__
1
,
^ L -d .
____-Lj--^.-_I-
~.~
~
~
~
-
_ I C -
L-.- -
Fig. 2.--9.m.r.
~
~~~
L.-___L--
spectrum of heptalene.
absorption of hydrogen (PtOn, cyclohexane-ethanol, 10-12 min.) to yield cis( ?)-bicyclo[5.5.0]dodecane (mol. wt. (mass spec.), 166))as indicated by identity of its v.p.c. retention time, infrared spectrum, and mass spectral cracking pattern with one of the cis, trans pair of products obtained from bicyclo [5.5.0]dodeca-1 1-ene-l-one6on catalytic hydrogenation, tosylation, lithium aluminum hydride reduction, and (iv) especially by the n.m.r. spectrum that contains proton absorptions only in the “vinylic” region (none in the lower field “aromatic” nor in the higher field “aliphatic” regions) and exhibits the uniquely characteristic spin-spin splitting pattern (Fig. 2) of an AB?Xz system (integrated intensities: A = 1.62, AB2 = 3.86, X2 = 4.14 protons) as expected for heptalene. In dilute carbon tetrachloride solution heptalene reacts instantaneously with bromine to give a yellow hygroscopic precipitate that is largely water-soluble; the similarity of these properties to those of tropenium bromide suggests that the major product is bromoheptalenium bromide. The limited data available allow only tentatiae conclusions to be made about the nature of the heptalene system. Its facile reactions with catalytic hydrogen, oxygen, acids and bromine, and the greater stability of 1-heptaleniurn ion than 1azulenium ion,j indicate that heptalene is a very reactive compound, and both more reactive and more basic than azulene. Preferential protonation a t the I-position6 shows that nucleophilic localization energy is lowest a t this position, and the higher field positions of the n.m.r. absorptions of X I protons that of 4 or B2 protons suggests that electron densities are highest a t the 1, 5 , 6 , and 10 positions. The location of all proton absorptions a t the high field end of the “vinylic” region probably is attributable to the lack of appreciable aromatic ring c~rrent.~ Its ultraviolet absorption maxima occur a t much shorter wave lengths (256, 352 m l ) than those predicted by the simple molecular orbital method for a fully conjugated system, but are in good agreement with those approximated by the weakly-interacting independent-systems method of Simpsons (245, allowed; 387 mp, allowed and forbidden); the long tail through the visible region may arise from a forbidden transition becoming partially allowed by molecular vibrations or twisting, from transitions according to the Franck-Con( 6 ) G Bilchi a n d 0. Jeger, Helo. C h i m . A c l a , 32, 638 (1949); w e are grateful to Professor A. G. Anderson, J r . , fur the 5.(l’-cycloheptenp1)-pentanoic acid used in this synthesis. (7) G. Berthier, B Pullman and J . Baudet. J . Chi7Jt. P h y s . , 49, 6~11 (1952), predicted a pronounced diminution of diamagnetic susceptibility in heptalene (and pentalene), in contrast t o the exaltations forind i n azulene and benzenoid aromatic compounds. (8) W. T. Simpson, J . .A;;z, C h e m . SOL.,7 7 , 6164 (1955); we are indebted to Professor Simpson for this suggestion and f o r t h e calculation of these values.
don principle from a ground state with a single or double potential energy minimum, or from limited resonance splitting of N and/or E states.s Provided these initial deductions are correct, heptalene more resembles a weakly-interacting cyclic polyene than a strongly-interacting aromatic system, presumably because of theoretical reasons or non-planarity . 19) Acknowledgment is gratefully made to t h e C. S. Army Research Ollice for partial support of this work, arid t o t h e Standard Oil Company of California for a fellowship held b y D J. B during a portion of this research.
