摘要:
1972 1069Crystal Structure of Compounds with (N-P),, Rings. Part X.l Deca-bromocyclopentaphosphazene, N 5P5Br10By Janneke G. Hartsuiker and A. J. Wagner," University of Groningen, Laboratory of Structural Chemistry,Groningen, The NetherlandsThe title compound, crystallizes in space-group P i with unit-cell dimensions a = 13.760(2), b = 9.352(1), c =10.329(1) A, u = 93.78(1), p = 122.57(1), y = 107.27(1)",Z = 2. The structure was solved from diffractometerdata (1 10 K) by the symbolic addition method and refined by block-diagonal least-squares analysis to R 0.1 2for 3260 observed reflexions. The molecule has a pseudo-mirror plane passing through a PBr, group and theopposite nitrogen atom. The ten-membered ring is puckered with a phosphorus atom deviating 0.54 b, from theaverage ring plane and with one re-entrant angle a t a nitrogen atom.The mean P-N and P-Br bond lengthsare 1.571 (1 7) and 2.175(6) A. The N-P-N angles fall into two different groups [mean values 120.5(9) and11 4.3(9)"] and there are three kinds of P-N-P angles [mean values 143.8(10), 136.6(11), and 131 *2(1 l)"].There are some short intermolecular Br . . * Br distances (3.45, 3.45, and 3.48 A).WE now report the crystal-structure analysis of deca-bromocyclopentaphosphazene N5P5BrI0, in order tocompare the molecular dimensions and the conformationof the ring with those of other medium-sized cyclo-pho~phazenes,~-~ in particular, the corresponding chloro-compound. The crystal and molecular structuresof the lower bromocyclophosphazenes, N,P,Br, andN,P,Br,,l have been described and the present workprovides an extension of the structural knowledge inthis series.EXPERIMENTALThe crystals used in the X-ray work were obtained byevaporating a solution of the compound in light petroleum(b.p.60-80 "C). Only few crystals were obtained, ofvarying shape and usually exposing a great many crystalplanes.Crystal Data.-N,P,Br,,, M = 1024-2, Triclinic, a =13.760(2), b = 9*352(1), c = 10*329(1) A, cc = 93.78(1),p = 122*57(1), y = 107.27(1)". U = 1027 A3, D, = 3-06(from ref. 6), 2 = 2, D, = 3.31 g. ~ m - ~ , F(000) = 920.Space-group Pi. A(Mo-K,) = 0.7107 A, ~(Mo-K,) =210.8 cm-l.Measurements.-The unit-cell constants were determinedfrom zero-level Weissenberg photographs around the a,b, and c axes, taken with Cu-K, radiation a t 110 K.Theunit-cell constants were calculated by a least-squaresprocedure and are approximately the same as those givenin ref. 6. D, was not measured, but D, for 2 = 2 agreesapproximately with D , from ref. 6. No systematicallyabsent reflexions were observed; on the basis of the dis-Part IX, H. Zoer and R. J. Wagner, Acta Cryst., 1972, B28,252.A. J. Wagner and -4. Vos, Acta Cryst., 1968, B24, 1423.N. L. Paddock, J. Trotter, and S. H. Whitlow, J . Chena.A. W. Schlueter and R. A. Jacobson, J . Chenz. SOC. ( A ) ,SOC. ( A ) , 1968, 2227.1968, 2317.tribution of the normalized structure-factors (see later)the space-group was taken to be Pi, rather than Pl.The intensities were measured on a Xonius automaticthree-circle single-crystal diffractometer with zirconiumfiltered molybdenum radiation at 110 K.The crystalwith dimensions ca. 0.20 x 0.17 x 0.12 mm was mountedwith c* parallel to the $-axis of the goniometer. The8-28 scanning method was used, with a scanning anglevarying from 1.7-2-3" in 28, a constant scanning speedof 2.4" rnin-l in 28, and stationary background measure-ments before and after the scan. 6077 independentreflexions in the range 0 < (sin8)/A < 0.70 A-1 weremeasured. One of three standard reflexions was monitoredevery 50 reflexions. The net intensities were calculatedfrom the measured peak and background counts and scaledto the same relative scale by means of the standard re-flexions. Reflexions for which the net intensity wasnegative, or for which it was <30% of the difference inbackground counts, were considered unobserved.For theremaining 3335 reflexions, the Lorentz and polarizationfactors were applied and the weights based on the counting-rate uncertainties were calculated. A correction forabsorption was computed.' For this correction thecrystal boundaries were described by 13 planes.Structure Determinution.