo Hua Zhang,a Thomas CliVord b and Eiichi Kimurab a Department of Chemistry, Nankai University, Tianjin 300071, P. R. China b Department of Medicinal Chemistry, School of Medicine, Hiroshima University, Kasumi 1-2-3, Minami-Ku, Hiroshima 734, Japan Two new dioxocyclam ligands, 4-(pyridin-2-ylmethyl)-1,4,8,11-tetraazacyclotetradecane-12,14-dione (L1) and 4,8-bis(pyridin-2-ylmethyl)-1,4,8,11-tetraazacyclotetradecane-12,14-dione (L2), have been synthesized and characterized.The solution chemistry of the CuII complexes with these ligands have been studied by potentiometric and spectroscopic titration, cyclic voltammetry (CV), UV/VIS and ESR spectral techniques.The CuII complex of L2 has been isolated as single crystals and the structure has been determined by X-ray diVraction analysis to be [CuL2][ClO4]2 1. In complex 1, the CuII atom is five-co-ordinate with the two pendant pyridyl nitrogens and the two tertiary amines forming a distorted trigonal-bipyramidal configuration in which the backbone oxygen O(2) co-ordinates to CuII while the two amido nitrogen groups remained non-deprotonated and non-co-ordinated.To our knowledge, this is the first example of CuII complexes of dioxotetraamines in which the backbone oxygen co-ordinates to the central metal by twisting the backbone. The chemistry of macrocyclic dioxotetraamines has received much attention and has been extensively studied in recent years.1–6 Typically, the 14-membered tetraamine macrocycles cyclam, monooxocyclam, and dioxocyclam (see below), like porphyrins and corrin, incorporate metal ions into their cavities to form stable complexes. Macrocyclic dioxotetraamines are unique chelators for some transition-metal ions. They bear dual structural features of macrocyclic tetraamines and oligopeptides and have many interesting properties and important functions.They can stabilize the higher oxidation states of some transition metals.1,2a,7,8 These properties have been applied to superoxide dismutase-like catalysts. The two amido groups in macrocyclic dioxotetraamines are equivalent, when co-ordinated to a 3d metal ion, they will be deprotonated simultaneously.As for dioxotetraamine, the presence of a non-deprotonated or a singly deprotonated complex is unlikely, therefore, complexes with this kind of ligand containing non-deprotonated amido groups were generally not considered in previous studies. So far, great eVort has been devoted to the incorporation of functionalized pendant groups into a saturated macrocyclic tetraamine structure (e.g. cyclam) to modify its conformational and the redox properties of the metal complex.9 However, up to now, only a few examples of the macrocyclic dioxotetraamines bearing functionalized pendant groups have HN NH NH HN HN NH NH HN O HN NH NH HN O O cyclam monooxocyclam dioxocyclam † E-Mail: buxh@public1.tpt.tj.cn ‡ Non-SI unit employed: G = 1024 T.been reported.2i,j,6,10,11 Herein, we report the synthesis and characterization of two novel macrocyclic dioxocyclam (dioxocyclam = 1,4,8,11-tetraazacyclotetradecane-12,14-dione) ligands bearing 2-pyridylmethyl as additional co-ordinating donor pendant(s) (see below), and their complexation properties with CuII as well as the crystal structure of one of their CuII complexes. In this complex, the two amido groups remain non-deprotonated, and the backbone oxygen atom co-ordinates to the central CuII ion by twisting the macrocycle backbone.To our knowledge, this is the first example for the dioxotetraamine metal complexes bearing such features.Experimental Materials and methods Most of the starting materials and solvents for syntheses were obtained commercially and purified prior to use. Dioxocyclam (L0, 1,4,8,11-tetraazacyclotetradecane-12,14-dione) was prepared according to the literature method.12 2-Chloromethylpyridine hydrochloride was purchased from Aldrich and used without further purification. Fourier-transform IR spectra were recorded on a 170SX (Nicolet) spectrometer; EI-MS was carried out on a VG ZAB-HS instrument.Elemental analyses were taken on a P-E 240C analyzer. Proton NMR spectra were recorded on a Bruker AC-P 200 spectrometer (200 MHz) at 25 8C, in CDCl3, with tetramethylsilane as the internal reference. The ESR spectra were measured on a Bruker ER-200-DSRC10 spectrophotometer. Methanol solutions of copper(II) HN NH N HN N O O HN NH N N N O O N L2 L12248 J. Chem. Soc., Dalton Trans., 1998, Pages 2247–2252 complexes [prepared by mixing equimolar amounts of the ligand and Cu(NO3)2 in methanol solution, and the pH value was adjusted to ª7 by methanolic NaOH solution] at room temperature and 112 K were used for the measurements. Cyclic voltammetric measurements were performed with a PARC Model 273 electrochemical apparatus in aqueous solution at 25 8C with 0.5 M Na2SO4 as supporting electrolyte and pure Ar gas was bubbled through the solution.The concentration of the complexes Cu(H22L0), Cu(H22L1) and Cu(H22L2) were kept at 2 × 1023 M.The pH was adjusted with concentrated NaOH or H2SO4 solution. The cyclic voltammograms at a scan rate of 100 mV s21 were evaluated graphically. A three-electrode system was employed: glassy carbon as working electrode, saturated calomel electrode (SCE) as a reference electrode and Pt as a counter electrode. All solutions for potentiometric titrations were made up with freshly redistilled water stored over nitrogen. High purity potassium nitrate, CuII salts, and sodium hydroxide were obtained from Aldrich and used without further purification. Standard HCl was obtained from Nacali Tesque.The concentration of copper(II) nitrate stock solution was determined by H4edta complexometric titration. Carbonate-free stock solutions of NaOH were prepared by dilution of a chilled 10 M stock solution with distilled, boiled water and then standardised potentiometrically by titrating with standard 0.1000 M HCl. A stock solution of HNO3 was prepared by diluting concentrated HNO3.All the other reagents for syntheses and analyses were of analytical grade. Potentiometric data for the determination of protonation and complexation constants were obtained from a Horiba F-16 pH meter equipped with Horiba reference and pH electrodes. Test solutions were thermostatted using a water-jacketed cell connected to an external circulating water bath (Komatsu- Yamato CTE-310 cooling bath). The UV/VIS spectra for spectrophotometric studies were measured at 25.0 8C using a Hitachi U-3500 spectrometer equipped with a thermoelectric cell-temperature controller coupled to a thermostatted circulating water bath as a secondary heat buVer (0.01). All measurements were conducted in aqueous solution at 25.0 8C.The electrode was calibrated by titration of standard base against standard acid as described earlier,13 defining pH as p[H] (pKw = 213.79), and the slope and oVset of the electrode calculated from the theoretical concentration of H1 at a single acidic and a single basic point.The linearity of electrode response and carbonate contamination of the standard NaOH solution (0.100 M) was determined by Gran’s method and was found to be less than 1%. All samples were kept under an argon atmosphere (supplied through a NaOH solution and distilled water wash-bottle). The solution temperature was maintained at 25.0 ± 0.1 8C and the ionic strength was maintained at 0.10 M with KNO3. Aqueous solutions (50 cm3) of the macrocyclic ligands (1 × 1023 M) in the presence or absence of CuII (3 × 1023 to 8 × 1024 M) were titrated with standard NaOH solution.Protonation constants were determined from a total of four curves (164 points) for L1 and three curves (131 points) for L2. Copper complexation constants were determined from three curves (104 points) for L1 and two curves (49 points) for L2. Syntheses 4-(Pyridin-2-ylmethyl)-1,4,8,11-tetraazacyclotetradecane- 12,14-dione (L1) and 4,8-bis(pyridin-2-ylmethyl)-1,4,8,11- tetraazacyclotetradecane-12,14-dione (L2). To a solution of 2-chloromethylpyridine hydrochloride (410 mg, 2.5 mmol) in 30 mL of deoxygenated dimethylformamide (DMF) was added dropwise to a solution of dioxocyclam (2.3 g, 10 mmol) in 80 mL of deoxygenated DMF in the presence of excess amount of fine and dried K2CO3 at ca. 80 8C. The resulting reaction mixture was heated and stirred at 80 8C for about 10 h under Ar. After filtration of the reaction mixture, the filtrate was evaporated to dryness and then dissolved in H2O and then extracted with CHCl3.The combined CHCl3 solutions were dried and evaporated, and then the residue was purified by column chromatography on silica gel by eluting with CH2Cl2– MeOH (100: 5). The product L1 was finally recrystallized from acetonitrile as colourless needles (320 mg, 40% based on 2-chloromethylpyridine hydrochloride). 