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

Download

Media
831-ubc_1992_spring_story_betty_anne.pdf [ 2.93MB ]
Metadata
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
831-1.0061756-fulltext.txt
Citation
831-1.0061756.ris

Full Text

THE TOTAL SYNTHESIS OF (±)-fi-PANASINSENE by BTIV-ANNE STORY  B.Sc., The University of Toronto, 1984 M.Sc., The University of Toronto, 1986 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES  (Department of Chemistry) We accept this thesis as conforming to the required standard  ThE UNIVERSITY OF BRITISH COLUMBIA October 1991 ©  Betty-Anne Story, 1991  In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department  or  by  his  or  her  representatives.  It  is  understood  that  copying  or  publication of this thesis for financial gain shall not be allowed without my written permission.  Department of  C/1EM(ST/  The University of British Columbia Vancouver, Canada  Date  DE-6 (2/88)  ?j’t1  /99/  11  ABSTRACT  This thesis describes a total synthesis of the sesquiterpenoid (±)-fl-panasinsene (31). Two different routes for the synthesis of a bicyclic enone of general structure 74 were investigated.  An unsuccessful attempt to  generate a synthetically useful enone 74 employed the Pauson Khand cyclization of an enyne 75 (R= Me; XX= S(CH S or MeS, p 3 ) 2 SO 4 H 6 MeC ) 2 . A second approach, which was based on the Weiss-Cook condensation of glyoxal (44) with dimethyl 3-oxoglutarate (45) led to the production of the dione 43, which was converted, via several steps, into the enone 159. The enone 1 59  was subjected to a methylenecyclohexane  annulation sequence.  Thus, copper (1)-catalyzed conjugate addition  of the Grignard reagent 7 to 159,  followed  by  intramolecular  alkylation of the resultant chloro keto ester, provided the tricyclic intermediate 171.  Sequential reduction of the keto function in 171,  deoxygenation of the resultant hydroxyl function, and hydrolysis of the ketal moiety gave rise to the keto ester 182. to  a photochemical  Wolff ring  contraction  Subjection of 182 reaction  sequence  provided a mixture of the diesters 200 and 201. Alkylation of the mixture of 200 and 201  with  methyl iodide, followed  by a  reduction-oxidation sequence, gave the dialdehyde 217.  Wolff  Kishner  reduction  of  21 7  resulted  in  the  simultaneous  deoxygenation of both of the carbonyl groups and successfully completed the synthesis of (±)--panasinsene (31).  111  R Me  Me  —  74  75  31 C 2 MeO  ::  H  H  C 2 MeO  44  45  43  BrMg 159  7  C4 2 MeO  H 171  182  C4 2 MeO  HO 200 and 201  217  iv  TABLE OF CONTENTS Page ABSTRACT  ii  TABLE OF CONTENTS  iv  LISTOFTABLES  vii  LIST OF FIGURES  viii  ABBREVIATIONS  x  ACKNOWLEDGEMENTS  xiii  I.  INTRODUCTION  1  1.1  The Rationale  1  1.2  The Problem  8  1 .3  Angularly Fused Terpenoids  9  1 .4  Isolation and Structural Elucidation of /3Panasinsene (31)  15  1.5  Previous Syntheses of f3-Panasinsene (31)  18  1.5.1  McMurry and Choy’s Synthesis of Racemic a and 13-Panasinsene (54 and 31)  18  Johnson and Meanwell’s Synthesis of (-)-f3Panasinsene and Its Enantiomer ((-)- and (+)-31)  20  1 .5.2  II.  DISCUSSION  22  Total Synthesis of (±)-J3-Panasinsene (31) 2.1  Retrosynthetic Analysis  22  V  2.2  An Approach to the Synthesis of (±)-J3Panasinsene (31) via the Use of the Pauson Khand Reaction  26  2.2.1  Background  26  2.2.2  Approach to the Synthesis of (±)-J3Panasinsene (31)  32  2.2.2.1  The 1,3-Dithianyl Function at C-7  35  2.2.2.2  The Methylthio-p-toluenesulfonyl Function at C-7  44  The Synthesis of (±)-/3-Panasinsene (31) via the Weiss-Cook Condensation Reaction  56  2.3.1  Background  56  2.3.2  Application of the Weiss-Cook Condensation Reaction to the Synthesis of (±)-J3Panasinsene (31)  60  2.3.2.1  Preparation of an Enone 74  60  2.3.2.2  Methylenecyclohexane Annulation on the Enonel59  71  Preparation of a Substrate (Intermediate 72) for Ring Contraction  82  Ring Contraction to Give a 4-5-6 Tricyclic Carbon Skeleton  1 00  Preparation of (±)-/-Panasinsene (31) via the Diacetate 213  11 7  Preparation of (±)-J3-Panasinsene (31) via a Woiff-Kishner Reduction  135  2.3  2.3.2.3 2.3.2.4  2.3.2.5  2.3.2.6 III.  CONCLUSION  1 49  vi IV.  V.  EXPERIMENTAL  151  4.1  General  151  4.2  Experimental Procedures for the Synthesis of (±)-13-Panasinsene (31) via the Weiss-Cook Condensation Approach  1 56  REFERENCES  191  vii  LIST OF TABLES Table 1  Page  The 75 MHz HETCOR Data for the Keto Ester Ketal 171  78  The 400 MHz COSY Data for the Keto Ester Ketal 171  81  3  The 400 MHz COSY Data for the Keto Ester 182  99  4  The 125 MHz HETCOR Data for the Diester 200  111  5  The 400 MHz COSY Data for the Diester 200  11 6  6  The 400 MHz COSY Data for the Diester 202  1 23  7  The 400 MHz COSY Data for the Diacetate 213  133  8  The 400 MHz NOESY Data for the Diacetate 213  1 34  9  A Comparison of the Spectral Data for Authentic and Synthetic J.3-Panasinsene (31)  1 39  The 125 MHz HETCOR Data for Synthetic (±)q3-Panasinsene (31)  144  The 400 MHz COSY Data for Synthetic (±)-fl-Panasinsene (31)  147  2  10  11  vii’  LIST of FIGURES Figure 1  Page  The 75 MHz broad band decoupled 13 C nmr spectrum of the enone diastereomer A  52  The 300 MHz 1H nmr spectrum of the keto ester ketal 171  75  The 75 MHz HETCOR spectrum of the keto ester ketal 171  77  The 400 MHz COSY spectrum of the keto ester ketal 171  80  5  The 300 MHz 1 H nmr spectrum of the keto ester 182  96  6  The 400 MHz COSY spectrum of the keto ester 182  98  7  The 125 MHz HETCOR spectrum of the diester 200  110  8  The 400 MHz 1 H nmr spectrum of the diester 200  112  9  The 400 MHz COSY spectrum of the diester 200  11 5  10  The 400 MHz 1 H nmr spectrum of the diester 202  119  1 1  The 400 MHz COSY spectrum of the diester 202  1 22  12  The 400 MHz ‘H nmr spectrum of the diacetate 213  130  13  The 400 MHz COSY spectrum of the diacetate 213  132  14  The 75 MHz broad band decoupled 13 C nmr spectrum of synthetic (±)-J3-panasinsene (31)  141  The 400 MHz 1 H nmr spectrum of synthetic (±)-f3-panasinsene (31)  142  2  3 4  15  ix  16  17  The 125 MHz HETCOR spectrum of synthetic (±)-fl-panasinsene (31)  143  The 400 MHz COSY spectrum of synthetic (±)-fl-panasinsene (31)  146  x  LIST OF ABBREVIATIONS Ac  Acetyl  AIBN  2,2’-azobisisobutyro nitrile  APT  attached proton test  aq  aqueous  br  broad  b.p.  boiling point  n-Bu  normal-butyl  t-Bu  tertiary-butyl  cat  catalyst,  cOSY  correlation spectroscopy  d  doublet  CR3  diethylene glycol  DMAP  4-(N, N-dimethylamino)pyridine  EXvE  1 ,2-dimethoxyethane  DMSO  dimethyl sulfoxide  E-  electrophile  equiv  equivalents  Et  ethyl  g  grams  glc  gas-liquid chromatography  h  hour(s)  HETOR  heteronuclear correlation spectroscopy  HMPA  hexamethylphosphoramide  Hz  hertz  i r  infrared  catalytic  xi  LAH  lithium aluminum hydride  LDA  lithium diisopropylamide  lit,  literature  m  multiplet  M  molar  Me  methyl  mg  milligram(s)  MHz  megahertz  mm  minute(s)  mmol  millimole(s)  mol  mole(s)  m.p.  melting point  Ms  methanesulfonyl  MS  mass spectrum (low resolution)  m/z  mass to charge ratio  N  normal  1’lvD  N-methylmorpholine N-oxide  nmr  nuclear magnetic resonance  nOe  nuclear Overhauser enhancement  NOESY  nuclear Overhauser enhancement spectroscopy  Nu:  nucleophile  p  page  PC  pyridinium chlorochromate  Ph  phenyl  ppm  parts per million  PTC  phenoxythiocarbonyl  Pyr  pyridine  xli  q  quartet  quant  quantitative  rel. i.  relative  rt  room temperature  s  singlet (nmr); strong (ir)  t  triplet  TBDMS  tertiary- b u t y I d I m e t h yl s ii y I  Tf  trifluoromethanesulfonyl  THE  tetrahydrofuran  THP  tetrahydropyranyl  TIPS  triisopropylsilyl  tIc  thin layer chromatography  TMEDA  N,N,N’, N’-tetramethylethylenediamine  TMS  tetramethylsilane  TMS-  trimethylsilyl  p-Tol  para-tolyl  p-Ts  para-toluenesulfonyl  v  very  w  weak  intensity  xlii  ACKNOWLEDGEMENTS This thesis would have been impossible were it not for the input, in various different ways, of many people. I am deeply indebted to my research supervisor, Professor Edward Piers, for his guidance throughout my research. His commitment to excellence is appreciated, as is his patience with my efforts in the laboratory. The members of the Piers research group during my sojourn here provided enriching friendships and many fruitful discussions. Veljko Dragojlovic’s loan of his computer greatly facilitated the writing of this thesis. Many thanks to Ms. Johanne Renaud, Ms. Renata Oballa, Mr. Todd Schindeler, Dr. Guy Plourde and Mr. Jacques Roberge for proof-reading this manuscript. The able technical assistance of the staff of the nmr, mass spectrometry and elemental analysis facilities was critical to the success of this research. Professor Scheffer’s research group kindly provided 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 the good and bad times and reminded me of the One who is the source of all knowledge. I have appreciated the financial independence and freedom to do research provided by an NSERC postgraduate scholarship.  1  I. INTRODUCTION  1.1. The Rationale.  The  synthesis of biologically useful, theoretióally interesting,  and stereochemically or structurally challenging  molecules  has  provided the basis for much of the research in synthetic organic chemistry.  Presently, the trend  “substances and  seems to be towards the study of  reactions relevant to life” 1  which means that  inhibitors for important enzymes or receptors 2 are becoming key target molecules.  The enantiomeric purity of a biologically active  product that has been made synthetically in a laboratory can have major  implications  in  biological  systems  enantiomer exhibits undesirable activity.  especially  if  one  Though in recent years  there has been a greater emphasis placed on the synthesis of enantiomerically pure compounds as compared to the synthesis of racemic  mixtures,  informative.  the  synthesis  of  racemic  mixtures  is  still  However, the successful synthesis of any molecule, in  a racemic or an enantiomerically pure form, depends largely on the chemist’s ability to analyze the target in a logical retrosynthetic* 3 manner, preparation.  and to develop a feasible synthetic route for its Despite careful retrosynthetic analysis, the actual  implementation of any route may be unsuccessful due either to the *  Retrosynthetic (or antithetic) analysis is a technique that has been developed in order  to transform the structure of the synthetic target into a logical sequence of progressively simpler structures which finally leads to simple or commercially available chemicals. Each step in the retrosynthetic direction (a transform) corresponds to a chemical reaction in the synthetic direction. 3  2  failure of a particular reaction or due to the lack of procedures for carrying out the desired transformation.  Consequently, alternative  routes must be envisaged in advance, particularly for reactions that may  be  synthetically  challenging.  In  where  cases  the  implementation of a route has failed, other options must be explored or  new  methods  must  be found  in  order to  circumvent the  obstruction. One of the currently fruitful areas of research with regard to developing new reactions is organometallic chemistry.  From an  organic chemist’s point of view, an important test of the utility of a particular organometallic reagent is its applicability to the solution of a given synthetic organic chemical problem. The use of organotin reagents in organic synthesis has been explored by many research groups including our own. 4 was first reported that to—substituted regioselective reaction with  1-alkynes  In 1983, it  (1) undergo a  (trimethylstannyl)copper (1)-dimethyl  sulfide 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 groups  tolerated in the reaction are halides, hydroxyls, and trialkylsilyl or tetrahydropyranyl ethers. that the  Subsequent research has demonstrated 2-trimethylstannyl-1-alkene reagents 46 and 57 can by  transmetallation  with  methyllithium  be  converted  into  the  corresponding lithio species, which can then be transformed into the corresponding Grignard or organocopper reagents. chloro-2-trirnethylstannyf-1-pentene  (5)  For example, 5-  was  reacted  with  methyllithium (THE, -78°C, 15 mm) to make the vinyllithium species  3  H  (CH)X 2 n  H  2 S 3 Me nCuSMe (2) (2 equiv) MeOH (60 equiv), THE, -63°C (12h)  (CH)X 2 -  1  —  3 SnMe  H  1 n=1-4 X= CI, OH, OTHP, OTBDMS  ===(\—_ ci  ===(“ci SnMe 3 4  3 SnMe  5  6. Treatment of 6 with anhydrous magnesium bromide produced the Grignard  reagent 7, which was found to undergo a copper (1)-  catalyzed  , 0.25 equiv) conjugate addition reaction to 2 (CuBrSMe  enones of general structure 8.  In the cases of enones with a  trisubstituted double bond, (e.g. 8, R or R’=Me), boron trifluoride etherate was added as an additional catalyst to efficiency of the conjugate addition.  improve the  The chloro ketones 9 were  cyclized with potassium hydride (THE, room temperature, 2 hours) to give exclusively the cis-fused where  subsequent  bicyclic products (1 0) for cases  equilibration  was  impossible  (R=Me).  If  equilibration of the bicyclic product was possible (R=H), then varying amounts of the trans-fused bicyclic ketone (11) were also obtained as the minor component of the product mixture (Scheme I 1).  The overall  utilization conjunctive  of  the  process described above demonstrates the chloro  vinylstannane  as  a  bifunctional  reagent* in which the two reactive sites have been  selectively, sequentially deployed. *  5  In these reactions, 5 serves as  A bifunctional conjunctive reagent is a reagent with two reactive sites which is incorporated in whole or in part into a substrate molecule to increase its structural  6 complexity.  4  , acceptor 2 -synth on 5 the synthetic equivalent of the 1 -pentene donor , 2 (d 5 synthon) a 12.  The overall result of the sequence of reactions,  in which 8 is converted into 10 and/or 11, is the annulation of a methylenecyclohexane unit onto a cyclic enone. CI  MeLi, THE -78°C (15mm)  .,Li  5  , THF, 2 MgBr  .MgBr  7  6  1) 2 CuBrSMe (0.25 equiv) 2) R R’> OEt 3 (BF ) 3) 2 H  0 R  11 10  9  Scheme  I-i  12  The methylenecyclohexane annulation process is an important tool in natural products synthesis because the methylenecyclohexane  5  ring  and  derivatives thereof  (part  structures  13, 14, 1 6-1 9,  Scheme 1-2) are common in terpenoid natural products.  Thus, the  olefinic function in the methylenecyclohexane annulated product (13) may be hydrogenated to give the corresponding methyl group present in 1 4 or it may be cyclopropanated to give 1 5.  The  cyclopropyl ring in 15 may then be hydrogenolyzed to give the gem dimethyl group found in 16.  An  acid-catalyzed double bond  isomerization in 13 would provide 17.  Alternatively, the methylene  group may be cleaved to give the ketone 18, or hydroxylated to give 19.  14  13  15  16  /  17  19 Scheme  1-2  The annulation procedure has been used in the syntheses of natural  products  with  the  bicyclic  9 axane  and  clerodane  skeletons 1 ’ 10 1 and has also been employed in the total synthesis of 2 (Scheme 1-3). the sesterterpenoid (±)-paIauolide  Each of these  6  syntheses involved the annulation of the methylenecyclohexane unit onto a cyclic enone with a trisubstituted double bond. synthesis  of  the  cyclopentenone vinylstannane  axane  skeleton,  conjugate  Thus, in the  addition  to  the  20 of the Grignard reagent 7 (derived from the 5) was catalyzed by copper (I)  sulfide complex and boron trifluoride etherate.  bromide-dimethyl The resultant chloro  ketone was cyclized as described earlier to give only the cis-fused bicyclic ketone 21, which was then transformed via a series of reactions into (±)-axamide-1  (22) and (±)-axisonitrile-1 (23). In a  similar manner, conjugate addition of the Grignard reagent 7 to the cyclohexenone  24, followed by cyclization of the resultant  chloroketone, gave a mixture of the cis- and trans-fused ketones in which the cis-fused product predominated.  bicyclic  The mixture  was converted by equilibration (KH/EtOH) to another mixture in which 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 for  the syntheses of the clerodane (±)-isolinaridiol diacetate (29) and the  sesterterpenoid  (±)-palauolide  (30), but a similar conjugate  addition! cyclization sequence led to the formation of a mixture of the cis- and trans-fused bicyclic ketones in which the cis-fused compound predominated.  Equilibration (t-BuOK/t-BuOH) of the  mixture led to the formation of another mixture in which the trans fused product (28) was the major component.  The bicyclic ketone  28 then was converted into (±)-isolinaridiol diacetate (29) and (±)palauolide (30).  7  (a) Axanes 9 0 Me  2 e  20  THE, -78°C 2) 2 OEt 3 BF 3)KH,THF  21  22 R=NHCHO (±)-Axamide-1  23 R=NC (±)-Axisonitrile-1  1 ” 10 (b) Clerodanes  0 1) 7, CuBrSMe , 2 OEt THE 3 BF , 2 -78°C KH, THE; EtOH, heat  24 0  Me  25  26 (±)-Stephalic Acid  1) 7, CuBrSMe , 2 0 3 BF , 2 THE Et 2) t-BuOK, t-BuOH  5.,Me 27  OAc  29 R= J,0Ac  28  (±)-Isolinaridiol diacetate  (C) A Sesterterpene 12  27  28  30 R= (±)-Palaulolide  Scheme  1-3  8  1.2. The Problem.  +  (1-2)  31  The angularly fused tricyclic sesquiterpene (-)-f3-panasinsene (31), isolated by Yoshihara and Hirose in 1975,13 attracted our attention due to (a) its novel structure, and (b) the presence of a methylenecyclohexane [3.2.O]heptane unit.  ring  cis-fused to a substituted bicyclo  It seemed reasonable to assume that the  methylenecyclohexane annulation  protocol  (vide supra, pp. 2-4)  utilizing the copper (1)-catalyzed conjugate addition of the Grignard reagent 7 could be extended from the use of monocyclic enone substrates having a trisubstituted double bond (20, 24, 27) to the use of a bicyclic enone substrate having a tetrasubstituted double bond.  If the methylenecyclohexane unit present in (-)-j3-panasinsene  (31) is disconnected and the remaining bicyclic portion is suitably functionalized, an enone such as 32 with a tetrasubstituted double bond results (equation 1-2).  Preparation of an enone similar to 32,  or its synthetic equivalent, then would be a key part of a possible synthesis of (±)-/3-panasinsene (31).  9  The strategy we wished to employ in the synthesis of (±)-(31), which would have the methylenecyclohexane  panasinsene  annulation as a key sequence, differs significantly from those of the two previously reported approaches (vide infra, p. 18).  1.3. Angularly Fused Terpenoids.  The carbon skeleton of (-)-f.3-panasinsene (31) is an example of one of a variety of angularly fused skeletons found in terpenoid natural products.  A few of the ring size combinations found in  terpenoids are depicted in Scheme 1-4.  The unifying feature of the  skeletons is the presence of a bridged spirane arrangement of 14 such that three variously sized carbocyclic rings share a rings common carbon atom. The synthesis of angularly fused terpenoids has been approached by a number of different methods which fall into three main categories.  In effect, the strategies may be divided according to  whether  monocyclic  a  annulation,  a  bicyclic  substrate substrate  is is  utilized employed  for  a  two  ring  for a  one  ring  annulation, or a monocyclic substrate is subjected to two sequential one ring annulations.  The approach chosen depends, of course, on  how the retrosynthetic analysis was performed on the target, particularly with respect to which rings were found to be strategic •  for  preservation  disconnection.  and  which  were  found  to  be  strategic  for  General factors to be considered in the antithetical  analysis include the sizes of the rings, the connectivities and  10  stereorelationships between the rings, the functional groups present and the availability of suitable precursors. 3 3-5-6  4-5-6  5 33 Cycloeudesmol’  4-5-6  31 13-Panasinsene 13  34 Panasinsanol B 16  5-5-5  4-5-6  7 35 Perforatone  36 PentaIenene 8  5-5-5  5-6-7  37 Retigeranic Acid 19  38 Gascardic Acid 20 H 2 CO  Scheme  1-4  11  In  the  first  synthetic  approach,  a  framework  tricyclic  is  assembled in one step from a monocyclic precursor having an appropriately substituted side chain (or side chains) which can undergo a cycloaddition reaction such as a carbene insertion, a 2÷2 cycloaddition, or a 2+2+1  cycloaddition.  Two examples of this  approach involve the syntheses of the terpenoids pentalenene 36 and retigeranic  acid  37.  The key step in Schore and Rowley’s  21 of (±)-pentalenene (36) was the octacarbonyldicobalt synthesis catalyzed 2+2+1  cyclization of the enyne 39  (Pauson-Khand  22 vide infra, p. 26) to make the angularly fused tricyclic cyclization, enone 40.  The enone 40 then was converted into 36 by a series of  chemical reactions (equation 1-3).  (1-3)  39  40  36  In Corey’s synthesis 23 of (±)-retigeranic acid (37) the key step in assembling the angularly fused triquinane portion of the target was a 2+2 cycloaddition reaction.  Thus, the carboxylic acid function  in 41 was converted into the corresponding ketene which underwent an intramolecular 2+2 cyclization to give 42.  Ring expansion of the  four membered ring, ring contraction of the six membered ring, and suitable functional group manipulations served to convert 42 into the racemic natural product 37 (equation 1-4).  Corey’s synthesis of  12  37 illustrates the fact that rings of sizes differing from those in the final product may be assembled initially due to the existence of a convenient route for their preparation and then, provided the appropriate methods exist, suitable ring expansions/contractions may be employed to create the desired ring size. which  a two  ring  annulation  performed was also  used  onto  for the  a  The approach in  monocyclic substrate  previous  is  syntheses of -  panasinsene (vide infra, p. 18).  (1-4)  ...IIII  H 2 CO 37 A second approach to the synthesis of angularly fused terpenoids involves the annulation of a bicyclic substrate to produce the required tricyclic skeleton.  An example of this approach is a  synthesis of pentalenene (36) different from that discussed above. Thus, the key skeleton-assembling step in the synthesis of (±)pentalenene by Piers and 24 Karunaratne was a methylenecyclo pentane annulation reaction on the tricyclic enone 47. The enone 47 was derived from the dione 43, which, in turn, was prepared from  13 H C 2 MeQ  Jls.CQ M 2 eh10 base  :>__+  (I -5)  acyclic precursors via the Weiss-Cook condensation 25 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 subjected to a series of chemical manipulations to furnish the enone 47. The enone  intermediate 47  annulation  to  underwent  the  methylenecyclopentane  provide the tetracyclic ketone  (48) which was  transformed by standard means into (±)-pentalenene (36) (Scheme I5).  The above described approach, in which the third ring is  annulated onto a bicyclic substrate to make a tricyclic skeleton, is the one we chose to use for the synthesis of j3-panasinsene (31). The sequential assembly of the rings making up the tricyclic framework is a third approach that has been employed to achieve the synthesis of angularly fused terpenoids.  The Boeckman synthesis ° 2  of gascardic acid (38) exemplifies this third approach. annulation of the six-membered  A one-pot  ring onto the enone 20 was  performed via conjugate addition of cuprate reagent 49, trapping of the resultant enolate anion with the enone 50, and a subsequent base-catalyzed aldol condensation to provide the key intermediate 51 (Scheme 1-6).  Introduction of a functionalized two carbon unit  at the a-carbon of the a,J3-unsaturated ketone function in 51 was a key objective in their synthesis.  However,  compound 51  proved to  14  H  47  46  t’CI çMgBr 1) 2 CuBrSMe ,  THF, -78°C 2) KH, THE H  36  48 Scheme  be  quite  Therefore, which  1-5  unreactive  towards  compound  51 was converted into the vinyl ether 52  subjected  to  a  conjugate  Claisen  additions  rearrangement to  reactions.  provide  53.  Intermediate 53 then was transformed into gascardic acid (38) via a  series  of  standard  chemical  reactions  which  included  intramolecular cyclization to form the 7-membered ring. interest to  note  (for future  reference)  that  an  It is of  steric congestion  contributed to the lack of reactivity towards conjugate addition reactions of the bicyclic enone 51. The examples chosen above, while illustrative of the different approaches to the synthesis of angularly fused terpenoids, also were chosen to demonstrate applications of reactions or strategies  15  studied as a part of the research reported herein.  Thus, the Pauson  Khand cyclization, the Weiss-Cook condensation, a ring contraction, and a conjugate addition to  a tetrasubstituted enone will be  discussed in greater detail at suitable points in the thesis. ‘——Cu Li, -78°C; -78° C .-20°C  1)  0  _fl’’ 2 ) TMS  20  3 C 4 (CH 2 ) H(OCH  o  50 -20°C 3) Base, MeOH  51  OMe  Ij s-collidine, ‘4  1 6 0°C  ‘4  CHO 2  38  53 Scheme  52 1-6  1.4. Isolation and Structural Elucidation of f3-Panasinsene (31).  In 1975 Yoshihara and Hirose’ 3 reported their isolation of the new  sesquiterpene  hydrocarbon,  J3-panasinsene  (31), from the  16  neutral portion of the volatile oil extract of the roots and rootlets of both fresh and commercial dried ginseng (Panax ginseng C.A. Meyer) from Japan and commercial dried ginseng rootlets from Korea. The molecular weight of 31 was found to be 204.  Signals for  three tertiary methyl groups and an exocyclic methylene group were found 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 at  1365 and 1360 cm’ in the ir spectrum indicated that two of the methyl groups were geminal.  Catalytic hydrogenation of 31 gave  two dihydro derivatives which were identical with two derivatives obtained by the catalytic hydrogenation of the endocyclic double bond isomer, a-panasinsene (54) (also isolated at the same time). Further confirmation of the structure of J3-panasinsene  (31) was  obtained by ozonolysis of 31 to give the known ketone (55), an intermediate in Parker’s synthesis of neoclovene (56).26  It was  found too, that 31, when treated with concentrated sulfuric acid in diethyl ether, rearranged to give a-panasinsene (54), a-neoclovene  (56), and j3-neoclovene (57). is  isomerized  to  It was known that caryophyllene (58)  cx-neoclovene  (56)  upon  treatment  with  concentrated sulfuric acid and Parker had postulated a cation with the panasinsene framework as an intermediate. 27  Therefore, based  on the spectroscopic evidence and the chemical behavior of the compound, the structure of J3-panasinsene (31) was established. The roots and rootlets of ginseng (Jen-shen) have been used in Chinese medicine for centuries, mention of jen-shen having been made in Chinese pharmaceutical works with traditions dating to the  17  later Han period.* with  28  Ginseng is still claimed to be a wonder remedy  anti-fatigue, anti-diabetic, anti-stress, and central  system  stimulant  and  sedative  However,  properties.  