Universi^ Micrxsilms International 300 N

INFORMATION TO USERS

This reproduction was made from a copy of a document sent to us for microfilming. While the most advanced technology has been used to photograph and reproduce this document, the quality of the reproduction is heavily dependent upon the quality of the material submitted.

The following explanation of techniques is provided to help clarify markings or notations which may appear on this reproduction.

1.Tlie sign or “target” for pages apparently lacking from the document photographed is “ Missing Page(s)” . I f it was possible to obtain the missing page(s) or section, they are spliced into the film along with adjacent pages. This may have necessitated cutting through an image and duplicating adjacent pages to assure complete continuity.

2. When an image on the film is obliterated with a round black mark, it is an indication of either blurred copy because of movement during exposure, duplicate copy, or copyrighted materials that should not have been filmed. For blurred pages, a good image o f the page can be found in the adjacent frame. If copyrighted materials were deleted, a target note will appear listing the pages in the adjacent frame.

3. When a map, drawing or chart, etc., is part of the material being photographed, a definite method of “sectioning” the material has been followed. It is customary to begin filming at the upper left hand comer o f a large sheet and to continue from left to right in equal sections with small overlaps. I f necessary, sectioning is continued again-beginning below the first row and continuing on until complete.

4. For illustrations that cannot be satisfactorily reproduced by xerographic means, photographic prints can be purchased at additional cost and inserted into your xerographic copy. These prints are available upon request from the Dissertations Customer Services Department.

5. Some pages in any document may have indistinct print. In all cases the best available copy has been filmed.

Universi^ MicrxSilms International 300 N. Zeeb Road Ann Arbor, Ml 48106

8426387

Galemmo, Robert Anthony, Jr.

SYNTHETIC STUDIES ON THE STRUCTURE OF SENOXYDENE: A SEQUENTIAL ANNULATION APPROACH TO ANGULAR TRIOUINANE SYNTHESIS

The Ohio Stale University Ph.D. 1984

University Microfilms i ntern eti 0n el300 N. zeeb Road, Ann Arbor, Ml 48106

PLEASE NOTE:

In all cases this material has been filmed in the best possible way from the available copy. Problems encountered with this document have been identified here with a check mark V

1. Glossy photographs or pages.

2. Colored illustrations, paper or print _____

3. Photographs with dark background _____

4. Illustrations are poor copy ______

5. Pages with black marks, not original copy.

6. Print shows through as there is text on both sides of page

7. Indistinct, broken or small print on several pages ______

8. Print exceeds margin requirements______

9. Tightly bound copy with print lost in spine______

10. Computer printout pages with indistinct print.

11. Page(s) 13(Fig. 1) lacking when material received, and not available from school or author.

12. Page(s) ______seem to be missing in numbering only as text follows.

13. Two pages numbered ______. Text follows.

14. Curling and wrinkled pages______

15. O t h e r ______

University Microfilms International

SYNTHETIC STUDIES ON THE STRUCTURE OF SENOXYDENE:

A SEQUENTIAL ANNULATION APPROACH TO

ANGULAR TRiqUINANE SYNTHESIS

DISSERTATION

Presented, in Partial Fulfillm ent of the Requirements for

the Degree Doctor of Philosophy in the

Graduate School of The Ohio State University

Robert Anthony Galemmo, J r ., B.A., M.S.

*****

The Ohio State University

1984

Reading Committee Approved by

Dr. Harold Shechter

Dr. John S. Swenton

Dr. Leo A. Paquette Dr. Leo A. Paquette, Advisor

Department of Chemistry To my Mother and Father, for giving me what i t takes

to see something through. ACKNOWLEDGEMENTS

I wish to gratefully acknowledge the advice and support of

Professor Leo Paquette while working under his direction. I am indebted to my many friends and associates in the Paquette group for their suggestions and technical assistance. Mary, B illy , Ken, Dan and Keith have been responsible for making my stay here a pleasant one.

Several individuals have made contributions to the preparation of this manuscript. Ken, Keith, Geoff, Dan, Craig and John have helped to correct my atrocious grammar and spelling. Kay's stamina and endurance in typing this work has served us both w ell. Finally I wish to thank the staff of the Venetian for giving me a quiet place to write on sunny days.

r r r VITA

May 14, 1954 ...... Born - Philadelphia, Pennsylvania

1976 ...... B.A., State University of New York,

College at Potsdam, Potsdam, New York.

1977-1981 ...... Teaching Associate, Department of

Chemistry, The Ohio State University,

Columbus, Ohio.

1981 ...... M.S., The Ohio State University,

Columbus, Ohio.

1981-1984 ...... Research Associate, Department of

Chemistry, The Ohio State University,

Columbus, Ohio.

MASTER THESIS

The Preparation Of Components Of The 0-Anti genic

Si de-Chain Of Pseudomonas Aeruginosa Type V

Advisor: Professor Derek Horton

PUBLICATIONS

”4.7.7-Trimethyl-cis-bicyclo[3.3.0]oct-3-en-2-one: A Potentially Useful Synthon for Triquinane Natural Products", Leo A. Paquette, Eugene Farkas and Robert A. Galemmo, J r ., J. Org. Chem., 46, 5434 (1981).

"Preparative Routes to Methyl 2-acetamido-2,6-dideoxy-a-D-glucopyrano- side", Robert A. Galenmo, Jr. and Derek Horton, CarbohydT Res. , 119, 231 (1983).

iv "Synthesis of the Alleged Structure of Senoxydene, the Triquinane Sesquiterpene Derived from Senecio Oxyodontus", Leo A. Paquette, Robert A. Galemmo, J r ., and James P. Springer, J. Am. Chem. Soc., 105, 6976 (1983).

FIELD OF STUDY Organic Chemistry TABLE OF CONTENTS

Page

DEDICATION...... 11

ACKNOWLEDGEMENTS ...... I l l

VITA ...... Iv

LIST OF FIGURES...... 1x

LIST OF TABLES...... x lll

A CAVEAT FOR THE CHEMIST...... xlv

CHAPTER

I. The Angular Triquinane Sesquiterpenes

1. Introduction ...... 1 2. Angular Triqulnanes ...... 2 3. Isocomene ...... 3 3.1 Isolation and Natural Occurrence ...... 3 3.2 Biogenesis...... 5 3.3 Syntheses ...... 6 4. Sllphlnene...... 10 4.1 Isolation and Natural Occurrence ...... 10 4.2 Biogenesis...... 11 4.3 Syntheses ...... 11 5. SI 1 phiperfolenes...... 15 5.1 Isolation and Natural Occurrence ...... 15 5.2 Biogenesis...... 16 5.3 Synthesis ...... 19 6. Rental e n e n e...... 20 6.1 Isolation and Natural Occurrence ...... 20 6.2 Biogenesis...... 21 6.3 Syntheses ...... 23

v1 TABLE OF CONTENTS (Cont.)

Page 7. Senoxydene ...... 27 7.1 Isolation and Natural Occurrence ...... 27 7.2 Biogenesis ...... 27

I I . Two Approaches to Senoxydene ( 56) Based Upon the Favorskii Reaction. 1. Statement of the Problem...... 29 2. Elaboration of a Favorskii Reaction Product ...... 30 2.1 Retrosynthesis and Methodology ...... 30 2.2 Synthetic W ork ...... 35 2.3 Previous Preparations of 6 6...... 39 3. Attempted Favorskii Ring Contraction of a Bicyclic Precursor ...... 41 3.1 Retrosynthesis and Methodology ...... 41 3.2 Synthetic W ork ...... 43

I I I . The Total Synthesis of Angular Triquinane 5^: Alleged Senoxydene. 1. Diketone Coupling Approach to ^ ...... 48 1.1 Retrosynthesis and Methodology ...... 48 1.2 Synthetic W ork ...... 51 1.2.1 Preparation of 7^ ...... 51 1.2.2 Some Transformations of 7 5 ...... 54 1.2.3 Alkylation of 8;4 and Di ketone Coupling. . 57 2. Synthesis of Triquinane ^ ...... 59 2.1 Retrosynthesis and Methodology ...... 59 2.2 Synthetic Work ...... 60 2.2.1 Preparation of ^ ...... 60 2.2.2 Preparation of ^ ...... 61 3. Alternative Synthesis of 5 6 ...... 69

v ii TABLE OF CONTENTS (Cont.)

Page

IV Preparation of Two Angular Triquinane Isomers of Alleged Senoxydene (56). 1. General Considerations ...... 71 2. Total Synthesis of Triquinane 9^ ...... 72 2.1 Retrosynthesis and Methodology ...... 72 2.2 Synthetic W ork ...... 73 2.2.1 Preparation of 1 0 £...... 73 2.2.2 Preparation of 9^...... 76 3. Approach to the Total Synthesis of Triquinane 96^ . . . 83 3.1 Retrosynthesis and Methodology ...... 83 3.2 Synthetic W ork ...... 84 3.2.1 Preparation of 109 ...... 84 3.2.2 Progress towards 96^...... 88

V Conclusion: A Review of the Evidence ...... 92

EXPERIMENTAL ...... 100

REFERENCES...... 172

v i i i LIST OF FIGURES

Figure Page T T Three classes of tr iq u ln a n e s...... 1

2. Five classes of angular triquinanes ...... 2

3. Proposed biosynthesis of isocomene and modhephene...... 5

4. Paquette and Han synthesis of isocomene ...... 6

5. Oppolzer's synthesis of Isocomene ...... 7

6. Pirrung's synthesis of isocomene ...... 8

7. Wender's meta-photocycloaddition and thermal rearrangement. 8

8. Dauben's synthesis of isocomene ...... 9

9. Silphinene derivatives ...... 10

10. Proposed biosynthesis of silphinene ...... 11

11. Paquete and Leone-Bay synthesis of silphinene ...... 13

12. Itô's synthesis of silphinene ...... 14

13. Biosynthesis of silphiperfolenenes ...... 16

14. Rearrangement of 8-a-hydroxypresi1 phi perfolene to ^ . . . 17

15. Paquette synthesis of optically active si 1 phiperfol-6-ene . 18

16. Absolute stereochemistry of si 1 phiperfol-6-ene and s ilp h in e n e...... 20

17. Pentalenene derivatives ...... 20

18. Ohfune's study of the formolysis of a protoilludyl cation e q u iv a le n t...... 21

19. Proposed biosynthesis of pentalenene and its morehighly oxidized derivatives...... 22

20. Paquette and Annis synthesis of pentalenene ...... 24

ix LIST OF FIGURES (Cont.)

Figure Page 21. Matsumoto's biomimetic preparation of pentalenic acid . . . 2F

22. Proposed biosynthesis of senoxydene from humulene ...... 27

23. Proposed biosynthesis of senoxydene from i^ocaryophylelene. 28

24. Retrosynthetic analysis of 5 6 ...... 30

25. Mechanism for the solvolytic cyclization of lactones to enones ...... 31

26. Tricyclic lactone 32

27, Favorskii rearrangement of pulegone dibromide and the cationic cyclization of pulegenic acid derivatives 32

28. Achmad and C avill's rearrangement of pulegone epoxide . 33

29. Reusch and Mattison's rearrangement of pulegone epoxide 34

30. Mechanism of the Favorskii rearrangement of pulegone epoxide 34

31. Preparation of 2-i sopropyli dene-4,4-dimethylcyclohexanone . 35

32. Preparation of bi cyclic lactone 6^...... 36

33. Derivatization of lactone M ...... 37

34. Matsumoto's synthesis of 6 ^ ...... 39

35. Thorén's synthesis of 6 £ ...... 40

36. Magnus synthesis of 6 ^ ...... 40

37. Retrosynthetic analysis of ^ ...... 41

38. Mukaiyama's method for the isopropylidenation of ketones . 42

39. Preparation of bi cyclo[4.3.OJnonanone 6 8 ...... 43

40. Unsuccessful 1sopropylidenation of 68 ...... 44

41. Successful isopropylidenation of 70 ...... 45

42. Epoxidation of 67^ ...... 46

X LIST OF FIGURES (Cont.)

Figure Page 43. Ring cleavage of epoxy ketone 7 ? ...... 47

44. Sequential annulation approach to SB ...... 48

45. Mechani'sm for the proposed coupling of di ketone 73 . . . . 49

46. Ene cyclization of e.ç-ethylenic ketones ...... 50

47. Preparation of 73 using the Ponaras Grignard reagent . . . 51

48. Thermal ene route to 7 5 ...... 53

49. Shapiro olefination of 7 3 ...... 54

50. Thermodynamic enolization of 7 3 ...... 55

51. Kinetic deprotonation of 7 3 ...... 55

52. Preparation of thiobutylmethylene ketone^ ...... 56

53. Methylcyclopentene annulation by diketone coupling reaction 57

54. Retrosynthetic analysis of ^ ...... 59

55. Preparation of allylic iodide 88 60

56. Preparation of 56^...... 61

57. Computer generated drawing of 9 4 £...... 63

58. 200 MHz NMR spectrum of triquinane 56^...... 65

59. 270 MHz ^H NMR spectrum of naturally occurring senoxydene. . 66

60. Two possible structures for senoxydene ...... 69

61. Itô's total synthesis of 5 6 ...... 70

62. Preparation of 97^ ...... 70

63. Retrosynthetic dissection of 9 ^ ...... 72

64. Ponaras Grignard route to 100 ...... 73

65. Thermal ene cyclization route to IJOO ...... 74

xi LIST OF FIGURES (Cont.)

Figure Page 66. Ground state and transition state arguments for the fa ilu re of the thermal ene cyclization of 104 to 100 . . 75

67. Methycyclopentenone annulation of 100 and transformation to 9 5 ...... TT ...... 76

68. The epimeric tricyclopentanoid ketones ...... 79

69. 300 MHz NMR spectra of 106 and 1 0 8 ...... 80

70. 300 MHz ^H NMR spectrum of triquinane 9^ ...... 82

71. Retrosynthetic dissection of triquinane 9 ]5 ...... 83

72. Hydrogenation ^ to 1 0 9 ...... 84

73. Ponaras Grignard route to 1,4 diketone 111_ ...... 85

74. Phenylsulfone route to 109 ...... 85

75. Preparation of thiobutylmethylene ketone 114 86

76. Progress towards the preparation of ^ ...... 87

77. Susceptibility of thiobutylmethylene ketones towards nucleophilic attack ...... 91

78. Range of chemical shifts for te rtia ry and secondary methyl groups in angular triquinanes ...... 93

x ii LIST OF TABLES

Table Page 1. Natural occurrence of isocomene and its derivatives .... 4

2. Natural occurrence of si 1 phiperfolenes ...... 15

3. Dehydrohalogenati on of 6 ^ ...... 38

4. Comparison of the NMR spectra of 56_and senoxydene . . . 67

5. Attempted ene reaction of 104 75

6. Aldol cyclization of 98 and 1 0 5 ...... 78

7. Partial comparison of the NMR spectra of 95 and senoxydene 81

8. Alkylation of l l ^ ...... 89

9. Comparison of chemical shifts of methyl groups on known triquinanes and senoxydene...... 95

x i i i A Caveat for the Chemist

"A. In investigating the principles of a body, we must not judge of them from a slig h t agreement with other known bodies, but they must be separated d irectly by analysis, and that analysis shall be confirmed by synthesis.

B. Analysis should chiefly be conducted in the humid way.

C. Such experiments should be instituted as are adapted to the discovery of the truth.

D. Experiments should be made with the utmost possible accuracy.

E. The experiments of others, particularly the more remarkable ones, should be candidly reviewed."

from Opuscula,

Torbern Olaf Bergman (1735-1784)

XI v CHAPTER I

The Angular Triquinane Sesquiterpenes

1. Introduction

Ongoing interest in the study of the chemotaxonomy of plant and microbial li f e has served not only to broaden our understanding of the biogenetic pathways operating in lower life forms, but also to entice synthetic chemists to reproduce Nature's molecular a rtis try in the laboratory. A major area of research in recent years has been in the isolation and synthesis of sesquiterpene triquinanes^.

Based upon their carbocyclic framework, the sesquiterpene tr iq u i nanes may be relegated to three broad classifications. The firs t, the q 7 "linear" triquinanes having a tric y c lo [6 .3 .0 .0 ’ ] undecane skeleton,

(Figure 1), are isolated from terrestrial microbes and represented by

Linear Triquinane Propel 1ane Angular Triquinane

Figure j . Three classes of triquinanes. 2 3 4 hirsutene , coriolin and their marine counterparts, the capnellanes . The

[3.3.3]propellanes comprise the second category with only modhephene^ and two oxidized derivatives^ possessing this unique structure. Mod hephene has been proposed to be the result of a shunt pathway in the biosynthesis of a member of the third class which features the tricy- c lo [6 .3 .0.0^ ’^]undecane ring system, i . e . , the "angular" triquinanes.

2. Angular Triquinanes

To date, five carbocyclic skeletons have been isolated and described as angular triquinanes^®. Illustrated below (Figure 2) are

isocomene silphinene

s11 phiperfol-6

pentalenene senoxydene

Figure 2. Five classes of angular triquinanes. the least oxidized olefins from which more highly oxygenated metabolites are believed to be derived.

The structural similarities of the five classes of angular triqui nanes were suggestive of closely related biosynthetic pathways. The synthetic challenge posed by these structures was more diverse. Each triquinane required a unique strategy for the regiocontrolled deployment of the four appended methyl substituents. This also prompted the development of ingenous approaches to the construction of their quaternary carbon atoms. Therefore, the preparation of these triquinanes has provided the impetus for the germination of new ideas in organic synthesis and for clever adaptation of the old.

3. Isocomene

3.1. Isolation and Natural Occurrence.

Isocomene (_1, Table I) was the firs t sesquiterpene containing the tric y c lo [6 .3 .0 .0 ^ ’^]undecane ring system uncovered in Nature. In 1977,

Zalkow reported its isolation from the volatile oil of Isocoma wrightii^.

Later that same year, Bohlmann found the identical triquinane in the roots of Berkheya radula^. In itia lly , the structure of isocomene was established by and NMR spectroscopy. Subsequent confirmation was 8 achieved by single crystal x-ray analysis of its diol derivative .

Since this original work, isocomene, its regioisomer B-isocomene, and six oxidized derivatives have been detected in 18 species of 6 geni of plants (Table 1). Table 1. Natural occurrence o f isocomene and it s d e riva tive s.

H, R£ “ Hg 5 R = CHgOH 2 OH, Rp = Hp 6 R 08200008208 (0 8 3 )2 «1 = 3 R, = H, Rg = 0 7 R 080 8 R OO2O83 plant derivative reference

Arni ci a amplexicaulis 3 9 Ar. chami ssanTT F 9 KL 1 ongi fol feT" F 9 Ar. mollis F 9 Sr7 Purryf F 9 Berkheya bergiana î 10 FT bi pinnat1 fida lT4 10 B. echinata '”1'" 10 B7 mori tlma T 10 B. multi jay a T 10 B7 radula T 7 B. rhapontica 0 10 FT setifera T.T 10 B. Fmberfâtâ 10 ü ^ lile p is sal ic i fo lia 1 ,5 ,1 , 7 ,8 11 Isocoma w r i^ t ii T 5 L i^ i urn eggersii 174 6 Puli caria dyserterica - g - 12 b *

i i

Figure 3. Proposed biosynthesis of isocomene (^) and modhepene (9^).

3.2 Biogenesis

Zalkow was the firs t to suggest that caryophyllene, present in significant amounts in Isocoma wrightii, was the biosynthetic precursor to isocomene (1^) and the novel[3.3.3]propel 1 ane co-isolated with viz. modhephene (9^)^ (Figure 3 ). I t was proposed that 1, and 9^ were derived from the same tricyclic carbocation intermediate and arose by means of two competing bond migrations. Bohlmann's studies have served to support this scheme. He reported finding caryophyllene, 1. and 9, in a 7,10 number of species of Berkheya 3.3 Syntheses.

A two year hiatus Intervened between the isolation of isocomene (1^) and the publication of synthetic efforts towards the natural product. 1 o Paquette and Han gained rapid access to the tric y c lic framework by the conjugate addition of the Grignard reagent of 3-bromopropionaldehyde ethylene ketal to a diquinane enone (Figure 4). After addition of methyllithium to the ketone, dehydration, and deketalization of the aldehyde, 10 was available to undergo stannic chloride-mediated ring closure. The resulting tricyclic alcohol was oxidized to dienone ^ which underwent stereoselective conjugate addition of lithium dimethyl cuprate from the convex face of 11. Wolff-Kishner reduction gave 1.

ÇH3 I. CHu''“3 ^ CuBr • MegS ISnCl4,C6H6 d d . 0 Z .C H ^ L i 2. Jones 3. SO C I2 , py 4. HOAc , HgO 10 1. LDA PhSeCI 2.MCPBA

(CHglgCuLi CH 2.H2Nr/H2, 0 H ~

Figure 4. Paquette and Han synthesis of isocomene. Me Me !. NaOMe 2 . KOI Bu 2B 0’C 0 3.1-A m O K, 0 CH3 I Me Me Me HgC 12 13

1. H2 , P» 2.K01BU, AmONO 3.NaOCI ,NHs 4. hv.MeOH

Me 1. L 1 AIH 4 Me TsOH 2. ArSeCN, CH2 CH2CI2 Me C Q r " ““' Me 3 .N0IO4 M . - < T " ' 4 .8 0 ’ C

Figure 5. Oppolzer's synthesis of isocomene.

Oppolzer^^ was able to prepare the methylated dienone ^ by

Robinson annulation and deconjugative méthylation in surprisingly good

(49%) yield (Figure 5). A thermal ene cyclization of 12 gave the tric y c lic ketone 1^ with a ll four of the isocomene stereocenters in their proper re la tiv e configuration. Hydrogenation of 13 and contraction of the interstitial 6-membered ring via the a-diazoketone gave tricyclic ester L4. Further manipulation of this intermediate gave

B-isocomene (4) which was readily isomerized to isocomene (_1 ). 0

2.H3O +

OEI 350 nm

15 1

Figure 6. Pirrung's synthesis of isocomene.

Pirrung envisioned a synthesis based upon an intramolecular [2 + 2] cycloaddition and carbonium ion chemistry to arrive at in seven steps and 40% yield^^ (Figure 6 ). The vinylogous ester, available from dihydroresorcinol in 2 steps, was elaborated to a dienone which was irradiated to give the [2 + 2] adduct 1^. The stereochemistry of the intramolecular photoaddition was directed by the vinyl methyl substituent on the si dechain of the dienone to give as a single product. Wittig olefination and acid-mediated cyclobutyl ring expansion gave isocomene (1 ).

Figure 7. Wender's meta-photocycloaddition and thermal rearrangement. An efficient route to isocomene utilizing an intramolecular meta- photocycloaddition of an olefin to an arene was developed in Wender's 1 fi laboratory (Figure 7). Arene olefin ^ was prepared by conjugate addition of (Ej-2-lithio-2-butene to methyl vinyl ketone followed by a reductive coupling of this product to o_-lithiotoluene. The meta- photocycloaddition of Jj5 gave only two of a possible 24 products. This selectivity was attributed to predictable electronic and conformational restrictions in the intermediate exciplex. The mixture of the photo products was thermally isomerized to and ultim ately hydrogenated to give 1 .

1. 6 H HCI u 2.K F.C H 3I

4 . T jO H , 0 ^ 0 oq ocetone V 18 1. HgNNHg , OH- 2. TsOH , CG"6 3. LDA , CH3T

1. H5"^SH, I.TsOH.CHgClg BF3 ElgO

2 CHgI.CoCOs 2 . PhgP *= CHg 0 HgO.CHgCN. %

20 19

Figure 8 . Dauben's synthesis of isocomene. 10

Dauben's route to Isocomene proceeded via a tric y c lic enone 17 developed in the e a rlie r Paquette synthesis (Figure 8). The key intermediate was monoprotected diquinane acid liB prepared in several steps by a Weiss-Cook condensation of 4,5-dioxohexanoate ethyl ester, hydrolysis, and selective ketone protection. After Wolff-Kishner reduction and deketalization, the appended carboxylic acid was condensed to form a 1,3-diketone which upon méthylation gave Once again, monoketalization, in this Instance with 1,2-ethanedithiol, was necessary before W ittig olefination. Paquette's enone was prepared in two steps from dithioketal

4. Silphinene.

4.1 Isolation and Natural Occurrence.

Silphinene (21_, Figure 9) was

found by Bohlmann in two geni of

plants. The roots of Si 1 phi urn

perfoliatum^^ and Liabiurn eggersii^

yielded the sesquiterpene whose

constitution was determined to be 2^

based on extensive NMR examination

and chemical transformations. The

21 Rg = H dioxygenated derivative ^ has also

2^ Rj^ = OH, Rg OCOCHgCHfCHgjg been detected in the roots of

Figure 9. Silphinene derivatives. Callilepis salicifolia^^. 11

H*.

Æ x - 8 <

23 24 21

Figure 10. Proposed biosynthesis of silphinene.

4.2 Biogenesis.

As in the case of isocomene, caryophyllene has been suggested as 18 the biogenetic origin of ^ . The isolation of 8 -a-hydroxypresilphiperfolene, a covalent derivative of carbonium ion provides important circumstantial support for this theory^^ (Figure 10). The proposed sequence culminates with a carbon-carbon bond migration within 2^ and proton loss in 24.

4.3 Synthesis.

19 Paquette and Leone-Bay developed an ite ra tiv e approach to ^ 20 based on the Marfat-Helquist annulation reagent (Figure 11). The e-hydroxy ketone was prepared by addition of this Grignard reagent with copper catalysis to 4 ,4-d1methylcyclopentenone. Deketalization and aldol 12 cyclization to the bicyclic 6-hydroxy ketone was achieved in one opera tion with aqueous acid. Mesylation of the aldol and elimination with diazabicycloundecene gave the kinetic product 2^. Methyllithium was added to 2^ to give the tertiary allylic alcohol that was transposed under oxidative conditions to enone 26^. The third ring was constructed by repeating this annulation scheme. Conjugate addition of the Marfat-

Helquist reagent to enone 2^ gave only the product of addition to the convex face of the molecule. Once again, acid hydrolysis sufficed to deliver the tricyclic 6-hydroxyketone 27^. A tric y c lic diene was prepared by standard methods and selectively epoxidized to give oxirane

28. The correct stereochemistry of the secondary methyl group was established by treatment of epoxide 28 with boron trifluoride etherate.

Wolff-Kishner reduction of the ketone product delivered 21_.

ltd's approach to silphinene employed the same bicyclo[3.3.0]- 21 octenone as Paquette, but arrived at 26 by a more laborious route

(Figure 12). Ketone derived from dicyclopentadiene, was trans formed in four steps to a tric y c lic primary alcohol. The appended primary alcohol was oxidized, e s te rifie d , and methylated to give 80.

