836 B. G. DEBOER, A.
ZALKIN, AND
D. H.TEMPLETON
are more nearly equivalent in the mass spectra of BBH18 and the tetraboranes. Formation of B4Hs as well as B4H10 is suggested by the fact that the absolute intensity of i n j e 48 remained constant while that of nz/e 50 decreased. The behavior of I(33+) relative to 1(52+) (Figure 3) can be explained by the formation of tetraborane(8) and tetraborane(l0) both of which would contribute more ion intensity t o inje 52 than to v z j e 53 and cause a relative decrease of I(53+). At high temperatures, pentaborane is formed and would contribute more ion intensity to m j e ti3 than to inle 52 causing the observed increase in the ratio I(53+)/1(52-), Although the ratios 1(37+)/1(52+)and 1(27')/1(52+) increase with increasing temperature, formation of dior triboranes is not necessarily indicated. Even though the absolute intensities of these peaks increase between approximately 130 and 170", relative intensity changes for these two peaks are considerably less than that for the tetraboranes. Fragmentation of the B4H8, B4Hl0, B,Hg, and the hexaborane, all of which are believed t o be formed, could account for the observed increases.
Inorganic Chemistry Conclusions
The unusual fragmentation pattern in the molecular beam inass spectrum of octaborane(l8) is strong evidence for a structure in which two B4H, groups are joined by a single B-B bond. An appearance potential of 11.2 0.1 eV determined for B4H9+also supports this structure. Pyrolysis of octaborane(l8) appears t o lead to tetraborane(8) and tetraborane(l0). Formation of pentaborane(9). possibly pentaborane(ll), and hexaborane(s) is indicated; however, their formation may result froin the pyrolysis of the tetraboranes. Increase with temperature of diboranp and triborane ion intensities is observed but can be due to fragmentation of other pyrolysis products. This work shows also that comparison of the molecular bean1 and total intensity mass spectra can help identify the pyrolysis products of a boron hydride. The conclusions drawn froin such a comparison and the pyrolysis study are in agreement for octaborane(l8).
+
CONTRIBUTION FROM THE DEPARTMENT O F CHEMISTRY AKD
LAWRENCE RADIATION LABORATORY, UNIVERSITY
O F CALIFORNIA, BERKELEY, CALIFORNI.4
94720
The Crystal and Molecular Structure of Phosphorus TrifluorideTris(difluoroboryl)borane, B,F, .PF,l BY BARRY G. D E B O E R , ~ALLAN ZALKIN,
AYD
DAVID H. TEMPLETOK
Received September 6, 1968 The crystal structure of phosphorus trifluoride-tris(difluoroboryl)borane, B I F ~PF,, has been determined by an X-ray diffract,ion study of single-cr>-stal specimens. The orthorhombic unit cell, space group Pnrna, with a = 13.803 rt 0.005 .A, b = 10.578 k 0.005 and c = 6.075 f 0.005 .A, contains four formula units. The calculated density is 1.82 g/cm3. The structure was solved by statistical methods and refined by full-matrix least squares to a conventional R of 0.306 for 703 data collected by counter methods (6.7% for the 603 nonzero data). The molecule consists of a central boron atom, tetrahedrally bonded to three BFz groups and the PF, group in such a way that the molecule has approximately 3m (Cav)point symmetry, one mirror of which is required by the crystal symmetry. Distances found (uncorrected for thermal motion) P-F, 1.51 .4 (all k0.015 8). are: B-F, 1.305 A; B-B, 1.68 *A; B-P, 1.825 8
a,
A&;
Introduction Timms has recently reported3 the synthesis of a number of novel boron fluoride compounds by reactions of the high-temperature species, boron monofluoride, on cold surfaces. Cocondecsation of BF and PF3 produces B4F,.PF3, a pyrophoric and water-sensitive substance which readily sublimes (nip 55", bp 74") to form well-developed, colorless crystals which are stable a t room temperature. The present X-ray diffraction
investigation was undertaken to verify the proposed structure
and to determine the bond distances and angles in this molecule. Experimental Section
(1) Work done under the auspices of the L S Atomlc Energy Commisslon ( 2 ) Sational Science Foundatlon graduate fellon, 1964-1967 (3) P L Tlmms, J A m Chem S O C ,89, 1629 (1967).
