,4ug. 20, 1938
BROMINATION OF QUADRICYCLOHEPTANEDICARBOXYLIC ACID
4355
[CONTRIBUTION FROM THE DEPARTMENT OF CHEMISTRY, UNIVERSITY OF COLORADO, AND CONTRIBUTION No. 2298 FROM THE GATESAND CRELLIN LABORATORIES OF THE CALIFORNIA INSTITUTE OF TECHNOLOGY ]
Bridged Polycyclic Compounds. VII. The Bromination of Quadricyclo [2,2,1,02,6,0335] heptane-2,3-dicarboxylicAcid1B2 BY STANLEY J. CRISTOLAND ROBERTT. LALONDE RECEIVED FEBRUARY 10, 1958 Quadricyc10[2,2,1,0~~~,0~*~] heptane-2,3-dicarboxylic acid was brominated in the dark to give exo-3-exo(or endo)-5-dibromotricyclo[2,2,1,0%~] heptane-2-endo-3-dicarboxylicacid. Demonstration of the structure of this dibromodicarboxylic heptaneacid was made possible by its conversion to the known y-lactone of exa-3-bromo-endo-5-hydroxytricyclo[2,2,1,0z~e] 2,endo-3-dicarboxylic acid. The y-lactone was converted to other known derivatives. The facile solvolysis of the dibromodicarboxylic acid is discussed in terms of its two possible structures.
In previous work2 i t was demonstrated t h a t quadricyclo[ 2,2, 1,02~603~5]heptane - 2,3 - dicarboxylic acid (I) could be obtained in good yield from its valence-bond tautomer bicycle[ 2,2,1]heptadiene2,3-dicarboxylic acid (11). The formation of the quadricyclenedicarboxylic acid, which is the first compound known with this ring system, has promoted a study of the addition of both symmetrical and unsymmetrical reagents to this compound. In this paper we wish to describe the mode of addition of bromine to I.
COOH
I I1 T o insure that the bromination of I took place only by an ionic mode of addition, the bromination was carried out in the dark in a mixed solvent of ethyl acetate and carbon tetrachloride. The bromination appeared to be exothermic. A 72% yield of a crude dibromide was obtained. This, when recrystallized, melted a t 205". The dibromide reacted immediately a t room temperature with ethanolic silver nitrate, did not react with sodium iodide in acetone, and was soluble in water and in a 5y0 aqueous solution of sodium bicarbonate. I t showed no strong ultraviolet light absorption in the region above 207 m p and gave a negative potassium permanganate test. These results indicated that this dibromide could not be represented by structures A or B, but that isomers of C are reasonable structures.
bonds. Lack of absorption in the region of 12.312.5 p would suggest6 t h a t the dibromide did not have the nortricyclene structure (C). However, it will be demonstrated t h a t this inference is incorrect. The dibromide was observed to react readily in warm water to give a solid melting a t 262" which had the empirical formula CSH7Br04. Evidence that this material was a bromolactonic acid was obtained from the infrared spectrum which showed an absorption band a t 5.54 p, an absorption characteristic of strained y - l a c t o n e ~ , ~and J an absorption band a t 5.89 p , a band which was assigned to the carboxylic acid carbonyl a b ~ o r p t i o n . ~This infrared absorption spectrum was identical in every way with the spectrum of an authentic sample of the y-lactone of exo-3-bromo-endo-5-hydroxytricyclo[ 2,2,1,02~8]heptane-2,endo-3-dicarboxylic acid (IV) which is reported to melt a t 265°.8t9 Absent in both spectra was the characteristic nortricyclene absorption in the region of 12.3-12.5 p.
.
In the infrared absorption spectrum of the dibromide was found a strong absorption band a t 5 . 8 0 ~ . This band is attributable to the carboxylic acid carbonyl absorption. Absence of absorption ~ . ~ indicated bands in the region of 6 . 0 - 6 . 4 0 ~ also that the dibromide did not possess olefinic double (1) This paper was presented a t t h e Sixteenth International Congress of P u r e and Applied Chemistry in Paris, France, in July, 1057. ( 2 ) Previous paper in series: S. J. Cristol and R . L. Snell, THIS J O U R N A L8,0 , 19.50 (1958). ( 3 ) L. J. Bellamy, "The Infra-red Spectra of Complex Molecules," John Wiley a n d Sans, Inc., New York, N. Y.,1954, p. 143. ( 4 ) Reference 3, pp. 32-34. ( 5 ) P. R . Schleyer, Abstracts of t h e 130th Meeting, American Chemical Society, Atlantic C i t y , N. J., September, 1956.