nervous which  components are responsible for which properties is still being investigated.  As yet, the biological activity of 13-panasinsene (31)  is unknown. 29  31  54  55  56  *  57  58  25 for jen-shen, The Han dynasty was in power 206 B.C.-250 A.D. in China. The entry 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 and benefits one’s wisdom. Consumed over a long time, it takes the material weight from the body and extends one’s years of life. Other names are ‘man’s bit’ and ‘demon’s cover’.”  18  1.5. Previous Syntheses of J3-Panasinsene  (31).  J3-Panasinsene (31) has been synthesized twice previously, by McMurry and Choy in 198O° and by Johnson and Meanwell in 1981.31 Both syntheses, as mentioned earlier, made use of the same general approach.  Thus, the two research groups introduced an unsaturated  side chain onto a cyclohexanone and performed a photochemical 2+2 cycloaddition reaction to assemble the tricyclic carbon framework. The details of the syntheses differ as outlined in the following description of their routes.  1.5.1. McMurry and Choy’s Synthesis of Racemic a- and $-Panasinsene (54 and 31).  ° of the panasinsenes (54 and 3 The McMurry and Choy synthesis 31) commenced with the alkylation of the sodium enolate of 2methylcyclohexanone  (59) with 1 -bromo-4-methyl-3-pentene (60)  to give 61 (Scheme 1-7). salt of dimethyl  Compound 61 was treated with the lithio  phenylthiomethyiphosphonate (62) to give a 9:1  4 to the mixture of vinyl sulfides which were oxidized with Nal0 corresponding sulfoxides (63).  Deconjugation of the sulfoxides with  dimsyl potassium provided the allyl sulfoxides 64. with  trimethylphosphite  alcohols (65).  provided  a  mixture  Treatment of 64  of the  key diene  Photolysis of 65 in diethyl ether in the presence of  copper (I) triflate gave the 2+2 cycloaddition product (66) as a mixture of epimeric alcohols which were oxidized to the ketone (55).  Ketone 55 was reacted with methyllithium and the resulting  19  alcohols were dehydrated to provide synthetic racemic a- and panasinsene (54 and 31).  2) Na10 4  61 S(Q)Ph —  —  3 S 2 KCH OCH  65  64  63  hv, CuOTf  H Pcc,  66  1) MeLi 2) SOC, /’ 2  55  +  5:2  54  31  Scheme  1-7  1-  20  1.5.2. Johnson and  Meanwell’s  Synthesis of (-)-J3-Panasinsene and  Its Enantiomer  ((-)- and (+)-31).  Johnson  Meanwell  panasinsene addition  of  and  began  their  31 of synthesis  ((-)-31) with the copper (1)-catalyzed 4-methyl-3-pentenylmagnesium  methyl-2-cyclohexen-1-one  (24) (Scheme 1-8).  bromide  (-)-J3-  conjugate (67) to 3-  The enolate anion  intermediate was trapped with formaldehyde and the resultant mixture of keto alcohols was converted to the corresponding keto tosylate mixture.  The tosylates were subjected to a base-catalyzed  elimination reaction to provide the enone 68.  Photolysis of 68 in  pentane provided racemic 55 by means of a 2+2 cycloaddition of the alkene functional group and the double bond of the enone.  Resolution  of the enantiomers occurred by reaction of the carbonyl function in 55 with (S)-(N-methylphenylsulfonimidoyl)methyllithium provide a mixture of diastereomers (70 and 71).  69 to  The diastereomers  were transformed separately into (+)- and (-)-/3-panasinsene ((+)-31 and (-)-31), respectively, by treatment with aluminum amalgam and acetic acid in wet THE.  21  MgBr 1)  2 (5 mol%) CuBrSMe  2) gaseous CH O 2 3) p-TsCI, Pyr 4) t-BuOK, t-BuOH  24  68 hv, pentane  NMe  II  (69)  55  70 AI/Hg,  HF  (+)-31  c 71 Al/Hg,  (-)-3 1 Scheme  1-8  22  II. DISCUSSION Total Synthesis of (±)-j3-Panasinsene  2.1.  31.  Retrosynthetic Analysis.  Our retrosynthetic analysis of (±)-J3-panasinsene 31 was guided by  two  main  strategies,  namely:  (a)  the  application  of the  methylenecyclohexane annulation transform described earlier (pp. 24) and (b) the utilization of the bicyclic enone 74 as a key intermediate.  The bicyclic enone 74 represents an important branch  point in the analysis since it may be derived from a variety of precursors.  The exact structure of the enone (i.e., the nature of R  and 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 Wolff rearrangement transform would “convert” 31 into the tricyclic ringexpanded ketone 72 (Scheme D-1).  The R group in ketone 72 was  expected to be either a methyl group as is present in the natural product or a methoxycarbonyl moiety.  There is ample precedent for  the synthetic conversion of a methoxycarbonyl function  into a  methyl group. 32 Retrosynthetic functional group manipulations and a functional group introduction would transform 72 into the ketone 73, in which XX is a carbonyl equivalent. *  Application of the  methylenecyclo  The numbering scheme utilized for (±)--panasinsene (31) is analagous to the one 16 for panasinsanols A and B. used by Iwabuchi and coworkers  23  M  15  14  :Q  12  Me le  72  31  cji R  ,SnMe 1 L 3 + H  73  74  5  Scheme D-1  hexane annulation transform to 73 would yield two fragments, the vinylstannane 5 and the key bicyclic enone 74 (Scheme D-1). well  documented  cuprate  in  reagents to  the  It is  literature that conjugate additions of  bicyclo[3.3.O]oct-1-en-3-ones occur on  convex face of the enone to give cis-fused  33 diquinanes.  the The  preference for the formation of cis-fused diquinanes is probably due to (a) the fact that trans-fused diquinanes are more strained than the cis-fused state  for  the  34 and (b) the likelihood that the transition isomers reaction  has  some  product-like  3 character.  Consequently, the transition state leading to the formation of the cis-fused product will be of lower energy than that leading to the trans-fused adduct, and the cis-fused diquinane will be formed  24  preferentially.  As outlined in the Introduction (pp. 2-4), cyclization  of the keto chloride intermediate in the methylenecyclohexane annulation sequence gives the cis-fused annulated product when further equilibration is impossible.  Thus, provided that the R group  was already installed on the enone double bond, the methylenecyclo hexane annulation would be expected to give rise to a product with the desired relative stereochemistry at the three chiral centers (C 1, C-4 and C-7) of /3-panasinsene (31). Retrosynthetic  analysis  of  the  important  approached in two different ways (Scheme D-2). is a methyl group,  74  enone  was  In the case where R  utilization of the Pauson-Khand cyclization  transform would lead to the linear enyne 75.  The enyne 75 can be  disconnected 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 assembled  synthetically by the sequential alkylation of an anion derived from 78 or 79 with the alkylating agents 76 and 77.  Utilization of 78 or  79 to prepare the enyne 75 would mean that, at an appropriate stage in the synthesis, it would be necessary to transform the dithioketal derived functions (i.e., XX=1,3-dithianyl or XX=MeS, 2 S0 p -Tol) into a carbonyl function (XX=O) in order to provide the tricyclic ketone 72 desired for the Wolff rearrangement reaction. alkylations  of  78 and 79  followed  by the  The  proposed  hydrolysis  of the  dithioketal derived functions to a carbonyl group are consistent with the previously reported use of 7836 and 7937 as masked acyl anion equivalents.  25  In the case where R is a methoxycarbonyl group, retrosynthetic removal of the double bond and a disconnection of the carbon methoxycarbonyl bond (bond a) in the enone 74 would provide the known keto ketal 46.24,38  In the synthetic direction, the reactions  are a methoxycarbonylation and a dehydrogenation, respectively, both of which are known processes.  H  74  H  46  +  %%%_•S 4 S  or  78 76  77 Scheme D-2  P-ToI 2 MeS\,,SO 79  26  2.2. 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 cycloaddition reaction of a hexacarbonyldicobalt alkyne complex with an alkene. During the reaction, one of the carbon monoxide ligands of the complex is used, and the product generated is a substituted cyclopentenone. 3 ’ 22 9  An example using the generalized alkyne 80  and ethylene is illustrative (equation D-1).  The alkyne reacts with  octacarbonyldicobalt 81 to give the hexacarbonyldicobalt complex 82.  Heating the complex with ethylene produces the cyclopentenone  83.  II  +  (CO) 2 Co 8  HC \  CH: heat  ‘°  (D-1)  H  80  8 1  82 (Dashes=CO)  83  The intramolecular version of the Pauson-Khand reaction was first reported by Croudace and Schore in 1981.° cyclization  of  hept-1-en-6-yne  bicyclo[3.3.O]oct-1-en-3-one contrast, the attempted  They found that  (84) at 95°C (4 days) produced  (85) in 31% yield (equation D-2).  cyclization of hex-1-en-5-yne  In  (86) to give  27 H___-_\  (CQ) 2 Co 8  +  84  CC, 95°C, C 3 (Me) C 2 H(Me) CH (4days)  (D-2)  85  H (D-3)  (CO) 2 Co 8  +  bicyclo[3.2.O]hept-1-en-3-one  (87) yielded only products from  alkyne trimerization (equation 0-3). Various substituents may be tolerated on the alkyne and alkene However, electron-withdrawing groups on the R, 2 (CO b 41 or on the alkyne moiety 39 R, or CN) 2 olefin (CHO, COR, CO  functional groups.  2 are detrimental to the cyclization due to the formation ) 4 2 CH(OEt) of dienes via a hydrogen migration.  This is illustrated by the  42 cyclization of the enyne 88 to give the diene 89 (equation D-4). The presence of electron-donating groups on the alkene moiety ) may be 45 44 or SR 43 or on the alkyne function (OR (OC(O)R or OR) beneficial. EtC  OEt TBDMSC  TBDMSC  Me + Me  88  (00) 2 Co 8  (D-4)  CH C 2 ,rt, 8h (83%)  NMO=N-methylmorpholine N-oxide  Me  89  28  The Pauson-Khand cyclization  reactions of enynes to give  bicyclo[3.3.O]oct-1 -en-3-ones have been studied to determine the effects on the cyclization of various substituents on the carbon chain linking the alkene and alkyne functions.  The yields, reaction  times and diastereoselectivities were affected.  Some of the  results, relevant to our attempted synthesis of (±)-/3-panasinsene (31), are presented in the following discussion.  It may be noted that  C-7 refers to the position on the carbon chain which becomes C-7 in the  bicyclo[3.3.O]oct-1-en-3-one  produced  in  the  reaction  (see  equation D-2). The yields of cyclizations of enynes with substituents on the carbon chain linking the alkene and alkyne functions are generally better than those of the less substituted cases, and the reaction times are  usually shorter.  Thus, for example,  cyclization  of the  TBL .R , 8 ( 2 Co CO) CC, heat  90 (R=H) 91 (R=Me)  (D-5)  90a (14%) 91a (78%)  enyne 90 occurred in only 14% yield, while cyclization of the enyne 91 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 reaction time when compared with the cyclization of the enyne 84 (4 days) (equation D-2).  Good yields are also obtained with enynes in which  29  the carbon chain between the alkene and alkyne functional groups is substituted only in the homopropargylic (C-7) position.  Cyclization  of the enyne 92 to give a mixture of 92a and 92b in 86% yield after 20 hours is a pertinent example (equation D-6). 47 The  improved  yields for the cyclizations of the enynes 91 and 92 (equations D-5 and 0-6) when compared with the enyne 84 (equation D-2) were proposed 4 ’ 46 7 to be due to the more favorable enthalpy and entropy of the reaction (Thorpe-Ingold effect). 48  TMS  \Me R  (20h)  H  92 2 R=CO E t  92a  H 55  : 45 (86%)  92b  The diastereoselectivities of the cyclizations of the enynes 9 1 and 92 differ significantly. 47  Thus, the cyclization of the enyne 91  gives rise to a very high selectivity in favor of the enone 91a, while cyclization of enyne  92  leads to virtually no selectivity.  The  results are as would be expected based on the steric effects of 1,3versus 1,4-pseudo diaxial interactions in the transition states (vide infra).  Further confirmation of the importance of steric effects to  the diastereoselectivity in the cyclization  may be obtained by  comparing the results of the cyclizations of enynes 91 and 9 3 (equation  The trimethylsilyl (TMS) group is more  sterically bulky than the methyl group.  As expected, the cyclization  30  TBDI  R  R TBDr  (CO) 2 Co , 8 (D-7) heptane  (20h)  a  91 R=TMS 93 R=Me  91a (79%)  93a  91b (3%) 93b (15%)  (50%)  reaction of the enyne 91 gave rise to a diastereoselectivity greater than that of the enyne 93 (79:3 versus 50:15). The working hypotheses for the mechanism of the Pauson-Khand cyclization as proposed by Magnus ° and Schore 5 51 invoke the same types of intermediates,* but the Magnus mechanism was developed to rationalize the stereoselectivity of the intramolecular reaction of various enynes including 91 and 93 (equation D-7).  Thus, according  to the Magnus proposal (Scheme D-3), alkene insertion into the internal C-Co bond of the hexacarbonyldicobalt complex 94 leads to the  formation  of  two  cobaltabicyclooctanes  metallocycles are likely to be cis-fused.  95 and 96.  Both  In the transition state  leading to metallocycle 95, the steric interactions between R 1 and 2 are minimized, but in the corresponding transition state for the R metallocycle  96 there is a severe 1,3-pseudo diaxial interaction  between R 1 and R . 2  Consequently, the pathway for the formation of  96 and thus, of the enone 100 is disfavored, particularly in the case of sterically bulky R’ groups.  For the metallocycle 95, insertion of  a carbon monoxide ligand into the indicated C-Co bond gives rise to the acyl-Co complex 97. *  Migration of the other C-Co bond to the  None of the various intermediates have, as yet, been isolated.  31  adjacent electrophilic carbonyl group produces 98.  The reductive  elimination of the cobalt carbonyl residue (likely 6 (CO) in 98 2 Co ) leads to the formation of the enone 99.  The exact identity of the  52 but in isooctane cobalt residue initially eliminated is uncertain, (CO) 4 [Co 5 ] 12 3 has been isolated and in aromatic solvents, such as benzene, 54 (PhH)] 4 [Co ( 9 CO) has been found (Scheme D-4).  94  95  96  1 100 1 R  H  Dashes=CO  98  99 Scheme D-3  32 CO) 6 ( 2 Co  -  12 Co(CO)  CO) 8 ( 2 \CO O  PhH  CO)g(PhH) (Co 4  PhH, heat  Scheme 0-4  H for enyne 94), R = In cases with substituents only at C-7 (i.e., 2 the steric interactions are less significant than those described above.  Thus, the main steric interactions are a 1,4-pseudo diaxial  4 and an interaction between the 3 or R 1 and R interaction between R . 4 3 or R metallocycle methylene and R  Understandably, the effects  are smaller than for 1,3- or 1,2-pseudo diaxial arrangements of  substituents.  Reduced steric interactions would lead then to a  reduced diastereoselectivity as is observed in the cyclization of the enyne 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 have been used in the syntheses of a variety of natural products and natural  product  1 6 7 29 55 precursors. ’ 4  It seemed that the  cyclization would provide a viable approach to the preparation of a bicyclic enone 74 which was a desired intermediate in the synthesis of (±)--panasinsene (31).  In one step a successful Pauson-Khand  reaction would transform an appropriately substituted enyne 75 into  33  the enone 74 (equation D-8) needed for the key methylenecyclo hexane annulation sequence (pp. 2-4).  R xx  xx  °  (D -8)  :67  According to the synthetic plan, the enyne 75 would have a methyl group on the alkyne function and would have a suitable functional group at C-7. purpose.  The function at C-7 would serve a two-fold  In the first place, by analogy to the examples given earlier  (equations D-5 and D-6 compared with equation D-2), it was hoped that the presence of substitution at C-7 would contribute to an acceptable reaction yield and reaction time for the cyclization. Secondly, the moiety at C-7 was to be used as a “handle” for future functional group manipulations at that position. nature of the functional group at C-7 was important.  Therefore, the Given that a  future step in the planned synthetic sequence (equation D-9) called for a carbonyl group at the position corresponding to C-7, the XX moiety on the enyne 75 would have to be transformable into a carbonyl group.  Also, the viability of the synthetic plan depended on  the stability of the XX moiety to the reaction conditions encountered before it was to be converted into a carbonyl group.  34  R  o  (D -9)  101 Potential functionalities at C-7 of the enyne 75, in terms of future usefulness in the synthesis, would include oxygenated groups (XX=OR,  H,  or  XX=2,2-dimethylpropan-1,3-dioxy)  or dithioketal  derived groups (XX=1 ,3-dithianyl, or XX=SMe, 2 S0 p -Tol).  To date,  there have been no reports of dithioketal derived moieties being employed as C-7 substituents of enynes subjected to the Pauson Khand cyclization.* If such groups were viable options, they would further expand the versatility of the cyclization due to the fact that the  hydrolysis  of  dithioketal  derived  functions  yields  the  corresponding carbonyl 36 group, 3 ’ 7 while desulfurization of the 1,3a 36 dithianyl group with Raney nickel provides a methylene group. Thus, the dithioketal derived functions (1,3-dithianyl and SMe, 2 S0 p Tol) at C-7 of enyne 75 were investigated with a view to their utility in the synthesis of (±)-3-panasinsene (31).  *  In the intermolecular Pauson-Khand reaction, a methylthioether tethered to the olefin function by a carbon chain has been used to enhance the regioselectivity of the reaction 56  35  2.2.2.1.  The 1,3-Dithianyl Function at C-7.  The ability to reversibly invert (umpolung) 57  the  normal  reactivity of an acyl carbon atom is a powerful tool in organic synthesis. 5 ’ 36 8  Thus, while acyl groups are generally attacked at  the electrophilic carbon by nucleophiles (for example, 102 gives 103, Scheme D-5), an umpolung causing group on the acyl carbon permits the atom to function as a nucleophile. 57  For example,  conversion of formaldehyde (102) into 1,3-dithiane (78) followed by deprotonation of 78 with n-butyllithium produces a nucleophilic anion. 104.  Treatment of the anion with an electrophile (E) provides The dithianyl group of 104 then may be hydrolyzed to produce  the aldehyde 1 05 or further deprotonated and reacted with an electrophile to give 106.  Hydrolysis of the dithiane function in 106  generates the ketone 107 (Scheme D-5).  A variety of electrophiles,  including alkyl halides, carbonyl compounds, small ring ethers and acylating reagents, may be employed to transform compound 78 into 104  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 N c 36 halosuccinimides.  The overall result of the reaction of the anion  of either 78 or 104 with an electrophile followed by hydrolysis of the dithiane function in the product is a nucleophilic acylation of the electrophile.  36  0  II  HS SH H”H Lewis Acid  102  1 )n-BuLi 2) E  Hydrolysis  H><H  E><H  78  104  0 II  105  1) n-Bu Li 2)E’  Nu:  0  OH  Hydrolysis  Nu—”H E’ 11 E”  103  107  106 Scheme D-5  We wished to exploit the reactivity of 1,3-dithiane (78) to prepare an enyne 75 on which to perform a Pauson-Khand cyclization to generate the corresponding enone 74 (equation D-1O; XX=1,3dithianyl).  Then, at an appropriate stage in the planned synthesis of  (±)-J3-panasinsene (31), the dithianyl function would be hydrolyzed to regenerate a carbonyl function.  The proposed reactions of 78  with allyl iodide (76) and with 1-iodo-2-butyne (77) appeared to be similar to alkylations reported earlier, 59 but in practice (vide infra) turned out to be somewhat problematic. R R— :XX (D -10)  - - - -  78  75  74  37  _JI  To 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), is ° 6 known.  Using a procedure similar to that reported for the  dithiane, a THE solution of commercially available a alkylations of 59 1,3-dithiane form the  (78) was treated with n-butyllithium  dithiane  anion.  at  “‘-25°C to  The solution of the anion was cooled  -78°C and allyl iodide (76) was added quickly.  to  Workup of the  reaction mixture and purification of the product led to the isolation of 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) 2) AIlyl iodide (76, 1.1 equiv), -78°C (2.5h); -78—* 0°C (1..5h)  SS  (D-1 1)  -  S..,.,..S  8  I  108  The ir spectrum (neat) of 2-allyl-1,3-dithiane 108  exhibited  absorptions due to the mono-substituted alkene at 3077 (w), 1640 (m), 991 (s), and 920 (vs) cm 61 In the 1 . 1 H nmr spectrum (400 MHz, CDCI ) 3 , 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, 2 CH=Ca ) , and at  38  ).* 2 5.82-5.92 (m, 1H, Ca=CH  The exact mass of the molecular ion  was found to be 160.0374 which is consistent with a molecular formula of 2 H, 7 C . S 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, a  dichioromethane solution of the commercially available alcohol 1 09 was treated with triphenylphosphine diiodide (1.1  equiv) in the  presence of triethylamine (1.1 equiv) to give the iodide 77 (equation D-12).  The ir spectrum of the iodide 77 exhibited an alkyne CC  stretch at 2235 cm, while the ‘H nmr spectrum (300 MHz, CDCI 1 ) 3 showed signals for the methyl group at 8 1 .83 (t, 3H, J for the methylene group at 3.68 (q, 2H, J  =  =  3 Hz) and  3 Hz).  OH  II  + PPh 3  +  ‘2  CI CH 2  (D-12)  N, rt, 3 Et (5-7h)  109  77  In order to prepare the enyne 75, a THE solution of 2-allyl-1,3dithiane (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-2butyne 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 signals for 108 are as follows: 3 1.6-3.0 (m, 8H), 4.1 (t, 1H, J= 6.6 Hz, -SCHS-) and 4.96.7 (m, 3H, -CH=CH ). 2  39  11 0 (an oil) and the allene isomer 111 approximately  equal  amounts  analysis) (equation D-13). generally  12-18%  (ratio,  (an oil) were formed in 1:1 .2,  respectively,  GLC  The isolated yield of each isomer was  because  the two  isomers  were  difficult to  separate from each other and from other by-products of the reaction  .  1) n-BuLi (1.1 equiv), THE 2)  108  110  The 1 H  nmr spectrum  (D-13)  +  ;5:ce  (400  111  MHz, CDCI ) of the alkyne 11 0 3  exhibited 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 5.875.97 (m, 1H).** In the ir spectrum (neat) the alkyne CC stretch occurred at 2235 (w) cm, 1 1638 (m) cm* 212.0697  while the alkene C=C stretch was at  The exact mass of the molecular ion was found to be  which  is  consistent with  the  molecular  formula  of  C 1 H . 2 S 1 6 The structure of the allene 111 was consistent with the ir and low resolution mass spectral data. *  1  H nmr,  Thus, in the 1 H  nmr  The purification procedure was not further optimized since the approach was ultimately abandoned. **Other hydrogen signals in the 1 nmr spectrum appeared at 3 1 .90-2.06 (m, 2H) and 2.76-2.93 (m, 8H). In the ir other absorptions due to the alkene were at 3076 (m), . 1 991 (m), 920 (s) cm-  40  spectrum (400 MHz, Cod ), the allene 111 exhibited signals for the 3 allenic methyl group at 8 1.82 (t, 3H, J  hydrogens at 4.88 (q, 2H, J  =  =  -2 Hz), for the allenic  2 Hz) and for the olefinic hydrogens at  5.13-5.20 (m, 2H), 5.79-5.90 (m, 1H).*  The ir spectrum of the allene  111 showed an allenic C=C stretch at 1953 (vs) cm 1 and an olefinic C=C stretch at 1639 (s) cm.  In the low resolution mass spectrum  the molecular ion was observed at 212 mass units (13%). The presence of allenes as by-products in the reactions of nucleophiles with propargylic alkylating reagents (and conversely, the presence of acetylenic by-products in similar reactions) is a persistent problem. 64 ’ 63  It is known that the alkyne/allene ratio in  the product can be influenced by a variety of factors which include: the solvent,  the temperature, the structure of the propargylic  substrate and the structure of the nucleophile. 65  Thus, for example,  in the reaction of a Grignard reagent with 112 (equation D-14), the acetylene/allene ratio in the product mixture generally was larger solvents. In c at higher temperatures and smaller in less polar 65 contrast,  similar  reactions  (using  1 ,4-dichloro-2-butyne)  of  alkyllithium reagents tended to give the opposite results in that raising the temperature led to a higher proportion of the allene in d 65 the product mixture.  Also, different products have been obtained  65 a 65 employed. depending on the nucleophile d  Thus, in the reaction  of methyllithium with 11 2 (X,Y=Cl) the product was mainly the allene *  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*  41  methylmagnesium bromide gave mainly the acetylene 113 (R=Me, Y=Cl).65d  2 CH  V  x  Nu:, solvent  Ii  +  (D-1 4)  R’)  112 Y= OMe, OH Alkyl X= Cl, Br Nu:= RMgBr  113  114  In our hands, various modifications in the reaction conditions for 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 (-78°C,  the solvent  (THE,  DME), (b)  varying the  --25°C), or (c) using an additive (none, HMPA).  results similar to obtained.  temperature However,  or worse than those described above were  In addition,  modifying the alkylating  reagent 77 by  replacing the methyl group with the more bulky triisopropylsilyl (TIPS) group 66 to give the iodide 116 did not improve the outcome of the reaction.  The option of changing the structure of the nucleophile  remained to be explored, but first the feasibility of the Pauson Khand cyclization reaction of an enyne with a dithioketal function at C-7 needed to be established.  42  n  —/ —  I  TIPS  77  —  _,,‘  116  115  The enyne 11 0 was subjected to Pauson-Khand cyclization conditions similar to those reported by Magnus. 46  Thus, the deep-red  hexacarbonyldicobalt alkyne complex 117 was formed by reacting the alkyne function in enyne 11 0 with octacarbonyldicobalt (1 .1 equiv) (Scheme D-6).  Purification of the crude product mixture by  rapid chromatography on a Florisil column gave the complex 117 in 68-78% yield.  The ir spectrum (KBr) of the complex exhibited  carbonyl absorptions at 2087 (m), 2045 (s), and 2016 (s) cm** The hexacarbonyldicobalt complex 117, dissolved in heptane and sealed in a resealable tube under a carbon monoxide atmosphere, was heated at 90-100°C for 14-18 hours.  The yield of the purified enone  118 (a pale yellow oil) was 5O% based on the enyne (or 70-80%, based on the isolated complex).  The yield and reaction time were  comparable to those reported previously (compare with equation D 7).  *  The absorptions due to the organic moiety were very weak in comparison with the carbonyl absorptions. Absorptions (br, vw) due to the olefinic group appeared at 967 and 931 cm, while the absorption due to the Co-C was at 519 cm1 . There was no 1 octacarbonyldicobalt present as evidenced by the lack of an absorption at 1864 cm1 due to bridging carbonyl Iigands. 43  43  (CO) 2 Co 6 S  8 ( 2 Co , C0) CO, heptane, rt  —  110  I  —  117 heptane, CO 90-100°C (14-18h)  Me  118 Scheme D-6  The structure of the enone 11 8 was confirmed by the ir, nmr and high resolution mass spectral data.  Thus, the ir spectrum of the  enone 118 exhibited absorptions at 1707 (vs) and 1672 (vs) cm1 which are characteristic of a conjugated cyclopentenone. ’ In the ‘H 6 nmr spectrum (300 MHz, CDCI ; traces of impurities present), the 3 vinylic methyl group appeared at 3 1.73 (br s, 3H) and the angular hydrogen appeared at 3.25-3.39 (m, 1H).