3290
CO%IMVNICATIONSTO THE
EDITOR
c‘ol. 7.5
While 2,3,5-tri-O-benzoyl-~-ribose shows little cupied by the leaving halogen3f4;( 3 ) the starting if any mutarotation in chloroform or aqueous halides and resulting anilines are not isomerized dioxane, methylation studies and comparisons under the reaction condition^^^-^; (4) no reaction between its rotation and those of some closely occurs in the benzene series with halides ( i e . , brorelated substances (both to be published in the m ~ m e s i t y l e n e , ~bromodurene5 ~ and 2-bromo-3near future) appear to justify the tentative con- methylani~ole~), where a hydrogen is not attached clusion that it belongs to the @-D-series. to the position adjacent to that occupied by the Acetylation of 2,3,5-tri-O-benzoyl-~-ribose in leaving halogen. These facts as well as the orienpyridine a t a low temperature afforded in 88% yield tation data for various substituents can be accommodated by an elimination-addition mechanism crystalline 1-0-acetyl-2,3,5-tri-O-benzoyl-~-ribose [m.p. 130-131’ (cor.); [cx]~’D +44.2’ (CHCls)]. involving a t least transitory existence of an elecSimilarly, benzoylation a t low temperature gave trically neutral “benzyne” intermediate (11). in 89% yield D-ribofuranose tetrabenzoate [m.p. 121-122’ (cor.); [ ~ ] * O D +17.0 (CHCb)]; the same substance was also obtained through the benzoylation of D-ribose in pyridine at 100’ although the yield in this case (11%) was low owing to the difficulty of separating the product from P-D-ribopyranose tetrabenzoate which is formed simultaneously. I I1 2,3,5-Tri-0-benzoyl-~-ribose has been used for the LTH3 synthesis of benzyl 0-D-ribofuranpside tribenzoate; it is possible that it may prove of general utility for J. 3. the synthesis of ribofuranosides. NATIONAL INSTITUTEOF ARTHRITIS AND METABOLIC DISEASES NATIONAL INSTITUTES OF HEALTH, PUBLIC HEALTH SERVICE DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE ROBERT K . NESS BETHESDA 14, MARYLAND HEWITT G . FLETCHER, JR. RECEIVED MAY 21, 1953
111
IV
As is evident from the above reaction sequence, a critical test of the proposed mechanism would be afforded by the reaction of chlorobenzene- l-C1* with potassium amide. If a symmetrical intermeREARRANGEMENT IN THE REACTION OF CHLORO- diate such as I1 were involved equal amounts of BENZENE-l-Cl4 WITH POTASSIUM AMIDE’ ani1i11e-l-C’~ (111) and aniline-2-Cl4 would be Sir: formed since C-1 and C-2 become eqfiivalent in 11. No satisfactory explanation has been published We have carried out the reaction of I6 with potasfor the rearrangements which often occur in the am- sium amide in liquid ammonia and obtained a 43% ination of “non-activated” aryl halides with alkali- yield of C14-labeledaniline. The C14in the product metal amides2 The pattern of the rearrangements was found to be distributed almost exactly as preshows a considerable disregard for the influences dicted for intermediate formation of 11. While this governing the usual aromatic substitutions and is experiment is not considered to “prove” the “benwell illustrated by the products obtained from the zyne” mechanism, it strongly indicates formation amination of the methoxy- and trifluoromethyl- of an intermediate in which the 1- and 2-positions halobenzenes. Although the methoxy and trifluor- of the ring are, or can become, equivalent? The omethyl groups orient oppositely in aromatic ni- only alternative is the occurrence of simultaneous tration, 0 - and m-methoxy- and trihoromethyl- rearranging and non-rearranging displacements in halobenzenes with alkali-metal amides yield ex- a ratio of almost exactly one to one. The utility of clusively m-substituted anilines, while the p-iso- intermediates like I1 in accounting for the pattern mers yield mixtures containing roughly equal of rearrangements with substituted halobenzenes amounts of m- and $-substituted a n i l i n e ~ . ~ will be demonstrated in a later paper. Besides the seemingly anomalous influence of An outline of the tracer experiments follows. substituents any mechanism proposed for the reac- The last steps were those developed by Loftfield.8 tion must be in accord with the following observa74, 3011 (4) R. A Benkeser and W. E Buting, THIS JOURNAL, tions : (1) the reactions are very rapid, even with (1952) (5) Unpublished experiments by Mr R. L. Harris. (2) chlorobenzene, in liquid ammonia at -33’; (6) Obtained from Tracerlab, Inc , on allocation from the U. S. the entering amino group has never been found Energy Commission. farther than one carbon away from the position oc- Atomic (7) Other possible symmetrical intermediates which would accom(1) Supported in part by the program of research of the U.S. Atomic Energy Commission. (2) The scope of this type of reaction has been investigated principally by Gilman and Bergstrom and their co-workers. For a review, see J. F. Bunnett and R.E. Zahler, Chem. Rem., IS, 273 (1951). (3) (a) H.Gilman and S. Avakian, THIS JOURNAL, 67, 349 (1945); (b) H.Gilman and R. H. Kyle, ibid.. 70, 3945 (1948); 74, 3027 (1952); (c) R.A. Benkeser and R. G. Severson, ibid., 71,3838 (1949); (d) C. W. Vaughaa, B.S. Thesis, M , I , T . ,1961; ( e ) L. A . Carlsmith, M.S, Thalir, M.T.T., m
a.
modate the C“-tracer experiment and fit the general character of the reaction to a more or less satisfactory degree are:
(8)
R.B. Loftfield, THISJOURNAL, 78, 4707 (1961).