1H NMR (CDCl3): d ª1.75 (2 H, m), 2.44–2.67 (8 H, m), 3.26 (2 H, s), 3.32 (2 H, qnt), 3.55 (2 H, qnt), 4.17 (2 H, s), 7.26–7.31 (2 H, m), 7.68–7.69 (2 H, m), 8.52–8.55 (1 H, m).IR (KBr pellet): 546w, 666w, 709w, 760m, 956w, 1043w, 1148m, 1309m, 1357m, 1431m, 1553s, 1629s, 1672s, 2795w, 2800m, 2933m, 3055w, 3325s cm21. EI-MS: M1 peak m/z = 319 (Mr = 319.41) (Found: C, 60.03; H, 8.16; N, 21.90. Calc. for C16H25N5O2: C, 60.17; H, 7.89; N, 21.93%). For the preparation of L2, a solution of 2-chloromethylpyridine hydrochloride (1.64 g, 10 mmol) in 50 mL of deoxygenated DMF was added dropwise to a solution of dioxocyclam (0.760 g, 3.33 mmol) in 50 mL of deoxygenated DMF in the presence of an excess of fine and dried K2CO3 at ca. 80 8C.The resulting mixture was heated and stirred at ca. 80 8C for about 10 h under Ar. After filtration of the reaction mixture, the filtrate was evaporated to dryness and then dissolved in H2O and then extracted with CHCl3. The combined CHCl3 solutions were dried and evaporated, and then, the residue was purified by column chromatography on silica gel by eluting with CH2Cl2–MeOH (100: 3).The product was finally recrystallized from acetonitrile as colourless needles (0.96 g, 70% based on dioxocyclam). 1H NMR (CDCl3): d ª1.75 (2 H, m), 2.44–2.83 (8 H, m), 3.26 (2 H, s), 3.32–3.35 (4 H, m), 3.69 (2 H, s), 7.19– 7.28 (4 H, m), 7.62–7.65 (2 H, m), 8.50–8.52 (2 H, m). IR (KBr pellet): 565w, 612w, 727w, 766m, 956w, 1049w, 1160m, 1316m, 1359m, 1433m, 1476w, 1518s, 1567m, 1641s, 1662s, 2801s, 2927m, 3063w, 3307s cm21.EI-MS: M1 peak m/z = 410 (Mr = 410.52) (Found: C, 64.25; H, 7.26; N, 20.23. Calc. for C22H30N6O2: C, 64.37; H, 7.37; N, 20.47%). [CuL2][ClO4]2 1. For complex 1 the single crystal suitable for X-ray analysis was obtained by mixing a 1 : 1 molar ratio of Cu(ClO4)2 and L2 in deoxygenated MeOH under reflux for ca. 15 min. The blue reaction mixture was then filtered. The crystal suitable for X-ray analysis was obtained upon slow evaporation of the solvent (Found: C, 39.51; H, 4.28; N, 12.22.Calc. for C22H30Cl2CuN6O10: C, 39.27; H, 4.49; N, 12.49%). IR (KBr pellet): 582w, 623m, 768w, 1068s, 1102vs, 1301w, 1446m, 1536w, 1573m, 1631s, 1695s, 2935w, 3125w, 3338s, 3430s (br) cm21. Crystallography A blue crystal (approximately 0.1 × 0.3 × 0.5 mm) of complex 1 was mounted on an Enraf-Nonius CAD-4 diVractometer equipped with a graphite-crystal monochromator situated in the incident beam for data collection.The determination of unit cells and the data collection were performed with Mo-Ka radiation (l = 0.710 73 Å). Unit-cell dimensions were obtained by least-squares refinements using 25 reflections in the q range 7.92–13.458. The intensities of reflections were measured in the w–2q scan mode in the range 2 < q < 238 at room temperature (299 ± 1 K). Crystal and instrument stabilities were monitored with a set of three standard reflections measured every 60 min, in all cases no significant variations were found.A total of 4035 reflections were collected and among them 3187 reflections are independent reflections, in which 2031 reflections with I > 2s(I) were considered to be observed and used in the succeeding refinements. The structure was solved by direct methods. All calculations were performed on an IBM 486 personal computer with the Siemens SHELXTL PC program package.14 The Cu atom was located from an E-map. The other non-hydrogen atoms were determined with successive Fourier diVerence syntheses.The final refinement was done by full-matrix least-J. Chem. Soc., Dalton Trans., 1998, Pages 2247–2252 2249 squares methods with anisotropic thermal parameters for non-hydrogen atoms. The refinement agreement factors are R = 0.064 and R9 = 0.056 {w = 1/[s2(F) 1 0.0001F2]}. The highest peak and hole on the final Fourier-diVerence map had a height of 0.75 and 20.68 e Å23, respectively. The hydrogen atoms were added theoretically, riding on the atoms concerned and refined with fixed thermal factors.We should note that the crystal was of poor quality even though several attempts were made in diVerent conditions to improve the quality. Crystal data for complex 1: C22H30Cl2CuN6O10, M = 672.96, monoclinic, space group P21/c, a = 8.962(2), b = 12.588(3), c = 23.940(5) Å, b = 92.508, U = 2698.2(1.9) Å3, Dc = 1.657 g cm23, Z = 4, F(000) = 1388, m = 10.75 cm21. CCDC reference number 186/989.Results and Discussion Synthesis of ligands and complexes The compound dioxocyclam was synthesized following the method reported by Tabushi et al.12 The new ligands were prepared according to Scheme 1. An excessive amount of dioxocyclam (L0) was used to obtain the monoalkylated product L1 which was purified by silica gel column chromatography and recrystallized from CH2Cl2–MeCN as colourless needles in 40% yield. The doubly-substituted ligand L2 was also obtained as a by-product and can be isolated by column chromatography from the monoalkylated product.The doubly-substituted product L2 can be prepared using an excessive amount of 2- chloromethylpyridine hydrochloride. The yield of L2 in this reaction was 70%. All the analytical and spectral data are in good agreement with the theoretical requirements of the new ligands. The new ligand L2 was treated with 1 equivalent of Cu(ClO4)2 in deoxygenated MeOH under reflux. The 1 : 1 pyridyl-pendant dioxocyclam–CuII complex was obtained as blue crystals.The results of elemental analysis (C, H, N) indicated a possible formula of [CuL2][ClO4]2. The co-ordination of one of the backbone oxygen atoms and the non-bonding of the amide nitrogens was also confirmed by the IR spectra of the complex [nC]] O: 1631 cm21 for C]] O(Cu) and 1695 cm21 for non-coordinated C]] O]. The non-deprotonated pyridyl-pendant dioxocyclam–CuII complex is air-stable in the solid state. Protonation and complexation Potentiometric titrations.Potentiometric pH measurements and computation of the protonation constants and CuII binding constants were carried out by procedures described previously. 13 Approximately 1 : 1 L : Cu molar ratios were used. The found/expected error, s, was found to be less than 12 in all calculations (more than one curve was used in each calculation). The titration curve for L1 in the presence of 3 equivalents of HNO3 with standard acid with and without CuII are shown in Fig. 1 and the corresponding titrations with L2 are plotted in Fig. 2. The ligands show titratable protons resulting from deprotonation at the pyridyl nitrogens and the amine ring nitrogens but the amide protons are too basic to be deprotonated. The final deprotonation of the L2 macrocycle occurs at Scheme 1 HN NH NH HN O O + N CH2 Cl•HCl L1 + L2 80 °C K2CO3–DMF L0 substantially lower pH than for L1 as this is a deprotonation from a tertiary amine rather than the secondary amine in L1.The diVerence in co-ordinating properties between L1 and L2 are evident in the titrations in the presence of CuII. Ligand L1 shows only one sharp inflexion point and one buVer region but L2 has two inflexion points and two buVer regions. The first acidic buVer region in L2 extends up to an inflexion point at 4 equivalents of base and then a more basic buVer region from 4 to 6 equivalents results from the deprotonation of the amide moieties.The first buVer region indicates the formation of a stable complex of L2 with the amide groups remaining protonated which is quite unusual for a macrocyclic amide. Presumably the ligand can co-ordinate through the pyridyl sidearms and the two ring amines to form a stable CuII complex. The structure of this species was investigated further in the spectrophotometric study below. Processing of the titration data in HYPERQUAD15 revealed the presence of a minor monodeprotonated L2 complex as well as the dideprotonated L2–CuII complex as the major species at high pH.Ligand L1 shows a rather diVerent species composition and distribution. The monodeprotonated CuII complex of L1 was a more important species and there was no neutral L1 complex although there was a protonated CuII complex. This will be due to the rather weaker complexing ability with only one pyridine side arm. The single armed macrocycle is unable to stabilize a neutral ligand complex with CuII outside the macrocyclic ring.However at low enough pH protonation of the ring enables CuII to be bound outside the ring through the pyridyl and one ring amine. At higher pH the