J.Org. Chem., Vol. 41, No. 3, 1976 473
A Spirobicyclic Pentaoxyphosphorane, (PO,)(C&&(C&&,) Reagents”, F1. E. Sievers. Ed., Academic Press, New York, N.Y., 1973, pp 143-158. (4) J. D. Roberts. “Nuclear Magnetic Resonance”, McGraw-Hill, New York, N.Y., 1959, p 63. (5) T. P. Forrest, D. L. Hooper, and S. Ray. J. Am. Chem. Soc., 96, 4286 (1974). (6) I. M. Armitage, L. D. Hall, A. G. Marshall, and L. G.Werbelow in ref 3, pp 322-323. (7) Program written by G. D. Smith, Montana State University. (8) M. R. Willcott, 111. and R. E. Davis in ref 3, pp 159-169. (9) For all PDlQM calculations, coordinates of all conformations were taken so that the oxygen atom and the four carbon atoms of the tetrahydrofuran ring were superposed in all conformations. Comparison of any two R factors is a comparison of the goodness of fit for two different positions for the substituents on the four nonfused carbon atoms of the sixmembered ring. (IO) We applied the Hamilton R-ratio tests” to observed R factors in the following way. The conformation or mix of conformations with the lowest R factor was deemed the “best fit conformation”. Any other conformation or mix of conformations was considered a “restrained conformation” with the number of protons unable to assume the “best fit conformatlon” defining the number of restrained parameters (b). Further, the number of experimental observatlons (n) was taken as the number of observed NMR peaks used in the calculation of slopes and the number of parameters ( p ) was taken as the number of protons in the molecule. , compared to the calculated RThe observed R-ratio, RrestlRbeSt~ h was ratio values, f?b,n--p.e, which have to be exceeded to reject the hypothesis that the two conformations are indistinguishable. The confidence level at which the hypothesis is rejected is (1 a)X 100%. Values of Rb.n--p,a, were taken from Hamilton‘* or calculated by the relation
-
Rb,n-p,e
= [ ( b h - p)Fa -i- 11
‘’*
as given by Hamilton.” For l a , n = 25, p = 12, b = 6. The conformational mix of 9 4 % open, 16% folded with the lowest R factor is not significantly different from the totally open conformation but is significantly different from any conformational mix of 75% or less open form at the 9 0 % confidence level The stioled area in Fioure 1 shows the limits of the 90% level. W.-C, Harkion, “Statistics in Physical Science”, Ronald Press, New York, N.Y., 1964, pp 157-162. Reference 11 pp 2 16-222. F. Johnson, Chem. Rev., 68,375 (1968). D. H. Wertz and N. L. Alllnger, Tetrahedron, 30, 1579 (1974). For I b n = 30, p = 16, 6 = 10. The 60%:40% conformational mix is significantly different from conformational mixes of 90% open or greater and 35% open and less at the 90% confidence level. The stipled area of Figure 1 shows the limits of the 90% level. R. E. R. Craig, A. C. Craig, and G. D. Smith, Tetrahedron Leff., 1189 (1975). (a) K. L. Servis, D. J. Bowler, and C. Ishii, J. Am. Chem. Soc., 97, 73 (1975); (b) K. L. Servis and D. J. Bowler, hid., 97, 81 (1975). J. Wolinsky and M. Senyek, J. Org. Chem., 33, 3950 (1968). P. Wilder, Jr., and C. V. A. Drinnan, ibid., 39, 414 (1974). For 2 n = 25, p = 20, b = 6. I. D. Blackburne, A. R . Katritzky, and Y. Takeuchi, J. Am. Chem. Soc., 96, 682 (1974). The compounds used in this study were prepared by established literature procedures, with physlcal and spectral data in agreement: l a , ref 1; Ib, A. P. Krapcho and B. P. Mundy, J. Heterocycl. Chem., 2, 355 (1965); 2, ref 13; 3, B. P. Mundy, Org. Prep. Proced. ht., 7, 21 (1975). I
Crystal and Molecular Structure of a Spirobicyclic Pentaoxyphospharane, (p05)(c6H4)2(c6%)
Raghupathy Sarma,*l* Fausto Ramirez,*lb and James F. Mareceklb Biochemistry and Chemistry Departments of the State University of New York at Stony Brook, Stony Brook, Neiu York 11794 Received July 15,1975 The crystal and molecular structure of a monoclinic form of spirodicatecholphenoxyphosphoranewas determined by single-crystal x-ray diffraction techniques. The space group is P21/, with four molecules in a unit cell of dimensions a = 6.910 A, b = 15.305 A, c = 14.779 A, fi 88.5O. The final R factor for 2840 independent reflections is 8.7%.The molecular structure does not correspond to an ideal or a “slightly distorted” trigonal bipyramid or tetragonal (square) pyramid. It is suggested that the phosphorus and its five oxygen ligands have the skeletal geometry of a 15O-turnstilerotation configuration. Compounds, with five oxygen ligands covalently bonded to phosphorus, e.g., the trimethyl phosphite-phenanthrenequinone adduct 1, have been extensively studied from structural2 and synthetic3 points of view. The geometry around the phosphorus in the analogous pentaoxyphosphorane 2 is TBP,4 as established by x-ray crystallography,j which revealed also the existence of severe crowding around the central atom. It was suggested5 that the relatively planar ring minimizes the crowding and contributes to the stability of this type of compound.