0-
c=o
IV Treatment of the broniolactonic acid with diazomethane gave the corresponding methyl ester, m.p. 120.5". The reported melting point of the ylactone of exo-3-bromo-endo-5-hydroxy-2-methoxycarbonyltricyclo[ 2,2,1,02,6]heptane -endo - 3-carboxylic acid (V) is 123°.8 The infrared spectrum of this ester showed absorption bands a t 5.52 p , an absorption characteristic of a strained y-lactone, and 'an absorption a t 5.76 p, an absorption which is within that region characteristic of ester carbonyl absorptions. (6) (a) J. D. Roberts, E. R . Trumbull, Jr., W. Bennett a n d R . Armstrong, THISJOURNAL,72, 3116 (1950): (b) E. R . Lippincott, ibid., 75, 2001 (1951); (c) C. D. Verh'ooy a n d C . S. Rondestvedt, ibid., 7 7 , 3583 (1955). (7) (a) J. A. Berson, ibid., 76, 4075 (1954); (b) P. Wilder, Jr., a n d A. Winston, ibid., 77, 5598 (1955). (8) K. Alder a n d F. Brochhagen, Be?., 87, 107 (1054). (9) W e have extended t h e endo-exo terminology beyond its original usage in bicyclo[2,2,1]heptane systemslo t o tricyclo[2,2,1,0Z~~lheptane systems b y referring configurations with respect to t h e unsubstituted methano bridge. T h i s system will be extendable t o nortricyclene derivatives with substituents on all three methano bridges if confipurations a r e always referred t o t h e CI-carbon a t o m (same side, e m , opposite side, e n d o ) . I t appears t o us t h a t iise of t h e terms endo and exo in these connotations is better t h a n coining new terms, even though t h e original derivationslo are not precisely applicable. (10) J. Bredt, J. p r n k f . Chem., 121, 153 (1920). (11) L. I. Bellamy, ref. 3, p. 153.
U'heii tlie bruinolactonic acid was hydrogenated However, cnrboxylate-ion participatioii is iiot the accordiiig to the method of Alder and Brochhagen.8 only factor which can be considered for the iacilc O.!J4 niole o f hydrogen was absorbed for each mole loss of bromide i o i i . The loss of broiiiitlc ion iroiii ()E bromolactonic acid. The isolated product was the dibromide might be also accompanied b y carboii the kriowii y-lactone of endo-5-hydroxytricyclo- participation. The enhanced solvolysis rates (11' 1 2 ) 2 ,I ,0L,6]lieptane-2,endo-3-carboxq.lic acid,2,8,12:3-nortricyclyl p-brc ~ i n o b e n z e r i e s u l i c ~ ~ ~cliloridc, ~tc, (I71)> n1.p. 198". The identity of the lactonic acid aiid bromidei6 have been attributccl to this effect. w;ts established by a mixed melting point detertni- Such participation would give rise to tlic res )ii;iiicc iiatioii with an authentic sample and a comparison stabilized iiitcrndiate V1Ia as showii lxlon-. F o r o f the infrared spectrum with that of an authentic such participation it is reasoiiablc to believe that sample. Thus our broniolactoiiic acid iiiay also be the leaving atoiii. or grouI), xiid the participatiilg assigiied structure IV. carbon atoms of the inolecule niust ha\
&
COOCH,
0--
c=o
0
L.
C=0
\. I
(Ji the possible dibi-uiiiidcs which might result from the additioii of bromine to the yuadricyclene derivative I oiily the dibroiiiides shown as structures I I I a or I I I b (C) could be reasonably expected to give the broniolactoiiic acid IV.
COOH
Br
coiiforiiiatioii; i. ( > . , iii structure \ - I ] , atoiiis X , C;,, C?, aiid Cg must be coplanar.17l 5 ?‘bus the s o volysis, with carbon participatioii. of I I I a and subsequerit formation of the bromolactonic acid IV would proceed through the interniediate 1'111. The solvolysis anti :jL1hse(lLieilt forniatioii o f the 1,roinolactonic acicl from I 1111 would l)rocce(l
COOH
11113
Ilia
COOH
!I:O
1t is interesting to iiote that the absorption iii the
VIII
COOh
region of 12.3-12.5 p which is presumably characteristic of tiortricyclene derivatives,6is not found in the three nortricyclene derivatives 111, IV and IT, which are doubly substituted a t Ca and'or are tetrasubstituted nortricyclenes. l 3 ;\t the present 3r COOH COOH I11 b 1x time we are unable to state a t what stage in substitution on riortricyclenes this absorption disappears, but thc trisubstituted lactonic acid V I and various through the intermediate IX. Of the two interdisubstituted derivatives of nortricyclene do show mediates, the formation of I X would be preferred over VI11 since the formation of the interincdiate this absorption. .\ deterinination of the pIi,'s for exo-3,cso(or VI11 would involve the development of a partial critfo) - 3-dibroniotricyclo[ 2 j 2 ,1 ,0?,6~heI)tane-",endo- positive charge on a carbon atom which is atljaceiit :I-dicarboxylic acid, I I I a or IIIb, showed that to a carbonyl carbon atoni. Thus only I I I b wouhl be likely to solvolyze with carboii participation aiiti, />I