*  In the high resolution mass  spectrum the exact mass was found to be 240.0644, which is consistent with the molecular formula, 2 0S C 1 H . 2 6 The Pauson-Khand cyclization of enyne 110 occurred with an acceptable yield in a reasonable length of time and the enone 11 8 was  isolable.  However,  the  overall  route was  synthetically  mediocre 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) cm . In the 1 1 H nmr  spectrum, signals due to the other hydrogens appeared at 31.58 (t, 1H, J (dd, 1H, J = 17, —6 Hz) and 2.83-3.20 (m, 7H).  2.06-2.17 (m, 3H), 2.70  =  12 Hz),  44  of the enyne 110.  The problems included: (a) the low yield in the  alkylation reaction with  1-iodo-2-butyne 77 and (b) the difficulty  in separating the desired alkylated product 11 0 from the allene Thus, it was decided to modify the  isomer 111 (equation D-13).  dithioketal function and use an oxidized dithioketal derivative, the methylthio-p-toluenesulfonyl function.  It was expected that the  more reactive anion (11 9) would show different (preferably more desirable) behavior in the alkylation reaction with the alkyne 77. MeS  -  -Tol S0 p 2  119 2.2.2.2. The Methylthio-p-toluenesulfonyl Function at C-7. c ’ dithioacetal S xides or 69 8 S,S-dioxides, 67 S-monosulfo 6 Several 3 including methyithiomethyl p-tolyl sulfone (79),37 have been used as masked carbonyl anion equivalents.  However, ketone synthesis via  the oxidized dithioacetal reagents, 120-122, may be accompanied by problems such as: (a) competative alkylation on a monoactivated c 37 alkyl group instead of at the doubly activated methylene position; a or (c) difficulties in the hydrolysis of 68 (b) little or no dialkylation; a 68 the oxidized dithioketal function to generate the carbonyl group. In  contrast,  for sulfone 79, site-selective  deprotonation  of  the  45  methylene group occurs with a variety of bases* and produces an anion  which  may  be  reacted  with  electrophiles  such  as  72 a,I3-unsaturated carbonyl compounds c, 37 aldehydes ,37c71 esters, 71 c b, 37 and alkyl halides.  The dithioacetal S,S-dioxide function in the  adducts so formed may be hydrolyzed to give an aldehyde carbonyl group or again treated with a base to form the a-anion.  Reaction of  the anion with a second electrophile followed by hydrolysis of the methylthio  p-toluenesulfonyl function  produces a ketone.  The  hydrolysis of the methylthio p-toluenesulfonyl group generally has been  performed  b 37 photochemically  or  using  acidic  l a,73 7 c, I, 37 Co nditions. PToI 2 MeS\/SO 79  MeS,,SOMe  MeS,SO M 2 e  EtS\/SOEt  12067  12169  12268  It appeared that methylthiomethyl p-tolyl sulfone (79) was a viable carbonyl anion equivalent and could be used in our strategy to prepare the enone 74 via the Pauson-Khand cyclization of an enyne 75 (equation D-15; XX=SMe, S0 p-Tol). 2  As was the case for 1,3-  dithiane (78) described earlier, the sulfone 79 was alkylated with allyl iodide (76) and 1-iodo-2-butyne (77) to prepare the desired enyne.  *  Bases used include: 50% aqueous NaOH-toluene/trioctylmethylammonium c 3 b 37 chIoride; 7 NaH-DMF; C3 37 70 NaHTHF; C 2 K i-PrOH 0 ;37c and n-BuLi-THF. 7  46  2 MeS,S p -Toi O  -  -  Methylthiomethyl  5 o  :  (D-1 5)  p-tolyl sulfone (79) (also available from the  Aldrich Chemical Co.), was prepared and recrystallized according to the procedure reported by Ogura and coworkers. 74  The product thus  obtained exhibited m.p. 83.5-85°C (lit. 82-83°C ) and 1 74 H nmr data in accord with the reported data. A cold (-78°C) THE solution of the sulfone 79 was deprotonated with n-butyllithium and the resultant anion was treated with allyl iodide (76) (equation D-16).  The purified monoalkylated material  (123) was obtained as a white solid (crude m.p. 37-38°C) in 7O75% yield.  The amount of the dialkylated product 124 obtained and  that of the starting sulfone 79 recovered varied depending on the scale of the reaction and the number of equivalents of allyl iodide (76) used.  Thus, in our hands, for a small scale reaction (-2 mmol  of 79) 1.1 equivalents of allyl iodide 79 each in -10% yield.  were found to give 124 and  However, in a large scale reaction (‘-10  mmol of 79), four equivalents of allyl iodide (76) were needed to obtain a similar result.  Other conditions generally led to the  formation of either 124 or 79 in larger amounts.  47  P-Tol 2 MeS\,SO  7 9  1) n-BuLi (1.1 equiv), MeS p-Tol MeS 2 S0 p-Tol 2 S0 THE, -78°C (lh) + %j ( D 1 6) 2) AIlyl iodide (76) J I (1.1 equiv), -78°C 1 (1.5h); -78—, -15°C (0.5h) 1 23 124  ><  -  In the 1 H nmr spectrum (300 MHz, CDCI ) of 1-methylthio-1-p3 toluenesulfonylbut-3-ene  (123) signals for the two methyl groups  were displayed at 32.24 (s, 3H) and 2.44 (s, 3H), while the signals for the three olefinic hydrogens were at 5.16-5.20 (m, 2H) and 5.735.87 (m, 1H).* ion peak at m/z  In the low resolution mass spectrum, the molecular =  256 2 0 1 H 12 (C ) S 6 was very small (0.3%), while the  base peak (101) corresponded to the loss of the p-toluenesulfonyl fragment, C S (m/z=155). 2 0 H 7 The  structure  of  the  diallylated  sulfone  1 24  was  deduced from its 1 H nmr spectrum (300 MHz, CDCI ). 3  readily  Thus, the  expected signals for the two methyl groups (s, 3H each) were at 3  2.25 and 2.36, while the signals for the olefinic hydrogens appeared at 5.12-5.20 (m, 4H) and 5.85-6.00 (m, 2H).**  In the ir spectrum  absorptions due to the C=C stretch and the sulfone asymmetric and symmetric S=O stretches were displayed at 1639, 1302 and 1147 , 1 cm  61 respectively.  In the low resolution mass spectrum, the  molecular ion appeared at m/z  *  =  296 (0.3%) and a major fragment  Other aliphatic hydrogen signals in the 1 H nmr spectrum of 123 appeared at 82.242.34 (m, 1H), 2.91-3.00 (m, 1H) and 3.71 (dd, 1H, J = 11, -4 Hz), while the aromatic hydrogens were at 7.36 (d, 2H, J = 8 Hz) and 7.84 (d, 2H, J = 8 Hz). ** The 1 H nmr signals for the remaining hydrogens of 124 were displayed at 32.602.81 (m, 4H), 7.33 (d, 2H,J = 8 Hz) and 7.84 (d, 2H, J = 8 Hz).  48  ) (C S 2 0 H corresponding to the loss of the p-toluenesulfonyl moiety 7 was observed at 141 (32%). The enyne 125 was prepared by treating a cold (-78°C) THE solution of the sulfone 1 23 with n-butyllithium and alkylating the  purified  isolated, (125)  was  The  1-iodo-2-butyne (77) (equation D-17).  resultant anion with  4-methylthio-4-p-toluenesulfonyloct-1 -en-6-yne  obtained  as  a  pale  oil  yellow  in  79-86%  yield.  Gratifyingly, none of the corresponding allene 126 was detected. MeS ,S0 p-ToI 2  p-ToI 2 S0  MeS 1) n-BuLi (1.25 equiv),  THE, -78°C (O.5h)  Ii  (D-17)  2) E801w  78—  15°C ( lh)  II  \  125  123  126 ) of a new 3 The presence in the ‘H nmr spectrum (300 MHz, CDCI methyl signal at 3 1.70 (t, 3H, J  =  -2Hz) for the acetylenic methyl  group and signals for two other methyl groups (s, 3H, each) at 2.31 and 2.45, as well as the other expected signals,* indicated that the desired 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= 8 Hz) and 7.87 (d, 2H, J = 8Hz).  6.01 (m, 1H), 7.35 (d, 2H, J  49  alkyne CC stretch at 2236 (very weak), and the sulfone S=O asymmetric and symmetric stretches (strong) at 13O2 and 1144 The molecular ion peak in the low resolution  61 cm, respectively.  mass spectrum was at m/z  =  308 mass units, consistent with the  6 0 0 1 C 2 H . S molecular formula, 2 Though similar reaction conditions were utilized, the results of the alkylation of the lithio anions of sulfone 123 and dithiane 109 with the iodide 77 were quite different.  Thus, for the anion 11 9  only the desired SN2 reaction was observed, while for the anion 115, both SN2 and SN2’ reactions occurred.  The differences in the  behaviors of the anions 119 and 115 are likely due to differences in the anion structure.  As mentioned earlier, reactions of Grignard and  organolithium reagents with propargylic substrates tend to give complementary results, but the provenance of this behavior is not fully  d 65 a, 63 understood.  However, our results confirm that the  structure of the nucleophile also plays an important role in the outcome of the reaction of a nucleophile with a propargylic halide.  MeS  SOToI  [1 119  E[y+ 115  50  The enyne 125 was subjected to the Pauson-Khand cyclization Cyclization of the enyne in this case can  reaction (Scheme D-7).  give rise to a mixture of diastereomers, but based on the results 47 and decribed earlier (equation D-6), the reported in the literature The enyne 1 25  diastereoselectivity was expected to be minimal.  was treated with octacarbonyldicobalt (1.2 equiv) in benzene. After  of  chromatography  the  crude  product  Florisil,  on  the  hexacarbonyldicobalt alkyne complex 127 was isolated as a deep red oil in 80-90% yield.  The complex was dissolved in benzene, and the  resultant solution was sealed in a resealable tube under a carbon monoxide atmosphere and then was heated at 80-90°C.  The mixture  H nmr analysis) was of diastereomers 128 and 129 (ratio -‘1:1, 1 formed in 46-54% yield based on the isolated complex 127, or in 37-49% yield based on the starting material (enyne 1 25).  The yield  for the cyclization of enyne 125 was a bit lower than that of the previous example (enyne 110, Scheme D-6), but the reaction also occurred somewhat more rapidly (5-12 hours versus 14-18 hours). The two diastereomers 128 and 129 could be separated by careful flash column chromatography or could be partially separated The  by fractional recystallization from dichloromethane-pentane. relative  stereochemistries  determined. diastereomer compounds.  of  the  two  compounds  were  not  Using the above solvent system, the less polar (A) Pure  crystallized A  was  first  from  recrystallized  mixtures from  both  of  diethyl  ether  dichoromethane (3:1) to produce prisms that exhibited m.p. 146.5) of A displayed the 3 148.5°C. The ‘H nmr spectrum (300 MHz, CDCI expected three singlets (3H each) for the methyl groups at 3 1.71  (br  51  SO2pToI  ::::  (CO) 2 Co 6 (1.::quhi),  125  127 PhH, CC, 80- 90°C (12h)  +  1  129  :1  128  Scheme D-7  s, allylic Me), 2.34 and 2.48.  The signal due to the angular hydrogen  appeared as a broad multiplet at 3 -‘3.28-3.40 (overlapped with a doublet 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 ‘ C nmr spectrum (75 3  MHz, CDCI ), fifteen signals were observed due to the fact that two 3 pairs of aromatic carbon atoms were magnetically equivalent (see figure 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 the bicyclic 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 ‘ C signals for the three methylene groups appeared at 335.8, 40.0 and 41.99; 3 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) and 130.94 (2.H).  52  0’ N ID+  0)-3  C  01  N  •  It) N N  o  -3  •  13  134  It3 ‘a 0 • U, —N c4* -) 10  ,_:)  132  - -  e  N It)  1. 82  80  78  76  74  N  ‘a 01 0l  ‘a “3 “5  ‘a  F..  It)  ‘a  ‘1  200  180  160  -  ‘a C  ‘a 0 N  220  ‘a F.  72 PP8I  140  1 120  100  80  60  40  20 PPM  Figure 1. The 75 MHz broad band decoupled 13 C nmr spectrum of the enone diastereomer A.  The diastereomer B was recrystallized from acetone/pentane and exhibited m.p. 149.5-152°C. In the 1 H nmr spectrum (300 MHz, ) of diastereomer B there were some slight differences from 3 CDCI that of A.  Thus, the signals for the three methyl groups (3H each)  appeared at 3 1.72 (br s, allylic Me), 2.36 and 2.47, while the angular hydrogen appeared at 3 —3.21-3.34 and the aromatic hydrogens reso nated at 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).  I  0  53  With the enones 128 and 129 in hand, the key methylenecyclo hexane annulation was investigated.  It was expected that the  reaction might be sluggish based on the results presented in the Introduction (pp. 2-4).  However, it was anticipated that the reaction  would be possible, since copper (1)-catalyzed conjugate additions have been performed on enones with tetrasubstituted double bonds. Thus, for example, Paquette and Han 75 found that the diquinane enone 130 reacted with the Grignard reagent 131 in the presence of a copper (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) , THE, 2 CuBrSMe S, -78°C (12h) 2 Me 130  :Q  (D-18)  132  The attempts to perform the copper (1)-catalyzed  conjugate  addition of the Grignard reagent 7 on the enones 128 and 129 were frustrated by the lack of solubility of the enones in THF, the solvent normally used in our laboratories 7 for the reaction.  The enones were  soluble in hexamethylphosphoramide (HMPA), an additive sometimes used  along  with  conjugate addition  trimethylsilyl  chloride  reactions of Grignard  to  improve  sluggish  76 (copper (I) reagents  catalysis) or of stoichiometric organocopper reagents. 77  However,  the amount of HMPA required to dissolve the enones was such that the polarity of the reaction solvent mixture would be significantly  54  It is known that 1,4-additions of cuprates to enones  increased.  78 (diethyl ether, dimethyl occur more readily in less polar solvents sulfide,  hydrocarbons)  trimethylsilyl  chloride)  and  that  retards  HMPA the  (in  the  of  absence  8 a 7 , 77 1 ,4additions.  Not  unexpectedly, the desired conjugate addition reaction of 7 to the enones 128 and 129 in a THF-HMPA solvent mixture (‘-24% v/v HMPA) was unsuccessful.  The reaction in the presence of TMSCI (2  equiv relative to the enones) also failed (equation D-19).  CI  Me.  (D-19) 133 and 134  128 and 129  Alkylated  methylthiomethyl  p-tolyl sulfones were reported to  undergo a facile hydrolysis of the thioether sulfone function to give the corresponding carbonyl compounds using a variety of mild reaction  73 ai 71 0, b, 37 conditions.  Therefore, it was decided to  hydrolyze the thioether sulfone function in 128 and 129 to the keto function in the enone 135.  The reactivities of the two carbonyl  groups in the enone 135 would be different, thus permitting the protection  of  the  unsaturated one.  saturated  ketone  in  the  presence  of  the  It was envisaged that the saturated ketone would  be reduced to the alcohol and protected as an ether.  The hydrolysis  reaction was attempted under several of the reported conditions 1 and C3 gel/CH 7 2 69 CuCI b 37 •silica a 2 (concentrated HCI/MeOH,  55  I but mixtures of products resulted and the H /MeOH 731 CuC / 0 2 ’), attempt was abandoned (equation D-20).  X  ‘-  o  ,  (D-20)  Tol  135  128 and 129  In summary, our use of the Pauson-Khand reaction to prepare the important bicyclic enone 74 (see Scheme D-1) on which we wished to carry out the methylenecyclohexane annulation was discontinued due to the unanticipated difficulties encountered which included; the poor yields in the alkylation of the 2-allyl-1,3-dithiane (109) with the propargylic iodide 77; the low solubilities of the enones 128 and 129 in THE; and the mixtures of products obtained in the attempted unmasking of the ketone carbonyl in the same enones. However, it was gratifying to find that the Pauson-Khand cyclization could be done with dithioketal derived functions at C-7 and that the yields and  reaction times were comparable to the previously  reported examples.  74  56  2.3. 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 3component (A+B+B or 2+3÷3) coupling reaction, 79 has proved useful in the synthesis of many natural and non-natural polyquinanes. 25 The reaction of a dialkyl 3-oxoglutarate (1 36) (2 equiv) with a 1,2dicarbonyl compound 137 in the presence of a base catalyst (or less commonly, an acid catalyst) produces, in high yield, a tetraalkyl cis bicyclo[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 the ester functions.  The 13-keto acid groups thus formed undergo a  spontaneous decarboxylation to generate the dione 139 (equation D 21). C 2 RO  C 2 RO 136  137  136  138  139  Many different tetraester and dione compounds (138 and 139) 25 may be produced via the Weiss-Cook condensation. of the R groups in 1 38  The identities  are determined by the ester of 3-  oxoglutarate which is employed; normally, R is either a methyl or a  57  The structure of the dicarbonyl compound utilized  t-butyl group.  determines the nature of the R’ and R” groups of 138 and 139; R’ and R” may be hydrogens, alkyl groups or aryl groups and may be either the same or different. (R’,  Thus, glyoxal (44, R’=R”=H), a-keto aldehydes  R”=H, alkyl/aryl), and acyclic or cyclic a-diketones  (R’,  R”=alkyl/aryl) may be utilized with the proviso that the bicyclic product may not be formed if very sterically bulky groups are used  (vide infra). ° 8  Also, dicarbonyl compounds with limited solubility in  the usual aqueous solvents (i.e., R’, R”=large alicyclic group) may be 81 employed if the reaction is performed in organic solvents. The utility of the Weiss-Cook condensation is due in part to the fact that both the tetraester intermediate 1 38 and the bicyclic dione 1 39 synthetic  are rich in functional groups, which permit further  manipulations  of the  molecule.  Different functional  groups at the 1- and 5-positions may be introduced by varying the structure of the initial dicarbonyl compound 1 37 and then by performing suitable synthetic manipulations on the R’ and R” groups. 25 at the 2-, Additional functional groups may be added selectively 4-, 6- and 8-positions by standard alkylation procedures or at the 3and 7-positions  via carbonyl group reactions.  It is possible to  introduce groups regioselectively at the 2- and 6- or the 2- and 8positions of the bisenol ether of the tetraester (e.g., 140) or of the dione 139.25  Also, as exemplified in Scheme D-8, selective  82 using potassium hydride and an electrophile (Mel, monoalkylations Eti, allyl iodide, etc.) have been performed on the bisenol ethers of tetra-t-butyltetraester  intermediates such  as  •82b 140  Subsequent  58 t-Bu 2 C0  t-BuO C  HO—OH  CH2,  Me0 (93%)  t-Bu 2 C0  C 2 t-BuO  140  141  KH, DMF, -60 to -50°C, Mel Me AcOH/HCI, MeO 87°C (82%)  o H  142  143 Scheme D-8  hydrolysis  and  decarboxylation  of the ester functions  of the  alkylated material,142, would provide the monoalkylated dione 143. One  restriction  on  the  versatility  of  the  Weiss-Cook  condensation reaction to produce diquinanes is that only the cis bicyclo[3.3.O]octane-3,7-dione  stereoisomer  is  produced.  The  preference for the formation of the cis isomer is reasonable since 83 for the reaction is based on a the presently accepted mechanism series of equilibria (Scheme D-9).  34 have Also, Boyd and coworkers  calculated that the trans isomer of bicyclo[3.3.O]octane is kcal/mole less stable than the cis isomer.  -6.5  Thus, even in the unlikely  event that any of the trans isomer were to be formed during the reaction,  it would rapidly  undergo a reverse reaction and eventually  59  R 2 CO 0 +  R’  R 2 CO  RK  HO CQR 137  CO R 2 145  144  146 H2O+4L  :0  + 144  148  147  H2OI R 2 CO  138  149  Scheme D-9 would be transformed into the thermodynamically more stable cis isomer. b for the Weiss-Cook condensation 3 aB 83 The proposed mechanism is thought to proceed as depicted  in Scheme D-9.  An aldol  condensation of one molecule of the mono-anion of the dialkyl 3oxoglutarate  (144) with one carbonyl group of a molecule of the  dicarbonyl compound 137 produces the hydroxy dione 145. A second  60  aldol reaction may generate the diol 146 which loses a molecule of water to produce the enone 147 (or possibly the dehydration may occur before the second aldol reaction and circumvent the did 146). Michael addition of a second molecule of the anion 144 to the enone 147 gives the hydroxy dione 148.  Loss of a molecule of water from  148 to give the enone 149, followed by a second Michael addition leads to the formation of the tetraester 138.  If appropriate R’ and  Oc the enone intermediate 8 R” groups are used (e.g. R’=R”=chexyl), 1 47 may be isolated as the reaction is sensitive to steric effects caused by the R’ and R” groups. ° 8  2.3.2. Application of the Weiss-Cook Condensation Reaction to the Synthesis of (±)-f3-Panasinsene (31).  2.3.2.1. Preparation of an enone 74. Form bond (R  Protect  R  H  150  74  If the Weiss-Cook approach were to be employed for the preparation 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 would  61  need to be protected to permit differentiation between the two carbonyls, and an appropriate R group would have to be introduced With regard to the  before the enone double bond was installed.  identity of the R group, it was thought that preparation of an enone 74 with R=methoxycarbonyl would provide several advantages (vide infra) in comparison with the preparation of an enone having the Thus, a synthetic sequence utilizing  ultimately desired R=methyl. the  monoalkylation  (Mel)  of  the  bisenol  ether  tetra-t-butyl  tetraester intermediate 140 to generate the methylated dione 143 (Scheme D-8) was not considered. Synthetically, a base-catalyzed Weiss-Cook condensation using glyoxal 44 and dimethyl 3-oxoglutarate 45, followed by an acidcatalyzed  hydrolysis!  decarboxylation  of  the  tetraester  84 intermediate (151) provided the known dione 43 (equation D-22). Me 2 CO  C 2 MeO  o+  +  /  Me 2 CO  C 2 MeO 45  NaOH,MeOHr 0  44  45  /HCI (1M), 151  /AcOH,  heat  (D-22)  Selective protection of one of the carbonyl functions in 43 to give the keto ketal 46 was achieved via an acid catalyzed reaction  62  of 43  with  2,2-dimethyl-1,3-propanediol  (152) by the method  reported by Moss and Piers (equation D-23). 85 ’ 38  Preparation of 46  usually involves a tedious chromatographic separation of 46 from the diketal 153 and the dione 43.  However, the purification was  simplified by loading the sample as a solid adsorbed on Celite onto the silica gel column and successively eluting compounds 153, 46 and 43 with diethyl ether-petroleum ether (2:1), diethyl ether-ethyl acetate (9:1) and ethyl acetate (neat), respectively.  The purified  keto ketal exhibited m.p. 46.5-47.5°C (literature 85 m.p. 48°C). H  46  (D-23)  +  43  152  153 +43  In order to transform the keto ketal 46 into the desired enone 154, it was necessary to introduce an appropriate R group and the enone double bond (equation D-24).  Depending on the identity of the  R group, formation of the double bond and the proposed conjugate addition reaction could be more or less expedited.  If R were a  methyl group, as is found  in the formal  in f3-panasinsene (31), then  63  (D-24)  46  154  dehydrogenation step to generate enone 154, the double bond theoretically could end up either exo- or endo- to the 5-membered ring (156 or 157, respectively, equation D-25).  Due to the strain  involved in introducing two new sp 2 centers into the five-membered ring to give 157, it was difficult to predict, a priori, whether or not the elimination of the selenoxide derived from 155 would generate the endo isomer 1 57 as the major product; however, others have noted the predominance of the endo isomer 86 for a variety of bicyclic lactones and 2,3-dialkylated cyclopentanones. less serious problem, was that the presence of the  Another,  methyl group on  the enone double bond would deactivate 87 the enone 157 towards the key copper (1)-catalyzed conjugate addition reaction of the Grignard  +  156  H  155  157  (D-25)  64  reagent 7.  On the other hand, if R were the methoxycarbonyl  function, there would be no possibility of the formation of an enone with an exocyclic double bond from the keto ester 158.  In addition,  a towards conjugate addition 78 the enone 159 would be activated reactions.  A disadvantage of utilizing R=methoxycarbonyl was that  at some stage during the synthesis, the methoxycarbonyl group would have to be deoxygenated to generate the corresponding methyl group.  However, deoxygenation reactions were already planned at  two stages of the synthesis (between 160 and 1 61  and between  162 and 31, relevant positions indicated, Scheme D-1O), so it was expected that at either point a double deoxygenation could be done. Consequently, it was decided that a methoxycarbonyl group rather than a methyl group would be employed as the R group in the synthesis of an enone 74 and that its deoxygenation would be performed in tandem with the deoxygenation of either the keto or the ester functions in 1 60 or 1 62, respectively.  7  158  159  65  R  160  161  I I  31  Scheme  ‘COMe 162 D-10  The keto ester ketal 1 58 (actually as a mixture of 1 63 and 1 64) was prepared via a modification of the procedure reported by Deslongchamps and coworkers. 88  Thus, a THE solution of the keto  ketal 46 was treated with potassium hydride, and the enolate anion thus generated was allowed to react with dimethyl carbonate to form, in -93% yield, an -.1.5:1 mixture ( H nmr analysis) of the keto 1 ester 163 and its ester enol tautomer 164 (equation D-26).  Due to  the fact that the mixture of 1 63 and 164  was not stable to  purification  material  by  flash  chromatography,  this  was  not  rigorously purified. The stereochemistry of the keto ester 163 was not proven, but was assumed to be that shown based on the following reasoning. Excess base present during the reaction would remove the proton  66  between the keto and ester functions to generate the corresponding enolate.  During the workup of the reaction, protonation of the  enolate would give a mixture of 1 63  and the epimer at C-2.  Equilibration of the mixture via the enol tautomer 164 would lead to the methoxycarbonyl group preferentially being on the sterically less congested, convex face 89 of the molecule. In the 1 H nmr spectrum of the crude mixture, the signals due to the methoxycarbonyl functions of 163 and 164 were displayed as singlets at 8 3.73 and 3.75, respectively, while the signal for the enol 0jL of 164 was displayed at 10.35.  In the high resolution mass  spectrum, the exact mass of the molecular ion of the mixture (1 63 and 164) was found to be 282.1459, which is consistent with the molecular formula 0 12 C 2 H . 5 5  Signals for fragments corresponding to  the 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 methoxycarbonyl containing intermediates which were prepared. C 2 MeO  H  46  H  163  1)KH,THF, 6O°C (2h) 2) (MeO) C0, 2 —60°C (1.5h)  +  HO  164  (D-26)  67  In order to generate the enone 159, a modification of the Reich  selenoxide syn elimination procedure developed by ° was employed. 9 coworkers  and  The required selenide was prepared by  treating a THF solution of a mixture of the keto ester 1 63 and its enol tautomer 164 with potassium hydride (1.3 equiv) and allowing the resultant enolate anion to react with benzeneselenenyl chloride (1 .35 equiv) at 0°C.  An -4:1 mixture (1 H nmr analysis, using the  ratio of the signals of the methoxy groups) of the epimeric selenides 165 and 166 was obtained in 86% yield. Ph  1) KH, THE, H  163 and 164  165 -  rt (40mm) 2) PhSeCI, 0°C (20mm)  1  166  The two epimers 165 and 166 could be distinguished readily by H nmr spectra of the mixture. three main signals in the 1  Thus, the  signals for the tertiary methyl groups of the major epimer 165 appeared at 8 0.89 and 0.97, while those of the minor epimer 1 66 were at 0.92 and 0.99.  The resonances due to the methoxycarbonyl  functions were at 3 3.71 and 3.51, respectively.  