摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1998, Pages 2253–2259 2253 Titanium imido complexes with tetraaza macrocyclic ligands Daniel Swallow, Jacqueline M. McInnes and Philip Mountford*,† Department of Chemistry, University of Nottingham, Nottingham, UK NG7 2RD Tetraaza macrocycle-supported tert-butyl titanium imido complexes [Ti(NBut)(Mentaa)] (n = 4 2 or 8 3; H2Mentaa = 6,8,15,17-tetra- or 2,3,6,8,11,12,17,18-octa-methyl-5,14-dihydrodibenzo[b,i][1,4,8,11]tetraazacyclotetradecine, respectively), [Ti(NBut)(Me4taen)] (5, H2Me4taen = 5,7,12,14-tetramethyl-1,4,8,11- tetraazacyclotetradeca-4,6,11,13-tetraene) and [Ti(NBut)(TTP)] (6, H2TTP = 5,10,15,20-tetratolylporphyrin) together with the N2O2-donor SchiV base analogue [Ti(NBut)(acen)] (7, H2acen = 4,9-dimethyl-5,8-diazadodeca- 3,9-diene-2,11-dione) were prepared in good yield from the readily available [Ti(NBut)Cl2(py)3] and the dilithium or disodium salts of the tetradentate ligands.The Ti]] NBut groups in 2 and 3 underwent imido group exchange reactions with anilines to form [Ti(NR)(Mentaa)] (n = 4, R = C6H3Me2-2,6 4, Ph, C6H4(NO2)-4, C6H4(NMe2)-4; n = 4 or 8, R = C6H4Me-4), and with H2E (E = O or S) to give the oxo and sulfido analogues [Ti(E)(Me4taa)].Compound 4 was also prepared in good yield from [Ti(NC6H3Me2-2,6)Cl2(py)3] and Li2[Me4taa]. Reaction of 2 with 2 or 1 equivalents of ROH (R = Me or C6H3Me2-2,6) or pinacol aVorded [Ti(OR)2(Me4taa)] and [Ti{OC(Me)2C(Me)2O}(Me4taa)] respectively.The crystal structures of 3 and 4 have been described. Dianionic tetraaza macrocycles and N2O2-donor SchiV bases have received much attention in recent years as potential alternatives to the ubiquitous bis(cyclopentadienyl) ligand set in early transition-metal chemistry.1–18 As part of an ongoing research program in transition-metal imido chemistry 4,19–21 we were interested to explore the opportunities that such ligands oVer in this area. We were especially attracted by the dibenzotetraaza[ 14]annulene systems Mentaa (n = 4 or 8), the chemistry of which has been reviewed.2,3 These tetraaza macrocycles are related to the porphyrins but diVer in several important respects.For instance, their N4 co-ordination cavity ‘hole size’ is ca. 0.1 Å smaller than that of porphyrins and they typically possess non-planar geometries. It is relevant to note that dibenzotetraaza[14]annulene ligands have already provided supporting environments for a number of transition-metal– and main-group metal–ligand multiple bonds.4–8,22 Cognisant of the very interesting reaction chemistry that GeoVroy and co-workers 5 found for the oxo- N H H N N N Me Me Me Me R R R R N N N N H H Tol Tol Tol Tol N H H N N N Me Me Me Me N H H N O Me Me Me Me O H2acen H2Me4taen R = H: H2Me4taa; R = Me: H2Me8taa H2TTP (Tol = C6H4Me-4) † P.Mountford is the Royal Society of Chemistry Sir Edward Frankland Fellow. E-Mail: Philip.Mountford@Nottingham.ac.uk (http://www.nottingham.ac.uk/~pczwww/Inorganic/PMount.html). titanium species [Ti(O)(Mentaa)] (n = 422 or 8), we prepared the isoelectronic imido analogues [Ti(NBut)(Mentaa)] (n = 4 2 or 8 3),23 together with homologous tetraaza[14]annulene-supported zirconium imido compounds.4,24 Since our preliminary communication 23 of a part of these initial studies, a number of other Group 4 macrocycle-supported imido derivatives have been described.25–27 In this contribution we describe in full the synthesis, properties and imido group exchange reactions of titanium imido complexes with dibenzotetraaza[14]annulene ligands, together with synthetic routes to a number of other macrocycle- and SchiV base-supported analogues.A part of this work has been communicated.23,28 Experimental General methods and instrumentation Manipulations were carried out under an atmosphere of dinitrogen or argon using either standard Schlenk-line or dry-box techniques.Solvents were pre-dried over molecular sieves and refluxed over potassium (tetrahydrofuran, hexane), sodium–potassium alloy (pentane) or calcium hydride (dichloromethane) under an atmosphere of dinitrogen and collected by distillation, CDCl3 and CD2Cl2 were dried over calcium hydride at room temperature (r.t.), distilled under reduced pressure and stored under N2 in Young’s ampoules in a dry-box. The NMR samples were prepared in a dry-box in Teflon valve (Young’s) 5 mm tubes.Proton and 13C NMR spectra were recorded on either a Bruker WM 250, Bruker AMX 500 or Bruker DPX 300 spectrometer at 298 K unless stated otherwise. The spectra were referenced internally to residual protio-solvent (1H) or solvent (13C) resonances and are reported relative to tetramethylsilane (d = 0). Chemical shifts are quoted in d (ppm) and coupling constants in Hz. Assignments were supported by DEPT-135 and DEPT-90, homo- and hetero-nuclear, one- and two-dimensional, and NOE experiments as appropriate.Mass spectra were recorded on either a VG Micromass 7070E or a AEI MS902 mass spectrometer. Elemental analyses were carried out by the analysis laboratory of this department or by Canadian Microanalytical Services Ltd. The compounds Li2[(Mentaa)] (n = 4 or 8), Li2[Me4taen], Li2[TTP]?2THF, Na2[acen] and [Ti(NR)Cl2(py)3] (R = But 1a or C6H3Me2-2,6 1b) were prepared as previously described.12,13,29–312254 J. Chem. Soc., Dalton Trans., 1998, Pages 2253–2259 Syntheses [Ti(NBut)(Me4taa)] 2.A cold (0 8C) solution of Li2[Me4taa] (3.50 g, 9.83 mmol) in THF (40 ml) was added dropwise to a cold solution of [Ti(NBut)Cl2(py)3] (4.18 g, 9.83 mmol) in THF (20 ml). The mixture was allowed to warm to r.t. and stirred for a further 12 h, after which the solvent was removed under reduced pressure. Dichloromethane (50 ml) was added, giving a red solution with a white precipitate (LiCl). The solution was filtered, reduced to ca. 20 ml and hexane (20 ml) was added. Compound 2 formed as a red solid that was washed with hexane (3 × 10 ml) and dried in vacuo. Yield: 3.18 g (70%). 1H NMR (CDCl3, 250.1 MHz, 298 K): d 7.32–7.36, 7.18–7.24 (2 × m, 2 × 4 H, C6H4), 5.13 [s, 2 H, NC(Me)CH], 2.40 [s, 12 H, NC(Me)CH], 0.26 (s, 9 H, But). 13C-{1H} NMR (CDCl3, 62.5 MHz, 298 K): d 160.1 [NC(Me)CH], 139.4 (CN of C6H4 of Me4taa), 124.3, 123.3 (2 × CH of C6H4), 103.3 [NC(Me)CH], 67.2 (NCMe3), 31.6 (NCMe3), 22.7 [NC(Me)CH] [Found (Calc.for C26H31N5Ti?CH2Cl2): C, 60.2 (59.4); H, 6.0 (6.1); N, 12.6 (12.8)%]. [Ti(NBut)(Me8taa)] 3. A THF (20 ml) solution of [Ti(NBut)- Cl2(py)3] (2.45 g, 5.70 mmol) was added dropwise to Li2[Me8taa] (2.40 g, 5.8 mmol) in THF (20 ml). The mixture was stirred for 16 h at r.t., after which the solvent was removed under reduced pressure. The red product was extracted with CH2Cl2 (200 ml), filtered through a pad of Celite to remove the LiCl and the solvent removed under reduced pressure.The resulting red-brown solid was recrystallised by solvent diVusion using CH2Cl2 and pentane at 278 8C to aVord the microcrystalline complex 3, which was dried in vacuo. Yield: 2.50 g (70%). DiVraction quality crystals of 3 formed overnight at 230 8C from a CH2Cl2 solution layered with hexane. 1H NMR (CDCl3, 300.1 MHz): d 7.13 (s, 4 H, C6H2Me2), 5.07 [s, 2 H, NC(Me)- CH], 2.39 [s, 12 H, NC(Me)CH], 2.32 (s, 12 H, C6H2Me2), 0.29 (s, 9 H, But). 13C-{1H} NMR (CDCl3, 125.7 MHz): d 159.7 [NC(Me)CH], 137.1 (CN of C6H2Me2 of Me8taa), 132.5 (CMe of C6H2Me2), 124.2 (CH of C6H2Me2), 102.9 [NC(Me)CH], 67.0 (NCMe3), 31.8 (NCMe3), 22.7 [NC(Me)CH], 20.0 (C6H2Me2) [Found (Calc.for C30H39N5Ti): C, 69.6 (69.6); H, 7.6 (7.6); N, 13.4 (13.5)%]. [Ti(NC6H3Me2-2,6)(Me4taa)] 4. Method (a): from 1b and Li2[Me4taa]. To a cold (0 8C) solution of 1b (0.200 g, 0.414 mmol) in CH2Cl2 (20 ml) was added cold Li2[Me4taa] (0.140 g, 0.414 mmol) in CH2Cl2 (20 ml).The mixture was allowed to warm to r.t. and then stirred for 24 h. Filtration and removal of the volatiles under reduced pressure gave spectroscopically pure 4 as a brown solid. Yield: 0.150 g (71%). Method (b): from 2 and 2,6-dimethylaniline. To a stirred solution of 2 (0.200 g, 0.433 mmol) in CH2Cl2 (15 ml) was added 2,6-dimethylaniline (0.190 ml, 1.52 mmol, 3.5 equivalents). The solution changed from light red to dark red over 3 d at r.t. after which the volatiles were removed under reduced pressure.The residue was crystallised from hexane–CH2Cl2 (6 : 1) at 225 8C overnight to give 4 as a brown solid that was washed with hexane–CH2Cl2 and dried in vacuo. Yield: 0.170 g (77%). DiVraction quality crystals of 4?H2NC6H3Me2-2,6 were grown at room temperature over several days from a CH2Cl2 solution of crude 4 layered with hexane. 1H NMR (CDCl3, 250.1 MHz): d 7.47–7.43, 7.31–7.25 (2 × m, 2 × 4 H, C6H4), 6.39 (d, J 7.5, 2 H, m-C6H3Me2), 6.16 (t, J 7.2, 1 H, p-C6H3Me2), 5.32 [s, 2 H, NC(Me)CH], 2.48 [s, 12 H, NC(Me)CH], 1.10 (s, 6 H, C6H3Me2). 13C-{1H} NMR (CDCl3, 62.5 MHz): d 161.1 [NC(Me)CH], 139.7 (CN of C6H4 of Me4taa), 129.5 (o-C6H3- Me2), 126.3 (m-C6H3Me2), 125.4, 124.3 (2 × CH of C6H4), 117.9 (p-C6H3Me2), 105.0 [NC(Me)CH], 23.5 [NC(Me)CH], 18.1 (C6H3Me2); note: the ipso carbon of the C6H3Me2 group was not observed. EI mass spectrum: m/z = 509 {M1}, 406 {M1 2 C8H7}, 390 {M1 2 C8H9N} [Found (Calc. for C30H31N5Ti): C, 69.2 (70.7); H, 6.1 (6.1); N, 13.0 (13.7)%].[Ti(NBut)(Me4taen)] 5. To a cold (0 8C) solution of [Ti- (NBut)Cl2(py)3] (0.740 g, 1.73 mmol) in THF (30 ml) was added cold Li2[Me4taen] (0.450 g, 1.73 mmol) in THF (30 ml) dropwise to give a colour change from orange to red. After 24 h the volatiles were removed under reduced pressure and the residues were extracted into CH2Cl2 (40 ml) and filtered. The volume was concentrated to 20 ml and hexane (20 ml) added. Compound 5 formed as a red solid, which was filtered, washed with hexane (2 × 10 ml), and dried in vacuo.Yield: 0.390 g (61%). 1H NMR (CDCl3, 250.1 MHz): d 4.70 [s, 2 H, NC(Me)CH], 4.21, 3.65 (2 × m, 2 × 2 H, NCH2), 1.98 [s, 12 H, NC(Me)CH], 0.89 (s, 9 H, But). 13C-{1H} NMR (CDCl3, 62.5 MHz): d 163.0 [NC(Me)CH], 98.1 [NC(Me)CH], 67.1 (NCMe3), 50.7 (NCH2), 32.8 [NC(Me)CH], 21.2 (NCMe3). EI mass spectrum: m/z = 365 {M1}, 350 {M1 2 CH3} [Found (Calc. for C18H31N5Ti): C, 59.7 (59.2); H, 8.3 (8.5); N, 16.1 (19.2)%]; repeated analyses did not lead to improved %N found which may be low due to titanium nitride formation during combustion.[Ti(NBut)(TTP)] 6. To a cold (0 8C) solution of [Ti(NBut)- Cl2(py)3] (0.100 g, 0.234 mmol) in THF (30 ml) was added Li2[TTP]?2THF (0.200 g, 0.234 mmol) in THF (30 ml) dropwise. The mixture was allowed to warm to r.t. and then stirred for 24 h to give a red-purple solution. The volatiles were removed under reduced pressure, and the purple residue was extracted with toluene–hexane (1 : 1, 30 ml) and concentrated to give 6 as a spectroscopically pure, purple solid after cooling to 225 8C overnight.Yield: 0.130 g (71%). The compound was characterised by comparison with previously described data.27 [Ti(NBut)(acen)] 7. To a cold (0 8C) solution of [Ti(NBut)- Cl2(py)3] (0.790 g, 1.8 mmol) in THF (20 ml) was added a slurry of Na2[acen] (0.500 g, 1.8 mmol) in THF (25 ml). On warming to r.t. the mixture turned a deep brown-red and was stirred at r.t.for a further 24 h. Volatiles were removed under reduced pressure and the dark brown-red residue was extracted into CH2Cl2 (40 ml) and quickly filtered. The volume was reduced and 7 formed as a brown solid on the addition of hexane (20 ml). Yield: 0.310 g (52%). 1H NMR (CDCl3, 300.1 MHz): d 5.19 [s, 2 H, MeC(O)CH(N)Me], 4.17, 3.67 (2 × m, 2 × 2 H, NCH2), 2.10, 2.08 [2 × s, 2 × 6 H, MeC(O)CH(N)Me and MeC(O)CH(N)Me], 0.85 (s, 9 H, But). 13C-{1H} NMR (CDCl3, 75.5 MHz): d 178.1 [MeC(O)CH(N)CMe], 168.6 [MeC(O)- CH(N)CMe], 101.9 [MeC(O)CH(N)CMe], 69.2 (NCMe3), 51.7 (NCH2), 31.8 (NCMe3), 24.9 [MeC(O)CH(N)CMe], 21.5 [MeC(O)CH(N)CMe].Elemental analyses were not obtained for this compound which decomposed in solvents from which recrystallisation was attempted. [Ti(NPh)(Me4taa)] 8, [Ti(NC6H4Me-4)(Me4taa)] 9, [Ti{NC6- H4(NO2)-4}(Me4taa)] 10, [Ti{NC6H4(NMe2)-4}(Me4taa)] 11 and [Ti(NC6H4Me-4)(Me8taa)] 12. These compounds were prepared according to method (b) above for 4 from complex 2 (except for the synthesis of 12 for which benzene was used as solvent) and the corresponding aniline with reaction times of 3–24 h.Yields after crystallisation from CH2Cl2–hexane at r.t. or 225 8C: for 8, 85; for 9, 81; for 10, 80; for 11, 75; for 12, 95%. Data for 8. 1H NMR (CDCl3, 250.1 MHz): d 7.50–7.46, 7.34– 7.31 (2 × m, 2 × 4 H, C4H4), 6.55 (apparent t, apparent J 7.9, 2 H, m-Ph), 6.23 (t, J 7.2, 1 H, p-Ph), 5.33 (d, J 8.1, 2 H, o-Ph), 5.29 [s, 2 H, NC(Me)CH], 2.47 [s, 12 H, NC(Me)CH]. 13C-{1H} NMR (CDCl3, 62.5 MHz): d 160.5 [NC(Me)CH], 138.2 (CN of C6H4 of Me4taa), 127.2 (m-Ph), 125.2, 123.7 (2 × CH of C6H4), 120.7 (o-Ph), 117.2 (p-Ph), 104.5 [NC(Me)CH], 22.7 [NC(Me)CH]; note: the ipso carbon of the Ph group was not observed [Found (Calc. for C28H27N5Ti): C, 68.5 (69.8); H, 5.7 (5.6); N, 14.4 (14.6)%]. Data for 9. 1H NMR (CDCl3, 250.1 MHz): d 7.49–7.45, 7.33– 7.29 (2 × m, 2 × 4 H, C6H4 of Me4taa), 6.36 (d, J 8.0, 2 H, m-C6H4Me), 5.28 [s, 2 H, 2 × NC(Me)CH], 5.27 (d, J 8.0, 2 H,J.Chem. Soc., Dalton Trans., 1998, Pages 2253–2259 2255 o-C6H4Me), 2.46 [s, 12 H, NC(Me)CH], 1.94 (s, 3 H, C6H4Me). 13C-{1H} NMR (CDCl3, 62.5 MHz): d 160.4 [NC(Me)CH], 158.9 (p-C6H4Me), 138.3 (CN of C6H4 of Me4taa), 127.7 (m-C6H4Me), 125.1, 123.7 (2 × CH of C6H4 of Me4taa), 120.4 (o-C6H4Me), 104.4 [NC(Me)CH], 22.7 [NC(Me)CH], 20.5 (C6H4Me); note: the ipso carbon of the C6H4Me group was not observed [Found (Calc.for C29H29N5Ti): C, 69.9 (70.3); H, 5.8 (5.9); N, 14.0 (14.1)%]. Data for 10. 1H NMR (CDCl3, 250.1 MHz): d 7.52–7.38 (overlapping 2 × m and d, 10 H, 2 × C6H4 of Me4taa and m-C6H4NO2), 5.43 [s, 2 H, NC(Me)CH], 5.07 (d, J 8.1, 2 H, o-C6H4NO2), 2.51 [s, 12 H, NC(Me)CH]. 13C-{1H} NMR (CDCl3, 62.5 MHz): d 165.2 (p-C6H4NO2), 160.8 [NC(Me)CH], 137.2 (CN of C6H4 of Me4taa), 126.1 (CH of C6H4 of Me4taa), 124.7 (m-C6H4NO2), 123.8 (CH of C6H4 of Me4taa), 120.3 (o-C6H4NO2), 113.2 (ipso-C6H4NO2), 105.5 [NC(Me)CH], 22.7 [4 × NC(Me)CH] [Found (Calc.for C28H26N6O2Ti): C, 64.0 (63.9); H, 5.3 (5.0); N, 16.0 (16.0)%]. Data for 11. 1H NMR (CDCl3, 250.1 MHz): d 7.50–7.46, 7.33–7.29 (2 × m, 2 × 4 H, C6H4 of Me4taa), 6.12 (d, J 8.7, 2 H, m-C6H4NMe2), 5.36 (d, J 8.7, 2 H, o-C6H4NMe2), 5.26 [s, 2 H, NC(Me)CH], 2.58 (s, 6 H, C6H4NMe2), 2.46 [s, 12 H, NC(Me)CH]. 13C-{1H} NMR (CDCl3, 62.5 MHz): d 160.4 [NC(Me)CH], 155.1 (p-C6H4NMe2), 143.3 (ipso-C6H4NMe2), 138.5 (CN of C6H4 of Me4taa), 125.0, 123.7 (2 × CH of C6H4 of Me4taa), 121.2 (m-C6H4NMe2), 113.8 (o-C6H4NMe2), 104.3 [NC(Me)CH], 42.1 (C6H4NMe2), 22.7 [NC(Me)CH].Satisfactory elemental analyses could not be obtained for this compound. Data for 12. 1H NMR (CDCl3, 250.1 MHz): d 7.37 (s, C6H6), 7.25 (s, 4 H, C6H2Me2 of Me8taa), 6.33 (d, J 8.0, 2 H, m-C6H4Me), 5.33 (d, J 8.0, 2 H, o-C6H4Me), 5.32 [s, 2 H, 2 × NC(Me)CH], 2.46 [s, 12 H, NC(Me)CH], 2.37 (s, 12 H, C6H2Me2), 1.98 (s, 3 H, C6H4Me). 13C-{1H} NMR (CDCl3, 62.5 MHz): d 160.3 (ipso-C6H4Me), 159.9 [NC(Me)CH], 136.0 (CN of C6H2Me2 of Me8taa), 134.2 (p-C6H4Me), 133.6 (CMe of C6H2Me2 of Me8taa), 127.7 (C6H6), 125.1 and 120.5 (o- and m-C6H4Me), 124.6 (4 × CH of Me8taa), 103.9 [2 × NC(Me)- CH], 22.8 and 20.1 [NC(Me)CH and C6H2Me2, respectively], 20.6 (C6H2Me2) [Found (Calc. for C33H37N5Ti): C, 70.8 (71.9); H, 6.7 (6.8); N, 11.8 (12.7)%]. [Ti(O)(Me4taa)] 13. To a red solution of [Ti(NBut)(Me4taa)] (0.100 g, 0.217 mmol) in THF (20 ml) at r.t.was added H2O (3.90 ml, 0.217 mmol). The colour changed instantly to yellow and a yellow solid precipitated which was filtered oV after 30 min, washed with THF (2 × 5 ml) and dried to give spectroscopically pure 13. Yield: 0.073 g (82%). The compound was characterised by comparison with previously described data.22 [Ti(S)(Me4taa)] 14. Dihydrogen sulfide was slowly passed through a stirred solution of [Ti(NBut)(Me4taa)] (0.100 g, 0.217 mmol) in CH2Cl2 (30 ml) for 1 min at r.t.to give 14 as a spectroscopically pure, orange solid. Yield: 0.080 g (87%). The compound was characterised by comparison with previously described data.22 [Ti(OMe)2(Me4taa)] 15. To a solution of [Ti(NBut)(Me4taa)] (0.200 g, 0.433 mmol) in CH2Cl2 (15 ml) was added methanol (0.10 ml, 2.5 mmol, 6 equivalents). After stirring for 24 h at r.t. the volume was reduced and hexane added. Cooling to 225 8C overnight gave 15 as a red solid which was washed with hexane (2 × 10 ml) and dried in vacuo.Yield: 0.12 g (62%). 1H NMR (CDCl3, 250.1 MHz): d 7.30–7.10 (overlapping 2 × m, 2 × 4 H, C6H4), 5.27 [s, 2 H, NC(Me)CH], 3.69 (s, 6 H, OMe), 2.39 [s, 12 H, NC(Me)CH]. 13C-{1H} DEPT-135 NMR (CDCl3, 62.5 MHz): d 124.9, 122.7 (CH of C6H4), 103.1 [NC(Me)CH], 62.6 (OMe), 23.6 [NC(Me)CH] [Found (Calc. for C24H28N4O2Ti): C, 63.2 (63.7); H, 6.3 (6.2); N, 12.3 (12.4)%]. [Ti(OC6H3Me2-2,6)2(Me4taa)] 16. To a solution of [Ti- (NBut)(Me4taa)] (0.200 g, 0.433 mmol) in CH2Cl2 (15 ml) was added 2,6-dimethylphenol (0.110 g, 0.866 mmol) in CH2Cl2 (15 ml).After 12 h at r.t. 16 formed as a red-brown microcrystalline CH2Cl2 solvate (by 1H NMR and elemental analysis) and was washed with a minimum volume of CH2Cl2–hexane (1 : 6) and dried in vacuo. Yield: 0.220 g (71%). 1H NMR (CDCl3, 250.1 MHz): d 7.22–7.11 (overlapping 2 × m, 2 × 4 H, C6H4), 6.48 (d, J 7.3, 4 H, m-C6H3Me2), 6.26 (t, J 7.3, 2 H, p-C6H3Me2), 5.50 [s, 2 H, NC(Me)CH], 2.43 [s, 12 H, C(Me)CH], 1.27 (s, 12 H, C6H3Me2). 13C-{1H} NMR (CDCl3, 62.5 MHz): d 165.7 (ipso- C6H3Me2), 158.5 [C(Me)CH], 140.4 (CN of C6H4 of Me4taa), 127.2 (m-C6H3Me2), 125.5 (o-C6H3Me2), 124.9, 123.2 (2 × CH of C6H4), 117.5 (p-C6H3Me2), 105.4 [NC(Me)CH], 24.4 [NC(Me)CH], 15.9 (C6H3Me2) [Found (Calc. for C38H40N4O2- Ti?CH2Cl2): C, 64.9 (65.3); H, 5.9 (5.9); N, 7.7 (7.8)%]. [Ti{OC(Me)2C(Me)2O}(Me4taa)] 17. To a solution of [Ti- (NBut)(Me4taa)] (0.100 g, 0.217 mmol) in CH2Cl2 (15 ml) was added pinacol (0.026 g, 0.217 mmol) in CH2Cl2 (15 ml).The solution immediately changed from red to orange and was stirred for 24 h at r.t. The volume was reduced, hexane (15 ml) added and the solution cooled to 225 8C overnight to give microcrystalline 17 as an orange CH2Cl2 hemi-solvate (by 1H NMR and elemental analysis) that was washed with hexane (2 × 5 ml) and dried in vacuo. Yield: 0.10 g (81%). 1H NMR (CDCl3, 250.1 MHz): d 7.35–7.13 (overlapping 2 × m, 2 × 4 H, C6H4), 5.32 [s, 2 H, NC(Me)CH], 2.39 [s, 12 H, NC(Me)CH], 0.37 (s, 12 H, O2C2Me4). 13C-{1H} NMR (CDCl3, 62.5 MHz): d 158.3 [NC(Me)CH], 141.5 (CN of C6H4 of Me4taa), 123.9, 123.3 (2 × CH of C6H4), 104.5 [NC(Me)CH], 92.1 (O2C2Me4), 25.8 (O2C2Me4), 23.6 [NC(Me)CH] [Found (Calc.for C28H34- N4O2Ti?0.5CH2Cl2): C, 62.7 (62.4); H, 6.6 (6.4); N, 10.4 (10.2)%]. Results and Discussion The ligand precursors used in this study are readily deprotonated with BunLi (for H2Mentaa and H2Me4taen),12,29 LiN- (SiMe3)2 (for H2TTP),30 or sodium hydride (H2acen) 13 to form the corresponding dilithium or disodium salts.In recent studies 19 we have found a number of diVerent classes of titanium imido compound can be prepared via chloride and/or pyridine substitution reactions of the readily-available synthons [Ti(NR)Cl2(py)3] (R = But 1a or C6H3Me2-2,6 1b).31 The reactions of 1 with dianionic N4- and N2O2-donor ligands are summarised in Scheme 1 and details of the preparation and characterisation of all the compounds are given in the Experimental section.The reaction of [Ti(NBut)Cl2(py)3] 1a with Li2[Mentaa] (n = 4 or 8) in THF procedes smoothly to aVord good yields of the red tetraaza macrocyclic derivatives [Ti(NBut)(Mentaa)] (n = 4 2 or 8 3). These compounds are also accessible from the bis(tertbutylpyridine) homologues [Ti(NBut)Cl2(Butpy)2] (Butpy = tertbutylpyridine), 23 and are air- and moisture-sensitive in solution and the solid state.The arylimido analogue [Ti(NC6- H3Me2-2,6)(Me4taa)] 4 was also prepared cleanly in an analogous fashion from [Ti(NC6H3Me2-2,6)Cl2(py)3] 1b and Li2[Me4taa] in dichloromethane. However, use of dichloromethane as reaction solvent for 2 and 3 gave lower yields of isolated product. The compounds 2–4 are the first macrocyclesupported imido complexes of Group 4. The NMR spectra for 2–4 are consistent with the proposed structures shown in Scheme 1. Interestingly, the 1H chemical shifts for the NBut group in 2 and 3 (d 0.26 and 0.29 respectively) and the ortho-methyl substituents for NC6H3Me2-2,6 in 4 (d 1.10) in CDCl3 occur at significantly higher field compared with the corresponding resonances 31 for 1a (d 0.92) and 1b (d 2.40) respectively in the same solvent.We attribute this to2256 J. Chem. Soc., Dalton Trans., 1998, Pages 2253–2259 shielding eVects of the o-phenylene aromatic rings of the Mentaa ligand (see below). The solid state structures 23,28 of 3 and 4 are shown in Figs. 1 and 2, and important molecular dimensions are summarised in Table 1. The structures contain approximately square-base pyramidal titanium centres with near-linear organoimido groups in the axial co-ordination sites. The macrocycle nitrogen donor atoms form the remainder of the co-ordination sphere and are eVectively coplanar (the maximum deviation from the least squares macrocyclic N4 plane is ca. 0.1 Å in both 3 and 4), and the Ti lies 0.76 (for 3) and 0.75 Å (for 4) above the N4 plane.The Ti]] ] Nimide bond lengths of ca. 1.72 Å are at the longer end of the Scheme 1 Synthesis of titanium imido complexes supported by tetraazamacrocyclic and acen ligands. (i) Li2[Mentaa] (n = 4 or 8), THF (for 2 and 3) or CH2Cl2, 0 8C then r.t., 12–24 h, ca. 70%; (ii) Li2[Me4taen], THF, 0 8C then r.t., 24 h, 61%; (iii) Li2[TTP]?2THF, THF, 0 8C then r.t., 24 h, 71%; (iv) Na2[acen], THF, 0 8C then r.t., 24 h, 52% N N R¢ R¢ Ti N N R¢ R¢ N R N N Ti N N N But N N Ti O O N But N N Ti N N N But Tol Tol Tol 5 6 7 R = But, R¢ = H 2 or Me 3 R = C6H3Me2-2,6, R¢ = H 4 Ti py Cl py Cl py N R R = But 1a or C6H3Me2-2,6 1b (ii) (iv) (i) (iii) Table 1 Selected distances (Å) and angles (8) for [Ti(NBut)(Me8taa)] 3 and [Ti(NC6H3Me2-2,6)(Me4taa)] 4 23,28 Ti]Nimide Ti ? ? ?N4 plane Ti]Nmacrocycle* Ti]Nimide]C Nimide]Ti]Nmacrocycle* 3 1.724(4) 0.76 2.070(4), 2.093(4), 2.089(4), 2.091(4) 164.3(3) 107.0(2), 114.3(2), 108.7(2), 115.1(2) 4 1.720(4) 0.75 2.064(4), 2.060(4), 2.084(4), 2.078(4) 175.4(4) 109.7(2), 109.9(2), 113.0(2), 112.4(2) * Values correspond to the atoms N(1), N(2), N(3), N(4) in that order in each instance.range for this linkage (range ca. 1.66–1.74 Å),19,32 but still imply a formal Ti]] ] Nimide triple bond.‡ The Mentaa ligands in 3 and 4 adopt the characteristic ‘saddle shape’,2,3 and the folding of the o-phenylene rings ‘up’ towards the imido groups is consistent with the shielding eVects seen in the 1H NMR spectra. The square-base pyramidal co-ordination geometry in 3 and 4 is well-established in titanium imido chemistry.19,32 However, the macrocyclic compounds described here have unusually large average N]] ] Ti]Nmacrocycle bond angles (average 111.38 for both 3 and 4) when compared with previous examples: [Ti(NBut)- Cl2(tmeda)] (tmeda = N,N,N9,N9-tetramethylethylenediamine; average N]] ] Ti]L = 103.58),33 [Ti2(NBut)2{m-O2P(OSiMe3)2}4] (average N]] ] Ti]L = 107.18),34 [Ti(NBut)Cl2(dipeda)] [dipeda = N,N9-diisopropylethylenediamine; average N]] ] Ti]L = 101.7(2)8],33 [ Ti(NBut)Cl2(OPPh3)2] [ average N]] ] Ti]L = 105.4(3)8],35 and [Ti(NPh)(TTP)] (average N]] ] Ti]Nmacrocycle = 104.38).36 This is apparently related to the small N4 ‘hole size’ of Mentaa (see Introduction section) which leads to a relatively large displacement of Ti from the N4 plane and so to larger average N]] ] Ti]Nmacrocycle bond angles.Possible electronic consequences of these structural constraints for [Ti(E)(Mentaa)] (E = NR or O) have been discussed by us elsewhere.28 We were interested to prepare other tetraaza macrocycle Fig. 1 Molecular structure of [Ti(NBut)(Me8taa)] 3 with hydrogen atoms omitted 23 Fig. 2 Molecular structure of [Ti(NC6H3Me2-2,6)(Me4taa)] 4 with hydrogen atoms omitted 28 ‡ Although for ease of representation all titanium-imido linkages in Schemes 1 and 2 are drawn ‘Ti]] NR’, the formal Ti]N bond order in the complexes [Ti(NR)(L)n] (R = But or aryl, n = 1 or 2; L = dianionic tetradentate ligand) described herein is generally best thought of as three (pseudo-s2 p4 triple bond) rather than as two.32J.Chem. Soc., Dalton Trans., 1998, Pages 2253–2259 2257 Scheme 2 Imido ligand exchange and protonolysis reactions of [Ti(NBut)(Mentaa)] (n = 4 2 or 8 3). (i) Slow stream of H2S, CH2Cl2, r.t., 1 min, 87%; (ii) H2NC6H2{(Ro)2(Rp)}-2,4,6 (ca. 1–4 equivalents), CH2Cl2, 18 h–3 d, 75–85%; (iii) 4-methylaniline (1.1 equivalents), C6H6, r.t., 3 h, 95%; (iv) H2O, THF, r.t., 30 min, 82%; (v) ROH (2 equivalents), CH2Cl2, r.t., 12–24 h, 62% (for 15) or 71% (for 16); (vi) pinacol, CH2Cl2, r.t., 24 h, 81% N N Ti N N N N N Ti N N N 12 But N N Ti N N O 13 N N Ti N N R = Me 15 or C6H3Me2-2,6 16 OR RO N N Ti N N S 14 N N Ti N N N Rp Ro = Me, R p = H 4 Ro = H, R p = H 8, Me 9, NO2 10, NMe2 11 Ro Ro N N Ti N N 17 O O R = H 2 or Me 3 (i) (iv) (iii) (ii) (vi) (v) Me Me Me Me R R R R titanium imido complexes (Scheme 1). Group 4 derivatives of Me4taen have recently been reported by Jordan and co-workers but no imido derivatives were described.11,12 We found that reaction of Li2[Me4taen] with [Ti(NBut)Cl2(py)3] 1a gave [Ti- (NBut)(Me4taen)] 5 in 61% yield after standard work-up.The compound 5 is analogous to 2 and 3, except that the o-phenylene rings have formally been replaced by ethylene linkages. The hydrogen atoms of these linkages appear as two mutually-coupled multiplets consistent with the proposed structure.The tert-butyl 1H NMR resonance for 5 appears at d = 0.89, somewhat upfield from the equivalent resonances for 2 and 3; this lends support to our view that the upfield shifts of the imido N-substituents in the dibenzotetraaza[14]annulene derivatives can be attributed to eVects of the o-phenylene rings. While this work was in progress, Woo and co-workers 27 described the synthesis of the porphyrin titanium imido complex [Ti(NBut)(TTP)] 6 from [Ti(TTP)Cl2] and LiNHBut.We have independently found that the same compound is accessible by treatment of [Ti(NBut)Cl2(py)3] 1a with Li2[TTP]?2THF in 71% recrystallised yield from toluene–hexane. This compares with a crude yield (recrystallised yield not reported) of 94% using the previously published method. There are only a few N2O2 SchiV base-supported imido compounds known.37–41 For comparison with the tetraaza macrocyclic systems we therefore prepared the complex [Ti(NBut)- (acen)] 7 from 1a and Na2[acen].This compound was obtained as a spectroscopically pure, brown solid in 52% yield. Attempts to obtain analytically pure samples were unsuccessful. However, the 1H and 13C NMR spectroscopic data are fully consistent with the structure proposed in Scheme 1. Thus the ethylene protons appear as a pair of mutually coupled multiplets, and the tert-butyl resonance occurs in the expected region (d 0.85) for a terminal Ti]] ] NBut linkage.Scheme 2 shows exchange and protonolysis reactions of the tert-butylimido ligand in [Ti(NBut)(Mentaa)] (2 and 3). Imide/ amine exchange reactions of tert-butylimido compounds with certain anilines appears now to be a widely applicable route to the corresponding arylimido homologues.31,42–48 We were interested to use this method for preparing dibenzotetraaza[14]- annulene-supported o-unsubstituted arylimido compounds since starting materials of the type [Ti(NC6H4R-4)Cl2(py)3] (R = H, Me or NO2) 31 are less convenient to use as synthons (due to their limited solubilities and stabilities) than [Ti(NBut)- Cl2(py)3] 1a in reactions with Li2[Mentaa] (cf.Scheme 1). Reaction of [Ti(NBut)(Me4taa)] 2 with 3.5 equivalents of 2,6-dimethylaniline in dichloromethane gave [Ti(NC6H3Me2- 2,6)(Me4taa)] 4 in 77% recrystallised yield (Scheme 2). The yield of 4 obtained this way is comparable to that from direct reaction of [Ti(NC6H3Me2-2,6)Cl2(py)3] with Li2[Me4taa] (71%, Scheme 1) and demonstrates the feasibility of the imide/aniline exchange route for these dibenzotetraaza[14]annulene derivatives.In a similar manner, reaction of 2 or 3 with either aniline itself or various 4-substituted anilines gave 75–95% yields of the corresponding arylimides, namely [Ti(NC6H4R-4)(Me4taa)] (R = H 8, Me 9, NO2 10 or NMe2 11) or [Ti(NC6H4Me-4)- (Me8taa)] 12. The NMR spectroscopic data for 8–12 are fully consistent with the proposed structures.The 1H NMR spectra show doublets for the o-hydrogens of the phenyl rings in the range ca. d 5.0–5.4 consistent with shielding eVects from the macrocycle aromatic rings.2258 J. Chem. Soc., Dalton Trans., 1998, Pages 2253–2259 Bergman and co-workers 46 have reported mechanistic studies for tert-butylimide/2,6-dimethylaniline exchange reactions of [Os(NBut)L] (L = h6-cymene). The reaction is proposed to go via a bis(amide) intermediate of the type [Os(NHBut)(NHR)L] (R = C6H3Me2-2,6) for which the rate of H-atom transfer to the most basic amide (in this case to NHBut from NHR) is much larger than that from NHBut back to NHR, leading to formation and release of ButNH2, giving [Os(NR)L].A similar mechanism presumably operates for tert-butylimide/aniline exchange in Scheme 2 although we were unable (as was the case in Bergman’s system) to observe equilibrium concentrations of the proposed bis(amide) intermediates [Ti(NHBut)(NHR)- (Me4taa)] when the reactions were followed by 1H NMR spectroscopy.For example, mixtures of [Ti(NBut)(Me4taa)] 2 and 2,6-dimethylaniline show resonances only for the starting materials and (with time) products 4 and ButNH2. Similarly, mixtures of 4 and 2,6-dimethylaniline show no NMR spectroscopic evidence for the bis(amide) [Ti(NHC6H3Me2-2,6)2- (Me4taa)]. This contrasts with the chemistry of the zirconium analogue of 4, namely [Zr(NC6H3Pri-2,6)(Me4taa)(py)], which reacts rapidly and irreversibly with 2,6-diisopropylaniline to form [Zr(NHC6H3Pri-2,6)2(Me4taa)].4 The diVering behaviour for Zr most likely reflects the larger covalent radius of the heavier congener.In addition to tert-butylimide/aniline exchange we have also observed arylimide/aniline exchange by a series of NMR tube experiments in CDCl3 [equations (1) and (2)]. Thus 1 : 1 mix- [Ti(NC6H4Me-4)(Me4taa)] 1 H2NPh [Ti(NPh)(Me4taa)] 1 H2NC6H4Me-4 (1) [Ti(NC6H4Me-4)(Me4taa)] 1 H2NC6H4(NO2)-4 [Ti{NC6H4(NO2)-4}(Me4taa)] 1 H2NC6H4Me-4 (2) tures of [Ti(NC6H4Me-4)(Me4taa)] 9 and aniline gave, after several days at ambient temperature, equimolar mixtures with [Ti(NPh)(Me4taa)] 8 and 4-methylaniline [equation (1)].However, when the analogous experiment was carried out with [Ti(NC6H4Me-4)(Me4taa)] 9 and 4-nitroaniline [equation (2)] near-quantitative formation of [Ti{NC6H4(NO2)-4}(Me4taa)] 10 and 4-methylaniline was observed. The crossover experiments in equations (1) and (2) suggest that in the presence of anilines, arylimide/bis(arylamide) equilibria exist in solution, and also support Bergman’s proposal that the relative basicities of the amide nitrogens in bis(amide) intermediates control the orientation of the equilibria.Interestingly, the complex [Ti(NBut)(TTP)] 6 prepared by Woo and co-workers 27 is reported not to undergo an imide exchange reaction with aniline, even though the expected product of such a process, namely [Ti(NBut)(TTP)], can be prepared by an alternative route.This observation clearly contrasts with the behaviour of the dibenzotetraaza[14]annulene complexes and might imply a greater accessibility of the titanium centre in 2 and its homologues by virtue of Ti lying further out of the N4 donor plane. It might also suggest a greater availability of the imido nitrogen lone pairs in the dibenzotetraaza[14]annulene systems, leading in turn to more facile amine to imide hydrogen transfer in the presumed first step of imide/aniline exchange.46 Such a hypothesis is supported by our previous computational studies of model square-base pyramidal systems: these predict an increase in negative charge at the imido nitrogen as the Nimide]] ] Ti]Nmacrocycle angle (average 111.38 in 3 and 4, and 104.38 in 6) is increased.28 The imido ligand in [Ti(NBut)(Me4taa)] 2 also undergoes exchange reactions with H2O and H2S to form the previously described 22 oxo– and sulfido–titanium complexes [Ti(E)- (Me4taa)] (E = O 13 or S 14) in excellent yield.Controlled conversion of the tetratolylporphyrin-supported phenylimido compound [Ti(NPh)(TTP)] to [Ti(O)(TTP)] has been described previously.36 The reaction of [Ti(NBut)(h5-C5H5)2(py)] with H2S to form the bis(m-sulfide) [Ti2(h5-C5H5)4(m-S)2] has also recently been reported.49 When the reaction of 2 with H2S was followed by 1H NMR spectroscopy there was no evidence for any intermediates.As in the case for imide/aniline exchange, we infer that H-atom transfer to the amido nitrogen of the most likely intermediate {namely [Ti(NHBut)(SH)(Me4taa)]} and subsequent elimination of RNH2 is very fast. In an attempt to prepare a model for the proposed amido intermediates [Ti(NHBut)(X)(Me4taa)] (X = RNH, OH or SH) in the imide/EH2 (E = RN, O or S) exchange process, the reaction of [Ti(NBut)(Me4taa)] 2 with MeOH and 2,6-dimethylphenol was carried out. For comparison, the reaction of 2 with pinacol was also studied (Scheme 2).Thus reaction of 2 with 2 equivalents of ROH or one of pinacol gave [Ti(OR)2(Me4taa)] (R = Me 15 or C6H3Me2-2,6 16) or [Ti{OC(Me)2C(Me)2O}- (Me4taa)] 17 in good yield. The compounds 15 and 16 are proposed to possess cis-(OR)2 geometries by analogy with the structurally characterised homologue [Ti(OSiMe3)2(Me4taa)].50 Addition of only 1 equivalent of 2,6-dimethylphenol to 2 did not give (by 1H NMR spectroscopy) any observable quantities of the mono(amide)–mono(aryloxide) species [Ti(NHBut)- (OC6H3Me2-2,6)(Me4taa)]; instead a ca. 50% conversion of 2 to 16 was observed. Since the proposed intermediate [Ti(NHBut)- (OR)(Me4taa)] in these processes does not possess any hydrogens for intramolecular transfer to NHBut (in contrast to that suggested in reactions of 2 with RNH2, H2O and H2S), we infer that intermolecular attack of a second equivalent of ROH at the NHBut amido nitrogen to form 16 must be substantially more favourable than attack at the imido nitrogen of another molecule of 2.This is consistent with previous work of Morrison and Wigley 51 who suggested that the nitrogen atoms in metal amides are much more basic than those of analogous imides. Conclusion We have described a unified route to three classes of tetraaza macrocycle-supported tert-butylimido compounds along with a related N2O2 SchiV base analogue. The imido/aniline, imido/ H2O and imido/H2S exchange reactions of the new [Ti(NR)- (Mentaa)] complexes have revealed important similarities and diVerences to previous porphyrin-supported titanium imides and dibenzotetraaza[14]annulene-supported zirconium imides.Unsuccessful attempts to prepare a mono(amide)–mono- (aryloxide) complex gave only the bis(aryloxide) 16, and so support the view that amido nitrogen lone pairs are more accessible than those of analogous imides. Acknowledgements This work was supported by grants (to P. M.) from the EPSRC, Leverhulme Trust and Royal Society.We thank the EPSRC also for a studentship (to D. S.) and the provision of an X-ray diffractometer. We also thank Professor K. L. Woo (Iowa State University) and Dr. S. D. Gray for helpful discussions and disclosure of results prior to publication. We acknowledge the use of the EPSRC Chemical Database Service at Daresbury Laboratory. References 1 H. Brand and J. Arnold, Coord. Chem. Rev., 1995, 140, 137. 2 F. A. Cotton and J.Czuchajowska, Polyhedron, 1990, 9, 2553. 3 P. Mountford, Chem. Soc. Rev., 1998, 27, 105. 4 G. I. Nikonov, A. J. Blake and P. Mountford, Inorg. Chem., 1997, 36, 1107. 5 C. E. Housemekerides, D. L. Ramage, C. M. Kretz, J. T. Shontz, R. S. Pilato, G. L. GeoVroy, A. L. Rheingold and B. S. Haggerty, Inorg. Chem., 1992, 31, 4453.J. Chem. Soc., Dalton Trans., 1998, Pages 2253–2259 2259 6 J. L. Kisko, T. Hascall and G. Parkin, J. Am. Chem. Soc., 1997, 119, 7609. 7 A. Klose, E. Solari, C.Floriani, N. Re, A. Chiesi-Villa and C. Rizzoli, Chem. Commun., 1997, 2297. 8 H. Schumann, Inorg. Chem., 1996, 35, 1808. 9 L. Giannini, E. Solari, S. De Angelis, T. R. Ward, C. Floriani, A. Chiesi-Villa and C. Rizzoli, J. Am. Chem. Soc., 1995, 117, 5801. 10 A. Martin, R. Uhrhammer, T. G. Gardner, R. F. Jordan and R. D. Rogers, Organometallics, 1998, 17, 382. 11 D. G. Black, R. F. Jordan and R. D. Rogers, Inorg. Chem., 1997, 36, 103. 12 D. G. Black, D. C. Swenson, R.F. Jordan and R. D. Rogers, Organometallics, 1995, 14, 3539. 13 E. B. Tjaden, D. C. Swenson, R. F. Jordan and J. L. Petersen, Organometallics, 1995, 14, 371. 14 P. G. Cozzi, E. Gallo, C. Floriani, A. Chiesi-Villa and C. Rizzoli, Organometallics, 1995, 14, 4994. 15 E. Solari, C. Floriani, A. Chiesi-Villa and C. Rizzoli, J. Chem. Soc., Dalton Trans., 1992, 367. 16 F. Franceschi, E. Gallo, E. Solari, C. Floriani, A. Chiesi-Villa and C. Rizzoli, Eur. J. Chem., 1996, 2, 1466. 17 F. Corazza, E. Solari, C. Floriani, A. Chiesi-Villa and C. Guastini, J. Chem. Soc., Dalton Trans., 1990, 1335. 18 J.-M. Rosset, C. Floriani, M. Mazzanti, A. Chiesi-Villa and C. Guastini, Inorg. Chem., 1990, 29, 3991. 19 P. Mountford, Chem. Commun., 1997, 2127 (Feature Article). 20 P. J. Stewart, A. J. Blake and P. Mountford, Inorg. Chem., 1997, 36, 1982. 21 P. J. Wilson, A. J. Blake, P. Mountford and M. Schröder, Chem. Commun., 1998, 1007. 22 V. L. Goedken and J. A. Ladd, J. Chem.Soc., Chem. Commun., 1982, 142. 23 S. C. Dunn, A. S. Batsanov and P. Mountford, J. Chem. Soc., Chem. Commun., 1994, 2007. 24 A. J. Blake, P. Mountford, G. I. Nikonov and D. Swallow, Chem. Commun., 1996, 1835. 25 M. J. Scott and S. J. Lippard, Organometallics, 1997, 16, 5857. 26 M. D. Fryzuk, J. B. Love and S. J. Rettig, Organometallics, 1998, 17, 846. 27 S. D. Gray, J. L. Thorman, L. M. Berreau and L. K. Woo, Inorg. Chem., 1997, 36, 278. 28 P. Mountford and D. Swallow, J. Chem.Soc., Chem. Commun., 1995, 2357. 29 S. de Angelis, E. Solari, E. Gallo, C. Floriani, A. Chiesi-Villa and C. Rizzoli, Inorg. Chem., 1992, 31, 2520. 30 S. D. Gray and L. K. Woo, personal communication. 31 A. J. Blake, P. E. Collier, S. C. Dunn, W.-S. Li, P. Mountford and O. V. Shishkin, J. Chem. Soc., Dalton Trans., 1997, 1549. 32 D. E. Wigley, Prog. Inorg. Chem., 1994, 42, 239. 33 T. S. Lewkebandra, P. H. Sheridan, M. J. Heeg, A. L. Rheingold and C. H. Winter, Inorg. Chem., 1994, 33, 5879. 34 D. L. Thorn and R. L. Harlow, Inorg. Chem., 1992, 31, 3917. 35 C. H. Winter, P. H. Sheridan, T. S. Lewkebandara, M. J. Heeg and J. W. Proscia, J. Am. Chem. Soc., 1992, 114, 1095. 36 L. M. Berreau, V. G. Young and L. K. Woo, Inorg. Chem., 1995, 34, 527. 37 R. L. Elliott, P. J. Nichols and B. O. West, Polyhedron, 1987, 6, 2191. 38 P. J. Nichols, G. D. Fallon, K. S. Murray and B. O. West, Inorg. Chem., 1988, 27, 2795. 39 W.-H. Leung, A. A. Danopoulos, G. Wilkinson, B. Hussain-Bates and M. B. Hursthouse, J. Chem. Soc., Dalton Trans., 1991, 2051. 40 W.-H. Leung, M.-C. Wu, K.-Y. Wong and Y. Wang, J. Chem. Soc., Dalton Trans., 1994, 1659. 41 We have recently reported complexes of the type [Ti(NR)L]n (R = But or 2,6-C6H3Me2-2,6; L = substituted salen; n = 1 or 2): J. M. McInnes, D. Swallow, A. J. Blake and P. Mountford, Inorg. Chem., 1998, submitted. 42 P. J. Stewart, A. J. Blake and P. Mountford, Inorg. Chem., 1997, 36, 3616. 43 M. C. W. Chan, J. M. Cole, V. C. Gibson and J. K. Howard, Chem. Commun., 1997, 2345. 44 A. Bell, W. Clegg, P. W. Dyer, M. R. J. Elsegood, V. C. Gibson and E. L. Marshall, J. Chem. Soc., Chem. Commun., 1994, 2547. 45 A. Bell, W. Clegg, P. W. Dyer, M. R. J. Elsegood, V. C. Gibson and E. L. Marshall, J. Chem. Soc., Chem. Commun., 1994, 2247. 46 R. I. Michelman, R. G. Bergman and R. A. Andersen, Organometallics, 1993, 12, 2741. 47 M. P. Coles, C. I. Dalby, V. C. Gibson, W. Clegg and M. R. J. Elsegood, Polyhedron, 1995, 14, 2455. 48 D. S. Glueck, J. Wu, F. J. Hollander and R. G. Bergman, J. Am. Chem. Soc., 1991, 113, 2041. 49 P. Mountford, J. Organomet. Chem., 1997, 528, 15. 50 C.-H. Yang, J. A. Ladd and V. L. Goedken, J. Coord. Chem., 1988, 19, 235. 51 D. L. Morrison and D. E. Wigley, J. Chem. Soc., Chem. Commun., 1995, 79. Received 8th April 1998; Paper 8/02686F

