Peroxidation and bromination of thujopsene

264 P. C. Traas et al. 1 Peroxidation and bromination of thujopsene

spectre du norcamphre d, -3 exo pris dans les m&mes conditions ne presente pas de massif A 6 = 6,40. Apres sublimation de la cetone, la pureti: a ttk contrdee par C.P.V. ( > 99 %).

3. Norcamphre d,-3 endo

Ce produit a t t t obtenu par dedeuteriation selective du norcamphre d,-3,3. Ce dernier a etC synthetisk par des echanges successifs dans D,O - dioxanne selon la methode de Lown”. D’apres les constantes de vitesses de dedeuteriation en milieu basique du norcamphre d,-3.3’ nous avons choisi les conditions experimentales suivantes : 4 g de norcamphre d,-3,3 sont places dans une solution de soude 0,2 N dans eau-dioxanne 50/50 pendant 1 h 40 mn a 25°C. Le melange reactionnel est extrait au pentane. La phase organique est sechke, concentrke. La sublimation conduit a 2,2 g de norcamphre dI-3-endo soit un rendement de 55%. La purete a ete contrBlte par C.P.V. (> 99 %). L’ktude du spectre de masse de la D.N.P. (F: 129”5) a montrk que le produit contient:

d o : 6,160,/,, d , : 82,10%, d,: 11,72%.

L’ktude R.M.N. effectuee en presence de Eu(dpm), dans les m&mes conditions que prtcCdemment montre la disparition presque totale d’un massif equivalent a 1 proton mais cette fois celui situt 6,25 6 correspondant au proton endo.

B. Methodes cinCtiques

1 . Etude cinttique de la dtcomposition des combinaisons bisulfitiques

Nous avons utilise les methodes prtcedemment decrites’,. La decomposition des combinaisons bisulftiques a Cte effectuee a 25,OO f 0.02”C dans I’eau a pH = 4,000 (AcOH/AcONa).

2. Cinktiques de riduction des cttones par Pion borohydrure

Les conditions experimentales ont ete celles utiliskes par Genesfez , dans le melange eau-dioxanne SO/SO en volume, 0,05 N en soude. La reaction est suivie a 25,OO f 0,05OC par spectrophotomktrie U.V. a une longueur d’onde (285 mp) pour laquelle seule la cetone de depart absorbe. On prepare une solution de cetone environ 0,05 M dans le solvant eau-dioxanne 50/50 (solution A) et une solution de BH,Na 0,7 M dans le m&me solvant (solution B). Cette derniere solution est soigneusement degazte. Les cinetiques sont effectutes en plaCant dans le cuve de reference 1 ml de la solution B et 1 ml de melange eau-dioxanne 50/50, et dans la cuve de mesure 1 ml de solution B et 1 ml de solution A. Les concentrations cinetiques deviennent alors :

[BH,], = 0,35 M [\C=O], = 0,025 M /

Nous sommes places dans des conditions ou la reaction est d’ordre apparent 1. Le temps zero est pris a I’introduction de la cetone dans la cuve de mesure.

C. Calculs

Le lissage des courbes a CtC effectut par la methode des moindres carres. L‘ensemble des calculs a Cte realist sur I’ordinateur I.B.M. 360-40 de 1’U.S.T.L.

” J . W. Lown, Can. J. Chem. 43, 3294 (1965). 2 2 P . Geneste, These de Doctorat es-Sciences, Montpellier, 1967.

P. Geneste et G. Lamaty, Bull. SOC. Chim. 669 (1968).

Introduction

Peroxidation and bromination of thujopsene

P. C. Traas, L. M. van der Linde and H. J. Takken

Naarden International Research Department, Naarden, The Netherlands (Received March 8th, I974 I

Abstract. We have confirmed that treatment of thujopsene (1) with peracetic acid and m-chloroperbenzoic acid occurs exclusively from the side (/I) towards the cyclopropane ring in the molecule, while osmium tetroxide reacts from the a- as well as from the P-side. It was found that reaction of the unisolated intermediate epoxide 2 with acetic acid and m-chlorobenzoic acid yields, besides the thujopsanones (4) and (5), 2~-hydroxy-3~-acetoxythujopsane (7) and 2fi-(m-chlorobenzoyloxy)-3fl-hydroxy thujopsane (3), respectively, and not 2a-acetoxy-3~-hydroxythujopsane (1 1) and 2a-(rn-chlorobenzoyloxy)-3~-hydroxythujopsane (10) in contrast with what was previously reported by other authors. Reaction of thujopsene with hypobromous acid and bromine yields, in both cases, 3-bromothujopsene (21).

