摘要:

1980 1317Pentagonal-bipyramidal Molybdenum- and Tungsten-( 11) Derivatives :Crystal and Molecular Structure of 1-3-~-AllyldicarbonyIchlorobis-(trimethyl phosphite)molybdenum(ii)By Brian J. Brisdon, Dennis A. Edwards, and Kathleen E. Paddick, School of Chemistry, University ofBath, Bath BA2 7AYMichael G. B. Drew, Department of Chemistry, The University, Whiteknights, Reading RG6 2ADCompounds of the type [MX(CO),(q3-C3H4R)L,] [M = Mo or W; X = halide; R = H, L = P(OMe), or P(OEt),;R = Me, L = P(OMe),] have been prepared. A crystal-structure determination has been-carried out on the titlecompound as a representative member of the series. Crystals are triclinic, space group Pi, with a = 13.309(11),b = 10.093(10), c = 8.148(7) A, a = 114.9(8), (3 = 93.1 (7), y = 100.6(7)', andZ = 2.A total of 1534 indepen-dent reflections above background were collected on a diffractometer and refined to R 0.072. The geometryaround the metal atom is best considered in terms of a distorted pentagonal bipyramid with a chlorine atom [Mo-CI2.606(6) A] and a carbonyl group [Mo-C 1.91 (3) A] in axial positions, and the equatorial girdle occupied by twophosphite ligands [Mo-P 2.432(6), 2.425(6) A], one carbonyl group [Mo-C 2.03(3) A], and the ally1 ligand whichoccupies two adjacent sites. The molecule has approximate C, symmetry with the mirror plane containing themeral atom, both carbonyl groups, and the central carbon atom of the ally1 group. The Mo-C(allyl) bond lengths are2.403(20), 2.347(24), and 2.407(22) A. Carbon-1 3 and 'H n.m.r.data indicate that this geometry is retained insolution below 230 K, but at room temperature the molecule is fluxional. The two cationic complexes [ hAO(CO),(qs-C,H,R)(P(OMe),},] [BF,] (R = H or Me) have also been characterised.MANY allyl complexes of the early transition elements areknown to be fluxional at room temperature, but in manyinstances the mechanisms of their rearrangement pro-cesses have not been estab1ished.ly2 A combination ofsingle-crystal X-ray diffraction, to reveal preferred solid-state configurations, and dynamic n.m.r. studies isdesirable when studying stereochemical non-rigidity.Using these methods the dynamic behaviour of [MoCl-(CO)2~q-C3H5)(Pl~,PCH,CH2PPh,)] 3 and [Mo(CO,)-(q-C,H,)(pd)(py)] (pd = pentane-2,4-dionate, py =pyridine) has been explained in terms of a trigonal twistmechanism, which involves rotation of the triangular faceformed by the halogen or pyridine nitrogen atom, and thedonor atoms of the bidentate ligand, relative to the faceformed by the allyl and two carbonyl groups.We wereinterested to ascertain whether the chelating ligandspresent in both these molecules had a major influence onthe structure and rearrangement of this type of complex,and whether a similar non-bond-breaking mechanismwould also apply to dynamic [M(CO),(q-C,H,)L,] (M =Mo or W) complexes containing only unidentate ligands.In order to spectroscopically monitor the maximumnumber of atoms directly bonded to the metal, we choseto investigate complexes containing unidentate yhos-phorus-donor ligands.As tertiary phospliines areknown to readily break the metal-ally1 linkage in thesespecies and form allylphosphonium ~ a l t s , ~ ? ~ the lessnucleophilic tertiary phosphites were used to preparecomplexes of the types [MX(CO),( q3-C3H4R) { P(OR'),},](M = Mo, R = H or Me, R' = Me or Et ; M = W, R = H,R' = Me or Et) and [MO(CO),(~~-C~H~R)(P(OM~)~},]-[BF,] (R = H or Me). Both series of complexes werefound to be stereochemfcally non-rigid at, or near, roomtemperature and were therefore suitable for such aninvestigation.EXPERIMENTALAll preparations were performed in an atmosphere of dryN, gas using solvents and liquid reagents freed from moistureand oxvyen by standard procedures. The phosphitesP(OR'), (13' = Me or Et) were left to stand over lead-sodium alloy for 1 week and then fractionally distilled fromfresh alloy and stored over molecular sieves.In addition,triethyl phosphite was fractionally distilled a t least twiceimmediately before use. The starting materials [MX(CO),-(q3-C3H4R)(NCMe),] (M = Mo or W; X = C1, Br, or I ;R = H or Me) were prepared from [M(CO),(NCMe),] byliterature methods.'Solution and mull i.r. spectra in the region 200-4 000 cm-'were recorded on a Perkin-Elmer 597 spectrophotometer.Far-i.r. spectra from 40 to 400 cm-l were obtained using aBeckmann RIIC FS 720 interferometer on samples pressedin Polythene discs. Hydrogen-1 and 31P n.m.r. wererecorded on a JEOL PS 100 instrument, and JEOL PTF-100and FX-90 spectrometers were used to obtain I3C n.m.r.spectra.Synthesis O~[MX(CO)~(~~-C~H,R)L~][M = Mo OY W; R =H, L == P(Ohle), OY P(OEt),; R = Me, L = P(OMe),].