ISSN:1477-9226    

DOI:10.1039/a802686f

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

DALTON J. Chem. Soc., Dalton Trans., 1997, Pages 2257–2262 2257 Hybrid ligands: synthesis, characterization and co-ordinative properties of a mixed phosphine–‚-ketophosphorus ylide Daravong Soulivong,a Catherine Wieser,a Monique Marcellin,a Dominique Matt,*,a Anthony Harriman a and Loïc Toupet c a Université Louis Pasteur, Ecole Européenne de Chimie, Polymères et Matériaux, Groupe de Chimie Inorganique Moléculaire, URA 405 CNRS, 1 rue Blaise Pascal, F-67008 Strasbourg Cedex, France b Université Louis Pasteur, Ecole Européenne de Chimie, Polymères et Matériaux, Laboratoire de Photochimie, 1 rue Blaise Pascal, F-67008 Strasbourg Cedex, France c Groupe Matière Condensée et Matériaux, Université de Rennes I, URA 40804 CNRS, Campus de Beaulieu, F-35042 Rennes Cedex, France Direct C-phosphination at the methyl carbon of the keto-stabilized ylide MeC(O)CH]] PPh3 was achieved using 2 equivalents of LiBu and 1 equivalent of PPh2Cl. The structure of the resultant phosphine–phosphorus ylide, Ph2PCH2C(O)CH]] PPh3, was shown by an X-ray diffraction study to comprise a 1 : 1 mixture of two rotamers built around the CH2]C(O) axis, the C(]] O)CH]] PPh3 moiety of both conformers adopting a cisoid form.A rotational barrier of 10 kJ mol21 between the two isomers was found using molecular-mechanics simulation. Oxidation of the ylide with sulfur resulted in quantitative formation of the corresponding phosphine sulfide Ph2P(S)CH2- C(O)CH]] PPh3.By treating 2 equivalents of Ph3P]] CHC(O)CH2PPh2 (L) with [{Pd(h3-C3H4Me-2)Cl}2], the P-monodentate complex [PdCl(h3-C3H4Me-2)L] 1 was formed quantitatively. Reaction of 2 equivalents of L with [{PdCl(C6H4CH2NMe2-o)}2] in tetrahydrofuran afforded, in quantitative yield, the stable cationic complex [Pd(C6H4CH2NMe2-o)L]Cl 2 where the hybrid ligand is P,O-bonded to the palladium, thus illustrating the nucleophilic character of the oxygen atom of the ylide. The BF4 2 analogue of the latter complex, 3, prepared in order to exclude counter-anion effects, was obtained by treating 1 with AgBF4.The nickel complexes [Ni(h5-C5Ph5)L]X (X = I 4 or Br 5), in which the ligand is bound as in 1 and 2, were obtained by treating the corresponding [Ni(h5-C5Ph5)X(CO)] complexes with the ylide L. Keto-stabilized phosphorus ylides of the type Ph3PCH]] C(O)R (R = alkyl, aryl or alkoxyl) have been well studied in coordination chemistry,1–11 primarily because of their potential application for preparation of ethylene 12–20 and acetylene polymerization catalysts.21 With platinum and the coinage metals, such compounds usually co-ordinate via the ylidic carbon atom,22–24 whereas the early transition metals show a preference for binding at the oxygen atom.25 This disparity in the binding mode is illustrated by the examples displayed in Scheme 1.It is interesting however that in certain platinum complexes co-ordination occurs through the oxygen atom,27 although in solution such structures tend to be unstable with respect to isomerization to the corresponding C-co-ordinated species.2 It is also known that low-valent metal atoms are capable of fragmenting carbonyl-stabilized ylides, leading in particular to acyl, alkylidene 28–31 or phosphinoenolato complexes.20,32 In contrast to phosphorus ylides where co-ordination does not take place at the phosphorus centre, the analogous phosphines bind to metals by way of the phosphorus(III) donor.Hybrid ligands comprising mixed phosphine and phosphorus ylide subunits have long been known and exhibit interesting co-ordinative properties.5 Somewhat surprisingly, there are no known ylides having the generic formula Ph3PCH]] C(O)CH2PR2, a compound that combines a phosphine fragment with an ambidentate phosphorus ylide moiety. The interest in such multitopic ligands arises from their potential to form either five-membered PIII,O (ketone) or PIII,C (ylidic C) metallacycles.Furthermore, such structures are relevant to the preparation of bimetallic polymerization catalysts since they possess two discrete co-ordination domains suitable for connecting two metal centres in close proximity. A unique feature of hybrid ligands of this type is that both termini possess the necessary functionality to form phosphinoenolates 33–43 around the central carbonyl. Using a previously developed methodology, 41 the targetted hybrid Ph2PCH2C(O)CH]] PPh3 has now been synthesized in moderate yield and its ability to bind late transition metals is demonstrated. A remarkable feature of this ylide is that it forms stable palladium complexes by way of binding to the oxygen atom, these being the second known examples of such entities.Scheme 1 Examples of C- and O-bonded, keto-stabilized phosphorus ylides taken from the literature (Ti,25 Ag,22 Pd,7 Au,4 Pt 26); thf = tetrahydrofuran Cl Pd PPh3 PPh3 C(O)Ph Ph(O)C Cl Au CH PPh3 C(O)Me PPh2 C(O)Ph Ph2P Pt Cl Cl thf Ti O Cl Cl Cl Cl Me PPh3 Ag O3ClO Ph3P P CO2Me PPh2 Cl Ph2 2 + – – – – + + + + + – P LnM O P LnM PPh3 PPh3 O R2 R2 – – + + type A type B2258 J.Chem. Soc., Dalton Trans., 1997, Pages 2257–2262 Experimental General All manipulations were carried out under an argon atmosphere using standard Schlenk-tube techniques. Solvents, including CDCl3, were dried over suitable reagents and freshly distilled under argon before use. Infrared spectra were recorded on an IFS 25 Bruker spectrometer.The 1H NMR spectral data were referenced relative to residual protiated solvents (d 7.27 for CDCl3), 13C relative to CDCl3 (d 77.0) and 31P relative to external 85% H3PO4. The mass spectrum of the compound Ph3P]] CHC(O)CH2P(S)Ph2 was recorded on a TSQ70 Finnigan MAT instrument while spectra of other compounds were recorded on a ZAB HF VG analytical instrument using m-nitrobenzyl alcohol as matrix. The complexes [{PdCl(C6H4CH2- NMe2-o}2],44 [{PdCl(C3H4Me)}2] 45 and [Ni(h5-C5Ph5)(X)(CO)] (X = Br or I) 46 were prepared according to published procedures, AgBF4 was obtained from Aldrich Chemicals and silica gel 60 was obtained from Merck.Elemental analyses were performed by the Service de Microanalyse CNRS/Université Louis Pasteur and the Service Central du CNRS-Vernaison. The calculations were performed with HYPERCHEM47 using PM3 parameters. Syntheses Ph3P]] CHC(O)CH2PPh2. A solution of LiBun (1.6 mol dm23) in hexane (8.30 cm3, 13.14 mmol) was added dropwise to a solution of Ph3PCH]] C(O)Me (2.000 g, 6.26 mmol) in thf (300 cm3) maintained at 278 8C.After 1 h a solution of PPh2Cl (0.725 g, 3.29 mmol) in thf (60 cm3) was added over a period of 1 h to the dark red mixture held at 278 8C. The mixture was then stirred for 15 h at room temperature, before being concentrated to ca. 20 cm3. Addition of 12 mol dm23 HCl (0.5 cm3) caused the red colour to disappear. After removal of the solvent, the residue was dissolved in the minimum volume of CH2Cl2 and adsorbed onto a silica gel column (30 g, height ca. 70 cm, diameter 4 cm, 70–230 mesh).The column was first eluted with CH2Cl2 in order to eliminate an impurity, then with MeOH–CH2Cl2 (1 : 9 v/v) to yield Ph3P]] CHC(O)CH2PPh2 [Rf(thf) = 0.71]. Following chromatography, the phosphine may contain small amounts of an unidentified impurity [Rf(thf) = 0.80]. The latter remains attached to the Schlenk flask as a thin oily layer so that pure Ph3P]] CHC(O)CH2PPh2 can be isolated as white powder (1.750 g, 56%), m.p. 130–131 8C. IR(KBr): 1545s cm21 [n(CO)]. NMR: 1H (C6D6), d 3.47 [d, 2 H, PCH2, 2J(PH) = 0.8], 4.08 [d br, 1 H, P]] CH, 2J(PH) = 26.7] and 6.79–7.80 (25 H, aromatic H); 31P-{1H}(thf–C6D6), d 15.7 [d, PPh3, 4J(PP) ª 2] and 214.6 [d, PPh2, 4J(PP) ª 2]; 13C- {1H}(C6D6), d 33.89 [pseudo t, PCH2, J(PC) ª 16, 3J(P9C) ª 16], 40.64 [dd, P]] CH, J(PC) ª 109, 3J(PC) ª 8], 116.28–131.70 (aromatic C) and 178.94 [pseudo t, C]] O, 2J(PC) ª 2J(P9C) ª 6 Hz] (Found: C, 78.75; H, 5.5. Calc.for C33H28OP2: C, 78.85; H, 5.6%). Mass spectrum (FAB): m/z 519 (M 1 O 1 H1, 100%). Ph3P]] CHC(O)CH2P(S)Ph2. Octasulfur (0.010 g, 0.30 mmol) was treated with a solution of Ph3P]] CHC(O)CH2PPh2 (0.100 g, 0.20 mmol) in CH2Cl2 (10 cm3). After 10 min the solution was concentrated to ca. 5 cm3 and pentane was added to afford Ph3P]] CHC(O)CH2P(S)Ph2 as a pale yellow powder (0.105 g, 98%), m.p. 146 8C (decomp.). IR(KBr): 1548s [n(C]] O)] and 632s cm21 [n(P]] S)]. NMR: 1 H (CDCl3), d 3.69 [dd, 2 H, PCH2, 2J(PH) = 14.6, 4J(PH) = 1.3], 4.22 [d, 1 H, P]] CH, 2J(PH) = 24.8] and 7.33–8.04 (25 H, aromatic H); 31P-{1H}(CH2Cl2–C6D6), d 14.5 [d, PPh3, 4J(PP) ª 3] and 40.3 [d, P(S)Ph2, 4J(PP) ª 3]; 13C- {1H}(CDCl3), d 47.27 [dd, PCH2, J(PC) ª 45, 3J(P9C) ª 14], Ph2P PPh3 O 55.07 [d, br, P]] CH, J(P9C) ª 107], 124.99–134.15 (aromatic C) and 180.79 [pseudo t, C]] O, 2J(PC) ª 2J(P9C) ª 3 Hz] (Found: C, 73.9; H, 5.25.Calc. for C33H28OP2S: C, 74.15; H, 5.3%). Mass spectrum (EI): m/z 534 (M, 18%). [PdCl(C3H4Me-2){Ph2PCH2C(O)CH]] PPh3}] 1. Solid [{Pd- Cl(C3H4Me-2)}2] (0.098 g, 0.25 mmol) was added to a solution of Ph2PCH2C(O)CH]] PPh3 (0.250 g, 0.50 mmol) in CH2Cl2 (20 cm3). After 2 h pentane was added, affording complex 1 as a pale yellow precipitate. The product was recrystallized from CH2Cl2–pentane (0.280 g, 80%), m.p. 113–114 8C. IR(KBr): 1543s cm21 [n(C]] O)].NMR: 1H (CDCl3), d 1.66 (s, 3 H, Me), 2.56 (s br, 1 H, CH of allyl), 3.10 (s br, 1 H, CH of allyl), 3.15 [d, 1 H, CH of allyl, 3J(PH) = 10], 3.63 [dd, 2 H, PCH2, 2J(PH) = 12, 4J(PH) = 1], 4.19 [d, 1 H, CH of allyl, 3J(PH) = 6], 4.42 [d, 1 H, P]] CH, 2J(PH) = 23] and 7.82–7.32 (25 H, aromatic H); 31P-{1H}(CDCl3), d 15.4 (s br) and 17.8 (s br); 13C-{1H}(CDCl3), d 22.96 (s, Me), 41.12 [pseudo t, PCH2, 1J(PC) ª 3J(P9C) = 16], 53.20 [d, P]] CH, J(PC) = 100 Hz], 59.68 (s, CH2 of allyl), 67.87 (s, CH2 of allyl), 125.31–133.73 (aromatic C 1 quaternary C of allyl) and 185.81 (s, CO) (Found: C, 59.65; H, 5.0.Calc. for C37H35ClOP2Pd?0.75 CH2Cl2: C, 59.4; H, 4.8%). Mass spectrum (FAB): m/z 663.1 ([M 2 Cl]1, 100%). [Pd(C6H4CH2NMe2-o){Ph2PCH2C(O)CH]] PPh3}]Cl 2. To a solution of Ph3P]] CHC(O)CH2PPh2 (0.134 g, 0.25 mmol) in CH2Cl2 (10 cm3) was added a solution of [{Pd(C6H4CH2- NMe2-o)Cl}2] (0.074 g, 0.12 mmol) in CH2Cl2 (10 cm3). After stirring for 3 h the solution was concentrated to ca. 5 cm3 and pentane was added affording complex 2 as a pale yellow precipitate (0.179 g, 92%), m.p. 127 8C (decomp.). IR: (KBr) 1512s [n(C]] O)], (CH2Cl2) 1520s cm21 [n(C]] O)]. NMR: 1H (CDCl3), d 2.18 [d, 6 H, NMe2, 4J(PH) = 2.5], 3.83 (d, 2 H, NCH2, 4J(PH) = 2.1], 3.91 [d, 2 H, PCH2, 2J(PH) = 12.5], 4.65 [d, 1 H, P]] CH, 2J(PH) = 22.9] and 6.44–7.89 (29 H, aromatic); 31P- {1H}(CDCl3), d 14.9 [d, PPh3, 4J(PP) = 3.5] and 37.0 [d, PPd, 4J(PP) = 3.5]; 13C-{1H}(CDCl3): d 46.56 [dd, PCH2, J(PC) = 30.6, 3J(PC) = 13.4], 49.39 (d, NMe2, 3J(PC) = 2.1], 62.55 [dd, P]] CH, J(PC) ª 104, 3J(PC) ª 8], 70.44 [d, NCH2, 3J(PC) = 2.9 Hz], 122.45–148.85 (aromatic C) and 185.93 (s br, C]] O) (Found: C, 64.35; H, 5.1; N, 1.8.Calc. for C42H40ClNOP2Pd: C, 64.8; H, 5.2; N, 1.8%). Mass spectrum (FAB): m/z 742 ([M 2 Cl]1, 100%). [Pd(C6H4CH2NMe2-o){Ph2PCH2C(O)CH]] PPh3}]BF4 3. Silver tetrafluoroborate (0.013 g, 0.06 mmol) was added to a solution of [Pd(C6H4CH2NMe2-o){Ph2PCH2C(O)CH]] PPh3}]Cl (0.050 g, 0.06 mmol) in CH2Cl2 (10 cm3).After filtration over Celite, the solution was concentrated and hexane added to afford complex 3 as pale yellow crystals (0.049 g, 94%), m.p. 127 8C (decomp.). IR(KBr): 1512s [n(C]] O)] and 1056s cm21 [n(B]F)]. The 1H and 31P NMR data are substantially the same as for 2 (Found: C, 54.75; H, 4.55; N, 1.45. Calc. for C42H40- BF4NOP2Pd?1.5CH2Cl2: C, 54.6; H, 4.55; N, 1.45%). The presence of CH2Cl2 was confirmed by NMR spectroscopy.Mass spectrum (FAB): m/z 742 ([M 2 BF4]1, 100%). [Ni(Á5-C5Ph5){Ph2PCH2C(O)CH]] PPh3}]I 4. To a solution of [Ni(h5-C5Ph5)I(CO)] 46 (0.131 g, 0.20 mmol) in thf (5 cm3) was added a solution of Ph2PCH2C(O)CH]] PPh3 (0.100 g, 0.20 mmol) in thf (5 cm3). After stirring for 1 h, the solution was concentrated to ca. 5 cm3 and pentane was added to yield a red precipitate. Recrystallization from CH2Cl2–pentane at 220 8C afforded complex 4 as red spindly crystals (0.160 g, 71%), m.p. 172 8C (decomp.).IR(KBr): 1521s cm21 [n(C]] O)]. NMR: 1H (CDCl3), d 3.61 [d, 2 H, PCH2, 2J(PH) = 9.8], 4.40 (d, 1 H, P]] CH, 2J(PH) = 33.6] and 7.42–9.31 (50 H, aromatic H); 31P-{1H}(CDCl3), d 12.8 (s br, PPh3) and 26.1 (s br, PNi); 13C-{1H}(CDCl3), d 48.28 [dd, PCH2, J(PC) = 30, 3J(PC) = 14],J. Chem. Soc., Dalton Trans., 1997, Pages 2257–2262 2259 65.49 [d, P]] CH, J(PC) ª 102 Hz], 109.37–133.11 (aromatic C) and 184.49 (s, C]] O) (Found: C, 67.3; H, 4.65. Calc.for C68H53INiOP2?CH2Cl2: C, 68.0; H, 4.55%). The presence of CH2Cl2 was confirmed by NMR spectroscopy. Mass spectrum (FAB): m/z 1005 ([M 2 I]1, 100%). [Ni(Á5-C5Ph5){Ph2PCH2C(O)CH]] PPh3}]Br 5. To a solution of [Ni(h5-C5Ph5)Br(CO)] (0.100 g, 0.16 mmol) in CH2Cl2 (20 cm3) was added Ph2PCH2C(O)CH]] PPh3 (0.081 g, 0.16 mmol). After stirring overnight the red solution was concentrated and Et2O was added to yield a brick-red precipitate. Compound 5 was recrystallized from CH2Cl2–Et2O and dried in vacuo for 3 d at 60 8C (0.175 g, 85%), m.p. 192–194 8C.IR: (KBr) 1506s [n(C]] O)], (Nujol) 1505s cm21 [n(C]] O)]. The 1H, 31P and 13C NMR spectra are essentially identical with those of 3 (Found: C, 75.2; H, 4.85. Calc. for C68H53BrNiOP2: C, 75.15; H, 4.9%). Mass spectrum (FAB): m/z 1005 ([M 2 Br]1, 100%). Crystallography Suitable single crystals of Ph3P]] CHC(O)CH2PPh2 were obtained by slow diffusion of hexane into a solution in dichloromethane.A crystal of approximate size 0.30 × 0.30 × 0.45 mm was studied on a CAD4 Enraf-Nonius diffractometer. The cell parameters were obtained by fitting a set of 25 high-q reflections. The data collection [tmax = 60s, h 0–19, k 0–19, l 224 to 24, intensity controls without appreciable decay (0.2%)] gave 10 230 reflections from which 4691 were independent (Rint = 0.019) with I > 2s(I). After Lorentz-polarization corrections, the structure was solved with direct methods which revealed many non-hydrogen atoms of the two molecules of the asymmetric unit.The remaining ones were found after successive scale-factor calculations and Fourier-difference syntheses. After isotropic (R = 0.10), then anisotropic refinement (R = 0.079), many hydrogen atoms were found from a Fourierdifference map (between 0.37 and 0.17 e Å23). The entire structure was refined by the full-matrix least-squares technique {on F; x, y, z, bij for P, C and O atoms and x, y, z fixed for H atoms; 650 variables; w = 1/s2(Fo)2 = [s2(I) 1 (0.04Fo 2)2]� �� } with the resulting R = 0.063 and Sw = 2.46 (residual Dr < 1.3 e Å23, residuals located around the phosphorus atoms).Atomic scattering factors were taken from ref. 48. All calculations were performed on a Digital Micro VAX 3100 computer with the MOLEN package.49 No absorption corrections were applied. Main crystal data are summarized in Table 1. Atomic co-ordinates, 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 CCDC for this material should quote the full literature citation and the reference number 186/522. Results and Discussion A rational synthetic procedure that should permit facile generation of the hybrid ligand from its natural precursors relies on the acidity of the methyl group in the simple ylide MeC- (O)CH]] PPh3.In fact phosphination of the methyl carbon of carbonyl compounds of the type MeC(]] O)R [R = aryl, alkoxy, dialkylamino or alkyl(aryl)amino] using a strong base and a chlorophosphine is now a well established method allowing the preparation of b-carbonyl phosphines of the type R92PCH2C- (O)R.39,41,50–61 Molecular orbital calculations concluded that the acidity of the terminal methyl group in the ylidic precursor was comparable to that of acetophenone, the prototypic reagent for usuch phosphination reactions.Trying to apply this methodology for the synthesis of Ph3P]] CHC(O)CH2PPh2, we found that treatment of Ph3PCH]] C(O)Me with 1 equivalent of LiBu in tetrahydrofuran at 278 8C followed by reaction with PPh2Cl (1 equivalent) afforded the required compound in only ca. 15% yield after work-up. In contrast acetophenone gives almost quantitative yields under the same conditions.62 The presence of large amounts of unchanged Ph3P]] CHC(O)Me in the product mixture suggests that much of the butyllithium is directly involved with the ylidic fragment and is unavailable for deprotonation of the methyl group.Consequently repeating the reaction with 2 equivalents of base caused a four-fold increase in yield after work-up (Scheme 2). It is noteworthy that by minor modification of the original synthetic approach41 the methodology has been extended to encompass the synthesis of highly polar b-ketophosphines.The compound Ph3P]] CHC(O)CH2PPh2 was characterized by microanalysis, mass and multinuclear (1H, 13C and 31P) NMR spectroscopy. The 31P NMR spectrum displays a doublet at d 214.6 [4J(PP) = 2 Hz], assigned as the phosphorus(III) atom, and another one at d 115.7 corresponding to the ylide moiety. The ylidic carbon resonance is found at d 40.64, while that of the PCH2 carbon lies at d 33.89. In the 1H NMR spectrum (200 MHz) the PCH2 protons give rise to a sharp doublet [2J(PH) = 0.8 Hz] at d 3.47, while the methine resonance appears as a relatively broad doublet centred at d 4.08.The 2J(PHylide) (26.7 Hz) and J(PCylide) (109 Hz) coupling constants are as expected for an sp2-hybridized carbon atom attached to the PPh3 moiety. Calculations indicate that the methylene group is not acidic and there is no indication that a second diphenylphosphino residue is incorporated into the molecular framework. X-Ray crystallography revealed that in the solid state Ph3P]] CHC(O)CH2PPh2 is evenly distributed between two rotamers (Fig. 1); important bond lengths and angles are given in Table 2. These rotamers differ according to the relative positioning of the carbonyl group which can be considered to be bisecting with respect to the adjacent methylene group. Taking the CH2C(O) fragment as reference, rotamer A has the carbonyl group bisecting the HCH group at an angle of 258 whereas B has the carbonyl displaced to an angle of 1658.Molecular mechanics simulations suggest an energy barrier of only ca. 10 kJ mol21 for interconversion between these rotamers. Owing to the multiple bond order within the PCCO fragment of the ketoylide part it is convenient to distinguish between these rotamers in terms of the geometry around the ylidic CC bond where both conformers adopt a cisoid arrangement. In fact this stereochemistry is normal for keto-stabilized ylides. It is noteworthy that the observed 31P chemical shift for the ylidic phosphorus atom (d 15) confirms that the cisoid configuration persists in solution.63 As expected, the Ph3P]C bond lengths, 1.708(5) Å in A and 1.723(6) Å in B, being typical for stabilized ylides,64,65 are signifi- cantly shorter than the P]C bond of the corresponding phosphine fragments [1.830(6) for A and 1.840(6) Å in B].Owing to conjugation, the carbonyl bonds [1.241(7) and 1.233(7) Å, respectively in A and B] are elongated while the intervening C]C bonds are significantly shortened [1.390(8) Å for C(19)– C(20) of A, 1.398(8) Å for C(59)–C(60) in B].In each isomer the ylidic phosphorus atom adopts a slightly distorted tetrahedral Scheme 2 (i) In thf at 278 8C, 2 equivalents LiBu, then PPh2Cl (1 equivalent), room temperature (r.t.), 15 h, then HCl; (ii) in CH2Cl2, r.t., S8 Ph2P O PPh3 P O PPh3 S Ph Ph MeC(O)CH=PPh3 (i ) (ii )2260 J. Chem. Soc., Dalton Trans., 1997, Pages 2257–2262 structure since the aryl groups are directed away from the P]] C bond.This has the effect of extending the C (ylide)]P]C (aryl) angles (average 1128 in A and B) while compressing the C (aryl)]P]C (aryl) bond angles (average 1078 in A and B). The usual pyramidal compression 66 is observed for the phosphino PIII (C]P]C angles lying between 99.0 and 102.88 in A, between 99.9 and 102.28 in B). Reaction of Ph3P]] CHC(O)CH2PPh2 with sulfur in CH2Cl2 gave the corresponding phosphine sulfide in quantitative yield (Scheme 2).Solid-state IR spectroscopy shows that the n(CO) bands of these two compounds appear at 1545 and 1548 cm21, respectively. These values, being approximately 100 cm21 lower than for a normal ketone, appear to be in keeping with strong conjugation between C]] O and P]] C bonds. Table 1 Crystal data and details of data collection for Ph2PCH2C- (O)CH]] PPh3 Formula C33H28OP2 M 502.54 Colour White Crystal system Monoclinic Space group P21/c (no. 14) a/Å 16.973(6) b/Å 16.774(5) c/Å 20.242(9) b/8 108.89(3) U/Å3 5452(3) Z 8 F(000) 2112 Dc/g cm23 1.224 m/cm21 1.775 Radiation Mo-Ka, graphite monochromated l/Å 0.710 73 Mode 2q flying step-scan T/K 294 q Range/8 1–25 Number of data collected 10 230 Number of data with I > 2s(I) 4691 Ra 0.063 R9 b 0.061 Goodness of fit c 2.46 a R = S(||Fo| 2 |Fc||)/S|Fo|.b R9 = [Sw(|Fo| 2 |Fc|)2/Sw|Fo|2]� �� . c [Sw(|Fo| 2 |Fc|)2/(No 2 Np)]� �� , where No, Np are the number of observations and parameters. Table 2 Selected bond distances (Å) and angles (8) for both isomers of Ph2PCH2C(O)CH]] PPh3 Molecule A P(1)]C(19) 1.708(5) C(20)]O(1) 1.241(7) P(1)]C(1) 1.811(6) C(20)]C(21) 1.530(8) P(1)]C(7) 1.824(5) P(2)]C(21) 1.830(6) P(1)]C(13) 1.799(6) P(2)]C(28) 1.843(6) C(19)]C(20) 1.390(8) P(2)]C(22) 1.848(6) C(7)]P(1)]C(19) 115.9(3) C(22)]P(2)]C(28) 99.0(3) C(1)]P(1)]C(19) 106.4(3) C(21)]P(2)]C(22) 101.1(3) C(13)]P(1)]C(19) 114.0(3) C(21)]P(2)]C(28) 102.8(3) C(1)]P(1)]C(13) 107.4(3) P(1)]C(19)]C(20) 121.6(3) C(13)]P(1)]C(7) 105.8(3) P(2)]C(21)]C(20) 114.0(4) C(7)]P(1)]C(1) 106.9(3) C(19)]C(20)]C(21) 115.6(5) Molecule B P(11)]C(59) 1.723(6) C(60)]O(11) 1.233(7) P(11)]C(53} 1.812(6) C(60)]C(61) 1.533(8) P(11)]C(47) 1.809(6) P(12)]C(61) 1.840(6) P(11)]C(41) 1.823(6) P(12)]C(68) 1.837(6) C(59)]C(60) 1.398(8) P(12)]C(62) 1.848(6) C(41)]P(11)]C(59) 113.6(3) C(62)]P(12)]C(68) 100.2(3) C(47)]P(11)]C(59) 116.9(3) C(61)]P(12)]C(62) 99.9(3) C(53)]P(11)]C(59) 106.2(3) C(61)]P(12)]C(68) 102.2(3) C(53)]P(11)]C(41) 107.7(3) P(11)]C(59)]C(60) 120.0(5) C(47)]P(11)]C(53) 107.3(3) P(12)]C(61)]C(60) 112.1(4) C(47)]P(11)]C(41) 104.6(3) C(59)]C(60)]C(61) 117.1(6) The compound Ph3P]] CHC(O)CH2PPh2 has three binding sites, namely the phosphorus(III) centre, the oxygen atom, and the ylidic C atom, which could facilitate mono- and bi-dentate complexation with suitable metals. In particular, the diphenylphosphino moiety remains readily accessible and the ubiquitous co-ordinative ability of such donor sites provides the opportunity to generate an inordinately wide range of metal complexes.This concept was demonstrated by virtue of the preparation and characterization of the P-monodentate complex 1 as a prototype. This allylic palladium complex was obtained in essentially quantitative yield from [{PdCl(h3-C3H4- Me-2)}2] (Scheme 3). The 31P NMR spectrum clearly resolves the two types of phosphorus and indicates that the phosphine centre is co-ordinated.More significantly, the carbonyl function appears at 1543 cm21 in the solid-state IR spectrum which remains very similar to that found for the free ylide. Such behaviour is a clear indication that the carbonyl group is not involved in the primary co-ordination sphere. Consequently it is interesting to compare this reaction with that described by Facchin et al.67 where the incoming allylic palladium fragment forms a metal–carbon bond with the ylidic carbon atom.This suggests that the phosphino residue shows greater affinity for the metal centre than does the ylidic subunit despite that the latter bears a high negative charge. In order to test the generality of the monodenticity of Ph3P]] CHC(O)CH3PPh2, we made use of a related cyclometallated palladium(II) dimer. Howen of 2 equivalents of the ylide with [{PdCl(C6H4CH2NMe2-o}2] in thf afforded complex 2 in quantitative yield (Scheme 3), but the expected 29 was not observed.In fact compound 29 can be considered to undergo an intramolecular substitution reaction in which the oxygen of the carbonyl group displaces the bound chloride ligand. We are aware of only one other example of a keto-stabilized phosphorus ylide O-bound to a palladium centre,27 but here the cationic precursor might be expected to facilitate formation of the observed product. In our case oxygen binding is more surprising and involves competition for the metal centre between chloride, oxygen and the ylidic carbon.Preferential co- Fig. 1 Molecular structure (MOLVIEW) of rotamers A and B of Ph3P]] CHC(O)CH2PPh2 showing the atom numbering schemeJ. Chem. Soc., Dalton Trans., 1997, Pages 2257–2262 2261 Scheme 3 (i) In CH2Cl2, [{Pd(h3-C3H4Me-2)}2] (0.5 equivalent); (ii) in CH2Cl2, [{PdCl(C6H4CH2NMe-o)}2] (0.5 equivalent); (iii) in CH2Cl2, AgBF4 (1 equivalent); (iv) in thf or CH2Cl2, [Ni(h5-C5Ph5)X(CO)] (1 equivalent) Ph2P O PPh3 Pd P Me O PPh3 Cl N Pd Me Me P O PPh3 N Pd Me Me P O PPh3 (i ) Cl- + (ii ) BF4 - Ni Ph Ph Ph Ph Ph P O P 3 Ph Ph Ph Ph Ph + 4 X = I 5 X = Br XPh Ph Ph Ph (iii ) Ph Ph 1 (iv ) 2 + ordination at oxygen rather than carbon seems likely because of the higher negative charge on the oxygen atom and also because of stereochemical crowding. Displacement of the chloride ligand could be promoted by a high trans influence of the s-bound aryl ring but even so a detailed understanding of the markedly disparate behaviour displayed by 2 and 29 remains elusive.Authentication of the assigned structure for complex 2 was obtained by elemental analysis, mass, NMR and IR spectroscopies. Critical information was obtained by solid-state IR in that the n(C]O) band appears at 1512 vs. 1545 for 1 and 1548 cm21 for the ylides. This decrease in the n(C]O) frequency is characteristic of keto-stabilized ylide ligands bound through the oxygen atom.1 Further support for an O-bound ketoylide relates to the displacement of the 13C signal of the ylidic carbon to lower field relative to the free ylide (d 62.55 for 2 against 40.64 for the ylide).2 Confirmation that the chloride anion of 2 is non-bonded was obtained by the facile formation of the corresponding BF4 2 salt 3 (Scheme 3).The latter complex had identical spectroscopic properties to those described for 2. Finally it should be mentioned here that Keim and coworkers 68 have recently demonstrated that P,O chelation of bcarbonyl phosphines may occur under forcing conditions to a palladium allyl centre.This suggests that conditions might be found that would enable P,O chelation with Ph3P]] CHC(O)- CH2PPh2 and an allylic fragment. A second illustration of the nucleophilicity of the oxygen atom in Ph3P]] CHC(O)CH2PPh2 was found by reaction with [Ni(h5-C5Ph5)I(CO)] (Scheme 3) where the chelated complex 4 was obtained in high yield. This product is a rare example of a keto-stabilized ylide nickel complex.That binding occurs through the oxygen atom was unequivocally demonstrated by the appearance of a single carbonyl band at 1521 cm21 in the IR spectrum. The corresponding bromide salt, 5, was obtained by treating the ylide with [Ni(h5-C5Ph5)Br(CO)]. It is noteworthy N Pd Me Me P Cl O PPh3 Ph2 2' + – that the n(C]] O) band of 5 appears at a slightly lower frequency (1505 cm21) than that of 4, possibly due to ion-pairing effects. In summary, the ligand presented in this work combines a phosphine donor with a highly polarized subunit, a ketostabilized ylide.Owing to the high negative charge localized on the ylidic function, this phosphine displays a marked tendency to form P,O-chelated complexes with palladium and nickel. Further investigations are in progress aimed at the use of such systems for the preparation of novel materials based on the presence of polar groups and/or the reactivity of the ylidic function. Acknowledgements We gratefully acknowledge Professor G.Facchin (Padova) for his interest in this work and for communicating results prior to publication. We are indebted to Johnson Mathey for a generous loan of palladium. References 1 U. Belluco, R. A. Michelin, R. Bertani, G. Facchin, G. Pace, L. Zanotto, M. Mozzon, M. Furlan and E. Zangrando, Inorg. Chim. Acta, 1996, 252, 355. 2 G. Facchin, L. Zanotto, R. Bertani, L. Canovese and P. Uguagliati, J. Chem. Soc., Dalton Trans., 1993, 2871. 3 J. Vicente, M.-T. Chicote, J. Fernández-Baeza, F. J. Lahoz and J. A. López, Inorg. Chem., 1991, 30, 3617. 4 J. Vicente, M. T. Chicote, I. Saura-Llamas, J. Turpin and J. Fernandez-Baeza, J. Organomet. Chem., 1987, 333, 129. 5 H. Schmidbaur, Angew. Chem., Int. Ed. Engl., 1983, 22, 907. 6 W. Malisch, H. Blau and F. J. Haaf, Chem. Ber., 1981, 114, 2956. 7 P. Bravo, G. Fronza and C. Ticozzi, J. Organomet. Chem., 1976, 111, 361. 8 M. Kato, H. Urabe, Y. Oosawa, T. Saito and Y. Sasaki, J.Organomet. Chem., 1976, 121, 81. 9 Y. Oosawa, H. Urabe, T. Saito and Y. Sasaki, J. Organomet. Chem., 1976, 122, 113. 10 S. Kato, T. Kato, M. Mizuta, K. Itoh and Y. Ishii, J. Organomet. Chem., 1973, 51, 167. 11 J. Buckle, P. G. Harrison, T. J. King and J. A. Richards, J. Chem. Soc., Chem. Commun., 1972, 1104. 12 G. Braca, A. M. Raspolli Galletti, M. Di Girolamo, G. Sbrana, R. Silla and P. Ferrarini, J. Mol. Catal., 1995, 96, 203. 13 K. Kurtev and A. Tomov, J. Mol. Catal., 1994, 88, 141.2262 J.Chem. Soc., Dalton Trans., 1997, Pages 2257–2262 14 P. Braunstein, Y. Chauvin, S. Mercier, L. Saussine, A. De Cian and J. Fischer, J. Chem. Soc., Chem. Commun., 1994, 2203. 15 A. Krentsel and L. A. Nekhaeva, Russ. Chem. Rev. (Engl. Transl.), 1990, 59, 1193. 16 G. A. Nesterov, G. Fink and V. A. Zakharov, Makromol. Chem. Rapid. Commun., 1989, 10, 669. 17 W. Keim, New J. Chem., 1987, 11, 531. 18 K. A. Ostoja Starzewski and J. Witte, Angew. Chem., Int.Ed. Engl., 1987, 26, 63. 19 U. Klabunde, T. H. Tulip, D. C. Roe and S. D. Ittel, J. Organomet. Chem., 1987, 334, 141. 20 W. Keim, F. H. Kowaldt, R. Goddard and C. Krüger, Angew. Chem., Int. Ed. Engl., 1978, 17, 466. 21 K. A. Ostoja Starzewski and L. Born, Organometallics, 1992, 11, 2701. 22 M. C. Gimeno, P. G. Jones, A. Laguna and M. D. Villacampa, J. Chem. Soc., Dalton Trans., 1995, 805. 23 M. C. Gimeno, P. G. Jones, A. Laguna, M. Laguna and I. Lázaro, J. Chem. Soc., Dalton Trans., 1993, 2223. 24 G.Facchin, R. Bertani, L. Zanotto, M. Calligaris and G. Nardin, J. Organomet. Chem., 1989, 366, 409. 25 J. A. Albanese, D. L. Staley, A. L. Rheingold and J. L. Burmeister, Inorg. Chem., 1990, 29, 2209. 26 M. Kato, H. Urabe, Y. Oosawa, T. Saito and Y. Sasaki, Chem. Lett., 1976, 51. 27 R. Usón, R. Forniés, R. Navarro, P. Espinet and C. Mendívil, J. Organomet. Chem., 1985, 290, 125. 28 D. Heineke and H. Vahrenkamp, Chem. Ber., 1993, 126, 365. 29 D. Heineke and H. Vahrenkamp, Chem.Ber., 1993, 126, 373. 30 D. Heineke, D. S. Bohle and H. Vahrenkamp, Chem. Ber., 1993, 126, 355. 31 A. J. Deeming, D. Nuel, N. I. Powell and C. Whittaker, J. Chem. Soc., Dalton Trans., 1992, 757. 32 H. Qichen, X. Minzhi, Q. Yanlong, X. Weihua, S. Meicheng and T. Youqi, J. Organomet. Chem., 1985, 287, 419. 33 J. Andrieu, P. Braunstein and F. Naud, J. Chem. Soc., Dalton Trans., 1996, 2903. 34 P. Braunstein, Y. Chauvin, J. Nähring, A. De Cian, J. Fischer, A.Tiripicchio and F. Ugozzoli, Organometallics, 1996, 15, 5551. 35 D. Matt, R. Ziessel, A. De Cian and J. Fischer, New J. Chem., 1996, 20, 1257. 36 D. Matt, M. Huhn, M. Bonnet, I. Tkatchenko, U. Englert and W. Kläui, Inorg. Chem., 1995, 34, 1288. 37 J. Bank, P. Steinert, B. Windmüller, W. Wolfsberger and H. Werner, J. Chem. Soc., Dalton Trans., 1996, 1153. 38 B. Demerseman and P. Crochet, Organometallics, 1995, 14, 2173. 39 D. Soulivong, R. Ziessel and D. Matt, J. Organomet.Chem., 1994, 474, 207. 40 D. Matt, M. Huhn, A. De Cian, J. Fischer, I. Tkatchenko and M. C. Bonnet, J. Chem. Soc., Dalton Trans., 1993, 1173. 41 S. E. Bouaoud, P. Braunstein, D. Grandjean, D. Matt and D. Nobel, Inorg. Chem., 1986, 25, 3765. 42 C. J. Moulton and B. L. Shaw, J. Chem. Soc., Dalton Trans., 1980, 229. 43 H. D. Empsall, S. Johnson and B. L. Shaw, J. Chem. Soc., Dalton Trans., 1980, 302. 44 A. C. Cope and E. C. Friedrich, J. Am. Chem. Soc., 1968, 90, 909. 45 W. D.Dent, R. Long and A. J. Wilkinson, J. Chem. Soc., 1964, 1585. 46 W. Kläui and L. Ramacher, Angew. Chem., Int. Ed. Engl., 1986, 25, 97. 47 HYPERCHEM, version 5, Hypercube Inc., Ontario, 1997. 48 International Tables for X-Ray Crystallography, D. Riedel, Boston, MA, 1983, Table A, p. 183. 49 C. K. Fair, MOLEN, An Interactive Intelligent System for Crystal Structure Analysis, Enraf-Nonius, Delft, 1990. 50 B. Guilbert, B. Demerseman, P. H. Dixneuf and C. Mealli, J. Chem. Soc., Chem. Commun., 1989, 287, 419. 51 L. Douce and D. Matt, C. R. Acad. Sci., Ser. 2, 1990, 310, 721. 52 D. A. Knight, D. J. Cole-Hamilton and D. C. Cupertino, J. Chem. Soc., Dalton Trans., 1990, 3051. 53 D. Matt and A. Van Dorsselaer, Polyhedron, 1991, 10, 1521. 54 R. K. Boeckman, jun., T. M. Kamenecka, S. G. Nelson, J. R. Pruitt and T. E. Barta, Tetrahedron Lett., 1991, 32, 2581. 55 S. D. Perera and B. L. Shaw, J. Organomet. Chem., 1991, 402, 133. 56 J. Andrieu, P. Braunstein and A. D. Burrows, J. Chem. Res., 1993, (S) 380. 57 D. Matt, N. Sutter-Beydoun, A. El Amiri, J.-P. Brunette, P. Briard and D. Grandjean, Inorg. Chim. Acta, 1993, 208, 5. 58 D. Matt, N. Sutter-Beydoun, J.-P. Brunette, F. Balegroune and D. Grandjean, Inorg. Chem., 1993, 32, 3488. 59 D. Soulivong, D. Matt, R. Ziessel, L. Douce and R. Deschenaux, Tetrahedron Lett., 1993, 34, 1151. 60 B. Demerseman, R. Le Lagadec, B. Guilbert, C. Renouard, P. Crochet and P. H. Dixneuf, Organometallics, 1994, 13, 2269. 61 J. Andrieu, P. Braunstein, A. Tiripicchio and F. Ugozzoli, Inorg. Chem., 1996, 35, 5975. 62 M. Huhn, D. Matt and P. Braunstein, Inorg. Synth., 1997, 31, 138. 63 J. M. Brittain and R. A. Jones, Tetrahedron, 1979, 35, 1139. 64 D. G. Gilheany, in The Chemistry of Organophosphorus Compounds, ed. F. R. Hartley, Wiley, Chichester, 1994, vol. 3, pp. 1–44. 65 S. M. Bachrach and C. I. Nitsche, in The Chemistry of Organophosphorus Compounds, ed. F. R. Hartley, Wiley, Chichester, 1994, vol. 3, pp. 273–302. 66 T. A. Albright, J. K. Burdett and M.-H. Whangbo, Orbital Interactions in Chemistry, Wiley-Interscience, New York, 1985. 67 G. Facchin, R. Bertani, M. Calligaris, G. Nardin and M. Mari, J. Chem. Soc., Dalton Trans., 1987, 1381. 68 G. J. P. Britvosek, W. Keim, S. Mecking, D. Sainz and T. Wagner, J. Chem. Soc., Chem. Commun., 1993, 1632. Received 14th February 1997; Paper 7/01057E