DEPARTUEITT OF CHEMISTRY UNIVERSITY OF ~YASHISGTON SEATTLE
L ? ,I
HYPJ. DAUBEX,J R . ~
DOMENICK J . BERTELLI~ RECEIVED OCTOBER9, 1961
n‘ASHISGTON
THE CONFIGURATIONS OF LEVOPIMARIC ACID AND a-PHELLANDRENE ; INTERPRETATION OF THEIR ROTATORY DISPERSIONS
Sw: Levopimaric (A6t5(14)-abietadienoic) acid (I) has been assigned the $-configuration a t C-I3 (steroid numbering, C-9) by Klynel from a Comparison of its [MID with that of 2,4-cholestadiene (11) (cf. 589 mp region of Fig. 1). Recently, however, Schuller and Lawrence2 have presented strong chemical Q0TAT?i7’c O’S?ERSION IN ETYONC,
[+’x
CY/?,4-CYOLESTAD~EhE
__
290
~.
3sc
400’
50C.
..
-
~l
ECS
A n‘p.
Fig. 1.
evidence for the a-orientation a t this center. In addition, other facts, such as formation of a common hydrogenation product3 from both I and neoabietic (A7(18)38(14)-13a - abietadienoic) acid4 can be viewed as favoring the a-configuration a t C-13 in 1.5
We have measured the complete Cotton-effect curves of I, 11, and several related compounds, and have observed an antipodal relationship between those of I and I1 (Fig. 1). Although a t first sight this result would seem to confirm the 13P-assignment, we wish to suggest an alternative interpreta(1) W. Klyne, J. Chem Soc., 3072 (1953). (2) W. H Schuller and R. V. Lawrence, Chenzisivy & + I n d u s l r y , 105 (1961); J . A m . Chem. S o c . , 83, 2563 (1961); c.f. also W. I). Lloyd and G . W. Hedrick, J . Ovg. Chem., 26, 2029 (1961). (3) C,f. J. C. W.Chien, J . A m . Chem Soc., 82, 4762 (1960). (4) Proof of configuration: C. Djerassi, R. Riniker a n d B Riniker, ibid.,78, 636.5 (1956). (.?I Epimerization of a n unqtable 8-configuration a t C-13 during hydrogenation would, of course, invalidate this partiriilar argument.
m
4661
COMMUNICATIONS TO THE EDITOR
Nov. 20, 1961
ID-A
m-B
Fig. 2.
COOH FOLDED FORM
H
E X T E N D E D FORM
tion which is compatible with the chemical evidence ( R - D I ENE H E L I X ) (L-DIENE HELIX) for the 13a-configuration. A fundamental analysis6 of the rotatory disperB I-A sion of optically active 1,3-cyclohexadienes has Fig. 3. revealed that the skewness imposed on such diene systems by structural factors constitutes the major data.3 The signal caused by them is shifted -0.17 element qf asymmetry responsible f o r the Cotton p.p.m. downfield from its position in thirteen reeffect, its contribution to optical activity far out- lated compounds, including, e.g., neoabietic acid weighing that of adjacent asymmetric centers. I n ( A 7 ( I 8 ) and pimaric acid Hence, this addition, it has been derived theoretically and dem- shift appears to be due to a specific interaction, onstrated experimentally6 that the direction of the possible only in I, between the methyl protons and Cotton effect so produced depends u p o n the sense of the entire diene, rather than to the presence of the helicity of the diene, a left-handed helix producing a A8(14)bond, isolated or c0njugated.l‘ negative effect. On the basis of this rule, the negaWhile this interpretation is permissive rather tive Cotton effect of I requires that its dienic sys- than decisive, i t removes the otherwise unresolved tem form a left-handed helix. It is apparent from discrepancy between chemical evidence and rotaDreiding models that this requirement is met not tional data. I t also emphasizes the importance of only by the 13P-configuration of I, but also by one making due allowance for conformational in(I-A)’ of two possible conformations with 13a (the fluences not immediately obvious. other one, I-B, predicts a positive Cotton effect). !Are wish to thank Dr. R. V. Lawrence, Naval Since the first of these possibilities is contrary to the Stores Research Laboratory, Olustee, Florida, for available chemical evidence, we feel that the second samples of resin