Sept. 5 , 1955
ACID-CATALYZED TRANSFORMATIONS OF $-SANTONIN
[CONTRIBUTION FROM THE CHEMICAL
LABORATORY OF THE UNIVERSITY OF
4609
CALIFORNIA, BERKELEY]
Acid-catalyzed Transformations of +-Santonin' BY WILLIAMG. DAUBEN, PAUL D. H A N C EAND ~ WILLIAMK. HAYES RECEIVED MARCH19, 1955 When +santonin is treated with 6 N hydrochloric acid, the lactonic hydroxyl group is eliminated and 1-keto-7-hydroxyAssK(lO)-santadienic acid (IV) (+santonic acid) is formed. This acid, upon heating, is transformed into a lactone with concomitant migration of the conjugated diene system and 1-keto -A2~4(10)-santadien-12,7-olide (XI) is obtained. \Vhen &-santonin is treated with 55% sulfuric acid a more extensive rearrangement occurs and the desmotropo-+santoniris which are formed possess structure X V .
The sesquiterpenic acid lactone, $-santonin (C15HtO04), has been studied extensively by Cocker and on the basis of his results structure I was postulated. As a consequence of some recent work in 0
0
tions gives rise to 1-methyl-7-ethylnaphthalene. This latter hydrocarbon also is obtained from 1ket0-7-hydroxy-A~('~)-santenic acidg,lo,ll (III), a
0
compound which is transformed readily into $-sarit~nin.~ #-Santonic acid possesses two ethylenic double bonds, a carbonyl group with an adjacent this L a b ~ r a t o r y ~an ~ *alternate ~~ structure I1 has3 methylene and a hydroxy group. The ultraviolet been assigned to the compound and this formulaspectrum of the acid shows an intense maximum a t tion differs from that originally postulated by the former workers3 in the location of the double bond 234.5 mp ( E 15,000); Cocker and Lipmans interand the free hydroxyl group. Such a reassignment preted this spectrum as t h a t of a disubstituted a,@unsaturated ketone and suggested that the two was found necessary in order to account for many of the transformations in the #-santonin series. olefinic linkages were not in conjugation. Such Recently, using different evidence, Cocker and his structural characteristics were difficult to reconcile co-workers6 have concurred with this new formula- with any compound derived from structure I1 and tion. I n continuing our study of this sesquiter- so attention was directed toward the nature of the penic acid lactone, some of the interesting reactions chromophoric system and the groupings lost in the it undergoes upon treatment with acid have been formation of the new unsaturated linkage in the reinvestigated and interpreted on the basis of the acid. #-Santonin possesses two possible sources of new structure 11. this new olefinic bond, the free hydroxyl a t C7 and IT'hen +santonin is allowed to react with dilute hydrochloric acid a t room t e m p e r a t ~ r e , ~ an. ~acidic the allylic lactonic oxygen function a t CS. The first material, #-santonic acid (C15HnoO4), is formed can be ruled out since i t has been shown that 1acid (111), a comand when the acid treatment is concentrated sulfu- ket0-7-hydroxy-A~(~~)-santenic pound possessing the original Ci-hydroxyl,' is staric acid a t 50°,719,10two isomeric phenolic lactones, (+)-@-desmotropo-$-santoninand (+)-a-isodesmo- ble to acid under conditions which rearrange $tropo-$-santonin (C1&1803), are obtained. Fur- anto on in.^ Thus, it must be concluded that $thermore, when #-santonic acid is treated under santonic acid is formed by conversion of the ?-lacthe latter condition^,^ i t also is transformed into tone of +santonin to an unsaturated acid. T h a t this new double bond is in conjugation with (+) -@-desmotropo-$-santonin. the original unsaturated linkage was established With regard to the structure of $-santonic acid, readily from the spectrum of the acid. It was found i t has been shown that the basic carbon skeleton of that #-santonic acid showed bands in the infrared $-santonin has not undergone rearrangement, (oil mull) a t 1707 cm.-', characteristic of a satusince a tetrahydro-$-santonic acid upon Clemmenrated six-membered ring ketone,12and a t 1605 and sen reduction followed by selenium dehydrogena1570 cm.