Samples of the compound, sealed into thin-walled glass capillaries, were provided by Professor Timms. Crystals formed quite readily as the compound was sublimed back and forth in the
F,B\
\
Vol. 8, No. 4,April 1969
PHOSPHORUS TRIFLUORIDE-TRIS(DIFLUOROBORYL)BORANE 837
capillaries under the very mild heating provided by a beam of light from a low-voltage lamp. Persistence seemed to be the only way to obtain a suitable crystal which did not have others nearby. Complete data sets were taken on two different crystals. The first crystal was a plate about 0.16 X 0.51 X 2.2 mm, where the intermediate dimension is the inside diameter of its capillary. A crystal this large %as acceptec' because there was no assurance of ever obtaining a better one and because of the low absorption coefficient, p, of 4.3 cm-l for Mo Kor radiation. The crystal was mounted with its long dimension, b axis, and capillary parallel to the instrument 'p axis. Preliminary oscillation and Keissenberg photographs gave the extinction rules, Laue symmetry, and rough cell dimensions. More accurate cell dimensions were derived from the Bragg angles of the hOO, OkO, and 001 reflections measured with a General Electric manual goniostat using molybdenum radiation and a Zr filter a t the receiving slit. The components of the or doublet were resolved ( ~ ( R l oK e l )0.70926 A) a t the highest angles, but the peaks were broad and the positions of reflections a t low x angles could not be accounted for except by assuming a 28 correction as a function of 'p. It is suspected that these difficulties were caused by the large size of the crystal, which w,s certainly longer and probably wider than the beam diameter, and by strains in the crystal (see below). I n spite of these troubles, a one-octant data set of 830 independent intensities, exclusivc of space group absences, was taken, using the same apparatus, by counting each reflection for 10 sec with crystal and counter stationary methods and a n approximately 3.5" takeoff angle to the X-ray tube anode. Of these, 173 TTere recorded as zero. Measurements were also made a t all of the locations in the region of the data set which correspond to space group absences. They were found to confirm the extinction rules of Okl, k $. 1 # 2n, and hkO, h # 2n. Individual backgrounds were measured for thoRe reflections which were seriously affected by streaking from other orders; otherwise, backgrounds were taken from a plot of background as a function of 28 a t various values of 'p and x. Three standard reflections which were measured a t frequent intervals during the course of data taking showed only small, random fluctuations. Measurements included reflections up to 28 = 50" ((sinO)/X = 0.595). Standard deviations mere assigned to the measurements of 5 % of the intensity or a fixed minimum (8, 5 , or 3 counts/sec, for low, intermediate, and high 28), whichever was greater. The structure was initially solved using this data set, which n a s then definitely seen to be of unacceptable quality. A second set of samples was provided by Professor Timms, and a better crystal %as sought. The second crystal had the shape of a cylinder -0.14 mm long and -0.15 mm in diameter (the inside diameter of its capillary). Nickel-filtered copper radiation n a s used in order to obtain as large a data set as possible from this smaller crystal. The smaller size of this crystal made the larger absorption coefficient, p , of 39 cm-' for Cu Ka radiation acceptable. The maximum variation in transmission factois is estimated to be less than 20%. The crystal was mounted with its cylinder axis, a, and capillary parallel t o the instrument 'p axis. The cell dimensions were determined as before, and found to predict correctly the position of off-axis reflections, in contrast to the trouble experienced with the first crystal. When a narrow receiving slit was being used t o obtain accurate cell dimensions, the peaks were sharp and well formed and the e doublet (X(Cu Kal) 1.5405 A) was resolved in a majority of cases. After the slit was removed, the peaks at low x angles showed a characteristic broad shape with a n irregular top in both 28 and w scans. At high x angles, a narrower, regular peak shape was maintained. This is again ascribed to distortions in the crystal, since the troubles with the two crystals (this and the previous one) seem t o be correlated with the direction of the capillary axis and not with the directions of the crystallographic axes. The strains might have been the result of growth on a curved surface or of slight flexings of the thinwalled capillaries as they were handled. However, it was found
that the calculated settings reproducibly gave the same part of the irregular peaks (and the tops of the regular ones) and that point counts taken a t the calculated settings on a few peaks a t x = 0 and 90 ' were in the same ratio as the integrated counts obtained by scanning through those peaks (in 28). A second data set of 703 intensities, including 100 recorded as zero, but not including space group absences, was taken in the same way as before. Backgrounds were likewise obtained as before, and, again, frequently measured standard reflections showed no systematic variations. The maximum value of 28 was 120' ((sin S)/X = 0,562). A standard deviation of 2.25 counts/sec or 7.5y0of the intensity, whichever WRS greater, was assigned to each measurement. Both data sets were corrected for Lorentz and polarization effects, but no corrections were made for absorption or extinction. All computations were performed on a CDC (3600 instrument. Zalkin's unpublished FORDAP and DISTAN programs were used for Fourier syntheses and distance and angle calculations, while a n unpublished Wilson-plot program by 1Iaddox and Maddox gave normalized structure factor magnitudes4 which were used in Long's sign-determining program6 as described below. The unpublished version of the Gantzel-Sparks-Trueblood leastsquares program which was used minimizes the function Zw(lkF,I - lFo1)2/2~lkFo/2, where F , and F , are the observed and calculated structure factors, k is the scale factor, and w is the weighting factor. For the early refinements, scattering factors for neutral phosphorus, fluorine, and boron were taken from standard tables.6 The phosphorus scattering factors were corrected for the real part of anomalous dispersion by +0.1 e- for the molybdenum data and +0.2 e- for the copper data, and the imaginary part was neglected. The final refinements were performed using the scattering factors listed by Cromer and Waber7 and both real and imaginary anomalous dispersion corrections given by Cr0mer.s a The anisotropic temperature factors used have the form: exp(-0.258,2,h,h,b,bl~,,), i, 1 = 1, 2, 3, where b, is the ith reciprocal cell length.
Results Four formula units of (BF2)3B.PF3are contained in the orthorhombic unit cell, space group Pnma, with a = 13.893 f 0.005 A, b = 10.578 k 0.005 A, and c = 6.075 0.005 A. The calculated density is 1.82 g/cm3. KO direct measurement of the density was attempted because of the highly reactive nature of the compound. The Laue syninietry and observed extinction rules correspond to space groups Pnnia and Pn21a. Choosing the former space group leads t o a successful solution with six atoms on the inirror plane in the fourfold special position set c : ( 5 , l/4, z; ' 1 2 x, lI4, - z ) and four more atoins in the general set of positions: (x, y, x; ' 1 2 - 5 , ' 1 2 y, ' 1 2 Z; x, 1 / z - y, 2; ' 1 2 x, y, ' 1 2 - z ) . Determination of Structure.-The Patterson function calculated from the first ( N o ) data set mas discouragingly featureless, consisting primarily of two positive regions a t y = 0 and l / 2 , with a negative region between (i.e., it was dominated by very strong OlcO reflections).
