UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The total synthesis of ±-[beta]-panasinsene Story, Betty-Anne 1992

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_1992_spring_story_betty_anne.pdf [ 2.93MB ]
JSON: 831-1.0061756.json
JSON-LD: 831-1.0061756-ld.json
RDF/XML (Pretty): 831-1.0061756-rdf.xml
RDF/JSON: 831-1.0061756-rdf.json
Turtle: 831-1.0061756-turtle.txt
N-Triples: 831-1.0061756-rdf-ntriples.txt
Original Record: 831-1.0061756-source.json
Full Text

Full Text

THE TOTAL SYNTHESIS OF (±)-fi-PANASINSENEbyBTIV-ANNE STORYB.Sc., The University of Toronto, 1984M.Sc., The University of Toronto, 1986A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFDOCTOR OF PHILOSOPHYinTHE FACULTY OF GRADUATE STUDIES(Department of Chemistry)We accept this thesis as conformingto the required standardThE UNIVERSITY OF BRITISH COLUMBIAOctober 1991© Betty-Anne Story, 1991In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.Department of C/1EM(ST/The University of British ColumbiaVancouver, CanadaDate ?j’t1 /99/DE-6 (2/88)11ABSTRACTThis thesis describes a total synthesis of the sesquiterpenoid(±)-fl-panasinsene (31).Two different routes for the synthesis of a bicyclic enone ofgeneral structure 74 were investigated. An unsuccessful attempt togenerate a synthetically useful enone 74 employed the PausonKhand cyclization of an enyne 75 (R= Me; XX= S(CH2)3or MeS, pMeC6H4SO2). A second approach, which was based on the Weiss-Cookcondensation of glyoxal (44) with dimethyl 3-oxoglutarate (45) ledto the production of the dione 43, which was converted, via severalsteps, into the enone 159.The enone 1 59 was subjected to a methylenecyclohexaneannulation sequence. Thus, copper (1)-catalyzed conjugate additionof the Grignard reagent 7 to 159, followed by intramolecularalkylation of the resultant chloro keto ester, provided the tricyclicintermediate 171. Sequential reduction of the keto function in 171,deoxygenation of the resultant hydroxyl function, and hydrolysis ofthe ketal moiety gave rise to the keto ester 182. Subjection of 182to a photochemical Wolff ring contraction reaction sequenceprovided a mixture of the diesters 200 and 201. Alkylation of themixture of 200 and 201 with methyl iodide, followed by areduction-oxidation sequence, gave the dialdehyde 217. WolffKishner reduction of 21 7 resulted in the simultaneousdeoxygenation of both of the carbonyl groups and successfullycompleted the synthesis of (±)--panasinsene (31).111MeO2CMeO2C45MeR3174::44Me —75HH43BrMg7MeO2C4159H171MeO2C4182HO200 and 201 217ivTABLE OF CONTENTSPageABSTRACT iiTABLE OF CONTENTS ivLISTOFTABLES viiLIST OF FIGURES viiiABBREVIATIONS xACKNOWLEDGEMENTS xiiiI. INTRODUCTION 11.1 The Rationale 11.2 The Problem 81 .3 Angularly Fused Terpenoids 91 .4 Isolation and Structural Elucidation of /3-Panasinsene (31) 1 51.5 Previous Syntheses of f3-Panasinsene (31) 1 81.5.1 McMurry and Choy’s Synthesis of Racemic aand 13-Panasinsene (54 and 31) 181 .5.2 Johnson and Meanwell’s Synthesis of (-)-f3-Panasinsene and Its Enantiomer ((-)- and(+)-31) 20II. DISCUSSION 22Total Synthesis of (±)-J3-Panasinsene (31)2.1 Retrosynthetic Analysis 22V2.2 An Approach to the Synthesis of (±)-J3-Panasinsene (31) via the Use of the PausonKhand Reaction 262.2.1 Background 262.2.2 Approach to the Synthesis of (±)-J3-Panasinsene (31) 322.2.2.1 The 1,3-Dithianyl Function at C-7 352.2.2.2 The Methylthio-p-toluenesulfonyl Functionat C-7 442.3 The Synthesis of (±)-/3-Panasinsene (31)via the Weiss-Cook Condensation Reaction 562.3.1 Background 562.3.2 Application of the Weiss-Cook CondensationReaction to the Synthesis of (±)-J3-Panasinsene (31) 602.3.2.1 Preparation of an Enone 74 602.3.2.2 Methylenecyclohexane Annulation on theEnonel59 712.3.2.3 Preparation of a Substrate (Intermediate 72)for Ring Contraction 822.3.2.4 Ring Contraction to Give a 4-5-6 TricyclicCarbon Skeleton 1 Preparation of (±)-/-Panasinsene (31)via the Diacetate 213 11 Preparation of (±)-J3-Panasinsene (31)via a Woiff-Kishner Reduction 135III. CONCLUSION 1 49viIV. EXPERIMENTAL 1514.1 General 1514.2 Experimental Procedures for the Synthesisof (±)-13-Panasinsene (31) via the Weiss-CookCondensation Approach 1 56V. REFERENCES 191viiLIST OF TABLESTable Page1 The 75 MHz HETCOR Data for the KetoEster Ketal 171 782 The 400 MHz COSY Data for the KetoEster Ketal 171 813 The 400 MHz COSY Data for the Keto Ester 182 994 The 125 MHz HETCOR Data for the Diester 200 1115 The 400 MHz COSY Data for the Diester 200 11 66 The 400 MHz COSY Data for the Diester 202 1 237 The 400 MHz COSY Data for the Diacetate 213 1338 The 400 MHz NOESY Data for the Diacetate 213 1 349 A Comparison of the Spectral Data for Authentic andSynthetic J.3-Panasinsene (31) 1 391 0 The 125 MHz HETCOR Data for Synthetic(±)q3-Panasinsene (31) 14411 The 400 MHz COSY Data for Synthetic(±)-fl-Panasinsene (31) 147vii’LIST of FIGURESFigure Page1 The 75 MHz broad band decoupled 13C nmr spectrumof the enone diastereomer A 522 The 300 MHz 1H nmr spectrum of the ketoester ketal 171 753 The 75 MHz HETCOR spectrum of the ketoester ketal 171 774 The 400 MHz COSY spectrum of the ketoester ketal 171 805 The 300 MHz 1H nmr spectrum of the keto ester 182 966 The 400 MHz COSY spectrum of the keto ester 182 987 The 125 MHz HETCOR spectrum of the diester 200 1108 The 400 MHz 1H nmr spectrum of the diester 200 1129 The 400 MHz COSY spectrum of the diester 200 11 510 The 400 MHz 1H nmr spectrum of the diester 202 1191 1 The 400 MHz COSY spectrum of the diester 202 1 221 2 The 400 MHz ‘H nmr spectrum of the diacetate 213 1301 3 The 400 MHz COSY spectrum of the diacetate 213 1321 4 The 75 MHz broad band decoupled 13C nmr spectrumof synthetic (±)-J3-panasinsene (31) 1411 5 The 400 MHz 1H nmr spectrum of synthetic(±)-f3-panasinsene (31) 142ix1 6 The 125 MHz HETCOR spectrum of synthetic(±)-fl-panasinsene (31) 1431 7 The 400 MHz COSY spectrum of synthetic(±)-fl-panasinsene (31) 146xLIST OF ABBREVIATIONSAc AcetylAIBN 2,2’-azobisisobutyro nitrileAPT attached proton testaq aqueousbr broadb.p. boiling pointn-Bu normal-butylt-Bu tertiary-butylcat catalyst, catalyticcOSY correlation spectroscopyd doubletCR3 diethylene glycolDMAP 4-(N, N-dimethylamino)pyridineEXvE 1 ,2-dimethoxyethaneDMSO dimethyl sulfoxideE- electrophileequiv equivalentsEt ethylg gramsglc gas-liquid chromatographyh hour(s)HETOR heteronuclear correlation spectroscopyHMPA hexamethylphosphoramideHz hertzi r infraredxiLAH lithium aluminum hydrideLDA lithium diisopropylamidelit, literaturem multipletM molarMe methylmg milligram(s)MHz megahertzmm minute(s)mmol millimole(s)mol mole(s)m.p. melting pointMs methanesulfonylMS mass spectrum (low resolution)m/z mass to charge ratioN normal1’lvD N-methylmorpholine N-oxidenmr nuclear magnetic resonancenOe nuclear Overhauser enhancementNOESY nuclear Overhauser enhancement spectroscopyNu: nucleophilep pagePC pyridinium chlorochromatePh phenylppm parts per millionPTC phenoxythiocarbonylPyr pyridinexliq quartetquant quantitativerel. i. relative intensityrt room temperatures singlet (nmr); strong (ir)t tripletTBDMS tertiary- b u t y I d I m e t h yl s ii y ITf trifluoromethanesulfonylTHE tetrahydrofuranTHP tetrahydropyranylTIPS triisopropylsilylt I c thin layer chromatographyTMEDA N,N,N’, N’-tetramethylethylenediamineTMS tetramethylsilaneTMS- trimethylsilylp-Tol para-tolylp-Ts para-toluenesulfonylv veryw weakxliiACKNOWLEDGEMENTSThis thesis would have been impossible were it not for theinput, in various different ways, of many people.I am deeply indebted to my research supervisor, ProfessorEdward Piers, for his guidance throughout my research. Hiscommitment to excellence is appreciated, as is his patience with myefforts in the laboratory. The members of the Piers research groupduring my sojourn here provided enriching friendships and manyfruitful discussions. Veljko Dragojlovic’s loan of his computergreatly facilitated the writing of this thesis. Many thanks to Ms.Johanne Renaud, Ms. Renata Oballa, Mr. Todd Schindeler, Dr. GuyPlourde and Mr. Jacques Roberge for proof-reading this manuscript.The able technical assistance of the staff of the nmr, massspectrometry and elemental analysis facilities was critical to thesuccess of this research. Professor Scheffer’s research group kindlyprovided the photolysis equipment for some of the experiments.The encouragement and prayers of my parents, the D. Lloyds,Mrs. Todd and Miss Standerwick, among others, saw me through thegood and bad times and reminded me of the One who is the source ofall knowledge.I have appreciated the financial independence and freedom todo research provided by an NSERC postgraduate scholarship.1I. INTRODUCTION1.1. The Rationale.The synthesis of biologically useful, theoretióally interesting,and stereochemically or structurally challenging molecules hasprovided the basis for much of the research in synthetic organicchemistry. Presently, the trend seems to be towards the study of“substances and reactions relevant to life”1 which means thatinhibitors for important enzymes or receptors2 are becoming keytarget molecules. The enantiomeric purity of a biologically activeproduct that has been made synthetically in a laboratory can havemajor implications in biological systems especially if oneenantiomer exhibits undesirable activity. Though in recent yearsthere has been a greater emphasis placed on the synthesis ofenantiomerically pure compounds as compared to the synthesis ofracemic mixtures, the synthesis of racemic mixtures is stillinformative. However, the successful synthesis of any molecule, ina racemic or an enantiomerically pure form, depends largely on thechemist’s ability to analyze the target in a logical retrosynthetic*manner,3 and to develop a feasible synthetic route for itspreparation. Despite careful retrosynthetic analysis, the actualimplementation of any route may be unsuccessful due either to the* Retrosynthetic (or antithetic) analysis is a technique that has been developed in orderto transform the structure of the synthetic target into a logical sequence ofprogressively simpler structures which finally leads to simple or commerciallyavailable chemicals. Each step in the retrosynthetic direction (a transform)corresponds to a chemical reaction in the synthetic direction.32failure of a particular reaction or due to the lack of procedures forcarrying out the desired transformation. Consequently, alternativeroutes must be envisaged in advance, particularly for reactions thatmay be synthetically challenging. In cases where theimplementation of a route has failed, other options must be exploredor new methods must be found in order to circumvent theobstruction.One of the currently fruitful areas of research with regard todeveloping new reactions is organometallic chemistry. From anorganic chemist’s point of view, an important test of the utility of aparticular organometallic reagent is its applicability to the solutionof a given synthetic organic chemical problem.The use of organotin reagents in organic synthesis has beenexplored by many research groups including our own.4 In 1983, itwas first reported that to—substituted 1-alkynes (1) undergo aregioselective reaction with (trimethylstannyl)copper (1)-dimethylsulfide complex (2) in the presence of 60 equivalents of methanol(THE, -63°C) to give the corresponding 2-trimethylstannyl-1-alkenes (3) in a good yields (equation l-1). Functional groupstolerated in the reaction are halides, hydroxyls, and trialkylsilyl ortetrahydropyranyl ethers. Subsequent research has demonstratedthat the 2-trimethylstannyl-1-alkene reagents 46 and 57 can bytransmetallation with methyllithium be converted into thecorresponding lithio species, which can then be transformed into thecorresponding Grignard or organocopper reagents. For example, 5-chloro-2-trirnethylstannyf-1-pentene (5) was reacted withmethyllithium (THE, -78°C, 15 mm) to make the vinyllithium species3H (CH)X__________Me3SnCuSMe2(2) (2 equiv) 2- 1H (CH)X—2 n MeOH (60 equiv), THE,1 -63°C (12h) H SnMe3n=1-4X= CI, OH, OTHP, OTBDMS===(“ci ===(\—_ ciSnMe3 SnMe34 56. Treatment of 6 with anhydrous magnesium bromide produced theGrignard reagent 7, which was found to undergo a copper (1)-catalyzed (CuBrSMe2, 0.25 equiv) conjugate addition reaction toenones of general structure 8. In the cases of enones with atrisubstituted double bond, (e.g. 8, R or R’=Me), boron trifluorideetherate was added as an additional catalyst to improve theefficiency of the conjugate addition. The chloro ketones 9 werecyclized with potassium hydride (THE, room temperature, 2 hours) togive exclusively the cis-fused bicyclic products (1 0) for caseswhere subsequent equilibration was impossible (R=Me). Ifequilibration of the bicyclic product was possible (R=H), thenvarying amounts of the trans-fused bicyclic ketone (11) were alsoobtained as the minor component of the product mixture (Scheme I1). The overall process described above demonstrates theutilization of the chloro vinylstannane 5 as a bifunctionalconjunctive reagent* in which the two reactive sites have beenselectively, sequentially deployed. In these reactions, 5 serves as* A bifunctional conjunctive reagent is a reagent with two reactive sites which isincorporated in whole or in part into a substrate molecule to increase its structuralcomplexity.64the synthetic equivalent of the 1 -pentene donor2,acceptor5-syn hon(d2,a5-synthon) 12. The overall result of the sequence of reactions,in which 8 is converted into 10 and/or 11, is the annulation of amethylenecyclohexane unit onto a cyclic enone.CIMeLi, THE.,Li .MgBr-78°C (15mm)6 71) CuBrSMe2(0.25 equiv)2) RR’>3) (BFOEt2010Scheme I-i12The methylenecyclohexane annulation process is an importanttool in natural products synthesis because the methylenecyclohexane5MgBr2, THF,HR1195ring and derivatives thereof (part structures 13, 14, 1 6-1 9,Scheme 1-2) are common in terpenoid natural products. Thus, theolefinic function in the methylenecyclohexane annulated product(13) may be hydrogenated to give the corresponding methyl grouppresent in 1 4 or it may be cyclopropanated to give 1 5. Thecyclopropyl ring in 15 may then be hydrogenolyzed to give the gemdimethyl group found in 16. An acid-catalyzed double bondisomerization in 13 would provide 17. Alternatively, the methylenegroup may be cleaved to give the ketone 18, or hydroxylated to give19.Scheme 1-2The annulation procedure has been used in the syntheses ofnatural products with the bicyclic axane9 and clerodaneskeletons10’1 and has also been employed in the total synthesis ofthe sesterterpenoid (±)-paIauolide2 (Scheme 1-3). Each of these14 13 15/1617 196syntheses involved the annulation of the methylenecyclohexane unitonto a cyclic enone with a trisubstituted double bond. Thus, in thesynthesis of the axane skeleton, conjugate addition to thecyclopentenone 20 of the Grignard reagent 7 (derived from thevinylstannane 5) was catalyzed by copper (I) bromide-dimethylsulfide complex and boron trifluoride etherate. The resultant chloroketone was cyclized as described earlier to give only the cis-fusedbicyclic ketone 21, which was then transformed via a series ofreactions into (±)-axamide-1 (22) and (±)-axisonitrile-1 (23). In asimilar manner, conjugate addition of the Grignard reagent 7 to thecyclohexenone 24, followed by cyclization of the resultantchloroketone, gave a mixture of the cis- and trans-fused bicyclicketones in which the cis-fused product predominated. The mixturewas converted by equilibration (KH/EtOH) to another mixture inwhich the trans-fused product (25) was the major component.Subsequent reactions converted the bicyclic ketone 25 into (±)-stephalic acid (26). A different cyclohexenone (27) was used forthe syntheses of the clerodane (±)-isolinaridiol diacetate (29) andthe sesterterpenoid (±)-palauolide (30), but a similar conjugateaddition! cyclization sequence led to the formation of a mixture ofthe cis- and trans-fused bicyclic ketones in which the cis-fusedcompound predominated. Equilibration (t-BuOK/t-BuOH) of themixture led to the formation of another mixture in which the transfused product (28) was the major component. The bicyclic ketone28 then was converted into (±)-isolinaridiol diacetate (29) and (±)-palauolide (30).71) 7, CuBrSMe2BF3OEt2,THE-78°CKH, THE; EtOH,0 1) 7, CuBrSMe25.,MeBF30Et,THE2) t-BuOK,Met-BuOH(C) A Sesterterpene12R=(±)-Palaulolide(a) Axanes90Mee2THE, -78°C2) BF3OEt20 3)KH,THF2122 R=NHCHO(b) Clerodanes10”(±)-Axamide-123 R=NC (±)-Axisonitrile-10heat24 2 5 26 (±)-Stephalic Acid27OAc28 29 R= J,0Ac(±)-Isolinaridioldiacetate27 28 30Scheme 1-381.2. The Problem.______+ (1-2)The angularly fused tricyclic sesquiterpene (-)-f3-panasinsene(31), isolated by Yoshihara and Hirose in 1975,13 attracted ourattention due to (a) its novel structure, and (b) the presence of amethylenecyclohexane ring cis-fused to a substituted bicyclo[3.2.O]heptane unit. It seemed reasonable to assume that themethylenecyclohexane annulation protocol (vide supra, pp. 2-4)utilizing the copper (1)-catalyzed conjugate addition of the Grignardreagent 7 could be extended from the use of monocyclic enonesubstrates having a trisubstituted double bond (20, 24, 27) to theuse of a bicyclic enone substrate having a tetrasubstituted doublebond. If the methylenecyclohexane unit present in (-)-j3-panasinsene(31) is disconnected and the remaining bicyclic portion is suitablyfunctionalized, an enone such as 32 with a tetrasubstituted doublebond results (equation 1-2). Preparation of an enone similar to 32,or its synthetic equivalent, then would be a key part of a possiblesynthesis of (±)-/3-panasinsene (31).319The strategy we wished to employ in the synthesis of (±)--panasinsene (31), which would have the methylenecyclohexaneannulation as a key sequence, differs significantly from those of thetwo previously reported approaches (vide infra, p. 18).1.3. Angularly Fused Terpenoids.The carbon skeleton of (-)-f.3-panasinsene (31) is an example ofone of a variety of angularly fused skeletons found in terpenoidnatural products. A few of the ring size combinations found interpenoids are depicted in Scheme 1-4. The unifying feature of theskeletons is the presence of a bridged spirane arrangement ofrings14 such that three variously sized carbocyclic rings share acommon carbon atom.The synthesis of angularly fused terpenoids has been approachedby a number of different methods which fall into three maincategories. In effect, the strategies may be divided according towhether a monocyclic substrate is utilized for a two ringannulation, a bicyclic substrate is employed for a one ringannulation, or a monocyclic substrate is subjected to two sequentialone ring annulations. The approach chosen depends, of course, onhow the retrosynthetic analysis was performed on the target,particularly with respect to which rings were found to be strategic• for preservation and which were found to be strategic fordisconnection. General factors to be considered in the antitheticalanalysis include the sizes of the rings, the connectivities and10stereorelationships between the rings, the functional groups presentand the availability of suitable precursors.337 Retigeranic Acid195-5-53-5-6 4-5-6 4-5-633 Cycloeudesmol’5 31 13-Panasinsene34-5-634 Panasinsanol B1635 Perforatone7 36 PentaIenene85-5-5 5-6-738 Gascardic Acid20CO2HScheme 1-411In the first synthetic approach, a tricyclic framework isassembled in one step from a monocyclic precursor having anappropriately substituted side chain (or side chains) which canundergo a cycloaddition reaction such as a carbene insertion, a 2÷2cycloaddition, or a 2+2+1 cycloaddition. Two examples of thisapproach involve the syntheses of the terpenoids pentalenene 36 andretigeranic acid 37. The key step in Schore and Rowley’ssynthesis21 of (±)-pentalenene (36) was the octacarbonyldicobaltcatalyzed 2+2+1 cyclization of the enyne 39 (Pauson-Khandcyclization,22 vide infra, p. 26) to make the angularly fused tricyclicenone 40. The enone 40 then was converted into 36 by a series ofchemical reactions (equation 1-3).(1-3)In Corey’s synthesis23 of (±)-retigeranic acid (37) the key stepin assembling the angularly fused triquinane portion of the targetwas a 2+2 cycloaddition reaction. Thus, the carboxylic acid functionin 41 was converted into the corresponding ketene which underwentan intramolecular 2+2 cyclization to give 42. Ring expansion of thefour membered ring, ring contraction of the six membered ring, andsuitable functional group manipulations served to convert 42 intothe racemic natural product 37 (equation 1-4). Corey’s synthesis of39 40 361237 illustrates the fact that rings of sizes differing from those inthe final product may be assembled initially due to the existence ofa convenient route for their preparation and then, provided theappropriate methods exist, suitable ring expansions/contractionsmay be employed to create the desired ring size. The approach inwhich a two ring annulation onto a monocyclic substrate isperformed was also used for the previous syntheses of -panasinsene (vide infra, p. 18).(1-4)...IIIIA second approach to the synthesis of angularly fusedterpenoids involves the annulation of a bicyclic substrate to producethe required tricyclic skeleton. An example of this approach is asynthesis of pentalenene (36) different from that discussed above.Thus, the key skeleton-assembling step in the synthesis of (±)-pentalenene by Piers and Karunaratne24 was a methylenecyclopentane annulation reaction on the tricyclic enone 47. The enone 47was derived from the dione 43, which, in turn, was prepared fromCO2H3713H:>__+MeQ2C Jls.CQ2Meh10base (I -5)acyclic precursors via the Weiss-Cook condensation25 of glyoxal(44) with the keto diester 45 (equation 1-5, and vide infra, p. 55).Monoketalization of the dione 43 yielded 46, which was subjectedto a series of chemical manipulations to furnish the enone 47. Theenone intermediate 47 underwent the methylenecyclopentaneannulation to provide the tetracyclic ketone (48) which wastransformed by standard means into (±)-pentalenene (36) (Scheme I-5). The above described approach, in which the third ring isannulated onto a bicyclic substrate to make a tricyclic skeleton, isthe one we chose to use for the synthesis of j3-panasinsene (31).The sequential assembly of the rings making up the tricyclicframework is a third approach that has been employed to achieve thesynthesis of angularly fused terpenoids. The Boeckman synthesis2°of gascardic acid (38) exemplifies this third approach. A one-potannulation of the six-membered ring onto the enone 20 wasperformed via conjugate addition of cuprate reagent 49, trapping ofthe resultant enolate anion with the enone 50, and a subsequentbase-catalyzed aldol condensation to provide the key intermediate51 (Scheme 1-6). Introduction of a functionalized two carbon unitat the a-carbon of the a,J3-unsaturated ketone function in 51 was akey objective in their synthesis. However, compound 51 proved to14t’CIçMgBr1), CuBrSMe2THF, -78°C2) KH, THEbe quite unreactive towards conjugate additions reactions.Therefore, compound 51 was converted into the vinyl ether 52which subjected to a Claisen rearrangement to provide 53.Intermediate 53 then was transformed into gascardic acid (38) viaa series of standard chemical reactions which included anintramolecular cyclization to form the 7-membered ring. It is ofinterest to note (for future reference) that steric congestioncontributed to the lack of reactivity towards conjugate additionreactions of the bicyclic enone 51.The examples chosen above, while illustrative of the differentapproaches to the synthesis of angularly fused terpenoids, also werechosen to demonstrate applications of reactions or strategiesH 4746H36Scheme 1-54815studied as a part of the research reported herein. Thus, the PausonKhand cyclization, the Weiss-Cook condensation, a ring contraction,and a conjugate addition to a tetrasubstituted enone will bediscussed in greater detail at suitable points in the thesis.1) ‘——Cu Li, -78°C;0_fl’’ -78° C .-20°C2 ) TMS (CH2)4CH(OCH320 o50-20°C3) Base, MeOHs-collidine,160°CCHO253Scheme 1-61.4. Isolation and Structural Elucidation of f3-Panasinsene(31).In 1975 Yoshihara and Hirose’3 reported their isolation of thenew sesquiterpene hydrocarbon, J3-panasinsene (31), from the51 OMeIj‘4‘438 5216neutral portion of the volatile oil extract of the roots and rootletsof both fresh and commercial dried ginseng (Panax ginseng C.A.Meyer) from Japan and commercial dried ginseng rootlets fromKorea.The molecular weight of 31 was found to be 204. Signals forthree tertiary methyl groups and an exocyclic methylene group werefound in the ‘H nmr spectrum at 3 0.74 (s, 3H), 0.86 (s, 3H), 1.08 (s,3H), 4.78 (d, 1H, J = 2 Hz) and 4.84 (d, 1H, J = 2 Hz). Absorptions at1365 and 1360 cm’ in the ir spectrum indicated that two of themethyl groups were geminal. Catalytic hydrogenation of 31 gavetwo dihydro derivatives which were identical with two derivativesobtained by the catalytic hydrogenation of the endocyclic doublebond isomer, a-panasinsene (54) (also isolated at the same time).Further confirmation of the structure of J3-panasinsene (31) wasobtained by ozonolysis of 31 to give the known ketone (55), anintermediate in Parker’s synthesis of neoclovene (56).26 It wasfound too, that 31, when treated with concentrated sulfuric acid indiethyl ether, rearranged to give a-panasinsene (54), a-neoclovene(56), and j3-neoclovene (57). It was known that caryophyllene (58)is isomerized to cx-neoclovene (56) upon treatment withconcentrated sulfuric acid and Parker had postulated a cation withthe panasinsene framework as an intermediate.27 Therefore, basedon the spectroscopic evidence and the chemical behavior of thecompound, the structure of J3-panasinsene (31) was established.The roots and rootlets of ginseng (Jen-shen) have been used inChinese medicine for centuries, mention of jen-shen having beenmade in Chinese pharmaceutical works with traditions dating to the17later Han period.* 28 Ginseng is still claimed to be a wonder remedywith anti-fatigue, anti-diabetic, anti-stress, and central nervoussystem stimulant and sedative properties. However, whichcomponents are responsible for which properties is still beinginvestigated. As yet, the biological activity of 13-panasinsene (31)is unknown.29* The Han dynasty was in power 206 B.C.