metal pops into the ring with concomitant deprotonation of the macrocyclic amides. Fig. 1 Titration of L1 (1 × 1023 M) with and without addition of CuII (3 × 1023 to 8 × 1024 M) with standard NaOH (0.100 M) Fig. 2 Titration of L2 (1 × 1023 M) with and without addition of CuII (3 × 1023 to 8 × 1024 M) with standard NaOH (0.100 M)2250 J.Chem. Soc., Dalton Trans., 1998, Pages 2247–2252 The equilibria are collected in Table 1 and the calculated species distribution diagrams for the CuII complexes of L1 and L2 are shown in Figs. 3 and 4, respectively. Spectrophotometric titration. The rather unusual coordinating behaviour of L2 with CuII prompted us to make a spectrophotometric study from which we could propose structures for the species identified in the potentiometric study.Under approximately the same conditions as for the potentiometric titrations, a visible spectrum (wavelength range 850–400 nm) was taken after each aliquot of standard sodium hydroxide was added. The absorbances were converted to molar extinctions and plotted out in Fig. 5. Unfortunately the changes in the absorbances were either insuYciently large or the accuracy of the absorbance measurements were not good enough to make a full HYPERQUAD calculation on the data, however, input of the potentiometrically determined equilibrium constants as a model for the solution chemistry allowed us to deconvolute the spectra in HYDRASP, a sub-program of HYPERQUAD,15 to give molar absorbances of the four absorbing species in Fig. 3 Speciation curves of the complexes of L1 in the presence of 1 equivalent of CuII Fig. 4 Speciation curves of the complexes of L2 in the presence of 2 equivalents of CuII Table 1 Potentiometrically determined equilibrium constants for the reaction of L1 or L2 with CuII and hydrogen ion * Reaction L1 1 H L1H L1H 1 H L1H2 Cu 1 L1 CuH21L1 1 H Cu 1 L1 CuH22L1 1 2H L2 1 H L2H L2H 1 H L2H2 L2H2 1 H L2H3 Cu 1 L2 CuL2 CuL2 CuH21L2 1 H CuL2 CuH22L2 1 2H OH2 H 1 OH Equilibrium quotient, K [L1H]/[L1][H] [L1H2]/[L1H][H] [CuH21L1][H]/[Cu][L1] [CuH22L1][H]2/[Cu][L1] [L2H]/[L2][H] [L2H2]/[L2H][H] [L2H3]/[L2H2][H] [CuL2]/[Cu][L2] [CuH21L2][H]/[CuL2] [CuH22L2][H]2/[CuL2] [OH][H]/[OH2] log K 9.154(2) 13.978(3) 5.77(2) 0.64(2) 7.190(7) 5.12(1) 1.86(5) 11.10(2) 26.38(4) 212.03(3) 213.79 * Charges omitted for clarity.solution. A graph of molar absorbances vs. wavelength for the absorbing species is shown in Fig. 6. The graph shows more clearly what can be inferred by inspection of the spectra in Fig. 5. The deprotonation of the amides result in a blue shift of the spectrum as the ligand field is increased. What is interesting however is the similarity of the molar absorbances of the neutral ligand complex and that for the monodeprotonated amide complex. This suggests a similarity of co-ordination geometry of the two species and indicates that CuII is also located outside the macrocyclic cavity in the monodeprotonated amide macrocycle as well.Possibly the neutral and monodeprotonated L2– CuII complexes have the structure given in Fig. 7, where CuII is in a distorted trigonal bipyramidal co-ordination environment chelated by the pyridyl sidearms, the macrocycle ring amines and the oxygen of one of the amide moieties. A deprotonation of the second amide results in the metal entering the ring to Fig. 5 Absorption coeYcient vs. wavelength for spectra taken over the course of a titration of L2 in the presence of CuII with standard NaOH Fig. 6 Calculation of absorption coeYcients of individual CuII complex species of L2 Fig. 7 Proposed structures of CuL2, Cu(H21L2) and Cu(H22L2) HN NH N N O O N N NH N O O N N Cu N N H Cu2+ –H+ NH N O O N N Cu N N N N N N N N Cu –O O – –H+J. Chem.Soc., Dalton Trans., 1998, Pages 2247–2252 2251 form the macrocyclic complex resulting in a strongly blueshifted spectrum commonly found with macrocyclic amides. Description of crystal structure The molecular structure of complex 1 is shown in Fig. 8. The important bond lengths and angles are listed in Table 2. From Fig. 8, it is obvious that in complex 1, the Cu atom is five-coordinate with O(2), N(1), N(4), N(16), N(26) and forms a distorted trigonal-bipyramidal configuration.The two amide nitrogens [N(2), N(3)], remained nondeprotonated and do not take part in the co-ordination with CuII. The Cu(1)]N(4) bond distance of 2.127(6) Å, is the longest one among the coordination bonds indicating weaker co-ordination due to Jahn– Teller eVects. The CuII ion is nearly in the plane formed by O(2), N(4), N(26). One of the oxygen atoms in the dioxocyclam backbone [O(2)] co-ordinates to Cu(1). This may be the first example of CuII complexes of macrocyclic dioxotetraamines in which a backbone oxygen co-ordinates to the central metal ion by twisting the backbone.The distances of N(2)]C(33) [1.309(10) Å] and N(3)]C(35) [1.343(11) Å] are obviously shorter than a normal C]N distance (1.47 Å) and show partial double bond character which may arise from the conjugation between O(2)]C(33)]N(2) and O(1)]C(35)]N(3). The dihedral angle between the pyridine pendants is 71.48, which means that two pyridine pendants are in close proximity perpendicularly to each other when Fig. 8 The ORTEP16 drawing of complex 1 with 35% probability thermal ellipsoids Table 2 Selected bond lengths (Å) and bond angles (8) for complex 1 with estimated standard deviations in parentheses Cu(1)]O(2) Cu(1)]N(4) Cu(1)]N(26) O(2)]C(33) N(1)]C(31) N(2)]C(32) N(3)]C(35) O(2)]Cu(1)]N(1) O(2)]Cu(1)]N(16) N(1)]Cu(1)]N(4) N(1)]Cu(1)]N(26) N(4)]Cu(1)]N(16) Cu(1)]N(1)]C(20) C(31)]N(1)]C(40) C(35)]N(3)]C(36) Cu(1)]N(16)]C(15) Cu(1)]N(26)]C(25) 2.092(6) 2.127(6) 2.061(7) 1.246(10) 1.518(11) 1.453(11) 1.343(11) 86.6(2) 90.5(3) 100.8(3) 82.4(3) 82.9(3) 104.6(5) 106.4(6) 125.2(7) 123.0(5) 130.7(5) Cu(1)]N(1) Cu(1)]N(16) O(1)]C(35) N(1)]C(20) N(1)]C(40) N(2)]C(33) N(3)]C(36) O(2)]Cu(1)]C(4) O(2)]Cu(1)]C(26) N(1)]Cu(1)]C(16) N(4)]Cu(1)]C(26) N(16)]Cu(1)]N(26) Cu(1)]N(1)]C(40) C(32)]N(2)]C(33) Cu(1)]N(4)]C(10) Cu(1)]N(26)]C(21) O(2)]C(33)]C(34) 2.014(6) 1.971(6) 1.215(10) 1.484(10) 1.505(11) 1.309(10) 1.453(10) 133.9(2) 101.9(2) 176.2(3) 124.1(3) 95.9(3) 109.2(5) 121.2(7) 103.1(5) 110.8(5) 119.7(7) co-ordinating to CuII.The dihedral angle between O(2)]N(4)] N(26) and the co-ordinated pyridine plane N(16)]C(11)]C(15) is 98.968, the other dihedral angle between O(2)]N(4)]N(26) and N(26)]C(21)]C(25) is 66.008. In the unit cell of complex 1, a perchlorate ion links with the macrocycle of the complex through hydrogen bonding. Electrochemical studies The cyclic voltammograms of the CuII complexes of L0 and L1 were examined in aqueous solution (0.5 M Na2SO4) at 25 8C, and the electrochemical data are summarized in Table 3.The cyclic voltammograms showed one quasi-reversible oxidation wave. The CuIII/II potential of Cu(H22L1), 10.83 V vs. SCE, is 0.19 V more positive than that for Cu(H22L0) under similar conditions (E2� 1 = 0.64 V vs. SCE),1a implying that the ligand L appended with 2-pyridylmethyl destabilizes the CuIII state compared with the unsubstituted L0.This behaviour can be interpreted as follows: the change from CuII state (d9) to CuIII state (d8, low spin) involves a drastic reduction of the metal ion radius and a change of electronic configuration (Scheme 2).6 In the new complex, the pyridine pendant co-ordinates to the central CuII from the apical site, and the CuII ion resides above the mean plane of the four basal nitrogens. The co-ordination of one pyridine pendant to CuII stabilizes the CuII ion. But when CuII is oxidized to CuIII, like NiII (d8, low spin), CuIII tends to adopt a square-planar co-ordination rather than a five-co-ordinate one, which means the pyridine pendant will not co-ordinate to CuIII.Since Nsubstitution by the pyridine pendant(s) increases the steric constraint of the macrocyclic ring and lowers the co-ordinative ability of the macrocycle, the CuIII ion in the new ligands are not stabilized to the same extent as in the unsubstituted dioxocyclam.The results are similar to that of the CuII complex of dioxocyclam appended with 8-methylquinoline.11 ESR Studies Fig. 9 represents the ESR spectra of Cu(H22L1) in MeOH solution at room temperature and 112 K in higher pH regions (in this region the doubly-deprotonated CuII complexes are formed). For complex Cu(H22L1), it can be seen that the spectrum of the complex is split into four equally spaced absorptions by the interaction with the CuII nucleus (I = ��� ) [Fig. 9(a)] Scheme 2 –N N– N HN CuII O O N co-ordinated to CuII d9 –N N– N HN CuIII O O N non-co-ordinated to CuIII