ISSN:1477-9226    

DOI:10.1039/a701057e

出版商:RSC    

年代:1997

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1998, Pages 2261–2266 2261 A new polymorph of LiZnPO4?H2O; synthesis, crystal structure and thermal transformation † Torben R. Jensen* Chemistry Department, University of Odense, DK-5230 Odense M, Denmark A new polymorph of lithium zinc phosphate hydrate, b-LiZnPO4?H2O, has been obtained by hydrothermal synthesis, and its crystal structure and thermal transformation investigated. The crystal structure can be viewed as a framework structure built from PO4 and ZnO4 tetrahedra with LiO4 tetrahedra and water molecules placed in eight-ring channels; b-LiZnPO4?H2O is apparently the only member of the zeolite ABW group of materials crystallising in space group P21ab.It has a doubled b axis compared to a known phase of LiZnPO4?H2O. The 31P MAS NMR spectrum showed two signals; d 6.94(2) and 7.88(2). The known phase of LiZnPO4?H2O transforms into b-LiZnPO4?H2O after treatment in 0.9 M LiNO3 (aq) at 80 8C and 10 kbar for 20 h.A mechanism for the pressure-induced displacive phase transition is proposed. Dry or hydrothermal heating of b-LiZnPO4?H2O at >100 8C gave dI-LiZnPO4. The enthalpy change for the dehydration with a maximum at 219(2) 8C was determined to be DH = 43(6) kJ mol21, using diVerential scanning calorimetry. Microporous materials have found extensive industrial use over the last decades, e.g. as gas absorbers, ion exchangers, catalysts and many other applications.1 Such materials with new framework structures and compositions are of great importance as their function is often correlated with the internal shape and size of cages and channels and the position of metal ions in the material.Among new materials a large number of phosphates and some arsenates have been characterised during the last decade.2 The chemistry of zinc orthophosphates shows a large variety of phases mainly due to the co-ordination flexibility of the zinc ions and their ability to form Zn]O]Zn linkages.3 An increasing number of orthophosphates are found in the system, ZnO– P2O5–cation, prepared by hydrothermal synthesis from aqueous or non-aqueous solutions using alkali-metal ions and/or organic amine ions as templates. There are several members of this group which are structural analogues to aluminosilicates, e.g.sodalite, zeolite-X and ABW,4 but also compounds with novel crystal structures with no naturally occurring mineral counter part, e.g.Zn2(HPO4)3?H3NCH2CH2NH3.5 A member of this family, Zn(H2PO4)(HPO4)?(CH3)4N, possesses very low framework density, 10.1 tetrahedral framework atoms (T) per 1000 Å3.6 In general microporous materials are metastable as they transform to more dense phases upon heating either dry or hydrothermal. Careful adjustment of synthesis conditions is crucial for preparation of new materials, e.g. stabilisation and isolation of intermediate phases in the synthesis of more dense phases. Hydrothermal preparation techniques have revealed the existence of several new lithium zinc phosphates and the crystallisation fields of the compounds are under further investigation.7 A polymorph of LiZnPO4 having a phenacite type structure (denoted e-LiZnPO4 in this work for distinction from the other polymorphs),8 a cristobalite related phase, dI-LiZnPO4,9 and a polymorph (denoted a9-LiZnPO4) 7 with a more complex crystal structure can all be prepared by hydrothermal methods.The compound a9-LiZnPO4 is possibly structurally related to a-LiZnPO4,7 which is the stable phase under ambient conditions prepared by solid state reaction.A common feature of the lithium zinc phosphates, a-, e- and dI-LiZnPO4 (and possibly also a9-LiZnPO4), are three-dimensional framework structures built from LiO4, ZnO4 and PO4 tetrahedra forming * E-Mail: trj@dou.dk † Non-SI unit employed: bar = 101 325 Pa. six-ring channels.7–10 The system Li3PO4–Zn3(PO4)2 was studied by means of solid state reactions and thermal analysis. The equilibrium phase diagram shows the existence of four lithium zinc phosphates with the compositions Li4Zn(PO4)2, Li9- Zn6(PO4)7, LiZnPO4 and ‘LiZn9(PO4)7’, and each of them may exist in one, two or three modifications.The phase diagram reveals that a-LiZnPO4 transforms to b-LiZnPO4 at 737 8C, g-LiZnPO4 at 1006 8C and melts at 1172 8C.11 It is not possible to quench b- or g-LiZnPO4 to ambient conditions as the phase transitions are reversible and the only available structural data is a powder pattern of b-LiZnPO4.9b A polymorph of Li4- Zn(PO4)2 and the mentioned phases of LiZnPO4 seem to be the only compositions from the phase diagram that can be prepared by means of hydrothermal synthesis.12 A water-containing lithium zinc phosphate, LiZnPO4?H2O, was described having a zeolite type ABW structure, i.e.isomorphous with zeolite Li-A(BW), LiAlSiO4?H2O.9a An accurate structure analysis of LiAlSiO4?H2O was performed by combining single crystal and powder diVraction data, using X-rays and neutron radiation, respectively.13 The structure can be viewed as a framework built from AlO4 and SiO4 tetrahedra giving four- and eight-ring channels in the crystallographic c direction and six-ring channels in a perpendicular direction. The lithium ions are co-ordinated to three framework oxygen atoms and one water oxygen atom.13 The polymorph of LiZnPO4?H2O9a found to be isomorphous with zeolite ABW is denoted a-LiZnPO4?H2O in the following.The polymorph of LiZnPO4?H2O discovered in this study has similar unit cell parameters to those of a-LiZnPO4?H2O, but the b axis is doubled, therefore, denoted b-LiZnPO4?H2O. The synthesis, crystal structure and thermal properties of b-LiZnPO4?H2O are now reported. Experimental Hydrothermal synthesis In a search for new materials the system Li2O–ZnO–P2O5– template was investigated by hydrothermal synthesis in aqueous and non-aqueous solutions using organic amine ions as templates.A synthesis was performed by dissolving LiOH?H2O (3.677 g), LiCl (3.709 g) and Zn(CH3CO2)2?2H2O (18.971 g) in water (100 cm3). Phosphoric acid, 85% H3PO4 (10.0 cm3), and then ethyldiisopropylamine (30.0 cm3) was slowly added with stirring giving a molar ratio of the reactants Li :Zn:PO4 : NEtPri 2 :water of 1.2 : 0.6:1:1.2:ª41, [PO4] ª 0.9 M, pH 4.552262 J. Chem. Soc., Dalton Trans., 1998, Pages 2261–2266 (measured 1 h after mixing).After stirring the gel for 1 h, powder diVraction showed the presence of a-LiZnPO4?H2O of poor crystallinity along with a trace of an unidentified phase. A portion of the gel, ca. 5 cm3, was left to stand for 3 months at room temperature, ca. 21 8C, before crystallisation was noticed and after 6 months the crystallisation was complete and the material (Sample I) was recovered by filtration and washed with deionised water. Another portion of the gel was heated at 55 8C for 3 months giving a-LiZnPO4?H2O as product.The rest of the gel was heated to 200 8C for 68 h in a Teflon-lined steel autoclave giving dI-LiZnPO4 as product. In order to investigate the crystallisation fields of lithium zinc phosphates formed in the system LiOH–Zn(CH3CO2)2–H3PO4– water, a number of syntheses were performed varying the amount of LiOH and the temperature. The LiOH and Zn(CH3CO2)2 were dissolved in water and H3PO4 was slowly added with stirring forming a white gel.The relative amount of lithium is given as a variable n, relative to the amount of phosphoric acid: n = n(Li)/n(PO4). The molar composition of the synthesis mixtures, Li :Zn:PO4:H2O, was n:0.52:1:ª20, where n = 1.49, 2.00, 2.48 or 2.99, [PO4] ª 1.8 M, 4.0 < pH < 7.2, thermal treatment at 5, 21 or 52 8C for 28 d and at 120 8C for 6 d. The phase identification was performed by combining information from 31P MAS NMR and powder diVraction and the products are displayed in Fig. 1. In one case a phase-pure sample of b-LiZnPO4?H2O was obtained (filled square) and the other samples contained impurities of b-Li3PO4 (open symbols). At high temperatures, e.g. 120 8C, a9-LiZnPO4 (circles) formed and at low temperatures a-LiZnPO4?H2O (triangles) was found as product. Optimisation of the conditions for the hydrothermal synthesis revealed that templates were in fact not essential for the preparation of b-LiZnPO4?H2O. Several syntheses in the system LiOH–LiCl–Zn(CH3CO2)2–H3PO4– (CH2OH)2 were performed in a similar way substituting the solvent water with ethylene glycol but gave only semiamorphous material with traces of a9-LiZnPO4 and b-Li3PO4 as products.A Teflon container holding a suspension of a-LiZnPO4?H2O (0.381 g) in 0.9 M LiNO3 (aq) (10 cm3) and H3PO4 (0.1 cm3) were heated in a PSIKA 20 kbar reactor at 80 8C and 10 kbar for 20 h, using oil as supporting fluid. The composition of the suspension was Li :Zn:PO4 :water = 2.6:0.58:1:ª142.The product was washed with ethanol (99.9%) and identified as b-LiZnPO4? H2O using powder diVraction. A Radiometer glass electrode combined with a calomel reference electrode (GK2401C) was used for pH measurements. The syntheses were performed using the following commercial chemicals: LiOH?H2O (Fluka, puriss p.a., >99%), LiNO3 Fig. 1 The crystallisation fields of lithium zinc orthophosphates formed in the system LiOH–Zn(CH3CO2)2–H3PO4–water, at varying amounts of LiOH, n, and temperature. In one case a phase-pure sample of b-LiZnPO4?H2O (j) was obtained.Open symbols indicate impurities of b-Li3PO4, and a-LiZnPO4?H2O is shown as triangles, b-LiZnPO4?H2O as squares and a9-LiZnPO4 as circles (Merck, >98%), LiCl (Merck, pro analysi, >99%), H3PO4 (85%, Fluka, extra pure), Zn(CH3CO2)2?2H2O (Fluka, purum p.a. >99.0%), ethyldiisopropylamine (Aldrich, >99%) and ethylene glycol (Fluka, puriss p.a., >99.5%).Powder diVraction Powder diVraction data for phase identification and refinement of unit cell parameters were obtained using a Siemens D5000 diVractometer equipped with a primary germanium monochromator (Cu-Ka1 radiation, l = 1.540 598 Å). The data were collected from 5 and up to 908 in 2q with a step length of 0.028 and counting time of 2–15 s per step. The observed powder pattern of b-LiZnPO4?H2O was distinctly diVerent from that of the known zeolite type ABW polymorph, a-LiZnPO4?H2O, as shown in Fig. 2. The trial and error indexing program TREOR14 revealed an orthorhombic unit cell of b-LiZn- PO4?H2O similar to the unit cell of a-LiZnPO4?H2O with a doubled b axis. The program CELLKANT15 was used to refine the unit cell parameters from the observed d spacings. The indexed powder pattern of b-LiZnPO4?H2O (39 reflections, I/I0 > 2%) is given in Table 1 using the refined unit cell dimensions of b-LiZnPO4?H2O, a = 10.030(1), b = 16.553(2) and c = 5.0120(5) Å, unit cell volume = 832.1(2) Å3.Single crystal diVraction A clear, colourless octahedron of b-LiZnPO4?H2O was selected from sample I. The data collection was performed on a Siemens SMART diVractometer equipped with a CCD detector and graphite monochromatised Mo-Ka radiation, l = 0.710 73 Å. Approximately one hemisphere of data was collected in frames covering 0.38 in w in three sets at diVerent f angles using a detector to crystal distance of 40.0 mm. Data were corrected for Lorentz-polarisation eVects and for absorption using Gaussian integration [transmission (maximum/minimum) 0.6383/0.3748] giving 1359 unique reflections.A structural model was found using direct methods in space group P21ab, no. 29. This setting was chosen to resemble the unit cell of a-LiZnPO4?H2O refined in space group Pna21, no. 33.9a Programs from the Siemens SHELXTLTM software package were used for structure solution and refinement.16 After anisotropic refinement it was possible to locate the positions of the four hydrogen atoms in the Fourier-diVerence maps.The H atom positions and thermal parameters were not refined; all other atoms were refined with anisotropic thermal parameters. Scattering factors of neutral atoms were applied throughout. The refinement converged at R = 0.0290 and R9 = 0.0717, with 146 parameters refined using full-matrix least squares based on all F2. Atomic coordinates are shown in Table 2. Drawings of the crystal structure were prepared using the program ATOMS.17 Fig. 2 Powder X-ray diVraction patterns of a- (bottom) and b-LiZnPO4?H2OJ. Chem. Soc., Dalton Trans., 1998, Pages 2261–2266 2263 Crystal data. LiO4PZn?H2O, M = 185.30, orthorhombic, space group P21ab (no. 29), a = 10.0224(9), b = 16.5596(15), c = 5.0126(5) Å, U = 831.93(13) Å3, Dc = 2.959, Dm = 3.01 g cm23 (measured using a pycnometer and water), T = 293 K, Z = 8, m(Mo-Ka) = 6.189 mm21, 3348 reflections measured, 1359 unique (Rint = 0.0362).CCDC reference number 186/1004. See http://www.rsc.org/suppdata/dt/1998/2261/ for crystallographic files in .cif format. Spectroscopic investigation Magic angle spinning 31P NMR spectra were recorded on a Varian UNITY-500 spectrometer [11.7 T, n(31P) = 202.332 MHz], using a Jakobsen MAS probe.18 The samples were placed in a cylindrical rotor (Si3N4, 5 mm outside diameter, volume 220 ml). The free induction decays were recorded with a spinning speed of 5.5 kHz and the resulting spectra had a digital resolution of 2.5 Hz.An aqueous solution of H3PO4 (85%, Fluka, extra pure) was used as external standard. Infrared spectra were measured on a Perkin-Elmer 1720 Fouriertransform spectrometer using pellets of KBr containing 1% of the compound. The spectral resolution was 4 cm21. Thermal investigation A thermogravimetric (TG) measurement was performed between 30 and 500 8C, using a SETARAM TG 92-12 instrument, a nitrogen atmosphere and a heating rate of 5 8C min21.The sample was kept in an open Al2O3 crucible. DiVerential scanning calorimetry (DSC) was performed using a Table 1 Indexed powder pattern of b-LiZnPO4?H2O h 121001212312021330210220104334244506124 k 2012420342435452405131277846268636803 10 5 l 001101110011111102122221101120112021302 dcalc/Å 6.3838 5.0152 4.3275 4.2872 4.1382 3.9422 3.5452 3.4796 3.1919 3.1001 3.0409 2.9826 2.7623 2.6923 2.6632 2.6365 2.6007 2.5060 2.4196 2.4055 2.2817 2.2214 2.1645 2.1386 2.0916 2.0691 1.9717 1.9587 1.9489 1.8562 1.7870 1.7402 1.6876 1.6225 1.5955 1.5858 1.5790 1.5719 1.5627 dobs/Å 6.3834 5.0173 4.3299 4.2842 4.1438 3.9433 3.5451 3.4827 3.1938 3.1009 3.0426 2.9826 2.7644 2.6918 2.6629 2.6352 2.6012 2.5074 2.4194 2.4052 2.2818 2.2229 2.1649 2.1384 2.0916 2.0688 1.9712 1.9593 1.9498 1.8562 1.7869 1.7399 1.6880 1.6224 1.5957 1.5856 1.5784 1.5720 1.5627 (I/I0)calc 100 4 12 79 36 22 16 2 58 69 895 16 24 7 32 16 858 22 28 14 35676857695 13 5 (I/I0)obs 100 77 74 46 18 15 3 49 65 794 18 21 6 30 16 749 22 36 15 46785857778 11 5 SETARAM DTA 92-16.18 instrument.The sample was placed in a platinum crucible and a-Al2O3 used as reference. The experiment was carried out between 20 and 900 8C using a heating and cooling rate of 5 8C min21 in an argon atmosphere. Two experiments were carried out using the same sample. The temperature and enthalpy change calibration was carried out using the low–high quartz transition and the dehydration of CaSO4?2H2O.Baseline fluctuations in the temperature range 20 to 130 8C are due to small diVerences between the set temperature and the actual temperature in the DSC instrument. Results and Discussion Thermogravimetric measurements were in accordance with the formula b-LiZnPO4?H2O, observed weight loss 9.1% at 120– 210 8C (calculated 9.7%). The density of b-LiZnPO4?H2O was measured using pycnometry and water to Dm = 3.01 g cm23, and the calculated density was Dc = 2.959 g cm23.The density of a-LiZnPO4?H2O is Dc = 2.886 g cm23.9a Phosphorus-31 MAS NMR spectroscopy on b-LiZn- PO4?H2O showed two signals with chemical shifts d(31P) 6.94(2) and 7.88(2), in accordance with the crystallographic data revealing two phosphorus atoms in the asymmetric unit. The two polymorphs of LiZnPO4?H2O thus have distinct phosphorus chemical shifts as a-LiZnPO4?H2O shows only one signal [at d(31P) 6.05(2)].9b Infrared spectra of the two polymorphs of LiZnPO4?H2O are illustrated in Fig. 3. The absorptions at ca. 3500 and ca. 1600 cm21 are due to stretch and bending vibrations in water molecules. A broad absorption in the spectrum of a-LiZnPO4?H2O around 3400 cm21 indicates the presence of adsorbed water, whereas multiple absorptions in the range 3300–3600 cm21 suggest more than one inequivalent crystal water molecule in b-LiZnPO4?H2O. Orthophosphates generally show a strong absorption band in the range 940–1120 and medium at 540–650 cm21, due to framework vibrations.19 DiVerences in this part of the spectra, Fig. 3, suggest diVerences in the framework structure of the two polymorphs, as will be discussed in the following. Crystal structure of ‚-LiZnPO4?H2O The crystal structure of b-LiZnPO4?H2O is built from regular tetrahedra of ZnO4 and PO4 and slightly distorted tetrahedra of LiO4. Selected bond lengths and angles are found in Table 3. Chains of alternating ZnO4 and PO4 tetrahedra run in the c direction.One half the chains have all tetrahedra pointing up (U) and the other chains have tetrahedra pointing down (D). Fig. 4(a) illustrates the connection between two chains forming zigzag four-ring chains parallel to the c axis. Fig. 4(b) displays how the double chains of four-ring chains are interconnected to Fig. 3 Infrared spectra of a- (A) and b-LiZnPO4?H2O (B)2264 J. Chem. Soc., Dalton Trans., 1998, Pages 2261–2266 form nets of six rings with the tetrahedra sequence UUUDDD, a characteristic feature of the ABW topology.6b,13 Two types of six rings appear trigonally and ovally distorted having the symmetry of twofold axes and a glide planes, respectively.The trigonal shape resembles the shape of six rings found in the structures of a-LiZnPO4?H2O and tridymite, whereas the oval resembles dI-LiZnPO4 and cristobalite (the sequence of tetrahedra in a six ring of the three latter compounds are UDUDUD).9b,20 The b,c projection, Fig. 4(b), consists of pseudo-hexagonal layers stacked with the sequence ABA to Fig. 4 Crystal structure of b-LiZnPO4?H2O: (a) the b,c projection, showing four-ring chains in the c direction; (b) connectivity of the fourring chains forming nets of six-ring channels with stacking sequence, ABA, in the (100) direction; (c) the a,b projection, showing four- and eight-ring channels in the c direction with water and Li atoms placed in the eight rings. The ZnO4 are grey and PO4 are dark tetrahedra, lithium ions are shown as medium size black circles and water molecules as small (H atoms) and large (O atoms) black circles ( a) ( b) ( c) form a three-dimensional framework structure, as illustrated in Fig. 4(c) viewed along the c axis, showing another characteristic feature of the ABW topology: four- and eight-ring channels in the (001) direction. The water molecules and lithium ions are placed in the eight-ring channels as shown in Fig. 4(c). Oxidation states, Vi, were calculated as a sum of bond valences, s, using the equation s = exp[(r0 2 r)/B], where r0 and B are empirical parameters [r0(P]O) = 1.617, r0(Zn]O) = 1.704, r0(Li]O) = 1.466, r0(H]O) = 0.882 and B = 0.37 Å] and r is the bond length from the refined structural model.21 The calculated oxidation states, Vi, are in Table 4, and the calculated values are in agreement with the expected oxidation states. The values of the oxygen atoms fall in three groups.The framework oxygen atoms co-ordinated to one P, Zn and Li each, O(1)–O(6), have calculated oxidation states, Vi, in the range 1.94–2.01.The framework oxygen atoms co-ordinated to two cations, O(7) and O(8), have lower Vi, 1.83 and 1.81, possibly due to a bond valence contribution from hydrogen bonding. The oxygen atoms in the water molecules, O(9) and O(10), gave values of 2.07 and 1.75, respectively. The low value of O(10) is due to long O(10)]H(3,4) bonds. Comparison with related zeolite ABW type materials A variety of diVerent materials adopt the zeolite ABW type structure and crystallise with at least six diVerent space group symmetries.22 A new modification of the ABW structure is adopted by b-LiZnPO4?H2O crystallising with space group symmetry P21ab.The material LiAlSiO4?H2O, zeolite Li- A(BW), and a-LiZnPO4?H2O crystallise in the space group Pna21, i.e. the twofold screw axes are placed inside and parallel to the four-ring chains. The tetrahedra in each four ring are pairwise symmetry related and all tetrahedra point in the same direction viewed along the polar c axis in the structure of a- (see Fig. 5) whereas in b-LiZnPO4?H2O one half point up and one half down, UUDD [see Fig. 4(c)]. In the structure of b-LiZn- PO4?H2O the twofold screw axes are in the centre of the six rings, and perpendicular to the four-ring chains, i.e. the tetrahedra in a four ring are crystallographically distinct. The connectivity of the tetrahedra is the same for LiAl- SiO4?H2O, a-LiZnPO4?H2O and b-LiZnPO4?H2O; they all have the characteristic ABW topology and as other lithiumcontaining zeolitic materials they have high framework densities, 19.0, 18.8 and 19.2 T/1000 Å3, respectively.Table 2 Atomic coordinates (×104) and equivalent isotropic displacement parameters (Å2 × 103) for b-LiZnPO4?H2O; Ueq is defined as one third of the trace of the orthogonalised Uij tensor. Parameters for H atoms were not refined Atom Zn(1) Zn(2) P(1) P(2) Li(1) Li(2) O(1) O(2) O(3) O(4) O(5) O(6) O(7) O(8) O(9) O(10) H(1) H(2) H(3) H(4) x 1521(1) 9184(1) 8494(2) 7232(2) 9007(17) 1768(13) 9584(5) 1915(5) 2393(5) 8326(5) 1116(5) 8865(5) 2181(5) 8531(5) 393(6) 261(6) 1026 84 529 9428 y 6912(1) 5564(1) 6581(1) 4082(1) 7524(8) 4971(7) 7041(3) 5974(3) 6764(3) 5740(3) 5463(3) 6510(3) 7953(3) 4552(3) 8247(3) 4260(4) 8000 8707 3788 4500 z 4606(2) 1108(2) 5875(3) 149(4) 923(20) 4694(24) 4330(10) 6867(9) 1078(9) 4615(10) 1360(10) 8823(9) 5676(10) 9731(10) 9846(11) 5348(11) 8687 9102 6369 5914 Ueq 10(1) 10(1) 7(1) 8(1) 16(3) 15(3) 13(1) 13(1) 11(1) 10(1) 12(1) 12(1) 14(1) 12(1) 22(1) 23(1) 90 90 90 90J.Chem. Soc., Dalton Trans., 1998, Pages 2261–2266 2265 Synthesis In a previous study a-LiZnPO4?H2O was found to crystallise from the system LiOH–Zn(NO3)2–H3PO4–water.4 Hydrothermal syntheses from the system LiOH–Zn(CH3CO2)2– H3PO4–water suggest that a-LiZnPO4?H2O is in fact a metastable intermediate phase that transforms to b-LiZnPO4?H2O, i.e.the system approaches higher density and higher thermodynamic stability: gel æÆ a-LiZnPO4?H2O æÆ b-LiZnPO4? H2O. The material a-LiZnPO4?H2O has only been obtained as a microcrystalline powder, whereas b-LiZnPO4?H2O can be obtained as larger crystals. This is probably due to the very diVerent timescales of nucleation and crystal growth for a- and b-LiZnPO4?H2O. Thermal investigation Phase transformations of a- and b-LiZnPO4?H2O were investigated by hydrothermal and dry heating.Hydrothermal heating of a-LiZnPO4?H2O at ultra high pressure (80 8C, 10 kbar, 20 h) produced b-LiZnPO4?H2O, i.e. approaching higher density. Fig. Table 3 Selected bond lengths (Å) and angles (8) for b-LiZnPO4?H2O. Symmetry transformations used to generate equivalent atoms: (i) x 2 1, y, z; (ii) x, y, z 2 1; (iii) x 1 1, y, z; (iv) x 1 ��� , 2y 1 ��� , z; (v) x 1 ��� , 2y 1 1, 2z 1 1; (vi) x 1 ��� , 2y 1 1, 2z; (vii) x 1 1, y, z 2 1; (viii) x 2 ��� , 2y 1 1, 2z 1 1; (ix) x 2 1, y, z 1 1 Zn(1)]O(7) Zn(1)]O(1i) Zn(1)]O(2) Zn(1)]O(3) Zn(2)]O(8ii) Zn(2)]O(5iii) Zn(2)]O(6ii) Zn(2)]O(4) P(1)]O(6) P(1)]O(7iv) P(1)]O(4) P(1)]O(1) P(2)]O(2v) P(2)]O(8ii) P(2)]O(3vi) P(2)]O(5vi) Li(1)]O(9vii) Li(1)]O(1) Li(1)]O(6ii) Li(1)]O(3iv) Li(2)]O(10) Li(2)]O(5) Li(2)]O(2) Li(2)]O(4viii) O(9)]H(1) O(9)]H(2) O(10)]H(3) O(10)]H(4) 1.922(5) 1.958(5) 1.964(5) 1.988(5) 1.927(5) 1.948(5) 1.967(5) 1.979(5) 1.528(5) 1.529(6) 1.538(5) 1.541(5) 1.532(5) 1.532(5) 1.537(5) 1.546(5) 1.91(2) 1.972(12) 1.987(13) 2.00(2) 1.943(14) 1.971(13) 1.992(13) 1.985(13) 0.952(5) 0.904(6) 0.972(6) 1.018(6) O(7)]Zn(1)]O(1i) O(7)]Zn(1)]O(2) O(1i)]Zn(1)]O(2) O(7)]Zn(1)]O(3) O(1i)]Zn(1)]O(3) O(2)]Zn(1)]O(3) O(8ii)]Zn(2)]O(5iii) O(8ii)]Zn(2)]O(6ii) O(5)]Zn(2)]O(6ii) O(8ii)]Zn(2)]O(4) O(5iii)]Zn(2)]O(4) O(6ii)]Zn(2)]O(4) O(6)]P(1)]O(7iv) O(6)]P(1)]O(4) O(7iv)]P(1)]O(4) O(6)]P(1)]O(1) O(7iv)]P(1)]O(1) O(4)]P(1)]O(1) O(2v)]P(2)]O(8ii) O(2v)]P(2)]O(3vi) O(8ii)]P(2)]O(3vi) O(2v)]P(2)]O(5vi) O(8ii)]P(2)]O(5vi) O(3vi)]P(2)]O(5vi) O(9vii)]Li(1)]O(1) O(9vii)]Li(1)]O(6ii) O(O(3iv) O(1)]Li(1)]O(3iv) O(6ii)]Li(1)]O(3iv) O(10)]Li(2)]O(5) O(10)]Li(2)]O(2) O(5)]Li(2)]O(2) O(10)]Li(2)]O(4viii) O(5)]Li(2)]O(4viii) O(2)]Li(2)]O(4viii) Li(1ix)]O(9)]H(1) Li(1ix)]O(9)]H(2) H(1)]O(9)]H(2) Li(2)]O(10)]H(3) Li(2)]O(10)]H(4) H(3)]O(10)]H(4) 105.3(2) 118.6(2) 109.1(2) 102.0(2) 112.7(2) 109.1(2) 106.7(2) 115.4(2) 105.5(2) 107.4(2) 112.8(2) 109.2(2) 108.1(3) 110.8(3) 109.6(3) 110.6(3) 109.1(3) 108.7(3) 109.9(3) 111.0(3) 108.6(3) 110.9(3) 107.5(3) 108.9(3) 106.6(8) 115.6(7) 97.9(6) 103.3(6) 116.2(6) 117.4(8) 97.8(6) 118.0(6) 98.2(6) 102.9(6) 130.8(6) 109.9(6) 112.8(6) 113.3(7) 109.8(5) 111.2(6) 84.3(6) 159.4(7) 5 shows the b,c projection of a-LiZnPO4?H2O with four-ring chains in the c direction and arrows indicate how a- transforms into b-LiZnPO4?H2O by a displacive phase transition [compare Fig. 5 and 4(a)]. Half the chains of tetrahedra (the lower chain in Fig. 5) are distorted as indicated by the arrows, i.e. the tetrahedra are rotated by ca. 458. The tetrahedra in the upper chain are all moved slightly. The mean angle between the framework tetrahedral atoms, T]O]T angles, in a- and b-LiZnPO4?H2O are similar, 125.2 and 128.38, respectively. Hydrothermal heating of a-LiZnPO4?H2O at T > 100 8C was found to produce dI-LiZnPO4,9 and the mechanism appeared to be solution mediated.23 The compound b-LiZnPO4?H2O also transforms into dI-LiZnPO4 upon hydrothermal treatment at >100 8C.The dry heating dehydration of b-LiZnPO4?H2O was investigated using diVerential scanning calorimetry giving an enthalpy change of DH = 43(6) kJ mol21, peak value at 219(2) 8C, for the irreversible phase transformation of b-LiZn- Fig. 5 Crystal structure of a-LiZnPO4?H2O, the b,c projection showing four-ring chains in the c direction. The arrows indicate a mechanism for a displacive phase transition of a- to b-LiZnPO4?H2O, i.e.the tetrahedra in the lower chain are rotated by ca. 458 and the upper chain is moved slightly downwards. The ZnO4 are shown as grey and PO4 as dark tetrahedra Table 4 Calculated oxidation states, Vi, for the structure of b-LiZn- PO4?H2O, using ref. 21 Cation Phosphorus Zinc Lithium Oxygen Framework Water No. 121212 123456789 10 Vi 5.01 4.97 2.02 2.03 1.04 1.02 1.99 2.00 1.94 1.96 1.98 2.01 1.83 1.81 2.07 1.75 Co-ordination number 444444 33333322332266 J.Chem. Soc., Dalton Trans., 1998, Pages 2261–2266 PO4?H2O into dI-LiZnPO4. A shoulder, at ca. 206 8C, was found on the DSC signal for the dehydration, Fig. 6, but the TG trace suggests a one-step process. A weak signal at 738(1) 8C shows the phase transition of dI-LiZnPO4 into b-LiZnPO4 in accordance with the equilibrium phase diagram.11 The DSC trace resembles a shift in baseline indicating that mainly a change in heat capacity is taking place.Cooling b-LiZnPO4 produces a-LiZnPO4 [a = 17.240(2), b = 9.768(1), c = 17.106(1) Å and b = 110.90(1)8, refined from powder diVraction data, 118 reflections] and the a- to b-LiZnPO4 phase transition is shown to be reversible in Fig. 6. The zeolites MAlSiO4?xH2O (M = Li, Na or Ag) show a large framework flexibility as the dehydrated polymorphs may retain the ABW topology and the aluminosilicates of Na and Ag have partly dehydrated ABW polymorphs (x = 0.8 and x = 0.8 and 0.68, respectively).24 Further heating of the dehydrated ABW type LiAlSiO4 produces a stuVed cristobalite-type structure,24b which is similar to the phase transition of a- and b-LiZnPO4? H2O to dI-LiZnPO4. Conclusion Investigation of the system LiOH–Zn(CH3CO2)2–H3PO4– water, revealed that a new phase of LiZnPO4?H2O can be prepared under ambient conditions. A known polymorph of LiZnPO4?H2O (ABW structure, space group Pna21) was found as an intermediate phase during the synthesis and transforms into b-LiZnPO4?H2O by hydrothermal heating at ultra high pressure.A mechanism for the displacive phase transition is proposed. Hydrothermal or dry heating of b-LiZnPO4?H2O at >100 8C produces dI-LiZnPO4. The crystal structure of b-LiZn- PO4?H2O can be viewed as framework structures built from PO4 and ZnO4 tetrahedra with LiO4 tetrahedra and water molecules placed in eight-ring channels running in the c direction. This compound is apparently the only member of the zeolite ABW group of materials crystallising in space group P21ab.Fig. 6 Thermal investigation of b-LiZnPO4?H2O showing the mass loss detected by thermogravimetry (upper graph) and diVerential scanning calorimetry; the lower and the middle DSC curves are the first and second heating and the upper DSC curve is one of the very similar cooling curves Acknowledgements R. G. Hazell, P. Norby (Aarhus University), P. C. Stein, O. Simonsen and E. M. Skou (Odense University) are thanked for valuable discussions.A Danish Technical Research Council Ph.D. grant is gratefully acknowledged. The Siemens SMART diVractometer at the Chemistry Department, Aarhus University was partly financed by Carlsberg fondet. I am grateful to M. B. Nielsen for help. The PSIKA 20 kbar reactor at the Chemistry Department, Odense University was financed by The Danish Natural Science Research Council. References 1 C. P. Grey, F. I. Poshni, A. F. Gualtieri, P. Norby, J. C. Hanson and D.R. Corbin, J. Am. Chem. Soc., 1997, 119, 1981; Yun-Jo Lee and Hakze Chon, J. Chem. Soc., Faraday Trans., 1996, 3453; H. van Bekkum, E. M. Flanigen and J. C. Jansen, Stud. Surf. Sci. Catal., 1991, 58. 2 (a) M. Estermann, L. B. McCusker, C. Baerlocher, A. Merrouche and H. Kessler, Nature (London), 1991, 352, 320; (b) V. Soghomonian, Qin Chen, R. C. Haushalter, J. Zubieta and C. J. O’Connor, Science, 1993, 259, 1596; (c) G. W. Noble, P. A. Wright and Å. Kvick, J. Chem. Soc., Dalton Trans., 1997, 4485; (d ) Pingyun Feng, Xianhui Bu and G.D. Stucky, Nature (London), 1997, 388, 735; (e) Sue-Lein Wang, Kuei-Fang Hsu and Yeu-Perng Nieh, J. Chem. Soc., Dalton Trans., 1994, 1681; ( f ) Pingyun Feng, Xianhui Bu and G. D. Stucky, Acta Crystallogr., Sect. C, 1997, 53, 997; ( g) T. R. Jensen, P. Norby, J. C. Hanson, E. M. Skou and P. C. Stein, J. Chem. Soc., Dalton Trans., 1998, 527. 3 Tianyou Song, M. B. Hursthouse, Jiesheng Chen, Jianing Xu, K. M. A.Malik, R. H. Jones, Ruren Xu and J. M. Thomas, Adv. Mater., 1994, 6, 679; ref. 2( g) and refs. therein. 4 T. E. Gier and G. D. Stucky, Nature (London), 1991, 349, 508. 5 Tianyou Song, Jianing Xu, Yu-e Zhao, Yong Yue, Yihua Xu, Ruren Xu, Ninghai Hu, Gecheng Wei and Hengqing Jia, J. Chem. Soc., Chem. Commun., 1994, 1171. 6 (a) W. T. A Harrison and L. Hannooman, Angew Chem., Int. Ed. Engl., 1997, 36, 640; (b) W. M. Meier and D. H. Olsen, Atlas of Zeolite Structure Types, 3rd edn., Butterworth-Heinemann, London, 1992. 7 T. R. Jensen, unpublished work. 8 Xianhui Bu, T. E. Gier and G. D. Stucky, Acta Crystallogr., Sect. C, 1996, 52, 1601. 9 (a) W. T. A. Harrison, T. E. Gier, J. M. Nicol and G. D. Stucky, J. Solid State Chem., 1995, 114, 249; (b) T. R. Jensen, P. Norby, P. C. Stein and A. M. T. Bell, J. Solid State Chem., 1995, 117, 39. 10 J. A. Gard, G. Torres-Trevino and A. R. West, J. Mater. Sci. Lett., 1985, 4, 1138; L. Elammari and B. Elouadi, Acta Crystallogr., Sect. C, 1989, 45, 1864. 11 G. Torres-Trevino and A. R. West, J. Solid State Chem., 1986, 61, 56. 12 T. R. Jensen and A. Nørlund Christensen, unpublished work. 13 P. Norby, A. Nørlund Christensen and I. G. Krogh Andersen, Acta Chem. Scand., Sect. A, 1986, 40, 500; E. Krogh Andersen and G. Ploug-Sørensen, Z. Kristallogr., 1986, 176, 67. 14 P.-E. Werner, L. Eriksson and M. Westdahl, J. Appl. Crystallogr., 1985, 18, 367. 15 N. O. Ersson, CELLKANT, Chemical Institute, Uppsala University, Uppsala, 1981. 16 G. M. Sheldrick, SHELXTL, Siemens Analytical X-Ray Systems Inc., Madison, WI, 1995. 17 E. Dowty, ATOMS, version 4.0, Shape Software, Kingsport, TN, 1997. 18 H. J. Jakobsen, P. Daugaard and V. Langer, J. Magn. Reson., 1988, 76, 162; U.S. Pat., 4 739 270, 1988. 19 R. A. Nyquist and R. O. Kagel, Infrared spectra of inorganic compounds, Academic Press, New York and London, 1971, p. 11. 20 A. Putnis, Introduction to mineral sciences, Cambridge University Press, 1992, p. 172. 21 I. D. Brown and D. Altermatt, Acta Crystallogr., Sect. B, 1985, 41, 244. 22 P. Norby, Ph.D. Thesis, University of Aarhus, 1989. 23 T. R. Jensen, P. Norby and J. C. Hanson, unpublished work. 24 (a) P. Norby and H. Fjellvåg, Zeolites, 1992, 12, 898; (b) P. Norby, Zeolites, 1990, 10, 193. Received 17th March 1998; Paper 8/02113I

ISSN:1477-9226    

DOI:10.1039/a802113i

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

DALTON J. Chem. Soc., Dalton Trans., 1997, Pages 2263–2266 2263 Synthesis of dinuclear gold-(I) and -(II) complexes by reaction of [AuI 2{Ï-(CH2)2PPh2}2] with protonic acids Manuel Bardají,a Elena Cerrada,a Peter G. Jones,b Antonio Laguna *,a and Mariano Laguna *,a a Departamento de Química Inorgánica, Instituto de Ciencia de Materiales de Aragón, Universidad de Zaragoza-C.S.I.C., 50009-Zaragoza, Spain b Institut für Anorganische und Analytische Chemie der Technischen Universität, Postfach 3329, 38023 Braunschweig, Germany The reaction of the bis(ylide) gold complex [Au2{m-(CH2)2PPh2}2] with strong acids (such as HClO4) in the presence of diphosphines L]L afforded a mixture of heterobridged gold(I) complexes [Au2{m-(CH2)2PPh2}(m-L]L)]1 (L]L = Ph2PCH2CH2PPh2 or Ph2PCH2PPh2) and [PPh2Me2]1.After addition of Li(tcnq), [Au2{m- (CH2)2PPh2}(m-L]L)]2[ClO4][tcnq] were obtained pure (tcnq = 7,7,8,8-tetracyanoquinodimethane). The crystal structure of the Ph2PCH2CH2PPh2 derivative shows an intramolecular gold–gold distance of 3.1294(9) Å.The use of weak protonic acids such as pyridine-2-thiol (C5H5NS) or 2-sulfanylbenzothiazole (C7H5NS2), in the presence of atmospheric oxygen, led to complexes of gold(II). The crystal structure of [AuII 2{m-(CH2)2PPh2}2(C5H4NS)2] shows a gold–gold bond length of 2.6686(13) Å. Dinuclear gold-(I) and -(II) complexes containing the bis(ylide) ligand (CH2)2PR2 2 have been extensively studied because of their high stability, which provides a handle for the study of gold–gold bonds and weak gold–gold interactions.1–5 Fackler and co-workers 6 have shown elegantly that the isomerization of trans-[Au2{m-(CH2)2PPhMe}2] to a cis–trans mixture is acid catalysed, via a nucleophilic attack of the gold atoms on the Lewis acids.However, the use of stronger protonic acids resulted in rupture of the dimer to give two monomeric units.7 Here we describe the reactions of [Au2{m-(CH2)2PPh2}2] with a strong acid in the presence of phosphine or diphosphine ligands, which afford dinuclear gold(I) derivatives, whereas the use of weak protonic acids (that are able to act as ligands) leads to gold(II) derivatives.The structures of [AuI 2{m-(CH2)2PPh2}- (m-dppe)]2[ClO4][tcnq] (dppe = 1,2-bis(diphenylphosphino)- ethane; tcnq = 7,7,8,8-tetracyanoquinodimethane) and [AuII 2{m-(CH2)2PPh2}2(C5H4NS)2] (C5H4NS = pyridine-2-thiolate) have been determined by X-ray diffraction analyses. Results and Discussion As stated above, the symmetrical ring cleavage of [Au2{m-(CH2)2PPh2}2] by hydrogen halides gives mononuclear derivatives [AuX(CH2PPh2Me)]. It has been suggested that these processes involve a dinuclear intermediate I with one bridging bis(ylide) ligand and terminal ylide and halide ligands.7 We have now carried out reactions between [Au2- {m-(CH2)2PPh2}2] and HClO4 in the presence of (i) triphenylphosphine or (ii) diphosphine [dppm = bis(diphenylphosphine)- methane or dppe] ligands (Scheme 1); NMR studies showed the presence of dinuclear gold(I) complexes (previously reported by us) 8 and phosphonium salts, although we could not separate the two components.These processes can evolve through the intermediate proposed if the second protonation takes place at the terminal ylide instead of the bridging bis(ylide) ligand. Ph2 P Au Au X MePh2P I Therefore, we tried to isolate this intermediate from the reaction of [Au2{m-(CH2)2PPh2}2] with triphenylphosphine and perchloric acid in a 1 : 1 : 1 molar ratio [reaction (iii), Scheme 1]; the result was a mixture of starting material and doubly protonated product, but not the intermediate species. In order to eliminate the phosphonium salt, metatheses (iv) with Li(tcnq) were carried out to afford complexes 1 and 2 as green solids from which the phosphonium salt can be eliminated by fractional crystallization (Scheme 1). Complexes 1 and 2 were characterized by elemental analyses and NMR spectroscopy (see Experimental section).The IR spectra show bands at 2180s, 2154s and 828m cm21 from the tetracyanoquinodimethane anion,9–11 1098s (br) and 623s cm21 from the perchlorate anion,12 587m (1) and 582m cm21 (2) assignable to Au]Cylide.13 In the positive-ion liquid secondary ion mass spectra (LSIMS) the parent ion [Au2{m-(CH2)2PPh2}- (m-L]L)]1 is the base peak at m/z = 991 and 1005 respectively. The nature of these double salts was confirmed by an X-ray diffraction analysis of complex 2.The asymmetric unit consists of one cation [Au2{m-(CH2)2PPh2}(m-L]L)]1, one half-occupied perchlorate anion (disordered over an inversion centre) and half a tetracyanoquinodimethane anion (also associated with Scheme 1 (i) HClO4 1 PPh3; (ii) HClO4 1 L]L, 2 [PPh2Me2][ClO4]; (iii) HClO4 1 PPh3 (1 : 1 : 1); (iv) Li(tcnq), 2 LiClO4 P P Au Au L L P Au Au L L P Au Au Ph2 P Au Au PPh3 MePh2P [ClO4][tcnq] 2 Ph2 Ph2 Ph2 ClO4 Ph2 (ii ) L-L = dppm 1, dppe 2 (iv ) Ph2 P Au Au PPh3 PPh3 ClO4 ClO4 (i ) (iii )2264 J.Chem. Soc., Dalton Trans., 1997, Pages 2263–2266 an inversion centre) (Fig. 1). Selected bond lengths and angles are collected in Table 1. The structure of the cation displays a nine-membered dimetallacycle in which the atoms Au(1,2), P(1,2), C(3,4) are approximately coplanar (mean deviation 0.13 Å) and P(3), C(1) and C(2) lie 0.9, 0.6 and 20.5 Å out of this plane. The intramolecular gold–gold distance is 3.1294(9) Å, which is longer than found in other ninemembered dinuclear gold(I) complexes: 2.867(1) Å in [Au2(mmnt)( m-dppen)] 14 [mnt = maleonitriledithiolate, dppen = 1,2- bis(diphenylphosphino)ethylene], 2.90 Å in [Au2(m-mnt)(mdadpe)] 14 [dadpe = 1-(diphenylarsino)-2-(diphenylphosphino)- ethane)], 3.026(1) Å in [NBu4][Au2(m-C3S5){m-(CH2)2PPh2}].15 The Au ? ? ?Au contact is however similar to that found in the ten-membered ring [Au2{m-S(CH2)3S}(m-dppm)] [3.128(1) Å].16 The bond lengths in the tcnq are very similar to those found in Rb(tcnq),17 as would be expected for tcnq2.The tcnq anions are essentially planar (mean deviation <0.04 Å) and are associated in planes at z = 0, coplanar rather than stacked; the shortest contact is N(2) ? ? ? N(2) (2 2 x, 1 2 y, 2 2 z) 3.45 Å. We have also realized reactions of [Au2{m-(CH2)2PPh2}2] with weak protonic acids such as pyridine-2-thiol (C5H5NS) or 2- sulfanylbenzothiazole (C7H5NS2), which can act as ligands either in the protonated or in the deprotonated form.Following the latter pattern, dinuclear gold(I) derivatives containing one bridging bis(ylide) ligand and two deprotonated S-donor ligands should be obtained, whereas mononuclear gold(I) Fig. 1 Structure of the cation of complex 2 in the crystal (H atoms omitted, radii arbitrary) showing the atom numbering scheme Table 1 Selected bond lengths (Å) and angles (8) for complex 2 Au(1)]C(3) Au(1)]Au(2) Au(2)]P(2) P(1)]C(21) P(2)]C(31) P(2)]C(2) P(3)]C(3) P(3)]C(61) C(3)]Au(1)]P(1) P(1)]Au(1)]Au(2) C(4)]Au(2)]Au(1) C(11)]P(1)]C(21) C(21)]P(1)]C(1) C(21)]P(1)]Au(1) C(31)]P(2)]C(41) C(41)]P(2)]C(2) C(41)]P(2)]Au(2) C(4)]P(3)]C(3) C(3)]P(3)]C(51) C(3)]P(3)]C(61) C(2)]C(1)]P(1) P(3)]C(3)]Au(1) 2.083(7) 3.1294(9) 2.271(2) 1.822(6) 1.816(6) 1.829(6) 1.774(7) 1.810(7) 172.2(2) 101.39(5) 88.5(2) 108.0(3) 103.5(3) 114.8(2) 107.6(3) 101.7(3) 113.4(2) 108.3(4) 111.0(3) 110.8(3) 113.1(4) 110.0(3) Au(1)]P(1) Au(2)]C(4) P(1)]C(11) P(1)]C(1) P(2)]C(41) P(3)]C(4) P(3)]C(51) C(1)]C(2) C(3)]Au(1)]Au(2) C(4)]Au(2)]P(2) P(2)]Au(2)]Au(1) C(11)]P(1)]C(1) C(11)]P(1)]Au(1) C(1)]P(1)]Au(1) C(31)]P(2)]C(2) C(31)]P(2)]Au(2) C(2)]P(2)]Au(2) C(4)]P(3)]C(51) C(4)]P(3)]C(61) C(51)]P(3)]C(61) C(1)]C(2)]P(2) P(3)]C(4)]Au(2) 2.276(2) 2.098(7) 1.819(7) 1.827(6) 1.823(7) 1.763(7) 1.798(7) 1.554(8) 84.1(2) 172.5(2) 98.69(5) 103.1(3) 110.2(2) 116.2(2) 105.2(3) 109.7(2) 118.4(2) 109.8(3) 110.1(3) 106.9(3) 113.3(4) 108.5(3) derivatives with one ylide ligand and one deprotonated S-donor ligand would be expected if symmetrical ring cleavage occurs.Unexpectedly the gold(II) derivatives 3 and 4 are obtained (Scheme 2). These complexes were obtained as orange solids, air and moisture stable at room temperature. Complex 3 is nonconducting in acetone solutions (4 is too insoluble). Their IR spectra show medium-intensity bands at 573 (3) and 571 cm21 (4) from Au]Cylide, and the absorptions from N]H or S]H have disappeared.The 31P-{1H} NMR spectrum shows a singlet at d 39.7 (s) for 3 (4 is too insoluble). The 1H NMR spectrum shows the ylide methylene proton resonance as a doublet at d 1.59, and the absence of the MeP resonance rules out the formation of phosphonium. The LSIMS mass spectrum of 3 shows the parent peak at m/z (%) = 1041 (8) whilst the base peak appears at m/z = 930, corresponding to the loss of one pyridine-2- thiolate ligand from the parent peak.It is known that by bubbling a Lewis acid such as SO2 into a solution of [Au2{m-(CH2)2PR}2] (R = Ph or Me) the coloured gold(I) adducts [Au2{m-(CH2)2PR}2(SO2)2] are obtained, for which crystal structures, albeit disordered, have been determined. 6 Therefore, in order to confirm the oxidation state of the gold centres, complexes 3 and 4 have been also synthesized by a well established method of obtaining dinuclear gold(II) derivatives: substitution reactions.Thus the neutral tetrahydrothiophene in the gold(II) complex [Au2{m-(CH2)2PPh2}2(tht)2][ClO4]2 can readily be displaced by addition of a freshly prepared solution of the corresponding anionic ligand (Scheme 2). As further confirmation, the crystal structure of complex 3 has been established by X-ray crystallography (Fig. 2). Selected bond lengths and angles are collected in Table 2. The structure determination shows an octadiauracycle with chair geometry, as is common for dinuclear gold(II) derivatives; the P atoms lie 0.76 Å out of the mean plane of the other ring atoms (mean deviation <0.01 Å).The molecule displays crystallographic inversion symmetry. The Au]Au bond length 2.6686(13) Å is similar to that found in other S-bonded gold(II) systems such as [Au2{m-(CH2)2PPh2}2(S2CNMe2)2] [2.6585(12) and 2.6444(8) Å] 18 and clearly shorter than that in [Au2{m- (CH2)2PPh2}2(SO2)2] [2.835(1) Å].6 The AuII]S bond 2.437(3) Å is also similar to that found in [Au2{m-(CH2)2PPh2}2(S2CNMe2) 2] [2.439(2) and 2.431(1) Å] and shorter than in [Au2- {m-(CH2)2PPh2}2(SO2)2] [2.587(5) Å].It is clear that the reactions of Scheme 2 involve oxidation, for which we suggest atmospheric oxygen as the agent; these processes are not sensitive to light and are accelerated by bubbling oxygen through the solutions, whereas they are prevented by the use of deoxygenated solvents. The first step could be an approach of the sulfur donor ligand to give an adduct (similar to that Scheme 2 tht = Tetrahydrothiophene P P Au Au P P Au Au L L P Au Au L PPh2Me Au N S S– N S– L L P P Au Au tht tht Ph2 Ph2 Ph2 Ph2 Ph2 Ph2 Ph2 2 HL 2 HL 2 HL [ClO4]2 1/2O2 2 NaL (3) 2 –2 tht L = [PPh2Me2] –2 NaClO4 –H2O (4)J.Chem. Soc., Dalton Trans., 1997, Pages 2263–2266 2265 reported for SO2) which should be very easy to oxidize. To the best of our knowledge that is the first example in which oxidation of gold(I) complexes to gold(II) is brought about by oxygen.Other examples affording gold(III) complexes have recently been reported.19 Reactions with perchloric acid take place irrespective of the use of distilled and deoxygenated solvents. Experimental Infrared spectra were recorded on a Perkin-Elmer 883 spectrophotometer, over the range 4000–200 cm21, using Nujol mulls between polyethylene sheets, 1H and 31P NMR spectra on a Varian UNITY 300 spectrometer in CDCl3 solutions with chemical shifts quoted relative to SiMe4 (external, 1H) or H3PO4 (external, 31P).Analyses (C, H and N) were performed with a Perkin-Elmer 2400 microanalyser. Conductivities were measured in acetone solution with a Philips PW 9509 apparatus. Melting points were measured on a Büchi apparatus and are uncorrected. Mass spectra were recorded on a VG Autospec using positive-ion LSIMS techniques. All reactions were carried out at room temperature. CAUTION: perchlorate salts or derivatives must be manipulated with care, as far as possible avoiding evaporation to dryness.Syntheses [Au2{Ï-(CH2)2PPh2}(L]L)]2[ClO4][tcnq] (L]L = dppm 1 or dppe 2). To a solution of [Au2{m-(CH2)2PPh2}2] 20 (0.082 g, 0.1 Fig. 2 The molecule of complex 3 in the crystal (H atoms omitted, radii arbitrary) showing the atom numbering scheme Table 2 Selected bond lengths (Å) and angles (8) for complex 3 Au]C(1I) Au]S S]C(11) P]C(2) P]C(21) C(11)]C(16) C(13)]C(14) C(15)]C(16) C(1I)]Au]C(2) C(2)]Au]S C(2)]Au]AuI C(11)]S]Au C(1)]P]C(31) C(1)]P]C(21) C(31)]P]C(21) P]C(2)]Au N]C(11)]S C(11)]N]C(13) C(13)]C(14)]C(15) C(11)]C(16)]C(15) 2.09(2) 2.437(3) 1.78(2) 1.77(2) 1.814(14) 1.36(2) 1.31(3) 1.41(2) 175.8(5) 83.4(4) 94.5(4) 109.1(5) 110.7(7) 111.2(6) 101.7(6) 108.8(7) 117.3(12) 115(2) 120(2) 118(2) Au]C(2) Au]AuI P]C(1) P]C(31) C(11)]N N]C(13) C(14)]C(15) C(1I)]Au]S C(1I)]Au]AuI S]Au]AuI C(1)]P]C(2) C(2)]P]C(31) C(2)]P]C(21) P]C(1)]AuI N]C(11)]C(16) C(16)]C(11)]S C(14)]C(13)]N C(14)]C(15)]C(16) 2.128(14) 2.6686(13) 1.715(14) 1.80(2) 1.35(2) 1.38(2) 1.35(3) 92.7(4) 89.5(4) 176.29(10) 107.7(8) 113.3(7) 112.3(6) 114.8(8) 124(2) 118.4(14) 125(2) 119(2) Symmetry transformation used to generate equivalent atoms: I 2x 1 1, 2y 1 1, 2z 1 1.mmol) in dichloromethane (30 cm3) was added HClO4 (0.2 mmol, 0.020 g) and L]L (0.1 mmol; dppm, 0.038 g; dppe, 0.039 g). After stirring for 2 h the solvent was removed, the residue dissolved in acetone (20 cm3) and Li(tcnq) 21 (0.021 g, 0.1 mmol) added.After 2 h the solvent was concentrated to ca. 5 cm3 and addition of ethanol (15 cm3) led to green solids, which were filtered off and dried in vacuo. Yields: 1, 80%; 2, 75%. NMR: 1H, 1; d 7.9–7.5 (34 H, m, Ph and tcnq), 3.6 (2 H, s, br, dppm) and 1.9 (4 H, sd, CH2P); 31P-{1H}, 1; d 36.5 (s br) and 34.8 (t, J 12.1); 1H, 2; d 7.8–7.5 (34 H, m, Ph and tcnq), 2.7 (4 H, s br, dppe) and 1.95 (4 H, s, CH2P); 31P-{1H}, 2; d 38.2 (d) and 35.2 (t, J 10.7 Hz).M.p. 205 (1), 110 8C (2) (Found: C, 46.95; H, 3.25; N, 2.45. Calc. for C90H76Au4- ClN4O4P6 1: C, 47.3; H, 3.35; N, 2.45. Found: C, 47.7; H, 3.3; N, 2.35. Calc. for C92H80Au4ClN4O4P6 2: C, 47.75; H, 3.3; N, 2.4%). LM = 226 (1), 234 ohm21 cm2 mol21 (2). [Au2{Ï-(CH2)2PPh2}2L2] (L = C5H4NS 3 or C7H4NS2 4). (a) To a solution of [Au2{m-(CH2)2PPh2}2] (0.082 g, 0.1 mmol) in dichloromethane (30 cm3) was added HL (0.2 mmol; C5H5NS, 0.022 g; or C7H5NS2, 0.033 g). After stirring for 9 h (3) or 2 d (4), complex 4 appeared as an insoluble orange solid, whereas 3 was obtained by concentration to ca. 5 cm3 and addition of diethyl ether (15 cm3). Complex 3 was washed with diethyl ether and 4 with dichloromethane. Yields: 3, 80%; 4, 85%. (b) To a solution of [Au2{m-(CH2)2PPh2}2(tht)2][ ClO4]2 22 (0.06 g, 0.05 mmol) in dichloromethane (30 cm3) was added a slight excess of freshly prepared NaL (0.11 mmol) in ethanol (10 cm3). After stirring for 90 min (3) or 3 h (4), complex 4 appeared as an insoluble orange solid, whereas 3 was obtained by concentration to ca. 5 cm3 and addition of diethyl ether (15 cm3); 3 was washed with diethyl ether and 4 with dichloromethane. Yields: 3, 91%; 4, 73%. NMR: 1H, 3; d 8.3–6.8 (28 H, m, Ph and C5H4NS) and 1.59 [8 H, d, 2J(HP) 10.3 Hz, CH2P]; 31P-{1H}, 3; d 39.7 (s). M.p. 160 (decomp.) (3), 215 8C (4), LM = 9 ohm21 cm2 mol21 (3) (Found: C, 43.75; H, 3.25; N, 2.6. Calc. for C38H36Au2N2P2S2 3: C, 43.85; H, 3.5; N, 2.7.Found: C, 43.35; H, 3.15; N, 2.4. Calc. for C42H36Au2N2P2S4 4: C, 43.75; H, 3.15; N, 2.4%). Crystallography Compound 2. Crystal data. C92H80Au4ClN4O4P6, M = 2314.74, triclinic, space group P1� , a = 9.278(3), b = 10.689(3), c = 22.192(5) Å, a = 101.00(2), b = 95.91(2), g = 105.11(2)8, U = 2058.5 Å3, Z = 1, Dc = 1.867 Mg m23, F(000) = 1115, l(Mo-Ka) = 0.710 73 Å, m = 7.3 mm21, T = 173 K. Data collection and reduction. Single crystals of compound 2 were obtained by slow diffusion of diethyl ether into a dichloromethane solution.A green prism with dimensions 0.80 × 0.20 × 0.12 mm was mounted on a glass fibre in inert oil and transferred to the cold gas stream of the diffctometer (Siemens R3). A total of 7707 intensities were measured to 2qmax 508, of which 7220 were independent (Rint = 0.027). An absorption correction based on y scans was applied, with transmission factors 0.72–1.00. Cell constants were refined from setting angles of 52 reflections in the range 2q 20–228.Structure solution and refinement. The structure was solved by the heavy-atom method and subjected to full-matrix leastsquares refinement on F 2 (program system SHELXL 93).23 All non-hydrogen atoms were refined anisotropically; hydrogen atoms were included by using a riding model. Refinement proceeded to wR(F 2) 0.0829 for all reflections, conventional R(F) 0.0347 and S(F 2) 1.03 for 503 parameters and 445 restraints (to light atom displacement parameters and local ring symmetry), maximum Dr 1.7 e Å23.Compound 3. Crystal data. C38H36Au2N2P2S2, M = 1040.618, monoclinic, space group P21/n, a = 10.791(3), b = 8.763(3), c = 19.407(4) Å, b = 104.78(2)8, U = 1774.4(9) Å3, Z = 2,2266 J. Chem. Soc., Dalton Trans., 1997, Pages 2263–2266 Dc = 1.948 Mg m23, F(000) = 996, l(Mo-Ka) = 0.710 73, m = 8.5 mm21, T = 173(2) K. Data collection and reduction. Single crystals of compound 3 were obtained by slow diffusion of diethyl ether into a dichloromethane solution.An orange plate with dimensions 0.75 × 0.20 × 0.02 mm was mounted as above (Siemens P4) and used to collect 3025 intensities (3013 independent) to 2qmax 508. Transmission factors 0.42–1.00. Structure solution and refinement. The structure was solved and refined as above: wR(F 2) 0.146, R(F) = 0.057, S(F 2) 1.085 for 208 parameters and 277 restraints, maximum Dr 2.8 e Å23. 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 CCDC for this material should quote the full literature citation and the reference number 186/528. Acknowledgements We thank the Dirección General de Investigación Científica y Técnica (PB94-0079 y PB95-0140) and the Fonds der Chemischen Industrie for financial support. References 1 H. Schmidbaur, Acc. Chem.Res., 1975, 8, 62. 2 J. P. Fackler, jun. and L. C. Porter, J. Am. Chem. Soc., 1986, 108, 2750. 3 M. Bardají, N. G. Connelly, M. C. Gimeno, P. G. Jones, A. Laguna and M. Laguna, J. Chem. Soc., Dalton Trans., 1995, 2245. 4 M. Bardají, P. G. Jones, A. Laguna and M. Laguna, Organometallics, 1995, 14, 1310. 5 J. P. Fackler jun., Polyhedron, 1997, 16, 1. 6 D. D. Heinrich, R. J. Staples and J. P. Fackler, jun., Inorg. Chim. Acta, 1995, 229, 61. 7 H. C. Knachel, C. A. Dettorre, H. J.Galuska, T. A. Salupo, J. P. Fackler, jun. and H. H. Murray, Inorg. Chim. Acta, 1987, 126, 7. 8 M. Bardají, N. G. Connelly, M. C. Gimeno, J. Jiménez, P. G. Jones, A. Laguna and M. Laguna, J. Chem. Soc., Dalton Trans., 1994, 1163. 9 W. Pukaki, M. Pawlack, A. Graja and M. Leqan, Inorg. Chem., 1987, 26, 1328. 10 J. S. Chapel, A. N. Bloch, W. A. Byden, M. Maxfield, T. O. Peolher and D. O. Cowan, J. Am. Chem. Soc., 1981, 103, 2442. 11 E. Cerrada, M. C. Gimeno, A. Laguna, M. Laguna, V. Orera and P. G. Jones, J. Organomet. Chem., 1996, 506, 203. 12 M. N. Gowda, S. B. Naikar and G. K. N. Reddy, Adv. Inorg. Chem. Radiochem., 1984, 28, 255. 13 H. Schmidbaur and R. Franke, Inorg. Chim. Acta, 1975, 13, 85. 14 R. M. Dávila, A. Elduque, R. J. Staples, M. Hartlass and J. P. Fackler, jun., Inorg. Chim. Acta, 1994, 217, 45. 15 E. Cerrada, P. G. Jones, A. Laguna and M. Laguna, Inorg. Chim. Acta, 1996, 249, 163. 16 R. M. Dávila, A. Elduque, J. Grant, R. J. Staples, M. Hartlass and J. P. Fackler, jun., Inorg. Chem., 1993, 32, 1749. 17 A. Hoesta, T. Spoelder and A. Vos, Acta Crystallogr., 1965, 18, 14. 18 M. Bardají, A. Blasco, J. Jiménez, P. G. Jones, A. Laguna, M. Laguna and F. Merchán, Inorg. Chim. Acta, 1994, 223, 55. 19 M. Bardají, M. C. Gimeno, P. G. Jones, A. Laguna, M. Laguna, F. Merchán and I. Romeo, Organometallics, 1997, 16, 1083. 20 J. D. Basil, H. H. Murray, J. P. Fackler, jun., J. Tocher, A. M. Mazany, B. Trzcinska-Bancroft, H. Knachel, D. Dudis, T. J. Delord and D. O. Marler, J. Am. Chem. Soc., 1985, 107, 6908. 21 D. S. Acker and W. R. Hertler, J. Am. Chem. Soc., 1962, 184, 3374. 22 R. Usón, A. Laguna, M. Laguna, J. Jiménez and P. G. Jones, J. Chem. Soc., Dalton Trans., 1991, 1361. 23 G. M. Sheldrick, SHELXL 93, Program for Crystal Structure Refinement, University of Göttingen, 1993. Received 3rd March 1997; Paper 7/01486D