-250 A.D. in China. The entry25 for jen-shen,purported to date to about that time is: “taste: sweet; [thermoinfluence:] slightly cold.Controls the filling of the five depots. Pacifies the spirit; fixes the hun- and p’o souls.Ends fright and agitation. Expels evil influences. Clears the eyes. Opens the heart andbenefits one’s wisdom. Consumed over a long time, it takes the material weight from thebody and extends one’s years of life. Other names are ‘man’s bit’ and ‘demon’s cover’.”31 545556 57 58181.5. Previous Syntheses of J3-Panasinsene (31).J3-Panasinsene (31) has been synthesized twice previously, byMcMurry and Choy in 198O° and by Johnson and Meanwell in 1981.31Both syntheses, as mentioned earlier, made use of the same generalapproach. Thus, the two research groups introduced an unsaturatedside chain onto a cyclohexanone and performed a photochemical 2+2cycloaddition reaction to assemble the tricyclic carbon framework.The details of the syntheses differ as outlined in the followingdescription of their routes.1.5.1. McMurry and Choy’s Synthesis of Racemic a- and $-Panasinsene(54 and 31).The McMurry and Choy synthesis3°of the panasinsenes (54 and31) commenced with the alkylation of the sodium enolate of 2-methylcyclohexanone (59) with 1 -bromo-4-methyl-3-pentene (60)to give 61 (Scheme 1-7). Compound 61 was treated with the lithiosalt of dimethyl phenylthiomethyiphosphonate (62) to give a 9:1mixture of vinyl sulfides which were oxidized with Nal04 to thecorresponding sulfoxides (63). Deconjugation of the sulfoxides withdimsyl potassium provided the allyl sulfoxides 64. Treatment of 64with trimethylphosphite provided a mixture of the key dienealcohols (65). Photolysis of 65 in diethyl ether in the presence ofcopper (I) triflate gave the 2+2 cycloaddition product (66) as amixture of epimeric alcohols which were oxidized to the ketone(55). Ketone 55 was reacted with methyllithium and the resulting19alcohols were dehydrated to provide synthetic racemic a- and 1-panasinsene (54 and 31).2) Na1046165 64hv,CuOTfHPcc,66S(Q)Ph— —KCH2SOC3631) MeLi2) SOC,/’+5:25531Scheme 1-754201.5.2. Johnson and Meanwell’s Synthesis of (-)-J3-Panasinsene andIts Enantiomer ((-)- and (+)-31).Johnson and Meanwell began their synthesis31 of (-)-J3-panasinsene ((-)-31) with the copper (1)-catalyzed conjugateaddition of 4-methyl-3-pentenylmagnesium bromide (67) to 3-methyl-2-cyclohexen-1-one (24) (Scheme 1-8). The enolate anionintermediate was trapped with formaldehyde and the resultantmixture of keto alcohols was converted to the corresponding ketotosylate mixture. The tosylates were subjected to a base-catalyzedelimination reaction to provide the enone 68. Photolysis of 68 inpentane provided racemic 55 by means of a 2+2 cycloaddition of thealkene functional group and the double bond of the enone. Resolutionof the enantiomers occurred by reaction of the carbonyl function in55 with (S)-(N-methylphenylsulfonimidoyl)methyllithium 69 toprovide a mixture of diastereomers (70 and 71). The diastereomerswere transformed separately into (+)- and (-)-/3-panasinsene ((+)-31and (-)-31), respectively, by treatment with aluminum amalgam andacetic acid in wet THE.21MgBr1) CuBrSMe2 (5 mol%)2) gaseous CH2O3) p-TsCI, Pyr4) t-BuOK, t-BuOH24NMeII68hv, pentane(69)5570AI/Hg,HFAl/Hg,71c(+)-31(-)-3 1Scheme 1-822II. DISCUSSIONTotal Synthesis of (±)-j3-Panasinsene 31.2.1. Retrosynthetic Analysis.Our retrosynthetic analysis of (±)-J3-panasinsene 31 was guidedby two main strategies, namely: (a) the application of themethylenecyclohexane annulation transform described earlier (pp. 2-4) and (b) the utilization of the bicyclic enone 74 as a keyintermediate. The bicyclic enone 74 represents an important branchpoint in the analysis since it may be derived from a variety ofprecursors. The exact structure of the enone (i.e., the nature of Rand XX) would depend on the route chosen for its synthesis.Suitable retrosynthetic functionalization of (±)-J3-panasi nsene(31),* followed by a carbon-carbon bond disconnection and the Wolffrearrangement transform would “convert” 31 into the tricyclic ring-expanded ketone 72 (Scheme D-1). The R group in ketone 72 wasexpected to be either a methyl group as is present in the naturalproduct or a methoxycarbonyl moiety. There is ample precedent forthe synthetic conversion of a methoxycarbonyl function into amethyl group.32Retrosynthetic functional group manipulations and a functionalgroup introduction would transform 72 into the ketone 73, in whichXX is a carbonyl equivalent. Application of the methylenecyclo* The numbering scheme utilized for (±)--panasinsene (31) is analagous to the oneused by Iwabuchi and coworkers16 for panasinsanols A and B.23M 151412________:QMelecjiL1,SnMe3+5 74Scheme D-1hexane annulation transform to 73 would yield two fragments, thevinylstannane 5 and the key bicyclic enone 74 (Scheme D-1). It iswell documented in the literature that conjugate additions ofcuprate reagents to bicyclo[3.3.O]oct-1-en-3-ones occur on theconvex face of the enone to give cis-fused diquinanes.33 Thepreference for the formation of cis-fused diquinanes is probably dueto (a) the fact that trans-fused diquinanes are more strained thanthe cis-fused isomers34 and (b) the likelihood that the transitionstate for the reaction has some product-like character.3Consequently, the transition state leading to the formation of thecis-fused product will be of lower energy than that leading to thetrans-fused adduct, and the cis-fused diquinane will be formed31 72RH7324preferentially. As outlined in the Introduction (pp. 2-4), cyclizationof the keto chloride intermediate in the methylenecyclohexaneannulation sequence gives the cis-fused annulated product whenfurther equilibration is impossible. Thus, provided that the R groupwas already installed on the enone double bond, the methylenecyclohexane annulation would be expected to give rise to a product withthe desired relative stereochemistry at the three chiral centers (C1, C-4 and C-7) of /3-panasinsene (31).Retrosynthetic analysis of the important enone 74 wasapproached in two different ways (Scheme D-2). In the case where Ris a methyl group, utilization of the Pauson-Khand cyclizationtransform would lead to the linear enyne 75. The enyne 75 can bedisconnected retrosynthetically in a number of ways. For example,disconnection of both bonds a- to the C=XX function (bonds b and b’)in enyne 75 would produce fragments which may be envisaged as 76,77 and 78 or 79. In theory, the enyne 75 could be assembledsynthetically by the sequential alkylation of an anion derived from78 or 79 with the alkylating agents 76 and 77. Utilization of 78 or79 to prepare the enyne 75 would mean that, at an appropriate stagein the synthesis, it would be necessary to transform the dithioketalderived functions (i.e., XX=1,3-dithianyl or XX=MeS, S02p-Tol) into acarbonyl function (XX=O) in order to provide the tricyclic ketone 72desired for the Wolff rearrangement reaction. The proposedalkylations of 78 and 79 followed by the hydrolysis of thedithioketal derived functions to a carbonyl group are consistent withthe previously reported use of 7836 and 7937 as masked acyl anionequivalents.25In the case where R is a methoxycarbonyl group, retrosyntheticremoval of the double bond and a disconnection of the carbonmethoxycarbonyl bond (bond a) in the enone 74 would provide theknown keto ketal 46.24,38 In the synthetic direction, the reactionsare a methoxycarbonylation and a dehydrogenation, respectively,both of which are known processes.7476 77+ S4%%%_•S78orMeS\,,SO2P-ToI79HH46Scheme D-2262.2. An Approach to the Synthesis of (±)-f3-Panasinsene (31)via the Use of the Pauson-Khand Reaction.2.2.1. Background.The Pauson-Khand reaction is a formal 2+2+1 cycloadditionreaction of a hexacarbonyldicobalt alkyne complex with an alkene.During the reaction, one of the carbon monoxide ligands of thecomplex is used, and the product generated is a substitutedcyclopentenone.22’39 An example using the generalized alkyne 80and ethylene is illustrative (equation D-1). The alkyne reacts withoctacarbonyldicobalt 81 to give the hexacarbonyldicobalt complex82. Heating the complex with ethylene produces the cyclopentenone83.II + Co2(CO)8 HC CH: ‘° (D-1)\ heatH8 0 8 1 82 (Dashes=CO)The intramolecular version of the Pauson-Khand reaction wasfirst reported by Croudace and Schore in 1981.° They found thatcyclization of hept-1-en-6-yne (84) at 95°C (4 days) producedbicyclo[3.3.O]oct-1-en-3-one (85) in 31% yield (equation D-2). Incontrast, the attempted cyclization of hex-1-en-5-yne (86) to give8327H___-_\+ Co2(CQ)8 CC, 95°C, (D-2)(Me)3CCH2H(Me)84(4days) 85H+ Co2(CO)8 (D-3)bicyclo[3.2.O]hept-1-en-3-one (87) yielded only products fromalkyne trimerization (equation 0-3).Various substituents may be tolerated on the alkyne and alkenefunctional groups. However, electron-withdrawing groups on theolefin (CHO, COR, CO2R, or CN)41 or on the alkyne moiety (CO2R,39bCH(OEt)242)are detrimental to the cyclization due to the formationof dienes via a hydrogen migration. This is illustrated by thecyclization of the enyne 88 to give the diene 89 (equation D-4).42The presence of electron-donating groups on the alkene moiety(OC(O)R or OR)43 or on the alkyne function (OR44 or SR45) may bebeneficial.EtC OEtTBDMSCMe+ Co2(00)8 CH2,rt, (D-4)Me 8h (83%) MeNMO=N-methylmorpholine N-oxide 8 9TBDMSC8828The Pauson-Khand cyclization reactions of enynes to givebicyclo[3.3.O]oct-1 -en-3-ones have been studied to determine theeffects on the cyclization of various substituents on the carbonchain linking the alkene and alkyne functions. The yields, reactiontimes and diastereoselectivities were affected. Some of theresults, relevant to our attempted synthesis of (±)-/3-panasinsene(31), are presented in the following discussion. It may be noted thatC-7 refers to the position on the carbon chain which becomes C-7 inthe bicyclo[3.3.O]oct-1-en-3-one produced in the reaction (seeequation D-2).The yields of cyclizations of enynes with substituents on thecarbon chain linking the alkene and alkyne functions are generallybetter than those of the less substituted cases, and the reactiontimes are usually shorter. Thus, for example, cyclization of the.R Co2(CO)8,CC, (D-5)heat90 (R=H) 90a (14%)91 (R=Me) 91a (78%)enyne 90 occurred in only 14% yield, while cyclization of the enyne91 with the additional gem-dimethyl substitution, occurred in 78%yield (equation D-5).46 In addition, the cyclization of the enyne 91(equation D-5) was complete in 20 hours, a decrease in reactiontime when compared with the cyclization of the enyne 84 (4 days)(equation D-2). Good yields are also obtained with enynes in whichTBL29the carbon chain between the alkene and alkyne functional groups issubstituted only in the homopropargylic (C-7) position. Cyclizationof the enyne 92 to give a mixture of 92a and 92b in 86% yield after20 hours is a pertinent example (equation D-6).47 The improvedyields for the cyclizations of the enynes 91 and 92 (equations D-5and 0-6) when compared with the enyne 84 (equation D-2) wereproposed46’7 to be due to the more favorable enthalpy and entropy ofthe reaction (Thorpe-Ingold effect).48TMS \MeR92 R=COEtThe diastereoselectivities of the cyclizations of the enynes 9 1and 92 differ significantly.47 Thus, the cyclization of the enyne 91gives rise to a very high selectivity in favor of the enone 91a, whilecyclization of enyne 92 leads to virtually no selectivity. Theresults are as would be expected based on the steric effects of 1,3-versus 1,4-pseudo diaxial interactions in the transition states (videinfra). Further confirmation of the importance of steric effects tothe diastereoselectivity in the cyclization may be obtained bycomparing the results of the cyclizations of enynes 91 and 9 3(equation The trimethylsilyl (TMS) group is moresterically bulky than the methyl group. As expected, the cyclization(20h) H H92a 55 : 45 92b(86%)30reaction of the enyne 91 gave rise to a diastereoselectivity greaterthan that of the enyne 93 (79:3 versus 50:15).The working hypotheses for the mechanism of the Pauson-Khandcyclization as proposed by Magnus5° and Schore51 invoke the sametypes of intermediates,* but the Magnus mechanism was developed torationalize the stereoselectivity of the intramolecular reaction ofvarious enynes including 91 and 93 (equation D-7). Thus, accordingto the Magnus proposal (Scheme D-3), alkene insertion into theinternal C-Co bond of the hexacarbonyldicobalt complex 94 leads tothe formation of two cobaltabicyclooctanes 95 and 96. Bothmetallocycles are likely to be cis-fused. In the transition stateleading to metallocycle 95, the steric interactions between R1 andR2 are minimized, but in the corresponding transition state for themetallocycle 96 there is a severe 1,3-pseudo diaxial interactionbetween R1 and R2. Consequently, the pathway for the formation of96 and thus, of the enone 100 is disfavored, particularly in the caseof sterically bulky R’ groups. For the metallocycle 95, insertion ofa carbon monoxide ligand into the indicated C-Co bond gives rise tothe acyl-Co complex 97. Migration of the other C-Co bond to theTBDIRR TBDrCo2(CO)8,heptane(20h)(D-7)a91 R=TMS 91a (79%) 91b (3%)93 R=Me 93a (50%) 93b (15%)* None of the various intermediates have, as yet, been isolated.31adjacent electrophilic carbonyl group produces 98. The reductiveelimination of the cobalt carbonyl residue (likely Co2(CO)6)in 98leads to the formation of the enone 99. The exact identity of thecobalt residue initially eliminated is uncertain,52 but in isooctane[Co4( O)12]53 has been isolated and in aromatic solvents, such asbenzene, [Co4( O)9PhH)]54has been found (Scheme D-4).94 95 961R1100Dashes=CO H98 99Scheme D-332Co2(CO)6 - Co(CO)12 PhH Co4(CO)g(PhH)\CO( )8 PhH, heatScheme 0-4In cases with substituents only at C-7 (i.e., R2=H for enyne 94),the steric interactions are less significant than those describedabove. Thus, the main steric interactions are a 1,4-pseudo diaxialinteraction between R1 and R3 or R4 and an interaction between themetallocycle methylene and R3 or R4. Understandably, the effectsare smaller than for 1,3- or 1,2-pseudo diaxial arrangements ofsubstituents. Reduced steric interactions would lead then to areduced diastereoselectivity as is observed in the cyclization of theenyne 92 (equation D-6) when compared with enynes 91 or 93(equation D-7).2.2.2. Approach to the Synthesis of (±)--Panasinsene (31).The inter- and intramolecular Pauson-Khand cyclizations havebeen used in the syntheses of a variety of natural products andnatural product precursors.55’214679 It seemed that thecyclization would provide a viable approach to the preparation of abicyclic enone 74 which was a desired intermediate in the synthesisof (±)--panasinsene (31). In one step a successful Pauson-Khandreaction would transform an appropriately substituted enyne 75 into33the enone 74 (equation D-8) needed for the key methylenecyclohexane annulation sequence (pp. 2-4).Rxx °:67xx (D -8)According to the synthetic plan, the enyne 75 would have amethyl group on the alkyne function and would have a suitablefunctional group at C-7. The function at C-7 would serve a two-foldpurpose. In the first place, by analogy to the examples given earlier(equations D-5 and D-6 compared with equation D-2), it was hopedthat the presence of substitution at C-7 would contribute to anacceptable reaction yield and reaction time for the cyclization.Secondly, the moiety at C-7 was to be used as a “handle” for futurefunctional group manipulations at that position. Therefore, thenature of the functional group at C-7 was important. Given that afuture step in the planned synthetic sequence (equation D-9) calledfor a carbonyl group at the position corresponding to C-7, the XXmoiety on the enyne 75 would have to be transformable into acarbonyl group. Also, the viability of the synthetic plan depended onthe stability of the XX moiety to the reaction conditions encounteredbefore it was to be converted into a carbonyl group.34oR (D -9)Potential functionalities at C-7 of the enyne 75, in terms offuture usefulness in the synthesis, would include oxygenated groups(XX=OR, H, or XX=2,2-dimethylpropan-1,3-dioxy) or dithioketalderived groups (XX=1 ,3-dithianyl, or XX=SMe, S02p-Tol). To date,there have been no reports of dithioketal derived moieties beingemployed as C-7 substituents of enynes subjected to the PausonKhand cyclization.* If such groups were viable options, they wouldfurther expand the versatility of the cyclization due to the fact thatthe hydrolysis of dithioketal derived functions yields thecorresponding carbonyl group,36’7 while desulfurization of the 1,3-dithianyl group with Raney nickel provides a methylene group.36aThus, the dithioketal derived functions (1,3-dithianyl and SMe, S02p-Tol) at C-7 of enyne 75 were investigated with a view to theirutility in the synthesis of (±)-3-panasinsene (31).* In the intermolecular Pauson-Khand reaction, a methylthioether tethered to the olefinfunction by a carbon chain has been used to enhance the regioselectivity of thereaction 56101352.2.2.1. The 1,3-Dithianyl Function at C-7.The ability to reversibly invert (umpolung)57 the normalreactivity of an acyl carbon atom is a powerful tool in organicsynthesis.36’58 Thus, while acyl groups are generally attacked atthe electrophilic carbon by nucleophiles (for example, 102 gives103, Scheme D-5), an umpolung causing group on the acyl carbonpermits the atom to function as a nucleophile.57 For example,conversion of formaldehyde (102) into 1,3-dithiane (78) followedby deprotonation of 78 with n-butyllithium produces a nucleophilicanion. Treatment of the anion with an electrophile (E) provides104. The dithianyl group of 104 then may be hydrolyzed to producethe aldehyde 1 05 or further deprotonated and reacted with anelectrophile to give 106. Hydrolysis of the dithiane function in 106generates the ketone 107 (Scheme D-5). A variety of electrophiles,including alkyl halides, carbonyl compounds, small ring ethers andacylating reagents, may be employed to transform compound 78 into104 or compound 104 into 106. The hydrolysis reactions(transformation of 104 and 106 into 105 and 107, respectively)most commonly are performed with mercuric salts or with Nhalosuccinimides.36 The overall result of the reaction of the anionof either 78 or 104 with an electrophile followed by hydrolysis ofthe dithiane function in the product is a nucleophilic acylation of theelectrophile.360 0II HS SH 1 )n-BuLi Hydrolysis IIH”H Lewis Acid H><H2) EE><H102 78 104 1051) n-Bu LiNu: 2)E’OH 0Nu—”HE”11’Hydrolysis103 107 106Scheme D-5We wished to exploit the reactivity of 1,3-dithiane (78) toprepare an enyne 75 on which to perform a Pauson-Khand cyclizationto generate the corresponding enone 74 (equation D-1O; XX=1,3-dithianyl). Then, at an appropriate stage in the planned synthesis of(±)-J3-panasinsene (31), the dithianyl function would be hydrolyzedto regenerate a carbonyl function. The proposed reactions of 78with allyl iodide (76) and with 1-iodo-2-butyne (77) appeared to besimilar to alkylations reported earlier,59 but in practice (vide infra)turned out to be somewhat problematic.- - -R—- -:XX (D -10)R78 75 7437_JI_______1To the best of our knowledge, the alkylation of 1 ,3-dithiane(78) with allyl iodide (76) has not been reported in the literature.However, the product of the reaction, 2-allyl-1,3-dithiane (108), isknown.6° Using a procedure similar to that reported for thealkylations of dithiane,59a a THE solution of commercially available1,3-dithiane (78) was treated with n-butyllithium at “‘-25°C toform the dithiane anion. The solution of the anion was cooled to-78°C and allyl iodide (76) was added quickly. Workup of thereaction mixture and purification of the product led to the isolationof 2-allyl-1,3-dithiane (108) in 63-73% yield (equation D-11).1) n-BuLi (1.1 equiv), THE-25 - -30°C (1.5h) SS (D-1 1)S..,.,..S 2) AIlyl iodide (76, 1.1 equiv),-78°C (2.5h);-78—* 0°C (1..5h) I8 108The ir spectrum (neat) of 2-allyl-1,3-dithiane 108 exhibitedabsorptions due to the mono-substituted alkene at 3077 (w), 1640(m), 991 (s), and 920 (vs) cm1.61 In the 1H nmr spectrum (400 MHz,CDCI3), signals were found for the-SCS- at 84.10 (t, 1H, J = 7 Hz)and for the olefinic hydrogens at 5.11-5.19 (m, 2H, CH=Ca2), and at385.82-5.92 (m, 1H, Ca=CH2).* The exact mass of the molecular ionwas found to be 160.0374 which is consistent with a molecularformula of C7H,2S.The second alkylating agent for the preparation of the enyne 75,the iodide 77, was synthesized from the corresponding alcohol(109) using a modification of a known procedure.62 Thus, adichioromethane solution of the commercially available alcohol 1 09was treated with triphenylphosphine diiodide (1.1 equiv) in thepresence of triethylamine (1.1 equiv) to give the iodide 77 (equationD-12). The ir spectrum of the iodide 77 exhibited an alkyne CCstretch at 2235 cm-1, while the ‘H nmr spectrum (300 MHz, CDCI3)showed signals for the methyl group at 8 1 .83 (t, 3H, J = 3 Hz) andfor the methylene group at 3.68 (q, 2H, J = 3 Hz).OHII + PPh3 + ‘2 CH2I (D-12)Et3N, rt,(5-7h)109 77In order to prepare the enyne 75, a THE solution of 2-allyl-1,3-dithiane (1 08) was treated first with n-butyllithium (1 .1 equiv) at-‘25-30°C (-‘3 hours) to form the anion and then with 1-iodo-2-butyne 77 (-‘1.2 equiv) to perform the alkylation. The desired alkyne* The signals due to the allylic methylene appeared at 32.52 (tt, 2H, J = 7, 1 Hz),while the dithiane methylene hydrogens appeared as multiplets at 1.80-1.92 (1H),2.08-2.16 (1H) and 2.80-2.94 (4H). According to the literature,60 the nmr signalsfor 108 are as follows: 3 1.6-3.0 (m, 8H), 4.1 (t, 1H, J= 6.6 Hz, -SCHS-) and 4.9-6.7 (m, 3H, -CH=CH2).3911 0 (an oil) and the allene isomer 111 (an oil) were formed inapproximately equal amounts (ratio, 1:1 .2, respectively, GLCanalysis) (equation D-13). The isolated yield of each isomer wasgenerally 12-18% because the two isomers were difficult toseparate from each other and from other by-products of thereaction .1) n-BuLi (1.1 equiv), THE2) ;5:ce+ (D-13)108 110 111The 1 H nmr spectrum (400 MHz, CDCI3) of the alkyne 11 0exhibited signals for the acetylenic methyl group at 3 1.84 (t, 3H, J =‘-2 Hz) and for the olefinic hydrogens at 5.17-5.25 (m, 2H) and 5.87-5.97 (m, 1H).** In the ir spectrum (neat) the alkyne CC stretchoccurred at 2235 (w) cm-1, while the alkene C=C stretch was at1638 (m) cm* The exact mass of the molecular ion was found to be212.0697 which is consistent with the molecular formula ofC11H6S2.The structure of the allene 111 was consistent with the 1 H nmr,ir and low resolution mass spectral data. Thus, in the 1 H nmr* The purification procedure was not further optimized since the approach wasultimately abandoned.**Other hydrogen signals in the 1 nmr spectrum appeared at 3 1 .90-2.06 (m, 2H) and2.76-2.93 (m, 8H). In the ir other absorptions due to the alkene were at 3076 (m),991 (m), 920 (s) cm-1.40spectrum (400 MHz, Cod3), the allene 111 exhibited signals for theallenic methyl group at 8 1.82 (t, 3H, J = -2 Hz), for the allenichydrogens at 4.88 (q, 2H, J = 2 Hz) and for the olefinic hydrogens at5.13-5.20 (m, 2H), 5.79-5.90 (m, 1H).* The ir spectrum of the allene111 showed an allenic C=C stretch at 1953 (vs) cm1 and an olefinicC=C stretch at 1639 (s) cm. In the low resolution mass spectrumthe molecular ion was observed at 212 mass units (13%).The presence of allenes as by-products in the reactions ofnucleophiles with propargylic alkylating reagents (and conversely,the presence of acetylenic by-products in similar reactions) is apersistent problem.63’4 It is known that the alkyne/allene ratio inthe product can be influenced by a variety of factors which include:the solvent, the temperature, the structure of the propargylicsubstrate and the structure of the nucleophile.65 Thus, for example,in the reaction of a Grignard reagent with 112 (equation D-14), theacetylene/allene ratio in the product mixture generally was largerat higher temperatures and smaller in less polar solvents.65c Incontrast, similar reactions (using 1 ,4-dichloro-2-butyne) ofalkyllithium reagents tended to give the opposite results in thatraising the temperature led to a higher proportion of the allene inthe product mixture.65d Also, different products have been obtaineddepending on the nucleophile employed.65a Thus, in the reactionof methyllithium with 11 2 (X,Y=Cl) the product was mainly theallene 11 4 (R=Me, Y=Cl), while the same reaction with* Other hydrogen signals in the ‘H nmr spectrum were found at 31.84-1.96 (m, 1H),2.01-2.09 (m, 1H), 2.64-2.72 (m, 4H), 2.92-3.00 (m, 2H). In the ir spectrum,other absorptions appeared at 3076 (s), 995 (vs), 917 (vs), 847 (vs) cm*41methylmagnesium bromide gave mainly the acetylene 113 (R=Me,Y=Cl).65dCH2V x Ii___________Nu:, solvent + (D-1 4)112 R’)Y= OMe, OH Alkyl 113 114X= Cl, BrNu:= RMgBrIn our hands, various modifications in the reaction conditionsfor the alkylation of the lithio anion of 2-aIIyl-1 ,3-dithiane (11 3)with 1-iodo-2-butyne (77) were made and included the following:(a) changing the solvent (THE, DME), (b) varying the temperature(-78°C, --25°C), or (c) using an additive (none, HMPA). However,results similar to or worse than those described above wereobtained. In addition, modifying the alkylating reagent 77 byreplacing the methyl group with the more bulky triisopropylsilyl(TIPS) group66 to give the iodide 116 did not improve the outcome ofthe reaction. The option of changing the structure of the nucleophileremained to be explored, but first the feasibility of the PausonKhand cyclization reaction of an enyne with a dithioketal function atC-7 needed to be established.42* The absorptions due to the organic moiety were very weak in comparison with thecarbonyl absorptions. Absorptions (br, vw) due to the olefinic group appeared at 967and 931 cm-1, while the absorption due to the Co-C was at 519 cm-1. There was nooctacarbonyldicobalt present as evidenced by the lack of an absorption at 1864 cm-1 dueto bridging carbonyl Iigands.43n__—/_,,‘I — TIPS —77 116115The enyne 11 0 was subjected to Pauson-Khand cyclizationconditions similar to those reported by Magnus.46 Thus, the deep-redhexacarbonyldicobalt alkyne complex 117 was formed by reactingthe alkyne function in enyne 11 0 with octacarbonyldicobalt (1 .1equiv) (Scheme D-6). Purification of the crude product mixture byrapid chromatography on a Florisil column gave the complex 117 in68-78% yield. The ir spectrum (KBr) of the complex exhibitedcarbonyl absorptions at 2087 (m), 2045 (s), and 2016 (s) cm**The hexacarbonyldicobalt complex 117, dissolved in heptane andsealed in a resealable tube under a carbon monoxide atmosphere, washeated at 90-100°C for 14-18 hours. The yield of the purified enone118 (a pale yellow oil) was 5O% based on the enyne (or 70-80%,based on the isolated complex). The yield and reaction time werecomparable to those reported previously (compare with equation D7).43Co2(CO)6S Co2(C0)8,CO,— I —heptane, rt110 117heptane, CO90-100°CMe (14-18h)118Scheme D-6The structure of the enone 11 8 was confirmed by the ir, nmrand high resolution mass spectral data. Thus, the ir spectrum of theenone 118 exhibited absorptions at 1707 (vs) and 1672 (vs) cm-1which are characteristic of a conjugated cyclopentenone.6’ In the ‘Hnmr spectrum (300 MHz, CDCI3; traces of impurities present), thevinylic methyl group appeared at 3 1.73 (br s, 3H) and the angularhydrogen appeared at 3.25-3.39 (m, 1H).* In the high resolution massspectrum the exact mass was found to be 240.0644, which isconsistent with the molecular formula, C12H60S2.The Pauson-Khand cyclization of enyne 110 occurred with anacceptable yield in a reasonable length of time and the enone 11 8was isolable. However, the overall route was syntheticallymediocre due to the practical difficulties involved in the synthesis* Further spectral data also was observed. Other absorptions in the ir appeared at 1413(s), 1310 (s), 1050 (m), 936 (w), 906 (w) and 667 (w) cm1. In the 1H nmrspectrum, signals due to the other hydrogens appeared at 31.58 (t, 1H, J = 12 Hz),2.06-2.17 (m, 3H), 2.70 (dd, 1H, J = 17, —6 Hz) and 2.83-3.20 (m, 7H).44of the enyne 110. The problems included: (a) the low yield in thealkylation reaction with 1-iodo-2-butyne 77 and (b) the difficultyin separating the desired alkylated product 11 0 from the alleneisomer 111 (equation D-13). Thus, it was decided to modify thedithioketal function and use an oxidized dithioketal derivative, themethylthio-p-toluenesulfonyl function. It was expected that themore reactive anion (11 9) would show different (preferably moredesirable) behavior in the alkylation reaction with the alkyne 77.MeS - S02p-Tol1192.2.2.2. The Methylthio-p-toluenesulfonyl Function at C-7.Several dithioacetal3S S-monosulfoxides67’8or S,S-dioxides,69including methyithiomethyl p-tolyl sulfone (79),37 have been used asmasked carbonyl anion equivalents. However, ketone synthesis viathe oxidized dithioacetal reagents, 120-122, may be accompaniedby problems such as: (a) competative alkylation on a monoactivatedalkyl group instead of at the doubly activated methylene position;37(b) little or no dialkylation;68 or (c) difficulties in the hydrolysis ofthe oxidized dithioketal function to generate the carbonyl group.68aIn contrast, for sulfone 79, site-selective deprotonation of the45methylene group occurs with a variety of bases* and produces ananion which may be reacted with electrophiles such asaldehydes ,37c71 esters,37c,72 a,I3-unsaturated carbonyl compounds71and alkyl halides.37b, The dithioacetal S,S-dioxide function in theadducts so formed may be hydrolyzed to give an aldehyde carbonylgroup or again treated with a base to form the a-anion. Reaction ofthe anion with a second electrophile followed by hydrolysis of themethylthio p-toluenesulfonyl function produces a ketone. Thehydrolysis of the methylthio p-toluenesulfonyl group generally hasbeen performed photochemically37b or using acidicCo nditions.37I,c7la,73MeS\/SO2PToI MeS,,SOMe MeS,SO2e EtS\/SOEt79 12067 12169 12268It appeared that methylthiomethyl p-tolyl sulfone (79) was aviable carbonyl anion equivalent and could be used in our strategy toprepare the enone 74 via the Pauson-Khand cyclization of an enyne75 (equation D-15; XX=SMe, S02p-Tol). As was the case for 1,3-dithiane (78) described earlier, the sulfone 79 was alkylated withallyl iodide (76) and 1-iodo-2-butyne (77) to prepare the desiredenyne.* Bases used include: 50% aqueous NaOH-toluene/trioctylmethylammoniumchIoride;37bNaH-DMF;70 NaHTHF;37CK2C03-i-PrOH ;37c and n-BuLi-THF.746MeS,SO2p-Toi --o5: (D-1 5)Methylthiomethyl p-tolyl sulfone (79) (also available from theAldrich Chemical Co.), was prepared and recrystallized according tothe procedure reported by Ogura and coworkers.74 The product thusobtained exhibited m.p. 83.5-85°C (lit. 82-83°C74)and 1H nmr data inaccord with the reported data.A cold (-78°C) THE solution of the sulfone 79 was deprotonatedwith n-butyllithium and the resultant anion was treated with allyliodide (76) (equation D-16). The purified monoalkylated material(123) was obtained as a white solid (crude m.p. 37-38°C) in 7O-75% yield. The amount of the dialkylated product 124 obtained andthat of the starting sulfone 79 recovered varied depending on thescale of the reaction and the number of equivalents of allyl iodide(76) used. Thus, in our hands, for a small scale reaction (-2 mmolof 79) 1.1 equivalents of allyl iodide were found to give 124 and79 each in -10% yield. However, in a large scale reaction (‘-10mmol of 79), four equivalents of allyl iodide (76) were needed toobtain a similar result. Other conditions generally led to theformation of either 124 or 79 in larger amounts.471) n-BuLi (1.1 equiv), MeS S02p-Tol MeS S02p-TolMeS\,SO2P-Tol THE, -78°C (lh) + >< D 12) AIlyl iodide (76) J I %j ( - 6)7 9 (1.1 equiv), -78°C 1(1.5h); -78—,-15°C (0.5h) 1 23 124In the 1H nmr spectrum (300 MHz, CDCI3) of 1-methylthio-1-p-toluenesulfonylbut-3-ene (123) signals for the two methyl groupswere displayed at 32.24 (s, 3H) and 2.44 (s, 3H), while the signalsfor the three olefinic hydrogens were at 5.16-5.20 (m, 2H) and 5.73-5.87 (m, 1H).* In the low resolution mass spectrum, the molecularion peak at m/z = 256 (C12H602S)was very small (0.3%), while thebase peak (101) corresponded to the loss of the p-toluenesulfonylfragment, C7H02S(m/z=155).The structure of the diallylated sulfone 1 24 was readilydeduced from its 1H nmr spectrum (300 MHz, CDCI3). Thus, theexpected signals for the two methyl groups (s, 3H each) were at 32.25 and 2.36, while the signals for the olefinic hydrogens appearedat 5.12-5.20 (m, 4H) and 5.85-6.00 (m, 2H).** In the ir spectrumabsorptions due to the C=C stretch and the sulfone asymmetric andsymmetric S=O stretches were displayed at 1639, 1302 and 1147cm1, respectively.61 In the low resolution mass spectrum, themolecular ion appeared at m/z = 296 (0.3%) and a major fragment* Other aliphatic hydrogen signals in the 1H nmr spectrum of 123 appeared at 82.24-2.34 (m, 1H), 2.91-3.00 (m, 1H) and 3.71 (dd, 1H, J = 11, -4 Hz), while thearomatic hydrogens were at 7.36 (d, 2H, J = 8 Hz) and 7.84 (d, 2H, J = 8 Hz).