|
1. |
Molecular structures of non-geminally substituted phosphazenes. Part II. Crystal structure of 2,cis-4,cis-6,cis-8-tetrachloro-2,4,6,8-tetraphenylcyclotetraphosphazatetraene |
|
Dalton Transactions,
Volume 1,
Issue 15,
1972,
Page 1651-1658
G. J. Bullen,
Preview
|
PDF (910KB)
|
|
摘要:
1972 1651Molecular Structures of Non-geminally Substituted Phosphazenes. PartI I. l Crystal Structure of 2,cis-4,cis-6,cis-8-Tet rachloro-2,4,6,8-tetra-phenylcyclotetraphosphazatet raeneBy G. J. Bullen * and P. A. Tucker, Department of Chemistry, University of Essex, Colchester, EssexCrystals of the title compound (or-modification. m.p. 190 " C ) are triclinic, a = 14-07, b = 11.67, c = 9.00 A,o! = 99.7. p = 88-5, y = 102.1 O , space group P i . Z = 2. The structure was solved by direct methods; atomicpositions were determined by least-squares refinement from X-ray diffractometer intensity data, the final R being0.056 for 451 9 reflections. The eight-membered phosphazene ring has an irregular crown conformation withP-N-P angles ranging from 133.1 to 142.0'. Mean bond lengths are : P-N 1 -570, P-CI 2.041, P-C 1.783, andC-C 1.39 A. The bond lengths are discussed in terms of the electronegativity of the exocyclic groups and the ringconformation is discussed in terms of steric factors.I s this series of papers the crystal structures of a numberof non-geminally substituted phosphazenes (I) arebeing studied in order to examine differences in con-figuration and conformation of the phosphazene ring inclosely related or isomeric compounds.In Part I(j-tvarts-N,P,(NHMe),Ph, f (11; R = NHMe) was shownto possess a chair-shaped ring with the approximatesymmetry 2/m and centrosymmetrically arranged sub-stituentsl N,P,C14Ph, (I; X = C1, Y = Ph) is known1 The following abbreviations will be used throughout thispaper : @-trans = 2,cis-4,trans-6,trans-8-, and cis = 2,cis-4,cis-G,cis-8-; the full cis-tvalzs nomenclature is taken from R.A.Shaw, B. W. Fitzsimmons, and B. C . Smith, Chent. Rev., 1962,62, 250.as three geometric isomers which have been assignedcis- or trans-structures from lH n.m.r. studies anddielectric constant measurement^.^ The molecule ofthe p-trans-isomer (11; R = C1) has been shown bycrystal-structure analysis to be similar in shape to thatof p-trans-N,P,(NHMe),Ph, with a ring having the chairconformation. The structure of cis-N,P,Cl,Ph, f (111),Part I, G. J. Bullen and P. R. Mallinson, J.C.S. Dalton,R. A. Shaw and C . Stratton, J . Chem. SOL, 1962, 5004.B. Grushkin, A. J . Berlin, J. M. McClanahan, and R. G.G. J . Bullen, P.R. Mallinson, and A. H. Burr, Chem. Conzm.,1972, 1412.Rice, Inorg. Chem., 1966, 5, 172.1969, 6911652 J.C.S. Daltondescribed here, was examined for comparison with thep-tram-isomer. cis-N4P4C14Ph4 exists in two crystallinemodifications which possess different melting pointsand X-ray powder patterns, and which we shall desig-nate a (m.p. 190 "C) and p (m.p. 226 "C). The conditionsfor the preparation of the a form and its conversion to3 have been rep~rted.~ Once formed, the p-modificationremains unchanged even at room temperature (for atleast a week and possibly longer) but reverts to a whenrecrystallised from solution a t room temperature.The crystal structure described here is that of thea-modification.EXPERIMENTALCrystal Dnta.-C,,H,,Cl,N,P,, Af = 629.7, Triclinic,cz = 14.07 & 0-02, A,a = 99.7 0.3", y = 102.1 f 0.2", U =1423 pi3, D, = 1-45 (by flotation), 2 = 2, D, = 1.47,F(000) = 640.Rlo-K, radiation, h = 0-7107 A; p(Mo-Ka)= 6.6 cn1-I.Suitable crystals of the a-modification were grown frombenzene-light petroleum (b.p. 40-60 "C) solutions a t roomtemperature. The forms commonly developed are thepiiiacoicls (OOl), {OlO), and (1IO). The crystals areroughly equidimensional in cross-section but slightlyelongated along c. The unit-cell dimensions quotedwere used in preference to those of the reduced cell (a =y = 99.4") in