+
+
+
+
(4) J. Iiarle and I. L. Iiarle, A d a Cryst., 21, 849 (1966). ( 5 ) R. E. Long, Ph.D. Thesis, University of California, Los Angeles, Calif.. 1965. (6) "International Tables for X-Ray Crystallography," Vol. 111, T h e Kynoch Press, Birmingham, England, 1962, p p 202-203, 214-215. ( 7 ) D. T. Cromer a n d J. T. Waber, Acta C ~ y s t . 18, , 104 (1965). (8) D. T . Cromer, ibin'., 1 8 , 17 (1965). (9) No significant changes i n eithex the structural parameters or t h e quality of agreement were found as a result of t h e change i n scattering factors.
838 B. G. DEBOER,A. ZALKIN,A N D D. H. TEMPLETOX
Inorganic Chemistry
TABLE I FINAL POSITIONAL AND THERMAL PARAMETERS IN B4Fe* PF3 Y
X
0.0998 0.0497 0.0524 0.2311 0.2788 0.2264 0.3694 0.2697 0.2131 0.3598
(l)* (3) (2) (5) (7) (4) (3) (4) (2) (2)
‘18
‘14 0.1403 ( 3 ) ‘14 l14
‘14 ‘14
0.3788 ( 5 ) 0.4693 ( 3 ) 0.4004 ( 3 )
a
0.5699 0.7031 0.4588 0.5805 0.3255 0.1424 0.2902 0.7122 0.7870 0.7500
Bit5
(4)
5.8 (1) (10) 8.6 (3) 7.8 (2) (7) (10) 4.9 (3) (14) 7.3 (5) (7) 15.0 (4) (8) 8.9 (3) (8) 7.8 (3) (5) 12.3 (2) (6) 9.3 (2)
B22
B3a
6.0 (1) 15.8 (5) 9.2 (2) 3.2 (3) 4.0 (3) 8.9 (3) 9.7 (3) 4 . 2 (2) 4.9 (1) 7.0 (2)
Biz
8.5 (1) O C 11.8 (4) 0 17.9 (4) -- 2 . 4 (1) 3.7 (3) 0 5.3 (5) 0 0 3 . 6 (2) 8.6 (3) 0 4.4 (2) 0.0 ( 3 ) 1 . 4 (1) 7.6 (2) 11.3 (2) -.1.8 (2)
Bta
B23
0 . 0 (1) 0 4.8 ( 3 ) 0 -2.7 (2) -2.5 (2) 0.4 ( 3 ) 0 1.0 (4) 0 -1.3 (2) 0 4.2 (2) 0 -0.2 (2) 0 . 3 (2) 0.5 (1) -1.9 (1) - 3 . 1 (2) - 1 . 7 (2)
[ *I1
/2
(0.33, 0.27, 0.27)d (0.45, 0.44, 0.26) (0.49, 0.37, 0.25) (0.26, 0.22, 0.21) (0.31, 0.25, 0.23) (0.44, 0.33, 0.20) (0.40, 0.35, 0.24) (0.32, 0.24, 0.23) (0.40, 0.33, 0.21) (0.41, 0.34, 0.24)
&az.
a The units are The number in parentheses is the standard deviation in the least significant digit as derived from the leastThese are the principal root-meansquares matrix. Values expressed without the use of a decimal point are fixed by symmetry square amplitudes of vibration, in order of magnitude. For orientations, see Figures 1 and 2.
Several attempts a t refining various trial structures guessed a t from this Patterson function met with no success whatever. We reasoned that the strong OkO reflections would arise only if some large fraction of the atoms were concentrated a t or near the same y coordinate, as would be tLe case if the molecule were bisected by the mirror plane in Pnma. With this in mind, the Wilson-plot program was used to calculate normalized structure factor ~nagnitudes,~ Eh, for use in the statistical method of sign determination (centrosymmetric case). For this purpose, we employed Long’s program5 mhich iteratively applies the relation: sEh sBkEnEpL_A,where s is t o be read “the sign of.” After enough of the signs had been 17-orlted out by hand to yield a good starting set (of E’s) for input, a run of Long’s program using 99 E’s 2 1.5 gave one set of signs which was more consistent than any other and which mas arrived a t in the felT-estpasses. A Fourier synthesis using the signed E’s as coefficients gave ten largest pealis, six of them in the mirror plane, which formed the expected molecule. Least-squares refinement of isotropic phosphorus, fluorine, and boron atonis in the appropriate positions quickly reduced the R factor to 0.29, where R = Z 1 ’ k F , 1 - 1 F,I 1 1 2 ,k F , . It was immediately evident that the OkO reflections had given intensities which were systematically too large by factors between 3.6 and 5.2. A refinement in which the OkO reflections were given zero Tveight did not reduce R, but when they mere omitted from the calculation of R, its value became 0.26. When all atoms were allowed anisotropic temperature factors, R fell to 0.180 on all data and 0.145 if the OkO’s were omitted. These values could not be improved upon and a rerun of Long’s program with the OkO reflections omitted gave the same structure a second time. At this time the first @Io) data set was discarded and the second crystal and second (Cu) data set were obtained as described above. The structure found from the first data set refined to R = 0.217 with isotropic temperature factors. When all a t o m were allowed anisotropic temperature factors, the structure gave the final R value of 0.093 on all 703 data and this
-
can be reduced to 0.067 by omitting the contributions from the 100 reflections which were recorded as having zero intensity. In the last cycle of the refinement, no parameter shifted by as much as 1% of the standard deviation calculated for it from the least-squares matrix. The standard deviation of an observation of unit weight, defined as [ Z L C (1 k F , [ - 1 F , ) 2 / ( u- v)]’ where u is the number of data and 1’ is the number of independent parameters refined, mas 1.33. Tests of slightly noncentric structures in the othcr space group (Pn21a) gave neither better agreement nor significant deviations from the higher symmetry. The magnitudes of the three largest pealis on a final difference Fourier were 0.50, 0.38, and 0.30 The final positional and thermal parameters are presented in Table I, while Table I1 lists 1 F , 1 and F,.