** The 1H nmr signals for the remaining hydrogens of 124 were displayed at 32.60-2.81 (m, 4H), 7.33 (d, 2H,J = 8 Hz) and 7.84 (d, 2H, J = 8 Hz).48corresponding to the loss of the p-toluenesulfonyl moiety (C7H02S)was observed at 141 (32%).The enyne 125 was prepared by treating a cold (-78°C) THEsolution of the sulfone 1 23 with n-butyllithium and alkylating theresultant anion with 1-iodo-2-butyne (77) (equation D-17). Theisolated, purified 4-methylthio-4-p-toluenesulfonyloct-1 -en-6-yne(125) was obtained as a pale yellow oil in 79-86% yield.Gratifyingly, none of the corresponding allene 126 was detected.MeS ,S02p-ToI MeS S02p-ToI1) n-BuLi (1.25 equiv),THE, -78°C (O.5h) (D-17)Ii 2) E801w II \78— 15°C ( lh)123 125The presence in the ‘H nmr spectrum (300 MHz, CDCI3) of a newmethyl signal at 3 1.70 (t, 3H, J = -2Hz) for the acetylenic methylgroup and signals for two other methyl groups (s, 3H, each) at 2.31and 2.45, as well as the other expected signals,* indicated that thedesired compound (125) had been prepared. The ir spectrum (neat),displayed absorptions due to the olefin C=C stretch at 1639, the* Other hydrogen signals appeared at 32.71-2.94 (m, 4H), 5.19-5.28 (m, 2H), 5.88-6.01 (m, 1H), 7.35 (d, 2H, J = 8 Hz) and 7.87 (d, 2H, J = 8Hz).12649alkyne CC stretch at 2236 (very weak), and the sulfone S=Oasymmetric and symmetric stretches (strong) at 13O2 and 1144cm, respectively.61 The molecular ion peak in the low resolutionmass spectrum was at m/z = 308 mass units, consistent with themolecular formula, C16H200S.Though similar reaction conditions were utilized, the results ofthe alkylation of the lithio anions of sulfone 123 and dithiane 109with the iodide 77 were quite different. Thus, for the anion 11 9only the desired SN2 reaction was observed, while for the anion115, both SN2 and SN2’ reactions occurred. The differences in thebehaviors of the anions 119 and 115 are likely due to differences inthe anion structure. As mentioned earlier, reactions of Grignard andorganolithium reagents with propargylic substrates tend to givecomplementary results, but the provenance of this behavior is notfully understood.63a,5 However, our results confirm that thestructure of the nucleophile also plays an important role in theoutcome of the reaction of a nucleophile with a propargylic halide.MeS SOToI[1 E[y+119 11550The enyne 125 was subjected to the Pauson-Khand cyclizationreaction (Scheme D-7). Cyclization of the enyne in this case cangive rise to a mixture of diastereomers, but based on the resultsreported in the literature47 and decribed earlier (equation D-6), thediastereoselectivity was expected to be minimal. The enyne 1 25was treated with octacarbonyldicobalt (1.2 equiv) in benzene.After chromatography of the crude product on Florisil, thehexacarbonyldicobalt alkyne complex 127 was isolated as a deep redoil in 80-90% yield. The complex was dissolved in benzene, and theresultant solution was sealed in a resealable tube under a carbonmonoxide atmosphere and then was heated at 80-90°C. The mixtureof diastereomers 128 and 129 (ratio -‘1:1, 1H nmr analysis) wasformed in 46-54% yield based on the isolated complex 127, or in37-49% yield based on the starting material (enyne 1 25). The yieldfor the cyclization of enyne 125 was a bit lower than that of theprevious example (enyne 110, Scheme D-6), but the reaction alsooccurred somewhat more rapidly (5-12 hours versus 14-18 hours).The two diastereomers 128 and 129 could be separated bycareful flash column chromatography or could be partially separatedby fractional recystallization from dichloromethane-pentane. Therelative stereochemistries of the two compounds were notdetermined. Using the above solvent system, the less polardiastereomer (A) crystallized first from mixtures of bothcompounds. Pure A was recrystallized from diethyl etherdichoromethane (3:1) to produce prisms that exhibited m.p. 146.5-148.5°C. The ‘H nmr spectrum (300 MHz, CDCI3) of A displayed theexpected three singlets (3H each) for the methyl groups at 3 1.71 (br51Co2(CO)6SO2pToI :::: (1.::quhi),____125 127PhH, CC,80- 90°C(12h)1 :1Scheme D-7s, allylic Me), 2.34 and 2.48. The signal due to the angular hydrogenappeared as a broad multiplet at 3 -‘3.28-3.40 (overlapped with adoublet at 3.45) and the aromatic hydrogens resonated at 7.39 (d, 2H,J = 8 Hz) and 7.90 (d, 2H, J = 8 Hz)*. In the ‘3C nmr spectrum (75MHz, CDCI3), fifteen signals were observed due to the fact that twopairs of aromatic carbon atoms were magnetically equivalent (seefigure 1). The signals for the three methyl groups were at 3 8.6,14.5 and 21.7, while the angular methine carbon resonated at 42.30.**The data was in accord with what one would expect for one of thebicyclic enone diastereomers.* The signals for the other hydrogens appeared at 81.54 (dd, 1H, J = 14, 11 Hz),2.07(dd, 1H, J = 18, 3 Hz), 2.66-2.74 (m, 2H), 3.05 (dd, 1H, J = 14, 9 Hz) and 3.45 (d,1H, J = 18 Hz).** The ‘3C signals for the three methylene groups appeared at 335.8, 40.0 and 41.99;those for the six quaternary carbons were at 76.6, 131.20, 134.1, 145.4, 174.5,209.0 (carbonyl); and those for the aromatic methynes were at 129.3 (2H) and130.94 (2.H).+129 128Figure 1. The 75 MHz broad bandenone diastereomer A.decoupled 13C nmr spectrum of theThe diastereomer B was recrystallized from acetone/pentaneand exhibited m.p. 149.5-152°C. In the 1H nmr spectrum (300 MHz,CDCI3) of diastereomer B there were some slight differences fromthat of A.appearedhydrogenresonatedThus, the signals for the three methyl groups (3H each)at 3 1.72 (br s, allylic Me), 2.36 and 2.47, while the angularappeared at 3 —3.21-3.34 and the aromatic hydrogensat 7.37 (d, 2H, J = 8 Hz) and 7.83 (d, 2H, J = 8 Hz).** The signals for the other hydrogens appeared at 3 2.09-2.20 (m, 3H), 2.63-2.76 (m,2H) and 3.50 (d, 1H, J = 19 Hz).52• It)No N-313 134 132 ,_:)- -82 80 78 76 74 72 PP8I‘a10’ NID+0)-3C 01N‘aC‘1• It3- ‘a0• U,—Nc4*-) 10‘a010l‘a0N‘a“3“5F..‘aIt)Ne It)1.N -‘aF.I220 200 180 160 140 120 100 80 60 40 20 PPM 053With the enones 128 and 129 in hand, the key methylenecyclohexane annulation was investigated. It was expected that thereaction might be sluggish based on the results presented in theIntroduction (pp. 2-4). However, it was anticipated that the reactionwould be possible, since copper (1)-catalyzed conjugate additionshave been performed on enones with tetrasubstituted double bonds.Thus, for example, Paquette and Han75 found that the diquinane enone130 reacted with the Grignard reagent 131 in the presence of acopper (I) salt to give the adduct 132 in 68% yield (equation D-18),but the reaction took 12 hours at -78°C.BrMg O_>(13i) (D-18)CuBrSMe2,THE,Me2S, -78°C (12h)The attempts to perform the copper (1)-catalyzed conjugateaddition of the Grignard reagent 7 on the enones 128 and 129 werefrustrated by the lack of solubility of the enones in THF, the solventnormally used in our laboratories7 for the reaction. The enones weresoluble in hexamethylphosphoramide (HMPA), an additive sometimesused along with trimethylsilyl chloride to improve sluggishconjugate addition reactions of Grignard reagents76 (copper (I)catalysis) or of stoichiometric organocopper reagents.77 However,the amount of HMPA required to dissolve the enones was such thatthe polarity of the reaction solvent mixture would be significantly130:Q13254increased. It is known that 1,4-additions of cuprates to enonesoccur more readily in less polar solvents78 (diethyl ether, dimethylsulfide, hydrocarbons) and that HMPA (in the absence oftrimethylsilyl chloride) retards the 1 ,4additions.778 Notunexpectedly, the desired conjugate addition reaction of 7 to theenones 128 and 129 in a THF-HMPA solvent mixture (‘-24% v/vHMPA) was unsuccessful. The reaction in the presence of TMSCI (2equiv relative to the enones) also failed (equation D-19).(D-19)Alkylated methylthiomethyl p-tolyl sulfones were reported toundergo a facile hydrolysis of the thioether sulfone function to givethe corresponding carbonyl compounds using a variety of mildreaction conditions.37b,071a3 Therefore, it was decided tohydrolyze the thioether sulfone function in 128 and 129 to the ketofunction in the enone 135. The reactivities of the two carbonylgroups in the enone 135 would be different, thus permitting theprotection of the saturated ketone in the presence of theunsaturated one. It was envisaged that the saturated ketone wouldbe reduced to the alcohol and protected as an ether. The hydrolysisreaction was attempted under several of the reported conditions(concentrated HCI/MeOH,37b69 CuCI2•silica gel/CH2C173 andCIMe.128 and 129 133 and 13455CuCI2/H0/MeOH731’), but mixtures of products resulted and theattempt was abandoned (equation D-20).X ‘- o , (D-20)TolIn summary, our use of the Pauson-Khand reaction to prepare theimportant bicyclic enone 74 (see Scheme D-1) on which we wishedto carry out the methylenecyclohexane annulation was discontinueddue to the unanticipated difficulties encountered which included; thepoor yields in the alkylation of the 2-allyl-1,3-dithiane (109) withthe propargylic iodide 77; the low solubilities of the enones 128and 129 in THE; and the mixtures of products obtained in theattempted unmasking of the ketone carbonyl in the same enones.However, it was gratifying to find that the Pauson-Khand cyclizationcould be done with dithioketal derived functions at C-7 and that theyields and reaction times were comparable to the previouslyreported examples.128 and 129 13574562.3. The Synthesis of (±)-J3-Panasinsene (31) via the Weiss-Cook Condensation Reaction.2.3.1. Background.The Weiss-Cook condensation, which has been described as a 3-component (A+B+B or 2+3÷3) coupling reaction,79 has proved usefulin the synthesis of many natural and non-natural polyquinanes.25The reaction of a dialkyl 3-oxoglutarate (1 36) (2 equiv) with a 1,2-dicarbonyl compound 137 in the presence of a base catalyst (or lesscommonly, an acid catalyst) produces, in high yield, a tetraalkyl cisbicyclo[3.3.O]octane-3,7-dione-2,4,6,8-tetracarboxylate 138.Heating the tetraester 1 38 with acid leads to the hydrolysis of theester functions. The 13-keto acid groups thus formed undergo aspontaneous decarboxylation to generate the dione 139 (equation D21).RO2CRO2C136 137Many different tetraester and dione compounds (138 and 139)may be produced via the Weiss-Cook condensation.25 The identitiesof the R groups in 1 38 are determined by the ester of 3-oxoglutarate which is employed; normally, R is either a methyl or a136 138 13957t-butyl group. The structure of the dicarbonyl compound utilizeddetermines the nature of the R’ and R” groups of 138 and 139; R’ andR” may be hydrogens, alkyl groups or aryl groups and may be eitherthe same or different. Thus, glyoxal (44, R’=R”=H), a-keto aldehydes(R’, R”=H, alkyl/aryl), and acyclic or cyclic a-diketones (R’,R”=alkyl/aryl) may be utilized with the proviso that the bicyclicproduct may not be formed if very sterically bulky groups are used(vide infra).8° Also, dicarbonyl compounds with limited solubility inthe usual aqueous solvents (i.e., R’, R”=large alicyclic group) may beemployed if the reaction is performed in organic solvents.81The utility of the Weiss-Cook condensation is due in part to thefact that both the tetraester intermediate 1 38 and the bicyclicdione 1 39 are rich in functional groups, which permit furthersynthetic manipulations of the molecule. Different functionalgroups at the 1- and 5-positions may be introduced by varying thestructure of the initial dicarbonyl compound 1 37 and then byperforming suitable synthetic manipulations on the R’ and R” groups.Additional functional groups may be added selectively25 at the 2-,4-, 6- and 8-positions by standard alkylation procedures or at the 3-and 7-positions via carbonyl group reactions. It is possible tointroduce groups regioselectively at the 2- and 6- or the 2- and 8-positions of the bisenol ether of the tetraester (e.g., 140) or of thedione 139.25 Also, as exemplified in Scheme D-8, selectivemonoalkylations82 using potassium hydride and an electrophile (Mel,Eti, allyl iodide, etc.) have been performed on the bisenol ethers oftetra-t-butyltetraester intermediates such as 140•82b Subsequent58t-BuO C C02t-BuHO—OH CH2,(93%)Me0t-BuO2C C02t-Bu141Meo AcOH/HCI, MeO87°C (82%)H143Scheme D-8hydrolysis and decarboxylation of the ester functions of thealkylated material,142, would provide the monoalkylated dione 143.One restriction on the versatility of the Weiss-Cookcondensation reaction to produce diquinanes is that only the cisbicyclo[3.3.O]octane-3,7-dione stereoisomer is produced. Thepreference for the formation of the cis isomer is reasonable sincethe presently accepted mechanism83 for the reaction is based on aseries of equilibria (Scheme D-9). Also, Boyd and coworkers34 havecalculated that the trans isomer of bicyclo[3.3.O]octane is -6.5kcal/mole less stable than the cis isomer. Thus, even in the unlikelyevent that any of the trans isomer were to be formed during thereaction, it would rapidly undergo a reverse reaction and eventually140KH, DMF,-60 to -50°C,Mel14259would be transformed into the thermodynamically more stable cisisomer.The proposed mechanism83aB3bfor the Weiss-Cook condensationis thought to proceed as depicted in Scheme D-9. An aldolcondensation of one molecule of the mono-anion of the dialkyl 3-oxoglutarate (144) with one carbonyl group of a molecule of thedicarbonyl compound 137 produces the hydroxy dione 145. A secondCO2R0 R’ CO2R+ RKCQRHOCO2R137 144 145+ 144:0146H2O+4L148H2OICO2R147149 138Scheme D-960aldol reaction may generate the diol 146 which loses a molecule ofwater to produce the enone 147 (or possibly the dehydration mayoccur before the second aldol reaction and circumvent the did 146).Michael addition of a second molecule of the anion 144 to the enone147 gives the hydroxy dione 148. Loss of a molecule of water from148 to give the enone 149, followed by a second Michael additionleads to the formation of the tetraester 138. If appropriate R’ andR” groups are used (e.g. R’=R”=chexyl),8O the enone intermediate1 47 may be isolated as the reaction is sensitive to steric effectscaused by the R’ and R” groups.8°2.3.2. Application of the Weiss-Cook Condensation Reaction to theSynthesis of (±)-f3-Panasinsene (31). Preparation of an enone 74.RIf the Weiss-Cook approach were to be employed for thepreparation of the key enone 74 in the synthesis of J3-panasinsene(31), then the R’ and R” groups in the dione 150 would be hydrogens.Furthermore, one of the two keto functions of the dione 150 wouldForm bond(R Protect150H7461need to be protected to permit differentiation between the twocarbonyls, and an appropriate R group would have to be introducedbefore the enone double bond was installed. With regard to theidentity of the R group, it was thought that preparation of an enone74 with R=methoxycarbonyl would provide several advantages (videinfra) in comparison with the preparation of an enone having theultimately desired R=methyl. Thus, a synthetic sequence utilizingthe monoalkylation (Mel) of the bisenol ether tetra-t-butyltetraester intermediate 140 to generate the methylated dione 143(Scheme D-8) was not considered.Synthetically, a base-catalyzed Weiss-Cook condensation usingglyoxal 44 and dimethyl 3-oxoglutarate 45, followed by an acid-catalyzed hydrolysis! decarboxylation of the tetraesterintermediate (151) provided the known dione 43 (equation D-22).84MeO2C CO2Meo+ +0NaOH,MeOHr /MeO2C CO2Me /HCI (1M),45 44 45 /AcOH,heat(D-22)Selective protection of one of the carbonyl functions in 43 togive the keto ketal 46 was achieved via an acid catalyzed reaction15162of 43 with 2,2-dimethyl-1,3-propanediol (152) by the methodreported by Moss and Piers (equation D-23).38’85 Preparation of 46usually involves a tedious chromatographic separation of 46 fromthe diketal 153 and the dione 43. However, the purification wassimplified by loading the sample as a solid adsorbed on Celite ontothe silica gel column and successively eluting compounds 153, 46and 43 with diethyl ether-petroleum ether (2:1), diethyl ether-ethylacetate (9:1) and ethyl acetate (neat), respectively. The purifiedketo ketal exhibited m.p. 46.5-47.5°C (literature85 m.p. 48°C).43 152In order to transform the keto ketal 46 into the desired enone154, it was necessary to introduce an appropriate R group and theenone double bond (equation D-24). Depending on the identity of theR group, formation of the double bond and the proposed conjugateaddition reaction could be more or less expedited. If R were amethyl group, as is found in f3-panasinsene (31), then in the formalH+46 (D-23)153+4363(D-24)dehydrogenation step to generate enone 154, the double bondtheoretically could end up either exo- or endo- to the 5-memberedring (156 or 157, respectively, equation D-25). Due to the straininvolved in introducing two new sp2 centers into the five-memberedring to give 157, it was difficult to predict, a priori, whether or notthe elimination of the selenoxide derived from 155 would generatethe endo isomer 1 57 as the major product; however, others havenoted the predominance of the endo isomer86 for a variety ofbicyclic lactones and 2,3-dialkylated cyclopentanones. Another,less serious problem, was that the presence of the methyl group onthe enone double bond would deactivate87 the enone 157 towards thekey copper (1)-catalyzed conjugate addition reaction of the Grignard155(D-25)46 154H+ 15615764reagent 7. On the other hand, if R were the methoxycarbonylfunction, there would be no possibility of the formation of an enonewith an exocyclic double bond from the keto ester 158. In addition,the enone 159 would be activated78 towards conjugate additionreactions. A disadvantage of utilizing R=methoxycarbonyl was thatat some stage during the synthesis, the methoxycarbonyl groupwould have to be deoxygenated to generate the corresponding methylgroup. However, deoxygenation reactions were already planned attwo stages of the synthesis (between 160 and 1 61 and between162 and 31, relevant positions indicated, Scheme D-1O), so it wasexpected that at either point a double deoxygenation could be done.Consequently, it was decided that a methoxycarbonyl group ratherthan a methyl group would be employed as the R group in thesynthesis of an enone 74 and that its deoxygenation would beperformed in tandem with the deoxygenation of either the keto orthe ester functions in 1 60 or 1 62, respectively.7 158 15965______‘COMeThe keto ester ketal 1 58 (actually as a mixture of 1 63 and1 64) was prepared via a modification of the procedure reported byDeslongchamps and coworkers.88 Thus, a THE solution of the ketoketal 46 was treated with potassium hydride, and the enolate anionthus generated was allowed to react with dimethyl carbonate toform, in -93% yield, an -.1.5:1 mixture (1H nmr analysis) of the ketoester 163 and its ester enol tautomer 164 (equation D-26). Due tothe fact that the mixture of 1 63 and 164 was not stable topurification by flash chromatography, this material was notrigorously purified.The stereochemistry of the keto ester 163 was not proven, butwas assumed to be that shown based on the following reasoning.Excess base present during the reaction would remove the protonR160 161I I31 162Scheme D-1066between the keto and ester functions to generate the correspondingenolate. During the workup of the reaction, protonation of theenolate would give a mixture of 1 63 and the epimer at C-2.Equilibration of the mixture via the enol tautomer 164 would lead tothe methoxycarbonyl group preferentially being on the stericallyless congested, convex face89 of the molecule.In the 1H nmr spectrum of the crude mixture, the signals due tothe methoxycarbonyl functions of 163 and 164 were displayed assinglets at 8 3.73 and 3.75, respectively, while the signal for theenol 0jL of 164 was displayed at 10.35. In the high resolution massspectrum, the exact mass of the molecular ion of the mixture (1 63and 164) was found to be 282.1459, which is consistent with themolecular formula C15H2205. Signals for fragments corresponding tothe loss of MeOH (M-32) and the loss of the methoxycarbonyl group(M-59) were also observed in the low resolution mass spectrum.Such signals were displayed by many of the other methoxycarbonylcontaining intermediates which were prepared.MeO2C H1)KH,THF, 163 (D-26)6O°C (2h)2) (MeO)C0,—60°C (1.5h)HOH46+16467In order to generate the enone 159, a modification of theselenoxide syn elimination procedure developed by Reich andcoworkers9°was employed. The required selenide was prepared bytreating a THF solution of a mixture of the keto ester 1 63 and itsenol tautomer 164 with potassium hydride (1.3 equiv) and allowingthe resultant enolate anion to react with benzeneselenenyl chloride(1 .35 equiv) at 0°C. An -4:1 mixture (1 H nmr analysis, using theratio of the signals of the methoxy groups) of the epimeric selenides165 and 166 was obtained in 86% yield.1) KH, THE, -rt (40mm)2) PhSeCI, 0°C(20mm)The two epimers 165 and 166 could be distinguished readily bythree main signals in the 1H nmr spectra of the mixture. Thus, thesignals for the tertiary methyl groups of the major epimer 165appeared at 8 0.89 and 0.97, while those of the minor epimer 1 66were at 0.92 and 0.99. The resonances due to the methoxycarbonylfunctions were at 3 3.71 and 3.51, respectively. Also, signals forPhH163 and 164165116668aromatic hydrogens at 37.53-7.57 and 7.63-7.65 were characteristicof 165 and 166, respectively. A small amount of the minor epimer(166) was obtained in pUre form and the expected molecularformula, C21H6O580Se, was confirmed by the presence of an ion witha mass of 438.0940 mass units in the high resolution massspectrum.The oxidation of the mixture of the selenides 165 and 166 tothe corresponding selenoxides had to be done with care as 2-alkoxycarbonyl-2-cyclopenten-1 -ones are sensitive to basecatalyzed epoxidation by hydrogen peroxide.9° In the case of 165and 166, only the major epimer 165 would be able to undergo thenormal selenoxide syn elimination.91 However, because compounds165 and 166 could not be separated by chromatography (silica gel),the minor epimer 166 was also present during the oxidation to givethe corresponding selenoxide. Thus, a dichloromethane solution ofthe mixture of the epimeric selenides was treated with 15% aqueoushydrogen peroxide (2.1 equiv) at 0°C (10 mm) and room temperature(20 mm) (equation D-28). A ‘H nmr spectrum of the crude productindicated unexpectedly that the product obtained was quite purewith just traces of aromatic compounds present.H20,CH2I,0°C (10mm);rt (20mm)H(D-28)165 15969A control experiment was done to try to determine the fate ofthe minor selenide epimer 166. Thus, a sample of 166 with lessthan 5% of the major epimer 165 present (H nmr analysis), butwhich contained some other impurities, was oxidized using ournormal procedure. A H nmr spectrum of the crude product showedthe presence in the mixture of the enone 159 and small amounts ofother compounds having aromatic, methoxycarbonyl and/or ketalfunctions present. The presence of enone 159 was surprising giventhat the normal syn elimination91 (as shown in formula 1 67) of theselenoxide cannot occur in compound 168. There have been reportsin the literature of the selenoxide elimination occurring in caseswhere the selenoxide and the J3—proton were anti to each other in thestarting material used for the reaction (for example, trans-2-(phenylselenenyl)-3-alkylcyclopentanones).92However, in suchcases, there was a proton a- to the selenoxide function. Thus, itwas believed that an in situ epimerization of the selenoxide functiongenerated occurred to give the syn arrangement of the selenoxideand the 13-proton and that then the normal elimination reaction tookplace. In the case of the selenoxide 168, a simple epimerization isimpossible, so a more complex process may be occurring. However,the mixture that was obtained from the oxidation of 166 was notH167 16870characterized further. The results of the oxidation of the selenide166 indicated that the presence of 166 during the oxidation!selenoxide elimination reaction of the major selenide epimer (165)would not be a significant problem.The enone 1 59 obtained from the selenoxide elimination wasnot stable to flash chromatography, so it was characterized withoutrigorous purification. In the 1H nmr spectrum, the signals for all thehydrogens of 159 were assigned based on decoupling experimentsand on the observations93 that in bicyclo[3.3.O]octanones, hydrogenson the convex face are deshielded relative to those on the concaveface and vicinal cis couplings between hydrogens a- to a ketofunction and the angular hydrogen are larger than the correspondingtrans couplings. Thus, the multiplet for the angular hydrogen (H-5b)at S 3.14-3.26 was coupled with the indicated coupling constants tothe signals at 2.25 (J = 4.0 Hz, H-4a), 2.79 (J = 6.5 Hz, H-4b), 1.48(J = 12.5 Hz, H-6a) and 2.70 (J = 8.0 Hz, H-6b). The geminalcouplings for H-4a and H-4b and for H-6a and H-6b were 18.0 Hz and12.5 Hz, respectively. The signal for the hydrogens at the 8-positionwas a broad singlet at 3.30 (2H). The methyl ester signal was at 610 9MeO2CHb Hb159713.85, while the resonances due to the tertiary methyl groups were at0.94 and 1.09 (s, 3H each). Two carbonyl absorptions were observedat 1750 and 1719 cm in the ir spectrum and were due,respectively, to the cyclopentenone and ester carbonyl stretches.61The molecular formula, C15H2005was consistent with the ion foundat 280.1311 in the high resolution mass spectrum.The key enone intermediate 159, representative of thegeneralized structure 74, (see p. 60) was one of the subtargets inthe synthesis of (±)-J3-panasinsene (31). Also, the enone 159appeared to be more suitable for the key methylenecyclohexaneannulation sequence than the previously synthesized enones 128 and129 (Scheme D-7, p. 51) had proved to be. Methylenecyclohexane Annulation on the Enone 159.The vinylstannane 5, prepared by a modificationb of previouslyreported procedures,5 was dissolved in THE and transmetallated at-78°C with a solution of methyllithium. The resulting vinyllithiumspecies 6 was transformed into the corresponding Grignard reagent7 by the addition of solid magnesium bromide etherate. Thesuccessive addition of the copper (I) bromide-dimethy) sulfidecatalyst (0.25 equiv) and a THE solution of the enone 159 gave anorange suspension which was stirred at -78°C for 25 minutes. Afteran appropriate workup, a mixture of the keto ester chloride 169 andits enol tautomer 170 was obtained in 94% yield (equation D-29). Itwas gratifying and not unexpected to find that the enone 159 was72SnMe3 MgBrvery reactive towards the conjugate addition reaction of theGrignard reagent 7. As seen earlier, similar reactions of enoneshaving tn- or tetrasubstituted double bonds were more sluggish andrequired either longer reaction times (see equation D-18) or thepresence of additives (see Schemes I-i and 1-3).The crude product mixture of 1 69 and 1 70 was not stable topurification by flash chromatography. Therefore, apart from a rapidfiltration through a short silica gel column to remove inorganic and1591) 7, CuBrSMe2THF, -78°C(25mm)2) aq NH4CI, rt169 (D-29)HO17073very polar organic material, it was characterized withoutpurification. A strong absorption characteristic of an enolized f3-dicarbonyl group61 was displayed at 1657 cm-1 in the ir spectrum.The keto and ester carbonyl stretches at 1754 and 1722 cm-1 wererelatively weak, which indicated that the enol tautomer 1 7 0predominated. The 1H nmr spectrum (CDCI3) also indicated that theenol 170 was the predominant component of the mixture. Thus, thesignal for the enol OH (br s at 3 10.79) integrated for -‘1 hydrogen.That the conjugate addition had been performed was furtherconfirmed by the presence in the 1H nmr spectrum of signals due tothe hydrogens a- to the chloride (part of a multiplet at 33.42-3.62)and due to the olefinic hydrogens at 64.74 and 4.82 (s, 1H each).With the keto ester mixture 169/170 in hand, the second stepin the methylenecyclohexane annulation, namely the cyclization ofthe keto chloride to obtain a tricyclic ketone, was performed.Surprisingly, the cyclization was problematic and, before successwas achieved, a variety of methods were tested (i.e., KH/THF/45°C;K2C03/ acetone! 50°C;94 K2C031 2-butanone! 