April 5 , 1950
BRIDGED POLYCYCLIC COMPOUNDS [CONTRIBUTION FROM THE
DEPARTMENT OF CHEMISTRY,
1655
UNIVERSITY OF COLORADO]
Bridged Polycyclic Compounds. VIII. The Bromination of Bicyclo [2,2,1] hepta-2,5-diene-2,3-dicarboxylic Acid1*’ BY STANLEY J. CRISTOLAND ROBERTT . LALONDE RECEIVED MAY12, 1958 The bromination of bicyclo~2,2,l]hepta-2,5-diene-2,3-dicarboxylic acid (I) in the dark gave a dibromide. This was hydrogenated t o a saturated dibromide which in turn was transformed into an anhydride, a brornolactonic acid and a methyl hept-5-ene-~ndo-cis-2,3-dicarboxylic acid. The ester. The saturated dibromide was debrominated to give bicycle( 2,2,1] bromolactonic acid was converted t o the known -,-lactone of endo-5-hydroxybicyclo[2,2,1]heptane-endo-cis-2,3-dicarboxylic acid. These transformations suggest that the bromination product is trans-6,B-dibromobicyclo[2,2,1]hept-2-ene-2,3dicarboxylic acid.
In connection with a study of the bromination of give the dibromides which are represented by quadricycloheptane - 2,3 - dicarboxylic acid,’ it structures C and D. (iv) Addition may occur in seemed interesting to investigate the bromination a manner analogous to the addition of bromine to of its valence-bond tautomer, bicyclo[2,2,1Ihepta- bicyclo[2,2, llheptadienes which gives, in part, 2,5-diene-2,3-dicarboxylicacid (I), the reactant cis- and trans-3,5-dibromonortricyclenes. The refrom which quadricycloheptane-2,3-dicarboxylic sulting dibromide from the bromination of I would acid was obtained.3 be E. The bromination of I was carried out in precisely The 242’ dibromide was unsaturated toward the same manner as was the bromination of its potassium permanganate, reacted with silver nitrate isomer, ;.e., in the dark and in a mixed solvent of only at the boiling point of the ethanol solvent, and ethyl acetate and carbon tetrachloride. The was inert toward sodium iodide in acetone. bromination gave a solid, m.p. 242’, (11) in 47% In the infrared absorption spectrum of the diyield. This material had a satisfactory analysis bromide was found an absorption peak a t 5.83 p , for CaHsBr204. which was assigned to the carboxylic acid carbonyl The addition of bromine to I may occur in any absorption,’ and absorption peaks at 6.04 and of the following theoretically possible ways for which 6.20 p. The latter two peaks are in the region of analogies are known. (i) Addition of bromine carbon-carbon double bond absorption .* to the As-double bond would result in the formation The positive test for unsaturation and the apof one or more of the stereoisomeric dibromides pearance of carbon-carbon double bond absorption represented by structure A. (ii) Addition to the in the infrared spectrum is evidence which definitely establishes that the dibromide is not repreR r ,COOH sented by structure E. The elimination of strtlctures B and C can be made from the observation that the characteristic &disubstituted carbon-carbon double bond abA B sorptions in the regions of 6.35 and 14.2 p9 are C absent, as well as from the ultraviolet absorption Br. spectrum of the 242’ dibromide. An absorption maximum at 250 mM, log 6 3.74, demonstrates that the A6-doublebond was brominated rather than the A2-doublebond, since only a carbon-carbon double bond in conjugation with the carboxylic acid carA2 double bond would result in those dibromides bonyl groups would give rise to strong absorption in represented by structure B. Addition of bromineto this region.’O For the general structure A, we may write now the A2-doublebond was considered the least likely as addition of bromine to double bonds which are three specific structures AI-AIII which in addition conjugated with electron-withdrawing groups, such to structure D gives a total of four structures to as carboxylic acid groups, is known to proceed with consider for the dibromide. Of the four probable difficulty compared to the addition of bromine to and L. Kaplan, ibid., 7 6 , 4072 (1954)l; (c) the y-lactone of onfi-7ethylene or to a double bond which is flanked by an bromo - endo - 5 - hydroxybicycIo[2,2,1]heptane- cndo cis - 2,3 - dielectron-donating group.* (iii) Bromination with a carboxylic acid [C. D. v e r Nooy and C. S. Kondestvedt, Jr., i b i d . , 7 7 , 3583 (1955) 1: (d) syn-5,7-dibromobicyclo [2.2,1]-2-heptene from bi normal Wagner-Meerwein rearrangement6 would cyclo[2,2,1 Jheptadiene [S. Winstein and M. Shatavsky. Chemistry b