Also, signals for  68  aromatic hydrogens at 37.53-7.57 and 7.63-7.65 were characteristic of 165 and 166, respectively.  A small amount of the minor epimer  (166) was obtained in pUre form and the expected molecular C 2 H 8 5 O 1 was confirmed by the presence of an ion with 6 0 formula, Se, a mass of 438.0940 mass units in the high resolution mass spectrum. The oxidation of the mixture of the selenides 165 and 166 to the corresponding selenoxides had to be done with care as 2alkoxycarbonyl-2-cyclopenten-1 -ones  are  sensitive  catalyzed epoxidation by hydrogen peroxide. ° 9  to  base  In the case of 165  and 166, only the major epimer 165 would be able to undergo the normal selenoxide syn elimination. 91  However, because compounds  165 and 166 could not be separated by chromatography (silica gel), the minor epimer 166 was also present during the oxidation to give the corresponding selenoxide.  Thus, a dichloromethane solution of  the mixture of the epimeric selenides was treated with 15% aqueous hydrogen peroxide (2.1 equiv) at 0°C (10 mm) and room temperature (20 mm) (equation D-28).  A ‘H nmr spectrum of the crude product  indicated unexpectedly that the product obtained was quite pure with just traces of aromatic compounds present. H  ,2 0 2 H CI CH , (D-28)  0°C (10mm); rt (20mm) 165  159  69  A control experiment was done to try to determine the fate of the minor selenide epimer 166.  Thus, a sample of 166 with less  than 5% of the major epimer 165 present (H nmr analysis), but which contained some other impurities, was oxidized using our normal procedure.  A H nmr spectrum of the crude product showed  the presence in the mixture of the enone 159 and small amounts of other compounds having aromatic, methoxycarbonyl and/or ketal functions present.  The presence of enone 159 was surprising given  that the normal syn elimination 91 (as shown in formula 1 67) of the selenoxide cannot occur in compound 168.  There have been reports  in the literature of the selenoxide elimination occurring in cases where the selenoxide and the J3—proton were anti to each other in the starting material used for the reaction (for example, 9 (phenylselen enyl)-3-alkyl 2 cyclopentanones).  trans-2-  However, in such  cases, there was a proton a- to the selenoxide function.  Thus, it  was believed that an in situ epimerization of the selenoxide function generated occurred to give the syn arrangement of the selenoxide and the 13-proton and that then the normal elimination reaction took place.  In the case of the selenoxide 168, a simple epimerization is  impossible, so a more complex process may be occurring. the mixture  that was  obtained  from the  oxidation of  H  167  168  However,  166 was not  70  characterized further.  The results of the oxidation of the selenide  166 indicated that the presence of 166  during the oxidation!  selenoxide elimination reaction of the major selenide epimer (165) would not be a significant problem. 10  9  C 2 MeO  Hb  Hb  159 The enone 1 59 obtained from the selenoxide elimination was not stable to flash chromatography, so it was characterized without rigorous purification. In the 1 H nmr spectrum, the signals for all the hydrogens of 159 were assigned based on decoupling experiments and on the observ 93 ations that in bicyclo[3.3.O]octanones, hydrogens on the convex face are deshielded relative to those on the concave face and vicinal cis couplings between hydrogens a- to a keto function and the angular hydrogen are larger than the corresponding trans couplings.  Thus, the multiplet for the angular hydrogen (H-5b) at S 3.14-3.26 was coupled with the indicated coupling constants to the signals at 2.25 (J  (J  =  =  4.0 Hz, H-4a), 2.79 (J  12.5 Hz, H-6a) and 2.70 (J  =  =  6.5 Hz, H-4b), 1.48  8.0 Hz, H-6b).  The geminal  couplings for H-4a and H-4b and for H-6a and H-6b were 18.0 Hz and 12.5 Hz, respectively. The signal for the hydrogens at the 8-position was a broad singlet at 3.30 (2H).  The methyl ester signal was at 6  71  3.85, while the resonances due to the tertiary methyl groups were at 0.94 and 1.09 (s, 3H each).  Two carbonyl absorptions were observed  at  in the ir spectrum and were due,  1750  and  1719  cm  respectively, to the cyclopentenone and ester carbonyl stretches. 61 The molecular formula, 5 0 1 C 2 H , 5 was consistent with the ion found 0 at 280.1311 in the high resolution mass spectrum. The  key  enone  intermediate  159,  representative  of the  generalized structure 74, (see p. 60) was one of the subtargets in the synthesis of (±)-J3-panasinsene (31). Also, the enone 159 appeared to be more suitable for the key methylenecyclohexane annulation sequence than the previously synthesized enones 128 and 129 (Scheme D-7, p. 51) had proved to be.  2.3.2.2. Methylenecyclohexane Annulation on the Enone 159.  The vinylstannane 5, prepared by a modificationb of previously reported procedures, 5  was dissolved in THE and  -78°C with a solution of methyllithium.  transmetallated at  The resulting vinyllithium  species 6 was transformed into the corresponding Grignard reagent 7 by the addition of solid magnesium bromide etherate. successive addition of the copper (I)  bromide-dimethy)  The sulfide  catalyst (0.25 equiv) and a THE solution of the enone 159 gave an orange suspension which was stirred at -78°C for 25 minutes.  After  an appropriate workup, a mixture of the keto ester chloride 169 and its enol tautomer 170 was obtained in 94% yield (equation D-29). was gratifying  and not unexpected to find that the  It  enone 159 was  72  3 SnMe  MgBr  very reactive towards the conjugate addition reaction of the Grignard reagent 7.  As seen earlier, similar reactions of enones  having tn- or tetrasubstituted double bonds were more sluggish and required either longer reaction times (see equation D-18) or the presence of additives (see Schemes I-i and 1-3).  1) 7, CuBrSMe , 2  169  (D-29)  THF, -78°C (25mm)  159  CI, rt 4 2) aq NH HO 170  The crude product mixture of 1 69 and 1 70 was not stable to purification by flash chromatography.  Therefore, apart from a rapid  filtration through a short silica gel column to remove inorganic and  73  very  polar  purification.  organic  material,  it  was  characterized  without  A strong absorption characteristic of an enolized f3-  1 in the ir spectrum. 61 was displayed at 1657 cmdicarbonyl group 1 were The keto and ester carbonyl stretches at 1754 and 1722 cmrelatively weak, which predominated.  indicated that the enol tautomer 1 7 0  ) also indicated that the 3 The 1 H nmr spectrum (CDCI  enol 170 was the predominant component of the mixture.  Thus, the  signal for the enol OH (br s at 3 10.79) integrated for -‘1 hydrogen. That the conjugate addition  had  been  performed was further  H nmr spectrum of signals due to confirmed by the presence in the 1 the 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 step  in the methylenecyclohexane annulation, namely the cyclization of the keto chloride to obtain a tricyclic ketone, was performed. Surprisingly, the cyclization was problematic and, before success was achieved, a variety of methods were tested (i.e., KH/THF/45°C; C0 32 K / C0 2-butanone! 80°C; 3 2 K 1 94 3 C0 acetone! 50°C; 2 K / 3 pentanone! -100°C; one-pot conjugate addition! cyclizations in the ). 95 presence of HMPA  Ultimately, it was found that the cyclization  of the crude keto ester chloride 1 69/170 worked best in hot 96 (5 equiv) as the base (-‘60°C) acetonitrile using cesium carbonate (equation D-30).  The product thus generated was more pure and was  produced in a better yield than that obtained from the other procedures.  After  chromatographic  purification  recrystallization, the tricyclic keto ester ketal 171  and  a  was obtained  as colorless crystals in 64% yield and exhibited m.p. 127.5-128.5°C.  74  3 C 2 Cs , 0 CH CN, 0 3 —60°C (20h)  (D-30)  169 and 170  171  The tricyclic keto ester ketal 171 was further characterized by ir spectroscopy, mass spectrometry (high and low resolution), an elemental analysis, and nmr spectroscopy ( H and 13 1 C, one and two 97 dimensional experiments).  Thus, in the ir spectrum, very strong  absorptions were observed at 1745 and 1723 1 cm- due to the keto and ester carbonyl stretches, respectively, and at 1115 cm due to the ketal C-C bonds.  Weak absorptions due to the exocyclic olefinic  methylene appeared at 1637 and 893 cm . 1  The high resolution mass  spectrum indicated that the molecular ion had a mass of 348.1929 mass  units  0 C 2 H . 5 0 8  which In  is  the  consistent  with  low resolution  the  molecular  formula,  mass spectrum, fragments  corresponding to the loss of MeC (M-31) and CO Me (M-59) were 2 observed. The  The analytical data also was in keeping with the formula. nmr spectrum of 171 (see figure 2) showed the expected  three singlets (3H, each) due to the tertiary methyl groups (Me-19 and 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 were  found 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)-13 in the various tricyclic intermediates synthesized ultimately becomes the corresponding methyl group (Me-13) in the synthetic $-panasinsene (31).  75  TMS  I  I  III  II  II  I  5  III  II  4  I  11111  III  3  I  II  II  II  1111111  1  2  II  P  OPPM  Figure 2. The 300 MHz 1 H 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 aliphatic  hydrogen signals also were displayed in the spectrum and are listed together with the  COSY ( H-’H COrrelation 1  correlations in Table 2.  ipectroscop..)  76 Hb  Ha Hb’ Me 20  0:  Hb  Hb  171 From a 13 C nmr APT (A.ttached Eroton jest)  experiment  performed on 171, the signals for the carbons corresponding to the methyl and methine groups could be assigned as follows; 822.23 and 22.32, the tertiary methyl groups (Q..H -19 and H 3 -20); 52.1, the 3 methyl ester 3 (.H 14’); and 38.65, the methine (H-4).  Some of the  signals due to the other carbons could be assigned based on their chemical shifts. Thus, the signal for the olefinic methylene (H 2 15) resonated at 3 112.7; the signal for the quaternary olefinic carbon (C-li) was at 145.1; and the signals for the keto (C-6) and ester (C-14) carbonyl groups were at 211.8 and 170.5, respectively. The rest of the carbon signals along with the HETCOR C 3 H 1 ( -’ , HETeronuclear Shift CORrelation) correlations are listed in Table 1. Based on the HETCOR and COSY data, all the hydrogen signals in the ‘H nmr spectrum and most of the carbon signals in the ‘ C nmr 3 spectrum could be reasonably assigned.  Thus, a HETCOR experiment  (see figure 3) confirmed the above ‘H and ‘ C assignments and 3 permitted the determination of the pairs of geminal hydrogens. For example, the carbon signal at 842.8 (H -2) correlated with 2 hydrogen signals at 1.88 (d, 1H, J  =  16.0 Hz, H-2a) and 2.95 (br d, 1H,  77  J  16.0 Hz, H-2b); the carbon signal at 38.74 (H -3) correlated 2  =  with hydrogen signals at 1.96 (distorted dd, 1H, J 3a) and 2.10 (distorted d, 1H, J  =  14.0, 6.0 Hz, H  14.0 Hz, H-3b); and the carbon correlated with only the 2-hydrogen  signal at 32.4 2 (...H 10)  =  multiplet at 2.26-2.39 (H-lOa and H-lob).  The other pairs of  geminal hydrogens were assigned in a similar manner, but the determination of which pair was at which position was based on the observed COSY correlations.  (PPM)  F2  140 130 120 110 100 90 80  CDCI3  70 60 50 40 30 20 I  I  5.0  4.5  4.0  I  I  I  I  I  I  3.5  3.0  2.5  2.0  1.5  1.0  O. ( 5 PPM)  Figure 3. The 75 MHz HETCOR spectrum of the keto ketal ester 171.  78  Table 1: The 75 MHz HETCOR Data for the Keto Ester Ketal 171. Hb Ha  19  Me Me 20  171 Position (C-x) 1 2 3 4 5 6 7 8 9 10 11 13 14 14’ 15 1 6/1 8 17 1 9/20  H (300 MHz) 5 ppm (H-x) 1  C (75 MHz) 13 Sppm 58.2a 42.8 38.74 38.65 39.9 211.8 68.Oa 30.6 23.8 32.4 145.1 109.0 170.5 52.1 112.7 71.8/72.4 30.0 22.23/22.32  -  -  1.88 (2a); 2.95 (2b) 1.96 (3a); 2.10 (3b) 2.76-2.84 (4b) 2.54 (5a or 5b); 2.69 (5b or 5a) -  -  -  -  1.54-1.71 (8a or 8b); 2.15-2.21 (8b or 8a) 1.72-1.80 (9a or 9b); 1.54-1.71 (9b or 9a) 2.26-2.39 (lOa and lOb)  a. Assignments may be interchanged.  -  -  -  -  -  -  3.74 (3H-14’) 5.07 (15a); 5.01 (15b) 3.40-3.57 (2H-1 6/2H-1 8) -  0.91/1.01  -  (3H-1 9/3H-20)  79  In the COSY spectrum of the keto ester ketal 171 (see figure 4), the spin system of the six-membered ring was separable from interrelated ones of the two five-membered rings.  Thus, key entry  points into the two main spin systems were the signals for the olefinic 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 system and the signal for the methine hydrogen at 2.76-2.84 (m, 1H, H-4b) which led into the five-membered ring systems.  The process  followed in making the assignments of the hydrogen positions by use of the COSY correlations is exemplified for the five-membered ring systems.  The signal for the methine hydrogen (H-4b) at 62.76-2.84  showed strong correlations (large cross peaks) to the distorted doublets at 1.96 (1H), 2.54 (1H),  2.69 (1H)  and weaker correlations  to the signals at 2.10 (distorted d, 1H) and at 2.95 (br d, 1H).  From  the 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 of geminal  hydrogens.  Since the signal at 62.95 also showed a weak  correlation to the signal for the olefinic hydrogen at 5.07 (H-15a), it seemed reasonable to assign the signal at 2.95 to H-2b and thus, the one at 1.88 to H-2a.  The signals at 6 2.54 and 2.64 were then  assigned 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 a and b-hydrogens based largely on reported observations of the magnitudes of relevant coupling constants in other systems. 93 The pairs of geminal hydrogens on the six-membered ring were assigned  80  to the appropriate positions starting from the cross peaks shown by the two olefinic hydrogens at 85.01 (H-15b) and 5.07 (H-15a).  ---  1.0 1.5  ,  —-——  —-—-——  -———  2.0  e  2.5 9 ..N —I  3.0  ——  3.5  —--  .+. 4.0 4.5 —  5.0  4.5  4.0  3.5  2.5  2.0  1.5  1.0  5.5 PPM  Figure 4. The 400 MHz COSY spectrum of the keto ester ketal 171.  81  Table 2: The 400 MHz COSY Data for the Keto Ester Ketal 171. Hb Ha Hb  19  Me  Hb  Hb  171  Position (H-x) 2a 2b 3a 3b 4b 5a or 5b 5b or 5a 8a or 8b 8b or 8a 9a or 9b 9b or 9a lOa and lOb 14’ 15a 15b  COSY Correlations Signal 8 ppm (multiplicity;a J; (H-x) number_of_H) 1.88 (d; 16.0; 1H) 2b 2.95 (br d; 16.0; 1H) 2a; 4b (W-coupling); l5ab 1.96 (distorted dd; 14.0, 3b; 4b 6.0; 1H) 2.10 (distorted d; 14.0; 3a; 4bb 1H) 2.76-2.84 (m; 1H) 3a; 3bb; 5b; 5a; 2b (W-coupling) 2.54 (distorted dd; 19.5, 5b or 5a; 4b 9.0; 1H) 2.69 (distorted dd; 19.5, 5a or 5b; 4b 9.0; 1H) 1.54-1.71 (m; 2H) 8b or Ba; 9b and 9a 2.15-2.21 (m; 1H) 8a or 8b; 9a and 9b; lOa or lOb (W-coupling) 1.72-1.80 (m; 1H) 9b or 9a; 8b or 8a; lOa and lOb 1.54-1.71 (m; 2H) 9a or 9b; Bb or 8a; lOa and lOb 2.26-2.39 (m; 2H) 9a and 9b; 8b or 8a (W coupling); 15a and 15b 3.74 (s; 3H) 5.07 (s; 1H) 15b; lOa or lObb; 2bb 5.01 (d; 1.0; 1H) 15a; lOa or lOb -  -  82  Table 2: continued. 16 and 3.40-3.57 (m; 4H) 18 19 or 20 0.91 (s; 3H) 20 or 19 1.01 (s; 3H)  20 or 19 (W-coupling) -  -  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 the goals of the project, that is, the synthesis of an enone with a tetrasubstituted double bond and its .utilization in the methylene cyclohexane annulation sequence.  It remained only to carry out  various functional group manipulations to arrive at the projected intermediate 72 on which to perform a ring contraction to assemble the correct tricyclic carbon framework of the natural product.  2.3.2.3.  Preparation  of  a  Substrate  (Intermediate 72) for Ring  Contraction.  In order to prepare the desired ketone substrate 72 for the planned ring contraction sequence, it was necessary to deoxygenate the keto group (and perhaps the methoxycarbonyl group) in 171 and to convert the ketal group into a keto function.  83  A double deoxygenation of the keto and ester functions to give the corresponding methylene and methyl groups was attempted first. Consequently, the keto and methoxycarbonyl groups of the keto ester ketal 171 were reduced with lithium aluminum hydride in diethyl ether to give a mixture of the epimeric diols 172 and 173 (ratio H nmr spectral analysis) in 80varied from -‘5:3 to >4:.<1 172:173, 1 95% yield (equation D-31).  In the 1 H nmr spectrum (300 MHz, CDCI ) 3  of a sample containing both 172 and 173, the signals for their respective 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), which  were converted to a triplet (J upon the addition of D 0. 2  =  -‘8 Hz) and a less complex multiplet  In the mixture, the signals for the CE OH 2  hydrogens appeared at 8 3.57-3.63 and 3.76-3.90 (m, m, 2H total).  LiAIH Et , 4 0, 2 0°C (0.5h); rt (2h)  172  (D-31)  +  171  H  173  The major epimer (172), which could be separated from the minor epimer by recrystallization of the mixture from 4:1 and then  84  from 7:1 petroleum ether-diethyl ether, exhibited m.p. 117-118°C. An ir spectrum of 172  (solution  in  chloroform,  polystyrene  reference) showed absorptions due to the hydroxyl group at 3630, the ketal group at 1125, and the olefin function at 1640 cm . In a 1 ‘H nmr spectrum (300 MHz, CDCI ) of the major epimer (172), the 3 signals for the methylene hydrogens of the hydroxymethyl group appeared 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 of 0.* Further confirmation of the identity of diol 172 was obtained 2 D from the mass spectra.  Thus, in the high resolution mass spectrum,  the molecular ion peak was found at 322.2136, which corresponds to the molecular formula, 0 10 C 3 H , 4 9 while in the low resolution mass spectrum, peaks corresponding to the loss of one and two molecules of water (M-18 and M-36, respectively) were observed. It was of interest to determine the stereochemistry at the 6position of the diol 172. A priori, one would have trouble predicting the stereochemical outcome of the reduction process, since both faces of the keto function in 171 are sterically hindered.  In key nOe  difference experiments (400 MHz, summarized in structure 172’) to determine the stereochemistry of 172, irradiation of the signal at 4.57 (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 enhancements of the signals at 3.79 (H-14b) and 4.57 (H-6a). *  Thus, it seemed  Signals due to the other hydrogens appeared in the 1 H 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), 2.262.40 (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).  85  reasonable to conclude that H-6a was cis to the hydroxymethyl group and that the stereochemistry at C-6 of the major product 172 was as assigned.  H  172  Various attempts to carry out the double deoxygenation reaction of both hydroxyl functions of 172 and 173 via derivatives of the diols were unsuccessful. super  hydride  Krishnamurthy and Brown have reported a  (lithium  triethylborohydride)  deoxygenation  98 utilizing the p-toluenesulfonate derivatives of alcohols. procedure Attempts  to  convert  the  diol  mixture  to  the  corresponding  bissulfonates by reaction with p-toluenesulfonyl chloride (p-TsCl, 3 equiv  or  2.2  equiv)  in  the  presence  of  a  base  (4-(N, N  dimethylamino)pyridine (DMAP), 3.5 equiv or KH, 3 equiv) in a suitable solvent (dichloromethane or THE) were unsuccessful when an unstable mixture of the mono- (mainly the primary sulfonate) and bissulfonates, unreacted diol and other uncharacterized materials was produced.  The sterically hindered nature of the secondary  alcohol was also implicated when the reaction of an acetonitrile solution of the diol 172 with phenoxythiocarbonyl chloride (PTC-Cl,  86  2.3 equiv) in the presence of DMAP (4.2 equiv) generated a complex mixture  of  products.  The  failure  to  form  the  bis-O  -  phenoxythiocarbonyl derivative of the diol 172 in a reasonable yield thwarted plans to employ the Robins radical deoxygenation reaction with tri-n-butyltin  99 hydride.  Sterically hindered alcohols have been converted into their  N,N,N’, N’-tetramethylphosphorodiamidate derivatives by a procedure ° 10 developed by Liu and coworkers. from the did  Thus, the bisalkoxide prepared  172 by its reaction with n-butyllithium was treated  with N,N-dimethylphosphoramidic dimethylamine (equation D-32).  dichloride  (174) followed by  A mixture of products resulted in  which the cyclic N,N-dimethylphosphoramidate  175 predominated.  The structure of 175 was deduced from its 1 H  nmr  spectrum  wherein 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  =  10  Hz, overlapped with a multiplet, 1H, at 2.69-2.9) and the tertiary methyl 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/ 12 but the results indicated (‘H nmr spectral 0°C) of 175 failed, analysis on the crude mixture) that the derivatized secondary hydroxyl function had been deoxygenated in preference to the primary hydroxyl group to generate a compound formulated as 176.  Other hydrogen signals were also observed: the methine hydrogen (H-6a) at o 5.09 (br t, 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 3.52 and 3.42 (2H each); the rest of the aliphatic hydrogens (12H) as multiplets at 1.4-2.4.  87  1) n-BuLl (2.5 equiv), TMEDA:DME (1:4), 0. 0°C (20-30mm) NP 2 2) 2 NP(0)C1 Me (Me) , (174, —10 equiv) 0°C rt (—19h) NH (—25 molar 2 3) Me equiv), 0°C (4h) —  172  (D-32)  H  175 and other products  0 176  It was apparent that a sequential deoxygenation of the keto and ester functions would be preferable as the reactivities of the primary and secondary hydroxyl groups were significantly different. Consequently, it was decided to postpone the deoxygenation of the methoxycarbonyl  group  until  the  end  of the  synthesis  and  to  deoxygenate it together with another methoxycarbonyl which would  be present at that stage.  Thus, the keto function in 171  was  reduced selectively with sodium borohydride in methanol (with a small amount of ethyl acetate to solubilize the keto ester ketal 171).  An ‘-2-3:1 mixture of the epimeric alcohols 177 and 178 was  produced, which proved to be difficult to separate by chromato graphy  on  silica  gel  or  by  fractional  recrystallization.  For  subsequent reactions it was desirable to have exclusively the major epimer (177) present (vide infra), so a variety of reducing agents  88  were tested to see if the ratio of epimers could be improved in favour of the major epimer.  Reduction procedures attempted  included: zinc borohydride in diethyl ether, which is known to reduce via chelation; 101 L-selectride in THF, which is expected to approach from the less hindered face of the carbonyl; 102  and  sodium  borohydride with varying amounts of cerium trichloride hexahydrate at  various  temperatures  in  methanol  (Luche’s  1 reagent).  03  Reductions using Luche’s reagent were the most consistent in terms of selectivity and yield.  Therefore, treatment of a methanolic  solution of 171 with sodium borohydride (1.3 equiv) in the presence cerium 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 mixture  of alcohol esters using Luche’s reagent was usually -5:1 with a ratio as good as 12:1  obtained upon occasion depending on small  variations in the reaction conditions.  (The ratios were based on the  relative integration area for the signals due to H-6a versus H-6b in the  1  H nmr spectrum).  Recrystallization of the mixture from ethyl  acetate-hexane generally improved the ratio slightly.  89  C1 2 MeO  HOi  NaBH , CeCI 4 3 6H 0 2  0:  MeOH, -48°C (lh)  177  (D-33)  +  171  H 178  The ir spectrum of an ‘-12:1 mixture of 177 and 178 displayed a strong absorption at 3511 cm 1 indicative of the presence of the 1 was alcohol function, while a very strong band at 1726 cm characteristic of the ester carbonyl group.  In the  1  H nmr spectrum  of the mixture, the signal due to the methoxycarbonyl group of both epimers appeared at 83.68 (s, 3H).  The carbinol hydrogen (H-6a)  adjacent to the hydroxyl group of 1 77 resonated as a triplet of doublets (J  =  9.0, 2.0 Hz) at 84.69 which simplified to a triplet (J  0. 2 9.0 Hz) upon the addition of D  =  On the other hand, H-6b of the  epimer 178 appeared as a multiplet at 84.11-4.14, which simplified 0. 2 somewhat upon the addition of D  The elemental analysis and high  resolution mass spectral analysis (performed on the mixture of 177 and 178) provided results consistent with the molecular formula, 20 C 3 H . 5 0 The stereochemistry at the 6-position was not determined for either the alcohol ester 177 or the alcohol ester 178.  However,  90  assuming  the  that  relative  chemical  shifts  and  the  signal  multiplicities of the hydrogen at the 6-position of the diol epimers 172 and 173 (vide supra, p. 83) and those of the alcohol esters 177 and 178 are comparable, then the relative stereochemistries of the alcohol esters 177 and 178 may be assigned.  Thus, the resonance of  the 6-hydrogen of the major diol epimer 172 was downfield from that of the 6-hydrogen in the minor epimer 173 and the signals appeared as a distorted triplet of doublets (J multiplet, respectively.  8, —4 Hz) and a  =  Similarly, the chemical shift of 6-hydrogen  of the major alcohol ester epimer (177) appeared downfield from that of the minor epimer 1 78 and the signals were a triplet of doublets (J  =  9.0, 2.0 Hz) and a multiplet, respectively.  Thus, it  seemed reasonable that compound 177 had the same configuration at the 6-position as did the diol 172.  Further evidence for the  depicted stereochemistry came from the fact that the epimer 177 was the major product in reductions using zinc borohydride, a reagent  which  tends  give  to  products  from  reduction  via  101 In the case of keto ester ketal 171, reduction via chelation. chelation with the ester function  would  be  expected to give  preferentially an alcohol possessing the configuration found at the 6-position of the compound 177. The deoxygenation of the secondary alcohol function in the alcohol esters 177 and 178 deoxygenation derivative  (PTC  of  the  corresponding  derivative)  procedure developed  was performed  by  prepared  Robins  using the radical  O-phenoxythiocarbonyl  via a modification of the  and coworkers. 99  The typical  procedures utilized by Robins and coworkers to convert secondary  91  alcohol groups into their PTC derivatives were to stir the alcohol with phenoxythiocarbonyl chloride (PTC-Cl) either in the presence of pyridine (3-4 equiv) in dichloromethane (2 hours) or, for more hindered alcohols, in the presence of 4-(N,N-dimethylamino)pyridine (DMAP, 2 equiv) in acetonitrile at room temperature (16 hours).  In  the most sluggish case they reported, 6-9 equivalents of DMAP were required.  The PTC derivatives were then deoxygenated using tn-n  butyltin hydride in the presence of a free radical initiator in warm toluene (75°C or at reflux, 3 hours). The  conditions  utilized  for  the  preparation  of  the  PTC  derivatives of alcohol esters 177 and 178 were more drastic than the most severe case reported by Robins and coworkers. 99 Thus, a solution of a mixture of 177 and 178 (ratio -5:1) in acetonitrile was converted into a mixture of the corresponding PTC derivatives by treatment of the former material with PTC-Cl (1 .5 equiv) in the presence of DMAP (8 equiv) at -70°C for 20 hours (equation D-34). The major epimer 179 was obtained in 76% yield, while the minor epimer 180, which was difficult to separate from an impurity, was not characterized nor used in further reactions.  