80°C; K2C03/ 3-pentanone! -100°C; one-pot conjugate addition! cyclizations in thepresence of HMPA95). Ultimately, it was found that the cyclizationof the crude keto ester chloride 1 69/170 worked best in hot(-‘60°C) acetonitrile using cesium carbonate96 (5 equiv) as the base(equation D-30). The product thus generated was more pure and wasproduced in a better yield than that obtained from the otherprocedures. After chromatographic purification and arecrystallization, the tricyclic keto ester ketal 171 was obtainedas colorless crystals in 64% yield and exhibited m.p. 127.5-128.5°C.74Cs203,CH3N, 0 (D-30)—60°C (20h)The tricyclic keto ester ketal 171 was further characterized byir spectroscopy, mass spectrometry (high and low resolution), anelemental analysis, and nmr spectroscopy (1H and 13C, one and two97dimensional experiments). Thus, in the ir spectrum, very strongabsorptions were observed at 1745 and 1723 cm-1 due to the ketoand ester carbonyl stretches, respectively, and at 1115 cm due tothe ketal C-C bonds. Weak absorptions due to the exocyclic olefinicmethylene appeared at 1637 and 893 cm1. The high resolution massspectrum indicated that the molecular ion had a mass of 348.1929mass units which is consistent with the molecular formula,C20H805. In the low resolution mass spectrum, fragmentscorresponding to the loss of MeC (M-31) and CO2Me (M-59) wereobserved. The analytical data also was in keeping with the formula.The nmr spectrum of 171 (see figure 2) showed the expectedthree singlets (3H, each) due to the tertiary methyl groups (Me-19and Me20)* at 8 0.91 and 1.01, and to the methoxycarbonyl group(Me-14’) at 3.74. The signals due to the two olefinic hydrogens werefound at 8 5.01 (H-15b) and 5.07 (H-iSa) and were, respectively, a* Note: The numbering system employed is based on the numbering scheme used for $-panasirisene (see Scheme D-1, compound 31). Thus, the position numbered C(H)-13in the various tricyclic intermediates synthesized ultimately becomes the correspondingmethyl group (Me-13) in the synthetic $-panasinsene (31).169 and 170 17175TMSI I III II I II III II I 11111 III I II II II 1111111 II P5 4 3 2 1 OPPMFigure 2. The 300 MHz 1H nmr spectrum of the keto ester ketal 171.doublet (J = 1 .0 Hz) and a singlet. The methine hydrogen (H-4b)appeared as a multiplet at ö2.76-2.84 (1H). The remaining aliphatichydrogen signals also were displayed in the spectrum and are listedtogether with the COSY (1H-’H COrrelation ipectroscop..)correlations in Table 2.76Me20From a 13C nmr APT (A.ttached Eroton jest) experimentperformed on 171, the signals for the carbons corresponding to themethyl and methine groups could be assigned as follows; 822.23 and22.32, the tertiary methyl groups (Q..H3-19 and H3-20); 52.1, themethyl ester (.H3-14’); and 38.65, the methine (H-4). Some of thesignals due to the other carbons could be assigned based on theirchemical shifts. Thus, the signal for the olefinic methylene (H2-15) resonated at 3 112.7; the signal for the quaternary olefiniccarbon (C-li) was at 145.1; and the signals for the keto (C-6) andester (C-14) carbonyl groups were at 211.8 and 170.5, respectively.The rest of the carbon signals along with the HETCOR (1H-’3C,HETeronuclear Shift CORrelation) correlations are listed in Table 1.Based on the HETCOR and COSY data, all the hydrogen signals inthe ‘H nmr spectrum and most of the carbon signals in the ‘3C nmrspectrum could be reasonably assigned. Thus, a HETCOR experiment(see figure 3) confirmed the above ‘H and ‘3C assignments andpermitted the determination of the pairs of geminal hydrogens. Forexample, the carbon signal at 842.8 (H2-2) correlated withhydrogen signals at 1.88 (d, 1H, J = 16.0 Hz, H-2a) and 2.95 (br d, 1H,HbHaHb’0:Hb Hb17177J = 16.0 Hz, H-2b); the carbon signal at 38.74 (H2-3) correlatedwith hydrogen signals at 1.96 (distorted dd, 1H, J = 14.0, 6.0 Hz, H3a) and 2.10 (distorted d, 1H, J = 14.0 Hz, H-3b); and the carbonsignal at 32.4 (...H2-10) correlated with only the 2-hydrogenmultiplet at 2.26-2.39 (H-lOa and H-lob). The other pairs ofgeminal hydrogens were assigned in a similar manner, but thedetermination of which pair was at which position was based on theobserved COSY correlations.F2 (PPM)1401301201101009080706050403020I I I I I I I I5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 O.5(PPM)CDCI3Figure 3. The 75 MHz HETCOR spectrum of the keto ketal ester 171.78Table 1: The 75 MHz HETCOR Data for the Keto Ester Ketal 171.19MeMe20Position 13C (75 MHz) 1H (300 MHz) 5 ppm (H-x)(C-x) Sppm1 58.2a- -2 42.8 1.88 (2a); 2.95 (2b)3 38.74 1.96 (3a); 2.10 (3b)4 38.65 2.76-2.84 (4b)5 39.9 2.54 (5a or 5b); 2.69 (5b or 5a)6 211.8- -7 68.Oa - -8 30.6 1.54-1.71 (8a or 8b); 2.15-2.21 (8b or 8a)9 23.8 1.72-1.80 (9a or 9b); 1.54-1.71 (9b or 9a)1 0 32.4 2.26-2.39 (lOa and lOb)11 145.1- -13 109.0- -14 170.5- -14’ 52.1 3.74 (3H-14’)15 112.7 5.07 (15a); 5.01 (15b)1 6/1 8 71.8/72.4 3.40-3.57 (2H-1 6/2H-1 8)17 30.0- -1 9/20 22.23/22.32 0.91/1.01 (3H-1 9/3H-20)HbHa171a. Assignments may be interchanged.79In the COSY spectrum of the keto ester ketal 171 (see figure 4),the spin system of the six-membered ring was separable frominterrelated ones of the two five-membered rings. Thus, key entrypoints into the two main spin systems were the signals for theolefinic hydrogens at 85.01 (d, 1H, J = 1.0 Hz, H-15b) and 5.07 (s, 1H,H-15a) which showed correlations in the six-membered ring systemand the signal for the methine hydrogen at 2.76-2.84 (m, 1H, H-4b)which led into the five-membered ring systems. The processfollowed in making the assignments of the hydrogen positions by useof the COSY correlations is exemplified for the five-membered ringsystems. The signal for the methine hydrogen (H-4b) at 62.76-2.84showed strong correlations (large cross peaks) to the distorteddoublets at 1.96 (1H), 2.54 (1H), 2.69 (1H) and weaker correlationsto the signals at 2.10 (distorted d, 1H) and at 2.95 (br d, 1H). Fromthe HETCOR results (see Table 1), the signals (a) at 6 2.95 and 1.88,(b) at 1.96 and 2.10 and (c) at 2.54 and 2.64 were due to pairs ofgeminal hydrogens. Since the signal at 62.95 also showed a weakcorrelation to the signal for the olefinic hydrogen at 5.07 (H-15a), itseemed reasonable to assign the signal at 2.95 to H-2b and thus, theone at 1.88 to H-2a. The signals at 6 2.54 and 2.64 were thenassigned to the hydrogens at position 5, a- to the ketone carbonyl,while the signals at 1.96 and 2.10 were assigned to H-3a and H-3b,respectively, a- to the ketal function with the assignment of the aand b-hydrogens based largely on reported observations of themagnitudes of relevant coupling constants in other systems.93 Thepairs of geminal hydrogens on the six-membered ring were assigned80to the appropriate positions starting from the cross peaks shown bythe two olefinic hydrogens at 85.01 (H-15b) and 5.07 (H-15a).,—-—— —-—-—— -———e9..N—I———--.+.2.0 1.5 1.0Figure 4. The 400 MHz COSY spectrum of the keto ester ketal 171.--- 4.5—4.0 3.5 2.55.5PPM81Table 2: The 400 MHz COSY Data for the Keto Ester Ketal 171.Position Signal 8 ppm COSY Correlations(H-x) (multiplicity;a J; (H-x)number_of_H)2a 1.88 (d; 16.0; 1H) 2b2b 2.95 (br d; 16.0; 1H) 2a; 4b (W-coupling); l5ab3a 1.96 (distorted dd; 14.0, 3b; 4b6.0; 1H)3b 2.10 (distorted d; 14.0; 3a; 4bb1H)4b 2.76-2.84 (m; 1H) 3a; 3bb; 5b; 5a; 2b (W-coupling)5a or 5b 2.54 (distorted dd; 19.5, 5b or 5a; 4b9.0; 1H)5b or 5a 2.69 (distorted dd; 19.5, 5a or 5b; 4b9.0; 1H)8a or 8b 1.54-1.71 (m; 2H) 8b or Ba; 9b and 9a8b or 8a 2.15-2.21 (m; 1H) 8a or 8b; 9a and 9b; lOa or lOb(W-coupling)9a or 9b 1.72-1.80 (m; 1H) 9b or 9a; 8b or 8a; lOa and lOb9b or 9a 1.54-1.71 (m; 2H) 9a or 9b; Bb or 8a; lOa and lOblOa and 2.26-2.39 (m; 2H) 9a and 9b; 8b or 8a (WlOb coupling); 15a and 15b14’ 3.74 (s; 3H)- -15a 5.07 (s; 1H) 15b; lOa or lObb; 2bb15b 5.01 (d; 1.0; 1H) 15a; lOa or lObHbHaHb19MeHb171Hb82Table 2: continued.16 and 3.40-3.57 (m; 4H) 20 or 19 (W-coupling)1819 or 20 0.91 (s; 3H)- -20 or 19 1.01 (s; 3H) 16 or 18 (W-coupling)a. The signals labelled s, d and dd may incorporate unresolved fine couplings.b. Small couplings observed.The synthesis of 171 represented the fulfillment of one of thegoals of the project, that is, the synthesis of an enone with atetrasubstituted double bond and its .utilization in the methylenecyclohexane annulation sequence. It remained only to carry outvarious functional group manipulations to arrive at the projectedintermediate 72 on which to perform a ring contraction to assemblethe correct tricyclic carbon framework of the natural product. Preparation of a Substrate (Intermediate 72) for RingContraction.In order to prepare the desired ketone substrate 72 for theplanned ring contraction sequence, it was necessary to deoxygenatethe keto group (and perhaps the methoxycarbonyl group) in 171 andto convert the ketal group into a keto function.83A double deoxygenation of the keto and ester functions to givethe corresponding methylene and methyl groups was attempted first.Consequently, the keto and methoxycarbonyl groups of the keto esterketal 171 were reduced with lithium aluminum hydride in diethylether to give a mixture of the epimeric diols 172 and 173 (ratiovaried from -‘5:3 to >4:.<1 172:173, 1H nmr spectral analysis) in 80-95% yield (equation D-31). In the 1H nmr spectrum (300 MHz, CDCI3)of a sample containing both 172 and 173, the signals for theirrespective epimeric carbinol hydrogens were displayed at 8 4.57(distorted td, J = -‘8, -‘4 Hz, H-6a) and at 3.91-3.97 (m, H-6b), whichwere converted to a triplet (J = -‘8 Hz) and a less complex multipletupon the addition of D20. In the mixture, the signals for the CE2O Hhydrogens appeared at 8 3.57-3.63 and 3.76-3.90 (m, m, 2H total).The major epimer (172), which could be separated from theminor epimer by recrystallization of the mixture from 4:1 and then171LiAIH4,Et20,0°C (0.5h);rt (2h)172+(D-31)H17384from 7:1 petroleum ether-diethyl ether, exhibited m.p. 117-118°C.An ir spectrum of 172 (solution in chloroform, polystyrenereference) showed absorptions due to the hydroxyl group at 3630,the ketal group at 1125, and the olefin function at 1640 cm1. In a‘H nmr spectrum (300 MHz, CDCI3) of the major epimer (172), thesignals for the methylene hydrogens of the hydroxymethyl groupappeared at 33.79 (dd, J = -‘11, 5 Hz, H-14b) and at 3.61 (br dd, J =-.11, 5 Hz, H-14a), which were converted, respectively, to a doublet(J = -1 1 Hz) and a broad doublet (J = -1 1 Hz) upon the addition ofD20.* Further confirmation of the identity of diol 172 was obtainedfrom the mass spectra. Thus, in the high resolution mass spectrum,the molecular ion peak was found at 322.2136, which corresponds tothe molecular formula, C19H3004while in the low resolution massspectrum, peaks corresponding to the loss of one and two moleculesof water (M-18 and M-36, respectively) were observed.It was of interest to determine the stereochemistry at the 6-position of the diol 172. A priori, one would have trouble predictingthe stereochemical outcome of the reduction process, since bothfaces of the keto function in 171 are sterically hindered. In key nOedifference experiments (400 MHz, summarized in structure 172’) todetermine the stereochemistry of 172, irradiation of the signal at4.57 (H-6a) led to an enhancement of the signal at 3.61 (H-14a),while irradiation of the signal at 3.61 (H-14a) led to enhancementsof the signals at 3.79 (H-14b) and 4.57 (H-6a). Thus, it seemed* Signals due to the other hydrogens appeared in the 1H nmr spectrum at 3 0.91 (s, 3H),1.01 (s, 3H), 1.32-1.61 (m, 2H), 1.66-1.95 (m, 5H), 1.99-2.13 (m, 4H), 2.26-2.40 (m, 3H), 2.71-2.79 (m, 1H), 3.42 (s, 2H), 3.44-3.56 (m, 2H), 4.82 and 4.85(s, s, 2H total).85reasonable to conclude that H-6a was cis to the hydroxymethyl groupand that the stereochemistry at C-6 of the major product 172 wasas assigned.Various attempts to carry out the double deoxygenation reactionof both hydroxyl functions of 172 and 173 via derivatives of thediols were unsuccessful. Krishnamurthy and Brown have reported asuper hydride (lithium triethylborohydride) deoxygenationprocedure98 utilizing the p-toluenesulfonate derivatives of alcohols.Attempts to convert the diol mixture to the correspondingbissulfonates by reaction with p-toluenesulfonyl chloride (p-TsCl, 3equiv or 2.2 equiv) in the presence of a base (4-(N, Ndimethylamino)pyridine (DMAP), 3.5 equiv or KH, 3 equiv) in asuitable solvent (dichloromethane or THE) were unsuccessful whenan unstable mixture of the mono- (mainly the primary sulfonate) andbissulfonates, unreacted diol and other uncharacterized materialswas produced. The sterically hindered nature of the secondaryalcohol was also implicated when the reaction of an acetonitrilesolution of the diol 172 with phenoxythiocarbonyl chloride (PTC-Cl,H172862.3 equiv) in the presence of DMAP (4.2 equiv) generated a complexmixture of products. The failure to form the bis-O -phenoxythiocarbonyl derivative of the diol 172 in a reasonable yieldthwarted plans to employ the Robins radical deoxygenation reactionwith tri-n-butyltin hydride.99Sterically hindered alcohols have been converted into theirN,N,N’, N’-tetramethylphosphorodiamidate derivatives by a proceduredeveloped by Liu and coworkers.10° Thus, the bisalkoxide preparedfrom the did 172 by its reaction with n-butyllithium was treatedwith N,N-dimethylphosphoramidic dichloride (174) followed bydimethylamine (equation D-32). A mixture of products resulted inwhich the cyclic N,N-dimethylphosphoramidate 175 predominated.The structure of 175 was deduced from its 1 H nmr spectrumwherein the signals for the olefinic hydrogens at 3 4.85 and 4.88 (s,1H each), the methyl groups on the nitrogen at 2.76 (d, 6H, J = 10Hz, overlapped with a multiplet, 1H, at 2.69-2.9) and the tertiarymethyl groups at 0.94 and 0.97 (s, s, 6H total) were in a 2:7:6 ratio.*The attempted double deoxygenation (lithium metal/ methylamine/0°C) of 175 failed,12 but the results indicated (‘H nmr spectralanalysis on the crude mixture) that the derivatized secondaryhydroxyl function had been deoxygenated in preference to theprimary hydroxyl group to generate a compound formulated as 176.* Other hydrogen signals were also observed: the methine hydrogen (H-6a) at o 5.09 (brt, 1H, J = —9Hz); the methylene hydrogens (2H-14) at 4.36 (br d, 1H, J = —10 Hz),and 4.15 (dd, 1H, J = —10, —22 Hz); the ketal methylene hydrogens as singlets at 3.52and 3.42 (2H each); the rest of the aliphatic hydrogens (12H) as multiplets at 1.4-2.4.871) n-BuLl (2.5 equiv),TMEDA:DME (1:4),0°C (20-30mm) 0.2) (Me)NP(0)C1, Me2NP(174, —10 equiv)0°C — rt (—19h)3) Me2NH (—25 molarequiv), 0°C (4h)It was apparent that a sequential deoxygenation of the keto andester functions would be preferable as the reactivities of theprimary and secondary hydroxyl groups were significantly different.Consequently, it was decided to postpone the deoxygenation of themethoxycarbonyl group until the end of the synthesis and todeoxygenate it together with another methoxycarbonyl which wouldbe present at that stage. Thus, the keto function in 171 wasreduced selectively with sodium borohydride in methanol (with asmall amount of ethyl acetate to solubilize the keto ester ketal171). An ‘-2-3:1 mixture of the epimeric alcohols 177 and 178 wasproduced, which proved to be difficult to separate by chromatography on silica gel or by fractional recrystallization. Forsubsequent reactions it was desirable to have exclusively the majorepimer (177) present (vide infra), so a variety of reducing agents172H (D-32)175 andother products017688were tested to see if the ratio of epimers could be improved infavour of the major epimer. Reduction procedures attemptedincluded: zinc borohydride in diethyl ether, which is known to reducevia chelation;101 L-selectride in THF, which is expected to approachfrom the less hindered face of the carbonyl;102 and sodiumborohydride with varying amounts of cerium trichloride hexahydrateat various temperatures in methanol (Luche’s reagent).103Reductions using Luche’s reagent were the most consistent in termsof selectivity and yield. Therefore, treatment of a methanolicsolution of 171 with sodium borohydride (1.3 equiv) in the presencecerium trichioride hexahydrate (0.53 equiv) at -48°C (equation D-33)led to the production of a mixture of the epimers 1 77 and 178 in-98% yield. The ratio of epimers (177 to 178) in the crude mixtureof alcohol esters using Luche’s reagent was usually -5:1 with a ratioas good as 12:1 obtained upon occasion depending on smallvariations in the reaction conditions. (The ratios were based on therelative integration area for the signals due to H-6a versus H-6b inthe 1 H nmr spectrum). Recrystallization of the mixture from ethylacetate-hexane generally improved the ratio slightly.89MeO2C1HOiNaBH4, CeCI3 6H20MeOH, -48°C (lh)The ir spectrum of an ‘-12:1 mixture of 177 and 178 displayed astrong absorption at 3511 cm1 indicative of the presence of thealcohol function, while a very strong band at 1726 cm1 wascharacteristic of the ester carbonyl group. In the 1 H nmr spectrumof the mixture, the signal due to the methoxycarbonyl group of bothepimers appeared at 83.68 (s, 3H). The carbinol hydrogen (H-6a)adjacent to the hydroxyl group of 1 77 resonated as a triplet ofdoublets (J = 9.0, 2.0 Hz) at 84.69 which simplified to a triplet (J =9.0 Hz) upon the addition of D20. On the other hand, H-6b of theepimer 178 appeared as a multiplet at 84.11-4.14, which simplifiedsomewhat upon the addition of D20. The elemental analysis and highresolution mass spectral analysis (performed on the mixture of 177and 178) provided results consistent with the molecular formula,C20H305.The stereochemistry at the 6-position was not determined foreither the alcohol ester 177 or the alcohol ester 178. However,0:171177+(D-33)H17890assuming that the relative chemical shifts and the signalmultiplicities of the hydrogen at the 6-position of the diol epimers172 and 173 (vide supra, p. 83) and those of the alcohol esters 177and 178 are comparable, then the relative stereochemistries of thealcohol esters 177 and 178 may be assigned. Thus, the resonance ofthe 6-hydrogen of the major diol epimer 172 was downfield fromthat of the 6-hydrogen in the minor epimer 173 and the signalsappeared as a distorted triplet of doublets (J = 8, —4 Hz) and amultiplet, respectively. Similarly, the chemical shift of 6-hydrogenof the major alcohol ester epimer (177) appeared downfield fromthat of the minor epimer 1 78 and the signals were a triplet ofdoublets (J = 9.0, 2.0 Hz) and a multiplet, respectively. Thus, itseemed reasonable that compound 177 had the same configuration atthe 6-position as did the diol 172. Further evidence for thedepicted stereochemistry came from the fact that the epimer 177was the major product in reductions using zinc borohydride, areagent which tends to give products from reduction viachelation.101 In the case of keto ester ketal 171, reduction viachelation with the ester function would be expected to givepreferentially an alcohol possessing the configuration found at the6-position of the compound 177.The deoxygenation of the secondary alcohol function in thealcohol esters 177 and 178 was performed using the radicaldeoxygenation of the corresponding O-phenoxythiocarbonylderivative (PTC derivative) prepared via a modification of theprocedure developed by Robins and coworkers.99 The typicalprocedures utilized by Robins and coworkers to convert secondary91alcohol groups into their PTC derivatives were to stir the alcoholwith phenoxythiocarbonyl chloride (PTC-Cl) either in the presence ofpyridine (3-4 equiv) in dichloromethane (2 hours) or, for morehindered alcohols, in the presence of 4-(N,N-dimethylamino)pyridine(DMAP, 2 equiv) in acetonitrile at room temperature (16 hours). Inthe most sluggish case they reported, 6-9 equivalents of DMAP wererequired. The PTC derivatives were then deoxygenated using tn-nbutyltin hydride in the presence of a free radical initiator in warmtoluene (75°C or at reflux, 3 hours).The conditions utilized for the preparation of the PTCderivatives of alcohol esters 177 and 178 were more drastic thanthe most severe case reported by Robins and coworkers.99 Thus, asolution of a mixture of 177 and 178 (ratio -5:1) in acetonitrilewas converted into a mixture of the corresponding PTC derivativesby treatment of the former material with PTC-Cl (1 .5 equiv) in thepresence of DMAP (8 equiv) at -70°C for 20 hours (equation D-34).The major epimer 179 was obtained in 76% yield, while the minorepimer 180, which was difficult to separate from an impurity, wasnot characterized nor used in further reactions. It was due to thedifficulty in purifying the minor epimer that the effort was made toobtain a good stereoselectivity in the reduction of the keto esterketal 171 to the alcohol esters 177 and 178.92PhOC(S)-CI P TDMAP, CH3N,-70°C (20h)The ir spectrum of the phenyl thionocarbonate 179 displayedabsorptions due to the carbonyl group at 1736 cm-1, the exocyclicmethylene group at 1639 and 888 and a monosubstituted aromaticring at 774 and 690 cm* In the ‘H nmr spectrum, the signal for themethoxycarbonyl group appeared as a singlet at 8 3.70 (3H), thesignal for H-6a was a triplet at 6.08 (J = 8.5 Hz, 1H), and thearomatic hydrogens gave rise to multiplets at 7.09-7.12 (2H), 7.25-7.30 (1H) and 7.38-7.43 (2H). The elemental analysis and highresolution mass spectral data were in accord with a molecularformula ofC27H3406S.The PTC derivative 1 79 was deoxygenated according to amodification of the procedure reported by Robins and coworkers.Thus, a solution of 179 in benzene (instead of toluene) was treatedwith tri-n-butyltin hydride (2.5 equiv) and the radical initiator,2,2’-azobisisobutyronitrile (AIBN, 0.18 equiv), and was heated underan argon atmosphere at 77°C for 20 hours (equation D-35). Thepurified deoxygenated product 181 was obtained in 72% yield.HH177 and 178(D-34)H179 and smallamount of epimer(180)93(D-35)‘20Recrystallized 181 (m.p. 87.5-89.5°C) showed the expectedabsorptions in the ir spectrum for the ester carbonyl at 1719, forthe ketal C-O at 1116 and for the exocycNc methylene group at 1638and 892 cm1. The H nmr spectrum, as expected, still showed thepresence of signals for the tertiary methyl groups as singlets at 60.95 and 0.97 (3H, each), the methoxycarbonyl function as a singletat 3.63 (3H) and the olefinic hydrogens as singlets at 4.87 and 4.92(1H each). The 13C nmr displayed the required 20 signals, includingone for the ester carbonyl carbon (C-14) at 3 176.3, the quaternaryolefinic carbon (C-il) at 149.1 and the olefinic methylene (.H2-15)at 109.8. From an APT experiment, it was possible to assign thetertiary methyl groups to the signals at 322.3 and 22.4 (.H3-19 and,H3-20), the methyl of the methoxycarbonyl group to the signal at51.2 (H3-i4’), and the methine (H-4) to the signal at 45.0. Theresults from the high resolution mass spectrum (found M=334.2141)n-Bu3SnH,AIBN, PhH—77°C (20h)179 1811916H18194and the elemental analysis were consistent with a molecularformula C20H thus further confirming that the 0-phenoxythiocarbonyl group had been replaced with a hydrogen. In thelow resolution mass spectrum the molecular ion and a fragment dueto the loss of the methoxycarbonyl group (M-59) were among thepeaks observed.The keto ester substrate for the ring contraction reactionsequence was prepared by deprotection of the ketal function in theketal ester 181. A procedure similar to the one reported by Mossand Piers38’85 was utilized. Thus, an acetone solution of the ketoketal 181 was treated with iN hydrochloric acid (0.5 equiv) at roomtemperature for 5.5 hours (equation D-36). After an appropriateworkup and purification, the keto ester 182 was obtained in -85%yield (equation D-36).The 1H nmr and ‘3C nmr spectra confirmed that the ketal groupof 181 had been removed. Thus, the signals for the tertiary methylgroups and the ketal methylene groups were not seen in the 1H nmrspectrum and the rest of the signals were consistent with thestructure of 182. Similarly, in the 13C nmr spectrum the ketalcarbon signals were replaced by a ketone carbonyl carbon signal (C-MeO2CH1811 N HCI (0.5 equivacetone, rt (5.5h)0 (D-36)1829513) at 6219.2 and the expected 19 other carbon signals required bythe structure of 182 were present. Thus, the ester carbonyl carbonsignal (C-14) resonated at 6 175.7, the olefinic methylene carbonsignal (.H2-15) was at 109.8 and the quaternary olefinic carbonresonance (C-li) appeared at 148.4. Based on a ‘3C nmr APTexperiment, signals at 642.1 and 51.6 could be assigned to H-4 andH3-i4’, respectively. Additional confirmation of the molecularformula, C,5H2003was provided by the low and high resolution massspectra as well as an elemental analysis.Further details of the molecular structure were obtained fromthe one and two dimensional ‘H nmr spectra and from decouplingexperiments. Thus, in the one dimensional ‘H nmr spectrum (seefigure 5), the signal due to the bridgehead methine (H-4b) wasreadily assigned to the multiplet at 6 2.84-2.92 (iH), the signal forthe methoxycarbonyl group (Me-i4’) appeared at 3.64 (s, 3H), and thesignals for the olefinic hydrogens were at 4.65 and 4.87 (s, 1H each,H-i5a and H-i5b, respectively). In decoupling experiments,irradiation of the signal at 6 2.88 (H-4b, the center of the multiplet)18296TMSFigure 5. The 300 MHz 1H nmr spectrum of the keto ester 182.PPMsimplified the doublet of doublets at 2.73 (J = 19.0, 1.0 Hz, 1H) to adoublet (J = 19 Hz), sharpened the multiplet at 2.25-2.42 (5H) andsimplified the multiplet at 1.38-1.50 (1H). Irradiation of the signalat 32.73 simplified the doublet of doublets at 2.08 (J = 19.0, 1.0 Hz,1H) to a doublet (J = 19 Hz) and the doublet of doublets at 2.19 (J =19.0, 0.5 Hz, 1H) to a broad singlet. Irradiation of the doublet ofdoublets at 32.08 sharpened the two doublets of doublets at 2.19 and2.73 and simplified the multiplet at 2.25-2.42. It was clear,F5 4 3 2 197therefore, that the signals at 82.73 and 2.19 were due to geminalhydrogens. However, neither signal showed a large enough couplingto H-4b to be due to hydrogens at the 3-position, hence the signalsat 82.73 and 2.19 were caused by hydrogens at the 2-position (H-2band H-2a, respectively). The signal at 8 2.08 also did not showcoupling to H-4b, but it and the one due to H-4b were coupled to partof the multiplet (5H) at 2.25-2.42. Consequently, the assumptionthat vicinal cis couplings between hydrogens a- to a keto functionand the angular hydrogen are larger than the corresponding transcouplings93 (see also p. 69) led to the assignment of the signal dueto H-3a to 62.08 and the signal due to H-3b to the multiplet at 2.25-2.42.The two dimensional 1H- COSY nmr spectrum (400 MHz)confirmed the above assignments and made it possible to assigntentatively most of the other hydrogen signals (see figure 6 andTable 3). As in the case of the COSY spectrum of the keto esterketal 171, the signals for the bridgehead methine (H-4b) and theolefinic hydrogens (H-15a and H-15b) of 182 provided key entriesinto the two spin systems of the five-membered rings and of thesix-membered ring, respectively. Thus, the signal assigned to H-4bat 8 2.84-2.92 showed correlations to the signals corresponding toH-2b (at 2.73, long range coupling), H-3b and H-5b (both part of themultiplet at 2.25-2.42) and H-5a (at 1.38-1.50). In turn, the signalat 6 1 .38-1 .50 (H-5a) showed other correlations to the signals at1.84 (br tt, 1H, J = 12.5, 3.5 Hz) and at 2.25-2.42 (m, 5H). Therefore,of the five hydrogens giving rise to the multiplet at 2.25-2.42, threewere assignable to H-3b, H-5b, and H-6a or H-6b. The signals for98the olefinic hydrogens at 34.65 (H-15a) and at 4.87 (H-15b) alsoshowed correlations to the multiplet at 2.25-2.42, so H-lOa and/orH-lOb were part of the multiplet. The remaining signals, at 3 1.52-1.65 (1H), 1.67-1.73 (1H), 1.78 (1H) and 1.98 (1H), could be assignedto the hydrogens of the six-membered ring, but specific assignmentswere not made. The partial assignments are listed in Table,-,i ::I I::e rp..p --5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5Figure 6. The 400 MHz COSY spectrum of the keto ester 182.99Table 3: The 400 MHz COSY Data for the Keto Ester 182.0. Position Signal S ppm COSY Correlations(H-x) (multiplicity;a J; number (H-x)of_H)2a 2.19 (dd; 19.0, 0.5; 1H) 2b; 3a (W-coupling)2b 2.73 (dd; 19.0, 10; 1H) 2a; 3b (W-coupling)b; 3a (Wcoupling); 4b (W-coupling)3a 2.08 (dd; 19.0, 1.0; 1H) 3bb; 2a (W-coupling); 2b (Wcoupling)3b 2.25-2.42 (m; 5H) C4b 2.84-2.92 (m; 1H) 2b (W-coupling); 3bb; 5a; 5bb5a 1.38-1.50 (m; 1H) 5bb; 4b; 6a; 6bb5b 2.25-2.42 (m; 5H) C6a or 6b 1.84 (br if; 12.5, 3.5; 1H) 6b or 6ab; 5a; 5bb6b or 6a 2.25-2.42 (m; 5H) Cd e flOa and/or 2.25-2.42 (m; 5H) ClOb14’ 3.64 (s; 3H)- -15a 4.65 (s; 1H) 15b; lOa and/or lObb15b 4.87 (s; 1H) 15a; lOa and/or lObba. The signals laoelled s, d, dd, may incorporated unresolved fine couplings.b. The hydrogen is part of the signal at 32.25-2.42 (m, 5H).c. The signal at 8 2.25-2.42 (m, 5H; 3b, 5b, 6b or 6a, lOa and/or lOb) showedcorrelations to 2b, 3a, 4b, 5a, Ga or 6b, as well as to the signals at 8 1.52-1.65, 1.67-1.73 and 1.98.d. The signals for 8a, 8b, 9a and 9b were not specifically assigned.e. Signals were observed at 1.52-1.65 (m, 1H), 1.67-1.73 (m, 1H), 1.78 (distorteddd, 1H, J = 13.5, 3.5 Hz) and 1.98 (dm, 1H, J = 13.5 Hz).f. The correlations were not determined.182100The synthesis of the keto ester 1 82 meant it was possible tostudy the one-carbon ring contraction of the functionalized fivemembered ring. Application of such a ring contraction to 182 wouldgenerate a carbon framework with ring sizes corresponding to thosein the target natural product. Ring Contraction to Give a 4-5-6 Tricyclic Carbon Skeleton.There are many different procedures available for performingone-carbon ring contractions.104 One common method employed isthe Wolff rearrangement of cyclic a-diazo ketones 183 (equation D37).105 The reaction can be performed thermally, photochemically orcatalytically. Generally, the mechanism for the photochemicalprocess is thought to involve the formation of a singlet carbene 184via the loss of a molecule of nitrogen in a first order rate process.The group a- to the carbonyl group migrates with its electron pair tothe carbene site and a ketene 185 is generated. If a nucleophilesuch as water, an alcohol, ammonia or an amine is present, it canadd to the ketene to produce the corresponding acid, ester or primaryor secondary amide.101heat/hv 6: p (D-37)catalyst(H2C)184 185A variety of procedures exist for the preparation of the desireda-diazo carbonyl compounds, but one of the most widely usedmethods is the “deformylation diazo group transfer”.106 In thisprocedure, a second activating group, usually a formyl group, isintroduced a- to the carbonyl group of a ketone (general formula,186) to give 187. During the introduction of the diazo group from adiazo transfer reagent (for example, p-toluenesulfonyl azide, pTsN3’°7)to give an a-diazo ketone 188, the formyl group is cleavedto form an amide, 189 (equation D-38).HCOORJL Hbase JL.N2 (D-38)-HC ,p-TsNH186 187 189 188Two main procedures have been developed to perform the diazogroup transfer reaction.108 In one method, the sodium salt 190resulting from the Claisen condensation of a carbonyl compound 186with alkyl formate/sodium alkoxide is treated with p-toluenesulfonyl azide (p-TsN3) to give the a-diazo ketone 188 (equation D39). In the second method, which is more commonly utilized forcycloalkanones, the tautomeric mixture of 191 and 192 is allowed183102to react with p-TsN3 in the presence of an organic base to generate183 (equation D-40).0HCOOftHp-TsN3 (D-39)R 0 Na186___H p-TsN3,base (D-40)183Diazo group transfer reagents other than p-TsN3 have been used(for example, p-carboxybenzenesulfonyl azide109, 2,4,6-triiso-propylbenzenesulfonyl azide° and methanesulfonyl azide111), butthe use of p-TsN3 has been most widespread. The diazo transferprocess may be described as the attack of the anion of theformylated carbonyl compound on a diazo transfer reagent consistingof N2 attached to a leaving group.’