-
(1) Previous paper in series: S. J. Cristol and R. T. LaLonde, THIS Industry, 56 (1956)l. (6) (a) S. Winstein and M. Shatavsky. ibid., 56 (1956); (b) L. JOURNAL, 80, 4355 (19.58). 78, 2819 Schmerling. L. P. Luvisi and R. W. Welch, Tsrs JOURNAL, (2) This paper was presented a t the Sixteenth International Congre85 of Pure and Applied Chemistry, Paris, France, July, 1957. (1956). (7) L. J. Bellamy, “The Infrared Spectra of Complex Molecules,” (3) S . J. Cristol and R. L. Snell, THIS JOURNAL, 76, 5600 (1954); 80, 1950 (1956). John Wiley and Sons, Inc., New York, N. Y.,1954, p. 143. (8) L. J. Bellamy, ibid., pp. 32-34. (4) J. Hine, “Physical Organic Chemistry,” McOraw-Hill Book Co., Inc., New York, N. Y.,1056, p. 211. (9) P. R. Schleyer, Abstracts of the 130th Meeting of the American Chemical Society, Atlantic City, N. J., September, 1956, p. 29-0. of : (a) the y-lactone of anti-7See for example the formation (5) bromu cndo-5-hydroxy-G-okabicyclo[2,2,l]heptane-en~o-cis-2,9-dicarb(10) (a) E. A. Braude and F. C. Nachod, “Determinations of Or. oxylic acid from 7-oxabicyclo(2,2.1)-5~heptene~exo-tir-2,3-~icarboxylic ganic Structures by Physical Methods,” Academic Press, Inc., New acid [R.B. Woodward and H. Baer, THISJOURNAL, 70, 1 I l i l (1946)l; York, N. Y..1955, pp. 147-148. (b) The corresponding bicyclo[2,2,1]2-heptene-2,3-dicarboxylic acid has Xmax 247 mp, log e 8.98. (b) ryn-2,7-dibromobicyc10[2,2.1 lbeptane from norbornene [ H . Kwart
-
STANLEYJ. CRISTOLAND ROBERT T. LALONDE
1656
AII
structures, A111 was considered the least probable on the basis of preferred exo-addition of reagents to bicyclic olefins." trans-Bromination leading to the dibromide represented by structure AI is the generally observed result for an ionic mode of addition which proceeds through a bromonium-ion intermediate.12 The dibromide which is represented by structure AII would arise by cis-bromination, a result which is uncommon when bromine adds to a double bond by an ionic mode of addition but which is reported to occur for a free-radical bromination of a bicyclic olefin.18 When the 242' dibromide was hydrogenated in ethanol with platinum oxide catalyst, one mole of hydrogen was absorbed per mole of dibromide to give 111, m.p. 193', anal. CQHloBrZOl; I11 gave a negative potassium permanganate unsaturation test and reacted immediately a t room temperature with a 2% ethanolic silver nitrate solution. The hydrogenation of the dibromide I1 should proceed in accordance with the exo-addition rule to give a saturated dibromodicarboxylic acid, 111, in which the carboxylic acid groups are both endo. The validity of the exo-addition rule is well established for analogous hydrogenations of bicyclic 01efins.l~ However, as will be discussed below, the stereochemistry of the hydrogenation is demonstrated in this paper to be in accordance with the ex0 addition rule, and it does not depend upon analogy. Treatment of the 193' solid (111) with acetic anhydride produced a solid, m.p. 139', which gave a satisfactory analysis for an anhydride, CsHsBrzOr, IV. The presence of the anhydride function was confirmed by noting two carbonyl absorption peaks at 5.60 and 5.37 p . These absorptions may be compared to the carbonyl absorptions of succinic anhydride which are found a t 5.61 and 5.36 p.l6 The saturated dibrornide 111, when treated with warm water, gave a solid, m.p. 254', which has a satisfactory analysis for a bromolactonic acid CpHoBrO,, V. The infrared absorption spectrum of this compound exhibited a strained ylactonecarbonyl absorption a t 5.59 pl6 and a carboxylic acid-carbonyl absorption at 5.83 p.'7 Treatment of this dibromide with diazomethane (11) K. Alder and G . Stein, Ann., 616, 185 (1935); 616, 183 (1936).