-', characteristic of a conjugated diene (1) For the preceding paper in this series see THISJ O U R N A L77, , system.I3 A conjugated unsaturated ketone, such 2451 (1955). as was postulated by Cocker and Lipman,8 on the (2) Recipient of the Dow Fellowship in Chemistry, Univ. of Calif, other hand, would be expected to absorb a t 1G63 1953-1954. (3) W. Cocker, B. E. Cross and D. H Hayes, C h e m i s l r y and I n d u s cm.-'.12 The presence of an isolated ketone was fry, 314 (1951),and earlier papers. established further by examination of the ultravio(4) W. G. Dauben and P . D. Hance, THIS J O U R N A L7, 6 , 3352 let spectrum of the 2,4-dinitrophenylhydrazone of (1953). the acid. It was found that this derivative retained (5) W.G. Dauben and P. D. Hance, ibrd., 77,606 (1955). (6) N. M.Chopra, W Cocker and J. T. Edwards, Chemistry a n d I n the characteristic 234.5 r n M band which now must d u s t r y , 1535 (1954); W.Cocker, ibid., 19 (1955). be attributed to a conjugated diene and, in addi(7) W. Cocker, B. E. Cross and C. Lipman, J . Chem. SOC.,959 I
I1
(1949). (8) W.Cofker and C. Lipman, ibrd., 1170 (1949). (9) G.R. Clemo and W.Cocker, ibid., 30 (1946). (10) W. Cocker and C. Lipman, ibid , 533 (1947).
I11
(11) This nomenclature is discussed in reference 4 and is based upon the hypothetical nucleus, santanic acid (a) (12) R N Jones, D. Humphrres. E Packard and K Dobriner, THIS JOURNAL, 11, 86 (1950)
3610
~ V I L L I A XG. DATJREK, P.\TJLD. HANCEAND WILLIAMK. HAYFS
tion, the material displayed a new band at 360 m p , characteristic of a saturated ketone 2,4-dinitrophenylliydrazone.' The chemical and spectral features allow either structure IV or V to be assigned to $-saiitonic acid The choice between these 0 I '\/
0
,+ , 0 1 1
A
dv
\H\
A,/011 I
/\J
COOFI
C00Il
IT.'
\'
two structures rests on the following considerations. First, employing the diene rules of Fieser," the calculated absorption for IV would be 239 m p and for V would be 244 mp as compared to the experimental value of 234.5 mp. Second, T' is an allylic alcohol and should be susceptible to oxidation by M n 0 2 t o yield a conjugated dienone.I5 Although no crystalline material could be isolated, examination of the spectrum of the reaction mixture during the course of the attempted oxidation did not show the presence of any absorption a t 314 mp where such a dienone would be expected to absorb. Thirdly, it has bee11 reported by Cocker and Lipman' that when $-santonic acid is allowed to stand for 7 days a t room temperature with 8% sodium hydroxide, an isomeric compound is formed. I t now has been found that this new compound, which is difficult to isolate in pure form, possesses an absorption maximum a t 330 mM, a value suggestive of a homoannular dienone. (The spectral characteristics of such a system are discussed below.) Such a transformation would be expected of a ,8,y,G,edienone, whereas the system which is present in V would not be expected to migrate under such conditions. On the basis of such evidence it is strongly suggestive that IV best represents the structure of $-santonic acid and this material should be designated l-keto-7hydr~xy-A~+(~~)-santadienic acid. Further evidence for the formation of a A? dienic system in the acid-catalyzed dehydration of the allylic lactone system was obtained by examination of the reaction of l-keto-A4(10''-santadiena product formed from $-santonin 12,5-olide (VI),> tosylate. LYhen VI was allowed to react with dilute hydrochloric acid under the usual conditions, a tri0
0
coorf
01 -I==O ,
\'I
\.I1
enic acid displaying an absorption maximum a t 240 mp ( E 10,150) was formed. Such an absorption is
characteristic of a heteroannular diene14 and from a structure such as VI only the formation of a AS,5(lO)dienic system will fit these spectral requirements (13) L. Dorfman, Chem. Reas., 63, 47 (19.53).