ISSN:1477-9226    

DOI:10.1039/a701486d

出版商:RSC    

年代:1997

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1998, Pages 2267–2273 2267 Nuclear magnetic resonance spectroscopic studies of pyridine methyl derivatives binding to cytochrome c Jun Lu,a,b Dejian Ma,a Jun Hu,a Wenxia Tang *,a and Dexu Zhub a State Key Laboratory of Coordination Chemistry, Nanjing University, Nanjing 210093, P.R. China b State Key Laboratory of Pharmaceutical Biotechnology, Nanjing University, Nanjing 210093, P.R. China The binding of pyridine methyl derivatives (2-, 3- and 4-methylpyridine) to horse heart ferricytochrome c (cyt c) by displacing methionine-80 was studied by 1H NMR spectroscopy to elucidate the eVects of the diVerent methyl substitution positions on the aYnity and kinetics of binding to cytochrome c and the hyperfine-shifted NMR signals of the ligand–cytochrome c complex.Two-dimensional exchange spectroscopy (2D-EXSY) showed that except for 2-methylpyridine (2-mpy) these pyridine derivatives can form stable complexes with cytochrome c.The complexes 3-mpy–cyt c and 4-mpy–cyt c exhibit diVerent hyperfine shift patterns compared to that of py–cyt c. The temperature dependence of the methyl resonances of 3-mpy–cyt c diVers from those of 4-mpy– and py–cyt c. Kinetic and equilibrium data for the binding of 3- and 4-mpy to cyt c have been obtained by 2D-EXSY. Based on these data a comprehensive comparison between the binding properties of these pyridine derivatives and those of pyridine towards cyt c was made.The 1H NMR resonances of 3-mpy–cyt c have also been assigned including the heme peripheral protons and some aliphatic and aromatic side chain protons. Ferricytochrome c (cyt c) can bind a wide range of ligands such as cyanide, azide, imidazole (Him) and pyridine (py) which displace Met-80 and ligate to the heme iron.1–7 Studies of these binding reactions will help obtain insight into the structural aspects of the heme surroundings and the source of the asymmetric electron spin density distribution of cytochrome c.6,7 Binding of pyridine to cytochrome c has been studied by NMR methods.6–9 In the present paper we have performed more comprehensive studies of the binding reactions of pyridine derivatives including 2-methyl- (2-mpy), 3-methyl- (3-mpy) and 4-methyl-pyridine (4-mpy) to cyt c employing 2D-EXSY (exchange spectroscopy).We have been investigating the eVects of diVerent positions of methyl substitution in pyridine on the aYnity and kinetics of binding of pyridine derivatives to cyt c and on the hyperfine-shifted NMR signals for ligand–cytochrome c complexes by comparing the shift patterns and temperature dependences and through the comparisons of kinetic and equilibrium constants obtained by 2D-EXSY.The 1H NMR resonances of 3-mpy–cyt c including those of the heme peripheral protons and some aliphatic and aromatic sidechain protons are further assigned. These protons are sensitive to the electronic state of the heme environment.The present studies thus help reveal the eVect of the diVerent methyl substitution positions in pyridine on the heme environment. Experimental NMR experiments Horse heart cytochrome c(VI) was obtained from Sigma Chemical Corp. and purified as previously described.10 It was dissolved in D2O and incubated at 60 8C for 3 h in order to exchange all the labile protons, and then lyophilized. 3- Methylpyridine was purified according to Heap et al.,11 2- methyl- and 4-methyl-pyridine were reagent grade used without further purification.The purity of these pyridine derivatives was checked by 1H NMR spectroscopy. The pH was adjusted by addition of small amounts of DCl or NaOD, with pH values not corrected for the isotope eVect. All NMR data were recorded on a Bruker Am500 spectrometer with an Aspect 3000 computer. All the data treatments were performed on a Silicon Graphics Indy workstation using the X-WINNMR software of Bruker Corp.Chemical shifts were calibrated with respect to 1,4-dioxane at d 3.743. Onedimensional NMR spectra were obtained using a presaturation pulse for elimination of the residual water resonance. Twodimensional exchange spectra (2D-EXSY) with a mixing time (tm) of 25 or 50 ms were acquired using the phase-sensitive NOESY pulse sequence12 over a 35 714.29 Hz bandwidth. All spectra were collected using 2048(t2) × 512(t1) data points with 160 scans for each t1 increment. After zero filling, which resulted in equal digital resolution in both dimensions, the time domain matrix was multiplied in both dimensions with the shifted sine bell function.The integral values of the two-dimensional peaks were obtained by calculating from the spectra using a square frame and then normalized according to SIij = 1. The same frame was used for estimating the average noise integral value in order to remove the noise eVects from the quantitative two-dimensional integration, and the Iij data were corrected before normalization.The equilibrium magnetization values were obtained by integration of the one-dimensional spectra and also normalized.13 Kinetics For a system involving chemical exchange between two sites it has been shown that the peak amplitude in 2D-EXSY spectra is related to the exchange rate constant k9, the relaxation rate and the mixing time tm by expression (1) 14 where A and R are given A = exp(2Rtm) (1) by the matrices (2) and (3). In A the quantities I11, I12 are two- A = )I11/M1 I21/M1 I12/M2 I22/M2) (2) R = )2R1 2 k91 k91 k921 2R2 2 k921) (3)2268 J.Chem. Soc., Dalton Trans., 1998, Pages 2267–2273 dimensional peak amplitudes measured in an experiment with mixing and normalized, and M1, M2 are the equilibrium magnetization values obtained from integration of the onedimensional spectra and also normalized. The quantity R contains the kinetic parameters to be determined, namely chemical exchange and longitudinal relaxation rates.It can be obtained directly by first diagonalizing A and then calculating the eigenvector matrix X and its inverse X21 so that XDX21 = A where D is the diagonal eigenvalue matrix. The solution to this equation is given15–17 by (4) where ln D = diag (ln li). Thus R can be directly calculated from A. R = 2 ln A tm = 2 X(ln D)X21 tm (4) In this paper, the binding of mpy to cyt c can be represented by reaction (5). The magnetization exchange between the cyt c 1 mpy k1 k21 mpy–cyt c (5) species is a first-order reaction (6).The relationships between Mcyt c k91 k921 Mmpy–cyt c (6) the magnetization exchange rate constants k9 and the reaction exchange rate constants k can be found in (7a) and (7b), and the k1 = k91/[mpy] (7a) k21 = k921 (7b) apparent equilibrium constants Kapp and the equilibrium constant K of the reaction can be calculated from equations (8a) and (8b) where ka is the dissociation constant of mpy.The Kapp = k1/k21 (8a) K = Kapp{1 1 ([H1]/ka)} (8b) thermodynamic values DH8, DS8, and the activation energy of the reaction of cyt c with mpy were obtained according to the Van’t HoV and Arrhenius equations (9) and (10). ln K = 2 DH8 RT 1 DS8 R (9) ln k1 = 2 Ea RT 1 ln A (10) Results and Discussion AYnity of cytochrome c for pyridine methyl derivatives Fig. 1 presents the hyperfine-shifted 1H NMR spectra of cytochrome c at pH 5.7, 303 K in the presence of 2-, 3- and 4-mpy respectively.It is obvious that the presence of 3- and 4-mpy significantly alters the hyperfine-shifted resonances which indicates the formation of stable complexes with cytochrome c. However, there is little deviation in the spectrum of cyt c after the addition of 1 mol dm23 2-mpy. In fact, even when the amount of 2-mpy was increased to 2 mol dm23 the spectrum still showed no critical alteration compared with the spectrum of native cyt c under the same conditions. This strongly suggests that 2-mpy cannot bind to cytochrome c which might be due to the 2-methyl group hindering the approach of 2-mpy to the heme plane.For 3-mpy binding to cyt c at 303 K and pH 5.7 the chemical exchange is a two-site spin system. According to the theory of kinetics studied by means of exchange spectroscopy,15–17 as shown above the reaction amplitude matrix A based on the integration of 8-CH3 is as in equation (11). From this, the A = )0.714 0.495 0.306 0.565) (11) kinetic matrix R is calculated, equation (12).Thus the magnet- R = )222.1 36.2 22.4 233.0) (12) ization exchange rate constants k91 = 36.2 s21 and k921 = 22.4 s21. According to equations (7a) and (7b) the reaction rate constants k1 = 90.5 dm3 mol21 s21 and k21 = 22.4 s21. Then the equilibrium constants Kapp and K are calculated by equations (8) and (9) and listed in Table 1. Employing the same method, the kinetic and equilibrium data for binding of 3- and 4-mpy to cyt c at diVerent temperatures at pH 5.7 were obtained and given in Table 1.From Table 1 it follows that the rate constants for 3- and 4-mpy for the forward and reverse reactions increase with increasing temperature, and so do the equilibrium constants. This is similar to the cases of pyridine binding to cyt c.8 Comparing the equilibrium constants for binding of pyridine derivatives with that of pyridine to cyt c at the same pH and temperature (see Table 2) we get the following relationship: 4-mpy > 3-mpy > py.This is consistent with their pka values: 5.23 (py), 5.68 (3-mpy), 6.03 (4-mpy) respectively.18 The DH8 and DS8 for binding of 3- and 4-mpy to cyt c are Fig. 1 Downfield hyperfine-shifted region of the 1H NMR spectra in D2O at 303 K, pH 5.7 of (a) cyt c with 1 mol dm23 2-mpy, (b) cyt c with 0.8 mol dm23 3-mpy–cyt c, (c) 0.6 mol dm23 4-mpy–cyt c and (d) native cyt c only; e, f, g, h denote 8-, 3-, 5- and 1-CH3 of 3-mpy–cyt c and 4-mpy–cyt c respectively Table 1 Rate and equilibrium constants for binding of 3- and 4-mpy to cyt c at diVerent temperatures 3-mpy 4-mpy T/K 293 296.5 298 300 303 303.5 308 313 k1 36.0 50.5 90.5 225 494 k21 15.8 17.9 22.4 34.8 46.1 Kapp 2.28 2.82 4.04 6.47 10.7 K 4.45 5.52 7.90 12.7 20.9 k1 13.0 24.9 54.7 143 k21 15.6 16.4 17.4 18.0 Kapp 0.828 1.52 3.14 7.93 K 2.56 4.68 9.70 24.5 The errors of the rate constants are in the range 5–10%.Units are: k1/ dm3 mol21 s21, k21/s21, K/dm3 mol21.J.Chem. Soc., Dalton Trans., 1998, Pages 2267–2273 2269 Fig. 2 Plots of ln K (a) and ln k1 (b) versus 1/T in the reaction of 3-mpy with cyt c obtained from Table 1 by least-squares fitting and listed in Table 2. Fig. 2 depicts plots of ln K and ln k1 versus 1/T for the reaction of 3-mpy and cyt c obtained by least-squares fitting. As seen from Table 2, the DH8 and DS8 for binding of py, 3- and 4-mpy to cyt c are both positive. Although DH8 > 0 is unfavourable for the binding reactions, the positive DS8 overweighs the negative eVect of DH8.This suggests that the reactions are driven by a favourable entropy change and the aYnity of pyridine derivatives for cyt c arises from the positive DS8. This is also the case for pyridine binding to cyt c.8 However the values of DS8 for the binding of 3- or 4-mpy to cyt c (272 and 548 J K21 mol21 respectively) are much larger than that for pyridine binding.8,9 This might be related to the disruption of some secondary structure elements of cyt c after the binding of 3- or 4-mpy.Previous studies had shown that the binding of imidazole to cyt c not only induced the movements of some sidechains of cyt c, but also resulted in the disappearance of the 310 helix (67–70), type II turn (75–78) and the distortion of the 50’s helix.19 The corresponding DS8 value for this system was 184 J K21 mol21. Being bulkier than imidazole, 4- or 3-mpy should result in larger DS8 values due to greater steric interactions.With the assistance of the molecular simulation software INSIGHT II20 we constructed a model of imidazole, 4- and Table 2 Thermodynamic parameters of the reactions of cyt c with py, 3- and 4-mpy respectively Reaction py and cyt c a 3-mpy and cyt c b 4-mpy and cyt c b K/dm3 mol21 1.40 c 7.9 c 24.5 d DH8/ kJ mol21 44 77.2 158 DS8/ J K21 mol21 148 272 548 a From ref. 8. b This work. c Obtained at 303 K. d Obtained at 303.5 K. Fig. 3 Portions of the two-dimensional EXSY spectrum of a mixture of 3-mpy–cyt c and cyt c at pH 5.7 and 303 K, mixing time 25 ms.The onedimensional spectrum is shown at the top. Resonances due to 3-mpy–cyt c and cyt c are labelled with B and N respectively. (a)–(c) Portions of downfield region, (d) upfield region2270 J. Chem. Soc., Dalton Trans., 1998, Pages 2267–2273 3-mpy binding to cyt c. It is found that imidazole can fit into the heme pocket more easily than 4- or 3-mpy. Furthermore, when 4-mpy binds to cyt c the 4-methyl group induces severe steric interactions with the peptide near Pro-71 because they are very close to each other.However, in the 3-mpy–cyt c system there is no such phenomenon. The 3-methyl group is located in the heme pocket and not so close to certain residues as that of 4-mpy. As a result, the binding of 4-mpy will induce more secondary structure changes or even gross conformational changes which result in the large DS8 for this system. As 3-mpy will not induce so large structural changes as 4-mpy, the DS8 of the 3-mpy–cyt c should be less than that for 4-mpy–cyt c.Further work on these calculations is in progress. The heme pocket provides a hydrophobic environment for the heme.21 This would be equivalent to a medium of low relative permittivity which results in a negative DH8 for the binding reactions of pyridine derivatives to cyt c.21 On the other hand, conformational changes due to steric interactions of the bound ligand (3- or 4-mpy) with the protein will contribute a positive DH8.22 The contribution to DH8 of the steric eVect of binding of 3- or 4-mpy to cyt c overweighs that of the hydrophobic heme environment, so the total DH8 is positive, reflecting the net eVect of the protein.Being a much bulkier ligand, 3- or 4-mpy would induce greater steric hindrance in the heme pocket and more positive DH8 than for pyridine are expected. The DH8 for 4-mpy binding to cyt c is much larger than that for 3-mpy, indicating that substitution at the 4 position of pyridine will induce a greater steric eVect than that at the 3 position.Hyperfine shift pattern and temperature dependence of cytochrome c complexes with pyridine methyl derivatives The 1H NMR spectra of cyt c, 3-mpy–cyt c and 4-mpy–cyt c are illustrated in Fig. 1. Fig. 3 shows portions of the 2D-EXSY spectrum of a mixture of cyt c and 3-mpy–cyt c at pH 5.7, 303 K. The exchange between the groups of the native species and of the substituted form can be visualized from the cross-peaks. In Fig. 3(a) and 3(b) four signals at d 29.37, 19.09, 18.38 and 12.79 show correlation with signals at d 33.87, 31.53, 10.34 and 7.21, respectively. The chemical shifts of the latter four signals belong to heme 8-, 3-, 5- and 1-methyl groups of cyt c.23,24 So the observation of the above cross-peaks unequivocally identi- fies the signals at d 29.27, 19.05, 18.38 and 12.84 as due to the 8-, 3-, 5- and 1-methyl groups of 3-mpy–cyt c.Employing the same method, these heme methyls of 4-mpy–cyt c are assigned at d 27.22, 19.89, 16.93 and 11.86 respectively at 303 K, pH 5.7 (figure not shown). Native cyt c has a pairwise pattern for the heme methyl shifts.25 The two most-shifted methyl resonances are 8- and 3-CH3. The shift values for the other two heme methyl signals are d 10.34 (5-CH3) and 7.21 (1-CH3).23,25 Upon binding of 3- or 4-mpy to cyt c, Met-80 is detached from the iron and the heme methyl shift pattern changes significantly. The heme methyl resonances of 3-mpy–cyt c and 4-mpy–cyt c have a spread of only 16.43 and 15.36 ppm respectively, compared to a spread of 26.66 ppm for cyt c.The most-shifted pair of the heme methyls in cyt c shifts upfield for 3-mpy–cyt c and 4-mpy–cyt c while the other pair of heme methyls shifts downfield. This results in the disappearance of the pairwise pattern which indicates that these complexes have higher symmetry for the electron spin distribution relative to cyt c.In comparison with py–cyt c, there is some diVerence between the chemical shift pattern of the heme methyls of 3- or 4-mpy–cyt c and that of py–cyt c.8,9 The major diVerence is seen for 3- and 5-CH3. The complex py–cyt c shows the hyperfine-shifted pattern 8 > 5 > 3 > 1, while for 3-mpy–cyt c and 4-mpy–cyt c the patterns are 8 > 3 > 5 >1. Such shift patterns primarily reflect a redistribution of the delocalized spin density among the four pyrroles.In cyt c the orientation of the axial Met is believed to determine the x and y magnetic axes.25,26 It is also indicated that the histidine orientation has a considerable influence on the asymmetry of the heme electronic structure.27 In the cytochrome c complex with py, 3- and 4-mpy it is the orientation of the axial histidine (His-18) and the corresponding bound ligand that determines the magnetic axes. As a 3- or 4-methyl group has a stronger steric interaction with the surrounding polypeptide near the heme, 3- or 4-mpy might adopt a somewhat diVerent orientation from that of pyridine in py–cyt c that aVects the hyperfine shift pattern.The hyperfine shift patterns of 3-mpy–cyt c and 4-mpy–cyt c are quite similar. Their diVerence also originates from the heme 3- and 5-CH3. At 303 K and pH 5.7 the chemical shift diVerence between 3- and 5-CH3 in 3-mpy–cyt c is 0.67 ppm while the value in 4-mpy–cyt c is enlarged to 2.96 ppm.This phenomenon reflects the diVerent electronic states of heme induced by pyridine substituted with methyl at diVerent positions. The temperature dependences of the heme methyl resonances for 3-mpy–cyt c and 4-mpy–cyt c are illustrated in Fig. 4 (Curie plots), from which it can be concluded that all the four heme methyls in 3-mpy–cyt c exhibit no deviation from Curie’s law while in the case of 4-mpy–cyt only 8- and 3-CH3 obey Curie’s law, the other methyls (5- and 1-CH3) exhibiting anti-Curie behaviour.The behaviour of the heme methyls of 4-mpy resembles that of native cyt c. It has been reported that the shifts of heme methyls 1 and 5 of native cyt c have an anti-Curie eVect, which increases with increasing temperature, while 3 and 8 exhibit normal Curie behaviour.28 The temperature depend- Fig. 4 Curie plots of 3-mpy–cyt c (a) and 4-mpy–cyt c (b) at pH 5.7J. Chem. Soc., Dalton Trans., 1998, Pages 2267–2273 2271 Fig. 5 Aliphatic region of the two-dimensional EXSY spectrum of a mixture of 3-mpy–cyt c and cyt c at pH 5.7 and 303 K, mixing time 50 ms. Resonances due to 3-mpy–cyt c and cyt c are labelled with B and N respectively Fig. 6 Aromatic region of the two-dimensional EXSY spectrum of a mixture of 3-mpy–cyt c and cyt c at pH 5.7 and 303 K, mixing time 50 ms. Resonances due to 3-mpy–cyt c and cyt c are labelled with B and N respectively ence of the hyperfine shifts is related to the energy separation between the ground and the excited levels, which, in turn, is modulated by interactions between the iron and the axial ligands.25 The anti-Curie eVect has been explained by a Boltzmann distribution between partially filled porphyrin 3e(p) molecular orbitals with an energy diVerence of 3 kJ mol21.292272 J.Chem. Soc., Dalton Trans., 1998, Pages 2267–2273 The diVerent temperature dependence of 4-mpy–cyt c from that of 3-mpy–cyt c also implies that the methyl substitution position in pyridine influences the behaviour of the mpy–cyt c complex.Assignments of some resonances of 3-mpy–cyt c Since the chemical shifts of many protons are expected to change in 3-mpy–cyt c and the native form of cyt c, chemical exchange correlation can be used to generate cross-peaks in a two-dimensional spectrum. As the full assignments of cyt c have been reported, exchange eVects can be used to determine and establish the reliability of the proton assignments for 3-mpy–cyt c.Fig. 3 presents the hyperfine shift region of the 2D-EXSY spectrum recorded at pH 5.7, 303 K with a mixing time of 25 ms. The propionic acid 7a(pro-7a) and 7a9 protons of native cyt c are found at d 18.11 and 11.45.23,30 In Fig. 3(c) the pro 7a-H shows an exchange cross-peak at d 5.85 which can be assigned to pro-7a of 3-mpy–cyt c. Similarly, pro-7a9H of cyt c gives an exchange cross-peak at d 7.65 which can be identified as due to pro-7a9 of 3-mpy–cyt c [Fig. 3(b)]. One of the Met-80 b-H of cyt c had been assigned to the peak at d 11.74,31 it shows an exchange peak at d 2.09, so the signal at d 2.09 can be assigned as Met-80 b-H of 3-mpy–cyt c [Fig. 3(b)]. The Met-80 e-CH3 of cyt c is correlated with the signal at d 1.98 31 [Fig. 3(d )]. Thus the latter resonance can be assigned to Met-80 e-CH3 of 3-mpy–cyt c. The above data on the exchange peaks of Met-80 indicate that in 3-mpy–cyt c the co-ordination bond Fe]S between Met-80 and heme is broken, and the side chain of Met-80 (e-CH3 and b-CH2) has left its original position in the heme pocket and is situated in the Table 3 Assigned resonances of 3-mpy–cyt c at 303 K and pH 5.7 d Assignment Heme 8-CH3 3-CH3 5-CH3 1-CH3 pro-7a pro-7a9 Met-80 e-CH3 Met-80 b-H Leu-68 d2 Leu-68 d1 Thioether-2 Leu-64 d2 Leu-64 d1 Leu-98 d2 Leu-98 d1 Leu-35 d2 Leu-35 d1 Ile-57 d Leu-94 d2 Leu-94 d1 Ile-85 d Ile-81 d Ile-81 g Ile-95 g Thr-47 g Thr-19 a Thr-19 b Thr-78 a Phe-82 p Phe-82 m Phe-82 o Tyr-97 C2 Tyr-97 C3 Trp-59 C5 Trp-59 C6 Trp-59 C7 Phe-36 p Phe-36 o 3-mpy–cyt c 29.27 19.05 18.38 12.84 5.85 7.65 1.98 2.09 21.39 20.58 22.10 20.32 20.61 20.55 20.01 20.42 0.03 0.38 0.03 0.39 0.30 0.05 0.13 0.36 0.99 5.52 5.86 5.83 6.65 6.36 7.13 6.57 6.42 6.46 6.78 7.03 6.96 7.11 cyt c 33.88 31.54 10.34 7.21 18.11 11.45 223.05 11.74 22.43 20.44 20.56 20.45 20.14 20.02 0.49 0.20 0.51 20.25 20.11 0.66 20.06 0.97 1.12 0.85 0.79 6.20 5.59 5.98 6.00 6.97 6.17 6.40 5.84 6.65 6.45 7.45 6.77 7.28 region where the paramagnetic contribution to the resonances from heme iron is minor.This movement causes the chemical shifts of the side chain of Met-80 to be close to those of the side chain of methionine in the random coil. In the diamagnetic region the 2D-EXSY experiment shows the major advantage of its greater selectivity over onedimensional double resonance experiments. This is illustrated in Fig. 5 which shows the aliphatic region of the 2D-EXSY spectrum where the cross-peaks arising from chemical exchange between methyl peaks of Leu (L), Ile (I) and Thr (T) are indicated.With the known assignments for cyt c 9,23,24,31,32 the corresponding resonances of 3-mpy–cyt c can be unambiguously assigned from these connectivities. For example, Leu-68 d2-CH3 of cyt c is known to resonate at d 22.43. In Fig. 5 the exchange cross-peak (d 21.39, 22.43) indicates the connectivity between the Leu-68 d2-CH3 resonance of cyt c and the corresponding resonance of 3-mpy–cyt c at d 21.39.Therefore the latter resonance can be assigned to the Leu-68 d2-CH3 protons of 3-mpy–cyt c. This approach accounts for 19 methyl resonances which have been assigned to the specific groups of 3-mpy–cyt c. The aromatic region of the two-dimensional EXSY spectrum of an equimolar mixture of 3-mpy–cyt c is shown in Fig. 6. The signal at d 6.17 of cyt c corresponds to the resonance at d 7.13 of 3-mpy–cyt c. This finding along with the previous assignment for cyt c 9,24,31,32 identifies the latter signal as due to Phe-82 o-H of 3-mpy–cyt c.Similarly, cross-peaks which are attributed from chemical exchange allow one to correlate resonances of the ligand-bound and the native species and to identify some aromatic ring protons of 3-mpy–cyt c. The assignments of heme-related protons are listed in Table 3. On the basis of the assignments of 3-mpy–cyt c, extensive resonance assignments can be made quickly from other twodimensional spectra.Comparing the chemical shifts of 3-mpy–cyt c with those of cyt c, some of the diVerences are near or larger than 1 ppm when passing from the native to the ligand-bound species such as Leu-68 d2, Phe-82 o, Ile-81 g, etc. This can be attributed to perturbations in electronic and molecular structure. In comparison with the 2D-EXSY spectrum of py–cyt c, the two spectra are quite similar although there exist some diVerences at Leu-64, Phe-36, etc.which indicate that 3-mpy–cyt c has a similar ligand-dependent conformation to that of py–cyt c, although its methyl group induces some diVerences. References 1 P. Georgs, S. C. Glauser and A. Schjter, J. Biol. Chem., 1967, 242, 1690. 2 N. Sutin and J. K. Yandell, J. Biol. Chem., 1972, 247, 6932. 3 P. D. Baker and A. G. Mauk, J. Am. Chem. Soc., 1992, 114, 3619. 4 W. Shao, H. Sun, Y. Yao and W. Tang, Inorg. Chem., 1995, 34, 680. 5 X. Hong and D. W. Dixon, FEBS Lett., 1989, 246, 105. 6 M. Smith and G. McLendon, J. Am. Chem. Soc., 1981, 103, 4912. 7 I. Morishima, S. Ogawa, T. Yonezawa and T. Iizuka, Biochim. Biophys. Acta, 1978, 532, 48. 8 W. Shao, Y. Yao, G. Liu and W. Tang, Inorg. Chem., 1993, 32, 6112. 9 G. Liu, Y. Chen and W. Tang, J. Chem. Soc., Dalton Trans., 1997, 795. 10 D. L. Bautigan, S. Ferguson-miller and E. Margolish, Methods Enzymol., 1978, 53D, 128. 11 J. G. Heap, W. J. Jones and J. B. Speakman, J. Am. Chem. Soc., 1921, 43, 1936. 12 G.Bodenhausen, H. Kogler and R. R. Ernst, J. Magn. Reson., 1984, 58, 370. 13 W. Shao and W. Tang, Spectrosc. Lett., 1994, 27, 763. 14 R. R. Ernst, G. Bodenhausen and A. Wokau, Principles of Nuclear Magnetic Resonance in One and Two Dimensions, Oxford University Press, 1983. 15 E. R. Johnston, M. J. Dellwo and J. Hendrix, J. Magn. Reson., 1986, 66, 399. 16 E. W. Abel, T. P. J. Coston, K. G. Orrell, V. Sik and D. Stephenson, J. Magn. Reson., 1986, 70, 34.J. Chem. Soc., Dalton Trans., 1998, Pages 2267–2273 2273 17 J. Lu, G. Liu, C. Yu, D. Zhu and W. Tang, Inorg. Chim. Acta, in the press. 18 Critical Stability Constants, R. M. Smith and A. E. Martell, Plenum Press, London, 1975, vol. 2, pp. 165–169. 19 G. Liu, W. Shao, X. Huang, H. Wu and W. Tang, Biochim. Biophys. Acta, 1996, 1277, 61. 20 INSIGHT II, version 97.0, MSI Corp., San Diego, CA, 1997. 21 D. C. Blumenthal and R. J. Kassner, J. Biol. Chem., 1980, 255, 5859. 22 M. M. M. Saleem and M. T. Wilson, Inorg. Chim. Acta, 1988, 153, 105. 23 H. Santos and D. L. Turner, FEBS Lett., 1986, 194, 73. 24 G. Williams, G. R. Moore, R. Porteus, M. N. Robinson, N. SoVe and R. J. P. Williams, J. Mol. Biol., 1985, 183, 409. 25 K. L. Bren, H. B. Gray, L. Banci, I. Bertini and P. Turano, J. Am. Chem. Soc., 1995, 117, 8067. 26 H. Senn, H. Bo�hme and K. Wüthrich, Biochim. Biophys. Acta, 1984, 789, 311. 27 D. L. Turner, Eur. J. Biochem., 1995, 227, 829. 28 H. Senn and K. Wüthrich, Biochim. Biophys. Acta, 1983, 743, 69. 29 D. L. Turner, Eur. J. Biochem., 1993, 211, 563. 30 A. J. Wand, D. L. Di Stefano, Y. Feng, H. Roder and S. W. Englander, Biochemistry, 1989, 28, 186. 31 Y. Feng, H. Roder, S. W. Englander, A. J. Ward and D. L. Di Stefano, Biochemistry, 1989, 28, 195. 32 J. D. Satterlee and S. Moensch, Biophys. J., 1987, 52, 107. Received 27th January 1998; Paper 8/00731D © Copyright 1998 by the Royal Society of Chemist

ISSN:1477-9226    

DOI:10.1039/a800731d

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

DALTON J. Chem. Soc., Dalton Trans., 1997, Pages 2273–2277 2273 Predominance of electron-withdrawing eVect over angular strain in the metal-promoted hydrolysis of 2,4,6-tris(2-pyridyl)-1,3,5-triazine Parimal Paul,* Beena Tyagi, Mohan M. Bhadbhade and Eringathodi Suresh Discipline of Coordination Chemistry & Homogeneous Catalysis, Central Salt & Marine Chemicals Research Institute, Bhavnagar—364 002, India The reaction of 2,4,6-tris(2-pyridyl)-1,3,5-triazine (tptz) and RhCl3?3H2O in refluxing ethanol–water (1 : 1) resulted in the hydrolysis of tptz to bis(2-pyridylcarbonyl)amide anion (bpca) and afforded a complex of composition [Rh(bpca)2][PF6] 1.However, hydrolysis of tptz did not occur when it was treated with RuCl3?3H2O under similar condition, yielding instead the complex [Ru(tptz)2][PF6]2?H2O 2. The molecular structures of 1 and 2 have been established by single-crystal X-ray analysis. In complex 1 the bis(2-pyridylcarbonyl)amido moiety functions as a tridentate ligand with nitrogen donor atoms and is bound to RhIII in a mutually perpendicular fashion forming a distorted-octahedral geometry around the metal ion.In complex 2 two tridentate tptz are co-ordinated to RuII in a similar manner as found for 1. From a comparison of bond lengths and angles in the co-ordination spheres of 1 and 2 it is suggested that the electron-withdrawing effect (LÆM) of the metal ion is the predominant factor, rather than angular strain at the carbonyl carbon atoms, responsible for hydrolysis of tptz.Electrochemical studies of 1 revealed a metal-based two-electron couple (RhIII]RhI) at 21.15 V and two ligand-based redox couples at 21.44 and 21.84 V. In the case of 2 the RuII]RuIII couple appears at 11.77 V and the ligand-based reduction at 20.62 and 20.80 V which are significantly positively shifted compared to those of free tptz. 2,4,6-Tris(2-pyridyl)-1,3,5-triazine (tptz) is of current interest because of its use as a spacer for designing multinuclear metal complexes.1–5 The compounds of this family, 2,4,6- triaryltriazines, are usually stable towards hydrolysis, concentrated mineral acid and temperatures above 150 8C are required for their hydrolytic reaction.6 In fact tptz has been used as an analytical reagent for various metal ions.7 A number of transition-metal and lanthanide complexes of it have also been reported.8–20 Lerner and Lippard21,22 found that tptz and a similar compound 2,4,6-tris(pyrimidin-2-yl)triazine (tpmtz) undergo hydrolytic reaction in the presence of CuII in aqueous media.Crystal structures of copper(II) complexes of the hydrolytic products of tpmtz and tptz have also been reported.21–24 On the basis of the Cu]N bond distances and angles at the carbonyl carbon atoms within the chelate ring it was suggested that co-ordination of tptz induces an angular strain allowing nucleophilic attack at the carbon atoms of the triazine ring by the solvent resulting in the hydrolysis of tptz.21,24 In the case of ruthenium there is no report of metal-promoted hydrolysis of tptz 1–5,19,20 and for mononuclear complexes, obtained by the reaction of K2[RuCl5] or RuCl3 with tptz, the electrochemistry of [Ru(tptz)2][ClO4]3,19 photophysical studies of [Ru(tptz)2]- [ClO4]2?3H2O20 and 1H NMR data for [Ru(tptz)2][PF6]2 2 have been reported.We were interested to study the reactions of tptz with hydrated RhCl3 and RuCl3 under similar conditions.In the case of RhIII, hydrolysis of tptz to bis(2-pyridylcarbonyl)amide anion (bpca) occurred under the reaction condition and for ruthenium a complex of RuII with intact tptz as found by earlier workers 2 was obtained. In this paper we report the crystal structures of both complexes together with electrochemical and spectroscopic data. To our knowledge this is the first report of a crystallographic characterisation of complexes where two hydrolysed tptz and two intact tptz are bound to metal ions.From the structural data we wish to emphasise mechanistic aspects of the hydrolytic reaction, particularly the role of the metal ion. Experimental Materials The compound tptz, ammonium hexafluorophosphate and tetrabutylammonium tetrafluoroborate were obtained from Aldrich, hydrated rhodium trichloride and ruthenium trichloride from Arora Matthey. Physical measurements Elemental analyses (C, H and N) were performed on a model 2400 Perkin-Elmer Elemental Analyser. Infrared spectra were recorded on a Bio-Rad FTS-40 spectrometer as KBr pellets, UV/VIS spectra on a model 8452A Hewlett-Packard Diode- Array spectrophotometer. Electrochemical experiments were performed with a model 273A EG & G Princeton Applied Research potentiostat.All experiments were conducted in an argon atmosphere with a platinum working electrode, an Ag– AgCl electrode as reference and 0.1 mol dm23 [NBun 4][BF4] as supporting electrolyte.Synthesis of complexes [Rh(bpca)2][PF6] 1. A mixture of tptz (312 mg, 1 mmol) and RhCl3?3H2O (132 mg, 0.5 mmol) in ethanol–water (1 : 1, 50 cm3) was refluxed for 30 h. The volume of the reaction mixture was reduced to ca. 25 cm3 by rotary evaporation and an aqueous solution (5 cm3) of KPF6 (368 mg, 2 mmol) was added. The precipitate was filtered off and washed with water and diethyl ether. Recrystallisation of complex 1 from boiling acetonitrile afforded pale yellow crystals. Yield: 72% (Found: C, 41.1; H, 2.3; N, 12.0.Calc. for C24H16F6N6O4PRh: C, 41.15; H, 2.3; N, 12.0%). [Ru(tptz)2][PF6]2?H2O 2. This complex was synthesized following the procedure of Thummel and co-workers.2 During purification on an alumina column eluting with acetonitrile– toluene (1 : 1) a small purple fraction was eluted first, then the2274 J. Chem. Soc., Dalton Trans., 1997, Pages 2273–2277 desired complex separated as a dark red band. Yield: 75% (Found: C, 41.9; H, 2.4; N, 16.2.Calc. for C36H26F12N12OP2Ru: C, 41.85; H, 2.55; N, 16.35%). Crystallography Preliminary data on the space group and unit-cell dimensions as well as intensity data were collected on an Enraf-Nonius CAD-4 X-ray diffractometer using graphite-monochromatised Mo-Ka radiation (l = 0.7107 Å) in the range q 2–238. Accurate cell dimensions were obtained using 25 reflections in the range q 8–128. The crystal orientation, refinement of cell parameters and intensity measurements were carried out using the program CAD-4 PC.25 Intensities were corrected for Lorentzpolarisation effects but not for absorption. The Lorentzpolarisation correction and data reduction were carried out using the NRCVAX program.26 The structures were solved by the heavy-atom method using the program SHELXS 86.27 Crystallographic data for complexes 1 and 2 are summarised in Table 1.For complex 1 the extensive disorder of the PF6 2 was obvious at the structure-solution stage itself.After locating all the nonhydrogen atoms of the cation in the Fourier-difference map, the peak for phosphorus appeared with the expected height of ª13 e Å23, whereas a large number of peaks appeared with heights ranging from 3 to 1.5 e Å23 joining the phosphorus atom and corresponding to F atoms. Isotropic full-matrix refinement of all the non-hydrogen atoms except F was carried out till convergence using the program SHELXL 93.28 Attempts were made to fit the orientational disorder of PF6 2 by analysing the Fourier-difference map.However, it was not possible to resolve the orientation of each PF6 2 in the lattice due to the presence of a large number of peaks with heights of ª2 e Å23 in the Fourier difference map. Therefore, all the peaks with heights >2 e Å23 were included at every stage only for their structure-factor contribution in the least-squares refinement, till the Fourierdifference map contained no significant peaks corresponding to F atoms and the peaks corresponding to H atoms started appearing.At this stage 24 peaks corresponding roughly to four major orientations of the PF6 2 with occupancy 0.25 were allowed to refine only for their thermal but not their positional parameters, as the peaks would merge into each other. The rest of the non-hydrogen atoms were refined anisotropically using the weighting scheme w = 1/s(|F|2)2 in both 1 and 2. Despite the extensive disorder of the PF6 2, the Fourier-difference map at this stage contained almost all the hydrogen atoms which could even be refined isotropically (except one which was refined using a riding model).The inclusion of these 24 disordered F atoms significantly improved the refinement parameters such as R, wR and goodness of fit. This fact can be taken as correct treatment of the anion disorder, although a constrained or rigid model refinement was not possible in this case. The structure solution and the least-squares refinement of compound 2 proceeded smoothly, with no disorder of PF6 2.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 CCDC for this material should quote the full literature citation and the reference number 186/532. Results and Discussion The reaction of rhodium trichloride and tptz in refluxing ethanol–water resulted in the hydrolysis of tptz to bis(2- pyridylcarbonyl)amide (bpca) anion and afforded a complex of composition [Rh(bpca)2][PF6] 1 with good yield.However, the hydrolysis of tptz did not occur when ruthenium trichloride reacted with tptz under similar conditions, and this reaction yielded a complex, [Ru(tptz)2][PF6]2?H2O 2, similar to that obtained by Thummel and co-workers,2 [Ru(tptz)2][PF6]2. The difference is only the water of crystallisation.The other complexes of RuII and RuIII, [Ru(tptz)2][ClO4]2?3H2O20 and [Ru(tptz)][ClO4]3,19 respectively were prepared by different methods. The ligand and complexes 1 and 2 are shown in Scheme 1. The IR spectrum of complex 1 exhibits a strong band at 1720 cm21 whereas 2 and free tptz do not show any band in that region. This band is assigned to n(C]] O) of the co-ordinated bpca chelate formed by the hydrolysis of tptz.21–24 Complexes 1 and 2 exhibit strong bands at 842 and 840 cm21 respectively, which can be assigned to n(PF6 2).Their electronic spectra in dimethylformamide (dmf) are shown in Fig. 1. Complex 1 exhibits absorption bands at 328 (9800) and 272 nm (e 20 900 dm3 mol21 cm21). The band of 2 in the visible region in dmf (504 nm; 22 300 dm3 mol21 cm21) compares well with the reported value of 501 nm for [Ru(tptz)2][ClO4]2 in water.20 In the UV region complex 2 shows a band at 280 nm (e 59 100 dm3 mol21 cm21). The high-energy bands at 272 and 280 nm for 1 and 2, respectively, are ligand centred and due to p–p* transition. 4,5 The band at 504 nm of 2 is due to a metal-to-ligand Scheme 1 Compound 2 was also reported earlier 2,20 N N N Rh N N N O O O O N N N Ru N N N N N N N N N N + 2+ N N N N N + MCl3 • 3H2O ethanol-water M = Rh M = Ru 1 2 tptz Fig. 1 Electronic absorption spectra of complexes 1 (a) and 2 (b) in dmfJ. Chem. Soc., Dalton Trans., 1997, Pages 2273–2277 2275 charge-transfer (MLCT) transition.20,29 That of 1 at 328 nm may be MLCT in character 30,31 but the possibility of s bond-toligand charge-transfer (SBLCT) cannot be ruled out.32,33 The electrochemical behaviour of the complexes are discussed in a later section. Crystal structures Complex 1.A perspective (ORTEP34) view of the complex cation along with the atom numbering is shown in Fig. 2. Selected bond distances and angles are presented in Table 2. The geometry of the rhodium(III) ion can be described as a distorted octahedron formed by the co-ordination of two bpca which functions as a tridentate ligand with nitrogen donor atoms.The complex cation has two perpendicular approximate planes of symmetry each containing the plane of one ligand and cutting the other through the Rh]N (amido) bond. The four nitrogen atoms N(2), N(4), N(5) and N(6) which comprise the equatorial co-ordination and the Rh show a high degree of Fig. 2 An ORTEP (50% probability) view with atom labelling scheme for the cation of complex 1; hydrogen atoms are omitted for clarity Table 1 Summary of crystallographic data and parameters for complexes 1 and 2 1 2 Formula M Crystal system Space group Crystal dimensions/mm a/Å b/Å c/Å a/8 b/8 g/8 U/Å3 Z Dc/g cm23 F(000) Total reflections Observed reflections [I > 2s(I)] Parameters refined Final (R1) (on F) Final (wR2) (on F2) C24H16F6N6O4PRh 700.31 Triclinic P1� 0.24 × 0.08 × 0.04 8.890(4) 12.053(3) 14.013(2) 109.48(1) 102.11(3) 98.12(3) 1347.2(7) 2 1.762 696 3537 2653 410 0.060 0.155 C36H26F12N12OP2Ru 1031.68 Orthorhombic Pbcm 0.04 × 0.14 × 0.06 11.043(2) 16.747(2) 21.751(2) 4022.7(2) 4 1.704 2056 2638 1923 344 0.031 0.094 R1 = S Fo| 2 |Fc /S|Fo|; wR2 = [Sw(Fo 2 2 Fc 2)2/Sw(Fo 2)2]� �� .planarity (maximum deviation 0.003 Å). The axial nitrogen atoms [N(1) and N(3)] and Rh make an angle of 162.78 which deviates significantly from the ideal value of 1808 as they are constrained to be part of the five-membered chelate rings.For the same reason, two angles at the equatorial base N(4)]Rh]N(5) and N(5)]Rh]N(6) having values 81.4(3) and 81.6(3)8, respectively, are significantly smaller than the ideal value of 908. The Rh]N bond distances, a shorter Rh]N (amido) and longer Rh]N (pyridyl) bonds (Table 2), are comparable to the values for copper(II) complexes with the same ligand.21–24 Though both the ligands are co-ordinated in a similar fashion there is a significant difference in their conformations. This can be seen by dividing each ligand into two halves.The two halves consisting of N(4), C(13)–C(18), N(5), O(3) and N(5), C(19)–C(24), N(6), O(4) of the bpca which comprises the equatorial base are quite planar (maximum deviation 0.03 Å) and exhibit an angle of 3.438 to each other. However, the two halves of the other ligand, N(1), C(1)–C(6), N(2), O(1) (maximum deviation 0.11 Å) and N(2), C(7)–C(12), N(3), O(2) (maximum deviation 0.10 Å) make a dihedral angle of 11.88.The PF6 2 in the molecule is severely disordered occupying four positions as discussed in the Experimental section. The molecular packing shows the C]H? ? ? O interactions listed in Table 3. Complex 2. An ORTEP view of the complex cation along with the atom numbering is shown in Fig. 3. The complex molecule occupies a special position (crystallographic two-fold axis) in the unit cell and therefore only half of it and the three symmetry-related co-ordinated nitrogen atoms, N(1A), N(2A) and N(3A), are labelled.The two PF6 anions also occupy the special positions in the unit cell. Selected bond distances and angles are presented in Table 2. The co-ordination of two tridentate tptz to the ruthenium(II) ion in a mutually perpendicular fashion provides a distortedoctahedral geometry. Three co-ordinated nitrogen atoms of one tptz and one [N(2A)] of the other ligand form the equatorial base of the octahedron; N(1) and N(3) of the latter are in axial positions.The four nitrogen atoms of the equatorial base and RuII show excellent planarity (maximum deviation 0.012 Å). The axial nitrogen atoms and RuII make an angle of 155.098 which is slightly smaller than that of complex 1. The two pyridyl rings and triazine moiety which are involved in co- Table 2 Selected bond distances (Å) and angles (8) for complexes 1 and 2 Complex 1 Rh]N(1) Rh]N(2) Rh]N(3) Rh]N(4) Rh]N(5) 2.015(7) 1.996(6) 2.024(7) 2.048(6) 1.993(6) Rh]N(6) C(6)]O(1) C(7)]O(2) C(18)]O(3) C(19)]O(4) 2.020(7) 1.215(9) 1.230(9) 1.197(10) 1.205(10) N(2)]Rh]N(4) N(4)]Rh]N(5) N(5)]Rh]N(6) N(6)]Rh]N(2) N(1)]Rh]N(3) N(2)]Rh]N(5) N(4)]Rh]N(6) N(5)]C(6)]C(5) 97.1(3) 81.4(3) 81.6(3) 100.0(3) 162.7(2) 178.4(3) 162.9(3) 111.8(6) N(2)]C(7)]C(8) N(5)]C(18)]C(17) N(5)]C(19)]C(20) N(2)]C(6)]O(1) N(2)]C(7)]O(2) N(5)]C(18)]O(3) N(5)]C(19)]O(4) 112.1(7) 111.1(7) 111.4(7) 128.0(7) 128.5(8) 128.7(9) 126.7(8) Complex 2 Ru]N(1) Ru]N(2) 2.099(3) 1.973(3) Ru]N(3) 2.100(3) N(2)]Ru]N(1A) N(1A)]Ru]N(2A) N(2A)]Ru]N(3A) N(3A)]Ru]N(2) N(1)]Ru]N(3) N(2)]Ru]N(2A) 103.47(11) 77.65(11) 77.44(11) 101.44(12) 155.09(11) 178.4(2) N(1A)]Ru]N(3A) N(2)]C(6)]C(5) N(2)]C(7)]C(8) )]N(4) N(2)]C(7)]N(5) 155.09(11) 112.6(2) 112.4(3) 122.5(3) 123.3(3)2276 J.Chem. Soc., Dalton Trans., 1997, Pages 2273–2277 Table 3 Hydrogen-bonding parameters in complexes 1 and 2 Complex Donor (D) Hydrogen Acceptor (A) D]A/Å H? ? ? A/Å D]H? ? ? A/8 Symmetry code 1 C(1) C(1) C(24) C(24) H(1) H(1) H(24) H(24) O(3) O(4) O(1) O(2) 2.97(1) 3.26(1) 3.39(1) 3.12(7) 2.37(8) 2.68(8) 2.57(8) 2.29(6) 129(7) 127(7) 138(6) 140(6) 1 2 x, 2y, 2z 1 2 x, 2y, 2z 1 2 x, 1 2 y, 1 2 z 1 2 x, 1 2 y, 1 2 z 2 O(1w) O(1w) C(9) C(1) C(2) C(12) H(10w) H(10w9) H(9) H(1) H(2) H(12) N(5) N(6) O(1w) F(5) F(4) F(1) 3.065(3) 2.947(5) 3.197(6) 3.109(5) 3.120(5) 3.319(5) 2.38(5) 2.16(5) 2.17(4) 2.73(4) 2.59(4) 2.46(4) 136(4) 150(4) 165(3) 111(3) 123(3) 164(4) x, y, z x, y, z x, y, z x, y, z 1 2 x, 2y, 1 2 z x, y, z ordination comprise a plane (maximum deviation 0.08 Å) but the free pyridyl ring is not in the same plane and is rotated by 6.48 about the C(13)]C(14) bond.The angles at the metal formed by the nitrogen atoms of the equatorial co-ordination sites are quite similar to those found around RhIII in 1. The Ru]N(2)/(2A) bond distances are comparable to the Rh]N (amido) distances but the Ru]N (pyridyl) bond lengths are slightly longer than the corresponding Rh]N (pyridyl) distances (Table 2).Unlike in complex 1, the PF6 anions do not show any disorder. The packing shows channels for the PF6 anion and a water molecule. There are hydrogen-bonding interactions with N(5), N(6) and C(9) with the oxygen atom O(1w) of the water molecule. The PF6 anion makes short contacts with C(1), C(2) and C(12) as shown in Table 3. Electrochemistry The Osteryoung square-wave and cyclic voltammograms of complex 1 in dmf show three reductions at 21.15 (DE = 92), 21.44 (87) and 21.84 V; the last peak is poorly resolved in the cyclic voltammogram but well resolved in square wave.The first reduction is metal-centred and is in fact a composite wave corresponding to two electrons and assigned to the RhIII]RhI couple.35–38 The resolution of the composite wave into two components at high scan rate (>10 V s21) as found for some complexes 35,36 could not be observed when the scan rate was increased to 20 V s21.The other two reductions are ligand based. It is interesting that in the cyclic voltammogram at low scan rate the anodic waves are small compared to the cathodic waves but they increased with increasing sweep rate. When the potential scan was reversed just after the first cathodic peak the first redox couple was almost reversible even at low scan rate (100 mV s21). These observations are consistent with no chemical change after metal reduction, but with a chemical change after ligand reductions.At high scan rate probably the electrochemical process becomes faster compared to the chemical change and thus the reversibility increases with increasing scan rate. The electrochemical study of complex 2 was carried out in acetonitrile because in dmf at positive potential a metal-based Fig. 3 An ORTEP view (50% probability) with the atom labelling scheme for the cation of complex 2; hydrogen atoms are omitted for clarity redox couple is obscured by the edge of the solvent window.The acetonitrile solution of 2 shows two reversible ligand-based redox couples at 20.62 (DE = 64) and 20.80 V (65 mV).4,5 Free tptz under the same conditions exhibits a reversible redox couple at 21.41 V, indicating a significant anodic shift of the reduction potential for both tptz co-ordinated to RuII. A quasireversible wave found at 11.77 V (DE = 120 mV) is assigned to the RuII]RuIII couple.4,5 Mechanistic aspects of hydrolytic reaction In copper(II) complexes with hydrolysed tptz and tpmtz the values of the angles at the carbonyl carbon atoms within the chelate ring, Aa (structure I of Scheme 2), are in the range 110.0–111.78,21–24 which is compressed considerably from the ideal value of 1208, and on the basis of these data it was suggested that the metal-induced angular strain at Ca (structure I) is responsible for the hydrolytic reaction.It is important to note that for mechanistic speculations the angular strain (Aa) at Ca of the hydrolysed ligands is cited,22,24 but nucleophilic attack occurs at the carbon atoms of the triazine ring of unhydrolysed tptz (Cb, structure II of Scheme 2).The geometry after hydrolysis of the ligand may not be the true picture at the time when intact tptz is bound to the metal ion. After hydrolysis Ca is no longer in a ring and therefore it is expected that the value of the N]Ca]O angle would be higher compared to N]Cb]N and this effect allows the pyridine moiety to come closer to the metal ion which results in a lower value of Aa and shorter M]N distances. In fact the angles N]Ca]O in the copper(II) complexes are in the range 128–1308 22–24 which is significantly high compared to the ideal value of 1208.Therefore, for the discussion of mechanism the angular strain at Cb should be considered. The rhodium(III) and ruthenium(II) ions have a charge difference of unity only, but probably that alone also cannot make much difference in the activation of the ligand,39 there are many examples of complexes of CuII and NiII which activate ligands making carbon atoms susceptible to nucleophilic attack.40–43 In complex 1 the values of Aa (average 111.68) and Rh]N distances (Table 2) are comparable to those of the copper(II) complexes.In the case of 2 the average value of Ab is 112.58 which is close to Aa indicating a similar angular strain at Cb to that found at Ca. Therefore, the stability of 2, which resists hydrolysis of tptz, cannot be explained by the strain alone.We suggest a different factor assists the hydrolytic reaction. The coordination of tptz to the metal ion leads to destabilisation of Scheme 2 Drawings showing the relevant structural portions of complexes 1 (I) and 2 (II) N N N N N Ru Cb Ab 112.6 112.4 N N N Rh Ca Aa O O 111.8 112.1 I IIJ. Chem. Soc., Dalton Trans., 1997, Pages 2273–2277 2277 the triazine ring by enhancing the electron deficiency (LÆM s donation) upon it and this makes Cb susceptible to nucleophilic attack. Though RhIII and RuII are isoelectronic, RuII has the ability to form p-back bonds to unsaturated ligands; this interaction serves to increase the electron density on the triazine ring, compensating at least partially for the s-electronwithdrawing effect; a similar interaction (p-back bonding) for RhIII is apparently not effective.Kinetics and NMR studies 38,44 have demonstrated that the rate of base hydrolysis of coordinated nitriles (which also requires electron deficiency at the carbon atom of the nitrile) in rhodium(III) complexes is more than four orders of magnitude greater than structurally analogous ruthenium(II) complexes.39 In the 1H NMR spectra the downfield shift of the nitrile resonances is larger for the rhodium( III) than for the analogous ruthenium(II) complexes indicating that there is less electron density at the nitrile protons of the former complexes.44 We also made similar observations for the chemical shift of the protons in complexes 1 and 2 attached to carbon atoms adjacent to co-ordinated nitrogens of pyridyl moieties.† For free tptz they appeared at d 8.98, and in 1 and 2 the same proton appeared at d 8.27 and 7.75, respectively. In some other rhodium(III) complexes with unhydrolysed tptz 45 it appeared in the range d 8.6–9.0.These data show that the shielding of these protons in the ruthenium(II) complex is higher than in the rhodium(III) complex, consistent with the electron deficiency on the protons and also at the attached carbon atoms for rhodium(III) being greater compared to that in ruthenium(II) complexes. It is worth mentioning that rhodium( III) ion is capable of promoting hydrolysis of two coordinated tptz.This contrasts with the inference drawn earlier that once the hydrolysed tptz is bound to CuII the metal ion loses its ability to promote hydrolysis of any more tptz.In the present case we, therefore, suggest that the electronwithdrawing effect of the metal ion is the predominant factor, rather than angular strain, in hydrolysis of metal-bound tptz. Acknowledgements We are grateful to the Department of Science & Technology (DST), Government of India for financial support. Our sincere thanks go to Professor P. Natarajan, Director of this institute, for his keen interest, encouragement and valuable suggestions in this work.References 1 N. C. Thomas, B. L. Iolcy and A. L. Rheingold, Inorg. Chem., 1988, 27, 3426. 2 S. Chirayil, V. Hegde, Y. Jahng and R. P. Thummel, Inorg. Chem., 1991, 30, 2821. 3 N. Gupta, N. Grover, G. A. Neyhart, P. Singh and H. H. Thorp, Inorg. Chem., 1993, 32, 310. 4 R. M. Berger and J. R. Holcombe, Inorg. Chim. Acta, 1995, 232, 217. 5 R. M. Berger and D. D. Ellis II, Inorg. Chim. Acta, 1996, 241, 1. 6 E. M. Smolin and L. Rapoport, S-Triazines and Derivatives, Interscience, New York, 1959, p. 163. 7 P. Collins, H. Diehl and G. F. Smith, Anal. Chem., 1959, 31, 1862; P. Collins and H. Diehl, Anal. Chim. Acta, 1960, 22, 125; H. Diehl, E. B. Buchanan, jun. and G. F. Smith, Anal. Chem., 1960, 33, 1117; C. C. Tsen, Anal. Chem., 1961, 33, 849; W. A. Embry and G. H. Ayres, Anal. Chem., 1968, 40, 1499; M. J. Janmohamed and G. H. Ayres, Anal. Chem., 1972, 44, 2263. † As there is no proton in the triazine ring, we have chosen these protons which after co-ordination of tptz to the metal ion would be most affected in view of the electron density upon it. 8 G. K. Pagenkopf and D. W. Margerum, Inorg. Chem., 1968, 7, 2514; F. H. Fraser, P. Epstein and D. J. Macero, Inorg. Chem., 1972, 11, 2031. 9 J. Prasad and N. C. Peterson, Inorg. Chem., 1971, 10, 88. 10 D. A. Durharm, G. H. Frost and F. A. Hart, J. Inorg. Nucl. Chem., 1969, 31, 571; J. V. Kingston, E. M. Krankovits, R. J. Magee, E. C. Watton and R. S. Vagg, Inorg. Nucl. Chem. Lett., 1969, 5, 445. 11 J. Halfpenny and R. W. H. Small, Acta Crystallogr., Sect. B, 1982, 38, 939. 12 D. Sedney, M. Kahjehnassiri and W. M. Reiff, Inorg. Chem., 1981, 20, 3476; P. J. Taylor and A. A. Schitt, Inorg. Chim. Acta, 1971, 5, 691. 13 R. S. Vagg, R. N. Warrener, E. C. Watton, Aust. J. Chem., 1969, 22, 141; H. A. Goodwin, R. N. Sylva, R. S. Vagg and E. C. Watton, Aust. J. Chem., 1969, 22, 1605; G. A. Barclay, R. S. Vagg and E. C. Watton, Acta Crystallogr., Sect. B, 1978, 34, 1833. 14 R. S. Vagg, R. N. Warrener and E. C. Watton, Aust. J. Chem., 1967, 20, 1841; G. A. Barclay, R. S. Vagg and E. C. Watton, Aust. J. Chem., 1969, 22, 643; Acta Crystallogr., Sect. B, 1977, 13, 3487. 15 P. Byers, G. Y. S. Chan, M. G. B. Drew, M. J. Hudson and C. Madic, Polyhedron, 1996, 15, 2845. 16 A. M. Arif, F. A. Hart, M. B. Hursthouse, M. Thornton-Pett and W. Zhu, J. Chem. Soc., Dalton Trans., 1984, 2449. 17 G. Y. S. Chan, M. G. B. Drew, M. J. Hudson, N. S. Isaacs, P. Byers and C.Madic, Polyhedron, 1996, 15, 3385. 18 V. Balzani, F. Bolletta, M. T. Gandolfi and M. Maestri, Top. Curr. Chem., 1978, 75, 1. 19 N. E. Tokel-Takvoryan, R. E. Hemingway and A. J. Bard, J. Am. Chem. Soc., 1973, 95, 6582. 20 C.-T. Lin, W. Bottcher, M. Chou, C. Creutz and N. Sutin, J. Am. Chem. Soc., 1976, 98, 6536. 21 E. I. Lerner and S. J. Lippard, J. Am. Chem. Soc., 1976, 98, 5397. 22 E. I. Lerner and S. J. Lippard, Inorg. Chem., 1977, 16, 1546. 23 A. Cantarero, J. M. Amigo, J.Faus, M. Julve and T. Debaerdemaeker, J. Chem. Soc., Dalton Trans., 1988, 2033. 24 J. Faus, M. Julve, J. M. Amigo and T. Debaerdemaeker, J. Chem. Soc., Dalton Trans., 1989, 1681. 25 CAD-4 Software, Version 5, Enraf-Nonius, Delft, 1989. 26 E. I. Gabe, Y. Le Page, I. P. Charland, F. L. Lee and P. S. White, J. Appl. Crystallogr., 1989, 22, 384. 27 G. M. Sheldrick, Acta Crystallogr., Sect. A, 1990, 46, 467. 28 G. M. Sheldrick, SHELXL 93, Program for refinement of crystal structures, University of Göttingen, 1993. 29 G. F. Strouse, J. R. Schoonover, R. Deusing, S. Boyde, W. E. Jones, jun. and T. J. Meyer, Inorg. Chem., 1995, 34, 473. 30 D. B. MacQueen and J. D. Petersen, Inorg. Chem., 1990, 29, 2313. 31 A. Juris, V. Balzani, F. Barigelletti, S. Campagna, P. Belser and A. V. Zelewsky, Coord. Chem. Rev., 1988, 84, 85. 32 P. Didier, I. Ortmans, A. K. Mesmaeker and R. J. Watts, Inorg. Chem., 1993, 32, 5239. 33 I. Ortmans, P. Didier and A. K. Mesmaeker, Inorg. Chem., 1995, 34, 3695. 34 C. K. Johnson, ORTEP, Report ORNL-5138, Oak Ridge National Laboratory, Oak Ridge, TN, 1976. 35 G. Kew, K. De Armond and K. Hanck, J. Phys. Chem., 1974, 78, 727. 36 G. Kew, K. Hanek and K. De Armond, J. Phys. Chem., 1975, 79, 1828. 37 C. M. Bolinger, N. Story, B. P. Sullivan and T. J. Meyer, Inorg. Chem., 1988, 27, 4582. 38 S. C. Rasmussen, M. M. Richter, E. Yi, H. Place and K. J. Brewer, Inorg. Chem., 1990, 29, 3926. 39 A. W. Zanella and P. C. Ford, Inorg. Chem., 1975, 14, 42. 40 S. Komiya, S. Suzuki and K. Watanabe, Bull. Chem. Soc. Jpn., 1971, 44, 1440. 41 S. Suzuki, M. Nakahara and K. Watanabe, Bull. Chem. Soc. Jpn., 1971, 44, 1441. 42 G. N. Storhoff and H. C. Lewis, jun., Coord. Chem. Rev., 1977, 23, 1. 43 P. Paul and K. Nag, Inorg. Chem., 1987, 26, 1586. 44 R. D. Foust, jun. and P. C. Ford, J. Am. Chem. Soc., 1972, 94, 5686. 45 P. Paul, B. Tyagi, A. K. Bilakhiya, M.-M. Bhadbhade and E. Suresh, unpublished work. Received 16th December 1996; Paper 6/08433H