order to avoid the large ct angle of the latter.X-Ray intensity data comprising the 13 layers of re-flections IzkO-12 were measured on a Philips PAILREDdiffractometer by use of monochromatised Mo-K, radia-tion.4519 reflections having I > 2a(I) were obtainedby measuring all reflections with sin O/A < 0.70 A-1, froma crystal of dimensions ca. 0.2 x 0.3 x 0.5 min. Theseintensities were corrected for Lorentz and polarisationeffects, but no absorption correction was applied as thelinear absorption coefficient is small.The least-squares refinement was carried out at theAtlas Computer Laboratory, Chilton, using the ' X-ray'63' programme package (J. M. Stewart, University ofMaryland Technical Report, TR 64 6). Atomic scatteringfactors were taken from ref. 6.Structure Determination.-The intensities were adjustedto an absolute scale by Wilson's method7 and normalisedstructure factor amplitudes, /El, were calculated.Statisticsfor these are given in Table 1. -4 comparison of the experi-b = 11-67 & 0.02, c = 9.00 * 0.010-3", p = 88-5Space group Pi (Ct, No. 2).14.07, b = 13.48, c = 9-00 A, = 121.4", p = 91-5",mental with the theoretical values for crystals with randomlydistributed atoms indicates that the space group is centro-symmetric and is therefore Pi. The phases of the re-flections were determined directly by computer applicationof the sign relationship s(EIL) = s(Ew En+,*) (s means' sign of ') to the 342 reflections with [El > 2-0. Thereflections used to specify the origin are given inTABLE 1Normalised structure-factor statisticsCalc.Found ' PIMeail E2 0.991 1.000Mean !El 0.782 0.79833.04.60.3TABLE 2Assigned phasesIt k I-I- 0 3 3-i- - 5 6 24- -6 3 4n -8 -1 1b 0 6 1c 6 -7 4Phase1 P11.0000.89636.91-80.01IEI2-612.262.952.592.2 12.88Table 2together with three other reflections with large I E ~ whosesigns were represented by the symbols a, b, and c.316Phases were determined in terms of these starting signsand symbols. A number of relationships suggested thatb and ac were positive. Since to have all signs positive isnot a likely solution, (E and c negative was chosen and anE-map calculated. This showed clearly the positions ofthe four phosphorus and four chlorine atoms in the asyni-metric unit. The carbon and nitrogen atoms were locatedfrom subsequent Fourier syntheses phased on the heavieratoms.Three cycles of isotropic full-matrix least-squaresrefinement, including refinement of individual layer scalefactors, reduced I? to 0.11.After the second cycle hydrogenatoms were included as a fixed contribution to the structurefactors, their positions being estimated from the moleculargeometry assuming C-H 1-08 A. They were given atemperature factor with B 5.0 Az. In subsequent cyclesof refinement anisotropic temperature factors were intro-duced and a weighting scheme w = (0.727 + 0-0151F01 +0.0001 9 I F,I 2)-1 was used. This weighting scheme gavesimilar average values of wA2 for ranges of increasinglFol and sin O/A. Before proceeding with the refinementall layers were put on the same scale as the layers withlow 1.Limitations on computer store prevented all para-meters from being refined simultaneously. They weretherefore divided into three blocks: one block each forthe parameters of the twelve carbon atoms in two phenylrings and a third block containing parameters associatedwith the chlorine, phosphorus, and nitrogen atoms. Eachblock was refined for two cycles, eight reflections havingw4A >10 (on absolute scale) being excluded from thesecond cycle. One cycle for refinement of individual6 G. J. Bullen and P. A. Tucker, Chem. Comm., 1970, 1186.6 ' International Tables for X-Kay Crystallography,' vol. 3,Kynoch Press, Birmingham, 1962, pp. 202, 203.7 A. J. C. Wilson, Nature, 1942, 150, 1521972layer scale-factors and three cycles of block-diagonal(3 x 3 normal matrices) refinement of the hydrogen atomco-ordinates reduced R to its final value of 0.056 for 4519reflections.RESULTSThe shape of the molecule and the numbering of theThe final atomic co-ordinates atoms are shown in Figure 1.FIGURE 1 Molecular shape and numbering of the atomsam1 thermal parameters with estimated standard deviationsare listed in Tables 3 and 4.Observed and calculatedstructure factors are given in Supplementary PublicationNo. 