S. WINSTEINAND
3244
JOSEPH
1850 A. it was necessary to use thin cells (0.1-cm. and 0.01cm. path length), a maximum phototube voltage (setting 4 or 5 ) , and a slit control of 25. I n the far ultraviolet region a scanning speed of 0.5 A.,'sec. with a chart speed of 2 in./
[CONTRIBUTION FROM
THE
SONNENBERG
Vol. 83
min. was use$; in the near ultraviolet region a scanning speed of 2.5 A./sec. with a chart speed of 2 in./min. was employed. Negligikle scattering was found in the instrument down to 1850 A.
DEPARTMENT OF CHEMISTRY, UNIVERSITY
OF
CALIFORNIA, LOS ANGELES24, CALIF.]
Homoconjugation and Homoaromaticity. IV. The Trishomocyclopropenyl Cation. A Homoaromatic Structure*i2 BY S. WISSTEIXAND JOSEPH SONNENBERG RECIXVED J A ~ U A R Y3 , 1961 3-Deuterated-3-bicyclo[3.1.0] hexariols have becm prepared and the corresponding toluenesulfonates acetolyzed in order to check for t h e occurrence of a uniquely syininetrical non-classical cation as a n intermediate in solvolysis of t h e cis-toluenesulfoimte. I n acetolysis of thr. trans-toluenesulfonate very little redistribution of deuterium is visible in the solvolysis product. In acetolysis of the c's-tuluerlesulfonate, however, deuterium is distributed equally over carbon atoms 1, .? and 5 in the product. Further, drutcriuui on the cyclopropane riiig of the initial cis-toluenesulfonate is also symmetrically distributed during acetolysis. The results obtained are uniquely consistent with the intervention of the trishomocyclopropeny~ cation as a n intermediate in svlvolysis of cis-3-bicyclo[3.1.0]hexyl toluenesulfonate. The theoretical relationship between trishomocyclopropenyl and cyclopropenyl cations is discussed, and some of the implications of the present results for organic chemistry are outlined. Regarding the trishomocyclopropenyl cation as the first example of a homoaromatic structure, the authors discuss a generalized concept of homoaromaticity.
The contrasting behavior of the cis- and trans3-bicyclo [3.l.O]hexyl toluenesulfonates as regards anchimeric acceleration, special salt effects, stereochemistry and olefin formation in acetolysis3 was an indication that the cis-epimer does indeed give rise to the symmetrical lion-classical cation 11. We had anticipated the possibility of such a struccation I on theoture for the 3-bicyclo 13.1.[~]liexyl retical grounds. Suitable isotopic labeling of the 3-bicyclohexyl ring system \vas olsviously the way to establish whether solvolysis does indeed involve an intermediate such as I1 with equivalent carbon atoms 1,3 and 5. I n 6.;paper we report and discuss the results of such a. study which prove that the 3-bicyclo [3.1.0]hexyl cation does indeed have the non-classical structure I I. Icegarding this so-called trishomocyclopropenyl cation as the firstrecognized homoaromatic structure, we go on to propose and discuss a generalized concept of hoiiioaro~naticity.~
I
11
Deuterium Labeling and Kinetic Isotope Factor.-The simplest way to label the bicyclo [3.1.0]hexyl ring system is by substitution of a deuterium atom for hydrogen on the carbinol carbon atom 3 . For this reason, 3-bicyclohexanone was reduced with lithium aluminum deuteride as shown in the Reaction Scheme. The bicyclohexanone was derived from oxidation of a mixture of cis- and truns(1) This research was supported by a grant from t h e Petroleum Research Fund administered by t h e American Chemical Society. Grateful acknowledgment is hereby made t o t h e donors of this fund. (2) T h e main results presented in this manuscript were reported in outline form: (a) in a preliminary Communication, S.Winstein, J. Sonnenberg and L. d e Vries, J. A m . Chem. Soc., 81, 6523 (1959); (b) by S. Winstein a t t h e Welch Foundation Conference on Molecular Structure and Orgenic Reactions, Houston, Tex., November 7-9, 1900. (3) S. Winstein and J. S o n u r u t e r r , iW., 83, 3235 (1961). (4) S. It'iustein, ibid.. 81, c52-1 (1959).