A
Spirobicyclic pentaoxyphosphoranes, e.g., 4 and 5, are also known, and are relatively stable.6 The introduction of an additional five-membered ring into the monocyclic pentaoxyphosphoranes is accompanied by a significant dispIacement of the 31P NMR chemical shift toward lower magnetic field, i.e., by a decrease in the shielding of the P nucleus by electrons, cf. 4 vs. 1 and 5 vs. 3. This effect could reflect significant differences in molecular structure between spirobicyclic and monocyclic pentaoxyphosphoranes, but there are no x-ray data on the molecular structure of 4 and 5. Six-membered rings do not cause this effect.637
4 2
R = CH3;&’= + 23.0 ppm R = C&; dlP = + 27.0 ppm
474 J. Org. Chem., Vol. 41, No. 3, 1976
Sarma, Ramirez, and Marecek
There seems to be a tendency toward the TBP geometry in five-coordinate phosphorus, as there is in five-coordinate antimonyS8Although there are as yet no x-ray crystallographic data for the caged polycyclic pentaoxyphosphoranegJO 6, the molecular structures of the two analogous compounds 7 and 8 are known.11J2 The skeletal geometry about the P atom is close to TBP in 7 and 8, but there are significant distortions in them from ideal D3h symmetry, e.g., 0(2)-P-0(3) = 106' and 0(5)-P-X(4) = 84' in 7 and 8, while 0(1)-P-0(5) = 170' in 7 and 168' in 8. It was suggestedl1J2 t h a t the molecules of phosphoranes 7 and 8, in their ground state in the crystal, have geometries which begin t o resemble those of TR c o n f i g ~ r a t i o n s . ~IfJ ~so, when those molecules are placed in solution, they are able t o undergo relatively rapid intramolecular permutational isomerization, since they require relatively little energy t o reach the barrier configuration of the TR mechanism.13
6
R = CF3;
$'P = + 42.4
ppm
5 0-
-7
x= x=
NCgkkj; b'p
=+56.8ppm
&=
c(cF~)~;
+17.9ppm
The purpose of this investigation was t o ascertain the effect exerted on the static stereochemistry of a pentaoxyphosphorane by the presence of two five-membered rings in the spiro configuration. T h e molecule chosen was the spirodicatecholphenoxyphosphorane,6b9.
6 -4
$'P= + 30.2 ppm
T h e results of x-ray crystallographic analysis of a number of spirobicyclic phosphoranes having combinations of P-0 and P-C14 (lo), P-0 and P-F14 ( l l ) ,P-S and P-C15 (12), and P-0, P-S, and P-C I6bonds have been interpreted in terms of more or less distorted TP4 skeletal geon-ietries.I7a
Sc::p3
a;:@2 CH3
2
x
=CH3 X=F
12
c
The TP geometry has also been suggested from x-ray data for certain spirodioxatricarbophosphoranes with fourand five-membered rings;18 in contrast, some monocyclic dioxatricarbophosphoranes with a four-membered ring have slightly distorted TBP configuration.lg Unfortunately, the effect of the spirobicyclic feature on the skeletal ge-
ometry of phosphoranes is obscured in these exampleslP16J8 by the circumstance that they have quite differe n t "ligand subset ~ y m m e t r y " , i.e., ~ ~ ,the ~ ~equatorial and the apical subsets of ligands cannot be made up of the same element, and electronic, as well as steric, factors may contribute t o the striking differences in molecular geometry in these cases. Several examples of phosphoranes with P-C bonds, and with certain combinations of P-C and P-0 bonds, are known to have slightly distorted TBP geometries according t o x-ray crystallography.21-28 Experimental Section Preparation of Spirodicatecholphenoxyphosphorane (9). This substance was prepared as described.6bThe crystals for x-ray analysis were obtained from a mixture of benzene and hexane at 20"; they had mp 108-110" and 631P +30.0 ppm vs. HsP04 (CH2C12 solution). Crystal Data. Compound 