The ester function in 30 was reduced to a methyl group in three steps and the internal ketal was unravelled to the iodoketone 31 with trimethyl s ily l iodide. The iodoketone ^ was transformed to in te r mediate 26 with diazabicycloundecene. From this point, Ito 's synthesis of the natural product was essentially the same as the Paquette route. 14

COOMe H I.HO'^OH,H+ H 0 - \ H 1. Jones 2.NoI0 4 ,O s 0 4 2. CHgNg

3 .N 0 BH4 3. L D A .C H g l H ô 4.HCI,MeOH çp OMe OM0

29 30

1. L iA IH 4 2.PCC 3.H2NNH2 . OH" A .M e jS iC i, NoI.CHjCN

DBU elher H 0

26 31 severol steps

Figure 12. ItÔ's synthesis of silphinene. 15

5. Si 1 phiperfolenes

5.1 Isolation and Natural Occurrence.

Bohlmann's examination of the roots of Silphium species yielded another variety of triquinane sesquiterpene. The si 1 phiperfolenes were isolated from the roots of S. perfoliatum and S. asteriscus as a mixture 18 1 of isomeric tricyclic olefins . Extensive H NMR analysis of these hydrocarbons and a number of synthetic derivatives determined them to be silphiperfol-6 -ene (32), 7aH-silphiperfol-5-ene (^ ), and

7BH-silphiperfol-5-ene (3^, Table 2 ). Since the appearance of this in itia l report, two oxygenated derivatives have been uncovered. The natural occurrence of the compounds has been summarized in Table 2.

Table 2. The natural occurrence of si 1 phiperfolenes.

32, R = H, M h = CH3 , Rg = H 36 35 R, 33, R = 0 «• «2 = CH3

plant derivative reference Espeletiopsis guacharnia 33 22 Li abum floribundum 36 6 Planatoa lychnophroids 34 23 Si 1 phi um asterisus 34,35 18 S. perfoliatum 32,35 18 16

34,35 32

Figure 13. Biosynthesis of the si 1 phi 1perfolenes,

5.2 Biogenesis.

As with isocomene and silphinene, caryophyllene was thought to be 1 ft the biosynthetic precursor of the si 1 phiperfolenes (Figure 13). The proposed biogenetic pathway for these triquinanes parallels the route for silphinene (see Figure 10) up to the rearrangement of carbocation

An alternative bond migration occurs in 23 giving rise to a secon dary carbonium ion that eventually collapses to the various silphiper- folene olefin s. Further support for the intermediacy of cation 2^ has come from the observed rearrangement of its covalent derivative,

8 -a-hydroxypresilphiperfolene, to 3^ upon treatment with acetic anhydride^^ (Figure 14). 17

Figure 14. Rearrangement of 8-a-hydroxypresilphiperfolene to 18

CH, CH,

CHc I. Brg 0 CH3 ' 03 2 .N0 OCH3 , ÇÙ, ^ COCH3 2 (CHg)gS ^^j^C«I0 CH3 erf CH3 OH

CH3 I.N o H . toluene

2 Li I , I. NoH . KH CH 3 COOCH3 Vk^OTt OMF CH, I, CuBr • MC2 S 2 .O 3 ; M e 2 S &H30 +

OH ' CH3-@-0CCI ^ C H 3 py CH, No2Cr04 2 à HOAc CH, 3.H2NNH2 , AC2O CH3 H OH- CH3 H

I . H i ,Pd-C I .C H j C L K H .C H jI 2.POCI3 EtjB

Figure 15. Paquette synthesis of optically active silphiperfol- 6-ene

(m i- 19

5.3 Synthesis.

The Paquette, Roberts and Drtina preparation of silphiperfol-6 -ene 24 was the firs t optically active synthesis of a triquinane sesquiterpene

(Figure 15). The optically pure keto ester ^ was available by

Favorskii rearrangement of the dibromide of (Rj-(+)-pulegone. Bicyclic

enone ^ was obtained from 37^ in four steps using a new annulation 20 reagent, l-tosyloxy-2-ethyl-2-propene. Marfat-Helquist annulation of

3^ gave a tric y c lic g-hydroxyketone which was transformed by standard methodology to enone 39_. S ilphiperfo l-6 -ene was prepared from 39 by a

sequence involving catalytic hydrogenation and clean o-methylation of

the ketone through its enol triethylboronate derivative. Methyl1ithium was added to the resulting methyl ketone and the alcohol was dehydrated

to furnish 3^.

As a consequence of this work, the absolute stereochemistry of

siIphiperfol-6 -ene was shown to be as in 3^ (Figure 15). With this knowledge, the absolute stereochemistry of the other si 1 phiperfolenes may be derived. Through the biosynthetic intermediate common to the

si 1 phiperfolenes and silphinene, we can also assign the absolute

configuration of silphinene (21, Figure 16). 20

32 21

Figure 16. Absolute stereochemistry of siIphiperfol-6-ene (32)

and silphinene (^ ).

6. Rentalenene

6.1 Isolation and Natural Occurrence.

The isolation and characterization of nr pentalenene (40) and pentalenic

acid (41)^^ (Figure 17) from the

mycelial cake of various species of

Streptomyces provided the Important

biogenetic link between the highly

40 = CHj, Rg = H oxidized pentalenolactones and humu-

41 R, = COoH, Ro = OH lene. The pentalenolactones have 1 Ù C Figure 17. Rentalene derivatives aroused much medicinal interest

because of their effectiveness as 25 antibiotics against both gram positive and gram negative organisms .

This broad spectrum of activity has been attributed to their ability to in h ib it the enzyme glyceraldehyde-3- phosphate dehydrogenase. 21

42 43 R= CHj, OH ^ ^ or CHg

Figure 18. Ohfune's study of the formolysis of a protoilludyl cation

equivalent.

6.2 Biogenesis

Ohfune's study of the formolysis of protoilludyl cation equivalents was important evidence for the evolution of pentalenene (40^) from

7 1 humulene (42) (Figure 18). The cation equivalents 4^ were prepared from humulene and upon solvolysis in refluxing formic acid yielded a mixture of pentalenene (40) and the bridged formate 4£ (40/44 = 3 /7 ).

Thus, 40 was synthesized before i t was isolated from Streptomyces. The accepted biogenetic pathway for pentalenene and its more highly oxidized metabolites is outlined in Figure 19^^. 22

/ 7 /

45 40

CO.H CO,H

41 pentalenolactone G pentalenolactone

Rj ,R2= 0

pentalenolactone H

^1= OH, Rg= H

Scheme 19. Proposed biosynthesis of pentalenene and its more highly

oxidized derivatives. 23

6.3 Synthesis.

28 Paquette and Annis published an approach to pentalenene that made use of the highly substituted bicyclo[3.3.0]octyl ketone 47 designed to allow for the selective deployment of functionality as the synthesis progressed (Figure 20). Ketone 47 was available from the addition of dichloroketene to silylenol ether after successive hydrolysis and ring expansion. The chlorohydrin functionality in 47 was used to unmask an a-chloro enone which readily entered into conjugate addition with lithio bis(3-butenyl)cuprate. A second chlorohydrin was generated by the addition of methyl magnesium bromide. The 3-butenyl appendage was subsequently ozonolyzed and the aldehyde carbonyl protected as its ethylene glycol acetal. This method allowed for the regiocontrolled installation of the double bond in 48 by dissolving metal reduction. The intact aldehyde side chain was subsequently unmasked for Lewis acid cyclization and oxidation to 4£.

With 49^ in hand, the remaining tasks were to in s ta ll the final secondary methyl group with the proper stereochemistry and to effect the fin al reduction to 40. The obvious approach, the addition of lithium dimethyl cuprate to a derived dienone, gave unexpectedly ketone 50, the product of approach from the concave face of the molecule! Reduction of this ketone, of course, gave epi-pentalenene. To remedy this situation,

Paquette and Annis prepared dienone Attempts to epimerize this toublesome methyl group using reductive methods under conditions of thermodynamic control (lithium in liquid aimonia, copper hydride) gave 24

I. CH 3 O H . H + Cl EtnSiO Cl HO .Cl p S iE t3 CIgCHCOCI 2 TsO H , CH3 OH XT ( C z W 3. C H gN g pentone 47 46 I Zn,HOAc 2(<^-^uLi

Cl w^CHO I. CHjMgBr 2.O 3 -.MegS 1.No, NH3 > d > 2. H3O + 3. H O -^ O H , H + 48 1. SnCI*. CgHg 2 .PCC

1. LDA.PhSeCI 2. HgOg 3. MegCuLi H gN N H g, 0 H~ 49 I. LDA.PhSeCI epi-pentalenene 2 HgOg

1.(Ph3P)3RhCI, El3SiH

2 .H g N N H g , 0 H “

40 51

Figuree 20. Paquette and Annis synthesis of pentalenene (4£) 25

I. Hg(N0g)2 « HO H HgO-THF

Z.KBr.HgO 3. NoBH4,02 DMF

L Acÿ.py E.PBr^,elher 3 .1 - AmONo, DMSO

HO HO, H Li I. Jones E1NH2-THF

54 53 52 BF^-EtgO CH2CI2 , -lo'c HQ H HO H HO H I.Se02.H20 i = KOH XWCOOMe V-COOH MeOH f 2. MnOg . / T~H HgO H KCN,MeOH HOAC H 41 55

Figure 21. Matsumoto's biomimetic preparation of pentalenic acid (_^) 26 only 50^. It was only by reduction with the very bulky reagent combina tion tris (triphenylphosphine)rhodium chloride and triethyl si lane that the desired epimer could be prepared. Wolff-Kishner reduction of this product gave pentalenene ( ^ ) .

Matsumoto's group reported an elegant biomimetic synthesis of 29 pentalenic acid (4^) from humulene (Figure 21). A mercuration- demercuration sequence gave two epimeric tric y c lic diols, which could independently be transformed to th e ir exomethylene derivatives 52 and

53. The unwanted epimer ^ was oxidized and reduced to The in te r nal cyclic ether was then cleaved with lithium in diethyl ami ne. Medium ring product 5^ contained the proper stereochemical alignment of the te rtia ry hydroxyl group, adjacent methine hydrogen, and trisubstituted olefin to simulate the transannular ring closure hypothesized for the biogenesis of 4£ and ^ (see 46 in Figure 19). The ring closure was initiated under aprotic conditions with boron trifluoride etherate to give triquinane 5^ in 20% y ield . Subsequent oxidation of the vinylogous methyl group gave pentalenic acid (41 ). 27

7. Senoxydene.

7.1 Isolation and Natural Occurrence.

In 1978 Bohlmann and Zdero reported the

isolation of an unidentified sesquiterpene from

3 0 the roots of Senecio oxyodontus . A year

la te r, based upon NMR studies of the hydro

carbon and an epoxide derivative, the sesqui

terpene senoxydene was formulated as an angular

triquinane possessing the unique carbocyclic

01 skeleton 5^ . Since these in itia l reports, neither senoxydene nor any derivative, has been found in Nature.

Figure 22. Proposed biosynthesis of senoxydene from humulene.

7.2 Biogenesis,

In terms of its biosynthesis, senoxydene (56) differed from other angular triquinanes isolated from plant sources. Bohlmann proposed that

31 56 arose from humulene rather than caryophyllene (Figure 22). This 28 suggestion seemed to be based largely upon structural considerations senoxydene could be derived from humulene by a less tortuous route.

At a later date, a second biogenetic route originating with 32 isocaryophyllene was devised for 56 (Figure 23).

Figure 23. Proposed biosynthesis of senoxydene from isocaryophyllene. CHAPTER I I

Two Approaches to Senoxydene (56) Based Upon the Favorskii

Reaction.

1.0 Statement of the Problem.

Our in terest in the total synthesis of senoxydene (56_) was due to two major considerations. F irs tly , the angular triquinane's structural assignment rested solely upon the interpretation of spectroscopic

evidence. An unambiguous preparative route to

^ would serve to confirm or disprove this pro

posal. Secondly, the stereocontrolled synthe

sis of this molecule demanded the solution of

four important problems. A strategy would have

ÊÊ. to be developed which created the central

quaternary carbon atom efficiently. Also, careful consideration must be given to the proper deployment of the four pendant methyl groups. Methodology that allowed for the in stallatio n of the secondary methyl group into the concavity of the bicyclo[3.3.0]- octane portion of 56^would be required. The regiochemistry of the double bond and the associated vinylogous methyl substituent must be controlled. F in a lly , the geminal dimethyl functionality had to be incorporated.

29 30

cOc

56 57 58

59 60

Figure 24. Retrosynthetic analysis of 56.

2.0 Elaboration of a Favorskii Reaction Product.

2.1 Retrosynthesis and Methodology.

Our first approach to the preparation of ^ envisioned tricyclic lactone 5^ as a reasonable precursor to the angular triquinane (Figure

24). Lactones of this type were known to undergo intramolecular

Friedel-Crafts acylation to give cyclopentenones under strongly acidic conditions'^. Achmad and Cavill have used polyphosphoric acid to prepare a bicyclo[3.3.0]octenone from the corresponding lactone in th eir synthesis of iridodial^’^. They suggested that the reaction proceeded through a carboxylic-phosphoric anhydride that cyclized to give the bicyclic enone (Figure 25). Eaton advocated a 10:1 methanesulfonic 31

pp/k

Figure 25. Mechanism for the solvolytic cyclization of lactones to

enones.

acid:phosphorous pentoxide mixture as a reagent superior to polyphos phoric acid, due to the greater solubility of organic compounds in this medium and the ease of manipulation . This procedure was the method of choice in the closure of a complex bisiactone to a dicyclopentenone in

Paquette's dodecahedrane synthesis^G.

A four carbon annulation of unsaturated bicyclic lactone 58 would give access to ^ (Figure 24). For this task, the Marfat-Helquist 20 procedure was considered . A fter conjugate addition of 2 -(2 -e th y l-l,3 - dioxolanyl)magnesium bromide and acid cyclization, a 6-hydroxy lactone should resu lt (Figure 26, R = H,OH). Following oxidation of the hydroxy lactone, the keto lactone (Figure 26, R = 0) would be converted to the o lefin by W ittig condensation (Figure 26, R = CHg). Catalytic hydro genation should set the correct relative stereochemistry of the secondary methyl group. 32

F in a lly, lactone ^ could be derived from

a hi cyclic lactone prepared by the cyclization

of 59 with aqueous acid (Figure 24). Func

tional i zed cyclopentenone 5^ should result from

the Favorskii ring contraction of the

Figure 26. Tricyclic o,e-dibromide of 2-isopropylldene-4,4—

lactone. dimethylcyclohexanone (60).

Ample precedent for these expectations lie with the extensive studies of the Favorskii rearrangement of pulegone dibromide by Achmad and C avill^^’^^, Marx and Norman^®, and several others^^ (Figure 27).

RO^ -Co' OR I______

Figure 27. Favorskii rearrangement of pulegone dibromide and the

cationic cyclization of pulegenic acid derivatives.

The mechanism advanced by these workers^^'^^b.c ^p^okes the formation of a cyclopropanone intermediate which suffers nucleophilic attack by hydroxide or alkoxide ion. Collapse of the tetrahedral intermediate is directed by the ejection of bromide ion. The resulting acid or ester was cyclized to the lactone with aqueous acid in a separate operation. 33

Pulegone has been transformed to pulegenic acid or ester in yields of 42-75% with base systems ranging from aqueous sodium hydroxide to sodium ethoxide in ethanol^^’^®’^^. The cationic cyclization of pulegenic ester in 10 ^hydrochloric acid was accomplished in 90% yield34'37. major drawback of the use of a di bromide in effecting this ring contraction has been its in s ta b ility . No worker has success fully isolated and characterized this compound.

A stable derivative of an a,6-unsaturated ketone that was known to undergo a Favorskii ring contraction was the corresponding epoxide^O.

Achmad and Cavill treated pulegone epoxide with sodium ethoxide in ethanol and obtained a 3:1 mixture of a bicyclic lactone and a trans.trans hydroxy acid in 20% yield (Figure 28)^^.

Figure 28. Achmad and C av ill's rearrangement of pulegone epoxide.

Reusch and Mattison separated two isomeric pulegone epoxides and demonstrated th e ir rearrangement with potassium t e r t-butoxide in glyme to be stereospecific^^ (Figure 29). Yields were not reported. 34

KCHBv

Figure 29. Reusch and Mattison's rearrangement of pulegone epoxide.

The synthetic u t ilit y of the rearrangement of pulegone epoxide was limited due to the formation of many side-products^^'^l (Figure 30).

The problems occurred in the collapse of the tetrahedral intermediate.

Unlike the dibromide rearrangement, this intermediate lacked an e ffic ie n t leaving group to direct the ring contraction. Three primary products were formed that underwent further transformations under the strongly basic conditions used.

-0 » 1 ? " -OR

-A

CO,R ?" I CAR o

Figure 30_^ Mechanism of the Favorskii rearrangement of pulegone

epoxide. 35

2.2 Synthetic Work.

The route for preparing M paralleled a previous synthesis of

(±)-pulegone^^. This approach was adopted for its procedural sim plicity and amenability to scale-up (Figure 31). Carbométhoxylation of

4 ,4-dimethylcyclohexanone with dimethyl carbonate and a mixture of sodium and potassium hydrides gave 61^ in 89% yield^^. Ketalization of

6^ and treatment with two equivalents of méthylmagnésium iodide produced

62. Exposure of 62_ to aqueous acid at reflux gave an 89% yield of

2-isopropylidene-4,4-dimethylcyclohexanone (60).

COOCH, , h o ^ o h , '

0=C(0CH3)2

CH3 CH3 T H F 'ÇH3

i l 62

CH) C H ; 0.

MCI _ . HgO, CH) CH3OH

Figure 31. Preparation of 2-isopropylidene-4,4-dimethylcyclohexanone 36

0 jL I Bfg, _Ü

A:' CH, CH3OH CH3

60 59

HCI

Figure 32. Preparation of b1cyclic lactone 6^.

The method of Marx and Norman was adapted to convert 1sopropylldene op ketone 6jO to the methyl ester of 5^ (Figure 32). Bromination of 60^ In ether solution buffered with sodium bicarbonate gave an unstable dibromide which was filte re d and added directly to a solution of sodium methoxlde In methanol. The Favorskii product 59 was Isolated a fter 12 hours at room temperature In 50% yield with the recovery of 27% of the starting ketone 60. Ester 5^ was cyclized with aqueous acid to lactone

63 In 82% y1eld34'37

The next step to be considered was the reglocontrolled Introduction of unsaturation Into 63^ to give the bicycllc lactone 58^ (Figure 33).

FunctionalIzatlon of 6^ as the a-bromolactone 64 was achieved by the 44 reaction of the lithium enolate of the lactone with 1,2-dlbromoethane . 37

Attempts to in itia te dehydrohalogenation of 6£ to 58 by a trans-diaxial elimination are summarized in Table 3. A variety of base and solvent systems were examined and, in a ll but two cases, unreacted 64 was obtained. In run 6, a mixture of lithium bromide and lithium carbonate heated in dimethylformamide at 158°C for 30 hours gave a 21% yield of unsaturated lactone Unfortunately, attempts to repeat this reaction gave only recovered a-bromolactone 6_4. Run 9 produced an unusual resu lt. With lithium fluoride and lithium carbonate in hot hexamethyl- phosphoric triam ide, nucleophilic attack upon the a-bromine atom by fluoride ion gave lactone 6^ in 48% yield. It appears that the anti-periplanar proton, lodged in the concavity of the bicyclic lactone adjacent to the geminal methyl substituted carbon atoms is too hindered to be abstracted by base.

— - CHj CHj

1.LDA.THF I.LDA.THF P2O5 . CH3SO3H, 2. E ie rj a.P h S eB r 3.H2O2. HOAc 50*C

0 Br CHj CH3 CHj'CHj CHj CH3

64 65 66

Figure 33. The derivatization of lactone 6^. 38

Table 3. Dehydrohalogenation of 54.

0. V ^CH3 0 tH , ------> \ 'CHa CHj C H g C H j C H j 64 58

Run Conditions Product Reference

1 DBU,CyHg,A recovered M 44 2 y-c o llid in e , 100° recovered 64 45

3 tBuOK/tBuOH.A recovered 64 —

LiBr/LiCOg/DMF 45

4 90°, lOh recovered &4 5 120°, lOh recovered 6£ 6 158°, 30h 58 in 21% y ield *

LIF/LiCOg/DMF 46

7 90°, 96h recovered 8 155°, 22h A complex mixture including 6j4

LiF/LiCOg/HMPT 46

9 90°, 48 h in 48% yield.

* Two attempts to reproduce this result failed and gave only recovered 64. 39

Selenenylation of the lithium enolate of 63^ followed by elimination gave only unsaturated lactone 6^^ (Figure 33). The identity of this undesired regioisomer was apparent from the sim plicity of its NMR spectrum. The spectrum featured only three sharp singlets at 6 2.32

(4H), 1.45 (6H) and 1.25 (6H), indicative of the high order of symmetry in 65y

Lactone §2 could be converted under Eaton's conditions (10:1 methanesulfonic acid:phosphorous pentoxide, 5Q°C) to b icyclo [3.0.0]- octenone 66 in 70% yield^^ (Figure 33).

'N oCN CHjLi W— t-Bu

Figure 34. Matsumoto's synthesis of 66.

2.3 Previous Preparations of 66^.

Matsumoto prepared 66^ as an intermediate in his synthesis of the formannosane skeleton (Figure 34). This approach began with a double displacement of a vicinal dibromocyclopentane with cyanide ion. The dicyano derivative was hydrolyzed to the dicarboxylic acid and converted to the bis(methylketone) with methyllithium. Cyclization of the diketone with potassium te r t-butoxide gave 6^. 40

i)Br^ OSOCIi OH

:><: a o "' ^

§§.

Figure 35. Thorén's synthesis of 66.

An improved procedure for the preparation of 6^ was reported by

Thorén in his total synthesis of the velleral skeleton^^ (Figure 35).

Favorskii ring contraction of 2-bromo-2-carbomethoxy-4,4-dimethylcyclo hexanone gave Matsumoto's diacid. The bis(acid chloride) was prepared from this diacid and treated with lithium dimethylcuprate to give the bis(methylketone) with more reproducible yields. Cyclization with ethanolic sodium hydroxide gave an excellent yield of 6^.

•>^CCC, ^ ^0

CICHgCHgCI SPh

P I.CHgLi a.HgCIg.HgO H3O +

Figure 36. Magnus synthesis of 66. 41

Magnus pioneered an interesting combination of vinyl si lane chemistry with the Nazarov cyclization to arrive at 66^^ (Figure 36)

Addition of the acid chloride to the vinyl si lane with s ilv er fluoro- borate gave an aryl sulfide derivative of a bicyclic enone. Upon reaction with methyl lithium and hydrolysis of this enone, 66 was obtai ned.

56 57 67

r~\ \ +

Figure 37. Retrosynthetic analysis of 56.

3.0 Attempted Favorskii Ring Contraction of ^ Bicyclic Precursor.

3.1 Retrosynthesis and Methodology.

With the in a b ility to prepare lactone 58^ from 6^, our in itia l plans

for the synthesis of the triquinane had to be abandoned. A new route

reordering the sequence of events was devised (Figure 37). I t was 42

proposed that a Favorskii rearrangement of a derivative of isopropylidene ketone 67 should lead to tric y c lic lactone Ketone 67^ would be available by the isopropylidenation of bicyclo[4.3.0]nonane 68.

Mukaiyama has developed a method of e ffic ie n tly performing this Cl Cl operation . Regiospecifically generated silyl enol ethers have been added to a mixture of titanium (IV) chloride and 2,2-dimethoxypropane in dichloromethane to give g-methoxy carbonyl compounds (Figure 38).

Treatment of the g-methoxy ketone with diazabicycloundecene produced the CO isopropylidene ketone .

O HjO" OBU

Figure 38. Mukaiyama's method for the isopropylidenation of ketones.

This reaction has been observed with many d ifferen t combinations of

s ily l enol ethers and dimethyl ketals. I t was hypothesized that the

Lewis acid and ketal form an electro philic complex which adds to the enol ether. A g-methoxy ketone intermediate chelated by the Ti(IV) atom

results. Acid hydrolysis gives the free condensation product.

A four carbon annulation method analogous to the Marfat-Helquist cyclization^^ was proposed for the preparation of bicyclic ketone 68

(Figure 37). The conjugate addition of 2-methyl-(2-ethyl-1,3-dioxo- 43

1anyl)magnesiuni bromide®'^ to 4 ,4-dimethyl cycl ohexenone^^ followed by

hydrolysis and cyclization of the appended ketal and catalytic

hydrogenation should give 68.

0 o H ) I. 1. BrMgx Cul ^ Hg/Pd-C

2. H3O+

59 68

Figure 39. Preparation of bicyclo[4.3.0]nonanone 68.

3.2 Synthetic Work.

The elaboration of 4,4-dimethylcyclohexenone to the bicyclic ketone

^ was accomplished in four steps and with 58% overall efficiency

(Figure 39). The Grignard of 2-methyl-2-(2-bromoethyl)-l,3-dioxolane was prepared according to the Ponaras procedure with three equivalents of magnesium and entrainment with 1,2-dibromoethane®^. With copper(I)

catalysis and a three-fold excess of this reagent, conjugate addition to

4,4-dim ethyl- cyclohexanone at -78°C was observed. The crude product 20 was cyclized with dilute hydrochloric acid in aqueous tetrahydrofuran

and after purification a 60% yield of b1cyclo[4.3.0]nonanone 69 was isolated. Catalytic hydrogenation of 69^ in glacial acetic acid gave 68

in 96% y ie ld . That bicyclic ketone 68 was a single isomer was evident 13 from the twelve lines observed in its C NMR spectrum. The secondary methyl group in 68 was assigned its o-configuration based upon the 44 expectation that hydrogen would approach from the less hindered e-face of the molecule. This assignment was confirmed by the NMR spectrum of 6^. The bridgehead methlne proton adjacent to the carbonyl (6 2.68) appeared as a triplet exhibiting a d s coupling constant of 7.5 Hz.

OTMS

l-LDA [ I > TiCl4 n .R. (Mi O)2CI*U2

Figure 40. Unsuccessful 1sopropylldenation of 68.

The 1sopropylldenation of M was the next synthetic operation to be attempted. The kinetic enolate of 68^ was generated with lithium diIsopropyl ami de at -78°C^® and trapped with trimethylsilyl chloride In 52 the presence of triethyl ami ne (Figure 40). The application of

Mukaiyama's method^^ for Lewis acid mediated aldol condensations to this silyl enol ether was unsuccessful. The enol ether failed to condense with the tltanlum(IV) chloride: 2,2-dlmethoxy-propane mixture; only the product of hydrolysis of the enol ether, ketone 68^, was observed.

The inspection of models suggested that cIs fusion of the blcyclo-

[4.3.0]nonanone ring system In the enol ether forced the «-oriented methyl substituent on the cyclopentane ring Into close proximity with the enollc oxygen and the site of condensation. It Is believed that the resultant crowding impeded formation of the chelated g-methoxy ketone

Intermediate, thereby curtailing reaction. The solution was to eplme- rlze 68 to the trans-fused ketone 70 and proceed from there (Figure 41). 45

N q OCH, I. TiC l4 (M«0l2CW*2 2.D B U

68 70 71 6 7

Figure 41. Successful 1sopropylidenation of the bicyclo[4.3.0]-

nonanone 70.

The epimerization of 6iB to TiO was achieved efficiently (85%) with a

c atalytic amount of sodium methoxide in methanol (Figure 42). The

reaction was complete in 8 hours, as determined by NMR. The methyl

doublet exhibited a considerable upfield shift (6 1.25 to 1.00) upon

alteration of the ring fusion.