20405 (23 pp., 1 microfiche).* Bond lengths are listedin Table 5 and bond angles in Table 6. Their estimatedstandard deviations were calculated from the formulae ofJeffrey and Cruickshank,* the maximum of ~ ( x ) , ~ ( y ) , and~ ( z ) being taken as the standard deviation Q ( Y ) of the atomicposition.The orientation and magnitudes of the principalaxes of the vibration ellipsoids for phosphorus, chlorine,and nitrogen atoms are given in Table 7. The anisotropicthermal parameters of the phenyl carbon atoms were usedin an analysis of the vibrations of the atoms in terms of arigid-body motion for each phenyl group. The transla-tional and librational tensors (Table 8 ) were calculated bythe least-squares procedure of Cruickshank with the originof the libration axes at the phosphorus atom to which thephenyl group is bonded (see Figure 2).z ,/ -0y\<IIGURE 2 Axial system used for analysis of the librationalmotion of the phenyl groups. The y axis is parallel to thephenyl ringDISCUSSIONRing Shu$e.-All the chlorine atoms in the moleculelie on the same side of the ring so that the compound isthe cis-isomer as suggested by Grushkin et uL3 Thering shape is different from that of any of the tetramericphosphazenes hitherto examined.The four nitrogenatoms are nearly coplanar (Table 9). The phosphorus* For details see Notice to Authors No. 7 in J. Chem. SOC. (.4),1970, Issue No. 20.1653atoms lie a t different distances from the mean plane of thenitrogen atoms but all on the same side of it. The ringTABLE 3Atomic co-ordinates (as fractions of unit-cell edges)with estimated standard deviations in parentheses%la0.2 11 80( 13)0.463 18( 13)0.27502 ( 14)0*02118(13)0*23124(10)0*41210(10)0-28776( 10)0.11 245(9)0.3397 (4)0.3 7 93 (4)0.1860(4)0*1469(4)0*2124(4)0*1257(7)0.1098 (8)0.1 8 14 (7)0-2668(6)0-2832(5)0.5 1 8 1 (4)0.541 4(5)0.6277 (7)0.685 8 (6)0.661 7( 6)0.57 66 (5)0.3 11 7(4)0*4058(5)0.4229 (6)0.3457 (7)0.25 13 (7)0.2338(6)0.0325(4)- 0*0458(6)- 0.0997(6)- 0.0857(6)- 0.0200(5)0-070( 6)0.045 (6)0- 163( 6)0.3 16(5)0.335(6)0-500(6)0.660( 6)0*740( 6)0-706(6)0.561 (6)0*461(5)0*492(5)0*365(6)0.184(6)0*163( 5 )0*064( 5)-0*062(6)- 0.153(6)-0.125(6)-0*005(6)0-0202(5)Y l b0.1 9757 (1 5)0.1 9983( 18)0*06802( 17)0*33406(13)0*29274( 13)0.10772( 13)0.3621 (5)0.2026(4)0-1408(5)0.3063 (4)0-4572 (6)0.4482 (8)0*5434(11)0-6444( 9)0-6548( 7)0.5596 (6)0-397 1 (6)0*5060( 7)0*5860( 8)0-5568(11)0.4481 (1 3)0*3668(9)0.0440(8)0*0622(7)0-OOSl(8)-0*0601(8)-0*0728(8)-0.0222(7)- 0.03 142 (1 5)0-20001 (12)0*2454(5)0.1922 (6)0.2247(7)0.3089(8)0.361 8(8)0*3306(7)0.364( 7)0*530( 7)0.7 1 O( 7)0.723 (6)0*574( 7)0-521(7)0*682( 7)0*607(7)0*432( 7)0.282 (7)0.1 13( 6)0-0 14( 6)- 0.081 (7)- 0.1 20( 7)- 0*037(6)0*134( 7)0.183(7)0*331(7)0*434( 7)0*372(7)zlc0*43662( 19)0.5381 l(21)0-69122(20)0.68074(21)0*61300( 17)0*72539( 17)0*86601(17)0.81 828( 16)0*6689(6)0-8387 (6)0.8908 (6)0*7288(6)0.5320( 7)0*4575( 12)0*3923( 15)0-4023(13)0-4750(11)0.5423(9)0*7978( 8)0.74 72 ( 10)0.8031 (1 3)0.9027 (1 4)0.9552( 14)0.9022( 10)1-0248 (7)1-0794(8)1.2037(8)1 -26 89 ( 10)1.2189(11)1.0926( 10)0*9625( 6)1*0927(8)1.1 998(9)1 * 1786( 1 1)1-0520( 12)0.9420(8)0*463(9)0.354(9)0-366(9)0.489 (8)0.6 1 3 (9)0.67 3 (9)0.762 (9)0*935(9)1.030(9)0*924(9)1-033 (8)1.249 (8)1 -372(9)1 *2 73 (9)1*059(8)1 * 1 14( 8)1 * 25 7 (9)1 -023(9)0.83 7 (9)1*293(9)The carbon atoms are numbered with two digits, the firstbeing that of the phenyl group to which the atom belongs,e.g. atoms C ( 11)-( 16) belong to phenyl group (1) : hydrogenatoms are numbered according