°9 The formyl group istransferred to the sulfonyl leaving group to give a sulfonamide (i.e.,189) while the diazo group is transferred to the carbonyl compoundto give the a-diazo carbonyl compound (188 or 183). For a-formylcycloalkanones 191 or their hydroxymethylene tautomers 192,diazo group transfer likely goes via the cyclic triazoline 1 9 3(Scheme D-11).1 2 Decomposition of the triazoline 193 can occur bytwo different pathways depending on the ring size.°6’12 The x01900188HO191 192103diazo cycloalkanone 183 and N-(p-toluenesulfonyl)formamide 189,products of the desired decomposition, are formed mainly by path aand are the only products observed for 5-, 7-, and 8-membered rings.It is also possible that the cz-diazo cycloalkanone 183 may beformed by the sequence 193 — 194 - 183. On the other hand, theformation of the p-toluenesulfonyl-2-oxo-cycloalkylcarbonamide196 via path b occurs to a certain extent for 6-, 9-, 10-, 11-, and12-membered rings. Thus, loss of dinitrogen from the intermediate194 is followed by a rearrangement of the intermediate 195 toproduce the amide 196.Application of the deformylation diazo group transfer process tothe keto ester 182 in order to prepare an a-diazo cycloalkanone forthe Wolff ring contraction required the synthesis of the formylatedketone 197. A variety of attempts to perform the formylation (KH/ethyl formate/ THE; NaH/ ethyl formate/ THE; NaH/ ethyl formate/diethyl ether;’3 NaH/ methyl formate/ diethyl ether;114 NaOMe/methyl formate/ diethyl ether; KH/ methyl formate/ THF), resultedeither in the formation of the product in low yields or in therecovery of the starting ketone 182. In the end, treatment of abenzene solution of the keto ester 182 with sodium t-amyloxide’’5(4 equiv) at room temperature, followed by the addition of methylformate (8 equiv) (equation D-41) led to the quantitative formationof the keto aldehydes 197a and 197b and the enol tautomer 198(ratio 1 :traces:8.5). The products 197 and 198 were not stable topurification by chromatography. Hence the mixture wascharacterized with traces of impurities present.104Hp-TsN3,CH2I,Et3NNPatha0(CI + HC’ p-TsNH‘2183 189p-TsPath b -N2N—p-TsNH—p-Ts195 196HO191 1H19H194Scheme D-11105MeO2C4Hb1) t-amyIONa, -7°C; 197a R=CHO, R’=HPhH, rI (1.5h)+ 197b R=H, R’=CHO (D-41)2) MeOCH,5°C —+ rt (17h) +An ir spectrum of the mixture of 1 97 and 1 98 indicated thepresence of the ester carbonyl, and the enolized 13-keto aldehydegroup with absorptions at 1725, and 1699, 1609 (broad) cm,respectively.61 Based on the H nmr spectrum of the formylatedmixture, the formylation of 182 occurred at the sterically lesshindered 3-position rather than at the 2-position and gave mainlythe enol 198, as evidenced by the change in the chemical shift forH-4b (8 2.84-2.92 in 182 and 3.20-3.23 in 198) and the presence inthe spectrum of 1 98 of a signal at 6 7.08 (s, 1 H) for the olefinichydrogen of hydroxymethylene function. The signal for the aldehydehydrogen of the major aldehyde epimer 197a (stereochemistry basedon steric considerations) was at 39.52 (s), while the presence oftraces of 1 97b were indicated by a singlet at 9.83. In the lowresolution mass spectrum, the molecular ion was found at 276 andfragments corresponding to the loss of CO (M-28), MeOH (M-32) andR’182198106CO2Me (M-59) were present; the latter fragment was the base peak.In the high resolution mass spectrum, the mass found for themolecular ion of 197/198 (M=276.1363) was consistent with theexpected formula, C16H2004.Initially, the a-diazo ketone 199 was prepared from theformylated keto/enol esters 197/198 using pTsN3* indichloromethane with triethylamine as the base.’ 16 Unfortunately,it proved impossible to separate unreacted p-TsN3 from the diazoketone, a problem which has been noted before.’11 Despite thepresence of the unreacted p-T5N3, the a-diazo ketone 199 wassubjected to the photochemical Wolff rearrangement inmethanol.3,Gb The reaction proceeded, albeit in very poor yields(<25% of the ring contracted product was obtained from thephotolysis), and an uncharacterized aromatic by-product proveddifficult to remove. The expected by-product, the amide 189 (seeequation D-38 or Scheme D-11), was obtained which indicated thatthe diazo transfer was indeed occurring.Taber and coworkers found that the use of methanesulfonylazide (MsN3) was advantageous”1 due to the fact that unreactedM5N3 can be separated from the a-diazo carbonyl compound bywashing an organic solution of the diazo compound with aqueous 10%sodium hydroxide solution. Consequently, MsN3 was preparedaccording to Danheiser’s 7a of Boyer’s procedure.’ 1 7bThe diazo group transfer reaction was then performed by treating adichloromethane solution of 197 and 198 with MsN3 in the presence* CAUTION: All sulfonyl azide compounds are potentially explosive and must be handledwith due care.107of triethylamine at 0°C (equation D-42).118 The product, a-diazoketone 199, was light-sensitive so the reaction mixture wasprotected from light and the subsequent manipulations wereexecuted in a dimly lit room. Partial removal of some impurities byan aqueous base/dichloromethane extraction was followed by a rapidchromatographic separation (silica gel) of 1 99 from more polarmaterial. Also, removal of the triethylamine (rotary evaporator)from the diazo ketone 199 was important since even traces oftriethylamine led to the formation of by-products in the photolysisreaction. An ir spectrum of 199 displayed the diazo group stretchat 2082 and the a-diazo carbonyl stretch at 1674 cm •6 1Absorptions due to the exocyclic methylene were at 1636 and 898cm-1.MsN3,Et3N,CH2I,0°C (4 :Q (D-42)(in the dark)Due to its instability, the a-diazo ketone 199 was used asquickly as possible after its preparation for the photochemical ringcontraction to prepare the diesters 200 and 201. Thus, 199 wasdissolved in deoxygenated distilled methanol in a quartz photolysistube and the tube was closed under an argon atmosphere. Generally,the photolysis reaction, using a medium pressure Hanovia mercury198 and theketo aldehydes 1 97199108lamp (450 Watt) with a Corex filter,”9 was complete in 30 minutesat 0°C (equation D-43). An -‘1.6:1 mixture of the diester epimers200 and 201 was obtained in a 40.5% overall yield from the ketoester 182.Usually, protonation of the ketene intermediate occurs from theless sterically hindered face of the ketene.’2° From an examinationof molecular models, the major epimer formed in the Wolffrearrangement of the diazo ketone 199 in the presence of methanolwas expected to be epimer 200.The mixture of epimers obtained was hard to separate and only asmall amount of the major epimer 200 was obtained uncontaminatedwith the minor epimer 201. However, the elemental analysis andhigh resolution mass spectral data for the mixture provided resultsconsistent with the molecular formula, C,6H2204. In the ‘H nmrspectrum, several signals readily distinguished the two epimers.MeOCi,.hv, MeOH,0°C (30mm)HN2199200+(D-43)12’Co2H H201109For example, a doublet of doublets resonated at 61.78 (J = 13.0, 8.0Hz) in the major epimer 200, and at 1.69 (J = 13.5, 7.0 Hz) in theminor epimer 201 . In addition, the signals for the epimericmethoxycarbonyl groups appeared as singlets (3H each) at 63.68(Me-13’) and 3.67 (Me-12’), respectively, while a triplet of doubletsresonated at 2.71 (J = 9.0, 3.0 Hz) for 200 and a doublet of doubletsof doublets was at 82.77 (J = 14.0, 5.0, 2.5 Hz) for the minor epimer.The major diester epimer (200) was characterized more fullyby one and two dimensional nmr experiments (13C and 1H) in order toassign reasonably all of the signals for the carbons and thehydrogens and to determine the relative stereochemistry at C-3.The broad band decoupled and APT 13C nmr spectra, as well as a 1H-13C HETCOR were obtained. Signals for only 15 of the 16 carbonatoms were displayed in the one dimensional 13C nmr spectra as thesignals for two methylenes resonated at the same frequency. Fromthe HETCOR 1H-3C correlations (see figure 7 and Table 4),assignments for the methines, the methoxycarbonyls and the geminalpairs of hydrogens were obtained. Thus, for example, the carbonsignals for the two methines at 6 36.2 (Q...H-3) and 46.8 (H-4)correlated to the hydrogen signals at 83.18 (dt, 1H) and 2.71 (br td,1H), respectively. Also, the carbon signal at 6 32.7 correlated tofour signals for hydrogens at 81.41-1.52, 1.89, 1.95-2.07 and 2.33,which indicated that two methylene carbons resonated at the sameposition. Other correlations between the carbon and hydrogensignals permitted the determination of the rest of the geminal pairsof hydrogens, but the assignments to specific positions weredependent on the COSY correlation results.110-25• D__0 0-50CDCI3_______ _--1000_________________________________________________________________________ppmI,,JIIuI.IIII I ,IIt, ill1,,,, 11111 IIppm 4 3 2 1Figure 7. The 125 MHz HETCOR spectrum of the diester 200.111Table 4: The 125 MHz HETCOR Data for the Diester 200.Position 13C (125 1H (500 MHz) Sppm (H-x)(C-x) MHz) Sppm1 52.4a - -2 26.6 2.26 (2a); 2.39-2.44 (2b)3 36.2 3.18 (3b)4 46.8 2.71 (4b)5 25.2 1.58-1.65 (5a); 1.95-2.07 (5b)6 37.2 2.43-2.50 (6a); 1.78 (6b)7 579 - -8 23.8 1.41-1.52 (8a or 8b); 1.58-1.65 (8b or 8a)9 32.7 1.89 (9a or 9b); 1.41-1.52 (9b or 9a)10 32.7 1.95-2.07 (lOa or lOb); 2.33 (10 b or lOa)11 147.9 - -13 and 14 174.0, 175.5 - -13’ and 14’ 51.37, 51.40 3.68 (13’); 3.62 (14’)15 107.7 4.93 (15a); 4.96 (15b)a. Signals may be interchanged.1H nmr decoupling and nOe experiments (500 and 400 MHz,respectively) were performed on a solution of the diester 200 inorder to determine the (a) identities of the hydrogen signals forhydrogens on the four-membered ring and (b) the relativeconfiguration at C-3 (see figure 8 for the normal 1H nmr spectrum).Thus, in decoupling experiments, irradiation of the signal at 83.18200112CHCI3PPMFigure 8. The 400 MHz H nmr spectrum of the diester 200.(H-3b) simplified the broad triplet of doublets at 2.71 (H-4b) to adistorted doublet of multiplets (J = 9 Hz), simplified the doublet ofdoublets at 2.26 (H-2a) to a doublet (J = 13 Hz) and simplified themultiplet at 2.39-2.44 (H-2b). Irradiation of the signal at 3 2.71 (H4b) simplified the multiplets at 1.95-2.07 (H-5b and one otherhydrogen) and at 2.39-2.44 (H-2b) and caused the doublet of triplets4.0 3:0 2.0 1.0113at 3.18 (H-3b) to collapse to a doublet of doublets (J = -‘9, -‘10 Hz).Irradiation of the signal at 6 2.26 (H-2a) simplified the multiplet at2.39-2.44 (H-2b) and caused the doublet of triplets at 3.18 (H-3b) tocollapse to a distorted triplet (J = 9 Hz). The above assignments ofthe signals for hydrogens on the four-membered ring were furtherconfirmed by nOe difference experiments (summarized in structure202’). Thus, irradiation of the signal at 3 2.71 (H-4b) led toenhancement of the signal at 3.18 (H-3b), while irradiation of thesignal at 3.18 (H-3b) led to enhancement of the signals at 2.39-2.44(H-2b), 2.71 (H-4b) and 4.93 (H-15a). Irradiation at 64.95 (betweenH-15a and H-15b) led to enhancements of the signals at 2.33 (H-lObor H-lOa), 2.39-2.44 (H-2b) and 3.18 (H-3b). From the nOe results,it was apparent that H-2b, H-3b and H-4b were on the same side ofthe four-membered ring and that H-3b and H-2b were spatially closeto the olefinic H-15a hydrogen. Therefore, the six-membered ringwas syn to the bridgehead methine (H-4b) as s found in the naturalproduct and the methoxycarbonyl group on the four-membered ringhad the configuration shown in 200.13’200’114The identification of other signals in the 1 H nmr spectrum wasfacilitated by combining the results summarized above with thosefrom the COSY and HETCOR spectra. For example, in the COSYspectrum (see figure 9), the signal for H-4b (32.71) showedcorrelations to the signals at 32.39-2.44 (H-2b), 3.18 (H-3b) and1 .95-2.07 (2H). In the foregoing discussion, the identities of thehydrogen signals at 32.39-2.44 and 3.18 had been established fromdecoupling and nOe experiments. Consequently, one of the hydrogensresonating at 3 1 .95-2.07 was H-5b. From the HETCOR results, thehydrogens gem inal to those at 6 1 .95-2.07 (m, H-5b and H-x)resonated at 1.58-1.65 (m, 2H) and 2.33 (1H). The signal at 62.33could be assigned to a hydrogen at the 10-position (H-lob or H-lOa)based on the nOe results. Thus, the signal for H-5a appeared as partof the multiplet at 6 1 .58-1 .65 and the other hydrogen at 1 .95-2.07was H-lOa or H-lOb. A similar combination of the various nmrresults led to the further assignment of the remaining carbon andhydrogen signals to the positions listed in Tables 4 and 5.115-d- . .-1—.—..i 1IuI ——______1. 4.5 4.0 3.5 3.0 2.5 2.0 1.5Figure 9. The 400 MHz COSY spectrum of the diester 200.116Table 5: The 400 MHz COSY Data for the Diester 200.Position Signal Sppm COSY Correlations(H-x) (multiplicity;a J; (H-x)number_of_H)2a 2.26 (dd; 13.5, 10.5; 1H) 2b; 3b2b 2.39-2.44 (m; 1H) 2a; 3b; 4b3b 3.18 (dt; 10.5, 9.5; 1H) 2a; 2b; 4b4b 2.71 (br td; 9.0, 3.5; 1H) 2b; 3b; 5b5a 1.58-1.65 (m; 2H) 5b; 6a5b 1.95-2.07 (m; 2H) 5a; 4b; 6a; 6b6a 2.43-2.50 (m; 1H) 6b; 5a; 5b6b 1.78 (dd; 13.0, 8.0) 6a; 5bBa or 8b 1.41-1.52 (m; 2H) b8b or Ba 1.58-1.65 (m; 2H) Ba or 8b; 9a and 9bC; lOb orlOa (W-coupling)9a or 9b 1.89 (dm; 10.5; 1H) 9b or 9aC; 8b or Ba; lOb or lOa9b or 9a 1.41-1.52 (m; 2H) blOa or lOb 1.95-2.07 (m; 2H) lOb or lOa; 9b or 9alOb or lOa 2.33 (dm; 13.5; 1H) lOa or lob; 8b or Ba (Wcoupling); 9a and 9b13’ 3.68 (s; 3H) - -14’ 3.62 (s; 3H) - -15a 4.93 (s; 1H) 15bl5b 4.96 (s, lH) 15aa. The signals labelled s, d, dd may incorporate unresolved fine couplings.b. Assignment of correlations is uncertain due to overlapped signals. The signal at 31.41-1.52 (m, 2H) showed correlations to 8b or 8a, 9a or 9b, and lOa and lOb.c. The signal for 9b or 9a is at the same position as the one for 8a or 8b. Thus, thecorrelations of 8b or 8a to 9b or 9a, and vice versa, are impossible to determine fromthe data.Ha2001172.3.2.5. Preparation of (±)-J3-Panasinsene (31) via the Diacetate213.The successful generation of the diesters 200 and 201 via theWolff rearrangement reaction set the stage for the performance ofthe final functional group manipulations to complete the synthesisof (±)--panasinsene (31). Thus, the introduction of a methyl groupat the 3-position of the diester mixture (200 and 201) followed bya double deoxygenation of the methoxycarbonyl functions wouldprovide the natural product.The methylated diesters 202 and 203 were prepared by treatinga cold (-78°C) THE solution of a mixture of the diesters 200 and201 (ratio -1 .6:1) first with a THF solution of lithiumdiisopropylamide to generate an anion at the 3-position, then withhexamethylphosphoramide* (HMPA, 1.6 equiv) and finally, withexcess methyl iodide. After an appropriate workup, an -‘18:1mixture (‘H nmr analysis) of the diester epimers 202 and 203 wasisolated in 65% yield (equation D-44). The two epimers could noteasily be separated at this stage so were characterized as themixture.The elemental analysis and the low and high resolution massspectroscopy performed on the diester mixture (202 and 203)provided results consistent with the molecular formula, C,7H2404.Details of the structure. of 202 were obtained from an ir spectrum,nmr spectra and one and two dimensional ‘H nmr experiments.* CAUTION: HMPA is known to be a potent carcinogen.1181) LDA, THE, -78°C(1 .5h)2) HMPA, -78°C3) Mel, -78°C, (O.5h);H CO2-78-,200 and 201 (50mm) MeO2CIg.14’Thus, in the ir spectrum of the mixture of 202 and 203, theabsorption for the ester carbonyl stretch was displayed at 1729cm-1, while the exocyclic olefinic function gave rise to absorptionsat 1642 and 887 cm. The broad band decoupled 13C nmr spectrumof the 18:1 mixture of 202 and 203 displayed the expected 17carbons for 202 and baseline signals (i.e., hardly distinguishablefrom the noise) for 203. From an APT experiment, the signal at 825.3 was assigned to the methyl group (.H3-12), while the signal at53.6 was assigned to the methine (.H-4). The methyl groups of themethoxycarbonyl functions resonated at 851.4 and 51.5 (H3-13’ andH3-14’). In the ‘H nmr spectrum (see figure 10), the signal for thenewly installed methyl group of the major epimer (202) appeared asa singlet at 31.42, while that of the minor epimer 203 was at 1118.The signals for the two methoxycarbonyl groups of 202 weresinglets at 63.61 (Me-14’) and at 3.69 (Me-13’), while the two1512Me,13’H1, CO2vb202 (D-44)+1512’Co2Me‘•‘ 13203119olefinic hydrogens appeared as singlets at 34.98 (H-15a) and 5.00(H-15b).PPMTMS5:0 4.5 3.5 3.0 2.5 1.5 0.5Figure 10. The 400 MHz ‘H nmr spectrum of the diester 202.120In order to determine the relative configuration at C-3, nOedifference experiments (summarized above in structure 202’) wereperformed on the mixture of 202 and 203. Thus, irradiation of thesignal at 31.42 (Me-12) led to enhancement of the signals at 2.26,-‘2.35, 3.69 (Me-13’) and 4.98 (H-15a). Due to unavoidableirradiation of part of the multiplet at 6 1 .46-1 .59, the signals at1.85 and 1.95-2.06 (H-lOa or H-lob) also showed enhancements.Irradiation of the signal at 34.98 (H-15a) led to enhancements of thesignals at 1.42 (Me-12) and at 2.26. From these results it wasapparent that the methyl group (Me-12) in the major epimer was inclose proximity to an olefinic hydrogen (H-15a) and, therefore, hadthe relative configuration depicted for 202. The configuration isthat expected if the methyl group had approached from the lesssterically hindered face of the anion obtained by deprotonation atthe 3-position of the diesters 200 and 201. Further evidence thatthe structure was correct was obtained once the identities of thehydrogens resonating at 62.26 and 2.35 were determined. In 202, H-202’2b and H-4b are cis to the methyl group (Me-12) and H-2b is in close121proximity to the olefinic hydrogen (H-15a). Consequently, the nOeresults were consistent with the assignment of the signal at 32.26(dd, 1H, J = 14.0, 3.0 Hz) to H-2b and the signal at -‘2.35 (part of a3H-multiplet at 2.28-2.39) to H-4b.HbMe020i1,14 14Confirmation of the assignments for H-2b and H-4b wasobtained from decoupling and/or COSY experiments. Thus, in adecoupling experiment, irradiation of the doublet at 32.45 (J = 14.0Hz, H-2a) led to the collapse of the doublet of doublets at 2.26 (H2b) to a distorted triplet (J = 3.0 Hz) indicating that the signals at2.26 and 2.45 were due to geminal hydrogens (J = 14.0 Hz). In a COSYexperiment (see figure 11), the signals for (a) the methyl group (Me-12) at 6 1.42 and (b) the olefinic hydrogens (H-i 5a and H-i 5b) at 34.98 and 5.00 provided key entries, via couplings unresolved in theone dimensional spectrum, into the spin systems of the interrelatedfour and five-membered rings and the six-membered ring,respectively. Thus, the signal at 6 1.42 (Me-12) showed acorrelation (W-coupling) to only the signal at 2.45 (H-2a), while thesignal at 2.45 (H-2a) showed a further correlation to the signal at2021221.52.0- 2.5-3.03.5- 4.0- 4.5-. 5.0PPti5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5PPMFigure 11. The 400 MHz COSY spectrum of the diester 202.2.26 (H-2b). In turn, the signal at 8 2.26 (H-2b) showed othercorrelations to the signals at 2.28-3.39 (H-4b and two otherhydrogens) and extremely small (long range) correlations to thesignal at 4.98 (H-15a). The correlations shown by the signal at S2.28-2.39 were too complex to further assign with certainty. Thus,the correlations due to the signals for the olefinic hydrogens (H-15aand H-15b) were examined. Apart from the correlation of the signalJ0123at 34.98 (H-15a) with the one at 2.26 (H-2b), both olefinichydrogens showed correlations to the multiplet at 1.95-2.06 (H-lOaor H-lOb and another H). Further tracing of the spin systems wasnot performed, but the COSY results, presented above andsummarized in Table 6, are consistent with the flOe and decouplingresults and confirm that the structure of the major epimer 202 isas assigned.Table 6: The 400 MHz COSY Data for the Diester 202.HbPosition Signal 3 ppm COSY Correlations(H-x) (multipli.city;a J; (H-x)number_of_H)2a 2.45 (d; 14.0; 1H) 2b; 12 (W-coupling)2b 2.26 (dd; 14.0, 3.0; 2a; 4b; 15a (long range coupling)1H)4b 2.28-2.39 (m; 3H) 2b; bc c dlOa or lOb 1.95-2.06 (m; 2H) 15a; 15b; b1 2 1.42 (s; 3H) 2a (W-coupling)1 3’ 3.69 (s; 3H) - -1 4’ 3.61 (s; 3H) - -15a 4.98 (s; 1H) 15b; 2b (long range coupling);lOa or lOb15b 5.00 (s; 1H) 15a; lOa or lOba. The signals labelled s, d, dd may incorporate unresolved fine couplings.b. Other correlations also were observed.202124Table 6: footnotes continued.c. The positions of H-x (x=5a, 5b, 6a, 6b, 8a, Bb, 9a, 9b and lOb or lOa) areuncertain.d. The correlations were not determined.A THE solution of the mixture of the diesters 202 and 203(ratio -‘18:1) was transformed into the corresponding mixture of thediols 204 and 205 by reduction with lithium aluminum hydride(equation D-45). After an appropriate workup and purification, themajor epimer 204 was obtained in 88% yield, while only traces ofthe minor epimer 205 were isolated.HRecrystallization of the major epimer 204 gave colorlessneedles with m.p. 139-139.5°C (sealed tube). Due to low solubilityin other solvents, the nmr spectra of 204 were obtained in acetoned6. In the 1H nmr spectrum (400 MHz) of the diol 204, the signal forthe methyl group (Me-12) appeared as a singlet (3H) at 6 1.11. Thesignals for the two hydroxyl hydrogens were triplets (1H, each) at 33.20 (J = 5.5 Hz) and at 3.32 (J = 5.0 Hz), both of which disappearedLiAIH4, THF,rt (1 .2h)202 and 203+HH(D-45)H12Hb205125upon the addition of D20. The signals for the Ca2OH hydrogens weredisplayed at 83.23-3.30 (m, 2H), 3.40 (dd, 1H, J = 10.5, 5.5 Hz) and3.46 (dd, 1H, J = 10.5, 5.0 Hz); upon the addition of D20, the multipletwas simplified and the two doublets of doublets gave rise todoublets (J = 10.5 Hz). The 13C nmr spectrum of 204 displayed theexpected 15 signals required for C,5H240. The elemental analysisand high resolution mass spectral data (found M=236.178O) werealso consistent with the molecular formula. In the low resolutionmass spectrum, the peak due to the molecular ion was very weak(0.4%) and peaks were found which corresponded to the loss of oneand two molecules of water (M-18 and M-36, respectively). The irspectrum of 204 indicated the presence of the hydroxyl groups by anabsorption at 3312 cm1.A small amount of the minor epimer 205 was isolated forcharacterization purposes. Thus, in the ‘H nmr spectrum of 205, thesignal for the methyl group (Me-13) appeared at 60.93 (s, 3H). Thesignal for one hydroxyl group was displayed as a triplet at 63.20 (J =5.5 Hz) and disappeared upon the addition of D20, while the otherhydroxyl hydrogen and both of the CU2OH signals were part of amultiplet at 8 3.30-3.49 (5H). The addition of D20 simplified themultiplet at 8 3.30-3.49 to give four distorted doublets at 3.26 (br d,J = 11.0 Hz), 3.31 (J = 10.5 Hz), 3.35 (J = 11.0 Hz) and at 3.43 (J =10.5 Hz). The exact mass of the molecular ion in the high resolutionmass spectrum (M= 236.1777) was consistent with the expectedmolecular formula of 205.Several factors were considered when choosing. which method touse for the double deoxygenation of the diol 204. In the first place,126both of the primary hydroxyl groups in 204 are neopentyl in natureand, therefore, the carbinol carbon atoms are very hindered. Thus,SN2-type reactions, which are susceptible to steric effects,121 werenot deemed feasible for use in the performance of thedeoxygenation.’22 On the other hand, radical reactions122’3 (unlikeSN2 processes) are not as susceptible to steric effects. In addition,the radical reactions occur under neutral conditions,122 so acid orbase sensitive groups are compatible with the conditions utilizedfor the reactions. One problem in applying radical reactions to thedeoxygenation of primary alcohols reflects the decreased stabilityof primary radicals in comparison with secondary radicals. Thus,the reaction conditions used are generally more drastic for thedeoxygenation of primary alcohols than for the same reaction ofsecondary alcohols.’22’124 For example, deoxygenation of thesecondary monothiocarbonylimidazolide 206 using tri-n-butyltinhydride to give 5a-cholestane (207) took 1.5 hours in refluxingtoluene.125 In contrast, deoxygenation of the primarymonothiocarbonylimidazolide 208 with the same reagent to give f3-amyrin 209 took 10 hours at 130°C in xylene’24 (equations D-46 andD-47, respectively).The possible application of a radical-based deoxygenation to thedouble deoxygenation of a derivative of the diol 204 led to theconsideration of how to separate the hydrocarbon product, J3-panasinsene, from the solvent and by-products of the reaction.Ideally, solvents, reagents and by-products with low boiling points(<60°C), or that were water soluble, or formed filterable solids were127deemed to be most suitable. Due to known problems in separatingthe product from reagents and by-products, triorganotin hydride126reductions or the more recent variations using silanes(triethylsilane,’27 tris(trimethylsilyl)silane,’ 28 or diphenylsilane129) were not attempted.Reductions using dissolving metals, such as those employed inIreland’s deoxygenation of phosphodiamidates,’29were expected tobe feasible, but problematic due to the possible over-reduction ofthe olefinic functional group to give the saturated deoxygenatedproduct. Thus, for example, Wai and Piers found that the stericallyhindered primary alcohol 210 could be deoxygenated via itsjjl) n-Bu3SnH, PhMe (O.5h)2) reflux (1.5h)206H207H1) n-Bu3SnH, xylene (2h)2) 130°C (lOh)H208(D-47)209128phosphorodiamidate derivative to give 211 ,12,131 but the reactionwas somewhat capricious and varying amounts of the over-reducedsaturated product 212 also were formed if conditions were notcarefully controlled (equation D-48).32I I(Me2(D-48)Pete133 and coworkers reported that the photolysis of acetatesin HMPA/water works well for acetate derivatives of primary andsecondary alcohols, while Collins and Munasinghe34 found that thereaction could also be successfully applied to the diacetates ortripivaloates derived from diols and triols, respectively. Thedeoxygenated product may be recovered by adding water to thereaction solvent and performing extractions with an organic solvent.In order to attempt Pete’s procedure, the diacetate 213 wasprepared. Thus, a cold (0°C) dichloromethane solution of the diol204, containing DMAP (-1.1 equiv) and pyridine (9 equiv), wastreated with acetyl chloride (6 equiv) (equation D-49). After anappropriate workup and purification, the diacetate 213 wasobtained in 85% yield.Li, MeNH2210 211 212129HO AcDMAP, Pyr,CH3OCI, (D-49)H CH2I,0°C (3h)In the ir spectrum of the diacetate 21 3, the presence of acarbonyl absorption at 1742 cm1 indicated that the replacement ofthe hydroxyl functions by acetate groups had occurred. The expectedabsorptions for the exocyclic methylene were displayed at 1636 and889 cm1. In the ‘H nmr spectrum (see figure 12), the signals forthe three methyl groups appeared at 6 1.15 (Me-12), 2.02 and 2.05(Me-13” and Me-14”), while the signals for the methylene hydrogensof the acetoxymethyl functions were found at 3.79 (dd, 1H, J = 11.0,1.0 Hz, H-14a), 3.86 and 3.89 (AB pair of d, 2H, J = 11.0 Hz, 2H-13)and 4.03 (d, 1H, J = 11.0 Hz, H-14b). The origin of the unexpectedsmall coupling in the signal for H-14a (J = 1.0 Hz) was explainedupon examination of a COSY spectrum of 213 (see figure 13) whichshowed a correlation (W-coupling) between the signals at 8 3.79 (H14a) and at 1.29-1.39 (m, 1H, H-8b or H-6b). From the shape of the14H3CC(O)O.14”13 13”OC(O)CHMe204 213Ii,213130TMS5.0 4.0 3.0 2.0 1.0 PPMFigure 12. The 400 MHz ‘H nmr spectrum of the diacetate 213.signal at ö 1 .29-1 .39 (m) in the normal 1 H nmr spectrum, it wasmore likely that the signal was due to H-8b rather than to H-6b.Other assignments of hydrogen signals were possible based on theresults of the COSY experiment. The resonance due to the methylgroup (Me-12) provided an entry into the spin system of the fourmembered ring, while the signals due to an olefinic hydrogen (H-15b)and H-14a provided the needed access into the spin system of thesix-membered ring. Thus, the signal at 3 1.15 (Me-12) showed a131correlation (W-coupling) to the signal at 1.72 (d, J = 13.0 Hz, H-2a).In turn, the signal at 6 1 .72 (H-2a) showed a correlation to the signalat 1.98 (dd, J = 13.0, 3.0 Hz, part of a 9H multiplet at 1.94-2.07).Due to the complexity of the correlation pattern of the signal at 61.94-2.07, the correlations due to the signals for ôlefinic hydrogensH-15a and H-15b at 4.84 and 4.96, respectively, were then examined.However, apart from correlations to each other, both showedcorrelations to the multiplet at 3 1 .94-2.07 (9H) indicating that thesignals for one or both of the allylic hydrogens (H-ba and/or H-lob)were part of the multiplet. Finally, the signal at 33.79 (H-14a) gaverise to a correlation to the signal at 1.29-1.39 (m, lH, H-8b). Thesignal at 6 1 .29-1 .39 (H-8b) also showed correlations to the signalsat 1.43 (br qt, lH, J = 13.0, 3.0 Hz) and 1.62-1.72 (m, 3H). Due to thepresence in the 1H nmr spectrum of several complex multiplets,other assignments of signals were not feasible.To confirm the assignments of the various hydrogen signalsgiven above (and in Table 7), a NOESY97 (two dimensional nOe)spectrum of 213 was obtained. The NOESY correlations aresummarized in Table 8. For example, the signal at 6 1.72 (H-2a)showed nOe enhancements to signals at 1.98 (H-2b), 3.79 (H-14a)and 3.86 and 3.89 (2H-13), while the signal at 1.98 (H-2b) showedenhancements to the signals at 1.15 (Me-12), 1.72 (H-2a) and 4.84(H-15a). The signal for the acetoxymethyl group at 63.86 and 3.89(AB pair of d, 2H-13) showed enhancements to the signals at 1.72(H-2a) and 1.15 (Me-l2). The signal for one of the hydrogens of theother acetoxymethyl group H-14a (at 63.79) displayed enhancementsto the signals at 4.03 (H-14b) and at 1.72 (H-2a). It thus appeared,132that the depicted structure for the diacetate 213 was correct andthat the assignments of the hydrogen signaJs were reasonable. 