The kinetic enolate of TiO was generated and trapped as silyl enol

ether 71 in quantitative yield. Condensation of 71 with titanium(IV)

chloride and 2,2-dimethoxypropane gave the crude e-methoxy ketone, which O CO was treated directly with diazabicycloundecene and 3A molecular sieves .

The bicyclo[4.3.0]nonanenone 67 was isolated in 58% yield after purifica

tio n.

A great deal of d iffic u lty was encountered in the bromination of 67.

Isopropylidene ketone 67^ did not consume elemental bromine when stirred

in ether for 12 hours at room temperature. When 67^was treated with

pyridinium bromide perbromide in acetic acid at 10°C, the solution turned

a bright blue a fter 30 minutes. Examination of this solution by thin

layer chromatography revealed that a ll of 67 was consumed but at least 46 six products had been formed. Subjection of this mixture to Favorskii 37 reaction conditions (sodium methoxide in methanol, reflux for 3hours ) only led to further decomposition. Evidently, any dibromide that may

have been formed was too unstable to survive the mild bromination conditions. Hence, it was very unlikely that clean Favorskii rearrangement would be observed.

chj CojH

67 72

Figure 42. Epoxidation of 67,

Due to the problems encountered in the preparation of the dibromide of 67_, recourse was made to the use of the stable a,e-epoxy ketone 72 in

the Favorskii ring contraction (Figure 42). While treatment of 67 with C? alkaline hydrogen peroxide gave only low yields (10%) of 7^ , buffered

peracetic acid in acetic acid delivered a 70% yield of this compound as a

single crystalline isomer^® (mp 92-92.5°C).

Three d iffe ren t reaction conditions for the Favorskii rearrangement were examined. Epoxide 72 was heated at reflux in ethanol with sodium

ethoxide^^. Thin layer chromatography of the reaction mixture revealed a m u ltip lic ity of products and much baseline m aterial. The NMR and IR

spectra were even less encouraging.

Epoxy ketone 7^ was treated under the mild conditions developed by

House and Gilmore for the rearrangement of pi peri tone oxide^®. Epoxide 47

7^ was stirred with potassium t e r t-butoxide in dimethoxyethane for 2 hours at 0°C and warmed to ambient temperature over several hours. The thin layer chromatogram of the reaction mixture displayed eight products, the major constituent being the starting epoxide. Examination of the

NMR and IR spectra confirmed the presence of 77. However, the IR spectrum revealed the presence of a conjugated enone (v 1700, 1640 cm"^) and a hydroxyl group (v 3400 - 3500 cm"^). A vinyl proton signal (6 5.2) was apparent in the NMR spectrum. A product derived from ring cleavage of the epoxide would be consistent with these data (Figure 43).

An attempt was made to enhance the reactivity of the potassium enolate by the addition of a crown ether. Epoxide 77 was stirred with a mixture of 5 equivalents of potassium hydride and a catalytic amount of

18-crown-6 in tetrahydrofuran. After 16 hours at room temperature and

5 hours at re flu x , only the epoxy ketone 72 was recovered.

Figure 43. Ring cleavage of epoxy ketone 72^.

From these results, it was evident that the trans-fused bicyclo-

[4.3.0]nonanone framework was too rig id to accommodate the formation of the cyclopropanone intermediate postulated for the Favorskii rear rangement of a,6-epoxy ketones. CHAPTER I I I

The Total Synthesis of Angular Triquinane 56: Alleged Senoxydene.

Tms

56 73 74

BrMg TMS A

BrMg' 75

Figure 44. Sequential annulation approach to 5^.

1.0 Diketone Coupling Approach to 56.

1.1 Retrosynthesis and Methodology.

In view of the d iffic u ltie s encountered with the Favorskii ring contraction strategies, our retrosynthetic dissection of 56 was reconsidered. A new approach involving two sequential four-carbon annulations of a cyclopentane nucleus containing a geminal dimethyl

substituent was conceived (Figure 44). An opportunity to develop a new methyl cyclopentene annulation scheme employing a diketone coupling

48 49

59 reaction developed by McMurry presented it s e lf with the disconnection of ^ to the 1,5 dicarbonyl compound 73.

McMurry has prepared cyclic olefins with ring sizes varying from 4 to 16 carbon atoms in synthetically useful yields (67-95%)^^. According to this method a dicarbonyl compound is slowly added to a suspension of finely divided titanium metal in 1,2-dimethoxyethane. The proposed mechanism for this transformation involves reduction of the dicarbonyl

T;(I) T i(l)

73 o = T i(n )

Figure 45. Mechanism for the proposed coupling of diketone 73.

compound to a diradical on the titanium metal surface®® (Figure 45).

The diradical then combines to form a glycol. The glycol intermediate can be isolated at low temperature; however, heating at reflux results in deoxygenation to the olefin®^. In the case of diketone 73, an olefin isomer of 56_, triquinane 73, would result, therefore migration of the double bond would ultim ately be required. 50

The vinyl si lane appendage of bicyclic ketone 7£ was to serve as a masked carbonyl functionality for 1,5 diketone 73 (Figure 44). Ketones have been generated regiospecifically from vinyl si lanes by a two-step 62 procedure . The vinyl si lane is epoxidized and hydrolyzed to the ketone with aqueous acid. Compound 7^ would result from the regioselective alkylation of bicyclo[3.3.0]octanone 75 with a reagent developed by 63 stork, (Ej-l-iodo-3-(trimethylsilyl)but-2-ene . This reagent has seen wide use as a methyl vinyl ketone equivalent in a modification of the 64 Robinson annulation .

Finally, two routes for arriving at bicyclic ketone 7,5 were planned

(Figure 44). An adaptation of the sequence used to prepare bicyclo- 54 [4.3.0]nonanone M employing the Ponaras Grignard reagent , 2-methyl-

2-(2-ethyl-l,3-dioxolanyl)magnesium bromide, should give 75. The second approach would exploit a method developed by Conia for the preparation of c is -perhydropentalenones sim ilar to 73®^. The exposure of

E,G-ethylenic ketones to high temperatures has resulted in ene cyclization of the enolized ketone and distal double bond to give methylcyclopentane annulated products (Figure 46).

0 320* c

Figure 46. Ene cyclization of e,ç ethylenic ketones to

ci s-perhydropentalenones. 51

An important feature of this reaction was the situation of the secondary methyl group of 77 inside the cavity of the cis-bicyclo[3.3.0]octanone.

The mechanism proposed by Conia is a two-step sequence in itia te d by the thermal enolization of the carbonyl compound®^. There follows a concerted six-electron process involving a 1,5 hydrogen shift and 3 bond formation between the a and e centers of the molecule. A suitable, e,ç-ethylenic ketone for the synthesis of bicyclic ketone 7^ would be prepared by the conjugate addition of 3-butenylmagnesium bromide to 4,4- dimethylcyclopentenone®^ (Figure 44).

0 cHo 0 0

W - „ II \J 0

78 79

0 KOtBu II Hg/Pd-C //

80 75

Figure 47. Preparation of 75^ using the Ponaras Grignard reagent.

1.2 Synthetic Work.

1.2.1 Preparation of 7^.

The route used to prepare bicyclo[4.3.0]nonanone 68^was easily modified for the synthesis of bicyclo[3.3.0]octanone 75. Conjugate 52 addition of 2-methyl-2-(2-ethyl-l,3-dioxolanyl)magnesium bromide to

4,4-dimethylcyclopentenone with copper(I) catalysis at -78°C gave a 67% yield of 78 (Figure 47). Keto ketal 78 and a catalytic amount of pyri dinium tosylate were heated at reflux in acetone:water (19:1, v:v) to give crude 1,6 diketone 7^. Exposure of 79^ to potassium te rt-butoxide in hot tert- butyl alcohol gave an 86% yield of e,y-unsaturated ketone

80. The strongly basic conditions required for the ring closure of 79 was in sharp contrast to the dilute aqueous acid used for the cycliza tion to bicyclo[4.3.0]nonanone 69^ (see Figure 39). Stirring 78 with dilute hydrochloric acid in aqueous tetrahydrofuran led to the isolation of diketone 7^ rather than dehydration to the corresponding bicyclic enone. Furthermore, the deconjugated product, enone BO^, not the a,B-unsaturated ketone, resulted from the vigorous conditions necessary for the aldol cyclization of di ketone 79^. Thus g.y-unsaturated enone 80 was the product of thermodynamic control.

A significant amount of ring strain was relieved by the isomeriza tion of the olefin from the a,g to the g,y position. The thermodynamic stabilization gained by isomerization of the double bond away from the bridgehead carbon atom outweighed destabilization due to deconjugation and lower substitution of the olefin.

As in the case of bicyclo[4.3.0]nonanone 68, catalytic hydrogena tion of 80 gave bicyclic octanone 7^ as a single isomer (Figure 47).

The a-configuration of the secondary methyl substituent of 75 was assumed on the basis of litera tu re precedent. Both 75 and its secondary methyl epimer have been isolated previously from the alkaline hydrogen 53

peroxide degradation of a- and B-pipitzols^®. The literature value for

the chemical s h ift of the doublet due to the secondary methyl

substituent in the NMR spectrum of 7^ (5 1.0) is in excellent

agreement with the value observed for our synthetic product (6 1.02).

In contrast, the chemical s h ift for the methyl doublet of the epimeric

degradation product was 6 1.11.

9 I Cu 1 h 20min. ^ -

81 75

Figure 48. Thermal ene route to 75.

A more convenient preparation of 75 employing Conia's thermal ene

cyclization was devised®^ (Figure 48). A 65% yield of purified e,ç-

ethylenic ketone 81^ was obtained from the conjugate addition of

3-butenyl magnesium bromide to 4,4-dimethylcyclopentenone with copper(I)

catalysis at -78°C. 4-(3-Butenyl)-3,3-dimethyl cyclopentanone (81) was

sealed under vacuum in Pyrex tubes and heated at 320°C for 80 minutes.

There resulted a 78% yield of purified 75, from this vigorous treatment.

The overall efficiency of this procedure was 51% of theory, contrasting

favorably with the 22% overall yield of the first route. 54

D " B u L f

t m e d a

75 8 3

Figure 49. Shapiro olefination of 7^.

1.2.2 Some transformations of 7^.

To explore further the chemistry of bicyclic ketone 7^ several transformations were attempted with interesting results. Forcing conditions were required to prepare £-toluenesulfonylhydrazone 8269

(Figure 49). A mixture of 7^, £-toluenesulfonylhydrazine, and

£-toluenesulfonic acid was heated at reflux in ethanol for 16 hours.

Submission of 82 to the Shapiro olefination conditions, four equivalents of £-butyllithium in tetramethylethylenediami ne at -30°C with a proton quench, gave olefin 83 in 81% yield^^. That 8^ was the trisubstituted olefin was apparent from the NMR spectrum. A single vinyl proton signal (s , IH) was observed at 6 5.4. This was unusual because the

Shapiro reaction generally delivers the less substituted olefin as the product^^.

A study of the enolization of ketone 7^ under conditions of thermo dynamic and kinetic control demonstrated that the direct alkylation of

75 could not be achieved with the desired regiospeelficity. Using the 55

conditions proscribed by House for the generation of the thermodyna mically favored (trimethylsilyl)enol ethers, viz., trimethylsilyl rp chloride and trie th y l ami ne heated at reflux in dimethylformamide , 7^ gave a single product which was shown by the NMR spectrum to be the

less substituted regioisomer (Figure 50).

T M S Cl » ° T M S Et3N ^

DM F

Figure 50. Thermodynamic enolization of 75.

The kinetic deprotonation of 7^ was found to be unsatisfactory also. Slow addition of 7^ to 1.1 equivalents of lithium diisopropyl- amide in tetrahydrofuran at -78°C^^ gave the kinetic enolate mixture,

CO which when trapped with trim ethylsilyl chloride and trie th y l ami ne produced two s ily l enol ethers (Figure 51). The IR spectrum of the mixture featured two olefinic adsorptions (u = 1645 and 1670 cm"^) of approximately equal intensity and gave a complex NMR spectrum.

OTMS . PTMS 1.LDA - 7 8 * C

2.TM SCI

Figure 51. Kinetic deprotonation of 7,5. 56

These results made apparent the necessity of blocking ketone 7^ in order to direct enolization to the more substituted carbon atom. For

this task, the thiobutylmethylene protecting group was chosen due to its

s ta b ility towards chromatography and the high re ac tiv ity of the enolates 72 formed from the ketone derivatives .

'•N0OCH3 C2H5 OCHO

2 . nBuSH H+ 75 m

Figure 52. Preparation of thiobutylmethylene ketone

Thiobutylmethylene ketone 84 was accessible from the a-formyl

derivative of 7^ (Figure 52). Formylation of 7^ was achieved in 93% y ield with a mixture of sodium methoxide and ethyl formate in benzene.

Treatment of the a-formyl ketone with n^butanethiol, £-toluenesulfonic

acid and magnesium sulfate in benzene gave a 72% yield of &4 a fte r 20

hours at reflux^^*^^. 57

SBu I. LDA ,- 3 0

TMS

84 85 R= CHSnBu

3-HO^o' ^ 74 R= Hg

1.MCPBA

2 .H3 O +

73 76

Figure 53. Methylcyclopentene annulation by a diketone coupling

reaction.

1.2.3 Alkylation of 84

With the preparation of thiobutylmethylene ketone ^ i t was now possible to control the regiochemistry of the alkylation of the c is-perhydropentalenone nucleus. By using a combination of Stork's allylic iodide, (E)-l-iodo-3-(trimethylsilyl)but-2-ene^^, and 8^ as the reactive nucleophilic component, we anticipated good yields of alkyla tion product (Figure 53). These hopes were realized with a 67% yield of purified 8^, after addition of the allylic iodide to the lithium enolate of 84 at 0°C and s tirrin g the mixture at room temperature for 3 hours.

Alkylation product ^ was heated at reflux with a mixture of 25% potassium hydroxide and diethylene glycol (3:4,v:v) for 48 hours to give 58 crude 74 in 70% yield^^’^^. To arrive at 1,5 diketone 73, a two step 62 sequence was mandated . The vinylsilane functionality of 74 was epoxidized with m;-chloroperbenzoic acid in dichloromethane in the presence of solid sodium bicarbonate. A quantitative yield of the epoxysilane was isolated with no indication of any competing Baeyer-Villiger oxidation.

The ketone functionality in 73 was introduced by hydrolysis of the epoxysilane with a l: l( v : v ) combination of methanol and 20% aqueous sulfuric acid. Diketone 73 was isolated in 74% overall yield for the two steps.

To attempt the McMurry coupling of diketone 73, the titanium metal catalyst was prepared by heating at reflux for 15 hours a mixture of tita n iu m (III) chloride and freshly prepared zinc-copper couple in dimethoxyethane^®. Diketone 73 was added to this suspension at room temperature over 6 hours and heated at reflux for 16 hours more. After s ilic a gel chromatography and further purification by vapor phase chromatography, a low yield (1%) of triquinane 76, a double bond isomer of was isolated. The sample of 76 isolated from this reaction mixture was impure, as evidenced by the NMR spectrum. However, the proper molecular ion for 7|5 was observed in the mass spectrum.

The titanium mediated coupling reaction was found to be unsatis factory for this application not only because of the low yield and impurity of 7^ but i t also frequently fa ile d . Of the five attempts at this coupling, only one resulted in the isolation of 73. These considerations, along with the necessity of undertaking the further transformation of 73 to double bond isomer 56_, led to the abandonment of this approach. 59

TMS,

56 86 87

SBu TMS

84 88

Figure 54. Retrosynthetic analysis of 56^.

2.0 Synthesis of Triquinane 56.

2.1 Retrosynthesis and Methodology.

Due to the in e ffic ie n t conversion of 7^ to triquinane 76 another method of constructing the methylcyclopentene ring of 56 was devised

(Figure 54). Aldol cyclization of 1,4 diketone 86 followed by appropriate reduction of the tric y c lic enone product should deliver

triquinane The 2-butanone segment of 86 would be unraveled from the

vinylsilane appendage of Ultimately, 87 would arise from thiobutyl methylene ketone 8^ and (E j-l-io d o -2-(trim eth ylsilyl)b u t-2-en e (88).

While isomeric with Stork's a llylic iodide, &B was a new compound and

its use would constitute a novel methylcyclopentenone annulation scheme. 60

I.CHsMgl Br

ClgSi-^ TMS—^ TMS—(\ ^ Z.Brg Br 89 OH l.MsCI TMS, I. Mg

2 . 2 T M S I

90 M

Figure ^5. Preparation of allylic iodide 88.

2.2 Synthetic Work.

2.2.1 The preparation of 88.

The synthetic route to allylic iodide 88 had a convenient starting

point in the form of known allylic alcohol 9c/^ (Figure 55). The

preparation of ^ presented in this work was a significant modification of the published routes. Trichlorovinylsilane was added to three equi valents of méthylmagnésium iodide in ethyl ether. Since the trim ethyl-

vinylsilane formed codistills with ethyl ether, the reaction mixture was quenched, washed with brine, and dried. Elemental bromine was added directly to the cooled ethyl ether solution and the Isolated dibromide gave an 80% yield of product for the two steps. The dibromide was shaken in diethyl ami ne for 20 hours at room temperature and a fter distillation a 47% yield of l-bromo-l-(trimethylsilyl)ethylene (89) was obtained. A lly lic alcohol 90 was prepared in 60% yield by the action of 61 the Grignard reagent of vinyl bromide 89 upon acetaldehyde. Mesylation of 92 and S.,2' displacement of the mesylate with sodium iodide in ace- 75 tone gave 88 in 60% yield for the two steps .

TMS.

SBu . L D A , - 3 0

TMS 2 88 91 R= CHSnBu

3 ^ R= "2

1.MCPBA KOIBu 2.H3O + THF

86 92

I. U / N H 3 POCI3 CsHs

93 R= 0 56 ^ R= H, OH

Figure 56. Preparation of

2.2 Preparation of 56.

The methodology employing a lly lic iodide 88 proved to be an effective means of methylcyclopentenone annulation. The lithium enolate of 84 was alkylated at -30°C In tetrahydrofuran to give a 45% yield of 62

^ after purification (Figure 56). Removal of the thiobutylmethylene

protecting group by heating with a mixture of 25% aqueous potassium

hydroxide and diethylene glycol (3:4,v:v) at reflux for 48 hours was 72 73 complicated by the formation of an unidentified side product ’ .

Consequently, a 47% purified yield of 87 was obtained. The regio-

specific alkylation of bicyclo[3.3.0]octanone 7^ was accomplished in 22%

overall yield. 62 As before, a two step sequence was required to transform vinyl

silane ^ into 1,4 diketone 86^, Epoxidation of 8_7 with m-chloroper-

benzoic acid in dichloromethane gave a quantitative yield of the

epoxysilane with no evidence of Baeyer-Villiger oxidation. Hydrolysis

of the epoxysilane to 8^ with a mixture of methanol and 20% aqueous

sulfuric acid (l:l,v:v) gave a 65% yield of purified product.

Formation of the fin al bond of the tric y c lic framework with 86_

proved to be a very efficient transformation (Figure 56). An excellent yield (82%) of tricyclic enone 9^ was obtained from the aldol cycliza

tion of ^ with two equivalents of potassium tert-butoxide in tetra

hydrofuran. When heated at 40°C, the reaction was complete within one

hour. The crude 92^ isolated was very pure, registering one peak by

vapor phase chromatography. This 9^ was customarily submitted to

reduction with lithium in liquid ammonia at reflux (-33°C) without

further p u rification. The product of this reduction was determined to

be a single compound from the NMR spectrum and assumed to be a ll

cis-fused tric y c lic ketone 93[. This assignment was based upon the

propensity of dissolving metal reductions to give the product of

thermodynamic c o n tro lS u b m is s io n of crude tric y c lic ketone 93 to 63 sodium borohydride reduction gave two epimeric alcohols. The epimers were separated by s ilic a gel chromatography and both were crystalline.

The product eluted f ir s t , 94A, melted at 50-52°C, while the alcohol with the longer retention time, melted over the range 78-80°C. Each was isolated in 32% yield.

Alcohol 94A was submitted to Dr. James P. Springer of Merck,

Sharpe, and Dohme for single crystal x-ray analysis. It was shown to be the 6-hydroxy epimer with the adjacent methyl substituent fixed in the

6-configuration (Figure 57). This implies that while reduction of

9^ with sodium borohydride was not stereoselective (a 1:1 mixture of epimers resulted), protonation of the lithium enolate formed by the

C2 016

CIS

Figure 57. Computer generated drawing of 94A. 64 dissolving metal reduction of tricyclic enone 9^ was. Apparently the proton approached the enolate at the convex face of this molecule exclusively, giving only one diasteromer of 9^. Consequently epimer 94a must have the a-hydroxy and e-methyl configuration.

With Dr. Springer's contribution to this project, we were now certain that appropriate dehydration of either 94A or 94B would result in formation of triquinane 56_. Dehydration of both alcohols, separately or as a mixture, with phosphorous oxychloride and pyridine in warm benzene gave ^ in modest yield^^ (Figure 56). However, comparison of the NMR spectrum of synthetic ^ (Figure 58) with the data provided 22 by Bohlmann and Zdero for the natural product (Figure 59) revealed important differences in the spectra of the two compounds (Table 4).

The most obvious divergence of the two spectra was observed in the chemical shifts of the protons of the geminal dimethyl substituent. In the natural product, the methyl proton singlets appear at 6 1.08 and

1.17 while in ^ the signals are shifted upfield to 6 0.93 and 0.99.

The protons assigned to the vinylogous methyl group in 56 were found to be downfield (6 1.66) of where they were observed (6 1.60) by Bohlmann and Zdero. Another important distinction was apparent in the a lly lic region of the spectra. In the a lly lic protons were found between 6

2.27 and 2.69; they appeared as four evenly spaced (15-18 Hz) but broad envelopes (Figure 58). For the natural product, the three allylic protons and one homoallylic proton were designated to be in this area.

The line shapes reported here were very different (Figure 59) with three complex signals at 6 2.55, 2.48 and 2.23 of an intensity of 1:1:2. x= impurity

cr> i 1 X 1 cn 5.0 4.0 3.0 2.0 1.0

Figure 58. 200 MHz NMR spectrum of triquinane 56. x= impurity

5.13

1.60 1.17 1.08 0.84 o> 2.32 o>

Figure 59. 270 MHz NMR spectrum of naturally occuring senoxydene. 67

Table 4. Comparison o f the H NMR spectra o f 56 and senoxydene.

Synthetic 56' Senoxydene ,31

Assignment 6 Assignment 6

14-H 0.85 (d.J = 7Hz) 14-H 0.84 12,13-H 0.93(s), 0.99(s) 12-H 1.08 12-H 1.17 5,11,10-H I.64-1.5(m), 1.45(d(br)), lO'-H 1.20 1.36{dd), l,05(m)C 15-H 1.65 15-H 1.60 10-H 1.66 7,9-H 1.88 (dd), 1.75 (ddq) 7-H 1.78 5'-H 1.9 (m) 9-H 1.98 4,5-H 2.32 1,4-H 2.58 (d(br)). 2.52(d). I'-H 2.48 2.35 (d) 1-H 2.55 2-H 5.13 (s(br)) 2-H 5.13 !S?il

a. Spectrum recorded at 200 MHz in CDCl- solution. b. Spectrum recorded at 270 MHz in CDCl, solution. c. Impurities occur in this region. 68

Impurities were present in the spectrum of synthetic ^ as well as the spectrum of the natural product provided by Bohlmann and Zdero.

Even though ^ was purified as a single peak by vapor phase chromato graphy and the usual precautions were taken to remove hydrochloric acid from the chloroform-d used in the NMR sample, a large singlet (6 1.52) was apparent in the NMR spectrum of 5^. The sample of may have been decomposing because this signal grew in intensity over several days. In the spectrum of senoxydene, impurities were apparent in the regions between 6 1.60 and 1.20, 1.08 and 0.84.

In a l l , there was only one conclusion that could be drawnfrom our study. The structure of senoxydene, as proposed by Bohlmann and Zdero, had been incorrectly formulated. The similarities observed in the two spectra were quite remarkable, however, and this prompted us to speculate upon the actual structure of this molecule. While there were differences in the line shapes observed for the three allylic protons of senoxydene, th e ir chemical shifts were very sim ilar to those observed for the corresponding protons in 5]6. The protons of the vinylogous methyl group of 55 did appear downfield by 0.06 ppm of where they were expected for the natural product, but here the lin e shapes of the two signals were very similar. These consideration suggested that senoxydene may have a methylcyclopentene ring oriented in the same manner as Finally, the chanical shift of 6 0.84 for the secondary methyl doublet in the natural product spectrum compared very favorably with 6 0.85 for the signal in 56. This implied that possibly there was a secondary methyl substituent in the same relative position in the natural product. The problem now reduced to determining where the 69 geminal dimethyl substituent may be located. Assuming the isoprene rule 78 to be a useful guide , we could generate two possible triquinane structures for the natural product (Figure 60).

95 96

Figure 60. Two possible structures for senoxydene.

I t was now our proposal to undertake the syntheses of 9^ and ^ to determine i f our prognostication was correct.

3.0 Alternative Synthesis of 56.

A few months a fter the publication of our work, ItO reported a preparation of 55^® by a rather lengthy route beginning with an intermediate prepared in 10 steps in his e a rlie r synthesis of silphinene^l (Figure 61). The NMR spectrum of EG synthesized by this route corresponded in a ll respects with our data. ltd 's conclusion supported our contention that the natural product had been incorrectly formulated. 70

I.OOj.AcOH i.Chjl : I. O3 Z-NqBH-j 3.LI0H.THF 3. NqH, 3.H3O H^O 0,H v J C H 3 I V-J ^ MsCI H3CO ^

I. MsCI l.t-BuOK Z. ,Pt02. I , CH5L! 3. BFg'Cl^O

OCH3 56

Figure 61. ItG's total synthesis of 56.

Ite considered the possibility that the natural product may be a

triquinane with an epimeric secondary methyl substituent, i.e ., 97^

(Figure 62). Triquinane 97^ was prepared from an intermediate used in the synthesis of 56^ in six steps. However, the NMR spectra of 97^ and

the natural product were not id en tical.

1. BF) Et,20 I. L1AIH4. _ 2. MsCI f-P'TsOH - — '-4 O ' ^ 3. ÜAIH4 ....OH 2. œ-CPBA 3. BFj-EtiO

OCH3 97

Figure 62. Preparation of 97^. CHAPTER IV

Preparation of Two Angular Triquinane Isomers

of Alleged Senoxydene ( 56).

I .0 General Considerations.

The desire to confirm or disprove the proposed alternatives to the alleged structure of senoxydene (5^) offered the opportunity to explore the advantages and lim itations of our approach to the synthesis of angular triquinanes. Hydrocarbons 95 and 96 were intriguing targets with which to evaluate the compatibility of the new methylcyclopentenone annulation scheme with their varied substitution patterns. Both com pounds would present unique challenges for the regiocontrolled alkyla tion of their bicyclo[3.3.0]octanone progenitors with the allylic iodide

- 96 annulation reagent. Preparation of the required bicyclic octanones entails a further extension of the technologies adopted in our prepara tion of alleged senoxydene (^ ). From the standpoint of synthetic chemistry, further development of more expedient routes to substituted angular triquinanes would be useful in its own right.

71 72

TMS

95 98 99

TMS BfMg

100 88

Figure 63. Retrosynthetic dissection of 95.