to the carbon atom to whichthey are attached.shape is thus best described as an irregular crown whichis rather flat at P(2) and P(4) (Figure 3), probablyowing to steric repulsion between chlorine atoms.,47, 335.8 G.A. Jeffrey and D. 11'. J. Cruickshank, Qitavl. Rev., 1953,D. CV. J. Cruickshank, Acta Cvyst., 1956, 9, 7541654 J.C.S. DaltonTABLE 4Anisotropic thermal parameters ( x lo4) with estimated standard deviations in parentheses *b2276.8 (1 4)120-1 (19)74.6( 14)98-6( 16)58*7( 11)63*9( 12)55*5( 11)56.1 (1 1)69(4)65(4)81(5)70(4)75(5)123(9)188 ( 14)13 1 (1 0)82(7)67(5)74(5)85(6)108(9)188( 14)1 94 ( 14)1 6 7 (1 0)124(8)160(9)113(8)104(8)59(5)95(7)77(5)94(6)113(8)111(8)124(9)101(7)b3,132*7(23)162.6 (26)140.2(24)148*6(25)102-9(19)114.3(20)113*4(19)99*8(18)146(8)147(8)154(8)129(7)115(8)301(19)3 93 (2 7)298 (20)238(15)195( 12)146(9)202 (1 3)2 66 ( 1 8)267(20)290(21)175( 12)133(9)137(10)174(13)2 23 (1 5)2 32 ( 14)143 (1 0)151(1 I)250( 17)291 (19)168( 11)121(10)99(7)* The temperature factor is in the form : exp (- bl,h2 - b2,k2 - b,,P - 2b1,hk - 2b1,hZ - 2b2,k2).TABLE 5 TABLE 6Bond angles (deg.) with estimated standard deviations Bond lengths (A) with estimated standard deviationsP(l)-Cl(I)P(2)-C1(2)P (3)-C1(3)P ( 4) -C1( 4)MeanP( 1)-C( 11)P( 2)-C(2 1)P( 3)-C(3 1)P(4)-C (4 1)Meanin parentheses *1.569(6) C( 1 I)<( 12)1*570(6) C ( 12)-C( 13)1-579(6) C( 13)-C( 14)1-556(6) C ( 14)-C( 15)1.560 (6) C (1 5)-C( 16)1*569(6) C(16)-C(ll)1 -5 7 7 (5)1.5701 *57 7 (5) c (2 1)-C(22)C1241-C(25) :[;;E:E;2-036(2)2-045(2)2-043(2)2.041 (2)2.0411-785(6)1-776(7)1 - 788 (6)1 - 7 82 (6)1.783c (25 j-c (26 jC (26)-C( 2 1)C( 3 1)<(32)C (32)-C (33) c (3 3)-c (34)C(34)-C(35)C(35)-C(36)C(36)-C(31)C (4 1)-C( 42)C (4 2)-C (4 3)C( 43)-c (44) c (44)-c (45)C(45)-C(46)C (46)-C (4 1)1-38( 3)1-40( 2)1*37(2)1 -36 (2)1-41(1)1-38(1)1-39(1)1-41 (1)1 -36 (2)1*40(2)1 -40(2)1-40(1)1*39(1)b42( 1)1*39(1)1-38( 1)1*42(1)1*39( 1)1.41 (1)1-38( 1)1*40( 1)1*38( 1)b39( 1)1*40(1)N(4)-P( 1)-N( 1)N ( I )-P( 2)-N (2)N( 2)-P( 3)-N (3)N (3)-P (4)-N (4)P(I)-N(l)-P(2)P(2)-N( 2)-P( 3)P( 3)-N( 3)-P( 4)P (4)-N (4)-P( 1)Cl(1)-P(1)-C(l1C1(2)-P(2)-C (2 1c1(3)-P( 3)-C( 3 1C1(4)-P (4)-c (4 1C1( I)-P(I)-N(l)C1( 1)-P( 1)-N (4)C1( 2)-P( 2)-N (2)C1( 2)-P( 2)-N ( 1)C1(3)-P( 3)-N (3)C1(3)-P(3)-N(2)C1(4)-P (4)-N (4)C1(4)-P (4)-N (3)in parentheses1 19*3(3)12 1*5(3)1 2 1 -0 (3)1 2 2.3 (3)137-3 (4)1 3 7- 6 (4)142-0(4)133*1(4)103-5(2)102*4(2)103-3(2)102-0( 2)109.3 (2)108.2 (2)106.8 (2)106-9(2)108-1 (2)107.8( 2)107.8 (2)1 07.4 (2)Mean 1.39C( 1 1)-P( 1)-N( 1)c(ll)-p(l)-N(4)1 0 7 q 3)107.6(3)mean 1-04 A.C(21)-P(2)-N(2) 109.2 (3)* The C-H bond lengths lie in the range 0.88-1-26 A,C (2 1 )-P( 2)-N (1) 1 08*4( 3)symmetrical and highly puckered crown conformation ~ ( 3 1 ) - ~ ( 3 ) - ~ ( 3 ) 107.0(3)is extremely unfavourable owing to the steric repulsions C(31)-P(3)-N(2) 108*2(3)between the four axial exocyclic groups.1° The flatten- ~ ~ i ~ ~ $ ) ~ { ~ ~ # ::;:# ., . I ~,10 N. L. Paddock, Quart. Rev., 1964, 18, 168.P( 1)-c ( 1 1)-c (1 2)P (2)-c (2 1 )-c (22)P (3)-c (3 1)-c (32)P( 1)-C( ll)-C( 16)P( 2)-C( 2 I)<( 26)P( 3)-C (3 1)-C( 36)P( 4)-C(41)<(42)P (4)-C (4 1)-C(46)C(l6)-C(ll)-C(l2)C( 1 1)-C( 12)-C( 13) c ( 12)-C( 13)-c ( 14)C( 13)-C( 14)-C( 15)C( 14)-C( 15)-C( 16)C ( 15)-C ( 16)-C( 1 1)C( 26)-C(2 1)-C( 22) c (2 1 )-c (2 2)-c (2 3) c (22)-C(29)-C( 24)C(23)-C(24)-C(25)C (24)-C (25)-C (26)C (2 5)-C (26)-C (2 1)C( 36)-C(3 1)-C( 32)C(3 1)-C( 32)-C( 33) c (32)-c (33)-c (34)C( 33)-C( 34)-C(35)C( 34)-C( 35)-C( 36)C( 35)-C( 36)-C( 3 1)C( 46)-C (4 l)-C(42) c (41)-c(42)-c( 43) c (42)-c (43)-c (44)C(43)-C(44)-C( 45)C (44)-C (45)-C( 46)C (45)-C(46)-C(41)119-1(6)120-7(5)119.0(6)119-0(6)120*1(5)118*5(5)119-9(4)119*3(5)120-2(7)119*2(11)121.7(11)119*6(9)119*4( 7)122.0(7)1 1 7.9 ( 8)120.7( 10)121-4( 11)119-3( 10)118*7(8)12 1-3 (6)118-9(6)11 9-8( 7)1 2 1 O( 8)1 19- 5 (8)119*3(8)120.8( 6)119*0(7)120*5( 7)120-0( 