The Total Synthesis of Natural Products. Volume 5

Total Synthesis Of Natural Products, Volume 5 Edited by John Apsimon Copyright © 1983, by John Wiley & Sons, Inc.

THE TOTAL SYNTHESIS OF NATURAL PRODUCTS

The Total Synthesis of Natural Products VOLUME 5

Edited by

John ApSimon Ottawa -Carleton Institute for Research and Graduate Studies in Chemistry and Department of Chemistry Carleton University, Ottawa

A WILEY-INTERSCIENCE PUBLICATION

JOHN WILEY & SONS New York Chichester

Brisbane

Toronto

Singapore

A NOTE TO THE READER

This book has been electronically reproduced tiom digital information stored at John Wiley & Sons,Inc. We are pleased that the use of this new technology will enable us to keep works of enduring scholarly value in print as long as there is a reasonable demand for them. The content of this book is identical to previous printings.

Copyright 0 1983 by John Wiley & Sons, Inc.

All rights reserved. Published simultaneously in Canada.

Reproduction or translation of any part of this work beyond that permitted by Section 107 or 108 of the 1976 United States Copyright Act without the permission of the copyright owner is unlawful. Requests for permission or further information should be addressed to the Permissions Department, John Wiley & Sons, Inc. Libnwy of Congress Cataloging in Publication Data: ApSimon, John. The total synthesis of natural products.

Includes bibliographical references. 1. Chemistry, OrganiGSynthesis. I. Title. QD262.A68 547l.2 72-4075 ISBN 0-471-09808-6 (v. 5 ) Printed in the United States of America 10987654321

Contributors to Volume 5 Samuel L. Graham, Department of Chemistry, University of California, Berkeley Clayton H. Heathcock, Department of Chemistry, University of California, Berkeley Michael C. Pirmng, Department of Chemistry, University of California, Berkeley Frank Plavac, Department of Chemistry, University of California, Berkeley Charles T. White, Department of Chemistry, University of California, Berkeley

Preface The art and science of organic synthesis has come of age. This is nowhere more apparent than in the synthetic efforts reported in the natural products area and summarized in the first four volumes of this series. This present volume describes the synthetic activities reported for a 10-year period only in the sesquiterpene field-evidence enough for the successful efforts of the synthetic organic chemist in recent years. Professor Clayton Heathcock and his colleagues have produced a masterly, timely and important contribution, the breadth of which necessitates a complete volume in the series. The sixth volume in this series is in an advanced stage of preparation and will contain updating chapters on the subject matter included in the first two volumes together with a description of synthetic efforts in the macrolide field. A seventh volume, covering diterpene synthesis, is in preparation. JOHN APSIMON Ottawa, Canada October 1982

vii

Contents The Total Synthesis of Sesquiterpenes, 1970- 1979

1

CLAYTON H. HEATHCOCK SAMUEL L. GRAHAM MICHAEL C. PIRRUNG

FRANKPLAVAC T. WHITE CHARLES Index

543

ix

Total Synthesis Of Natural Products, Volume 5 Edited by John Apsimon Copyright © 1983, by John Wiley & Sons, Inc.

Total Synthesis of Sesquiterpenes, 1970-79 CLAYTON H. HEATHCOCK, SAMUEL L. GRAHAM, MICHAEL C. PIRRUNG, FRANK PLAVAC, AND CHARLES T. WHITE Department of Chemistry, UniversiQ of California, Berkeley, California

1.

2.

Introduction Acyclic Sesquiterpenes A. Farnesol and Farnesene B. Terrestrol, Caparrapidiol, and Caparrapitriol C. Juvenile Hormone D. Sinensals E. Fokienol, Oxonerolidol, and Oxodehydronerolidol F. Gyrindal G . Sesquirosefuran and Longifolin H. Davanafurans I

5 6 6 9 11 17 22 24 25 21

2

Contents I.

3.

4.

Dendrolasin, Neotorreyol, Torreyal, Ipomeamarone, Freelingyne, and Dihydrofreelingyne Monocyclic Sesquiterpenes A. acurcumene, Dehydro-a-curcumenes, Curcuphenol, Xanthorrihizol, Elvirol, Nuciferal, or-Turmerone, Curcumene Ether, and Sydowic Acid B. Sesquichamaenol C. Bisabolenes, Lanceol, and Alantone D. a-Bisabolol, adisabololone, Deodarone, Juvabione, and Epijuvabione E. Deoxytrisporone, Abscisic Acid, and Latia Luciferin F. Caparrapi Oxide, 3P-Bromo-8-epicaparrapi Oxide, Ancistrofuran, Aplysistatin, and a- and @-Snyderols G. Isocaespitol H. Lactaral I. y-Elemene, @-Elemenone,Shyobunone, and Isoshyobunone J. Saussurea Lactone and Temsin K. Vernolepin and Vernomenin L. Pyroangolensolide and Fraxinellone M. Ivangulin, Eriolanin, and Phytuberin N. Hedycaryol, Preisocalamendiol, Acoragermacrone, Costunolide, Dihydrocostunolide, Dihydroisoaristolactone,and Periplanone-B 0. Humulene Bicarbocyclic Sesquiterpenes; Hydronaphthalenes A. Eudesmanes (1) Occidol, Emmotin H, Rishitinol, and Platphyllide (2) a-Cyperone, P-Cyperone, /I-Eudesmol, and 8-Selinene (3) Juneol, 10-Epijuneol, and 4-Epiaubergenone (4) Cuauhtemone (5) P-Agarofuran, Norketoagarofuran, and Evuncifer Ether (6) Rishitin and Glutinsone (7) Occidentalol (8) Santonin, Yomogin, Tuberiferine, Alantolactone, Isotelekin, Dihydrocallitrisin, and Frullanolide B. Cadinanes (1) Aromatic Cadinanes (2) d a d i n e n e , yz-Cadinene, a-Amorphene, Zonarene, and Epizonarene (3) a-Cadinol and Torreyol C. Drimanes (1) Driman-8-01, Driman-8,1l-diol, and Drim-8-en-7-one (2) Confertifolin, Isodrimenin, Cinnamolide, Drimenin, Futronolide, Polygodial, and Warburganal

28

35 35 46 47

55

68

75 81 84 84 90 93 107 109 114 122 124 124 124 129 134 136 137 140 143 149 157 157 161 166 169 169 170

5.

Contents

3

(3) Pallescensin A D. Eremophilanes (1) Valencene, Nootkatone, 7-Epinootkatone, Isonootkatone, and Dihydronootkatone (2) Fukinone and Dehydrofukinone (3) Isopetasol, Epiisopetasol, and Warburgiadone (4) Eremophilone ( 5 ) Furanoeremophilanes (6) Cacolol E. Miscellaneous Hydronaphthalenes (1) Valeranone and Valerane (2) Khusitine and 8-Gorgonene F. Hydronaphthalenes Containing an Additional Cyclopropane Ring Other Bicyclic Sesquiterpenes A. Isolated Rings (1) Taylorine and Hypacrone (2) Cuparene, a-Cuparenone, and 8-Cuparenone (3) Laurene and Aplysin (4) Trichodiene, Norketotrichodiene, 12,13-Epoxytrichothec-9-ene, Trichodermin, and Trichodermol ( 5 ) Debromolaurinterol Acetate B. Bridged Systems (1 1 Camphorenone, Epicamphorenone, a-Santalene, a-Santalol, 8-Santalene, epi-@-Santalene,@-Santalol,and Sesquifenchene (2) a-bans-Bergamotene C. Spirocyclic Systems (1) Spirovetivanes (2) Acoranes (3) Axisonitrile-3 (4) Chamigrenes D. Fused Ring Compounds: 3,6 (1 Bicycloelemene (2) Sirenin and Sesquicarene E. Fused Ring Compounds: 5,s (1 1 Pentalenolactone F. Fused Ring Compounds: 5,6 (1) Hypolepins and Pterosin B (2) Bakkenolide A (3) Oplopanone (4) Picrotoxinin G. Fused rings: 5,7 (1) Guaiazulenes: Bulnesol, a-Bulnesene, Guaiol, Dehydrokessane, and Kessanol

178 180 180 188 192 195 202 212 215 215 218 221 228 228 228 230 235 238 248 249 249 263 264 264 284 306 306 313 313 314 318 318 323 323 325 327 330 333 333

4

6.

7.

Contents (2) Guaianolides: Dihydroarbiglovin and Estafiatin (3) Guaiazulenes with an Additional Cyclopropane Ring: Cyclocolorenone, 4-Epiglobulo1, 4-Epiaromadendrene, and Globulol (4) Pseudoguaianolides: The Ambrosanolide Family; Deoxydarnsin, Darnsin, Ambrosin, Psilostachyin, Stramonin B, Neoambrosin, Parthenin, Hymenin, Hysterin, Damsinic Acid, and Confertin (5) Pseudoguaianolides: The Helenanolide Family; Helanalin, Mexicanin, Linifolin, Bigelovin, Carpesiolin, Aromaticin, and Aromatin (6) Other Hydroazulenenes: Duacene, Daucol, and Carotol (7) Other Hydroazulenes: Velleral, Pyrovellerolactone, and Vellerolactone H. Fused Ring Compounds: 6,7 (1) Himachalenes (2) Perforenone (3) Widdrol I. Fused Ring Compounds: 4,9 (11 Isocaryophyllene Tricarbocyclic and Tetracarbocyclic Sesquiterpenes A. Fused Systems (1) Illudol, Protoilludanols, and Protoilludenes (2) Marasmic Acid and Isomarasmic Acid (3) Hirsutic Acid C, lsohirsutic Acid, Hirsutene, and Coriolin (4) Isocomene B. Bridged Systems (1) Gymnomitrol (2) Copacamphor, Copaborneol, Copaisoborneol, Copacamphene, Cyclocopacamphene, Ylangocamphor, Ylangoborneol, Ylangoisoborneol, Sativene, Cyclosativene, cis-Sativenediol, Helminthosporal, and Sinularene (3) Longifolene, Longicyclene, Longicamphor, and Longiborneol (4) Copaene, Ylangene, and Longipinene (5) isocyanopupukeanenes (6) Patchouli Alcohol and Seychellene (7) Zizaene (tricyclovetivene), Zizanoic Acid, Epizizanoic Acid, and Khusimone (8) a-Cedrene, Cedrol, A2-Cedrene and Cedradiene (9) Quadrone (10) Isolongifolene (1 1) Ishwarone and Ishwarane Sesquiterpene Alkaloids

341 344

347 369 377 381 384 384 388 389 391 391 393 393 394 398 405 41 9 424 424

429 446 452 455 460 473 484 488 492 493 498

Introduction A. Illudinine B. Deoxynupharidine, Castoramine, Deoxynupharamine, and Nupharamine C. Dendrobine References

5 498 500 510 520

1. INTRODUCTION

The first total synthesis of a sesquiterpene was Ruzicka's farnesol synthesis, communicated in 1923.' In Volume 2 of this series, we reviewed the sesquiterpene total syntheses which had been published since that time, up to the middle of 197OS2That review, covering a 47-year period and including about 300 papers, required 361 pages. In the intervening decade since our initial survey of the field there has been a veritable explosion of activity. In this chapter, we review a further 533 papers dealing with the total syntheses of over 260 different sesquiterpenes. We have made an effort to include all papers dealing with sesquiterpene total synthesis which appeared in the literature through the end of 1979. In addition, we have added a few papers which were inadvertently omitted from the first installment of this review, and have included a few which were either published while the review was under preparation during 1980 or were communicated to us in the form of preprints during that time. Although some of the 1970-1979 papers are improved routes to molecules previously prepared by total synthesis, most of them are new. The general organization of the earlier review' has been followed, with some modification. In general, we have grouped the syntheses according to the number of carbon rings:. acyclic, monocyclic, bicyclic, and tri- and tetracyclic. Compounds containing a cyclopropane ring are generally included with the class which would contain the molecule with the cyclopropane ring absent. This arbitrary decision has been made since many of these syntheses are simple extensions of syntheses of a parent with addition of the cyclopropane ring being an additional terminal step. In addition, the review now includes a separate section for sesquiterpene alkaloids.

6

Acycllc Sesqulterpenes

As before, not all relay total syntheses are included. The general rule

of thumb is that a relay synthesis is included only if the Anal product differs in carbon skeleton from the starting material. Thus, conversion of santonin into a germacrane or elemane would be included, but conversion into another eudesmane would not. The core of the review is the flow charts, which outline the syntheses. We have described the syntheses in words, sometimes rather succinctly and sometimes in more detail. We have attempted to point out novel chemistry or unusual synthetic strategy and have sometimes offered a brief critique of the synthesis. One of the most interesting aspects of a field such as sesquiterpene synthesis is comparison of the various strategies which different workers have employed for a given target. Consequently, we have been more verbose in discussing such comparative syntheses in several cases, such as occidentalol, the vetivanes, the acoranes, the pseudoguaianolides, vernolepin, gymnomitrol, and dendrobine. For the purpose of comparing the efficiency of different syntheses, we generally use the criteria of number of steps, overall yield, and the number of isomer separations required in the synthesis.

2. ACYCLIC SESQUITERPENES

A. Farnesol and Funesene Corey and Yamamoto have reported the e.dgant synthesis of trans, mns-farnesol which is outlined in Scheme l.3 The synthesis features a method for stereospecific synthesis of olefins from poxido phosphonium ylides and aldehyde^.^ Thus, the phosphorane derived from salt 2 is treated first with aldehyde 3 at low temperature to give the p-oxido phosphonium salt 4, which is deprotonated and treated with formaldehyde to obtain allylic alcohol 5, uncontaminated by the trans&diastereomer. The allylic hydroxyl is removed by the reduction of the bisulfate ester and the terminal hydroxyl is deprotected to obtain farnesol (7).

Farnesol and Farnesene

7

2

I

+PPh3

HO

h:

IH- [

I. C~H~JN-SO~

2. LiAlH4

6

Scheme 1.

\5

OTHP

\

0-25'

H3ofLA-An '

OH

7

Corey-YamamotoSynthesis of Farnesol

Pitzele, Baran, and Steinman, of Searle Laboratories in Chicago, have studied the alkylation of the dianion of 3-methylcrotonic acid (81, with geranyl bromide (Scheme After addition of the geranyl bromide,

8

10

Scheme 2.

Searle Synthesis of Methyl Farnesate

8

Acyclic Sesquiterpenes

methyl iodide is added to obtain the methyl esters. Isomers 10, 11, and 12 are obtained in a ratio of 2.3:2.1:1.0; methyl farnesate (11) of 89% isomeric purity may be obtained by low pressure chromatography in 26% yield, based on geraniol. 0. P. Vig and co-workers report a synthesis of p-farnesene (17) wherein the dianion of acetoacetic ester is alkylated with geranyl bromide and the resulting p-keto ester transformed into a butadiene unit as shown in Scheme 3.6 It is not quite clear from their paper just what they synthesized, since both geraniol and p-farnesene are depicted as having Z double bonds.

13

14

I . H30+

0 \

IS

\

Ph3P=CH;

/ -

16

IT

Scheme 3.

Vig’s Synthesis of p-Farnesene

Otsuka and his co-workers at Osaka University have reported the most direct sesquiterpene synthesis yet -direct trimerization of isoprene (Scheme 41.’ Several catalysts were found which give a preponderance of the linear trimers 17-19. The best system for production of p-farnesene

Terrestral, Caparrapidiol, and Caparrapitriol

9

19

Scheme 4.

Otsuka's p-Farnesene Synthesis

(17) utilizes [NiCl(r13-C3H5)12-A~(n-CgH,3)3 and t-BuOK. If the reaction is stopped at 30% conversion of the isoprene, p-farnesene comprises 57% of the product. Unfortunately, preparative glpc is required to separate 17 from its isomers.

B. Terrestrol, Caparrapidiol, and Caparrapitriol Terrestrol, (35b2,3-dihydrofarnesol (201, is the marking perfume of the small bumble bee. Caparrapidiol (21) and caparrapitriol (22) are plant sesquiterpenes which contain centers of chirality.

OH 20

21

22

Ahlquist and Stallberg-Stenhagen of the University of Goteborg in Sweden have synthesized both enantiomers of terrestrol by way of the Kolbe electrolysis of homogeranic acid (23) with the enantiomers of monomethyl 3-methylglutarate (24, Scheme 5).* Ester 25 is obtained in 8Yo yield, based on homogeranic acid.

10

Acyclic Sesquiterpews

I

U

C

O

O

H

+-

HOOC-COOMr

23

24 LiAIH,

I

U\ C

\ O

O

rlrctrolyrir

M

a

26

20

H

Ahlquist-Stiillberg-Stenhagen Synthesis of Terrestrol

Scheme 5.

A synthesis of caparrapidiol by 0. P. Vig is summarized in Scheme 6.9 The question of diastereoisomerism in the formation of 21 is not addressed by the authors, who simply state that “...The identity of the synthesized compound was established by comparing its IR and NMR (spectra) with those reported in literature.”

27

13

30

29

31

Scheme 6.

21

Vig’s Carrapidiol Synthesis

Juvenile Hormones

11

Weyerstahl and Gottschalk, at the Technical University of Berlin, have synthesized caparrapitriol as shown in Scheme 7.'' As in the Vig synthesis of caparripidiol, the German group makes no mention of a diastereomeric mixture in the addition of vinyllithium to methyl ketone 35. However, in this case the final trio1 is obtained as a sharp-melting solid (mp 78-79°C) in 9096 yield! Chromatography on starch provides one pure enantiomer of caparripitriol.

32

34

22

35

Scheme 7.

Weyerstahl-Gottschalk Synthesis of Caparrapitriol

C. Juvenile Hormones The Cl,- and C,,-Cecropiu juvenile hormones (36 and 37) (JH), although not sesquiterpenes, are included because their structures are so similar to those of the acyclic sesquiterpenes. Although 37 was not characterized until 1967 and 36 until 1968, a total of IS syntheses had been reported by 1972.

COOMe

36: R = M e

37: R * E t

12

Acyclic Sesquiterpnes

Corey and Yamamoto have utilized the &oxidophosphonium ylide method for the synthesis of both Cl,- and C,,-JH, as shown in Scheme 8.3 Intermediate 40 is converted via aldehyde 41 into tetraene 42, which

38

39

40

Ph3P=CH2

41

I . diimidt

/

OTHP

2. H P +

4 steps

Scheme 8.

uo 43

42

-

-

&&&c0oue 37

Corey-YarnarnotoSynthesis of Juvenile Hormones

is selectively reduced to obtain alcohol 43. This material has previously been converted into C18-JH." The C,,-JH 36 is prepared from 40 along the same lines as are used to convert alcohol 5 into farnesol (see Scheme 1).

Findlay and MacCay at New Brunswick, and Bowers at the Agriculture Research Service in Beltsville have reported full details of stereorandom syntheses of both 36 and 37.Iza Their C,,-JH synthesis had previously been published in preliminary form and was discussed in Volume 2 of this series.'2b The New Brunswick-Beltsville C,,-JH synthesis is essentially the same as the Schering synthesis of C,,-JH.I3

Juvenile Hormones

13

Cochrane and Hanson of the University of Sussex have reported two C,,-JH syntheses.14 Their first, summarized in Scheme 9, is modeled closely on the Julia nerolidol synthesis.” Bromide 46 is obtained as a 3:l

44

46 ( c + t )

45

4 9 (4 isomera)

2 . MsMgI 3. H30+

47

48 ( c + t )

SO ( 4 isomers) I . NBS, H20

NaH

51 ( 8 isomera)

37

Scheme 9.

Juvenile Hormone: Sussex Synthesis A

mixture favoring the unnatural E stereoisomer. The second cyclopropyl carbinol solvolysis (48+49) also produces a bad stereoisomer mixture, giving 594/0 of 3E and 41% of 32 compounds. Analysis at the stage of dienone 50 showed the 22, 24 EZ, and EEstereoisomers to be present in a ratio of 16:43:11:30. A final Horner-Wadsworth-Emmons olefination (5W51) affords a mixture of all eight stereoisomers, of which the natural EEZisomer is less than 10%. The Sussex group also reports a somewhat more stereoselective synthesis (Scheme 10). The starting unsaturated bromide 53 is prepared as a 3:l mixture favoring the

Acyclic Sesquiterpenes

14

53 ( c + t )

52

Scheme 10.

Juvenile Hormone: Sussex Synthesis B

undesired Estereoisomer. The second double bond is introduced by a Wittig reaction, which proceeds in an essentially stereorandom fashion, as expected. The final double bond is introduced by the Corey procedure.16 Analysis of ester 51 showed it to be an approximately equimolar mixture of the four stereoisomers having 2E stereochemistry. The desired isomer comprised 22% of the mixture. A Zoecon group headed by C. A. Henrick has prepared the C,,-JH from trans-geranylacetone (56) as shown in Scheme 11. This substance is converted into methyl farnesate (111, which is then degraded to aldehyde 58. The epoxide moiety is introduced via chloroketone 60 by a method adapted from Johnson’s earlier C,,-JH synthesis. Since this synthesis starts with rransgeranylacetone 6 6 ) , the C, double bond is homogenous. The C, linkage is established in the Wadsworth-Emmons reaction. The reaction gives a 2:l mixture favoring the desired 2E stereoisomer which is obtained in pure form by distillation. Although the Stanford group originally reported that the epoxide construction occurs with 92% stereoselectivity, Henrick and co-workers were only able to obtain 36 as an 82:18 mixture with its C,,-C,, trans isomer.

**

’*

Juvenile Hormones

15

I. NBS-H20

56

II

57

58

59

2. KpCOx - -

*

60

Scheme 11.

Zoecon Synthesis of C,,-JH

The Zoecon group has reported two methods for synthesis of C18JH.19The first (Scheme 12) begins with methylheptenone (611, which is converted into methyl geranate (62). Although this reaction shows only modest stereoselectivity, the 2E stereoisomer is conveniently isolated in pure form by distillation of the crude product. The terminal double bond is cleaved and the resulting aldehyde is treated with the Grignard reagent derived from 2-bromo-I-butene to obtain allylic alcohol 64. The C, double bond stereochemistry is established by Claisen rearrangement (96% stereoselectivity). After selective reduction of the saturated ester function, the synthesis is completed as in Scheme 11. Again, the final hormone is obtained as an 82:18 mixture of cis and trans isomers.

16

OHCA

Acyclic Sesquiterpenes

'

O 63O

,+ COOMe

M

e

''

A

2. LCirA0 3' H

MeOOC,&&COOMe

65

4 5teps

C

CH3C(OMe)

A

O

O

M

e

66 -Me 37

Scheme 12.

First Zoecon Synthesis of C,,-JH

The second Zoecon synthesis (Scheme 13) starts with cyclopropyl carbinol 45, which is solvolyzed to unsaturated chloride 67 as a 3:1 mixture of E and 2 isomers. The mixture of isomers is oxidized by singlet oxygen to obtain allylic alcohol 68 as the major product of a 55:39:6 mixture of isomers. After separation of the mixture, 68 is subjected to Claisen rearrangement using the orthoacetate method to obtain chloroester 69. As usual, the stereoselectivity in this reaction is excellent, only 4% of the Z stereoisomer is produced. The C,-C, double bond geometry is established by adding the cuprate derived from 71 to methyl 2-butynoate to obtain 66.

Sinensals

45

67

17

68

71

72

66

Scheme 13.

Second Zoecon Synthesis of C,,-JH

D. Sinensals The sesquiterpene aldehydes a- and p-sinensal (73 and 74) are important contributors to the aroma and taste of Chinese orange oil. Buchi and

73

74

Wuest have reported the stereorational synthesis of the a isomer (73) which is outlined in Scheme 14.20 The stereochemistry of the Cgdouble bond is assured by the use of the diene alcohol 75 as the starting

I8

Acyclic Sesquiterpenes

Scheme 14.

Buchi’s a-Sinensal Synthesis

material. The synthesis features a novel [2,3]-sigmatropic rearrangement of the ammonium ylide derived from 80 to form amino nitrile 81 (3:2 mixture of diastereomers). Stereochemistry at the C, double bond is established in the final Cope rearrangement; 73 and its 22 diastereomer are produced in a 2:3 ratio. The latter isomer is quantitatively isomerized to the more stable 2E isomer 73 by heating with potassium carbonate. A BASF group headed by Werner Hoffmann has reported a synthesis which affords a mixture of the two sinensals, as well as modifications which allow the production of either pure isomer.*’ The first synthesis (Scheme 15) begins with chloroaldehyde 83, which contains the eventual C, double bond. The chain is elaborated to 88 by two cycles of the basic Nazarov-Ruzicka-Isler synthesis (vinyl Grignard, Carroll reaction).22

Sinensals

83

19

84

Scheme 15.

First BASF Synthesis of Sinensals

Dehydration of 88, followed by deprotection of the aldehyde affords aand P-sinensals in a ratio of 2:l. The EIZratio at the C, double bond in 73 is not stated. A modified synthesis which yields no p-sinensal is shown in Scheme 16. The C,-C,,segment is assembled as shown, using 48% HBr-

6H

n-BuLi

90

89

85

Br-

4-OHC-

O H C 73

Scheme 16.

91

92

BASF Synthesis of Pure a-Sinensal

/

/

20

Acyclic Sesquiterpenes

the Julia method. Bromo diene 90 is obtained as an 85:15 mixture of stereoisomers favoring the desired E isomer. The final Wittig coupling affords a-sinensal (73) as a 1:l mixture with its C, diastereomer 92. The other BASF modification (Scheme 17) leads to p-sinensal (741, uncontaminated by a-sinensal, again as a 1:l mixture with the C,diastereomer (99). The required bis-unsaturated halide 96 is isolated from a 7:3 mixture of 95 and 96 by formation of the sulfolene 97. The remainder of the synthesis follows the same lines as are used to prepare the a isomer.

97

96

96

I 98

OHC

+

74

Scheme 17.

O W % , 99

BASF Synthesis of Pure p-Sinensal

A final synthesis of p-sinensal, from Hiyama's group in Kyoto, is summarized in Scheme 18.23 The synthesis starts with myrcene, a frequently-used precursor for the preparation of 0-sinensal. After oxidation of the more reactive trisubstituted double bond, epoxide 101 is subjected to Crandall-Rickborn isomerization to an allylic alcohol, which is

Sinensals

101

100

I . LDA 2. t-BuMe2SiCI

3.

A

+

nooc

u,

21

OAc I02

I . LiAIH4

2 . PCC

OHC

I03

104

I05

74

Scheme 18.

Hiyama Synthesis of p-Sinensal

acetylated and subjected to Claisen rearrangement to obtain 103, apparently with good stereoselectivity. The terminal unsaturated aldehyde function is introduced by a method developed in Hiyama’s group, whereby the carbanion derived from I,l-dibromo-2ethoxycyclopropane is added to aldehyde 104 at low temperature (-95°C). The resulting adduct (106) is solvolyzed in basic ethanol to obtain the unsaturated acetal:

106

The synthesis of p-sinensal is completed by reductive removal of the acetoxy function, which occurs with double bond isomerization to the more stable position.

22

Acyclic Sesquiterpenes

E. Fokienol, Oxonerolidol, and Oxodehydronerolidol Fokienol (107)’ 9-oxonerolidol (lo@, and 9-oxo-5,8-dehydronerolidol (109) are relatives of the simpler ner~lidol.*~ Fokienol is a stereochemically more complex problem than is nerolidol, since it may exist as four racemates.

I07

109

I08

0. P. Vig’s synthesis of racemic fokienol is outlined in Scheme 19.25 The synthesis begins with a Wadsworth-Emmons reaction of keto acetal 110, which affords unsaturated ester 111 as a 60:40 mixture of Eand 2

Me0

I.

Me0

2 . NoCI, DMSO

A

114

113

I07

Scheme 19.

Vig’S Synthesis of Fokienol and Oxonerolidol

Fokienol, Oxonerolidol, and Oxodehydronerolidol

23

stereoisomers. Vig separates the mixture and carries on only with the correct 2E diastereomer, which is elaborated by straightforward steps into aldehyde 115. The most impressive step in this synthesis is the Wittig reaction on the &y-unsaturated aldehyde 115, which is reported to occur stereospecifically, in good yield, and without enolization or prior conjugation of the unsaturated aldehyde, to give 107. Aldehyde 115 is converted into oxonerolidol (108) by Grignard addition and oxidation. Bohlmann and Krammer have synthesized 9-oxo-5,8-dehydronerolidol (109) as shown in Scheme 20.26 The synthesis starts with the protected unsaturated alcohol 116, which was used as a mixture of cis and trans isomers. The eventual C, double bond is established by reduction of acetylene 117 by chromous hydroxide. The mixture of C, stereoisomers is separated by chromatography after preparation of acid 122. The synthesis is completed by reaction of the correct stereoisomer with 2-methyl-1-propenyllithium.

116 ( c + t )

117 ( c + t )

119 (2 isomers)

118 ( 2 isomers)

HO

A-&G M ~ O Z _OHc., ~ 0

0

~

separatec e q

121 ( 2 isomers)

120 (2 isomers)

122

109

Bohlmann-Gammer Synthesis of 9-0xo-5,8- dehydronerolidol

Scheme 20.

24

Acyclic Sesquiterpenes

F. Gyrindal The norsesquiterpene gyrindal (123) is a defense secretion of the whirligig water beetle. Its synthesis has been reported by Meinwald, Opheim, and Eisner, of Corne1Iz7 and by Miller, Katzenellenbogen, and Bowles,

123

of IlIinois.28The two syntheses, which are essentially identical, are outlined in Scheme 21. They differ only in the protecting group used for geraniol- the Cornell group employed the acetate whereas the Illinois group used the mesitoate-and in the method used for reducing the triple bond-the Cornell group used Li/NH, whereas the Illinois group used LiA1H4-NaOMe. The Illinois team reports a much higher overall yield (9.6% vs. 1.7%).

OLi

125

124

127

128

MnOe_

W

C

H

0 I23

Scheme 21.

Synthesis of Gyrindal

O

Sesquinosefurnn and Longifolin

25

Kato et al., have described the synthesis of oxocrinol (129),29 a norsesquiterpene from marine algae (Scheme 22). The synthesis is conceptually identical to the synthesis of geranylgeraniol by Altman, Ash, and (E, El-allylic chloride 131 was coupled with sulfone M a r ~ o n Altman’s .~~ anion 132. After reductive cleavage of the sulfone moiety and deprotection, oxocrinol (129) was obtained in 18% overall yield.

I30

I33

131

I . Li/EtNH

2_ 2. HCI

I29

Scheme 22,

Kato’s Synthesis of Oxocrinol

G. Sesquirosefuran and Longifolin

Sesquirosefuran (134) and longifolin (135) are the first 2,3-substituted furans in the sesquiterpene series. Sesquiterpene 134 has been prepared by three groups. 31-33 All three syntheses rely on coupling geranyl

I34

135

26

Acyclic Sesqulterpenes

bromide with a 3-methyl-2-furylmetallic reagent (Scheme 23). The exact reagents employed have been the f ~ r y l l i t h i u m the , ~ ~ bis-f~rylmercury,~~ and the furylmagnesium bromide.33 The highest coupling yield (40%) is

134

Scheme 23.

Synthesis of Sesquirosefuran

reported for the furyllithium reagent. A related synthesis of longifolin (135) is outlined in Scheme 24.34 The final coupling proceeds in only 4% yield.

1

(Et0)2PCH2COOEt

136

NH40Ac

t

&i&CO,Et

I37

138

I40

139

135

Scheme 24.

Synthesis of Longifolin

Davanafurans

21

H. Davanafurans The davanafurans (141-144) a-- a set of stereois meric farnesene derivatives which contribute to the characteristic odor of Davana oil. The principal component is isomer 141. Thomas and Dubini have synthesized

141

142

143

I44

the four isomers starting with linalool oxides of known stereochemistry.35 The synthesis of isomers 141 and 142 is summarized in Scheme 25. The (-)-cis-linalyl oxide 145 is converted by standard methods into the cis ketone 146, which is treated with 5-methyl-2-furyllithium to obtain a mixture of unstable tertiary alcohols. The alcohol mixture is hydrogenolyzed using LiAIH,-AICl, to obtain a mixture of 141 and 142 in a ratio of 4:l. Similar transformation of (+)-trans-linalyl oxide affords

I . PhCOCl

145

147

Scheme 25.

I46

141

142

Thomas-Dubini Synthesis of Davanafurans

28

Acyclic Sesquiterpenes

143 and 144 in a ratio of 3 : l . Although the absolute stereostructures of the davanafuran group are established by their synthesis from linalool of known absolute configuration, the relative stereochemistry within the cis and trans pairs is still open to question. That is, does the major natural diastereomer correspond to 141 or 142? The major product of the reduction shown in Scheme 25 is identical with the major natural davanafuran. Thomas and Dubini argue that the major product in this reduction should have structure 141 on the basis of transition state 148. However, the alternate formulation 149, which leads to 142, would seem to be more consistent with the Cram-Felkin model for asymmetric induction.

Ar

$MeA r

MI

li

H

148

9

n I49

141

H - A I ~

-

Ar

H

H

I42

I. Dendrolasin, Neotorreyol, Torreyal, Ipomeamarone, Freelingyne, and Dihydrofreelingyne The common structural unit in this group of sesquiterpenes is the psubstituted furan ring. Several syntheses of dendrolasin (1501, neotorreyol (1511, torreyal (152), and ipomeamarone (153) were recorded in our original review.36 The first three present the challenge of double bond stereochemistry. In ipomeamarone there is the similar problem of achieving cis,trans selectivity about the tetrahydrofuran ring. Freelingyne (154) and dihydrofreelingyne (155) are much more challenging synthetic targets. In addition to their highly unsaturated nature, both

Dendrolasin, Neotorreyol, Torreyal, etc.

29

have two trisubstituted double bonds and 155 has a disubstituted double bond. 0

0

\

H

\

O\

I50

I/

U

151

153

154

155

Takahashi has reported a synthesis of dendrolasin which is summarized in Scheme 26.37The synthesis features a novel method for construction of the furan ring, in which intermediate 159, a mixture of cis and trans

&

B

~

~

NaOEt

0 I56

-0Ef

NaOH EtOOC

OEt

I57

160

Scheme 26.

A

HzO-EtoH

I 50

Takahashi Synthesis of Dendrolasin

30

Acyclic Sesquiterpaes

isomers, is irradiated under acidic conditions to produce the ethoxybutenolide 160. A two-stage reduction process converts 160 into dendrolasin. Kondo and Matsumoto have accomplished the interesting syntheses of 150-152which are outlined in Scheme 27. The starting point is myrcene (loo), which is converted by two consecutive singlet oxygen oxidations into the endoperoxide 163. The initial oxidation provides 162 and its isomer 161 in a ratio of 2:3; the desired isomer is obtained in 36% yield by chromatography. The furan ring is formed from endoperoxide 163 by base-catalyzed scission of the peroxide, followed by acid-catalyzed cyclization. Reaction of allylic alcohol 164 with thionyl chloride gives trunsw-chloroperillene (165),which is used to prepare 150-152as shown.

I00

161

& @ 2.I . I-BuOLi. H2S04 OH

@ Soci2__ OH

164

163

165

166

150: X = H 151: X * O H

Scheme 27.

162

152

Kondo-MatsurnotoSynthesis of Dendrolasin, Neotorreyol, and Torreyal

Dendrolasin, Neotorreyol, Torreyal, etc.

31

Ipomeamarone (153) and its trans isomer epi-ipomeamarone (167) are toxic metabolites produced by molds which infect sweet potatoes. The

I67

chief synthetic problem in this series, control of stereochemistry about the tetrahydrofuran ring, has still not been solved, as all syntheses to date are StereoRons~ci~c. A synthesis by Burka, Wilson, and Harris of Vanderbilt is summarized in Scheme 28.39 Beginning with ethyl 3furoylacetate (168), keto acetate 172 is constructed in a straightforward manner. Wadsworth-Ernmons reaction of 172 affords enone 173. Hydrolysis of the acetate occurs with concomitant cyclization to an equimolar mixture of epi-~~meamarone (167) and i ~ m e ~ a r o (153). ne Although the equilibrium constant for isomerization of iporneamarone to epi-ipomeamarone is 1, it was found that the formation of the isomer in

I72

I71

)JJ+ 0

NoOH_

q&+v.&6 H

OAc

173

Scheme 28.

167

I53

Burka-Wilson-Harris Synthesis of Ipomeamarone

32

Acyclic Sesquiterpenes

this synthesis is not under thermodynamic control, but rather that the two isomers are formed from the hydroxy enone precursor with equal rates. Kondo and Matsumoto have applied their method for synthesis of furans to the preparation of ipomeamarone and epi-ipomeamarone as shown in Scheme 29.40 In this case the required diene is not commercially available, so it is synthesized by the addition of 2-butadienylmagnesium chloride to aldehyde 174. The singlet oxygen oxidation and two-stage conversion of the resulting endoperoxide to the furan proceed well (70% for the three steps). However, it should be noted that simple addition of 3-furylmagnesium bromide to 174 would probably provide 176 directly. The tetrahydrofuran ring is established by treatment of 176 with N-iodosuccinimide. The cyclization shows essentially no stereoselectivity; the two cis and trans isomers are produced in a ratio of 2:3. The synthesis is completed in a conventional manner, although

175

174

I76

NIS NaHCO,

CI

dioxane

178 ( c + t )

177 ( c + t )

179 ( c + 1 )

Scheme 29.

I53

HMPT

167

Kondo-Matsumoto Synthesis of Ipomeamarone

Dendrolasin, Neotorreyol, Torreyrl, etc.

33

alkylation of 2-lithio-1,3-dithiane by the neopentyl iodide 177 is noteworthy. The final product, a 2:3 mixture of 153 and 167 may be equilibrated by base to a 1:1 mixture of the two sesquiterpenes. The first synthesis of freelingyne (154) was reported by Ingham, Massy-Westropp, and Reynolds, of the University of Adelaide.41 This synthesis is summarized in Scheme 30. Reaction of citraconic anhydride with acetylmethylene triphenylphosphorane affords a mixture of isomeric enol lactones 180-183. Although isomer 182 can be obtained in pure form, the desired isomer 183 cannot easily be separated from 181. However, isomerization of pure 182 affords an easily separable 2:l mixture of 182 and 183. The 3-fury1 acetylenic ester 187 is prepared from 3-furoic acid as shown in the scheme, and is then converted into propargylic

HOOC

EtOOC

I85

184

260' (vacuum)

I86

& ,

EtOOC

I87

189

Scheme 30.

I88

154

Adelaide Synthesis of Freelingyne

34

MonocPrbocyciic Sesquiterpenes

bromide 188. Reformatsky coupling of 188 with 183 affords an alcohol (1891, which is dehydrated by vacuum pyrolysis of the derived acetate. Freelingyne (154) is produced, along with two stereoisomers. Compound 154 may be obtained in 24% yield by chromatography. The Adelaide group has also prepared dihydrofreelingyne (155) as outlined in Scheme 31.42 The synthesis starts with propargylic ester 187, which is elaborated in a conventional series of steps to 155. The synthesis is highly stereoselective, as both Wittig reactions give only a single isomer.

I92

Scheme 3 1.

celite

I93

Adelaide Synthesis of Dihydrofreelingyne

Knight and Pattenden of Nottingham have reported a synthesis of freelingyne which employs a final step similar to that used to form the final double bond in the Adelaide dihydrofreelingyne synthesis.43 The Nottingham synthesis is summarized in Scheme 32. Chloroepoxide 194 reacts with sodium acetylide to produce alcohol 195 as a single isomer. The stereochemistry of 195 was proven by X-ray analysis of a derivative.

a-Curcumene, Dehydroa-curcumenes, Curcuphenoi, etc.

I95

194

THP&*'

196

&OHC , ,) ,

I . H' 30 2. MnOe

THPO&@

I97

\

I98

154

Scheme 32.

35

199

201

Nottingham Synthesis of Freelingyne

The hydroxyl is protected and the acetylene is iodinated. Iodo acetylene 197 is coupled with 3-furylcopper by a Stephens-Castro reaction. After deprotection and oxidation, aldehyde 199 is condensed with phosphorane 200 to obtain freelingyne (154) and its stereoisomer 201 in a 3 2 ratio. The lack of stereoselectivity in this reaction is in sharp contrast to the high selectivity observed by Massy-Westropp in his dihydrofreelingyne synthesis.

3. MONOCARBOCYCLIC SESQUITERPENES A.

a-Curcumene, Dehydrea-curcumenes, Curcuphenol, Xanthorrhizol, Elvirol, Nuciferal, or-Turmerone, Curcumene Ether, and Sydowic Acid

Perhaps the simplest monocarbocyclic sesquiterpenes are the aromatic members of the bisabolane family. Examples of those that have been

36

Monocarbocyclic Sesquiterpenes

synthesized in this period are a-curcumene (11, the dehydrocurcumenes 2 and 3, ar-tumerone (41, nuciferal (5), curcuphenol (a), xanthorrhizol (7),elvirol (8), curcumene ether (91, and sydowic acid (10). For most of these compounds the synthetic task is merely one of assembling the carbon skeleton. Compounds 3 and 5 present a stereochemical problem in their double bonds. Although several are chiral, none have been synthesized in optically active form.

)+q $ +q I

2

3

o+ 4

OHC

5

6

9

7

8

10

Krapcho and Jahngen have synthesized a-curcumene (1) as shown in Scheme The synthesis serves only as a vehicle to demonstrate the utility of olefination via a p-lactone. However, it seems more unwieldy than simple olefination of aldehyde 11 by the Wittig reaction.45 Hall and co-workers of Rutgers (Newark) have worked out a one-pot synthesis of a-curcumene wherein ptolylmagnesium bromide is added to methylheptenone (14) in THF, ammonia is added, excess lithium is

a-Curcumene, Dehydro-a-curcumenes, Curcuphenol, etc.

ACOOH

2 LDA

dCHOh I I

37

AOH &OOH

I2

13

Scheme 1.

I

Krapcho-Jahngen a-Curcumene Synthesis

added, and the resultant solution is rapidly quenched with NH,Cl (Scheme 2).46 a-Curcumene is obtained in 90% yield. The synthesis is an adaptation of one introduced earlier by Birch,47 which requires two operations and affords a-curcumene in only 35% yield.

Scheme 2.

Hall's Synthesis of a-Curcumene

Dehydro-a-curcumene (2) has been prepared by Vig et al. in the unexceptional route shown in Scheme 3.48 Vig and his co-workers have also

38

Monocarbocyclic Sesquiterpenes

Ph,P=CHzCOOEt

&),

I . LiAIH4 COOE+ 2. Collins +

IS

16

Ph,P=C(CH&+ CHO

I7 2

Scheme 3.

Vig’s Synthesis of Dehydro-a-curcumene

synthesized the dehydrocurcumene isomer 3, as shown in Scheme 4. The key step is Wittig olefination of aldehyde 19; no mention is made of cis,trans isomerism in the product.

18

19

A 3

Scheme 4.

Vig’s Synthesis of Dehydrocurcumene 3

Grieco and Finkelhor have reported a nice two-step synthesis of UP turmerone (Scheme 51.” The dianion of keto phosphonate 20 is alkylated with benzylic bromide 21 and the resulting keto phosphonate is condensed with acetone to give the sesquiterpene.

a-Curcumene, Dehydro-a-curcumenes, Curcuphenol, etc.

39

4

Scheme 5 .

Grieco's Synthesis of apTurmerone

Park, Grillacea, Garcia, and Maldonado, of the University of Mexico City, reported the synthesis of apturmerone shown in Scheme 6.'* The interesting feature of this synthesis is reaction of dienone 26 with lithium dimethylcuprate; addition occurs only at the less hindered double bond.

Mo2CuLL

o*

26 4

Scheme 6.

Garcia-Maldonado Synthesis of apTurmerone

40

Monocrrbocyclic Sesquiterpeoes

Gosselin, Massoni, and Thuillier, of the University of Caen, have synthesized apturmerone from bromide 27. The derived dithio ester 28 is alkylated with methallylmagnesium bromide, and the resulting dithioketal hydrolyzed. Conjugation of the double bond affords apturmerone (Scheme 7).52 It should be noted that Mukherji has previously converted bromide 27 into 4 in a single stepsS3

30

Scheme I.

4

Thuillier Synthesis of opTurmerone

An interesting synthesis of nuciferal has been accomplished by Gast and Naves (Scheme 8LS4 The synthesis begins with an imaginative synthesis of phosphonium salt 32 from 1-ethoxy-2-methylbutadiene(31). Formation of the stabilized ylide in the presence of a - p tolylpropionaldehyde (33) affords a 4 5 5 5 mixture of dienes 34 and 35, which is hydrogenated to give nuciferal (S), along with 15% of the tetrahydro product. As in other nuciferal syntheses,” as well as in syntheses of the sinensalsS6 and t ~ r r e y a l ~ ’ control * ~ ~ , of stereochemistry at the trisubstituted double bond is not a real problem, as the E

a-Curcumene, Dehydro-a-curcumenes, Curcuphenol, etc.

* 'PPh,

Br-

OHC'

31

OEt

d

CHO

35

Scheme 8.

H

O

33

NaOAc EtOH

32

+

\C

41

Hfpd%

J+ OHC

34

OHC

5

Gast-Naves Synthesis of Nuciferal

stereoisomer is apparently much more stable than the 2 stereoisomer. This apparent steric bulk of the aldehyde functional group is surprising when one considers the probable preferred conformations of the two isomer~:~~

&

'HR

H

H

H H H

Since angle A is much more acute than angle B (109°C vs. 120"C), one might expect the E stereoisomer to experience more crowding. However, whatever the explanation, the difference in energy between the two stereoisomers seems to be substantial, since there has never been a report of any 2 stereoisomer under equilibrium conditions.

42

Monocrrbocyclic Sesquiterpenes

Phenols 6-8 do not pose any stereochemical problems. They have mostly been synthesized by straightforward adaptations of methods previously used for the synthesis of the curcumenes. Curcuphenol (6) was synthesized by McEnroe and Fenical as shown in Scheme 9.60 The synthesis is characterized by its excellent overall yield (86%). An interesting chemical transformation is the quantitative dehydration of the tertiary alcohol 37 upon treatment with acetic anhydride in pyridine at room temperature. A synthesis of xanthorrhizol (71, an isomer of curcuphenol, had been recorded earlier by Mane and Krishna Rao.61 This synthesis, outlined in Scheme 10, requires eleven steps beginning with ketone 39. It is clear that the Fenical approach (Scheme 9 ) is greatly superior. Elvirol (8) is interesting because of its divergence from the head-to-tail isoprene rule. It has been synthesized twice, by Bohlmann and Koring,62 who first isolated the compound, and also by Dennison, Mirrington, and Stuart.63 The Bohlmann synthesis is outlined in Scheme 11. Although it is a good deal longer than other syntheses of bisabolenetype sesquiterpenes, it does employ rather different methodology. Unfortunately, many of the steps proceed in low yield. For example, the

Scheme 9.

McEnroe-Fenical Synthesis of Curcuphenoi

a-Curcnmene, Dehydro-a-curcumenes, Curcuphenol, etc.

"moH

Br$ooE'

Meo&

H2-Pd/Cb

\

43

Mro&cooE,

COOEt

40

39

I . LiAIH, 2. per3

Br

41

I . CH2(COOEt)2 NaOEt 2. NaOH,H20 3. CsHsN. A

I . CHzN2 t 2. MeMgI

42

43

OOH

OH

45

44

7

62

Mane-Kirshna Rao Synthesis of Xanthorrhizol

Scheme 10.

,.A

I - =I + . B H

I

Ph PBr

2 ,

62..pcHo 62. 2. H202

2. DHP, H+

OH

47

46

48

OTHP

OTHP

Ph3P@OMew

Me,$-OL

49

~

51

50

HjOt

HO .

52

Scheme 11.

I

%

P

4

~

8

Bohlmann Synthesis of Elvirol

OM8

44

Monocnrbocyclic Sesquiterpenes

overall yield for the last four steps is only 0.3%. It is not clear why the two-stage Khronke process for converting alcohol 48 into aldehyde 50 must be employed rather than direct oxidation. The Dennison-Mirrington-Stuart synthesis, outlined in Scheme 12, employs the method first introduced by Birch4’ and subsequently employed by Hall and c o - ~ o r k e r for s ~ ~the synthesis of curcumene. The synthesis is efficient and was easily adapted to a synthesis of isomer 56, thus allowing an unambiguous assignment of structure for the natural product.

The final members of this class which have been synthesized are curcumene ether ( 9 ) and sydowic acid (10). Curcumene ether was synthesized by Vig et al. as outlined in Scheme 13.64 Keto ester 57, obtained in the straightforward manner shown, is treated with excess Grignard reagent to obtain diol 58, which is cyclized by treatment with sulfuric acid. The final step proceeds in Soo/b yield. Vijayasarathy, Mane,

kBr C6H5L i

14

&s

53

54

Na, NH3 __L

55

8

Scheme 12.

Dennison-Mirrington-Stuart Synthesis of Elvirol

a-Curcumene, Dehydro-a-curcumenes, Curcuphenol, etc.

45

and Krishna Rao employed essentially the same method to synthesize sydowic acid (10, Scheme 14).65 The crucial oxidation to give 62 proceeds in only 21% yield. A number of other oxidation methods were examined, but none provide the acid in better yield. MeMgI

*

2. EtOH, COOEI

57

Jyp-0””; 9

OH

58

Vig’s Synthesis of Curcumene Ether

Scheme 13.

a.

I. AIClf

+

0

2. MIOH, H2S04

COOMe

59

bH

60

HOOC

Scheme 14.

10

Krishna Rao Synthesis of Sydowic Acid

46

Moaocubocyclic Sesquiterpeaes

B. Sesquichomaenol This interesting seco-cadinane (63) is the only member of its skeletal

63

class. A synthesis by Takase and co-workers66is outlined in Scheme 15. The synthesis follows well-precedented lines and provides the racemic natural product in reasonable yield.

65

64

a 66

67

48% HBr,

A 68

Scheme 15.

A 63

Takase Synthesis of Sesquichamaenol

Bisabolenes, Lanced, and Alantone

47

C. Bisabolenes, Lanceol, and Alantone A number of isomeric bisabolenes occur in nature. Those which have been synthesized are a-bisabolene (69), p-bisabolene (70), Zy-bisab-

olene (71), By-bisabolene (72), and isobisabolene (73). The related oxygenated bisabolenes lanceol (74) and a-alantone (75) will also be discussed in this section. Within this group, the simpler synthetic targets

Jqqd" 70

69

73

72

71

OH 74

75

are 70 and 73, as there is no diastereomer problem. With a-bisabolene (as), lanceol (74), and a-alantone ( 7 9 , there is a trisubstituted double bond which complicates any synthesis. With y-bisabolene, it has recently been found that both stereoisomers, with regard to the tetrasubstituted double bond, occur in nature. A problem in interpreting much of the synthetic literature is that questions of stereochemistry have often been ignored. Several syntheses have been published which should lead to a mixture of diastereomers, but the authors seem to beg the issue of isomeric purity or even of isomer identity. Two syntheses of a-bisabolene (69) have been published. These syntheses, by Vig and c o - ~ o r k e r and s ~ ~Teisseire and co-workers6* typify the stereochemical problem. The synthesis by Via et al. is outlined in Scheme 16. It was reported that ketone 76, the Diels-Alder product from butadiene and methyl vinyl ketone, reacts with triethyl phosphonoacetate anion to give only the Z diastereomer 77. This compound is transformed into a substance purported to be a-bisabolene as shown in the scheme.

48

MonocarbocycllcSesquiterpenes

Scheme 16.

Vig’s Synthesis of a-Bisabolene

The synthesis of a-bisabolene by Teisseire et al. is a simple Wittig coupling of ketone 76 with phosphorane 79 (Scheme 17).65 Their paper notes that the Wittig reaction gives only a single isomer, which they presumed to be 80, since that was the then-accepted structure of abisabolene. Delay and Ohloff reinvestigated both the Vig and Teisseire

Scheme 17.

Teisseire’s Synthesis of a-Bisabolene

syntheses in 1979.69 They found that the Teisseire Wittig coupling is indeed highly stereoselective, giving the two diastereomers in a 95:s ratio. However, they also showed that the major product, and hence the natural product itself, is the Zstereoisomer 69. Interestingly, the related Wittig coupling of ketone 76 with saturated phosphorane 81 gives mainly the E stereoisomer 82. In their reinvestigation of the Vig synthesis,

Bisabolenes, Lanced, and Alantone

81

76

82

3 :I

49

83

Delay and Ohloff found that the major isomer produced in the triethyl phosphonoacetate reaction on ketone 76 is the E isomer (~OYO),rather than the Zisomer as alleged by Vig and co-workers. Thus, it seems that the synthesis outlined in Scheme 16 provides the natural product as the minor component of a 4:l mixture. Vig has also published a synthesis of P-bisabolene (70) during the period of this review;’Oa the same investigator has previously published two other syntheses of the same t e r ~ e n e . ~The ’ ~ synthesis (Scheme 18) follows a path which is essentially identical to one of Vig’s earlier syntheses of 7 0 the only difference is the use of a P-keto sulfoxide (86) rather than a p-keto ester. Isobisabolene (73) was also prepared by Vig through a simple modification of the synthesis shown in Scheme 18, where 84 is methylenated by the Wittig reaction, rather than subjected to the Grignard-dehydration sequence.

87

Scheme 18.

88

Vig’s Third Synthesis of fi-Bisabolene

70

50

Mooocarbocyclic Sesquiterpenes

The y-bisabolenes (71 and 72) are interesting because of their tetrasubstituted double bonds. The first reported synthesis came from Vig et al. (Scheme 19).71 The question of stereochemistry is totally ignored. The initial reaction probably gives both diastereomers. Aside from the problem of stereochemistry, the most interesting feature of this synthesis is the first step: Wadsworth-Emmons reaction on the sensitive &y-unsaturated ketone 89.

89

90

91 COOMe

&'

C H 2 ( C O O M e ) ~ &cooMe I-BuOK

92

94

Scheme 19.

I , NoCN, DMSO ~

2. LiAIHg

93

71 (+72?1

Vig's Synthesis of the y-Bisabalols

Teisseire and co-workers reported a synthesis of the mixture of 71 and 72 (Scheme 20).68 The synthesis begins with terpinolene (95) and proceeds via alcohol 97. The double bond geometry is set in the Claisen rearrangement, which provides 98 and 99 in a 60:40 ratio. Faulkner and Wolinsky independently executed the same synthesis, except that they prepared alcohol 97 by addition of isopropenylmagnesium bromide.72 Both the Teisseire and Faulkner teams report the separation of the final y-bisabolene stereoisomers, but neither team was able to assign stereostructures to either the synthetic products or the natural compound. Subsequently, Faulkner and Wolinsky developed the alternate synt-hesis

Bisabolenes, Lnnceol, and Alantone

Scheme 20.

51

Teisseire’s Synthesis of the y-Bisabolenes

of Scheme 2 ~ The ’ ~Krapcho method44 was applied to acid 100 and methylheptenone to obtain a mixture of diastereomeric P-hydroxy acids 101 and 102 which was separated by fractional crystallization. Each isomer was converted into the corresponding y-bisabolene. The structure I.

-

PLDA

f ractionol cryatalliration

100

72

104

Scheme 21. Faulkner-Wolinsky Synthesis of the y-Bisabolene Stereoisomers

52

Monocarbocycllc Sesquiterpenes

of 101 was determined by single-crystal X-ray analysis, thus establishing the stereostructures of the two terpenes, both of which are natural products. Crawford, Erman, and Broaddus, of the Procter and Gamble Company, have studied the metallation of limonene with a complex of nbutyllithium and tetramethylethylenediamine (TMEDA) .74 The reaction provides simple syntheses of several bisabolene class sesquiterpenes, as shown in Scheme 22. The Wadsworth-Emmons reaction used in the synthesis of lanceol (74) actually gives the E and Z stereoisomers in a ratio of 8:l. This ratio can be changed to 5:3 by ultraviolet irradiation. The two isomers were separated and each was converted into the corresponding lanceol isomer. The authors cast a shadow on the assigned stereostructure of the natural alcohol, since it was determined by oxidation to the aldehyde. As discussed earlier, it is known that such

105

I08

1

-

106

70

f

OH

I07

I. Collins

2. E t 2 O 3 A C 0 O E t 3. LiAIH4

74

Collins

75 109

Scheme 22. Procter and Gamble 8-Bisabolene, Lanceol and Alantone Syntheses

Bisabolenes, Lanceol, and Alantone

53

aldehydes show a strong preference for the E configuration and readily isomerize under the conditions used to accomplish the lanceol oxidation. Unfortunately, spectra for natural lanceol were not publi~hed.'~ A synthesis of (E)-lanceol has also been reported by Cazes and Julia of the C. N. R. S. in Paris. 76 The synthesis (Scheme 23) employs the Crawford method for activation of limonene and utilizes the novel ketene dithioacetal 110 as an isoprene equivalent. The synthesis provides lanceol in an overall yield of 30%.

Ill

6U

74

Scheme 23.

Cazes-Julia Synthesis of Lanceol

Krishna Rao and Alexander have reported the efficient synthesis of alantone which is summarized in Scheme 24.77 The synthesis begins with free radical addition of carbon tetrachloride to limonene; the product, a result of addition, followed by elimination of HCl, is 112. Vigorous hydrolysis of this material affords the crystalline acid 113, which is itself a natural product. Acid 113 reacts with a large excess of isobutenyllithium to give a-alantone (75). Although the conversion of 113 to product is low (only 35%) the yield of 75 based on unrecovered starting material is good (79%).

54

Monoelrbocyclic Sesqnlterpenes

113

75

Scheme 24.

Kirshna Rao Synthesis of Alantone

Another synthesis of a-alantone, by Babler, Olsen, and Arnold, begins with addition of vinyllithium to ketone 76 (Scheme 25).'* The resulting allylic alcohol is rearranged to the allylic acetate 115, which is said to be at least 90% E stereoisomer. The synthesis is completed in a straightforward manner.

2I .. NOOH, Coilinr

Pno ILMg? fl 117

116

I . Collins 2. MOO-

___c

75

Scheme 25.

Babler-Olsen- Arnold a-Alantone Synthesis

a-Bisabalol, a-Bisabololone, Deodarone, err.

55

D. a-Bisabolol, u-Bisabololone, Deodarone, Juvabione, and EpiJuvabione

Of the bisabolene family of sesquiterpenes, the most challenging are those which contain two asymmetric carbons, one in the cyclohexene ring and one in the side chain. This remains a major problem, although considerable progress has been made toward stereoselective syntheses of this type of compound. Five natural products of this kind have been synthesized-a-bisabolol (1181, a-bisabololone (1191, deodarone (1201, juvabione (1211, and epijuvabione (122). Four syntheses are done 0

121

I22

under conditions of good stereocontrol-Ficini’s 1974 epduvabione synthesis, Evans’ 1980 epijuvabione synthesis, and the 1979 syntheses of a-bisabolol by Schwartz and Kakisawa. Several other syntheses follow routes which require separation of mixtures of diastereomeric intermediates. In addition, there have been the normal quota of syntheses which blithely ignore the stereochemical problem altogether. There is yet to be a synthesis which produces diastereomerically homogeneous deodarone (120). Indeed, the relative configuration of this sesquiterpene is still unknown. a-Bisabolol (118) was first prepared by Ruzicka in 1925 by a method which undoubtedly gives a mixture of the two dia~tereomers.~~ The next report of the synthesis of alcohol 118 was by Manjarrez and Guzman.80

56

Monocrrboryclic Sesquiterpenes

However, it has been noted that the material prepared by these workers is probably a mixture of double bond isomers, methyl group positional isomers, and diastereomers." Rittersdorf and Cramer prepared a-bisabolol as shown in Scheme 26, which is essentially the same as the earlier Ruzicka synthesis.82 The initial Diels-Alder reaction of isoprene and methyl vinyl ketone gives ketones 76 and 123 in a ratio of 2:3. However, as in the Ruzicka synthesis, the mixture was not separated but was subjected to Grignard addition to obtain 124 and 125, each as a mixture of diastereomers. Rittersdorf and Cramer were able to separate the diastereomers by glpc of their trimethylsilyl derivatives. As expected, they are formed in equal amounts. The cyclization of (2Z,6E)-farnesyl phosphate (126d and (6E)-nerolidyl phosphate (126b) was also studied. Among the products formed are the two a-bisabalol isomers 118 and 127. The 118:127 ratio is 39:61 from phosphate 126a and 4 5 5 5 from phosphate 126b.

12Sb

I27

Gutsche, Maycock, and Changsl studied the acid-catalyzed cyclization of farnesal and nerolidol. To prepare an authentic specimen of a-bisabolol, they modified the earlier Ruzicka and Manjarrez-Guzman synthesis. Because pure ketone 76 could not easily be obtained from the Diels-Alder reaction of isoprene and methyl vinyl ketone, they prepared it as shown in Scheme 27. Preparation of 76 by this method is tedious and gives the product in poor yield, but it is isomerically pure. The Grignard alkylation of 76 gives a mixture of diastereomers 118 and 127.

a-Bisabalol, a-Bisabololone, Deodarone, etc.

I24

Scheme 26.

125

Ruzicka-Ritterdorf-CramerSynthesis of a-Bisabalol

COOH I . MeLi A, 7 days 0

2. H,O+,E

76

lie, 127

c

I28

Scheme 27.

lie, 127

Gutsche-Maycock-Chang Synthesis of a-Bisabolol

57

58

Monocsrbocyclic Sesquiterpenes

NoOAc-NaI 129

131

I30

4 -

Scheme 28.

Knoll-Tamm Synthesis of a-Bisabolols

MI

\\

A

133 (+2E)

134

1-

3 steps

135

118

Scheme 29. Kakisawa-Schwartz Synthesis of (6SR,7 SR)-a-Bisabolol

Me1 +

a-Bisabalol, a-Bisabololone, Deodarone, etc.

59

Two groups have prepared radio-labeled a-bisabolols for biosynthesis experiments. Forrester and Money prepared a mixture of 118 and 127 having the 14C in the C-6 position.83 Kn611 and Tamm prepared a mixture of the diastereomers starting with the two pure enantiomers of 4-acetyl-1-methylcyclohexene (76), which were prepared from optically active limonenes (129).84 The synthesis of one of the optically active diastereomer mixtures is shown in Scheme 28. The first syntheses of the pure diastereomers of a-bisabolol were reported in 1979 by Schwartz and Swansons’ of Florida State University, and Iwashita, Kusumi, and Kakisawa,86 of the University of Tsukuba. The syntheses are identical and employ intramolecular cycloaddition of a nitrone derived from farnesal. Synthesis of the 6SR, 7SR diastereomer from (6Zbfarnesal is outlined in Scheme 29. The nitrone 133 is prepared by treating (6Zbfarnesal (132) with N-methylhydroxylamine. Both 132 and 133 are mixtures of 2Eand 2 2 stereoisomers. Cycloaddition of 133 gives a mixture of diastereomeric oxazolidines 134. Although 134 is a mixture of diastereomers, the relative stereochemistry at C-6 and C-7 is the same, since this is determined by the stereochemistry of the C-6 double bond. The two groups complete the synthesis by different routes, although each involves reductive cleavage of the C1-nitrogen bond. The 6SR,7RS diastereomer 127 was prepared by applying the same sequence to (6EI-farnesal. The natural a-bisabolol is identical with 118. Since Kn6ll and Tamm had prepared a-bisabolol from (S)-limonene (1291, the natural alcohol is shown to be 6S,7S In 1974, Gopichand and Chakravarti reported the synthesis of a-bisabololone (119) and deodarone (120) which is summarized in Scheme 30.87 The side chain is attached to ketone 76 by an interesting method discovered by Suga, Watanabe and Fujita.88 Although it was not determined, the product is certainly a mixture of diastereomers. Synthetic deodarone, also a mixture of diastereomers, is obtained by acid-catalyzed cyclization of 119. Vig and co-workers reported another stereorandom synthesis of 119 and 120 in 1976 (Scheme 31).89 The synthesis utilizes P-keto ester 138, which has previously been used as a precursor to P-bisabolene and lanceot.90

60

Monocrrbocyclic Sesquiterpenes

I . lithium naphthalmidt

COOH

2. MaOH, H+

4 4

4 6OOMa

76

H2Cr04t

MeMgI

136

2 HClO

‘0

OH I37

120

I19 + isomer

Scheme 30. Gopichand-Chakravarti Synthesis of a-Bisabolol-3-one and Deodarone

3. Collins

138

H

139

I40

I . MeMpI 2 . Jones

141

Scheme 3 1.

119

+ isomer

I20

Vig’s Synthesis of a-Bisabololone and Deodarone

a-Bisabalol, a-Bisabololone, Deodarone, etc.

61

A third stereorandom synthesis of 119 was reported by Malanco and Maldonado in 1976 (Scheme 32).91 The synthesis begins with enone 76,

I44

I19

+ isomer

75

Scheme 32. Malanco-Maldonado Synthesis of a-Bisabolene and Alantone

which is converted into a diastereomeric mixture of limonene epoxides (142) of unknown composition. The epoxide ring is opened by Stork’s protected cyanohydrin method to obtain lactone 143, which is converted into bisabololone (119) by the method shown in the scheme. Dehydration of the mixture of diastereomeric a-bisabololones provides racemic alantone (75). Kergomard and Veschambre also prepared the a-bisabololones from the limonene epoxides, but they separated the isomers first.Y2Thus, they obtained each a-bisabololone diastereomer in a pure state and determined the configuration of the natural product. The synthesis of the natural isomer is shown in Scheme 33. An identical sequence starting with the (SRI-epoxide afforded the other diastereomer of 119, which was clearly distinguishable from the natural product. Unfortunately, the two diastereomers have not been cyclized to obtain the two deodarone diastereomers, which would determine the stereostructure of 120.

62

Monoearbocyclic Sesquiterpenes

145

146

4

KOH/MeOH

147

0

119

Scheme 33.

Kergomard-VeschambreSynthesis of (-)-aBisabololone

The last reported a-bisabololone synthesis, due to Park, Grillasca, Garcia, and Mald~nado,’~ is similar to their synthesis of ar-turmerone (see Scheme 6 ) . In this case the crossed aldol condensation is applied to ketone 76 to afford directly a mixture of diastereomeric a-bisabololones.

119 +isomer

The correct stereochemistry of juvabione (121) and epijuvabione (122) has only recently been elucidated. It turns out that both diastereomers are natural products. ( +1-Juvabione (121, the R,R diastereomer) was the first isolated and appears to be the most widely distributed stereoisomer occurring in balsam and Douglas fir. Epijuvabione (122) has been isolated only once-from a tree of uncertain species by Cerny and cow o r k e r ~ Most . ~ ~ of the earlier syntheses of the juvabiones were not stereocontrolled, although some allowed separation of diastereomers at an intermediate stage.95 The two isomers are exceedingly difficult to dis-

a-Bisabalol, a-Bisabololone, Deodarone, etc.

63

tinguish spectroscopically, since they have virtually identical infrared and 'H-NMRspectra. As reported in Volume 2 of this series,95 Pawson and co-workers synthesized all four stereoisomers of the juvabiones from the enantiomeric limonenes. An intermediate in one of the syntheses was shown by X-ray analysis to be alcohol 148. Since this alcohol was converted into a compound identical with a sample obtained from c e r n ~ , ~ ~ the stereochemistry of the Cerny natural product was established as 122.

I22

148

However, it was not appreciated at that time that the Cerny sample was epijuvabione, and not juvabione itself. Thus, the original structural assignment to juvabione ( R , R ) was correct. This situation provides a good deal of confusion in the literature, as all the published work between 1968 and 1976 have the structures reversed. The first stereocontrolled synthesis of one of the diastereomers was achieved by Ficini and co-workers in 1973 (Scheme 34).96 The synthesis begins with cycloaddition of ynamine 149 to cyclohexenone to give the aminocyclobutene 150. Treatment of this compound, first with water and then with aqueous acetic acid, results in clean conversion to diastereomerically pure keto acid 151. The stereochemistry is presumably established by kinetic protonation of 150 from the convex face to yield the immonium ion 155. The side chain is attached by the StorkMaldonado method, alkylating lithiated a-alkoxy nitrile 153 with

I50

I55

64

MonocarboeyclicSesquiterpenes

I22

Scheme 34.

Ficini's Stereocontrolled Epijuvabione Synthesis

bromide 152. After regeneration of the cyclohexanone carbonyl function, the synthesis of (*)-epijuvabione (122) is completed by standard transformations. In 1973 Farges and Veschambre reported a synthesis of a mixture of juvabione and epijuvabione (Scheme 35)97 which begins with (-)-perillic alcohol (156). The synthesis is essentially identical to the earlier Hoffmann-La Roche synthesis95 and provides a 2:l mixture of (-bepijuvabione (122) and (-)-juvabione (121)' the enantiomers of the natural products.

a-Bisabalol, a -Bisabololone, Deodarone, etc.

I. p-TICI 2. NaCN

HC104_ RO

RO

MeOOC

++ ent-121

Scheme 35.

65

0

MsOOC

?i! 0

mt-122

Farges-VeschambreSynthesis of the Juvabiones

An interesting juvabione synthesis was reported in 1976 by Negishi, Sabanski, Katz, and Brown (Scheme 36).98 The synthesis starts with the commercially-available perillartine (1571, which is converted via the corresponding nitrile into methylperillate (158). This compound is hydroborated using thexylisobutylborane and the intermediate adduct is carbonylated to obtain a 1:l mixture of juvabione (121) and epijuvabione (122). Although stereorandom, the synthesis is remarkably short (four operations) and produces the mixture of 121 and 122 in 53% overall yield.

66

Monocarbocyclic Sesquiterpenes

A

Hot

I57

Me00

f

u

I . Ac20

2 . KOH

-

4. d*p 158

\O

121

Scheme 36.

3. H202, NaOAc

0

MeOOC

122

Negishi-Brown Synthesis of the Juvabiones

An elegant stereocontrolled synthesis of epijuvabione was reported by Evans and Nelson early in 1980.99 The synthesis (Scheme 37) begins with addition of an organozinc reagent derived from l-methoxy-2-butyne to cyclohexenone; diastereomeric methoxy alcohols 159 and 160 are produced in a ratio of 2:l. After anti-reduction of the triple bond, the 2:l mixture of 161 and 162 is rearranged as the potassium alkoxide. Diastereomers 163 and 164 (each a mixture of double bond isomers) are obtained in a ratio of 1O:l. The mixture is carbomethoxylated and the resulting mixture of p-keto esters deoxygenated by a variant of the Bamford-Stevens reaction to obtain a p-y-unsaturated ester. Basecatalyzed conjugation of the latter affords 165, which is converted into ( Abepijuvabione (122) by straightforward steps. The final product is contaminated by about 4% of (*)-juvabione (121). An approach to a stereospecific epijuvabione synthesis has recently been disclosed by Larsen and Monti.’Oo The method has been applied to the synthesis of alcohol 148, an intermediate in the Pawson ~ynthesis.~’ The Monti approach (Scheme 38) establishes the relative stereochemistry by kinetic protonation of the enolate derived from bicyclic ketone 168. After hydrochlorination of the double bond, the ring is opened by Grob fragmentation. The intermediate aldehyde is reduced in sifu to give racemic alcohol 148.

a-Bisabalol, a-Blsnbololone, Deodarone, etc.

I . n-BuLi

2. ZnClp CH3CPCCH20Me

% fi R R2

LiAIH4 ___t NaOMe

* /@

3'

KH 1100

161: RI = MeO, R2.H 162: R I C H Re = Me0

159: R I =MeO,

R2'H 160: A1 = H R 2 = Me0

I63

164

2I .. PLDA hm-NH2

OMe 2 . Jones

MeOOC

3 . NoOMe

MeOOCh

C

O

I65

121

Evans-Nelson Synthesis of (f)-Epijuvabione

&2. I . LMe1 DA

2. I . LDA H30+

Me&

~

'&

I68

167

169

HCI H20-pentone

H

170

Scheme 38.

H

166

122

Scheme 37.

O

148

Monti's Synthesis of an Epijuvabione Precursor

67

68

Monocorbocyclic Sesqulterpenes

E, Deoxytrisporone, Latia Luciferin, and Abscisic Acid A number of sesquiterpenes have the 1,3,3-trirnethyl-2-cyclohexenyl (cyclocitral) system, which is also common to the retinoids and carotenoids. Biogenetically, this skeleton presumably arises by a cationic mechanism such as:

A number of marine sesquiterpenes have recently been isolated which have the same skeleton, but with bromine at C-4:

In this section, we will discuss the synthesis of deoxytrisporone (1721, latia luciferin (173), and abscisic acid (174). A great deal of closely related synthetic work has been done in the retinoid and carotenoid areas. For a treatment of this research, please refer to other sources.

172

173

I74

A synthesis of deoxytrisporone (172, Scheme 39), by Torii and Uneyama lo' utilizes the novel cyclocitral synthon 176, which is formed by acid-catalyzed cyclization of 175. The ring oxygen is introduced in good yield by selenium dioxide oxidation and the remaining carbons of the side chain are introduced by alkylation of the dianion of hydroxy sulfone 177 with bromo ester 178. After elimination of benzenesulfinate ion, the oxidation states of the two functionalized positions are adjusted to obtain deoxytrisporone (172).

Deoxytrisporone, Latia Luciferin, and Abscisic Acid

69

OH

176

I75

! T178E 2

PH

OH I

GOH Lo+ Ph02S

Li A ' H 4*

177

I

' I79

I80

(IK,,. O

+

I82

181

0

I72

Scheme 39.

Torii-UneyamaSynthesis of Deoxytrisporone

Latia luciferin (173) has been synthesized by Sum and Weiler"' as shown in Scheme 40. The synthesis employs an interesting method for repetitive introduction of isoprene units, alkylation of acetoacetic ester dianion, followed by cyclization and reaction of the derived enol phosphate with lithium dimethylcopper. The sequence is utilized twice, first in the synthesis of methyl cyclogeranate (187), and again in attaching the The synthesis is completed by Baeyer-Villiger side chslin (188-190). oxidation of aldehyde 191 to obtain enol formate 173. Abscisic acid (abscisin 11, 174) is an important plant hormone. It was ~ ~order to first synthesized in 1965 by Cornforth and c o - w ~ r k e r s 'in

70

Monocarbocyclic Sesquiterpenes

I83

184

I86

I87

189

I85

I88

GCH0 * Rc, 190

OCHO

191

Scheme 40.

173

Sum-Weiler Synthesis of Latia Luciferin

confirm its assigned structure. The synthesis consisted of photooxygenation of acid 193 (obtainable from @-ionone, 1921, followed by isomerization of the resulting endoperoxide. The hormone is obtained in 7% yield from acid 193. Mousseron-Canet and co-workers lo4 examined the same procedure and made some modifications which improved the yield of 174 to 50%.

I92

I93

194

MOOlOO*C 0.01N NaOH

174

Deoxytrisporone, Latin Luciferin, and Abscisic Acid

71

Roberts and co-workers discovered that a-ionone (195) undergoes oxidation by t-butyl chromate in f-butyl alcohol to afford hydroxy dione 196 in 23-279'0 yield.''' This substance is subjected to WadsworthEmmons reaction and hydrolysis to obtain abscisic acid and its isomer 197. 0 NOH

I95

196

I74

I97

Findlay and MacKay worked out an alternative preparation of intermediate 196 beginning with p-ionone (192). lo' This modification produces 196 in 439'0 yield.

192

198

I99

196

72

Monocnrbocyclic Sesquiterpenes

The trans,trans isomer 197 has also been prepared by Okuma as outlined in Scheme 41.'" Emmons reaction on ketone 202 affords esters 203 (minor) and 204 (major). Hydrolysis and reduction of isomer 203 gives the abscisic acid isomer 197, but hydrolysis of the cis isomer 204 affords the enol lactone 205. OLi

200

13.

203 I . OH2 . H,O+ HZ

197

Scheme 41.

202

201

204

1

I . OH2. n,o+

205

Okuma's Synthesis of an Abscisic Acid Isomer

As part of a project to determine the absolute configuration of abscisic acid, Mori carried out the synthesis with optically active hydroxy dione 196 (Scheme 42).Io8 The synthesis begins with a surprising selective reduction: dione 206 is hydrogenated selectively at its more hindered carbonyl. What is even more surprising is that the product is the trans isomer 207. This ketol is resolved by formation of an ester with 38-acetoxyetienoyl chloride. After separation of the diastereomeric

Deoxytrisporone, Latin Luciferin, and Abscisic Acid

73

esters, isomer 208 is converted by the multi-step route shown into unsaturated alcohol 214. Epoxidation of the derived acetate affords cis epoxide 215 (20% yield) and trans epoxide 216 (6.4% yield). The latter product is converted into the dextrorotatory enantiomer of hydroxy dione 196, which had previously been converted into (+)-abscisic acid.

LiA l H(OR)

206

207

I . DHP, H +

209

I.

A,\

THPO.,

I . Jones

--

211

-

HO.

I. A c p

AcO\

2. POC13

212

I. LiAIH4 2. DDP

THPO.,

210

WgBr

2. H,0+

208

213

"--& 0 2. I . m-CPBA* AcoO

214

Scheme 42.

Aco~' &,o

21 5

Mori Synthesis of (+)-I96

Y

OAc

74

Monocarbocyclic Sesquiterpenes

19s

217

t - BuOC rOjH

218

o'Q+cooEt 219(E) + 2 2 0 ( 2 )

Scheme 43.

Oritani-Yamashita Synthesis of Abscisic Acid

22 I

222

224

Scheme 44.

I 22 3

225

Hoffman-LaRoche Synthesis of (+)-Abscisic Acid

A synthesis of optically active ethyl abscisate (220) has also been reported by Oritani and Yamashita (Scheme 43).'09 The synthesis begins with optically active a-ionone (1951, which undergoes oxidation by

Caparrnpi Oxide, 3/3-Bromo-8-epicnparrapi Oxide, etc.

75

selenium dioxide with retention of optical activity to give keto alcohol 217. The optical purity of the product has not been published. This material is converted to ester 218 (2:l mixture of E and Z isomers), which is oxidized by r-butyl chromate to obtain esters 219 and 220 in yields of 57% and 6Y0, respectively. Unfortunately, the minor product corresponds to abscisic acid. Kienzle, Mayer, Minder, and Thomas, of Hoffman-LaRoche in Basel, have reported the synthesis of the pure enantiomers of abscisic acid, as outlined in Scheme 44.”’ The synthesis begins with optically active keto ether 221, which is condensed with the acetylide 222. The diastereomeric diols 223 are separated and independently converted into the two enantiomers of abscisic acid. The major isomer of 223 (80%) corresponds to natural abscisic acid.

F. Caporrapt Oxide, 3fl-Bromo-8-epicaparrapt Oxide, Ancistrofuran, Aplysistatin, and a- and fl-Snyderols The syntheses of these compounds are a bit more complex than those discussed in the previous section because of stereochemical complications. Caparrapi oxide (226), 3p-bromo-8-epicaparrapi oxide (227), and ancistrofuran (228) each contain three stereocenters within the fused ring system. These are, therefore, the easier targets for synthesis. Aplysistatin (229) contains these three centers as well as a fourth. Although a-snyderol (230) and P-snyderol (231) have only three stereo-

” ;I

B& ,f

H

H 227

226

228

H

229

230

231

76

Monocarbocyclic Sesquiterpenes

centers, one is in the side chain, well-removed from the ring. quently, these are the most challenging targets of this group. tunately, neither the relative nor absolute stereochemistry snyderols is known. Thus, the two syntheses recorded to date address the issue of stereochemistry.

ConseUnforof the do not

A synthesis of caparrapi oxide (226) reported in 1975 by Cookson and Lombardi (Scheme 45)"' begins with dihydro ionone (2321, which is condensed with sodium acetylide and cyclized to obtain a 1:4 mixture of 234 and 235. Semihydrogenation of 234, the minor isomer, affords

racemic caparrapi oxide. A subsequent complete study of the crucial cyclization was carried out by Lombardi and Cookson in collaboration with workers at Sandoz and Firmenich in Their results are shown below. Unfortunately, the maximum amount of the caparrapi oxide stereoisomer

-*:

@" w m @" H

234

q&: qT:

H

235

k

k

236

237

16%

80%

I%

3%

6%

25%

40010

29%

2 0%

24%

29%

27%

233

238

239

which may be realized is 20%. That there is any difference in the product ratio at all is of interest. In isomers 238 and 239 the relative stereochemistry of two of the asymmetric carbons is already established. However, in cyclization of 233, all the relative stereochemistry is established

Cnparrapi Oxide, 3&Bromo-8-epicapnrrapi Oxide, etc.

233

232

234

226

Scheme 45.

77

235

Cookson-Lombardi Caparraapi Oxide Synthesis

+

+

in one step. The ratio (235 236):(234 237) = 4:1, suggests substantial asymmetric induction in the initial protonation of 233:

(&p 233

’OH

Hoye and Kurth have reported the synthesis of 3P-bromoThe known 8-epicaparrapi oxide, which is summarized in Scheme 46. 8-hydroxy ester 240 is cyclized by treatment with mercuric trifluoroacetate. The cyclization is specific with regard to the stereochemistry of addition. That is, the E alkene 240 yields only trans products 241. However, there is no diastereoface selectivity, as 241 is an equimolar mixture of two diastereomers. The isomers are separated and the correct one is converted into 3/3-bromo-8-epicaparrapi oxide by conventional methods. The side chain stereocenter does not give any asymmetric induction in attack by the mercury cation in the cyclization reac-

78

Monocnrbocyclic Sesquiterpenes

I . Hg(OCOCF3)

OOMe

2. KEr, Br2

0 240 /COOMI I . DIBAL 2. Separate

Br H

Br

H

241

242

H

227

Scheme 46.

Hoye-Kurth Synthesis of 3fi-Brorno-8-epicaparrapi Oxide

tion. This is not surprising, but it is unfortunate that advantage cannot be taken of the inherent asymmetric induction involved in formation of the tetrahydropyran ring which was observed by Lombardi and Cookson in their caparrapi oxide synthesis, as this would yield the correct stereochemistry for 3fl-bromo-8-epicaparrapi oxide. Ancistrofuran (228) is the principal component of the defensive secretion by soldiers of the West African termite Ancistrorermes cauirhorax It has been synthesized by Baker, Briner, and Evans as shown in Scheme 47.'14 Homocitral reacts with a-furyllithium to give a 3:2 mixture of alcohols 243 and 244. The diastereomers are separated and the double bond is epoxidized, the resulting epoxide is reduced, and the tetrahydrofuran ring is closed by treatment of the mixture of diols 246 with tosyl chloride. Ancistrofuran is produced along with a single cis-fused isomer, 241.

1 % p-

Caparrapi Oxide, 3~-Bromo-8-epicaparrapiOxide, etc.

g

c

H

o

+

243

+

1

' -

245

244

246

228

Scheme 47.