d8 low spin square planar oxidation reduction distorted square pyramidal Table 3 Electrochemical data of the CuII complexes of L0 and L1 * Ligand pH Epa/V Epc/V DEp/mV E2� 1 /V L0 ª5.5 0.58 0.71 130 0.64 L1 ª6 0.77 0.88 90 0.83 * Cyclic voltammograms in H2O at 25 8C, in V vs.SCE; scan rate 100 mV s21. The concentration of the complexes of L0 and L1 were kept at 2 × 1023 M.Reversible redox voltammograms have not been obtained for CuII–L2 complexes.2252 J. Chem. Soc., Dalton Trans., 1998, Pages 2247–2252 at room temperature. The isotropic ESR parameters are listed in Table 4. When these solutions are frozen at 112 K, ESR spectra characteristic of nearly axial symmetry are observed, which are very similar to that of Cu(H22L0) and Cu(H22L2) complexes. The approximate ESR parameters of complex Cu(H22L1) are graphically evaluated as g|| = 2.188, g^ = 2.050, A|| = 188 G (1.933 × 1022 cm21), and A^ = 28 G (2.70 × 1023 cm21), where g^ = (3giso 2 g||)/2 and A^ = (3Aiso 2 A||)/2.17 It is obvious that the observed A|| values decrease and g|| values increase from Cu(H22L0) to Cu(H22L1) and Cu(H22L2).The tendencies for A|| to decrease and g|| to increase have been taken as parameters to measure the lowering of the strength of inplane ligand fields under the tetragonal basal square arrangement of CuII complexes.18 Therefore, ESR spectra also support the weakened in-plane bonding in complexes Cu(H22L1) and Cu(H22L2). This is in consistent with the CV measurements.The ESR parameters of these complexes also indicate a dx2 2 y2 ground-state of CuII in these complexes. Acknowledgements We gratefully acknowledge the financial support from the National Natural Science Foundation of China (No. 29771022) and Tianjin Natural Science Foundation, China (to X. H. B.). The ESR spectra and CV were carried out at the National Key Laboratory on Coordination Chemistry in Nanjing University, China.References 1 (a) E. Kimura, J. Coord. Chem., 1986, 15, 1; (b) E. Kimura, Pure Appl. Chem., 1986, 58, 1461; (c) E. Kimura, Crown Compounds Toward Future Applications, ed. S. R. Cooper, VCH, New York, 1992, ch. 6. Fig. 9 The X-band ESR spectra of complex Cu(H22L1) in MeOH at (a) 298 K, and (b) 112 K Table 4 The ESR parameters of the CuII complexes of L0, L1 and L2 in methanol solution at room temperature and 112 K * 298 K 112 K Ligand L0 L1 L2 giso 2.086 2.096 2.103 Aiso/G (103 cm21) 96 (9.41) 81 (7.98) 72 (7.12) g|| 2.173 2.188 2.194 g^ 2.043 2.050 2.058 A||/G (102 cm21) 208 (2.124) 188 (1.933) 184 (1.897) A^/G (103 cm21) 40 (3.84) 28 (2.70) 16 (1.65) * A^ was calculated according to the formula 3Aiso = A|| 1 2A^ and g^ according to the formula 3giso = g|| 1 2g^. 2 (a) M.Kodama and E. Kimura, J. Chem. Soc., Dalton Trans., 1979, 325; (b) M. Kodama and E. Kimura, ibid., 1979, 1783; (c) M.Kodama and E. Kimura, ibid., 1981, 694; (d ) K. Ishizu, J. Hirai, M. Kodama and E. Kimura, Chem. Lett., 1979, 1045; (e) R. Machida, E. Kimura and M. Kodama, Inorg. Chem., 1983, 22, 2055; ( f ) E. Kimura, T. Koike, R. Machida, R. Nagai and M. Kodama, Inorg. Chem., 1984, 23, 4181; ( g) E. Kimura, C. A. Dalimunte, A. Yamashita and R. Machida, J. Chem. Soc., Chem. Commun., 1985, 1041; (h) E. Kimura, Y. Lin, R. Machida and H. Zenda, ibid., 1986, 1020; (i) E.Kimura, S. Korenari, M. Shionoya and M. Shiro, ibid., 1988, 1166; ( j) E. Kimura, M. Shionoya, M. Okamoto and H. Nada, J. Am. Chem. Soc., 1988, 110, 3679; (k) M. Shionoya, E. Kimura and Y. Iitaka, ibid., 1990, 112, 9237. 3 R. W. Hay, M. P. Pujari and F. McLaren, Inorg. Chem., 1984, 23, 3033; L. Fabbrizzi, Inorg. Chem., 1985, 4, 33; L. C. Siegfried and T. A. Kaden, J. Phys. Org. Chem., 1992, 549; L. Fabbrizzi, A. Perotti, A. Profumo and T. Soldi, Inorg. Chem., 1986, 25, 4256; L.Fabbrizzi, F. Forlini, A. Perotti and B. Seghi, Inorg. Chem., 1984, 23, 807; L. Fabbrizzi, T. A. Kaden, A. Perotti, B. Seghi and L. Seigfried, Inorg. Chem., 1986, 25, 321; G. D. Santis, L. Fabbrizzi, M. Licchelli and P. Pallavicini, Coord. Chem. Rev., 1992, 120, 237. 4 E. Kimura and H. Nada, Tanpakushitsu Kakusan Koso, 1988, 16, 2914; R. Machida, E. Kimura and M. Kodama, Inorg. Chem., 1983, 22, 2055; E. Kimura and R. Machida, Yuki Gosei Kagaku Kyokaishi, 1984, 42, 407. 5 B. J. Hathaway, in Comprehensive Coordination Chemistry, ed. G. Wilkinson, Pergamon Press, Oxford, 1987, 5, p. 533. 6 X. H. Bu, D. L. An, Y. T. Chen, M. Shionoya and E. Kimura, J. Chem. Soc., Dalton Trans., 1995, 2289; X. H. Bu, X. C. Cao, D. L. An, R. H. Zhang, T. CliVord and E. Kimura, ibid., 1998, 433. 7 L. Fabbrizzi and A. Poggi, J. Chem. Soc., Chem. Commun., 1980, 646; L. Fabbrizzi, A. Perotti and A. Poggi, Inorg. Chem., 1983, 22, 1411; L. Fabbrizzi, A. Perotti, A. Profumo and T.Soldi, Inorg. Chem., 1986, 25, 4256; R. Hay, R. Rembi and W. Sommerville, Inorg. Chim. Acta, 1982, 59, 147; Y. D. Lampeka and S. P. Gavrish, J. Coord. Chem., 1990, 21, 351. 8 D. W. Margerum and G. D. Oven, in Metal Ions in Biological Systems, ed. H. Sigel, Marcel Dekker, New York, 1981, vol. 12, p. 75; T. R. Wagner and C. J. Burrows, Tetrahedron Lett., 1988, 29, 5091; T. R. Wagner, Y. Fang and C. J. Burrows, J. Org. Chem., 1989, 54, 1584. 9 See, for example, E. Kimura, J. Inclusion Phenom., 1989, 7, 183; E. Kimura, Pure Appl. Chem., 1989, 61, 823; E. Kimura, Tetrahedron, 1992, 48, 6175; P. V. Bernhardt and G. A. Lawrence, Coord. Chem. Rev., 1990, 104, 297; T. A. Kaden, Crown Compounds Toward Future Applications, ed. S. R. Cooper, VCH, New York, 1992, ch. 8; E. Kimura, X. H. Bu, M. Shionoya, S. Wada and S. Maruyama, Inorg. Chem., 1992, 31, 4542; X. H. Bu, Y. T. Chen, M. Shionoya and E. Kimura, Polyhedron, 1994, 13, 325; E. Kimura, S. Wada, M. Shionoya and Y. Okazaki, Inorg. Chem., 1994, 33, 770. 10 E. Kimura, T. Koike, R. Machida, R. Nagai and M. Kodama, Inorg. Chem., 1984, 23, 4181; E. Kimura, T. Koike, H. Nada and Y. Iitaka, ibid., 1988, 27, 1036. 11 X. H. Bu, D. L. An and Y. T. Chen, M. Shionoya, E. Kimura, J. Inclusion Phenom., 1997, 27, 245. 12 I. Tabushi, H. Okino and Y. Kuroda, Tetrahedron Lett., 1976, 4339. 13 E. Kimura, T. Kioke, K. Uenishi, M. Hedigar, M. Kuramoto, S. Joko, M. KY. Iitaka, Inorg. Chem., 1987, 26, 2975. 14 G. M. Sheldrick, SHELXTL PC, Siemens Analytical X-Ray Instruments, Inc., Madison, WI, 1990. 15 P. Gans, A. Sabatini and A. Vacca, Talanta, 1996, 43, 1739. 16 C. K. Johnson, ORTEP, Report ORNL-5138, Oak Ridge National Laboratory, Oak Ridge, TN, 1976. 17 K. Miyoshi, H. Tanaka, E. Kimura, S. Tsuboyama, S. Murata, H. Shimizu and K. Ishizu, Inorg. Chim. Acta, 1983, 78, 23. 18 A. S. Brill, Molecular Biology Biochemistry and Biophysics. 26. Transition Metals in Biochemistry, ed. A. Kleinzeller, Springer- Verlag, New York, 1977, p. 43. Received 5th December 1997, revised manuscript received 25th March 1998; Paper 8/02327A

ISSN:1477-9226    

DOI:10.1039/a802327a

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

DALTON J. Chem. Soc., Dalton Trans., 1997, Pages 2249–2255 2249 Regioselective reactions of isothiocyanates with the titanocene vinylidene fragment [Ti(] C] CH2)(Á-C5Me5)2]. Crystal and molecular structure of [Ti{SC(] NC6H11)C] CH2}(Á-C5Me5)2] † Rüdiger Beckhaus,*,a Javier Sang,a Trixie Wagner a and Uwe Böhmeb a Department of Inorganic Chemistry, Technical University Aachen, Professor-Pirlet-Strasse 1, D-52062 Aachen, Germany b Department of Inorganic Chemistry, Leipziger Strasse, Technical University Bergakademie Freiberg, D-09596 Freiberg, Germany The titanocene vinylidene intermediate [Ti(]] C]] CH2)(h-C5Me5)2] 1 formed by ethane or methane elimination from [Ti(CH2CH2C]] CH2)(h-C5Me5)2] 2 or [Ti(CH]] CH2)Me(h-C5Me5)2] 3 respectively, reacted with isothiocyanates RNCS (R = C6H11 a, Ph b or But c) by a [2 1 2] cycloaddition, to give the titanathietane complexes [Ti{SC(]] NR)C]] CH2}(h-C5Me5)2].The crystal structure of the R = C6H11 complex has been determined.In all cases the regioisomer in which the sulfur atom is bonded to titanium is observed as the primary product. Upon heating in the presence of pyridine a rearrangement to the regioisomeric titanacyclobutane derivative [Ti{C(]] NR)SC]] CH2}(h-C5Me5)2] was observed. The regioselectivity of the formation of the complexes 5 and 7 is discussed on the basis of ab initio calculations at the Hartree–Fock level of theory, with an effective core potential basis set. Recent interest in ambidentate ligands, especially the thiocyanate ion which can co-ordinate via the sulfur or the nitrogen atoms or both, probably results from two principal considerations.The ambidentate nature of SCN2 may be interpreted in terms of sulfur being a ‘soft’ and nitrogen a ‘hard’ base. When it is the only ligand present in a complex its mode of bonding generally follows the hard (M]NCS) or soft (M]SCN) pattern throughout the Periodic Table. However, the nature of other ligands in a complex may determine whether the metal functions as a hard ion and forms isothiocyanato complexes, or as a soft ion and forms thiocyanato complexes.Steric factors in bulky ligands also may alter the nature of thiocyanate coordination. 