-The structure was solved bythe symbolic addition method.* An average isotropictemperature factor was calculated by the method ofWilson9 and used in the calculation of the normalizedstructure-factors IEl . Quantities characterizing the dis-tribution of the IEI values are given in Table 1. Compari-son with the theoretical values of these quantities showsPI to be the more probable space-group.5 H.Zoer and6 G. E. Coxon, D. B. Sowerby, and G. C. Tranter, J . Chcm.W. R. Busing and H. A. Levy, Acta Cryst., 1957, 10, 180.J. Karle and I. L. Karle, Acta Cryst., 1966, 21, 849.A. J . C. Wilson, Natvre, 1942, 150, 151.J. Wagner, Acta Cryst., 1970, B26, 1812.SOC., 1965, 56971070 J.C.S. DaltonFor the 249 reflexions having \El > 2, all triplets ofreflexions to be used in the C2 relation, s(E=+~e) 21s(EH) - s(E=t), were sought and their triple products cal-culated. Table 2 lists the reflexions which were chosenTABLE 1Distribution of the iiormalized structure-factorsObserved CalculatedCcntrosymm. Yon-centrosymm.E12) 0.994 1.000 1.000 El;- 1> 0.940 0.968 0.736 0.886E > 1 (%) 32.8 32.0 36.8E > 2 (yo) 4.2 5-0 1.80.806 0.798E > 3 ( % ) 0.2 0.3 0.01TAIZLE 2Starting set for phase cleterniinatioiiSign__ lieflexion IEI 5, 1, 11 3.9154, 2, 6 3.828 -.5 , 4, :I 3.676 __14, 4, (i 3.335 d.L 2, 6 3.307 b5, 1, 9 3- 194 C6, !), 1 2.523 d,-t o determine the origin, together with the reflexions towhich a symbolic sign was given. Starting with this set,it was possible to give a real or symbolic sign to all 249reflexions, under the condition that a sign was acceptedonly, when the probability of the corresponding relationwas a t least 98%. Using the sign indications for a, b, c,and d having the highest probability ( b = +, d = +,a = c) an E map was calculated on the assumption b = d = + ,a = c = -.This map proved to be correct and revealedthe position of the bromine and the phosphorus atoms.After a structure-factor calculation based on the positionsof the heavy atoms, the positions of the nitrogen atomswere found from an observed Fourier synthesis.Refinement.-The crystal structure was refined by block-diagonal least-squares analysis ,lo in which the functionCw[lF,I - klFC1l2, summed over 3335 independent re-flexions, was minimized. Besides refinement of the posi-tions of the atoms, the thermal motions were refinedanisotropically for the bromine and phosphorus atoms ,and isotropically for the nitrogen atoms.After the first stages of the refinement R was 0.17 andthe weighted factor, R', was 0.09 [R' = {Xw/AF12/CwlF,12}*].At this point, 75 reflexions having IFc] values( 5 and lFol values of ca. 35 were removed from the re-finement , because of apparent inaccuracies in the measuredstructure-factors. The refinement was continued with3260 reflexions and the final values of R and R' were 0.12and 0.08. In the last cycle the changes in the positionaland thermal parameters were < 0.1 cr. The atomic scatter-ing factors were taken from ref. 11. The weightingscheme used was w = {wcl + $lFo12}:1, where w, is theweight from counting statistics and p is a constant; p =0.0005 in order to make <wlAF12), averaged over groupsof reflexions, approximately independent of IF/.The final atomic co-ordinates with the standard devi-ations as determined from the least-squares residuals are* For details of Supplementary Publications sec Notice toAuthors No.7 in J. Chew. SOC. (A), 1970, Issue No. 20.given in Table 3. Parameters of the anisotropic andisotropic temperature factors are given in Tables 4 and 5,TABLE 3Fractional atomic co-ordinates with standarddeviations in