J . 0 , ~ Chein., . 13, 800 (1948). (15) For bibliography of this reaction, see F. Sondheimer, C . Amendolla and G. Rosenkranz, THISJoURx.AL. 76, 5930,5932 (1953). 114) L. F. Fieser, hf. Fieser a n d S. Rajagopalan,
Vol. 77
since a A4(10)f6-system would give rise to a conjugated triene with the original A7-bond and A 2 , 4 ( 1 0system would be a homoannular dienone. Xccordingly, structure VI1 must be assigned to the trienic acid. It is of interest to note that the ultraviolet maximum has shifted from 234 mp in I V to 2-10 inp in VII. Such would be expected since it has been shownI6 that a 1,4-dihydrobenzene system in such a position to a chromophore displays a transannular interaction which induces a bathochromic shift. I t is of further interest to note that the uniqueness of structure VI1 lends further support to the position of the double bond in VI, a feature which was e m ployed to establish the stereochemical relationship between the original hydroxyl group and the side chain of $-santonin as being trans. The formation of the A3 5(10)-dienestructure indicates that the acid-catalyzed elimination is rate controlled for in VI1 either the establishment of a conjugated triene or a dienone would be thermodynamically favored. RIechanistically, the reaction can be viewed as proceeding by protonization of the lactonic carbonyl group of $-santonin and cleavage of the allylic carbon-oxygen bond to give rise to the allylic carbonium ion irIII, Stabilization of the positive charge by hyperconjugation points to structure VIIIb as being of lower energy than T'IIIa. The structure of the product will be C
,A I
4 / O H
I
-+
controlled by the rate of ionization of the hydrogens on the carbon atom adjacent to the carbonium ion. When the santenic acid (IV) is heated to 200", it is transformed into a lactone which upon alkaline saponification does not regenerate the starting acid. The lactone (as an oil mull) showed absorption in the infrared a t 1665 cm.-l (conjugated ketone"), 1635 and IGOO cm.-' (conjugated dieneI2) and a t 1770 c m - l (saturated five membered l a ~ t o n e ' ~ ) . Furthermore, the lactone displayed an absorption at 320 mp ( e 5210) also demonstrating that an isomerization of the unsaturated system had occurred.18 Absorption in this region has been shown to be characteristic for hornoannular dienones. For (16) E. A. Braude, E. R . H . Jones, I;. Sondheirner and J H . T o o good, J . Chem. Soc., GO7 (1919) (17) R . S. Rasmussen and R. R . Hrattain, THISJ O U R K A L , 7 1 , l O i R (1019), and 1. F. Grove and H A . Willis, J . Chenz. S O L ,8 5 7 (1931). (18) Cocker a n d Lipman' have reported t h a t 1-keto-7-hydroxyAS.S['0)-santadienic acid displayed a weak inHection pomt a t 320 rnu i f 200) while t h e material employed in t h e present work i s non-absorbing in this region. T h e absorption noticed by the former workers would suggest t h a t their acid possessed about 4 % of a conjugateil homoannular dienone system.
ACID-CATALYZED TRANSFORMATIONS OF $-SANTONIN
Sept. 5, 1955
4611
example, Wessely’g and Cavil120have prepared compounds such as IX and X by the action of lead tetraacetate on phenols and the spectra of such com-
The demonstration of the presence of a dienone system by chemical means was attempted by reduction of X I with sodium borohydride to the homoannular dienol (XII). It was found, how0 0 ever, that in order to free the reduction product from the intermediate borate ester acid hydrolysis was required. Such reaction conditions brought about an aromatization of X I 1 and the only product isolated was the lactone X I I I . This lactone (in CS2 R z IX R* x solution) showed absorption in the infrared a t 1785 ) at pounds containing two alkyl groups on the conju- cm.-’ (saturated five-membered l a ~ t o n e ’ ~and gated system fall in the range 312-318 mp ( E 4000).21 1875 cm.-’ (characteristic of a 1,2,3,4-tetrasubstiAlso, it is characteristic of such a system to show an tuted benzene ringz4). Furthermore, in the ultraextinction coefficient of the order of 4000. The lac- violet, XI11 possessed a maximum a t 267 m p ( E 295) tone in the +santonin series which could arise by which is characteristic of a tetralin structure.zj lactonization of the free hydroxyl with the car- Although the demonstration of a homoannular diboxyl group of IV and concomitant migration of ene was not possible by this method, the formation the diene system into conjugation with the carbonyl of the benzenoid lactone devoid of a free oxygen function ( ~ ~ ) - clearly demonstrates the presence of a group and thus possess structure XI, l - k e t ~ - A ~ ~ ~ readily