MOLECULAR STRUCTURE OF PHOSPHORUS OXYTRICHLORIDE
Jan. 5, 19.54
[CONTRIBUTIONFROM
THE
2(i 1
DEPARTMENT O F CHEMISTRY, PURDUE UNIVERSITY]
An Investigation of the Molecular Structure of Phosphorus Oxytrichloride by
Electron Diffraction1 BY GORDON R. BADGLEY AND R. L. LIVINGSTON RECEIVED JULY 24, 1953 The molecular structure of phosphorus oxytrichloride has been reinvestigated by electron diffraFtion using the visual COTrelation procedure. The following structural parameters were obtained: P-Cl = 1.996 i 0.02 A , , P-0 = 1.45 & 0.05 A. and L ClPCl = 103.5 =I=1'. These values are significantly different from those obtained in an earlier investigation but are in excellent agreement with a recent microwave determination.
Introduction An early electron diffraction investigation2of the molecular structure of POCI, gave P-C1 = 2.02 f 0.03 A., LClPCl = 106 f 1" with the P-0 distance assumed to be 1.58 A. A recent microwave study3produced the following parameters: P-CI = 1.99 f 0.02 A., LClPCl = 103.6 & 2' and P-0 = 1.45 f 0.03 8. The discrepancy between these results suggested a reinvestigation of the molecular parameters of Poc13 by electron diffraction.
last fractionation was chosen for photographing and the purity was estimated at better than 99%. A series of photographs was obtained using the method described by Brockway.4 Several exposures were made on each of several Kodak 33 plates using a camera distapce of 107.8 mm. and an electron wave length of 0.05938 A . (as determined from transmission patterns of zinc oxide). Measurements of three of the best plates by three independent observers led to the 90 values listed in Table I and the visual curve shown in Fig. 1.
+
Experimental The sample of POCls was prepared by numerous distillations of the commercial product. The middle cut of the
VIS e
TABLE I QUANTrTATIVE ELECTRON DIFFRACTION DATAFOR POc1, Max
1 2 2 3 3 4 4 5 5 6 6 7 7
8 8 9 9
gob8
15.13 18.38 21.68 25.35 28.42 30.55 33.99 37.76 41.62 48.39 52.99 56.23 58.90 61.33 64.14 67.86 71 41 74.38 77.40 81.35 83.22 87.84 91.34 95.05 97.66 100.31 102.25 Av. Av. dev. I
10 10 11
11 12 12
13 13 14 14
B6
Min.
P/PO
(0.938) (0.964) (0.991) (0.999) (0.993) (0.967) (0.982) 0.991 0.998 1.004 0.993 1.005
.... .... 0.991 0.998 1.005 1.019 1,010 0.981 1.006 0.997 1.000 0.995 0.997 (0.995) (1.011) 0.999 0.007
B5 c4
CS dqa
(0.934) (0,958) (0.991) (1 ,000) (0.989) (0.945) (0.975) 0.987 1.010 1,000 0.988 0.996 (1.007) (0.987) 0.994 0.996 1.001 1.013 1.001 0.983 1.005 0.996 .997 .993 ,991 ( ,992) (1.009) 0.997 0.007
(1) From the M.S. Thesis of G. R. Badgley, Purdue University, 1951-1953. (2) L. 0. Brockway and J. Y. Beach, THISJOURNAL, 60, 1836 (1038). (3) Q . Williams, I. Sheridan and W. Gordy, J . Chcm. P h y s . , 20, 164 (1862).
A4 B2 c8 D8
I
,
l
I
,
l
2
,
l
3
,
l
,
l
4
r,A. Fig. 1.-Intensity
and radial distribution curves for POCIS.
Correlation Procedure Theoretical intensity curves were calculated according to the equation
using the punched card m e t h ~ d . ~Assuming Can symmetry, the P-C1 distance was fixed a t 2.00 A. and the remaining parameters (P-0 distance and LCIPCI) were varied. The range of parameters covered is shown in Fig. 2; the dotted line encloses the parameter field of acceptable models. Of these, the curves for models Bb and Cd are shown in Fig. 1 (4) L. 0. Brockway, REIS.Mod. Phys., 8, 231 (1936). ( 5 ) P.A. Schaffer, V. Schomaker and L. Pauling, J. Chem P h y s , 14, 659 (1946)
CORDOX R. BADGLEY N I.D
2 02
The limits of acceptable shapes for the unresolved pair, maxima seven and eight, are indicated by these curves.