ISSN:1477-9226    

DOI:10.1039/a608433h

出版商:RSC    

年代:1997

数据来源: RSC

摘要:

DALTON J. Chem. Soc., Dalton Trans., 1997, Pages 2279–2291 2279 Control of the reactivity of trans-[Mo2(cp)2(CO)(Y]] Z)(Ï-SR)2] (cp = Á-C5H5; Y]] Z = CO or CN2) by the sulfur substituents (R = Me, Pri, But, Ph or CF3). Crystal structure of trans- [Mo2(cp)2(CO)(CNMe)(Ï-SCF3)2] † Marie-Laurence Abasq,a David L. Hughes,b François Y. Pétillon,a Roger Pichon,a Christopher J. Pickett b and Jean Talarmin*,a a URA CNRS 322 ‘Chimie, Electrochimie Moléculaires et Chimie Analytique’, Université de Bretagne Occidentale BP 809, 29285 Brest Cédex, France b Nitrogen Fixation Laboratory, John Innes Centre, Colney Lane, Norwich NR4 7UH, UK Reaction of trans-[Mo2(cp)2(CO)2(m-SR)2] (cp = h-C5H5; R = Me, Pri, But, Ph or CF3; syn and anti isomers) with cyanide ion gave the corresponding cyanide complexes trans-[Mo2(cp)2(CO)(CN)(m-SR)2]2, except with R = But where no reaction was observed.For R = CF3, two isomers having a syn orientation of the sulfur substituents were obtained.The nature of the R groups is shown to have a crucial influence on the site of the reaction of the cyanide complexes with Me3O1. Complexes where R = Me, Pri (syn and anti isomers) or Ph (anti isomer) were Smethylated, whereas N-methylation was observed for R = Ph (syn isomer) or CF3 (anti and both syn isomers). This is ascribed to electronic effects of the R groups which control the site of methylation by switching the reaction from orbital control (S-methylation) to charge control (N-methylation).For R = CF3, the R groups also affect the reaction of the dicarbonyl precursor with a Y]] ] Z substrate since the preferred site of attack is different for Y]] ] Z = CN2 and Y]] ] Z = RNC. In previous papers of this series we have shown that reactive intermediates can be generated electrochemically from stable precursors possessing the {Mo2(m-SR)n} core (n = 2 or 3).1,2 In the absence of substrates, deactivation of these sites leads to the thermodynamically stable complex trans-[Mo2(cp)2(CO)2- (m-SMe)2] (cp = h5-C5H5; trans relates to the positions of the CO and cp ligands).However, this complex possesses an interesting reactivity. In particular, its reaction with isocyanides produces the substituted [Mo2(cp)2(CO)(CNR)(m-SMe)2] derivatives and the reaction can be reversed by CO (Scheme 1).2 Furthermore, as we show herein, complex 1a reacts with cyanide ion. This suggested that trans-[Mo2(cp)2(CO)2(m-SMe)2] might be used as a platform at which various isocyanides could be assembled through the successive reactions of cyanide ions and different electrophilic reagents.Many studies have been concerned with the synthesis of isocyanides in the co-ordination sphere of metal centres and this has been reviewed recently.3 The interest of trans-[Mo2(cp)2(CO)2(m-SMe)2] in this context lies in the fact that the new isocyanide molecule can be easily released by a reaction of the complex with carbon monoxide, Scheme 1.The crucial question was whether or not electrophilic reagents attack the co-ordinated CN2 group thus enabling a cycle such as that shown in Scheme 2. The results presented here demonstrate that the cyanide complex [Mo2(cp)2(CO)(CN)(m-SMe)2]2 2a does indeed react with Me3O1 but that the site of attack is a sulfur lone pair rather than the bound cyanide. We also deduce the factors which control the site of attack by studying the reactions of various new cyanide complexes related to 2a with Me3O1.A preliminary account of this work has been published.4 Results Synthesis and methylation of [Mo2(cp)2(CO)(CN)(Ï-SMe)2]2 2a Synthesis. The complex trans-[Mo2(cp)2(CO)2(m-SMe)2] 1a † Electrochemistry of dinuclear, thiolate-bridged transition-metal compounds. Part 9. Part 8: F. Y. Pétillon, S. Poder-Guillou, P. Schollhammer and J. Talarmin, New J. Chem., in the press. exists as two isomers which differ by the relative orientation of the sulfur substituents,5,‡ syn and anti, the spectroscopic and redox data for which are listed in Tables 1 and 2 respectively.Cyclic voltammetric monitoring of the stepwise addition of CN2 to a solution of trans,syn-[Mo2(cp)2(CO)2(m-SMe)2] syn-1a in MeCN–[NBu4][PF6] shows that the reaction [equation (1)] Scheme 1 d = Mo(cp); R = But, xylyl or CH2Ph Scheme 2 ‡ The spectroscopic data for trans,syn-[Mo2(cp)2(CO)2(m-SMe)2] are in very good agreement with those reported by Li and Curtis.6 The other isomer was assigned by these authors as cis,syn-[Mo2(cp)2(CO)2(m- SMe)2].2280 J.Chem. Soc., Dalton Trans., 1997, Pages 2279–2291 [Mo2(cp)2(CO)2(m-SMe)2] 1 CN2 æÆ syn-1a anti-1a [Mo2(cp)2(CO)(CN)(m-SMe)2]2 1 CO (1) syn-2a anti-2a leads to the formation of a 1 : 1 derivative, resistant to further substitution.4 The reaction carried out with the trans,anti analogue gives similar results. The assignment of the reaction product as [Mo2(cp)2(CO)- (CN)(m-SMe)2]2 2a is confirmed by spectroscopic data, Table 1.The 13C NMR spectrum shows the presence of one CO ligand instead of two in the precursor and, although it is not detected in the 13C NMR spectrum, the co-ordination of CN2 is evidenced by infrared spectroscopy. The retention of the syn orientation of the methyl groups on the bridging sulfur atoms in syn-2a is confirmed by the singlet for the methyl protons in the 1H NMR spectrum (Table 1), and the trans geometry of the dicarbonyl precursor is also maintained (see Discussion section).The cyanide complex syn-2a is characterized [Fig. 1(b)] by reversible oxidation and reduction processes occurring at slightly different potentials for the syn and anti isomers (Table 2). Whereas the oxidation of the parent dicarbonyl is an irreversible process [Fig. 1(a)], that of the cyanide derivative is Fig. 1 Cyclic voltammetry of a 2.2 mmol dm23 solution of trans- [Mo2(cp)2(CO)2(m-SMe)2] syn-1a under N2 before (a) and after (b) the addition of 1 equivalent [NBu4][CN]; the curve in (c) was recorded after treatment of solution (b) with CO; (d) was obtained after the solution (b) had been flushed with N2 (MeCN–[NBu4][PF6] electrolyte; vitreous carbon electrode, n = 0.2 V s21) reversible on the cyclic voltammetry (CV) time-scale; however, in the presence of more than 1 equivalent CN2 the oxidation loses its reversibility.One reduction step (instead of two for the dicarbonyl precursor) is observed in the potential window available in MeCN or tetrahydrofuran (thf) electrolyte.This illustrates the effects of substitution of CN2 for CO: first, the presence of CN2 stabilizes the oxidized form of the complex (formally MoIII–MoII) with respect to CO and secondly, as expected from the substitution of CO by a donor ligand, the redox processes are shifted towards negative potentials with respect to the redox steps of the dicarbonyl precursor, by about 900 mV. The magnitude of the potential shift is attributed, at least in part, to the negative charge carried by the cyanide complex.In agreement with the potential shift, the infrared spectrum of [Mo2(cp)2(CO)(CN)(m-SMe)2]2 shows the band of the remaining CO ligand at 1780 cm21, that is ca. 70 cm21 lower than for the parent dicarbonyl complex (Table 1): this indicates a more important Mo to CO back donation due to the presence of the net donor cyanide ligand at the neighbouring metal centre. The reaction with CN2 can be reversed partially upon treatment of a solution of the cyanide complex with CO: as shown in Fig. 1(c), an equilibrium mixture containing the dicarbonyl and the cyanide complexes in a ca. 2 : 1 ratio is obtained under 1 atm (ca. 105 Pa) CO.§ That CN2 is only partially displaced whereas RNC is totally displaced under similar conditions shows that the cyanide ligand is more strongly retained than isocyanides RNC (R = But, xylyl or CH2Ph) 2 by the {Mo2(cp)2(CO)(m-SMe)2} core.Flushing the CO-saturated solution with N2 or Ar regenerates the cyanide complex syn-2a, Fig. 1(d). A third isomer of [Mo2(cp)2(CO)(CN)(m-SMe)2]2, syn-2a9 can be prepared electrochemically, as described previously for isocyanide complexes.2 Controlled-potential electrolysis (CPE) of [Mo2(cp)2(CO)2(m-SMe)3]1 performed in the presence of CN2 produces this new isomer (Figs. 2 and 3), which is accessible only by the electrochemical route. From Figs. 2(a) and 2(b) it can be seen that the addition of cyanide ions modifies the CV of [Mo2(cp)2(CO)2(m-SMe)3]1 in much the same way as did the addition of isocyanide:2 the redox systems of the intermediate cis-[Mo2(cp)2(CO)2(m-SMe)2] are replaced by new peaks due to syn-2a9.The cyclic voltammogram recorded after electrolysis (1.3–1.7 F mol21 starting material, N2 purge¶) shows the redox processes of the electrogenerated complex at E1/2 red = 22.77 V and E1/2 ox = 21.01 V [Table 2, Fig. 2(c)].No analytical data could be obtained for this complex which could not be separated from the supporting electrolyte. However, that syn-2a9 is a cyanide complex with syn-Me substituents is demonstrated by the following experiments. Purging CO through a solution of it [Fig. 3(a), 3(b)] and flushing the resulting solution with N2 [Fig. 3(c)] converted syn-2a9 into syn-2a, which can be obtained directly from trans,syn-[Mo2(cp)2(CO)2(m-SMe)2] and CN2 (Scheme 3). Indeed, the cyanide complex in Fig. 3(b) is no more syn-2a9 but syn-2a, as demonstrated by the reduction potential E1/2 red = 22.86 V, instead of E1/2 red = 22.77 V in Fig. § We thank Dr. J. N. Verpeaux for drawing our attention to the fact that the equilibrium is not shifted by the reduction of the dicarbonyl complex at the electrode. This is shown by the fact that the ratio of the oxidation to reduction peak currents of the cyanide complex is almost the same in Figs. 1(b) and 1(c). The loss of reversibility of the oxidation of the cyanide complex, in Fig. 1(c), is due to the presence of free cyanide (released by co-ordination of CO): when a small excess of cyanide is added to the dicarbonyl complex the oxidation loses reversibility, due to a reaction of the oxidized complex with CN2. ¶ Nitrogen is purged through the catholyte in order to remove the CO released on binding of CN2 or RNC to cis-[Mo2(cp)2(CO)2(m-SMe)2]; CO was shown2 to catalyse the cis/trans isomerization of [Mo2(cp)2- (CO)2(m-SMe)2]; therefore, the nitrogen purge during electrolyses prevents the formation of trans-[Mo2(cp)2(CO)2(m-SMe)2] and of products derived therefrom.J.Chem. Soc., Dalton Trans., 1997, Pages 2279–2291 2281 Table 1 Spectroscopic data NMR (d, J/Hz; CDCl3) IR (cm21) Complex 1H 13C-{1H} (CH2Cl2) Dicarbonyls syn -1a syn-[Mo2(cp)2(CO)2(m-SMe)2] a anti-1a anti-[Mo2(cp)2(CO)2(m-SMe)2] a syn-1b syn-[Mo2(cp)2(CO)2(m-SPri)2] anti-1b anti-[Mo2(cp)2(CO)2(m-SPri)2] syn-1c syn-[Mo2(cp)2(CO)2(m-SBut)2] b syn-1d syn-[Mo2(cp)2(CO)2(m-SPh)2] anti-1d anti-[Mo2(cp)2(CO)2(m-SPh)2] syn-1e syn-[Mo2(cp)2(CO)2(m-SCF3)2] anti-1e anti-[Mo2(cp)2(CO)2(m-SCF3)2] Cyanides c syn-2a K[syn-Mo2(cp)2(CO)(CN)(m-SMe)2] syn-2e9 K[syn-Mo2(cp)2(CO)(CN)(m-SCF3)2] syn-2e anti-2e K[anti-Mo2(cp)2(CO)(CN)(m-SCF3)2] syn-2d [NBu4][syn-Mo2(cp)2(CO)(CN)(m-SPh)2] Isocyanides d 4a syn-[Mo2(cp)2(CO)(CNMe)(m-SCF3)2] Brown-yellow 4b syn-[Mo2(cp)2(CO)(CNMe)(m-SCF3)2] Red anti-[Mo2(cp)2(CO)(CNMe)(m-SCF3)2] Green syn-[Mo2(cp)2(CO)(CNBut)(m-SCF3)2] Red anti-[Mo2(cp)2(CO)(CNBut)(m-SCF3)2] Green 5.44 (s, 5 H, cp) 5.29 (s, 5 H, cp) 2.36 (s, 6 H, SMe) 5.40 (s, 10 H, cp) 2.29 (s, 6 H, SMe) 5.47 (s, 5 H, cp) 5.27 (s, 5 H, cp) 2.43 (m, 2 H, CH) 1.49 (d, 6 H, J = 6.7, CH3) 1.38 (d, 6 H, J = 6.7, CH3) 5.40 (s, 10 H, cp) 2.37 (m, 2 H, CH) 1.49 (d, 6 H, J = 6.7, CH3) 1.29 (d, 6 H, J = 6.7, CH3) 5.49 (s, 5 H, cp) 5.15 (s, 5 H, cp) 1.28 [s, 18 H, C(CH3)3] 7.33–7.17 (m, 10 H, Ph) 5.63 (s, 5 H, cp) 5.28 (s, 5 H, cp) 7.37–7.15 (m, 10 H, Ph) 5.42 (s, 10 H, cp) 5.58 (s, 5 H, cp) 5.35 (s, 5 H, cp) 5.50 (s, 10 H, cp) 5.19 (s, 5 H, cp) 4.25 (s, 5 H, cp) 2.13 (s, 6 H, SMe) 5.20 (s, 5 H, cp) 4.81 (s, 5 H, cp) 5.34 (s, 5 H, cp) 4.53 (s, 5 H, cp) 5.31 (s, 5 H, cp) 4.63 (s, 5 H, cp) 7.54–6.95 (Ph) 5.04 (s, 5 H, cp) 4.34 (s, 5 H, cp) 5.31 (s, 5 H, cp) 5.25 (s, 5 H, cp) 3.36 (s, 3 H, MeNC) 5.47 (s, 5 H, cp) 4.94 (s, 5 H, cp) 3.34 (s, 3 H, MeNC) 5.44 (s, 5 H, cp) 5.07 (s, 5 H, cp) 3.38 (s, 3 H, MeNC) 5.45 (s, 5 H, cp) 4.87 (s, 5 H, cp) 1.24 (s, 9 H, ButNC) 5.40 (s, 5 H, cp) 5.03 (s, 5 H, cp) 1.24 (s, 9 H, ButNC) 252.0, 246.5 (CO) 91.1 (cp) 32.6 (SCH3) 249.5 (CO) 91.0 (cp) 34.3 (SCH3) 250.8, 246.3 (CO) 91.6, 90.7 (cp) 50.1 (SCH) 26.0, 25.8 (CH3) 249.2 (CO) 91.1 (cp) 52.4 (SCH) 26.2, 26.0 (CH3) 261.4, 247.0 (CO) 91.7, 90.5 (cp) 47.3 [C(CH3)3] 32.6 (CH3) 248.4, 242.9 (CO) 147.6, 130.3, 128.0, 126.3 (C6H5) 92.0, 91.6 (cp) 246.8 (CO) 148.0, 130.6, 128.1, 126.5 (C6H5) 92.0 (cp) 241.2, 239.8 (CO) 136.2 (q, JCF = 320, CF3) 92.0, 91.1 (cp) 241.0 (CO) 136.4 (q, JCF = 320, CF3) 91.5 (cp) 243.6 (CO) 89.1, 87.3 (cp) 32.7 (SCH3) 239.6 (CO) 90.0, 89.3 (cp) 238.6 (CO) 152.3 (CN) 90.3, 88.8 (cp) 237.8 (CO) 151.1 (CN) 90.1, 89.2 (cp) 244.0 (CO) 89.6, 88.3 (cp) 239.7 (CO) 138.2 (JCF = 321, CF3) 90.2, 89.5 (cp) 32.0 (CH3NC) 237.8 (CO) 177.0 (CNCH3) 136.6 (JCF = 322, CF3) 90.7, 89.5 (cp) 31.05 (CH3NC) 237.8 (CO) 181.8 (CNCH3) 137.6 (JCF = 320, CF3) 137.0 (JCF = 323, CF3) 90.3, 89.96 (cp) 31.6 (CH3NC) 238.1 (CO) 168.7 (CNBut) 136.7 (JCF = 322, CF3) 90.5, 89.4 (cp) 57.5 [C(CH3)3] 30.1 [C(CH3)3] 238.1 (CO) 176.6 (CNBut) 137.7 (JCF = 321, CF3) 137.2 (JCF = 323, CF3) 90.2, 89.8 (cp) 58.3 [C(CH3)3] 30.6 [C(CH3)3] n(CO) 1855, 1885 (sh) n(CO) 1840 n(CO) 1855 n(CO) 1845 n(CO) 1845, 1880 n(CO) 1855, 1905 n(CO) 1855 n(CO) 1890, 1950 n(CO) 1900 n(CN) 2060 n(CO) 1780 n(CN) 2040 n(CO) 1885 n(CN) 2070 n(CO) 1800 n(CN) 2140 n(CO) 1860 n(CN) 2120 n(CO) 1860 n(CN) 2125 n(CO) 1855 n(CN) 2110 n(CO) 1860 a See ref. 6. b The 1H NMR data are in agreement with those reported by Benson et al.7 c Solvent for NMR was CD3CN. d The d(19F) values in CDCl3 are: 241.0; 242.9; 239.7, 244.0; 241.8; 239.7, 242.5, respectively.2282 J. Chem. Soc., Dalton Trans., 1997, Pages 2279–2291 Table 2 Redox potentials (V vs. ferrocene–ferrocenium) as measured from CV experiments (MeCN–NBu4PF6) Complex E1/2 red1 E1/2 red2 Ep ox Ref. Dicarbonyls syn-1a syn-[Mo2(cp)2(CO)2(m-SMe)2] Brown anti-1a anti-[Mo2(cp)2(CO)2(m-SMe)2] Green syn-1b syn-[Mo2(cp)2(CO)2(m-SPri)2] Brown anti-1b anti-[Mo2(cp)2(CO)2(m-SPri)2] Green syn-1c syn-[Mo2(cp)2(CO)2(m-SBut)2] Green syn-1d syn-[Mo2(cp)2(CO)2(m-SPh)2] Brown anti-1d anti-[Mo2(cp)2(CO)2(m-SPh)2] Green syn-1e syn-[Mo2(cp)2(CO)2(m-SCF3)2] Red-brown anti-1e anti-[Mo2(cp)2(CO)2(µ-SCF3)2] Green Cyanides syn-2a syn-[Mo2(cp)2(CO)(CN)(m-SMe)2]2 anti-2a anti-[Mo2(cp)2(CO)(CN)(m-SMe)2]2 syn 2a9 syn-[Mo2(cp)2(CO)(CN)(m-SMe)2]2 syn-2b syn-[Mo2(cp)2(CO)(CN)(m-SPri)2]2 anti-2b anti-[Mo2(cp)2(CO)(CN)(m-SPri)2]2 syn-2d syn-[Mo2(cp)2(CO)(CN)(m-SPh)2]2 anti-2d anti-[Mo2(cp)2(CO)(CN)(m-SPh)2]2 syn-2e syn-[Mo2(cp)2(CO)(CN)(m-SCF3)2]2 syn-2e9 syn-[Mo2(cp)2(CO)(CN)(m-SCF3)2]2 anti-2e anti-[Mo2(cp)2(CO)(CN)(m-SCF3)2]2 S-Methylated cyanides 3a [Mo2(cp)2(CO)(CN)(m-SMe2)(m-SMe)] e 3a9 [Mo2(cp)2(CO)(CN)(m-SMe2)(m-SMe)] f 3b [Mo2(cp)2(CO)(CN){m-S(Me)Pri}(m-SPri)] g 3d [Mo2(cp)2(CO)(CN){m-S(Me)Ph}(m-SPh)] h Isocyanides 4a syn-[Mo2(cp)2(CO)(CNMe)(m-SCF3)2] 4b syn-[Mo2(cp)2(CO)(CNMe)(m-SCF3)2] anti-[Mo2(cp)2(CO)(CNMe)(m-SCF3)2] syn-[Mo2(cp)2(CO)(CNBut)(m-SCF3)2] anti-[Mo2(cp)2(CO)(CNBut)(m-SCF3)2] [Mo2(cp)2(CO)(CNMe)(m-SPh)2] [Mo2(cp)2(CO)(CNBut)(m-SPh)2] 21.94 21.91 21.92 21.95 21.93 21.97 21.94 21.83 21.83 21.58 21.57 22.86 22.88 22.77 22.95 22.95 22.72 22.70 22.56 b 22.45 22.51 c 22.46 d 22.18 22.13 22.23 22.07 21.86 21.98 21.91 22.03 21.94 22.20 c 22.20 c 22.27 22.25 22.23 22.26 22.27 22.33 22.4 22.12 22.14 22.02 22.0 —————————— 22.78 c 22.76 c 22.84 c 22.64 c 22.18 22.19 22.16 22.3 c 22.15 — 22.36 c 20.03 20.10 20.05 20.07 0 20.03 20.01 20.04 0 0.50 0.43 20.89 a 20.96 a 21.01 a 20.87 a 20.93 a 20.77 a 20.85 a 20.43 a 20.52 a 20.48 a 20.28 20.37 20.23 20.18 10.02 10.13 10.06 10.13 10.06 20.24 20.27 This work 2 This work 2 This work This work This work This work This work This work This work 88 This work This work a Reversible one-electron oxidation.b Ill defined peaks.c Irreversible. d At 220 8C. e E1/2 red3 22.87 V. f Overlapping reduction peaks around 22.8 V. g E1/2 red3 22.95 V. h E1/2 red3 22.73 V. 3(a) [the oxidation in Fig. 3(b) is less reversible due to the presence of free cyanide, see above]. This shows that the deco-ordination of CN2 from syn-2a9 is irreversible, and the presence of syn-2a under CO is a consequence of the equilibrium between this complex and the dicarbonyl syn-1a [compare Figs. 3(b) and 1(c)]. Exactly the same processes were observed for the ButNC analogues of syn-2a9 and syn- Scheme 3 2a,2 and, therefore, we assign to syn-2a9 the same geometry as for the isocyanide analogue, that is trans-syn with the cyanide ligand opposite to the methyl groups. Reaction with Me3O1. The addition of the methylating agent [OMe3][BF4] to a MeCN–[NBu4][PF6] solution of the cyanide complex syn-2a leads to changes in the CV, characterized by complete loss of the oxidation at E1/2 = 20.89 V and by the appearance of new reduction (E1/2 red1 = 22.18, Ep red2 = 22.79, E1/2 red3 = 22.87 V) and oxidation processes (Ep ox = 20.28 V) (Fig. 4). The primary processes of the methylated product are observed at potentials ca. 700 mV more positive than those of the cyanide precursor, which is consistent with neutralization of the negative charge of the latter. The assignment of the methylated product 3a as the mixed thiolate–thioether-bridged cyanide complex is based upon spectroscopic data, and comparison of the cyclic voltammetry with that of an authentic sample of [Mo2(cp)2(CO)(CN)(m-SMe)(m-SMe2)] 3a (Table 2), synthesized as shown in Scheme 4.8 The 13C NMR spectrum of complex 3a shows the resonance of the CN ligand at d 152.6, and the CN band in the infrared is observed at 2080 cm21.8 These are diagnostic of cyanide ligands in this type of complexes, and we will show below that isocyanides are characterized by 13C NMR resonances around d 170–180 and by infrared bands in the range 2100–2140 cm21.The assignment of the product as [Mo2(cp)2(CO)(CN)(m-SMe)- (m-SMe2)] is confirmed by the fact that, unlike [Mo2(cp)2(CO)- (CNR)(m-SMe)2] (R = But, xylyl or CH2Ph),2 the methylatedJ. Chem. Soc., Dalton Trans., 1997, Pages 2279–2291 2283 compound does not regenerate the parent dicarbonyl under a carbon monoxide atmosphere. There is no detectable difference in the redox potentials of the methylated products obtained from either syn or anti isomers of the cyanide complex: this suggests that a single product is formed according to Scheme 5. As for syn- and anti-2a, the methylation of syn-2a9 yields a thiolate–thioether-bridged complex.8 This is shown by the electrochemical data of the product, 3a9, which are quite analogous to those of 3a (Table 2), and totally different from those of an isocyanide complex; furthermore, the reaction of 3a9 with CO does not generate a dicarbonyl complex.The question which arises then is: how to favour the alkylation of the cyanide ligand? The first, obvious, answer is to alkylate the sulfur lone pairs before treating the resulting product with cyanide and Me3O1.However, as shown elsewhere, the Fig. 2 Cyclic voltammetry of (a) a 1.6 mmol dm23 solution of [Mo2(cp)2(CO)2(m-SMe)2]1, (b) in the presence of 1 equivalent [NBu4][CN] and (c) after electrolysis in the presence of 1 equivalent [NBu4][CN] with a nitrogen purge through the catholyte (mercury-pool cathode; electrolysis potential = 22.1 V) (MeCN][NBu4][PF6] electrolyte; vitreous carbon electrode, n = 0.2 V s21) Scheme 4 isocyanide complexes [Mo2(cp)2(CO)(CNR)(m-SMe)(m-SMe2)]1 and [Mo2(cp)2(CO)(CNR)(m-SMe2)2]21 do not significantly release RNC on reaction with CO.8 We have consequently looked for other ways of promoting the methylation of the cyanide ligand in thiolate-bridged compounds [Mo2(cp)2- (CO)(CN)(m-SR)2]2.In this context, we have investigated the influence of the steric and electronic properties of the R substituents on the site of electrophilic attack.Influence of the nature of the sulfur substituents of [Mo2(cp)2- (CO)(CN)(Ï-SR)2]2 on the site of electrophilic attack by Me3O1 The substituents on the bridging sulfur atoms may influence the alkylation of the cyanide complexes, via both their steric and their electronic properties. Replacing the methyl groups on the sulfur atoms by bulkier tert-butyl or isopropyl substituents may result in a change in the orientation of the reaction of the electrophile.On the other hand, a modification of the electronic properties of the sulfur substituents may be responsible for a diversion of the electrophilic attack. Fenske and Milletti 9 have reported Fig. 3 Cyclic voltammetry of (a) a 1.6 mmol dm23 solution of trans- [Mo2(cp)2(CO)(CN)(m-SMe)2]2 2a9 produced as shown in Fig. 2(c), (b) after purging CO through the solution in (a) (CV under CO) and (c) after flushing solution (b) with N2 (MeCN–[NBu4][PF6] electrolyte; vitreous carbon electrode, n = 0.2 V s21) Scheme 52284 J.Chem. Soc., Dalton Trans., 1997, Pages 2279–2291 that the reaction of a complex with a nucleophile can be either orbitally controlled when the energy difference between the HOMO (highest occupied molecular orbital) of the nucleophile and the LUMO (lowest unoccupied molecular orbital) of the complex is negligible, or charge controlled when this energy gap is larger.In the first situation the LUMO (or an orbital close to it in energy and of correct symmetry) of the complex is attacked, in the second the reaction takes place at the site where the charge is localized. This can be extended to the present case of electrophilic attack of [Mo2(cp)2(CO)(CN)(m-SR)2]2 by Me3O1. The alkylation of a sulfur lone pair for R = Me suggests that the reaction is under orbital control, and that the sulfur lone pairs are close to the HOMO in energy and have correct symmetry for an interaction with the electrophile’s LUMO.The replacement of the methyl substituents by electron-withdrawing groups such as CF3 will stabilize the HOMO of the complex as well as the sulfur lone pairs, and will thus increase the energy gap between these orbitals and the LUMO of Me3O1: this might switch the reaction from orbitally to charge controlled and consequently favour the alkylation of the cyanide ligand. In order to discriminate between the steric and the electronic effects of the R substituents, R groups with similar electronic properties are needed to probe the steric influence and vice versa, substituents of similar size will be used to check on their electronic effect on the orientation of the reaction.We have therefore prepared the dicarbonyls trans-[Mo2(cp)2(CO)2- (m-SR)2] (R = But or Pri) and investigated their redox behaviour in order to check whether or not this substitution induces an important change in the electronic properties of the complexes.We have also synthesized trans-[Mo2(cp)2(CO)2(m-SCF3)2], since the similar size of the CF3 and alkyl substituents makes interferences of steric effects unlikely. Synthesis and electrochemistry of trans-[Mo2(cp)2(CO)2- (Ï-SR)2] (R = Pri, But, CF3 or Ph). The complexes were obtained from the reaction of the molybdenum dimer [Mo2(cp)2(CO)n] (n = 4 or 6) with the disulfide RSSR or with the thiol in the presence of triethylamine (see Experimental section).The syn and anti isomers were separated by column chromatography or recrystallization. The ButS2 and PhS2 complexes were already known7,10 but the electrochemical and/or spectroscopic properties of the syn and anti isomers had not been reported. The complexes with R = CF3 or Pri are new. The 1H, 13C NMR, and infrared data are listed in Table 1. Several X-ray structural studies of [M2(cp)2(CO)2(m-SR)2] have shown that the complexes are always in a trans geometry.From the crystal analysed by X-ray diffraction for M = Mo and R = But, the R groups were shown to adopt a mutually syn orientation,7,10 whereas for M = Mo, R = Ph10 and M = W, R = Pri 11 the R substituents were found in anti positions in the solid. All the trans-[Mo2(cp)2(CO)2(m-SR)2] complexes undergo two Scheme 6 R = Me syn- and anti-1a, Pri syn- and anti-1b, But syn-1c, Ph syn- and anti-1d, CF3 syn- and anti-1e reversible, one-electron reduction steps (Scheme 6) and one irreversible multielectron oxidation in a MeCN electrolyte (Table 2).The addition of two electrons to the Mo]Mo antibonding LUMO12 leads to cleavage of one metal–metal bond in the dianion, in agreement with the 18-electron rule. There is almost no difference between the reduction potentials of the syn and anti isomers of these complexes (Table 2), which do not interconvert in solution at room temperature. The redox potentials are sensitive to the nature of R: as expected, complexes with the more electron-withdrawing substituents are easier to reduce (cf.Table 2, CF3 and Ph). The alkanethiolate derivatives all show similar redox potentials, which confirms that the R substituents (R = Me, But or Pri) have similar effects on the redox orbitals of the complexes: the SMe, SPri and SBut analogues can therefore be used to probe the effect of increasing the size (with no effect of the electronic properties) of the sulfur substituents on the reactivity of the compounds.Reaction of trans-[Mo2(cp)2(CO)2(Ï-SR)2] (R = Pri, But, CF3 or Ph) with CN2. The complexes trans-[Mo2(cp)2(CO)2(m-SR)2] react with 1 equivalent [NBu4][CN] at room temperature [reaction (2)] except for R = But where no reaction is detected. The [Mo2(cp)2(CO)2(m-SR)2] 1 CN2 æÆ [Mo2(cp)2(CO)(CN)(m-SR)2]2 1 CO (2) cyanide complexes were not stable in the solid state and only the compounds with R = CF3 2e or Ph syn-2d were isolated and characterized spectroscopically (Table 1).The isopropyl derivatives and anti-2d were formed in the CV cell and their redox potentials are listed in Table 2. The 13C NMR resonance of the cyanide ligand and the infrared n(CN) are in the region expected for cyanide complexes, and the presence of the remaining CO in the complexes is also evident on the 13C NMR and infrared spectra (Table 1). The substitution of one CO by one cyanide ligand does not affect the disposition of the R substituents.In the 13C NMR spectrum of the benzenethiolate derivative syn-2d four resonances only are observed in the range d 125–150, which are characteristic of the equivalent phenyl substituents. The NMR spectrum of the crude product of the reaction of trans,syn-[Mo2(cp)2(CO)2(m-SCF3)2] with 1 equivalent of [NBu4][CN] shows the presence of two isomers of the cyanide complex (60 : 35), both different from the single product of the reaction of CN2 with the trans,anti isomer (Table 1, Scheme 7).The redox potentials of the cyanide complexes (Table 2) are influenced by the nature of the R substituents in the same way as for the dicarbonyl precursors, with a slight difference between the syn and anti isomers. Influence of the steric and electronic properties of the substituents R on the site of methylation of [Mo2(cp)2(CO)(CN)(Ï-SR)2]2 (R = Pri, CF3 or Ph). Steric effect. As the bulkiest substituent used in this study, R = But, prevents the formation of the cyan- Scheme 7J.Chem. Soc., Dalton Trans., 1997, Pages 2279–2291 2285 ide complex, the steric effect of R upon the site of alkylation can be estimated only from a comparison of the reaction carried out with the MeS2 and PriS2 analogues. The NMR data (Experimental section) and the electrochemical behaviour (Table 2) of the product formed on addition of Me3O1 to [Mo2- (cp)2(CO)(CN)(m-SPri)2]2 are similar to those of [Mo2(cp)2- (CO)(CN)(m-SMe)(m-SMe)2] [compare Figs. 4 and 5(a)]. The two-electron reduction of this complex led to S]C bond cleavage 8 giving rise to a product identified as [Mo2(cp)2(CO)(CN)- (m-SPri)2]2 by its redox processes (E1/2 red = 22.95, E1/2 ox ª 20.9 V). The CN resonance in the 13C NMR spectrum (Experimental section) of the methylated (PriS)2 complex is observed at d 150.6 and is quite different from that of ButNC (d 183.4) in [Mo2(cp)2(CO)(CNBut)(m-SPri)2]. This shows that the methylated product is [Mo2(cp)2(CO)(CN){m-S(Me)Pri}(m-SPri)] 3b.Consistent with the methylation at a sulfur lone pair, the methylated PriS2 complex does not regenerate the parent Fig. 4 Cyclic voltammetry of [Mo2(cp)2(CO)(CN)(m-SMe)(m-SMe2)] 3a obtained by the successive additions of 1 equivalent [NBu4][CN] and of Me3O1 to a 2.2 mmol dm23 solution of trans-[Mo2(cp)2(CO)2(m-SMe)2] syn-1a (MeCN–[NBu4][PF6] electrolyte; vitreous carbon electrode, n = 0.2 V s21) Fig. 5 Cyclic voltammetry of trans-[Mo2(cp)2(CO)(CN){m-S(Me)R}- (m-SR)] obtained by successive additions of 1 equivalent [NBu4][CN] and of Me3O1 to trans-[Mo2(cp)2(CO)2(m-SR)2]: (a) R = Pri 3b; (b) R = Ph 3d (MeCN–[NBu4][PF6] electrolyte; vitreous carbon electrode, n = 0.2 V s21) dicarbonyl on reaction with CO, in contrast with [Mo2(cp)2- (CO)(CNBut)(m-SPri)2].Electronic effect. This was probed by comparing the reactivity of the anti and of both syn isomers of [Mo2(cp)2(CO)(CN)- (m-SCF3)2]2 to that of the methanethiolate analogues.The reaction of [Mo2(cp)2(CO)(CN)(m-SPh)2]2 with Me3O1 has also been investigated; in this case, both the size and the electronic properties of the sulfur substituents are different from those of the methyl groups, and it is difficult to assess which of these parameters is responsible for the observed reactions. The addition of Me3O1 to [Mo2(cp)2(CO)(CN)(m-SPh)2]2 produces two different species. When the cyanide complex is prepared from trans,syn-[Mo2(cp)2(CO)2(m-SPh)2] syn-1d the major product of the methylation is an isocyanide derivative: 1H, 13C NMR [d 183.5 (CN)] and infrared spectroscopies show the presence of the isocyanide ligand (Table 1).The complex reacts with CO to regenerate the dicarbonyl parent almost quantitatively (90–100%) (Figure 6), and this is diagnostic of isocyanide complexes such as [Mo2(cp)2(CO)(CNR9)(m-SR)2] (R9 = But, xylyl or CH2Ph; R = Me).2 In addition, the 13C NMR data and the electrochemical behaviour of the MeNC derivative are similar to those of the authentic isocyanide complex prepared from syn-1d and 1 equivalent ButNC (Tables 1, 2 and Experimental section).These data demonstrate unambiguously that [Mo2(cp)2(CO)(CN)(m-SPh)2]2 syn-2d undergoes methylation preferentially at the cyanide nitrogen atom. The product formed by successive additions of CN2 and Me3O1 to trans,anti-[Mo2(cp)2(CO)2(m-SPh)2] anti-1d was generated in the CV cell only. The formation of a S-methylated compound as the major product is demonstrated by its electrochemical behaviour, which is similar to that of the other [Mo2(cp)2(CO)(CN){m-S(Me)R}(m-SR)] compounds described above [R = Me, Pri; Figs. 4, 5(a) and 5(b), Table 2].This is also confirmed by a comparison of the redox potentials of the methylated complex with those of [Mo2(cp)2(CO)(CN)- {m-S(Me)Ph}(m-SPh)] prepared from the addition of 1 equivalent cyanide to [Mo2(cp)2(CO)2{m-S(Me)Ph}(m-SPh)]1, obtained according to a procedure analogous to that described in Scheme 4.The origin of a syn/anti dependence of the Fig. 6 Cyclic voltammetry of trans,syn-[Mo2(cp)2(CO)(CNMe)(m- SPh)2] (a) under Ar and (b) after flushing solution (a) with CO (CV under CO) (MeCN–[NBu4][PF6] electrolyte; vitreous carbon electrode, n = 0.2 V s21)2286 J. Chem. Soc., Dalton Trans., 1997, Pages 2279–2291 Scheme 8 methylation site will be discussed in more detail below (Discussion). The product resulting from the addition of Me3O1 to [Mo2(cp)2(CO)(CN)(m-SCF3)2]2 anti-2e displays spectroscopic [13C NMR: d 181.8 (CN), Table 1] and electrochemical (Table 2) characteristics, as well as the reactivity with CO (Scheme 8) typical of an isocyanide complex.The retention of the anti orientation of the CF3 groups is evidenced by 19F NMR spectroscopy (Table 1). Furthermore, the assignment of the reaction product as [Mo2(cp)2(CO)(CNMe)(m-SCF3)2] is supported by a comparison of the electrochemical and spectroscopic data for the above complex with those of an isocyanide derivative obtained from the reaction of trans-anti-[Mo2(cp)2(CO)2- (m-SCF3)2] with ButNC [13C NMR, d 176.6 (CN); IR, n(CN) 2110 cm21; Tables 1, 2].The alkylation of the syn isomers of [Mo2(cp)2(CO)(CN)- (m-SCF3)2]2 syn-2e and -2e9 also produces syn isomers of an isocyanide complex, characterized spectroscopically (Table 1). Again, the CN resonance in the 13C NMR spectrum (d 177.0) and the infrared band [n(CN) 2140 cm21] establish that methylation occurred at the CN ligand. This is supported by a comparison of these data with those of the single tert-butyl isocyanide complex obtained from trans,syn-[Mo2(cp)2(CO)2- (m-SCF3)2] with ButNC [d 168.7 (CN), n(CN) 2125 cm21, Tables 1, 2].Cyclic voltammetric monitoring of the reaction of the mixture of complexes syn-2e and -2e9 with Me3O1 shows that the initially major isomer of the isocyanide complex, 4a, reduces reversibly in two separate one-electron steps and undergoes an irreversible oxidation (Table 2).The 1H NMR spectrum of 4a shows two cp signals with a small separation [Dd(cp) = 0.06 ppm, Table 1]. A significant amount of a second isomer, 4b, is also present. The latter is characterized by its redox potentials (Table 2) and by a wider separation of the cp signals with Dd(cp) = 0.53 ppm (Table 1). Upon treatment of the solution with CO the dicarbonyl precursor is regenerated in 70–80% yield, and a 4a æÆ 4b isomerization is induced as shown by the increase in the oxidation peak due to 4b at 10.13 V.Flushing the solution with N2 converts the dicarbonyl complex into the initially minor isomer 4b whereas 4a is not restored. A similar irreversible CO-induced conversion has been observed 2 for two syn isomers of [Mo2(cp)2(CO)(CNBut)- (m-SMe)2], the 1H NMR spectra of which also showed cp separation [Dd(cp) = 0.06 and 0.52 ppm] very similar to those of 4a and 4b. Therefore, there exist two types of [Mo2(cp)2(CO)- (CNR9)(m-SR)2] complexes with a mutually syn orientation of the R groups. Complexes of the first type, by far the more frequent (R = Me, R9 = But, xylyl or CH2Ph;2 R = Pri, R9 = But; R = Ph, R9 = Me or But; R = CF3, R9 = Me 4b or But), are characterized by a reversible displacement of the isocyanide ligand by CO.Only two complexes of the second type, characterized by the irreversible substitution of R9NC by CO, have been obtained so far (R = Me, R9 = But;2 R = CF3, R9 = Me 4a).It should be noted that the latter are not formed in the reaction of trans,syn-[Mo2(cp)2(CO)2(m-SR)2] with isocyanides but are obtained from the cis isomer of this complex (R = Me, R9 = But),2 or via the reaction of a cyanide complex with Me3O1 (R = CF3, R9 = Me 4a, see above). In order to establish the geometry of the isocyanide complexes of the first type mentioned above, we undertook an X-ray structural determination on a crystal of 4b. Crystal structure of [Mo2(cp)2(CO)(CNMe)(m-SCF3)2], isomer 4b Although trans-[Mo2(cp)2(CO)2(m-SR)2] complexes have been known for about twenty years,5,7 their reactivity had not been studied until recently,2,6,13 and [Mo2(cp)2(CO)(CNMe)(m-SCF3)2] is the first isocyanide complex derived from the above dicarbonyl to be characterized by X-ray crystallography.A perspective view of the molecule 4b is shown in Fig. 7, and bond lengths and angles are listed in Table 3. The CO and CNMe ligand are trans with respect to the Mo2S2 core, with the isocyanide ligand situated on the same side of the core as the syn- CF3 groups. Therefore, the trans,syn geometry of the dicarbonyl precursor is maintained in the isocyanide complex.The Mo]Mo separation, 2.597(1) Å, is consistent with a metal– metal double bond which confers a closed-shell configuration on the metal centres and is similar to that found in other trans- [Mo2(cp)2(CO)2(m-SR)2] complexes,7,10,11 the principal molecular dimensions of which are in Table 4.A comparison of the M]C(O) and C]O bonds in these complexes suggests a more important dp–p*(CO) back bonding in 4b. This is consistent with the spectroscopic data [n(CO) and d(13CO)] for the different complexes. As in the structure of trans,syn-[Mo2(cp)2(CO)2(m-SBut)2],7 there is a pseudo-mirror plane of symmetry in the molecule of complex 4b; the carbonyl and isocyanide ligands lie in this plane, each of the cp rings lies astride the plane, and the two SCF3 groups are closely related by it.Whereas, however, the cp rings are approximately staggered in orientation with respect to each other in the dicarbonyl complex, they are almost eclipsed in 4b. The pseudo-symmetry does not extend beyond the Fig. 7 View of the molecule trans,syn-[Mo2(cp)2(CO)(CNMe)(m- SCF3)2], isomer 4b, showing the atomic numbering schemeJ. Chem. Soc., Dalton Trans., 1997, Pages 2279–2291 2287 molecule, however; intermolecular contacts are at normal van der Waals distances except perhaps for two shorter F ? ? ? H contacts at 2.38 and 2.49 Å to hydrogen atoms of cp rings.The trans,anti complexes in Table 4 adopt a centrosymmetrical arrangement, with a planar Mo2S2 core. The trans,syn Table 3 Selected bond lengths (Å) and angles (8) in trans,syn- [Mo2(cp)2(CO)(CNMe)(m-SCF3)2], isomer 4b with estimated standard deviations (e.s.d.s) in parentheses a (a) About the molybdenum atoms Mo(1)]Mo(2) Mo(1)]S(1) Mo(1)]S(2) Mo(1)]C(11) Mo(1)]C(12) Mo(1)]C(13) Mo(1)]C(14) Mo(1)]C(15) Mo(1)]C(16) Mo(1)]C(1x) Mo(2)]Mo(1)]S(1) Mo(2)]Mo(1)]S(2) Mo(2)]Mo(1)]C(16) Mo(2)]Mo(1)]C(1x) S(1)]Mo(1)]S(2) S(1)]Mo(1)]C(16) S(1)]Mo(1)]C(1x) S(2)]Mo(1)]C(16) S(2)]Mo(1)]C(1x) C(16)]Mo(1)]C(1x) 2.597(1) 2.391(2) 2.381(2) 2.311(7) 2.317(7) 2.333(7) 2.361(7) 2.341(7) 1.920(6) 2.02 57.3 b 56.9 b 77.9(2) 168.0 113.7(1) 87.5(2) 122.5 86.2(2) 120.1 114.0 Mo(2)]S(1) Mo(2)]S(2) Mo(2)]C(21) Mo(2)]C(22) Mo(2)]C(23) Mo(2)]C(24) Mo(2)]C(25) Mo(2)]C(26) Mo(2)]C(2x) Mo(1)]Mo(2)]S(1) Mo(1)]Mo(2)]S(2) Mo(1)]Mo(2)]C(26) Mo(1)]Mo(2)]C(2x) S(1)]Mo(2)]S(2) S(1)]Mo(2)]C(26) S(1)]Mo(2)]C(2x) S(2)]Mo(2)]C(26) S(2)]Mo(2)]C(2x) C(26)]Mo(2)]C(2x) 2.397(2) 2.378(2) 2.257(8) 2.310(9) 2.369(10) 2.326(10) 2.263(8) 2.069(6) 1.99 57.0 b 57.0 b 101.8(2) 148.1 113.6(1) 93.5(2) 121.3 92.5(2) 117.9 110.1 (b) In the CF3S2 ligands S(1)]C(1) C(1)]F(11) C(1)]F(12) C(1)]F(13) Mo(1)]S(1)]Mo(2) Mo(1)]S(1)]C(1) Mo(2)]S(1)]C(1) S(1)]C(1)]F(11) S(1)]C(1)]F(12) F(11)]C(1)]F(12) S(1)]C(1)]F(13) F(11)]C(1)]F(13) F(12)]C(1)]F(13) 1.821(7) 1.315(8) 1.350(10) 1.299(10) 65.7 b 110.0(2) 113.3(3) 115.1(5) 111.4(6) 104.5(6) 110.5(6) 108.4(7) 106.4(6) S(2)]C(2) C(2)]F(21) C(2)]F(22) C(2)]F(23) Mo(1)]S(2)]Mo(2) Mo(1)]S(2)]C(2) Mo(2)]S(2)]C(2) S(2)]C(2)]F(21) S(2)]C(2)]F(22) F(21)]C(2)]F(22) S(2)]C(2)]F(23) F(21)]C(2)]F(23) F(22)]C(2)]F(23) 1.836(7) 1.314(8) 1.316(9) 1.321(10) 66.1 b 110.5(2) 113.4(3) 114.7(5) 111.3(6) 107.4(6) 108.7(5) 106.6(7) 107.9(7) (c) In the carbonyl ligand C(16)]O(17) 1.167(8) Mo(1)]C(16)]O(17) 173.7(7) (d) In the MeNC ligand C(26)]N(27) N(27)]C(28) 1.133(8) 1.412(8) Mo(2)]C(26)]N(27) C(26)]N(27)]C(28) 177.2(5) 178.2(7) a C(1x) and C(2x) are the centroids of the cp rings.b The e.s.d. is less than 0.058. complexes, however, normally have a puckered core with the S atoms displaced towards the side of the thiolate ligands, and the degree of non-planarity depending principally on steric factors within the molecule.In trans-[Mo2(cp)2(CO)(PMe3)(m-SMe2)2]- [BF4]2,13 with bridging thioether ligands, interactions between the ligands appear minimal and the puckering of the core is significant; we measure the puckering by (i) the mean (absolute) torsion angle in the Mo2S2 ring, here 7.98, and (ii) the mean (absolute) displacement from the mean plane through the four atoms, here 0.076 Å. In trans,syn-[Mo2(cp)2(CO)2(m-SBut)2] there are short contacts around both carbonyl C atoms and the butyl groups appear forced outwards from the central pseudo-mirror plane; there is very little puckering here, with a mean torsion angle of 1.68 and displacement 0.015 Å.In our complex the interligand distances are marginally longer than in the dicarbonyl complex, i.e. a sterically slightly less crowded molecule, and the core is more puckered with a mean torsion angle of 7.58 and displacement 0.071 Å. Another distinctive feature in the thiolate complexes is the mean S ? ? ? S]C angle: in the dicarbonyl complex, this is 124.7(5)8 whereas in 4b there is a tight folding back of the SCF3 ligands towards the central plane and a mean S ? ? ? S]C angle of 112.2(2)8. Discussion Electronic control of the site of methylation by the sulfur substituents: orbital vs.charge control of the reaction The results reported above demonstrate that the nature of the substituents on the bridging sulfur atoms is crucial to the orientation of the reaction of Me3O1 with the cyanide complex [Mo2(cp)2(CO)(CN)(m-SR)2]2.The steric effect of the R groups can prevent the formation of the cyanide complex (R = But), but the orientation of the methylation reaction of [Mo2(cp)2- (CO)(CN)(m-SR)2]2 is not modified by the size of the sulfur substituents for R = Me or Pri: in both cases, methylation occurs at a sulfur lone pair. On the contrary, the electronic properties of R affect the site of methylation: electron-withdrawing substituents such as CF3 (and to a lesser extent Ph) promote alkylation at the cyanide ligand.This is consistent with the reaction being switched from orbital to charge control upon substitution of R = CF3 for Me. The syn/anti arrangement of the sulfur substituents has no effect on the site of methylation of the cyanide complexes, except when R = Ph. In this case, an isocyanide complex is obtained as the major product from the syn isomer, whereas the anti isomer is preferentially methylated at sulfur.Again, the difference may have an electronic origin. The redox potentials in Table 2 show that the syn isomer is harder to oxidize than the anti analogue by 80 mV: the energy gap between the LUMO of the electrophile and the HOMO of the complex increases on going from the anti to the syn isomer. Provided the sulfur lone pairs experience a similar shift in energy, this stabilization might be sufficient to switch the alkylation from an Table 4 Comparison of selected crystallographic dimensions of [Mo2(cp)2(CO)(Y]] ] Z)(m-SR)2] complexes Complex M]M/Å M]C(O)/Å M]S/Å M]S]M/8 S]M]S/8 Ref.trans,syn-[Mo2(cp)2(CO)2(m-SBut)2] trans,anti-[Mo2(cp)2(CO)2(m-SPh)2] trans,anti-[W2(cp)2(CO)2(m-SCHMe2)2] trans,syn-[Mo2(cp)2(CO)(CNMe)(m-SCF3)2] 2.616(2) 2.616(3) 2.569(6) 2.602(5) 2.597(1) 1.946(3) 1.930(4) 1.95(1) 1.920(6) 2.422(2) 2.415(4) 2.420(4) 2.413(2) 2.422(5) 2.413(5) 2.405(8) 2.42(1) 2.412(4) 2.411(3) 2.391(2) 2.381(2) 2.397(2) 2.378(2) 65.43(5) 65.64(5) 65.4(1) 65.8(1) 64.4(2) 65.3(1) 66.1(1) 65.7(1) 114.51(6) 114.37(6) 114.5(1) 114.7(1) 115.6(2) 114.7(1) 113.7(1) 113.6(1) 7 10 10 11 This work2288 J.Chem. Soc., Dalton Trans., 1997, Pages 2279–2291 orbital (anti) to a charge control (syn). The borderline between the two types of control of the methylation of the cyanide complexes by Me3O1 would thus be situated in the range where the oxidation potential is between 20.85 and 20.77 V (Fig. 8). Pertinent to this point, it has been demonstrated that the electronic properties of the sulfur substituents affect the reaction of the [Mo2(cp)2(m-S2CH2)(m-SMe)(m-SR)] and [Mo2(cp)2- (m-S2CH2)(m-S2CRR9)] complexes with protons.14 Whereas the reaction led to a one-electron oxidation of [Mo2(cp)2(m-S2CH2)- (m-SMe)(m-SR)] for R = Me or Pri, it resulted in rapid S]C bond cleavage in the complexes where R = C4H3S, CH(Me)Ph or CH2Ph.14a In the case of [Mo2(cp)2(m-S2CH2)(m-S2CRR9)] a 110 mV shift of the oxidation potentials is enough to switch the reaction with H1 from a one-electron oxidation (R = R9 = Me) to a C]S bond cleavage (R = H, R9 = CO2Me).14b For these examples, the concept of orbital vs.charge control could still account for the orientation of the reaction with H1: oneelectron oxidation suggested that protonation initially occurred at the metal centre.14 This is consistent with an orbitally controlled reaction since the HOMO of the complexes is mainly metal in character.15 Substitution of electron-releasing by withdrawing groups at the bridging sulfur will stabilize the HOMO and therefore increase the gap between this orbital and the electrophile’s LUMO: this should favour a charge control of the reaction. The cleavage of S]C bonds suggest that the reaction of complexes possessing electron-withdrawing substituents with H1 could be charge controlled since these substituents stabilize a carbanionic character in the thiolate carbon.14 Other examples showing the electronic influence of ligands on the site of protonation or alkylation include [Fe(CO)3L- (SR)]2 and [Fe(CO)2L2(SR)]2 complexes [L = CO or P(OEt)3; R = alkyl or aryl],16 and the protonation of [FeH(CN)(R2PCH2- CH2PR2)2] (R = Et, Ph or p-tolyl).17 Site of attack of trans-[Mo2(cp)2(CO)2(Ï-SR)2] by CN2 and RNC Concerning the geometry of the cyanide and isocyanide complexes, the following information is available.(i) There are two types of isocyanide and cyanide complexes having a syn arrangement of the sulfur substituents, while only one for the anti analogue. This is due to the fact that in trans,syn-[Mo2(cp)2(CO)2(m-SR)2] the two faces of the Mo2S2 core, e.g. the two metal sites, are not equivalent; this can lead to isomers A and B, which differ by the situation of the Y]] ] Z ligand with respect to the R groups (Scheme 9), whereas in the trans,anti derivative, isomers A and B are identical.(ii) The more common isocyanide and cyanide complexes with syn R groups are obtained in the reaction of the substrate with trans,syn-[Mo2(cp)2(CO)2(m-SR)2], which is regenerated Fig. 8 Schematic representation of the relationship between E1/2 ox of trans-[Mo2(cp)2(CO)(CN)(m-SR)2]2 complexes and the site of methylation by Me3O1 by treating the cyanide or isocyanide product, [Mo2(cp)2- (CO)(Y]] ] Z)(m-SR)2], with CO; although synthesized differently, 4b and [Mo2(cp)2(CO)(CNMe)(m-SPh)2], which show the same reactivity and similar electrochemical and spectroscopic characteristics as those of the authentic isocyanide complexes, belong to this group.Therefore, all these compounds can be assigned the same geometry as that of the structurally characterized 4b (A in Scheme 9). (iii) The uncommon isocyanide and cyanide complexes (syn) are not obtained from trans,syn-[Mo2(cp)2(CO)2(m-SR)2]. They undergo an irreversible reaction with CO, which produces this dicarbonyl complex; consistent with (ii), flushing the solution with dinitrogen or argon leads to the common isocyanide and cyanide complexes {see Scheme 3, the 4a æÆ 4b conversion, and the similar reaction observed with [Mo2(cp)2(CO)(CNBut)- (m-SMe)2]}.2 There are only four examples of the uncommon cyanide and isocyanide complexes so far: one isomer of [Mo2(cp)2(CO)(CNBut)(m-SMe)2],2 syn-2a9, 4a and its cyanide precursor syn-2e9.The first two were produced in the reaction of the substrate Y]] ] Z (ButNC or CN2) with the electrogenerated cis-[Mo2(cp)2(CO)2(m-SMe)2] 2,8 and were assigned a trans geometry with the Y]] ] Z ligand opposite to the Me groups (B in Scheme 9). This is consistent with the fact that the reaction of cis-[Mo2(cp)2(CO)2(m-SMe)2] with Y]] ] Z = CO produces the trans isomer.2 Therefore, the trans,syn-[Mo2(cp)2(CO)2(m-SR)2] complexes appear to favour attack of Y]] ] Z on the same side of the Mo2S2 plane as the R groups (see I in Scheme 10), that is on the more crowded face of the complex, which leads to an A-type derivative.The attack shown in II (Scheme 10), which would produce a compound with a B geometry, might be avoided because of a repulsion of the nucleophile by the sulfur lone pairs; this is consistent with the fact that when the attack according to I is sterically hindered (R = But) no cyanide or isocyanide complex is formed at all. However, when R is CF3 and Y]] ] Z = CN2 the approach as I appears to be less favourable since the precursor of complex 4b (e.g.a cyanide complex like A) is initially the minor product (ca. 35%) formed in the reaction of syn-1e with CN2. This might be due to the fact that, in this case, the substrate attacking either side of the Mo2S2 plane experiences repulsions, by the fluoride ligands or by the sulfur lone pairs. On the contrary, when Y]] ] Z = CNR, the single product has a A-type geometry. Therefore, the trans,syn-[Mo2(cp)2(CO)2(m-SCF3)2] complex is able to discriminate between cyanide and isocyanide substrates. This is yet another type of effect of the sulfur substituents upon the reactivity of trans-[Mo2(cp)2(CO)2(m-SR)2] complexes.Scheme 9 Scheme 10J. Chem. Soc., Dalton Trans., 1997, Pages 2279–2291 2289 Conclusion The results reported in this paper lead to the conclusions summarized below. (1) Our approach to methylate the cyanide ligand rather than the sulfur lone pairs of trans-[Mo2(cp)2(CO)(CN)(m-SR)2]2 was based on the concept of orbital vs.charge control of a reaction. 9 We have shown that the R groups have a pronounced influence on the reactivity of the trans-[Mo2(cp)2(CO)(Y]] ] Z)- (m-SR)2] complexes (Y]] ] Z = CO or CN2). Their steric bulk can prevent the reaction of the complex with cyanide or isocyanide (R = But, Y]] ] Z = CO), but more importantly their electronic properties can lead to a completely different orientation of the reaction of the cyanide complex with a methylating agent (R = Me or Pri, S-methylation; CF3, N-methylation), which is consistent with the above concept.In the case of the methylation reaction of trans-[Mo2(cp)2(CO)(CN)(m-SPh)2]2 this is even more spectacular, since the anti isomer is methylated at sulfur, whereas the syn isomer produces an isocyanide complex. The present results demonstrate that the redox potentials of the complexes are largely affected by the electronic properties of the R groups, since the different substituents used allow the oxidation potentials to be varied in a range extending over 0.58 V.Therefore, the sulfur substituents exert a control of the reactivity of the complexes via their effect on the redox potentials, and consequently the choice of particular R groups can allow a selective activation of a given site in a complex which has several potentially reactive centres. As the actual site of attack of trans-[Mo2(cp)2(CO)(CN)(m-SR)2]2 by an electrophile depends on the energy gap between the HOMO of the complex and the LUMO of the electrophile, it is not certain that other electrophiles would react with trans-[Mo2(cp)2(CO)(CN)- (m-SR)2]2 as Me3O1 was shown to do.(2) The sulfur substituents make possible a recognition of CN2 and RNC by trans,syn-[Mo2(cp)2(CO)2(m-SCF3)2] since the favoured site of attack of the complex by these substrates is different. For the other trans,syn-[Mo2(cp)2(CO)2(m-SR)2] complexes this discrimination is not observed.(3) The displacement of the isocyanide ligand by CO is typical of all the complexes trans-[Mo2(cp)2(CO)(CNR9)(m-SR)2] (R9 = But or xylyl, R = Me; R9 = But, R = Pri, Ph or CF3; R9 = Me, R = Ph or CF3) studied so far. It appears to be a convenient way to distinguish an isocyanide complex from its S-alkylated isomer [Mo2(cp)2(CO)(CN){m-S(R9)R}(m-SR)]. Furthermore, whether this reaction is reversible or not is a criterion to assign the complex a geometry of type B or A, respectively.Experimental Methods and materials All the experiments were carried out under an inert atmosphere, using Schlenk techniques for the syntheses. Tetrahydrofuran (thf ) was purified as described previously.8 Acetonitrile (Carlo Erba or BDH, HPLC grade) was used as received. The preparation and the purification of the supporting electrolyte [NBu4][PF6] and the electrochemical equipment were as described previously.8 Infrared spectra were obtained with a Perkin-Elmer 1430 and 1H and 13C NMR spectra on a Bruker AC300 spectrophotometer.Shifts are relative to tetramethylsilane as internal reference. The mass spectra were obtained on a GC/MS Hewlett-Packard 5995C apparatus. Chemical analyses were performed by the Centre de Microanalyses du CNRS, Vernaison. The complexes [Mo2(cp)2(CO)4],18 [Mo2(cp)2(CO)2- (m-SMe)(m-SMe2)][BF4],13 [Mo2(cp)2(CO)(CN)(m-SMe)- (m-SMe2)] 3a 8 and [Mo2(cp)2(CO)3(m-SMe3)][PF6] 19 were synthesized as described in the literature.Syntheses The complexes trans-[Mo2(cp)2(CO)2(m-SR)2] (R = Me, But or Ph) have been obtained following different routes.5,7,10 We prepared them by reaction of [Mo2(cp)2(CO)6] with the disulfides for R = Me and Ph, or with the thiol for R = But. trans-[Mo2(cp)2(CO)2(Ï-SR)2]. R = Pri. A toluene (30 cm3) solution of [Mo2(cp)2(CO)6] (1 g, 2.04 mmol) was heated (110 8C) in the presence of an excess of PriSSPri (2–3 equivalents) for 48 h.The solution was taken to dryness and the residue dissolved in CH2Cl2–hexane (1 : 4) and chromatographed on a silica gel column. The syn and anti isomers were eluted with CH2Cl2–hexane (3 : 7). The isomers were separated by recrystallization from cold acetonitrile (yield: 55%): syn-1b, brown; anti-1b, green (Found: C, 41.3; H, 4.6; S, 10.9. Calc. for C18H24Mo2O2S2: C, 40.9; H, 4.6; S, 12.15%). Mass spectrum: m/z 528, M1; 485, [M 2 Pri]1; 457, [M 2 Pri 2 CO]1; 386, [M 2 2Pri 2 2CO]1; and 321, [M 2 2Pri 2 2CO 2 cp]1.R = Me or Ph. The same procedure as described above was followed. The reaction was complete after ca. 2 h (yield 70%). trans-[Mo2(cp)2(CO)2(m-SPh)2] syn-1d: mass spectrum: m/z 596, M1; 568, [M 2 CO]1; 540, [M 2 2CO]1 and 386, [M 2 2CO 2 2Ph]1. R = But, syn-1c. A solution of [Mo2(cp)2(CO)6] (0.6 g, 1.22 mmol) in toluene (50 cm3) was heated at 100 8C in the presence of an excess of ButSH and NEt3 for 72 h under vacuum.The solution was taken to dryness and the residue dissolved in CH2Cl2–hexane (3 : 17) and chromatographed on a silica gel column. The syn isomer was eluted with CH2Cl2–hexane (1 : 4). A small amount of the anti isomer was eluted with CH2Cl2– hexane (1 : 3) (total yield ca. 30%). syn-1c: green; mass spectrum: m/z 556, M1; 499, [M 2 But]1; 471, [M 2 But 2 CO]1; 443, [M 2 But 2 2CO]1 and 386, [M 2 2But 2 2CO]1. R = CF3, syn- and anti-1e. A solution of [Mo2(cp)2(CO)4] (1.6 g, 3.74 mmol) in toluene (40 cm3) was heated (110 8C) in the presence of an excess of CF3SSCF3 (2–3 equivalents) for 72 h. The solution was taken to dryness and the residue dissolved in the minimum volume of CH2Cl2 and chromatographed on a silica gel column.The syn isomer (red-brown) was eluted with CH2Cl2–hexane (1 : 4) and the anti isomer (green) with CH2Cl2– hexane (3 : 7) (total yield: 50%) (Found: C, 29.3; H, 1.9. Calc. for C14H10F6Mo2O2S2: C, 29.0; H, 1.75%). syn-1e: mass spectrum: m/z 580, M1; 511, [M 2 CF3]1; 483, [M 2 CF3 2 CO]1; 455, [M 2 CF3 2 2CO]1 and 386, [M 2 2CF3 2 2CO]1. The anti isomer gave exactly the same spectrum.K[trans-Mo2(cp)2(CO)(CN)(Ï-SR)2] (R = Me or CF3) and [NBu4][trans-Mo2(cp)2(CO)(CN)(Ï-SR)2] syn-2d (R = Ph). To a solution of trans,syn-[Mo2(cp)2(CO)2(m-SR)2] (R = Me, 0.32; CF3, 0.25 mmol) in acetonitrile (30 cm3) was added 1 equivalent KCN in a minimum volume of water–MeCN (1 : 5) at room temperature. The solution instantly turned from brownish to orange.After 15 min of stirring the solvent was evaporated under vacuum and the residue extracted by acetonitrile (20 cm3). The volume of the solution was reduced to 5 cm3 and an equivalent volume of diethyl ether was added to precipitate the complex. After filtration, the residue was stirred with pentane (3 × 5 cm3). syn-2a: orange solid, yield 80%. syn-2e and -2e9: orange solid, yield 80%; the mass spectral peak corresponding to the complex anion (m/z 578, Ma) was not observed, others at m/z 455 ([Ma 2 CO 2 CN 2 CF3]1), 386 ([Ma 2 CO 2 CN 2 2CF3]1) and 321 ([Ma 2 CO 2 CN 2 2CF3 2 cp]1).A similar procedure was followed to obtain the anti analogue of the SCF3 derivative (orange solid, yield 80%), and also the SPh derivative, except that in this case [NBu4][CN] was used in place of KCN. syn-2d: orange solid, yield 80%; the mass spectral peak corresponding to the complex anion (m/z 594, Ma) was not observed, others at m/z 540 ([Ma 2 CO 2 CN]1), 462 ([Ma 2 CO 2 CN 2 C6H6]1), 386 ([Ma 2 CO 2 CN 2 2Ph]1) and 321 ([Ma 2 CO 2 CN 2 2Ph 2 cp]1).The derivatives syn- and anti-2b were generated in situ in the electrochemical cell only. Cyclic voltammetric monitoring of a2290 J. Chem. Soc., Dalton Trans., 1997, Pages 2279–2291 solution of trans,syn-[Mo2(cp)2(CO)2(m-SBut)2] in the presence of cyanide showed that no reaction occurred. The cyanide complexes were found unstable in the solid state under nitrogen or argon and did not give satisfactory microanalyses. [Mo2(cp)2(CO)(CN){Ï-S(Me)R} (Ï-SR)] (R = Pri or Ph).To a solution of trans,syn-[Mo2(cp)2(CO)2(m-SPri)2] (0.12 g, 0.23 mmol) in MeCN (25 cm3) was added 1 equivalent KCN in a minimum volume of water–MeCN (1 : 5). After 15 min of stirring at room temperature, 1 equivalent [OMe3][BF4] (0.034 g) in MeCN (10 cm3) was added. An instant colour change was observed. After the solvent was evaporated under vacuum, the residue was extracted by thf (2 × 10 cm3) in order to remove K[BF4].The volume of the solution was reduced to 4 cm3 and pentane (8 cm3) was added to precipitate complex 3b. After filtration, the residue was washed several times with pentane. Orange-brown solid, yield 60%. NMR (CDCl3): 1H, d 5.49 (s, 5 H, cp), 4.85 (s, 5 H, cp), 3.25 (m, 1 H, J = 6.9), 2.88 (s, 3 H, SMe), 2.65 (m, 1 H, J = 6.9), 1.72 (d, 3 H, J = 6.9, CH3), 1.62 (d, 3 H, J = 6.9, CH3), 1.48 (d, 3 H, J = 6.7, CH3) and 1.38 (d, 3 H, J = 6.7 Hz, CH3); 13C-{1H}, d 233.3 (CO), 150.6 (CN), 89.2, 88.9 (cp), 57.7, 50.7 [CH(CH3)2], 34.2 S(CH3), 25.5, 24.5, 20.2, 19.8 [CH(CH3)2].IR(CH2Cl2, cm21): 2070 (CN) and 1835 (CO). FAB mass spectrum: m/z 541, M1. Complex 3d was obtained by addition of 1 equivalent KCN to [Mo2(cp)2(CO)2{m-S(Me)Ph}(m-SPh)][BF4] prepared as follows. To a solution of [Mo2(cp)2(CO)2(m-SPh)2] (0.22 g, 0.37 mmol) in MeCN (70 cm3) was added [OMe3][BF4] (0.54 g, 1 equivalent). After 4 h of stirring at room temperature the volume of the solution was reduced to 10 cm3 and ether (ca. 20 cm3) was added to precipitate the product. The solution was filtered and the product washed several times with ether and with pentane. Yield 60% (Found: C, 42.7; H, 3.5; S, 8.8. Calc. for C25H23BF4Mo2O2S2: C, 43.0; H, 3.3; S, 9.2%). Mass spectrum: the peak corresponding to the complex anion (m/z 611, Ma) was not observed; other peaks at m/z, 540 ([Ma 2 2CO 2 Me]1), 462 ([Ma 2 2CO 2 C6H6]1) and 386 ([Ma 2 2CO 2 Me 2 2Ph]1).To a solution of [Mo2(cp)2- (CO)2{m-S(Me)Ph}(m-SPh)][BF4] (0.0824 g, 0.11 mmol) in MeCN (50 cm3) was added KCN (0.077 g, 1 equivalent). An instant change from greenish brown to dark orange was observed. The solution was taken to dryness and the residue stirred with thf and filtered to remove K[BF4]. The volume of the thf extract was reduced to 4 cm3 and pentane (8 cm3) was added. The solution was filtered and the solid washed several times with pentane.Yield ca. 60%. Mass spectrum: m/z 609, M1; 540, [M 2 CO 2 CN 2 Me]1; 463, [M 2 CO 2 CN 2 Me 2 Ph]1 and 386, [M 2 CO 2 CN 2 Me 2 2Ph]1. trans-[Mo2(cp)2(CO)(CNMe)(Ï-SR)2] (R = Ph or CF3). To a solution of trans,syn-[Mo2(cp)2(CO)2(m-SR)2] (0.33 mmol) in MeCN (20 cm3) was added 1 equivalent KCN in a minimum volume of water–MeCN (1 : 5). After 10 min of stirring at room temperature, 1 equivalent [OMe3][BF4] was added. After the solvent was removed under vacuum, the isocyanide complex formed was extracted by ether (3 × 10 cm3) and washed by cold pentane (2 × 5 cm3).The two isomers of trans,syn- [Mo2(cp)2(CO)(CNMe)(m-SCF3)2] were separated by column chromatography: the mixture of isomers was dissolved in the minimum volume of CH2Cl2 and 4b (syn isomer, garnet-red solid) was eluted with CH2Cl2–hexane (1 : 3) and 4a (syn isomer, brown-yellow solid) with CH2Cl2–hexane (3 : 7), yield 60%. 4b (Found: C, 30.5; H, 2.2; N, 2.4. Calc. for C15H13F6Mo2NOS2: C, 30.35; H, 2.2; N, 2.35%): mass spectrum: m/z 593, M1; 578, [M 2 Me]1; 524, [M 2 CF3]1 or [M 2 CO 2 MeNC]1; 455, [M 2 2CF3]1 or [M 2 CO 2 MeNC 2 CF3]1; 386, [M 2 CO 2 MeNC 2 2CF3]1 and 321 [M 2 CO 2 MeNC 2 2CF3 2 cp]1.R = Ph, syn isomer: garnet-red solid, yield 60%. A similar procedure was followed to obtain the anti isomer of trans-[Mo2(cp)2(CO)(CNMe)(m-SCF3)2] from trans,anti-[Mo2- (cp)2(CO)2(m-SCF3)2] as a green solid in 60% yield. trans-[Mo2(cp)2(CO)(CNBut)(Ï-SR)2] (R = Pri, Ph or CF3).To a solution of trans-[Mo2(cp)2(CO)2(m-SR)2] (R = Pri syn-1b, Ph syn-1d, CF3 syn- and anti-1e) (0.1 g) in MeCN (20 cm3) was added 1 equivalent ButNC. An instant reaction was observed. The solvent was removed under vacuum and the residue washed twice with cold pentane (5 cm3). R = Pri, syn isomer: red solid, yield 80%; mass spectrum: m/z 583 (M1), 541 ([M 2 CMe2]1), 455 ([M 2 CO 2 But 2 Pri]1) and 386 ([M 2 CO 2 CNBut 2 2Pri]1). R = Ph, syn isomer: red solid, yield 85–90%; mass spectrum: m/z 651 (M1), 623 ([M 2 CO]1), 540 ([M 2 CO 2 CNBut]1), 463 ([M 2 CO 2 CNBut 2 Ph]1), 386 ([M 2 CO 2 CNBut 2 2Ph]1) and 321 ([M 2 CO 2 CNBut 2 2Ph 2 cp]1).R = CF3, syn isomer: red solid, yield 85–90%; mass spectrum: m/z 635 (M1), 566 ([M 2 CF3]1), 538 ([M 2 CO 2 CF3]1), 455 ([M 2 CO 2 CNBut 2 CF3]1), 412 ([M 2 CO 2 But 2 2CF3]1) and 386 ([M 2 CO 2 CNBut 2 2CF3]1). Anti isomer: green solid, yield 85–90%; mass spectrum: m/z 635 (M1), 566 ([M 2 CF3]1), 538 ([M 2 CO 2 CF3]1), 455 ([M 2 CO 2 CNBut 2 CF3]1), 412 ([M 2 CO 2 But 2 2CF3]1) and 386 ([M 2 CO 2 CNBut 2 2CF3]1).Crystallography Crystal data. C15H13F6Mo2NOS2 4b, M = 593.3, monoclinic, space group P21/n (equivalent to no. 14), a = 15.436(2), b = 8.895(1), c = 14.470(1) Å, b = 92.325(9)8, U = 1985.0(4) Å3, Z = 4, Dc = 1.985 g cm23, F(000) = 1152, m(Mo-Ka) = 14.9 cm21, T = 293 K, l(Mo-Ka) = 0.710 69 Å. Crystals are deep red plates, obtained by recrystallization from pentane at 210 8C.One, ca. 0.10 × 0.43 × 0.83 mm, was mounted on a glass fibre and, after preliminary photographic examination, transferred to an Enraf-Nonius CAD4 diffractometer (with monochromated radiation) for determination of accurate cell parameters (from the settings of 25 reflections, q = 10–118, each centred in four orientations) and for measurement of diffraction intensities (3497 unique reflections to qmax = 258; of these 2627 were ‘observed’ with I > 2sI).During processing: corrections were applied for Lorentzpolarization effects, absorption (by semi-empirical y-scan methods) and to eliminate negative net intensities (by Bayesian statistical methods). No crystal deterioration was observed. The structure was determined by the heavy-atom method using the SHELX 76 program20 and refined on F by full-matrix least-squares methods. Hydrogen atoms on the cp rings were included in idealized positions, those in the methyl group were located in difference maps and refined with geometrical constraints; the isotropic thermal parameters of all were refined freely.The non-hydrogen atoms were allowed anisotropic thermal parameters. Refinement to convergence was rapid, with R = 0.056 and Rg = 0.076 20 for all 3497 reflections, weighted w = (sF2 1 0.00375F2)21. In the final difference map the highest peaks (to ca. 1.0 e Å23) were all close to the molybdenum atoms. Scattering factors for neutral atoms were taken from ref. 21. Computer programs used in this analysis have been noted above or in Table 4 of ref. 22, and were run on a DEC-MicroVAX 3600 machine in the Nitrogen Fixation Laboratory. 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 CCDC for this material should quote the full literature citation and the reference number 186/508. Acknowledgements The CNRS (Centre National de la Recherche Scientifique) and the BBSRC are acknowledged for financial support to thisJ. Chem. Soc., Dalton Trans., 1997, Pages 2279–2291 2291 work. Professors A. Darchen (Ecole Nationale Supérieure de Chimie de Rennes), H. des Abbayes (URA CNRS 322, Brest) and Dr. J. N. Verpeaux (Ecole Normale Supérieure, URA CNRS 1679, Paris) are acknowledged for useful discussions and comments. M.-L. A. is grateful to Ministère de l’Enseignement Supérieur et de la Recherche for providing a studentship. References 1 C. Le Floch, F. Y. Pétillon, C. J. Pickett and J. Talarmin, J. Organomet. Chem., 1990, 390, C39. 2 F. Gloaguen, C. Le Floch, F. Y. Pétillon, J. Talarmin, M. El Khalifa and J. Y. Saillard, Organometallics, 1991, 10, 2004. 3 W. P. Fehlhammer and M. Fritz, Chem. Rev., 1993, 93, 1243. 4 M. L. Abasq, F. Y. Pétillon and J. Talarmin, J. Chem. Soc., Chem. Commun., 1994, 2191. 5 F. Y. Pétillon, J. L. Le Quéré, J. Roué, J. E. Guerchais and D. W. A. Sharp, J. Organomet. Chem., 1981, 204, 207; J. L. Le Quéré, Thesis, Brest, 1982. 6 P. Li and M. D. Curtis, Inorg. Chem., 1990, 29, 1242. 7 I. B. Benson, S. D. Killops, S. A. R. Knox and A. J. Welch, J. Chem. Soc., Chem. Commun., 1980, 1137. 8 M. L. Abasq, F. Y. Pétillon, P. Schollhammer and J. Talarmin, New J. Chem., 1996, 20, 1221. 9 R. F. Fenske and M. C. Milletti, Organometallics, 1986, 5, 1243 and refs. therein. 10 S. E. Nefedov, A. A. Pasynskii, I. L. Eremenko, G. A. Papoyan, L. I. Rubinstein, A. I. Yanoskii and Y. Y. Struchkov, Zh. Neorg. Khim., 1993, 38, 76. 11 A. Shaver, B. Soo Lum, P. Bird, E. Livingstone and M. Schweitzer, Inorg. Chem., 1990, 29, 1832. 12 M. El Khalifa, J. Y. Saillard, F. Gloaguen, C. Le Floch, F. Y. Pétillon and J. Talarmin, New J. Chem., 1992, 16, 847. 13 N. Kuhn, E. Zauder, R. Boese and D. J. Bläser, J. Chem. Soc., Dalton Trans., 1988, 2171. 14 (a) L. L. Lopez, P. Bernatis, J. Birnbaum, R. C. Haltiwanger and M. Rakowski DuBois, Organometallics, 1992, 11, 2425; (b) P. Bernatis, R. C. Haltiwanger and M. Rakowski DuBois, Organometallics, 1992, 11, 2435. 15 D. L. Dubois, W. K. Miller and M. Rakowski DuBois, J. Am. Chem. Soc., 1981, 103, 3429. 16 W.-F. Liaw, C. Kim, M. Y. Darensbourg and A. L. Rheingold, J. Am. Chem. Soc., 1989, 111, 3591; M. Y. Darensbourg, W.-F. Liaw and C. G. Riordan, J. Am. Chem. Soc., 1989, 111, 8051. 17 P. I. Amrhein, S. D. Drouin, C. E. Forde, A. J. Lough and R. H. Morris, Chem. Commun., 1996, 1665. 18 M. D. Curtis and R. J. Klinger, J. Organomet. Chem., 1978, 161, 23. 19 M. B. Gomes de Lima, J. E. Guerchais, R. Mercier and F. Y. Pétillon, Organometallics, 1986, 5, 1952. 20 G. M. Sheldrick, SHELX 76, Program for crystal structure determination, University of Cambridge, 1976. 21 International Tables for X-Ray Crystallography, Kynoch Press, Birmingham, 1974, vol. 4, pp. 99 and 149. 22 S. N. Anderson, R. L. Richards and D. L. Hughes, J. Chem. Soc., Dalton Trans., 1986, 245. Received 6th February 1997; Paper 7/00873B

