1250
Vol. S6
COMMUNICATIONS TO THE EDITOR
be the result, not of an unusually high exo rate with a normal endo rate, as is commonly assumed, b u t of a normal exo rate with a n unusually slow endo rate." But what is the factor t h a t might be responsible for the slow endo rate? An examination of the preferred path for the leaving group in the endo derivative reveals t h a t this path brings the departing group very close to the opposite side of the rigid endocyclic system. Indeed, if one constructs an intimate ion pair with a representative anion, such as chloride, situated along the perpendicular to the face of the carbonium ion a t C-2, a t distances equal to the sum of the two ionic radii, the anion is observed to be severely crowded against the 5,6-bridge.l 2 It is apparent t h a t the chloride ion must depart along some other path which is more favorable sterically, b u t more costly in requiring greater separation of charges in the ion pair. In conclusion, the results clearly indicate t h a t a high exo-endo rate ratio in a norbornyl derivative does not provide a unique basis for concluding t h a t the derivative undergoes ionization with participation of the 1,6-bonding pairs to form a nonclassical norbornyl cation. Acknowledgment.-This investigation was assisted by Project No. XT(ll-1)-70, supported by the ,Itornic Energy Commission, and by Grant 19878, provided by the National Science Foundation. (11) H . C. Brown, "The Transition State," Special Publication T o . 16, T h e Chemical Society, London, 1962, p p . 140-158, 174-178 (12) This hypothesis of steric hindrance to ionization in rigid bicyclic systems is readily susceptible t o experimental t e s t , and such tests are underway in research in progress with D r . I Rothherg.
R . B. WETHERILL LABORATORY HERBERT C. BROWN PURDUE UNIVERSITY FRANK J. CHLOUPEK INDIANA MIS-HONREI LAFAYETTE, RECEIVED JANUARY 6, 1964
The Diamagnetic Anisotropy of a Borazine Ring. Correction
A
Sir: The importance of data on the diamagnetic anisotropy of borazines1,2 with respect to the problem of the aromaticity of these compounds3 renders necessary t h e correction of some errors made in the paper by U'atanabe, Ito, and Kubo,' especially in the paragraph headed "Evaluation of Diamagnetic Anisotropy." T h e experimental molar susceptibilities of boron compounds listed in the compilation by FO@xlaclearly show t h a t Pascal's rule concerning the additivity of atomic susceptibilities, X A , for obtaining the molar susceptibilities, X M , does not hold f o r boron compounds. This might be due to the fact t h a t Pascal calculated the X A values4b of nitrogen, oxygen, hydrogen, and chlorine from X M values of carbon compounds, so t h a t these values do not reflect a n y effects specific to boron compounds, such as back coordination. Moreover, Pascal's value for the atomic susceptibility of boron,5 X B = -7.2 X li1-6, was obtained solely from the X M of B(OCH3)a. I t remains to be proved t h a t this value could indeed be used as such for calculating the X M (1) H . Watanabe, K . Ito, and SI. K u b o , J A m . Chem. SOL., 82, 3294 (1960) ( 2 ) K . 1,onsdale and E. Toor, A r i a C r y s f . , 12, 1048 (1959). (9) J . C Sheldon a n d B C. S m i t h , Quari Rev. (London), 1 4 , 200 (1960). ( 4 ) ( a ) C,. Foex, "Tables d e Constantes et Donnees NumCriques. 7. Constantes Sklectionnees, Diamagnetisme et Paramagn&isme," Masson et Cie, Paris, 1957, pp. 28-29; (b) ibrd.,p. 222. ( 5 ) P. Pascal, Compt. rend., 218, 57 (1944).
values of boron compounds. It is, therefore, apparently as yet impossible to calculate the diamagnetic susceptibility of the borazine molecule, its plane parallel to the magnetic field, xi,from x = ~ X B :ZXN (jXH as has been done.''6 The X B required for calculating x 1 ' was obtained from the X M of BC13. However, the value -67.0 X 10P6 for the molar susceptibility of boron trichloride listed in this paper' is incorrect. According to these authors,' this value was taken from the compilation of foe^.^ who, however, does not list the -67.0 X 1 0 - 6 value, but gives only the value -59.9 X this being an experimental value given by Pasca1.j The value' XM = -67.0 X 10V6 is probably an approximation of the xhf = -66.9 X calculated by Pascal5 by adding the x.4 of C1, - 19.9 X 1i)Y6,and the X A of B, -7.2 X lOF, i.e., - ( 3 X 19.9 7.2) X 1 0 P = -66.9 X 10k6. I t is very diHicult to justify this procedure for obtaining X B , ~which amounts to estimating X B from an experimentally nonexisting value for the X M of BC13, -67.0 X 1 0 F . Moreover, as a result of an arithmetical error, these whereas from authors' obtained X B = - 3 . 2 X their values they should have obtained X B = -(G7.0 - 3 X 19.6) X = -S.2 X 1 0 F . This value, X B = -8.2 X 1 0 - 6 , is the X B from B(OCHa)3, X B = - 7 . 2 x 10-6, modified b y adding and subtracting different x.4 values of chlorine. The erroneous value X B = -3.2 X was used by those authors' for calculating the X I ,of borazine
+
+
+
Xi
+
f 3XN AXH - ( 3 X 3.2 f 3 X 3.3 -3T.l X
= 3XB =
=
+ 6 X 2.93) X
Had the value X B = -8.2 X IOe6 been used, the whereas the would have amounted to -52.1 X would yield X I ,= -49.1 X value X B = -7.2 X I t must here again be stressed that the X N (ut least) was chosen' quite arbitrarily. Thus no significance a t all should be attributed to the "corrected" x values -52.1 X 1W6or -49.1 X lop6. T h e molar susceptibility of borazine as determined,' x h f = -49.6 X 1 V 6 , and the x 1 values -52.1 X or -49.1 x 1W6,yield the following values for the diamagnetic anisotropy, A x , respectively AX = 3(XM
-
Xmi)
A x = -.7(49.6-52.1)
X 10-6 = + 7 . 5 X 1 O P
and Ax
=
--3(49.6-49.1) X 10-6
= -1.5
X 10-6
I t is thus clear that the otherwise important conclusions,' based on the value Ax = -36 X 1i)-6,1are as yet unsubstantiated. On the other hand, the other values of Ax, + i . 5 X 10-6 and -1.5 X 1W6, create the impression of borazine not possessing a n y diamagnetic anisotropy. However, from the very moderately accurate data of Lonsdale and Toor2 for the diamagnetic anisotropy of B-trichloroborazine, A x = - 18 x 1 0 - 6 , i t seems t h a t borazine and its other derivatives would show some diamagnetic anisotropy. Obviously this problem requires further investigation. (6) Neither can t h e x , , of borazine be assumed t o be equal t o t h e x I I of t h e isoelectronic benzene molecule as suggested.' T h u s , t h e calculated xll-values of molecules isoelectronic with benzene, such a s pyridine. pyrimidine, pyrazine, and sym-triazine, differ appreciably from t h e x , . of benzene.
DEPARTMENT O F I S O R C A T I C A S D h N A L Y T I C A L CHEMISTRY THEHEBREW UNIVERSITY K . .I MGSZKAT JERUSALEM, ISRAEL RECEIVEDDECEMBER 3 , 1963