The chemistry of the unsaturated tricyclic sesquiterpene thujopsene has evoked a great deal of interest because of its ubiquity in nature and the presence of a reactive vinylcyclo- propane function. Much attention has been given to reactions of the double bond and their regio- and stereo-selectivity. In the present paper we describe the peroxidation with peracetic and m-chloroperbenzoic acid, the fission products of the primarily formed unstable epoxide and the isolation of a keto acid 6 as a new by-product in this peroxidation reaction. Similar reactions have recently been studied by Acharya and Brown’ and Ohlofl’ with essentially different conclusions regarding the regio- and stereo-chemistry.

Because of the analogy of the protonated epoxide with a bromonium ion the addition of hypobromous acid and of bromine to thujopsene was also studied.

A. Stereochemistry of reaction products from thujopsene and peracetic acid

When thujopsene is treated with peracetic acid in methylene chloride the epimeric thujopsanones 4 and 5 (in a ratio of

S. P . Acharya and H . C. Brown, J. Org. Chem. 35, 3874 (1970). * G. Ohloff; H . Strickler, B. Wilhalm, G. Borer and M . Hinder,

Helv. Chim. Acta. 53, 623 (1970).

Recueil, Journal of the Royal Netherlands Chemical Society, 93/9-10. September-October 1974 265

5: I), resulting from the acid catalyzed spontaneous rearrangement of the intermediate epoxide 2, are isolated in 60% yield. Even when the peroxidation reaction is performed under basic conditions such as when using benzonitrilel hydrogen peroxide',' or in a two layer system of m-chloroperbenzoic acid in methylene chloride/aqueous sodium bicarbonate4 we could not isolate the epoxide 2 as such but only the ketones 4 and 5. As a minor by-product the keto acid 6 was isolated in 2% yield. The formation of this compound can be ascribed to Baeyer Villiger oxidation of the ketones 4 and 5 to lactone 8. Further oxidation of this lactone yields the keto acid. The latter can be cyclized by hydrochloric acid to yield the y-lactone 9t in which the cyclopropane ring has been ruptured. The formation of the *major by-product, the hydroxy acetate 7, particulary important in the determination of the stereo- selectivity of the peroxidation, is initiated by protonation of the epoxide and subsequent cis-attack of an acetate ion, followed by acyl migration. To this product of acetolysis we assigned the structure 2fi-hydroxy-3~-acetoxythujopsane *. This assignment is based on the following evidence: i. The appearance of a one proton multiplet at 4.65 pprn and a methyl signal at 1.28 ppm is only consistent with the presence of a secondary acetate and a tertiary hydroxyl group and not with a tertiary acetate and a secondary hydroxyl group. This is further supported by reaction of the hydroxyl group with trichloroacetyl isocyanate6*' which causes a downfield shift of the C(2) methyl group signal from 1.28 to 1.77 ppm and no appreciable change in the position of the proton multiplet at 4.65 ppm. i i . Dehydration of the hydroxy acetate 7 with phosphoryl chloride in pyridine yields the two isomeric unsaturated esters 12 and 13 as expected from a tertiary alcohol. (From an isomeric 3-hydroxy-2-acetoxy compound only one unsaturated acetate can be formed).

mo" OCOCH] Qy+ (p ,COOH

7 4/ 5

Scheme I

t In their report on the elucidation of the structure of thujopsene, the properties but not the structures of compounds 6 and 9 were described by Akiyoshi and Nagahamas.

* The numbering throughout this paper is in accordance with the generic name for thujopsene, 2,4aj,8,8-tetramethyI- l,laa,4,4a,5,6, 7,8-octahydrocyclopropa[d]naphthalene as given in Chemical Abstracts. For convenience, however, the trivial name thujopsene is retained.

G . B. Payne, Tetrahedron 18, 763 (1962). W. K . Anderson and T. Veysoylrc, J. Org. Chem. 38, 2267 (1973). S . Akiyoshi and S . Nayahama, Bull. Chem. SOC. Jap. 30,886 (1957). V. W. Goodletr, Ann. Chem. 37, 431 (1965).

' I . R . Treham and C. Monder, Tetrahedron Letters 1968, 67.

Having resolved the regioselectivity of the epoxide fission# with acetic acid it was of interest to examine the stereochemistry of the acetolysis product in order to deduce the stereochemistry of the initial epoxide i.e. to establish whether the epoxide ring is up (p) or down (a). In order to solve this problem we eliminated water from the hydroxy acetate 7 by means of phosphoryl chloride (Scheme 2).

ococw, ococHl

7 12 13

I OMO ion"

14 15 4 I5

Scheme 2

Two isomeric acetates 12 and 13 were obtained in a ratio of 1 : 1, separated by preparative gas chromatography and easily distinguished by their NMR spectra. Upon hydrolysis, the enol acetate 13 afforded a mixture of the epimeric ketones 4 and 5 in a ratio of 1: 1. Hydrolysis of the other acetate 12 afforded the unsaturated alcohol 15. The /I configuration of the hydroxyl group at C(3) in 15* was assigned by NMR spectroscopy using Eu(FOD), as shift reagent and was based on the following evidence (Fig. 1): i . The shift sequence** HIB > HIaa > H'", H'O. ii. The magnitude of coupling ( J = 10 Hz) of H3' and H40, agreeing with the diaxial relationship between those t w o protons. iii. The strong coupling of H3" with the allylic protons' at C(2') ( J = 2 - 3 Hz), corresponding with a dihedral angle of about 90° between the planes H3"C3Cz and C2C2'H2'.