It was due to the  difficulty in purifying the minor epimer that the effort was made to obtain a good stereoselectivity in the reduction of the keto ester ketal 171 to the alcohol esters 177 and 178.  92  PhOC(S)-CI P T DMAP, CH CN, 3 -70°C (20h)  H H  (D-34) H  177 and 178  179 and small amount of epimer  (180)  The ir spectrum of the phenyl thionocarbonate 179 displayed absorptions due to the carbonyl group at 1736 cm, the exocyclic 1 methylene group at 1639 and 888 and a monosubstituted aromatic ring at 774 and 690 cm* In the ‘H nmr spectrum, the signal for the methoxycarbonyl group appeared as a singlet at 8 3.70 (3H), the signal for H-6a was a triplet at 6.08 (J  =  8.5 Hz, 1H), and the  aromatic hydrogens gave rise to multiplets at 7.09-7.12 (2H), 7.257.30 (1H) and 7.38-7.43 (2H).  The elemental analysis and high  resolution mass spectral data were in accord with a molecular formula of 6 0 2 C 3 H S 7 4 . The PTC derivative 1 79 was deoxygenated according to a modification of the procedure reported by Robins and coworkers. Thus, a solution of 179 in benzene (instead of toluene) was treated with tri-n-butyltin  hydride (2.5 equiv) and the radical initiator,  2,2’-azobisisobutyronitrile (AIBN, 0.18 equiv), and was heated under an argon atmosphere at 77°C for 20 hours (equation D-35). purified deoxygenated product 181 was obtained in 72% yield.  The  93  n-Bu S 3 nH,  (D-35)  AIBN, PhH —77°C (20h)  179  181 19  ‘20 16  H  181 Recrystallized  181  (m.p. 87.5-89.5°C) showed the expected  absorptions in the ir spectrum for the ester carbonyl at 1719, for the ketal C-O at 1116 and for the exocycNc methylene group at 1638 and 892 cm . The H nmr spectrum, as expected, still showed the 1 presence of signals for the tertiary methyl groups as singlets at 6 0.95 and 0.97 (3H, each), the methoxycarbonyl function as a singlet at 3.63 (3H) and the olefinic hydrogens as singlets at 4.87 and 4.92 (1H each).  The 13 C nmr displayed the required 20 signals, including  one for the ester carbonyl carbon (C-14) at 3 176.3, the quaternary olefinic carbon (C-il) at 149.1 and the olefinic methylene 2 (.H 15) at 109.8.  From an APT experiment, it was possible to assign the  tertiary methyl groups to the signals at 322.3 and 22.4 3 (.H 19 and ,H 3 20), the methyl of the methoxycarbonyl group to the signal at -i4’), and the methine (H-4) to the signal at 45.0. 3 51.2 (H  The  results from the high resolution mass spectrum (found M=334.2141)  94  and the elemental formula  analysis were consistent with a  20 C H  thus  further  confirming  that  molecular the  phenoxythiocarbonyl group had been replaced with a hydrogen.  0In the  low resolution mass spectrum the molecular ion and a fragment due to the loss of the methoxycarbonyl group (M-59) were among the peaks observed. The keto ester substrate for the ring contraction reaction sequence was prepared by deprotection of the ketal function in the ketal ester 181.  A procedure similar to the one reported by Moss  and 38 Piers 8 ’ 5 was utilized.  Thus, an acetone solution of the keto  ketal 181 was treated with iN hydrochloric acid (0.5 equiv) at room temperature for 5.5 hours (equation D-36).  After an appropriate  workup and purification, the keto ester 182 was obtained in -85% yield (equation D-36).  C 2 MeO 1 N HCI (0.5 equiv  0 (D-36)  acetone, rt (5.5h) H  181  182  H nmr and ‘ The 1 C nmr spectra confirmed that the ketal group 3 of 181 had been removed.  Thus, the signals for the tertiary methyl  groups and the ketal methylene groups were not seen in the 1 H nmr spectrum and the rest of the signals were consistent with the structure of 182.  Similarly, in the 13 C nmr spectrum the ketal  carbon signals were replaced  by a ketone carbonyl  carbon signal (C-  95  13) at 6219.2 and the expected 19 other carbon signals required by Thus, the ester carbonyl carbon  the structure of 182 were present.  signal (C-14) resonated at 6 175.7, the olefinic methylene carbon 15) was at 109.8 and the quaternary olefinic carbon (.H signal 2 Based on a ‘ C nmr APT 3  resonance (C-li) appeared at 148.4.  experiment, signals at 642.1 and 51.6 could be assigned to H-4 and i4’, respectively. H 3  Additional confirmation of the molecular  0 was provided by the low and high resolution mass 0 2 H 5 C, , formula, 3 spectra as well as an elemental analysis.  182  Further details of the molecular structure were obtained from the one and two dimensional ‘H nmr spectra and from decoupling experiments.  Thus, in the one dimensional ‘H nmr spectrum (see  figure 5), the signal due to the bridgehead methine (H-4b) was readily assigned to the multiplet at 6 2.84-2.92 (iH), the signal for the methoxycarbonyl group (Me-i4’) appeared at 3.64 (s, 3H), and the signals 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)  96  F  TMS  5  4  3  PPM  1  2  Figure 5. The 300 MHz 1 H nmr spectrum of the keto ester 182.  simplified the doublet of doublets at 2.73 (J doublet (J  =  =  19.0, 1.0 Hz, 1H) to a  19 Hz), sharpened the multiplet at 2.25-2.42 (5H) and  simplified the multiplet at 1.38-1.50 (1H).  Irradiation of the signal  at 32.73 simplified the doublet of doublets at 2.08 (J 1H) to a doublet (J  =  =  19.0, 1.0 Hz,  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 of  doublets at 32.08 sharpened the two doublets of doublets at 2.19 and 2.73 and simplified the multiplet at 2.25-2.42.  It was clear,  97  therefore, that the signals at 82.73 and 2.19 were due to geminal hydrogens.  However, neither signal showed a large enough coupling  to H-4b to be due to hydrogens at the 3-position, hence the signals at 82.73 and 2.19 were caused by hydrogens at the 2-position (H-2b and H-2a, respectively).  The signal at 8 2.08 also did not show  coupling to H-4b, but it and the one due to H-4b were coupled to part of the multiplet (5H) at 2.25-2.42.  Consequently, the assumption  that vicinal cis couplings between hydrogens a- to a keto function and the angular hydrogen are larger than the corresponding trans 93 (see also p. 69) led to the assignment of the signal due couplings to H-3a to 62.08 and the signal due to H-3b to the multiplet at 2.252.42. The two dimensional 1 H- COSY nmr spectrum (400 MHz) H confirmed the above assignments and made it possible to assign tentatively most of the other hydrogen signals (see figure 6 and Table 3).  As in the case of the COSY spectrum of the keto ester  ketal 171, the signals for the bridgehead methine (H-4b) and the olefinic hydrogens (H-15a and H-15b) of 182 provided key entries into the two spin systems of the five-membered rings and of the six-membered ring, respectively.  Thus, the signal assigned to H-4b  at 8 2.84-2.92 showed correlations to the signals corresponding to H-2b (at 2.73, long range coupling), H-3b and H-5b (both part of the multiplet at 2.25-2.42) and H-5a (at 1.38-1.50).  In turn, the signal  at 6 1 .38-1 .50 (H-5a) showed other correlations to the signals at 1.84 (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, three were assignable to H-3b, H-5b, and H-6a or H-6b.  The signals for  98  the olefinic hydrogens at 34.65 (H-15a) and at 4.87 (H-15b) also showed correlations to the multiplet at 2.25-2.42, so H-lOa and/or H-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 assigned to the hydrogens of the six-membered ring, but specific assignments were not made.  The partial assignments are listed in Table 3.  , ,i  -  :: I  1.5  2.0  ::  I  2.5  :: e  r  3.0  3.5  p. 4.0  4.5 .p  -  -  5.0  4.5  4.0  3.5  3.0  2.5  5.0 2.0  1.5  Figure 6. The 400 MHz COSY spectrum of the keto ester 182.  PPM  99 Table 3: The 400 MHz COSY Data for the Keto Ester 182.  0  182 .  Position  2a 2b  Signal S ppm (multiplicity;a J; number of_H) 2.19 (dd; 19.0, 0.5; 1H) 2.73 (dd; 19.0, 10; 1H)  3a  2.08 (dd; 19.0, 1.0; 1H)  (H-x)  COSY Correlations (H-x) 2b; 3a (W-coupling) 2a; 3b (W-coupling)b; 3a (W coupling); 4b (W-coupling) 3bb; 2a (W-coupling); 2b (W coupling)  C 2.25-2.42 (m; 5H) 3b 4b 2.84-2.92 (m; 1H) 2b (W-coupling); 3bb; 5a; 5bb 5bb; 4b; 6a; 6bb 1.38-1.50 (m; 1H) 5a C 5b 2.25-2.42 (m; 5H) 1.84 (br if; 12.5, 3.5; 1H) 6a or 6b 6b or 6ab; 5a; 5bb C 6b or 6a 2.25-2.42 (m; 5H) d e f C lOa and/or 2.25-2.42 (m; 5H) lOb 14’ 3.64 (s; 3H) 15a 4.65 (s; 1H) 15b; lOa and/or lObb 15b 4.87 (s; 1H) 15a; lOa and/or lObb a. 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) showed correlations to 2b, 3a, 4b, 5a, Ga or 6b, as well as to the signals at 8 1.52-1.65, 1.671.73 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 (distorted dd, 1H, J = 13.5, 3.5 Hz) and 1.98 (dm, 1H, J = 13.5 Hz). f. The correlations were not determined. -  -  100  The synthesis of the keto ester 1 82 meant it was possible to study the one-carbon ring contraction of the functionalized five membered ring.  Application of such a ring contraction to 182 would  generate a carbon framework with ring sizes corresponding to those in the target natural product.  2.3.2.4. Ring Contraction to Give a 4-5-6 Tricyclic Carbon Skeleton.  There are many different procedures available for performing one-carbon ring contractions. 104  One common method employed is  the Wolff rearrangement of cyclic a-diazo ketones 183 (equation D 37).105  The reaction can be performed thermally, photochemically or  catalytically.  Generally, the mechanism for the photochemical  process is thought to involve the formation of a singlet carbene 184  via 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 to the carbene site and a ketene 185 is generated.  If a nucleophile  such as water, an alcohol, ammonia or an amine is present, it can add to the ketene to produce the corresponding acid, ester or primary or secondary amide.  101  heat/hv catalyst  6:  (D-37)  p  (H C 2 )  183  184  185  A variety of procedures exist for the preparation of the desired a-diazo carbonyl compounds, but one of the most widely used methods is the “deformylation diazo group transfer”. 106  In this  procedure, a second activating group, usually a formyl group, is introduced a- to the carbonyl group of a ketone (general formula, 186) to give 187.  During the introduction of the diazo group from a  diazo transfer reagent (for example, p-toluenesulfonyl azide, p ’° to give an a-diazo ketone 188, the formyl group is cleaved 3 TsN ) 7 to form an amide, 189 (equation  HCOORJL  D-38).  base H  2 JL.N  (D-38)  -HC ,p-Ts NH  186  187  189  188  Two main procedures have been developed to perform the diazo group transfer reaction. 108  In one method, the sodium salt 190  resulting from the Claisen condensation of a carbonyl compound 186 with alkyl formate/sodium alkoxide is treated with p-toluene sulfonyl azide (p-TsN ) to give the a-diazo ketone 188 (equation D 3 39).  In the second method, which is more commonly utilized for  cycloalkanones, the tautomeric mixture of 191 and 192 is allowed  102  3 in the presence of an organic base to generate to react with p-TsN 183 (equation D-40). 0  0  HCOOft  H  R 0 Na  3 p-TsN  (D-39)  190  186  188  0 HO  p-TsN base , H 3  191  (D-40)  183  192  Diazo group transfer reagents other than p-TsN 3 have been used (for example, p-carboxybenzenesulfonyl propylbenzenesulfonyl  azide 09, 1  2,4,6-triiso-  azide° and methanesulfonyl 111 azide but ),  the use of 3 p-TsN has been most widespread.  The diazo transfer  process may be described as the attack of the anion of the formylated carbonyl compound on a diazo transfer reagent consisting of N 2 attached to a leaving group.’° 9 The formyl group is transferred to the sulfonyl leaving group to give a sulfonamide (i.e., 189) while the diazo group is transferred to the carbonyl compound to give the a-diazo carbonyl compound (188 or 183). For a-formyl cycloalkanones  191  or their hydroxymethylene tautomers 192,  diazo group transfer likely goes (Scheme D-11). 112  via the cyclic triazoline 1 9 3  Decomposition of the triazoline 193 can occur by  two different pathways depending on the ring size.° 2 The x 1 ’ 6  103  diazo cycloalkanone 183 and N-(p-toluenesulfonyl)formamide 189, products of the desired decomposition, are formed mainly by path a and are the only products observed for 5-, 7-, and 8-membered rings. It is also possible that the cz-diazo formed by the sequence 193  —  194  cycloalkanone -  183.  183 may be  On the other hand, the  formation of the p-toluenesulfonyl-2-oxo-cycloalkylcarbonamide 196 via path b occurs to a certain extent for 6-, 9-, 10-, 11-, and 12-membered rings.  Thus, loss of dinitrogen from the intermediate  194 is followed by a rearrangement of the intermediate 195 to produce the amide 196. Application of the deformylation diazo group transfer process to the keto ester 182 in order to prepare an a-diazo cycloalkanone for the Wolff ring contraction required the synthesis of the formylated ketone 197.  A variety of attempts to perform the formylation (KH/  ethyl formate/ THE; NaH/ ethyl formate/ THE; NaH/ ethyl formate/ diethyl  3 NaH/ methyl formate/ diethyl ether; ether;’ 114  NaOMe/  methyl formate/ diethyl ether; KH/ methyl formate/ THF), resulted either in the formation of the product in low yields or in the recovery of the starting ketone 182.  In the end, treatment of a  keto ester 182  with sodium t-amyloxide’’ 5  benzene solution of the  (4 equiv) at room temperature, followed by the addition of methyl formate (8 equiv) (equation D-41) led to the quantitative formation of the keto aldehydes 197a and 197b and the enol tautomer 198 (ratio 1 :traces:8.5). purification  by  The products 197 and 198 were not stable to  chromatography.  Hence  characterized with traces of impurities present.  the  mixture  was  104  H p-TsN , 3 CI CH , 2 N 3 Et  HO 191  1  N H 19  Patha  0 p-Ts + HC’ NH  (CI  ‘2  183  H  194 Path b  189  p-Ts  2 -N  N—p-Ts  NH—p-Ts  196  195  Scheme D-11  105  MeO C 2 4  R’ Hb 1) t-amyIONa, -7°C; 197a R=CHO, R’=H PhH, rI (1.5h) + 197b R=H, R’=CHO (D-41) 2) MeO CH, 2 + 5°C —+ rt (17h) 182  198  An ir spectrum of the mixture of 1 97 and 1 98  indicated the  presence of the ester carbonyl, and the enolized 13-keto  aldehyde  group with absorptions at 1725, and 1699, 1609 (broad) cm, 61 respectively.  Based on the H nmr spectrum of the formylated  mixture, the formylation of 182  occurred at the sterically less  hindered 3-position rather than at the 2-position and gave mainly the enol 198, as evidenced by the change in the chemical shift for H-4b (8 2.84-2.92 in 182 and 3.20-3.23 in 198) and the presence in the spectrum of 1 98 of a signal at 6 7.08 (s, 1 H) for the olefinic hydrogen of hydroxymethylene function.  The signal for the aldehyde  hydrogen of the major aldehyde epimer 197a (stereochemistry based on steric considerations) was at 39.52 (s), while the presence of traces of 1 97b were indicated by a singlet at 9.83.  In the low  resolution mass spectrum, the molecular ion was found at 276 and fragments corresponding to the loss of CO (M-28), MeOH (M-32) and  106  Me (M-59) were present; the latter fragment was the base peak. 2 CO In the high resolution mass spectrum, the mass found for the molecular ion of 197/198 (M=276.1363) was consistent with the 10 C 2 H . 4 6 expected formula, 0 Initially, the a-diazo formylated  keto/enol  ketone esters  was prepared from the * in 3 197/198 using pTsN 199  dichloromethane with triethylamine as the base.’ 16  Unfortunately,  it proved impossible to separate unreacted p-TsN 3 from the diazo ketone, a problem which has been noted before.’ 11  Despite the  presence of the unreacted p-T5N , the a-diazo ketone 199 was 3 subjected  to  ,Gb 3 methanol.  the  photochemical  Wolff  rearrangement  in  The reaction proceeded, albeit in very poor yields  (<25% of the ring contracted product was obtained from the photolysis), and an uncharacterized aromatic by-product proved difficult to remove. The expected by-product, the amide 189 (see equation D-38 or Scheme D-11), was obtained which indicated that the diazo transfer was indeed occurring. Taber and coworkers found that the use of methanesulfonyl azide (MsN ) was advantageous” 3 1 due to the fact that unreacted 3 can be separated from the a-diazo carbonyl compound by M5N washing an organic solution of the diazo compound with aqueous 10% sodium hydroxide solution. according to Danheiser’s  Consequently, MsN 3 was prepared 7a  of Boyer’s procedure.’  1 7b  The diazo group transfer reaction was then performed by treating a dichloromethane solution of 197 and 198 with MsN 3 in the presence *  CAUTION: All with due care.  sulfonyl azide compounds are potentially explosive and must be handled  107  of triethylamine at 0°C (equation D-42). 118 ketone  199,  protected  was light-sensitive so the  from  light and  executed in a dimly lit room.  the  The product, a-diazo reaction  subsequent  mixture was  manipulations were  Partial removal of some impurities by  an aqueous base/dichloromethane extraction was followed by a rapid chromatographic separation (silica gel) of 1 99 from more polar material.  Also, removal of the triethylamine (rotary evaporator)  from the diazo ketone 199 was important since even traces of triethylamine led to the formation of by-products in the photolysis reaction.  An ir spectrum of 199 displayed the diazo group stretch  at 2082 and the a-diazo  carbonyl  stretch  at  1674  cm  •6 1  Absorptions due to the exocyclic methylene were at 1636 and 898 . 1 cm-  , 3 3 MsN Et, N  CI 0°C 2 CH ,  :Q  (4  (D-42)  (in the dark)  198 and the  199  keto aldehydes 1 97  Due to its instability, the a-diazo ketone 199 was used as quickly as possible after its preparation for the photochemical ring contraction to prepare the diesters 200 and 201.  Thus, 199 was  dissolved in deoxygenated distilled methanol in a quartz photolysis  tube and the tube was closed under an argon atmosphere.  Generally,  the photolysis reaction, using a medium pressure Hanovia mercury  108  lamp (450 Watt) with a Corex filter,” 9 was complete in 30 minutes at 0°C (equation D-43).  An -‘1.6:1 mixture of the diester epimers  200 and 201 was obtained in a 40.5% overall yield from the keto ester 182.  MeOCi,. H hv, MeOH, 0°C (30mm)  (D-43)  200 +  2 N  199 12’  2 Co H  H  201 Usually, protonation of the ketene intermediate occurs from the less sterically hindered face of the ° 2 ketene .’ of molecular models,  the  From an examination  major epimer formed  in  the Wolff  rearrangement of the diazo ketone 199 in the presence of methanol was expected to be epimer 200. The mixture of epimers obtained was hard to separate and only a small amount of the major epimer 200 was obtained uncontaminated with the minor epimer 201.  However, the elemental analysis and  high resolution mass spectral data for the mixture provided results consistent with the molecular formula, 0 22 In the ‘H nmr H 6 C, . 4 spectrum, several signals readily distinguished the two epimers.  109  For example, a doublet of doublets resonated at 61.78 (J Hz) in the major epimer 200, and at 1.69 (J minor epimer 201  =  =  13.0, 8.0  13.5, 7.0 Hz) in the  In addition, the signals for the epimeric  .  methoxycarbonyl groups appeared as singlets (3H each) at 63.68 (Me-13’) and 3.67 (Me-12’), respectively, while a triplet of doublets resonated at 2.71 (J  =  9.0, 3.0 Hz) for 200 and a doublet of doublets  of 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 fully by one and two dimensional nmr experiments C H) in order to 13 and 1 ( assign  reasonably all of the signals for the carbons and the  hydrogens and to determine the relative stereochemistry at C-3. The broad band decoupled and APT 13 C nmr spectra, as well as a 1 HC HETCOR were obtained. 13  Signals for only 15 of the 16 carbon  C nmr spectra as the atoms were displayed in the one dimensional 13 signals for two methylenes resonated at the same frequency. the  HETCOR  From  H- correlations (see figure 7 and Table 4), 1 C 13  assignments for the methines, the methoxycarbonyls and the geminal pairs of hydrogens were obtained.  Thus, for example, the carbon  signals 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 to  four 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 same position.  Other correlations between the carbon and hydrogen  signals permitted the determination of the rest of the geminal pairs of hydrogens,  but the assignments to specific positions were  dependent on the COSY correlation results.  110  -25 •  D  0  0  -50  3 CDCI  -  -100  0  ppm I,,JIIuI.IIII  ppm  I  4  ,IIt,  ,,, ill , 1  3  II  11111  2  1  Figure 7. The 125 MHz HETCOR spectrum of the diester 200.  111  Table 4: The 125 MHz HETCOR Data for the Diester 200.  200 Position C (125 13 (C-x) MHz) Sppm 52.4a 1 2 26.6 3 36.2 4 46.8 25.2 5 37.2 6 579 7 23.8 8 32.7 9 10 32.7 11 147.9 13 and 14 174.0, 175.5 13’ and 14’ 51.37, 51.40 107.7 15 a. Signals may be interchanged.  H (500 MHz) Sppm (H-x) 1 -  -  2.26 (2a); 2.39-2.44 (2b) 3.18 (3b) 2.71 (4b) 1.58-1.65 (5a); 1.95-2.07 (5b) 2.43-2.50 (6a); 1.78 (6b) -  -  1.41-1.52 (8a or 8b); 1.58-1.65 (8b or 8a) 1.89 (9a or 9b); 1.41-1.52 (9b or 9a) 1.95-2.07 (lOa or lOb); 2.33 (10 b or lOa) -  -  -  -  3.68 (13’); 3.62 (14’) 4.93 (15a); 4.96 (15b)  H nmr decoupling and nOe experiments (500 and 400 MHz, 1 respectively) were performed on a solution of the diester 200 in order to determine the (a) identities of the hydrogen signals for hydrogens  on  the  four-membered  ring  and  (b)  the  relative  configuration at C-3 (see figure 8 for the normal 1 H nmr spectrum). Thus,  in decoupling experiments,  irradiation of the  signal at 83.18  112  3 CHCI  4.0  2.0  3:0  PPM  Figure 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 a distorted doublet of multiplets (J  =  9 Hz), simplified the doublet of  doublets at 2.26 (H-2a) to a doublet (J multiplet at 2.39-2.44 (H-2b).  =  13 Hz) and simplified the  Irradiation of the signal at 3 2.71 (H  4b) simplified the multiplets at 1.95-2.07 (H-5b and one other hydrogen) and at 2.39-2.44 (H-2b) and caused the doublet of triplets  1.0  113  at 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 at 2.39-2.44 (H-2b) and caused the doublet of triplets at 3.18 (H-3b) to collapse to a distorted triplet (J  =  9 Hz).  The above assignments of  the signals for hydrogens on the four-membered ring were further confirmed by nOe difference experiments (summarized in structure 202’).  Thus, irradiation of the signal at 3 2.71  (H-4b)  led to  enhancement of the signal at 3.18 (H-3b), while irradiation of the signal 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 (between  H-15a and H-15b) led to enhancements of the signals at 2.33 (H-lOb or 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 of the four-membered ring and that H-3b and H-2b were spatially close to the olefinic H-15a hydrogen.  Therefore, the six-membered ring  was syn to the bridgehead methine (H-4b) as s found in the natural product and the methoxycarbonyl group on the four-membered ring had the configuration shown in 200.  13’  200’  114  The identification of other signals in the  1  H nmr spectrum was  facilitated by combining the results summarized above with those from the COSY and HETCOR spectra.  For example, in the COSY  spectrum (see figure 9), the signal for H-4b (32.71)  showed  correlations to the signals at 32.39-2.44 (H-2b), 3.18 (H-3b) and 1 .95-2.07 (2H).  In the foregoing discussion, the identities of the  hydrogen signals at 32.39-2.44 and 3.18 had been established from decoupling and nOe experiments.  Consequently, one of the hydrogens  resonating at 3 1 .95-2.07 was H-5b.  From the HETCOR results, the  hydrogens gem inal to those at 6 1 .95-2.07 (m, resonated at 1.58-1.65 (m, 2H) and 2.33 (1H).  H-5b and H-x)  The signal at 62.33  could 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 part  of the multiplet at 6 1 .58-1 .65 and the other hydrogen at 1 .95-2.07 was H-lOa or H-lOb.  A similar combination of the various nmr  results led to the further assignment of the remaining carbon and hydrogen signals to the positions listed in Tables 4 and 5.  115  -  1.5 d  -  .  .-  2.0  1 —  2.5 .  3.0 i  1IuI  —  ..  3.5  —  4.0  4.5 5.0 —______  .50.  4.5  4.0  3.5  3.0  2.5  2.0  1.5  Figure 9. The 400 MHz COSY spectrum of the diester 200.  116  Table 5: The 400 MHz COSY Data for the Diester 200.  Ha  200 Position (H-x) 2a 2b 3b 4b 5a 5b 6a 6b Ba or 8b 8b or Ba 9a 9b lOa lOb  or or or or  9b 9a lOb lOa  Signal Sppm (multiplicity;a J; number_of_H) 2.26 (dd; 13.5, 10.5; 1H) 2.39-2.44 (m; 1H) 3.18 (dt; 10.5, 9.5; 1H) 2.71 (br td; 9.0, 3.5; 1H) 1.58-1.65 (m; 2H) 1.95-2.07 (m; 2H) 2.43-2.50 (m; 1H) 1.78 (dd; 13.0, 8.0) 1.41-1.52 (m; 2H) 1.58-1.65 (m; 2H) 1.89 (dm; 1.41-1.52 1.95-2.07 2.33 (dm;  10.5; 1H) (m; 2H) (m; 2H) 13.5; 1H)  COSY Correlations (H-x) 2b; 3b 2a; 3b; 4b 2a; 2b; 4b 2b; 3b; 5b 5b; 6a 5a; 4b; 6a; 6b 6b; 5a; 5b 6a; 5b b Ba or 8b; 9a and 9bC; lOb or lOa (W-coupling) 9b or 9aC; 8b or Ba; lOb or lOa b lOb or lOa; 9b or 9a lOa or lob; 8b or Ba (W coupling); 9a and 9b  13’ 3.68 (s; 3H) 14’ 3.62 (s; 3H) 15a 4.93 (s; 1H) 15b l5b 4.96 (s, lH) 15a a. The signals labelled s, d, dd may incorporate unresolved fine couplings. b. Assignment of correlations is uncertain due to overlapped signals. The signal at 3 1.41-1.52 (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, the correlations of 8b or 8a to 9b or 9a, and vice versa, are impossible to determine from the data. -  -  -  -  117  Preparation of (±)-J3-Panasinsene (31) via the Diacetate  2.3.2.5.  213. The successful generation of the diesters 200 and 201 via the Wolff rearrangement reaction set the stage for the performance of the final functional group manipulations to complete the synthesis Thus, the introduction of a methyl group  of (±)--panasinsene (31).  at the 3-position of the diester mixture (200 and 201) followed by a double deoxygenation of the methoxycarbonyl functions would provide the natural product. The methylated diesters 202 and 203 were prepared by treating a cold (-78°C) THE solution of a mixture of the diesters 200 and 201  (ratio  -1 .6:1)  first  with  a  THF  solution  of  lithium  diisopropylamide to generate an anion at the 3-position, then with hexamethylphosphoramide* (HMPA, 1.6 equiv) and finally, with excess methyl iodide.  After an appropriate workup, an -‘18:1  mixture (‘H nmr analysis) of the diester epimers 202 and 203 was isolated in 65% yield (equation D-44).  The two epimers could not  easily be separated at this stage so were characterized as the mixture. The elemental analysis and the low and high resolution mass spectroscopy performed on the diester mixture (202 and 203) 4 0 2 H 7 C, . provided results consistent with the molecular formula, 4 Details of the structure. of 202 were obtained from an ir spectrum, nmr spectra and one and two dimensional ‘H nmr  *  CAUTION: HMPA is known to be a potent carcinogen.  experiments.  118  15 12  1) LDA, THE, -78°C (1 .5h) 2) HMPA, -78°C + 3) Mel, -78°C, (O.5h); 2 CO H -78 200 and 201 (50mm) CIg. 2 MeO  Me 13’  ,  , 1 H  vb 2 CO  202  (D-44)  -,  15  14’  12’  2 Co Me ‘•‘  203  13  Thus, in the ir spectrum of the mixture of 202 and 203, the absorption  for the ester carbonyl stretch was displayed at 1729  , while the exocyclic olefinic function gave rise to absorptions 1 cmat 1642 and 887 cm.  