---A solution of [hlX(CO),(q3-C3H4R)(NCMe),] (1 .O mmol)dissolved in acetone (10 cm3) was treated with P(OR'), (2.0mmol) a t rooin temperature.After 1 h the solution wasreduced to low bulk in vucuo. The precipitated product wasfiltered off, washed with a little cold acetone (-20 "C), andvacuum dried. The P(OEt), complexes crystallised withdifficulty and an alternative procedure was adopted toobtain pure products. After mixing, the solvent wascompletely removed and the residue evacuated for 2 h.Approximately 2 cm3 of acetone were added at -20 "C aridthe mixture stored at -30 "C overnight.The product wasfiltered off and washed as above.Attempts to prepare triphenyl phosphite analogues bythis procedure were unsuccessful.[Mo( CO) ,( q3-C3H4R){ P( OMe) ,},I [ BJ?,] .-The salt Ag [ BF,](1.0 mmol) was added to a solution of [MoC1(CO),(q3-C3H4R){ P(OMe),},] (1 mmol) dissolved in acetone (20 cm3).The mixture was stirred for 10 min, filtered, and the filtrateadded to a solution of P(OMe), (ca. 1.5 mmol) in acetone( 5 cm3) at room temperature. After 0.5 h, the solution wasevaporated a t reduced pressure to a volume of 6 cm3 andtreated dropwise with Et,O to initiate precipitation. Theproduct was filtered off and recrystallised from CHC1,-Et,O1318 J.C.S. DaltonThe conductivity (A) of a mol dm-3 solution of (9;Crystal Data.-Cl,H2,C1Mo0,P,, M = 476.5, Triclinic,I< = H).in nitromethane was 92.4 S cm2 mol-1.u = 13.309(11), b = 10.093(10), c = 8.148(7) A, 01 =ll4.9(8), (3 = 93.1(7), )/ = 100.6(7)", U = 965.1 A 3, D,(flotation) = 1.64 g cme3, Z = 2, D, = 1.64 g ~ m - ~ , F(000)= 484, A(Mo-K,) = 0.710 7 A, p(Mo-K,) = 10.0 cm--l, spacegroup PI from the successful structure determination.A crystal of approximate size 0.4 x 0.3 x 0.2 mm wasmounted with a* parallel to the instrument axis of a G.E.XRD 5 diffractometer which was used to measure diffractionintensities and unit-cell dimensions. The apparatus wasequipped with a manual goniostat scintillation counter andpulse-height analyser. The stationary-crystal-stationary-counter method was used with a 4" take-off angle and acounting time of 10 s.Individual backgrounds were takenfor those reflections whose counts were seriously affected bythe streaking of other orders. For other reflections, back-grounds were taken from plots of background as a functionof 28. Several standard reflections monitored during thecourse of the experiment showed no significant changes inintensities. 2 135 Independent reflections were measuredfor 28 < 40", of which 1 534 with I > 3o(I) were used insubsequent refinement. No absorption or extinction cor -rections were applied.Structure Determination.-The positions of the molyb-denum atoms were obtained from the Patterson map andthe positions of remaining atoms were found from Fourieranalysis. The hydrogen-atom positions of the allyl ormethyl groups could not of course be calculated ; some wereobserved in a difference Fourier but refinement was notsuccessful, and therefore they were not included in the finalTABLE 1deviations in parenthesesAtomic co-ordinates ( x lo4) with estimated standardAtom X Y Z7 735( 1) 7 758(2) 5 076(2)6 608( 5) 7 308(8) 7 391(8)7 504(4) 10 212(6) 5 506(8)7 583(4) 5 l65(6) 2 912(7)7 174(12) 3 821(16) 3 461(21)6 845(11) 4 488( 17) 981(20)8 795(19) 3 308(28) 1026(33)Mo(l)P(1)P(2)O(21)O(22)O(23)C(21)8 708( 11) 4 835( 17) 2 391(20)final positions are given in Table 1, and bond lengths andangles in Table 2.Structure-factor tables and anisotropicthermal parameters are given in Supplementary PublicationNo.SUP 22774 (11 pp.).*RESULTS AND DISCUSSIONNeutral Com$exes .