After this paper was submitted for publication our attention was drawn to another example of a reversed structure assignation based on incorrect interpretation of NMR data'. Our experiments showed that in this case the initially formed adduct of 2-phenyI-2- (2-phenylcyclopropyl)oxirane and m-chlorobenzoic acid can be isolated but is immediately transesterified when treated with aqueous base (Analogous to Scheme 4).

* The alcohol 15 shows spectral properties identical with those reported by Ohloff' and Ifolo for the alcohol obtained by photooxidation of thujopsene.

** Under no circumstances can the shift sequence alone be taken as evidence for the configuration of the hydroxyl group in 15 because the magnitude of the shift is not only dependent on the distance between the hydroxyl group and the proton but also on the angle between the principal axis of the complex and the vector from the metal ion to the proton''.

' J . A. Donnelly, . I . G . Hoey, S . O'Brien and J . O'Grady, J. Chem. SOC. (Perkin I), 1973, 2030. M . Barfield and B. Chakruharti, Chem. Rev. 69, 757 (1969).

l o S . Iro, H . Takeshita and T. Muroi, Tetrahedron Letters 1969, 309 1. B. L. Shapiro, J . R. Hlubucek and L. F. Johnson. J. Am. Chem. SOC. 93, 3281 (1971).

266 P. C. Traas et al. 1 Peroxidation and bromination of thujopsene

These observations are only in accord with an equatorial fl hydroxyl group at C(3) in the steroidal conformation. With the configuration at C(3) in 15 well established it was necessary to solve the configuration at C(2) in the hydroxy acetate 7 in order to deduce the stereochemistry of the initially formed epoxide. In view of the accompanying formation of a large quantity of enol acetate, diagnostic for an axial hydroxyl group trans-related to a proton in the "P"-position''*13, a cis-relationship between the hydroxyl and the acetate group in 7 would be predicted. Moreover, in substrates where at least one unsaturated substituent (phenyl or a l k e n ~ l ' ~ - ' ~ ) is attached to the epoxide ring, opening of the epoxide ring with retention of configuration has been observed. Direct chemical evidence for the cis- configuration in 7 (and 14 respectively) was obtained by an independent synthesis of the cis-diol from thujopsene by oxidation with osmium tetroxide. Although it is well known that osmium tetroxide adds to double bonds in a cis-way, predominantly from the less hindered side' ' . I 8 and all reactions with thujopsene so far studied i.e. epoxidation'.' and hydroboration' occur from the fl-side, we obtained, after reduction of the osmic ester with sodium sulfite, a mixture of both possible cis-diols 14 and 16 in a ratio 1: 1 (Scheme 3).

1 14 16

Scheme 3

Separation of the two isomeric cis-diols 14 and 16 was accomplished by liquid solid chromatography. Comparison of one of the diols obtained from the osmium tetroxide oxidation (Scheme 3) with the diol obtained by saponification of the hydroxy acetate 7 (Scheme 2) clearly showed the identity of these diols (IR, NMR and mixed melting point). The possibility of inversion of configuration during the alkaline saponification of the hydroxyacetate 7 was ruled out by the observation that the reduction of 7 with sodium dihydrobis(2-methoxyethoxy)alurninate gave the same diol. On the basis of these spectral and chemical evidences, the structure 2fl-hydroxy-3fl-acetoxythujopsane 7 was assigned to the acetolysis product of the intermediate epoxide 2 and not structure I 1 with a reversed regio- and stereo-chemistry as was reported by Ohloff*. (cf. Scheme 1)

B. Stereochemistry of reaction products from thujopsene and m-chloroperbenzoic acid

Treatment of thujopsene with m-chloroperbenzoic acid in chloroform yields 3-isothujopsanone 4 [with the two methyl groups and the cyclopropane ring all directed to the same (p) side]. Again no epoxide could be isolated. A small amount of the keto acid 6 is also formed. The major by-product is 2~-(m-chlorobenzoyloxy)-3fl-hydroxythujopsane 3. From the

l 2 H . Sulser, J . R . Scherer and K . L. Stevens, J. Org. Chem. 36,2422

l 3 D. H. R . Burton, A. Da. S . Cameos-Neves and R. C . Cookson,

l 4 J . H . Brewster, J. Am. Chem. SOC. 78, 4061 (1956). I s , D. Y. Curtin, A. Bradley and' Y.\ G. ffendrickson, ibid. 78, 4064

(1956). I' G. Berti, F . Bottari, G. Lippi and B. Macchia, Tetrahedron 24,

1959 (1968). I ' J . March, Advanced Organic Chemistry: Reactions, Mechanisms

and Structure Mc. Graw-Hill Book Company New York 1968 p. 616. C. Djerassi and J . Fishmun, J. Am. Chem. SOC. 77, 4291 (1955).