parentheses2:0*4053( 5)0*3387(5)0- 1425 (5)0*1682( 5)0.2476(6)0*3382(14)0*2669( 17)0.102 1 (1 6)0- 1674( 16)0.3739 (1 6)0*6079( 2)0.361 7(3)0.5247(2)0-2666 (3)0*1642( 2)0*0768(2)0.362 7 (2)0.1327( 2)0.301 6( 2)- 0*0129( 2)Y0.7 9 12 (6)0.6459(6)O.3368( 6)0*3355(6)0*6980(6)0.6732( 18)0*4666(20)0.3191 (1 8)0.4666( 17)0*7202( 18)0.8849 (3)0.9968 (3)0.72 10( 3)0-7 898 (3)0.3499 (3)0.121 8( 2)0.1 1 19( 2)0.3644 (2)0.7 120( 2)O*5005(2)0.0911(6)0*2966(6)0.1352(6)- 0.1263 (6)- 0.231 8( 6)0.1485( 17)0*2764( 2 1)- 0*0399( 19)- 0.2 166( 18)- 0*0776( 19)0*2563(3)0.0882 (3)0-6164(3)0-3466( 3)0.13 13( 3)0.2014( 3)0.02 76 (3)- 0.3063(3)-0.3877(3)- 0*3678(3)ParametersfactorsTABLE 4(A: x 10-4) of t ~ i e anisotropic temperature ' exp[-i2x2(h2a*2Ul1 4- - - --+21zita*b*~,, +- - .)I with standard deviations in parenthesesAtom U,, U,, U,, 2U1, 2U,, 2U,,P(l) 177(28) 59(22) 183(26) 86(41) llO(39) 298(48)191(28) 90(23) 128(26) 87(42) 86(39) 218(46)132(26) 77(22) 136(26) 74(41) 124(38) 156(44)181(28) 77(23) 136(26) 91(42) 39(38) 200(46)162(27) 83(23) 87(23) 81(41) 44(36) 146(43):$; 584(18) 184(11) 478(16) 500(23) 393(21) 877(29)239(12) 211(11) 173(10) 29(18) 142(17) 163(19) :${ 560(17) 233(11) 372(13) 519(23) 332(20) 762(27)229(12) 435(14) 283(12) 331(22) 300(21) 393(22) ::{:{ 250(12) 124(9) 238(11) 97(18) 211(17) 187(20) ::[:! 190(11) 132(9) 262(11) 190(17) 178(16) 314(19)214(11) 176(10) 227(10) 178(17) 2S7(16) 288(19)::[!!I) 282(12) 226(10) 212(11) 226(19) 122(17) 379(20)193(12) 306(12) 176(11) -34(19) -42(18) 227(19)340(14) 112(10) 221(11) 43(18) -30(16) 341(21)TABLE 5Parameters (A2) of the isotropic temperature factorsexp[ --B (sin28)/h2] with standard deviations in parenthesesAtom B Atom €30.4 (0.2)N(2) 3) 1.2 (0.3)0-8 (0-3)N(4) 1.0 (0.3) N(l) 1-7 (0-3) N(6)respectively.Observed and calculated structure-factorsare listed in Supplementary Publication No SUP 20339(1 8 pages , 1 microfiche).*A comparison of the estimated standard deviations ofthe structural parameters with those obtained in the struc-ture analysis of N,P,Br,l indicates that the errors arehigher in the case of N,P,Br,, by a factor 2 or 3. Althoughthe number of reflexions measured in proportion to thenumber of parameters adjusted (47 : 1 for N,P,Br,, 21 : 1 forN,P,Br,,) is much larger for N,P,Br,, the ratio for N,P,Br,,l o D. W. J. Cruickshank, ' Computing Methods and the PhaseProblem in X-ray Crystal Analysis,' Pergamon Press, Oxford,1961,I1 P. A. Doyle and P. S. Turner, A d a . Cryst., 1968, A24, 3901972 107 1is still a reasonable number. We think, therefore, that themain origin of the relatively low accuracy of the presentstructure is the occurrence of many weak reflexions andthe difficulties experienced in obtaining accurate valuesfor their intensities.Also, the complex shape of thecrystal presented difficulties in connection with the ab-sorption correction.DESCRIPTION AND DISCUSSIONIn the crystal structure two molecules occupy ageneral position in the triclinic unit-cell of space-groupPi. A view of one molecule is shown in Figure 1.Br(4 PBrQ)FIGURE 1 View of a molecule projected along a line havingdirection cosines -0.96693, 0, and -0.26882 with respect tothe orthogonal axes a, c x a, and ce-+-+* 3The central part of the molecule is formed by apuckered ten-membered ring of alternately phosphorusand nitrogen atoms.Data on least-squares planesthrough various sets of ring atoms are given in Table 6.The most interesting point is the position of the atomP(4), which deviates 0-54 A from the best plane throughall ring atoms, or 0-76 A from the best plane throughthe remaining nine atoms. As a consequence of thelocation of P(4), the bromine atom Br(8) has shortnon-bonded distances to three of the ring atoms, viz.3.33, 3.64, and 3-71 A to the atoms N(l), P(5), andP(3). These contacts are shorter than the usuallyaccepted van der Waals distances l2 (3.45 Br - - - N,3-85 A Br 9 P). The special location of P(4) is alsoreflected in the dihedral angles * which can be assignedto the ring bonds and which are included in Figure 1.These angles further indicate that the (N-P), ring has are-entrant angle at the atom N(1).The bond lengths and valence angles with theirstandard deviations are given in Figure 2.The tenindependent P-Br bonds vary from 2.156 to 2.195 A,and, on the basis of the least-squares standard deviations(0.006 A) do not all have the same length. The variationof the P-Br bond lengths around the molecule is rather* For sign convention, see Ref. 1.l2 L. Pauling, ' The Nature of the Chemical Bond,' CornellUniversity Press, Ithaca, New York, 1960.irregular. No correlation has been found with thepositions and orientations of particular atoms or groupsof atoms of surrounding molecules, i.e. the longerP-Br bonds cannot be explained by intermolecularbonding interactions, as has been possible for thecrystal structure of N,P,Br,.,The P-N bond lengths vary from 1.541-1.606 A,mean 1.571(17) A.The largest deviations from themean are shown by the P-N bonds at either side of theatom N(2), but the mean length of the bonds P(2)-N(2)and P(3 N(2) (1.574 A) is the same as the mean lengthrelatively large temperature factor (Table 5) and,apparently, its position could not be determined veryaccurately. We believe, therefore, that all P-N bondsmay be considered as being equal.The valence angles (Figure 2) inside the ring have avery regular pattern. There are two different groups ofN-P-N angles [means 114.3 (9) and 120.5(9)"] and three(1.570 1- ) of the eight remaining bonds. N(2) has aTABLE 6Equations of least-squares planes through various setsof ring atoms and distances (A> from these planesEquations of planes :*(I) All ring atoms included in the calculation:-0-6804X + 0.7212Y + 0.13062 = 0.6661(11) All ring atoms, except P(4) :-0.6297X + 0.7602Y + 0.16992 = 0.9704(111) Phosphorus atoms only :(IV) Nitrogen atoms only :-0.7464X + 0.6698Y + 0*08702 = 0,3082-0.6777X + 0.7936Y + 0.19102 = 1.1161Distances from planes(1) (11)0-048 0.104'(') 0.127 0.1860.114 -0.013P(4) -0,644 -0.746P(2)P(3)0.146 0.014"1) p(6) 0.116 0.1370.034 0.226N(3) N(4) 0,324 0.139N(2) -0.266 -0.286N(6) -0.289 -0.314(111)-0.107- 0.034- 0.3290.2130.2670.0060.41 70.610- 0.326-0.313(IV)0.2890.3820.009- 0.8060.0110.296-0.160- 0.0130.088- 0.2094+ * The equations are referred to the orthogonal axes a,-> 3*G x a, and:; X, Y, 2 are expressed in A.groups of P-N-P angles [means 131.2(11), 136.6(11),and 14343(10)"].It follows from the equivalence ofall P-N bond lengths that there is no dependenceof the length of a P-N bond on the value of the P-N-Pangle adjacent to it. Such a correlation has beendescribed by Schlueter and Jacobson for the structureof the related compound N,P,Cl,,. They found adecrease in the P-N bond length accompanying anincrease in the angle P-N-P. However, it was shownby Zoer and Wagner 1 that a relationship between angleP-N-P and P-N bond length in cyclophosphazenescannot be postulated in general.The bond lengthsand valence angles in the N,P,Br, molecule furthersupport this latter point of view1072 J.C.S. DaltonMean values of bond lengths and valence angles inthree bromocyclophosphazene molecules are comparedin Table 7. There is a good agreement between theTABLE 7Mean values (and individual standard deviations) ofbond lengths (A) and valence angles (") in bromo-c yclophosphazenesP-N P-Br BrP-Br N-P-N P-N-PN,P,Br, * 1.576 (8) 2-162 (4) 102.1 (1) 118.6 (6) 121.4 (6)N4P4Br, t 1.675 (6) 2-171 (2) 103-9 (1) 120.1 (4) 131.0 (4)N,P,Br,, 1.571 (17) 2-175 (6) 103.3 (2) 116.8 (9) 136.9 (11)* Ref. 6. t Ref. 1.P-N bond lengths, the P-Br bond lengths, and theangles Br-P-Br, but, as expected, there are somevariations among the angles N-P-N or P-N-P.It is clear from the data presented in Figures 1 and 2that the