2.0 Total Synthesis of Triquinane 95.

2.1 Retrosynthesis and Methodology

The sequential annulation strategy for the synthesis of alleged senoxydene ( 56) was fle x ib le enough to accommodate the development of a retrosynthetic plan for 95^ (Figure 63). As before, a 1,4 diketone, viz.

98, was to be the intermediate from which triquinane ^ would be derived.

This diketone was to be prepared from vinyl si ly l ketone 99_, a product of the alkylation of bicyclo[3.3.0]octanone 100 with allylic iodide §8^.

Octanone 100 was to originate from the four-carbon annulation of 80 5 ,5-dimethylcyclopent-2-enone It was anticipated that a route .54 exploiting the Ponaras Grignard reagent and a second scheme extending the thermal ene cyclization of r.Ç-ethylenic ketones should serve w e ll. 73

0 O BrMg n , H3 O+

101 102

p 0

KOlBu Hz/Pd-C

103 100

Figure 64. Ponaras Grignard route to 100.

2.2 Synthetic Work

2.2.1 Preparation of 100

The route employing the Ponaras Grignard reagent for the prepara tion of bicyclo[3.3.0]octanone 7^5 was extended to 100^ (Figure 64).

Copper(I) promoted conjugate addition to 5,5-dimethylcyclopent-2-enone at -78°C gave a quantitative yield of 101. Keto ketal 101 was converted to diketone 102 in 97% yield by heating at reflux in aqueous acetone with pyridinium tosylate. In accordance with earlier experience, 1,6 diketone

102 was cyclized with potassium ter^-butoxide in hot t e r t - butanol to give e,y-unsaturated ketone 103 as the product of thermodynamic control.

Catalytic hydrogenation of 103 resulted in the formation of octanone 100 due to the exclusive approach of hydrogen from the convex face of the enone. The assignment of the a-configuratlon to the secondary methyl 74 substituent in 100 was supported by the NMR spectrum. Conia has made the observation that signals assigned to the protons of sim ilar methyl substituents in c is -perhydropentalenones usually lie at 6 1.0 or 81 higher fie ld . In this case the corresponding doublet was found at 6

0.99, in excellent agreement with this generalization. In contrast, the signals for corresponding secondary methyl substituents having the

G-configuration were reported to be relatively deshielded by 0.10 ppm.

V ' A ^

104 100

Figure 65. Thermal ene cyclization route to 100.

The four-step sequence employing the Ponaras Grignard reagent gave a 58% overall yield of 100, but the attempt to extend the thermal ene cyclization to the synthesis of this compound was not successful (Figure

65). The requisite y,ç-ethylene ketone, 104, was prepared in 55% yield by the conjugate addition of 3-butenylmagnesium bromide to 5 ,5-dimethylcyclopent-2-enone at -30°C. All attempts to effect the ene cyclization of 104 using either sealed tubes or a flow oven resulted in recovery of the starting material or decomposition (Table 5).

Two rationales may be advanced to account for the reluctance of 104 to undergo ene cyclization. I t can be argued that steric crowding by the geminal dimethyl substituent in 104 prevents the o lefin ic appendage from achieving the conformation necessary fo r reaction (Figure 66).

A ltern atively, steric congestion in the transition state of ^ to TOO 75

Table 5. Attempted ene reaction o f 104.

Sealed tube experiments

Temperature (°C) Contact time (hours) Result

320" 1.3 recovered 104

370f 3.0 recovered 104

450^ 3.0 recovered 104

450^ 12.0 recovered 104

Flow oven experiments

Temperature (°C) Flow rate (mL/min) Result

380° 4 recovered 104

450° 6 recovered 104

610° 5 recovered 104

750° 6 decomposi ti on

may raise the activation energy of the 6-electron concerted ring closure above the point where the y,ç-ethylenic ketone thermally decomposes. In any event, Conia has reported the failure of the thermal ene reaction 82 for the similarly disubstituted 3,3-dimethylhept-6-en-2-one

V— W

104 100

Figure 66. Ground state and transition state arguements for the

fa ilu re of the thermal ene cyclization of 104 to 100. 76

TMS,

I. LDA ,- 3 0

TMS 2 .

100 99 88

1.MCPBA NoOElt

2 .H3 O + EtOH 105

1. Hz.Pj-C^ POCI3 2.L:AIH+

106 R= 0 95 107 R= H, OH

Figure 67. Preparation of triquinane 99^.

2.2.2 Preparation of 95.

The methylcyclopentenone annulation of 100 and the subsequent reduction and dehydration to 9^ was accomplished, but not without some interesting complications. The regiocontrolled alkylation of this cis-perhydropentalenone was greatly simplified because the position of the geminal dimethyl substituent eliminated the need to proceed through the thiobutylmethylene derivative (Figure 67). Deprotonation of 100^ with lithium diisopropyl ami de at -30°C and quenching the enolate with 88 gave a 51% yield of pure 99. The two-step transformation of 99 to 98 77 delivered 1,4 diketone in 94% yield. However, the aldol cyclization of

98 to enone 105 was troublesome (Table 6 ). Forcing conditions (runs 2,

5 and 6b) gave mostly polymeric material and l i t t l e 105, while mild treatment (runs 1, 3, 4, 6a and 7) led to recovery of diketone 98^. A useful yield of IjM (43%) was obtained by heating diketone 9^ at reflux with 20% sodium ethoxide in ethanol solution (run 8).

Two methods were used to reduce enone 105 with interesting results.

Lithium in liquid atmonia reduction of IjM gave two ketones in a 5:1 ratio and 38% combined y ie ld . Catalytic hydrogenation of 1 ^ produced a single ketone identical to the major component of the dissolving metal reduction in 76% y ie ld . Assuming that catalytic hydrogenation would give the ketone resulting from the approach by hydrogen to the less hindered face of 105, this product was formulated as KI6_ (Figure 67 and

68). Consequently, the minor constituent of the dissolving metal reduction must have been ketone 108 (Figure 68), differing from 106^ by the configuration of the methyl substituent adjacent to the carbonyl functionality. These assignments were confirmed by epimerizing pure 10^ to a mixture of 106 and 108 with sodium methoxide in methanol. 78

Table 6 . Aldol cyclization of 98 to 105.

Run Conditions Results

1_ t-BuOK, t-BuGH, THF recovered 98 45°, Ih

t-BuOK, t-BuOH, THF polymeric material 45°, 15 h

,83 20% aqueous NaOH recovered 98 reflux, 48 h

40% aqueous NaOH recovered 98 reflux, 48 h

t-BuOH, t-BuOK 10% yield of 105 re flu x , Ih polymeric material

t-BuOK, t-BuOH §1 R .T ., 16h then - no reaction (tic) 6b 40°, 48h - low yield of 105, recovered 102, and polymer

LiNtTMSjg, THF - no reaction (tic) R .T., 48 h

20% NaOEt in EtOH 43% yield of 105 re flu x , 8h 79

The subtle structural difference

between 106 and 108 led to remarkable

changes in the anisotropic effect of the

carbonyl group upon the four methyl

substituents. This was reflected by

drastic changes in the NMR spectra of

the two ketones (Figure 69). The doublet

assigned to the protons of the secondary

106 Ri = H, Rg = CH3 methyl substituent in the methylcyclopen-

108 Rj = CH3 , Rg = H tane ring of 106 exhibited a downfield

Figure 68. _ The epimeric shift of 0.13 ppm (6 1.01, = 6.7 Hz)

tricyclopentanoid ketones, re lative to the corresponding signal in the spectrum of 108 (6 0.88, ^ = 6.7 Hz). The signal for the protons of the secondary methyl substituent adjacent to the carbonyl function in

106 revealed a similar deshielding (6 1.20, ^ = 7 Hz) when compared with the same resonance in 108 (6 1.13, J_ = 7.5 Hz). Both shielding and deshielding effects were observed for the two singlets due to the protons of the geminal dimethyl substituent of 106 (6 0.81 and 1.09) relative to those of 108 (6 0.92 and 1.03).

The tric y c lic ketone was reduced and dehydrated without further incident (Figure 67). In practice, only ketone 106 from the catalytic hydrogenation was submitted to lithium aluminum hydride reduction. The product was the pair of epimeric alcohols 107 which were separable by chromatography on s ilic a gel. The epimer of 107 that was eluted fir s t was successfully dehydrated with phosphorous oxychloride in warm benzene 80

106 108

.0 2.0 1.0

Figure 69. 300 MHz NMR spectra of 106 and 81

77 1 to triquinane 9^ in good yield . A comparison of the H NMR spectrum O] of 9^ (Figure 70) with the data supplied by Bohlmann and Zdero for the natural product demonstrated that the two compounds were not identical

(Table 7 ).

Table 7. Partial comparison of the NMR spectra of 9^ and senoxydene'

Synthetic 95^ Senoxydene^

6 0.76 (d). 3H 6 0.84 (d), 3H

6 0.67 (s), 3H 6 1.08 (s), 3H

5 0.98 (s), 3H 6 1.17 (s), 3H

6 1.57 (s,(br)), 3H 6 1.60 (ddd), 3H

6 5.15 (d), IH 6 5.13 (s(br)), IH a. Spectrum was recorded at 300 MHz in CDCl^ solution. b. Spectrum was recorded at 270 MHz in CDClg solution. 00 À. ro 5.0 4.0 3.0 2.0 1.0 0.0

Figure 70. 300 MHz NMR of triquinane 95. 83

° TMS

96 109 (W3 26

110 111

Figure 71. Retrosynthetic dissection of angular triquinane 9^.

3.0 Approach to the Total Synthesis of Triquinane 96

3.1 Retrosynthesis and Methodology

The plan for the synthesis of tricyclo[6.3.0.0^’^]undecene 96 v/as based upon the extention of our methylcyclopentenone annulation procedure to c is-perhydropentalenone 109 (Figure 71). Due to the methyl substitution pattern of 109, the straightforward conjugate addition strategy employed in the preparations of bicyclic ketones 7^ and 100 would be inadequate for this ketone. Two routes were devised that would enable us to construct the trimethylcyclopentane substitution pattern in

109. The most direct plan was to hydrogenate enone 26, prepared by both

Paquette^^ and Itdf^ in th eir syntheses of silphinene. The second 84 proposal required the thermal ene cyclization of e,ç-ethylenic ketone llpGS, This would be the f ir s t example of a thermal ene cyclization to a c is -perhydropentalenone with a geminal dimethyl substituent on the olefin side chain of the starting material. Ketone 110 would be obtained by the aldol cyclization of 1,4 diketone 111 and the subsequent dissolving metal reduction of the resulting dienone.

m Hg/Pd-C

26 109

Figure 72. Hydrogenation of ^ to 109.

3.2 Synthetic Work

3.2.1 Preparation of 109

Bicyclic enone ^ was prepared according to the procedure of

Paquette and Leone-Bay^^. Catalytic hydrogenation of 2^ gave 10£ as a single product in 92% yield (Figure 72). The chemical shift of the doublet assigned to the protons of the secondary methyl substituent in the NMR spectrum of 109 was 5 0.88, a value which confirmed the a- 81 configuration of this methyl group . 85

PDC O

112 R= H. OH 0 113 R= 0 0

111

Figure 73. Ponaras Grignard route to 1,4 diketone 111.

Two routes to diketone 111 were devised. The Ponaras Grignard reagent was added to 2,2-dimethylpent-4-enal®^ to give a 78% crude yield of hydroxy ketal 112 (Figure 73). Oxidation of n2 with pyridinium dichromate delivered keto ketal 113®^. Deblocking of 113 with pyridinium tosylate in 19:1 acetone:water (v:v) gave 111. The quality of the 1,4 diketone that was obtained was poor and a tedious chromatography on silica gel was required to purify this material for use in the next step.

Consequently, this preparation of 111 was abandoned. 86

Q 2. N o \H g n

113

-r TT - 2.LI/NH, 0

111 110 109

Figure 74. Phenylsulfone route to 109.

A more satisfactory route to 111 made use of the coupling of lithiated 2-(2-phenylsulfonylethyl)-2-methyl-l,3-dioxolane®^, developed by Kondo and Tunemoto®^, with 2 ,2-dimethylpent-4-enoyl chloride^®

(Figure 74). Reductive desulfonylation of the coupled product with 6% sodium.-mercury amalgam gave 113. A fter aqueous acid hydrolysis, diketone 111 was obtained in 41% overall yield for the three-step sequence. Diketone Hl_ was cyclized with aqueous base and the resulting dienone reduced to e,;-ethylenic ketone 110 with lithium in liquid ammonia. The ene reaction of lljO gave cls-perhydropentalenone 109 in

67% y ie ld .

I.N0OCH3 ^ o SBu CgHgOCHO

2. nBuSH H+ 109 114

Figure 75. Preparation of thiobutylmethylene ketone 114. 87

Relying on our previous experience with c is-perhydropentalenone 75, we anticipated the need to prepare the thiobutylmethylene derivative of

109 to control the regiochemistry of enolate formation in the forth coming alkylation step (Figure 75). An o-formyl ketone was prepared from 109 in 90% yield with a mixture of sodium methoxide and ethyl formate in benzene. Submission of this compound to the action of n^butanethiol and £-toluenesulfonic acid in benzene at the reflux 72 73 temperature gave thiobutylmethylene ketone 114 in 70% y ield ’ .

TMS,

,0 SBu

= R TMS

88 115 R= CHSnBu 114 3-hq^ o'^ qh . XOH 126 R= Hr

1.MCPBA KOIBu 2.H3O + THF

117 118

Figure 76. Progress towards the preparation of 9]5, 88

3.2.2 Progress towards ^

Time consuming d iffic u ltie s encountered in the alkylation of thio butylmethylene ketone frustrated efforts to complete the preparation of triquinane 96_in time for inclusion in this dissertation (Figure 76).

I t was necessary to screen a variety of alkylation conditions to devise a procedure that delivered 115 in useful yields (Table 8). In runs 1, 2, 4,

5, and 5, was added to solutions of strong base at temperatures above

-78°C. After quenching with an excess of allylic iodide 88, low yields

(0 to 21%) of the alkylation product 115 were obtained. Even more troubling was the observation that unreacted 114 was never recovered from these tr ia ls ; apparently the ketone was being decomposed under the reaction conditions. In run 3, a small quantity of unreacted 114 (12% of the total amount) and product 1_15 (13% yield ) were isolated when thiobutylmethylene ketone lj4 was added to a lithium diisopropyl ami de solution and quenched with 88 at -78°C. These results were interpreted to suggest that low temperature (ca. -78°C and below) and a more s te rica lly encumbered base may prevent the destruction of the thiobutylmethylene ketone. Consequently, the enolate of 114 was generated with lithium hexamethyldisilazide at -78°C and reacted with 2 equivalents of 88 at

-78°C (run 7). 89

Table 8. Alkylation of 119

Run Conditions for Conditions for Yield of 114 recovered deprotonation alkylation 115 (%) of 114* with 88

1 LDA,-30"C,THF 1.1 equiv. of 188 15 -30°C to R.T.,16h

2 LDA,-20°C,THF 1.1 equiv. of 88 7 -20°C to R.T.,3h

3 LDA,-78°C,THF 1.1 equiv. of 88 13(15)* 12 -78°C to R.T.,16h

4 PhgCLi,-30°C,THF 1.5 equiv.of 88 0 -30°C to -40°C,3h

5 LDA,-78°C,THF 2 equiv. of 88 19 -20°C to -40°C,3.5h

6 LiN(TMS)2,-78°C, 2 equiv. of ^ 21 THF -30°C to -40°C,4h

7 LiN(TMS)2,-78°C, 2 equiv. of ^ 23(61) 60 THF -78°C, 24h

8®^ 1) KN(TMS)2,-78°C, 1.5 equiv. of 88^ low most of THF -30°C, 2h product 2) EtgB

9®® KOtBu, t-BuOH 6 equiv. of ^ low most of R .T., Ih product a. Ketone 114 added over 1 h except in runs 8 and 9 where 114 was added in one portion. b. Numbers in brackets are yields corrected for recovered 114. 90

From this mixture a 23% yield of 115 (61% when corrected for recovered

114) and 60% of unreacted 114 were isolated. While this was not an ideal solution to the problem of alkylating 114, the recovery of reasonable amounts of unreacted starting ketone unharmed was an important improvement. The unconsumed 11^ was routinely recycled in later alkylations.

Product 11^ was converted to vinyl s ily l ketone 116 under surprisingly 17 mild conditions (Figure 76). By heating 115 in a mixture of 10% aqueous potassium hydroxide and diethylene glycol (3:4, v:v) for 18 hours, a 77% yield of 116 was obtained. This compared favorably with the corresponding deprotection of 91 to 87^ in the synthesis of alleged senoxydene (see

Figure 56). The earlier reaction required more vigorous conditions (25% aqueous potassium hydroxide and diethylene glycol [3 :4 ,v ;v ], heated at reflux fo r 48 hours), gave a low yield of 87_ (45%), and was plagued by formation of an unidentified side product.

The difference in the behavior of 91^ and 115 towards hydrolysis may be accounted for by considering the susceptibility of each of the thiobutylmethylene ketones to the 1,4 addition of hydroxide ion (Figure

77). The close proximity of the geminal dimethyl substituent to the reacting center in 9^ evidently exerted a retarding effect upon hydroxide ion addition. Of course, this situation did not pertain to

115; therefore fa c ile base hydrolysis was observed. This reasoning can be extended to explain the d iffic u ltie s encountered in the alkylation of

114. With unhindered access to the thiobutylmethylene functionality of

114, at temperatures above -78°C the strong base used acted as a nucleo phile, adding readily to this material and in itia tin g decomposition 91

Sq-Bi

91 114 R = H

116 R = 2-(trimethyl silyl)-2-butenyl

Figure 77. Susceptibility of thiobutylmethylene ketones towards

nucleophilic attack.

rather than causing deprotonation. A sim ilar proclivity for 1,4 addition of hindered base to a thiobutylmethylene ketone was noted by OQ Marshall in his studies on the synthesis of (+)valeranone . He reported the isolation of a product resulting from the addition of lithium triphenylmethide to a thiobutylmethylene substituted decalone.

The elaboration of 11^ to the la s t compound prepared in this sequence, tric y c lic enone .118, proceeded without incident (Figure 76).

The vinyl si lane appendage of was converted in two steps to the

2-butanone side chain of 1,4 diketone 117 in 30% overall yield®^.

Diketone 117, was transformed to 118 in 79% yield with potassium te rt- butoxide and t e r t - butyl alcohol in hot tetrahydrofuran. All compounds up to this point have been fully characterized; to complete this synthetic route to triquinane 96 only the reduction and dehydration steps remain to be accomplished. CHAPTER V

Conclusion: A Review of the Evidence

Bohlmann and Zdero unequivocally established that senoxydene is an o lefin ic hydrocarbon with the molecular formula 6^ 5^24 by their observation of the molecular ion in the mass spectrum of the natural product . The other evidence upon which structure 56_was proposed consisted of the 270 MHz ^H NMR of the natural product in chloroform-^ and benzene-dg, suitable decoupling of these spectra, a s h ift reagent study of an epoxide derivative, and the infrared spectrum.

Unfortunately, as we have demonstrated, th e ir interpretation of these data was fa u lty.

A review of the ^H NMR chemical shifts of protons associated with the te rtia ry , secondary, and vinyl methyl groups of fifte e n triquinane hydrocarbons is compiled in Table 9. Some of this information is presented as a composite of the ranges observed for chemical shifts of te rtia ry and secondary methyl protons at various positions of angular triquinanes (Figure 78). Comparison of these facts with the data supplied by Bohlmann and Zdero (see Table 9, firs t entry) leads to the following conclusions:

92 93

1) the presence of a methylcyclopentene group in this natural product seems likely. The value of the chemical shift reported by

Bohlmann and Zdero for the vinyl proton (6 5.13) and the protons of the vinylogous methyl (6 1.60) lies well within the normal ranges for this part structure.

2) a slight shielding effect is experienced by methyl protons found in the cavity of a ci s -perhydropentalenene portion of an angular triquinane relative to its epimer (e.g., alleged senoxydene versus e p i-alleged senoxydene, pentalenene versus e p i-pentalenene and 7aH- versus 7gH-silphiperfol-5-ene). This effect is small compared to the range of chemical shifts found for the secondary methyl groups of different triquinane skeletons (Figure 78), therefore this observation is not useful for the a priori assignment of the configuration of a given secondary methyl substituent.

3) the composite of chemical shifts for protons of te rtia ry methyl groups illustrates (Figure 78) that the signals reported by

te rtia ry methyl substituents secondary methyl substituents

4 :6 0.94-0.96 iiiH

6 0.97 SO. 93-0.99 gem dimethyl gem dimethyl

« 0 .9 3 -1 .0 8 angular methyl Figure 78. Range of chemical shifts for te rtia ry and secondary methyl groups in angular triquinanes. The a and g designation refers to the relative configuration of the secondary methyl groups. 94

Bohlmann and Zdero as geminal dimethyl substituents are very atypical.

The chemical s h ift of the singlet at 6 1.08 falls into the range where

"angular" methyl substituents are found. The second "geminal" methyl signal has a shift of 6 1,17; this resonance is further downfield than any that has been observed for a te rtia ry methyl of an angular triquinane.

4 ) i t is probable that senoxydene is not an angular triquinane.

The closest agreement between the chemical shifts of the methyl singlets in the NMR of the natural product (6 1.08 and 1.17) and an existing triquinane is with two signals found in the linear triquinane capnellene (6 1.08 and 1.16). However, these singlets were never assigned to specific te rtia ry methyl groups in this compound, so i t is difficult to suggest what component of this structure may be present in senoxydene. 95

Table 9. Comparison o f chemical s h ifts o f methyl groups on known triquinanes and senoxydene.

Assignments A B ------C D Ref.

naturally occurring , , senoxydene 1.08(s), 1.17(s), 0.84(d), 1.60(ddd) 31

Angular Triquinanes a l1eged senoxydene (56)

0.93 (s), 0.99 (s) 0.85 (d) 1.65 (s[br])

J = 6.7 Hz b “ ^ CDCI3 , 300 MHz

epi-alleged senoxydene (97)

0.92 (s). 0.96 (s) 0.90 (d) 1.66(s[br]) 79

J = 7 Hz

CDCI3 , 270 MHz

0.98 (s) 0.67 (s)® 0.76 (d) 1.57 (s[b r])

J = 6.6 Hz

CDCI3 , 300 MHz 95

Table 9 (continued)

Ref. pentalenene

0.97 ( 6H,s) 0.88 (d) 1.60 (s[br]) 25

J = 7 Hz

epi-pentalenene

0.98 (6 H,s) 0.93 (d) 1.64 (s[br]) 28

0 = 7 Hz */iHM C b d

isocomene

1.03 (s) 0.86 (d) 1.03 (s) 1.56 (d)

J_= 7 Hz J = 1.5 Hz

CDCI3 . 270 MHz

si 1 phinene 1.08 (s) 0.93 (s) 0.98 (s) 0.82 (d) 18 J = 7 Hz d // CDCI3 , 270 MHz a 97

Table 9 (continued)

Ref.

silphiperfol-6-ene

l.OO(s) 1.52 (s[br]) 1.55 (s[br]) 0.96 (d) 18

CDCI3 . 270 MHz

7aH-si1 phi perfol-5-ene

0.97 (s) 1.67 (dd) 0.98 (d) 0.94 (d) 18

d CDCI3 , 270 MHz

7BH-si1 phiperfol-5-ene

d 0.93 (s) 1.56 (dd) 0,00 (d) 0.96 (d) 18

CDCI3 , 270 MHz 98

Table £ (continued)

A Ref.

Linear Triquinanes.

Hirsutene

0.92(6H, s), 1.05 (3H. s) 4.72

CDCl^, 100 MHz

capnellene

0.98 (s) 1.08 (s) 1.16 (s) 4.82,4.94(m) 90 a b CDClg, 100 MHz

capnellane

b d 0.91 (s) 1.00 (s) 1.19 (s) 0.93 (d) 90 ± = 5.4 Hz

CDCI3 , 100 MHz 99

Table 9. (continued)

A B C D Ref.

^(8i9)_cgpneTiane

0.94(s),1.03(s),l.l(s) 1.67 (s[br]) 91

Propen ane modhephene

0.97(6H, s) 0.99 (d) 1.58 (d)

J = 5.5 Hz J = 1.5 Hz

CCl^, 60 MHz

a: This high fie ld s h ift arises from the methyl group that lies within the shielding region of the o le fin , therefore, i t was not considered a representative chemical shift and omitted from Figure 78. EXPERIMENTAL

Melting points were determined in open capillaries with a Thomas-

Hoover apparatus and are uncorrected. Infrared spectra were recorded on a Perkin-Elmer Model 467 instrument. Proton magnetic resonance spectra were obtained with Varian EM-360, Varian EM-390, Bruker WP-200, and

Bruker WM-300 spectrometers. Mass spectra were determined on AEI MS-9

and Kratos MS-30 spectrometers at an ionization potential of 70eV.

Elemental analyses were performed by Scandinavian Microanalytical

Laboratory, Herlev, Denmark.

100 101

2-Carbomethoxy-4.4-Dimethylcyclohexanone (61^).

Dimethyl carbonate (17 mL) in tetrahydrofuran

(50 ml) was added dropwise to sodium hydride C 00C H 3 (10.2 g of a 50% oil suspension prewashed with CH3 "tv'cHj toluene and tetrahydrofuran) and the mixture

was heated to reflux. A portion (2 mL) of a

solution of 4 ,4-dimethylcyclohexanone (9.0 g ,

71 mmol) in tetrahydrofuran (20 mL) was added slowly. A washed slurry of potassium hydride (0.6 g) in tetrahydrofuran (9 mL) was added. The remainder of the ketone solution was added with continued heating during

45 min. The reaction mixture was heated for another 1.5 h and cooled to

10°C. Acetic acid (75 mL) and brine (100 mL) were added, followed by ether (250 mL) and solid sodium bicarbonate. The ether phase was washed with brine, dried, and evaporated. Distillation of the residue at

55-61°C and 0.3 to rr gave 11.67 g (89%) of 61,; IR (CCl^, cm"^): 2900

(b r), 1740 (s ), 1710 (s ), 1650 (s ), 1610 (s ), 1430 (m), 1350 (s ), 1275

(s ), 1220 (m), 800 (s ); NMR (CDCI 3 , 90 MHz) 6 3.72 (s, 3H), 2.27 (t,

J = 7 Hz, 2H), 2.02 (br s, 2H), 1.40 (t, J, = 7Hz, 3H), 0.97 (s, 6H), m/z calcd (M*) 184.1099, obs 184.1106.

Anal. Calcd fo r C^gHjgO^: C, 65.19; H, 8.75. Found: C, 65.27; H, 8.72, 102

Methyl (± )- 8 ,8-Dimethyl-1,4-dioxaspiro[4.5]decane-6-carboxylate.

A stirred solution of ^ (106 g, 0.58 mol),

r~ \ ethylene glycol (65 mL), and p-toluenesulfonic o o "X^co^cH) acid (0.5 g) in benzene (195 mL) was heated at

reflux with azeotropic removal of water for 72 h.