For further references see S. J. Cristol and R . P. Arganbright, THIS JOURNAL,
79, 6039
(1957).
(12) I. Roberts and G . E. Kimball, ibid.. 69, 947 (1937). (13) J. A. Berson and R . Swidler. ibid., 7 6 , 4060 (1954). (14) Some examples of the preference of cxo-addition of hydrogen, In the presence of a catalyst, to bicyclic olefins are as follows: (a) bicyclo[2,2,l]hept-2-ene-2,3-dicarboxylicacid gave bicyclo[2,2,1]-
heptane-cndo-ci~-2,3-dicarboxylic acid (K.Alder and G. Stein, An#., 636, 183 (19313)l; (b) bicyclo[2,2,1 ]hept-2-ene-2-carhoxylic acid gave bicyclo[2,2,l]heytane-c~ido-2-carboxylic acid [K. Alder and G. Steiu, i6id.l; (c) 2,3-dimeth?lbicyclo[2,2,1]hept-2-enegave cndo-cis-2.3dimethylbicyclo[2,2,l]heptane [E. Deussen. J . P i a k f . Ckcm., 121 114, 113 (1930). and K. Alder and W. Roth, Be?., 87, 161 (194511. (15) L. J . Bellamy, ref. 7, p. 111. (16) J. A. Berson, Tars JOURNAL, 76, 4976 (1954). (17) L. 1. Bellamy, ref. 7, p. 143.
VOl. 81
gave a solid dimethyl ester, C1lHl,BrzOc (VI), which melted at 125'. The bromination of the olefin, bicyclo[ 2,2,1]hept - 5 - ene - endo - cis - 2,3 - dicarboxylic anhydride (VII), when carried out in ethyl acetate, has been reported to give two products.la In addition to exo-cis-5,6-dibromobicyclo[2,2,1] heptane-endo-cis2,3-clicarboxylic anhydride (VIII), a second isomer was isolated by conversion to its methyl ester,
m.p. 119'. From what is known of the course of addition of bromine to bicyclic olefins, it was likely that this ester was either the trans-5,6-dibromoendo-cis-2,3-dicarbomethoxycompound VI or the product of a normal Wagner-Meerwein rearrangement, that is, the exo-5-anti-7-dibromo-exo-cis2,3-dicarbomethoxy compound I X.