COMMUNICATIONS TO THE EDITOR
July 5, 1953
3291
p>() =z() c1
CH3 H / The presence of the grouping -CH-C-Cin HNOi Hz the unsaturated ring of cedrene is well established.2 Raney Ni The size of that unsaturated ring (ring I, fig. 11) has incorrectly been deduced to be five-membered I by previous investigator^^,^^ on the basis of obOH 0 servations which appeared to us to point clearly to a six-membered ring. W e have confirmed our deduction by examination of the infrared spectrum of the anhydride of the bicyclic C13 diacid, norcedrenedicarboxylic acid (NCDA),6 which proved to be that of a glutaric anhydride, with peaks at 5.57 and 5.67 p, while a succinic anhydride, derived from a five-membered ring I, would have absorption maxima at 5.40 and 5.63 p.6 We have further prepared by the action of phenylmagnesium brom" VI1 ide on dimethyl norcedrenedicarboxylate a diVI VI (as the dibenzenesulfonamide) and VI1 (as chenyl lactone, m.p: 173-173.5', which had its barium carbonate) had respectively 51.8 f 1% and infrared band a t 5.70 1,in confirmation of the six43.1 f 1% of the radioactivity of V (as the semi- membered lactone structure corresponding to a carbazone or 2,4-dinitrophenylhydrazone). . In a glutaric acid. The two carboxyls of NCDA are attached to a the corblank degradation of aniline-l-C14(III),6tg responding percentage activity figures for VI and ring, the size of which we have proved by examinaVI1 were 0.2 f 1 and 96.7 f 1%, respectively. tion of the infrared spectrum of the anhydride of Since in the blank degradation the barium carbon- the monocyclic CII dibasic acid obtained by further ate activity was consistently low, we consider the degradation of NCDAe4 Again this proved to be activity of VI to be the most reliable index of the the anhydride of a glutaric acid and ring I1 is thus established to be five-membered. The posiamount of rearrangement in the amination of I. tion of the gem dimethyl group shown in fig. I1 DEPARTMENT O F CHEMISTRY AND JOHN D. ROBERTS'^ LABORATORY FOR NUCLEAR HOWARD E. SIMMONS, JR. is considered elsewhere.' With respect to ring I11 SCIENCE AND ENGINEERING L. A . CmLsmTn two facts pointed strongly to the arrangement MASSACHUSETTS INSTITUTE OF C. WHEATON VAUGHANindicated in structure 11: The isolation in high TECHNOLOGY, CAMBRIDGE 39, MASS. yield of p-cymene from the catalytic dehydrogenaRECEIVED MARCH12, 1953 tion of cedrene,*.indicating the probable presence (9) Since I was prepared from 111 by Tracerlab, Inc., using the of a potential isopropyl group para to the methyl Sandmeyer reaction, this degradation constitutes a proof of the isotope group of ring I; and the obvious biogenetic relaposition assignment for I. tionship to P-curcumene,of which cedrol is a formal (10) Crellin Laboratory, California Institute of Technology, Pasacyclization product. In any event, this surmise is dena 4, Calif. confirmed by the fact that the dehydroacid obtained from the C11 dibasic a ~ i d ~is9oxidized ~ to THE STRUCTURE OF CEDRENE a,a-dimethylhomotricarballylic acid, an important Sir: result, a possible structural implication of which The tricyclic sesquiterpenes cedrene, ClSH.24, was incorporated, together with previous erroneous and cedrol, C16H260,have been the object, since structural deductions, into the last published their isolation by Walter in 1841,' of a large number proposals for the structure of cedrene.3 The full of chemical investigations2 which have led to the development of the necessarily sketchy outline proposal of no fewer than thirteen different struc- presented here will be given in a subsequent paper.g tures for the tertiary alcohol cedrol and the related CHEMICAL LABORATORIES anhydro compound, cedrene. STORK HARVARD UNIVERSITY GILBERT We have elucidated the structures of these two CAMBRIDGE 38, MASSACHUSETTS sesquiterpenes which can now be represented by I CHANDLER LABORATORY (cedrol) and I1 (cedrene). COLUMBIA UNIVERSITY RONALD BRESLOW "2
OH
114
NEW YORK 27, NEWYORK RECEIVED JUNE10, 1953
(4). P1. A. Plattner, G. W. Kusserow and H. Kliiui, Hcls. Chim. Acta,
as, 1345 (1942). I
I
CHs
CHs
I
I1
(1) Ph. Walter, Ann., 89, 247 (1841). (2) Reviews of the work in this field up to 1947 are found in J. Simonsen and D. H. R. Barton, The Terpenes, Vol. 111, Cambridge University Press, London (1952); see also S. H. Harper, Ann. Repfs. Chcm. SOC.,44, 143 (1948). (3) The last publiihed propwals are by PI.A. Pluttner, Ckimia, 8, 148 (IQ48).
( 4 4 L. Ruzicka, P1. A. Plattner and G. W. Kusserow, ibid., 36, 85 (1942). (5) P1. A. Plattner and H. Klaui, Hels. Chim. A c f a , 36, 1553 (1943). (6) We have established that the position of the diagnostically more important lower wave length band is not affected by the degree of substitution. (7) G. Stork and R. Breslow, Tms J O U R N ~ 16, , 3292 (1953). (8) W. Treibs, Bcr., 68, 1041 (1985). (9) Structures I and I1 for adrol and cedrene were first presented formally in a lecture given by one of YI st Hsrvsrd on Junusry 17, 1868.