The product was partitioned between ether and

water. The aqueous phase was extracted with ether and the combined organic layers were washed with sodium bicarbonate solution and brine, dried and evaporated. The residue was distilled to give 117.6 g (89.5%) of the k etal, bp 100-105°C at 0.35 to rr. The colorless liquid solidified upon distillation, mp 39-44°C; IR (CCl^, cm‘^): 2980, 2900, 1750, 1550, 1440, 1370, 1350, 1200, 1070, 770; NMR

(CDCI3 , 90MHz) 6 3.9 (m, 4H), 3.65 (s, 3H), 2.85 (dd, J^= 10 and 4 Hz,

IH), 2.1 - 1.2 (series of m, 6H ), 1.0 (s, 6H) ; m/^ calcd (M*) 228.1361, obs 228.1358. 103

( ± ) - g ,ct , 8 ,8-Tetramethyl-1 ,4-dioxaspiro[4. 5]decane-6-methanol (62).

To an ether solution containing méthylmagnésium

CHj iodide (0.51 mol) was added dropwise a solution

of the ketal ester (43.0 g, 0.19 mol) in ether 0“'T I (120 m l). After completion of the addition, CH3 the reaction mixture was heated at reflux for 2

h, cooled in ic e, and treated with ammonium chloride solution. Additional water was added to dissolve the inorganic salts, the layers were separated, and the aqueous phase was extracted with ether. The combined organic layers were washed with brine, dried, and evaporated. There was obtained 40.55 g (94%) of6 ^ as a colorless solid, mp 65-68 °C (from petroleum ether); IR (CCl^, cm"^): 3510, 2950,

1450, 1390, 1350, 1190, 1170, 1110, 1050, 940; NMR (CDCI 3 , 90 MHz) 6

4.66 (s, IH), 3.96 (m, 4H), 2.21 - 1.35 (series of m, 7H), 1.2 (s, 3H),

1.15 (s , 3H), 0.98 (s , 3H), 0.92 (s, 3H); m/z calcd (M*) 228.1725, obs

228.1732.

Anal. Calcd for CigHg^Og: C, 68.38; H, 10.59. Found: C, 68.24; H,

10.59. 104

2-Isopropy1i dene-4,4-dimethylcyclohexanone (60).

A solution of 6^ (26.96 g, 0.12 mol), concen-

trated hydrochloric acid (2.2 mL), water (100

mL), and methanol (166 mL) was heated at reflux CH3 with stirring for 2 h, cooled, and neutralized

with solid sodium bicarbonate. The major portion

of the methanol was evaporated under reduced pressure, water was added, and the mixture was extracted with ether.

The aqueous layer was further extracted with ether and the combined organic layers were washed with brine, dried, and evaporated. D istilla tion of the residue afforded 16.35 g (83%) of 68 as a colorless oil, bp

47-49°C (0.75 torr); IR (CCl^, cm”^): 2940, 2840, 1685, 1615, 1450,

1360, 1275, 1125; NMR (CDCI 3 , 90 MHz) 6 2.4 (t, J = 6 Hz, 2H), 2.30

(s, 2H), 1.95 (s, 3H), 1.75 (s, 3H), 1.65 (t, J = 6 Hz, 2H), 1.1 (s,

6H); m/z calcd (M+) 166.1358, obs 166.1362.

Anal. Calcd for C^iH^gO; C, 79.46; H, 10.91. Found: C, 79.21; H,

10.97. 105

Methyl (±)-2-Isopropy1i dene-4,4-di methylcyclopentanecarboxylate (59).

A mixture of &0 (5.0 g, 30 mmol) and sodium

° bicarbonate (1 g) in ether (30 ml) was CHjOC^ _ "CH3 blanketed with nitrogen and cooled to -10 C.

CHj During 30 min, bromine (5 g, 30 mmol) was added

and the reaction mixture was stirred for an

additional hour and filtered. The filtrate was added to a solution of sodium methoxide in methanol (prepared from 2.5 g of sodium and 70 mL of methanol) which was stirred for 12 h and poured into a mixture of dilute hydrochloric acid and ice. The product was extracted into ether. The combined organic layers were washed with saturated sodium bicarbonate and brine solutions, dried, and evaporated.

The residual oil (5.07 g), consisting of a mixture of 5^ and 6£ , was

separated into its components by high pressure liquid chromatography on

a Waters Prep 500 instrument (silica gel, elution with 2% ethyl acetate

in petroleum ether). There was isolated 1.36 g (27%) of 6jO and 2.13 g

(50% based on recovered starting material) of 5^ as a colorless o il; IR

(CCI4 , cm-1): 2940, 2890, 1735, 1460, 1430; NMR (CDCI 3 , 90 MHz) 5

3.65 (s, 3H), 2.22 (s, 2H), 2.05 - 1.6 (m, 3H), 1.64 (s, 3H), 1.55 (s,

3H), 1.10 (s, 3H), 0.86 (s, 3H); m/^ calcd (M+) 196.1463, obs 196.1458.

Anal. Calcd for ^i2^2CPz' 10.24. Found; C, 72.88;

H, 10.12. 106

(±)-cis-Hexahydro-3,3,5,5-tetramethy1-lH-cyc1openta[c]furan-l-one (63)

A solution containing ^ (2.13 g, 11 mol), con

centrated hydrochloric acid (8.5 mL), methanol

(25 mL) and water (4 mL) was heated at the

reflux temperature for 8 h. After cooling,

most of the methanol was removed under reduced

pressure and the residue was partitioned between ether and water. The aqueous phase was extracted with ether and the combined organic layers were washed with saturated sodium bicarbonate and brine solutions prior to drying and solvent evaporation. There was obtained 1.62 g (82%) of M as a brown o il which was purified by s ilic a gel high pressure liquid chromatography (elution with 6% ethyl acetate in petroleum ether). The pure lactone was a colorless crystalline solid, mp

42-43°C; IR (CCl^, cm"^): 2950, 2860, 1770, 1450, 1190, 760; NMR

(CDCI3 , 90 MHz) 6 3.25 (m, IH), 2.75 (m, IH), 1.85 (m, 2H), 1.6 (s, 2H),

1.5 (s, 3H), 1.4 (s, 3H), 1.15 (s, 3H), 1.0 (s, 3H); m/z calcd (M+-CH 3 )

167.1072, obs 167.1066.

Anal. Calcd for CjiH^8°2' 72-49; H, 9.95. Found: C, 72.32; H,

9.95. 107

(±)-cis-6a-Bromohexahydro-3,3,5,5-tetramethy1-lH-cyc1openta[c]furan-l- one (6£ ).

To a cold (-78°C) solution of the lithium eno-

^ la te of 63^ (300 mg, prepared by addition of 63

0^ j " ' ^ c h 3 to 1.1 equivalents of lithium di isopropyl ami de

in tetrahydrofuran [20 mL] at -20°C) was added

1 ,2-dibromoethane (3 mL). The reaction mixture

was stirred at room temperature for 2 h, evaporated under reduced pressure, and treated with water. The aqueous mixture was extracted with ether. The usual workup gave an orange oil

(550 mg) which was purified by preparative tic chromatography on s ilic a gel plates (elutionwith 20:1 petroleum ether:ethyl acetate). There was obtained 260 mg (61%) of 64_ as an off-white solid, mp 71-75°C; IR (CCl^, cm 'l): 2960, 2860, 1760, 1460, 1385, 1370, 1270; NMR (CDCI 3 , 90 MHz)

6 3.15 (dd, J = 12 and 6 Hz, IH), 2.35 (m, 2H), 1.7 (s, 3H), 1.55 (s,

2H), 1.35 (s, 3H), 1.20 (s, 3H), 1,05 (s, 3H); m/z calcd (M+) 245.0178, obs 245.0186. 108

(±)-3,3a.4,5-Tetrahydro-3,3,5,5-tetraniethy1-lH-cyc1openta[c]furan-l~one

(58).

Bromide 6£ (44 mg, 0.17 mol) was added to a

mixture of dry lithium bromide (30 mg) and

lithium carbonate (40 mg) in N,N-dimethyl forma-

mi de (2 mL) and heated to 158°C for 30 h. The

reaction mixture was poured into water and

extracted with petroleum ether. The organic phase was washed with water, dried, and evaporated to give an o il.

Further purification by preparative tic (elution with petroleum etherrethyl acetate, 20:1) gave recovered &4 (10.1 mg, 23%) and ^ (6.5 mg, 21%); IR (CCl^, cm"^): 2920 (s ), 2860 (s ), 1770 (s ), 1670 (w), 1190

(s ), 780 (s ); NMR (CDCI 3 , 200 MHz) 6 6.35 (d, ^ = 3 Hz, IH), 3.5 (m,

IH), 1.49 (s, 2H), 1.26 (s, 3H). 1.24 (s, 3H), 1.23 (s, 3H). 1.19 (s,

3H).

Repeated attempts to reproduce this experiment fa ile d . 109 l-.)i f'’ s..Hexahydro-3,3,5.5-tetramethy1-6a-(pheny1se1eny1 j-lH-cyclopenta-

CcJfuran-l-one.

A solution of 6^ (200 mg, 1.10 mmol) in dry

tetrahydrofuran (4 mL) was added to a cold PhSe o (-78°C) lithium diisopropyl ami de solution [pre 'I pared from diisopropyl ami ne (0.32 mL, 2.6 mmol)

and jr-butyllithium (0.86 mL of 1.55 M in

hexane, 1.30 m o l) by stirrin g for 10 min at

-20°C] and stirred for 30 min. Phenylselenyl bromide (300 mg, 1.10 m ol) dissolved in tetrahydrofuran (6 mL) was added and the mixture was stirred at -78°C for 2 h before being poured into ice-cold dilute hydrochloric acid. The aqueous phase was extracted with ether and the combined organic layers were washed with saturated sodium bicarbonate and brine solutions prior to drying. Solvent evaporation left an oil (220 mg) from which the selenide was obtained by preparative tic on silica gel plates. In addition to the phenylselenyl lactone (90 mg, 25%), there was recovered 25 mg of 6^; IR (CCl^, cm"^): 3300, 2960,

2280, 1730, 1580, 1390, 1260, 1160, 1140, 1020, 950, 900, 680; NMR

(CDCI3 , 90 MHz) 6 7.75 - 7.15 (m, 5H), 2.75 (t, J = 8 Hz, IH ), 2.0 (s,

2H), 1.64 (s, 3H), 1.62 - 1.5 (m, 2H), 1.32 (s, 3H), 1.05 (s, 3H). 0.98

(s, 3H); m/^ calcd (M*) 338.0785, obs 338.0777. 110

3,4,5,6-Tetrahydro-3,3,5,5-tetrainethy1-lH-cyc1openta[c]furan-l-one (65)

To a solution of the phenylselenyl lactone

(34.1 mg, 0.10 m ol) in carbon tetrachloride

(15 mL) was added a 4:1:4 mixture of water:

acetic acid:30% hydrogen peroxide (v:v:v, 4.5

ml). The biphasic mixture was heated at reflux

for 7.5 h, poured into saturated sodium bicar bonate solution, and admixed with dichloromethane. The layers were separated and the organic phase was washed with water, dilute hydro chloric acid, water, and brine. Following drying and solvent evapora tion, there was obtained 10.5 mg (60%) of 65 as a colorless crystalline solid, mp 65-69°C; IR (CCl^, cm"^): 2960, 1770, 1330, 1290; NMR

(CDCI3 , 90 MHz) 6 2.32 (s, 4H), 1.46 (s, 6H), 1.25 (s, 6H ); m/z calcd

(M+) 180.1150, obs 180.1157. Ill

(±)-cis-4 ,5 .6 ,6 a -T e tra hyd ro -3 ,5 .5 -trim e th y1 -l(3 a H )“ Penta1enone ( 6 6 )

Into a solution of phosphorus pentoxide (4.5 g)

In tnethanesulfonic acid (57 g ), prepared by

CHj heating with stirring at 80°C under nitrogen, was

CH3 added lactone 6^ (400 mg, 2.2 mmol) in small portions. The dark reaction mixture was heated

at 50°C for 48 h and added dropwise to water.

The resulting aqueous suspension was extracted wth dichloromethane (6 x) and the combined organic layers were washed successively with sodium bicarbonate solution, water, and brine. Drying and solvent evaporation le f t a brown o il (298 mg) which was filte re d through a s ilic a gel plug

(6% ethyl acetate in petroleum ether) to give 238 mg (70%) of 6^; IR

(CCI4 , cm 'l): 2950, 1700, 1620; NMR (CDCI 3 , 90 MHz) 6 5.5 (s, IH),

3.4 - 2.8 (m, 2H), 2.02 (s, 3H), 2.0 - 1.1 (m, 4H), 1.03 (s, 6H) ; m/z calcd (M+) 164.1201, obs 164.1199. 112

2- ( 2-Bromoethy 1 ) -2-methyl- 1 ,3-di oxol ane (.92).

Methyl vinyl ketone (144.0 g, 2 mol) was dis

solved in ethylene glycol (500 g, 8 mol) and

Br^s3<° cooled to 10°C. Gaseous hydrogen bromide (182.0

g, 2 mol) was introduced while the temperature

of the solution was maintained below 40°C. This

mixture was extracted with petroleum ether ( 6x ).

The combined petroleum ether layers were washed with brine, dried and evaporated. The residue was distilled (50-52°C, 0.2 torr) to give the bromoketal (100 g, 26%); NMR (CDCI3 , 60 MHz) 6 4.0 (s, 4H), 3.4 (t,

J = 8 Hz, 2H), 2.2 (t, J = 8 Hz, 2H), 1.3 (s, 3H). 113

(±)-5,6.7.7a-Tetrahydro-3,7.7-trimethy1inden-4(2H)-one (69).

A solution of 2-(2-bromoethyl)-2-methyl-1,3-

di oxol ane (33.8 g, 180 m o l) and 1,2- dibromo-

ethane (5 mL) in tetrahydrofuran (15 mL) was

added over 6 h to magnesium shavings (13.0 g,

540 mg-at) in tetrahydrofuran (250 mL). In a

separate flask, copper iodide (4.40 g, 23 mmol) in dimethyl sulfide (20 mL) was cooled to -78°C and the Grignard solu tion was transferred by canula to this mixture. 4,4-Dimethylcyclo- hexenone (7.50 g , 60.0 mmol) in tetrahydrofuran (12 mL) was added over 8 h at -78°C. The reaction mixture was maintained at -78°C for 4 h and at ambient temperature for 4 h. At this point, amonium chloride solution was added, the mixture was filtered, and the filtra te was extracted with ether. The ether layer was washed with sodium bicarbonate solution and brine, dried, and evaporated. The residue was dissolved in a mixture of tetrahydrofuran (600 mL), water (60 mL), and concentrated hydrochloric acid (12 mL), then stirred for 40 h. This mixture was neutralized with solid potassium carbonate, evaporated, and extracted with ether. The ether solution was washed with sodium bicarbonate solution and brine, dried, and evaporated. Purification by hplc (elution with petroleum ether:ethyl acetate, 30:1) gave 69^ (6.33 g, 60%); IR (neat, cm"^): 2940

(s ), 1678 (s ), 1618 (m); NMR (CDCI 3 , 90 MHz) 5 2.8 (m, IH ), 2.5 - 2.2

(m, 4H), 2.1 (m, 3H), 2.0 - 1.3 (m, 4H), 0.97 (s, 3H), 0.82 (s, 3H); m/z calcd (M+) 178.1358, obs 178.1362. 114

(±)-(3R*,3aR*,7aS*)-Tetrahydro-3,7.7-tr1methy1-4(3aH)-indanone ( 6^ ) .

Ketone 6£ (2.69 g, 15.1 m ol) and 10% palladium

on carbon (300 mg) in acetic acid (50 mL) was

shaken under an atmosphere of hydrogen (50 psi)

for 12 h. The solution was filte re d and added

to a mixture of ether and water. The biphasic

mixture was neutralized with solid potassium carbonate. The aqueous layer was extracted with ether and the ether solution was washed with sodium bicarbonate solution and brine prior to drying and solvent evaporation. There was obtained 2.61 g (96%) of 68^;

IR (neat, cm"^): 2940 (s ), 2865 (s ), 1710 (s ), 1460 (m). 1365 (m), 900

(s ), 710 (m); NMR (CDCI 3 , 90 MHz) 5 2.68 (t, J = 7.5 Hz, IH), 2.55 -

1.4 (m, lOH), 1.24 (d, J. = 6Hz, 3H), 1.25 (s , 3H), 0.93 (s , 3H); NMR

(CDCI3 ) ppm 215.13, 54.50, 53.83, 38.72,37.93, 34.78, 31.13, 29.74,

28.64, 27.79, 25.85, 16.32; m/z calcd (M*) 180.1514, obs 180.1519. 115

(±)-(3R*.3aS*,7aS*)-Tetrahydro-3,7.7-trimethy1-4(3aH)-indanone (70)

Ketone 68^ (4.83 g, 27 mmol) was stirred with

sodium methoxide in methanol (from 1.0 g of Na

in 100 mL of methanol) for 8 h. The mixture

was reduced to half its volume, poured into

anmonium chloride solution and extracted with

ether. The ether solution was washed with

sodium bicarbonate solution and brine, dried, and evaporated to give

4.19 9 (85%) of 70; IR (neat, cm"^): 2940 (s ), 2060 (m), 1710 (s );

NMR (CDCI3 , 90 MHz) 6 2.5 - 1.4 (m, IlH ), 1.02 (s, 3H), 0.92 (d, J =

4Hz, 3H), 0.90 (s, 3H); m/z calcd (M+) 180.1514, obs 180.1519.

Anal. Calcd for 0^2^20°' 79-94; H, 11.18. Found; C, 79.98; H,

11.24. 116

Trimethylsilyl enol ether 71.

Ketone 70 (1.0 g, 5.56 trnioT) in tetrahydrofuran

(10 mL) was added to a cold (-78°C) solution of

lithium diisopropyl ami de (7.23 mmol) in tetra

hydrofuran (30 mL) over 1 h. This solution was

treated with a portion (5.6 mL) of a mixture of

trimethylsilyl chloride (3.4 mL), triethylamine

(4 mL), and tetrahydrofuran (8 mL). This mixture was spun in a centrifuge tube to settle the white precipitate formed and the supernatant that was added to the enolate solution was free of any precipitate. The reaction mixture was stirred at ambient temperature for 1 h and poured into petroleum ether. The petroleum ether solution was washed with sodium bicarbonate solution and brine, dried, and evaporated to give 7^ (1.43 g,

100%); IR (neat, cm'^): 2940 (s ), 2860 (m), 1640 (m), 1245 (s ), 1195 (s ),

895 (s ), 830 (s ); NMR (CCl^/CHgClg, 90 MHz) 6 4.3 (m, IH), 3.4 (m,

IH ), 2.1 - 1.1 (m, 8H), 1.0 (d, 2 Hz, 3H), 0.85 (s, 3H), 0.75 (s,

3H), 0.10 (s, 9H); m/z calcd (M+) 252.1909, obs 252.1915. 117

(±)-(3R*,3aS*,7aS*)-Tetrahydro-5-i sopropyli dene-3,7,7-trimethyl-4(3aH)■ indanone (67).

A solution of Tl, (1.43 g, 5.56 imol) in dichlo

romethane (10 mL) was added to a cold (-78°C)

solution of 2,2-dimethoxypropane (0.75 ml, 6.5

mmol) and titanium(IV) chloride (2.75 mL, 5.56

mol) in dichloromethane (50 mL). The reaction

mixture was allowed to warm to room temperature during 5 h then treated carefully with water at -78°C. The mixture was poured into water and extracted with ether. The ether phase was washed with water and brine, dried, and evaporated. The residue was admixed with diazabicycloundecene (1.52 g, 10 mmol) and molecular sieves (3Â,

1.25 g) in dichloromethane (25 mL) and was heated at reflux for 8 h.

This mixture was filtered, poured into 10% hydrochloric acid, and extracted with ether. The ether solution was washed with sodium bicar bonate solution and brine, dried, and evaporated. Purification by hplc

(elution with petroleum ether:ethyl acetate, 40:1) gave ^ (700 mg, 58%);

IR (CCI4 , cm 'l): 2940 (s ), 2860 (s ), 1685 (s ), 1625 (m), 1450 (m), 1360

(m); NMR (CDCI3 , 90 MHz) 6 2.5 - 2.0 (m, 4H), 1.9 (s, 3H), 1.75 (s,

3H), 1.6 - 1.2 (m, 5H), 1.10 (d, = 6 Hz. 3H), 1.00 (s , 3H), 0.95 (s ,

3H); m/z calcd (M'*') 220.1827, obs 220.1833.

Anal. Calcd for Ci^Hg^O: C, 81.76; H, 10.98. Found: C, 81.46; H,

10.98. 118

(±)-(3R*.3aS*.7aS*)-Tetrahydro-3.3',3 '.7.7-pentamethy1spiro[indan-5(4H)

2 '- 0x 1ran]-4-one (7^).

Ketone 67_ (260 mg, 1.18 mmol) in acetic acid (5

mL) previously saturated with potassium acetate

was cooled to 15°C. To this solution was added

38% peracetic acid In acetic acid (222 mg, 1.18

mmol). The reaction mixture was thawed to

ambient temperature and stirred for 24 h. The reaction mixture was poured Into ether and water neutralized with solid potassium carbonate and extracted with ether. The ether layer was washed with sodium bicarbonate solution and brine, dried, and evaporated to give 77, m.p. 72-74°C (210 mg, 75%). An analytical sample, m.p.

91.0-91.5°C, was prepared by recrystalIzatlon from petroleum ether; IR

(CCI4 , cm 'l): 2940 (s ), 2860 (m), 1725 (s ), 1455 (m), 1380 (m), 1370 (m),

1235 (w); NMR (CDCI3 , 90 MHz) 6 2.5 - 1.5 (m, 9H), 1.45 (s, 3H), 1.25

(s, 3H), 1.15 (s, 3H), 1.10 (d, J_= 6 Hz, 3H), 1.10 (s , 3H); NMR

(CDCI3 ) ppm 206.52, 70.15, 63.48, 60.20, 57.35, 46.36, 33.56, 31.80,

31.37, 29.25, 24.64, 20.39, 19.66, 19.30; m/z calcd (M*) 236.1776, obs

236.1784.

Anal. Calcd for CjgHg^Og: C, 76.23; H, 10.24. Found; C, 76.07; H,

10.21. 119

(±)-3,3-Dimethyl-4-[2-(2-methy1-l.3-dioxo1an-2-y1)ethy1]cyc1opentanone

(&).

2-(2-Bromoethyl)-2-methyl-l,3-dioxolane (15 g, 76

^ iranol) and 1,2 dibromoethane (2.5 mL) in

tetrahydrofuran (10 mL) was added over 6 h to

0 o magnesium shavings (5.4 g, 240 mg-at) in \_J tetrahydrofuran (200 mL). In a separate flask,

copper iodide (1.30 g, 8 mmol) in dimethyl

sulfide (10 mL) was cooled to -78°C and the

Grignard solution was introduced via canula. 4 ,4-Dimethylcyclopentenone

(4 9 , 36.4 mmol) in tetrahydrofuran (100 mL) was added over 3 h and the reaction mixture was maintained at -78°C for 15 h, quenched with ammonium chloride solution, filtered, and extracted with ether. The ether phase was washed with ammonium chloride solution and brine, dried, and evaporated. The residue was d is tille d to give TiB (5.50 g, 67%), bp

100-107°C at 0.45 to rr; IR (neat, cm"^) 2940 (s ), 2875 (s ), 1740 (s ),

1510 (m), 1370 (s ), 1210 (s ), 1020 (s ), 940 (m), 845 (s ); NMR (CDCI 3 ,

60 MHz) 6 4.0 (s, 4H), 2.6 - 1.4 (m, 7H), 2.2 (s, 2H), 1.3 (s, 3H). 1.2

(s, 3H), 0.9 (s, 3H); m/z calcd (M+-CH 3) 211.1334, obs 211.1338. 120

(±)-3,3-Dimethyl-4-(3-oxobuty1)cyc1opentanone (79).

Ketal 78_ (1.00 g, 44 mmol) and pyridinium

0 tosylate (200 mg) in acetone (28 mL) and water (2

ml) was heated at reflux for 6 h. The reaction

mixture was evaporated and the residue was

diluted with ether. The ether phase was washed

with sodium bicarbonate solution and brine, dried, and evaporated to give 79 (600 mg, 75%); IR (neat, cm"^): 2950

(s ), 2860 (m), 1740 (s ), 1710 (s ), 1455 (m), 1400 (m), 1360 (s ), 1255

(m), 1160 (m); NMR (CDCI3 , 60 MHz) 6 2.8 - 1.4 (m, 9H), 2.2 (s. 3H),

1.2 (s, 3H), 0.9 (s, 3H); m/z calcd (M+) 164.1201, obs 164.1198. 121

(±)-ci s-3,3a ,4,6a-Tetrahydro-3,3. 6 -trim ethyl-1[2H]-pentalenone (80)

Diketone 7£ (1.85 g, 10.2 nmol) was heated at

60°C for 5 h with potassium t e r t - butoxide (3.44

g, 30.6 nmol) in te r t-butanol (100 m l). The

te r t- butanol was removed in vacuo and the

residue was partitioned between ether and water.

The organic phase was washed with 5% hydrochloric

acid, sodium bicarbonate solution, and brine, and evaporated to give 80 (1.43 g, 86%); IR (neat, cm“^): 3050 (w ), 2860 (s ),

2840 (s ), 2760 (m), 1740 (s ); NMR (CDCI 3 , 200 MHz) 5 5.35 (m, IH ),

3.13 (d, J = 9 Hz, IH), 2.71 (dd, J = 16 and 8 Hz, IH), 2.4 (m, 2H), 2.24

(d, = 16 Hz), 2.0 (d, = 16 Hz), 1.77 (s, 3H), 1.09 (s, 3H), 1.05

(s , 3H); NMR (CDCI3 ) ppm 216.8, 137.0, 126.9, 62.6, 52.0, 51.0, 36.9,

33.9, 30.3, 24.3, 15.2; m/z calcd (M+) 164.1201, obs 164.1211.

Anal. Calcd for Cj^H^gO: C, 80.44; H, 9.82. Found: C, 80.34; H, 9.88. 122

(±)-4-(3-Buteny1)-3.3-dimethylcyclopentanone (81).

4-Bromo-l-butene (48.6 g, 0.36 m ol) in tetra-

n hydrofuran (80 ml) was added to magnesium

shavings (17.6 g, 0.72 g-at) over 30 min. The

reaction mixture was diluted with tetrahydrofuran

(80 mL), heated at reflux for 1 h, cooled to

-25°C, and treated with a solution of the copper bromide-dimethyl sulfide complex (24.6 g, 0.12 mol) in dimethyl sulfide

(160 mL). The blue solution was stirred for 1 h. 4 ,4-Dimethylcyclopen- tenone (20 g, 0.18 mol) in tetrahydrofuran (80 mL) was added over 2 h and the reaction mixture was allowed to s tir at ambient temperature for 15 h, quenched with ammonium chloride solution, and filte re d through a C elite pad. The filtra te was diluted with ether, washed repeatedly with ammonium chloride solution and once with brine, dried, and evaporated.