Discussion The B4FE PF3 molecule is shown in Figure 1. The duplicate atom labels are caused by the crystallographic mirror plane n hich bisects the molecule and contains Fl(P), P, B ( l ) , B(2), Fl(2), and 1:2(2). The a t o m F2(P), P, B(1), and B(3) all lie in another plane to within the accuracy of this determination; Fl(3) and F2(3) deviate from it by only a small (-3”) rotation of the BF2 group about the B-B bond. This second plane makes an angle of 120 1 1’ with the mirror plane so that the molecule has approximately 3111 (C3%)point symmetry, only one mirror of which is required by the crystal symmetry. The fluorines on phosphorus atoms are in a staggered configuration with respect to the BYz groups on B(1), and the angles at B(1) are tetrahedral to within the experimental accuracy. Distances and angles found in this structure are presented and compared with those found for related compounds in Table 111. These interatonzic distances must be corrected for the effect of thermal motion in order to arrive at quantities which are comparable with the results of studies by other methods (e.g., spectroscopic and electron diffraction methods). This is
“01. 8, No.
4, April
1969
PHOSPHORUS TRIFLUORIDE-TRIS(DIFLUOROBORYL)BORANE 839 TABLE I1 OBSERVED AND CALCULATED STRUCTURE FACTORS (X9.0) for B&‘6. PFB K FOB FCA H.L0, 0 2 162-952 4 484 484 6 545-552 8 450 460
4 5 6 7 8 9 10
430-419 1 15 - 6 I1 0 -13 2 1 9 6 I98 H t L 5, 2 190-194 1 6 1 137 3 19 0 0 0 I1 14 103 - 9 8 1 58 6 2 4 23 -34 9 3 -87 5 11 1 9 2 1 0 1 94 186-180 6 6 5 62 3 2 3 -40 IO 148-147 25 20 7 30 33 4 37 - 3 5 H . LS 0, 1 11 $6 * a 8 5 0 -49 5 110 1 0 3 1 383 4 1 1 H I L = 2, 1 9 22 1 9 6 7 6 84 3 164 8 5 8 0 1 2 8 133 H l L I 3 . 5 7 17 i a 5 190 177 1 136 137 0 90 -90 8 30 -50 7 0 -11 2 403-416 1 6 7 -52 9 34 4 2 9 60 1 0 3 n 9 - m IO 0 -3 2 0 1 11 9 0 9 2 4 148 1 5 1 3 4 9 32 H - L 51 3 H,L= 0, 2 5 101 104 4 3 9 -34 0 8 5 -80 6 2 6 - 1.~ 5 o 777-866 . . 5 0 34 1 134 1 3 7 7 115-113 6 1 7 22 2 6 3 -64 2 351-330 4 115-118 a 2 3 34 1 50 48 3 61 -64 6 8 9 93 H,L3, 6 4 0 -16 9 6 7 -59 8 0 -32 i o a 3 -98 0 0 -23 5 1 9 77 10 5 4 -64 I 1 1 9 -19 6 38 -30 1 0 1 H.L0. 3 H l L l 2, 2 7 0 20 2 0 9 i i a -83 0 747 7 5 1 3 16 -26 8 0 -2 3 543-528 1 564-589 4 54 -64 9 1 6 12 5 I 5 2 156 2 272 264 5 0 -5 H.L. 5, 4 7 sa -98 3 44 -50 H.L4. 0 0 1 6 10 9 0 21 4 66 - 6 5 0 30 20 I 11 6 6 1 783-199 H,L. 0. 4 5 2 1 -16 0 -8 2 0 160 164 6 33 44 2 i i a 188 3 111-126 2 230 2 1 3 7 302 2 8 3 3 19 8 1 4 118-109 4 147-129 8 104 103 4 111 100 5 35 4 2 6 46 5 1 9 30 -20 5 199 1 9 1 6 2 5 24 7 6 1 -65 8 24 20 10 0 -14 6 2bG 2 8 1 HlL. 0. 5 1 1 26 26 7 2 8 9 287 8 1 6 11 1 164 1 8 3 H.L. 2, 3 8 0 5 H I L I 5, 5 3 7 8 -85 0 1 5 1 136 9 30 -11 0 1 7 16 1 37 2 1 IO I 1 -30 1 11 -4 5 9 7 104 7 63 -80 2 74 - 1 6 I 1 4 1 41 2 65 56 H I L = 0. 6 3 168 1 5 0 H.L* 4. 