13. The 400 MHz COSY spectrum of the diacetate 213.133Table 7: The 400 MHz COSY Data for the Diacetate 213.Positiona Signal Sppm (multiplicity;b COSY Correlations (H-x)(H-x) J; number of H)2a 1.72 (d; 13.0; 1H) 2b;C 12 (W-coupling)2bC 1.98 (dd; 13.0, 3.0; 1H) d8b 1.29-1.39 (m; 1H) 14a (W-coupling)elOa or lObC 1.94-2.07 (m; 9H) l5ad12 1.15 (s; 3H) 2a (W-coupling); 13 (WCo u p11 n g)1 3 3.86 and 3.89 (AB pair of d; 12 (W-coupling); f11.0; 2H)13” and 14” 2.02 (s, 3H); 2.05 (s; 3H) - -14a 3.79 (dd; 11.0, 1.0; 1H) 14b; 8b (W-coupling)14b 4.03 (d; 11.0; 1H) 14a15a 4.84 (s; 1H) 15b15b 4.96 (s; 1H) 15a; lOa or lObCa The assignments of the hydrogens H-x (x= 4b, 5a, 5b, 6a, 6b, 8a, 9a, 9b, and lOb orba) were not feasible.b The signals labelled s, d or dd may also incorporate unresolved fine couplings.c The hydrogen is part of the signal at 3 1.94-2.07 (m, 9H; 2b, ba or lOb, Me-13”,Me-14” and 1 other H).d The signal at 8 1.94-2.07 showed correlations to: 1.72 (d, 2a), 2.27 (br dd, 1H),1.62-1.72 (m, 3H), 1.54 (dd, 1H), 1.43 (qt, 1H), 2.21-2.27 (m, 1H), 1.15 (s,Me-12), 3.86 and 3.89 (AB pair of d, 2H-13), 4.84 (s, 15a) and 4.96 (s, 15b).e The signal also showed correlations to 1.43 (br qt, 1H) and 1.67-1.72 (m, 3H).f There is also a correlation to 3 1.94-2.07 (m, 9H).Hb213134Table 8: The 400 MHz NOESY Data for the Diacetate 213.Position Signal Sppm (multiplicity; J; number NOESY Correlations(H-x) of H) (H-x)2a 1.72 (d; 13.0; 1H) 2b; 13; 14a2b 1.98 (dd; 13.0, 3.5; 1H) 2a; 12; 15a12 1.15 (s; 3H) 2b; 13; 15a13 3.86 and 3.89 (AB pair of d; 11.0; 2H) 2a; 1214a 3.79 (dd; 11.0, 1.0; 1H) 14b; 2a14b 4.03 (d; 11.0; 1H) 14a15a 4.84 (s; 1H) 15b; 2b; 1215b 4.96 (s; 1H) 15aThe preparation of the diacetate 213 set the stage for thecrucial double deoxygenation reaction to prepare the natural product,(±)-/3-panasinsene (31) via Pete’s procedure.133 Thus, solutions ofthe diacetate 213 in HMPA-H20 (-p95:5) were photolyzed using eithera single low pressure mercury lamp (emission at 253.7 nm) or aRayonet reactor (16 such lamps) until no more starting material wasdetected (‘7-9 hours). Dilution of the reaction mixture with water,followed by an extraction with pentane and removal of the solventled to the isolation of low yields of (±)-f3-panasinsene (31), whichwas usually contaminated by a small amount of a compound withvery similar properties. The impurity could be removed byHb2Hb213135chromatography of the mixture on silica gel impregnated with silvernitrate (1.25 g AgNO3/ 5.00 g of 70-230 mesh silica gel),’31 but theyields of the target compound were too low (<10%) to make theprocedure feasible. After several unsuccessful attempts to improvethe results, a different approach to the double deoxygenation of thediol 204 was undertaken. Preparation of (±)-/3-Panasinsene (31) via a Wolff-KishnerReduction.There are various methods known for the reduction of aldehydeor keto groups to the corresponding methylene groups, but the twomost common methods are the Clemmensen and Wolff-Kishnerreductions.’35 The Clemmensen reduction uses zinc amalgam andaqueous, or in some cases gaseous, HCI. f3-Panasinsene (31) isknown to rearrange in acid to give neoclovene 56 (vide supra,Introduction),’3 so the Clemmensen reduction would not be useful inthis case. The Wolff-Kishner reduction, in contrast, involvesconversion of an aldehyde or ketone of general structure 214 to thehydrazone 21 5 and heating the hydrazone with base to give thedeoxygenated product 216 (equation D-50). In a one pot procedure,the aldehyde or ketone may be heated with hydrazine hydrate and abase to produce the deoxygenated product.1360RRH2NNH0,heat214Nbase,R R’ heat215Recently, Roberge136 found that the use of anhydrous hydrazine*as a cosolvent, rather than just as a reagent,137 to form thehydrazone improved the yields of the reaction. Thus, heating asolution of the ketone in an -1 :2 mixture of anhydrous hydrazinediethylene glycol at 130-140°C generated the hydrazone. Removalof the excess hydrazine by a reduced pressure distillation, followedby reaction of the hydrazone with potassium hydroxide at 200-210°C, led to the formation of the deoxygenated product. Yieldswere reduced and reaction times were increased if the excesshydrazine was not removed before the base was added. The productcould be isolated by the addition of water to the cooled reactionmixture followed by extractions with an organic solvent. Theprocedure seemed suitable for the deoxygenation of the dialdehyde217, presumed to be available from the diol 204.* Anhydrous hydrazine is explosive in the presence of oxidizing agents (including air)and must be handled with great care.NH2(D-50)216H I0217137The dialdehyde 217 was prepared from the dial 204 via a Swernoxidation.138 Thus, the dial 204, in a mixture of dichioromethaneand dimethyl sulfoxide (DMSO), was allowed to react with a mixtureof DMSO and oxalyl chloride. Treatment of the mixture withtriethylamine, followed by an appropriate workup and a rapidfiltration of a solution of the product through a silica gel column,led to the isolation of the dialdehyde 217 in 98% yield (equation D51). The product thus obtained was pure enough to be characterized.The ir spectrum of the dialdehyde 217 showed the expectedabsorptions for the aldehyde functions at 2723 (w) and at 1718 (vs)cm’, while the absorptions for the exocyclic methylene weredisplayed at 1638 and 894 cm* In the 1H nmr spectrum, the signalfor the methyl group (Me-12) appeared at 5 1.36 (s, 3H), while thesignals for the aldehyde hydrogens were at 9.50 (s, 1H) and 9.65 (s,1H). In the low resolution mass spectrum, the molecular ion (1.9%)and fragments corresponding to the loss of one and two formylgroups (M-29 and M-58, respectively) were observed. The exactmass of 232.1459, found for the molecular ion of 217 in the highresolution mass spectrum, was consistent with the formula,C15H200.1) DMSO, (COd)2CH2I, -78°C (40mm)H —OH2) Et3N, -78—, —0°C(70m in)2043H CHO(D-51)217138The dialdehyde 217 was then submitted to the Wolff-Kishnerdeoxygenation procedure using the modifications developed byRoberge.’36 Thus, a solution of the dialdehyde 217 in diethyleneglycol and anhydrous hydrazine (5:3 diethylene glycol-hydrazine)was heated at ‘-135°C for 1.5 hours under an argon atmosphere.After removal of excess hydrazine and addition of base (potassiumhydroxide), the mixture was heated at ‘-200°C for 7.5 hours. Cooling,followed by an aqueous workup, led to the isolation of crude (±)-J3-panasinsene (31) in ‘-76% yield (equation D-52). A reduced pressuredistillation of this material resulted in a 46% yield of the syntheticnatural product 31 in >97% purity (gic analysis).1) H2NNH,DEG,—135°C (1.5h)2) KOH, DEG,-200°C (7h)The purified synthetic (±)-j3-panasinsene (31) displayedspectral characteristics similar to those reported in the literature.Unfortunately, all our attempts to obtain samples of natural13’6 orsynthetic30’1 f3-panasinsene or the appropriate spectra forcomparison purposes were unavailing. There is no doubt, however,that (±)-13-panasinsene (31) was the compound synthesized via thereaction sequence described in this thesis. This contention is borneout by a comparison of the data obtained for the newly synthesized(±)-J3-panasinsene (31) with the data reported in the literature1314MeMeH CHO217(D-52)1331139(see Table 9), and by an examination of the other spectral data thatwere obtained for the synthetic natural product.Table 9. A Comparison of the Spectral Data for Authentic andType of Data Authentic (-)-J3- Synthetic (±)-J3-Panasinsenea PanasinsenebInfrared 3080, 1620, 1365, 3088, 1636, 1377,(cm1) 1360, 1260, 1080, 1365, 1269, and 885930, and 8851H nmrc 0.74 (s; 3H) 0.75 (s; 3H; Me-14)o ppm (multiplicity; J; 0.86 (s; 3H) 0.86 (s, 3H; Me-13)number of H; Hx) 1.08 (s; 3H) 1.07 (s, 3H; Me-12)4.78 (d; J = 2 Hz; 1 H) 4.80 (d; J = 1.5 Hz;1H; H-15a)4.84 (d; J = 2 Hz; 1 H) 4.91 (dd; J = 1.5, 1.5Hz; 1H; H-15b)Low Resolution Mass 204 (Mj 204 (Mt, 29%)Spectrum 189, 175, 161 (base 189 (23), 175 (14),Peak (% relative peak), 146, 133, 122, 162 (16), 161 (100),intensity) 109, and 107 147 (19), 133 (44),122 (46), 1 19 (23),107 (47), 105 (40),91 (38)a. From reference 13. No otner data were reported.b. From this thesis.c. The spectrum of authentic 31 was recorded on a JEOL-JNM-C-60 60 MHzspectrometer (solvent not identified).13 The spectrum of synthetic 31 was recorded ona Bruker WH 400 MHz spectrometer in CDCI3.The data for the authentic and synthetic J3-panasinsene (31) ascompiled in Table 9 are similar, but a few differences may be noted.Firstly, according to the ir spectral data, the synthetic sample didnot display significant absorptions at 1080 or 930 cm’ as werereported for the natural product and the absorption due to theexocyclic methylene (C=C stretch) occurred at different positions(1620 and 1636 cm). The reasons for the difference are uncertain.Synthetic j3-Panasinsene (31).140Secondly, from the ‘H nmr spectral data, the signal for one of theolefinic methylene hydrogens differs significantly with regard toboth position and multiplicity (at 3 4.84 (d) versus at 4.91 (dd)). Thediscrepancies in the nmr results may be mainly due to the differentinstruments used (60 MHz versus 400 MHz). Also, the solvent used inthe literature was not identified. Finally, the low resolution massspectral patterns differ particularly with regard to the peak at 146(literature) or at 147 (synthetic). The disagreement may be due to atypographical error in the literature. Other mass spectraldifferences probably are unimportant as no contradiction is involved.The discrepancies between the literature data and the data from thesynthetic J3-panasinsene as summarized in Table 9 are minor incomparison with the similarities and thus, it may be concluded that31 was synthesized.Further confirmation of the identity of the synthetic (±)-f3-panasinsene (31) was procured in the form of a high resolution massspectrum, ‘3C nmr spectra (broad band and APT), as well as severalone and two dimensional nmr spectra (flOe, HETCOR and COSY). Amolecular formula of C,5H24, as expected for 13-panasinsene, wasindicated by the presence of a molecular ion with an exact mass of204.1875 in the high resolution mass spectrum of 31. The numberof carbons was also implied by the presence of 15 carbon signals inthe broad band decoupled ‘3C nmr spectrum (see figure 14). Based onthe chemical shifts of the signals, those at 8108.3 and 152.3 couldbe assigned to the olefinic methylene (H2-15) and to the quaternaryolefinic carbon (-11), respectively. According to a 13C nmr APTexperiment, four signals (at 8 52.60, 18.2, 24.85 and 30.64) were due0(.4a0II.r, flta,2• rI JJbfl$i a141Hb HaHb31to either methine or methyl carbons. The signal at 6 52.60 wasclearly due to the methine, H-4, while the other three signalscorresponded to the signals of the three methyl groups in 31.CDCI3’0(.4 C160 140 120 100 80 60 40 20 0 PPMFigure 14. The 75 MHz broad band decoupled 13C nmr spectrum ofsynthetic (±)-$-panasinsene (31).142The ‘H nmr spectrum (see figure 15) of our synthetic 31, inaddition to the signals mentioned in Table 9, showed signals foranother 13 hydrogens, of which only two signals overlapped in anindistinguishable manner (81.35-1.41, m, 2H). In order to identifythe methine (H-4b) and the pairs of geminal hydrogens in the ‘H nmrspectrum, a HETCOR experiment was performed and yielded thePPM3.0AJJJ2.0Figure 15. The 400 MHz ‘H nmr spectrum of synthetic (±)-J3-7.5 7.0 5.0 4.0panasinsene (31).143results summarized in Table 10 (see also figure 16). For example,the methine carbon signal (H-4) at 6 52.60 correlated with thehydrogen signal at 2.10 (br dd, 1H, J = 8.5, 3.0 Hz, H-4b), while acarbon signal at 35.7 correlated with two hydrogen signals at 1.46(d, 1H, J = 12.5 Hz, H-2a) and at 1.96 (dd, 1H, J = 12.5, 3.0 Hz, H-2b),thus indicating that the pair of hydrogens was geminal.H20Figure 16. The 125 MHz HETCOR spectrum of synthetic (±)-j3-panasinsene (31). (* Folded peaks. ** Ti noise).50MO150ppm144Table 10: The 125 MHz HETCOR Data for Synthetic (±)-J3-Panasinsene(31).Position ‘3C (125 MHz) 1H (500 MHz) ö (H-x)(C-x)1 52.77 - -2 35.7 1.46 (2a); 1.96 (2b)3 45.6a - -4 52.60 2.10 (4b)5 24.72 1.63-1.68 (5a); 1.87-1.97 (5b)6 41.1 1.73 (6a); 1.35-1.41 (6b)7 30.53a - -8 36.0 1.27 (8a); 1.42-1.48 (8b)9 25.10 1.35-1.41 (9a or 9b); 1.57-1.63 (9b or 9a)10 33.8 2.00 (lOa or lOb); 2.18 (lOb or lOa)11 152.3 - -12 30.64 1.07 (3H-12)1 3 24.85 0.86 (3H-13)14 18.2 0.75 (3H-14)15 108.3 4.80 (15a); 4.91 (15b)a Signals may be interchanged.The precise assignments of the positions of the geminal pairs ofhydrogens and the methyl groups were made based on the COSYcorrelations observed for the signals (see figure 17 and Table 11).Entry into the spin systems was afforded by the signal for themethine at 3 2.10 (H-4b) and the signal for olefinic hydrogen at 4.9131145(H-15b). Thus, the signal at 8 2.10 (br dd, 1H, J = 8.5, 3.0 Hz, H-4b)showed correlations to the signals at 1.96 (dd, 1H, J = 12.5, 3.0 Hz,H-2b, W-coupling) and at 1.87-1.97 (m, 1H, H-5b). The signal at 81.46 (1H) for the hydrogen geminal to the one at 1.96 (1H) showed acorrelation to the methyl group at 1.07 (Me-12), while neither thesignal at 1.87-1.97 (H-5b) nor the one due to the geminal hydrogen(at 1.63-1.68) did so. Thus, the former pair of signals (i.e., at 8 1.46and 1.96) was assigned to the 2-position and the latter pair (i.e., at 81.63-1.68 and 1.87-1.97) was assigned to the 5-position. Thesignals arising from the hydrogens at the 6-position were thenidentified by determining with which other signals the signals dueto H-5a and H-5b showed correlations. Based on the nOe resultsdescribed below and from consideration of molecular models, the a-and b-hydrogens on carbons 5 and 6 were assigned as listed in Table11. Similarly, the spin system of the six-membered ring wasentered by way of an olefinic hydrogen at 34.91 (dd, 1H, J = 1.5, 1.5Hz, H-15b). The methyl groups were assigned based on the fact thatthe signals at 8 1.07 and 0.86 (Me-12 and Me-13, respectively)showed W-coupling to each other, while the signals at 1 .07 and 0.75(Me-12 and Me-14, respectively) showed W-coupling to the signalsfor two ring hydrogens at 1.42-1.48 (a multiplet overlapping thedoublet at 1.46). The W-coupling between the signals at 8 1.07 (Me12) and at 0.86 (Me-13) meant that the two signals were due to thegeminal methyl groups and thus, that the signals at 6 1 .07 and 0.75correlated, respectively, to the signals at 1 .46 (d, H-2a) and at 1 .42-1.48 (m, H-8b). 17. The 400 MHz COSY spectrum of synthetic (±)-J3-1460.51.0. o.I • ‘a0o‘‘oT’5.5 5.0 4.5 4.0 3.5 3.0PPM2.5 2.0 1.5 1.0 0.55.5PPMpanasinsene (31).147Table 11: The 400 MHz COSY Data for Synthetic (±)-13-Panasinsene(31).2Position Signal o ppm COSY Correlations(H-x) (multiplicity;a J; (H-x)number_of_H)2a 1.46 (d; 12.5; 1H) 2b; 12 (W-coupling)2b 1.96 (dd; 12.5, 3.0; 1H) 2a; 4b4b 2.10 (br dd; 8.5, 3.0; 1H) 2b; 5b5a 1.63-1.68 (m; 1H) 5b; 6a5b 1.87-1.97 (m; 1H) 5a; 4b; 6b; 6ab6a 1.73 (dd; 12.0, 7.0; 1H) 6b; 5a; 5bb6b 1.35-1 .41 (m; 2H) 6a; 5b8a 1.27 (dm; 12.5; 1H) 8b; 9a and 9b; lOa or lOb (Wcoupling)8b 1.42-1.48 (m; 1H) 8a; 9b or 9a; 14 (W-coupling)9a or 9b 1.35-1.41 (m; 2H) 9b or 9a; Ba; lOa and lOb9b or 9a 1.57-1.63 (m; 1H) 9a or 9b; Ba; Bb; lOa and lOblOa or 2.00 (br td; 12.5, 4.0; lOb or lOa; 9b and 9a; 15blOb 1H)lOb or 2.18 (dm; 12.5; 1H) lOa or lOb; 9a and 9b; 8a (WlOa coupling)12 1.07 (s; 3H) 2a (W-coupling); 13 (Wcoupling)13 0.86 (s; 3H) 12 (W-coupling)1 4 0.75 (s; 3H) 8b (W-coupling)15a 4.80 (d; 1.5; 1H) 15b15b 4.91 (dd; 1.5, 1.5; 1H) 15a; lOa or lOba Signals labelled s, d, or dd may also incorporate unresolved fine couplings.b Small correlations observed.31148In nOe difference experiments (summarized in structure 31’),irradiation of the singlet at 3 1.07 (Me-12) led to enhancement ofthe signals at 1.96 (H-2b), 2.10 (H-4b) and 4.80 (H-15a), whileirradiation of the singlet at 30.86 (Me-13) led to enhancements ofthe signals at 1.46 (d, H-2a) and 1.63-1.68 (m, H-5a). Irradiation ofthe singlet at 30.75 (Me-14) led to enhancement of the doublet at1.46 (H-2a). Thus, the assignments for the methyl groups wereconsistent with the other nmr data.The results outlined above for the synthetic (±)-J3-panasinsene(31) provide further information about the spectroscopic propertiesof the natural product and confirm the identity of the syntheticmaterial.149IlL CONCLUSIONThe work summarized above and outlined in Scheme D-12,constitutes a successful total synthesis of (±)-13-panasinsene (31)in fourteen steps from the keto ketal 46. While the approach to thesynthesis of (±)qi-panasinsene (31) via the Pauson-Khandcyclization reaction was unsuccessful, the Weiss-Cook condensationprovided a viable alternative for the synthesis of an enone (159)with a tetrasubstituted double bond. The key methylenecyclohexaneannulation sequence previously developed in our laboratories wassuccessfully applied to the enone 159 and efficiently provided thetricyclic keto ester ketal 171 which has the desired relativestereochemistry at the three chiral centers. Functional groupmanipulations followed by a Wolff rearrangement providedintermediates 200 and 201 with the required tricyclic carbonskeleton. Further reactions, including the double deoxygenation ofthe sterically hindered aldehyde functions in the dialdehyde 217,generated synthetic (±)--panasinsene 31 . The previouslyunreported 13C, HETCOR, COSY and nOe data were obtained for 31 asa part of this research and provide a more complete picture of thespectroscopic characteristics of the natural product.150kmMenScheme D-12a. KH, THF, 600C; dimethyl carbonate, —60°C (—93%), b. KH, THF, rt; PhSeCI, 0°C(86%), C. H2O, CH2I, 0°C; rt (—quant.), d. 5-chloro-1-pentenyl-2-magnesiumbromide, CuBrSMe THE, -78°C (—94%), e. Cs2O3,CH3N, —60°C (64%), f. NaBH4,CeCl36H2O, MeOH, -48°C (98%), g. PTC-CI, DMAP, CH3N, —70°C (76%), h. n-Bu3SnH,AIBN, PhH, —77°C, (74%), i. 1 N HCI (aq), acetone, rt (85%), j. i) t-amylONa, PhH,—7°C; ii; MeO2CH, 5°C to rt (—quant.); ii) MsN3,Et3N, CH2I,0°C (in the dark); iii) hv,MeOH, 0°C (40.5%), k. LDA, THE, -78°C; HMPA, -78°C; Mel, -78°C to —5°C (65%), I.LiAI H4, THF, rt (88%), m. DMSO, oxalyl chloride, CH2I -78°C; Et3N, -78°C to 0°C(98%), n.2NNH DEG,—135°C; KOH, —200°C (46%).171 169 and 170 159177 and 178 179t02Me202HO‘a200 and 201 182OHCa204Me217 31151IV. EXPERIMENTAL4.1. General.Proton nuclear magnetic resonance (1H nmr) spectra wererecorded on either a Varian XL-300 or a Bruker WH-400 nmrspectrometer using deuterochioroform as the solvent and tetramethylsilane (TMS) or the proton of the residual chloroform (6 7.26)as the internal standard, unless otherwise noted. Signal positionsare given in parts per million (8) from TMS. Coupling constants (J)are given in Hertz (Hz). The multiplicity, number of protons,coupling constant(s), and assignments (when known) are given inparentheses. Abbreviations used are: s, singlet; d, doublet; t, triplet;q, quartet; m, multiplet; br, broad.Carbon nuclear magnetic resonance (13C nmr) spectra wererecorded on a Varian XL-300 nmr spectrometer at 75.3 MHz, or on aBruker AM-400 spectrometer at 100 MHz, or on a Bruker AMX-500spectrometer at 125.8 MHz using deuterochloroform as the solvent,unless otherwise noted. Signal positions are given in parts permillion (6) relative to the chloroform signal at 6 77.0.61 Signalmultiplicities were determined by the Attached Proton Test (APT)experiment.Two dimensional spectra were recorded on the Bruker WH-400nmr spectrometer (COSY, NOESY), or a Varian XL-300 spectrometer(HETCOR) using a dual probe or a Bruker AMX-500 nmr spectrometer(HETCOR) employing an inverse detection probe. References were as152indicated above or, in some cases, residual chloroform at 3 7.24 wasused for the reference and a correction factor (x + 0.02 ppm) wasapplied to the data.Infrared (ir) spectra were recorded on a Perkin-Elmer 1710Fourier Transform Spectrophotometer with internal calibration.Abbreviations used are: s, strong; v, very; w, weak.Low resolution mass spectra (LRMS) were recorded on a KratosMS8ORFA spectrometer. High resolution mass spectra (HRMS) wererecorded on a Kratos/AEI MS 50 spectrometer.Elemental analyses were performed on a CARLO ERBA CHNelemental analyzer, Model 1106, or a Schoniger’s Oxygen Flask(analysis of sulfur).Melting points (uncorrected) were measured on a Fisher-Johnsmelting point apparatus, unless otherwise noted. Distillationtemperatures (uncorrected) are indicated as air-bath temperaturesof Kugelrohr distillations, unless otherwise noted.Gas-liquid chromatography (glc) was performed on either aHewlett-Packard model 5880A or 5890 capillary gas chromatograph,each having a flame ionization detector and a fused silica column,either 20 m x 0.21 mm coated with cross-linked SE-54 (formerinstrument) or -‘25 m x 0.20 mm coated with 5% phenyl-methylsilicone (latter instrument).Thin layer chromatography (tic) was performed on commercially• available aluminum backed silica gel plates (E. Merck, type 5554).Visualization was accomplished using ultraviolet light, a 5%solution of ammonium molybdate in 10% aqueous sulfuric acid (wlv),or a solution of phosphomolybdic acid in ethanol (20%, w/v).153Conventional column and flash139 chromatography were done on 230-400 mesh silica gel (E. Merck, Silica Gel 60).Unless otherwise stated, all reactions were performed under anatmosphere of dry argon using dry solvents in flame dried glassware.Liquid reagents or solutions of compounds were added via syringe,unless otherwise noted.Cold temperatures were maintained by use of the followingbaths: 5-10°C, water/(ice); 0°C, ice/water; -20°C and -48°C, aqueouscalcium chloride/CO2 (27.0 g calcium chloride/ 100 mL water; 46 gcalcium chloride! 100 mL water, respectively); -78°C, acetone/CO2.Temperatures were measured in degrees Celsius.Solvents and ReayentsSolvents and reagents were dried and purified using standardprocedures.14°Tetrahydrofuran (THF) and diethyl ether were distilled fromsodium benzophenone ketyl. Benzene, dichloromethane and dimethylsulfoxide were distilled from calcium hydride. Petroleum etherrefers to a hydrocarbon mixture with b.p. 30-45°C (from thedistillation of a commercially available mixture with b.p. 30-60°C).Diisopropylamine, triethylamine, hexamethylphosphoramide,pyridine and acetonitrile were distilled from calcium hydride.Anhydrous hydrazine (explosive in the presence of oxidizing agents)was prepared by refluxing hydrazine hydrate over an equal weight ofsodium hydroxide pellets for 2 h and distilling under a flow of154argon.141 Methanesulfonyl azide (MsN3, potentially explosive) wasprepared from distilled methanesulfonyl chloride (distilled fromPCI5) and sodium azide, according to the Danheiser et aI.hl7modification of the procedure of Boyer et aI.,hl7b and was usedwithout distillation.A 0.78 M benzene solution of sodium t-amyloxide was preparedaccording to the procedure of Conia’42 and was standardized usingaqueous hydrochloric acid in ethanol’ (phenolphthalein indicator).Solutions of methyllithium in diethyl ether and n-butyllithium inhexanes were obtained from Aldrich Chemical Co., Inc. and werestandardized using the procedure of Kofron and Baclawski.143 An“O.32 M tetrahydrofuran solution of lithium diisopropylamide wasprepared by the reaction of diisopropylamine (1.1 equiv) and nbutyllithium (1 equiv) in THE at -78°C (-45 mm). The solution wasnot standardized, and was warmed to 0°C immediately prior to use.Methanol and t-amyl alcohol were distilled from magnesium.Diethylene glycol was distilled from sodium.Oxalyl chloride was distilled before use. Acetyl chloride wasdistilled from PCI5. Dimethyl carbonate was dried over 4Amolecular sieves and was distilled before use. Methyl formate wasdistilled from phosphorus pentoxide. lodomethane was passedthrough a short column of flame dried basic alumina before use.Copper (I) bromide-dimethyl sulfide complex was prepared bythe method described by Wuts.144 Magnesium bromide-etherate wasprepared by the reaction of magnesium metal with 1,2-dibromoethane in diethyl ether, followed by the removal of the155diethyl ether under reduced pressure (-‘0.2 Torr) at roomtemperature.5-C hloro-2-tri methylstannyl- 1 -pentene (5) was preparedaccording to a previously reported modificationOb of the originalprocedure.5 Thus, the reaction mixture was stirred at -78°C for-‘6.5-7 hours, instead of at -63°C for 12 hours and no methanol wasadded during the reaction. The purified product exhibited theexpected H nmr spectrum.The keto ketal 46 was prepared according to the proceduredescribed by Moss38’85 except that the purification of the productwas modified. Thus, a mixture of the keto ketal 46, diketal 153 anddiketone 43 (glc ratio -‘57:28:13, 18.5 g, adsorbed on 36 g of Celite)obtained by the reaction of the diketone 43 with 2,2-dimethyl-1 ,3-propanediol (152), was subjected to flash chromatography on silicagel (400 g, elution with 2:1 diethyl ether-petroleum ether to elutethe diketal 153; elution with 9:1 diethyl ether-ethyl acetate toelute the keto ketal 46; and elution with ethyl acetate to elute thediketone 43). The appropriate fractions were combined to yield 7.5g of the keto ketal 46 as a white solid, m.p. 46.5-47.5°C (lit. m.p.48°C)85 which was spectroscopically (1H nmr) identical with thematerial reported by Moss. The combined recovery of the diketal153 and the diketone 43 was 8.1 g.Distilled solvents were deoxygenated by bubbling argon throughthe stirred solvent for at least 1 h.All other reagents were commercially available and were usedwithout further purification.1564.2. Experimental Procedures for the Synthesis of (±)-J3-Panasinsene (31) via the Weiss-Cook CondensationApproach.Preparation of the Bicyclic Keto Ester Ketals 163 and 164.MeHOA stirred suspension of potassium hydride (0.945 g, 23.6 mmol,2.6 equiv, freed from mineral oil by washing with three 5 mLportions of dry THE) in 43 mL of dry THE, under an argonatmosphere, was warmed briefly to -50°C. The mixture wasallowed to cool to room temperature and a solution of the keto ketal46 (2.002 g, 8.93 mmol) in 2 mL of dry THE was added, with threerinses of dry THE (8 mL total). The mixture was heated at -‘60°C for2 h. To the resultant orange-tan suspension was added quickly drydimethyl carbonate (2.1 mL, 2.2 g, 25 mmol, 2.8 equiv). After thedark-colored mixture had been heated at -‘60°C for a further 1.5 h,it was cooled to 0°C (ice bath). The vigorously stirred reactionmixture was treated with a mixture consisting of 100 mL ofsaturated aqueous ammonium chloride (pH 5), 100 mL of ice, and100 mL of chloroform. The aqueous layer was acidified (1 Nhydrochloric acid) to pH 6-7 and the phases were separated. TheHMe163H164157aqueous phase was extracted with two 50 mL portions ofchloroform. The combined chloroform extracts were washed withbrine, dried (anhydrous sodium sulfate), filtered, and concentratedto give the keto ester 163 and the ester enol tautomer 164 (2.352g, 93%; ratio 1.5:1, ‘H nmr analysis) as an orange oil* whichsolidified slowly to an off-white waxy solid. This material wasused without further purification in the next reaction.The crude material, which consisted of a mixture of 1 63 and164 exhibited ir (neat): 1756 (m), 1729 (s), 1661 (s), 1621 (m),1281 (s), 1202 (s), 1113 (vs) cm-1; ‘H nmr (300 MHz): 30.946, 0.952(s, s, tertiary Me of 1 63), 0.93, 0.98 (s, s, tertiary Me of 164)(combined tertiary Me, 6H), 1.59-1.74 (m, 1H), 1.79-1.86, 2.03 (m,dm (J = 8.0 Hz), 1H total), 2.20-2.42 (m, 3H), 2.54-2.81, 2.86-2.99,3.12-3.32 (m, m, m, 4H total), 3.43-3.50 (m, 4H, both ketal-Cjz12groups), 3.73, 3.75 (s, s, 3H total, CO2Me, 163 and 164,respectively), 10.35 (br s, 0.1H, enol CE); MS m/z (% rel. mt.): 282(M, 41), 250 (12), 226 (10), 223 (44), 213 (17), 181 (22), 167 (37),165 (49), 164 (53), 154 (36), 153 (27), 128 (47), 121 (27), 69(100). Exact Mass calcd. for C,5H220:282.1467; found: 282.1459.In some runs, the oil was deep red. In these cases, the crude product was dissolved indiethyl ether and the resultant solution was filtered rapidly through a short column ofsilica gel (—3X by weight, elution with diethyl ether).158Preparation of the Keto Ester Selenides 165 and 16&MeTo a stirred suspension of potassium hydride (0.455 g, 1.3mmol, 1.3 equiv, freed from mineral oil by washing with three 3 mLportions of dry THE) in 35 mL of dry THE, under an argonatmosphere, was added, over a period of 10 mm, a solution of themixture of the keto esters 163 and 164 (ratio -‘1.5:1, 2.403 g, 8.51mmol) in 5 mL of dry THE, with two rinses of dry THE (10 mL total).After the mixture had been stirred for 40 mm at room temperature,the brown suspension was cooled in an ice bath and a solution ofbenzeneselenenyl chloride (2.20 g, 11.5 mmol, 1.35 equiv) in 2 mL ofdry THF, with several rinses of dry THF (11 mL total), was addedquickly. The orange reaction mixture was stirred at 0°C for 20 mmand then was pipetted carefully (over a period of 10 mm) into avigorously stirred mixture consisting of ice (15 mL), saturatedaqueous sodium bicarbonate (25 mL), and a 1:1 pentane-diethylether mixture (50 mL). The organic layer was separated and theaqueous phase was extracted twice with 1:1 pentane-diethyl ether(75 mL total). The combined organic extracts were washed withbrine (50 mL), dried (anhydrous sodium sulfate), filtered andconcentrated to yield an orange oil (4.098 g, >100%). UnreactedMeO2C HH