ISSN:1477-9226    

DOI:10.1039/a700873b

出版商:RSC    

年代:1997

数据来源: RSC

摘要:

DALTON J. Chem. Soc., Dalton Trans., 1997, Pages 2293–2303 2293 Synthesis of fluorophenyl derivatives of iron, molybdenum and tungsten via B(C6F5)3 and unusual carbon–fluorine bond reactions † Alexander N. Chernega, Andrew J. Graham, Malcolm L. H. Green,* Jane Haggitt, Julian Lloyd, Christian P. Mehnert, Nils Metzler and Joanne Souter Inorganic Chemistry Laboratory, South Parks Road, Oxford, UK OX1 3QR The reaction between B(C6F5)3 and [Fe(h-C5H5)(CO)2Me] gave the unexpected product [Fe{C6F4C(O)Me-2}- (h-C5H5)(CO)] 1 which reacts with the donor molecules L = PMe3, PPh3 or ButNC giving [Fe{C6F4C(O)Me-2}- (h-C5H5)L].* Likewise [{Fe[C6F4C(O)Me-2](h-C5H5)}2(m-dppe)]* was formed from 1 and Ph2P(CH2)2PPh2 (dppe).The formation of the compound where L = PMe3 is shown to proceed via initial formation of [Fe{C6F4C(O)- Me-2}(h-C5H5)(CO)(PMe3)]*. The reaction between B(C6F5)3 and [M(h-C5H5)(CO)3Me] (M = Mo or W) gave the compounds [M{C6F4C(O)Me-2}(h-C5H5)(CO)2]. The compounds [M{C6F3X-5-C(O)Me-2}(h-C5H5)(CO)2]* (M = Mo or W) where X = H were prepared from the compounds where X = F via unusual C]F bond-activation reactions.The compound [Fe{C6F3H-5-C(O)Me-2}(h-C5H5)(PMe3)] was prepared by photolysis of a mixture of [Fe{C6F3H-5-C(O)Me-2}(h-C5H5)(CO)] and PMe3. The asterisk indicates the crystal structure has been determined. The strong Lewis acid molecule tris(pentafluorophenyl)boron, B(C6F5)3,1 has recently attracted considerable interest as a cocatalyst in homogeneous Ziegler–Natta catalysis.2 It is readily available and convenient to handle.Thus, unlike BF3 which is a volatile gas and readily hydrolyses giving HF, the compound B(C6F5)3 is a soluble white crystalline solid which is only slowly hydrolysed. We decided to explore its reactions with organotransition- metal compounds which might, in different ways, act as nucleophiles to the boron centre. This has led to unexpected reactions and we have recently reported that with [Fe(h-C5H5)- (CO)2Me] which leads to most unusual products, namely the compound [Fe{C6F4C(O)Me-2}(h-C5H5)(CO)] 1.3 We have also shown that B(C6F5)3 readily co-ordinates to the oxygen of terminal oxometal groups.4 Here we describe further reactions of compound 1 and related studies with molybdenum and tungsten compounds.Results and Discussion The Lewis acid molecule B(C6F5)3 was first prepared many years ago,1 however the two reported syntheses have some disadvantages. The first method employed was to treat a pentane solution of pentafluorophenyl bromide with n-butyllithium at low temperatures to generate Li(C6F5). The latter is thermally highly sensitive and detonates at ca. 220 8C. The resulting suspension of the lithium salt was treated with boron trichloride to form LiCl and B(C6F5)3. The reported yield of the crude product was between 30 and 50%, but in our hands, when working on a several gram scale, the overall yield after purification was much lower at ca. 20%. The compound can also be prepared 1 using the thermally stable Grignard reagent Mg(C6F5)Cl in Et2O and BF3?Et2O. The yield of the compound Et2O?B(C6F5)3 was 88%. This preparation is unsatisfactory due to the difficulty of removal of co-ordinated diethyl ether to give the desired solvent-free molecule B(C6F5)3. It was decided to seek a modified procedure and improved preparation for solvent-free B(C6F5)3. The reaction vessel shown in the Experimental section was designed to enable the addition of the thermally sensitive suspension of Li(C6F5) to a solution of BBr3 at 278 8C.Maintaining the entire reaction vessel at 278 8C is an essential feature of this synthesis.2 This † Non-SI units employed: Torr ª 133 Pa, eV ª 1.60 × 10219 J. method, which is described in detail in the Experimental section, gave pure B(C6F5)3 in ª50% yield on a 20 g scale. The reactions between the compound B(C6F5)3 and the methyl derivatives [Fe(h-C5H5)(CO)2Me], [Mo(h-C5H5)(CO)3- Me] and [W(h-C5H5)(CO)3Me] have been studied.Treatment of a toluene solution of [Fe(h-C5H5)(CO)2Me] with 1 equivalent of B(C6F5)3 gave air-sensitive black-green crystals of the compound [Fe{C6F4C(O)Me-2}(h-C5H5)(CO)] 1 in 79% yield. Similarly, treatment of [Mo(h-C5H5)(CO)3Me] with B(C6F5)3 gives moderately air-sensitive red crystals of [Mo{C6F4C(O)Me- 2}(h-C5H5)(CO)2] 2. The reaction between [W(h-C5H5)(CO)3- Me] and B(C6F5)3 gave purple crystals of [W{C6F4C(O)Me- 2}(h-C5H5)(CO)2] 3.The crystal structures of 2 and 3 have been determined and they are closely similar. The molecular structure of 3 is shown in Fig. 1 and selected distances and angles for 2 and 3 are given in Table 1. A discussion of these structures will be given later in the text. The analytical and spectroscopic data for 1–3 and all the other new compounds described in this work are given in Table Table 1 Selected bond lengths (Å) and angles (8) for the compounds 2, 3, 7 and 13 2 3 7 13 M]C(6) M]C(7) M]C(8) M]O(3) C(14)]O(3) C(10)]F(4) C(11)]F(3) C(12)]F(2) C(13)]F(1) M]Cpcentroid W]P(1) W]C(1) 1.949(4) 2.003(4) 2.172(3) 2.141(2) 1.258(4) 1.347(4) 1.353(4) 1.340(4) 1.357(4) 1.999 1.957(12) 2.003(11) 2.171(10) 2.141(7) 1.283(14) 1.355(12) 1.392(12) 1.366(13) 1.335(12) 2.003 1.950(4) 2.005(4) 2.162(4) 2.145(2) 1.236(4) 1.353(4) 1.344(5) — 1.364(4) 1.990 1.900(14) 2.107(13) 2.078(9) 1.254(12) 2.480(3) 2.390(13) O(3)]W]P(1) C(6)]M]O(3) C(8)]M]O(3) M]O(3)]C(14) O(3)]C(14)]C(9) C(14)]C(9)]C(8) C(9)]C(8)]M C(6)]M]C(7) C(6)]M]C(8) 120.0(1) 73.2(1) 121.4(2) 117.1(10) 113.5(3) 115.5(2) 77.0(1) 78.0(1) 120.9(4) 73.1(3) 120.8(7) 116.3(3) 112.6(10) 116.1(7) 76.9(4) 78.3(4) 119.24(12) 72.00(11) 122.1(2) 116.3(3) 112.6(3) 116.0(3) 76.9(2) 77.82(14) 77.4(2) 118.34 71.8(4) 125.3(9) 114.1(11) 109.7(11) 118.9(9) — 77.4(5)2294 J.Chem. Soc., Dalton Trans., 1997, Pages 2293–2303 Table 2 Analytical and spectroscopic data Fc Ff M C Fe Fd or Hd d c b a f e Compounda NMR data b 1 [Fe{C6F4C(O)Me-2}(h-C5H5)(CO)] Black-green C, 49.8 (49.45); H, 2.45 (2.4) Mass (EI): 340, M1; 312, [M 2 CO]1 (base peak); 297, [M 2 CO 2 CH3]1; 247, [M 2 CO 2 C5H5]1 IR (Nujol): 1968vs, 1630m, 1581m 2 [Mo{C6F4C(O)Me-2}(h-C5H5)(CO)2] Red C, 44.35 (44.1); H, 2.0 (2.0); Mo, 23.2 (23.5) Mass (EI): 413 (410), M1; 385 (382), [M 2 CO]1; 357 (354), [M 2 2CO]1 IR (KBr): 1975vs, 1877vs, 1632s, 1573s, 1504s 1H:c 2.13 (d, 3 H, JFH = 4.2, COCH3), 4.80 (s, 5 H, C5H5) 13C-{1H}:c 27.7 (d, JFC = 8.5, COCH3), 93.5 (s, C5H5), 123.7 (m, Cb), 136.2 (dm, 1JFC = 242, Ce), 137.6 (dm, 1JFC = 252, Cd), 139.2 (dm, 1JFC = 248, Cf), 150.5 (dm, 1JFC = 243, Cc), 183.2 (dm, JFC = 45, Ca), 203.4 (m, COCH3), 246.8 (d, J = 5.0, CO), 253.2 (s, CO) 19F:c 2164.9 (m, 1 F, Fe), 2156.8 (m, 1 F, Fd), 2138.4 (m, 1 F, Ff), 2115.7 (m, 1 F, Fc) 3 [W{C6F4C(O)Me-2}(h-C5H5)(CO)2] Purple C, 36.2 (36.3); H, 1.6 (1.6); W, 38.9 (37.1) Mass (EI): 502 (496), M1 IR (KBr): 1966vs, 1864vs, 1636s, 1567s, 1505s 1H:c 2.80 (d, 3 H, JFH = 3.8, COCH3), 5.07 (s, 5 H, C5H5) 13C-{1H}:c 26.7 (d, JFC = 8.4, COCH3), 92.1 (s, C5H5), 124.3 (m, Cb), 136.0 (dm, 1JFC = 246, Ce), 140.2 (dm, 1JFC = 266, Cd), 150.0 (dm, 1JFC = 263, Cf), 152.2 (dm, 1JFC = 233, Cc), 178.1 (dm, JFC = 40, Ca), 202.7 (pseudo-t, JFC = 6.1, COCH3), 236.6 (d, J = 4.7, CO), 245.6 (s, CO) 19F:c 2165.3 (m, 1 F, Fc), 2157.1 (m, 1 F, Fd), 2140.0 (m, 1 F, Ff), 2116.0 (m, 1 F, Fc) 4 [Fe{C6F4C(O)Me-2}(h-C5H5)(ButNC)] Dark purple C, 54.7 (54.7); H, 4.45 (4.3); Fe, 14.2 (14.1); N, 3.5 (3.5) Mass (FAB1): 395, M1 (base peak) IR (Nujol): 2044s, 1629m, 1560w, 1530m 5 [Fe{C6F4C(O)Me-2}(h-C5H5)(CO)(PMe3)] Black-purple C, 49.3 (49.1); H, 4.2 (4.1); Fe, 13.95 (13.4); P, 7.7 (7.4) IR (Nujol): 1968s, 1627s, 1547m, 1504s 6 [Fe{C6F4C(O)Me-2}(h-C5H5)(PMe3)] Purple Mass (FAB1): 388, M1 (base peak); 323, [M 2 C5H5]1; 312, [M 2 PMe3]1; 247 IR (Nujol): 1626s, 1553m, 1502s 1H:d 1.02 (m, 9 H, PCH3), 2.54 (m, 3 H, COCH3), 4.28 (s, 5 H, C5H5) 13C-{1H}:d 17.2 (d, JPC = 24.4, PCH3), 27.6 (s, COCH3), 76.1 (s, C5H5), 130.9 (m, Cb), 133.0 (m, Ce), 140.7 (m, Cd), 149.6 (m, Cf), 156.1 (m, Ce), 195.6 (m, Ca), 206.7 (s, COCH3) 19F:d 2170.6 (m, 1 F, Fe), 2159.1 (m, 1 F, Fd), 2140.2 (m, 1 F, Ff), 2120.5 (m, 1 F, Fc) 31P-{1H}:d 31.9 (s, 1 P) 7 [Mo{C6F3H-5-C(O)Me-2}(h-C5H5)(CO)2] Dark red e Mass (EI): M1, [M 2 CO]1 IR (KBr): 1975vs, 1876vs, 1624vs, 1564vs, 1519s 1H:c 2.81 (d, 3 H, JFH = 4.2, COCH3), 5.44 (s, 5 H, C5H5), 6.89 (m, 1 H, Hd) 13C-{1H}:c 28.9 (d, JFC = 9.2, COCH3), 94.7 (s, C5H5), 104.5 (dd, JFC = 21 and 37, Cd), 129.7 (dm, JFC = 18, Cb), 146.0 (dm, 1JFC = 256, Ce), 150.9 (dm, 1JFC = 255, Cf), 164.2 (dm, 1JFC = 260, Cc), 183.1 (dm, JFC = 49, Ca), 205.0 (s, COCH3), 247.9 (s, CO), 253.6 (s, CO) 19F:c 2146.9 (m, 1 F, Fe or Ff), 2144.9 (m, 1 F, Ff or Fe), 292.3 (m, 1 F, Fc) 8 [W{C6F3H-5-C(O)Me-2}(h-C5H5)(CO)2] Green e IR (KBr): 1965vs, 1864vs, 1726s, 1559s 1H:c 2.92 (d, 3 H, JFH = 4.0, COCH3), 5.58 (s, 5 H, C5H5), 6.85 (m, 1 H, Hd) 13C-{1H}:c 27.8 (d, COCH3), 92.9 (s, C5H5), 104.3 (m, Cd), 130.3 (m, Cb), 145.5 (dm, 1JFC = 243, Ce), 150.3 (dm, 1JFC = 271, Cf), 164.1 (dm, 1JFC = 237, Cc), 178.4 (dm, JFC = 64, Ca), 204.3 (m, COCH3), 238.3 (s, CO), 246.5 (s, CO) 19F:c 2147.5 (m, 1 F, Fe), 2146.5 (m, 1 F, Ff), 292.0 (m, 1 F, Fc) 9 [W{C6F3H-5-C(O)Me-2}(h-C5H5)(CO)(PMe3)] Blue violet C, 38.5 (38.8); H, 3.0 (3.45); P, 5.7 (5.9) Mass (FAB): 526, M1 IR (KBr): 1802vs, 1632s, 1586m, 1540m 1H:c 1.54 [d, JPH = 9.0, 9 H, P(CH3)], 2.99 (d, 3 H, JFH = 3.0, COCH3), 5.17 (s, 5 H, C5H5), 6.44 (m, 1 H, Hd) 13C-{1H}:c 19.2 [d, JPC = 29.0, P(CH3)], 26.7 (d, JFC = 7.7, COCH3), 91.3 (s, C5H5), 100.0 (m, Cd), 128.9 (m, Cb), 144.2 (dm, 1JFC = 236, Ce), 150.2 (dm, 1JFC = 249, Cf), 165.4 (dm, 1JFC = 250, Cc), 190.4 (m, COCH3), 193.2 (dm, Ca), 240.6 (d, JPC = 10.7, CO) 19F:c 2151.1 (m, 1 F, Fe or Ff), 2149.7 (m, 1 F, Ff or Fe), 291.3 (m, 1 F, Fc) 31P-{1H}:c 11.3 (s) 10 [Fe{C6F3H-5-C(O)Me-2}(h-C5H5)(PMe3)] Black purple C, 51.4 (51.9); H, 4.8 (4.9); Fe, 15.1 (15.1); P, 8.3 (8.4) Mass (FAB1): 370, M1 (base peak) IR (KBr): 1619s, 1547s, 1505s 1H:d 1.00 [m, 9 H, P(CH3)], 2.56 (m, 3 H, COCH3), 4.28 (s, 5 H, C5H5), 6.78 (m, 1 H, Hd) 13C-{1H}:d 17.1 [d, JPC = 25.6, P(CH3)], 27.6 (s, COCH3), 76.1 (s, C5H5), 103.5 (m, Cd), 137.1 (m, Cb), 143.1 (m, Ce), 149.6 (m, Cf), 169.8 (m, Cc), 196.4 (m, Ca), 207.4 (s, COCH3) 19F:d 2153.0 (m, 1 F, Fe), 2146.6 (m, 1 F, Ff), 296.4 (m, 1 F, Fc) 31P-{1H}:d 32.3 (s)J.Chem. Soc., Dalton Trans., 1997, Pages 2293–2303 2295 Table 2 (Continued ) Compounda NMR data b 11 [Fe{C6F4C(O)Me-2}(h-C5H5)(PPh3)] Black-purple C, 65.2 (64.8); H, 3.7 (4.0); Fe, 9.2 (9.7); P, 5.2 (5.3) Mass (FAB1): 574, M1 (base peak); 509, [M 2 C5H5]1; 383 IR (Nujol): 1734m, 1628m, 1562w 1H:c 1.75 (m, 3 H, COCH3), 4.32 (s, 5 H, C5H5), 6.94 [m, 9 H, P(C6H5)], 7.35 [m, 6 H, P(C6H5)] 13C-{1H}:c 27.1 (s, COCH3), 77.2 (s, C5H5), 127.8, 129.4 [m, P(C6H5)], 134.2 (m, Ce), 141.1 (m, Cd), 149.8 (m, Cf), 156.3 (m, Cc), 193.7 (m, Ca), 207.2 (s, COCH3), Cb not observed 19F:c 2167.1 (m, 1 F, Fe), 2154.3 (m, 1 F, Fd), 2136.4 (m, 1 F, Ff), 2115.4 (m, 1 F, Fc) 31P-{1H}:c 76.0 (s) 12 [{Fe[C6F4C(O)Me-2](h-C5H5)}2(m-dppe)] Purple C, 61.5 (61.1); H, 4.2 (3.9); Fe, 10.6 (10.9) Mass (FAB1): 1022, M1 1H:d 1.28 [m, 4 H, P(CH2)2P], 2.06 (s, 3 H, COCH3), 4.09 (s, 5 H, C5H5), 6.93, 7.21, 7.39 [m, 20 H, P(C6H5)] 13C-{1H}:d 21.6 [s, P(CH2)2P], 27.6 (s, COCH3), 76.5 (s, C5H5), 128.4, 129.7 [m, P(C6H5)], 131.3 (m, Cb), 132.6 [br m, P(C6H5)], 134.0 (m, Ce), 140.8 (m, Cd), 149.5 (m, Cf), 155.5 (m, Cc), 194.0 (m, Ca), 207.3 (s, COCH3), Cb not observed 19F:d 2170.1 (m, 1 F, Fe), 2158.2 (m, 1 F, Fd), 2139.4 (m, 1 F, Ff), 2120.1 (m, 1 F, Fc) 31P-{1H}:d 69.2 (s) a Analytical data given as found (calculated) in %.Mass spectral data given as: m/z (assignment), IR data (cm21) as KBr discs or Nujol mulls, as indicated. b At probe temperature. Data given as: chemical shift (d) (multiplicity, relative intensity, J in Hz, assignment). The data for compounds 1, 4 and 5 are given in ref. 3. c In C6D6. d In CD2Cl2. e Analysis not made; characterised by spectroscopic data only. 2. For many of the compounds the complete assignment of the 13C and 19F NMR spectra was assisted by two-dimensional NMR experiments, including F]F COSY (correlation spectroscopy), C]F heteronuclear single-quantum correlation (HSQC) and heteronuclear multiple-bond coherence (HMBC) experiments.The close similarities between the NMR and IR spectra of 1 and those of 2 and 3 strongly suggest 1 has the same five-membered MC2CO ring that is found in 2 and 3. This is further supported by the crystal structures of several derivatives of 1, as described below. The formation of compounds 1–3 appears to involve a formal insertion of a tetrafluorobenzyne C6F4 fragment, arising from B(C6F5)3, into the M]C bond of a acetylmetal group.To help to elucidate the reaction mechanism, boron-containing products of the reaction were isolated and characterised. Thus, the preparation of 1 was repeated under the same conditions, but the highly volatile products of the reaction mixture were collected in a cold trap and then treated with 4-methylpyridine. After removal of solvent a white powder was recovered. The 1H and 11B-{1H} NMR spectra of this residue showed the formation of the 4-methylpyridine adducts C6H7N?BF2(C6F5) and C6H7N?BF3.The observed chemical shifts were identical with the values quoted in the literature (see the Experimental section). 5 Owing to the lower volatility of BF(C6F5)2 it was not Fig. 1 Molecular structure of [W{C6F4C(O)Me-2}(h-C5H5)(CO)2] 3; that of 2 is almost superimposable possible selectively to evaporate it from the initial reaction mixture.However, this mixture was monitored by 11B-{1H} NMR spectroscopy which showed the formation of bands assignable to BF(C6F5)2. The reaction between the compound [Mo- (h-C5H5)(CO)3Me] and B(C6F5)3 was monitored by 11B NMR spectroscopy and the data also showed the formation of BF(C6F5)2, BF2(C6F5) and BF3. In light of the observation of these boron-containing products, the reaction of the compound [Fe(h-C5H5)(CO)2Me] was repeated using only 1/3 equivalent of B(C6F5)3.The yield of 1 under these conditions was essentially the same as before but it was necessary to double the reaction time. We infer that during the formation of 1–3 the B(C6F5)3 successively loses C6F4 fragments to the metal centre with formation of the compounds BF(C6F5)2, BF2(C6F5) and, finally, BF3. On the basis of these preliminary observations, we tentatively propose a mechanism for the formation of 1, as shown in Scheme 1. The first step is the co-ordination of the Lewis acid B(C6F5)3 to the oxygen atom of a CO ligand thereby promoting migration of the methyl group.In related systems other Lewis acids such as (AlBr3)2 have been shown to promote the migration of a methyl group to give an acyl ligand.6 It is envisaged that the resulting vacant co-ordination site on the Fe is then occupied by a fluorine atom of a weakly co-ordinating h2-C6F5 group of B(C6F5)3. Similar co-ordination of B(C6F5)3 to metal centres has been reported.7–10 The next proposed step involves fragmentation of the B(C6F5)3 ligand by cleavage of a C]F bond and a B]C bond leading to the elimination of BF(C6F5)2, as shown in Scheme 1.Finally, the co-ordinated benzyne C6F4 fragment inserts into the metal–acyl group giving the neutral compound 1. The five-membered rings formed by the atoms MC3O in compounds 1–3 are unusual and the reactions of these compounds with potential donor ligands were studied with a view to displacing the acyl oxygen from the metal.Treatment of the compound [Fe{C6F4C(O)Me-2}(h-C5H5)(CO)] 1 with tertbutyl isocyanide gave air-sensitive dark purple crystals of [Fe{C6F4C(O)Me-2}(h-C5H5)(ButNC)] 4 in 45% yield. This compound is thermally stable, sublimes at 75 8C under vacuum (1021 Torr) and melts without decomposition in a sealed capillary at 85 8C. It is moderately soluble in pentane and very soluble in Et2O, benzene, CH2Cl2 and tetrahydrofuran (thf). Its crystal structure has been determined and the molecular structure is shown in Fig. 2(a). Selected distances and angles are given in Table 3. The structure shows that the tert-butyl isocyanide group has replaced the terminal carbonyl in 1 and a fivemembered FeC3O ring is still present. Attempts to cleave the Fe]O bond in 1 by treatment with an excess of tert-butyl isocyanide gave only 4. In contrast to the formation of 4, treat-2296 J. Chem. Soc., Dalton Trans., 1997, Pages 2293–2303 ment of 1 with 1 equivalent of trimethylphosphine gives airsensitive purple crystals of the compound [Fe{C6F4C(O)- Me-2}(h-C5H5)(CO)(PMe3)] 5 which it appears is formed by a simple addition of a trimethylphosphine ligand and displacement of the co-ordinating acyl oxygen.Compound 5 melts without decomposition at 95 8C and is soluble in all common organic solvents. Its crystal structure has been determined and the molecular structure is shown in Fig. 2(b). Selected distances and angles are given in Table 3.The structure shows that both the PMe3 and CO ligands are bound to the iron centre and the oxygen atom of the acyl group is no longer co-ordinated to the metal. On heating a heptane solution of the compound [Fe{C6- F4C(O)Me-2}(h-C5H5)(CO)(PMe3)] 5 carbon monoxide was evolved and [Fe{C6F4C(O)Me-2}(h-C5H5)(PMe3)] 6 was formed in 74% yield. The black-purple crystals are sensitive to air and melt without decomposition at 92 8C in a sealed capillary. The crystal structure of 6 has been determined and the molecular structure is shown in Fig. 2(c). Selected distances and angles are given in Table 3. The structure is closely similar to those of compounds 2–4 and includes the by now familiar fivemembered MC3O ring system. Thus it appears that compound 5 is an intermediate in the formation of 6. Indeed, refluxing a heptane solution of 1 with an excess of PMe3 leads directly to the formation of 6, as shown by monitoring the infrared spectra of the reaction mixture.Compound 2 was treated with trimethylphosphine in the expectation that the co-ordinated acyl oxygen or a CO ligand would be displaced. However instead an unexpected and specific activation of a C]F bond occurs giving purple crystals of [Mo{C6F3H-5-C(O)Me-2}(h-C5H5)(CO)2] 7. Its crystal structure has been determined; the molecular structure is shown in Fig. 3 and selected distances and angles are given in Table 1. The molecular structure is closely similar to that of 2 except Scheme 1 A possible mechanism for the formation of compound 1 that the F atom attached to the C(12) carbon has been replaced by a hydrogen atom.The anisotropic thermal parameters of all fluorine atoms are unexceptional. The spectroscopic properties of 7 are closely similar to those of 2. For example, the IR spectra of 2 and 7 were almost identical and the carbonyl region for both is consistent with the presence of an Mo(CO)2 fragment. Further, the 19F NMR spectrum of 7 confirms the presence of only three fluorine atoms and in the 1H NMR spectrum there is a multiplet at d 6.89 which may be assigned to a hydrogen attached to the C(12) carbon of the aryl ring.The conversion of 2 into 7 requires the presence of the PMe3 since refluxing 2 in Fig. 2 Molecular structures of (a) [Fe{C6F4C(O)Me-2}(h-C5H5)- (ButNC)] 4, (b) [Fe{C6F4C(O)Me-2}(h-C5H5)(CO)(PMe3)] 5 and (c) [Fe{C6F4C(O)Me-2}(h-C5H5)(PMe3)] 6J. Chem. Soc., Dalton Trans., 1997, Pages 2293–2303 2297 toluene for 48 h in the absence of PMe3 showed no evidence for the formation of 7.Under similar conditions, no reaction occurred between 2 and triphenylphosphine. This suggests the trimethylphosphine is directly responsible for the removal of the fluorine during the formation of 7. The reactions of 2 and 3 are given in Scheme 2. The reaction between [W{C6F4C(O)- Me-2}(h-C5H5)(CO)2] 3 and PMe3 gave the green tungsten compound [W{C6F3H-5-C(O)Me-2}(h-C5H5)(CO)2] 8 and the similarity of the spectroscopic data showed that 8 and 7 were very closely similar and therefore we propose they are isostructural. Photolysis of 3 in the presence of an excess of trimethylphosphine causes displacement of a carbonyl ligand giving the blue-violet compound [W{C6F3H-5-C(O)Me-2}(h-C5H5)(CO)- (PMe3)] 9, together with some 8; these may be separated by column chromatography on alumina.It seems likely that the formation of 9 from 3 proceeds via 8 as an intermediate.Following the discovery of C]F bond-activation reactions which lead to the formation of compounds 7 and 8, the photolysis of a refluxing toluene solution of 1 and trimethylphosphine was investigated. Chromatography of the reaction mixture allowed the separation of the purple-black crystalline compound [Fe{C6F3H-5-C(O)Me-2}(h-C5H5)(PMe3)] 10 in 26% yield. The air-sensitive crystals of 10 are soluble in common organic solvents and melt without decomposition at 90 8C.Table 3 Selected interatomic distances (Å) and angles (8) for compounds 4–6, 11 and 12 Compound 4 Fe]O Fe]C(1) Fe]C(20) Fe]Cpcen O]C(7) C(1)]C(2) C(1)]C(6) C(21)]N 1.964(2) 1.938(2) 1.811(2) 1.712 1.247(3) 1.378(4) 1.424(3) 1.453(3) O]Fe]C(1) O]Fe]C(20) C(1)]Fe]C(20) Fe]O]C(7) Fe]C(1)]C(6) O]C(7)]C(6) Fe–C(20)]N C(1)]C(6)]C(7) 81.05(8) 95.22(8) 87.30(9) 117.6(2) 112.9(2) 115.6(2) 117.4(2) 112.4(2) 5 Fe]P Fe]C(1) Fe]C(9) Fe]Cpcen O(7)]C(7) C(1)]C(2) C(1)]C(6) C(9)]O(9) 2.195(1) 2.017(4) 1.744(4) 1.722 1.207(5) 1.377(6) 1.414(6) 1.158(5) P]Fe]C(1) P]Fe]C(9) C(1)]Fe]C(9) C(1)]C(6)]C(7) Fe]C(1)]C(6) O(7)]C(7)]C(6) Fe]C(9)]O(9) C(6)]C(7)]O(7) 93.0(1) 86.4(1) 99.4(2) 122.5(3) 124.9(3) 121.5(4) 172.2(4) 121.5(4) 6 Fe]P Fe]C(9) Fe]O Fe]Cpcen O]C(15) C(9)]C(14) C(14)]C(15) P]C(6) 2.188(1) 1.929(3) 1.946(3) 1.701 1.262(5) 1.436(5) 1.433(6) 1.816(5) P]Fe]C(9) C(9)]Fe]O P]Fe]O C(9)]C(14)]C(15) Fe]C(9)]C(14) O]C(15)]C(14) Fe]O]C(15) C(10)]C(9)]C(14) 88.2(1) 81.7(1) 95.30(9) 112.3(3) 112.0(3) 115.8(3) 117.4(2) 114.1(3) 11 Fe]P Fe]C(1) Fe]O Fe]Cpcen O]C(7) C(1)]C(6) C(6)]C(7) P]C(111) 2.220(1) 1.939(5) 1.980(3) 1.717 1.256(5) 1.442(6) 1.440(7) 1.844(4) P]Fe]C(1) C(1)]Fe]O P]Fe]O C(1)]C(6)]C(7) Fe]C(1)]C(6) O]C(7)]C(6) Fe]O]C(7) C(2)]C(1)]C(6) 93.7(1) 81.6(2) 88.8(1) 112.9(4) 112.5(3) 115.9(4) 116.9(3) 115.0(4) 12 Fe(1)]P(1) Fe(1)]C(6) Fe(1)]O(1) Fe(1)]Cpcen O(1)]C(12) C(6)]C(11) C(11)]C(12) P(1)]C(14) 2.186(2) 1.921(6) 1.956(4) 1.692 1.264(7) 1.429(9) 1.42(1) 1.822(6) P(1)]Fe(1)]C(6) C(6)]Fe(1)]O(1) P(1)]Fe(1)]O(1) C(6)]C(11)]C(12) Fe]C(6)]C(11) O(1)]C(12)]C(11) Fe]O(1)]C(12) C(7)]C(6)]O(11) 93.1(2) 81.3(2) 92.9(1) 112.3(6) 113.4(6) 116.0(6) 117.0(5) 115.3(7) Cpcen denotes the centroid of the C5H5 ring.The spectroscopic data show that, as for 7 and 8, there is a hydrogen attached to the Cd carbon atom of the aromatic C6 ring. The reactions of 1 described above are shown in Scheme 3. Analogous photolysis reactions on mixtures of compound 1 with PPh3 or dppe (Ph2PCH2CH2PPh2) did not result in the substitution of fluorine Fd by hydrogen.Instead, displacement of the terminal carbonyl occurred. A toluene solution of a mixture of 1 and triphenylphosphine was refluxed for 18 h, until the n(CO) absorption of the carbonyl ligand of 1 had disappeared, to give black-purple crystals of [Fe{C6F4C(O)Me-2}(h-C5H5)- (PPh3)] 11, in 42% yield. Similarly, when a mixture of 1 and dppe in toluene at 60 8C was photolysed for 15 h dark purple crystals of [{Fe[C6F4C(O)Me-2](h-C5H5)}2(m-dppe)] 12 were prepared in 34% yield. The crystals of 11 and 12 are moderately stable in air and only show signs of decomposition after several hours.They dissolve sparingly in pentane and heptane and are very soluble in benzene, CH2Cl2 and thf. Compound 11 sublimes at 105 8C under vacuum (1021 Torr) and melts without decomposition at 172 8C. The crystal structures of 11 and 12 have been determined and the molecular structures are shown in Fig. 4(a) and 4(b), respectively. Selected distances and angles are given in Table 3. In the light of the crystal structure data an explanation for the preference for fluorine substitution at the C(12) (or Cd) position of 4, 6, 11 and 12 is offered below. The compound [W{C6F4C(O)Me-2}(h-C5H5)(CO)2] 3 was Fig. 3 Molecular structure [Mo{C6F3H-5-C(O)Me-2}(h-C5H5)(CO)2] 7 Scheme 2 (i) M = Mo, B(C6F5)3 in toluene for 42 h, yield 25%; M = W, same except 48 h, 75%; (ii) excess of PMe3 in toluene, photolysis for 7.5 h then room temperature (r.t.) for 16.5 h, chromatography on alumina, ca. 25% 8 and 25% 9; (iii) excess of PMe3 in toluene, reflux for 3 h, stir at r.t. for 15 h, 25%2298 J. Chem. Soc., Dalton Trans., 1997, Pages 2293–2303 Scheme 3 (i) L = ButNC, in pentane at r.t. for 18 h, yield 45%; L = PPh3, in toluene, reflux for 18 h, 42%; (ii) PMe3 in heptane at r.t. for 15 h, 45%; (iii) dppe (0.5 equivalent) in toluene under photolysis for 15 h at 60 8C, 34%; (iv) excess of PMe3 in toluene at reflux under photolysis for 3 d, 26%; (v) excess of PMe3 in heptane under reflux for 15 h, 74% treated with dppe and then the product was subsequently photolysed in the presence of PMe3.After chromatography a few crystals of a blue-violet compound were obtained and the crystal structure was determined. The molecular structure of 13 is shown in Fig. 5 and selected interatomic distances and angles are given in Table 1.The structure shows that the dppe ligand has replaced one of the terminal carbonyl ligands and is Fig. 4 Molecular structures of (a) [Fe{C6F4C(O)Me-2}(h-C5H5)- (PPh3)] 11 and (b) [{Fe[C6F4C(O)Me-2](h-C5H5)}2(m-dppe)] 12. The H atoms are omitted for clarity bonded to the metal centre in a monodentate fashion via one phosphorus atom. The usual bidentate co-ordination of dppe may be prevented by steric constraints. In addition to dppe coordination, the aromatic fluorine atom meta to the metal centre has been substituted by hydrogen.This substitution is presumed to arise from the presence of PMe3. The bond lengths and angles in compound 13 compare favourably with those obtained for [Fe{C6F4C(O)Me-2}(h-C5H5)(PPh3)] 11. For example, the acyl C]] O bond length is found to be 1.254(12) Å as compared with 1.256(5) Å in 11. The acyl O–metal bond in 13 is 2.078(9) Å compared to 1.980(3) Å in 11. Since only a few crystals of 13 were available further characterisation was not made.Spectroscopic studies The labelling of the C6F4 atoms used in the assignments of the NMR spectra are shown in Table 2. For the most part the assignments are straightforward. However it is interesting that the 1H NMR spectra showed that the resonances due to the methyl hydrogens of the acyl group occur either as doublets (1, 4, 2, 3, 7) or multiplets (5, 6, 10, 11) between d 1.85 and 2.56. The four fluorine atoms Fc–f of compound 11 were irradiated in turn whilst the 1H NMR spectrum of the acyl methyl group was monitored.It was found that irradiation of the fluorine at d 2136.4 led to the collapse of the methyl multiplet. This selective 1H-{19F} irradiation experiment suggests that the splitting of Fig. 5 Molecular structure of [W{C6F3H-5-C(O)Me-2}(h-C5H5)- (CO){P(Ph)2CH2CH2PPh2}] 13J. Chem. Soc., Dalton Trans., 1997, Pages 2293–2303 2299 the methyl signals is the result of through-space coupling.The 1H-{19F} difference nuclear Overhauser effect (NOE) spectrum of 3 also indicates that the coupling arises from a through-space interaction with a 19F nucleus. A 19F]19F COSY experiment on 3 gave the order of d(19F) 2140.0, 2165.3, 2157.1 and 2116.0 for the C6F4 fluorines and 1H-{19F} selective-irritation experiments identified the 19F signal at d 2140.0 as due to Ff (JFH = 3.8 Hz). A 19F]13C multiple-bond correlation (19F]13C HMBC) experiment further indicates the Ff atom is the one adjacent to the carbon atom bearing the acetyl group (Ca), and the 19F signal at d 2116.0 is due to the Fc atom.Also, a 19F]13C single-bond correlation (19F]13C HSQC) experiment for 3 also indicates the same assignment of the four C6F4 fluorines, as shown in Table 2. Independent support for this assignment comes from a similar NMR study of 7, which also indicated a close spatial proximity of the Ff fluorine atom as giving rise to the splitting of the 1H NMR signal of the acyl methyl group.The assignments of the carbon resonances of the C6F4 fragment of compound 1 were achieved by 19F]13C singlebond correlation experiments. The signals of atoms Cc–Cf were assigned with the help of the cross-peaks of the related fluorine resonances. The remaining signals of atoms Ca and Cb of the C6F4 fragment could be assigned on the basis of their chemical shifts. Thus Ca bound to the acyl group is expected to be less shielded than Cb which is bound to the metal atom.Comparison of the chemical shifts of the CdF carbon of compounds 1–3 and 6 with those of the CdH carbons of 7–9 show that, as expected, the latter are shifted upfield, by 33, 36 and 37 ppm, respectively. The resonances of the carbon atoms of the acyl groups in compounds 1–12 were observed between d 206 and 211. For 5, where the acyl group is not co-ordinated to the iron centre, the acyl carbon signal was observed in the same range indicating that the co-ordination of the oxygen atom to the metal has little influence on the shielding of the acyl carbons.The IR spectra of compounds 1–12 reveal strong bands between 1581 and 1547 cm21 characteristic of n(CO) of acyl groups. The absorption of the non-co-ordinating acyl group in 5 occurs at 1547 cm21 compared to 1581 cm21 for [Fe{C6- F4C(O)Me-2}(h-C5H5)(CO)] 1. Structural aspects In compounds 2–4, 6, 7 and 11–13 the atoms of the fivemembered MC3O ring are essentially coplanar and deviations from planarity do not exceed 0.5 Å; the MC3O rings lie essentially coplanar with the condensed benzene ring.For example, the angle between the FeC3O ring and the C6 rings for 6 and 12 are 2.4 and 1.38, respectively. The bond lengths of the C]] O groups in 4, 6, 11 and 12 are 1.247(3), 1.262(5), 1.256(5) and 1.264(7) Å respectively, and these are longer than in the compound 5 [1.207(5) Å]. The latter distance is typical for related non-co-ordinating C]] O groups, values for which occur in the range 1.19–1.22 Å.11 The carbon–carbon bond distance between the C6 ring and the carbon of the acyl C]] O group [Ca]C (acyl)] in compounds 4, 6, 11 and 12 are unusually short with the values 1.446(4), 1.433(6), 1.440(7) and 1.42(1) Å, respectively.The Ca]C (acyl) bond length in 5 is longer at 1.514(6) Å which is characteristic of a single bond11 between an aromatic ring and a sp2- hybridised carbon atom. These shorter bond distances can be understood in terms of a contribution of the resonance form B shown in Scheme 4. This would develop a positive charge on the Cd carbon which may account for its susceptibility to nucleophilic substitution of the Fd fluorine as observed in the formation of 7, 8 and 10.We note that the substitution of fluorine atoms by hydrogen on aryl rings using organic and organometallic nucleophiles is known12 but, to our knowledge, there is no previous report of the use of tertiary phosphines for this transformation.13 In conclusion, we have identified surprising reactions, namely the formation of the metallacycle MC3O group in a high-yield reaction which incorporates a benzenoid C6F4 moiety derived from B(C6F5)3, and a high-yield and selective substitution of a fluorine of the aromatic C6F4 group by a hydrogen atom.Experimental All manipulations of air- and moisture-sensitive materials were carried out under dinitrogen using either standard Schlenk techniques, or in an inert-atmosphere dry-box containing dinitrogen. The dinitrogen was purified by passing over 4 Å molecular sieves, and either BASF catalyst or MnO.Solvents were predried by standing over 4 Å molecular sieves and then refluxed and distilled under an atmosphere of dinitrogen from sodium–potassium alloy [pentane, light petroleum (b.p. 40– 60 8C), toluene, benzene], and stored over freshly activated molecular sieves (4 Å). Deuteriated solvents for NMR studies were refluxed and condensed from potassium (C6D6, [2H8]- toluene) or CaH2 (CD2Cl2), then stored in Young ampoules.The NMR data were recorded using a Bruker AM 200 or 300 (1H, 13C) and Varian Unity Plus 500 spectrometer (1H, 11B, 13C, 19F, 31P, variable-temperature and two-dimensional experiments). Most two-dimensional experiments were performed using z gradients with a version of the pulse sequences. An unambiguous assignment of all 19F and 13C NMR signals was carried out for compounds 3 and 8 (see text).The signals of 2, 7 and 5 were assigned by analogy. Standards: SiMe4 (1H, 13C); internal solvents as secondary standards, d(1H)(CHCl3) 7.24, (C6D5H) 7.15, d(13C)(CDCl3) 77.0, (C6D6) 128.0; CFCl3 (19F, fluorobenzene as a secondary standard, d 2116.0); external BF3?OEt2 (11B). Chemical shifts (d) in ppm; a positive sign indicates a downfield shift relative to the standard. Coupling constants are given as absolute values. The IR spectra were recorded as Nujol mulls on NaCl plates or KBr discs, using a Perkin-Elmer 1510 Fourier-transform interferometer. Mass spectra were measured using electron-impact (EI, 70 eV, the calibration of the instrument is unreliable and varies ±1%) and fast atom bombardment techniques. Elemental analyses were performed by the analytical department of this laboratory.The compounds [Fe(h-C5H5)(CO)2Me] and [M(h-C5H5)- (CO)3Me] (M = Mo or W) were prepared as previously described.14 Preparations B(C6F5)3.A modification of a previously published method1 was used. CAUTION: extreme precautions were exercised due to the explosive nature of the intermediate lithium salt Li(C6F5) which detonates above ca. 220 8C. It is essential to maintain the reaction mixture at 278 8C until the reaction with BBr3 has been completed. The use of safety screens is advised. The apparatus used is shown in Fig. 6. A solution of bromopentafluorobenzene (53.5 g, 0.217 mol) in light petroleum (b.p. 40–60 8C) (1 l) was treated with a solution of n-butyllithium in hexane (86.8 cm3, 2.5 mol dm23, 0.217 mol) at 278 8C.The resulting white suspension was stirred for 1 h and then added to a solution of boron tribromide (72.0 cm3, 1 mol dm23, 0.072 mol) in light petroleum (0.3 l) at 278 8C. After additional stirring for 15 h at room temperature the mix- Scheme 4 Resonance structures suggested from the interatomic distances, showing the development of a positive charge on Cd in B2300 J.Chem. Soc., Dalton Trans., 1997, Pages 2293–2303 ture was filtered through a bed of Celite to give a clear colourless solution. A white microcrystalline solid was obtained after the volume of the filtrate was concentrated to 30 cm3. The supernatant was decanted off and the resulting white solid dried in vacuo. Further purification by vacuum sublimation (140 8C, 1021 Torr) gave white crystals. Yield: 18 g (48%). [Fe{C6F4C(O)Me-2}(Á-C5H5)(CO)] 1. A red solution of [Fe(h-C5H5)(CO)2Me] (4.34 g, 22.6 mmol) in toluene (30 cm3) was treated with B(C6F5)3 (11.57 g, 22.6 mmol) in toluene (120 cm3) and was stirred for 15 h at room temperature.Completion of the reaction was revealed by IR spectroscopy [replacement of the n(CO) absorptions of the starting material at 2009 and 1952 cm21 by that of the product at 1957 cm21]. The resulting orange-green mixture was evaporated to dryness, and purified by alumina column chromatography at 225 8C. Elution with Et2O–pentane (1 : 10) gave a green-brown fraction, from which compound 1 was isolated as a microcrystalline solid after removal of the volatiles in vacuo.Extraction into pentane and crystallisation at 280 8C yielded green-brown crystals. Yield: 6.03 g (79%). The volatiles of the reaction mixture were isolated in a cold trap (77 K) and were treated with an excess of 4-methylpyridine to form Lewis acid–base adducts with the boron-containing byproducts. The clear solution was warmed to room temperature and the volatiles were evaporated under reduced pressure to give a white residue.The components of the residue were identified as BF3?NC6H7 and BF2(C6F5)?NC6H7 by 1H and 11B NMR spectroscopy. BF3?NC6H7: 1H NMR (C6D6) d 1.46 (s, 3 H, NC5H4CH3), 6.18 (d, 2 H, 3J = 6.0, NC5H4aCH3) and 8.14 (d, 2 H, 3J = 6.0, NC5H4bCH3); 11B NMR (C6D6) d 1.0 (q, 1 B, J = 11.0 Hz, BF3). BF2(C6F5)?NC6H7: 1H NMR (C6D6) d 1.50 (s, 3 H, NC5H4CH3), 6.28 (d, 2 H, 3J = 6.3, NC5H4aCH3) and 8.22 (d, 2 H, 3J = 6.3, NC5H4bCH3); 11B NMR (C6D6) d 3.9 (t, 1 B, J = 47.9 Hz, BF2C6F5).