(1971).

J. Chem. SOC. 1956, 3500.

results of the acetolysis of epoxide 2 one would also predict a cis-relationship between the hydroxy and the m-chlorobenzoyloxy groups in 3. Unexpected difficulty was en- countered in the unsuccessful separation of this product from 3-isothujopsanone. All attempts to isolate the hydroxy m-chlorobenzoate 3 from the reaction mixture by liquid solid chromatography failed and only the isomeric 2fl-hydroxy-3fl-(m-chlorobenzoyloxy)- thujopsane (19), the product of intramolecular transesterification, could be isolated in low yield. Also on standing in thecrude epoxidation mixture (residual m-chloro(per)benzoic acid still present) transesterification from 3 to 19 could be observed. As a further complication 19 is gradually trans- formed into the epimeric ketones 4 and 5 (via the enol) and traces of the allylic alcohol 15. The same reaction sequence is also observed under neutral conditions, but is then far more sluggish (Scheme 4). Because this reaction sequence proceeds also in the absence of any added water, even in carbon tetrachloride solutions in the presence of molecular sieves a 100% conversion is observed, we believe that formation of 4, 5 and 15 does not proceed via dehydration of 19 and subsequent hydrolysis of the two possible unsaturated m-chlorobenzoates (analogous to 12 and 13 Scheme 2). A more reasonable mechanism for the formation of 4,s and 15 proceeds via loss of m-chlorobenzoic acid from 3, which is in equilibrium with its congener 19*. To avoid the isolation of the hydroxy m-chlorobenzoate 3, the crude epoxidation mixture was saponified directly with an aqueous solution of potassium hydroxide in methanol and the diol was isolated by column chromatography. Comparison of this diol with the diol obtained from the peracetic acid oxidation after alkaline hydrolysis (Scheme 2) and that of the osmium tetroxide oxidation (Scheme 3) clearly showed the identity of these diols. The same diol was synthesized by saponification of the transesterified hydroxy m-chlorobenzoate 19 and by reduction with sodium dihydrobis(2-methoxyethoxy)aluminate of 3 and 19, respectively present in the crude peroxidation mixture. Consequently, the structure 2fl-(m-chlorobenzoyloxy)-3fl-hydroxythujopsane was assigned to 3. This means that the epimeric structure 2a-(m-chlorobenzoyloxy)-3~-hydroxythujopsane (lo), as sug- gested by Acharya and Brown', is incorrect (cf. Scheme 1).

4;5 15

Scheme 4

C. Discussion

Thujopsene can adopt either a steroidal or a non-steroidal conformation", arising from its cis-decalin structure. Molec- ular models suggest that approach from the a-side in the steroidal and the fl-side in the non-steroidal conformation are highly unfavourable for attack by a reagent with large steric requirements.

* An analogous ketone formation, however, only under strong basic conditions, was observed from 7-hydroxy-2,lO-dioxatri- cycl0[4,3,1,O~~~]dec-6-yl brosylatel'.

l 9 Dr. I . Ernest (Woodward Research Institute Basel). Private

'O C. Djerassi, Optical Rotation Dispersion Mc. Graw-Hill Book Communication.

Company New York 1960 p. 186.

Recueil, Journal of the Royal Netherlands Chemical Society, 93/9-10, September-October I974 267

i?

Steroidal conformation Non-steroidal conformation of thujopsene of thujopsene

Quantitatively little is known about the equilibrium between these conformations of thujopsene in the ground state. The exclusive formation of 3p-hydroxythujopsane in the hydroboration of thujopsene has been taken as an indication that the steroidal conformation predominates largely’*’’. Because of the low free energy of activation and large steric requirements, epoxidation shows a certain similarity to hydroboration and hence approach of the peracid from the P-side and formation of the P-epoxide is to be expected. The isolation of 3-isothujopsanone 4t, obtained from the initially formed epoxide via a stereoselective hydride and the cis-dihydroxythujopsane derivatives 3 and 7 found by us, substantiate the P-configuration of this epoxide. It appears, that this “rule of fl attack” is infringed in the osmium tetroxide oxidation because of the formation of large amounts of the a, a-diol 16 (Scheme 3) besides the 0, a-diol 14 (molar ratio 1 : 1). The a-reactivity can be explained by the Curtin-Hammett principle24. Epoxidation and especially hydroboration, provide a class of reactions with a low free energy of activation, which is small compared with that for interconversion between the two thujopsene conformers. Product formation is therefore controlled by the ratio of conformers in the ground state. However, in the case of the oxidation with osmium tetroxide the free energy of activation for product formation is large compared with that for interconversion. In accordance with the Curtin-Hammett principle, product formation then is independent of the relative energy levels of the steroidal and non-steroidal conformations and depends only on the relative energy levels of the transition states leading to the products. Because the transition state for the a-attack on the non-steroidal conformation is reached more easily than that for P-attack on the more highly populated steroidal conformation, both stereoisomers may be formed at approximately equal rates. Apparently this case arises when osmium tetroxide adds to the double bond in thujopsene from the a-side in the non-steroidal conformation and from the p-side in the steroidal conformation. The ring opening of the epoxide is of special interest because it features an example of an epoxide in which a cyclopropyl substituent is directly attached to the epoxide ring (Fig. 2).