\
P-0
DlStQnCe
P-GI
Distance
,
142 -144 -2138 _140- _ - -146- 148 150 1.52 1.54 00 2 00 200 2.00 2.00 2.00 2.00 2.00 2.00
1701. ;Ti
R. L. LIVINGSTON
A radial distribution curve6 was calculated using the function rD(r) =
Cqmax e-bg2
sin (?r/10)qr
Pmin
where e-bq' = 0.10 for 9 = 100. This curve is shown in Fig. 1; its strongest peaks correspond to the values listed opposite R D in Table XI.
6
TABLE I1
INTERATOMIC DISTANCESIX POCla
ACCEPTABLE
FROM
MODELS XIodel
Bs B4
B5 Be C4 Cb C8 c 7
RD" Fig. 2.-Parameters
for calculated models of POCl,;.
Tn setting the limits of the acceptable models, the unsatisfactory models of the parameter field were rejected for the following reasons. Curves for models Ad, -46 and A s were rejected because of the discrepancy in the shape of maxima seven and eight. Curves AI,A2 and -43 are satisfactory as regards these features but are not correct in that the intensity of maximum eleven is much too low, being much lower than in curve &. The latter curve is shown in Fig. 1. The main discrepancy in curve B2 is also the low intensity of maximum eleven; curve C3is better in this respect but the shape of the fifth peak is entirely wrong in this curve. Curve €37 has maxima seven and eight wrong and maximum ten is virtually absent. In curve CS the seventh maximum is almost completely missing and the eleventh maximum is too intense. Of this group, curves Bz and CSare shown in Fig. 1 since they are actually somewhat more nearly acceptable than C3and &. Of the 105" models, D?, D4 and D6 give rise to curves with a completely wrong shape for maximum five. D8 is nearly satisfactory in this respect but has maximum seven too weak, maximum eleven too strong and minimum eleven too deep as compared with minimum ten. Curve Dg is shown in Fig. 1. In addition to the above, curves were calculated for a few models in which account was taken of vibrations of the atoms. 115th reasonable values of vibrational damping factors, no significant changes in the agreement of curves resulted. The mean 9/90 values for models B5 and C5 are listed in Table I. The values in parentheses are not included in computing the averages because the measurements on these features were not considered to be sufficiently reliable. The average deviation from the mean of the 9/90 values for all acceptable models was 0.006 or 0.007.
P-0
r-Cl
0-c1
1.421 1.440 1.459 1.479 1.437 1.456 1.474 1.492 1.445
2.002 2.000 1.998 1.998 1.996 1.994 1.992 1.990 1.986
2.913 2.920 2.937 2.947 2.904 2.911 2.928 2.945 2.916
Cl-CI
L C l P-CI
3.133 3.130 3.127 3.127 3.144 3.141 3.137 3.134 3.110
1U3'
103' 103' 103" 104' 104' 104' 104'
RESULTSOF THIS INVESTICATIOX P-0 distance 1.45 & 0.054. 1.99s =!= 0.02 P-C1 distance 0-C1 distance 2.92 =!= 0 . 0 5 -?. C1-C1 distance 3 . 1 3 i0.03 A. C1-P-C1 angle 103.5 =!= l o Radial distribution curve.
+
a
The values of the interatomic distances listed for the acceptable models in Table I1 are those obtained by multiplication of the assumed distances by the average of q/qo for each model. The final results, based on a model midway between B5 and C5, are also listed in Table 11.
Discussion The results of this investigation are in excellent agreement with those given by the microwave determination (see Introduction), The present results are not, however, in agreement with those of the earlier electron diffraction study.? A model corresponding t o the parameters of the earlier diffraction work gives a curve which is in complete disagreement with the visual curve shown in Fig. 1. In the earlier investigation, the diffraction patterns did not extend t o as large scattering angle as the ones obtained here and, in addition, the interpretation of the shapes and intensities of maxima seven and eight were different in the two investigations. The latter fact accounts for the difference in shape parameters observed previously and in this investigation. Gordy3 has calculated, for comparison with the observed distances, the length of the P-Cl bond using the Schomaker-Stevenson rule and the length of the P-0 bond; only the P-CI distance is in good agreement with the calculated value and a possible explanation of the decrease in the P-0 distance has been suggested. LAFAYETTE, INDIANA (6) L. Pauling and
L.0. Brockway, THISJOURXAI.,
57,2684 (1935).