1 3 46 42 4 66 -64 0 30 -34 0 76 1 9 4 1 7 -12 2 0 -4 5 86 7 7 1 1 4 -15 5 5 1 -48 4 1 6 -8 6 169-172 2 373-312 6 16 18 H,L. I, 1 7 30 14 7 2 1 -21 3 9 1 -14 0 626-639 a 52 6 1 4 87 - 8 0 H,L5, 6 I 405 3 9 0 9 24 33 5 30 2 6 0 16 4 2 305 286 10 6 192-199 0 -2 1 2 8 40 3 194-113 H.L. 2, 4 7 1 4 5 147 2 0 -8 4 110 1 2 7 0 21 10 8 30 3 8 3 15 -6 5 1 2 11 I 122 1 2 0 9 a0 i i 4 33 39 6 31 18 2 183-191 6, 0 1 0 34 4 1 H v L 7 112-111 3 0 -1 11 4 1 4 5 o 543-511 4 5 2 5 1 H.L= I 592-574 8 129-128 4. 2 9 76 70 5 24 1 9 0 1 9 -10 2 401 3 7 7 10 6 5 6 0 6 70 - 6 7 1 25 -33 3 216-216 I1 46 -49 7 60 -70 2 3 1 -25 4 222-234 5 5* -56 H.L. I, 2 8 0 -1 3 9 6 -95 0 121-114 9 0 -11 4 258-249 s 9 9s 1 115-105 H,L. 2, 5 5 0 -13 1 44 - 5 0 0 0 39 6 47 -54 a 193-197 2 319 370 3 292 277 1 114 -90 1 P4 - 9 9 9 30 -40 4 99-106 2 55 -65 a 6 5 -57 1 0 7 0 72 5 ia*-ia2 3 39 30 9 74 -a2 H.16, 1 8‘ 91 6 1 5 1 153 1 0 23 2 9 0 263 261 7 107 1 2 3 5 i5 -3; H.L. 4. 3 1 52 -49 8 9 9 -99 6 17 - 1 4 0 0 22 2 3 2 5 335 9 25 -44 7 28 3 1 1 187-200 3 164 166 10 11 9 H i l i 2. 6 2 2 8 22 4 14 1 2 I 1 I5 2 5 0 0 -8 3 220 234 5 33 -36 H.L. 1. 3 I 17 15 4 229 228 6 2 1 -6 0 367-369 5 120-120 2 29 2 7 7 1 7 18 1 112 1 0 5 3 44 - 3 4 6 0 -7 E 30 -18 2 153 1 5 1 4 16 - 1 1 7 0 4 9 5 2 -61 3 13 - 9 5 30 1 8 a 17 19 10 1 6 1 4 4 76 -73 H,L3, 1 9 7 1 -10 HtL6. 2 5 146 1 4 0 0 290-275 I O 1 5 -20 0 191-194 6 8 1 74 I 585 579 H.LI 4, 4 I 1 9 7 205 1 64 1 2 2 42 -43 0 52 -65 2 a i 99 8 5 5 -42 3 71 -74 I 86 -a7 3 73 -68 9 34 30 4 1 4 3 136 2 77 a i 4 39 -40 10 23 2 6 5 103 98 3 0 10 5 260 260 H.L1, 4 6 1 4 -76 4 0 -18 6 142 139 0 194-191 7 90 - 8 5 5 11 5 7 24 -8 I 5 t -65 8 0 17 6 82 BO 8 2 5 -13 2 43 5 2 9 9 7 86 1 55 5 1 9 29 1 6 3 54 5 1 1 0 0 1 0 8 0 2 4 217-223 11 I6 - 2 0 9 15 1 6 H 1IO L - 2 6, 1 22 3 5 106 -95 H.L31 2 HIL= 4, 5 0 214-278 6 8C 71 0 116 1 6 4 0 30 -22 I 78 -16 1 30 2 9 I 298 301 1 30 32 2 0 -22 8 11 -25 2 269-243 2 0 -2 3 62 8 1 9 1 6 -19 3 283-284 3 5 5 -46 4 40 -46 H.L1. 5 4 112-113 4 25 -26 5 123-129 0 9 3 96 5 40 41 5 0 18 6 25 6 1 0 19 6 43 4 1 6 23 -26 7 46 5 3 2 64 -53 1 105-109 7 0 -8 a 30 -21 3 25 3 8 4 9 30 H I L 4, 6 9 16 -9 4 7a 56 9 35 3 6 0 33 2 8 H.L= 6. 4 5 67 -65 I O 24 1 9 1 17 20 0 98 1 0 2 6 38 - 3 5 I 1 26 -30 2 33 -19 I 5 5 -51 7 40 -48 H i L = 3s 3 3 16 -6 2 81 -79 8 0 14 0 156 1 6 3 4 15 7 3 38 - 3 9 H.L= Is 6 1 26 3 5 H , L = 51 1 4 6 7 62 0 42 37 2 19 -24 0 324 332 5 1 7 -18 I 0 -5 3 14 9 I 233 210 6 52 -55 4 156 152 2 170-176 7 0 0 5 128-134 3 152 159 3 0 22 8 0 1 4 6 2 13 6 7 6 -74 4 1 8 3 185 H.L. 6, 5 5 0 -6 7 99 -a5 5 191-193 0 11 2 6 H.L. 2, 0 n n 3 6 141-137 1 58 -66 0 662-653 1; 7 4 3 -48 2 25 1 8 1 356-339 IO 31 - 3 0 a 96 90 3 69 71 2 408-406 9 39 4 8 HIL. 3, 4 4 0 10 3 475 489 0 82 - 7 7 10 5 5 -53 5 37 -35 ~~
Figure 1.-The B ~ F ~ s P molecule, FI which is bisected by a crystallographic mirror plane which contains Fl(P), P, B(l), B(2), F1(2), and F2(2). The thermal elipsoids have been reduced to 30% probability contours for the sake of clarity.