3-bicyclohexanols, designated A-OH and F-OH, respectively. A large portion of the deuterated cis-alcohol B-OH was separated from the contaminating trans-epimer by chromatography on alumina and then purified further by way of the crystalline acid phthalate. Combustion analysis of the deuterated cis-alcohol B-OH showed it to have 9 S5 atom yo excess deuterium or 9S.5yo of theoretical for one deuterium atom per molecule. Saponification of the lion-crystalline portion of the acid phthalate of the deuterated alcohol gave rise to a BO :40 cis-trans mixture of deuterated alcohols. A small amount of deuterated transalcohol (2-013 was isolated by preparative vapor phase chromatography, but this sample was still slightly impure. Rates of acetolysis of the toluenesulfonates of thc deuterated and ordinary cis-3-bicyclo [3.1.0]liexanols €3-OH and A-OH were determined simultaneously a t 50.0°. From these rate runs, summarized inore fully in the Experimental section, a StiIall kinetic isotope effect is indicatetl, (bjr ~ K D ) being equal to 1.05. Deuterium Scrambling and Spectra.--\cetolysis ot the toluenesulfonate B-OTs of the deuterated cis-alcohol B-OHat 50' in acetic acid solvent 0.10 Al,Tin sodium acetate and reduction of the resulting acetate with lithium aluminum hydride in the usual way3 gave rise to n cis-alcohol C-OH. The infrarcd spwtr:t o f the three cis-alcohols ;\-OH, U-OH ant1 C-OH, suinniarized in the Experiniental section, show qualitatively that solvolysis of H-OTs is accompanied by the type of deuterium scrambling expected from a symmetrical non-classical intermediate 11. The substitution of deuterium for hydrogen on the carbinol carbon atom 3 of the cis-3-bicyclo[3.1.O]hexanol causes the appearance of a number of new infrared absorption bands, while others disappear or are shifted. The two most prominent new absorption bands in alcohol B-OH are the C-D stretching absorption a t 2151 cn1.l' and an absorption a t 723 cm-'. These two barids can be
THETRIS-HOMOCYCLOPROPENYL CATION
Aug. 5, 1961
3245
REACTIONSCHEME
HwoHdoH H
F-OH
I
LiAID4
f
H A-OH
(D)Kl
H
I
H
E-OH
assigned to the deuterium on the carbinol carbon atom 3 . The strong absorption band a t 745 cm.-l in alcohol A-OH is absent in B-OH, and it can, therefore, be assigned to hydrogen on the carbinol carbon atom. The infrared spectrum of the product alcohol COH shows all the bands of the deuterated B-OH as well as some additional ones. The 2151 and 723 cm.-1 bands for carbinyl deuterium occur with decreased intensities in C-OH. The additional bands for C-OH not present for B-OH include the carbinyl hydrogen band a t 743 cm.-l and two others, namely, at 226BG and 070 cm.-l. As pointed out below, the latter two are quite clearly to be assigned to tertiary cyclopropyl deuterium. Inspection of the infrared spectrum of cis-3bicyclo [3.1.01hexanol in the 3.2-3.6 ~r.region using lithium fluoride optics reveals three frequencies logically associated with cyclopropyl hydrogen atoms, 3070, 3028 and 3000 c n ~ . - ~ .For the tmmepimer the corresponding frequencies arc a t 300-3, 3032 and 2995 cm.-l. By analogy with the inI'hase and out-of-phase vibrations at crc. 2926 and 2533 crn.-l for CH2 groups and the band a t cn. 2890 cm.-' for tertiary C-H observed in ordinary tinstrained aliphatic compounds,6 one can ascribe the 3070 and 3000 cm.-' frequencies for the cis:!.Soc., 7 6 , 18 ( I 9 34) . (14) S. Winstein. e l nl.. ibid., (a) 70, 838 (lM8); ( b ) 7 0 , 3528 (1948); (c) 78, 4347, 4%54 (19%); ( d ) El, 4309 (1939). (16) S.Winstein, H. hI. Walborsky and IC. Schreiber, i b i L L72, 57!i,j (1950). (16) S. Winstein, Expevieniia Suppl. I I , 137 (1935). (17) ( a ) S. Winetein, M. Shatavsky, C. Xorton and R. B. Woodrrnrd, J . A m . Chem. Soc., 77, 4183 ( l g 5 5 ) ; (b) S. Winstein and h1. Shatavskj-, i h i d . . 78, 592 (10.56). (18) (a) E. Hdckel, Z.P h y s i k , 70, 201 (1031); (b) J. D . Roberts. A. Streitwieser, Jr., and C. M. Regan, J . A,?%.Chem. Soc., 74, 4570 (1952); (c) S. L. Menatt and J . D Roberts, J . O m . Chem., 24, 1336 (1939). (19) (a) R . nreslom and C. Yuan, J . A m . ChLm. Sor., 80, ,XI!) (1958); ( L J I < . Breslow and I I . Hover, ibid., 82, 2644 (1960).