Fig. 2

t The concomitant formation of some thujopsanone 5 in the treatment of thujopsene with peracetic acid is due lo epimerisation of the initially formed 3-isothujopsanone 4 under the reaction conditions.

“ S . P . Acharya and H . C . Brown, J. Org. Chern. 35, 3874 (1970)

’’ H. B. Henbest and T. T. Wrigley, J. Chem. SOC. 1957, 4596. 2 3 D. N. Kirk and V. Petrov, ibid. l%O, 4657. 2 4 D. Y. Curtin, Record Chem. Prog. Kresge-Hooker Sci. Lib. 15,

ref 14.

1 I 1 (1954).

It has been recognized that ring scission of phenyl substituted epoxides proceeds very often with retention of configura- tion25-28 , especially with organic acids in aprotic solvents of low polarity. On the basis of a certain analogy of the cyclopropane ring with the phenyl group one might predict that the ring scission in 2 occurs with retention of configuration. We believe that the observed formation of cis-ring-fission products from the reaction of the epoxide with acetic as well as with rn-chlorobenzoic acid involves a cleavage of the weaker cyclopropylcarbinyl bond via a six-membered ring process in which bond-breaking, facilitated by the cyclopropane ring, has proceeded to a large extent in the transition state before attack of the anion* (i.e. acetate or m-chlorobenzoate). Furthermore, the steric hindrance by the axial C(5) methylene group, not only dictates the exclusive formation of the a-epoxide but also excludes a common S,2 reaction. Whereas the hydroxyacetate 17 is rapidly trans- formed to the more table*^**^*^^ 7 under the reaction conditions the transesterification of the hydroxy-rn-chlorobenzoate 3 to 19 is far more sluggish (Scheme 4). The slow migration of the m-chlorobenzoyl group in 3 with respect to the acetyl migration in 17 must be ascribed to steric repulsions in the transition state of the former as is revealed by molecular models**. This transesterification involves a shift of an acetyl (or rn-chlorobenzoyl) group from one equatorial position to another. When the transesterification involves a shift from an equatorial to an axial position it has been reported to be

D. Hypobromous acid and bromine additions to thujopsene

The ease with which the protonated epoxide rearranges to thujopsanone and the analogy of this species with a bromonium ion prompted us to investigate the aptitude of the thujopsane bromonium ion 20 to rearrange to 3-bromothujopsene 21.

J 1 20 21

x : &.on

Addition of hypobromous acid, generated in situ from N-bromosuccinimide and water, to thujopsene affords 3-bromothujopsene 21 rather than the common addition product. Besides 21, considerable quantities of oxidized sesquiterpenoids were isolated. Oxidation can be suppressed by lowering the redox potential of the hypobromous acid by performing the reaction under alkaline condition^^^, which increased the yield of 21 from 45 to 60”,:,.

* Ohloff3’ claimed to have isolated the trans-hydroxy acetate from the acetolysis of 2,3-epoxycarane.

** Experimental evidence obtained from model systems3’ de- monstrates that on the basis of electronic effects the rate of transesterification of the m-chlorobenzoate should exceed that of the acetate. ’’ G. Berti, F. Botrari, B. Macchia and F. Macchia, Tetrahedron 21, 3277 (1965).

2 6 G. Berri, B. Macchia and F . Macchia, ibid. 24, 1755 (1968). ’’ G. Belluci, G. Berti, B. Bettoni and F . Macchia, J. Chem. SOC.

(Perkin 2) 1973, 292. H. Riuiere, Bull. SOC. Chim. France 1064, 97.

(1 962). z 9 G. Berti, F . Bottari and B. Macchia, Ann. Chim. Rome 52, I101

30 Idem, Tetrahedron 20, 545 (1964). 3 1 G. Ohloffand W. Giersch, Helv. Chim. Acta. 51, 1328 (1968). ” J. Koskikallio, The Chemistry of Carboxylic acid and Esters,

S . Patai, Editor, Intersience-publishers. London 1969 p. 104. 3 3 F. See/, Grundlagen der analytischen chemie, Verlag Chcmie

GMBH. Weinheim/Bergstr. 1960 p. 333.