rearranged dienol system, and for the first santadien-12,7-olide, would adequately account for time unequivocally establishes the adjacent posithe spectral features observed.22s23 tions of the carbonyl and angular methyl group. OH 0 With a compound such as X I I I , being free of an oxygen function in the aromatic ring, i t should be possible to interrelate +santonin and santonin. +II With the establishment of the structure of 1ket0-7-hydroxy-A~,~(~~)-santadienic acid (IV) , the nature of the desmotropo-$-santonins can be deterXI mined. As discussed earlier, when either $-santonin or the santenic acid I V i s allowed to react a t 50’ with 55% sulfuric acid, phenolic lactones arc XI11 formed and structure XIV has been assigned t o (19) F. Wessely and P. Sinwell, J f o n a l s h . , 81, 1055 (1950), and F. I
~
Wessely and J. Koltan, i b i d . , 84, 291 (1953). 120) G. 11’. K. Cavill, E. R. Cole, P. T . Gilham and D . J. McHugh, J . Chem. S O L .278.5 , (1954). (21) There appears t o h e an appreciable effect on the maxima (as well as on t h e displacement of the maximum of the chromophoric system with substitution) of such compounds by t h e nature of the groups not o n the conjugated system. For example, with t h e unsubstituted compounds ( R i , Rz = HI, I X absorbs a t 292 mp and X a t 306 mp.19990 I n I X , addition of an alkyl group t o t h e chromophoric system produces a bathochromic shift of 10 mp while in X, a value of 6 mp is found. Furthermore, we have examined t h e compound ( b ) , prepared originally by v. Auwers and Keil ( B e ? . , 36, 4207 (1902)),
0
b and have found t h a t the maximum appears a t 299 mu. Of further interest is t h e fact t h a t t h e extinction coefficient of (b) is 19,000, a n effect which must h e attributed t o a strong interaction between t h e carbonyl and t h e dichloromethyl groupings. (22) Recently, a homoannular dienone has been prepared in t h e ursolic acid series ( H e i v . Chim. A c t o , 3 7 , 2173 (195rj) and its spectral characteristics (Amnx 315 mp, c 6300) are in excellent agreement with those of t h e simple models. (23) T h e lower extinction coefficient possessed by t h e homoannular dienones a s compared t o t h e heteroannular type would be expected by comparison of the length of t h e dienone systems in these two types of compounds. In general, “the greater t h e polarity of the molecule in the ground state, t h e greater the intensity of t h e absorption” (F. A. Miller in Gilman’s “Organic Chemistry,” Vol. 111, John Wiley and Sons, Inc., New York, N. Y . , p. 163), and other things being equal, t h e square of t h e length of t h e conjugated system is proportional t o t h e integrated absorption (R.S. Mulliken, J . Chcm. P h y s . , 7 , 364 (1939). R. B. Turner and D . M Voitle, THIS JOURNAL, 73, 1403 (1951)). T h e length of t h e system (evaluated in a scale drawing and using t h e values found in L. Pauling, ”The N a t u r e of t h e Chemical Bond,” Corne11 University Press, Ithaca, N. y . , 1945, p. 164) in a heteroannular dienone such as methyl 1,7-diketo-A‘c’Q)*6-santadienate is 4.5 A. and in t h e homoannular dienone X I is 3.6 A. T h e ratio of t h e square
OH I
OH
Xv
these desmotropo corn pound^.^^^^^^ The basic carbon skeleton as well as the position of the phenolic hydroxyl group was established by C o ~ k e r ~ who ~a~~ showed that when the phenolic lactones were dehydrogenated over palladium, 2,4-dimethyl-6-ethyl1-naphthol was formed. The position of the lactonic fusion, however, was assigned only on the basis of the earlier, incorrect formula for $-santonin. Since acid IV, which is devoid of the hydroxyl a t Cg,can be transformed into the desmotropo-$+antonins, the position of the lactonic fusion warrants re-examination. It follows, however, that if the phenolic lactones do possess structure XIV, the lactonic hydroxyl group is not that originally present in $santonin but must be formed by lactonization of an unsaturated acid. It has been shown by infrared spectra that these of these lengths is 2.25 and the ratio of the integrated ahsorptions (found by weighing paper patterns of t h e extrapolated maxima) f o r these two compounds is 2.18, in excellent agreement with theory. (24) C. W . Young, R. B. DuVall and N . Wright, A n d . Chcm., 2 3 , 709 (1951). (2.5) R . A. Friedel and M. Orchin, “Ultraviolet Spectra of Aromatic Compounds,” J. Wiley and Sons, Inc., New Y o r k , N. Y., 1951. i2fi) W. Cocker, B. E. Cross, A . K . Fateen, C. Lipman. E. R. S t u a r t , W H. Thompson and D. R . A. Whyte. J Chem. Sac., 1781 (1950). (27) 14’. Cocker, A . K. Fateen and C. Lipman, ibid., 929 (1951).