-React ion of tertiary p hosp hi t eswith the molybdenum- or tungsten-nitrile starting mater-TABLE 2Molecular dimensions, distances (A) and angles (")Mo(l)-Cl( 1) 2.606(0) P(2)-0(22) 1.613(15)Mo( 1)-P( 1) 2.432(6) P(2)-O(23) 1.609( 15)Mo(l)-P(2) 2.425(6) O( 21)-C( 21) 1.604(28)Mo( 1)-C( 1) 2.403( 20) 0 (2 2)-C ( 2 2) 1.503 ( 29)Mo( 1)-C(2) 2.347(24) O( 23)-C( 23) 1.479( 28)Mo( 1)-c(3) 2.407(22) O( 11)-C( 11) 1.521(30)Mo(l)-C(4) 1.914(25) 0(12)-C(12) 1.475(28)Mo( l)-c(5) 2.035(27) 0(13)-C(13) 1.435(31)1.403 (36)c(1)-c(2) 1.384(37)P(l)-O(ll) 1.589(16)c(2)-c(3) 1.192(26)P( 1)-O( 12) 1.603( 16)1.099( 25)P ( 1 )-0 ( 13) c (4) -0 (4)P( 2)-O( 2 1) 1.632( 16) C(5)-0(5)1.6 1 8 ( 1 6)Cl(l)-Mo(l)-P(l) 100.79(21) C(4)-Mo(l)-C(5) 101.9(8)C1( l)-Mo( 1)-P(2) 98.71 (21) Mo( 1)-P( 1)-O( 11) 111.1(5)P( l)-Mo(l)-P(2) 146.12(20) Mo(l)-P( 1)-O( 12) 119.8(6)C1( l)-Mo( 1)-C( 1) 80.0(6) Mo( 1)-P( 1)-0( 13) 117.7(6)P(1)-MO(1)-C(1) 75.0(6) O(ll)-P(1)-0(12) 102.2(8)P( ~)-Mo( l)-C( 1) 135.9( 6) O( 1 1)-P( 1)-O( 13) 106.0( 8)Cl(l)--M0(l)-C(2) 94.6(6) 0(12)-P(1)-0(13) 97.9(8)P( l)-Mo( l)-C( 2) 102.6( 7) Mo( 1)-P( 2)-0 (2 1) 1 1 1.5( 5)P( ~)-Mo( 1)-C (2) 103.2 (7) Mo( 1)-P( 2)-0 (22) 1 20.7( 6)C(l)-Mo(l)-C(2) 34.3(8) Mo(l)-P(2)-0(23) 119.3(6)C1( l)-Mo( 1)-C( 3) 80.5(6) O( 21)-P( 2)-O( 22) 99.7( 8)P( l)-Mo(l)-C(3) 135.4(6) 0(21)-P(2)-0(23) 105.1(8)P(2)-Mo(l)-C(3) 75.1(6) 0(22)-P(2)-0(23) 97.7(8)C( l)-Mo( l)-C( 3) 61.2( 8) P( 2)-O( 21)-C( 2 1) 120.7( 13)C(2)-Mo( 1)-C(3) 33.8(8) P(2)-0(22)-C(22) 119.8(13)Cl(l)-Mo(l)-C(4) 178.3(6) P(2)-0(23)-C(23) 123.2(14)P(l)-Mo(l)-C(4) 80.9(5) I'(1)-O(l1)-C(l1) 119.8(13)C(l)-Mo(l)-C(4) 100.9(8) P(1)-0(13)-C(13) 122.4(16)C(2)-Mo(l)-C(4) 85.6(9) Mo(l)-C(l)-C(2) 70.7(13)C(3)-Mo(l)-C(4) 98.7(9) Mo(l)-C(2)-C(l) 75.0(13)Cl(l)-Mo(l)-C(5) 78.0(6) Mo(l)-C(2)-C(3) 75.4(14)P(l)-Mo( 1)-C(5) 78.8(6) C(l)-C(2)-C(3) 122.8(26)P(2)-Mo(l)-C(5) 78.5(6) Mo(l)-C(3)-C(2) 70.7( 13)C( l)-Mo( 1)-C(5) 141.8(9) Mo( l)-C(4)-0(4) 178.7( 18)C( ~)-Mo( 1 )-C( 5) 1 72.5( 9) 1 75.4( 20)C(3)-Mo(l)-C(5) 142.7(9)P( ~)-Mo( l)-C( 4) 79.6( 5) P( 1)-0( 1 2) -C( 1 2) 1 22.2( 14)MO ( 1)-C( 5)-O( 5 )C(22) 6 063(20) 3 470i32) 3 723(42) ials at room temperature yielded a series of substitutedC(23) 6 909( 2 1) 5 214(31)O(11) 8 578(11) 11 294(16)O( 12) 7 040(13) 11 218(17) 7 264(20) There was no evidence of nucleophilic attack of the6 729(12) 10 284(16) 961(20) phosphites on the allyl ligand even in refluxing MeCN, 8 623(21) 12 925(29) 6 107(37)O(13)5 986(18) 10 733(28) 7 592(36) Me,CO, or MeOH in the presence of excess of phosphite, C(11)6 915(22) 9 804(31) 2 101(36) in marked contrast to the analogous reactions of tertiaryC(12)9 8 816(18) 312(20) 9 8 348(34) 499( 27) :::[:!\ phosphines.6~6 However, the spectral properties of theC(13)C( 1)8 898(19) 6 866(32) 6 501(35) whole series of complexes were inconsistent with the C(2)8 562(18) 8 027(21) 3 342( 30) pseudo-octahedral geometry which has been establishedC(3)2 283(25)727(28) by single-crystal X-ray diffraction studies for 15 other 9 095(14) 8 195(20)C(4)6 278(22) 7 219(24)O(4)5 467(16) 6 924(20) 3 098(26) complexes containing the M(CO),(?-C3H5) entity; 3p4~10-13 C(5)O(5)consequently a crystal-structure determination wasAll atoms were refined anisotropically t o R carried out on [MoC1(CO) ( -C H ) p OMe),},], (11, as arepresentative member of the series.A preliminary Calculations were doneat the University of London Computer report of this structure has appeared re~ent1y.l~ The weighting scheme used was w) = 1 for F, <Scattering factors and SoZid-State Strzcctzcre.-The structure consists Of-264(25)5 682(21) complexes [MX(C0)2(?3-C3H,R){P(OR')3),I (Table 3) -refinement.0.072 by full-matrix least squares.using SHELX 76Centre.40 and w& = 4 0 / ~ , for F, ,dispersion corrections were taken from ref.9. The final discrete units of [MOCl(Co)2(?