9: C18H1305P; monoclinic; P21/,; a = 6.910 & 0.01 A, b = 15.305 & 0.01 A, c = 14.779 f 0.01 A, 0 = 88.5O; 2 = 4; d calcd = 1.445 g/cm3;d meas = 1.437 g/cm3. Data Collection and Structure Determination. The crystals were extremely sensitive to atmospheric moisture and dissolved readily during exposure. This was prevented by sealing the crystal in a capillary tube filled with a small amount of PnOa at one end of the tube. The crystal used for data collection had dimensions 0.7 X 0.4 X 0.4 mm. Intensity data were collected using a computer controlled CAD4 automatic diffractometer using Cu K a (1.542 A) radiation. Nearly 3000 reflections were collected using 0-28 scan with a scan width of 1 . 2 O . Four reflections were measured periodically to monitor any crystal deterioration. No such effects were observed during the entire data collection. Ninety-six separate measurements were made for each scan; the first 16 and the last 16 measurements were summed to provide the background measurements for each reflection. There were approximately 2840 reflections with IFd2 > 20((Fd2)with u estimated from counting statistics. Since the crystal was sealed inside a capillary tube, an empirical absorption correction was applied;29 this correction was determined by performing an azimuth scan for a reflection occurring at a x value of approximately 90". The variation in intensity of such a reflection for the different azimuth angles is dependent on the thickness of the crystal traversed by the incident and the reflected beams and this variation is used to calculate the transmission factor for the crystal for all the other reflections. The transmission factor varied from a value of 1.00 to approxifnately 1.30 during the azimuth scan. Structure amplitudes derived in the usual way were used in calculating a sharpened Patterson map3which was used in deducing the position of the phosphorus atom.'Remaining atoms were located from a series of structure factors followed by weighted Fourier syntheses calculated using wlFdexp a as coefficients, where the ~ Q is ! ~the ) ; observed ~~ weighting factor LO = tanh ( I F ~ ~ F P ~ / Z IFd structure amplitude, IF4 is the contribution from the known atoms, and Qj corresponds to the unknown atoms. The structure was refined by full matrix least squares, minimizing the function ZLO(AF)~. The weighting scheme used was of%he'formw = l/(a IFd clFd2) where a and c were of the order dF2'Fmi, and 2/Fm,,, respectively.31The atomic scattering factors for all the nonhydrogen atoms were taken from a standard source. The final leastsquares cycle was performed with anisotropic temperature parameters for all the nonhydrogen atoms. No attempt was made to locate the positions of the hydrogen atoms. Refinement was stopped with an R value of 8.7% for 2840 reflections. A final difference Fourier synthesis P&sd - Pc&d was calculated. The maximum electron density in this synthesis was 0.383 e/A3, indicating that no atoms other than hydrogen remain to be located. The final positional and thermal parameters are given in Table I. The observed and calculated structure factors are given in Table 11.
+
+
Discussion Molecular S t r u c t u r e of 9. The interatomic d i s t a n c e ~ 3 ~ and bond angles and their standard deviations are listed in Table 111. Some intramolecular distances between nonbonded atoms are collected in Table IV. Equations of leastsquares planes, and deviations of certain atoms from these
J . Org. Chem., Vol. 41, No. 3, 1976 475
A Spirobicyclic Pentaoxyphosphorane, (P05)(C6H4)2(C6&)
Table I11 Bond Distances and Angles in Spirodicatecholphenoxyphosphorane(9) A. In PO, Group