The residue was purified by hplc (elution with petroleum ether:ethyl acetate, 30:1) to give 8^ (18.6 g, 65%); IR (neat, cm“^): 2980 (s ), 2780

(s ), 1750 (s ), 1650 (m), 1465 (m), 1375 (m), 910 (s ); NMR (CDCI 3 , 300

MHz) 6 5.82 (m, IH ), 5.00 (dd, ^ = 1 9 and 1 Hz, 2H), 2.43 (g., J = 9.6 Hz,

IH), 2.17 (m, IH), 2.13 (s, 2H), 2.04 - 1.80 (m, 3H), 1.67 (m, IH), 1.26

(m, IH), 1.14 (s, 3H), 0.88 (s, 3H).

Anal. Calcd for C, 79.46; H, 10.91. Found: C, 79.21; H,

10.88 . 123

(±)-(3aR*,6S*,6aS*)-Hexahydro-3,3,6-trimethy1-l(2H)-pentalenone (75).

A. Catalytic Hydrogenation of 80.

Ketone 80 (1.43 g, 8.72 mmol) in absolute ethanol o (50 mL) was shaken under an atmosphere of hydro

gen (50 psi) in the presence of 10% palladium on

carbon catalyst (300 mg) for 9 h. The solution

was filte re d through a pad of C elite and

evaporated to give 7^ (1.13 g, 78%). Further purification by chromatography on s ilic a gel (45 g, 100-200 mesh, elution with 2% ethyl acetate in petroleum ether) gave pure 75 (0.75 g, 52%) as a single isomer; IR (neat, cm“^): 2960 (s), 2880 (m), 1735 (s), 1460 (m),

1380 (m), 1150 (w); NMR (CDCI 3 , 200 MHz) Ô 2.68 (dd, J = 10 and 1.2

Hz, IH), 2.4 (m, 2H), 2.11 (dd, J_= 16.4 and 1.5 Hz, IH), 1.93 (dt, J =

16 and 1.5 Hz, IH), 1.8 - 1.65 (m, 3H), 1,27 - 1.14 (m, IH), 1.13 (s,

3H), 1.02 (d, J = 6 Hz, 3H), 0.99 (s, 3H); NMR (CDCI3 ) ppm 220.9,

56.2, 53.8, 53.5, 37.6, 36.0, 35.6, 30.8, 27.3, 25.1, 16.4; m/z calcd

(M+) 166.1358, obs 166.1361.

Anal. Calcd for ^22^2 8 0 - C, 79.46; H, 10.91. Found: C, 79.26; H,

10.98.

B. Ene Reaction of 81.

Twenty pyrex glass tubes containing 8I_ (500 mg each, 60.2 mmol) were sealed under vacuum and heated at 320°C for 80 min. The combined product was purified by hplc (elution with 1% ethyl acetate in petroleum ether) to give 75 (7.85 g, 78%). 124

Tosylhydrazone of 75.

A solution of 7^ (1.00 g, 6.02 mmol)« tosylhy

drazine (1.4 g) and £-toluenesulfonic acid

N N H T s (lOOmg) is absolute ethanol (25 mL) was heated

at reflux for 16 h, transferred to an Erlenmeyer

flask, and refrigerated for 2 days. The

crystalline product was filte re d and dried to give 82^ (1.05 g, 52%) mp 153-155°C. The mother liquor was condensed and refrigerated to yield a second crop of ^ (0.40 g, 20%), mp 147.5-151°C;

IR (KBr, cm 'l): 3190 (s ), 2930 (s ), 2860 (m); 1590 (w ), 1440 (m), 1400

(m), 1320 (s ), 1160 (s ), 660 (m) ; NMR (acetone-^, 90 MHz) 5 7.9 - 7.3

(m, 4H), 5.7 (s. IH), 3.0 (t, J = 6 Hz, IH), 2.8 (s, IH), 2.6 (s. 3H),

2.3 - 1.3 (m, 7H), 1.2 (s, 3H), 0.9 (s, 3H), 0.7 (d, 6 Hz, 3H); m/z calcd (M+) 334.1715, obs 334.1723. 125

Shapiro Reaction of 82.

Hydrazone 82 (300 mg, 0.90 mmol) was added to

n-butyllithium (2.25 mL of 1.6 M in hexane, 3.6

mmol) in tetramethylethylenediami ne (5 mL) at

-30°C. The reaction mixture was stirred at

-30°C for 1 h and at ambient temperature for 1

h, quenched with water, and diluted with petroleum ether. The organic phase was washed with copper sulfate solu tion, sodium bicarbonate solution, and brine prior to drying. The solvent was removed to give 8^ (110 mg, 81%); IR (CCl^, cm"^): 3020 (w ),

2920 (s ), 2860 (m), 1450 (m), 1350 (m) ; NMR (CDCI 3 , 60 MHz) 6 5.4 (s,

IH), 3.3 (t, J = 10 Hz, IH), 2.4 - 1.2 (m, 5H), 1.3 (s, 2H), 1.1 (s, 6 H),

1.0 ppm (d, ^ = 1 0 Hz, 3H); tn/^ calcd (M^) 150.1408, obs 150.1412. 126

(±)-(3aR*,6S*,6aS*)-Hexahydro-2-[(Z)-hydroxymethy1ene]-3.3, 6 -trimethyl-

I(2H)-penta1enone .

Ketone 7^ (100 mg, 0.60 m o l) in benzene (4 mL)

was added to sodium methoxide (330 mg, 5.0 mol)

I? OH in benzene. The reaction mixture was cooled to

0°C and ethyl formate (355 mg, 0.40 mL, 4.8

mmol) was added. This mixture was stirred at

ambient temperature for 48 h. Ether was added and the suspension was extracted with water followed by 1 potassium hydroxide solution (3 x ). A fter acidification (concentrated hydrochloric acid), the aqueous solution was extracted with ether (4 x) and the combined organic layers were dried and evaporated to give the hydroxymethylene ketone (110 mg, 93%); IR (neat, cm”^): 3500 - 3000 (s ), 2960

(s ), 2880 (m), 1740 (s ), 1680 (s ), 1670 (s ), 1560 (m), 1500 (m), 1485

(m), 1470 (m), 1110 (m); NMR (CDCI 3 , 90 MHz) 6 9.7 (br s, IH), 7.3 (s,

IH), 3.1 (t, J = 7 Hz, IH), 2.55 - 2.05 (m, 3H), 1.95 - 1.45 (m, 3H),

1.22 (s, 6 H), 1.1 (d, J = 7 Hz, 3H); m/z calcd (M*) 194.1307; obs

194.1313, 127

(±)-(3aR*.6S*,6aS*)-2-[(Buty1thio)methy1ene]hexahydro-3,3,6-trimethy1- l(2H)-penta1enone (84).

A mixture of the hydroxymethylene ketone (110 mg,

0.56 mmol), magnesium sulfate (400 mg, 3.53

0 sBu m o l) , p-toluenesulfonic acid (10 mg, .05 mmol),

and jv-butanethiol (0.5 mL, 5.0 mmol) was heated

in benzene (20 mL) at reflux for 20 h. The

reaction mixture was filtered and evaporated, and the residue was taken up in ether. This solution was washed with sodium bicarbonate solution and brine, dried, and evaporated to give &4 (100 mg,

72%); IR (neat, cm"^): 2970 (s ), 2940 (s ), 2880 (s ), 1695 (s ), 1575 (s ),

1460 (s ), 1390 (w), 1380 (w ), 1370 (m), 1230 (m). 1165 (m), 1135 (m);

NMR (CDCI3 , 200 MHz) 6 7.31 (s, IH), 2.78 (g_, J = 10 Hz, 3H), 2.4 - 2.2

(m, 3H), 1.74 - 1.64 (m, 4H), 1.49 - 1.35 (m, 3H), 1.38 (s, 3H), 1.21 -

1.09 (m, 8H), 0.94 (5 , ^ = 12 Hz, 3H); m/z calcd (M*) 266.1704, obs

266.1716. 128

(±-(3aR,6R*,6aR*)-2-[(Buty1th1o)methy1ene3hexahydro-3,3, 6 -trimethyl- 6a-

[ (E)-3(tri methyl si1y1)-2-butenyi]-l(2H)-pentalenone (85).

Compound 84 (100 mg, 0.40 nmol) in tetrahydro-

7*^^ furan (5 ml) was added over 1 h to a solution of

lithium diisopropyl ami de (1.1 equiv, 0.44 nmol;

from 0.27 mL of 1.58 M iv-butyllithium in hexane

and 0.06 ml of diisopropyl ami ne in tetrahydro

furan (5 ml) at 0°C. The enolate was quenched with a tetrahydrofuran (3 mL) solution of (Ej-l-iodo-3-(trim ethylsilyl)-

2-butene (114.3 mg, 0.45 mmol). The reaction mixture was warmed to ambient temperature over 3 h, poured into water (20 mL), and extracted with ether (40 mL). The ether phase was washed with sodium bicarbonate solution and brine, dried, and evaporated. The residue was fractionated by preparative tic (elution with petroleum etherrethyl acetate, 40:1) to give 85 (100 mg, 67%); IR (CCl^, cm'^): 2960 (s ), 2940 (s ), 2880 (s ),

1690 (s ), 1620 (w), 1580 (s ), 1250 (s ). 840 (s ). 750 (s ), 740 (m); NMR

(CDCI3 , 200 MHz) 6 7.30 (s , IH ), 5.74 (m, IH ), 2.83 ( t , J = 7.2 Hz, 2H)

2.42 (m, 2H), 2.16 (m, IH), 2.0 - 1.4 (m, 9H), 1.69 (s, 3H), 1.34 (s,

3H), 1.23 (s, 3H), 1.15 - 0.90 (m, 6H ), 0.05 (s , 9H); m/z calcd (M*)

392.2569, obs 392.2579. 129

(±)-(3aR*,6R*,6aR*)-Hexahydro-3,3, 6 -trimethyl-6a-[(E)-3-(trimethyl s i1y1)•

2-buteny1] -I(2H)-pentalenone (74).

Compound 8^ (80 mg, 0.20 mmol) was heated at

TW5 reflux in aqueous 25% potassium hydroxide solu

tion (3 mL) and diethylene glycol (4 mL) for 48

h. The reaction mixture was diluted with brine

and extracted with ether. The organic phase was

washed with sodium bicarbonate solution and brine, dried, and evaporated to give 74 (40 mg, 70%); IR (CCl^, cm~^):

2980 (s ), 2960 (s ), 2880 (m), 1730 (s ), 1620 (w ), 1445 (m), 1250 (s ), 850

(s ), 840 (s ); NMR (CDCI 3 , 200 MHz) 6 5.70 (m, IH), 2.52 (d, J = 7 Hz,

IH), 2.45 (d, = 8.7 Hz, IH), 2.24 (d, ^g = 8.7 Hz, IH), 2.13 (m,

IH ), 2.02 (d . J = 4 Hz, 2H), 2.0 - 1.7 (m, 2H), 1.65 (d, J = 2Hz, 3H),

1.3 - 1.1 (m, 2H), 1.05 (s, 3H), 0.98 (s, 3H), 0.85 (d, J = 6.5 Hz, 3H),

0.01 (s, 9H).

Anal. Calcd for CigHggSiO: C, 73.90; H, 11.03. Found: C, 73.63;

H, 10.96. 130

(±)-(3aR*.6R*,6aR*)-6a-[2.3-Epoxy-3-(trimethylsilyl)buty1]hexahydro-3.3.6- trimethyl -1(2H)-pentalenone .

To compound 74 (54.1 mg, 0,19 mmol) in dichlo-

TMs romethane (3 mL) at 0°C was added solid sodium A' / ^ \ bicarbonate (50 mg) followed by buffer washed

^ — A mnchloroperbenzoic acid (50 mg, 0.79 mmol). The

' reaction mixture was stirred at 0°C for 2 h and at

ambient temperature for 16 h. After a ll of 74 was consumed (tic analysis) the reaction mixture was diluted with ether and washed with sodium sulfite solution, sodium bicarbonate solution, and brine prior to drying and evaporation. There was obtained 60 mg (100%) of the epoxide; IR (CCl^, cm"^): 2970 (s ), 2880 (m), 1735 (s ), 1460 (m), 1250

(s ), 860 (s ), 845(s); NMR (CCl^/CHgClg, 90 MHz) 6 2.9 - 1.5 (m, IIH),

1.9 (s, 3H), 1.2 (d, J = 3Hz, 3H), 1.1 (s, 3H), 0.80 (s, 3H), 0.10 (s,

9H); m/z calcd (wf) 308.2171; obs 308.2179. 131

(±)-(3aR*,6R*.6aR*)-Hexahydro-3,3,6-tr1methy1-6a-(3-oxobuty1 )-l(2H)-pent al enone (73).

The epoxysilane (60 mg, 0.19 mmol) was heated at

reflux for 15 h in absolute methanol (5 mL) and

20% aqueous sulfuric acid (5 mL). The reaction

mixture was diluted with water and extracted

with ether. The organic phase was washed with

sodium bicarbonate solution and brine, dried, and evaporated to give 73 (34 mg, 74%); IR (CCl^, cm"^): 2970 (s ), 2880

(m), 1730 (b r), 1460 (m), 1160 (m) ; NMR (CDCI 3 , 200 MHz) 6 2.7 - 1.1

(m, 12H), 2.15 (s, 3H), 1.09 (s, 3H), 1.00 (s, 3H), 0.87 (d, 1= 6 .8 Hz,

3H); m/z calcd (M"*") 236.1776, obs 236.1782. 132

(±)-(IR*,3aR*, 8 aR*)-1,2,3,3a.4,5 J ,8 -Octahydro-l,4,4, 6 -tetramethylcyclopenta[c]penta1ene (76).

The titanium metal catalyst was prepared by

heating at reflux a mixture of titanium(III)

chloride (500 mg, 3.35 mmol) and freshly pre

pared zinc-copper couple (500 mg, 7.7 m o l) in

dimethoxyethane (25 ml) for IS h. Diketone 73

(70 mg, 0.30 m o l) in dimethoxyethane (10 mL) was added over 6 h and the resulting suspension was heated at reflux for

16 h. The reaction mixture was filtered and evaporated. The solid residue was leached with a small amount of petroleum ether. After evaporation, the residue (80 mg) was chromatographed (mplc, elution with

1% ethyl acetate in petroleum ether) and a fraction with a very high R^

(0.90 by tic ) was isolated (40.1 mg). The major component of this fraction was separated by vpc (5 f t x % in , 10% SE-30 on Chromsorb W,

135°C) to give 76 (0.60 mg, 1%); NMR (CDCI 3 , 200 MHz) 6 2.18 - 1.18

(m, lOH), 1.65 (s, 2H), 1.51 (s, 3H), 0.96 (s, 3H), 0.84 (d, 7 Hz,

3H), 0.80 (s, 3H); m/z calcd (M*) 204.1878, obs 204.1883. 133

74 l-Bromo-l-(trimethy1si1y1)ethy1ene (89)

Méthylmagnésium iodide (0.80 mol) was prepared

by adding iodomethane (113.56 g, 0.80 mol) dis- Br TMs —^ solved in ether (200 mL) to magnesium shavings

(I9.44g, 0.80 g-at) over 2 h. Trichlorovinyl-

silane (40.38 g, 0.25 mol) in ether (50 mL) was

added over 1 h and the reaction mixture was heated at reflux for 3 h. Hydrochloric acid (5%) was added, the layers were partitioned, and the organic phase was washed with sodium bicar bonate solution and brine before drying. To this cooled (0°C) solution was added bromine (39.95 g , 12.8 mL, 0.25 mol) over 2 h. The reaction mixture was washed with sodium thiosulfate solution and brine, dried, and evaporated to give 1,2 dibrom o-l-(trim ethylsilyl)ethane (49.6 g,

80%).

The dibromide was shaken with diethylamine (27.8 g, 40 mL, 0.40 mol) at ambient temperature for 20 h. The reaction mixture was poured into ether and water and the organic phase was washed with 5% hydrochloric acid, sodium bicarbonate solution, and brine prior to drying and solvent evaporation. Distillation at 63-66°C and 112 torr gave 15.63 g (47%) of 89; NMR (CCl^, 60 MHz) 6 6.0 (d, ^ = 4 Hz, 2H), 0.1 (s, 9H). 134

74 2-(Trimethyl s i1y1)but-l-en-3-o1 (90)

Bromide 89 (128 g, 0.71 mol) in tetrahydrofuran

(500 ml) was added dropwise to magnesium shavings

^ (19 g, 0.78 g-at). The mixture was heated at

reflux for 30 min before acetaldehyde (34.3 g,

0.78 mol) in tetrahydrofuran (100 mL) was added.

The reaction mixture was heated at reflux for

15 min, cooled, treated with 5% hydrochloric acid, and diluted with ether. The ether phase was washed with sodium bicarbonate solution and brine, dried, and distilled to give 55.7 g (60%) of 90^ (bp 80-82°C, 30

to rr); NMR (CCl^, 60 MHz) 6 5.4 (m, 2H) 4.3 (g., 0_ = 6 Hz, IH ), 2.3 (br m, IH), 1.1 (d, 1= 6 Hz, 3H), 0.1 (s , 9H). 135

(E)-l-Iodo-2-(trimethy1si1y1)but-2-ene (8 8 ).

To a solution of 9jD (11.0 g, 76.4 nmol) and

triethylamine (18 ml) in dichloromethane (100 ml)

Tivts cooled to 0°C was added methanesulfonyl chloride

(6 .6 mL, 83 m ol) dissolved In dichloro

methane (25 mL). The reaction mixture was

stirred at 0°C for 1 h and poured Into Ice water. The organic phase was washed with 5% hydrochloric acid, sodium bicarbonate solution, and brine prior to drying and solvent evaporation.

There was Isolated 15.06 g (89%) of the mesylate; IR (neat, cm”^): 2960

(s ), 1350 (s ), 1250 (s ), 1175 (s ), 900 (s ), 835 (s ); NMR (CCl^, 60

MHz) 6 5.6 (m, 2H), 5.1 (m, IH), 2.7 (s, 3H), 1.4 (d, J = 6 Hz, 3H), 0.1

(s, 9H).

The mesylate (15.06 g, 68 mmol) In acetone (500 mL) was heated at reflux with sodium Iodide (12.15 g, 81 mmol) for 15 h. The reaction mixture was filtered and evaporated, and the residue was dissolved In ether. The ether solution was washed with sodium thiosulfate solution, sodium bicarbonate solution, and brine, dried, and evaporated to give 88

(11.42 g, 66%); IR (neat, cm'^): 2940 (m), 1600 (m), 1240 (s), 1185 (s),

1140 (s ), 830 (s ); ^H NMR (CDCI 3 , 200 MHz) 6 6.01 (g., i = 6 Hz, IH ), 3.99

(s, 2H), 1.72 (d, J = 7.1 Hz, 3H) 0.13, (s, 9H); m/z calcd (M+) 253.9990, obs 253.9984. 136

(±)-(3aR*,6R*,6aR*)-2-[(Butylthio)methy1ene]hexahydro-3,3, 6 -trimethyl- 6a-

[(E)-2-(trimethy1sily1)-2-butenyl]-l(2H)-penta]enone (91).

Compound ^ (3.60 g, 14.1 mmol) in tetrahydro-

TMs furan (40 mL) was added over 1 h to a -30°C

solution of lithium di isopropyl ami de [15.51 /9 mmol, from 9.7 ml of 1.6 M n_-butyllithium in

^ hexane and diisopropyl ami ne (2.2 mL)] in te tra

hydrofuran (80 mL). Iodide 88 (4.11 g, 16.2 m o l) in tetrahydrofuran (40 mL) was added dropwise at -30°C. The reaction mixture was allowed to s tir at ambient temperature during 3 h and partitioned between ether and 5% hydrochloric acid solution. The ether layer was washed with sodium bicarbonate solution and brine, dried, and evaporated to give 6.38 g of residual o il. Chromatography of this material (hplc, elution with petroleum ether:ethyl acetate, 1 0 0 :1 ) gave 91 (2.47 g, 45%); IR (CCl^, cm"^): 2970 (s ), 2880 (s ), 1690 (s ),

1660 (m), 1570 (s ), 1460 (s ), 1250 (s ), 850 (s ); NMR (CDCI 3 , 200 MHz)

6 7.22 (s, IH), 6.05 7Hz, IH), 2.76 (t, J_ = 12 Hz, 2H), 2.5 - 0.8

(m, 18H), 1.67 (d, J = 7 Hz, 3H), 1.27 (s, 3H), 1.15 (s, 3H), 0.01 (s,

9H); m/z calcd (M+) 392.2969, obs 392.2576. 137

(±)-(3aR*.6R*.6aR*)-Hexahydro-3.3,6-tnmethy1-6a-[(E)-2-(trimethy1si1y1 )■

2-butenyl]-1 (2H)-pentalenone (87).

Compound 91^ (2.47 g, 6.3 m o l) in 25% potassium

TMS hydroxide solution (90 mL) and diethylene glycol

(120 mL) was heated at reflux for 48 h. The

reaction mixture was diluted with brine and

extracted with ether. The ether layer was washed

with sodium bicarbonate solution and brine, dried and evaporated to give 1.67 g of o il. This material was purified by preparative thin layer chromatography (elution with petroleum ether:ethyl acetate, 20:1) and there was isolated 840 mg (46%) of IR (neat, cm"^):

2970 (s ), 2860 (s ), 1730 (s ), 1650 (w), 1245 (m), 830 (s ); NMR (CDCI 3 ,

200 MHz) 6 6.05 (g_, J = 7Hz, IH), 2.57 (d, = 12 Hz, IH), 2.23 (d,

= 12 Hz, IH), 2.3 - 1.00 (m, 8H), 1.65 (d, J = 7 Hz, 3H), 0.98 (s, 3H),

0.90 (s, 3H), 0.78 (d, J = 6 .8 Hz, 3H), 0.01 (s, 9H) ; m/z calcd (M+-CH 3 )

277.1988, obs 277.1997.

Anal. Calcd for C^gHggSiO: C, 73.90; H, 11.03. Found: C, 73.69; H,

10.91. 138

(±)-(3aR*.6R*.6aR*)-6a-[2,3-Epoxy-2-(triinethy1si1yT)buty1]hexahydro-3,3,6- tri methyl- 1 ( 2H)-pentalenone.

To compound 87^ (770 mg, 2.64 irniol) in dichloro

methane (40 mL) with sodium bicarbonate (0.50 g) \ O /\ at 0°C was added buffer washed mnchloroperbenzoic o y ^ J l acid (650 mg, 4.0 mmol). The reaction mixture

was stirred at room temperature for 16 h and

additional m^chloroperbenzoic acid (0.5 equiv) was added. When the oxidation was complete (tic analysis), ether was added. The organic phase was washed with sodium s u lfite solution, sodium bicarbonate solution, and brine prior to drying and evaporation. There was obtained 840 mg (100%) of the epoxysilane; IR (neat, cm"^): 2970 (s ),

2860 (s ), 1725 (s ), 1450 (m), 1245 (s ), 830 (s ); NMR (CCl^/CHgClg, 90

MHz) 5 2.8 (g_, 5Hz, 2H), 2.8 - 0.5 (m, 15 H), 1.1 (s , 6 H), 0.10 (s,

9H); m/z calcd (M+) 308.2171, obs 308.2179. 139

(±)-(3aR*.6R*,6aR*)-Hexahydro-3.3. 6 -trimethyl-6a-(2-oxobuty1)-l(2H)■ pental enone (.86 ).

The epoxysilane (840 mg, 2.64 mmol) was heated at

reflux in 20% sulfuric acid (50 ml) and methanol

(50 ml) for 16 h. The reaction mixture was

diluted with water and extracted with ether. The

ether phase was washed with sodium bicarbonate

solution and brine, dried, and evaporated. The residue was purified by mplc (elution with petroleum ether:ethyl acetate,

20:1) to give 86 (400 mg, 65%); IR (neat, cm"^): 2980 (s ), 2870 (s ), 1720

(s, br), 1460 (s), 1410, (s). 1180 (s); NMR (CDCI 3 , 200 MHz) 6 2.91 ( d, = 17 Hz, IH), 2.57 (d, J^g = 17 Hz, IH), 2.5 (m, 3H), 2.1 (m, 2H),

1.92 (m, 3H), 1.2 - 1.0 (m, 2H), 1.10 (s, 3H), 1.00 (t, 0 = 7.6 Hz, 3H),

0.99 (s, 3H), 0.82 (d, J = 6.7 Hz, 3H) ; m/z calcd (M+) 236.1778, obs

236.1782.

Anal. Calcd C^gHg^Og: C, 76.22; H, 10.24. Found: C, 76.61; H, 10.23, 140

(±)-(5aR*.8R*.8aR*)-4,5,5a,6.7,8-Hexahydro-3.5.5.8-tetramethylcyc1openta

[c]-penta1en-2(lH)-one (92).

Diketone 86_ (330 mg, 1.40 mmol) in tetrahydro

furan (30 mL), t e r t - butanol (2 ml) and potassium

t e r t - butoxide (330 mg, 2.28 mmol) was heated at

4G°C for 1 h. The reaction mixture was diluted

with ether, and washed with sodium bicarbonate

solution and brine. After drying and evapora tio n , there was isolated 250 mg (82%) of 9^; IR (neat, cm"^); 2980 (s ),

2870 (s ), 1710 (s ), 1660 (s ); 1450 (m), 1190 (m); NMR (COCI 3 , 200 MHz)

6 2.62 (d, = 17 Hz, IH), 2.41 (d, J^g = 17 Hz, IH), 2.32 - 1.55 (m,

7H), 1.74 (d, J = 2.3 Hz, 3H), 1.19 (m, IH), 1.12 (s, 3H), 0.80 (s, 3H),

0.71 (d, J = 7.1 Hz, 3H); m/z calcd (M+) 218.1671, obs 218.1676.

Anal. Calcd for C^gHggO: C, 82.44; H, 10.16. Found: C, 82.15; H,

10.13. 141

(±)-(3R*,3aS*,5aR*, 8 R*,8 aS*)-Octahydro-3.5.5, 8 -tetramethylcyclopenta[c]-

penta1en-2(lH)-one (9^).

Enone 9^ (93.8 mg, 0.43 mmol) in tetrahydrofuran

(3 mL) was added to lithium metal (30 mg, 4.3

mg-at) in liquid ammonia (50 mL). The reaction

mixture was allowed to reflux for 30 min before

isoprene (2 mL) and methanol (2 mL) were added.

The anmonia was allowed to evaporate and the

residue was partitioned between ether and water. The ether phase was washed with sodium bicarbonate solution and brine, dried, and evaporated

to give 95 mg of 9^ (100%); IR (neat, cm"^): 2960 (s), 2870 (s), 1740 (s),

1450 (m), 1180 (m); NMR (CDCI 3 , 300 MHz) 6 2.42 (m, 2H), 2.2 - 1.2

(series of m, lOH), 1.6 (d, ^ = 7 Hz, 3H), 1.1 (s, 3H), 1.0 (d, ^ = 7.5

Hz, 3H), 0.90 (s, 3H); NMR (CDCI3 ) ppm 64.28, 57.97, 52.42, 47.49,

46.43, 45.56, 41.27, 35.89, 31.93, 26.21, 23.81, 14.08, 10.92; m/z calcd

(M+) 220.1827, obs 220.1805.

Sodium Borohydride Reduction of (93).