The broad band decoupled 13 C nmr spectrum  of the 18:1 mixture of 202 and 203 displayed the expected 17 carbons for 202 and baseline signals (i.e., hardly distinguishable from the noise) for 203.  From an APT experiment, the signal at 8  25.3 was assigned to the methyl group (.H -12), while the signal at 3 53.6 was assigned to the methine (.H-4).  The methyl groups of the  methoxycarbonyl functions resonated at 851.4 and 51.5 (H -13’ and 3 -14’). 3 H  In the ‘H nmr spectrum (see figure 10), the signal for the  newly installed methyl group of the major epimer (202) appeared as a singlet at 31.42, while that of the minor epimer 203 was at 1118. The signals for the two methoxycarbonyl groups of 202 were singlets at 63.61  (Me-14’) and at 3.69 (Me-13’), while the two  119  olefinic hydrogens appeared as singlets at 34.98 (H-15a) and 5.00 (H-15b).  TMS  5:0  4.5  3.5  3.0  2.5  1.5  0.5  Figure 10. The 400 MHz ‘H nmr spectrum of the diester 202.  PPM  120  202’ In order to determine the relative configuration at C-3, nOe difference experiments (summarized above in structure 202’) were performed on the mixture of 202 and 203.  Thus, irradiation of the  signal 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  unavoidable  irradiation of part of the multiplet at 6 1 .46-1 .59, the signals at 1.85 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 the signals at 1.42 (Me-12) and at 2.26.  From these results it was  apparent that the methyl group (Me-12) in the major epimer was in close proximity to an olefinic hydrogen (H-15a) and, therefore, had the relative configuration depicted for 202.  The configuration is  that expected if the methyl group had approached from the less sterically hindered face of the anion obtained by deprotonation at the 3-position of the diesters 200 and 201.  Further evidence that  the structure was correct was obtained once the identities of the hydrogens resonating at 62.26 and 2.35 were determined.  In 202, H-  2b and H-4b are cis to the methyl group (Me-12) and H-2b is in close  121  proximity to the olefinic hydrogen (H-15a).  Consequently, the nOe  results 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 a  3H-multiplet at 2.28-2.39) to H-4b. Hb  1 0 2 Me0 , i  14  14  202 Confirmation  of the assignments for H-2b and  obtained from decoupling and/or COSY experiments.  H-4b was Thus, in a  decoupling experiment, irradiation of the doublet at 32.45 (J  =  14.0  Hz, H-2a) led to the collapse of the doublet of doublets at 2.26 (H 2b) to a distorted triplet (J  =  3.0 Hz) indicating that the signals at  2.26 and 2.45 were due to geminal hydrogens (J  =  14.0 Hz). In a COSY  experiment (see figure 11), the signals for (a) the methyl group (Me12) at 6 1.42 and (b) the olefinic hydrogens (H-i 5a and H-i 5b) at 3 4.98 and 5.00 provided key entries, via couplings unresolved in the one dimensional spectrum, into the spin systems of the interrelated four  and  five-membered  respectively.  rings  and  the  Thus, the signal at 6 1.42  six-membered (Me-12)  ring,  showed  a  correlation (W-coupling) to only the signal at 2.45 (H-2a), while the signal at  2.45 (H-2a) showed  a further correlation  to the  signal at  122  J  1.5  2.0  0  2.5  -  3.0  -  3.5  4.0  -  4.5  -  5.0  -.  PPti 5.0  4.5  4.0  3.0  3.5  2.5  2.0  1.5  PPM  Figure 11. The 400 MHz COSY spectrum of the diester 202.  2.26 (H-2b).  In turn, the signal at 8 2.26  correlations to the signals at 2.28-3.39  (H-2b)  showed  other  (H-4b and two other  hydrogens) and extremely small (long range) correlations to the signal at 4.98 (H-15a). The correlations shown by the signal at S 2.28-2.39 were too complex to further assign with certainty.  Thus,  the correlations due to the signals for the olefinic hydrogens (H-15a and H-15b) were examined.  Apart from the correlation of the signal  123  at 34.98  (H-15a)  with  the  one  at 2.26  (H-2b),  both  olefinic  hydrogens showed correlations to the multiplet at 1.95-2.06 (H-lOa or H-lOb and another H).  Further tracing of the spin systems was  not  COSY  performed,  but the  results,  presented  above  and  summarized in Table 6, are consistent with the flOe and decoupling results and confirm that the structure of the major epimer 202 is as assigned.  Table 6: The 400 MHz COSY Data for the Diester 202. Hb  202 Position (H-x) 2a 2b 4b c lOa or lOb 1 2 1 3’ 1 4’ 15a  Signal 3 ppm (multipli.city;a J; number_of_H) 2.45 (d; 14.0; 1H) 2.26 (dd; 14.0, 3.0; 1H) 2.28-2.39 (m; 3H) c 1.95-2.06 (m; 2H) 1.42 (s; 3H) 3.69 (s; 3H) 3.61 (s; 3H) 4.98 (s; 1H)  COSY Correlations (H-x) 2b; 12 (W-coupling) 2a; 4b; 15a (long range coupling) 2b; b d 15a; 15b; b 2a (W-coupling) -  -  -  -  15b; 2b (long range coupling); lOa or lOb 1H) lOa or lOb (s; 15a; 15b 5.00 a. The signals labelled s, d, dd may incorporate unresolved fine couplings. b. Other correlations also were observed.  124  Table 6: footnotes continued. c. The positions of H-x (x=5a, 5b, 6a, 6b, 8a, Bb, 9a, 9b and lOb or lOa) are uncertain. 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 the diols 204 and 205 (equation D-45).  by reduction with lithium aluminum hydride  After an appropriate workup and purification, the  major epimer 204 was obtained in 88% yield, while only traces of the minor epimer 205 were isolated. H  H  LiAIH THF, , 4 rt (1 .2h)  (D-45) + H  202 and 203  H 12  Hb  205 Recrystallization of the major epimer 204 needles with m.p. 139-139.5°C (sealed tube).  gave  colorless  Due to low solubility  in other solvents, the nmr spectra of 204 were obtained in acetone H nmr spectrum (400 MHz) of the diol 204, the signal for . In the 1 6 d the methyl group (Me-12) appeared as a singlet (3H) at 6 1.11.  The  signals for the two hydroxyl hydrogens were triplets (1H, each) at 3 3.20 (J  =  5.5 Hz) and at 3.32 (J  =  5.0 Hz), both of which disappeared  125  upon the addition of D 0. The signals for the Ca 2 OH hydrogens were 2 displayed at 83.23-3.30 (m, 2H), 3.40 (dd, 1H, J 3.46 (dd, 1H, J  =  =  10.5, 5.5 Hz) and  10.5, 5.0 Hz); upon the addition of D 0, the multiplet 2  was simplified and the two doublets of doublets gave rise to doublets (J  =  10.5 Hz). The 13 C nmr spectrum of 204 displayed the  expected 15 signals required for 0 24 H 5 C, . 2  The elemental analysis  and high resolution mass spectral data (found M=236.178O) were also consistent with the molecular formula.  In the low resolution  mass spectrum, the peak due to the molecular ion was very weak (0.4%) and peaks were found which corresponded to the loss of one and two molecules of water (M-18 and M-36, respectively).  The ir  spectrum of 204 indicated the presence of the hydroxyl groups by an absorption at 3312 cm . 1 A small amount of the minor epimer 205 was isolated for characterization purposes.  Thus, in the ‘H nmr spectrum of 205, the  signal for the methyl group (Me-13) appeared at 60.93 (s, 3H).  The  signal for one hydroxyl group was displayed as a triplet at 63.20 (J  =  0, while the other 2 5.5 Hz) and disappeared upon the addition of D hydroxyl hydrogen and both of the CU OH signals were part of a 2 multiplet at 8 3.30-3.49 (5H).  The addition of D 0 simplified the 2  multiplet 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).  =  10.5 Hz), 3.35 (J  =  11.0 Hz) and at 3.43 (J  =  The exact mass of the molecular ion in the high resolution  mass spectrum (M= 236.1777) was consistent with the expected molecular formula of 205. Several factors were considered when choosing. which method to use for the double deoxygenation of the diol 204.  In the first place,  126  both of the primary hydroxyl groups in 204 are neopentyl in nature and, therefore, the carbinol carbon atoms are very hindered.  Thus,  121 were SN2-type reactions, which are susceptible to steric effects, not  deemed  feasible  for  use  in  the  performance  of  the  23 (unlike reactions 1 ’ On the other hand, radical 122  22 deoxygenation.’  SN2 processes) are not as susceptible to steric effects.  In addition,  122 so acid or the radical reactions occur under neutral conditions, base sensitive groups are compatible with the conditions utilized One problem in applying radical reactions to the  for the reactions.  deoxygenation of primary alcohols reflects the decreased stability of primary radicals in comparison with secondary radicals.  Thus,  the reaction conditions used are generally more drastic for the deoxygenation of primary alcohols than for the same reaction of For example, deoxygenation of the  secondary  24 alcohols.’ 1 ’ 22  secondary  monothiocarbonylimidazolide  hydride to give 5a-cholestane 125 toluene.  In  contrast,  monothiocarbonylimidazolide  206 using tri-n-butyltin  (207) took 1.5 hours in refluxing deoxygenation  of  the  primary  208 with the same reagent to give f3-  24 (equations D-46 and amyrin 209 took 10 hours at 130°C in xylene’ D-47, respectively). The possible application of a radical-based deoxygenation to the double deoxygenation of a derivative of the diol 204 led to the consideration of how to separate the hydrocarbon product, J3panasinsene, 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 were  127  SnH, PhMe (O.5h) 3 jjl) n-Bu 2) reflux (1.5h) H  207  206  (D-47) 1) n-Bu SnH, xylene (2h) 3 2) 130°C (lOh)  H H  209  208  Due to known problems in separating  deemed to be most suitable.  126 the product from reagents and by-products, triorganotin hydride reductions  or  the  more  recent  variations  27 tris(trimethylsilyl)silane,’ 28 (triethylsilane,’  silanes  using or  diphenyl  silane 29) were not attempted. 1 Reductions using dissolving metals, such as those employed in 29 were expected to Ireland’s deoxygenation of phosphodiamidates,’ be feasible, but problematic due to the possible over-reduction of the olefinic functional group to give the saturated deoxygenated product. hindered  Thus, for example, Wai and Piers found that the sterically primary  alcohol  210  could be deoxygenated  via its  128  phosphorodiamidate derivative to give 211 ,12,131 but the reaction was somewhat capricious and varying amounts of the over-reduced saturated product 212 also were formed if conditions were not carefully controlled (equation D-48). 32  I  I  2 (Me Li, MeNH 2  210  (D-48)  211  212  133 and coworkers reported that the photolysis of acetates Pete in HMPA/water works well for acetate derivatives of primary and secondary alcohols, while Collins and Munasinghe 34 found that the reaction could also be successfully applied to the diacetates or tripivaloates derived from diols and triols,  respectively.  The  deoxygenated product may be recovered by adding water to the reaction solvent and performing extractions with an organic solvent. In order to attempt Pete’s procedure, the diacetate 213 was prepared.  Thus, a cold (0°C) dichloromethane solution of the diol  204, containing DMAP (-1.1 equiv) and pyridine (9 equiv), was treated with acetyl chloride (6 equiv) (equation D-49). appropriate  workup  obtained in 85% yield.  and  purification,  the  diacetate  After an 213  was  129  HO  Ac DMAP, Pyr, COCI, 3 CH  Me  (D-49)  CI 0°C (3h) CH , 2  H  213  204  In the ir spectrum of the diacetate 21 3, the presence of a carbonyl absorption at 1742 cm 1 indicated that the replacement of the hydroxyl functions by acetate groups had occurred.  The expected  absorptions for the exocyclic methylene were displayed at 1636 and . 1 889 cm  In the ‘H nmr spectrum (see figure 12), the signals for  the three methyl groups (Me-13”  appeared at  6 1.15 (Me-12),  2.02 and 2.05  and Me-14”), while the signals for the methylene hydrogens  of the acetoxymethyl functions were found at 3.79 (dd, 1H, J 1.0 Hz, H-14a), 3.86 and 3.89 (AB pair of d, 2H, J and 4.03 (d, 1H, J  =  11.0 Hz, H-14b).  =  =  11.0,  11.0 Hz, 2H-13)  The origin of the unexpected  small coupling in the signal for H-14a (J  =  1.0 Hz) was explained  upon examination of a COSY spectrum of 213 (see figure 13) which showed a correlation (W-coupling) between the signals at 8 3.79 (H 14a) and at 1.29-1.39  (m, 1H, H-8b or H-6b).  From the shape of the  Ii, 14  CC(O)O. 3 H 14”  13  13”  3 OC(O)CH  213  130  TMS  5.0  4.0  3.0  2.0  1.0  Figure 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 was more likely that the signal was due to H-8b rather than to H-6b. Other assignments of hydrogen signals were possible based on the results of the COSY experiment.  The resonance due to the methyl  group (Me-12) provided an entry into the spin system of the four membered ring, while the signals due to an olefinic hydrogen (H-15b) and H-14a provided the needed access into the spin system of the six-membered ring.  Thus, the signal at 3 1.15 (Me-12) showed a  PPM  131  correlation (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 signal at 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 6 1.94-2.07, the correlations due to the signals for ôlefinic hydrogens H-15a and H-15b at 4.84 and 4.96, respectively, were then examined. However, apart from correlations to each other, both  showed  correlations to the multiplet at 3 1 .94-2.07 (9H) indicating that the signals 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) gave  rise to a correlation to the signal at 1.29-1.39 (m, lH, H-8b).  The  signal at 6 1 .29-1 .39 (H-8b) also showed correlations to the signals at 1.43 (br qt, lH, J  =  13.0, 3.0 Hz) and 1.62-1.72 (m, 3H).  Due to the  presence in the 1 H nmr spectrum of several complex multiplets, other assignments of signals were not feasible. To confirm the assignments of the various hydrogen signals given above (and in Table 7), a NOESY 97 (two dimensional nOe) spectrum of 213 was obtained. summarized in Table 8.  The NOESY correlations are  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) showed enhancements 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 the  other acetoxymethyl group H-14a (at 63.79) displayed enhancements to the signals at 4.03 (H-14b) and at 1.72 (H-2a).  It thus appeared,  132  that the depicted structure for the diacetate 213 was correct and that the assignments of the hydrogen signaJs were reasonable.  1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 PPM  PPM  Figure 13. The 400 MHz COSY spectrum of the diacetate 213.  133  Table 7: The 400 MHz COSY Data for the Diacetate 213.  Hb 213 Positiona (H-x) 2a 2bC 8b lOa or lObC 12  Signal Sppm (multiplicity;b J; number of H) 1.72 (d; 13.0; 1H) 1.98 (dd; 13.0, 3.0; 1H) 1.29-1.39 (m; 1H) 1.94-2.07 (m; 9H) 1.15 (s; 3H)  13  3.86 and 3.89 (AB pair of d; 11.0; 2H) 2.02 (s, 3H); 2.05 (s; 3H) 3.79 (dd; 11.0, 1.0; 1H) 4.03 (d; 11.0; 1H) 4.84 (s; 1H) 4.96 (s; 1H)  13” and 14” 14a 14b 15a 15b  COSY Correlations (H-x) 2b;C 12 (W-coupling) d 14a (W-coupling)e l5ad 2a (W-coupling); 13 (W Co u p11 n g) 12 (W-coupling); f -  -  14b; 8b (W-coupling) 14a 15b 15a; lOa or lObC  a The assignments of the hydrogens H-x (x= 4b, 5a, 5b, 6a, 6b, 8a, 9a, 9b, and lOb or ba) 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).  134  Table 8: The 400 MHz NOESY Data for the Diacetate 213.  Hb 2  Hb 213 Position Signal Sppm (multiplicity; J; number NOESY Correlations (H-x) (H-x) of H) 2a 1.72 (d; 13.0; 1H) 2b; 13; 14a 2b 1.98 (dd; 13.0, 3.5; 1H) 2a; 12; 15a 12 1.15 (s; 3H) 2b; 13; 15a 13 3.86 and 3.89 (AB pair of d; 11.0; 2H) 2a; 12 14a 3.79 (dd; 11.0, 1.0; 1H) 14b; 2a 14b 4.03 (d; 11.0; 1H) 14a 15a 4.84 (s; 1H) 15b; 2b; 12 15b 4.96 (s; 1H) 15a  The preparation of the diacetate 213 set the stage for the crucial double deoxygenation reaction to prepare the natural product, 133 (±)-/3-panasinsene (31) via Pete’s procedure.  Thus, solutions of  the diacetate 213 in HMPA-H 0 (-p95:5) were photolyzed using either 2 a single low pressure mercury lamp (emission at 253.7 nm) or a Rayonet reactor (16 such lamps) until no more starting material was detected (‘7-9 hours).  Dilution of the reaction mixture with water,  followed by an extraction with pentane and removal of the solvent led to the isolation of low yields of (±)-f3-panasinsene  (31), which  was usually contaminated by a small amount of a compound with very  similar properties.  The  impurity could  be  removed  by  135  chromatography of the mixture on silica gel impregnated with silver nitrate (1.25 g AgNO / 5.00 g of 70-230 mesh silica gel),’ 3 31 but the yields of the target compound were too low (<10%) to make the procedure feasible.  After several unsuccessful attempts to improve  the results, a different approach to the double deoxygenation of the diol 204 was undertaken.  2.3.2.6. Preparation of (±)-/3-Panasinsene (31) via a Wolff-Kishner Reduction.  There are various methods known for the reduction of aldehyde or keto groups to the corresponding methylene groups, but the two most common methods are the Clemmensen and Wolff-Kishner 35 reductions.’  The Clemmensen reduction uses zinc amalgam and  aqueous, or in some cases gaseous, HCI.  f3-Panasinsene (31) is  known to rearrange in acid to give neoclovene 56 (vide supra, 3 so the Clemmensen reduction would not be useful in Introduction),’ this case.  The Wolff-Kishner reduction,  in contrast,  involves  conversion of an aldehyde or ketone of general structure 214 to the hydrazone 21 5 and heating the hydrazone with base to give the deoxygenated product 216 (equation D-50).  In a one pot procedure,  the aldehyde or ketone may be heated with hydrazine hydrate and a base to produce the deoxygenated product.  136 2 NH  0 RR  NNH 0 2 H , heat  214  N R’  R  base, heat  (D-50)  215  216  Recently, Roberge 136 found that the use of anhydrous hydrazine* as a cosolvent, rather than just as a reagent, 137 to form the hydrazone improved the yields of the reaction.  Thus, heating a  solution of the ketone in an -1 :2 mixture of anhydrous hydrazine diethylene glycol at  130-140°C generated the hydrazone.  Removal  of the excess hydrazine by a reduced pressure distillation, followed by reaction of the hydrazone with potassium hydroxide at 200210°C, led to the formation of the deoxygenated product.  Yields  were reduced and reaction times were increased if the excess hydrazine was not removed before the base was added.  The product  could be isolated by the addition of water to the cooled reaction mixture followed by extractions with an organic solvent.  The  procedure seemed suitable for the deoxygenation of the dialdehyde 217, presumed to be available from the diol 204.  H  I0  217  *  Anhydrous hydrazine is explosive in the presence of oxidizing agents (including air) and must be handled with great care.  137  The dialdehyde 217 was prepared from the dial 204 via a Swern 138 oxidation.  Thus, the dial 204, in a mixture of dichioromethane  and dimethyl sulfoxide (DMSO), was allowed to react with a mixture of DMSO and oxalyl chloride. triethylamine, followed  by an  Treatment of the mixture with appropriate workup and a rapid  filtration of a solution of the product through a silica gel column, led to the isolation of the dialdehyde 217 in 98% yield (equation D 51).  The product thus obtained was pure enough to be characterized.  H  —OH  1) DMSO, (COd) , 2 CI -78°C (40mm) CH , 2 2) Et N, -78 3 —0°C  (D-51)  —,  (70m in)  204  H  3  CHO  217  The ir spectrum of the dialdehyde 217 showed the expected absorptions for the aldehyde functions at 2723 (w) and at 1718 (vs) cm’, while the absorptions for the exocyclic methylene were displayed at 1638 and 894 cm* In the 1 H nmr spectrum, the signal for the methyl group (Me-12) appeared at 5 1.36 (s, 3H), while the signals 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 formyl groups (M-29 and M-58, respectively) were observed.  The exact  mass of 232.1459, found for the molecular ion of 217 in the high resolution 20 H 5 . 2 10 C  mass  spectrum,  was  consistent  with  the  formula,  138  The dialdehyde 217 was then submitted to the Wolff-Kishner deoxygenation 36 Roberge.’  procedure using the modifications developed by Thus, a solution of the dialdehyde 217 in diethylene  glycol 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 (potassium hydroxide), the mixture was heated at ‘-200°C for 7.5 hours.  Cooling,  followed by an aqueous workup, led to the isolation of crude (±)-J3panasinsene (31) in ‘-76% yield (equation D-52).  A reduced pressure  distillation of this material resulted in a 46% yield of the synthetic natural product 31 in >97% purity (gic analysis).  Me H  CHO  14  1) 2 NNH DEG, H , —135°C (1.5h) 2) KOH, DEG, -200°C (7h)  Me  (D-52) 13  217 The  purified  31 synthetic  (±)-j3-panasinsene  (31)  displayed  spectral characteristics similar to those reported in the literature. Unfortunately, all our attempts to obtain samples of 13 natural 1 ’ 6 or synthetic 3 ’ 30 1 f3-panasinsene  or  the  comparison purposes were unavailing.  appropriate  spectra  for  There is no doubt, however,  that (±)-13-panasinsene (31) was the compound synthesized via the reaction sequence described in this thesis.  This contention is borne  out by a comparison of the data obtained for the newly synthesized (±)-J3-panasinsene  (31) with the data reported in the literature 1  3  139  (see Table 9), and by an examination of the other spectral data that were obtained for the synthetic natural product. Table 9. A Comparison of the Spectral Data for Authentic and Synthetic j3-Panasinsene (31). Type of Data Authentic (-)-J3Synthetic (±)-J3Panasinsenea Panasinseneb Infrared 3080, 1620, 1365, 3088, 1636, 1377, ) 1 (cm 1360, 1260, 1080, 1365, 1269, and 885 930, and 885 0.74 (s; 3H) 0.75 (s; 3H; Me-14) H nmrc 1 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, 1.5 Hz; 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 MHz spectrometer (solvent not identified). 13 The spectrum of synthetic 31 was recorded on a Bruker WH 400 MHz spectrometer in CDCI . 3  The data for the authentic and synthetic J3-panasinsene (31) as compiled in Table 9 are similar, but a few differences may be noted. Firstly, according to the ir spectral data, the synthetic sample did not display significant absorptions at 1080 or 930 cm’ as were reported for the natural product and the absorption due to the exocyclic methylene (C=C stretch) occurred at different positions (1620 and 1636 cm).  The reasons for the difference are uncertain.  140  Secondly, from the ‘H nmr spectral data, the signal for one of the olefinic methylene hydrogens differs significantly with regard to both position and multiplicity (at 3 4.84 (d) versus at 4.91 (dd)).  The  discrepancies in the nmr results may be mainly due to the different instruments used (60 MHz versus 400 MHz). Also, the solvent used in the literature was not identified.  Finally, the low resolution mass  spectral patterns differ particularly with regard to the peak at 146 (literature) or at 147 (synthetic). typographical  error  in  the  The disagreement may be due to a literature.  Other  mass  spectral  differences probably are unimportant as no contradiction is involved. The discrepancies between the literature data and the data from the synthetic J3-panasinsene as summarized in Table 9 are minor in comparison with the similarities and thus, it may be concluded that 31 was synthesized. Further confirmation of the identity of the synthetic (±)-f3panasinsene (31) was procured in the form of a high resolution mass spectrum, ‘ C nmr spectra (broad band and APT), as well as several 3 one and two dimensional nmr spectra (flOe, HETCOR and COSY).  A  molecular formula of , 24 as expected for 13-panasinsene, was H 5 C, indicated by the presence of a molecular ion with an exact mass of 204.1875 in the high resolution mass spectrum of 31.  The number  of carbons was also implied by the presence of 15 carbon signals in C nmr spectrum (see figure 14). Based on 3 the broad band decoupled ‘ the chemical shifts of the signals, those at 8108.3 and 152.3 could -15) and to the quaternary 2 be assigned to the olefinic methylene (H olefinic carbon (-11), respectively.  According to a 13 C nmr APT  experiment, four signals (at 8 52.60, 18.2, 24.85 and 30.64) were due  141  Hb  HaHb  31 to either methine or methyl carbons.  The signal at 6 52.60  was  clearly due to the methine, H-4, while the other three signals corresponded to the signals of the three methyl groups in 31.  CDCI3’  a  0 I  I.  r,  2  a, t fl  •  I  r  JJ  0 (.4  0 (.4  bfl$i 160  140  120  C  a 100  80  60  40  20  C nmr spectrum of Figure 14. The 75 MHz broad band decoupled 13 synthetic (±)-$-panasinsene (31).  0  PPM  142  The ‘H nmr spectrum (see figure 15) of our synthetic 31, in addition to the signals mentioned in Table 9, showed signals for another 13 hydrogens, of which only two signals overlapped in an indistinguishable manner (81.35-1.41, m, 2H).  In order to identify  the methine (H-4b) and the pairs of geminal hydrogens in the ‘H nmr spectrum,  7.5  a HETCOR experiment  7.0 5.0  4.0  was performed and yielded the  3.0  AJJJ 2.0  PPM  Figure 15. The 400 MHz ‘H nmr spectrum of synthetic (±)-J3panasinsene (31).  143  For example,  results summarized in Table 10 (see also figure 16).  the methine carbon signal (H-4) at 6 52.60 correlated with the hydrogen signal at 2.10 (br dd, 1H, J  =  8.5, 3.0 Hz, H-4b), while a  carbon 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. 0 2 H  50  MO  150  ppm  Figure 16. The 125 MHz HETCOR spectrum of synthetic (±)-j3panasinsene (31). (* Folded peaks.  **  Ti noise).  144  Table 10: The 125 MHz HETCOR Data for Synthetic (±)-J3-Panasinsene (31).  31  Position ‘ C (125 MHz) 3 (C-x) 52.77 1 35.7 2 45.6a 3 4 52.60 24.72 5 41.1 6 30.53a 7 36.0 8 25.10 9 10 33.8 152.3 11 30.64 12 24.85 13 18.2 14 108.3 15 a Signals may be interchanged.  H (500 MHz) ö (H-x) 1 -  -  1.46 (2a); 1.96 (2b) -  -  2.10 (4b) 1.63-1.68 (5a); 1.87-1.97 (5b) 1.73 (6a); 1.35-1.41 (6b) -  -  1.27 (8a); 1.42-1.48 (8b) 1.35-1.41 (9a or 9b); 1.57-1.63 (9b or 9a) 2.00 (lOa or lOb); 2.18 (lOb or lOa) -  -  1.07 (3H-12) 0.86 (3H-13) 0.75 (3H-14) 4.80 (15a); 4.91 (15b)  The precise assignments of the positions of the geminal pairs of hydrogens and the methyl groups were made based on the COSY correlations observed for the signals (see figure 17 and Table 11). Entry into the spin systems was afforded by the signal for the methine at 3 2.10 (H-4b) and the signal for olefinic hydrogen at 4.91  145  (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 H-2b, W-coupling) and at 1.87-1.97 (m, 1H, H-5b).  =  12.5, 3.0 Hz,  The signal at 8  1.46 (1H) for the hydrogen geminal to the one at 1.96 (1H) showed a correlation to the methyl group at 1.07 (Me-12), while neither the signal 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 1.46  and 1.96) was assigned to the 2-position and the latter pair (i.e., at 8 1.63-1.68 and 1.87-1.97) was assigned to the 5-position.  The  signals arising from the hydrogens at the 6-position were then identified by determining with which other signals the signals due to H-5a and H-5b showed correlations.  Based on the nOe results  described below and from consideration of molecular models, the aand b-hydrogens on carbons 5 and 6 were assigned as listed in Table 11.  