I . Separate

2 . m-CPBA on 243

Li

1

79

247

Baker-Briner-EvansSynthesis of Ancistrofuran

Aplysistatin (229), a brominated sesquiterpene from the sea hare, has been synthesized by Hoye and Kurth (Scheme 48).'15 Homogeranyl tosylate (248) is used to produce the a-phenylthioester 249. Aldol condensation of this ester with 0-benzyllactaldehyde or benzyloxyacetaldehyde affords diastereomeric esters 250 (2:l ratio). The isomers are separated and the major one cyclized by the same method employed for the synthesis of 3/3-bromo-8-epicaparrapi oxide. As in that synthesis, the cyclization shows no diastereoface selectivity; each pure diastereomer of 250 affords a 1:l mixture of diastereomeric cyclization products (251). After oxidative elimination of the a-phenylthio group, the primary carbinyl ether is eliminated by treatment with trityl fluoborate. The reaction occurs with concomitant lactonization to afford aplysistatin directly.

80

Monocarbwyclic Sesquiterpenes

P h S(-1 ACOOMt

W T s

248

floEt I . LDA,ZnCI2 2. PhAOACHO

249

251

250

H

Br

Br

252

Figure 48.

229

Hoye-Kurth Synthesis of Aplysistatin

The snyderols (230 and 231) present a fascinating synthetic problem. Although two syntheses have been reported, the problem of stereochemistry has yet to be addressed. Kato and co-workers have prepared the snydetols as shown in Scheme 49.' l 6 Treatment of nerolidol (253) with the interesting brominating agent 254 affords a- and P-snyderols (230 and 231) each in 2% yield.

253

230

Scheme 49.

Kato Synthesis of the Snyderols

231

Isocaespitol

81

A synthesis by Gonzilez and co-workers of the University of La Laguna, in Tenerife, Spain, is summarized in Scheme Methyl farnesate is cyclized by NBS and cobaltous acetate to give bromo ester 255 in 12% yield. The side chain is elaborated as shown. A noteworthy reaction is the clean double bond isomerization accompanying the conversion of allylic bromide 256 to alcohol 231, which is presumably formed as a diastereomeric mixture.

& * .,& COOMc

46

COOMc

I-EuOH

HOAc

% ::::;

Br

255

H

256

S i02. H20 Br

231

Scheme 50.

Gonza'lez Synthesis of P-Snyderol

G . Isocaespitol Isocaespitol (257) and caespitol (258) are halogenated sesquiterpenes which are produced by the marine seaweed Laurencia caespirosa. Each of the compounds possess six stereocenters, three in each ring, thus presenting an interesting problem for synthesis which has yet to be solved. A nonstereoselective synthesis of 257 has been reported by

257

258

82

Monocrrbocyclic Sesquiterpenes

Gonzdlez and his co-workers (Scheme 5 l ) . ’ l8 The synthesis begins with trans-P-terpineol (2591, which is ozonized. The resulting ketone is condensed with trimethyl phosphonoacetate to obtain a 1:l mixture of 261

259

26 I (+50% E-isomer)

260

rrcl

268 (+120/.isomer)

269

Br

Scheme 51.

‘0H 257

270

‘OH

27 I

Gonzilez Synthesis of Isocaespitol

Isocaespitol

83

and its E stereoisomer in 78% yield. The separated 2 diastereomer is converted into a-bromo ester 264, which is a 5050 mixture of stereoisomers at the brominated position. Epoxidation of alkenes 265 and 266 is remarkably stereoselective; each isomer gives a 7:l mixture of products. The authors proposed that the alkenes each adopt a preferred conformation (265a,266a) in which the bromine shields one face of the double bond. The epoxides are opened by the tertiary hydroxy when the

)rOH

265a

266a

mixture of 267 and 268 is treated with acid-washed alumina. The tertiary ring hydroxy undergoes concomitant dehydration, leading to a mixture of alkenes 269 and 270. Isomer 269, didehalocaespitol, is obtained by fractional crystallization of the mixture. Compound 269 reacts with BrCl to give isocaespitol (257) and its isomer 271 in a ratio of 1:3. The observed regiochemistry of the halogenation step presumably reflects a preference of the electrophilic bromine to coordinate from the less hindered face of the double bond in 269:

i)H

OH

The isocaespitol synthesis is interesting, but it leaves much room for improvement in terms of reaction selectivity. First, the preparation of 261 proceeds with only 50% stereoselectivity, as does the preparation of 264. The oxidation to 267 and 268 shows good stereoselectivity (88%). Thus, to this point the stereochemical yield is 0.5 x 0.5 x 0.88 = 22%. Relative stereochemistry is maintained in the last two steps, but the final halogenation gives a regiochemical yield of only 25%.

84

Monocarbocyclic Sesquiterpenes

H. Lactaral Lactaral (277) is interesting because its novel carbon skeleton departs from the biogenetic isoprene rule. It has been synthesized by Froborg, Magnusson, and ThorCn"' as outlined in Scheme 52. The synthesis begins with the commercially available diester 272, which is converted into the monoprotected diol 273 (42% yield). The mesitoate ester 274 is coupled with bromide 275 by treating the mixture with lithium metal. The mixed coupling gives lactarol (276) in poor yield (44 mg of 276 from 1.8 g of ester 274).

272

273

274

276

277

Scheme 5 2 .

Froberg-Magnusson-Thor6nSynthesis of Lactaral

I. y-Elemene, &Elemenone, Shyobunone, and Isoshyobunone A number of elemanoid sesquiterpenes were synthesized in the 1970s. The simplest examples are y-elemene (2781, /3-elemenone (2791, shyobunone (2801, and isoshyobunone (281).

y-Elemene, P-Elemenone, Shyobunone and Isoshyobunone

278

279

280

85

281

Kato, Kurihara, and Yoshikoshi have synthesized y-elemene as summarized in Scheme 53.l2' The known unsaturated ketal ester 282, obtainable in three steps from the Wieland-Miescher ketone, is allylically oxidized and the resulting P-acetoxy enone transformed by reductive alkylation and mesylation into the P-mesyloxy ketone 285. This material undergoes smooth reduction and Grob fragmentation upon being

282

283

285

287

284

286

288

289

278

Scheme 53.

Yoshikoshi's Synthesis of y-Elemene

86

Monocarbocyclic Sesquiterpenes

refluxed with LiAIH, in 1,2-dimethoxyethane. The resulting unsaturated alcohol 286 is converted by precedented steps into geigerenone (288), which is transformed into y-elemene by way of the dibromomethylene derivative 289. Majetich, Grieco, and Nishizawa reported the synthesis of j3-elemenone which is outlined in Scheme 54.12' The synthesis begins with the monoketal 290, which is converted into alkene 293 by established procedures. The double bond is cleaved by ozone and the bis primary alcohol 294 is converted into diene 296 by the selenoxide elimination method. The synthesis employs a novel method for introducing the

290

291

292

3 . Li-NH3 ' H

293

+

294

295

298

296

+

SMe

299

Scheme 54.

297

279

Grieco's p-Elemenone Synthesis

y-Elemene, &Elemenone, Shyobunone and Isoshyobunone

87

isopropylidene group. The lithium enolate of 297 is treated successively with CS, and methyl iodide to give a 1:3 ratio of 298 and 299. After separation of the isomers, 299 is treated with lithium dimethylcuprate to obtain p-elemenone. The first synthesis of shyobunone was recorded in 1970 by Kato, Yamamura, and Hirata.122a7bThe synthesis (Scheme 5 5 ) begins with a-santonin (3001, which is converted by standard methods into ketal diol 301. This substance is converted by straightforward steps into ketone 303. A noteworthy feature of this synthesis is the reported exclusive formation of enol acetate 304 upoii treatment of 303 with isopropenyl acetate. The remainder of the synthesis is straightforward. Another

300

30I

302

I . MeS0,Cl

2. No1 I

303

HbAc

304

%306

305 I.LiAIH4

2.

JOne8

160°

280

Scheme 55.

308

Yamamura-Kato-Hirata Synthesis of Shyobunone

88

Monocrrbocyclic Sesquiterpenes

noteworthy feature is the thermal conversion of shyobunone (280) into preisocalamendiol (308). Although the equilibrium between a divinylcyclohexane and a 1,5-cyclodecadiene normally favors the former isomer, in this case, equilibrium is apparently shifted by thermal 1,5-hydrogen transfer in the initial Cope rearrangement product 309:

q&

+H$$&-q&-309

CH2 OH

H H

280

309

Friter has reported the synthesis of shyobunone which is summarized in Scheme 56.'23 The synthesis begins with a clever use of the ester enolate Claisen rearrangement on neryl senecioate (310), which affords unsaturated acids 311 and 312 in a 4:l ratio. The cyclohexane ring is formed by cyclization of the acid chloride with stannic chloride and the resulting p-chloro ketones are dehydrochlorinated to obtain a mixture of

310

312

31 3

316

Scheme 56.

31 5

314

2 80

318

Frdter's Synthesis of Shyobunone and 6Epishyobunone

y-Elemene, @-Elemenone.Shyobunone and Isoshyobunone

89

enones 315 and 316. The major isomer is reduced by triphenyltin hydride to obtain shyobunone (280) and 6-epishyobunone (318) in a ratio of 7:3. The sequence may by carried out with geranyl senecioate to obtain a 1:4 mixture of 311 and 312. Williams and Callahan have published a synthesis of (-1-shyobunone, which begins with [2 21 cycloaddition of 1-methylcyclobutene and (Spiperitone (Scheme 57).'24 The photoadduct 319 is reduced to alcohol 320, which is thermolyzed as a xylene solution in a sealed tube at

+

319

320

OH

32 I

322

323

324

SnC I

(-1-280

Scheme 57.

Williams-Callahan Synthesis of (-1-Shyobunone

250°C. The thermolysis product is a complex mixture consisting of the elemanoids 321 (lo%), 322 (13%), 323 (27%), and the aldehyde 324 (43%). Treatment of the latter isomer with stannic chloride affords alcohol 321, which is oxidized to enantiomeric shyobunone. Aldehyde 324 presumably arises from 321-323 by a retro-ene reaction:

90

Monocarbocycllc Sesquiterpenes

Isoshyobunone has been prepared by Alexandre and Rouessac as outlined in Scheme 58,125a9b Enedione 326, prepared as shown, is cyclized to a 2:3 mixture of isomers 327 and 328. The major isomer is separated and converted into 330, which adds vinylmagnesium bromide under copper catalysis to afford a 4:l mixture of isoshyobunone (281) and its 6-epimer 331. L C O O E t

Ct

'-OH)Z; H+

325

2. L M g C l 3. H30+

& 326

& N d t

4 +

321

I . dirtillation 2. HCOOE1,NaH 3. n-BUSH, H +

328

329

28 I

Scheme 58.

330

33 I

Alexandre-Rouessac Synthesis of Isoshyobunone

J. Saussurea Lactone and Temsin

Saussurea lactone (332) and temsin (333) have both been synthesized as part of projects associated with the total synthesis of vernolepin (334). The synthesis of saussurea lactone by Ando and co-workers is summarized in Scheme 59.126a3bIt was carried out to test the epoxy mesylate

q-Q-:

Saussurea Lactone and Temsin

A i

Ab

332

333

91

o@ 334

0

fragmentation (338-339) as a possible method for construction of the vernolepin A-ring. Unsaturated ketal 336, obtained in six steps from a-santonin (3001, is hydrolyzed and the resulting enone is reduced with lithium tri-t-butoxyaluminohydrideto give a 4: 1 mixture of isomeric alcohols. Epoxidation of the mixture gives comparable amounts of two epoxides (337). These are separated and the major isomer is converted into mesylate 338. The fragmentation is carried out by boiling a toluene solution of 338 and aluminum isopropoxide. The lactone ring undergoes concomitant trans-esterification. The allylic hydroxy group is removed by lithium-ammonia reduction, and after reoxidation of the resulting hemiacetal, saussurea lactone is obtained.

300

336

-

Ms

337

I . separate 2. MsCl

I . KOH ~

q-q; -“5h3. Ac,O

HO

338

339

A&

Li”H3e

AcO

340

Scheme 59.

34 I

332

Ando’s Synthesis of Saussurea Lactone

92

Monocarbucyclic Sesquiterpenes

Grieco and co-workers synthesized temsin by the route shown in Scheme 60.127bThe synthesis begins with octalone 342, which is converted into the keto ketal 343 in straightforward steps.'27a The ring-B functionality is introduced in the same manner as was used for Grieco's earlier vernolepin synthesis (see Scheme 61). Thus, epoxidation of dienone 344 gives the a-epoxide 345. Lithium-ammonia reduction of 345 gives an enolate, which is quenched with ammonium chloride while

342

&

343

344

3.H202

I . OBMeOH 2. MI$

4. Jones 3. CH2N2

2. I.NH4CI Li/*"b&

Iko

3. A c g

345

dAc

346

313

Scheme 60.

333

Grieco's Synthesis of Temsin

Vernolepin and Vernomenin

93

excess reducing agent is still present. The P-hydroxyketone so produced undergoes subsequent reduction to give the diequatorial diol, which is acetylated to give 346. The trans-fused lactone ring is formed and alkylated to give the @methyl isomer. After ozonolysis of the A-ring double bond, the resulting diol 350 is dehydrated to obtain 352. The eliminations are accomplished by successive formation of the alkyl o-nitroselenoxide, which undergoes in sifu elimination. This is a weakness of the synthesis because simultaneous bis elimination fails, and the four-step successive process proceeds in only 33% yield. Finally, the C-10 stereochemistry is corrected by formation of the lactone enolate, which is protonated from the &face to afford the correct stereoisomer. The complete synthesis requires 26 steps from octalone 342.

K. Vernolepin and Vernomenin In 1968, the late Professor S. M. Kupchan reported the isolation and characterization of the interesting elemanoids vernolepin (334) and vernomenin (354). In addition to its interesting structure, compound 334 originally attracted attention because of its cytotoxic and antitumor h

334

354

activity. It soon became apparent that many natural products having the a-methylenebutyrolactone unit show comparable biological activity, and that most of the members of the class are not significant medicinal leads. Nevertheless, vernolepin captured the imaginations of synthetic chemists, and for the next decade, the intense synthetic effort which was directed at this one compound was out of all proportion to its importance.

Monocarbocyelic Sesquiierpenes

94

Bp6,

2. I. N a H , M THF d

~

&

"M"po"n",

&e

3. Collins

366

355

351

362

36 I

383

hAC

I . G. I S e C N

I.

03,cn2ci2

MIOH 2 . NaBH4 3 . CH2N2

MeOOC

MaOOC

dAc COOMe

f

I.

"

OAc OOMa

365

364

0 ~

356

H+-

2 . p-TsOH A OAc COOMe

c

360

36?

"

I LDA,THF 2 . cn20 HMPT ~

brnp 369

0

31 I

334

Scheme 61.

on

no

ST0

3?2

& 3se

354

Grieco's Vernolepin Synthesis

+

Vernolepin and Vernomenin

95

The first synthesis to be completed by the end of the decade was due to Grieco and co-workers (Scheme 61).128a3bThe unsaturated ketal 356, obtainable in three steps from 2-ethoxycarbonylcyclohexanone,is methylated and hydroborated to obtain the cis-fused decalone 357, which is isomerized to the more stable isomer 358. The ring-B functionality is introduced using the method previously discussed in connection with the Grieco temsin synthesis (Scheme 60). After cleavage of ring A, and dehydration of the resulting hydroxyethyl side chain, the interesting intermediate 365, which has five contiguous chiral centers, is obtained. Cleavage of the methoxy group affords the 6-valerolactone 366. Hydrolysis of the esters, followed by relactonization, gives a 3:l mixture of prevernolepin (367) and prevernomenin (368). After protection of the free hydroxy group, the two methylene groups are introduced by hydroxymethylation, mesylation, and elimination. The Grieco synthesis requires a total of 31 steps. A key feature was the demonstration that the two methylene groups can be introduced simultaneously at the end of the synthesis. The synthesis of Danishefsky and co-workers is summarized in Scheme 62.129a3b It begins with a Diels-Alder cycloaddition of methyl 1,4-cyclohexadienecarboxylate and the versatile diene 373. The ring-B functionality is introduced by making use of the angular carboxy group. The first maneuver in this regard is iodolactonization, which introduces the eventual free hydroxy group. After dehydroiodination, the lactone ring is saponified and the allylic hydroxy group used to direct epoxidation to the a-face of the double bond. Direct epoxidation of lactone 376 affords the diastereomeric epoxide in poor yield. The ring-A cyclohexenone is converted into the 6-valerolactone of vernolepin by oxidation to aldehyde ester 378, which is reduced and lactonized to afford dilactone 379. A similar process has been used by Chavdarian and Heathcock to prepare vernolepin ana10gs.l~' The first key to the Danishefsky synthesis is the ability to protect the 6-valerolactone as an ortholactone (380) while the y-butyrolactone is reduced and the resulting aldehyde is methylenated to obtain epoxy alcohol 381. Compound 381 reacts with a large excess of dilithioacetate to give a single dihydroxy acid which is esterified to obtain 382. This regiospecific ring opening is the second important feature of

96

Monocarbocyclic Sesquiterpenes

FOOH

2. NaOH TMSO

37 3

K*3

O

k

NaHCOc

374

I . separate 2. LDA 3. Me2N+=CH2 4. Me1

5. NoHC03

6H 368

367

334

Scheme 62.

b

354

Danishefsky’s Vernolepin Synthesis

Vernoiepin and Vernomenin

97

the Danishefsky synthesis. If one assumes diaxial opening of the epoxide ring, then the reactive conformer of 381 is 381a. It has been suggested that conformation 381a is favored over the alternative 381b mainly because the incoming nucleophile must encounter the axial oxygen of the ortholactone function in its attack upon the latter conformation. However, chelation of a lithium cation between the epoxide and

381 a

381 b

alkoxide oxygens may play an important role in stabilizing conformation 381a, for it has been found that the tetrahydropyranyl ether of 381 fails to react under conditions which suffice to convert 381 into 382. This failure is unfortunate, as a successful opening would have allowed regiocontrol of the subsequent lactonization. As it is, lactonization of 368 yields a 2:l mixture of prevernolepin (367) and prevernomenin (368), which is separated chromatographically and then methylenated using Eschenmoser’s Mannich salt to obtain either vernolepin (334) or vernomenin (354). This process for introducing the methylene groups represents an improvement over the Grieco method, as it does not require prior protection of the free hydroxy group. The overall yield for the bis-methylenation by Danishefsky’s method is 3 I%, compared to 12% in the Grieco synthesis. In all, Danishefsky’s synthesis requires 22 steps from diene 373. Schlessinger’s vernolepin synthesis (Scheme 63) 13’ differs from the Grieco and Danishefsky syntheses because it does not proceed through decalin intermediates. The synthesis begins with construction of the symmetric P-diketone 386, which is protected as the methyl enol ether and reduced to alcohol 387. A feature of this sequence is protection of the ketone in 386 by conversion into its enolate ion while the ester is reduced. Acetal 388 is obtained as a 1:l diastereomeric mixture by the

98

Monocarbocyclic Sesquiterpenes

I . LDA,

&COOEt

I. L D A ,

HMPT

!!!EL

HMPT

383

384

388

E t o o ~ : O E +

H,O

389

385

390

392

39 I

387

,OMe DIBAL,

Me&aH+

H 393

H 394

I . separate 2.m-CPBA 3. MeOACIr NaH

M~

395

I . NaN02.H+ ____)

2. AcpO 3. 60°

Me

396

397

I . PhSH

BF3,Et20 * 2 . CAN 3. Jones

7

-' -+ 367

0 398

Scheme 63.

Schlessinger's Vernolepin Synthesis

Vernolepin and Vernomenin

99

method indicated in the scheme. The intramolecular alkylation of 388 is remarkably stereoselective; one diastereomer of 388 reacts and one does not. The result has been explained on steric grounds. Treatment of the mixture of 389 and 390 with zinc-copper couple in aqueous DMSO gives alcohol 387 and unreacted 389. This specific vicinal elimination is surprising, as it appears that methanol could just as easily have been eliminated from iodo acetal 390, yielding enol ether 399. The ring-B functionality is introduced by reduction of enone 392, which affords

399

alcohols 393 and 394 in a ratio of 3:2. The desired isomer is epoxidized and the secondary hydroxy group is protected to obtain epoxy ether 395. The epoxide ring is opened by dilithioacetoacetate to obtain 396. The successful opening of the hydroxy-protected 395 is an important feature of Schlessinger’s synthesis, as it allows the first regiospecific synthesis of vernolepin. Recall that in Danishefsky’s synthesis, the tetrahydropyranyl ether of 381 does not undergo epoxide opening. The synthesis is completed by a clever degradation of the @-keto ester side chain (Beckman fragmentation) and a novel oxidative conversion of the methoxy tetrahydropyran ring A into the 6-valerolactone ring. The endpoint of the Schlessinger synthesis is the protected form of prevernolepin (398), which can presumably be converted into the natural product by deprotection and bis methylenation. Although the Schlessinger synthesis is the first to achieve regio-control with regard to the ring-B functionality, it is longer than Danishefsky’s (25 steps, allowing 3 steps to convert prevernolepin into vernolepin) and suffers from lack of stereocontrol in two steps 387-388 392-393. Isobe’s synthesis132 is summarized in Scheme 64. Aldehyde 403, obtained from p-methoxybenzyl alcohol by Birch reduction, ketalization, and oxidation, is alkylated with ally1 iodide and the resulting aldehyde is reduced to alcohol 404. Esterification with ethyl malonyl chloride, hydrolysis, and intramolecular Michael addition affords the bicyclic keto

100

Monncrrbocyclic Sesquiterpeaes

. _

400

401

403

402

404

405 /CHO

-

1 . [cHzOH)2, H+ \ 2. 03.CH2C12 3. E?3N

I. NOW4

J Q 2. MsCl

3. PhSeEtOOC

406

407

do

I. ( I - B u o o c ) ~ ~

I ’ 03’-20’

2 . 500

:A

3. H30+

EtOOC

408

2. DBU

EtOOC

409

410

411

I . M~O+SN~

2. m-CPBA

m-CPBA +

412

413

Y

41 4

415

4

4

I . MeOH, OH-

2. Et2NH, CH20

3 . NaOAc, HOAc 334

Scheme 64.

354

Isobe’s Vernolepin Synthesis

416

Vernolepin and Vernomenin

101

lactone 406. Careful monobromination, followed by base-catalyzed cyclization affords the cyclopropyl ketone 407, which is ketalized for the sequence of steps necessary to excise a carbon from the angular group. The resulting angular vinyl compound 410 is subjected to Knoevenagel condensation with di-t-butyl malonate and the double bond isomerized to the P , y isomer 411. The ring-B functionality is now.established by a wonderfully intricate and convoluted sequence involving opening of the cyclopropyl ring with inversion of configuration, Mislow-Evans rearrangement of the resulting allylic sulfoxide and epoxidation, affording 414. Treatment of 414 with sodium cyanoborohydride effects epoxide opening providing the cyclic cyanoborate ester 415. Addition of diborane to 415 leads to a metathesis reaction in which the cyano ligand is replaced by hydride. Subsequent intramolecular conjugate reduction leads to 416. If 415 is treated with excess sodium cyanoborohydride in wet HMPA, the C-7 epimer of 416 is produced, apparently by an intermolecular reaction. The synthesis is completed by hydrolysis of the ester functions, Mannich condensation, and elimination to give a mixture of vernolepin (22%), and vernomenin ( < 5 % ) . The Isobe synthesis is similar to Schlessinger's in that it begins with a 4'4-disubstituted cyclohexanone. It requires a total of 28 steps. As of this writing, other vernolepin syntheses have not been completed. However, a number of model studies and partial syntheses have been published. Since these demonstrate alternative approaches to the problem, they shall be discussed briefly. Clark and H e a t h ~ o c k 'added ~ ~ lithium divinylcuprate to hydrindenone 417 and silylated the resulting enolate to obtain ether 418 (Scheme 65). Selective ozonolysis, reduction of the initial ozonide by sodium borohydride, and lactonization affords 8-valerolactone 419, which is methylenated by Grieco's procedure to obtain the simple ring-A analog 420. An alternative approach by Heathcock and co-workers is summarized in Scheme 66.'34 Alkylation of enol ether 421 by r-butyl bromoacetate, followed by Wittig methylenation yields the dienyl ether 423, which is oxidized by m-CPBA in aqueous ethanol to obtain allylic alcohol 424. Protection of the hydroxy group, conjugate addition of vinylmagnesium bromide, and lactonization yields keto lactone 425.

102

MonocarbocyclicSesgulterpenes

I , (fi12CuLi 2. Me3SiCI

I. 03,MeOH T

M

S

O

a

2. NaBH4

3. H30+

I

41 7

+

0

418

H 419

420

Scheme 65.

Clark-Heathcock Approach to Ring-A W

42 I

O

E

'

Ph3P=CH2 c

422

423

424

I , Me3SiCI,Et3N

2. fiMgBr,CuI 3. H ~ O +

*

J@ H

425

Scheme 66.

Wege-Clark-Heathcock Synthesis of a Vernolepin Precursor

A model study for establishment of the ring-BC functionality was carried out by the Heathcock group and is outlined in Scheme 67.13' Epoxidation of enone 427 is stereospecific, as is reduction of epoxy ketone 428. The hydroxy group is protected and the epoxide ring opened by an

Vernolepin and Vernomenin

$H

I-BuMepSiCI

~

c; 3

;I ’‘b

429

430

43 I

I , Et2AICECSiMeL 2. AgN03,EtOH 3. CN-

432

Scheme 67.

103

433

Heathcock’s Ring-BC Model

acetylenic alane, to give the trans-ethynyl alcohol 431, which is transformed into the trans-fused a-methylenebutyrolactone 432 using Norton’s cyclocarbonylation process. Although the reaction yield is low (only 21%), this approach avoids the necessity of introducing the a-methylene unit at a final stage of the synthesis. Note that this synthesis yields the wrong stereochemistry at the secondary hydroxy position, which must be inverted in order to obtain products of the correct series. However, the stereochemistry at this position is important for the epoxide ring opening. The isomeric epoxy ether 434 does not react, presumably for conformational reasons.

434

104

Monocrrboryclic %spuiterpenes

An approach by Torii, Okamoto, and Kadono of Okayama University is summarized in Scheme 68.'36 It is essentially a streamlined version of Isobe's synthesis, except that the angular vinyl group is introduced from the beginning, rather than by employing an ally1 group as in Isobe's original work. Intermediate 440 could presumably be converted into vernolepin and vernomenin as is 406 in Isobe's synthesis.

42 I

435

436

COOMI

437

439

438

'H

MoOOC

440

Scheme 68.

Tori-Okamoto-Kondo Approach to Vernolepin

Scheffold's approach is outlined in Scheme 69.13' The Scheffold synthesis starts with 3'5-dimethoxybenzoic acid (441)' which is transformed by Birch reduction and alkylation into bromide 443. This substance is transformed into triene 445 by Cope elimination of amine oxide 444. Reduction of the carboxy group and hydrolysis of the enol ether functions yields the symmetrical diketone 446, which is similar to an intermediate in the Schlessinger synthesis (387, Scheme 63). Zutterman, de Clerq, and Vandewalle have explored a similar approach outlined in Scheme 70.'38 The Birch reduction product of benzoic acid

Vernolepin and Vernomenin

105

serves as the starting material, leading eventually to the diazomalonate 451, which is converted into the cyclopropane 452. Opening of the cyclopropane ring, as in the Isobe approach, affords 453.

H

o

o

Me

c

I . Na,NHL CH2N2

~2.

Meoocy -

OMe

44 I

442

443

M e O IO 4T M e

0

OMe

444

Scheme 69.

450

445

446

Scheffold’s Approach to Vernolepin

451

452

NOMI

453

Scheme 70.

Vandewalle’s Approach to Vernolepin

106

Monocarbocyclic Sesqulterpeaes

454

455

456

I

I Nan

\

OSiRJ

/

+

1 OSiR3

4 57

458

(n-BuI4N+F-

0&COOMe dSiR3

0

4 H

I. LDA 2. 3 Me2N+=CH MbI

4 . NoHC03

463

462

0 464

Scheme 71.

Marshall-Flynn Synthesis of a Vernolepin Isomer

Fraxinellone and Pyroangolensolide

107

Marshall and Flynn have reported a synthesis (Scheme 71) which leads * ~ ~ monoketal to compound 464, a positional isomer of ~ e r n o l e p i n . The of 1,4-~yclohexanedioneis bis alkylated by way of its piperidine enamine to obtain the symmetrical dialkylated ketone 455. Reduction of the ketone of 455 is stereospecific; after deketalization and protection, the symmetrical intermediate 456 is produced. This intermediate is converted into a P-keto ester, which is alkylated stereospecifically from the face opposite the vicinal chloroallyl group to obtain diester 457. Reduction with lithium aluminum hydride at low temperature gives a 7:3 mixture of isomers 458 and 459. The mixture is separated by treatment with triethyl orthoacetate; the major isomer forms a cyclic orthoacetate (4601, while the minor isomer does not react. The resulting mixture of 460 and 459 is easily separated and the orthoacetate 460 is hydrolyzed to obtain pure 458. The primary alcohol is dehydrated by Grieco’s method, oxidative elimination of the o-nitrophenylselenide. After deprotection of the hydroxy group, the y-butyrolactone ring spontaneously forms and the resulting dilactone 463 is bis methylenated by Danishefsky’s method. Several attempts to transform 464 into vernolepin by transposition of the hydroxy group were not successful.

L. Fraxinellone and Pyroangolensolide Fraxinellone (465) and pyroangolensolide (466) are probably not actual sesquiterpenes, but metabolic products related to the diterpene limonin (467). Nevertheless, the syntheses of 465 and 466 which have been

a Q

po H

465

466

467

108

Monocarbocyclic Sesqulterpenes

reported are typical of work in the sesquiterpene area and will be discussed in this section. Tokoroyama and associates have reported syntheses of both compounds. Their fraxinellone synthesis is summarized in Scheme 72.140 Treatment of methyl vinyl ketone with ethoxyethynylmagnesium bromide affords a 5:4 mixture of diene esters 468 and 469. The mixture is heated with methacrolein to obtain adduct 470 and unchanged cis ester 469 (which may be isomerized to an equilibrium mixture of 468 and 469 and recycled). Aldehyde 470 is treated with 3-furyllithium (prepared from 3-bromofuran) to obtain a 7 5 mixture of lactones 471 and 472. Chromatographic separation and double-bond isomerization yields ( f 1-fraxinellone (465). + EtOCPCMgBr 0

-

,L i

470

3%q - C O O E t

c 47 I

468

+

7’ /

469

472

465

Scheme 72.

Tokoroyama’s Fraxinellone Synthesis

A synthesis of the somewhat more challenging target pyroangolensolide is outlined in Scheme 73.14’ The synthesis begins with the hydrindanone 474, which is prepared from 2,6-dimethylcyclohexanone by a Wichterle-type annelation. Two successive lead tetraacetate oxidations of 474 serve to cleave the cyclopentenone ring and establish the hemilactol

Irangulin, Eriolanin, and Phytuberin

473

+o

Pb(OAc), HOAC, H 2 0

~

109

474

&o

4_ NBS CHCl3 CCI

@ 0

475

476

466

Scheme 73.

477

470

Tokoroyama's Pyroangolensolide Synthesis

functionality. Bromination of 476 with N-bromosuccinimide in a mixture of CCI, and CHCI, proceeds with concomitant elimination of HBr, giving the diene 477 in 64% yield. Treatment of this substance with excess 3-furyllithium gives a 7:3 mixture of pyroangolensolide (466) and its stereoisomer 478.

M. Ivangulin, Eriolanin, and Phytuberin Ivangulin (4791, eriolanin (4801, and phytuberin (481) are secoeudesmanes which have been synthesized in the period of this review. The principal challenge in the synthesis of ivangulin and eriolanin is relating the stereochemistry of the methyl group located in the acyclic portion to the stereocenters in the six-membered ring. A similar problem exists in the synthesis of phytuberin, as can be seen in the openchain form 481a. In each case, the problem has been solved by first establishing the relative stereochemistry in an intermediate decalin, followed by cleavage of one ring.

110

Monocrrbocyclic Sesquiterpenes OH

0. 481

479 480

OHC *OH

Grieco's ivangulin synthesis is summarized in Scheme 74.'42 Simmons-Smith methylenation of the homoallylic alcohol 483 occurs from the (Y face of the molecule. Electrophilic cleavage of the cyclopropane establishes the methyl group required for the eventual side chain. The relative stereochemistry of this center is governed by the thermodynamic preference for the p (equatorial) configuration. The elements of the eventual y-butyrolactone ring are introduced in an interesting way, by [2+21 cycloaddition of dichloroketene to diene 486. The addition is regio- and stereospecific, giving, after dechlorination, only cyclobutanone 487. After a series of straightforward functional group manipulations, the A ring is opened by Baeyer-Villiger oxidation, giving hydroxy ester 489. The tertiary alcohol is eliminated and the resulting unsaturated epoxide 490 is reduced with lithium in ammonia. Unfortunately, the desired 491 and 492 are produced in a ratio of 5:4. However, the major product may be converted into alkene 492 by acetylation and reduction. The only remaining noteworthy step is the BaeyerVilliger conversion of the cyclobutanone ring into the y-butyrolactone ring, which is accomplished using alkaline t-butyl hydroperoxide. The synthesis, although long (27-29 steps), is marked by two inventive

Irangulin, Eriolanin, rnd Phytuberin

111

maneuvers, introduction of the side-chain methyl group as a cyclopropane ring, and the method used to establish the y-butyrolactone funct ionali ty.

462

403

404

PhpI. C12CHCOCI

Ph-

I , p-TsNHNH2

2. L D A , T H F 3. n20

Ph-

Et3N

b

2 . Zn,HOAcb

465

468

H+ 2 . L i , NH3

3 . Collins

H

I . m-CP6Ab 2 . K2C03, MIOH

407

MeOOC

469

466

I. A c g

MeOOC

no

HO

491

490

492

494

493 I. H$+

2. Jones 3. CH2N2 4. DBU

*

MaOOC

419

Scheme 74.

Grieco's Ivanqulin Synthesis

Monocrrkyclic Sesquiterpeaee

112

Grieco's eriolanin synthesis (Scheme 75) is a modification of the ivangulin synthesis. 143 It begins with alcchol 483, which is converted into diketone 499 by a sequence of operations similar to that employed in the ivangulin synthesis. It was hoped that both Baeyer-Villiger oxidations and the epoxidation could be combined into one step. However, this did not prove possible, as peroxyacid oxidation affords the wrong epoxide (see 488-489 in Scheme 72). Nevertheless, the four-step sequence 499-501 was found to give the epoxy dilactone with good selectivity. The remainder of t h e synthesis is straightforward, although n?

n?

I. CM212 Zn(Cu) 2 . MC104* CM2C12

R3Si?

I .

483

496

wo 498

ci2cncoci El

I . p-TlNHNH 2 . n-BuLi 3. I - B u M ~ ~ S I C I

N

3 2 . In,HOAc 3. n30+

49?

499

500 I . n30+

m-CPBA+

&

* 501

R3Si0

0.

503

502

504

tYr),o 1.

HCOOH

2 . On-

OHCO

505

0

HO

506

480

Scheme 75.

Grieco's Eriolanin Synthesis

Irangulin, Eriolanin, and Phytuberin

113

the events accompanying the epoxide opening are of some interest. The transformation is carried out with formic acid in chloroform, which also removes the t-butyldimethylsilyl group from the side-chain hydroxy, which is subsequently formylated. The secondary alcohol, which undergoes formylation more slowly, is thus left for conversion to the methacrylate ester. The synthesis requires a total of 27 steps from the Wieland-Miescher enedione (482).

T. Masamune and co-workers, of Hokkaido University in Sapporo, have synthesized phytuberin (4811, beginning with a-cyperone (507, Scheme 76).’44 Oxidation of the kinetic enolate of 507 under Vedejs’ conditions, followed by protection of the resulting acyloins, gives an inseparable mixture of keto ethers 508 and 509. Lithium

I . LDA 2. MOOS 3. I-BuMe2SiCl*

0

508

507

Li(s-Bu)3BH or LiAlH4

*

% s i-i

+

I . I-Bu00H VO(ocadp 2. 3. MICl Li-NH3

b s & ;

HO

510

512

Scheme 76.

509

511

4. H,O+ 5. M i 0 2

513

Masamune’s Phytuberin Synthesis

514

114

Monocrrbocyclic Sesqulterpenes

tri- sec-butylborohydride selectively reduces 508 to 510, leaving 509 unchanged. The mixture of 509 and 510 is separated and 509 is reduced by lithium aluminum hydride to obtain 511. Both 510 and 511 are epox-

idized with Sharpless' reagent, the epoxy alcohols mesylated, and the epoxy mesylates reduced under Birch conditions. After deprotection of the C, alcohols, and oxidation of the resulting allylic alcohols, (+)-p-rotunol (512) is obtained. The six-step sequence from 508 or 509 to 512 proceeds in about 50% overall yield. The A-ring is cleaved after a second application of the Vedejs oxidation, affording the unsaturated lactone 515. Reduction of the two carbonyl groups and dehydration yields phytuberin. The multistep sequence for converting 507 into 512 is necessary because diene 516, obtained from 507 in 72% yield via the Bamford-Stevens reaction, reacts with singlet oxygen from the a-face, yielding a-rotunol (517, 88%) after rearrangement of the resulting peroxide. The synthesis requires 14 steps overall, beginning with a-cyperone. 2I . MoLi P-TSNHNH~

q+

&&

*LO

517

516

N . Hedycaryol, Preisocalamendiol, Acoragerrnacrone, Costunolide, Dihydrocostunolide, Dihydroisoaristolactone, and Periplanone B

The germacrane class of sesquiterpenes continues to present one of the most difficult challenges to synthesis. In the last ten years, seven compounds have been synthesized: hedycaryol (5181, preisocalamendiol (519), acoragermacrone (5201, costunolide (5211, dihydrocostunolide (522), dihydroisoaristolactone (5231, and periplanone B (524).

519

520

52 I

Hedycarol, Preisocnlamendiol, Acoragermacrone, etc.

522

523

115

524

Hedycaryol (518) was first synthesized in 1968 by Wharton and cow o r k e r ~ . The ' ~ ~ synthesis (Scheme 77) starts with dimethyl 4-hydroxyisophthalate ( 5 2 9 , a readily available by-product of the manufacture of salicylic acid. Hydrogenation of the ring followed by acetylation gives a mixture of isomeric triesters which is heated at 260°C to obtain unsaturated diester 527 in 32% overall yield for the three steps. Diester 527 I , H2-Ru02

t

MsOOC

COOMe 2' "2'

525

COOMe

526

& & kH & MeOOC

XOOMe+ 530

532

OH

533

535

Scheme 77.

528

529

I .LIAIH~

H

2. NoI04

534

$

H

53 I

1-3

+

OH

HO

527

__e P-T~CI

H

518

Wharton's Synthesis of Hedycaryol

116

Monocrrbocyclic Sesquiterpenes

reacts with methyllithium to give a diol (528), which is partially dehydrated by heating in dimethyl sulfoxide. Diels- Alder addition of diene 529 to methyl a-acetoxyacrylate affords a mixture of diastereomeric adducts (5301, which is reduced and cleaved to obtain a mixture of octalones 531. Methylation of the thermodynamically-favored enolate derived from 531 affords the trans- and cis octalones 532 and 533 in a ratio of 4 : l . The mixture is reduced and the major reduction product (534) converted into tosylate 535. The decalin ring is fragmented by Marshall’s method, giving hedycaryol (518) in 60% yield. Ito has prepared hedycaryol starting with (2E16E)-farnesol as shown in Scheme 78.146a Oxidation of the terminal double bond of sulfide 537 by van Tamelen’s method gives epoxide 538 which undergoes a remarkably efficient cyclization to a 7 5 mixture of cyclodecadienes 539 and 540. After chromatographic separation, the major isomer (539) is subjected to reduction by lithium in ethylamine to obtain an inseparable mixture of 541 and 518 (2:3 ratio). Thermolysis of the mixture affords trans-elemenol (542) stereospecifically. Although the sequence is plagued by lack of specificity in two steps, the direct cyclization of 538 to 540 and 539 is nevertheless an impressive reaction that could point the I . NBS

PhS 638

SPh

538

537

518

Scheme 78.

542

116’sConversion of Farnesol into Hedycarol

Hedycarol, Preisocalamendiol, Acorrgermrcrone, etc.

117

way to more effective germacrane syntheses in the future. In a subsequent investigation, Ito and his collaborators have studied the conversion of (2E,6Z)-farnesol (543) into the (6Z)-hedycaryols 544 and 545.146b

-w+q

k+ OH

543

544

545H

W. C. Still has reported the concise synthesis of preisocalamendiol (519) and acoragermacrone (520) outlined in Scheme 79.147 The (546) is converted into the monoterpene isopiperitenone truns-l,2-divinylcyclohexenol547, which is subjected to oxy-Cope rearrangement under Evans’ conditions to obtain isoacoragermacrone (548). Formation of the kinetic enolate of 548, followed by acetic acid quench, yields acoragermacrone (519) in only three steps from isopiperitone! The cis double bond in 548 is isomerized to the less stable trans geometry of 520 by an ingenious method involving conjugate addition of trimethylstannyllithium, followed by trapping of the resulting enolate with trimethylsilyl chloride to obtain enol ether 549. Mild oxidation of 549 yields preisocalamendiol (520). The reader is referred to the original article for a discussion of the rationale for the isomerization of 548 to 520.

546

54 7

548

i

I . Me3SnLi 2. MefSiCl

520

549

Scheme 79. Still’s Preisocalamendiol and Acoragermacrone Syntheses

519

Monocarbocyclic Sesquiterpenes

118

Grieco's synthesis of costunolide is recorded in Scheme 80.14*The well-known keto lactone 550, obtained in two steps from santonin, is subjected to a sequence of reactions more or less identical to that employed in the Grieco temsin synthesis (see Scheme 58, 348-3521 to obtain (554). Thermolysis of 555 is accomplished by injection into a gas chromatograph with an oven temperature of 200°C. An equilibrium mixture is produced from which 521 and 555 may be obtained by preparative tlc in 200/0 and 42% yield, respectively. This appears to be a rather unusual example of a cyc1odecadiene.-divinylcyclohexane equilibrium in which the equilibrium does not overwhelmingly favor the sixmembered ring isomer.

300

I.

550

55 I

I.

O3 I H

553

554

Scheme 80.

555

52 I

Grieco's Synthesis of Costunolide

Fujimoto, Shimizu, and Tatsuno converted santonin into dihydrocostunolide as shown in Scheme 81.'49 Chloro epoxide 556, obtained in three steps from santonin (30% yield), is reduced by zinc to diene 557,

Hedycnrol, Prelsocrlnmendiol, Acorngermncrone, etc.

300

55s

561

Scheme 8 1.

587

556

558

119

522

Fujimoto-Shimizu-Tatsumo Synthesis of Dihydrocostunolide

which is photolyzed. The initially formed cyclodecatrienol 558 isomerizes to the cyclodecadienone 559, which is obtained as a 5:l:l mixture of stereoisomers. Hydrogenation of the more accessible double bond of the major isomer affords a keto lactone (5601, which is reduced and dehydrated to obtain dihydrocostunolide (522). The elimination of the secondary mesylate derived from 561 is noteworthy, in that DBU and other amine bases are ineffective. The only base which was found to accomplish the desired elimination was tetra-n-butylammonium oxalate (TBAO) . Lange and McCarthy have investigated the use of the photo adduct 564, obtained from cyclobutenecarboxylic acid and (-)-piperitone (563), for preparation of germacranes (Scheme 821.'' Reduction of the ketone carbonyl of 564 gives a hydroxy ester which spontaneously lactonizes to give 565. Thermolysis of this compound affords the dihydroisoaristolactone 523 (see also Scheme 57).

120

Monocrrbaeyclic Sesquiterpeaes

562

563

564

565

523

Lange-McCarthy Synthesis of Dihydroisoaristolactone

Scheme 82.

I . Me3SnLi

h

2. A c p

2. Me3SiCI RO/

I . Me2CuLit

"83'"-

2. m-CPBA OAc 568 I . +BuMepSiCI

= 570

I . eMgBr

6

'OR

RoO

RO'

OAc

567

566

4 . m-CPBA

&

569

I-BKUHOOH:

571

4 . H20,

573

Scheme 83.

572

524

Still's Periplanone B Synthesis

Hedycnrol, Preisocalarnendiol, Acoragerrnacrone, etc.

121

Periplanone-B, the sex excitant of the American cockroach, was isolated (200 pg) and characterized in 1976 as a germacrane of structure 524, but of unknown stereochemistry. In 1979, Still reported a brilliant series of syntheses in which he prepared three of the four possible diastereomers and determined the correct stereochemistry for the phermone.151 The synthesis of the actual natural product is summarized in Scheme 83. Aldol condensation of cyclohexenone 566 with crotonaldehyde affords, after acetylation, adduct 567. The enone is protected by addition of trimethylstannyllithium and the derived trimethylsilyl enol ether 568 is alkylated by lithium dimethylcuprate. After regeneration of the cyclohexenone, vinylmagnesium bromide is added and the resulting 1,2-divinylcyclohexenol is subjected to oxy-Cope rearrangement. The crude reaction product is silylated and the silyl enol ether oxidized to obtain a-hydroxy ketone 570. The secondary hydroxy group is protected and the primary hydroxy deprotected by mild acidic hydrolysis. Dehydration of the primary alcohol is accomplished by Grieco’s method, elimination of the derived o-nitroselenoxide. The resulting enone 571 is epoxidized to 572 and the ketone methylenated using Corey’s reagent to obtain bis-epoxide 573. Deprotection of the hydroxy group and oxidation affords periplanone-B. The stereochemical outcome of both the epoxidation and methylenation steps (571-572, 572-5731 is understandable in terms of preferred conformations for the cyclodecadienone (571a) and epoxycyclodecenone (572a) which suffer only peripheral attack. It is assumed that the inherent preference of a 1,3-butadiene for

OR

111a

572 a

the s-trans conformation plays an important role in assuring that 571 and 572 adopt the required conformation shown. When the oxidation and methylenation steps are carried out on 574 (the protected form of intermediate 570) isomer 575 is produced, presumably as a consequence of another preferred conformation.

122

Monoearbocyclic Sesquiterpenes

171

874

0. Humulene Vig and co-workers have reported a synthesis of humulene (585) which utilizes nickel carbonyl coupling of a bis-allylic bromide to form the eleven-membered ring (Scheme 84)."* Although Corey had earlier used

578

579

I

580

581

582

583

585

Scheme 84.

Vig's Synthesis of Humulene

584

Humulene

123

this method in the first humulene synthesis,lS3 the Corey synthesis proceeds through the 5 Z stereoisomer of 584 and therefore yields 4,5-cis-humulene1 which must then be isomerized to humulene. The Vig synthesis yields the correct stereoisomer without necessity of equilibration and is also much more efficient (8 steps rather than 14). Yamamoto and co-workers have reported an extremely interesting synthesis of humulene (Scheme 85).lS4 Bromo-ester 586, obtained in two steps from geranyl acetate, is used to alkylate the dianion of ethyl 2-methyl-3-oxobutanoate to obtain keto diester 587. The sodium salt of 587 is cyclized by Trost’s method to obtain cyclic P-keto ester 588 in 45% yield. The carbonyl groups are reduced and the resulting diol converted into oxetane 589 by way of a monotosylate. The second key reaction of this imaginative synthesis is base-catalyzed ring openivg of the oxetane, which yields trienol 590. The primary alcohol is deoxygenated in a straightforward sequence of steps to obtain humulene.

V

O

E

t

-

+

AC

586

588

Scheme 85.

Meoocm OAc

587

589

Yamamoto’s Humulene Synthesis

124

4.

Bicarbocycllc Sesquiterpenes; Hydronaphthalenes

BICARBOCYCLIC SESQUITERPENES; HY DRONAPHTHALENES

The largest single group of sesquiterpenes from the standpoint of syntheses which have been reported, is the hydronaphthalene group. In this section, we consider this group of compounds, further subdivided into four major classes: eudesmanes (l), cadinanes (2), drimanes (3), and eremophilanes (4). Finally, we consider a few syntheses of

cbyq I

2

@cp, 4

3

miscellaneous hydronaphthalenes containing an additional cyclopropane ring. A. Eudesmanes (1) Occidol, Emmotin H, Rishitinol, and Platphyllide

The first group of eudesmanes which we now consider is the rearranged ring-A aromatic compounds occidol (5), emmotin H ( 6 ) and rishitinol (7) and the desmethyl aromatic eudesmane platphyllide (8).

I

6

7

Eudesrnanes

125

The simplest of the group is occidol, in that there is no stereochemical problem and no real skeletal challenge. Three different syntheses, all by T.-L.Ho, have been reported (Scheme The first synthesis begins with 5,8-dimethyl-cr-tetralone (91, requires three steps, and affords occidol in 45% overall yield. The second synthesis, published a year later, begins with 3,6-dimethylphthalic anhydride, requires four steps, and provides occidol in only 7% overall yield. The key step is Diels-Alder cycloaddition of the o-quinodimethane derived from a,a'-dibromoprehnitene and methyl acrylate. A year later, Ho reported

9

5

10

I2

II

5

13

@

COOMe

I5

Scheme 1.

ahoH 14

5

The T.-L. Ho Occidol Syntheses

126

Bicarbacyclic Sesquiterpenes; Hydronaphthalenes

his third synthesis, which begins with the Diels-Alder reaction between benzoquinone and butadiene. This synthesis requires six steps, and gives occidol in about 14% overall yield. Emmotin-H is interesting for its novel 1’2-naphtho quinone structure. Krishna Rao has recently reported the straightforward synthesis summarized in Scheme 2.’’’ Friedel-Crafts cyclization of diacid 16 provides a mixture of keto acids 17 and 18 from which 17 may be crystallized in 63% yield. Oxidation of the derived methyl ester with selenium dioxide gives o-quinone 19, which is reduced and acetylated to obtain 20. Treatment of triester 20 with excess methylmagnesium bromide, followed by air oxidation, gives emmotin-H (6).

COOH

\

- @, PPA

COOH

+wH \

18

I7

16

I . separate

2. MeOH, H +

OOMe

3. seop

COOMe

20

19

6

Scheme 2.

Krishna Rao’s Emmotin-H Synthesis

Tomiyama and co-workers have synthesized rishitinol, an interesting hydroxy occidol isolated from fungus -infected white potatoes. lS9 The synthesis is outlined in Scheme 3. Successive alkylation of malonic ester with benzylic chloride 21 and ethyl chloroacetate affords a triester (23),

Eudesmanes

NaCH(COOE1) OOEt

127

Na CI~COOEt

22

21

&

"

%.

I . NaOHc

2. HsO'

@

OH

3. Ac20

OOEt

18

24

23

+ OoH

2. EtOH, H+ 3. NaBH,

OOEt

25

MWI,

I . separate

2. P205 (on257

OOEt

26

+& OH

27

28

7

Scheme 3.

Tomiyama's Synthesis of Rishitinol

which is converted into anhydride 24. Friedel-Crafts cyclization affords tetralone 17 and indanone 18 in a ratio of 2:1, exactly as was later found by Krishna Rao for cyclization of the succinic acid 16 (see Scheme 2). The tetralone 17 is separated by fractional crystallization, esterified and reduced to obtain a mixture of hydroxy ester 25 and lactone 26. Dehy-

128

Bicarbocyclic Sesquiterpenes; Hydronrphthaienes

dration of the former compound, followed by addition of methylmagnesium iodide, yields dehydrooccidol (281, which undergoes regioselective and stereoselective hydroboration to afford rishitinol. Bohlmann's synthesis of the nor-eudesmane platphyllide is outlined in Scheme 4.I6O It begins with 1-naphthoic acid, which is hydrogenated and converted into amide 30. This substance is converted into lactone 31 by oxidation with lead tetraacetate (22% yield). After methanolysis of the lactone, the hydroxy group is oxidized and the isopropanol side chain introduced by Mukaiyama's titanium tetrachloride mediated aldol condensation. After reduction of the keto ester and reoxidation of the primary alcohol, lactones 37 and 38 are obtained in a ratio of 2:l. The minor isomer is dehydrated to obtain platphyllide.

30

29

31

I . LDA

~

2. Me3SiCI MeOOC

MeOOC

OH

32

MeOOC

0

33

OTMS 34

__c

MeOOC

HO

0 35

37

Scheme 4.

38

36

H

OH

8

Bohlmann's Synthesis of Platphyllide

Eudesmanes

129

(2) a -Cyperone, /3 -Cyperone, P -Eudesmol, and /3 -Selinene

a-Cyperone (39) has a stereochemical complexity relating the isopropenyl substituent to that of the angular methyl group. P-Cyperone (40) is even simpler, as there is only one stereocenter. P-Eudesmol (41) and P-selinene (42) have three stereocenters and present somewhat more complex problems of synthesis. Extensive early work was directed at these simple eudesmanes and is reported in Volume 2 of this series. The new syntheses have involved novel approaches to the carbon skeleton.

39

40

41

42

Although it only contains two stereocenters, a-cyperone (39) presents a stereochemical problem since simple Robinson annelation of 2,s-disubstituted cyclohexanones affords mainly the diastereomeric octalone having the two substituents both axial:161

Caine and Gupton provided an interesting solution to this problem, shown in Scheme 5.16* Michael addition of (-1-2-carone (43) to ethyl vinyl ketone affords diketone 44. Stereoselectivity is good because of the gem-dimethyl group which hinders attack from the top face of 43. Treatment of 44 with ethanolic HCI causes ring opening of the cyclopropyl ketone and aldol cyclization. The three-step synthesis provides 39 in 46% overall yield and is much more efficient than the earlier Pinder synthesis. 16'

130

Bicsrbocyclic Sesquiterpenes; Hydronsphthrlenes

44

43

45

39

Caine-Gupton Synthesis of a-Cyperone

Scheme 5.

Gammill and Bryson have reported the synthesis of P-cyperone outlined in Scheme 6.163 The synthesis begins with enone 47, available in two steps from dihydroresorcinol (46). Enone 47 is alkylated with dibromide 48, a homolog of the familiar Wichterle reagent, to obtain enone 49, which is treated with isopropylmagnesium bromide under copper catalysis, to obtain 50. Hydrolysis of the vinyl bromide and aldol cyclization occur concomitantly upon treatment of 50 with perchloric acid in refluxing formic acid; P-cyperone is obtained in about 18% overall yield for the five-step synthesis beginning with 46.

no,

% Br

2 steps

0

+

46

AMaBr* CuCN

0

so

Scheme 6.

SBU

SBU ::L+r+

9-;b,.:i;+ 47

49

48

o& 40

Gammill-Bryson Synthesis of p-Cyperone

Eudesmanes

131

Huffman and Mole have reported a synthesis of P-eudesmol which does not involve Robinson annelation as the method of constructing the decalin nucleus (Scheme 7).'64 The synthesis begins with keto acid 52,

8- --L *qQ a 0

OOH

CHO

Me0

OOH

POOH MOO

Me

52

bl

I . Li/NH3

2. H$+

3. reparate

53

Me2CuLi

COOH

Ph,P=CH2 QMSO

55

54

2. MOO-, MeOH

3. OH-,MeOH

56

Scheme 7.

57

Huffman-Mole Synthesis of 8-Eudesmol

available in four steps (22% yield) from o-anisaldehyde. After Clemmensen reduction, the methoxy acid 53 is subjected to Birch reduction to obtain a complex mixture of acids. The desired unsaturated keto acid 54 is obtained in 25% yield after chromatography. Addition of lithium dimethylcuprate to 54 gives a mixture of three keto acids which is subjected to Wittig methylenation to obtain acids 56. This mixture of acids is esterified with diazomethane (mineral acid causes isomerization of the double bond) and the resulting mixture of methyl esters is equilibrated by sodium methoxide in methanol. Saponification affords methylene acid 57, which had previously been converted into P-eudesmol. The overall yield for the 11 steps from aldehyde 5 1 to acid 57 is only 1.4%, even though the authors advertise the synthesis as short and efficient.165

132

Bicarbocyclic Sesquiterpenes; Hydronaphthalenes

Carlson and Zey have reported a similar approach to the eudesmanes (Scheme 8) .166 Stobbe condensation of di-t-butyl glutarate and keto ester 58 affords 59, which is deprotected and dehydrated to obtain 60. Cyclization affords keto acid 54. The overall yield for the three-step synthesis of 54 (35% from keto ester 58) is superior to that employed by Huffman and Mole (Scheme 7, 5 % from aldehyde 51). Addition of lithium dimethylcuprate to enone 61 provides a 48:24:17:11 mixture of the four diastereomers of 62, with the desired stereoisomer being the least abundant. Base-catalyzed equilibration changes the initial ratio to 14: 10:18:58. Saponfication and chromatography gives pure keto acid 63, which has previously been converted into P-eudesmol. 16’

59

58 COOH ___t

p-TaoH t o Iuenr

I . PPA OOMe 2. H 3 0 +

HOOC%

60

CH2N2 *

*

P

C

O

O

H

0 54

MrpCuLi COOMe

0

61

I. NoOMI-MIOH

P

O

O

M

e

2. KOH-H20 MeOH

P

62

Scheme 8.

O

O

H

63

Carlson-Zey Synthesis of p-Eudesmol

Wolinsky and co-workers have synthesized p-selinene as outlined in Sc’leme 9.16* Unsaturated chloride 65, obtained in two steps from limonene, is acylated with vinylacetyl chloride (66) to obtain a mixture of enone 67 (34%) and enone 68 (33%). After separation of the mixture, enone 67 is hydrogenated, and dehydrochlorinated to obtain ketal

Eudesmnnes

67

65

64

133

%

68

69

I-BuOKe OMSO

v Scheme 9.

Ph~P3CH2e 42

71

70

Wolinsky’s Synthesis of p-Selinene

70. Hydrolysis followed by Wittig methylenation affords j3-selinene (42).

Stereochemically, ketone 71 is obtained as a 9:l mixture with its

72

diastereomer 72. The relative stereochemistry of the side chain and the angular methyl center is set in the cyclization leading to 67: or

1-

The other enone (681, which is suggested to be formed by carbonium ion rearrangements, was converted into occidol.

134

Blcarbacyclic Sesquiterpenes; Hydronrphthnlenes

(3) Juneol, 10-Epijuneol, and 4-Epiaubergenone

This group of eudesmanes is slightly more complex than those considered in the last two sections in that each has four stereocenters. Although both juneol (73) and lO-epijuneol (74) are natural products, 4-epiaubergenone (75) has not itself been found in nature.

73

74

75

Schwartz and co-workers have synthesized juneol from methyl farnesate (76) as outlined in Scheme Cyclization of 76 is brought about using Lucas' reagent, to obtain a mixture consisting mainly of chloro ester 77. Treatment of this mixture first with aqueous ZnCI,, then with ZnCl, in refiuxing benzene affords a mixture of diastereomeric esters 78, accompanied by 16% of a double bond isomer

I . ZnC12,

COOMe

COOMo

77

76

78

80

79

81

A

82

Scheme 10.

Schwartz Synthesis of Juneol

73

Eudesmrnes

135

and 17% of a bicyclic isomer. The mixture of isomers 78 is reduced to obtain a 1:l mixture of diastereomeric alcohols 79. After chromatographic separation, each isomer is oxidized to the corresponding aldehyde. Cyclization of the two aldehydes is brought about by treatment with stannic chloride in benzene. Each isomer undergoes ring closure with retention of stereochemistry at the formyl group. However, each yields a mixture of stereoisomers. The desired isomer 82 is obtained in 18% yield by chromatography of the cyclization product from isomer 80. Selective hydrogenation of the isopropenyl double bond and photoisomerization of the endocyclic double bond yields ( f) -juneol. Although this synthesis produces the desired natural product, it leaves a great deal of room for improvement in selectivity. Wender and Lechleiter have carried out an elegant synthesis of 10-epijuneol which is outlined in Scheme 11.170 Photoaddition of methyl cyclobutenecarboxylate (83) to piperitone (84) provides adduct 85. Deoxygenation by the Bamford-Stevens reaction gives alkene 86, which is reduced, hydroborated, and oxidized to obtain diol 87. This substance is selectively phosphorylated on the primary hydroxy group and the resulting monophosphate reduced by lithium naphthalenide to obtain 10-epijuneol (74). The synthesis is short, efficient, and completely stereoselective.

I . p-TsNHNHZ 2. NoH.PhMe

+ qCOOM~

83

MeOOC

85

84

87

Scheme 11.

74

Wender-Lechleiter Synthesis of 10-Epijuneol

136

Blcnrbocyclic Sesquiterpenes; Hydronaphthalenes

R. B. Kelly and co-workers at the University of New Brunswick synthesized 4-epiaubergenone, an isomer of the natural product, as outlined in Scheme 12.”’ Ketone 88 is subjected to a straightforward sequence of steps to obtain racemic dihydrocarissone (91), which is brominated and dehydrobrominated to obtain racemic enone 75. The optically active form of this material is obtained by base-catalyzed equilibrium of natural aubergenone.

@

/dsb””

I . Li-NH3, ElOH ~

I . IEiO)ZCO, NaH

2 . NaH,

0

PhCH2CI

3.H30+

88

2. NoBH, 3. POCI,, C5H5N

Ph

89

I . Hz-Pd/C

Ph-0

2. MeLi ‘ H

‘ H

90

91

I . PhMe3N+Br; 2. L i B r , L i 2 C 0 3 , DMF

75

Scheme 12.

Kelly Synthesis of 4-Epiaubergenone

(4) Cuauhtemone

Cuauhtemone (98) is a sesquiterpene growth inhibitor which is isolated from a Mexican medicinal shrub. It has been synthesized by Goldsmith and Sakano as shown in Scheme 13.’72 Monoketal 88 (see also Scheme 12) is reduced and the resulting enolate is deoxygenated by Ireland’s method to obtain unsaturated ketal 92. Hydrolysis and acylation with diethyl carbonate affords a P-keto ester (93) which is hydroxylated to obtain a 3:l mixture of diastereomeric diols, with 94 the major isomer.

Eudesmrnes

88 I.

92

137

93

oso4

MeLi

2. Separate

HO"

H0 ' 94

96

95

97

98

Goldsmith Synthesis of Cuauhtemone

Scheme 13.

The enone system is introduced by addition of methyllithium to the salt of the protected P-keto ester 95, followed by dehydration of the resulting 0-hydroxy ketone.*73 Isomers 96 and 97 are produced in a 2:l ratio. Deprotection of 96 affords cuauhternone (98). (5)

p -Agarohran, Norketoagarofuran, and Evuncfler Ether

P-Agarofuran (99), norketoagarofuran (loo), and evuncifer ether (101) are related structurally in that each is a cyclic ether.

qq@Q 0

99

00

100

I

H

101

Bic8rbacyciic Sesquiterpenes; Hydronrphthrlenes

138

Bifchi and Wuest have reported the synthesis of /3-agarofuran summarized in Scheme 14. Dihydrocavone (102) is first hydrated and then hydrogenated to obtain hydroxy ketone 103.'74 Alkylation with 5-iOdO1-pentyne affords a mixture of diastereomers from which the major isomer 104 may be obtained in 41% yield by crystallization. Treatment of

)4,

0

I . H$O,-H& 2. HO-?I/C

102

-%6

:2** #

H0*

%OH

103

104

I . EtMgBr 2. Ma3SiCI

CI*

n-Bu$nH+ Mo3Si0

105

Mo3Si

106

R I07

Scheme 14.

p-MoC6H4SO3H

MaCN-H20

c

(@ 99

Buchi and Wuest Synthesis of 8-Agarofuran

104 with PCI, provides the chloro ether 105. After silylation of the terminal acetylene, the A-ring is closed by free radical cyclization. Desilylation affords p-agarofuran (99). This imaginative synthesis is one of the rare examples of the use of free radical chemistry in natural product synthesis. In Volume 2 of this series, we reported a synthesis of norketoagarofuran by Heathcock and Kelly.'75 Although the synthesis was brief, the final steps turned out to be capricious and irreproducible. Kelly subsequently reported an alternative, reproducible route (Scheme 15) .176 Treatment of unsaturated acid 108 with peroxyformic acid yields a mixture of diol monoformates which is saponified to obtain dihydroxy acid

Eudesmiines

&).

.

I . HCOSH 2. OH-

‘COOH

Hb OH

108

‘.

NaOMr,,

dioxanr

A

COOMe

a

.

Jonra

‘COOH

I10

109

I. (CH20H)2, H+ 2. MrO-,MrOH

-

139

2.

H+

I . MeLi

112

Itt

too

Scheme 15.

Kelly’s Synthesis of Norketoagarofuran

109. Jones oxidation gives a ketol (110) which is ketalized and the resulting ethylene glycol ester trans-esterified to obtain 111. Treatment

of this hydroxy ester with sodium methoxide in refiuxing dioxane using a Soxhlet extractor containing calcium hydride causes isomerization and lactonization. Lactone 112 is treated with methyllithium and the resulting dihydroxy ketal hydrolyzed with acid to obtain norketoagarofuran (100).

Evuncifer ether (101), a defense secretion of the West African soldier termite, has been prepared by Baker, Evans and McDowell as summarized in Scheme 16.17’ Hydroxy ketone 103178 is annelated using 1-chloro-3-pentanone to obtain enone 114. Deoxygenation of this material affords epi-y-eudesmol (1151, which is oxymercurated to obtain a 3: 1 mixture of 101 and isodihydroagarofuran (116).

Bicarbocyclic Sesquiterpenes; Hydronnpbtbalenes

140

%/

qoH

1 . Zn-NoOH 2. H ~ ( o A ~ ) ~ - H ~ o + 3. NoBH, 0

0

I02

HCI-H2S04

aoH :: h

1

w0

OH

103

::*,re*

OH

113

(&

I . Hg(OAcI2

2. NaBH4

0

3. Li-NH3

114

101

Scheme 16.

C

OH

I15

116

Baker-Evans-McDowell Synthesis of Evuncifer Ether

(6) Rishitin and Giutinosone

Rishitin (117) and glutinosone (118) are nor-eudesmanes having four asymmetric centers. Both have been synthesized by Masamune and coworkers at Hokkaido University in Sapporo.

now no

‘ H

117

118

The rishitin ~ y n t h e s i s ” ~uses santonin as a relay (Scheme 17). Although long (19 steps), the conversion of santonin into rishitin proceeds with good stereocontrol and embodies some exceedingly clever methodology. Removal of the angular methyl group with establishment of the C,-C,, double bond is accomplished by a modified Barton reaction (122-123) followed by decyanation of the ally1 nitrile 124. The latter transformation is rarely used in synthesis.

Eudesmnnes

141

I20

119

I . NOCl

2. hw

3.

I H 121

I\,THF,A OH

122

I . Na,EtOH, toluene, A

2. MsCI,

Ac

2. l , H + OAc

OAc

123

124

125

Scheme 17.

117

Hokkaido Synthesis of Rishitin

The Hokkaido synthesis of glutinosone'*' is summarized in Scheme 18. The synthesis begins with Robinson annelation on keto diester 126 to obtain octalone 127. The C, oxygen is introduced by epoxidation of the 2,3-enol acetate, as in the rishitin synthesis. However, in the present case, the acetoxy ketone 130 is equilibrated with its isomer 131 by heating with tetramethylammonium acetate; an approximately equimolar mixture of the two isomers is obtained. Treatment of the triester 132 with a large excess of methylenetriphenylphosphoraneaffords a mixture

Bicarbocyclic Sesquiterpenes; Hydronaphthalenes

142

EtOOC

0n

o o E t 126

I . Ac20

2 . m-CPBA

- fiooE 4%- ofi 0

' H

fi - AOO 'fiOoE+ Acm I27

0::

128

p-TaOH

Ace**

OOEt

H

I

' H

I29

* -Me4N'AcOA

OOEt

I30

lCH;O+H)zr

AcO

' H

I H

131

OOEt

I32

133: X ' C H 2 134: X = 0

I35

I . KOH 2. PhSCBF4

HO

118

Scheme 18.

Hokkaido Synthesis of Glutinosone

of lactones 133 (22%) and 134 (500/0). Treatment of the latter compound with methylenetriphenylphosphoraneaffords more of the desired isopropenyl compound 133 (88Yo). Hydrolysis of 133, followed by lead tetraacetate decarboxylation gives a mixture of unsaturated compounds (135). Cleavage of the ketal and conjugation of the enone system provides glutinosone (118) along with large quantities of phenolic byproducts. The overall yield of glutinosone in this 14-step synthesis is 2.a0j0.

Eudesmanes

143

(7) Occidentalol (+)-Occidentalol (136) is of interest as a rare eudesmane in which the angular methyl center has the (Y configuration. Its synthesis was first

OH

H

I36

achieved by Hortmann (Scheme 191.'81*'82The Hortmann synthesis begins with (+)-carvone (1371, which is converted into ester' 138 in a 10-step sequence. 183 The A-ring diene system is introduced by straightforward reactions, giving the trans-fused diene 140. Photoisomerization

0ACOOMe I H I38

I37

I39

+

@OOMO

142

P Cdl O O M e 143

I-BuOK

I

MeLi

'COOMe 145

144

Scheme 19.

I36

Hortmann Synthesis of Occidentalol

144

Bicrrbocycllc Sesquiterpenes; Hydronaphthalenes

at -78°Caffords the truns,cis,?runs-cyclodecatriene141, which undergoes disrotatory closure lo a mixture of cis-fused dienes 142 and 143. Treatment of the latter with methyllithium provides (+)-occidentalol (136). Base-catalyzed epimerization of isomer 142 gives a 2 5 1 mixture of isomers 144 and 142. Isomer 144 reacts with methyllithium to afford (-) -occidentalol (145). Antipode 145 has also been prepared by Ando and co-workers in a relay synthesis from ~ a n t o n i n . ' *The ~ Ando synthesis is summarized in Scheme 20. The synthesis begins with the vinyl decalone 146, which had

I . Na1O4-

0

COOMc

119

147

146 L i b , Li2C03

'r%ooMe 0

Br2'HoAce

* 0@OOMe

DMF

A1203-C5H5Nc

HOa

O

O

M

e

OOMe

*loo

I50

,Zb @

IS1

145

Scheme 20.

NoBH4_

Ando Synthesis of (-)-Occidental01

*'COOMe

152

Eudesmanes

145

previously been prepared from santonin in seven steps.'85 The steps used to convert keto ester 147 into (-)-occidentalol are very similar to those employed by Hortmann. Shorter syntheses of (+I-occidentalol were reported in 1972 by Deslongchamps and co-workers in Sherbrooke186 and by Amano and Heathcock in Berkeley.18' The latter synthesis was reported in Volume 2 of this series. 188 The Deslongchamps synthesis is summarized in Scheme 21. The synthesis proceeds via epicarrisone (1531, which is hydrogenated to establish the cis-fused decalone system. The diene system is introduced by a precedented series of steps. The most interesting transformation is selective dehydration of the allylic alcohol in 157. In fact, this transformation gives occidental01 in 62% yield, along with 26% of recovered 157.

I53

I55

154

I56

157

no" I36

Scheme 21.

Deslongchamps Synthesis of (*)-Occidentalol

146

Blcarbocycllc Sesqulterpenes; Hydronaphthalenes

A synthesis of (*:)-occidentalol has been reported by Watt and Corey (Scheme 22) .189 The synthesis begins with an unusual “inverse-electron demand” Diels-Alder reaction between pyrone 158 and alkene 159. The initial adduct extrudes CO,, providing diene 160 in 25-40% yield. The ester group is reduced to methyl by a sequence previously developed in Corey’s laboratory and the side chain is introduced by a WadsworthEmmons reaction. Hydrolysis of the enol thioether 164 gives a 3:l mixture of epimeric methyl ketones. Addition of methyllithium to the major isomer yields racemic occidentalol.

IS8

159

I62

163

165

164

Scheme 22.

161

160

I36

Watt-Corey (*)-Occidental01 Synthesis

Eudesmanes

147

The final occidentalol synthesis, which also yields the racemate, is that of Marshall and Wuts (Scheme 23).lgo Alkylation of the dianion resulting from Birch reduction of m-toluic acid affords diene acid 167, which is converted into ester 170 by a sequence of straightforward steps. Conversion of 170 to malonate 171 is surprisingly difficult, and may only be accomplished using lithium 2,2,6,64etramethylpiperidide (LTMP) as the base. Unsaturated aldehyde 172 undergoes smooth cyclization in a remarkably stereospecific reaction, affording diene acylal 173 as the only isomer. Straightforward conversion of 173 provides racemic occidentalol. The six published occidentalol syntheses are compared in Table 1.

I . Na-NH3 1-BuOH

42

I . A020

LiTMP

CHo 2. MeOH,) p-TaOH

COOMe

169

I70

2. L A H

HClOq CHO

COOEt 171

Ac20-EtOAr

I72

174

Scheme 23.

Marshall-Wuts Synthesis of

ClCOOEt

p-

'CH( OAC 12

173

I56

(zk 1-Occidentalol

148

Bicarbocyclic Sesquiterpenes; Hydronaphthalenes

Table 1. Comparison of Occidentalol Syntheses

Authors

Product

Steps

Overall Yield

)-occidentalol (-)-occidentalol

17 18

0.37% 0.13%

(-1-occidentalol

16

a

Starting Material

Hortman, et al.

( 0

Ando, el al.

+

0

Deslongchamps, e l al.

(+

)-occidentalol

10

5.8%

Heat hcock and Amano

( + )-occidentalol

8

2.3%

( k ) -occidentalol

9

1.5-2.4%

Watt and Corey

0

55:

Marshall and Wuts

a. Yields not provided in paper

Eudesmnnes

149

(8) Santonin, Yomogin, Tuberiferine, Alantolactone, Isotelekin, Dihydrocallitrisin, and Frullanolide The eudesmanolides have been particularly inspiring with regard to synthesis.19' During the period of this review, syntheses of a-santonin (119), yomogin (175), tuberiferine (1761, alantolactone (177), isotelekin (1781, dihydrocallitrisin (179), and frullanolide (180) have been accomplished.

q-*

0

;Ab

119

I75

I76

177

178

I79

I80

Marshal and Wuts have carried out the synthesis of a-santonin which is summarized in Scheme 24. 192 The synthesis begins with methylation of the dianion produced by Birch reduction of mtoluic acid to obtain dienoic acid 181, which is converted into bromide 184 by standard transformations. Alkylation of ethyl thioethoxymethyl sulfoxide with 184 affords a thioacetal monosulfoxide which is hydrolyzed with concomitant cyclization to dienol 185. Alkylation of the derived ketone occurs from the 0-face, affording keto ester 186 with good stereoselectivity. Reduction of 186 with lithium aluminum hydride affords an equimolar mixture of epimeric diols. After chromatographic separation, the desired isomer is oxidized by silver carbonate on Celite to obtain the trans-fused

Bicarbacyclic Sesqulterpenes; Hydronrphthrlenes

150

I66 T

O

181 o

E

t

I82 Li

@

I . Li-NH3, EtOH .c

Br

2 . PhiP,NBS

I. EtSAjEt

2. HCIO4

184

183

I . NCS-MOeS " :%OOEt>

G

C

O

O

E

l

I . LiAIH4 3. 2. AQ~COJ, roporolo, COlltO

189

Scheme 24.

dcg

I85 LDA, MoI,

H O

I87

I86

I88

*

119

Marshall-Wuts Synthesis of (f)-a-Santonin

lactone 187. Although methylation occurs stereospecifically from the pface, if the enolate of the methylated lactone is formed and quenched, the a-methyl epimer 189 is produced in good yield. Singlet-oxygen oxidation of 189 affords a-santonin. Caine and Hasenhuettl synthesized yomogin as is summarized in Scheme 2S.'93 The octalindione 191, available by a straightforward six step synthesis from diester 190, is alkylated as its enamine to obtain diketo ester 192. Selective reduction of the saturated ketone affords the cis-fused lactone 193, which is methylenated by Grieco's procedure. The final double bond is smoothly introduced by DDQ oxidation of enone 194.

Eudesmanes

I90

191

.

I (CH20H)2, H+

K-relrctridr

2. LDA 3. CHeO

____t

THF, -78'

0

0

I92

5 . HjO'

193

194

*

4 . MICI,C&N

COOEt

Scheme 25.

151

It5

Caine-Hasenhuettl Synthesis of (*I-

- amogin

Tuberiferine was synthesized by Grieco and Nishizawa as outlined in Scheme 26.'94 Enone 195, obtained in 85% yield by acid-catalyzed Robinson annelation of 2-methylcyclohexanone, is ketalized by reaction with ethylene glycol. Since 195 is known to give a 1:l mixture of the 4,5 and 5,6 isomers upon ketali~ation,'~~ the low isolated yield obtained by Grieco and Nishizawa (57%) is understandable. Alkene 196 is converted into trans-fused decalone 198 by the method developed by Marshall in his P-eudesmol synthesis.196 The side-chain destined to become the butenolide ring is introduced by alkylation of 198 with methyl bromoacetate; the desired stereochemistry is achieved by equilibration of the acetic ester group to the equatorial position. The stereochemistry of the oxygenated carbon is established as equatorial by lithium-ammonia reduction. The final transformations, methylenation of the lactone and introduction of the ring-A unsaturation, are accomplished using selenoxide methodology.

152

Bicsrbocyclic Sesquiterpenes; Hydronnphthnlenes

I99

198

20 I

200

Scheme 26.

176

Grieco-Nishizawa Synthesis of

( *bTuberiferine

Posner and Loomis have developed a synthesis of enones 205 and 206 as shown in Scheme 27.19' Compound 205 has been converted into alantolactone (177) by nine further transformations. 19* The real utility of the Posner-Loomis approach is for the synthesis of the furanoeudesmane attractylon (207). Enone 206 has been converted into 207 by a straightforward three-step sequence.'99 Thus, the Posner-Loomis synthesis of

207

207 from 202 requires only eight steps, compared with sixteen by the

Minato-Nagasaki route.'99

Eudesmanes

205

Scheme 27.

153

206

Posner-Loomis Approach to Eudesmanes

Miller’s synthesis of racemic isotelekin is summarized in Scheme 28.200 The synthesis is laid out along well-precedented classical lines and proceeds with high stereoselectivity and affords excellent yields at each step. Noteworthy features are the unexpected trans-esterification which occurs in conversion of 214 to 215 and the use of the Mannich method for methylenation of the lactone. Schultz and Godfrey carried out the interesting synthesis of dihydrocallitrisin which is shown in Scheme 29.201a*bEnol ether 217 is condensed with butenolide 218 and the resulting product brominated to obtain 219 (52% overall yield). Reductive removal of the bromine provides the deconjugated diene 220, which is reduced by sodium cyanoborohydride to obtain the cis-fused lactone 221. Hydroboration occurs predominantly from the &face as Oxidation of the secondary alcohol and epimerization of the cis to the trans decalone affords 223. The remaining carbons are introduced by Wittig methylenation and methylation of 224. The stereochemistry of the lactonic methyl group is

154

Bicarkyclic Sesquiterpenes; Hydronaphthalenes

I . Li-NH, 2. (EtO)$JOCI

Na

0

3. Li-EtNHz

208

(2 H

209

211

n

I.

210

+

2. MeOOCABr 3 . NaOAc. HOAc

212

'OM@

t-BuCOO'

I . NaBH 2. HCI '&

t-Buc~~*

213

214

215

216

I . CHP, MezNH

2. Me1 3 . DMF, 180.

178

Scheme 28.

Miller-Behare Synthesis of (*1-Isotelekin

established in the decarbomethoxylation step, which is carried out by heating with cyanide ion in DMF. Intermediate 219 has also been converted into (f)-7,8-epialantolactone (2251, an isomer of the natural product 177.

225

Eudesmanes

22 I

220

155

222 I . NaH, Me1 2. NaCN,HMPf

I . Joner 2. NaOMe ~

223

224

119

Scheme 29.

Schultz-GodfreySynthesis of (*1-Dihydrocallitrisin

Frullanolide has been synthesized by two groups. The first synthesis was by Still and Schneider (Scheme 30).*03 Octalone 227 is prepared from 2,6-dimethylcyclohexanone by a route involving acid-catalyzed Michael addition and base-catalyzed aldolization. The two-step process is necessary because other direct annelation procedures fail. Epoxidation occurs from the &face to give an epoxy ketone which is rearranged by the Wharton procedure to obtain the cis-fused octalol 228. The elements of the methylenebutyrolactone moiety are introduced by an inventive application of Ireland's enolate Claisen rearrangement. Iodolactonization, followed by elimination of HI, completes this original synthesis.

156

Bicarbocyclic Sesquiterpenes; Hydronaphthalenes

ho

I . H202,0H2. N2H4 *

2 . NoOEt, EtOH

I.C

I

226

227

0 0

2I .' LEt3SiCI DA * 3.

A

L

n

228

I . KOH

& y c o o M O

4. Me2S0,. K2CO3 MeOH

I

3. DBU

I80

230

229

Scheme 30.

Still-Schneider Frullanolide Synthesis

Yoshikoshi and co-workers synthesized frullanolide as shown in Scheme 31 .*04 The synthesis employs the novel annelation reagent 232, which is used to convert P-keto ester 231 into adduct 233 (33% yield). The main problem with reagent 232 is that it does not work with simple ketones. Thus, in this case, an extra six steps are required to convert the angular methoxycarbonyl group to a methyl group. The overall yield

231

232 COOMr

I . DlBAL

233

234

I . LAH

I . LOA

4. Jonra

H+

235

Scheme 3 1.

236

Yoshikoshi's Synthesis of Frullanolide

I80

Cadinanes

157

of frullanolide is 2.6% for the 12 steps from 232. This is to be compared with 8.3% for the 11 steps of the Still synthesis, which begins with commercially available material.