268 P. C . Traas et al. Peroxidation and bromination of thujopsene

During the addition of bromine to thujopsene in chloroform a spontaneous evolution of hydrogen bromide is observed and 21 can be isolated in 25 % yield. This low yield is caused by rearrangement of the starting material under the influence of liberated hydrogen bromide34. The yield can be increased to 60% when the reaction is performed in ethyl acetate with acetamide as hydrogen bromide s~avenger '~. Neither from the reaction of thujopsene with bromine nor from its reaction with hypobromous acid were we able to isolate the corresponding adducts. The only product which could be isolated was in both cases the substitution product 2lt together with rearranged starting material. These results can be explained on the basis of the formation of a bridged species 20 in which abstraction of a proton from C(3) is favoured over abstraction from C(2'). The reason that the epoxide, besides rearrangement to ketones 4 and 5, gives rise to formation of some adduct by nucleophilic attack at C(2) while the bromonium ion does not, can be contributed to the larger steric requirements of the latter. Moreover the common trans attack is highly unfavourable because of steric hindrance by the axial C(5) methylene group.

Experimental

NMR spectra were measured on a Varian Associates A-60A instrument, using tetramethylsilane as internal reference; unless stated otherwise, spectra were measured of lo:< v/v carbon tetrachloride solutions. Liquid solid chromatographic separations were performed on a modified ALC 201 Waters instrument equipped with a 2m 1/8" column filled with silica gel Merck 53-74 p. Gas liquid chromatographic separations were performed on a Varian Aerograph 9 G P 3 gas chromatograph, using 20% of OV 17 on Embacel. Melting points were determined on a Reichert Thermopan melting point microscope and are uncorrected.

Peroxidation of thujopsene

A . With peracetic acid

To a stirred suspension of thujopsene (1) (20.4 g; 0.1 mole) and sodium acetate (10.2 g; 0.8 mole) in methylene chloride (50 ml) was added with vigorous stirring a solution of peracetic acid in acetic acid (21.0 g; 0.1 mole) over a period of Ih at 25". After the addition was complete, stirring was continued for Ih at room temperature. The sodium acetate was filtered off and washed with methylene chloride. The filtrate was washed with water until free from acetic acid and dried (MgS04). The solvent was removed under reduced pressure. Pentane (25 ml) was added to the oily residue and the solution was stored overnight in a refrigerator. The crude crystalline material was filtered off and a fractional crystallization from petroleum ether (4G-60") yielded: (i) 0.5 g (2 %) of I-acetyl-4,4,8-trimethyl-8-(carboxymethyl)spiro- [2,5]octane 6. m.p. 165-166" (Lit.' 164165"). NMR (CDCI,): 6 0.71 (3H, s); 1.00 (3H, s); 1.26 (3H, s); 2.04 (IH, d, J = 13 Hz); 2.37 (3H, s); 3.07 (lH, d, J = 13 Hz); 10.27 ppm ( lH, s). (ii) 8.4 g (30 %) of 2fi-hydroxy-3fi-acetoxythujopsane 7 m.p. 108-9" (Lit.' 108-9"). NMR: 6 0.2-0.9 (2H, m); 0.55 (3H, s); 1.00 (3H, s); 1.10 (3H, s); 1.28 (3H, s); 1.99 (3H, s); 4.65 ppm ( lH, X-part of ABX-system, JAX + JBx = 16 Hz). N-(Trichloroacety1)carbamate NMR: 6 0.24.9 (2H, m); 0.61 (3H, s); 1.02 (3H, s); 1.16 (3H, s); 1.77 (3H, s); 2.05 (3H, s); 4.77 (IH, X-part of ABX-system, JAX + JBX = 16 Hz); 8.87 pprn (lH, s). The mother liquor was distilled to give 13.2 g (60%) of a mixture of the thujopsanones 4 and 5' b.p. 110-1 12"/1 mm.

t In all entries NMR analysis of the crude reactior! mixtures were carried out before base washings and distillations in order to exclude misleading information because of dehydrobromination or dehydrations respectively during the work up procedure.

# NMR analysis indicated that aheady before distillation, the reaction mixture consisted of a mixture of the ketones 4 and 5. No effort was made to separate the two epimeric ketones because this had already been achieved by Acharya' and Ohloff'.

34 M . Ito, K . Abe, H. Takeshita and M . Yatagai, Bull. Chem. SOC. Jap. 44, 2168 (1971).

" K. Ziele and H. Meyer, Chem. Ber. 82, 275 (1949).

B. With peracetic acid in acetic acid*

The reaction was carried out as described under A, except that methylene chloride was replaced by acetic acid. The work up procedure was as follows: the crude peroxidation residue was saponified with potassium hydroxide and chromatographed on silica. Gradient elution with pentanelether afforded: (i) 14.3 g (65 %) of a mixture of the thujopsanones 4 and 5. (ii) 6.9 g (29 %) 2fi,3fi-dihydroxythujopsane (14), m.p. 79-80" (Lit.* 78-80"), NMR: 6 0.254.9 (2H, m); 0.55 (3H, s); 1.01 (3H, s); 1.09 (3H, s); 1.37 (3H, s); 3.33 pprn ( lH, X-part of ABX-system, J,, + J, , = 16 Hz).