Vol. 77
WILLIAMG. DAUBEN, PAUL D. HANCEAND WILLIAMK. HAYES
4612
lactones are of the saturated y-type (1766 cm.-', CHC1,) and upon this basis structure X V would be equally acceptable. A characteristic feature of structure XIV not possessed by XV, however, is the lactonic fusion with a benzylic hydroxyl, a feature also found in the desmotroposantonins (XVI) . These latter benzylic lactones display a special re-
$63.0" (c 0.80, alc.) (lit.'#* m.p. 172-173", [ a ] l f l+65.4', ~ for the monohydrate). A,naZ. Calcd. for C16H2001: C, 68.16; H, 7.63; neut. equiv., 264.3. Found: C, 68.01; H , 7.76; neut. equiv., 266. The 2,4dinitrophenylhydrazone was prepared by dissolving 0.20 g. of the acid in 5 ml. of glacial acetic acid and while a stream of nitrogen was passed over the surface of the solution, 0.155 g. of 2,4dinitrophenylhydrazine was added. The heating was continued for 5 minutes a t steambath temperature. After cooling and addition of water to turbidity, the product crystallized. The derivative was recrystallized from methanol-water, m.p. 190.0-191.6 (lit.' m.p. 193-194'). Anal. Calcd. for C Q ~ H ~ ~ O C,~ 54.54; N ~ : H, 5.67; N, 12.12. Found: C, 55.09; H, 5.54; N,11.87. 1-Ket~-Az~~(~~)-santadien-lZ ,7-olide (XI) .-1-Keto-7hydro~y-A3~~(~~)-santadienic acid (0.6 9.) was heated, under a n atmosphere of nitrogen, for 10 minutes a t 200". After cooling, the glassy residue was triturated with ether in order to induce crystallization. The crystals were filtered, washed with aqueous sodium bicarbonate solution and water and then recrystallized twice from ethanol. The yellow crpstalline lactone was obtained in 77% vield (0.5 E.)* m.D. 192.8-194.0", [LYY]~'D+113" (c l.od,"dc.) (lit.? i . p . 19i1920. 1 ~ 1 1 9 + ~i i n o ) . ~. Q-( I ,4-Dimethyl-7:hydroxy-5,6,7,8-tetrahydro-6-naphthyl)propionic Acid Lactone (XIII).-An aqueous methanolic solution of 2.0 g. of l-ket0-A2~~(~0)-santadien-l2,7-olide and 0.46 g. of sodium borohydride was allowed to stand six days a t room temperature and then diluted with 6 N hydrochloric acid until a n oily phase separated. The mixture was then concentrated under reduced pressure on a steam-bath until a cloudiness appeared. Acetic acid was added to the hot mixture to redissolve the oil and the concentration continued until a semi-crystalline material was deposited. Thc solid was collected by filtration and recrystaAlized from bcnzene-ligroin; yield 0.4 g., m.p. 133-142 . Further recrl-stallization yielded white needles, m.p. 142-143 An additional quantity of the material can be obtained by further concentration of the acidified reaction mixture folloivcd by chromatography over alumina. l-Ket0-A3,5(~0),7-~atatrienic Acid (VII).-l-Ket0-A~~'~'~7santadien-12,5-olideS(0.1 g.) was powdered and shaken with 6 N hydrochloric acid under nitrogen for 24 hours. The crude solid reaction product mas recrystallized from aqueous methanol as described for IV, yield 0.9 g . , m . p . 