[Mo{C6F4C(O)Me-2}(Á-C5H5)(CO)2] 2. A solution of B(C6F5)3 (1.57 g, 3.08 mmol) in toluene (25 cm3) was added to a stirred solution of [Mo(h-C5H5)(CO)3Me] (0.8 g, 3.08 mmol) in toluene (30 cm3). The solution rapidly changed from golden yellow to dark red. It was left to stir for 42 h. The reaction mixture was blood-red and the IR spectrum indicated that all the starting material had reacted.The solvent was removed under reduced pressure to give a dark red solid which was then Fig. 6 Apparatus used for the preparation of solvent-free B(C6F5)3 placed on alumina (ca. 3 g). The coated alumina was added to a low-temperature (225 8C) chromatography column which had been filled with light petroleum and alumina (25 cm). The product was washed with light petroleum (3 × 100 cm3) and slow elution with light petroleum–Et2O (10 : 1) facilitated collection of a red band.The solvent was removed from the eluate under reduced pressure and the solid obtained recrystallised from warm heptane to give red crystals of compound 2. Yield: 25%. [W{C6F4C(O)Me-2}(Á-C5H5)(CO)2] 3. A solution of B(C6- F5)3 (1.47 g, 2.87 mmol) in toluene (25 cm3) was added to a stirred solution of [W(h-C5H5)(CO)3Me] (1 g, 2.87 mmol) in toluene (30 cm3). The solution immediately turned from yelloworange to dark red. It was left to stir and the reaction progress was monitored by solution IR spectroscopy.After stirring for 48 h the solution was dark purple and the IR spectrum indicated that the starting material had reacted. The solvent was removed under reduced pressure to give a dark purple solid which was then placed on alumina. The coated alumina was added to a low-temperature (225 8C) chromatography column which had previously been filled with pentane and 25 cm of alumina. The product was washed with pentane (3 × 100 cm3) and elution with pentane–Et2O (1 : 1) developed a deep purple band which was collected.The solvent was removed under reduced pressure and the residue recrystallised from hot heptane to give purple crystals (m.p. 175–176 8C; sublimation point 95 8C, 0.1 Torr) of compound 3. Yield: 75%. [Fe{C6F4C(O)Me-2}(Á-C5H5)(ButNC)] 4. An orange-brown solution of [Fe{C6F4C(O)Me-2}(h-C5H5)(CO)] 1 (230 mg, 0.68 mmol) in pentane (100 cm3) was treated with tert-butyl isocyanide (56.5 mg, 0.68 mmol) and stirred for 18 h at room temperature.Completion of the reaction was monitored by IR spectroscopy [replacement of the n(CO) absorption of the starting material at 1957 cm21 by the n(C]] ] NBut) absorption at 2038 cm21]. The resulting purple mixture was evaporated to dryness and purified by column chromatography on alumina at 225 8C. Compound 4 was eluted with Et2O–pentane (1 : 4) and the volatiles of the purple eluate were removed in vacuo. Extraction into pentane and crystallisation at 280 8C yielded dark purple crystals.Yield: 120 mg (45%). [Fe{C6F4C(O)Me-2}(Á-C5H5)(CO)(PMe3)] 5. An orangebrown solution of compound 1 (410 mg, 1.21 mmol) in heptane (50 cm3) was treated with trimethylphosphine (96 mg, 1.21 mmol) and stirred for 15 h at room temperature. The resulting purple solution was evaporated to dryness and vacuum sublimation (70 8C, 1021 Torr) gave a microcrystalline solid. Extraction into heptane and crystallisation at 280 8C yielded purple crystals.Yield: 120 mg (45%). Compound 5 melts without decomposition at 100 8C in a sealed capillary and sublimes at 85 8C under vacuum (1021 Torr). It is soluble in common organic solvents. [Fe{C6F4C(O)Me-2}(Á-C5H5)(PMe3)] 6. Method A. A purple solution of compound 5 (110 mg, 0.26 mmol) in heptane (30 cm3) was refluxed for 15 h. The elimination of CO from the compound was revealed by IR spectroscopy, as in the synthesis of compound 4. The solvent was evaporated under reduced pressure, and extraction of the residue into heptane and crystallisation at 280 8C yielded purple crystals of 6.Yield: 75 mg (74%). Method B. A solution of compound 1 (550 mg, 1.62 mmol) and trimethylphosphine (1 cm3; 735 mg, 9.6 mmol) in heptane (50 cm3) was refluxed for 15 h until the reaction was complete (IR monitoring). The volatile materials were removed under reduced pressure and the purple residue was purified by column chromatography on alumina at 225 8C.Elution with Et2O– pentane (1 : 20) gave a purple fraction, from which the com-J. Chem. Soc., Dalton Trans., 1997, Pages 2293–2303 2301 pound 6 was isolated as a microcrystalline solid after removal of the solvent in vacuo. Yield: 200 mg (32%). [Mo{C6F3H-5-C(O)Me-2}(Á-C5H5)(CO)2] 7. A blood-red solution of [Mo{C6F4C(O)Me-2}(h-C5H5)(CO)2] (0.06 g, 1.47 mmol) in toluene (40 cm3) was treated with PMe3 (1.22 cm3, 11.79 mmol) and the mixture was left to stir at r.t. for 5 h.It became purple and the solution IR spectrum showed little change in the CO region. The mixture was refluxed for 3 h and then left to stir at r.t. for 16 h. The solvent was removed under reduced pressure and the dark purple residue chromatographed on alumina at 225 8C. Elution with light petroleum gave a purple band. Further elution with light petroleum–Et2O (10 : 1) developed this band further and also initiated a slow-moving blue band towards the top of the column. The purple band was collected, the solvent was removed under reduced pressure and the residue recrystallised from hot heptane to yield dark red crystals of compound 7.Yield: 65%. [W{C6F3H-5-C(O)Me-2}(Á-C5H5)(CO)2] 8 and [W{C6F3H- 5-C(O)Me-2}(Á-C5H5)(CO)(PMe3)] 9. A claret-purple solution of [W{C6F4C(O)Me-2}(h-C5H5)(CO)2] 3 (0.4 g, 0.81 mmol) in toluene (60 cm3) was treated with PMe3 (1.03 cm3, 9.95 mmol). The mixture was photolysed under reflux using an ultraviolet lamp (500 W) for 7.5 h.The solution was then allowed to stir for 16.5 h, giving a violet-purple solution with a black deposit. The solvent was removed under reduced pressure and the dark violet-black residue chromatographed at 225 8C. Elution with light petroleum (3 × 100 cm3) and then with light petroleum– Et2O (2 : 1) developed a claret-purple band. Increasing the solvent ratio to 1 : 1 also developed a second blue band. The two bands were neatly separated by very slow elution and collected separately. The solvent in each case was removed under reduced pressure to yield green and blue-violet solids.The green solid was recrystallised from hot heptane to give green crystals of compound 8. The blue-violet solid was also recrystallised from hot heptane to give a blue-violet microcrystalline powder of 9. [Fe{C6F3H-5-C(O)Me-2}(Á-C5H5)(PMe3)] 10. An orangebrown solution of [Fe{C6F4C(O)Me-2}(h-C5H5)(CO)] 1 (490 mg, 1.44 mmol) and an excess of trimethylphosphine (ª3 cm3) in toluene (50 cm3) were refluxed and photolysed using a ultraviolet lamp (500 W) for 3 d.The volatile materials of the bluepurple reaction mixture were removed under reduced pressure and the residue was purified by column chromatography on alumina at 225 8C. Elution with Et2O–pentane (1 : 1) gave a purple fraction and after removal of the solvent in vacuo the compound [Fe{C6F3H-5-C(O)Me-2}(h-C5H5)(PMe3)] 10 was isolated as a microcrystalline solid. The residue was extracted into boiling heptane and cooling of the solution to 280 8C gave black-purple crystals.Yield: 140 mg (26%). Compound 10 can also be prepared in the same manner starting from 6, in 74% yield. [Fe{C6F4C(O)Me-2}(Á-C5H5)(PPh3)] 11. Compound 1 (700 mg, 2.06 mmol) and triphenylphosphine (540 mg, 2.06 mmol) were refluxed in toluene (50 cm3) for 18 h until the reaction was complete. It was monitored by IR spectroscopy [disappearance of the n(CO) absorption of the starting material at 1957 cm21].The volatiles of the resulting purple solution were removed under reduced pressure and the residue was purified by column chromatography on alumina at 225 8C. Elution with Et2O– pentane (1 : 2) gave a purple fraction which was evaporated to dryness in vacuo and crystallisation from a solution of pentane at 280 8C gave dark purple crystals. Yield: 500 mg (42%). [{Fe[C6F4C(O)Me-2](Á-C5H5)}2(Ï-dppe)] 12. An orangebrown solution of compound 1 (295 mg, 0.87 mmol) and 1,2- bis(diphenylphosphino)ethane (173 mg, 0.43 mmol) in toluene (50 cm3) was photolysed using an ultraviolet lamp (500 W) for 15 h at 60 8C until the reaction was complete (IR monitoring). An important feature is that exactly half an equivalent of dppe must be used to avoid bidentate co-ordination of the phosphine ligand.The solvent was removed under reduced pressure and the residue was extracted into pentane–Et2O (1 : 1) and filtered through a bed of alumina. The volatiles of the filtrate were evaporated under reduced pressure and the residue was extracted into boiling heptane.Crystallisation at 280 8C yielded purple crystals of compound 12. Yield 150 mg (34%). [W{C6F3H-5-C(O)Me-2}(Á-C5H5)(CO){P(Ph)2CH2CH2- PPh2}] 13. The compound [W{C6F4C(O)Me-2}(h-C5H5)(CO)2] (1.14 g, 2.30 mmol) was dissolved in toluene (50 cm3) and Ph2PCH2CH2PPh2 (dppe) (0.916 g, 2.30 mmol) was added. The mixture was refluxed for 5.5 h after which time the solution IR spectrum showed no change in the carbonyl region.The mixture was left to stir at room temperature for 7 d. No reaction was obvious so the solvent was removed under reduced pressure and the solid obtained was placed on alumina (ca. 5 g). The coated alumina was added to a low-temperature (225 8C) chromatography column filled with light petroleum and 20 cm of alumina. The column was washed with light petroleum (3 × 100 cm3) and elution with light petroleum–Et2O (1 : 1) gave a purple band.The solvent was removed under reduced pressure and recrystallisation from boiling heptane gave claret-purple crystals. Trimethylphosphine (1.03 cm3, 9.95 mmol) was added to a stirred solution of the above crystals in toluene (40 cm3) and the mixture photolysed for 7 h using a medium-pressure ultraviolet lamp (500 W). After this time the solution had turned to dark blue-violet and the carbonyl region of the IR spectrum showed the formation of a new product. The solvent was removed under reduced pressure and the solid obtained was suspended on alumina.The coated alumina was added to a lowtemperature (225 8C) chromatography column which had been filled with light petroleum and 20 cm of alumina. The product was washed with light petroleum (2 × 100 cm3) and eluted with light petroleum–Et2O (10 : 1). This developed a blue-violet band which was collected by gradually increasing the solvent ratio to 1 : 1. The solvent was removed under reduced pressure and recrystallisation from boiling heptane gave blue-violet crystals of the compound [W{C6F3H-5-C(O)Me-2}(h-C5H5)(CO)- {P(Ph)2CH2CH2PPh2}] 13.Only a few crystals were obtained (yield ca. 4%) and their characterisation was by X-ray crystallography only. Crystallography X-Ray-quality crystals were obtained by slow cooling of hot saturated solutions of the compounds in pentane (4, 6, 11) and heptane (2, 3, 5, 7, 12, 13) to 280 8C. They were sealed in a Lindemann capillary (0.5–0.7 mm) under dinitrogen.For 2, 5–7 and 12 the data collection was carried out on an Enraf-Nonius CAD4 diffractometer. The unit-cell parameters were calculated from the setting angles of 25 strong high-angle carefully centred reflections. Three reflections were chosen as intensity standards and measured every 3600 s of X-ray exposure time and three orientation controls were measured every 147–200 reflections. The data were measured using graphite-monochromated Mo- Ka radiation (l = 0.710 69 Å) with an w–2q scan mode.The ratio of the scanning rates was w: q = 1 : 2. Data were corrected for Lorentz-polarisation effects. For 3, 4 and 13 the data collection was carried out on a Delft Instruments FAST TV areadetector diffractometer, equipped with a rotating-anode FR 591 generator (50 kV, 10 mA); further details are described in ref. 15. For compound 6 a correction for crystal decay (ca. 13%) was applied during processing the data set. For 6 and 12 data were also corrected using an empirical absorption correction.16 In 122302 J.Chem. Soc., Dalton Trans., 1997, Pages 2293–2303 Table 4 Crystallographic details for compounds 2, 3, 7 and 13 2 3 7 13 Formula Colour Formula weight Crystal size/mm Crystal system Space group a/Å b/Å c/Å a/8 b/8 g/8 U/Å3 Z Dc/g cm23 F(000) m/cm21 T/K Lattice segment qmax/8 for data No. reflections: total unique in refinement Rmerge Parameters Data: parameter ratio Minimum, maximum transmission Ra/R9 b R1 [F > 4s(F)] wR2 (all data) Goodness of fit Maximum, minimum peaks in difference maps/e Å23 C15H8F4MoO3 Red 408.16 0.33 × 0.36 × 0.24 Triclinic P1� 8.3420(4) 8.9687(7) 11.1874(9) 69.126(6) 75.691(5) 63.515(5) 698.47 2 1.94 400 9.47 293(2) ±h, ±k, ±l 25.0 2836 2437 2072 [I > 3s(I)] 0.013 240 9 : 1 [F > 3s(F)] 0.88, 1.14 0.0272/0.0325 ——— 0.70, 20.40 C15H8F4O3W Red 496.06 0.28 × 0.14 × 0.14 Triclinic P1� 8.206(8) 8.905(4) 11.104(9) 68.44(8) 75.48(8) 63.83(4) 673.5(9) 2 2.446 464 8.636 120 ±h, ±k, ±l 29.9 1913 (after red.) 1913 1810 [F > 4s(F)] 0.0 198 9.7 : 1; 9.1 : 1 [F > 4s(F)] 0.616, 1,0 — 0.0467 0.1325 1.098 1.794, 21.355 C15H9F3MoO3 Purple 390.16 0.5 × 0.25 × 0.15 Triclinic P1� 8.285(2) 8.933(1) 11.176(2) 68.79(1) 74.55(2) 63.78(2) 683.3(2) 2 1.896 384 10.04 293(2) ±h, ±k, 1l 29 2144 (after red.) 2144 1882 [F > 4s(F)] 0.0 200 10.7 : 1; 9.4 : 1 [F > 4s(F)] 0.575, 1.0 — 0.0315 0.0706 1.072 0.832, 20.669 C40H33F3O2P2W Blue-violet 848.49 0.14 × 0.14 × 0.07 Monoclinic P21/c 11.577(2) 19.525(4) 15.389(3) 90.00 97.20(3) 90.00 3451.1(1) 4 1.633 1680 3.491 150 ±h, ±k, ±l 25.0 8529 5015 (Rint = 0.0784) 2404 [F > 4s(F)] 0.078 434 11.6 : 19; 5.5 : 1 [F > 4s(F)] 0.72, 1.0 — 0.0494 0.1028 0.645 1.020, 20.835 a R = S(|Fo| 2 |Fc|)/S|Fo.b R9 = [wS(|Fo| 2 |Fc|)2/Sw|Fo|2]� �� . Table 5 Crystal data and experimental details for compounds 4–6, 11 and 12 4 5 6 11 12 Formula M Crystal size/mm Crystal system Space group a/Å b/Å c/Å a/8 b/8 g/8 U/Å3 Z Dc/g cm23 F(000) m/cm21 T/K Total data collected Total unique data Total observed data R Merge No.parameters Observations/ parameters Chebychev weighting scheme parameters Minimum, maximum residual peak/e Å23 RR 9 b C18H17F4FeNO 395.18 0.19 × 0.22 × 0.40 Triclinic P1� 8.4663(7) 10.5039(7) 10.9915(8) 98.927(6) 106.067(6) 104.746(6) 881.2 2 1.49 404 8.97 293 4372 3838 2322 [I > 3s(I)] 0.020 226 10.3 6.64, 23.52, 5.57 20.23, 0.22 0.033 0.031 C17H17F4FeO2P 416.14 0.07 × 0.22 × 0.18 Monoclinic P21/c 8.586(2) 13.215(2) 15.712(3) 104.70(1) 1724.39 4 1.60 848 10.10 120 6712 2597 1795 [I > 2s(I)] 0.061 226 7.9 15.1, 29.4, 11.6 20.36, 0.83 0.043 0.048 C16H17F4FeOP 388.12 0.31 × 0.40 × 0.74 Monoclinic P21/c 8.577(2) 9.222(3) 20.916(7) 92.53(2) 1652.9(8) 4 1.56 792 10.45 293 3826 3585 2053 [I > 3s(I)] 0.061 208 9.9 21.90, 25.94, 18.70 20.39, 0.39 0.043 0.051 C31H23F4FeOP 574.34 0.13 × 0.18 × 0.25 Triclinic P1� 9.391(3) 10.195(6) 13.26(1) 95.23(4) 103.13(4) 93.49(3) 1226.9 2 1.56 588 7.29 120 4014 2629 2018 [I > 2s(I)] 0.068 253 8.0 6.06, 0.60, 3.51 20.41, 0.67 0.047 0.049 C52H40F8Fe2O2P2 1022.52 0.34 × 0.42 × 0.43 Monoclinic P21/n 10.667(4) 22.408(3) 19.185(5) 93.92(2) 4575(2) 4 1.48 2088 7.74 293 7970 7525 4499 [I > 3s(I)] 0.030 595 7.6 10.90, 217.70, 8.19, 27.70 20.67, 0.54 0.055 0.051 all hydrogens were located in the difference maps whilst for 6 only H atoms of the Me were located (positions of the H atoms of the C5 ring were calculated). All hydrogens of 6 and 12 were included in the final refinement with fixed positional and thermal parameters.Neutral atom scattering factors were taken from the usual sources.17J. Chem. Soc., Dalton Trans., 1997, Pages 2293–2303 2303 For compounds 2, 3, 7 and 13 the non-hydrogen atoms were located by Patterson (2, 13) and direct methods (3, 7) and Fourier-difference syntheses. A DIFABS18 absorption correction was carried out on all three data sets. The hydrogen atoms were placed in calculated positions in the final cycles of refinement.The structures were refined using full-matrix least squares with anisotropic thermal parameters for all nonhydrogen atoms. All crystallographic calculations were carried out using the CRYSTALS19 program package on a MicroVax 3800 computer (2, refinement against F) or with the SHELXL 93 20 software on a personal computer (3 and 7, refinement against F2).The crystallographic data are given in Tables 4 and 5. For 2, a Chebychev weighting scheme was applied (parameters: 8.89, 25.46, 6.92); 21 for 3 (n = 0.0779) and 7 (n = 0.0403) the weighting scheme was w21 = s2Fo 2 1 (nP)2 1 0.0P, where P = (Fo 2 1 2Fc 2)/3. Only the relevant R values are given in Table 4. Atomic scattering factors were taken from the usual sources.17 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 CCDC for this material should quote the full literature citation and the reference number 186/518. Acknowledgements We thank Dr. J. Cook for helpful discussions and BASF AG/ Studienstiftung des Deutschen Volkes for a post-doctoral fellowship (to N. M.). This work was financially supported by the EPSRC. Access to a Brüker AM 200 NMR instrument (courtesy of Dr. P. Grebenik, Oxford Brookes University) is gratefully acknowledged. We thank the EPSRC Mass Spectroscopy Service (Dr. J. A. Ballantine) for essential assistance. References 1 A. G. Massey, A. J. Park and F. G. A. Stone, Proc. Chem. Soc., 1963, 212; A. G. Massey and A. J. Park, J. Organomet. Chem., 1964, 2, 245. 2 H. H. Brintzinger, D. Fischer, R. Mülhaupt, B. Rieger and R. Waymouth, Angew. Chem., 1995, 107, 1225; Angew. Chem., Int. Ed. Engl., 1995, 34, 1143. 3 M. L. H. Green, J. Haggitt and C. P. Mehnert, J. Chem. Soc., Chem. Commun., 1995, 1853. 4 J. R. Galsworthy, M. L. H. Green, M. Müller and K. Prout, J. Chem. Soc., Dalton Trans., 1997, 1309. 5 H. Nöth and B. Wrackmeyer, Nuclear Magnetic Resonance Spectroscopy of Boron Compounds, Springer, Heidelberg, 1978. 6 F. Correa, R. Nakamura, R. E. Stimson, J. R. L. Burwell and D. F. Shriver, J. Am. Chem. Soc., 1980, 102, 5112; S. B. Butts, E. M. Holt, S. H. Strauss, N. W. Alcock, R. E. Stimson and D. F. Shriver, J. Am. Chem. Soc., 1980, 102, 5093; B. Butts, E. M. Holt, S. H. Strauss, N. W. Alcock, R. E. Stimson and D. F. Shriver, J. Am. Chem. Soc., 1979, 101, 5864. 7 L. Jia, X. Yang, A. Ishihara and T. J. Marks, Organometallics, 1995, 14, 3135. 8 A. R. Seidle, R. A. Newmark, W. M. Lamanna and J. C. Huffman, Organometallics, 1993, 12, 1491. 9 B. Temme, G. Erker, J. Karl, H. Luftmann, R. Fröhlich and S. Kotila, Angew. Chem., 1995, 107, 1867. 10 X. Yang, C. L. Stern and T. J. Marks, Angew. Chem., Int. Ed. Engl., 1992, 31, 1375. 11 F. A. Allen, O. Kennard, D. Watson, L. Brammer, G. Orpen and R. Taylor, J. Chem. Soc., Dalton Trans., 1987, S1. 12 L. Kiplinger, T. G. Richmond and C. E. Osterberg, Chem. Rev., 1994, 94, 373. 13 C. Mehnert, D.Phil Thesis, Oxford, 1996. 14 T. S. Piper and G. Wilkinson, J. Inorg. Nucl. Chem., 1956, 3, 104; R. Birtwhistell, P. Hackett and A. R. Manning, J. Organomet. Chem., 1978, 157, 239. 15 S. R. Drake, M. B. Hursthouse, K. M. Abdul Malik and S. A. S. Miller, Inorg. Chem., 1993, 32, 4653. 16 A. T. C. North, D. C. Phillips and F. S. Matthews, Acta Crystallogr., Sect. A, 1968, 24, 151. 17 International Tables for X-Ray Crystallography, Kynoch Press, Birmingham, 1974. 18 N. Walker and D. Stuart, Acta Crystallogr., Sect. A, 1983, 39, 158. 19 CRYSTALS, D. J. Watkin, J. R. Carruthers and P. W. Betteridge, Oxford University, 1985. 20 SHELXL 93, G. M. Sheldrick, University of Göttingen, 1993. 21 J. S. Rollet, Computing Methods in Crystallography, Pergamon, Oxford, 1965. Received 10th February 1997

ISSN:1477-9226    

DOI:10.1039/a700925i

出版商:RSC    

年代:1997

数据来源: RSC

摘要:

DALTON J. Chem. Soc., Dalton Trans., 1997, Pages 2305–2307 2305 Alkyne addition across imido and sulfido moieties: molecular structure of the 1-thio-2-iminoene complex [Mo{Á2-PhNC(R)] C(R)S}{Á1,Á3-SC(R)] C(R)C(NEt2)S}(S2CNEt2)] (R = CO2Me) Glyn D. Forster and Graeme Hogarth * Chemistry Department, University College London, 20 Gordon Street, London, UK WC1H 0AJ Thermolysis of [Mo(h2-S2)(NPh)(S2CNEt2)2] with dimethyl acetylenedicarboxylate afforded the crystallographically characterised 1-thio-2-iminoene complex [Mo{h2-PhNC(R)]] C(R)S}{h1,h3-SC(R)]] C(R)C(NEt2)S}- (S2CNEt2)] (R = CO2Me), resulting from alkyne addition across sulfido and phenylimido moieties and further insertion of alkyne into the sulfur–carbon bond of a dithiocarbamate ligand.Complexes containing the 1,2-dithiolene ligand have attracted widespread attention primarily due to their ability to form extended p systems resulting from delocalisation of electrons between the metal and ligand.1 In this context, their electronic structures have provoked a great deal of controversy, the dithiolene formulation being the most apt description when bound to low-valent metal centres, while a 1,2-dithiolato form is preferred at high-valent centres.Similarly, a wide range of complexes containing the isoelectronic a-diimine ligand have also been prepared, although here the neutral diimine formulation appears generally to be the best ligand description.2 Somewhat surprisingly, given the extensive chemistry of 1,2- dithiolene and a-diimine ligands, hybrid 1-thio-2-iminoene complexes are relatively rare and appear to be limited specifically to the 2-aminobenzenethiolate ligand.Complexes of this type include [Ni(HNC6H4S)2],3 [Mo(HNC6H4S)2(S2CNEt2)],4 [Mo(HNC6H4S)3],5 [Co(h-C5H5)(HNC6H4S)],6 [Tc(HNC6- H4S)3],7 [TcO(HNC6H4S)2]2 8 and [Re(HNC6H4S)3].9 This limitation to the benzene backbone is a consequence of the use of commercially available 2-aminobenzenethiol as the ligand precursor.While a number of synthetic methods have been adopted towards the synthesis of 1,2-dithiolene complexes, one often used is the addition of activated alkynes across two sulfi- do moieties 10 or the sulfur–sulfur bond of a disulfide ligand.11,12 Potentially, this approach could be modified to produce 1-thio- 2-iminoene complexes via alkyne addition across sulfido and imido moieties, however to date this has not been achieved. Dithiocarbamate is widely employed in transition-metal chemistry as a stable ligand which is capable of supporting metal centres in a wide range of oxidation states.13 There are, however, an increasing number of examples in which the ligand is found to act in a non-innocent fashion, generally resulting from the cleavage of one or both carbon–sulfur bonds.14,15 Recently, Young et al.12 described novel transformations of a dithiocarbamate ligand in [MoO(h2-S2)(S2CNR92)2] (R9 = Me or Et) upon reaction with the activated alkyne dimethyl acetylenedicarboxylate (dmad). At elevated temperatures, two 1,2- dithiolato complexes, [Mo{h2-SC(R)]] C(R)S}{h3-C(NR92)SC- (R)]] C(R)S}(S2CNR92)] and [Mo{h2-SC(R)]] C(R)S}{h1,h3- SC(R)]] C(R)C(NR92)S}(S2CNR92)] (R = CO2Me), were formed.Both result from addition of alkyne across a disulfide ligand and also via insertion of further alkyne into a sulfur–carbon bond of the dithiocarbamate. This latter process was termed ‘melding’ and affords novel new bidentate ligands at the molybdenum(VI) centre. We recently reported a novel route to a wide range of imidodisulfido complexes [Mo(h2-S2)(NR)- (S2CNEt2)2].15 During our attempts to use imido complexes in carbon–nitrogen bond-formation reactions, and prompted by the report of Young et al.12 detailed above, we investigated the reaction of [Mo(h2-S2)(NPh)(S2CNEt2)2] with dmad.In a recent paper 15 we briefly described that this gave a mixture of products, the major component of which was the 1-thio-2- iminoene complex [Mo{h2-PhNC(R)]] C(R)S}{h1,h3-SC(R)]] C(R)C(NEt2)S}(S2CNEt2)] 1 (R = CO2Me).Since such complexes have not previously been reported, we sought to substantiate this formulation and herein report further details of this reaction together with the crystal structure of 1. Results and Discussion As previously described,15 thermolysis of [Mo(h2-S2)(NPh)- (S2CNEt2)2] with an excess of dmad in toluene affords bright blue [Mo{h2-PhNC(R)]] C(R)S}{h1,h3-SC(R)]] C(R)C(NEt2)- S}(S2CNEt2)] 1 (R = CO2Me) in approximately 40% yield after chromatography (Scheme 1).In our earlier report, a further green product was thought to be an isomer of 1, however we now believe that the latter is a mixture of 1 and an as yet unidentified yellow product which does not contain an NPh group. Slow diffusion of methanol into a saturated dichloromethane solution of 1 afforded small blue crystals suitable for X-ray crystallography, the results of which are summarised in Fig. 1 and Table 1. The molecule consists of a highly distorted eight-co-ordinate molybdenum centre ligated by three chelating ligands via five sulfur, one nitrogen and two carbon atoms. One of the dithiocarbamates remains intact and bond lengths and angles are within the expected ranges. An equivalent of alkyne has inserted into a carbon–sulfur bond of the second dithiocarbamate ligand to give a novel, 3-diethylamino-1,2-bis- (methoxycarbonyl)-3-thioxoprop-1-ene-1-thiolate ligand.This is bound to molybdenum via an heteroallylic-type interaction [Mo]S(4) 2.3729(10), Mo]C(43) 2.184(3), Mo]C(42) 2.312(3) Å] and also through a further sulfur atom [Mo]S(3) 2.5582(10) Å]. It is analogous to the related N,N-dimethyl ligand crystallographically characterised by Young et al. 12 in [Mo{h2-SC- (R)]] C(R)S}{h1,h3-SC(R)]] C(R)C(NMe2)S}(S2CNMe2)] (R = Scheme 12306 J. Chem. Soc., Dalton Trans., 1997, Pages 2305–2307 CO2Me) and bond lengths and angles do not vary significantly between the two.The third chelating ligand is formed via addition of alkyne across the sulfur and phenylimido moieties, the fate of the second sulfur atom being unknown. The new 1-thio-2-iminoene ligand binds to the molybdenum centre with a bite angle of 79.49(8)8 and both molybdenum–sulfur [Mo]S(5) 2.3382(9) Å] and molybdenum–nitrogen bonds [Mo]N(3) 2.125(3) Å] are relatively short, being indicative of some multiple-bond character.Indeed the former lies close to those found in Young’s related dithiolato complexes [average Mo]S 2.340(3) Å] 12 indicating that the thiolato (I) rather than a thiolene (II) formulation is more apt. The Mo]N bond is, however, considerably longer than those found in the related molybdenum(V) [average Mo]N 2.001(2) Å] 4 and molybdenum(VI) [average Mo]N 1.997(8) Å] 5 2-aminobenzenethiolate ligands, while the Mo]S bond is somewhat shorter. Indeed, the difference between these two bond lengths (n) in 1 of 0.213 Å is significantly less than found for the 2-aminobenzenethiolate ligand at MoVI (n 0.421 Å),5 MoV (n 0.371 Å),4 TcVI (n 0.35 Å)7 and CoIII (n 0.349 Å),6 but comparable to that at TcV (n 0.21 Å).8 The reason for this difference is not immediately apparent, but may in some part be due to the relatively bulky phenyl group at nitrogen in 1 and/or the trans influence of the melded ligand. An imino formulation I is, however, supported by the approximate planarity of the Mo]S(5)]C(32)]C(33)]N(3) ring [maximum deviation 0.15 Å by S(5)], the planar nature of N(3) and the short carbon– carbon bond [C(32)]C(33) 1.367(5) Å].The latter is close to values of 1.33(1) and 1.343(5) Å found by Young in the related dithiolato complexes and is indicative of substantial doublebond character. The phenyl substituent lies approximately at right angles to the latter which may be a result of adverse steric interactions with the methyl ester bound to C(33).This is further manifested in the 1H NMR spectrum with the appearance of a complex multiplet in the aromatic region even at 50 8C, which, while difficult to interpret precisely, is clearly indicative of the inequivalence of all phenylic protons, and presumably results from the restricted rotation about the carbon– nitrogen bond. The oxidation state of the molybdenum centre in complex 1 Fig. 1 Molecular structure of complex 1 is not clear, however based on its diamagnetism and the effective oxidation of sulfido and imido centres upon alkyne addition a 14 formulation appears most probable.In conclusion, complex 1 represents the first example of a non-benzene 1-thio-2-iminoene complex which is formed via the novel addition of an alkyne across sulfido and imido moieties. Structural data and physical properties suggest that this hybrid ligand is more akin to a 1,2-dithiolato than an a-diimine ligand. Experimental Complex 1 was prepared as previously described.15 Small blue well formed crystals were grown upon slow diffusion of methanol into a saturated dichloromethane solution at room temperature.A single crystal was mounted on a glass fibre and all geometric and intensity data were taken from this sample using an automated four-circle diffractometer (Nicolet R3mV) equipped with Mo-Ka radiation (l = 0.710 73 Å). The lattice parameters were identified by application of the automatic indexing routine of the diffractometer to the positions of a number of reflections taken from a rotation photograph and centred by the diffractometer. The w–2q technique was used to measure reflections in the range 5 < 2q < 508.Three standard reflections (remeasured every 97 scans) showed no significant loss in intensity during data collection. The data were corrected for Lorenz-polarisation effects and empirically for absorption. The unique data with I > 2.0s(I) were used to solve and refine the structure.The structure was solved by direct methods and developed by using alternative cycles of least-squares refinement (based on F 2) and Fourier-difference synthesis. All nonhydrogen atoms were refined anisotropically. Hydrogens were placed in idealised positions (C]H 0.96 Å) and assigned a common isotropic thermal parameter (U = 0.08 Å2). Final Fourier-difference maps were featureless and contained no peaks greater than 1.00 e Å23. Structure solution used the SHELXTL PLUS program package 16 on a micro Vax II computer.The methyl carbons on the dithiocarbamate ligand were disordered over two sites with equal occupancies, C(3)/ C(3a) and C(5)/C(5a). Important crystallographic parameters; triclinic, space group P1� , a = 10.7178(17), b = 12.8823(14), c = 12.9820(21) Å, a = 96.38(1), b = 95.03(1), g = 91.37(1)8, U = 1773.4(4) Å3, Z = 2, Dc = 1.50 g cm23, F(000) = 824, m(Mo- Ka) = 6.90 cm21, crystal dimensions 0.62 × 0.48 × 0.21 mm, R(R9) = 0.041(0.045) for the 5168 unique reflections and 406 variables, 0.049(0.056) for all 6173 data.Atomic coordinates, thermal parameters, and bond lengths and angles have been deposited at the Cambridge Crystallographic Data Centre (CCDC). See Instructions for Authors, Table 1 Bond lengths (Å) and angles (8) in complex 1 Mo]S(1) Mo]S(3) Mo]S(5) Mo]C(42) S(1)]C(1) C(1)]N(1) C(6)]N(2) C(42)]C(43) N(3)]C(33) S(5)]C(32) C(20)]C(21) C(22)]C(23) C(24)]C(25) S(1)]Mo]S(2) S(5)]Mo]N(3) N(3)]C(33)]C(32) Mo]S(5)]C(32) 2.5414(11) 2.5582(10) 2.3382(9) 2.312(3) 1.704(5) 1.334(7) 1.308(4) 1.446(4) 1.349(5) 1.737(4) 1.372(7) 1.347(11) 1.408(9) 68.23(4) 79.49(8) 119.8(3) 99.87(13) Mo]S(2) Mo]S(4) Mo]N(3) Mo]C(43) S(2)]C(1) S(3)]C(6) C(6)]C(42) C(43)]S(4) C(32)]C(33) N(3)]C(20) C(21)]C(22) C(23)]C(24) C(25)]C(20) S(3)]Mo]S(4) Mo]N(3)]C(33) C(33)]C(32)]S(5) 2.5507(11) 2.3729(10) 2.125(3) 2.184(3) 1.717(5) 1.699(4) 1.493(5) 1.735(3) 1.367(5) 1.447(5) 1.392(7) 1.366(11) 1.392(7) 125.13(3) 118.9(2) 117.1(3)J.Chem. Soc., Dalton Trans., 1997, Pages 2305–2307 2307 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 186/525. Acknowledgements This work was supported by a grant from the SERC (studentship to G. D. F.). References 1 U. T. Mueller-Westerhoff and B. Vance, Comprehensive Coordination Chemistry, eds. G. Wilkinson, R.D. Gillard and J. A. McCleverty, Pergamon, Oxford, 1973, vol. 2, p. 595; J. A. McCleverty, Prog. Inorg. Chem., 1968, 10, 49; S. Alvarez, R. Vincente and R. Hoffmann, J. Am. Chem. Soc., 1985, 107, 6253. 2 G. van Koten and K. Vrieze, Adv. Organomet. Chem., 1982, 21, 151. 3 R. H. Holm, A. L. Balch, A. Davison, A. H. Maki and T. E. Berry, J. Am. Chem. Soc., 1967, 89, 2866. 4 K. Yamanouchi and J. H. Enemark, Inorg. Chem., 1978, 17, 1981. 5 K. Yamanouchi and J. H. Enemark, Inorg.Chem., 1978, 17, 2911. 6 E. J. Miller, A. L. Rheingold and T. B. Brill, J. Organomet. Chem., 1985, 282, 399. 7 J. Baldas, J. Boas, J. Bonnyman, M. F. Mackay and G. A. Williams, Aust. J. Chem., 1982, 35, 2413. 8 G. Bandoli and T. I. A. Gerber, Inorg. Chim. Acta, 1987, 126, 205. 9 J. K. Gardner, N. Pariyadath, J. L. Corbin and E. I. Steifel, Inorg. Chem., 1978, 17, 2911. 10 J. T. Goodman, S. Inomata and T. B. Rauchfuss, J. Am. Chem. Soc., 1996, 118, 11 674; C. M. Bolinger, T.B. Rauchfuss and A. L. Rheingold, J. Am. Chem. Soc., 1983, 105, 6321; D. Coucouvanis, D. Toupadakis, J. D. Lane, S. M. Koo, C. G. Kim and A. Hadjikyriacou, J. Am. Chem. Soc., 1991, 113, 5271; A. A. Eagle, S. Harben, E. R. T. Tiekink and C. G. Young, J. Am. Chem. Soc., 1994, 116, 9749. 11 D. Coucouvanis, A. Hadjikyriacou and M. G. Kanatzidis, Polyhedron, 1986, 5, 349; T. R. Halbert, W.-H. Pan and E. I. Stiefel, J. Am. Chem. Soc., 1983, 105, 5476; T. B. Rauchfuss, D. P. S.Rodgers and S. R. Wilson, J. Am. Chem. Soc., 1986, 108, 3114; C. M. Bolinger, J. E. Hoots and T. B. Rauchfuss, Organometalics, 1982, 1, 223; M. Rakowski DuBois, B. R. Jagirdar, S. Dietz and B. C. Noll, Organometallics, 1997, 16, 294. 12 C. G. Young, X. F. Yan, B. L. Fox and E. R. T. Tiekink, J. Chem. Soc., Chem. Commun., 1994, 2579. 13 A. M. Bond and R. L. Martin, Coord. Chem. Rev., 1984, 54, 23. 14 L. Ricard, J. Estienne and R. Weiss, J. Chem. Soc., Chem. Commun., 1972, 906; R. S. Herrick, S. Neiter-Burgmayer and J. L. Templeton, J. Am. Chem. Soc., 1983, 105, 2599; D. C. Brower, T. L. Tonker, J. R. Morrow, D. S. Rivers and J. L. Templeton, Organometallics, 1986, 5, 1093; J. R. Morrow, T. L. Tonker and J. L. Templeton, Organometallics, 1985, 4, 745; P. F. Gilletti, D. A. Femec, F. I. Keen and T. M. Brown, Organometallics, 1992, 11, 4008; A. Mayr, G. A. McDermott, A. M. Dorries and A. K. Holder, J. Am Chem. Soc., 1986, 108, 310; P. R. Heckley, D. G. Holah and D. Brown, Can. J. Chem., 1971, 49, 1151; A. J. Deeming and R. Vaish, J. Organomet. Chem., 1993, 460, C8. 15 T. A. Coffey, G. D. Forster and G. Hogarth, J. Chem. Soc., Dalton Trans., 1996, 183. 16 G. M. Sheldrick, SHELXTL PLUS, program package for structure solution and refinement, version 4.2, Siemens Analytical Instruments Inc., Madison, WI, 1990. Received 27th February 1997; Paper 7/0139