9)120*8(9)118*9(6)119*9(101972 1655ing a t P(2) and P(4) relieves this by moving Cl(2) andCl(4) outwards so that the C1.Cl distances (3.67-4.05 A, Figures 1 and 4), exceed the van der Waalsdiameter of a chlorine atom. As shown by the torsionangles the symmetry of the ring approximates to wz, themirror plane passing through P(l) and P(3).TABLE 7Amplitudes of thermal vibration (A) along the principal axes of the vibration ellipsoids.Each principal axis is specifiedby its direction cosines I, m, and n referred to orthogonal axes a', b', and c, where b' lies in the bc plane0-2680.2410.209Cl(2) 0.3270.2360.1900.2840.2510.1990-3040-2520.1850.2140.1940.1740.2230.1980.164'C1(1)Cl(3)Cl(4)P(1)P(2)E0-809- 0.4990-312-0.1020.803- 0.5880.7530-064- 0.655- 0.48 1- 0.4260-7660.0090.9160.4010.1150.0620.992 .TYZ- 0.507- 0.3220.8000.7230.4650.51 1- 0.399- 0.3770.8360.8230.0800.5620.702-0.2910.650-0.474 -0.8800~000n0.2990.8050.5130.3740.6270.5240.6550.545- 0.3010.9010.3130.7120.2760.8730.470- 0.683- 0.646-0.130TABLE 8Tij (Hi2) and oiJ (deg.2) for the phenyl rings, withestimated standard deviations in parenthesesRing (1) Ring (2) Ring (3) Ring (4)TI, 0.070(4) 0*041(2) 0-065(3) 0-053(8)T,, 0.041(7) 0-059(4) 0-042(6) 0.043(5)T3, 0.034(10) 0-036(6) 0*039(8) 0.030(7)T12 - 0.009(4) -0.001(2) O.OOO(3) 0.003(3)TI, - 0*003(4) 0.009(2) 0*002(3) 0*009(3)Tz, - 0*003(7) 0.006(4) 0.011 (6) - 0.007(5)a11 75(19) 49( 11) 54( 15) 41(14)0 2 2 14(2) 6(1) 5(2) 6(2)0 3 3 8P) 21(1) 9(1) 11(1)w12 16(2) -W) - 4 m 4(2)- 2(3) w13 2(4) - 2(2) lO(3)c'123 - 3(2) 1(1) 2(2) -4(UTABLE 9Parameters of mean planes through sets of atoms and insquare brackets distances (A) of atoms from the planes.The equation of a plane is Zx + my + wz = p withco-ordinates and distances (A) referred to the ortho-gonal axes a', b', and c, where b' lies in the bc planeI TPL n PPlane (1) : N(1)-(4) 0.119 0.690 0.714 6.644"(1) -0.022, N(2) 0.022, N(3) -0.022, N(4) 0.022, P(l)-0.551, P(2) -0.122, P(3) -0.411, P(4) -0.0931Plane (2): C(l1)-(16) -0.395 0.469 0.790 4.169[C(11) -0.002, C(12) -0.002, C(13) 0.004, C(14) -0.003,C(15) -0.002, C(16) 0.004, P(l) -0.0221Plane (3) : C(21)-(26) -0.511 0.526 0.680 2.452[C(21) -0.007, C(22) 0.003, C(23) 0.004, C(24) -0.007,C(25) 0.003, C(26) 0.004, P(2) -0.0691Plane (4) : C(31)-(36) -0.118 0.866 0,486 3.647[C(31) -0.014, C(32) 0.015, C(33) 0.002, C(34) -0.021,C(35) 0.022, C(36) - 0.005, P(3) - 0.0381Plane ( 5 ) : C(41)-(46) 0.721 0.586 0.369 4-950[C(41) -0.006, C(42) 0.004, C(43) 0.004, C(44) -0.009,C(45) 0.007, C(46) 0.001, P(4) -0.09110.2120.1980.1670.2100.1770.1630.246N(l) Q.2030.196N(2) 0.2490.2000.1930.267N(3) 0.2080.1700.2450.2100.177P(3)P(4)N (4)E0.0790.761- 0.644- 0.107- 0.3200.942-0.116- 0.3240.939-0.181-0.4010.8980,2930.6160.7310-2690.8760.4011120.2670.6070.7490.7050.6430-2980.4650.8180.3400.3690.8190.4400.61 10.4670.6270.617- 0.639- 0.475920.961-0.231-0.1550.701- 0.696- 0.1570.878- 0.476- 0.0560.912- 0.41 1 o*ooo0.735- 0.6340-2400.7310.085- 0.677In the p-trans-isomer of N4P4Cl,Ph4 the molecule iscentrosymmetric and the ring has a chair conf~rmation.~There have been attempts to correlate differences ofring shape with electronic factors lo but here we have-44 P(4)CL Po d3"""3) "21n$1( b lFIGURE 3 (a) Shape of the phosphazene ring; (b) torsionangles (deg.) in the ringdifferent ring shapes in two isomers whose only otherdifference is one of the cis-trans type.The difference inshape therefore seems more likely to arise from stericfactors. In both isomers the more bulky phenylgroups occupy the equatorial positions where they pointoutwards and away from each other. If the conforma-tions were other than those actually found it would notbe possible to.place all the phenyl groups in equatorialpositions. We therefore suggest that the two conforma-tions are uniquely suited to the respective geometricisomers1656 J.C.S. DaltonBond Lengths and AizgZes.