-C3H,){P(OMe),),] as2 1 3 5{ 1319shown in Figure 1, which also gives the atomic number-ing scheme. The structure has C, symmetry withinexperimental error, with the mirror plane containing themetal atom, both carbonyl groups, the chlorine atom,and the central carbon of the allyl group. The moststriking feature of the structure is the P(l)-Mo-P(2)angle of 146.1(2)", which is not an angle observed inprevious [Mo(CO),(-~~-C~HJL,]~~ structures where theleast-squares plane through the pentagonal girdle showsa maximum deviation of 0.57 A for an atom from the six-atom plane [plane (1) in Table 41.However there isa good explanation for this distortion from the idealgeometry.Our starting point is plane (2) in Table 4 which showsthat Mo, P(1), P(2), C(1), and C(3) are almost co lanar(maximum deviation 0.17 A) and that C(5) is 1.14 1 fromTABLE 3Analytical data (calculated values in parentheses), melting points, yields, and selected i.r. dataAnalysis (%) Carbonyl i.r. ------ M.?. Yield ,-A%-Compound C H (OO/ C) (%)27.5 (27.2) 4.8 (4.9) 76-78 58 1985, 186025.4 (25.5) 4.5 (4.5) 85-88 51 1986, 1855( 3 ) [MoI (co) 2 ( T-caHti) {P(OMc) 3121 23.2 (23.3) 4.0 (4.1) 77-79 88 1 985, 1 860(4) [MoCl(CO) z (713-C4H7) {P(OMe)3)2l 29.2 (29.4) 6.1 (6.1) 95-97 38 1992, 1 84735.1 (36.4) 6.1 (6.3) 38-41 46 1985, 185623.2 (23.4) 4.1 (4.1) 95-97 57 1974, 184831.5 (31.5) 5.2 (6.4) 40-52 61 1967, 1843(9) [Mo(CO)~(~-C~H,){P(OMC)~},][BF~] 25.7 (25.8) 5.0 (6.0) 104 72 2 010, 1023(1) [~~oC~(Co),(?-~,H,){P(0Me)3}21( 2 ) [MoBr(CO) a(rl-CaH6) {P(OMe) 31 a1( 5 ) [MoC1(CO)2(~-C,I~,){P(O~t)3~,1( 7) lwcl (CO) (7)-C3H6) P(OMe) d(8) [WCl (CO) a(rl-C,H,) {P(OEt) 31 8131.4 (31.3) 5.4 (5.4) 46-48 50 1983, 1859 (6) [M~I(CO)~(T-C~HS){P(OE~)~}ZI(decomp.)(10) [Mo(CO),(q3-C4H,){P(OMe),},][BF4] 26.1 (27.1) 5.0 (5.2) 79-82 71 2 000, 1 917* In CHCl,.data *others1960m1 963m1960 (sh)1 060 (sh)1944 (sh)1942 (sh)M(CO)2 v(M-X)angle (") /cm-l109 337108 151113 120109 230232110 108111 239112 227114112allyl and carbonyl groups occupy three fac sites in apseudo-octahedron, and the L-Mo-L angles vary between71 and 92°.3*4~10-13 Indeed the geometry observed for (1)is best considered in terms of a pentagonal bipyramid(p.b.) with Cl(1) and C(4) in axial positions [C1(1)-Mo-C(4) 178.3(6)"], and the allyl ligand occupying twoadjacent sites in the equatorial girdle.* The root meansquare (r.m.s.) deviation 115 of the so-ordination spherefrom the ideal polyhedron is 0.16 A, a large value in-dicating severe distortion from the p.b. However, thedeviation from other seven-co-ordinate polyhedra isconsiderably larger.The distortions of (1) from an ideal p.b.are indicatedby the angles subtended by the Mo-Cl(1) bond withMo-C(S), Mo-C(1), and Mo-C(3) which are 78.0(6),80.0(6), and 80.5(6)0, and with Mo-P(l) and Mo-P(2)which are 100.8(2) and 98.7(2)".Calculation of the* In these q3-allyl structures, all three carbon atoms are in-volved in some degree of bonding with the metal. While the allylgroup can be considered as formally bidentate, from a purelystructural point of view it can be described as bi- or uni-dentate.Our choice is made on the grounds of whethcr the resulting geo-metry can best be explained in terms of a six- or sevcn-co-ordin-ate polyhedron. In the case of the M(CO),(q3-allyl) structuresusually described as fuc-octahedral, there is little to choosebetween this description and that of a capped trigonal prism withthe bidentate allyl occupying the unique edge.ls In these struc-tures the plane of the allyl group is approximately parallel to aplane containing the metal and four donor atoms and thus thecarbon atoms do not distort the octahedron to any appreciableextent.A plane through Mo, C(1), C(3) is close to just one otherdonor atom, that trans to the allyl group. (Thus for example inref. 4, the trans atom is 0.08 and the other four donor atomsover 1.45 A from this plane.) By contrast in (l), as shown inTable 4, both phosphorus atoms are close to the plane of Mo, C( l),and C(3) while C(5) is also within 1 A. Thus the allyl group has amajor effect on the geometry of the complex and the only reason-able description of the geometry is as a seven-co-ordinate penta-gonal bipyramid with the allyl group bidentate, Furthermore,the allyl ligand, with a small normalised bite of 1.02, is wellsuited to occupy the equatorial girdle of such a structure.16this plane (which intersects the allyl plane at an angle of78.7").The position of C(5) relativc to plane (2) isdetermined by the positions of the OMe groups. Studyof the torsion angles shows that the six C(1)-Mo-P(1)-O(1m) and C(3)-Mo-P(2)-0(2m) angles (m = 1-3) are(322)FIGURE 1 Structure of [MoC~(CO)~(~-C,H,){P(OM~),),]all within 6" of 60, -60, or 180". Thus the P-0 bondsare staggered with respect to the allyl group, a conform-ation which maximises OMe contacts