C. With m-chloroperbenzoic acid

To a solution of 1 (20.4 g; 0.1 mole) in chloroform (75 ml) was added m-chloroperbenzoic acid (23.6 g; 0.11 mole) in chloroform (300 ml) over a period of 25 min. External cooling was applied to keep the temperature between 20 and 23". After addition was complete the reaction mixture was cooled to -200 and left overnight. The m- chloroperbenzoic acid was filtered off and the filtrate was washed with 5 % sodium hydroxide (4 x 50 ml), water (2 x 50 ml) and dried (K,CO,). The solvent was removed to give 25.0 g of an oily residue. Approximate composition according to NMR analysis 70% of 4 and 30% of 3. Distillation of the oil yielded 20.1 g (91.4%) of a mixture of the epimeric ketones 4 and 5 (Approximate composition according to NMR analysis 50% of 4 and 50% of 5) contaminated with a small amount of the ally1 alcohol 15. The crude material from another entry was used for LSC separation and provided: (i) A mixture of the thujopsanones 4 and 5. (ii) Keto acid 6. (iii) 2fi-hydroxy-3fi-m-(Chlorobenzoyloxy)thujopsane 19. NMR: 6 1.39 (3H, s); 4.94 ( lH, X-part of ABX-system, J,, + J,, = 16 Hz); 7.3-8.1 ppm (4H, m).

D. With m-chloroperbenzoic acid in methylene chloridelaqueous

Thujopsene was oxidized by the method as described by Anderson4. Yield: 88 % of 4.

E. With benzonitrilelhydrogen peroxide

Thujopsene was oxidized by the method as described by Payne, and Acharya'. Yield: 60% of 4.

sodium bicarbonate

1,3,3-Trimethyl-2-hydroxy-2(3-oxobutyl)cyclohexaneacetic acid lactone (9)

To a stirred solution of 6 (0.63 g ; 2.5 mmole) in ether (25 ml) was added a catalytic amount of concentrated hydrochloric acid. After 15 min the etheral solution was neutralized and dried with solid potassium carbonate. Evaporation of the solvent yielded 035 g (87.3 %) of crystalline material. An analytical sample was crystallized from petroleum ether, m.p. 54-55" (Lit.' 55-56O). NMR: 6 1.08 (6H, s); 1.26 (3H, s); 1.92 (IH, d, 5 = 1 7 Hz); 2.12 (3H, s); 2.6 (2H, m); 2.64 ppm (IH, d, J = 17 Hz).

2-Methylene-3fi-acetoxy-4afi,8,8-trimethyl- laa,2,3,4,4a,5,6,7,8- decahydrocyclopropald1naphthalene (12) and lacetoxy- thujopsene (13)

To a stirred solution of 7 (2.8 g; 0.01 mole) in pyridine (25 ml) was added phosphoryl chloride (1.7 g; 0.011 mole) at 65-75" over a period of 15 min. After the addition was complete, stirring was continued at 750 for i h . After cooling the reaction mixture was poured on to ice (50 g) and extracted with hexane (3 x 20 ml). The organic layer was washed and dried (MgSO,). The solvent was removed at reduced pressure to give 1.9 g (72.5 %) of a mixture of 12 and 13 as an oily residue (Approximate composition according to NMR analysis 50% of 12 and 50% of 13). Separation by GLC yielded both unsaturated acetates.

* This experiment shows that the special function of an apolar solvent cage as claimed by Brewster14 is not essential for the stereochemical course of the solvolysis reaction.

Recueil, Journal of the Royal Netherlands Chemical Society. 93/9-10, September-October 1974 269

NMR: 12 6 0.3-1.0 (2H, m); 0.62 (3H, s); 1.05 (3H, s): 1.13 (3H, s); 2.00 (3H, s); 4.85 ( 1 H, broad s); 4.95 ( 1 H, broad s); 5.38 ppm (1 H, m, W i h = 25 Hz). 13 6 0.5-1.0 (2H, m); 0.62 (3H, s); 1.10 (3H, s); 1.15 (3H, s); 1.65 (3H, s); 2.01 pprn (3H, s).