molecule N5P5Brlo has a pseudo-mirror plane,cyclophosphazene compound, as realized in its crystalstructure, will be the result of very subtle intra- andinter-molecular energy effects (neglecting entropyeffects) .The foregoing argument gives, in a qualitative sense,the reason why two such closely related compoundsas N5P5Brlo and N,P5Cllo have so different crystaland molecular structures.In this case, it would bedifficult to determine precisely what energy effectsare responsible for the different structures. It can beshown, however, that these effects are not of a simpleintramolecular steric nature. The N5P,Brlo ring con-formation can be combined with chlorine atoms assubstituents (P-Cl 1-98 A, in the direction of the P-Brbonds) without obtaining too close C1 C1 contacts.Conversely, the N5P,C1, conformation with bromineatoms as substituents (P-Br 2.17 A) also leads to asterically acceptable structure.FIGURE 2 Bond lengths (A) and valence angles (") with standard deviations are given in parenthesespassing through the atoms N(l) and P(4) and orientatedperpendicular to the ring.The corresponding chloro-phosphazene N,P,Cl,, also has a pseudo-mirror planenormal to the (N-P), ring, but, apart from this, the twomolecular conformations are entirely different. Thering in N5P5CllO is nearly planar and is characterizedby two re-entrant P-N-P angles, situated at eitherside of the pseudo-mirror plane.The bonding in cyclophosphazenes has been discussedby Craig and Paddock 1 3 9 1 4 and by Cruickshank.15The model of a double system of d,+, bonds, extendingthroughout the cyclophosphazene ring and constructedfrom mutually perpendicular and equally strong com-ponents in each particular P-N bond, is generallyaccepted now.Owing to the double x bonding systemthe total x overlap in a P-N bond will be highly in-dependent of the dihedral angle of that bond, andenergy differences between different conformationsof a cyclophosphazene ring may be assumed to be small.As a consequence, the molecular conformation of al3 D. P. Craig and N. L. Paddock, J . Chem. SOC., 1962, 4118.l4 N. L. Paddock, Quart. Rev., 1964, 18, 168.l5 D. W. J . Cruickshank, J .Chem. SOL, 1961, 5486.Some short intermolecular Br * N and Br 9 - Brdistances are tabulated in Table 8. The three shortestBr - - * Br distances (3.45, 3.45, 3.48 A) suggest a vander Waals radius for bromine of cn. 1.73 A, i.e. 0.2 ATABLE 8Short intermolecular distances (A) between bromineatoms of the standard molecule and nitrogen orbromine atoms of surrounding molecules (Br - - - N <3.70 A, Br * * - Br < 3.90 A)Br(3) * - N(2I) 3.57 Br(3) * * . Br(6I) 3.45Br(5) * * N(4II) 3.38 Br(4) * - Br(6VII) 3-85Br(9) * - - N(4III) 3.61 Br(6) - . - Br(9II) 3.80Br(l0) * - - N(2Iv) 3.42 Br(6) - - * Br(GVII1) 3-81Br( 1) - - - Br(8V) 3.84 Br(7) - * Br(7IX) 3.48Br(1) * - - Br(9VJ) 3-86 Br(7) - - Br(9'11) 3-84Br(2) * * - Br(8Vx1) 3.45 Br(8) * * Br(l0v) 3.70positions, relative to the reference molecule at x , y , z :II1Roman numerals as superiors refer to the following equivalent- x + 1, -y + 1, -2 + 1- x , --y + 1, --zI11 - x , -y + 1, --z - 1 VIII - x , -y, -2IV x , y , z - 1 v --x + 1, -y + 1, - zsmaller than that radius given by Pauling.12 It maybe noted that the intramolecular non-bonded contactsVI --x + 1, -y + 2, -2VII x , y + l , zIX - x , -y, - z - 1972 1073in NSP5BrI0, as well as the intermolecular Br - Brdistances in the crystal structure of N,P,Br,,l are consis-tent with the small value of 1.73 A.We thank Professor Aafje Vos for her interest in thiswork, Dr. T. Spoelder-Migchelsen and H. Zoer for helpduring the experimental stages, the Netherlands Organiz-ation for the Advancement of Pure Fhsearch (z.W.0.)for financial support for this work, and Dr. D. B. Sowerbyfor a sample of the compound.[1/1964 Received, 22nd October, 1971
ISSN:1477-9226
DOI:10.1039/DT9720001069
出版商:RSC
年代:1972
数据来源: RSC