-The P-N bond lengthsare equal within experimental error and fall within therange of lengths found for tetrameric phosphazenes.Ansell and Bullen have correlated the bond length withI-- b sin as A O C L P 0 N,C '-FIGURE 4 Projection of the structure down the c axis, showingintramolecular C1 * - C1 and intermolecular distances (A).The phenyl group numbers are given in parenthesesthe mean electronegativity of the exocyclic ligandatoms.ll Wagner has improved on this by use of theorbital electronegativities of groups of atoms l2 butregards N4P4F, as a special case which does not fit thecorrelation.In Figure 5 we have added to the sevenpoints plotted by Wagner [in Figure l(b) of ref.121 afurther five points for the non-geminal derivativesstudied in our laboratory and the geminal phosphazenesN4P,F,Me2 and N4P4F4Me4. We suggest, on the basisof this larger body of data, that the correlation caninclude N4P4F, but, because of the limited accuracyof the bond lengths, it is difficult to define by a curve.We prefer instead just to indicate the general trend by abroken line.Bond lengths and angles for the cis- and p-traits-isomers of N4P4C14Ph4 are compared in Table 10. The11 G. B. Ansell and G. J. Bullen, J . Chem. SOC. ( A ) , 1971, 2498.12 A. J. Wagner, J . Inorg. Nuclear Chern., 1971, 33, 3988.1-3 G. J. Bullen, J . Chem. Soc. ( A ) , 1971, 1460.1' R. Hazekamp, T. Migchelsen, and A. Vos, Actu Cryst., 1962,16 A. J.Wagner and A. Vos, Acta Cryst., 1968, B24, 707.16 A. W. Schlueter and R. A. Jacobson, J. Chem. SOC. ( A ) ,15, 639.1968, 2317.data for the p-traits-isomer are taken from a preliminaryreport and their limits of error are rather large. Never-theless, mean bond lengths and angles in the twoisomers appear to be equal despite the difference inconformation. The mean P-Cl length in cis-N,P,Cl,Ph,(2-041 -hi) is significantly longer than in the compounds[NPCl,],, (n = 3,13 4,l4~l5 and 5 16) and is, indeed, longerthan the Schomaker-Stevenson estimate for a P-Clsingle bond (2-01 A). The mean P-C bond length1 I I I I J2.5 3 0 3.5 4.0Electronegativi t y (Pauling scale)FIGURE 5 I'ariation of the mean cyclic P-N bond length (i) intetrameric phosphazenes with the mean orbital electro-negativity of the exocyclic groups.Possible error is repre-sented as f 2 . 5 ~ . Exocyclic groups in N,P,X, or N,P,X,Y,:X, Me, (Af. W. Dougill, J . Chenz. Soc., 1961, 5471) ; B, (NMe,),(G. J. Bullen, J . Chem. Sor., 1962, 3193); C, Br, (H. Zoer andA. J. Wagner, Acla C~yst., 1972, B28, 252); D, (OMe), (ref.11); E, C1, ( K form) (ref. 14); I;, C1, (T form) (ref. 15);G, F, (ref. 21); H, (NHMe),Ph, (ref. 1 ) ; I, Cl,(NMe,), (G. J.Bullen and P. A. Tucker, unpublished work) ; J, Cl,Ph, (thiswork); I(, F,Me, [W. C. Marsh and J. Trotter, J . Cheiia.SOC. ( A ) , 1971, 5693; L, N,P,F,?tIe, (ref. 20)(1.783 -hi) is on the other hand shorter than in p-trans-N,P4(NHMe),Ph4 (1.808 A) 1 or the trimeric derivativesN3P,Ph, (1.804 gem-N3P3C1,Ph4 (1,792 A) ,laand gein-N,P3C1,Ph2 (1.788 A) l9 in which pairs ofTABLE 10Comparison of bond lengths (A) and angles (deg.) inthe cis- and p-trans-isomers of N,P,Cl,Ph,c i s aP-N 1.556-1.579 P-c 1 * 7 76-1.788P-CI 2-036-2.045N-P-N 119.3-122.3P-N-P 133.1- 1 42.0C1-P-C 102.0- 103.50 This work.b Ref.p-trans b1.77, 1-802.03, 2.04119.5, 121.3132.4, 138.6102.2, 105.11 56-1 * 604.phenyl groups are attached to phosphorus. Theseobservations may be rationalised by consideration of thephosphorus environment shown in (IV). If the groupR1 is capable of electron donation (e.g. R1 = NMe, or Ph)and the group R2 of electron withdrawal (e.g. R2 = C1)resonance structures of the type (V) will be important1969, B25, 316.1967, 21, 375.1966, 19, 693.17 F. R.Ahmed, P. Singh, and W. H. Barnes, Acta Cryst.,18 F. R. Ahmed, M 7 . H. Barnes, and N. V. Mani, Actu Cryst.,19 F. R. Ahmed, W. H. Barnes, and N. V. Mani, Acta Cryst.1972in the overall bonding scheme, leading to shortenedP-R1 and