H165 166159benzeneselenenyl chloride was separated from the crude product byflash chromatography on silica gel (211 g, elution with 2:1petroleum ether-ethyl acetate). A normally unseparated mixture(ratio 4:1, ‘H nmr analysis) of the epimeric selenides 165 and 166(3.185 g, 86%) was obtained as an orange oil. The mixture was usedfor the next reaction without further purification. An -5:1 epimericmixture of the isomers exhibited ir (neat): 1752 (vs), 1730 (vs),1120 (vs), 744 (w), 692 (w), 669 (w) cm-1; nmr (300 MHz): 3 0.89(s, 3H, tertiary Me, 165), 0.92 (S, tertiary Me, 166), 0.97 (s, 3H,tertiary Me, 165), 0.99 (s, tertiary Me, 166), 1.54-1.62 (m, 2H),1.90-2.03 (m, 2H), 2.23-2.34 (m, 3H), 2.44 (dd, 1H, J = 14.5, 7.0 Hz),2.83-3.01 (m, 3H), 3.39 (s, 2H, ketal-CU2), 3.45 (s, -‘2H, ketal -C2-, partially burying a m, 166), 3.51 (s, COjj, 166), 3.71 (s, 3H,COj, 165), 7.29-7.36 (m, -3H, aromatic Ca), 7.53-7.57 (m, 2H,aromatic CJL, 165), 7.63-7.65 (m, aromatic Cth 166).The minor, undesired isomer (166), was less polar and smallamounts could sometimes be isolated pure using thechromatographic procedure described above (vide supra). The minorisomer 166, exhibited ir (neat): 1751 (vs), 1728 (vs), 1117 (vs),742 (s), 693 (m) cm1; 1H nmr (300 MHz): 60.92 (s, 3H, tertiary Me),0.99 (s, 3H, tertiary), 1.90-2.03 (m, 2H), 2.24-2.31 (m, 2H), 2.41(dd,1H, J= 14.0, 9.0 Hz), 2.60-2.78 (m, 2H), 3.29-3.46 (m, 5H, bothketal -Cj- groups and a bridgehead Ca), 3.51 (s, 3H, CO2L’j, 7.29-7.38 (m, 3H, aromatic CU), 7.63-7.65 (m, 2H, aromatic CU). ExactMass calcd. forC21H6O580Se: 438.0946; found: 438.0940.160Preparation of the Enone Ester 15910 9Hb HaF4Ha 14159To a cold (0°C) stirred solution of the mixture of keto esterselenides 165 and 166 (ratio -‘5:1, 2.437 g, 5.57 mmol) in 20 mL ofdistilled dichloromethane was added, over a period of 7 mm, anaqueous hydrogen peroxide solution (2.4 mL of a 15% solution, 11.7mmol, 2.1 equiv). After the mixture had been stirred for 10 mm at0°C and for -‘20 mm at room temperature, 10 mL of water wasadded and the phases were separated. The dichloromethane layerwas washed with water (10 mL). Then each aqueous phase wasextracted with dichloromethane (two 2 mL portions). All thedichloromethane layers were combined, dried (anhydrous sodiumsulfate), filtered and concentrated to yield a yellow-orange oilcontaining a solid. The crude product was diluted with diethyl ether(6 mL) and the mixture was filtered through Celite (elution withdiethyl ether) to remove any benzeneselenenic acid. The eluate wasconcentrated to yield the enone ester 159 (1.589 g, >100%, due tosmall amounts of impurities, including some aromatic products) asan orange oil which was not further purified. The crude productexhibited ir (neat): 1750 (vs), 1719 (vs), 1653 (w), 1274 (m), 1149161(w), 1112 (s) cm1; 1H nmr (300 MHz): 80.94 (s, 3H, tertiary Me),1.09 (s, 3H, tertiary Me), 1.48 (t, 1H, J= 12.5 Hz, H-6a), 2.25 (dd,1H, J= 18.0, 4.0 Hz, H-4a), 2.70 (ddd, 1H, J= 12.5, 8.0, 1.0 Hz, H-6b),2.79 (dd, 1H, J= 18.0, 6.5 Hz, H-4b), 3.14-3.26 (m, 1H, H-5b), 3.30(br s, 2H, H-8a and H-8b), 3.45-3.60 (m, 4H, both ketal-Ca2groups), 3.85 (s, 3H, CO2M.). Some minor signals due to impuritieswere also present. In 1H nmr decoupling experiments (400 MHz),irradiation of the signal at 8 1.48 (H-6b) simplified the ddd at 2.70(H-6a) to a br d (J = 8.0 Hz); irradiation of the signal at 52.25 (H4a) simplified the dd at 2.79 (H-4b) to a br d (J = 6.5 Hz); andirradiation of the multiplet at 6 3.14-3.26 (H-5b) simplified the t at1.48 (H-6b) to a d (J = 12.5 Hz), the dd at 2.25 (H-4a) to a d (J = 18.0Hz), the ddd at 2.70 (H-6a) to a br d (J = 12.5 Hz), and the dd at 2.79(H-4b) to a d (J = 18.0 Hz); MS m/z (% rel. mt.): 280 (Mt, 36), 248(13), 194 (12), 180 (21), 163 (40), 162 (26), 135 (23), 134 (27),121 (20), 69 (100). Exact Mass calcd. forC15H2005:280.1311; found:280.1311.Preparation of the Keto/Enol Ester Chlorides 169/170.CI,MeMeC[MeHHO’169 170162To a cold (-78°C) stirred solution of 5-chloro-2-trimethyl-stannyl-1-pentene 5 (2.585 g, 9.67 mmol, 1.37 equiv) in 60 mL ofdry THE, under an argon atmosphere, was added a solution ofmethyl-lithium in diethyl ether (1.55 M, 7.20 mL, 11.2 mmol, 1.59equiv). After the solution had been stirred for 20 mm at -78°C,anhydrous magnesium bromide etherate (2.886 g, 11.2 mmol, 1.59equiv) was added in one portion. The white suspension was stirredfor 20 mm at -78°C and then copper bromide-dimethyl sulfidecomplex (0.366 g, 1.78 mmol, 0.25 equiv) was added in one portion.The pale yellow suspension was stirred for 20 mm at -78°C and thena solution of the ester enone 159 (1.98 g, 7.04 mmol) in 3 mL of dryTHF, with three rinses of dry THE (9 mL total), was added over 5mm. The orange suspension was stirred for 25 mm at -78°C, andthen was treated with saturated aqueous ammonium chloridesolution (pH ‘-6, 90 mL) and diethyl ether (90 mL). The cooling bathwas removed and after the mixture had been stirred for 10 mm atroom temperature, the phases were separated. The aqueous phasewas extracted with three 60 mL portions of diethyl ether. Thecombined organic extracts were washed with brine (90 mL), dried(anhydrous magnesium sulfate), filtered and concentrated to yield agreen-brown oil which was quickly filtered through a short silicagel column (4.8 g, elution with diethyl ether). Concentration of theeluate yielded the keto/enol ester chlorides 169/170 (2.568 g,94%) as a brown oil which was not further purified, but was useddirectly in the next reaction.The crude keto/enol ester chlorides 169/170 (which existedmainly as the enol tautomer, 170) exhibited ir (neat): 1754 (w),1631722 (w), 1657 (vs), 1619 (s), 1258 (s), 1218 (s), 1116 (vs), 806(w) cm-1; 1H nmr (300 MHz): 3 0.90 (s, 3H, tertiary Me), 1.02 (s, 3H,tertiary Me), 1.70 (dd, 1H, J= 12.5, 8.5 Hz), 1.88-2.04 (m, 2H), 2.07-2.43 (m, 6H), 2.51 (dd, 1H, J= 14.5, 1.5 Hz), 2.74 (dd, 1H, J= 18.0,8.0 Hz), 3.42-3.62 (m, 6H,-Ca2CI and both ketal -Cjj.2 groups), 3.74(s, 3H, CO2f), 4.74 (s, 1H, C=Cjj2), 4.82 (s, 1H, C=Ck12), 10.79 (br s,1H, enol Ca). Some minor signals due to impurities were alsopresent. MS m/z (% rel. mt.): 384 (Mt, 3), 352 (7), 324 (3), 317 (4),307 (5), 281 (5), 266 (11), 249 (11), 203 (18), 161 (15), 154 (17),141 (16), 135 (16), 129 (32), 128 (100), 121 (17). Exact Masscalcd. forC20H90351: 384.1704; found: 384.1704.Preparation of the Tricyclic Keto Ester Ketal 171.Me20To a stirred solution of the keto/enol ester chlorides 169/170 (1.351 g, 3.51 mmol) in 21 mL of freshly distilled acetonitrile,under an argon atmosphere, was added, in one portion, cesiumcarbonate (5.685 g, 17.4 mmol, 5.0 equiv). The suspension wasHaHb’19Me0Hb Hb171164heated at 58-61°C for 20 h. The color changed from dark red-orangeto light brown by the end of the reaction. After the mixture hadbeen cooled to room temperature, cold water (45 mL) and diethylether (40 mL) were added. The deep red aqueous layer wasseparated, was extracted with diethyl ether (40 mL), was acidifiedto pH -‘8 with hydrochloric acid (1 N, -“20 mL) and then wasextracted with three 40 mL portions of diethyl ether. The combinedethereal extracts were washed with brine (two 50 mL portions),dried (anhydrous magnesium sulfate), filtered and concentrated toyield the crude tricyclic keto ester ketal 171 (1.133 g, 93%) as ared-brown oily solid. The crude product (dissolved in 4:1 petroleumether-ethyl acetate (-‘8 mL) and dichloromethane (1.5 mL)) waspurified by flash chromatography on silica gel (135 g, elution with4:1 petroleum ether-ethyl acetate). Concentration of theappropriate fractions yielded colorless crystals (0.846 g, 69%) of97% purity (GLC analysis) which were then recrystallized (threecrops) from hot ethyl acetate and cold petroleum ether (initial ratio-“1:3, additional petroleum ether added twice at 15 mm intervals inportions -“double the ethyl acetate volume) to yield colorlesscrystals (0.787 g, 64.3%). The crystalline material thus obtainedexhibited m.p. 127.5-128.5°C; ir (KBr): 3088 (vw), 1745 (vs), 1723(vs), 1637 (w), 1115 (vs), 921 (m), 893 (w) cm-1; 1 nmr (300 MHz):80.91 (s, 3H, Me-19 or Me-20), 1.01 (s, 3H, Me-20 or Me-19), 1.54-1.71 (m, 2H, H-Ba or H-8b and H-9b or H-9a), 1.72-1.80 (m, 1 H, H-9aor H-9b), 1.88 (d, 1H, J = 16.0 Hz, H-2a), 1.96 (distorted dd, 1H, J =14.0, 6.0 Hz, H-3a), 2.10 (distorted d, 1H, J = 14.0 Hz, H-3b), 2.15-2.21 (m, 1H, H-Bb or H-8a), 2.26-2.39 (m, 2H, H-lOa and H-lob),1652.54 (distorted dd, 1H, J = 19.5, 9.0 Hz, H-5a or H-5b), 2.69(distorted dd, 1H, J= 19.5, 9.0 Hz, H-5b or H-5a), 2.76-2.84 (m, 1H,H-4b), 2.95 (br d, 1H, J = 16.0 Hz, H-2b), 3.40-3.57 (m, 4H, 2H-16and 2H-18), 3.74 (s, 3H, Me-14’), 5.01 (d, 1H, J = 1.0 Hz, H-15b), 5.07(s, 1H, H-15a); 13C nmr (75 MHz): 322.23 (H3-19 orH3-20), 22.32(H3-20 or H3-19), 23.8 (.H2-9), 30.0 (-17), 30.6 (H2-8), 32.4(.H2-10), 38.65 (.H-4), 38.74 (.H2-3), 39.9 (H2-5), 42.8 (H2-2),52.1 (H3-14’), 58.2 (-1 or -7), 68.0 (.-7 or .Q-1), 71.8 (H2-16 or..H2-18), 72.4 (.Q..H2-18 or H2-16), 109.0 (-13), 112.7 (H2-15),145.1 (-11), 170.5 (..-14), 211.8 (-6); MS m/z (% rel. intj: 348(M, 14), 317 (4), 289 (4), 280 (15), 279 (59), 247 (8), 231 (9), 203(11), 193 (26), 165 (22), 161 (25), 155 (22), 129 (61), 128 (100),105 (40). Exact Mass calcd. for C20H80: 348.1936; found:348.1929. Anal. calcd. forC20H805:C 68.94, H 8.10; found: C 69.01,H 8.15.For the HETCOR and COSY data, see Tables 1 and 2, PP. 78 and81, respectively.PreDaration of the Alcohol Ester Ketals 177 and 178.‘MeMeMe177 178166To a stirred solution of the keto ester ketal 171 (152.5 mg,0.438 mmol) in 8.8 mL of reagent grade methanol, under an argonatmosphere, was added cerium trichioride hexahydrate (82.1 mg,0.232 mmol, 0.53 equiv). When the solution became homogeneous(-‘1 mm), it was cooled to -48°C and stirred for 4 mm. Solid sodiumborohydride (21.? mg, 0.560 mmol, 1.3 molar equiv) was added inone portion to the now white suspension. The mixture bubbledvigorously and then cleared somewhat. After the mixture had beenstirred for 1 h at -‘-48°C, the cooling bath was removed and, 2 mmlater, -‘1 N hydrochloric acid (370 .tL, 0.37 mmol, -‘1.2 equiv) wasadded, followed 1 mm later by 10 mL of cold water and 10 mL ofpentane. The mixture was stirred at 0°C for 5 mm, then a further10 mL of pentane was added and the phases were separated. Theaqueous phase was extracted three times with pentane. Thecombined pentane extracts were concentrated to yield a mixture ofa white solid and an oil, which was dissolved in pentane (‘-30 mL).The solution was dried (anhydrous magnesium sulfate), filtered andconcentrated to yield the epimeric alcohols 177 and 178 (150.9mg, 98%; ratio 5.5:1, ‘H nmr analysis) as a white solid. The alcoholscould be recrystallized from an -‘1:2 mixture of ethyl acetatehexane to yield colorless needles which exhibited m.p. 105-107°C(ratio of epimers -‘8:1, ‘H nmr analysis). A 12:1 mixture of epimers(‘H nmr analysis) exhibited m.p. 105-106.5°C. The ratio of epimersin the unrecrystallized mixture varied from -‘4.5:1 to -‘12:1 (1H nmranalysis) depending on the scale of the reaction and slight changesin reaction conditions.167A 12:1 mixture of alcohols 177 and 178 exhibited ir (KBr):3511 (s), 3087 (w), 1726 (vs), 1636 (m), 1240 (s), 1193 (s), 1124(vs), 1101 (vs), 886 (s) cm’; ‘H nmr (400 MHz): 30.90 (s, -3H,tertiary Me), 0.99 (s, -‘3H, tertiary Me), 1.47-1.63 (m, 2H), 1.66-1.74(m, 2H, includes 1.70 (d, J= 16.0 Hz)), 1.79 (dd, 1H, J = 13.5, 7.5 Hz),1.87-2.00 (m, 3H), 2.14 (d, 1H, J= 13.5 Hz), 2.20-2.31 (m, 1H), 2.35-2.42 (m, 2H, includes 2.35 (d, J = 2.0 Hz, Oki. of 177, exchanged withD20)), 2.45-2.51 (m, 1H), 2.65 (d, 1H, J = 16.0 Hz), 3.42-3.54 (m, 4H,both ketal-C±12 groups), 3.68 (s, -3H, C02M.j, 4.69 (td, 1 H, J = 9.0,2.0 Hz, H-6a, simplified to a t (J = 9.0 Hz) upon D20 exchange), 4.85(d, 1H, J= 1.0 Hz, C=C±12), 4.88 (br s, -‘lH, C=Ckj2 both epimers).Signals due to the minor epimer, 178, appeared at: 0.92 (s, tertiaryMe),0.98 (s, tertiary Me), 4.11-4.14 (m, H-6b, simplified upon D20exchange), 5.00 (br s, C=C±12); MS m/z (% rel. mt.) 350 (M, 17), 280(5), 279 (7), 264 (10), 194 (15), 187 (26), 161 (16), 145 (18), 135(22), 129 (70), 128 (100), 107 (16), 105 (24). Exact Mass calcd. forC20H305:350.2093; found: 350.2087. Anal. calcd. for C20H305:C68.85, H 8.63; found: C 68.64, H 8.64.Preparation of the Phenyl Thionocarbonate 179.179168To a stirred solution of the mixture of the alcohols 178 and179 (ratio -5:1, 295.8 mg, 0.844 mmol) in 7.7 mL of dry, freshlydistilled acetonitrile, under an argon atmosphere, was added 4-(N,N-dimethylamino)pyridine (DMAP, 830.5 mg, 6.80 mmol, 8 equiv).The solution was cooled to 10°C (cold water bath) andphenoxythiocarbonyl chloride (180 iiL, 225 mg, 1.27 mmol, 1.5equiv) was added. Within 1 mm a precipitate formed and, after aperiod of 8 mm, the cooling bath was removed and the mixture washeated at 67-72°C for 20 h. The reaction mixture was cooled andthe solvent was removed under reduced pressure to yield a tan solid.This material was suspended in water (25 mL) and the resultantmixture was extracted with ethyl acetate (50 mL, then three 25 mLportions). The combined ethyl acetate extracts were washed,successively, with 1 N hydrochloric acid (30 mL, 20 mL, rapidly),water (30 mL), saturated aqueous sodium bicarbonate (30 mL) andbrine (two 30 mL portions). The organic phase was dried (anhydroussodium sulfate) and concentrated to yield the crude phenylthionocarbonate 179 as an oil (463.4 mg, > 100% due to thepresence of impurities). The crude product was dissolved in aminimum volume of dichloromethane and purified by flashchromatography on silica gel (84 g, elution with 9:1 petroleumether-ethyl acetate). The appropriate fractions were combined andconcentrated to yield the pure phenyl thionocarbonate 179 (310.7mg, 76%) as one epimer. The purified product thus obtained could berecrystallized from a minimum volume of hot ethyl acetate and coldpetroleum ether to give colorless plates which exhibited m.p. 151-152.5°C; ir (KBr): 3091 (w), 3066 (w), 1736 (s), 1639 (w) 1594 (w),1691395 (m), 1296 (vs), 1236 (vs), 1169 (s), 1124 (s), 1106 (s), 888(m), 774 (m), 690 (m) cm; ‘H nmr (300 MHz): 8 0.95 (s, 3H,tertiary Me), 0.97 (s, 3H, tertiary Me), 1.69-1.74 (m, 3H), 1.82-1.88(m, 2H), 1.95-2.06 (m, 2H), 2.12 (d, 1H, J = 13.5 Hz), 2.24-2.55 (m,3H), 2.60-2.68 (m, 2H), 3.40-3.49 (m, 2H, ketal-CU2), 3.53 (br s,2H, ketal-CU2), 3.70 (s, 3H, CO2j4&j, 4.93 (br s, 2H, C=CU2), 6.08 (t,1H, J= 8.5 Hz, H-6a), 7.09-7.12 (m, 2H, aromatic CU), 7.25-7.30 (m,1 H, aromatic CE), 7.38-7.43 (m, 2H, aromatic Ca); MS m/z (% rel.mt.): 486 (M, 0.9), 333 (25), 273 (11), 247 (28), 187 (53), 159 (22),145 (29), 129 (43), 128 (100), 117 (11), 105 (10). Exact Masscalcd. for C27H3406S: 486.2076; found: 486.2082. Anal. calcd. forC27H3406S:C 66.66, H 7.06, S 6.59; found: C 66.56, H 7.10, S 6.54.Preparation of the Ester Ketal 181.Me20To a stirred solution of the phenyl thionocarbonate 179 (117.0mg, 0.240 mmol) in 2.4 mL of dry, degassed benzene, under an argonatmosphere, was added tri-n-butyltin hydride (160 p.L, 173 mg,0.595 mmol, 2.5 equiv) and recrystallized 2,2’-azobisisobutyro-Hb181170nitrile (AIBN, 5.9 mg, 43 p.mol, 0.18 equiv). After argon had beenpassed over the surface of the reaction mixture for 10 mm, thesolution was heated at 75-79°C for 20 h. The reaction mixture wascooled and the solvent was removed under reduced pressure to yieldan oil which was purified by flash chromatography on silica gel (44g, elution with 10:1 petroleum ether-ethyl acetate). Theappropriate fractions were combined and concentrated to yield theslightly impure ester ketal 181 (74.4 mg, 92%). The pure esterketal 1 81 (58.0 mg, 72%) was obtained by combining andconcentrating the appropriate fractions after further flashchromatography on silica gel (17.9 g, elution with 10:1 petroleumether-ethyl acetate; 5.5 g, elution with 15:1 petroleum ether-ethylacetate). The product could be recrystallized from hot hexane togive colorless plates, which exhibited m.p. 87.5-89.5°C; ir (KBr):1719 (vs), 1638 (w), 1203 (m), 1179 (s), 1163 (s), 1116 (s), 892(m) cm1; 1H nmr (300 MHz): ö 0.95 (s, 3H, tertiary Me), 0.97 (5, 3H,tertiary Me), 1.53-1.73 (m, 5H), 1.78 (d, 1H, J = 15.5 Hz), 1.82-2.09(m, 4H), 2.23-2.40 (m, 3H), 2.50-2.58 (m, 1H), 2.68 (d, 1H, J= 15.5Hz), 3.41-3.52 (m, 4H, both ketal-Cj[2 groups), 3.63 (s, 3H, CO2M..),4.87 (s, 1H, C=Cj), 4.92 (s, 1H, C=CE2); 13C nmr (75 MHz): 3 22.3(.H3-19 or .H3-20), 22.4 (.H3-20 or .H3-19), 23.4 (.H2), 26.6 (.H2),29.9 (-17), 32.8 (.H2), 33.8 (H2), 34.7 (H2), 39.7 (j, 43.8 (j,45.0 (H-4), 51.2 (.H3-14’), 59.2 (.-1 or -7), 60.0 (-7 or .-1),71.7 (H2-16 or .H2-18), 72.4 (H2-18 or H2-16), 109.8 (.H2-15),110.0 (.-13), 149.1 (-11), 176.3 (.-14); MS m/z (% rel. mt.): 334(M, 10), 275 (12), 189 (12), 145 (14), 131 (14), 129 (40), 128(100), 117 (15), 105 (28). Exact Mass calcd. forC20H304:334.2144;171found: 334.2141. Anal. calcd. forC20H304:C 71.82, H 9.04; found: C72.11, H 9.19.Preparation of the Keto Ester 182.To a stirred solution of the ketal ester 181 (29.5 mg, 88.4i.Lmol) in 2.5 mL of reagent grade acetone was added 1 Nhydrochloric acid (44 iL, 44 .tmol, 0.5 equiv). The solution wasstirred at room temperature for 5.5 h and then was added to 5 mL ofwater. The aqueous suspension was extracted with four 5 mLportions of diethyl ether. The combined ether phases were washedwith water (5 mL), brine (5 mL), dried (anhydrous magnesiumsulfate), and concentrated to yield the crude keto ester 1 82 as acolorless oil (22.4 mg, >100% due to the presence of impurities).The crude product was purified by flash chromatography on silicagel (5.5 g, elution with 3:1 petroleum ether-ethyl acetate) and theappropriate fractions were combined to yield a colorless oil whichwas distilled (115-1 20°C/0.3 Torr) to provide the pure keto ester182172182 (18.6 mg, 84.7%) as a colorless oil. The oil could berecrystallized from pentane to yield colorless needles whichexhibited m.p. 43-43.5°C; ir (KBr): 3080 (m), 1729 (vs)1 1633 (m),1236 (vs)1 1157 (vs)1 903 (s), 890 (m) cm’; ‘H nmr (300 MHz): S1.38-1.50 (m, 1H, H-5a), 1.52-1.65 (m, 1H), 1.67-1.73 (m, 1H), 1.78(distorted dd, 1H, J = 13.5, 3.5 Hz) overlapped with 1.84 (br tt, 1H, J= 12.5, 3.5 Hz, H-6a or H-6b), 1.98 (dm, 1H, J= 13.5 Hz), 2.08 (dd,1H, J = 19.0, 1.0 Hz, H-3a), 2.19 (dd, 1H, J = 19.0, 0.5 Hz, H-2a),2.25-2.42 (m, 5H, H-3b, H-5b, H-6b or H-6a, H-lOa and/or H-lob),2.73 (dd, 1H, J= 19.0, 1.0 Hz, H-2b), 2.84-2.92 (m, 1H, H-4b), 3.64(s, 3H, Me-14’), 4.65 (s, 1H, H-15a), 4.87 (s, 1H, H-15b). In ‘H nmrdecoupling experiments (400 MHz), irradiation of the dd at 52.08(H-3a) sharpened the two dd at 2.19 (H-2a) and 2.73 (H-2b), andsimplified the m at 2.25-2.42. (H-3b); irradiation of the dd at 52.19(H-2a) simplified the dd at 2.73 (H-2b) to a br d (J = 1 Hz) andsharpened the dd at 2.08 (H-3a), the m at 2.25-2.42 (H-3b) and them at 2.84-2.92 (H-4b); irradiation of the dd at 3 2.73 (H-2b)simplified the dd at 2.08 (H-3a) to a d (J = 19 Hz) and the dd at 2.19(H-2a) to a br s, and sharpened the two multiplets at 2.25-2.42 (H3b) and 2.84-2.92 (H-4b); and irradiation at 52.88 (the center of them at 2.84-2.92 (H-4b)), simplified the dd at 2.73 (H-2b) to a d (J =19 Hz), sharpened the m at 2.25-2.42 (H-3b and H-5b) and simplifiedthe m at 1.38-1.50 (H-5a). ‘3C nmr (75 MHz): 823.6 (H2), 28.5(.H2), 32.4 (H2), 33.5 (.H2), 35.1 (H2), 42.1 (H-4), 43.0 (H2),45.9 (H2), 51.6 (H3-14’), 57.3 (-1 or -7), 58.0 (.-7 or -1),109.8 (Q.H2-15), 148.4 (-11), 175.7 (-14), 219.2 (-13); MS m/z(% ret. mt.): 248 (M, 34), 220 (11), 216 (21), 191 (34), 190 (17),173189 (100), 188 (76), 180 (30), 161 (32), 147 (39), 145 (29), 133(26), 131 (88), 130 (51), 119 (43), 117 (28), 107 (28), 105 (60).Exact Mass calcd. for C15H2003:248.1412; found: 248.1416. Anal.calcd. forC15H2003:C 72.55, H 8.12; found: C 72.62, H 8.15.For the COSY data, see Table 3, p. 99.Preparation of the Diesters 200 and 201.a) Preparation of the formylated keto esters 197 and 1 98MeO2Ci:Q197a R=CHO, R’=H197b R=H, R’=CHOTo a cold (5-10°C) stirred solution of the keto ester 182 (160.5mg, 0.646 mmol) in 2.1 mL of dry benzene, under an argonatmosphere, was added, dropwise, a benzene solution of sodium tamyloxide (0.78 M, 3.3 mL, 2.58 mmol, 4 equiv). After a period of 6mm the cooling bath was removed, and the solution was stirred atroom temperature for 1.5 h. The mixture was recooled to 5-10°Cand freshly distilled methyl formate (330 iiL, 311 mg, 5.17 mmol, 8equiv) was added quickly and the reaction mixture was stirred for afurther 17 h at 5C to rt. The solvent was removed under reduced1981 74pressure to yield a reddish-orange oil which was dissolved inaqueous sodium hydroxide (10 mL of a 0.25 N solution). The solutionwas extracted with 2 portions of dichioromethane. The basicaqueous phase was acidified with 1 N hydrochloric acid (-4 mL) andthe resultant mixture was extracted with dichloromethane (10 mL,four 5 mL portions). The combined dichioromethane solutions werewashed with brine (8 mL), dried (anhydrous magnesium sulfate), andconcentrated to yield an isomeric mixture of the formylated ketoesters 197 and 198 (185.7 mg, >100% due to the presence of minorimpurities; ratio 1 :-‘8.5, 1 H nmr analysis) as a yellow oil. Themixture was not purified further, but was used directly in the nextreaction. The crude formylated keto esters 197 and 198 exhibitedir (neat): 1725 (vs), 1699 (s), 1609 (s, broad), 1227 (s), 1202 (s),1157 (vs), 1084 (m), 893 (w), 800 (w) cm-1; 1 H nmr (400 MHz): 31.48-1.77 (m, 5H), 1.78-1.89 (m, 1H), 1.90-1.95 (m, 1H), 2.23-2.42(m, 5H), 2.48 (d, 1H, J = 18.5 Hz), 2.77 (d, 1H, J = 18.5 Hz), 3.20-3.23(m, bridgehead proton), 3.63 (s, 3H, CO2M), 4.64 (distorted d, 1H, J= 1.5 Hz, C=Cj12), 4.84 (d, 1H, J = 1.5 Hz, C=CU2), 7.08 (s, 1H,C=CaOH). Signals due to the major aldehyde isomer (197a)appeared at 31.98-2.04 (m), 3.68 (s, CO2M.), 4.49 and 4.78 (d, d, J =‘1.5 Hz, C=Cjj2), 9.52 and 9.83 (s, 5, CjjO, 197a and 197b,respectively); MS m/z (% rel. mt.): 276 (M, 46), 248 (12), 244 (10),217 (100), 216 (59), 208 (30), 189 (33), 188 (36), 187 (34), 173(36), 147 (33), 145 (38), 131 (43), 129 (30), 117 (31), 115 (30),105 (46). Exact Mass calcd. for C16H2004: 276.1361; found:276.1 363.175b) Preparation of the cx-diazoketone 199.MeO2C1To a stirred solution of the crude formylated keto ester mixture197 and 198 (ratio 1:-8.5, 71.3 mg, 0.258 mmol) in 1.9 mL of drydichloromethane at 0°C, under an argon atmosphere, was addedmethanesulfonyl azide* (M5N3, 29 iL, 40.6 mg, 0.335 mmol). After 3mm, freshly distilled dry triethylamine (54 .tL, 39 mg, 0.387 mmol)was added. The reaction mixture was fully protected from light andstirred at 0°C for 4 h. The following experimental manipulationswere performed in a dimly lit room. The reaction mixture wastreated with 40 drops of water and 40 drops of 6% aqueouspotassium hydroxide and then was stirred for 5 mm. Distilleddichloromethane (4 mL) was added and the phases were separated.The aqueous phase was extracted with 4 more portions ofdichioromethane. The combined organic phases were washed with asmall amount of brine, dried (anhydrous magnesium sulfate) andconcentrated. The resultant off-white solid and yellow oil weredissolved in a small volume of 1:1 pentane-diethyl ether and theresulting suspension was filtered quickly through a short column ofsilica gel (0.4 g, elution with 1:1 pentane-diethyl ether). The eluate199CAUTION: All sulfonyl azides are potentially explosive.176was concentrated* to yield the a-diazo ketone 199 (44.5 mg, 62%) asa yellow oil which exhibited ir (neat): 2082 (vs), 1727 (s), 1674 (s),1636 (w), 1352 (m), 1247 (m), 1157 (m), 898 (w) cm1. Thismaterial was not stable and, therefore, was used immediately in thenext reaction.c) Preparation of the diesters 200 and 201.The crude a-diazo ketone 199 was dissolved in 10.4 mL ofdeoxygenated, distilled methanol (-‘0.025 M solution, based on 71.3mg, 0.258 mmol of formylated keto esters 197 and 198) in a quartzphotolysis tube (dimensions: 8 cm X 1.5 cm). After argon had beenpassed over the solution for 10 mm, the vessel was closed with aglass stopper and placed as close as possible to a medium pressureHanovia mercury lamp (450 watt, Corex filter in a water-cooledquartz jacket). The reaction mixture was photolyzed at 0°C (the* If all the triethylamine was not removed from the product at this stage, the yield of thediester mixture from the photolysis reaction was reduced due to side reactions.200Hb201177lamp and reaction vessel were immersed in a cooling bath) for 30mm. In order to follow the progress of the reaction by TLC or irspectroscopy, the reaction vessel could be opened under a stream ofargon and a small aliquot removed. After the solvent had beenremoved, a mixture of the ring contracted diesters 200 and 201(36.5 mg, 51%) was obtained as a colorless oil (ratio -1.6:1, gicanalysis). The crude product was purified by flash chromatographyon silica gel (5.5 g, elution with 8:1 hexanes-ethyl acetate). Theappropriate fractions were combined to yield the less polar majordiester epimer 200 (5.6 mg) as a solid and a mixture of the diesters200 and 201 (23.5 mg) as a colorless oil (29.1 mg total, 40.5%,from the keto ester 182). The diester 200 could be recrystallizedfrom pentane to yield colorless prisms which exhibited m.p. 39.5-40°C; ir (KBr): 3101 (w), 1728 (vs), 1640 (w), 1439 (s), 1244 (s),1199 (s), 1156 (s), 883 (m) cm;1H nmr (400 MHz): 81.41-1.52 (m,2H, H-Ba or H-8b and H-9b or H-9a), 1.58-1.65 (m, 2H, H-5a and HBb or H-Ba), 1.78 (dd, 1H, J = 13.0, 8.0 Hz, H-6b), 1.89 (dm, 1H, J =10.5 Hz, H-9a or H-9b), 1.95-2.07 (m, 2H, H-5b and H-lOa or H-lob),2.26 (dd, 1H, J = 13.5, 10.5 Hz, H-2a), 2.33 (dm, 1H, J = 13.5 Hz, HlOb or H-lOa), 2.39-2.44 (m, 1H, H-2b), 2.43-2.50 (m, 1H, H-6a),2.71 (br td, 1H, J = 9.0, 3.5 Hz, H-4b), 3.18 (dt, 1H, J = 10.5, 9.5 Hz,H-3b), 3.62 (s, 3H, Me-14’), 3.68 (s, 3H, Me-13’), 4.93 (s, 1H, H-l5a),4.96 (s, 1H, H-15b); 13C nmr (75 MHz): 623.8 (H2-8), 25.2 (H-5),26.6 (.H2-2), 32.7 (.H2-9 and H2-10), 36.2 (.H-3), 37.2 (.H-6),46.8 (QH-4), 51.37 (H3-13’ orH3-14’), 51.40 (.H3-14’ or .H3-13’),52.4 (.-1 or-7), 57.9 (Q-7or.-1), 107.7 (H2-15), 147.9 (-11),174.0 (.-13 or -14), 175.5 (-14 or -13). In nOe difference nmr178experiments (400 MHz), irradiation at 3 2.71 (H-4b) led toenhancement at 3.18 (H-3b); irradiation at 3.18 (H-3b) led toenhancements at 2.39-2.44 (H-2b), 2.71 (H-4b) and 4.93 (H-15a);irradiation at 4.95 (H-15a/H-15b) led to enhancement at 2.33 (HlOb or H-lOa), 2.39-2.44 (H-2b) and 3.18 (H-3b). In 1H nmrdecoupling experiments (500 MHz), irradiation of the dt at 3 3.18(H-3b) simplified the td at 2.71 (H-4b) to a distorted dm (J = 9 Hz),simplified the rn at 2.39-2.44 (H-2b) and simplified the dd at 2.26(H-2a) to a d (J = 13 Hz); irradiation of the td at 2.71 (H-4b)simplified the dt at 3.18 (H-3b) to a dd (J = -‘9, -‘10 Hz) andsimplified the mutiplets at 2.39-2.44 (H-2b) and at 1.95-2.07 (H-5band another H); and irradiation of the dd at 2.26 (H-2a) simplifiedthe m at 2.39-2.44 (H-2b) and simplified the dt at 3.18 (H-3b) to adistorted t (J = 9 Hz). MS m/z (% rel. mt.): 278 (M, 1.8), 247 (19),246 (100), 219 (34), 218 (58), 187 (39), 186 (35), 165 (36), 160(28), 159 (77), 158 (24), 145 (24), 133 (54), 132 (22), 131 (26),119 (21), 117 (33), 107 (20), 105 (29). Exact Mass calcd. forC16H2204:278.1518; found: 278.1527.The ‘-1.6:1 mixture of epimers 200 and 201 (colorless oil)exhibited ir (neat): 1732 (vs), 1639 (w), 1435 (m), 1241 (m), 1200(m), 1180 (m), 1157 (m), 889 (w) cm1;H nmr (400 MHz): 31.40-1.65 (m, ‘-4H), 1.69 and 1.78 (dd, dd, 1H total, J= 13.5, 7.0 Hz; J =13.0, 8.0 Hz, 201 and 200, respectively), the latter dd overlappedwith 1.81-2.08 (m, 2H), which overlapped with 2.06-2.16 (m, <1H,201), 2.23-2.35 (m, 2H), 2.39-2.53 (m, 2H), 2.71 and 2.77 (td, ddd,1H total, J = 9.0, 3.5 Hz; J = 14.0, 5.0, 2.5 Hz, 200 and 201,respectively), 3.18 (dt, <1H, J= 10.5, 9.5 Hz, 200), 3.62 (s, 3H, Me-17914’), 3.67 and 3.68 (s, s, 3H total, epimeric CO2,201 (Me-12’) and200 (Me-13’), respecUvely), 4.92-4.96 (m, 2H, C=C2). MS m/z (%rel. mt.): 278 (M, 3.2), 247 (23), 246 (100), 219 (27), 218 (39), 187(44), 186 (48), 165 (26), 160 (25), 159 (63), 145 (21), 133 (53),131 (24), 119 (22), 117 (27), 107 (22), 105 (29). Exact Mass ca!cd.for C16H2204:278.1518; found: 278.1524. Anal. calcd. for C16H2204:C 69.04, H 7.97; found: C 69.30, H 8.03.For the HETCOR and COSY data, see Tables 4 and 5, pp. 111 and116, respectively.Preparation of the Diesters 202 and 203.To a cold (-78°C), stirred solution of lithium diisopropylamide(0.33 M, 630 p.L, 208 .tmol, 4.0 equiv) in dry THF, under an argonatmosphere, was added (via a cannula) a solution of the mixture ofdiesters 200 and 201 (ratio 2:1, 14.1 mg, 50.7 iimol) dissolved in215Hb202Hb 13203180200 iL of dry THE and rinsed in with three portions of dry THE (‘-3001iL total). After the solution had been stirred at -78°C for 1.5 h,hexamethylphosphoramide* (HMPA, 14 jiL, 14.4 mg, 80.5 iimol, 1.6equiv) was added and the solution was stirred for 12 mm. Freshlydried iodomethane (48 iLL, 109 mg, 771 jimol, l5equiv) was thenadded quickly and the reaction mixture was stirred for a further 30mm at -78°C. The reaction mixture was warmed to ‘-5°C over aperiod of 50 mm and then was treated with saturated aqueousammonium chloride and dilute aqueous sodium thiosulfate (justenough to decolorize the solution). The mixture was extracted withdiethyl ether (4 portions). The combined ethereal extracts werewashed with 2 portions of brine, dried (anhydrous magnesiumsulfate) and concentrated. The resultant yellow oil was dissolved ina small amount of diethyl ether and the solution was filteredthrough a short column of silica gel (‘-0.4 g, elution with diethylether). The colorless eluate was concentrated to yield 12.9 mg(87%) of a mixture of the crude diesters 202 and 203 (12.9 mg,87%; ratio ‘-18:1, 1H nmr analysis), which was purified bychromatography on silica gel (0.9 g, elution with 15:1 hexanes-ethylacetate; repeated 2-3 times until the material obtained consistedonly of a mixture of the two diesters 202 and 203). Theappropriate fractions were combined to yield 9.6 mg (65%) of amixture of the pure diesters 202 and 203 (9.6 mg, 65%; ratio ‘-18:1,nmr analysis) as a colorless oil. The mixture of diesters 202 and203 exhibited ir (neat): 1729 (vs), 1642 (w), 1456 (w), 1240 (rn),* CAUTION: HMPA is known to be a potent carcinogen.1811146 (s), 887 (w) cm-1; H nmr (400 MHz): 31.42 (s, 3H, Me-12),1.46-1.59 (m, 3H), 1.63-1.68 (m, 1 H), overlapped with 1.68 (dd, 1 H, J= 13.0, 7.5 Hz), 1.85 (dm, 1H, J = 11.0 Hz), 1.95-2.06 (m, 2H, H-lOa orH-lOb and another H), 2.26 (dd, 1H, J= 14.0, 3.0 Hz, H-2b), 2.28-2.39(m, 3H, H-4b and 2H), 2.45 (d, 1H, J = 14.0 Hz, H-2a), 3.61 (s, 3H, Me-14’), 3.69 (s, 3H, Me-13’), 4.98 (s, 1H, H-15a), 5.00 (s, 1H, H-15b).Signals due to the minor isomer (203) appeared at 31.18 (s, Me-13),3.60 (s, CO2j), 3.66 (s, CO2j), 4.85 (s, C=Cj[2), 4.87 (s, C=C1j2). In‘H nmr decoupling experiments (400 MHz), irradiation of the signalat 3 1 .85 (dm, 1 H, J = 11.0 Hz) led to simplification of the multipletsat 1.46-1.59 (3H) and 2.28-2.39 (H-4b and 2H); irradiation of the mat 32.00 (H-lOa and H-lob) led to simplification of a d in the signalat 1 .46-1.59 (m, 4H), the collapse of the dd at 1.68 to a distorted d(J= 13.0 Hz) and simplification of the m at 2.28-2.39 (H-4b, and 2H);and irradiation of the signal at 3 2.45 (H-2a) simplified the dci at2.26 (H-2b) to a distorted t (J = 3.0 Hz). In nOe differenceexperiments (400 MHz), irradiation of the singlet at 6 1 .42* ppm(Me-12) led to enhancement of the signals at 1.85, 1.95-2.06 (H-lOaand H-lOb), 2.26 (H-2b), -‘2.35 (H-4b), 3.69 (Me-13’), and 4.98 (H15a); and irradiation of the signal at 64.98 (H-15a) led toenhancement of the signals at 2.26 (H-2b) and 1.42 (Me-12); 13C nmr(75 MHz): 624.7 (.H2), 25.3 (H-12), 26.3 (H2), 32.1 (H2), 33.1(H2), 33.3 36.1 (H2), 42.0 (C), 49.9 (C), 51.4 (H3-13’, orQ..H3-14’), 51.5 (Q.H3-14’, or H3-13’), 53.6 (H-4), 58.2 (C), 109.2(.H2-15), 149.8 (-11), 175.6 (-13 or .