B. Cadinanes (1) Aromatic Cadinanes A number of aromatic cadinanes (cadalenes and calamenes) have been synthesized. Compounds 237-241, constituents of elm wood, have been prepared by Krishna Rao and c o - w o r k e r ~ Mansonone . ~ ~ ~ ~ ~ ~D ~(242) and mansonone G (2431, interesting natural o-quinones, have also been prepared by Krishna Rao, 2073208as has the structurally related furan pyrocurzerenone (244) .209 Lacinilene C methyl ether (2451, a compound which may be responsible for byssinosis, a debilitating disease prevalent in the textile industry, has been prepared by McCormick and coworkers.210

Me0 237: R =Me 238: R = CHO

239: R * Me 240: R = CHO

243

CHO

241

244

242

245

Krishna Rao's synthesis of the phenolic calamenes 237 and 238 is summarized in Scheme 32.*05 A noteworthy feature of this synthesis is cyclization of 247 to 248, in which it appears that a secondary, rather than a tertiary, carbonium ion is involved. The reason for this unusual behavior cannot be formation of a six- rather than a seven-membered

158

Bicarbocyclic Sesquiterpenes; Hydronaphthalenes

246

248

I . DMF, P O C I S 2. H$4NHCONH2,

CH=NNHCONH2

C5H5N

E"j\ 249

I. (COOH),

2. C r , b N * H C I 170- 180'

@

I

Me

250

@OH

CHO

238

CSH5N* HC I 170-180'

p

H

237

Alexander-Krishna Rao Synthesis of Phenolic Calamenes

Scheme 32.

ring, since @so-cyclizationof the tertiary carbonium ion could lead to the reasonable spirocyclic ion 251, which would probably rearrange to a mixture of 252 and 253. Dehydrogenation of 249 is effected by heating with

MI 251

252

253

Cadinanes

159

@

CH=NHNHCONHe Me

254

2,3-dichloro-5,6-dicyanoquinoneto afford the cadalene 254 in 55% yield. This material is converted into cadalenes 239 and 240 by methods similar to those used to convert 249 into 237 and 238. Cadalene 241 was prepared from the known tetralin 25S211 as shown in Scheme 33. The conversion of tetralin 255 to 13-tetralone 257 is carried out using a method developed by Stork for his early diterpene synthesis work. The most noteworthy step in the synthesis is conversion of ketone 257 to the hydroxynaphthoic acid 258, which occurs in unspecified yield upon treatment of 257 with dimethyl oxalate and sodium methoxide.

255

( CNoOMc OOMd2

-

256

@ioH 258

2. I . MaOH, LAH

-

257

HL

Me

@@ CHeOH

259

Collin8

MI

A

CHO

24 I

Scheme 33. Viswanatha-Kirshna Rao Synthesis of 3-Methoxy-7-hydroxycadalenal

Bicarbocyclic Sswuiterpenes; Hydronaphthalenes

160

Krishna Rao synthesized pyrocurzerenone (244) as shown in Scheme 34.209 Friedel-Crafts cyclization of the readily available acid 260 occurs with concomitant deisopropylation to provide tetralone 261. The furan

26 I

260

262

Scheme 34.

263

244

Viswanatha-Krishna Rao Synthesis of Pyrocurzerenone

ring is introduced in good yield by cyclization of the aryloxy ketone 262. Pyrocurzerenone is converted into mansonone D by catalytic hydrogenation, followed by successive oxidations with chromic acid and selenium d i ~ x i d e . ~A” similar sequence may be used to convert the cadalene 265 into mansonone G.*O* I . Cr03, HOAc

244

Me0

2. SeOZ,EtOH

&, 265

*

264

242

1 . CrO,, CH3COOH CHSCHLCOOH

2. H e r - H O A C

no 266

243

Cadinanes

161

McCormick’s synthesis of lacinilene C methyl ether (Scheme 35)210 begins with keto ester 267, available in quantity by Friedel-Crafts acylation of o-cresol methyl ether. The most noteworthy feature of the synthesis is the method of introducing the unusual functionality of ring A. The three-step sequence 270-245 occurs in 27% overall yield.

E+oocFoM 267

268

2 . Jonas-

Scheme 35.

(2)

McCormick’s Synthesis of Lacinilene C Methyl Ether

Q Xadinene, y

Xadinene, a -Amorphene, Zonarene, and Epizonarene

At the time of the writing of Volume 2 in this series, little synthetic work had been directed at the nonaromatic cadinanes. In the intervening 10 years, a considerable amount of research in this area has appeared. Syntheses of the hydrocarbons z -cadinene (272), y 2 -cadinene (273), a -amorphene (2741, zonarene (279, and epizonarene (276) are discussed in this section.

162

Bicorbocyclic Sesquiterpenes; Hydronrphtbrlenes

27 2

27 4

273

275

276

Vig's synthesis of Q -cadinene is summarized in Scheme 36.212The initial Diels-Alder adduct of diene 278 and methyl vinyl ketone has the acetyl group axial, the result of endo addition. Base-catalyzed epimerization affords isomer 280, in which the acetyl substituent occupies the equatorial position. Wittig methylenation provides diene 281, which is selectively hydrogenated and hydroborated, oxidized, and hydrolyzed to

277

278

280

279

281

282

Ph3P=CH2

2 . Collins 3. HjO+

0

h

HA

i

283

Scheme 36.

272 ( *1-r-Cadinene:

Vig Synthesis

Cadinanes

163

obtain dione 283, the racemic version of an intermediate previously employed by Burk and Soffer in a synthesis of ( + ) - ~ - c a d i n e n e . ~For l~ comparison, the Burk-Soffer synthesis is reproduced in Scheme 37. In addition to providing ( + b e -cadinene, a simple modification provides (- 1 - y p d i n e n e (273).

EtO

P h 3 P s C H L Et

, H

&pJ

"A

A

204

yi

285 2. P ~I . ~ P x C H ~

3. SOCI2-CsHSN

272

Scheme 37.

27 3

(*)-r-Cadinene and (-)-y2-Cadinene: Burk-Soffer Synthesis

Like the cadinenes just discussed, a-amorphene (274) presents the challenge of three stereocenters-two at the ring junction and the third at the isopropyl substituent. Unlike the cadinenes, a-amorphene possesses a cis ring fusion. In principal, a-arnorphene could be prepared from the Burk-Soffer intermediate 284. However, it was found that epimerization of the ring junction position occurs under the conditions of the Wittig reaction, thus leading into the trans-fused cadinene series. Remarkably, Burk and Soffer have also reported that 284 undergoes epimerization even in its reaction with methyllithi~rn.~'~ Clearly this reaction invites further study. The only synthesis of an amorphene to date is the Mirrington-Gregson synthesis of a-amorphene, summarized in Scheme 38.215Diels-Alder

164

9

Bicnrbocyclic Sesquiterpenes; Hydronaphthalenes CI +

I . HCOOEt, NaH 2 . N a O M e , Me1

EtOH, N a O S

&N-)

3. NoOH 286

287

289

288

291

Scheme 38.

292 ( f )-aArnorphene:

290

274

Mirrington-Gregson Synthesis

addition of 1-methyl-] ,3-cyclohexadiene and a-chloroacrylonitrile provides an adduct (2861, which is hydrolyzed by Evans’ procedure to bicyclooctenone 287. Methylation, via the formyl ketone, affords an epimeric mixture (288) which reacts with the Grignard reagent derived from (8-1-brorno-3-methy1-1-buteneto give a mixture of tertiary alcohols 289 and 290 in a ratio of 3:l. The alcohols are separated and the minor isomer pyrolyzed at 300°C to obtain the cis-fused octalone 291. Deoxygenation of the more highly substituted enol phosphate provides (*)-a-amorphene. This synthesis provides a very nice example of the solution of a difficult problem by the judicious application of two modern methods- the oxy-Cope rearrangement and the Ireland deoxygenation sequence. It is unfortunate that the synthesis is marred by the lack of stereoselectivity in the methylation step. Zonarene (275) and epizonarene (276) are simpler targets than the cadinenes and a -amorphene, since there are only two stereocenters. Of the two, zonarene is the more challenging target, since the methyl group

Cadinanes

165

is axial. The only synthesis to date is one due to Yamamura (Scheme 39).216 The synthesis starts with dienone 293, previously prepared by Yamamura and co-workers by modification of a-santonin (119).216b

119

-

293

294

+ 295

296

298

299

Scheme 39.

I . Seporote 2. H z - P t O r

HO"

297

275

300

Yamamura's Synthesis of (+)-Zonarene

Acid-catalyzed cyclization affords diol 294, which is epoxidized by rn-chloroperoxybenzoic acid in ether to provide epoxy diol 295. Selective dehydration of one of the two tertiary alcohols is achieved by POCI, -pyridine, but some isomerization occurs, leading to a mixture of unsaturated hydroxy ethers 296 and 297 in a ratio of 40:27. The isomers are separated and 296 is hydrogenated to obtain hydroxy ether 298, which is reduced to diol 299. Dehydration of this substance also gives rise to a mixture-(*) -zonarene (275) being accompanied by unsaturated alcohol 300. Although Yamamura's synthesis will be recorded as the first successful synthesis of zonarene, it leaves much room for improvement. The synthesis requires over 25 steps from suntonin and requires separation of isomers at several stages.

166

Bicnrbocyclic Sesquiterpenes; Hydronrpbtbrlenes

A short synthesis of (4-1-epizonarene (276) from (-bmenthone (301) is summarized in Scheme 40.2'7 A noteworthy feature of this syn-

thesis is the Mannich alkylation of 301, which occurs at the methylene, rather than the methine position. Annelation of ring A is accomplished by reaction of enone 302 with acetoacetic ester. Enone 303 is produced under equilibrating conditions; the stereochemistry at the three stereocenters corresponds to the more thermodynamically stable product.

Scheme 40.

(3)

a -Cadino1 and

a -Cadino1

Kulkarni's Synthesis of (+I-Epizonarene

Torreyol

(304d and torreyol (304b) present the additional synthetic

challenge of the tertiary alcohol group-a fourth stereocenter compared to the cadinenes and a -amorphene.

304

a: R = a-ti b: R = @ - H

Cadinanes

167

A synthesis of (*)-a-cadinol, by Caine and Frobese, is outlined in Scheme 41.*'* Enone 306, obtained as a 3:2 mixture of stereoisomers, is dehydrogenated to trienone 307 which is reduced by transfer hydrogenation to stereochemically pure 306 in which the angular methyl and 0 dioaone 305

-

308 I . L .I / N H ~

+pH 4

I . LiAIH4 L 2. Jonas

HOAc

306

0

307

306

2. H202

0

+

o@oH

309

EtLi

L

2. PhSSPh

f

4

i

311

313

Scheme 41.

o

H

2. HgCl2, MeCN/H20 3. KOH,MaOH

Ho h P! !

PhS

510

€(OH, CYCIOhexene, A

312

304a

314

Caine-Frobese Synthesis of (f)-a-Cadinol

isopropyl groups are cis. This compound is converted to the crossconjugated cyclohexadienone 308. Photochemical rearrangement of the latter provides acetoxy enone 309, which is transformed into hydroxy enone 310. Lithium ammonia reduction of 310, followed by sulfenylation of the resulting enolate, gives a 3:l mixture of epimeric sulfides 311 in 769'0 yield. The mixture is treated with excess ethyllithium and the resulting mixture of diols (312) is cleaved by lead tetraacetate. After

168

Bicarbocyclic Sesquiterpenes; Hydronaphthalenes

hydrolysis, aldol condensation and dehydration gives 3-0x0-a-cadinol (313). Wolff-Kishner reduction of this compound gives (*1-a-cadinol (304d and its isomer 314 in a ratio of 2:l. Taber and Gunn reported the synthesis of torreyol which is summarized in Scheme 42.219 Aldehyde 314, prepared by alkylation of the piperidine enamine of isovaleraldehyde with ethyl acrylate, is condensed with the Wittig reagent derived from rnethylallyltriphenylphosphonium chloride to give exclusively the E-stereoisomer 315. Ester 315 is converted to an aldehyde, which reacts with vinylmagnesium bromide to give a diastereomeric mixture of trienols (316). Oxidation of this material at 0 "C gives trienone 317, which spontaneously cyclizes to a 9:l mixture of diastereomeric octalones 318 and 319. Treatment of the mixture with methyllithium provides a mixture of (*)-torreyo1 (304b,64% based on the 318 in the starting material) and stereoisomers.

COOEl

2. HzO

EtOOC 314

+

315

4 qoH +

319

Scheme 42.

isomers

304 b

Taber-Gunn Synthesis of (st )-Torreyo1

Drimanes

169

C. Drimanes (1) Driman-8-01, Driman-8,1 I -diol, and Drim-&en- 7-one

Driman-8-01 (320), driman-8,ll-diol (3211, and drim-8-en-7-one (322) are constituents of Greek tobacco. Enzell and co-workers have prepared

all three of these compounds220i221from drim-7-en-11-01 (323, bicyclofarnesol), which has been prepared previously.222Scheme 43 outlines the synthesis of alcohol 320 and diol 321. Bicyclofarnesyl acetate undergoes stereoselective epoxidation from its a face and epoxy acetate 324 reacts with lithium aluminum hydride to give diols 321 and 325 as a 1:9 mixture. Deoxygenation of the minor isomer is accomplished by reduction of its monomesylate to obtain alcohol 320. For synthesis of enone

boc c$:

2. I . m-CPBA Ac20,CsHSN H

323

LiAIH4c

H

325

2. i , rrparote MsCI,CSHSN

3. LiAIH,

+

H

H

324

..OH

Scheme 43.

O H -& :

321

&--OH

H

320

Enzell's Synthesis of Driman-8-01 and Driman-8,1 I-diol

Bicarbocyclic Sesquiterpenes; Hydronaphthalenes

170

322 (Scheme 44) epoxy acetate 324 is hydrolyzed and the resulting alcohol oxidized using Brown’s two-phase conditions to obtain B, y-epoxy aldehyde 327. Base-catalyzed elimination, followed by reduction of the

doc&: KOH, MIOH+

H 324

H2Cr04 rthrr/H20

~

cb-& dOH H

326

Ac20-CsHsN;

k

&o

2. I . KOH,MeOH+ NaBH4

H 327

I , Jones 2 . Zn-HOAc

+

H

328

329

Scheme 44.

&& H 322

Enzell’s Synthesis of Drim-8-en-7-one

aldehyde function affords diol 328. Careful acetylation yields a mixture of products, of which monoacetate 329 is the major product. Oxidation and reductive removal of the acetoxy group provides drim-8-en-7-one (322).

(2) Confertgolin, Isodrimenin, Cinnamolide, Drimenin, Futronolide, Polygodial, Isotadeonal, and Warburganal

Confertifolin (330), isodrimenin (3311,cinnamolide (3321, and drimenin (333) are simple drimanic lactones. Futronolide (334) is related to 331 by having an additional 7a hydroxy group. Polygodial (335) and isotadeonal (336) are dialdehydes having the drimane and epidrimane stereochemistry. Warburganal (337) is an oxygenated relative of polygodial. The discovery of warburganal, which has potent antifeedant activity against the African army worm, KB cytotoxicity, and antimicrobial properties, has sparked a resurgence of activity in synthesis of drimanic sesquiterpenes.

Drimanes

H

H

330

331

H

332

171

& H

333

&O H

334

335

336

337

Because of its importance in several of the syntheses discussed in the following pages, we reiterate a basic reaction in the drimane field-the acid-catalyzed cyclization of farnesic acid to give the bicyclofarnesic acids. This reaction, which was examined extensively by Schinz, Stork, and E s ~ h e n m o s e produces r ~ ~ ~ predominantly drimanic acid (338), with 8-epidrimanic acid (339) being a minor product; typical yields are 35% and Solo, respectively.

H

H

338

339

The first syntheses of confertifolin and isodrimenin were carried out by Wenkert.224 In connection with his project to synthesize warburganal, Oishi and his co-workers developed the synthesis outlined in Scheme 45.225 The synthesis begins with dihydro-P-ionone, which is converted by successive treatment with methyl orthoformate and pyridinium bromide in methanol into the keto acetal 340. Compound 340 is subjected to a Darzens condensation and the intermediate epoxide is treated with acid to obtain a furan ester (342). Friedel-Crafts cyclization of this substance provides a tricyclic ester, which is saponified to 343. It is noteworthy that the cyclization is stereospecific, affording only the transfused isomer. After decarboxylation, furan 344 is oxidized by lead

172

Bicrrbocycllc Sesquiterpenes; Hydroarpbthrlenes

340

342

& H

Ac

A. 170°. or

+

_____t

B. I . KOH 2. Jon08 3. N o B H ~

345

Scheme 45.

344

343

k

H

330

331

( *bConfertifolin and (*)-Isodrirnenin:

Oishi Synthesis

tetraacetate to obtain a crystalline diacetate (345, undetermined stereochemistry). Thermolysis of this compound yields (*bconfertifolin (330, 74%) and (f)-isodrimenin (331, 17%). To obtain isodtimenin as the major product, the diacetate is first saponified and the resulting diol oxidized to (*)-winterin (346). Reduction of (*)-winterin by sodium

346

borohydride in THF produces (kbisodrimenin (81%) along with 14% of (*)-confertifolin. The Oishi synthesis is reasonably short (10-12 steps from @ionone) and produces 330 and 331 in reasonable yields (about 13% overall). Furthermore, the reactions employed appear to be adaptable to large scale applications.

Drimanes

173

The group headed by Kitahara at Sendai has developed syntheses of (*bdrimenin, (*1-cinnamolide, and (*bpolygodial. Although part of

that work was reported in Volume 2 of this series,226the full synthesis is repeated in Scheme 46, since the applications to (*)-cinnamolide and (*1-polygodial have appeared since 1970.227*228 The Sendai group begins with cyclization of monocyclofarnesic acid to give mainly drimanic acid, which is esterified to obtain methyl cyclofarnesate (347). Oxidation of 347 with singlet oxygen produces allylic alcohol 348 in 33% yield. Treatment of 348 with aqueous sulfuric acid provides (*bdrimenin (333) in a direct four-step route from 346 (15% overall yield). Reduction of 333 gives a diol which is smoothly oxidized by manganese diox-

&

& &

*4;::;;

H

H 333

332

CrO,*CsHsN-

(MeO),CHW p-TSOH

H

& H 352

Scheme 46.

I . NaOH, MeOH 2. CH2NL, HCI

'8

A;;20+

I.LiAIH4+

2. MnOe

H

350

O

H

349

351

&HO H

335

(*1-Drimenin, (*1-Cinnamolide, and (*1-Polygodial:

Sendai Syntheses

174

Bicrrbocyclic Sesquiterpenes; Hydronaphthalenes

ide to (*)-cinnamolide (332). Saponification of the latter compound gives a sodium salt which is treated with ethereal diazomethane and then slowly acidified with ethereal HCI to obtain hydroxy ester 349 admixed with the starting lactone 332. The mixture is oxidized and the resulting aldehyde converted into an acetal (351). At this stage 351 is separated from unreacted lactone 332 by chromatography. Crystalline 351 is obtained in 3096 overall yield based on (*)-cinnamolide. Reduction of the ester function, followed by manganese dioxide oxidation provides the monoacetal 352, which is hydrolyzed with oxalic acid in aqueous acetone to obtain (f)-polygodial (335). The Sendai group has also used methyl bicyclofarnesate in a synthesis of futronolide (3341, as summarized in Scheme 47.229The singlet oxygen oxidation of 347 is followed by isomerization of the double bond and oxidation of the secondary alcohol to obtain unsaturated keto ester 353. Ketalization, allylic bromination, and hydrolysis provide keto lactone 355. Reduction of the ketone affords entirely the equatorial alcohol

COOMa

COOMe

COOMe

H

H 354

353

347

&y0E3+& A

ti

355

H

H

357

356

H 356

Scheme 47.

+

&*OH

H

334

( *)-Futronolide: Kitahara's Synthesis

Drimanes

175

356, which is epimeric with (*)-futronolide at C-7. However, the crystalline bromide 357 undergoes solvolysis in aqueous dioxane to provide a 1:l mixture of 356 and (*)-futronolide (3341, which may be separated by repeated crystallization. The first successful synthesis of warburganal to be published was that of Ohsuka and Matsukawa of Osaka City University.23o The synthesis (Scheme 48) begins with methyl 9-epibicyclofarnesate (358). It is important to recall that cyclization of either farnesic acid (or its esters) or monocyclofarnesic acid (or its esters) affords this stereoisomeric bicyclic acid as the minor product, usually in yields on the order of 5%. However, the first step of the Ohsuka-Matsukawa synthesis, allylic oxidation with selenium dioxide, does not work with methyl bicyclofarnesate. Thus, the synthesis must be carried out starting with the less abundant epimer. Reduction of acetal ester 359, followed by Collins oxidation, provides an acetal (360) which is hydrolyzed by aqueous acid to obtain (k)-isotadeonal (336). Alternatively, the enolate derived from 360 is

I . LiAIH4 2. Collins

H

H

H+

358

360

359

H+,H,O

HO f H O

LHMDS Mo05,HMPT

CHO

H 361

A\

W t

H+

H

338

H

337

Scheme 48.

H

(f)-Isotadeonal and (*)-Warburganal: Ohsuka-Matsukawa Synthesis

O

176

Blcarbacyclic Sesquiterpenes; Hydronapbthrlenes

oxidized with MOO, to obtain the hydroxylated aldehyde 361 in 24% yield. The oxidation is stereospecific, the bulky oxidant apparently being directed to the a-face of the molecule by the angular methyl group. Hydrolysis of 361 provides (*)-warburganal (337). Although this synthesis is short (six steps) and produces warburganal in reasonable yield (9%overall), it is marred by the scarcity of the starting material. However, if the problems associated with the allylic oxidation of the more abundant 347 could be solved, this approach to warburganal would be very attractive, especially in light of Nakanishi's experience with the hydroxylation step (see Scheme 50 below). Oishi and co-workers have synthesized ( f )-warburganal beginning with (*)-isodrimenin (331) as shown in Scheme 49.231Allylic oxidation

33 I

355

-

362

366

367

I . NaOH

2 . DMSO. DCC

368

Scheme 49.

337

Warburganal: Oishi Synthesis

Drimrnes

177

affords 355, which had been prepared earlier by the Kitahara group (see Scheme 47). Intermediate 355 is converted by several standard steps into enone 363, which is epoxidized using alkaline hydrogen peroxide to obtain epoxy ketone 364. Wolff-Kishner reduction occurs with concomitant elimination to afford unsaturated triol 365, which only needs to have its two primary alcohol functions oxidized to obtain warburganal. Unfortunately, such a simple solution does not succeed, and it is necessary to put triol 365 through an intricate series of protection, deprotection, and oxidation steps to achieve the desired goal. In any event, the eight-step sequence works, and (*)-warburganal (337) is produced in 38% overall yield from epoxide 364. The major loss seems to come in the silylation of 365, which is not completely selective. In all, the Oishi synthesis requires 25 steps from dihydro-P-ionone (see Scheme 44) and produces (*)-warburganal in an overall yield of about 2.6%. Nakanishi’s synthesis of warburganal (Scheme begins with a Diels-Alder reaction between diene 369 and dimethylacetylenedicarboxylate, a reaction employed earlier by Brieger in a synthesis of ( f ) - ~ i n t e r i n Allylic . ~ ~ ~ oxidation of the adduct 370, obtained in 83% yield, affords dienone 371. Catalytic hydrogenation of 371 provides the trans-fused decalone 372. The overall yield of 372 is 76% from diene 369. The keto function is removed and the necessary C,-C, double bond is introduced by a straightforward method and the resulting diester 373 is reduced to obtain diol 374 (which may be oxidized to polygodial, 335). Again, an intricate sequence of protection, deprotection, and oxidation steps leads to the monoprotected polygodial (3771, which is hydroxylated by the Vedejs procedure to obtain the axial alcohol. Hydrolysis of the acetal provides (*)-warburganal in 85% yield for the last two steps. It is not clear why the Nakanishi hydroxylation is so superior to that reported by Ohsuka and Matsukawa (see Scheme 481, which proceeds in only 24%. Different acetals and different bases were employed by the two groups. Overall, the Nakanishi synthesis provides (*)-warburganal in 16 steps (15.7% yield) from diene 369. Of the three syntheses of this interesting compound, this is clearly the most efficient at this stage. However, note that application of Nakanishi’s hydroxylation conditions in the Ohsuka-Matsukawa route would raise the overall yield of that synthesis to 32%.

178

Bicsrbocyclic Sesquiterpenes; Hydroosphthrlenes

369

370

H2- Pd/C

&Me

0

&: H

37 I

LiAIH4,

~

COOMa

3.DEU

H

372

&OH

H

373

I . I-BuMe2SiCI+

Scheme SO.

OSiR3

2 . AcZO

I . HOAc, THF 2. MnOe

H

374

376

I . NaBH4 2 . MtCI,Et3N

375

377

( *)-Warburganal: Nakanishi’s Synthesis

337

(3) Pallescensin-A Pallescensin A (379) does not fit into any of the normal sesquiterpene skeletal classes. It has been suggested that it may arise from the acyclic precursor 378.234 Because of its structural similarity to the drimane class, and because the key reaction employed for its synthesis is one of the characteristic reactions leading into the drimane family, we consider pallescensin-A at this point. Nasipuri and Das constructed the sesquiterpene in a straightforward manner234 by coupling geranyl chloride (380) with 3-furylmethylmagnesium chloride to obtain diene 378 (Scheme 51). Treatment of this material with boron trifluoride affords (f)-palles-

Drimrnes

380

Scheme 51.

179

H 379

378

Nasipuri-Das Synthesis of (*)-Pallescensin A

censin-A in 84% yield. The cyclization step is stereospecific; only the trans-fused diastereomer is produced. A related cyclization was employed by Matsumoto and Usui in a synthesis of (+I-pallescensin-A (Scheme 52).235 The cyclohexene derivative 383 is prepared in a 2-step sequence beginning with R-(-)-a-cyclocitral (381). Cyclization of 383 is effected by aluminum chloride. Unfortunately, (i)-pallescensin-A is produced as the minor component of a 2:l mixture, in 10% yield.

38 I

382

H 384

Scheme 52.

383

H

379

Matsumoto-Usui Synthesis of (*)-Pallescensin A

180

Bicrrbocyclic Sesqulterpenes; Hydronaphtbrlenes

D. Eremophilanes (1) Valencene, No0 tka to ne, 7-Epinootka tone, Isonootka to ne, and Dihydronootkarone

The eremophilane family of sesquiterpenes has received a great deal of attention during the past 10 years. In this section, we review the work on valencene (3851, nootkatone (3861, 7-epinootkatone (387)’ and isonootkatone (388). In addition, we describe a synthesis of dihydronootkatone (389). Although this substance is not a natural product, its synthesis is styled after the probable biogenesis of the eremophilanes.

385

387

386

oT O

388

+

q

389

Full papers reporting Marshall’s syntheses of (*1-nootkatone (386)236 and (*1-isonootkatone (388)237 have appeared. Since these syntheses have been reported in Volume 2 of this series,238 they will not be repeated. Odom and Pinder explored the direct route to nootkatone which is summarized in Scheme 53.239 Keto acid 390 is converted into (*)-dihydrosylvecarvone (394) by the straightforward sequence depicted. Annelation of ketone 394 with 3-penten-2-one affords (*)-7-epinootkatone (387) and (f)-nootkatone (386) in a ratio of 9:l. Sased on earlier observations by Piers,240 Takagi and co-workers improved the ratio of nootkatone and 7-epinootkatone to 1:l by converting ketone 394 into the 6- n-butylthiomethylene derivative before Robinson annelation (Scheme 54).241 Actually, at the time of Takaga’s paper,

Eremophilanes

390

391

393

394

Scheme 53.

181

392

387

386

Odom-Pinder Synthesis of (+ )-7-Epinootkatone and (fbNootkatone

___t.

NaOMe

394

395

-“cq; H20

I I

386

Scheme 54.

I-AmOK, I - AmOH,

396

+

O

90°

R

387

Takagi’s Synthesis of (+ bNootkatone

P i r ~ d e rhad ~ ~already ~ provided a superior solution to the problem of stereochemistry at C7. The problem was solved (Scheme 5 5 ) by simply delaying introduction of the C, side chain until after the octalin system has already been formed, so that the desired equatorial disposition of the C, substituent can be established by equilibration. The required starting material is monoketal 397, which is transformed into 398. The carbonyl group is removed by standard steps, affording ketal 399. After hydrolysis, the acetonyl group is introduced by a modified Wadsworth-

182

Bicarbacyclic Sesquiterpenes; Hydronaphthalenes

397

398

399

I. H~O+ 0

2.

ii

3. H g C I z , M I C N , +O

4 . bare

385

400 1-BuOCrO3H

386

Scheme 55.

Pinder's Synthesis of ( *bvalencene and (*t)-Nootkatone

Emmons reaction and ketone 400 is then subjected to Wittig methylenation to obtain (*bvalencene. Allylic oxidation restores the enone function, giving (fhootkatone (386). In 1973, Dastur reported the imaginative synthesis of (*)-nootkatone which is outlined in Scheme 56.243 Birch reduction of 3,4,5-trimethylanisole affords diene 401, which is heated with methyl acrylate in the presence of dichloromaleic anhydride to provide adduct 402. Although products arising both from exo and endo transition states are formed, addition appears to occur only trans to the secondary methyl in the active diene, 403. Thus, although the product is a mixture of epimers at the

Eremophllrnes

HCooHc

o%cHo

NaOH,

406

386

Scheme 56.

183

qoH AIZOs collidine, A *

407

388

Dastur's Synthesis of (*)-Nootkatone

methoxycarbonyl position, the secondary methyl group is completely syn to the double bond. This is important, as this relative stereochemistry governs the relative stereochemistry of the vicinal methyl groups. The allylic methyl is functionalized by selenium dioxide oxidation and the resulting aldehyde is methylenated to obtain diene 404. After addition of methyllithium, tertiary alcohol 405 is solvolyzed in formic acid to obtain formate 406. The conversion of 405 may be formulated as:

184

Bicnrbocyclie Sesqulterpenes; Hydronaphthalenes

The synthesis is completed by saponification of the formate and dehydration of the tertiary alcohol by heating it with alumina and collidine. The final step provides a 3:l mixture of (f)-nootkatone (386) and (*)-isonootkatone (388). Heathcock and Clark have modified the Dastur synthesis as shown in Scheme 57.244 The enol ether of dihydroorcinol (408) is alkylated by the Stork-Danheiser procedure fist with methyl iodide and then with pienyl bromide. The product is a 4:l mixture of diastereomers of which 409 is

2. 1 . H30+ -Lib

3. rrparatr

oq 408

409

HCOOH

A

o

P

410

Scheme 57.

O

C

H

O

406

Heathcock-Clark Modification

of the Dastur Nootkatone Synthesis

the major isomer. This isomer is separated chromatographically and treated with vinyllithium to obtain, after hydrolysis, trienone 410. Formolysis of 410 affords the Dastur enone formate (406). Hiyama and co-workers have developed an interesting stereospecific synthesis of ( k h o o t k a t o n e from keto ester 412, which they prepare by Diels-Alder addition of methyl acrylate to diene 411 (Scheme 58).245 This material is condensed with the dilithium salt of 3-butyn-2-01 to obtain 413 as a diastereomeric mixture. The methyl ester undergoes saponification at some point during the treatment. Compound 413 reacts with sulfuric acid in methanol to give 414, a 3:2 mixture of stereoisomers at the methoxycarbonyl center. The stereoselectivity in this ring

’

TMSO

+

Eremophilnnes

185

“aooMe OLi I

A

A- HJO+_

‘COOMe

*\Li

2. H+

411

412

PH

7

\v

H

OH

O

a OOH

413 I . KOH,MeOH 2. MeLi 3. Ph3P=CH2 4. PCC

Na0H4-CeC13

H2S04-MIOH

*

&

*

b%COOMe

btCOOMe 415

414 Br2CH OH

BrfGit

I . n-BuLi+ 2. H2S04 41 7

416

386

Scheme 58.

Hiyarna’s Synthesis of (*:)-Nootkatone

closure is interesting and may be regarded as a conrotatory closure of the protonated dienone:

-H+

ketonize

~

h

O

O

M

e

Bicarbocyclic Sesquiterpenes; Hydronaphthalenes

186

Reduction of the enone is accomplished with NaBH,-CeCI,, to afford alcohol 415, a mixture of several diastereomers. The ester is saponified, the resulting hydroxy acid treated with excess methyllithium and the acetyl group thus produced epimerized and methylenated by Wittig reaction. Oxidation of the secondary alcohol affords dienone 416 in isomerically pure form. The five-membered ring is expanded by addition of dibromomethyllithium, and conversion of the resulting adduct with n-butyllithium. The product is a &y-unsaturated ketone, which is conjugated with acid to obtain (*hootkatone. The overall synthesis requires 11 steps and proceeds in 17% yield. A final nootkatone synthesis, which produces the natural enantiomer, is that of Yoshikoshi and his co-workers (Scheme 59).246 Nopinone (4191, conveniently obtained by ozonolysis of 8-pinene (418) is converted into silyl ether 420, which is employed in Mukaiyama’s version of

&

O,_

Me3SiCI DMF, Et3N

418

419

421

Me3si0&6I . -CHOP TiCI,

2 . p-TsOH

420

422

423

O y MIOH,

I . aaparole

2. HCI-HOAC 424

426

Scheme 59.

425

386

388

Yoshikoshi’s Synthesis of (*.)-Nootkatone

Eremophilanes

187

the aldol condensation to obtain enone 421. Application of the Sakurai reaction to this enone affords diastereomeric ketones 422 in a ratio of 3:l. The ketone is then methylated via its sodium enolate; alkylation occurs exclusively anti to the gem-dimethyl bridge, providing isomers 423. Ozonolysis gives diones 424 and 425 in a 3:l ratio. The isomers are separated and the major isomer cyclized. The cyclobutane ring suffers concomitant ring opening to give 426. Dehydrochlorination provides (+)-nootkatone (386) and (-)-isonootkatone (388) in a ratio of about 9:l. The synthesis provides 386 in eight steps, 13% overall yield. The final synthesis we discuss in this section is that of (+)-dihydronootkatone by Caine and Graham (Scheme 60).247 Annelation of (+)-carone with methyl vinyl ketone affords enone 428 exclusively. Stereoselectivity is assured in this annelation by the steric hindrance provided by the gem-dimethyl groups. Treatment of 428 with HCl in ethanol yields chloro enone 429, which is dehydrochlorinated to obtain dienone 430. Compound 430 is transformed into the methylated dienone 431 by a straightforward sequence and 431 is subjected to photoisomerization to obtain tricyclic dienone 432. Catalytic hydrogenation

427

428

430

431

433

Scheme 60.

429

432

389

Caine-Graham Synthesis of (+)-Dihydronootkatone

188

Bicarbocyclic Sesquiterpenes; Hydronaphthalenes

reduces both double bonds. Stereoselectivity in reduction of the cyclopentenone double bond results from adsorption of the molecule on the less hindered exo-face of the bicyclo i3.1 .Olhexane system comprising rings A and B of 432. Acid-catalyzed rearrangement of 433 provides (+I-dihydronootkatone (389). The final step may be formulated as follows:

A similar angular methyl migration may be involved in the biogenesis of the eremophilanes.

(2) Fukinone and Dehydrofukinone Syntheses of fukinone (434) and dehydrofukinone (435) were reported in Volume 2 of this series.24* Full details on Piers' synthesis of H

434

435

(*hfukinone have subsequently been published.249 Shortly thereafter,

Torrence and Pinder reported a similar synthesis of (kbfukinone, which is summarized in Scheme 61.250The method employed by Torrence and Pinder to reach enone 441 is essentially identical to that used by Piers and Smillie. A somewhat different method was used to introduce the isopropylidine unit. Whereas the Piers route from 441 to (*bfukinone requires nine steps, the Pinder approach accomplishes the same task in only five steps. In all, the Torrence-Pinder synthesis requires 13 steps and proceeds in about 10% overall yield.

Eremophilanes

189

Ph

I

T

O

H

0

I . Hp-cot.

NOH

2. H 2 C r 0 F 436

C

~

2. PhNHMe

437

O

O

H

2. hydrolyrir 438

- Po cp" Ac20

G

1 . B,/VCOOEt bare c

I . HCOOEt.

I . reparate 2 . MeLi

440

439

(MeOl2CO, NOH

44 I

H2-Cat. W

O

O

M

e

442

443

441

Scheme 61.

444

434

Torrence-Pinder Synthesis of (f)-Fukinone

About the same time, Marshall and Cohen reported their synthesis of (stbfukinone, which is outlined in Scheme 62.251 o-Anisaldehyde is elaborated into cyclohexenone 449, which, after purification as its 1,4 adduct with piperidine, is treated with methyllithium to obtain allylic alcohol 450. Formolysis of 450, after the manner of Johnson, results in the bicyclic formate 451, which is transformed into unsaturated acetate 452. After allylic oxidation to enone 453, the second methyl group is introduced by cuprate addition to obtain keto ester 454. Stereoselectivity in this addition is a result of the folded nature of the cis-octalin system. In such molecules, reagents invariably attack trigonal carbons adjacent to the bridgeheads from the convex face. After removal of the keto function, the acetate is hydrolyzed, the resulting alcohol is oxidized, and the

Bicarbocyclic Sesquiterpenes; Hydronaphthalenes

190

446

447

448

I . LiAIH4

449

4 51

450

.OAc

‘OAc

HOAC, A c e 0

452

O.+.A \‘C

453

- 63I

Mo2CuLi

454

. W0lff-

Kishner

2 . Jones

OAc

3. PhjCLi Ac~O

2‘ 436

455

cAO:& 457

458

Scheme 62.

& 434

Marshall-Cohen Synthesis of ( *)-Fukinone

ketone so produced is converted into enol acetate 455. The ensuing transformations serve to transport the carbonyl group to C, while introducing the necessary isopropylidine unit at C,. This interesting construction was employed because more direct methods were unsuccessful. For example, the following sequence fails in the last step, as unsaturated alcohol 463 cannot be oxidized to (*bfukinone:

Eremophilanes

459

460

462

191

461

463

In all, the Marshall-Cohen synthesis requires 19 steps from o-anisaldehyde and produces 434 in only 0.25% yield. In 1979, Torii and co-workers reported a second synthesis of (f1-dehydrofukinone, which is outlined in Scheme 63.252 Diels-Alder reaction of unsaturated keto ester 464 and butadiene provides octalone 465 as a mixture of cis and tram ring junction isomers. The mixture may be equilibrated by base to the pure trans isomer 466. The cis disposition of the vicinal methyl groups results from a steric bias in the Diels-Alder reaction due to the secondary methyl group in the dienophile. The ketone is ketalized and the ester function is reduced and converted into a mesylate. After deprotection of the ketone, base treatment of the resulting keto mesylate 468 results in the formation of tricyclic ketone 469. Lithium-ammonia reduction of the latter compound gives unsaturated ketone 470, in which the angular methoxycarbonyl group has finally been reduced to a methyl. Eight further steps are required to convert this compound into enone 441, an intermediate employed in both the Piers and Pinder (fbfukinone syntheses. The Torii synthesis provides a vivid illustration of a trap into which synthetic chemists are prone to fall. It often happens that a worker discovers an interesting new method and, in an honest attempt to demonstrate the utility of the method, proceeds to apply it to solve a problem which has already been solved in a more satisfactory manner by some other, simpler method. In the present case, the Piers-Pinder synthesis of enone 441 requires only eight steps from a readily available aromatic precursor and provides the compound in >3W! overall yield. The Torii procedure necessitates 15 steps from a less accessible starting material, and gives 441 in less than

Bicnrbocyclic Sesquiterpenes; Hydronapbtbalenes

192

464

I . MsCl 2 . HCIO,,

____c

&&

465

466

MIO_

1 -

I . LiA IH 4

Li-NH

2. ~ - C P B A +

MI

H20

468

@

'O

Q,"+

2I . L i - N H j t

(+ &#

469

470

2. I . PCC H30+-

TH&s.OH

47 I

~

r,znclL

w6"-qg I . MsCl

2.

A

474

Scheme 63.

2. I . A.DMSO MsCl

473

472

441

467

435

Torii's Synthesis of (*bDehydrofukinone

10% overall yield. However, the Torii synthesis does make one important contribution to methodology in this class of compounds. The introduction of the isopropylidine function by directed aldol condensation and dehydration appears to be the most direct method employed to date.

(3) Isopetasol, 3-Epiisopetasol, and Warburgiadione

Isopetasol (475), 3-epiisopetasol (4761, and warburgiadione (477) are more highly oxidized fukinone derivatives. All three compounds have

Eremophilanes

193

HO'

475

476

477

been prepared by Yamakawa and co-workers as shown in Scheme 64.253 Dione 479 is obtained from 2,3-dimethylphenol via 2,3-dimethylbenzoquinone. Michael addition of 479 to methyl vinyl ketone gives a mixture of adducts 480, 481, and 482 in a ratio of 20:20:1. After chromatographic separation, aldols 481 and 482 are treated with acid to obtain enone 483, along with a small amount of the bicyclo[3.3.llnonane isomers 484. Enone 483 may be prepared from 479 in this manner in about 259'0 yield. After selective ketalization of the unconjugated carbonyl, the

2. NaHSO3 3. H2-cat. 4. Jones

HO 478

+

0 479

@'"

0

482

480

2. 1 . reporate+ (on 481 p-TsOH and 482)

+

481

I . reparate 2. (CH20H12, H+-

o& 484

483

I . (MeO)gO, 2. NaH, MeLi

n+.*oJqq OH

485

486

487

NaEy IDDQ

HO"

475

476

477

Scheme 64. Yamakawa's Synthesis of (*1-Isopetasol, (*)-3-Epiisopetasol and (*)-Warburgiadione

194

Bicarbocyclic Sesquiterpenes;Hydronaphthalenes

isopropylidine unit is introduced by methodology which is now more or less standard in the fukinone field (cf. the Piers and Pinder fukinone syntheses). Enedione 487 is selectively reduced to give (*)-isopetasol (475) and (*)-3-epiisopetasol (476) in a ratio of 1 5 . Alternatively, 487 may be dehydrogenated with dichlorodicyanobenzoquinone to obtain (*bwarburgiadione (477). Torii and co-workers have published the synthesis of (*)-isopetasol, which is outlined in Scheme 65.254 The synthesis is similar to the Torii dehydrofukinone synthesis, in that the decalin nucleus is reached by a Diels-Alder reaction and also because a cyclopropyl ketone plays a key role. In this case, however, cyclopropyl ketone 493 is cleaved reduc-

488

489

494

490

495

497

496

498

Scheme 65.

4 . KOH,>20

Torii's Synthesis of (*)-Isopetasol

475

Eremophilanes

195

tively and the resulting enolate methylated to obtain 494, an intermediate with the required vicinal methyl groups. Although the stereochemistry resulting from the reductive methylation is trans, the C2-methyl center is easily corrected by base equilibration to the more stable cis isomer, in which the secondary methyl group is equatorial. Lithium in ammonia reduction of ketone 494 affords the thermodynamically favored (equatorial) alcohol 495. The remaining steps are unexceptional and (*)-isopetasol is obtained in 19 steps in an overall yield of about 4%. (4) Eremophilone

Eremophilone (499) has stood as a special symbol to the natural products chemist of the perversity of nature. It was isolated by Simonson and co-workers in 1932. Simonson proposed an erroneous structure, ruling out the correct structure because of the isoprene rule. Later, how-

499

ever, on the basis of a suggestion by Robinson, Simonson reconsidered and eventually advanced the correct structure, which was subsequently confirmed by Grant’s X-ray analysis of a derivative. At the time of the writing of Volume 2 in this series, no successful syntheses of eremophilone had been achieved, even though it was the oldest and most wellknown member of the group. The principle synthetic challenge is controlling the stereochemistry of the isopropenyl group, which must be axial. Since 1970, there have been three total syntheses, and a fourth synthesis which leads to a key intermediate in one of the others. The first successful synthesis was reported by Ziegler and Wender in 1974 (Scheme 66).255 The synthesis begins by conjugate addition of lithium divinylcuprate to 3,4-dimethylcyclohexenone. The cis disposition of the vicinal methyl groups is assured by the well-established preference

Bicarbocyclic Sesquiterpenes; Hydronaphthalenes

196

50I

SO0

502

503

so4

2. HOAc

I . rrparatr 2 . NoOH, MeOH

SO5

506

508

I . Hz02, OH-

2. HONNHz 3. Collins

499

Scheme 66.

( f )-Eremophilone:

Ziegler-Wender Synthesis

for such conjugate additions to occur trans to a substituent at C,. Unsaturated ketone 501 is elaborated by straightforward methods into alcohol 504, which is subjected to Claisen rearrangement using the butyl vinyl ether technique. The ingenious plan is marred by the fact that the Claisen rearrangement is essentially stereorandom; the two diastereomers are produced in a ratio of 55:45. The desired isomer is the minor product of the reaction. After separation, the two diastereomers may be independently converted into ( f )-eremophilone (499) and its C,-epimer. Subsequently, Ziegler and his co-workers developed an alternative synthesis which, while not totally stereoselectiv,e, provides ( f ) -eremophilone with about 7046 control of stereochemistry of C,.256 This synthesis

Eremophilanes

197

(Scheme 67) begins with the same unsaturated ketone (501) used in the first eremophilone synthesis. It is epoxidized by the three-step sequence shown (direct epoxidation of 501 fails) and epoxy ketone 510 is cyclized by treatment with potassium r-butoxide. As it turns out, it is in this step

so9

501

I . separate 2. Mn02 *

"2'"

512

&

bl I

610

#'kHO

*'ACOOMe

513

Collins_

517

Scheme 67.

514

I . MeOH,NaOH, 2 . 260%

51 8 ( f )-Eremophilone:

& 508

Ziegler's Stereoselective Variant.

that the eventual stereochemistry of C, is effectively established. Endo and exo compounds 511 and 512 are produced in a ratio of about 70:30. The isomers are separated and the major isomer 511 is oxidized to aldehyde 513, which is converted by Wittig reaction to unsaturated ester 514. After protection of the ketone group, the isopropenyl unit is added by way of the cuprate reagent; compound 515 is the sole product of the

198

Bicubocyclic Sesquiterpenes; Hydronaphthalenes

reaction. The stereochemical outcome of this reaction is understandable in terms of the following depiction of the endo-cyclopropylacrylate:

Attack of the organometallic reagent from the front face of the double bond is severely hindered by the dioxolane ring. Standard transformations serve to convert 515 into keto alcohol 516, which is ring-opened by lithium in ammonia to give keto alcohol 517. Oxidation, ring closure, and dehydration provide enone 508, an intermediate in the first Ziegler synthesis. The minor cyclopropyl carbinol 512, when submitted to the same sequence of steps, provides enone 508 and its C,-epimer in a ratio of 15:85. In 1975, McMurry and co-workers reported a completely stereoselective synthesis of (*)-eremophilone (Scheme 68).257 Ketone 394,

519

387

394

I . Ac,O,

2. I . m-CPBA LICIO,,A-

-

522

52 I

520

523

Scheme 68.

( *)-Eremophilone:

&+

499

McMurry's Synthesis

Eremophilanes

199

prepared from P-pinene in five steps by a procedure introduced by Van der Gen, is condensed with 3-penten-2-one to give 7-epinootkatone 387 (see Scheme 53). Enone 387 is converted into diene 521, which is epoxidized and the mixture of epoxides rearranged with lithium perchlorate to give the cis-fused enone 522 (itself a minor natural product) and the trans isomer 523 in a ratio of 3:l. Conjugation of the mixture provides (~ )-eremophilone t (499). Although (+)-@-pinene is employed as starting material, racemization occurs in the course of its conversion to ketone 394. The complete synthesis from P-pinene requires 13 steps and gives eremophilone in about 3% overall yield. Ficini and Touzin synthesized (f)-eremophilone as shown in Scheme 69.258 Keto ester 524 reacts with excess methyllithium to give keto alcohol 525 after hydrolysis. The tertiary alcohol is protected and the elements of the second ring are added in the form of an organometallic reagent. In this step, the necessary cis disposition between the angular methyl and the isopropenyl groups is established with high selectivity. Since the protecting groups on the two alcohols are the same, four additional steps are required to reach keto alcohol 528, even though the overall change amounts to no more than hydrolysis of one of the two ether groups. The primary hydroxy is oxidized and the keto aldehyde is aldolized to obtain 530, a mixture of diastereomers at the secondary hydroxy position. Parallel studies had shown that catalytic hydrogenation of the exocyclic methylene group, either at this stage or after dehydration of the aldol, would probably be stereorandom. Consequently, it is necessary to arrange the functionality in the molecule so that this double bond can be reduced chemically. This requires that the ketone function be protected. Ficini and Touzin chose to reduce it temporarily to an alcohol. Of course, this choice in turn requires that the existing secondary alcohol be protected in order to differentiate it from the secondary alcohol to be produced by reduction of the ketone. For this purpose the ethoxyethyl group is again called into service. After the proper manipulations, the reduction of 531 turns out to be stereoselective, as anticipated. The keto function is regenerated, the protecting groups removed, and diol 533 is heated with alumina at 280°C to obtain a mixture of (*)-eremophilone (499, 25% yield) and the deconjugated enone 522.

200

Bicnrbocyclic Sesquiterpenes; Hydronaphthnlenes

524

526

529

526

530

Li-EtNHZb

I . Jones 2 . H,o+*

OR 531

532

499

533

Scheme 69.

(+)-Eremophilone:

522

Ficini-Touzin Synthesis

This synthesis provides a graphic illustration of the extent to which organic chemists have come to rely upon protecting groups. Of the sixteen steps in the synthesis, seven are taken up with introducing or removing hydroxy protecting groups. The final synthesis of eremophilone was reported in 1979 by the Firmenich group of Na'f, Decorzant and Thommen (Scheme 70).259 Enone

Eremophilanes

534

250°

o&

535

OOE t

538

IS MeLic

+

537'

& - ($& OOEt

ptoluene -YH

OOEt

boH & 539

540

SOCI2-C5H5N+

54 I

Scheme 70.

201

508

(*)-Eremophilone:

Firmenich Synthesis

534 is carboxylated and the resulting P-keto acid 535 is condensed with aldehydo ester 536 to give, after decarboxylation, intermediate 537. When this material is heated at 250"C, intramolecular Diels-Alder addition occurs, providing a mixture of trans- and cis-fused adducts 538 and 539. It is suggested that isomer 538 is the kinetic product and that some epimerization occurs under the conditions of the cycloaddition. The mixture is conjugated by treatment with acid to obtain enone ester 540, which reacts with excess methyllithium to give keto alcohol 541. It is remarkable that the ketone carbonyl does not react under these conditions. Dehydration of 541 provides the Ziegler-Wender enone 508. The relative configurations of the three chiral centers are established in the intramolecular Diels-Alder reaction, which is proposed to proceed by way of the following transition state:

Me

202

Bicarbocyclk Sesquiterpenes;Hydronaphthalenes

(5) Furanoeremophilanes

There exists a large family of eremophilanes in which the three-carbon side chain is incorporated into a furan or butenolide ring. Those that have been synthesized in the past 10 years are furanoeremophilane (542), furanoligularone (5431, eremophilenolide (544, tetrahydroligularenolide (5451, 9,l O-dehydrofuranoeremophilane (5461, ligularone (547), isoligularone (5481, euryopsonol (5491, epi-euryopsonol (5501, 3/3-hydroxyfuranoeremophilane (5511, and 3/3-furanoligularanol (5521.

$q@q$qel 542

543

544

546

547

548

545

549

HO

no 550

551

562

A common synthetic problem in this group is establishing the cis stereochemistry of the vicinal methyl groups. In addition, the ring junction stereochemistry must be controlled- usually to produce the cis isomer. The additional stereocenter in butenolides 544 and 545 actually presents no problem, since it is defined in the thermodynamically more stable configuration and may be equilibrated via the hydroxyfuran tautomer. Compounds 547-550 present a problem in regioselectivity, in placing the 0x0 function at the proper site. As will be seen, this problem has not been solved, and the syntheses to date give mixtures of regioisomers which must be separated. Finally, compounds 549-552 present the

Eremophilanes

203

problem of the additional chiral center at the secondary hydroxy group. This problem has also not been solved satisfactorily. The first reported synthesis of one the compounds in this group was Piers' synthesis of (*beremophilenolide (544) .260 The synthesis was later extended to produce kt 1-tetrahydroligularenolide (545) as Enone 441, used previously in the Piers and Pinder fukinone syntheses262 is converted into /3-keto ester 556 by the rather circuitous sequence indicated in Scheme 71. Torrence and Pinder report that

cp"

HCOOEt+ NoOMe

\

553

441

555

H

- cat .*

EtOH

OH

- vo DDQ

CHO

554

556

-

+ oc",*

*coo"

I . separate

2. p-780H.

- -

C6H6

658

559

560 H

561

Scheme 71.

544

H

545

Piers Synthesis of ( f )-Eremophilenolide and (k.)-Tetrahydroligularenolide

557

204

Bicarbocyclic Sesquiterpenes; Hydronaphthalenes

enone 441 may be directly carbomethoxylated to give a /3-keto ester corresponding to 556. However, the Piers sample is a solid, mp 108"C, while the specimen obtained by Torrence and Pinder is reported to be an oil. Because of reports that similar enones undergo direct carbomethoxylation both at C, and at C,, the purity of Pinder's product is questionable and Piers and co-workers chose the more complex route. Compound 556 is alkylated and the resulting keto diester is hydrolyzed and decarboxylated to give enone acid 557. Catalytic hydrogenation provides a mixture of diastereomeric keto acids 558 and 559. The ratio of 558 to 559 depends upon the catalyst employed. With Pd/C, the ratio of 558 to 559 is 3:2; with RWC, the ratio is 2:3. The two isomers are separated and converted independently into butenolides 544 and 545. In 1973, Takahashi and Tatee reported a second synthesis of (*)-tetrahydroligularenolide (545). 263 The synthesis (Scheme 72) begins with the known keto ether 562, which is obtained from dihydroresorcinol in six

565

566

567

I . separate

I

3. n,o+.

Scheme 72.

I4

A

545

Takahashi's Synthesis of (*bTetrahydroligularenolide

Eremophilanes

205

steps. Compound 562 is converted into keto ketal 563, which is methylated with methyllithium and the resulting alcohol is dehydrated to obtain a mixture of alkenes 564. Catalytic hydrogenation appears to be highly stereoselective, with ketone 565 being the only stereoisomer produced. The overall yield from 562 to 565 is 16%. Methoxycarbonylation of 565 yields a mixture of P-keto esters, which is separated chrornatographically to obtain 75% of 566 and 12% of 567. Alkylation of the former isomer with ethyl a-iodopropionate gives, after hydrolysis and decarboxylation, ( r t )-tetrahydroligularenolide (545). The alkylation proceeds in only 30% yield, the principal by-product being the O-alkylated isomer. Kitahara and his co-workers reported the synthesis of (*beremophilenolide (544) and (*1-furanoeremophilane (542) which is summarized in Scheme 73.264 Octalone 568, obtained by the Diels-Alder reaction of 2-methylcyclohexenone and butadiene, reacts with methyllithium to produce a 2:l mixture of diastereomers 569 and 570. When the mixture is subjected to oxymercuration conditions, only the major isomer reacts, leading to ether 571. Treatment of 571 with acetyl chloride in the presence of mercuric acetate as catalyst yields a chloro acetate, which, upon treatment with base, is converted into unsaturated alcohol 572. After oxidation of the secondary alcohol and protection of the resulting ketone, the double bond is hydroborated and the resulting ketal alcohol is oxidized to 574. Since hydroboration occurs from the convex face of the cis-octalin system, the stereochemistry of the C, methyl group is adjusted by epimerization with base. As shown by the following projection formulas, the interaction labeled a must be more destabilizing than the one labeled h

206

Blcrrbocyclic Sesquiterpenes; Hydronrphthrlenes

569

568

572

570

573

542

Scheme 73.

Kitahara's Synthesis (k)-Erernophilenolide and ( f )-Furanoeremophilane

The carbonyl group is removed by the Bamford-Stevens method. After deprotection, the three-carbon side chain is introduced by a Reformatsky reaction. After dehydration, a mixture of alkenes (576) is produced. Allyiic oxidation with f-butyl chromate occurs regiospecifically, giving enone 577, which is reduced by sodium borohydride to obtain (*)-eremophilenolide (544). Further reduction of the lactone, followed by dehydration, provides (*1-furanoeremophilane (542).

Eremophilanes

207

In 1976, Bohlmann and co-workers reported syntheses of (*bligularone (547), (*)-euryopsonol (549) and (f1-epieuryopsonol The synthesis of ligularone (Scheme 74) begins with p-cresol, which is converted by a nine-step sequence into furanoquinone 584. Diels-Alder

578

& 580

\I

1

579

581

Fremy'r Salt

1 1 /

OH

584

583

586

587

Scheme 74.

++ & 582

AcO

HCI,ethor MrOH+

I I

585

547

Bohlmann's Synthesis of ( *bLigularone

reaction of this material with 3-acetoxy-1,3-pentadiene provides adduct 585. Upon hydrolysis, the secondary methyl center undergoes epimerization, giving the key intermediate 586. This compound reacts with excess ethanedithiol to give the bis-dithioketal 587 in 58% yield; the more hindered carbonyl does not react. Desulfurization of 587 with Raney nickel gives (*1-ligularone.

Bicarbocycllc Sespuiterpenes; Hydronapbthrlenes

208

For the synthesis of (f)-euryopsonol (549) and (*)-epieuryopsonol (550) (see Scheme 79, the key trione 586 is carefully reduced with NaAlH2(OCH,CH20CH,), to obtain a mixture of alcohols, which is acetylated. Fractional crystallization of the mixture provides acetates 588

588

586

I.

589

rrparotr

P

2 . NaBH4

AcO

590

H&

2. Ph3P12 3. OH-

AcO

591

+

H

549

Scheme 75.

550

Bohlmann's Synthesis of ( *)-Euryopsonol

and ( f )-Epieuryopsonol

(25Oh yield) and 589 (13% yield) in pure form. The major isomer is reduced by sodium borohydride to obtain 590 (63% yield) and 591 (26% yield). Again, the two isomers are separated by crystallization. The minor isomer is deoxygenated to obtain (k)-euryopsonol (549) and (f1-epieuryopsonol (550) in 36% and 11% yields, respectively. Although the three carbonyl groups in 586 are differentiated in their reactivity toward various reagents, the difference in reactivity is not sufficient for the synthesis of 549 in a highly selective manner. Yamakawa and Satoh have also explored the use of triketone 586 for the synthesis of various furanoeremophilanes.266 As shown in Scheme 76, they prepared 586 in a similar manner to that employed by Bohlmann and co-workers. In agreement with Bohlmann's observation, Yamakawa found the C, carbonyl to be more reactive toward ethylene glycol; the crystalline 3-ketal is obtained in 88% yield. However, no

Eremophilanes

584

586

-

209

592 H QAC

2. Ac20-

693

594

195

I . (HSCHzIz, BFS

2. Raney Ni

0 596

Scheme 76.

547

Yamakawa's Synthesis of ( *)-Ligularone

regioselectivity is observed in reduction of this rnonoketal, isomers 592 and 593 being produced in equal amounts. The isomers are separated and isomer 593 is converted into dione 596 by removal of the hydroxy group. For the synthesis of (f)-ligularone (547), advantage is again taken of the greater reactivity of the C3-carbonyl. Yamakawa and Satoh have also converted triketone 586 into alcohols 551 and 552, as shown in Scheme 77. Hydrolysis of isomer 592 (see Scheme 76) gives diketo alcohol 597 in 75% yield. Deoxygenation of this substance gives isomeric diones 598 (20% yield), 599 (20% yield), and 600 (30% yield). Reduction of 598 by sodium borohydride occurs with good stereoselectivity, providing keto alcohol 601 in 81% yield, along with 7% of its C1-epimer. Further reduction of 601 gives diol 602, which is converted to a diacetate and hydrogenolyzed to obtain (J?)-3@-hydroxyfuranoerernophilane (551). Similar treatment of isomer 599 gives keto alcohol 603 (93% yield), which is further reduced and hydrogenolyzed to obtain (st -3P-furanoligularanol (522). Diones 598 and 599 have also been converted into compounds 542, 543, and 546.

210

Bkarbocycllc Sesquiterpenes; Hydronrpbthslenes

I'"

592

597

O

[o&+o&+&' 5998

599

I 600

NOBH 4

HO

HO

& 1

603

601

1

LiAIH4

LiAIH4

H&

H

602

1

I . Ac~O.C&N

2 . Birch

H

551

Scheme 77.

603

1

I . AcZO. %HSN

2 . Birch

H

552

Yamakawa's Synthesis of ( f )-3~-Hydroxyfuranoeremophilane and (*1- 3~-Furanoligularanol

Eremophilanes

211

The final synthesis in this area is that due to Yoshikoshi and his coworkers, which produces (*)-ligularone and (*)-isoligularone (Scheme 78) .267 Enedione 604, conveniently prepared by Diels-Alder reaction of 2-methyl-cyclohexenone and the Danishefsky diene is ketalized at the nonconjugated carbonyl and treated with lithium dimethylcuprate to obtain keto ketal 605. As in the Marshall fukinone synthesis, the cisstereochemistry of the vicinal methyl groups results from the folded nature of the octalone. After removal of the exposed carbonyl group, the other carbonyl is deprotected and converted into enone 607. This

9 ""3 '$

I . (CH20H12, H,' 2 . Me2CuLi

2. I . H30 A +r

+

'& U

Me

so5

604

607

606

606

SPh

'SPh 610

611

I. NaI04, ditto

547

Scheme 78.

548

Yoshikoshi's Synthesis of (fbLigularone and ( f )-lsoligularone

212

Bicarbocyclic Sesqulterpeaes; Hydronaphthalenes

material is subjected to alkaline epoxidation, and the mixture of diastereomeric epoxides is reduced with lithium in ammonia. The resulting mixture of diols is oxidized to dione 608. Michael reaction of 608 with unsaturated nitro compound 609 provides 610 and 611, each a mixture of diastereomers, in a ratio of 1:2. Oxidation of 610, followed by thermolysis of the resulting sulfoxide gives (*)-liguIarone (547) in 47% yield. Similar treatment of isomer 611 provides (*)-isoligularone (548) in 54% yield. (6) Cacalol

Cacalol (612) is an example of a sesquiterpene in which a methyl group has undergone a net 1,3-migration. Three syntheses of this novel compound have been reported. In spite of its apparent simplicity, cacalol has proven to be a formidable target. As will be seen, none of the three syntheses are very efficient.

& .' 1

612

The first published synthesis (Scheme 79) came from Inouye, Uchida, and Kakisawa of Tokyo Kyoiku University.268 2-Methoxy-5-methylphenol is subjected to Friedel-Crafts reaction with y-valerolactone to give an acid which is methylated and cyclized to obtain ketone 614. The carbonyl group is removed by catalytic hydrogenolysis and the resulting tetralin 615 is acylated to obtain 616. Unfortunately, compound 616 is the minor product of this reaction, being obtained in only 20% yield. The major product, obtained in 53% yield, is the ester PI. :wed by demethylation of the C1-methoxy group, followed by acetylation of the resulting phenol. The methyl ethers are removed with BBr, and the resulting bisphenol is converted to the bis-aryloxyacetic ester 617, which is hydrolyzed and the resulting diacid is cyclized by treatment with acetic anhydride. Removal of the other carboxymethyl group is accomplished

Eremophilanes OMe

0

OMe

61 3

HOAc,

2. MezSO4

HClOg

213 OMe

614

2. K2C03'

c

Br*COOMe

615

616

1 . OH-

0

I . LiAIH4

2. AcZO,

NaOAc 61 7

619

618

QH

61 2

Scheme 79.

(rt )-Cacalol: Tokyo Synthesis

by conversion into bromide 619, which fragments upon being metallated, thus giving (f)-cacalol (612). Walls' synthesis (Scheme 80)269also begins with a Friedel-Crafts reaction, but with 2-methoxy-Cmethylphenol. Use of BF, allows both bonds to be formed, resulting in tetralone 620, albeit in only 13% yield. The elements of' the furan ring are introduced by Claisen rearrangement of crotyl ether 621. Ozonolysis of the resulting rearrangement product gives hemiacetal 623, which undergoes dehydration upon being heated with phosphoric acid. Clemmensen reduction of 624 removes the carbony1 group and provides (f)-cacalol (612).

214

Bicarbocyclic Sesquiterpenes; Hydronaphthalenes

620

62 I

623

622

624

?H HOAc

612

Scheme 80.

(*)-Cacalol: Walls' Synthesis

Huffman's synthesis (Scheme 81) 270 utilizes p-cresol for the FriedelCrafts cyclization (which proceeds in only 26% yield), thus necessitating the subsequent introduction of the other ring oxygen. After Clemmensen reduction of the tetralone carbonyl, tetrahydronaphthol 626 is oxidized at C, by acylation and Bayer-Villiger oxidation. The eventual furan ring is added by alkylation of phenol 629 with a-chloroacetoacetic ester. However, compound 630 is obtained in only 14% yield. Although Walls found that a compound analogous to 630, but lacking the ethoxycarbonyl group, fails to cyclize, Huffman was successful in cyclizing 630. After the appropriate further manipulations, ( Ahacalol (612) is produced. Of the three syntheses, the latter appears to be the least efficient, requiring nine steps and producing (*I-cacalol in only 0.17% overall yield. The Walls synthesis requires six steps and gives the final product in 1.5% yield. The Tokyo synthesis requires 13 steps. Although yields are not reported for most of the steps, one of the key steps, FriedelCrafts acylation of 615, occurs in only 20% yield.

Miscellaneous Hydronaphthalenes

MeZSOe K2C03

#

625

2. I . m-CPBA, KOH

($f

628

629

626

215

627

etL NOH

I . KOH, MeOH

630

63 I

632 OH

DMF,

A

Scheme 81.

61 2

(*1-Cacalol: Huffman’s Synthesis

E. Miscellaneous Hydronaphthalenes (1) Valeranone and Valerane

Valeranone (633) is a novel sesquiterpene which, although isoprenoid, does not fit the “head-to-tail isoprene rule.” It has attracted considerable synthetic attention, and several synthetic approaches were reported in Volume 2 of this series. In this section, we will discuss one additional synthesis of 633, as well as two papers which deal with the synthesis of the parent hydrocarbon, valerane (634).

216

Bicarbocyclic Sesquiterpenes; Hydronaphthalenes

633

634

In 1973, Banerjee and Angadi published the synthesis outlined in Scheme 82.271 Conjugate addition of methylmagnesium bromide to the well-known octalone 635 provides keto ester 636, which is converted by sodium hydride and diethyl carbonate into 637. The ketone is removed by desulfurization of the derived dithioketal and the resulting diester 638 is treated with excess methyllithium. After oxidation of the secondary alcohol, hydroxy ketone 639 is produced. Dehydration of 639 yields a

(EtO)2C0,

NaH

*

COOEt EtOOCO

AC

636 I . (CH2SH12, BF3 2. Roney Ni

COOEt

637 I . x s MeLi 2 . Collins +

KHS04,

A

c

EtOOC

638

640

Scheme 82.

639

633

64 I

Banerjee’s Synthesis of (& )-Valeranone

mixture of four unsaturated ketones (640), which is hydrogenated to obtain 2OYo (*)-valeranone (633) and 80% of its diastereomer 641. In light of this result, the author’s claim of a “stereoselective total synthesis of (*)-valeranone ...” seems specious.

Miscellaneous Hydronaphthalenes

217

Posner and his co-workers reported the interesting synthesis of valerane outlined in Scheme 83.272 Enol ether 642 is converted into cyclohexenone 643 by a well-precedented method. This material is

dT

0

OTHP

2. MsCI,CsHsN

I . THPO-MgC’ 2. H30+

642

643

I Me2CuLi

Reduction:

I

‘.f

2. HMPT,

@.,,

00, 2hrs

644

645

Scheme 83.

634

Posner’s Synthesis of ( *)-Valefane

transformed into enone bromide 644, which is subjected to conjugate addition with lithium dimethylcuprate. The resulting enolate is treated with hexamethylphosphoric triamide whereupon internal alkylation occurs to produce decalone 645 in 25% yield. Removal of the carbonyl group provides (*1-valerane (634). Although this synthesis is brief, it is not clear how it would be applied to valeranone itself, the only important compound of the valerane class. Baldwin and Gawley explored the interesting synthesis outlined in Scheme 84.273 Photoaddition of 1,2-dimethylcyclohexene and p-keto aldehyde 646 produces keto aldehyde 647 in quantitative yield. Acidcatalyzed aldolization gives enone 648, which is hydrogenated to decalone 649. Wittig reaction of 649 gives alkene 650, which is hydrogenated to a mixture of 40% (f)-valerane (634) and 60% of its diastereomer (674). Attempts to significantly alter this ratio by hydrogenating other alkene isomers were not successful. Thus, Baldwin confirms Banerjee’s finding that catalytic hydrogenation of an alkene is not the way to establish the correct valerane stereochemistry.

218

Bicarbocyelic Sesquiterpenes; Hydronaphthalenes

Scheme 84.

Baldwin’s Synthesis of (f)-Valerane

(2) Khusitene and fl -Gorgonene

Khusitene (675) is a nor-cadinane which occurs in Indian vetiver oil. The nonisoprenoid P-gorgonene (676) may be a rearranged eudesmane. In this section, we discuss two syntheses of the former and one of the latter.

675

676

The first synthesis of (f)-khusitene came from the laboratories of 0. P. Vig at the Punjab University in Chandigark. The Vig synthesis is summarized in Scheme 85.274 Keto aldehyde 676, obtained from the piperidine enamine of butanal and ethyl acrylate, is condensed with methallyltriphenylphosphorane to obtain diene ester 677. This material undergoes a Diels-Alder reaction with ethyl acrylate to produce adduct 678. After Dieckman cyclization, decarboxylation, and Wittig methylenation, (f1-khusitene is obtained.

Miscellaneous Hydronaphthalenes

219

LOOEt

,COOEt

677

676

678

t-BuOK ~ - B ~ o H ~

679

680

6 75

Scheme 85.

Vig's First (*)-Khusitene Synthesis

In 1977 Vig reported the alternative synthesis outlined in Scheme 86.275 Diene aldehyde 686, prepared by the straightforward route depicted, is condensed with vinylmagnesium bromide to obtain an alcohol (687) which is oxidized by manganese dioxide. The resulting trienone

Yo-

EtQOC

E t o o C ~ c o o E&CHO, t

E,OOC

68 I

I. LiAlHg 2. PCC ~

I . fCH,OH),, H+ 'c 2. NaCI,DMSO

682

OHC?

b P P h S t

E

t

O

683

686

\

687

Scheme 86.

p

9'"

685

684

O

688

Vig's Second ( *1-Khusitene Synthesis

675

220

Bicarbocyclic Sesquiterpenes; Hydronaphthalenes

undergoes spontaneous intramolecular Diels-Alder addition, giving an octalone. Although the question of stereochemistry is not addressed, the predominant isomer is presumably 688, based on Taber's experience with the analogous isopropyl compound (see Scheme 42). Wittig methylenation occurs with epimerization to the nuns-octalin system, as usual. Thus, the final product of this synthesis is probably diastereomer 675, although the stereostructure of natural khusitene is still unknown. ( f )-P-Gorgonene has been prepared by Boeckman and co-workers as shown in Scheme 87.276 The known decalone 692 is prepared by wellestablished reactions. Chlorination of 692, followed by dehydrochlorination with hot quinoline affords enone 693. Conjugate addition of isopropenylmagnesium bromide provides a 7:2:1 mixture of ketones 694, 695, and 696. The major isomer, 694, is isolated in 45% yield. Methylenation of 694 by the Peterson method provides (f)-P-gorgonene (676).

I . SO2CIZ, CCI4

2. quinoline,

A * 693

I

69 4

695

I. Me3SiCH2MgCI 2. HOAc

676

Scheme 87.

Boeckman's ( f )-p-Gorgonene Synthesis

696

Hydronaphthalenes Containing an Additional Cyclopropane Ring

221

F. Hydronaphthalenes Containing an Additional Cyclopropane Ring In this section, we discuss syntheses of p,P-cycloeudesmol (697), p,a-cycloeudesmol (6981, a,p-cycloeudesmol (6991, mayurone (700), thujopsene (701), and thujopsadiene (702).

%OH

697

700

H&@

698

70 I

%OH

699

702

There are four diastereomers corresponding to the cycloeudesmol structure. In 1974, an antibiotic substance was isolated from a marine alga by Fenical and Sims and proposed to have the cycloeudesmol structure. However, the relative stereochemistry of the natural product was not determined. This proposal prompted several groups to synthesize various isomers. As it turned out, none of the synthesized isomers 697-699 are identical with the natural product. Hence, natural cycloeudesmol is suspected to have one of the a,a-structures 703.

703

Moss and co-workers synthesized the P,p and P,a isomers 697 and 698 as shown in Scheme For synthesis of the p,p diastereomer, the known278unsaturated ketal 704 is employed. It is known that this compound undergoes addition reactions preponderantly from the top face of the double bond.278 Thus, reaction with dichlorocarbene under conditions of phase transfer catalysis gives a dichlorocyclopropane which is reduced by sodium in ammonia to tricyclic ketal 705. After ketal hydrolysis, the carbonyl group is reduced to give predominantly alcohol 706. Replacement of the hydroxy group, by reaction of the tosylate with

222

Bicarbocyclic Sesqulterpenes; Hydronaphthalenes

704

705

706

.

I KOH, (CH2OH)2

2. HCI

I . P-TICI

-.CN

2 . KCN, HMPT

MeLi

3. M e I , OH-, HMPT

a

0

707

0

.

e

708

697

y

0

709

I . LiAIH4 2. Simmonr- * Smith 3. Jones

Scheme 88.

bcf3-.f

0

.

I Wolff-Kirhner 2 . m-CPBA

710

Moss Synthesis of Cycloeudesmol Isomers

cyanide ion, is accompanied by some loss of stereochemistry at C,. However, this is of no consequence as nitrile 707 undergoes ready epimerization upon hydrolysis, yielding after esterification, methyl ester 708. Reaction of this compound with methyllithium gives ( f 1-fi,fi-cycloeudesmol (697). For synthesis of the P,a diastereomer, Moss and co-workers employed octalone 709, readily prepared from dihydrocarvone and methyl vinyl

Hydronaphthalenes Containing an Additional Cyclopropane Ring

223

ketone. It is well-known that dienone 709 has the isopropenyl and angular methyl groups trans.279 Reduction of the carbonyl gives the equatorial allylic alcohol which undergoes assisted Simmons-Smith cyclopropanation to provide, after oxidation, the cyclopropyl ketone 710. The HC1' HoAcb

@*f

0

OH

o,

I . (CH,OH),,

H+

2. LiAIH(f-BuOt3

713

712 I . Simrnons-

Smith 2. Collins

H

e

0

714

I , WOlff Kishner 2. H30'

o&y

*

71 5

714

717

I . Simmonr-

I . Wolff-

Smith

Kishner

HO' 718

719

720

Scheme 89.

699

Ando Synthesis of Cycloeudesmol Isomers

224

Bicarbocyclic Sesquiterpenes; Hydronaphthalenes

carbonyl group is removed and the isopropenyl group is hydrated to obtain ( f 1-P,a-cycloeudesmol (698). Ando and co-workers prepared all three isomers 697-699.280 As shown in Scheme 89, the synthesis of the &/3 diastereomer is different from that employed by Moss. Compound 712281 is dehydrated with subsequent epimerization of the acetyl group to give enedione 713. Ketalization of the saturated carbonyl, followed by reduction of the enone gives the equatorial allylic alcohol 714. The remaining steps are similar to those used by Moss for the /3,a isomer. The &p stereochemistry is obtained by the assistance of the /3-hydroxy in the Simmons-Smith reaction. To make the a,/3 isomer, Ando and co-workers first invert the allylic alcohol by the Mitsunobu method. Simmons-Smith reaction on 718 gives the a-cyclopropyl ketone 719. The synthesis is completed in the normal manner to give (*)-a,/3-cycloeudesmol (699). The Japanese group prepared the B,a diastereomer 698 by the identical sequence employed by Moss and co-workers. Caine and co-workers prepared the p,p isomer by the efficient route summarized in Scheme 90.282Chloroenone 429 (see Scheme 60) is sol-

72I

429

I , Simmons-Smith

2. Jones

H

Wolff -Kishner

*

722

0

&OAc

723

697

Scheme 90.

Caine Synthesis of (+)-p,P-Cycloeudesmol

Hydronaphthalenes Containing an Additional Cyclopropane Ring

225

volyzed to obtain acetate 721. The remaining steps in the synthesis are essentially the same as are used in the other syntheses discussed.* McMurry and Blaszczak have synthesized (fbmayurone (700) as outlined in Scheme 91.283Aldehyde 724 (an intermediate in Buchi’s thujopsene synthesis)284is converted into an ester, which is alkylated with ally1 COOMe

724

725

726

700

702

3 . H30+ 701

Scheme 91.

McMurry’s Synthesis of (+ )-Mayurone,

(*)-Thujopsene and (*)-Thujopsadiene

‘In 1981, Ando and co-workers synthesized the fourth stereoisomer (a,a) and found that it is also not identical with the natural product [M. Ando, S. Sayama, and K. Takese, Chem. Lett., 377 (1981)l. The structure is now known to be:

[T. Suzuki, A. Furusaki, H. Kikuchi, E. Kurosawa, and C. Katayama, Tetrahedron Lett., 22, 3423 (198111.

226

Bicnrbocyclic Sesquiterpenes; Hydronnphtbnlenes

bromide. The ally1 group is cleaved to an acid (723, which is converted into a diazoketone (728). Copper-catalyzed decomposition of 728 gives a carbene which inserts in the double bond to give 729. This step is based on Biichi's earlier o b ~ e r v a t i o n ~that ' ~ such intramolecular carbene insertions provide the correct thujopsene stereochemistry. The novel transformation in the McMurry synthesis is the oxidative decarboxylation of y-keto ester 729, which is brought about by treating the acid with lead tetraacetate. The ester function thus serves as a latent double bond. The product (*bmayurone (700) is converted into ( I t 1-thujopsadiene (702) and (*1-thujopsene (701) by literature procedures. Smith and co-workers have also prepared kt)-mayurone (700), as summarized in Scheme 92.285The Smith synthesis is a showcase for diazoketone chemistry, such intermediates being used on three occasions. The first use is in preparation of ester 725 (an intermediate in the McMurry synthesis). In this case, diazoketone 734 is rearranged under

7 30

731

I. CCOCl)~

2. CH2N2

732

Cu(acac)L w

N

7 34

2

7 33

COOMe

A,MeOH

725

2. CH2Np

735

2. Raney-Ni 739

Scheme 92.

740

Smith's Synthesis of (+-)-Mayurone

700

Hydronaphthalenes Containing an Additional Cyclopropane Ring

227

the influence of Cu (acac)*, presumably via the intermediate bicyclopentanone 741. In the second use, diazoketone 736 is subjected to photochemical Wolff rearrangement as a method for homologation of the

734

741

742

side chain. Finally, diazo ketone 739 is subjected to copper-catalyzed decomposition, leading to cyclopropyl ketone 740, as in the McMurry synthesis. It is interesting to compare the efficiency of the four syntheses which have been reported in this group. Since the earlier workers did not prepare mayurone, we compare on the basis of thujopsene. The 1963 Dauben synthesis of thujopsene requires seven steps from octalone 743 7 rteps

(7.3%)

0 743

7 0I

and gives thujopsene in 7.3% yield. Buchi’s 1964 synthesis requires nine steps from cyclocitral (744) and proceeds in 0.8% overall yield. The 9 steps

744

701

McMurry synthesis is 16 steps from cyclocitral and gives the final product in 32% yield. Finally, the Smith synthesis also requires 16 steps and

acHo 16 steps

7izr

744

gives (f1-thujopsene in 5.7% overall yield.

701

228

Other Bicnrbocycllc Sesquiterpenes

Go0q& 16 rtepr

r 30

701

5. OTHER BICARBOCYCLIC SESQUITERPENES

A. Isolated Rings (1) Taylorine and Hypacrone

Taylorine (756) was synthesized from ( f )-A3-carene (745) by Nakayama and co-workers as is outlined in Scheme 93.286The steps to ester aldehyde 747 proceed without incident. Unfortunately, the Wittig reaction leading to 748 is complicated by the formation also of the acyclic isomer 749. The two isomers are obtained in yields of 38% and 17%, respectively, after “repeated chromatography”. Intermediate 748 is elaborated as shown in the scheme. The cyclpentenone ring is added by the six-step sequence 751-754 in 16% overall yield. The synthesis is completed by a low yield (13%) Wittig reaction on 754, followed by oxidation of the side-chain hydroxy to obtain (-1 -taylorine. Although all yields are not published, the overall yield for this 15-step conversion of carene into taylorine is no greater than 0.5%. Hypacrone (7631, a novel seco-iliudoid from the fern Hypolepis puncrara Mett., was synthesized by Hayashi and co-workers by the brief synthesis summarized in Scheme 94.287 Unsaturated acid 757, conveniently prepared by alkylation of the dianion of isobutyric acid, is cyclized by way of its acyl chloride to cyclopentenone 758. This material is converted into the dienol ether 759, which is condensed by the method of Mukaiyama with diketone 760. Adduct 761 is obtained in 30% yield after chromatography on silica gel. Dehydration of 761 occurs upon

Isolated Rings

,A

I . 0, 2 . MeOH,* H+

2. KZC03, MeOH 3. PCC

MeOOC/.'& H

745

I . PhC03H+

229

FHO

MeOOC/'&H H

746

747

748

749

750

751

752

753

754

755

Scheme 93.

756

Nakayama's Synthesis of (-)-Taylorine

preparative gas chromatography at 190°C to give a dienone in 60% yield. Not unexpectedly, the more stable E diastereomer 762 is the sole produce of the reaction. Sensitized photolysis of 762 produces a photostationary state (1:l) of 762 and hypacrone (7631, the Zdiastereomer. At the time of publicationz8' isomers 762 and 763 had not been separated, and the production of hypacrone was inferred only from the 'H-NMR spectrum of the mixture.

230

Other Bicnrbocyclic Sesquiterpenes

757

758

759

&qpJfj..-., 0 76 I

762

Scheme 94.

763

Hayashi’s Synthesis of Hypacrone

(2) Cuparene, a -Cuparenone, and p -Cuparenone During the period of this review, the cuparanes cuparene (7641, a -cuparenone (765) and p -cuparenone (766) have been synthesized. The synthetic challenge is the same in each case-creation of the adjacent quaternary centers. A variety of methods have been used to solve this problem.

764

765

766

Vig and co-workers have reported a synthesis of (*1-cuparene (Scheme 95).288 The synthesis begins with addition of the p-tolyl Grignard reagent to p-keto ester 767 to give an alcohol which is allegedly dehydrated to 768. However, their report should be taken with caution, since 768 is described as a “conjugated ester” and ascribed a carbonyl absorption of 1710 cm-l. The remaining methyl group is said to be introduced by coupling of the tertiary, benzylic bromide 769 with lithium dimethylcuprate.

Isolated Rings

770

231

771

76 4

Scheme 95.

Vig’s Synthesis of (*)-Cuparene

Bird, Yeong, and Hudec, in England, have prepared (fbcuparene as shown in Scheme 96.289 Cyclohexene 772, obtained from 2,2-dimethylcyclohexanone and p-tolyllithium, is epoxidized to obtain 773. Upon treatment with BF, etherate, 773 isomerizes to a mixture of aldehyde 774 (22%) and ketone 775 (25%). Wolff-Kishner reduction of 774,

772

773

774

I . separate

2. Wolff-Kirhner

764

Scheme 96.