with the allylhydrogen atoms. Consequentially they are also stag-gered with respect to the Mo-C(4) bond, but not to Mo-Cl(1) and Mo-C(5). The smallest torsion angles involv1320 J.C.S.Daltoning these latter two bonds are C1( l)-Mo-P(n)-O(n2) 15.3,-15.9" for n = 1,2 and C(5)-Mo-P(n)-O(n3) -28.1,29.1" for n = 1,2. I t is interesting and significant thatthe Mo-P(n1)-O(n1) angles are considerably less (by anaverage of 8.1") than the Mo-P(n2)-0(n2) and Mo-P(n3)-O(n3) angles. Clearly this is because these O(n1) atomsare not involved in close contacts with either Cl(1) orA consequence of the staggering of the phosphite OMegroups with respect to the allyl group is that O(23) andTABLE 4Least-squares planes with deviations (A) of atoms from theplanes. Atoms not contributing to the planes aremarked with an asteriskPlane (1): Mo -0.06, P(l) -0.42, P(2) -0.39, C(l) 0.17, C(3)Plane (2): Mo -0.17, P(l) 0.08, P(2) 0.05, C(l) -0.01, C(3) 0.05,C(4).0.12, C(5) 0.57, C(2) * -0.47Cl(1) * -2.67, C(2) * 0.68, C(4) * 1.67, C(5) * -1.14, O(23) *0.09, O(13) * 0.09Plane (3) : Mo, C(1), C(3) 0.00, P ( l ) * 0.36, P(2) * 0.22, C(5) -0.80,C(2) * 0.64Angles between the allyl group C(1), C(2), C(3) and planes (1)and (2) are 68.0 and 78.7" respectively.O(13) are very close to the girdle plane (Table 4).IfC(5) were in the girdle plane (as in an ideal pentagonalbipyramid) it would be involved in impossibly shortcontacts with O(13) and 0(23), so atom C(5) moves out ofthe plane by over 1 A to give C(5)-Mo-P(n)-O(nS) torsionangles of -28.1,29.1" and C(5) 0(13), O(23) distances-. -.FIGURE 2 Disposition of the P(OMe), groups. A projection ofthe molecule onto the approximate mirror plane through atomsMo, C1(1), C(4), 0(4), C(5), 0 ( 5 ) , and C(2).In this projectionC(l), C(3) are coincident as are the two P(OMe), groups. Thedisposition of the OMe groups maximises 0 * allyl contactsof 2.96, 2.99 A respectively. This movement of C(5)causes the axial atoms Cl(1) and C(4) to move out of theperpendicular with respect to the girdle plane (Figure 2).Even so the Cl(1)-Mo-P(n) angles are 100.8(2), 98.7(2)"and together with the long Mo-Cl(1) bond [2.606(6) Acompared to a mean of 2.52 A in ten molybdenum(I1)seven-co-ordinate structures] l5 ensure that the smalltorsion angles C1( l)-Mo-P(n)-O(n2) do not lead to shortC1 0 distances. Other distances in the complex areas expected except that the Mo-C(1),C(3) bonds a t2.403(20), 2.407(22)A are ca.0.1 A longer than the meanmolybdenum-ally1 distances in the fac-octahedral struc-tures (see Table 3 in ref. 4). This is presumably a con-sequence of the crowding in the present molecule. TheMo-C(carbony1) bond lengths are not significantlydifferent particularly as the Mo O(4) and Mo O(5)distances are comparable.We have shown therefore that the bulkiness of theP(OMe), groups is a cause of the distortions from theideal seven-co-ordinate pentagonal-bipyramidal geo-metry, and it seems likely that steric effects are alsoimportant in determining this particular polytopal form,since severe crowding would occur in the usual pseudo-octahedral structure found for all other [Mo(CO),-(q-C,H,)L,]n* complexes.However, such a geometry isby no means impossible and is adopted by [MoCl(CO),-(q-C,H,) (Ph,PCH,CH,PPh,)] in which the cone angle ofthe phosphorus ligand (128") is considerably larger thanthat of P(OMe), (107").l6 Consequently we believe thatelectronic considerations are also important in establish-ing the polytopal form. With only one exception, allprevious structure determinations on this class of com-plex have been carried out on molecules containing bi-or tri-dentate ligands, which prevent the formation of astructure analogous to (1). The one exception is[Mo(CO),(q-C3H5) (NCMe),]+ which adopts a pseudo-octahedral structure lo with the usual lac arrangement ofcarbonyl and allyl ligands, which we previously sug-gested to be electronically favoured in other complexes,all of which lack additional strong x-acceptor ligands.However, for the phosphite complexes, not only are theresteric problems associated with a pseudo-octahedralstructure, but a t least one phosphite ligand would betrans to a carbonyl group and hence competing as ax acceptor.In view of the small energy difference be-tween t he various idealized seven-co-ordinate geo-metries,15 the stereochemistry of these molybdenum-and tungsten-(11) derivatives may well be particularlysensitive to ligand properties.Solution Behuviour.