lengthened P-R2 bonds. The increasinglengths of the P-Cl and P-C bonds in the series hT3P3c16,genz-N3P,C1,Ph2, genz-N3P3C1,Ph,, N3P3Ph6 l7 can beaccounted for similarly, the effects being transmittedthrough the nitrogen atoms in the phosphazene ring.With cis-N,P4C14Ph, the effects are more pronouncedbecause the electron-donating phenyl and -withdrawingchloro-groups are both attached to the same phosphorusatom. Consequently the P-C1 bonds are longer and theP-C bonds shorter than in these other compounds.The bond angles are in general similar to those inother phosphazenes.The main interest attaches tothe P-N-P angles, whose individual values differ fromeach other by amounts far greater than their experi-mental error. Several tetrameric phosphazenes containP-N-P angles in the range 131-134". Larger angles,as here, appear characteristic of molecules containingchlorine or fluorine : T-N4P,C18 138",15 p-trans-N,P,Cl,-and are clearlv connected with the approach to planarityin the pliosphazene ring.Oric.1-itatioiz of Phe123d Groz@s.-At phosphorus atomsP(2), P(3), and P(4) the phenyl groups are orientateds!-mmetrically, i.e. with their planes perpendicular tothe adjoining C1-P-C plane (Figure 6a).Each chlorinePh, 139",, N,P4I;,?lle2 143.3, 146.7°,20 N4P4F, 147*2°,21C(161(c ISewinan projections down carbon-phosphorus bonds :(a) cis-S,P,Cl,Ph, a t P(2), (b) P-tuaizs-N,P,(NHMe),Ph,, and(c) cis-S4P,C1,Ph, at P( 1)FIGURE 6atom is then almost equidistant from the two sides ofthe neighbouring benzene ring, e.g. Cl(2) - - C(22)3.70, and Cl(2) - C(26) 3-73 A. These distances areonly slightly greater than that expected for a van derWaals separation of chlorine and carbon and it thereforeseems reasonable to suggest that the symmetricalplacing of the phenyl groups results from equalisationzo W. C. Marsh and J. Trotter, J . Chew SOC. ( A ) , 1971, 573.*l H. McD. McGeachin and F. R. Tromans, J . Chem. SOC.,1961, 4777.1657of chlorine-carbon repulsions, particularly as thissymmetrical orient ation is also found in p-trans-N,P4C1,-Ph,22 but not in p-trans-N,P,(NHMe),Ph,.In thelatter all the phenyl groups are orientated asymmetric-ally so that they are eclipsed with one of the nitrogenatoms of the ring (Figure 6b).l This difference betweenthe chloro- and methylamino-derivatives could be attri-buted to the difference in spatial requirements betweena chlorine atom and the more asymmetric methyl-amino-group.The symmetrical placing of phenyl groups in cis-N,P,Cl,Ph, also means that the x system of the groupmakes the maximum overlap with the phosphorus3 d , ~ orbital whose axis lies in the C1-P-C plane (Craigand Paddock co-ordinate system 23). Although suitablecombinations of 3d orbitals for overlap can be foundwhatever the orientation of the phenyl x system, theuse of 3dZ2 for the exocyclic groups is perhaps particularlyadvantageous because this is the only 3d orbital ofphosphorus not already involved in x bonding in thephosphazene ring.% Since, however, the symmetricorientation is also sterically favoured one cannot saywhether x-bonding requirements exert any significantinfluence on the orientation.The occurrence of asym-metric orientations in p-trans-N,P,(NHMe),Ph, sug-gests that they do not, though one must remember thathere the presence of an exocyclic nitrogen atom willlessen x interaction between the phenyl group andthe phosphorus. A decision between the steric andelectronic arguments could more likely be made froma knowledge of the molecular structure of a non-geminalN4P4F4Ph4, since here the electronegative fluorine atomsare also small.At P(l) the phenyl group is not orientated symmetric-ally, the C1-P-C-C torsion angle being 57" (Figure 6c)and the chlorine-carbon contacts unequal :Cl(1) - - * C(12) 3.37, and Cl(1) * C(16) 4.08 A.In thiscase it seems that intermolecular repulsions affect thesituation (see later).Intermolecular Distances.