I-OOCHa Br
M C O O C H J COOCH3
VI
IX
B&coocH3 COOCH, X
Consideration of the method of preparation suggested that our methyl ester, m.p. 125', was the same trans-5,6-dibromo compound VI or the different rearranged 5,7-dibromo-endo-cis-2,3-dicarbomethoxy compound X. In a mixed melting point determination of our dibromo methyl ester with that obtained by Berson and Swidler,'* a melting point of 90' was obtained. Thus the two dibromo methyl esters are not the same. If that dibromo methyl ester which was obtained by Berson and Swidler is either of the two probable structures which we have written for it, then the proof (see below) that our methyl ester is .the trans-5,6-&bromo methyl ester VI demonstrates that the Berson and Swidler dibromo methyl ester must be represented by the rearranged structure IX. The conversion of the 193' dibromide I11 to a 139' anhydride IV, a 254' bromolactonic acid V, and a 125' methyl ester VI makes it permissible to eliminate all but structure AI as possible structures for the 242' unsaturated dibromide, 11. The elimination of structure AII may be made conclusively, as the hydrogenation of this dibromide would lead to the known exo-cis-5,6-dibromobicyclo(2,2,1)heptane-endo-cis-2,3-dicarboxylic acid, m.p. 188°,19 and the subsequent transformations of this acid would have led to the corresponding known anhydride, m.p. 205-20G0,1a~1Q~20 the corresponding y-lactone of exo-6-bromo-endo-5-hydroxybicyclo(2,2,1)heptane-endo-cis-2,3-dicarboxylicacid, m.p. 157-159' and the corresponding dimethyl ester, m.p. 81.5°.21 (18) We are Indebted to Professor Berson for this material. (18) K. Alder and 0 . Stein, Ann., 604, 247 (1933). (20) R. Kwart and L. Kaplan, TEISJOURNAL, 76, 4078 (1954). (21) J. A. Berson, ibid., 76, 5748 (1954).
BRIDGED POLYCYCLIC COMPOUNDS
April 5, 1959
Elimination of structure A111 may be made on theoretical grounds. Hydrogenation of A111 would lead to XI, whose bromine atoms do not have the appropriate stereochemistry to permit ready lactonization either with participation by the carboxylate anionz2 or with solvolytic assistance by carbon.2a
-
Br t &COOH d
XI11
1657
COOH
COOH d
C
O
O
H
XIV Br dOOH dibromodicarboxylic acid during the course of XI the debromination. Kwart and Kaplan have reStructure D was eliminated on the basis that upon ported the nearly complete conversion of syn-5,7hydrogenation and subsequent lactonization, the dibromobicyclo(2,2,l)heptane to trans-2,3-dibromoin the presence of hydrogen known y-lactone of anti-7-bromo-endo-5-hydroxy- bicy~lo(2~2,l)heptane bicyclo(2,2,1) heptane - endo - cis - 2,3 - dicar- bromide or stannic bromide.a0 However, had a boxylic acid, m.p. 213°,24 would have been pro- similar Wagner-Meerwein rearrangement occurred duced. However, in order to demonstrate with with XIII, it would have led to the exo-dicarboxylic certainty the non-identity of the 242' dibromide acid XV which on debromination would yield the with structure D, its hydrogenation product I11 exo-acid XIV rather that the endo-acid XI1 which was dehalogenated with zinc in ethyl acetate. was in fact obtained. The formation of bicyclo(2,2,l)hept-5-ene-endocis-2,3-dicarboxylic acid (XII) confirms the identity of the 242' dibromide I1 with structure AI. As the melting point of XI1 is in question,26the unsaturated acid which was obtained from the Br xv The work which has been reported thus far constitutes a proof that the 242' dibromide I1 can be represented by structure AI only if the hydrogena+ Zn &COOH tion of this dibromide took place in accordance I I COOH COOH with the ex0 addition rule."," If this were not XI1 111 true, then structure D might be consistent with the debromination was converted by hydrogenation to observations.a1 endo-Hydrogenation of D would bicyclo(2,2,l)heptane - endo - cis - 2,3 - dicarboxylic give the saturated exo-dicarboxylic acid XVI, and acid.2e Treatment