130.9132.0", [CL]~'D +29.6" ( C 0.30, a h . ) . Anal. Calcd. for Cl5Hl8O3: C, 73.15; H, 7.37; neut. cquiv., 246. Found: C, 72.75; H, 7.29; neut. equiv., 246. Treatment of this material with joy0sulfuric acid a t 50' gave only a dark, uncharacterizable oil. ( +)-p-Desmotropo-$-santonin (XV).6-Pomdered $santonin (0.50 g . ) was stirred into 6 ml. of 50% sulfuric acid and kept a t 50" For 20 hours. The reaction mixture was poured into water, extracted with chloroform and the organic phase washed with sodium bicarbonate solution and water. Removal of the solvent and crystallization from alcohol gave i . 2 7 g. (58%) of the crude phenolic lactone, m.p. 175-180 . Recrystallization from ethanol-water and ethyl acetate-hexane gave pure ( )-Pdesmotropo-$santonin, m . p . 186.2-188.5', [ c Y ] * ~ D+68.7" ( c 0.95, alc.) (1it.G m.p. 185-186",[ C Y ] ' ~ Dj-68'). Attempted Zinc-Acetic Acid Reduction of ( )-p-Desmotropo-$-santonin.-The lactone (0.20 g.) was refluxed with 0.5 g. of zinc dust and 1ml. of acetic acid for three hours. The acetic acid was evaporated and the residue dissolved in chloroform. Extraction of the solution with sodium bicarbonate solution yielded no acidic fraction. The remaining neutral material gave 0.17 g. of starting lactone, m.p. 184.3186.9". O
activity in that upon treatment with zinc and acetic acid, they are reduced to desmotroposantonous acids (XVII).z8 It has been found that the desmotropo+-santonins are recovered unchanged from a similar zinc and acetic acid treatment. Such evidence strongly suggests that XV represents the structure of the desmotropo-+-santonins and the lactonic hydroxyl in them is that originally present in $santonin a t Ci. The sequence of acid-catalyzed transformations in this series can be outlined (shown below) as being (1) elimination of the allylic lactone to form a A3,5(10)-santadienic acid derivative, 0
0
'I
I1
/Z'/ \,:/OH
/\'/,.,/OH
(//AV/\\/ ;
--+
8
'
\/v\J
I
---f
O.
COOH
O---O
0
OH
'9y-J I
/.2 /\/*I1
-+
I
\
5
v
\
/
I
COOH
1
'k.0
(2) migration of the diene into conjugation to the C1-carbonyl group to form a A2s4(10)-santadienic acid and (3) a double Wagner-Meerwein (dienonephenol) rearrangement with concomitant lactonization to the phenolic lactone.
Experimentalz9 1-Keto-'l-hydro~y-~3~5~~0)-santadienic Acid (IV).-Finely powdered $-santonin (3.0 g.) and 15 tnl. of 6 N hydrochloric acid were shaken for 24 hours in a stoppered flask filled with nitrogen. The flask was chilled and the crude product llization was conducted so as to avoid excessive heating or contact with oxygen. The crude acid was dissolved in the minimum amount of hot ethanol, and after cooling to room temperature water was added until a turbidity was produced. After crystallization was complete a t room temperature, the solution was cooled and the acid filtered. After repetition of this process there was obtained 2.4 g. (80'%) of the pure product, m.p. 173.4-174.7", [(Y]"D (28) J . Simonsen and D. H. R. Barton, "The Terpenes," Vol. 111, Cambridge Univeisity Press, Cambridge, England, 1951, pp. 219-328. (29) All analyses were performed I,y the Microanalytical 1,aboratory u i the College of Chemistry, Uiliver\ity of California, Berkeley. melting points are corrected.
All
+
+
Acknowledgment.-The authors wish to thank Messrs. T. and H. Smith, Ltd., Edinburgh, Scotland, for kindly supplying the +-santonin used in this work. BERKELEY 4, CALIFORNIA