1 In the case of electron-poor transition metals the N-bonding mode of the thiocyanate ion is dominant,2 but also examples of S-bonding modes are known.2b Additionally to the ambivalent character in different end-on co-ordination modes of the SCN2 ion, side-on co-ordinations are known for isothiocyanates RSCN.3 The ambivalent behaviour can also be observed in [2 1 2] cycloaddition reactions of isothiocyanates and metal–ligand double bonds.Thus C]] N4 and C]] S addition products 4c,5a–c are found. The large variety of reactions involving the titanocene vinylidene intermediate [Ti(]] C]] CH2)(h-C5Me5)2] 1 6 and carbon dioxide, ketenes, isocyanates,7 transition-metal carbonyls,8 nitriles and phosphaalkynes,9 alkynes 10a,b and carbodiimides 11 leads to four-membered titanacycles 4 of high thermal stability (Scheme 1).The regioselectivity can be explained in accordance with the polarities of the unsaturated compounds used in the cycloadditions towards the polarized titanium–carbon double bond (Tid1]] Cd2) of 1. Large differences in the partial charge of the unsaturated substrates (e.g. isocyanates, nitriles, alkynes) lead to stereochemically pure compounds, the more negative carbon being bonded to titanium.7,9,10a,b Regioisomers are obtained by using substrates with small differences in their partial charges.In a series of such studies we were interested in the behaviour of isothiocyanates RNCS towards the vinylidene intermediate 1, in order to determine the accessibility of titanathietanes 5. Owing to lower differences in the polarity of the † Non-SI unit employed: Eh ª 4.36 × 10218 J. NCS unit in isothiocyanates, compared to isocyanates, the formation of the regioisomers 6 and 7 should also be possible.Results and Discussion The vinylmethyl derivative 3 reacts with 1 equivalent of isothiocyanates RNCS (R = C6H11 a, Ph b or But c) at room temperature by liberation of methane to give the metallacyclobutanes 5a–5c (Scheme 2) which can be isolated as brown crystals of high thermal stability [m.p. 110 (decomp.) 5a, 139–140 (decomp.) 5b and 129 8C (decomp.) 5c]. The products 5a and 5b with the sulfur atom in the a position are formed exclusively.When using ButNCS the reaction product contains 5c and the regioisomer 7c in a 10 : 1 ratio. Isomers of 5, indicating [2 1 2] cycloadditions involving the C]] N bond (forming 6), are not found during the synthesis. The mass spectra of 5 exhibit the expected molecular peaks and the formation of a (h-C5Me5)2- Ti]] S fragment in a first step, similar to the fragmentation behaviour of the titanaoxetanes 4b, where a (h-C5Me5)2Ti]] O fragment is also observed.The 1H and 13C NMR data for the titanathietanes 5 are listed Scheme 1 (h-C5Me5)2Ti X Y (h-C5Me5)2Ti Me Heat Heat – C2H4 – CH4 2 3 (h-C5Me5)2Ti C CH2 1 X Y (h-C5Me5)2Ti X Y OOONP CR NR C=O C=NR C=ML CR CR CR C=NR n 4 4a 4b 4c 4d 4e 4f 4g2250 J. Chem. Soc., Dalton Trans., 1997, Pages 2249–2255 Table 1 NMR Data for the titanathietanes 5a C2 (h-C5Me5)2Ti S C3 C1 N R H2 H1 1H (JHH/Hz) 13C R C5Me5 H1 (endo) H2 (exo) Db C5Me5 C1 C2 C3 5a 5b 5c C6H11 Ph But 1.69 1.64 1.68 3.91 (d, 2.8) 4.06 (d, 2.4) 3.74 (d, 2.8) 6.92 (d, 2.8) 7.06 (d, 2.4) 6.97 (d, 2.8) 3.01 3.00 3.23 12.4 124.6 12.4 125.0 12.7 124.9 110.1 112.2 109.5 196.0 195.9 195.7 149.7 153.4 149.0 a Listed in ppm vs.SiMe4; solvent was C6D6 and temperature = 25 8C. b D = |d(H2) 2 d(H1)|. Table 2 Selected bond distances (Å) and angles (8) in complex 5a compared to those of similar structure types Ti C=NC6H11 X e b a d c a b d g S Ti X S b a d c a b d g 5a (X = S) 4b (X = O) S Ti S b a d R 10 S S S S aTi S c a d 11 Ti SH SH b c a d b a b a 12 8 (X = S) 9 (X = SiMe2) Complex a b c d e Ref. 5a 4b 89 10 11 12 2.466(1) 1.983(2) 2.413(4) 2.454(1) 2.418(3) 2.422(1) 2.409(2) 2.156(3) 2.121(3) 2.413(4) 2.454(1) 2.389(3) 2.448(1) 2.418(3) 1.484(5) 1.477(4) 2.041(5) 2.110(1) 1.76(1) 2.059(2) — 1.795(3) 1.348(3) 2.041(5) 2.110(1) 1.76(1) 2.058(1) — 1.320(4) 1.325(4) ————— This work 7(a) 13(a) 13(b) 13(c) 13(d) 14 a b g d Ref. 5a 4b 89 10 11 12 70.45(9) 67.6(1) 84.44(9) 87.1 81.4(1) 94.6 94.8(1) 99.5(2) 87.2(2) 76.34 83.4 97.4(4) —— 109.2(2) 107.9(2) 105.3(1) 105.5 ——— 80.9(1) 96.7(2) 76.34 83.4 95.9(4) —— This work 7(a) 13(a) 13(b) 13(c) 13(d) 14 in Table 1.Especially the value of the difference D of the chemical shifts of H1 and H2 in 5a–5c appears to be a characteristic feature of the new complexes. The low-field signals in the 13C NMR spectra are consistent with the metal-bonded C2 atom (d 195.7–196.0), whereas the chemical shifts of C1 and C3 are in the expected range.The structure of compound 5a was confirmed by X-ray analysis. The ORTEP12 plot is shown in Fig. 1, relevant bond distances and angles, compared to those of similar structure (h-C5Me5)2Ti S C N R (h-C5Me5)2Ti N C S (h-C5Me5)2Ti C S R N R 6 5 7 types, in Table 2. The crystal chemistry of this and related titanacyclobutanes (4b,7a 4d, 4e 9) will be the subject of a forthcoming paper.7b The geometry of 5a shows a planar titanacycle. Thietanes are characterized by an angle of 154(4)8 between the CCC and the CSC planes.15 The solid-state structural data for 5a are consistent with a metallacycle formalism.The Ti]C(2) bond is longer as in the titanacyclobutane 2 [2.068(6) Å] 6c and also longer than in titanacyclobutenes [4f (R = Me), 2.104(3) Å].10a More interesting is the slight elongation of the Ti]S bond compared to those of other sulfur-containing metallacycles 8–10 13 and particularly to the non-cyclic titanocene sulfide [Ti(SH)2(h-C5Me5)2] 12.14 Generally, the small ring size of 3 + R N C S – CH4 (h-C5Me5)2Ti S C N R C6H11 Ph But 5a 5b 5c R Scheme 2J.Chem. Soc., Dalton Trans., 1997, Pages 2249–2255 2251 heterotitanacyclobutanes Ti]X]CC (X = O, N or S) leads to longer Ti]X distances as in larger rings or non-cyclic compounds, due to the lowering of XÆTi p-bonding interactions as a result of smaller Ti]X]C angles in four-membered rings. Thus the Ti]S bond distances increase from 12 [2.409(2)] to the fivemembered rings 11 and 10 [2.422(1), 2.418(3)] and to the fourmembered rings 8, 9 and 5a [2.413(4), 2.454(1) and 2.466(1)].The exocyclic C]] C double bond C(1)]C(2) [1.320(4) Å] is relatively short, compared to those in 4f [1.377(4) Å, R = Me,10a 1.342(5) Å, R = CCSiMe3 or SiMe3 10b], but similar to 4d [1.337(3)], 4e [1.326(8)] 9 and 4f [1.322(8), R = SiMe3 or Ph].10a The value of the distance e is indicative of the reactivity of the methylenetitanacycles. A long distance characterizes stable complexes, a shorter distance points to a tendency to cycloreversion reactions (forming 1), as found for 2 (e = 1.321 Å),6c 4e and 4f.Additionally the C(2)]C(3) bond in 5a is longer than in 4b. As a consequence of the larger sulfur atom in 5a compared to the oxygen in the oxetanes 4b, a longer distance Ti ? ? ? C(3) of 2.881(3) in 5a compared to 2.52 Å in 4b is found. The value of the angle a in 5a (Table 2) is in the expected range for fourmembered titanacycles 6a exhibiting a lower value compared to the more sulfur-rich compounds 8 and 9 and especially to larger rings or non-cyclic titanocene derivatives like 11 and 12, as discussed before.The orientation of the substituents in the cyclobutanes can be attributed to the polarity of the isothiocyanate molecule and the strongly nucleophilic a-C atom in the vinylidene 1 (Tid1]] Cd2]] CH2). However, the nitrogen atom in RNCS exhibits the most negative partial charge (Pauling electronegativity values: S, 2.5; C, 2.5; N, 3.0) and a N-co-ordination mode A is expected.Calculated electron densities for the optimized structure of MeNCS (Mulliken values, basis set 6-31G*) confirmed that the most negative charge was on the nitrogen atom (S, 20.21; C, 10.31; N, 20.45). On the other hand, the primary coordination mode A exhibits a larger space requirement (a) compared to the S-co-ordination mode B. In particular, the rod-like shape of isothiocyanate molecule RNCS seems to be the reason for a preferred S-end-on co-ordination mode and formation of the CS-cycloaddition products 5 instead of a CN cycloaddition to the regioisomer 