COMMUNICATIONS TO THE EDITOR
492
Vol. 53
The present a reement factor is R = 2 11 FO I - 1 Foll/Z: I Fof = 0.14 for the 889 observed reflections. Acknowledgments.-We wish to thank Dr. 51. F. Hawthorne for supplying us with several B9H15R derivatives until this favorable X-ray problem was chosen. We also thank the Office of Naval Research and the Office of Ordnance Research for support.
3.3; 0, 26.7; mol. wt., 240; found C, 70.0; H, 3.2. 0,26.5; mol. wt., 219 (cryoscopic in benzene); a single, sharp and strong band a t 5.68 p. 2,2'Diphenic anhydride was produced in over SOTo yield from the reaction of the peroxide I1 with triphenylph~sphine~ (or trimethyl phosphite). 2,2'Diphenic acid was obtained in the iodimetric assay of the peroxide 11. .A crystalline 1:1 adduct IV can be isolateda from the reaction of phenanthrenequinone I with DEPARTMENT OF CHEMISTRY FREDERICK E. WAX HARVARD UNIVERSITY PACLG. SIMPSON trimethyl phosphite. Ozonolysis of the adduct IY CAMBRIDGE 38, MASSACHUSETTS WILLIAMN. LIPSCOMB in methylene chloride as described above gave a RECEIVED NOVEMBER 18, 1960 mixture shown to contain diphenoyl peroxide (11, ca. 77% yield), phenanthrenequinone (I, ca. 205& yield) and trimethyl phosphate (111, ca. 95Yo yield), A NEW SYNTHESIS OF CYCLIC DIACYL PEROXIDES. DIPHENOYL PEROXIDE FROM PHENANTHRENE- by a combination of infrared spectrometric and iodimetric assays. (Pure peroxide, 11, was again QUINONE oia A PHOSPHORANE DERIVATIVE' isolated in 46-50y0 yields after one recrystallizaSir: tion.) The product of a comparable ozonolys~s~ We have developed a convenient method for the of the crystalline 1: 1 adduct derived from benzil preparation of certain cyclic diacyl peroxides,2 for instance diphenoyl peroxide (11), a substance in and trimethyl phosphite consists of benzoyl peroxide (ca. 3iVc) benzil (ca. 52Cj,) and trimethyl which the peroxidic linkage is part of an eight- phosphate. -4 possible mechanism for this new membered ring. reaction has bee11 a d ~ a n c e d . ~ *
'\,
' ,,
\
CI '
O( tl I' y- or I f
U(?H
I
The results of a typical experiment are: a mixture of phenanthrenequinone (I, 4.16 g.), trimethyl phosphite (2.48 g., one mole equivalent; an excess should be avoided) and methylene chloride (100 1111.) was kept 0.5 hr. a t room temperature, under nitrogen. The pale yellow solution was cooled to -70" and treated with ozone from a Welsbach T-23 ozonator. The reddish-brown color of the solution was discharged sharply as the ozone consumption reached the value of one mole equivalent. The reaction mixture was flushed with nitrogen and filtered to remove some phenanthrenequinone (0.4 1 8.). The methylene chloride was removed in ziacuo a t teiiiperatures not exceeding 20". The residue was extracted with cold methanol, which left 3.8 g. of crude peroxide, contaminated with some phenanthrenequinone. From this material, 2.4 g. (Z070 over-all yield) of crystalline diphenoyl peroxide (11) of over 9G% purity (iodimetric assay) was obtained after one recrystallization from methylene chloride-methanol (a solution containing 10 ml. of methylene chloride and 10 ml. of methanol per gram of crude peroxide was concentrated in zxzci~o,below 20"). Diphenoyl peroxide (11) is a colorless substance which can be preserved a t low temperature but which explodes violently when heated to ca. 70" or under impact. Calcd. for C14H804: C, 70.0; H, (1) Acknowledgment is made t o t h e Donors of the Petloleuin Research Fund administered by the American Chemical Society for partial support of this Research (Grant 2813-A) and t o t h e National Cancer Institute of t h e Xational Institutes of Health (Grant CY-47C9). (2) For a discussion of cyclic diacyl peroxides, see F D. Greene and W W. Rees, J. A m . Chem. Soc., 82, 893 (1960) aud references therein.
(3) M. A. Greenbaum, D. B. Denney and A . K. Hoffinann, i b i d . , 78, 2563 (1956). (4) F.Ramirez and N. B. Desai, i b i d . , 82, 2A,52 (1900). ( 5 ) F. Ramirez, R . B. Mitra and N. B.Desai, i b i d . , 82, 57133(1900).
DEPARTMENT OF CHEMISTRS FAL~STO RAMIREZ STATE UNIVERSITYOF XEWYORK iY.B. DESAI LOSC ISLAXD CENTER R . B. MITRA OYSTERBAY, S . Y . I Z E C E I V SOT.~EXBER E~ 29, l O G O DEHYDROHALOGENATION OF SIMPLE ALKYL HALIDES BY STRONG BASE ; EVIDENCE OF CARBENE INTERMEDIATES. EXTENT OF e-ELIMINATION'
Sir: I t was shown tliat iieoalkyl halides react wit11 strong base via n-eliininntion and subsequent inscrtiou t o give cyclopropanes. S i n ~ eisobutyl ~ ~ , ~chloride and soc1iu:ii or propyl sodimi yields rnctliylcyclopropaue ant1 isobutyleiie as major products it was of interest to determine the niechanisiii b;: which they are for-iiied. ~\lethylcyclopropanecaii arise via a-elimination, whereas isobut)-lene ma!result by either E2 (,8) elimination andl'or a-eliniination, followed by hydride transfer. (1) (a) Presented in part a t the Mfeting-in-Miniatilre. New York Section, 4.C S . , March 11. 1900, paper H-9. (b) J. G. Berger, hf S. Thesis, New York University, June 19GO. (2) L. Friedman and J . 0. Berger, J . A i n . Chein. Soc.. 83, 500 (1961). (3) (a) F. E . Condon and D. E . Smith, ibid., 69, 965 (1947).
(b)
Recently, W. Kirmse, German Chemical Society Local Section Meeting, Mainz-Weisbaden, July 21, 1960; abstracted in Angcza. Chrm., 72, 716 (1900). reported t h a t a,a-dideuteroisobutyl chloride reacts with sodium or potassium but not with lithium t o give methylcyclopropane B Z U u-elimination and isobutylene by concurrent a n d P elimination. These results complement those of t h e present investigation. (c) X O ~ AED D E D I N PROOP.-For a complete account see W. Kirmse and W. van E. Doering, Tetrahedron, 11, 266 (1960). (II