particularly true in view of the large thermal parameters found here. Large thermal parameters are consistent with the observations that this is a low-melting, readily sublimable crystal. The accuracy of thermal corrections is poor, however, because of the ambiguity as to the correct model on which t o base their calculation. The minimum and “riding-model” corrections1° in Table I11 are included t o indicate the approximate magnitude of the calculated corrections.
i
‘
i
0 -ii
i 0
Figure 2.-Molecular packing in B 4 F 6 . P F , . This is one layer, approximately Ma thick, seen in perspective. The other layers are related to the one shown by the a glide perpendicular to c . (10)W. R. Busing and H. A. Levy, Acta Cryst., 17, 142 (1964).
6 HvL0
0 6. 16
5 6
6
2 1
226 -13 1 -1 3 0 5 H,L. 0 1 0 74 , 9 7 1
H
4 3 -41 17 15 0 -2 15 19 8s 5 38 -38 38 37
17 I 148-139 2 60 -56 3 224 218 4 64 - 5 9 5 232-231 6 1 6 -23 7 127 126 8
i
I 3
I1 0 -2 16 -28 H
67
3 8 -40 16 - I 4 H.LI 7. 2 0 245 255 I 0 -7 2 147-159 3 148-161 4 56 60 5 97 9 1 6 93 - 9 0 7 6 0 -64 8 62 5 6
IO
9
ia H.L.
0
H
-I
2 i -3; 1, 3
0 2 1 3 214 94 H 2I 13051 -38 3 5 9 58 4 34 35 5 34 - 3 1 6 46 -39 1 17 - 2 1 8 55 50 9 15 2 H,L* 7, 4 0 8 0 72 I ha 7 1 H 2 114-113 3 0 -5 4 111 1 1 6 5 46 4 1 6 2 9 -34 1 0 -9 a 36 34 H,L= 7, 5 h 0 83 -76 I 0 -19 2 0 3 3 34 -27 4 1 7 -8 5 O 1 1 H 6 15 8 H,L7, 6 P 1 5 18 1 30 -30 2 0 6 H,L= 8. 0 o 8 5 a6 1 1 8 4 176 2 104-105 3 88 -85
*
90
PO
H
5 2 9 9 310 6 146-134 7 6 3 -10 8 70 6 9 9 5 3 47 I0 0 9 H.L.0 181-184 8, 1
I 14 9 2 111-119 3 1 8 1 183 4 a 4 -97 5 129-128 6 38 4 5 i 1 5 -58 8 0 -8 9
0
b
1
10 1 5 23 H.L. 8, 2 0 145-153 I 47 5a 2 15 -12 3 198-199 4 1 6 9 159 5 56 45 6 17 1 1 7 99 -95 8 4 5 -48 9 3 1 30 H,L8. 3 0 104-107 I 7 5 72 2 225 2 3 1 3 113-102 4 8 3 -91 5 63 6 1 6 70 69 7 17 10 8 5 9 -57 HIL= 8. 4 0 1 7 18 1 46 -37 2 25 3 3 3 9 38
H
H
H
H
0 91 - 1I 6 5 66 2 6 -28 110 1 1 1 57 58 182-182 5 4 -52 8 5 80 0 48 24 -11 16 -15 91 2 16 -3 209-211 48 - 5 1 220 220 3a 44 9 3 -91 1 7 -4 11 68 2 3 13 41 - 4 8 9, 3 104-111 2 9 -27 24 25 0 7 49 - 4 0 49 -48 4 s 47 23 I1 30 -28 9, 4 39 21 49 -41 0 3 3 9 -40 0 8 16 -5 1 6 -12 91 5 8 6 76 28 29 2 3 -14 0 -1 15 I1 10, 0 27-102 95 a3 42 -36 68 66 196 203 6 5 -51 43 -44 7 7 -80 0 11 15 - 4 1 10. 2a 3 4 16 -4 3 6 -30 6 0 -65 1 2 8 129 2 5 29 17 I 0 3 6 0 2 10, 2 90 93 96 -90 9 9 -99 24 6 6 5 65 17 17 9 1 -92 58 54 3 1 22 10, 3 30 2 9 1 1 1 104 0 3 25 - 2 3 63 -72 I 1 15 16 -20 48 -41 10, 4 6 9 -55 24 22 30 25 0 12 0 10 0 -I 10, 5 0 9 3 1 28 15 -14 11. 1 38 -35 68 63 5 9 50 8 6 -85
4
5 6 7 8 H.L= 0 1 2 3
9 1 -96
52 0 63 0 -23 1 5 -14 11. 2 6 2 -58 46 -40 39 3 1 0 -7
46
17 0 -3 15 0 11 0 -6 11, 3 o 78 83 L 30 -29 2 1 7 -16 3 0 7 4 17 -3 5 16 -5 6 0 -12 HrL= 11, 4 0 2 9 -29 1 33 -22 2 16 1 8 3 36 4 4 4 26 -22 HILI 12. 0 0 105-104 1 72 - 7 5 2 25 11 3 4 3 32 4 25 18 5 6 0 -56 6 23 - 2 1 7 51 51 H,L. 12, I 0 70 6 9 1 0 9 2 17 6 3 50 -47 4 52 - 6 1 0 IO 5 6 0 -21 7 0 -11 H l L l 12, 2 0 58 55 1 11 2 5 2 2 5 -26 3 16 71 4 42 50 5 33 -25 6 3 1 -30 H,L= 12, 3 0 30 2 8 1 42 38 2 5 9 -51 3 1 7 -14 4 5 1 49 5 0 10 H.L= 12, 4 0 22 -29 1 2 7 -33 2 0 8 3 0 -2 13, 1 H.L0 17 2 1 I 46 37 2 17 -12 3 7 1 -71 4 0 -5 5 56 56 6 1 5 16 H , L = 13. 2 0 7 3 -80 1 0 1 2 63 63 3 0 5 4 42 -40 5 1 5 -5 H.L= 1 3 . 3 0 16 11 1 0 4 2 0 -6 3 22 -13 4 15 -7 H,L= 14. 0 0 24 - 7 1 24 -16 2 5 5 55 3 5 7 64 4 42 -36 5 47 -51 H v L - 14. 1 0 0 2 I 29 -25 2 0 -12 3 4 3 47 4 22 1 1 5 14 - 1 4 H.L= 1 4 , 2 0 5 3 -55 1 23 -22 2 22 1 7 3 22 26 H,L. 14. 3 0 4 8 50 H.L- 15. 1 0 4 3 48 1 1 5 -10 2 41 -50 4 5 6 7 H.L.