-In t erat omic contacts likelyto determine the molecular packing are marked on theprojection of the crystal structure shown in Figure 4.The C - * * C and Cl - - * C distances are unexceptional.The C1. . . C1 distance across a centre of symmetry(3.38 A), though short, is similar to the shortest inter-molecular contacts in solid chlorine (3.34 & 0.04Phenyl rings (1) and (2) are very approximately coplanar(see Table 9), so that near the middle of the unit cellvan der Waals contact between neighbouring moleculesis between pairs of rings [ring (1) in one molecule andring (2) in the other] which are almost parallel.Itseems likely that the twist of phenyl ring (1) aboutthe P-C bond mentioned earlier is forced by the packingtogether of these rings. It may be that in the high-melting polymorph a rearrangement of the moleculesallows a more symmetrical phenyl group orientation a t22 A. H. Burr, personal communication.23 D. P. Craig and N. L. Paddock, J . Chem. SOC., 1962, 4118.24 R. L. Collin. Ada Crvst.. 1952. 6.4311658 J.C.S. DaltonP(l) and possibly also a more symmetrical phosphazenering (with symmetry closer to mm2).Molecular 0scilZations.The thermal vibrations ofthe chlorine atoms show considerable anisotropy withthe minor axes of the thermal ellipsoids lying approxi-mately along the P-Cl bonds (Table 7). An attemptto rationalise the phosphorus and chlorine atom vibra-tions in terms of a rigid-body oscillation was not success-ful. On the other hand such an analysis of the in-dividual phenyl groups did give sensible results althoughsome of the tensor components have large estimatedstandard deviations (Table 8). The standard deviationswere found to be least when the origin of the librationalaxes was put at or near the phosphorus atom but somemay be large because of the difficulty of guaranteeingthat the librational axes intersect at the phosphorus atom.As very few of the off-diagonal elements of the tensorsare significantly different from zero, the translationsand librations can be described satisfactorily in termsof the natural axes of the groups shown in Figure 2.For the phenyl groups (2), (3), and (4) the order ofmagnitude all > w33 ? w22 may be rationalised in termsof the effect of atomic displacements on the bonding.Rotation about axis (1) reduces only the (weak) xoverlap between carbon and phosphorus, while rotationabout axis (3) reduces CJ overlap and rotation aboutaxis (2) reduces 'both.The larger all for phenyl group(1) may reflect a weaker carbon-phosphorus x overlapconnected with the twist of the phenyl ring about theP-C bond.The calculation of the o tensors for the phenyl ringsallows correction of the bond lengths in the P-Phgroups for librational effects. However as the correc-tions to the atomic co-ordinates prove to be comparableto or less than the estimated standard deviations of theco-ordinates, they have been disregarded.In conclusion we suggest that the cyclic and exocyclicbond lengths in tetrameric pliosphazenes are consider-ably influenced by the extent of electron withdrawalfrom or donation to phosphorus by the exocyclic groups,that in certain instances the orientation of one exo-cyclic group is determined largely by the nature of thesecond group attached to the same phosphorus atom,and that with bulky exocyclic groups the conformationof the phosphazene ring is determined primarily byintramolecular steric factors but in a crystal inter-molecular forces niay sometimes be the over-ridingfactor.We thank Professor R. A. Shaw and Dr. C. Stratton forsupplying a sample of the compound (111), the Universityof Essex Computing Centre and the Atlas ComputerLaboratory, Chilton, for the use of their facilities, theS.R.C. for the award of a research studentship (to P. A. T.),and N. Lewis for the preparation of diagrams.[2/484 Received, 1st Madh, 1972
ISSN:1477-9226
DOI:10.1039/DT9720001651
出版商:RSC
年代:1972
数据来源: RSC
|
|