Similarly, the spin system of the six-membered ring was  entered by way of an olefinic hydrogen at 34.91 (dd, 1H, J Hz, H-15b).  =  1.5, 1.5  The methyl groups were assigned based on the fact that  the 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 signals for two ring hydrogens at 1.42-1.48 (a multiplet overlapping the doublet at 1.46).  The W-coupling between the signals at 8 1.07 (Me  12) and at 0.86 (Me-13) meant that the two signals were due to the geminal methyl groups and thus, that the signals at 6 1 .07 and 0.75 correlated, respectively, to the signals at 1 .46 (d, H-2a) and at 1 .421.48 (m, H-8b).  146  0.5 .  I  •  o.  ‘a  0  1.0 1.5 2.0  o  2.5 3.0 3.5 4.0 4.5  ‘‘o T’  5.5  5.0  5.0  4.5  4.0  3.5  3.0  2.5  2.0  1.5  1.0  5.5 0.5  PPM  Figure 17. The 400 MHz COSY spectrum of synthetic (±)-J3panasinsene (31).  PPM  147  Table 11: The 400 MHz COSY Data for Synthetic (±)-13-Panasinsene (31).  2  31 Position (H-x) 2a 2b 4b 5a 5b 6a 6b 8a 8b 9a or 9b 9b or 9a lOa or lOb lOb or lOa 12  Signal o ppm (multiplicity;a J; number_of_H) 1.46 (d; 12.5; 1H) 1.96 (dd; 12.5, 3.0; 1H) 2.10 (br dd; 8.5, 3.0; 1H) 1.63-1.68 (m; 1H) 1.87-1.97 (m; 1H) 1.73 (dd; 12.0, 7.0; 1H) 1.35-1 .41 (m; 2H) 1.27 (dm; 12.5; 1H) 1.42-1.48 (m; 1H) 1.35-1.41 (m; 2H) 1.57-1.63 (m; 1H) 2.00 (br td; 12.5, 4.0; 1H) 2.18 (dm; 12.5; 1H) 1.07 (s; 3H)  COSY Correlations (H-x) 2b; 12 (W-coupling) 2a; 4b 2b; 5b 5b; 6a 5a; 4b; 6b; 6ab 6b; 5a; 5bb 6a; 5b 8b; 9a and 9b; lOa or lOb (W coupling) or 9a; 14 (W-coupling) 8a; 9b 9b or 9a; Ba; lOa and lOb 9a or 9b; Ba; Bb; lOa and lOb lOb or lOa; 9b and 9a; 15b lOa or lOb; 9a and 9b; 8a (W coupling) 2a (W-coupling); 13 (W coupling) 12 (W-coupling) 8b (W-coupling) 15b 15a; lOa or lOb  0.86 (s; 3H) 13 0.75 (s; 3H) 14 4.80 (d; 1.5; 1H) 15a 4.91 (dd; 1.5, 1.5; 1H) 15b a Signals labelled s, d, or dd may also incorporate unresolved fine couplings. b Small correlations observed.  148  In nOe difference experiments (summarized in structure 31’), irradiation of the singlet at 3 1.07 (Me-12) led to enhancement of the signals at 1.96 (H-2b), 2.10 (H-4b) and 4.80 (H-15a), while irradiation of the singlet at 30.86 (Me-13) led to enhancements of the signals at 1.46 (d, H-2a) and 1.63-1.68 (m, H-5a).  Irradiation of  the singlet at 30.75 (Me-14) led to enhancement of the doublet at 1.46 (H-2a).  Thus, the assignments for the methyl groups were  consistent with the other nmr data.  The results outlined above for the synthetic (±)-J3-panasinsene (31) provide further information about the spectroscopic properties of the natural product and confirm the identity of the synthetic material.  149  IlL CONCLUSION  The 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. synthesis  of  While the approach to the  (31) via  (±)qi-panasinsene  the  Pauson-Khand  cyclization reaction was unsuccessful, the Weiss-Cook condensation provided a viable alternative for the synthesis of an enone (159) with a tetrasubstituted double bond.  The key methylenecyclohexane  annulation sequence previously developed in our laboratories was successfully applied to the enone 159 and efficiently provided the tricyclic keto ester ketal  171 which  has  the  stereochemistry at the three chiral centers. manipulations  followed  intermediates  200 and 201  skeleton.  by  a  Wolff  desired  relative  Functional group  rearrangement  provided  with the required tricyclic carbon  Further reactions, including the double deoxygenation of  the sterically hindered aldehyde functions in the dialdehyde 217, generated  synthetic  (±)--panasinsene  31  .  The previously  unreported 13 C, HETCOR, COSY and nOe data were obtained for 31 as a part of this research and provide a more complete picture of the spectroscopic characteristics of the natural product.  150  159  169 and 170  171  179  177 and 178  k e t0 M 2  182  200 and 201  202 HO ‘a  Me  OHCa  n  m Me  204  31  217  Scheme  D-12  a. KH, THF, 600C; dimethyl carbonate, —60°C (—93%), b. KH, THF, rt; PhSeCI, 0°C CI 0°C; rt (—quant.), d. 5-chloro-1-pentenyl-2-magnesium CH , O, 2 2 (86%), C. H , 4 CN, —60°C (64%), f. NaBH 3 CO CH 2 Cs , , THE, -78°C (—94%), e. 3 2 bromide, CuBrSMe SnH, 3 CN, —70°C (76%), h. n-Bu 3 , MeOH, -48°C (98%), g. PTC-CI, DMAP, CH 6H 3 CeCl O 2 AIBN, PhH, —77°C, (74%), i. 1 N HCI (aq), acetone, rt (85%), j. i) t-amylONa, PhH, CI 0°C (in the dark); iii) hv, CH , N, 2 3 , Et 3 CH, 5°C to rt (—quant.); ii) MsN 2 —7°C; ii; MeO Mel, -78°C to —5°C (65%), I. HMPA, -78°C; THE, -78°C; MeOH, 0°C (40.5%), k. LDA, N, -78°C to 0°C 3 I C CH , -78°C; Et , THF, rt (88%), m. DMSO, oxalyl chloride, 2 4 LiAI H NNH DEG,—135°C; KOH, —200°C (46%). H , (98%), n. 2  151  IV. EXPERIMENTAL  4.1. General.  H nmr) spectra were 1 Proton nuclear magnetic resonance ( recorded on either a Varian XL-300 or a Bruker WH-400 nmr spectrometer using deuterochioroform as the solvent and tetra methylsilane (TMS) or the proton of the residual chloroform (6 7.26) as the internal standard, unless otherwise noted. are given in parts per million (8) from TMS. are given in Hertz (Hz).  Signal positions  Coupling constants (J)  The multiplicity, number of protons,  coupling constant(s), and assignments (when known) are given in parentheses.  Abbreviations used are: s, singlet; d, doublet; t, triplet;  q, quartet; m, multiplet; br, broad. 13 nmr) spectra were ( Carbon nuclear magnetic resonance C recorded on a Varian XL-300 nmr spectrometer at 75.3 MHz, or on a Bruker AM-400 spectrometer at 100 MHz, or on a Bruker AMX-500 spectrometer at 125.8 MHz using deuterochloroform as the solvent, unless otherwise noted.  Signal positions are given in parts per  million (6) relative to the chloroform signal at 6 77.0.61  Signal  multiplicities were determined by the Attached Proton Test (APT) experiment. Two dimensional spectra were recorded on the Bruker WH-400 nmr 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 as  152  indicated above or, in some cases, residual chloroform at 3 7.24 was used for the  reference and a correction factor (x  +  0.02 ppm) was  applied to the data. Infrared (ir) spectra were recorded on a Perkin-Elmer 1710 Fourier Transform  Spectrophotometer with  internal  calibration.  Abbreviations used are: s, strong; v, very; w, weak. Low resolution mass spectra (LRMS) were recorded on a Kratos MS8ORFA spectrometer.  High resolution mass spectra (HRMS) were  recorded on a Kratos/AEI MS 50 spectrometer. Elemental analyses were performed on a CARLO ERBA CHN elemental analyzer, Model 1106, or a Schoniger’s Oxygen Flask (analysis of sulfur). Melting points (uncorrected) were measured on a Fisher-Johns melting  point apparatus,  unless otherwise  noted.  Distillation  temperatures (uncorrected) are indicated as air-bath temperatures of Kugelrohr distillations, unless otherwise noted. Gas-liquid chromatography (glc) was performed on either a Hewlett-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 (former  instrument) or -‘25 m x 0.20 mm coated with 5% phenyl-methyl silicone (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).  153  139 chromatography were done on 230Conventional column and flash 400 mesh silica gel (E. Merck, Silica Gel 60). Unless otherwise stated, all reactions were performed under an atmosphere 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 following baths: 5-10°C, water/(ice); 0°C, ice/water; -20°C and -48°C, aqueous calcium chloride/CO 2 (27.0 g calcium chloride/ 100 mL water; 46 g . 2 calcium chloride! 100 mL water, respectively); -78°C, acetone/CO Temperatures were measured in degrees Celsius.  Solvents and Reayents  Solvents and reagents were dried and purified using standard ° 14 procedures. Tetrahydrofuran (THF) and diethyl ether were distilled from sodium benzophenone ketyl.  Benzene, dichloromethane and dimethyl  sulfoxide were distilled from calcium hydride.  Petroleum ether  refers to a hydrocarbon mixture with b.p. 30-45°C (from the distillation of a commercially available mixture with b.p. 30-60°C). Diisopropylamine, pyridine  and  triethylamine,  acetonitrile were  hexamethylphosphoramide,  distilled  from  calcium  hydride.  Anhydrous hydrazine (explosive in the presence of oxidizing agents) was prepared by refluxing hydrazine hydrate over an equal weight of sodium hydroxide pellets for 2 h and distilling under a flow of  154  141 argon.  Methanesulfonyl azide (MsN , potentially explosive) was 3  prepared from distilled  methanesulfonyl chloride  (distilled from  a 7 ) and sodium azide, according to the Danheiser et aI.hl 5 PCI  b and was used 7 modification of the procedure of Boyer et aI.,hl without  distillation.  A 0.78 M benzene solution of sodium t-amyloxide was prepared according to the procedure of Conia’ 42 and was standardized using aqueous hydrochloric acid in ethanol’ (phenolphthalein indicator). Solutions of methyllithium in diethyl ether and n-butyllithium  in  hexanes were obtained from Aldrich Chemical Co., Inc. and were standardized using the procedure of Kofron and Baclawski. 143 An “O.32 M tetrahydrofuran solution of lithium diisopropylamide was prepared by the reaction of diisopropylamine (1.1 equiv) and n butyllithium (1 equiv) in THE at -78°C (-45 mm).  The solution was  not 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. distilled  from  . 5 PCI  Acetyl chloride was  Dimethyl carbonate was dried over  molecular sieves and was distilled before use. distilled from phosphorus pentoxide.  4A  Methyl formate was  lodomethane was passed  through a short column of flame dried basic alumina before use. Copper (I) bromide-dimethyl sulfide complex was prepared by the method described by Wuts. 144  Magnesium bromide-etherate was  prepared  magnesium  by  the  reaction  of  metal  with  1,2-  dibromoethane in diethyl ether, followed by the removal of the  155  diethyl  ether  under  reduced  pressure  (-‘0.2  Torr)  at  room  temperature. 5-C hloro-2-tri methylstannyl- 1 -pentene  (5)  was  prepared  according to a previously reported modificationOb of the original 5 procedure.  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 was added during the reaction.  The purified product exhibited the  expected H nmr spectrum. The keto ketal 46 was prepared according to the procedure described by Moss 85 except that the purification of the product ’ 38 was modified.  Thus, a mixture of the keto ketal 46, diketal 153 and  diketone 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 ,3propanediol (152), was subjected to flash chromatography on silica gel (400 g, elution with 2:1 diethyl ether-petroleum ether to elute the diketal 153; elution with 9:1  diethyl ether-ethyl acetate to  elute the keto ketal 46; and elution with ethyl acetate to elute the diketone 43).  The appropriate fractions were combined to yield 7.5  g of the keto ketal 46 as a white solid, m.p. 46.5-47.5°C (lit. m.p. 85 which was spectroscopically ( 48°C) H nmr) identical with the 1 material reported by Moss.  The combined recovery of the diketal  153 and the diketone 43 was 8.1 g. Distilled solvents were deoxygenated by bubbling argon through the stirred solvent for at least 1 h. All other reagents were commercially available and were used without further purification.  156  4.2. Experimental Procedures for the Synthesis of (±)-J3Panasinsene (31) via the Weiss-Cook Condensation Approach.  Preparation of the Bicyclic Keto Ester Ketals 163 and 164. Me  H Me  HO H  164  163  A stirred suspension of potassium hydride (0.945 g, 23.6 mmol, 2.6 equiv, freed from mineral oil by washing with three 5 mL portions of dry THE) in 43 mL of dry THE, under an argon atmosphere, was warmed briefly to  -50°C.  The mixture was  allowed to cool to room temperature and a solution of the keto ketal 46 (2.002 g, 8.93 mmol) in 2 mL of dry THE was added, with three rinses of dry THE (8 mL total). 2 h.  The mixture was heated at -‘60°C for  To the resultant orange-tan suspension was added quickly dry  dimethyl carbonate (2.1 mL, 2.2 g, 25 mmol, 2.8 equiv).  After the  dark-colored mixture had been heated at -‘60°C for a further 1.5 h, it  was cooled to 0°C (ice bath).  mixture  The vigorously stirred reaction  was treated with a mixture consisting of 100 mL of  saturated aqueous ammonium chloride (pH 5), 100 mL of ice, and 100 mL of chloroform.  The aqueous layer was acidified (1  hydrochloric acid) to pH 6-7 and the phases were separated.  N The  157  aqueous  phase  chloroform.  was  extracted  with  two  50  mL  portions  of  The combined chloroform extracts were washed with  brine, dried (anhydrous sodium sulfate), filtered, and concentrated to give the keto ester 163 and the ester enol tautomer 164 (2.352 g, 93%; ratio 1.5:1, ‘H nmr analysis) as an orange oil* which solidified slowly to an off-white waxy solid.  This material was  used without further purification in the next reaction. The crude material, which consisted of a mixture of 1 63 and 164 exhibited ir (neat): 1756 (m), 1729 (s), 1661 (s), 1621 (m), 1281 (s), 1202 (s), 1113 (vs) cm; ‘H nmr (300 MHz): 30.946, 0.952 1 (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 -Cjz1 2 groups),  3.73,  3.75  (s, s,  3H  total,  Me, 163 and 164, 2 CO  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 5 0 2 H C, : 2 282.1467; found: 282.1459.  In some runs, the oil was deep red. In these cases, the crude product was dissolved in diethyl ether and the resultant solution was filtered rapidly through a short column of silica gel (—3X by weight, elution with diethyl ether).  158  Preparation of the Keto Ester Selenides 165 and 16& C 2 MeO  H  Me  H  H  165  166  To a stirred suspension of potassium hydride (0.455 g, 1.3 mmol, 1.3 equiv, freed from mineral oil by washing with three 3 mL portions of dry THE) in 35 mL of dry THE, under an argon atmosphere, was added, over a period of 10 mm, a solution of the mixture of the keto esters 163 and 164 (ratio -‘1.5:1, 2.403 g, 8.51 5 mL of dry THE, with two rinses of dry THE (10 mL total).  mmol) in  After the mixture had been stirred for 40 mm  at room temperature,  the brown suspension was cooled in an ice bath and a solution of benzeneselenenyl chloride (2.20 g, 11.5 mmol, 1.35 equiv) in 2 mL of dry THF, with several rinses of dry THF (11 mL total), was added quickly.  The orange reaction mixture was stirred at 0°C for 20 mm  and then was pipetted carefully (over a period of 10 mm) into a vigorously stirred mixture consisting of ice (15 mL), saturated aqueous sodium bicarbonate (25 mL), and a 1:1 pentane-diethyl ether mixture (50 mL).  The organic layer was separated and the  aqueous phase was extracted twice with 1:1 pentane-diethyl ether (75 mL total). brine (50  The combined organic extracts were washed with  mL), dried (anhydrous sodium sulfate), filtered and  concentrated to yield an orange oil (4.098 g, >100%).  Unreacted  159  benzeneselenenyl chloride was separated from the crude product by flash chromatography on  silica gel  petroleum ether-ethyl acetate).  (211  g,  elution with 2:1  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 used for the next reaction without further purification.  An -5:1 epimeric  mixture 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 -CU -), 3.45 (s, -‘2H, ketal 2  -  -, partially burying a m, 166), 3.51 (s, COjj, 166), 3.71 (s, 3H, 2 C 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 small amounts  could  sometimes  be  isolated  pure  using  the  chromatographic procedure described above (vide supra). The minor isomer 166, exhibited ir (neat): 1751 (vs), 1728 (vs), 1117 (vs), 742 (s), 693 (m) 1 cm H ; 1 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, both ketal -Cj- groups and a bridgehead Ca), 3.51 (s, 3H, CO L’j, 7.292 7.38 (m, 3H, aromatic CU), 7.63-7.65 (m, 2H, aromatic CU). Exact Mass calcd. for O Se: C 2 H 8 5 1 438.0946; found: 438.0940. 6 0  160  Preparation of the Enone Ester 159 10  9  Hb  HaF4Ha  14  159  To a cold (0°C) stirred solution of the mixture of keto ester selenides 165 and 166 (ratio -‘5:1, 2.437 g, 5.57 mmol) in 20 mL of distilled dichloromethane was added, over a period of 7 mm, an aqueous hydrogen peroxide solution (2.4 mL of a 15% mmol, 2.1 equiv).  After the mixture had been stirred for 10 mm  0°C and for -‘20 mm  at  at room temperature, 10 mL of water was  added and the phases were separated. was  solution, 11.7  washed with water (10 mL).  The dichloromethane layer  Then each aqueous phase was  extracted with dichloromethane (two 2 mL portions).  All the  dichloromethane layers were combined, dried (anhydrous sodium sulfate), filtered and concentrated to yield a yellow-orange oil containing a solid.  The crude product was diluted with diethyl ether  (6 mL) and the mixture was filtered through Celite (elution with diethyl ether) to remove any benzeneselenenic acid. concentrated to yield  The eluate was  the enone ester 159 (1.589 g, >100%, due to  small amounts of impurities, including some aromatic products) as an orange oil which was not further purified.  The crude product  exhibited ir (neat): 1750 (vs), 1719 (vs), 1653 (w), 1274 (m), 1149  161  (w), 1112 (s) cm ; 1 1 H 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 -Ca 2 groups), 3.85 (s, 3H, CO M.). 2 were also present.  Some minor signals due to impurities  In 1 H 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 (H  =  4a) simplified the dd at 2.79 (H-4b) to a br d (J  =  6.5 Hz); and  irradiation of the multiplet at 6 3.14-3.26 (H-5b) simplified the t at 1.48 (H-6b) to a d (J  =  12.5 Hz), the dd at 2.25 (H-4a) to a d (J  Hz), the ddd at 2.70 (H-6a) to a br d (J (H-4b) to a d (J  =  =  18.0  =  12.5 Hz), and the dd at 2.79  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. for 0 10 C 2 H : 5 5 280.1311; found: 280.1311.  Preparation of the Keto/Enol Ester Chlorides 169/170. CI,  C[ Me  Me Me HO’  H  169  170  162  To 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 of dry THE, under an argon atmosphere, was added a solution of methyl-lithium in diethyl ether (1.55 M, 7.20 mL, 11.2 mmol, 1.59 equiv).  After the solution had been stirred for 20 mm  at -78°C,  anhydrous magnesium bromide etherate (2.886 g, 11.2 mmol, 1.59 equiv) was added in one portion. for 20 mm  The white suspension was stirred  at -78°C and then copper bromide-dimethyl sulfide  complex (0.366 g, 1.78 mmol, 0.25 equiv) was added in one portion. at -78°C and then  The pale yellow suspension was stirred for 20 mm  a solution of the ester enone 159 (1.98 g, 7.04 mmol) in 3 mL of dry THF, with three rinses of dry THE (9 mL total), was added over 5 mm.  The orange suspension was stirred for 25 mm  at -78°C, and  then was treated with saturated aqueous ammonium chloride solution (pH ‘-6, 90 mL) and diethyl ether (90 mL).  The cooling bath  was removed and after the mixture had been stirred for 10 mm room temperature, the phases were separated.  at  The aqueous phase  was extracted with three 60 mL portions of diethyl ether.  The  combined organic extracts were washed with brine (90 mL), dried (anhydrous magnesium sulfate), filtered and concentrated to yield a green-brown oil which was quickly filtered through a short silica gel column (4.8 g, elution with diethyl ether).  Concentration of the  eluate yielded the keto/enol ester chlorides 169/170 (2.568 g, 94%) as a brown oil which was not further purified, but was used directly in the next reaction. The crude keto/enol ester chlorides 169/170 (which existed mainly as the enol tautomer, 170) exhibited ir (neat): 1754 (w),  163  1722 (w), 1657 (vs), 1619 (s), 1258 (s), 1218 (s), 1116 (vs), 806  (w) cm; 1 1 H 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), 2.072.43 (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, -Ca CI and both ketal -Cjj. 2 - groups), 3.74 2  (s, 3H, CO f), 4.74 (s, 1H, C=Cjj 2 ), 4.82 (s, 1H, C=Ck1 2 ), 10.79 (br s, 2 1H, enol  Ca).  present.  MS m/z (% rel.  Some minor signals due to impurities were also  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  Mass  calcd. for C1: C 2 H 3 5 0 0 384.1704; found: 384.1704. 9 5  Preparation of the Tricyclic Keto Ester Ketal 171.  Ha  19 Me  Hb’  Me 20  0  Hb  Hb  171  To 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, cesium carbonate (5.685 g, 17.4 mmol, 5.0 equiv).  The suspension was  164  heated at 58-61°C for 20 h.  The color changed from dark red-orange  to light brown by the end of the reaction.  After the mixture had  been cooled to room temperature, cold water (45 mL) and diethyl ether (40 mL) were added.  The deep red aqueous layer was  separated, was extracted with diethyl ether (40 mL), was acidified to pH -‘8 with hydrochloric acid (1  N, -“20 mL) and then was  extracted with three 40 mL portions of diethyl ether.  The combined  ethereal extracts were washed with brine (two 50 mL portions), dried (anhydrous magnesium sulfate), filtered and concentrated to yield the crude tricyclic keto ester ketal 171 (1.133 g, 93%) as a red-brown oily solid.  The crude product (dissolved in 4:1 petroleum  ether-ethyl acetate (-‘8 mL) and dichloromethane (1.5 mL)) was purified by flash chromatography on silica gel (135 g, elution with 4:1  petroleum  ether-ethyl  acetate).  Concentration  of  the  appropriate fractions yielded colorless crystals (0.846 g, 69%) of 97% purity (GLC analysis) which were then recrystallized (three crops) from hot ethyl acetate and cold petroleum ether (initial ratio -“1:3, additional petroleum ether added twice at 15 mm  intervals in  portions -“double the ethyl acetate volume) to yield colorless crystals (0.787 g, 64.3%).  The crystalline material thus obtained  exhibited 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), 1.541.71 (m, 2H, H-Ba or H-8b and H-9b or H-9a), 1.72-1.80 (m, 1 H, H-9a or 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),  165  2.54 (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-16  and 2H-18), 3.74 (s, 3H, Me-14’), 5.01 (d, 1H, J  =  1.0 Hz, H-15b), 5.07  (s, 1H, H-15a); 13 C nmr (75 MHz): 322.23 (H -19 or H 3 -20), 22.32 3 -20 or H 3 (H -19), 23.8 (.H 3 -9), 30.0 (-17), 30.6 (H 2 -8), 32.4 2 -10), 38.65 (.H-4), 38.74 (.H 2 (.H -3), 39.9 (H 2 -5), 42.8 (H 2 -2), 2 52.1 (H -14’), 58.2 (-1 or -7), 68.0 (.-7 or .Q-1), 71.8 (H 3 -16 or 2 -18), 72.4 (.Q..H 2 ..H -18 or H 2 -16), 109.0 (-13), 112.7 (H 2 -15), 2 145.1 (-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 0 C 2 H : 5 0 348.1936; found: 8  348.1929. Anal. calcd. for 0 C 2 H : 5 0 C 68.94, H 8.10; found: C 69.01, 8 H 8.15. For the HETCOR and COSY data, see Tables 1 and 2, PP. 78 and 81, respectively.  PreDaration of the Alcohol Ester Ketals 177 and 178. Me  ‘Me  Me  177  178  166  To 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 argon atmosphere, 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 sodium  borohydride (21.? mg, 0.560 mmol, 1.3 molar equiv) was added in one portion to the now white suspension. vigorously and then cleared somewhat.  The mixture bubbled  After the mixture had been  stirred for 1 h at -‘-48°C, the cooling bath was removed and, 2 mm later, -‘1 N hydrochloric acid (370 .tL, 0.37 mmol, -‘1.2 equiv) was added, followed 1 mm pentane.  later by 10 mL of cold water and 10 mL of  The mixture was stirred at 0°C for 5 mm, then a further  10 mL of pentane was added and the phases were separated.  The  aqueous phase was extracted three times with pentane.  The  combined pentane extracts were concentrated to yield a mixture of a white solid and an oil, which was dissolved in pentane (‘-30 mL). The solution was dried (anhydrous magnesium sulfate), filtered and concentrated to yield the epimeric alcohols 177 and 178 (150.9 mg, 98%; ratio 5.5:1, ‘H nmr analysis) as a white solid.  The alcohols  could be recrystallized from an -‘1:2 mixture of ethyl acetate hexane 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 epimers in the unrecrystallized mixture varied from -‘4.5:1 to -‘12:1 ( H nmr 1 analysis) depending on the scale of the reaction and slight changes in reaction conditions.  167  A 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), 2.352.42 (m, 2H, includes 2.35 (d, J  =  2.0 Hz, Oki. of 177, exchanged with  0)), 2.45-2.51 (m, 1H), 2.65 (d, 1H, J 2 D  =  16.0 Hz), 3.42-3.54 (m, 4H,  both ketal -C±1 - groups), 3.68 (s, -3H, C0 2 M.j, 4.69 (td, 1 H, J 2 2.0 Hz, H-6a, simplified to a t (J  =  =  9.0,  0 exchange), 4.85 2 9.0 Hz) upon D  (d, 1H, J= 1.0 Hz, C=C±1 ), 4.88 (br s, -‘lH, C=Ckj 2 , both epimers). 2 Signals due to the minor epimer, 178, appeared at: 0.92 (s, tertiary Me),0.98 (s, tertiary Me), 4.11-4.14 (m, H-6b, simplified upon D 0 2 exchange), 5.00 (br s, C=C±1 ); MS m/z (% rel. 2  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). 20 350.2093; found: 350.2087. C 3 H : 5 0  Exact Mass calcd. for  Anal. calcd. for 0 20 C C 3 H : 5  68.85, H 8.63; found: C 68.64, H 8.64.  Preparation of the Phenyl Thionocarbonate 179.  179  168  To a stirred solution of the mixture of the alcohols 178 and 179 (ratio -5:1, 295.8 mg, 0.844 mmol) in 7.7 mL of dry, freshly distilled 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)  and  phenoxythiocarbonyl chloride (180 iiL, 225 mg, 1.27 mmol, 1.5 equiv) was added.  Within 1 mm  a precipitate formed and, after a  period of 8 mm, the cooling bath was removed and the mixture was heated at 67-72°C for 20 h.  The reaction mixture was cooled and  the solvent was removed under reduced pressure to yield a tan solid. This material was suspended in water (25 mL) and the resultant mixture was extracted with ethyl acetate (50 mL, then three 25 mL portions).  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) and brine (two 30 mL portions). sodium  sulfate)  thionocarbonate  and concentrated to  yield  179 as an oil (463.4 mg,  presence of impurities). minimum  The organic phase was dried (anhydrous  volume  of  the crude >  100% due to the  The crude product was dissolved in a dichloromethane  and  purified  chromatography on silica gel (84 g, elution with 9:1 ether-ethyl acetate).  phenyl  by  flash  petroleum  The appropriate fractions were combined and  concentrated to yield the pure phenyl thionocarbonate 179 (310.7 mg, 76%) as one epimer.  The purified product thus obtained could be  recrystallized from a minimum volume of hot ethyl acetate and cold petroleum ether to give colorless plates which exhibited m.p. 151152.5°C; ir (KBr): 3091 (w), 3066 (w), 1736 (s), 1639 (w) 1594 (w),  169  1395 (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 -CU -), 3.53 (br s, 2 2H, ketal -CU -), 3.70 (s, 3H, CO 2 j4&j, 4.93 (br s, 2H, 2 2 C=CU ) , 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). calcd. for 6 0 2 C 3 H S 7 4 : 486.2076; found: 486.2082.  Exact Mass Anal. calcd. for  0 2 C 3 H S 6 7 4 : 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.  Me 20  Hb  181  To a stirred solution of the phenyl thionocarbonate 179 (117.0 mg, 0.240 mmol) in 2.4 mL of dry, degassed benzene, under an argon atmosphere, was added tri-n-butyltin hydride (160 p.L, 173 mg, 0.595 mmol, 2.5 equiv) and recrystallized 2,2’-azobisisobutyro-  170  nitrile (AIBN, 5.9 mg, 43 p.mol, 0.18 equiv).  