2B, 38-Dihydroxythujopsane (14)

(i) from 7

A mixture of 7 (2.8 g; 0.01 mole), potassium hydroxide (1.12 g; 0.02 mole) in aqueous ethanol was refluxed for lh. The reaction mixture was cooled and poured into water (100 ml) and extracted with ether (3 x 100 ml). The combined extracts were washed and dried (MgSO,). Evaporation of the solvent yielded 2.2 g of a crystalline product. Crystallization from pentane yielded 2.0 g of 14 (84 7 0 ; m.p. 78-80",

( i i ) from the crude m-chloroperbenzoic acid peroxidation mixture

A mixture of the crude product obtained from the peroxidation of I (2.04 g; 0.01 mole), potassium hydroxide (2.8 g; 0.05 mole) in ethanol (30 ml) was refluxed for lh. The reaction mixture was cooled and poured into water (150 ml) and extracted with ether (2 x 50 ml). The combined organic layers were washed and dried (K,CO,). Evaporation of the solvent yielded an oily residue which was chromatographed on silica. Gradient elution with pentane/ ether afforded: ( i ) 1.3 g (6076) of a mixture of the thujopsanones 4 and 5. (ii) 0.08 g (3 7;) of a compound to which we tentatively assigned the structure 2-(acetylhydroxymethylene)4a~,8,8-trimethyl-1,1aa,2,3,4, 4a,5,6,7,8-decahydrocyclopropa[d]naphthalene: m.p. 75-77O. NMR: 6 0.24.9 (2H, m); 0.61 (3H, s); 1.03 (3H, s); 1.12 (3H, s); 2.22 (3H, s, disappears with NaOD/D,O in CH,OD); 3.90 ppm (IH, s, disappears with D,O). (iii) 0.6 g (25 y:,) of 14.

2-Methylene-3~-hydroxy4a/?,8,8-trimethyl-I, la1,2,3,4,4a,5,6,7,8- decahydrocyclopropa[d]naphthalene (15)

Compound 15 was obtained from 12 using the same saponification procedure as described for the synthesis of 14 from 7. Yield: 85"{,; m.p. 85-86" (Lit.' 84-86). NMR: 6 0.3-1.0 (2H, m); 0.60 (3H, s); 1.04 (3H. s); 1.12 (3H, s); 4.13 ( 1 H, m, W ih = 23 Hz); 4.95 ( I H, broad s); 5.10 ppm (1 H, broad s).

3-Thujopsanone (4) and (5)

The thujopsanones 4 and 5 were obtained from 13 using the same saponification procedure as described for the synthesis of 14 liom 7. Yield: 84?,; b.p. 110 112"/1 mm.

2j,3B-Dihydroxythujopsane (14) and 2c(, 3a-dihydroxythujopsane (16)

A solution of 1 (0.4 g; 2 mmole), osmium tetroxide (0.56 g; 2.2 minole) and pyridine (0.6 ml) in ether (50 ml) was stirred at ambient tempe-

rature for two weeks. To the solution was added a saturated aqueous solution of sodium sulfite (150 ml) and refluxed for 2h. The aqueous layer was extracted with ether (2 x 25 ml). The combined organic layers were washed and dried (MgSO,) and the solvent removed at reduced pressure. The crude crystalline mixture of the diols 14 and 16 was separated by LSC to give: ( i ) 0.19 g (40%) of 14. ( i i ) 0.21 g (44%) of 16. NMR: 6 0.14.2 (IH, m); 0.34.7 (IH, m); 0.77 (3H, s); 0.83 (3H, s); 1.03 (3H, s); 1.29 (3H, s); 3.17 ppm ( lH, X-part of ABX-system).

3-Bromothujopsene (21)

A. Hypobromous acid addition to 1

A suspension of 1 (20.4 g; 0.1 mole) and N-bromosuccinimide ( I 9.6 g; 0.11 mole) in 2N-aqueous solution of potassium hydroxide (100 ml) was vigorously stirred for 2h. External cooling was applied to keep the temperature below 30". The organic layer was separated and the aqueous layer was extracted with ether (50 ml). The combined extracts were washed, dried (MgSO,) and concentrated. The oily residue was distilled to give 16.6 g (59 04) of 21, which slowly crystallized from ethanol, m.p. 39-40", NMR: 6 0.5-1.0 (2H, m); 0.61 (3H, s); 1.09 (3H, s); 1.13 (3H, s); 1.93 ppm (3H, s).

B. Bromination of I Bromine (16.0 g ; 0.1 mole) was slowly added to a solution of 1 (20.4 g; 0.1 mole) and acetamide (18.0 g; 0.4 mole) in ethyl acetate (300 ml) over a period of Ih. at 200. As soon as the addition was complete the solution was filtered and the filtrate was washed with a saturated aqueous solution of sodium bicarbonate (100 ml), water (100 ml), dried and concentrated. Distillation afforded 17.5 g (62 7;) of 21.

Acknowledgements

We are greatly indebted to Prof. Dr. Th. J. de Boer, Labo- ratory for Organic Chemistry, University of Amsterdam and to Mr. H . Boelens for their stimulating interest and helpful discussiors and to Drs. H. J. W o b b r n for Liquid Solid Chromatographic separations.

Note added in press

A "looser" transition state for osmylation than for epoxidation has been put forward by Berti to explain differences i n sensitivity to repulsive interaction and stereocliemical out- come. [G. Berti in: Topics in Stereochemistry 7. 12.3 (1973).]

Documents

Peroxidation and bromination of thujopsene