of the debrominated product subsequent lactonization would produce the bromowith acetic anhydride gave a second derivative, lactonic acid XVII. In addition, zinc-promoted bicyclo(2,2,l)hept - 2 - ene - endo - cis - 2,3 - di- dehalogenation of XVI, assuming the transformacarboxylic anhydride.26 tion analogous to XI11 would give the endo-acid Had the 242' dibromide been the dibromide XII. represented by structure D, it is conceivable that its hydrogenated product XIII, upon treatment with zinc in ethyl acetate, could have been dehalogenated. Such a 1,3-dehalogenation, where zinc semes as the electron source, has been reported in XVI the conversion of pentaerythrityl tetrabromiden." and 1,l-bis-(bromomethyl)-cyclopropaneg~ to methBr&=o -HLB = o ylenecyclobutane. However, it seems likely that such a debromination, if possible a t all in the COOH Ni COOH bicycloheptane ring system, would lead to the exoXVII XVIII acid XIV rather than the endo-acid XII, assuming the transformations shown in XIII. However, hydrogenolysis of the 254' bromolacIt was also considered possible that the presence tonic acid V over Raney nickel catalyst gave the of zinc bromide might promote isomerization of the known 7-lactone of endo-5-hydroxybicyclo(2,2,1)heptane-endo-cis-2,3-dicarboxylicacid (XIX), map. (22) G . N. Lewis, "Valence and the Structure of Atoms and Molecules," Chemical Catalog Co.. (Reinhold Publ. Corp.), New York, and mixed m.p. with an authentic sample,8s202.5N. Y.,1923, p. 113. 203.5'. The infrared spectra of this sample and (23) A. Streitwieser, Jr., Chem. Revs., 66, 703 (1956). the authentic sample were identical. Had the 254O (24) C. D. Ver Nooy and C. S. Rondestvedt, Jr., ibid.. 77, 3683 bromolactonic acid been that bromolactonic acid (1955). represented by structure XVII, hydrogenolysis (2.5) A 177-17RO m.p. is reported by 0. Diels and K. Alder, Ann., 460,QS (1928). A 193-19.1' m.p. is reported by D. Craig, TEISJoun- would have given the new lactonic acid XVIII NAL. 75, 4889 (1951). rather than XIX. (20) 0. Diels and K. AIder, Ann., 460. 98 (1928).
ir
COOH
(30) R. Kwart and L. Kaplan, THISJOURNAL, 76, 4072 (1964). (27) W. Shand, V. Schomaker and J. R. Flscha, TEIS J O ~ A L , 66, 636 (1944). (31) It may be noted that endo-hydrogenation of the &bromides (28) M. J. Murray and E. H. Stevenson, ibid., 66, 812 (1944). reprwented by formulas AII and AIII would give compounds that (29) 1. Shaker and N. Slobodin. J . Can. Chcm. U.S.S.R., 21, 2006 would not yield bromolactonic acids readily. (1951). (32) K. Alder and C. Stein, Ann., 614, 1 (1934).
1658
STANLEY J. CRISTOL AND ROBERT T. LALONDE
Vol. 81
the resulting solution 18.0 g. (0.111 mole) of bromine in 100 ml. of carbon tetrachloride was added with stirring. The reaction mixture was allowed t o stand at room temperature for 5 hours. This reaction was carried out in a 3-liter brown \ , bottle which was covered with aluminum foil and which was 0-e-0 provided with a magnetic stirrer. The solid which had precipitated out of solution was fil. The conversion of the 254' bromolactonic acid to tered and dried; 15.2 g. of solid (40.3% yield), m . ~ 232238", was recovered. The filtrate was distilled until only the known lactonic acid XIX made it apparent an oil remained. This oil was triturated with 170 ml. of that the hydrogenation of this dibromide had oc- carbon tetrachloride. I n time the tot:tl quantity of oil discurred aceording to the eso addition rule. Thus, solved in the carbon tetrachloritlc. ,\111~wi11gthis solution we have shown that the structure of the 242' di- t o stand ovrrniglit :it room tcniperaturc rcsriltetl in t h e forrn:ition o f 2.24 g . ( ? . I r ;) of :L solid, i i i . 1 , . 2 1 0 bromide, obtained from the bromination of bi- lntion of the solvent frrirn the filtrxtc gavr 20.8 g . of id, cyclo(2,2,l)heptadiene-2,3-dicarboxylic acid ( I ) , i\n infrarrtl :ihsorption sp