Bird-Yeong-Hudec Synthesis of (It 1-Cuparene

775

232

Other Bicarbcyclic Sesquiterpenes

obtained by preparative thin layer chromatography, provides (*1-cuparene. De Mayo and Suau synthesized (f)-cuparene by the inventive route outlined in Scheme 97, which incorporates the photocyclization of thione

776

779

777

778

3

780

78 I

Wolff-Kithnrr-

782

Scheme 97.

764

De Mayo’s Synthesis of (rtbcuparene

778 as a key step.290 The crude benzylic thiol 779 is treated with mercuric acetate to accomplish the elimination of hydrogen sulfide. The remainder of the synthesis is straightforward, proceeding through cyclopentanone 781, which is methylated to establish the second quaternary center. Kametani and co-workers prepared the de Mayo ketone 781 as summarized in Scheme 98.291 Coupling of 5-methyl-2-furyllithium with p-xylyl bromide affords a furan (783), which is hydrolyzed to dione 784. The aldol closure of the latter substance is regioselective, due to the greater acidity of the benzylic protons. The second methyl is introduced by cuprate addition, leading to ketone 781. Wenkert and co-workers have synthesized (f )-a-cuparenone (765) as shown in Scheme 99.292The synthesis begins with aldehyde 786, which is obtained in 40% yield by the chromyl chloride oxidation of p-cymene.

233

Isolated Rings

I . n-BuLi

H2°*H2S04,

783

-

M02CuLi+

base

9

784

aee Scheme 97t

785

fl

781

Scheme 98.

764

Kametani’s Synthesis of (f)-Cuparene

Decomposition of diazoacetone in the presence of enol ether 788 provides a mixture of diastereomeric cyclopropyl ketones (789), which is ring-opened by treatment with HCI. The resulting keto aldehyde is aldolized to give enone 790, which is methylated to obtain 791. Saturation of the double bond affords (f)-a-cuparenone.

&(786

\

Cu,

A

I. HCI

\

Me

788

787

2. KOH,

MeOH

789

79 I

Scheme 99.

790

765

Wenkert’s Synthesis of (*)-a-Cuparenone

234

Other Bicarbocyclic Sesquiterpenes

(*)+?-Cuparenone has been prepared by Mane and Krishna Rao as

outlined in Scheme The Grignard reagent derived from chloride 792 undergoes conjugate addition with the Knoevenagel adduct from acetone and ethyl cyanoacetate. Hydrolysis and decarboxylation of the

CN

I . Mg 2 . wCN

2 . KOH,

792

I . SOClZ

cu2s04b A

&Nz

2 . CHpN2

795

Scheme 100.

A\’

&

&cooH 794

766

Krishna-Rao’s Synthesis of (*)-B-Cuparenone

crude adduct (793) provides acid 794 in 50% overall yield. The carbene derived from diazoketone 795 undergoes insertion into the tertiary, benzylic C-Hbond to give (f)-fl-cuparenone (766) in 39% yield, based on acid 794. It should be noted that this reaction is one of only a small number of examples of the insertion of an a-keto carbene into a u-bond. The synthetic P-cuparenone is accompanied by several byproducts, no doubt resulting from other reactions of the intermediate carbene. Casares and Maldonado have reported the extremely interesting synthesis of (f)-fl-cuparenone, which is summarized in Scheme 101.294 The synthesis begins with an extraordinary reaction-coqbgate addition of the preformed enolate of protected cyanohydrin 796 to mesityl oxide. The adduct 797 is obtained in 7040 yield. This material is converted into the known enone 799, which fails to undergo conjugate addition with lithium dimethylcuprate. Consequently, the final methyl is introduced in an indirect manner, via cyclopropyl ketone 802. The reductive cleavage

Isolated Rings

235

NHO 796

798

7 97

799

801

802

800

803

766

Scheme 101.

Casares-Maldonado Synthesis of ( *)-P-Cuparenone

of 802 provides a striking demonstration of the importance of orbital overlap in this reaction; the peripheral bond is cleaved even though cleavage of the ring fusion bond would lead to a benzylic radical.

(3) Laurene and Apbsin Laurene (804) and aplysin (805) are related to the simpler cuparanes, but have suffered a methyl migration. Both molecules present the problem of relative stereochemistry. In laurene, the methyls must be made trans, the less stable relative configuration. In aplysin, the secondary methyl is endo with respect to the bicyclo[3.3.0loctane system, also the less stable configuration.

Other Bicarbocyclic Sesquiterpenes

236

801

804

An attempted synthesis of 804 by Irie and co-workers went astray when oxidation of alcohol 806 with Sarett's reagent gave the epimerized ketone 807.295 Subsequently, McMurry and von Beroldingen modified Sorettb

#--OH

806

807

the basic hie approach, as summarized in Scheme 102.296The requisite relative stereochemistry is established by hydroboration of cyclopentene

808

811

Collins

~

812

fl

-I . PhSCH2Li, 2 . seporote

816

Scheme 102.

810

809

fi

813

FSPh

H

(PhCO)zO

804

McMurry's Synthesis of (*)-Laurene

+

Isolated Rings

819

817

820

821

823

Scheme 103.

237

822

805

Ronald’s Synthesis of (+)-Aplysin

811; a 4:l mixture of diastereomers 812 and 813 is produced. McMurry found that, unlike the Sarett procedure (compare 806+807), Collins oxidation of the hydroxy group is not accompanied by epimerization of the methyl group; a 4:1 mixture of ketones (814) is obtained. Although simple Wittig methylenation of this mixture causes ex.tensive epimerization, ketones 814 react with phenylthiomethyllithium without isomerization. The diastereomers are separated at this point and the major isomer is eliminated by Coated procedure to obtain (f1-laurene (804). Ronald has synthesized (*)-aplysin by the novel route shown in Scheme 103.297Tigloyl chloride reacts with acetylene and aluminum chloride to give a 3:2 mixture of the chlorocyclopentenone 817 and its diastereomer. After separation of these isomers, by silica gel chromatography, compound 817 is condensed with the aryllithium reagent 818 to obtain chlorohydrin 819. Treatment of this material with hot KOH affords allylic alcohol 820, which is treated with phosphorus tribromide

238

Other Bicarbocyclic Sesquiterpenes

and the crude product immediately added to excess methylmagnesium bromide. Compound 821 appears to be the only isomer produced in this remarkable sequence, in 20% yield. The correct sense of chirality at the final stereocenter is established by isomerization of the double bond to the more substituted position, followed by catalytic hydrogenation to give (*bdebromoaplysin (823). Bromination of 823 provides (fl-aplysin. (4) Trichodiene, Norketotrichodiene, 12,13-Epoxytrichothec- 9-ene,

Trichodermin, and Trichodermol The trichothecane antibiotics are a group of oxygenated sesquiterpenes produced as metabolites of various fungi. They have elicited considerable interest due to their cytotoxic and phytotoxic effects. Syntheses have been recorded for trichodiene (824, a presumed biosynthetic intermediate), norketotrichodiene (829, 12,13-epoxytrichothec-9-ene(8261, trichodermin (827a). and trichodermol (827b).

824

825 PR

826

8 2 7 0 R Ac 827b R = H

The first publication in the area was the Colvin-Raphael synthesis of (k1-trichodermin, which sets important precedents for further efforts in As shown in Scheme 104, the synthesis begins with the preparation of the key lactone 832 from p-cresol methyl ether. Two

Isolated Rings

033

836

034

036

032

OEt I . NaBH4 2. Na-NH3

239

HOEt & 037

I . H30t ' ,

\

H

,COOH

2. Collins

'.OHOH

acH0 H

039

030

040

PH

LiAlH(I-BuO)3, H

+

H

H

fl

@

041

PAC

?H

Ph3P=CH2

H

044

Scheme 104.

C5H5N

H

042

@

I. m - c p ~ ~

043 pAc

@

2. ACpO

H

845

I , separate 2. Ac20,

H

8270

Colvin-Raphael Synthesis of (+1-Trichodermin

240

Other Bicarbocyclic Sesquiterpenes

routes to this important intermediate were developed, one proceeding by way of cyclopropane 830, and the other by way of the Diels-Alder adduct 833. Methylation of 832 provides a single diastereomeric product, presumably 837. This lactone is manipulated, via hemiketal 838, to acetal 839. Upon hydrolysis of the acetal, cyclization occurs. Oxidation of the resulting hydroxy aldehyde affords keto acid 840, which reacts with acetic anhydride to give the enol lactone 841. Reduction of the latter compound gives a mixture of keto aldehyde 812 (52%) and aldol 843 (7%). Unfortunately, compound 842 cannot be induced to cyclize to provide more of aldol 843. The reason for this failure is not clear, since the various diastereomers of 842 should be in equilibrium under the reaction conditions. However, it is not certain that 843 even arises by aldol cyclization. It is at least conceivable that it may form by “alkoxide-assisled [1,31sigmatropic rearrangement ”:

The Colvin-Raphael synthesis is completed by methylenation of the acetate of 843, followed by epoxidation of alcohol 845. Good selectivity is observed in the epoxidation, presumably because the proximate hydroxy group guides the reagent to the exocyclic double bond. In 1974, Fujimoto and co-workers reported a synthesis of (&)-12,13-epoxytrichothec-9-ene, a metabolite identical to trichodermin but lacking the acetoxy function.299 The synthesis begins with the Michael addition of p-keto ester 846 onto crotonaldehyde (Scheme 105). The resulting aldehyde is converted into an acetal and the enone carbonyl is reduced by the Meerwein-Pondorf method. Upon acidification, the cis-fused dihydropyran 848 is obtained, apparently as a single stereoisomer. Although the stereochemistry at the secondary methyl center is unimportant, as it is destined to be lost at a later stage, it is noteworthy that a single isomer is produced in the Michael reaction. The stereochemistry is the same as that observed by Marshall in a

p 2

1 , &CHO b

NaOMe

I , MeerweinPondorf 2 . H30+

2. (CH20H)2, H+ 847

&+& mH

848

8 50

849

851

O

H

852

I . (PhO)3PCH31-, + HMPT,

H

A

2 . Raney N i

241

H

846

H

a

Isolated Rings

I . separate

-

‘OH

&OH

2. Os04-KC103 3. N a I 0 4

@

NaOMe

H

853

P:C;CHp+

0 H

855

854

856

m-CPBA CH2C12 H 826

Scheme 105.

Fujimoto’s Synthesis of (~)-12,13-Epoxytrichothec-9-ene

related reaction.300 Epoxidation of 849 occurs on the more nucleophilic enol ether double bond. Pyrolysis of the resulting mixture of diastereomeric epoxides provides ketone 850. Allylation occurs, rather unexpectedly, on oxygen. However, Claisen rearrangement rectifies the situation and affords a 2:l mixture of diastereomers 851 and 852. The isomers are separated by silica gel chromatography and the ally1 double bond of the major isomer is cleaved. Treatment of the resulting product (853) under basic conditions affords tricyclic aldol 854 in 90% yield. The

242

Other Bicarbocyclic Sesquiterpenes

aldol is deoxygenated and the ketone is methylenated by the Wittig method. Epoxidation of 856 provides 826 (30% yield), along with 40% of recovered 856. No comment is made regarding oxidation of the other double bond. However, later work by Kamikawa suggests that the two double bonds of 856 undergo oxidation at comparable rates (see Scheme

106).

In 1976, Masuoka and Kamikawa reported a second synthesis of (~~)-12,13-epoxytrichothec-9-ene (Scheme 106).301 Photoaddition of ketal 857 onto 3-methyl-2-cyclopentenone provides a mixture of adducts, from which 858 may be obtained in 16% yield. Acetolysis of adduct 858

yields 859, in which the relative stereochemistry of the two asymmetric carbons has been neatly arranged. Dione 859 is manipulated by a straightforward sequence of steps into p-keto ester 861, which undergoes Mannich condensation at C, of the cyclopentanone to provide 862 in 37% yield. After removal of the ethoxycarbonyl group, the enone is reduced to a mixture of diastereomeric allylic alcohols (864). The hydroxy group is acetylated and the dithioketal hydrolyzed to obtain 865. Upon removal of the acetate, a mixture of 866 (60% yield) and 867 (12% yield) is obtained. The mixture is separated and the hydroxy group in the major isomer inverted to obtain 868. Reaction of the latter with methylmagnesium iodide, followed by acid, yields diene 856. Oxidation of 856 gives a 1:l mixture of (*)-12,13-epoxytrichothec-9-ene (826) and 870.302 In 1980, Still and Tsai reported the outstanding piece of synthetic craftsmanship which we summarize in Scheme 107.303Dienyl ether 871 reacts with benzoquinone to give the endo adduct 872, which is regiospecifically ring-contracted to cyclopentenone 874 by alkaline epoxidation, followed by Favorskii rearrangement. A second stereoselective epoxidation, followed by lithium-ammonia reduction provides trio1 875, in which the relative stereochemistry of the eventual secondary hydroxy and the two methyl-bearing centers is correctly established. Selective acetylation of the primary hydroxy gives acetate 876, which is reductively cleaved by photolysis in aqueous HMPT. The next four steps serve to rearrange the protecting groups and set up the desired 1,3-diol system for fragmentation (878+879), which unmasks the trichothecane carbon

243

Isolated Rings

ese

859

FOOEt

COOEt

2. I . HgC12,CdC031 Ac20 + MoCN, ti20

H

0

O

c __c OH-

A

0

BAC # 86s

I . separate

2. MsCI+ 3. EtqN O A c -

+

o&--OH

0

866

&H

0

e6e

869

m-CPBA

+

\

826

em

H+

I

___L

H

----t

HO

/

om

eio

Scheme 106. Masuoka-Kamikawa Synthesis of (*)-12,13-Epoxytrichothec-9-ene

\

H es6

Other Blcarbocyclic Sesquiterpenes

244

I . LDA

t -BuOOH triton B

*

2. I - B u M o ~ S I C I OSiR3

171

a72

I

PH

OH

OSiR3

OSiR3

I-BuOOH,

Iriton B 2 LI-NH3 OSiR3

881

880 pCoPh

pC0Ph

pocls-

I PhCOCl 2 Collinr

___)

882 ?COPh I . separate

-.

2. Phf=CH2 3 m-CPBA

H

H

884

a05

OH

H

828

Scheme 107.

Still’s Synthesis of (*)-Trichodermol

Isolated Rlngs

245

skeleton. The introduction of the oxygen bridge between the two carbocycles now remains to be done. This is accomplished by stereoselective epoxidation of the alcohol derived by hydrolysis of benzoate 879. The resulting epoxide (880) undergoes acid-catalyzed ring-opening to a trio1 which undergoes intramolecular Michael addition to give 881. The specificity of the ring-closure is interesting as two other tricyclic ethers (886 and 887) might also have formed by addition of the other two hydroxy groups to the enone. However, each of these compounds is of

886

887

necessity a trans-fused bicyclo (3.3.0loctane system, which would be considerably less stable than the bridged system in the actual product. The synthesis is completed by addition of methyllithium to the carbonyl group, protection of the less hindered secondary hydroxy group, and oxidation of the other. Dehydration of the tertiary alcohol gives a 7:l mixture of the desired alkene 884 and its isomer 885. The reason for this fortunate selectivity is not clear. It may simply be that the acidity of the C4-hydrogen is slightly greater than that of the C2-hydrogen because of the inductive effect of the proximate oxygen. The synthesis is completed

in the same manner as by Colvin and Raphael in their earlier trichodermin synthesis. In all, the Still synthesis requires 21 steps from dienyl ether 871 and proceeds in an overall yield of only 0.5%. For comparison, the Colvin-Raphael synthesis (Scheme 104) requires 17-19 steps

246

Other Bicarbacyclic Sesquiterpenes

and gives (*t)-trichodermol in an overall yield of 0.006%. Thus, although the Still synthesis represents an advance, it leaves considerable room for improvement in the area. In 1976 there appeared three papers reporting syntheses of trichodiene and the related ketone, norketotrichodiene. Welch and co-workers reported two syntheses of diene 824.304a3bIn the first (Scheme 1081, the Colvin-Raphael lactone 832, which was prepared by Welch and Wong in

H

S

H

832

OH

T

H

P

888

p - TsOH

___c

889

890

I . separate

2. Ph3P=CH2

825

Scheme 108.

89 I

824

Welch’s First Synthesis of ( *1-Trichodiene

an alternate way,30Sa*b is alkylated first with methyl iodide and then with the tetrahydropyranyl ether of 4-bromo-1-butanol to obtain an adduct (888) as a 96:4 mixture of diastereomers. After removal of the protecting group, the allylic oxygen is eliminated by reduction with calcium in ammonia. Unfortunately, as it develops, this reduction gives two isomers, of which the desired is the minor component. Straightforward elaboration of hydroxy acid 889 eventually gives a 1:2 mixture of

Isolated Rings

247

(khorketotrichodiene (825) and its isomer 891. Wittig methylenation of the former compound yields (*I-trichodiene (824). In an alternative approach (Scheme 1091, the cyclohexene double bond is introduced by lithium-ammonia reduction of a diene (8951, which gives a 4:l mixture of ketones 825 and 897, this time favoring the right one.

S

O

T

H

p

I.

;;--;OH,

2.

cro3

COOMe

3. CH2N2

H

H

886

&

COOMe

‘OH

893 I . LiAIH4.

2. Li-NH3

892 I . HCI, CH2C12

B

c

O

o

M

2. collidineC

e

No N (Si Me3I2

t

-& DBU

col II dine,

A

0’

dH & do .-.. 894

I . Cr03 2. separate

896

Scheme 109.

+

825

895

0

897

Welch’s Second Synthesis of (rt 1-Trichodiene

Yamakawa and co-workers have prepared (*horketotrichodiene (825) as shown in Scheme 1 Photocycloaddition of enones 898 and 899 yields a mixture of numerous products from which adducts 900 and 901 may be isolated. Reverse aldol reaction of 900 yields 902, which is tosylated and reduced to hydroxy ketone 904. Dehydration provides 825. In spite of the poor yield in the first step, this synthesis provides ketone 825 in a very direct manner and is clearly competitive with the Welch synthesis.

248

Other Bicarbocyclic Sesquiterpenes

I . rrparatr

2. KOH

HO

898

899

902

900

903

901

904

826

Scheme 1 10.

Yamakawa's Synthesis of (*)-Norketotrichodiene

(S) Debromolaurinterol Acetate

In 1973, Mirrington and co-workers reported a synthesis of (*)-debtomolaurinterol acetate (9131, a possible precursor to a number of racemic sesquiterpenes including laurinterol (915).307 The early part of the syn-

thesis (Scheme 1111, preparation of ketone 907 from the anisole derivative 905, is almost identical to work published by Hirata and co-workers in 1969.308Dehydration of the alcohol resulting from methylation of 907 gives an inseparable mixture of alkenes 908 and 909. The ensuing Simmons-Smith methylenation of the mixture gives a complex mixture of products including both (f)-laurinterol methyl ether (910) and its diastereomer 911 in a 1:3 ratio. Again, the isomers turn out to be

Bridged Systems

908

909

912

Scheme 11 1.

91 I

910

913

249

914

Mirrington’s Synthesis of ( *)-Debrornolaurinterol Acetate

inseparable, so the bromine and the methyl are removed, and the phenol acetylated to give a mixture of 913 and 914, which can be separated by preparative gas chromatography. The enantiomerically homogeneous version of 913 has been converted into (-)-laurinterol previously.

B. BriQed Systems (1) Camphorenone, Epicamphorenone, a -Santalene, a -Santalol,

p -Santakne, epi+ -Santalene,p -Santalol, and Sesquifenchene

A number of sesquiterpenes have the bicyclo[2.2.llheptane or closely related tricyclo(2.2.1.O}heptane skeleton. Those which have been synthesized in this period are camphorenone (9161, epicamphorenone (917), a-santalene (9181, u-santalol (9191, p-santalene (9201, epi-p-santalene (9211, @-santalol (922) and sesquifenchene (923).

250

Other Bicarbocyclic Sesquiterpenes

916

917

919

91 8

92 I

920

922

923

Money and co-workers, at the University of British Columbia, in Vancouver, have synthesized (*)-camphorenone and (*)-epicamphorenone as shown in Scheme 1n 3 0 9 Racemic dihydrocarvone (924), after ketalization, is metallated and allowed to react with ethylene oxide to obtain alcohol 926. After replacement of the hydroxy group with chlorine, the ketal is removed and the ketone converted into an enol acetate. Cyclization of 929 is brought about by boron trifluoride in wet dichloromethane to yield a mixture of chloro ketones 930 and 931 in 55-600/0 yield (ratio not specified). After protection of the carbonyl, the remaining three carbons of the side chain are introduced by reaction of the derived Wittig reagent with acetone. The final product is a mixture of (*)-camphorenone (916) and ( *bepicamphorenone (917), which is separated by preparative glpc. Compound 917 was subsequently converted into (f)-p-santalene (9201, (f)-a-santalene (9181, (*bcopacamphor (934) and (f1-ylangocamphor (935). Compound 918 was converted into ( f )-epi-P-santalene (921) .3'0 Since these syntheses involve

934

93 5

251

Bridged Systems

924

925

926

928

927

929

93 I

930

932

933

I . Ph3P 2. DMSO, NOH

=.

JL

4. H30+

Scheme 112.

916

917

Money's Synthesis of (-+)-Camphorenone

essentially the conversion of one sesquiterpene to another, the reader is referred to the original source for details. In 1973, Money and co-workers reported a synthesis beginning with (+)-Pbromocamphor (936, Scheme 113) which affords (+)-epicamphorenone (917), and (+ 1-epi-@-santalene(921).3l 1 A parallel sequence beginning with (- )-8-iodocamphor (939) provides (-)-camphorenone (916) and (-)-@santalene (920). As will be seen from the scheme, the key transformation is coupling of the neopentyl iodide in each case with the r-allylnickel reagent derived from the reaction of nickel carbonyl with prenyl bromide. Note that the Wagner-Meerwein rearrangement of the tosylate derived from camphorenone leads to the epi-@-santalene

252

Other Bicubocyclic Sesquiterpenes

(r-

I . (CHzOH12, H+

I.

2 . No1 , DMSO

2. H30+

936

NiBrlz

937

L i ( 1 - Bu013Al H

H

O

938

S

917

p-TsCI

92 I

Scheme 11 3. Money’s Synthesis of (-)-Camphorenone, ( )-Epicamphorenone, (-)-P-Santalene. and (+ )-Epi-P-santalene

+

stereochemistry while the analogous tosylate derived from the epicamphorenone gives rise to normal p-santalene stereochemistry. In 1970, Corey and co-workers had reported a synthesis of (+)-a-santalol (919, Scheme 114) beginning with (-)-w-bromotricyclene (940).312 The key to the synthesis is reduction of propargylic alcohol 943 by diisobutylaluminum hydride, followed by iodination of the intermediate vinyl alane. Since the alkyne reduction occurs by an anti mechanism under these conditions, the resulting vinyl iodide has the 2 stereochemistry (944). This is unfortunate, for if it had the Estereochemistry only one . ~ ~it ~is, the more operation would be necessary to obtain a - s a n t a l ~ l As CH20H group must be converted into CH, and the I into CH,OH. These manipulations require an additional five steps. The final product is (+)-a-santalol (919).

Bridged Systems

940 @A OH

I . n-BuLi 2. CH*0

943

941

942

I . n-BuLic 2. D I B A L 3. I,

I. MsCl 2 . LiBr

-

253

944

91 9

Scheme 114.

Corey's Synthesis of (+)-a-Santalol

Julia and Ward have also employed (- 1-r-bromotricyclene in In syntheses of (+)-a-santalene and (+)-a-santalol (Scheme 1 this case the side chain is attached by displacement of the neopentyl bromide with sodium benzenesulfinate in DMF to obtain a sulfone (941) which is alkylated either with prenyl bromide to prepare a-santalene or with the dichloride 944 to prepare a-santalol. In each case the benzenesulfenate group is removed by reduction of an aryl-alkyl sulfone (942 or 946) with sodium amalgam. Takagi and co-workers have reported the conversion of (-)-r-bromotricyclene into (+)-a-santalol by a direct coupling using the Ir-allylnickel

Other Bicnrboryclic Sesquiterpenes

254

gr

PhSO;Na+

DMF

940

941

942

91 6

2s

943

S02Ph

SOZPh

941

&a

cc;

I . n-BuLi

2.

Na(Hg) EtOH-THF)

I . n-BuLi

*

~

945

-

S08h

NaOAc

946

944

EtOH- + Na(Hg) HMPT

Scheme 115.

S

O

H

919

Julia's Synthesis of (+)-a-Santalene and (+)-a-Santalol

derivative 947 as the source of the side chain (Scheme 116).3'5 The coupling reaction is carried out in HMPT and affords ethers 948 and 949 in 45% yield. Unfortunately, the desired stereoisomer is the minor component of a 3:2 mixture. Debenzylation of the mixture gives a mixture of (+)-a-santalol (919) and its E stereoisomer (950) which may be separated by silica-gel chromatography. In 1978, Monti and Larsen at the University of Texas published the very interesting synthesis of ( *)-a-santalene which is shown in Scheme 1 17?16 Unlike the other syntheses discussed so far, this synthesis does not begin with the tricyclene nucleus, but rather with the unsaturated

255

Bridged Systems

b v Ph

948

947

940

949

919

Scheme 116.

950

Takagi’s Synthesis of (+)-a-Santalol

acid 951. Intramolecular acylation provides the bicyclo[3.2.1 loctenone 952317 which undergoes smooth photoisornerization to tricyclic ketone 953 (71% yield). Photochemical Wolff rearrangement of diazo ketone 954 provides the ring-contracted acid 955, which is alkylated as its dianto obtain acid 956. Conversion of ion with S-iodo-2-methyl-2-pentene the carboxy group to methyl completes the synthesis of (f)-a-santalene.

951

952

hv Nonco3, H20. THF

I . LDA,THF2. n-BuLi

955

3. &I

953

-

954 I . LiAIH4 2 . p-TsCI

“

O

956 ° F

3. LiEf3BH. THF

918

Scheme 117.

Monti-Larsen Synthesis of (*)-a-Santalene

-

256

Other Bicarbocycllc Sesquiterpeoes

In 1979, Bertrand and co-workers reported the Diels-Alder reaction of cyclopentadiene and the allenic ester 957 (Scheme 1 18).3’8 Two adducts (958 and 959) are produced. Reduction of the more reactive norbornene

0+

/H

CHZ=C=C,

COOMr

CgHg

957

&ooMe 958

+

&foMe

H,-NiB EtOH

959

I . LIAIH,

3 . Wolff960

THF, HMPT

Scheme 11 8.

Kirhnrr

96 1

920

Bertrand’s Synthesis of (rt )-&Santalene

double bond using Brown’s nickel boride catalyst provides a mixture of esters (960) which is alkylated with 5-iodo-2-methyl-2-pentene.As in Corey’s earlier p-santalene work319 the alkylation is highly stereoselective and the exo isomer 961 is the sole alkylation product. Conversion of the methoxycarbonyl group to methyl provides (It 1-p-santalene (920).

Christenson and Willis, of the Fritzsche, Dodge and Olcott Company, have reported a synthesis of 8-santalene and 8-santalol beginning with racemic camphene (Scheme 1 1 9).320 Epoxidation with peroxyacetic acid yields two epoxides (963 and 964) in 87% yield. Treatment of the mixture with the dilithio derivative of acetic acid, followed by acidification yields a mixture of exo and endo spiro lactones 965 and 966. When this mixture of lactones is stirred in concentrated sulfuric acid, rearrangement occurs to give a mixture of two lactones in a 5:l ratio. The major lactone (967) is isolated in 45% yield by silica gel chromatography. After reduction to the aldehyde oxidation state, the isopropylidine unit is added by Wittig reaction to obtain alcohol 969. Dehydration of this

Bridged Systems

@

CH3COSH

905

I . LiCH2C02Li

+

903

962

2. H+

904

966

907

97 0

967

257

97 I

I . PhSP=CHCHS 2. n-BuLi 3. CH2O

922

Scheme 119. Christenson-Willis Synthesis of (*)-b-Santalene and (f)-@-Santalol

material yields (*)-P-santalene. For preparation of 8-santalol, lactone 967 is treated with p-toluenesulfonic acid and ethanol in refluxing benzene, whereupon unsaturated ester 970 is produced. The side-chain construction is completed as in the earlier Corey-Yamamoto synthesis of (+)-a-santa~ol.~~~

258

Other Bicrrbocyclic Sesquiterpenes

In 1979, Baumann and Hoffmann, of BASF in Ludwigshafen, reported a synthesis of (*)-fl-santalol (Scheme 120).322 The synthesis begins with a Diels-Alder reaction between cyclopentadiene and (Z)-4-chloro2-methyl-2-pentena1, an intermediate in the BASF vitamin A synthesis. Surprisingly, the reaction gives exclusively the exo addition product 972,

DMSO

eH0 && 97 2

-=! H20

974

97 3 I . HCOOMe, NaOMe ~

2 . MeOH, H+

97 5

976

977

978

Scheme 120.

922

Baumann-Hoffmann Synthesis of

979 (zt

)-P-Santalol

which is obtained in greater than 80% yield. After hydrogenation of the double bond, the aldehyde is protected and the chloride eliminated. After regeneration of the aldehyde function, intermediate 974 is obtained. Mixed aldol condensation of aldehyde 974 with 2-butanone gives enone 975 in 84% yield. The remaining carbon is added by formylation; after acetalization and hydrogenation of the double bond, com-

Bridged Systems

259

pound 977 is obtained. Reduction of the carbonyl group gives a hydroxy acetal which is treated with acid to obtain an a,P-unsaturated aldehyde. This substance is depicted by Baumann and Hoffmann as the Estereoisomer 978, which is a reasonable assignment based on analogy to simpler examples.323 Inexplicably, the product of reduction of this aldehyde is depicted as (*)-p-santalol (922). However, the spectra reported for the product seems to be more consistent with the spectral properties of 922 than with the E stereoisomer 979. In particular, the 'H-NMR CH20H resonance of 922 occurs at 4.06 ppm while that of 979 occurs at 3.85 Baumann and Hoffmann report that their alcohol has this resonance at 4.03 ppm. Until this confusion is clarified, the BaumannHoffmann synthesis should be considered with caution. In 1963, Bhattacharyya isolated a new sesquiterpene from Vuleriunu Wuulichi and assigned it a bergamotene structure. After all four bergamotene racemates had been synthesized, it was clear that Bhattacharyya's hydrocarbon has a different structure and Corey and coworkers proposed the sesquifenchene structure 923.324 In 1972, Bessiere-Chretien and Grison confirmed this assignment by the synthesis outlined in Scheme 121.325Ether 98d26 undergoes skeletal rearrangement upon treatment with boron trifluoride etherate in acetic anhydride to provide acetoxy alcohol 981. The derived endo-norbornyl tosylate undergoes Wagner-Meerwein rearrangement upon solvolysis. After

& BFA,fc,o'

I . HOAc,A

& O A c

980

983

Scheme 121.

p-T'CI:

W

A

OH

OTs

98 I

982

C

2' 3. Lp-TsCI iA1H4y 4. No1

923

Bessiere-Chretien-Grison Synthesis of (+Mesquifenchene

260

Other Bicnrbocyclic Sesquiterpeoes

reduction of the ester, the primary alcohol is iodinated to obtain 983. The prenyl side chain is attached using the m-ally1 nickel procedure to obtain ( + hesquifenchene. Three years later, Grieco and his co-workers reported two additional ~~~ first (Scheme 1221, norsyntheses of (*) - s e s q ~ i f e n c h e n e . ~In~ ’the bornadiene is converted in two steps following a procedure developed by Corey in connection with prostaglandin synthesis into the tricyclic keto acid 984. Three further steps suffice to convert this compound into the bromo ester 985, which is converted into the enolate ion and then methylated to obtain 986 and 987. It turns out that the bromine is more effective than the ketal oxygen at hindering the approach of the electrophile to the planar enolate and the 986 to 987 ratio is 15:l. The major isomer is debrominated and the methoxycarbonyl group transformed into CH,I as shown. The remaining carbons of the side chain are introduced by Julia’s method (see Scheme 115) to obtain (*I-sesquifenchene.

I , HBr-HOAc 2 . CHZNO

2 atepa

&

p-TtOH

984

+Me&

COOMe

985

COOMe Br

I . reparate 2 . n-Bu3SnH 3. N o 1

Scheme 122.

Grieco’s First Synthesis of (It)-Sesquifenchene

Bridged Systems

261

In his second synthesis (Scheme 123) Grieco begins with endodicyclopentadiene, which is selectively hydrogenated using the Brown nickel boride catalyst to obtain 991. Hydroboration of this alkene, followed by direct chromic acid oxidation of the resulting alkylborane provides a 3:2 mixture of isomeric ketones 992 and 993. These are easily separated, since 993 forms a bisulfite addition product while 992 does not. Baeyer-Villiger oxidation gives a lactone (9941, which is reduced to an aldehyde and condensed with isopropylidenetriphenylphosphorane to obtain unsaturated alcohol 995. After oxidation, the resulting ketone is methylated to obtain ketone 996. (This ketone had previously been prepared in two steps from n ~ r b o r n a n o n e . )Reduction ~~~ occurs to give

991

992

994

995

I . p-TrCI 2. HOAc

LiAIHq,

3. NoOH

99 3

.

4. Jones

996

997

999

Scheme 123.

923

Grieco's Second Synthesis of (kbSesquifenchene

+

262

Other Bicarbocycllc Sesquiterpenes

1001

I000

-

C P - f

c

n

o

m

1004

2

+g

h: %'% 100s

1002

I006

+

H2S

1007

1001

&?"

I.p-T'C'b

2 No1.a

OH

I008

+C%

2I AgN03 srporotc

1012

'' NOH

HCooE'*

OH

+cf-&b T : . ;

p&o

1014 I . NaBH4

%cHo

2. H2O.HCI-

2 . C5H5N,S03 3. LiAIH4

1016

1016

1017

Scheme 124.

OH

1011

1010

1013 I . NOW4

2 . N O H , ) \ I*

OH

loo*

Corey's Synthesis of (*)-a-rrunsBergamotene

+

Bridged Systems

263

the endo alcohol 997, which is rearranged as its tosylate to obtain a mixture of acetates. Conversion of these compounds into ketones gives a 7:3 mixture of ketones 998 and 999. Methylenation of the major isomer gives (fbsesquifenchene (923). It is not clear that either Grieco synthesis represents an improvement over the Bessiere-Grison synthesis.

(2)

a -trans-Bergamotene

In 1975, Corey, Cane, and Libit reported a synthesis of (f)-a-transbergamotene (1017).324The synthesis begins with geranyl acetate, which is ozonized to acetoxy aldehyde 1000 (Scheme 124). Straightforward manipulations provide the allylic bromide 1002 which is coupled with a vinyl cuprate reagent to obtain allylamine 1003. Oxidation of this compound by the modified Polonovski method gives an a,P-unsaturated aldehyde (10041, which is converted by Wittig reaction into triene 1005. Photochemical isomerization of this compound gives bicyclo[2.2.1]hexane derivatives 1006 and 1007 in a ratio of 5:3. The mixture is hydroxylated, converted into monotosylates, and iodinated to obtain 1010 and 1011, which are separated by silica gel chromatography. Rearrangement of the major isomer 1010 gives isomeric ketones 1012 and 1013 in yields of 55% and 22Y0, respectively. Again, the isomers are separated by chromatography and the major isomer converted into unsaturated ketone 1014. A series of six straightforward steps completes the synthesis of (*))-a-trans-bergamotene (1017). Although the synthesis outlined in Scheme 124 was the first reported synthesis of a transbergamotene, it left much room for improvement, since it requires over 20 steps and requires isomer separations at two stages. An improvement was recorded by Larsen and Monti in their 1977 report of the synthesis outlined in Scheme 1 X 3 * ' Bicyclic enone 952 is hydroxylated and the resulting secondary-tertiary diol is sequentially mesylated and then converted into the trimethylsilyl ether 1019. The enolate ion undergoes intermolecular alkylation faster than intramolecular alkylation at -4O"C, to give ketone 1020, which is cyclized to the

Other Bicnrbocyclic Sesquiterpenes

264

6H

~TMS

ioie

952

I020

I022

1019

1021

H20

Scheme 125.

1023

1017

Monti-Larsen Synthesis of (It)-a-irunsBergamotene

tricyclic ketone 1021. Haller-Bauer fragmentation provides 1022, which has the correct a-rrans-bergamotene skeleton. Unfortunately, the amide is very unreactive and a further six steps are required to convert it into a methyl group. The Monti synthesis is an outstanding contribution, even with this minor flaw.

C. Spirocyclic Systems (1) Spirovetivanes

In the time since 8-vetivone (1024) and its congeners were recognized

to have the spiro (4.51decane rather than the hydroazulene skeleton,328 the members of this class have become a favorite target for synthesis. In

Spirocyclic Systems

265

Volume 2 of this series, we documented three syntheses of vetivane sesquiterpenes. In this volume, we report nine more syntheses of 1024 (a compound which is rapidly becoming the “most synthesized” sesquiterpene), in addition to syntheses of agarospirol (1029, hinesol (1026), P-vetispirene (10271, a-vetispirene (1028), solavetivone (1029), and anhydro-a-rotundol (1030). In addition, we call attention to the full paper reporting the Marshall-Brady synthesis of (f1-hinesol (1026)329 which was discussed earlier.330 O

%;

y

%

i 6 H

1025

1024

(&$ 1027

1026

1028

1030

1029

In 1973, Stork, Danheiser, and Ganem published an astoundingly simple synthesis of (*)-p-vetivone (Scheme 126).33’ The ethyl enol ether

I . LDA,THF, I

I

lo31

I. MeLi 2. H30+

2, I032

3. LDA

1033

1024

I . LMoLi, HMPT 2. MsC1,LiCI

LiAlHq

CI

OH

1035

1034

Scheme 126.

1032

Stork’s Synthesis of (*)-@-Vetivone

266

Other Bicarbocyclic Sesquiterpenes

of dihydroorcinol (1031) is deprotonated by LDA in THF containing one equivalent of HMPT. Dichloride 1032 is added, followed by more LDA. Spiroannelated enol ether 1033 is obtained as one stereoisomer in 45% yield. Addition of methyllithium and acid hydrolysis gives (*)-@vetinone (1024) in an overall yield of 27% for the three steps. The requisite dihalide is prepared from diethyl isopropylidinesuccinate as shown in the scheme. Also published in 1973 was a synthesis by McCurry and Singh leading to (*bp-vetivone and (f)-a-vetispirene (Scheme 127).332 Alkylation of the pyrrolidine enamine of aldehyde 1036 with crotyl bromide affords 1037 as a 1:1 mixture of diastereomers. Since the alkylation occurs first on nitrogen, and the resulting N-crotylenammonium ion then undergoes [3.3]sigmatropic rearrangement, only the terminal alkenes are produced. The diastereomeric mixture is converted by a well-precedented sequence to unsaturated aldehyde 1040 which undergoes Prins cyclization to a mixture of 1041 and 1042, which is separated by chromatography. Moffatt oxidation of 1041 is accompanied by conjugation of the double bond, leading to (*)-@-vetivone (1024). Alternatively, the mesylate of 1041 may be eliminated to obtain (*)-a-vetispirene (1028). In an earlier publication, McCurry and co-workers had synthesized alcohol 1043 and studied its formolysis. Predictably, alkylation of the allylic cation occurs on the face of the cyclohexane double bond trans to the methyl group, leading to ( *)-lO-epi-@-vetivone(1044) in 40% yield.

n

1043

1044

Spirocyclic Systems

2I . NCS, MeLi DMS,

H

Et3N 3. MeLi

1037

I036

Ho% 1039

I. Acp 2. SOCle* CSHSN 3. LiAIH4 4. NCS, OMS, Et3N

I . R2BH 2. H202, OH-

1038

SnCl

4_ CHZC12

OH%1040

1041

1024

Scheme 127.

%

267

1042

ioza

McCurry's Synthesis of (*)-p-Vetivone and (i)-a-Vetispirene

Yamada and co-workers have developed the rather long, but highly stereoselective synthesis of ( f 1-p-vetivone which we summarize in Scheme 128.333The starting point aldehyde is 1045 (obtained in six steps from m-cresol methyl ether). After homologation, the resulting

268

Other Bicarbocyck Sesquiterpenes

C-0

MeoF3

I . Li-NH3e

I . PS*OCH3

I - BuOH

2. ( C H 2 0 H ) ~ . H + b

OOH 1046

COOH 1045

1047

1049

1048 I . NaBH, 2 . MsCI,

I . (CH20H)2. H+ 2. KOH, MoOH 3 . Collins 4 , (MeO)zCO, NOH

2. H30+

c

3. NoOMe, 1051

I050

I052

1053

””$““r o% Br ’ 1054

Scheme 128.

r I024

Yamada’s Synthesis of (*)-p-Vetivone

aldehyde is protected, and the acetal acid 1046 is reduced by Birch’s procedure to obtain compound 1047. Hydrolysis of the acetal gives an aldehyde which undergoes aldol condensation at the y-carbon of the a,B-unsaturated ketone to generate the [5.4]spirodecane system. Studies on simple model compounds had shown that this condensation is reversible, and that the initial vinylogous aldol is prone to rearrange to other

Spirocyclic Systems

269

products. However, in this case the carboxy group traps the initial product, leading to lactone 1048. Protection of the ketone carbonyl, reduction of the lactone, and regeneration of the enone system is followed by spontaneous cyclization to give a tricyclic alcohol, which is acetylated to obtain keto ester 1049. Precedented transformation of this material into a,p-unsaturated ester 1051 is followed by catalytic hydrogenation, The hydrogenation is highly stereoselective, the sole product being stereoisomer 1052. After conversion of the methoxycarbonyl group to isopropylidene, the ether bridge is eliminated by reaction with triphenylphosphine dibromide, giving 1054, which is readily transformed into (f)-p-vetivone (1024). In a companion paper334Yamada and co-workers describe conversion of intermediates from the p-vetivone synthesis into (*1-hinesol (l026), (f)-a-vetispirene (10271,and (f )-P-vetispirene (1028). These conversions, which are all rather routine, are summarized in Scheme 129. In 1977, Yamada reported utilization of compound 1049 for synthesis of (*I-solavetivone (10291, a stress metabolite from infected potato tubers.335 A key step in this synthesis is conjugate addition of isopropenylmagnesium bromide to enone 1060. An approximate equimolar mixture of diastereomers is obtained. The isomers were readily separated by chromatography and the complete stereostructure of one isomer was elucidated by X-ray analysis. This isomer was used to complete the synthesis outlined in Scheme 130, thus establishing the stereostructure of the natural product. The synthesis was complicated by the necessity of protecting the isopropenyl double bond, which otherwise was found to undergo extensive isomerization. In 1974, Caine and Chu reported another approach to the p-vetivone problem (Scheme 131).336 (fbNootkatone (386) is converted into the cross-conjugated dienone 1065, and the isopropenyl double bond is isomerized by conversion into the tertiary bromide, which is eliminated under E,C conditions to obtain 1066. Successive photoisomerizations, first in dioxane, then in aqueous acetic acid, lead to trienone 1068,which is selectively hydrogenated to a 7:3 mixture of (*)-p-vetivone (1024) and its diastereomer 1044.

270

Other Bic8rbocyclk Sesqulterpenes

M--cooMo I. H~O+ 2. lPhO@MoIBFS * Etz0

c

I . Zn-HOAc

_____c

2 . (CHzSH)z,

‘COOMe

BF3

I055

1052

I . Raney Ni 2. MaLi

H ‘r
1051

c%

I . Roney Ni 2. 3. MoLi H’, ~ O ~ Z W I :

COOMe

1058

I . Zn, HOAc

2 . NaBH, 3. H + , b m t e n r )

I’

0s

1057

H

‘COOMe

1055

Scheme 129.

3

‘COOMe

1059

1027

I . MeLi

1028

Yamada’s Synthesis of ( *bHineso1, (*h-vetispirene, and ( *1-p-Vetispirene

Spirocyclic Systems

271

1062

1061

1063

1064

H

1029

Scheme 130.

Yamada's Synthesis of (f1-Solavetivone

In the same year, Pesaro and co-workers reported the synthesis summarized in Scheme 132.337 The synthesis begins with (f)-a-terpineol (l069), which is ozonized to obtain hemiacetal 1070. This compound cyclizes upon treatment with acid to provide the acetylcyclopentene 1071. Catalytic hydrogenation of the less substituted double bond gives 1072, which is elaborated into unsaturated nitrile 1074. The derived acyl halide undergoes intramolecular acylation to give ketone 1075 (see also Scheme 127, 1040-1041). The cross-conjugated dienone 1076 undergoes coqjugate addition when treated with several methylcopper reagents to give (kbp-vetivone and its diastereomer. Although isomer 1024 is invariably the major isomer, the highest selectivity (1024 to 1044 = 5:l) is obtained using IM~CUB~I~IL~(~-BU~NH)~I+.

272

Other Blcubocycllc Sesqulterpenes

1044

I024

Scheme 13 I .

Caine's Synthesis of (*)-p-Vetivone from ( *bNootkatone

I069

1070

1073

I076

Scheme 132.

1071

I372

I075

1074

1024.

Pesaro's Synthesis of (*)-P-Vetivone

1044

Spirocyclic Systems

273

Trost and co-workers have studied the applicability of their cyclobutanone spiroannelation procedure for the synthesis of spirovetivanes as outlined in Scheme 133.338 Treatment of 2,6-dimethylcyclohexenone (1077) with 1-lithiocyclopropyl phenyl sulfide gives an oxaspiropentane

n

NMe2

I079

-

I076

1077

I080 I. MeI.MeCN, H20

NaOMe, MeOH retlux 4 days

STr STs, KOAc EtOH, A

1061

1082

~

2. NaOH

I063

-3

I . NH4CN. DMF 2. HCI, MeOH

H ‘COOMe

1085

1064

Scheme 133.

+

?

COOMe ‘H

I066

Trost’s Synthesis of a Spirovetivane Intermediate

(1078) which rearranges upon treatment with lithium fluoborate to cyclobutanone 1079. The process is highly stereoselective, the reagent attack-

ing solely from the side of the cyclohexenone ring opposite the secondary methyl and the ring-expansion occurring with effective inversion of configuration at the spiro carbon. The a-trimethylenedithio derivative 1081 is prepared via vinylogous amide 1080. Treatment of 1081 with methyliithium gives a tertiary alcohol, which is treated with sodium methoxide in refluxing methanol for four days to effect ring cleavage. Ketone 1083 is obtained in 91% yield. The dithiane is hydrolyzed and the resulting keto aldehyde is cyclized to obtain dienone 1084. Conjugate addition of cyanide gives two stereoisomers, which are hydrolyzed

274

Other Bicarbocyclic Sesquiterpeoes

WB J% PAC :: il :!$:: e J

k 1087

Ac20

I . LiAIH4 2 . P-TsCI 3. NoCN

2 . Zn, BrCH2COOEi 3. HgO, 1 2 , hW “ 03

COOMI

1089

1088

@ J

*:iM:

OOMa

COOMa

COOMe

1091

I090

=&J

‘J: kg!, H+

COOEt

I . LiAlH4

2 . p-TsCI

3 . PhS02CH,Li+ COOMI

p 4

1092

I . H C I , CHCIS 2. ~ - T ~ c I , c ~ H ~ N ~ 3. Collins

S02Ph

1094

1093

1095

1096 I . m-CPBA,

@&

2. Burgoso

H

H

1098

1097

I099

1026

Scheme 134.

LiAIH4 c

1100

Magnus Synthesis of (+)-Hinesol

1101

Spirocycllc Systems

275

to obtain a difficult to separate 2:l mixture of keto esters 1085 and 1086. Although these intermediates have not been taken further, removal of the ketone carbonyl and addition of methyllithium should provide (*bhinesol (from 1085) and (*bagarospirol (from 1086). Magnus and his co-workers have reported a synthesis (Scheme 134) which leads to a mixture of k b h i n e s o l (1026), (+)-lO-epi-hinesol (10991, and two exocyclic double bond isomers (1100 and 1101).339The synthesis proceeds through the bridged ether 1088, which is homologated to ester 1089. Wagner-Meerwein rearrangement provides the norbornane derivative 1090, which is elaborated to a,p-unsaturated ester 1092. Conjugate addition of lithium dimethylcuprate gives a 1:l mixture of diastereomers 1093. Another homologation ensues, leading to sulfone 1094. After deprotection of the 1,3-diol system, the primary alcohol is tosylated and the secondary one is oxidized. The stage is now set for the key maneuver of the synthesis. The sulfone anion adds to the carbonyl group, generating a tertiary alkoxide which is well disposed for Grob fragmentation. The result is a spiro(4.51decane (1096) in which the absolute configurations of the spiro carbon and the isopropenyl carbons are established correctly. Unfortunately, the secondary methyl carbon is not correctly fixed, and the ultimate product is a mixture of (+)-hinesol (1026) and (+)-10-epi-hinesol (1099), along with smaller amounts of the exocyclic double bond isomers 1100 and 1101. Silica gel chromatography suffices to separate 1026 1099 (26% yield) from 1100 1101 (l8Y0 yield) and treatment of the latter fraction with acid causes quantitative rearrangement to 1026 1099, thus increasing the yield of the endocyclic isomers to about 45%. However, 1026 and 1099 can be separated only by conversion to the mixture of acetates, which are separable by glpc. Hydrolysis of the separated acetate affords pure (+)-hinesol (1026). It should be noted that (+I-hinesol is the enantiomer of the natural product. Dauben and Hart developed a synthesis (Scheme 135) which efficiently leads to most of the known spiro~etivanes.~~' Enol ether 1031 (see also Scheme 126) is formylated to obtain 1102. This substance gives an anion which reacts with phosphonium salt 1103 to give the spiroannelated product 1104. It is noteworthy that the alkylation is highly

+

+

+

Other Bicarbocyelic Sesquiterpenes

276

NoH,HMPT

HCOOEI I

NoH

*

CHO

HZ-Pd/C

PPh2BFi

COOEI

1103

1102

1031

E’oq

D(cOOEt

1104

I . MeLi

H

‘COOEt

y 2. BF3

‘ri‘

I

1026

1030

1024

9 ‘YH 9 I . LIAIH, 2. H+,C6H6

1028

- - - - - - - - - -- - - - - - ‘lo-

benzene* H+

OOEt

1104

1107

I . MeLi

L2.L

23:;$Nn5O3,

I027

Scheme 135.

Dauben-Hart Synthesis of Vetivanes

I

\-

1108

Spirocyclic Systems

277

stereoselective, occurring trans to the secondary methyl group. Thus, intermediate 1104 is obtained in stereohomogeneous form. Catalytic hydrogenation of 1104 is also stereoselective, although in this case the stereochemical outcome is perhaps more fortuitous than predictable. As in the Stork synthesis (Scheme 1261, intermediate 1105 reacts with methyllithium to form the cyclohexenone system. In the present case, of course, the ethoxycarbonyl group is simultaneously changed to the isopropanol side chain. Straightforward transformations then lead to (*1-anhydro-p-rotundol (1030), (f)-p-vetivone (1024), (*)-hinesol (10261, and (f)-&vetispirene (1028). By a slight modification, k1-a-vetispirene (1027) is also readily available. Ramage has utilized R-(+)-3-methylcyclohexanone as the starting point for a synthesis of (-bagarospirol (1029, the natural enantiomer (Scheme 136).341 The ketone is first blocked on the less hindered side and then exhaustively cyanoethylated. After hydrolysis and esterification, keto diester 1110 is obtained. Dieckmann cyclization and removal of the methoxycarbonyl group affords the spiro[5.5]undecanedione 1111. One of the two carbonyl groups is severely hindered and selective Wittig reaction leading to 1112 is possible. Acidcatalyzed ketalization is accompanied by double bond isomerization. Unfortunately, isomers 1113 and 1114 are obtained in a 3:2 ratio. The isomers are separated by chromatography on silver nitrate impregnated alumina. Ozonolysis of the major isomer gives a keto aldehyde which cyclizes to 1115. Catalytic hydrogenation gives two diastereomers (11161, which are oxidized to acids, esterified, and deketalized. Although 1117 and 1118 (formed in a ratio of 7:3) may be separated by chromatography on alumina, this is not necessary since both react with methylenetriphenylphosphorane to give 1119. Apparently the two isomers are in equilibrium under the conditions of the Wittig reaction and isomer 1117 reacts faster, due to the hindered environment of the ketone carbonyl in 1118. Isomerization of the double bond gives 1120, which is converted into (-bagarospirol in the standard manner. In 1976, Caine and co-workers published a synthesis of ( f )-a-vetispirene by a route involving a photochemical cyclohexadienone rearrangement as a key step (Scheme 137).342 Photolysis of

278

Other Bicrrrbocyclic Sesquiterpenes

$.. - qcooM

0 PhN,

I . @CN Triton 0

I . HCOOEt NaH 2. PhNHMe

I

I . Na,C6H6 2. L i I , D M c

3. MeOH. H+

3 I

A

bOOMe

1110

1109

(CH20H12*H+m

+

benzene

I

Ill1

Ill2

1114

1113

p-TaOH

-COOMe

OOMe 1117

Ill8

'H 1119

bH 1120

Scheme 136.

1025

Rarnage's Synthesis of (-)-Agarospirol

Spirocyclic Systems

2. I . Pb(OAc1, KOH, 02

0

Hf& I

qo 1121

-t::..:

-

I Me1 BuOK c

1123

*0AC

1125

OMc

1126

I . LDA

0

I

1130

I I33

1132

Scheme 137.

%

2 . ; 3

2. PhSrCI,

1129

hv HOAc

I

I . Li,

1124

0

ao"'

1122

MI

279

3. "202

m0

hv dioxanr ~

I

1131

1127

Caine's Synthesis of (f)-a-Vetispirene

methoxycyclohexadienone 1123 in acetic acid provides the spiro[4.5ldecane intermediate 1124, which is elaborated by a tedious but straightforward sequence of reactions into the valuable spiro (4.51decenone 1127. Addition of isopropenylmagnesium bromide and dehydration gives (*)-a-vetispirene (1027) and its isomer 1128 in a ratio of 3:2. The two isomers may be separated by preparative glpc. In a later

284

Other Bicrrboeycllc Sesquiterpnes

1155 with ethyl crotonate gives cyclohexenone ester 1156. Removal of the ketone carbonyl is accomplished by desulfurization of the dithioketal. Alkylation of the enolate ion of 1157 with ethyl 3-bromopropionate is stereoselective, but proceeds in only 19% yield. The resulting diester 1158 is subjected to acyloin condensation and the a-hydroxy ketone SO obtained is reduced to obtain the Caine-Biichi enone 1127. This material is elaborated by the indicated sequence to obtain a 3:2 mixture of (Itbhinesol (1026) and (f)-agarospirol (1025). A minor modification allows production of ( f )-a-vetispirene (1027) from acyloin 1159.

Because of the large number of spirovetivane syntheses that have been published, it is appropriate to compare them against one another. This we do in Table 2. It includes the three syntheses (Marshall’s p-vetivone, Marshall’s hinesol, and Deslongchamps’ agarospirol) which were reviewed in Volume 2 of this work. From the standpoint of efficiency, the Stork synthesis is by far the best. However, it should be noted that this synthesis is easily applicable only to p-vetivone. The Dauben synthesis is almost as efficient in the number of steps, although not in overall yield. However, the Dauben synthesis is highly adaptable and easily leads to almost all members of the class. Many of the syntheses are clearly not useful for producing the end compound in any quantity. Such is the case with the Marshall hinesol synthesis, which was done principally to determine the stereostructure of hinesol. Other impractical syntheses serve only to illustrate some particular chemistry, such as Buchi’s interesting utilization of fulvene 1135 or Trost’s application of his cyclobutanone spiroannelation. At this point, it would seem that it will be hard to improve upon the Stork and Dauben syntheses. However organic chemists seem to have a particular fascination for spirocyclic systems, and we can probably expect the parade of spirovetivane syntheses to continue.

(2) Acoranes The acoranes also contain the spiro(4.51decane nucleus, but have a different substitution pattern from the spirovetivanes. In this group also a large number of syntheses have been recorded. In this section, we

Spirocyclic Systems

281

Addition of lithium dimethylcuprate to 1135 regenerates a cyclopentadienide ion, which adds to the ketone carbonyl. The reaction is performed in ether solution at -20°C and gives a single carbinol, 1136. Reduction of acetate 1137 by diimide gives the crystalline dihydro compound 1138 in 57% yield along with 5-10°! of the tetrahydro derivative. Epoxidation of the unsaturated alcohol is highly stereoselective (presumably syn to the hydroxy group). Rearrangement of the epoxide with n-butyllithium provides the spiro f4.51decenone 1127. For preparation of (*)-p-vetivone, 1127 is transformed by p-toluenesulfonylmethyl isocyanide to nitrile 1140. Conversion of this substance to the hinesolagarospirol mixture (1141) to (*)-p-vetivone follows a precedented line. Torii and co-workers have generated the spiro [4.5ldecane system by base-catalyzed cyclization of a 4- (4-hydroxypheny1)butyl tosylate, a reaction discovered by Winstein and Baird in 1957 (Scheme 139).345a*b The required substrate is built up efficiently from phenolic ester 1142 (obtained in four steps from m-cresol methyl ether and succinic anhydride). Treatment of 1144 with potassium t-butoxide provides 1145 as a

OH

1142

1143

1024

Scheme 139.

1044

Torii’s Synthesis of (*)-@-Vetivone

282

Other Bicrrbocycllc Sesquiterpenes

mixture of diastereomers. Addition of lithium dimethylcuprate occurs at the less substituted double bond of the cyclohexadienone system, giving 1146 as a mixture of several stereoisomers. Isomerization of the double bond occurs with rhodium chloride. The product is a 1:l mixture of (*)-@-vetivone (1024) and its diastereomer 1044, which may be separated by hplc. In connection with his investigation of the reaction of enol ethers with diazomethyl ketones, Wenkert carried out the synthesis outlined in Scheme 140.292Compound 1147 (available by the Diels-Alder reaction of piperylene and crotonaldehyde) is converted into the enol ether 1148, which is conjugated by treatment with strong base. Diazoacetone is decomposed in the presence of 1149 to obtain a 1:3 mixture of the desired cyclopropane 1150 and its isomer 1151. Electrophilic cleavage of the cyclopropane rings gives a mixture of 1152 and 1153. The mixture is separated and 1152 is aldolized to obtain 1154, an intermediate in Marshall’s synthesis of (*) - @ - ~ e t i v o n e . ~ ~ ~

1147

1150

1148

1151

1149

1152

1153

I . separate

____)

2. KOH-MeOH

1154

Scheme 140.

Wenkert’s Synthesis of a ( *bp-vetivone Precursor

Spirocyclic Systems

283

The final spirovetivane synthesis we discuss is a 1979 synthesis leading to (*bhinesol, (f)-agarospirol, and (f)-a-vetispirene (Scheme 141).347a9b Inubushi and co-workers begin with acetylacetone which is converted into diene 1155 by House’s method. Diets-Alder reaction of

I I55

1156 I . No.

1157

I , MsCl 2. Zn,NH4CI

~

%

1159

I . Ph3PeOMe 2. H3O+

I . MeLi 2. Jones

HO

1127

1128

6H

1025

1026

%cHo I I59

I . MeLi 2. Jones 3. MeLi 4. H+, benzene

II 6 0

Inubushi’s Synthesis of (*)-Hinesol, Scheme 141. (+ )-Agarospirol, and (f)-a-Vetispirene

1027

284

Other Bicnrbocyclic Sesquiterpenes

1155 with ethyl crotonate gives cyclohexenone ester 1156. Removal of the ketone carbonyl is accomplished by desulfurization of the dithioketal. Alkylation of the enolate ion of 1157 with ethyl 3-bromopropionate is stereoselective, but proceeds in only 19% yield. The resulting diester 1158 is subjected to acyloin condensation and the a-hydroxy ketone so obtained is reduced to obtain the Caine-Buchi enone 1127. This material is elaborated by the indicated sequence to obtain a 3:2 mixture of (kl-hinesol (1026) and (*)-agarospirol (1025). A minor modification allows production of (*)-a-vetispirene (1027) from acyloin 1159. Because of the large number of spirovetivane syntheses that have been published, it is appropriate to compare them against one another. This we do in Table 2. It includes the three syntheses (Marshall’s p-vetivone, Marshall’s hinesol, and Deslongchamps’ agarospirol) which were reviewed in Volume 2 of this work. From the standpoint of efficiency, the Stork synthesis is by far the best. However, it should be noted that this synthesis is easily applicable only to p-vetivone. The Dauben synthesis is almost as efficient in the number of steps, although not in overall yield. However, the Dauben synthesis is highly adaptable and easily leads to almost all members of the class. Many of the syntheses are clearly not useful for producing the end compound in any quantity. Such is the case with the Marshall hinesol synthesis, which was done principally to determine the stereostructure of hinesol. Other impractical syntheses serve only to illustrate some particular chemistry, such as Buchi’s interesting utilization of fulvene 1135 or Trost’s application of his cyclobutanone spiroannelation. At this point, it would seem that it will be hard to improve upon the Stork and Dauben syntheses. However organic chemists seem to have a particular fascination for spirocyclic systems, and we can probably expect the parade of spirovetivane syntheses to continue.

(2) Acoranes The acoranes also contain the spiro[4S]decane nucleus, but have a different substitution pattern from the spirovetivanes. In this group also a large number of syntheses have been recorded. In this section, we

Spirocyclic Systems

285

review syntheses leading to acorone (11611, isoacorone (11621, acorenone (11631, acorenone B (11641, a-acorenol (11651, p-acorenol (11661, P-acoradiene (11671, y-acoradiene (also known as a-alaskene, 11681, 8-acoradiene (also known as p-alaskene, 11691, and the unnamed acoradiene (1170).

1161

1162

I I63

I I66

1167

1168

II64

1169

I I65

1170

There are several synthetic challenges in this group. One is clearly the construction of the spiro[4S]decane nucleus. As will be seen, this turns out to be especially difficult and many methods have been used. For acorone and isoacorone, arranging the relative stereochemistry of the methyl and isopropyl groups in the five-membered ring is no problem, since the isopropyl group is epimerizable and the natural products have the more stable trans stereochemistry at these two centers. The other methyl group is also epimerizable (the thermodynamic ratio of 1161 to 1162 is 7:31. None of the syntheses reported to date have dealt with this problem; all give a mixture of 1161 and 1162. For the acorenones (1163 and 1164) the methyl and isopropyl groups must be cis, and this is usually dealt with by establishing the cis stereochemistry on a cyclopentane precursor before constructing the cyclohexenone ring. In the acorenols (1165 and 11661, as in the acorone isomers, the two side chains are trans. However, the major problem, one that is common to all members of the group, is establishing the correct sense of chirality of the spiro carbon relative to the centers in the cyclopentane ring. Because the sixmembered ring is pseudosymmetric, this stereochemical problem sometimes presents itself as a regiochemical problem. For example, the

286

Other Blcarbocyellc Sesquiterpenes Tabk 1

Principal Investigator

Year

Marshall

1968

Marshall

1969

Startin# Material

Product

Isomer Steps Yield Separations

0

15

4.5%

( ibhinesol

31

<0.1%

1(1:4)"

(*)-agarospirol

18

3.0%

1(1:1)

(*)-B-vetivone

M'yq 0

Deslongchamps 1970

Stork

1973

McCurry

1973

Yarnada

1973

Caine

-'0OEt

(*)-&vetivone

OHCY

1974

Pesaro

1974

Trosi

1975

2(1:1)

11 12

15% 15%

(*)-@-vetivone (*)-hinesol (*)-a-vetispirene (i)-@-vetispirene

23

23 23 25

10.996 11.9% 8.6% 5.3%

0 0 0 0

6

16.0%

1(7:3)

CHO

(i)-hinesol I

0

(*)-@-vetivone (*)-pvetispirene

(*)-/3-vetivone

coon

27%

13'

-8.0

1(2:1)

287

Spirocyclic Systems Table 2 PrhICip81 Invest i~ator Year

Starting Materiri

Magnus

Product

Isomer Steps Yield Se~aratlons

( f 1-hinesol

Dauben

E’ovo (+)-p-vetivone

1975

( f )-hinesol

(*)-a-vetispirene (f)-P-vetispirene

Ramage

1975

(-)-agarospirol

Caine

1916

(f)-a-vetispirene

Biichi

1976

(f)-p-vetivone

1978

(*)-@-vetivone

26

0.2%

7 8

7.4% 8.1%

1(1:1)

0

8

9.0%

1(9:1) 1(3:2)

9

6.4%

0

18

c

1 (3:2)

8

11.0%

1(3:2)

14

8.5%

0

Tori

Wenkert

lnubushi 1979

1(1:3)’

oxooE, I

a. b. c. d.

20d <0.2%

(*:)-hinesol (+)-agarospirol (f)-a-vetispirene

Synthesis completed with minor isomer. Convergent; 5 steps from unsaturated diester 1034 (Scheme 126). No yields given. Projected on basis of Marshall synthesis from 1154.

12 12 12

0.3% 0.3% 0.03%

1(3:2)

0

288

Other Blcubaeyclk Sesqulterpenes

acorenols (1165 and 1166) can be interconverted by isomerization of the double bond from one position to another. However, this process is equivalent to inversion of the spiro carbon. The same relationship exists between the acorenones (1163 and 1164) and the acoradienes 1168 and 1169. As we will show, many of the reported syntheses do not address this interesting problem. The first published report of a synthetic approach to the acorane sesquiterpenes came from Conia and co-workers, who prepared and studied the pyrolysis of ketone 1171.348 Intramolecular ene cyclization occurs via the enol, giving four spiro[4SIdecanes (11721175) in yields of 2046, 30%, 8%, and 22%, respectively. Conversion of 1173 into acorone or isoacorone would require ketone transposition and oxidation of the cyclopentane double bond. However, this has not been carried out.

-’0

L

1171

1172

1173

J

I I74

1175

Pinder reported a serious attempt to synthesize acorone that failed at a late stage.349 Because the basic approach has some merit, and may be adaptable, this unsuccessful attempt is summarized in Scheme 142. p-Methylacetophenone (1176) is converted by a rather tedious nine-step sequence into acid 1178, which is reduced by the Birch method. The side chain is further elaborated to obtain 8-keto ester 1181. This substance undergoes intramolecular Michael addition, affording a diketo ester of the gross structure 1182. Although no attempt is made to control stereochemistry in this synthesis, Pinder notes that 1182 is probably mostly one stereoisomer since the material is “mostly crystalline”.

Spirocyclic Systems

I . Birch 2. CH2N2 3 . H+

I . (MeOIzCO, No H

2. NoOH 3 . McLi

1179

. % MsOOC

2. HQCI2, HgO, H 2 0 , MIOH

NoOMe,

lie1

289

E'1182

1183

Scheme 142.

Pinder's Unsuccessful Approach to (*)-Acorone

Indeed, since the Michael addition is probably reversible, all four stereocenters in 1182 are effectively epimerizable. Thus, it might not be surprising if one or two stereoisomers predominate heavily. The first report of a successful acorane synthesis came from Kasturi and Thomas at the Indian Institute of Science in B a t ~ g a l o r e The .~~~ reported synthesis (Scheme 143) begins with p-methoxyisobutyrophenone (1184) which is converted by steps similar to those used by Pinder into enone 1187. Note that the relative stereochemistry of the methyl and isopropyl positions is established in reduction of methyl ketone 1186-undoubtedly 1187 is a 1:l mixture of diastereomers. Cyclization of 1188 gives 1189. In this step, a new asymmetric carbon is created but it is removed immediately when the double bond is hydrogenated. Chirality of the spiro carbon is restored when the tertiary

290

Other Bicarbcyclic Sesquiterpenes

&

2. P B r 3

2. H Z - C O ~ . 3. tip+

OMb

1184

&

1 . LiAIH4

I . EtOOCCH&N+

OM'

I185

3. (COOHI2

OM0

1187

1186

2I . M H2-coti eMgI I

@

0

1189

& :%, It88

-H20_

4. - H e r 1191

0

Scheme 143.

3. Br*

H 1190

1163

I I64

*

3. Mg 4 . CH3CH0 5 . COI

0 II92

I I93

Kasturi-Thomas Synthesis of Acorenone Isomers

alcohol 1190 is dehydrated. Four final steps convert the cyclohexene ring to the correct cyclohexenone oxidation level. The final product is probably a nearly equimolar mixture of the two acorenones (1163 and 1164) along with the unnatural isomers 1192 and 1193. In 1972, Naegeli and Kaiser, of Givaudan in Switzerland, isolated acoradiene 1170 from vetiver oil. As part of their structure elucidation, they synthesized the compound as shown in Scheme 144.3s1 Dehydrolinalool (1194) cyclizes thermally to a mixture of alcohols 1195 which is converted by Claisen rearrangement into ketone 1196. Catalytic hydrogenation of the isopropenyl double bond is followed by addition of vinylmagnesium bromide. The resulting allylic alcohol undergoes

Spirocyclic Systems

1195

1194

I I97

291

I I96

1198

1170

1199

& / &+&+&* OH

benzene

I200

Scheme 144.

1201

1202

I203

Naegeli-Kaiser Synthesis of Acoradiene 11 70

stereospecific cyclization when treated with stannic chloride. Acoradiene 1170 and the hexahydroazulene 1199 are formed in a ratio of 3:7. Similar cyclization of the dehydro derivative 1200 is much cleaner, affording a hydrocarbon mixture containing 75-80% of acoratriene 1201 in 65% yield. The principal by-products in this case are the double bond isomer 1202 and the simple dehydration product 1203. In 1973 Ramage and co-workers at Liverpool reported a synthesis leading to (-)-a-acorenol and (+)-fl-acorenol (enantiomers of the actual natural products).3S2 The syntheses (Scheme 145) begin with the diastereomeric unsaturated ketals 1113 and 1114 which have been discussed earlier in connection with the Ramage agarospirol synthesis (see Scheme 136). In the present case, the ketal grouping of 1113 is hydrolyzed and the resulting ketone is nitrosated. Treatment of the oximino ketone with chloramine gives a diazo ketone which undergoes photochemical Wolff rearrangement. The resulting acid is esterified to obtain methyl ester 1205. Addition of methyllithium to 1205 yields

292

Other Bicnrkyclic Sesquiterpenes

1113

1204

1114

Scheme 145.

I205

1207 (rnl-1166)

Ramage's Synthesis of (-)-a-Acorenol and ( + )-P-Acorenol

(- 1-a-acorenol, the enantiomer of natural a-acorenol (1165). Applica-

tion of the same sequence to unsaturated ketal 1114 affords (+)-p-acorenol (1207). The apparent stereospecificity of the Wolff rearrangement probably results from equilibration during esterification of the initially formed acids. There is no reason to expect kinetic stereoselectivity in hydration of the ketenes (1208 and 1209) which intervene in the

1208

1209

two cases. In fact, if there is some steric factor which favors addition to one face of the ketene double bond in 1208, it should favor the other face of 1209. In the same year, Marx and Norman also published a synthesis leading to the enantiomers of two acorane sesquiterpenes, (- by-acoradiene and (+) - 6 - a ~ o r a d i e n eThey . ~ ~ ~later ~ extended the synthesis to (- bacorone

Spirocyclic Systems

293

and (+I-isoacorone, the enantiomers of 1161 and 1163.353bThe Marx synthesis (Scheme 146) begins with (R)-(+)-pulegone (1210) which is

&@

:?OE+b

z":,

&cooE'

1210

1211

\

1

1213

1214

1212

1216

1215

COOEt

HOAc)

x

I . (CH20H12, H+ 2. LiAIH, 3. n30+

1217

ditto

1219

1220

1221

(enf- 1168)

1223 (ent-II6I)

Scheme 146.

1222 (ant-1169)

1224

-

(ent 1162)

Marx's Synthesis of (-)-Acorone, (+)-Isoacorone, (-1-y-Acorodiene and (+)-8-Acoradiene

294

Other Bicrrboryclic Sesquiterpenes

transformed into 8-keto ester 1212 by a literature procedure. This compound is converted by a further four-step sequence into the methylene cyclopentanone 1214 which undergoes Diels-Alder reaction with isoprene to give a mixture of stereoisomers 1215-1218. Although the thermal cycloaddition ratio is poor (45:25:20:10), the stannic chloride-catalyzed addition is more selective, giving 1215 to 1218 in a ratio of 69:27:3:1 in 48% yield. The major isomers 1215 and 1216 were separated by preparative hplc and used to complete the synthesis. Addition of isopropyllithium to 1215, followed by dehydration of the resulting alcohol gives a 7:3 mixture of 1219 and 1220, which is found to be the enantiomer of y-acoradiene (1168). Similar treatment of 1216 gives dienes 1221 and 1222, the enantiomer of 6-acoradiene (ll69), also in a ratio of 73. Hydroboration of both double bonds in diene 1219 yields a mixture of diastereomeric diols, which is oxidized to obtain a mixture of diones 1223 and 1224-the enantiomers of natural acorone (1161) and isoacorone (11621,respectively. Although the Marx synthesis played a role in correlating the acoradienes with acorone and isoacorone, it is poor as a source of any of the compounds because of the number of isomer separations which are necessary. In 1973, Oppolzer provided a synthesis of (*)-p-acorenol (1166) and (k)-p-acoradiene (1167).354aThe basic scheme was later extended to provide syntheses of (f1-acorenone B (1164)354band (fbacorenone (1163).354cThe Oppolzer synthesis (Scheme 147) exploits the intramolecular ene reaction of diene 1226 to create the spiro l4.51decane skeleton. Diastereomers 1227 and 1228 are produced in a ratio of 1:l and can be equilibrated by base to a 3:2 mixture. They are separated by silica gel chromatography. Note that the ene reaction can only give one relative stereochemistry for the secondary methyl and spiro carbons. The trans isomer (1228) is converted by allylic oxidation into enone 1229, which is treated with one equivalent of methyllithium to provide a tertiary alcohol. The corresponding methoxymethyl ether is pyrolyzed to obtain cyclohexene 1231. Treatment of ester 1231 with methyllithium gives (*)-@-acorenol (1166) which is dehydrated by heating with alumina to obtain (k:)-p-acoradiene (1167). For the synthesis of (*I-acorenone B, the cis isomer 1227 is used. Conversion to the terti-

Spirocycllc Systems

I . MeLi 2. ~ - T ~ o H , * '6%

1233

I

1236

1163

Scheme 147.

Oppolzer's Syntheses of Acoranes

295

296

Otber Bicarbocyclic Sesquiterpenes

ary alcohol, which is dehydrated by heating with alumina at 220°C gives diene 1232. The more reactive isopropenyl double bond is hydrogenated and the cyclohexene allylically oxidized to obtain an enone which is acetoxylated to give 1234. Treatment of 1234 with methyllithium gives a diol which reacts with p-toluenesulfonic acid in benzene to give (f)-acorenone B (1164). The unusual dehydration leading to 1164 apparently involves a pinacol-type rearrangement:

+-* HO

HO+

For preparation of (*1-acorenone alkene 1233 is hydroborated using disiamylborane to obtain ketone 1235. Introduction of a methyl and the double bond provides (st )-acorenone (1163). Wolf and Kolleck reported the stereorandom synthesis of (f)-acorenone B, which we summarize in Scheme 148.355The synthesis

1164

Scheme 148.

