COMMUSICATIONS TO THE EDITOR
July 20, 1962
2837
sense) as a methyl or ethyl group, and in quite a separate class from the t-butyl group.
ably assignable to a double bond stretching mode. In contrast, other unsaturated compounds in this DEPARTMENT OF CHEMISTRY NORMAX L. ALLINGER series exhibit a very strong Raman band in the WAYNESTATE UNIVERSITY LESLIEA. FREIBERG vicinity of 1625-50 cm.-’ (see below) and the DETROIT2, MICHIGAN SHIH-EN Hu absence of such a band effectively eliminates I1 and RECEIVED APRIL 19, 1962 I11 as potential candidates5 This confirms our previous conclusions, based on preliminary X-ray symmetry consideration^.^ TETRAPHENYLCYCLOBUTADIENE DERIVATIVES. IV.1 LLOCTAPHENYLCUBANE”*A DIMER OF Further X-ray diffraction experiments on the TETRAPHENYLCY CLOBU+ADIENE needle-like single crystals of I have shown the Sir: lattice to be body-centered tetragonal, with space We wish to report compelling evidence that the group Ci,-141/a, established without ambiguity. previously reported2 tetraphenylcyclobutadiene di- Cell constants were derived from single-cryostaland mer of m.p. 430’ is “octaphenylcubane” (I), the from powder data:,a = 19.489 f 0.019 X.; c = first organic compound to incorporate eight carbon 10.647 A 0.011 A. The morphological needle atoms bonded to each other symmetrically in the axis is [OOl]; the only developed faces are [loo]. The measured density, by flotation and by disshape of a cube. placement, is 1.173 0.008 g. ~ m . - ~the ; density calculated on the basis of four molecules of (C,H6)8 per cell is 1.171 g. ~ m . - ~ Reasonably . complete three-dimensional diffracted-intensity data have been gathered about both [loo] and [ O O l ] with the Weissenberg technique, and a full structure determination is in progress by D.R.P. Arguments I Ph based upon the X-ray evidence indicate that discrete molecules of formula CS6H4,, exist in the I crystalline state, and that each such molecule is a t Dimerization of the radical intermediate from the a position of symmetry within the cell. Though thermal decomposition of (4-bromo-1,2,3,4-tetra- the present state of the X-ray investigation does not phenyl-cis& l13-butadienyl)-dimethyltin bromide unequivocally establish the structure of the dimer leads to I in 85% yield.2 Its non-identity with as I, none of the data is in contradiction to this hexaphenylbenzene, from which i t cannot be dis- structure and, in the absence of an equally satistinguished on the basis of analysis or melting point, factory alternative structure, the total body of chemwas established by comparison with an authentic ical and physical evidence appears quite compelling. sample prepared by the method of DiltheyS3 The extreme insolubility of this dimer precluded By analogy with the dimerization products of sub- normal chemical studies; reasonable solubility was stituted cyclobutadienes obtained by previous obtained only in refluxing diphenyl ether, solutions worker^,^ the most likely structure for our dimer of which yield well-formed crystals on cooling and might be octaphenylcyclooctatetraeneKe (11) or from which its molecular weight was established the octaphenyltricyclooctadiene, III.6b However, ebulliometrically8 in agreement with the X-ray result above. Exposure to hot concentrated sulfuric acid leads to a violet solution which yields a Ph Ph ph P h G P h water-soluble ring-sulfonated mixture, and warming with concentrated nitric acid leads to indisPh - Ph criminate nitration and an intractable mixture of Ph Ph Ph Ph Ph Ph products. In an attempt to prepare the possible precursors, I1 I11 I1 and 111, of the cubane I, i t was found that two the saturated character of the dimer is unequiv- unrelated reactions unexpectedly yielded the same ocally established by its solid-state Raman spec- new product. Thus, while the reaction of equivtrum6 which exhibits a strong phenyl band a t 1605 alent quantities of 1,4-dilithio-cis,cis-tetraphenylem.-’ (also present in the solid state infrared spec- butadiene and 1,4-diiodo-cis,cis-tetraphenylbutatrum a t 1603 cm.-l) but contains no band reason- diene (route (1)) is expected to produce the D 2 h (“tub”) form of octaphenylcyclooctatetraene (IIa) (1) Part 111, H. H. Freedman, J . OYE.Chem., in press. (2) H. H. Freedman, J . A m . Chem. SOC.,83, 2195 (1961). and the treatment of 3,4-dibromotetraphenyl(3) W. Dilthey and G. Hurtig, Chem. Ber., 61, 2004 (1934). cyclobuteneg with phenyllithium in etherla (route
*
4
phi,
11
(4) (a) R. Criegee and G. Louis. Chcm. Ber., 90, 417 (1957); (b) R. Criegee. et al., ibid., 93, 1553 (1960); (c) M.Avram, et al., Tetrahedron Letters. 1, 21 (1961); (d) E. H.White and H. C. Dunathan, Abstracts of the 134th A.C.S. Meeting, Chicago, Ill., 1958, p. 41-P; ( e ) M. Avram, et al., Chem. Ber., 93, 1789 (1960). ( 5 ) (a) A recent publication (E. H. Braye, W. Hubel and I . Caplier, J . A m . Chem. SOC.,83, 4406 (1961)), reports the formation of a hydrocarbon (Ci4Hio)x of m.p. 421’ which appears to be identical with our dimer and, on the basis of its method of formation and ultraviolet spectrum, is assumed by the authors to be 11. (b) Structure I11 has been assigned to the dimer by M. Tsutsui (Chem. and I n d . , 780 (1962). ( 6 ) The authors are greatly indebted to Dr. John Evans for supplying the Raman data.
(7) Private