After argon had been  passed over the surface of the reaction mixture for 10 mm, the solution was heated at 75-79°C for 20 h. The reaction mixture was cooled and the solvent was removed under reduced pressure to yield an oil which was purified by flash chromatography on silica gel (44 g,  elution  with  10:1  petroleum  ether-ethyl  acetate).  The  appropriate fractions were combined and concentrated to yield the slightly impure ester ketal 181 (74.4 mg, 92%). ketal 1 81  (58.0  concentrating  the  mg,  72%) was obtained  appropriate  fractions  The pure ester  by combining after  further  and flash  chromatography on silica gel (17.9 g, elution with 10:1 petroleum ether-ethyl acetate; 5.5 g, elution with 15:1 petroleum ether-ethyl acetate).  The product could be recrystallized from hot hexane to  give 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 H nmr (300 MHz): ö 0.95 (s, 3H, tertiary Me), 0.97 (m) cm ; 1 1 tertiary Me), 1.53-1.73 (m, 5H), 1.78 (d, 1H, J  =  (5,  3H,  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= 15.5 Hz), 3.41-3.52 (m, 4H, both ketal -Cj[ - groups), 3.63 (s, 3H, CO 2 M..), 2 4.87 (s, 1H, C=Cj), 4.92 (s, 1H, 2 C nmr (75 MHz): 3 22.3 C=CE ) ; 13 (.H 3 -20 or 3 3 19 or 3 .H 20), 22.4 (.H ), 26.6 2 2 .H 19), 23.4 (.H (.H ) , 29.9 (-17), 32.8 2 (.H ) , 33.8 (H ), 34.7 2 2 (H ) , 39.7 (j, 43.8 (j, 45.0 (H-4), 51.2 3 (.H 14’), 59.2 (.-1 or -7), 60.0 (-7 or .-1), 71.7 2 (H 16 or 2 -18 or H 2 .H -16), 109.8 2 2 18), 72.4 (H (.H 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. for 4 0 2 C 3 H : 0 334.2144;  171  found: 334.2141. Anal. calcd. for 4 0 2 C 3 H : 0 C 71.82, H 9.04; found: C 72.11, H 9.19.  Preparation of the Keto Ester 182.  182 To a stirred solution of the ketal ester 181  (29.5 mg, 88.4  i.Lmol) in 2.5 mL of reagent grade acetone was added 1 hydrochloric acid (44 iL, 44 .tmol, 0.5 equiv).  N  The solution was  stirred at room temperature for 5.5 h and then was added to 5 mL of water.  The aqueous suspension was extracted with four 5 mL  portions of diethyl ether. The combined ether phases were washed with water (5 mL), brine (5 mL), dried (anhydrous magnesium sulfate), and concentrated to yield the crude keto ester 1 82 as a colorless oil (22.4 mg, >100% due to the presence of impurities). The crude product was purified by flash chromatography on silica gel (5.5 g, elution with 3:1 petroleum ether-ethyl acetate) and the appropriate fractions were combined to yield a colorless oil which was distilled (115-1 20°C/0.3 Torr) to provide the pure keto ester  172  182  The oil could be  (18.6 mg, 84.7%) as a colorless oil.  recrystallized  from  pentane to  yield colorless  needles which  1 1633 (m), exhibited m.p. 43-43.5°C; ir (KBr): 3080 (m), 1729 (vs) 1 903 (s), 890 (m) cm’; ‘H nmr (300 MHz): S 1236 (vs) 1 1157 (vs) 1.38-1.50 (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 nmr  decoupling 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), and simplified 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) and  sharpened the dd at 2.08 (H-3a), the m at 2.25-2.42 (H-3b) and the m at 2.84-2.92 (H-4b); irradiation of the dd at 3 2.73 simplified the dd at 2.08 (H-3a) to a d (J  =  (H-2b)  19 Hz) and the dd at 2.19  (H-2a) to a br s, and sharpened the two multiplets at 2.25-2.42 (H 3b) and 2.84-2.92 (H-4b); and irradiation at 52.88 (the center of the m 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 simplified the m at 1.38-1.50 (H-5a).  , 28.5 (H ) C nmr (75 MHz): 823.6 2 3 ‘  , (H ) , 42.1 (H-4), 43.0 2 (H ) , 35.1 2 (.H ) , 33.5 2 (H ) , 32.4 2 (.H ) 2 -14’), 57.3 (-1 or -7), 58.0 (.-7 or -1), 3 , 51.6 (H (H ) 45.9 2 -15), 148.4 (-11), 175.7 (-14), 219.2 (-13); MS m/z 2 109.8 (Q.H (% ret.  mt.):  248 (M, 34), 220 (11), 216 (21), 191 (34), 190 (17),  173  189 (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 3 0 1 C 2 H : 5 248.1412; found: 248.1416. 0  Anal.  calcd. for 3 0 1 C 2 H : 5 C 72.55, H 8.12; found: C 72.62, H 8.15. 0 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 98  MeO C 2 i :Q  198  197a R=CHO, R’=H 197b R=H, R’=CHO  To a cold (5-10°C) stirred solution of the keto ester 182 (160.5 mg, 0.646 mmol) in 2.1  mL of dry benzene, under an argon  atmosphere, was added, dropwise, a benzene solution of sodium t amyloxide (0.78 M, 3.3 mL, 2.58 mmol, 4 equiv). mm  After a period of 6  the cooling bath was removed, and the solution was stirred at  room temperature for 1.5 h.  The mixture was recooled to 5-10°C  and freshly distilled methyl formate (330 iiL, 311 mg, 5.17 mmol, 8 equiv) was added quickly and the reaction mixture was stirred for a further 17 h at 5C to rt.  The solvent was removed under reduced  1 74  pressure to yield a reddish-orange oil which was dissolved in aqueous sodium hydroxide (10 mL of a 0.25 N solution). The solution was extracted with 2 portions of dichioromethane.  The basic  aqueous phase was acidified with 1 N hydrochloric acid (-4 mL) and the resultant mixture was extracted with dichloromethane (10 mL, The combined dichioromethane solutions were  four 5 mL portions).  washed with brine (8 mL), dried (anhydrous magnesium sulfate), and concentrated to yield an isomeric mixture of the formylated keto esters 197 and 198 (185.7 mg, >100% due to the presence of minor impurities; ratio 1 :-‘8.5,  1  H nmr analysis) as a yellow oil.  The  mixture was not purified further, but was used directly in the next reaction.  The crude formylated keto esters 197 and 198 exhibited  ir (neat): 1725 (vs), 1699 (s), 1609 (s, broad), 1227 (s), 1202 (s), ; 1 H nmr (400 MHz): 3 1 1157 (vs), 1084 (m), 893 (w), 800 (w) cm1.48-1.77 (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, CO M), 4.64 (distorted d, 1H, J 2 =  1.5 Hz, C=Cj1 ), 4.84 (d, 1H, J 2  C=CaOH).  =  1.5 Hz, C=CU ), 7.08 (s, 1H, 2  Signals due to the major aldehyde isomer (197a)  appeared at 31.98-2.04 (m), 3.68 (s, CO M.), 4.49 and 4.78 (d, d, J 2 , 9.52 and 9.83 (s, C=Cjj ) ‘1.5 Hz, 2 respectively); MS m/z (% rel.  mt.):  5,  =  CjjO, 197a and 197b,  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). 276.1 363.  Exact  6 276.1361; found: 0 0 1 C 2 H : Mass calcd. for 4  175  b) Preparation of the cx-diazoketone 199.  C1 2 MeO  199  To a stirred solution of the crude formylated keto ester mixture 197 and 198 (ratio 1:-8.5, 71.3 mg, 0.258 mmol) in 1.9 mL of dry dichloromethane at 0°C, under an argon atmosphere, was added methanesulfonyl azide* (M5N , 29 iL, 40.6 mg, 0.335 mmol). After 3 3 mm, freshly distilled dry triethylamine (54 .tL, 39 mg, 0.387 mmol) was added.  The reaction mixture was fully protected from light and  stirred at 0°C for 4 h.  The following experimental manipulations  were performed in a dimly lit room. The reaction mixture was treated with 40 drops of water and 40 drops of 6% aqueous potassium hydroxide and then was stirred for 5 mm. Distilled dichloromethane (4 mL) was added and the phases were separated. The  aqueous  dichioromethane.  phase was  extracted  with  4  more  portions of  The combined organic phases were washed with a  small amount of brine, dried (anhydrous magnesium sulfate) and concentrated.  The resultant off-white solid and yellow oil were  dissolved in a small volume of 1:1 pentane-diethyl ether and the resulting suspension was filtered quickly through a short column of silica gel (0.4 g, elution with 1:1 pentane-diethyl ether). CAUTION: All sulfonyl azides are potentially explosive.  The eluate  176  was concentrated* to yield the a-diazo ketone 199 (44.5 mg, 62%) as a yellow oil which exhibited ir (neat): 2082 (vs), 1727 (s), 1674 (s), 1636 (w), 1352 (m), 1247 (m), 1157 (m), 898 (w) cm . 1  This  material was not stable and, therefore, was used immediately in the next reaction.  c) Preparation of the diesters 200 and 201.  Hb  200  201  The crude a-diazo ketone 199 was dissolved in 10.4 mL of deoxygenated, distilled methanol (-‘0.025 M solution, based on 71.3 mg, 0.258 mmol of formylated keto esters 197 and 198) in a quartz photolysis tube (dimensions: 8 cm X 1.5 cm).  After argon had been  passed over the solution for 10 mm, the vessel was closed with a glass stopper and placed as close as possible to a medium pressure Hanovia mercury lamp (450 watt, Corex filter in a water-cooled quartz 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 the diester mixture from the photolysis reaction was reduced due to side reactions.  177  lamp and reaction vessel were immersed in a cooling bath) for 30 mm.  In order to follow the progress of the reaction by TLC or ir  spectroscopy, the reaction vessel could be opened under a stream of argon and a small aliquot removed.  After the solvent had been  removed, a mixture of the ring contracted diesters 200 and 201 (36.5 mg, 51%) was obtained as a colorless oil (ratio -1.6:1, gic analysis).  The crude product was purified by flash chromatography  on silica gel (5.5 g, elution with 8:1 hexanes-ethyl acetate).  The  appropriate fractions were combined to yield the less polar major diester epimer 200 (5.6 mg) as a solid and a mixture of the diesters 200 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 recrystallized from pentane to yield colorless prisms which exhibited m.p. 39.540°C; ir (KBr): 3101 (w), 1728 (vs), 1640 (w), 1439 (s), 1244 (s), 1199 (s), 1156 (s), 883 (m) cm; H nmr (400 MHz): 81.41-1.52 (m, 1 2H, H-Ba or H-8b and H-9b or H-9a), 1.58-1.65 (m, 2H, H-5a and H Bb 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, H  lOb 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); 13 -8), 25.2 (H 2 -5), 2 C nmr (75 MHz): 623.8 (H 26.6 (.H -2), 32.7 (.H 2 -9 and H 2 -10), 36.2 (.H-3), 37.2 (.H 2 -6), 2 46.8 (QH-4), 51.37 (H -13’ or H 3 -14’), 51.40 (.H 3 -14’ or .H 3 -13’), 3 52.4 (.-1 or-7), 57.9 (Q-7or.-1), 107.7 (H -15), 147.9 (-11), 2 174.0 (.-13 or -14), 175.5 (-14 or -13).  In nOe difference nmr  178  experiments  MHz),  (400  irradiation  at  3 2.71  (H-4b) led to  enhancement at 3.18 (H-3b); irradiation at 3.18 (H-3b) led to enhancements 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 (H lOb or H-lOa), 2.39-2.44 (H-2b) and 3.18 (H-3b).  In 1 H  nmr  decoupling 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) and  simplified the mutiplets at 2.39-2.44 (H-2b) and at 1.95-2.07 (H-5b and another H); and irradiation of the dd at 2.26 (H-2a) simplified the m at 2.39-2.44 (H-2b) and simplified the dt at 3.18 (H-3b) to a distorted 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. for  0 1 C 2 H : 4 6 278.1518; found: 278.1527. 2 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) cm H nmr (400 MHz): 31.40; 1 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 overlapped with 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-  179  14’), 3.67 and 3.68 (s, s, 3H total, epimeric CO , 201 (Me-12’) and 2 200 (Me-13’), respecUvely), 4.92-4.96 (m, 2H, 2 C=C ) . 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). for 4 0 1 C 2 H : 6 278.1518; found: 278.1524. 2  Exact Mass ca!cd.  Anal. calcd. for 4 0 1 C 2 H : 6 2  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 and 116, respectively.  Preparation of the Diesters 202 and 203.  15  2  Hb  Hb  202  203  13  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 argon atmosphere, was added (via a cannula) a solution of the mixture of diesters  200 and 201  (ratio 2:1, 14.1  mg, 50.7 iimol)  dissolved in  180  200 iL of dry THE and rinsed in with three portions of dry THE (‘-300 iL total). After the solution had been stirred at -78°C for 1.5 h, 1 hexamethylphosphoramide* (HMPA, 14 jiL, 14.4 mg, 80.5 iimol, 1.6 equiv) was added and the solution was stirred for 12 mm.  Freshly  dried iodomethane (48 iLL, 109 mg, 771 jimol, l5equiv) was then added quickly and the reaction mixture was stirred for a further 30 mm  at -78°C.  The reaction mixture was warmed to ‘-5°C over a  period of 50 mm  and then was treated with saturated aqueous  ammonium chloride and dilute aqueous sodium thiosulfate (just enough to decolorize the solution). diethyl ether (4 portions).  The mixture was extracted with  The combined ethereal extracts were  washed with 2 portions of brine, dried (anhydrous magnesium sulfate) and concentrated.  The resultant yellow oil was dissolved in  a small amount of diethyl ether and the solution was filtered through a short column of silica gel (‘-0.4 g, elution with diethyl ether).  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,  H 1  nmr  analysis),  which  was  purified  by  chromatography on silica gel (0.9 g, elution with 15:1 hexanes-ethyl acetate; repeated 2-3 times until the material obtained consisted only of a mixture of the two diesters 202 and 203).  The  appropriate fractions were combined to yield 9.6 mg (65%) of a mixture 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 and  203 exhibited ir (neat): 1729 (vs), 1642 (w), 1456 (w), 1240 (rn),  *  CAUTION: HMPA is known to be a potent carcinogen.  181  1146 (s), 887 (w) cm; H nmr (400 MHz): 31.42 (s, 3H, Me-12), 1 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 or  H-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, 2 CO j ), 3.66 (s, 2 CO j ), 4.85 (s, 2 C=Cj[ ) , 4.87 (s, 2 C=C1j ) . In ‘H nmr decoupling experiments (400 MHz), irradiation of the signal at 3 1 .85 (dm, 1 H, J  =  11.0 Hz) led to simplification of the multiplets  at 1.46-1.59 (3H) and 2.28-2.39 (H-4b and 2H); irradiation of the m at 32.00 (H-lOa and H-lob) led to simplification of a d in the signal at 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 at 2.26 (H-2b) to a distorted t (J  In nOe difference experiments (400 MHz), irradiation of the singlet at 6 1 .42* ppm =  3.0 Hz).  (Me-12) led to enhancement of the signals at 1.85, 1.95-2.06 (H-lOa and H-lOb), 2.26 (H-2b), -‘2.35 (H-4b), 3.69 (Me-13’), and 4.98 (H 15a); and  irradiation of the signal at 64.98  (H-15a)  led  to  enhancement of the signals at 2.26 (H-2b) and 1.42 (Me-12); 13 C nmr (75 MHz): 624.7 2 (.H ) , 25.3 (H -12), 26.3 2 3 (H ) , 32.1 2 (H ) , 33.1 (H ) 2 , 33.3  36.1 2 (H ) , 42.0 (C), 49.9 (C), 51.4 (H -13’, or 3  -14’), 51.5 (Q.H 3 Q..H -14’, or H 3 -13’), 53.6 (H-4), 58.2 (C), 109.2 3 -15), 149.8 (-11), 175.6 (-13 or .-14), 176.2 (.-14 or -13); 2 (.H *  Unavoidable irradiation of part of the multiplet at 1.46-1 .59 also occurred as the signals were too close.  182  MS 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). 292.1674; found: 292.1675.  Exact Mass calcd. for 0 14 C 2 H : 4 7  Anal. calcd. for 0 14 C 2 H : 4 7 C 69.84, H  8.27; found: C 69.58, H 8.51. For the COSY data, see Table 6, P. 123.  Preparation of the Diols 204 and 205 HO 1  •QH  A 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 a cannula (with four 200 iL rinses of THE), over a period of 15 mm, to a 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 argon atmosphere.  The cooling bath was removed and the mixture was  stirred at room temperature for 1.2 hours.  Then sodium sulfate  decahydrate was added cautiously to react with the excess lithium aluminum hydride.  The mixture was stirred for a few minutes  before it was filtered through a short column of Elorisil (-‘0.28 g, elution with THE and small amounts of diethyl ether).  The solvent  183  was removed under reduced pressure to give the crude product (35 mg, >100% due to the presence of impurities) which was purified by flash chromatography on silica gel (6 g, elution with diethyl ether). The sample was loaded on the column by dissolving it in a mixture of acetone (‘-250 p,L) and hot ethyl acetate (‘-800 p.L).  The appropriate  fractions 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-pentane to yield colorless needles which exhibited m.p. 139-139.5°C (sealed tube; 204 sublimed at 137°C on a Fisher-Johns melting point apparatus); ir (KBr): 3312 (vs), 1635 (m), 1401 (s), 1039 (s), 1024 ; ‘H nmr (400 MHz, acetone-d 1 , external TMS): 81.11 6 (s), 885 (m) cm(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 the acetone peak), 2.16-2.22 (m, 2H), 3.20 (t, 1H, J = 5.5 Hz, 01±, 0* ), 3.23-3.30 (m, 2H, Ca 2 0 2 exchanged with D OH, simplified upon D 2 exchange), 3.32 (t, 1 H, J  =  0), 3.40 (dd, 2 5.0 Hz, 01± exchanged with D  0 2 1H, J= 10.5, 5.5 Hz, C 0H, simplified to a d (J= 10.5 Hz) upon D 2 exchange), 3.46 (dd, 1H, J= 10.5, 5.0 Hz, Ca OH, simplified to a d (J= 2 0 exchange), 4.81 (d, 1H, J= 1.5 Hz, C=C1j 2 10.5 Hz) upon D ), 4.89 2 C nmr (125 MHz, acetone 3 (distorted dd, 1H, J= 1.5, 1.5 Hz, C=Ca ); ‘ 2 ): 825.1, 25.58, 25.62, 30.0 (buried in the acetone Me), 31.0, 34.3, 6 d 0H-13 or .H 2 0H-14), 2 36.5 (C), 38.1, 51.9 (C), 52.4 (C), 52.6, 63.3 (.H 0H-13), 108.6 (.H 2 -15), 153.7 (.-11); MS m/z 2 0H-14 or .H 2 67.7 (.H  *  The addition of D 0 caused the chemical shifts of non-exchanged hydrogens to change. 2  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. for 0 14 C 2 H : 2 5 236.1776; found: 236.1780; Anal. calcd. for 0 14 C 2 H : 2 5 C 76.23, H 10.23; found: C 76.25, H 10.24. A very small amount of the minor (less polar) epimer 205 (an oil), which was isolated from repeated flash chromatography of the combined  impure  fractions  from  several  reactions  using  the  procedure described above, exhibited ir (neat): 3305 (s), 1635 (m), 1448 (m), 1053 (s), 1030 (s), 1000 (m), 887 (m) cm H nmr (400 ; 1 1 MHz, acetone-d ): 80.93 (s, 3H, Me-13), 1.27-1.31 (m, 1H), 1.41 (qt, 6 1H, J= 13.0, 4.0 Hz), 1.52-1.64 (m, 3H), 1.72 (dd, 1H, J= 13.5, 7.0 Hz), 1.80-2.24 (m, -7H, includes acetone peak), 3.20 (t, 1 H, J  =  5.5  Hz, 01± exchanged with D 0), 3.30-3.49 (m, 5H, simplifies upon D 2 0 2 exchange 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=CE ), 4.87 (distorted dd, 1H, J= 1.5, 1.5 Hz, C=Ca 2 ); MS m/z 2 (% 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. for 0 14 C 2 H : 2 5 236.1776; found: 236.1777.  185  Preparation of the Diacetate 213.  H  Hb  213  To a cold (0°C), stirred solution of the diol 204 (5.3 mg, 22 tmol) and 4-(N,N-dimethylamino)pyridine (DMAP, -‘3 mg, -‘25 j.tmol, -‘1.1  equiv) in 800 p.L of dry dichloromethane, under an argon  atmosphere, was added dry pyridine (16.5 .tL, 16.1 mg, 0.20 mmol, 9 equiv) followed 5 mm  later by freshly distilled acetyl chloride (9.6  tL, 10.6 mg, 0.14 mmol, 6 equiv). The mixture was stirred at 0°C for 3 h and then diethyl ether (8 mL) and 0.15 N hydrochloric acid (1.3 mL) were added. The aqueous phase was extracted with two portions  of diethyl ether.  The combined ethereal phases were washed with  water (1.5 mL), saturated sodium bicarbonate (1.5 mL), and brine (two -‘1 .5 mL portions) and then were dried (anhydrous magnesium sulfate) and concentrated to yield the crude diacetate 213 (7.4 mg, >100% due to small amounts of impurities) as an oil.  The crude  diacetate 213 was purified by chromatography on silica gel (0.9 g, elution with 1:1 pentane-diethyl ether).  The appropriate fractions  were combined and concentrated to yield the diacetate 213 (6.1 mg, 85%) as a colorless oil which exhibited ir (neat): 1742 (vs), 1636  186  (w), 1238 (vs), 1034 (m), 889 (w) cm; 1 H nmr (400 MHz): 31.15 (s, 3H, Me-12), 1.29-1.39 (m, 1H), 1.43 (br qt, 1H, J (dd, 1H, J  =  =  13.0, 3.5 Hz), 1.54  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-lOb and H-x), 1.98 (dd, 1H, J  =  13.0, 3.0 Hz, H-2b), 2.02 (s, 3H, Me-13” or  Me-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, 2H13), 4.03 (d, 1H, J  =  11.0 Hz, H-14b), 4.84 (s, 1H, H-iSa), 4.96 (s, 1H,  H-15b); ‘ C nmr (125 MHz): 3 20.88, 20.92, 24.56, 24.67, 25.2, 30.3, 3 30.7, 33.3, 33.7  -13 or 2 (Q.), 37.1, 49.1 (j, 51.3, 51.9 (j, 66.0 (H  -i4), 69.1(H 2 H -14 or H 2 -13), 109.4 (H 2 -15), 150.8 (-11), 2 171.2 (.-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). calcd. for 0 18 C 2 H : 4 9 320.1988; found: 320.1994.  Exact Mass  Anal. calcd. for  18 C 2 H : 4 0 9 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 and 134, respectively.  187  Preparation of the Dialdehyde 217.  OHCi  217  To a cold (-78°C), stirred solution of dry dimethyl sulfoxide (DMSO, 54 pL, 0.76 mmol) in 300 p.L of dry dichioromethane, under an argon atmosphere, was added freshly distilled oxalyl chloride (34 .tL, 49 mg, 0.39 mmol).  After the mixture had been stirred for 30  mm, a solution of the diol 204 (10.0 mg, 42 p.mol) in 40 p.L of dry DMSO and 500 iiL of dry dichioromethane was added via a cannula and rinsed in with dry dichloromethane (700 i.tL total). mixture was stirred for 40 mm  The reaction  at -78°C and then dry triethylamine  (240 .iL, 174 mg, 1.72 mmol) was added and the mixture was warmed to —0°C over a period of 70 mm.  Water (—3 mL) was added  and the product was extracted with dichloromethane (5 mL, four 3 mL portions).  The combined organic phases were dried (anhydrous  magnesium sulfate) and concentrated to give a mixture of an oil and a white solid.  The crude product was suspended in diethyl ether and  the mixture was filtered through a short silica gel column (—0.2 g, elution with diethyl ether) to remove the solid.  The eluate was  concentrated to yield the dialdehyde 217 (9.6 mg, 98%) as a colorless oil which was not further purified.  The dialdehyde 217  thus obtained exhibited ir (neat): 3091 (w), 2723 (w), 1718 (vs),  188  1638 (w), 1459 (w), 894 (w) cm; 1 H 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 2.51 (dd, 1H, J  =  8.5, 3.0 Hz), 2.67 (d, 1H, J  =  =  13.0, 2.0 Hz),  13.5 Hz), 5.01 (s, 1H,  C=CH. ) 2 , 5.05 (s, 1H, 2 C=C ) , 9.50 (s, 1H, CaO), 9.65 (s, 1H, CaO); MS  m/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. for  0 1 C 2 H : 2 5 232.1463; found: 232.1459. 0  Preparation of (±)-J3-Panasinsene (31).  31  The crude dialdehyde 217 (9.6 mg, 41 j.imol), was dissolved in a  mixture of 250 j.tL of dry diethylene glycol and 150 pL of anhydrous hydrazine (151 mg, 4.73 mmol). for -‘5 mm  Argon was passed over the mixture  and then the mixture was heated, under an argon atmos  phere, at 133-138°C for 1.5 h.  The reaction mixture was cooled to  room temperature and then most of the water and excess hydrazine  189  were removed via distillation (-65°C, ‘-15 Torr) over a period of 20 mm.  Crushed potassium hydroxide pellets (48 mg, 0.85 mmol) were  added and the mixture was heated under an argon atmosphere at ‘-190-210°C for 7.5 h.  After the reaction mixture had been cooled to  room temperature, ‘-250 tL of water was added and the product was extracted with pentane (-‘3 mL total).  The combined pentane  extracts were dried (anhydrous magnesium sulfate) and most of the solvent was removed via distillation at atmospheric pressure.  The  residue was then filtered through a short column of silica gel (‘-0.2 g, elution with pentane) to remove polar impurities.  Most of the  pentane was removed from the eluate by distillation (atmospheric pressure) through a short Vigreux column and the last traces of solvent were removed by a Kugelrohr distillation (heated up to 80°C) to yield crude (±)-f3-panasinsene (31) (‘-6.4 mg, ‘-76%). was  distilled  then  (80-90°C,  100 Torr) to yield pure (±)-fl  panasinsene (31) (3.9 mg, 46%) as a colorless oil. f3-panasinsene  The product  The distilled (±)-  (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 with  1.96 (dd, 1H, J  =  12.5, 3.0 Hz, H-2b), 2.00 (br td, 1H, J  H-lOa or H-lOb), 2.10 1H, J  =  (  br dd, 1H, J  =  =  12.5, 4.0 Hz,  8.5, 3.0 Hz, H-4b), 2.18 (dm,  12.5 Hz, H-lOb or H-ba), 4.80 (d, 1H, J  (dd, 1H, J  =  =  1.5 Hz, H-15a), 4.91  -14), 3 1.5, 1.5 Hz, H-15b); 13 C nmr (75 MHz): 8 18.2 (H  190  -13), 25.10 (H 3 24.72 (H -5), 24.85 (H 2 -9), 30.53 (.-7 or .-3), 2 30.64 (H -12), 33.8 (H 3 -1O), 35.7 (H-2), 36.0 (H 2 -8), 41.1 2 -6), 45.6 (.-3 or -7), 52.60 (.H-4), 52.77 (-1), 108.3 (H 2 (H 2 15), 152.3 (..-11);  In nOe difference experiments (400 MHz),  irradiation of the signal at 8 0.75 (s, Me-14) led to enhancement of the signal at 1.46 (d, H-2a); irradiation of the signal at 0.86 (s, Me13) led to enhancement of the signals at 1.46 (d, H-2a), and at 1.631.68 (m, H-5a); and irradiation of the signal at 1.07 (s, Me-12) led to enhancement of the signals at 1.96 (dd, H-2b), 2.10 (br dd, H-4b) and at 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 : 14 C 2 H 5 204.1878;  found: 204.1875. For the HETCOR and COSY data, see Tables 10 and 11, pp. 144 and 147, respectively.  191  V. REFERENCES  mt.  Ed. EngI. 29, 1209 (1990).  1.  G. WHITESIDES. Angew. Chem.  2.  0. SEEBACH. Angew. Chem.  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 British Columbia. (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).  mt.  Ed. EngI. 29, 1320 (1990).  192 14.  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. Tetrahedron Lett. 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 by A. 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.  193  29.  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. John Wiley & Sons, New York. 1985. pp. 1101-2. (b) Esters may also be converted to the corresponding aldehydes or alcohols which may deoxygenated directly or via derivatives; see Ibid. pp. 1096-8 and pp. 392-393, 395, respectively, and references therein.  33.  For some examples of copper-catalyzed conjugate additions to enones 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).  194 41.  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 of cobalt complexes, see: K. NAKAMOTO. Infrared and Raman Spectra of 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. DENICOLA and 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. YANUCK  and 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).  195 54.  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. THOMSON and 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. Spectrometric Identification 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 the Carbon-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. (1967).  PREMPREE.  J. Am. Chem. Soc. 89, 6177  (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. Tetrahedron  Lett. 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. Tetrahedron Lett. 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).  197 77.  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 references  therein. 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  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.  G. P. P0LLINI. Synthesis, 316 (1973).  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).  199  1 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 Chemistry Vol. 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 and Synthesis. Academic Press, Orlando, Florida. 1986. pp. 372383 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).  200  115. 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) (b) (c) (d)  D. P. CURRAN. Synthesis, 417, D. P. CURRAN. Synthesis, 489, M. RAMAIAH. Tetrahedron, 43, C. WALLING. Tetrahedron, 41,  (1988). (1988). 3541 (1987). 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 separation problems 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).  201  128. 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. 94, 5098 (1972).  HENGARTNER.  J. Am. Chem. Soc.  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 of Laboratory Chemicals. Second Edition. Pergamon Press, Oxford. 1980.  202 1 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).  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

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"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0061756/manifest

Comment

Related Items