To ketone 9^ (95 mg, 0.43 mmol) in cold (0°C) methanol (20 mL) was

added sodium borohydride (165 mg, 4.3 mmol). The reaction mixture was

stirred for 15 h, quenched with 5% hydrochloric acid solution, and

extracted with ether. The ether solution was washed with sodium bicar

bonate solution and brine, dried, and evaporated to give ^ (80 mg, 84%).

Two epimers were isolated by mplc (elution with petroleum etherrethyl ace

tate, 40:1): 94A mp 50-52°C (30 mg, 32%) and 9^mp 78-80°C (30 mg, 32%). 142

(±)-(2R*,35*,3aR*,5aS*, 8 S*,8 aR*)-Decahydro-3,5.5, 8 -tetramethylcyclopenta-

[c]penta1en-2-o1 (94A) .

IR (CCI^, cm'l): 3620 (w), 2980 (s), 2870 (s),

1450 (m), 900 (m); NMR (CDCI3 , 200 MHz) 6 4.19

OH (g., J = 3 Hz, IH). 2.38 - 1.05 (m, 13H), 1.01 (d,

7.5 Hz, 3H), 0.99 (s, 3H), 0.91 (d, J = 6.3

Hz, 3H), 0.79 (s, 3H); NMR (CDCI 3 ppm

65.00, 63.91, 54.01, 48.54, 47.40, 45.70, 42.42,

41.87, 35.97, 31.54, 26.57, 23.72, 14.16, 11.20;

m/z calcd (M+-CH 3 ) 207.1749, obs 207.1714.

(±)-(2R*.3R*,3aS*,5aR*,8R*,8aS*)-Decahydro-3,5.5,8-tetramethylcyclopenta-

[c]pentalen-2-ol (948).

IR (CCl^, cm'l): 3620 (w), 2980 (s), 2870 (s),

1450 (m), 1270 (m), 1050 (m); ^H NMR (CDCI 3 , 200

OH MHz) 6 3.83 (^ , 0 = 8.0 and 5.6 Hz, IH), 2.33 -

1.09 (m, 13H), 1.02 (d, J = 7.2 Hz, 3H), 1.00 (d,

J = 7.2 Hz, 3H), 0.96 (s, 3H), 0.78 (s, 3H) ;

NMR ( 00013) ppm 65.20, 52.96, 51.41, 48.69,

46.99, 46.46, 44.95, 35.73, 35.48, 30.97, 26.60,

23.49, 13.49; m/z calcd (M+) 222.1984, obs

222.2003. 143

(±)-(lR*,3aR*.5aS*,8aR*)-l,2,3.3a,4.5,5a.8-0ctahydro-1.4.4,6-tetramethy1' cyclopenta[c]penta1 ene (5i6).

Alcohol 94 (15.4 mg, 0.07 mmol) in benzene (3 mL)

was heated at 50°C with phosphorous oxychloride

(33 mg, 0.21 mmol) and pyridine (166 mg, 2.1

mmol) for 15 h. The reaction mixture was poured

into water (3 mL) and the benzene phase was

washed with sodium bicarbonate solution and brine prior to drying. The hydrocarbon was isolated from the benzene solution by vpc (5 f t . x % in column of 10% SE-30 on Chromosorb W, 140°C) to give ^ (1 mg, 7%); NMR (CDCI 3 , 200 MHz) 6 5.13 (s, IH), 2.58 (br d,

9 Hz, IH), 2.52 (d, 16.5 Hz, IH), 2.35 (d, 16.5 Hz, IH), 1.88

(dd, J = 10 and 2.5 Hz, IH), 1.75 (ddq, 12.5, 5.0, and 6 .6 Hz, IH ),

1.65 (s, 3H), 1.64 - 1.5 (m, 3H), 1.45 (br d, J = 13 Hz, IH), 1.36 (dd, J

= 13 and 9.2 Hz, IH), 1.05 (m, IH), 0.99 (s, 3H), 0.93 (s, 3H), 0.85 (d, J

= 7 Hz, 3H); m/z calcd (M"*") 204.1878, obs 204.1883. 144

(±)-2 ,2-Dimethyl-4-[2-(2-methy1-l,3-dioxo1an-2-y1)ethy1]cyc1opentanone

(M l)'

2-(2-Bromoethyl)-2-methyl-1,3-dioxolane (40.2 g,

^ 0.205 mol) with 1,2-dibromoethane (12 mL) in

tetrahydrofuran (115 mL) was added over 2 h to

Q Q magnesium shavings (15 g, 0.618 g-at) in te tra - \ / hydrofuran (300 mL). This mixture was cooled to

-78°C, copper iodide (3.9g, 0.0207 mol) in dimethyl

sulfide (30 mL) was added and the blue solution was stirred for 30 min. 5 ,5-Dimethylcyclopent-2-enone (11.38 g, 0.103 mol) in tetrahydrofuran (100 mL) was introduced over 2 h and the reaction mixture was warmed to ambient temperature and stirred for 15 h. Ammonium chloride solution was added with cooling (-78°C), followed by filtration and dilution of the filtrate with ether. The ether phase was washed with ammonium chloride solution and brine, dried, and evaporated to give 101

(29.45 g, 100%); IR (neat, cm“^): 2980 (s ), 2880 (s ), 1730 (s ), 1455 (s ),

1380 (s ), 1220 (s ), 1070 (s ); NMR (CDCI 3 , 300 MHz) 6 3.92 (d, ^ = 4 Hz,

4H), 2.55 - 1.35 (m, 9H), 1.31 (s, 3H), 1.06 (s, 3H), 1.00(s. 3H) ; m/z calcd (M+) 211.1334, obs 211.1329.

Anal. Calcd for C, 68.98; H, 9.80. Found: C, 68.49; H, 9.73. 145

(±)-2,2-Dimethyl-4-(3-oxobuty1)cyc1opentanone (102).

Ketal 1 ^ (29.45 g, 0.103 mol) with pyridinium

^ tosylate (4.6 g) in acetone:water (19:1, 780 mL)

was heated at reflux for 18 h. The reaction

mixture was evaporated and the residue was taken u up in ether. The ether phase was washed with

sodium bicarbonate solution and brine, dried, and evaporated to give 18.12 g (97%) of 102; IR (neat, cm“^): 2980 (s ), 2950

(s ), 2880 (s ), 1740 (s ), 1720 (s ), 1460 (s ), 1370 (s ), 1130 (s ); NMR

(CCI4 , 90 MHz) 6 2.8 - 1.2 (m, 9H), 2.1 (s, 3H), 1.05 (s, 3H), 0.95 (s,

3H); m/z calcd (M+) 164.1201, obs 164.1191.

Anal. Calcd for 72.48; H, 9.96. Found: C, 71.96; H,

10.00. 146

(±)-cis-3.3a.4,6a-Tetrahydro-2>2.6-trimethy1-l(2H)-penta1enone (103).

Diketone 10^ (18.12 g, 99.6 iranol) with potassium

t e r t - butoxide (34 g, 298.8 nrniol) in t e r t - butanol

(1.5 L) was heated at 50°C for 5 h. The reaction

' mixture was evaporated, 5% hydrochloric acid was

added, and the acidic aqueous layer was extracted

with ether. The ether phase was washed with sodium bicarbonate solution and brine, dried, and evaporated. Purifica tion by elution of the residue with 5% ethyl acetate in petroleum ether on a column of s ilic a gel (800 g) gave IjM (11.6 g, 71%); IR (neat, cm-1):

3040 (w), 2980 (s ), 2870 (s ), 1740 (s ), 1650 (w), 1460 (m), 1390 (m), 1100

(m), 810 (m); NMR (CDCI 3 . 300 MHz) 6 5.31 (s, IH), 3.21 (m, IH), 2.90

(m, IH), 2.58 (m, IH), 2.05 (m, 2H), 1.76 (ddd, J = 2.8, 2.6 and 1.6 Hz,

3H), 1.44 (dd, 1= 12.8 and 9.5 Hz, IH), 1.01 (s, 6H); m/z calcd (M*)

164.1201, obs 164.1181; m/z calcd (M^-CHg) 149.0996, obs 149.0942. 147

(±)-(3aR*,6R*,6aR*)-Hexahydro-2,2.6-trimethy1-l(2H)~penta1enone (100).

Enone 103 (10.6 g, 64.6 mmol) with 10% palladium

on carbon (1.1 g) in absolute ethanol (100 mL)

was shaken under an atmosphere of hydrogen (50

psi) for 48 h. The reaction mixture was filtered

and the filt r a t e was evaporated to give 100 (9.0

g, 84%); IR (neat, cm"^): 2960 (s ), 2880 (s ),

1730 (s ). 1460 (s ), 1380 (m), 1360 (m), 1150 (m), 860 (m); NMR (CDCI 3 .

300 MHz) 6 2.78 (m, 2H), 2.28 (m, IH). 2.00 (m. IH), 1.80 (m, 2H), 1.58

(m, IH), 1.30 (m, 2H), 1.02 (s, 3H), 1.00 (s, 3H), 0.99 (d, ±= 7 Hz, 3H):

NMR (CDCI3 ) ppm 225.1, 53.9, 47.8, 44.7, 38.1, 37.6, 34.5, 31.9, 25.8,

22.8, 17.3; m/z calcd (M+) 166.1358, obs 166.1362.

Anal. Calcd for C^iHjgO: C, 79.46; H, 10.91. Found: C, 79.23; H,

11.05. 148

(±)-4-(3-Buteny1)-2,2-dimethylcyclopentanone (104).

4-Bromobutene (2.43 g, 18 mmol) in tetrahydro-

n furan (10 ml) was added dropwise to magnesium

shavings (0.88 g, 36 mg-at). Tetrahydrofuran

(10 mL) was added and this mixture was heated at

reflux for 30 min, cooled to -30°C and treated

with a solution of copper bromide-dimethyl

sulfide complex (1.23 g , 6 mmol) in dimethyl sulfide (20 mL). The blue

solution was stirred for 1 h, 5,5-dimethylcyclopent-2-enone (1.00 g, 9 mmol) was added over 1 h, and the reaction mixture was warmed to ambient

temperature and stirred for 15 h. To this solution was added 5%

hydrochloric acid followed by dilution with ether and filtration. The

filtra te was washed with ammonium chloride solution and brine, dried, and evaporated. Purification by mplc (elution with petroleum ether:ethyl acetate, 30:1) gave 104 (820 mg, 55%); IR (neat, cm"^): 3090 (m), 2980

(s ), 2940 (s ), 2880 (s ), 1745 (s ), 1645 (m), 1510 (m), 910 (s ); NMR

(CCI4 , 90 MHz) 6 5.9-5.5 (m, IH), 5.1 - 4.8 (m, 2H), 2.6 - 1.2 (m, 9H),

1.05 (s, 3H), 0.95 (s, 3H) ; m/z calcd (M"*") 166.1357, obs 166.1322. 149

(±)-(3aR*.6R*.6aR*)-Hexahydro-2,2,6-trimethy1-6a-[(E)-2-(tnmethy1si1y1 )

2-buteny1]-l(2H)-pentalenone (99).

Ketone 100 (5.00 g, 30.1 mmol) in tetrahydro

furan (50 mL) was added over 1 h to a cooled TMS (-30°C) solution of lithium di isopropyl ami de

[from 33.1 mmol of diisopropyl ami ne (4.75 mL)

and 33.1 mmol of n_-butyllithium (20.7 mL of a

1.6 M hexane solution)] in tetrahydrofuran (250

mL). To the enolate solution was added iodide

88 (11.5 9 , 45 mmol) in tetrahydrofuran (20 mL) and the reaction mix ture was warmed to ambient temperature over 3 h, poured into 5% hydro chloric acid, and extracted with ether. The ether layer was washed with sodium bicarbonate solution and brine, dried, and evaporated. Purifica tion by mplc gave 99 (4.5 g, 51%); IR (neat, cm"^): 2980 (s ), 2880 (s ),

1730 (s ), 1610 (m), 1460 (m), 1380 (m), 1360 (w ), 1260 (s ), 850 (s );

NMR (CDCI3 , 300 MHz) 6 6.07 (g_, 0_ = 6 .8 Hz, IH ), 2.57 (g^ J = 6 Hz, IH ),

2.49 (d, = 13 Hz, IH ), 2.34 (d, = 13 Hz, IH ), 1.90 (m, 2H), 1.73

(d, J = 6 .8 Hz, 3H), 1.55 (m, 2H), 1.30 (m, 3H), 1.00 (s, 3H), 0.98 (s,

3H), 0.91 (d, J = 7 Hz, 3H), 0.09 (s, 9H); m/z calcd (M+-CH 3 ) 277.1998, obs 277.2021.

Anal^ Calcd for CjgHggOSi; C, 73.90; H, 11.03. Found: C, 73.83;

H, 11.03. 150

(±)-(3aR*,6R*,6aR*)-6a-[2,3-Epoxy-2-(tnmethy1si1y1)buty1]hexahydro-2.2.

6-tr1methy1-l(2H)-pentalenone.

Vinylsilane 99 (4.0g, 13.7 mmol) with solid

sodium bicarbonate ( 8 .0 g) in dichloromethane TMS A (100 mL) was cooled to 0°C and buffer washed m-chloroperbenzoic acid (3.5 g, 20.6 mmol) was

added. The reaction mixture was stirred at

ambient temperature for 15 h, poured into

potassium sulfite solution, and extracted with ether. The ether phase was washed with sodium bicarbonate solution and brine. After drying and evaporation, 4.13 (98%) of the epoxysilane was Isolated; IR (neat, cm"^): 2970 (s ), 2880 (s ), 1730 (s ), 1460 (s ),

1380 (m), 1250 (m), 840 (s ); NMR (CCl^/CHgClg, 90 MHz) 6 3.0 - 1.4

(m, IIH), 1.2 (d, J = 5 Hz, 3H), 1.05 (s, 3H), 0.95 (s, 3H), 0.90 (d, ^

= 7 Hz, 3H), 0.10 (s , 9H); m/z calcd (M*) 308.2171, obs 308.2189. 151

(±)-(3aR*.6R*,6aR*)-Hexahydro-2.2,6-trimethy1-6a-(2-oxobuty1)-l(2H)-pen ta l enone (98).

The epoxysilane (4.13 g, 13.4 mmol) in methanol

(100 mL) and 20% sulfuric acid (100 mL) was

heated at reflux for 2 h. The reaction mixture

was diluted with ether and washed with sodium

bicarbonate solution and brine. After drying

and evaporation, the residue was purified by

mplc (elution with 5% ethyl acetate in petroleum ether) to give 3.0 g (95%) of 98^; IR (neat, cm"^): 2980 (s ),

2790 (s),1730 (s ), 1720 (s ), 1460 (m), 1380 (m), 1360 (m); NMR

(CDCI3 , 300 MHz) 6 3.02 (d, ;^g = 17 Hz, IH), 2.65 (l. i = 7 Hz, IH),

2.46 (d, = 17 Hz, IH), 2.35 (g_, J^= 7.3 Hz, 2H), 2.15 (dd, J = 13.7 and 9.8 Hz, IH), 1.8 (m, 3H), 1.6 (m, IH), 1.4 (m, 2H), 1.20 (s, 3H),

1.03 (s , 3H), 0.98 ( t , J = 7.3 Hz, 3H), 0.87 (d, J_= 7 Hz, 3H); m/z

calcd (M+) 236.1776, obs 236.1769.

Anal. Calcd for C^5 H2402= C, 76.21; H, 10.24. Found: C, 76.11; H,

10.25. 152

(±)-(5aR*,8R*.8aR*)-4,5,5a ,6,7,8-Hexahydro-3,4,4, 8 -tetramethylcyclopenta[c]-penta1en-2(lH)-one. (IjM ).

Diketone 98 (2.0 g, 8.47 mmol) in 20% sodium

ethoxide in ethanol [prepared from 345 mg-at of

sodium metal (7.95 g) and absolute ethanol (150

mL)] was heated at reflux for 10 h. The reac

tion mixture was evaporated, 5% hydrochloric

acid solution was added to the residue, and this mixture was extracted with ether. The ether phase was washed with sodium bicarbonate solution, dried, and evaporated. The residue was purified by mplc (elution with 9% ethyl acetate in petroleum ether) to give 105 (800mg, 43%); IR (neat, cm"^): 2970 (s ), 2880 (s ), 1710 (s ),

1640 (s ), 1460 (s ), 1380 (s ); NMR (CDCI 3 , 300 MHz) 6 2.57 (d, =

13 Hz, IH ), 2.40 (m, IH ), 2.39 (d, ^ g = 13 Hz, IH ), 2.0 (m, 2H), 1.80

(s, 3H), 1.6 (m, 5H), 1.34 (s, 3H), 1.24 (s, 3H). 0.78 (d, J = 7 Hz,

3H); NMR (CDCI3 ) ppm 210.05, 191.79, 65.22, 57.41, 51.12, 47.89,

47.46, 45.71, 40.84, 33.46, 31.49, 29.47, 26.02, 18.10, 8.47; m/z calcd (M+)

218.1671, obs 218.1668. 153

A. Dissolving Metal Reduction of Enone 105.

An ether solution (10 mL) of tricyclic enone 105 (680 mg, 3.12 mmol) was added dropwise to lithium metal (220 mg, 31.2 mg at) in liquid ammonia (100-150 mL, d is tille d from sodium) at -33°C. The reaction mixture was stirred for 2 h at -33°C, then quenched with solid ammonium chloride. The liquid ammonia was evaporated and the residue was treated with 5% aqueous hydrochloric acid and extracted into ether. The combined ether extracts were washed with saturated aqueous sodium bicarbonate solution and brine, dried, and evaporated. Chromatography of the residue (530 mg) on silica gel (elution with 4.5% ethyl acetate in petroleum ether) gave the two epimeric tric y c lic ketones: 150 mg of

106 as a crystalline solid (mp 45-48°C) and 30 mg of 108, for a combined yield of 38%. 154

(±)-(3R*,3aR*,5aR*,8R*,8aR*)-Octahydro-3,4,4,8-

tetramethy1cyc1openta[c]penta1en-2(lH)- one

(106);

IRCCCl^.cm"^): 2940 (s ). 2870 (m), 1740 (s ), 1460

(m), 1200 (m); NMR (CDCl^, 300 MHz) 6 2.53

(d, = 18 Hz. IH), 2.53 (m, IH), 2.38 (g_, J

= 7 Hz, IH ), 2.30 (d, = 18 Hz, IH ), 2.05

(d, J = 6.7 Hz, IH), 1.8 - 1.0 (m, 7H), 1.20 (d, ±= 7Hz, 3H), 1.09 (s,

3H), 1.01 (d, J = 6.7 Hz, 3H), 0.81 (s, 3H) ; m/z calcd (M+) 220.1827, obs 220.1837.

Anal. Calcd for CigHg^O: C, 81.76; H, 10.99. Found; C, 81.91; H,

10.82. 155

(±)-[3R*,3aS*.5aS*, 8 S*,8 aS*)-Octahydro-3.4.4. 8 -

tetramethylcyclopenta-[c]pentalen-2(lH)-one

(108):

IR(CCl4 ,cm"l): 2930 (s ), 2860 (s ), 1740 (s ),

1460 (m), 1180 (m), 1120 (m); NMR (CDCl 3 ’ 300 MHz) 6 2.70 (d, v^g = 16.8 Hz, IH), 2.43

(q, J = 7.5 Hz, IH), 2.27 (m, IH), 2.20 (d, ^g

= 16.8 Hz, IH ), 1.9 - 1.1 (m, 8H), 1.13 (d, ^ =

7.5 Hz, 3H). 1.03 (s, 3H), 0.92 (s, 3H), 0.88 (d, 6.7 Hz, 3H); m/z calcd (M+) 220.1827, obs 220.1870.

B. Catalytic Hydrogenation of Enone 105.

A methanol solution of tric y c lic enone 105 (92.0 mg, 0.42 mmol) was shaken with 5% palladium on carbon catalyst (40 mg) under an atmosphere of hydrogen (50 psi) fo r 24 h. The mixture was filte re d through a

Celite pad and evaporated to give 71.1 mg (76%) of 106. 156 ( ±)- (2R*,3S*,3aS*,5aS*, 8 S*,8 aS*)-Decahydro-3,4,4, 8 -tetramethylcyclopenta

[c]penta1en-2-o1 (107):

To tric y c lic ketone 10^ (71.1 mg, 0.32 rmol) in

ether solution (10 mL) was added lithium alumi

num hydride (3.2 mmol, 122 mg). This mixture

was stirred for 24 h at ambient temperature

a fter which the excess hydride was destroyed by addition of ethyl acetate followed by 5% aqueous hydrochloric acid. The layers were separated and the aqueous phase was extracted with ether.

The combined organic phases were washed with saturated aqueous sodium bicarbonate solution and brine, dried, and evaporated. The residue

(55.4 mg) was chromatographed on s ilic a gel (elution with 10% ethyl acetate in petroleum ether) and three components were isolated. The f ir s t to elute was starting ketone 106 (10 mg, 14%). The second compound, the pure tric y c lic alcohol 107 (20.5 mg, 34%), was followed by an impure alcohol (9.8 mg, 16%) presumably an epimer of 107.

107: IR (CCI 4 , cm-1): 3640 (w ), 2980 (m), 2940 (s ), 2880 (s ), 1460

(m), 1120 (s ), 910 (s ); NMR (CDCI3 , 300 MHz) 6 3.29 (m, IH), 2.40 -

1.15 (m, 13H), 1.14 (s, 3H), 1.12 (d, J_= 7.3 Hz, 3H), 1.07 (s, 3H),

0.88 (d, J^= 6.7 Hz, 3H); m/z calcd (M^) 222.1984, obs 222.1937.

The epimer of 107: IR (CCl^, cm"^) 3640 (w ), 2940 (s ), 2860 (s ),

1460 (s ), 1070 (s ); ^H NMR (CDCI 3 , 300 MHz) of this compound was complex due to many impurities; m/^ calcd (M^) 222.1984, obs 222.1964. 157

(±)-(IR*,3aR*,5aS*, 8 aR*)-1.2.3,3a,4.5,5a, 8 -Octahydro-l.5.5, 6 -te tra - methylcyclopenta[c]penta1ene (95)

Pure alcohol 107^ (9.2 mg, 0.041 mmol) 1n

benzene (3mL) was heated at 50°C for 2h with

,11'' H phosphorous oxychloride (18.9 mg, 0.123 mmol)

and pyridine (0.92 ml, 1.23 irniol). The reac

tion mixture was poured into 5% aqueous hydro

chloric acid and the separated benzene layer was washed with saturated aqueous sodium bicarbonate solution and brine.

The f ilt r a t e was dried and evaporated to give6 mg of crude 9^ (74%).

Further purification by elution through a plug of silica gel with

pentane gave 3 mg of the olefin (37%); NMR (CDClg, 300 MHz) 6 5.15

(d, J = 1.4 Hz, IH ), 2.46 (ddd, ± = 2.4, 6 , and 12 Hz, IH ), 2.21 (dd, J

= 7.2 and 19.2 Hz, IH), 2.10 (s[br], IH), 2.0 (ddd, 1.2, 3.6, and 12

Hz, IH ), 1.7 - 0.9 (m, 6H), 1.57 (s,[br], 3H), 0.98 (s, 3H), 0.76 (d, J_

= 6 .6 Hz, 3H), 0.67 (s, 3H); m/z calcd (M"^) 204.1878, obs 204.1989;

calcd (M+-1) 203.1799, obs 203.1848; calcd (M*-CHj) 189.1643, obs

189.1677. 158

4,4-Dimethyl-l-(2-methy1-1.3-dioxo1an-2-y1)-6-hepten-3-o1 (112)

2-(2-Bromoethyl)-2-methyl-1,3-dioxolane (3.97

g, 20.7 mmol) and 1,2-dibromoethane (300 mg,

1.6 mmol) In tetrahydrofuran (4 ml) was added

HO dropwise to magnesium shavings (1.42 g , 59.2

mg-at). The reaction mixture was diluted with

tetrahydrofuran (12 ml) and stirred for 30 min.

A tetrahydrofuran solution (5 ml) of 2 ,2 -dimethylpent-4-enal (3.5 g,

31.1 mmol) was added and the reaction mixture was stirred for 15 h, quenched with ammonium chloride, and diluted with ether. The organic

phase was washed with sodium bicarbonate solution and brine. After

drying and evaporation, the residue (3.5 g, 78%) could be purified by mplc (19% ethyl acetate in petroleum ether), but usually crude 112 was

suitable for use in the next reaction; IR (neat, cm"^): 3450 (s), 2960

(s ), 2880 (s ), 1380 (s ), 1050 (s ); NMR (CDCI 3 , 90 MHz) 6 6.1 - 5.6

(m, IH), 5.1 - 4.9 (m, 2H), 4.0 (s. 4M), 2.3 - 1.5 (m, 6 H), 1.3 (s, 3H),

0.9 (s , 6 H); m/z calcd (M^-CgH^O) 182.1307, obs 182.1384. 159

(±)-4,4-Dimethy1-I-(2-methy1-1.3-dioxo1an-2-y1)-2-(pheny1su1fony1)- 6 - hepten-3-one.

To 2-(2-phenylsulfonylethyl)-2-methyl-1,3-dioxo-

lane^S (g.04 g, 35.5 mmol) in tetrahydrofuran

(100 mL) and tetramethyl ethyl enedi ami ne (8.2 g.

70.6 mmol) at -78°C was added tv-buty11ithiurn

(44 ml of 1.6 M in hexane, 70.6 mmol). This

mixture was transferred via syringe to 2 , 2- dimethylpent-4-enoic acid chloride (5.16 g, 35.5 mmol) in tetrahydro furan :hexamethyl phosphoric triamide (4 :1 , 100 mL) at -78°C. The reaction mixture was maintained at -78°C for 1 h, thawed to ambient temperature over 15 h, quenched with ammonium chloride solution, and extracted with ether. The organic phase was washed with brine, dried, and evaporated to give the phenylsulfonyl ketone (15.0 g, 100%); IR

(neat, cm"^): 3080 (m), 2990 (s ), 2900 (s ), 1710 (s ), 1645 (m), 1330

(s ); NMR (CCl^, 90 MHz) 6 7.6 (m, 5H), 5.9 - 5.5 (m, IH), 5.2 - 4.9

(m, 2H), 4.6 (dd, J = 8 and 3 Hz, IH), 3.9 - 3.5 (br s, 4H), 2.4 (d, J

7 Hz, 2H), 2.2 - 1.9 (m, 2H), 1.3 (s, 6 H), 1.2 (s, 3H); m/^ calcd

(M^-CgH^o) 284.0719, obs 284.0725. 160

4,4-Dimethy1-l-(2-tnethy1-1,3-dioxolan-2-y1 )-6-hepten-3-one (113).

A. Oxidation of Ketal Alcohol 112:

To crude 112 (13.7 g, 51 m ol) in dichloro-

methane (350 ml) was added pyridinium

dichromate (139 g, 0.36 mmol) and the mixture

was stirred mechanically for 48 h, diluted with

ether (2 L), and filtered through Celite. The

filtr a te was eluted through a column of s ilic a gel (75 g) and evaporated to give 10.1 g, (100%) of ketone 1 ^ ; IR

(neat, cm"^): 2980 (s ), 2880 (s ), 1730 (s ), 1650 (w ), 1270 (s ); NMR

(CDCI3 , 90 MHz) 6 6.1 - 5.6 (m, IH), 5.1 - 4.9 (m, 2H), 3.9 (s, 4H), 2.6

( t , J = 8 Hz, 2H), 2.4 (d, J^= 6 Hz, 2H), 1.9 ( t , J = 8 Hz, 2H), 1.4 (s,

3H), 1.2 (s, 6 H); m/z calcd (M+) 226.1568, obs 226.1584.