-In Nu j ol mulls, compounds(1)-(8) each exhibited the expected two A' carbonylmodes at approximately 1980 and 1850 cm-1. Thesolution spectra of several of the complexes dissolved inCCI,, CHCl,, or CH,Cl, however revealed additional bandsbetween 1940 and 1960 cm-l (Table 1) which did notincrease in intensity with time.Evaporation of thesolutions regenerated the original two-band solid-statespectra indicating that a second isomeric species isformed in solution and is responsible for these spectralchanges. This feature complicated the calculation ofMo(CO), bond angles from the relative intensities of thetwo major carbonyl bands,17 but in those complexeswhere such calculations could be performed, values of108-113" were obtained (Table 3). These values are inreasonable agreement with the solid-state Mo(CO), angl1321of 102" in (l), and considered together with the n.m.r.data indicate that the whole series of neutral complexeshave the same basic structure.The n.m.r.spectra of (1) below 230 K were in completeaccord with the solid-state structure, with the exceptionof additional weak absorptions probably attributable tothe isomeric species mentioned earlier. Thus for themajor isomer 13C n.m.r. measurements at 170 K showedequivalence of the ends of the allyl group and of themethyl groups of the phosphite ligands (Table 5 ) . Thetwo carbonyl groups are inequivalent, with both carbonsstrongly coupled with the two phosphorus atoms. Inother phosphine-substituted carbonyl complexes ofsimplified and approximated to overlapping AX spectrain which each syn-proton was strongly coupled to onlyone phosphorus ( J 7.0 Hz), while at room temperatureboth syn protons were coupled equally with two phos-phorus atoms (J 3.5 Hz) resulting in an A2X2 pattern forthe lH-decoupled spectrum. Preparation of the 2-methylallyl derivative (4) in which homonuclear couplingeffects were absent resulted in simpler spectra withsimilar P-H coupling constants, but a higher temperaturedependence.Thus complex (4) was rigid on the lHn.m.r. time scale at rooin temperature and the limitinghigh-temperature A2X2 spectrum was not reached untilca. 335 K, under which conditions decompositionTABLE 513C41H) n.m.r. data13C [6/p.p.m. relative to SiMe,, J(C-P)/Hz]r 7 hAlly1(O,/"C) Carbonyl- 90 233.0 (t, 35.9)225.0 (t, 27.2)0 228.4 (br)32 228.0 (t, 29.3)- 30 231.8 (t, 35.9)225.4 (t, 26.7)- S O 225.7 (dt, 33.3, 19.8)216.4 (dt, 58.0, 21.4)32molybdenum and tungsten, tram 31P-13C coupling hasbeen reported to be larger than cis coupling,l* and con-sequently the more suongly coupled triplet centred at233.0 p.p.m. is tentatively assigned to the axial carbonylgroup in view of the slightly larger P(n)-Mo-C(4) anglescompared with P(n)-Mo-C(5) (Table 2).By 260 K thecarbonyl signals had coalesced into a single broad re-sonance at 228.4 p.p.m., which sharpened into a triplet byroom temperature showing the carbonyls to be equiva-lent, and coupled equally with both phosphorus atoms.The free energy of activation of this rearrangement pro-cess, calculated by Kessler's procedure,19 was found to be55 5 kJ mol-1. The terminal allyl carbons, whichshowed a downfield shift of 10-15 p.p.m.compared withother Group 6 complexes of a similar t ~ p e , ~ , ~ remainedequivalent throughout. In addition, for this one room-temperature spectrum, couplings between lH and 13C inthe co-ordinated allyl and phosphite ligands were inea-sured. For the former ligand, values of lJ(C2-H) =161.1 and average lJ(C1.3-H) = 161.6 Hz are similar tothe limited data reported on other complexes containingthe allyl group 20921 and are consistent with sp2 hybridis-ation of the allylic carbon atoms. For the phosphiteligand lJ(C-H) = 147.5 Hz which differs little from thefree-ligand value (144.5 Hz).The lH n.m.r. spectrum of (1) was typical of sym-metrical q3-allyl complexes, but significant P-HA,,(allyl) coupling a t all temperatures resulted in two over-lapping AA'XX' patterns rather than the doublet (ordoublet of triplets) usually observed