The authors can find no chemical or molecularpacking-forces explanation for the apparent difference (-0.04 A) between the axially and equatorially directed B-F distances. Since this difference could have arisen from a small (-0.02 A 5 2 ~ shift ) of the two borons, the authors are not willing to assert that it is real in the absence of supporting evidence. The up and down orientations of the BF2 groups were unexpected because the F-F distances between different groups would be greater if the BF2 groups were in a
I
840 B. G. DEBOER,A. ZALKIN,AND D. H. TEMPLETON
Inorganic Chemistry
TABLEI11
DISTANCES AND ANGLESIN B4Fs P F 3 AND COMPARISON COMPOUNDS Atoms
Dist, A
B(2)-F1(2) 1.330 (10)" B (2)-F2(2) 1.276 (10) B(3)-F1(3) 1.319 (6) B (3)-F2 (3) 1.293 (7) Av 1.305 k 0.015b (Thermal correction = 0.009, 0.051)c 1.685 (10) B(l)-B(2) 1.668 (7) ~(1)-~(3) AV 1.677 k 0.015 (Thermal correction = 0.001, 0.013) B(l)-P 1.825 (7) 1.825 k 0.015b (Thermal correction = 0.005, 0.026) P-Fl(P) 1.525 (6) P-F2 (P) 1.496 (4) Av 1.511 k 0.015 (Thermal correction = 0.013, 0.066)
B (21-B (3) B (3)-B(3) F1(2)-F2(2) F1(3)-F2(3) F1 (P)-F2 (P) F2 (P)-F2 (P) F2 (2)-F2 (3) F2 (3)-F2 (3) F1(2)-F2(P) F1(3)-F1 (P) F1(3)-F2(P)
Comparison dist, A
Comparison compd
1.30 1.32 i 0.035
1.67 i 0.045
1.836 k 0.012
H3B * PF3'
1.535 k 0.02 1.538 It 0.008
H3B * PFa
Intramolecular, Nonbonded Distances 2.719 (9) 2.725 (11) 2.180 (7) 2.176 (4) 2.23 i 0.024 2.339 (7) 2.321 (7) 3.217 (5) .3.181 (7) 3.299 (6) 3.246 (4) 3.210 (5)
PF3g
B2F4
Intermolecular F-F Contactsh F2 (P)-F2 (3) F1 (P)-F2(3) F1(2)-F1(3) F2 (P)-F2 (2) F2(2)-F1(3) F1(2)-F1(3) F2(P)-F2(3) F1 (P)-Fl(2) F 1(3)-F 1(3) F1(3)-F2(3) F2 (P)-F2 (P) F 1(2)-F2 (3) All others Atoms
F 1(2)-B (2)-F2 (2) F1(3)-B(3)-F2(3) Av B( 1)-B (2)-F1(2) B ( 1)-B (2)-F2 (2) B(l)-B(3)-Fl(3) B (1)-B (3)-F2(3) Av B (2)-B(l)-B (3) B (3)-B (1)-B(3) B (2)-B (l)-P B (3)-B (1)-P Av B(l)-P-Fl(P) B (1)-P-F2 (P) Av F1 (P)-P-FB(P) F2 (P)-P-F2(P)
hv
3.088 3.092 3.174 3.178 3.183 3.209 3.236 3.244 3.271 3.292 3.343 3.413 >3.5
(5) (5) (4) (6) (3) (4) (5) (7) (2) (5) (7) (5)
Angle, deg
113.6 (7) 112.8 (5) 113.2 k 1 123.6 (7) 122.8 (7) 124.5 (5) 122.7 (5) 123.4 k 1 108.3 (4) 109.5 (5) 111.1 (5) 109.8 (3) 109.7 2 1 115.2 (3) 117.2 (2) 116.2 i 1 101.5 (2) 101.8 (3) 101.7 1
+
[3.0951 [3.11] [3.1251 [3.22] [3.27] [3.28] [3.28] [3.30] [3.34] [3.37] [3.45]
Comparison angle, degk
Comparison compd
11151
[122.5]
BzFa
(Tetrahedral)
109.5
100 i 2 99.8 1
*
Vol. 8, No. 4,April 1969
METALDIOXIDES WITH RUTILE-RELATED STRUCTURES 841 TABLEI11 (Footnotes)
The number in parentheses is the standard deviation in the least significant digit as calculated from the standard deviations of coordinates. * The standard deviations of 0.015 A and 1’ listed for average valuee are based upon comparisons of distances expected to be the same and the authors’ past experience. They are believed to be better representations of the accuracy of these results than The two corrections for thermal motion given are a minimum correction those calculated from the standard deviations of coordinates. and a “riding-model” correction, respectively. Both are much less than the maximum possible correction. (See text and ref 10.) d “Tables of Interatomic Distances and Configuration in Molecules and Ions,” Special Publication S o . 11, The Chemical Society, Burlington House, London, 1958, p 11118. e L. Trefonas and W. N. Lipscomb, J . Chem. Phys., 28, 54 (1958). f R . L. Kuczkowski Q. Williams, J. Sheridan, and IT.Gordy, ibid., 20, 164 (1952) (microand D. R. Lide, Jr.. ibid., 46, 357 (1967) (microwave study). wave and electron diffraction study). Numbers in brackets are the authors’ calculation from cell dimensions and positional parameters given (X-ray crystallographic study). Q