6.Furthermore, isothiocyanates normally undergo nucleophilic attack at the carbon Fig. 1 An ORTEP drawing of [Ti{SC(]] NC6H11)C]] CH2}(h-C5Me5)2] 5a (30% ellipsoids) S (h-C5Me5)2Ti S S S (h-C5H4Me)2Ti S SiMe2 S (h-C5H5)2Ti S R SH (h-C5Me5)2Ti SH 8 9 10 S (h-C5H5)2Ti S S S S 11 12 R = C(O)N(H)CH2Ph atom and not, like many other carbon–sulfur double bondcontaining molecules, thiophilic attack.16 Heating compound 5a in the presence of pyridine at 80 8C for 20 min results in isomerization to 7 (85%).The pyridine is not incorporated in the reaction product, but without it no isomerization takes place. As mentioned before, the reaction of 1 with ButNCS leads to a small amount of 7c (5c : 7c = 10 : 1), which cannot be separated. Owing to the relatively long Ti]S and the C(2)]C(3) bond distances in 5a, a reactive Ti]S bond is expected, indicating the possibility of a cycloreversion of 5.Thus the formation of ring-opened intermediates 13 and 14 seems to be possible, which can rearrange to the second CScycloaddition product 7. During the isomerization the colour changes from brown to light yellow. The NMR spectra clearly indicate the four-membered ring structure of 7, which can be seen from the exo-CH2 group at d 4.19 and 7.91 for 7a. In the 13C NMR spectrum a second titanium-bonded carbon atom can be observed (d 204.5, 7a), instead of the signal d 149.7 for the b-C]] N carbon atom in 5a.To understand the reaction course 5 æÆ 7, the geometries of the titanium complexes 15, 16 and 17/18 were investigated by ab initio calculations at the restricted Hartree–Fock (RHF) level of theory (see Experimental section). In our ab initio calculations the C5Me5 groups of the real molecules 5 and 7 were replaced by chloride ligands, which has been shown in other studies to provide a good theoretical substitute for the actual bent metallocene system.17 The optimized geometries and the atomic charge distributions and bond-overlap populations of 15, 16 and 18a are shown in Figs. 2 and 3. The geometry of 16 is in good agreement with the structure found for 5a. The total energies of the optimized structures 15, 16 and 18a are shown in Table 3. The calculation of single-point energies by secondorder Møller–Plesset perturbation (MP2) results in a lowering of the total energy of these systems.This stabilization effect is similar for 16 and 18a. The smaller molecule 15 shows a smaller effect, as expected. Scheme 3 (h-C5Me5)2Ti C CH2 S C NR (h-C5Me5)2Ti C CH2 RN C S 7a 5a Heat 13 14 Ti S CH2 Cl Cl Ti S C Cl Cl Ti C S R R N Me N Me Ti S R R C N Me 15 16 17a = Cl 17b = C5H5 18a = Cl 18b = C5H52252 J. Chem. Soc., Dalton Trans., 1997, Pages 2249–2255 Table 3 Total energies (E) and energy differences of complexes 15, 16 and 18a [Eh; kJ mol21 (in parentheses)] 15 16 18a E(16) 2 E(18a) ERHF EMP2/RHF ERHF 2 EMP2/RHF 2115.582 53 2116.071 11 20.488 58 2131.259 45 2131.921 67 20.662 22 2131.284 75 2131.938 37 20.653 62 0.025 3 (66.4) 0.016 7 (43.8) Whereas structure 16 converged well during optimization, the starting geometry 17a led straightforwardly to the optimized structure 18a.The fully optimized geometry of 18a is more stable than 16 (66.4 kJ mol21 at the restricted Hartree– Fock level, 43.8 kJ mol21 at the MP2 level).The geometries of the four-membered titanacycles 15 and 16 are similar. Replacement of the CH2 group in 15 by a C]] NMe group in 16 leads to elongation of the Ti]C, Ti]S, Ti]C and C]] C (exomethylene group) bond lengths. The bond lengths S]C and C]C are shorter in 16 by 0.048 and 0.038 Å, respectively. This reflects the peculiar effect of the C]] NMe group on the titanathietane ring in 16: on one hand there are longer bonds between titanium and its ligands, on the other shorter bonds between the C]] NMe group and its neighbours.The repulsive interaction Fig. 2 Optimized geometries of complexes 15, 16 and 18a at the RHF level; distances in Å (20.362) between the a- and b-carbon in the four-membered ring is the most striking feature of the Mulliken-overlap population analysis of 16 (Fig. 3). Therefore it becomes probable that the kinetic product 16 undergoes a ring-opening reaction at the internal C]C bond with rearrangement to the thermodynamic stable product 18a.It was found in our experiments that 5a rearranges to 7 upon heating. The structure 18a represents a titanium thioketene complex with co-ordinated methyl isocyanide. Obviously the formation of complex 18a reflects the degradation of isothiocyanates to isocyanides typical of latetransition- metal SCNR complexes,3a,18a,b but also in the case of Fig. 3 Atomic charge distributions and bond-overlap populations of complexes 15, 16 and 18a obtained by Mulliken population analysis at the RHF levelJ.Chem. Soc., Dalton Trans., 1997, Pages 2249–2255 2253 Lewis-acid metallocenes Si(C5Me5)2. 5a The existence of h2-C,S bonded thioketene derivatives 19 as calculated in 18a and comparable structures are also known in the case of titanium complexes. 20 However, in the case of the formation of 7 from 5 there are no hints regarding the presence of a co-ordinated isocyanide in the IR spectra [expected for n(C]] N) in Ti]CN]R complexes with RNC as s donors (2200,21 2260–2310 cm21 22) or as p-acceptor ligands (2038, 1937 cm21 23)].Only the typical n(CN) band of an iminoacyl structure 24 is observed in the case of 7a (1541 cm21), which is only consistent with the model structure 17 and 7 in the real molecules. The mass spectra of 7a shows the expected molecular peak. Comparison of complex 18a with the titanocene thioformaldehyde complex [Ti(h2-SCH2)(h-C5H5)2(PMe3)] 19, which was isolated and structurally characterized some years ago,20a shows that there are great similarities in the co-ordination geometry of the C]S fragment at titanium. The Ti]S and C]S bond lengths exhibit only small differences of 0.056 and 0.039 Å, respectively.The bond length Ti]C in the model complex 18a is shorter than in 19 by 0.244 Å. The angles C]Ti]S are similar with 46.88 in 18a and 43.38 in 19. The differences between 18a and 19 can be attributed to the different donor molecules CNMe and PMe3 and to the substitution of the C5H5 ligand by Cl atoms.To gain a better insight into the influence of the spacer ligand chloride in comparison with any cyclopentadienyl system, we performed the optimization of the cyclopentadienyl-substituted model complexes 17b and 18b with basis set STO-3G.‡ We found that the structure 17b is a minimum in this system for this primitive basis set (Fig. 4). Therefore we conclude that the chloride ligands with their higher electronegativity and different space filling than C5H5 are the reason for the unexpected minimum structure 18a in our Cl2Ti model system. Conclusion We have demonstrated that the reaction of the titanocene Fig. 4 Optimized geometry of complex 17b (RHF level, basis set STO-3G); distances in Å ‡ This basis set and cyclopentadienyl instead of pentamethylcyclopentadienyl ligands for the geometry optimization of 17b and 18b were chosen in order to obtain results in a reasonable time.vinylidene intermediate [Ti(]] C]] CH2)(h-C5Me5)2] 1 with isothiocyanates yields titanathietane complexes 5. Only one regioisomer is formed, where the sulfur atom is in the a position of the metallacycle. The regioselectivity can be explained in terms of the shape of the isothiocyanate and the stereochemistry of 1. On heating in the presence of pyridine, 5 can be isomerized to a second C]] S cycloaddition product 7. C]] N-Cycloaddition products are not observed.The isomerization of 5 to 7 shows that the lower polarity of the C]] S unit in the RNCS molecule allows the formation of a second isomer. In this context the behaviour of titanathietanes is quite different from that of titanaoxetanes, which react to give Ti]] O (classical behaviour) or by cycloreversion to give Ti]] C bonds (non-classical behaviour). 