for the syn-ally1proton signal.At low temperatures this spin system,C2 c1ca Me Phosphi te114.4 (s) 76.0 (s) 53.6 (d, 4.6)113.6 ( s ) 75.1 ( s ) 53.6 (br)112.9 ( s ) 74.7 ( s ) 53.1 (d, 4.9)130.0 (s) 75.4 ( s ) 25.1 (s) 53.2 (d, 4.6)103.1 (s) 62.1 ( s )101.3 (s) 61.2 (s)53.6 (d, 9.2)54.5 (d, 9.2)54.3 (br)occurred. The remaining Complexes in Table G showedsimilar effects in their lH n.m.r. spectra, with coalescencetemperatures (as determined from the syn-proton signals)in the same range (235-260 K) as ( l ) , and considerablylower than the more sterically hindered 2-methylallylderivative (4).The triethyl phosphite complexes [ ( 5 ) , (6), and (S)]showed additional tempera t ure-dependen t phenomenain their lH 1i.m.r.spectra. At room temperature theinethylene proton signals of the phosphite ligandsappeared as a pseudo-quintet, since 3J(P-H) z 3J(H-H)caused overlapping of the expected quartet of doublets,whereas at low temperatures rotation about the P-0-Cbonds was prevented resulting in two sets of overlappingquintets of equal intensity. The molecular symmetry ofthe neutral complex is such that both P atoms are chemi-cally equivalent, and in solution at low temperature the3lP n.m.r. spectrum of (1) consisted of a single broadband centred at 181 p.p.m. downfield from 85% H3P0,.At room temperature no significant changes occurred inthe spectrum, although we were unable to observe finestructure on the JEOL PS 100 instrument at either highor low temperatures.These n.m.r.studies show that all the neutral com-plexes undergo a fast rearrangement in solution at roomtemperature or above, which results in equivalence of thecarbonyl groups, equivalence of the phosphite ligands,and coupling of the allyl syn-protons with two equivalentphosphorus atoms with coupling constants equal to theaverage of those at the low-temperature limit. Theseobservations are inconsistent with a simple allyl-rotatio1322mechanism such as that found for [Mo(CO),(q-C,H,)-(T-C,H,)],~~ since this could not result in the carbonylgroups becoming equivalent. Addition of free P(OMe),had no effect on the lH n.m.r.spectrum of (l), and there-fore a rearrangement involving dissociation of thephosphite ligand seems unlikely. Since no informationcould be obtained on any involvement of the axialhalogen atoms in the rearrangement process, and as thespectra of the neutral complexes were complicated bysignals from isomeric species, we cannot at present definethe rearrangement mechanism further.CatioTbic Derivatives .-Two [ Mo (CO) ,( q3-C,H4R) -(P(ONe),),][BF4] (R = H or Me) complexes (9) and (10)31P(6/p.p.m.relatike toHaPo,)181179177187176173150168J.C.S. Daltona broad singlet. The lH n.m.r. spectra of (7) below 240K contained two doublets of intensity ratios 1 : 2 centredat 6 3.66 and 3.78 p.p.m. respectively which wereassigned to the P(OMe), ligands.Two 31P decouplingfrequencies were found. One decoupled the phosphitesignal at 6 3.66 p.p.m. [the unique P(OMe),] and also theoverlapping doublet of doublets centred at 6 2.96 p.p.m.arising from the anti-ally1 protons [3J(P-H) = 12 Hz],the other decoupled the phosphite signal at 6 3.78 p.p.m.[P(OMe), trans to allyl] and slightly sharpened the signalfrom the anti-ally1 protons, suggesting weak couplingwith these protons also. As the temperature was raisedfrom -30 "C the two pliosyhite signals coalesced andTABLE 6Hydrogen-1 and 31P-(1H} n.m.r. data at 303 K'H [6/p.p.m. relative to SiMe,, J(H-H) and J(P-H)/Hz]H,,ai H,vn3.24 (d, 13.3)3.49 (d, 13.2)3.17 (s) 4.35 (d, 7.0)3.23 (d, 13.5)3.77 (d, 13.5) *2.87 (d, 13.5)2.81 (br d, 12) *3.13 (br, ca.I(;) *4.52 (d, 7.9, 3.5)4.42 (dt, 7.8, 3.5) * 4.21 (dt, 7.4, 3.5)4.40 (dt, 7.5, 3.3)2.88 (d, 13.5) 4.33 (dt, 8.0, 3.7)4.22 (dt, 7.8, 3.6)* Hidden by P(OR'), signal.H or Me4.92 (m)4.74 (m)4.64 (m)1.81 (s)4.85 (m)4.60 (m)4.90 (m)2.03 (br)4.95 ( I l l )4.08 ( I l l )Phosphitc3.85 (d, 11.5)3.80 (d, 11.4)3.79 (d, 11.0)3.82 (d, 11.8)1.33 (t, 7.0). 