XJ
propeller-like configuration. The observation that each fluorine has at least two intermolecular F-F contacts shorter than any intramolecular one (except between two fluorines bonded to the same atom) shows that the fluorine contacts within the molecule are not short enough to control the configuration. The BF, orientations are evidently governed by the meshing together of fluorine atoms on adjacent molecules (see Figure 2 ) . The existence of intermolecular F-F contacts shorter than the intramolecular ones also indicates that the B-P distance is not controlled by F-F interactions. This, and the fact that the B-P distance found here is not significantly different from that found for H3B ‘PF3, indicates that BF2 will not only formally replace hydrogen3 but is also quite similar to hydrogen in its electronic effects on the rest of the molecule.
E. I. DU
F O N T DE
NEMOURS
Inspection of the magnitudes and directions of the principal axes of the thermal ellipsoids reveals that the greatest anisotropy is in the motion of the fluorine atoms and that the largest amplitudes are in directions which are more or less perpendicular to bond directions. A considerable component of this motion may consist of librations of the PF3 and BF2 groups about the directions of their bonds to the central boron atom, but bending vibrations also appear to be important. Acknowledgment.-We thank Professor Timms for providing the samples which made this work possible and for providing them in such a convenient form. We also thank Professor Timins and N r . Ralph Kirk for helpful discussions regarding the significance of these results.
CONTRIBUTION No. 1463 FROM THE CENTRALRESEARCH DEPARTMENT, COMPANY, EXPERIMENTAL STATION, \vILI\IINGTON, DELAWARE19898
AND
Crystal Chemistry of Metal Dioxides with Rutile-Related Structures BY D . B. ROGERS, R. D. SHANNON, A. W. SLEIGHT,
AND
J. L. GILLSON
Received M a y $0, 1968 I n order to attempt a systematic correlation of the crystal chemistries of the transition metal dioxides with rutile-related structures, single crystals were grown of the oxides MOz, where M = Mn, 310,Ru, W, Re, Os, or Ir, and certain electrical transport and crystallographic parameters were accurately determined. Trends of these parameters within the rutile family are shown to be closely associated with variations in the occupancy of the transition metal d shells. Qualitative, one-electron energy level diagrams are discussed that appear to rationalize the electrical and crystallographic properties of these materials.
Introduction Many metal dioxides, particularly those of the transition metals, crystallize in structural types that are closely related t o that of the rutile form of TiOz (Figure 1). The properties of rutile were extensively discussed by Grant,’ and structural reviews of inost of the related dioxides and of compounds possessing the similar trirutile structure have been given by (1) F. A. Grant, Rev. Mod. Phys., 31, 646 (1959).
B a d and B a ~ e r .The ~ dioxides of Cr, Os, R u , Ir, Sn, and T a are tetragonal and isomorphous with rutile; those of V, Nb, N o , Tc, and W adopt distorted variants of the rutile structure in which the metal atoms occur in pairs along the c axis of the rutile pseudocell. Most of the latter oxides are monoclinic and are characterized by a tilt of the metal atoiii doublets in the [loo] direction of the rutile subcell. SbOz, however, is tetragonal and (2) W.H. Baur, Acta Crust., 9, 515 (1936). (3) G. Bayer, Ber. Deut. Keram. Ges., 39, 5 3 5 (1962).