acids, and several colleagues for one with its 13a-configuration is likely to be correct. helpful discussions. An analogous situation obtains in (-)-a-phellanT h e downfield shift3 in compounds with a Ala bond (e.g., palusdrene (111), our only other case (out of eleven com- tric(11) acid) is normal for a double bond a t this distance and hence not pounds studied so far) of a seemingly wrong predic- relevant (cf. J. S . Shoolery and >T. I,Rogers, J . Am. Chem. SOL.,80, tion of the sign of a Cotton effect from the helicity 5121 (1968)). (12) (a) Visiting Scientist, X.I.H., Summer, 1961; (b) Department rule.8 Here, a positine Cotton effect is expected if the isopropyl group has the quasi-equatorial con- of Chemistry, T h e University of Kansas, Lawrence, Kansas. ISSTITUTE OF ALBERTilr.BURGSTAHLER~~ formation 111-A usually assumed to be the preferred NATIONAL ZIFFER ARTHRITIS AXD METABOLIC DISEASES HERMAX one. The observed negative effect suggests that the BETHESDA 14, MARYLAND VLRICHWEISS alternative conformation 111-B with a quasi-axial RECEIVED SEPTEMBER 18, 1961 isopropyl group is in fact correct; in this case the diene system assumes the required left-handed helicity . OPTICAL ACTIVITY IN SKEWED DIENES It seems significant that both in 111-B and in 1 - 4 S i r : the protons of an alkyl group (17-methyl in I, isoIn connection with our interest in dissymmetric propyl in 111) are spatially close to the center of the unsaturated system. This suggests some type of chromophores,l#’ we have considered the optical interaction which stabilizes an otherwise unfavor- activity associated with skewed dienes3 It will condition of the be recognized that butadiene would be optically able c o n f o r m a t i ~ n . A ~ ~ special ~~ protons on C-17 of I is actually suggested by n.m.r. active in any of its nonplanar forms and that comparable dissymmetric diene conformations may (6) A. Moscowitz, E. Charney, U. Weiss and H. Ziffer, J . A m . Chem. be achieved permanently when the conjugated Soc., 83, 4081 (1961). moiety is part of a rigidly fixed ring system, e.g., (7) Models of I-A actually show a n almost planar diene; it is plausible t o assume, however, t h a t in reality this system is more in lumisterol (I).
I-
strongly twisted t o avoid serious crowding of t h e hydrogen atoms a t C-4 and C-5. (8) T h e absolute configuration of I11 appears well established; c.f., inlev alia, the discussion by K. Freudenherg and W. Lwowski, A n n . , 687,213 (19541, and literature quoted there. (9) Cf. also t h e case of isomenthone: (a) C. Djerassi, “Optical Rotatory Dispersion,” McGraw-Hill Book Co., New York, N. Y., 1960, PP. 106, 107; (b) C. Djerassi, el ol., J . A m . Chem. Soc., 83,3334 (1961). (10) T h e positive Cotton effect observed for I1 clearly eliminates a possible “folded” conformation resembling I-A (left-handed helix). I n I 1 t h e absence of a n alkyl group a t C-9 precludes t h e stabilizing interaction assumed for I-A a n d 111-B, a n d hence t h e extended conformation (cf. I-B)with a right-handed helix prevails.
(1) K . Mislow, iif. A. W . Glass, A. Moscowitz and C. Djerassi, J . Am. Chem. Sot., 83, 2771 (1961). ( 2 ) 4.Moxowitz, Tetuaheduon, 13, 48 (1961). (3) While this manuscript was in preparation, the publication of R . Deen a n d H. J. C. Jacobs, Koninkl. N e d e r l . Akadcmie Vart Wetenschappen Amstevdam, 64, 313 (1961), came t o our attention. From both theoretical and experimental considerations (see also R. Deen, Thesis, Leiden, September, 1961), these authors conclude t h a t “ a n asymmetrically situated butadiene system will either in itself or due t o its environment bring about optical activity correlated with its longest wavelength transition.” T h e present work is in agreement with and supports this conclusion.