ASD
(c,,I:2,L.
Cyclopropcn yl Trishornoc)-clopropcii!,I Cyclopentadicnitlc H~ t~m ~t a h~o n ~ ~ ? c ~ c ~ ~ ~ ~ ~ ~ ~ l ~ ~ t ~ l i e I l i ~ c 13enzeiic I I esahomobcil Z C I ic Trop\.liuiii IIcl,talioiiiotl~)j,?l i > , i i i i ,./,:\,
sI \’
XI1
XV
XVI
XVII
_-.___
( 2 0 ) ( a ) LV. v. E. Doering and L H. Knox, ibid , 76, 3203 ( I 951) ; ( 1 , ) see \V. v . E. Doering and H. Krauch, AEgew. Chem., 68, 6l;l (196fi) for an excellent review of this aud related mailers.
THETRIS-HOMOCYCLOPROPEKYL CATION
Aug. 5 , 1961
represent one of the “Kekul6” structures for hexaIiomobenzene, and the heptahoniotropyliuni ion XIV could be formulated as the product of ionization of a material with structure XVTI. The question whether species XII-XIV will prove to be truly homoaromatic is one of balance between quantum-mechanical delocalization energy and the compression energy necessary to force classical structures (like the “ISekulC” structures of type XVI) into the same geometry. It is quite clear that quantum-mechanical delocalization energies in homoaromatic cases XII-XIV will be less than in aromatic cases, perhaps ca. 4070 as large, 2 b , 1 3 , 2 1 hut the compression energies may not be large enough to preclude formation of hybrid homoaromatic structures. The balance seems no more discouraging for XII-XIV than for the trishoniocyclopropenyl cation 11, and we have experiments under way in an effort to prepare new homoaromatic compounds. Homoconjugation and Homoaromaticity.-The concept of homoconjugation first arose in considering the behavior of cholesteryl and z-cholesteryl derivatives in solvolytic reactions. 1391d A non-classical intermediate VI was visualized which was termed “homoallyl. ” The latter designation is made clearer with formulas XVIIIa-c which portray a carbonium ion with a p-olefinic group and show explicitly the overlapping atomic p-orbitals q and the on the two olefinic carbon atoms Ca and C cationic carbon atom CI. The idea behind the honioallyl designation is that a methylene group (Cz)is a poor insulator of conjugation if the proper rotational positions about the C1-Cz and C2-C3 bonds are assumed. 115th proper rotational positions, there is very appreciable 1,3-orbital overlap13 of a type intermediate between v and T . Semiempirical molecular orbital c a l c ~ l a t i o n s ~ sug~ gest substantial stabilization from electron delocalization.
XVIIIC
From the above point of view, one may say that the homoallyl cation XVIII is hoinoconjugatively stabilized. In conjugation there is electron delocalization over adjacent carbon atoms. Komoconjugation involves electron delocalization across intervening carbon atoms, a single intervening carbon atom in the case of cation XVIII. Such homoconjugation, involving electron delocalization across intervening carbon atoms, is present, also, (21) (a) C. F. Wilcox, Jr.. S. Winstein and W. G. McMillan, J. A m . Chcm Soc., 82, 5450 (1900); (b) S. Winstein and R . Piccolini, unpublished work.