4.16 (q, 7.0)1.30 (t, 7.0), 4.10 (4, 7.0)1.32 (t, 7.2), 4.00 (q, 7.3)3.71 (d, 11.0)3.86 (br d, ca. 8)3.76 (a, 11.8)were readily prepared from the neutral halogeno-com-plexcs by an analogous method to that described byPowell 23 for the preparation of related cationic complexes[ Mo( CO),( q3-C,H4R) (bipy)L] [BF,] (bipy = 2,2'-bipy-ridyl).Infrared v(C0) band-intensity measurementsindicated M(CO), angles similar to those of the neutralphosphite complexes and far too large for the usualpseudo-octahedral structure found for other cations ofthe type [Mo(CO),(q-C,H,)L,] +. Furthermore, low-temperature lH [and 13C on complex (9)] n.m.r. dataclearly show two chemical environments for the phosphiteligands in the ratio 2 : 1 , and a symmetrically bondedallyl group. Consequently it seems reasonable to pre-sume that at low temperatures the cationic species alsoadopt a basically pentagonal-bipyramidal structurederived from (1) by replacement of the halogen by tri-methyl phosphite, although different distortions fromthose of (1) are anticipated. Unequivocal confirmationcould not be obtained froin the 13C n.m.r.spectrum of(9). A clearly defined doublet of triplets, with 31Pcoupling constants in the ratio ca. 3 : 1, as expected fora combination of trans- and cis-phosphorus atoms,l8could be assigned to the axial carbonyl groups in (9), butthe position and fine structure of the other carbonylsignal is more tentative (Table 5) because of interferencefrom spurious resonances, presumably due to decompo-sition products. The room-temperature spectrum waseven less informative as no carbonyl signals could bedetected, and the two phosphite resonances collapsed tosharpened to a doublet cciitred at 6 3.71 L3J(P-H) = 11Hz], but did not exchange with added P(OMe), indicatinga non-dissociative phosphite rearrangement.The anti-allyl proton signal was also temperature dependent andbecame a broad doublet by 300 K. Unfortunately theresonances froin the syn-ally1 protons completelyobscured by the phosphite signals at all temperatures.The lH n.m.r. of (10) showed similar features, but as withthe neutral 2-metliylallyl derivatives the fully dynamicstate was reached at a relatively high temperature(>350 K) , when decomposition became significant.We thank A. W. Johans for his assistance with thecrystallographic investigations and 13. Wood for the extensivelH n.1n.r. measurements. We are also indebted to Dr. M.Murray (School of Chemistry, University of Bristol) andMr. 1'. Beynon [JEOT, (U.K.) Ltd.] for n.ni.r. xneasure-ments, and to the S.1I.C. for the award of a studcntship (to[!)/1765 Received, 2nd Nouember, 19701K. E. P.).REFERENCESK. Vrieze in ' Dynamic Nuclear Magnetic ResonanceSpectroscopy,' eds. L. M. Jackinan and 1;. A. Cotton, AcademicPress, New York, 1975, p. 441.2 E. G. Hoffman, R. Kallwcit, G. Schroth, K. Seevogel, W.Stempfle, and G. Wilke, J . Organometallic Chem., 1975, 97, 183.3 J . W. Faller, D. A. Haitko, R. D. Adams, and D. F. Chodosh,J. Amer. Chem. Soc., 1979, 101, 865.B. J. Brisdon and A. A. Woolf, J.C.S. Dalton, 1978, 291.ti H. Friedel, I. W. Renk, and H. tom Dieck, J . OrganometallicChem., 1971, 26, 2476 B. J. Brisdon and K. E. Paddick, J . Organometallic Chem.,7 R. G. Hayter. J. Organometallic Chem., 1968, 13, C1.8 G. M. Sheldrick, personal communication.9 International Tables for X-Ray Crystallography,’ Kynoch10 M. G. B. Drew, B. J. Brisdon, and M. Cartwright, lnorg.11 M. G . B. Drew and G. F. Griffin, Acta Cryst., 1979, Ba,12 K. R. Breakell, S. J. Rettig, D. L. Singbeil, A. Storr, and J.A. J. Graham, D. Alerigg, and B. Sheldrick, Cryst. Struct.l4 M. G. B. Drew, B. J. Brisdon, D. A. Edwards, and K. E.1978, 149, 113.Press, Birmingham, 1975, vol. 4.Chim. Acta, 1979, 86, 127.3036.Trotter, Canad. J . Chem., 1978, 56, 2099.Comm., 1977, 6, 253.Paddick, Inorg. Chim. A d a , 1979, 85, L381.1323lS M. G. B. Drew, Progr. Inorg. Chem., 1977, 23, 67.16 C. A. Tolman, Chem. Rev., 1977, 77, 313.17 L. E. Orgel, Inorg. Chem., 1962. 1, 75.I* P. S. Braterman, D. W. Milne, E. W. Randall, and E.l9 H. Kessler, Angew. Chem. Intemat. Edn., 1970, 9, 219.20 B. E. Maim, R. Pietropaolo, and B. L. Shaw, J . C . S . Dalton,21 M. H. Chisholm and S. Godleski, Progr. Inorg. Chem., 1976,22 J. W. Faller, Chin-Chun Chen, M. J. Mattina, and A.z3 P. Powell, J. Organometallic Chem., 1977, 129, 175.Rosenberg, J.C.S. Dalton, 1973, 1027.1973, 2391.$30, 299.Jakubowski, J . Organometallic Chenz., 1973, 52, 361

ISSN:1477-9226    

DOI:10.1039/DT9800001317

出版商:RSC    

年代:1980

数据来源: RSC