-14), 176.2 (.-14 or -13);* Unavoidable irradiation of part of the multiplet at 1.46-1 .59 also occurred as thesignals were too close.182MS m/z (% rel. mt.): 292 (M, 2.4), 261 (25), 260 (100), 233 (22), 232(42), 201 (34), 200 (55), 173 (87), 172 (30), 145 (51), 133 (47), 131(45), 117 (36), 107 (32), 105 (52). Exact Mass calcd. for C17H2404:292.1674; found: 292.1675. Anal. calcd. for C17H2404:C 69.84, H8.27; found: C 69.58, H 8.51.For the COSY data, see Table 6, P. 123.Preparation of the Diols 204 and 205HO•QHA solution of a mixture of the diesters 202 and 203 (ratio-18:1, 41.8 mg, 0.143 mmol) in 400 j.tL of dry THE was added via acannula (with four 200 iL rinses of THE), over a period of 15 mm, toa cold (0°C) stirred solution of lithium aluminum hydride (12.7 mg,0.335 mmol, -‘2 molar equiv) in 800 tL of dry THE, under an argonatmosphere. The cooling bath was removed and the mixture wasstirred at room temperature for 1.2 hours. Then sodium sulfatedecahydrate was added cautiously to react with the excess lithiumaluminum hydride. The mixture was stirred for a few minutesbefore it was filtered through a short column of Elorisil (-‘0.28 g,elution with THE and small amounts of diethyl ether). The solvent1183was removed under reduced pressure to give the crude product (35mg, >100% due to the presence of impurities) which was purified byflash chromatography on silica gel (6 g, elution with diethyl ether).The sample was loaded on the column by dissolving it in a mixture ofacetone (‘-250 p,L) and hot ethyl acetate (‘-800 p.L). The appropriatefractions were combined and concentrated to yield the pure diol 204(29.9 mg, 88%, major epimer) as a solid.The diol 204 could be recrystallized from ethyl acetate-pentaneto yield colorless needles which exhibited m.p. 139-139.5°C (sealedtube; 204 sublimed at 137°C on a Fisher-Johns melting pointapparatus); ir (KBr): 3312 (vs), 1635 (m), 1401 (s), 1039 (s), 1024(s), 885 (m) cm-1; ‘H nmr (400 MHz, acetone-d6 external TMS): 81.11(s, 3H, Me-12), 1.24 (tdd, 1H, J= 13.5, 4.5, 1.0 Hz), 1.40 (qt, 1H, J=13.0, 3.5 Hz), 1.49-1.63 (m, 2H), 1.66 (d, 1H, J= 12.5 Hz), 1.79-2.00(m, 5H), 2.08 (br td, 1H, J= 13.0, 4.5 Hz, partially buried under theacetone peak), 2.16-2.22 (m, 2H), 3.20 (t, 1H, J = 5.5 Hz, 01±,exchanged with D20* ), 3.23-3.30 (m, 2H, Ca2OH, simplified upon D20exchange), 3.32 (t, 1 H, J = 5.0 Hz, 01± exchanged with D20), 3.40 (dd,1H, J= 10.5, 5.5 Hz, C20H, simplified to a d (J= 10.5 Hz) upon D20exchange), 3.46 (dd, 1H, J= 10.5, 5.0 Hz, Ca2OH, simplified to a d (J=10.5 Hz) upon D20 exchange), 4.81 (d, 1H, J= 1.5 Hz, C=C1j2), 4.89(distorted dd, 1H, J= 1.5, 1.5 Hz, C=Ca2);‘3C nmr (125 MHz, acetoned6): 825.1, 25.58, 25.62, 30.0 (buried in the acetone Me), 31.0, 34.3,36.5 (C), 38.1, 51.9 (C), 52.4 (C), 52.6, 63.3 (.H20H-13 or .H20H-14),67.7 (.H20H-14 or .H20H-13), 108.6 (.H2-15), 153.7 (.-11); MS m/z* The addition of D20 caused the chemical shifts of non-exchanged hydrogens to change.184(% rel. mt.): 236 (M, 0.4), 218 (4.8), 206 (8.6), 205 (36), 200 (3.2),187 (60), 147 (78), 146 (52), 145 (46), 133 (45), 131 (40), 109 (42),105 (69), 91 (100). Exact Mass calcd. forC15H240:236.1776; found:236.1780; Anal. calcd. forC15H240:C 76.23, H 10.23; found: C 76.25,H 10.24.A very small amount of the minor (less polar) epimer 205 (anoil), which was isolated from repeated flash chromatography of thecombined impure fractions from several reactions using theprocedure described above, exhibited ir (neat): 3305 (s), 1635 (m),1448 (m), 1053 (s), 1030 (s), 1000 (m), 887 (m) cm1; 1H nmr (400MHz, acetone-d6):80.93 (s, 3H, Me-13), 1.27-1.31 (m, 1H), 1.41 (qt,1H, J= 13.0, 4.0 Hz), 1.52-1.64 (m, 3H), 1.72 (dd, 1H, J= 13.5, 7.0Hz), 1.80-2.24 (m, -7H, includes acetone peak), 3.20 (t, 1 H, J = 5.5Hz, 01± exchanged with D20), 3.30-3.49 (m, 5H, simplifies upon D20exchange to give four distorted d at 3.26 (br d, J = 11.0 Hz), 3.31 (J =10.5 Hz), 3.35 (J= 11.0 Hz) and 3.43 (J= 10.5 Hz)), 4.79 (d, 1H, J=1.5 Hz, C=CE2), 4.87 (distorted dd, 1H, J= 1.5, 1.5 Hz, C=Ca2); MS m/z(% rel. mt.): 236 (M, 2.3), 218 (3.6), 206 (13), 205 (69), 200 (1.7),187 (54), 159 (28), 147 (68), 146 (36), 145 (43), 133 (32), 131 (37),123 (25), 119 (27), 117 (27), 109 (31), 107 (23), 105 (61), 91 (100).Exact Mass calcd. forC15H240:236.1776; found: 236.1777.185Preparation of the Diacetate 213.To a cold (0°C), stirred solution of the diol 204 (5.3 mg, 22tmol) and 4-(N,N-dimethylamino)pyridine (DMAP, -‘3 mg, -‘25 j.tmol,-‘1.1 equiv) in 800 p.L of dry dichloromethane, under an argonatmosphere, was added dry pyridine (16.5 .tL, 16.1 mg, 0.20 mmol, 9equiv) followed 5 mm later by freshly distilled acetyl chloride (9.6tL, 10.6 mg, 0.14 mmol, 6 equiv). The mixture was stirred at 0°C for3 h and then diethyl ether (8 mL) and 0.15 N hydrochloric acid (1.3mL) were added. The aqueous phase was extracted with two portionsof diethyl ether. The combined ethereal phases were washed withwater (1.5 mL), saturated sodium bicarbonate (1.5 mL), and brine(two -‘1 .5 mL portions) and then were dried (anhydrous magnesiumsulfate) and concentrated to yield the crude diacetate 213 (7.4 mg,>100% due to small amounts of impurities) as an oil. The crudediacetate 213 was purified by chromatography on silica gel (0.9 g,elution with 1:1 pentane-diethyl ether). The appropriate fractionswere combined and concentrated to yield the diacetate 213 (6.1 mg,85%) as a colorless oil which exhibited ir (neat): 1742 (vs), 1636HHb213186(w), 1238 (vs), 1034 (m), 889 (w) cm; 1H nmr (400 MHz): 31.15 (s,3H, Me-12), 1.29-1.39 (m, 1H), 1.43 (br qt, 1H, J = 13.0, 3.5 Hz), 1.54(dd, 1H, J = 12.5, 7.5 Hz, partially buried under water), 1.62-1.72 (m,3H), 1.72 (d, 1H, J = 13.0 Hz, H-2a), 1.83 (dt, 1H, J = 7.0, 12.5 Hz),1.94-2.07 (m, 9H, which includes: 1.94-2.07 (m, 2H; H-lOa or H-lOband H-x), 1.98 (dd, 1H, J = 13.0, 3.0 Hz, H-2b), 2.02 (s, 3H, Me-13” orMe-14”), 2.05 (s, 3H, Me-14” or Me-13”)), 2.21-2.27 (m, 1H),overlapped with 2.27 (br dd, 1H, J = 9.0, 3.0 Hz), 3.79 (dd, 1H, J =11.0, 1.0 Hz, H-14a), 3.86 and 3.89 (AB pair of d, 2H, J= 11.0 Hz, 2H-13), 4.03 (d, 1H, J = 11.0 Hz, H-14b), 4.84 (s, 1H, H-iSa), 4.96 (s, 1H,H-15b); ‘3C nmr (125 MHz): 3 20.88, 20.92, 24.56, 24.67, 25.2, 30.3,30.7, 33.3, 33.7 (Q.), 37.1, 49.1 (j, 51.3, 51.9 (j, 66.0 (H2-13 orH2-i4), 69.1(H2-14 or H2-13), 109.4 (H2-15), 150.8 (-11),171.2 (.-13’ or -14’), 171.3 (-14’ or Q.-13’); MS m/z (% rel. mt.):320 (Mt, 9.7), 260 (4.8), 247 (4.6), 201 (ii), 200 (47), 187 (92), 185(34), 172 (60), 159 (43), 157 (23), 146 (52), 145 (52), 133 (28), 131(47), 120 (54), 119 (27), 117 (31), 105 (72), 91 (100). Exact Masscalcd. for C19H28: 320.1988; found: 320.1994. Anal. calcd. forC19H2804:C 71.22, H 8.81; found: C 71.48, H 8.82.For the COSY and NOESY data, see Tables 7 and 8, pp. 133 and134, respectively.187Preparation of the Dialdehyde 217.OHCiTo a cold (-78°C), stirred solution of dry dimethyl sulfoxide(DMSO, 54 pL, 0.76 mmol) in 300 p.L of dry dichioromethane, under anargon atmosphere, was added freshly distilled oxalyl chloride (34.tL, 49 mg, 0.39 mmol). After the mixture had been stirred for 30mm, a solution of the diol 204 (10.0 mg, 42 p.mol) in 40 p.L of dryDMSO and 500 iiL of dry dichioromethane was added via a cannula andrinsed in with dry dichloromethane (700 i.tL total). The reactionmixture was stirred for 40 mm at -78°C and then dry triethylamine(240 .iL, 174 mg, 1.72 mmol) was added and the mixture waswarmed to —0°C over a period of 70 mm. Water (—3 mL) was addedand the product was extracted with dichloromethane (5 mL, four 3mL portions). The combined organic phases were dried (anhydrousmagnesium sulfate) and concentrated to give a mixture of an oil anda white solid. The crude product was suspended in diethyl ether andthe mixture was filtered through a short silica gel column (—0.2 g,elution with diethyl ether) to remove the solid. The eluate wasconcentrated to yield the dialdehyde 217 (9.6 mg, 98%) as acolorless oil which was not further purified. The dialdehyde 217thus obtained exhibited ir (neat): 3091 (w), 2723 (w), 1718 (vs),2171881638 (w), 1459 (w), 894 (w) cm; 1H nmr (400 MHz): 31.26-1.49 (m,6H, includes 1.36 (s, 3H, Me-12)), 1.71-1.76 (m, 2H), 1.90-2.10 (m,4H), 2.24 (td, 1H, J= 13.0, 7.5 Hz), 2.34 (dq, 1H, J = 13.0, 2.0 Hz),2.51 (dd, 1H, J = 8.5, 3.0 Hz), 2.67 (d, 1H, J = 13.5 Hz), 5.01 (s, 1H,C=CH.2), 5.05 (s, 1H, C=C2), 9.50 (s, 1H, CaO), 9.65 (s, 1H, CaO); MSm/z (% rel. mt.): 232 (M, 1.9), 214 (1.4), 204 (13), 203 (23), 175(23), 147 (34), 145 (23), 135 (25), 134 (25), 133 (62), 131 (28), 119(32), 117 (24), 107 (28), 105 (56), 91 (100). Exact Mass calcd. forC15H200:232.1463; found: 232.1459.Preparation of (±)-J3-Panasinsene (31).The crude dialdehyde 217 (9.6 mg, 41 j.imol), was dissolved in amixture of 250 j.tL of dry diethylene glycol and 150 pL of anhydroushydrazine (151 mg, 4.73 mmol). Argon was passed over the mixturefor -‘5 mm and then the mixture was heated, under an argon atmosphere, at 133-138°C for 1.5 h. The reaction mixture was cooled toroom temperature and then most of the water and excess hydrazine31189were removed via distillation (-65°C, ‘-15 Torr) over a period of 20mm. Crushed potassium hydroxide pellets (48 mg, 0.85 mmol) wereadded and the mixture was heated under an argon atmosphere at‘-190-210°C for 7.5 h. After the reaction mixture had been cooled toroom temperature, ‘-250 tL of water was added and the product wasextracted with pentane (-‘3 mL total). The combined pentaneextracts were dried (anhydrous magnesium sulfate) and most of thesolvent was removed via distillation at atmospheric pressure. Theresidue was then filtered through a short column of silica gel (‘-0.2g, elution with pentane) to remove polar impurities. Most of thepentane was removed from the eluate by distillation (atmosphericpressure) through a short Vigreux column and the last traces ofsolvent were removed by a Kugelrohr distillation (heated up to 80°C)to yield crude (±)-f3-panasinsene (31) (‘-6.4 mg, ‘-76%). The productwas then distilled (80-90°C, 100 Torr) to yield pure (±)-flpanasinsene (31) (3.9 mg, 46%) as a colorless oil. The distilled (±)-f3-panasinsene (31) exhibited ir (neat): 3088 (w), 1636 (m), 1377(m), 1365 (w), 1269 (w), 885 (s) cm; H nmr (400 MHz): 80.75 (s,3H, Me-14), 0.86 (s, 3H, Me-13), 1.07 (s, 3H, Me-12), 1.27 (dm, 1H, J= 12.5 Hz, H-8a), 1.35-1.41 (m, 2H, H-6b and H-9a or H-9b), 1.46 (d,1H, J = 12.5 Hz, H-2a), overlapped with 1.42-1.48 (m, 1H, H-8b),1.57-1.63 (m, 1H, H-9b or H-9a), 1.63-1.68 (m, 1H, H-5a), 1.73 (dd,1H, J = 12.0, 7.0, H-6a), 1.87-1.97 (m, 1H, H-5b), overlapped with1.96 (dd, 1H, J = 12.5, 3.0 Hz, H-2b), 2.00 (br td, 1H, J = 12.5, 4.0 Hz,H-lOa or H-lOb), 2.10 ( br dd, 1H, J = 8.5, 3.0 Hz, H-4b), 2.18 (dm,1H, J = 12.5 Hz, H-lOb or H-ba), 4.80 (d, 1H, J = 1.5 Hz, H-15a), 4.91(dd, 1H, J = 1.5, 1.5 Hz, H-15b); 13C nmr (75 MHz): 8 18.2 (H3-14),19024.72 (H2-5), 24.85 (H3-13), 25.10 (H2-9), 30.53 (.-7 or .-3),30.64 (H3-12), 33.8 (H2-1O), 35.7 (H-2), 36.0 (H2-8), 41.1(H2-6), 45.6 (.-3 or -7), 52.60 (.H-4), 52.77 (-1), 108.3 (H2-15), 152.3 (..-11); In nOe difference experiments (400 MHz),irradiation of the signal at 8 0.75 (s, Me-14) led to enhancement ofthe signal at 1.46 (d, H-2a); irradiation of the signal at 0.86 (s, Me-13) led to enhancement of the signals at 1.46 (d, H-2a), and at 1.63-1.68 (m, H-5a); and irradiation of the signal at 1.07 (s, Me-12) led toenhancement of the signals at 1.96 (dd, H-2b), 2.10 (br dd, H-4b) andat 4.80 (d, H-15a). MS m/z (% rel. mt.): 204 (M, 29), 189 (23), 175(14), 162 (16), 161 (100), 147 (19), 133 (44), 122 (46), 119 (23),107 (47), 105 (40), 91 (38). Exact Mass calcd. for C15H24: 204.1878;found: 204.1875.For the HETCOR and COSY data, see Tables 10 and 11, pp. 144 and147, respectively.191V. REFERENCES1. G. WHITESIDES. Angew. Chem. mt. Ed. EngI. 29, 1209 (1990).2. 0. SEEBACH. Angew. Chem. mt. Ed. EngI. 29, 1320 (1990).3. (a) E. J. COREY. Angew. Chem. mt. Ed. EngI. 30, 455 (1991).(b) E. J. COREY and X.-M. CHENG. The Logic of Chemical Synthesis.John Wiley & Sons, New York. 1989. pp. 1-98.4. M. PEREYRE, J.-P. QUINTARD and A. RAHM. Tin in Organic Synthesis.Butterworths. London. 1987.5. (a) E. PIERS and J. M. CHONG. Can. J. Chem. 66, 1425 (1988).(b) E. PIERS and J. M. CHONG. J. Chem. Soc. Chem. Commun. 934(1983).6. (a) E. PIERS and V. KARUNARATNE. Tetrahedron, 45, 1089 (1989).(b) E. PIERS and V. KARUNARATNE. J. Chem. Soc. Chem. Commun. 935(1983).7. E. PIERS and B. W. A. YEUNG. J. Org. Chem. 49, 4567 (1984).8. B. M. TROST. Acc. Chem. Res. 11, 453 (1978).9. E. PIERS, B. W. A. YEUNG, and S. J. RETTIG. Tetrahedron, 43, 5521(1987).10. (a) E. PIERS and F. F. FLEMING. J. Chem. Soc. Chem. Commun. 1665(1989).(b) F. F. FLEMING. Ph.D. Thesis. The University of BritishColumbia. (1990).11. E. PIERS and J. S. M. WAI. J. Chem. Soc. Chem. Commun. 1245(1988).12. E. PIERS and J. S. M. WAI. J. Chem. Soc. Chem. Commun. 1342(1987).13. K. YOSHIHARA and Y. HIR0sE. Bull. Chem. Soc., Jpn. 48, 2078(1975).19214. L. A. PAQUETTE. Top. Curr. Chem. 79, 41(1979). (see p. 108).15. E. Y. CHEN. J. Org. Chem. 49, 3245 (1984).16. H. IWABUCHI, M. YOSHIKURA, Y. IKAWA and W. KAMISAKO. Chem. Pharm.Bull. 35, 1975 (1987).17. A. G. GONZALEZ, J. M. AGUIAR, J. D. MARTIN and M. NORTE. TetrahedronLett. 2499 (1975).18. H. SETO and H. YONEHARA. J. Antibiot. 33, 92 (1980).19. M. KANEDA, R. TAKAHASHI, Y. IITAKA and S. SHIBATA. Tetrahedron Lett.4609 (1972).20. R. K. BOECKMAN, Jr., D. M. BLUM, S. D. ARTHUR. J. Am. Chem. Soc.101, 5060 (1979).21. N. E. SCHORE and E. G. ROWLEY. J. Am. Chem. Soc. 110, 5224 (1988).22. P. L. PAUSON. In Organometallics in Organic Synthesis. Edited byA. de Meijere and H. torn Dieck. Springer-Verlag, Berlin. 1987.p. 233.23. E. J. COREY, M. C. DE5AI and T. A. ENGLER. J. Am Chem. Soc. 107,4339 (1985).24. (a) E. PIERS and V. KARUNARATNE. J. Chem. Soc. Chem. Commun. 959(1984).(b) E. PIERS and V. KARUNARATNE. Can. J. Chem. 67, 160 (1989).25. A. K. GUPTA, X. Fu, J. P. SNYDER and J. M. COOK. Tetrahedron, 47,3665 (1991) and references therein.26. T. F. W. McKILL0P, J. MARTIN, W. PARKER, J. S. ROBERTS and J. R.STEVENSON. J. Chem. Soc. C, 3375 (1971).27. W. PARKER, R. A. RAPHAEL and J. S. ROBERTS. Tetrahedron Left. 2313(1965).28. P. U. UNSCHULD. Medicine in China (A History of Pharmaceutics).University of California Press, Berkley. 1986. pp. 22, 64, 212,234-235.19329. J. P. Hou. Comparative Medicine East and West, 5, 123 (1977).30. J. E. MCMURRY and W. CH0Y. Tetrahedron Lett. 21, 2477 (1980).31. C. R. JOHNSON and N. A. MEANWELL. J. Am. Chem. Soc. 103, 7667(1981).32. (a) J. MARCH. Advanced Organic Chemistry. Third Edition. JohnWiley & Sons, New York. 1985. pp. 1101-2.(b) Esters may also be converted to the correspondingaldehydes or alcohols which may deoxygenated directly orvia derivatives; see Ibid. pp. 1096-8 and pp. 392-393, 395,respectively, and references therein.33. For some examples of copper-catalyzed conjugate additions toenones to give cis-fused diquinanes see:(a) L. A. PAQUETTE. Top. Curr. Chem. 119, 1 (1984). (pp. 93, 121,129).(b) reference 24.34. S.-j. CHANG, D. MCNALLY, S. SHARY-TEHRANY, Sr. M. J. HICKEY and R. H.BoyD. J. Am. Chem. Soc. 92, 3109 (1970).35. E. J. COREY and N. W. BOAZ. Tetrahedron Left. 26, 6015 (1985).36. (a) reference 32. pp. 421-422, 331.(b) B.-T. GROBEL and D. SEEBACH. Synthesis, 357 (1977).(c) 0. W. LEVER, Jr. Tetrahedron, 32, 1943 (1976).37. (a) K. OGURA. Pure and AppI. Chem. 59, 1033 (1987).(b) K. OGURA, K. OHTSUKI, M. NAKAMURA, N. YAHATA, K. TAKAHASHI and H.IIDA. Tetrahedron Lett. 26, 2455 (1985).(c) K. OGuRA, N. YAHATA, K. HASHIZUME, K. TSUYAMA, K. TAKAHASHI and H.IIDA. Chem. Left. 767 (1983).38. E. PIERS and N. Moss. Tetrahedron Left. 26, 2735 (1985).39. (a) N. E. SCHORE. Chem Rev. 88, 1081 (1988).(b) P. L. PAUSON. Tetrahedron, 41, 5855 (1985).(c) P. L. PAUSON and I. U. KHAND. Ann. N. Y. Acad. Sci. 295, 2(1977).40. N. E. SCHORE and M. C. CROUDACE. J. Org. Chem. 46, 5436 (1981).19441. I. U. KHAND and P. L. PAUSON. J. Chem. Soc. Chem. Commun. 379(1974).42. S. SHAMBAYATI, W. E. CROWE and S. L. SCHREIBER. Tetrahedron Lett.31, 5289 (1990).43. M. C. CROUDACE and N. E. SCHORE. J. Org. Chem. 46, 5357 (1981).For further information about the infrared absorptions ofcobalt complexes, see: K. NAKAMOTO. Infrared and Raman Spectraof Inorganic and Coordination Compounds. Fourth Edition.Wiley-Interscience, New York. 1986. pp. 291-308.44. M. POcH, E. VALENTI, A. MOYANO, M. A. PERICAS, J. CASTRO, A. DENICOLAand A. E. GREENE. Tetrahedron Lett. 31, 7505 (1990).45. L. DAALMAN, R. F. NEWTON, P. L. PAUSON and A. WADSWORTH. J. Chem.Res. (S), 346 (1984).46. P. MAGNUS, L. M. PRINcIPE and M. J. SLATER. J. Org. Chem. 52, 1483(1987).47. P. MAGNUS, C. EXON and P. ALBAUGH-ROBERTSON. Tetrahedron, 41,5861 (1985).48. D. F. DETAR and N. P. LUTHRA. J. Am. Chem. Soc. 102, 4505 (1980).49. C. EXON and P. MAGNUS. J. Am. Chem. Soc. 105, 2477 (1983).50. (a) P. MAGNUS and L. M. PRINcIPE. Tetrahedron Left. 26, 4851(1985).(b) P. MAGNUS and D. P. BECKER. J. Am. Chem. Soc. 109, 7495(1987).51. (a) B. E. LA BELLE, M. J. KNUDSEN, M. M. OUvI5TEAD, H. HOPE, M. D. YANUCKand N. E. SCHORE. J. Org. Chem. 50, 5215 (1985).(b) S. E. MACWHORTER, V. SAMPATH, M. M. OLMSTEAD and N. E. SCHORE. J.Org. Chem. 53, 203 (1988).52. V. RAUTENSTRAUCH, P. MEGARD, J. CONESA and W. KUSTER. Angew. Chem.tnt. Ed. EngI. 29, 1413 (1990) and references therein.53. I. U. KHAND, G. R. KNOX, P. L. PAUSON, W. E. WATTS and M. I. FOREMAN. J.Chem. Soc. Perkin Trans. I, 977 (1973).19554. I. U. KHAND, G. R. KNox, P. L. PAUSON and W. E. WATTS. J. Chem. Soc.Chem. Commun. 36 (1971).55. (a) M. E. PRICE and N. E. SCHORE. Tetrahedron Lett. 30, 5865(1989).(b) D. C. BILLINGTON, P. BLDON, I. M. HELPS, P. L. PAUSON, W. THOMSONand D. WILLISON. J. Chem. Res. (S), 326 (1988).56. M. E. KRAFFT. J. Am. Chem. Soc. 110, 968 (1988).57. (a) D. SEEBACH and M. KOLB. Chem. md. (London), 687 (1974).(b) D. SEEBACH. Angew. Chem. tnt. Ed. EngI. 18, 239 (1979).58. Reference 32 (a) pp. 418-419 and references therein.59. (a) D. SEEBACH and E. J. COREY. J. Org. Chem. 40, 231 (1975).(b) D. SEEBACH. Synthesis, 17 (1969).60. C. WESTERLUND. Tetrahedron Lett. 23, 4835 (1982).61. R. M. SILVERSTEIN, G. C. BASSLER and T. C. MORRILL. SpectrometricIdentification of Organic Compounds. Fourth Edition. John Wiley& Sons, New York. 1981.62. G. A. WILEY, R. L. HERSHKOWITZ, B. M. REIN and B. C. CHUNG. J. Am.Chem. Soc. 86, 964 (1964).63. For two review articles see:(a) F. THERON, M. VERNY and R. VESSIERE. In The Cemistry of theCarbon-Carbon Triple Bond. Part 1. Edited by S. Patai.Wiley-lnterscience, Chichester. 1978. pp. 410-413.(b) D. R. TAYLOR. Chem. Rev. 67, 317 (1967).64. (a) C. N. BUSH and D. E. APPLEQUIST. J. Org. Chem. 42, 1076 (1977).(b) K. FUJIWARA, H. SAKAI and M. HIRAMA. J. Org. Chem. 56, 1688(1991).65. (a) T. L. JACOBS and P. PREMPREE. J. Am. Chem. Soc. 89, 6177(1967).(b) J. GORE and M. L. ROUMESTANT. Tetrahedron Lett. 891 (1970).(c) R. GELIN, S. GELIN and M. ALBRAND. Bull. Soc. Chim., Fr. 4146(1971).196(d) J. IOSSIPHIDES, C. TROYANOWSKY and A. TSAMANTAKIS. C. R. Acad.Sci. C, 272, 1724 (1971).66. (a) E. J. COREY and C. ROCKER. Tetrahedron Left. 23, 719 (1982).(b) J. A. MARSHALL, B. S. DEHOFF and S. L. CROOKS. Tetrahedron Lett.28, 527 (1987).67. (a) G. SCHILL and P. R. JONES. Synthesis, 117 (1974).(b) K. OGURA, Y. ITO and G.-i. TSUCHIHASHI. Bull. Chem. Soc. Jpn. 52,2013 (1979).(c) K. OGURA and G.-i. TSUCHIHASHI. Tetrahedron Lett. 3151 (1971).68. (a) J. E. RICHMAN, J. L. HERRMANN and R. H. SCHLESSINGER. TetrahedronLett. 3267 (1973).(b) J. L. HERRMANN, J. E. RICHMAN, P. J. WEPPLO and R. H. SCHLESSINGER.Tetrahedron Lett. 4707 (1973).69. K. OGURA, M. Suzuki, J.-i. WATANABE, M. YAMASHITA, H. IIDA and G.-i.TSUCHIHASHI. Chem. Left. 813 (1982) and references therein.70. K. OGURA, K. OHTSUKI, K. TAKAHASHI and H. IIDA. Chem. Lett. 1597(1986).71. (a) K. OGURA, T. TSURUDA, K. TAKAHASHI and H. IIDA. TetrahedronLett. 27, 3665 (1986).(b) K. OGURA, N. YAHATA, M. MIN0GucHI, K. OHTSUKI, K. TAKAHASHI and H.IIDA. J. Org. Chem. 51, 508 (1986).72. K. OGURA, T. UCHIDA, T. TSURUDA and K. TAKAHASHI. Tetrahedron Lett.28, 5703 (1987).73. (a) Y. MURATA, K. INOMATA, H. KINOSHITA and H. KOTAKE. Bull. Chem.Soc. Jpn. 56, 2539 (1983).(b) K. OGURA, T. IIHAMA, S. KIuCHI, T. KAJIKI, 0. KOSHIKAWA, K.TAKAHASHI and H. IIDA. J. Org. Chem. 51, 700 (1986).74. K. OGURA, N. YAHATA, J.-i. WATANABE, K. TAKAHASHI and H. IIDA. Bull.Chem. Soc. Jpn. 56, 3543 (1983).75. L. A. PAQUETTE and Y.-K. HAN. J. Am. Chem. Soc. 103, 1835 (1981).76. Y. HORIGUCHI, S. MATSUZAWA, E. NAKAMURA and I. KuwAJIMA.Tetrahedron Lett. 27, 4025 (1986).19777. E. NAKAMURA, S. MATSUZAWA, Y. H0RIGucHI and I. KUWAJIMA.Tetrahedron Lett. 27, 4029 (1986).78. (a) H. 0. HOUSE and J. M. WILKINs, J. Org. Chem. 43, 2443 (1978).(b) G. HALLNEMO and C. ULLENIUS. Tetrahedron, 39, 1621 (1983).(c) S. H. BERTZ and G. DABBAGH. Tetrahedron, 45, 425 (1989).79. G. H. POSNER. Chem. Rev. 86, 831 (1986).80. (a) G. KUBIAK, J. M. CooK and U. WEISS. Tetrahedron Lett. 26, 2163(1985).(b) G. KUBIAK, J. M. Cook and U. WEiss. J. Org. Chem. 49, 561(1984).(c) K. AVASTHI, M. N. DESHPANDE, W.-C. HAN, J. M. Cook and U. WEiss.Tetrahedron Lett. 22, 3475 (1981).81. P. ASHKENAZI, J. KETrENRING, S. MIGDAL, A. L. GUTMAN and D. GINSBURG.Helv. Chim. Acta, 68, 2033 (1985).82. (a) S. H. BERTZ, G. LANNOVE and J. M. COOK. Tetrahedron Lett. 26,4695 (1985).(b) K. SAMBASIVARAO, G. KUBIAK, G. LANNOYE and J. M. Cook. J. Org.Chem. 53, 5173 (1988).(C) A. K. GUPTA, J. M. COOK and U. WEISS. Tetrahedron Lett. 29,2535 (1988).(d) X. Eu and J. M. COOK. Tetrahedron Lett. 31, 3409 (1990).83. (a) R. MITSCHKA, J. OEHLDRICH, K. TAKAHASHI, J. M. Cook, U. WEISS and J.V. SILVERTON. Tetrahedron, 37, 4521 (1981).(b) S. YANG-LAN, M. MUELLER-JOHNSON, J. OEHLDRICH, D. WICHMAN, J. M.Cook and U. WEISS. J. Org. Chem. 41, 4053 (1976).(C) T. HUDLICKY and J. D. PRICE. Chem. Rev. 89, 1467 (1989).84. S. H. BERTZ, J. M. Cook, A. GAwI5H and U. WEISS. Org. Syn. 64, 27(1986).85. N. Moss. Ph.D. Thesis. The University of British Columbia. 1985.86. For some examples see:(a) D. LIonA, C. S. BARNUM and M. SAINDANE. J. Org. Chem. 46, 4301(1981) and references therein.(b) P. A. GRIECO, C. S. PoGoNowskl and S. BURKE. J. Org. Chem. 40,542 (1975) and references therein.198(C) D. L. J. CLIVE. Tetrahedron, 34, 1049 (1978) and referencestherein.87. (a) See reference 78a and references therein.(b) For an example, see also, S. J. DANISHEFSKY and N. MANTLO. J.Am. Chem. Soc. 110, 8129 (1988).88. L. RUEST, G. BL0uIN and P. DESLONGCHAMPS. Syn. Commun. 6, 169(1976).89. (a) J. WRIGHT, G. J. DRTINA, R. A. ROBERTS and L. A. PAQUETTE. J. Am.Chem. Soc. 110, 5806 (1988).(b) L. A. PAQUETTE and Y.-K. HAN. J. Am. Chem. Soc. 103, 1831(1981).90. H. J. REICH, J. M. RENGA and I. L. REICH. J. Am. Chem. Soc. 97, 5434(1975).91. K. B. SHARPLESS, M. W. YOUNG and R. F. LAUER. Tetrahedron Lett.1979 (1973).92. G. ZIMA, C. BARNUM and D. LIOTTA. J. Org. Chem. 45, 2736 (1980)and references therein.93. N. E. SCHORE and M. J. KNUDSEN. J. Org. Chem. 52, 569 (1987).(The spectrocopic data must be interpreted with care).94. A. BARCO, S. BENETTI and G. P. P0LLINI. Synthesis, 316 (1973).95. R. J. K. TAYLOR. Synthesis, 364 (1985) and references therein.96. Based on a suggestion made by Dr. C. K. Lu.97. A. E. DEROME. Modern NMR Techniques for Chemistry Research.Pergamon Press, Oxford. 1987.98. S. KRISHNAMURTHY and H. C. BROWN. J. Org. Chem. 41, 3064 (1976).99. M. J. ROBINS, J. S. WILSON and F. HANSSKE. J. Am. Chem. Soc. 105,4059 (1983).100. H.-J. LIu, S. P. LEE and W. H. CHAN. Can. J. Chem. 55, 3797 (1977).1991 01. (a) T. NAKATA and T. OIsHI. Tetrahedron Lett. 21, 1641 (1980).(b) W. J. GENSLER, F. JOHNSON and A. D. B. SLOAN. J. Am. Chem. Soc.82, 6074 (1960).102. H. C. BROWN and S. KRISHNAMURTHY. J. Am. Chem. Soc. 94, 7159(1972).103. (a) A. L. GEMAL and J.-L. LUCHE. J. Am. Chem. Soà. 103, 5454(1981).(b) H. B. KAGAN and J. L. NAMY. Tetrahedron, 42, 6573 (1986).104. D. REDMORE and C. D. GUTSCHE. In Advances in Alicyclic ChemistryVol. 3. Edited by H. Hart and G. J. Karabatsos. Academic Press,New York. 1971. P. 1.105. H. MEIER and K.-P. ZELLER. Angew. Chem. mt. Ed. EngI. 14, 32(1975).1 06. M. REGITZ and G. MAAS. In Diazo Compounds, Properties andSynthesis. Academic Press, Orlando, Florida. 1986. pp. 372-383 and references therein.107. (a) Y. UEDA, G. ROBERGE and V. VINET. Can. J. Chem. 62, 2936(1984).(b) W. VON E. DOERING and C. H. DEPUY. J. Am. Chem. Soc. 75, 5955(1953).108. M. REGITZ. Synthesis, 351 (1972).109. J. B. HENDRICKSON and W. A. WOLF. J. Org. Chem. 33, 3610 (1968).110. L. LOMBARDO and L. N. MANDER. Synthesis, 368 (1980).111. D. F. TABER, R. E. RUCKLE, JR. and M. J. HENNESSY. J. Org. Chem. 51,4077 (1986) and references therein.112. M. REGITZ. Angew. Chem. mt. Ed. Engl. 6, 733 (1967).113. W. G. DAUBEN and D. M. WALKER. Tetrahedron Lett. 23, 711(1982).114. K. B. WIBERG, B. L. FURTEK and L. K. OLLI. J. Am. Chem. Soc. 101,7675 (1979).200115. V. K. RAO and M. NAGARAJAN. J. Org. Chem. 54, 5678 (1989).116. (a) M. REGITZ and J. RUTER. Chem. Ber. 101, 1263 (1968).(b) K. B. WIBERG, L. K. OLLI, N. GOLEMBESKI and R. D. ADAMS. J. Am.Chem. Soc. 102, 7467 (1980).117. (a) R. L. DANHEISER, R. F. MILLER, R. G. BRisBols and S. Z. PARK. J. Org.Chem. 55, 1959 (1990).(b) J. H. BOVER, C. H. MACK, N. GOEBEL and L. R. MORGAN, JR. J. Org.Chem. 23, 1051 (1958).118. G. STORK and R. P. SZAJEWSKI. J. Am. Chem. Soc. 96, 5787 (1974).119. T. N. WHEELER and J. MEINWALD. Org. Syn. 52, 53(1972).120. S. WOLFF and W. C. AGOSTA. J. Chem. Soc. Chem. Commun. 118(1981) and references therein.121. See reference 32 (a) pp. 256-259, 298-320.122. W. HARTWIG. Tetrahedron, 39, 2609 (1983).123. (a) D. P. CURRAN. Synthesis, 417, (1988).(b) D. P. CURRAN. Synthesis, 489, (1988).(c) M. RAMAIAH. Tetrahedron, 43, 3541 (1987).(d) C. WALLING. Tetrahedron, 41, 3887 (1985).124. D. H. R. BARTON, W. B. MOTHERWELL and A. STANGE. Synthesis, 743(1981).125. D. H. R. BARTON and S. W. MCCOMBIE. J. Chem. Soc. Perkin Trans. I,1574 (1975).126. A couple of examples of ways to get around the separationproblems are found in the following:(a) J. LIGHT and R. BRESLOW. Tetrahedron Left. 31, 2957 (1990).(b) D. N. HARPP, M. GINGRAS, T. AIDA and T. H. CHAN. Synthesis, 1122(1987).127. J. N. KIRWAN, B. P. ROBERTS and C. R. WILLIS. Tetrahedron Left. 31,5093 (1990).201128. B. GIESE, B. KOPPING and C. CHATGILIALOGLU. Tetrahedron Lett. 30,681 (1989).129. D. H. R. BARTON, D. 0. JANG and J. C. JASZBERENYI. Tetrahedron Lett.31, 4681 (1990).130. R. E. IRELAND, D. C. MUCHMORE and U. HENGARTNER. J. Am. Chem. Soc.94, 5098 (1972).131. J. S. M. WA!. Ph. D. Thesis. The University of British Columbia.1988.132. Dr. J. S. M. WAI. personal communication.133. (a) H. DESHAYES, J.-P. PETE, C. PORTELLA and D. SCHOLLER. J. Chem.Soc. Chem. Commun. 439 (1975).(b) J.-P. PETE and C. PORTELLA. Synthesis, 774 (1977).(c) H. DESHAYES, J.-P. PETE and C. PORTELLA. Tetrahedron Lett. 2019(1976).134. P. M. COLLINS and V. R. Z. MUNASINGHE. J. Chem. Soc. Chem. Commun.927 (1977).135. (a) See reference 32 (b) pp. 1096-1098 and references therein.(b) G. W. KABALKA and J. H. CHANDLER. Syn. Commun. 9, 275 (1979)and references therein.136. J. Y. ROBERGE. personal communication.1 37. W. CARRUTHERS. Some Modern Methods of Organic Synthesis.Third Edition. Cambridge University Press, Cambridge. 1986. p.483.138. A. J. MANCUSO, S.-L. HUANG and D. SWERN. J. Org. Chem. 43, 2480(1978).139. W. C. STILL, M. KHAN and A. MITRA. J. Org. Chem. 43, 2923 (1978).140. D. D. PERRIN, W. L. F. ARMAREGO and D. R. PERRIN. Purification ofLaboratory Chemicals. Second Edition. Pergamon Press, Oxford.1980.2021 41. L. F. FIESER and M. FIESER. Reagents for Organic Synthesis. Vol. 1.John Wiley and Sons, New York. 1967. p. 434.142. J.-M. CONIA. Soc. Chim. Fr., Mem. 537 (1950).143. W. G. KOFRON and L. M. BACLAWSKI. J. Org. Chem. 41, 1879 (1976).144. P. G. M. WUTS. Syn. Commun. 11, 139 (1981).


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items