Table I A. Atomic Coordinates and Their Standard Deviations (in Parentheses) Atom P 01 02 03 04 05
c1
c2 c3 c4 c5 C6 c7 C8 c9 c10
c11 c12 C13 C14 C15 C16 C17 C18
Atom B11 P 01 02 03 04 05 C1 C2 C3 C4 C5 C6 C7 C8 C9 C10 C11 C12 C13 C14 C15 C16 C17 C18
X
z
Y
0.3153 (0) 0.0817 ( 2 ) -0.1440 ( 7 ) 0.3530 (3) 0.4036 ( 3 ) 0.1522 (7) 0.2426 ( 3 ) 0.0619 ( 7 ) 0.2650 (3) 0.0236 (7) 0.3046 (3) 0.3157 ( 7 ) -0.1759 (11) 0.4233 ( 4 ) -0.0010 (11) 0.4507 ( 4 ) -0.1027 (12) 0.1896 ( 4 ) 0.1828 (11) 0.2262 ( 4 ) 0.2470 (4) 0.3540 (11) 0.5217 ( 5 ) 0.0078 (13) 0.5596 ( 5 ) - 0 . 1 7 0 8 (14) 0.5309 ( 5 ) -0.3497 (14) 0.4582 (5) -0.3548 (12) 0.1061 (5) -0.0976 (16) -0.2661 (18) 0.0542 ( 6 ) -0.4275 (17) 0.0804 ( 6 ) 0.1634 ( 6 ) -0.4252 (16) -0.2603 (14) 0.2183 (5) 0.5321 (12) 0.2183 ( 5 ) 0.1609 (5) 0.5336 (13) 0.1404 ( 5 ) 0.3594 (14) 0.1783 (13) 0.1718 ( 5 ) B. Thermar Parameters B22
0.0194 Q.0038 0.@213 0.0051 0.0219 0.0041 0.0259 0.0050 0.0233 0.0043 0.0219 0.0050 0.0258 0.0037 0.0237 0.0037 0.0285 0.0042 0,0240 0.0035 0.0250 0.0036 0.0298 0.0040 0.0329 0.0042 0.0329 0.0042 0.0274 0.0045 0.0422 0.0038 0.0528 0.0038 0.0446 0.0047 0.0394 0.0051 0.0341 0.0043 0.0240 0.0050 0.0305 0.0044 0.0344 0.0043 0.0330 0.0046
0.4739 ( 0 ) 0.4841 ( 2 ) 0.4159 ( 3 ) 0.3972 ( 3 ) 0.5708 (2) 0.4904 ( 2 ) 0.4254 ( 4 ) 0.3878 ( 4 ) 0.3906 ( 4 ) 0.6065 ( 4 ) 0.5617 ( 4 ) 0.3267 (5) 0.3071 (5) 0.3469 ( 5 ) 0.4092 ( 5 ) 0.4297 ( 6 ) 0.4212 ( 7 ) 0.3770 ( 6 ) 0.3384 ( 6 ) 0.3447 ( 6 ) 0.5854 ( 5 ) 0.6623 (5) 0.7095 ( 5 ) 0.6846 ( 4 )
B33
B23
B13
B12
0.0030 0.0035 0.0041 0.0035 0.0033 0.0036 0.0031 0.0037 0.0032 0.0033 0.0033 0.0047 0.0053 0.0054 0.0042 0.0063 0.0074 0.0069 0.0071 0.0058 0.0043 0.0044 0.0041 0.0034
-0.0020 -0.0002 -0.0014 -0.0023 -0.0008 -0.0022 -0.0006 -0.0016 -0.0005 -0.0010 -0.0021 -0.0002 0.0002 0.0012 0.0015 0.0021 -0.0047 -0.0045 -0.0029 -0.0019 -0.0012 -0.0020 0.0010 0.0008
0.0041 0.0035 0.0040 0.0052 0.0051 0.0024 0.0013 0.0030 0.0033 0.0026 0.0029 0.0019 0.0002 -0.0028 0.0001 -0.0043 -0.0053 -0.0034 -0.0069 -0.0079 0.0003 -0.0004 0.0006 0.0035
0.0002 0.0018 0.0016 -0.0017 0.0009 0.0023 -0.0010 -0.0003 -0.0014 -0.0003 0.0003 0.0005 -0.0008 -0.0015 -0.0012 0.0002 0.0003 -0.0016 -0.0021 -0.0006 0.0002 -0.0002 0.0001 0.0008
planes, are given in Table V. The dihedral angles formed by pairs of these planes are shown in Table VI. The data in Tables 111-VI demonstrate that the skeletal geometry about the phosphorus atom in the spiropentaoxy: phosphorane 9 does not resemble a regular T B P , or even a reasonably distorted TBP. This is brought out by Figure 1, which is a drawing of the molecule of 9 generated by computer from the experimental data, and by formula I, Figure 2, which is the hypothetical ideal T B P arrangement of the PO5 group. In formula I, the five-membered rings formed by the phosphorus atom and the two catechol bidentate ligands, A and C , are placed in apical-equatorial skeletal positions in accordance with previous observations5 in the related system 2. A superficial examination of the data suggests, a t first, the skeletal geometry of a regular TP for compound 9. However, a more detailed analysis reveals significant deviations from ideal TP symmetry. This is shown in Figure 3
Bond distances,
Bond angles, deg
O(1)-P-O( 5 ) 160.02 (2.3) P-0(1) 0(2)-P-0(4) 151.36 (2.1) P-0(2) 0(3)-P-0(4) 105.37 (1.0) P-0(4) O(2)-P-O( 3 ) 103.26 (1.0) P-0(5) O( 1)-P-O( 3 ) 101.99 (1.0) P-0(3) 0(3)-P-0(5) 97.98 (0.9) 0(4)-P-0(5) 92.38 (0.8) 91.47 (0.8) O( 1)-P-O( 2) O( 2)-P-O( 5 ) 83.25 (0.8) O( 1)-P-O( 4 ) 83.04 ( 0 . 8 ) B. In the Phenoxy Ligand (Ring B)
122.51 (1.8) 0(3)-C(3) 120.37 (2.6) 117.9 5 (2.5)