Wolf's Synthesis of (kt)-Acorenone B

Spirocyclic Systems

297

begins with (*)-camphorone (1237) which reacts with 4-pentenylmagnesium bromide to give alcohol 1238. This substance undergoes slow solvolysis to give a mixture of diastereomeric formates (1239) in 32% yield. Hydrolysis of the ester and oxidation of the resulting secondary alcohol gives a 1:1 mixture of diastereomeric ketones (1240). After hydrogenation of the double bond, a mixture of isomers (1241) is obtained in which the desired isomer constitutes 46%. The mixture is separated by preparative tlc and the desired ketone methylated. After introduction of the double bond, (f)-acorenone B is obtained. It is a little perplexing that the a-cyclization leading to 1239 is not more stereoselective. One would have thought that the methyl substituent would have provided some steric hindrance. For example, see the Diels-Alder reaction of enone 1214 (Scheme 144). Wolf et al. have also reported a similar study leading to the unnatural product 4-epiacorenone B.356 Trost and co-workers have applied “secoalkylation” to the synthesis of (*bacorenone B as shown in Scheme 149.357 2-Methylcyclopentanone is converted into 2-isopropyl-5-methylcyclopentanone(1244) by a threestep sequence featuring exhaustive methylation of the p-acetoxy enone 1243 by lithium dimethylcuprate. Formation and rearrangement of the oxaspiropentane gives a single cyclobutanone, 1245. Since 1244 is a mixture of cis and trans isomers, equilibration must occur under conditions of the ylid reaction. It is reasonable that the less hindered cis isomer reacts faster than the trans and it is also reasonable that attack of the ylid occurs exclusively trans to the two alkyl groups flanking the carbonyl. After conversion of 1245 to its a-formyl derivative, acidic conditions promote reverse Claisen reaction and dehydration to the enol lactone 1246. This material is reduced by diisobutylaluminum hydride to a lactol which is oxidized by chromic acid to lactone 1247. Addition of methyllithium to 1247 gives a new hemiketal which reacts with ethanedithiol to give 1248. After oxidation of the primary alcohol, regeneration of the ketone, and base-catalyzed aldolization, spiro[4.5ldecenone 1250 is reached. This key intermediate is converted into (fbacorenone B (1164) by a four-step sequence featuring a novel use of sulfenylation for transposing the enone chromophore by one posi-