6a,b,8 A similar comparison of different reactivities between homologous compounds can be made between aza- and phospha-titanacyclobutenes.For the former only one regioisomer is observed; the latter exhibit also a,b regioisomers.9 This behaviour can also be attributed to the lower polarity of the C]] X bond used in cycloaddition reactions. Experimental General considerations The preparation and handling of the described compounds were performed under rigorous exclusion of air and moisture under a nitrogen atmosphere, using standard vacuum-line and Schlenk techniques. All solvents were dried with appropriate drying agents and distilled under a nitrogen atmosphere.Deuteriated solvents were degassed by freeze–pump–thaw cycles and dried over molecular sieves (3, 4 Å) prior to use. Proton and 13C NMR spectra were recorded on a Varian Unity 500 spectrometer. Chemical shifts are reported in ppm and referenced to residual protons in deuteriated solvents (C6D6, d 7.15 for 1H NMR, d 126.96 for 13C NMR). Mass spectra were recorded on a Finnigan MAT 95 mass spectrometer, infrared spectra as KBr pellets on a Perkin-Elmer 1720X FT-IR spectrometer.Elemental analyses were carried out at the Analytische Laboratorien in Lindlar, Germany. The titanocene complexes [Ti(CH]] CH2)Me(h-C5Me5)2] 8,25 and [Ti(CH2CH2C]] CH2)(h-C5- Me5)2] 26a,b were prepared by literature procedures. The isothiocyanates were obtained from Aldrich. Preparations (4-Cyclohexylimino)(3-methylene)-2,2-bis(Á5-pentamethylcyclopentadienyl)- 1-thia-2-titanacyclobutane 5a. To a solution of complex 3 (366.0 mg, 1.015 mmol) in hexane (40 cm3) was added cyclohexyl isothiocyanate (150.6 mg, 1.066 mmol) at 278 8C.The mixture was slowly warmed to room temperature and stirred overnight. The hexane was partially removed under reduced pressure until crystallization took place (0 8C), yielding 5a after decantation as brown, octahedral crystals (m.p. 110 8C), suitable for structure determination. Yield 0.265 g, 54% (Found: C, 72.25; H, 9.2; N, 2.75; S, 6.35.C29H43NSTi requires C, 71.75; H, 8.95; N, 2.9; S, 6.6%). NMR (C6D6): 1H (500 MHz), d 1.28–1.37 (1 H, t/m, J 3.5), 1.51–1.58 (2 H, t/m, J 3.5), 1.58–1.65 (1 H, m), 1.80–1.85 (1 H, m), 1.85–1.92 (3 H, m), 2.17 (2 H, m) (all CH2 ring), 1.69 [30 H, s, C5(CH3)5], 3.91 [1 H, d, J 2.8, ]] CHH(cis)], 4.20 (1 H, t/t, J 9.9/4.0, NCH) and 6.92 [1 H, d, J 2.8 Hz, ]] CHH(trans)]; 13C-{1H} (125 MHz), d 12.4 [C5(CH3)5], 25.7, 26.8, 34.2 (all CH2 ring), 60.2 (]] NCH), 110.1 (]] CH2), 124.6 [C5(CH3)5], 149.7 (C]] N) and 196.0 (Ti]C]] ).Electron impact (EI) mass spectrum (111 8C): m/z 485 (M, 28), 460 (40), 350 [(C5Me5)2Ti]] S, 33], 318 [(C5Me5)2Ti, 100], 279 (10), 271 (30), 256 (17), 215 [(C5Me5)Ti]] S, 20], 181 (19), 159 (12), 136 (C5Me5H, 39), 119 (53), 105 (26), 91 (23), 83 (11), 77 (11) and 55 (18%); exact mass 485.2596 (C29H43NSTi), calculated 485.2596. IR (KBr): 3076w, 3022w, 2985w, 2956m, 2927vs, 2848s, 2722w, 1630w, 1539vs [n(C]] N)], 1491m, 1448m, 1379vs, 1343w, 1255w, 1180w, 1165w, 1120m, 1058w, 1019m, 1006m, 979m, 960m,2254 J.Chem. Soc., Dalton Trans., 1997, Pages 2249–2255 902m, 891w, 843w, 792w, 713m, 627w, 619w, 593w, 556m, 475w, 458w, 434w and 406m cm21. (3-Methylene)-2,2-bis(Á5-pentamethylcyclopentadienyl)(4- phenylimino)-1-thia-2-titanacyclobutane 5b. A reaction of complex 3 (271.2 mg, 0.7525 mmol) in hexane (30 cm3) with phenyl isothiocyanate (106.8 mg, 0.7901 mmol) was carried out using the same procedure as in the case of 5a.Owing to the low solubility of 5b in hexane the reaction mixture was filtered, yielding 5b as brown microcrystals, m.p. 139–140 8C (decomp.). Yield 0.310 g, 86% (Found: C, 71.0; H, 8.1; N, 2.7. C29H37NSTi requires C, 72.65; H, 7.8; N, 2.9%). NMR (C6D6): 1H (500 MHz), d 1.64 [30 H, s, C5(CH3)5], 4.06 [1 H, d, J 2.4, ]] CHH- (cis)], 7.01 (1 H, t/m, J 7.2, p-H), 7.06 [1 H, d, J 2.4, ]] CHH(trans)], 7.37 (1 H, t/m, J 8.0, m-H) and 7.50 (2 H, d/m, J 8.5 Hz, o-H); 13C-{1H} (125 MHz), d 12.4 [C5(CH3)5], 112.2 (]] CH2), 122.7 (m-C), 125.0 [C5(CH3)5], 128.3 (p-C), 129.8 (o-C), 153.4 (C]] N), 156.8 (ipso-C) and 195.9 (Ti]C]] ).EI mass spectrum (151 8C): m/z 479 (M, 19), 445 (3), 389 (3), 380 (13), 353 (23), 350 [(C5Me5)2Ti]] S, 9], 345 (9), 318 (23), 317 (26), 297 (13), 279 (14), 265 (100), 250 (63), 235 (16), 218 (60), 194 (6), 167 (21), 162 (36), 147 (51), 134 (68), 130 (57), 121 (47), 119 (92), 105 (43), 93 (30), 91 (36), 77 (48) and 57 (35%); exact mass 479.2128 (C29H37NSTi), calculated 479.2126.IR (KBr): 3074w, 3060w, 3029w, 2959m, 2905s, 1819w, 1593s, 1530vs [n(C]] N)], 1485m, 1431m, 1378vs, 1262s, 1222s, 1165w, 1103m, 1069w, 1021m, 1002w, 991m, 908s, 846m, 803s, 765s, 695vs, 622w, 608m, 548m, 500w and 420m cm21. (4-tert-Butylimino)(3-methylene)-2,2-bis(Á5-pentamethylcyclopentadienyl)- 1-thia-2-titanacyclobutane 5c. A reaction of complex 3 (205.1 mg, 0.5691 mmol) in hexane (20 cm3) with tert-butyl isothiocyanate (68.8 mg, 0.598 mmol) was carried out using the same procedure as in the case for 5a.Similar work-up gave 5c as yellow-brown needles, m.p. 129 8C (decomp.). Yield 0.191 g, 73% (Found: C, 69.6; H, 9.55; N, 2.85. C27H41NSTi requires C, 70.55; H, 9.0; N, 3.05%). NMR (C6D6): 1H (500 MHz), d 1.68 [30 H, s, C5(CH3)5], 1.82 [9 H, s, NC(CH3)3], 3.74 [1 H, d, J 2.8, ]] CHH(cis)] and 6.97 [1 H, d, J 2.8 Hz, ]] CHH(trans)]; 13C-{1H} (125 MHz), d 12.7 [C5(CH3)5], 29.8 [C(CH3)3], 54.8 [C(CH3)3], 109.5 (]] CH2), 124.9 [C5(CH3)5], 149.0 (C]] N) and 195.7 (Ti]C]] ).EI mass spectrum (91 8C): m/z 459 (M, 8), 402 (4), 356 (13), 350 [(C5Me5)2Ti]] S, 19], 337 (22), 318 (76), 317 (76), 279 (8), 277 (9), 268 (12), 221 (29), 202 (24), 149 (27), 143 (100), 135 (72), 121 (47), 119 (91), 105 (35), 91 (30), 88 (27), 87 (30), 71 (33) and 57 (47%); exact mass 459.2441 (C27H41NSTi), calculated 459.2439. IR (KBr): 3040w, 2962vs, 2901vs, 1808w, 1556vs [n(C]] N)], 1490m, 1475w, 1452m, 1432m, 1379vs, 1351s, 1262w, 1227s, 1209s, 1109s, 1065w, 1020s, 962s, 915w, 901s, 803m, 772w, 711w, 651w, 595w, 556m, 539m, 505w and 470w cm21.(2-Cyclohexylimino)(4-methylene)-3,3-bis(Á5-pentamethylcyclopentadienyl)- 1-thia-3-titanacyclobutane 7a. Complex 5a (140 mg, 0.288 mmol) was dissolved in pyridine (30 cm3) (dried over KOH and distilled before use) and heated to 80 8C for 15 min. During that time the mixture changed from brown-red to vivid yellow. The pyridine was partially removed under reduced pressure until crystallization took place, yielding 7a as a yellow powder.Yield 85% according to NMR spectroscopy. NMR (C6D6): 1H (500 MHz), d 1.01–1.11 (1 H, m), 1.13–1.22 (1 H, m), 1.25–1.35 (1 H, m), 1.45–1.51 (3 H, m), 2.24 (4 H, m) (all CH2 ring), 1.67 [30 H, s, C5(CH3)5], 4.19 [1 H, d, J 3.0, ]] CHH(cis)], 4.86 (1 H, t/t, J 7.5/4.0, NCH) and 7.91 [1 H, d, J 30 Hz, ]] CHH(trans)]; 13C-{1H} (125 MHz), d 12.2 [C5(CH3)5], 25.4, 26.1, 32.4 (all CH2 ring), 54.3 (]] NCH), 121.1 (]] CH2), 122.6 [C5(CH3)5], 198.5 (Ti]C]] ) and 204.5 (C]] N).EI mass spectrum (111 8C): m/z 486 (M 1 H, 2), 414 (5), 351 (5), 317 [(C5Me5)2- Ti]H, 11], 305 (10), 272 (21), 170 (24), 136 (C5Me5H, 100), 121 (64), 119 (31), 105 (27), 98 (17), 93 (11), 91 (15), 88 (14), 71 (13) and 55 (11%); exact mass 485.2596 (C29H43NSTi), calculated 485.2596. IR (KBr): 3640s, 2978w, 2925vs, 2847s, 1541s [n(C]] N)], 1494m, 1449s, 1381vs, 1344m, 1261w, 1247w, 1192w, 1166m, 1132w, 1099w, 1065w, 1020w, 1002vs, 971m, 929m, 891m, 799s, 717s, 697s, 626w, 597w, 557s, 478m and 428m cm21.Ab initio and modelling calculations The geometry of MeNCS was fully optimized at the restricted Hartree–Fock level of theory with the 6-31G* standard basis set,28a,b those of complexes 17b and 18b with basis set STO- 3G28 and titanium complexes 15, 16 and 18a with an effective core potential (ECP) basis set.29a,b The ECP replaces the innermost core orbitals for titanium and all core orbitals for the main-group elements (C, N, S, Cl).For titanium, orbitals 3s, 3p, 3d, 4s and 4p were treated explicitly by a double-z quality sp and a quadruple-z quality d basis set. For the main-group elements, ns and np were treated explicitly by a double-z basis set. It has been shown that this ECP basis set is suitable for transitionmetal compounds of various kinds.30a–c Although geometries are predicted accurately at the RHF level, energetics are expected to be poor if correlation energy is ignored.The correlation contribution was taken into consideration with single-point energy calculations (at the geometries obtained at the RHF level) according to Møller–Plesset secondorder perturbation theory (MP2/RHF).31 The atomic charges have been calculated from the fully optimized structures by Mulliken population analysis. The calculations have been carried out using the program packages SPARTAN 3.1 32 on an IBM RS6000-355 and GAMESS33 on a CONVEX-C3420 computer.Crystallography Geometry and intensity data were collected on an Enraf- Nonius CAD4 diffractometer equipped with a graphite monochromator. A summary of crystallographic data, data collection and refinement parameters is given in Table 4. The monoclinic setting with a unique c axis was chosen for easier comparison between the structure of complex 5a and the related but not isotypic structure of 4b (R = C6H11) 7a which has similar unit-cell dimensions [a = 9.904(3), b = 13.780(3), c = 19.342(4) Å, b = 95.07(2)8, U = 2629(2) Å3].In both structures pairs of molecules are arranged in layers; they differ, however, with respect to the stacking of these layers along the c Table 4 Crystal data and parameters of structure refinement for complex 5a Formula M Space group (no.) Crystal symmetry a/Å b/Å c/Å g/8 U/Å3 Z Dc/g cm23 m/cm21 T/8C Radiation (l/Å) Crystal dimensions/mm Measured reflections Scan range/8 Unique observed reflections [I > 1.0s(I)] Parameters refined RR 9 * Goodness of fit C29H43NSTi 485.64 P21/n (14) Monoclinic 10.559(1) 13.252(1) 18.903(3) 90.60(1) 2644.9(9) 4 1.219 35.78 20 Cu-Ka (1.5418) 0.7 × 0.5 × 0.4 4931 5 < q < 65 3247 289 0.054 0.064 1.500 * Weighting scheme w21 = s2(Fo).J.Chem. Soc., Dalton Trans., 1997, Pages 2249–2255 2255 direction. A detailed comparison of these structures on the molecular and packing level will be published in a forthcoming paper.7b Although the crystal was sealed in a glass capillary, the intensities of three regularly measured check reflections indicated considerable anisotropic crystal decay (ca. 50%) which was taken into account by scaling the intensity data to the closest standard. Owing to this crystal instability the empirical absorption correction had to be performed by the DIFABS program34 (minimum correction 0.822, maximum 1.165) after completion of the isotropic structure model.The structure was solved by direct methods35 and refined on structure factors with the local version of the SDP program suite.36 In the full-matrix least-squares refinement (based on F), all non-hydrogen atoms were assigned anisotropic displacement factors and hydrogen atoms were included riding on the corresponding carbon atoms [C]H 0.98 Å, Uiso(H) = 1.3Ueq(C)]. The highest fluctuations in a final Fourier-difference map amounted to 0.25 e Å23. Atomic coordinates, thermal parameters, bond lengths and angles have been deposited at the Cambridge Crystallographic Data Centre (CCDC).See Instructions for Authors, J. Chem. Soc., Dalton Trans., 1997, Issue 1. Any request to the CCDC for this material should quote the full literature citation and the reference number 186/527. Acknowledgements This work was supported by the Deutsche Forschungsgemeinschaft and the Fonds der Chemischen Industrie. We thank Professor Dr. G. E. Herberich for his continuing interest.References 1 D. W. Meek, P. E. Nicpon and V. I. Meek, J. Am. Chem. Soc., 1970, 92, 5351; J. E. Huheey and S. O. Grim, Inorg. Nucl. Chem. Lett., 1974, 10, 973; G. J. Palenik, W. L. Steffen, M. Mathew, M. Li and D. W. Meek, Inorg. Nucl. Chem. Lett., 1974, 10, 125; G. R. Clark and G. J. Palenik, Inorg. Chem., 1970, 9, 2754; J. L. Burmester, Coord. Chem. Rev., 1968, 3, 225; 1966, 1, 205; J. L. Burmester and F. Basolo, Inorg. Chem., 1964, 3, 1587; A. Wojcicki and M.F. Farona, Inorg. Chem., 1964, 3, 151; F. Basolo, J. L. Burmester and A. J. Poe, J. Am. Chem. Soc., 1963, 85, 1700. 2 (a) J. L. Burmester, E. A. Deardorff, A. Jensen and V. H. Christiansen, Inorg. Chem., 1970, 9, 58; (b) I. Jibril, S. T. Abu-Orabi, S. A. Klaib, L. Zsolnai and G. Huttner, J. Organomet. Chem., 1994, 467, 189; (c) A. Chiesi-Villa, A. G. Manfredotti and C. Guastini, Acta Crystallogr., Sect. B, 1976, 32, 909; (d) R. S. P. Coutts and P. C. Wailes, Aust. J. Chem., 1966, 19, 2069. 3 (a) H. Werner, Coord. Chem. Rev., 1982, 43, 165; (b) R. Flügel, O. Gevert and H. Werner, Chem. Ber., 1996, 129, 405; (c) A. Antinolo, S. Garcia-Lledo, J. M. de Iiarduya and A. Otero, J. Organomet. Chem., 1987, 335, 85. 4 (a) V. C. Gibson, C. Redshaw, W. Clegg and M. R. J. Elsegood, J. Chem. Soc., Chem. Commun., 1994, 2635; (b) H. Brunner, M. M. Kubicki, J.-C. Leblanc, C. Moise, F. Volpato and J. Wachter, J. Chem. Soc., Chem. Commun., 1993, 851; (c) H. Werner and B. Strecker, J.Organomet. Chem., 1991, 413, 379. 5 (a) P. Jutzi, D. Eikenberg, A. Möhrke, B. Neumann and H.-G. Stammler, Organometallics, 1996, 15, 753; (b) H. Suzuki, N. Tokitoh, S. Nagase and R. Okazaki, J. Am. Chem. Soc., 1994, 116, 11 578; (c) G. Erker, M. Mena, C. Krüger and R. Noe, J. Organomet. Chem., 1991, 402, 67. 6 (a) R. Beckhaus, Angew. Chem., 1997, 109, 694; Angew. Chem., Int. Ed. Engl., 1997, 36, 686; (b) R. Beckhaus, J. Oster, J. Sang, I. Strauß and M. Wagner, Synlett, 1997, 241; (c) R.Beckhaus, S. Flatau, S. I. Troyanov and P. Hofmann, Chem. Ber., 1992, 125, 291; (d) R. Beckhaus, in Organic Synthesis via Organometallics (OSM 4), Proceedings of the Fourth Symposium in Aachen, 15–18 July 1992, eds. D. Enders, H.-J. Gais and W. Keim, Vieweg, Braunschweig, 1993, pp. 131–149. 7 (a) R. Beckhaus, I. Strauß, T. Wagner and P. Kiprof, Angew. Chem., 1993, 105, 281; Angew. Chem., Int. Ed. Engl., 1993, 32, 264; (b) T. Wagner, U. Englert, I. Strauß, J.Sang and R. Beckhaus, unpublished work. 8 R. Beckhaus, J. Oster and T. Wagner, Chem. Ber., 1994, 127, 1003. 9 R. Beckhaus, I. Strauß and T. Wagner, Angew. Chem., 1995, 107, 738; Amgew. Chem., Int. Ed. Engl., 1995, 34, 688. 10 (a) R. Beckhaus, J. Sang, T. Wagner and B. Ganter, Organometallics, 1996, 15, 1176; (b) R. Beckhaus, J. Sang, U. Englert and U. Böhme, Organometallics, 1996, 15, 4731. 11 R. Beckhaus and M. Wagner, unpublished work. 12 C. K. Johnson, ORTEP, Report ORNL-5138, Oak Ridge National Laboratory, Oak Ridge, TN, 1976. 13 (a) P. H. Bird, J. M. McCall, A. Shaver and U. Siriwardane, Angew. Chem., 1982, 94, 375; Angew. Chem., Int. Ed. Engl., 1982, 21, 384; (b) D. M. Giolando, T. B. Rauchfuss and G. M. Clark, Inorg. Chem., 1987, 26, 3080; (c) F. E. Hahn and W. W. Seidel, Angew. Chem., 1995, 107, 2938; Angew. Chem., Int. Ed. Engl., 1995, 34, 2700; (d) E. F. Epstein and I. Bernal, Chem. Commun., 1970, 410. 14 F. Bottomley, D. Drummond, G. O.Egharevba and P. S. White, Organometallics, 1986, 5, 1620. 15 H. Lettau, Chemie der Heterocyclen, Deutscher Verlag der Grundstoffindustrie, Leipzig, 1980. 16 A. Ohno, K. Nakamura, Y. Shizume and S. Oka, Bull. Chem. Soc. Jpn., 1977, 50, 1003; E. Schaumann and W. Walter, Chem. Ber., 1974, 107, 3562; P. Beak and J. W. Worley, J. Am. Chem. Soc., 1972, 94, 597. 17 I. Hyla-Kryspin, S. Niu and R. Gleiter, Organometallics, 1995, 14, 964; A. K. Rappe and W. A. Goddard III, J. Am. Chem.Soc., 1982, 104, 297; T. H. Upton and A. K. Rappe, J. Am. Chem. Soc., 1985, 107, 1206; A. K. Rappe, Organometallics, 1987, 6, 354. 18 (a) R. O. Harris, J. Powell, A. Walker and P. V. Yaneff, J. Organomet. Chem., 1977, 147, 217; (b) H. Werner, S. Lotz and B. Heiser, J. Organomet. Chem., 1981, 209, 197. 19 K. Seitz and U. Behrens, J. Organomet. Chem., 1985, 288, C47. 20 (a) J. W. Park, L. M. Henling, W. P. Schaefer and R. H. Grubbs, Organometallics, 1990, 9, 1650; (b) H. G. Alt, K.-H. Schwind and M. D. Rausch, J. Organomet. Chem., 1987, 321, C9. 21 F. E. Hahn and T. Lügger, J. Organomet. Chem., 1995, 501, 341. 22 M. Bochmann, L. M. Wilson, M. B. Hursthouse and R. L. Short, Organometallics, 1987, 6, 2556. 23 L. B. Kool, M. D. Rausch, H. G. Alt, H. E. Engelhardt and M. Herberhold, J. Organomet. Chem., 1986, 317, C38. 24 L. D. Durfee and I. P. Rothwell, Chem. Rev., 1988, 88, 1059. 25 G. A. Luinstra and J. H. Teuben, Organometallics, 1992, 11, 1793. 26 (a) R. Beckhaus, in Synthetic Methods of Inorganic and Organometallic Chemistry, ed. W. A. Herrmann, Georg Thieme, Stuttgart, in the press; (b) R. Beckhaus, K.-H. Thiele and D. Ströhl, J. Organomet. Chem., 1989, 369, 43. 27 (a) A. Hendriksson and M. Sundbom, Theoret. Chim. Acta, 1972, 27, 213; (b) M. M. Francl, W. J. Pietro, W. J. Hehre, J. S. Binkley, D. J. DeFrees and J. A. Pople, J. Chem. Phys., 1982, 77, 3654. 28 W. J. Pietro and W. J. Hehre, J. Comput. Chem., 1983, 4, 241. 29 (a) W. J. Stevens, H. Krauss and M. Basch, J. Chem. Phys., 1984, 81, 6026; (b) W. J. Stevens, M. Krauss, H. Basch and P. G. Jasien, Can. J. Chem., 1992, 70, 612. 30 (a) T. R. Cundari, Organometallics, 1993, 12, 4971; (b) T. R. Cundari, J. Am. Chem. Soc., 1994, 116, 340; (c) T. R. Cundari, J. Am. Chem. Soc., 1992, 114, 7879. 31 C. Møller and M. S. Plesset, Phys. Rev., 1934, 46, 618. 32 SPARTAN 3.1, Wavefunction, Inc., Irvine, CA, 1993. 33 M. W. Schmidt, K. K. Baldrige, J. A. Boatz, J. H. Jensen, S. Koseki, M. S. Gordon, K. A. Nguyen, T. L. Windus and S. T. Elbert, QCPE Bull., 1990, 10, 52. 34 N. Walker and D. Stuart, Acta Crystallogr., Sect. A, 1983, 39, 158. 35 G. M. Sheldrick, SHELXS 86, Program for Structure Solutions, University of Göttingen, 1986. 36 SDP, Version 5.0, Enraf-Nonius, Delft, 1989. Received 4th March 1997; Paper 7/01499F

ISSN:1477-9226    

DOI:10.1039/a701499f

出版商:RSC    

年代:1997

数据来源: RSC