3249
in the trishomocyclopropenyl cation I1 and the other visualized homoaromatic species XII-XIV. The striking behavior of the cis-3-bicyclo [3.1.0]hexyl system has confirmed our anticipations on theoretical grounds that such homoconjugation could be important in systems possessing no ?r-electrons in the original classical structures. If we consider the stepwise insertion of the methylene groups which convert species like the cyclopropenyl cation V and benzene X to their homoaromatic counterparts I1 and X I I I , the designation bishomocyclopropenylz2is appropriate for cation VIII, and the pseudo-aromatic structure for tropilidene visualized by DoeringZ3 can be termed monohomobenzene. Only by explicit stepwise insertion of methylene groups may we appreciate the full ramifications of all the possible molecules which may represent discrete species. For illustration, monohomocyclopentadienide ion X X is a conceivable anion from proton removal from X I X . The homoaromatic species I1 and XIIXIV, with a methylene group inserted between all alternate CH groups, are unique in having complete equivalence of the classical contributing structures. While a systematic nomenclature system will ultimately be desirable, the designation “perhomoaromatic” may be a useful one to distinguish cases like I1 and XII-XIV from analogs with fewer interposed methylene groups. 2 4
sx
SIX
Experimental cis-Alcohol B-OH.-3-Bicyclo[3.1 .O] hexanone was obtained by oxidation of cis-3-bicyclo [3.1.O] hexano13 containing a small proportion of trans-epimer with chromic trioxide in pyridine.’ Reduction3 of the ketone with lithium aluminum deuteride in ether a t -78” pave rise t o a 5.0-g. quantity of crude deuterated alcohol. From chromatography on alumina3 three main fractions were obtained: (a) 2.9 g. of the essentially pure cis-alcohol (maximum of 4yo of the trans-alcohol); (b) an intermediate fraction which was mostly the cis-alcohol; and (c) the remaining residue which came off with the ether eluent. Each alcohol fraction was converted separately t o acid ~ h t h a l a t e . ~ ”Fraction a yielded 7.1 g. of crude acid phthalate, m.p. 118-121.5’. After two recrystallizations from ether-pentane, 4.9 g. (67%) of product, m.p. 122.5125.5’, was obtained. A second crop of 1.3 g., m.p. 118-123’, was :dso obtained. The combined crude acid phthalates from fractions b and c amounted t o 1.2 g. of material, m.p. 120123.5’. The mother liquors from all the recrystallizations were combined and saved for later use in obtaining the deuterated tram-alcohol. Saponification3 of the 4 9 g. of acid phthalate, m.p. 122.5-125.5”, yielded 1.7 g. (87%) of the deuterated czsalcohol, b.p. 67-68’ (17.0 mm). The center fraction of 1.4 g., b.p. 67.5’ (17.0 mm.), ZZ5.’D 1.4770 was used in all the analytical work. A v.p.c. analysis indicated the presence of only the cis-alcohol, no more than 0.597, of the trans-alcohol being present. Deuterium combustion analysis25 indicated 9.85 atom yo excess deuterium. (22) W. G. Woods, R. A. Carboni and J. D. Roberts, J. A m . Cham. Soc., ‘78, 5653 (1956). (23) W.v. E.Doering, e l a l . , ibid., 78, 5448 (1956). (24) This type of distinction is also desirable in t h e non-cyclic cases, such as those based on allyl. It is necessary t o consider bishomoallyl as well as hnmnallyl. (25) Bi Di J. Kemeth, Urbana, Ill.
3250
S.\ I r ~ .nu JOSEPH ~ ~SOLSENBERG ~ ~ ~ \
Anal. Calcd. for C6HSDO: C, 72.08; H and D , 11.18. Found: C, 72.88; H and D, 11.09. Saponification of the combined lower melting acid phthalates ( 2 . 5 g.) led t o 0.90 g. of alcohol. I V.P.C. analysis indicated contamination b y only 27, OF the trans-isomer, so this material was used in the preparation of the deuterated cis-toluenesulf onate. Conversion of deuterated cis-alcohol t o toluenesulfonate3 led t o a 7Ot7, yield of product, m.p. 51.5-52.0", in addition to a second crop, m.p. 51.0-51.5", obtained from the last mother liquor. Anal. Calcd. for ClaHljD03S:C, 61.63; H a n d I), 6.76. Found: C, 61.79; H and D , 6.52. Acetolysis of B-0Ts.-The toluenesulfonate (xn.1). 51 .O52.0°, 1.25 g., 5.0 mmoles) and 65 ml. of anhydrcius acetic acid, 0.01 X in acetic anhydride and 0.1005 .If in sodium acetate, was heated a t 50.0" for 48 hours in a sealed flask. After the acetolysis products were subjected t o the standard extraction and reduction proccdure,3 a V.P.C. analysis of the undistilled alcohol product indicated the presence of more than 997; of cis-alcohol. Distillation of the residue yielded 0.45 g. of cis-alcohol C-01-1, h.p. 6466' (18.5 mm.). C-Ketone a n d cis-Alcohol D-OH.--Alcohol C-OH (ca. 0.45 g.) in pyridine (5 ml.) was oxidizedl c.ith chromic trioxide (2.3 g.) in pyridine (23 nil.! and worked u p in the usual way. T h e product in ether was anallzed by v.P.c., and this showed the presence of the desired ketone, a relatively large amount of pyridine, and several minor impurities of times. Xo bicyclic alcohol was detected. The ketone iii ether was reduced3 with lithium aluminum hydride a t -7s" to yield 0.35 g. of alcohol product, b.p. 65-66" (16 mtn.). A V.P.C. analysis indicated a 90: 10 ci.