298

Other Bkrrbocyclic Sesquiterpenes

$

4 1243

I'

~~~oE'*,

2. P-TIOH

1245

1244

I . DIBAL

2 . H2Cr04

1246

1248

Scheme 149.

BF3

1247

1249

1251

I . MeLi

____c_

1250

1252

1164

Trost's Synthesis of (&:)-AcorenoneB

tion. The use of the cyclopropylsulfurane to achieve spiroannelation is a novel strategy. However, the 14 steps required to reach 1250 by this procedure are rather laborious. As will be seen, intermediates of this type are more easily accessible by direct Robinson annelation of the readily available aldehyde 1255 (see Schemes 150 and 151). White and co-workers developed the interesting stereoselective synthesis of (-bacorenone B which is documented in Scheme 150.358The synthesis begins with (R)-(+)-limonene (12531, which is transformed

Spirocyclic Systems

299

I . SOCI.

1257

1258

1260

I259

1261

I262

1164

Scheme 150.

White’s Synthesis of (-1-Acorenone B

by a precedented four-step sequence into unsaturated aldehyde 1255. Reformatsky reaction of this aldehyde with unsaturated bromo ester 1256 gives a lactone (1257) which is subjected to successive hydrogenation and hydrogenolysis to obtain unsaturated acid 1258. This compound is transformed into a diazoketone, which is decomposed by copper powder. The carbene attacks the double bond exclusively trans to the isopropyl group, giving rise to tricyclic ketone 1260. Electrophilic cleavage of the cyclopropane ring gives enone 1261, which undergoes

300

Other Bicorbocyclic Sesquiterpenes

1266

1268

Scheme 151.

3. Jones

1267

1161

1162

Dolby’s Synthesis of ( *)-Acorone and (*)-lsoacorone

hydrogenation trans to the isopropyl group. After introduction of the double bond, (-bacorenone B (1164) is obtained. In 1977, McCrae and Dolby reported the nice synthesis of (f)-acorone and (*)-isoacorone, which we outline in Scheme 1S1.359 Unsaturated aldehyde 1263 (from isoprene and acrolein) is alkylated with 1-trimethylsilyl-3-bromo-1-propyne by Stork’s metalloenamine method. The resulting acetylenic aldehyde 1264 is hydrated to a 1,Cketo aldehyde which undergoes subsequent aldolization to produce the spirocyclic dienone 1265. The isopropyl and methyl groups are introduced by exhaustive methylation of 1266 with lithium dimethylcuprate. Hydroboration of enone 1268, followed by oxidation with Jones’ reagent gives a 55:45 mixture of (f)-acorone (1161) and (f)-isoacorone (1162). Pure 1161 may be obtained by direct crystallization from this mixture. Lange has reported a very direct synthesis of (-bacorenone (Scheme 152) .360 The synthesis begins with a,punsaturated aldehyde, 1255 (see

Spirocyclic Systems

1255

I269

HOAC

-&

301

I270

2. I . p-TtOH, MeMgI C6H6 9

Scheme 152.

A

1163

Lange's Synthesis of (-)-Acorenone

Scheme 150). Hydrogenation occurs from the side of the molecule trans to the isopropyl to give an epimeric mixture of aldehydes (1269) which is subjected to Robinson annelation with methoxymethyl vinyl ketone to obtain 1270 in 35% yield. In spite of the low yield, this represents a direct and highly stereoselective entry into the acorane skeleton. Treatment of 1270 with methylmagnesium iodide gives a tertiary alcohol which is refluxed with p-toluenesulfonic acid in benzene to obtain (*hacorenone (1163). Pesaro and Bachmann have reported a synthesis of (-)-amrenone and (-)-acmenone B which is very similar to the Lange synthesis (Scheme 153).36* In their synthesis, the Robinson annelation is carried out with methyl vinyl ketone and the yield is lower (only lo%, or 20% based on reacted aldehyde). Manipulation of the cyclohexane ring features a novel hydroboration step. The initial borane adduct, being a 8-hydroxyborane, undergoes elimination to give a new cyclohexene which is hydroborated again. Upon oxidation, alcohol 1273 is produced. Introduction of the double bond leads to (-bacorenone (1163). To prepare isomer 1164 the allylic alcohol 1272 is dehydrated to diene 1275, which is epoxidized at the more nucleophilic and less hindered double bond. Treatment of the epoxide with p-toluenesulfonic acid results in rearrangement to the P,y-unsaturated enone 1276. The sequence from 1272

Other Bicarbocyclic Sesqulterpenes

302

&OH

Jones_

1273

1272

&

I27 I

1269

1272

I . Br2,CH2CI2 2 . DMF, Liecof

I I63

1274

1275

I276

p-TsOH

PhMe, A

0

1164

Scheme 153.

Pesaro's Synthesis of (-)-Acorenone and (-)-Acorenone B

to 1276 is efficiently carried out in one pot by treating 1272 with a mixture of p-toluenesulfonic acid and rn-chloroperoxybenzoic acid in methylene chloride. Coqjugation of 1276 provides (zk bacorenone B (1164). Martin and Chou have prepared (f)-acorone and (*)-isoacorone by a route which is essentially the same as the Dolby route (Scheme 154).362

Spirocyclic Systems

1278

I279

1161

Scheme 154.

303

I I62

Martin's Synthesis of (*)-Acorone and (*bkoacorone

The principal difference is in the method used for adding the cyclopentenone ring. Rather than the propargyl group which Dolby uses, Martin employs 2-chloro-3-iodopropene and does the alkylation using the enamine of aldehyde 1263. A similar method is also used to add the methyl and isopropyl groups to the five-membered ring. As in the Dolby synthesis, the crucial stereochemistry at the secondary methyl center results from attack of the cuprate reagent over the face of the cyclopentenone chromophore syn to the cyclohexene double bond. Table 3 compares the successful syntheses of acorane sesquiterpenes. Note that the only syntheses which do not require isomer separations are those leading to the acorenones, which are clearly the simpler synthetic tasks.

304

Other Blcarbocycllc Sesquiterpenes

Table 3 Principal investigator Year Kasiuri

starting Material

Product

1972

Steps Yield

Isomer Separations

18

-

5

-12%

I (3:7)

(-)-a-acorenol ( + )-p-acoreno\

I5

b

1 (1:l)C

(-)-y-acoradiene (+)-&awradiene (-)-acorone ( + Lisoacorone

10 10 12 12

0.9% 2(7:3. 3:7d) 0.9% 2(3:7", 3:$) 0.5% 3(7:3, 73, 1:I) 0.5% 3(7:3,73, I : ] )

( *)-p-acorenol ( *)-p-acoradiene

7 8 9 11

5.5 5.5

1(1:1) 1(1:1)

14.5%

1(3:2)

(*)-acorenone B

9

1.2%

1(1:1)

(*Lacorenone B

18

4.4%

0

( *Lacorenones

l~l~l~l~l)a

OM*

Naegeli

1972

Ramage

1973

Marx

1973

(-bacoradiene 1170

0"

1975

Oppolzer

1973 1975 I977

Wolf

b

COOEt

1975

(*)-acorenone B (*)-acorenone

11.3%

1(3:2)

+O

Trosi

1975

305

Spirocyclic Systems

Table 3 Prlnclpal Investinator Year White

1976

Dolby

1977

Lange

1977

Pesaro

1978

Martin

a.

Startlnp Material

I

Product (-Lacorenone

Stem Yleid

B

(*)-acorone

4r

1978

(-1-acorenone

(- 1-acorenone (-bacorenone

B

Isomer Se~aratlons

15

1.5%

0

10 10

4.9% 4.9%

1(1:1)

1(1:1)

8

12.2%

0

11 10

2.3% 3.4%

0 1(3:1)

( *bacorone

The paper does not consider stereochemistry. However, a consideration of the chemistry suggests that 4 isomers are produced in approximately equal amounts.

b.

Not reported.

c.

The two isomers are each employed, one for one target and one for the other.

d.

The minor isomer is employed.

306

Other Bicarbacyclic Sesquiterpenes

(3) Axisonitrile-3 (

+ )-Axisonitrile

(1280) is an unusual sesquiterpene isonitrile which

occurs in the marine sponge Axinella cannabina. Caine and Deutsch

I280

have synthesized the enantiomer of 1280 as shown in Scheme 155.363 (+)-Dihydrocarvone is converted by a known procedure to the bicyclic ketol 1281,364 which is hydrogenated to obtain 1282. Oxidation of the dienyl acetate 1283 by the Kirk-Wiles procedure gives 42% of the enone arising from simple hydrolysis of 1283 and 56% of a 6:l mixture of alcohols 1284 and 1285. The major isomer is obtained by chromatography on Florisil. After protection of the hydroxy group, the A’** double bond is introduced by the selenoxide method. Irradiation of 1287 in anhydrous dioxane gives the lumiproduct 1288, which is hydrogenated and methylenated to obtain vinyl cyclopropane 1290. Lithium in ethylamine reduces 1290 with inversion of configuration at the methyl center, giving the spiro [4.5ldecene 1291. The remaining steps are required to change the hydroxy group into the isonitrile function of the proper relative stereochemistry. This is a fine example of a well-planned stereorational synthesis. The sole point of uncertainty is the stereochemical outcome of the vinyl cyclopropane reduction, and this aspect was tested prior to the synthesis outlined in Scheme 155 with simple analogs of 1290. (4) Chamigrenes

Chamigrenes contain the spiro (5.51undecane nucleus. In Volume 2 of this series, we reported syntheses of chamigrene (1294) and ( 51-a-chamigrene (1295). We now report three more syntheses of

spirocyclic Systems

1281

AcO

m-CPBA

1283

2.

A

q/ +oa OH

OH

1284

1285

I . LDA

I . reporote

OMe

1282

hw

3. HI04

,H+

OH

I287

1286

I . p- TrCl , C A N

Li -EtNH2

2. KN3, 18-crown-6 OH

3. L i A I H 4

1291

I292

I . HCOOH, AcpO 2. ~ - T I C I . C ~ H S N

NC

1293

Scheme 155.

307

Caine-Deutsch Synthesis of

( *)-Axisonitrile-3

308

Other Bicarbocyclic Sesquiterpenes

(f)-a-chamigrene as well bromo-a-chamigrene (1296).

I294

as

two

syntheses

1295

of

(*)-lo-

1296

In 1974, White, Torii, and Nogami reported the synthesis of (*)-a-chamigrene which we summarize in Scheme 1S6.365The synthesis begins with acid 1298, which is obtained in one step from anisole and 4,4-dimethylbutyrolactone. Birch reduction, ketalization, conversion to the acid chloride, and reaction of the latter with diazoethane provide the diazo ketone 1300. Copper-catalyzed decomposition of 1300 affords t:'icyclic ketone 1301 in almost 30% overall yield from acid 1298. Lithium-ammonia reduction converts 1301 quantitatively into

OMe

I . ICH2OH)2, H+ 2. NaH. (COCI)2

I . Birch 2 . 3% HCI

1298

3. C H I J C H N ~

I300

1299

1301

I302

I . MnCI,

I . MeLi 2. DMSO,OOO

C5HC.N 2. DMS0,60° 3. HSO+

1303

I304

Scheme 156.

1295

White's Synthesis of (kba-chamigrene

Spirocyclic Systems

309

spiro[5.5lundecanone 1302. Reduction of the carbonyl gives the cis alcohol 1303, which is dehydrated by heating the derived mesylate in DMSO at 60°C. Ketone 1304 is methylated to obtain a diastereomeric mixture of tertiary alcohols, which is dehydrated by heating in DMSO to obtain (*)-a-chamigrene (1295). The 12 step synthesis proceeds in an overall yield of about 14%. Friter studied the conversion of cis- and trans-p-ionol (1305) into (*)-a-chamigrene as shown in Scheme 157.366Treatment of 1305 with

I306

1305

1307

I308

I . 100-1100 2 . separate 1310

I309

Scheme 157.

I295

Friter's Synthesis of

( *1-a-Chamigrene

1O0/o sulfuric acid in a mixture of ether and acetone gives a mixture of tetraenes 1306-1309. The composition of the mixture depends upon whether cis- or trans-1305 is employed:

1306 1307 1308 1309

trans- 1305 70.5 22.5 4.5 2.5

cis- 1305

54.5 36.0 7.0 2.5

Other Bicrrbocyclic Sesquiterpeaes

310

When the mixture of tetraenes is heated at 100-11O"C, electrocyclic closure of isomers 1307 and 1309 occurs to afford (kbdehydroa-chamigrene (13101, which may be obtained in a pure state by chromatography. If trans-1305 is employed, the isolated yield of 1310 is 19%; if cis-1305 is used, 1310 is obtained in 25%. Hydrogenation of 1310 using Lindlar's catalyst provides (*1-a-chamigrene (1295) and an isomer in a 3:l ratio in quantitative yield. Pure 1295 can be obtained by chromatography. In 1979, Iwata, Yamada, and Shinoo at Osaka University reported a synthesis of ( f 1-a-chamigrene which is almost a duplicate of White's 1974 synthesis (Scheme 158).367 The only point of substantial difference in the two syntheses is that Yamada and co-workers carried out the carbene cycloaddition on the phenol ring (1312) leading to the spiro dienone 1313 while White and co-workers used an alkene (1300) and must reductively cleave the cyclopropane ring to reach the spiro [5.5lundecane system. The modification is not especially effective, since Yamada reported a yield of only 20% in the cyclization while White obtained a yield of at least 45% for the two steps from 1300 to 1302.

I . Ac20 2 . SOCl2

1311

OH

.

3 . CH3CHN2 4. Na2C03, McOH

CuCI2

1:

OH

.

I . NaBHq, 2. Hp-Pd/C

A*C6H6

1312

1313

OH

2 . DMSO. 1314

0 1304

Scheme 158.

I . MeMgI 2. Si02-RC13* 1295

Yamada's Synthesis of (kl-a-charnigrene

Spirocyclic Systems

311

10-Bromo-a-chamigrene (1296) is the simplest of several halogenated chamigrenes of marine origin. In early 1976, Wolinsky and Faulkner reported a synthesis of the compound, although it had not actually been found in nature at that time.368 Later, in 1976, Fenical and Howard discovered the compound in the red algae Laurencia pacific^.'^^ The Wolinsky-Faulkner synthesis (Scheme 159) begins with geranylacetone (1319, which is treated with stoichiometric amounts of bromine and silver fluoborate to effect cyclization; bromo ether 1316 is obtained in 20% yield. Treatment of 1316 with p-toluenesulfonic acid in benzene

1316

1315

1318

1317

1296

Scheme 159. Wolinsky-Faulkner Synthesis of (*1-10-Bromo-a-chamigrene

causes rearrangement to unsaturated ketone 1317. Treatment of the latter with vinylmagnesium bromide gives a tertiary alcohol (1318) which cyclizes upon treatment with acid to produce (*)-lO-bromoa-chamigrene in 26% yield. The cyclization is stereospecific, only one diastereomer being produced, and the synthetic product is identical with the natural product. In a later publication, Ichinose and Kato also reported the synthesis of (*1-10-bromo-a-chamigrene (Scheme 160).370This synthesis also starts with geranylacetone, which is converted into a mixture of the 2E and 22

312

Other Bicsrbocyclic Sesquiterpenes

1315

1319

COOMe

I320

I . separate ____)

2. A I H 3

Br

I323

Br

I324

1296

1297

Scheme 160. Ichinose-Kato Synthesis of ( f ) - 10-Bromo-a- chamigrene

diastereomers of methyl farnesate (1319 and 1320). Although the composition is not stated in the paper, it is known that this transformation normally gives a 2 E : 2 Z ratio of about 3:l. The isomers are separated by chromatography and the minor isomer (1320) is treated with 2,4,4,6-tetrabromocyclohexadienone in nitromethane. A mixture of cyclization products (1321-1323) is produced, from which 1321 can be obtained in 28% yield by crystallization from pentane. The ester group is reduced and alcohol 1324 is cyclized by treatment with iodine in refluxing benzene (conditions employed earlier by Kitahara in a synthesis of (*) - ~ - c h a m i g r e n e ) . ~ ” From the cyclization mixture, (*)-lobromo-a-charnigrene may be obtained in 6% yield by “repeated high

Fused Ring Compounds: 3,6

313

pressure liquid chromatography.” It is not at all clear that this synthesis, which requires three isomer separations and proceeds in an overall yield of no more than 0.4%, is an improvement over the Wolinsky-Faulkner synthesis, which gives the sesquiterpene in an overall yield of approximately 4%.

D. Fused Ring Compounds: 3,6 (1) Bicycloelemene

Bicycloelemene (1325) is a constituent of Bulgarian peppermint oil and also of the cold pressed peel oil of Cirrus junos. Although its full stereostructure has not been proven, it is probably as depicted based on

1325

analogy to simple elemanes. Vig and co-workers have synthesized 1325 as shown in Scheme 161.372Known keto-acid 1326 is transformed by the identical sequence used earlier in the synthesis of ( *)-P-elemene373into dibromide 1332, which is cyclized by the method first used by Corey for the synthesis of (*)-elern01.~~~ The purity and identity of the synthetic 1325 was established by tic and glpc. The yield of pure 1325 was not given. However, in Corey’s synthesis of elemol, the desired diastereomer was produced in 24% yield.374It should be noted that the extra rigidity of 1332 as a result of the cyclopropane ring may provide an entropic advantage for cyclization in this case.

314

Other Bicrrbocycllc Sesquiterpenes

I . EtOH, H+ 2 . (~H,oH),.H+~

1326

2. Collins

H I327

&

COOEt

H

1330

k$

COOEt

0

II

(Et 012PVCOOEt NOH

EtOOC

wBr9

I . LiAIHq 2 . PerS

*

1331

Ni(C0)4*

1332

Scheme 161.

1325

Vig's Synthesis of (*)-Bicycloelemene

(2) Sirenin and Sesquicarene The interesting sesquiterpene sperm attractant sirenin (1333) and its related hydrocarbon sesquicarene (1334) have attracted a good deal of synthetic attention. In Volume 2 of this series, we reported five

Fused Ring Compounds: 3,6

315

OH I333

1334

syntheses of 1333 and four syntheses of 1334. Rapoport has published Intermediate 1335 has the full details of his first sirenin synthesis.375t376 now been resolved via the diastereomeric ketals formed with optically active 2,3-butanediol. The two enantiomers corresponding to 1335 have

I335

been converted into (-)-sirenin and (+)-sirenin by the procedure described earlier for the racemic series.375 Correlation of natural (-)-sirenin with natural (+)-sesquicarene shows that the two have the same absolute configuration, as shown by 1333 and 1334. Garbers and co-workers prepared unsaturated ester 1343, an intermediate in the Rapoport synthesis of ( f ) - ~ i r e n i n ;from ~ ~ methyl (*)-perillate (1336) as shown in Scheme 162.377 The synthesis is based on the observation that certain alkenes undergo effective Prins addition when treated with chloromethyl ethers and Lewis acids. Thus, 1336 reacts with allyl chloromethyl ether to give a mixture of diastereomeric adducts 1337. Cyclization gives cyclopropanes 1338 and 1339 in a ratio of 2:3 in 30% yield, based on 1336. After cleavage of the allyl ether and transesterification with methanol, alcohols 1340 and 1341 are separated by chromatography and the minor isomer is converted into tosylate 1342. Tosylate displacement is faster than conjugate addition when 1342 is treated with lithium diisobutenylcuprate and ester 1343 is produced. This material has been converted into (&)-sirenin in three steps.

316

Other Bicsrbocyclic Sesquiterpenes

I-BuOK COOMe

BF3

COOMe

I336

1337

I . AcpO, B F j COOMI

TrO

A

COOMI

A ki

2 steps

COOMe

1342

*

2. MeONo, MeOH

COOMe

1333

1343

Scheme 162.

Garber’s Synthesis of (*)-Sirenin

Although the yield and the isomer ratio in the key reactions (1336-1338 + 1339) are not especially good, this route makes sirenin available in only eight steps from a relatively available natural product. Although the Garbers synthesis was done with racemic methyl perillate, the optically active ester is available and it is known that it undergoes Prins reaction without racemization. Thus, the synthesis in Scheme 162 could probably be applied to prepare (-)-sirenin.

Fused Ring Compounds: 3,6

317

However, Hiyama and co-workers have discovered a route to sirenin and sesquicarene which is even more effective (Scheme 163).378 The 7,7-dichloronorcarane 1344, available in three simple steps from cyclohexanone, is added to an ether solution of the cuprate reagent prepared

3 steps

i

"

9

0

.

( L ) p c u L i

23.. Me1 H30+

0 1344

1335

Scheme 163.

Hiyama's Synthesis of (+)-Sirenin

from S-bromo-2-methyl-2-pentene.After 1 hour at -2O"C, excess methyl iodide is added and the mixture is kept at room temperature overnight. After hydrolysis of the ketal, ketone 1335 is obtained as a single stereoisomer in 44% yield! The reaction apparently involves displacement of one chlorine by the cuprate reagent to give a 7-chloro7-alkylnorcarane, which undergoes transmetallation to the 7-lithio-7-alkyl compound. The stereoselectivity of the process must result from attack of the cuprate reagent at the more accessible exo face to give 1345, which is stereospecifically transmetallated to give 1346 in which the lithium is endo. The latter compound must undergo displacement on methyl iodide with retention of configuration at the cyclopropyl center. Since ketone 1335 has been converted into both sirenin and sesqui-

318

Other Bicnrbocyclic Sesquiterpenes

4

1345

I346

I347

~ a r e n e , ~ ' ~the , ~ Hiyama '~ contribution makes both terpenes much more accessible.

E. Fused Ring Compounds: 5 3 (1) Pentalenolactone

Pentalenolactone (1348) is an antibiotic sesquiterpene containing a bicyclo[3.3.0]octane carbon framework. In addition to the construction of the intricate framework itself, there are several interesting stereochemical problems which must be reckoned with. The first is arranging

1348

for the secondary methyl group to be endo with respect to the bicyclo [3.3.0]octane unit. The second problem is at C,-the hydroxymethyl group must be exo at this position. Finally, the most difficult problem is the stereochemistry at C,, the spiro carbon.

Fused Ring Compounds: 5,5

319

The first successful synthesis was reported in 1978 by Danishefsky and co-workers (Scheme 164).380 The readily available diene 1349 is

&

MeOOC MeOOC

1 3

Y0: H L OH-

4 / \ 0 / 4

/

& 0

I350

I349

I.

050,

2

Pb(0Ac)T

Me I

COOMe

2 . Rorenmund

0

COOMe

COOH

4 . H+

&) I357

1356 I

2

(Me2N)2CHO+ S i 0 2 , H20

3 NoBH, 4 MsCI,C5H5N 5 C5H5N. A

I . LDA

2. PhSICl

*

&'OM.

3, NoIOq*

A00

0

I358 I

1359

Scheme 164.

DIBAL

2

t-BuOOH,

3 4

Jones KOH,H20 THF

1348

( f )-Pentalenolactone: Danishefsky's Synthesis

320

Other Bicrrbocyclic Sesqulterpenes

hydroxylated and the resulting exo diol is protected as the acetonide. The esters are saponified and the diacid dehydrated to obtain anhydride 1350. Diels-Alder reaction with the Danishefsky diene occurs in nearly quantitative yield. When the resulting adduct is treated with barium hydroxide a series of transformations occurs, resulting in an acid, which is methylated to obtain 1351. Scission of the cyclohexenone ring by hydroxylation and lead tetraacetate oxidation gives a lactol (1352) which is reduced. After saponification of the ester, acidification provides lactonic acid 1353. The carboxy group is transformed into an aldehyde by Rosenmund reduction of the corresponding acyl chloride. Wittig reaction occurs without interference from the lactone. After hydrolysis of the acetonide, the resulting diol is oxidized to obtain diacid 1354, which is selectively esterified to 1355. The carboxy group is converted into an acyl chloride which undergoes Darzens acylation to yield 1356, in which the bicyclo [3.3.0]octane nucleus is finally established. Wittig methylenation gives a diene (1357) which is hydrogenated using Wilkinson’s catalyst to obtain solely the endo methyl diastereomer 1358. The lactone ring is methylenated by a five-step sequence commencing with a Vilsmeier-type formylation. The second cyclopentene double bond is introduced via the selenoxide to obtain deoxypentalenolactone methyl ester (1359). Direct epoxidation of the methylene lactone with alkaline hydrogen peroxide fails, providing instead the 9-epi compound. However, if the lactone carbonyl is first reduced and the resulting hemiacetal is epoxidized using the Sharpless reagent, the correct stereochemistry is obtained at C,. The synthesis is completed by oxidation of the hemiacetal back to the lactone and saponification of the methyl ester. The synthesis requires 33 steps and proceeds in 0.7%overall yield to the methyl ester of pentalenolactone. A second synthesis was communicated by Schlessinger and co-workers in 1 980.381 The Schlessinger synthesis (Scheme 165) commences with the enol ether of 2-methyl-l,3-~yclopentanedione(13601, which is alkylated by the Stork-Danheiser method with ally1 bromide. The kinetic enolate is then regenerated and added to diethyl maleate. The product (1361) is a 1:l mixture of diastereomers. The bicyclo~3.3.0loctanesystem is now established by reaction of 1361 with sodium hydride in the

Fused Ring Compounds: 5,5

321

NaH,

EtOOC~COOEt

I360

1361

I . (Me3Si)2NK

'OM'

2 . c02 3. CHzNz

bOOMe

I362

I . NaBH4 2 . MsCl 3. collidine,)

A

1363

I . DIBAL

I364

I . 031

2. H ~ O +

HO

3. Mn02

I

t

CHo

I . (MOO)~CH H+, MeOH 2. Ph#=CH2*

1365

:.,

:;:h3Pl3RhC1,

, ,

3. Jones 4. CHZNz

2. CH20, EtZNH.

Me0 1386

-

1387

-&COOMm

4 steps

1348

(Scheme 164)

I359

Scheme 165.

(*)-Pentalenolactone: Schlessinger's Synthesis

322

Other Bicubocycllc Sesquiterpeoes

presence of dimethyl carbonate. This is perhaps the key reaction in the Schlessinger sequence, as any number of things might (but don’t) go 1362 is wrong. The resulting highly functionalized bicyclo~3.3.0~octane thus obtained in a two-step sequence which proceeds in 68% overall yield. Note that, although 1361 is a 1:l mixture of diastereomers, 1362 is obtained as a single diastereomer. It is clear from the yield in the Claisen step that epimerization occurs at some stage. The cyclopentanone is carbonated by reaction of its enolate (formed under conditions of thermodynamic control) with CO,. Esterification of the resulting P-keto acid with diazomethane gives 1363. The more reactive carbonyl is reduced and the resulting P-hydroxy ester is eliminated to obtain 1364. At this point, all three carbonyl groups are reduced by diisobutylaluminum hydride and the enol ether is hydrolyzed, with subsequent dehydration of the P-hydroxycyclopentanone. The sequence is completed by oxidation of the allylic alcohol with manganese dioxide. This important three-step series of reactions proceeds in 65% overall yield. Ozonolytic cleavage of the ally1 double bond gives a hemiacetal (1365) which is treated with methyl orthoformate and then with methylenetriphenylphosphorane to obtain diene 1366. This compound is hydrogenated using Wilkinson’s catalyst, as in Danishefsky’s synthesis. Surprisingly, both the endo and the exo methyl diastereomers are obtained. After removal of the protecting groups, oxidation of the aldehyde and hemiacetal, and esterification, 1367 is obtained as a 2:l mixture of isomers, with the endo isomer predominating. It is postulated that the endo methoxycarbonyl group in Danishefsky’s hydrogenation substrate (1357) provides significant assistance in directing the catalyst to the exo face of the bicyclo[3.3.0~octane system. The methylene group is introduced by carbonating the lactone with magnesium methyl carbonate, and treating the resulting acid with aqueous formaldehyde and diethylamine. Methylene lactone 1359 can be converted into ( *bpentalenolactone by the Danishefsky method. The Schlessinger synthesis is considerably shorter than the Danishefsky synthesis (24 steps) and gives pentalenolactone methyl ester in 1.2% overall yield.

Fused Ring Compounds: 5'6

323

F. Fused Ring Compounds: 5,6 (1) Hypolepins and Pterosin E

The hypolepins are a group of indanones which were isolated from the fern Hypolepis puncruta Mett. Three compounds, hypolepin A (1368a), hypolepin B (1368b)' and hypolepin C (1368~)comprise the class. Hypolepin B has been converted into hypolepin A (POCl,-C,H,N) and hypolepin C (r-BuOK-CH31), thus correlating the three compounds. x* I368a: X = Cl 1368b: X = OH 1 3 6 8 ~ :X = OCH3

Hayashi and co-workers synthesized hypolepin B as shown in Scheme Ester 1369 is acylated with acetyl chloride to obtain ketones 1370

COOEt

COOEt

+& OOEt

A

I369 I . separate 2. ( i - P r O L A I ,

82H6

8

/'-Pro5

0

% 1371

I370

+Ti7

'

"Y

I372

I.

soc12,

I . 2.4-DNP 2 . 82H6

CSHJN 1 A

2. OH3. ( C F 3 C 0 ) 2 0 4 . t-BuOK, Me1

3. I373

Scheme 166.

n2o2.-onb

4. HCOOH, cuco3. I000

I368b

Hayashi's Synthesis of Hypolepin B

324

Other Bicubocyclic Sesquiterpenes

and 1371 in a ratio of 2:l. The major isomer is transesterified using aluminum isopropoxide in isopropyl alcohol. Reduction of the ketone with diborane gives an alcohol (1372) which is dehydrated by thionyl chloride in refluxing pyridine. Hydrolysis of the ester and cyclization of the resulting acid gives an indanone which is permethylated to obtain 1373. The carbonyl is protected as the 2,4-dinitrophenylhydrazoneand the vinyl group is hydrated by hydroboration and oxidation. Hydrazone hydrolysis is accomplished by heating with formic acid and in the presence of copper carbonate to give hypolepin B (1368b). Pterosin E (1374)is a nor-sesquiterpene which is structurally related to the hypolepins. It may be isolated from the leaves of bracken, Pteridium aquilinum var. Lariusculum, along with a large number of other indanones of similar structure. For a time there was disagreement as to the position of the two-carbon side chain on the nucleus (C, or C6). To

4

I374

settle this question Krishna Rao synthesized pterosin E as shown in Scheme 1 67.383 3,5-Dimethylbenzyl bromide is used to alkylate diethyl methylmalonate to obtain 1374, which is chloromethylated to 1375. The

NaOEl

L

C

O

O COOEt E t

'"''2*

HC I

I375

I . DMF NaCN.

5

___)

H0O*OH

C1&COOEt \

H

O

I376

O

S

2 . H30+

1377

Scheme 167.

1374

Krishna Rao's Synthesis of (*)-Pterosin E

COOEt

Fused Ring Compounds: 5,6

325

benzylic chloride is displaced and the cyano-diester hydrolyzed and decarboxylated to diacid 1376. Friedel-Crafts cyclization results from treatment of 1376 with polyphosphoric acid; the product is ( f I-pterosin E (1372).

(2) Bakkenolide A Bakkenolide A (1378) is an interesting spiro lactone which has been isolated from the flower stems of Petasites japonicus Max. It was suggested that it may arise biogenetically from the eremophilane fukinone (434) by some path such as the following:384

I378

In fact, Hayashi, Nakamura, and Mitsuhashi succeeded in preparing bakkenolide A from fukinone in just this way (Scheme 168).385 Alkaline epoxidation followed by treatment of the resulting epoxide with ethanolic sodium hydroxide gives a mixture of acids which is esterified with diazomethane. Compounds 1379-1381 are produced in a ratio of 42:42:16. Column chromatography gives pure 1379, which is dehydrated to obtain 1382. This compound is oxidized by selenium dioxide in acetic acid to obtain 33% of bakkenolide A and 57% of aldehyde 1383. Treatment of the latter compound with sodium borohydride gives more bakkenolide A. Ester 1382 has also been prepared by Naya and co-workers by a similar route from f ~ k i n o n e . ~ ~ ~

326

Other Bicarbocyclic Sesquiterpenes

434 I . reparote 2 . S0Cl2.

1381

H I382

I378

Scheme 168.

I380

I379

I383

Hayashi’s “Biomimetic” Synthesis of Bakkenolide A

Evans and co-workers prepared (*I-bakkenolide A by the novel method summarized in Scheme 169.387a,bThe 4.5:l mixture of ketone 1384, prepared by the Piers method,388 is subjected to Lemieux-Johnson oxidation to obtain diones 1385 and 1386. The major isomer is separated by column chromatography and cyclized by treatment with strong base. Hydrogenation of the resulting enone gives the cis-fused hydrindanone 1387. This compound reacts with isopropenyllithium to give a mixture of alcohols which is treated with PBr, to obtain allylic bromide 1388, which is probably a mixture of E and Z diastereomers. Treatment of 1388 with the sodium salt of p-toluenesulfonylS-methylcarbazate gives the hydrazone 1389. Treatment of the latter with sodium hydride initiates the sequence of events shown leading to dithioester 1390. The key event is the highly stereoselective [2,31sigmatropic rearrangement of the intermediate carbanion 1389a Compound 1390 is hydrolyzed and oxidized by selenious acid to obtain (*)-bakkenolide A (1377). In an earlier paper, Sarma and Sarkar reported the preparation of ketone 1387 by the same route.389

Fused Ring Compounds: 5,6

&y0 0

os04& NaIO,

1384

+

&I . separate 2. 1-BuOK 3. Hz.Pd/C

00

1385

1386

I

1387

00

327

~NHTS

Br

1389

I388

1

NaH,

J

13890

I390

Scheme 169.

1389b

1378

Evans' Synthesis of (*I-Bakkenolide A

(3) Oplopanone

Oplopanone (1391) is a rearranged cadinane which has been synthesized by Caine and Tuller at Georgia Institute of Technology390 and by Taber and Korsmeyer at Vanderbildt University.391It appears at the outset that there may be a major problem in creating the trans-fused hydrindanone

328

Other Bicarbocyclic Sesquiterpenes

nucleus. However, ketone 1392 had been prepared in the course of the structure proof and was found to be thermodynamically stable. Like-

1391

1392

wise, the epimerizable acetyl side chain of oplopanone was previously shown lo be in the more stable position. Thus, the stereochemical problem reduces to controlling the relative stereochemistry of C,, C, and c7a. Caine and Tuller applied photochemical isomerization of a cyclohexadienone, a much used reaction by Caine and his co-workers for the synthesis of other sesquiterpene systems. The Caine-Tuller synthesis (Scheme 170) begins with Robinson annelation of 2-isopropyl-5-methylcyclopentanone (1393)392;unfortunately, enone 1394 is produced in only

0 I. Mao&OMa E l O K , EtOH 2. KOH, EtOH

1393

I reporate 2 hv,HOAc

.

PAC

o@

:z

0

1:

tyZH)

MI

MI

1394

1395

I.NaBb

~

@OH

3 . Li-EtNH2

Me0 I396

1397

Scheme 170.

I.

kiYHc

2. AcpO I392

I398

Caine-Tuller Sythesis of

1391

( *) -0plopanone

Fused Ring Compounds: S,6

329

17% yield. Although 1394 is a homogenous substance, the dienone produced by oxidation with selenium dioxide in f-butyl alcohol is a 5:l mixture of diastereomers. The isomers are separated by careful silica gel chromatography and the major isomer is photolysed in acetic acid. A single acetoxy ketone, 1396, is produced in 91% yield. Thus, the three important asymmetric centers of oplopanone are established in the correct sense. Removal of the oxygen is accomplished by a precedented sequence. Ketone 1392 is indeed found to be the thermodynamically more stable isomer, since careful hydrolysis initially yields an isomer of 1392 which rearranges to it upon extended treatment with acid or upon refluxing with sodium methoxide in methanol. The acetyl side chain is introduced into 1392 via the acetylene 1397. After hydration of the triple bond, reductive cleavage of the acetoxy group provides (*)-oplopanone (1391). In all, the synthesis requires 13 steps and proceeds in 0.6% overall yield. The Taber-Korsmeyer synthesis is summarized in Scheme 171. Reductive alkylation of the o-methoxybenzoic acid 1399 with p-bromophene-

Hoocd I . Li/NH3 THF

P

b

h

O

y

g7Brt * P

h

DMS,THF

2. P h v B r

3. H 3 0 +

1399

Br

9

OHC\,*

t-BuOK,

@ OHC

Scheme 17 1.

2. I . MeLi Hg(OAC12

*.* 1401

&OH

H20, THF 3. NaBH, 4. PCC

I-BuOH

1404

1400

O

1405

Taber-Korsmeyer Synthesis of

1391

( f 1-Oplopanone

3

330

Other Bicnrbocycllc Sesquiterpenes

tole gives cyclohexenone 1400 which adds vinylmagnesium bromide stereoselectively trans to the isopropyl group. Equilibration of the epimerizable position affords the more stable all-trans isomer 1401 in which three of the stereocenters of oplopanone are correctly established. After Wittig methylenation and cleavage of the phenoxy group, the more accessible double bond is selectively hydroborated to obtain bromo alcohol 1403. This compound is oxidized to aldehyde 1404 which is cyclized by potassium r-butoxide. In this reaction, a fourth stereocenter is fixed thermodynamically since, as discussed earlier, the natural configuration at this position is the more stable. The final stereocenter is introduced by oxymercuration of the exocyclic methylene group, which apparently occurs by attack of water on the mercurinium ion solely from the equatorial direction. The Taber-Korsmeyer synthesis requires 1 1 steps and produces 1391 in 2.9% overall yield. (4) Picrotoxinin

Picrotoxinin (14061, a constituent of the long-known “picrotoxin”, is one of the more complex sesquiterpenes to have been synthesized. Of its numerous interesting structural features, not the least intriguing and challenging from a synthetic standpoint is the “prow-stern” interaction between the methyl and isopropenyl groups on the cyclohexane boat. The first and only successful synthesis of picrotoxinin was reported by Corey and Pearce in 1979.393 Several syntheses of the sesquilerpene alkaloid dendrobine, which presents similar but less challenging synthetic problems, will be discussed in section 7.

1408

The Corey-Pearce synthesis (Scheme 172) begins with (-)-cawone (1407) which is converted into the dimethylhydrazone and alkylated with

Fused Ring Compounds: 5,6 I . MeZNNH2, CF3COOH

a

2. L D A ,

0

ErJk,

1407

I

*

'on 1410

+

@ Meor

2. HCI,

Me

+

"Me2

"Me2

1408

HMPT

@

2' . separate . ~ :

n

nczcLI,

111

'* OCOP h

1412

4

I . PhCOCl

--OH

C5H5N 2. NaOCl

.

Er& --OCOPh

1410

1417

1420

1491

"OCOPh

1413

kOOH

1419

H20

1409

~ ~ ~ I * ~

1411

Scheme 172.

331

1400

Coery-PearceSynthesis of Picrotoxinin

332

Other Bicarbocyclic Sesqulterpenes

the dimethyl acetal of 3-bromopropanal. The hydrazones 1408 and 1409, obtained as a 3:2 mixture, are hydrolyzed and cyclized to obtain the bicyclic aldols 1410 and 1411 in a 3:2 ratio. The mixture of benzoates is separated by preparative hplc and the more abundant isomer 1412 is used for the remaining steps of the synthesis. Addition of lithium acetylide to ketone 1412 occurs exclusively from one face of the carbonyl, yielding alcohol 1413, probably as a result of steric hindrance by the isopropenyl group on the other face of the carbonyl. At this point the newly created hydroxy is added to the isopropenyl double bond by treatment of 1413 with N-bromosuccinimide. Bromo ether 1414 is formed stereoselectively in quantitative yield. This transformation is the key to the synthesis. The ether bridge not only serves to protect the isopropenyl double bond, but it also locks the cyclohexene ring, which it spans, into one rigid conformation. At a later stage, this rigidity turns out to be quite important. The next steps of the synthesis are concerned with converting the ethynyl group into an acetaldehyde unit, which is protected as a dithioacetal (1415). The benzoate is now hydrolyzed and the resulting secondary alcohol oxidized to a ketone, which is further oxidized to the diosphenol 1416. Mercuric oxide suffices both to hydrolyze the dithioacetal and bring about aldol cyclization to yield the tetracyclic compound 1417. After benzoylation of the secondary alcohol the a-diketone is cleaved to obtain diacid 1418. At this point, the synthesis almost foundered, as numerous straightforward methods for bringing about the required bis lactonization failed due to the extremely hindered nature of the double bond. However, cyclization occurs when diacid 1418 is treated with lead tetraacetate in acetonitrile; dilactone 1419 is obtained in 99% yield. Again, it is worth noting at this point that the bromo ether moiety is essential for the success of this reaction. Without this bridge, the molecule would undoubtedly exist in a conformation (1422) in which lactonization is impossible. The remaining steps proceed without HOOC I

Hofpx 'OCOPh

I422

Fused Ring Compounds: 5,7

333

incident. After elimination of benzoic acid, the double bond is epoxidized and the bromo ether cleaved to obtain picrotoxinin. The synthesis requires only 18 steps, and proceeds in a very creditable 7.8% overall yield. All in all, it is a very professional piece of work.

G. Fused Rings: 5,7

(1) Guaiazulenes; Bulnesol, a -Bulnesene, Guaiol, Dehydrokessane, and Kessanol

In Volume 2 of this series, a notably brief section was devoted to the synthesis of hydroazulenic sesquiterpenes. We noted that general methods for the synthesis of the hydroazukne skeleton were few, and problems of stereochemical control about this flexible nucleus were largely unsolved. During the 10 years that have elapsed since our original review article, the progress in this area has been truly impressive and some 30 papers have appeared detailing the synthesis of one or more natural products. In this section we begin by discussing the synthesis of the guaiazulenes bulnesol (1423), a-bulnesene (14241, guaiol (1425) and the kessane derivatives 1426 and 1427.

H 1423

1424

1426

1425

1427

334

Other Bicarboeyclic Sesquiterpenes

Andersen has reported the synthesis of bulnesol (1423) outlined in Scheme 1 73.394 Homologation of photocitral-A (1428) to 1429 and subsequent conversion to the diene-aldehyde 1430 were carried out in a precedented manner. When 1430 is exposed to perchloric acid and acetic

HO

I . LiAIH, 2. p-TaCI 3. NOCH(COOEt)2

1428

I . NOH COOE’ COOEt

3. C r 0 3

G

1429

1431

C

H

O

1430

I432

OH

1433

1423

Scheme 173.

Andersen’s Synthesis of

1434

( *bBulnesol

anhydride, two cyclization products 1431 and 1432 are obtained in 27% and 56% yields, respectively. After separation, 1432 is converted to the esters 1433, an equilibrium mixture of C-7 stereoisomers. Treatment of this mixture with methyllithium affords bulnesol (1423) and 7-epibulnesol (1434) in a ratio of 7:3. During the period of this review, full experimental details of a bulnesol synthesis previously abstracted395a by us have appeared.395b For completeness we include these references here.

Fused Ring Compounds: J,7

335

Mehta and Singh have prepared a-bulnesene (1424) in a brief synthesis from patchouli alcohol (1435) (Scheme I 74).396 Lead tetraacetate oxidation of 1435 affords a mixture of products from which 1436 may be

I . NaBH,

Pb(OAc),

2. p-TsCI, pyridine

1435

KOAc/HOAc

1437

Scheme 174.

-ch,

1436

la?

Mehta-Singh Synthesis of

1424

( f )-a-Bulnesene

isolated in approximately 20% yield. After reduction and tosylation, solvolytic rearrangement of 1437 affords a-bulnesene (1424) in unspecified yield. Solvolytic rearrangement of cis-fused and trans-fused decalyl tosylates has previously been employed for the synthesis of bulnesol and a-bulnesene.397 Guiaol (1425) has attracted a great deal of interest among synthetic chemists. However, the stereospecific synthesis of this molecule has been elusive. In Marshall’s approach398 (Scheme 175), a series of simple transformations is employed to convert the known octalone 1438 to keto-aldehyde 1441. After an aldol condensation and dehydration, 1442 is converted to its enolate with trityllithium. Addition of this enolate to ethanol containing sodium borohydride followed by mesylation affords 1443, in which the double bond has migrated to the six-membered ring. Note that the stereochemical relationship between the substituents at what will be C-7 and (2-10 in guaiol has been established by the direction of kinetic protonation of the enolate, apparently quite efficiently.

Other Bicarbocyclic Sesquiterpenes

336

I . A c ~ OH+ ,

2. NaBH4,

pyridina

*

EtOH

2. Li/NH3.

HO

1439

EtOH

I440

3. PhCH2CI.

.qHZph

I . 03

Na2C03

2 . Zn/HOAc

EtOH, H20

NaH

I . PhSCLi

CHO

I441

*

2. NaBH4, EtOH

1442

3. MSCl

I . LiAIH,

I . Li/NH3

2 . MeCl

1443

&, I444

Z ! Br

4. CHzN2

I446

1445

MsLi

COOMs

1447

1425

Scheme 175.

1448

Marshall's Synthesis of (i:)-Guaiol

After further transformation to 1444, the stage is set for an ingenious solvolytic rearrangement affording the hydroazulene 1445. The mechanism of this solvolysis presumably involves participation of the 7r system of the double-bond giving the cyclopropyl carbinyl cation 1449 which rearranges to 1450, the more stable tertiary cation. The latter

Fused Ring Compounds: 5,7

337

species suffers nucleophilic attack by solvent leading to 1445, which is converted to ester 1447 as shown in the scheme. Unfortunately, a 1:1 +

(J?

1449

HOL\c_ -H+

1445

1450

mixture of isomers at C-7 is obtained and thus addition of methyllithium gives an equimolar mixture of guaiol (1425) and 7-epiguaiol (1448). Buchanan and Young reported the preparation of guaiol outlined in Scheme 176.399 Condensation of Mannich base 1451 with 2-methylcyclopentanone under thermal conditions affords a 68% yield of 1452, where Michael reaction has occurred at the less substituted carbon of the cyclopentanone. Upon treatment with HCl, 1452 undergoes aldol condensation and lactonization. Lactone 1453, a mixture of stereoisomers, is converted to a single enone 1454 in 35% overall yield upon heating

1451

%3

1452

-

H+ MeOH A

PPA

l0O0C

1454

1453

I . MeLi 2. H z I P d I C

(+ other products)

1425

Scheme 176.

Buchanan-Young Synthesis of (*)-Guaiol

338

Otber Bicnrbocycllc Sesquiterpenes

with polyphosphoric acid. After cleavage of the vinylogous P-diketone 1454 with acidic methanol, ester 1455 is produced as a 1:l mixture of stereoisomers. After exposure of this mixture to methyllithium and hydrogenolysis of the resulting allylic alcohol, a mixture of four stereoisomers is obtained from which guaiol (1425) can be isolated by gas chromatography in 10%yield. The final guaiol synthesis is by Andersen and Uh (Scheme 177).400 This synthesis is founded on an observation made earlier in the Andersen bulnesol synthesis (Scheme 173). When diene aldehyde 1430 is treated with acetic anhydride and perchloric acid, a minor cyclization product 1431 is isolated. This material results from stereospecific hydride migration in a cyclization intermediate such as 1456. It was therefore

I430

reasoned that cyclization of 1457 would afford a guaiol precursor. When the cyclization reaction was carried out, 1460 was in fact a minor product. In a series of transformations identical to those employed in the

bulnesol synthesis (Scheme 173) the cyclization products were converted to a mixture of four products of which guaiol (1425) was a minor component.

6

- dOAC I . t-BuOK

2. LiAlH4

1458

1459

+

HC104

p-TsCI, etc.

~

(see Scheme 173)

$Q

I460

1423 (44%)

Fused Ring Compounds: 5,7

&

339

CHO

1457

(see Scheme 173)

m

-0Ac

1432

I434 (16%)

Scheme 177.

1425 ( 1 1 % )

Andersen’s Synthesis of

1448 (14%) (+1-Guaiol

and Related Compounds

Two syntheses of kessane derivatives have appeared in the literature. Liu and Lee prepared dehydrokessane (1426) by the route shown in Scheme 178.401 Photocycloaddition of 1461 and 1462 followed by acidcatalyzed elimination of acetic acid affords tricyclic enone 1463 in 60% yield. Copper-catalyzed conjugate addition of methylmagnesium bromide occurs stereospecifically. After reductive removal of the ketone, the esters in 1464 are reduced with lithium aluminum hydride. Fragmentation occurs upon treating the intermediate diol with tosyl chloride in pyridine yielding hydroazulene 1466. After introduction of the isopropylo1 group at C-7, as shown in the scheme, the ether bridge is closed by oxymercuration. It is interesting to note that stereoisomer 1467 is the principal product of the sequence shown (57% yield). After reduction of 1468, the resulting alcohol is treated with POCI,. Elimination and double bond migration occur under the reaction conditions leading to dehydrokessane (1426).

340

Other Bicrrbocyclic Sesquiterpenes

0

0

Ac6 1461

I

LtAIH4,

2

p-TsCI, pyrtdine

-*u

3 . Roney Ni

OAc

1462

[

COOMI

- @ ''

~ ~ ~ 6 ~ ~ c o b

2. NaH 3. MeLi

1465

1467

Scheme 178.

OAc

1464

1463

pyrldine

+'

1466

1468

0 I426

Liu's Synthesis of (*)-Dehydrokessane

Andersen has synthesized kessanol (1427) from photocitral-A (1428) as shown in Scheme 179.402 Straightforward transformations afford unsaturated aldehyde 1471. Prins cyclization and silylation give a mixture of products from which 1472 can be isolated from 1428 in approximately 14% overall yield. The double bond in 1472 is protected as its dibromide and the amide is converted to the acetate by nitrosation. After regeneration of the double bond by zinc in acetic acid, 1473 is isolated in 20% yield. The isopropyl side chain is fabricated as shown, and the ether linkage closed by oxymercuration. The correct stereochemistry at C-8 is established by conversion to the ketone 1476 and subsequent stereoselective reduction with an unspecified hindered borane.

Fused Ring Compounds: 5,7

341

I. NCVCOOEt 2. NaBH,, MeOH

dk

1428

I . KOH, MIOH, 2 . Collins

1471

.

4 3.. ZnIHOAc CCI4, N2°4 A

*.

I . SnC14 w

+.Aicl

I470

QAi+

9; HAc

I . Br2

1469

Si+

2.

3. separate

q:+ I

fl;r~Me

, I :

I

I

HAc

1472

si+

AcO

1473

I . MeLi

I

+II

2. 3. PCC MeLi

CHO

*

1474

I . Hg(OAc12

"

2. 3. NaBH4 HOAc, H20b 4 . Jones

I471

Scheme 179.

hind ere d

&-..OH

1476

1427

Andersen's (f)-Kessanol Synthesis

(2) Guaianolides: Dihydroarbiglovin and Estqfiatin

At this juncture it is appropriate to note that in all of the syntheses thus far discussed in this section, problems of stereocontrol have remained largely unsolved. In the syntheses of the guaianolides dihydroarbiglovin (1477) and estafiatin (1478) these stereochemical problems are solved by employing the highly elaborated decalinic sesquiterpene santonin (1479)

342

Other Bicubocyclic Sesquiterpenes

I477

I478

1479

as a starting material. Marx has prepared dihydroarbiglovin (1477) by way of santonin derivative 1480 (Scheme 180),403 the preparation of which we described in Volume 2 of this series.404 Methanolysis of the

1477

Scheme 180.

Marx's Synthesis of Dihydroarbiglovin

tertiary acetate 1480 followed by dehydration with thionyl chloride gives the diene 1481 as the major product. Selective hydrogenation of 1481 is readily accomplished, and is stereospecific. Allylic oxygenation with r-butyl chromate gives dihydroarbiglovin (1477). In CrabbC's synthesis 405 of (-1-estafiatin (1478) an alternative synthesis of 1481 is employed, and this material is converted to the natural product (Scheme 181). The most interesting aspect of this synthesis is the use of a large excess of sodium borohydride in pyridine to effect

Fused Ring Compounds: 5,7

I . KOH,

343

I . NaBH4,

*

pyridine 2 . H2O. NaBH4, pyr i dine

H20 2. HCI 3. S0CI2

b I482

HMPT,

A

H

*

1483

I . LDA

I'

I

m-CPBA

3. H 2 0 ~

1484

-

1488

IS78

Scheme 181.

Crabbe's Synthesis of (-1-Estafiatin

reduction of enone 1483 in the presence of the exocyclic double bond. After 1,4-reduction is complete, water is added to the reaction mixture leading to 1484 as the isolated product. Dehydration of 1484 is accomplished in approximately 25% yield by the agency of hot HMPT,and the synthesis is completed as shown in the scheme. The final epoxidation is stereo- and chemoselective giving (-bestafiatin in 51% yield along with about 10% of the P-epoxide.

344

Other Bicarbocyclic Sesquiterpenes

(3) Guaiazulenes with an additional cyclopropane ring: Cyclocolorenone, 4-Epiglobulol, 4-Epiaromadendrene, and Globulol. A few guaiane derivatives containing a cyclopropane ring are known and four such materials have been synthesized: cyclocolorenone (14861,globulol (1487a),4-epiglobulol (1487b)and 4-epiaromadendrene (1488).

1486

14870

4a methyl

1487b 4b

1488

methyl

Earlier approaches to cyclocolorenone (1486) confronted the problem of the C-1 epimer of 1486 being thermodynamically favored over the natural isomer.406 Thus Buchi’s attempted synthesis of 1486 failed when it was found that the reaction conditions required to convert 1489 to 1486 resulted in isomerization to the la isomer 1490. Caine and Ingwal-

I489

1490

son have devised an elegant photochemical solution to this problem shown in Scheme 182.407 Maalione (1491) is converted by formylation, dehydrogenation and Jones oxidation to the dienone-acid 1492. Irradiation of this cross-conjugated cyclohexadienone causes rearrangement and concomitant decarboxylation to afford 1493 directly. Hydrogenation of this product gives 1486 in good yield.

Fused Ring Compounds: 5,7

mH . H

I . HCOOEt

H

NaOMe

3. Jones

o

z

c

~

''H

0 '

hu

H

1491

345

H'

1492

I493

1

H,, Pd/C O=QH

I486

Scheme 182.

Caine-Ingwalson Synthesis of (-)-Cyclocolorenone

Marshall has prepared globulol (1487a) by solvolytic cyclization of a cyclodecandienol (Scheme 183).408 The starting material has previously been prepared by Yoshikoshi in connection with his bulnesol synthesis,409 and is available (eight steps) in good overall yield from the

0

&

I . MsCI, pyridine

8 steps, 0

1494

* NoOMe, MeOH

HO

1495 p-N02C,,H,COCI

I . BH3 2.

2. Li(l-Bu0)3AIH

HO I

1497

1496 H201

*

ArCO

dioxane I

I498

I SO0

I499

Scheme 183.

Marshall's (+)-Globulol Synthesis

14810

346

Other Bicarbocyclic Sesquiterpenes

Wieland-Miescher ketone (1494). Mesylate 1496 is hydroborated and subjected to the action of sodium methoxide which causes fragmentation via the intermediate boronate ester. After conversion to the p-nitrobenzoate ester, 1498 is solvolyzed in buffered aqueous dioxane leading to the hydroazulene 1499. The trans-fused ring system and the presence of unsaturation in the seven-membered ring serve to fix the conformation of this system (149911) such that reaction with dibromocarbene occurs stereoselectively from the convex face of the molecule yielding 1500. Methylation of 1500 affords (*)-globulol (1487a).

CH3

I4990

Caine and Gupton have exploited the well-known photochemical transformation of cross-conjugated cyclohexadienones to prepare 4-epiglobulol (148713) and 4-epiaromadendrene (1488, Scheme 184h4'' Compound 1501 was prepared as an intermediate in the Caine a-cyperone synthesis (see Scheme 5 ) . Treatment of 1501 with base induces cyclization and the product is dehydrogenated with DDQ. Dienone 1502 undergoes the expected santonin-type rearrangement upon photolysis in aqueous acetic acid. Dissolving metal reduction affords material with the P-configured C-4 methyl group. Deoxygenation affords 4-epiglobulol (14878). Dehydration with thionyl chloride gives mostly the desired 4-epiaromadendrene (1488) in addition to a 10-chloro derivative which is dehydrohalogenated with sodium acetate in acetic acid, furnishing a further quantity of 1488.

Fused Ring Compounds: 5.7

1501

2. I . Wolff L i / N H-Kirhner 3

H;

**ti

1487b

Scheme 184.

1503

1502

$Q ::E;,-

347

Q (M HI' 1488

Caine-Gupton Synthesis of (-)-4-Epiglobulol

(4) Pseudoguaianolides: The Ambrosanolide Family; Deoxydamsin,

Damsin, Ambrosin, Psilostachyn, Slramonin B, Neo-ambrosin, Parthenin, Hymenin, Hysterin, Damsinic Acid, and CoMertin. The tremendous success in the total synthesis of pseudoguaianolides is remarkable. Prior to 1975 only a single synthesis of a pseudoguianolide structure had been achieved, Hendrickson's preparation of the nonnaturally-occurring compound 1504. In the ensuing five year period some 21

I504

papers appeared detailing the syntheses of more than a dozen of these natural products. In part, this sudden flurry of interest was prompted by the observation that many pseudoguaianolides, in common with other sesquiterpene a-methylene lactones, possess antineoplastic and cytotoxic activity.

348

Other Bicarbocyclic Sesquiterpenes

The pseudoguaianolides have been classified into two broad subgroups depending on the configuration of the secondary (C-10) methyl group. Those with the 10-a configuration are named helenanolides while those with the 10-p configuration are called ambrosanolides. We consider the total synthesis of members of the latter class first. The ambrosanolides that have yielded to syntheses are the non-naturally occurring compound deoxydamsin (1505). damsin (15061, neoambrosin (1507), parthenin (1508), hymenin (1509), ambrosin (1510), stramonin B (15111, psilostachyin C (15121, damsinic acid (15131, confertin (1514) and hysterin (1515).

I 505

I506

I so9

1510

QCOOH 151 3

I sor

1508

hAC I512

1511

,OH

1514

1515

Marshall's synthesis of deoxydamsin (Scheme 1851, while not leading to a natural product, is an important contribution to pseudoguaianolide synthesis since the methodology developed here has found wide application in other ~ y n t h e s e s . ~ The " well-known hydroxy enone 1516 is con-

Fused Ring Compounds: 5,7

349

1518

1519

OMs I . HzO;!, -OH 2. N H ~ N H ~ , A )

0

OH

3. MsCl

I517

1516

-

I . Jones 2. AcpO,

I m-CPBA 2. L i A l H 4

NoOAc, A

dH I520

I522

1521

1523

I524 I . NOH,

HCOOEt

2. NoBH, 3 . p-TsCI,

4. I525

Scheme 185.

pyridine pyridine,A

I505

Marshall’s Synthesis of (*)-rl-Deoxydarnsin

verted in an unexceptional fashion to the cyclodecadienol 1519. Cyclization (to 1520) occurs readily in acetic acid, a reaction which finds precedent in Marshall’s earlier work leading to globulol (see Scheme 183). Epoxidation of 1520 occurs stereoselectively, trans to the angular methyl group. Reduction of the epoxide is regiospecific, the angular methyl group directing the hydride reagent to the less hindered (C-7) position.

350

Other Bicarbocyclic Sesquiterpenes

After oxidation of the secondary alcohol and dehydration, a mixture of alkenes 1522 is obtained. The two-carbon fragment required for the lactone is added by alkylation with methyl bromoacetate. Subsequent hydrogenation affords 1523 with excellent stereoselectivity. Again the angular methyl group appears to exert a dominant effect on the stereochemical outcome of reactions on the pseudoguiane framework. After saponification of 1523 the product acid is converted, via the enol-lactone, to the butenolide 1524. The stereochemistry at C-6 may arise from the preferred direction of kinetic protonation of the intermediate enollactone or may reflect a thermodynamic preference for the p-configuration. Hydrogenation of 1524 is stereoselective, affording 1525 in 90% yield. The lactone is methylenated by the method of Minato and Horibe providing deoxydamsin (1505). Unfortunately the strategy employed in the Marshall deoxydamsin synthesis is not readily adapted to the preparation of naturally occurring pseudoguaianolides, all of which are oxygenated at C-4. Kretchmer’s synthesis of damsin (Scheme 186) is the first of the successful syntheses of naturally occurring p s e u d o g ~ a i a n o l i d e s Octalone .~~~ 1530, prepared as shown in the scheme, affords triketone 1531 upon ozonolysis. This substance undergoes sequential aldolization, dehydration, and methylation upon exposure to mild base in the presence of methyl iodide. The major product of this transformation is 1532 (36% yield from 1526) which is separated from its C-7 epimer (18% yield) by column chromatography and crystallzation. Reduction of 1532 affords a triol. Selective oxidation of the primary alcohol and saturation of the double bond provides hydroxy lactone 1534 in moderate yield. The conversion of 1534 to damsin (1506) is uneventful. Vandewalle’s synthesis of damsin (Scheme 187) provides an interesting construction of the hydroazulene ring system.413 Photocycloaddition of 2-methylcyclopentenone (1536) with 1537 affords the tricyclic ketone 1538. After reduction of the carbonyl and desilylation, the intermediate vicinal diol is oxidatively cleaved to obtain 1539. Protection of the secondary alcohol and base-catalyzed epimerization affords a 6:l mixture of the trans-fused hydroazulene 1540 and its cis diastereomer. The mixture is separated and the major isomer is carried on, in a manner remin-

Fused Ring Compounds: 5,7

351

I . MeLi COOH

Me

2 . HZl Pd'C*

I . Zn, BrCH2COOMe

Me0

I527

I526

2. H p . Pd/C

COOH

Me0

1528

I . Li/NH3

b

2 . (C02H)p. 3. CH2N2

Me0

3. KOH

'COOH

I529

'COOMe 1530

Me1 2. separate

aoH . .

0 @-COOMe

COOMe

I532

1531

2. I * He, p t * O2 Pd/C$ $ ( H

2. I . (CHz0H)2 Jones

@

HO

1533

0

I . NaH, HCOOEt 2. N o B H ~ 3. H', HzO 4. p-TsCI, pyridine 5. pyridine,A

Scheme 186.

I534

1535

I506

Kretchmer-Thompson Synthesis of (*)-Damsin

352

Other Bicarbocyclic Sesquiterpenes

1536

0- @

P 23. I ., NaOMe separate H+

I539

1538

I537

I . LDA,

0rCH2COOMe

Ph3PCH2,

2. 3. KOH Ac20, NaOAc*

THp@

T H P ~

1540

1541 I . Jones 2 . ( C H ~ O H.H+* )~

THP(

I542

1543

I544

I . NoH, HCOOEt 2. NOBH4

3. H + , H 2 0 4. p - T s C I ,

pyridine

5. pyridine,A

Scheme 187.

I506

Vandewalle’s Synthesis of (*)-Damsin

iscent of Marshall’s deoxydamsin synthesis, to damsin (1506). Of special note in this sequence is the cleavage of the tetrahydropyranyl ether of 1542 under hydrogenation conditions. In an extension of this work, Vandewalle has also reported the conversion of 1543 to neoambrosin (15071, parthenin (1508) and hymenin

Fused Ring Compounds: 5,7

353

(1509) as shown in Scheme 188.4’4 Ketalization of the enone 1545 (derived as shown from 1543) occurs with concomitant migration of the double bond. After methylenation 1546 is obtained. The latter may be

9

I . Jones 2 . PhSeCl

Hd

3. 03

4. pyridine,

1543

I soa

I . H+,

(HOCH2)2

P

2 . NaH, HCOOEt

A

P

3. NaBH4

4 . p-TsCI, pyridinr.

IS45

A I546

I so9

1507

Scheme 188. Vandewalie’s Synthesis of (*)-Neoambrosin, (*:)-Parthenin and (*)-Hymenin

hydrolyzed directly to yield neoambrosin (1507). Epoxidation of 1546 affords a mixture of two ketal epoxides which undergo hydrolysis and elimination to a 7:3 mixture of y-hydroxy enones parthenin (15081, and hymenin (1509) upon attempted chromatography on silica. Grieco has contributed impressively in the pseudoguaianolide area. His first report deals with the total synthesis of damsin (1506, Scheme 189).415 Technology developed by Grieco and others416 in connection with prostaglandin synthesis is used to convert norbornadiene to the functionalized bicyclo [2.2.1 lheptane 1547. This material undergoes

354

Other B i c u k y c i i c Sesquiterpenes

db

I . DBU

5 rlepr

2. L i A I H 4 ___)

I547

I.

$,

1548

2. L D A , Me1

I. Hg,PtOz 2. L i A I H 4

I , HZOpt -OH 2. C H

H+

____)

THp%

~

N

~

I . PhCH28r,

CN

OH

NOH 2. D l B A L 3. NaBH4

OTHP

m

~

3. p - T G I , pyridine 4 . NaCN

1551

I550

I549

+

3. HOAC,HZO

2 . MeLi

(50Xyleldi

&HP PhCH2

I552

I . MeOH,

TsOH 2. p-TsCI, pyridine 3. Jones

-

I553

I . LDA, Br02. 0 3 3. Jones

I. NaI 2. LHMDS PhCHp

I555

I554

2. Hz,?d/C

2. MsCl 3. DBU 4 . Jones

PhCHZ

1556

I557

Scheme 189.

I506

Grieco’s (*)-Damsin Synthesis

remarkably stereoselective alkylation upon treatment with lithium diisopropylamide and methyl iodide. Note that the ring fusion stereochemistry of the eventual product is established in this reaction. The

Fused Ring Compounds: 5,7

355

alkylation product is converted to the methyl ketone (1548) by treatment with methyllithium. Subsequent transformations lead to 1549 which is protected as the tetrahydropyranyl ether. The crucial C-10 methyl group is now introduced stereospecifically by akylation of the derived enolate from the less hindered endo face. Baeyer-Villiger oxidation of 1550 followed by esterification leads to 1551 in which three of the five asymmetric centers of damsin are correctly established. The keto-tosylate 1554, derived from 1551 in ten straightforward steps, is converted to the iodide and the seven-membered ring is closed by alkylation. The conversion of 1555 to damsin relies on the same strategy employed by Marshall in his synthesis of deoxydamsin. Grieco has also converted damsin into several other natural products, thus completing total syntheses of these compounds. Ambrosin (1510) is prepared by acid catalyzed selenylation of damsin followed by selenoxide elimination. Stramonin B (1511)417 is prepared by conversion of damsin to the corresponding epoxide in two steps followed by introduction of unsaturation to the five-membered ring as outlined above. Finally, modified Baeyer-Villiger oxidation of damsin leads to psilostachyin (1512).415 I . PhSeCl EtOAc

/-

2. NaI04

I . 0s04

2. p-TsCI 3. PhSeCl 4. N ~ I O ~

\\

PhSeOOH, "202

356

Other Bicirbocyclic Sesquiterpenes

Schlessinger has provided an interesting and novel solution to the problem of preparing a functionalized hydroazulene nucleus, exemplified here by his total synthesis of (*)-damsin (Scheme 190).4’8 The known enone 1558 is converted by Barton’s nitrone version of the Beckman rearrangement to the enamide 1559. Treatment of 1558 with dimethyl lithiomethylphosphonate affords an intermediate aldimino-pketophosphonate anion, which is hydrolyzed to the aldehyde 1560. The latter compound readily cyclizes to the enone 1561. The stereochemistry of the ring fusion is determined in the conversion of 1559 to 1560. The 0

II

H

I . LiCHQ(OMd2

I . MeNOH) 2 . P-TsCI

I-Bu

2 . NoOAc, H 2 0

I-BUO

I686

LHMOS, Ma3SiCI

2 . Pd(OAc)2

. @

I . H2O2,-0H

2. NZH*

I-Bu

0

I-BUO $>(OMe12

I560

1589

1501

I.

.

.

I504

I563

1562 I . Hp. Rh/AI#3

I565 I. HC104,

COO(1-Bu)

2. Ac20, NoOAc,A

3 . Me3SiCI

TMS

1567

IS66 I . StllGS

2 . CH20, Mc2NH 3 . Jonir

I506

Scheme 190.

Schlessinger’s ( f ) -Damsin Synthesis

Fused Ring Compounds: 5,7

357

intermediate in that conversion (1569) undergoes kinetic protonation trans to the angular methyl group. This is presumably an intramolecular process, the proton being delivered from the same side of the molecule that the P-ketophosphonate group resides on. Addition of lithium

I569

dimethylcuprate and reintroduction of a 9,lO-double bond affords 1562. Hydrogenation of this enone is highly stereoselective affording the compound with the 10-p configured methyl group, 1563. Deprotonation of 1563 occurs regioselectively, allowing the preparation of 1564. This compound is converted by Wharton’s method to 1565 and then, in a precedented fashion, to damsin (1506). The use of the Stiles’ reagentMannich sequence was found by Schlessinger to be a superior method for the introduction of the a-methylene group. Two syntheses of the simplest naturally occurring pseudoguaiane, damrelies on sinic acid (1513), have been reported. Lansbury’s a cation-acetylene cyclization to form the seven-membered ring (Scheme 191). The required substrate for this cyclization, 1572, is prepared from 2-methylcyclopentane-1,3-dione (1570) as shown in the scheme. Formic acid induces cyclization to the hydroazulene, 1573. This cyclization is noteworthy, in that the seven-membered ring ketone is formed to the exclusion of the possible six-membered ring aldehyde 1578. In most

I578

ring forming reactions, enthalpic and entropic considerations favor formation of a cyclohexane ring over a cycloheptane. However, formation of 1579 proceeds via a primary vinyl cation and thus enthalpy strongly

Other Bicrrbocyclie Sesqulterpenes

358

I571

I570

1572

T M S y COOMe

H2-Pd,

m COOMe I576

C - 7 epimer (20%)

,-Bu&

COOMe

t-Bu

+

I575

I574

I573

I577

I . 8088, C H 2 0 2. MaCI. pyr i d ine 3 . DBU

I . H+ COOMe

m

2. oxidation 3. roponification

1513

Scheme 191.

Lansbury’s

( *1-Damsinic Acid

Synthesis

favors formation of 1573 via the secondary vinyl cation. The seven.. membered ring carbonyl of 1573 is selectively protected as the dithioketal and the cyclopentanone is reduced. After thioketal hydrolysis and protection of the secondary alcohol as the r-butyl ether, hydrogenation of the double bond affords a 6:l mixture of two stereoisomers. The

Fused Ring Compounds: 5,7

359

major product, 1575, was isolated by crystallization. The lithium derivative of methyl trimethylsilyl acetate converts 1575 to 1576 in nearly quantitative yield. Hydrogenation of 1576 affords 1577 in 7O-8O0/o yield along with 10-20% of the C-7 epimer. Unfortunately, the two isomers proved to be inseparable. The conversion of 1577 to damsinic acid (1513) is straightforward. Wender's imaginative approach to damsinic acid (1513) employs a divinylcyclopropane rearrangement to fuse a seven-membered ring onto an existing five-membered ring (Scheme 192).420 Reaction of enol ether 1579 with the lithium reagent 1580 (a mixture of stereoisomers) affords 1581 as a mixture of isomers. Upon thermolysis, isomer 1581c, in which the vinyl groups are cis on the cyclopropane ring, readily rearranges to the desired hydroazulene 1582. The trans isomer l58lt does

I579

1581 t

1580

I582

I583

I585

Scheme 192.

1581c

1584

Ill3

Wender's (*)-Damsinic Acid Synthesis

360

Otber Bicarbocyclic Sesquiterpenes

not give the desired rearrangement, a homo-[1,51hydrogen migration leading to 1586 being preferred. The preparation of stereochemically

I586

homogeneous 1581c is possible but is somewhat laborious. Wender makes use of the observation that the stereoisomers of 1581 may be equilibrated through photolysis. Since the rearrangement of 1581~to the desired hydroazulene occurs at temperatures where 1581t is stable to thermolysis, simultaneous irradiation and heating converts both stereoisomers cleanly to 1582. The ketone is protected as its ethylene ketal and the diene is oxidized to a mixture of the linear dienone 1583, and the cross-conjugated dienone oxygenated at C-9. The major isomer, 1583, isolated in 60% overall yield from 1582, is converted to 1584 with 90% stereoselectivity. Conversion of the latter to damsinic acid (1513) follows precedented lines. Several syntheses of the pseudoguaianolide confertin (1514) have been reported. The first is by Marshall42' (Scheme 193). Enone 1587 is alkylated and reduced to 1588. The allylic alcohol moiety allows regioand stereoselective epoxidation of the tetra-substituted olefin. Reductive elimination of the derived mesylate provides alcohol 1589. SimmonsSmith cyclopropanation is also regio- and stereoselective, the directive effect of the allylic hydroxyl group again being exploited. The key intermediate 1590 is then prepared by oxidative cleavage of the ally1 group. Solvolysis of 1590 in aqueous perchloric acid leads to the hydroazulene 1591. The stereochemical course of this reaction is of i .rest. Compound 1591 may exist in either of two conformations 1SYla or 1591b. The latter is presumably favored in this equilibrium since the '

fqh \

COOH

7 \

"p

t -0uo

t-Bu

15910

I591 b

Fused Ring Compounds: 5,7

&

?*

I . LDA, ! 2. LiAIH4

..-

I-BuO

1587

I . m-CPBA, 2. MsCl

I588 I . SimmonrSmith

t-Bu

I-Bu

1589

t-Bu

1590

I592

$ +

OH

t-BUO

1591

I593

2. I . CFjCOOH MnOp 3. OH-, i-PrOH

1594

H &

CrOS py r id ino

1595

1514

Scheme 193.

Marshall’s Synthesis of (*)-Confertin

361

Other Bicarbocyclic Sesquiterpenes

362

acetic acid side chain is equatorial. Moreover, only in conformation 1591b can the carboxyl group be involved in nucleophilic participation in a concerted process in which the cyclopropane ring is cleaved. Thus rearrangement via 1591b is energetically favored. In any event, the synchronous nature of this rearrangement ensures the stereospecific introduction of the C-8 oxygen in the p-configuration as is required for the synthesis of confertin. The correct stereochemistry at C-7 is established

I . I-BuLi

2 . +=-cu 2. Collins 4 . 6rCH2COOMe

I597

I596

@d

COOMc

I599

OOMe

I . TsOH. MeOH

2.

DCC,

CF,COOH, OMSO

3. CH30S02F I600

I602

1601

'%

1514

1603

Scheme 194.

Semmelhack's

(*)-Confertin

Synthesis

1598

Fused Wing Compounds: 5,7

363

by conversion of 1591b to the butenolide 1592, which is hydrogenated to obtain 1593 in 70% yield. The methylenation of 1593 is carried out by initial carbonation and subsequent reduction of the p-keto ester anion affording the unsaturated diol 1594. The allylic alcohol is oxidized and the t-butyl ether cleaved with trifluoroacetic acid. Mild hydrolysis serves to cleave the intermediate trifluoroacetate ester providing 1595. Collins oxidation provides confertin (1514). Semmelhack's unique approach to hydroazulene synthesis is shown in Scheme 194.422 The preparation of 1598 from 1596 proceeds uneventfully. Wadswoth-Emmons reaction of 1598 with 1599 (generated in situ from trimethyl 2-phosphonoacrylate and sodium isopropylmercaptide) provides 1600 as a 1:4 mixture of E and Z stereoisomers. The methoxymethyl group is cleaved and Moffatt oxidation provides the corresponding aldehyde. The sulfonium salt 1601 is prepared by the agency of methyl fluorosulfonate. The salt is converted by Ni" to a mixture of a-methylene lactones. The major isomer 1602 is formed in approximately 25% yield from 1600. Unfortunately, selective hydrogenation of 1602 to confertin (1514) is not possible, necessitating the sequence shown. Confertin (1514) is produced in low yield by this method. Schlessinger has prepared confertin by the route shown in Scheme 195.4'8 Alkylation of 1563 (Scheme 190) is regiospecific, and the resulting ester is converted to the enollactone 1604. Hydrogenation affords

@,

I . LHMDS, ICH$OOBu-t 2. A c ~ O HC104 ,

t-0u

&

AcO 1563

I605

Scheme 195.

I . H p , RhIA1203 2. KOH 3. Me3SiCl

1604

1514

Schlessinger's Synthesis of

( f )-Confertin

Other Bicarbocyclic Sesquiterpenes

364

the desired lactone 1605 in 63% yield overall. Conversion of the latter to confertin (1514) is straightforward. Heathcock and co-workers have also prepared the key intermediate 1604 and provided an alternate synthesis of confertin from it (Scheme 196).423The well-known enone 1606 is converted to the tosylate 1607 in a simple sequence. The latter readily rearranges to 1608 (after reacetylation of the alcohol). Direct conversion of 1608 to the key enone 1609 is accomplished by the action of trimethylaluminum and methyllithium. The mechanism for this transformation is interesting. Apparently 1612 is initially formed by P-elimination. Trimethylaluminum facilitates this

&

1610

1609

I. LDA CH2&CH3)21-

I . H2-Rh/AI2O3

2 . -OH

*

Ac

H 1611

1604

2. M e 1 3. NaHC03, H20, EtOAc 4 . Jotire

I514

Scheme 196.

Heathcock-Graham-DelMar Synthesis of (*)-Confertin

b

Fused Ring Compounds: 5,7

365

process by coordinating with the ketal oxygens. Subsequent addition of methyllithium provides 1613 which hydrolyzes directly to 1609 at pH 9 in 63% yield. A variant of the Marshall-Schlessinger strategy affords the

1612

1613

lactone 1611 which is converted to confertin (1514) by Poulter’s methylenation sequence and Jones’ oxidation. Wender’s confertin synthesis (Scheme 197) proceeds from the intermediate 1583, the preparation of which we have discussed previously in connection with the damsinic acid synthesis (see Scheme 192). Epoxidation and modified Wittig olefination afford 1614 which is converted by

@,

I . H202, NaOH

I . H30’.

2. (Et0~2P061COOEt

2. NaOH

I583

@

-COOEt

1614

___)

2. Me1

Pd/C

1615

3. NaHCO3 1617

1616

1514

Scheme 197.

Wender’s (*)-Confertin Synthesis

366

Otber Bicarbatcyclic Sesquiterpenes

treatment with acid to a hydroxy lactone and thence by base treatment to the enol lactone 1615. Hydrogenation of this intermediate proceeds with a high degree of stereoselectivity, lactone 1616 comprising greater than 80% of the product mixture. Thus the angular methyl group again proves valuable in dictating asymmetry in the generation of new stereocenters on the pseudoguaiane framework. Poulter’s olefination sequence is employed to prepare the a-methylene lactone 1617 which affords confertin (1514) upon hydrolysis of the ketal. The preparation of the C-10 hydroxymethyl pseudoguaianolide, hysterin (1519, has been reported by Vandewalle and co-workers (Scheme 198).424The key intermediate 1540 (see Scheme 187) is converted to enol ether 1618 by Wittig reaction. This material is elaborated by precedented means to the butenolide 1619. Reduction of 1619 is only moderately stereoselective and, surprisingly, affords the 1Oa-isomer 1620 as the major component of a 7:3 mixture (compare the reduction of 1542, Scheme 187). It is also noteworthy that the tetrahydropyranyl ether is hydrogenolytically cleaved during the course of this reduction. The mixture of 1620 and 1621 is carried on in the manner shown to a mixture of 10-epihysterin (1625) and hysterin (1515). Heathcock and co-workers have recently developed a route to the ambrosanolides which proceeds via cis-fused hydroazulene intermediates (Scheme 199).425 Cyclohexenone is converted into 2,4-dimethylcycloheptenone (1628) by a procedure developed by Ito. When enone 1626 is subjected to copper-catalyzed conjugate addition of Grignard reagent 1627, trans adduct 1628 and its cis diastereomer are obtained in a ratio of 85:15 (both are produced as 1:l epimeric mixtures at C-2). Ozonolysis of 1628 gives a keto aldehyde which cyclizes to an aldol upon treatment with HCI in acetic acid. Under the conditions of the cyclization the aldol is acetylated to give a keto acetate. The acetyl group is removed and the resulting aldol (1629) is protected as the tetrahydropyranyl ether. At this point, the product is purified chromatographically to remove the small amount of product resulting from the 3,4-cis isomer of 1628. Alkylation of ketone 1630 occurs stereospecifically, affording ketone 1631. Base-catalyzed epimerization of 1631 gives the diastereomeric ketone 1635 (cf. compound 1467, Scheme

Fused Ring Compounds: J,7

Ph$CHOMe

BrCHZCOOMe 2. KOH, MeOH 3. AcZO, NoOAC,

b

THPd

1618

I540

1619

,OMa

&

+

Pt/C

NaBHq,

I . Jones

Hd

I620

1621

I . L=N:ItA,

H $

2. K2C.03, MIOH

2. Me1 3. NoHCOS

no

sa$

Scheme 198.

b

@

1624

PH

Ac

I622 I . Ace0

____)

1623

367

t

Ac

Vandewalle’s Synthesis of (*)-Hysterin

368

Other Bicarbwyclic Sesquiterpenes

I626 1627 2 . HCI,HOAc,

C u I , BF3

p-TSOH

3. MeO-,MeOH

1628

1630

1629

1631 I . Me3SiOTf, Et3N, benzene * 2. N B S , T H F

I . 03,MeOH

3. CoC03. DMAC

THP

1632

1634

Scheme 199.

1633

1545

Heathcock-Germroth-Tice Synthesis of

(*)-Parthenin

Fused Ring Compounds: S,?

I63 I

369

1635

178). Reduction of ketone 1631 with diisobutylaluminum hydride in ether at -78°C gives predominantly (>95%) one alcohol, 1632. Ozonolysis of the prenyl group in methanol gives a cyclic acetal, which is deprotected and oxidized to obtain keto lactone 1633. A double bond is introduced into the five-membered ring and the resulting cyclopentenone is ketalized to obtain the deconjugated ketal 1545, which had been previously converted into (*)-parthenin and (fj-hymenin (see Scheme 188). (5) Pseudoguaianolides: The Helenanolide Family; Helenalin, Mexicanin, Linfolin, Bigelovin, Carpesiolin, Aromaticin, and Aromatin.

Notably fewer syntheses of pseudoguaianolides in which the C-10 methyl group is of the a-configuration- the so-called helenanolides-have been reported. The seven members of this family that have thus far yielded to total synthesis are helenalin (16361, bigelovin (1637), mexicanin I (16381, linifolin A (16391, carpesiolin (1640), aromaticin (16411, and aromatin (1642).

I836

1637

1640

1641

1638: R = H 1639: R = A c

1642

Other Bicarbaryclic Sesquiterpenes

370

Grieco's synthesis of helenalin (1636) is shown in Scheme 200.426a The key cyclopentenol 1551 (prepared in 14 steps from norbornadiene, Scheme 189) is converted to the lactone 1643 in four steps. During the lactonization reaction, the methyl group undergoes epimerization to the I . NoH.

OH

.

I . DIBAL

PhCHpBr 2 . TIOH, MIOH 3. KOH, EtOH 4 . p-TsCI, DBU

+OM8

-.
3. Collins

1551

POT

1643

I. HCI, H 2 0 , THF

2. KOH, MeOH

3. 0 . H '

CACHE0

PhCH2

1644

1645

I . I-BuOOH, -OH b 2. NaBH,. EtOH

PhCHZO

2. I . NaBH4 MsCl 3 . MeOH, TsOH) 4 . Jones 5. DBU

I. LiCH2COOLi PhCH20

1646

'6

2. Li/NHS

3. H+

.

1647

I. 8 . H '

2. L D A , CH20 3 . MsCl 4 . DBU 1648

I . AcOH H 0 A

2. MnO,

THP

1649

rn OH

1636

Scheme 200.

Grieco's (rt )-Helenalin Synthesis

Fused Ring Compounds: 5,7

371

more stable equatorial position. The sequence of reduction, Wittig olefination and oxidation furnishes 1644. The enol ether is hydrolyzed, the resulting keto aldehyde is cyclized by aldol condensation and the aldol protected to obtain 1645. Enone 1646 is prepared from 1645 in the straightforward five-step sequence shown. Epoxidation of 1646 is stereoselective as is sodium borohydride reduction to 1647. Epoxide 1647 is opened by treatment with dilithio acetate, the benzyl group is removed by lithium/ammonia and the resulting product acidified to obtain the cis lactone 1648. After protection of the alcohols, the a-methylene unit is introduced in a typical sequence furnishing 1649. Deprotection and oxidation of the allylic alcohol provide (f)-helenalin (1636) in 1.5% overall yield and 40 steps from norbornadiene. The remarkable achievement of an average yield of 90% per step is a tribute to the tenacity and skill of the four chemists who carried out this work. In an extension of the foregoing methodology it was discovered that reduction of the enone 1646 prior to epoxidation allows the preparation of hydroxy epoxide 1650, a precursor to bigelovin (Scheme 201).426b H!

a

I . LiAIH,

2 . m-CPBA

PHCH2P

I . LiCH2COOLi

m

2. Li/NH3 3. HCI

O

1646

1650

H!

THP

~THP

4. Q,H+

H !

I . LDA,CH20,

I. M I I O ~ ~

2. MsCl

2. A c p

3. DBU

4. HOAC.Hz0

OH

1651

dH 1652

H

!

1637

Scheme 201.

Grieco's (*)-Bigelovin Synthesis

c

372

Other Bicarbocyclic Sesquiterpenes

The conversion of 1650 to bigelovin (1637) follows an established line. One interesting observation in that sequence is the regiospecificity of formation of the lactone ring which occurs exclusively to the C-6 hydroxy group. In Scheme 202 the further conversion of 1650 to mexicanin I (1638) and subsequently to linifolin A (1639) is shown.427Reduction of 1653 is

p- OH

I . NoBH4

I . LiCHZCOOLi

23.. H Jonas +

2. Q.H+) PhCH2

PhCH2

1650

1653 I . Q,H+

2. L i / N H S 3. DCC PhCH2

1654

-

2. LDA.CHzO 3. MsCl, pyridine 4. DBU

I655

1656

1638

Scheme 202.

1639

Grieco’s Synthesis of (*t)-Mexicanin-I and (*)-Linifolin A.

stereoselective, hydride delivery occurring trans to the angular methyl group. The C-6 alcohol is protected at this point since it was discovered that the acid 1657 cyclizes exclusively to the cis lactone 1658:

1657

1658

Fused Ring Compounds: 5,7

373

The conversion of 1654 to mexicanin I (1630) follows a now familiar line. Linifolin A (1639) is readily prepared from 1638 by acetylation. Schlessinger’s synthesis418 of helenalin (Scheme 203) is much more direct than Grieco’s approach. The enone 1561, prepared as shown in Scheme 190, undergoes copper-catalyzed conjugate addition of methylmagnesium bromide in a highly stereoselective fashion yielding 1659. A straightforward sequence converts 1659 to 1660. Deprotonation of 1660

-

0

2. I . (HOCH2)2b HCI,MeOH

M r;e:hl

I-Bu

.

1561 I. NOH, PhSSPh 2. m-CPBA 3. A

3. PCC

I-Bu

0

I659

I . H+,H20

I . LHMDS Me3SiCl

2. H+,

2. Pd(OAc)2)

Q*

4. D l B A L

1660

I*

@! *

H

2. T I B A

THP

THP

THP

1661

I . LiCH2COOLi, 2. H30+

\d 1662

I. 6H

1648

Scheme 203.

Stiles

2. CH20. Me2NH 3. Mn02

1636

Schlessinger’s Synthesis of ( f bHelenalin

374

Other Bicarbocyclic Sesquiterpenes

with lithium hexamethyldisiiazide is regiospecific, allowing formation of the enone 1661. Conversion of the latter to helenalin (1636) follows a precedented line ( vide supra). Heathcock and co-workers also prepared the Schlessinger ketone 1659 by dissolving metal reduction of the key enone 1609 (Scheme 196) used in their confertin synthesis.423 Vandewalle4** has found that hydrogenation of 1663 (prepared from 1541, Scheme 187) over palladium on charcoal takes a different stereochemical course than the reduction of 1542, affording the 10-a compound 1664 in approximately 66% yield (Scheme.204). The 10-p com-

H+

I . He. PdIC

EtOH

2. srporotr

1541

I664

1663

1665

1666

--OH

I DIBAL 2. m-CPBA

I . TsNHNH2 2. MeLi

1667

I . LiCH2COOLI:,

@-

2 . CH2N2 COOMe I668

I . KOH,

2 . MIOH, Ac~O

&

1669

I. LDA.

+

H2CN(CH3)zI;

2. 3. M N oe H 1 CO~ 4 . H+.H20

1670

Scheme 204.

@ bH

1640

Vandewalle's Synthesis of

(~ bcarpesiolin t

Fused Ring Compounds: 5,7

375

pound is formed in 10-20% yield. No unambiguous reason for this behavior has been provided. The major isomer may be isolated by chromatography and is converted by Bamford-Stevens reaction to unsaturated alcohol 1665. The alcohol is oxidized and the resulting ketone protected. Allylic oxygenation gives enone 1667, which Vandewalle independently discovered may be converted to the a-epoxy-a-hydroxy system 1668 (compare Grieco’s bigelovin synthesis, Scheme 201 ). The preparation of carpesiolin (1640) from 1668 finds ample precedent in other schemes of this section. Lansbury’s synthesis of aromaticin (Scheme 205)429 proceeds from ketone 1575 which had been prepared earlier in connection with the synthesis of damsinic acid (see Scheme 191). Sulfinylation of 1575 under equilibrating conditions affords 1671. Pummerer rearrangement gives the unsaturated sulfide 1672. Equilibration of the methyl group is now possible using diazabicyclononane as base catalyst. (Attempts to carry out the equilibration without the presence of the thiophenyl substituent fails since the @,A enone 1679 is thermodynamically favored in that system). Having served its purpose, the vinyl sulfide moiety is removed,

1679

affording 1674. Reaction of 1674 with the anion derived from methyl trimethylsilylacetate provides an a,@unsaturated ester which is deconjugated by kinetic deprotonation followed by kinetic protonation of the resulting ester dienolate. The ester 1675 is the sole product of this sequence. Hydroboration-oxidation is regio- and stereoselective, but simultaneously causes reduction of the ester to the primary alcohol. Selective reoxidation provides the lactone 1676. A novel a-methylenation sequence is then employed to prepare 1677 which is converted to aromaticin (1641) in four straightforward steps.

Other Bicarbocyclic Sesquiterpenes

&,

376

&iph

PhSOZMe,, NOH

I-BU

1671

‘. @Ph I-Bu

1672

H

H

Ran;

‘,

I . Me3SiCHLiCOOMeb

Ni,

2. LDA

3. HOAc

I-Bu

1673

1674 I . LOA, BrCHZOMe

I . BH3,THF

2. t i 2 0 2

I-BuR

C

SPh

I-Bu

I-Bu

1575

&-

MeS03H, AC2 0

O

O

M

e

3.

Pt,op

1675

*

2. I-BuOK.

I-Bu

*

1676

@+

__t

2I .. PCC H+

I-Bu 1677

%

I . PhSeCl, H+ 2 . NaI0,

____)

1678

1641

Scheme 205.

Lansbury’s (*)-Aromaticin Synthesis

Lansbury has also prepared a r ~ r n a t i n(16421, ~ ~ ~ which differs from aromaticin (1641) only in configuration at C-8. Lactone 1676 is subjected to the sequence shown in Scheme 206 which produces the desired cis lactone 1680. The conversion of the latter to aromatin is carried out as shown in the scheme.

Fused Ring Compounds: 5,7

I . KOH, MeOH 2. M I S O ~ C I ,

I . Stiles rn 2. CHeO, EtflH, HOAc

E’SN

I-BU

377

Is-cs

1680

gee

Scheme 205

I-Bu

1681

Scheme 206.

1642

Lansbury’s Arornatin Synthesis

(6) Other Hydroazulenes: Daucene, Daucol, and Carotol.

The three sesquiterpenes daucene (1682)’ daucol (1683) and carotol (1684) all possess the unusual carotane skeleton, and the biosynthesis of these materials clearly is unique among the hydroazulenes. The more

1682

1683

1684

common guaiane derivatives are formed from farnesyl pyrophosphate (1685) by initial cyclization to a 10-membered ring intermediate (1686) and subsequent formation of the hydroazulene nucleus. On the other

378

Other Bicarbocyclic Sesquiterpenes

y

o

.

*

- r,

-

~

H

1686

guaione

(prrudoguoionr)

hand, the carotane skeleton appears to form by simultaneous closure of the five- and seven-membered rings via the hydroazulenic cation 1687. A 1,3-hydride shift is proposed to lead to 1688 which is the precursor for the isolated materials.430

Yamasaki was the first to prepare one of the carotanes, daucene (Scheme 207).431 Monoepoxidation of (+blimonene 1689 and mild acid hydrolysis provides the diol 1690. Periodate cleavage of 1690 gives a keto-aldehyde which is cyclized to the unsaturated aldehyde 1691 in approximately 25% overall yield. The aldol condensation of 1691 with acetone proceeds in 37% yield to trienone 1692. Selective reduction of the less-substituted, conjugated double bond is effected by treatment with triphenyltin hydride. The isopropenyl group is then saturated to afford 1693. The ethynyl carbinol derived from 1693 is converted to the allylic alcohol 1198 which cyclizes to daucene 1682 (42% yield) upon exposure to formic acid. Four unidentified products (48% yield) were also reported. At about the same time Naegeli and Kaiser reported the synthesis of racemic 1198 (see Scheme 144). They found that exposure of 1198 to stannic chloride in benzene affords essentially only two compounds, the

Fused Ring Compounds: 5,7

,

I . Na104 2. piprridinr.

2. I % H2SO4,

4

A

"20

,

HOAc

1690

1689

379

1691

I . PhjSnH 2 . He, P d K )

I . P-Na 2. H2, Lindlar's catalyst

I692

1693

HCOOH,

1198

*0 1682

Scheme 207.

Yamasaki's Synthesis of (-)-Daucene

major component (70% yield) being identified as (fhdaucene. The minor component is (fbacoradiene (1170).

q s p-& +

SnC1?+

1198

1682

1170

Within a few weeks after the appearance of the Naegeli and Kaiser report, Levisalles and co-workers reported the total syntheses of daucene (1682), daucol (1683) and carotol (1684) (Scheme 208).432 (+)-Dihydrocarvone (1694) is converted in the straightforward manner shown to diol 1697. After acetylation of the secondary alcohol, oxidation with lead tetraacetate cleaves the isopropyl group to afford, after saponification, the

380

Other Bicarbacyclic Sesquiterpenes

I.

04

4%

, ElO-,

P - NOtC&CO+

*

0

E120 2 . KOH,EtOH

1694

1695

@.;a

I . I-BuOK, Me1

I. Ac~O

2. L i A I H ,

0

2. 3. Pb(OAc14) -OH

HO@.*&H

1696

1697

A"

I . MnO2

2. L i / N H r

H

H

2. AC2O 3. separate

H

1699

1698

Jq-p*p+

AcO

p:: p I700

H+_

1701

2;

3. reparale

1703

P

1682

p- NOzC&C03H

1684

Scheme 208.

I702 I . p-N02C&CO3H

2. L i I E t N H 2

b

P:: 1683

Levisalle's Synthesis of Three Carotanes

Fused Ring Compounds: 5,7

381

diol 1698. The allylic alcohol is selectively oxidized and the resulting enone reduced. Diazomethane ring expansion of 1699 gives an inseparable 2:l mixture of keto alcohols. Chromatography of the derived acetates, however, allows isolation of the desired ketone 1700 which is the minor component of the mixture. After hydrolysis of the acetate, phosphorus pentachloride induces dehydration which is accompanied by skeletal rearrangement affording a mixture of 1701 and 1702. These are isomerized by acid to 1703. Addition of methylmagnesium iodide to 1703 and subsequent dehydration affords a mixture of compounds from which daucene (1682) may be isolated in unspecified yield. Epoxidation of daucene followed by reduction affords carotol (1684) in approximately 5% yield, some daucene, and four unidentified alcohols. Upon further oxidation, carotol is converted to daucol (1683) in unspecified yield.

(7) Other Hydroazulenes: Velleral and Pyrovellerolactone Several sesquiterpenes with the unique lactarane skeleton have been isolated. Three of these, pyrovellerolactone (17041, velleral (17051, and vellerolactone (1706) have been successfully synthesized by Froborg and M a g n u s s ~ n In . ~referring ~ ~ ~ ~to~ the ~ ~ original literature, the reader is

1704

1705

I706

cautioned that the earlier papers dealing with these three substances433a,b contain erroneous assignments of the structures of the natural products and of many of the synthetic intermediates. The last paper of the series433c provides the correct structures and with that knowledge we have corrected the earlier reports in editing them for inclusion in this section. The first report details the preparation of pyrovellerolactone (1704, Scheme 209).433aCycloalkylation of 1707 with bischloromethylfuran (1708) provides 1709 in high yield. After reduction and mesyla-

382

Other Bicarbocyclic Sesquiterpenes

1707

-+ 3 . HOAc. NaOAc

I foe

I709

H

i . separate

1710

H+_

2. - 2 e - , ~ $

dH

1712

1711

1704

Scheme 209.

1713

Magnusson’s First Synthesis of

(f1-Pyrovellerolactone

tion, solvolysis leads to a mixture of hydroazulenes 1710 (53%) and 1711 (26%). (This method for synthesizing the hydroazulene ring system was developed by Marshall in connection with his bulnesol synthesis and has been discussed in detail in Volume 2 of this series434.) For the synthesis of pyrovellerolactone the minor isomer (1711) is electrochemically oxidized. Subsequent dehydration of 1712 affords pyrovellerolactone 1704 in 12% yield along with the regioisomer 1713 in approximately l0h yield. A successful approach to the synthesis of all three natural products was reported in two papers in 1976433band in 1978433c(Scheme 210). The known /3-keto ester 1714 is condensed with 2-nitro-1-butene and subsequently decarboxylated. Nef reaction gives the diketone 1716 in 47% overall yield from 1714. Aldol cyclization occurs in fair yield and the

Fused Ring Compounds: $7

a

Em ,

I . NOH, 2. HCI, HOAc

aCOOM~ 1714

I . NOOH, 2. HCI

1715

1716

I

I . NaOH, EtOH

2. Hp, RhmAlp03 1717

1718

2. I. R--'-COOMe A

*

&COOMI

3. BH3*THF 1719 a. R = CH (0Me)p b. R=CH#Ac

2 . MnO2

3.,'H

H20

17190

I705

I706

I704

Scheme 21 0. Magnusson-Froborg Synthesis of Velleral, Vellerolactone and Pyrovellerolactone

383

384

Otber Bicarbacyclic Sesquiterpenes

product is hydrogenated to 1717. Enamine formation also leads to epimerization of the methyl group to the more stable exo configuration. Cycloaddition of the appropriate acetylenic ester occurs smoothly, the intermediate cyclobutene is thermolyzed and the product enamine cleaved by borane in THF. For the synthesis of velleral (1709,the oxidation state of the ester in 1719a is first adjusted and the acetal subsequently hydrolyzed. Vellerolactone (1706) is derived from 1719b by sequential treatment with hydroxide, to saponify the esters, and acid to catalyze lactonization. Heating vellerolactone in refluxing toluene for 4 hours causes isomerization to pyrovellerolactone (1704).

H. Fused Ring Compounds: 6,7 (1) Himachalenes

Several syntheses of a- and p-himachalene (1720 and 1721) and ar-himachalene (1722) have been reported. The Wenkert-Naemura syn-

1720

1721

I722

thesis (Scheme 21 1) employs an intramolecular Diels-Alder reaction to establish the bicyclic ring skeleton.435 The requisite trienone is elaborated from aldehyde 1723 by the straightforward sequence shown in the scheme. Cyclization is brought about with Lewis acid catalysis; cisfused enone 1728 is apparently the only isomer produced in the reaction. This compound reacts with methyllithium to give 7-isohimachalol (1729),which is dehydrated by phosphorus oxychloride and pyridine to a 4:l mixture of a-himachalene (1720) and p-himachalene (1721). Pure a-himachalene may be prepared by allowing ketone 1728 to react with methylenetriphenylphosphorane.

Fused Ring Compounds: 6,7

k

C

H

0

I723

I . Bf+OOEt Zn 2. POCIS,C~HSN) 3. H,, P d K , piparidine, EtOH

385

I . I 0,

h

c

o

o

E

t

2. KHSO,,Ab

1724

I . LiAIH4 hCooEt

b

2. CrO3(CSH,N)zw

C

H

0

&-

I726

1725

+A+

AICI3

1727

MaLi

1728

H \\

I729

Scheme 21 1.

I720

I72 I

Wenkert-Naemura Synthesis of the Himachalenes

In 1974, Mehta and Kapoor reported the synthesis summarized in Scheme 21 2.436 The synthesis begins with ( E)-o-bromolongifolene (1730), available in two steps from natural longifolene. Solvolysis of 1730 in anhydrous trifluoroacetic acid gives a mixture of isomeric hydrocarbons and bromo ester 1731 (53% yield). The mechanism of this interesting reaction involves Wagner-Meerwein rearrangement, followed by a transannular hydride shift. Hydrolysis of the ester and oxidation of the resulting secondary alcohol gives ketone 1732 (55% yield), accompanied by 27% of a diketone which is formed by over-oxidation of 1732. The enolate formed upon treatment of bromo ketone 1732 with dimsyl sodium undergoes fragmentation to give a mixture of three isomeric

386

Other Bicarbcyclic Sesquiterpenes

@%+p-@ PCOCF,

DMSO NaH

1730

1733

1731

1732

1734 I . chloranil

2. PIK,

A

1721

1722

Scheme 212. Mehta-Kapoor Synthesis of (*)-Himachalene Dihydrochloride and ( f )- mHimachalene

dienones, of which isomer 1733 is the major component (75%). WolffKishner reduction of 1733 gives diene 1734, a himachalene isomer. This compound gives a dihydrochloride, which has previously been converted into P-himachalene (1721). The most interesting aspect of this synthesis is the conversion of diene 1734 to (+I-ar-himachalene (1722), a transformation which apparently occurs with complete retention of optical actiuig. From this result it can be deduced that the chloranil dehydrogenation of 1734 gives only triene 1735, since any 1736 formed would lead to loss in optical activity. Furthermore, the aromatization step must take

Fused Ring Compounds: 6,f

387

place by suprafacial transfer of hydrogen from one end of the allylic system to the other, in a process not allowed on the basis of the Woodward-Hoffman rules.

1735

1734

1722

I

I736

Piers' synthesis of racemic p-himachalene is outlined in Scheme 213.437 Unsaturated acetal 1737, obtained from acrolein and 2,2-dimethyl-l,3-propanediol,is allowed to react with bromoform and sodium hydroxide under phase-transfer conditions to obtain dibromocyclopropane 1738. When a cold solution of 1738 is treated with n-butyllithium in the presence of methyl iodide, a 6:l mixture of isomers 1739 and 1740 is obtained. The isomers are separated and the major isomer (isolated in 74% yield) is hydrolyzed to obtain aldehyde 1741. After Wittig reaction, bromide 1742 is converted into a cuprate reagent with is added to 3-iodocyclohexenone. Substituted cyclohexenone 1743 is obtained in quantitative yield; thermolysis of 1743 in refluxing xylene provides enone 1744, which is methylated and selectively hydrogenated to obtain 1746. The ketone is removed via the derived enol phosphonate (Ireland method) to obtain (*)$-himachalene (1721).

388

Other Bicarbocyclic Sesquiterpenes

6 C H O

B

d6, H+_

+

r

n-BuLi

M ~ I , THF,)

H

h

I737 I . srparotr,

+

1739

H20

&pch 1743

1742

B.Pr,,,

Phg+,

2 . HCOOH,

1740 I . I-BuLi

HMPT

1738

1741

(&

/ MLDA II-

0;

1744

1741

Scheme 213.

1746

Pier’s Synthesis of

r.

1721

(*)-p-Himachalene

(2) Perforenone Perforenone (1752) is a bicyclic sesquiterpene of marine origin. It has been synthesized by Gonziilez and co-workers as summarized in Scheme 214.438 Homoallylic bromide 1747, obtained from methyl cyclopropyl ketone by Julia’s method, is metalated and condensed with methyl vinyl ketone to obtain dienol 1748. Cyclization of this substance occurs upon treatment with N-bromosuccinimide. When bromo ether 1749 is heated

Fused Ring Compounds: 6,7

1749

389

1751

1750

NaH. THF I

I752

Scheme 214.

1753

Gonzdlez’ Synthesis of (*)-Perforenone