ISSN:1477-9226    

DOI:10.1039/a701398a

出版商:RSC    

年代:1997

数据来源: RSC

摘要:

DALTON J. Chem. Soc., Dalton Trans., 1997, Pages 2309–2314 2309 The promotion of air oxidation and carbonyl substitution reactions of [(Á5-C5H5)2Mo2(CO)4(Ï-RCCR9)] by trimethylamine N-oxide or nitrosonium tetrafluoroborate: the synthesis of [(Á5-C5H5)2Mo2(O)2- (Ï-O)(Ï-RCCR9)] (R 5 R9 = CO2Me or Ph; R 5 H, R9 5 Ph; or R 5 H, R9 5 CO2Me) and [(Á5-C5H5)2Mo2(CO)3(NO)(Ï-RCCR9)]- [BF4] (R 5 R9 5 CO2Me, Me or H; or R 5 H, R9 5 CO2Me) Joanne C.Stichbury, Martin J. Mays,* John E. Davies, Paul R. Raithby and Gregory P. Shields Department of Chemistry, University of Cambridge, Lensfield Road, Cambridge, UK CB2 1EW Trimethylamine N-oxide promotes the air oxidation of [(h5-C5H5)2Mo2(CO)4(m-RCCR9)] to afford the novel organometallic oxo complexes [(h5-C5H5)2Mo2(O)2(m-O)(m-RCCR9)] [R = R9 = CO2Me or Ph; R = H, R9 = Ph; or R = H, R9 = CO2Me).Trimethylamine N-oxide also promotes the near quantitative substitution of one of the carbonyl groups in [(h5-C5H5)2Mo2(CO)4(m-MeO2CC2CO2Me)] by either ButNC or (MeO)3P (L), giving [(h5-C5H5)2- Mo2(CO)3(L)(m-MeO2CC2CO2Me)] at room temperature. The addition of [NO][BF4], in place of Me3NO, to solutions of [(h5-C5H5)2Mo2(CO)4(m-RCCR9)] under the same reaction conditions, does not promote air oxidation but leads instead to the new dimolybdenum nitrosyl complexes [(h5-C5H5)2Mo2(CO)3(NO)(m-RCCR9)][BF4] (R = R9 = CO2Me, Me or H; or R = H, R9 = CO2Me).The structure of [(h5-C5H5)2Mo2(CO)3(NO)(m-MeO2CC2CO2- Me)][BF4] has been determined by X-ray crystallography. Owing to their numerous catalytic applications, organometallic oxo complexes of molybdenum have received much attention.1 Dinuclear organometallic oxo complexes of molybdenum have been prepared previously by oxidation of the corresponding organometallic carbonyl complexes,2 but the synthesis of a dinuclear organometallic molybdenum oxo complex incorporating a transverse bridging alkyne has not been reported.As part of our investigations into the chemistry of alkyne bridged Group 6 metal complexes 3 we have now attempted the synthesis of such an oxo complex from the alkyne bridged dimolybdenum carbonyl complexes [(h5-C5H5)2Mo2(CO)4(m-RCCR9)] (R = R9 = CO2Me or Ph; R = H, R9 = Ph; or R = H, R9 = CO2Me).The conventional method of preparing organometallic oxo complexes from carbonyl complexes involves a thermally or photochemically induced reaction in the presence of air or oxygen.4 However, the complexes [(h5-C5H5)2Mo2(CO)4- (m-RCCR9)] are stable under such reaction conditions and it was decided to use trimethylamine N-oxide in the presence of oxygen, in an attempt to labilise the carbonyl ligands and facilitate their replacement by oxo groups.The ability of amine Noxides to act as decarbonylation agents for carbonyl substitution in transition-metal complexes is well documented.5,6 Results and Discussion Herein we report the synthesis of the new complexes [(h5- C5H5)2Mo2(O)2(m-O)(m-RCCR9)] (R = R9 = CO2Me 2 or Ph 3; R = H, R9 = Ph 4; or R = H, R9 = CO2Me 5), via the air oxidation of [(h5-C5H5)2Mo2(CO)4(m-RCCR9)] 7 in the presence of trimethylamine N-oxide.In the absence of air, oxidation does not take place but, if ButNC or (MeO)3P is present, the trimethylamine N-oxide promotes monosubstitution of [(h5-C5- H5)2Mo2(CO)4(m-RCCR9)] to form [(h5-C5H5)2Mo2(L)(CO)3- (m-RCCR9)] [R = R9 = CO2Me; L = ButNC 7 or (MeO)3P 8] in virtually quantitative yields at room temperature. A preliminary account of part of this work has been published.8 Synthesis of [(Á5-C5H5)2Mo2(O)2(Ï-O)(Ï-RCCR9)] (R = R9 = CO2Me 2 or Ph 3; R = H, R9 = Ph 4; or R = H, R9 = CO2Me 5) The complexes [(h5-C5H5)2Mo2(O)2(m-O)(m-RCCR9)] (R = R9 = CO2Me 2 or Ph 3; R = H, R9 = Ph 4 or R = H, R9 = CO2Me 5) were prepared by air oxidation of the parent alkyne bridged carbonyl complexes [(h5-C5H5)2Mo2(CO)4(m-RCCR9)] 7 in a 1 : 1 acetonitrile–dichloromethane solution containing an excess of freshly sublimed trimethylamine N-oxide (Scheme 1).Air oxidation of a similar solution in the absence of trimethylamine N-oxide resulted only in decomposition products and starting material after the same reaction time. In our preliminary communication 8 we reported the molecular structure of 2, [(h5-C5H5)2Mo2(O)2(m-O)(m-MeO2CC2CO2- Me)], which was determined by X-ray diffraction.The quasi-tetrahedral structure of the central Mo2C2 moiety in 2 corresponds to that of the related parent carbonyl complexes [(h5-C5H5)2Mo2(CO)4(m-RCCR9)] (R = R9 = H, Et or Ph).7,9 Complex 2 has been fully characterised by IR, 1H and 13C-{1H} NMR spectroscopy, fast atom bombardment (FAB) mass spectrometry and microanalysis as described in the Experimental section.The complexes [(h5-C5H5)2Mo2(O)2(m-O)(m-RCCR9)] (R = R9 = Ph 3; R = H, R9 = Ph 4; or R = H, R9 = CO2Me 5) were characterised by IR, 1H and 13C-{1H} NMR spectroscopy Scheme 1 (i) Excess Me3NO, air, 1 :1 MeCN:CH2Cl2, 298 K2310 J. Chem. Soc., Dalton Trans., 1997, Pages 2309–2314 and FAB mass spectrometry, the data being closely comparable with those for 2.Although a yield of 31% was obtained for complex 2 (R = R9 = CO2Me), the yields of the analogous complexes 3, 4 and 5 with less electron withdrawing alkyne substituents were poor, the highest being 17% for 5 (R = H, R9 = CO2Me). It appears that the yield of [(h5- C5H5)2Mo2(O)2(m-O)(m-RCCR9)] is maximised by electron withdrawing substituents on the alkyne, perhaps because this reduces p-back bonding from the metal to the carbonyl ligands thus increasing their lability.This suggested that the yield might be improved further by the introduction of electron withdrawing substituents onto the cyclopentadienyl ligands in addition to those on the alkyne. To test this suggestion, the complexes [(h5-C5H4CO2Me)2Mo2(CO)4(m-RCCR9)] 10 (R = R9 = CO2Me and R = H, R9 = CO2Me) were synthesised and oxidised in the presence of trimethylamine N-oxide.However, the complexes rapidly suffered total decomposition to give a mixture of intractable ‘isopolymolybdates’.11 This decomposition is perhaps initiated by nucleophilic attack of the trimethylamine Noxide on the carbomethoxy ring substituents, although the CO2Me groups on the alkyne in [(h5-C5H5)2Mo2(CO)4(m- RCCR9)] (R = R9 = CO2Me or R = H, R9 = CO2Me) are presumably not attacked in this way.An alternative explanation is that the oxidised products or reaction intermediates are destabilised by the presence of the ring substituents which may withdraw electron density from the metal centres to an extent which weakens the bonds between these centres and the alkyne ligand.The oxo ligands in complexes 2–5 could originate from the oxygen released by decomposition of excess trimethylamine N-oxide or from the air bubbled through the solution. In order to investigate this, a ‘control’ reaction was carried out on [(h5- C5H5)2Mo2(CO)4(m-MeO2CC2CO2Me)] 1 using similar reaction conditions, but purging the solution of 1 with argon rather than air. After 18 h the colour of the solution had changed and the IR spectra, taken in situ, showed carbonyl vibrations not characteristic of the starting material.The new complex giving rise to these absorptions may result from substitution of one or more carbonyl ligands of 1 by acetonitrile and would then have the formula [(h5-C5H5)2Mo2(CO)(42x)(CH3CN)x(m-MeO2CC2- CO2Me)] 6. A complex, thought to be [(h5-C5H5)Mo(CO)2- (CH3CN)I], was similarly obtained by dissolution of [(h5- C5H5)Mo(CO)3I] in acetonitrile in the presence of (CH3)3NO.12 Addition of ButNC or (MeO)3P to 6 in situ gave almost quantitative yields of the monosubstituted complexes [(h5-C5H5)2Mo2- (CO)3(L)(m-RCCR9)] [L = ButNC or (MeO)3P], described in more detail below.This result suggests that x = 1 in the acetonitrile complex 6, but unfortunately it could not be isolated.On exposure to air it formed mainly intractable decomposition products and a small quantity of the oxo complex 2. The probable formation of an acetonitrile substituted complex, instead of direct formation of 2 in the above reaction, demonstrates that the air purge is necessary in order for the oxo species to form and hence that the oxo ligands are derived from this source.Complex 2, [(h5-C5H5)2Mo2(O)2(m-O)(m-MeO2CC2CO2Me)], is stable towards substitution, thermolysis and photolysis. It was not possible to replace the oxo ligands with CO, even at high pressure and temperature [100 atm (1013250 Pa) CO, 150 8C]. Protonation with HCl and trifluoroacetic acid was investigated for comparative purposes with the corresponding known reactions of the parent carbonyl species 13 but treatment with either acid resulted in immediate decomposition. A well known route to the preparation of imido species is the treatment of an analogous oxo complex with an isocyanate.14 No reaction was observed on treating complex 2 with phenyl isocyanate under reduced pressure at 120 8C for several days, conditions similar to those used by Green and co-workers 15 to prepare bis(cyclopentadienyl) dimolybdenum imido complexes from their corresponding dimolybdenum oxo analogues.The use of trimethylamine N-oxide in the synthesis of [(Á5-C5- H5)2Mo2(CO)3(L)(Ï-RCCR9)] [L = ButNC or (MeO)3P] under controlled, mild conditions It was of interest to investigate the extent to which ligands other than oxygen might replace the carbonyl groups in [(h5- C5H5)2Mo2(CO)4(m-MeO2CC2CO2Me)] 1 in the presence of trimethylamine N-oxide.Acetonitrile solutions of 1 were treated separately with excess ButNC or (MeO)3P in the presence of Me3NO, under nitrogen. In both cases a single monosubstituted product [(h5-C5H5)2Mo2(CO)3(L)(m-RCCR9)] [R = R9 = CO2Me; L = ButNC 7 or (MeO)3P 8] was obtained in high yield at room temperature (Scheme 2).These complexes have been fully characterised by IR, 1H and 13C-{1H} NMR spectroscopy, FAB mass spectrometry and microanalysis as described in the Experimental section. A thermolytic synthesis giving a 71% yield of [(h5-C5Me5)2Mo2(CO)3(ButNC)(m-MeO2CC2CO2Me)], analogous to 7, has previously been described.16 However the milder conditions afforded here by the use of trimethylamine N-oxide enable near quantitative yields to be achieved.The complex [(h5-C5H5)2Mo2(CO)3{P(OMe)3}(m-HCCH)], analogous to 8, has previously been synthesised photolytically as part of a study on the cleavage of monophosphine ligands at a dimolybdenum centre, although the yield was not reported.17 Complexes 2–5 are the first complexes reported in which more than one CO group in [(h5-C5H5)2Mo2(CO)4(m-RCCR9)] has been replaced by other ligands, without transformation of the alkyne bridge. To determine whether ligands other than oxo groups could multiply substitute the CO groups in [(h5- C5H5)2Mo2(CO)4(m-RCCR9)], the monosubstituted complexes 7 and 8 were treated with respectively an excess of either ButNC or (MeO)3P under more forcing conditions.In each case the complexes were (i) refluxed in acetonitrile in the presence of excess (CH3)3NO or (ii) refluxed in toluene. No evidence for further substitution was obtained in any of the experiments. It is possible that steric factors may hinder multiple substitution of [(h5-C5H5)2Mo2(CO)4(m-RCCR9)] by such ligands as ButNC or (MeO)3P.A crystallographic determination of the molecular structure of the unsubstituted complexes [(h5-C5H5)2- Mo2(CO)4(m-RCCR9)] (R = R9 = H, Et or Ph) showed that one of the carbonyl ligands adopts a semi-bridging site close to the metal–metal bond.7,9 This was taken as evidence of a high degree of internal crowding in the molecule due to the large solid angle required by the h5-C5H5 ligands.Multiple substitution of [(h5-C5H5)2Mo2(CO)4(m-RCCR9)] by ButNC or (MeO)3P may also be unfavourable due to additional p-back bonding to the remaining three carbonyl ligands on monosubstitution of [(h5-C5H5)2Mo2(CO)4(m-RCCR9)] by ButNC or (MeO)3P. This additional p-back bonding will reduce the lability of the CO groups towards further substitution. The reason oxo ligands can multiply substitute CO groups in [(h5-C5H5)2Mo2(CO)4(m-RCCR9)] may be that substitution of a CO ligand by O, which increases the formal oxidation state of the metal, thereby reduces the electron density around the metal centre.In turn, this will reduce the p-back bonding to the remaining CO ligands, facilitating further substitution. There is no evidence for the formation of any mixed carbonyl–oxo species during the course of the oxidation reactions.It follows that once one Scheme 2 R = R9 = CO2Me. (i) Excess Me3NO, ButNC (7) or (MeO)3P (8), MeCN, 298 KJ. Chem. Soc., Dalton Trans., 1997, Pages 2309–2314 2311 CO is replaced by O any mixed carbonyl oxo intermediates are unstable with respect to formation of the final product. The attempted oxidation of [(Á5-C5H5)2MoW(CO)4(Ï-RCCR9)] The mixed-metal transverse alkyne bridged complex [(h5- C5H5)2MoW(CO)4(m-RCCR9)] has been prepared previously and its reactivity investigated.13,18 It was of interest to investigate the susceptibility of this complex towards air oxidation in the presence of trimethylamine N-oxide since if the integrity of the complex were to be maintained during the oxidation, a single heterobimetallic oxo product [(h5-C5H5)MoW(O)2(m-O)- (m-RCCR9)] would result.In fact, air oxidation of [(h5- C5H5)2MoW(CO)4(m-MeO2CC2CO2Me)] in the presence of trimethylamine N-oxide led only to the formation of decomposition products and [(h5-C5H5)2Mo2(O)2(m-O)(m-RCCR9)] 2. It follows that metal–metal bond cleavage and dissociation into monometallic fragments must have occurred under the reaction conditions.Even so, the mixed-metal oxo complex might have been expected as one product, and the fact that it was not obtained indicates that it may not be stable. The reaction of [(Á5-C5H5)2Mo2(CO)4(Ï-RCCR9)] with [NO]- [BF4] to form the monosubstituted products [(Á5-C5H5)2Mo2- (CO)3(NO)(Ï-RCCR9)][BF4] (R = R9 = CO2Me 9, Me 10 or H 11; or R = H, R9 = CO2Me 12) In an attempt to find a carbonyl labilising agent which might be more efficient than trimethylamine N-oxide in converting [(h5- C5H5)2Mo2(CO)4(m-RCCR9)] into [(h5-C5H5)2Mo2(O)2(m-O)- (m-RCCR9)], the use of [NO][BF4] was explored.This reagent was considered because of its potential ability to bring about chemical oxidation of one or both of the metal centres of [(h5- Scheme 3 (i) Excess [NO][BF4], CH2Cl2, 298 K Fig. 1 Molecular structure of the cation [(h5-C5H5)2Mo2(CO)3(NO)(m- RCCR9)]1 in complex 9. Only one component of the disordered OMe group is shown for clarity C5H5)2Mo2(CO)4(m-RCCR9)] and thereby to reduce the degree of metal p-back donation to the carbonyl ligands. It had previously been found that chemical oxidation of [(h5-C5Me5)2- Mo2(CO)3(ButNC)(m-MeO2CC2CO2Me)] with [(h5-C5H5)2Fe]- [PF6] gave a sensitive violet product, shown by in situ IR and ESR spectroscopy to be the monocation [(h5-C5Me5)2Mo2- (CO)3(ButNC)(m-MeO2CC2CO2Me)]1.16 It is known that the compound [NO][BF4] may act either as a chemical oxidising agent or as the source of the nitrosyl ligand NO1, a twoelectron donor ligand, isoelectronic with CO.19 Treatment of [(h5-C5H5)2Mo2(CO)4(m-RCCR9)] with [NO]- [BF4] in the presence of air, did not lead to the formation of oxo products analogous to those obtained in the corresponding reaction of [(h5-C5H5)2Mo2(CO)4(m-RCCR9)] with Me3NO. Instead, the sole products obtained were the monosubstituted nitrosyl species [(h5-C5H5)2Mo2(CO)3(NO)(m-RCCR9)][BF4] (R = R9 = CO2Me 9, Me 10 or H 11; or R = H, R9 = CO2Me 12) (Scheme 3).It was not established in these reactions whether one-electron chemical oxidation of [(h5-C5H5)2Mo2(CO)4- (m-RCCR9)] by [NO][BF4] takes place and is then followed by reaction of the oxidised product with NO generated as part of the redox process, or whether the carbonyl ligands are directly substituted by NO1. The molecular structure of 9 has been determined by X-ray crystallography and is shown in Fig. 1 with selected bond lengths and angles listed in Table 1. To our knowledge, this is the first example of an alkyne bridged bis(cyclopentadienyl) dimolybdenum complex containing a nitrosyl ligand. Indeed there are relatively few known examples of dimolybdenum nitrosyl complexes,20 the only other dimolybdenum nitrosyls to have been prepared by the use of [NO][BF4] being the recently reported complexes [Mo2(m-H){m-Ph2P(CH2)nPPh2}(CO)7(NO)] (n = 1–4).21 Complex 9 displays the same quasi-tetrahedral central Mo2C2 moiety as is present in 2, while the Mo]Mo bond length [2.9762(9) Å] is comparable with those reported for the parent carbonyl complexes [(h5-C5H5)2Mo2(CO)4(m-RCCR9)] [R = R9 = H 2.984(1); Et 2.977(1); Ph 2.956(1) Å].7,9 In the parent complexes the most noteworthy structural aspect is the lack of any symmetry in the arrangement of the ligands around the Mo2C2 core with each of the four CO groups lying in a different environment, one of them in a semi-bridging position.The unsymmetrical arrangement is believed to result from internal crowding in the molecule due to the large solid angle required by the h5-C5H5 ligands.Because of the crowding, one CO group is forced into the region close to the Mo]Mo bond, opposite the bridging alkyne, and in consequence the rest of the CO ligands are arranged in an unsymmetrical fashion. As an example, in the complex [(h5-C5H5)2Mo2(CO)4(m-HCCH)] the metal to carbonyl bond lengths were found to be 1.993(4), 1.989(3), 1.951(4) and 1.953(4) Å, the shortest distance being that for the metal to semi-bridging carbonyl.7 The angles Mo]C]O for the four carbonyls were found to be 178.5(3), 179.3(4), 178.9(5) and 168.5(3)8, the most acute angle being that for the semi-bridging carbonyl. In complex 9 the unsymmetrical arrangement of ligands is still apparent, the substituted nitrosyl ligand having taken the semi-bridging site; this is not unexpected since the nitrosyl ligand is a better p acceptor than the CO ligands. The metal to carbonyl bond lengths in complex 9 were found to be 2.051(9), 2.046(9) and 2.019(9) Å with a molybdenum to semibridging nitrosyl distance of 1.824(6) Å.The Mo]C]O angles in complex 9 were observed to be 176.3(9), 176.4(7) and 177.2(8)8, while the metal–semi-bridging nitrosyl angle was found to be considerably more acute [162.9(6)8].Complex 9 was also characterised by IR, 1H and 13C-{1H} NMR spectroscopy, FAB mass spectrometry and microanalysis. The IR spectrum in dichloromethane solution was compatible with substitution of the semi-bridging carbonyl by a nitrosyl. In the parent species [(h5-C5H5)2Mo2(CO)4- (m-MeO2CC2CO2Me)] 1 carbonyl absorptions are observed at2312 J.Chem. Soc., Dalton Trans., 1997, Pages 2309–2314 2022, 1969 and 1849 cm21, the lowest frequency absorption being attributed to the semi-bridging carbonyl. The IR spectrum of complex 9 possessed two terminal carbonyl absorptions, shifted to higher frequencies at 2071 and 2032 cm21 due to the positive charge on the complex, but the semi-bridging carbonyl absorption observed in 1 was no longer present.The nitrosyl absorption 22 was observed at 1636 cm21. The analogous complexes [(h5-C5H5)2Mo2(CO)3(NO)- (m-RCCR9)][BF4] (R = R9 = Me 10 or H 11; or R = H, R9 = CO2Me 12) were characterised by IR and 1H NMR spectroscopy, FAB mass spectrometry and microanalysis. By comparison with the data for complex 9 it was determined that the nitrosyl ligand occupies the semi-bridging site in each case.In no case was there any evidence for multiple substitution of the carbonyl ligands by the nitrosyl group or for the formation of oxo products. To determine whether the remaining carbonyl ligands in [(h5- C5H5)2Mo2(CO)3(NO)(m-MeO2CC2CO2Me)][BF4] 9 were labile towards substitution by (i) oxo or (ii) phosphine ligands, an acetonitrile solution of 9 was purged with air for 18 h.After this time it was found that no reaction had occurred other than some decomposition. Separate acetonitrile solutions of 9 were treated with (MeO)3P or Ph3P and stirred under nitrogen. No substitution products were obtained, even after the temperature Table 1 Selected bond lengths (Å) and angles (8) for [(h5-C5H5)2- Mo2(CO)3(NO)(m-MeO2CC2CO2Me)][BF4] 9 Mo(1)]Mo(2) Mo(1)]C(3) Mo(1)]C(4) Mo(1)]C(7) Mo(1)]C(8) Mo(1)]C(16) Mo(1)]C(17) Mo(1)]C(18) Mo(1)]C(19) Mo(1)]C(20) Mo(2)]N(1) Mo(2)]C(2) C(4)]Mo(1)]C(3) C(4)]Mo(1)]C(7) C(4)]Mo(1)]C(8) C(3)]Mo(1)]C(7) C(3)]Mo(1)]C(8) C(7)]Mo(1)]C(8) C(3)]Mo(1)]Mo(2) C(4)]Mo(1)]Mo(2) C(7)]Mo(1)]Mo(2) C(8)]Mo(1)]Mo(2) C(16)]Mo(1)]Mo(2) C(17)]Mo(1)]Mo(2) C(18)]Mo(1)]Mo(2) C(19)]Mo(1)]Mo(2) C(20)]Mo(1)]Mo(2) N(1)]Mo(2)]C(2) N(1)]Mo(2)]C(7) N(1)]Mo(2)]C(8) C(2)]Mo(2)]C(7) C(2)]Mo(2)]C(8) C(7)]Mo(2)]C(8) N(1)]Mo(2)]Mo(1) C(2)]Mo(2)]Mo(1) C(7)]Mo(2)]Mo(1) C(8)]Mo(2)]Mo(1) C(14)]C(13)]Mo(2) C(12)]C(13)]Mo(2) C(13)]C(14)]Mo(2) C(15)]C(14)]Mo(2) C(14)]C(15)]Mo(2) C(17)]C(18)]Mo(1) C(19)]C(18)]Mo(1) C(20)]C(19)]Mo(1) 2.9762(9) 2.046(9) 2.019(9) 2.147(7) 2.150(7) 2.350(8) 2.302(8) 2.279(7) 2.296(8) 2.354(8) 1.824(6) 2.051(9) 83.1(4) 79.0(3) 90.0(3) 85.5(3) 122.5(3) 37.5(3) 94.0(2) 125.3(2) 46.5(2) 46.7(2) 93.2(2) 115.5(2) 150.3(2) 137.8(3) 103.9(2) 95.6(3) 100.0(3) 107.1(3) 113.7(3) 76.5(3) 37.2(3) 62.0(2) 91.1(3) 46.1(2) 46.2(2) 74.0(5) 73.0(5) 71.8(5) 72.4(5) 73.5(5) 73.1(4) 72.7(4) 74.9(5) Mo(2)]C(7) Mo(2)]C(8) Mo(2)]C(11) Mo(2)]C(12) Mo(2)]C(13) Mo(2)]C(14) Mo(2)]C(15) O(1)]N(1) O(2)]C(2) O(3)]C(3) O(4)]C(4) C(7)]C(8) C(11)]Mo(2)]Mo(1) C(13)]Mo(2)]Mo(1) C(12)]No(2)]Mo(1) C(15)]Mo(2)]Mo(1) C(14)]Mo(2)]Mo(1) O(1)]N(1)]Mo(2) O(2)]C(2)]Mo(2) O(3)]C(3)]Mo(1) O(4)]C(4)]Mo(1) C(8)]C(7)]C(6) C(8)]C(7)]Mo(1) C(6)]C(7)]Mo(1) C(8)]C(7)]Mo(2) C(6)]C(7)]Mo(2) Mo(1)]C(7)]Mo(2) C(7)]C(8)]C(9) C(7)]C(8)]Mo(1) C(9)]C(8)]Mo(1) C(7)]C(8)]Mo(2) C(9)]C(8)]Mo(2) Mo(1)]C(8)]Mo(2) C(12)]C(11)]Mo(2) C(15)]C(11)]Mo(2) C(11)]C(12)]Mo(2) C(13)]C(12)]Mo(2) C(11)]C(15)]Mo(2) C(17)]C(16)]Mo(1) C(20)]C(16)]Mo(1) C(16)]C(17)]Mo(1) C(18)]C(17)]Mo(1) C(18)]C(19)]Mo(1) C(19)]C(20)]Mo(1) C(16)]C(20)]Mo(1) 2.160(7) 2.167(7) 2.305(9) 2.323(9) 2.313(8) 2.341(9) 2.328(8) 1.195(8) 1.111(10) 1.122(10) 1.135(9) 1.381(10) 138.7(3) 158.1(3) 167.1(3) 124.9(3) 133.0(3) 162.9(6) 176.3(9) 177.2(8) 176.4(7) 132.4(6) 71.4(4) 137.9(5) 71.7(4) 129.8(5) 87.4(2) 133.2(7) 71.1(4) 129.7(5) 71.4(4) 138.0(5) 87.2(3) 73.3(6) 73.2(5) 71.9(5) 72.2(5) 71.5(5) 70.7(5) 72.7(5) 74.4(5) 71.3(4) 71.4(5) 70.3(5) 72.5(5) of the solutions was elevated and the reaction mixtures refluxed for several hours.The experiments were repeated in the presence of trimethylamine N-oxide to determine whether the presence of Me3NO would labilise the carbonyl ligands.However, in each case, complete decomposition had taken place after several hours and no new products could be isolated after shorter reaction periods. Experimental Except where stated, all reactions were carried out under a nitrogen atmosphere, using standard Schlenk techniques. Solvents were distilled under nitrogen, from appropriate drying agents, and degassed prior to use.The starting materials [(h5- C5H5)2Mo2(CO)4(m-RCCR9)] (R = R9 = CO2Me or Ph; R = H, R9 = Ph; or R = H, R9 = CO2Me) were prepared by literature methods.7 Trimethylamine N-oxide was purchased from Aldrich and sublimed under vacuum prior to use.All other chemicals were purchased from commercial suppliers and used without further purification. Column chromatography was performed on Kieselgel 60 (70–230 or 230–400 mesh). Products are given in order of decreasing Rf values. The instrumentation used to obtain spectroscopic data has been described previously.23 Proton and 13C-{1H} NMR spectra were recorded in CDCl3 at 293 K.Chemical shifts are given in ppm on the d scale relative to SiMe4 (d 0.0); the solvent resonance was used as an internal standard. In all 31P spectra, chemical shifts are given relative to P(OMe)3 with upfield shifts negative. The 13C and 31P spectra were 1H gated decoupled. Preparations [(Á5-C5H5)2(O)2(Ï-O)(Ï-MeO2CC2CO2Me)] 2. To [(h5-C5- H5)2Mo2(CO)4(m-MeO2CC2CO2Me)] 1 (1.0 g, 1.74 mmol), dissolved in 1 : 1 acetonitrile–dichloromethane (150 cm3), was added an excess of freshly sublimed trimethylamine N-oxide (0.39 g, 5.22 mmol, 3 equivalents).The orange-red solution was stirred at room temperature whilst being purged gently with air for 24 h, more solvent being added as necessary. At the end of this time a slight change in the colour of the solution to orange-yellow could be detected.The solvent was removed in vacuo, the residue redissolved in dichloromethane and purified by column chromatography. Elution with dichloromethane gave a red–orange band due to [(h5-C5H5)2Mo2(CO)4- (m-MeO2CC2CO2Me)] 1 (0.21 g, 14%). Further elution with ethyl acetate–acetone (1 : 1) gave an orange-yellow band which, on removal of solvent, gave a powdery orange solid [(h5- C5H5)2Mo2(O)2(m-O)(m-MeO2CC2CO2Me)] 2 (0.41 g, 31%).Analytically pure samples of 2 were prepared by recrystallisation from dichloromethane at 24 8C (Found: C, 37.34; H, 3.10. C16H16Mo2O7 requires C, 37.52; H, 3.15%); nmax(CH2Cl2)/ cm21 1692 (COOMe); (Mo]] O) and (Mo]O]Mo) 894, 799, 645; dH(400 MHz, CDCl3) 3.79 (6 H, s, CO2CH3), 6.14 (10 H, s, C5H5); dC(400 MHz, CDCl3) 52.5 (2 C, s, CO2CH3), 89.7 (10 C, s, C5H5), 212.6 (2 C, s, CO2CH3) [Mo]C2(CO2Me)2 were not observed]; m/z 513 [(M1) 1 H].The reaction was repeated using separate solutions of [(h5- C5H5)2Mo2(CO)4(m-RC2R9)] (R = R9 = Ph, 1.06 g, 1.74 mmol; R = H, R9 = Ph, 0.94 g, 1.74 mmol; R = H, R9 = CO2Me, 0.90 g, 1.74 mmol) in 1 : 1 acetonitrile–dichloromethane (150 cm3).The solutions were treated with an excess of freshly sublimed trimethylamine N-oxide (0.39 g, 5.22 mmol, 3 equivalents) and stirred at room temperature for 24 h whilst being purged gently with air, more solvent being added as required. In each reaction the orangeyellow product was isolated from unreacted starting materials and decomposition products by column chromatography, using ethyl acetate–acetone (1 : 1) as the eluent.The products were obtained as yellow-orange solids, [(h5-C5H5)2Mo2(O)2(m-O)(m-RC2R9)] [R = R9 = Ph 3 (0.104 g, 11%); R = H, R9 = Ph 4 (0.114 g, 14%); R = H, R9 = CO2Me 5 (0.132 g, 17%)].J. Chem. Soc., Dalton Trans., 1997, Pages 2309–2314 2313 Complex 3. nmax(CH2Cl2)/cm21 (Mo]] O) and (Mo]O]Mo) 899, 805, 640; dH(400 MHz, CDCl3) 6.07 (10 H, s, C5H5), 7.28– 7.51 (10 H, m, C6H5); dC(400 MHz, CDCl3) 92.2 (10 C, s, C5H5), 127.0–136.7 (12 C, m, C6H5), [Mo]C2Ph2 were not observed]; m/z 549 [(M1) 1 H].Complex 4. nmax(CH2Cl2)/cm21 (Mo]] O) and (Mo]O]Mo) 894, 810, 636, 628; dH(400 MHz, CDCl3) 6.34 (1 H, s, HCCPh), 6.89 (10 H, s, C5H5), 7.34–7.58 (5 H, m, C6H5); dC(400 MHz, CDCl3) 89.7 (5 C, s, C5H5), 91.3 (5 C, s, C5H5), 125.5–134.2 (6 C, m, C6H5) [Mo]C2(Ph)(H) were not observed]; m/z 473 [(M1) 1 H].Complex 5. nmax(CH2Cl2)/cm21 1690 (COOMe); (Mo]] O) and (Mo]O]Mo) 899, 787, 659, 643; dH(400 MHz, CDCl3) 3.84 (3 H, s, CO2CH3), 6.25 (1 H, s, HC2CO2Me), 6.43 (10 H, s, C5H5); dC(400 MHz, CDCl3) 52.2 (1 C, s, CO2CH3), 88.9 (10 C, s, C5H5), 225.2 (1 C, s, CO2CH3) [Mo]C2(CO2Me)(H) were not observed]; m/z 455 [(M1) 1 H].[(Á5-C5H5)2Mo2(CO)(42x)(MeCN)x(Ï-MeO2CC2CO2Me)] 6. A solution of [(h5-C5H5)2Mo2(CO)4(m-MeO2CC2CO2Me)] 1 (0.5 g, 0.87 mmol) in acetonitrile (100 cm3) was treated with an excess of freshly sublimed trimethylamine N-oxide (0.33 g, 4.35 mmol, 5 equivalents). The solution was stirred at room temperature for 18 h whilst purging gently with argon.The orange-brown product, thought to be [(h5-C5H5)2Mo2(CO)(42x)(MeCN)x- (m-MeO2CC2CO2Me)] could not be isolated from unreacted starting materials and decomposition products without itself decomposing, forming small quantities of [(h5-C5H5)2Mo2- (O)2(m-O)(m-MeO2CC2CO2Me)] 2 as a side product. nmax(CH3CN)/cm21 1922, 1815, 1684 (CO). [(Á5-C5H5)2Mo2(CO)3(ButNC)(Ï-MeO2CC2CO2Me)] 7.A solution of [(h5-C5H5)2Mo2(CO)4(m-MeO2CC2CO2Me)] 1 (1 g, 1.74 mmol) in acetonitrile (150 cm3) was treated with an excess of freshly sublimed trimethylamine N-oxide (0.39 g, 5.22 mmol, 3 equivalents). An excess of ButNC was added (0.46 cm3, 4.35 mmol, 2.5 equivalents) and the orange-red solution then stirred at room temperature for 2 h. The solvent was removed in vacuo, the residue redissolved in dichloromethane and purified by column chromatography.Elution with 1 : 1 hexane–ethyl acetate gave an orange-red band which was evaporated in vacuo to give a crystalline orange solid [(h5-C5H5)2Mo2(CO)3(ButNC)- (m-MeO2CC2CO2Me)] 7 (1.1 g, 96%) (Found: C, 45.86; H, 4.13; N, 2.33. C24H25Mo2NO7 requires C, 45.66; H, 3.99; N, 2.22%); nmax(CH2Cl2)/cm21 1937, 1824, 1684 (CO); 2154 (CN); dH(400 MHz, CDCl3) 1.42 [9 H, s, C(CH3)3], 3.70 (3 H, s, CO2CH3), 3.74 (3 H, s, CO2CH3), 5.18 (5 H, s, C5H5), 5.26 (5 H, s, C5H5); dC(400 MHz, CDCl3) 30.45 [3 C, s, C(CH3)3, 51.7 (1 C, s, CO2CH3), 52.5 (1 C, s, CO2CH3), 87.7 (5 C, s, C5H5), 88.4 (5 C, s, C5H5), 210.2 (2 C, s, CO2CH3) [Mo]C2(CO2Me)2, Mo]CO and Mo]C(CH3)3 were not observed]; m/z 632 [(M1) 1 H], [(M1) 2 nCO] (n = 1–3). [(Á5-C5H5)2Mo2(CO)3{P(OMe)3}(Ï-MeO2CC2CO2Me)] 8.A solution of [(h5-C5H5)2Mo2(CO)4(m-MeO2CC2CO2Me)] 1 (1 g, 1.74 mmol) in acetonitrile (150 cm3) was treated with an excess of freshly sublimed trimethylamine N-oxide (0.39 g, 5.22 mmol, 3 equivalents). An excess of P(OMe)3 was added (0.49 cm3, 4.35 mmol, 2.5 equivalents) and the orange-red solution then stirred at room temperature for 6 h.The solvent was removed in vacuo, the residue redissolved in dichloromethane and purified by column chromatography. Elution with 1 : 1 hexane–ethyl acetate gave an orange-red band which was evaporated in vacuo to give a crystalline orange solid [(h5-C5H5)2Mo2(CO)3{P(OMe)3}- (m-MeO2CC2CO2Me)] 8 (1.1 g, 95%) (Found: C, 39.16; H, 3.59; P, 4.79. C22H25Mo2O10P requires C, 39.31; H, 3.75; P, 4.61%); nmax(CH2Cl2)/cm21 1930, 1826, 1686 (CO); dH(400 MHz, CDCl3) 3.54 (3 H, s, CO2CH3), 3.79 (3 H, s, CO2CH3), 3.92 [9 H, s, P(OCH3)3], 5.12 (5 H, s, C5H5), 5.30 (5 H, s, C5H5); dC(400 MHz, CDCl3) 51.8 (1 C, s, CO2CH3), 52.0 (1 C, s, CO2CH3), 52.9 [3 C, s, P(OCH3)3], 91.1 (5 C, s, C5H5), 93.1 (5 C, s, C5H5) [Mo]C2(CO2Me)2 and Mo]CO were not observed]; dP(250 MHz, CDCl3) 21.2 [P(OMe)3]; m/z 673 [(M1) 1 H], [(M1) 2 nCO] (n = 1–3).[(Á5-C5H5)2Mo2(CO)3(NO)(Ï-RCCR9)][BF4] (R = R9 = CO2- Me 9, Me 10 or H 11; or R = H, R9 = CO2Me 12). Separate solutions of [(h5-C5H5)2Mo2(CO)4(m-RCCR9)] 1 [R = R9 = CO2Me (0.50 g, 0.87 mmol), R = R9 = Me (0.42 g, 0.87 mmol), R = R9 = H (0.40 g, 0.87 mmol) or R = H, R9 = CO2Me (0.45 g, 0.87 mmol)] in dichloromethane (75 cm3) were each treated with an excess of [NO][BF4] (0.15 g, 1.28 mmol, 1.5 equivalents).Each of the solutions was stirred at room temperature for 30 min. After this time, diethyl ether (30 cm3) was added to precipitate the product. The product was collected by filtration and recrystallised from the minimum of 1 : 1 diethyl ether– dichloromethane to afford yellow-brown crystals of [(h5- C5H5)2Mo2(CO)3(NO)(m-RCCR9)][BF4] [R = R9 = CO2Me 9 (0.52 g, 90%), R = R9 = Me 10 (0.44 g, 87%), R = R9 = H 11 (0.43 g, 91%) or R = H, R9 = CO2Me 12 (0.46 g, 87%)].Complex 9 (Found: C, 34.56; H, 2.39; N, 1.89. C19H16BF4Mo2NO8 requires C, 34.32; H, 2.42; N, 2.10%); nmax(CH2Cl2)/cm21 2071, 2032, 1710 (CO); 1636 (NO); dH(400 MHz, CDCl3) 3.86 (3 H, s, CO2CH3), 3.98 (3 H, s, CO2CH3), 5.88 (5 H, s, C5H5), 6.02 (5 H, s, C5H5); dC(400 MHz, CDCl3) 52.57 (1 C, s, CO2CH3), 52.89 (1 C, s, CO2CH3), 92.61 (5 C, s, C5H5), 93.35 (5 C, s, C5H5) [Mo]C2(CO2Me)2 and 3 Mo]CO were not observed]; m/z 579 [(M1) 1 H], [(M1) 2 nCO] (n = 1–3).Complex 10 (Found: C, 35.60; H, 3.00; N, 2.27. C17H16- BF4Mo2NO4 requires C, 35.39; H, 2.79; N, 2.43%); nmax- (CH2Cl2)/cm21 2044, 1995, 1970 (CO); 1697 (NO); dH(400 MHz, CDCl3) 2.21 (3 H, s, H3CC2CH3), 2.33 (3 H, s, H3CC2CH3), 5.94 (5 H, s, C5H5), 6.40 (5 H, s, C5H5); m/z 490 [(M1)], [(M1) 2 nCO] (n = 1–3).Complex 11 (Found: C, 33.00; H, 2.43; N, 2.36. C15H12- BF4Mo2NO4 requires C, 32.82; H, 2.20; N, 2.55%); nmax- (CH2Cl2)/cm21 2058, 2040, 2006 (CO); 1624 (NO); dH(400 MHz, CDCl3) 3.56 (2 H, s, br, HCCH), 3.98 (3 H, s, CO2CH3), 5.88 (5 H, s, C5H5), 6.02 (5 H, s, C5H5); m/z 462 [(M1)], [(M1) 2 nCO] (n = 1–3).Complex 12 (Found: C, 33.80; H, 2.85; N, 2.20. C17H16- BF4Mo2NO4 requires C, 33.64; H, 2.66; N, 2.31%); nmax- (CH2Cl2)/cm21 2066, 2025, 1987, 1723 (CO); 1652 (NO); dH(400 MHz, CDCl3) 3.96 (3 H, s, CO2CH3), 5.88 (5 H, s, C5H5), 5.93 (1 H, s, HC2CO2Me), 6.12 (5 H, s, C5H5); m/z 521 [(M1) 1 H], [(M1) 2 nCO] (n = 1–3).Crystallography Crystal data and data collection parameters for complex 9. C19H16BF4Mo2NO8, Mr = 665.02, monoclinic, space group P21 / n (no. 14), a = 13.382(2), b = 13.655(2), c = 14.770(2) Å, b = 116.074(7), U = 2424.3(5) Å23 (by least-squares refinement on diffractometer angles from 25 centred reflections, 15 < q < 208), Z = 4, Dc = 1.822 Mg m23, m(Mo-Ka) = 1.109 mm21, semiempirical absorption correction based on y scans, relative transmission factors 0.92–1.00, F(000) = 1304, orange– brown block with dimensions 0.40 × 0.40 × 0.20 mm.Data were collected by the w–2q scan method on a Rigaku AFC5R four-circle diffractometer at room temperature using graphitemonochromated Mo-Ka radiation (l = 0.710 73 Å) in the range 7.40 < 2q < 55.18, 1h, 1k, ±l; 3 standard reflections showed no significant variation in intensity; 4466 reflections of which 4263 were independent (Rint = 0.042) and used in all calculations.Structure solution and refinement. The structure was solved using direct methods 24 and subsequent Fourier-difference2314 J. Chem. Soc., Dalton Trans., 1997, Pages 2309–2314 syntheses, and refined by full-matrix least squares on F2 with anisotropic thermal parameters for all non-hydrogen atoms.25 The nitrosyl ligand was identified by comparison of thermal displacement parameters.Hydrogen atoms were placed in geometrically idealised positions and refined using a riding model or as rigid methyl groups. In the final cycles of refinement a weighting scheme of the form w21 = [s2(Fo 2) 1 (0.0371P)2 1 8.56P], where P = (Fo 2 1 2Fc 2)/3, was employed, which produced a flat analysis of variance.Final R(F) = 0.0465, wR(F2) = 0.1053 for 2833 observed reflections [I > 2s(I)], R(F) = 0.1036, wR(F2) = 0.1278 for all data; 311 parameters, 28 restraints (to thermal and positional parameters of BF4 2 anion and disordered OMe group); S = 1.038, maximum D/s = 0.001.Maximum peak and hole in final Fourier-difference map 0.86 and 20.66 e Å23 respectively. 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 CCDC for this material should quote the full literature citation and the reference number 186/516.Acknowledgements We wish to thank the EPSRC for financial support (to J. E. D., G. P. S. and J. C. S.). We are grateful to the Cambridge Crystallographic Data Centre for a CASE award (to G. P. S.). References 1 R. A. Sheldon and J. K. Kochi, Metal-Catalyzed Oxidations of Organic Compounds, Academic Press, New York, 1981. 2 F. Bottomley and L. Sutin, Adv. Organomet. Chem., 1988, 28, 339 and refs. therein. 3 G. Conole, K. A. Hill, M. McPartlin, M. J. Mays and M. J. Morris, J. Chem. Soc., Chem. Commun., 1989, 688; G. Conole, M. McPartlin, M. J. Mays and M. J. Morris, J. Chem. Soc., Dalton Trans., 1990, 2359. 4 W. A. Herrmann, R. Serrano and H. Bock, Angew. Chem., Int. Ed.Engl., 1984, 23, 383; A. H. Klahn-Oliva and D. Sutton, Organometallics, 1984, 3, 1313; N. G. Bokiy, Yu. V. Gatilov, Yu. T. Struchkov and N. A. Ustynyuk, J. Organomet. Chem., 1973, 54, 213. 5 W. Hieber and A. Lipp, Chem. Ber, 1959, 92, 2085; M. O. Albers and N. J. Coville, Coord. Chem. Rev., 1984, 53, 227; T.-Y. Luh, Coord. Chem. Rev., 1984, 60, 255. 6 H. Alper and J. T. Edward, Can.J. Chem., 1970, 48, 1543. 7 W. I. Bailey jun., M. H. Chisholm, F. A. Cotton and L. A. Rankel, J. Am. Chem. Soc., 1978, 100, 5764. 8 J. C. Stichbury, M. J. Mays, P. R. Raithby, M.-A. Rennie and M. R. Fullalove, J. Chem. Soc., Chem. Commun., 1995, 1269. 9 W. I. Bailey jun., F. A. Cotton, J. D. Jamerson and J. R. Kolb, J. Organomet. Chem., 1976, 121, C23; W. I. Bailey jun., D. M. Collins and F. A. Cotton, J. Organomet. Chem., 1977, 135, C53. 10 J. C. Stichbury and M. J. Mays, Inorg. Chim. Acta, in the press. 11 F. A. Cotton and G. Wilkinson, Advanced Inorganic Chemistry, Wiley Interscience, New York, 5th edn., 1988, p. 818. 12 D. J. Blumer, K. W. Barnett and T. L. Brown, J. Organomet. Chem., 1979, 173, 71. 13 G. A. Acum, M. J. Mays, P. R. Raithby and G. A. Solan, J. Organomet. Chem., 1995, 492, 65. 14 W. A. Nugent and B. L. Haymore, Coord. Chem. Rev., 1980, 31, 123. 15 M. L. H. Green and K. J. Moynihan, Polyhedron, 1986, 5, 921; M. L. H. Green, G. Hogarth, P. C. Konidaris and P. Mountford, J. Chem. Soc., Dalton Trans., 1990, 3781. 16 C. Coates, N. G. Connelly and M. C. Crespo, J. Chem. Soc., Dalton Trans., 1988, 2509. 17 G. R. Doel, N. D. Feasey, S. A. R. Knox, A. G. Orpen and J. Webster, J. Chem. Soc., Chem. Commun., 1986, 543. 18 G. A. Acum, M. J. Mays, P. R. Raithby and G. A. Solan, J. Chem. Soc., Dalton Trans., 1995, 3049 and refs. therein. 19 K. G. Caulton, Coord. Chem. Rev., 1975, 14, 317. 20 P. Legzdins, S. J. Rettig and J. E. Veltheer, J. Am. Chem. Soc, 1992, 114, 6922; P. Legzdins, S. J. Rettig, J. E. Veltheer, R. J. Batchelor and F. W. B. Einstein, Organometallics, 1993, 12, 3575; G. R. Clark, D. Hall and K. Marsden, J. Organomet. Chem., 1979, 177, 411 and refs. therein; D. Seddon, W. G. Kita, J. Bray and J. A. McCleverty, Inorg. Synth., 1976, 16, 24. 21 J. T. Lin, Y. C. Lee, H. M. Kao, T.-Y. Dong and Y. S. Wen,, J. Organomet. Chem., 1994, 465, 199. 22 D. M. P. Mingos and D. J. Sherman, Adv. Inorg. Chem., 1989, 34, 293. 23 A. J. M. Caffyn, M. J. Mays and P. R. Raithby, J. Chem. Soc., Dalton Trans., 1991, 2349. 24 SHELXTL-PLUS Version 4.0, Siemens Analytical X-Ray Instruments Inc., Madison, WI, 1990. 25 G. M. Sheldrick, SHELXL 93, University of Göttingen, 1993. Received 3rd February 1997; Paper 7/00761B

ISSN:1477-9226    

DOI:10.1039/a700761b

出版商:RSC    

年代:1997

数据来源: RSC