B. Reduction of Phenylsulfone:

To the phenylsulfone prepared above (15.0 g, 35.3 mmol) in methanol

(100 mL) was added disodium hydrogen phosphate (20 g) and 6% sodium mercury amalgam (62.7 g ). The reaction mixture was stirred for 3 h and filtered. The filtrate was diluted with ether and washed with water and brine. After drying and evaporation, 6.2g (78%) of 113 was isolated. 161

6 ,6 -Dimethyl-8-nonene-2,5-dione (111).

Ketoketal (6.2 g, 27.4 iranol) with

pyridinium tosylate (750 mg) in acetone (95 mL)

Q and water 5 mL) was heated at reflux for 18 h.

The reaction mixture was evaporated and the

residue was diluted with ether. The ether

solution was washed with sodium bicarbonate

solution and brine, dried, and evaporated.

Purification by mplc (10% ethyl acetate in petroleum ether) gave 111

(3.43 g, 70%) as a clear o il; IR (neat, cm"^): 3080 (w), 2980 (s ), 2920

(s ), 1720 (s ), 1700 (s ), 1640 (w), 1470 (m), 1370 (s ); NMR (CCl^, 90

MHz) 6 5.90 - 4.80 (m, 3H), 2.65 (s, 4H), 2.25 (d, = 6 Hz, 2H), 2.15

(s, 3H), 1.1 (s, 6H); m/z calcd (M+) 182.1307, obs 182.1328.

Anal. Calcd fo r C^^H^gOg: C, 72.48; H, 9.96. Found: C, 72.21; H,

10.02. 162

3 -(1 ,1 -Di methyl-3 -buteny1)-2-cyc1open ten-1-one.

Diketone Hl^ (5.0 g, 27.5 mmol) was heated at

reflux in 13% sodium hydroxide solution for 36

h. The reaction mixture was saturated with

sodium chloride solution and extracted with

ether. The organic phase was washed with

brine, dried, and evaporated to give 4.04 g

(90%) of the dienone; IR (neat, cm”^): 3080 (w),2980 (s), 2940 (s), 1720

(s ), 1650 (w), 1600 (s ), 1470 (m), 1190 (m); NMR (CCl^, 90 MHz) 6

5.85 (m, IH), 5.7 - 4.8 (m, 3H),2.55 (m, 2H), 2.4 - 2.1 (m, 4H), 1.2 (s,

6H); m/z calcd (M*) 164.1203, obs 164.1179. 163

(±)-3-(l,l-Diniethy1-3-buteny1)cyc1opentanone (110).

The dienone (2.1 g, 12.8 mol) and water (0.112

mL, 0.5 equiv) in tetrahydrofuran (20 mL) was

added dropwise to a solution of lithium (210

mg, 28 mg-at) in liquid amonia (200 mL,

distilled from sodium) at -78°C. The reaction

mixture was maintained at -78°C for 4 h, solid ammonium chloride was added, the liquid ammonia was allowed to evaporate, and ether was added. The organic phase was washed with brine, dried, and evaporated. Purification by mplc (elution with 5% ethyl acetate in petroleum ether) gave 110 (1.0 g, 50%); IR (neat, cm"^): 3080 (w ), 2980

(s ), 1740 (s ), 1640 (w), 1470 (m), 1170 (m); NMR (CCl^, 90 MHz) 6 5.9

- 4.8 (m, 3H), 2.2 - 1.8 (m, 5H), 1.95 (d, J = 6Hz, 2H), 1.3 - 1.1 (m,

2H), 1.9 (s, 6 H).

Anal. Calcd for C^iH^gO: C, 79.45; H, 10.92. Found: C, 79.18; H,

11.05. 164

(±)-(3aR *, 6 R*,6aR*)-Hexahydro-4,4, 6 -trimethyl -1(2H)-pentalenone (109).

A. Catalytic Hydrogenation of 26:

a,e-Unsaturated enone 2^ (100 mg, 0.61 m ol) in

absolute ethanol (10 mL) with 10% palladium on

carbon (30 mg) was shaken under an atmosphere

of hydrogen (50 psi) for 16 h. The reaction

mixture was filte re d and evaporated to give 109

(90 mg, 92%); IR (neat, cm‘ ^): 2880 (s ), 1740 (s ), 1470 (s ), 1170 (m) ;

NMR (CDCI3 , 300 MHz) 6 2.63 (m, IH). 2.48 (m, IH), 2.25 (m, IH), 2.15

(m, 2H), 1.85 (m, IH), 1.5 (m, 2H), 1.2 (m. IH), 0.96 (s, 3H), 0.92 (s,

3H), 0.88 (d, ^ = 7 Hz, 3H); NMR (CDCI 3 ) ppm 222.2, 54.7, 54.2,

47.8, 41.5, 40.5, 34.8, 29.5, 24.7, 23.7, 18.5; m/z calcd (M^-CH 3 )

151.1123, obs 151.1147.

Anal. Calcd for Cj^H^gO: C, 79.46; H, 10.91. Found: C, 79.45; H,

10.95,

B. Ene Reaction of 110:

Pyrex tubes containing 110 (2 tubes, 300 mg each, 3.6 mmol), were sealed under vacuum and heated at 320°C for 80 min. Purification of the combined product by mplc (elution with 5% ethyl acetate in petroleum ether) gave 1JD9 (400 mg, 67%). 165

(± )- (3aR* , 6 R*,6 aR*)-Hexahydro-2-[(Z)-hydroxymethylene]-4.4, 6 -trim e th yl - l(2H)-pentalenone.

Ketone 10^ (400 mg, 2.4 mmol) and sodium

methoxide (745 mg, 14.4 mmol) in benzene (20 mL)

ÿP OH was treated with ethyl formate (511.3 mg, 7.2

m ol). The reaction mixture was stirred for 48

h and poured into water. The benzene layer was

extracted with 1 sodium hydroxide solution and the combined aqueous layers were acidified (concentrated hydrochloric acid) and extracted with the ether. The organic phase was washed with brine, dried and evaporated to give the hydroxymethylene ketone (420 mg, 90%); IR (CCl^, cm"^): 3500 - 3000 (s ), 2980 (s ), 2940 (s ), 2880

(s ), 1750 - 1710 (s ), 1G70 (s ), 1610 (s ), 1470 (s ), 1190 (s ); NMR

(CCI4 , 90 KHz) 6 9.4 - 9.0 (m, IH ), 7.25 (m, IH ), 3.0 - 1.2 (m, 5H), 1.0

(s, 6H), 0.95 (d, J = 5 Hz, 3H); m/z calcd (M+) 194.1307, obs 194.1285. 166

(±)-(3aR*,6R*.6aR*)-2-[(Buty1thio)methy1ene]hexahydro-4.4.6-tnmethyT-

1(2H)-pentalenone (114).

A mixture of the hydroxymethylene ketone (420

mg, 2.16 mmol), magnesium sulfate (0.5 g ), and

0 sBu £-toluenesulfonic acid (75 mg) in benzene (20

mL) was treated with rv-butanethiol (0.35 mL, 3.3

mmol), heated at reflux for 15 h, and

evaporated. The residue was dissolved in ether

and the organic solution was washed with sodium bicarbonate solution and brine, dried, and evaporated. Purifica tion by mplc (elution with 5% ethyl acetate in petroleum ether) gave 390 mg (70%) of U 4 ; IR (neat, cm"^): 2960 (s ), 2930 (s ), 2870 (s ), 1690

(s ), 1580 (s ), 1460 (s ), 1200 (s ); NMR (CDCI 3 , 300 MHz) 6 7.32 (m,

IH ), 3.0 - 2.0 (m, 8H), 1.7 - 1.2 (m, 8H), 1.04 (s, 3H), 0.99 (s, 3H),

0.91 (d, 2 = 6 Hz, 3H): m/^ calcd (M*) 266.1704, obs 266.1751. 167

(± )- (3aR*, 6 S*,6 aS*)-2-[(B utylthi o)niethy1ene]hexahydro-4.4, 6 -tri methyl-6 a-

[ (E)-2-tri methyl si1y1)-2-butenyl]-l(2H)-pentalenone (115).

A solution of 11£ (300 mg, 1.17 nmol) in te tra

TMS hydrofuran (10 mL) was added during 1 h to a

cold (-78°C) solution of lithium hexamethyldisi-

lazide [from 1.3 mmol of hexamethyldisilazane

(0.27 mL) and 1.3 nmol of jv-butyllithium (0.84

mL of a 1.55 M hexane solution)] in tetrahydro furan (20 mL). Iodide 88 (594 mg, 2.6 mmol) in tetrahydrofuran (3 mL) was added and the reaction mixture was maintained at -78°C for 24 h,

poured into 5% hydrochloric acid, and extracted with ether. The ether

solution was washed with sodium bicarbonate solution and brine, dried and evaporated. Purification by mplc (elution with petroleum ether:ethyl acetate, 40:1) gave 105.1 mg (61% based on recovered 119) of 115 and

188.1 mg of 114; IR (CCl^, cm“^): 2960 (s), 2930 (s), 2870 (m), 1695 (s),

1570 (s ), 1455 (m), 1250 (s)m NMR (CDCI 3 , 300 MHz) 5 7.20 (m, IH ),

5.04 (m, IH), 2.84 (t, J = 9 Hz, 2H), 2.8 - 2.0 (m, 6H), 1.73 (d, J = 6 .8

Hz, 3H), 1.68 - 1.15 (m, 5H), 1.24 (d, J = 5.9 Hz, 3H). 1.04 (s, 3H),

1.0 - 0.8 (m, 4H), 0.92 (s, 3H), 0.03 (s, 9H); m/z calcd (M+) 392.2569, obs 392.2570. 168

(±)-(3aR*,6S*,6aS*)-Hexahydro-4,4,6-trimethy1-6a-[(E)-2-(trimethy1siTyT)'

2-buteny1]-l(2H)-pentalenone (116).

Ketone 115 (171 mg, 0.44 mmol) in 10% potassium

TMS, hydroxide solution (6 mL) and diethylene glycol

(BmL) was heated at reflux for 18 h. The

reaction mixture was poured into brine and

extracted with ether. The ether solution was

washed with brine, dried, and evaporated to provide 100 mg (77%) of 116; IR (neat, cm“^): 2970 (s ), 1730 (s ), 1600

(m), 1460 (m), 1250 (s ), 850 (s ); NMR (CDCI 3 , 300 MHz) 6 6.05 (g., J =

6.7 Hz, IH), 2.57 (d, = 13.4 Hz, IH), 2.33 (d, = 13.4 Hz, IH),

2.26 - 1.75 (m, 4H), 1.72 (d, J = 6.7 Hz, 3H), 1.63 - 1.45 (m, 2H), 1.30

- 1.15 (m, 2H), 1.03 (s, 3H), 1.01 (s , 3H), 0.95 (d, J = 6.9 Hz, 3H),

0.04 (s, 9H); m/z calcd (M*) 292.2222, obs 292.2222.

Anal. Calcd for CigHggOSi: C, 73.90; H, 11.03. Found; C, 73.88;

H, 10.99. 169

(±)-(3aR*,6S*.6aS*)-6a-[2.3-Epoxy-2-(trimethy1si1y1)buty1]hexahydro-4,4.

6- t r i methyl- 1 ( 2H)-pentalenone.

To vinylsilane 116^ (280 mg, 0.95 mmol) and solid

sodium bicarbonate (1 g) in dichloromethane (20

mL) at 0°C was added buffer washed nwchlorope-

benzoic acid (246 mg, 1.43 mmol). The reaction

mixture was stirred at ambient temperature for 15

h, a fter which the progress of the reaction was examined by tic. If the reaction was incomplete, it was treated with additional nnchloroperbenzoic acid (0.5 equiv) and this procedure was repeated if necessary. The reaction mixture was poured into sodium sulfite solution and extracted with ether. The ether layer was washed with sodium bicarbonate solution and brine, dried, and evaporated. There was isolated 190 mg (65%) of the epoxysilane; IR (neat, cm"^): 2970 (s ),

2930 (s ), 2870 (s ), 1765 (s ), 1460 (s ), 1270 (s ); NMR (CDClg/CHgClg/

90 MHz) 6 4.0 - 1.5 (m, IlH), 1.5 - 0.8 ( m, 12H), 0.10 (s, 9H); m/z calcd (M^-CgHg) 280.1859, obs 280.1834. 170

(-)-(3aR*.6S*.6aS*)-Hexahydro-4,4, 6 -trimethyl-6a-(2-oxobuty1)-l(2H)- pentalenone-(117).

The epoxysilane (190 mg, 0.62 m o l) in methanol

(10 mL) and 20% sulfuric acid (10 mL) was heated

at reflux for 15 h. The reaction mixture was

poured into water and extracted with ether. The

ether phase was washed with sodium bicarbonate

solution and brine, dried and evaporated.

Purification by mplc (elution with 12% ethyl acetate in petroleum ether) gave 1J7 (60 mg, 40%); IR (CCl^, cm"^): 2980 (s ), 2880 (s ), 1740 (s ),

1730 (s ), 1470 (s ), 1420 (m), 1390 (s ), 1130 (s ), 740 (m); NMR (CDCI 3 ,

300 MHz) 6 3.28 (d, = 17.6 Hz, IH), 2.65 (d, = 17.6 Hz, IH), 2.5

- 2.3 (m, 3H), 2.2 - 1.95 (m, 4H), 1.9 - 1.7 (m, IH), 1.58 g_, J = 5.7 Hz,

IH), 1.18 (g., J = 13.2 Hz, IH), 1.05 (s, 3H), 1.02 (s, 3H), 0.98 (t, J =

8 .8 Hz, 3H), 0.83 (d, 6 .6 Hz, 3H); m/z calcd (M+) 236.1776, obs

236.1787. 171

(± ) - ( 5aR*,8S*,8aS*)-4,5,5a.6 ,7 ,8-Hexahydro-3, 6 ,6 ,8-tetramethylcyclo-

penta[c]-penta1en-2(lH)-one. (118) .

Diketone 117^ (100 mg, 0.42 nrniol) with potassium

te r t-butoxide (100 mg, 0.84 mmol) in te tra

hydrofuran (20 mL) and te r t-butanol (0.5 mL)

was heated at 45°C for 1 h. The reaction

mixture was poured into 5% hydrochloric acid

and extracted with ether. The ether phase was washed with sodium bicarbonate solution and brine. After drying and

solvent removal, 72.1 mg (79%) of 11^ was isolated; IR (CCl^, cm"^): 2980

(s ), 1710 (s ), 1670 (s ), 1460 (s ), 1380 (m), 740 (s ); NMR (CDCI 3 , 300

MHz) 6 2.71 (d, = 17 Hz, IH), 2.4 (d, = 17 Hz, IH), 2.6 - 1.5 (m,

8H), 1.70 (d, J = 2 Hz, 3H), 1.05 (s, 6H), 0.69 (d , 0 = 7 Hz, 3H); m/z

calcd (M*) 218.1620, obs 218.1685. REFERENCES

1. a) L.A. Paquette, Topics Curr. Çhem., 79, 43 (1979); b) L.A. Raquette, Ibid. , 119, (l984)'.

2. S. Nozoe, J. Furukawa, U. Sankawa, and S. Shibata, Tetrahedron Lett., 195 (1976).

3. T. Takeachi, H. linuma, J. Iwanaya, S. Takahashi, T. Takita, and S. Umezawa, J. A ntibiotics, 22, 215 (1969).

4. See: K.E. Stevens and L.A. Paquette, Tetrahedron L e tt., 22, 4393 (1981) and the references cited thereTrû

5. L.H. Zalkow, R.N. Harris I I I , and N .I. Burke, J. Nat. Prod., 42, 96 (1979).

6 . F. Bohlmann, C. Zdero, R. Bohlmann, R.M. King, and H. Robinson, Phytochemistry. 19, 579 (1980).

7. F. Bohlmann, N. Le Van, and J. Pickardt, Chem. B er., 110, 3777 (1977).

8 . L.H. Zalkow, R.N. Harris I I I , D. Van Deveer and J.A. Bertand, J.C.S. Chem. Commun., 456 (1977).

9. R. Schmitz, A.W. Frahm, and H. Kating, Phytochemistry, 19, 1477 (1980).

10. F. Bohlmann, N. Le Van, T.V.C. Pham, J. Jacupovic, A. Schuster, V. Zobel, and W.H. Watson, Phytochemistry, 18, 1831 (1979).

11. F. Bohlmann and C. Zdero, Phytochemistry, 2^, 139 (1982).

12. F. Bohlmann and C. Zdero, Phytochemistry, 2iO, 2529 (1981).

13. L.A. Paquette and Y.K. Han, J. Org. Chem., 44, 4014 (1979); J. Am. Chem. Soc., 103, 1835 (1981).

14. W. Oppolzer, K. Belttig, and T. Hudlicky, Helv. Chim. Acta, 62, 1493 (1979); Tetrahedron, 37, 4359 (1981).

15. M.C. Pirrung, J. Am. Chem. Soc., 101. 7130 (1979); 103, 82 (1981).

16. P.A. Wender and G.B. Dreyer, Tetrahedron. 37, 4445 (1981).

172 173

17. W.G. Dauben and D.M. Walker, J. Org. Chem., 46, 1103 (1981).

18. F. Bohlmann and J. Jakupovic, Phytochemistry. 19, 259 (1980).

19. A. Leone-Bay and L.A. Paquette, J. Org. Chem., 47, 4173 (1982); L.A. Paquette and A. Leone-Bay, ^ AmTTRemT§oc7, lUS", 7352 (1983). 20. A. Marfat and P. Helquist, Tetrahedron L e tt., Ï Î Ï 7 (1978); S.A. Bal, A. Marfat and P. Helquist, J. Org."tïïëm. . 4^, 5045 (1982).

21. T. Tsunoda, M. Kodama, and S. ItO , Tetrahedron L e tt., 83 (1983).

22. F. Bohlmann, H. Suding, J. Cuatrecasas, H. Robinson, and R.M. King, Phytochemistry, 19, 2399 (1980).

23. F. Bohlmann, C. Zdero, H. Robinson, and R.M. King, Phytochemistry, n , 465 (1982). ------

24. L.A. Paquette, R.A. Roberts, and G. Drtina, J. Am. Chem. Soc., in press.

25. H. Seto and H. Yonehara, J. A ntibiotics, 92 (1980).

26. H. Seto, T. Sasaki, J. Uzawa, S. Takeuchi, and H. Yonehara, Tetrahedron L e tt., 45, 4411 (1978).

27. Y. Ohfune, H. Shirahama, and T. Matsumoto, Tetrahedron L e tt., 33, 2869 (1976).

28. 6 .D. Annis and L.A. Paquette, J. Am. Chem. Soc. 104, 4504 (1982); L.A. Paquette and G.D. Annis, TBicITT idS, 7358 (1 W 3 ).

29. K. Sakai, T. Ohtusa, S. Misumi, H. Shirahama, and T. Matsumoto, Chem. L e tt., 355 (1981).

30. F. Bohlmann and C. Zdero, Phytochemistry, 17, 1591 (1978).

31. F. Bohlmann and C. Zdero, Phytochemistry. 18, 1747 (1979).

32. J.S. Roberts, Terpenoids Steroids, 10, 104 (1981).

33. J.M. Conia and M.L. Leriverend, Bull. Soc. Chim. (France), 2981 (1970).

34. S.A. Achmad and G.W.K. C a v ill, Austr. J. Chem., 18, 1989 (1965).

35. P.E. Eaton, G.R. Carlson, and J.T. Lee, J. Org. Chem., 38, 4071 (1973). ~

36. L.A. Paquette, M.J. Wyvratt, 0. Schallner, J.L. Muthard, W.J. Begley, R.M. Blankenship, and D. Balogh, J. Org. Chem., 44, 3616 (1979); L.A. Paquette, R.J. Ternansky, O.WT Balogh, and Kentgen, J. Am. Chem. Soc., 105, 5446 (1983). 174

37. S.A. Achmad and G.W.K. C a v ill, Austr. J. Chem., 16, 858 (1963).

38. J.N. Marx and L.R. Norman, J. Org. Chem., 40, 1602 (1975).

39. (a) 0. Wallach, Liebigs Ann., 414, 233 (1917). (b) J. Wolinsky, H. Wolf, and T. Gibson. J . Org.inîem. . 28, 274 (1963). (c) J. Wolinsky and D. Chan, IbTT.T W , 41 (1 9 ^ ). (d) P. Yates, M.J. Jorgenson, and P. Singh, J. Am. Chem. Soc., 91 , 4739 (1969).

40. H.O. House and W.F. Gilmore, J. Am. Chem. Soc., 83, 3972 (1961).

41. W Reusch and P. Mattison, Tetrahedron, 23, 1953 (1967).

42. C. Black, G.L. Buchanan, and A.W. Jarvie, J. Chem. Soc. , 2971 (1956).

43. L. Ruest, G. Blouin, and P. Deslongchamps, Synth. Commun., 6 , 169 (1976).

44. A.E. Greene, J.C. M uller, and G. Ourisson, J. Org. Chem., 39, 186 (1974).

45. E.J. Corey and A.G. Hortmann, J. Am. Chem. Soc. 87, 5736 (1965).

46. P.F. King and L.A. Paquette, Synthesis, 4_, 279 (1979).

47. K.B. Sharpless, R.F. Lauer, and A.Y. Teranishi, J. Am. Chem. Soc., 95, 6137 (1973).

48. K. Miyano, Y. Ohfune, S. Azuma, and T. Matsumoto, Tetrahedron Lett., 1545 (1974).

49. T. Fex, J. Froborg, G. Magnusson, and S. Thorén, J . Org. Chem., 41, 3518 (1976).

50. P. Magnus, D.A. Quagliato, and J.C. Huffman, Organometallics, 1, 1240 (1982).

51. (a) T. Mukaiyama, K. Banno, and K. Narasaka, J. Am. Chem. Soc.. 96. 7503 (1974). (b) T. Mukaiyama and M. HayasTnTl^ëm. Lett., 15 TÎ974).

52. H.O. House, L.J. Czuba, M. G all, and H.D. Olmstead, J. Org. Chem., 2324 (1969).

53. A. Ishida and T. Mukaiyama, Chem. L e tt., 1167 (1975).

54. A.A. Ponaras, Tetrahedron L e tt.. 3105 (1976).

55. M.E. Flaugh, T.A. Crowell, and O.S. Farlow, J. Org. Chem.. 5399 (1980). 175

56. J. D'Angelo, Tetrahedron, 32, 2979 (1976).

57. R.L. Wasson and H.O. House, Org. Syn. CoTT. Vol. IV , 552.

58. H.M. Walton, J. Org. Chem., 1161 (1957).

59. J.E. McMurry, M.P. Fleming, K.L. Kees and L.R. Krepski, J. Org. Chem., 43, 3255 (1978); J.E. McMurry and M.P. Fleming, Ib id ., 41, 5 # T 197 5 ’) .

60. R. Dams, M. Malinowski, I . Westdorp and H.Y. Geise, J. Org. Chem., 4V, 248 (1982).

61. J.E. McMurry, M.G. S ilv e s tri, M.P. Fleming, T. Hoz, and M.W. Grayston, J. Org. Chem., 43, 3249 (1978).

62. G. Stork and M.E. Jung, J. Am. Chem. Soc., 96, 3682 (1974).

63. G. Stork, M.E. Jung, E. Colvin and Y. Noel, J. Am. Chem. Soc., 96, 3684 (1974).

64. M.E. Jung, Tetrahedron, 32, 3 (1976).

65. P. Beslin, R. Bloch, G. Moinet and J.M. Conia, Bull. Chim. Soc. (France)., 508 (1969); for a review see: J.M. Conia' and F. LeFerchec, Synthesis, 1 (1975).

6 6 . P. LePerchec, F. Rouessac and J.M. Conia, Bull. Soc. Chim. (France), 826 (1967).

67. P. Magnus and M.S. Nobbs, Synth. Comun., 10 , 273 (1980).

6 8. F. Walls, J. P ad illa, J. Joseph-Nathan, F. G ira l, M. Escobar and J. Romo, Tetrahedron, 22, 2387 (1966).

69. G.R. Newkome and D.L. Fishel, J. Org. Chem., 31, 677 (1966).

70. J.E. Stemke and F.T. Bond, Tetrahedron L e tt., 1815 (1975).

71. R.H. Shapiro, Org. React., 23, 405 (1976).

72. R.E. Ireland and J.A. Marshall, J. Org. Chem., 27, 1615 (1962), Ib id ., 1620 (1962).

73. R.D. Little, G.W. Muller, M.G. Venegas, G.L. Carroll, A. Bukhari, L. Patton and K. Stone, Tetrahedron, 4371 (1981).

74. T.H. Chan, W. Mychajlowskij, B.S. Ong, and D.N. Harpp, J. Org. Chem., 43, 1526 (1978).

75. T.H. Chan and B.S. Ong, J. Org. Chem., « , 2994 (1978). 176

76. G. Stork and S.D. Darling, J. Am. Chem. Soc., 1761 (1974).

77. B.D. Challard, H. Hikino, G. Lange, G. Kornis, and P. DeMayo, J. Org. Chem., 34, 794 (1969).

78. J.H. Richards and J.B. Hendrickson, Biosynthesis of Terpenes, Steroids And Acetogenins, W.A. Benjamin, Inc. (l”96T).

79. T. Tsunoda, Y. Kafaasawa, S. ltd and M. Kodama, Tetrahedron L e tt., 773 (1984).

80. (a) P.L. Creeger, J. Am. Chem. Soc., 89, 2500 (1967); (b) W.C. Agosta and A.B. Smith, I I I , J. Am. Chem. Soc., 93, 5513 (1971).

81. J. Drouin, F. Leyendecker and J.M. Conia, Tetrahedron, .36, 1203 (1980).

82. G. Moinet, J. Brocard and J. M. Conia, Tetrahedron L ett., 4461 (1972).

83. T.S. Sorensen and K. Rajeswari, J. Am. Chem. Soc., 93, 4222 (1971).

84. E.J. Corey and G. Schmidt, Tetrahedron L e t t ., 399 (1979).

85. M. Julia and B. Badet, Bull. Chim. Soc. (France), 1363 (1975).

86. K. Kondo and D. Tunemoto, Tetrahedron L e tt., 1397 (1975).

87. E. Negishi and M.J. Idacavage, Tetrahedron L e tt., 845 (1979).

88. D.A. Evans, C.L. Sims and G.C. Andrews, J. Am. Chem. Soc., 99^, 5453 (1977).

89. J.A. Marshall, G.L. Bundy and W.I. Fanta, J. Org. Chem.. 3913 (1968).

90. E. Ayanoglu, T. Gebreyesus, C.M. Beechan and C. Djerassi, Tetrahedron L e tt., 1671 (1978).

91. A.M. Birch and G. Pattenden, Tetrahedron L ett., 991 (1982).

92. E. Brown and R. Dhal, Bull. Chim. Soc. (France), 4292 (1972).