with DBN, elimination of HBr occurs to give en01 ether 1750 which undergoes Claisen rearrangement to cycloheptenone 1751. Robinson annelation occurs in poor yield (29%) giving a 2:l mixture of diastereomeric enones, of which (*)-perforenone (1752) is the major isomer. The noteworthy feature of this synthesis is the imaginative method employed to construct the cycloheptenone ring. (3) Widdrol Danishefsky has synthesized widdrol (1764) as shown in Scheme 215.439 The synthesis mainly serves as a vehicle to demonstrate the use of a cyclic version of Ireland’s ester enolate Claisen rearrangement to generate a cycloheptane ring (i.e., 1761-1763). The synthesis begins with cyclocitral (1754)’ which is converted into unsaturated ester 1757 by well-precedented transformations. Allylic bromination of this substance

Other Bicarbocyclic Sesquiterpenes

390

I755

1754

1756

1757

I

1758

I760

1759

I . LDA 2. Me1 3. LDA, TBDMSCI

\\ I762

1761

I.

COOH

I763

Scheme 21 5.

OSi+

I

soc12

2 . m-CPBA 3. L I A I H ,

I764

Danishefsky-Tsuzuki Synthesis of (*)-Widdrol

gives a diastereomeric mixture of allylic bromides (1758), which is oxidized to give epoxides 1759. Reductive elimination of 1759 gives a single alcohol (1760), showing that epoxidation of 1758 occurs only from the face of the double bond cis to the methyl group. The derived lactone 1761 is then subjected to the Ireland reaction to obtain acid 1763. The carboxy group is transformed into hydroxy by a "carboxy inversion reaction" giving racemic widdrol (1764).

Fused Ring Compounds: 4,9

391

I. Fused Ring Systems: 4,9 (1) Isocaryophyllens

In Volume 2 we reported a model study by Gras, Maurin, and Bertrand This work has now been aimed at the total synthesis of ~aryophyllene.~~' extended to the synthesis of (+I-isocaryophyllene shown in Scheme 21 6.441 1,5-Cyclooctadiene is converted by two successive carbene reactions into alkene 1767, which is resolved by partial hydroboration with the borane derived from (-)-a-pinene. The dextrorotatory enantiomer

0

I . CI,C:

1. ~rzc:

2 . Li, I-BuOH)

2. n-BuLi

1765

9

resolution,

I766

1767

I . LIAIH, 2. p-T,CI* (+)

1767

I768

1770

1772

Scheme 216.

Bertrand-Gras Synthesis of

TsO 1769

1771

1773

(fbIsocaryophyllene

392

Other Bicsrbocyclic Sesquiterpenes

so produced is subjected to cycloaddition with dimethylketene to obtain tricyclic enone 1768. Although this compound is a diastereomeric mixture, each diastereomer has the R configuration at the 4,9-ring fusion position. The enone is reduced and the resulting alcohol converted to tosylate 1769. Reduction of this allylic tosylate gives a mixture of double bond isomers. However, upon preparative glpc at 150"C, double bond isomerization occurs giving principally isomer 1770. The derived ketone 1771 is pyrolyzed at 360°C to give enone 1772. Wittig methylenation gives (+I-isocaryophyllene. A key reaction in this synthesis is the thermal rearrangement of cyclopropane 1771 to olefin 1772. The isomerization actually affords three products, but 1772 is the major product and is isolated in 56% yield. Devaprabhakara and co-workers have capitalized on the Gras-Bertrand approach by the modification outlined in Scheme 217.442 In this approach, 1,5-cyclooctadiene is first converted to allene 1774, which is converted into dienone 1775. Photoannelation of this compound with

1765

1774

I776

I775

1777

I772

I773

Scheme 217.

Devaprabhakara's Synthesis of

(*)-Isocaryophyllene

Fused Systems

393

isobutylene gives 1776 which is subjected to Simmons-Smith methylenation to obtain ketone 1777 (presumably a mixture of 1771 and its cisfused isomer). The synthesis is completed by the Gras-Bertrand method. Devaprabhakara has reported an alternate synthesis of enone 1775.443

6. TRICARBOCYCLIC AND TETRACARBOCYCLIC SESQUITER PENES

A. Fused Systems In this section, we will discuss syntheses of tricarbocyclic systems in which the three rings are fused side-to-side. Syntheses of cubebol (1) and the cubebenes (2 and 31, which were discussed in Volume 2 of this series, have been reported in full. 444-446 New syntheses which will be discussed are those of illudol (41, the protoilludanols 5 and 6, the protoilludenes 7 and 8, marasmic acid (91, hirsutene (lo), hirsutic acid C (ll),coriolin (12) and isocomene (13).

?H

4

OH

5

6

7

394

Tricnrbocyclic and Tetracarbocyclic Sesquiterpenes

9

I2

10

13

(1) Illudol, Protoilludanols, Protoilludenes

In Volume 2 of this series we reported a synthesis of the protoilludane skeleton from the group of T. Matsumoto at Hokkaido University in Sapporo.447The synthesis has subsequently been extended to a synthesis ~ with ~ ~ of (*)-illudol (4) as shown in Scheme l.448Keto ester 1 4 reacts allylrnagnesium bromide to give a diol (15) which is transformed by the three-step sequence shown into dihydroxy ester 16. Cleavage of the diol gives an intermediate diketo ester which closes to give a mixture of aldol products 17 and 18. Dehydration provides an unsaturated keto ester (19) which is reduced by sodium hydridobis(2-methoxyethoxy)aluminate to obtain diol 20 (47Yo) accompanied by 17% of its diastereomer. Treatment of diol 20 with acetone and acid yields a keto ketal (21) which is reduced and deketalized to obtain (*)-illudol (4). The stereochemistry of the two secondary alcohols was assigned on the basis of reduction of the corresponding ketones 19 and 21 from the less hindered direction in each case. The synthesis provided the first concrete evidence in favor of the relative stereochemistry of illudol at these centers.

Fused Systems

395

I5

14

NaI04

xOEt

'-

EtO

EtO

&$ #*

OEt

Et09Et

?H I . Nbenzene aAIHz(dVo~)Z,

2 . H30+

T

O

H I

bH 21

Scheme 1.

4

Matsumoto's Synthesis of (*)-Illudol

In 1975 the Matsumoto group reported a synthesis of the protoilludanols 5 and 6 and the protoilludene 7 (Scheme 2).449 Enone 22447 undergoes photochemical cycloaddition with ethylene to provide the cis,anti,cis tricyclic ketone 23 in 75% yield. Wittig methylenation of 23

396

Trtcarbocyclic and Tetracarbocyclic Sesquiterpenes

23

22

24

TI(CI0,) H20

25

5 Ph3P=C Hg

I. H ~ ~ O A C ) ~ H 0,THF

2 2 . NOSH, 6

7

Scheme 2.

Matsumoto’s Synthesis of Protoilludanes

gives alkene 24, which is ring-expanded with thallium perchlorate to obtain ketone 25. This substance is converted into alcohols 5 and 6 and alkene 7 by the straightforward methods illustrated in the scheme. Takeshita and co-workers have also investigated the synthesis of illudanes by photoaddition of ethylene and 1,l-dimethoxyethylene to a hydrinden~ne.~’’In the Takeshita approach (Scheme 3) the hydrindenone 27 is prepared by DeMayo’s method. The four-membered ring is added by photoaddition of 1,l-dimethoxyethylene. Keto ketal 28 is treated with methylmagnesium iodide and the resulting alcohol is converted into unsaturated thioketal 5 by treatment with 1,2-ethanedithiol

Fused Systems

397

27

26

29

28 Ronry Ni

Y

-

H

q

30

Scheme 3.

-

H

8

Takeshita's Synthesis of Protoillud-7-ene

and boron trifluoride. Desulfurization of 30 provides protoilludene 8. Related studies showed that trans-fused hydrindenone 31 undergoes photoaddition to ethylene but that its cis-fused isomer 32 does not. CH =CH u hv

--H

Me00

COOMe 31

I

COOMr

32

398

Tricarboeyclic and Tetracarbocyclic Sesquiterpenes

(2) Marasmic Acid and Isomarasmic Acid Marasmic acid (9) has been the synthetic target of a number of investigations. De Mayo and co-workers took the imaginative photochemical approach outlined in Scheme 4.45' Spiro[4.2lheptene 33 is irradiated in the presence of the enol acetate of 1,2-~yclopentanedione to obtain adduct 35, which is hydrogenolyzed to 36. After introduction of a double bond in the cyclopentanone ring, the acyloin is rearranged by treatment with 1% methanolic sodium hydroxide. The product of this rearrangement, hydroxy ketone 38, is cleaved with lead tetraacetate. After esterification, the enone system is regenerated by treatment of 39 with acid. Enone 40 is subjected to oxidation with singlet oxygen to obtain a hydroperoxide (41) which is photolyzed in the presence of vinylene carbonate. Catalytic hydrogenation reduces the hydroperoxy group and ketone carbonyl, giving 42. The secondary hydroxy is protected as the pivalate and the tertiary alcohol dehydrated to obtain unsaturated ester 43. Addition of diazomethane to 43 affords a pyrazoline which loses nitrogen upon irradiation to give 44. After saponification of the carbonate linkage, cleavage of the resulting glycol, and elimination of pivalic acid methyl isomarasmate (45) is obtained. Although the stereostructure of 45 at the allylic position was not rigorously defined, the compound is a single isomer and is definitely different from methyl marasmate (46), obtained from authentic marasmic acid. The stereo-

COOMI

46

chemistry shown in Scheme 4 is based on the assumption that the hydrindane skeleton is cis-fused. Since 45 and 46 were not interconverted under equilibrating conditions, 45 must have the cyclopropane ring trans to the angular hydrogen, as shown.

Fused Systems

33

34

+

,

I . CsHsNH,Bri

2.

DEN

a

I . HZ-Pt

A

A

35

36

&Me

HO-,MeOH,

2. I . Pb(OAct4, CH2N2

ti

37

39

I . hu,

38

40

@

41

I.

____)

2. HL-PtlRh

3\cocI,

2. POCIS,

& HO COOMe

CsHsN

OOMe

42

I . CHON2

2. hu

399

*

COOMe 4

44

Scheme 4.

43

I . OH-

2. NaI0,

*

3. Et3N, benzene,

A

wH: COOMe

45

de Mayo’s Synthesis of Methyl k )-Isomarasmate

400

Tricarbocyclic and Tetracarbocyclic Sesquiterpenes

In 1973, Wilson and Turner reported an approach to marasmic acid which utilized as a key reaction addition of diazomethane to a bicyclic triester (47).452 After photolysis of the adduct, a single cyclopropane was formed and was presumed to have structure 48.

In 1976, Greenlee and Woodward reported an investigation which casts doubt on the Wilson-Turner stereochemical interpretation^.^'^ The Greenlee-Woodward approach (Scheme 5 ) involved construction of intermediate 53 by addition of diene 52 to dimethyl acetylenedicarboxylate. As in the Wilson-Turner work, Greenlee and Woodward found that photolysis of the initial pyrazoline affords a single cyclopropane, which was subjected to straightforward manipulations to obtain anhydride 55. Reduction with disodium tetracarbonylferrate gives a 1:l mixture of lactols, which were separated and converted into their acetates. Hydrolysis of one of these isomers gave isomarasmic acid (58), which was converted by esterification into methyl isomarasmate, identical with the de Mayo sample. Thus, it is clear that addition of diazomethane to the a,P-unsaturated ester shows a strong preference for addition anti to the angular hydrogen, both in de Mayo's intermediate 43 and in the Greenlee-Woodward intermediate 53. A convincing explanation for this highly stereoselective addition has not been advanced. Greenlee and Woodward solved the problem of the cyclopropane stereochemistry with the elegant and highly economical synthesis summarized in Scheme 6.453Reduction of dienal 51 gives dienol 59, which is treated with bromomethylmaleic anhydride at room temperature to obtain a 1:l mixture of lactone acids 60 and 61. The derived t-butyl esters are treated with potassium t-butoxide to obtain a single cyclopropane (62), which is produced in 44% overall yield from dienol 59. This material is converted into (f)-marasmic acid (9) by the sequence of

Fused Systems

cCHo * @

H(OEt12

CH(OEt)3,H+

I.

z:F*’ * WCH0

2 2. NaOAc. HOAc

51

50

49

kOOMe

@

52

--CH(OMe12 COOMe

53

I . NaOH, MeOH

2.

A

*

COOMe

55

54 %:Me)2

+

2. I . separate Ace0

$0” .’

3. K2C03* MeOH 4. H30+ 57

56

58

Scheme 5.

401

45

Greenlee-Woodward Synthesis of ( f Hsomarasmic Acid

Tricarbocyclic and Tetracarbocyclic Sesqulterpenes

402

rCHo rH DIBAL,

x q p x EA,”+ 7 q”” 51

69

Br

60

I.

Hood

B,

I-BUOOC

OH

63

62

61

65

64

66

Scheme 6.

I-BUOOC

9

Greenlee-Woodward Synthesis of (*)-Marasmic Acid

Fused Systems

403

steps shown in the scheme. The key step in the successful GreenleeWoodward synthesis presumably involves intramolecular Diels-Alder reaction of esters 67 and 68, which are formed by reaction of alcohol 59 with the anhydride. The fact that both adducts eventually afford marasmic acid shows that both cycloadditions proceed by way of an endo transition state:

67

In all, the Greenlee-Woodward synthesis requires only 12 steps from the known aldehyde 49 and is completely regio- and stereoselective.

In 1980, Boeckman and KO reported the marasmic acid synthesis which

is summarized in Scheme 7.454Unsaturated aldehyde 69 is transformed into the monoprotected dimethyladipaldehyde 70, which is transformed by Wadsworth-Emmons reaction into unsaturated ester 71. Further elaboration of this substance provides aldehyde 74, which is condensed with phosphonate 75 to obtain the crucial intermediate 76. The intramolecular Diels-Alder reaction is achieved by heating 76 in toluene in a sealed tube. Under these conditions isomers 77 and 78 are produced in a 1:l ratio. This unfortunate situation is a result of poor stereoselectivity in the Diels-Alder reaction, which appears to proceed by way of endo and exo transition states with equal facility. The isomers are separated after conjugation of the double bond of 77 (the isomer resulting from exo addition coqjugates under the conditions of cycloaddition). After conversion of the primary acetoxy group into a better leaving group, the

Tricnrbocyclic and Tetracnrbocyclic Sesquiterpenes

404

I MeOH,n+ 2. a H j T H F 4 . Collins

6Q

B'

70

70

79

COOMe

ao

a4

Scheme 7.

Boeckmann-Ko Synthesis of (f1-Marasmic Acid

Fused Systems

405

cyclopropane ring is formed in the same manner as was employed by Greenlee and Woodward. The remainder of the synthesis involves some rather nice functional group manipulations, resulting in the eventual synthesis of (f1-marasmic acid (9). The Boeckman-Ko synthesis requires a total of 20 steps and requires separation of diastereomers after the Diels-Alder step. Although this isomer separation is necessary, Boeckman and KO were able to convert isomer 78 by an alternate sequence of reactions also into (f1-marasmic acid.

(3) Hirsutic Acid, Isohirsutic Acid, Hirsutene, and Coriolin The first published work directed toward the synthesis of hirsutane Alkylation sesquiterpenes came from Lansbury in 1971 (Scheme of the known 3-methyl-3-methoxycarbonylcyclopentanone with 3-bromo-2-butanone provides diketo esters 86 and 87 in 50% and 10% yields, respectively. Base-catalyzed aldolization of the major isomer is accompanied by rapid double bond isomerization, yielding keto acid 88 as a diastereomeric mixture. After esterification and hydrogenation of the double bond, ketone 89 is subjected to Claisen alkylation using (Z)-2-chloro-2-butenol. Isomers 91 and 92 are produced in 55% and 36% isolated yields. The major isomer is treated with 90% sulfuric acid to hydrolyze the vinyl chloride and then with potassium t-butoxide to effect cyclization. Acidic dehydration of aldol 93 affords enone ester 94. Although compound 94 is a potential precursor to hirsutic acid C, the substance has apparently not been converted further. Subsequently, Lansbury and co-workers reported a modification of this basic approach which provides isohirsutic acid (991, a substance formed by rearrangement of hirsutic acid C.456 The synthesis begins with the keto ester 89, which is alkylated by the Claisen route with allylic alcohol 95 to obtain a 3:2 mixture of keto esters 96 and 97 (Scheme 9 ) . After separation, the major isomer (96) is treated with 90% sulfuric acid. Hydrolysis of the vinyl chloride is followed by elimination of propanethiol, leading to enedione 98. This material is treated with mild base to effect ring closure to (*)-isohirsutic acid (99). There is some

Tricarbocyclic and Tetracarbocyclic Sesquiterpenes

406

@

MeOOC’

6 85

,

M e O O C V

87

263=

2 . H2-Pd/C-

I . separate 2 . base

o.

86

I . CH2N2

HOOC

+

Me00

H

89

&yl 190-

Me00C

I

A

90

I . HZS04

2. I-BuOK

*

&k A

H+

MeOOC’ ti

l.i

93

Scheme 8.

+ Meooc,&

MeOOC’

CI

k

91

92

-

MeOOC’

ti 94

Lansbury’s Approach to Hirsutic Acid

uncertainty regarding the identity of this material with isohirsutic acid derived from natural sources. Lansbury and co-workers report that the spectral data obtained from synthetic 99 “...correspond only in part with those reported”. However, the naturally derived isohirsutic acid may have been impure.

Fused Systems

Y r

k

I. H

e

6 3 '+

-

MaOOC'

2. heat

89

97

407

k 96

nooc-w 98

H

II

nooc-Q-$y 99

Scheme 9.

Lansbury's Synthesis of (*)-Isohirsutic Acid

The first successful synthesis of hirsutic acid C came from Matsumoto and co-workers in 1974.4571458As shown in Scheme 10, the dione monoketa! 100 is methylenated and the resulting alkene allowed to react with the carbenoid obtained from methyllithium and a,a-dichloromethyl methyl ether. Cyclopropanes 101 and 102 are obtained in a ratio of 2 1 . The isomers are separated and the major isomer (101) is treated with acid to effect the ring opening of the methoxycyclopropane unit. The resulting aldehyde is oxidized and esterified to obtain keto ester 103. The cyclopentanone is indirectly methylated and the product alkylated with methallyl chloride to obtain, after ozonolytic cleavage, dione 105. Aldolization is brought about with potassium t-butoxide, leading to

Tricarbocycllc and Tetrncarbocyclic Sesquiterpenes

408

100

101

I . separate 2. H30+

MaOOC'

3. Jones 4 . CHzNz

I. l L C l bare

2. ~ - B ~ s H H+) , 3. Ranry Ni I03

0

MeOOC'

2 . 0,

HCOOEt NaH

-

I BuOK

I02

MaOOC'

I04

MeOOC'

I06

I05

MeOOC'

2 . DMF, L i I *

H

HooC'

H

107

I09

Scheme 10.

I08

II

Matsumoto's Synthesis of (*)-Hirsutic Acid C

Fused Systems

409

enone ester 106. Installation of the a-methylene function is accomplished by formylation, followed by condensation with formaldehyde. After nucleophilic cleavage of the ester function, dienone 108 is treated with hydrogen peroxide in alkaline methanol. The major product of this epoxidation, complicatic acid (109, 6Wo) is accompanied by 20% of a P-methoxy ketone, which can be converted back into 108 by treatment with sodium hydroxide. Reduction of 109 with sodium bordhydride provides (*bhirsutic acid (11). The Matsumoto synthesis is reasonably efficient (16 steps from ketone 100) and is marred by the necessity of only one isomer separation. However, the non-stereocontrolled step occurs early in the synthesis. Trost and co-workers have also synthesized (*I-hirsutic acid C,by the route outlined in Scheme 11.459 In the Trost synthesis, the desired relative stereochemistry is achieved by employing a tetracyclic intermediate (118) which is ozonized to generate the tricyclic system. Compound 118 is prepared from the known cyano ketal 110 by the multi-step sequence illustrated in the scheme. The most noteworthy feature of this series of steps is the use of Stettler’s method to close the cyclopentanone ring (115-117). Ozonolysis of 118 provides tricyclic dialdehyde 119, which is necessarily of the correct stereostructure, since only one stereoisomer of 118 is capable of existence. After reduction of the aldehyde groups to methyls, the lactone ring of 120 is manipulated so as to obtain dione 105, an intermediate in the Matsumoto synthesis. As has already been stated, the Trost synthesis is of necessity stereospecific. However, a high price is paid, as the synthesis requires 33 steps, more than twice the number necessary in the Matsumoto synthesis. Tatsuta and co-workers have developed a route to hirsutanes which has been applied in the synthesis of (f)-hirsutene As shown in Scheme 12, 2,2-dimethylcyclohexanone is converted into unsaturated ketal 121 by the Garbisch procedure. After introduction of an allylic acetoxy group, 122 is irradiated in the presence of the enol acetate of 2-methyl-l,3-~yclopentanedioneto obtain tricyclic adduct 123 in 35% yield. After reduction of the carbonyl and protection of the resulting secondary alcohol, the two ester groups are hydrolyzed. The key reaction of the Tatsuta approach is solvolytic rearrangement of the mono-

Tricarbocyclic and Tetracarbocyclic Sesqulterpenes

410

I . Et#, toluene, A 2. H2-Pd/BoC03* 3. B r e O O E t , t n

I . LDA, Cog 2 . KOH,MeOH

2 . CH2N2

3. --P-COOM.

112

Ill

110

114

* P

I. K2C03,MOOH 2. NOEHq

N

I . PCC

2 . Et3N

"$1-

N+

I10

I

I . HCl,MeOH

2.

o3

3 . MepS

H

II?

+

* B

3. PhsP, MeOOCN=NCOOMe

MeOOC

118

I . MeSH, EF3

McOOC O

H

W

*

2 . Roney Ni

MeOOC'

H

I . KOH

____c 2 . H30+

3 . MeLi 4 . PCC

t-BuOK-

MeOOC'

4

e

O

O

C

O

W

5 stepr

H

108

Scheme 11.

M

I06

Trost's Synthesis of (*)-Hirsutic Acid C

I'

Fused Systems

I . Br2.

(CH20Hl2

H+ 2. I-BuOK

w

8-Q I . NBS 2. AgOAc

hu

OAc

I22

121

I . p-TtiCI, CSHSN 2. K2C03, acetone

2. M e P C I

&

Ac

.

124

I23

.&

F

p”0’

2. I . Na1,Zn DMF H30+

li

/

I . Hz-Pd/C MeOH ____) 2. LiAIH,

126

I25

I . NoH,CSz

I . H 0’ s 2 . PCC

Me1

2 . n-Bu3SnH

li

H I27

H

I28

H

I29

Scheme 12.

10

Tatsuta’s Synthesis of

( *)-Hirsutene

411

412

Tricarbocyclic and Tetracarbocyclic Sesquiterpenes

tosylate of diol 124, which provides epoxide 125. Deoxygenation of this material gives an alkene (126), which is hydrogenated and reduced to obtain monoprotected diol 127. For a synthesis of hirsutene, the superfluous hydroxy group must be removed, and this is accomplished by the Barton procedure. Deprotection and oxidation give a ketone (129) which is methylenated to obtain (fbhirsutene (10). Although this synthesis is long (19 steps), it is easily adapted to the more challenging target, coriolin. Such an adaptation (Scheme 13) was reported by Tatsuta in early 1980.461 Epoxide 125 is deoxygenated and the methoxymethyl and ketal groups removed to obtain unsaturated keto alcohol 130. Catalytic osmylation followed by ketalization of the resulting diol gives 131. The stereoselectivity of this oxidation results from the fact that attack on the other face of the double bond would give a highly strained nuns-bicyclo~3.3.0]octanesystem. After oxidation of 131, the diketone 132 is bis-sulfenylated by reaction with methyl o-nitrophenyl disulfide and the resulting dithioketal is transetherified by reaction with thallium trinitrate in methanol. Compound 133 is thereby obtained in 54% overall yield for the two-step process. After reaction of the more reactive carbony1 group with methyllithium the remaining ketone is reduced by lithium in ammonia to establish the desired exo stereochemistry at C-3. Hydrolysis of the acetonide, selective acetylation of the two secondary alcohols and dehydration of the tertiary alcohols give dienone 137. After removal of the protecting groups, the epoxide functions are installed by reaction with alkaline hydrogen peroxide. Tatsuta and co-workers make no comment regarding the stereoselectivity of the final epoxidation. However, it is presumably poor, based on analogy with the results obtained by other workers (see below). However, as the quoted yield for the last two steps is only 6W0,i t is reasonable to assume that other stereoisomers may also have been produced. In all, the Tatsuta synthesis requires 24 steps from 2,2-dimethylcyclohexanone.

Danishefsky and co-workers have reported the synthesis of (f1-coriolin which is summarized in Scheme 14.462 Conjugate addition of P-keto ester 138 to 5,s-dimethylcyclopentenone followed by acid-catalyzed aldoiization gives enedione 139 which undergoes Diels-Alder reaction

9

Fused Systems

I. N DM a 1F, Z n ___)

2. H2S04,

acetone

125

& ti

I

. 0504,N M O

2. Me2C(OMe)gf

I

H+

I30

I32

131

I33

2_ Ac 0

TFA

*Me

OMe

413

__c

I

H

on

OAc

I35

I34

137

Scheme 13.

I36

12

Tatsuta's Synthesis of (*t)-Coriolin

with 3-trimethylsilyloxy-1,3-pentadiene. The adduct (140) is degraded by the interesting sequence shown to obtain triketone 144. The overall result of this eight-step sequence is introduction of the 2-butanone-3-yl group to the eventual C-I position. Aldolization is accomplished by treatment of 144 with potassium t-butoxide. The a,fl-unsaturated ketone is decotljugated and the resulting enedione is selectively reduced

Tricarbocyclic and Tetracarbocyclic Sesqulterpenes

414

I40

I39

n

n

I41

142

H

I

I40

I44

&../

I46

3 LOA

PhSOSPh

n 140

147

I

m-CPEA

2 . Et0Ac.A

-

I . m-CPBA 2 . PCC

___c

H

___)

LI-NHL

t-BuOK,

2 . nOAc 3. Dl0AL

H'

A

143

-

, "

ti$+.,

OHooc

149

-

n I50

181

I,

-y.b;s-

t-~~oon, VO~ococ)~

2 . Sorratf

H

OH

I2

Scheme 14.

Danishefsky's Synthesis of (*)-Coriolin

IS9

Fused Systems

415

by reaction with diisobutylaluminumum hydride. Lithium-ammonia reduction of the remaining carbonyl group gives a diol (147) whicb is epoxidized. Selective oxidation with the Corey-Suggs reagent gives keto epoxide 148. Treatment of this material with three equivalents of lithium diisopropylamide and phenyl thiophenylsulfonate results in sulfenylation at C-11 and elimination of the Plyepoxy ketone, yielding intermediate 149. After elimination, enedione 150 is partially epoxidized to obtain epoxy enone 151. Further epoxidation of this compound was found to give a 7 5 mixture of (f)-coriolin (12) and its diastereomer (153). However, reduction of 151 gives a triol (152) which is oxidized

I53

by the Sharpless method to give only the spiroepoxide of the correct relative configuration. Selective oxidation of the resulting diepoxy triol using Sarrett’s reagent has been reported by Umezawa. The remarkable aspect of Danishefsky’s synthesis is that the entire 24 steps are carried out without the necessity of using a single protecting group. Considering the number of different functional groups, this is a signal achievement. Ikegami and co-workers have also reported a synthesis of (f)-coriolin (Scheme 15).463 1,3-Cyclooctadiene is converted into enone 154 by an eight-step sequence in 35% overall yield. Addition of lithium dimethylcopper occurs to give 155, which is subjected to Conia alkylation to obtain 156. Lithium-ammonia reduction places the new secondary alcohol predominantly in the more stable exo position. However, both diastereomers (157 and 158) are obtained in a 5:2 ratio. After exchange of protecting groups, the secondary alcohol is oxidized and the final ring is added by the three-step sequence 15+160 . The noteworthy feature of this series of steps is use of a Wacker oxidation to convert the ally1 group into an acetonyl group. Enone 160 is converted into dienone 162, which is deconjugated to obtain 163. Peracid epoxidation of this material

416

Tricarbocyclic and Tetracarbocyclic Sesquiterpenes

THP?

I . PhSeCl

2 . H202 3. HOAc-+

t-BuOK

H20

n 100

101

I . I-BuOK

*

m-CPBA

2 . HOAc

104

DBN_

Danishefsky, H

on I80

Scheme 15.

ti

on lZ

Ikegami's Synthesis of (*)-Coriolin

Fused Systems

417

provides an epoxide (164) which undergoes elimination upon treatment with DBN. At this juncture, the Ikegami synthesis converges with Danishefsky’s route. In all, the synthesis requires 28 steps and involves one isomer separation. Greene has recently reported the remarkably efficient synthesis of hirsutene (10) which is outlined in Scheme 16.464 The basic reaction employed in the Greene approach is a three-carbon annelation, which is

I . NoBH4 2. Cr(C104)eb

I . NoBH4

CHzN2,

2. Cr(C104)2b

ti

Et20

166

CI

ti

CCISCOCI,

ti

ti

165

y

4 ti

169

I68

167

I . HC104,

HCOOH

H

Scheme 16.

10

Greene’s Synthesis of (ItbHirsutene

accomplished by cycloaddition of a chloro ketene to an alkene, followed by diazomethane ring expansion. The resulting chloro ketone is transformed into a new alkene by a two-stage reduction (165-166). Repetition of the process gives a vinyl chloride (169) which is hydrolyzed and methylenated to obtain (&)-hirsutene (10).

418

Tricarbocyclic and Tetracarbocyciic Sesquiterpenes

At this point, we discuss a final synthetic endeavor in the hirsutane area. Although it has not yet led to the synthesis of a natural product, the novelty of the chemistry justifies its inclusion. Little and Miller converted 3,3-dimethylglutaric anhydride, via the known hemiacetal 170, into unsaturated ester aldehyde 171 (Scheme 17).465 Condensation of 171 with cyclopentadiene gives fulvene 172, which reacts with di(2,2,2-trichloroethyl) azodicarboxylate to give an adduct, which is selectively hydrogenated to 173. The trichloroethyl groups are removed by controlled potential electrolysis and the resulting hydrazine is oxidized to azo compound 174. When 174 is refluxed in acetonitrile, nitrogen is

I . Ph3PeC00Me

I. NoBH~

OH

0

2 . PCC

CHO 171

170

172

**COOMI

-[ + MeCN

I . electrolysis

H

173

~

c

reflux

2. KJFIICN)~

174

175

H"

dOOMe

I76

Scheme 17.

Little-Miller Hirsutane Synthesis

o

o

-

M

~

Fused Systems

419

extruded to give a 1,3-diyl (175) which is trapped by the alkene linkage. The tricyclic product (176) is produced in 50% yield based on biscarbamate 173.

(4) Isocomene The tricyclic hydrocarbon isocomene (13) has attracted several synthetic efforts. The first published synthesis was that of Oppolzer and coworkers (Scheme Annelation of 2-methylcyclopentanone with dienone 177 produces a hydrindenone (178) which is methylated to

13

obtain 179 and its C-4 epimer 180 in a ratio of 5:l. When 179 is heated to 280°C, intramolecular ene reaction occurs to give tricyclic ketone 181. This reaction neatly establishes the two remaining stereocenters of isocomene. Unfortunately, the yield in this process is only 17%. The synthesis is completed by ring-contraction, which is effected by photochemical Wolff rearrangement of diazo ketone 182, reduction of the resulting methoxycarbonyl group, introduction of unsaturation, and isomerization of the double bond of 185 into the ring. In all, Oppolzer's synthesis requires 11 steps and affords (f)-isocomene in 0.7% overall yield. Pirrung's synthesis is summarized in Scheme 19.467 Stork-Danheiser methylation of the well-known enol ether 186 is followed by addition of the Grignard reagent 187. The resulting dienone (188) undergoes smooth photoaddition to give tricyclic ketone 189 in 77% yield. Although ketone 189 is highly hindered, and undergoes predominant enolization when treated with methyllithium, it reacts with methylenetriphenylphosphorane under forcing conditions to give hydrocarbon 190. Treatment of this compound with p-toluenesulfonic acid in benzene

420

Tricrrbocyciic and Tetrrcrrbocyciic Sesquiterpenes

qo"gvv, +

MeO-/MeOHI

I. 2. t-BuOK

benzene

I77

178

-

2. I-BuOK, AmON0

2 . 280.

I80

179

181

COOMe

MeOH

.

3. N a O C I , N H 3

I . LiAIH, 2. o-N02C6H4SeCN,e CSHSN

I83

182

I . NaIO,

p-TsOH ___)

184

Scheme 18.

@

I85

13

Oppolzer's Synthesis of (*)-Isocomene

brings about Wagner-Meerwein rearrangement and affords ( f)-isocomene (13). The synthesis requires seven steps and proceeds in 34% overall yield. Paquette's isocomene synthesis (Scheme 20)468 begins with a Yoshikoshi-type annelation of enol ether 191 with 2-nitro-l-butene. Codugate addition of Grignard reagent 193 to the resulting enone 192 gives a keto acetal (194) which is converted into unsaturated aldehyde 195 by the efficient method illustrated in the scheme. The key step of

Fused Systems

I89

Scheme 19.

421

13

190

Pirrung's Synthesis of (*)-Isocomene

NO n

I92

191

0

I95

($

196

Wolff-Kishner,

198

I97

Scheme 20.

Paquette's Synthesis of

3. m-CPBA 4. A

@I3

(*1-1socomene

422

Tricarbocyclic and Tetracarbocyclic Sesquiterpenes

the Paquette synthesis is the intramolecular Prins reaction which provides unsaturated alcohol 196 as a mixture of diastereomers. Straightforward manipulations convert 196 into dienone 197 which reacts stereospecifically with lithium dimethylcopper to afford unsaturated ketone 198. Wolff-Kishner reduction completes an efficient synthesis (12 steps, 22% overall yield). The most recent synthesis of isocomene is that of Dauben and Walker (Scheme 21).469 Compound 199 is alkylated with ethyl /3-bromopropionate and the thioacetal is hydrolyzed to obtain diketo ester 201. Condensation of this material with dimethyl acetonedicarboxylategives a diketo pentaester (202) which is hydrolyzed, decarboxylated, and reesterified to obtain diketo ester 203. Wolff-Kishner reduction of monoketal 204 followed by deprotection gives keto acid 205. Compound 205 undergoes smooth cyclization under acidic conditions to provide @-diketone 206 which is methylated to give the sensitive nonenolizable fl-diketone 207. Thioketalization of the more reactive carbony1 gives 208, which is converted into alkene 209 by reaction with methylenetriphenylphosphorane under forcing conditions. After isomerization of the double bond into the ring, the dithiolane group is cleaved to give unsaturated ketone 210, an intermediate in the Paquette synthesis. The Dauben-Walker synthesis requires 19 steps and produces isocomene in 1.6% overall yield. An earlier report claiming the synthesis of i s ~ c o m e n e ~is~ 'probably erroneous.47 1

Fused Systems

Me&@

COOMe

DMF

COOMe

I . 6rVHCI,A

Meon

2. KF, MIX,

& MaOOC

I . n+ Mr$(CH2OH)2

Z . n+,acetone

o&

423

*

POP

COOMe

0

&PO3

n30+

2. I . Wolff-Kirhnrr

p-TaOH

benzene

908

204

206

e07

POI

210

209

13

Scheme 21.

Dauben-Walker Synthesis of (kbisocornene

424

Tricarbocyciic and Tetracarbocyciic Sesquiterpenes

B. Bridged Systems

(I)

Gymnominol

The interesting 4,8-methanoperhydroazulene skeleton of gymnomitrol (211) has elicited a number of interesting synthetic investigations. The

21 I

fist published synthesis was that of Coates and co-workers (Scheme 22).472 Dione 212, available in 52% yield by condensation of biacetyl with dimethyl acetonedicarboxylate, is deoxygenated by Ireland’s method and the resulting unsaturated ketone hydrogenated to obtain ketone 213. After introduction of an a-cyano group, 214 is alkylated by reaction with the acetal of acrolein in refluxing benzene; the ethoxyallylated product 215 is obtained in 76% yield. After formation of the ethylene glycol acetal, the cyano group is removed reductively. The resulting enolate is trapped with trimethylsilyl chloride, then regenerated and methylated to obtain 217, which is hydrolyzed to keto aldehyde 218. Somewhat surprisingly, compound 218 is reluctant to undergo aldol condensation, giving an equilibrium mixture consisting of approximately 2-3 parts of 218 to one part of bridged ketol upon treatment with sodium carbonate in aqueous methanol. Consequently the aldehyde is oxidized and the resulting acid converted into enol lactone 219. Reduction of 219 with diisobutylaluminum hydride produces aldol 220, perhaps as a result of alkoxide promoted [1,3)sigmatropic rearrangement of the initially formed enol hem iacetal :

425

Bridged Systems

212

213

215

214

I . Jones

HCI

& 21 7

216

DIBAL,

acetone

OH

219

218

220 I . LiAlH4

*

2. separate

22 I

222

223

21 I

Scheme 22.

Coates' Synthesis of (*t)-Gymnomitrol

After oxidation of 220 to dione 221, the final carbon is added by reaction of the more reactive carbonyi with methyllithium. Dehydration of the resulting tertiary alcohol gives a 1:l mixture of exocyclic and endocyclic

Tricarbocycllc and Tetracarbocyclic Sesquiterpenes

426

double bond isomers 222 and 223. After reduction of the remaining carbony1 the isomers are separated to obtain (f)-gymnomitrol (211). The synthesis requires 16 steps, proceeds in 2% overall yield, and requires one isomer separation. The imaginative approach to gymnomitrol (Scheme ~ 3 ) ~by ' BClchi ~ and Chu begins with aldehyde 224, which is cleaved by Baeyer-Villiger oxidation to phenol 225. Oxidation of this material in methanol affords trimethoxycyclohexadienone 226. This material is condensed with 1,2-dirnethylcyclopentene in the presence of stannic chloride. The initially formed, highly fragile diketone is reduced with sodium borohydride to provide keto alcohols 227 and 228, along with a larger amount of phenol 225. Isomer 227 is obtained in 10% yield by crystallization of the crude product from ether. After catalytic hydrogenation, the secondary alcohol is protected and the methoxy group removed reductively. The

224

225

227

226

228

230

Scheme 23.

Buchi-Chu Synthesis of

229

211

(*:)-Gymnornitro1

Brldeed Systems

427

synthesis is completed by Wittig methylenation and deprotection. Overall, the Biichi-Chu synthesis requires nine steps and proceeds in 2.4% overall yield. Welch and Chayabunjonglerd synthesized gymnomitrol as shown in Scheme 24.474 Enone 231, obtained by straightforward annelation of 2-methylcyclopentanone, is allowed to react with lithium dimethylcopper and the resulting enolate is alkylated to obtain ketone 232.. Methylation of this material occurs from the convex face to give ketone 233, which is transformed into keto ester 234. Dieckman cyclization is accomplished by treatment of 234 with lithium hexamethyldisilazide. The resulting enolate is trapped by reaction with t-butyldimethylsilyl chlaride to give keto ether 235. After reduction of the carbonyl group and protection of the resulting secondary alcohol, the ketone is regenerated to obtain 236. The final carbon is introduced by Wittig methylenation to give, after deprotection, (f) -gymnornitro1 (211). The Welch synthesis is the

'. KOH-EtOH

-

v *

I . (%)2BH

233

3. Jones 4 . CHzNz

A

** W

O

O

M

e

2. H ~ O +

-

LHMDS

I I

+Si-CI

234

I . Ph3P=CHe

I . NoBHq

2. 0 OMO.H+ 3. ( ~ - B u ) ~ N +* F-

232

23 I

23%

I

*

236

Scheme 24.

*

Welch-Chayabunjonglerd Synthesis of

211

(*1-Gymnornitro1

428

Tricarbocyclic nud Tetracarbocyclic Sesquiterpeaes

highest-yield route thus far developed, providing 211 in 7% overall yield for the 14 steps from 2-methylcyclopentanone. Paquette's synthesis of gymnomitrol (Scheme 2 9 4 7 5 is in many respects similar to the Coates synthesis (Scheme 22). As in the Coates synthesis, the starting point is the Weiss diketone (2121, which is deoxygenated by Wolff-Kishner reduction of the monoketal. The third ring is added by conjugate addition of 2-trimethylsilylvinylmagnesiumbromide to enone 237, followed by methylation of the resulting enolate. Peracid oxidation of 238 provides keto aldehyde 218. As was also noted by Coates and co-workers, Paquette and Han found 218 to be resistant to aidolization. However, conditions were found whereby 218 can be converted to the extent of 43% into aldol 220. The remainder of the Paquette-Han synthesis is identical to the Coates synthesis. Overall, the synthesis involves 11 steps and gives 211 in 2% yield. I . (CHpOH 12. H + 2. o l f t -Kirhner 3. W H1O+

212

-+p*c+ PhNHpMe, CF3COz

21s

2. M e I . HMPT

A

238

4 & 220

+

222

Scheme 25.

237

218

22 I 2. I . reporoteb LiAIH4

223

Paquette-Han Synthesis of

211

(*bGymnomitrol

Bridged Systems

429

(2) Copacomphor, Copaborneol, Copaisoborneol, Copacamphene, Cyclocopacamphene, Ylangocamphor, Ylangoborneol, Ylangoisoborneol, Sativene, Cyclosativene, cis-Sativenediol, Helminthosporal, and Sinularene This group of tricyclic compounds is characterized by the presence of a bicyclo[2.2.l]heptane system with a further three carbon bridge. The “copa” series includes copacamphor (239), copaborneol (2401, copaisoborneol (241), copacamphene (2421, and cyclocopacamphene (243).* The “ylango” series includes ylangocamphor (244), ylangoborneol (2451, ylangoisoborneol (246), ylangocomphene (sativene, 247), and cycloylangocamphene (cyclosativene, 248). In addition to these 10 compounds, all of which have been synthesized, we shall discuss syntheses of cis-sativenediol (2491, and its conversion product helminthosporal (250). Finally, although sinularene (251) is not biogenetically related to the foregoing group of compounds, its similar structure justifies its consideration in this section.

239

24 I

240

242

* The prefixes “copa” and “ylango” are derived from the tricyclic hydrocarbons copaene and ylangene (vide i d r u ) and specify the configuration of the isopropyl-bearing carbon relative to the remainder of the molecule. Many of the compounds to be discussed in this section occur in nature in both enantiomeric forms. However, some syntheses that will be discussed employ optically active starting materials while others use racemic reactants. In cases where a specific enantiomeric series was used, we have illustrated with the correct formulas. In syntheses which used racemic reactants, we have utilized the enantiomeric series which is most convenient for drawing structures.

430

Tricrrboeyclic and Tetracarbocyclic Sesquiterpenes

I

243

244

249

2 50

251

In Volume 2, we outlined McMurry’s synthesis of (fbsativene and he reported an adaptation of his basic ( f ) - c y c l ~ s a t i v e n e .In~ ~1971, ~ approach to the synthesis of (&:)-copacamphene (Scheme 26).477Unsaturated ketone 252, an intermediate in the sativene synthesis,476is epoxidized stereoselectively from the convex face to give epoxy ketone 253. When this material is treated with dimsylsodium in dimethyl sulfoxide,

Bridged Systems

252

A

255

I . srparatr v

2. Hz-PdIC 3. Collins

431

254

2. SOCIZ, 258

Scheme 26.

259

McMurry’s Synthesis of

242

247

(fKopacamphene

intramolecular alkylation occurs to afford keto alcohol 254. Dehydration of 254 gives a 7:3 mixture of 255 and 256, which is separated chromatographically. The crux of the synthesis is the establishment of the correct stereochemistry of the isopropyl group. It is clear from molecular models that catalytic hydrogenation of either 255 or 256 should occur to produce the “ylango” rather than the “copa” configuration at the isopropyl-bearing carbon. Indeed, experiments bore out this prediction. In order to circumvent this problem, ketone 255 was reduced by lithium in ammonia. A 3:2 mixture of epimeric alcohols 257 is obtained, with the exo diastereomer predominating. Hydrogenation of this substance in hexane solution gives an 85:15 mixture of ketones 258 and 259. Introduction of the methylene group gives a similar mixture of (f)-copacamphene (242) and (f1-sativene (247). The stereoselectivity of the

432

Tricnrbocyclic and Tetrncarbocyclic Sesquiterpenes

hydrogeration presumably results from bonding of the secondary hydroxy group to the catalyst surface and is precedented by earlier observations made in the Heathcock copaene synthesis.478 Piers and his co-workers have made extensive contributions to synthetic methodology in the copacamphor and ylangocamphor series. The Piers syntheses of (+)-copacamphor (2391, (+)-copaborneol (2401, and (+)-copaisoborneol (241) is summarized in Scheme 27.479a3e (+)-Carvomenthone is converted into the n-butylthiomethylene derivative 260, which is alkylated with ethyl a-iodopropionate. Alkylation occurs predominantly from the equatorial direction, affording keto acid 261 as a mixture of diastereomers at the side-chain asymmetric center. After reesterification, the keto ester is subjected to Dieckmann cyclization to obtain P-diketone 262. Hydrogenation of the derived enol acetate 263 gives keto ester 264. The next eight steps are expended in a somewhat tedious homologation of 264 to tosylate 267. After oxidation of the secondary alcohol to a ketone, the final ring is closed by intramolecular alkylation. The product, (+I-copacamphor (2391, is converted into (+)-copaborneol (240) by reduction with sodium in ethanol and into (+I-copaisoborneol (241) by reduction with lithium aluminum hydride. For the synthesis of (-hopacamphene and (- )-cyclocopacamphor (Scheme 28),479b7' dione 262 is partially reduced to hydroxy ketone 269. The remaining carbonyl is then subjected to Bamford-Stevens conditions. Reoxidation provides unsaturated ketone 270. The standard homologation sequence provides unsaturated alcohol 272. The derived tosylate is solvolyzed by treatment with silica gel in pentane to obtain (-)-copscamphene (242). Tosylhydrazone 273, prepared from alcohol 272, is converted by methyllithium into a diazo compound which undergoes cycloaddition with the double bond to provide pyrazoline 274. Photochemical extrusion of nitrogen gives (-)-cyclocopacamphene (243). For synthesis of the ylango series, Piers and co-workers condense (+I-dihydrocarvone with ethyl vinyl ketone (Scheme 29). It is well established that such annelations proceed by axial alkylation. Therefore, the methyl and isopropenyl groups in 275 are trans rather than cis, as they are in intermediate 261 (see Scheme 27). Compound 275 is transformed into dienone aldehyde 277 by standard methodology. After

Bridged Systems

433

I . )-&OK,

26 I

260

262 2. I . LiAIH4 Ph3P&OMe

263

%cHo

Ph3PmCHZ

3. H30+ 4 . K2CO3,MtOH

% 266

26 I

267

268

2 39

Scheme 27. Piers' Synthesis of ( ~ 1-Copacomphor, t ( f ) -Copaborneol, and (*I- Copaisoborneol

264

434

Tricuboeyclic and Tebrc8rboeyciic Sesquiterpeoes

262

269

270

I . PhjPAOMI

I . diaiornyl-

3 . K2CO3,MrOH* 4 . Ph3P=CH2

2. H202,OH'*

2 . H30'

boronr

27 I

272

242

I . Collinr

2. P-TaNHNH2

273

Scheme 28.

274

243

Piers' Synthesis of (*1-Copacamphene and (-bCyclocopacamphene

introduction of an axial methyl group with lithium dimethylcopper, the cyclohexenone ring is degraded by ozonolysis to obtain keto acid 279, a diastereomer of 261. The remainder of the synthesis follows lines similar to those used for synthesis of the copa series.

Bridged Systems

-

0

o@y OH

I . Hz-Pd/C

2. barb

0

e7a

276 OAc

I . HCOOEt,

I . Me2CuLi

bore

2. DDQ

2. AcCl

OH"*0

0

OH-

277

279

' . Ph3PcCH2z

2 . diriamylboranb 3. H 2 0 2

435

278

WH A:

28 I

280

~

~

~

~

D

s

w I

282

;$.

244

C a - N H j I L i AIH,

*a 249

Scheme 29. Piers' Synthesis of (-)-Ylangocamphor, (-bYlangoborneo1, and (-bYlangoisoborneo1

246

436

Tricarbocyclic and TetrncarbocyclicSesquiterpenes

For the synthesis of (+)-sativene (247) and (+)-cyclosativene (248), keto aldehyde 281 is elaborated as in the syntheses of (-)-copacamphene and (- )-cyclocopacamphene (Scheme 30) .479d7g

~

H

I . LiAIHq

Ph3P=CH2, O

2.

soc12, CSHSN

281

3. HzO2,OH4 rtepr (Scheme 28)

I

I . p-TsCI 2. S i 0 2

-

247

248

Scheme 30.

Piers' Synthesis of

(+ 1-Sativeneand

(+)-Cyclosativene

Money and co-workers have prepared (*)-copacamphor (239) and (*)-ylangocamphor (244) beginning with (k)-camphorenone (283) as shown in Scheme 31.480 Epoxidation of the double bond gives a diastereomeric mixture of epoxides (284), which undergoes cyclization

Bridged Systems

benzene m-CPBA

*

437

$3

t-BuOK,

0

284

283

HO

285

286

I . separate 2 . p-TsCI, CSHSN

287

288

I . separate 2. Hz-Pt

239

286

Scheme 31.

244

Money's Synthesis of (*)-Copacarnphor and ( *)-Ylangocamphor

when treated with potassium t-butoxide. The resulting keto alcohols 285 and 286, formed in a 1:l ratio, are separated by preparative glpc. Dehydration of isomer 285 affords unsaturated ketones 287 and 288 in a 7:3 ratio. The isomers are once again separated by preparative glpc and the major isomer is hydrogenated to obtain (*1-copacamphor (239). Similar treatment of isomer 286 yields (*)-ylangocamphor (244). Baldwin and Tomesch have synthesized (f)-cyclosativene as shown in Scheme 32.481Diels-Alder addition of 2,3-dimethyl-l,3-~yclopentadiene (generated in situ from alcohol 289) and propynal affords an adduct

Tricubocyclic and Tetracarbocycllc Sesquiterpenes

438

nn

+

HCZCCHO

4

CHO 289

I . LiAlH4

CHO

290

291

*

2. PhsP*Br*

3. H C o C L i

;F3cv9 $ 292

293 A A c

2. I . HCOOEt, H30+

MeeCuLi

~

MOO-

3. A c e 0

OCHZCF3 294

2 96

Scheme 32.

295

297

248

Baldwin-Tomesch Synthesis of

248

(*1-Cyclosativene

(290) which undergoes base-catalyzed addition of ally1 alcohol to give aldehyde 291. After conversion of the formyl group to propargyl, the secondary alcohol is deprotected and tosylated. The cation obtained by solvolysis of tosylate 293 is trapped by the proximal alkyne a-system to form the cyclosativene skeleton. The remaining steps are concerned with introducing the necessary isopropyl group with the proper stereochemistry. After hydrolysis of enol ether 294, the resulting ketone is formylated and converted into enol acetate 295. Reaction of 295 with excess lithium dimethylcopper gives ketone 296, which has the “copa”

Bridged Systems

439

stereochemistry at the isopropyl position. The ketone is reduced and the resulting alcohol converted into a mesylate which is reduced by Masamune’s method. The product of this reduction is a 3:2 mixture of alkene 297 and (*)cyclosativene. Catalytic hydrogenation of the mixture gives pure (*hyclosativene. Bakuzis and co-workers have reported the synthesis of ( fhsativene and (*bcopacamphene which is summarized in Scheme 33.482 Methylation of the known bromo ketone 298 gives ketones 299 as a 3:l mixture of exo and endo isomers. The magnesium enolate is alkylated with ally1 bromide to give a 3:4:1 mixture of ether 300 and ketones 301 and 302.

298

2 99

300

30I

302

303

I . NCS.CC14

2. CU’~, HpO

36OC 305

304

306

Scheme 33.

307

247

242

Bakuzis’ Synthesis of W-Sativene and (*)-Copacamphene

440

Tricrrbocyclic and Tetrrcarbocycllc Sesquiterpenes

After separation, the major product is converted into aldehyde 304 by the interesting method illustrated. Bromide 305 is treated with trin-butylstannane to give a 3:2 mixture of diastereomeric ketones 306 and 307, which are converted by McMurry’s method476 into ( fbsativene (247) and (f1-copacamphene (242). The noteworthy feature of this synthesis is use of free-radical technology in two different transformations (301-303 and 305 306 307). The synthesis by Matsumoto and co-workers of (fbsativene and ( *1-copacamphene is shown in Scheme 34.483 Racemic piperitone is irradiated with 1,l-dimethoxyethylene to obtain a mixture of diastereomeric adducts 308 When these adducts are passed through a glpc column containing the weakly acidic solid phase Shimalite at 200°C both lose methanol and rearrange to the same 5545 mixture of 309 and 310. When this mixture is subjected to Wittig reaction with methoxymethylenetriphenylphosphorane only isomer 310 reacts, leading to a mixture of 309 and diether 311, which can be separated by preparative glpc. Alternatively, the mixture can be hydrolyzed and the products (keto aldehyde 312 and a diketone derived from hydrolysis of 309) separated by silica gel chromatography. The remainder of the synthesis is very similar to the route employed by Piers (see Scheme 30). Under more vigorous conditions, the less reactive diastereomer 309 also undergoes Wittig reaction. The resulting keto ether may be converted by the same sequence of steps into ( f )-copacamphene (242). Piers, McMurry, and Matsumoto have all adapted their syntheses in order to prepare the interesting antibiotic cis-sativenediol (249). The Piers synthesis (Scheme 35)484 begins with unsaturated alcohol 285 (see Scheme 301, which is oxidized to a keto aldehyde. Intramolecular Prins reaction occurs when this material is treated with trifluoroacetic acid. The resulting alcohol is oxidized by Collins reagent to give unsaturated ketone 314. In fact, the overall conversion of 285 to 314 can be carried out without workup at the intermediate stages. Hydroxylation of the enolate of 314 with Vedejs’ reagent occurs from the exo face of the bicyclo I2.2.11heptane system, leading to hydroxy ketone 315. Reduction of this material affords an equimolar mixture of (*)-cis-sativenediol (249) and its trans diastereomer 316.

-

+

Bridged Systems

441

300

-%+ -7 Ph3P&OMe

309

Me

*OMe

OMe

*

303

+

310

I . H30+

CHO

2. separate

I . Ph3PfiOMe

2 . H30+

OMe

31 I

*

TCHO

3. base

312

309

Scheme 34.

313

242

Matsumoto's Synthesis of (rtbsativene and (*t)-Copacamphene

442

Tricarbocyclic and Tetracarbocyclic Sesquiterpenes LDA, OH

I

Collinsc 2 . TFA 3 Collins

314

285

+

315

249

316

Scheme 35.

Piers’ Synthesis of (*I- ciPSativenedio1

McMurry’s synthesis (Scheme 36)485 begins with hydrazone 317, an intermediate in the McMurry sativene synthesis.486 The hydrazone is cleaved using titanous ion in 97% yield. After introduction of a double bond by the Reich-Sharpless procedure, the secondary alcohol is tosylated. Internal alkylation of 320 gives the cyclopropyl enone 321, which undergoes vinyl cyclopropane rearrangement to enone 322 when passed through a quartz tube at 450°C. Osmylation occurs from the exo face of the bicyclo[2.2.llheptane system to give a diol (323) which is converted into acetonide 324. After introduction of the methylene group, the acetonide is hydrolyzed to obtain (*I- cis-sativenediol. Matsumoto’s synthesis of (*)-cis-sativenediol and (*)-helminthosporal is summarized in Scheme 37.483 Keto aldehyde 312 is converted into an unsaturated ketone by selective Wittig methylenation of the more reactive aldehyde carbonyl. Dibromide 325 (presumably a 1:l mixture of diastereomers) undergoes internal alkylation when treated with potassium r-butoxide to give bromo ketones 326. Further treatment of

Bridged Systems

317

318

319

32 I

320

443

>-

322 080,

249

Scheme 36.

324

McMurry's Synthesis of

1

323

(*)cidativenediol

326 with base affords unsaturated ketone 322, an intermediate in the McMurry synthesis. This material is converted into (*)-cis-sativenediol and (*1-helminthosporal (250) by the straightforward methods shown in the scheme. The final synthesis we discuss in this section is the Collins-Wege synthesis of (f)-sinularene (Scheme 38),486 which is patterned on Money's synthesis of copacamphor (see Scheme 31). Diels-Alder adducts 328 and 329 are treated with concentrated sulfuric acid to obtain lactone 330, which is converted into unsaturated ketal 333 by precedented procedures. The ketal is hydrolyzed, the double bond is epoxidized, and ring closure effected to obtain a 1:l mixture of diastereomeric hydroxy ketones (see

444

Tricarbocyclic and Tetracarbocyclic Sesquiterpenes I . Ph3P=CH2 c 2 . Br,

312

r

t-BuOKw

325

326

322

249

2 . MeO-

250

Scheme 37.

Matsumoto’s Synthesis of ( f 1- cisSativenedio1 and (rt )-Helminthosporal

Scheme 31). These are separated and the exo isomer is dehydrated by pyrolysis of its acetate at 450°C. Unsaturated ketones 335 and 336, obtained in a ratio of 9 5 5 , are separated by preparative glpc. Catalytic hydrogenation of the major isomer gives nor-ketosinularene (3371, which is converted into ( f h i n u l a r e n e by the method indicated.

Bridged Systems

H

333

334

+

@y

335

336

Scheme 38.

2. 1 . reparate, H2-Pt02

445

I . separate 2. AcCI 3 . 450'

@.< 337

;:;* 3. 450'

Collins-Wege Synthesis of (*t)-Sinularene

446

Tricarboeyclic ind Tetracrrboeyclic Sesquiterpenes

(3) Longgolene, Longiwclene, Longicamphor, and Longiborneol The “longi” tricyclic sesquiterpenes have a typical four-carbon bridge which bears a gem-dimethyl group. The ones we discuss in this section are longifolene (3381, longicamphor (339), longiborneol (3401, and longicyclene (341). The pioneering effort in this series was Corey’s synthesis of ~ o n g i f o ~ e n e . ~ ~ ’

338

339

340

34 I

In 1972, McMurry and Isser reported a second synthesis of (*)-longifolene (Scheme 39).488 The first four steps, beginning with the Wieland-Miescher ketone (3421, parallel McMurry’s sativene, 476 and copacamphene syntheses (see Scheme 26). Unfortunately, dehydration of the methyl carbinol derived from ketone 343 gives only a 7:3 mixture of unsaturated ketones 344 and 345. The major isomer is epoxidized and the resulting keto epoxide cyclized to obtain hydroxy ketone 347, which undergoes acid-catalyzed dehydration to 348. The cyclohexene ring is expanded via dibromocyclopropane 349 and bromo alcohol 350. After dehalogenation and oxidation of 351 to an enedione, the gemdimethyl unit is completed by coqjugate addition of lithium dimethylcopper. The initially formed enolate, 351% undergoes intramolecular aldolization to give 352. This unexpected event is actually used to

Bridged Systems

447

k 342

343

345

344

346

Br

AgC104

"9 '6 .'"i$ 9 347

348

_____)

acetone, H20

349

Li-NH3/MeOH,

3 50

HO

352

311

I . NaBH4 2. MsCl

I-BuOK

(Ph3P)sRhCI H2

___.)

~

H6

353

Scheme 39.

354

McMurry-Isser Synthesis of (rtbLongifolene

Tricarbocycllc 8nd TetracnrbocyclicSesquiterpeoes

448

advantage, since the free carbonyl can readily be reduced and esterified to obtain the diol monomesylate 353. Grob fragmentation of this material regenerates both the carbonyl group and the cycloheptane ring. The remaining steps to longifolene are straightforward adaptations of known methodology. In all, McMurry's synthesis entails 18 steps, somewhat more than are required in the Corey synthesis (14 steps). The overall yield is on the order of 5-6%. Johnson and co-workers have reported the very unusual longifolene synthesis which is summarized in Scheme 40.489 The cuprate reagent prepared from acetylenic iodide 356 is added to 24sopropylidinecyclopentanone and the resulting enolate is acetylated to obtain enol acetate 357. Bromination-dehydrobromination gives enone 358, which is reduced to allylic alcohol 359. n-Cyclization occurs when 359 is treated with trifluoroacetic acid in ether at 0°C to give, after hydrolytic workup,

.

I . t-BuLi 2. ( ~ - B U ] ~ P * C U I ~

3. 356

$ 359

ether,

00

I

f_

. gH

4. AcCl

I

AG2 6 I

357

358

$

co? MeqN

NaZnBrZ. B HSCN

LIAIH~+

p-TIOH,

CF3C00H

36 I

360

362

Scheme 40.

355

Johnson's Synthesis of

338

(fbLongifolene

Bridged Systems

449

tricyclic alcohol 360. This remarkable transformation presumably occurs by the following mechanism:

The unwanted bridgehead hydroxy is removed by taking advantage of the ease of solvolysis of 7-norbornenyl systems. Thus, when 360 is treated with sodium cyanoborohydride in the presence of zinc bromide, alkene 361 is formed in 94% yield. Treatment of 361 with p-toluenesulfonic acid causes isomerization of the double bond to the exocyclic position. Oxidation of the double bond gives a ketone, which is methylated to obtain 355, an intermediate in both the Corey and McMurry syntheses. The Johnson synthesis requires 11 steps and proceeds in almost 20% overall yield. An even more efficient longifolene synthesis was reported in 1978 by Oppolzer and Godel (Scheme 41).490 Acylation of the morpholine enamine of cyclopentanone with 3-cyclopentenecarbonyl chloride gives an unsaturated P-diketone (363). Acylation of this compound with benzyl chloroformate gives an enol ether (364) which is photolyzed to obtain the tetracyclic keto ester 365. Hydrogenolysis of the benzyl group gives a hydroxy ketone which undergoes in situ reverse aldolization to afford diketone 366. The remaining six steps are given over to introducing the other four carbons of longifolene. The gem-dimethyl groups are installed by Wittig methylenation of the more reactive carbonyl, Simmons-Smith cyclopropanation, and catalytic hydrogenolysis. The synthesis is then completed as in the Johnson synthesis. The overall yield for the 10 step synthesis is an impressive 26%. Welch and Waters have synthesized (*1-longicyclene (3411, (A)-longicamphor (339) and (*)-longiborneol (340) as shown in Scheme 42.491 The synthesis begins with (f)-tetrahydroeucarvone (370), which is readily prepared from (- I-carvone as shown. Alkylation with 4-chloro2-pentene gives unsaturated ketone 371 in 80% yield. Cleavage of the

450

Tricarbocyclic and Tetracarbocyclic Sesquiterpenes

363

365

Scheme 41.

364

366

367

Oppolzer-Godel Synthesis of (d-Longifolene

double bond gives keto acid 372 (mixture of diastereomers) which is cyclized to enol acetate 373. Reduction with diisobutylaluminum hydride gives bridged ketol 374 (see also Scheme 22 and accompanying discussion). After dehydration of 374, the resulting unsaturated ketone is manipulated more or less as in the Piers approach to copacamphene and cyclocopacamphene (see Scheme 28). The overall yield for the .I9steps from (-)-cawone to (It)-longicyclene is 8-10'36. For the synthesis of (*)-longiborneol, 23 steps are required, resulting in an overall yield of about 9%.

Bridged Systems

369

37 I

370

372

374

373

376

375

339

340

Welch-Waters Synthesis of (*)-Longicyclene, Scheme 42. (*)-Longicamphor, and (*)-Longiborneol

451

452

Tricarbocycllc and Tetracarboeyclic Sesquiterpenes

(4) Copaene, Ylongene, and Long@inene a-Copaene (380), 8-copaene (3811, a-ylangene (38J, 8-ylangene 3831, a-longipinene (3841, and 8-longipinene (385) are sesquiterpene analogs of the common monoterpenes a- and 8-pinene. In Volume 2 of this

380

38 I

382

383

304

385

series, we discussed a synthesis of a-copaene and a-ylangene in which skeleton is formed by base-catalyzed ring the tricycl0[4.4.0.0~~~]decane closure of a decalinic keto t ~ s y l a t e : ~ ~ ~

In 1973, Corey and Watt reported a similar synthesis of these sesquiterpenes (Scheme 43).493 The novel feature of the Corey-Watt synthesis is the method used to synthesize the decalinic intermediate 387. Inverse electron demand Diels-Alder reaction of pyrone 386 and 4-methyl3-cyclohexenone gives adduct 387 in yields of up to 40%. The reaction is believed to involve the intermediate enol form of the ketone, since several other 1-methylcyclohexene derivatives do not react with pyrone 386. Compound 387 is transformed by a straightforward sequence of steps into dienone 389, which is reduced with lithium in ammonia.

Bridged Systems

0

453

2 . m-CPBA

OOMe

386

387

300

I . LIAIH,

2 ClCOOEt pyrtdine 3. DMS0,Ac20

389

No H 2. Mg,MeOH

3 . p-TSNHNH2 4 . NaBH,

CN

392

I . Zn, HOAc ~

309

391

390

2. Na, 3. p - T t C I 4 . H30'

~

Tb & 300

302

Me!CH;NaL

393

394

6 steps

30 I

Scheme 43.

303

Corey-Watt Synthesis of (*)-Copaene and (*)-Ylangene

After esterification with p-toluenesulfonyl chloride and deprotection, keto tosylate 390 is obtained. Cyclization of 390 is accomplished in the same manner as was previously used in the Heathcock synthesis.492 Unfortunately, tricyclic ketone 391 is obtained in only 10-38% yield. The isopropyl group is introduced by Wadsworth-Emmons reaction of 391 with a-diethylphosphonopropiononitrile, followed by reduction of the

454

Trlcarbocyclic and Tetracarbocyclic Sesquiterpenes

d - & h0 & 5 rtepr

I . AcS,

NaBHt

2. CrOg

0

OH

HOAc

395

342

397

396

m-CPBA CH2Cl2

___)

398

399

404

403

406

405

384

Scheme 44.

385

Yoshikoshi’s Synthesis of

(*)-Longipinene

Bridged Systems

455

conjugated double bond with magnesium in methanol. The nitrile group is then reduced to methyl. The final product is a 1:l mixture of (f)-a-copaene (380) and (*)-a-ylangene (381). Zinc-acetic acid reduction of 389 gives an exocyclic double bond isomer of 390 (393) which is cyclized to unsaturated ketone 394. The latter substance is transformed by an identical sequence of steps into a 1:l mixture of (fbp-copaene (381) and (*)-@-ylangene (383). A disadvantage of the Corey-Watt synthesis is that it allows no control over the stereochemistry of the isopropyl group, in contrast to the Heathcock synthesis, which can be carried out so as to select either copaene or ylangene. Yoshikoshi’s synthesis of the longipinenes (Scheme 44)494 starts out with the Wieland-Miescher ketone (3421, which is transformed into keto acetate 396 by well-established procedures. Allylic alcohol 397 is epoxidized and the resulting epoxy alcohol transformed into hydroxy mesylate 399. Treatment of this material with lithium-aluminum hydride results in Grob fragmentation and reduction of the resulting ketone. Compound 400 is transformed into dienone 401 and thence into triene 402. Sensitized photoisomerization of 402 gives tricyclic alkene 403 in 75% yield.49S The remaining steps are concerned with expanding the fivemembered ring and introducing the final carbon. The final product is a 5:3 mixture of (*)-a-longipinene and (*)-P-longipinene. The overall yield of the two terpenes is 6% for the 20 steps from the known keto ester 396. (5) Isocyunopupukeanunes

9-Isocyanopupukeanane (407) and 2-isocyanopupukeanane (408) are substances which are produced by a sponge and apparently serve certain nudibranchs as defensive substances. Compounds 407 and 408 present several interesting synthetic challenges. First, there is the problem of generating the interesting tricyclo[4.3.1.03371decanenucleus. There is also the problem of the isopropyl group, which has the thermodynamically unfavorable endo stereochemistry. Finally, there is a problem in

456

Tricarbocyclic and Tetracarbocyclic Sesquiterpenes

407

408

establishing the correct stereochemistry of the isocyano groups in both compounds. Two synthetic approaches have been published. Corey’s synthesis of 9&isocyanopupakeanane is shown in Scheme 45.496 Unsaturated ester 409 undergoes conjugate addition of isopropylmagnesium chloride to give compound 410, which is cyclized to indanone 411 by treatment with polyphosporic acid. Ketone 411 is converted into ester 412 by the p-toluenesulfonylmethyl isocyanide method. Methylation of 412 gives a 6:l mixture of diastereomers 413 and 414. This is an important step as it is here that the eventual endo configuration of the isopropyl group is established. After cleavage of the methoxy groups the resulting phenolic acid is hydrogenated over Nishimura’s catalyst to give lactone 415 in 37% yield. The stereoselectivity of the reduction presumably results from absorption of the face of the aromatic ring opposite the isopropyl group. Keto tosylate 416, prepared from 415 by a straightforward method, is cyclized by treatment with base to give ketone 417. Hydrogenation of the derived oxime occurs from the face opposite the angular methyl group to provide amine 418 in 80% yield. Overall, the Corey synthesis provides isonitrile 407 in 3.6% yield for the 16 steps from unsaturated ester 409. Yamamoto’s synthesis of 9-isocyanopupukeanane is shown in Scheme 46.497 Allylic alcohol 420 is converted by Claisen rearrangement into unsaturated aldehyde 421. This material is further transformed into triene 424 which undergoes Diels-Alder reaction when heated to 160°C. After deprotection, hydroxy ketone 425 is obtained in quantitative yield. The ketone is protected and the secondary alcohol oxidized to obtain ketone 426. In the Yamamoto approach, the isopropyl stereochemistry is

Bridged Systems

409

410

OMe

2. KOH, H202,

+

O

M

I

+

412

MeOOC &OM'

&I. LiAIH4

I. BarS

~

2. H 2 - I r

2 413

LDA,MeI

___)

HMPT

H20, EtOH 3. CHzN2

411

G

457

414

*

2. P-TsCI 3. PCC

41 5

OTs

&xo

-4

I. HzNOH

I -BuOK

2. H 2 - I r b

416

417

418

I . AcOCHO

2. MsCI.

407

Scheme 45.

Cotey's Synthesis of (+)-9-Isocyanopupukeanane

Tricarbocyclic and Tetracrrbocyclic Sesquiterpenes

458 0

OH

419

4 20

421

o*

OTHP

422

OTHP

Me3SiCl LDA

423

425

OTHP

424

426

418

Scheme 46.

. Me3si0m

Yamamoto’s Synthesis of

427

407

( f 1-9-Isocyanopupukeanane

established by catalytic hydrogenation of diene 427, which occurs exclusively from the more accessible exo face. The isonitrile stereochemistry is established by the same method as is used in the Corey synthesis. The Yamamoto synthesis of 407 requires 18 steps and proceeds in 6% overall yield.

Bridged Systems

459

For the preparation of 2-isocyanopupukeanane, Corey and Ishiguro transformed lactone 415 into keto aldehyde 428 (Scheme 47).498 Intramolecular aldolization of these materials affords a mixture of diastereomeric aldols 429 in 98Oh yield. The keto group is removed by Raney nickel desulfurization of the derived dithioketal and the hydroxy group oxidized to obtain ketone 430. The amino group is introduced as in the earlier Corey synthesis. However, in this case hydrogenation of the oxime gives a 1:l mixture of two diastereomers. The isomers are separated after formation of the formamides. Isomer 431 is dehydrated to ( f )-2-isocyanopupukeanane (408).

& 4

.

I. LiAIH4,

I , ICH2SH)pr

THF 2 . PCC

BF3 2 . Raney Ni 3. PCC

* H$

415

4 29

428

?

I . H2NOH

H

OH@(

+

2I . separate MsCI,

2 . Hp-Pt/Rh 3. AcOCHO

430 H

CsYN

H

431

432

cq H

406

Scheme 47.

Corey's Synthesis of (~)-2-Isocyanopupukeanane

*

460

Tricnrbocyclic and Tetracnrbocyclic Sesquiterpenes

(6) Patchouli Alchohol and Seycheilene

Patchouli alcohol (433) the major component of patchouli oil, is one of the longest known terpenes. Its congeners, seychellene (4341,and norpatchoulcnol (435) have the same tricyclic skeleton. Previous syntheses of 433 by Buchi and Danishefsky and of 434 by Piers were discussed in

433

435

434

Volume 2 of this series.499Full details of the Piers seychellene synthesis have subsequently appeared in print.500 In 1970 Schmalzl and Mirrington reported a synthesis of (*) - s e y ~ h e l l e n e , ~which ~ * ~ ~was ~ later adapted for a synthesis of (*)-patchouli alcohol.501 The patchouli alcohol synthesis (Scheme 48) is similar in many respects to the Danishefsky synthesis.502 The synthesis begins with Diels-Alder reaction of 1,3-dimethyl1,3-cyciohexadiene and methyl vinyl ketone, which gives unsaturated ketone 436 as the major isomer of an 8:l mixture. Reformatsky reaction followed by acetylation of the initially formed complex affords a mixture of P-acetoxy esters (437)which is dehydrated by treatment with sodium ethoxide. The resulting diastereomeric mixture (438)is reduced with lithium in ammonia to give a 4:l mixture of diastereomeric alcohols 439 and 440. Schmalzl and Mirrington explain the good stereoselectivity observed in this reduction by assuming that the preferred conformation of the intermediate radical anion is 43811 and that protonation occurs from the face away from the angular methyl group. After separation of

'Lj----H' EtO 4380

-

Et

$d.

-

439

Bridged Systems

&+a,-&

t

. Br*COOEt,

461

Zn

2 . ACCI. PhNMbE A 0

*

%Ac

COOEt

437

436

438

439

440

433

Scheme 48.

Mirrington-Schmalzl Synthesis of ( f )-Patchouli Alcohol

439 and 440, the synthesis of patchouli alcohol is completed by conversion of 439 into a ketone (4411, which is then methylated to obtain 442. The derived iodide 443 is cyclized by heating with sodium in THF at 100°C; racemic patchouli alcohol (433) is obtained in about 40%yield.

462

Tricarbocyclic and Tetracarbocyclic Sesquiterpenes

This basic approach has also been applied to a synthesis of ( * h e y chellene as shown in Scheme 49. Unsaturated alcohol 439 is acetylated and hydrochlorinated to obtain 444 which is dehydrochlorinated by treatment with sodium ethoxide. Somewhat surprisingly, the exocyclic alkene 445 constitutes no more than 60% of the mixture of alkenes produced. More hindered bases (e.g., potassium triethylmethoxide) were found to give even more of the endocyclic isomer 439. After separation of 445 from 439 by chromatography on silver nitrate-impregnated alumina, the primary alcohol is protected and the alkene oxidized by the Lemieux-Johnson method. After deprotection, the primary alcohol is esterified with p-toluenesulfonyl chloride. Cyclization is accomplished by treating the resulting keto losylate with potassium triphenylmethide. Ketone 448 has previously been converted by Piers into (*I-seychellene. SO3 ~

9 I . Ac~O

2. HCI

HO

AcO

HO

444

439

445

I . separate

2. PhsCCI 3. 0104, NaIO,

* 3. Ph3C-K+

441

446

448

Scheme 49.

434

Mirrington-Schmalzl Synthesis of

(*Meychellene

Bridged Systems

463

Yoshikoshi and co-workers have prepared ( fheychellene by a route incorporating an intramolecular Diels-Alder reaction as the key step in constructing the tricyclic skeleton (Scheme 50).'04 Alkylation of 2,3-dimethylcyclohexenonewith 3-methyI-1,S-diiodopentaneand sodium amide in ammonia gives unsaturated ketone 449 as a 1:l mixture of diastereomers in 35% yield. Treatment of this material with two equivalents of N-bromo-succinimide in carbon tetrachloride provides bromodienone 450 which is reduced with chromous chloride. Elimination of bromine is accompanied by deconjugation. The diene system is conjugated by treatment of 451 with acid. The third double bond is introduced by Cope elimination of the dimethylamine oxide at 430°C in a glpc apparatus. The intermediate trienone (453) undergoes in siru Diels-Alder cyclization affording unsaturated ketone 454. Hydrogenation of 454 gives ketone 448, which is transformed by the Piers procedure into (f)-seychellene.

NaNH2.NH3

449

FYI

450

p-TsOH

I

EtOH

451

2. H202 3. 430'

452

454

448

Scheme 50.

Yoshikoshi's Synthesis of

453

434

(*)-Seychellene

464

Tricarboryclic and Tetracarbocyclic Sesquiterpenes

Naf and Ohloff have reported the interesting and economical synthesis of ( f )-patchouli alcohol shown in Scheme S1.’05 Cyclohexadienone 455, the starting material for Danishefsky’s synthesisSo2 is alkylated with 3-methyl-Cpentenyllithium to obtain a 1:1 mixture of diastereomeric alcohols 456 and 457 in 59% yield. When a decalin solution of this mixture containing 5% potassium f-butoxide is heated in a sealed tube for 24 hours, alcohol 458 is obtained in 60% yield based on isomer 457. The intramolecular Diels-Alder reaction appears to be remarkably stereospecific, since no adduct derived from diastereomer 456 is

L

L

i

c

6 6

decalin. 280°

+

I-BuOK

4 55

456

458

Scheme 5 1 .

457

433

Naf-Ohloff Synthesis of (*)-Pathcouli Alcohol

obtained. The role of the base in the Diels-Alder reaction is not understood, although it appears to be necessary. It has been suggested that this may be an example of “alkoxide-accelerated cycloaddition,” analogous to the well-known alkoxide-accelerated sigmatropic rearrangements. The base may also serve to catalyze interconversion of 456 and 457 by alkoxide-accelerated electrocyclic reaction:

456

ent-457

Bridged Systems

465

Fra’ter has also employed an intramoleculer Diels-Alder reaction in a synthesis of (*)-seychellene (Scheme 52).’06 In Frdter’s synthesis, phenoxide 459 is alkylated with bromide 460 to obtain a 1:l mixture of diastereomers 461 in 80% yield. When this mixture is heated at 80°C, cycloaddition occurs to give a 3:l mixture of tricyclic ketones 462 and 463 in 60% yield. Catalytic hydrogenation of the major isomer affords a 2:l mixture of saturated ketones 464 and 465. When methyllithium is

459

462

46 I

460

463

464

466

Scheme 52.

Friter’s Synthesis of

465

434

( f 1-Seychellene

added to this mixture only 465 reacts, giving alcohol 466 in 97% yield (based on isomer 465). Solvolytic rearrangement of this material gives (fbseychellene in 58% yield. Even though isomers are produced in two key steps, this short synthesis still provides seychellene in 7% overall yield. Jung and McCombs have synthesized (*t)-seychellene as shown in Scheme 53.’” Enone 467 is transformed into the trimethylsilyloxycyclohexadiene 468 which undergoes regio- and stereoselective Diels-Alder

Tric8rbocyclic rod Tetr8e8rboeydic fksquiterpenes

466

?!%+

471

470

472

H

d

474

473

434

Scheme 53.

Jung-McCombs Synthesis of

(*1-Seychellene

reaction to give adduct 469 in 75% yield. Homologation of this material gives allylic alcohol 471 which is hydrogenated to a 1:l mixture of hydroxy ketones 472 and 473. These compounds are converted into bromo ketones, which undergo cyclization to a mixture of diastereomeric ketones. After purification by silica gel chromatography the correct isomer is converted into (*)-seychellene by Piers' method. The Jung synthesis, even though stereorandom, requires only 10 steps and gives 434

Bridged Systems

467

in about 10% overall yield. A similar approach has been applied by Spitzner in a synthesis of 7-desmethyl-l l-norseychellanone (475).'08

475

Yamada and co-workers have synthesized (*heychellene and (f)-patchouli alcohol by a route involving a common intermediate.'09 For the synthesis of seychellene (Scheme 54) the known keto acid 476 is

477

416

I . Li-NH1 2. H 3 0 + 3. NoOMe

Me

I . H30+ 2. I - B u O K ) 478

479

480

448

Scheme 54.

481

434

Yamada's Synthesis of (*.)-Seychellene

Tricarbocyclic and Tetracarbecyclic Sesquiterpenes

468

transformed into 477 by Wittig methylenation and hydrogenation. After conversion of the carboxy group to a protected aldehyde, the aromatic ring is reduced by the Birch procedure to obtain enone 478 as a mixture of diastereomers. After deacetalization, the enone aldehyde is treated with potassium t-butoxide. Diastereomeric hydroxy ketones 479 and 480 are obtained in 23% and 3% yield, respectively. The major isomer is deoxygenated by Raney nickel desulfurization of the dithioketal. Alcohol 481 is oxidized with chromium trioxide in pyridine and the resulting ketone is methylated to obtain 448, an intermediate in the Piers seychellene synthesis. For the synthesis of (*)-patchouli alcohol (Scheme 5 5 ) , Yamada and co-workers oxidize the ketone derived from 481 by the Vedejs procedure. After protecting the alcohol as the methyl ether, the gemdimethyl group is introduced via alkene 483 and spirocyclopropane 484. The protecting group is removed by oxidation with chromium trioxide in acetic acid. The Yamada syntheses are considerably longer and give lower overall yields than previous syntheses of both seychellene and patchouli alcohol. For seychellene, 12 steps are required to give the terpene in 5% yield. Patchouli alcohol is obtained in only 0.5% yield in 16 steps.

48 I

482

484

Scheme 5 5 .

483

485

433

Yamada's Synthesis of (*)-Patchouli Alcohol

Bridged Systems

469

The interesting norsesquiterpene norpatchoulenol (435) has been found to be the compound actually responsible for the characteristic scent of patchouli alcohol. It was first synthesized by Teisseire by the lengthy route summarized in Scheme 56.510 Addition of methylmag-

486a

___._c I . LiAlH4

&o:

Ph:Cl pyridine. ~

2. separate &OH H

4;:"'

li 487

486 b

H

ti

488 I . 8pHg

NaH

Ph C-K+ s

-.cp,, 2. CrO

___c

PhCHpCl

H

H

489

-

acetone

'&. -

L i , EtNHz * ''0-P

490

Collins

'0 H

h

491

492

I . LiAIH, 2. p-BrCgH4COCI

493

Scheme 56.

Teisseire's Synthesis of

435 ( *1-Norpatchoulenol

470

Trlcarbocycllc and Tetracarborycllc Sesquiterpeoes

nesium iodide to 3-methylcyclohexenone followed by acid-catalyzed dehydration of the resulting alcohol gives a 5545 equilibrium mixture of dienes 486n and 486b Treatment of this mixture with refluxing acrolien gives Diels-Alder adduct 487 and unreacted diene 486b, which is subjected to acidic equilibration to regenerate a mixture of 486a and 486b Reformatsky reaction of 487 gives a 3:2 mixture of p-hydroxy esters which is reduced to a similar mixture of diols. In order to work with homogeneous substances, the mixture is separated by crystallization of 488, the major isomer. The two hydroxy groups are separately protected, to obtain unsaturated diether 489, which is hydroborated with oxidative workup to obtain ketone 490. After methylation and removal of the trityl group, the primary alcohol is converted into a tosylate and thence into an iodide. The resulting halo ketone is cyclized by the same method employed by Danishefsky in his patchouli alcohol synthesis to obtain 491. Debenzylation provides a diol which is oxidized to keto alcohol 492. Reduction of this compound provides the epimeric keto alcohol, which is converted by pyrolysis of the p-bromobenzoate into (f)-norpatchoulenol (435). Oppolzer and Snowden have synthesized (f)-norpatchoulenol by the interesting route shown in Scheme 57."' Dienone 455 is allowed to react with 3-triethylsilyloxypentadienyllithium (494) and the initial adduct silylated to obtain 495. After desilylation of the enol ether with KF in methanol, trienone 496 is obtained in 47% overall yield. When a benzene solution of 496 is heated in a sealed tube at 230°C a highly regioselective Diels-Alder cyclization occurs to produce unsaturated keto alcohol 497. After saturation of the double bond, 498 is converted into (*bnorpatchoulenol by Shapiro reaction. Gras has reported an approach to the synthesis of norpatchoulenol ' ~ method has which is based on the Piers seychellene ~ y n t h e s i s . ~The not yet been used to synthesize the natural product, but has been applied in a synthesis of desoxynorpatchoulenol as shown in Scheme 58. Oxidation of the readily-available phenol 499 gives a benzoquinone (500) which undergoes regio- and stereospecific Diels-Alder reaction with 2,4-pentadienol. The resuiting adduct (501)is converted into enone 504

Bridged Systems

KF

Me3SiCl

MeOH

OSiEt3 455

OSiMe3

471

494

495

I . 230.

benzene

2. H ~ O +

496

497

4913

I . P-TSNHNH~ 2. MeLi

435

Scheme 57.

Oppolzer-Snowden Synthesis of

(*bNorpatchoulenol

by methodology originally introduced by Woodward in his pioneering steroid synthesis. The gem-dimethyl unit is completed by conjugate addition of lithium dimethylcopper to 504. After replacing the tetrahydropyranyl group with tosyl, the third ring is closed using dimsylsodium as base. Wolff-Kishner reduction of tricyclic ketone 507 provides racemic desoxynorpatchoulenol (508).

Tricarbocyclic and Tetracsrbocyclic Sesquiterpenes

472

@

Me0

OH

500

499

1.

Q,n+

____)

2 . NaBH,, E t OH ,.Oo

Me0

501

b - b I. A c ~ O . OMAP 2. Li-NHI. t-BuOil 3. NH4CI

Me0

‘OTHP

503

Me CuLi 2

I. H30+

2. p-TICI

0

‘OTHP

‘OTHP

504

J3

0

P

505

DMSO

‘OTr

506

507

Scheme 58. Gras Synthesis of (*)-Desoxynorpatchouli Alcohol

508

MeLi

Bridged Systems

473

(7) Zizane (tricyclovetivene), Zizanoic Acid, Epizizanoic Acid, and Khusimone Vetiver oil contains a group of methanohydroazulenes which includes zizaene (tricyclovetivene, 5091, zizanoic acid (SlO), epizizanoic acid (512), and khusimone (514). These four natural products and the unnatural isomers of epizizaene (511 and 513) have yielded to synthesis. Yoshikoshi's synthesis of epizizanoic acid, which was discussed in Volume 2 of this series, has now been reported in full.509

509: R = Me 510: R = COOH

511: R = M e 512: R = COOH

613

61 4

474

Tricarbocyclic and Tetracarbocyclic Sesquilerpenes

MacSweeney and Ramage have synthesized methyl zizanoate as shown in Scheme 59.'14 The approach is similar in many respects to

620

I . LiAIH4

2 . Jones

824

626

I . MeMgI

2 . CH2N2 3 . P0CIJ, C&N, 86.

HOO

627

Scheme 59.

626

@+

Me00

628

MeOO@

620

MacSweeney-Ramage Synthesis of Zizanoic Acid

Bridged Systems

475

Yoshikoshi’s synthesis of epizizanoic acid, which begins with methyl 1-camphenecarb~xylate.~In the MacSweeney-Ramage synthesis the Grignard reagent derived from bromide 515 is condensed with (+)-camphor to give a mixture of alcohols 516 and 517. The mixture is not separated but is smoothly rearranged by silica gel to provide the arylcamphene 518 in 30% overall yield. The low yield is due to competing enolization in the Grignard addition step. After Lemieux- Johnson cleavage of the methylene group, the resulting ketone is subjected to Birch reduction to obtain 519. Isomerization to the coqjugated diene system is brought about by heating with Wilkinson’s catalyst. Ozonolysis and Jones oxidation provide a diastereomeric mixture of diketo esters which is cyclized by base to obtain a 2:3 mixture of enones 522 and 523. The isomers are separated chromatographically and the minor isomer deoxygenated by desulfurization of its dithioketal. The resulting unsaturated ester (524) is hydroxylated and the derived diol 525 rearranged by Biichi’s methodS16 to obtain keto ester 526. This is a key step in the synthesis. A similar rearrangement was employed by Yoshikoshi and co-workers in their epizizanoic acid ~ y n t h e s i s . ~However, ’~ in that case potassium t-butoxide was employed as the base. The stronger base leads to equilibration of both epimerizable centers. By using more weakly basic conditions, MacSweeney and Ramage are able to realize complete stereospecificity in the rearrangement. In order to bring about saponification of 526 without epimerization, it is reduced with lithium aluminum hydride and the resulting diol is reoxidized with Jones’ reagent. The methylene group is introduced by addition of methyllithium and dehydration of the resulting tertiary alcohol. The product is a 2:3 mixture of methyl isozizanoate (528) and methyl zizanoate (529) which are separated by chromatography. A similar sequence of steps beginning with epimer 523 provides methyl epizizanoate. Alternatively, this epimer may be isomerized to a 2:3 mixture of 522 and 523, thus providing more of the former isomer for conversion into ester 529. In all, the MacSweeney-Ramage synthesis requires 18 steps, also requires two isomer separations (in each case the minor isomer is used), and affords the final product in an overall yield of less than 1%.

Tricarboryclic and Tetracarbocyclic Sesquiterpenes

476

Wiesner and co-workers have synthesized zizaene (509) and its two diastereomers 511 and 513 as shown in Scheme 60.'17 The synthesis begins with indanol 530, which is alkylated via Claisen rearrangement of the ally1 ether. The resulting phenol is methylated to obtain 531, which

I . 0504,

NaC103

2 185. 3 . Me2SO4, NoOH

2. NaI04 3. (CH20H)2,

n+

630

532

631

-

Ti u

I Birch 2 . onallc acid

I . seporotc

___)

2. Ac20.

C5H5N

3 . Hz-Pd/C

ef & I

pyrrolidine

I

01A C O O E I)

2

NOH

Smith

I

630

639

641

Scheme 60.

Ph3P=CHZ

2 Simmonr-

540

611

Wiesner's Synthesis of (*)-Zizaene and its Stereoisomers

Bridged Systems

&

3 steps

on 634

&

477

I . OSOq

___)

2. N a I 0 4

6MO

541

613

543

544

545

509

Scheme 60, Continued. Wiesner’s Synthesis of ( f )-Zizaene and its Stereoisomers

is converted into acetal 532. Birch reduction followed by gentle hydrolysis yields the unconjugated enone 533. Upon treatment with 80% acetic acid cyclization occurs to give a 2:3 mixture of hydroxy ketones 534 and 535. The hydroxy configuration was not ascertained in either case. After separation of the isomers, compound 535 is acetylated and hydrogenated to give 536. As indicated, hydrogenation occurs exclusively on the face of the cyclopentene ring anti to the methyl group. The ketone is masked and the unwanted ester function removed by hydrolysis, conversion of the alcohol to a xanthate, pyrolytic elimination, and catalytic hydrogenation. After deprotection of the carbonyl, ketone 538 is obtained. The three remaining carbons are introduced by Stork alkylation with ethyl bromoacetate, Wittig methylenation of the resulting keto acid, and Simmons-Smith cyclopropanation. After catalytic hydrogenolysis, the acid is subjected to modified Hunsdiecker decarboxylation and the resulting bromide eliminated to obtain (fbepizizaene (511).

478

Tricarbocyclfc and Tetracarbocyclic Sesquiterpenes

Isomer 534 is subjected to a similar sequence to give keto acetate 542. Note that, as in the case of 535, hydrogenation occurs anti to the methyl group. The remaining steps are identical to those employed for the synthesis of 511; the final product is (fbepizizaene 513. For the preparation of (*)-zizaene itself, isomer 513 is cleaved by the LemieuxJohnson method to obtain ketone 543, which is epimerized by treatment with base to obtain a 1:l mixture of 543 and 544. The isomers are separated and the latter isomer converted into (f)-zizaene (509) as shown. The Wiesner synthesis of (f)-zizaene requires 30 steps, requires two isomer separations, and proceeds in an overall yield of 0.009%. Coates and Sowerby have also synthesized (*1-zizaene (Scheme 61).’18 As in the MacSweeney-Ramage approach, Coates and Sowerby employ a bicyclo l2.2. llheptane starting material. Addition of 3-phenylpropylmagnesium bromide to norbornanone gives an alcohol (546) which is rearranged in acetic acid to acetate 547. The secondary methyl group is introduced by a six-step sequence which gives a 2:l mixture of diastereomers 550 and 551. After Wolff-Kishner deoxygenation, the isomers are separated and the major acetate (552, obtained in 25% overall yield from 547) is degraded to acetoxy ester 553 in 30-3546 yield. By a routine series of steps 553 is converted into keto amide 555. This compound is further transformed into a diazo ketone which undergoes smooth cyclization to a single tricyclic ketone, 556. The high degree of stereoselectivity observed in the cyclization step is considered to be due to exo-face attack on the carbonyl group, leading to diazonium alkoxide 557. The gem-dimethyl group is introduced by reduction of the n-butylthiomethylene ketone 558 with lithium in ammonia, followed by methylation of the resulting enolate. The product of this procedure is found to

5 57

556

Bridged Systems

OH

546 I . NES, CHC13

2. caco3,* DMAC

Ph

I. NEA,

P

h

548

547

Ph

H20 2. H2Cr04* 3. COCO, d DMAC

p ? l 550

540 I . 03,HOAc

P

a

h

I. (COCI)

00 2 . HZCr04

q

2. NH3 3. L i A I H o

HOOC-

552 I . Ac20, c5H5Nb 2 . KOH 3. Collins

553

A

c

N

q

I. "$48 NaOAc CH2Cl2)

'zN

&-

554

I . NOH,

EtOCHO

2. n-BUSH,

2. NoOMe, MIOH

H

'& 55*

Scheme 61.

-

p TSOH

566

555

558

551

643

Li,Et20,

&

544

so*

Coates-Sowerby Synthesis of (*1-Zizaene

479

480

Tricrrbocyclic and Tetracarbocyclic Sesquiterpenes

be isomer 543, presumably as a result of epimerization in the course of either the formylation or methylation steps. Equilibrium between 543 and 544 is established by treatment of 543 with sodium methoxide in methanol. Although equilibrium slightly favors isomer 543, isomer 544 can be obtained in 64% yield by several cycles of equilibration, chromatography, and reequilibration of the unwanted isomer. The methylene group is introduced via the S-phenylthiomethyl benzoate 559 which is reductively eliminated to give ( f )-zizaene. The Coates-Sowerby synthesis requires 25 steps, requires 2 isomer separations, and affords the end product in 0.2% overall yield. Piers and Banville have prepared zizaene starting with cyclopropyl enone 560 as outlined in Scheme 62?19 The synthesis begins with thermal vinylcyclopropane rearrangement to give, after base-catalyzed conjugation of the double bond, enone 561. Conjugate addition of lithium divinylcopper occurs from the face of the double bond opposite the methyl group, providing unsaturated ketone 562. After introduction of a second double bond, thiophenol is added to the enone and the resulting ketone is converted into ketal 564. The vinyl group is hydroborated and the resulting alcohol converted into a tosylate. After oxidation of the sulfide to a sulfone, the final ring is closed and the benzenesulfonyl group removed by reduction with aluminum amalgam. The ketone is unmasked to obtain tricyclic ketone 567, which is treated with sodium methoxide in methanol to obtain a 2:l mixture of 567 and 556, which is an intermediate in the Coates synthesis. The Piers synthesis requires 22 steps from enone 560, involves 2 isomer separations, and provides (fbzizaene in an overall yield of 0.8% (assuming a quantitative separation of isomers 567 and 556). Khusimone (514) has been synthesized by Buchi and c o - w o r k e r ~ ~ ~ ' and by Liu and Chan.'*l Biichi's synthesis (Scheme 63) begins with the Diels-Alder reaction of a-chloroacrylonitrile and isoprene. The mixture of isomers 568 and 569, formed in a ratio of 7:3, is dehydrochlorinated and the pure diene 570 isolated by fractional distillation. Addition of 5-lithio-2-methyl-2-pentene gives trienone 571. Since this material does not undergo intramolecular Diels-Alder reaction, it is ketalized with 1,2-propanediol and the resulting diastereomeric ketals are heated at

Bridged Systems

56 I

560

481

562

OTS

&

566

565

- $J I . separate

MOO;

2 . 7 stops (Coo tes)

7

567

Scheme 62.

556

509

Piers-BanvilleSynthesis of (*)-Zizaene

250°C. The resulting adducts are separated and hydrolyzed to give tricyclic ketones 573 and 574 in a ratio of 1 3 . The minor isomer is rear-

ranged by treatment with p-toluenesulfonic acid in refluxing benzene; (*)-isokhusimone is obtained in 80% yield. Double bond isomerization is achieved by oxidation with singlet oxygen to give a mixture of allylic

482

Trlcarbocyclic and TetrncarbocyclicSesquiterpenes

dN6 568

I . DBN 2. separote

CN

+

569

bLL DCN 2I ’. H30+

E HO+ H m

57 1

570

-

c6H61A

+

ton 5 7 3 )

2. separote 3. HjO+

573

576

Scheme 63.

& 512

p-TSOH,

I . 250’

~

~

@ 575

574

577

514

Buchi’s Synthesis of (*)-Khusimone

alcohols 576. Reduction with zinc and hydrochloric acid gives a 3:7 mixture of (*)-epikhusirnone (577) and (&)-khusirnone (514). The synthesis requires 10 steps, involves 2 isomer separations, and gives 514 in 1.7% overall yield. The Liu-Chan synthesis of ketone 514 is summarized in Scheme WS2* The starting material is (-)-camphor-10-sulfonic acid, which is fused with sodium hydroxide to obtain (-1-a-campholenic acid (579) in 50% yield. After esterification, unsaturated ester 580 is ozonized and the resulting keto-aldehyde cyclized to cyclohexenone 581. Photoaddition of

Bridged Systems

483

NaOH

fusion) 578

579

580

'I

585

584

586

Scheme 64.

514

Liu-Chan Synthesis of (-1-Khusimone

ketene dimethyl acetal, followed by hydrolysis, gives a 5:8 mixture of diketo esters 582 and 583. After separation, the cyclobutanone carbonyl is selectively ketalized. The ester is hydrolyzed, methyllithium is added to the ketone, and the acid is reesterified. The tertiary alcohol is dehydrated, the ester reduced, and the resulting primary alcohol converted into chloride 585. After deprotection the ketone is ring expanded by treatment with ethyl diazoacetate and boron trifluoride. The resulting keto ester (586) is treated with ethanolic sodium hydroxide to obtain (-bkhusimone. The synthesis requires 16 steps, involves one isomer separation, and provides the optically active natural product in 2.7% overall yield.

484

(8)

Tricarbocyclic and Tetracnrbocyclic Sesqulterpenes a -Cedrene,

Cedrol, A -Cedrene and Cedradiene

In Volume 2 of this series, we recounted four syntheses leading to a-cedrene (587) and cedrol (588). In this section, we now discuss several additional syntheses of these tricyclic sesquiterpenes, as well as a synthesis leading to A*-cedrene (589) and cedradiene (590).

5 87

589

588

590

Naegeli and Kaiser have prepared A*-cedrene (589) and cedradiene (590) as shown in Scheme 65.’22 Thermolysis of dehydrolinalool at 200°C gives a mixture of dienols 591 and 592 which is converted into dienone 593 by treatment with methyl isopropenyl ether and acid.

592

59 I

593

595

Scheme 65.

594

590

589

Naegeli-Kaiser Synthesis of A2-Cedrene

Bridged Systems

485

Allylic alcohol 594 is cyclized to acoratriene (595) by treatment with stannic chloride in benzene. Further treatment of 595 with p-toluenesulfonic acid in benzene gives, in 60% yield, a hydrocarbon mixture of which the major component (50%) is cedradiene (590). Partial hydrogenation of 590 provides 589. Corey and Balanson have applied a wcyclization method developed by Stork and co-workers to the synthesis of racemic cedrone, the precursor to a-cedrene and cedrol (Scheme 66).523 Addition of 6-lithio-2-methyl2-heptene to enone 596 gives, after hydrolytic workup, enone 597. Diisobutylaluminum hydride reduction gives allylic alcohol 598, presumably as a 1:1 diastereomeric mixture. Simmons-Smith cyclopropanation, followed by Collins oxidation, gives cyclopropyl ketone 599. Cyclization is brought about by acetyl methanesulfonate. Diastereomers 600 and 601 are obtained as a 1:l mixture in 85% yield. Compound 600 has previously been converted into (*1-a-cedrene and (*)-cedr01.'~~

596

-0

PH

Smith Simmonr-

597

mQd*

Collins,

598

599

600

Scheme 66. Corey-BalansonSynthesis of (*)-a-Cedrene and (*)-Cedrol

60I

486

Tricarbocyclic and Tetracarbocyclic Sesquiterpenes

Lansbury's synthesis of ( *)-a-cedrene is summarized in Scheme 67.525Addition of lithium reagent 603 to ketone 602 gives alcohol 604 as a diastereomeric mixture. Formolysis gives a mixture of cis and trans ketones 605 and 606, with the latter predominating. After equilibration with sodium methoxide, the trans isomer 606 constitutes 85% of the mixture. Free-radical chlorination at the less hindered cyclohexyl tertiary hydrogen gives a mixture consisting principally of the diastereomeric chlorides 607. Regioselective dehydrochlorination is brought about by heating the alkoxide obtained by addition of methyllithium to 607.526 The unsaturated alcohol so produced (608) is formolyzed to give a complex mixture of hydrocarbons in which (*)-a-cedrene (587) is present in 80% yield.

L

i

dCI

603

HCOOH

HO

602

604

Y

$1 I . NaOMe 2. PhIC12, hv

0

605

w

606

'((OH

608

Scheme 67.

601 HCOOH

A

587

Lansbury's Synthesis of ( *)-a-Cedrene

Bridged Systems

487

Buchi’s khusimone synthesis also leads to racemic a-cedrene (Scheme 68).520 Unsaturated ketone 574, the major isomer produced in the intramolecular Diels-Alder cyclization of 572 (see Scheme 631, undergoes acid-catalyzed rearrangement to give a mixture consisting of 15% isokhusimone (575) and 40% of norcedrenone (609). Wittig methylenation of the latter compound gives cedradiene 610, which is reduced by diimide to a mixture of (*1-a-cedrene (587) and (*)-Zepi-a-cedrene (611).

@ /

____.)

p-TsOH benzene

574

&Q

+

@

575

610

Scheme 68.

I . separate 2. Ph3P=CHZb

609

5a 7

611

Blichi’s Synthesis of (*)-a-Cedrene

Breitholle and Fallis have prepared (f)-cedrol and (kl-a-cedrene by a route involving intramolecular Diels-Alder reaction as the key maneuver (Scheme 69).527Methylheptenone is reduced and converted into tosylate 612. Alkylation of sodium cyciopentadienide with this material yields triene 613, which is heated in xylene (sealed tube) at 155°C for 50 hours. Under these conditions a 1:l mixture of diastereomeric alkenes 614 is produced in 74% yield. Hydroboration-oxidation affords ketones 615, which are ring expanded by Demjanov rearrangement. Ketones 616 and 617 are produced in a ratio of 6:l. After addition of methyllithium the isomers are separated chromatographically to obtain (*bcedrol (588). Dehydration of 588 affords (*t)-a-cedrene (587).

488

Tricarbocyclic and Tetracarbocyclic Sesquiterpenes

-

I . LiAIH, 2. p-TIC1 612

-

613

617

616 !

I . MeLi 2 . separate

H

588

587

Scheme 69. Breitholle-Fallis Synthesis of kt)-Cedrol and (*t)-u-Cedrene

(9) Quadrone

The antibiotic quadrone (618) is a member of a growing group of sesquiterpenes containing the perhydropentalene nucleus (coriolin, hirsutanes, isocomene, pentalenolactone, gymnomitrol, etc.). As we have seen earlier in this review, these compounds have captured the imaginations of synthetic chemists, and a number of successful syntheses have resulted.

Bridged Systems

489

618

The only published quadrone synthesis to date is that of Danishefsky and co-workers (Scheme 70).528 Conjugate addition of vinylmagnesium bromide to enone 620, followed by alkylation of the resulting enolate with bromide 619, gives keto ester 621. Ketalization gives diketal ester 622. The vinyl group is fashioned into a P-bromoethyl group by the four-step sequence indicated. After liberation of the two carbonyl groups, cyclization is brought about by sodium methoxide in methanol. Mukiayama addition of 1- r-butoxy-1-f-butyldimethylsilyloxyethylene to 624 provides keto diester 625 which is hydrolyzed, decarboxylated, reesterified, and ketalized to afford 626. This material is properly constituted for formation of the third carbocycle. Treatment of the corresponding iodide with lithium hexamethyldisilazide affords, after ketal hydrolysis, keto ester 627. It is interesting that only the axial methoxycarbonyl isomer is produced. This probably results from a preferred conformation of the enolate derived from ester 626. The observed product would result from conformation 62611;626b would lead to the diastereomer 628.. There is no obvious reason for this stereospecificity. With the ester group thus placed in its proper configuration, it is next hydrolyzed to protect against epimerization. It turns out that the keto acid derived from 627 undergoes enolization or enolate formation exclusively toward C-4, probably due to relief of a non-bonded interaction between a C-4 hydrogen and the proximate methyl group. Thus, it is necessary to block this position so that the hydroxymethyl group can be introduced at C-2. This is accomplished by the Sharpless-Reich procedure, which gives enone 629. Treatment of this *It is known that enolization of such esters in the presence of hexamethylphosphoric triamide affords the enolate in which the alkoxide group is trans to the hydrogen of the double bond.

490

Tricrrbocyclic and Tetrrcrrbocyclic Sesquiterpeoes

Br

623

622

I , HCI,A 2. C t i z N ~

0

3. ( c H ~ o H ) ~ ~ ti+

MeOOC

624

625 Br

626

629

627

630

Scheme 70.

63 I

Danishefsky's Synthesis of (*)-Quadrone

618

Bridged Systems

6260

491

627

material with three equivalents of lithium diisopropylamide and gaseous formaldehyde affords a hydroxy acid (630) which is hydrogenated to obtain 631. High stereospecificity is observed in the hydrogenation step since reaction from the other face of the double bond would produce a trans-fused perhydropentalene system. Acid-catalyzed dehydration of 631 affords enone 632, the probable biologically active agent. However,

631

- AH benzene

632

+

&H

618

633

further treatment of 632 with acid yields primarily 633, an isomer of quadrone. It is found that when either enone 632 or p-hydroxy ketone 631 is heated at 190-195°C for 5 minutes, the sole product is (*Iquadrone (618).

492

Tricarbocyclic and Tetracarbocyclic Sesquiterpenes

Isolongifolene (634) is actually not a natural product. Instead, it is the product of acid-catalyzed rearrangement of longifolene. In Volume 2,

634

we reported an eight-step synthesis of 634 from camphene-1-carboxylic acid.529 Piers and Zbozny have recently reported an alternative synthesis of 634 (Scheme 71).530The key reaction of the Piers-Zbozny synthesis is

fi0

2 . PhSP=CHZ* Me1

635

a (% COOMe

I . I-EuOK

I ' H202 2. 3. Ac20

636

638

EHw

COOMe

~

N 637

639

moAC I . LiAIH4 2 . p-TsCI 3 . H30+

*

f-EuOK

0

64 I

640

+

I . C H NH

I

5 -5

f3'3 2. DMAC 3. Me2CuLi

I . DDQ

0 642

8 F3 2 . Raney N i

Scheme 71.

644

643

*

634

Piers-Zbozny Synthesis of

(f1-Isolongifolene

O

A

C

Bridged Systems

493

self-alkylation of keto tosylate 641, which occurs solely at the y-position of the extended enolate to give 642. Cyclization at the a-position would give a strained bridgehead alkene. Intermediate 641 is prepared by the straightforward sequence of steps shown in the scheme. The most noteworthy transformation is the method used for the allylic oxidation (637-638). When oxidation with chromium failed, it was found that the Finucane-Thompson procedure provides enone 638 in 96% yield. In all, the synthesis requires 17 steps and provides 634 in about 1% overall yield. It is considerably less efficient than Dev’s eight-step synthesis, which gives an overall yield of

(I I) Ishwarone and Ishwarane Ishwarone (645) and the corresponding hydrocarbon ishwarane (646) are two members of the relatively small class of tetracarbocyclic sesquiterpenes. Several syntheses of these interesting compounds have been reported.

645

646

Kelly and co-workers have synthesized (f)-ishwarane as shown in Scheme 72.531a-cEnone 647 undergoes photoaddition of allene to give an unsaturated ketone which is ketalized to obtain 648. Epoxidation gives a mixture of epoxides which is reduced to obtain alcohols 649. Acid-catalyzed deprotection of the ketone is followed by rearrangement of the resulting aldol to the more stable bicyclo [2.2.0loctane aldol 650, which is obtained as a diastereomeric mixture. Dehydration and reduction gives a mixture of unsaturated alcohols (651). The alcohol is protected as a benzyl ether, the alkene is hydroborated and the resulting organoborane oxidized to the ketone. Deprotection gives hydroxy ketone 652. The corresponding tosylates are both cyclized by treatment

Tricarhyclic and Tetracarbocycllc Sesqulterpenes

494

OH

I

648

647

HCI, H 2 0 THF

o&

1 . benzene, P-TsOH 2. LiAIH4

649 I , PhCHZCI. 2. NOH BzHg

HC+&

3. Jones

650

4 . H2-Pd/C

651

Kishner 652

653

Scheme 72.

646

Kelly’s Synthesis of (*.)-Ishwarane

with methylsulfinyl carbanion, yielding a ketone (653) which is reduced to (*)-ishwarane (646). The fact that diastereomeric enolate 652a cyclizes is not surprising, since it is constituted for nucleophilic substitution with inversion of configuration:

6520

However, enolate 6528 cyclizes with rerenrion of configuration at the tosylate-bearing carbon:

Bridged Systems

495

652 b

Kelly and co-workers propose that this diastereomer may react by initial displacement of tosylate by a molecule of the solvent, which is in turn displaced by the enolate to give 653:

It is also suggested that elimination may occur first, followed by cyclization to the homoenolate:

Whereas the Kelly synthesis requires 14 steps for the conversion of enone 647 into ishwarane, Cory and McLaren discovered a way that the same conversion can be done in only four steps (Scheme 73).532Conjugate addition of lithium dimethylcopper gives ketone 654, which is converted into alkene 655 by Grignard addition followed by dehydration of the resulting tertiary alcohol. Alkene 655 is accompanied by 20% of its A'-isomer. The key transformation is addition of dibromocarbene to give an intermediate dibromocyclopropane (656) which is treated with methyllithium to generate a carbene (657). Cyclization occurs regiospecifically, affording ( f)-ishwarane (646) in 26% yield.

496

Tricrrbocyclic rod TetrrcrrbocyclicSesquiterpenes

I

I . MeMgI 2. HZSOq

I '

I

647

654

CBr4,MoLi -70.

I' 655

*

I

I

656

657

Scheme 73.

646

Cory's Synthesis of (*:)-Ishwarane

An application of this basic approach to the synthesis of (fbishwarone

is summarized in Scheme 74.532Enol ether 658 is converted into enone 659, which undergoes conjugate addition of the cuprate reagent derived from 4-lithio-2-methyl-l-butene.The product (660) is epoxidiied and

the resulting keto epoxide cyclized by treatment with potassium t-butoxide. The product is a mixture of alcohols 661 and 662. Dehydration with 50% aqueous sulfuric acid gives a 2:l mixture of unsaturated ketones 663 and 664 in 59% yield. Although this mixture can be separated chromatographically, this is not necessary at this stage, since isomer 664 does not react with dibromocarbene. Thus, treatment of the mixture with bromoform in the presence of 50% aqueous NaOH gives cyclopropane 665, which is treated with methyllithium in ether at -30°C to obtain racemic ishwarone.

Bridged Systems

658

659

CHBrS_ NaOH

660

66 2

661

&:; I

I

I

665

Scheme 74.

497

4 66 3

MeLi ether,

~

- 30°

664

645

Cory’s Synthesis of

(*1-Ishwarone

Piers and Hall have synthesized (f)-ishwarone as outlined in Scheme

75.533Ketal aldehyde 666, an intermediate in the Ziegler eremophilone

synthesis, is converted into acetylenic alcohol 667, which is deprotected and hydrogenated to the cis-allylic alcohol 668. The derived mesylate is cyclized to give unsaturated ketone 669. The ketal is allowed to react with dimethyl diazomalonate to give cyclopropane diester 670. This compound is converted into dichloro ketone 671 by a standard four-step sequence. Base-catalyzed cyclization gives the chloroishwarone 672, which is reduced with lithium triethylborohydride. Reoxidation of the resulting alcohol provides racemic ishwarone.

Sesquiterpene Alkalolds

498

3. C H 2 0

I

666

1 1 -

66 7

668

I . LiAIH4 I . MsCI

2 . 1-BuOK

I 1 I

669

,I:&

I-BuOK,

'

1

1

A 67 I

2. N ~ C ( C O O M O ) ~ . Cu, A

' I

I

A

pgC'I . LiEtSBH 2 . PCC

672

Scheme 75.

HMPT

670

645

Piers-Hall Synthesis of (k)-lshwarone

7. SESQUITERPENE ALKALOIDS A. Illudinine

The fungal metabolite illudinine (1) has been synthesizec. by Woodward and Hoye as outlined in Scheme 1.534 Friedel-Crafts acylation of indane with @-chloropropionyl chloride gives a ketone (3) which cyclizes upon treatment with hot sulfuric acid to give a mixture of tricyclic ketones 4 and 5. After methylation and Clemmensen reduction, the resulting isomeric hydrocarbons are separated by careful distillation. Compound 6

Illudinine

3

2

499

4

6

I . n-BuLi

I

,

2. B(OMe13

t-BuOK, M e 1

3 . HOAc 4 . He02

2 . Clemmensen 3. separate

Br

6 OH

I . K2CO3, Me2SO4

Cr03, HOAc

*

rn

2 . Mg, THF 3. ClCOOMe

I I . separate 23.. NoBHq p-TsOH

*

9:::24- ow NH40Ac

OH

COOMe

HOAc*

COOMe

II

N/

Scheme 1.

KOH

E1OHd.OOMe

dOOH

12

I

Woodward-Hoye Synthesis of Illudinine

m

500

Sesquiterpeae Alkaloids

is obtained in 44% yield, accompanied by 19% of the isomer derived

from ketone 5. After bromination of the ring, one aromatic position is oxidized by way of the boronate ester. The resulting phenol is methylated and the other aryl bromide replaced by a methoxycarbonyl group. Benzylic oxidation provides a mixture of ketones 7, 8, 9, and 10 in yields of 36%, 14%, 8%, and 6%, respectively. The major oxidation product is converted into dialdehyde 11 and thence into pyridine 12 by standard procedures. Saponification of the ester function affords illudinine.

B. Deoxynupharidine, Castoramine, Deoxynupharamine, and Nupharrmime The genus Nuphur produces a number of sesquiterpene alkaloids of which deoxynupharidine (131, castoramine (141, and nupharamine (15) are characteristic. These compounds and the deoxy analog of nupharamine (16) have been synthesized.

I3

14

15: R = OH 16: R = H

The first synthesis of deoxynupharidine was reported by Kotake and Kaneko and their co-workers (Scheme 2).535a-d2,s-Dimethylpyridine is heated with formalin at 160°C to obtain diol 17, which is acetylated with concomitant dehydration to obtain allylic acetate 18. Alkylation of malonic ester with this substance provides compound 19, which is hydrogenated, hydrolyzed, and decarboxylated to obtain amino acid 20. The corresponding ethyl ester is acetylated with 3-furancarbonyl chloride to 21. When 21 is heated with soda lime, there is produced an enamine (22) which is hydrogenated to a mixture of (st 1-deoxynupharidine (13) and its diastereomer 23. Isomers 13 and 23 are readily separated by chromatography on alumina.

Deoxynupharldlne, Castoramhe, Deoxynupharamine, and Nupharrmine HO

&@,

OH

CH2(C00Et)2m

+A

NOH

17

18

22

21

13

Scheme 2.

501

23

Kotake-Kaneko Synthesis of (*1-Deoxynupharidine

Arata’s approach to deoxynupharidine is summarized in Scheme 3.536 Substituted malonate 24 is hydrogenated over Adarns’ catalyst and the resulting amino diester heated to obtain lactam 25. After hydrolysis, the mixture of diastereorneric amino acids is separated and isomer 26 is converted into lactam 27, of the proper configuration for conversion into racemic deoxynupharidine. Claisen acylation of 31 with ethyl 3-furoate gives 8-keto amide 28 which is hydrolyzed and decarboxylated to amino

502

Sesquiterpene Alkaloids

P5

24

HOOC-’b H

*.

I . EtOH,H+ 2. A

NOH

9.. 27

26

HCI, H20

A

*

I . NaBH, 2 . soc12

P=

3. KZCOJ

29

28

&I.. +

24

13

Scheme 3.

Arata’s Synthesis of

( f bDeoxynupharidine

ketone 29. After reduction of the carbonyl group, the resulting alcohol is allowed to react with thionyl chloride. Cyclization is brought about by treatment with potassium carbonate. kI-Deoxynupharidine (131, presumably accompanied by some of its diastereomer 24, is obtained and purified by recrystallization of the perchlorate salt. A similar approach was reported by Jezo and c o - ~ o r k e r s . ’ ~ ~ Bohlmann and co-workers carried out an extensive stereochemical investigation which fully clarified the structure of de~xynupharidine.’~~ As shown in Scheme 4, ester 30 is alkylated successively with methyl

Deoxynupharidine, Castoramine, Deoxynupharamine, and Nupharamine

32

I . NaOH

GCoc'-

2. CHzN2

*

33

MeOOC

503

34

roda lime+ 280.

35

b.

+

b

reduction,

36

isomers

13

Scheme 4.

Bohlmann's Synthesis of

(*bDeoxynupharidine

iodide and ethyl P-iodopropionate to obtain diester 30, which is elaborated by straightforward methods into ester 32. Catalytic hydrogenation gives a mixture of the four diastereomeric lactams 33, which are hydrolyzed to obtain the mixture of amino acid salts 34. ShottenBaumann acylation with 3-furancarbonyl chloride gives an amide ( 3 9 , which is cyclized to enamine 36 by heating with soda lime. Catalytic hydrogenation of 36 over palladium on barium sulfate gives a mixture of

504

Sesquiterpene Alkaloids

four diastereomers. All four of these isomers were shown by infrared spectroscopy (Bohlmann bands) to have the “trans” perhydroquinolizidine structure. These structures are presumably 13, 23, 37, and 38. ( f )-Deoxynupharidine (13) is obtained in 7% yield by chromatographic

U

I

23

37

Ar

38

separation of the mixture. When enamine 36 is protonated and the resulting immonium salt reduced with sodium borohydride, six diastereomeric amines are produced. In addition to isomers 13, 23, 37, and 38, the “cis” perhydroquinolizidine isomers 39 and 40 are also obtained under these conditions.

Deoxynupharidine, Castoramine, Deoxynupharamine, and Nupharnmlne

505

H

I

39 n

I

40

Wr6bel and Dabrowski synthesized (*1-deoxynupharidine as shown in Scheme 5.539 Addition of ethyl propiolate to ketone 41 gives an adduct (42) which is converted into amino ester 43 and thence into Bohlmann’s lactam (33). Addition of 3-furyllithium to 33, followed by dehydration, gives enamine 36, which is hydrogenated as in the Bohlmann synthesis. Subsequently, Wr6bel and co-workers worked out the first stereoselective synthesis of (fbdeoxynupharidine (Scheme 6).5393-Acetylfuran (44) is converted into enol ether 45, which is converted into quinolizidinium salt 46 under acidic conditions. Reduction of this material with sodium cyanoborohydride provides diastereomeric amino esters 47 and 48 in 56% and 11% yield, respectively. Base-catalyzed equilibration of the mixture gives a single isomer which is converted into (f)-deoxynupharidine (13) by standard methods. The synthesis of (*hastoraxnine (Scheme 7)540 by Bohlmann and co-workers utilizes the Wr6bel-Dabrowski method for introduction of the furyl group. The quinolizidine is constructed by a classical sequence starting with the keto diester 50. Lactam ester 51 is reduced by lithium borohydride to a mixture of diastereomeric hydroxy lactams (52). After protection of the hydroxy group, the furyl group is introduced and the resulting enamine hydrogenated. (f)-Castoramine (14) is obtained in approximately 10% yield by chromatography. In addition to 14, three

Sesquiterpene Alkaloids

506

I . H2-Pt

HCWXOOEt boar

EtOOC’

3. H2-Pt 42

41

b

EtOOC

$”

0

43

H

36

33

2 -Pd

13

+.a*

Wr6bel-Dabrowski Synthesis of (+ bDeoxynupharidine

Scheme 5 .

I . NaH,

,

I . CFSCOOH 2. 45% HBr, HOAc

HCOOE? DMF

45

44

.

NaBH3CN EtOH

I

EtOH

Br46

48

47

I . LiAIH4

3 . LiAIH4

47

Scheme 6.

I3

Wr6bel’s Synthesis of (+bDeoxynupharidine

Deoxynupharldlne, Castornmlne, Deoxynupharamine, and Nupharamlne

507

H ___)

h

EtOOC

M

e

2

MII_

50

H2-Pd

2. H,O+

8

O ,H

/OH

53

Scheme 7.

+ isomers

14

Bohlmann's Synthesis of

(fKastoramine

crystalline stereoisomers are obtained in comparable yields. These three isomers are believed to have structures 54-56.

54

55

56

508

Sesquiterpene Alkaloids

Nupharamine (15) is a monocyclic congener of deoxynupharidine. The first synthetic study reported led to the deoxy analog 16 (Scheme Beginning with acetoacetic ester and the acid chloride 57, the monocyclic lactam (63) is built up in more or less the same manner as utilized by Bohlmann in the deoxynupharidine and castoramine syntheses (see Schemes 4 and 7). Hydrolytic ring opening of 63 gives an amino ester

-

COOEt

NoOEt

O

18

57

59

Jf),

COOEt er/vCOOEt

HCI, H 2 0

____)

A

EtOOC

HOOC

*

60

\OH

2 . A

I . HCI, H 0

2 . EtOH, HCI

$H

63

62

1 1

d o C I

v .

EtOOC

64

Scheme 8.

I . HzNOH,

2. CHzN2

61

I . H2-PtO2 MeOOC

EtOOC

Ti?

6

0i;X 16

b

I , soda lime.

A

2. H +

c

3. NoBH,

61

+

isomers

Arata's Synthesis of

(*bDeoxynupharamine

Deoxynuphnridine, Cnstoramine, Deoxynuphnrnmine, and Nupharnmine

509

(641, which is acylated with 3-furancarbonyl chloride to obtain amide ester 65. Ring closure is accomplished by the Kotake-Kaneko method. After reduction of the resulting enamine, a mixture of (fbdeoxynupharamine and its stereoisomers is obtained. Wr6bel and co-workers have prepared (*hupharamine itself (Scheme 9).539 The synthesis begins with a Hantzsch pyridine construction, leading from 3-acetylfuran to the furylpyridine 67. After conversion of the ethoxycarbonyl group to methyl, the more acidic methyl position is alkylated with methallyl chloride and the pyridine ring reduced with sodium in ethanol. Diastereomers 71 and 72 are obtained as a 3:l mixture in 12% yield. After hydration of the alkene linkage, the isomers are separated, giving a 3:l mixture of (*)-3-epinupharamine (73) and ( f h u p h a r a m i n e (15).

HgEtb &cHo

(-f

F'

ECOOEtb

\ N

I . LiAlH,

2. 60% Her

66 67

-

&*),

68 +

, NH

71

6-1 69

\

70

I . HCOOH, HCIO4

, NH

72

Scheme 9.

\

2. NaHC03,* MeOH 3. separate 73

16

Wr6bel's Synthesis of (*)-Nupharamine

510

Sesquiterpene Alkaloids

C. Dendrobine Dendrobine (74) was first isa..ded by Suzuki and co-workers from the stem of the ornamental Chinese orchid Dendrobium nobile ( Orchiducue). It is a component of the Chinese herbal drug Chin-Shih-Hu, which is used as a tonic and antipyretic. An interesting feature of the dendrobine

74

molecule is the fact that the cyclohexane ring must exist in a boat conformation, with alkyl groups at the bowsprit positions. At the outset of any dendrobine synthesis, it would appear that this requirement would constitute a formidable hurdle. However, investigations of the alkaloid showed that hydroxy acid 75, obtainable from dendrobine by hydrolysis, undergoes ready acid-catalyzed lactonization to regenerate the alkaloid.

* x

\

74

e Hoe* \

71

The first successful total synthesis of dendrobine was that of Yamada The synand co-workers, which was completed in 1972 (Scheme thesis utilizes a method developed by Johnson for construction of cisfused hydrindanones. Thus, treatment of diketo acid 78 with potassium r-butoxide results in Michael addition, followed by intramolecular aldolization and provides acids 79 and 80 in a ratio of 8:l. The stereochemistry of this process is interesting. The 8:l ratio may reflect the ratio of

Dendrobine

-08

MI

I . AC,O,H+

" O3

OM.*

MaO

COOH

3 . PhsP=(

I-BuOH

COOH

3. H30+

4 . H+

76

-

511

77

78 ,OAc

t-BuOK

HO'

la

COOH

OOH

I. separate 2 . CHeN2 3. Ac20, H+

AcO"

:A

COOMa

80

79

81

I . HCOOMI, NoOMa

NOH, p l y m r

2. CH2N2 3. n-BuSH

I b

COOMI

86

87

86

88

Scheme 10.

74

Yamada's Synthesis of (*t)-Dendrobine

512

Sesquiterpene Alkaloids

trans and cis isomers of the precursor 78. Alternatively, the entire process may be under thermodynamic control, and the diastereomer ratio may reflect the relative stability of the products:

80

79

After separation, the major isomer is esterified and treated with acetic anhydride to obtain an enol acetate (81) which is ozonized to obtain 82. The nitrogen is inserted at this point by reaction of the imidazolide with methylamine. The resulting carbinolamide is brominated and the resulting bromo ketone cyclized by treatment with sodium hydride. Although it is apparently fortuitous, the regiospecific bromination of 83 is perhaps the key step in the Yamada synthesis. The isopropyl group is introduced by a method developed by Coates, that is, by addition of excess lithium dimethylcopper to the a-n-butylthiomethylene derivative 85. The resulting keto ester, 86, is equilibrated by treatment with sodium hydride in glyme, leading to a 1:l mixture of 86 and 87. Although the full stereochemistry of isomer 86 was not elucidated, it was shown by nmr evidence that the methoxycarbonyl group had retained its configuration. The likely structure of this compound is that in which the cyclohexane

MeOOC

m a

ring is in a boat conformation (Baa). The synthesis is completed by reduction of keto ester 87 to the hydroxy ester, which is cyclized. The amide carbonyl is removed by Borch’s method giving (*I-dendrobine.

Dendrobine

513

Inubushi and co-workers prepared (*1-dendrobine as shown in Scheme 11.545The known enedione 89 is stereospecifically reduced and converted into an enone tosylate (90) which is treated with cyanide in dimethyl sulfoxide. Although Inubushi and co-workers hoped to obtain the inverted nitrile by this method, the displacement occurs with apparent retention of configuration, undoubtedly as a result of subsequent epimerization of the initial product. The same cyano enone is obtained by Robinson annelation of cyano ketone 92 with methyl vinyl ketone. After saturation of the double bond of 91, a double bond is reintroduced by bromination-dehydrobromination. Enones 95 and 96 are obtained in a 1:3.5 ratio. Although the undesired major isomer may be recycled to 94, this is an exceedingly poor step. Nevertheless, the minor isomer 95, obtainable in 21% yield by silica gel chromatography, is utilized to continue the synthesis. The enone is ketalized and the cyano function hydrolyzed with alkali. Acidic hydrolysis of the ketal provides a 3:l mixture of enone acid 97 and keto lactone 9 8 the latter substance is obtained in 22% yield by chromatography on silica gel. Treatment of the keto lactone with methylamine gives keto lactam 99, which is allowed to react with isopropylmagnesium bromide. Unfortunately, enolization of the ketone is the sole result at room temperature. At -60°C to -70°C some addition occurs and alcohol 100 may be obtained by fractional crystallization in 17% yield. Resubmission of the mother liquors to the Grignard reaction gives more alcohol. After 4-5 cycles of this procedure, alcohol 100 may be obtained in a total yield of 74%. Dehydration provides a 1:l:l mixture of three alkenes (101). This mixture is treated under Prevost hydroxylation conditions. After treatment of the crude product with silica gel, allylic acetates 102 and 103 are obtained in yields of 109'0 and 20%, respectively. The minor isomer is converted into enone 104. Nagata cyanation of 104 yields a mixture of p-cyano ketones 105, 106, and 107 in isolated yields of 1846, 20% and 29%, respectively. The isomers are separated and the major isomer converted into keto ester 87. The remainder of the Inubushi synthesis is identical to the Yamada synthesis. Kende's synthesis of dendrobine is outlined in Scheme 12.546The known triacetate 108 is saponified and oxidized to hydroxy benzoquinone

514

Sesquiterpeoe Alkaloids

9s

I . 12,AgOAc

2

100

*"

101

3. Colllnr

I02

103

88

Scheme 1 1 .

Et2AICN

2 . KOH

__I)

HOAc silica

I04

74

Inubushi's Synthesis of (*:)-Dendrobine

Dendrobine

110

109

108

112

114

113

118

I16

COOMc

118

EOOMo I19

74

Scheme 12.

Kende's Synthesis of (rtbDendrobine

515

516

Sesqulterpene Alkaloids

109, which undergoes regiospecific Diels-Alder cycloaddition with butadiene to afford adduct 110 in 95% overall yield for the three steps. After methylation of the hydroxy group, the cyclohexene double bond is cleaved to obtain dialdehyde 111, which is cyclized to a 1:l mixture of unsaturated aldehydes 112 and 113, each of which is obtained in 25% overall yield from 110. The isomers are separated and compound 112 reductively aminated by treatment with methylamine and sodium cyanoborohydride. The resulting product is a mixture of alcohol 114 and tricyclic amine 115, which are obtained in yields of 26% and 3546, respectively. The unexpected formation of tricyclic 115 in this reaction probably results from initial 1,Creduction of aldehyde 112, followed by reductive amination of the resulting saturated aldehyde. The secondary amine so formed must condense to give an immonium ion which suffers reduction to produce 115. Conversion to enone 116 is routine. Kende and co-workers introduce the methoxycarbonyl group by codugate addition of lithium divinylcopper, to obtain 117 in 80% yield. Cleavage of the vinyl group is accomplished by oxidation with ruthenium tetroxide. After esterification there is produced a keto ester (stereochemistry of the isopropyl group unknown). Treatment of 118 with sodium methoxide gives a mixture of which the desired keto ester 119 constitutes only 30%. Compound 119 is isolated in 43% yield, based on unrecovered keto ester. Sodium borohydride reduction and acidification provide (*)-dendrobine in 46% yield. The final successful synthesis of dendrobine, which was reported in 1978,547is that of Roush. As shown in Scheme 13, the Roush synthesis begins with a straightforward construction of triene ester 125, which is silylated and heated in refluxing toluene to accomplish intramolecular Diels-Alder cycloaddition. Isomers 126-129 are obtained in yields of The two major isomers, 126 and 39%, 36%, 1% and !I% respectively. , 127, both result from endo addition: OH

126. I27

517

Dendroblne

-

lZ4

lZ6

1 . BSA,

toluene

A

2. H30+

I LO

COOMe

tOOMa

127

120

COOMo

kOOMo

I29

EOOMo

COOMo

I30

139

133

2. I . Zn,HOAc (COCI),

NH ,,-flN 134

I . MaCl

3. L i A l H ( t - B u 0 ) 3 D

2. MONH2, OMSO

U

k k o

EOOYo

I37

I36

>pH +

2 . m-CPBA

COOMo I .I 3 Jones 0

I . separate

___)

2. Zn-HOAc

\

COOMo

I39

2 . NaBH4

COOMe

I40

Scheme 13.

3. silico

74

Roush's Synthesis of (It)-Dendrobine

518

Sesqulterpene Alkaloids

After separation, these isomers are oxidized and the hydrindenone isomerized to the more stable cis form by treatment with silica gel. Methylation with potassium r-butoxide and methyl iodide provides an enone (131). This is converted into a 4:l mixture of nitriles 133 and 134 by treatment with p-toluenesulfonylisoyanide and potassium r-butoxide. After separation, the major isomer is hydrolyzed to an unsaturated amide, which is bromolactonized to obtain 135. Reductive elimination gives an unsaturated acid, which is reduced via its acid chloride to obtain alcohol 136. The methylamino group is introduced by displacement on the derived mesylate and the resulting secondary amine (137) protected as a carbamate. Epoxidation of the double bond is accomplished by rn-chloroperoxybenzoic acid, but diastereomeric epoxides 138 and 139 are produced in yields of 40 and 47%, respectively. The minor isomer is deprotected and cyclized to obtain 140. Inversion of the secondary alcohol is accomplished by oxidation and reduction of the resulting ketone. Spontmeous lactonization results when the hydroxy ester is chromatographed on silica gel. At this point, it is instructive to compare the four dendrobine syntheses which have been completed to date (Table 4). The most efficient, from the standpoint of step economy, is the Kende synthesis. However, this synthesis gives a lower overall yield and requires a more unfavorable isomer separation than does the Yamada synthesis. The highest overall yield is obtained in the Roush synthesis, but this synthesis is also quite long. The least efficient synthesis, on all counts, is the Inubushi method. It is clear that there is still room for improvement in the synthesis of this interesting alkaloid. Unsuccessful approaches to dendrobine have emanated from three other l a b o r a t o r i e ~ . ~ ~ ~ - ~ ~ ~

a.

Roush, 1978

Kende, 197

[nubushi. 1972

Yamada, 1972

Investlgator

1 .o%

0.6%

I5

23

0.00008%

24

0.8%

It is necessary to employ the minor isomer from this separation.

Me0

19

Table 4 Comparlson 01 Dcndrobinc Syntheses Starting Overall Isomer Point Steps Yield

2 0 :1,3:7a)

4(1:3.5.81:3,a 1:2,'3:2:2)

2(8:1,1:1)

Separations

520

References

REFERENCES 1. L. Ruzicka. Helv. Chim. ACIU,6, 492 (1923).

2. C. H.Heathcock in The Total Synthesis ofNatural Products, Volume 2, edited by J. ApSimon, John Wiley and Sons, Inc., New York, 1973. 3. E. J. Corey and H. Yamamoto, J. Am. Chem. Soc., 92, 6637 (1970). 4. (a) E. J. Corey and H.Yamamoto, J. Am. Chem. SOC.,92, 3523 (1970). (b) E. J. Corey, J. I. Shulman, and H. Yamamoto, Tetrahedron Lett., 447 (1970).

5 . B. S. Pitzele, J. S. Baran, and D. H.Steinman, Terrahedron, 32, 1347 (1976). 6. 0. P. Vig, A. K.Vig, and S . D. Kumar, lndion J. Chem., 13, 1244 (1975). 7. S. Akutapwa, T.Taketomi, and S. Otsuka, Chem. Left., 485 (1976). 8. L. Ahlquist and S. Stallberg-Stenhagen, Acta Chem. Scand., 25, 1685 (1971). 9. 0. P. Vig, A. K. vig and S. D. Kumar, IndianJ. Chem., 13, 1003 (1975). 10. F. J. Gottschalk and P. Weyerstahl, Chem. Ber., 108, 2799, (1975). 11. See ref. 2, Scheme 9, pages 209-210. 12. (a) J. A. Findlay. W. D. MacKay, and W. S . Bowers, J. Chem. SOC. (C), 2631, (1970). (b) See ref. 2, Scheme 15, page 219. 13. See ref. 2, Scheme 16, page 219. 14. J. S. Cochrane and J. R. Hanson, J. Chem. Soc.. Perkin I, 361 (1972). 15. See ref. 2. Scheme 5, page 205. 16. See ref. 2, Scheme 9, pages 209-210. 17. R. J. Anderson, C. A. Henrick, J. B. Siddall, and R. Zurfluh, J. Am. Chem. Soc., 94, 5379 (1972).

18. See ref. 2, Scheme 11, pages 213-214. 19. 20. 21. 22. 23. 24. 25. 26. 27.

C. A. Henrick, F. Schaub, and J. B. Siddall, J. Am Chem. SOC.,94, 5374 (1972). G. Biichi and H.Wuest, J. Am. Chem. Soc., 96, 7573 (1974). M.Baumann, W.Hoffmann, and H. Pommer, Leibigs Ann. Chem.. 1626 (1976). See ref. 2, Schemes 2 and 3, pp, 202-203. T. Hiyama, A. Kanakura, H. Yamamoto, and H. Nozaki, Tetrahedron Len, 3051 (1978). See ref. 2, pages 200-205. 0. P. Vig, S. D. Sharma, M. L. Sharma, and S . S. Bari, Indian J. Chem., 14B, 926 (1976). F. Bohlmann and R . Krammer, Chem. Ber., 109, 3362 (1976). J. Meinwald, K. Opheim, and T. Eisner, Tetrahedron Lett., 281 (1973).

References

521

28. C. H. Miller, J. A. Katzenellenbogen, and S . B. Bowlus, Tetrahedron Lett., 285 (1973). 29. T. Kato, H. Takayanagi, T. Uyehara and Y. Kitahara, Chem. Lett.. 1009 (1977). 30. L. J. Altman, L. Ash, and S. Marson, Synthesis 129 (1974). 31. Y. Gopichand, R. S. Prasad, and K. K. Chakravarti, Tefruhedron Lett., 5177 (1973). 32. S. Kumazawa, K. Nishihara, T. Kato, Y. Kitahara, H. Kornae, and N. Hayashi, Bull. Chem. SOC.Jupufi 47. 1530 (1974). 33. A. Takeda, K. Shinhama, and S. Tsuboi, Bull. Chem. SOC.Jupufi 50, 1903 (1977). 34. N. Fukamiya and S. Yasuda, Chem. Ind. (London), 126 (1979). 35. A. F. Thomas and R. Dubini, Helv. Chem. Acre 57, 2066 (1974). 36. See ref. 2, pages 227-233.

37. S. Takahashi, Synth. Commun., 6, 331 (1976).

38.

K. Kondo and M. Matsumoto,

Tetrahedron Lett., 391 (1976).

39. L. T. Burka, B. J. Wilson, and T. M. Harris, J. Org. Chem., 39, 2212 (1974). 40. K. Kondo and M. Matsumoto, Tetrahedron Lett., 4363 (1976). 41. C. F. Ingham, R. A. Massy-Westropp and G. D. Reynolds, Ausr. J. Chem., 27, 1477 (1974).

42. C. F. Ingham and R. A. Massy-Westropp, Aust. J. Chem.. 27. 1491 (1974). 43. D. W. Knight and G. Pattenden, J. Chem. Soc., Perkin I, 641 (1975). 44. A. P. Krapcho and E. G. E. Jahngen, Jr., J. Org. Chem., 39, 1322 (1974). 45. See ref. 2, Scheme 6, page 243.

46. S. S. Hall, F. J. McEnroe, and H. -J. Shue, J. Org. Chem., 40, 3306 (1975).

47. See ref. 2, Scheme 6, page 243 48. 0. P. Vig, R. C. Anand, A. Singh, and J. P. Salota, Indiun J. Chem., 8, 953 (1970). 49. 0. P. Vig, A.

K. Vig, and 0. P. Chugh,

Indian J. Chem., 10, 571 (1972).

50. P. A. Grieco and R. S . Finkelhor, J. Org. Chem., 38, 2909 (1973).

51. 0. S. Park, Y. Grillasca, G. A. Garcia, and L. A. Maldonado, Synth. Commun., 7, 345 (1977). 52. P. Gosselin, S . Masson, and A. Thuillier, J. Org. Chem., 44, 2807 (1979). 53. See ref. 2, Scheme 13, page 249. 54. G. Gast and Y. -R. Naves, Helv. Chim. Act6 54, 1369 (1971). 55. See ref. 2, Schemes 15-16, pages 250-251. 56. See ref. 2, Schemes 21-22. pages 224-227. 57. See ref. 2, Scheme 23, page 228.

522

References

58. See Section 2 of this article, Scheme 27. 59. 0. Errner and S . Lifson, Tetrahedron, 30, 2425 (1974). 60. F. J. McEnrot and W. Fenical, Tetrahedron, 34, 1661 (1978).

61. R. B. Mane and G . S.Krishna Rao, Indian J. Chem., 12, 938 (1974).

62. F. Bohlrnann and D. Kornig, Chem. Ber., 107, 1777 (1974).

63. N. Dennison, R. N. Mirrington, and A. D. Stuart, Ausr. J. Chem., 28, 1339 (1975). 64. 0. P. Vig. ( 1 973).

H. Kurnar, J. P. Salota, and

S. D. Sharrna, Indian J. Chem., 11, 86

65. P. R . Vijayasarathy. R. B. Mane, and G . S. Krishna Rao, J. Chern. SOC.,Perkin I, 34 (1977). 66. M. Ando, S. Ibe, S. Kagabu, T. Nakagawa, T. Asao, and mun., 1538 (1970).

K. Takase,

Chem. Com-

67. 0. P. Vig. J . Chander, and B. Ram, IndianJ. Chem.. 12, 1156 (1974). 68. E. Givaudi, M. Plattier and P. Teisseire, Recherrheg 19, 205 (1974). 69. F. Delay and G . Ohloff, Helv. Chim. Acfa., 62. 369 (1979).

70. (a) 0. P. Vig, S. D. Sharrna, K. L. Matta. and J. M. Sehgal, J. Indian Chem. SOC., 48, 993 (1971). (b) See ref. 2. Scheme 2, page 237. 71. 0. P. Vig, B. Ram, C. P. Khera, and J. Chander, Indian J. Chem., 8, 955 (1970).

72. D. J. Faulkner and L. E. Wolinsky, J. Org. Chem., 40, 389 (1975).

J. Org. Chem., 41, 697 (1976). 74. R. J. Crawford, W. F. Erman, and C. D. Broaddus, J. Am. Chem. SOC.,94, 4298

73. L. E. Wolinsky and D. J. Faulkner, (1972).

75. A. Manjarrez, T. Rios, and A. Guzrnin, Tetrahedron, 20, 333 (1964). 76. B. Cazes and S. Julia, Tetrahedron Lerr., 4065 (1978). 77. J . Alexander and G . S. Krishna Rao, Indian J. Chem., 11, 859 (1973).

78. J. H. Babler, D. 0. Olsen, and W. H. Arnold, J. Org. Chem., 39, 1656 (1974). 79. See ref. 2, Scheme I , page 235. 80. A. Manjarrez and A. Guzrnin, J. Org. Chem., 31, 348 (1966); see also ref. 2, page 236.

81. G . D.Gutsche, J. R. Maycock, and C. T. Chang, Terrahedron, 24, 859 (1968). 82.

w.Rittersdorf and F. Cramer,

Tetrahedron, 24, 43 (1968).

83. J. M Forrester and T. Money, Can. J. Chem., 50, 3310 (1972). 84. W. Knoll and C. Tamm, Helv. Chim. Acfa., 58, 1162 (1975).

85. M.A. SchwartzandG. C. Swanson, J. Ore. Chem., 44, 953 (1979).

86. T. Iwashita, T. Kusurni, and H. Kakisawa, Chem. Leu., 947 (1979).

References 87.

Y. Gopichand and K. K. Chakravarti,

523

Tetrahedron Lett., 3851 (1974).

88. K. Suga, S. Watanabe, and T. Fujita, Aust. J. Chem., 25, 2393 (1972).

89. 0. P. Vig, S. S. Bari, S. D. Sharma, and M. Lal, Indian J. Chem, 14B, 929 (1976).

90. See ref. 2, Schemes 2 and 3, pages 237 and 239.

91. F. L. Malanco and L. A. Maldonado, Synth. Commun., 6, 515 (1976).

92. A. Kergomard and H. Veschambre, Tetrahedron Leu., 4069 (1976). 93. (a) I. H. Rogers, J. F. Manville, T. Sahota, Can. J. Chem., 52, 1192 (1974). (b) J.F. Manville, ibid 53, 1579 (1979); ibid 54, 2365 (1976). 94. V. Cernf, L. DolejS, L. Labler, F. S6rm, and K. Sla’ma, Tetrahedron Lett., 1053 (1967); Coil. Czech. Chem. Commun., 32, 3926 (1967). 95. See ref. 2, pages 253-262.

96. J. Ficini, J. d’Angelo, and J. Noiri, J. Am. Chem. Soc., 96, 1213 (1974).

97. G. Farges and H. Veschambre, Bull. SOC.Chim. Frunce, 3172 (1973). 98. E. Negishi, M. Sabanski, J. -J. Katz, and H. C. Brown, Tetrahedron, 32, 925 (1 976). 99. D. A. Evans and J. V. Nelson, J. Am. Chem. Soc., 102, 774 (1980).

100. S. D. Larsen and S. A. Monti, Synth. Comrnun., 9, 141 (1979). 101. K. Uneyama and S. Torii, Tetrahedron Lett., 443 (1976). 102. F. W. Sum and L. Weiler, Tetrahedron Lett., 707 (1979).

103. J. W. Cornforth, B. V. Milborrow, and G. Ryback, Nature, 206, 715 (1975). 104. M. Mousseron-Canet, J. -C. Mani, J. -P. Dalle, and J. -L. Olivi, Bull. SOC.Chim. France, 3874 (1966). 105. D. L. Roberts, R. A. Heckman, B. P. Hege, and S. A. Bellin, J. Org. Chem., 33, 3566 (1968). 106. J. A. Findlay and W. D. MacKay, Can. J. Chem., 49, 2369 (1971). 107. K. Ohkuma, Agr. Bid. Chem. (Tokyo), 29, 962 (1965); 30, 434 (1966). 108. K. Mori, Tetrahedron, 30, 1065 (1974). 109. T. Oritani and K. Yamashita, Tetrahedron Lett., 2521 (1972).

110. F. Kienzle, H. Mayer, R. E. Minder, and W. Thommen, Helv. Chim. Acta., 61, 2616 (1978). 111. R. C. Cookson and P. Lombardi, Gun. Chim. ftal., 105, 621 (1975).

112. P. Lombardi, R. C. Cookson, H. P. Weber, W. Renold, A. Hauser, K, H. Schulte-Eke, B. Willhalm, W. Thomner, and G. Ohloff, Helu. Chim. Act% 59, 1158 (1976).

113. T. R. Hoye and M. J. Kurth, J. Org. Chem., 44, 3461 (1979).

114. R. Baker, P. H. Briner, and D. A. Evans, J. Chem. SOC.,Chem. Commun., 981 (1978).

524

References

115. T. R. Hoye and M . J. Kurth, J. Am. Chem. Soc., 101 5065 (1979).

116. T. Kato, 1. Ichinose, A. Kamoshida, and Y. Kitahara, J. Chem. SOC. Chem. Commun., 518 (1976). 117. A. 0. Gonzilez, J. D. Martin, C. Perez, and M. A. Rami'rez, Tetrahedron Lett., 137 (1976). 1 1 8. A. 0. Gonzilez, J. D. Marti'n, and M. A. Meliin. Tetrahedron Lett., 2279 (1976). 119. J. Froborg, G. Magnusson, and S. Thoren, Acta Chem. Stand., B28, 265 (1974). Perkin I, 2740 (1979). 120. M. Kato, H. Kurihara, and A. Yoshikoshi. J. Chem. SOC., 121. G . Majetich, P. A. Grieco, and M. Nishizawa, J. Org. Chem., 42,2327 (1977). 122. (a) K. Kato, Y. Hirata, and S. Yamamura, Chem. Commun., 1324 (1970). (b) K. Kato, Y. Hirata, and S. Yamamura, Tetrahedron, 27, 5987 (1971). 123. G. Fra'ter, Helv. Chim. Acta, 61, 2709 (1978).

124. J. R . Williams and J. F. Callahan, J. Chem. Soc.. Chem. Commun., 404 (1979). 125. (a) C. Alexandre and F. Rouessac. J. Chem. SOC.,Chem. Commun., 275 (1975). (b) C. Alexandre and F. Rouessac, Bull. SOC.Chim. France, 117 (1977). 126. (a) M. Ando, A. Akahane, and K. Takase, Bull. Chem. SOC. Japan, 51, 283 (1978). (b) M. Ando, K. Tajima, and K. Takase, Chem. Lett., 617 (1978). 127. (a) P. A. Grieco and M. Nishizawa, J. Chem. Soc.. Chem. Commun., 582 (1976). (b) M. Nishizawa, P. A. Grieco, S. D. Burke, and W. Metz, ibid., 76 (1978). 128. (a) P. A. Grieco. M. Nishizawa, S. D. Burke, and N. Marinovic, J. Am. Chem. Soc., 98, 1612 (1976). (b) P. A. Grieco, M. Nishizawa, T. Oguri, S. D. Burke, and N. Marinovic, ibid., 99, 5773 (1977). 129. (a) S. Danishefsky, T. Kitahara, P. F. Schuda, and S. J. Etheredge, J. Am. Chem. SOC., 98, 3028 (1976). (b) S. Danishefsky, P. F. Schuda, T. Kitahara, and S. J. Etheredge, ibid., 99, 6066 (1977). 130. C. G. Chavdarian and C. H. Heathcock, J. Org. Chem., 40,2970 (1975). 131. G. R. Kieczykowski and R. H. Schlessinger, J. Am. Chem. Soc., 100, 1938 (1978). 132. (a) H. lio, M. Isobe, T. Kawai, and T. Goto. Tetrahedron, 35, 941 (1979). (b) M. Isobe, H. lio, T. Kawai, and T. Goto, J. Am. Chem. SOC.,100, 1940 (1978). (c) H. Iio, M. lsobe, T. Kawai, and T.Goto, ibid, 101, 6076 (1979). 133. R. D. Clark and C. H. Heathcock, Tetrahedron Lett., 1713 (1974); J. Org. Chem., 41, 1396 (1976). 134. P. M. Wege, R. D. Clark, andC. H. Heathcock, J. Org. Chem., 41, 1396 (1976). 135. C. G. Chavdarian, S. L. Woo, R. D. Clark, and C. H. Heathcock, Tetrahedron Leu., 1769 (1976). 136. S. Torii, T. Okamolo, and S. Kadono, Chem. Lett., 495 (1977). 137. R . Scheffold, L. RCvCs, J . Aebersold, and A. Schaltegger, Chimia, 30, 57 (1976).

References

525

138. F. Zutterman, P. de Clerq, and M. Vandewalle, Tetrahedron Lett., 3191 (1977). 139. J. A. Marshall and G. A. Flynn, J. Org. Chem., 44, 1391 (1979). 140. J. J. Looker, D. P. Maier, and T. H. Regan, J. Org. Chem., 37, 3401 (1972). 141. Y. Fukuyama, T. Tokoroyama, and T. Kubota, Tetrahedron Lett., 4869 (1973). 142. P. A. Grieco, T. Oguri, C. -L. J. Wang, and E. Williams, J. Org. Chem., 42, 4113 (1977). 143. P.A. Grieco, T. Oguri, S. Gilman, and G. T. De Titta, J. Am. Chem. Soc., 100, 1616 (1978). 144. A. Murai, M. Ono, A. Abiko, and T. Masamune, .I. Am. Chem. SOC., 100, 7751 (1978). 145. P. S. Wharton, C. E. Sundin, D. W. Johnson, and H. C. Kluender, J. Org. Chem., 37, 34 (1972). 146. (a) M. Kodama, Y. Matsuki, and ShB It6, Tetrahedron Lett., 1121 (1976). (b) M. Kodama, S. Yokoo, H. Yamada, and ShB It;, ibid., 3121 (1978). 147. W. C. Still, J. Am. Chem. SOC.,99, 4186 (1977). 148. P. A. Grieco and M. Nishizawa, J. Org. Chem., 42, 1717 (1977). 149. Y. Fujimoto, T. Shimizu, and T. Tatsuno, Tetrahedron Letr., 2041 (1976). 150. G. Lange and F. C. McCarthy, Tetrahedron Lett., 4749 (1978). 101, 2493 (1979). 151. W. C. Still, J. Am. Chem. SOC.,

152. 0. P. Vig, B. Ram, K. S. Atwal, and S. S. Bari, Indian J. Chem., 148, 855 (1976).

153. See ref. 2, Scheme 38, pages 280-281. 154. Y. Kitagawa, A. Itoh, S. Hashimoto, SOC.,99, 3864 (1977).

H.Yamamoto, and H. Nozaki, J. Am.

Chem.

155. T. -L.Ho, Chem. and Ind., 487 (1971). 156. T. -L. Ho, Can. J. Chem., 50, 1098 (1972).

157. T. -L. Ho, J. Chem. SOC., PerkinI, 2579 (1973).

158. P. A. Reddy and G. S. Krishna Rao, J. Chem. Soc., Perkin 4 237 (1979).

159. (a) N. Katsui, A. Matsunaga, K. Imaizumi, T. Masamune, and K. Tomiyama, Tetrahedron Lett., 83 (1971). (b) N. Katsui, A. Matsunaga, K. Imaizumi, T. Masamune, and K. Tomiyama, Bull. Chem. SOC. Japan, 45, 2871 (1972).

160. F. Bohlmann and E. Eickeler, Chem. Ber., 112, 2811 (1979). 161. See ref. 2, Scheme 5, page 292. 162. D. Caine and J. T. Gupton, 111, J. Org. Chem., 39, 2654 (1974).

163. R. B. Gammill and T. A. Bryson, Synrh. Commun., 6, 209 (1976). 164. (a) J. W. Huffman and M. L. Mole, Tetrahedron Letr., 501 (1971). (b) J. W. Huffman and M. L. Mole, J. Org. Chem., 37, 13 (1972).

526

References

165. While it is true that the Marshall synthesis provides acid 57 in less than 1% overall yield from 2-methylcyclohexanone (ref. 2, Scheme 4, page 290), the Heathcock-Kelly synthesis affords /3-eudesmol in more than 5% overall yield from 2-methylcyclohexanone (ref. 2, Scheme 6, page 293). 166. R. G. Carlson and E. G. Zey, J. Org. Chem., 37, 2468 (1972). 167. C. H. Heathcock and T. R. Kelly, Terrohedron, 24, 1801 (1968). 168. 9. D. MacKenzie, M. M. Angelo, and J. Wolinsky, J. Org. Chem., 44, 4042 (1979). 169. M. A. Schwartz, J. D. Crowell, and J. H. Musser, J. Am. Chem. Soc., 94, 4361 ( 1972). 170. P. A. Wender and J. C. Lechleiter. J. Am. Chem. Soc., 100, 4321 (1978). 171. R. 9. Kelly. S. J. Alward, K. Suryanarayana Murty, and J. 9. Stothers, Con. J. Chem., 56, 2508 (1978). 172. D.J. Goldsmith and 1. Sakano, J. Org. Chem., 41, 2095 (1976). 173. See also ref. 2, Scheme 57, pages 372-373. 174. G. Biichi and H.Wuest, J. Org. Chem., 44, 546 (1979). 175. See ref. 2, Scheme 17, page 303. 176. T.R. Kelly, J. Org. Chem..37, 3393 (1972). 177. R. Baker, D. A. Evans, and P.G. McDowell, J. Chem. Soc., Chem. Commtm., 111 (1977). 178. See also ref. 2. Scheme 22, page 309. 179. A. Murai, K. Nishizakura, N. Katsui, and T. Masamune, Tetrahedron Lett., 4399 (1975).

180. A. Murai, H. Taketsuru, K. Fujisawa. Y. Nakahara, M. Takasugi, and T. Masamune, Chem. Lert., 665 (1977). 181. (a) A. G. Hortmann, 4th Midwest Regional Meeting of the American Chemical Society, Manhattan, Kansas, November 1, 1968. (b) A. G. Hortmann, D. S. Daniel, and J. E. Martinelli, J. Org. Chem., 38, 728 (1973). 182. See ref. 2, page 309. 183. A. G. Hortmann, J. E. Martinelli, and Y.Wang, J. Org. Chem., 34, 732 (1969). 184. M. Ando, K. Nanaumi, T. Nakagawa, T. Asao, and K. Takase, Tetrahedron Letr., 3891 (1970). 185. See ref. 2, Scheme 19, pages 305-306. 186. M. Sergent, M. Mongrain, and P. Deslongchamps, Can. J. Chem., 50, 336 (1972). 187. Y.Amano and C. H. Heathcock, Con J. Chem., 50, 340 (1972). 188. See ref. 2. Scheme 22, pages 309-310. 189. D. S. Watt and E. J. Corey, Tetrahedron Lett., 4651 (1972).

References

527

190. J. A. Marshall and P. G. M. Wuts, J. Org. Chem., 42, 1794 (1977). 191. See ref. 2, pages 315-326.

192. J. A.. Marshall and P. G. M.Wuts, J. Org. Chem., 43, 1086 (1978).

193. D. Caine and G. Hasenhuettl. Tetrahedron Lett., 743 (1975). 194. P. A. Grieco and M.Nishizawa, J Chem. SOC.,Chem. Commun., 582 (1976). 195. D. Becker, N. C. Brodsky, and J. Kalo, J. Org. Chem., 43, 2557 (1978). 196. See ref. 2, Scheme 4, page 290. 197. G. H. Posner and G. L. Loomis, J. Org. Chem., 38, 4459 (1973). 198. See ref. 2, Scheme 30, pages 326-327. 199. See ref. 2, Scheme 23, pages 311-312. 200. R. B. Miller and E. S. Behare, J. Am. Chem. Soc., 96, 8102 (1974). 201. (a) J. D. Godfrey and A. G. Schultz, Tetrahedron Lett., 3241 (1979). (b) A. G. Schultz and J. D. Godfrey, J. Org. Chem., 41, 3494 (1976). 202. See ref. 2, Schemes 4-7, pages 290-294 and Scheme 2, page 405. 203. W. C. Still and M.J. Schneider, J. Am. Chem. Soc., 99, 948 (1977).

204. (a) F. Kido, R. Maruta, K. Tsutsumi, and A. Yoshikoshi, Chem. Lett., 311 (1979). (b) F. Kido, K. Tsutsurni, R. Maruta, and A. Yoshikoshi, J. Am. Chem. Soc., 101, 6420 (1979). 205. J. Alexander and G. S. Krishna Rao, Tetrahedron, 27, 645 (1971). 206. V. Viswanatha and G. S. Krishna Rao, Tetrahedron Lett., 243 (1974).

207. V. Viswanatha and G. S. Krishna Rao, Tetrahedron Lett., 247 (1974).

208. V. Viswanatha and G. S. Krishna Rao, Indian J. Chem., 11, 974 (1973). 209. V. Viswanatha and G. S. Krishna Rao, J. Chem. SOC.,Perkin I, 450 (1974). 210. J. P. McCormick, J. P. Pachlatko, and T. R. Schafer, Tetrahedron Lett., 3993 (1978). 211. P. B. Talukdar, J. Org. Chem., 21, 506 (1956). 212. 0. P. Vig, S. D. Sharma, G. L. Kad, and M. L. Sharma, Ind. J. Chem., 13, 764 (1975). 213. See ref. 2, Scheme 38, page 336. For full experimental details see L. A. Burk and M. D. Soffer, Tetrahedron, 32. 2083 (1976).

214. M. D. Soffer and L. A. Burk, Tetrahedron Lett., 21 1 (1970). 215. (a) R. P. Gregson and R. N. Mirrington, J. Chem. Soc., Chem. Commun., 598 (1973). (b) R. P. Gregson and R. N. Mirrington, Aus. J. Chem., 29, 2037 (1976).

216. (a) M. Iguchi, M. Niwa, and S. Yamamura, Bull. Chern. SOC.J a p 4 46, 2920 (1973). (b) K. Kato, Y. Hirata, and S. Yamamura, Tetrahedron, 27, 5987 (1971).

528

References

217. V. K. Belavadi and S. N. Kulkarni, Indian 1. Chem., 14B, 901 (1976). 218. D. Caine and A. A. Frobese, Tetrahedron Lett., 3107 (1977).

219. D. F. Taber and B. P. Gunn, J. Am. Chem. Soc., 101, 3992 (1979). See also the 0. P. Vig 1977 (*)-kushitene synthesis, Scheme 85. 220. J. R. Hlubucek, A. J. Aasen. S -0. Almquist, and C. R. Enzell, Acra Chem. Scund., B28, 289 (1974). 221. A. J. Aasen, C. H. G. Vogt, and C. (1975).

R. Enzell,

Acra. Chem. Scand., B29, 51

222. See ref. 2, Schemes 41-42, pages 340-342. 223. See ref. 2, pages 339-340. 224. See ref. 2, Scheme 46, pages 347-348. 225. H. Akita, T. Naito, and T. Oishi, Chem. Lett., 1365 (1979). 226. See ref. 2, Scheme 44, page 344.

227. (a) T. Suzuki, M. Tanemura, T. Kato, and Y. Kitahara, Bull Chem. SOC.Jupua 43. 1268 (1970). (b) H. Yanagawa, T. Kato, and Y. Kitahara, Synthesis 257 (1 970). 228. T. Kato, T. Suzuki, M. Tanemura, A. S . Kumanireng, N. Ototani, and Y. Kitahara, Tetruhedron Lett., 1961 (1971).

229. T. Kato, T. Iida, T. Suzuki, Y. Kitahara, and K. H. Overton, Terruhedron Lert., 4257 (1972). 230. A. Ohsuka and A. Matsukawa, Chem. Lett., 635 (1979).

231. T.Nakata, H.Akita, T. Naito, and T. Oishi, J. Am. Chem. Soc., 101, 4400 (1979). 232. S.P. Tanis and K. Nakanishi, J. Am. Chem. Soc., 101, 4398 (1979). 233. See ref. 2, Scheme 48, page 352.

234. D. Nasipuri and G. Das, J. Chem. Soc., Perkin I, 2776 (1979). 235. T. Matsumoto and S. Usui, Chem. Lert., 105 (1978). 236. J. A. Marshall and R. A. Ruden, J. Org. Chem., 36. 594 (1971). 237. J. A. Marshall and T.

M.Warne, Jr. J. Org. Chem., 36, 178 (1971).

238. See ref. 2, Schemes 53 and 54, pages 364-366.

239. H. C. Odom, Jr. and A. R. Pinder, J. Chem. Soc., Perkin I, 2193 (1972). 240. See ref. 2, pages 371-372. 241. Y. Takagi. Y. Nakahara, and M. Matsui, Teiruhedroa 34, 517 (1978).

242. H. M. McGuire, H. C. Odom, Jr., and A.R. Pinder, J. Chem. Soc., Perkin 4 1879 (1974). 243. K. P. Dastur, J. Am. Chem. Soc., 95, 6509 (1973); 96, 2605 (1974).

244. C. H. Heathcock and R. D. Clark, unpublished results.

References

529

245. T.Hiyama, M. Shinoda, and H.Nozaki, Tetrahedron Lett., 3529 (1979). 246. T. Yanami, M. Miyashita, and A. Yoshikoshi, J. Chem. SOC.,Chem. Commun., 525 (1979). 247. D. Caine and S. L. Graham, Tetrahedron Lett., 2521 (19761, and private communication. 248. See ref. 2, Schemes 57 and 60, pages 372 and 370. 249. E. Piers and R. D. Smillie, J. Org. Chem., 35, 3997 (1970). 250. A. K. Torrence and A. R. Pinder, Tetrohedron Lett., 745 (1971).

251. J. A. Marshall and G. M. Cohen, J. Org. Chem., 36, 877 (1971). 252. S. Torii, T.Inokuchi, and T.Yamafuji, Bull. Chem. SOC.Jopon, 52, 2640 (1979). 253. (a) K. Yamakawa, 1. Izuta, H. Oka, and R. Sakaguchi, Tetruhedron Lett., 2187 (1974). (b) K. Yamakawa, I. Izuta, H. Oka, R. Sakaquchi, M.Kobayashi, and T. Satoh, Chem. Phurm. Bull., 21, 331 (1979). 254. S. Torii,

T.Inokuchi, and K. Kawai, Bull. Chem. SOC. Jopon, 52, 861 (1979).

255. F. E. Ziegler and P. A. Wender, Tetrahedron Lett., 449 (1974).

256. F. E. Ziegler, G. R. Reid, W. L. Studt, and P. A. Wender, J. Org. Chem., 42, 1991 (1977). 257. J. E. McMurry, J. H. Musser, M. S. Ahmad, and L. C. Blaszczak, J. Org. Chem., 40, 1829 (1975). 258. J. Ficini and A. M . Touzin, Tetrohedron Lett., 1081 (1977). 259. F. Naf, R. Decorzant, and W. Thornmen, Helv. Chim. Act4 62, 114 (1979). 260. E. Piers, M. B. Geraghty, and R. D. Smillie, J. Chem. SOC.,Chem. Commun., 614 (19711. 261. E. Piers and M. B. Geraghty, Cun. J. Chem.. 51, 2166 (1973). 262. See ref. 2, Scheme 57, pages 372-373 and Scheme 61 of this section. 263.

T.Tatee and T.Takahashi, Chem. Lett., 929 (1973); Bull. Chem. SOC. Japan, 48, 281 (1975).

264. I. Nagakura, S. Maeda, M. Ueno, M. Funamiru, and Y. Kitahara, Chem. Letr., 1143 (1975). 265. F. Bohlmann, H. -J. Forster, and C. H. Fischer, Ann., 1487 (1976). 266. K. Yamakawa and T. Satoh, Chem. Phorm. Bull. (Tokyo) 25, 2535 (1977); Heterocycles, 8, 221 (1977); Chem. Phorm. Bull. (Tokyo), 26, 3704 (1978). 267. M. Miyashita, T.Kumazawa, and A. Yoshikoshi, Chem. Lett., 163 (1979). 268. Y. Inouye, Y. Uchida, and H. Kakisawa, Chem. Lett., 1317 (1975). 269. F. Yuste and F. Walls, Aust. J. Chem., 29, 2333 (1976). 270. J. W. Huffman and R. Pandian, J. Org. Chem., 44. 1851 (1979). 271. D. K. Banerjee and V . B. Angadi, Ind. J. Chem., 11, 511 (1973).

530

References

272. G. H. Posner, C. E. Whitten, J. J. Sterling, and D. J. Brunelle, Tetrahedron Lett., 2591 (1974). 273. S. W. Baldwin and R. E. Gawky, Tetruhedron Lett., 3969 (1975). 274. 0. P. Vig. S. D. Sharma. 0. P. Chugh, and A. K. Vig, Ind. J. Chem., 12, 1050 (1 974). 275. 0. P. Vig, I. R. Trehan, and R. Kumar, Ind. J. Chem., 15B, 319 (1977). 276. R. K. Boeckman, Jr. and S. M. Silver, Tetruhedron Lert., 3497 (1973); J. Org. Chem., 40, 1755 (1975). 277. R. A. Moss, E. Y. Chen, J. Banger, and M. Matsuo, Terruhedron Leff., 4365 (1 978). 278. See ref. 2, Scheme 4, page 290. 279. See ref. 2, page 286. 280. M. Ando, S. Sayama, and K. Takase, Chem. Lett., 191 (1979). 281. See ref. 2. Scheme 5 , page 292. 282. D. Caine, P. Chen, A. S. Frobese, and J. T. Gupton, J. Org. Chem., 44, 4981 (1979). 283. J. E. McMurry and L. C. Blaszczak, J. Org. Chem., 39, 2217 (1974). 284. See ref. 2, Scheme 70, page 394. 285. S. J. Branca, R. L. Lock, and A. B. Smith, Ill., J. Org. Chem., 42, 3165 (1977). 286. M. Nakayama, S. Ohira, S. Shinke. Y. Matsushita, A. Matsuo, and S. Hayashi, Chem. Lert., 1249 (1975). 287. Y. Hayashi, M. Nishizawa, and T. Sakan, Chem. Lert., 387 (1975). 288. 0. P. Vig, R. K. Parti, K. C. Gupta, and M. S. Bhatia, Ind. J. Chem., 11, 981 (1973). 289. C. W. Bird, Y.C. Yeong, and J. Hudec, Synthesis, No.1, 27 (1974). 290. P. de Mayo and R. Suau, J. Chem. Soc., Perkin I, 2559 (1974). 291. T. Kametani, M. Tsubuki, and H. Nemoto. Heterocycles, 12, 791 (1979). 292. E. Wenkert, B. L. Buckwalter, A. A. Craveiro, E. L. Sanchez, and S. S. Sathe, J. Am. Chem. Soc., 100. 1267 (1978). 293. R. B. Mane and G. S. Krishna Rao, J . Chem. Soc., Perkin I, 1806 (1973). 294. A. Casares and L. A. Maldonado, Synth. Commun., 6, 11 (1976). 295. T. hie, T. Suzuki, Y. Yasunari, E. Kurosawa, and T. Masamune, Terruhedron, 25, 459 (1969). 296. J. E. McMurry and L. A. von Beroldingen, Terruhedron, 30, 2027 (1974). 297. R. C. Ronald, Terruhedron Letr.. 4413 (1976).

References

531

298. (a) E. W. Colvin, R. A. Raphael, and J. S. Roberts, Chem. Commun., 858 (1971). (b) E. W. Colvin, S. Malchenko. R. A. Raphael, and J. S . Roberts, J. Chem. SOC. Perkin I, 1989 (1973). 299. Y. Fujimoto, S. Yokura, T.Nakamura, T. Morikawa, and T. Tatsuno, Tetrahedron Lett., 2523 (1974).

300. See ref. 2, Schemes 53 and 54, pages 364-365. 301. N. Masuoka and T. Kamikawa, Tetrahedron Lett., 1691 (1976). 302. Note that the structure of 870, which is apparently a single diastereomer, is drawn incorrectly in ref. 301. 303. W. C. Still and M. -Y. Tsai, J. Am. Chem. Soc., 102, 3654 (1980). 304. (a) S. C. Welch, A. S. C. P. Rao, and R. Y. Wong, Synth. Commun., 6, 443 (1976). (b) S. C. Welch, A. S. C. P. Rao, and C. G. Gibbs, ibid., 485 (1976).

305. (a). S. C. Welch and R. Y. Wong, Tetrahedron Lett., 1853 (1972). Welch and R . Y. Wong, Synth. Commun., 2, 291 (1972).

(b). S. C.

306. K. Yamakawa, R. Sakaguchi, T. Nakamura, and K. Watanabe, Chem. Len, 991, (1976). 307. G. 1. Feutrill, R. N. Mirrington and R. J. Nichols, Aust. J. Chem., 26, 345 (1973). 308. K. Yamada, H. Yazawa. D. Uemura, M. Toda, and Y. Hirata, Tetrahedron 25, 3509 (1969). 309. G. L. Hodgson, D. F. MacSweeney and T. Money, Chem. Commun., 766 (1971). 310. G. L. Hodgson, D. F. MacSweeney, and T. Money, Tetrahedron Lert., 3683 (1972). 311. G. L. Hodgson, D. F. MacSweeney, R. W. Mills, and T. Money, J. Chem. Soc., Chem. Commun., 235 (1973). 312. E. J. Corey, H. A. Kirst, and J. A. Katzenellenbogen, J. Am. Chem. Soc., 92, 6314 (1970). 313. See ref. 2, Scheme 6, page 205. 314. M. Julia and P. Ward, Bull. SOC.,Chem. France, 3065 (1973). 315. K. Sato, S. Inoue, Y. Takagi, and S. Morii, Bull. Chem. SOC.J a p R 49, 3351 (1 976). 316. S. A. Monti and S. D. Larsen, J. Org. Chem., 43, 2282 (1978). 99, 8015 (1977). 317. S. D. Larsen and S. A. Monti, J. Am. Chem. SOC.,

318. M. Bertrand, H. Monti, and K.C. Huong, Tetrahedron Lett., 15 (1979). 319. See ref. 2, Scheme 44, page 483. 320. P. A. Christenson and B. J. Willis, J. Org. Chem., 44, 2012 (1979). 321. See ref. 2, Scheme 50, page 489. M.Baumann and W. Hoffmann, Liebigs Ann. Chem., 743 (1979).

322.

532

References

323. See ref. 2, Schemes 15 and 16, pages 250-251. 324. E. J. Corey, D. E. Cane, and L. Libit, J. Am. Chem. SOC.,93, 7016 (1971).

325. Y. Bessiere-Chretien and C. Grison, C. R. Acad. Sci., Ser C, 275, 503 (1972). 326. See ref. 2, Scheme 26, page 447.

327. (a) P. A. Grieco and Y. Masaki, J. Org. Chem., 40, 150 (1975). (b) P. A. Grieco and J. J. Reap, Synth. Commun., 5, 347 (1975). 328. J. A. Marshall, N. H. Andersen and P. C. Johnson, 1. Am. Chem. SOC.,89, 2748 (1967). 329. J. A. Marshall and S. F. Brady, J. Org. Chem., 35, 4068 (1970). 330. See ref. 2, Scheme 37, pages 469-470. 331. G.Stork, R. L. Danheiser, and B. Ganem, J. Am. Chem. SOC.,95, 3414 (1973).

332. (a) P. M. McCurry, Jr., and R. K. Singh, Tefruhedron Lett., 3325 (1973). (b) M. McCurry, Jr., R. K. Singh, and S. Link, ibid., 1155 (1973).

P.

333. K. Yamada, H. Nagase, Y. Hayakawa, K. Aoki, and Y. Hirata, Tetrahedron Lett., 4963 (1973). 334. K. Yamada, K. Aoki, H. Nagase, Y. Hayakawa, and Y. Hirata, Tetrahedron Left., 4967 (1973). 335. K. Yamada, S. Goto, H.Nagase, and A. T. Christensen, J. Chem. Soc., Chem. Commun., 554 (1977). 336. D. Caine and C -Y.Chu, Tetrahedron Lett., 703 (1974).

337. G. Bozzato, J -P. Bachmann, and M.Pesaro, J. Chem. Soc.. Chem. Commun., 1005 (1974).

338. B. M. Trost, M. Preckel, and L. M. Leichter, J. Am. Chem. SOC.,97, 2224 (1975). 339. (a) D. Buddhasukh and P. Magnus, 1. Chem. Soc., Chem. Commun., 952 (1975). (b) D. A. Chass. D. Buddhasukh, and P. D. Magnus, J. Org. Chem., 43, 1750 (1978). 340. W. G. Dauben and D. J. Hart, J. Am. Chem. Soc., 97, 1622 (1975); 99, 7307 (1977). 341. M. Deighton, C. R. Hughes, and R. Ramage, J. Chem. SOC.,Chem. Commun., 662 (1 975).

342. D. Caine, A. A. Boucugnani, S. T. Chao, J. 8. Dawson, and P. F. Ingwalson, J.

Org. Chem., 41, 1539 (1976). 343. D. Caine, A. A. Boucugnani and W. R. Pennington, J. Org. Chem., 41, 3632 (1976). 344. G . Biichi, D. Berthet, R. Decorzant, A. Grieder, and A. Hauser, J. Org. Chem., 41, 3208 (1976).

345. (a) K. Uneyama, K. Okamoto, and S. Torii, Chem. Lett., 493 (1977). (b) S. Torii, K. Uneyama. and K. Okamoto, Bull. Chem. SOC.J u p u ~51, 3590 (1S78).

References

533

346. See ref. 2, Scheme 36, page 467. 347. (a) T. Ibuka, K. Hayashi, H. Minakata. and Y. Inubushi, Tetrahedron Lett., 159 (1979). (b) T. Ibuka, K. Hayashi, H. Minakata, Y. Ito, and Y. Inubushi, Can. J. Chem., 57, 1579 (1979).

348. J. M. Conia, J. P. Drouet, and J. Gore, Tetruhedron, 27, 2481 (1971).

349. A. R. Pinder, S. J. Price, and R. M. Rice, J. Org. Chem., 37, 2202 (1972).

350. T. R. Kasturi and M. Thomas, Indian J. Chem., 10, 777 (1972). 351. P.Naegeli and R. Kaiser, Tetrahedron Letr., 2013 (1972).

352. I. G. Guest, C. R. Hughes, R. Ramage, and A. Sattar, J. Chem. Soc., Chem. Commun., 526 (1973). 353. (a) J. N. Marx and L. R. Norman, Tetrahedron Lett., 4375 (1973). (b) J. N. Marx and L. R. Norman, J. Org. Chem., 40, 1602 (1975). 354. (a) W. Oppolzer, Helv. Chim. Act4 56, 1812 (1973). (b) W. Oppolzer and K. K. Mahalanabis, Tetrahedron Lett., 3411 (1975). (c) W. Oppolzer, K. K. Mahalanabis and K. Battig, Helu. Chim. Act4 60, 2388 (1977).

355. H. Wolf and M. Kolleck, Tetrahedron Lett., 451 (1975).

356. H. Wolf, M. Kotteck, K. Claussen and W. Rascher, Chem. Ber., 109, 41 (1976). 357. B. M. Trost, K. Hiroi, and N. Holy, J. Am. Chem. Soc., 97,5873 (1975). 358. J. F. Ruppert, M. A. Avery, and J. D. White, J. Chem. Soc., Chem. Commun., 978 (1976). 359. D. A. McCrae and L. Dolby, J. Org. Chem., 42, 1607 (1977). 360. (a) G. L. Langc, W. J. Orrom, and D. J. Wallace, Tetrahedron Leu., 4479 (1977). (b) G . L. Lange, E. E. Neidert, W. J. Orrom, and D. J. Wallace, Can. J. Chem., 56, 1628 (1978). 361. M. Pesaro and J -P. Bachmann, J. Chem. Soc., Chem. Commun., 203 (1978). 362. S . F. Martin and T. Chou, J. Org. Chem., 43, 1027 (1978).

363. D. Caine and H. Deutsch, J. Am. Chem. Soc., 100, 8030 (1978). 364. See ref. 2, page 286.

365. J. D. White, S. Torii, and J. Nogami, Tetrahedron Leu., 2879 (1974). 366. G. Friter, Helv. Chim. Act4 60, 515 (1977).

367. C. Iwata, M. Yarnada, and Y. Shinoo, Chem. Pharm. Bull., 27, 274 (1979). 368. L. E. Wolinsky and D. J. Faulkner, J. Org. Chem., 41, 597 (1976).

369. W. Fenical and

B. M. Howard, Tetrahedron Lett., 2519 (1976).

370. 1. lchinose and T. Kato, Chem. Lett., 61 (1979).

371. See ref. 2, Scheme 29, pages 452-453. 372. 0. P. Vig, B. Ram, S. S . Bari, and S . D. Sharma, Indiun J. Chem., 14B, 852 (1976).

534

373. 374. 375. 376. 377. 378. 379. 380. 381. 382. 383. 384. 385. 386. 387. 388. 389. 390. 391. 392. 393. 394.

References

See ref. 2, Scheme 32, page 272. See ref. 2, Scheme 33, page 273. See ref. 2, Scheme 20, pages 435-437. J. J. Plattner and H. Rapoport, J. Am. Chem. Soc., 93, 1758 (1971) C. F. Garbers, J. A. Steenkamp, anji H. E. Visagie, Tetrahedron Lett., 3753 (1975). K. Kitatani, T.Hiyama, and H. Nozaki, J. Am. Chem. Soc., 98, 2362 (1976). See ref. 2, Scheme 16 pages 429-431. S. Danishefsky, M. Hirama, K. Gombatz, T. Harayama, E. Berman, and P. Schuda. J. Am. Chem. Soc., 100, 6536 (1978); ibid., 101, 7020 (1979). W. H. Parsons, R. H. Schlessinger, and M. L. Quesada, J. Am. Chem. Soc., 102, 889 (1980). Y. Hayashi, M. Nishizawa, S. Harita, and T. Sakan. Chem. Lett., 375 (1972). M.E. N. Nambudiry and G . S. Krishna Rao, J. Chem Soc., Perkin I, 317 (1974). Y. Kitahara, N. Abe, T. Kato, and K. Shirahata. Nippon Kagaku Zasshi, 90, 221 ( 1 969). K. Hayashi, H. Nakamura, and H. Mitsuhashi, Chem. Phorm. Bull., 21, 2806 (1973). K. Naya, M. Hayashi, I. Takagi, S. Nakamura, and M. Kobayashi, Bull. Chem. SOC. Jopn, 45, 3673 (1972). (a) D. A. Evans and C. L. Sims Tetruhedron Left., 4691 (1973). (b) D. A. Evans, C. L. Sims, a n d G . C. Andrews, J. Am. Chem. Soc., 9 9 , 5453 (1977). See ref. 2, Scheme 65, page 386. S. Sarma and B. Sarkar, Indian J. Chem., 10, 950 (1972). D. Caine and F. N. Tuller, J. Org. Chem., 38, 3663 (1973). D. F. Taber and R. W. Korsmeyer, J. Org. Chem., 43. 4925 (1978). K. Sisido, S. Kurozumi, K. Utimoto. and T.Isida, J. Org. Chem, 31, 2795 (1966). E. J. Corey and H. L. Pearce, J. Am. Chem. Soc., 101, 5841 (1979). N. H. Andersen and H. Uh, Synth. Commun., 3, 115 (1973).

395. (a) See ref. 2, pages 404-407. (b) C. H. Heathcock and R. Ratcliffe, J. Am. Chem. Soc., 93, 1746 (1971). 396. G . Mehta and B. P. Singh, Tetrahedron Letr., 4495 (1975). 397. See ref. 2 , pages 402-410. 398. (a) J. A. Marshall and A. E. Greene, J. Org. Chem., 37, 982 (1972). (b) J. A. Marshall, A. E. Greene, and R. Ruden, Tetrahedron Left., 855 (1971). (c) J. A. Marshall and A. E. Greene, ibid., 859 (1971).

References

535

399. G.L. Buchanan and G. A. R. Young, J. Chem. SOC.,Perkin I, 2404 (1973); J. Chem. Soc., Chem. Commun., 643 (1971). 400. N. H. Andersen and H. S. Uh, Tetrahedron Lett., 2079 (1973). 401. H -J. Liu and S . P. Lee, Tetrahedron Lett., 3699 (1977). 402. N.H. Andersen and F. A. Golec, Jr., Temhedron Lett., 3783 (1977). 403. J. N. Marx and S . M. McGaughey, Tetrahedron, 28, 3583 (1972). 404. See ref. 2, pages 413-416. 405. M.T. Edgar, A. E. Greene, and P. CrabbC, J. Org. Chem., 44, I59 (1979). 406. See ref. 2, pages 417-423. 407. D. Caine and P. F. Ingwalson, J. Org. Chem., 37, 3751 (1972). 408. J. A. Marshall and J. A. Ruth, J. Org. Chem., 39, 1971 (1974). 409. See ref. 2, Scheme 4, page 408. 410. D. Caine and J. T.Gupton, J. Org. Chem., 40, 809 (1975). 411. J. A. Marshall and W. R. Snyder, J. Org. Chem., 40, 1656 (1975). 412. R. A. Kretchmer and W. J. Thompson, J. Am. Chem. SOC.,98, 3379 (1976). 413. P. DeClercq and M. Vandewalle, J. Org. Chem., 42, 3447 (1977). 414. P. Kok, P.DeClercq, and M. Vandewalle, Bull. Soc., Chem. Belg., 87, 615 (1978). 415. P. A. Grieco, Y. Ohfune, and G. Majetich, J. Am. Chem. SOC., 99, 7393 (1977). 416. (a) R. Peel and J. K. Sutherland, J. Chem. SOC., Chem. Commun., 151 (1974). (b) J. S. Bindra, A. Grodski, T. K. Schaaf, and E. J. Corey, J. Am. Chem. Soc.. 95, 7522 (1973). 417. P. A. Grieco, T. Oguri, S. Burke, E. Rodriguez, G. T. DeTitta, and S. Fortier, J. Org. Chem., 43, 4552 (1978). 418. (a) M.R. Roberts and R. H. Schlessinger J. Am. Chem. Soc., 101, 7626 (1979). (b) G. 1. Quallich and R. H.Schlessinger, ibid., 101, 7627 (1979). 419. P. T. Lansbury and A. K. Serelis, Tetrahedron Lett., 1909 (1978). 420. P. A. Wender, M. A. Eissenstat, and M. P. Filosa, J. Am. Chem. Soc., 101, 2196 (1979). 421. J. A. Marshall and R. H. Ellison, J. Am. Chem. Soc., 98, 4312 (1976). 422. M. F. Semmelhack, A. Yamashita, J. C. Tomesch, and K. Hirotsu J. Am. Chern. SOC.,100, 5565 (1978). 423. C. H.Heathcock, E. G. Delmar, and S. L. Graham, J. Am. Chem. Soc., 104, 1907 (1982). 424. M. Demuynck, P. DeClercq, and M. Vandewalle, J. Org. Chem., 44, 4863 (1979). 425. c. H. Heathcock, T. C. Germroth and C. M. Tice. J. Am. Chern. SOC., 104, 0000 (1982).

536

References

426. (a) Y. Ohfune, P. A. Grieco, C. -L.Wang, G.Majetich, J. Am. Chem. Soc., 100, 5946 (1978). (b) P. A. Grieco, Y. Ohfune, and G. Majetich, J. Org. Chem., 44, 3092 (1979). 427. P. A. Grieco, Y.Ohfune, and G. Majetich. Tetrahedron Leu., 3265 (1979).

428. P. Kok, P. DeClercq, and M. Vandewalle, J. Org. Chem., 44, 4553 (1979). 429. (a) P. T. Lansbury and D. Hangauer, Tetrahedron Leu., 3623 (1979). (b) P. T. Lansbury, D. G.Hangauer, and J. P. Vacca, J. Am. Chem. Soc., 102, 3964 (1980). 430. M. Soucek, Coll. Czech. Chem. Commun., 27, 2929 (1962). 431. M. Yamasaki, J. Chem. SOC.,Chem. Commun., 606 (1972). 432. (a) H. deBroissia, J. Levisalles, and H. Rudler, J. Chem. Soc., Chem. Commun., 855 (1972). (b) idem., Bull. Soc. Chim. Fr., 4314 (1972). 433. (a) J. Froborg, G. Magnusson, and S . ThorCn, J. Org. Chem., 40, 1595 (1975). (b) T. Fex, J. Froborg, G. Magnusson, and S.ThorCn, ibid 41, 3518 (1976). (c) J. Froborg and G.Magnuson, J. Am. Chem. Soc., 100, 6728 (1978). 434. See ref. 2, pages 396-402. 435. E. Wenkert and K. Naemura, Synth. Commun., 3, 45 (1973). 436. G.Mehta and S. K. Kapoor, J. Org. Chem., 39, 2618 (1974).

437. E. Piers and E. H.Ruediger, J. Chem. Soc., Chem. Commun., 166 (1979).

438. A. G. GonzPlez, J. Darias, J. D. Martin, and M. A. Meliln, Tetrahedron Lett., 481 (1978). 439. S. Danishefsky and K. Tsuzuki, J. Am. Chem. Soc., 102, 6891 (1980). 440. See ref. 2, Scheme 42, page 480. 441. M. Bertrand and J. -L.Gras, Terrahedron 30, 793 (1974). 442. A. Kumar, A. Singh, and D. Devaprabhakara, Tetrahedron Left., 2177 (1976). 443. A. Kumar and D. Devaprabhakara. Synthesis 461 (1976). 444. E. Piers, R. W. Britton, and W. de Waal, Can. J . Chem., 49, 12, 1971. 445. A. Tanaka, R. Tanaka, H. Uda, and A. Yoshikoshi, J. Chem. Soc., Perkin I, 1721 (1972).

446. S. Torii and T. Okamoto, Bull. Chem. SOC.Jupa4 49, 771 (1976).

447. See ref. 2, Scheme 17, page 538. 448. T. Matsumoto, K. Miyano, S. Kagawa, S. Yii, J. Ogawa, and A. Ichihara, TefrahedronLeft., 3521 (1971). 449. Y. Ohfune, H. Shirahama, and T. Matsumoto, Tetrahedron Left., 4377 (1975). 450. H.Takeshita, H. Iwabuchi, 1. Kouno, M. lino, and D. Nomura, Chem. Len., 649 (1 979). 451. D. Helmlinger, P. de Mayo, M. Nye, L. Westfelt, and R. B. Yeats, Tetrahedron Lett.. 349 (1970).

References

537

452. S. R. Wilson and R. B. Turner, J. Org. Chem., 38, 2870 (1973). 453. W. J. Greenlee and R. B. Woodward, J. Am. Chem. SOC.,98, 6075 (1976).

454. R. K. Boeckrnan, Jr. and S. S . KO,J. Am. Chem. SOC.,102, 7146 (1980).

455. (a) P. T. Lansbury, N. Y. Wang, and J. E. Rhodes, Tetrahedron Lett, 1829 (1971). (b) P. T. Lansbury and N. Nazarenko, ibid., 1833 (1971). 456. P. T. Lansbury, N. Y.Wang, and J. E. Rhodes, Tetrahedron Lett.;2053 (1972).

457. H. Hashimoto, K. Tsuzuki, F. Sakan, H. Shirahama, and T. Matsurnoto, Tetrahedron Lett., 3745 (1 974). 458. F. Sakan, H. Hashimoto, A. Ichihara, H. Shirahama, and T. Matsurnoto, Tetrahedron Lett., 3703 (1971). 459. B. M. Trost, C. D. Shuey, F. DiNinno, Jr. and S. S . McElvain, J. Am. Chem. Soc., 101, 1284 (1979).

460. K. Tatsuta, K. Akimoto, and M. Kinoshita, J. Am. Chem. Soc., 101, 6116 (1979). 461. K. Tatsuta, K. Akimoto, and M. Kinoshita, J. Antibiot., 33, 100 (1980). 462. (a) S. Danishefsky, R. Zarnboni, M. Kahn, and S. J. Etheredge, J. Am. Chem. SOC.,102, 2097 (1980). (b) S. Danishefsky and R. Zamboni, Tefrahedron Lett., 3439 (1980).

463. M. Shibasaki, K. Iseki, and S. Ikegami, Tetrahedron Leu., 3587 (1980).

464. A. E. Greene, Tetrahedron Left., 3059 (1980).

465. R. D. Little and G. W. Miller, J. Am. Chem. SOC., 101, 7129 (1979).

466. W. Oppolzer, K. Battig, and T. Hudlicky, Helu. Chim. Acta, 62, 1493 (1979).

467. M. C. Pirrung, J. Am. Chem. SOC., 101, 7130 (1979); 103. 82 (1981). 468. L. A. Paquette and Y. K. Han, J. Org. Chem., 44, 4014 (1979). 469. W. G. Dauben and D. M. Walker, J. Org. Chem., 46, 1103 (1981).

470. S. Chatterjee, J. Chem. SOC.,Chem. Commun., 620 (1979). 471. Ref. 468, footnote 16. See also J. Cornforth, Tetrahedron Lett., 709 (1980). 472. R. M. Coates, S. K. Shah, and (1979).

473. G. Biichi and P.

R. W. Mason, J. Am. Chem. SOC., 101, 6765

-S.Chu, J. Am. Chem. SOC.,101, 6767 (1979).

474. S. C. Welch and S. Chayabunjonglerd, J. Am. Chem. SOC.,101, 6768 (1979). 475. Y.-K.Han and L. A. Paquette, J. Org. Chem., 44, 3731 (1979)

476. 477. 478. 479.

See ref. 2, Scheme 11, pages 526-527. J. E. McMurry, J. Org. Chem., 36, 2826 (1971). See ref. 2, page 523. (a) E. Piers, R. W. Britton, R. J. Keziere, and R. D. Srnillie, Can. J. Chem., 49, 2620 (1971). (b) E. Peirs, R. W. Britton, R. J. Keziere, and R. D. Smillie, ibid.,

538

References 49, 2623 (1971). (c) E. Piers, M. B. Geraghty, F. Kido, and M. Soucy, Synth. Commun., 3, 39 (1973). (d) E. Piers, M. B. Geraghty, and M. Soucy, ibid., 3, 401 (1973). (e) E. Piers, R. W. Britton, M. B. Geraghty, R. J. Keziere, and R. D. ) Peirs, R. W. Britton, M. B. Smillie, Can. J. Chem., 53, 2827 (1975). (IE. Geraghty, R. J. Keziere, and F. Kido, ibid., 53, 2838 (1975). (g) E. Piers, M. B. Geraghty, R. D. Smillie, and M. Soucy, ibid., 53, 2849 (1975).

480. (a) G. L. Hodgson, D. F. MacSweeney, and T. Money, Tetrahedron Lett., 3683 (1972). (b) G. L. Hodgson, D. F. MacSweeney, and T. Money, J. Chem. Soc., Perkin I, 2113 (1973). (c) C. R. Eck, G.L. Hodgson, D. F. MacSweeney, R. W. Mills, and T. Money, ibid., 1938 (1974).

481. S. N. Baldwin and J. C. Tomesch, Tetrahedron Lett., 1055 (1975).

482. P. Bakuzis. 0. 0.S. Campos, and M. L. F. Bakuzis, J. Org. Chem., 41, 3261 (1976).

483. M. Yanagiya, K. Kaneko, T. Kaji, and T. Matsurnoto, Tetrahedron Leu., 1761 (1979). 484. (a) E. Piers and H. -P. Isenring, Synth. Commun., 6, 221 (1976). (b) E. Piers and H. -P. Isenring, Can. J. Chern., 55, 1039 (1977). 485. J. E. McMurry and M. G . Silvestri, J . Org. Chem., 41, 3953 (1976). 486. P. A. Collins and D. Wege, Ausi. J. Chem., 32, 1819 (1979). 487. See ref. 2, Scheme 9, pages 518-519.

488. J. E. McMurry and S. J. Isser, J. Am. Chem. Soc., 94, 7132 (1972).

489. R. A. Volkmann, G . C. Andrews, and W. A. Johnson, J. Am. Chem. Su. . 91, 4777 (1975).

490. W. Oppolzer and T. Godel, J. Am. Chem. Soc., 100, 2583 (1978).

491. (a) S. C. Welch and R. L. Walters, Synth. Comrnun., 3, 15 (1973). (b) S. C. Welch and R . L. Walters, ibid., 3, 419 (1973). (c) S. C. Welch and R. L. Walters, J . Org. Chem., 39, 2665 (1974). 492. See ref. 2, pages 521-525.

493. E. J. Corey and D. S.Watt, J. Am. Chem. Soc., 95, 2302 (1973).

494. (a) M. Miyashita and A. Yoshikoshi, Chem. Commun., 1091 (1971). (b) M. Miyashita and A. Yoshikoshi, J. Chem. Soc., Chem. Commun., 1173 (1972). (c) M. Miyashita and A . Yoshikoshi, J. Am. Chem. Soc., 96, 1917 (1974).

495. See also C. H. Heathcock and R. A. Badger, Chem. Commun., 1510 (1968) and C. H. Heathcock, R. A. Badger, and R. A. Starkey, 1. Org. Chem., 37, 231 (1972).

496 E. J. Corey, M . Behforouz, and M. Ishiguro, J. Am. Chem. Soc., 101, 1608 ( 1979).

497.

H.Yamamcito and H. L. Sham, J. Am.

Chem. Soc., 101, 1609 (1979).

498. E. J. Corey and M. Ishiguro, Tetrahedron Lett, 2745 (1979).

References

539

499. See ref. 2, pages 492-503. 500. E. Piers, W. de Waal and R. W.Britton, J. Am. Chem. Soc., 93, 5113 (1971). 501. (a) K. J. Schmalzl and R. N. Mirrington, Tetrahedron Lett., 3219 (1970). (b) R. N. Mirrington and K. J. Schmalzl, J. Org. Chem., 37, 2871 (1972). (c) R. N. Mirrington and K. J. Schmalzl, ibid., 37, 2877 (1972). 502. See ref. 2, Scheme 2, pages 497-498. 503. See ref. 2, Scheme 3, pages 501-502. 504. (a) N. Fukamiya, M. Kato, and A. Yoshikoshi, Chem. Commun., 1120 (1971). (b) N. Fukamiya, M. Kato. and A. Yoshikoshi, J. Chem. Soc.. Perkin /, 1843 (1973). 505. F. Naf and G . Ohloff, Helv. Chim. Act@ 57, 1868 (1974). 506. G. Friter Helv. Chim. Acta, 57, 172 (1974). 507. M. E. Jung and C. A. McCombs, J. Am. Chem. Soc., 100, 5207 (1978). 508. D. Spitzner, Tetrahedron Lett., 3349 (1978). 509. K. Yamada, Y. Kyotani, S. Manabe, and M. Suzuki, Tetrahedron, 35, 293 (1979). 510. P. Teisseire, P. Pesnelle, B. Corbies, M. Plattier, and P. Manpetit, Recherches, 19, 69 (1974).

511. W. Oppolzer and R. L. Snowden, Tetrahedron Left., 3505 (1978).

512. J. -L. Gras, Terrahedron Lett., 4117 (1977). 513. F. Kido, H. Uda and A. Yoshikoshi, J. Chem. SOC.,Perkin I, 1755 (1972). 514. D. F. MacSweeney and R. Ramage, Tetrahedron, 27, 1481 (1971). 5 15. See ref. 2, Scheme 8, pages 5 15-5 16. 516. See ref. 2, Scheme 10. pages 417-418.

517. A. Deljac, W.D. MacKay, C. S. J. Pan, K. J. Wiesner, and K. Wiesner, Can. J. Chem., 50, 726 (1972). 518. R.

M.Coates and R. L. Sowerby,

J. Am. Chem. SOC., 94, 5386 (1972).

519. E. Piers and J. Banville, J. Chem Soc., Chem. Commun., 1138 (1979).

520. G. Biichi, A. Hauser. and J. Limacher, J. Org. Chem.. 42, 3323 (1977) 521. H. -J. Liu and W. H. Chan, Can. J. Chem., 57, 708 (1979). 522. P. Naegeli and R. Kaiser, Tetrahedron Lett., 2013 (1972). 523. E. 1. Corey and R. D. Balanson, Tetrahedron Lett., 3153 (1973). 524. See ref. 2, Scheme 6, page 510. 525. P. T. Lansbury, V. R. Haddon, and R. C. Stewart, J. Am. Chem. ,Sot., 96, 896

(1974). 526. See also ref. 2, Scheme 7 and accompanying discussion, pages 512-514.

540 527. (a) E. G . Breitholle and A. G. Fallis, Can. J. Chem., 54, 1991 (1976). (b) E. G. Breitholle and A. G. Fallis, J. Org. Chem., 43, 1964 (1978). 528. S. Danishefsky, K. Vaughn, R. Gadwood, and K. Tsuzukim J. Am. Chem. Soc., 102, 4262 (1980). 529. See ref. 2, Scheme 19, pages 542-543. 530. E. Piers and M. Zbozny, Con. J. Chem.. 57, 2249 (1979). 531. (a) R. B. Kelly and J. Zamecnik, Chem. Commun., 1102 (1970). (b) R. B. Kelly, 1. Zamecnik, and B. A. Beckett, ibid. 479 (1971). (c) R. B. Kelly, J. Zamecnik, and B. A. Beckett, Can. J. Chem., 50. 3455 (1972).

532. (a) R. M. Cory and F. R. McLaren, J. Chem. Soc., Chem. Commun., 587 (1977).

533. 534. 535.

536. 537. 538. 539. 540.

(b) R. M. Cory, D. M. T. Chan, F. R. McLaren, M. H. Rasmussen, and R. M. Renneboog, Tetrahedron Lett., 4133 (1979). E. Piers and T. -W. Hall, 1. Chem Soc., Chem. Commun., 880 (1977). R. B. Woodward and T. R. Hoye, J. Am. Chem. Soc., 99, 8007 (1977). (a) T. Kaneko, I. Kawasaki and T. Okamoto, Chem. and Ind. (London), 1191 (1959). (b) M. Kotake, 1. Kawasaki, T. Okamoto, S. Kusumoto and T. Kaneko, Bull. Chem. Soc. Japn.. 32, 892 (1959). (c) M. Kotake, 1. Kawasaki, T. Okamoto, S. Kusurnoto and T. Kaneko, Ann. Chem., 636, 158 (1960). (d) M. Kotake, I. Kawasaki, S. Matsutani, S. Kusumoto and T. Kaneko, Bull. Chem. SOC. Japn, 35, 1494 (1962). (a) Y.. Arata and T. Nakanishi, J. Phurm. Soc. Jupn, 80, 855 (1960). (b) Y. Arata, T. Nakanishi, and Y. Asaoka, Chem. Pharm. Bull. (Tokyo), 10, 675 (1962). 1. Jezo, M. Karvas and K. Tihlarik, Chem. Zuesti 15, 283 (1961). F. Bohlmann, E. Winterfeldt, P. Studt, H. Laurent, G. Boroschewski, and K. -M. Kleine, Chem. Ber., 94, 3151 (1961). (a) J. T. Wrbbel and 2. Dabrowski, Roczniki Chem., 39, 1239 (1965); (b) J. Szychowski, A. Leniewski and J. T. Wr6be1, Chem. ondind. (London), 273 (1978). F. Bohlmann, E. Winterfeldt, H. Laurent, and W. Ude, Tetrahedron, 19, 195 (1963).

541. Y. Arata, T. Ohashi, Z. Okomura, Y. Wada, and J u p 4 82, 782 (1962); 83, 79 (1963).

M. Ishikawa, J.

Pharm. Soc.

542. J. Szychowski, J. T. Wr6bel and A. Leniewski, Can. J. Chem., 55, 3105 (1977). 543. H. Suzuki, I. Keimatsu and M. lto, J. Pharm. SOC.Jupun 52, 162 (1932). 544. (a) T. Hayakawa, H. Nakamura, K. Aoki, M. Suzuki, K. Yamada, and Y. Hirata, Tetrahedron, 27, 5157 (1971). (b) K. Yamada, M.Suzuki, Y. Hayakawa. K. Aoki, H. Nakamura, H.Nagase, and Y. Hirata, J. Am. Chem. Soc., 94, 8278 (1972). (c)

M.Suzuki, Y. Hayakawa, K. Aoki, H. Nagase, H. Nakamura, K. Yamada, and Y. Hirata, Tetrahedron Lett., 331 (1973).

References

541

545. (a) Y. Inubushi, T. Kikuchi, T. Ibuka, K. Tanaka, I. Saji, and K. Tokane, J. Chem. SOL, Chem. Commun., 1251 (1972). (b) Y. Inubushi, T. Kikuchi, T. Ibuka, K.Tanaka, I. Saji, and K. Tokane, Chem. Pharm. Bull., 22, 349 (1974). 546. A. S. Kende, T. J. Bentley, R. A. Mader, and D. Ridge, J. Am. Chem. Soc., 96, 4332 (1974). 547. W. R. Roush, J. Am. Chem. Soc., 100, 3599 (1978). 548. K. Yamamoto, I. Kawasaki, and T. Kaneko, Tetrahedron Lefr., 4859 (1970). 549. D. N. Brattesani and C. H. Heathcock, J. Org. Chem., 40, 2165 (1975). 550. R. F. Borch, A. J. Evans, and J. J. Wade, J. Am. Chem. Soc., 99, 1612 (1977).

Total Synthesis Of Natural Products, Volume 5 Edited by John Apsimon Copyright © 1983, by John Wiley & Sons, Inc.

Index Abscisic acid, 68, 69 Acetoacetic ester, dianion of, 8,69 a-Acetoxyacrylate, Diels Alder Addition of, I16 Acetoxy function reductive removal of, 2 I Acetyl methanesulfonate, 485 Acetylmethylene triphenylphosphorane, reaction with citraconic anhydride,.33 Acoradiene, 291,379 P-Acoradiene, 285,294 y-Acoradiene, 285,292 8-Acoradiene, 285.292 Acoradiene, unnamed, 285 Acoragermacrone, I 14, I 17 Acoranes, 284 a-Acorenol. 285.29 I P-Acorenol, 285.29 I. 294 Acorenone, 285,294,300 Acorenone B, 285,294 Acorone, 285,292,300,302 Acyloin condensation, 284 Adams’ catalyst, 500 African army worm, 170 8-Agarofuran, 137 Agarospirol, 265,275,277.283 Alantolactone, 149, 152 Alantone, 53 a-Alantone, 47 a-Alaskene, 285 P-Alaskene, 285 Aldol condensation, 79, 121, 128, 168, 186, 268 crossed, 62 Alkylation with I-trimethylsilyl-3-bromoI-propyne, by Stork’s metalloenamine method, 300 Alkyl o-nitroselenoxide, 93 a-AUylnickel reagent from reaction of nickel carbonyl with prenyl bromide, 25 1

Aluminum amalgam, 480 Aluminum chloride, 179 Aluminum isopropoxide, 91, 324 Ambrosanolides, 347 Ambrosin, 347,355 American cockroach, 121 Amino nitrile from [2.3]-sigmatropic rearrangement of ammonium ylide, 18 a-Amorphene. 161, 163 Ancistrofuran, 75. 78 Ancistrotermes cavithorax, 78 Anhydro-P-rotundol, 265.277 Aplysin, 235,237 Aplysistatin. 75. 79 Aromatic cadinanes, 157 Aromaticin, 369, 375 Aromatin. 376,396 Attractylon, 152 Axinella cannabina, 306 Axisonitrile, 306 Baeyer-Villiger oxidation, 69, 110, 355,426 Bakkenolide A, 325 Balsam. 62 elam ford-Stevens reaction: deoxygenation by, 114, 135 variant of, 66 Barton reaction, modified, 140 Beckman fragmentation, 99 Beckman rearrangement, nitrone version, 356 Benzenesulfinate ion, elimination of, 68 Benzyl chloroformate, 449 a-trans-Bergamotene, 263 Bicarbocyclic sesquiterpenes, 124 Bicycloelemene, 313 Bicyclofarnesol. 169 Bigelovin, 369,372 Bisabolane family. 35

543

544

Index

a-Bisabolene, 47 8-Bisabolene, 47,49 y-Bisabolenes, 47, 50 a-Bisabolol, 55 a-Bisabololone, 55,59 a-Bisabolols, radio-labeled, 59 Bischloromethyl furan, 381 Bisulfate ester, reduction of, 6 Boron trifluoride, 179 Boron trifluoride etherate, 259 10-Bromo-a-chamigrene, 308.3 I 1 3/3-Bromo-9-epicaparrapioxide, 75, 77 (E)-o-Bromolongifolene, 385 N-Bromo succinamide, 81 cyclization with, 388 Bulgarian peppermint oil, 3 13 a-Bulnesene, 333,335 Bulnesol, 333 /-Butyl chromate, 71, 75 allylic oxidation with, 206,342 3-Butyn-2-01dilithium salt of, 184 Cacalol, 212 Cadalenes. I57 Cadinanes, 157 y,-Cadinene, 16I , 163 E-Cadinene, 161. 162 a-Cadinol, 166 Caespitol, 81 Calamenes, 157 Calcium in ammonia, 246 Camphorenone, 249,436 Caparrapidiol, 10 Caparrapi oxide, 75.76 Caparripitriol, 1 1 ( + )-Carone, 187 Carotol, 377, 379, 381 Carpesiolin, 369.375 Carroll reaction, 18 Carvone, 143,330,449 Castoramine, 500,505 Cecropia. Cl,- and C,, juvenile hormones

(JH). 11

Cedradiene, 484 a-Cedrene, 484,485,486 A’-Cedrene, 484 Cedrol, 484,485,487 a-Chamigrene, 306 Chinese orange oil, 17 Chloranil, 386 /runs-w-Chloroperillene, 30 Chromium trioxide, in acetic acid, 468 Chromous hydroxide, reduction by, 23 Chromyl chloride oxidation, 232 Cinnamolide, 170, 173 Cirrusjunos, 3 13 Claisen acylation. 501

Claisen rearrangement, 15,21.50,213,241, 290,389,456,476 of eater enolate, 389 orthoacetate method, 16 stereorandom, 196 Cobaltous acetate, 81 Complicatic acid, 409 Confertifolin. 170, 172 Confertin, 347,360 Copabomeol, 429,432 Copacamphene, 429,430,432,439 Copacamphor, 250.429,432,436 a-Copaene, 452 8-Copaene, 452 Copaisobomeol, 429,432 Cope elimination, of amine oxide, 104 Cope rearrangement, 18.88 Copper carbonate, 324 Coriolin, 405,412 Costunolide, 114, I18 Cram-Felkin model, for asymmetric induction, 28 Crandall-Rickborn isomerization, 20 Cu(acac), catalyzed rearrangement of diazoketone, 226 Cuauhtemone, 136 Cubebenes, 393 Cubebol, 393 Cuparene, 230 a-Cuparenone, 230.232 8-Cuparenone, 230,234 Curcumene, 44 a-Curcumene. 35,36 Curcumene ether, 35.44 Curcuphenol, 35,42 Cyanide: conjugate addition of, 273 in dimethyl sulfoxide, 5 13 Cyclization, acid catalyzed, 30,68 12 + 2) cycloaddition. of I - methylcyclohexanone and(s)-piperitone, T B Cyclocitral, 389 Cyclocolorenone. 344 Cyclocopacamphene, 429 Cyclocopacamphor, 432 a.,9-Cycloeudesmol. 221,224 /?,o-Cycloeudesmol, 22 I , 224 /3,/3-Cycloeudesmol,22 I Cyclopropyl carbinol solvolysis, 13 Cyclosativene, 429,436,437 Cycloylangocamphene, 429 a-Cyperone, 113, 129 /3-Cyperone, 129, I30 Damsin, 347,350,356 Damsinic acid, 347, 357 Darzens acylation, 320

Index Darrens condensation, 171 Daucene, 377,381 Daucol, 377,379.38 1 Davanafurans, 27 Davana oil, 27 DBP, 375.417 DDQ, 159,346 DDQ oxidation, 150 ( f)-Debromoaplysin, 238 ( f)-Debromolaurinterol acetate, 248 Dehydro-a-curcumene, 35,37 Dehydrofukinone, 188 9,IO-Dehydrofuranoeremophilane,202 Dehydrokessane, 333,339 Demjanov rearrangement, 487 Dendrobine, 510 Dendrobium nobile (Orchidacae), 5 10 Dendrolasin, 29 Deoxydamsin, 347 Deoxygenation. Ireland’s method, 136 Deoxynupharamine, 500,509 Deoxynupharidine. 500,502 Deoxytrisporone. 68 DesoxynorpatehoulenoI, 47 1 Dibromocarbene, addition of, 495 I,l-Dibromo-2-ethoxycyclopropane, carbanion of, 2 I Dibromomethyllithium, 186 Dichlorocarbene, under conditions of phase transfer, 22 I Dichloroketene, (2 + 21 cycloaddition of, 1 10 Dichloromaleic anhydride, 182 a,a-Dichloromethyl methyl ether. 407 Didehalocaespitol, 83 Dieckmann cyclization, 218,277,427,438 Diels-Alder addition, of a-acetoxyacrylate, I16 Diels-Alder reaction, 95, 125, 126, 146, 162, 177, 184, 191, 194, 207,256,258,282, 283,294,320, 384,413,437,456,460, 466,470.481,516 with Danishefsky diene, 21 1 with ethyl acrylate, 218 intramolecular, 201,220,403,463,464, 466,487 inverse electron demand, 452 methyl vinyl ketone and isoprene, 56 a-Diethylphosphonopropionitrile,453 Dihydroarbiglovin, 341 Dihydrocallitrisin, 149, 153 Dihydrocarissone, 136 Dihydrocavone, 138,306,379,432 Dihydrocostunolide, 114. 118 (3S)-2,3-DihydrofarnesoI(20), 9 Dihydrofreelingyne, 28,34 Dihydro-P-ionone, 76, 171. 177 Dihydroisoaristolactone, 114, I19

545

Dihydronootkatone, 180, 187 Dihydrosylvecarvone, 180 Diimide, 487 Diisobutylaluminum hydride, 252 Dilithio acetate, 95,371 Dilithioacetoacetate. 99 Dimethyl diazomalonate, 497 Dimethyllithiomethylphosphonate,356 Dimsylsodium, 385,431,471 Disiamylborane, 296 Disodium tetracarbonylferrate, 400 Douglas fir, 62 Driman-8-01, 169 Driman-8,1 I-diol, 169 Drimanes, 169 Drimanic acid, I7 I (*)-Drimenin. 170, 173 Drim-7-en-ll-ol, 169 Drim-8-en-7-one, 169 y-Elemene, 84 8-Elemenone, 84, 86 Elm wood, 157 Elvirol, 35,42 Emmons reaction, 72 Emmotin-H, 124, 126 Endoperoxide. singlet oxygen oxidation of, 32 Ene reaction, intramolecular, 294 Enolate Claisen rearrangement, 155 Enolate ester Claisen rearrangement, 88 Enol phosphate, 69 ( *)-7,8-epialanolactone, 154 4-Epiaromadendrene, 344,346 4-Epiaubergenone, 134, 136 9-Epibicyclofarnesale, I75 7-Epibulnesol, 334 Epicamphorenone, 249 Epicarrisone, 145 2-Epi-a-cedrene, 487 8-Epidnmanic acid, 171 Epi-y-eudesmol, 139 Epieuryopsonol, 202,207 4-Epiglobulol, 344,346 7-Epiguaiol, 337 10-Epi-hinesol,275 10-Epihysterin, 366 Epiipomeamarone, 3 1 3-Epiisopetaso1, 192 10-Epijuneol, 134, 135 Epijuvabione, 55.62 7-Epinootkatone, 180, 199 3-Epinupharamine, 509 Epi-8-santalene, 249 6-Epishyobunone, 89 Epizizaene, 473,477 Epizizanoic acid, 473

546

Index

Epizonarene, 161 164

12,13-Epoxytrichothec-9-ene, 238,240

Eremophilanes, 180 Eremophilenolide, 202,205 Eremophilone, 195 Eriolanin, 109, I12 Eschenmoser’s Mannich salt. 97 Estafiatin, 34 I , 342 Ethoxyethynylmagnesium bromide, I08 1 -Ethoxy-2-methylbutadiene. phosphonium salt from. 40 Ethyldiazoacetate, 483 Ethyl-3-fury1 acetate, 31 Ethyllithium, 167 Ethyl malonyl chloride, 99 Ethyl thioethoxymethyl sulfoxide, 149 Eudesmanes, 124 Eudesmanolides, 149 8-Eudesmol, 129, 13 I Euryopsonol. 202,207 Evuncifer ether. 137, 139 Farnesal: acid catalyzed cyclization, 56 nitrone derived from, 59 8-Farnesene, 8 (2E-60-Fameso1, 6, I16 (22-6E)-Farnesyl phosphate, cyclization of, 56 Favorskii rearrangement, 243 Fokienol. 22 Fraxinellone, 107 Freelingyne. 28. 33 Friedel-Crafts cyclization, 171 Friedel-Crafts reaction: with 2-methoxy-4-methyl phenol. 213 with y-valerolactone. 212 Frullanolide, 149, 155 Fukinone. 188 Furanoeremophilane, 202.205 3~-Furanoligularanol.202,209 Furanoligularone, 202 3-Furyllithium. 78. 108, 109 3-Furylmethylmagnesium chloride. 178 Futronolide. 170. I74 Geraniol, 24 Geranyl acetate, 123 Geranyl acetone, 3 1 I Geranyl senecionate, 89 Globulol, 344 Glutinosone, 140 8-Gorgonene, 2 IS, 220 Greek tobacco. 169 Grignard reagent: from 2-bromo-1 -butene. I5 from (E)-l-brorno-3-methyl-I -butene, 164

Grob fragmentation, 66, 85,275,448,455 Guaianolides, 341 Guaiazulenes, 333 Guaiol, 333,335 Gymnomitrol. 424 Gyrindal, 24 Haller-Bauer fragmentation, 264 62-Hedycaryols, 114, I 17 Helenalin, 369,371,373 Helenolides, 347, 369 Helminthosporal. 429,442 a-Himachalene, 384 8-Himachalene, 384,387 Hinesol, 265,269, 275. 283 Hirsutene, 405,409.4 I7 Hirsutic acid, 405 Hirsutic acid C, 407 Homocitral, 78 Homogeranic acid, 9 Homogeranyl tosylate, 79 Homer-Wadsworth-Ernmons olefination, 13 Humulene, 122 Hydronaphthalenes, 124 3~-Hydroxyfuranoeremophilane,202.209 Hydroxy sulphone. alkylation of. dianion of. 68 Hymenin, 341,352,369 Hypacrone, 228 Hypolepin A, 323 Hypolepin B, 323 Hypolepin C, 323 Hypoleprs punctata. 323 Hypolepis punetato Melt., 228 Hysterin, 347, 366 Illudinine, 498 Illudol, 394 Indian vetiver oil, 218 N-lodosuccinimide. 32 a-lonone, 74 p-lonone, 71 8-lonone, acid obtained from, 70 Ipomeamarone. 28.31.32 Ishwarane, 493 Ishwarone, 493,496 Isoacorone. 285,293,300,302 Isobisabolene, 47,49 Isobutenyllithium, 53 Isocaespitol. 81 Isocaryophyllene, 39 I Isocomene, 419 2-lsocyanopupukeanane, 455 9-lsocyanopupukeanane. 455,456 lsodihydroagarofuran, I39 Isodrimenin, 170, 172, 176 7-lsohimachalol, 384

Index lsohirsutic acid, 405 Isokhusimone, 48 I, 487 Isoligularone, 202,2 I I Isolongifolene, 492 lsomarasmic acid, 398 Isonootkatone, 180, 184, 187 lsopetasol, 192. 194 Isopiperitone, I17 Isoprene, trimerization of, 8 Isopropenyllithium, 326 lsopropenylmagnesium bromide, conjugate addition of, 130, 220,269 lsopropylidene group, novel method for introducing, 87 Isopropylidenetriphenylphosphorane,26 I lsopropyl magnesium chloride, PhSCu catalyzed addition of, 456 Isoshyobunone, 84,90 Isotadeonal, 170. 175 Isotelekin, 149, 153 Ivangulin, 109 C,,-JH, 12 C,,-JH. from rrans-geranylacetone. 14 C,,- and C,"-Jtl, synthesis of /I-oxidophosphonium ylide method, I2 C,,-JH, 12, 15, 16 Juneol, 134 Juvabione, 55.62 Kessanol. 333,340 Ketene dithioacetal, isoprene equivalent, 53 Keto phosphonate. dianion of, 38 K husimone, 473.480 Khusitene, 218 Knoevenagel condensation, 234 with di-t-butyl malonate, 101 Kolbe electrolysis, homogeranic acid, 9 Lacinilene C methyl ether, 157, 161 Lactaral. 84 Lanceol, 47, 53 Latia lucifenn, 68,69 Laurencia caespitosa, 8 1 Lauremia pac$ca. 31 I Laurene, 235 ( =k )-Laurinterol methyl ether. 248 Lead tetraacetate: decarboxylation, I42 oxidation with, 108, 128 Lemieux-Johnson cleavage, 475 Lemieux-Johnson oxidation, 326 Ligularone, 202,207,2 I I Limonene. 132 free radical addition of carbon tetrachloride, 53

547

metallation of, with n-butyllithium and tetramethyl-ethyl-enediamine, 52 Limonene epoxides, 61 Limonin, 107 Linalool oxides, 27 Linifolin A, 369,372 Lithiated a-alkoxy nitrile, Stork-Maldonado method, 63 I -Lithio-cyclopropyl phenyl sulfide, 273 2-Lithio- I ,3-dithiane, alkylation by neopentyl iodide, 33 Lithio-methyl trimethylsilyl acetate. 359 Lithium acetylide, addition of, 332 Lithium aluminium hydride-aluminium chloride, hydrogenolysis with, 27 Lithium aluminium hydride-sodium methowide, 24 Lithium diisobutenylcuprate. 3 15 Lithium dimethylcuprate, 21 I. 275 displacement of bromide with, 23 I Lithium divinylcuprate, 101. 195. 480, 516 Lithium ethylamine, 116.306 Lithium fluoborate, 273 Lithium hexamethyldisilazide, 374, 427, 489 Lithium metal, 84 Lithium naphthalenide, reduction by, 135 Lithium perchlorate. 199 Lithiurn 2,2,6,6-letramethylpiperidide (LTMP). 147 Lithium tri-I-butoxyaluminohydride. 9 I Lithium tri-sec-butylborohydride, I 14 Lithium triethylborohydride, 497 Longiborneol, 446,449 Longicarnphor, 446.449 Longicyclene, 446. 449 Longifolene, 446 Longifolin, 25 a-Longipinene, 452 P-Longipinene, 452 Lucas reagent, 134 McMurray reaction. 440 Magnesium methyl carbonate, 322 Mannich alkylation. 166,242,243 Mansonone D, 157, 160 Mansonone G, 157, 160 Marasmic acid, 398 Mayurone. 22 1,225 [MeCuBrJ-[Li(i-Bu,NH)?]. , 27 I Mercuric oxide, as aldol catalyst, 32 Mercuric trifluoroacetate, 77 Methallylmagnesium bromide, 40 Methallytriphenylphosphoniumbromide, 218 Methallytriphenylphosphonium chloride, Wittig reagent from, 168

548

lodex

Methoxymethylenetriphenylphosphorane, 440 Methylation, Greico’s procedure, 150 Methyl bicyclofarnesate, 174 Methyl-2-butynoate. cuprate addition to. 16 3-Methylcrotonic acid, alkylation of dianion of, 7 Methyl cyclogeranate, 69 Methylenation, Mannich method, 153 Methylenetriphenylphosphorane. 141, 142 Methyl famesate. 8, 81, 134 Methyl Ruorosulfonate, 363 3-Methyl-2-furyllithium, 26 5-Methyl-2-furylIithium. 27, 232 3-Methyl-2-furylmetallic reagent: bisfurylmercury. 26 f‘urylmagnesium bromide, 26 Methyl geranate, 15 Methyl isopropenyl ether, 484 Methyl isozizanoate, 475 Methylmagnesium bromide, copper-catalyzed conjugate addition of, 339.373 Methyl o-nitrophenyl disulfide, 412 Methyl perrilate. 65 2-Methyl- I -propenyllithium, 23 Methylsulfinyl carbanion, 494 Methyl trimethylsilyacetate, anion of, 375 Methyl zizanoate, 474 Mexicanin I, 369, 372 Mexican medicinal shrub, I36 Michael addition, 1’29, 193,240, 245,337, 510 acid catalyzed, 155 intramolecular. 99 with unsaturated nitro compound, 212 Mislow/Evans rearrangement, of allylic sulfoxide, 101 MOO,, molybdenium pentoxide, 176 Monocarboxyclic sesquiterpenes, 35 Monocyclofamesic acid, 173 Mukiayama addition, of I-f-butoxy-I-I-butyldimethylsilyloxyethylene, 489 Myrcene, 20 singlet oxygen oxidation, 30 Nagata cyanation. 513 Neoambrosin, 347. 352 Neotorreyol, 28 Nerolidol, 22.80 acid catalyzed cyclization of, 56 (6E)-Nerolidyl phosphate, cyclization of, 56 Neryl senecionate. 88 Ni (COD),, 363 Nickel boride catalyst, 256,261 Nickel carbonyl coupling, of bis-allylic bromide, 122 Nishimura’s catalyst, hydrogenation with, 456

o-Nitroselenoxide, elimination of, 121 Nootkatone, 180, 182, 186,269 Norketoagarofuran, 137, 138 Norketotrichodiene, 238,246 Norpatchoulenol, 469 Nuciferal, 35, 40 Nupharamine, 500,508,509 Nuphar spp., 500 Occidol, 124, 133 ( + )-Occindentalol, 143 Oplopanone, 327 Organozinc reagent, from I-methoxy2-butyne, 66 Oxidation: f-butyl chromate, 7 I with lead tetraacetate, 128 Lemieux-Johnson method, 462 modified Polonovski method, 263 with selenium dioxide. 126 singlet oxygen, 16, I50 van Tamelen’s method. I16 Oxidative elimination. of o-nitrophenylselenide. 107 Oxocrinol, 25 9-0~0-5,8-dehydronerolidol,22, 23 9-Oxonerolidol, 22 Oxy-Cope rearrangement, 117, 121, 164 Palladium, on barium sulfate, 503 Pallescensin A, 178 Parthenin, 347,352,369 Patchouli alcohol, 335,460,464,469 Pentalenolactone, 3 18 4-Pentenylmagnesium bromide, 247 Perforenone, 388 Perillartine, 65 (-)-Perillic alcohol, 64 Penplanone-B, 114, 121 Peroxyformic acid, 138 Pefasilesjaponicus Max., 325 Phenylthiomethyllithium, 237 Phenyl thiophenylsulfonate, 415 Photocitral-A, 340 Photooxygenation, 70 Phytuberin, 109, I13 Picrotoxinin, 330 8-Pinene, 186, 199 (-)-Pipendine, I19 Pipendine enamine: alkylation of, 107 of isovaleraldehyde, 168 Platphyllide, 124, 128 Polygodial, 170, 173 Polyphosphoric acid, 338 Potassium fluoride, in methanol, 470 Potassium triethylmethoxide, 462

Index Potassium triphenylmethide, 462 Potato tubers, 269 Preisocalamendiol, 17.88, I14 Prevernolepin, 95 Prevemomenin, 95 Prevost hydroxylation, 5 13 Prins reaction, 3 15, 422 intramolecular, 266,340 Protected cyanohydrin, conjugate addition of, 234 Protoilludanols, 394 Protoilludenes, 394 Pseudoguaianolides, 369 Psilostachyin, 355 Psilostachyin C. 347 Pteridium aquilinum var. Latiusculum, 324 Rerosin E, 324 R-( +)-Pulegone, 293 Pummerer rearrangement, 375 Pyridinium bromide, 171 Pyroangolensolide, 107 Qrocumerenone, 157, 160 Pyrovellerolactone, 38 I

549

Reformatsky reaction, 34,206,299,430,470 Reverse Claisen reaction, 297 Rhodium chloride, 282 Rishitin, 140 Rishitinol, 124 Robinson annelation, 141, 180,301,328.389, 513 acid-catalyzed, 151 (+)-fl-RotunoI, I14 Ruthenium tetroxide, 516

Sharpless reagent, 320 Shimalite, solid phase, acidic, 440 Shotten-Baumann acylation, 503 Shyobunone, 84.87 Silver carbonate, on celite, 149 Silver tluoborate, 31 1 Simmons-Smith cyclopropanation, 110,223, 224,248,360,393.449,485 a and Sinensal, 17, 18 Sinularene, 429,443 Sirenin, 3 14 a-Snyderol, 75 P-Snyderol, 75 Snyderols, 80 Sodium acetylide, reaction with chloro epoxide, 34 Sodium benzenesulfinate, 253 Sodium bis (methoxyethoxy), aluminium hydride, 208 Sodium borohydride-cerium trichloride, 186 Sodium cyanoborohydride, 101, 153,505 in methylamine, 5 16 in presence of zinc bromide, 449 Solavetivone, 265,269 Spirovetivanes, 264 Stannic chloride, 88, 135,378,426 stereospecific cyclization with, 291 Stephens-Castro reaction, with 3-furylcopper, 35 Stereospecificsynthesis of olefins, from B-oxido phmphonium ylides and aldehydes, 6 Stobbe condensation, 132 Stramonin B, 347,355 Sulphore, reductive cleavage of, 25 Sydowic acid, 35,44

a-Santalene, 249,253, 254 8-Santalene, 249 a-Santalol, 249, 252 8-Santalol, 249,258 a-Santonin. 87,91, 118, 140, 144, 149. 165, 34 I Sativene, 429,430,436,439 cis-Sativenediol, 429,440 Saussurea lactone, 90 Seco-cadinane, 46 Selenium dioxide, oxidation with retention of optical activity, 75 Selenium dioxide oxidation, 68, 126 Selenoxide elimination, 86 /3-Selinene, 129, 132 Sesquicarene, 3 14,317 Sesquichamaenol, 46 Sesquifenchene, 249,263 Sesquirosefuran, 25 Seychellene, 460,462,465

Taylorine, 228 Temsin, 90.92 a-Terpineol, 271 rrans-/3-Terpineol, 82 Terpinolene, 50 Terrestrol, 9 Tetra-n-butylarnmonium oxalate (TBAO), 119 Tetrahydroeucaryone, 449 Tetrahydroligularenolide, 202,204 Tetramethylammonium acetate, 14I Thallium perchlorate, 396 Thallium trinitrate, 412 Thexylisobutylborane, 65 Thujopsadiene, 221,226 Thujopsene, 22 I, 226 Titanium tetrachloride, 128 p-Toluenesulfonylisoyanide, and potassium r-butoxide, 518 p-Toluenesulfonyl-S-methylcarbazate, sodium salt of, 326

Quadrone, 488

550

Index

p-Toluenesulfonylmethyl isocyanide, 28I,

456

p-Tolyl Grignard reagent, 230 p-TolyUithium. 23 I p-Tolylmagnesium bromide, addition to methylheptanone. 36 Torreyal. 28 Torreyol. 166,168 Transfer hydrogenation. 167 Tri-n-butylstannane. 440 Trichodermin, 238 Trichodermol, 238 Trichodiene. 238,246 Tricyclovetivene. 473 Triethylphosphonoacetate anion, 47 3-Triethylsilyloxypentadienyllithium. 470 Trimethylaluminum. 364 Trimethyl phosphonoacetate, 82 3-Trimeih~lsilyiallymagnesiumbromide, 428 Trimethvlstannvllithium. 121 conjugate a d h i o n of, 117 Triphenylphosphjne dibromide. 269 Triphenyltin hydride. 89.378 Tritylfluoroborate. 79 Trityllithium, 335 Tuberiferine, 149.I5 I ur-Tumerone. 35.38 Valencene, 180 Valerane, 2I5,2I7 Valeranone. 2I5 Valeriana Waalichi, 259 VeUeral. 38 I 384 Vellerolactone. 381 384 Vernolepin, 90,93,95 Vernomenin. 93 a-Vetispirene. 265.266.269.277.283 8-Vetispirene, 265,269.277 Vetiver oil. 290.473 p-Vetivone. 264.265,281 Vinylacetyl chloride. acylalion with. 132 Vinyl cyclopropane. rearrangement. 442 I

Vinyllithium. addition of, 54 Vinylmagnesium bromide, copper catalyzed addition of, 90. 101,329 retro-ene reaction, 89 Wacker oxidation, 415 Wadsworth-Emmons reaction, 181,363,403 Wagner-Meenvein rearrangement. 25 I, 259.

275.385,420

Warburganal, 170,175 Warburgiadione. 192 West African soldier termite, 139 Whirligig water beetle, 24 White potatoes, fungus infection, 126 Wichterle annelation, 108 Wichterle reagent. bromine homolog of, 130 Widdrol, 389 Wieland-Miescher ketone, 85 Wilkinson’s catalyst, 320,322,475 I. *)-Winterin. 172 , Wittig olefinalion, via @lactone, 36 Wittig reaction, 186,2I7 selective, 277 stereoselective, 23.34 Wolff-Kishner reduction, 386 WoltT rearrangement, 419 photochemical, 227.25s.29 I Xanthorrhizol, 35,42 a-Ylangene. 452 8-Ylangene. 452 Ylangocamphene. 429 Mangocamphor, 250,429,436 Yomogin, 149,150 Zinc chloride. 134 Zinc-copper couple. in aqueous DMSO, 99 Zizane, 473 Zizaene. 478 Zizanoic acid, 473 Zonarene. 161,164